Organic Synthesis [4th Edition] 9780128008072, 9780128007204

Table of contents :
Content:
Front Matter,Copyright,About the Author,Preface to the Fourth Edition,Preface to the Third Edition,Preface to the First Edition. Why I Wrote This Book!,Common AbbreviationsEntitled to full textChapter 1 - Retrosynthesis, Stereochemistry, and Conformations, Pages 1-60
Chapter 2 - Acids, Bases, and Addition Reactions, Pages 61-95
Chapter 3 - Functional Group Exchange Reactions: Aliphatic and Aromatic Substitution and Elimination Reactions, Pages 97-160
Chapter 4 - Acids, Bases, and Functional Group Exchange Reactions: Acyl Addition and Acyl Substitution, Pages 161-183
Chapter 5 - Functional Group Exchange Reactions: Protecting Groups, Pages 185-213
Chapter 6 - Functional Group Exchange Reactions: Oxidations, Pages 215-307
Chapter 7 - Functional Group Exchange Reactions: Reductions, Pages 309-418
Chapter 8 - Synthetic Strategies, Pages 419-482
Chapter 9 - Functional Group Exchange Reactions: Hydroboration, Pages 483-518
Chapter 10 - Functional Group Exchange Reactions: Selectivity, Pages 519-546
Chapter 11 - Carbon-Carbon Bond-Forming Reactions: Cyanide, Alkyne Anions, Grignard Reagents, and Organolithium Reagents, Pages 547-603
Chapter 12 - Carbon-Carbon Bond-Forming Reactions: Stabilized Carbanions, Organocuprates, and Ylids, Pages 605-657
Chapter 13 - Nucleophilic Species That Form Carbon-Carbon Bonds: Enolate Anions, Pages 659-742
Chapter 14 - Pericyclic Reactions: The Diels-Alder Reaction, Pages 743-800
Chapter 15 - Pericyclic Reactions: [m+ n]-Cycloadditions, Sigmatropic Rearrangements, Electrocyclic, and Ene Reactions, Pages 801-861
Chapter 16 - Carbon-Carbon Bond-Forming Reactions: Carbocation and Oxocarbenium Ion Intermediates, Pages 863-915
Chapter 17 - Formation of Carbon-Carbon Bonds Via Radicals and Carbenes, Pages 917-980
Chapter 18 - Metal-Mediated, Carbon-Carbon Bond-Forming Reactions, Pages 981-1023
Chapter 19 - Combinatorial and Process Chemistry, Pages 1025-1054
Subject Index, Pages 1055-1079
Disconnection Index, Pages 1081-1083

Citation preview

ORGANIC SYNTHESIS

ORGANIC SYNTHESIS FOURTH EDITION MICHAEL B. SMITH

AMSTERDAM • BOSTON • HEIDELBERG • LONDON NEW YORK • OXFORD • PARIS • SAN DIEGO SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO Academic Press is an imprint of Elsevier

Academic Press is an imprint of Elsevier 125 London Wall, London EC2Y 5AS, United Kingdom 525 B Street, Suite 1800, San Diego, CA 92101-4495, United States 50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom Copyright © 2017, 2010 Michael Smith. Published by Elsevier Inc. All rights reserved. First and second edition copyrighted by : Copyright © 2002, 1994 Mc Graw- Hill Companies Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein). Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library ISBN: 978-0-12-800720-4

For information on all Academic Press publications visit our website at https://www.elsevier.com/

Publisher: John Fedor Acquisition Editor: Katey Birtcher Editorial Project Manager: Anneka Hess Production Project Manager: Anitha Sivaraj Cover Designer: Greg Harris Typeset by SPi Global, India

About the Author

Professor Michael B. Smith was born in Detroit, Michigan in 1946 and moved to Madison Heights, Virginia in 1957, where he attended high school. He received an A.A. from Ferrum College in 1967 and a BS in chemistry from Virginia Polytechnic Institute in 1969. After working for 3 years at the Newport News Shipbuilding and Dry Dock Co. in Newport News VA as an analytical chemist, he entered graduate school at Purdue University. He received a PhD in Organic Chemistry in 1977, under the auspices of Professor Joe Wolinsky. He spent 1 year as a faculty research associate at the Arizona State University with Professor G. Robert Pettit, working on the isolation of cytotoxic principles from plants and sponges. He spent a second year of postdoctoral work with Professor Sidney M. Hecht at the Massachusetts Institute of Technology, working on the synthesis of bleomycin A2. Smith began his academic career at the University of Connecticut in 1979, where he is currently professor of chemistry. In 1986 he spent a sabbatical leave in the laboratories of Professor Leon Ghosez, at the Universite Catholique de Louvain in Louvain-la-Neuve, Belgium, as a visiting professor. Smith’s research interests focus on (1) the synthesis and structure-proof of bioactive lipids obtained from the human dental pathogen Porphyromonas gingivalis. (2) The development of indocyanine-imidazole dye conjugates that show an affinity for cancerous hypoxic tumors, allowing imaging of the tumors using near-infrared fluorescence spectroscopy. In addition to this research, he is the author of the fifth, sixth, and seventh editions of March’s Advanced Organic Chemistry. He is also the author of an undergraduate textbook in organic chemistry titled Organic Chemistry. An Acid-Base Approach, now in its second edition. He is the editor of the Compendium of Organic Synthetic Methods, Volumes 6–13. He is the author of Organic Chemistry: Two Semesters, in its second edition, which is an outline of undergraduate organic chemistry to be used as a study guide for the first organic course. He has authored a research monograph titled Synthesis of Non-alpha Amino Acids, in its second edition.

ix

Preface to the Fourth Edition

The fourth edition of Organic Synthesis has been revised, reorganized, and rewritten from front to back. I want to thank all who used the book in its first three editions. It is the same graduate level textbook that past users are familiar with, but specific examples have been updated and the book has been reorganized. Several long chapters have been split to make the concepts more focused. The chapter sequence has been reorganized, and a new chapter was created that will focus on organometallic chemistry. The molecular modeling problems that were prominent in the third edition have been removed. While I remain convinced that molecular modeling problems offer new insights into certain aspects of chemical reactivity, conformational analysis, and stereoselectivity, this feature in the third edition did not attract much interest and has been removed from the new edition. Many new examples of reactions have been added, specifically from the literature dating from 2008 to 2015. Apart from these changes, which will be described in detail below, the fourth edition remains what it was always intended to be. It is a graduate level textbook of organic chemical reactions, with a bias toward total synthesis, that targets first and second year graduate students. However, advanced undergraduate organic chemistry courses can certainly use the book. There are now 19 chapters in the new edition rather than the 14 in the third edition. The former Chapter 2 has been split into three chapters. In this new edition, Chapter 2 is now “Acids and Bases and Addition Reactions.” Chapter 3 is now “Functional Group Exchange Reactions. Aliphatic and Aromatic Substitution, and Elimination Reactions.” Chapter 4 is “Acids, Bases, and Functional Group Exchange Reactions. Acyl Addition and Acyl Substitution.” These changes were made to make this undergraduate organic chemistry review material more manageable. Chapter 7 in the third edition was the protecting groups chapter. Since most of that chemistry involves functional group exchange reactions, the protecting group discussion in the fourth edition is Chapter 5. Chapters 3 and 4 on oxidation and reduction from the third edition are retained as the new Chapters 6 and 7. The hydride reduction discussions in Chapter 7 have been reorganized to focus on borane and alane, and their structural modification that leads to new reagents. Catalytic hydrogenation and dissolving metal reductions follow. Note that the student syntheses for Chapter 14 in the 1st-3rd editions has been deleted in the fourth edition. The synthetic strategy chapter has been moved to Chapter 8 in the fourth edition, but the section that introduced combinatorial chemistry has been removed, and is now part of the new Chapter 19. The synthetic strategy chapter is now followed by Chapter 9, “Functional Group Exchange Reactions. Hydroboration,” and Chapter 10, “Functional Group Exchange Reactions. Selectivity.” These discussions were Chapters 5 and 6, respectively in the first to third editions. Chapter 8 in the third edition focused on Cd disconnections and carbanion reagents, whereas Chapter 9 in the third edition focused on enolate anion chemistry. The former Chapter 8 has been split into two chapters (Chapters 11 and 12) for the fourth edition: Chapter 11, “Carbon-Carbon Bond-Forming Reactions. Cyanide, Alkyne Anions, Grignard Reagents, and Organolithium Reagents” and Chapter 12, “Carbon-Carbon Bond-Forming Reactions. Stabilized Carbanions and Ylids.” Chapter 13 in the fourth edition is “Nucleophilic Species That Form Carbon-Carbon Bonds. Enolate Anions.” Chapter 11 in the third edition discussed all pericyclic reactions, which made for a very long chapter. In the fourth edition, these discussions have been split into two chapters (Chapters 14 and 15). Chapter 14 is “Pericyclic Reactions. The Diels-Alder Reaction,” and Chapter 15 is “Pericyclic Reactions: [m+n]-Cycloadditions, Sigmatropic Rearrangements, Electrocyclic and Ene Reactions.” The discussion of carbocation-driven reactions in Chapter 12 of the third edition is now in Chapter 16 of the fourth edition. There is a significant change, however, because the discussion of organopalladium, organocopper, and other organometallic chemistry have been consolidated and moved to a new chapter, Chapter 18. The discussions of radical and carbene chemistry in Chapter 13 of the third edition are now found in Chapter 17 of the fourth edition. As noted, Chapter 18 is a new chapter in the fourth edition, “Metal-Mediated, Carbon-Carbon Bond-Forming Reactions.” This new chapter consolidates modern organometallic reactions from several chapters found in the third edition, primarily 8, 12, and 13, for a more focused presentation of this important area of chemistry.

xi

xii

PREFACE TO THE FOURTH EDITION

Chapter 19, “Combinatorial and Process Chemistry” is new to the fourth edition. The discussion of combinatorial chemistry has been moved from the synthetic strategy chapter to this new chapter. In addition, a new section has been added on Process Chemistry, and a third new section added that discusses flow process chemistry. Homework in each chapter has not been extensively revised. This decision was taken to minimize the introduction of errors by retaining most of the homework found in the third edition. Further, there has been extensive reorganization of the chapters but most homework problems were deemed completely suitable for the fourth edition. The answers to most problems are available in pdf format as an on-line Student Solutions Manual that is available from the Elsevier website for the book. Many of the homework problems that deal with synthesis do not contain leading references for the answers, but as in previous editions, leading references are provided for synthesis problems where it is appropriate. Note that the synthetic strategy chapter in particular, Chapter 8, does not offer specific answers but rather leading references. Students are encouraged to discuss their answers to any synthetic problem with their instructor as there are usually several correct approaches, especially for complex targets. With the exception of a few scanned figures, all drawings in this book were prepared using ChemDraw Professional, version 15.0.0.106, provided by PerkinElmer. I thank Perkin-Elmer for this gift, and Ms. Julia Bracken in particular. All molecular model graphics are rendered with Spartan’10, version 1.0.1, provided by Wavefunction, Inc. I thank Warren Hehre and Sean Ohlinger for this gift, and for generously sharing their expertise over the years. I express my gratitude to all of those who were kind enough to go through the first, second, and third editions and who supplied me with comments, corrections, and suggestions. I give special thanks to Peter Wuts, PhD (Kalexsyn) and Professor Tim Jamison (MIT) who reviewed the process chemistry and flow chemistry sections of Chapter 19, respectively, and who provided the inspiration for those new sections. I thank my students, who provided the inspiration over many years for this book. They have also been my best sounding board, allowing me to test new ideas and organize the text as it now appears. I thank my friends and colleagues who have provided countless suggestions and encouragement over the years, particularly George Majetich (Georgia), Frederick Luzzio (Louisville), Spencer Knapp (Rutgers), and Phil Garner (Washington State). You have all helped more than you can possibly know, and I am most grateful. A special thanks to my wife Sarah whose patience and understanding made the work possible, and to my son Steven. Every effort has been made to keep the manuscript error free. Where there are errors, I offer my apologies and take complete responsibility. If there are corrections and/or suggestions, please let me know by Email or normal post. Any error list will be posted on the Elsevier website for this book. I thank Ms. Jill Cetel for her work in the early days of this project, and especially Ms. Katey Birtcher, who worked to develop the manuscript in its present form, and Anitha Sivaraj, who made production of this book possible. A special thank you goes to Ms. Jeanette Stiefel, whose care and expertise ensured the highest quality of this book. Thank you again for using this new edition. I hope that it is useful to you in your studies. Michael B. Smith University of Connecticut, Storrs, CT, United States March, 2016

Preface to the Third Edition

The new edition of Organic Synthesis has been revised and rewritten from front to back. I want to thank all who used the book in its first and second editions. The book has been out of print for several years, but the collaboration of Warren Hehre and Wavefunction, Inc. made the third edition possible. It is the same graduate level textbook past users are familiar with, with two major exceptions. First, the book has been revised and updated. Second, molecular modeling problems are included in a manner that is not obtrusive to the theme of understanding reactions and synthesis. A total of 64 molecular modeling problems are incorporated into various discussions, spread throughout 11 of the 13 reactions-synthesis oriented chapters. Spartan models for each problem are provided on an accompanying CD, and Spartan Model is included. These features will allow the reader to manipulate each model and, in most cases, change or create model compounds of interest to the reader. It is our belief that the selected molecular modeling problems will offer new insights into certain aspects of chemical reactivity, conformational analysis, and stereoselectivity. Updated examples are used throughout the new edition when possible, and new material is added that make this edition reflect current synthetic methodology. The text has been modified in countless places to improve readability and pedagogy. This new edition contains references taken from more than 6100 journal articles, books, and monographs. Of these references, more than 950 are new to this edition, all taken from the literature after 2002. More than 600 updated or new reactions have been added. There are several entirely new sections that discuss topics missing in the second edition. These include SN2 type reactions with epoxides; the Burgess Reagent; functional group rearrangements (Beckmann, Schmidt, Curtius, Hofmann, and Lossen); oxidation of allylic carbon with ruthenium compounds; a comparison of LUMO-mapping with the Cram model and Felkin-Anh models in Chapter 4; electrocyclic reactions; [2.3]-sigmatropic rearrangement (Wittig rearrangement); and consolidation of CdC bond forming reactions of carbocations and nucleophiles into a new section. Homework in each chapter has been extensively revised. There are more than 800 homework problems, and more than 300 of the homework problems are new. Most of the homework problems do not contain leading references for the answers. The answers to all problems from Chapters 1–9, and 11–13 are available in an on-line Student Solutions Manual for this book. As in previous edition, a few leading references are provided for the synthesis problems in Chapter 10. Although answers are given for homework that relates to all other chapters, in Chapter 10 most problems do not have answers. The student is encouraged to discuss any synthetic problem with their instructor. With the exception of scanned figures, all drawings in this book were prepared using ChemDraw, provided by CambridgeSoft, and all 3D graphics are rendered with Spartan, provided by Wavefunction, Inc. I thank both organizations for providing the software that made this project possible. I express my gratitude to all of those who were kind enough to go through the first and second editions and supply me with comments, corrections, and suggestions. For this new edition, special thanks and gratitude are given to Warren Hehre. Not only did he design the molecular modeling problems, but also included the CD with solutions to the problems and the accompanying software are provided by Warren. My thanks go deeper than that however. Warren, thorough Wavefunction, Inc., is publishing this edition and without his hard work and help this book would not have been possible. He has also helped me think about certain aspects of organic synthesis in a different way because of the modeling, and I believe this has greatly improved the book and the approaches presented in the book. Special thanks are also given to Ms. Pamela Ohsan, who converted the entire book into publishable form. Once again, without her extraordinary efforts, this third edition would not be possible. Finally, I thank my students, who have provided the inspiration over the years for this book. They have also been my best sounding board, allowing me to test new ideas and organize the text as it now appears. I thank my friends and

xiii

xiv

PREFACE TO THE THIRD EDITION

colleagues who have provided countless suggestions and encouragement over the years, particularly Spencer Knapp (Rutgers), George Majetich (Georgia), Frederick Luzzio (Louisville), and Phil Garner (Washington State). You have all helped more than you can possibly know, and I am most grateful. A special thanks to my wife Sarah and son Steven, whose patience and understanding made the work possible. Michael B. Smith Storrs, Connecticut April, 2010

Preface to the First Edition. Why I Wrote This Book!

A reactions oriented course is a staple of most graduate organic programs, and synthesis is taught either as a part of that course or as a special topic. Ideally, the incoming student is an organic major, who has a good working knowledge of basic reactions, stereochemistry, and conformational principles. In fact, however, many (often most) of the students in a first year graduate level organic course have deficiencies in their undergraduate work, are not organic majors and are not synthetically inclined. Does one simply tell the student to “go away and read about it,” giving a list of references, or does one take class time to fill in the deficits? The first option works well for highly motivated students with a good background, less well for those with a modest background. In many cases, the students spend so much time catching up that it is difficult to focus them on the cutting edge material we all want to teach. If one exercises the second option of filling in all the deficits, one never gets to the cutting edge material. This is especially punishing to the outstanding students and to the organic majors. A compromise would provide the student with a reliable and readily available source for background material that could be used as needed. The instructor could then feel comfortable that the proper foundations have been laid and push on to more interesting areas of organic chemistry. Unfortunately, such a source of background material either is lacking altogether or consists of several books and dozens of review articles. I believe my teaching experience at UConn as just described is rather typical, with a mix of nonorganic majors, outstanding and well-motivated students, and many students with weak backgrounds who have the potential to go on to useful and productive careers if time is taken to help them. Over the years I have assigned what books were available in an attempt to address these problems, but found that “graduate level textbooks” left much to be desired. I assembled a large reading list and mountains of handouts and spent half of my life making up problems that would give my students a reasonable chance at practicing the principles we were discussing. I came to the conclusion that a single textbook was needed that would give me the flexibility I craved to present the course I wanted to teach, but yet would give the students the background they needed to succeed. As I tried different things in the classroom, I solicited the opinions of the graduate students who took the course and tried to develop an approach that worked for them and allowed me to present the information I wanted. The result is this book. I hope that it is readable, provides background information, and also provides the research-oriented information that is important for graduate organic students. I also hope it will be of benefit to instructors who face the same challenges I do. I hope this book will be a useful tool to the synthetic community and to graduate level education. From talks with many people I know that courses for which this book is targeted can be for either one or two semesters. The course can focus only on functional groups, only on making carbon-carbon bonds, or some combination of both (like my course), or only on synthesis. I have tried to organize the book in such a way that one is not a slave to its organization. Every chapter is internally cross-referenced. If the course is to focus upon making carbon-carbon bonds, for example, there are unavoidable references to oxidation reagents, reducing agents, stereochemical principles, etc. When such a reaction or principle appears, the section and chapter where it is discussed elsewhere in the book is given “in line” so the student can easily find it. It is impossible to write each chapter so it will stand alone, but the chapters are reasonably independent in their presentations. I have organized the book so that functional groups are discussed in the first few chapters and carbon-carbon bond formation reactions are discussed in later chapters, making it easier to use the one book for two different courses or for a combined course. The middle chapters are used for review and to help the student make the transitions from functional group manipulations to applying reactions and principles and thence to actually building molecules. I believe that a course devoted to making carbon-carbon bonds could begin with Chapter 8, knowing that all pertinent peripheral material is in the book and readily available to the student. The ultimate goal of the book is to cut down on the mountains of handouts, provide homework to give the student proper practice, give many literature citations to tell the student exactly where to find more information, and allow the instructor to devote time to their particular focus.

xv

xvi

PREFACE TO THE FIRST EDITION. WHY I WROTE THIS BOOK!

This book obviously encompasses a wide range of organic chemistry. Is there a theme? Should there be? The beautiful and elegant total syntheses of interesting and important molecules published by synthetic organic chemists inspired me to become an organic chemist and I believe that synthesis focuses attention on the problems of organic chemistry in a unique way. To solve a synthetic problem, all elements of organic chemistry must be brought to bear: reactions, mechanism, stereochemistry, conformational control, and strategy. Synthesis therefore brings a perspective on all aspects of organic chemistry and provides a theme for understanding it. The theme of this book is therefore the presentation of reactions in the context of organic synthesis. Wherever possible, examples of a given reaction, process, or strategy are taken from a published total synthesis. The disconnection approach is presented in the first chapter, and as each new functional group transform and carbon-carbon bond forming reaction is discussed, the retrosynthetic analysis (the disconnect products for that reaction) is given. An entire chapter (Chapter 10) is devoted to synthetic strategies, and Chapter 14 provides examples of first year students’ first syntheses. I believe that this theme is a reasonable and useful device for presenting advanced organic chemistry. The text is fully referenced to facilitate further study, and (where feasible) the principal researcher who did the work is mentioned by name, so the student can follow that person’s work in the literature and gain even more insight into a given area. As far as it is known to me, the pioneering work of the great chemists of the past has been referenced. Many of the “named reactions” are no longer referenced in journals, but when they are first mentioned in this book, the original references are given. I believe the early work should not be lost to a new generation of students. In many cases I have used 3-D drawings to help illustrate stereochemical arguments for a given process. I give the structure of each reagent cited in the text, where that reagent is mentioned, so a beginning student does not have to stop and figure it out. This is probably unnecessary for many students, but it is there if needed. This is a reaction-oriented book, but an attempt is made to give brief mechanistic discussions when appropriate. In addition, some physical organic chemistry is included to try to answer the obvious if unasked questions: why does that alkyl group move, why does that bond break, why is that steric interaction greater than the other one, or why is that reaction diastereoselective? Most of all, a student needs to practice. Chapters 1–13 have end-of chapter problems that range from those requiring simple answers based on statements within the text to complex problems taken from research literature. In a large number of cases literature citations are provided so answers can be found. The first part of the book (Chapters 1–4) is a review of functional group transforms and basic principles: retrosynthesis, stereochemistry, and conformations. Basic organic reactions are covered, including substitution reactions, addition reactions, elimination reactions, acid/base chemistry, oxidation, and reduction. The first two chapters are very loosely organized along the lines of an undergraduate book for presenting the functional group reactions (basic principles, substitution, elimination, addition, acyl addition, and aromatic chemistry). Chapter 1 begins with the disconnection approach. I have found that this focuses the students’ attention on which reactions they can actually apply and instantly shows them why it is important to have a larger arsenal of reactions to solve a synthetic problem. This has been better than any other device I have tried and that is why it is placed first. Most of the students I see come into our program deficient in their understanding of stereochemistry and conformational control, and so those topics are presented next. Some of this information is remedial material and where unneeded can be skipped, but it is there for those who need it (even if they will not admit that they do). Chapter 2 presents a mini-review of undergraduate organic chemistry reactions and also introduces some modern reactions and applications. Chapter 3 is on oxidation and Chapter 4 is on reduction. Each chapter covers areas that are woefully under-emphasized in undergraduate textbooks. Chapter 5 covers hydroboration, an area that is discussed in several books and reviews. I thought it useful to combine this material into a tightly focused presentation which (1) introduces several novel functional group transforms that appear nowhere else and (2) gives a useful review of many topics introduced in Chapters 1–4. Chapter 6 reviews the basic principles that chemists use to control a reaction rather than be controlled by it. It shows the techniques chemists use to “fix” the stereochemistry, if possible, when the reaction does not do what it is supposed to. It shows how stereochemical principles guide a synthesis. An alternative would be to separate stereochemistry into a chapter that discusses all stereochemical principles. However, the theme is synthesis, and stereochemical considerations are as important a part of a synthesis as the reagents being chosen. For that reason, stereochemistry is presented with the reactions in each chapter. Chapter 6 simply ties together the basic principles. This chapter also includes the basics of ring-forming reactions. Chapter 7 completes the first part of the book and gives a brief overview of what protecting groups are and when to use them. The second half of the book focuses on making carbon-carbon bonds. It is organized fundamentally by the disconnection approach. In Chapter 1, breaking a carbon-carbon bond generated a disconnect product that was labeled as Cd (a nucleophilic species), Ca (an electrophilic species), or Cradical (a radical intermediate). In some cases, multiple bonds were disconnected, and many of these disconnections involved pericyclic reactions to reassemble the target. The

PREFACE TO THE FIRST EDITION. WHY I WROTE THIS BOOK!

xvii

nucleophilic regents that are equivalent to Cd disconnect products are covered in Chapters 8 and 9, with the very important enolate anion chemistry separated into Chapter 9. Chapter 10 presents various synthetic strategies that a student may apply to a given synthetic problem. This information needs to be introduced as soon as possible, but until the student “knows some chemistry,” it cannot really be applied. Placement of synthetic strategies after functional group transforms and nucleophilic methods for making carbon-carbon bonds is a reasonable compromise. Chapter 11 introduces the important Diels-Alder cyclization, as well as dipolar cycloadditions and sigmatropic rearrangements that are critically important to synthesis. Chapter 12 explores electrophilic carbons (Ca), including organometallics that generally react with nucleophilic species. Chapter 13 introduces radical and carbene chemistry. Chapter 14 is included to give the student a taste of a first-time student proposal and some of the common mistakes. The point is not to reiterate the chemistry but to show how strategic shortsightedness, poor drawings, and deficiencies in overall presentation can influence how the proposal is viewed. It is mainly intended to show some common mistakes and also some good things to do in presenting a synthesis. It is not meant to supersede the detailed discussions of how and why a completed elegant synthesis is done but to assist the first-time student in preparing a proposal. The goal of this work is to produce a graduate level textbook, and it does not assume that a student should already know the information, before the course. I hope that it will be useful to students and to the synthetic community. Every effort has been made to keep the manuscript error-free. Where there are errors, I take full responsibility and encourage those who find them to contact me directly, at the address given below, with corrections. Suggestions for improving the text, including additions and general comments about the book are also welcome. My goal is to incorporate such changes in future editions of this work. If anyone wishes to contribute homework problems to future editions, please send them to me and I will, of course, give full credit for any I use. I must begin my “thank yous” with the graduate students at UConn, who inspired this work and worked with me through several years to develop the pedagogy of the text. I must also thank Dr. Chris Lipinski and Dr. David Burnett of Pfizer Central Research (Groton, CT) who organized a reactions/synthesis course for their research assistants. This allowed me to test this book upon an “outside” and highly trained audience. I am indebted to them for their suggestions and their help. There are many other people to thank. Professor Janet Carlson (Macalester College) reviewed a primeval version of this book and made many useful comments. Professors Al Sneden and Suzanne Ruder (Virginia Commonwealth University) classroom tested an early version of this text and both made many comments and suggestions that assisted me in putting together the final form of this book. Of the early reviewers of this book I would particularly like to thank Professor Brad Mundy (Colby College) and Professor Marye Anne Fox (University of Texas, Austin), who made insightful and highly useful suggestions that were important for shaping the focus of the book. Along the way, many people have helped me with portions or sections of the book. Professor Barry Sharpless (Scripps) reviewed the oxidation chapter and also provided many useful insights into his asymmetric epoxidation procedures. Dr. Peter Wuts (Upjohn) was kind enough to review the protecting group chapter (Chapter 7) and helped me focus it in the proper way. Professor Ken Houk (UCLA), Professor Stephen Hanessian (Universite de Montreal), Professor Larry Weiler (U. of British Columbia), Professor James Hendrickson (Brandeis), Professor Tomas Hudlicky (U. Florida), and Professor Michael Taschner (U. of Akron) reviewed portions of work that applied to their areas of research and I am grateful for their help. Several people provided original copies of figures or useful reprints or comments. These include Professor Dieter Seebach (ETH), Professor Paul Williard (Brown), Professor E.J. Corey (Harvard), Dr. Frank Urban (Pfizer Central Research), Professor Rene Barone (Universite de Marseilles), and Professor Wilhelm Meier (Essen). Two professors reviewed portions of the final manuscript and not only pointed out errors but made enormously helpful suggestions that were important for completing the book: Professor Fred Ziegler (Yale) and Professor Douglass Taber (U. of Delaware). I thank both of them very much. There were many other people who reviewed portions of the book and their reviews were very important in shaping my own perception of the book, what was needed and what needed to be changed. These include: Professor Winfield M. Baldwin, Jr. (U. of Georgia), Professor Albert W. Burgstahler (U. of Kansas), Professor George B. Clemens (Bowling Green State University), Professor Ishan Erden (San Francisco State University), Professor Raymond C. Fort, Jr. (U. of Maine), Professor John F. Helling (U. of Florida), Professor R. Daniel Little (U. of California), Professor Gary W. Morrow (U. of Dayton), Professor Michael Rathke (Michigan State University), Professor Bryan W. Roberts (U. of Pennsylvania), Professor James E. Van Verth (Canisius College), Professor Frederick G. West (U. of Utah), and Professor Kang Zhao (New York University). I thank all of them. I must also thank the many people who have indulged me at meetings, at Gordon conferences, and as visitors to UConn and who discussed their thoughts, needs, and wants in graduate level education. These discussions helped shape the way I put the book together.

xviii

PREFACE TO THE FIRST EDITION. WHY I WROTE THIS BOOK!

Finally, but by no means last in my thoughts, I am indebted to Professors Joe Wolinsky and Jim Brewster of Purdue University. Their dedication and skill taught me how to teach. Thank you! I particularly want to thank my wife Sarah and son Steven. They endured the many days and nights of my being in the library and the endless hours on the computer with patience and understanding. My family provided the love, the help, and the fulfillment required for me to keep going and helped me to put this project into its proper perspective. They helped me in ways that are too numerous to mention. I thank them and I dedicate this work to them. Michael B. Smith

Common Abbreviations

Other, less common abbreviations are given in the text when the term is used.

O Ac

acetyl

acac AIBN All Am aq Ar ax

acetylacetonate azobisisobutyronitrile allyl amyl (dCH2 (CH2)3CH3) aqueous aryl axial

Me

B

9-borabicyclo[3.3.1]nonylboryl

9-BBN BINAP BINOL Bn

9-borabicyclo[3.3.1]nonane (2R,3S)-2,20 -bis(diphenylphosphino)-1,10 -binapthyl 1,10 -bi-2-naphthol benzyl (dCH2Ph)

Boc

tert-butoxycarbonyl

Bom Bpy Bu Bz CAN ccat

benzyloxymethyl 2,20 -bipyridyl, 2,20 -bipyridine n-butyl (dCH2CH2CH2CH3) benzoyl ceric ammonium nitrate ((NH4)2Ce(NO3)6) cyclocatalytic

Cbz

carbobenzyloxy

Chap Chirald CIP cod cot Cp CSA CTAB

chapter(s) (2S,3R)-(+)-4-dimethylamino-1,2-diphenyl-3-methylbutan-2-ol Cahn-Ingold-Prelog 1,5-cyclooctadiene 1,3,5-cyclooctatriene cyclopentadienyl camphorsulfonic acid cetyltrimethylammonium bromide (C16H33NMe+3 Br)

Cy (c-C6H11)

cyclohexyl

°C DABCO d dba DBE DBN DBU DCC DCE DCM DDQ % de DEA

temperature in degrees Celsius 1,4-diazabicyclo[2.2.2]octane day(s) dibenzylideneacetone 1,2-dibromoethane (BrCH2CH2Br) 1,5-diazabicyclo[4.3.0]non-5-ene 1,8-diazabicyclo[5.4.0]undec-7-ene 1,3-dicyclohexylcarbodiimide (c-C6H11dN]C]Ndc-C6H11) 1,2-dichloroethane (ClCH2CH2Cl) dichloromethane 2,3-dichloro-5,6-dicyano-1,4-benzoquinone % diastereomeric excess diethylamine (HN(CH2CH3)2)

O Ot-Bu

O OCH2 Ph

xix

xx

COMMON ABBREVIATIONS

DEAD DET (DHQD)2PHAL) (DHQ)2PHAL) DHP DIAD dibal DIPEA diphos (dppe) diphos-4 (dppb) DIPT DMA DMAP DMB DMDO DME

diethyl azodicarboxylate (EtO2CdN]NCO2Et) diethyl tartrate 1,4-dichlorophthalazine adduct with dihydroquinidine 1,4-dichlorophthalazine adduct with dihydroquinone dihydropyran diisopropyl azodicarboxylate diisobutylaluminum hydride ((Me2CHCH2)2AlH) diisopropylethylamine 1,2-bis(diphenylphosphino)ethane (Ph2PCH2CH2PPh2) 1,4-bis(diphenylphosphino)butane (Ph2P(CH2)4PPh2) diisopropyl tartrate dimethylacetamide 4-(N,N-dimethylamino)pyridine 3,4-dimethoxybenzyl ether 3,3-dimethyldioxirane dimethoxyethane (MeOCH2CH2OMe)

DMF

N,N0 -dimethylformamide

DMS DMSO (solvent) dmso (ligand) dppb dppe dppf dppp dr dvb e2 EA % ee EE er Et EDA EDTA equiv ESR FMO fod Fp FVP GC gl 1 H NMR h hν 1,5-HD HMDS HMPA HMPT HOMO HPLC HSAB IP Ipc2BH i-Pr IR IUPAC LICA (LIPCA) LDA LHASA LHMDS LTMP LUMO

dimethyl sulfide dimethyl sulfoxide dimethyl sulfoxide 1,4-bis(diphenylphosphino)butane (Ph2P(CH2)4PPh2) 1,2-bis(diphenylphosphino)ethane (Ph2PCH2CH2PPh2) (1,10 -bis(diphenylphosphino)ferrocene) 1,3-bis(diphenylphosphino)propane (Ph2P(CH2)3PPh2) diastereomeric ratio divinylbenzene electron transfer electron affinity % enantiomeric excess 1-ethoxyethoxy (EtO(Me)CHd) enantiomeric ratio ethyl (dCH2CH3) ethylenediamine (H2NCH2CH2NH2) ethylenediaminetetraacetic acid equivalent(s) electron spin resonance frontier molecular orbital tris-(6,6,7,7,8,8,8)-heptafluoro-2,2-dimethyl-3,5-octanedionate cyclopentadienylbis(carbonyl iron) (ferrocene) flash vacuum pyrolysis gas chromatography glacial proton nuclear magnetic resonance hour (hour) irradiation with light 1,5-hexadienyl hexamethyldisilazide hexamethylphosphoramide ((Me2N)3P]O) hexamethylphosphorus triamide ((Me2N)3P) highest occupied molecular orbital high performance liquid chromatography Hard-Soft Acid-Base ionization potential diisopinocampheylborane isopropyl (dCH(Me)2) infrared International Union of Pure and Applied Chemistry lithium isopropylcyclohexylamine lithium diisopropylamide (LiN(i-Pr)2) logic and heuristics applied to synthetic analysis lithium hexamethyldisilazide (LiN(SiMe3)2) lithium 2,2,6,6-tetramethylpiperidide lowest unoccupied molecular orbital

O H

NMe2

COMMON ABBREVIATIONS

mcpba Me MEM Mes min MOM Ms MS MTM MVK NAD NADP NADPH NAP napth NBD NBS NCS NIS Ni(R) NMO N.R. Nu (Nuc) OBs Oxone

P

or

meta-chloroperoxybenzoic acid methyl (dCH3 or Me) β-methoxyethoxymethyl (MeOCH2CH2OCH2d) mesityl (2,4,6-tri-Me-C6H2) minutes methoxymethyl (MeOCH2d) methanesulfonyl (MeSO2d) molecular sieves (3 or 4 Å) methylthiomethyl methyl vinyl ketone (MeSCH2d) nicotinamide adenine dinucleotide nicotinamide adenine dinucleotide phosphate reduced nicotinamide adenine dinucleotide phosphate 2-napthylmethyl napthyl (C10H8) norbornadiene N-bromosuccinimide N-chlorosuccinimide N-iodosuccinimide Raney nickel N-methylmorpholine N-oxide no reaction nucleophile O-benzenesulfonate 2 KHSO5  KHSO4  K2SO4 polymeric backbone

PCC PDC PEG pet

pyridinium chlorochromate pyridinium dichromate poly(ethylene glycol) petroleum

Ph

phenyl

PhH PhMe phen Phth

benzene toluene 1,10-phenanthroline phthaloyl

Pip

piperidino

Piv PMB PMP PPA PPTS Pr

pivaloyl p-methoxybenzyl p-methoxyphenyl polyphosphoric acid p-toluenesulfonic acid n-propyl (dCH2CH2CH3)

Py

pyridine N

PyBOP quant Red-Al rt sBu sBuLi s salen SEM SET Siamyl (Sia)2BH SOMO TASF TBAF TBDMS or TBS TBDPS TBHP (t-BuOOH)

(benzotriazol-1-yloxy)tripyrrolidinophosphonium hexafluorophosphate quantitative yield [(MeOCH2CH2O)2AlH2]Na room temperature sec-butyl (CH3CH2CH(CH3)) sec-butyllithium (CH3CH2CH(Li)CH3) second bis(salicylidene)ethylenediamine 2-(trimethylsilyl)ethoxymethyl single electron transfer sec-isoamyl ((CH3)2CHCH(CH3)d) disiamylborane singly occupied molecular orbital tris-(diethylamino)sulfonium difluorotrimethyl silicate tetrabutylammonium fluoride (n-Bu4N+ F) tert-butyldimethylsilyl (t-BuMe2Si) tert-butyldiphenylsilyl (dBuPh2Si) tert-butylhydroperoxide (Me3COOH)

N

xxi

xxii TES t-Bu TEBA TEMPO Tf (OTf) TFA TFAA ThexBH2 THF (solvent) Thf (ligand) THP TIPS TMEDA TMS TMP Tol TPAP Tr TRIS Ts(Tos) UV VSEPR Xc

COMMON ABBREVIATIONS

triethylsilyl tert-butyl (dCMe3) triethylbenzylammonium (Bn(Et3)3N+) tetramethylpiperidinyloxy free radical triflate (dSO2CF3 (dOSO2CF3)) trifluoroacetic acid (CF3COOH) trifluoroacetic anhydride ((CF3CO)2O) thexylborane (tert-hexylborane) tetrahydrofuran tetrahydrofuran tetrahydropyranyl triisopropylsilyl N,N,N0 ,N0 -tetramethylethylenediamine (Me2NCH2CH2NMe2) trimethylsilyl (dSi(CH3)3) 2,2,6,6-tetramethylpiperidine tolyl (4-(Me)C6H4) tetrapropylammonium perruthenate trityl (dCPh3) triisopropylphenylsulfonyl tosyl ¼ p-toluenesulfonyl (4-(Me)C6H4SO2) ultraviolet valence shell electron pair repulsion chiral auxiliary

C H A P T E R

1 Retrosynthesis, Stereochemistry, and Conformations 1.1 INTRODUCTION The total synthesis of complex molecules demands a thorough knowledge of reactions that form carbon-carbon bonds, as well as those that change one functional group into another. The largest number of chemical reactions used in a synthesis involve the manipulation of functional groups. Further, the synthesis of a molecule is rarely successful unless all aspects of chemical reactivity, functional group interactions, conformations, and stereochemistry are well understood. Today the term organic synthesis encompasses an enormous variety of chemical reactions. Planning and using organic transformation to put together a molecule is certainly an important aspect of organic synthesis. A thorough understanding of the many organic reactions, reagents, and chemical transformations that are now known is required to accomplish this goal. As noted, the practice of organic synthesis requires an understanding of chirality and the stereochemistry of molecules, both for developing a synthetic strategy and for the choice of reactions and reagents used for various chemical transformations. Conformational analysis of each molecule, from starting material to final product, must be understood because chemical reactivity and stereochemistry are often influenced by the conformation. Perhaps the most important component of planning an organic synthesis is a thorough and intimate knowledge of chemical reactions and reagents. If one knows only one reagent to convert an alcohol to a ketone, and if that reagent does not work for a given system, there is no alternative. On the other hand, if one knows 30 different reagents for that transformation, there are many alternatives if one of them does not work. Perhaps more importantly, understanding the 30 reagents allows one to better plan a synthesis to use a certain reagent that will maximize the chance that the synthetic sequence will go as planned. The same comment applies to making carbon-carbon bonds. Presumably, a synthesis begins with a starting material of a few carbon atoms, and reactions will add carbon fragments to increase the complexity of the molecules as it is transformed in many steps to the final target. Understanding different reactions and reagents that form different types of carbon-carbon bonds is therefore essential. This book has the title Organic Synthesis, but from the preceding paragraph it is clear that organic synthesis begins and ends with reactions. The goal of this text is to explain and provide examples of the many reactions that manipulate functional groups (functional group exchange reactions), as well as those that form carbon-carbon bonds. Examples of various reactions are provided, taken from published organic syntheses, to provide an example and also to show the context of how they are used. Before discussing these reactions, it is important to present a brief overview of structural features that are important in planning a synthesis. A full discussion of strategies for total synthesis will be introduced in Chapter 8. Reviews of stereochemistry and conformational analysis are provided in this chapter for the same reason. The operating paradigm is that these concepts are an important part of chemical reactivity with respect to why some reactions work better than others, and why some reagents are better suited for a given application. Changing one functional group into another is defined as a functional group interchange (FGI). Simple examples are the SN2 reaction of 3-bromopentane with the nucleophilic cyanide ion of NaCN to form 2-ethylbutanenitrle, and the E2 reaction of 2-bromo-2-methylpentane via reaction with the basic KOH to yield 2-methylpent-2-ene. The first transformation changes a halide to a cyano group (alkyl halide to nitrile), whereas the second transformation changes a halide to an alkene (alkyl halide to alkene). Such functional group exchange reactions are important because they incorporate

Organic Synthesis http://dx.doi.org/10.1016/B978-0-12-800720-4.00001-5

1

Copyright © 2017 Michael Smith. Published by Elsevier Inc. All rights reserved.

2

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

key functionality into the final target, but they are also used to “set up” the molecule for making a carbon-carbon bond (DMF ¼ N,N-dimethylformamide). Br

CN

NaCN DMF

3-Bromopentane

2-Ethylbutanenitrile KOH , EtOH

Br 2-Methylpent-2-ene

2-Bromo-2-methylpentane

A reaction that brings reactive fragments together to form a new bond between two carbon atoms is classified as a carbon-carbon bond-forming reaction, and it is clear that such reactions are used to increase the size of a molecule. Such reactions are obviously critical to total synthesis. An example is the reaction of a carbon nucleophile with an alkyl halide in an SN2 reaction. The conjugate base formed when prop-1-yne reacts with a strong base (e.g., sodium amide) is an alkyne anion. Under SN2 reaction conditions (Section 3.2.1), this alkyne anion is a carbon nucleophile that reacts with (2S)-bromobutane with inversion to yield (4R)-methylhex-2-yne. Note that a new carbon-carbon bond has been formed, that the new alkyne product has more carbon atoms, and it is more complex than either prop-1-yne or (2S)bromobutane. (4R)-Methylhex-2-yne has a stereogenic carbon atom and the SN2 reaction has proceeded with 100% inversion of configuration, so there is chirality transfer from the enantiopure (2S)-bromobutane to (4R)methylhex-2-yne. With respect to synthetic planning, the carbon-carbon bond-forming reaction was chosen to control stereochemistry in the product. This example was chosen to show why it is important to understand a given reaction, the reactivity of the chosen reagent with the substrate, and possible stereochemical consequences before choosing that chemical reaction (THF ¼ tetrahydrofuran, solvent). Br

H

NaH2+JNH2

: – Na+

(2S)-Bromobutane

THF

THF

Prop-1-yne

(4R)-Methylhex-2-yne

Today, the relationship of two molecules in a synthesis is commonly shown using a device known as a transform, defined by Corey and Cheng1 as: “the exact reverse of a synthetic reaction to a target structure.” Using the conversion of prop-1-yne to (4R)-methylhex-2-yne, the target structure is (4R)-methylhex-2-yne (Fig. 1.1), and it is the molecule that is the object of the synthesis. The transform for this synthetic step is, therefore, (4R)-methylhex-2-yne ) prop-1-yne, which requires that one mentally break the highlighted bond (bond a) in (4R)-methylhex-2-yne (represented by the dashed line) leads to fragments prop-1-yne and (2S)-bromobutane and in this process bond a is said to be disconnected. The “backward arrow” ()) is used to indicate that prop-1-yne is the starting material for the preparation of (4R)methylhex-2-yne. a : – Na+

Transform (4R)-Methylhex-2-yne

Synthesis

Br

+ (2S)-Bromobutane

Br

: – Na+ THF

FIG. 1.1

Transform versus synthesis.

How is the disconnection approach useful in planning a synthesis? In part, this statement suggests that it is important to understand why bond a in Fig. 1.1 was chosen for the disconnection. The choice of bond a is based on a thorough 1

Corey, E. J.; Cheng, X. The Logic of Chemical Synthesis; Wiley–Interscience: New York, NY, 1989.

3

1.1 INTRODUCTION

knowledge of the chemical properties of (4R)-methylhex-2-yne. When bond a was disconnected, it is with the understanding that bond a will be made during the synthesis by a specific chemical reaction. To understand the structural characteristics of (4R)-methylhex-2-yne that led to the disconnection of bond a, the chemical reaction(s) required to form that bond must be known and understood. Disconnection of a bond in the target leads to so-called disconnection fragments or disconnection products, which are not real compounds. Disconnection of bond a in (4R)-metylhex-2-yne, for example, does not lead directly to real molecules, but to fragments 1 and 2. These disconnection fragments are intended to point the chemist toward a chemical reaction between two reactive partners. The structures of 1 and 2 allow certain assumptions to be made that will correlate any disconnection fragment with a real molecule a • (4R)-Methylhex-2-yne

1

+

•

2

To convert 1 and 2 to real fragments, another operation is required that correlates each disconnect fragment with what is known as a synthetic equivalent. The process used for this correlation is discussed in Section 1.2. Once actual molecules have been identified as capable of undergoing a chemical reaction that will make the target, in this case (4R)methylhex-2-yne, the synthesis can proceed. In effect, disconnection of the target to yield fragments “works backward” to discover the starting materials that are required to make the target. Working backward in this manner is termed retrosynthetic analysis or retrosynthesis,2 defined by Corey2 as “a problem-solving technique for transforming the structure of a synthetic target molecule to a sequence of progressively simple materials along a pathway that ultimately leads to a simple or commercially available starting material for chemical synthesis.” In principle, the synthesis is the “reverse” of the disconnection. The use of a “backward arrow” ()) indicates disconnection of the target to give disconnect fragments. This individual disconnection is usually referred to as a transform. The retrosynthesis shown is a single disconnection that points to a single reaction, but to complete a real synthesis starting from a given target, reagents must be provided. In this case, the alkyne is treated with a suitable base to generate the corresponding alkyne anion, which subsequently reacts with the chiral alkyl halide. It is important to point out that a retrosynthetic analysis rarely correlates with the exact reverse track with simple reagents to synthesize the target. For molecules with multiple functionality, particularly complex natural products, the idea of doing a retrosynthetic analysis and simply providing reagents for each disconnection to convert the starting material to the target is usually problematic. A given disconnection may not be possible unless a functional group is changed or modified. Commonly, there are steps that simply do not work using available reagents or those suggested by literature precedent. In addition, reactions may give poor yields or the wrong stereochemistry. There may also be unanticipated interactions of functional groups and unexpected requirements for protecting groups (see Chapter 5). In short, the approach shown here is a first step, intended to begin the retrosynthesis process, and to think about what reactions may be appropriate to put them together again. The discussion given here is intended to show the importance of a basic understanding of organic chemistry. Other important concepts in organic chemistry must be brought to bear, including stereochemistry and conformational theory. The issues of stereochemistry and conformation with respect to organic reactions can be illustrated by the simple transform (2R)-(hydroxymethyl)cyclohexan-(1S)-ol ) (2R)-(hydroxymethyl)cyclohexan-1-one, which is arguably a single chemical reaction (a reduction, see Section 7.5). This disconnection demands the use of a reaction that will provide the relative stereochemistry (trans) shown in (2R)-(hydroxymethyl)cyclohexan-(1S)-ol. This demand cannot be satisfied unless the proper spatial relationship of the functional groups in the target is known prior to making the choice for a chemical reaction. That relationship is the conformation of the cyclohexanone starting material (see Section 1.5.2), which can be difficult to see using the two-dimensional (2D) structures shown. The molecular model shown for (2R)-(hydroxymethyl)cyclohexan-1-one provides a better perspective of the relative positions of all atoms in the molecule (the stereochemical and conformational relationships; see Sections 1.4 and 1.5). In this case, the molecular model shows that the hydroxyl group is positioned on one side of the molecule, relative to the carbonyl group. If a chelating metal is used in a hydride reduction reagent (e.g., zinc borohydride; see Section 7.5), coordination with the hydroxyl group (3) will deliver the hydride primarily from the same face as the CH2OH unit, leading to the observed

2

Reference 1, p 6.

4

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

trans stereochemistry in (2R)-(hydroxymethyl)cyclohexan-(1S)-ol. In this example, the conformation of the molecule provides a model that helps to plan a desired stereochemical outcome of the reaction. O

OH OH

OH

(2R)-(Hydroxymethyl)cyclohexan-1-one

(2R)-(Hydroxymethyl)cyclohexan-(1S)-ol

H2 B O

HO

BH2 Zn

Zn(BH4)2

O

H

O

H3

H

O+

HO

HO

H

H H 3

(2R)-(Hydroxymethyl)cyclohexan-(1S)-ol

(2R)-(Hydroxymethyl)cyclohexan-1-one

Three important factors will be reviewed in this chapter: (1) disconnection and retrosynthesis, (2) stereochemistry, (3) conformational analysis. This review will not formally introduce synthesis (synthetic strategies will be presented in Chapter 8),3 so the goal is to understand concepts that place the use of reactions in proper context.

1.2 THE DISCONNECTION PROTOCOL When retrosynthetic theory is introduced as part of a course, it is easy to become concerned about making the “correct” disconnection rather than focusing on the chemical reactions and concepts required to form the disconnected bond in the synthesis. In reality, the choice of a disconnection is usually dictated by the ability to form a given bond, in the context of stereochemistry and selectivity, and not the other way around. The synthesis of organic molecules dates to the nineteenth century, but the work of Perkin, Robinson, and others in the early twentieth century demonstrated the importance of synthetic planning.1 In the 1940s and 1950s, Woodward, Robinson, Eschenmoser, Stork, and others clearly showed how molecules could be synthesized in a logical and elegant manner. In the 1960s, Corey identified the rationale behind his syntheses, and such logical synthetic plans (termed retrosynthetic analyses) are now a common feature of the synthetic literature.2 The disconnection approach is now used to teach synthesis, and Warren4 has several books that describe this approach in great detail. There are several other books that describe syntheses and approaches to total synthesis.5 Several different strategies for the synthesis of organic molecules are available (see Chapter 8) and all are useful. All disconnection approaches assign priorities to bonds in a molecule and disconnect those bonds with the highest priorities, as with Corey’s strategic bond analysis (see Section 8.5).1,3 Smith6 described a simple method where the priorities are based only on the relative ability to chemically form the bond broken in the disconnection, based on known reactions. When these strategies are applied to a first synthesis in an introductory course, the first issue raised after a disconnection is what to do with the disconnect products. 3

(a) Corey, E. J.; Wipke, W. T. Science 1969, 166, 178. (b) Corey, E. J.; Long, A. K.; Rubinstein, S. D. Science 1985, 228, 408. (c) Corey, E. J.; Howe, W. J.; Pensak, D. A. J. Am. Chem. Soc. 1974, 96, 7724. (d) Corey, E. J. Q. Rev. Chem. Soc. 1971, 25, 455. (e) Corey, E. J.; Wipke, W. T.; Cramer, R. D., III; Howe, W. J. J. Am. Chem. Soc. 1972, 94, 421. (f ) Corey, E. J.; Jorgensen, W. L. J. Am. Chem. Soc. 1976, 98, 189.

4 (a) Warren, S. Organic Synthesis: The Disconnection Approach; John Wiley & Sons: Chichester, 1982. (b) Warren S. Workbook for Organic Synthesis: The Disconnection Approach; John Wiley & Sons: Chichester, 1982. (c) Warren, S. Designing Organic Synthesis: A Programmed Introduction to the Synthon Approach; John Wiley: Chichester, 1978. Also see (d) Wyatt, P.; Warren, S. Organic Synthesis, Strategy and Control; John Wiley & Sons, Inc.: Hoboken, NJ, 2007. 5

(a) Hudlicky, T.; Reed, J. W. The Way of Synthesis. Evolution of Design and Methods for Natural Products; Wiley–VCH: Weinheim, 2007. (b) Fuhrhop, J.-H.; Li, G. Organic Synthesis. Concepts and Methods; Wiley–VCH: Weinheim, 2003.

6

(a) Smith, M. B. J. Chem. Educ. 1990, 67, 848. (b) Smith, M. B. J. Chem. Educ. 1996, 73, 304. Also see (c) D’Angelo, J.; Smith, M. B. Hybrid Retrosynthesis. Organic Synthesis using Reaxys and SciFinder; Elsevier: Amsterdam, 2015.

5

1.2 THE DISCONNECTION PROTOCOL

In a typical introductory organic chemistry course, there are two fundamental types of synthesis problems. In the first, both the starting material and the target are specified. In the second, only the final target is given and the synthetic chemist must deduce the starting material. This latter case usually poses a more difficult problem. The fundamentals of a retrosynthetic analysis will be illustrated with the first type of problem, and a discussion of the second type will be delayed until Chapter 8.

O 2,6-Dimethylheptan-3-one

2-Methylprop-1-ene

a

b O

4A

O

4B

In the disconnection of 2,6-dimethylheptan-3-one ) 2-methylprop-1-ene (isobutene), the alkene is the designated starting material. Since the starting material is specified, the disconnections are limited to those that will lead back to isobutene. Therefore, the four carbon atoms of the alkene must be “located” in 2,6-dimethylheptan-3-one, which will define the carbon-carbon bonds that must be disconnected for the retrosynthesis. The presence of two methyl groups in isobutene limits the carbon atoms that correlate with the target, but there are two different locations where the four carbons may be found in the target (see 4A and 4B). If the pattern shown in 4A is chosen, CdC bond a must be disconnected. If the pattern in 4B is chosen, then the CdC bond b must be disconnected. Disconnection of bond a in 4A leads to two fragments (disconnect products), 5 and 6. Similarly, disconnection of bond b in 4B leads to disconnect products 7 and 8. Both disconnections must be considered, but structures 5–8 are not real molecules, so the relative merit of each disconnection cannot be properly evaluated using these fragments. Only real molecules can be correlated with the viability of a real reaction. To assist in this process, two assumptions will be made: (1) The key carbon-carbon bonds will be formed by a small subset of reactions and (2) the bonds will be made by reactions involving polarized or ionic intermediates. a • •

O

O

5

6

b •

• O

O 7

8

The first assumption that key carbon-carbon bonds will be formed by a small subset of reactions is based on those reactions that are usually presented in a typical sophomore organic chemistry course. Examples are the carbon-carbon bond-forming reactions shown in Table 1.1.6a The second assumption that bonds will be made by reactions involving polarized or ionic intermediates is based on the observation that all reactions in Table 1.1 except entry 10 (the DielsAlder reaction, see Section 14.5.1) involve highly polarized or ionic intermediates. Based on these two assumptions, it is reasonable to conclude that the disconnection of bond a or b in 2,6dimethylheptan-3-one will lead to ionic or polarized precursors, and the requisite carbon-carbon bond will be formed using one of the reactions found in Table 1.1. In other words, fragments 5–8 should be polarized if possible, and then correlated with a polarized or ionic real molecule.

X

d a

a d

9

d

6

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

The concept of nucleophilic and electrophilic atoms in ionic and polarized intermediates is well known. Polarized bond notation (e.g., Cδ+dBrδ and CδdLiδ+) is commonly used in describing the polarization of such bonds, and such polarization commonly correlates with the reactivity of these bonds. Using these observations, Seebach7 used structure 9 to formalize a bond-polarization model. The sites marked d in 9 represent donor sites or nucleophilic atoms. The sites marked a are acceptor sites and correspond to electrophilic atoms. TABLE 1.1 Carbon-Carbon Bond-Forming Reactions 1. Cyanide 2. Alkyne anions (Acetylides) 3. Organometallics (not Cu or enolate anions) M = MgX, Li, ZnX,...

–

C LN C LC–R'

R—X + R—X +

(a) RM + R'—X

R—R'

O

(b) RM +

(Section 11.2)

R—CL N R—CLC–R'

–

R1

OH (Section 11.4.3.1)

R2

R

R2

(Section 11.3) (Section 11.4.2.1)

R1

O

OH

(Section 11.4.5)

(c) RM + 4. Organocuprates

R1 (a) R'—X + R2CuLi O (b) R'M + + R'—X (a)

R1

O– (b)

R' O

R 7. Friedel-Crafts acylation

Ar—H +

8. Friedel-Crafts alkylation

Ar—H + R—X

R' R3

O R2

10. Diels-Alder reaction

R1 O

O

+

X X R4

OH

R

R2

R'

(Section 13.4.1)

(Section 13.4.2) (Section 16.4.4)

Ar

Ar—X R3

R2

O

R O R'

(Section 12.3.2.2) (Section 13.3.1)

O

O

+ R

R1

R'

R O

9. Wittig reaction

R

O

R 6. Enolate anion condensation

O

R' R

–

(Section 12.3.1.1)

R—R'

O– 5. Enolate anion alkylation

R1

R

(Section 16.4.2) R4

+

(Section 12.5.1.1) PPh3

R1

R2

+

(Section 14.5.1)

Reprinted with permission from Smith, M.B. J. Chem. Educ. 1990, 67, 848. Copyright © 1990 American Chemical Society.

Bond polarization induced by the heteroatom extends down the carbon chain, due to the usual inductive effects that are a combination of through-space and through-bond inductive effects.8 The X moiety is typically a functional group or an electron-withdrawing atom (e.g., oxygen or a halogen). The electrophilic carbon adjacent to X is designated Ca (an acceptor atom) since proximity to the δ electronegative atom (X) induces the opposite polarity. Similarly, C2 is a donor atom (Cd), but less polarized than X (this carbon is further away from the electrons that induce the bond polarization), and C3 is a weak acceptor atom. As a practical matter, the effect is negligible beyond C4 and will be ignored. If this protocol is applied to disconnect fragments 5–8, either carbon at the point of disconnection can be assigned as a donor (d) or an acceptor (a), giving four different possibilities. As pointed out several times, disconnect fragments are not real molecules and another step is required before the best donor-acceptor pair can be chosen. Each fragment must

7 8

Seebach, D. Angew. Chem. Int. Ed. Engl. 1979, 18, 239.

(a) Baker, F. W.; Parish, R. C.; Stock, L. M. J. Am. Chem. Soc. 1967, 89, 5677. (b) Golden, R.; Stock, L.M. J. Am. Chem. Soc. 1966, 88, 5928. (c) Holtz, H. D.; Stock, L. M. J. Am. Chem. Soc. 1964, 86, 5188. (d) Branch, G. E. K.; Calvin, M. The Theory of Organic Chemistry; Prentice Hall: New York, NY, 1941, Chapter 6. (e) Ehrenson, S. Prog. Phys. Org. Chem. 1964, 2, 195. (f ) Roberts, J. D.; Carboni, C. A. J. Am. Chem. Soc. 1955, 77, 5554. (g) Clark, J.; Perrin, D. D. Q. Rev. Chem. Soc. 1964, 18, 295.

7

1.2 THE DISCONNECTION PROTOCOL

be correlated with a synthetic equivalent. Table 1.26a provides a list of common synthetic equivalents, leading to a definition of a synthetic equivalent as a molecular fragment that is equivalent to a real molecule by virtue of its chemical reactivity.9 A Cd site on an unfunctionalized carbon, for example, is equivalent to the Cδ of a Grignard reagent (see Section 11.4) and a Ca site on an unfunctionalized carbon is equivalent to the electrophilic carbon on an alkyl halide. Each synthetic equivalent in Table 1.2 is based on an ionic or highly polarized nucleophilic substitution or acyl addition reaction. Using Table 1.2, fragment 5 is the equivalent of a Grignard reagent (isobutylmagnesium bromide; see Section 11.4.1). The equivalent for the accompanying fragment 6 is the α-chloroketone, 1-chloro-3-methylbutan-2-one. Similarly, the synthetic equivalent of 7 is an alkyl halide, 1-bromo-2-methylpropane (another halide can be used), and the equivalent of 8 is an enolate anion, 3-methylbut-1-en-2-olate (see Section 13.3.1). In the actual syntheses, the Grignard reagent will react with the α-chloroketone to form the requisite bond, and the enolate anion will react with the alkyl halide. Both reactions will produce 2,6-dimethylheptan-3-one. Note that 1-bromo-2-methylpropane is derived from the 2-methylprop-1-ene starting material, but this transformation requires functional group modification. To accomplish this modification, the chemical relationship between the CdCdBr and C]C must be known. In other words, what reagents are required to transform an alkyl halide into an alkene, and vice versa.

TABLE 1.2 Common Synthetic Equivalents for Disconnect Products a

a

C

R

or R R

R

R a C CH2

a C CH2

and O

R

R

R

O–

R

L

d C C OH R R

R

O

R

H

R and

d C C

L

H

H C C

O

O

R

Cl C H OR

O

R C OR and R C OH L R R

C C R

R

(Enolate anion) R

R

R

L

a C C OH

R R

R C C O R

d

R S

L

R

R

R C

R

R d H and R C C O

d

O

R

R Li

R R C MgX or

L

R R C Li or

R

R a

S (Acyl anion equivalent)

R R C CuLi or

R R C PR3

R 2

R

R

R C

R C C

R

R

L

(X = Cl, Br, I, OTs, OMs, OTf, ...) (See abbreviations page)

R C X

R

R a

R C

O

L

R C R

(For the Wittig reaction)

R Reprinted with permission from Smith, M.B. J. Chem. Educ. 1990, 67, 848. Copyright © 1990 American Chemical Society.

9

See Ref. 1, p 30.

8

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

Fig. 1.26a provides a functional group reaction diagram to interconvert one functional group into another by known chemical pathways. Since an alkene and a halide are chemically related by the reactions indicated, either could be used in the disconnect product since one can easily be converted to the other at the appropriate time. C

C

CR2

O

t

u

x

C

k

C C

X

i

j

C

s

O

a b

m

l

C

n

CH2 CH2 –

o d

r

C

C

p

q

C

h

OH c

–COOH

X

f

g

C

e

C

O –C

w

v

X

C O a. LiAlH4 or NaBH4 or H2/cat c. NaOH/H2O or i. PPh3, RCO2 –, MeO2C-N N-CO2Me (the Mitsunobu reaction) e. HX g. i. Br2 ii. NaNH2 i. NaNH2/NH3 k. H3O+ m. N2H4/KOH or Zn(Hg)/HCl o. LiAlH4 or H2/cat q. H2/cat s. Excess H2 /cat u. Me2S+ CH2 – w. i. B2H6 ii. NaOH/H2O2 or i. Hg(OAc)2 ii. NaBH4

C

b. CrO3 d. PCl5 or SOCl2 (X = Cl); PBr3 (X = Br); i. NaH ii. –OH iii. H 3O+ RX (X = OR); i. PBr 3 or RSO 2Cl ii. K-phthalimide; hydrolysis or NaN3; H2/cat (X = NH2) f. NaOH/H2O or RO–/ROH h. H2/cat or Na/NH3 /EtOH j. HX l. PCl5 n. i. RSO2Cl ii. LiAlH4 (only one carbon reacts) p. X2/h or Heat r. 2 H2/cat t. Ph3P=CR2 v. RCO3H x. i. O3 ii. Me2S or H2O2

FIG. 1.2 Functional group exchange reaction wheel. A visual reminder that functional groups can be interconverted. Reprinted with permission from Smith, M.B. J. Chem. Educ. 1990, 67, 848. Copyright © 1990 American Chemical Society.

In most undergraduate organic chemistry textbooks, the functional group approach used means that groups with related chemical properties may be presented in different semesters. Although alkenes and carbonyl both react as a base in the presence of an acid, for example, these groups are presented at different places and at different times during a typical course. This approach can make the relationship of one functional group to another difficult to see. Fig. 1.26a is provided as a visual reminder of these relationships, which are often essential for completion of a total synthesis. In the context of this chapter, Fig. 1.2 illustrates chemical relationships that can be used to understand FGI reactions. In the example at hand, the relationship between a bromide and an alkene is apparent from the table (CdCdBr ) C]C). It is also possible to write a functional group exchange based on the alkene as C]C ) CdCdBr. a •

•d

MgBr

Isobutylmagnesium bromide 5 and •

Cl

O O 1-Chloro-3-methylbutan-2-one 6

d •

a Br

1-Bromo-2-methylpropane 7

O– O 3-Methylbut-1-en-2-olate 8

It is important to emphasize that Fig. 1.2 is not intended as a device to memorize specific reactions, but rather to introduce the idea of using various functional groups in a synthesis by recognizing their synthetic relationships. Returning to the disconnection of 2,6-dimethylheptan-3-one, synthetic sequences for both disconnections a and b

9

1.3 BOND PROXIMITY AND IMPLICATIONS FOR CHEMICAL REACTIONS

are shown in Fig. 1.3. The best route probably involves conversion of 2-methylprop-1-ene to 1-bromomethylpropane and subsequent reaction with the enolate anion, which is derived from 3-methylbutan-2-one. This synthesis is considered to be better than the second one based on disconnection b since it involves simpler reagents and is shorter (fewer chemical steps). The synthetic chemist must decide which is best, however, based upon his/her own experience and goals. The best route is a subjective judgment, although it usually makes more sense to follow a short and simple route rather than a longer and more complicated one.

HBr

Br

t-BuOOt-Bu Bond a

Br O

O−

LiN(i–Pr)2, THF

O

1. Li°

−78 °C

2. CuI Bond b

or

CuLi

2 1. B2H6

OH CrO3

2. NaOH H 2O 2

FIG. 1.3

aq H+

CO2H

O

SOCl2

Cl

Syntheses of 2,6-dimethylheptan-3-one.

1.3 BOND PROXIMITY AND IMPLICATIONS FOR CHEMICAL REACTIONS How is the synthetic analysis presented in Section 1.2 related to a study of chemical reactions? In all of these disconnections, the disconnect products were converted to actual compounds in order to predict which chemical reactions might work. In all cases, the disconnected bonds were either the one directly attached to the functional group (CdX) or the one next to it (CdCdX). These observations can be used to understand what reaction types are important for making CdC bonds. When a functional group (X) is attached to a carbon atom, the bond polarity extends to about the third bond, as indicated in 9. If the assumption is made that most reactions involve highly polarized species, then the three important bonds are the CdX bond (called the α-bond), the adjacent one (the β-bond), and the γ-bond, as seen in 10. If the α-bond is disconnected, fragments X and 11 are obtained where a nucleophilic X group attacks the electrophilic carbon. A simple example is the reaction of azide ion (N3) with 1-iodopropane to yield 1-azidopropane (see Section 3.2.1). If the β-bond is disconnected, fragments X and 12 are obtained where the natural bond polarization suggests a reaction in which a nucleophilic species attacks the electropositive carbon of the CdX unit. Assume the functional group is C]O, where addition of a Grignard reagent to an aldehyde fits this description (see Section 11.4.3). Finally, disconnection of bond γ leads to fragments 13, and conjugate addition of an organocuprate to a α,β-unsaturated ketone fits this type of disconnection, where X is C]O; (see Sections 12.3.1 and 12.3.2.2). (X O, N, S, C O, C N, etc.) X

C

C 10

C Disconnect

Disconnect Disconnect

X X

C

C 11

C

X C

C

C 12

C

C

C

13

The point of this short section is to show how disconnection of molecules that contain polarized functional groups can be correlated with recognition of the natural bond polarization characteristics. This analysis is based on a variety of common reactions, which is important not only for planning a synthesis, but for understanding how chemical reactions work. This general theme will be elaborated in many chapters to follow.

10

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

1.4 STEREOCHEMISTRY The concept of chirality and absolute configuration is introduced early in undergraduate organic chemistry courses. For that reason, this section is intended as a review and it is assumed the reader has some familiarity with the concepts. This section will discuss stereoisomers. The discussion will begin with 2-chlorobutane and 1-chlorobutane, which have the same empirical formula, but are clearly different molecules. They are isomers: two or more molecules that have the same empirical formula. They are different molecules! The term isomer does not adequately characterize the relationship of the atoms and groups in 2-chlorobutane and 1-chlorobutane. The term used to define the relationship between these two alkyl chlorides is regioisomer: two or more molecules with the same empirical formula, but with a different attachment of the atoms (different connectivity). Another type of isomerism occurs when two molecules have the same empirical formula and also have the same connectivity of atoms, but are different molecules. Such molecules differ only in the relative spatial positions of the atoms, and are called stereoisomers.10 The following sections will discuss different types of stereoisomers and their characteristic properties. CH3 CH3CH2

Cl

Cl

*C

CH3CH2CH2

C*

H H

H 2-Chlorobutane

1-Chlorobutane

1.4.1 Absolute Configuration in Chiral Nonracemic Molecules When a carbon atom is bound to four different atoms or groups in a tetrahedral arrangement, that carbon is said to be stereogenic or chiral. Other atoms may be stereogenic, but the number and type of attached atoms or groups will vary with the valence of the atom. In this chapter, the focus will be exclusively on carbon. The C* in 2-chlorobutane is a stereogenic center. The C* in 1-chlorobutane is not stereogenic, since that carbon has two identical atoms (H) attached to it. Why is identification of C* in 2-chlorobutane as a stereogenic center important? The answer appears when the mirror image of 2-chlorobutane is drawn (see Fig. 1.4). Note that the chlorine atom in A is reflected into the chlorine atom in A0 . Further, a hydrogen atom reflects to a hydrogen atom, a methyl (Me,CH3) to a methyl, and an ethyl (Et,C2H5) to an ethyl, but when an attempt is made to superimpose A and A0 one on the other, they are found to be two different molecules. 2-Chlorobutanes A and A0 are not superimposable by any rotation of bonds, or positioning of the molecules. In other words, they are different molecules. More precisely, the molecules have no symmetry (they are asymmetric)11 as can be seen in the molecular model, which is an attempt to superimpose A and A0 . It is clear that while the Me and Et groups superimpose, the Cl and H atoms do not. This statement can be confirmed with any manual or electronic model, by making models of both enantiomers and trying to make all atoms and groups superimpose. They do not! Merging the two structures shows that it is not possible to make the Cl and H atoms superimpose

A A 2-Chlorobutane

CH3

A Cl A

H3CH2C

Cl H

CH3

H3CH2C H

Comparing A and A shows that Et reflects to Et, C to C, and Me to Me but Cl does not reflect to Cl and H does not reflect to H

FIG. 1.4

The superimposed mirror images

Mirror images of 2-chlorobutane and an attempt to superimpose mirror images.

10

For a general discussion of stereoisomers, see Eliel, E. L.; Wilen, S. H.; Mander, L. N. Stereochemistry of Organic Compounds; Wiley-Interscience: New York, NY, 1994; pp 49–70.

11

For a general discussion of symmetry, symmetry operators, and their importance in organic chemistry, see Ref. 10, pp 71–99.

11

1.4 STEREOCHEMISTRY

While A and A0 do not superimpose, they are clearly isomers, and they are closely related since they are mirror images. The two 2-chlorobutanes are not regioisomers because the connectivity of atoms is the same. The position of the atoms and groups are different, so these two isomers are stereoisomers (molecules that differ in their spatial arrangement of atoms but have the same points of attachment).10 The two 2-chlorobutanes A and A0 are further distinguished as a special type of stereoisomer called enantiomers, that is, stereoisomers that are nonsuperimposable mirror images. Therefore, the term enantiomer specifies the relationship of the two stereoisomers. If a structure and its mirror image are superimposable by rotation or any motion other than bond making and breaking, then they are identical (a single molecule and not enantiomers; 1-chlorobutane is an example). Make a model of 1-chlorobutane and its mirror image and demonstrate that they are superimposable. When molecules contain more than one stereogenic center, a new type of stereoisomer known as a diastereomer will be present, but this discussion will wait until Section 1.4.2. 2-Chlorobutanes A and A0 are different molecules, but they both have the name 2-chorobutane. Clearly, each must have a unique identifier that can be used with that name, allowing one enantiomer to be distinguished from the other. The method used is the (R/S) system, which employs a set of priority rules developed by Cahn, Ingold, and Prelog, the so-called CIP selection rules.12 In this system, each atom attached to the stereogenic center is assigned a priority (a–d, where a is the highest priority atom and d is the lowest priority atom). Once a priority (a–d) is assigned to each group or atom attached to the stereogenic center, a protocol is required to utilize differences in the priorities. This protocol is shown in Fig. 1.5A, where the viewer positions the molecule in such a way that it is possible to sight down the C*dd bond, with the d atom projected away from the viewer, as shown. In fact, the atom with assigned priorities must be turned so that the d group is projected behind the plane of this page, with the viewer positioned in front of the page. This action leads to the representation shown in Fig. 1.5B, which can be correlated with a steering wheel, where an imaginary line a ! b ! c is drawn to represent the wheel. This steering-wheel model is then inspected to determine if the a ! b ! c line is projected around the wheel following a clockwise (right) or a counterclockwise (left) path. a

a (S)

(R)

d

d c

(A)

b c

b a

a (R)

(S) d b

(B)

d c

Steering-wheel model

c

b

Steering-wheel model

FIG. 1.5 The steering-wheel model. (A) The model applied to (S) and (R) absolute configurations. (B) Determining (S) and (R) on a “steering wheel.” In part, reprinted with permission from Cahn, R.S. J. Chem. Educ. 1964, 41, 116. Copyright © 1964 American Chemical Society.

If the a ! b ! c sequence is clockwise, the enantiomer is assigned the (R) configuration. Conversely, if it is counterclockwise, that enantiomer is assigned the (S) configuration. To reiterate, this model requires that the molecule must be rotated so that d is projected away from the observer prior to sighting down the C*dd bond. Determining the relative priority (a–d) is accomplished by a series of rules just mentioned, the so-called CIP selection rules.12–14

12

(a) Prelog, V.; Helmchen, G. Angew. Chem. Int. Ed. 1982, 21, 567. (b) Cahn, R. S. J. Chem. Educ. 1964, 41, 116 (see p 508). (c) Cahn, R. S.; Ingold, C.; Prelog, V. Angew. Chem. Int. Ed. Engl. 1966, 5, 385. (d) Cahn, R. S.; Ingold, C. K. J. Chem. Soc. (London) 1951, 612. (e) Cahn, R. S.; Ingold, C. K.; Prelog, V. Experientia 1956, 12, 81.

13 14

Ref. 12, pp 101–112.

(a) IUPAC Commission on Nomenclature of Organic Chemistry Pure Appl. Chem. 1974, 45, 13. Also see (b) Hirschmann, H. Hanson, K. R. Tetrahedron 1974, 30, 3649.

12

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

1. Atoms with a higher atomic number precede atoms of lower atomic number O>N>C>H 2. Isotopes with a higher mass number precede isotopes of lower mass number This rule applies to isotopes (e.g., tritium, deuterium, and protium) 3H > 2H > 1H. a

Br

a

b

Cl Me

d

H

a

( R)

d

c

d c

b

c

b

1-Bromo-1-chloroethane

These two rules are illustrated by the first example, 1-bromo-1-chloroethane, where the priorities based on rules 1 and 2 lead to Br > Cl > Me > H and the a ! b ! c rotation is clockwise, giving an (R) configuration. Similarly, in ethan-1-d-1-ol, the priority is O > C > D > H and the a ! b ! c sequence is also clockwise for an (R) configuration, for (1R)-bromo-1-chloroethane. Note the use of a Fischer projection for ethan-1-d-(1R)-ol, where the horizontal lines project out of the paper toward you, and the vertical line projects behind the paper. Fischer projections are not used as much today.

H

a

a

OH d

D

c b

Me

(R) d b

c

Ethan-1-d-1-ol

A problem arises when two of the atoms attached to C* cannot be prioritized by rules 1 and 2. An example is 1-aminohexan-3-ol, where the analysis based on rules 1 and 2 leads to the following priority for atoms attached to C*: O > C]C > H. Although the groups are different, the same atom (C) is attached to C*, so there is no difference in priority: C*: O (a) > CCHH]CCHH > H (d). A new rule is required.

*

NH 2

HO H 1-Aminohexan-3-ol

c

b

b

(S)

d a

d

c

a

3. When there are two or more identical atoms, proceed atom by atom down the highest priority chain to the first (or next) point of difference and apply rules 1 and 2 to determine the priority. It is necessary to proceed further down each carbon chain to find a point of difference, using rule 3. Comparing these two chains leads to a point of different at the second carbon removed from C*, yielding [C*dCCHHdCNHH ¼ b] and [C*dCCHHdCHCHH ¼ c]. At this point of difference, a–d can be assigned to distinguish these groups, for an (S) configuration so it is 1-aminohexan-(3S)-ol. In another example, 1-bromo-7-methyltetradecan-7-ol shows the hydroxyl oxyen and the methyl group to be (a) and (d), respectively, but to determine (b) and (c) requires inspection of each chain to the sixth carbon, where there is a point of difference. This priority establishes the (S) configuration, for 1-bromo-7-methyltetradecan-(7S)-ol. OðaÞ > C  C  C

C*dCHHH ¼ ðdÞ

C*dCCHHdCCHHdCCHHdCCHHdCCHHdCBrHH ¼ ðbÞ C*dCCHHdCCHHdCCHHdCCHHdCCHHdCCHH ¼ ðcÞ

13

1.4 STEREOCHEMISTRY

Me

HO Br

a

*

b

a

d

(S)

d

c

b

1-Bromo-7-methyltetradecan-7-ol

c

In 3,5-dimethoxy-2,4-dimethylheptane, a different problem arises for assignment of priorities. The methyl carbon atom (CC) and the hydrogen atom (CH) are assigned the priorities (c) and (d) relative to the two CO atoms. To determine the a and b priorities, the higher priority atoms connected to the oxygen atoms must be used, with the recognition that oxygen has a higher priority than carbon or hydrogen. Using rule 3, the two pathways that contain OCH are examined for a point of difference. Although the highest priority chains are along the two CdOdC pathways (the oxygen chains), both groups are OCdCHHH. In other words, they are identical and there is no point of difference, so they cannot use them to establish the priority. In such a case, the lower priority carbon chain must be used (in this case, the i-Pr and Et groups rather than the methoxy groups), giving C*dCOCHdCCCH for (a) and C*dCOCHdCCHH for (b), leading to 3,5-dimethoxy-2,(4S)-dimethylheptane. MeO

Me

c

c

OMe a

b

(S)

d

d

H

b

a

3,5-Dimethoxy-2,4-dimethylheptane

A different problem presents itself in 4-methylhex-5-ene-1,3-diol, where the priorities of the atoms attached to C* at the first point of difference are O > C  C > H, but there are different numbers of carbon atoms at that point. Remember that priorities are assigned to the atoms, not the group. Rules 1 and 2 are not sufficient to distinguish the two carbon atoms, and rule 3 does not resolve the problem, but there are a different number of atoms at the point of difference. Another rule is required. c HO

b

b

(S)

d

*

HO H 4-Methylhex-5-ene-1,3-diol

a

d

a

c

4. When the highest priority atoms at the first point of difference are identical, compare the number of priority atoms at that point to distinguish them. When identical atoms are attached, another priority scheme is required. Based on carbon substituents attached to C*, the priorities by this rule are CCCC > CCCH > CCHH > CHHH and priorities based on oxygen and carbon substituents attached to C* are: COOO > COOC > COCC > CCCC Examination of 4-methylhex-5-ene-1,3-diol reveals that the two carbon atoms attached to C* are CCHH and CCCH. The carbon atoms have equal priority, but using rule 4, two carbon atoms have a higher priority than one. Therefore, the secondary carbon has a higher priority (CCCH) than the primary carbon (CCHH), so the absolute configuration is (S), for 4-methylhex-5-ene-1,(3S)-diol. O O

Cl

*

b

( R)

d a

c

3-(2-Chloroethoxy)-3-isopropoxyheptane

Another example is the ketal, 3-(2-chloroethoxy)-3-isopropoxyheptane, which shows a priority of O]O > C]C. Rule 4 is required to prioritize the two “O” groups and the two “C” groups, but there is another problem. As one proceeds down the carbon chain, the ethyl (CCHH) and butyl (CCHH) groups have the same atoms attached directly to C*. To determine the proper priorities, follow each chain to the next carbon of both the ethyl group and the butyl group to

14

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

find a point of difference. In this case, the point of difference is CHHH for ethyl versus CCHH for butyl. Similarly, both of the oxygen atoms attached to C* show an identical OCHH, but a point of difference is reached by following the carbon chains attached to each oxygen atom, OdCCHH for O-chloroethyl versus OdCCCH for Od i-Pr. The priority analysis based on finding a point of difference when the atoms attached to the stereogenic carbon are identical leads to [C*dOdCCHH ¼ b], [C*dOdCCCH ¼ a], and [C*dCCHHdCHHH ¼ d], [C*dCCHHdCCHH ¼ c] so the absolute configuration is (R) for (3R)-(2-chloroethoxy)-3-isopropoxyheptane. A common misconception for rules 3 and 4 arises from an incorrect interpretation. In 4,5,5-trimethylnonan-1-ol, the methyl carbon and the hydrogen atom are assigned (c) and (d), respectively. An analysis by rule 3 predicts CCCH > CCHH. The CCHH chain (dCH2CH2CH2OH) contains an oxygen atom. The CH(Me)CH2CH2CH2CH3 (CCCH) chain does not contain a heteroatom. The rules dictate that the priority be determined at the first point of difference, which places the focus on the atom and not the group. The presence of the oxygen is, therefore, irrelevant and the CCCH chain will have priority over the CCHH chain because the first point of difference is reached before oxygen is encountered. The absolute configuration is (S) for (4S),5,5-trimethylnonan-1-ol. Me Me

c

Me OH

* H

c

CHHH

CCHC

dH

(S)

d

CHHC

a

b

b

a

4,5,5-Trimethylnonan-1-ol

Cyclic compounds are treated by the same rules, but determining different groups in a ring requires a different perspective. Each arm of the ring is viewed as a different group and then compared atom-by-atom. 3-Chlorocycloheptanone, for example, has Cl(a) > C  C > H(d). The arm that includes the carbonyl carbon is C*dCCHHdCOO°C (b), in contrast with the arm that does not include the carbonyl carbon is C*dCCHHdCCHH (c). The configuration is (R), so the name is (3R)-chlorocycloheptanone. Cl H

a

*

(R)

a

d

c

d b

b

c

O 3-Chlorocycloheptanone

In 14 [4-chloro-2-(1-(4-chloro-2-hydroxycyclohexyl)ethyl)cyclohexan-1-ol], the methyl and hydrogen are again (c) and (d), but the rings pose a problem, one arm of the ring points to chlorine and one to hydroxyl. Since the secondary hydroxy carbon is closest to C*, the bottom arm of each ring is followed rather than the top arm. By this route, the left ring is highest priority: (C*dCCCHdCCOHdCCHHdCClCH vs. C*dCCCHdCCOHdCCHHdCCHH), and the absolute configuration is (S). Therefore, the name is 4-chloro-2-((1S)-(4-chloro-2-hydroxycyclohexyl)ethyl)cyclohexan-1-ol.

Cl

Me * H OH HO 14

Cl

c

c a

b d

(S)

d a

b

(3R)-Chlorocycloheptanone contained a carbonyl, and the analysis treated that carbon as COOC. Where did the “second oxygen” come from? Multiply bonded atoms such as this are found in alkenes, alkynes, carbonyls, and nitriles, and they require yet another rule. 5. If an atom is attached to another by a multiple bond (double or triple) both atoms are considered to be duplicated. The duplicated atom (X°) is considered to have a valence of zero and has a lower priority than a real atom (X), if that is the only point of difference.

15

1.4 STEREOCHEMISTRY

This rule will convert the following common functional groups into their phantom counterparts for consideration in the priority scheme. C*dCH]CH2^C*dCHCC° dCC°HH

C*dCð]OÞdC^C*dCOO°CdOCC°

C*dC ^CdMe ^C*dCCC°C°dCCC°CdCHHH

C*dCO2 H ^C*dCOO°OdOCC°

C*dC ^N ^ C*dCNN°N°dNCC°C° Generally, both X° and X are simply treated as real atoms. If the only choice is between a duplicated atom X° and a real atom (X), however, the duplicated atom has a lower priority (in this case, all other rules have failed to establish the priorities). With the tert-butyl fragment, for example, proceeding down the chain to the first point of difference leads to tert-butyl with a lower priority than the alkynyl fragment: C*dC ^CdMe vs: C*dCMe3

C*dCCC°CdCCC°C > C*dCCCCdCHHH

The C and C° are taken as equal in this analysis and rule 4 determines the priority. We compare CCC°C with CHHH, which gives the alkyl fragment a higher priority. Another example of this rule is the analysis of 6-methylhept-1-en-4ol, which shows O(a) > C  C > H(d). The alkenyl arm is C*dCCHHdCCHC° and the alkyl arm is C*dCCHHdCCCH. In this case, the rules are unable to distinguish the priority of these two fragments. One proceeds to the next atom (CC°HH), which is compared with CHHH (]CH2 vs. CH3). The former atom is higher in priority and becomes group (b). The absolute configuration is (S), for 6-methylhept-1-en-(4S)-ol HO H

a

* b

a

d

(S)

d c

b

c

6-Methylhept-1-en-4-ol

The International Union of Pure and Applied Chemistry (IUPAC) Commission assembled a list of common substituents, sorted by the increasing order of sequence rule preference (number 76, iodo, is the highest priority and number 1, hydrogen, is the lowest priority in that table).14a With this table two or more substituents can be evaluated in order to determine their relative priority using the CIP selection rules.15 Higher numbers have a higher priority. Iodo (76) has a higher priority than bromo (75) and hydrogen (1) has the lowest priority. Similarly, allyl (10) has a higher priority than propyl (4), acetylenyl (21, C^CH) has a higher priority than tert-butyl (19), and hydroxyl (57) has a higher priority than amino (43). In principle, a nitrogen atom can be stereogenic if there are three different alkyl groups on nitrogen, where the fourth group is the lone pair of electrons on nitrogen. The nitrogen may be a stereogenic center, but rapid inversion at nitrogen leads to the mirror image (see N-ethyl-N-methylpropan-2-amine) being present at the same time. Because of this facile racemization such compounds are not optically active unless this fluxional inversion can be inhibited. Approximately 2  1011 inversions occur each second for ammonia.16 The energy barrier for this inversion is somewhat higher in amines (see Table 1.3),17,18 due the presence of alkyl groups on nitrogen that are bulkier than the hydrogen atoms in ammonia. Nonetheless, alkyl amines undergo rapid inversion.19 The magnitude of the energy barrier to inversion in amines is determined by the inter-group bond angle (α, C1dN-C2), and the corresponding bending force constant.17 Table 1.317,18 shows values of the parameter α for ammonia, phosphine, trimethylamine and trimethylphosphine, along with an estimate for the energy barrier to inversion. Older estimates of these energy barriers are also included.18 The inversion at nitrogen in alkyl amines cannot be stopped at the reaction temperatures usually employed in organic chemical reactions. When inversion is impossible due to structural features of a molecule such as those found in bicyclic and polycyclic amines, the nitrogen is stereogenic and enantiomers are observed. 15

Cahn, R. S. J. Chem. Educ. 1964, 41, 116.

16

Smith, M. B. March’s Advanced Organic Chemistry, 7th ed.; John Wiley & Sons, Inc.: Hoboken, NJ, 2013, pp 127–128.

17

Koeppl, G. W.; Sagatys, D. S.; Krishnamurthy, G. S.; Miller, S. I. J. Am. Chem. Soc. 1967, 89, 3396.

18

Kincaid, J. F.; Henriques, F. C., Jr. J. Am. Chem. Soc. 1940, 62, 1474.

19

(a) Mislow, K. Pure Appl. Chem. 1968, 25, 549. (b) Rauk, A.; Allen, L. C.; Mislow, K. Angew. Chem. Int. Ed. Engl. 1970, 9, 400. (c) Lambert, J. B. Top. Stereochem. 1971, 6, 19.

16

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

Me

Me Et

••

••

N

Me2HC

N

CHMe2

Et N-Ethyl-N-methylpropan-2-amine

The energy barrier for inversion is low for second-row elements (C, O, N), and rapid inversion occurs. With elements in the third row (e.g., P and S), however, inversion is slow at ambient temperature and those molecules may exist as enantiomers. Methylpropylphenylphosphine is configurationally stable at 25°C, although it rapidly inverts at 130°C.20b At 130°C, this barrier was measured to be 30.7 kcal (128.4 kJ) mol1,20a for a rate of inversion of 3.34  105 s1.21 At 130°C, cyclohexyl(methyl)(propyl)phosphane showed a rate of inversion of 0.043  105 s1 and the rate for methyl(prop-1-enyl)phenylphosphine was 1.44  105 s1 [Ea ¼ 32.2 kcal (134.6 kJ) mol1].21 TABLE 1.3 Energy Barrier to Inversion of Amines and Phosphines a

Compound NH3

106.77°

PH3

109.0°

PMe3

100°

b

Reference 17

Einversion(kJ mol–1)

5.58a

23.2a

11b

46.0b

27a

112.9a

47b

196.5b

7.46a

31.2a

15b

62.7b

20.4a

85.3a

57b

283.3b

93.3°

NMe3

a

Einversion(kcal mol–1)

Reference 18

Reprinted with permission from Koeppl, G.W.; Sagatys, D.S.; Krishnamurthy, G.S.; Miller, S.I. J. Am. Chem. Soc. 1967, 89, 3396. Copyright © 1967 American Chemical Society, and from Kincaid, J.F.; Henriques, F. C., Jr. J. Am. Chem. Soc. 1940, 62, 1474. Copyright © 1940 American Chemical Society.

d

a

(R) •• b

d c

P

CCHH

HHHC c

b

Me

CCCC° a

P

C3H7

Methyl(phenyl)(propyl)phosphane

••

Me

Ph

••

C3H7 C6H11

P Me Ph

Cyclohexyl(methyl)(propyl)phosphane Methyl(prop-1-enyl)phenylphosphine

If a nitrogen atom is a stereogenic center, it must be assigned a configuration by the CIP selection rules. An example is the bridgehead nitrogen in ()-castoramine (15), shown in Fig. 1.6,22 which is incapable of inversion at nitrogen because of its rigid bicyclic structure.23 Castoramine is also drawn such that the trans-orientation of the electron pair is indicated, which is preferred to the cis-orientation by  2.4 kcal (10.0 kJ) mol1.23 This barrier effectively locks the molecule into the trans-conformation. In this case, the electron pair must be considered a group, but the first five rules do not allow it to be assigned a priority. A sixth rule is necessary. 20

(a) Horner, L.; Winkler, H.; Rapp, A.; Mentrup, A.; Hoffmann, H.; Beck, P. Tetrahedron Lett. 1961, 161. (b) Horner, L.; Winkler, H. Tetrahedron Lett. 1964, 461. 21

Baechler, R. D.; Mislow, K. J. Am. Chem. Soc. 1970, 92, 3090.

22

LaLonde, R. T.; Muhammad, N.; Wong, C. F.; Sturiale, E. R. J. Org. Chem. 1980, 45, 3664.

23

(a) Aaron H. S.; Ferguson, C. P. Tetrahedron Lett. 1968, 6191. (b) Aaron H. S. Chem. Ind. (London) 1965, 1338.

17

1.4 STEREOCHEMISTRY

6. Lone electron pairs receive an atomic number of zero and are assigned the lowest priority. The duplicated C, O, and N atoms from rule 5 have a higher priority rather than an electron pair. For ()-castoramine (15), rule 6 sets the priority such that the stereogenic nitrogen center is (R). Just as a nitrogen atom can be stereogenic, a phosphorus atom in a phosphine (e.g., ethylpropylphenylphosphine) can be considered chiral at temperatures up to 130°C, as noted in Table 1.3. The lone-pair electrons have the lowest priority by rule 6, and the absolute configuration is (R). The six rules for determining the priority of groups on stereogenic centers can be applied to any molecule. Natural products are particularly interesting, since they are the targets of many synthetic endeavors. As noted above, the absolute and relative configuration of such molecules must be known if they become synthetic targets, and must be factored into the retrosynthetic plan. The Amaryllidaceae alkaloid crinine (16) has four stereogenic centers, including the ringfused nitrogen atom (see Fig. 1.6). The absolute stereochemistry for each stereogenic center, as the molecule is drawn, is indicated where each is treated as if it were an individual and isolated atom. In other words, focus on the carbon bearing the OH group, treat it as an individual tetrahedral carbon with four attached groups (independent of the fact they are all part of one molecule) and then assign priorities. Once done, that carbon atom is determined to have an absolute configuration of (R). This fundamental approach is used for all the stereogenic centers, and can be applied to any molecule, regardless of the complexity. In some cases, assigning priorities can be tedious and occasionally confusing, but there are additional rules that can be applied to cover all contingencies. Such rules will not be discussed further as the six presented in this section are sufficient for the purposes of this book.

b

CCC°

CJCCHC

N

a

(R)

d

••

CH3 H

CCCHJCCCH

CCHHJCCCH c

8 9

7 10

6

N1

H 5

H3C H

2

3

N

H

••

b

H

4

OH

••

OH O

d a

O

c

(R)

15 b

b

(S)

C

CCCH CCHH CCHH CCC

a

CCCH

a

H CCCH O

H

CCHH

d

c

d

NCC

OH

c

O a

CHHC CCCC

CCCH CCHH N CCHH ••

c

H O (S)

16

(R) N ••

c b

CCCC CCHH

d

H

d

b

NCC a

FIG. 1.6

Determination of stereogenic centers in 15 and 16.

The rules given above allow the (R) or (S) configuration to be assigned for relatively simple molecules, and both 15 and 16 are included as relatively simple molecules. At first glance, applying these rules to complex molecules appears to be daunting. Compound 17, for example, is more complex, and assigning all stereogenic centers by the manual methods presented in this section is at best time-consuming, at worst difficult, and prone to error. Modern computational methods, with algorithms to assess the absolute configuration, make such assignments rapid and rather easy. One simply draws the structure with the correct stereochemical relationships in the model, and the program makes the correct assignment. Indeed, the computer-generated assignments are shown in 18–20 using Spartan.

18

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

CH2CH2Cl MeO

O H CH2NH2

(R)

MeO

(S)

(R)

CO2H

O

(R)

Cl (R)

(R)(R) (R)(R)

(Z)

(S)

HO

(E) (S)

(R)

(S)

(R)

(S)(S)(R)

O O

O

HO

O HO Me

(R)

OH

(R) (S)

(R)

O

Me

(R)

(R)

HO

H

H

(S)

HO Me O

Me

(Z) (E)

(S) (R) (R)

O H

O

H

(R)

H

O

(S)

H OH

H

H H O

O

H O

O H

H

(R) (S) (R)

(R) (S) (S) (R)

(R)

(S) (S) (R)

(R) (S)

N

O

N

(R) (S)

(R) (R)

HO

OH

H

Me

Me

(Z)

(S)

(E)

Me

19

HO

(E)

18

MeO H O O (S) (R) Me O

H

O

O

(R)

(R)

(S)

O

H

(R)

(R)

(E)

(R) (E)

(R)

O

N (R)

O Me

O

O

(R)

17

(R)

Cl

(Z)

Me

(R)

(R) (R)

(R)

(S)

(S) (R) (R)

(S)

Me

(S) (R) (R)

(R)

COOH

(S)

OHC

(S)

(S) (S) (S) (S)

OH

HO

(S)

H

(R) (S) (R) (S) (R)

O H

O H H

H

O (S) (S) (S) (S) (R)

H O H (S) Me

OH H O O (R) (S) (S)

Me

OH

20

(Z)

(Z)

Try assigning all stereocenters using the rules just discussed. Compound 18 is the oligocyclopropane FR-900848,24 and 19 is the macrolide toxin pectenotoxin 2.25 A similar analysis of the absolute configuration is shown for all stereogenic centers in ciguatoxin, 20.26

1.4.2 Diastereomers When there is more than one stereogenic center in a molecule, the maximum number of possible stereoisomers is predicted by the 2n rule (for n stereogenic centers there is a maximum of 2n stereoisomers). Therefore, two stereogenic centers in a molecule (e.g., 2-bromo-2-phenylpentan-3-ol) lead to four possible stereoisomers. (2S,3S)-2Bromo-2-phenylpentan-3-ol and its mirror image, (2R,3R)-2-bromo-2-phenylpentan-3-ol, are stereoisomers and they are enantiomers. The stereoisomers [(2S,3R)-2-bromo-2-phenylpentan-3-ol] and its mirror image [(2R,3S)-2bromo-2-phenylpentan-3-ol] are also enantiomers. Inspection of (2S,3S)-2-bromo-2-phenylpentan-3-ol and (2S,3R)-2bromo-2-phenylpentan-3-ol shows that they are also isomers, and stereoisomers, but they are not mirror images nor are they superimposable. These two stereoisomers are diastereomers. Indeed, the definition of a diastereomer is a 24

Yoshida, M.; Ezaki, M.; Hashimoto, M.; Yamashita, M.; Shigematsu, N.; Okuhara, M.; Kohsaka, M.; Horikoshi, K. J. Antibiot. 1990, 43, 748.

25

(a) Yasumoto, T.; Murata, M.; Oshima, Y.; Sano, G. K.; Matsumoto, J. Tetrahedron 1985, 41, 1019. (b) Murata, M.; Sano, M.; Iwashita, T.; Naoki, H.; Yasumoto, T. Agric. Biol. Chem. 1986, 50, 2693. (c) Sasaki, K.; Wright, J. L. C.; Yasumoto, T. J. Org. Chem. 1998, 63, 2475. (d) Suzuki, T.; Beuzenberg, V.; Mackenzie, L.; Quilliam, M. A. J. Chromatogr. A 2003, 992, 141. (e) Miles, C. O.; Wilkins, A. L.; Samdal, I. A.; Sandvik, M.; Petersen, D.; Quilliam, M. A.; Naustvoll, L. J.; Rundberget, T.; Torgesen, T.; Hovgaard, P.; Jensen, D. J.; Cooney, J. M. Chem. Res. Toxicol. 2004, 17, 1423. (f) Halim, R.; Brimble, M. A.; Merten, J. Org. Lett. 2005, 7, 265. 26

(a) Scheuer, P. J.; Takahashi, W.; Tsutsumi, J.; Yoshida, T. Science 1967, 155, 1267. (b) Murata, M.; Legurand, A. M.; Ishibashi, Y.; Fukui, M.; Yasumoto, T. J. Am. Chem. Soc. 1990, 112, 4380. (c) Satake, M.; Morohashi, A.; Oguri, H.; Oishi, T.; Hirama, M.; Harada, N.; Yasumoto, T. J. Am. Chem. Soc. 1997, 119, 11325. (d) Hamajima, A.; Isobe, M. Org. Lett. 2006, 8, 1205.

19

1.4 STEREOCHEMISTRY

stereoisomer with two or more stereogenic centers that are not superimposable and not mirror images. By this definition, the (2R,3R)- and (2R,3S)-stereoisomers are also diastereomers. Indeed, the (S,S) stereoisomer is a diastereomer of the (R,S) and (S,R) stereoisomers. Likewise, the (R,R) stereoisomer is a diastereomer of the (R,S) and (S,R) stereoisomers. This rather tedious comparison is meant to show the stereochemical relationship of the four stereoisomers, with an emphasis on identification of diastereomers. For any molecule, the 2n rule predicts the maximum number of stereoisomers, and there cannot be >2n. However, there are instances where there are fewer stereoisomers than the number of stereoisomers predicted by the 2n rule. An example is cis-1,2-cyclopentanediol [(1R,2S)-cyclopentane-1,2-diol], where a plane of symmetry bisects the molecule. Due to this plane of symmetry, the mirror image is also (1R,2S)-cyclopentane-1,2-diol. Both of these structures are superimposable so they represent only one compound. In other words, there is only one structure, not two, and the two structures shown are not enantiomers but the same compound. In effect, one-half of the molecule is the mirror image of the other one-half due to the symmetry of the molecule. Symmetry of this type leads to fewer stereoisomers than the maximum predicted by the 2n rule. Remember that 2n predicts the maximum number of stereoisomers. There can never be more, but there may be fewer. H

Br

Br

H

HO Et Me

Ph

Me Et

Ph

H Et

OH

HO

(2R, 3R)-2-Bromo-2phenylpentan-3-ol

(2S, 3S)-2-Bromo-2phenylpentan-3-ol

Br Me

Ph (2S, 3R)-2-Bromo-2phenylpentan-3-ol

H

Br

Et Me OH Ph (2R, 3S)-2-Bromo-2phenylpentan-3-ol

A molecule with two or more stereogenic centers that has a superimposable mirror image is called a meso compound. The cis-diol, (1R,2S)-cyclopentane-1,2-diol, is a meso compound, but its diastereomer, the trans-diol, exists as enantiomers because the two mirror images are not superimposable. These two enantiomers are (1S,2S)-cyclopentane-1,2-diol and (1R,2R)-cyclopentane-1,2-diol. There are two stereogenic carbons and the 2n rule predicts four stereoisomers, but because one stereoisomer is a meso compound there are only three stereoisomers, not four. The meso compound is a diastereomer of both enantiomers.

OH

OH

OH

OH

OH

OH

OH

OH (1S,2S)-Cyclopentane-1,2-diol

(1R,2S)-Cyclopentane-1,2-diol

(1R,2R)-Cyclopentane-1,2-diol

Meso compounds are possible for acyclic molecules with more than one stereogenic center. The mirror image of (2R,3S)-2,3-dibromobutane, for example, is superimposable and it is the same compound. Clearly, (2R,3S)-2,3dibromobutane is a meso compound, but the diastereomer (2S,3S)-dibromobutane has a nonsuperimposable mirror image, (2R,3R)-dibromobutane, so they are enantiomers. 2,3-Dibromobutane has three stereoisomers despite having two stereogenic centers. Note that changing the configuration of one stereogenic center (2R ! 2S) cannot be done by rotation about any bond, but only by making and breaking bonds, which is a chemical reaction. Also note that changing the configuration of one of the stereocenters leads to the diastereomer, but changing the configuration of both stereocenters leads to the mirror image. H H

Br

Br Me

Br

H

H H

Br Me

Me (2R,3S)-2,3-Dibromobutane

Me

Br

Br H Me

H

Br H Me

Br

Me Me (2S,3S)-2,3-Dibromobutane (2R,3R)-2,3-Dibromobutane

Different nomenclature systems have been developed based on the relationship of the atoms or groups on the stereogenic centers. One system for distinguishing diastereomers labels them threo and erythro. Unfortunately, there

20

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

is more than one definition for threo and erythro. Winstein and Lucas27 gave the following definition of these diastereomers: In a compound with two asymmetric carbons that has two common ligands and a third that differs, the isomers that would be meso if the third ligand were identical are erythro diastereomers. Here is an alternative definition: If two asymmetric carbons have only one ligand in common, then the other four ligands are paired in the same commonsense way and isomers that would have equal pairs eclipsed in any conformation are erythro.28 Eliel, Mislow, and their coworkers29 defined erythro and threo in terms of Fischer projections. The aldol products (see Section 13.4.1) (1S,2R)-1-hydroxy-2-methyl-1-phenylpentan-3-one and (1S,2S)-1-hydroxy-2-methyl-1-phenylpentan-3-one, and the ester-aldehyde condensation products methyl (2R,3S)-3-hydroxy-2,4-dimethylpentanoate and methyl (2S,3S)-3hydroxy-2,4-dimethylpentanoate30 are shown with the erythro and threo notation. OH

O

OH

Erythro

Ph

O

Me

Threo

Ph Me

CO2Me

CO2Me Me

(1S,2 R)-1-Hydroxy-2-methyl1-phenylpentan-3-one

Me Erythro

Threo

OH

OH

(1S,2S)-1-Hydroxy-2-methyl1-phenylpentan-3-one

Methyl (2R,3S)-3-hydroxy2,4-dimethylpentanoate

Methyl (2S,3S)-3-hydroxy2,4-dimethylpentanoate

To alleviate the confusion of the various uses and perceptions of the erythro/threo notation,31 Masamune et al.32 proposed the terms syn and anti to describe the relative stereochemistry of diastereomers. In this notation, adjacent groups on the same side of an extended (zigzag) structure are syn, and those on opposite sides are anti. This model is illustrated with three carboxylic acids that have the syn/anti labels, (2S,3S)-2,3-dimethylpentanoic acid, (2S,3R,4R)3-hydroxy-2-methyl-4-phenylpentanoic acid, and (2S,4R)-2,4-dimethylhexanoic acid.32,33 An extended structure is used as the basis of the model. Me

Me C C Anti 2 3 CO2H Me (2S,3S)-2,3-Dimethylpentanoic acid

Me C C Syn; C C Anti 2 3 3 4

CO2H OH (2S,3R,4R)-3-Hydroxy-2-methyl4-phenylpentanoic acid

Me

Ph

Me

C 2C 4 Anti

CO2H (2S,4 R)-2,4-Dimethylhexanoic acid

The clearest way to show differences is to use the absolute configuration (R) or (S) nomenclature for each stereogenic center, as shown in the names of each example. Throughout this book, the correct configuration will be cited or the syn and anti terminology will be used.

There are molecules that have several stereogenic carbon atoms, but the molecules are symmetrical. Two examples are (2R,3r,4S)-2,3,4-trihydroxypentanedioic acid (known as ribaric acid) and (1s,3R,5S)-3,5-dimethylcyclohexan-1-ol. Note the use of lower case r and a lower case s in the name. These carbon atoms are identified as asymmetric carbon atoms. These tetrahedrally coordinated carbon atoms are bonded to four different entities, two and only two of which have the same constitution, but opposite chirality sense.14b The r/s descriptors are used for pseudo-asymmetric carbon atoms that are invariant on reflection in a mirror (i.e., r remains r, and s remains s), but are reversed by the exchange of any two entities (i.e., r becomes s, and s becomes r) 27

Winstein, S.; Lucas, H. J. J. Am. Chem. Soc. 1939, 61, 1576, 2845.

28

(a) Lucas H. J.; Schlatter, M. J.; Jones, R. C. J. Am. Chem. Soc. 1941, 63, 22. (b) Cram, D. J. J. Am. Chem. Soc. 1952, 74, 2149. (c) Curtin, D. Y.; Kellom, D. B. J. Am. Chem. Soc. 1953, 75, 6011. (d) House, H. O. J. Am. Chem. Soc. 1955, 77, 5083.

29

(a) Eliel, E. L.; Wilen, S. H.; Mander, L. N. Stereochemistry of Organic Compounds; Wiley–Interscience: New York, NY, 1994. (b) Eliel, E.L. Stereochemistry of Carbon Compounds; McGraw-Hill: New York, NY, 1962. (c) Mislow, K. Introduction to Stereochemistry; W.A. Benjamin: New York, NY, 1965. 30

Meyers, A. I.; Reider, P. J. J. Am. Chem. Soc. 1979, 101, 2501.

31

Seebach, D.; Prelog, V. Angew. Chem. Int. Ed. 1982, 21, 654.

32

Masamune, S.; Ali, Sk. A.; Snitman, D. L.; Garvey, D. S. Angew. Chem. Int. Ed. 1980, 19, 557.

33

Heathcock, C. H. In Asymmetric Synthesis; Morrison, J. D., Ed.; Academic Press: New York, NY, 1984; Vol. 3.

21

1.4 STEREOCHEMISTRY

O

OH

OH

O

HO

OH OH

OH

(2R,3r,4S)-2,3,4-Trihydroxypentanedioic acid (Ribaric acid)

(1s,3R,5S)-3,5-Dimethylcyclohexan-1-ol

1.4.3 Chiral Molecules Without a Stereogenic Center (Molecules Containing a Chiral Axis) A few classes of organic molecules have a chiral axis although they do not have a stereogenic center.34 The mirror image of such a molecule is not superimposable, which means it is possible to have enantiomers without the presence of a stereogenic center. Four important classes of compounds that exhibit this property are biaryls (e.g., 21), alkylidene cyclohexanes (22), substituted allenes (e.g., 23), and substituted spiranes [e.g., 3,9-diphenylspiro[5.5]tridecane (24)].15,34 Chiral biaryls are important chiral catalysts and are used in many reactions (see Sections 7.9.3, 7.10.1, 7.10.8, and 14.10.2). Obviously, it is important to determine the absolute configuration of both the chiral reactants and the asymmetric products resulting from their use. R3

R1 R3

R1

R3

R1

C R4 R4

R2

C

H

H

C

R4

Ph

R2

Ph

R2 21

22

24

23

Allenes are common partners in pericyclic reactions (see Sections 14.5, 15.2, and 15.4) and they are chiral partners in some of these reactions. Alkylidene cyclohexanes are produced by phosphorus ylids upon reaction with cyclohexanone derivatives (see Section 12.5.1) and the potential for creating asymmetric products is a key consideration in planning a synthesis of such compounds. a a

b

a

X

X

X b

H

Me

b

a

b

C

• d

c

25

c

26A

c

a

C L

b

d

b L

a

L

C

d c

d

d

c

Y

Y

26B

26C

d

c 27

Y

a b

d c

27A

27B

No chiral atom is present in the molecules described above, so the CIP rules described in Section 1.4.1 do not directly apply. Chiral molecules that have no stereogenic center are evaluated by recognizing the presence of what is known as an extended tetrahedron (26A).35 A normal tetrahedral carbon atom is shown by 25, and distortion of the bond lengths for atoms a ! d yield an extended tetrahedron (26A). Rather than a chiral atom, 26A contains a chiral axis (see X⋯Y in 26B), which can be used to assign priorities. This model requires that 26B not be interconvertible with its mirror image 26C (i.e., rotation about the chiral axis X⋯Y must not interconvert 26B and 26C). The allene, (S)-4-ethyl-5-methylhexa-2,3-diene, is a chiral molecule with a chiral axis because there are four different groups at each corner of an extended tetrahedron. These groups can be assigned priorities a ! d by the same rules used for molecules with a stereogenic center. Rotation of the extended tetrahedron such that the d-atom which leads to an a ! b ! c sequence in front, and could be used for determining the absolute configuration. This protocol is incorrect. There is no appropriate angle from which to view the extended tetrahedron, so there is no C*dCd axis from which to view the molecule. Therefore, the CIP rules were modified to accommodate the extended tetrahedron.35 (1) The top 34

For a general discussion of this concept and molecules that exhibit this property, see Ref. 10, pp 1119–1190.

35

Ref. 10, pp 1119–1122.

22

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

edge of the extended tetrahedron is prioritized a,b and the bottom edge of the extended tetrahedron is also prioritized a,b. (2) Near groups precede far groups when viewed from the top of the extended tetrahedron. With these modified rules, 4-ethyl-5-methylhexa-2,3-diene is converted to model 28, where the top a,b pair in the model is associated with 1,2 (a ¼ 1, b ¼ 2). Similarly, the bottom a,b pair is assigned 3,4 (a ¼ 3, b ¼ 4) since top has priority over bottom. The model is rotated to put (4) to the rear and follows the order 1 ! 2 ! 3, when yields a counterclockwise pathway and an (S) configuration, and (S)-4-ethyl-5-methylhexa-2,3-diene. H

Me

a

C

1

b

C

L

1

2

L

2

2

L

1

L

C Et b

3

a 4

(S)-4-Ethyl-5-methylhexa-2,3-diene

4

3

4

(S)

28 a

H

Me

3

1

b

2

C

L

C

L

C Me

4

b

H

3

a

( R)-Penta-2,3-diene

( R)

This new model can also be used with allene penta-2,3-diene, where only two different groups are present. The methyl-hydrogen priorities are shown in the model and are converted to the 1 ! 4 priority scheme, which places (4) to the rear, yielding an (R) configuration.15, The name is (R)-penta-2,3-diene. Molecules such as this can also be evaluated a priori using the Lowe-Brewster rules.36 Eliel et al.37 point out that the configuration of allenes and alkylidene cycloalkanes can be predicted by these rules in most case, but the model fails for many spirans. Chiral axes occur in cyclic molecules with an exocyclic alkylidene moiety, such as the molecule marked (R,Z)-1-(2carboxyl)-4-methylcyclohexane, which has two different groups at C4 of the cyclohexyl system and two different groups attached to the π-bond. The COOO > H priority for CO2H and H on the alkene is straightforward, and this constitutes the top of the extended tetrahedron. The cyclohexyl arms are in the plane of the π-bond, but the methyl and hydrogen at C4 of the cyclohexane ring are in a different plane that constitutes the bottom of the extended tetrahedron. Since the methyl carbon has priority over the hydrogen the extended tetrahedron model leads to the (R)-configuration.35 HOOC

H

a

L CH 3 H ( R, Z )-1-(2-Carboxyl)4-methylcyclohexane

1

b

2

L a

b

( R) 3

4

Biaryls [e.g., (R)-50 -chloro-20 ,5-dimethoxy-60 -methyl-[1,10 -biphenyl]-2,30 -dicarboxylic acid] can be analyzed with the extended tetrahedron model. The top aromatic ring is prioritized as a,b for the 2,6 substituents, as is the bottom ring (see the model). The near-far rule leads to an (R)-configuration.18

36

(a) Lowe, G. Chem. Commun. 1965, 411. (b) Brewster, J. H. Top. Stereochem. 1967, 2, 1. (c) Ref. 10, pp 1129–1132.

37

Ref. 10, p 1091.

23

1.4 STEREOCHEMISTRY

MeO

b

HC

C

CO2 H

H Me

CC

CO2 H

Cl

L

L

OMe

2

a

CCC

C

C

1

1

L

CO

( R)

L 3

a

b

2

4

4

3

(R)-5 -Chloro-2 ,5-dimethoxy-6 -methyl-[1,1 biphenyl]-2,3 -dicarboxylic acid

Cl

CO2H

Me

OMe

H

b C

a C CO

CC

2

L

L C

( R)

3

CCC

HC

2

L

CO2H

MeO

1

1

4

4

3

a

b

A reasonable question asks which aryl ring is on top and which is on the bottom of the extended tetrahedron. Do the rules accommodate these two orientations? The second structure is identical to the first one, except that the former structure has been rotated by 180°. Analysis of both structures leads to an (R)-configuration. Prioritizing top and bottom separately accounts for rotating the molecule in this manner, but the top must be assigned before prioritizing the molecule and must not be changed after the process has begun. An example of a biaryl system used in synthesis is the chiral reducing agent BINAL-H (an abbreviation for 29).38 The extended tetrahedron reveals an (R) configuration. Note that binding the Al atom into a ring, as shown in 29, effectively locks the aromatic rings into a single configuration and does not allow racemization unless the CdCdCdCdAldOd ring is disrupted.

b O

1

1

2

H L

Al O

2

a L

OEt b

a

( R)

L 4

3 4

3

29

1.4.4 (E/Z) Isomers Alkenes are an important class of molecules that figure prominently in many organic reactions. One of the impor3 3 tant features of an alkene is a π-bond, and while rotation is possible about a Csp dCsp bond, rotation is impossible about a π-bond. This particular structural feature is apparent if one is asked to draw the structure of hex-3-ene. However, there are two possible structures, one with both ethyl groups on the same side of the C]C unit, and the other with the two ethyl groups on opposite sides. They are not superimposable, nor are they mirror images, but they are clearly isomers. Rotation is not possible around the C]C moiety, so the position of the ethyl groups and the hydrogen atoms cannot be interconverted by rotation. The two different hex-3-enes are stereoisomers. Although there is no stereogenic center, 30 and 31 are stereoisomers, and they are formally considered to be diastereomers. In this particular case, the C]C unit has two attached groups (ethyl), which may be on the same side of the

38

(a) Noyori, R.; Tomino, I.; Tanimoto, Y.; Nishizawa, M. J. Am. Chem. Soc. 1984, 106, 6709. (b) Noyori, R.; Tomino, I.; Yamada, M.; Nishizawa, M. J. Am. Chem. Soc. 1984, 106, 6717.

24

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

C]C or on opposite sides. When an alkene contains identical groups and those groups are on the same side, it is named as a cis-alkene. When the two identical groups are on opposite sides, it is named as a trans-alkene. Alkene 30 is cis-hex-3-ene and 31 is trans-hex-3-ene. These terms are part of the nomenclature, as shown.

30

31

The terms cis and trans are applied in a straightforward manner for simple alkenes (e.g., cis- and trans-hex-3-ene or cisand trans-1,2-dibromoethene). For 3-(2-bromopropan-2-yl)-5-hydroxy-4-isopropylhex-3-enoic acid, however, no two groups are the same and the cis-trans nomenclature does not apply. A more general nomenclature system is required, the (E/Z) system.14a,39 The CIP selection rules are used to assign the priority of the groups attached to each carbon of the double bond. Similarly, a and a0 are on opposite sides in 32 and this arrangement is given the designation (E) (from entgegen ¼ opposite). With this system, the hydroxyl-bearing carbon of C4 is assigned the highest priority (CCHO vs. CCCH for the i-Pr carbon). Analysis of C3 shows the bromine-bearing carbon to be the highest priority (CCCBr vs. CCHH). In 32, a and a0 have the higher priorities and b and b’ have the lower, as shown in Fig. 1.7. Since a and a0 are on the same side of the double bond, they are given the designation (Z) (from zusammen ¼ together). Therefore, the name is (E)-3-(2-bromopropan-2-yl)-5-hydroxy-4-isopropylhex-3-enoic acid. This nomenclature is generally applicable to all alkenes that do not have identical groups on one of the alkenyl carbons. Hex-3-ene (30) is hex-3(Z)-ene and 31 is hex-3(E)-ene. The (E/Z) system will be used extensively throughout this book, but the cis-trans designations will be used for simple molecules and to describe relative stereochemistry. For example, the carbonyl group (C1) and the hydroxyl-bearing moiety (C5) in (E)-3-(2-bromopropan-2-yl)-5-hydroxy-4-isopropylhex-3-enoic acid are cis to each other.

O

OH 2

6

5

1

OH

3 4

Br

(E)-3-(2-Bromopropan-2-yl)-5-hydroxy-4-isopropylhex-3-enoic acid O

OH

2

a

5 6

CCHO > CCCH a b

• 4

a

b

•

b

b 32

FIG. 1.7

a

3

4

1

•

•

OH

3

a

b 33

Br

a′

CCCBr > CCHH b′

Priorities for determining (E)- and (Z)-isomers in 3-(2-bromopropan-2-yl)-5-hydroxy-4-isopropylhex-3-enoic acid.

1.4.5 Prochiral Centers There are molecules that do not possess a stereogenic center, but generate a product with a stereogenic center after a chemical reaction. In terms of planning a reaction, predicting the absolute configuration of any and all stereogenic centers in the target, and all products is important. Such predictions cannot be made unless the face of the molecule from which the reagent will approach during the reaction. If it approaches from one face, the (R)-enantiomer is generated; if it approaches from the opposite face the product is the (S)-enantiomer. It is obvious that butan-2-one, and 3-ethylpent-2-ene do not contain a stereogenic center. If butan-2-one reacts with a Grignard reagent (e.g., phenylmagnesium bromide, PhMgBr), however, the product is a racemic alcohol with enantiomers (R)-2-phenylbutan-2-ol and (S)-2-phenylbutan-2-ol (see Section 11.4.3). The reaction of 3-ethylpent-2-ene with a borane [e.g., 9-BBN (9-borabicyclo[3.3.1]nonane); see Section 9.2] followed by oxidation, yields a racemic alcohol as the major product, enantiomers (S)-3-ethylpentan-2-ol and (R)-3-ethylpentan-2-ol (see Section 9.4.1).

39

Ref. 10, pp 541–543.

25

1.4 STEREOCHEMISTRY

O

Ph

1. PhMgBr, THF

HO

OH

Ph

+

2. H3O+

(S)-2-Phenylbutan-2-ol

( R)-2-Phenylbutan-2-ol

Butan-2-one

HO

H

H

1. 9-BBN

OH

+

2. H2O2, NaOH

3-Ethylpent-2-ene

( R)-3-Ethylpentan-2-ol

(S)-3-Ethylpentan-2-ol

In both of these examples, the reaction produced a product that contains a stereogenic carbon. The ketone and alkene are described as prochiral,40 and a working definition of prochirality was provided by Hanson41: If a chiral assembly is obtained when a point ligand in a finite non-chiral assembly of point ligands is replaced by a new point ligand, the original assembly is prochiral. A ligand is simply a group attached to the prochiral atom (a, b, c, d). A point ligand for a prochiral center has two characteristics: “(1) any two point ligands may be identical or non-identical, and (2) two point ligands may not occupy the same position in space.”41 Addition of a point ligand to a prochiral center creates a new chiral center.41 If the ligands are assigned a priority a-b-c-d by the CIP selection rules, configurations for each face of the prochiral center can be obtained. This concept is illustrated by a prochiral atom represented by tetrahedron 34 in Fig. 1.8, which undergoes a chemical reaction to generate a new stereogenic center, yielding either 35 or 36. In this example, 36 represents (S) stereogenic center and is formed by replacement of a1 with b. Similarly, replacing a2 generates (R) stereogenic center 35. The a1 and a2 atoms are described by the terms pro-S and pro-R, respectively. A pro-R center is a molecule with a ligand in a prochiral molecule whose replacement leads to an (R) center. Analogously, a proS center has a ligand in a prochiral molecule whose replacement leads to an (S) center. In 34, ligand a2 is pro-R and a1 is pro-S, and an example is D-glyceraldehyde. The methylene group adjacent to the chiral center is prochiral, and replacement of Hb with a group X yields (S) center in 37 (assuming X has a priority of c with H ¼ d). Replacement of Ha yields (R) center (in 38) and that hydrogen is pro-R. a2

a1

a1 b

• d

c

•

34

35

Ha

Hb

Hb

X CHO

HO CHO

HO

OH

OH D-Glyceraldehyde

(R)-2,3-Dihydroxypropanal

FIG. 1.8

Ha

36 X

OH

Ha

d

c

CHO H

•

OR

d

c

a2

b

b

X

X

Hb

OH

CHO

HO 38

37

OH

Determination of prochirality.

Hanson41 described rules that accommodate most situations observed with prochiral centers. The reader is referred to this work for specific examples that do not yield to a simple analysis. The intent of this section is to familiarize the reader with nomenclature and the identification of pro-R and pro-S sites in reactions. As noted above, typical prochiral centers are the carbonyl of an unsymmetrical ketone or an aldehyde, and the double bond of an alkene. These functional groups do not contain a pro-R or pro-S group, but it is clear that delivery of a fourth point ligand from one face or the other will lead to an (R) or (S) stereogenic center, as in conversion of 39 to 40 and/or 41 in Fig. 1.9. If the carbonyl group is oriented as in ketone 1-chloro-5-methylhexan-3-one-A, priorities can be assigned to the three atoms connected to the prochiral atom based on the CIP rules. 40

Ref. 10, pp 465–488.

41

Hanson, K. R. J. Am. Chem. Soc. 1966, 88, 2731.

26

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

X

R1

X

R1

a

b

O R2

R1

X

OH

HO

R2 39

R2 40

L

41

CHH

CCH

O

X

CHH

CJC c

c

HHCl

CJC b

L

b

a

O a

Cl

b

L Cl Si

a

O

Re

1-Chloro-5-methylhexan-3-one-B

1-Chloro-5-methylhexan-3-one-A

c

FIG. 1.9 Acyl addition to prochiral carbonyl compounds.

For 1-chloro-5-methylhexan-3-one-A, the a ! b ! c priority is counterclockwise, which correlates to (S). It is not really an (S) configuration, of course, since it is not a stereogenic center, so the (S) sequence is termed si, from the Latin sinister.41 When the incoming group approaches the π-bond from this face, with the orientation shown in Fig. 1.10, it is called the si face. If the molecule has the opposite orientation in a reaction, the reagent approaches from the other face and the priority sequence a ! b ! c is clockwise or (R), which is termed re, from the Latin rectus. That face is the re face.41 Cl

Cl

Cl

[H –] H

b

a

OH

From si

re

si

HO

H

From re

1-Chloro-5-methylhexan-3-one-C

(R)

FIG. 1.10

O

[H –]

(S)

The re and si faces of 1-chloro-5-methylhexan-3-one.

In this example, face a (see 1-chloro-5-methylhexan-3-one) is the si face and face b is the re face. Attack of hydride (see Sections 7.4, 7.6.1, and 7.9.2) from face a (the si face) leads to the (R) alcohol and attack from face b (the re face) leads to the (S) alcohol. The configuration of the final product depends on the priority of the new group added to the prochiral center. Similar terminology can be applied to alkenes, as with (Z)-1-bromoprop-1-ene. In this case, there are two prochiral centers to be considered (C1 and C2 of the C]C bond), and re or si is assigned to each carbon, as shown in Fig. 1.11. For (Z)-bromoprop-1-ene, the top faces of both C1 and C2 are re (re-re) and the bottom faces are si (si-si) c

b

re a

FIG. 1.11

H

CH3

H3C

1

C

C

C

C

H

Br 2 H (Z)-1-Bromoprop-1-ene

C Br

C

b re H

a

c

The re and si faces of (Z)-1-bromoprop-1-ene.

1.4.6 Definitions of Selectivity In the preceding sections, various types of molecules were classified as regioisomers or stereoisomers (further categorized as diastereomers and enantiomers). When there are two different functional groups in a molecule, a given reagent may react preferentially with one rather than the other. Such a reaction is sometimes termed chemoselective. Oxidation of 42 with manganese dioxide (MnO2, see Section 6.2.5.3) gave a 50% yield of 43 and 25% of 44.42 Manganese dioxide showed a preference for oxidation of the secondary allylic alcohol at the expense of the primary alcohol. The reagent selectively reacted with one alcohol moiety over the other, and such selectivity is termed chemoselectivity: of two or more reactive functional groups, one reacted preferentially to yield the major product in the mixture. When both products are formed, but one is formed in greater proportions, the term selective applies. Contrast this reaction with Gribble’s reduction of ketoaldehyde 4-benzoylbenzaldehyde with tetrabutylammonium triacetoxyborohydride [Bu4 N + BHðOAcÞ3  , where Ac ¼ acetyl and Bu ¼ butyl, see Section 7.8.2], to yield [4-(hydroxymethyl)phenyl] (phenyl)methanone in 88% yield.43 42

Hlubucek, J. R.; Hora, J.; Russell, S. W.; Toube, T. P.; Weedon, B. C. L. J. Chem. Soc. Perkin Trans. 1974, 1, 848.

43

Nutaitis, C. F.; Gribble, G. W. Tetrahedron Lett. 1983, 24, 4287.

27

1.4 STEREOCHEMISTRY

MnO2

HO

+

CH2Cl2 2d

OH

O

O

OH

CHO 43

42 Ph

Ph

4 Bu4N+ BH(OAc)3-

CHO

44

PhH , 80°C, 1 d

O

OH

O

4-Benzoylbenzaldehyde

[4-(Hydroxymethyl)phenyl](phenyl)methanone (88%)

If the result had been different, and the aldehyde carbonyl was reduced and a portion of the ketone carbonyl was also reduced, the reaction would be chemoselective rather than chemospecific. If there had been 100% reduction of the aldehyde, but 0% of the ketone, the reaction would be termed chemospecific. Trost defined these terms: “of two or more reactive functional groups, only one reacts (specific), or one predominates (selective).” Reduction of the alkenyl ketone, hept-6-en-2-one, yielded the alkenyl alcohol, alcohol hept-6-en-2-ol, where only the carbonyl reacted with sodium borohydride (NaBH4). Since 0% of the alkene reacted, this reaction is chemospecific. O

OH

1. NaBH4 2. aq NH4Cl

Hept-6-en-2-one

Hept-6-en-2-ol

As noted above, the use of the terms selective and specific for giving a preponderance of a given product or only that product, respectively, can be applied for all reactions involving stereochemistry. If a reaction can produce two or more regioisomers, it is regioselective or regiospecific. The second-order (E2) elimination (see Section 3.5.1) of the racemic bromide 3-bromo-2-methylpentane gave a mixture of 2-methylpent-2-ene and 4-methylpent-2-ene with 2-methylpent-2-ene being the major product, so the reaction is regioselective. Elimination of the enantiopure (S,S)diastereomer [(2S,3S)-2-bromo-2,3-diphenylbutane], however, gave only the (Z)-alkene, (Z)-2,3-diphenylbut-2-ene, with none of the (E)-isomer and none of the alkene formed by removal the α-hydrogen on the methyl. The reaction is regiospecific and also stereospecific. This result was confirmed by reaction of the (R,S)-diastereomer (2R,3S)-2-bromo-2,3diphenylbutane under E2 conditions, which gave only (E)-2,3-diphenylbut-2-ene. If (2S,3S)-2-bromo-2,3-diphenylbutane and (2R,3S)-2-bromo-2,3-diphenylbutane yield an unequal mixture of (Z)-2,3-diphenylbut-2-ene and (E)-2,3-diphenylbut2-ene, the reaction would be stereoselective. One product is formed, so the reaction is stereospecific. Br KOH , EtOH

+ 3-Bromo-2-methylpentane

2-Methylpent-2-ene

Me Br

Ph

4-Methylpent-2-ene

Me

Me

Ph

Ph

KOH , EtOH

H Ph (2S,3S)-2-Bromo-2,3-diphenylbutane Me

Br

Me

(Z)-2,3-Diphenylbut-2-ene

Ph

Me

Ph

Ph

Me

KOH , EtOH

H Me (2R,3S)-2-Bromo-2,3-diphenylbutane Ph

(E )-2,3-Diphenylbut-2-ene

28

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

Similar terminology can be applied to formation of diastereomers when substituted alkenes react with certain reagents (e.g., Br2). Smith44 cited the reactions of bromine with maleic and fumaric acid. On addition of bromine to maleic acid, a racemic mixture of (2R,3R)-2,3-dibromosuccinic acid was formed. The reaction of bromine with fumaric acid gave only (2R,3S)-2,3-dibromosuccinic acid, the meso compound. In both cases, the reaction was diastereospecific. If maleic acid yields predominantly (2R,3R)-2,3-dibromosuccinic acid, with only a trace of the diastereomer [(2R,3S)2,3-dibromosuccinic acid], the reaction would be diastereoselective rather than diastereospecific. The term stereoselectivity is applied to chemical reactions. If a reaction produces at least one substance that is not a stereoisomer of the major product, that reaction cannot be stereospecific, but at most stereoselective.45 If stereoisomeric starting materials react to give a single stereoisomeric product, the reaction is stereospecific, but if another stereoisomer is also produced (giving a mixture of products) the reaction is stereoselective. These terms apply to all types of stereoisomers, including enantiomers (enantioselective and enantiospecific), diastereomers (diastereoselective and diastereospecific), and regioisomers (regioselective and regiospecific). Reduction of the keto-ester, methyl (R)-2-methyl-3-oxo-3-phenylpropanoate, with zinc borohydride [Zn(BH4)2, see Sections 7.5 and 7.9.2] gave 98% reduction with a >99:1 preference for methyl (2R,3R)-3-hydroxy-2-methyl-3-phenylpropanoate over methyl (2R,3S)-3-hydroxy-2-methyl-3-phenylpropanoate. The reaction produced two diastereomers, and it is diastereoselective.45 Since traces of methyl (2R,3S)-3-hydroxy2-methyl-3-phenylpropanoate are produced, the reduction cannot be diastereospecific.

HO2C

Br2

CO2H

Br CO2H (2R,3R)-2,3-Dibromosuccinic acid Br H

H

CO2H

Br2

Br H (2R,3S)-2,3-Dibromosuccinic acid

CO2H

H

CO2H

H

Maleic acid HO2C

H

Br

H

H

HO2C

Fumaric acid

The final definition concerns formation of enantiomers, and the terms enantioselective and enantiospecific are used. If a reaction produces an unequal mixture of enantiomers, it is enantioselective. If it generates only one enantiomer of two possibilities, it is enantiospecific. The baker’s yeast reduction (see Section 7.12.6) of hexane-2,4-dione gave (S)-5hydroxyhexan-3-one with >99 %ee (S).46 (Here %ee means percent of enantiomeric excess.) A 0 %ee means a 50:50 mixture (racemic mixture), 50 %ee means a 75:25 mixture, and 90 %ee means a 95:5 mixture. The predominance of the (S)-enantiomer makes this reaction highly enantioselective. Me

Me 1. Zn(BH4)2

Ph

CO2Me

Me

Ph

Ph

+

CO2Me

2. H3O+

CO2Me

OH

O

OH

Methyl (2R,3R)-3-hydroxy-2methyl-3-phenylpropanoate (98%)

Methyl (R)-2-methyl-3-oxo3-phenylpropanoate

Methyl (2R,3S)-3-hydroxy-2methyl-3-phenylpropanoate

The selectivity terms introduced in this section will be used throughout the book for reactions that generate stereoisomers. In addition to %ee, % de (diastereomeric excess, defined in the same way as ee, but for diastereomers), or % dr (diastereomeric ratio) will be used throughout. O

O

Baker's yeast

Hexane-2,4-dione

44

OH

O

(S)-5-Hydroxyhexan-3-one

Ref. 16, p 1002.

45

Oishi, T. In New Synthetic Methodology and Functionally Interesting Compounds; Yoshida, Z., Ed.; Kodansha/Elsevier: Tokyo/Amsterdam, 1986; pp 81–98. 46

Bolte, J.; Gourcy, J.-G.; Veschambre, H. Tetrahedron Lett. 1986, 27, 565.

1.5 CONFORMATIONS

29

1.5 CONFORMATIONS Organic molecules can, for the most part, be categorized as having tetrahedral, trigonal, or digonal (linear) geometry around each sp3, sp2, or sp hybridized carbon atom, respectively. When organic molecules absorb energy from the environment, translational motion and collisions dissipate much of that energy. Molecules can also dissipate excess energy by molecular vibrations, including rotation or twisting about all carbon-carbon single bonds. Generally, the more energy absorbed, the more facile will be the rotation. For acyclic molecules, the three-dimensional (3D) nature of carbon compounds, and rotation around carbon-carbon bonds leads to different spatial orientations of atoms attached to each carbon atom called rotamers. Different rotamers are not different structures, but different orientations of atoms in the same molecule. Analysis of the spatial relationships of the atoms and groups can yield information about these respective interactions as they rotate around the carbon-carbon bond. Indeed, the shape that a molecule assumes is determined by understanding the rotamer population for all bonds, which leads to the overall shape (conformation) of the molecule. Most molecules have more than one conformation, but one or at least a small subset will usually constitute the low energy conformation(s). The shape associated with a molecule is usually taken to be the lowest energy conformation of that molecule. For cyclic and polycyclic molecules, complete rotation about each CdC bond is not possible, but partial rotation known as pseudorotation allows the molecule to dissipate energy. Pseudorotation and interactions of the various atoms and groups will lead to at least one and often several energy minima that are drawn to represent the shape (conformation) of that cyclic molecule. In chemical reactions, the shape of the molecule will influence the way an incoming reagent interacts with it and this can have a significant effect on both reactivity and stereochemical induction, especially with the more rigid cyclic and polycyclic molecules. This information is used to determine how reagents will approach the molecule, and even the stereochemistry of certain reactions. This section will discuss the fundamental conformational preferences of simple acyclic and cyclic molecules. These principles will be applied throughout the book for discussions of chemoselectivity and stereoselectivity. An example of the link between conformation and stereoselectivity is illustrated with bicyclic ketone 45, which has a rather rigid bicyclic structure. In such a case, a simple drawing is often sufficient to conclude that one face is more sterically hindered than the other. Based on this model, the direction of hydride attack (see Section 7.9.1) is predicted to be from the more open face of the molecule, which is along the direction of the arrow in 45 as it is drawn. Delivery of hydride from this face leads to equatorial alcohol 46 as the only isomer isolated in 99% yield. This reaction was taken from the total synthesis of the microbial immunosuppresive agent FR901483, reported by Weinreb and coworkers.47 Predicting stereochemistry in reactions of acyclic and monocyclic molecules is problematic, however, since they do not assume a single, rigid conformation. They are quite flexible and all but the simplest molecules may exist in a variety of different shapes or conformations. Indeed, the shape of such molecules depends on the torsional angles involving the various single bonds. Therefore, it is not possible to unambiguously identify which carbonyl face is likely to be the less crowded in a flexible ketone, and to thereby identify the stereochemistry of a reaction product. The second example using ketone 47 is far more typical, where there are several possible conformations. Each has a different 3D structure with different hindered or exposed regions. When 47 was reduced with NaBH4 (Section 7.4), a 1:1.7 mixture of diastereomeric alcohols 48 and 49 was formed, taken from Jan and Liu’s synthesis of (+)-ricciocarpin A.48 Even this relatively simple molecule shows enough conformational flexibility that a mixture of diastereomers is formed. In Section 7.9.2, the Cram model and the Felkin-Anh model will be discussed, and used to predict selectivity in acyclic systems. However, both models assume that one key conformation is present in order to make the prediction. Identification of the “relevant” conformer or set of conformers is a necessary first step to anticipate product selectivity, and a conformational analysis is required.

47

Kropf, J. E.; Meigh, I. C.; Bebbington, M. W. P.; Weinreb, S. M. J. Org. Chem. 2006, 71, 2046.

48

Jan, N.-W.; Liu, H.-J. Org. Lett. 2006, 8, 151.

30

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

O

O O N

Ph

H

NaBH4 , MeOH

OH

N

Ph

45

46 (99%)

I

I O

I OH

NaBH4 , MeOH

OH +

rt , 3 h (79%)

O 47

O

48

49

O

1.5.1 Conformations of Acyclic Molecules 1.5.1.1 Conformations of Simple Alkanes Rotation about all single bonds in an acyclic molecule will lead to low-energy rotamers for each bond, and a lowenergy conformation of the molecule is effectively the sum of low-energy rotamers for all bonds in the molecule. The number of different conformations (conformers) therefore depends on the number of single bonds. A simple rule of thumb is that each single bond multiplies the number of possible conformers by three. Thus, a molecule with one single bond has three conformers, a molecule with two single bonds has nine conformers, and so on. This particular process cannot be used to elaborate conformers for flexible rings, and they will be discussed in Section 1.5.2. In the final analysis, even a reasonably sized organic molecule can exhibit hundreds to thousands of possible conformations. Thus, identifying the relevant shape cannot be accomplished by simply looking at a single drawing or by manipulation of a simple model. This section will use molecular modeling as a tool to help understand conformational analysis, because it allows more detailed analysis than would otherwise be possible. This treatment relies heavily on data from quantum chemical calculations. Properly used, reliable predictions can be made about what shapes flexible molecules adopt and, in so doing, provide insight into how they are likely to react. Important conformational preferences must first be established for relatively simple organic molecules. Once these principles have been established, molecular modeling allows the analysis of more complex molecules. Sawhorse Diagram H

H H

H

H H

H

Ethane-anti

H

H H

H H

Newman Projection

H Ethane-syn

H

H H

•

HH

H

H

• H

H Ethane-anti

H H Ethane-syn

Ethane is perhaps the simplest example of an acyclic molecule capable of rotation about two carbon atoms. If one could freeze this motion at different positions, the result would be different rotamers. There are two common models used to represent rotamers: sawhorse diagrams and Newman projections (the dot and circle model shown).29 The ethane rotamer marked “anti” is shown as both a sawhorse diagram and a Newman projection reveals that the CdH bonds and the hydrogen atoms attached to each carbon are as far removed from each other as possible. This arrangement minimizes the steric effect that arises when the hydrogen atoms are close in space. The electronic interactions of the electrons in the CdH bonds are also minimized since the bonds are as far apart as possible. The “syn” rotamer sharply contrasts with the anti rotamer, since the CdH bonds and the hydrogen atoms attached to each carbon are as close together as possible in the syn rotamer. The traditional view is that nonbonded interactions between the two eclipsing hydrogen atoms and the electronic repulsion of the CdH bonds destabilize this rotamer, making it higher in energy. There are other repulsive forces, and

31

1.5 CONFORMATIONS

15 syn-Ethane

rel. E (kJ/mol)

10

5 anti-Ethane

anti-Ethane

0 0

FIG. 1.12

syn-Ethane

120 240 HCCH torsion angle

360

Energy barrier for ethane over a rotation of 360°.

a review categorized destabilizing interactions as Pauli exchange steric repulsion, hyperconjugation, and relaxation energy changes.49 Skeletal relaxation effects play a role.49 An analysis of ethane shows that the energy of each rotamer depends on the torsion angle about the carbon-carbon bond. Beginning with the syn rotamer, rotation of the CdC bond from 0° to 60° leads to an anti rotamer. Likewise, rotation of 60° converts the anti rotamer to a new syn rotamer. Complete (360°) rotation about the CdC bond (see Fig. 1.12) leads to three identical energy minima that correlate with three identical staggered structures (anti-ethane), and three identical energy maxima that correlate with three identical eclipsed structures (syn-ethane). The difference in energy between the staggered and eclipsed structures is measured to be 2.97 kcal (12 kJ) mol1,50 and it constitutes an energy barrier to rotation about the CdC single bond. This energy barrier is small enough that there is no significant hindrance to bond rotation at normal temperatures. In general, there is a much higher percentage of the lower energy staggered (anti) rotamer.

Propane

2-Methylpropane

syn-Butane-A

49 50

2,2-Dimethylpropane

syn-Butane-B

Goodman, L.; Pophristic, V.; Weinhold, F. Acc. Chem. Res. 1999, 32, 983.

(a) Gunstone, F. D. Guidebook to Stereochemistry; Longman: London, 1975; p 56. (b) Ref. 10, p 599. (c) Ref. 28b, p 125. (d) Kagan, H. Organic Stereochemistry; Halsted Press/Wiley: New York, NY, 1979; p 50. (e) Hine, J. Physical Organic Chemistry, 2nd ed.; McGraw-Hill: New York, NY, 1962; p 36.

32

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

Propane, with two identical carbon-carbon bonds, can give rise to nine identical energy minima and nine identical energy maxima. As with ethane, there is a higher percentage of low-energy rotamers relative to high-energy rotamers. 2-Methylpropane has three identical CdC bonds that can give rise to 27 identical sets of energy minima and maxima, and 2,2-dimethylpropane has four identical CdC bonds that can give rise to 81 identical sets of energy minima and maxima. Staggered rotamers are expected to be lowest in energy. Butane has a new structural feature that must be addressed. The C1dC2 and the C3dC4 bonds are identical in terms of their substitution pattern (CH3 and CH2Et), but the C2dC3 bond is different with each carbon having one methyl and two hydrogen atoms. Rotation can occur about both C1dC2 and C3dC4 (H3CdCH2dCH2dCH3), as seen by comparing syn-butane-A and syn-butane-B. Rotation of the central carbon-carbon bond in n-butane (C2dC3) leads to three energy minima and three energy maxima, as shown in Fig. 1.13. Two of the minima (so-called gauche conformers) are identical, while the third minimum (the so-called anti conformer) is different. Likewise, two of the energy maxima (connecting gauche and anti minima) are the same, while the third maximum (connecting the two gauche minima) is different. The magnitude of the barriers connecting anti and gauche conformers depends on the involved rotamers (anti to gauche or gauche to anti), the difference being the same as the energy difference between anti and gauche n-butane conformers. The maximum barrier to rotation is estimated to be 4.5–4.9 kcal (19–21 kJ) mol1,51 which represents the steric interaction of the two overlapping methyl groups. The other energy barrier represents each Me-H eclipsing interaction and is estimated to be 3.6 kcal (15 kJ) mol1.52 The energy estimated for each gauche conformation is 0.96 kcal (4 kJ) mol1,50b an indication that the methyl groups are close, but do not eclipse.

27 syn-Butane

rel. E (kJ/mol)

18

Eclipsed-butane anti-Butane

Staggered-butane

9

Staggered-butane anti-Butane

0 0

240 120 CCCC torsion angle

360 Eclipsed-butane

FIG. 1.13

syn-Butane

Energy barrier for butane over a rotation of 360°.

Butane can be categorized as a dimethyl substituted ethane. Energy profiles for rotation about the central carboncarbon bond for 1,2-dichloroethane and other “disubstituted ethanes” are qualitatively similar, in that they show two gauche minima and one anti minimum (see Fig. 1.14). The relative energies of gauche and anti conformers and rotation barriers depend on the substituents. Anti conformers are lower in energy, and are usually preferred over gauche arrangements, as seen for 1,2-dichloroethane and for 1-chloropropane (A and B, respectively, in Fig. 1.14). One reason

51

(a) Ref. 50c. (b) Ref. 50a, p 57. (c) Ref. 10, p 602. (d) Ref. 29b, p 126.

52

(a) Ref. 50a, p 57. (b) Ref. 50c, pp 48–53.

33

1.5 CONFORMATIONS

for this preference is a desire to minimize unfavorable nonbonded contacts between the substituents. Another factor in some cases is electrostatics, and it is known that bond dipoles align so the total dipole moment is removed. For example, in a lower energy rotamer of difluoroethane the δ + C dFδ bond dipoles can align such that they subtract rather than add.

B

42

28

rel. E (kJ/mol)

rel. E (kJ/mol)

A

14

0

18

9

0 0

120

240

360

C1CCC1 torsion angle

FIG. 1.14

27

0

120

240

360

CCCC1 torsion angle

Energy barrier for substituted ethanes over a rotation of 360°.

1.5.1.2 The Boltzmann Distribution: Average Properties of Flexible Molecules To understand the conformations associated with any sample of n-butane, it is necessary to average the properties of the anti and gauche conformers. More generally, to find the average value of a property (a) of a flexible molecule, it is necessary to sum over all possible different conformers, taking into account the value of the property for that conformer (a), as well as both the number of times that the conformer appears (n) and its Boltzmann weight (w). X a¼ ai n i w i The Boltzmann weight depends on the energy of the P conformer relative to the energy of the lowest energy conformer (Δe), and on the temperature (t): wi ¼ exp(ΔEi/kT)/ exp(ΔEi/kT), where k is the Boltzmann constant. An energy difference of 0.96 kcal (4 kJ) mol1 leads to a Boltzmann weight of 0.1 (10%) at room temperature, an energy difference of 1.9 kcal (8 kJ) mol1 to a weight of 0.05 (5%), and a difference of 2.87 kcal (12 kJ) mol1 to a weight of 0.01 (1%). Only rarely will more than a few of the possible conformers have Boltzmann weights in excess of 1% and contribute significantly to the equilibrium. For molecules like ethane and propane, where all conformers are the same, the average is independent of temperature. For n-butane, where the anti and the 2 equivalent gauche conformers contribute, the average depends on temperature. At very low temperatures, the average rotamer population will be dominated by the lowest energy (anti) conformer, but it will limit to an equal weighting of both conformers as the temperature increases. Note that measurements of some quantities [e.g., the proton nuclear magnetic resource (1H NMR) spectrum], yield averages of possible conformations, while others [e.g., the infrared (IR) spectrum] provide information about individual conformers. In the latter case, low Boltzmann weights will almost always preclude actually “seeing” any but the few lowest-energy conformers. Whether a particular measurement yields an average or discrete quantities depends on the time scale of the underlying physical process. Relaxation of magnetic spin (the basis of 1H NMR) is typically slower than conformational equilibrium at normal temperatures, while molecular vibration (the basis of IR spectroscopy) is much faster. The time scale of the experimental measurement must be established before interpreting the result. Energy barriers for various molecules are quantified in terms of enthalpy (H°), and calculation of the energy of a given rotamer allows us to estimate the relative population of that rotamer (see Boltzmann distribution above). The relative energy of the conformations can be correlated with the relative percentage of each rotamer. An enthalpy difference of 1 kcal (4.186 kJ) mol1 between the anti rotamer and the next most populous rotamer corresponds to the presence of 72% of the anti isomer at room temperature.53 A line drawing of butane is typically drawn in an all-anti conformation. This zigzag or extended conformation of butane is typically used to represent all straight-chain alkanes. 53

(a) Ref. 29b, pp 131–133.

34

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

n-Tridecane, for example, is also drawn as the all-anti conformation, the extended conformation. Virtually all acyclic hydrocarbon chains are assumed to exist primarily in this extended conformation. This assumption is, of course, not 100% correct and while tridecane probably has a high percentage of the zigzag conformation, it is certainly not the only conformation in a real distribution. Indeed, tridecane is more flexible and will have several conformations, and the contribution of the different conformations can be estimated by determining the Boltzmann distribution. When drawing molecules, however, straight-chain alkanes or alkyl fragments are usually drawn in this fully extended form. In reality, the conformational picture is much more complex, but the zigzag drawing is a useful convention for drawing and probably represents a relatively high percentage of all available conformations. H

H H

H H

H H

H H

H H

H

H H H Butane

H H H

H H H

H H

H H

H H

L

H

Tridecane

1.5.1.3 Heteroatom Substituents Thus far, the discussion has been focused on rotation about carbon-carbon bonds. The same principles apply to rotation about carbon-heteroatom bonds, particularly carbon-nitrogen and carbon-oxygen bonds found in many organic molecules. Consider methylamine and methanol. The nitrogen in organic amines is sp3 hybridized. Three of the hybrid orbitals are used to form single bonds to other atoms, and there is a nonbonded pair of electrons (a lone pair). Similarly, the oxygen in alcohols or ethers is sp3 hybridized. In both cases, two of the hybrids are used in bonds to other atoms, leaving two lone pairs. This observation implies that both sp3 nitrogen and oxygen centers are roughly tetrahedral, if the nonbonded electrons are counted. This valence-shell electron-pair repulsion (VSEPR) model is consistent with the fact that nitrogen centers in amines are pyramidal and oxygen centers in alcohols and ethers are angular (bent). Energy plots for methylamine (A) and methanol (B) are similar to that for ethane, as seen in Fig. 1.15. All of these models show three identical sets of energy minima and maxima. Rotational barriers are somewhat smaller than that in ethane: 1.91 kcal (8 kJ) mol1 in methylamine and 0.96 kcal (4 kJ) mol1 in methanol.

FIG. 1.15 Energy barrier for substituted amines, alcohols, and ethers over a rotation of 360°.

If a hydrogen atom on the nitrogen atom in methylamine is replaced with an alkyl group, the result is a secondary amine. Similarly, if a hydrogen atom on the oxygen atom in methanol is replaced with an alkyl group, the result is an

35

1.5 CONFORMATIONS

ether. When analyzing primary or secondary amines, alcohols, or ethers there is little change to the fundamental nature of the energy curves seen for the parent compounds, although there are changes in the details (energy barriers). Extending the carbon chain of methylamine by one carbon atom leads to ethylamine, and extending the carbon chain of methanol by one carbon atom leads to ethanol. These modifications lead to energy curves that show two different minima when compared to the parent molecules (C for ethylamine in Fig. 1.15 and D for ethanol). Note that anti-gauche energy differences for both molecules are very small. Internal H-bond

anti-1,2-Ethanediol

syn-1,2-Ethanediol

Internal H-bond

gauche-1,2-Ethanediol

Groups capable of hydrogen bonding provide a stabilizing interaction that can compensate for the destabilizing interaction due to steric repulsion in the absence of a hydrogen-bonding solvent. It is possible to draw both an antiand syn-conformation for 1,2-ethanediol (ethylene glycol), where the hydroxyl group is capable of hydrogen bonding. Rotating the molecule from the syn-conformation to the gauche-conformation relieves the eclipsing interaction, but maintains the internal hydrogen bonding. In the absence of a hydrogen-bonding solvent, the gauche rotamer is expected to be the low-energy conformation. If ethylene glycol is in a solvent that is unable to hydrogen bond, the diol will hydrogen bond with itself as shown. If a hydrogen-bonding solvent (e.g., water) is used, intermolecular hydrogen bonding will compete with intramolecular hydrogen bonding, and intermolecular hydrogen bonding will predominate based on an analysis of statistics (entropy). H H

O Me

Zn(BH4)2

H B

H

C5 H11 HO

H

H

O O

H

H

C5 H11 H

(S)-2-Hydroxyoctan-3-one

H B H

Zn

Me 50

Another way to stabilize an eclipsed or gauche conformation is to coordinate a heteroatom substituent with a metal ion (chelation). Oishi and coworker’s54 reduction of (S)-2-hydroxyoctan-3-one with zinc borohydride proceeds via a chelated species, 50 (Section 7.5). Chelation of Zn to the hydroxyl and carbonyl groups effectively immobilizes the reactive components into a single conformation or even a single rotamer in the transition state required for reaction, as shown in 50. This fixed conformation sets the position of the methyl and hydrogen at the α-carbon, which leads to facial bias, and the hydride is delivered from the less hindered face over the hydrogen in 50 (see Sections 7.5 and 7.9.2). Since transition metal salts usually behave as Lewis acids, the presence of a heteroatom that functions as a Lewis base (e.g., O, S, N, or P; see Section 2.3) will lead to chelation. The most favored acyclic conformation is usually a gauche or analogous rotamer. 1.5.1.4 Heteroatom-Heteroatom Bonds Many molecules have heteroatom-heteroatom bonds (e.g., NdN in hydrazines or OdO in peroxides). The energy profile for hydrazine (see Fig. 1.16) is different from those of previous systems, containing two identical minima corresponding to arrangements in which the nitrogen lone pairs are perpendicular, and two different maxima. The higher maximum is 10.52 kcal (44 kJ) mol1 above the minima and corresponds to an arrangement in which both NH bonds eclipse and the two lone pairs eclipse. The other maximum is only 3.11 kcal (13 kJ) mol1 above the minima, and it is very broad. It corresponds to a range of conformers centered around a structure in which the two lone pairs are anti to each other. Unfavorable lone pair-lone pair interactions are much larger than bond-bond or bond-lone-pair interactions even if the lone pairs point away from each other. The best way to 54

Nakata, T.; Tanaka, T.; Oishi, T. Tetrahedron Lett. 1983, 26, 2653.

36

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

minimize them is to keep the lone pairs perpendicular. Note that the preference due to lone-pair interactions is much larger than preferences previously noted due to bond-bond and bond-lone-pair interactions. Hydrogen peroxide provides another example of the consequences that arises from the interaction of lone electron pairs. The energy curve (in Fig. 1.16) shows a pair of identical minima with torsional angles 120° and 240°, and it shows two different maxima. The higher value is 9.56 kcal (40 kJ) mol1 above the minima and corresponds to a syn conformer (HOOH angle ¼ 0°), while the lower value is 1.2 kcal (5 kJ) mol1 above the minima and corresponds to an anti conformer (HOOH angle ¼ 180°). In summary, hydrazine and hydrogen peroxide, (e.g., ethane, methylamine, and methanol) are both described in terms of a single structure. 40

45

rel. E (kJ/mol)

rel. E (kJ/mol)

30 30 Hydrazine

15

20

Hydrogen peroxide

10

0

0 0

FIG. 1.16

120 240 :NN: torsion angle

360

0

120 240 HOOH torsion angle

360

Energy barrier for substituted amines, alcohols, and ethers over a rotation of 360°.

1.5.1.5 Bonds Connecting sp3 and sp2 Hybrids: Propene and But-1-ene The planar sp2 carbon atoms in an alkene are relatively easy to distinguish, but what effect does the C]C unit have on the adjacent sp3]sp2 bond? A plot of energy versus the C]CdCdH torsional angle in prop-1-ene (A in Fig. 1.17) shows three identical minima and three identical maxima, just like that for ethane. Even the rotation barrier is similar [1.91 kcal (8 kJ) mol1 in prop-1-ene vs. 2.87 kcal (12 kJ) mol1 in ethane]. However, the methyl group in propene sees a different environment than the methyl group in ethane. If the vinylic CdH bond is staggered, then the C]C unit must eclipse and vice versa.

rel. E (kJ/mol)

A

9

6 Prop-1-ene

3

0 0

120

240

360

CCCH torsion angle

B

rel. E (kJ/mol)

12

8 But-1-ene

4

0 0

120

240

CCCC torsion angle

FIG. 1.17 Rotation of prop-1-ene and but-1-ene about 360°.

360

37

1.5 CONFORMATIONS

The minima in prop-1-ene correspond to a conformation in which one of the methyl CdH bonds will eclipse the carbon-carbon double bond, but two methyl carbons are staggered relative to the alkene CdH bond. The plot for the central carbon-carbon single bond in but-1-ene (B in Fig. 1.17) closely resembles the curve for n-butane. It shows three minima, two that are identical corresponding to arrangements in which a (methylene) CdH bond eclipses the carbon-carbon double bond, and the third only 0.72 kcal (3 kJ) mol1 higher in energy in which the CdC single bond eclipses the double bond. There is a general rule that single bonds eclipse double bonds.

FIG. 1.18

Rotation of acetic acid and methyl acetate about 360°.

Much larger conformational preferences involving single and double bonds can occur. The most conspicuous are the CdO bonds in carboxylic acids and carboxylic acid esters, between sp2 (carbonyl group) and sp3 (oxygen) centers. Energy curves for CdO bond rotation in acetic acid and in methyl acetate (A and B, respectively, in Fig. 1.18) both show two minima corresponding to syn (O]CCH and O]CCC ¼ 0°) and anti (O]CCH and O]CCC ¼ 180°) arrangements, and a pair of identical maxima in between. The syn conformer is preferred by nearly 14.34 kcal (60 kJ) mol1 for both molecules. In both conformers, C]O and OdH (OdMe) bonds eclipse, but the C]O and CdO bond dipoles subtract in the syn-conformer and they add in the anti-conformer (see A and B in Fig. 1.19).

O O Syn

Acetic acid

Anti

H

O O

Dipoles add A

FIG. 1.19

H

Dipoles subtract

B

Relationship of dipole moment and rotamers in acetic acid.

1.5.1.6 Bonds Connecting sp2 Hybrids: Buta-1.3-diene and Styrene A final case is presented with a focus on the carbon-carbon single bond connecting two sp2 hybridized atoms, such as those found in conjugated dienes or styrene derivatives. The conventional wisdom is that the two double bonds need to be coplanar in order to maximize “conjugation.” While trans planar buta-1,3-diene may represent an energy minimum, the corresponding cis planar conformer is not (see A in Fig. 1.20). There is an energy minimum nearby (CCCC torsion angle 40°) that is 2.87 kcal (12 kJ) mol1 higher than a cis form (an energy maximum), and 2.87 kcal (12 kJ) mol1 higher in energy than the trans form. The energy barrier to rotation about C2dC3 has been shown to be 5 kcal (21 kJ) mol1.55 These two rotamers arise by rotation about the C2dC3 single bond in buta-1,3-diene with the two π-bonds cis or with the two π-bonds trans. Rotamer s-cis-buta-1,3-diene is called the s-cis conformation, and s-trans-buta-1,3-diene is the s-trans-conformation. When the diene has one or more alkyl substituents, (E,E)-, (Z,E)-, and (E,Z)-dienes are possible. The s-cis conformation of s-cis-(Z,Z)-hexa-1,3-diene is even less favorable than the s-trans conformation [s-trans-(Z,Z)-hexa-1,3-diene], 55

Ref. 50c, p 54.

38

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

FIG. 1.20 Rotation of buta-1,3-diene and styrene about 360°.

when compared with buta-1,3-diene.56 In s-cis-(Z,Z)-hexa-1,3-diene, the Me-Me interaction is quite apparent, and although the s-trans conformer [s-trans-(Z,Z)-hexa-1,3-diene] has two methyl-hydrogen interactions, it is lower in energy than the Me-Me interaction in s-cis-(Z,Z)-hexa-1,3-diene. Similar expectations apply to styrene (see B in Fig. 1.20), where there is only one unique conformer. The vinyl and phenyl groups are nearly coplanar and the energy barrier through planar styrene is tiny. With the advent of coupling reactions (e.g., the Heck reaction, Section 18.4.1), such compounds have gained greater importance.

s-cis-Buta-1,3-diene

s-trans-Buta-1,3-diene

s-cis-(Z,Z)-Hexa-1,3-diene

s-trans-(Z,Z)-Hexa-1,3-diene

1.5.2 Conformations of Cyclic Molecules Steric and electronic interactions influence the rotation about single bonds within acyclic molecules, leading to a relatively small population of low-energy rotamers. Cyclic molecules have additional energy barriers to conformational mobility relative to acyclic molecules. The CdC bonds in cyclic compounds are not capable of complete rotation, but they undergo what is known as pseudorotation, largely by twisting and bending around the various carbon-carbon bonds in the ring. Depending on the flexibility of the ring, which is related to the size of the ring, many conformations are possible. As with acyclic compounds, there are usually a small number of low-energy conformations that dominate the population. When compared with acyclic molecules, an important difference is observed in cyclic molecules. Since 360° is impossible, deformation of the bond angles away from 109°280 as the molecule attempts to rotate around the CdC bonds leads to an increase in energy for that conformation. Bond distortion can be severe when the ring is small. Cyclopropane, for example, has a CdCdC bond angle of 60°, which induces great strain in the molecule. This type of strain is called Baeyer or angle strain, formally defined as the increase in energy of cyclic compounds that arises from the deformation of the optimum valence angle of 109°280 for sp3 carbon or 120° for sp2 carbon. In general, small rings have higher Baeyer strain when compared to larger rings. In addition, the strain inherent to a small ring is often observed in ring-opening reactions that release that strain energy as the ring is broken. Cyclopropane, for example, has great strain that leads to a higher ground-state energy. Formation of three-membered rings is common. The higher ground-state energy imposes an energy barrier to formation of such rings, but provides an “energy assist” for opening of those rings in a chemical reaction.55 56

See Squillacote, M. E.; Liang, F. J. Org. Chem. 2005, 70, 6564.

1.5 CONFORMATIONS

39

If a reaction occurs at cyclopropane, the CdC bond is significantly easier to break when compared to acyclic alkanes because the strain energy is released when the bond is broken. When forming a cyclopropane ring by cyclization, the strain energy is a major contributor to the Eact, forming an energy barrier to formation of that ring. Apart from strain energy, such reactivity is partly explained by deformation of σ-bonds in cyclopropane, which leads to significant p character (the hybridization is Csp2.3). The increased p character is reflected in the diminished electron density of the CdC bonds in cyclopropane, which is slightly displaced from linearity between the nuclei.57 Due to this type of hybridization, many ring-opening reactions of cyclopropane mimic the chemistry of alkenes. The four-carbon cyclic compound is cyclobutane with bond angles of 90°, and in planar cyclopentane with bond angles of 108°. Cyclobutane has significant Baeyer strain, less than in cyclopropane, but more than in cyclopentane.

Cyclopropane

Planar cyclopentane

Planar cyclobutane

Puckered cyclobutane

Cyclopropane, planar cyclobutane, and planar cyclopentane exhibit a second major type of strain, which is similar to that observed in acyclic molecules. All CdH bonds in these planar molecules are eclipsed, with severe nonbonded interactions of the eclipsed hydrogen atoms and electronic repulsion of the eclipsed bonds. Rings with this type of strain are said to have Pitzer or bond-opposition strain, which is defined as the increase in energy for a compound arising when adjacent bonds are eclipsed, bringing the attached atoms into close spatial proximity. Cyclic molecules generally do not exist in the planar form due to the Pitzer strain. Although rotation of 360° about the CdC bonds is not possible, excess energy is dissipated by twisting and bending, which leads to partial rotation (pseudorotation) that will minimize Pitzer strain.57 One or two low-energy conformations are usually drawn to represent the entire molecule, although as the ring becomes larger and the flexibility increases more low-energy conformations are possible. Baeyer strain is relatively small in cyclobutane relative to planar cyclobutane,58 and pseudorotation leads to a bent or puckered conformation as the low energy conformation.59

Envelope cyclopentane

Planar cyclohexane

Chair cyclohexane

Chair cycloheptane

57

(a) Bernett, W. A. J. Chem. Educ. 1967, 44, 17. (b) de Meijere, A. Angew. Chem. Int. Ed. Engl. 1979, 18, 809. (c) Ref. 10, pp 676–678. (d) Ref. 29b, pp 204–306 and 124–179.

58

(a) Dunitz, J. D.; Schomaker, V. J. Chem. Phys. 1952, 20, 1703. (b) Rathjens, G. W., Jr.; Freeman, N. K.; Gwinn, W. D.; Pitzer, K. S. J. Am. Chem. Soc. 1953, 75, 5634.

59

(a) Ref. 10, p 676. (b) Ref. 29b, p 248.

40

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

Pseudorotation in cyclopentane leads to a low-energy conformation, the so-called envelope shape.59 Although Baeyer strain is increased in envelope cyclopentane relative to planar cyclopentane,60 the decrease in Pitzer strain more than compensates, and envelope cyclopentane is the low-energy conformation. This observation suggests that Pitzer strain is a greater contributor to the conformational instability of planar cyclopentane than the contribution of Baeyer strain. Analysis of planar cyclohexane reveals extensive Pitzer strain and the bond angles would be 120°, introducing Baeyer strain. Pseudorotation leads to the familiar chair conformation as the low-energy conformation. Chair cyclohexane has virtually no Baeyer strain with bond angles of 109°280 . Cyclohexane also has a so-called boat conformation, as well as a higher energy “twist-boat” conformation. There is the potential for a so-called half-chair conformation as well. In the boat conformation, there is a cross ring (transannular) interaction of the 1,4-hydrogen atoms (referred to as flagpole hydrogen atoms). In other words, these hydrogen atoms are not on adjacent carbon atoms and the interaction is essentially a through-space interaction, so it is a type of transannular strain. Cycloheptane has similar bond angles,61 but is more flexible, and there are several low-energy conformations, although only chair cycloheptane is shown. Cycloheptane also has a boat form that is very close in energy to a chair form.62 The lowest energy form of cycloheptane, however, is the twist-chair conformation.63 H

H

H

H

Boat cyclohexane

Twist–boat cyclohexane

Half-chair cyclohexane

Twist–chair cycloheptane

Boat cycloheptane

As mentioned, a “chair” conformation is favored for cyclohexane, in which all six carbons are equivalent. The hydrogen atoms divide into six equatorial hydrogen atoms and six axial hydrogen atoms, marked in Fig. 1.21. There is an alternating pattern of three axial and three equatorial hydrogen atoms on each side of the ring, which is top and bottom as it is drawn. However, the fact that only one proton resonance is seen in the room temperature 1H NMR spectrum of cyclohexane suggests a rapid conformation change in which equatorial and axial hydrogen atoms exchange. An energy profile for the curve (Fig.1.21) shows the two identical chair structures as the starting and ending points. The curve contains three energy maxima. Two are identical “half-chairs” that connect the chair and twist-boat forms, while the boat is the lower energy maximum peak that appears to be linked to two twist-boat forms. Half–chairs ax

ax

ax eq

eq

Boat

ax

ax

eq eq

ax

eq eq

eq

eq eq

eq eq ax

eq ax

ax

Chair Chair cyclohexane

FIG. 1.21

ax

Twist–boats

ax

Conformational mobility of cyclohexane.

60

Kilpatrick, J. E.; Pitzer, K. S.. Spitzer, R. J. Am. Chem. Soc. 1947, 69, 2483.

61

(a) Ref. 10, pp 686–689; (b) Ref. 29b, p 252.

62

Allinger, N. L. J. Am. Chem. Soc. 1959, 81, 5727.

63

Wiberg, K. B. J. Org. Chem. 2003, 68, 9322.

ax

Chair Chair cyclohexane

41

1.5 CONFORMATIONS

The chair cyclohexane is 6.69 kcal (28 kJ) mol1 lower in energy than the twist-boat (or just twist) conformation, suggesting that the latter will have little influence on the properties of cyclohexane. The interconversion of chair cyclohexane into the twist-boat form can be viewed as a restricted rotation (pseudorotation) about the ring bonds. Correspondingly, the interconversion of the twist-boat intermediate into the other chair form can be viewed as rotation about the opposite ring bond. Pseudorotation

The various types of strain associated with cyclic molecules plays a role in chemical reactions that form those rings. The relative energy of each conformation is determined primarily by relief of Baeyer and Pitzer strain, which contribute to the energy required for intramolecular cyclization reactions. In a cyclization reaction, two reactive ends of a molecule come together to form a ring. In the transition state of that cyclization, the molecule will assume the conformation of the ring being formed. Therefore, as two reactive ends of the acyclic fragment come together, the strain inherent to the ring product becomes important (see Section 4.5). Strain in this transition state is important for the formation of the cyclic product. Fig. 1.22A shows the relative reactivity for formation of lactones by cyclization of ω-hydroxy acids of ring size C3–C17.64 A reactivity maximum appears at C5 and a reactivity minima at C3 and C8.64 Fig. 1.22B shows the enthalpy (ΔH) of several cyclic alkanes, with a minimum at C6 and a maximum at C9.65 There are two other important features of Fig. 1.22B. Rings of C8–C13 members are significantly higher in energy than the small rings (C3 ! C7) or the large rings (C14). These medium-size rings are extremely difficult to form using intramolecular cyclization.64 The other feature of Fig. 1.22B is the higher energy required to form rings with an odd number of atoms when compared to the energy required to form rings with an even number of atoms. Compare cyclohexane and cycloheptane to see the

2

1

0

–1

ΔH #(kcal/mol)

log kintra

22

–2

20

18

16 –3 inter

14 –4 3

(A)

5

7

9

11 13 15 17 19 21 23 Ring size

5

(B)

10

15 Ring size

20

FIG. 1.22

(A) Reactivity profile for lactone formation. (B) Enthalpy of cycloalkanes [CnH2n]. Reprinted with permission from Illuminati, G.; Mandolini, L. Acc. Chem. Res. 1981, 14, 95. Copyright © 1981 American Chemical Society. 64 65

Illuminati, G.; Mandolini, L. Acc. Chem. Res. 1981, 14, 95. Also see Ref. 10, p 680.

Also see (a) Dunitz, J. D.; Prelog, V. P. Angew. Chem. 1960, 72, 896. (b) Prelog, V. In Perspectives in Organic Chemistry; Todd, A. R., Ed.; Interscience: New York, NY, 1956; p 96.

42

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

difference. Cycloheptane has an odd carbon that is not easily accommodated by the low-energy chair (all gauche) form, and its presence leads to an increase in Pitzer strain.64,65 Similar effects are seen for all odd carbon rings. As the ring size increases to eight members and higher, increased flexibility leads to an increasing number of lower energy conformations. Indeed, an inspection of cyclooctane reveals there are several conformations, including crown cyclooctane, boat-chair cyclooctane, boat-boat cyclooctane,62,66 as well as twist-boat-chair and twist chair-chair conformations.67 These latter conformations are higher in energy than the crown cyclooctane.68 Wiberg63 reported calculations that show the boat-chair to be the lowest energy conformer, with the twist-boat-chair, twist-chair-chair and the crown as the next higher energy conformations. The boat-boat conformation is relatively high in energy.62 A crown conformation is observed for cyclodecane (crown cyclodecane),65b,69 which conforms to the same extended chair conformation found in diamond (see Fig. 1.23). 7 6

5

8

3 4

1

2

Crown cyclooctane

Boat–chair cyclooctane

Boat–boat cyclooctane

Crown cyclodecane

The fact that crown-cyclodecane62 fits on this diamond lattice66 is usually taken to be a demonstration of stability, and indeed it is an important low-energy conformation. The odd-membered ring in cyclononane shows a twist, also found in cycloheptane when compared to chair cyclohexane58 and a slight increase in Pitzer strain. The twist in crownlike cyclononane makes it unable to superimpose on the diamond lattice. Even large even-membered rings (e.g., cyclooctadecane) can show this chair-like or crown-like conformation in a low energy form, but not oddmembered rings. Cyclodecane

Cycloundecane

FIG. 1.23 Cyclodecane and cycloundecane on a diamond lattice.

Another type of steric strain is found in C8–C13 cyclic compounds. This interaction is best observed in models of cyclopentane-cyclooctadecane that look down on the top of each ring, where a cavity in the center of each ring is clearly visible. If the conformation of a ring brings atoms in close proximity, within the cavity of the ring, this is known as a transannular interaction.64 Hydrogen atoms do not intrude into the internal cavities of envelope cyclopentane or chair cyclohexane, so there are no transannular interactions.

Envelope cyclopentane

Chair cyclohexane

Chair cycloheptane

In cycloheptane, two of the hydrogen atoms can move into positions that are slightly within the cavity, contributing to modest transannular strain and partially accounting for the higher energy of cycloheptane. In crown-cyclooctane, pseudorotation moves hydrogen atoms that are on opposite sides of the ring into the cavity, contributing to a significant transannular interaction, and a significant increase in the inherent energy for that ring. There is also significant transannular strain in cyclononane and crown-cyclodecane. As the ring size increases, the cavity becomes larger, and 66

Bellis, H. E.; Slowinski, E. J. Spectrochim. Acta 1959, 15, 1103.

67

Pauncz, R.; Ginsburg, D. Tetrahedron 1960, 9, 40.

68

(a) Ref. 10, pp 765–766. (b) Ref. 29b, p 253. (c) Allinger, N.L.; Hu, S. J. Am. Chem. Soc. 1961, 83, 1664.

43

1.5 CONFORMATIONS

the net energy of the ring decreases after reaching a maximum at cyclodecane. Although the potential for a transannular interaction is quite high in cyclooctadecane, the large cavity and increased flexibility allow the transannular hydrogen atoms to sweep past each other upon pseudorotation. The cavity inside the cyclooctadecane ring is large enough to accommodate these atoms without generating a large energy gradient.

Crown cyclooctane

Crown cyclodecane

Cyclooctadecane

Cyclononane

If a ring has unfavorable dihedral angles,69 there is another type of strain termed I-strain (internal strain).70 Formally, I-strain involves the change in strain when going from a tetrahedral to a trigonal carbon or vice versa (as in oxidizing an alcohol to a ketone; Section 6.2). In rings where the tetrahedral bond angles bring the transannular substituents into close proximity, conversion to a trigonal-planar carbon relieves this interaction somewhat and lowers the energy of the system.71 An example is the oxidation of cyclooctanol to cyclooctanone (see Section 6.2). The reverse process (the reduction, see Sections 7.4, 7.6.1, and 7.10.4: cyclooctanone ! cyclooctanol) is more difficult due to the increased I-strain, as well as increased transannular strain in the alcohol.72

OH

O

CrO3 , H+

L

L

H Cyclooctanol

Cyclooctanone

The presence of a heteroatom, a C]C unit, or a carbonyl influences the conformation of medium-size rings, but the changes are often subtle. The calculated low-energy conformation of cyclooctane is compared with those of cis-cyclooctene, the ether oxocane, and cyclooctanone. Although cis-cyclooctene is somewhat flattened, the conformations of the other three eight-membered rings are rather similar.73 69

Prelog, V. J. Chem. Soc. 1950, 420.

70

(a) Brown, H. C.; Fletcher, R. S.; Johannesen, R. B. J. Am. Chem. Soc. 1951, 73, 212. (b) Brown, H. C.; Borkowski, M. J. Am. Chem. Soc. 1952, 74, 1894.

71

Prelog, V. Bull. Chim. Soc. Fr. 1960, 1433.

72

Ref. 29b, pp 266–269.

73

Pawar, D. M.; Moody, E. M.; Noe, E. A. J. Org. Chem. 1999, 64, 4586.

44

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

•

•

•

• •

•

•

•

•

•

•

• •

Cyclooctane

•

O

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

O

Oxocane

cis-Cyclooctene

Cyclooctanone

Estimating the low-energy conformations of large ring compounds (macrocycles) is more complicated and arguably more difficult than a similar analysis of the compounds discussed in Sections 1.5.1 and 1.5.2. Since this class of compounds includes macrolide antibiotics,74a important commercial products (e.g., muscone), and also crown ethers,74b there is a great deal of interest in macrocycles.74a Both the chemical and physical properties of macrocycles depend on the conformations of the large ring.75 An analysis of conformations of small- and medium-ring compounds was reported that used an analysis of the sign of torsional angles,76 and was combined with the use of dihedral maps.77 Clyne and Weiler78 produced a polar map analysis of macrocyclic compounds that yielded conformational information. This method has also been applied to the conformational analysis of macrocyclic ethers.79 Another way to represent this conformation is based on Dale’s numerical system describing the number of bonds found between corner atoms.80c Dale’s system is sometimes called wedge notation.81 Unfortunately, calculations can be difficult, and the method does not seem to be widely used. The system is based on the idea that even-membered rings can exist in four quadrangular (four-cornered system) conformations (see 51). (1) All four sides contain an odd number of bonds (a–d ¼ odd). (2) Two adjacent sides are odd with the others even (ab odd, cd even; or bc odd, ad even). (3) Two adjacent sides are odd with the others even (ac odd, bd even; or bd odd, ac even). (4) All sides contain an even number of bonds (a ! d ¼ even). For oddmembered rings, the best conformation arises when there are three (a triangular system) or five corners (a quinguangular system). When there are three corners, all three sets can be odd, or two can be even with one odd and one side must have one or more convex faces. For five corners, all can be odd, two adjacent or next to adjacent can be even with the other three odd, or one can be odd and four even, so there must be one or more concave faces. –80

+80

°

b

°

°

–55

+55 •

•

° °

°

c

a

–155

°

°

+155

d

+55 °

51

+80

–55 °

–80

52

3

10

°

° –135

13 +55

+55

•

•

–135 +55 ° ° °

+55

° +55

°

° +55

–135 6

° 53

Cyclotetradecane

The actual notation is illustrated for the diamond form of cyclodecane, (52). The dihedral angles were measured from a Dreiding model™, although such measurements can now be done with appropriate computer software, and the corners were determined for a given conformation. For 52, the angles for the four corners are +55/+80 and 55/80. This method takes the number of bonds on each side, starting with the short side and gives the lowest

74

(a) Paterson, I.; Mansuri, M. M. Tetrahedron 1985, 41, 3568. (b) Hayward, R. C. Chem. Soc. Rev. 1983, 12, 285. (c) Still, W. C.; Galynker, I. Tetrahedron 1981, 37, 3981 and references cited therein. 75

(a) Dale, J. Angew. Chem. Int. Ed. Engl. 1966, 5, 1000. (b) Idem Top Stereochem. 1976, 9, 199. (c) Idem Acta Chem. Scand. 1973, 27, 1115.

76

(a) Bucourt, R. Top Stereochem. 1974, 8, 159. (b) DeClerq, P. J. Tetrahedron 1984, 40, 3729. (c) Idem Tetrahedron 1981, 37, 4277. (d) Toromanoff, E. Tetrahedron 1980, 36, 2809.

77

Ogura, H.; Furuhata, K.; Harada, Y.; Iitaka, Y. J. Am. Chem. Soc. 1978, 100, 6733.

78

(a) Ounsworth, J. P.; Weiler, L. J. Chem. Educ. 1987, 64, 568. (b) Keller, T. H.; Neeland, E. G.; Rettig, S.; Trotter, J.; Weiler, L. J. Am. Chem. Soc. 1988, 110, 7858.

79

Clyne, D. S.; Weiler, L. Tetrahedron 2000, 56, 1281.

80

(a) Dale, J. Acta Chem. Scand. 1973, 27, 1115, 1130. (b) Dale, J. Acta Chem. Scand. 1973, 27, 1149. (c) Bj€ ornstad, S. L.; Borgen, G.; Dale, J.; Gaupset, G. Acta Chem. Scand. Ser. B 1975, B29, 320.

81

Ref. 12, pp 763–764, 766–769.

45

1.5 CONFORMATIONS

possible combination of numbers. For 52, there are 2, 3, 2, and 3 bonds, respectively, and this is [2323]-decane. Nonane has an odd number of atoms, and has three corners. Conformation 53 represents nonane and it has three corners with angles determined to be +55/+55, leading to the notation [333]-nonane. For cyclotetradecane, the lowest energy conformation is shown, the [3434]-conformation. When substituents are attached to large rings, they tend to occupy only exterior positions to minimize the large transannular interactions.82 A fully substituted atom of a macrocycle will usually occupy a corner position since this is the only position where it does not cause severe transannular strain.80 For both substituted and unsubstituted macrocycles, the conformation of the ring will impose restraints on the torsional angles available to each unit of carbons (four carbons define a torsional angle). The use of polar maps of torsional angles may be of value for determining the conformation of many macrocyclic compounds. The 14-membered 3-keto lactones, for example, have been synthesized and their conformations analyzed using this technique.83 This ability to predict conformational bias is important to the stereochemical outcome of chemical reactions involving macrocyclic rings.

1.5.3 A1,3-Strain and G-Strain The preceding sections focused on one or two low-energy conformations of cyclic compounds, but in fact several conformations for each ring may lie close together in energy. To obtain a complete picture several conformations of equal, close, or higher energy must be considered. The conformational mobility of cyclohexane includes, for example, the two low-energy chair conformations, six degenerate twist-boat conformations, and six degenerate boat conformations.84,85 If all the energies for the various conformations are known, it should be possible to estimate the relative percentage of each. Since the chair conformation usually constitutes the majority of this conformer population, a typical assumption is usually made to ignore the other conformations for an initial estimation. Eliel et al.86 estimated the energy difference between the boat and chair forms of cyclohexane to be 4 kcal (16 kJ) mol1, which means that “only one molecule in 1000 will be in the boat form” at 298 K (25°C). Note that the boat form exists in certain bridged cyclohexane molecules (e.g., trans-1,2-di-tert-butylcyclohexane), which contains 12% of the boat conformation.86,87 H

H H

H H

H trans-1,2-Di-tert-butylcyclohexane Boat

Diaxial chair

Diequatorial chair

The discussions of cyclohexane given in Section 1.5.2 made it clear that the chair conformations are usually lower in energy. The boat and two chair conformations are shown for trans-1,2-di-tert-butylcyclohexane. It is clear that there is a steric interaction in the boat because the two tert-butyl groups are in close proximity, whereas in the diaxial chair those groups are on opposite sides of the ring. Interestingly, the interaction of the tert-butyl groups is somewhat relieved in the twist-boat conformation, which exists alongside the more stable diequatorial chair,86 but the structures shown indicate that the diequatorial conformation is lower in energy than the diaxial conformation. Why is the diaxial chair higher in energy than the diequatorial chair, and why is the low-energy form the boat conformation or the twist-boat conformation mentioned? The answer reveals two major sources of strain in chair cyclohexanes, which have both axial and equatorial bonds. In cyclohexane, this interaction of the axial hydrogen atoms on either side of the ring is minimal

82

Dale, J. Stereochemistry and Conformational Analysis; Verlag-Chemie: New York, NY, 1978.

83

Neeland, E. G.; Ounsworth, J. P.; Sims, R. J.; Weiler, L. J. Org. Chem. 1994, 59, 7383.

84

(a) Lowry, T. H.; Richardson, K. S. Mechanism and Theory in Organic Chemistry; 3rd ed.; Harper & Row: New York, NY, 1987; p 141. (b) Squillacote, M.; Sheriden, R. S.; Chapman, O. L.; Anet, F. A. L. J. Am. Chem. Soc. 1975, 97, 3244. (c) Hirsch, J. A. Concepts in Theoretical Organic Chemistry; Allyn and Bacon: Boston, 1974; pp 249–252. 85

Barton, D. H. R. Experientia 1950, 6, 316.

86

Eliel, E. L.; Wilen, S. H.; Mander L. N. Stereochemistry of Organic Compounds: John Wiley & Sons: New York, NY, 1994; pp 688–690.

87

Allinger, N. L.; Freiberg, L. A. J. Am. Chem. Soc. 1960, 82, 2393.

46

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

because hydrogen atoms are relatively small, as seen in the space-filling model shown in Fig. 1.24 for chair cyclohexane. Replacing the hydrogen atoms at the C1, C3, and C5 axial positions with substituents leads to conformation in which there is a competition for the space on that side of the ring, giving rise to a different type of transannular interaction. This transannular interaction is known as A-strain, or A1,3-strain, or a 1,3-diaxial interaction (see Fig. 1.24). When one hydrogen atom is replaced with a methyl group (chair methylcyclohexane), the interaction between methyl at C1 and the hydrogen atoms at C3 and C5 is greater, leading to the transannular interaction shown. This interaction is much higher in energy relative to the hydrogen-hydrogen interactions at C1, C3, and C5 in chair cyclohexane. The methyl-hydrogen interaction is A1,3-strain. In chair 1,3-dimethylcyclohexane and in chair 1,3,5-trimethylcyclohexane, A1,3-strain increases significantly as the size of the substituent increases. There are two chair cyclohexane conformations that are in equilibrium, and A1,3-strain destabilizes the conformation with more axial substituents, shifting the equilibrium to the lower energy chair form that has the most substituents in the equatorial position and the fewest in the axial position. Minimum transannular interaction

Chair cyclohexane

A1,3-strain

Chair 1,3-dimethylcyclohexane

Chair methylcyclohexane

Chair 1,3,5-trimethylcyclohexane

1,3

FIG. 1.24 The A -strain in cyclohexane, methylcyclohexanes.

Returning to diaxial di-tert-butylcyclohexane, each axial tert-butyl group will interact with the two axial hydrogen atoms on that side of the ring, leading to severe A1,3-strain that is estimated to be 12 kcal (50.2 kJ) mol1 or 6 kcal (25.1 kJ) mol1 per tert-butyl group.88 There is no question that in most substituted cyclohexane derivatives, the equilibrium for the two chair conformations will shift to favor the chair with the substituents in the equatorial position, minimizing or eliminating A1,3-strain. Closer inspection of diequatorial di-tert-butylcyclohexane reveals that A1,3-strain has been eliminated, but the 1,2diequatorial tert-butyl groups are in close proximity, producing a new type of strain called G-strain. The energy of this interaction (the G-value)89 has been measured to be 2.5 kcal (10.47 kJ) mol1 per tert-butyl group (see Fig. 1.25). This type of steric interaction is analogous to the interactions found in the gauche butane conformation, and for that reason is often referred to as a gauche interaction (G-strain). If cyclohexane is taken as a model, the adjacent trans-diequatorial hydrogen atoms for chair cyclohexane in Fig. 1.25 have little or no interaction, but replacing them with methyl (see chair trans-1,2-dimethylcyclohexane) leads to

88

Winstein, S.; Holness, N. J. J. Am. Chem. Soc. 1955, 77, 5562.

89

Corey, E. J.; Feiner, N. F. J. Org. Chem. 1980, 45, 765.

47

1.5 CONFORMATIONS

G-strain

G-strain

Adjacent diequatorial hydrogens

Chair cyclohexane

FIG. 1.25

Chair cyclohexane

Chair trans-1,2-dimethyl cyclohexane

Chair trans-1,2-di-tertbutylcyclohexane

The G-strain in chair 1,2-dimethylcyclohexane and chair 1,2-di-tert-butylcyclohexane.

G-strain. The interaction is very large when the groups are trans-diequatorial, as found with the two tert-butyl groups in chair di-tert-butylcyclohexane, and the significant G-strain destabilizes that conformation. If both chair conformations have high-energy interactions, the molecule will distort to minimizes these interactions, and 1,2-di-tert-butylcyclohexane exists largely in either the boat or the twist-boat. It will be apparent throughout this book that the conformation of a molecule has an important influence on the reactivity and stereochemistry of many reactions. Strain energies that influence the conformation of a molecule can be quantified in some cases, and the percentage of each conformation can be calculated. With such information, better predictions of reactivity and stereochemistry are possible. The following protocols are presented in an attempt to reinforce the idea that conformational analysis is a critical part of synthetic analysis, and to introduce the most basic approaches to conformational analysis. This fundamental idea is best illustrated by focusing on substituted cyclohexane derivatives. When a reagent approaches a cyclohexane ring, the conformation of the ring will influence how the reagent interacts with any functional group on that ring. Indeed, the populations of the two chair conformations influence the relative rate and the stereochemical outcome of a given reaction. The relative populations of both chair conformations in cyclohexane derivatives can be predicted with some accuracy by calculating the A- and/or G-values. For a monosubstituted cyclohexane, two chair forms are possible (54 and 55) and they are in a dynamic equilibrium. If X is an alkyl group or a heteroatom substituent, the diaxial steric interaction (A-strain) in 54 will be larger than in 55, and a higher percentage of 55 is expected at equilibrium. This nonbonded interaction can be measured in terms of F° (H°), the conformational free energy.90 Since there are two conformations of differing energies, this difference is represented by (ΔF° or ΔH°),84a and it is possible to relate this energy term to the free energy of the system by: ΔG° ¼ ΔH°  TΔS°. For a given temperature, the entropy term will be small relative to ΔF° (taken to be ΔH° hereafter) and a second assumption can be made, that ΔS° ¼ 0, leading to ΔG°  ΔH°. H

X H

H

H

Keq

H

X 54

55

The ΔS° term is not zero, but is usually measured in calories (or joules) and the ΔH° term in kilocalories (kJ), so ignoring ΔS° will introduce only a small error into the calculation. Lowry and Richardson’s91 list of ΔH° and S°298 values for many bonds demonstrates this difference. In the CdMe bond, for example, ΔH° is 10.08 kcal (42.2 kJ) mol1 and S°298 is 30.4 cal (127.2 J) mol1. Similarly, ΔH° for cyclopropane is listed as 27.6 kcal (115.5 kJ) mol1 and S°298 is 32.1 cal (134.3 J) mol1.91 The ΔH° term is related to ΔG°, the free energy, which is related to the equilibrium constant by the expression: ΔG° ¼  2.303 RT log Keq. The conformational equilibrium between 54 and 55 is represented by Keq, which can be calculated from the percentage of the two conformers: Keq ¼ (%54)/(%55), where it is assumed that the %54 + %55¼ 100%. Note that a general discussion of the Curtin-Hammett principle,90c–f which assumes that the product composition is related to the relative concentrations of the conformers,90 may be of value in any discussion of this equilibrium reaction. For this process to work, some estimation of F° (hereafter called H°) must be made. In connection with the development of the interactive synthesis computer program Logic and Heuristics Applied to Synthetic Analysis (LHASA, 90

(a) Eliel, E. L. J. Chem. Educ. 1960, 37, 126. (b) Ref. 10, pp 694–698. (c) Ref. 29b, pp 234–239. (d) Ref. 84a, pp 138–140. (e) Hammett, L. P. J. Am. Chem. Soc. 1937, 59, 96. (f ) Hammett, L. P. Physical Organic Chemistry; 2nd ed.; McGraw-Hill: New York, NY, 1970; p 347ff. (g) Jaffe, H. H. Chem. Rev. 1953, 53, 191. (h) Hansch, C.; Leo, A. J. Substituent Constants for Correlation Analysis in Chemistry and Biology; John Wiley: New York, NY, 1979. (i) Ref. 10, p 654. 91

Ref. 84a, pp 164–165.

48

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

Section 8.4.2),1,3 Corey and Feiner89 developed a protocol for determining the conformational energies of cyclohexane derivatives, and assembled a list of H-values for various substituent interactions, which is shown in Table 1.4.89 This table is related to a similar one, first assembled by Hirsch and coworkers.92,93 TABLE 1.4 Corey’s A-, G-, and U-Values for Calculating Conformational Energies of Cyclohexane Derivatives X

X X1

HH G

A Group H

X1

A (kcal

mol–1)a

0

X1

XH

or

X X2

U G (kcal mol–1)a

U (kcal mol–1)a

0

0

0

0

F

0.2 (0.84)

Cl

0.4 (1.67)

0.5 (2.09)

0.4 (1.67)

Br

0.4 (1.67)

0.8 (3.35)

0.4 (1.67)

I

0.4 (1.67)

1.0 (4.19)

0.4 (1.67)

PR3

1.6

(6.7)

1.6 (6.7)

1.6

SR

0.8 (3.35)

0.5 (2.09)

0.8 (3.35)

S(O)R

1.9 (7.95)

2.7 (11.30)

1.9 (7.95)

S(O2)R

2.5 (10.47)

3.5 (14.65)

2.5 (10.47)

OR

0.8 (3.35)

0.2 (0.84)

0.8 (3.35)

NH3+

2.0 (8.37)

0.5 (2.09)

2.0 (8.37)

NR3+

2.1 (8.79)

0.5 (2.09)

2.1 (8.79)

NHR

1.3 (5.44)

0.3 (1.26)

1.3 (5.44)

NK

0.5 (2.09)

0.1 (0.42)

0.5 (2.09)

NL

0.2 (0.84)

0.1 (0.42)

1.2 (5.02)

NO2

1.1 (4.61)

0.3 (1.26)

0.5 (2.09)

CL

0.2 (0.84)

Aryl

3.0 (12.56)

1.2 (5.02)

0.5 (2.09)

CO2–

2.0 (8.37)

0.5 (2.09)

2.0 (8.37)

CHO

0.8 (3.35)

0.3 (1.26)

0.8 (3.35)

CK

1.3 (5.44)

0.2 (0.84)

0.9 (3.77)

CR3

6.0 (25.11)

2.5 (10.47)

6.0 (25.11)

CHR2

2.1 (8.79)

0.8 (3.35)

2.1 (8.79)

CH2R

1.8 (7.54)

0.4 91.67)

1.8 (7.54)

a

0

(6.7)

1.2 (5.02)

The value in parentheses is the energy in kJ mol–1

Reprinted with permission from Corey, E.J.; Feiner, N.F. J. Org. Chem. 1980, 45, 765. Copyright © 1980 American Chemical Society.

Using data from Table 1.4, the energy for a given conformation is taken to be the sums of all monoaxial interactions (A-values), any 1,2-diequatorial interactions (G-values) and any multiple diaxial interactions (defined by Corey as U-values).84 Simple doubling or tripling of the appropriate A-value did not give correct results in this latter case

92 93

Hirsch, J. A. In Topics in Stereochemistry; Allinger, N. L., Eliel, E. L., Eds.; Wiley–Interscience: New York, NY, 1967; Vol. 1, pp 199–222.

(a) Ref. 29b, pp 236–237. (b) Barrett, J. W.; Linstead, R. P. J. Chem. Soc. 1936, 611. For a list of energy values for several functional groups, see Ref. 10, pp 696–697.

49

1.5 CONFORMATIONS

and the more accurate U-values were required. Note that the A- and U-values are identical in many cases. They usually differ if the substituent contains unsaturation. OH THF

CH3

HO Cl

Cl CH3

56

H3C

CH3 57

The protocol for calculating the percentage of each conformation is illustrated by the two chair conformations of (1R,2R,3S,4R)-2-chloro-3,4-dimethylcyclohexan-1-ol, 56 and 57. Using Table 1.4, the pertinent interactions are UOR, the two UCH2R and UCl in 56, and the two GCH2R, GOR and GCl in 57. ΔH° ¼ Hproduct  Hreactant ¼ H57  H56 ¼ (GCH2R + GCl + GCH2R + GOR)  (UOR + UCH2R + UCH2R + UCl) ¼ (0.4 + 0.5 + 0.4 + 0.2)  (0.8 + 1.8 + 1.8 + 0.4) ¼ (1.5)  (4.8) ¼  3.3 kcal (13.8 kJ) mol1. If ΔH°  ΔG°, then assume ΔG° ¼  3.3 ¼  2.303 RT log Keq. At 25°C, 2.303 RT ¼  1.364. Therefore, ΔG° ¼  3.3 ¼  1.364 log Keq so log Keq ¼  3.3/1.364 ¼ 2.42 and Keq ¼ 102.42 ¼ 262.6. Since %56 + %57 ¼ 100%, where 57 ¼ 1  56, and if Keq ¼ (57)/(56) then Keq ¼ (1  57)/(56) and Keq (56) ¼ 1  56); so, 262.6 (55) ¼ 1  (56), and 262.6 (56) + (56) ¼ 1 or 262.6 + 1 (56) ¼ 1, which leads to (56) (263.6) ¼ 1 and (56) ¼ 1/263.6 ¼ 0.0038, where (57) ¼ 100  0.38%. Therefore, there is 0.38% of 56 and 99.62% of 57. Remember that two major assumptions were made for this calculation: (1) all conformations except 56 and 57 were ignored, and (2) ΔS° was assumed to be zero. As first shown by Barton,85 the relative percentage of a given conformation directly influences the rate of a reaction. As summarized by Eliel,90a the equilibrium between 54 and 55 is given by the equilibrium constant (K), and the rate of conversion of each conformation to product is given by k54 and k55, as developed by Winstein and Holness88 and Eliel.90a The overall rate will be Rate ¼ k [C], where k is the observed specific rate and [C] is the stoichiometric concentration of the cyclohexane.90a Rate ¼ k54 [54] + k55 [55] for this system, and the value k is related to the equilibrium constant as shown in the following: k54 - k X

Keq =

Keq

k54

k55

Product (P)

Product (P) X 54

and k =

55

k - k55 k55 + k54 K + 1

The percentage of a given conformation is an essential factor in determining the rate and viability of a given reaction and these calculations allow the synthetic chemist to estimate conformational populations. Just as the minor chair conformation influences the rate, so other conformations (boat, twist-boat, etc.) may exert an influence, depending on their relative populations. In addition to the values in Table 1.4, used for calculations of substituted cyclohexane derivatives,83 Corey and Feiner89 developed formulas for several other cyclic molecules. Incorporation of two halogens with a 1,2- or 1,4-diaxial relationship on a cyclohexane ring led to a higher proportion of the diaxial conformer94 due to the more favorable electrostatic interactions of the lone-pair electrons.95 Corey’s models in this case were 58 and 59, and the equation used to calculate the conformational energy is E ¼ 1⁄2 ðAX + AY Þ. Introduction of a π-bond or a heteroatom in the cyclohexane ring changes the energy relationships, since the presence of a π-bond flattened the ring. The usual U- and G-values were used for monosubstituted or disubstituted derivatives (e.g., 60 or 61), respectively. The presence of a carbonyl or an aldehyde unit led to a diminished value for A, and the presence of a spirocyclic or fused ring structure also diminished A. Introduction of an oxygen in the ring (to form a pyran) led to a diminished value for A.

94

Jensen, F. R.; Bushweller, C. H. Adv. Alicyclic Chem. 1971, 3, 139.

95

Wood, G.; Woo, E. P.; Miskow, M. H. Can. J. Chem. 1969, 47, 429.

50

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

X

X And Y

58

59

Y

R1 R

R

R

R1 60

R

61

Corey and Feiner89 also developed parameters to evaluate boat conformations, which are often important contributors to the conformational population of cyclohexane derivatives (see Section 1.5.2). The energies for boat conformations with one or more substituents in the flagpole positions use a new term (b): E ¼ (b1 + b2) AR for 62, and E ¼ (b1 + b2) (UR + UR0 ) for 63, derived from the bowsprit-flagpole interactions (RdH in 62 or R1dR in 63). Introduction of a π-bond, a cyclopropane, or a cyclobutane ring flattens the boat, leading to increased destabilization. Fused five-membered rings and larger fused rings do not flatten the ring of interest. R

H

R

R1

b1

b1

Me

b2

b2

62

63

Me Me L

Br

(3S,6S)-3-Bromo-6-isopropyl1-methylcyclohex-1-ene

64A

64B

These structural types emphasize that the presence of π-bonds and heteroatoms will influence the overall conformation. “Flattening” of a ring is often an important structural feature of natural products and other attractive synthetic targets. In (3S,6S)-3-bromo-6-isopropyl-1-methylcyclohex-1-ene, the π-bond in a cyclohexene moiety flattens the ring in 64A and its other low-energy conformer, 64B. These are actually half-chair conformations analogous to a half-chair cyclohexane (see Section 1.5.2).96 The isopropyl group (i-Pr) is less stable in the pseudo-axial position and more stable in the pseudo-equatorial position.97 The i-Pr $ Br interaction is largely alleviated in 64B. This effect is compatible with 64B as the major conformation and higher relative energy for 64A. Similar conformational effects arise in other functionalized systems, including reactive intermediates. Conversion of cyclohexanone to its enolate anion (65A, see Section 13.2) flattens the ring (see 65B), and leads to a cyclohexene-like conformation. Note the conformational similarity of 64A to cyclohexene model 60. The dimethylamino enamine (see Section 13.6) of cyclohexanone (66A) also shows this half-chair conformation (see 66B). In these cases, the lone-pair electrons (on oxygen in 65 and on nitrogen in 66) overlap with the π-bond. 96 97

(a) B€ oeseken, J.; de Rijck van der Gracht, W. J. F. Rec. Trav. Chim. 1937, 56, 1203. (b) Ref. 29b, p 239.

(a) Ref. 29b, p 240. (b) Dauben, W. G.; Pitzer, K. S. In Conformational Analysis in Steric Effects in Organic Chemistry; Newman, M.S., Ed.; John Wiley; New York, NY, 1956; pp 38–39.

51

1.5 CONFORMATIONS

LDA

L

THF

O

O–

65A

65B

NMe2 R

L

L N

R

Me

Me

R = Me

66B

66A

When compared to chair cyclohexane, the presence of the trigonal-planar carbonyl in cyclohexanone removes the 1,3-diaxial hydrogen atom interactions and, as seen in the models for Corey’s calculations, partially flattens the ring. The conversion of alcohol 67 to ketone 68 illustrates this point. Although 67 is an equilibrating mixture with an axial or equatorial hydroxyl group, upon oxidation to the ketone the flattening effect of the carbonyl is easy to see. Energetically, the R $ OH interaction (U-value) in 67 is removed upon oxidation to 68. In the ketone, the carbonyl effectively eclipses the α equatorial hydrogen atoms, further lowering the conformational energy.98 Ketone 68 exists as an equilibrating conformational mixture favoring the R group in an equatorial position. Exocyclic methylene compounds [e.g., 1-(tert-butyl)-4-methylenecyclohexane] exhibit the same flattening effect as observed when a carbonyl group is attached to a cyclic ketone. An interesting effect is seen in the conversion of this alkene to (1s,4s)-1-bromo-4-(tert-butyl)-1-methylcyclohexane or (1r,4r)-1-bromo-4-(tert-butyl)-1-methylcyclohexane, where A strain is increased in the products, relative to the methylene group. Further, the greater A strain of methyl [1.8 kcal (7.5 kJ) mol1 in Table 1.4 vs. 0.4 kcal (1.7 kJ) mol1] for Br suggests that (1r,4r)-1-bromo-4-(tert-butyl)-1methylcyclohexane may predominate after addition of HBr to the alkene (see Section 2.5.1). This analysis ignores the relative trajectories of approach of the reagents (see Section 10.3) and other steric interactions, however. OH H

H

CrO3 , H

R R

+

H

O

H R

R

OH

O 67 H

68

CH2 H

1-(tert-Butyl)-4-methylenecyclohexane

HBr

H

Br H

H CH3

(1s,4s)-1-Bromo-4-(tert-butyl)1-methylcyclohexane

+

CH3 H Br

(1r,4r)-1-Bromo-4-(tert-butyl)1-methylcyclohexane

In addition to the endocyclic double bond usually found in an enolate anion (e.g., 65A), enolate anions can be formed that have an exocyclic double bond (also see Section 13.5.4). Relief of strain will presumably assist formation of the enolate anion, and may influence the final stereochemistry of groups in the product when the enolate anion reacts with a suitable reagent (see Sections 13.3.1 and 13.4).99 This effect can also be seen in the equilibration of the cis-ester ethyl (1s,4s)-4-(tert-butyl)cyclohexane-1-carboxylate to the trans-ester, ethyl (1r,4r)-4-(tert-butyl)cyclohexane-1-carboxylate, via the planar enolate (69).96 This equilibrium, accomplished by treating ethyl (1s,4s)-4-(tert-butyl) cyclohexane-1-carboxylate with a base followed by a proton source, favors ethyl (1r,4r)-4-(tert-butyl)cyclohexane-1carboxylate due to reduction of the A strain of the axial carboethoxy group in [ACOOR ¼ 1.20 kcal (4.99 kJ) mol1]. 98

(a) Allinger, N. L.; Blatter, H. M. J. Am. Chem. Soc. 1961, 83, 994. (b) Allinger, N. L.; Freiberg, L. A. J. Am. Chem. Soc. 1962, 84, 2201.

99

Ref. 50a, p 67.

52

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

O– H

CO2Et H

H

Base

H

Ethyl (1s,4s)-4-(tert-butyl-)cyclohexane-1-carboxylate

OEt H

H+

H

H H CO2Et

Ethyl (1r,4r)-4-(tert-butyl)cyclohexane-1-carboxylate

69

Six-membered rings containing oxygen (pyran derivatives based on tetrahydro-2H-pyran) are an integral part of carbohydrates [e.g., α-D-glucopyranose, (2S,3R,4R,5S,6R)-6-(hydroxymethyl)tetrahydro-2H-pyran-2,3,4,5-tetraol]. Less substituted pyrans show similar effects. The presence of the α-alkoxy group in pyran (70) leads to a significant effect known as the anomeric effect.100 An alkyl group on cyclohexane or pyran usually prefers the equatorial position due to increased A strain. When an alkoxy group is attached to pyran, however, it prefers the axial position. This effect is probably due to dipolar interactions of the oxygen lone pairs (see 70A and 70B).101 The anomeric effect is evident in glucose, where α-D-glucopyranose accounts for a significant portion of the conformational equilibrium. HO O

O

OH

HO HO

O H H

- D-Glucopyranose (2S,3R,4 R,5S,6 R)-6-(Hydroxymethyl)tetrahydro-2H-pyran-2,3,4,5-tetraol

••

••

O ••

••

O ••

OH Tetrahydro-2H-pyran

R

••

H

O ••

R

70A

••

70B

The influence of substituents on conformational stability is also evident in amines (e.g., N-ethyl piperidine, also known as 1-ethylpiperidine). The chair conformation can be drawn with the ethyl group axial (N lone pair is equatorial as in 71A) or equatorial (N lone pair is axial as in 71B). The greater A-value for ethyl than for the lone pair suggests a predominance of 71B in this equilibrium mixture. This preference for the larger group being in an equatorial position is analogous to the effects seen in pyrans and cyclohexanes.

N

N •• 1-Ethylpiperidine 71A

••

71B

The presence of bulky substituents, or substituents capable of chelation or hydrogen bonding can alter conformational populations,102,103 as first seen with ethylene glycol in Section 1.5.1.3. In 72A [4,40 -((3aR,5R,6R,7aR)-2,2dimethylhexahydrobenzo[d][1,3]dioxole-5,6-diyl)bis(piperazine-2,6-dione)], there are two effects that influence the conformation.104 Bulky substituents lead to large A- and G-strain in both chair conformations. In addition, internal hydrogen bonding of the glutarimide moieties helps stabilize the twist-boat conformation shown in 72B, which is the lowest energy conformation of this molecule. The molecular model shows the twist in the cyclohexane moiety, as well as the proximity of the two diketopiperazine units. The twist-boat conformation is also observed in the absence of internal hydrogen bonding. As confirmed by X-ray crystallography, the main conformation for 72 is a twist-boat.105

100

(a) Lemieux, R. U. Pure Appl. Chem. 1971, 27, 527. (b) Angyal, S. J. Angew. Chem. Int. Ed. Engl. 1969, 8, 157. (c) Kirby, A. J. The Anomeric Effect and Related Stereoelectronic Effects at Oxygen; Springer–Verlag: New York, NY, 1983.

101

David, S.; Eisenstein, O.; Salem, L.; Hehre, W. J.; Hoffmann, R. J. Am. Chem. Soc. 1973, 95, 3806.

102

Ref. 16, p 200.

103

(a) Stolow, R. D. J. Am. Chem. Soc. 1961, 83, 2592. (b) Stolow, R. D.; McDonagh, P. M.; Bonaventura, M. M. J. Am. Chem. Soc. 1964, 86, 2165.

104

Witiak, D. T.; Wei, Y. J. Org. Chem. 1991, 56, 5408.

105

Columbus, I.; Cohen, S.; Biali, S. E. J. Am. Chem. Soc. 1994, 116, 10306.

53

1.5 CONFORMATIONS

H O

N

O

H

H

O

O

N

H N

N

H

O

N Me

O

O

O

N H

N

Me

O

H

O O

O

N H

72B

72A

As mentioned in Section 1.5.2, the reduction of cyclooctanone shows how a given conformation can influence reactivity. Treatment of cyclooctanone with lithium aluminum hydride (LiAlH4, see Sections 7.6.1 and 7.9.4) generates the alcohol (cyclooctanol). The alcohol will have greater transannular strain than the ketone. The reduction may be somewhat sluggish relative to similar reduction of cyclohexanone. The oxidation, however, converts cyclooctanol to cyclooctanone using chromium trioxide (see Section 6.2.1), and the newly introduced carbonyl flattens the ring slightly, which leads to diminished strain. As the product is somewhat more stable than the starting material, the oxidation is rather facile. Similarly, formation of the enolate (cyclooctanone ! 73) with lithium diisopropylamide (LDA, see Section 13.2) should be facile due to relief of transannular strain for formation of the planar enolate moiety. A subsequent alkylation step (73 ! 2-methylcyclooctan-1-one, see Section 13.3.1) may be sluggish, however, since the methyl group in 2-methylcyclooctan-1-one may introduce a bit more strain in the product (remember there are several other low-energy conformations for the eight-membered ring) than in the original unsubstituted cyclooctanone. Me

O

O

O

LDA

Li+

H

1. LiAlH4 2. H3O+

OH THF

MeI

CrO3 , H+

2-Methylcyclooctan-1-one

Cyclooctanone

73

Cyclooctanol

This section has shown the importance of understanding and predicting the conformational preferences for many types of cyclic molecules. The importance of this conformational bias to reactions will be seen throughout most chapters in this book.

1.5.4 Conformations in Polycyclic Molecules O O

L

L

Cyclohexene oxide

The conformational constraints in monocyclic molecules are greatly increased in bi-, tri-, and polycyclic molecules. A simple case is cyclohexene oxide, where the planar three-membered ring flattens the ring as shown. Similar effects are seen in cyclohexene, and also occur when cyclopropane is fused to a ring, so bicyclo[3.1.0]hexanes show this flattening effect. An example is cis-carane, which exists primarily in the conformation shown. Note that the geminal dimethyl groups are perpendicular to the plane of the cyclopropane ring. The cis-isomer is expected to show a Me-Me interaction that is not present in trans-carane. Me Me

Me L Me

L

H

Me Me cis-Carane

cis-Carane

trans-Carane

54

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

The conformational constraints peculiar to [m.n.0] alkanes (n 6¼ 1) can be illustrated by comparing the cis- and transisomers of bicyclo[3.3.0]octane (octahydropentalene), bicyclo[4.3.0]nonane (hydrindane), as well as bicyclo[4.4.0]decane (decalin). Each five- and six-membered ring tends to assume an envelope conformation or a chair conformation, respectively, in their lowest energy form. cis-Bicyclo[3.3.0]octane shows a bent conformation with the two envelope shapes relieving some of the cross-ring interactions.90,106 trans-Bicyclo[3.3.0]octane shows a more extended or open structure, and there is some distortion of the five-membered ring. Compound trans-bicyclo[3.3.0]octane is 6 kcal (25.1 kJ) mol1 higher in energy than cis-bicyclo[3.3.0]octane.107 Both cis- and trans-isomers are easily prepared, however.108

Bicyclo[3.3.0]octane

Bicyclo[4.3.0]nonane

cis-Bicyclo[3.3.0]octane

Bicyclo[4.4.0]decane

trans-Bicyclo[3.3.0]octane

Hydrindanes have the five-membered ring in the envelope conformation with the six-membered ring in a chair. The cis-indane (cis-bicyclo[4.3.0]nonane) shows a 1,3-diaxial interaction with substituents, and there is a gauche-like interaction. In the trans-isomer (trans-bicyclo[4.3.0]nonane), the 1,3-diaxial interaction is diminished, and the extended conformation in trans-bicyclo[4.3.0]nonane is slightly lower in energy [1 kcal (4.19 kJ) mol1] than cis-bicyclo[4.3.0] nonane.109 Similar analysis of cis-decalin (cis-bicyclo[4.4.0]decane) and trans-decalin (trans-bicyclo[4.4.0]decane) leads to the same bent and extended forms, with both six-membered rings in a chair conformation.110 In cis-bicyclo[4.4.0] decane, one ring can assume a boat-like structure in some conformations. There is a 1,3-diaxial interaction in substituted cis-bicyclo[4.4.0]decane, and the overall energy is lower due to decreased nonbonded interactions.111 Eliel et al.112 described the conformations of several other bicyclic and polycyclic compounds.112

cis-Bicyclo[4.3.0]nonane

cis-Bicyclo[4.4.0]decane

trans-Bicyclo[4.3.0]nonane

trans-Bicyclo[4.4.0]decane

106

See Ref. 10, p 776.

107

(a) Ref. 10, pp 774–775. (b) Ref. 29b, p 274

108

Owen, L. N.; Peto, A. G. J. Chem. Soc. 1955, 2383.

109

(a) Ref. 10, pp 776–779. (b) Ref. 29b, p 275. (c) Browne, C. C.; Rossini, F. D. J. Phys. Chem. 1960, 64, 927.

110

Turner, R. B. J. Am. Chem. Soc. 1954, 74, 2118.

111

(a) Ref. 29b, p 279. (b) H€ uckel, W. Annalen 1925, 441, 1. (c) H€ uckel, W.; Friedrich, H. Annalen 1926, 451, 132.

112

Ref. 10, pp 780–793.

55

1.5 CONFORMATIONS

Incorporation of a π-bond from an alkene or aryl moiety into a bicyclic system leads to significant flattening of the ring. Examples are 9-decalene (1,2,3,4,5,6,7,8-octahydronaphthalene) and the benzene-containing derivatives 1,2,3,4tetrahydronaphthalene and 2,3-dihydro-1H-indene. The planar alkene moiety in 9-decalene (see the molecular model) forces the four allylic carbons to be coplanar and each ring more or less behaves as if it were cyclohexene. The planar benzene ring in 1,2,3,4-tetrahydronaphthalene imposes similar conformational constraints (see the molecular model), and the carbons of the nonaromatic ring assume a conformation similar to cyclohexene. In 2,3-dihydro-1H-indene, the nonplanar cyclopentane ring (see the molecular model) behaves more or less like cyclopentene. The presence of three or more fused rings is rather common in nature.

9-Decalene

2,3-Dihydro-1H-indene

1,2,3,4-Tetrahydronaphthalene

The natural product hirsutene113 shows the conformational bias imposed by three fused five-membered rings (see the molecular model). A last example of this phenomenon is seen in the steroid nucleus (74). Each of the cyclohexane moieties will exist in a chair conformation and the cyclopentane will be in an envelope conformation, as seen in the conformational rendering. If a carbonyl or alkene moiety is included the carbonyl carbon is sp2 hybridized, so the sixmembered ring will be flattened at that carbon. As seen in the conformational drawing, each trans-fused cyclohexane ring resembles trans-decalin and the four fused rings impart great conformational rigidity to the molecule. Only the terminal cyclohexyl and cyclopentyl rings have some mobility. The interaction R1-R2-Me of the cyclohexane ring is important, as are the R3-Me and R-R4 interaction of the cyclopentane ring. CH2 Me

H

Me

Me H

H Hirsutene Me R3

Me

R3

R4

H

H

C Me

L

D

H

R1

Me R1

H

H

A

R4

H R2

B H

H R2

113

74

H

For leading references relating to the structure and synthesis of this molecule, see (a) Hua, D. H.; Venkataraman, S.; Ostrander, R. A.; Sinai, G. -Z.; McCann, P. J.; Coulter, M. J.; Xu, M. R. J. Org. Chem. 1988, 53, 507. (b) Curran, D. P.; Raciewicz, D. M. Tetrahedron 1985, 41, 3943.

56

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

Examples of the conformational variations in steroids are seen by examination of cholesterol and β-estradiol.114 When the A ring is aromatic, as in the estradiol the A/B ring system is flattened, inducing significant distortion in the C ring as well (see the two molecular models). Changes in conformation for polycyclic systems can often be predicted using the simple five- and six-ring models discussed in Section 1.5.2. Indeed, it is reasonable to assume that the simple concepts introduced for monocyclic rings (e.g., cyclopentane, cyclohexane, cyclohexene, and cyclohexanone) can be applied with little change to complex molecules containing those units, at least for a first estimate. Throughout this book the analysis of simple rings will be used to understand the conformational bias and reactivity of larger and more difficult synthetic targets. Me Me Me

H H

H

HO

Cholesterol

Me

OH

H A HO

H B

H

-Estradiol

1.6 CONCLUSION This chapter has reviewed the most fundamental concepts of stereochemistry and conformational analysis, with some applications to more complex problems. For students who have had a good undergraduate organic chemistry course, it is simply a review since these concepts are integral to the understanding of organic reactions. The concept of disconnection and retrosynthetic analysis is a prelude to a discussion of synthetic strategies in Chapter 8. The disconnection method and pertinent transforms will, however, be presented in virtually every chapter as new reactions are introduced. The introduction to organic reactions will be continued in the functional group exchange reactions in Chapters 2–4, where the reactions introduced in a typical first organic chemistry course will be reviewed. Some of the concepts will be expanded and updated to include more synthetically useful reactions. Oxidation and reduction will be presented in separate chapters to focus attention on the vast number of reactions that fall under these categories. HOMEWORK

1. For each of the following molecules calculate the percentage for both chair conformations using the values in Table 1.4. Calculate using the temperatures 150°C for (a), 25°C for (b), (c), and (d).

114

Fieser, L. F.; Fieser, M. Steroids; Van Nostrand Reinhold: New York, NY, 1959.

57

1.6 CONCLUSION

Cl

NH2

Me2HC

H3C

Me-CLC

O

Me3C

OMe

MeO

MeO

(A)

Me

Cl

Ph

MeO

(B)

(C)

(D)

2. Determine the absolute configuration [(R) or (S)] for every stereogenic center in the following molecules: H OH

OH

OH

O

H OH

N

H

N OSO3- OH

S HO

HO HO

OH Kotalanol

(A)

O

Lepistine

(B)

O

O

H

(C)

(–)-Crinipellin A

3. Determine the absolute configuration for every stereogenic center in the following molecules: O

OAc BzO O O

OAc

BzO

OAc N

O O O

OH

N

O

O

O

MeO2C

OAc

(–)-4-Hydroxyzinowol

Amphidinolide X

(A)

(+)-Lapidilectine B

(C)

(B)

4. Draw both chair conformations and one twist-boat conformation for trans-1,2-di(triphenylmethyl)cyclohexane. 5. Using the reaction wheel (Fig. 1.2) give reasonable syntheses, including reagents and all intermediates (no mechanisms).

OH

OH

(A)

C N

(B)

CO2Me

Br

CHO

(C)

(D) CH2

(E)

OH

CH3 OH

CN

CO2Me CONH2

(F)

6. Use the Compendium of Organic Synthetic Methods (Vol. 13) to give three different reactions (with literature references and reactions) for each of the following: (a) Acids from nitriles (b) Aldehydes from nitriles (c) Amides from halides (d) Amines from nitriles

58

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

(e) Ethers from halides (f ) Halides from amines (g) Ketones from olefins (h) Alkenes from aldehydes 7. Draw molecule A in what should be the low-energy conformation. Briefly discuss the conformation for the highlighted ring in molecule A. Also discuss the conformational preference (axial or equatorial) for both methyl groups (1 and 2), as well as the hydroxyl.

Me2

HO Me1 A

8. In each of the following, disconnect the indicated bond and draw the disconnect products. Assign d/a based on any natural bond polarization. If there is no bond polarization, assign a reasonable d/a. Correlate the disconnect products with real reagents using the concept of synthetic equivalents in Table 1.2.

OH

O

(A)

(C)

(B) O CN

(E)

(D)

(F)

9. Determine the absolute configuration (R or S) for each chiral axis in the following molecules:

OMe Cl

Et

H

CH2Cl

NH2

CH2CH2Cl

N-H H

(A)

(B)

CHMe2

O

OH

C Br

Cl

H

Me2HC

Me

CMe3 HOH2C

(D)

(C)

OH

(E)

10. Determine the absolute configuration for each stereogenic center in the following molecules:

Me O Me

Me

O Me

(A)

Me H

Mg

Br

CKO

Ph Me Me H Me

H

(B)

P

Me

Me

Ph3P

(C)

O O

O O

P

O

H H

Fe

Et

Me

O H

O

(D)

Ph

59

1.6 CONCLUSION

11. The conversion of the hydroxy-acid shown to the corresponding lactone is rather slow. When pushed under aqueous acid conditions, epimerization at the carbon bearing the hydroxyl can occur. Draw the lactone and explain the observations noted in this question. t-Bu H Me CO2H H

OH

H

Me

12. Explain why cis-1,3-cyclohexanediol exists mainly in the diaxial conformation in aqueous ethanol solution, when by Corey’s ΔG calculations (see Table 1.4), the diequatorial conformation should be lower in energy. 13. Determine the correct re/si label for each prochiral atom in the following molecules:

H

O

Ph N ••

N—H

OH Ph

(A)

(B)

(C)

Relative Energy / kcal mol–1

14. Draw the three most significant anti-rotamers of 1-phenylethan-1-ol (R ¼ Me in the accompanying diagram), and correlate them with the rotamers marked a–c on the energy curve.

OH

8

H H

6

H R Ph f Rotamer b

4 2

Rotamer c Rotamer a

0 0

100

200 f / degree

300

15. (a) Draw the (2S, 3R,4R) and the (2S, 3S, 4S) derivative of molecule A. (b) Draw both the erythro/threo and the syn/anti forms of molecule B.

OH Br Br A

Br B

Cl

60

1. RETROSYNTHESIS, STEREOCHEMISTRY, AND CONFORMATIONS

16. Classify each of the following reactions as stereospecific/stereoselective and/or regiospecific/regioselective: H

H

H

1. LiAlH4

+

2. H3O+

H

(A)

H

O Br H

H Br

But,

OH

8 : 2

H

NaI , Acetone

H I

NaI , Acetone

I H

OH

(B) Me

1. LiN(i-Pr)2

H

2. MeI

(C)

O

O

H

+ 90 : 10

Me O

17. Explain why the ground-state energy of cyclodecane is higher than that of cyclooctadecane. Why is cyclopentadecane higher in energy than cyclohexadecane? 18. Draw the boat conformation of cis-,trans-,trans-1,2,3,4-tetraisopropylcyclohexane.

C H A P T E R

2 Acids, Bases, and Addition Reactions 2.1 INTRODUCTION In Section 1.1, it was stated that most of the actual chemical reactions in a synthesis are those that incorporate or change functional groups. Such reactions are known as functional group exchange reactions, and this chapter will review two major reaction types that are involved in functional group exchanges. The two reactions discussed here include acid-base reactions and addition reaction (see below). Acid-base reactions constitute all or part of many traditional functional group exchange reactions and processes1 and, in fact, most so-called addition reactions are actually acid-base reactions. • Acid-base reactions HA + Bƒƒƒ!HB + A • Addition reactions C]C + XdXƒƒƒ!XdCdCdX Acid-base reactions are among the most general that are known and will be discussed in Section 2.2. They are used to generate reactive intermediates that can be part of other reaction types. Both Lewis and Brønsted-Lowry bases donate two electrons to an acid. A Brønsted-Lowry base donates two electrons to a proton, whereas a Lewis base donates two electrons to an atom other than hydrogen. Acid-base reactions are integral to many reactions, although they are not always easy to describe by a specific transform since they may be an adjunct to the desired transformation. An example is the use of conjugate bases of mineral acids (e.g., HBr, HI) as nucleophiles in substitution reactions. The cyanide and azide ion are also typical nucleophiles, generated by the acid-base reactions: HCN ! CN and HN3 ! N3  (see Section 3.2). Relatively strong bases, which are usually the conjugate bases of weak acids, are typically required to initiate E2 type reactions (Section 3.5.1). Ethoxide and amide bases are typical examples, generated by the acid-base reactions EtOH ! EtO and HNEt2 ! LiNEt2 . Addition reactions involve the transformation of sp hybridized carbons to sp2 or to sp3, or the transformation of sp2 hybridized carbon atoms to sp3. Many reagents that add to alkenes are acids (HCl, HOBr, etc.), and the reaction mechanism for these reactions involves an acid-base reaction in which the alkene reacts as a base, donating two electrons to the acid (H+). Addition reactions that add diatomic bromine, chlorine, and so on, to alkenes, are common in synthesis, and the alkene moiety in such reactions may be categorized as a Lewis base. Such reactions will be discussed in Section 2.5.

2.2 BRØNSTED-LOWRY ACIDS AND BASES Many organic molecules react via an initial acid-base reaction that is formally classified by the Brønsted-Lowry definition. Remember that a Brønsted-Lowry acid is defined as a proton donor and a Brønsted-Lowry base is a proton acceptor. The focus of this definition is an understanding that there is a proton transfer. However, to best understand organic reactions, the focus must be changed to understand that a base “accepts” a proton from the acid by donating two electrons to the proton to form a new covalent bond. In other words, a Brønsted-Lowry base donates two electrons to a hydrogen atom to form a new bond. In principle, any molecule that donates electrons to a protonic acid to form a new bond to that hydrogen atom is a base. Understanding the properties of an acid-base equilibrium is useful for an understanding of the mechanistic details of many different reactions. 1

Many examples of these types of reactions are found in Larock, R. C. Comprehensive Organic Transformations, 2nd ed.; Wiley-VCH: NY, 1999.

Organic Synthesis http://dx.doi.org/10.1016/B978-0-12-800720-4.00002-7

61

Copyright © 2017 Michael Smith. Published by Elsevier Inc. All rights reserved.

62

2. ACIDS, BASES, AND ADDITION REACTIONS

2.2.1 Acidity in Organic Molecules 2.2.1.1 The Importance of Ka and pKa The reaction of HA and B to yield BH+ and A is the general expression written for acid-base reactions. The acid (HA) reacts with a base (B) to yield the conjugate acid HB+ (CA) and a conjugate base, A (CB). The base (B) donates two electrons to the proton (H+) to form a new BdH bond with formation of the product HB+ (CA). Note that CA is also an acid and since A is a base (CB), these two molecules react with each other to regenerate the original acid (A) and the original base (B). In other words, there are two competing acid-base reactions at equilibrium: (A)+ (B) to yield (CA) + (CB), and (CA)+ (CB) to yield (A) + (B). The position of the equilibrium is given by the equilibrium constant K, where Ka is used for acid-base reactions. A large value of Ka means that HA reacted with B to produce BH+ and A so there is a higher concentration of those products relative to the starting materials. The equilibrium is shifted to the right and HA is categorized as a stronger acid. A small value of Ka indicates little reaction of HA with B and there is a higher concentration of HA and B relative to CA and CB. The equilibrium is shifted to the left, which means that HA did not react with the base (B) to any great extent and HA is considered to be a weaker acid. The position of the equilibrium (the value of Ka) is dependent on many factors, including the solvent. A convenient way to express the relative strength of an acid is pKa, which is inversely proportional to Ka. A large Ka is equivalent to a small pKa and as the pKa decreases, the acid strength increases. Remember that the definition of a strong acid (large Ka) implies that the reaction of HA and B is more facile than the reverse reaction of BH+ and A. Conversely, a weak acid (small Ka or large pKa) implies that the reaction of BH+ and A is more facile than that of HA and B. Ka

 + HA + B ƒƒ! ƒƒ BH + A

Ka ¼

½HB + ½A  ½HA½B :

Where pKa ¼  log Ka and K ¼ 10pKa. When comparing two acids, several factors can be used to evaluate relative acid strength. In an acid-base reaction, an acid is on both sides of the equilibrium (A and CA), and comparing the relative strength of A and CA can be used to determine the position of the equilibrium, Ka. Conversely, knowledge of Ka allows the chemist to make an estimate of the relative concentrations of A and CA for that acid-base reaction. An equilibrium reaction can be understood only when both sides of the reaction are examined. In other words, examine the relative strength of the acid HA and base, but also the relative strength of CA and CB. A convenient measure of the strength of CA and CB is to examine the stability and relative reactivity of the conjugate base. A more stable CB is less reactive (with CA), consistent with a larger Ka, whereas a less stable CB is more reactive (with CA), which is consistent with a smaller Ka. 1. In general, a strong acid generates a weak conjugate base and, conversely, a weak acid generates a strong conjugate base. It is convenient to categorize acids with a larger pKa (e.g., water, alcohols, amines) as weak and acids with smaller pKa (e.g., HCl, HNO3) as strong. In general, acids with a pKa < 3 are strong and those with a pKa > 15 or so are weak. Those acids with pKa values between 3 and 15 are usually classified as moderate in strength. Such categorization is, of course, rather arbitrary, but it is useful for many organic reactions. There is an old axiom: A strong acid generates a weak conjugate base. Indeed, the acid HCl is a strong acid, and it generates the weak base, Cl:. Formation of an acid on the right of this reaction that is weaker than HCl shifts the equilibrium to the right (toward the conjugate acid and base), which is consistent with the fact that the HCl-base reaction is more favorable than the H-Base-Cl reaction.

HCl +

Base

HJBase +

Cl–

Ammonia is a weak acid that generates a strong conjugate base, NH2  . Formation of a base on the right of this reaction that is stronger than ammonia is consistent with the fact that the NH3-base reaction is shifted to the left (toward ammonia). If the amide anion is more reactive (less stable), it will react with H-Base to regenerate ammonia and the Base, shifting the equilibrium to the left, which is consistent with a smaller Ka.

NH3 +

Base

H—Base +

NH2-

2.2 BRØNSTED-LOWRY ACIDS AND BASES

63

2. Across the periodic table, acidity of an XdH species generally increases from left to right.2 Weak acid CH3 < NH3 < H2 O < HF Strong acid ð 50Þ ð 31Þ ð15:7Þ ð3:17Þ pKa Strong conj: base CH3  > NH2  > HO > F Weak conj: base This observation effectively puts attention on the strength of the HdA bond. Going across the periodic table, the size of atoms does not increase, and indeed there is a contraction in atomic radius from left to right. Therefore, the strength of HdO, HdN, and HdF will not be influenced very much by changes in size. However, there is a difference in electronegativity across the periodic table. Assuming that a bond to the more electronegative atom is more polarized and easier to break, HdF should be easier to break than HdO, which is easier to break than HdN. This observation is correct, and consistent with HF as the strongest acid. It is therefore useful to assume that across the periodic table, changes in electronegativity are more important for a given bond. Differences in electronegativity must also be influential for stability (reactivity) of the conjugate base when going across the periodic table. The more electronegative atom should be less able to donate electrons and therefore be a weaker base and less reactive (more stable). Examination of the conjugate bases F, HO, and NH2  shows that F is a weaker base than HO, which is a weaker base than NH 2 . This trend is consistent with the fact that fluorine is more electronegative than oxygen, which is more electronegative than nitrogen. The fact that fluoride ion is predicted to be a weaker base is consistent with a larger Ka for the reaction of HF. A necessary consequence of the trend in acidity is that basicity increases from right to left across the periodic table. 3. Acidity increases going down the periodic table, despite a decrease in electronegativity (Ref. 3,2). HFð3:17Þ < HClð7Þ < HBrð9Þ < HIð10Þ and, H2 Oð15:74Þ < H2 S ð7:00Þ Up and down the periodic table there are significant differences in the size of the atoms, and an assumption can be made that differences in the size of the atoms plays a more important role than differences in electronegativity for the Ka of acids HdX. The covalent radius of I (135 pm; 1.35 Å) is much greater than that of F (71 pm; 0.71 Å),4 and the bond length for HdF is 92 pm (0.92 Å) and that for HdI is 161 pm (1.61 Å). The fluorine atom is smaller than the iodine atom, so the internuclear distance for HdF is smaller than the internuclear distance for the HdI bond, which is consistent with a stronger covalent bond for HdF. The bond dissociation energy of the HdI bond is 71.3 kcal (298.3 kJ) mol1 and the HdF bond is 136.2 kcal (569.7 kJ) mol1.5 In other words, it is easier to break the HdI bond, which is consistent with HI being a stronger acid when compared to HF. The greater size of the iodine atom leads to a greater HdI bond distance and, due to the resultant decrease in internuclear electron density, a weaker bond. The acid strength increases, since the weaker IdH bond is more easily ionized, which ignores solvation effects, however, as well as product stability, which are critical to this analysis. Both the starting materials and the products must be examined to determine the position of the equilibrium, and thereby the relative acidity. The conjugate bases in these reactions are the fluoride and iodide ions. The ionic radius of the iodide ion (I) is reported to be 215 pm (2.15 Å) and that of the fluoride ion (F) is 136 pm (1.36 Å).4 It is clear that the iodide ion is significantly larger than the F ion. The larger I ion has a greater surface area when compared to the fluoride ion and is capable of dispersing the charge over a larger area. If the charge is more dispersed, it is more difficult to donate two electrons, so the I ion is more stable and less reactive. In other words, the I ion is a weaker base relative to the F ion. In addition, larger ions are also more easily solvated, which also contributes to greater stability. Greater stability and poorer reactivity with the conjugate acid is consistent with a larger Ka for HI, and HI is a stronger acid than HF. A consequence of this trend for acid strength is that base strength increases going up the periodic table. The three concepts discussed in this section allow a quick inspection of a simple acid (HX) to estimate its relative acid strength. Experimental analysis of the acids commonly encountered in organic reactions requires knowledge of at least three factors: (1) electronic effects, (2) resonance effects, and (3) solvent effects. 2

Lowry, T. H.; Richardson, K. S. Mechanism and Theory in Organic Chemistry, 3rd ed.; Harper & Row: NY, 1987; pp 297–298.

3

Smith, M. B. March’s Advanced Organic Chemistry, 7th ed.; John Wiley & Sons: NJ, 2013; pp 314–315.

4

Huheey, J. E. Inorganic Chemistry, Principles of Structure and Reactivity; Harper & Row: NY, 1972; Table 5.1, pp 184–185.

5

Haynes, W. M. CRC Handbook of Chemistry and Physics, 94th ed.; CRC Press: Boca Raton, FL, 2013–2014; pp 9-67–9-68.

64

2. ACIDS, BASES, AND ADDITION REACTIONS

2.2.1.2 Inductive and Resonance Effects The presence of an electron-donating or an electron-withdrawing heteroatom (or group) on a carbon induces bond polarization through adjacent carbon atoms down a carbon chain, as illustrated by 1 (first introduced in Section 1.2).6 This phenomenon is known as an induced dipole. Electron-donating groups release electrons via inductive or field effects, leading to the so-called +I effect. Conversely, an electron-withdrawing group removes electron density, the so-called I effect. With respect to an acid, the +I effect lowers the acid strength, whereas the I effect raises it. Electronic effects are described by the inductive model,7,8 and also by the field-effect model.7,9 The inductive model assumes that substituent effects are propagated by the successive polarization of the bonds between the substituent and the reaction site, as in 1. This effect is transmitted through the σ-bond network (σ inductive effect), as well as the π-bond network (π inductive effect).8 The field-effect model assumes that the polar effect originates in bond dipole moments, and is propagated according to the classical laws of electrostatics. The appropriate description of this effect is the Kirkwood-Westheimer model,9 in which the molecule is treated as a cavity of low dielectric constant submerged in a solvent continuum.7a Y

••

••

• •d− X

CO 2H

d+

d+

CO 2H

X

d−

1

2

3

Attempts to distinguish inductive and field effects have been frustrated by the lack of a molecule that provides an unambiguous answer.7b Such a molecule would require that the low dielectric cavity of the field-effect model10 be occupied by the chemical bonds important to the inductive-effect model.8 For example, O-chlorophenylpropionic acid is weaker than expected by the inductive model.10 If a small attenuation factor is adopted for through-bond transmission of the polar effect, both models predict similar results.11 Stock and coworkers reported the pKa for derivatives of 27c and 3,7b which allow the field effect to be studied. This study determined that the dominant term associated with the change in pKa was the positive charge, transmitted through the cavity.12a This finding is in general agreement with the Kirkwood-Westheimer model,9 which suggests that the field effect is the dominant term.7 When X is an ammonium salt, there is a I effect,13 which lowers the pKa (strengthens the acid) relative to acetic acid.14 Due to the localized character of the σ electrons, the inductive effect in this system diminishes with increasing distance between separation of the substituent and reaction center. The change in pKa is actually due to a combination of inductive, steric, and hyperconjugative influences. Using simple acyclic carboxylic acids to illustrate these phenomena, chloroacetic acid (pKa ¼ 2.87)15 can be compared with propionic acid (pKa ¼ 4.76), since there is no change in the number of α-hydrogen atoms, no steric or hyperconjugative effects, and an inductive effect is the only significant contribution to pKa.12 The I effect of chlorine induces a significant reduction in pKa (that compound is a stronger acid). Electron-releasing groups with a +I effect will have the opposite effect and lead to a decrease in acidity.12 Many organic compounds are categorized as acids with pKa values that are approximately water (pKa ¼ 15.75),16 including alcohols, carboxylic acids, phenols, and 1,3-dicarbonyl compounds. Simple unfunctionalized examples of these organic molecules are weaker than the familiar mineral acids HCl or H2SO4. 6

(a) Seebach, D. Angew. Chem. Int. Ed. Engl. 1979, 18, 239; (b) Cram, D. J.; Cram, J. M. Chem. Forsch. 1972, 31, 1 (Chem. Abstr. 77:163690d 1972).

7

(a) Baker, F. W.; Parish, R. C.; Stock, L. M. J. Am. Chem. Soc. 1967, 89, 5677; (b) Golden, R.; Stock, L. M. J. Am. Chem. Soc. 1966, 88, 5928; (c) Holtz, H. D.; Stock, L. M. J. Am. Chem. Soc. 1964, 86, 5188. 8

Branch, G. E. K.; Calvin, M. The Theory of Organic Chemistry; Prentice Hall: NY, 1941; Chapter 6.

9

(a) Kirkwood, J. G.; Westheimer, F. H. J. Chem. Phys. 1938, 6, 506; (b) Westheimer, F. H.; Kirkwood, J. G. J. Chem. Phys. 1938, 6, 513.

10

(a) Roberts, J. D.; Carboni, R. A. J. Am. Chem. Soc. 1955, 77, 5554; (b) also see Wells, P. R.; Adcock, W. Aust. J. Chem. 1965, 18, 1365.

11

Ehrenson, S. Progr. Phys. Org. Chem. 1964, 2, 195.

12

(a) Dewar, M. J. S.; Grisdale, P. J. J. Am. Chem. Soc. 1962, 84, 3539, 3548; (b) Murrell, J. N.; Kettle, S. F. A.; Tedder, J. M. Valence Theory; John Wiley: London, 1965; Chapter 16. 13

Brown, H. C.; McDaniel, D. H.; Hafliger, O. In Determination of Organic Structures by Physical Methods; Braude, E. A.; Nachod, F. C., Eds.; Academic Press: NY, 1955; p 569.

14

Reference 13, see Chapter 14 therein.

15

Bolton, P. D.; Hepler, L. G. Qated. Rev. Chem. Soc. 1971, 25, 521.

16

Ballinger, P.; Long, F. A. J. Am. Chem. Soc. 1960, 82, 795. Also see Olmstead, W. N.; Margolin, Z.; Bordwell, F. G. J. Org. Chem. 1980, 45, 3295.

65

2.2 BRØNSTED-LOWRY ACIDS AND BASES

Some molecules can behave as either an acid or a base, and are categorized as amphoteric. Alcohols are amphoteric and in the presence of a stronger acid, they will react as a base. The reaction of methanol with HCl, for example, leads to an onium ion (formally, an oxonium ion that is named methyloxonium chloride), and it is the conjugate acid of methanol. Methanol has a pKa of 15.516 and the pKa of the oxonium ion conjugate acid is about 2.17 Therefore, the conjugate acid is a much stronger acid when compared to the alcohol, the equilibrium lies toward methanol, which is categorized as a base in this reaction. Alcohols also react as acids in the presence of a strong base, illustrated by the reaction of methanol with NaNH2 to yield sodium methoxide (the conjugate base) and ammonia (the conjugate acid). Alcohols will react as an acid in the presence of a base to yield a conjugate acid that is weaker than the alcohol. When the amide anion and methanol are mixed together there will be two acids in the equilibrium mixture, methanol and ammonia. Methanol has a pKa of 15.516 and the pKa of ammonia has been estimated to be 41.18 Therefore, the equilibrium is expected to be on the side of methoxide and the conjugate acid is on the side of ammonia. In this reaction, methanol reacts as an acid. In general, alcohols have pKa values of 16–18, and alcohols are classified as weak acids. H

H Cl−

+ H C O H

Na+

−NH

H ClJH

H

Methyloxonium

H H C H

2

H C H

O

H

Methanol

O− Na+ +

NH3

Sodium methanolate (sodium methoxide)

As noted previously, the acidity of a molecule cannot be properly estimated without considering the products formed by the reaction. It has been established that dispersal of charge on the larger iodide ion, relative to the fluoride ion, leads to greater stability and lower basicity. Therefore greater charge dispersal in the conjugate base is consistent with a larger Ka. Another way to disperse charge is by resonance delocalization. Indeed, resonance disperses charge over several atoms rather than over a larger surface of one atom. Therefore, dispersal of charge via resonance delocalization leads to greater stability and lesser reactivity for a conjugate base. In other words, a resonance-stabilized conjugate base is less basic because it shows poor reactivity with the conjugate acid when compared to a base that is not stabilized by resonance. If the reaction with the conjugate acid is poor, the equilibrium lies to the right, consistent with a stronger acid and base, and a weaker conjugate acid and base. O

H H

C

C

−

O

H

H

Na+ OH

O

H H

C

C

H Acetic acid

O−

H O−

H Na+

C

C

O

+

H2O

H

Sodium acetate

The effect of resonance is clear when acetic acid and methanol are compared as acids. Acetic acid (pKa ¼ 4.76)19 owes its greater acidity, in part, to the resonance-stabilized acetate anion, whereas methanol (pKa ¼ 15.5)16 generates the methoxide anion. The charge is localized on the oxygen atom in methoxide, but the charge is delocalized over three atoms in the acetate anion. In other words, the acetate anion is resonance stabilized, less reactive, and a weaker base. The increased stability and decreased basicity of the conjugate base of acetic acid (the acetate anion) leads to a larger Ka, which means a lower pKa for acetic acid relative to methanol. Resonance effects that influence pKa can also be transmitted via the π-bonds of aromatic rings. A comparison of phenol (pKa ¼ 9.98), 4-nitrophenol (pKa ¼ 7.2) and 2,6-dinitrophenol (pKa ¼ 3.6)20 shows that the presence of a nitro group at the ortho- and para-positions leads to an enhancement in acidity, and delocalization of electron density in the conjugated phenoxide base is a major contributing factor. Removal of the acidic proton of phenol leads to a resonance-stabilized phenoxide anion. The presence of a nitro group at C4 leads to more extensive delocalization of the charge on the nitro group (more resonance contributors) and this increased resonance stability makes 4-nitrophenoxide (the conjugate base) more stable and 4-nitrophenol more acidic. A second nitro group at C2 or C6

17

Smith, M. B. March’s Advanced Organic Chemistry, 7th ed.; John Wiley & Sons, Inc.: NJ, 2013; p 315.

18

Bordwell, F. G.; Drucker, G. E.; Fried, H. E. J. Org. Chem. 1981, 46, 632.

19

Haynes, W. M. CRC Handbook of Chemistry and Physics, 94th ed.; CRC Press: Boca Raton, FL, 2013–2014; pp 5–94.

20

Bos, M.; van der Linden, W. E. Anal. Chim. Acta 1976, 332, 201.

66

2. ACIDS, BASES, AND ADDITION REACTIONS

(as in 2,6-dinitrophenol) will lead to even greater resonance delocalization, greater stability, and enhanced acidity for that phenol. In other words, stabilization of the anion by nitro substituents at C2, C4, and/or C6 is reflected in the pKa values. OH

O

O

O

O

Base

Phenoxide anion

Phenol OH

O

O

O

O

O

Base

O

N

O

O

N

N O

4-Nitrophenol

O

O

O

N

O

O

N

O

O

N

O

4-Nitrophenoxide anion

Most reactions take place in solution rather than in the gas phase,21 and the factors that account for the enormous differences between solution and gas-phase chemistry are operationally defined as solvation.22 Protic solvents facilitate charge separation, and solvation of both the conjugate acid and base can shift the equilibrium to the right (the compound is more acidic). It is important to understand that water is more effective at facilitating charge separation and solvation when compared to most other solvents, including other protic solvents (e.g., ethanol or methanol). The acidity of a compound changes with the polarity of the solvent, and is typically greater in protic solvent when compared with aprotic solvents. A comparison of the pKa values of acetic acid (4.76), benzoic acid (4.20), and 4-nitrophenol (7.15) compares the pKa values in water. In dimethyl sulfoxide, (DMSO, solvent) the pKa of acetic acid is 11.4, benzoic acid is 10.0, and 4-nitrophenol is 9.9. And in acetonitrile, acetic acid has a pKa of 22.3, benzoic acid is 20.7, and 4-nitrophenol is 7.0.22 The pKa of compounds in various solvents has been reported.21 Mineral acids and bases, and the conjugate acids and bases, are usually poorly soluble in aprotic solvents. Organic acids (e.g., carboxylic acids and sulfonic acids), on the other hand, are usually soluble in many aprotic solvents, as well as in alcohol solvents. For this reason, it is usually necessary to ascertain acid-base equilibrium in aprotic solvents. Ionization and solvation of ionic species is relatively poor in many protic solvents other than water, although very poor in aprotic solvents. Some aprotic solvents (e.g., DMSO or DMF) are exceptions. If an acid-base reaction occurs in an aprotic solvent, there is a problem that influences the extent of ionization (i.e., the value of Ka). Aprotic solvents effectively solvate only positively charged species and not anions or other negatively charged species. Therefore, ionization (separation of charge) is more difficult in an aprotic solvent when compared to a protic solvent. In general, water or aqueous solvents are the best choice for ionization reactions. There are some exceptions, and ionization is more facile in DMSO and DMF. Indeed, DMSO is a superior solvent for ionization reactions. Acetic acid, benzoic acid, and 4-nitrophenol all show enhanced acidity in the more polar DMSO, relative to DMF. Solvating ability is not the only factor, however. Acidity is diminished in methanol (pKa of acetic acid, 9.6; benzoic acid, 9.1; 4-nitrophenol, 11.2)22 due to decreased hydrogen-bonding and weaker solvation when compared to acidity measurements in water. The basicity of the solvent is an important factor. A comparison of pKa data in the aprotic solvents DMSO, and acetonitrile shows that the more basic the solvent, the greater the apparent acidity of HA. This effect is due, in part, to disruption of the internal hydrogen-bonding of HA.23 An important reason for understanding that compounds have different pKa values in different solvents involves chemical reactions. Differences in pKa allow the use of different bases of different base strength, which may have important implications for a given reaction. Indeed, the relative acidity of a molecule will influence the chemical reactions that may occur in different solvents. In some cases, this knowledge may prevent unwanted side reactions while in other cases improved yields may be realized.

21

Arnett, E. M. Acc. Chem. Res. 1973, 6, 404.

22

(a) Clare, B. W.; Cook, D.; Ko, E. C. F.; Mac, Y. C.; Parker, A. J. J. Am. Chem. Soc. 1966, 88, 1911; (b) Kolthoff, I. M.; Chantooni, M. K., Jr.; Bhowmik, S. J. Am. Chem. Soc. 1968, 90, 23. 23

Grunwald, E.; Ralph, E. K. Acc. Chem. Res. 1971, 4, 107.

67

2.2 BRØNSTED-LOWRY ACIDS AND BASES

There are organic compounds that react as weak acids, but they have pKa values that are greater than that of water. Such reactions require the use of a rather strong base, however. The hydrogen on the α-carbon of a ketone (e.g., butan2-one and similar ketones) are important weak acids in organic chemistry (see Sections 13.3–13.5). In this case, butan-2one (the acid) reacts with sodium ethoxide (the base), the conjugate base is an enolate anion (sodium but-2-en-2-olate), and ethanol is the conjugate acid. Ethanol (pKa  15.5)24 is a stronger acid than the ketone (pKa  19–20),25 and will react with the basic enolate anion to regenerate the ketone and shift the equilibrium back to the left. The net result is an equilibrium solution that contains the enolate anion ethanol, butan-2-one, and NaOEt (Section 13.2.5).26 As in any acid-base reaction, a ketone does not function as a Brønsted-Lowry acid unless a sufficiently strong base is present to remove and accept the acidic proton. +

NaOEt

O− Na+

O Butan-2-one

Sodium ethoxide

+

Sodium but-2-en-2-olate

EtOH Ethanol

In later chapters (particularly Chapters 11–13), carbon acids (molecules in which the acidic proton is attached to a carbon atom, CdH) including terminal alkynes, ketones, aldehydes and esters will be used in acid-base reactions. As with other acid-base reactions, the solvent plays an important role in the pKa of carbon acids (Section 13.3), so the choice of solvent is important in any reaction that involves those acids. Phenylacetylene (PhC^CH, ethynylbenzene), for example, has a pKa of 15.8 in diethyl ether, but 26.5 in DMSO.27 Phenylacetylene reacts with sodium amide to generate the alkyne anion as the conjugate base and ammonia as the conjugate acid. In this particular example, the difference in pKa is explained by the extent of ion-pairing or ion aggregation. The equilibrium for the deprotonation of phenylacetylene is also shifted to the right when the solvent is changed from cyclohexylamine (pKa ¼ 20.5) to diethyl ether due to increased stability of the ion pair (for the potassium anion, PhC^CK+) or the ion aggregate (PhC^CK+)n.27 Na+

−NH 2

CLC

C L C: −Na+

H

+

NH3

Ethynylbenzene

The fundamental concept of inductive and field effects introduced for carboxylic acids can be applied to weak acids, including carbon acids (e.g., the α-hydrogen of ketones). The reaction of 6-allyl-3-isobutoxycyclohex-2-en-1-one with the strong base lithium diisopropylamide (LDA; see Section 13.2.2) in the Zoretic et al.28 synthesis of megaphone illustrates this premise. Both Ha and Hb in 6-allyl-3-isobutoxycyclohex-2-en-1-one are acidic, but Ha has a lower pKa (it is the stronger acid) due to its proximity to the carbonyl (Sections 13.1 and 13.2) and is removed preferentially. The conjugating effect of the π-bond makes Hb acidic (it is a vinylogous acid), but that acidity is diminished due to the distance from the carbonyl. Formation of the enolate anion of 6-allyl-3-isobutoxycyclohex-2-en-1-one was followed by alkylation (Sections 3.2.1 and 13.3.1) to provide methyl 2-(1-allyl-4-isobutoxy-2-oxocyclohex-3-en-1-yl)acetate in good yield. O

O

Ha

1. LDA (1.2 equiv) THF , –78°C

O

MeO2C

2. BrCH2CO2Me THF , –78°C rt

Hb 6-Allyl-3-isobutoxycyclohex-2-en-1-one

O Hb Methyl 2-(1-allyl-4-isobutoxy-2oxocyclohex-3-en-1-yl)acetate

24

Haynes, W. M. CRC Handbook of Chemistry and Physics, 94th ed.; CRC Press: Boca Raton, FL, 2013–2014; pp 4–94.

25

See Zook, H. D.; Kelly, W. L.; Posey, I. Y. J. Org. Chem. 1968, 33, 3477.

26

(a) House, H. O. Modern Synthetic Reactions, 2nd ed.; W.A. Benjamin: NY, 1972; pp 452–595; (b) House, H. O.; Phillips, W. V.; Sayer, T. S. B.; Yau, C. C. J. Org. Chem. 1978, 43, 700; (c) Etheredge, S. J. J. Org. Chem. 1966, 31, 1990; (d) House, H. O.; Czuba, L. J.; Gall, M.; Olmstead, H. D. J. Org. Chem. 1969, 34, 2324; (e) House, H. O.; Gall, M.; Almstead, H. D. J. Org. Chem. 1971, 36, 2361.

27

Bordwell, F. G.; Matthews, W. S. J. Am. Chem. Soc. 1974, 96, 1214.

28

Zoretic, P. A.; Bhakta, C.; Khan, R. H. Tetrahedron Lett. 1983, 24, 1125.

68

2. ACIDS, BASES, AND ADDITION REACTIONS

It is apparent that acidity in organic molecules is influenced by several factors. Among these, inductive effects, solvent effects, and effects that influence the stability of both the acid and the conjugate base are important. The base is obviously important since the strength of an acid is directly dependent on the strength of the base.

2.2.2 Basicity in Organic Molecules Since the base is part of the acid-base equilibrium, the relative strength of a base is determined by the position of the equilibrium and is influenced by the same factors discussed for acids.29 For common bases, base strength increases going to the left and up in the periodic table. An example follows: F > Cl > Br > I Base strength can be expressed by the pKa of its conjugate acid, BH+ according to the equation + BH + ƒƒƒ! ƒƒƒ B : + H Ka ¼

½B :½H +  BH +

The inductive and field effects lead to +I (electron-donating groups, base strengthening) and I effects (electronwithdrawing groups, base weakening). Examination of amines shows that electron-releasing groups intensify the electron density around nitrogen, and to a first approximation enhances basicity. Electron-withdrawing groups should have the opposite effect. Resonance effects (+R) also apply, as described earlier for acids. Resonance effects are also known as mesomeric effects (+R is +M; R is M), as described in the discussion of aromatic substitution in Section 3.10. Electrondonating groups can be characterized as having +R effects (base strengthening), and electron-withdrawing groups as having R effects (base weakening). When comparing 4-nitroaniline and 4-methylaniline, for example, the methyl group is electron releasing (+R), but the nitro group is electron withdrawing (R). The +R effect of the methyl is transmitted through the aromatic ring, intensifying the electron density on nitrogen and increasing the basicity. Commonly used bases can be categorized as being +I or I, +R or R effects.30 Alkoxy (RO d) and amino groups (R2Nd) are categorized as +I (base strengthening), whereas halogens, SO2R, COR, CO2R, and Ph are categorized as I (base weakening). For groups attached to an aromatic ring, halogens, OH, OR, NR2, and alkyl are +R and base strengthening, whereas NO2, CN, CO2R, Ph, and SO2R are categorized as R and base weakening. The I and R effects may operate in opposite directions. An example is a slight resonance effect of a group in the meta position of an aromatic nucleus, where the inductive effect is rather large. In the para position, the resonance effect is also large (see Section 3.10.3). Apart from the +I and I and +R and R effects, which are electronic effects, steric hindrance plays a role in base strength. There are two categories of steric hindrance around a basic atom that influence relative base strength. 1. Primary steric effect. Steric impedance to protonation occurs when bulky groups surround the basic center. When this results in greater strain in the cationic conjugate acid produced upon protonation than in the neutral molecule, the base is weaker.31 2. Secondary steric effect. There is hindrance to solvation and it is base weakening.13

The primary steric effect is credited with the greater basicity of aniline relative to 2,6-dimethylaniline, due to greater steric inhibition for donation of electrons to a proton. The secondary effect is probably responsible for the difference in basicity between N,N-diethylaniline (pKa ¼ 6.65) and N,N-dimethylaniline (pKa ¼ 5.8).13 The two ethyl groups inhibit solvation of the protonated conjugated acid,14 which is the ammonium salt. Solvation varies with the available space around the basic center, and should be less for pyridine when compared with aniline. The diminished basicity of tertiary amines relative to secondary amines in solution has been attributed to this effect.30a In the preceding discussion, it was assumed that the inversion of base strength in solution was primarily a solvation phenomenon. Indeed, this effect is largely the result of hydrogen-bonding in an aqueous solvent when comparing ammonia (NH3 ! NH4 + ) with four sites for hydrogen-bonding in the ammonium salt versus a tertiary amine (R3 N ! R3 NH + ), with only one site for hydrogen-bonding in the ammonium salt. A larger number of alkyl groups on nitrogen leads to fewer protons in the ammonium salt, and diminished hydrogen-bonding. This structural feature 29

Clark, J.; Perrin, D. D. Q. Rev. Chem. Soc. 1964, 18, 295.

30

(a) Trotman-Dickenson, A. F. J. Chem. Soc. 1949, 1293; (b) Arnett, E. M.; Jones, F. M., III; Taagepera, M., Henderson, W. G., Beauchamp, J. L.; Holtz, D.; Taft, R. W. J. Am. Chem. Soc. 1972, 94, 4724. 31

Taft, R. W., Jr. Steric Effects in Organic Chemistry; John Wiley: NY, 1956; Chapter 13.

69

2.2 BRØNSTED-LOWRY ACIDS AND BASES

works in opposition to the electron-releasing effect of the increasing number of alkyl groups and dispersal of charge by the inductive effect.30 It has been known for many years that the basicity of amines in solution [expressed as pKa of the conjugate acid (BH+)] takes the order:22 NH3 < R3N < RNH2  R2NH. This anomalous order is understood in terms of two opposed influences, one base strengthening and the other base weakening. The base-strengthening effect is the electronreleasing inductive effects of the alkyl groups on nitrogen. This effect should make tertiary amines more basic when compared to secondary or primary amines. As described immediately above, the other major influence is solvation. Solvation is greatest for the ammonium salt produced from primary amines, leading to greater stabilization and enhanced basicity. Using high-pressure mass spectrometry, where there are no solvation effects, Munson32 demonstrated the order of gas-phase basicity is NH3 < RNH2 < R2NH < R3N. Since most reactions that will be discussed in this book are done in solution, understanding the order of basicity in solvents is critical. Any discussion of relative acidity or basicity must specify the solvent. The actual order of basicity for amines in solvents, however, is a combination of inductive and solvation effects. Steric hindrance around the electron pair on nitrogen also plays an important role. When all of these factors are accounted for, secondary amines tend to be the most basic, as shown. There is a great deal of structural variety for the amines used in synthesis,33 and the pKa of several common amines have been determined in THF at 25°C,34 including 4-(N,N-dimethylamino)pyridine (DMAP), proton sponge (N1,N1, N8,N8-tetramethylnaphthalene-1,8-diamine), quinuclidine, (1-azabicyclo[2.2.2]octane), DBU (1,8-diazabicyclo[5.4.0] undec-7-ene or 2,3,4,6,7,8,9,10-octahydropyrimido[1,2-a]azepine),33b 1,4-diazabicyclo[2.2.2]octane (DABCO), triethylamine, tributylamine, and tribenzylamine.34 The pKa value of DMAP in THF is 0.61, Proton Sponge is 2.15, quinuclidine is 0.15, DBU is 3.78, DABCO is 0.80, triethylamine is 2.11, tribenzylamine is 2.41, and tributylamine is 3.40.34 The pKa values in DMSO and acetonitrile were also determined in this work, all from proton-transfer ion pairs of the type BH+ with acidic indicator hydrocarbons.34 The values obtained in THF differ substantially from the ionic pKa values for NH+ in DMSO or acetonitrile. In 2000, Streitwieser and Kim34 stated that “at the present time amines cannot be placed quantitatively on any of the ion-pair acidity scales currently in use for neutral acids in THF.” The realization that it may be necessary to use different bases in different solvents, despite the fact that all are considered to be good bases in organic reactions, makes it clear why a quantitative comparison of bases is useful. Me2N N

NMe2

NMe2 N

(DMAP) 4-(N,N-Dimethylamino)pyridine

(Proton sponge) N1,N1,N8,N8-Tetramethylnaphthalene-1,8-diamine

Quinuclidine (1-azabicyclo[2.2.2]octane)

N

N N

N

(DABCO) [DBU (1,8-diazabicyclo-[5.4.0]-undec-7-ene] 1,4-Diazabicyclo[2.2.2]octane 2,3,4,6,7,8,9,10-Octahydropyrimido[1,2-a]azepine

Many different types of organic bases are used in organic chemistry, as is made clear by the structural variety of the bases just described. Inorganic bases are also used, and they are commonly the conjugate bases of weak acids (e.g., water) and ammonia (e.g., hydroxide and amide: OH and NH2). Deprotonation of alcohols leads to alkoxide bases (RO), which are very useful in organic chemistry as bases and nucleophiles. Methoxide (MeO), ethoxide (EtO), and tert-butoxide (Me3CO) are very common, and are usually used in the alcohol solvents from which they were made (methoxide in methanol, tert-butoxide in tert-butanol, etc.). Water (pKa ¼ 15.8) or alcohols (pKa  16–18) are typically

32

Munson, M. S. B. J. Am. Chem. Soc. 1965, 87, 2332.

33

For synthetic examples, see (a) Achab, S.; Das, B. C. J. Chem. Soc. Chem. Commun. 1983, 391; (b) Oediger, H.; M€ oller, Fr. Angew. Chem. Int. Ed. Engl. 1967, 6, 76.

34

Streitwieser, A.; Kim, Y.-J. J. Am. Chem. Soc. 2000, 122, 11783 and references cited therein.

70

2. ACIDS, BASES, AND ADDITION REACTIONS

deprotonated using sodium hydride (NaH), potassium hydride (KH), sodium and potassium hydroxide (NaOH, KOH), sodium metal [Na(0)], and sodium or potassium amide (NaNH2 and KNH2). Many neutral organic compounds, particularly amines and phosphines, are bases but they are generally weaker than ionic bases (e.g., NaOH or the alkoxide bases). Alcohols are organic bases, but they are amphoteric compounds, as mentioned in Section 2.2.1.2. Alcohols are typically weaker bases than amines, and alkoxides are weaker bases than amide bases. If an alcohol reacts as a base with a strong acid the conjugate acid is an oxonium ion. In the reaction of pentan-3-ol and H+, for example, the initial product is an oxonium ion (pentan-3-yloxonium). In a subsequent reaction, the oxonium ion intermediate reacts with the nucleophilic chloride ion at carbon, with loss of the leaving group water to yield 3-chloropentane as the final product (see Section 3.4). As noted, an acid catalyst that is strong enough to react with the alcohol must be a stronger acid than the OH unit of the alcohol. This reaction can be related to the generic acidbase reaction, where [HA] is the acid catalyst (HCl), [B] is the oxygen atom of pentan-3-ol, the conjugate base is the chloride ion [sA] (the gegenion of the acid catalyst), and the conjugate acid [HB+] is Et2CH]OH+2 . Note that an oxonium ion, the conjugate acid derived from an alcohol, is a significantly stronger acid than the ammonium salt, the conjugate acid derived from an amine. HCl Cl−

+H+

O

O

−H+

H

H

Cl

H

Pentan-3-yloxonium

Pentan-3-ol

I−

HI O

O CH3

3-Methoxypentane

3-Chloropentane

H

O

−

CH3

I

Methyl(pentan-3-yl)oxonium

+ CH3I

H Pentan-3-ol

Ethers are bases, and although they are weaker bases when compared with amines they are slightly stronger bases than most alcohols. In general, ethers are relatively unreactive compounds except for very strong acids (e.g., HI or HBr), and they are commonly used as solvents. Diethyl ether and THF are weak bases in the presence of a strong acid and THF is a stronger base than diethyl ether. The relative base strength can be quantified by comparing the experimentally determined acidity of the conjugate acid derived from diethyl ether and from THF. The conjugate acid of diethyl ether is [Et2OH+],which has a pKa ¼  3.12,35 which is a stronger acid when compared to the protonated form of THF, which has a pKa ¼  2.08.35 Tetrahydrofuran is recognized as a stronger base, and leads to a weaker conjugate acid relative to a comparative reaction with diethyl ether. In reactions where the ether behaves as a Lewis base, as in Grignard reactions (Section 11.4.1) or hydroboration (Section 9.2.1), understanding the relative basicity of the solvent is important. Ethers do react with HI or HBr to give an alcohol and an alkyl halide. 3-Methoxypentane reacts with HI, for example, to form an oxonium ion [methyl(pentan-3yl)oxonium], which reacts with the iodide counterion at the less substituted methyl group, via an SN2 reaction, to yield iodomethane and pentan-3-ol. While HBr reacts similarly, HCl is significantly less reactive and acids that are weaker than HCl generally do not react at all. Amines are amphoteric and react as acids, but only in the presence of very strong bases that generate a conjugate acid that is significantly weaker than the amine. Amines are acids in the presence of Grignard reagents (R]MgX) and organolithium reagents (R]Li, Sections 11.4.1 and 11.6.9), for example, which give amide bases (R2NM+) along with an alkane as the very weak conjugate acid (R]H). In a typical acid-base reaction, butyllithium is a strong base that reacts with diethylamine, which is the acid, and the conjugate base is lithium diethylamide (Et2NLi) and the butane is the conjugate acid. Similarly, diisopropylamine yields lithium diisopropylamide (i-Pr2NLi, often called LDA) and butane, and 2,2,6,6-tetramethylpiperidine yields lithium 2,2,6,6-tetramethylpiperidide and butane.36 Butane has a pKa of >40, and amines have pKa values 35, so the amine is a relatively strong acid in this system and formation of the amide base is straightforward. These so-called amide bases are very strong bases indeed, and will be used in several applications in Chapters 11 and 12, and especially in Chapter 13. 35

(a) Arnett, E. M.; Wu, C. Y. J. Am. Chem. Soc. 1960, 82, 4999; (b) Arnett, E. M.; Wu, C. Y. J. Am. Chem. Soc. 1962, 84, 1680, 1684; (c) Deno, N. C.; Turner, J. O. J. Org. Chem. 1966, 31, 1969.

36

Olofson, R. A.; Dougherty, C. M. J. Am. Chem. Soc. 1973, 95, 581.

71

2.3 LEWIS ACIDS

BuJLi N

N

+

H

Li

Diisopropylamine

Lithium diisopropylamide

BuJLi N

N

H 2,2,6,6-Tetramethylpiperidine

+

BuJH

BuJH

Li Lithium 2,2,6,6-tetramethylpiperidide

The parent amine can be considered as amphoteric, and triethylamine is a relatively strong organic base in a reaction with HCl, forming triethylammonium chloride. However, it is a very weak base in the presence of a weak acid (e.g., H2O), where the equilibrium favors the amine rather than the ammonium salt. Tertiary amines [e.g., triethylamine (NEt3) or pyridine (C5H5N, Py)] are probably the most commonly used as organic bases when an amine base is required. 1H-Imidazole is another amine base that is used in systems requiring a base that is weaker than NEt3 or Py. The stronger amide bases discussed above are used for deprotonation of weak acids (ketones, alkynes, esters, sulfides, etc., see Chapters 11–13). The amide bases are often called non-nucleophilic bases due to their poor reactivity as nucleophiles in SN2 (Sections 13.2.1 and 13.2.2) and in acyl addition reactions (Section 4.2). Their diminished nucleophilic strength results from the bulky alkyl substituents about the basic nitrogen, which inhibit approach to carbon, but do not encumber approach to hydrogen. Two other non-nucleophilic bases are 1,5-diazabicyclo[4.3.0]non-5-ene (DBN, which is 1,5-diazabicyclo[4.3.0]non-5-ene or 2,3,4,6,7,8-hexahydropyrrolo[1,2-a]pyrimidine)33b and DBU (1,8-diazabicyclo[5.4.0]undec-7-ene, or 2,3,4,6,7,8,9,10-octahydropyrimido[1,2-a]azepine from above),33a in which the bridgehead nitrogen is sterically encumbered. When a very powerful base is required for deprotonation (pKa of the acid is >25), bases such as an organolithium reagent are used most often (Section 11.6.9). Commercially available reagents [e.g., butyllithium (BuLi), methyllithium (MeLi), and phenyllithium (PhLi)] are commonly used in synthesis (see Section 11.6). N

N

H

N N

1 H–Imidazole

(1,5-Diazabicyclo[4.3.0]non-5-ene , DBN) 2,3,4,6,7,8-Hexahydropyrrolo[1,2-a]pyrimidine

The acid-base concepts described in this section focus on bases that react with protons, the classical Brønsted-Lowry definition of a base. There are other types of electron-deficient or electron-donating molecules that do not involve transfer of protons. Such compounds are called Lewis acids or bases, respectively. Section 2.3 will discuss Lewis acids.

2.3 LEWIS ACIDS There are two working definitions of acidity and basicity. A Lewis acid is an electron- pair acceptor and a BrønstedLowry acid is a proton donor. For basicity, a Lewis base is an electron-pair donor and a Brønsted-Lowry base is a proton acceptor. Section 2.2.1cused primarily on the Brønsted-Lowry definition, and reactions of protonic acids (H+). This definition contrasts with a Lewis acid, which is an atom other than hydrogen that accepts two electrons from an electron-donating species (the Lewis base). Perhaps the best-known Lewis acids are Group 13 atoms (e.g., Boron and Aluminum), which are electron deficient because they can form only three covalent bonds and remain neutral. They can satisfy the octet rule by accepting an electron pair from a Lewis base, forming an single product, a charged complex called an “ate” complex. Such complexes will be of particular importance in hydride reductions (Sections 7.2, 7.3, and 7.6) and in hydroboration reactions (see Sections 9.2 and 9.4). Acid

MXn + BBase ƒƒƒƒƒƒƒƒƒƒ!MXn  B +

}ate} Complex

Transition metal salts are usually characterized by the presence of a metal atom that is capable of assuming multiple valences, and in many cases they can function as a Lewis acid. The relative order of Lewis acidity is not always straightforward since an electrophilic metal can have different valences and ligands (the groups or atoms attached to the metal).

72

2. ACIDS, BASES, AND ADDITION REACTIONS

Lewis acids usually take the form MXn, where X is the ligand (a halogen atom, an amine or phosphine, CO, cyclopentadienyl, an alkene, etc.). The metal is M and n in Xn is the normal valency of M. The acid MXn exhibits acidic properties in an equilibrium reaction with the base B. Formation of a conjugate acid-base adduct (MX B+, the “ate” complex mentioned above) is characteristic of Lewis acid-base reactions. If chelation effects and double-bond formation between M and any given atom of the base are excluded, three general statements can be made concerning acid strength.37 1. In MXn (n < 4), acidity arises from the central atom’s requirement for completion of an outer-electron octet by accepting one or more pairs of electrons from the base. Acidity is diminished when two electron pairs are required. There is a smaller energy gain upon receipt of the first pair and an accumulation of negative charge on M if two pairs are received. Therefore, Group 13 acids are more acidic than transition metal acids: BF3 > AlCl3 > FeCl3. 2. The acidity of M will decrease within any group with increasing atomic volume (effectively, with increasing atomic number) owing to the weaker attraction between nuclear charge and incoming electron pairs (Fajans’ rules).37a The result of these effects leads to the order: BX3 > AlX3 > GaX3 > InX3 : 3. In general, the energies of different atomic orbitals lie closer together with increasing atomic number, because orbital contractions arising from the electronegativity of the nucleus tend to decrease with increasing atomic number. As a result, there is more effective overlap of hybridized orbitals. The availability of d-orbitals will be easier (especially with d outer orbitals) and more effective, the heavier the element. This availability of d-orbitals leads to a decrease in acidity, as in the series (B ! Al ! Ga ! In) in statement 2. These three statements can be used to estimate the strength of Lewis acids according to their group in the periodic table.37S a. Groups 11 (Cu, Ag, Au) and 12 (Zn, Cd, Hg) usually give covalent species MX2: (ZnCl2, CdI2, etc.); M has four electrons in the outer group and needs two additional pairs. By statement 1, these should have only moderate acidity and by statement 2: ZnCl2 > CdCl2 > HgCl2. b. Groups 13 (B, Al, In, Tl) and 3 (Sc, Y, La, Ac) by statement 1 tend to form the most acidic compounds (MX3). The metal (M) requires only one electron pair to complete the octet (BF3, AlCl3, GaBr3, etc.). Statement 2 is then applied. BF3 > AlCl3 > GaBr3 > InCl3 > TlCl3 and ScCl3 > YCl3 > LaCl3 c. Groups 14 (C, Si, Ge, Sn, Pb) and 4 (Ti, Zr, Hf ) form covalent species (e.g., MX4). There are no acidic properties when M ¼ C, and when M ¼ Si only weak acidic properties are observed. d. Groups 15 (N, P, As, Sb, Bi) and 5 (V, Nb, Ta) usually take the form MX3 or MX5. When M is nitrogen, these compounds are not Lewis acids, and indeed they are bases. When M is phosphorus the molecule is a weak base, but the octet can expand to yield weakly acid properties in some reactions. In general, MX5 compounds are more acidic than the Groups 15 and 5 MX3 compounds due to the expanded octet and the utilization of d orbitals. Note that both SbX5 and NbX5 are powerful Lewis acids. e. Groups 16 (O, S, Se, Te, Po) and 6 (Cr, Mo, W) have the normal covalent character of MX2. There are close parallels with Groups 14 and 4. Both OX2 and SX2 have virtually no acidity and TeX2 is very weak. Species arising from abnormal valences, (e.g., TeX4), can exhibit enhanced acidity since only one electron pair is then required to complete the outer shell of 12. f. Groups 17 (F, Cl, Br, I, At) and 7 (Mn, Tc, Re) resemble 14 and 4 and 16 and 6 in that the lower members are only weakly acidic. The higher members can expand the octet when attached to electron-withdrawing substituents, and  38 have enhanced acidity (I 3 and PhdICl ). With these six general statements as a guide, it is possible to order the more common Lewis acids by decreasing acid strength: BX3 > AlX3 > FeX3 > GaX3 > SbX5 > InX3 > SnX4 > AsX5 > SbX3 > ZrX4

37

(a) Satchell, D. P. N.; Satchell, R. S. Chem. Rev. 1969, 69, 251; (b) Idem Q. Rev. Chem. Soc. 1971, 25, 171.

38

Andrews, L. J.; Keefer, R. M. J. Am. Chem. Soc. 1952, 74, 4500.

2.4 HARD-SOFT ACID-BASE THEORY

73

The relative strength of Lewis acids is important to many major classes of organic reactions utilized in synthesis. In Sections 11.6.1 and 12.4, Lewis acids will be used in connection with several important carbon-carbon bond forming reactions, including Friedel-Crafts type reactions39 (Sections 16.4 and 16.5) and Diels-Alder reactions,40,41,42 (Section 11.5.1), as well as other pericyclic reactions (Chapter 15). Changing the Lewis acids can have dramatic effects on the rate of reaction, regio- and stereoselectivity, and product distribution in both the Diels-Alder and Friedel-Crafts reactions.39 If the Lewis acid is too reactive, degradation of products or starting materials can be a serious problem. For these reasons, several Lewis acids covering a wide range of acid strengths should be available for a synthesis. Aluminum chloride is a very reactive and unselective catalyst, reacting with virtually all functional groups that have Lewis base properties.41,43 Zinc chloride (ZnCl2), however, is a mild and selective catalyst in reactions, where halides or alcohols are required to react selectively with an olefinic (alkenyl) double bond.39,44 Tin tetrachloride (SnCl4) is a very mild catalyst that can be used for acylation of reactive aromatic nuclei (e.g., thiophene). Aluminum chloride is too strong for use in the catalysis of reactions involving thiophene and extensive decomposition accompanies the process.39,45 Tin tetrachloride has also been applied to Friedel-Crafts reactions of active aromatics,39,46 which cannot tolerate vigorous reaction conditions, as with furan derivatives, for example.39,47 These few examples clearly demonstrate that a good working knowledge of the relative acid strength of Lewis acids is essential for understanding many important reactions in synthesis.

2.4 HARD-SOFT ACID-BASE THEORY 2.4.1 Hard and Soft Acids and Bases There is an alternative view of reactions that relies on the Lewis acid-base definition, and is used to classify a wide range of organic reaction types. It is known as Hard-Soft Acid-Base Theory.48,49 Cations are classified as Lewis acids, anions are Lewis bases, and salts are viewed as acid-base complexes. When one writes the normal acid-base equation, certain conventions are observed. A0 + A  Bƒƒƒ!A0  B + A implies that A0 is a stronger acid than A, and B0 + A  Bƒƒƒ!A  B0 + B implies that B0 is a stronger base than B. For A + Bƒƒƒ!A  B where the equilibrium constant is defined by log K ¼ ½SA  ½SB  and the S terms represent the relative strengths, which are functions of the environment and the temperature. There are solvent effects that make it difficult to establish a universal order of acid or base strength50 due to variations in the parameter S, as well as other factors. 39

Olah, G. A., Friedel Crafts and Related Reactions; Interscience Publishers: NY, 1963; Vol. I, pp 201–366.

40

(a) Inukai, T.; Kasai, M. J. Org. Chem. 1965, 30, 3567; (b) Iukai, T.; Kojima, T. J. Org. Chem. 1967, 32, 869, 872.

41

For references dealing with coordination of the Lewis acid to ester dienophiles, see (a) Lappert, M. F. J. Chem. Soc. 1961, 817; (b) Lappert, M. F. J. Chem. Soc. 1962, 542.

42

(a) Houk, K. N.; Strozier, R. W. J. Am. Chem. Soc. 1973, 95, 4094; (b) Alston, P. V.; Ottenbrite, R. M. J. Org. Chem. 1975, 40, 1111; (c) Ansell, M. F.; Nash, B. W.; Wilson, D. A. J. Chem. Soc. 1963, 3012.

43

Thomas, C. A. Anhydrous Aluminum Chloride in Organic Chemistry; Reinhold Publishers: NY, 1941.

44

Nenitzescu, C. D.; Isacescu, D. A. Berichte 1933, 66, 1100.

45

Goldfarb, I. J. Russ. Phys. Chem. Soc. 1930, 62, 1073 (Chem. Abstr. 25: 2719, 1931).

46

Stadnikoff, G.; Banyschewa, A. Berichte 1928, 61, 1996.

47

Gilman, H.; Calloway, N. O. J. Am. Chem. Soc. 1933, 55, 4197.

48

(a) Pearson, R. G. Hard and Soft Acids and Bases; Dowden, Hutchinson and Roe Inc.: Stroudsburg, PA, 1973; (b) Pearson, R. G. J. Am. Chem. Soc. 1963, 85, 3533; (c) Pearson, R. G.; Songstad, J. J. Am. Chem. Soc. 1967, 89, 1827; (d) Idem J. Org. Chem. 1967, 32, 2899.

49

(a) Pearson, R. G. J. Chem. Educ. 1968, 45, 581; (b) Idem J. Chem. Educ. 1968, 45, 643.

50

(a) Lewis, G. N. Valence and the Structure of Atoms and Molecules; The Chemical Catalog Co.: NY, 1923; (b) Lewis, G. N. J. Franklin Inst. 1938, 226, 293.

74

2. ACIDS, BASES, AND ADDITION REACTIONS

If the acid (A) is chromic acid, for example, there is a different basicity order than if A is cuprous ion. Usually, the simple equation shown above is inadequate and requires additional terms log K ¼ ½SA ½SB  + ½sA ½sB 

(2.1)

where s is a parameter for the strength of each acid and base, and is called the softness parameter. The Edward’s equation51 (Eq. 2.2) can be of value for determining the relative strength of Lewis acids and is written as follows: log

K ¼ αEn + βH K0

(2.2)

where H is a proton basicity factor: H ¼ 2.74 + pKa for water, at 25°C. The term En is a redox factor and is equal to E° + 2.60 [En ¼ E° + 2.60), where E° is the standard oxidation potential for the base, defined by the reaction: B ƒƒƒ!B + + 2e There are obviously many acids that could be used to define strong and weak, but the proton (H+) and an alkyl mercuric ion (RHg+) have roughly opposite affinities for binding bases. The values of the parameters for the Edward’s equation for H+ are α ¼ 0.000 and β ¼ 2.000 and for MeHg+ the values are α ¼ 5.786 and β ¼  0.031.49,52 Therefore, the reactions of these two acids with Lewis bases define the relative strengths of the Lewis bases, which gives rise to two observations. (1) Bases in which the donor atom is O, N, or F prefer to coordinate to the proton. (2) Bases in which the donor atom is P, S, I, Br, Cl, or C prefer to coordinate to mercury. In Group 1, the donor atoms are highly electronegative, have a low polarizability, and are hard to oxidize. These are called HARD BASES (valence electrons are bound tightly). In Group 2, the donor atoms are of low electronegativity, are highly polarizable, and are easy to oxidize.48c These are called SOFT BASES (valence electrons are loosely bound).48c Recognition of these categories form the basis of the HSAB (hard-soft acid-base)principle, initially developed by Pearson48 and also described by Ho.52 In separate reactions with hard bases, there are two distinct classes of acids called HARD ACIDS and SOFT ACIDS. For HARD ACIDS,48c acceptor atoms are small in size, have a high positive charge, and do not contain unshared pairs of electrons in their valence shell. They are of low polarizability and of high electronegativity. For SOFT ACIDS,48c acceptor atoms are large in size, have a low positive charge, and contain unshared p or d electrons in the valence shell. They are highly polarizable and of low electronegativity. In order to rank these acids according to their relative strength, the criteria of Schwarzenbach53a and Ahrland and coworkers53b were used. Hard acids form complexes with the following relative stability:52 N O F

≫ ≫ >

P S Cl

> > >

As Se Br

> > >

Sb Te I

Similarly, soft acids react in the following manner to produce stable complexes:52 N O F

≪ ≪
<
<
R3 N R3 P ðsofterÞ I > Br > Cl > F ðharderÞ 2. The more p character, the softer the acid. ðsofterÞ sp3 C > sp2 C > sp C ðharderÞ 3. The more electron releasing groups on the moiety, the harder the acid. ðharderÞ ðCH3 Þ3 C + > ðCH3 Þ2 HC + > CH3 H2 C + > H3 C + ð softerÞ Metals require a minimum of a half-filled outer d shell to be good acceptors,55 emphasizing the importance of d electrons for metal ions. A comparison of the series Ca ! Zn suggests that their ions should become harder due to increasing nuclear charge. One also expects an increase in electronegativity, and therefore an increase in hardness. In fact, the increase in d electrons leads to increasing softness in this series. The softness of a base is defined by the equilibrium established in the reaction of MeHg+ and BH+: K

+ + MeHg + ðaqÞ + BH + ðaqÞ ƒƒƒ! ƒƒƒ MeHgB ðaqÞ + H ðaqÞ

If K ≫ 1, the base is soft. Conversely, if K is 1, the base is hard. Table 2.2 shows common hard and soft bases.49a,54,56 The nature of the outer groups on the acceptor atom is important. The hard acid BF3 possesses hard fluoride ions and readily adds to hard bases. This contrasts with BH3, which is a soft acid, where soft hydride ions readily add to soft anions. Soft bases tend to group together on a given central atom as do hard ligands. There is a mutual stabilizing effect called symbiosis.57 55

Ahrland S. Struct. Bonding (Berlin) 1966, 1, 207.

56

Reference 52, p 6.

57

Jorgensen, C. K. Struct. Bonding (Berlin), 1966, 1, 234.

76

2. ACIDS, BASES, AND ADDITION REACTIONS

In the case under consideration, the hard fluoride ligands form a complex that makes boron essentially B+ and hard. The soft hydride donates negative charge to the central boron via a covalent bond, or by simple polarization, making the boron more neutral and soft. It appears that the actual charge on the central atom, rather than the formal TABLE 2.2 Classification of Hard and Soft Basesa Hard Acids

Borderline Cases

Soft Acids

C6H5NH2, C 5H5N, N3 N2,

H –, R–, CH2KCH2, C6H6 , CN –,

O –, ROH, RO–, R2O, AcO–,

NO –, SO2 –, Br–,

RNC, CO, SCN–, R3P, (RO) 3P

CO3 –, NO3–, PO4 –, SO4 –,

R2S, RSH, RS–, S2O3 –, I–

NH3, RNH2, N2H4, H2O,

–OH,

–,

ClO 4–, F– (Cl –) a

In all cases, the molecules are listed in descending order (the first entry in each column is the hardest or softest). Reprinted with permission from Pearson, R.G. J. Chem. Educ. 1968, 45, 581. Copyright © 1968 American Chemical Society.

charge, determines the degree of softness. Pearson48a pointed out “while the HSAB principle has proved useful in many ways, it has been justifiably criticized because of the lack of a precise definition of hardness and the inability to assign numbers to this property.”58 The hardness parameter (η) is usually defined in terms of electronegativity, as in Eq. (2.3)59 η¼

1 @ 2 E 1 @M ¼ 2 @N 2 2 @N

(2.3)

where @E and @N are obtained from the Mulliken electronegativity (χ M).60 Pearson59 used an operational definition of hardness (Eq. 2.4). η¼

IA 2

(2.4)

In this definition, I is the ionization potential and A is the electron affinity. Note that a hard molecule will have a large value of η. For a reaction A + B ! A:B, the following expression can be used to decide which is the acid, A or B:58 ðIA  AB Þ  ðIB  AA Þ ¼ 2 ðX°A  X°B Þ

(2.5)

If this quantity is positive, transferring an electron from B to A requires less energy than from A to B. The A group is then a Lewis acid. The direction of electron flow can be determined by the difference in absolute electronegativity, whose magnitude is a measure of the driving force for the transfer.58 In most cases, there will be electron transfer in both directions. Eq. (2.6) yields the total energy when the amount of electron transfer in both directions is near unity (1). ðIA  AB Þ + ðIB  IA Þ ¼ 2 ðηA + ηB Þ

(2.6)

If both A and B are hard, then ηA and ηB will be large, and transfer of electrons in both directions will be disfavored. If A and B are soft, however, ηA and ηB will be small, favoring electron transfer.58 Table 2.359 provides typical values for electronegativity (X°) and for hardness (η) for several cations, anions, and neutrals. From Table 2.3, it is apparent that Al3+ (η ¼ 45.8 eV) is a hard acid, whereas Ag+ (η ¼ 6.9 eV) is a soft acid. To illustrate neutral molecules, hydrogen (H2) is harder than iodine (I2), and chloromethane (η ¼ 7.5 eV) is harder than iodomethane (η ¼ 4.7 eV). Trimethylamine (η ¼ 6.3) is a harder base than trimethylphosphine (η ¼ 5.9).

2.4.2 The HSAB and Molecular Orbital Theory Frontier molecular orbital (FMO) theory (Section 14.2) correlates molecular orbital energies and orbital coefficients with the relative reactivity of molecules. In a molecule, the orbital containing the highest energy bonding electrons is called the highest occupied molecular orbital (HOMO) and the lowest lying orbital that does not contain electrons, but can 58

Pearson, R. G. J. Chem. Educ. 1987, 64, 561.

59

Parr, R. G.; Pearson, R. G. J. Am. Chem. Soc. 1983, 105, 7512.

60

Mulliken, R. S. J. Chem. Phys. 1934, 2, 782.

77

2.4 HARD-SOFT ACID-BASE THEORY

TABLE 2.3 Electronegativity and Hardness Parameters Ion/Molecule

X° (eV)

h (eV)

Ion/Molecule

X° (eV)

h (eV)

CATIONS

Na+

26.2

21.1

Fe +

23.4

7.3

Ag+

14.6

6.9

Zn

+

28.8

10.8

I+

14.8

4.3

Pd +

26.2

6.8

Mn +

24.4

9.3

Hg +

26.5

7.7

Ca +

31.6

19.7

Al +

74.2

45.8

4.4

6.0

4.4

6.1

NEUTRALS BF3

7.8

7.8

SO3

7.2

5.5

N2

7.0

8.6

H2S

4.3

6.4

Cl2

7.0

4.6

C 6 H6

4.0

5.2

SO2

6.7

5.6

PH3

4.0

6.0

H2

6.7

8.7

H2O

3.1

9.5

O2

6.3

5.9

MeCl

3.8

7.5

CO

6.1

7.9

NH3

2.9

7.9

C 4H5N

I2

6.0

3.4

PMe3

2.8

5.9

Pt

5.6

3.5

Me2S

2.7

6.0

MeI

4.9

4.7

Me2O

2.0

8.0

HCl

4.7

8.0

NMe3

1.5

6.3

ANIONS F–

3.40

7.0

Br–

3.36

4.2

HO –

1.83

5.7

I–

3.06

3.7

5.3

H–

0.74

6.8

4.9

MeS–

1.9

3.1

4.7

t-Bu–

-0.3

3.6

1.1

4.1

3.1

3.9

H2N – H3C – Cl – HS – H2P –

0.74 0.08 3.62 2.30 1.25

4.1

C 6H 5

4.3

–

NO2

–

Reprinted with permission from Parr, R.G.; Pearson, R.G. J. Am. Chem. Soc. 1983, 105, 7512. Copyright © 1983 American Chemical Society.

receive them is the lowest unoccupied molecular orbital (LUMO). Both HOMO and LUMO are labeled in Fig. 2.1.58 The difference in energy between these orbitals determines the reactivity of many types of reactions. Experimentally, the energy of the HOMO is the ionization potential (I) and the energy of the LUMO is the electron affinity (A), leading to the equations: I ¼ εHOMO and A ¼ εLUMO In Fig. 2.1, the electronegativity (X°) is taken as the midpoint (average) between the HOMO and LUMO energies.48 If the HOMO is 10 eV and the LUMO is +1 eV, then the midpoint is 5.5 eV, and the value of X° is 1  5.5 eV ¼  4.5 eV. The parameter η will then be 5.5 eV (see Fig. 2.1), which allows one to define a hard molecule or ion as having a large energy gap between the HOMO and LUMO (ΔHOMO-LUMO).58 A soft molecule or ion will have a small ΔHOMO-LUMO.58 Soft acids and bases have properties that lead to a high-energy HOMO and a low-energy LUMO (a small energy gap).

78

2. ACIDS, BASES, AND ADDITION REACTIONS

Hard acids and bases have the opposite properties: a low-energy HOMO and a high-energy LUMO (a large energy gap). This generalization can be applied to cations, anions, and radicals, as well as neutral molecules. Radicals contain a singly occupied molecular orbital (SOMO) as shown in Fig. 2.2.58 The electron affinity term in this case arises from adding a second electron to the SOMO and X° is the average energy of the two electrons in that orbital. The quantity I-A is the average value of the interaction energy of the two electrons.61

+1 eV

LUMO

0 eV

0 eV

h

E

−X

–5 eV h

–10 eV

–10 eV CH2− HOMO

FIG. 2.1 Correlation of HOMO and LUMO with electronegativity and hardness. Reprinted with permission from Pearson, R.G. J. Chem. Educ. 1987, 64, 561. Copyright © 1987 American Chemical Society.

LUMO 0 eV

E h

−X SOMO

FIG. 2.2 Energy diagram for radicals. Reprinted with permission from Pearson, R.G. J. Chem. Educ. 1987, 64, 561. Copyright © 1987 American Chemical Society.

The HOMO-LUMO energy gap is usually the lowest energy electron absorption band. Soft acids and bases tend to absorb light closer to the visible region than hard acids and bases. Alkenes are chromophoric and it is typical for them to have high-energy HOMOs and low-energy LUMOs. Note that a chromophore is a group that selectively absorbs light, resulting in formation of a color. In general, unsaturation increases softness. It is also apparent that soft molecules are more reactive than hard molecules (due to a smaller energy gap I-A) in unimolecular and bimolecular reactions. These principles allowed Pearson to define hardness (η) as follows:58 A hard molecule resists changes in its electron charge cloud, both the total amount of charge and the charge distribution in space. A soft molecule has an easily changed electron distribution.

2.5 ACID-BASE REACTIONS OF ALKENES AND ALKYNES (ADDITION REACTIONS) Traditional acid-base reactions (e.g., the reaction of HCl and NaOH) are relatively easy to identify. It is possible to describe many organic reactions as acid-base reactions if the definition is expanded to include very weak acids and very strong bases, or very weak bases with very strong bases. In this section, the π-bond of an alkene or an alkyne will be used as a weak base, primarily in reactions with strong mineral acids, Lewis acids, and with other electrophilic atoms. The π-bond of an alkene reacts as a Brønsted-Lowry base with a protonic acid, donating an electron pair to 61

Klopman, G. J. Am. Chem. Soc. 1964, 86, 1463.

2.5 ACID-BASE REACTIONS OF ALKENES AND ALKYNES (ADDITION REACTIONS)

79

the proton. A new CdH σ-bond is formed to one carbon of the former C]C unit as the π-bond is broken. Both electrons of the π-bond are used to form the CdH bond, leaving the other carbon of the C]C unit electron deficient. This carbon atom becomes a carbocation that is subject to further reactions. Remember that a cation is an electron-deficient species that has a formal charge of +1. A carbon atom that bears a positive charge is called a carbocation (also known as a carbenium ion). The central carbon atom of a carbocation is clearly electron deficient, and this charged species has only three covalent bonds, is high in energy, unstable, highly reactive, and difficult to isolate in most cases. In other words, it is an intermediate. R

R C R A generic carbocation

The electron-deficient carbon of a carbocation is sp2 hybridized, so it has trigonal-planar geometry. The positive charge is localized on carbon, and that charge is associated with an empty porbital on that carbon. An empty p-orbital is considered to be the region in space above and below the plane of the carbon and attached atoms where electron density can be accepted to form a new bond. A carbocation will react with another species that can donate two electrons to form the fourth bond to make carbon tetravalent and give it eight electrons, which satisfies the valence requirements of carbon. In other words, a carbocation can react with a nucleophile. The traditional category for a reaction in which a new atom or group is transferred to each carbon of the π-bond, forming two new sp3 bonds as the π-bond is broken, which constitutes an addition reaction since the atoms are “added” to the π-bond.62 Traditionally, addition reactions have been subdivided into three mechanistic categories. (1) Those involving free and metal-stabilized ions as intermediates. (2) Those involving symmetrically bridged ions as intermediates. (3) Addition via a four-center transition state. In this section, addition reactions will be discussed as either Brønsted-Lowry acid-base reactions of Lewis acid-base reactions. The main examples of four-center transition state additions are hydroboration. If an alkene reacts with BH3, for example, the alkene reacts as a Lewis base and the boron is the Lewis acid, generating an organoborane product. Reactions of alkenes with boron compounds will be discussed in Chapter 9.

2.5.1 Alkenes and Alkynes as Brønsted-Lowry Bases The π-bond of an alkene reacts as a Brønsted-Lowry base in the presence of a protonic acid (e.g., HCl or HBr). An alkene only reacts with relatively strong acids, so it is considered to be a weak base. Indeed, alkenes do not react with water, alcohols, amines, and even carboxylic acids react slowly or not at all if the pKa is >4–5.63 Acids with a pKa  2–3 usually react quickly, and mineral acids (e.g., HCl, HBr, or HI) react rapidly. A simple example is the reaction of cyclopentane with HCl. Initial reaction of the π-bond leads to a highly reactive carbocation intermediate (cyclopentylium chloride), which reacts with the nucleophilic chloride ion (also see Section 3.2.2) to yield the final product, chlorocyclopentane. Note that a catalytic amount of a mineral acid such as sulfuric acid can be used to generate the reactive carbocation in situ, which then reacts with even weak nucleophiles such as water or an alcohol. It is usually assumed that the initial reaction of an alkene with a Brønsted-Lowry acid will form of a solvent-separated ion pair, but this mechanistic detail depends on the solvent. Nucleophiles will react rapidly with solvent separated carbocations, but it is known that a so-called tight ion pair intermediate is a possible intermediate, and it can react in the substitution step to give the same product.

H—Cl Cyclopentene

62 63

H Cl− Cyclopentylium

H

Cl Chlorocyclopentane

For examples see Ref 2, pp 629–666, 912–915.

(a) Peterson, P. E.; Tao, E. V. P. J. Org. Chem. 1964, 29, 2322; (b) LaLonde, R. T. J. Org. Chem. 1962, 27, 2276; (c) Alo see Isenberg, N.; Grdinic, M. J. Chem. Educ. 1969, 46, 601.

80

2. ACIDS, BASES, AND ADDITION REACTIONS

The net result of this reaction is addition of H and Br across the π-bond, which is the derivation of the traditional name “addition reaction.” Since the sp2 carbon of a carbocation is planar, allowing the nucleophilic bromide to approach from either face, the addition reaction is expected to yield a racemic product. As noted above, alkenes are weak Brønsted-Lowry bases and react only with strong acids HX (HCl, HBr, H2SO4, HNO3, HClO4, HBF4, etc.), which react with alkenes to form an alkyl halide, sulfate, nitrate, perchlorate, or tetrafluoroborate, and so on. However, the hydrogen sulfate, nitrate, perchlorate, and tetrafluoroborate anions are very weak nucleophiles and react very slowly with the cation intermediate. If the product does form, alkyl hydrogen sulfates, nitrates, or perchlorates are often unstable, and undergo elimination reactions (Sections 3.5 and 3.6) or decompose. In some cases the instability of the products means that they cannot be formed at all, which suggests that the carbocation may be the dominant species in situ. In such cases, another nucleophile can be added that will react preferentially with the cation. The nucleophile can be water (it may come from an aqueous solvent), and reaction with the carbocation intermediate will yield an alcohol product. If a more nucleophilic reagent (e.g., KCN or NaN3) is added to the reaction, a nitrile or an azide will be formed (Section 3.2.2). From a synthetic viewpoint, this option greatly expands the utility of the reaction, allowing introduction of a nucleophilic unit (Y) that is not a part of the initial acid (HX). A simple example of this concept is the hydration reaction of methylcyclohexene. An initial acid-base reaction with a catalytic amount of sulfuric acid in an aqueous medium generates the tertiary carbocation (methylcyclohexan-1-ylium hydrogen sulfate), which reacts with the best and most prevalent nucleophile. In this case at is water. The nucleophilic atom is the oxygen of water, and reaction with C+ leads to the oxonium salt intermediate, (1-methylcyclohexyl) oxonium hydrogen sulfate. Loss of a proton from the acidic oxonium ion (pKa   2), in an acid-base reaction, yields the final product, 1-methylcyclohexanol. In other words, the oxonium ion is such a strong acid that water or unreacted methylcyclohexane will react as the base. The overall reaction generates an alcohol from an alkene, where water is the nucleophile. Remember that alkenes do not react with water in the absence of the acid catalyst, which is necessary to generate the requisite carbocation intermediate that reacts with water. OH2

cat H2SO 4 −

H2O

Methylcyclohexene

HSO4

Methylcyclohexan-1-ylium hydrogen sulfate − H+

O H

OH

H HSO4− (1-Methylcyclohexyl)oxonium hydrogen sulfate

1-Methylcyclohexanol

The reaction of mineral acids with alkenes was discovered long before the mechanism of this reaction was understood. It was observed that reactions of alkenes and acids always produced the more substituted product. In other words, addition of HCl, HBr, or HI to substituted alkenes always led to an alkyl halide, where the hydrogen atom was attached to the less substituted carbon and the nucleophile was attached to the more substituted carbon of the initial π-bond. Without understanding a carbocation mechanism, this reaction was known as Markovnikov addition,64,63c of simple alkenes, in honor of Vladimir Markovnikov who discovered this reaction. Nowadays, the reaction of an alkene with an acid HX is said to be highly regioselective. The basis of this regioselectivity is, of course, formation of the more stable carbocation, which is the normal course of all reactions with an acid and an alkene if the intermediate is a carbocation. A simple example is the reaction of 2-methylpent-2-ene with HBr, which gave a mixture of 2-bromo-2-methylpentane and 3-bromo-2-methylpentane, but 2-bromo-2-methylpentane is the major product, often >80–90% of the product distribution. Br 2-Methylpent-2-ene

+

2-Bromo-2-methylpentane MAJOR

Br 3-Bromo-2-methylpentane MINOR

In the reaction of unfunctionalized alkenes with acids (e.g., HX), the key to predicting the regiochemistry is to identify the more stable carbocation (or carbon radical) intermediate. The presence of a functional group or a heteroatom substituent can lead to a different outcome by stabilizing or destabilizing the carbocation. Generating a carbocation on a 64

Markovnikoff, V. Compt. Rend. 1875, 81, 668.

2.5 ACID-BASE REACTIONS OF ALKENES AND ALKYNES (ADDITION REACTIONS)

81

carbon atom that has a positive charge or a δ+ dipole is very unstable, for example, due to the repulsive interaction of like charges. Hydrobromic acid reacted with the conjugated π-bond of methyl acrylate to yield methyl 3-bromopropanoate as the final product, an anti-Markovnikov addition.65 This result can be predicted by comparing the possible carbocation intermediates. Addition of H+ to the π-bond generates the primary carbocation 3-methoxy-3-oxopropan-1-ylium as the precursor to methyl 3-bromopropanoate. Alternatively, reaction to yield the secondary carbocation 1-methoxy-1-oxopropan-2-ylium is disfavored because the positive charge is adjacent to the δ+ carbon of the carbonyl. Such an intermediate is destabilized and formation of 3-methoxy-3-oxopropan-1-ylium will be highly favored under these conditions. HBr

CO2Me

CO2Me

0°C

1-Methoxy-1-oxopropan-2-ylium

Methyl acrylate + Br−

CO2Me 3-Methoxy-3-oxopropan-1-ylium

Br

CO2Me

Methyl 3-bromopropanoate

The reaction of an alkene with HCl is possible without using the mineral acid directly. A mixture of ethanol and acetyl chloride or chlorotrimethylsilane in water, for example, will convert an alkene to the corresponding alkyl halide. The reaction of (R)-2-methyl-5-(prop-1-en-2-yl)-cyclohex-2-en-1-one [carvone] with acetyl chloride in ethanol, for example, gave a 96% yield of (R)-5-(2-chloropropan-2-yl)-2-methylcyclohex-2-en-1-one.66 This transformation also illustrates the poor reactivity of the conjugated C]C unit relative to the unconjugated alkene. In a different experiment, careful addition of H2SO4 to NaCl will generate HCl gas, which will react in situ with an added alkene. In all cases, the carbocation mechanism for addition of acids (e.g., HX) to an alkene yields the more stable carbocation, and the substitution product will result from trapping that carbocation with the nucleophile. Generation of HBr by similar methods is complicated by formation of Br2, which often contaminates the HBr, although it can be removed by passage through copper turnings. Cl

AcCl/EtOH rt , 1.5 h

O

O (R)-2-Methyl-5-(prop-1-en-2-yl)cyclohex-2-en-1-one [Carvone]

(R)-5-(2-Chloropropan-2-yl)-2methylcyclohex-2-en-1-one (96%)

Once a carbocation is formed as a result of an acid-base reaction, it is subject to a skeletal rearrangement that usually occurs before any reaction with a nucleophile. This rearrangement is a chemical reaction that is under thermodynamic control, and tends to yield the more stable carbocation, that leads to the final major product. An example is the reaction of 3-methylpent-1-ene with HCl, which gave 3-chloro-3-methylpentane as the major product rather than 2-chloro-3-methylpentane, would be formed if the initially formed secondary carbocation (3-methylpentan-2-ylium chloride) reacted with the chloride ion. Although 3-methylpentan-2-ylium chloride is indeed the initially formed carbocation, the observed product (3-chloro-3-methylpentane) can only be formed from a different carbocation, 3-methylpentan-3-ylium chloride. The explanation for this experimental observation is to say that 3-methylpentan-2-ylium chloride, which is a secondary carbocation, rearranges to 3-methylpentan-3-ylium chloride, a tertiary carbocation. 1,2–H Shift

H—Cl

3-Methylpent-1-ene

3-Methylpentan-2-ylium Cl−

3-Methylpentan-3-ylium 65

Mozingo, R.; Patterson, L. A. Org. Synth. Coll. 1955, 3, 576.

66

Yadav, V. K.; Babu, K. G. Eur. J. Org. Chem. 2005, 452.

Cl

3-Chloro-3-methylpentane

82

2. ACIDS, BASES, AND ADDITION REACTIONS

It is known that a secondary carbocation is more stable than a primary carbocation by 12–15 kcal (50–63 kJ) mol1 and a tertiary carbocation is more stable than a secondary carbocation by 12–15 kcal (50–63 kJ) mol1.67 Therefore, the rearrangement shown is an exothermic process by 12–15 kcal (50–63 kJ) mol1. This migration of a hydrogen atom from one carbon to the adjacent carbon is formally a rearrangement, and this chemical reaction is labeled as a 1,2-hydride shift. The 1,2 number simply means that the hydrogen atom migrates from one carbon to an adjacent carbon. In reality, the hydrogen atom does not migrate by itself, but rather the bonding electrons to the hydrogen atom move in a process defined as a 1,2-hydride shift. The energetic driving force for this reaction is conversion of a less stable secondary carbocation to a more stable tertiary carbocation. The sequence shown for the conversion of 3-methylpent-1-ene to 3-chloro-3-methylpentane, showing all intermediates, constitutes the mechanism of the reaction. An explanation is in order for the statement that the hydrogen atom moves, but the electrons in the bond connecting the migrating hydrogen atom to the carbon move and essentially take the hydrogen atom along. Note the single bond between the sp2 hybridized carbon atom (C+) and the adjacent sp3 carbon. Rotation about this C]C single bond is possible (see Section 1.5.1) and in one rotamer, the p-orbital of the positively charged carbon atom is parallel to the adjacent sp3 hybrid orbitals of the C]H bond. The substituent can only be transferred if the electrons in the σ bond are parallel to the p orbital of the carbocation.68 If we use the formation of 3-chloro-3-methylpentane as an example, the bonding electrons in the adjacent and parallel sp3 orbital of the C]H bond in the initially formed carbocation, 3-methylpentan-2-ylium, begin to migrate toward the electron-deficient carbon. This migration leads to transition state 4, with an empty p-orbital on the tertiary carbocation with an adjacent sp3 C]H bond. As the electrons in the bond continue to migrate, the hydrogen atom is carried along as the bonding electrons migrate, and the initial sp3 bond begins to rehybridize to a p-orbital in the new carbocation, 3-methylpentan-3-ylium, and the initial orbital begins to rehybridize to a new sp3 orbital. This migration is a chemical reaction, and the migration generates a more stable (more highly substituted) carbocation from a less stable carbocation, so it is exothermic by 12–15 kcal (50–63 kJ) mol1. H

H

H

H Me 3-Methylpentan-2-ylium

Me

Me

Me

H

Me 4

Me

H

3-Methylpentan-3-ylium

The rearrangement mechanism can be expressed in terms of the rate of the rearrangement reaction (krearr) versus the rate of reaction with the chloride ion (kCl). If 3-chloro-3-methylpentane is the major product from above, the rearrangement to the tertiary carbocation from the initially formed secondary carbocation must be faster than the reaction of the chloride ion with the secondary carbocation. In other words, krearr > kCl (see Section 16.2.3). The rearrangement just described is not limited to migration of a hydrogen atom. As well as hydrogen shifts, alkyl groups can move (called 1,2-alkyl shifts) and aryl groups can be transferred (called 1,2-aryl shifts). An example is the rearrangement of 3,3-diethylpentan-2-ylium to 3-ethyl-4-methylhexan-3-ylium. Interestingly, in 2,3-dimethylbutan-2-ylium, hydrogen was shifted in preference to a bulkier methyl group, but in 3,3-diethylpentan-2-ylium one of the adjacent carbon atoms has three hydrogen atoms and the other carbon has three attached ethyl groups. In this case, migration of a hydrogen from the methyl group would yield a less stable primary carbocation, and is not a competitive process. To form a more stable tertiary carbocation, one of the ethyl groups must migrate and the 1,2-ethyl shift occurs to yield 3-ethyl-4-methylhexan-3-ylium. In general, however, the smaller substituent migrates preferentially, as in hydrogen versus ethyl in 2,3-dimethylbutan-2-ylium. The energy barrier to transfer a hydrogen atom is lower than that for alkyl or aryl shifts, although the energy barrier for a 1,2-methyl shift is estimated to be CH3CO2H

OH CH3CO2H In water

>

CH3CO2H

In THF

OMe MeO

2. Offer a mechanistic explanation for the following transformation: HO OH

Amberlyst-15 CHCl3 , –20°C

O

Amberlyst is an acidic resin

3. The following reaction yields a mixture of the alcohol and the ether, in the proportions given later in this question. Only 0.03 equiv of oct-1-ene were used. In each case, a 1:1 ratio of water to alcohol was used. 1: HgðOAcÞ2 , H2 O=ROH Sodium dodecyl sulfate C6 H17 CH ¼ CH2 ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ! C6 H17 CHCH3 + C6 H17 CHCH3 2: NaBH4 j j OH OR The reaction with 1.0 equiv of octan-1-ol (to 0.03 equiv of oct-1-ene) gave a 98:2 mixture of alcohol/ether in 50% aq. THF. The use of 10 equiv of octan-1-ol in 0.3 M THF in water gave a 52:48 mixture of alcohol/ether. Explain the implications of the 52:48 mixture for the relative nucleophilicity of water and octan-1-ol. Explain why increasing the amount of octan-1-ol leads to more ether product.

93

2.6 CONCLUSION

4. Two products are formed in the following reaction, A and B. (a) Offer a reasonable explanation of how B could arise in this reaction. (b) Would the use of the 4-chlorophenyl derivative, rather than the 4-methoxy derivative help or hurt formation of B? Why or why not?

Cl

Cl2

Cl

MeO

Cl

MeO

Cl

+ MeO

A

B

5. Explain each of the following: (a) H3N+CH2CO2H is a stronger acid than H3N+(CH2)4CO3H (b) O2N

is a stronger acid than

OH

OH

6. Offer a mechanistic explanation for the following transformation: H

H H H

NBS , CH2Cl2

OH

H

Br

H

–25°C to rt

O

7. Calculate the pKa given the following values for Ka: (a) 6.35  106 (b) 12.1  107 (c) 18.5  1012 (d) 9.2  103 (e) 10.33  108 (f ) 0.08  103 8. Calculate the Ka given the following values for pKa: (a) 6.78 (b) 3.2 (c) 23.5 (d) 10.3 (e) 35.8 (f ) 11.1 9. Draw complete reactions for (a) propanoic acid and (b) methanesulfonic acid with (i) NaOH, and then (ii) NaNH2, showing all starting materials and all products. (a) Propanoic acid (b) Methanesulfonic acid 10. Careful analysis shows that carboxylic acid A is more acidic than carboxylic acid B. Suggest a reason for this observation.

Me

Cl

Me

H

CO2H

CO2H

H A

Cl

B

94

2. ACIDS, BASES, AND ADDITION REACTIONS

11. Draw all resonance structures for those anions that are resonance stabilized and indicate which are not resonance stabilized.

O

O

(A)

Et

O−

−O

(B)

O−

S

Et

(C)

O

O−

(D)

O

−NMe

(E)

−O

(F)

Cl O O

12. Alcohols are known to be amphoteric. Predict whether propan-1-ol will be an acid or a base or will be neutral in the presence of each of the following: (a) NaOH (b) HCl (c) Water (d) Ethanol (e) NaNH2 (f ) Butan-2-one (g) Methane (h) H2SO4 13. In a solvent that is not capable of hydrogen-bonding, maleic acid (A) is known to be more acidic than fumaric acid (B). Explain!

HO2C

HO2C

CO2H

CO2H

A

B

14. Which of the following should react faster with BF3, trimethylamine, or trimethylarsine (Me3As)? Explain! 15. Draw the product expected from a reaction of mercuric chloride (HgCl2) and dimethylamine. 16. Give the products formed in each of the following reactions:

(A)

HI

Br2

(B)

CCl4

(C)

cat H+

I2 CCl4

(D) H2O , Ether

cat H+

(E)

(F)

HCl , THF

H2O/Ether

17. Briefly discuss why 2,3-dimethylbut-2-ene might react faster with HCl than with but-2-ene. Draw the mechanism for both reactions as part of your answer.

95

2.6 CONCLUSION

18. Give the major product of the following reactions: HCl

(A)

HBr

(B) Br2 , CCl4

cat p-TsOH

(C)

(D) EtOH

(E)

HCl

I2 , CCl4

(F)

HOCl

(G)

Br2 , CCl4

(H)

aq THF

HCl

(J)

cat p-TsOH

(I) EtOH Heat

19. Show a synthesis of 2,3-dibromo-2,3-diiodohexane from hex-2-yne. 20. Give a complete mechanism for the following reaction: O H2O , Ether cat H+

21. Give the complete mechanism for the following transformation: cat H+ H2O

OH

22. Give the major product for each reaction. HgO , H2SO 4

(A)

aq Acetone

HO

OTES

O

(B)

1 Hg(OAc)2 , rt THF , H2O 2. NaBH4 , 0°C

OTES

H H

(C) HO

HO O

H H

1. Hg(OAc)2 2. NaBH4

C H A P T E R

3 Functional Group Exchange Reactions: Aliphatic and Aromatic Substitution and Elimination Reactions 3.1 INTRODUCTION It was stated in Section 1.1 that most of the actual chemical reactions in a synthesis are those that incorporate or change functional groups. Such reactions are known as functional group exchange reactions. This chapter will review two major reaction types that are involved in functional group exchanges: substitution and elimination reactions. There are many reagents that correlate with these reactions.1 • Substitution reactions RX + Aƒƒƒ!RA + X • Elimination reactions XdCdCdYƒƒƒ!C ]C + X dY Aliphatic substitution (SN1 or SN2) is a reaction where one functional group attached to an aliphatic (sp3 hybridized) carbon is replaced by another atom or group called a nucleophile (nucleophilic aliphatic substitution). These reactions are discussed in Section 3.2. Elimination involves conversion of a saturated moiety containing a leaving group to a molecule with a multiple bond. Reagents that are classified as bases are essential for so-called elimination reactions of alkyl halides (E2 reactions; Section 3.5.1). Note that another class of elimination reactions includes polar species (e.g., sulfoxides, amine oxides, and selenoxides), which undergo thermal syn elimination (Section 3.7). Another class of substitution reaction involves aromatic rather than aliphatic compounds. Both electrophilic aromatic substitution (SEAr) and nucleophilic aromatic substitution (SNAr) are categorized under this heading. This chapter will conclude with a review and a discussion of both types of these reactions, which are used to prepare aromatic derivatives.

3.2 ALIPHATIC SUBSTITUTION REACTIONS Replacement of one atom or functional group for another is referred to as substitution. The two major types of substitution usually involve reaction at an sp3 carbon (nucleophilic aliphatic substitution) or at an sp2 carbonyl carbon (nucleophilic acyl substitution). Substitution reactions at an sp3 carbon can be mechanistically categorized as bimolecular or unimolecular. Bimolecular substitution formally refers to the fact that the reaction follows second-order kinetics. As a practical matter, these reactions involve collision of a nucleophile (electron rich) with a carbon that is electron poor because it is attached to a polarizing leaving group. Unimolecular substitution is also derived from the rate expression because the reaction follows first-order kinetics, but requires ionization of a substrate bearing a leaving group to form a carbocation intermediate as the rate-determining step, and must occur before reaction with a nucleophile. The former reaction is labeled SN2 and the latter is labeled SN1. Both SN2 and SN1 reactions are extremely useful for the introduction of heteroatom and carbon functional groups into a carbon skeleton. 1

Many examples of these types of reactions are found in Larock, R. C. Comprehensive Organic Transformations, 2nd ed.; Wiley-VCH: New York, NY, 1999.

Organic Synthesis http://dx.doi.org/10.1016/B978-0-12-800720-4.00003-9

97

Copyright © 2017 Michael Smith. Published by Elsevier Inc. All rights reserved.

98

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

3.2.1 Aliphatic Bimolecular Nucleophilic Substitution When a nucleophile collides with an sp3 carbon that is part of a polarized bond bearing a leaving group (an electrophilic carbon), the generalized reaction shown is possible. In this transformation, Nuc is a nucleophile, C is an sp3 carbon, and X is a suitable leaving group. This reaction is called nucleophilic aliphatic substitution.2 +

Nuc:

C

Nuc

X

C

+

X:–

3.2.1.1 The SN2 Reaction In reactions where the substrate (see 1) has a leaving group (X) connected to an sp3 carbon, substitution by a nucleophile leads to a product in which X has been replaced by the nucleophile to yield 3. Substitution is defined here as a chemical reaction in which one atom or group in a chemical compound is replaced by another atom or group. This process can be described as a bimolecular (it follows a second-order rate equation) substitution that involves a nucleophile, and the mechanistic descriptor is SN2. Reactions labeled as SN2 may be observed in several different functional group exchange reactions and the substrate and the reagent can be either simple or complex. All SN2 reactions can be represented by the simple reaction sequence: 1 ! 3, but it has been determined experimentally that there is no intermediate.3 Such a reaction is synchronous, proceeds with no intermediate, so the transition state must be examined to understand the reaction. When the reaction occurs at an sp3 carbon, this substitution reaction requires a pentacoordinate transition state (2).4b H Nuc

H

H

+

+

X

R2

Nuc

X

R1

R2 R1

1

2

Nuc

R2 R1 3

The determination that the course of the reaction can be described by 2 is based on an analysis of many reactions that convert 1 to 3, with many different nucleophiles. All show a lack of evidence for an intermediate. This transition state structure is taken to be the logical midpoint of the reaction, where the nucleophile-carbon bond is partially formed and the CdX bond is partially broken. The reaction is initiated by collision of the nucleophile with the δ+ carbon, during which two electrons are transferred to that carbon in order to form a new covalent bond. At the same time, the CdX bond is broken and both of those electrons are transferred to the more electronegative heteroatom-leaving group, completing the substitution process. When the nucleophile reacts with 1, it approaches the carbon atom bearing the leaving group 180 degrees away from the leaving group, which is referred to as backside attack. Approach of the nucleophile from the side of the molecule bearing the group X is unfavorable due to electrostatic and steric repulsion. Approach from the side opposite X (backside or antiattack) will minimize these repulsive forces and is energetically preferred. Transition state 2 is consistent with the following observations: 1. Backside attack dictates that inversion of configuration must occur in the final product, relative to the starting material. Inversion can be detected when the starting halide possesses a stereogenic carbon bearing a leaving group, as in the conversion of 1 ! 3 (R1, R2 ¼ H, alkyl, aryl). 2. Since the reaction requires a collision of the nucleophile and the substrate,4a the rate of reaction is proportional to the concentrations of both reactants and the nature of the solvent. An SN2 reaction is termed bimolecular because it is described by the rate expression: (Rate ¼ k [N] [substrate]), where N is the nucleophile. The rate of reaction is also dependent on the strength of the nucleophile.4 Protic solvents slow the reaction, making aprotic solvents the preferred reaction medium.

2

For examples, see Ref. 1, pp 611–618 (halogenation of hydrocarbons), 667–676 (interconversion of halides), 779–784 (amines from alkyl halides), 889–910 (ethers by alkylation of alcohols), 970–972 (alcohol by halide substitution), and pp 1717–1808 (alkylation and substitution of nitriles and acid derivatives).

3

Lowry, T. H.; Richardson, K. S. Mechanism and Theory in Organic Chemistry, 3rd ed.; Harper & Row: New York, NY, 1987; pp 297–298.

(a) Ref. 3, pp 177–181. (b) Cowdrey, W. A.; Hughes, E. D.; Ingold, C. K.; Masterman, S.; Scott, A. D. J. Chem. Soc. 1937, 1252. For a discussion of solvation energy, see (c) van Bochove, M. A.; Bickelhaupt, F. M. Eur. J. Org. Chem. 2008, 649. 4

.

99

3.2 ALIPHATIC SUBSTITUTION REACTIONS

3. The rate of reaction is fast with MeBr and slower with secondary alkyl halides. Tertiary halides (e.g., tert-butyl bromide, Me3CdBr) do not undergo substitution via this mechanism at any reasonable rate, so for all practical purposes there is no reaction. The structure of the alkyl halide is critically important for an SN2 reaction, as in the generic reaction of RdBr with KI.5,6 If the reaction with bromomethane is taken as the standard, bromoethane reacts 82 times slower, 2-bromopropane 3300 times slower, and 2-bromo-2-methylpropane 18,180 times slower. It is clear that as the steric bulk around the carbon being attacked by the nucleophile increases, the rate of reaction slows. This observation correlates directly with an increase in activation energy required to achieve the pentacoordinate transition state, 2. With tertiary halides, steric hindrance in the transition state makes the activation energy so high that the rate of an SN2 reaction is so slow that for all practical purposes there is no reaction. The neopentyl system, which appears to be primary, is in fact so sterically hindered (2, R1 ¼ H, R2 ¼ CMe3) that the rate of an SN2 reaction is nearly 17 times slower than that of a tertiary halide and 3 million times slower than bromomethane. Methyl halides easily attain the transition state (2, R1 ¼ R2 ¼ H) and an SN2 reaction is very facile. Interestingly, the π-bond participates in expulsion of the leaving group in 2 for allyl and benzyl halides, and both halides react faster than bromomethane (23 times faster for allyl bromide and 4 times faster for benzyl bromide).5 RdBr + KdIƒƒƒ!RdI + KdBr The SN2 reaction is a powerful method for incorporating functional groups into a carbon skeleton, and because of a clean backside attack, the stereochemistry at one carbon atom can be controlled and predicted. The SN2 transition state required for the reaction imposes several important synthetic restrictions. The poor reactivity of sterically hindered halides dictates that primary or secondary substrates be used, but inversion of configuration makes those reactions suitable for manipulating stereochemistry. However, stereochemical control via inversion of configuration is only important if the stereochemistry of the initial substrate can be controlled. One approach that deals with these limitations exploits the facile reactivity of primary substrates, but the primary reaction center is part of a chiral molecule with a preexisting stereogenic center. An example is the SN2 reaction of the cyanide ion with the primary tosylate of 4 to yield nitrile 5 (where NAP ¼ napthylmethyl) in 80% overall yield, taken from Gauthier and coworker’s7 synthesis of fragments of Burkholderia pseudomallei and Burkholderia mallei capsular polysaccharide. Note the use of the crown ether 18-crown-6 to facilitate the substitution reaction by sequestering the potassium ion. An SN2 reaction could be used to introduce nitrogen into a molecule using an amine nucleophile. However, there are several problems associated with this approach, including elimination, salt formation, and reversibility. The conversion of 4 to 5 used cyanide, which introduces nitrogen into the molecule without the complications associated with amine nucleophiles. A second chemical step (a reduction) is required to unmask the amine moiety, however. The nucleophilic azide anion is another alternative. An example is the conversion of the hydroxyl unit in 6 to azide 7 (BOC ¼ t-butoxycarbonyl) in 74% yield via initial conversion to the methanesulfonate ester. The SN2 reaction proceeded with clean inversion of configuration in Ichikawa’s and coworker’s8 synthesis of quinaldopeptin. Note that the polar aprotic solvent DMF was used to maximize the yield of SN2 products and suppress side reactions (e.g., elimination, Section 3.5.1). Alkyl azides can be explosive, however, so care should be exercised when an azide is used. OTs ONAP O

BnO TBSO

KCN , MeCN 18-crown-6

SPh

CN ONAP O

BnO TBSO

Bn = Benzyl TBS = tert-Butyldimethylsilyl

4

SPh 5 (88%) NHBoc

NHBoc CO2t-Bu 1. MsCl , NEt3 , CH2Cl2

CO2t-Bu

2. NaN3 , DMF , 70°C

N3

OH

7 (79%)

6

(a) Streitwieser, A., Jr. Chem. Rev. 1956, 56, 571. (b) Streitwieser, A., Jr. Solvolytic Displacement Reactions; McGraw-Hill: New York, NY, 1962; p 13. (c) Ref. 6, p 343.

5

6

Hine, J. Physical Organic Chemistry, 2nd ed.; McGraw-Hill: New York, NY, 1962; p 203.

7

Kenfack, M. T.; Bl
eriot, Y.; Gauthier, C. J. Org. Chem. 2014, 79, 4615.

8

Katayama, K.; Okamura, T.; Sunadome, T.; Nakagawa, K.; Takeda, N.; Shiro, M.; Matsuda, A.; Ichikawa, S. J. Org. Chem. 2014, 79, 2580. .

100

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

Typical SN2 reactions involve functional group transforms (e.g., Br ! OMe, OTs ! N3 ! NH2, and Cl ! I), but other types of nucleophilic reactions are possible, including reactions of enolate anions with alkyl halides (Section 13.3.1), which proceed via an SN2 pathway.9 A classical reaction that involves an SN2 reaction is the Williamson ether synthesis,10 illustrated by the conversion of the OH units in 8 to bis(ether) 9 in 90% yield, via SN2 displacement of the alkoxide units with iodomethane, in Banwell and coworker’s11 synthesis of ribisin C. The acidic OH units reacted with the base, sodium hydride, to yield the alkoxide moieties in situ, which subsequently reacted with iodomethane. In this reaction, an alkoxide is the nucleophile. OH Br

OMe OH

NaH , THF , MeI

Br

OMe

0–18°C

O

O

O

8

O

9 (90%)

Intramolecular Williamson ether syntheses are possible. In a synthesis of galanthamine, Tu and coworkers12 showed that treatment of 10 with the base (DBU) converted the dOdSiMe2t-Bu group to the alkoxide (RO), which displaced the bromide to yield the benzofuran unit in 11, in 90% yield. The use of microwave irradiation in conjunction with phase-transfer catalysis is also quite effective in these reactions.13 CHO

CHO

O

O DBU , DMSO

O

Br t-BuMe2SiO

O OMe

O

10

OMe

11 (90%)

The success or failure of an SN2 reaction depends on the strength of the nucleophile, as well as on the nature of the substrate it reacts with. Nucleophilic strength is measured as a function of both the substrate (s) and the strength of the nucleophile. The parameter known as nucleophilicity (n) is a measure of nucleophilic strength, and is given by the SwainScott equation14a [log (k/k0) ¼ (s) (n)], where n ¼ 0 for water at 25°C, s ¼ 2.00 for bromomethane,14a and n is a reagent parameter measuring nucleophilic power in a particular system. The parameter s is the susceptibility of the substrate (here a halide) to being attacked (it measures the power of a substrate to discriminate between various nucleophiles). An electrophilicity index has been established based on electronegativity divided by chemical hardness.14b,c The important qualitative message of this equation is that nucleophilic strength varies with the substrate. Table 3.1 gives a partial list of nucleophiles,15 rank ordered by use of the Swain-Scott equation14 for their reaction with bromomethane. This list constitutes a general order of nucleophilic strength in SN2 reactions. It is known that the nucleophilicity order differs for SN2 reactions when compared with acyl addition to a carbonyl (see Section 4.2). Several parameters contribute to nucleophilic strength, but a few general statements can be made. (1) A nucleophile with a negative charge is stronger than its conjugate acid. NH2  > NH3 and HO > HOH. (2) In the same row, nucleophilicity parallels basicity, and R3 C > R2 N > RO > F . Rules 1 and 2 predict the following order for nucleophilic strength: NH2  > RO > HO > R2 NH > ArO > N3 > C5 H5 N > F > HOH > ClO4 

9

For a synthetic example, see Marshall, J. A.; Ellison, R. H. J. Am. Chem. Soc. 1976, 98, 4312.

(a) Williamson, A. W. J. Chem. Soc. 1851, 4, 229. (b) Dermer, O. C. Chem. Rev. 1934, 14, 385 (see p 409). (c) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-101.

10

11

Lan, P.; Banwell, M. G.; Ward, J. S.; Willis, A. C. Org. Lett. 2014, 16, 228.

12

Hu, X.-D.; Tu, Y. Q.; Zhang, E.; Gao, S.; Wang, S.; Wang, A.; Fan, C.-A.; Wang, M. Org. Lett. 2006, 8, 1823.

For a review of microwave activation in phase-transfer catalysis, see Deshayes, S.; Liagre, M.; Loupy, A.; Luche, J.-L.; Petit, A. Tetrahedron 1999, 55, 10851. 13

(a) Swain, C. G.; Scott, C. B. J. Am. Chem. Soc. 1953, 75, 141. (b) Parr, R. G.; Szentpály, L. V.; Liu, S. J. Am. Chem. Soc. 1999, 121, 1922. (c) Maynard, A. T.; Huang, M.; Rice, W. G.; Covel, D. G. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 11578.

14

15

Edwards, J. O.; Pearson, R. G. J. Am. Chem. Soc. 1961, 84, 16. .

101

3.2 ALIPHATIC SUBSTITUTION REACTIONS

TABLE 3.1 Nucleophilic Strength With Bromoethane as Ordered by the Swain-Scott Equation Nucleophile 2–

S2O3

–

PO4

2–

CO3 Br–

n

NO2

p-NO2

PhO–

–

HCO2 F–

–

2–

n

Nucleophile

n

6.35

SO3

5.67

–CN

5.13

5.06

I–

4.93

–SCN

4.80

4.36

HO–

4.23

–OPh

4.16

3.92

–OPhp-Cl

3.86

4.02 –

Nucleophile

–

N3

–

3.71

HPO3

3.08

Cl– –

2.62

HSO3

1.88

2–

HSO4

2.1

HOH

0.01

2–

3.52

B2O7

2.99

AcO–

ClO3

2.2

–

2.1

HNO3

2.1

–

2.76 –

H2PO4

2.3

SO4

3.45

–

ClO4

2.1

2.1

Reprinted with permission from Edwards, J. O.; Pearson, R. G. J. Am. Chem. Soc. 1961, 84, 16. Copyright © 1961 American Chemical Society.

Nucleophilic reactions are usually under kinetic control, whereas acid-base reactions are under thermodynamic control. Under kinetic control, the most nucleophilic species will react faster in substitution reactions. Nucleophilic strength for a given substituent can be measured in terms of the rate of the SN2 reaction.16 The relative rates of the nucleophiles shown in Table 3.2 were determined by reaction with iodomethane in an SN2 reaction.17 Electronic effects are important as illustrated by methoxide, where the presence of the electron-releasing methyl group on oxygen makes methoxide more nucleophilic relative to hydroxide. The rate of the SN2 reaction of sodium hydroxide with iodomethane is 1.3  104 M1 s1, whereas the rate with sodium methoxide with iodomethane is 2.51  102 M1 s1.17 TABLE 3.2 Relative Rates of Reaction of Nucleophiles With Iodomethane k (¥ 10–3 M–1 s–1)

Nucleophile MeO–

0.251

F–

0.00005

NHEt2

2.2

PhO–

0.073

NH3

0.041

Py

0.022

Reprinted with permission from Pearson, R. G.; Sobel, H.; Songstad, J. J. Am. Chem. Soc. 1968, 90, 319. Copyright © 1968 American Chemical Society.

A species is less nucleophilic if there is resonance delocalization of the electron density. An example is the phenoxide anion, where the electrons on oxygen are delocalized into the adjacent phenyl ring, away from oxygen. Consequently, oxygen cannot donate electrons as effectively. In other words, phenoxide is less nucleophilic than a typical alkoxide that is incapable of resonance (e.g., methoxide). (3) Going down the periodic table, nucleophilicity increases, but basicity decreases: 
 I > Br > Cl > F this ordering is solvent dependent , and ðRS > RO Þ :

(a) Hudson, R. F.; Green, M. J. Chem. Soc. 1962, 1055. (b) Bender, M. L.; Glasson, W. A. J. Am. Chem. Soc. 1959, 81, 1590. (c) Jencks, W. P.; Gilchrist, M. Ibid. 1965, 90, 2622.

16

17

Pearson, R. G.; Sobel, H.; Songstad, J. J. Am. Chem. Soc. 1968, 90, 319. .

102

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

Comparing the rate of reaction of MeI with PhSe (7000  103 M1 s1), PhS (1070  103 M1 s1), and PhO (0.073  103 M1 s1) shows the trend mentioned in (3).17 Likewise, trialkylphosphines are more nucleophilic than trialkylamines (R3P > R3N). The rate of reaction of Et3P with iodomethane is 1.29  103 M1 s1, whereas the rate for Et3N is 0.595  103 M1 s1.17 These latter reactions are particularly useful in organic synthesis. Ph3 P Triphenylphosphine

+ CH3 I ƒƒƒƒƒƒ! Ph3 P +dCH3 I Methyltriphenyl phosphonium iodide

Just as amines react as nucleophiles, so do phosphines, R3P. An example is the reaction of iodomethane with triphenylphosphine, which yields methyltriphenylphosphonium iodide via an SN2 reaction. The reaction of trialkylphosphines with alkyl halides is particularly useful since the resultant phosphonium salts are easily converted to a phosphonium ylid on treatment with a suitable base (Section 12.5.1). Ylids are the reactive species in the well-known Wittig olefination reaction,18 which will be discussed in Section 12.5.1.1. A related SN2 process involves reaction of a trialkylphosphite with an alkyl halide, the Arbuzov reaction (sometimes called the Michaelis-Arbuzov reaction).19 Triethylphosphite reacts with iodomethane to yield the phosphonium salt, triethoxy(methyl)phosphonium iodide. Heating generates the monoalkyl phosphonate ester (diethyl methylphosphonate) along with iodoethane. This type of phosphonate ester can be converted to an ylid and used in the Horner-Wadsworth-Emmons olefination20 (Section 12.5.1.3). EtO

EtO

MeI

P

EtO P EtO

EtO EtO

Triethyl phosphite

I Me

O

Heat

EtO P EtO

Triethoxy(methyl)phosphonium iodide

Me

+

EtI

Diethyl methylphosphonate

Substitution by amines or phosphines with an alkyl halide is clearly an SN2 type process. Normally, the presence of water in the reaction medium suppresses the rate of SN2 reactions because they usually involve the reaction of a charged nucleophile and a neutral substrate. An exception can occur, however, if both starting materials are neutral, but the product is charged, as in the reaction of amines with alkyl halides. This reaction generates transition state 12 with a δ+ nitrogen and a δ halogen. A protic solvent will separate the developing charge in this transition state, ultimately leading to the solvent separated ions +NRH2R0 and X. Solvation and ion separation accelerates the SN2 process in the preparation of ammonium salts, and the use of water as the solvent or as a cosolvent is usually the best choice for such reactions. R1 RNH2

+

R2 R3

R1

H X

R

H

+

N H

X R2 R3

R

R1

N H

R2 R3

12

The SN2 reaction with amines and alkyl halides forms ammonium salts. As mentioned for the reactions of 4 and 6, there are problems associated with the reaction of an amine with an alkyl halide. Nonetheless, the reaction can be used. Elimination reactions are possible when the ammonium salt has a hydrogen atom on the carbon β to the nitrogen atom. This elimination can be exploited to synthesize alkenes in the Hoffmann elimination sequence, which will be discussed in Section 3.7.1.21 Triphenylamine reacts with iodomethane to yield methyltriphenylammonium iodide (MePh3N+ I) and alkyl ammonium salts such as this can be used as phase-transfer catalysts.22 Phase-transfer ammonium salts often (a) Wittig, G.; Sch€ ollkopf, U. Berichte 1954, 87, 1318. (b) Wittig, G.; Haag, W. Berichte 1955, 88, 1654. (c) Trippett, S. Q. Rev. Chem. Soc. (London) 1963, 17, 406. (d) Wittig, G. Acc. Chem. Res. 1974, 7, 6.

18

(a) Michaelis, A.; Kaehne, R. Berichte 1898, 31, 1048. (b) Arbuzow, A. E. J. Russ. Phys. Chem. Soc. 1906, 38, 687. (c) Idem Chem. Zentr. 1906, II, 1639. (d) Arbuzow, B. A. Pure Appl. Chem. 1964, 9, 307. (e) Kosolapoff, G. M. Org. React. 1951, 6, 273 (see pp 276–277).

19

20

(a) Wadsworth, W. S.; Emmons, W. D. J. Am. Chem. Soc. 1961, 83, 1733; (b) Boutagy, R.; Thomas, R., Jr. Chem. Rev. 1974, 74, 87.

(a) Hofmann, A. W. Annalen 1851, 78, 253. (b) Hofmann, A. W. Annalen 1851, 79, 11. (c) Brown, H. C.; Moritani, I. J. Am. Chem. Soc. 1956, 78, 2203. (d) Cope, A. C.; Trumbull, E. R. Org. Reactions, 1960, 11, 317. (e) Hofmann, A. W. Berichte 1881, 14, 659. (f) Wall, E. N.; McKenna, J. J. Chem. Soc. B, 1970, 318. (g) Wall, E. N.; McKenna, J. J. Chem. Soc. C 1970, 188. (h) Bach, R. D.; Bair, K. W.; Andrezejewski, D. J. Am. Chem. Soc. 1972, 94, 8608. (i) Archer, D. A.; Booth, H. J. Chem. Soc. 1963, 322.

21

(a) Jones, R. A. Aldrichimica Acta 1976, 9, 35. (b) Starks, C. M. J. Am. Chem. Soc. 1971, 93, 195. (c) Starks, C. M.; Owens, R. M. J. Am. Chem. Soc. 1973, 95, 3613. (d) Herriott, A. W.; Picker, D. J. Am. Chem. Soc. 1975, 97, 2345.

22

.

103

3.2 ALIPHATIC SUBSTITUTION REACTIONS

contain large alkyl groups (dodecyl, hexadecyl, and octadecyl), and small groups are also incorporated in many salts. Tetrabutylammonium chloride is a useful phase-transfer reagent, for example. The alkylammonium salts are readily prepared by the reaction of an amine (e.g., tridecylamine) with chloromethane.23 ðC10 H21 Þ3 N + CH3 Cl ƒƒƒƒ! ðC10 H21 Þ3 N +dCH3 Cl Tridecylamine TridecylðmethylÞammonium chloride Despite reactivity problems, amine alkylation of this type finds its way into synthesis. An interesting example is taken from Loh and Lee’s synthesis of ()-epibatidine,24 in which the intramolecular displacement of one bromide in 13 gave an 85% yield of 14. Polyalkylation is sometimes a problem when amines or ammonia react with alkyl halides, and amine surrogates are often used, including phthalimide. Cl

NH2

H N

Cl

N

Br

N

MeCN, 82°C

HN

Br

N Br

Br 13

Cl

14 (85%) O

Br

NH2

O

O

O

N

NH

NH2NH2

3 Steps

O

Cs2CO3 , DMF

I 1-(Bromomethyl)-4-iodobenzene

(CH 2)3CH2OH

(CH2)3CH2OH 15

NH

+

BuOH Reflux

16

NH O 2,3-Dihydrophthalazine-1,4-dione (phthalhydrazide)

In a synthesis of a mitochondrial complex 1 inhibitor, Radeke et al.25 prepared phthalimide [15, [2-(4-(4-hydroxybutyl)benzyl)isoindoline-1,3-dione]] from the reaction of the phthalimide anion with a benzylic bromide [1-(bromomethyl)-4-iodobenzene]. Initial coupling with phthalimide gave the phthalimide derivative in 86% yield by an SN2 reaction, followed by three steps to convert the aryl iodide to the hydroxyl-butyl unit in 2-(4-(4-hydroxybutyl)benzyl)isoindoline-1,3-dione. Final treatment with hydrazine in butanol heated at reflux liberated the amino group in 16 [4-(4-(aminomethyl)phenyl)butan-1-ol] in 84% yield, along with phthalhydrazide (2,3-dihydrophthalazine-1,4-dione) as a byproduct. (4) Greater steric hindrance diminishes nucleophilicity. Comparing SN2 reactions of alkyl halides with pyridine and 2,6-dimethylpyridine (2,6-lutidine) illustrates this point. The rate of reaction of pyridine with methyl iodide is 0.022  103 M1 s1. The two methyl groups proximal to the nitrogen atom in 2,6-lutidine sterically hinder approach of nitrogen to the carbon of iodomethane, and the rate of reaction is 50 times slower, 0.00042  103 M1 s1.17 As pyridine approaches the electrophilic carbon, the ortho hydrogen atoms interact with the alkyl groups on that carbon, but this steric hindrance is minor when compared to the ortho methyl groups in 2,6-lutidine. Steric hindrance makes lutidine less nucleophilic than pyridine. 3.2.1.2 The Mitsunobu Reaction A key feature of a SN2 reaction is inversion of configuration at the electrophilic center. Therefore, the SN2 reaction is an effective method for inversion of the stereocenter with incorporation of a different functional group. Alcohols are convenient synthetic intermediates, and an SN2 reaction of these useful substrates is most attractive. In some cases, reactions that yield alcohols with good enantioselectivity (e.g., Section 7.9) lead to the incorrect stereochemistry for

23

See (a) Starks, C. M.; Owens, R. M. J. Am. Chem. Soc. 1973, 95, 3613. (b) Starks, C. M. J. Am. Chem. Soc. 1971, 93, 195.

24

Lee, C.-L. K.; Loh, T.-P. Org. Lett. 2005, 7, 2965.

25

Radeke, H.; Hanson, K.; Yalamanchili, P.; Hayes, M.; Zhang, Z.-Q.; Azure, M.; Yu, M.; Guaraldi, M.; Kagan, M.; Robinson, S.; Casebier, D. J. Med. Chem. 2007, 50, 4304. .

104

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

the target. Changing that sterochemistry can be important if there are few synthetic alternatives. However, the OH moiety is a poor leaving group (Section 3.6.4). If one sets a goal of inverting the stereochemistry of pentan-(2S)-ol to yield pentan-(2R)-ol, the fact that hydroxyl is a very poor leaving group is a problem. Specifically, the hydroxide ion, or any other nucleophile, will not displace OH directly. Conversion of the OH unit into an ester, sulfonate ester, or phosphonate ester derivative often allows an SN2 type displacement to proceed. If the OH unit is converted to a halide or sulfonate ester, however, treatment with hydroxide can lead to elimination as a major competitive reaction (Section 3.5.1). The use of a nucleophile such as acetate would minimize elimination, but acetate is a poor nucleophile in substitution reactions. As with acetate, there are times when a nucleophile is too weak for facile SN2 displacement, and substitution is also difficult if the substrate contains a poor leaving group. A successful and highly useful solution to this problem is the Mitsunobu reaction.26 This procedure is typically used to invert the stereochemistry of an alcohol by converting a poor leaving group to an excellent leaving group for subsequent reaction with a nucleophile. A synthetic example shows the conversion of (R)-alcohol 17 to (S)-alcohol 18 in 72% overall yield using diethyl azodicarboxylate (DEAD), p-nitrobenzoic acid, and triphenylphosphine, followed by hydrolysis of the resulting ester with potassium carbonate in methanol in a synthesis of the major component of mosquito oviposition attractant pheromones by Wang and coworkers.27 OH O

C10 H21

O

1. DEAD , p-NO2PhCOOH PPh3 , THF , rt , 4 h 2. K2CO3, MeOH rt , 10 min

OH

(83%)

O

C10 H21

(87%)

O

18 (72%)

17

As initially formulated, the process involved reaction of DEAD with triphenylphosphine (Ph3P) to form 19. Product 19 is a dipolar ion, and it reacts with HX, which is present in the initial reaction, to yield phosphonium salt 20. In the presence of an alcohol, there is a reaction that generates alkoxyphosphonium salt 21 and a diimide, diethyl hydrazine-1,2-dicarboxylate. Displacement of triphenylphosphine oxide (an excellent leaving group) by X (the nucleophile) yields the substitution product, RdX. The X moiety in HX can be a variety of groups, including carboxylate, azide, imido, and R (an active methylene compound, e.g., malonate or an α-cyanoester).26

EtO2C

N

N

CO2Et

PPh3

EtO2C

DEAD ROH

H

N N CO Et 2 PPh3

HX

EtO2C

19 EtO2C

NH

NH CO2Et +

X

Ph3P

Diethyl hydrazine1,2-dicarboxylate

OR

R–X

N N CO Et 2 PPh3 X 20

+

Ph3P=O

21

There are many variations. In one, taken from De Brabander and coworker’s28 synthesis of gustastatin, selective etherification of one phenolic OH unit in 22 gave benzyl ether (23) in 73% yield. Note the use of DIAD (diisopropyl azodicarboxylate) in this reaction. Lepore and He29 showed that ether formation via the Mitsunobu reaction is facilitated with sonication. Goswami and coworkers30 reported an example of inversion of configuration for an alcohol in an enantioselective synthesis of cytospolide P, where the hydroxyl unit in (3S,4S)-3-(benzyloxy)dec-1-en-4-ol was converted to the p-nitrobenzoate ester with clean inversion of configuration, and then hydrolyzed to (3S,4R)-3-(benzyloxy) dec-1-en-4-ol in 78% yield.

26

(a) Mitsunobu, O. Synthesis 1981, 1. (b) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-62.

27

Dong, H.-B.; Yang, M.-Y.; Zhang, X.-T.; Wang, M.-A. Tetrahedron: Asymmetry, 2014, 25, 610.

28

Garcia-Fortanet, J.; Debergh, J. R.; De Brabander, J. K. Org. Lett. 2005, 7, 685.

29

Lepore, S. D.; He, Y. J. Org. Chem. 2003, 68, 8261.

30

Chatterjee, S.; Guchhait, S.; Goswami, R. K. J. Org. Chem. 2014, 79, 7689. .

105

3.2 ALIPHATIC SUBSTITUTION REACTIONS

Recent work to improve this reaction has focused on several aspects of the reaction,31 including the use of alternative reagents, phase-switching modifications of phosphine, azodicarboxylate, and nucleophilic components are emphasized and also the separation of products from byproducts32 using fluorous compounds. HO

Ph

O O OH

BnO

Ph

O

BnOH , PPh3

Ph

O

DIAD , THF

OH

O 22

Ph

O

23

OBn

OBn

1. p-Nitrobenzoic acid , DIAD PPh3 , Toluene

C5H11

C5H11

0°C to rt , 2.5 h 2. K2CO3 , MeOH , 30 min 0°C – rt

OH (3S,4S)-3-(Benzyloxy)dec-1-en-4-ol

OH (3S,4 R)-3-(Benzyloxy)dec-1-en-4-ol (78%)

Hydroxyl can be replaced with other functional groups using the Mitsunobu reaction. The reaction of succinimide with propan-2-ol under Mitsunobu conditions gave N-isopropyl succinimide in 73% yield.33 Mitsunobu coupling can be used to insert functional groups that might be difficult to prepare under other conditions. As just discussed, reaction of an alcohol and an acid34 generates an ester. Azides are also conveniently prepared, as illustrated in Jia and Zhoui’s35 synthesis of ()-goniomitine, in which alcohol 24 was converted to azide 25 in 81% yield using Mitsunobu conditions. Note that the reagent diphenylphosphoryl azide (DPPA) was used, since it is often more efficient than reaction with KN3 or NaN3.36 Alcohols react with phthalimide under Mitsunobu conditions to yield the N-substituted phthalimide, which can be unmasked to the primary amine by reaction with hydrazine, as noted above.37 This latter reaction of phthalimide with alkyl halides followed by treatment with hydrazine is a variation of a classical reaction called the Gabriel synthesis.38 The use of hydrazine in this manner is called the Ing-Manske modification39 of the Gabriel synthesis. The conversion of 15 to 16 in Section 3.2.1.1 is an example of this modification. The use of Mitsunobu conditions simply expands the reaction to include alcohols as starting materials. The same transformation has been accomplished using TsNHCO2t-Bu to incorporate a nitrogen moiety.40 OAc

OAc N

PPh3 , DEAD THF

N

O

O

PhO P PhO

O

N3

N3

HO 24

25 (81%)

31

Dembinski, R. Eur. J. Org. Chem. 2004, 2763.

32

Dandapani, S.; Curran, D. P. Chem. Eur. J. 2004, 10, 3131.

33

Ref. 21, p 5.

34

For an example, see Keck, G. E.; McHardy, S. F.; Murry, J. A. J. Org. Chem. 1999, 64, 4465.

35

Zhou, S.; Jia, Y. Org. Lett. 2014, 16, 3416.

36

See Lal, B.; Pramanik, B. N.; Manhas, M. S.; Bose, A. K. Tetrahedron Lett. 1977, 18, 1977.

37

For examples, see (a) Brosius, A. D.; Overman, L. E.; Schwink, L. J. Am. Chem. Soc. 1999, 121, 700. (b) Mulzer, J.; Scharp, M. Synthesis 1993, 615.

38 (a) Gabriel, S. Berichte 1887, 20, 2224. (b) Gibson, M. S.; Bradshaw, R. W. Angew. Chem. Int. Ed. Engl. 1968, 7, 919. (c) Mundy, B. P.; Ellerd, M. G.; Fabaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; John Wiley & Sons, Inc.: New York, NY, 2005; p 264. (d) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-36. 39

Ing, H. R.; Manske, R. H. F. J. Chem. Soc. 1926, 2348.

40

Trost, B. M.; Oslob, J. D. J. Am. Chem. Soc. 1999, 121, 3057. .

106

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

Another modification of the Mitsunobu reaction illustrates two features that can be useful in synthesis. In a synthesis of ()-stenine by Fujioka et al.,41 alcohol 26 was treated with triphenylphosphine and DIAD, and cyclization occurred via an internal substitution reaction to yield the bicyclic compound, (27), in 55% yield. This example shows that Mitsunobu conditions can be used for intramolecular reactions, and also that a protected amine (e.g., a sulfonamide) can participate in the substitution reaction. NHNs NNs

DIAD , 2 PPh3 , rt 15 min 1,4-Dioxane

OH

O O

26

27 (55%)

The Mitsunobu reaction is a clever and highly useful exploitation of the SN2 reaction that allows one to control the stereochemistry of a given stereogenic center (see Chapter 10). It also allows one to correct the stereochemistry if an alcohol is produced that has the opposite stereochemistry from that desired (see Section 10.2.2).

3.2.1.3 The SN2’ Reaction Normal SN2 reactions were illustrated in the previous sections. In general, allylic halides react faster with nucleophiles than do simple alkyl halides (see Section 3.2.1) because the π-bond can participate in expulsion of the leaving group in the pentacoordinate transition state (see 2). If the carbon atom bearing the leaving group becomes too sterically hindered in an allylic system, however, the normal SN2 path is inhibited, and an alternative mode of attack occurs at the end of the π-system. Such attack is termed an SN20 reaction42 (nucleophilic bimolecular substitution with allylic rearrangement). An SN20 reaction is possible because the CdX unit polarizes adjacent bonds, leading to extended bond polarization: δ + C]Cδdδ + CdXδ , allowing the nucleophile to attack either the carbonyl carbon or the terminal carbon of the C]C unit. This extension of reactivity due to conjugating π-bonds is called vinylogy. The SN2 and SN20 pathways often compete, which can lead to annoying mixtures of regioisomers. Nonetheless, the SN20 reaction is synthetically useful in some cases. Formation of piperidine derivative 1-(4-isopropylcyclohex-2-en-1-yl)piperidine, via the SN20 displacement of the allylic 3,5-dichlorobenzoate (6-isopropylcyclohex-2-en-1-yl 3,5-dichlorobenzoate), is an excellent example of this process.43 The nucleophile attacks the π-bond since the O-allylic carbon is sterically encumbered. The π-bond assists the removal of the leaving group in an SN20 reaction.

••

N

H Cl

H

S N2

O H

N

H

130°C

O

H

Cl

6-Isopropylcyclohex-2-en-1-yl 3,5-dichlorobenzoate

1-(4-Isopropylcyclohex-2-en-1-yl)piperidine

41

Fujioka, H.; Nakahara, K.; Kotoku, N.; Ohba, Y.; Nagatomi, Y.; Wang, T.-L.; Sawama, Y.; Murai, K.; Hirano, K.; Oki, T.; Wakamatsu, S.; Kita, Y. Chem. Eur. J. 2012, 18, 13861.

42

Magid, R. M. Tetrahedron 1980, 36, 1901.

(a) Stork, G.; White, W. N. J. Am. Chem. Soc. 1953, 75, 4119. (b) Stork, G.; White, W. N. Ibid. 1956, 78, 4609. (c) Stork, G.; Kreft, A. F., III. Ibid. 1977, 99, 3850, 8373.

43

.

107

3.2 ALIPHATIC SUBSTITUTION REACTIONS

An SN20 reaction typically proceeds with a high degree of stereochemical control, as in the nucleophilic displacement of 6-isopropylcyclohex-2-en-1-yl 3,5-dichlorobenzoate by piperidine to yield 1-(4-isopropylcyclohex-2-en-1-yl)piperidine.44 Bordwell and coworkers45 described the SN20 reaction with amines as second order, but invoked a tight ion pair to explain the observed transformation. A tight ion pair will have significant ionic character, but the ions are not solvent separated. In work by Smith and coworkers,46 the reaction of piperidine with the tert-butyl derivative ((1-bromo-2,2dimethylpropyl)cyclopropane) gave the expected (E)-1-(5,5-dimethylhex-3-en-1-yl)piperidine in 88% yield, but also 8% of 1-(1-cyclopropyl-2,2-dimethylpropyl)piperidine. Formation of this latter product is not compatible with a simple SN2 displacement due to the great steric hindrance imposed by the neopentyl-like structure, but the tight ion-pair mechanism explains the presence of an SN2 product very well. This latter study with cyclopropylcarbinyl halides supports this mechanistic rationale.46 When it can be applied as the major process, stereochemical and regiochemical control makes the SN20 reaction a useful tool in synthetic planning. With other allylic substrates, the possibility of this reaction as a competitive side reaction must be acknowledged as a process that can diminish the yield of the SN2 product. Br

N

H

(1-Bromo-2,2-dimethylpropyl)cyclopropane

1-(5,5-Dimethylhex-3-en-1-yl)piperidine Me

Me

HO

1-(1-Cyclopropyl-2,2-dimethylpropyl)piperidine Me

Me

Me PBr3

Me

N

+

N

130°C

Me

Me PhS– Na+

Me

Me Me

Br

Me Me Me

Me 28

SPh 30

29

A synthetic example of an SN20 reaction is the conversion of 28 to tertiary bromide 29 with phosphorus tribromide. o and coworker’s47 Subsequent reaction with the sodium salt of thiophenol gave an SN20 displacement to yield 30 in It^ synthesis of neocembrene. Organocuprates (Section 12.3.1) react via an SN20 type process in some cases.48 As will be discussed in Section 12.3.1.2, the organocuprate reaction may actually proceed by copper-stabilized radical intermediates, rather than a formal SN20 mechanism.49 In a synthesis of ()-exiguolide, Roulland and coworkers50 reacted phosphonate ester 31 with lithium dimethylcuprate (see Section 12.3.1.1), generated in situ, to yield 32 in 96% yield. Note that the substitution reaction proceeded with good diastereoselectivity for 32. Magnesium cuprates also undergo SN20 reactions.51 An SN20 like reaction has also been observed in reactions of organocuprates with cyclopropylcarbinyl halides, where the strained cyclopropane ring behaves similarly to the π-bond of an alkene.52 (a) Stork, G.; Kreft, A. F., III. J. Am. Chem. Soc. 1977, 99. 3850. 3851. (b) Stork, G.; White, W. N. J. Am. Chem. Soc. 1975, 97, 4119. (c) Stork, G.; White, W. N. J. Am. Chem. Soc. 1956, 78, 4609.

44

(a) Bordwell, F. G.; Pagani, G. A. J. Am. Chem. Soc. 1975, 97, 118. (b) Bordwell, F. G.; Mecca, T. G. Ibid. 1975, 97, 123, 127. (c) Bordwell, F. G.; Wiley, P. F.; Mecca, T. G. Ibid. 1975, 97, 132.

45

46

(a) Smith, M. B.; Hrubiec, R. T.; Zezza, C. A. J. Org. Chem. 1985, 50, 4815. (b) Hrubiec, R. T.; Smith, M. B. Tetrahedron Lett. 1983, 24, 5031.

47

Kodama, M.; Matsuki, Y.; It^ o, S. Tetrahedron Lett. 1975, 3065.

48

For older examples, see (a) Kreft, A. Tetrahedron Lett. 1977, 1035. (b) Trost, B. M.; Tanigawa, Y. J. Am. Chem. Soc. 1979, 101, 4413.

(a) Johnson, C. R.; Dutra, G. A. J. Am. Chem. Soc. 1973, 95, 7777, 7783. (b) Ashby, E. C.; DePriest, R. N.; Tuncay, A.; Srivastava, S. Tetrahedron Lett. 1982, 23, 5251.

49

50

Cook, C.; Liron, F.; Guinchard, X.; Roulland, E. J. Org. Chem. 2012, 77, 6728.

51

Belelie, J. L.; Chong, J. M. J. Org. Chem. 2002, 67, 3000.

(a) Hrubiec, R. T.; Smith, M. B. Tetrahedron 1984, 40, 1457; (b) Idem J. Org. Chem. 1984, 49, 385; (c) Posner, G. H.; Ting, J. -S.; Lentz, C. M. Tetrahedron 1976, 32, 2281.

52

.

108

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

TBSO O

TBSO

(2 MeLi/CuI)

O

Me

Ether , –10°C to rt

P

OEt TBDPS = tert-Butyldiphenylsilyl

OEt

OTBDPS

OTBDPS

31

32 (96%, >95:5 dr)

The disconnections for this section include the following: R

R

C N

R

R

X

X

S N2

S N2

C N

3.2.1.4 SN2 Type Reactions With Epoxides Although ethers are generally unreactive, epoxides are the exception and they are highly reactive due to the strain inherent to a three-membered ring, and also due to the bond polarization imparted by the oxygen atom. Many nucleophiles can open the three-membered ring to yield a substituted or functionalized alcohol. Nucleophilic opening of an epoxide by hydroxide ion in an aprotic solvent is best described as an SN2 reaction. The nucleophile attacks the less sterically hindered carbon atom in 2-isopropyloxirane to yield an alkoxide, 1-hydroxy-3-methylbutan-2-olate. Subsequent protonation yields a diol, 3-methylbutane-1,2-diol. This example is representative of epoxide ring-opening reactions in aprotic solvents. In Feldman and Saunders’53 synthesis of ()-agelastatin A, the epoxide (R)-trimethyl (3-(oxiran-2-yl)prop-1-yn-1-yl)silane reacted with sodium azide, and the resulting alkoxide was protonated with aqueous ammonium chloride to yield an azido alcohol, (R)-1-azido-5-(trimethylsilyl)pent-4-yn-2-ol. O O :OH

Me2HC

OH H3O+

THF

Me2HC

Me2HC OH

2-Isopropyloxirane

OH

1-Hydroxy-3-methylbutan-2-olate

3-Methylbutane-1,2-diol

OH O

Me3Si

1. NaN3

Me3Si

2. aq NH4Cl

N3 (R)-Trimethyl(3-(oxiran-2-yl)prop-1-yn-1-yl)silane

(R)-1-Azido-5-(trimethylsilyl)pent-4-yn-2-ol

In the presence of a Lewis acid, ion-pair formation is possible, or even conversion to a carbocation, which is followed by attack of the nucleophile at the more substituted site. In Baskaran and coworker’s54 synthesis of indolizidine 167B, initial reaction of the epoxide unit in 33 with a Lewis acid (BF3) generated an oxonium ion. Subsequent reduction of the azide gave a primary amine, and attack of the ion-paired system at the more hindered site gave 34 in 50% overall yield. O

N3

HO N

1. BF3•OEt2 , –78°C 2. NaBH4

H 33

53

Feldman, K. S.; Saunders, J. C. J. Am. Chem. Soc. 2002, 124, 9060.

54

Reddy, P. G.; Varghese, B.; Baskaran, S. Org. Lett. 2003, 5, 583.

34

.

109

3.2 ALIPHATIC SUBSTITUTION REACTIONS

3.2.2 Unimolecular Nucleophilic Substitution: The SN1 Reaction Unimolecular substitution, which is ionization to a carbocation followed by reaction with a nucleophile, is generally less useful in synthesis. However, a carbocation-forming reaction may be the best specific choice to yield a particular functional group exchange. In Chapter 16, many examples will be discussed that involve carbocation intermediates. Unimolecular substitution occurs as a side reaction for many reactions done in aqueous media, and the extent of ionization can influence the yield, stereochemistry, and regiochemistry of the final product. The SN2 reaction is not the only pathway available for an alkyl halide. Ionization to a carbocation is a competitive process in reactions that have water as a solvent or cosolvent, where the substrate has a good leaving group attached to a carbon atom, and when the substrate is tertiary or secondary. If a nucleophile is present as the carbocation is formed, they react to form a substitution product. This two-step process of ionization followed by reaction with a nucleophile follows first-order kinetics since the initial ionization step has a higher activation barrier. Ionization to a carbocation is, therefore, much slower than the reaction of the nucleophile and the highly reactive carbocation. Therefore, this reaction is labeled as a unimolecular nucleophilic aliphatic substitution, or an SN1 reaction. A simple example is the reaction of 2-bromo-2-methylbutane with an excess of sodium azide in aq THF. An SN2 reaction of azide with the tertiary alkyl halide is not possible (see Section 3.2.1.1), but in the aqueous solvent, ionization becomes competitive to generate the tertiary carbocation, 2-methylbutan-2-ylium. This carbocation is solvated by water, which not only assists in the ionization, but also stabilizes the carbocation and bromide ion products by solvation. Subsequent reaction of the nucleophilic bromide ion with 2-methylbutan-2-ylium leads to the final isolated product, 2-azido-2-methylbutane, via an SN1 pathway. Ionization to the carbocation followed by reaction with the nucleophile constitutes the SN1 mechanism. –N 3

Excess NaN3

Br 2-Bromo-2-methylbutane

N3

THF/H2O

2-Methylbutan-2-ylium

2-Azido-2-methylbutane

An SN1 reaction can occur in any polar protic solvent (e.g., methanol, ethanol, or acetic acid), but ionization is significantly slower in these solvents relative to water because they are not as efficient for the solvation and separation of ions. If ionization is relatively slow, faster reactions (SN2 or elimination processes) often predominate if those reactions are possible. It is convenient to assume that SN1 reactions occur only in aqueous media, and use this assumption to predict product distributions. As with any assumption, this one fails to accurately predict the products in many cases, but it provides a good working model to begin analyzing reactions. The ultimate test, of course, is the actual experimental result. The mechanistic rationale described above results from many years of experimental observations about the SN1 reaction (statements 1–6).55 (1) The rate of reaction with tertiary halides is much faster than with primary halides due to more facile ionization to a more stable carbocation.56 The rate of reaction of KI with bromomethane or bromoethane in aqueous media are about the same, but 2-bromopropane reacts 11.6 times faster, and 2-bromo-2-aqueous reacts 1.2  106 times faster.56 (2) The rate equation can be described using only the concentration of the halide and not that of the nucleophile. In the reaction of 2-bromo-2-methylbutane, the rate-determining (slow) step is ionization of the CdBr bond and the fast second step (reaction of azide with 2-methylbutan-2-ylium) does not greatly influence the overall rate. The rate expression is Rate ¼ k [RX] for this unimolecular reaction.48 (3) Reaction with chiral halides proceeds with significant racemization rather than with inversion. (4) A cationic intermediate (e.g., 2-methylbutan-2-ylium) is formed rather than the transition state noted for an SN2 reaction (see Section 3.2.1.1). (5) The reaction often proceeds with rearrangement of the carbon skeleton of the organic substrate (6) Unless water is a solvent or cosolvent, the initial ionization reaction is rather slow and may not be competitive with other possible reactions. As indicted in (1), a carbocation can be formed by ionization of an alkyl halide, and its stability is critical to the overall rate of the SN1 process.57 A general ranking of carbocations by their relative stability is57 55

(a) Bateman, L. C.; Hughes, E. D.; Ingold, C. K. J. Chem. Soc. 1940, 960, 1011. (b) Harris, J. M. Prog. Phys. Org. Chem. 1974, 11, 89.

56

Ref. 5b, p 43.

57

(a) Olah, G. A.; Olah, J. A. Carbonium Ions; Olah, G. A., Schleyer, P. v. R., Eds.; John Wiley: New York, NY, 1969; Vol. 2, pp 715–782. (b) Richey, J. M. The Chemistry of Alkenes; Zabicky, J., Ed.; Interscience: New York, NY, 1970; Vol. 2, p 44.

.

110

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

R3 C + > H2 C]CHCH2 +  PhCH2 + > R2 CH + > RCH2 + > CH3 + This ranking is consistent with the number of electron-releasing carbon groups attached to the positive center. This inductive effect of three alkyl groups on the tertiary carbocation will partially diminish the point charge of the sp2 hybridized carbon atom, leading to greater stability of the tertiary carbocation relative to the primary, which has only one electron-releasing group. H—Cl

Cl—H

Cl Cl

Cyclobutylmethylium

Methylcyclobutan-1-ylium

Methylenecyclobutane

1-Chloro-1-methylcyclobutane

In an ionization reaction, it is reasonable to assume that the more stable carbocation will be formed as the major product of a reaction because the activation energy to formation is lower. The reaction of HCl with methylenecyclobutane illustrates this point, where the final product is 1-chloro-1-methylcyclobutane, used by Fitjer and Mandelt58 in a synthesis of cuparene. The π-bond reacts as a base, donating a pair of electrons to the acid (H+ ¼ HCl). As the new CdH bond is formed, a carbocation is generated at the other carbon of the π-bond, and there are two possible carbocations, methylcyclobutan-1-ylium and cyclobutylmethylium. The electron-releasing effects of the alkyl groups make the tertiary carbocation (1-chloro-1-methylcyclobutane) more stable than the primary carbocation (cyclobutylmethylium). The reaction will favor formation of the more stable tertiary carbocation, which then reacts with the nucleophilic chloride ion (generated by cleavage of HCl after transfer of the proton to the alkene) to yield the major product, 1-chloro-1-methylcyclobutane. The stability of the intermediate carbocation will determine formation of the major product in a unimolecular (ionization) reaction. Another way to diminish the net charge at the sp2 hybridized carbon of a carbocation is by charge dispersal due to resonance. The greater the dispersal of charge, the more stable the carbocation. When more atoms are involved in the charge dispersal, there are more resonance contributors and greater stability for this carbocation. As always, greater stability means that the carbocation is less reactive, whereas instability indicates the carbocation is more reactive. Both allyl and benzyl cations are resonance stabilized and are more stable than primary alkyl or secondary alkyl carbocations. In general, tertiary alkyl carbocations are probably more stable than a primary allylic carbocation, although a secondary or tertiary allylic carbocation will be more stable than a tertiary alkyl. Carbocation stability is also discussed in Section 16.2.1. C C

C C

C

C

Allyl carbocation

CH2

CH2

CH2

CH2

Benzyl carbocation

This relative ease of formation of allylic and benzylic carbocations is due, in part, to participation of the π-bond in the ionization process, but primarily by stabilization of the intermediate by resonance. The net charge at the cationic carbon is diminished by resonance delocalization in both the allyl and benzylic carbocation, making them more stable and easier to form. Although they are relatively stable, they are highly reactive intermediates and easily react with nucleophiles in an SN1 reaction. In a generic ionization reaction of an alkyl halide, carbocation 35 is formed as the product and it reacts with the best available nucleophile, which may be Y from the ionization process, the solvent, or a different added nucleophile. It is assumed that this intermediate is a solvent separated carbocation rather than an ion-pair. The central carbon of carbocation 35 is sp2 hybridized and trigonal planar and, as mentioned in Section 2.5.1, is conveniently viewed as being a

58

Mandelt, K.; Fitjer, L. Synthesis 1998, 1523.

.

111

3.2 ALIPHATIC SUBSTITUTION REACTIONS

trisubstituted carbon atom with an empty p orbital. Since 35 is trigonal planar, an attacking nucleophile can approach from either face. Therefore, reaction should generate a racemized substitution product. R R R

Y

R

R

R

R

:Nuc

– Y– RR

Nuc

+

Nuc

R

R R

35

This simple planar model is misleading because stereochemical retention or inversion is often observed in SN1 reactions. If the initial ionization generates a tight ion pair (represented by 36), then the leaving group (Y) provides some steric hindrance to the attacking nucleophile and the inverted substitution product may predominate, although it is usually not the exclusive product. If the leaving group (Y) can chelate to or otherwise coordinate with the nucleophile, then an ion pair (e.g., 37) may result, and delivery of the nucleophile will be from the same face as the leaving group. Such ion-pairing will lead to a product with retention of configuration. Depending on the solvent, an SN1 reaction may have 37 or 36 as an intermediate, but in general the reaction proceeds with significant if not complete racemization at the electrophilic center in acyclic and simple monocyclic compounds. H

R

H

Inversion

Nuc

R

Y

R

R

R

Nuc

R

Y

Retention

Nuc

R

R

H

Nuc

R

Y R

R

Nuc R

37

36

Relief of steric strain influences the rate of SN1 reactions, although the effect is relatively small. Structure 38 represents a generic alkyl chloride, and ionization will generate carbocation 39. When R is large, the planar carbocation 39 will likely have less steric hindrance than the tetrahedral precursor 38, accelerating the ionization process. Remember, however, that the carbocation is a high energy and reactive intermediate, so 38 is always lower in energy than 39. However, EtMe2CCl is ionized to the corresponding carbocation 1.7 times faster than Me3CCl, Et2MeCCl is ionized 2.6 times faster than Me3CCl, and Et3CCl is ionized 3.0 times faster than Me3CCl.59g These ionization rates illustrate that as the steric crowding in the halide increases, the rate of ionization is accelerated.59g This type of crowding, called back strain by Brown,59d contributes to the increased rate of ionization observed with hindered halides.60d R1 Cl

R

R R2

R1

R2 38

R=R1 =R2 = Me k =1 R=Et, R1 =R2 = Me k = 1.7 R=R1 =Et, R2 = Me k = 2.6 R=R1 =R2 = Et k = 3.0

39

In reactions with nucleophiles, primary halides commonly yield only SN2 reactions, but do not yield an SN1 reaction due to the high activation energy associated with formation of a primary carbocation. In effect, primary halides do not undergo SN1 reactions. Tertiary halides never undergo SN2 reactions, but can undergo SN1 reactions. Secondary halides yield a mixture of bimolecular and unimolecular products in aqueous solvents, but the relative amounts of each type of product are dependent on the nucleophile, the substrate, and (of course) the ionizing power of the solvent. Ionization to the planar sp2 carbon of the carbocation may be difficult for a polycyclic molecule because the requisite flattening of the ring system may be difficult. If a carbocation is generated from a polycyclic molecule, racemization does not always occur because the conformational constraints (see Section 1.5.4) may allow only one face to react with an attacking nucleophile. Adamantan-1-ol, for example, yields a carbocation (adamantan-2-ylium) upon treatment with aq HBr. The sp2 hybridized carbon atom prefers a planar geometry, but this requirement demands that the rigid (a) Fry, J. L.; Engler, E. M.; Schleyer, P. v. R. J. Am. Chem. Soc. 1972, 94, 4628. (b) Schleyer, P. v. R.; Nicholas, R. D. Ibid. 1961, 83, 182. (c) Tanida, H.; Tsushima, T. Ibid. 1970, 92, 3397. (d) Brown, H. C. J. Chem. Soc. 1956, 1248. (e) Brown, H. C.; Borkowski, M. J. Am. Chem. Soc. 1952, 74, 1894. (f) Brown, H. C.; Fletcher, R. S. J. Am. Chem. Soc. 1949, 71, 1845. (g) Harris, J. M.; Wamser, C. C. Fundamentals of Organic Reaction Mechanisms; John Wiley: New York, NY, 1976; p 118.

59

(a) Fittig, R. Ann. Chem. 1860, 114, 54. (b) Collins, C. J. Q. Rev. 1960, 14, 357. (c) Mundy, B. P.; Otzenberger, R. D. J. Chem. Educ. 1971, 48, 431. (d) B€ uchi, G.; MacLeod, W. D., Jr.; Padilla, O. J. J. Am. Chem. Soc. 1964, 86, 4438.

60

.

112

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

skeletal structure of adamantan-2-ylium be partially flattened. In addition, the polycyclic structure of adamantan-2ylium sterically inhibits approach of the nucleophilic bromide ion from any direction except the top of the molecule as it is drawn. As a result, ionization of adamantan-1-ol is somewhat slow, and reaction with the bromide ion leads to exclusive formation of 1-bromoadamantane, where the bromine atom is attached only to the exo-face of the polycyclic ring system.61 In general, reactions that generate carbocation intermediates can be highly stereoselective in substitution reactions, if there is steric bias for one face over the other. In the absence of such steric bias, these reactions often show poor-tomoderate stereoselectivity. Other examples of carbocation formation in polycyclic systems will be discussed in Chapter 16. Br

OH Br HBr H 2O

Adamantan-1-ol

Adamantan-2-ylium

1-Bromoadamantane

3.3 HETEROATOM-STABILIZED CARBOCATIONS When heteroatoms are attached to a positive carbon, donation of electrons (back-donation) to the electron-deficient center leads to greater stability. Two common examples are oxygen-stabilized carbocations (e.g., 40, which is known as an oxocarbenium ion, and sulfur-stabilized cations, e.g., 41). In both cases, back-donation from the heteroatom leads to resonance-stabilized carbocations, as shown. In general, both 40 and 41 are more stable than a simple tertiary carbocation, but the relative stability really depends on the nature of the groups of the sp2 carbon atom. Direct protonation of a carbonyl is the most common method for generating oxocarbenium ions. Protonation of the carbonyl moiety of ketones or aldehydes, and also acid derivatives is the most common way to generate an oxocarbenium ion intermediate (see Section 4.2). Another method for generating this carbocation is by the reaction of an acid with a vinyl ether. The reaction of 1-methoxy-2-methylprop-1-ene, for example, yields 1-methoxy-2-methylpropan-1-ylium rather than the tertiary carbocation 1-methoxy-2-methylpropan-2-ylium. Clearly, the oxygen-stabilized carbocation, known as an oxocarbenium ion, is more stable. As noted, the oxocarbenium ion is resonance stabilized, with the two resonance contributors shown: 1-methoxy-2-methylpropan-1-ylium and methyl(2methylpropylidene)oxonium. Similar results are observed upon protonation of thiocarbonyls or thioenol ethers. R C

R

O

C

R

O

C

40 Me

O Me

Me 1-Methoxy-2-methylpropan-2-ylium

R

S

C

S

41 Me

H+

O Me

Me

O Me

Me

Me 1-Methoxy-2-methylprop-1-ene

1-Methoxy-2-methylpropan-1-ylium

Me

O Me

Me Methyl(2-methylpropylidene)oxonium

3.4 SUBSTITUTION BY HALOGEN Halogen can be introduced into a molecule is several ways, including substitution of one halogen atom for another via an SN2 reaction (Section 3.2.1.1). This particular method is limited to the extent that more nucleophilic halide ions (e.g., the iodide ion) will displace a less nucleophilic halide (e.g., chloride or bromide), but the reverse reaction is generally not viable.

61

Dubowchik, G. M.; Padilla, L.; Edinger, K.; Firestone, R. A. J. Org. Chem. 1996, 61, 4676.

.

113

3.4 SUBSTITUTION BY HALOGEN

The Finkelstein reaction is highly useful for converting various halides into an iodide62 in which an alkyl chloride, bromide, mesylate, or tosylate is treated with NaI or KI to produce alkyl iodides via an SN2 reaction (Section 3.2.1.2). In a synthesis of sandresolide B, Trauner and coworkers63 converted mesylate 42 to iodide 43 in 84% yield with sodium iodide in acetone. Conversion to an iodide is required when a better leaving group is required for a given transformation. Since alkyl iodides are known to decompose upon long standing, the Finkelstein reaction allows the iodide to be generated in situ and used immediately in a subsequent reaction. Alcohols are directly converted to iodides using (PhO)3P+CH3I in DMF.64 Me MeO2SO

Me

H NaI , Acetone

I

H

85°C

Me

Me

42

43 (84%)

An important alternative method involves replacement of the OH unit of an alcohol with halogen. In addition, the hydrogen atom of an alkyl fragment, particularly an allylic or benzylic fragment can be replaced with halogen by a free radical mechanism. These latter two transformations will be examined in this section.

3.4.1 Halogenation of Alcohols Alcohols are converted to the corresponding alkyl halide by an acid-base reaction with HBr, HCl, or HI via an intermediate oxonium ion.65 This transformation is illustrated by a simple example in which 1-methylcyclohexan-1-ol reacts with HBr to yield oxonium ion (1-methylcyclohexyl)oxonium as the initially formed intermediate. This acidbase reaction is possible due to the amphoteric nature of the alcohol. Loss of H2O from the oxonium ion via ionization yields the tertiary methylcyclohexyl carbocation, methylcyclohexan-1-ylium. This carbocation reacts with the gegenion of the acid (Br in this case) to yield 1-bromo-1-methylcyclohexane. Me

H O

H

Me HBr

O

Me H

Me

Br

Br–

– H 2O

+ H+

1-Methylcyclohexan-1-ol

(1-Methylcyclohexyl)oxonium

Methylcyclohexan-1-ylium

1-Bromo-1-methylcyclohexane

Reaction of butan-1-ol with HBr also yields the expected primary bromide (1-bromobutane), but the reaction follows an SN2 mechanism, with bromide displacing the +OH2 moiety of the intermediate butyloxonium. Water (HOH) is a good leaving group in this reaction, whereas OH is not. However, formation of the oxonium ion provides a good leaving group, OH+2 . Ionization of butyloxonium to yield a primary carbocation is energetically unfavorable, which precludes an SN1 reaction. The SN2 transition state is favorable and leads to 1-bromobutane as the product. In general, tertiary alcohols react via a SN1 pathway, primary alcohols react via a SN2 pathway. In aqueous media, secondary alcohols usually react via a cationic mechanism, although a competitive SN2 mechanism also operates. If no solvent is used (neat), or if the reaction occurs in aprotic solvents, the SN2 mechanism dominates. In aqueous media, ionization and the SN1 pathway either dominates or becomes highly competitive. It is clear that the mechanism depends on the specific substrate and the reagents used, as well as on the reaction conditions. As a functional group exchange, however, reaction of primary, secondary, or tertiary alcohols with HBr or HCl is an excellent method for the preparation of alkyl bromides or chlorides. (a) Finkelstein, H. Berichte 1910, 43, 1528. (b) Ingold, C. K. Structure and Mechanism in Organic Chemistry, 2nd ed.; Cornell University Press: London, 1969; p 435. (c) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-31.

62

63

Chen, I. T.; Baitinger, I.; Schreyer, L.; Trauner, D. Org. Lett. 2014, 16, 166.

64

See Su, Q.; Dakin, L. A.; Panek, J. S. J. Org. Chem. 2007, 72, 2, in a synthesis of leucascandrolide A.

65

For examples, see Ref. 1, pp 689–696.

.

114

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

H

O

HBr

H

O

+ Br–

Butan-1-ol

Br

– H 2O

H

1-Bromobutane

Butyloxonium

Several halogenating reagents other than HBr, HCl, or HI have been discovered that convert alcohols to halides,65 and sulfur or phosphorus halides or oxyhalides are the most common reagents. Two of the most common are thionyl chloride (SOCl2) and thionyl bromide (SOBr2). Thionyl chloride reacts with an alcohol [e.g., (R)-butan-2-ol] to form an alkyl chlorosulfite ((R)-sec-butyl sulfurochloridite). Subsequent loss of ClSO 2 generates a tight ion pair (represented by 44), and chloride is transferred to carbon intramolecularly with retention of configuration, yielding (R)-2chlorobutane.66 The ion pair does not dissociate sufficiently to allow an SN1 type reaction. The mechanistic designator for this process is (Substitution Nucleophilic Internal), SNi, and requires that the leaving group possess an atom capable of being transferred to the carbon substrate.67 When primary alcohols are treated with thionyl chloride, the chloride is usually formed in excellent yield. An example is taken from Yang and coworker’s68 synthesis of an unnamed antifungal tricyclic o-hydroxy-p-quinone methide diterpenoid. In this work, benzylic alcohol (45) reacted with thionyl chloride to yield a 93% yield of benzylic chloride (46). O Cl

HO

S

Cl

O

H

S

O S

Cl

O

Cl

Cl

H

O

(R)-sec-Butyl sulfurochloridite

(R)-Butan-2-ol

– SO2

H

H

(R)-2-Chlorobutane

44

OMe

OMe

MeO

MeO SOCl2 CH2Cl2

OH

MeO Cl

45

46 (93%)

If the alcohol substrate has a stereogenic center, addition of an amine base (e.g., pyridine or triethylamine) after initial formation of the chlorosulfite ester, leads to inversion of configuration. Initial reaction with ethyl (S)-2-hydroxy-2-phenylacetate, for example, gave the chlorosulfite, ethyl (2S)-2-((chlorosulfinyl)oxy)-2-phenylacetate. Pyridine is added to the reaction, and it reacts as a base with the HCl byproduct to yield pyridinium hydrochloride. The nucleophilic chloride ion is now in solution rather than being lost as HCl, and it displaces the chlorosulfite unit in an SN2 reaction (see 47) that yields ethyl (R)-2-chloro-2-phenylacetate with inversion of configuration. The reaction of an alcohol with thionyl chloride and pyridine is called the Darzens’ procedure.69 A modification that improves the stereoselectivity premixes the alcohol with thionyl chloride and pyridine, and the resulting sulfinate is treated with thionyl bromide.70 Direct reaction of the alcohol with thionyl bromide converts alcohols to the bromide,71 but the extent of inversion or retention is not as clean as with thionyl chloride. An acid byproduct is produced, however, and sensitive systems (particularly allylic alcohols) may be subject to acid-catalyzed rearrangements and side reactions.

66

(a) Lee, C. C.; Clayton, J. W.; Lee, D. G.; Finlayson, A. J. Tetrahedron 1962, 18, 1395. (b) Lee, C. C.; Finlayson, A. J. Can. J. Chem. 1961, 39, 260.

67

(a) Smith, M. B. March's Advanced Organic Chemistry, 7th ed.; John Wiley & Sons, Inc.: Hoboken, NJ, 2013; p 408. (b) Lewis, E. S.; Boozer, C. E. J. Am. Chem. Soc. 1952, 74, 308.

68

Huang, J.; Foyle, D.; Lin, X.; Yang, J. J. Org. Chem. 2013, 78, 9166.

69

Darzens, G. Compt. Rend. 1911, 152, 1601.

70

Frazer, M. J.; Gerrard, W.; Machell, G.; Shepherd, B. D. Chem. Ind. 1954, 931.

71

Elderfield, R. C.; Kremer, C. B.; Kupchan, S. M.; Birstein, O.; Cortes, G. J. Am. Chem. Soc. 1947, 69, 1258.

.

115

3.4 SUBSTITUTION BY HALOGEN O

Ph HO

CO2Et

Cl

S

Ph Cl

CO2Et O

H

H

S

Ph O N

Cl S

Cl

O

O Ethyl (S)-2-hydroxy2-phenylacetate

CO2Et – SO2

– HCl

Ethyl (2S)-2-((chlorosulfinyl)oxy)-2-phenylacetate

Ph

CO2Et

H H

H Cl N

Cl

Ethyl (R)-2-chloro2-phenylacetate

47

Phosphorus halides react with primary, secondary, and tertiary alcohols to yield the corresponding alkyl halide. Apart from thionyl chloride or bromide, the most common reagents are phosphorus trichloride (PCl3), phosphorus pentachloride (PCl5), phosphorus oxychloride (phosphoryl chloride, POCl3), phosphorus tribromide (PBr3), and phosphorus pentabromide (PBr5). The chloride reagents give little or no rearrangement in reactions with alcohols when compared with HBr or HI. Some rearrangement may be observed with secondary alcohols, even with the phosphorus reagents. When allylic alcohols are subjected to acidic conditions, formation of the allyl cation can lead to regioselectivity problems, as well as polyalkylation or SN20 reaction byproducts. Most of the halogenating reagents described can give deleterious side reactions. Such problems can be circumvented or alleviated by using a different type of halogenating reagent. Magid et al. 72 showed that allylic alcohols (e.g., 3-methylbut-2-en-1-ol) are converted to the corresponding allylic chloride (1-chloro-3-methylbut-2-ene) by reaction with a mixture of triphenylphosphine (PPh3) and hexachloroacetone. The active chlorinating agent is probably chlorotriphenylphosphonium chloride (CldPPh+3 Cl), formed in situ. O Cl3C

OH

CCl3 PPh3

3-Methylbut-2-en-1-ol

Cl 1-Chloro-3-methylbut-2-ene

Phosphorus trichloride is commonly used for the preparation of acid chlorides,73 but less often for the conversion of alcohols to chlorides because other reagents are more practical. This choice contrasts with phosphorus tribromide, which is a common reagent for conversion of alcohols to bromides.74 However, problems persist in reactions with sensitive alcohols. An alternative and convenient method for converting an alcohol to an alkyl bromide uses a mixture of carbon tetrabromide and triphenylphosphine. An example is the conversion of (R)-4,8-dimethylnon-7-en-2-yn-1-ol to (R)-9-bromo-2,6-dimethylnon-2-en-7-yne in 90% yield, taken from Lee and Geum’s75 synthesis of panaginsene. Triphenylphosphine and NCS (see Section 3.4.2) converts allylic alcohols to allylic chlorides.76 Conversion of an alcohol to a mesylate followed by treatment with LiBr is a mild method for preparing bromides,77 and treatment with LiCl yields the chloride.78 A mixture of carbon tetrachloride and hexamethylphosphorus triamide (HMPT) has been used to convert alcohols to chlorides.79

(a) Magid, R. M.; Fruchey, O. S.; Johnson, W. L.; Allen, T. G. J. Org. Chem. 1979, 44, 359. (b) Magid, R. M.; Fruchey, O. S. J. Am. Chem. Soc. 1977, 99, 8368.

72

73

Allen, C. F. H.; Barker, W. E. Org. Synth. Coll. Vol. 2, 1943, 156.

74

Fieser, L. F.; Fieser, M. Reagents for Organic Synthesis; John Wiley: New York, NY, 1967; Vol. 1, p 873.

75

Geum, S.; Lee, H.-Y. Org. Lett. 2014, 16, 2466.

76

For an example from a synthesis, see Lan, J.; Li, J.; Liu, Z.; Li, Y.; Chan, A. S. C. Tetrahedron: Asymmetry 1999, 10, 1877.

For examples from syntheses, see (a) Clive, D. L. J.; Hisaindee, J. Org. Chem. 2000, 65, 4923. (b) Huang, A. X.; Xiong, Z.; Corey, E. J. J. Am. Chem. Soc. 1999, 121, 9999.

77

78

Smith, A. B., III; Wan, Z. J. Org. Chem. 2000, 65, 3738.

79

For an example taken from a synthesis of (+)-australine, see White, J. D.; Hrnciar, P. J. Org. Chem. 2000, 65, 9129.

.

116

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

CBr4 , PPh3

OH

Br

(R)-4,8-Dimethylnon7-en-2-yn-1-ol

(R)-9-Bromo-2,6-dimethylnon- (90%) 2-en-7-yne

The bromination method using triphenylphosphine and either CBr4 or elemental bromine is a particularly mild method for this transformation. Cyclopropylcarbinyl alcohols are acid sensitive substrates, for example, usually giving ring-opening products or cyclobutane derivatives on treatment with acid (see Section 16.2.3).80 When 1-cyclopropyl-2-methylpropan-1-ol was treated with PPh3 and bromine, however, a 79% yield of (1-bromo-2-methylpropyl)cyclopropane was obtained.81,82 In this work, it was essential that the bromide-triphenylphosphine-alcohol complex be heated in order to generate this bromide, since mixing the reagents without heating gave only starting material.81 As with other halogenating agents, primary alcohols react to yield mainly the substitution product, the alkyl halide. Phosphorus pentabromide can be used with secondary alcohols in place of PBr3 since much less rearrangement occurs.83 Bromination of aromatic rings has been reported using POBr3, as in the multikilogram-scale synthesis of a TRPV1 antagonist by Cleator et al.,84 in which 6-hydroxy-5-(trifluoromethyl)nicotinonitrile was converted to 6-bromo-5-(trifluoromethyl)nicotinonitrile in >70% yield. In this particular case, 6-hydroxy-5-(trifluoromethyl)nicotinonitrile was generated in situ rather than isolated, and then treated with POBr3. OH

Br

1. PPh3 , DMF Pentane 2. Br2 ; Distil

1-Cyclopropyl-2-methylpropan-1-ol NC

CF3

(1-Bromo-2-methylpropyl)cyclopropane

POBr3 , MeCN

NC

CF3

80°C

N

OH

N

6-Hydroxy-5-(trifluoromethyl)nicotinonitrile

Br

6-Bromo-5-(trifluoromethyl)nicotinonitrile

(>70%)

In combination with pyridine, phosphoryl chloride (POCl3) reacts with alcohols to form an intermediate that undergoes elimination to yield an alkene (see Section 3.5.1 for elimination reactions). The elimination generally yields the best results with tertiary alcohols, although secondary alcohols work as well. An example is taken from a synthesis of amphidinolide B2 by Carter and coworkers,85 in which the reaction of allylic alcohol 48 with PBr3 and pyridine gave bromide 49 in 99% yield. A mixture of thionyl chloride and pyridine can also lead to elimination.86 Substitution by chloride ion sometimes competes with elimination when secondary alcohols are substrates. Treatment of alcohol methyl 10-hydroxytridec-12-ynoate with POCl3 and pyridine, for example, gave the chloride methyl 10-chlorotridec-12-ynoate rather than the conjugated ene-yne.87

80

Descoins, C.; Samain, D. Tetrahedron Lett. 1976, 745.

81

Hrubiec, R. T.; Smith, M. B. J. Org. Chem. 1984, 49, 431.

(a) Kirmse, W.; Kapps, M.; Hager, R. B. Chem. Ber., 1966, 99, 2855. (b) Wiley, G. A.; Hershkowitz, R. L.; Rein, B. M.; Chung, B. C. J. Am. Chem. Soc. 1964, 86, 964. 82

83

Fieser, L. F.; Fieser, M. Reagents for Organic Synthesis; John Wiley: New York, NY, 1967; Vol. 1, p 866.

84

Cleator, E.; Scott, J. P.; Avalle, P.; Bio, M. M.; Brewer, S. E.; Davies, A. J.; Gibb, A. D.; Sheen, F. J.; Stewart, G. W.; Wallace, D. J.; Wilson, R. D. Org. Process Res. Dev. 2013, 17, 1561. 85

Lu, L.; Zhang, W.; Nam, S.; Horne, D. A.; Jove, R.; Carter, R. G. J. Org. Chem. 2013, 78, 2213.

86

For an example taken from a synthesis of (–)-cacospongionolide F, see Demeke, D.; Forsyth, C. J. Org. Lett. 2003, 5, 991.

87

Truscheit, E.; Eiter, K. Liebigs Ann. Chem. 1962, 658, 65.

.

117

3.4 SUBSTITUTION BY HALOGEN

Et3SiO Me HO

Et3SiO Me

Ph

Ph

SOCl2 , Py PhCH3 , –78 °C

OSi(i-Pr)3

OSi(i-Pr)3

48

49 (99%)

(CH2)8CO2Me

(CH2)8CO2Me

POCl3 Py

OH Methyl 10-hydroxytridec-12-ynoate

Cl Methyl 10-chlorotridec-12-ynoate

Although most phosphorus chlorides and bromides are commercially available, it is sometimes necessary to prepare the appropriate reagent, as it is needed. The phosphorus iodides have poor shelf lives because they are unstable and decompose under mild conditions, so they are usually prepared in situ, or immediately prior to use by reaction of red phosphorus with iodine. Using phosphorus and I2 is a common method for the conversion of aliphatic alcohols to aliphatic iodides. An example is the conversion of cetyl alcohol (hexadecan-1-ol) to cetyl iodide (1-iodohexadecane) in 85% yield.88 A related method is illustrated by treatment of (S)-5-methyl-2-(prop-1-en-2-yl)hex-5-en-1-ol with triphenylphosphine, iodine, and imidazole. In this example, taken from George and coworker’s89 synthesis of (+)garcibracteatone, the primary alcohol unit was converted to (S)-3-(iodomethyl)-2,6-dimethylhepta-1,6-diene in 71% yield. Phosphorus bromides can also be prepared from phosphorus and bromine. Reaction of propane-1,2,3-triol with red phosphorus and bromine, for example, gave 1,3-dibromopropan-2-ol.90 C15 H31

OH

2 Pred , 3 I2

C15 H31

150°C

HO

I

1-Iodohexadecane (85%)

Hexadecan-1-ol

I2 , PPh3 , CH2Cl2

I

Imidazole , rt

(S)-5-Methyl-2-(prop-1-en-2-yl)hex-5-en-1-ol

(S)-3-(Iodomethyl)-2,6-dimethyl(71%) hepta-1,6-diene

There are other functional group transformations that yield alkyl halides. A classical transformation converts carboxylates to alkyl bromides ðC  CO2  ! C  BrÞ. This transformation is called the Hunsdiecker reaction91 (it has also been called the Borodin reaction),92 and is most useful for the preparation of secondary halides. Note that Borodin93 was a noted composer of classical music, as well as a chemist. The silver salt of a carboxylic acid is heated with bromine to yield the bromide via decarboxylation. An example is conversion of the silver salt of cyclohexane carboxylic acid [silver (I) cyclohexanecarboxylate] to bromocyclohexane.94 In a different process, phosphorus and bromine react with acids in

88

(a) Ref. 83, p 862; (b) Hartman, W. W.; Byers, J. R.; Dickey, J. B. Org. Synth. Coll. Vol. 2 1943, 322.

89

Pepper, H. P.; Tulip, S. J.; Nakano, Y.; George, J. H. J. Org. Chem. 2014, 79, 2564.

McElroy, W. T.; DeShong, P. Tetrahedron 2006, 62, 7155. For the use of PBr5 in this type of reaction, see Kaslow, C. E.; Marsh, M. M. J. Org. Chem. 1947, 12, 456. 90

(a) Hunsdiecker, H.; Hunsdiecker, Cl. Berichte 1942, 75, 291. (b) Johnson, R. G.; Ingham, R. K. Chem. Rev. 1956, 56, 219. (c) Wilson, C. V. Org. React. 1957, 9, 332 (see p 341). (d) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-47. For a microwavae-promoted version, see Bazin, M.-A.; El Kihel, L.; Lancelot, J.-C.; Rault, S. Tetrahedron Lett. 2007, 48, 4347.

91

92

Borodine, A. Annalen 1861, 119, 121.

93

See the biography by (a) Abraham, G.; Seroff, V. I. The Mighty Five, 1948. See also, (b) Zetlin, M. O. The Five (tr. 1959).

94

(a) Eliel, E. L.; Haber, R. G. J. Org. Chem. 1959, 24, 143. (b) Marvell, E. N.; Sexton, H., Ibid. 1964, 29, 2919.

.

118

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

the Hell-Volhard-Zelinsky reaction.95 An example is the conversion of cyclopropanecarboxylic acid to the α-bromo acid bromide 1-bromocyclopropane-1-carbonyl bromide, which was reacted with 4-aminopyridine to yield 1-bromo-Nmethyl-N-(pyridin-4-yl)cyclopropane-1-carboxamide in Storey and Ladwa’s synthesis of 5-azaoxindoles.95e CO2 – Ag+

Br

Br2 , Heat

+

Silver(I) cyclohexanecarboxylate

N

cat Pred

OH

AgBr

Br

2.5 Br2

O

N

NHMe

Br N

EtN(i-Pr)2 , CH2Cl2

Br Cyclopropanecarboxylic acid

+

Bromocyclohexane O

O

CO2

Me 1-Bromo- N-methyl-N-(pyridin4-yl)cyclopropane-1-carboxamide

1-Bromocyclopropane1-carbonyl bromide

3.4.2 Allylic and Benzylic Halogenation Bromine or chlorine can be introduced into a molecule by processes that generate radical intermediates.96 The reaction of bromine or chlorine with allylic and benzylic compounds, in the presence of heat, light, and/or radical initiators (Sections 17.3 and 17.4) leads to resonance-stabilized free radicals that react with additional bromine or chlorine to yield the corresponding halide. Addition of diatomic bromine to methyl 2-methyl-3-nitrobenzoate, in the presence of benzoyl peroxide and with photochemical initiation, gave benzylic bromide methyl 2-(bromomethyl)-3-nitrobenzoate in high yield, in a synthesis of indole alkaloids.97 This reaction is just another example of the well-known free radical substitution reaction of alkanes with chorine or bromine in the presence of light (hν) or heat. The resonance stability of an allylic or a benzylic radical is a significant contributor to the benzylic or allylic selectivity. In nonallylic or benzylic systems, radical chlorination of alkanes by chlorine is unselective, converting virtually all CdH bonds in the molecule to CdCl. Bromination, on the other hand, is more selective in such systems, showing a great propensity to react with tertiary positions (tertiary CdH) to yield tertiary bromides, as in the conversion of 2-methylpentane to 2-bromo-2-methylpentane. Chlorination or bromination of allylic and benzylic sites is usually a highly useful synthetic transformation. CO2Me CH3

CO2Me Br

Br2 , CCl4 , h Benzoyl peroxide

NO2

NO2

Methyl 2-methyl3-nitrobenzoate

CH3

Methyl 2-(bromomethyl)3-nitrobenzoate

CH3 Br

Br2 , h

CH3 2-Methylpentane

CH3 2-Bromo-2-methylpentane

Halogenation of allylic and benzylic CdH compounds is more convenient using the readily available NBS or NCS with radical initiators, heat and/or light,98 in what is called the Wohl-Ziegler reaction.98f–h,99 Reactions with NBS are (a) Hell, C. Berichte 1881, 14, 891. (b) Volhard, J. Annalen 1887, 242, 141. (c) Zelinsky, N. Berichte 1887, 20, 2026. (d) Watson, H. B. Chem. Rev. 1930, 7, 180. (e) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-42. (e) Storey, J. M. D.; Ladwa, M. M. Tetrahedron Lett. 2006, 47, 381.

95

96

For examples, see Ref. 1, pp 611–615.

97

S€ oderberg, B. C.; Chisnell, A. C.; O'Neil, S. N.; Shriver, J. A. J. Org. Chem. 1999, 64, 9731.

(a) Goldfinger, P.; Gosselain, P. A.; Martin, R. H. Nature (London) 1951, 168, 30. (b) McGrath, B. P.; Tedder, J. M. Proc. Chem. Soc. 1961, 80. (c) Pearson, R. E.; Martin, J. C. J. Am. Chem. Soc. 1963, 85, 354. (d) Russell, G. G.; DeBoer, C.; Desmond, K. M. Ibid. 1963, 85, 365. (e) Walling, C.; Rieger, A. L.; Tanner, D. D. Ibid. 1963, 85, 3129. (f) Wohl, A. Berichte 1919, 52, 51. (g) Ziegler, K. Sp€ath, A.; Schaaf, E.; Schumann, W.; Winkelmann, E. Annalen 1942, 551, 80. (h) Djerassi, C. Chem. Rev. 1948, 43, 271.

98

99

The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-102. .

119

3.4 SUBSTITUTION BY HALOGEN

commonly carried out in carbon tetrachloride heated at reflux, with addition of a catalytic amount of radical initiator (e.g., AIBN or benzoyl peroxide, Section 17.3), or with an incandescent lamp held close to the reaction. Heating AIBN leads to homolytic cleavage and formation of a cyano radical (Me2CdCN), which initiates the radical chain process (Section 17.3). This chain process generates the allylic (or benzylic) radical that reacts with bromine (or additional NBS) to yield the allylic or benzylic bromide product and another radical chain carrier.98 Chlorination reactions with NCS proceed by a similar mechanism. A synthetic example is taken from Hu and coworker’s100 synthesis of asterredione, where the benzylic position relative to the indole unit in 50 was converted to 51 in >95% yield by reaction with NBS and AIBN. O

O

N

Br

N

O N-Bromosuccinimde (NBS)

Cl

O N-Chlorosuccinimide (NCS) Br

CO2Me

CO2Me NBS , AIBN CCl4

N

N

Boc

Boc

50

51 (>95%)

Ketones that have an enolizable hydrogen atom can be halogenated at the α-position (the carbon adjacent to the carbonyl carbon atom) with bromine, chlorine, NBS, or NCS.101 The reaction probably proceeds via addition of X2 to the enol form of the carbonyl (Sections 2.5.1, 13.2.1, and 13.8.1). Elimination of HX from the addition product generates a new enol, which tautomerizes to the α-haloketone.101 Reaction of cyclohexanone with bromine, for example, yields 2-bromocyclohexan-1-one,102,103 and reaction with NCS yields 2-chlorocyclohexan-1-one.102 Halogenation of carbonyl compounds is often accompanied by secondary reactions. The initially formed α-halo carbonyl can react with another equivalent of halogen to yield the α,α-dihalocarbonyl. In the bromination of cyclohexanone, 2,2-dibromocyclohexan-1-one was also produced. The mono-halogenated product usually predominates with ketones, but the extent of dihalogenation depends on the substrate and reaction conditions. Halogenation of lactams using this method often yields a α,α-dihalide. 2-Pyrrolidinone, for example, reacted with NCS or PCl5, to yield a mixture of 3-chloro- and 3,3-dichloro-2-pyrrolidinone.104 When phosphorus halides are used, the dichloride is usually the major product, but the monochloride can be formed in reasonable yield using NCS. O

O

O Cl

NBS or Br2

NCS

O Br

+

Br Br

2-Chlorocyclohexan-1-one

Cyclohexanone

2-Bromocyclohexan-1-one

2,2-Dibromocyclohexan-1-one

In unsymmetrical ketones, halogenation usually occurs at the more highly substituted α-position, but regioselectivity can be a problem. Heating an unsymmetrical ketone with tert-butyl bromide and DMSO at 65°C is a mild and selective solution to the problem,105 where the more substituted bromide is the major product. Another problem

100

Gai, S.; Zhang, Q.; Hu, X. J. Org. Chem. 2014, 79, 2111.

101

Catch, J. R.; Hey, D. H.; Jones, E. R. H.; Wilson, W. J. Chem. Soc. 1948, 276.

See (a) Sreedhar, B.; Surendra Reddy, P.; Madhavi, M. Synth. Commun. 2007, 37, 4149. (b) Arbuj, S. S.; Waghmode, S. B.; Ramaswamy A. V. Tetrahedron Lett. 2007, 48, 1411. 102

103

See Meshram, H. M.; Reddy, P. N.; Sadashiv, K.; Yadav, J. S. Tetrahedron Lett. 2005, 46, 623.

104

Elberling, J. A.; Nagasawa, H. T. J. Heterocyclic Chem. 1972, 9, 411.

105

Armani, E.; Dossena, A.; Marchelli, R.; Casnati, G. Tetrahedron 1984, 40, 2035. .

120

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

in halogenation reactions of carbonyls is elimination (note that POCl3 is produced as a side product when PCl5 is used; see Section 3.4.1). When cyclopentanone reacted with PCl5, for example, 1-chlorocyclopentene was the major product, via elimination, but 1,1-dichlorocyclopentane was also formed by addition of chloride ion to the α-chlorocarbocation.106 Similarly, cyclohexanone gave 1-chlorocyclohex-1-ene.107 O

PCl5 , CH2Cl2

Cl

– POCl3

1-Chlorocyclohex-1-ene

The transform to convert alcohols to halides follow: R

R Cl [Br , I]

OH

R

R

3.5 ELIMINATION REACTIONS There are several types of reactions in which a halogen, a sulfonate ester, or another leaving group is lost from a molecule, along with a hydrogen atom or sometimes a functional group, to generate a carbon-carbon double bond. The loss of two atoms or groups from the starting material leads to the term, elimination reaction. Elimination reactions generally produce alkenes or alkynes. This section will examine methods for the formation of such molecules via elimination reactions.108 A monograph is available that describes several different preparations of alkenes.109

3.5.1 Bimolecular 1,2-Elimination: The E2 Reaction Elimination is a major class of functional group-exchange reactions in which a hydrogen atom and a leaving group are removed from a molecule, generating in a π-bond. However, removal of the hydrogen atom in such a reaction is actually an acid-base reaction. The reaction proceeds with loss of the elements of H and X, which leads to the name elimination. The characteristic features of nucleophiles and their reactions were discussed in Section 3.2.1. In many cases, the nucleophile used in a substitution reaction is also a good Brønsted-Lowry base. Although there is strong attraction between the δ+ dipole on the carbon bearing the leaving group and the base, collision with this carbon is unproductive in tertiary substrates due to the high activation energy for the SN2 transition state (see Section 3.2.1.1). In such molecules, bond-polarization induced by an attached halogen atom extends down the carbon chain to the β-hydrogen, which has a δ+ dipole. Collision of the base with this β-hydrogen is an acid-base reaction that removes the hydrogen, as indicated in 52. As the base forms a bond with the β-hydrogen atom, the electron density in that CdH bond migrates from the β-carbon atom toward the δ+ dipole of the carbon bearing the X group. As the leaving group departs, a new π-bond is formed and the product is an alkene. This reaction follows second-order kinetics, and is given the designation of E2 reaction (bimolecular elimination). The E2 reaction is synchronous rather than stepwise,110 so there is no intermediate, and all experimental observations can be explained by a transition state (e.g., 53). Note, however, that a “unified rule for elimination” has been proposed that predicts regioselectivity for a wide range of substrates.111 This rule appears to apply best when ion-pairing occurs, and is not as useful in cases where there is no ion paring.

106

Charpentier-Morize, M.; Sansoulet, J. Bull. Chim. Soc. Fr. 1977, 331.

107

Baldwin, J. E.; Burrell, R. C. J. Org. Chem. 1999, 64, 3567.

108

For examples, see Ref. 1, pp 251–263, 569–580.

109

Williams, J. M. J., Ed., Preparation of Alkenes; Oxford Press: Oxford, 1996.

110

(a) Saunders, W. H., Jr.; Cockerill, A. F. Mechanisms of Elimination Reactions; Wiley-Interscience: New York, NY, 1973. (b) Bordwell, F. G. Acc. Chem. Res. 1972, 5, 374. (c) Dhar, M. L.; Hughes, E. D.; Ingold, C. K.; Masterman, S. J. Chem. Soc. 1948, 2055, 2058, 2065. (d) Sneen, R. A. Acc. Chem. Res. 1973, 6, 46. (e) Bordwell, F. G. Ibid. 1970, 3, 281. 111

Gevorkyan, A. A.; Arakelyan, A. S.; Cockerill, A. F. Tetrahedron 1997, 53, 7947.

.

121

3.5 ELIMINATION REACTIONS

R2 R1

X

R2

R2

R1 +

R1

X

H

+

R4

H Base

R3

R4

Base

R4

R3

52

R3

53

The elimination sequence begins with an acid-base reaction, where the acid is a hydrogen atom attached to the β-carbon relative to the leaving group (the β-hydrogen on the β-carbon, see 52). It has been determined that as the added base reacts with the β-hydrogen, the favored rotamer that leads to the E2 transition state will have the leaving group and the β-hydrogen in an anti relationship (i.e., 180 degrees apart).110 The requirement for an anti relationship makes the stereochemistry of the alkene product predictable. Rotation about the CdC bond is not possible when bond making and bond breaking begins, and the spatial relationships of the groups in the transition state (53) will be the retained in the alkene product. In 52, the relative position of the alkyl groups in the anti transition state (note that R1 and R3 are on the same side) is retained in transition state 53 and in the alkene as the π-bond is formed. In other words, the E2 reaction is stereospecific. The stereospecificity of the reaction is apparent when (2S,3R)-2-bromo-2-phenylbutane, drawn as its Fischer projection, is treated with a strong base (e.g., tert-butoxide). Only one diastereomer, (Z)-2-phenylbut-2-ene, is formed via E2 elimination. The rotamer that leads to the E2 transition state required to form this product has the bromine leaving group and the β-hydrogen in an anti conformation, as shown in 54. The relative positions of phenyl and methyl are syn in 54, and rotation is not possible in the transition state so elimination of the elements of H (to tert-butoxide) and Br as the bromide ion yields (Z)-2-phenylbut-2-ene, exclusively. Similar reaction with the (2R,3R)-diastereomer of 2-bromo-3-phenylbutane yields only (E)-2-phenylbut-2-ene. Br H

Me

Me

H

O – K+

O – K+

OH

Me

Ph

H

H

Me

Me

H

– Br–

Br

Ph

Me

Ph syn

((2S,3 R)-2-Bromo-2-phenylbutane

54

(Z)-2-Phenylbut-2-ene

Some alkenes have more than one β-hydrogen atom, and when there are two different β-hydrogen atoms, Ha and Hb, there is the possibility of forming two different alkene products (regioisomers). Removal of Hb in (2R,3S)-2-bromo-3-ethylhexane, for example, will lead to a double bond between C2dC3 [(Z)-3-ethylhex-2-ene], but removal of Ha will lead to a double bond between C1dC2 (3-ethylhex-1-ene). These two alkenes are regioisomers. While it may not be obvious, Ha is slightly more acidic than Hb since it is attached to the less substituted carbon. The base should react with the more acidic hydrogen atom, which would lead to the less substituted alkene, 3-ethylhex-1-ene, but that is not the major product. Acid-base reactions are under kinetic control, where the more acidic hydrogen atom is preferentially removed. If 3-ethylhex-1-ene is not the major product, the E2 reaction is not under kinetic control. The observed major product is the more substituted alkene, 3-ethylhex-2-ene, which is the thermodynamically more stable alkene product because the C]C has more electron-releasing substituents. If the thermodynamic product is formed preferentially, it is reasonable to assume that the E2 reaction is under thermodynamic control. In E2 reactions, the major product is always the more highly substituted alkene. This observation can be explained by the Hammond postulate112 (if two states, as, e.g., a transition state and an unstable intermediate, occur consecutively during a reaction process and have nearly the same energy content, their interconversion will involve only a small reorganization of the molecular structures).113 To produce the more substituted alkene product, the transition state for this reaction (53) must be product-like (a late transition state), so that factors stabilizing the alkene product will have a strong influence on the transition state. Alkyl groups that are attached to an sp2 carbon release electrons to the π-bond and enhance the strength of the π-bond. A more highly substituted alkene is therefore more stable, and a relative order of alkene stability is R2 C]CR2 > R2 C]CHR > RCH]CHR > R2 C]CH2 > RCH]CH2 > H2 C]CH2 .114 112

Farcasiu, D. J. Chem. Educ. 1975, 52, 76.

113

Harnmond, G. S. J. Am. Chem. Soc. 1955, 77, 334.

(a) Brown, H. C.; Moritani, I. J. Am. Chem. Soc. 1953, 75, 4112. (b) Brown, H. C.; Moritani, I. Ibid. 1956, 78, 2203. (c) Brown, H. C.; Nakagawa, M. Ibid. 1956, 78, 2197.

114

.

122

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

If the Hammond postulate is applied to the observation that 3-ethylhex-1-ene is the major product, the transition state leading to that product must be favored (lower activation energy) over the transition state of 3-ethylhex-1-ene. The E2 reaction is an acid-base reaction, and all acid-base reactions are equilibrium processes and under thermodynamic control. Therefore, it is reasonable to assume that formation of the more stable product is favored where a trisubstituted alkene is thermodynamically more stable than a monosubstituted alkene. Indeed, formation of the more stable alkene is characteristic of E2 reactions, and is usually termed Saytzeff (Zaitsev) elimination.115 Formation of the more substituted alkene is favored, and this term is also associated with E1 reactions.

Et

O – K+

Hb H

H

H

C3H7(Et)H bC

H

– Ha

Ha

C3H7 Br

H H (2R,3S)-2-Bromo-3-ethylhexane

3-Ethylhex-1-ene O – K+

Et

H – Hb

CH2Hb

Et

(Z)-3-Ethylhex-2-ene

CH2Ha 2

3

C3H7

1

H

Hb

Br C3H7

(2R,3S)-2-Bromo-3-ethylhexane

Cyclic halides pose a problem for E2 reactions because the E2 transition state demands an anti relationship for the β-hydrogen atom and the leaving group, and attaining the appropriate conformation may be problematic with many cyclic structures.116 While the E2 transition state is relatively easy to attain in five-membered rings, six-membered rings are problematic. The anti transition state for an E2 reaction in the cyclohexane system can only be attained if the β-hydrogen and the leaving group are trans-diaxial. Before considering an E2 reaction for (1R,3R)-1-bromo-3-methylcyclohexane, both equilibrating chair conformations (1R,3R)-1-bromo-3-methylcyclohexane-A and (1R,3R)-1-bromo-3-methylcyclohexane-B must be examined. Since elimination of bromine requires a trans-diaxial β-hydrogen atom when a base is added, only (1R,3R)-1bromo-3-methylcyclohexane-B has Hb trans-diaxial when the bromine is axial. Although there are two regioisomeric alkenes, 4-methylcyclohex-1-ene and 3-methylcyclohex-1-ene, only 3-methylcyclohex-1-ene can be formed by an E2 reaction. When the bromine is equatorial in 3-methylcyclohex-1-ene, there are no β-hydrogen atoms that can attain an anti conformation relative to the bromine, so an E2 elimination is impossible. H Ha

H Br

Br Me

H

Hb

(1R,3R)-1-Bromo-3methylcyclohexane-A

Me

+

H Me

Me

O– K+

Ha

OH

Hb

(1R,3R)-1-Bromo-3methylcyclohexane-B

Et

4-Methylcyclohex-1-ene

Et

Et

H

Br H

H Br (1R,2r,3S)-2-Bromo-1,3diethylcyclohexane-A

3-Methylcyclohex-1-ene

Et H

(1R,2r,3S)-2-Bromo-1,3diethylcyclohexane-B

(1R,2r,3S)-2-Bromo-1,3-diethylcyclohexane poses a different problem in that there are two β-hydrogen atoms at C2 and C6 relative to the carbon bearing the bromine atom. There are two chair conformations,

115

de la Mare, P. B. D. Progr. Stereochem. 1954, 1, 112.

116

Ref. 83, pp 189–194.

.

123

3.5 ELIMINATION REACTIONS

(1R,2r,3S)-2-bromo-1,3-diethylcyclohexane-A and (1R,2r,3S)-2-bromo-1,3-diethylcyclohexane-B, but (1R,2r,3S)-2bromo-1,3-diethylcyclohexane-B is the thermodynamically more stable due to decreased A strain (Section 1.5.2). In (1R,2r,3S)-2-bromo-1,3-diethylcyclohexane-A, the two ethyl groups are axial and both β-hydrogen atoms are equatorial and, therefore, not susceptible to base induced elimination. In (1R,2r,3S)-2-bromo-1,3-diethylcyclohexane-B both β-hydrogen atoms are axial, but the bromine is equatorial, and again E2 elimination is not possible. Treatment of (1R,2r,3S)-2-bromo-1,3-diethylcyclohexane-A with base will not lead to an E2 reaction. The conformational characteristics of a cyclohexane, and possibly other ring systems, must be evaluated carefully because they may have a significant influence on reactivity in an E2 reaction. The leaving group in an E2 reaction can be a chloride, a bromide, or an iodide, and also sulfonate esters [e.g., a mesylate (dOSO2Me), a tosylate (dOSO2p-MedC6H4), or a triflate (dOSO2CF3)]. The previously mentioned base DBN is a weak nucleophile due to steric hindrance about the nitrogen atom117a and can be used to induce an E2 type reaction with primary halides. Even primary halides, which give mostly SN2 reactions with nucleophilic bases, usually give an alkene in good yield with this base. Methanesulfonates participate in E2 reactions, as in a synthesis of syringolin A by Ichikawa and coworkers,118 in which initial conversion of the alcohol unit to yield mesylate 55 was followed by reaction with DBN to yield 56 in >64% overall yield. In a synthetic example taken from Alvarez-Manzaneda’s et al.119 synthesis of (+)-austrodoral, the related base DBU reacted with acetate 57 to give conjugated ketone 58 in 95% yield. Note that it does not remove Hb with elimination of acetate (an E2 process) nor Hc to generate the ester enolate (Section 9.2). The nitrogen atom in the base DBN is sterically hindered, and it is a poor nucleophile. This poor nucleophilicity means that the normal nucleophilic acyl addition to a carbonyl group in an ester or a ketone is very slow, and not competitive with the elimination process. These two nonnucleophilic bases give moderate-to-good yields of alkene in E2 reactions with alkyl halides, and little or no substitution products.

O DMB HN

N H

N

O

CO2t-Bu N H

DMB DBN , CH2Cl2 , rt , 2 d

HN

N H

N

O O MsO

CO2t-Bu N H

O O

O

O

NH

NH

55

56 ( >64%) O

OAc

O Reflux , 12 h

DBU , Benzene

57

58 (95%)

The most common bases used in E2 reactions are hydroxide (in H2O) or alkoxides in an alcohol solvent (sodium methoxide in methanol, sodium ethoxide in ethanol, potassium tert-butoxide in tert-butanol).110 Amide bases (e.g., sodium amide or lithium diethylamide) can be used in ammonia or amine solvents. In these latter cases, the bases are also good nucleophiles, but in reactions with tertiary halides the elimination process dominates. With secondary halides, substitution can compete and with primary halides, SN2 may be the major reaction, but this depends on the nature of the base-nucleophile.

For synthetic examples, see (a) Achab, S.; Das, B. C. J. Chem. Soc. Chem. Commun. 1983, 391. (b) Oediger, H.; M€ oller, Fr. Angew. Chem. Int. Ed. Engl. 1967, 6, 76.

117

118

Chiba, T.; Hosono, H.; Nakagawa, K.; Asaka, M.; Takeda, H.; Matsuda, A.; Ichikawa, S. Angew. Chem. Int. Ed. 2014, 53, 4836.

119

Alvarez-Manzaneda, E.; Chahboun, R.; Barranco, I.; Cabrera, E.; Alvarez, E.; Lara, A.; Alvarez-Manzaneda, R.; Hmamouchi, M.; Es-Samti, H. Tetrahedron 2007, 63, 11943.

.

124

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

Specialized amine bases are commonly used, including DBN, DBU, and sterically hindered amines (e.g., 2,2,6,6tetramethylpiperidine). The weaker base imidazole is important for use in elimination reactions that involve sensitive substrates. Note that sulfonate esters are good leaving groups, commonly used in elimination reactions. It is clear that a variety of substrates with different leaving groups can be converted to alkenes via elimination. β-Elimination of sulfones is also known.120 A general transform for the E2 reaction follow: R

R

R

R

H R

R

Where X = halogen, OH, OR, OSO2R

X R

R

Vicinal dibromides can be treated with base in such a way that both Br atoms are lost in a stepwise manner to generate an alkene. Many different reagents have been used for this transformation, including potassium tert-butoxide.121 Recently, indium metal was shown to be an effective reagent.122 The E2 reaction can be applied to the synthesis of alkynes. When a vinyl halide is treated with a strong base, loss of HX by what is essentially an E2 process leads to formation of a triple bond. In a simple example, taken from Mori and coworker’s123 synthesis of the sphingosine derivative sulfobacin B, 12-methyltridec-1-ene was treated with bromine in dichloromethane to yield dibromide 1,2-dibromo-12-methyltridecane. When this dibromide was treated with potassium tert-butoxide in petroleum (pet) ether, in the presence of 18-crown-6, initial elimination gave vinyl bromide 2-bromo-12-methyltridec-1-ene. Subsequent reaction with the base initiated an E2 reaction that gave alkyne 12-methyltridec-1-yne in 72% yield for both chemical steps. Note that this method of alkyne formation from vicinal bromides can lead to isomerization of the alkyne unit to yield internal alkynes rather than terminal alkynes via formation of a transient allene product. Higher reaction temperatures and a large excess of base should be avoided. Br (CH 2)7

CHMe2

Br2

(CH 2)7

CH2Cl2

CHMe2

Br 12-Methyltridec-1-ene

Me3CO –K+ 18-Crown-6 Pet ether

1,2-Dibromo-12-methyltridecane Br

Me3CO –K+ (CH 2)7

(CH 2)7

CHMe2

2-Bromo-12-methyltridec-1-ene

CHMe2

12-Methyltridec-1-yne (72%)

An alternative method for preparing vinyl bromides can be incorporated into a synthesis of alkynes when combined with an olefination reaction (see Section 12.5.1). In the Lee et al.124 synthesis of (+)-violapyrone C, aldehyde 59 was treated with carbon tetrabromide and triphenylphosphine to yield vinyl dibromide 60 in >65%. Subsequent treatment with the strong base butyllithium (Section 12.5.1) led to elimination and an alkyne anion, and workup gave the alkyne 61 in 96% yield. This sequence is commonly known as the Corey-Fuchs procedure).125 Note that this sequence is compatible with the presence of reactive functionality (e.g., a tosylate) leaving group.

120

For a review of desulfonylation reactions, see Nájera, C.; Yus, M. Tetrahedron 1999, 55, 10547.

(a) Butcher, T. S.; Detty, M. R. J. Org. Chem. 1998, 63, 177. (b) Malanga, C.; Mannuccki, S.; Lardicci, L. Tetrahedron 1998, 54, 1021. (c) Khurana, J. M.; Maikap, G. C. J. Org. Chem. 1991, 56, 2582. (d) Savoia, E.; Tagliavini, C.; Trombini, C.; Umani-Ronchi, A. J. Org. Chem. 1982, 47, 876. (e) Allred, E. L.; Beck, B. R.; Voorhees, K. J. J. Org. Chem. 1974, 39, 1426. 121

122

Ranu, B. C. Eur. J. Org. Chem. 2000, 2347 (see pp 2348–2349).

123

Takikawa, H.; Nozawa, D.; Kayo, A.; Muto, S.-e.; Mori, K. J. Chem. Soc. Perkin Trans. 1 1999, 2467.

124

Lee, J. S.; Shin, J.; Shin, H. J.; Lee, H.-S.; Lee, Y.-J.; Lee, H.-S.; Won, H. Eur. J. Org. Chem. 2014, 21, 4472.

125

Corey, E. J.; Fuchs, P. L. Tetrahedron Lett. 1972, 36, 3769.

.

125

3.5 ELIMINATION REACTIONS

Me

Me

Me Me

CBr4 , PPh3

Me H

1. BuLi , –78°C to rt

H

0 °C to rt

Me

2. H2O

Br Br 60 (>65%)

O 59

61 (96%)

The E2 disconnections to prepare alkynes follow: R R

X

R R

Where X = halogen, OH, OR, OSO2R

R

R

R

CHO

3.5.2 Unimolecular 1,2-Elimination: The E1 Reaction The E2 reaction follows second-order kinetics, and the rate of reaction is proportional to the concentration of both the halide and the base. Under different reaction conditions, usually involving aqueous media, the reaction can follow firstorder kinetics with a slow ionization step to form a carbocation, followed by a rapid acid-base reaction. In most cases (1) water is the solvent, (2) a good leaving group is present, (3) a weak base is used, and (4) molecular rearrangements are often observed. As with the SN1 reaction in Section 3.2.2, the rate of the reaction is given by the expression: Rate ¼ k1 [RX] + k2 [nucleophile or base]. The slow, and rate-determining step is ionization, followed by a very fast removal of the β-hydrogen atom from the carbocation intermediate. Therefore, in this mechanism, there is no dependence on the concentration of the base and the observed overall rate is dependent only on the halide, leading to Rate ¼ k [RX]. A simple example based on the generic mechanism of ionization to a carbocation and removal of a proton involves heating (3R,4S)3-bromo-4-methylhexane in aqueous media, in which the water assists in ionizing the CdBr bond to generate a solvent separated ion pair [(R)-4-methylhexan-3-ylium bromide]. The resulting carbocation and bromide ions are solvated by water. The β-hydrogen atom of the carbocation [Ha in (R)-4-methylhexan-3-ylium bromide] is adjacent to a full positive charge, and polarized to a greater extent than the β-hydrogen in (3R,4S)-3-bromo-4-methylhexane. Note that (R)-4methylhexan-3-ylium bromide may not be completely solvent separated, but is assumed here for simplicity. A base is required to remove Ha, but Ha in the carbocation (R)-4-methylhexan-3-ylium bromide is a stronger acid than Ha in (3R,4S)-3-bromo-4-methylhexane, so a weaker base is required for that deprotonation. Indeed, the water present in the reaction medium will suffice, and the electrons in the CdHa bond migrate toward the positive carbon to form a new π-bond in the alkene product, 3-methylhex-3-ene. Rotation about the CdC bond in (R)-4-methylhexan-3-ylium bromide prior to removal of Ha will lead to a mixture of rotamers, and 3-methylhex-3-ene will be formed as a mixture of (E/Z)-isomers. The intermediacy of the planar carbocation allows facile attack at both faces of the carbocation, which leads to significant racemization at that carbon if the halide was enantiopure. In general, the thermodynamically more stable product will predominate (called Saytzeff elimination; Section 3.5.1). The reaction is unimolecular in mechanism, and a nucleophile substitutes for the halide leaving group, so it is termed an E1 reaction (unimolecular elimination).126,127 The SN1 reaction usually competes with the E1 process in aqueous media since most bases used in this reaction are nucleophilic. Only when the gegenion is a very weak nucleophile will the E1 process dominate and produce substantial amounts of the alkene product. An example is formation of an alkene when alcohols react with concentrated sulfuric or perchloric acid, as when 9-methyl-9-azabicyclo[3.3.1]nonan-2-ol was treated with concentrated sulfuric acid to yield alkene 9-methyl-9-azabicyclo[3.3.1]non-2-ene.127 Nonoxidizing acids can be used, as in the treatment of 3-hydroxy-1,2-diphenylbutan-1-one with p-toluenesulfonic acid in benzene heated at reflux to give an 86% yield of (E)-1,2-diphenylbut-2-en-1-one, in the Shiina et al.128 synthesis of tamoxifen. It is not always clear if the reaction proceeds by an E1 or an E2 mechanism since the oxonium ion derived from 3-hydroxy-1,2-diphenylbutan-1-one can undergo an E2 reaction in the presence of sodium sulfate.

126

(a) Seib, R. C.; Shiner, V. J., Jr.; Sendijarevi
c, V.; Humski, K. J. Am. Chem. Soc. 1978, 100, 8133. (b) McLennon, D. J. Q. Rev. Chem. Soc. 1967, 21, 490.

127

Willst€ atter, R.; Waser, E. Berichte 1911, 44, 3423.

128

Shiinaa, I.; Sanoa, Y.; Nakata, K.; Suzuki, M.; Yokoyama, T.; Sasaki, A.; Orikasa, T.; Miyamoto, T.; Ikekita, M.; Nagahara, Y.; Hasome, Y. Bioorg. Med. Chem. 2007, 15, 7599.

.

126

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

Br

6

5

Me

OH –

H

– Ha

4

H 2

1

Me

Br

H 2O

3

Ha

Ha Base (R)-4-Methylhexan-3-ylium bromide

(3R,4S)-3-Bromo-4-methylhexane

Me

OH

1. concentration H2SO 4

N

H Et

Me

Et

3-Methylhex-3-ene

Me N

2. Basify

9-Methyl-9-azabicyclo[3.3.1]non-2-ene

9-Methyl-9-azabicyclo[3.3.1]nonan-2-ol

Such mechanistic blurring is common, and additives (e.g., sodium sulfate) are often used to drive the reaction to the elimination product. If a carbocation intermediate is present, rearrangements are possible. Although sometimes useful, the E1 reaction is more commonly observed as a side reaction in SN1 processes, or in E2 reactions in aqueous or alcohol media. If there is a mixture of E1 and E2 in a reaction, there is the possibility of E1 reactions when carbocation intermediates are generated in protic solvents.

3.6 CHARACTERISTICS OF SUBSTITUTION AND ELIMINATION REACTIONS It is apparent from Sections 3.2 and 3.5 that many factors influence the course of nucleophilic reactions, including the nature of the nucleophile, its electron-donating capability, the solvent, the substrate, and the nature of the leaving group. When substitution (Section 3.2.1) and elimination (Section 3.5.1) in an aliphatic substrate are discussed, these reactions compete if the nucleophiles used are also bases (e.g., MeO). It would be very useful to have a list of parameters for such situations that allow one to make predictions. This section will focus on several factors that influence both nucleophilic and elimination reactions. Analysis of these factors lead to key assumptions that allows one to predict the major product in many cases.

3.6.1 The Solvent A solvent should solubilize all reactants and also absorb excess heat that may be liberated by the reaction. A very important property of a solvent is polarity, which largely determines its ability to solvate and separate ions (solvation). Solvation is an important factor in most nucleophilic reactions, and is very important in acid-base reactions, as seen in Section 2.2. A good measure of the ability to separate ions is the dielectric constant.4,129 Common solvents may be organized into two categories: polar or nonpolar, and then protic or aprotic.3 Normally, a polar solvent is one with a substantial dipole and a nonpolar solvent tends to have a small dipole, or none at all. Another measure of polarity is the dielectric constant, which is the ability of a solvent to conduct charge and, in effect, to solvate and separate ions. For substitution reactions, a high dielectric constant is associated with a polar solvent and a low dielectric is associated with a less polar solvent. A protic solvent is one that contains an acidic hydrogen (OdH, NdH, SdH, essentially a weak Brønsted-Lowry acid), whereas an aprotic solvent does not contain an acidic hydrogen. If a solvent is protic, it will have a polarized bond to a hydrogen atom. However, there are differences in polarity that are important to the progress of different reactions. Water has a dielectric constant of 78.5, for example, compared to 24.55 for ethanol, 32.7 for methanol, 16.9 for ammonia, and 6.15 for acetic acid.130 Clearly, water is the more polar based on the criterion of ion separation. Common aprotic solvents include carbon tetrachloride with a dielectric constant of 2.24, diethyl ether (4.34), THF (7.58), acetone (20.7), acetonitrile (37.5), DMF (36.71), and DMSO (46.68).3 This data shows that even the polar DMSO is significantly less efficient at ion separation when compared to water. Both the oxygen and hydrogen atom of the polarized OH unit in the protic solvents water or alcohols solvate cations via the δO and anions via the δ+H via hydrogen bonding. Therefore, both cations and anions are solvated, which leads 129

(a) Parker, A. J. Q. Rev. Chem. Soc. 1962, 16, 163. (b) Parker, A. J. Chem. Rev. 1969, 69, 1.

(a) Arnett, E. M.; Wu, C. Y. J. Am. Chem. Soc. 1960, 82, 4999. (b) Arnett, E. M.; Wu, C. Y. Ibid. 1962, 84, 1680, 1684. (c) Deno, N. C.; Turner, J. O. J. Org. Chem. 1966, 31, 1969.

130

.

3.6 CHARACTERISTICS OF SUBSTITUTION AND ELIMINATION REACTIONS

127

to ionization and solvation. This ability contrasts with the negative dipole of aprotic solvents that can solvate cations, but the positive dipole is sterically hindered and anion solvation is poor. If the solvent is water or contains water, it is an excellent solvent for the solvation and separation of ions, so unimolecular processes, which involve ionization to carbocations (see Section 3.2.2), may be competitive with bimolecular reactions. If water is not the solvent or present as a cosolvent, ionization is much slower. Ionization certainly occurs in alcohols or solvents (e.g., acetic acid), but the process is so slow that other reactions usually occur faster. It is useful to assume that ionization (unimolecular reactions) will be competitive in water, but not in other solvents, leading to the assumption that bimolecular reactions should dominate in solvents other than water. This statement is clearly an assumption, and it is not entirely correct since ionization can occur in ethanol, acetic acid, and so on, but the assumption is remarkably accurate in many simple reactions and allows one to begin making predictions about nucleophilic reactions. In those cases where the assumption fails and the products are different from those predicted, the analysis will lead to a better understanding of the overall process. The solvation ability of water is based on its ability to surround the nucleophilic and electrophilic centers (by hydrogen bonding to the electrophilic H or the nucleophilic O: HOdH⋯X and H2O⋯X+), making collision more difficult in a bimolecular process. For ionization to compete with bimolecular processes, the counterion should be a weak nucleophile and/or a weak base in most cases. If the counterion is too basic, it can induce elimination that can be faster than ionization. If the counterion is too nucleophilic, bimolecular substitution processes will be faster. Coordination of the positive and negative dipoles with C and X “pulls” the CdX bond, increasing the bond polarity, and allowing greater solvation by the water. Solvation of the polarized atoms eventually leads to separation of the developing charges, which accelerates the ionization process and stabilizes the ion products. The net result is ionization of the CdX bond into a carbocation and the X anion, both solvated by water. In an aprotic solvent (e.g., diethyl ether), solvation of the cation is possible, but the anion is not solvated very well. In practical terms, the solvent does not help to pull the leaving group away in the rate-determining step, so ionization is much slower and generally not competitive with bimolecular processes. Even in protic solvents (e.g., ethanol), the dielectric is so small that charge separation is inefficient, again slowing the rate-determining ionization step in a unimolecular reaction. In nucleophilic substitution reactions of alkyl halides, a pentacoordinate transition state is formed at the midpoint of this bimolecular process. No intermediate has ever been observed. If water is the solvent, separating charges will hinder formation of the transition state, which will slow both bimolecular substitution and elimination reactions. A large number of nucleophiles and bases are ionic, and they will be solvated by water. A solvated nucleophile must penetrate its own solvent sheath, as well as any solvation sphere associated with the electrophile to reach the carbon for collision in a bimolecular process, which effectively reduces its nucleophilicity. Nucleophilic strength is, therefore, dependent on the solvent. Aprotic solvents solvate cations, but the nucleophile is not well solvated. Therefore, the nucleophile has virtually no solvent sheath and can approach the electrophilic center with greater ease, effectively increasing the nucleophilicity. Using an aprotic solvent therefore allows the maximum rate possible for the bimolecular process. In all cases, the more polar the solvent, protic or aprotic, the more pronounced will be the solvation effect.

3.6.2 The Nucleophile-Base In a bimolecular process, the stronger the nucleophile, the greater the attraction for δ+C, which is favored in aprotic solvents. When the nucleophile is also a base, the rate of elimination will increase relative to substitution as base strength increases. In general, high nucleophilicity and/or high basicity will lead to a competition between SN2 and E2. If the attacking species is a strong nucleophile, but a weak base, the reaction will proceed almost entirely by SN2 when primary and secondary substrates are involved, but tertiary substrates show elimination or no reaction at all depending on the strength of the base. With low-nucleophilicity, high-basicity species, E2 dominates for tertiary and secondary substrates. Even primary halides can undergo elimination with a nonnucleophilic base (e.g., DBN or DBU). For unimolecular reactions, unimolecular substitution (SN1) is favored with strong, nonbasic nucleophiles in aqueous media. Unimolecular elimination (E1) processes are favored with bases of weak-to-moderate strength that are also poor nucleophiles. Such bases include the hydrogen sulfate and the perchlorate anion, which are resonance stabilized and very poor nucleophiles. In general, the SN1 reaction is faster than the E1 reaction since the cationic carbon is most strongly attracted to the electron-rich center, but this assumption is complicated by the fact that the SN1 product is often unstable and/or difficult to isolate. The hydrogen sulfate or perchlorate SN1 products are examples of unstable substitution products, and in such a case, the E1 process may lead to the major product.

.

128

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

Solvent plays a role in the competition between SN2 and E2 reactions (see Section 3.6.3). In general, protic solvents (e.g., ethanol or methanol) favor elimination reactions over substitution. Substitution reactions are favored over elimination reactions in aprotic solvents (e.g., diethyl ether or THF), however.

3.6.3 The Substrate In general, bimolecular processes are dependent on the nature of the substrate (the molecule containing the electrophilic center). Bimolecular processes generally dominate reactions of primary alkyl halides. Substitution reactions tend to be faster than elimination for primary halides, even in protic solvents. Exceptions occur when a very hindered base, or nucleophile, (e.g., DBN or DBU) is used (see Section 3.5.1), since approach of those reagents to the carbon bearing the leaving group is difficult. In reactions of tertiary halides, unimolecular processes dominate in protic solvents (especially water and aqueous solvents), where substitution is usually faster than elimination. A carbocation is formed in the ionization process and substitution involves nucleophilic attack directly at a positive carbon, whereas elimination involves removal of an acidic hydrogen two bonds removed from the positive carbon. Clearly, a nucleophilic species is strongly attracted to the most positive center and that should lead to the major product. In aprotic solvents, tertiary halides do not undergo SN2 reactions and E2 reactions occur only when a suitable base is present. For secondary halides in aqueous solvents, unimolecular and bimolecular processes compete, and the result is often a mixture of products. In polar aprotic solvents, bimolecular processes are usually faster. If a strong base is present and a protic solvent is used, bimolecular elimination is usually preferred to bimolecular substitution, but this is another assumption. In general, elimination is preferred in protic solvents and substitution in aprotic solvents, for secondary halides and sulfonate esters. In reactions of NaOEt in ethanol the rate of an E2 reaction is 42 times faster for a tertiary bromide when compared to a primary bromide.6 Under the same conditions, the SN2 reaction is >1700 times faster for the primary when compared to the tertiary bromide. For a secondary bromide, under these conditions, the E2 reaction is four times faster.

3.6.4 The Leaving Group A leaving group (X) is the atom or group of atoms displaced in a substitution or elimination reaction. A good leaving group should have a relatively weak and polarized CdX bond. After departure, X should be a very stable ion or a stable neutral molecule (e.g., water or ammonia). If this ion or molecule can be effectively solvated, its leaving group ability is enhanced. In general, gegenions derived from strong acids are good leaving groups and those derived from weak acids are poor leaving groups. Typical good leaving groups that generate ions after their departure are Br, I, OTs, OMs (Ms ¼ methane sulfonge), and Cl. The typical poor leaving groups are OH, NH2, and acetate, which is significantly better than hydroxyl or amino. Both OH and NH2 are poor leaving groups, but alcohols can be protonated to generate oxonium ions, where water is a good leaving group. Similarly, amines can be protonated to form an ammonium ion and is a good leaving group. Halides are probably the most common leaving groups. A general order of leaving group ability for halides is I > Br > Cl ≫ F. Large atoms (e.g., iodide) form a CdI unit, which has a longer bond than CdF, so one expects iodide to be a better leaving group than fluoride. The longer bond is generally weaker due to less electron density between the nuclei, and easier to break (it is actually displaced by the nucleophile). After the halide departs, the larger iodide is better solvated by the solvent, diminishing the net charge of the ion and making it more stable, which assists the overall reaction. The less basic the group, the more easily it departs. In the reaction of haloethanes with ethoxide, iodide “leaves” >1200 times faster than chloride and 1.45 times faster than bromide.131 Bromine “leaves” >830 times faster than chloride and tosylate leaves 1.24 times better than iodide, and 1500 time faster than chloride.3 Tosylate [dOSO2dC6H2(4-CH3)] and mesylate (dOSO2dMe) groups are derived from sulfonic acids. They are good leaving groups. These resonance-stabilized anions (RSO 3 ) have a diffuse charge and are poor nucleophiles. Indeed, sulfonate esters possess several qualities of a good leaving group: (1) resonance stabilization of the departed ion, (2) relief of steric strain upon leaving due to the bulk of the group, and (3) the relative weakness of the CdO (sulfonate) bond. 131

Ref. 3, p 374.

.

129

3.7 SYN-ELIMINATION REACTIONS

In general, good leaving groups are large, of low nucleophilicity, more stable after departure, and of low electronegativity. High polarizability with large atoms (e.g., iodide or sulfur) is usually associated with weaker bonds and better leaving group ability. Poor leaving groups are small, very nucleophilic, and highly electronegative.

3.7 SYN-ELIMINATION REACTIONS Both E2 and E1 reactions require a base and are intermolecular processes. For E2 reactions, the hydrogen atom removed by the base must have an anti relationship to the leaving group. Another type of elimination process is possible if the basic atom is part of, or tethered to the substrate, but the β-hydrogen atom can only be removed if it eclipses the basic atom in a syn-conformation. If the tethered base is also a leaving group, removal of the syn β-hydrogen atom is possible via an intramolecular process. This elimination process requires higher reaction temperatures since the requisite synconformation is higher in energy (Section 1.5.1). Such reactions are termed syn-elimination.21 There are several examples of this reaction involving different tethered bases that can also serve as leaving groups, and several are named reactions.

3.7.1 Hofmann Elimination In a syn-elimination, experimental results show that the less stable alkene is formed, and that the reaction is under kinetic rather than thermodynamic control. The thermodynamically controlled E2 reactions require an anti transition state. The kinetically controlled syn-elimination, however, requires a syn transition state in which the β-hydrogen and the leaving group assume an eclipsed conformation. When there is more than one β-hydrogen atom, the lowest energy syneclipsed rotamer will lead to the major product. An example of this process is the Hofmann elimination,21,132 illustrated by conversion of 2-bromo-3-methylbutane to 3-methylbut-1-ene. The bromide is first converted to an ammonium bromide [(3-methylbutan-2-yl)trimethylammonium bromide] by an SN2 reaction with trimethylamine. This ammonium salt does not possess a basic atom or group (the bromide counterion is a weak base) so elimination is not facile. The bromide moiety must be replaced with a stronger base (e.g., OH), before elimination can occur. Reaction with silver oxide (Ag2O) in the presence of 1 equiv of water (water cannot be the solvent or a cosolvent) yields the corresponding trimethylalkylammonium hydroxide [(3-methylbutan-2-yl)trimethylammonium hydroxide]. Hb

Hb NMe3

Ha

Ha NMe3+ Br–

Br 2-Bromo-3-methylbutane

(3-Methylbutan-2-yl)trimethylammonium bromide Ag2O H 2O

Hb Heat

Ha NMe3

+

OH –

(3-Methylbutan-2-yl)trimethylammonium hydroxide

Hb

–Ha

3-Methylbut-1-ene

If water is present in excess, the ionic base would be solvent separated from the ionic ammonium moiety, and an E2 reaction would result. In the absence of an ion separating solvent, however, the basic hydroxide is tethered to the ammonium ion by ionic bonding. The hydroxide ion can react with the β-hydrogen atom only when that CdH bond eclipses the C  NMe3 +  OH bond. When hydroxide reacts with this hydrogen, water is formed along with a π-bond via expulsion of trimethylamine. In some cases, there is evidence that a nitrogen ylid intermediate is generated (see Section 12.5.3).133 This alternative analysis of the Hofmann elimination puts the focus on a syn-conformation, but intramolecular deprotonation occurs via an N-ylid (see Section 12.5.3). In most cases, direct removal of a hydrogen atom by hydroxide accounts for the product.134 132

DePuy, C. H.; King, R. W. Chem. Rev. 1960, 60, 431.

(a) Cope, A. C; Mehta, A. S. J. Am. Chem. Soc. 1963, 85, 1949. (b) Baldwin, M. A.; Banthorpe, D. V.; Loudon, A. G.; Waller, F. D. J. Chem. Soc. B 1967, 509. (c) Ingold, C. K. Proc. Chem. Soc. 1962, 265. (d) Banthorpe, D. V.; Hughes, E. D.; Ingold, C. K. J. Chem. Soc. 1960, 4054.

133

(a) Julian, P. L.; Meyer, E. W.; Printy, H. C. J. Am. Chem. Soc. 1948, 70, 887. (b) Banwell, M. G.; Austin, K. A. B.; Willis, A. C. Tetrahedron 2007, 63, 6388.

134

.

130

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

It is possible to use the standard E2 model to evaluate the selectivity of the Hofmann elimination, when protic solvents (e.g., water or ethanol) are employed. In such a case, solvent separated ammonium and hydroxide ions are expected with an anti transition state for removal of a β-hydrogen atom. First, note that in 2-bromo-3-methylbutane there are two acidic β-hydrogen atoms, Ha and Hb. The anti conformations for removal of Ha and also Hb in 2-bromo-3-methylbutane are considered. The anti conformation 62A leads to the less substituted alkene 3-methylbut-1-ene via removal of Ha. The anti conformation 63A leads to the more substituted alkene 2-methylbut-2-ene via removal of Hb. In 63A there is a gauche interaction (see circled groups) that raises the energy of the resulting transition state relative to that from 62A, where this interaction is less. Conformation 62A is lower in energy and its transition state yields the major product, the less substituted alkene pent-1-ene.

H

H

H H

Hb

Ha

OH Hb NMe3

OH

NMe3 Ha

NMe3

NMe3

•

CH2Ha –OH

HO–

62A

Hb

Me2HC

63A

H

• H H

Me

H

62B

Me Me 63B

If the solvent is aprotic or if there is no solvent at all (neat), one expects the ammonium hydroxide moiety will exist as a tight ion pair (no solvent separation). The only facile mechanism for removal of a β-hydrogen is an intramolecular process that demands a syn transition state. In 2-bromo-3-methylbutane, removal of either hydrogen atom via an intramolecular process requires a syn-conformation, so 62B and 63B are the two important eclipsed conformations that must be considered. The methyl-methyl and methyl-hydrogen interactions in 63B (see () are significantly higher in energy than the hydrogen-hydrogen and isopropyl-hydrogen interactions that arise in 62B (see Section 1.5.1). Removal of Ha from the lower energy 62B will lead to the favored lower energy transition state and the observed major product 3-methylbut-1-ene. A synthetic transformation taken from the Banwell et al. synthesis of (+)-hirsutic acid uses the Hofmann elimination and also illustrates an alternative synthetic approach to this reaction. Reaction of iodomethane with the dimethylamine unit in 64 leads to the trimethylammonium salt 65. The usual Hofmann elimination sequence, with syn-elimination from 65, gave the alkylidene derivative, 66.134b This variation is important since using an amine as a Hofmann precursor is often more convenient than relying on an SN2 reaction of an amine and a halide to produce the requisite ammonium salt (Section 3.2.1). H

H H

MeO2C

MeO2C

MeO2C

12 MeI , Et2O

H

O

CH2Cl2 18°C , 12 h

H

NMe2

O

Basic alumina CH2Cl2 , 18°C 30 min

H

O

NMe3 I

64

65

66

An important conclusion from the preceding discussion is that solvent effects are quite important in this reaction. A solvent that promotes complete solvent separation of the ammonium and hydroxide ion will favor E2 and anti elimination, whereas a solvent that favors a tight ion pair will favor syn elimination. In water, one anticipates extensive ion separation, but in ethanol or methanol, solvent separation is less efficient and the elimination may proceed by more than one mechanistic pathway. Vital to the success of the Hofmann elimination is the incorporation of an amine that does not possess β-hydrogen atoms (e.g., trimethylamine or triphenylamine). Triethylamine has β-hydrogen atoms on all three ethyl units and reaction with 3-iodo-2-methylpentane will yield ammonium salt 67 where all four alkyl substituents attached to nitrogen have a β-hydrogen that can be removed by a base. Elimination via the most energetically favorable conformation will lead to the smallest and least substituted alkene, ethylene (from the ethyl units).133a,b After loss of ethylene, the other product is the tertiary amine, N,N-diethyl-2-methylpentan-3-amine. There are variations of this reaction that are synthetically useful. Ammonium bromide 68, for example, gave a quantitative yield of 69 when heated with potassium tert-butoxide, in a synthesis of haouamine A by Baran and coworkers.135 135

Burns, N. Z.; Jessing, M.; Baran, P. S. Tetrahedron 2009, 65, 6600. .

131

3.7 SYN-ELIMINATION REACTIONS

Heat

N

– CH 2=CH2

N

OH

N, N-Diethyl-2-methylpentan-3-amine

67 OMe

OMe

OMe

OMe

KOt-Bu , 80°C

MeO

MeO N

N

Br– OMe

OMe

MeO

MeO 68

69 (quant)

Syn-elimination can be extended to functional groups other than those required for Hofmann elimination, but a common feature is the presence of a basic atom connected to the molecule that is also a leaving group. The basic end of the molecule in the new system should be negatively charged or at least negatively polarized. As the dipole moment of the basic end of the molecule increases, the attraction for the δ+ proton should increase, leading to greater facility of the syn-elimination via transfer of hydrogen to the basic atom. For these new systems, the combination of a good base and a good leaving group in the molecule should lead to lower reaction temperatures. If one can build a leaving group into the molecule that is unstable to the reaction conditions as it departs, fragmentation to other products should drive the reaction to completion under milder conditions. A number of syn-elimination reactions commonly used in synthesis exhibit the structural features just presented.

3.7.2 Amine Oxide Pyrolysis (Cope Elimination) Using the Hofmann elimination as a prototype, modification of the leaving group can facilitate syn-elimination. Specifically, the hydroxide ion (the base) and the amine unit (the leaving group) pair can be replaced with a basic atom or group that is part of the leaving group. Cope et al.136 found that incorporation of an amine N-oxide (R3NdO) satisfied this new criterion (see N,N-dimethyl-3-phenylbutan-1-amine oxide). The oxygen atom is a negative dipole that reacts as a base, and removal of the β-hydrogen via a syn conformation leads to cleavage of the CdN bond and loss of the resulting neutral leaving group, N,N-dimethylhydroxylamine. The base is part of the molecule and not a separate molecule, so E2 elimination is not competitive. However, the syn-elimination reaction generally requires higher reaction temperatures relative to an E2 process. Heating amine oxides to produce alkenes is known as Cope elimination,21d,132,136 and a typical reaction temperature is 120°C, significantly lower than that required for the Hofmann elimination, as shown for the conversion of N,N-dimethyl-3-phenylbutan-1-amine oxide to but-3-en-2-ylbenzene.21d A Cope elimination requires the synthesis of N-oxides, and these are usually formed by oxidation of the corresponding amine with hydrogen peroxide or m-chloroperoxybenzoic acid (mcpba) (Section 6.9.2).137 In a synthesis of virosaine A, Gademann and coworkers138 oxidized amine 70 to N-oxide 71 with meta-chloroperoxybenzoic acid (TBDPS ¼ tert-butyldiphenylsilyl). In the presence of silica gel, and without heating, Cope elimination occurred to yield 72 in 77% yield. Note that alkenyl-N-methylhydroxylamines can undergo a reverse Cope elimination to yield pyrrolidine or piperidine N-oxides.139 This elimination reaction is not limited to dimethylamines, but the temperatures required for elimination will vary with the structure of the amine oxide.

(a) Cope, A. C.; Foster, T. T.; Towle, P. H. J. Am. Chem. Soc. 1949, 71, 3929. (b) Cope, A. C.; Pike, R. A.; Spencer, C. F. Ibid. 1953, 75, 3212. (c) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-19.

136

137

(a) Cope, A. C.; Ciganek, E. Org. Syn. Coll. Vol. 4 1963, 612. (b) Craig, J. C.; Purushothaman, K. K. J. Org. Chem. 1970, 35, 1721.

138

Miyatake-Ondozabal, H.; Bannwart, L. M.; Gademann, K. Chem. Commun. 2013, 49, 1921.

139

Ciganek, E.; Read, J. M., Jr.; Calabrese, J. C. J. Org. Chem. 1995, 60, 5795. .

132

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

Me Ph

N H

Me

Me

120°C z– NMe2OH

Me

O

N, N-Dimethyl-3-phenylbutan-1-amine oxide

But-3-en-2-ylbenzene O

O H

Ph

O

H

mcpba CH2Cl2 rt

OTBDPS

O OTBDPS

N

N

O 70

71 O

O H SiO2 CH2Cl2

HO

O N

N

OTBDPS

HO

HO

TBDPSO 72

3.7.3 Ester Pyrolysis Other functional groups can be incorporated into a molecule that will facilitate syn-elimination via intramolecular removal of a β-hydrogen. A simple example is an acetoxy group (dOCOMe), where the carbonyl oxygen of the ester functions as a tethered base.132,140 The oxygen is a weak base and loss of the acetoxy group leads to acetic acid, which is a poorer leaving group relative to an amine from an ammonium salt or a N,N-dialkylhydroxylamine from an amine oxide. Therefore, syn-elimination of acetates (with loss of acetic acid) requires higher reaction temperatures. On the other hand, acetates are readily available by reaction of an alcohol with acetic anhydride or acetyl chloride and pyridine (or another base). Me O

O Ph

D

H

Ph

– MeCOOH

H H

Ph 450°C

D

Ph

(E)-1-d-1,2-Diphenylethene

(1R,2 R)-1,2-Diphenylethyl-2-d acetate Me O

O

Ph

D H

Ph

H

H

Ph

450°C – MeCOOD

H

Ph (1R,2S)-1,2-Diphenylethyl-2-d acetate

(E)-1,2-Diphenylethene

For (1R,2R)-1,2-diphenylethyl-2-d acetate, heating to 450°C was required for elimination to (E)-1-d-1,2-diphenylethene, and in this alkene the deuterium was retained.140b,141 Similar pyrolysis of the enantiomer [(1R,2S)-1,2-diphenylethyl-2-d acetate)] gave (E)-1,2-diphenylethene in which the deuterium was lost. This finding is completely consistent with elimination of hydrogen or deuterium from the β-carbon by the acetoxy group from the lowest energy syn-rotamer.132,140c 140

(a) Alexander, E. R.; Mudrak, A. J. Am. Chem. Soc. 1950, 72, 1810. (b) Idem Ibid. 1950, 72, 3194. (c) Curtin, C. Y.; Kellom, D. B. Ibid. 1953, 75, 6011.

141

Jung, M. E.; Lyster, M. A. J. Org. Chem. 1977, 42, 3761.

.

133

3.7 SYN-ELIMINATION REACTIONS

The importance of the syn-relationship of a β-hydrogen atom and the leaving group is seen in hydrindanes (1R,2R)2-methyl-2,3-dihydro-1H-inden-1-yl acetate and (1S,2R)-2-methyl-2,3-dihydro-1H-inden-1-yl acetate. Both acetates gave 2-methyl-1H-indene upon pyrolysis. However, the syn-derivative [(1R,2R)-2-methyl-2,3-dihydro-1H-inden-1-yl acetate] required temperatures at least 200°C higher than the anti acetate [(1R,2R)-2-methyl-2,3-dihydro-1H-inden-1-yl acetate].140b A synthetic example taken from the Basabe et al.142 synthesis of (+)-lagerstronolide from (+)-sclareol, illustrates two key points with this reaction. When diacetate 73 was heated on silica gel (a variation of the fundamental reaction), a mixture of three alkenes were formed in 90% yield, which were difficult to separate and used in subsequent reactions as a mixture. It is likely that 74 was the favored product and isomerized to the other products under the reaction conditions. Note that such isomerization of the exo-methylene group is common with heating. Note also that the acetate unit at the tertiary carbon eliminated in preference to the acetate at the secondary carbon, which is also typical in this type of syn-elimination. Me

650°C

Me

450°C

Me

OAc

OAc

(1R,2 R)-2-Methyl-2,3-dihydro1 H-inden-1-yl acetate

2-Methyl-1H-indene

(1S,2 R)-2-Methyl-2,3-dihydro1 H-inden-1-yl acetate

OAc

OAc

SiO2 , 100°C

OAc

OAc

+

+

OAc H

H

73

74

H

H

The high temperatures required for acetate pyrolysis greatly limit the synthetic utility of this reaction. Improved leaving group ability and a greater basicity of the tethered base would improve the reaction. Departure of the leaving group to generate a product that was more stable or decomposed after its departure would also facilitate elimination. The sulfur analog of a carbonate is called a xanthate, Od(C]S)dSR, and was found to satisfy the stated goals for improved syn-elimination. The xanthate group is a better leaving group than acetate, but sulfur is a poorer base than oxygen, although it is a better nucleophile.18,19,143 Elimination occurred at significantly lower temperatures relative to an acetate (200°C), and the neutral molecule (S)-methyl S-hydrogen carbonodithioate was lost as a byproduct. This product fragmented to COS and methanethiol, making the elimination essentially irreversible. This reaction is known as xanthate ester pyrolysis.132,140a,b,144 S 200°C

H S Me

+

+

O

H H a

S H

S 75

Me

O

(R)-3-(tert-Butyl)cyclohex-1-ene

1-(tert-Butyl)cyclohex-1-ene

(S)-Methyl S-hydrogen carbonodithioate

(64%)

In a specific example, the thiocarbonyl group of xanthate (75) fragmented to (R)-3-(tert-butyl)cyclohex-1-ene and 1-(tert-butyl)cyclohex-1-ene, with the less substituted (R)-3-(tert-butyl)cyclohex-1-ene being the major isomer.

142

Basabe, P.; Bodero, O.; Marcos, I. S.; Diez, D.; de Román, M.; Blanco, A.; Urones, J. G. Tetrahedron 2007, 63, 11838.

143

(a) Wells, P. R. Chem. Rev. 1963, 63, 171. (b) Koskikallio, J. Acta Chem. Scand. 1969, 23, 1477, 1490.

144

Nace, H. R. Org. React. 1962, 12, 57.

.

134

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

This reaction is one step of the Chugaev elimination143,145 used for conversion of alcohols to alkenes.143 The xanthate ester was prepared from the corresponding alcohol by reaction with carbon disulfide to yield a thioanion ðROCS2  Þ, followed by reaction with iodomethane, as shown for 75. A synthetic example is taken from a synthesis of ()-blepharocalyxin D by Willis and coworkers,146 in which alcohol 76 was converted to xanthate ester 77. Subsequent heating in xylene at reflux gave 78 in >70% yield. Ph Ph

OTBS

Ph 1. NaH , THF

OH

2. CS2 , MeI

Ph

OTBS

S

O

Ph

NaHCO3 Xylene

SMe

Reflux

Ph Ph

OTBS Ph

Ph 77

76

78 (>70%)

Epoxy-xanthate esters undergo an elimination ring-opening reaction. Epoxy-alcohol 79 was converted to xanthate ester 80 in quantitative yield using standard conditions. The elimination step used triethylsilane (Et3SiH), a radical initiator (e.g., AIBN), and treatment with tetrabutylammonium fluoride (TBAF). Under these conditions, opening of the epoxide ring was accompanied by elimination to yield 81 in 72% yield in Ogasawara and coworker’s147 synthesis of (+)-frontalin. This elimination is related to the Barton-McCombie reaction involving reduction of xanthates under radical conditions (see Section 7.11.7).

O

O

NaH , CS2 MeI , THF

AIBN , EtSiH 85–90°C Then

C9H19 C9H19

MeS

O

OH 79

S

80 (quant)

OH

Evaporation then TBAF, THF

C9H19 81 (72%)

3.7.4 Sulfoxide and Selenoxide Pyrolysis A useful change in the syn-elimination precursor incorporates the polarized oxygen of a sulfoxide.148 If a sulfide is oxidized with peroxide or with sodium meta-periodate (NaIO4),149 the resulting sulfoxide has a highly polarized SdO bond (see Section 6.9.1). The negatively polarized oxygen of sulfoxides (S-oxides) can function as a tethered base with the OdSR moiety serving as a leaving group. If the β-hydrogen and the SdO moiety assume an eclipsed conformation, the basic oxygen removes the hydrogen to generate PhSdOH (benzenesulfenic acid, where sulfur is bonded to a phenyl group). This product is unstable to the reaction conditions and decomposes, facilitating the elimination. An example is taken from a synthesis of paracaseolide A by Kraus and Guney,150 in which the sulfide unit in 82 was oxidized to the corresponding sulfoxide (83) in situ with mcpba (Section 6.9.1) and heating gave 84 in 90% yield. In addition to phenyl sulfoxides, methyl sulfoxides [dS(O)Me] can also be prepared, and they undergo synelimination when heated, but the temperature required for pyrolysis is  150°C. For phenyl sulfoxides, elimination is more facile, and heating to only 80°C will usually lead to elimination.

(a) Tschugaeff, L. Berichte 1899, 32, 3332. (b) DePuy, C. H.; King, R. W. Chem. Rev. 1960, 60, 441 (see p 444). (c) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-17.

145

146

Cons, B. D.; Bunt, A. J.; Bailey, C. D.; Willis, C. L. Org. Lett. 2013, 15, 2046.

147

Kanada, R. M.; Tankguchi, T.; Ogawawara, K. Tetrahedron Lett. 2000, 41, 3631.

(a) Grieco, P. A.; Reap, J. J. Tetrahedron Lett. 1974, 1097. (b) Montanari, F. Int. J. Sulfur Chem. C 1971, 6, 137 (Chem. Abstr. 76:24243b 1972). (c) Trost, B. M.; Salzmann, T. N. J. Org. Chem. 1975, 40, 148. (d) Kingsbury, C. A.; Cram, D. J. J. Am. Chem. Soc. 1960, 82, 1810.

148

(a) Grieco, P. A.; Miyashita, M. J. Org. Chem. 1974, 39, 120. (b) Mitchell, R. H. J. Chem. Soc. Chem. Commun. 1974, 990. (c) Sharpless, K. B.; Young, M. W.; Lauer, R. L. J. Am. Chem. Soc. 1973, 95, 2697. (d) Sharpless, K. B.; Young, M. W.; Lauer, R. L. Tetrahedron Lett. 1973, 1979. (e) Jones, D. N.; Mundy, D.; Whitehouse, R. D. J. Chem. Soc. Chem. Commun. 1970, 86. (f ) Reich, H. J.; Renga, J. M.; Reich, I. L. J. Org. Chem. 1974, 39, 2133. (g) Sharpless, K. B.; Young, M. W. Ibid. 1975, 40, 947.

149

150

Guney, T.; Kraus, G. A. Org. Lett. 2013, 15, 613.

.

135

3.7 SYN-ELIMINATION REACTIONS

O

O Ph

O Me

O

S

H O

O

C12 H23

Me

C12 H23

O

O

H O

O

H

Me

C12 H23

Me

H

O O H

rt – 75 °C

C12 H23

Me

S

H

Me

mcpba , 2 h CH2Cl2

O

Ph

O

82

O

C12 H23

H O

83

C12 H23

84 (90%)

Just as sulfides are oxidized to sulfoxides, selenides (RdSedR) can be oxidized to selenoxides.151 Analogous to a sulfoxide, heating a selenoxide leads to thermal syn-elimination to yield the less substituted alkene. The increased polarity of the SedO bond of the selenoxide, relative to the SdO bond of the sulfoxide, and the loss of the unstable RdSedOH leads to even lower temperatures for thermal syn-elimination (typically 0–25°C). Elimination of PhSeOH from 85 gave the exocyclic-methylene derivative (86) as the major product, for example, rather than the endocyclic alkene.152a This elimination reaction occurred at 0°C. Indeed, selenoxide elimination is a particularly attractive method for the generation of sensitive alkenes, such as the less stable exocyclic methylene lactone (86). Note that selenium compounds are toxic and care should be exercised in the handling and disposal of all materials. O

O

0°C

O

O Se

CH2 85

+

Ph–Se–OH

Ph

O

H

86 O

O 7% H2O2 , Py

PhSe

87

88

There are many synthetic examples that use this technique. In a synthesis of cyclobakuchiol A by Kobayashi and coworkers,153 selenide 87 was treated with aq H2O2 in the presence of pyridine to yield the selenoxide, which eliminated PhSeO2H under the reaction conditions to give alkene 88 in good yield. Selenides can undergo another type of elimination reaction, when the selenium unit is attached to a carbon adjacent to a thiocarbamate group as in 89. Radical induced elimination with AIBN in the presence of a hydrogen atom donating species (e.g., tributyltin hydride, see Section 17.5.4) gave the elimination product 90 in 82% yield, which was used in the Comins and Kuethe154 synthesis of (+)-cannabisativine. It is also known that n-alkyl aryl selenides that have an ortho substituent on the aromatic ring are superior reagents for oxidation and syn-elimination when compared to those compounds having a para substituent.155 Differences in structure therefore play a role in the reaction.

151

See, for example, Qi, J.; Roush, W. R. Org. Lett. 2006, 8, 2795.

152

(a) Prelog, V.; Zalán, E. Helv. Chim. Acta 1944, 27, 535. (b) Prelog, V.; H€afliger, O. Ibid. 1950, 33, 2021.

153

Kawashima, H.; Kaneko, Y.; Sakai, M.; Kobayashi, Y. Chem. Eur. J. 2014, 20, 272.

154

Kuethe, J. T.; Comins, D. L. Org. Lett. 2000, 2, 855.

155

Sayama, S.; Onami, T. Tetrahedron Lett. 2000, 41, 5557.

.

136

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

S N

O

N

PhSe CO2Me

N C5H11

CO2Me

N

H

BnO

H

BnO

Bu3SnH , AIBN

C5H11

O O

O O 89

90 (82%)

The retrosynthetic transform for syn-elimination is that shown for the antielimination reactions except that the less substituted alkene is the target: R

R

R

R

X

3.7.5 Burgess Reagent Another reagent, the (carboxysulfamoy1)triethylammonium hydroxide inner-salt methyl ester (91, the Burgess reagent),156 follows a syn-elimination mechanism, and is quite useful for introducing an alkene unit into a substrate. Reaction of an alcohol with 91 leads to 92, and heating initiates the elimination to yield the alkene. The Burgess reagent does not appear to be as selective for syn-elimination as the preceding reagents. As stated directly by Burgess et al.,156 there is a cis stereochemical constraint and Saytzeff elimination is observed in aliphatic cases. Rearrangement is observed in some cases. Nonetheless, the Burgess reagent can be quite valuable in fine organic synthesis. In a synthesis of speciosin G by Macías and coworkers,157 alcohol 93 was heated with 91, in benzene at reflux, to yield an 80% yield of 94. CO2Me O HO R

R

O

H R

R

+

Et3N

S

O N

O

N

O

CO2Me

R R

R 92

OMOM

HNEt3 H

R R

91

HO

S

R R

R

OMOM 91 , Benzene Reflux , 60 h

OMOM

OMOM

93

94 (80%)

3.8 1,3-ELIMINATION (DECARBOXYLATION) Sections 3.5–3.7 focused on 1,2-elimination reactions, where the hydrogen atom was attached to a carbon atom β to the carbon atom bearing a leaving group. The syn-elimination reactions in Section 3.7 showed that intramolecular removal of an acidic hydrogen leads to loss of a leaving group and formation of a π-bond. Elimination is also possible (a) Burgess, E. M.; Penton, H. R., Jr.; Taylor, E. A. J. Org. Chem. 1973, 38, 26. (b) Atkins, G. M., Jr.; Burgess, E. M. J. Am. Chem. Soc. 1968, 90, 4744. (c) Crabb
e, P.; León, C. J. Org. Chem. 1970, 35, 2594. For a chiral Burgess reagent, see (d) Leisch, H.; Saxon, R.; Sullivan, B.; Hudlicky, T. Synlett 2006, 445. (e) Sullivan, B.; Gilmet, J.; Leisch, H.; Hudlicky, T. J. Nat. Prod. 2008, 71, 346.

156

157

Guerrero-Vásquez, G. A.; Chinchilla, N.; Molinillo, J. M. G.; Macías, F. A. J. Nat. Prod. 2014, 77, 2029.

.

137

3.8 1,3-ELIMINATION (DECARBOXYLATION)

when an acidic hydrogen atom and a leaving group are separated by three atoms, in what is termed 1,3-elimination, which also involves an intramolecular process. When a carboxyl carbon is β to a π-bond (as in a carbonyl, alkene, or aryl) intramolecular transfer of a proton to the oxygen of the carbonyl (or to the sp2 carbon of alkenes or aryls) is possible via a six-center transition state. Proton transfer requires a syn orientation of the carbonyl oxygen and hydrogen. This thermal process results in cleavage of the bond connecting the carboxyl carbon, loss of carbon dioxide, and formation of a new π-bond. This 1,3-elimination process is called decarboxylation and it is a common reaction of β-keto acids, β-carboxyl esters, and 1,3-diacids (malonic acid derivatives). A simple example is the elimination of carbon dioxide from 3-oxopentanoic acid to yield an enol, but-1-en-2-ol, which tautomerizes to butan-2-one (keto-enol tautomerism; see Sections 2.5.1, 2.5.2, and 13.2). Note that 3-oxopentanoic acid is drawn in a way that emphasizes the transfer of the acidic hydrogen atom to the oxygen atom of the ketone moiety via a six-center transition state. H O

O O

Heat

C

O

+

H

O

O

O 3-Oxopentanoic acid

Butan-2-one

But-1-en-2-ol

O

O OEt OEt

OEt

NaOEt , EtOH Br

1. aq NaOH 2. Neutralize (H +)

OEt O Diethyl (E)-2-(hex-4-en-1-yl)malonate (72%)

O Diethyl malonate O O H O HO

160°C

OH HO

– CO2

(E)-2-(Hex-4-en-1-yl)malonic acid

(E)-Octa-1,6-diene-1,1-diol

O HO (E)-Oct-6-enoic acid (94%)

Decarboxylation begins with an internal acid-base reaction, where the acid is the OdH unit of the carboxylic acid, and the base is the oxygen of the carbonyl β to the acid moiety. Decarboxylation is possible because of the facile loss of a neutral leaving group carbon dioxide. The malonic ester synthesis (Section 13.3.1) generates substituted malonic esters that can be saponified to diacids. In one example, treatment of diethyl malonate with sodium ethoxide and then (E)-6bromohex-2-ene gave a 72% yield of diethyl (E)-2-(hex-4-en-1-yl)malonate.158 Saponification with aq KOH gave a 99% yield of (E)-2-(hex-4-en-1-yl)malonic acid, which gave a 94% yield of (E)-oct-6-enoic acid when heated to 160°C for 5 h, in Clive and Hisaindee’s158 synthesis of brevioxime. The initially formed product of decarboxylation is an enol, in this case diethyl (E)-2-(hex-4-en-1-yl)malonate, and this enol form of the acid tautomerized to the isolated product, (E)oct-6-enoic acid. Krapcho et al.159 developed a mild decarboxylation procedure of esters. In a synthesis of the antimalarial alkaloid myrioneurinol by Weinreb and Nocker,160 diester 95 was heated with LiCl in aq DMSO, and decarboxylation gave a 90% yield of 96. This mild procedure is referred to as Krapcho decarboxylation.161 A modified procedure has been reported, illustrated by a synthesis of 14-epiamphilectadiene by Shenvi and Pronin,162 in which heating 97 with LiI and 2,6-lutidine gave 98 in 50% yield.

158

Clive, D. J.; Hisaindee, S. J. Org. Chem. 2000, 65, 4923.

(a) Krapcho, A. P.; Weimaster, J. F.; Eldridge, J. M.; Jahngen, E. G. E., Jr.; Lovey, A. J.; Stephens, W. P. J. Org. Chem. 1978, 43, 138. (b) Krapcho, A. P. Synthesis 1982, 805, 893.

159

160

Nocket, A. J.; Weinreb, S. M. Angew. Chem. Int. Ed. 2014, 53, 14162.

161

(a) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-53. (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 380–381. 162

Pronin, S. V.; Shenvi, R. A. J. Am. Chem. Soc. 2012, 134, 19604.

.

138

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

O

O

O

H

O

H LiCl , DMSO/H2O

CO2Me

O N

O

155–160°C

N

Bn 95

96 (90%) Me

Me O

Bn

H

H

O

LiI , 2,6-Lutidine

H

H 150°C

MeO2C

H

O

O 98 (50%)

97

Decarboxylation is also observed with conjugated carboxylic acids (e.g., (E)-2-(but-2-en-1-ylidene)malonic acid]. In conjugated acids, the C]C unit attacks the acidic proton of the COOH unit via a six-centered transition state leading to decarboxylation. The alkene unit can be even further removed from the acid, as in the Smith et al.163 synthesis of zampanolide in which (E)-2-(but-2-en-1-ylidene)malonic acid was heated to 130°C in quinoline for 3 h. The proximal C]C unit functioned as the internal base, and decarboxylation led to a 79% yield of (2Z,4E)-hexa-2,4-dienoic acid. In general, the π-bond of an alkene is a much weaker base than the oxygen of a carbonyl so the hydrogen transfer is less efficient. This fact usually means that higher reaction temperatures are required for the decarboxylation and additives (e.g., quinoline) are often required. H O O

O

H

130°C , 3 h

CO2H

+

Quinoline

C O

CO2H (E)-2-(But-2-en-1-ylidene)malonic acid

(2Z,4E)-Hexa-2,4- (79%) dienoic acid

Decarboxylation of conjugated acids proceeds with bond migration. Such bond migration must be accounted for in any synthetic plan. Another source of π-bonds are aromatic rings. Transfer of the acidic hydrogen of the acid to the π-bond of an aryl carbon is more difficult than with simple alkenes. First, the π-bond is a weaker base due to resonance delocalization and second, accepting the hydrogen requires that the aromatic nature of the benzene ring be disrupted. For both reasons, higher energy is required for the reaction, which means higher reaction temperatures for thermolysis of 99 relative to benzylic systems (e.g., 101). The decarboxylation proceeds by the usual six-center transition state to yield benzylidene intermediate 100, which is unstable to the thermal conditions of the reaction and quickly aromatizes to 101. H O

H H

Heat

O

– CO2

R

R

R 99

100

101

3.9 1,3-ELIMINATION (GROB FRAGMENTATION) Both 1,2-elimination reactions and 1,3-decarboxylation require a reaction with an acidic hydrogen atom, which is accompanied by concomitant bond cleavage that releases a leaving group and forms a π-bond. Other types of elimination reactions are possible if the electron-donating species and leaving group are properly positioned. If the basic 163

Smith, A. B., III.; Safonov, I. G.; Corbett, R. M. J. Am. Chem. Soc. 2002, 124, 11102.

.

139

3.9 1,3-ELIMINATION (GROB FRAGMENTATION)

atom and leaving group are separated by at least three carbons and the bonds that are made and broken assume an antirelationship, elimination is possible. HN N

N

Ag

••

N

••

Br

+ OH– – CH2O

••

102

103

104

105

Prelog and Klyne164 first observed an elimination reaction in work that solved the structure of quinine and other Cinchona alkaloids. This work involved degradation studies that included treatment of 102 with KOH and silver nitrate. Silver ion reacted with the bromide to yield a carbocation (103). A subsequent 1,3-elimination ring-opening reaction, induced by the nitrogen lone pair, as shown, led to iminium salt 104. In the presence of aqueous hydroxide in the reaction medium, addition to the iminium salt led to loss of formaldehyde and formation of the final product, 105. Transfer of the nitrogen lone-electron pair toward the leaving group (or the carbocation center) required the NdC bond and the CdC+ bond to have an anti relationship. This anti relationship is taken from Prelog’s proposed synchronous mechanism for the 1,3-elimination, and can be generalized to say that “both the CαdX bond and the orbital of the nitrogen lone pair be antiperiplanar to the CβdCγ bond.”164,165 Grob et al.166 later expanded this 1,3-elimination to include amino halides and sulfonates. The synchronous nature of the reaction was confirmed by Grob165 and is illustrated by the elimination of the tosylate group in 106 and 107. Elimination of the 5β-isomer (106) gave the (E)-isomer of 1-methyl-2,3,4,7,8,9-hexahydroazecin-1-ium, whereas elimination of the 5α-isomer gave the (Z)-isomer of 1-methyl-2,3,4,7,8,9-hexahydroazecin-1-ium, along with 108. The stereochemistry of the products is best explained if the pertinent orbitals are anti-parallel in the transition state leading to the elimination. The reaction rates for both of these substrates were rather high, but the 5α-isomer (107) reacted much slower and gave a mixture of the substitution product (108) in addition to the elimination product (1-methyl-2,3,4,7,8,9-hexahydroazecin-1-ium). This latter result led to a more general conclusion that the usual SN2 reaction expected with a tosylate and a nucleophile (e.g., hydroxide) will occur unless the rate of the synchronous elimination is fast. Although the reaction was discovered by Prelog, Grob’s contributions to this reaction led to its bearing his name, the Grob fragmentation.167 Me

TsO

N

AgNO3

N

NaOH

Me

H

(1E,5E)-1-Methyl-2,3,4,7,8,9hexahydroazecin-1-ium

106 OH Me H N

AgNO3

+

NaOH

N

OTs

N

Me

Me 107

108

(1Z,5Z)-1-Methyl-2,3,4,7,8,9hexahydroazecin-1-ium

164

Klyne, W.; Prelog, V. Experientia 1960, 16, 521.

165

Grob, C. A. Angew. Chem. Int. Ed. Engl. 1969, 8, 535 and references cited therein.

166

Grob, C. A.; Kiefer, H. R.; Lutz, H. J.; Wilkens, H. J. Helv. Chim. Acta 1967, 50, 416.

167

(a) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-39. (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 288–289.

.

140

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

Both ionization and elimination processes are possible in these systems, as illustrated by generic structure 109. Compound 109 yields the elimination products (iminium ion 110 and the alkene) with a rate constant kf and the ionization product 111 with a rate of ki. This latter pathway leads to substitution. The synchronous fragmentation can occur only if the kf is comparable with or greater than the rate of ionization ki for the γ amino carbon (kf ki).166 If the stereochemical requirements are satisfied, kf is the dominant term. This increase in reactivity due to synchronous fragmentation is referred to as the frangomeric effect, and expressed by f ¼ kf/ki. This effect is not limited to amines or carbocations derived from alkyl halides and silver salts. Suitable derivatives of 1,3-diols also undergo this fragmentation.168 kf N

C

C

C

N

C

+

C

C

+

X–

X 110

ki N

109

C

C

C

111

A synthetic example is taken from the Paquette et al.169 synthesis of jatrophatrione, in which mesyl-alcohol 112 was treated with potassium tert-butoxide to yield the alkoxide (see 113A). Fragmentation of alkoxide 113 (see molecular model 113B, where the hydrogen atoms were removed for clarity) leads to ring opening and formation of an alkeneketone. As shown, the mesylate leaving group is properly oriented relative to the alkoxide for a Grob fragmentation leading to 114, in 98% yield. Me Me HO

Me

t-BuO–

O

Me

H

OCH2Ph

Me

Me H

Me Me

O

Me

H

Me

Me

Me OCH2Ph

Me

Me

MeO2SO

MeO2SO

Me

OCH2Ph

112

113A

114 (98%)

O– –

–OSO

2Me

113B

Another example is taken from the Baran and coworkers170 synthesis of vinigrol in which 115 was treated with potassium hexamethyldisilazide (KHMDS) to yield alkoxide 116. The mesylate leaving group is positioned to allow the Grob fragmentation, which gave 117 in 93% yield. A radical-induced Grob-type fragmentation has also been reported in a synthesis of terpenoids.171 The alkoxide required to initiate a Grob fragmentation can be generated in several ways. Direct reaction of an alcohol with a suitable base to form the alkoxide is one way, as seen in 112. An alternative route is carbanion addition to a carbonyl followed by a Grob fragmentation, as was used in the Magnus and Buddhsukh synthesis of hinesol.172

168

Zimmerman, H. E.; English, J., Jr. J. Am. Chem. Soc. 1954, 76, 2285, 2291, 2294.

169

Paquette, L. A.; Yang, J.; Long, Y. O. J. Am. Chem. Soc. 2002, 124, 6542.

170

Maimone, T. J.; Voica, A.-F.; Baran, P. S. Angew. Chem. Int. Ed. 2008, 47, 3054.

171

Lange, G. L.; Gottardo, C.; Merica, A. J. Org. Chem. 1999, 64, 6738.

172

(a) Buddhsukh, D.; Magnus, P. J. Chem. Soc. Chem. Commun. 1975, 952. (b) Chass, D. A.; Buddhsukh, D.; Magnus, P. D. J. Org. Chem. 1978, 43, 1750.

.

141

3.10 AROMATIC SUBSTITUTION

OMs

OMs KHMDS

H

H

H

H

– OMs

115

116

117 (93%)

O R

O O

R R1 CO2H

O

O

OH

O

H

H

R

R R2

X

R1 R2 OH

CO2H R

R

CO2H CO2H

O O

R

R CO2H

3.10 AROMATIC SUBSTITUTION In many ways, the principles of substitution, elimination, and addition converge in aromatic systems in what is generically called aromatic substitution.173 Addition to electrophilic centers, substitution of carbocations, nucleophilic displacement, and elimination of leaving groups are all mechanistic features of various aromatic substitution reactions.

3.10.1 Defining Aromatic Substitution The substitution reactions discussed in Sections 3.2–3.4 occurred at an sp3 hybridized carbon, and direct substitution at an sp2 carbon in a neutral molecule was not observed. Clearly, substitution occurs via carbocations, which have an sp2 carbon atom, but carbocations are highly reactive intermediates. Alkenes, which contain sp2 hybridized carbon atoms, react with Brønsted-Lowry acids and the C]C unit reacts as a base. In contrast to alkenes (e.g., cyclohexene) and dienes (e.g., cyclohexadiene), which readily react with HCl or HBr, benzene does not react. One must conclude that benzene is a significantly weaker base when compared to alkenes and dienes. To react as a base, benzene must donate the π-electrons from the ring to HCl or HBr and the fact that it does not react demonstrates a special stability. To react similarly to cyclohexadiene, benzene must be “cyclohexatriene,” which would have different bond lengths due to alternating CdC and CdC units. Experimentally, all six CdC bonds in benzene are identical in bond length and strength. Benzene is not “cyclohexatriene,” and there must be a special stability associated with three π-bonds confined to a six-membered ring. In such a system, the π-electrons are stabilized by a special type of resonance delocalization that is known as aromaticity. Benzene is an aromatic hydrocarbon that does not react with reagents that readily react with alkenes or dienes. Substitution reactions are possible, but the conditions are much different. If benzene is mixed with CldCl at ambient temperature and pressure, for example, there is no reaction. For a substitution reaction to occur, benzene must react as a Lewis base, donating two electrons to the halogen to form a new covalent bond. However, benzene does not react with Cl2 or Br2 so it is a very weak Lewis base and benzene does not react as a Brønsted-Lowry base when mixed with HCl or HBr. The fact that benzene is resonance stabilized (aromatic) accounts for the fact that the electrons will not be donated, because such a reaction would disrupt the aromatic character of benzene. This constraint imposes a high activation barrier to reaction. The poor reactivity of benzene can be changed if the neutral molecules X2 or HX are replaced with a formal cation (X+). In the presence of a cation, benzene is a strong enough Lewis base to react, forming a new CdX bond. This reaction must disrupt the aromatic π-cloud, so there is an activation barrier that must be overcome. In the reaction of benzene with an electrophilic species (X+), a carbocation intermediate is formed, which is resonance stabilized. The resonance stability of this intermediate facilitates the reaction with the cation X+. The intermediate carbocation formed when benzene reacts with X+ is sometimes referred to as a benzenonium ion,174 and old terminology called it a Wheland intermediate, especially when drawn in the abbreviated form shown.175 The more common and proper term for this carbocation is an arenium ion. This term will be used throughout. Loss of 173

For examples, see Ref. 1, pp 129–134, 619–629, 703–705, 759–761, and 911–912.

174

Stock, M. M. Aromatic Substitution Reactions; Prentice Hall: Englewood Cliffs, NJ, 1968; p 21.

175

Wheland, G. W. J. Am. Chem. Soc. 1942, 64, 900.

.

142

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

a proton from the arenium ion will yield a substituted benzene derivative, in what is essentially an E1 process. This elimination is very rapid because a stable aromatic ring will be regenerated, which is an exothermic process. The E1 analogy is reasonable when it recognized that the proton on the carbon adjacent to the cationic center in an arenium ion is acidic. O’Ferrall and coworkers176 have estimated the pKa of a proton in an arenium ion to be about 24.3. In this reaction, the electrophilic species (X+) replaced an aromatic hydrogen atom via an electrophilic process, and is termed electrophilic aromatic substitution. It is labeled as an SEAr reaction.177 X X+

X

H

X

H

H

Arenium ion X H

X

– H+

"Wheland intermediate"

The benzene ring donates electrons to an electrophilic species in an SEAr reaction, but is it possible for an electronrich species to donate electrons to a benzene ring? Electron donation to an unsubstituted benzene ring has a substantial activation energy barrier because an electron-rich species must react with the electron-rich benzene ring. The reaction of an electron-rich species Y is possible if a good leaving group (X) is attached to the aromatic ring, but reaction conditions are often harsh. The reaction of an aryl halide (PhX) with a nucleophile (Y), for example, would produce a resonance-stabilized carbanion (118) as an intermediate. The activation energy for the formation of 118 is relatively high in energy and typically requires high temperatures, high pressures, and long reaction times. Loss of X from 118 yields the final substitution product, PhY. The overall reaction is called nucleophilic aromatic substitution, labeled as an SNAr reaction. X

X

X

X

Y–

Y

Y

– X–

Y

Y 118

An alternative mechanism for nucleophilic aromatic substitution is known. When an unactivated aryl halide (ArX) is treated with a strong base, removal of the ortho hydrogen and subsequent loss of the halide ion (X) yields a highly reactive intermediate called benzyne, which is the parent compound of a class of intermediates known as arynes. Benzyne will be attacked by nucleophiles in a reaction that opens the π-bond that is not part of the aromatic cloud, to generate a carbanion (119). Protonation completes the sequence to yield the aromatic substitution product, ArdB. The chemistry of benzyne, and arynes in general, is >100-years old,178 and these reactive intermediates remain quite important in a variety of synthetic transformations.179 X

•• B:

B:

B Benzyne

+ H+

B

119

The following sections will discuss the three aromatic substitution reactions SEAr, SNAr, and benzyne reactions. This subject is usually presented in detail in undergraduate textbooks. The treatment varies greatly with the textbook, however, and leading references are usually absent. Nonetheless, the essentials of these reactions are typically discussed in an undergraduate organic chemistry course, so the discussion in this section is intended to be a review.

176

McCormack, A. C.; McDonnell, C. M.; More O'Ferrall, R. A.; O'Donoghue, A. C.; Rao, S. N. J. Am. Chem. Soc. 2002, 124, 8575.

177

Taylor, R. Electrophilic Aromatic Substitution; John Wiley & Sons: Chichester, 1990.

178

Stoermer, R.; Kahlert, B. Ber. Dtsch. Chem. Ges. 1902, 35, 1633.

179

Wenk, H. H.; Winkler, M.; Sander, W. Angew. Chem. Int. Ed. 2003, 42, 502. .

143

3.10 AROMATIC SUBSTITUTION

3.10.2 Electrophilic Aromatic Substitution: The SEAr Reaction As pointed out in Section 3.10.1, benzene is a weak base and normally will not react with an electrophilic species (e.g., HCl). The poor reactivity is further illustrated by mixing benzene with diatomic chlorine. Contrary to the reactions of cyclohexene or cyclohexa-1,3-diene, benzene does not react with chlorine. If benzene is mixed with Cl2 in the presence of ferric bromide (FeBr3) or aluminum chloride (AlCl3), however, there is a reaction and chlorobenzene is formed in good yield, along with HCl. It is known that diatomic chlorine reacts with a Lewis acid (e.g., ferric bromide) to generate an “ate” complex, Cl + FeCl4  . It is reasonable to assume that benzene reacted with the “ate” complex, and specifically with the cationic Cl+ since it does not react with diatomic chlorine. Cl2

No reaction

Benzene Cl

Cl2 , AlCl3

Benzene

Chlorobenzene

In the absence of a Lewis acid, benzene can react to form a weak charge-transfer complex with chlorine (see 120), but mixing benzene and chlorine does not yield a substitution reaction. A strong Lewis acid (e.g., AlCl3) is required for reaction, and it is known that a charge-transfer complex (e.g., 121) is formed between the Lewis acid and chlorine rather than with benzene.180 In 121, the chlorine-chlorine bond is sufficiently weakened that it can be attacked by benzene, transferring one chlorine atom to benzene (with disruption of the aromatic system) and transfer of the other chlorine to AlCl3 (see 122). For all practical purposes, 121 can be treated as Cl+ in its reaction with benzene, and benzene reacts more or less as a Lewis base. The stronger the Lewis acid (see Section 2.3), the more easily the electrophilic species X+ will be generated, and the faster the reaction with benzene. The product is an arenium ion intermediate, which loses a proton to yield chlorobenzene, as described in Section 3.10.1.181 Analysis of the overall conversion of benzene to a substituted benzene derivative shows that a chlorine or bromine atom, or a nitro or sulfonyl group, has replaced one H on the benzene ring. Therefore, this reaction is a substitution (S). The arenium ion mechanism is shown for the reaction of benzene with Cl+ to yield chlorobenzene, so it is an electrophilic (E) reaction, and it clearly involves an aromatic species (benzene or aryl, Ar). This type of reaction is labeled electrophilic aromatic substitution (SEAr) and it accounts for most of the chemistry of benzene and its derivatives. Note that the validity of a charge-transfer complex mechanism for chlorination of alkyl benzenes has been questioned, and it is solvent dependent.181a However, in this book the SEAr mechanism is assumed to be valid for all electrophilic aromatic substitution reactions unless otherwise stated. Cl Cl 120

Cl

HCl

AlCl3–

Cl 121

H

Cl

AlCl3–

122

In an SEAr chlorination reaction, the rate-determining step (the slowest) is conversion of the charge-transfer complex to the arenium ion complex (121 ! 122).177 Formation of 121 and elimination of the proton from the arenium ion are all fast processes. Bromine reacts in a manner identical to chlorine,181b,182 producing a Br+ complex as a reactive intermediate in the charge-transfer complex. Indeed, the reaction of bromine, benzene, and aluminum chloride yields bromobenzene. Typical Lewis acid catalysts for this reaction are iodine, AlCl3, SbCl3, PCl3, PCl5, and SnCl4.183 When benzene is treated with a mixture of nitric and sulfuric acids, nitrobenzene is the product,184 via a nitronium ion 180

Ref. 174 p 27.

(a) de la Mare, P. B. D. Acc. Chem. Res. 1974, 7, 361 and references cited therein. (b) Le Page, L.; Jungers, J. C. Bull. Soc. Chim. Fr. 1960, 525. (c) Olah, G. A.; Kuhn, S. J.; Hardie, B. A. J. Am. Chem. Soc. 1964, 86, 1055.

181

182

Ref. 177, pp 377–380.

183

(a) Price, C. C. J. Am. Chem. Soc. 1936, 58, 2101. (b) Idem Chem. Rev. 1941, 29, 37. (c) Price, C. C.; Arntzen, C. E. J. Am. Chem. Soc. 1938, 60, 2835.

184

(a) Hoggett, J. G.; Moodie, R. B.; Penton, J. R.; Schofield, K. Nitration and Aromatic Reactivity; Cambridge University Press: Cambridge, 1971. (b) Schofield, K. Aromatic Nitration; Cambridge University Press: Cambridge, 1980. .

144

3. FUNCTIONAL GROUP EXCHANGE REACTIONS

(NO+2 ).185 Benzene also reacts with sulfuric acid [or sulfuric acid saturated with sulfur trioxide (SO3), which is known as fuming sulfuric acid] to give benzenesulfonic acid.186 Benzenesulfonic acid is also formed by reaction of benzene with chlorosulfonic (ClSO3H)187 or fluorosulfonic acid (FSO3H).188 Cl+

Cl

Cl

Cl

H

H

H

Benzene

Cl

– H+

Arenium ion

Chlorobenzene

Note that the aromatic stability of benzene must be disrupted in order to form an arenium ion, but this intermediate is resonance stabilized. This intermediate is sufficiently stable that the reaction proceeds first to give the arenium ion and then substitution products. It is important to state that an arenium ion is resonance stabilized, but not aromatic. In principle, benzene will react with any cation, including carbocations. Carbocations are generated by several different methods (see Sections 2.5.1 and 16.3). If benzene reacts with a carbocation, electrophilic aromatic substitution will yield an arenium ion intermediate and the product is an alkyl benzene. Alkyl benzenes are typically called arenes. The reaction of benzene and its derivatives with carbocations is generically called the Friedel-Crafts reaction. FriedelCrafts reactions take two fundamental forms, Friedel-Crafts alkylation and Friedel-Crafts acylation. Friedel-Crafts alkylation requires formation of an alkyl carbocation in the presence of benzene or a benzene derivative. Carbocations can be produced by the reaction of an alkyl halide with a Lewis acid. A simple example illustrates this reaction. When benzene reacted with benzyl chloride in the presence of AlCl3, diphenylmethane was isolated in 59% yield.189 To form a carbocation from benzyl chloride, the chlorine atom reacts as a Lewis base with AlCl3 to form an “ate” complex, PhCH2 + AlCl . Benzene reacts with this carbocation via a SEAr reaction. There are other methods to generate a carbocation that can be used in Friedel-Crafts alkylation reactions. These methods will be examined in Section 16.4. 0:4AlCl3 , 020∘ C PhH + PhCH2 Cl ƒƒƒƒƒƒƒƒƒƒƒƒƒ!

Benzene

Benzyl chloride

PhCH2 Ph Diphenylmethane ð59%Þ

If a stable carbocation (e.g., the benzylic or a tertiary cation) is the intermediate, the reaction is relatively straightforward. If a primary or secondary carbocation is formed, however, rearrangement to a more stable cation may occur, and rearrangement will occur before the reaction with benzene (see Sections 2.5.1 and 3.2.2). When 1-bromopropane reacts with aluminum chloride (AlCl3) in the presence of benzene, the isolated product is isopropylbenzene (cumene), not n-propylbenzene. The initial reaction of 1-bromopropane and aluminum chloride yields a primary carbocation (propan-1-ylium), but to explain formation of isopropylbenzene as the major product, benzene must react with a secondary carbocation propan-2-ylium. The initially formed propan-1-ylium must rearrange to the more stable secondary cation via a 1,2-hydride shift, before it reacts with benzene. This type of 1,2-hydride shift was introduced in Section 2.5.1 and rearrangement is always assumed to occur before reaction of the nucleophile with the carbocation. After the rearrangement, propan-2-ylium reacts with benzene to yield arenium ion 123. Subsequent loss of a proton by the E1 reaction yields the final product, cumene. In other words, this reaction proceeds by the normal SEAr mechanism, except for the rearrangement prior to reaction with the benzene ring. Assume that primary carbocations obtained from primary halides will always rearrange before they react with benzene. Br

H AlCl3

1-Bromopropane

Propan-1-ylium

Propan-2-ylium

123

Cumene

Another type of Friedel-Crafts reaction involves reaction of benzene with an acid chloride to generate a carbocation intermediate that is different from the simple alkyl carbocations discussed in Section 2.5.1. When benzene reacted with

(a) Westheimer, F. H.; Kharasch, M. S. J. Am. Chem. Soc. 1946, 68, 1871. (b) Martinsen, H. Z. Phys. Chem. 1905, 50, 385. (c) Idem Ibid. 1907, 59, 605. (d) Marziano, N. C.; Sampoli, M.; Pinna, F.; Passerini, A. J. Chem. Soc. Perkin Trans. 2 1984, 1163.

185

(a) Brand, J. C. D.; Jarvie, A. W. P.; Horning, W. C. J. Chem Soc. 1959, 3844. (b) Cerfontain, H.; Sixma, F. L. J.; Vollbracht, L. Recl. Trav. Chim. Pays-Bas 1963, 82, 659. (c) Cerfontain, H. Ibid. 1961, 80, 296. (d) Idem Ibid. 1965, 84, 551. 186

187

(a) Harding, L. J. Chem. Soc. 1921, 1261. (b) Levina, L. I.; Patrakova, S. N.; Patruskev, D. A. J. Gen. Chem. USSR 1958, 28, 2427.

188

Gillespie, R. J. Acc. Chem. Res. 1968, 1, 202.

189

Fieser, L. F.; Fieser, M. Advanced Organic Chemistry; Reinhold Publishers: New York, NY, 1961; p 650. .

145

3.10 AROMATIC SUBSTITUTION

butanoyl chloride in the presence of AlCl3, the isolated product was a ketone; butyrophenone (1-phenylbutan-1-one) in 51% yield.190 If benzene and butanoyl chloride were heated without the AlCl3, there was no reaction. The only way to explain this reaction is that there must be a reaction between AlCl3 and the acid chloride prior to reaction with benzene. Formation of 1-phenylbutan-1-one is an aromatic substitution, and based on the knowledge that an arenium ion must be an intermediate. O O

PhH, AlCl3

Cl Butanoyl chloride

1-Phenylbutan-1-one

The chlorine atom in butanoyl chloride reacts as a Lewis base with the aluminum atom of AlCl3 to form an acyl cation (124), which is known as an acylium ion. An acylium ion is resonance stabilized, with the two resonance contributors shown. It is sufficiently stable that it does not rearrange, but it is reactive enough to react with benzene. Benzene reacts with acylium ion 124 via a SEAr reaction, the same as any other cation, to yield arenium ion 125. Loss of a proton leads to the final product, 1-phenylbutan-1-one. The net result of this reaction is electrophilic aromatic substitution of a hydrogen atom on benzene with an acyl group (a carbonyl group), and it is called Friedel-Crafts acylation (see Section 16.4.4).

O

O

AlCl3

H

O

Cl

O

Butanoyl chloride

124

125

3.10.3 The SEAr Reactions of Substituted Benzenes When substituents are present on the aromatic ring, SEAr reactions can proceed with great selectivity. Majetich et al.,191 for example, showed that aniline reacted with HBr/DMSO to yield a 76% yield of 4-bromoaniline, in a remarkably selective reaction. Bromination of 2,4-dihydroxybenzoic acid gave a 75% yield of 5-bromo-2,4-dihydroxybenzoic acid in a synthesis of lateriflorone by Theodorakis and coworkers.192 Note that the highly activated ring in 2,4-dihydroxybenzoic acid was brominated without the use of a Lewis acid. Such reactivity is common when strong activating groups (e.g., OH and OR) are present on the aromatic ring. In a synthesis of 11,12-demethoxypauciflorine, Magnus et al.193 treated 4-formyl-2-methoxyphenyl acetate with fuming nitric acid, and obtained an 82% yield of 4-formyl-2-methoxy-3-nitrophenyl acetate. Br

OH

OH Br2, AcOH 5 h, 258°C

HO2C

HO2C

OH

OH 5-Bromo-2,4-dihydroxybenzoic acid (75%)

2,4-Dihydroxybenzoic acid CHO

CHO Fuming HNO3 HO > NHR > NH2 >

+ d– Z

> O

R

N

> Alkyl

R H (or R)

Faster rate of reaction

Slower rate of reaction Electron-Releasing Groups

ortho, para Products are major BENZENE (reference compound) Halogens (F, Cl, Br,. I) react slower than benzene, but give ortho-para products React slower than benzene d+ X

+

Electron-Withdrawing Ability O


EtO > MeO > HO > > F > HOH > Br > I N3 

4.2.2 Nucleophilic Acyl Addition to Carbonyls In general, addition of carbon nucleophiles (e.g., a Grignard reagent or an alkyne anion) to the carbonyl of an aldehyde or ketone is irreversible. This statement means that the equilibrium greatly favors the product. This result contrasts with the reaction of heteroatom nucleophiles with ketones or aldehydes, which is generally reversible. Oxygen and nitrogen nucleophiles (e.g., alcohols or amines) often require an acid catalyst in order to establish an equilibrium that leads to good yields of acyl addition products. All of these reactions are discussed in any undergraduate organic chemistry book, so this discussion will primarily focus on addition of strong nucleophiles. Nucleophilic addition to a carbonyl3 can be reversible if the attacking nucleophilic species is also a good leaving group in the alkoxide product (leaving groups are discussed in Section 3.6.4). Such nucleophiles are termed weak nucleophiles, and water and hydroxide are examples. When water adds to the acyl carbon of cyclohexanone, the product is an alkoxide, 1-oxoniocyclohexan-1-olate, but H2O is a good leaving group in the presence of the alkoxide unit, making the reaction reversible as shown. In other words, the reaction of water is unfavorable, and water is categorized as a weak nucleophile for that reason. If the reaction of water with cyclohexanone is examined, addition to the carbonyl generates an oxonium ion, 1-oxoniocyclohexan-1-olate. This oxonium ion contains the –OH2+ unit, which is the excellent leaving group water. Loss of water is facile, making the addition of water reversible. The hydrate, cyclohexane-1,1-diol, can be generated from the oxonium ion via an intramolecular acid-base reaction. If the hydrate is formed, it is unstable. Hydrates that have a hydrogen atom on the α-carbon are known to be unstable, losing water to form an enol (cyclohex-1-en-1-ol). Enols are unstable (see Section 13.2.1), and undergo keto-enol tautomerization that regenerates the original ketone or aldehyde. In this case, formation of cyclohexane-1,1-diol loses water to form cyclohex-1-en-1-ol, which tautomerizes to cyclohexanone. It is assumed that all hydrates are unstable, but hydrates derived from aldehyde or ketones that do not have an α-hydrogen atom are stable. Choral (Cl3CCHO), for example, reacts with water to form a stable, and crystalline hydrate known as choral hydrate, Cl3CCH(OH)2. 2

Koskikallio, J. Acta Chem. Scand. 1969, 23, 1477, 1490. Also see Harris, J. C.; Kurz, J. L. J. Am. Chem. Soc. 1970, 92, 349.

3

For examples see Ref. 1, pp 1125–1188.

163

4.2 NUCLEOPHILIC ACYL ADDITION AND SUBSTITUTION

O

O

OH2

Cyclohexanone

HO

OH2

1-Oxoniocyclohexan-1-olate

O

OH

OH

Cyclohexane-1,1-diol

Cyclohex-1-en-1-ol

Cyclohexanone

Unstable

Just as water (HOH) is a weak nucleophile because of the poor and reversible reaction with a carbonyl, ethanol, or other alcohols (ROH) are also weak nucleophiles because acyl addition to cyclohexanone is reversible. The product of a reaction of an aldehyde or ketone and an alcohol is an acetal. Formally, a ketal (the product formed from a ketone) is a subclass of an acetal. In this book, the term acetal will be used for the product derived from an aldehyde and ketal will be used for the product derived from a ketone. One method to improve the reaction is to protonate the acyl unit of cyclohexanone by adding an acid catalyst. Protonation of cyclohexanone, for example, yields the resonance-stabilized oxocarbenium ion 3 (Sections 3.3 and 16.2.1) that reacts with ethanol to yield 4. Loss of a proton regenerates the acid catalysts and forms a hemiacetal, 1-ethoxycyclohexan-1-ol. All steps are reversible, but in the presence of the acid catalyst and ethanol solvent (i.e., a large excess of ethanol), the hemiketal is present in the equilibrium mixture. Protonation of OH in 1-ethoxycyclohexan-1-ol yields (1-ethoxycyclohexyl)oxonium. Loss of water yields oxocarbenium 5, and subsequent addition of a second molecule of ethanol yields the reactive intermediate 6. Loss of a proton yields the diethyl ketal of cyclohexanone (1,1-diethoxycyclohexane). Removal of water from this reaction mixture shifts the equilibrium toward the ketal product. O

O

+H+

H

OH

HO

Et

OH

–H+

OEt

O

HOEt

H

Cyclohexanone

3

1-Ethoxycyclohexan-1-ol

4 Et

H2O

O

OEt

Et

OEt

H

5

(1-Ethoxycyclohexyl)oxonium

O

OEt

HOEt

EtO

OEt

1,1-Diethoxycyclohexane

6

Ketals and acetals are common protecting groups for the carbonyl group (Section 5.3.3.1). When thiols are used in this reaction, dithioketals or dithioacetals are formed. Amines react reversibly with aldehydes or ketones. Primary amines react to yield imines and secondary amines react to yield enamines (Section 13.6). Although acyl addition is usually reversible, Le Chatlier’s principle applies, so addition of excess reagent (e.g., ethanol) or removal of a product (water) will drive the equilibrium from ketone to ketal (or another addition product). Conversely, if the ketal is treated with an acid catalyst in the presence of excess water (e.g., aq acid), the equilibrium will shift to favor the ketone. In other words, the ketal will be hydrolyzed to the ketone (see Section 5.3.3.1). As a practical matter, acyl addition becomes irreversible if the nucleophile generates a product that has a strong covalent bond with a poor leaving group in the product. When Na+C^CMe reacts with a carbonyl, as in the formation of 1, a carbon–carbon bond is formed. A carbon group is an extremely poor leaving group for the reverse reaction, so the acyl addition is essentially irreversible. A specific example is the reaction of cyclohexanone with the sodium salt of prop-1-yne to yield alkoxide 7. A hydrolysis step generates the final alcohol product, 1-(1-propynyl)cyclohexanol. In reactions such as this, the nucleophilic strength of the incoming reagent, as well as the groups attached to the acyl carbon, play a major role. O

O

CLCMe H3O+

NaCLCMe

Cyclohexanone

7

HO

CLCMe

1-(1-Propynyl)cyclohexanol

164

4. ACIDS, BASES, AND FUNCTIONAL GROUP EXCHANGE REACTIONS

Other important carbon nucleophiles include Grignard reagents (RMgX, Section 11.4.3.1) and organolithium reagents (RLi, Section 11.6.4), as well as enolate anions (Section 13.4.1). All of these acyl addition reactions are irreversible and yield an alkoxide product that is hydrolyzed to yield an alcohol. This reaction with a Grignard reagent is known as a Grignard reaction, which is an extremely important reaction in organic synthesis. An example is taken from a synthesis of gabosine J by Vidyasagar and Sureshan,4 in which aldehyde 8 reacted with vinylmagnesium bromide to generate alkoxide 9. Aqueous hydrolysis of the alkoxide, an acid-base reaction, yields the alcohol product, (10), in 82% yield. CHO

O– +MgBr

BnO OBn

BnO

CH2KCHMgBr , THF

aq H+

OBn

BnO

OH BnO OBn

BnO

BnO

BnO

BnO 8

BnO 9

10 (82%)

The reaction of a ketone with a strong base can lead to a CdC bond cleavage reaction. Discovered by Semmeler,5 Haller and Bauer6 developed the reaction illustrated by reaction of the ketone (1-methylcyclohexyl)(phenyl) methanone with NaNH2 to yield an amide, 1-methylcyclohexane-1-carboxamide. Acyl addition of the amide anion to the carbonyl of the ketone led to an alkoxide intermediate, amino(1-methylcyclohexyl)(phenyl)methanolate, and loss of the anion Ph gave the amide product. Treatment with aqueous acid (known as the workup) leads to protonation of the anion to yield benzene. This reaction is known as the Haller-Bauer reaction.7 Me Ph

Me

NH2

Me

Ph

NH2

NH2

Ph

+

PhH

O

O

O

Amino(1-methylcyclohexyl) (phenyl)methanolate

(1-Methylcyclohexyl) (phenyl)methanone

H

+

1-Methylcyclohexane 1-carboxamide

4.2.3 Nucleophilic Acyl Substitution Carboxylic acid derivatives (e.g., acid chlorides) and esters have an interesting structural feature that differentiates them from a ketone or aldehyde. Acid chlorides and esters have a carbonyl that bears a Cl atom or an OR group attached to the carbonyl carbon atom, respectively.8 In a reaction with a nucleophile, these groups lead to another difference when compared to an aldehyde or a ketone. When a nucleophile adds to the carbonyl of carboxylic acid derivatives (e.g., acid chlorides and esters), the resulting alkoxide intermediate, but it is given the special name tetrahedral intermediate because it bears the leaving groups Cl or OR. The CdCl or CdOR bonds are relatively easy to break. Indeed, the chlorine of an acid chloride, or the OR group of an ester, become leaving groups in the tetrahedral intermediate. Loss of Cl or RO, which are rather stable ionic entities, regenerates the C]O unit and the overall result of this process is a substitution reaction in which the nucleophile replaces chlorine on the carbonyl. This reaction with carboxylic acid derivatives (acid chlorides, esters, anhydrides, amides) is therefore called nucleophilic acyl substitution. A generalized example is shown for the reaction of a Nuc with a carboxylic acid derivative (X ¼ Cl, OR, O2CR, NR2) to yield a tetrahedral intermediate, which loses the leaving group (X) to yield a new carbonyl product. O–

O R

O

– X–

Nuc

R

X

X Nuc

R

Nuc

Tetrahedral intermediate

4

Vidyasagar, A.; Sureshan, K. M. Eur. J. Org. Chem. 2014, 2349.

5

Semmler, F. W. Berichte 1906, 39, 2577.

6

Haller, A.; Bauer, E. Compt. Rend. 1908, 147, 824.

(a) Hamlin, K. E.; Weston, A. W. Org. React. 1957, 9, 1. (b) Mehta, G.; Venkateswaran, R. V. Tetrahedron 2000, 56, 1399. (c) Paquette, L. A.; Gilday, J. P. Org. Prep. Proceed. Int. 1990, 22, 167.

7

8

For examples see Ref. 1, pp 1717–1808, 1929–1994.

165

4.2 NUCLEOPHILIC ACYL ADDITION AND SUBSTITUTION

Many types of nucleophiles can be used, including hydroxide, alkoxides, alcohols, or amines, which tend to form the tetrahedral intermediate reversibly, for the same reasons as noted in Section 4.2.2. A simple example is the acyl addition of hydroxide ion to benzoyl chloride, which yields sodium chloro(hydroxy)(phenyl)methanolate as the tetrahedral intermediate, and loss of the Cl ion leads to benzoic acid. This reaction is formally known as hydrolysis of the acid derivatives. Under these basic conditions, the initially formed acid is converted to the carboxylate anion, requiring a neutralization step to recover the acid. The acid or base hydrolysis of esters, acid chlorides, anhydrides, or amides will give the corresponding carboxylic acid. O – Na+

O HO

–

Cl

OH

Benzoyl chloride

O – Cl

O

–

HO

–

O – Na+

OH

Cl

Benzoic acid

Sodium chloro(hydroxy) (phenyl)methanolate

Sodium benzoate

Acyl substitution also forms the basis for conversion of a carboxylic acid to an acid chloride with thionyl chloride, for conversion of an acid chloride. Phosphorus pentachloride is commonly used for the preparation of acid chlorides,9 as with thionyl chloride.10 An acid chloride or an ester can be converted to another ester by reaction with an alcohol, and an acid chloride or an ester can be converted to an amide by reaction with an amine. Other reagents are available to facilitate acyl substitution. A very effective method for the conversion of a carboxylic acid to a methyl ester, for example, is the reaction with diazomethane (CH2N2, see Section 17.9.3). The reactivity of carboxylic acids can be largely predicted by the leaving group ability of the group attached to the carbonyl carbon. Therefore acid chlorides are the most reactive, and acid anhydrides are approximately as reactive. Acid chlorides are used most often, however, so acid anhydrides will not be discussed here. All of the acid derivatives react with aqueous acid or with aqueous hydroxide, followed by neutralization with aqueous acid, to yield the parent carboxylic acid. A carboxylic acid (e.g., propanoic acid) reacts with thionyl chloride to yield propanoyl chloride, and reaction of this acid chloride with aqueous acid regenerates propanoic acid. Acid chlorides react with alcohols to give the corresponding ester, as in the conversion of propanoyl chloride to the corresponding methyl propanoate. The reaction of propanoyl chloride with methylamine yields the amide, N-methylpropanamide. Similarly, methyl propanoate reacts with the amine to yield the same amide. These simple examples summarize the acyl substitution reactions of acid derivatives, and examples will appear throughout this book. O OH Propanoic acid

SOCl2 aq H+

O

CH3OH

Cl Propanoyl chloride

O

CH3NH2

OCH3 Methyl propanoate

O NHCH3 N-methylpropanamide

CH3NH2

Acyl addition can be reversible when the leaving group ability of the acyl derivative is close to that of the new nucleophile, as in the transesterification reaction of ethyl butanoate and methanol. The acid-catalyzed reaction of ethyl butanoate and methanol generates a tetrahedral intermediate bearing both OMe and OEt, and the equilibrium is driven to the methyl ester only if a large excess of methanol is present. Recognizing the mechanistic details of this process allows one to control the reaction, making a process like transesterification useful in synthesis. An interesting transesterification example is taken from a synthesis of the banana volatile (S)-2-pentyl (R)-3-hydroxyhexanoate by Smonou and coworkers.11 In this work, an enzyme identified as immobilized Candida antarctica lipase B (CAL-B, Novozym 435) was used rather than an acid catalyst, and methyl 3-oxohexanoate reacted with (S)-pentan-2-ol under these conditions to give a 90% yield of (S)-pentan-2-yl 3-oxohexanoate. Other enzymatic reactions will be discussed in Section 7.12.6.

9

Fieser, L. F.; Fieser, M. Reagents for Organic SynthesisL. F.; Fieser, M; Wiley: New York, 1967; Vol. 1, p 873.

10

For an example from a synthesis see Mori, K.; Tashiro, T.; Sano, S. Tetrahedron Lett. 2000, 41, 5243.

11

Kallergi, M.; Kalaitzakis, D.; Smonou, I. Eur. J. Org. Chem. 2011, 3946.

166

4. ACIDS, BASES, AND FUNCTIONAL GROUP EXCHANGE REACTIONS

O

HO

O

O

(S)-Pentan-2-ol

CH3

O

CAL-B , 45°C , 1 d

O Methyl 3-oxohexanoate

O (S)-Pentan-2-yl 3-oxohexanoate (90%)

CAL-B = Immobilized Candida antarctica lipase B (CALNovozym 435)

Carbon nucleophiles, “R3C”, include Grignard reagents or alkyne anions (see Section 11.4), which react with acid derivatives to form the tetrahedral intermediate in which there are strong CdC or CdH bonds, and the H or C units are not considered to be leaving groups. The tetrahedral intermediate derived from an acid chloride or ester, for example, will have a leaving group. After loss of the leaving group, the product of the reaction should therefore be a ketone, as shown. The presence of a leaving group in acid derivatives and therefore in the tetrahedral intermediate leads to this fundamental difference in reactivity between acyl addition and substitution. O

R3C

R

O

–

–X

R

X

O

–

X CR3

R

CR3

Tetrahedral intermediate

A ketone product in this reaction may compete for reaction with the nucleophile, which can cause problems with the reaction. This issue will be discussed for specific reactions in later chapters, but it can be illustrated here. Nucleophilic addition of Grignard reagents to an ester is one example where acyl addition yields a product that can react with the reagent.12 Initial addition of PhMgBr to the carbonyl of methyl butanoate yields the tetrahedral intermediate, 1-methoxy-1-phenylbutan-1-olate, which loses methoxide to yield the ketone, 1-phenylbutan-1-one (see Section 11.4.3.2). Ketones are slightly more reactive than the ester starting material. Remember that after one half-life for the reaction, both the ketone product and ester starting material are available in the reaction mixture, so there is a competition for reaction with the Grignard reagent. The 1-phenylbutan-1-one product reacts with phenylmagnesium bromide to yield 1,1-diphenylbutan-1-olate. Treatment with aqueous acid leads to an acid-base reaction that yields a tertiary alcohol, 1,1-diphenylbutan-1-ol (see Section 11.4.2.2). In a typical reaction, all of the ester and Grignard reagent will be consumed, and there will be a mixture of the ketone and alcohol products. If an excess of the Grignard reagent is used, the alcohol is usually the only product (see Section 11.4.3.2). OMe

OMe O

–O

PhMgBr

Methyl butanoate

Ph

Ph

–MeO

–

Ph –O

O PhMgBr

1-Methoxy-1-phenyl- 1-Phenylbutan-1-one butan-1-olate

Ph

Ph

H3O

+

1,1-Diphenyl butan-1-olate

HO

Ph

1,1-Diphenyl butan-1-ol

Esters, anhydrides, and amides yield similar products via acyl substitution. The best leaving group is chloride followed closely by carboxylate (acid chlorides and anhydrides, respectively) and then the alkoxy group from an ester. Amides have an NR2 leaving group and are the least reactive. A useful approximation correlates the reactivity of the carbonyl moiety with the leaving group propensity of the ligand attached to carbon. This leads to the usual order of reactivity to nucleophilic acyl substitution as follows: O

O

O

R

Cl

O

O

>

> R

O

R

> R

OR1

R

N(R1 )2

Therefore, acid chlorides and anhydrides should react with Grignard reagents to yield a higher percentage of alcohol, whereas an ester might yield a mixture of ketone and alcohol products. For the most part, this analysis is 12

Kharasch, M. S.; Reinmuth, O. Grignard Reactions of Nonmetallic Substances; Prentice-Hall: Englewood Cliffs, NJ, 1954; pp 549–766, 846–869.

167

4.3 CONJUGATE ADDITION

correct, but the product distribution depends on the reactivity of the Grignard reagent, the nature of the acid derivative, the solvent, and temperature of the reaction. If a large excess of the nucleophile (e.g., excess Grignard reagent) is used, most reactions are driven to give the tertiary alcohol as the major (often the only) product. In later chapters, the problem of reactivity of the ketone product with organometallic reagents will be circumvented by use of a dialkylcadmium reagent (Section 11.4.3.2)13 or an organocuprate (Section 12.3.1.4). In some cases, specialized amides can give good yields of the acyl addition product. In a synthesis of (+)-seimatopolide A, Prasad and Revu14 reacted ester 11 with methoxymethanamine in the presence of isopropylmagnesium chloride to yield the amide 12 in 94% yield. N-methoxy-N-methyl amide (e.g., 12) is referred to as a Weinreb amide (Section 11.4.3.2).15 Subsequent reaction with nonylmagnesium bromide gave ketone 13 in 84% yield. Grignard reactions with acid derivatives will be discussed further in Section 11.4.2.2. O

O

OTBS NH(OMe)Me•HCl

O

O

OTBS

C9H19 MgBr

O

Me

N

i-PrMgCl , THF 0°C , 1 h

EtO

OTBS THF , 0°C 1 h

O C9H19

OMe

11

12 (94%)

13 (84%)

4.3 CONJUGATE ADDITION Conjugated ketones, aldehydes, esters, and related molecules have a carbonyl unit connected directly to the π-bond of an alkene (as in 14).16 Inductive effects make the terminal carbon of the alkene electrophilic, and subject to attack by a nucleophile in what is formally a vinylogous acyl addition. In effect, the C]C can react as a vinylogous carbonyl due to the inductive effects. This type of reaction is referred to as conjugate addition or Michael addition17 (see Sections 11.4.4, 12.3.2.2, and 13.7) and it is facilitated by formation of a resonance-stabilized product, the enolate anion (15). An acidbase reaction with aqueous acid yields the final addition product, ketone 16. Conjugate addition competes with normal nucleophilic attack at the acyl carbon (acyl addition), and the major product is usually determined by steric hindrance at the carbonyl or at the alkene carbon, and by the nature of the nucleophile. When the R group in 14 is hindered, attack at the acyl carbon can be difficult, leading to more conjugate addition. Nuc

O

O

O R 14

H3O

Nuc

Nuc R 15

+

O Nuc R

R 16

When R in 14 is small (e.g., the hydrogen atom in conjugated aldehydes), normal 1,2-addition to the carbonyl (acyl addition) is usually preferred. Acyl addition is typically preferred when nucleophiles (e.g., alkyne anions and Grignard reagents) react with conjugated aldehydes, but organocuprates give conjugate addition with conjugated aldehydes and with most α,β-unsaturated carbonyl derivatives.18,19 These reactions will be discussed in Section 13.7, but an example is shown here, taken from a synthesis of stemaphylline N-oxide by Lindsley and coworkers,20 in which conjugate amide 17 was treated with lithium dimethylcuprate to yield 90% of 18.

13

(a) Jones, P. R.; Desio, P. J. Chem. Rev. 1978, 78, 491. (b) Cason, J.; Fessenden, R. J. Org. Chem. 1960, 25, 477.

14

Prasad, K. R.; Revu, O. J. Org. Chem. 2014, 79, 1461.

(a) Nahm, S.; Weinreb, S. M. Tetrahedron Lett. 1981, 22, 3815. (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; p 866.

15

16

For examples see Ref. 1, pp 1567–1616, 1809–1842.

17

Bergmann, D.; Ginsburg, D.; Pappo, R. Org. React. 1959, 10, 179.

18

(a) Posner, G. H. An Introduction to Synthesis Using Organocopper Reagents; Wiley-Interscience: New York, 1980. (b) Mandeville, W. H.; Whitesides, G. M. J. Org. Chem. 1974, 39, 400. (c) House, H. O.; Wilkins, J. M. Ibid. 1978, 43, 2443. (d) Whitesides, G. M.; San Filippo, J., Jr.; Casey, C. P., Jr.; Panek, E. J. J. Am. Chem. Soc. 1967, 89, 5302.

19

(a) Davis, R.; Untch, K. G. J. Org. Chem. 1979, 44, 3755. (b) Stork, G.; Isobe, M. J. Am. Chem. Soc. 1975, 97, 6260, 4745.

20

Schulte, M. L.; Turlington, M. L.; Phatak, S. S.; Harp, J. M.; Stauffer, S. R. Lindsley, C. W. Chem. - Eur. J. 2013, 19, 11847.

168

4. ACIDS, BASES, AND FUNCTIONAL GROUP EXCHANGE REACTIONS

TBDPSO

O

TBDPSO

O

O

Me2CuLi THF

N

O

O N

O

Me3SiCl , –78°C

Ph

Ph

17

18 (90%)

Conjugate addition will be a key feature of the Robinson annulation,17,21 which will be discussed in Section 13.7.3. In this important reaction, a ketone enolate reacts with a conjugated ketone to yield a cyclohexenone derivative.

4.4 FUNCTIONAL GROUP MANIPULATION BY REARRANGEMENT There are several common functional group manipulations that involve a skeletal rearrangement. Many of these reactions involve carbonyl compounds. This section will discuss a few examples that correlate with named reactions, both to illustrate the principle and to show the details of these specific transformations.

4.4.1 Beckmann Rearrangement Oximes are readily obtained by the reaction of an aldehyde or ketone with hydroxylamine (NH2OH). When oximes are treated H2SO4, PCl5, and related reagents, a rearrangement occurs to yield substituted amides in what is called the Beckmann rearrangement.22 Initial protonation of the hydroxyl group of an oxime yields 19, and subsequent rearrangement displaces water as a leaving group to yield nitrilium salt 20. Nitrilium salts react with water with loss of a proton to yield an amide.23 O R

N–OH

NH2OH

R1

R

R1

N–OH2

H+

R1 R1

R

An oxime

19

N

R

O

+ H2O; – H+

RHN

– H 2O

20

R1

An amide

The group that migrates is generally the one anti to the hydroxyl unit, and this observation has been used as a method of determining the configuration of the oxime. However, the syn group may migrate in some oximes, especially where R and R0 are both alkyl. In most cases, the oxime can undergo isomerization under the reaction conditions before migration takes place, which leads to an isomer.24 Therefore, it is possible to get mixtures of the two different amides from oximes derived from unsymmetrical ketones. Note that a hydrogen atom seldom migrates, and the oxime can tolerate most alkyl groups in this reaction. In alkyl aryl ketones, the aryl group generally migrates preferentially. Beckmann rearrangement of an oxime of a cyclic ketone (e.g., cyclohexanone oxime), leads to ring enlargement under these conditions, with formation of a lactam, azepan-2-one (caprolactam) in this case.25 Beckmann rearrangements have also been carried out photochemically.26 21

duFeu, E. C.; McQuillin, F. J.; Robinson, R. J. Chem. Soc. 1937, 53.

For reviews, see (a) Gawley, R. E. Org. React. 1988, 35, 1; McCarty, C. G. In The Chemistry of the Carbon–Nitrogen Double Bond; Patai S., Ed.; John Wiley: NY, 1970; pp 408–439. (b) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-7. (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 80–81. Also see (d) Nguyen, M. T.; Raspoet, G.; Vanquickenborne, L. G. J. Am. Chem. Soc. 1997, 119, 2552. 22

23

Donaruma, L. G.; Heldt, W. Z. Org. React. 1960, 11, 1, see p 5.

24

Lansbury, P. T.; Mancuso, N. R. Tetrahedron Lett. 1965, 2445.

25

Vinnik, M. I.; Zarakhani, N. G. Russ. Chem. Rev. 1967, 36, 51.

(a) Izawa, H.; de Mayo, P.; Tabata, T. Can. J. Chem. 1969, 47, 51. (b) Cunningham, M.; Ng Lim, L. S.; Just, T. Can. J. Chem. 1971, 49, 2891. (c) Suginome, H.; Yagihashi, F. J. Chem. Soc. Perkin Trans. 1977, 1, 2488.

26

169

4.4 FUNCTIONAL GROUP MANIPULATION BY REARRANGEMENT

HN

PCl5

HO—N

O Azepan-2-one (caparolactam)

Cyclohexanone oxime

Esters of oximes (R3C]NdOR1) undergo the Beckmann rearrangement, and many organic and inorganic acids can be used to initiate the reaction. Under the correct conditions, cycloalkane carboxylic acids rearrange to form lactams. Azacyclododecan-2-one, for example, was prepared from cycloundecan-2-carboxylic acid by reaction with nitrosyl sulfuric acid in chlorosulfonic acid.27 A side reaction with some oximes is the formation of nitriles in what is called an abnormal Beckmann rearrangement.28 Another useful variation of the Beckmann rearrangement is the direct treatment of a ketone with hydroxylamine-Osulfonic acid in the presence of an acid (e.g., formic acid), or with sulfuric acid in a second step. Initial formation of an oxime-O-sulfonic acid is followed by in situ formation of an oxime, which rearranges in the presence of the acid.29 In the example shown, 2-pentylcyclohexan-1-one reacted with hydroxylamine-O-sulfonic acid and formic acid to yield (((2-pentylcyclohexylidene)amino)oxy)sulfonic acid, and rearrangement in situ gave a 71% yield of 7-pentylhexahydroazepin-2-one (7-pentylazepan-2-one).30 A synthetic example of the Beckmann rearrangement is taken from the Moody and coworker’s31 synthesis of the naphthoquinone-azepinone core of hygrocin B and divergolide C. Oxime 21 was treated with tosyl chloride to give the O-tosyl oxime, (22), in 94% yield. Subsequent Beckmann rearrangement induced by treatment with AlCl3 gave lactam 23 in 83% yield.31 O

OSO3H

N C5H11 H2N–OSO3H

O

C5H11

N

HCOOH

2-Pentylcyclohexan-1-one

H C5H11

(((2-Pentylcyclohexylidene)amino)oxy)sulfonic acid

OH

7-Pentylazepan-2-one (71%)

OTs

N

H N

N

O

AlCl3 , CH2Cl2

TsCl , NaH THF

MeO Me

CO2Et

21

CO2Et

MeO Me

22 (94%)

MeO Me

CO2Et 23 (83%)

4.4.2 Schmidt Rearrangement The addition of hydrazoic acid to carboxylic acids, aldehydes, ketones, alcohols, and alkenes all yield amides, and all are known as the Schmidt reaction.32 The reaction with carboxylic acids is probably the most common, involving initial formation of an acyl azide, which rearranges with loss of nitrogen to yield a transient isocyanate. Subsequent reaction with water gives a carbamic acid, which decarboxylates to yield the amine product.33 Sulfuric acid and many (a) Imperial Chemical Industries LTD Belgian Patent 616 544, Oct. 17, 1962 (Chem. Abstr. 1963, 59, 452). (b) Nagasawa, H. T.; Elberling, J. A.; Fraser, P. S.; Mizuno, N. S. J. Med. Chem. 1971, 14, 501.

27

(a) Hill, R. K.; Conley, R. T. J. Am. Chem. Soc. 1960, 82, 645 and see (b) Leuckart, R.; Bach, E. Berchte 1887, 20, 104. (c) Schroeter, G.; Ibid. 1911, 44, 1201. (d) Perkin, W. H., Jr.; Titley, A. F. J. Chem. Soc. 1921, 19, 1089. (e) Glover, W. H. Ibid. 1908, 93, 1285. (f ) Price, C. C.; Mueller, G. P. J. Am. Chem. Soc. 1944, 66, 634. (g) Bartlett, P. D.; Stiles, M. Ibid. 1955, 77, 2806. (h) Hatch, M. J.; Cram, D. J. Ibid. 1955, 76, 38.

28

29

(a) Smith, P. A. S. J. Am. Chem. Soc. 1948, 70, 323. (b) Sanford, J. K.; Blair, F. T.; Arroya, J.; Sherk, E. W. J. Am. Chem. Soc. 1945, 67, 1941.

30

Duhamel, P.; Kotera, M.; Monteil, T.; Marabout, B.; Davoust, D. J. Org. Chem. 1989, 54, 4419.

31

Nawrat, C. C.; Kitson, R. R. A.; Moody, C. J. Org. Lett. 2014, 16, 1896.

32

(a) Banthorpe, D. V. In The Chemistry of the Azido Group; Patai, S., Ed.; John Wiley: NY, 1971; pp 405–434. (b) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-83. (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 574–575.

33

Koldobskii, G. I.; Ostrovskii, V. A.; Gidaspov, B. V. Russ. Chem. Rev. 1978, 47, 1084.

170

4. ACIDS, BASES, AND FUNCTIONAL GROUP EXCHANGE REACTIONS

Lewis acids can be used as catalysts in this reaction. Under normal reaction conditions, the isocyanate is not isolated. When hydrazoic acid reacts with a ketone, the product of the rearrangement is an amide.34 O R

O

HN3 , H+

R

OH

R

N3

N

C

HNR

– N2

Isocyanate

Acyl azide O

– CO2

OH

RNH2

Carbamic acid O

HN3 , H+

NHR1

R

R1

R

O

H 2O

O

Dialkyl ketones and cyclic ketones generally react faster than alkyl aryl ketones, and these react faster than diaryl ketones. Cyclic ketones give lactams,35 as observed in the Beckmann rearrangement, and a simple example is the reaction of cyclohexanone with HBr and sodium azide. A 63% yield of azepan-2-one (hexahydroazepine-2-one; caprolactam) was obtained in water, and a 66% yield was obtained in acetic acid in a reaction with HBr and NaNO2.36 A variation of the Schmidt reaction treated the ketone with hydrazoic acid, another acid, and also obtained a lactam. Under these conditions 2-isopropylcyclopentan-1-one reacted with HN3/H2SO4 at 3–7°C to yield a 63% yield of 6-isopropylpiperidin-2-one.37 O

O

H

HBr , NaNO2

N

H 2O

Azepan-2-one (63%)

Cyclohexanone

O O HN3 , H2SO 4

N

H

3–7°C

2-Isopropylcyclopentan-1-one

6-Isopropylpiperidin-2-one (63%)

This reaction is rarely applied to aldehydes, because the Schmidt reaction will usually give a nitrile as the product. Nitriles can be observed as by products with ketones as well. Regioselectivity can be a problem when unsymmetrical cyclic ketones are converted to the lactam. When 2-methylcyclohexanone reacted with sodium azide in the presence of polyphosphoric acid,38 3-methylazepan-2-one and 7-methylazepan-2-one were formed.39 The reaction often proceeds with good regioselectivity, however. A synthetic example illustrates an intramolecular variation. In a synthesis of antofine by Gu and coworkers,40 azido-aldehyde 24 was treated with trifluoroacetic acid (TFA), which induced an intramolecular Schmidt reaction to yield a mixture of 68% of 25 and 17% of 26. In this synthesis, 25 was easily separated and converted to antofine. A useful variation of the Schmidt reaction involves reaction of a silyl enol ether of a cyclic ketone with TMSN3,41 followed by photolysis of the product with ultraviolet light to yield a lactam.

Koldobskii, G. I.; Tereschenko, G. F.; Gerasimova, E. S.; Bagal, L. I. Russ. Chem. Rev. 1971, 40, 835. (b) Beckwith, A. L. J. In The Chemistry of Amides; Zabicky, J., Ed.; John Wiley: NY, 1970; pp 137–145. 34

35

Krow, G. R. Tetrahedron 1981, 37, 1283.

36

Smith, P. A. S. J. Am. Chem. Soc. 1948, 70, 320–323.

37

Shechter, H.; Kirk, J. C. J. Am. Chem. Soc. 1951, 73, 3087.

38

Conley, R. T. J. Org. Chem. 1958, 23, 1330.

39

Overberger, C. G.; Parker, G. M. J. Polym. Sci., A 1968, 6, 513.

40

Yi, M.; Gu, P.; Kang, X.-Y.; Sun, J.; Li, R.; Li, X.-Q. Tetrahedron Lett. 2014, 55, 105.

41

Evans, P. A.; Modi, D. P. J. Org. Chem. 1995, 60, 6662.

171

4.4 FUNCTIONAL GROUP MANIPULATION BY REARRANGEMENT

OMe

OMe

MeO

OMe

MeO CHO

MeO

TFA , DCM

+

0°C

N

O

OHC N3

MeO

MeO

N H

MeO

24

25 (68%)

26 (17%)

4.4.3 Related Rearrangements An intermediate in the Schmidt rearrangement has been shown to be an acyl azide, and the intermediate isocyanate is usually not isolated under those conditions. The Schmdit rearrangement is generally done in an aqueous medium, but without aqueous conditions the isocyanate is a useable intermediate. Indeed, the thermal rearrangement of acyl azides to isocyanates is known as the Curtius rearrangement.42 Once formed, subsequent reaction of an isocyanate with water, an alcohol, or an amine lead to an amine, a carbamate, or an acylurea.43 In general, acyl azides are prepared by treatment of acylhydrazines (hydrazides) with nitrous acid. Lewis acids can catalyze the Curtius rearrangement, but they are not required. The Curtius rearrangement is relatively common in synthesis. In a synthesis of ()-lundurine B by Nishida and coworkers,44 acid 27 was treated with DPPA to yield the acyl azide, 28 in >90% yield. In the presence of tert-butanol and triethylamine, the product was the N-Boc protected amine 29, via reaction of the alcohol with an intermediate isocyanate. O R R N3 Acyl azide OBn H MeO

HO O

N

P

PhO PhO

C

O

Isocyanate

O

OBn

OBn

H

H

N3

O O

– N2

MeO

NEt3, t-BuOH MS 4 Å , 80°C 62.5 h

27

MeO

O

HN O

N3 O 28 (>90%)

O

O t-Bu

O

29

When an amide is treated with sodium hypobromite (NaOBr; NaOH and Br2) an isocyanate is formed,45 and subsequent hydrolysis liberates an amine with one less carbon than the starting amide, in what is known as the Hofmann rearrangement.46 Ureas and acylureas are sometimes formed in this reaction. N-acyl derivatives of hydroxamic acids47 yield isocyanates when treated with base, or upon heating, in the Lossen rearrangement.48 Similarly, aromatic acyl halides are converted to amines when treated with hydroxylamine-O-sulfonic acid.49

42

(a) Banthorpe, D. V. In The Chemistry of the Azido Group; Patai, S., Ed.; John Wiley: NY, 1971; pp 397–405. (b) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-21. (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 192–193.

43

(a) Pfister, J. R.; Wyman, W. E. Synthesis 1983, 38. See also (b) Capson, T. L.; Poulter, C. D. Tetrahedron Lett. 1984, 25, 3515.

44

Hoshi, M.; Kaneko, O.; Nakajima, M.; Arai, S.; Nishida, A. Org. Lett. 2014, 16, 768.

45

Sy, A. O.; Raksis, J. W. Tetrahedron Lett. 1980, 21, 2223.

(a) Wallis, E. S.; Lane, J. F. Org. React. 1946, 3, 267–306. (b) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-46. (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 326–327. 46

47

Bauer, L.; Exner, O. Angew. Chem. Int. Ed. Engl. 1974, 13, 376.

48

Salomon, C. J.; Breuer, E. J. Org. Chem. 1997, 62, 3858.

49

Wallace, R. G.; Barker, J. M.; Wood, M. L. Synthesis 1990, 1143.

172

4. ACIDS, BASES, AND FUNCTIONAL GROUP EXCHANGE REACTIONS

These reactions yield the following retrosynthetic transforms: O

O

O

O

RJNH2 R

OH

R

NHR

R

O NH

R

4.5 MACROCYCLIC COMPOUNDS Cyclic compounds were introduced in Section 1.5.2, and many molecules discussed in this book involve monocyclic compounds that contain a variety of functional groups. For the most part, these monocyclic compounds will be 3 to 8-membered rings. There are compounds that have much larger rings, however, including important 9 to 20-membered ring compounds. Such monocyclic compounds, including those of even larger ring size are generically known as macrocycles, or macrocyclic compounds. Macrocyclic lactones, or macrolides, are important members of this class of compounds, in large part because many macrolides have important biological properties (e.g., antibiotic activity). Lactones are, of course, cyclic esters and are generated by intramolecular acyl addition reactions. Three examples include exaltolide, which was isolated by Ciamician and Silber50 in 1896, and is a widely used macrocyclic musk lactone.51 A 10-membered macrolide, modiolide A, was isolated from a cultured broth of the fungus Paraphaeosphaeria sp., which was separated from a marine horse mussel.52 Fluvirucin B1 is a 14-membered macrolactam produced by Actinomadura vulgaris and has been shown to have moderate-to-good antifungal and antiviral activities.53 Macrolactonization has become an important reaction in the synthesis of natural products.54 This chapter discusses the chemistry of carbonyl compounds, and this section will introduce methods for the preparation of macrolides via ring-closing reactions, and also introduce issues related to cyclization reactions involving large rings. O

O

NH

O

O

O OH

HO OH Exaltolide

Modiolide A

Fluvirucin B1

4.5.1 Macrocyclic Ring Closures Formation of carbocyclic rings will be discussed in some later chapters in connection with the appropriate carbon– carbon bond-forming reactions. Macrolactonization, however, is a functional group exchange process that involves acyl substitution reactions. Large lactone rings are an important feature of many natural products, but the formation of a large ring poses unique problems. This section will discuss the problems and solutions for macrocyclic cyclization reactions. The principles discussed here for preparing large ring lactones are applicable to most other macrocyclization reactions. Illuminati and Mandolini55 described the ring-closing reactions of bifunctional chain molecules. In the 1920s and 1930s, Ruzicka et al.56 and Ziegler et al.57 studied macrocyclic reactions. Formation of macrocyclic rings usually requires an intramolecular cyclization reaction of a bifunctional molecule (e.g., 30), where cyclization yields the monocyclic product, 31. In this model, X and Y are reactive functional groups that generate a new bond, represented by Z (which may contain X, Y, or both). 50

Ciamician, G.; Silber, P. Ber. 1896, 29, 1811.

51

William, A. S. Synthesis 1999, 10, 1707.

52

Tsuda, M.; Mugishima, T.; Komatsu, K.; Sone, T.; Tanaka, M.; Mikami, Y.; Kobayashi, J. J. Nat. Prod. 2003, 66, 412.

(a) Naruse, N.; Tenmkyo, O.; Kawano, K.; Tomita, K.; Ohgusa, N.; Miyaki, T.; Konishi, M.; Oki, T. J. Antibiot. 1991, 44, 733. (b) Naruse, N.; Tsuno, T.; Sawada, Y.; Konishi, M.; Oki, T. J. Antibiot. 1991, 44, 741. 53

54

Parenty, A.; Moreau, X.; Campagne, J.-M. Chem. Rev. 2006, 106, 911.

55

Illuminati, G.; Mandolini, L. Acc. Chem. Res. 1981, 14, 95.

56

Ruzicka, L.; Stoll, M.; Schinz, H. Helv. Chim. Acta 1926, 9, 249.

57

Ziegler, K.; Eberle, H.; Ohlinger, H. Annalen 1933, 504, 94.

173

4.5 MACROCYCLIC COMPOUNDS

Y

X

Z

30

31

An intermolecular reaction competes with cyclization where initial coupling generates the dimeric species (32). Repeated intermolecular reactions yield oligomers or polymers (33). Ruggli58 discovered that high-substrate concentrations favor polymerization, while low concentrations favor cyclization. The rate of cyclization is a function of the structure of the open-chain precursor and that of the product-like transition state. The activation energy for ring closure is largely determined by the strain energy of the final ring.59 As shown in Sections 1.5.2 and 1.5.C, strain energy is due to “(1) bond opposition forces due to imperfect staggering (Pitzer strain), (2) deformation of ring bond angles (Baeyer strain), and (3) transannular strain due to repulsive interactions between atoms across the ring when they are forced close to each other.”55,60 As the chain length increases for a cyclization reaction, the probability of the chain termini approaching each other decreases (negative ΔS{ due to less freedom of internal rotation around single bonds of the molecular backbone when the disordered, open-chain precursor is converted to the ring-shaped transition state).55 In general, the ring product is a good model of the transition state of cyclization (a ring-shaped transition state) for all but the shortest chains.61 Y

Y

Y Y

+

Z

Z X

Z

X X

n

X

32

33

When large rings are formed that are free of strain, there is minimal strain energy in the transition state. In short chains, an advantage in terms of entropy is offset by an increase in enthalpy due to extremely large strain energies. Ziegler62 first used this principle of a ring-shaped transition state to generate large-membered rings in what is known as the high-dilution method. Cyclization of 34 to 35 was accomplished by slow addition (6.6  104 mol L1 d1) of bromoacid to a solvent containing potassium carbonate (K2CO3) or hydroxide,63 for example. For n ¼ 9, a 77% yield of 35 was obtained.63 The main side product of this process was the dimeric ester 36. Br (CH 2)n

Br

O O (CH2)n

kinter

CO2H

(CH 2)n

(CH 2)n

O

kintra

O

HO2C

36

34

35

Bromocarboxylate (34) can be cyclized to a lactone, and this simple reaction will be used to illustrate cyclization processes in general. An intramolecular cyclization reaction of 34 (governed by kintra) will generate the lactone (35), but an intermolecular reaction (kinter) will yield the dimeric ester (36).55 The relative preference of a reaction is given by the effective molarity (EM) ¼ kintra/kinter.64 The EM is supposed to be the first-order rate constant for ring closure times the secondorder rate constant for reaction between chain ends (if they were not connected), but the EM for five- and six-membered rings exceeds the real concentration.55 Two parameters similar to the EM are useful for predicting conditions under which a ring can be synthesized, free of significant polymerized byproducts.65 The first is the parameter Cα/Mo (see Eq. 4.1), where kR is the rate of ring formation of a cyclic monomer and kP is the rate of formation of cyclic dimer. If the initial concentration is less than unity (α < 1), the yield is not < 55%. For medium rings, kR/kP should be < 0.1 M.66 58

(a) Ruggli, P. Annalen 1912, 392, 92. (b) Idem Ibid. 1913, 399, 174. (c) Idem Ibid. 1916, 412, 1.

59

Liebnan, J. F.; Greenberg, A. Chem. Rev. 1976, 76, 311.

60

Allinger, N. L.; Tribble, M. T.; Miller, M. A.; Wertz, D. H. J. Am. Chem. Soc. 1971, 93, 1637.

61

Galli, C.; Mandolini, L. Eur. J. Org. Chem. 2000, 3117.

62

Ziegler, K. In Methoden der Organischen Chemie (Houben-Weyl); M€ uller, E., Ed.; Georg Thieme Verlag: Stuttgart, 1955; Vol. 4/2.

63

(a) Hunsdiecker, H.; Erlbach, H. Chem. Ber. 1947, 80, 129. (b) Stoll, M. Helv. Chim. Acta 1947, 30, 1393.

64

Illuminati, G.; Mandolini, L.; Masci, B. J. Am. Chem. Soc. 1977, 99, 6308.

65

Galli, C.; Mandolini, L. Gazz. Chim. Ital. 1975, 105, 367.

66

Illuminati, G.; Mandolini, L.; Masci, B. J. Am. Chem. Soc. 1974, 96, 1422.

174

4. ACIDS, BASES, AND FUNCTIONAL GROUP EXCHANGE REACTIONS

Cα 1 Mo ¼ ln ð1 + 2αÞ where α ¼ Mo kR =kP 2α

(4.1)

and kR and kP are defined by kR

Monomer ! Monomeric complex kP 2 Monomer ! Dimer Similarly, the parameter ηc/ηm (the yield of cyclic substance obtained by high-dilution conditions) can be used, where ηm is the total number of moles of bifunctional substrate, ηc is the number of moles of monomeric ring product produced, and νf is the constant feed rate that is measured in mol L1 s1.65 The dimensionless parameter β is a measure of the feed rate, and the relative formation of cyclic monomer is a function of the feed rate.65 ηc 2 2 ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ pffiffiffiffiffiffiffiffiffiffiffiffi ηm 1 + 8β 1 + 8v k =ðk Þ2 2.5 2 1.5 1 0.5 0 −0.5 −1 −1.5 −2 −2.5 −3 −3.5 −4

22 21

ΔS (eu)

20 19 18 17 16 15 14

2

(A)

(4.2)

R

23

ΔH (kcal mol–1)

log k(intra)

f p

4

6

8 10 12 14 16 18 20 22 24 Ring size

2

4

(B)

6

0 −2 −4 −6 −8 −10 −12 −14 −16 −18 −20 −22 −24

8 10 12 14 16 18 20 22 24 Ring size

2

(C)

4

6

8 10 12 14 16 18 20 22 24 Ring size

FIG. 4.1 For lactone formation: (A) reactivity profile, (B) ΔH{ profile, and (C) ΔS{ profile. Reprinted with permission from Illuminati, G.;

Mandolini, L. Accts. Chem. Res., 1981, 14, 95. Copyright © 1981 American Chemical Society.

The dependence on ring size for lactone formation is illustrated by Fig. 4.1A (the reactivity for lactone formation), Fig. 4.1B (the ΔH{ profile for lactone formation), and Fig. 4.1C (the ΔS{ profile for lactone formation).55 The quantity ΔH{intra  ΔH{inter is a useful measure of the strain energies involved in ring formation of lactones. As the chain length increases, ΔH{ decreases.67 For very large rings (C18–C23), ΔH{intra  ΔH{inter and the intramolecular reaction is nearly strain free. For ring sizes of C3–C16, ring formation is strain dependent with ΔH{ maximal at C3 and C8 [ 8 kcal (33.5 kJ) mol1]. For C3–C8 rings, the cis conformation predominates due to the increased strain inherent to a trans conformation. As shown in Fig. 4.1C, ΔS{ does not show a regular pattern, but is more favorable for large rings because of a “looseness in the ring due to low frequency out of plane bending motions.”55,68 The variation in ΔS{ indicates the importance of individual ring features, with maxima occurring at ring sizes of C8, C12, and C16. For ring sizes of C3–C6, the ΔS{ values are relatively insensitive to ring size. Solvation around the reactive end plays a major role in C3 and C4 rings, but for C5 rings and larger, such solvation is less important.

4.5.2 Synthetic Approaches to Macrocyclic Lactones There are many examples of biologically important, naturally occurring lactones.69 High-dilution techniques can be used, as mentioned above. In the final step of a synthesis of amphidinolide P, Trost et al.70 simply heated a dilute 67

Mandolini, L. J. Am. Chem. Soc. 1978, 100, 550.

68

O'Neal, H. E.; Benson, S. W. J. Chem. Eng. Data 1970, 15, 266.

For a review of macrolactonization in the total synthesis of natural products, See (a) Parenty, A.; Moreau, X.; Campagne, J.-M. Chem. Rev. 2006, 106, 911. Also see (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 402–403. For a review of syntheses of medium-sized ring lactones, see Shiina, I. Chem. Rev. 2007, 107, 239.

69

70

Trost, B. M.; Papillon, J. P. N.; Nussbaumer, T. J. Am. Chem. Soc. 2005, 127, 17921.

175

4.5 MACROCYCLIC COMPOUNDS

solution of 36 in hexane to give an 84% yield of amphidinolide P (37). In this reaction, the eight-membered ring lactone was opened to yield a seco-acid, which closed to the macrocyclic lactone, and the hydroxyl group remaining from the eight-membered ring lactone attacked the ketone moiety to generate the hemiketal unit in 37. O

O 0.001 M in hexane reflux , 1 h

O O

O HO

O

HO

O

O 36

37 (84%)

A variety of cyclization techniques have been developed, but all are based on the idea that the carbonyl end of a ω-substituted acid is “activated” to facilitate attack by or at the other functionalized end. Trifluoroacetic anhydride reacts with the carboxylic acid moiety of a seco-acid to form a mixed anhyride, which activates the carbonyl to attack by the hydroxyl moiety to give the macrolactone.71 Cyclization has also been observed using a mixture of TFA and trifluoroacetic anhydride.72 The use of the reagents 2-methyl-6-nitrobenzoic anhydride (MNBA) and dimethylaminopyridine constitutes the so-called Shiina macrolactonization, 73 This procedure was used for the cyclization of 38 to 39 in 57% yield, in a synthesis of calcaripeptide C by Das and Goswami.74 O

OH CO2Me

O N

Ph

O

1. LiOH•H2O , THF/H2O

O

2. MNBA , DMAP, CH2Cl2, rt

N H

N

OTIPS Ph

38

O

O

N H

OTIPS

39 (57%)

Another macrocyclization technique has been developed based on formation of a mixed anhydride. In the Yamaguchi protocol75 a seco-acid is treated with 2,4,6-trichlorobenzoyl chloride, and the resulting mixed anhydride is heated with DMAP in toluene. In Nagasawa’s and Kuwahara76 synthesis of lactimidomycin, for example, seco-acid 40 was converted to lactone 41 in >90% yield using the Yamaguchi protocol. A variation of this procedure prepares the mixed anhydride by reaction of the seco-acid with Boc anhydride (also see Section. 5.3.4.3), followed by heating with DMAP in toluene.77

OTES

SePh CO H OH 2

1. 2,4,6-Cl3C6H2COCl NEt3

OTES

SePh

2. DMAP

40

O

O

41 (>90%)

Mercuric trifluoroacetate [Hg(OCOCF3)2] is an effective reagent for the cyclization of hydroxy thio-esters.78 Masamune et al.79 used this cyclization procedure for the conversion of 42 to 44 (in 90% yield) in a synthesis of zearalenone dimethyl ether. Masamune et al.79,80 developed this procedure (the Masamune protocol) to complete the For examples, see (a) Taub, D.; Girotra, N. N.; Hoffsommer, R. D.; Kuo, C. H.; Slates, H. L.; Weber, S.; Wendler, N. L. Tetrahedron 1968, 24, 2443. (b) Idem J. Chem. Soc. Chem. Commun. 225.

71

72

Baker, P. M.; Bycroft, B. W.; Roberts, J. C. J. Chem. Soc. C 1967, 1913.

73

(a) Shiina, I.; Kubota, M.; Oshiumi, H.; Hashizume, M. J. Org. Chem. 2004, 69, 1822. (b) Wu, Y.; Yang, Y.-Q. J. Org. Chem. 2006, 71, 4296.

74

Das, S.; Goswami, R. K. J. Org. Chem. 2014, 79, 9778.

(a) Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M. Bull. Chem. Soc. Jpn. 1979, 52, 1989. (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 710–711.

75

76

Nagasawa, T.; Kuwahara, S. Org. Lett. 2013, 15, 3002.

77

Nagarajan, M.; Kumar, V. S.; Rao, B. V. Tetrahedron 1999, 55, 12349.

78

Masamune, S.; Yamamoto, H.; Kamata, S.; Fukuzawa, A. J. Am. Chem. Soc. 1975, 97, 3513.

79

Masamune, S.; Kamata, S.; Schilling, W. J. Am. Chem. Soc. 1975, 97, 3515.

80

Masamune, S.; Kim, C. U.; Wilson, K. E.; Spessard, G. O.; Georghiou, P. E.; Bates, G. S. J. Am. Chem. Soc. 1975, 97, 3512.

176

4. ACIDS, BASES, AND FUNCTIONAL GROUP EXCHANGE REACTIONS

total synthesis of methymycin, where one step involved treatment of a thioester with mercuric salts. Mercury(II) has an affinity for bivalent sulfur and it binds the thioester unit as well as the terminal hydroxyl group, as in 43. This coordination brings the reactive termini in 43[HO and S(R)C]O] into close proximity, and the hydroxyl oxygen can attack the carbonyl carbon of the thioester to give a cyclized product, lactone 44. Thioester 42 was formed from the corresponding phosphonate ester of the acid. A carboxylic acid is first converted to the phosphonate ester, represented by generic structure 45. Subsequent reaction of 45 with the Tl salt of 2-methyl-2-propanethiol [TlSC(Me)3] gave 46. This generic example shows a mild and general procedure for the preparation of thioesters, and was used to prepare 42. MeO

MeO

MeO

OMe O St-Bu OH O

OMe t-Bu S Hg O O H O C O MeF3C

Hg(O2CCF3)2 25°C , 5 min

O

Me

O

OMe O O Me O

O 42

O 43 (90%)

O

O R

OH

R

O O P OEt OEt

44 (90%) O

Tl+ –St-Bu

R

45

St-Bu 46

Diethyl azodicarboxylate (EtO2CdN¼NdCO2Et, DEAD) is a key reagent in the Mitsunobu reaction (Section 3.2.1.2), which has also been used for macrolactonization.81 A synthetic example is the reaction of 47 with DEAD, and PPh3, which gave a >35% yield of 48 in the Williams and Smith82 synthesis of (+)-18-epi-lacraunculol A. O O

PMB N

S

O O

O

O

PMB N

DEAD , PhMe

HO

PPh3

MeO

S

O O

O

O

MeO O

47

OH

48 (>35%)

A variety of specialized reagents have been developed for macrolactonization reactions. Two commonly used reagents for the cyclization of ω-hydroxy acids, as shown, include the Corey-Nicolaou reagent83 (2,20 -dipyridyl disulfide, 49), and the Mukaiyama reagent84 (2-chloro-1-methylpyridinium iodide, 50). Both reagents tend to yield the lactone in good-to-excellent yield for several ring sizes, and the relative percentage of lactone to diolide seems to improve for larger rings when compared to small rings. Indeed, the use of 49 gave 41% of the diolide for cyclization of a 7-membered ring and 30% of diolide for formation of a 10-membered ring. Similarly the use of 50 led to 93% of the diolide for a 6-membered ring, 34% for a 7-membered ring, and 24% for a 10-membered ring. These results contrasts with formation of a 14-membered ring in which the use of 49 yielded 80% of the lactone and the use of 50 yielded 84% of the lactone. Related reagents include imidazole disulfide (51) [2,20 -dithio-(4-tert-butyl-1-isopropylimidazole)]85 and imidazole 52 [N-(trimethylsilyl)imidazole].86

81

Kurihara, T.; Nakajima, Y.; Mitsunobu, O. Tetrahedron Lett. 1976, 2455.

82

Williams, B. D.; Smith, A. B., III. Org. Lett. 2013, 15, 4584.

(a) Corey, E. J.; Nicolaou, K. C. J. Am. Chem. Soc. 1974, 96, 5614. The actual cyclization has been called the Corey-Nicolaou macrocyclization. (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 184–185.

83

Mukaiyama, T.; Usui, M.; Saigo, K. Chem. Lett. 1976, 49. For a synthetic example taken from a synthesis of the sorangiolides, see Das, S.; Abraham, S.; Sinha, C. Org. Lett. 2007, 9, 2273.

84

85

Corey, E. J.; Brunelle, D. J. Tetrahedron Lett. 1976, 3409.

86

Bates, G. S.; Diakur, J.; Masamune, S. Tetrahedron Lett. 1976, 4423.

177

4.5 MACROCYCLIC COMPOUNDS

t-Bu

I N Me S

N

S

N

49

50

N S S N i-Pr 51

Cl

t-Bu

N

N

N i-Pr

N SiMe 3

52

Diolides are commonly observed during the formation of medium-size lactones, where transannular interactions lead to poor yields. Cyclization to form a 10-membered lactone is difficult, for example, and under high dilution conditions the formation of the 20-membered diolide is usually favored. Indeed, diolides are common contaminants in reactions that form 6 to 11-membered ring lactones.

N

HO

O

CO2H

N

S S

S O

N

N H

HO

N H

O

S

N H

O O

O 54

53

49

S

55

S

+ O

Pyridine-2(1H )-thione

The Corey-Nicolaou procedure reacted 49 with a seco acid to form a thioester (53), which is in equilibrium with the protonated form. Hydrogen bonding leads to a template effect (see 54), that allows nucleophilic acyl attack by the hydroxy group to yield 55. Displacement of thiopyridone [pyridine-2(1H)-thione] gives the lactone. A synthetic example that used 49 converted 56 to 57, in 62% yield, in the Kang et al.87 synthesis of (+)-pamamycin-607. The addition of cupric bromide in a second step facilitated the macrocyclization.88

O

O

HO

O C3H7

2. CuBr2 , MeCN

O

O

1. 49 , PPh3 , MeCN

O O

O

O

O

C3H7

OO

CO2H N3

C3H7

N3 56

C3H7

57 (62%)

0

Another useful reagent is 1,1 -carbonyldiimidazole (59). Hydroxy acid 58 was converted to 60 (40% yield) in the White et al.89 model studies for a synthesis of erythromycin B. This reagent promotes formation of an initial imidazolium intermediate to activate the carbonyl to attack. White et al.89 also reported the cyclization of 58 by treatment with tosyl chloride and triethylamine under dilute conditions (0.01–0.02 M), and obtained a 52% yield of 60.89 TsCl , NEt3 , PhH , 0.01– 0.02 M (52%)

O O

Na tert-amylate PhH (40%)

O N N

CO2H 58

O O

N

HO

O

N O

59

O

O

60

Dibutyltin oxide (n-Bu2SnO) was reported to be a tin-template driven extrusion lactonization reagent.90a,b This reagent effects cyclization in mesitylene, heated at reflux with ω-hydroxy acid, giving 22–63% of the corresponding 13–17-membered ring lactone (0% for n ¼ 7, 63% for n ¼ 14, and 60% for n ¼ 15). The reaction works via formation 87

Kang, S. H.; Jeong, J. W.; Hwang, Y. S.; Lee, S. B. Angew. Chem. Int. Ed. 2002, 41, 1392.

88

Kim, S.; Lee, J. I. J. Org. Chem. 1984, 49, 1712.

89

White, J. D.; Lodwig, S. N.; Trammell, G. L.; Fleming, M. P. Tetrahedron Lett. 1974, 3263.

(a) Steliou, K.; Szczygielska-Nowosielska, A.; Favre, A.; Poupart, M. A.; Hanessian, S. J. Am. Chem. Soc. 1980, 102, 7578. (b) Steliou, K.; Poupart, M. A. Ibid. 1983, 105, 7130. (c) Shanzer, A.; Mayer-Shochet, N.; Frolow, F.; Rabinovich, D. J. Org. Chem. 1981, 46, 4662. (d) Shanzer, A.; Libman, J.; Gottlieb, H. E. Ibid. 1983, 48, 4612. 90

178

4. ACIDS, BASES, AND FUNCTIONAL GROUP EXCHANGE REACTIONS

of tin macrocycle (62), via initial complexation of the Sn reagent to the carbonyl unit and the hydroxyl group (as in 61). Thermally induced loss of water is accompanied by extrusion of n-Bu2SnO to yield the lactone. (CH 2)n

OH (CH 2)n

O O H

n-Bu2SnO

O – H 2O

O H O Sn n-Bu n-Bu

CO2H

(CH 2)n

O – n-Bu2SnO

O

(CH 2)n

O

Sn n-Bu O n-Bu

61

62

The reaction requires stoichiometric amounts of dibutyltin oxide to ensure reasonable reaction times. The reagent does not induce racemization at the stereogenic center bearing the hydroxyl group, so it is amenable to asymmetric macrolactonization. Cyclization of 63 to 64, for example, proceeded in 44% yield and was used in Hanessian and coworkers90 synthesis of ricinelaidic lactone. HO H

CO2H

C6H13

O Bu2SnO , 72 h , Reflux

O C6H13

63

H 64

Trost and Chisholm91 reported another macrolactonization method. Reaction of ω-hydroxy acid 65 with ethoxyethyne and the Ru catalyst shown was followed by treatment with camphorsulfonic acid in dilute solution to yield a 69% yield of 66. This two-stage macrolactonization protocol is effective for generating 14-membered rings and larger, and does not involve treatment with base. O

O OH OH

1. EtOCLCH , Toluene 2% [RuCl 2( p-cymeme)]2

O

2. 10% Camphorsulfonic acid 0.005 M

65

66 (69%)

Interestingly, the yield of macrocyclic product usually depends on the method used. Clearly, there are several reagents that one can use for this reaction, and the choice is usually determined by the ease of reaction and the yield. This finding is illustrated in a synthesis of ()-spinosyn A, reported by Frank and Roush.92 Cyclization of seco-acid 67 with the Mukaiyama reagent (50) or under Yamaguchi conditions gave relatively poor yields of macrocycle 68. Mitsunobu conditions gave better results, using either diethyl azodicarboxylate or diisopropyl azodicarboxylate. Significant amounts of hydrazide were formed using the standard Mitsunobu conditions.92 The use of DIAD for suppression of this side product had been reported in the literature,93 but the yield of 68 was diminished. The use of yet another coupling reagent PyBOP (benzotriazol-1-yloxy)tripyrrolidinophosphonium hexafluorophosphate)94 gave the best results, although the related PyBrOP95 gave significantly poorer results.

91

Trost, B. M.; Chisholm, J. D. Org. Lett. 2002, 4, 3743.

92

Frank, S. A.; Roush, W. R. J. Org. Chem. 2002, 67, 4316.

93

Evans, D. A.; Ratz, A. M.; Huff, B. E.; Sheppard, G. S. J. Am. Chem. Soc. 1995, 117, 3448.

94

Coste, J.; Frerot, E.; Jouin, P.; Castro, B. Tetrahedron Lett. 1991, 32, 1967.

95

Frerot, E.; Coste, J.; Pantaloni, A.; Dufour, M.-N.; Jouin, P. Tetrahedron 1991, 47, 259.

179

4.6 CONCLUSION

CO2H Rhamnose-O

TBSO Br O-Rhamnose

Macrocyclization

O O

TBSO Br

OH 67

68 Reagent

Yield (%)

59, NEt3 2,4,6-Trichlorobenzoyl chloride , DMAP PPh3 , DEAD PPh3 , DIAD PyBOP , DMAP PyBOP , DMAP

33 38 54 48 37 70

4.6 CONCLUSION This chapter has shown three powerful functional group exchange reactions, all involving carbonyl compounds, which concludes the introduction of functional group exchange reactions that included Chapters 2 and 3. Chapter 5 will discuss protecting groups, as a prelude to discussions of synthesis, and it will be seen that most of the reactions used for protecting groups are functional group exchange reactions. Chapters 6 and 7 will discuss the use of oxidation and reduction reactions in organic chemical transformations, which will be followed by hydroboration in Chapter 9. Chapter 10 will discuss methods for controlling stereochemistry. HOMEWORK

1. Give the major product for each of the following transformations involving conversion of a carboxylic acid or a derivative into another acid derivative. N(CH2Ph)2

SiMe3 O

O

SOCl2 , MeOH

H 25°C

Et

O OH

HO

(C)

MeOH , H2SO 4

(B)

O

Et Cl

2 equiv i-PrNH2 , CH2Cl2

Cl

1. DCC , DMAP

1. SOCl2 , PhH

Me (CH 2)9

2. Et2NH , PhH

(F)

MeO

CO2H

2.

H2 N

Ph Me

OMe CH2CH2CO2Et N H

O O MeO O

Heat

O O

(G)

rt

0°C

(D)

O

OMe CO2H

(E)

Ether

HO2C

CO2H

(A)

(COCl) 2 , NEt3

N

O

(H)

O

NH3 , MeOH/THF 70°C , 2 d

180

4. ACIDS, BASES, AND FUNCTIONAL GROUP EXCHANGE REACTIONS

2. Offer a mechanistic explanation for the following transformation: O

O O

H

MeOH , 1 N HCl

H

rt

O

CO2Me

3. Suggest a reasonable mechanism for this reaction. OBn O

OH LiOH , EtOH–H2O

O

OBn

N

N O

OH

4. Explain why these two saponification reactions lead to different products (ring closed vs. ring opened). O O

O

but

2. dil H3O+

O

CO2H

O

1. aq NaOH

1. aq NaOH

OH

2. dil H3O+

O

5. Inspection of Section 4.5.2 suggests that the Mukaiyama reagent gives better yields of lactone with simple ω-hydroxy acids than the Corey-Nicolaou reagent. Offer an explanation that accounts for this observation. 6. Give the major product for each reaction. BnO

Cl

HO

(A)

1. Cl

Me Me TBSO

Me S

O

O

N

Cl Cl

Me

NEt3 , THF , 0°C 2. DMAP , Toluene rt

CO2H

Me CO2Me

O

1. KOH , MeOH 2. 2,4,6-Trichlorobenzoyl chloride DMAP , NEt3

MeO O O

OH

(B) 7. Show a synthesis for the following retrosynthesis: O OH

O (CH 2)10 (CH 2)10

8. The reaction of sodium cyanide and cyclohexanone gives a poor yield of the cyanohydrin. You attend a lecture and learn that silicon compounds (e.g., Me3SiCl) react with methoxide to form Me3SiOMe. In another flash of inspiration, you mix cyclohexanone, KCN, and Me3SiCl. What product do you anticipate will be formed from this reaction? Draw a reaction sequence that leads to your product.

181

4.6 CONCLUSION

9. The reaction of heptan-3-one and butan-1-amine leads to the corresponding imine. Draw it. In order to convert this imine back to heptan-2-one, you heat the imine with aqueous acid. Write out the complete mechanism for conversion of this imine back to heptan-3-one under these conditions. 10. Molecule A (called talaromycin B) is a natural product that contains a ketal unit. Molecule B is a derivative of talaromycin B. Suggest an acyclic precursor that can be transformed to B by a simple chemical reaction. Show that reaction and briefly discuss why B could be formed from this precursor.

O

O O

HO

O

PhH2CO

OCH2Ph

OH

B

A

11. When cyclohexanone is treated with an acid catalyst in the presence of both propane-1,3-diol and octane-1,8-diol, one product predominates via reaction of only one of the diols. Draw this product and briefly discuss your choice (specifically why it works and the other does not). 12. Give the product formed from each of the following reactions: OH

Pyrrolidine , cat H+

Ph

O

(A)

cat

O

(D)

NH2

O O

cat H+

(B)

CHO

HO

Ph

NH

O

OH

(E)

cat H+

O

cat H+

(C)

H+

Ph

O

H 2O cat H+

(F)

13. Laetrile (A) is a compound that was believed to be effective against cancer. What do you think will happen to A when it is ingested and hits the 6 N HCl in your stomach. Draw a mechanism and the products. CO2H O O

HO HO

HO

CN Ph

A

14. In the Bayer-Villiger reaction of nonan-4-one to butyl pentanoate with trifluoroperoxyacetic acid, a large excess of sodium acetate is sometimes added as a buffer. Explain why this is necessary. 15. Briefly explain why saponification of 1,1-dimethylethyl butanoate is much slower than the saponification of methyl butanoate. 16. An attempt to convert ethyl butanoate to isopropyl butanoate was done as follows: Ethyl butanoate was dissolved in methanol and treated with 5 equiv. of isopropyl alcohol (propan-2-ol) and an acid catalyst. When the reaction was analyzed, there was essentially no ethyl butanoate, 68%)

Alcohols can be converted to a tert-butyl ether28 [OdC(CH3)3, dOdt-Bu] by reaction of the alcohol with 2-methylprop-2-ene (isobutylene) and an acid catalyst (H2SO4 or BF3).29 This procedure takes advantage of the facile formation and relative stability of a tertiary cation, which allows subsequent reaction with the alcohol in a SN1 type reaction (Section 3.2.2) to yield the ether. A limitation on this reaction is that other functionality in the molecule must be compatible with acidic conditions. As a protecting group, tert-butyl is stable to pH 1–14, as well as to nucleophiles, organometallics, hydrides, catalytic hydrogenation, oxidations, and dissolving metal reductions. The group can be removed with strong aqueous acid (75%) O 4 Steps

O TiCl4 , CH2Cl2

O

O

0°C – rt

MEMO

OBn OBn 43

HO

OH OH 44 (78%)

37

(a) Corey, E. J.; Trybulski, E. J.; Melvin, L. S., Jr.; Nicolaou, K. C.; Secrist, J. A.; Lett, R.; Sheldrake, P. W.; Falck, J. R.; Brunelle, D. J.; Haslanger, M. F.; Kim, S. F.; Yoo, S. J. Am. Chem. Soc. 1978, 100, 4618, 4620. (b) Corey, E. J.; Hopkins, P. B.; Kim, S.; Yoo, Y.; Nambiar, K. P.; Falck, J. R. Ibid., 1979, 101, 7131.

38

Reference 2(a), pp 19, 296–298; Reference 2(b), pp 27–29; Reference 2(c), pp 41–44; Reference 2(d), pp 49–53; Reference 2(e), pp 57–60.

39

Corey, E. J.; Gras, J.-L.; Ulrich, P. Tetrahedron Lett. 1976, 809.

40

Gupta, P.; Kumar, P. Eur. J. Org. Chem. 2008, 1195.

195

5.3 PROTECTING GROUPS FOR ALCOHOLS, CARBONYLS, AND AMINES

A variation of the OMEM group is the 2-(trimethylsilyl)ethoxymethyl ether (OdCH2OCH2CH2dSiMe3, OSEM),41 developed by Lipshutz et al.42 The trimethylsilyl group was incorporated so it could be used as a trigger for the facile deprotection of the alcohol. The OSEM group is attached to an alcohol by reaction with 2-(trimethylsilyl) unig’s base. The SEM group is stable in a pH range of ethoxymethyl chloride (Me3SiCH2CH2OCH2Cl) and H€ 4–12. Some nucleophiles and many Lewis acids react with this group, but it is generally stable to hydrides and to many organometallics (e.g., butyllithium), as well as most oxidizing agents. The presence of silicon makes the group sensitive to fluoride (as with silyl protecting groups, see Section 5.3.1.2), and treatment with tetrabutylammonium fluoride (n-Bu4N+F, TBAF) usually regenerates the alcohol, but not always. Indeed, the OSEM group is difficult to remove in some cases. However, in the Choshi, Hibino and coworker’s43 synthesis of carbazomadurin A, the phenolic OH units in 45 were protected as the bis(SEM) derivative (46) in 98% yield. This fragment was coupled to a vinyl borinate via Pd(0) coupling (see Section 18.4.3) to yield 47. Final treatment with TBAF removed both SEM groups to yield 48 in 71%. CHO

CHO

OH Cl

CH3 HO

N H

O

OSEM

SiMe3

O

CH3

i-Pr2NEt , CH2Cl2

OTf

O

SEMO

45

N H

B

Pd(PPh3)4 , DMF Na2CO3 , 3 M aq

OTf

46 (98%) CHO

CHO

OSEM CH3

SEMO

OH CH3

TBAF , HMPA

N H

N H

HO

100°C

48 (71%)

47 (86%)

Tetrahydropyranyl ethers (OTHP, see 49)44 are acetal-protecting groups that can be found in many older organic syntheses. An alcohol reacts with dihydropyran (3,4-dihydro-2H-pyran) in the presence of an acid (e.g., p-toluenesulfonic acid) to yield the THP derivative (49) via an oxonium ion.45 One problem with this protecting group is the fact that a new stereogenic center is created when 49 is formed. If ROH is chiral, 49 will exist as diastereomers, which may make identification, isolation, and purification difficult if this protecting group is carried through many synthetic steps. This group is stable to base (pH 6–12), but unstable to aqueous acid and to Lewis acids. It is relatively stable to nucleophiles and organometallics. It is also stable to hydrogenation, hydrides, and oxidizing agents.45 It is removed with aqueous acid or methanolic tosic acid.46 Greene2a cited a reference reporting that explosions have occurred when THP ethers are treated with diborane and then basic H2O2,47 or with 40% peroxyacetic acid.46 – H+

ROH , H+

O

O

O

O

R

H

3,4-Dihydro-2H-pyran

41

O

O

R

49

Reference 2(a), pp 20–21, 296–298; Reference 2(b), pp 30–31; Reference 2(c), pp 45–48; Reference 2(d), pp 54–58; Reference 2(e), pp 63–67.

(a) Lipshutz, B. H.; Pegram, J. J. Tetrahedron Lett. 1980, 21, 3343. (b) Lipshutz, B. H.; Moretti, R.; Crow, R. Ibid., 1989, 30, 15. (c) Lipshutz, B. H.; Miller, T. A. Ibid., 1989, 30, 7149.

42

43

Hieda, Y.; Choshi, T.; Fujioka, H.; Hibino, S. Eur. J. Org. Chem. 2013, 7391.

44

Reference 2(a), pp 21–22, 296–298; Reference 2(b), pp 31–34; Reference 2(c), pp 49–52; Reference 2(d), pp 59–68.; Reference 2(e), pp 69–80

This report is taken from (a) Bernady, K. F.; Floyd, M. B.; Poletto, J. F.; Weiss, M. J. J. Org. Chem. 1979, 44, 1438. (b) Miyashita, M.; Yoshikoshi, A.; Grieco, P. A. Ibid. 1977, 42, 3772. 45

46

Corey, E. J.; Niwa, H.; Knolle, J. J. Am. Chem. Soc. 1978, 100, 1942.

47

See Meyers, A. I.; Schwartzman, S.; Olson, G. L.; Cheung, H.-C. Tetrahedron Lett. 1976, 2417.

196

5. FUNCTIONAL GROUP EXCHANGE REACTIONS: PROTECTING GROUPS

An example involving this protecting group is taken from Stecko’s48 synthesis of lacosamide. The alcohol unit in ethyl (S)-2-hydroxypropanoate was protected as the THP derivative in 50 in 75% yield. The conversion to 51 required four steps, including conversion of a primary alcohol unit to a methyl ether. Final treatment with methanolic acetyl chloride liberated the secondary hydroxyl group in (S,E)-5-methoxypent-3-en-2-ol in 78% yield. , p-TsOH

OH

O

CO2Et

CH2Cl2 . –78°C

Ethyl (S)-2-hydroxypropanoate

OTHP CO2Et

1. dibal , CH2Cl2 , –78 °C 2. NaH , (EtO) 2P(O)CH2CO2Et 3. dibal , CH2Cl2 , –78 °C 4. NaH , MeI , THF

50 (75%) OTHP

5% AcCl , MeOH

OH

rt

OMe 51

OMe (S, E)-5-Methoxypent-3-en-2-ol (78%)

Wuts and Greene49 classified another generic type of acetal protecting group as a substituted ethyl ether. A common member of this group is 1-ethoxyethyl ether [OdCH(OCH2Me)Me, OEE]. As with the OTHP group, diastereomers can be generated when the OEE group is incorporated in a molecule with the same problems. The protected derivative will exist as diastereomers, since there are two stereogenic centers. The group is usually attached to the alcohol via reaction with ethyl vinyl ether under acid conditions (HCl or tosic acid).50 The group is stable to a relatively narrow pH range of 6–12, but it is compatible with reactions of nucleophiles, organometallics, hydrogenation, hydrides, and most oxidants. It is sensitive to acid conditions and to Lewis acids. The most common methods for cleavage to the alcohol are treatment of the OEE ether with aq acetic acid,50a or with aq HCl in THF.50b Nagumo and coworkers51 used this protecting group in a synthesis of the macrolide sekothrixide. 5.3.1.2 Silyl Ether Protecting Groups Silyl ethers, with the generic structure OSiR3, have become extremely important for the protection of alcohols. Variations in the R group lead to significant differences in the stability of the protecting group. In general, silyl ethers can be cleaved with aqueous base or acid, but the rate of hydrolysis for a secondary silyl ether is significantly slower than that of a primary silyl ether.52 In addition, alkyl alcohols are more reactive than phenolic derivatives, and the order of reactivity for other silane-protected functional groups is COOSiR3 > NHSiR3 > CONHSiR3 > SSiR3.53 Steric factors play a major role in the relative stability of the OSiR3 group in cleavage reactions. The prototype of this class of protecting groups is the trimethylsilyl ether [OdSi(Me)3, OTMS].54 The group is usually attached to the alcohol by reaction with chlorotrimethylsilane (Me3SiCl) in the presence of an amine base (e.g., triethylamine or pyridine),55 although Wuts and Greene2 list several other methods. This group is sensitive to many aqueous conditions, and the TMS group is sometimes cleaved in an aqueous workup procedure.56 The group is not very stable to organometallics (e.g., Grignard reagents), or to nucleophiles, hydrogenation, or hydrides, but it can be used with some oxidants. The sensitivity of OTMS to these reagents is often the result of the aqueous or acid workup procedures that are required for a given reaction. If strictly anhydrous conditions are used, the TMS group can be used for short-term protection.

48

Stecko, S. J. Org. Chem. 2014, 79, 6342.

49

Reference 2(a), pp 25, 296–298; Reference 2(b), pp 414–415; Reference 2(c), pp 709–711; Reference 2(d), pp 74–75; Reference 2(e), pp 87–89.

(a) Chládek, S.; Smrt, J. Chem. Ind. (London) 1964, 1719. (b) Meyers, A. I.; Comins, D. L.; Roland, D. M.; Henning, R.; Shimizu, K. J. Am. Chem. Soc. 1979, 101, 7104. 50

51

Terayama, N.; Yasui, E.; Mizukami, M.; Miyashita, M.; Nagumo, S. Org. Lett. 2014, 16, 2794.

52

(a) Reference 2(a), p 39. (b) McInnes, A. G. Can. J. Chem. 1965, 43, 1998.

53

(a) Reference 52(a). (b) Cooper, B. E. Chem. Ind. (London) 1978, 794.

54

Reference 2(a), pp 40–42; Reference 2(b), pp 68–73; Reference 2(c), pp 116–121; Reference 2(d), pp 171–178; Reference 2(e), pp 208–217.

55

(a) Corey, E. J.; Snider, B. B. J. Am. Chem. Soc. 1972, 94, 2549. (b) Sweely, C. C.; Bentley, R.; Makita, M.; Wells, W. W. Ibid., 1963, 85, 2497.

56

Reference 2(a), pp 296–298; Reference 2(b), pp 414–415; Reference 2(c), pp 709–711; Reference 2(d), pp 171–178; Reference 2(e), pp 208–217.

197

5.3 PROTECTING GROUPS FOR ALCOHOLS, CARBONYLS, AND AMINES

Under aprotic conditions, fluoride ion (e.g., tetrabutylammonium fluoride) readily attacks silicon and cleaves the OdSi bond.52a A synthetic example is the protection of the alcohol moiety in 52, using trimethylsilyl triflate (Tf ¼ dSO2CF3) and 2,6-lutidine to yield 53 in 78% yield, in Yang and coworker’s57 synthesis of pseudolaric acid A. This group was sufficiently stable to allow the next nine steps to yield 54, which also contained an OMEM ester. Treatment with tetrabutylammonium fluoride removed the OTMS group to yield 55, and hydrolysis of the OMEM ester gave pseudolaric acid A in 80% yield from 55. MeO2C

MeO2C

TMSOTf , 2,6-Lutidine CH2Cl2

MeO2C

MeO2C

Me H

Me

Me

OH

H

Me

OSiMe3 53 (78%)

52 O

O

9 Steps TBAF , THF

O MEMO2C Me

O H

OSiMe3

MEMO2C Me

54

OH

H 55

In general, the hydrolysis of the OTMS group is so facile that it is unsuitable for most alcohol moieties that require long-term protection. A more stable silyl derivative is the triethylsilyl group (OdSiEt3, OTES).58 This group is attached by reaction of the alcohol with chlorotriethylsilane (Et3SiCl) and Py.59 The OTES group is generally more stable than OTMS, and in particular it is less sensitive to water. Aqueous acetic acid cleaves OTES and also alcoholic tosic acid,54 but fluoride can also be used for removing it. In a synthesis of (+)-heronapyrrole C by Brimble and coworkers,60 the two alcohol fragments in 56 were converted to the bis-OTES derivative 57, in 85% yield, using triethylsilyl triflate and imidazole. After five synthetic steps, compound 58 was obtained. Treatment with TBAF in THF deprotected the OTES alcohol unit to yield 59, and treatment with camphorsulfonic acid deprotected the N-Boz unit, where Boz ¼ benzyloxymethyl, to give (+)-heronapyrrole C in 80% yield from 58. Me H O

Me

Me Me

OH

Et3SiOTf, Imidazole

Me

H O

Me

OSiEt3

5 Steps

DMF , 0°C

OSiEt3

OH 56

57 (85%) Me

O

Me

H O

Me Me

OSiEt3 TBAF , THF

O

Me

H O

Me OH

0°C – rt

O2N

O2N

OSiEt3

N

N

OH

Boz

Boz 58

N-Boz = N-benzoyloxymethyl

59 (>80%)

57

Xu, T.; Li, C.-C.; Yang, Z. Org. Lett. 2011, 13, 2630.

58

Reference 2(a), p 44; Reference 2(b), pp 72–73; Reference 2(c), pp 121–123; 709–711; Reference 2(d), pp 178–183; Reference 2(e), pp 218–224.

59

Hart, T. W.; Metcalfe, D. A.; Scheinmann, F. J. Chem. Soc. Chem. Commun. 1979, 156.

60

Ding, X.-B.; Furkert, D. P.; Capon, R. J.; Brimble, M. A. Org. Lett. 2014, 16, 378.

198

5. FUNCTIONAL GROUP EXCHANGE REACTIONS: PROTECTING GROUPS

A triisopropylsilyl ether [OSi(CH(Me)2)3, OTIPS],61 is prepared via the reaction of the alcohol with triisopropylchlorosilane (i-Pr3SiCl)62 and imidazole. The group is stable in the pH 2–14 range and it is generally impervious to nucleophiles, organometallics, hydrogenation (except in acetic acid), hydride, and oxidizing agents. It is usually removed with fluoride (TBAF or aq HF). In an example taken from the Cook and coworker’s63 synthesis of ()-macroine, the primary alcohol unit in diol 60 was selectively protected with triisopropylsilyl chloride in the presence of 2,6luitidine to yield the OTIPS derivative 61 in 90% yield. Four synthetic steps gave the OTIPS protected intermediate 62 and deprotection of the OTIPS groups with TBAF gave 63 in 86% yield. H

H

N Me

N

TIPSOTf, 2,6-Lutidine

N

CH2Cl2

H

H

H

OH

Me

H 60

OSi(i-Pr)3

N

4 Steps

H H

61 (90%) H Me N Me

H

H

H OSi(i-Pr)3

N

H OH

TBAF , THF reflux

Me N Me

H

N

H

62

H O

O 63 (86%)

One of the most used silyl ether protecting groups is tert-butyldimethylsilyl [OdSi(Me)2t-Bu, abbreviated as OTBDMS, although it has been abbreviated as OTBS].64 The most common method for attaching this group reacts the alcohol with tert-butyldimethylsilyl chloride using imidazole65a or dimethylaminopyridine65b as the base. The analogous silyl triflate (t-BuMe2SiOSO2CF3) is also useful for generating this ether from alcohols, with an amine base.65c The group is stable to base and reasonably stable to acid (pH 4–12) relative to the other silyl protecting groups, and also to nucleophiles, organometallics, hydrogenation (except in acidic media), hydrides, or oxidizing agents. It can be removed with aqueous acid, but fluoride ion is the most common cleavage method (tetrabutylammonium fluoride,65 tetrabutylammonium chloride, and KF,66 as well as aq HF67). Fukuyama and coworker’s68 used this protecting group in a synthesis of ()-lepenine, for the alcohol unit in 64. Conversion to the tert-butyldimethylsilyl ether by reaction the silyl triflate gave 65 in 91% yield. Six synthetic steps were required to synthesize 66, and removal of the TBS group with TBAF in THF gave 67 in 93% yield.

61

Reference 2(a), pp 50, 296–298; Reference 2(b), pp 74–75; Reference 2(c), pp 125–127; 709–711; Reference 2(d), pp 183–187; Reference 2(e), pp 225–229. 62

Ogilvie, K. K.; Thompson, E. A.; Quilliam, M. A.; Westmore, J. B. Tetrahedron Lett. 1974, 2865.

63

Liao, X.; Zhou, H.; Yu, J.; Cook, J. M. J. Org. Chem. 2006, 71, 8884.

64

Reference 2(a), pp 44–46, 296–298; Reference 2(b), pp 77–83; Reference 2(c), pp 127–141; 709–711; Reference 2(d), pp 189–211; Reference 2(e), pp 231–256. (a) Corey, E. J.; Venkateswarlu, A. J. Am. Chem. Soc. 1972, 94, 6190. (b) Chaudhary, S. K.; Hernandez, O. Tetrahedron Lett. 1979, 99. (c) Corey, E. J.; Cho, H.; R€ ucker, C.; Hua, D. H. Ibid., 1981, 22, 3455.

65

66

Carpino, L. A.; Sau, A. C. J. Chem. Soc. Chem. Commun. 1979, 514.

67

Newton, R. F.; Reynolds, D. P.; Finch, M. A. W.; Kelly, D. R.; Roberts, S. M. Tetrahedron Lett. 1979, 3981.

68

Nishiyama, Y.; Han-ya, Y.; Yokoshima, S.; Fukuyama, T. J. Am. Chem. Soc. 2014, 136, 6598.

199

5.3 PROTECTING GROUPS FOR ALCOHOLS, CARBONYLS, AND AMINES

OMe OH

OMe O

H

OMe OTBS t-BuMe2SiOTf 2,6-Lutidine

O

H

CH2Cl2

Me

OMe

Me

N

N

Et

Et

64

65 (91%) HO OTBS H

6 Steps

OH

H Me

HO OH H

TBAF , THF 65 °C

OH

H Me

N Et

N Et 67 (93%)

66

A protecting group that is used with great regularity is the tert-butyldiphenylsilyl group [O-SiPh2C(Me)3, OTBDPS].69 Hanessian and Lavallee70 developed this protecting group when it was discovered that the OTBDMS group was sensitive to hydrogenation. The TBDPS group is attached to the alcohol by reaction with tert-butyldiphenylsilyl chloride (t-BuPh2SiCl) and imidazole.70 It has stability characteristics similar to dimethyl-tert-butyl derivatives, but is more stable to Lewis acids. An example is the synthesis of ()-crinipellin A by Lee and coworkers,71 in which alcohol 68 was converted to TBDPS derivative 69 in 96% yield. The conversion to 70 required 10 steps and deprotection with TBAF gave 71 in >80% yield. MeO OH

MeO

OMe OTBDPS

TBDPSCl , imidazole

•

OMe

•

DMAP , CH2Cl2 , rt

68

69 (96%)

H 10 Steps

H TBAF , THF 60 °C

OTBDPS

OH

70

71 (>80%)

5.3.1.3 Ester Protecting Groups Esters, O-esters (e.g., OdCOR), constitute another class of alcohol protecting groups where COR is an acyl group. Ester protecting groups are limited in scope due to their susceptibility to nucleophilic acyl substitution, hydrolysis, and reduction. Several esters are commonly used in synthesis, however, including acetates, benzoates, or mesitoate esters. Although many esters may be used in synthesis, this discussion will focus only on the more common derivatives.

69

Reference 2(a), pp 47–48, 296–298; Reference 2(b), pp 83–84; Reference 2(c), pp 141–145; 709–711; Reference 2(d), pp 211–215; Reference 2(e), pp 257–262. 70

Hanessian, S.; Lavallee, P. Can. J. Chem. 1975, 53, 2975. (b) Idem Ibid., 1977, 55, 562.

71

Kang, T.; Song, S. B.; Kim, W.-Y.; Kim, B. G.; Hee-Yoon Lee, H.-Y. J. Am. Chem. Soc. 2014, 136, 10274.

200

5. FUNCTIONAL GROUP EXCHANGE REACTIONS: PROTECTING GROUPS

Acetates (OAc)72 are formed by reaction of an alcohol with acetic anhydride or acetyl chloride and pyridine or triethylamine. Acetates are stable to pH from 1 up to 8, and to ylids as well as organocuprates, catalytic hydrogenation, borohydrides, Lewis acids, and oxidizing agents. Hydrolysis with acid or base (saponification) cleaves the ester to the alcohol and an acid. Reaction with LiAlH4 reductively cleaves the acetate to the alcohol and ethanol. Acetates have found widespread use in synthesis. In the synthesis of ()-isoavenaciolide by Ariza and coworkers,73 diol 72 was acetylated to yield 73 in 97% yield. Only short-term protection (two steps) was required to generate 74, and deprotection with aq H2SO4 liberated the diol units, which cyclized with the two carboxylic acid moieties to yield bis-lactone 75 in 71% yield. OH

OAc

OH S

C8 H17

Ac2O

OAc S

C8 H17

73 (97%)

72

OAc 2 Steps

O

OAc aq H2SO 4

C8 H17

CO2H CO2H

Dioxane

O H

C8 H17

74

H O

O

75 (71%)

Alcohols react with benzoyl chloride (PhCOCl) or benzoic anhydride [(PhCO)2O], with Py or NEt3, to yield the benzoyl derivative (benzoate ester, OBz).74 Benzoates are more stable to hydrolysis than acetates, with stability from pH 1 up to 10–11. Treatment with 10% K2CO3 (pH 11), for example, will hydrolyze the OBz group. Benzoyl esters are more stable to reactions with nucleophiles (e.g., cyanide and acetate, ylids, or organocuprates). Benzoate esters also resist catalytic hydrogenation, as well as reaction with borohydrides or with oxidizing agents. The group is usually cleaved by basic hydrolysis or by reduction with LiAlH4. Three additional protecting groups in this category are the mesitoate ester (OCOdC6H2-2,4,6-trimethyl, OMes),75 the pivaloyl ester (OCOdt-Bu, OPiv)76,77 and the p-methoxybenzoyl ester [OCOdC6H4d(4-Me)]. The mesitoate ester is formed by reaction of an alcohol with mesitoyl chloride in the presence of pyridine or triethylamine.78 This ester is stable to hydrolysis (pH 1–12), nucleophilic attack, organometallics, catalytic hydrogenation; reaction with boranes, borohydrides, Lewis acids, or oxidation. This stability is due to the two ortho methyl groups, which block attack to the acyl carbon. Both LiAlH473 and concentrated alcoholic potassium tert-butoxide at,73,79 will cleave the group. Cleavage of the CdO bond is occasionally a problem with LiAlH4.80 Similar reactivity is observed with OPiv, but the precursor is pivaloyl chloride, t-BuCOCl.81

72 Reference 2(a), pp 53–55, 300–302; Reference 2(b), pp 88–92; Reference 2(c), pp 150–160; 713–716; Reference 2(d), pp 223–239; Reference 2(e), pp 273–297. 73

Santos, D.; Ariza, X.; Garcia, J.; Lloyd-Williams, P.; Martínez-Laporta, A.; Sánchez, C. J. Org. Chem. 2013, 78, 1519.

74

Reference 2(a), pp 61–62, 300–302; Reference 2(b), pp 100–103; Reference 2(c), pp 173–178; 713–716; Reference 2(d), pp 255–262; Reference 2(e), pp 315–325.

75

Reference 2(a), pp 63–64, 300–302; Reference 2(b), pp 103–104; Reference 2(c), pp 178–179; 713–716; Reference 2(d), p 263; Reference 2(e), p 325.

76

(a) Reference 2(c), pp 170–173; 713–716. (b) Reference 2(d), pp 250–254; Reference 2(e), pp 310–314.

77

For a synthetic example, taken from a synthesis of elutherobin, see Chen, X.-T.; Bhattacharya, S. K.; Zhou, B.; Gutteridge, C. E.; Pettus, T. R. R.; Danishefsky, S. J. J. Am. Chem. Soc. 1999, 121, 6563.

78

Corey, E. J.; Achiwa, K.; Katzenellenbogen, J. A. J. Am. Chem. Soc. 1969, 91, 4318.

79

(a) Reference 73. (b) Gassman, P. G.; Schenk, W. N. J. Org. Chem. 1977, 42, 918.

80

For cleavage of esters, see Robins, D. J.; Sakdarat, S. J. Chem. Soc. Perkin Trans. 1981, 1, 909.

81

For a synthesis of ophiobolin A that used pivaloyl protecting groups, see Tsuna, K.; Noguchi, N.; Nakada, M. Chem. – Eur. J. 2013, 19, 5476.

201

5.3 PROTECTING GROUPS FOR ALCOHOLS, CARBONYLS, AND AMINES

5.3.2 Protection of Diols 1,2-Diols are obviously alcohols, but the vicinal nature of the hydroxyls allows them to be protected as cyclic ketals. When a 1,2-diol (e.g., butane-2,3-diol) reacts with a ketone (e.g., acetone in the presence of an acid catalyst), a 1,3dioxolane (2,2,4,5-tetramethyl-1,3-dioxolane) is formed. Ketals (e.g., 2,2,4,5-tetramethyl-1,3-dioxolane) that are derived from acetone are called acetonides (isopropylidene ketals).82 A 1,3-diol will generate a six-membered ring acetonide, which is a 1,3-dioxane derivative. Me

Me

OH

Me

O

Me

O

Acetone cat H+

OH

Butane-2,3-diol

Me Me

2,2,4,5-Tetramethyl-1,3-dioxolane

Common methods for acetonide formation are reaction of the diol with 2-methoxy-1-propene in the presence of an acid (e.g., anhydrous HBr)83 or reaction of acetone with an acid catalyst.84 The group is stable to base, but not to acid (pH 4–12). It is stable to nucleophiles, organometallics, catalytic hydrogenation, hydrides, and oxidizing agents. It can be cleaved with aq HCl, with acetic acid,85 or with p-toluenesulfonic acid in methanol.86 This group has been used extensively in the manipulation of carbohydrates. Acetonides are useful in other synthetic applications, as in the protection of the 1,2-diol unit found in 76, in Ostermeier’s and Schobert87 synthesis of (+)-chloriolide, which was converted to 77 in 90% yield by reaction with dimethoxypropane and tosic acid in acetone. Intermediate 78 was prepared in 11 synthetic steps, and then treated with HCl in THF to yield diol 79. This solution of the hemiacetalphosphonium salt was layered with CH2Cl2 and neutralized with NaOH to generate the ylid (see Section 12.5.1.1), which underwent spontaneous Wittig cyclization to (+)-chloriolide (80) in 65% yield. PPh3

O OH HO O

OH

Me2C(OMe)2 TsOH , Acetone, rt

OH

O

11 Steps

O O

O O O

76

77 (90%)

78 O

PPh3

O HCl/THF

O

O

NaOH

O

HO

rt , 20 h

HO O

OH OH

79

80 (65%)

Diols react with other aldehydes or ketones to yield a wide array of ketals or acetals. Two of the more common derivatives are benzaldehyde (to yield benzylidene acetals) and cyclohexanone (to yield cyclohexylidene ketals). In a synthesis of ()-dinemasone B by Meng and coworkers,88 triol 81 was converted to the benzylidene acetal (82) via an acid-catalyzed reaction with benzaldehyde dimethyl acetal, and then converted to 83 in four synthetic steps. The diol was deprotected by reduction with borane, in this case removing the protecting group and generating 84, 82

Reference 2(a), pp 76–78, 304–306; Reference 2(b), pp 123–127; Reference 2(c), pp 207–215; 717–719; Reference 2(d), pp 306–318; Reference 2(e), pp 394–410.

83

Corey, E. J.; Kim, S.; Yoo, S.; Nicolaou, K. C.; Melvin, L. S., Jr.; Brunelle, D. J.; Falck, J. R.; Trybulski, E. J.; Lett, R.; Sheldrake, P. W. J. Am. Chem. Soc. 1978, 100, 4620.

84

(a) Schmidt, O. Th. Methods Carbohydr. Chem. 1963, 2, 318. (b) de Belder, A. N. Adv. Carbohydr. Chem. 1965, 20, 219.

85

Lewbart, M. L.; Schneider, J. J. J. Org. Chem. 1969, 34, 3505.

86

(a) Ichihara, A.; Ubukata, M.; Sakamura, S. Tetrahedron Lett. 1977, 3473. (b) Kimura, J.; Mitsonobu, O. Bull. Chem. Soc. Jpn. 1978, 51, 1903.

87

Ostermeier, M.; Schobert, R. J. Org. Chem. 2014, 79, 4038.

88

Xue, X.; Yin, Z.; Meng, X.; Li, Z. J. Org. Chem. 2013, 78, 9354.

202

5. FUNCTIONAL GROUP EXCHANGE REACTIONS: PROTECTING GROUPS

in 94% yield, as the O-benzyl derivative. A more common method for deprotection uses hydrogen with a Pd/C catalyst, via hydrogenolysis of the benzylic position (Section 7.10.6). O TBDPSO

HO

OH

O

PhCH(OMe)2 p-TsOH , MeCN

TBDPSO

rt , 8 h

OH

O 4 Steps

HO

O

OBn

Ph

OBn

81

82 (89%) H

Me

O

H

O O

H

H OBn 83

O O

H

Me

O

H OH

BH3•THF , rt

Ph

Bu2BOTf , 2 h CH2Cl2

O O

H

H OBn

O

Ph

84 (94%)

Wuts and Greene2 discuss many other ketal protecting groups that have been used in synthesis, but acetonide is probably the most common. Cyclopentylidene or cyclohexylidene89 and methylene90 have been used, as well as a benzophenone ketal.91 Discussions of methods for protecting alcohols in this Section have been extensive, due to the wide range of alcohol protecting groups. Since the hydroxyl is capable of being transformed into a wide range of other functional groups, its protection is of particular importance. More than one hydroxyl moiety is commonly incorporated into a molecule, so the requirement for several selective protecting groups is obvious. The following Sections will deal with protection of the other two key functional groups, the carbonyl group found in ketones and aldehydes, and also the nitrogen found in amines.

5.3.3 Protection of Aldehydes and Ketones As with alcohols, developing methods for the protection of ketones or aldehydes must begin with understanding the interfering portion of these molecules. The main reaction of a carbonyl is nucleophilic acyl addition. The oxygen of the carbonyl can also function as a base in the presence of a suitable acid, leading to an oxocarbenium ion (Sections 4.2.2 and 16.2.1) and a number of reactions. To protect against these reactions, the electrophilic carbon and the π-bond of the carbonyl must be removed by conversion to a new functional group that is easily converted back to the carbonyl. One method is to reduce the C]O unit to an alcohol (see Sections 7.4 and 7.6.1, and then protect the OH unit as described in Section 5.3.1. When it is appropriate in the synthesis, the alcohol is deprotected, and then oxidized back to the aldehyde or ketone. Such a protocol requires four synthetic steps. When feasible, protection procedures should require only two steps: protection and deprotection. With this limitation in mind, the major method for protecting ketones and aldehydes is conversion to a ketal or acetal, using alcohol or diol reactants. A variation of this reaction uses thiol or dithiol reactants to generate dithioketals or dithioacetals. In both cases, treatment with acid (or Lewis acids with the dithio derivatives) readily converts the ketals or acetals back to the ketone or aldehyde. It is important to note that in previous Sections, alcohols were protected as ketals or acetals by reaction with ketones or aldehydes, but in this Section the carbonyl group will be protected as the ketal or acetal by reaction with alcohols. The group is the same. The difference is which functional group is the object of attention and isolation. 5.3.3.1 Ketals and Acetals Ketals and acetals are formed by reaction of the carbonyl with alcohols (e.g., methanol or ethanol) under anhydrous conditions, in the presence of an acid catalyst. It is obvious that many alcohols can be used to generate acetals and ketals, but methanol and ethanol are probably the most common ones used. This choice is determined by the fact that the yields of product with these reagents are high, and the lower molecular weight alcohol byproducts are easily removed after deprotection. Another attractive feature of the methoxy unit is that it shows simple singlets in the 1 H NMR spectrum that usually do not obscure important signals from the remainder of protected substrate. In general, 89

Reference 2(d), p 318; Reference 2(e), pp 410–414.

90

Reference 2(d), pp 300–302; Reference 2(e), pp 385–388.

91

Reference 2(d), p 344; Reference 2(e), pp 444–445.

203

5.3 PROTECTING GROUPS FOR ALCOHOLS, CARBONYLS, AND AMINES

dry HCl (g), sulfuric acid,92 BF3OEt2 or p-toluenesulfonic acid (PPTS) are the catalysts of choice. The addition of molecular sieves to adsorb water usually improves the yield of ketal or acetal.93 Diethyl and dimethyl ketals and acetals are generally stable in a pH range of 4–12, but they are sensitive to strong aqueous acid and Lewis acids. They are stable to nucleophiles, organometallics, catalytic hydrogenation, hydrides, and most oxidizing agents (they do react with ozone).94 Conversion of the ketal or acetal back to the ketone or aldehyde is accomplished by treatment with aqueous acids [e.g., trifluoroacetic acid95 or oxalic acid (HOOCdCOOH)].96 A popular alternative for regenerating the carbonyl from a ketal or acetal is transketalization, whereby acetone is added along with PPTS (or another acid catalyst). Acetone generates a new ketal (2,2-dimethoxypropane) in this reaction, liberating the originally protected carbonyl derivative.97 Other acetals or ketals can be used as well. EEO

OMe

CHO

MeOH , amberlyst 15E

HO

85

OMe 86

In the Kuwahara et al.98 synthesis of a sesquiterpene isolated from the pheromone gland of a stink bug, Tynacantha marginata Dallas, conjugated aldehyde 85 was treated with methanol in the presence of an acidic resin to yield the dimethyl acetal 86. Under these reaction conditions, the OdEE unit was deprotected to yield the alcohol. Diethoxy acetals are also quite useful. A synthetic example using a diethyl acetal is taken from Tanino and coworkers99 synthesis of ()-coriolin. Formation of the anion derived from 2-methylpropanenitrile (isobutyronitrile) and reaction with the diethyl acetal of 2-bromoacetaldehyde (2-bromo-1,1-diethoxyethane) gave 4,4-diethoxy-2,2-dimethylbutanenitrile. The nitrile unit was elaborated to the methylthio-ketone unit in 5,5-diethoxy-3,3-dimethyl-1-(methylthio)pentan-2one, and treatment with sulfuric acid regenerated the aldehyde, 3,3-dimethyl-5-(methylthio)-4-oxopentanal. CN CN

LDA , HMPA Br

Isobutyronitrile

1. MeSCH2Li 2. AcOH

OEt

OEt OEt

2-Bromo-1,1-diethoxyethane

OEt 4,4-Diethoxy-2,2-dimethylbutanenitrile O O

SMe OEt

H2SO 4

SMe CHO

OEt 5,5-Diethoxy-3,3-dimethyl-1(methylthio)pentan-2-one

3,3-Dimethyl-5-(methylthio)4-oxopentanal

The 1,2- and 1,3-diols, which form cyclic ketals or acetals [1,3-dioxolanes or 1,3-dioxanes as discussed for protection of diols, see above], are particularly important for the protection of aldehydes and ketones. 1,3-Dioxolane derivatives are formed by the reaction of the carbonyl with ethane-1,2-diol, whereas the reaction with propane-1,3-diol yields 1,3dioxanes. Dioxane protected ketones are usually hydrolyzed back to the carbonyl faster than the analogous dioxolane protected derivatives, but dioxolane protected aldehydes are hydrolyzed faster than dioxane protected.100 Despite this difference in reactivity, dioxolanes appear to be the most commonly used protecting group. A carbonyl group reacts with ethane-1,2-diol (ethylene glycol) and PPTS, BF3OEt2, or oxalic acid to yield the ethylenedioxy ketal or 92

Cameron A. F. B.; Hunt, J. S.; Oughton, J. F.; Wilkinson, P. A.; Wilson, B. M. J. Chem. Soc. 1953, 3864.

93

Roelofsen, D. P.; Wils, E. R. J.; Van Bekkum H. Recl. Trav. Chim. Pays-Bas 1971, 90, 1141.

94

Reference 2(a), pp 116–120, 312–314; Reference 2(b), pp 178–183; Reference 2(c), pp 297–304; 725–727; Reference 2(d), pp 435–444, 1009–1016; Reference 2(e), pp 554–685.

95

Ellison, R. A.; Lukenbach, E. R.; Chiu, C.-W. Tetrahedron Lett. 1975, 499.

96

Huet, F.; Lechevallier, A.; Pellet, M.; Conia, J. M. Synthesis 1978, 63.

97

Colvin, E. W.; Raphael, R. A.; Roberts, J. S. J. Chem. Soc. Chem. Commun. 1971, 858.

98

Kuwahara, S.; Hamade, S.; Leal, W. S.; Ishikawa, J.; Kodama, O. Tetrahedron 2000, 56, 8111.

99

Mizuno, H.; Domon, K.; Masuya, K.; Tanino, K.; Kuwajima, I.. J. Org. Chem. 1999, 64, 2648.

100

(a) Newman, M. S.; Harper, R. J. J. Am. Chem. Soc. 1958, 80, 6350. (b) Smith, S. W.; Newman, M. S. Ibid. 1968, 90, 1249, 1253.

204

5. FUNCTIONAL GROUP EXCHANGE REACTIONS: PROTECTING GROUPS

acetal (a 1,3-doxolane).101 Reaction of the carbonyl with propane-1,3-diol (propylene glycol) under similar conditions generates the 1,3-dioxane, a propylenedioxy ketal or acetal.102 Cyclic ketals are sensitive to acid and are generally stable to pH 4–12 and to reactions with nucleophiles, organometallics, hydrogenation (except in acetic acid), hydrides and oxidizing agents, but they do react with Lewis acids. The most common method for cleavage of the dioxane or dioxolane derivatives is treatment with aqueous acid, including HCl in THF103 and aqueous acetic acid,104 and DDQ is another method for deprotection.105 In a synthesis of ()-cladospolide B by Wang et al.,106 diol 87 was treated with 2,2-dimethoxypropane and an acid catalyst to give a 94% yield of dioxane 88. After five synthetic steps, the dioxolane group in 89 was converted to the diol 90 and acetone with methanolic HCl, in 88% yield. OTBS

OH

(S)

(S)

OH

OTBS

Me2C(OMe)2 , p-TsOH

(R)

CH2Cl2 , rt

CO2Et

O

5 Steps

(S) (S)

O (R)

CO2Et

87

88 (94%) O

O

O

O (S)

cat HCl , MeOH

(R)

O

OH (S)

O

rt

(R)

(S)

OH (S)

89

90 (88%)

A variation in ketal protection introduces alkyl substituents on the diol moiety. Barrett and coworkers,107 for example, used 2,2-dimethyl-1,3-propanediol to protect one of the ketone units of the quinone in 91 (forming 92), in studies aimed at a synthesis of lactonamycin. Seven synthetic steps led to 93, and deprotection with aqueous acetic acid regenerated the carbonyl in quinone 94, in 70% yield. O O HO

7 Steps

OH

H+

O

O

O 91

92 t-BuO2C

t-BuO2C

O H

O

BnO2C

BnO2C AcOH , H2O

H

O

O

93

Air

O

94 (70%)

101

Reference 2(a), pp 124–125; Reference 2(b), pp 188–195, 185–186; Reference 2(c), pp 307–308; 725–727; Reference 2(d), pp 454–466; Reference 2(e), pp 585–601.

102

Reference 2(a), pp 121–126, 312–314; Reference 2(b), pp 188–195, 185–186; Reference 2(c), pp 307–308; 725–727; Reference 2(d), pp 449–452; Reference 2(e), pp 578–581. (a) Grieco, P. A.; Nishizawa, M.; Oguri, T.; Burke, S. D.; Marinovic, N. J. Am. Chem. Soc. 1977, 99, 5773. (b) Grieco, P. A.; Yokoyama, Y.; Withers, G. P.; Okuniewicz, F. J.; Wang, C.-L. J. J. Org. Chem. 1978, 43, 4178.

103

104

Babler, J. H.; Malek, N. C.; Coghlan, M. J. J. Org. Chem. 1978, 43, 1821.

105

Tanemura, K.; Suzuki, T.; Horaguchi, T. J. Chem. Soc. Chem. Commun. 1992, 979.

106

Wang, W.-K.; Zhang, J.-Y.; He, J.-M.; Tang, S.-B.; Wang, X.-L.; She, X.-G.; Pan, X.-F. Chin. J. Chem. 2008, 26, 1109.

107

Henderson, D. A.; Collier, P. N.; Pave, G.; Rzepa, P.; White, A. J. P.; Burrows, J. N.; Barrett, A. G. M. J. Org. Chem. 2006, 71, 2434.

205

5.3 PROTECTING GROUPS FOR ALCOHOLS, CARBONYLS, AND AMINES

5.3.3.2 Dithioketals and Dithioacetals The use of cyclic ketals and acetals in conjunction with acyclic carbonyl derivatives offers some selectivity for protection of more than one carbonyl. Expanding the methodology to include the sulfur analog [dithioketals and dithioacetals, R2C(SR0 )2] gives even more flexibility. Both cyclic and acyclic dithio derivatives are known and used, but cyclic dithioketals and acetals are more common. A ketone or aldehyde reacts with 2 equiv of a thiol (mercaptan) and an acid (typically HCl and BF3).108,109a The main advantage of these groups is their stability to acid. They are stable from pH 1–12 and compatible with nucleophiles, organometallics, hydrides, and oxidizing agents, although CrO3 can oxidize the sulfur.110 Dithioacetals and ketals are particularly reactive with mercuric and silver salts, and are relatively unreactive with many other Lewis acids.102 Mercuric chloride under aqueous conditions is the most commonly used method for conversion back to the carbonyl.111 As with the oxygenated analog, the use of cyclic dithioketals and acetals is very prevalent. Reaction with 1,2-ethanedithiol generates 1,3-dithiolanes, and 1,3-propanedithiol yields 1,3-dithianes. Boron trifluoride etherate (BF3OEt2) is commonly used as an acid catalyst to form dithianes and dithiolanes from ketones and aldehydes and dithiols.109 These protecting groups are similar in their stability to the acyclic derivatives. Cyclic dithioketal and dithioacetal derivatives are relatively insensitive to acid, showing stability from pH 1–12. They are stable to nucleophiles, organometallics, hydrides, and many oxidizing agents.110 They are subject to hydrogenolysis and the bivalent sulfur atoms in dithianes can poison heterogeneous hydrogenation catalysts (Section 7.10.1). They are also sensitive to strong Lewis acids [e.g., aluminum chloride (AlCl3) and mercuric salts]. This latter reaction is the basis of the most common cleavage reaction, the Corey-Seebach procedure.112 Other cleavage reagents include Vedejs’ use of boron trifluoride etherate in aq THF containing mercuric oxide (HgO),112c NBS,113 iodine in DMSO,114 ceric ammonium nitrate [Ce(NH4)2(NO3)6, CAN],115 and iodomethane in aqueous media.116 One of the best cleavage procedures is Corey’s use of NCS and silver nitrate in aqueous acetonitrile.117 There are many synthetic examples that use dithiane or dithiolane protecting groups, in combination with many deprotection protocols. In Geum’s and Lee118 synthesis of panaginsene, aldehyde 95 was converted to dithiane 96 by reaction with 1,3-propanedithiol and magnesium bromide, in 62% yield. This sequence required 11 synthetic steps to generate 97, and the dithiane group was removed with iodine and bicarbonate to yield 98 in 76% yield. OTBDPS SH

OTBDPS

11 Steps

I2

H

SH

MgBr2

CHO

95

H

NaHCO3

S

S

S

S O

96 (62%)

97

98 (76%)

Other 1,2- and 1,3-dithiols have been used as protecting groups, but those discussed here are the most common. Wuts and Greene2 discuss other methods for protection of ketones and aldehydes, including cyanohydrins, hydrazones, oximes, oxazolidines, or imidazolidines.119 Most of these are rather specialized and will not be discussed in this general presentation. 108

Zinner, H. Chem. Ber. 1950, 83, 275.

109

(a) Fujita, E.; Nagao, Y.; Kanelo, K. Chem. Pharm. Bull. 1978, 26, 3743. (b) Corey, E. J.; Bock, M. G. Tetrahedron Lett. 1975, 2643.

110

Reference 2(a), pp 129–132, 312–314; Reference 2(b), p 198; Reference 2(c), pp 333–340; 725–727; Reference 2(d), pp 477–500; Reference 2(e), pp 615–648.

111

English, J., Jr.; Griswold, P. H., Jr. J. Am. Chem. Soc. 1945, 67, 2039.

112

(a) Seebach, D.; Corey, E. J. J. Org. Chem. 1975, 40, 231. (b) Seebach, D. Synthesis 1969, 17. (c) Vedejs, E.; Fuchs, P. L. J. Org. Chem. 1971, 36, 366.

113

(a) Cain, E. N.; Welling, L. L. Tetrahedron Lett. 1975, 1353. (b) Corey, E. J.; Erickson, B. W. J. Org. Chem. 1971, 36, 3553.

114

Chattopadhyaya, J. B.; Rama Rao, A. V. Tetrahedron Lett. 1973, 3735.

115

Ho, T.-L.; Ho, H. C.; Wong, C. M. J. Chem. Soc. Chem. Commun. 1972, 791.

116

(a) F
etizon, M.; Jurion, M. J. Chem. Soc. Chem. Commun. 1972, 382. (b) Takano, S.; Hatakeyama, S.; Ogasawara, K. Ibid., 1977, 68.

117

Corey, E. J.; Erickson, B. W. J. Org. Chem. 1971, 36, 3553.

118

Geum, S.; Lee, H.-Y. Org. Lett. 2014, 16, 2466.

119

Reference 2(a), pp 141–147, 312–314; Reference 2(b), pp 210–223; Reference 2(c), pp 348–364; Reference 2(d), pp 506–528; Reference 2(e), pp 650–685.

206

5. FUNCTIONAL GROUP EXCHANGE REACTIONS: PROTECTING GROUPS

5.3.4 Protection of Amines The last major type of functional group to be discussed will be amines. The interfering feature of this group is the lone pair of electrons on nitrogen, which makes the molecule basic and nucleophilic. If the amine reacts with alkyl groups, the electron pair is used to form a bond to carbon to generate an ammonium salt. Alkylation, basicity, or oxidation are not possible for ammonium salts, but they are not often used as protecting groups since these charged species are incompatible with many reactions, and solubility can be a problem. Ammonium salts have a NdH unit, and they are weak acids, which is another complication. In addition, ammonium salts are good leaving groups for substitution and elimination reactions. A more useful protection method generates a derivative in which the electron pair is completely or partially delocalized, reducing the basicity and nucleophilicity of the nitrogen. The three most common methods of amino group protection involve conversion of primary and secondary amines to a tertiary amine (usually benzyl or trialkylsilyl), conversion to an amide or a carbamate, and to a lesser extent, conversion to specialized sulfonamides.120 Only the first two of these methods will be discussed. 5.3.4.1 N-Alkyl and N-Silyl Protecting Groups Benzyl (NdCH2Ph, NdBn)121 is a common alkyl protecting group for nitrogen. An amine is treated with benzyl chloride or bromide, usually in the presence of a base [e.g., potassium carbonate (K2CO3) or hydroxide].122 This group is stable to acid and base (pH 1–12) and to nucleophiles, organometallics, and hydrides, but reacts with Lewis acids (as a Lewis base). The NdC bond is subject to hydrogenolysis by catalytic hydrogenation or dissolving metals. Hydrogen and Pd/C123 (Section 7.10.6) or sodium in liquid ammonia124 (Section 7.11.4) are the two most common methods of cleavage. Most of the other N-alkyl groups used are less general.125 A synthetic example using the N-benzyl protecting group, although it is used to protect an amide, is taken from Nocket’s and Weinreb synthesis of myrioneurinol.126 The reaction of amide 99 with silver oxide and benzyl bromide gave N-benzyl derivative 100 in 73% yield. After 14 synthetic steps, to yield 101, treatment with Na in NH3 led to deprotection of the N-benzyl group to yield 102 in 79% yield. It has been reported that N-benzyl group can be removed in the presence of O-benzyl by catalytic transfer hydrogenation in the presence of a hydrogen donor, cyclohexa-1,4-diene.127 H

O N H

H

CO2Me

CO2Me 14 Steps

BnBr , Ag2O

O

TBAI , CaSO4 , rt

N

(TBAI = Tetrabutylammonium iodide)

99

Bn

100 (73%) OMOM

H

Na/NH3 , ether

O

OMOM

H O

–78°C

N

N H

Bn SiMe3 101

120

SiMe3 102 (79%)

For a review of nitrogen protecting groups, see Theodoridis, G. Tetrahedron 2000, 56, 2339.

121

Reference 2(a), pp 272–273, 332–334; Reference 2(b), pp 335–338; Reference 2(c), pp 620–621; 745–747; Reference 2(d), pp 814–818; Reference 2(e), pp 1042–1048.

122

Velluz, L.; Amiard, G.; Heymès, R. Bull. Soc. Chim. Fr. 1954, 1012.

123

Hartung, W. H.; Simonoff, R. Org. React. 1953, 7, 263.

124

du Vigneaud, V.; Behrens, O. K. J. Biol. Chem. 1937, 117, 27.

125

Reference 2(a), pp 268–275, 332–334; Reference 2(b), pp 321–341; Reference 2(c), pp 573–586; 619–620; Reference 2(d), pp 803–809; Reference 2(e), pp 1025–1060.

126

Nocket, A. J.; Weinreb, S. M. Angew. Chem. Int. Ed. 2014, 53, 14162.

127

Bajwa, J. S.; Slade, J.; Repic, O. Tetrahedron Lett. 2000, 41, 6025.

207

5.3 PROTECTING GROUPS FOR ALCOHOLS, CARBONYLS, AND AMINES

The benzyl protection group can be applied to primary or secondary amines. Gathergood and Scammells128 incorporated a primary amine precursor in the synthesis of psilocin, by reaction of but-3-yn-1-yl 4-methylbenzenesulfonate with dibenzylamine to yield N,N-dibenzylbut-3-yn-1-amine in 56% yield. The dibenzylated species functions as a primary amine surrogate, where both hydrogen atoms attached to the nitrogen have been removed. A Pd catalyzed coupling reaction with aryl iodide 103 (Section 18.2.3) gave 104, and catalytic hydrogenation with a Pd/C catalyst removed both benzyl groups to yield 105 in 83% yield. 0.2 Pd(OAc)2 , 0.4 PPh3 Et4NCl , i-Pr2NEt DMF , 80°C

Bn2NH , MeOH

OTs

NBn2

Reflux , 2 d

MeO

I NHCO2t-Bu

But-3-yn-1-yl 4-Methylbenzenesulfonate

N, N-Dibenzylbut- (56%) 3-yn-1-amine

103

OMe

OMe NBn2 SiMe3

N

NH2

20% Pd(OH)2/C 40 psi H2

CO2t-Bu

N

SiMe3

CO2t-Bu

104

105 (83%)

A simple silyl protecting group is trimethylsilyl (N-SiMe3, N-TMS), which is formed by reaction of the amine with chlorotrimethylsilane and triethylamine or pyridine. The trimethylsilyl group is sensitive to water or alcohol solvents, and the reaction of Me3SiNEt2 and alcohols yields ROSiMe3. The N-TMS group is therefore used under anhydrous conditions.129 In a synthesis of ()-21-isopentenylpaxilline by the Smith and Cui130 reaction of aniline derivative 4-bromo-2-methylaniline with MeLi (Section 8.5) and excess chlorotrimethylsilane gave an 84% yield of 106, blocking both hydrogen atoms on the primary amine. Alkylation (Section 11.6) gave 107, and treatment with ethanolic HCl gave 108 in 82% yield for the last two steps. Br

Br MeLi , TMSCl

NH2

SiMe3

t-BuLi , THF

N

THF , 0°C

Br

SiMe3

4-Bromo-2methylaniline

106 (84%)

1 N HCl , EtOH

SiMe3

for 2 steps

N 107

SiMe3

NH2 108 (82%)

5.3.4.2 N-Acyl Protecting Groups Conversion to an amide is a common method for the protection of amines. N-Acetyl is the best known of the amide protecting groups, and N-acylamines are known as acetamide derivatives (N-COMe, N-Ac).131 Reaction of acetic anhydride or acetyl chloride with an amine, in the presence of a base (e.g., Py or triethylamine) will generate the acetamide.132 Acetamides are sensitive to strong acid and base, but are stable in the pH range 1–12. Nucleophiles and many organometallics react with the N-acetyl group. Organolithium reagents are unreactive, although reactive 128

Gathergood, N.; Scammells, P. J. Org. Lett. 2003, 5, 921.

129

(a) Reference 2(a), pp 283; Reference 2(b), pp 69–71. (b) Pratt, J. R.; Massey, W. D.; Pinkerton, F. H.; Thames, S. F. J. Org. Chem. 1975, 40, 1090.

130

Smith, III, A. B.; Cui, H. Org. Lett. 2003, 5, 587.

131

Reference 2(a), pp 251–252, 328–330; Reference 2(b), pp 351–352; Reference 2(c), pp 552–555; 741–743; Reference 2(d), pp 775–779; Reference 2(e), pp 993–998.

132

Barrett, A. G. M.; Lana, J. C. A. J. Chem. Soc. Chem. Commun. 1978, 471.

208

5. FUNCTIONAL GROUP EXCHANGE REACTIONS: PROTECTING GROUPS

Grignard reagents do react. This group can be reduced by catalytic hydrogenation, borane, or borohydride reducing agents (not LiAlH4), and can be oxidized by many oxidizing agents. Reactions that use Lewis acids (e.g., glycosylation), may cause problems if this protecting group is present. The two major methods for conversion of N-acetamides to the amine are treatment with aqueous acid133 and treatment with triethyloxonium tetrafluorobo135 ). Reduction with LiAlH4 yields the ethylamine derivative, but cleavage is rate134 (Et3O+BF 4 , Meerwein’s reagent 136 sometimes observed. In a synthesis of ()-securinine by Honda et al.,137 the N-acetyl group was used to protect 109 as an acetamide, 110. Conversion to 111 in five steps was followed by cleavage with trifluoroacetic acid to yield the amine group in 112. Et

HO

H

N

H H

109

Et

HO

Ac2O , NEt3 CH2Cl2 , rt

N

Ac

110 (quant) O

O 5 Steps

H

O

CH2Cl2

N Ac

TFA , rt

Br 111

H N H

O

Br

112

A simple modification of the structure has dramatic effects on the stability of the protecting group. If trifluoroacetyl is used rather than acetyl, the trifluoroacetamide (N-COCF3, N-TFA)138 is generated. The trifluoroacetyl group is usually attached by reaction of the amine with trifluoroacetic anhydride [(CF3CO)2O] in the presence of triethylamine or Py.132,139 This group retains its stability to acid, but is more sensitive to base (pH 1–10) than the acetamide analog. It is stable to anhydrous bases or tetraalkylammonium hydroxides. The group reacts with most nucleophiles, organolithium reagents, LiAlH4, and borohydrides via nucleophilic acyl substitution, but is stable to oxidation, Lewis acids, borane, and most catalytic hydrogenation conditions. It is usually removed with potassium carbonate in aqueous methanol,140 by reduction with sodium borohydride141 or by ammonia in methanol.142 In a synthesis of microsclerodermin E by Zhu and Ma,143 the primary amine unit in 113 was converted to the trifluoroacetamide (114), in 62% yield. After 7 steps, 115 was deprotected with ethanoic sodium borohydride to give a quantitative yield of 116.

133

Dilbeck, G. A.; Field, L.; Gallo, A. A.; Gargiulo, R. J. J. Org. Chem. 1978, 43, 4593.

134

Hanessian, S. Tetrahedron Lett. 1967, 1549.

135

(a) Meerwein, H. Org. Synth. Coll. 1973, 5, 1080. (b) For reactions with amides and lactams see Borch, R. F. Tetrahedron Lett. 1968, 61.

136

Gaylord, N. G. Reductions With Complex Metal Hydrides; Ellis Horwood Ltd.: Chichester, UK, 1964; pp 544–546.

137

Honda, T.; Namiki, H.; Kaneda, K.; Mizutani, H. Org. Lett. 2004, 6, 87.

138

Reference 2(a), pp 254–255, 328–330; Reference 2(b), pp 353–354; Reference 2(c), pp 556–558; 741–743; Reference 2(d), pp 781–783; Reference 2(e), pp 1000–1002.

139

Green, J. F.; Jham, G. N.; Neumeyer, J. L.; Vouros, P. J. Pharm. Sci. 1980, 69, 936.

140

(a) Quick, J.; Meltz, C. Ibid., 1979, 44, 573. (b) Schwartz, M. A.; Rose, B. F.; Vishnuvajjala, B. J. Am. Chem. Soc. 1973, 95, 612.

141

Weygand, F.; Frauendorfer, E. Chem. Ber. 1970, 103, 2437.

142

Imazawa, M.; Eckstein, F. J. Org. Chem. 1979, 44, 2039.

143

Zhu, J.; Ma, D. Angew. Chem. Int. Ed. 2003, 42, 5348.

209

5.3 PROTECTING GROUPS FOR ALCOHOLS, CARBONYLS, AND AMINES

O O

O O

O

(CF3CO)2O , Py

O

PhO2S PhO2S

O

O HN

7 Steps

CF3

NH2 O 114 (62%)

113

O

OMOM

O

OH HN

O NaBH4 , EtOH

CF3

OMOM

O

OH NH2

O EtO 115

EtO

116 (quant)

Benzamides (NdCOPh, N-Bz)144 are formed by the reaction of an amine with benzoyl chloride in pyridine or triethylamine.145 This group is stable to pH 1–14, nucleophiles, organometallics (not RLi), hydrogenation, hydrides (except LiAlH4, and borane), and oxidizing agents. Cleavage is accomplished with 6 NHCl or HBr in acetic acid,146 hot and concentrated aq NaOH,147 and reduction with dibal (Section 7.6.1).148 5.3.4.3 N-Carbamate Protecting Groups The carbamates (NdCOOR) are a related class of protecting groups for nitrogen.149 Several different carbamates have been used for the protection of amino acids in peptide synthesis. One of the most popular is the tert-butyl carbamate (tert-butoxycarbonyl, [NCOC(Me)3], N-Boc).150 Commercially available BOC-ON [(Me)3CdOdCOCO2N]C(CN)Ph] reacts with amines, in the presence of another base (e.g., triethylamine) to yield the N-Boc derivative.151 Similarly, reaction with di-tert-butyldicarbonate ([(CH3)3CCO]2CO) under basic conditions yields the BOC protected amine.152 This group is sensitive to strong anhydrous acid (stable to pH 1–12) and trimethylsilyl triflate (Me3SiOTf ), but it is stable to nucleophiles, organometallics (including organolithium reagents although Grignard reagents can react by nucleophilic acyl addition), hydrogenation (except under acidic conditions), hydrides, oxidizing agents (not Jones’ conditions), aqueous acid, and mild Lewis acids. The Boc group is usually removed by treatment with aq HCl153 or with anhydrous trifluoroacetic acid.154 In a synthesis of spirobacillene A by Taylor and coworkers,155 the indole amine unit in 117 was converted to the N-Boc derivative 118, in >98% yield. Using (Boc)2O. After eight synthetic steps to yield 119, treatment with trifluoroacetic acid cleaved the Boc group to give an 82% yield of the amine 120.

144

Reference 2(a), pp 261–263, 328–330; Reference 2(b), pp 355–356; Reference 2(c), pp 560–561; 741–743; Reference 2(d), pp 785–787; Reference 2(e), pp 1004–1006.

145

White, E. Org. Synth. Collect. 1973, 5, 336.

146

Ben-Ishai, D.; Altman, J.; Peled, N. Tetrahedron 1977, 33, 2715.

147

Khorana, H. G.; Turner, A. F.; Vizsolyi, J. P. J. Am. Chem. Soc. 1961, 83, 686.

148

Gutzwiller, J.; Uskokovi
c, M. J. Am. Chem. Soc. 1970, 92, 204.

149

Reference 2(a), pp 222–248, 324–326; Reference 2(b), pp 327–330; Reference 2(c), pp 518–525; 737–739; Reference 2(d), pp 707–718; Reference 2(e), pp 907–989.

150

Reference 2(a), pp 324–326; Reference 2(d), pp 725–735; Reference 2(e), pp 930–946.

151

Itoh, M.; Hagiwara, D.; Kamiya, T. Bull. Chem. Soc. Jpn. 1977, 50, 718.

152

Tarbell, D. S.; Yamamoto, Y.; Pope, B. M. Proc. Natl. Acad. Sci. 1972, 69, 730.

153

Stahl, G. L.; Walter, R.; Smith, C. W. J. Org. Chem. 1978, 43, 2285.

154

Lundt, B. F.; Johansen, N. L.; Vølund, A.; Markussen, J. Int. J. Pept. Protein Res. 1973, 12, 258.

155

Unsworth, W. P.; Cuthbertson, J. D.; Taylor, R. J. K. Org. Lett. 2013, 15, 3306.

210

5. FUNCTIONAL GROUP EXCHANGE REACTIONS: PROTECTING GROUPS

O

O HO I

HO

I (Boc) 2O , DCM NEt3 DMAP

N

TFA , DCM

8 Steps

N

H

N

Boc

N O

Boc 117

118 (>98%)

O

H

119

120 (82%)

Another widely used carbamate is the benzyl carbamate (NdCO2CH2Ph, N-Cbz, carbobenzyloxy),156 which Greene pointed out has been used since 1932.157 The group is attached by the reaction of an amine with benzyl chloroformate (PhCH2O2CCl) in the presence of base (aqueous carbonate or triethylamine). The group is very stable to acid and base (pH 1–12), to nucleophiles, to milder organometallics (it reacts with organolithium reagents and Grignard reagents), to milder Lewis acids, and to most hydrides (it reacts with LiAlH4). The group is sensitive to hydrogenolysis by catalytic hydrogenation, which is the primary means of cleaving this group (Pd/carbon is the usual catalyst).151,158 In a synthesis of aspidospermidine, Matsuo and coworkers159 protected N-benzylamine (121) as the Cbz derivative 122 in 99% yield by treatment with CbzCl. After three steps, 123 was obtained, and catalytic hydrogenation removed the Cbz group to yield 124, in quantitative yield. O Et

Et

Et

3 Steps

CbzCl

N Bn 121

DCM , rt

O

N

N

Cbz 122 (99%)

O Et H2 , Pd/C

Cbz 123

EtOH , rt

O N H 124 (quant)

5.4 CONCLUSION It is obvious that the preceding discussions are not exhaustive, but they are representative of the methods used by synthetic chemists to protect key functional groups. The monographs by Wuts and Greene2 also discuss methods for protection of alkynes,160 phosphates,161 carboxylic acids,162 amides,163 and thiols.164 These monographs also contain detailed discussions of other protecting groups that can be used for alcohols, amines, ketones, and aldehydes. There are obviously other situations that demand a protecting group, but the cases presented above give a useful arsenal of protecting groups, as well as a sample of the situations that require their use. One should always be cautious with protecting groups, remembering that each protection-deprotection sequence adds two steps to a synthesis that diminishes the overall yield. Protecting and deprotecting steps often involve the use of an excess of reagents. These reagents are sometimes cheap, and the yields are usually high. Also, one-pot and simultaneous deprotection saves steps. New protecting groups for specific and general situations are always being developed, and a search of the current literature is always useful.

156

Reference 2(a), pp 239–241, 324–326; Reference 2(b), pp 335–338; Reference 2(c), pp 531–537; 737–739; Reference 2(d), pp 748–756; Reference 2(e), pp 961–971.

157

(a) Reference 2(a), p 239. (b) Bergmann, M.; Zervas, L. Berichte 1932, 65, 1192.

158

Meienhofer, J.; Kuromizu, K. Tetrahedron Lett. 1974, 3259.

159

Kawano, M.; Kiuchi, T.; Negishi, S.; Tanaka, H.; Hoshikawa, T.; Matsuo, J.-I.; Ishibashi, H. Angew. Chem. Int. Ed. 2013, 52, 906.

160

Reference 2(d), pp 927–933; Reference 2(e), pp 1194–1202.

161

Reference 2(d), pp 934–985; Reference 2(e), pp 1203–1262.

162

Reference 2(b), pp 224–276; Reference 2(c), pp 369–453; 729–731; Reference 2(d), pp 533–646; Reference 2(e), pp 686–836.

163

Reference 2(a), pp 152–192, 315–317; Reference 2(b), pp 349–362; Reference 2(c), pp 632–647; Reference 2(e), pp 1151–1193.

164

Reference 2(a), pp 195–222, 319–321; Reference 2(b), pp 277–308; Reference 2(c), pp 454–493; 733–735; Reference 2(d), pp 647–695; Reference 2(e), pp 837–894.

211

5.4 CONCLUSION

HOMEWORK

Synthesis involves many reactions and many of the following problems may involve reactions from chapters yet to come. Using protecting groups requires multifunctional molecules, and the use of reactions in succeeding chapters is unavoidable. The focus is on the functional group exchange reactions to incorporate and then deprotect the various groups. 1. In each of the following, discuss the relative merits of protecting functional groups or finding a chemoselective reagent for (i) reduction with hydrides (Sections 7.3–7.8), (ii) oxidation with Jones reagent or PCC (Sections 6.2.1 and 6.2.2), (iii) catalytic hydrogenation (Section 7.10), and (iv) reaction with MeMgBr (Section 11.4). O

O

O

(a)

(b)

(c)

(d)

O

CHO

OH

CHO

OH

CO2H

2. In the transformation of A to B, and then to C, taken from the synthesis of 7-deoxypancratistatin, deprotection was followed by conversion of a lactone to a lactam. Provide a mechanism for the deprotection of both the OdMOM group and the dioxolane, giving a triol, and also for the lactone-lactam conversion in the second step. OMOM O

O

OH

OH

O

O

O

OH

Dowex-H+

K2CO3 , MeOH

O HN

HO

OH

CF3

HN

O

CF3

OH

O

OH NH

O

O O

O

O

O

O B

A (racemic)

C (racemic)

3. Explain the following transformation: HO C12 H25

Me2C(OMe)2 , cat TsOH

O

O

O

O CO2Me

PhH , reflux

C12 H25

212

5. FUNCTIONAL GROUP EXCHANGE REACTIONS: PROTECTING GROUPS

4. In each of the following, give the major product. Show stereochemistry where it is appropriate. OH

H

(A)

Excess Ac2O

OH

O

3. I2 , NaHCO3

CHO

CO2Me

2

HS

SH

Propylene glycol cat TsOH , PhH/THF

Br

(D)

O

BF3

MeO2C

OH N3

(E)

1. Na , NH3 2. CrO3 , H2SO 4

H H

O

(C)

O

N

(B) BnO

Cbz Excess H2 , Pd/C

BnO

Boc

N

Pd black , HCO2NH4

N

(F) HO

O

CF3SO3H

OAc BzO

OH

(G)

BBr3 , CH2Cl2

O

CO2Me

MeO2C

(H) Py–TsOH

OH

O

OAc

O

2,2-Dimethoxypropane cat TsOH , CH2Cl2

NHBoc

–78

0°C

NH OMe

OMe O

OH

(I)

OAc

BBr3

(J)

CH2Cl2

MeO2C OH

(K)

3 equiv PhCHO

O

4 equiv CF3CO2H PhMe , –20 0°C

O

OBz

OH

(L)

1. 60% AcOH 2. TBDPSCl , DMAP , Py

O O

3. Me2C(OMe)2 , TsOH

OH

H

CHO

H 6 equiv BBr3

(M)

N

MeO N

H

CH2Cl2 rt –78°C

H N

OH

OSiMe2t-Bu OEt

H

Me

Me

(O)

(N)

NaH , BnBr

OH THF

1. Me3SiCH2CH2OCH2Cl Hünigs base , CH2Cl2 2. 80% AcOH

213

5.4 CONCLUSION

5. For each of the following, determine if protection is required in the given sequence. If so, choose the protecting group(s), explain your choice(s), and provide all necessary reagents for the entire sequence. OH Br

OH

OH

OH (B)

(A)

OH Me

OH

Me

O

O Me

Me

OH

OH

OH OH

CO2H

CHO

HO

(C) OH

(D)

OH

OHC OH Me

Me

Me

Me H

(E)

n-C4H9

O OH

OH

6. In each case, complete the synthesis showing all reagents and all intermediate products. O

OH

O

OH

MeO

(A)

(B)

OH

OH

OH

HO

OH EtO OH

O

(C)

OH

O

O O

Ph

OMe OMe OH

O

O

OTBS

O

S

BnO

S O

HO

OH

H

OCPh3

(H)

CO2Et

N

N H

O

C14 H29

CHO Ot-Bu

OPiv

O HOOC

O

N

CO2H

CHO

(I)

OBn

H

(F)

O

OTBS

OBn

(E)

OMOM

O

O

O

O

(G)

CHO

(D)

O

O

O

(J)

OH

OMOM Me3Si

O Me3Si CHO

OH

OAc

HO

(L)

(K) BnO

TBDPSO

MOMO

O OAc

TBDPSO

O OH

Me

Me Me

OPMB

Me

OPMB

C H A P T E R

6 Functional Group Exchange Reactions: Oxidations 6.1 INTRODUCTION Oxidation is a very important class of chemical reactions, which in organic chemistry can be defined in several ways. A broad definition by Sheldon and Kochi1 states “oxidation in organic chemistry refers to either (1) the elimination of hydrogen atoms, as in the sequential dehydrogenation of ethane, or (2) the replacement of a hydrogen atom bonded to carbon with another more electronegative element (e.g., oxygen) in the following series of oxidative transformations of methane:” CH4




→ CH3 OH




→ CH2 O




→ HCO2 H




→ CO2 Oxidation can also be defined as the reaction of an element with oxygen,2 analogous to the second criterion stated above. A more general definition, applied most often in inorganic chemistry, involves the loss of one or more electrons from an atom or group.2 Soloveichik and Krakauer3 listed five criteria that could be used to identify an oxidation process for a given organic reaction, and arrange compounds by their ability to be oxidized. 1. 2. 3. 4. 5.

Pauling’s electronegativity scale.4 The oxidation number of the parent substance of each homologous series under consideration. The ratio of bond moment to bond length, μL
1 (net charge).5 Hammett’s sigma function as revised by Taft (Taft’s σ*).6 Chemical reactions of the substance involved, particularly hydrolysis, which does not provide fundamental changes in oxidation stage if certain conventions are maintained.3

Determination of the oxidation state of a molecule is useful for categorizing a reaction as an oxidation or a reduction. It is also useful to determine if a starting material is in a higher or a lower oxidation state relative to the product. A practical view of oxidation associates a loss of electrons with the change in oxidation state of an atom, and electron loss can be determined by a device known as the oxidation number.7 A simplified version of this approach was used by

1

Sheldon, R. A.; Kochi, J. K. Metal-Catalyzed Oxidations of Organic Compounds; Academic Press: New York, NY, 1981; p 6.

2

Yalman, R. G. J. Chem. Educ. 1959, 36, 215.

3

Soloveichik, S.; Krakauer, H. J. Chem. Educ. 1966, 43, 532.

4

Pauling, L. Nature of the Chemical Bond, 3rd ed.; Cornell University Press: Ithaca, NY, 1940; pp 221–231.

5

(a) Del Re, G. Electronic Aspects of Biochemistry; Pullman, B., Ed.; Academic Press: New York, NY, 1964; pp 221–231; (b) Smyth, C. P. Dielectric Behavior and Structure; McGraw-Hill: New York, NY, 1955; pp 244–245, 247. 6

(a) Taft, Jr., R. W. J. Am. Chem. Soc. 1952, 74, 2729; (b) Taft, Jr., R. W. J. Am. Chem. Soc. 1958, 75,4231; (c) Taft, Jr., R. W. Steric Effects in Organic Chemistry; Newman, M. S., Ed.; John Wiley: New York, NY, 1956 (chapter 13); pp 556–675.

7

Holleran, E. M.; Jespersen, N. D. J. Chem. Educ. 1980, 57, 670.

Organic Synthesis http://dx.doi.org/10.1016/B978-0-12-800720-4.00006-4

215

Copyright © 2017 Michael Smith. Published by Elsevier Inc. All rights reserved.

216

6. FUNCTIONAL GROUP EXCHANGE REACTIONS

Hendrickson et al.8 and later by Pine9 in their undergraduate organic textbooks. Only those atoms most commonly encountered in simple organic molecules are considered: H, O, C, N, and the halogens.8,9
1 for hydrogen 0 for carbon (bonds to same atom) +1 for a covalent bond to a heteroatom (usually oxygen or halogen) The oxidation number should be determined only for those atoms that are modified or changed as a result of the reaction. If the sums of the oxidation numbers of the various atoms in the starting material and in the product are compared, the change in oxidation number will indicate an oxidation or a reduction. An example is the oxidation of propan-1-ol to propanal, which is subsequently oxidized to propanoic acid in a second step. In propan-1-ol, the carbon bearing the OH group has an oxidation number of
1 (C is 0, each H is
1, and O is +1). The oxidation number of carbon in propanal is +1 (C is 0, H is
1, O is +1, and O, from the second bond to the carbonyl, is +1). The change in oxidation number is
1 ! +1, for a net loss of two electrons, and an oxidation. Remember that as the net charge becomes more positive, this is associated with the loss of electrons, which have a negative charge. Similarly, the carbon of interest in propanoic acid has an oxidation number of +3 (C is 0, the three bonds to oxygen total +3). Conversion of propanal to propanoic acid involves the loss of two additional electrons (+1 ! +3), and it is an oxidation. Only the carbon bearing the oxygen atoms was examined in this example since the other carbon atoms are not changed during the course of the reaction.

H

–1 OH C–

+1

H –1

2 e– Lost –1

O

2 e– Lost

O

2 e– Lost

+3 H 0

OH

O 0

The oxidation of but-2-ene to the corresponding epoxide (see Section 6.4) involves modification of two carbon atoms. The oxidation number of each carbon comprising the π-bond is
1 (C and C ¼ 0, and H is
1), and the oxidation number of each epoxy carbon in the product is 0 (C and C ¼ 0, O is +1 and H is
1), for a change of one electron (
1 ! 0) for each of those two carbons. The net change for each molecule is loss of two electrons and an oxidation. In the following sections, the emphasis will be on oxidation reactions that transform one functional group into another. Important transformations will be the oxidation of alcohols to carbonyls, oxidation of alkenes to diols or epoxides, and oxidative cleavage of alkenes to carbonyl derivatives. A few methods will be discussed for the oxidation of alkyl fragments, (e.g., oxidation of a methyl group to hydroxymethyl, an aldehyde, or a carboxylic acid). The oxidation of sulfur and nitrogen compounds will also be briefly discussed. Oxidation reactions are generally categorized as functional group exchange reactions. Therefore, classification of a reaction as an oxidation in organic synthesis can be correlated with the disconnection-transform theme of this book. The reader is referred to Hudlický’s10a excellent monograph for a thorough synthetic discussion of oxidation reactions.

6.2 ALCOHOLS TO CARBONYLS (CHdOH ! C]O) One of the most important oxidations in organic chemistry converts alcohols to aldehydes, ketones, or carboxylic acid derivatives. Many oxidizing agents can be used, but the particular product formed depends on the structure of the alcohol, as well as the reagent.11 Primary alcohols are initially oxidized to aldehydes, which can be isolated, but that

8

Hendrickson, J. B.; Cram, D. J.; Hammond, G. S. Organic Chemistry, 3rd ed.; McGraw-Hill: New York, NY, 1970.

9

Pine, S. H. Organic Chemistry, 5th ed.; McGraw-Hill: New York, NY, 1987.

10

(a) Hudlický, M. Oxidations in Organic Chemistry; American Chemical Society Monograph 186, American Chemical Society: Washington DC, 1990; also see (b) Stewart, R. In Oxidation in Organic Chemistry, Part A; Wiberg, K., Ed.; Academic Press: New York, NY, 1965; p 11

11

Larock, R. C. Comprehensive Organic Transformations, 2nd ed.; Wiley-VCH: New York, NY, 1999; pp 1234–1256.

6.2 ALCOHOLS TO CARBONYLS (CHdOH ! C]O)

217

aldehyde can be oxidized further to a carboxylic acid. Secondary alcohols can be oxidized to ketones, but oxidative cleavage (a four-electron oxidation) can occur if the oxidation conditions are too harsh. This section will begin with the most common alcohol oxidizing reagents, which are based on Cr(VI). A discussion of other reagents, including the use of DMSO reagents, the Dess-Martin periodinane reagent, tetrapropylammonium perruthenate (TPAP, Pr4 N + RuO4
), and others will follow.

6.2.1 Oxidation With Chromium (VI) Chromium (VI) is a strong oxidizing agent. There are several Cr species that can be used for oxidation, including several that arise in aqueous solvents. 6.2.1.1 The Reaction of Alcohols With Chromium (VI) Primary alcohols are oxidized to aldehydes and secondary alcohols are oxidized to ketones using Cr(VI) reagents. In some cases, Cr(VI) reagents can oxidize initially formed aldehydes to a carboxylic acid. In order to understand these functional group transformations, the Cr reagent must be understood. Oxidation of an alcohol by Cr(VI) is accompanied by reduction of the chromium [Cr(VI) ! Cr(III)]. There are many Cr(VI) reagents.12,13 The inorganic reagent chromium trioxide is commonly used, which is a polymeric species usually written as (CrO3)n. In aqueous media, chromium trioxide exists in equilibrium with several other Cr(VI) species, including H2CrO4, HCrO4
, CrO4 2
, HCr2O7, Cr2 O7 2
, H2Cr2O7, and HCr2 O7
.14 Of the Cr(VI) species listed, dichromate is the strongest. Oxidation of an alcohol requires that both the alcohol and the chromium oxidant are soluble or partly soluble in the reaction medium, so most oxidations are carried out in aqueous media. When the reaction medium is water, several Cr(VI) species are present. At high dilution, the excess of water shifts the overall equilibrium toward dichromate.14,15 Dichromate is generated by “complex formation between the two acid chromate ions to yield a dichromate anion.”16 Formation of this complex and its relative concentration are dependent on the pKa of the acid. In highly concentrated solutions (less water), polymeric chromium trioxide (polychromates) and chromic acid are the predominate species.15 In dilute solutions, the concentration of Cr(VI) is independent of the acid, but only if the acidity is below the apparent pKa of chromic acid.16 Modification of the acid will increase the pKa of the complex as the electron-withdrawing power of the acid decreases. The electrons on the oxygen atoms attached to Cr are less available for protonation,16 leading to a larger dissociation constant,17 which suggests that the position of the equilibrium depends on the acid. Indeed, adding a different acid (HA) to the mixture will influence the position of the overall equilibrium according to the reaction18: HCrO4
+ 2H + + A





→ HCrO3 A + H2 O If the acid is more effective at withdrawing electrons from the complex (HCrO3A), the pKa for the complex increases. There is, therefore, a correlation with the strength of the mineral acid and the HCrO3A species:16,17 H3 PO4 < HCl < H2 SO4 < HClO4

12

Wiberg, K. B. In Chapter 2 of Oxidation in Organic Chemistry, Part A; Wiberg, K. B., Ed.; Academic Press: New York, NY, 1965; pp 69–70.

13

(a) Bystr€ om, A.; Wilhelmi, K. A. Acta Chem. Scand. 1950, 4, 1131; (b) Hanic, F.; Stempelóva, D. Chem. Zvestii 1960, 14, 165 (Chem. Abstr. 54: 20402a, 1960).

14

(a) Reference 12a, p 71; (b) Neuss, J. D.; Rieman, W. J. Am. Chem. Soc. 1934, 56, 2238; (c) Tong, J. Y. P.; King, E. L. J. Am. Chem. Soc. 1953, 75, 6180; (d) Davies, W. G.; Prue, J. E. Trans. Faraday Soc. 1955, 51, 1045; (e) Howard, J. R.; Nair, V. S. K.; Nancollas, G. H. Trans. Faraday Soc. 1958, 54, 1034; (f ) Schwarzenbach, G.; Meier, J. J. Inorg. Nuc. Chem. 1958, 8, 302; (g) Bailey, N.; Carrington, A.; Lott, K. A. K.; Symons, M. C. R. J. Chem. Soc. 1960, 290; (h) Sasaki, Y. Acta Chem. Scand. 1962, 16, 719.

15

Freedman, M. L. J. Am. Chem. Soc. 1958, 80, 2072.

16

Lee, D. G.; Stewart, R J. Am. Chem. Soc. 1964, 86, 3051.

17

Reference 10b, p 72.

18

(a) Boyd, R. H. J. Am. Chem. Soc. 1961, 83, 4288; (b) Boyd, R. H. J. Phys. Chem. 1963, 67, 737.

218

6. FUNCTIONAL GROUP EXCHANGE REACTIONS

The structure of the oxidizing species is obviously important for determining the mechanism of oxidation of an alcohol to a carbonyl with Cr(VI). As mentioned above, in aqueous solutions CrO3 generates CrO4 2
, HCrO4
, and HCr2 O7 2
as the primary oxidizing species, along with other Cr species.19 At pH 1, CrO3 is almost completely ionized.19 Since organic molecules are being oxidized with Cr reagents, the focus tends to be on the organic molecule (the alcohol in this case), and Wiberg and Mukerjee20 presented the following scheme to represent the oxidation of a secondary alcohol to a ketone: R2 CHOH + CrðVIÞ



→ R2 C ¼ O + CrðVIÞ 

R2 CHOH + CrðIVÞ



→ R2 COH + CrðIIIÞ 

+ CrðVIÞ



→ R2 C ¼ O + CrðVÞ

R2 COH

R2 CHOH + CrðVÞ



→ R2 C ¼ O + CrðIIIÞ R2 CHOH + CrðVIÞ



→3 R2 C ¼ O + 2CrðIIIÞ Reprinted with permission from Wiberg, K.B.; Mukerjee, S.K. J. Am. Chem. Soc. 1974, 96, 1884. Copyright © 1974 American Chemical Society.

It is clear that Cr(VI), Cr(IV), and Cr(V) species are involved in the overall oxidation process, but there are no balanced equations in this scheme, and the Cr species are undefined. The Cr(VI) species are usually HCrO-4 and CrO3, the Cr(V) is usually HCrO4
and the Cr(IV) is usually HCrO3
.20 The oxidation requires both water (H2O) and an acid catalyst (H+), and a balanced equation for the oxidation of propan-2-ol is19 OH 3

+

2 HCrO4–

+

O

8 H+

3

2 Cr3+

+

+

8 H2O

There is strong evidence that the oxidation begins by formation of a chromate ester,21 which collapses to the carbonyl compound and a Cr(IV) species in the rate-determining step.19,22,23 The driving force of the reaction is reduction of Cr(VI) to Cr(IV).24 Westheimer and Novich23 determined that the rate term for the oxidation of propan-2-ol was 2 v ¼ ka ½HCrO4
 ½R2 CHOH ½H +  + kb ½HCrO4
 ½R2 CHOH ½H +  If sulfuric acid is used as the acid in an aqueous medium containing chromium trioxide and propan-2-ol, the initial chromate ester is probably 1, which arises from reaction of HCrO4
and H2SO4 (HA in the reaction of HCrO4
with 2H+ and A
).16 When the reaction medium is water that contains no external acid, the chromate ester reacts with propan-2-ol to form a chromate ester, i-PrOCrO3H.25 HCrO4–

OCrO3H

OH

O

HOCrO2OSO3–

O

Cr OSO3– O 1

In the reaction of propan-2-ol, decomposition of the chromate ester i-PrOCrO3H involves removal of the proton attached to the oxygen-bearing carbon. Westheimer and coworkers22,23,26c initially proposed that water behaves as a base to remove this hydrogen, although later work showed that addition of Py to the reaction (a stronger base than water) did not enhance the rate of oxidation.26

19

(a) Westheimer, F. H.; Nicolaides, W. J. Am. Chem. Soc. 1949, 71, 25; (b) Westheimer, F. H. Chem. Rev. 1949, 45, 419 (see p 427).

20

(a) Wiberg, K. B.; Mukerjee, S. K. J. Am. Chem. Soc. 1974, 96, 1884; (b) Rahman, M.; Rocek, J. J. Am. Chem. Soc. 1971, 93, 5462, 5455.

21

(a) Reference 12, p 161 and references cited therein; (b) Kl€ aning, U. Acta Chem. Scand. 1957, 11, 1313 and Kl€aning, U. Acta Chem. Scand. 1958, 12, 576.

22

Brownell, R.; Leo, A.; Chang, Y. W.; Westheimer, F. H. J. Am. Chem. Soc. 1960, 82, 406.

23

(a) Westheimer, F. H.; Novich, A. J. Chem. Phys. 1943, 11, 506; (b) Reference 12, see p 159 and Ref. 102 cited therein.

24

Reference 12, p 167.

25

Reference 12, p 162.

26

(a) Rocek, J.; Krupicka, J. Chem. Ind. (London) 1957, 1668; (b) Rocek, J. Collect. Czech. Chem. Commun. 1960, 25, 1052; (c) Westheimer, F. H.; Chang, Y. W. J. Phys. Chem. 1959, 63, 438.

219

6.2 ALCOHOLS TO CARBONYLS (CHdOH ! C]O)

R R

C

CrO3H R

H

•• ••

H2O

R O

C

+

H3O+

+

CrIV

O

1A R C R

O O Cr

R OH

R

H O

C

+

CrIV

O

2

R

H

H

OH

O

Cr OH

O

R

H

R

CrIV

O

H 3

OH2 • CH III R O Cr

OH

C

O

C R

R

H

H

4

5

R R

C

+

CrIV

O

At least two mechanisms have been proposed for this step. Westheimer proposed an intermolecular attack by water, as mentioned (see 1A), but later retracted this proposal.26c,27 Kwart and Francis27 proposed an intramolecular reaction for removal of the proton (see 2) by an oxygen on the chromate ester. Rocek26b discounted both mechanisms for several reasons, including the important observation that protonation should not lead to rate enhancement with a species (e.g., 3), but in fact it does. Rocek26b proposed the mechanism shown for 3, in which oxygen atoms from the chromic acid removed hydrogen atoms, both from carbon and from oxygen. Rahman and Rocek28 later proposed formation of a coordination complex (e.g., 4, shown exactly as reported in the paper), presumably formed from Cr(IV) in a similar way to 4, that decomposed to a radical species (5). Further oxidation of 5 generated the products. There is a steric component to this reaction. As the steric bulk around the carbon bearing the OH group increases, removing that hydrogen becomes increasingly difficult and the overall rate of oxidation will be diminished.29 The relative rate of oxidation of endo-borneol compared to exo-borneol, for example, is 25.0:49.1.30a This difference is correlated with the more facile removal of Ha from a chromate ester complex, where Ha is on the exo face than from a complex, where Ha is on the more hindered endo face. Another example of how steric hindrance influences an oxidation can be seen in cyclohexyl derivatives by comparing (1R,2R)-2-(tert-butyl)cyclohexan-1-ol (relative rate 50.6) and (1S,2R)-2-(tert-butyl)cyclohexan-1-ol (relative rate 10.7).30a In both (1R,2R)-2-(tert-butyl)cyclohexan-1-ol and (1S,2R)-2-(tert-butyl)cyclohexan-1-ol, the bulky tert-butyl group will occupy an equatorial position in the lowest energy conformation (Section 1.5.2). Removal of the equatorial Ha from the chromate ester of (1R,2R)-2-(tert-butyl)cyclohexan-1-ol is inhibited by a large steric interaction with the adjacent tert-butyl group (G-strain, Section 1.5.3). This rate difference is also observed with modified Cr reagents (e.g., pyridinium chlorochromate, Section 6.2.2.2).30b Me

Me

Me

H

Me

H Ha

H

OCrO3H

endo-Borneol

OCrO3H H

Ha

exo-Borneol Hb

HO Ha

H

OH

H

(1R,2 R)-2-(tert-Butyl)- (1S,2 R)-2-(tert-Butyl)cyclohexan-1-ol cyclohexan-1-ol 27

Kwart, H.; Francis, P. S. J. Am. Chem. Soc. 1959, 81, 2116.

28

Rahman, M.; Rocek, J. J. Am. Chem. Soc. 1971, 93, 5455.

29

Reference 12, p 165.

30

(a) Reference 12, p 166; (b) Suggs, J. W., Ph.D. Thesis Harvard University, 1976, Chapter III (see Dissertation Abstracts On-Line).

220

6. FUNCTIONAL GROUP EXCHANGE REACTIONS

From this discussion, it is apparent that oxidation of alcohols with Cr(VI) depends on several factors. The solvent and any acidic additives (or acidic solvents) will influence the rate of the oxidation, as will the structure of the alcohol. It is therefore not surprising that there are many variations in reaction conditions for Cr(VI) oxidations of alcohols. As the substrate and the oxidizing requirements of the Cr are changed, the reaction medium will be changed. Since most organic compounds are insoluble in water, a cosolvent is usually required to dissolve both the Cr reagent and the alcohol substrate. This solvent must be resistant to oxidation so acetic acid or acetone is commonly used. For the alcohol ! carbonyl conversion, several Cr(VI) reagents can be used, including chromium trioxide in water or aqueous acetic acid catalyzed by mineral acid, sodium dichromate in aqueous acetone catalyzed by mineral acid, sodium dichromate in acetic acid, the CrO3 Py complex, and tert-butyl chromate.31 Low-molecular-weight alcohols usually have sufficient solubility in water, so no cosolvent is required for oxidation with Cr reagents. The oxidation of propan-2-ol in water gives excellent yields of acetone, although the miscibility of acetone in water makes its isolation difficult. The presence of other oxidizable groups on the organic substrate (e.g., alkenes, sulfides, phenolic, and amines) can lead to side reactions in the oxidation of water soluble alcohols, significantly lowering the yield of carbonyl products. Sulfides are oxidized to sulfoxides, amines to hydroxylamines, and phenols to quinones, for example. Phenylalkyl carbinols [PhCH(OH)R] are subject to oxidative cleavage with chromium trioxide in aqueous acid, as observed in the oxidation shown.32 Oxidation of 1-phenyl-2-methylpropan-1-ol gave 1-phenyl-2-methylpropan-1-one and only 6% cleavage to benzaldehyde and propan-2-ol. Oxidation of 1-phenyl-2,2-dimethylpropane-1-one, however, gave 60% cleavage, presumably due to greater facility for generating a tertiary cation in the latter case.32b Cleavage of this type is greatly suppressed by addition of manganous ion. Small amounts of cleavage products can be observed with hindered carbinols, but this is a problem only when a stable carbocation can be produced.33 OH RCHO

+

H+

R

HCrO4–

C

H

O O

CrO3H

R

H

C O H

+ CrIV

O H

6

or OH R

C H

OH O

OH Cr

O

O

R

C

+

CrIV

O

7

Oxidation of primary alcohols leads to aldehydes in moderate-to-good yield. Aldehydes are relatively easy to oxidize to the corresponding carboxylic acid, however. In most cases, the oxidation can be stopped at the aldehyde, but small amounts of the acid are a common byproduct. Heating and long reaction times lead to increased amounts of the acid, and in some cases the aldehyde is the minor product. Where feasible, removal of the aldehyde as it forms will minimize side reactions. When the reaction is pushed to give the carboxylic acid, there are at least two reasonable mechanistic rationales for this conversion.34 Both mechanisms involve formation of a chromate ester. Removal of the α-hydrogen (analogous to the alcohol-to-aldehyde conversion) either by an external base (as in 6), or intramolecularly (as in 7), generates the carboxylic acid. This oxidation may also occur by initial formation of a hydrate, followed by removal of the α-hydrogen from 8. This mechanism yields the acid moiety, although a radical process that generates 9 is also possible.34

31

Reference 12, p 142.

32

(a) Reference 12, p 143 and Ref. 4 cited therein; (b) Hampton, J.; Leo, A.; Westheimer, F. H. J. Am. Chem. Soc. 1956, 78, 306.

33

Lee, D. G. In Oxidation, Vol. 1, Augustine, R. L, Ed.; Marcel-Dekker: New York, NY, 1969; p 58.

34

Reference 12, p 174.

221

6.2 ALCOHOLS TO CARBONYLS (CHdOH ! C]O)

O R

C

+ H2O

R

C

H

OH H OH

O HO

8 OH R

C

OH

H

R

OH

C•

OH

+ CrIV

O

OH

HCrO4–

C

Cr O

or

R

OH CrIV

R

+ CrV

C O

OH 9

With some primary alcohols, it is possible to establish an equilibrium between the initially formed aldehyde and the precursor alcohol, in which unreacted alcohol is converted to a hemiacetal (10), which can be further oxidized to an ester.35 This transformation is often observed in the oxidation of lower molecular weight primary alcohols. Oxidation of butan-1-ol to butyl butyrate with Cr(VI), for example, is an Organic Syntheses preparation.36 RCH2OH

+

RCHO

CrVI

OH RCH2OH

+

RCHO

RHC

OCH2R

10 OH RHC

+

RCO2CH2R

CrVI

OCH2R

Some of the problems that accompany oxidation in water can be alleviated or moderated by using an organic cosolvent. Oxidation in aqueous acetic acid will suppress the oxidative cleavage to some extent, as well as greatly improving the solubility of the organic substrate. It is possible to oxidize a wider variety of alcohols in this solvent. Bowman et al.37b found that CrO3 in anhydrous acetic acid was superior to CrO3 in aqueous acetic acid for the oxidation of saturated alcohols. Oxidation of allylic alcohols in anhydrous acetic acid gave yields that were comparable to those obtained in aqueous media.37 This improvement may be due to formation of an acetyl chromate ion (AcOCrO3
), analogous to chromate complexes formed with other acids (see HCrO3A). Such a species may increase the electron-accepting power of Cr.16,38

H2CrO4 + AcOH

AcOCrO3– + H2O

There are several variations in Cr(VI) reagents and reaction conditions that can be used for the oxidation of alcohols. Variations include modification of the solvent, as well as structural modification of the Cr reagent. These modifications allow more control in oxidation reactions of alcohols. 6.2.1.2 Jones Oxidation Organic solvents other than acetic acid can be used with aqueous chromium trioxide. Acetone has been used as a cosolvent in a dilute sulfuric acid solution, and oxidation of alkynyl carbinols was improved when compared to other procedures known at that time. Secondary alcohols are oxidized to ketones, and primary alcohols can be oxidized to

35

(a) Reference 12, p 143; (b) Mosher, W. A.; Preiss, D. M. J. Am. Chem. Soc. 1953, 75, 5605.

36

Robertson, G. R. Org. Synth. Coll. 1941, 1, 138.

37

(a) Reference 12, p 152; (b) Bowman, M. I.; Moore, C. E.; Deutsch, H. R.; Hartman, J. L. Trans. Kentucky Acad. Sci. 1953, 14, 33 (Chem. Abstr. 48:1250b, 1954). 38

(a) Cohen, M.; Westheimer, F. H. J. Am. Chem. Soc. 1952, 74, 4387; (b) Symons, M. C. R. J. Chem. Soc. 1963, 4331.

222

6. FUNCTIONAL GROUP EXCHANGE REACTIONS

either an aldehyde or a carboxylic acid. This chromium trioxide/acetone/sulfuric acid reagent is often referred to as the Jones reagent, and oxidation of alcohols with this reagent is called Jones oxidation.39 Jones oxidation is especially useful for molecules that contain alkenyl or alkynyl groups.40 Oxidation of alcohols is usually faster in acetone than in acetic acid, and using a large excess of acetone protects the ketone product from further oxidation.41 An example is the oxidation of testosterone to give ketone 11 in 73% yield, taken from a synthesis of formestane by Martin et al.42 The secondary alcohol was oxidized to the ketone and the primary alcohol was simultaneously oxidized to the carboxylic acid. OH

O CrO3, Acetone/H2O

H

H

H2SO 4 , 20°C

H

H

H

O

H

O 11 (73%)

Testosterone O

O CrO3, Acetone/H2O

HO

O

H2SO 4

O

2-(Hydroxymethyl)-4H-pyran-4-one

O

OH 4-Oxo-4H-pyran-2-carboxylic acid (61%)

Despite the propensity of aldehydes to be oxidized further with Cr(VI), many primary alcohols can be oxidized to aldehydes using Jones’ oxidation,43 especially with low temperatures and short reaction times, but the reaction conditions are strongly acidic and the final product is sometimes a carboxylic acid. Direct conversion of a primary alcohol to an acid is indeed possible, as in the oxidation of 2-(hydroxymethyl)-4H-pyran-4-one to 4-oxo-4H-pyran-2-carboxylic acid, in Koert and coworker’s44 synthesis of lodopyridone. However, there are many examples in which Jones oxidation has been used in the presence of reactive functional groups. If the starting material is an aldehyde rather than an alcohol, Jones oxidation yields the carboxylic acid. An example is seen in Hu and Panek’s45 synthesis of (
)-motuporin, in which aldehyde [methyl (4S,5S,E)-4azido-5-methyl-6-oxohex-2-enoate] was oxidized to the acid, (2S,3S,E)-3-azido-6-methoxy-2-methyl-6-oxohex-4-enoic acid in 88% yield. Jones oxidation is compatible with complex molecules that contain a variety of functional groups (e.g., alkenes, tertiary alcohols, esters, ketones, oxetanes, and amides).46 N3

O

N3

O Jones oxidation

H

CO2Me

HO

Me Methyl (4S,5S, E)-4-azido-5methyl-6-oxohex-2-enoate

39

CO2Me Me

(2S,3S, E)-3-Azido-6-methoxy-2methyl-6-oxohex-4-enoic acid (88%)

The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-49.

40

(a) Heilbron I.; Jones, E. R. H.; Sondheimer, F. J. Chem. Soc. 1949, 604; (b) Haynes, L. J.; Heilbron I.; Jones, E. R. H.; Sondheimer, F. J. Chem. Soc. 1947, 1583; (c) Heilbron I. M.; Jones, E. R. H.; Sondheimer, J. J. Chem. Soc. 1947, 1586.

41

Reference 12, p 145.

42

Martin, G. D. A.; Narvaez, J.; Marti, A. J. Nat. Prod. 2013, 76, 1966.

43

(a) Hurd, C. D.; Meinert, R. N. Org. Synth. Coll. 1943, 2, 541; (b) Fossek, W. Monatsh 1881, 2, 614 and 1883, 4, 660; (c) Bouveault, L.; Rousset, L. Bull. Soc. Chim. Fr. 1894, 11, 300; (d) Jacobson, M. J. Am. Chem. Soc. 1950, 72, 1489; (e) Sauer, J. Org. Synth. Coll. 1963, 4, 813.

44

Burckhardt, T.; Harms, K.; Koert, U. Org. Lett. 2012, 14, 4674.

45

Hu, T.; Panek, J. S. J. Am. Chem. Soc. 2002, 124, 11368.

46

Magri, N. F.; Kingston, D. G. I. J. Org. Chem. 1986, 51, 797.

223

6.2 ALCOHOLS TO CARBONYLS (CHdOH ! C]O)

6.2.2 Modified Chromium (VI) Oxidants Apart from changing the solvent, mixing certain additives with a Cr(VI) reagent will generate a new oxidizing reagent in situ. The fundamental Cr(VI) reagent can also be chemically modified prior to being used as an oxidizing agent. This section will examine both of these approaches.47a HO

O

Me

O H O

+

H

O

Me

N

Cr N

CrO3 Py

O

H O H

O

O O

O 12

13

14 (89%)

6.2.2.1 Chromium Trioxide-Pyridine In 1948, Sisler et al.47b isolated and characterized a reasonably stable complex from the reaction of chromium trioxide and pyridine. Sisler did not use this reagent for the oxidation of organic molecules, but Sarett and coworkers48 recognized its utility in the synthesis of steroids. In this connection, the alcohol unit in 12 was oxidized to 14 in 89% yield. The reagent, which probably has the trigonal-bipyramidal structure shown in 13, proved useful for the oxidation of primary and secondary alcohols even in the presence of double bonds and thioethers. The oxidation usually required Py as a solvent and was referred to as Sarett oxidation49 for many years. Care must be exercised in preparing the reagent. The solution “must always be prepared by cautious addition of chromium trioxide to pyridine, which has been carefully purified by distillation from potassium permanganate.”50 Reversing this order of addition may cause the mixture to ignite spontaneously.51

A second problem with Sarett oxidation was the difficulty in isolating products from pyridine, which was typically used in excess or even as the solvent. On the other hand, an advantage of the technique, as mentioned above, is that alkenes, ketals, sulfides, and tetrahydropyranyl ethers are oxidized much slower than alcohols and competitive side reactions are rare.52 Oxidation of secondary alcohols is usually facile, but oxidation of primary aliphatic alcohols is sluggish, and low yields of the aldehyde are common.53 A modification, introduced by Collins et al.,54 was applied to the oxidation of alcohols and has come to be known as Collins oxidation. This modification was developed to deal with the problem of poor yields in the oxidation of primary alcohols to aldehydes, and to improve the isolation of the carbonyl products. The reagent formed by reaction of chromium trioxide and pyridine was first removed from the pyridine solvent and added to dichloromethane, and this mixture was treated with the alcohol. This procedure proved to be a superior reagent for the oxidation of alcohols, especially sensitive allylic alcohols. This modified oxidation typically required a 5:1 or 6:1 ratio of complex to alcohol, and reaction occurred at ambient temperatures.54 Cyclohexanol was oxidized to cyclohexanone in 98% yield by this method, and heptan-1-ol was oxidized to heptanal in 93% yield. The yield of aldehyde products from primary alcohols was significantly better, but a large excess of the reagent was required and problems relating to isolation of the product persisted. Nonetheless, Collins oxidation became a mainstay of organic synthesis. An example is taken from the synthesis of hybocarpone 47

(a) For a review of alcohol oxidation using oxochromium(VI)-amine reagents, see Luzzio, F. A. Org. React. 1998, 53, 1; (b) Sisler, H. H.; Bush, J. D.; Accountius, O. E. J. Am. Chem. Soc. 1948, 70, 3827.

48

Poos, G. I.; Arth. G. E.; Beyler, R. E.; Sarett, L. H. J. Am. Chem. Soc. 1953, 75, 422.

49

The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-83.

50

Reference 33, p 60.

51

Holum, J. R. J. Org. Chem. 1961, 26, 4814.

52

Reference 12, pp 154–158.

53

Korytnyk, W.; Kris, E. J.; Singh, R. P. J. Org. Chem. 1964, 29, 574.

54

Collins, J. C.; Hess, W. W.; Frank, F. J. Tetrahedron Lett. 1968, 3363.

224

6. FUNCTIONAL GROUP EXCHANGE REACTIONS

by Nicolaou and Gray,55 in which oxidation of 15 gave ketone 16 in 86% yield. Allylic alcohols are easily oxidized to the corresponding conjugated ketone or aldehyde with this reagent, and Urones and coworkers56 used this type of oxidation in a synthesis of (
)-hyrtiosal. Collins oxidation can be used with groups that are more sensitive to hydrolysis or to over oxidation. These include lactols, as illustrated by the Ando et al.57 conversion of lactol 17 to lactone 18. Ratcliffe et al.58 further improved this procedure by directly preparing solutions of 17 with the chromium trioxide-pyridine complex. The procedure used today typically involves addition of chromium trioxide to dichloromethane and pyridine, followed by addition of the alcohol, but it is still referred to as Collins oxidation. OMe OH

OMe O

Me CO2Me

Me

CrO3•2 Py, CH2Cl2

CO2Me

MeO

MeO OMe

OMe

15

16 (86%)

Me

Me Collins'

Me Me

Me

oxidation

O

O

Me

O

OH 18

17

An example of the oxidation of a non-benzylic secondary alcohol is taken from Kahn and Ahmad’s59 synthesis of convolutamine H, in which 19 was treated with chromium trioxide and pyridine in dichloromethane at room temperature to give 20 in 84% yield. Oxidation of benzylic and allylic alcohols are known to give good yields of the corresponding carbonyl compound. MeO

OMe Br

Br

MeO Br

CrO3, Py, rt

Br

OBn OH

Br

CH2Cl2, rt

19

Br

Br

OMe Br OBn O 20 (84%)

6.2.2.2 Pyridinium Chlorochromate and Pyridinium Dichromate The need for improved selectivity in the oxidation of primary alcohols, and greater ease for isolation of products, prompted further research into the nature of Cr(VI) reagents. Corey and Suggs60 found that addition of pyridine to a solution of chromium trioxide in aq HCl led to crystallization of a solid reagent pyridinium chlorochromate (PCC). This reagent was shown to be superior for the conversion of primary alcohols to aldehydes, but less efficient than the Collins oxidation when applied to allylic alcohols using dichloromethane as the solvent.61 Oxidation of heptan-1-ol with PCC in dichloromethane gave 78% of heptanal, for example. Another example is taken from Isobe and Chuang’s62 synthesis of the right segment of solanoeclepin A, in which ent-Hajos-Parrish ketone was oxidized by PCC to ketone 21 in 73% yield. As stated by Corey, PCC is an effective oxidant in dichloromethane although aqueous chlorochromate species are not very effective oxidants.61 Oxidation of secondary alcohols to ketones is straightforward. 55

Nicolaou, K. C.; Gray, D. L. F. J. Am. Chem. Soc. 2004, 126, 607.

56

Basabe, P.; Diego, A.; Díez, D.; Marcos, I. S.; Urones, J. G. Synlett 2000, 1807.

57

(a) Ando, M.; Akahane, A.; Takase, K. Bull. Chem. Soc. Jpn. 1978, 51, 283; (b) Ando, M.; Tajima, K.; Takase, K. Chem. Lett. 1978, 617.

58

Ratcliffe, R.; Rodehorst, R. J. Org. Chem. 1970, 35, 4000.

59

Khan, F. A.; Ahmad, S. J. Org. Chem. 2012, 77, 2389.

60

Corey, E. J.; Suggs, J. W. Tetrahedron Lett. 1975, 2647.

61

(a) Banerji, K. K. Bull Chem. Soc. Jpn. 1978, 51, 2732; (b) Banerji, K. K. J. Chem. Soc. Perkin Trans. 1978, 2, 639.

62

Chuang, H. -Y.; Isobe, M. Org. Lett. 2014, 16, 4166.

225

6.2 ALCOHOLS TO CARBONYLS (CHdOH ! C]O)

OH

O CH2Cl2

N+

ClCrO3–

+

NaOAc

H

AcO

PCC

AcO 21 (73%)

ent-Hajos-Parrish ketone OH

OH

PCC, CH2Cl2

+

22

23

O

24 (90%)

A drawback to the use of PCC is its mild acidity. Note the use of sodium acetate to buffer the oxidation of ent-Hajos-Parrish ketone to 21. It is inferior to Collins oxidation when the oxidation products or the starting alcohol are acid sensitive. This acidity can be used to synthetic advantage, however, and Corey and Boger63 described its use to promote oxidative cationic cyclizations. The reaction of citronellol with PCC in dichloromethane was accompanied by a cationic cyclization reaction to yield (
)-pulegone.60,64a Oxidative transpositions can also occur.64b In studies toward a synthesis of guanacastepene A, Mehta et al.65 oxidized a mixture of 22 and 23 with PCC to give a 90% yield of the rearranged ketone 24. Oxidations using PCC on alumina66 have also been reported, presumably to minimize acidcatalyzed rearrangement. The PCC reagent has been used for the conversion of oximes to ketones,67 and for the direct oxidation of tetrahydropyranyl protected alcohols to aldehydes.68 Problems caused by the acidity of PCC can be largely eliminated, by using the more neutral reagent pyridinium dichromate (PDC), 26. Although first used by Coates and Corrigan,69 it was not exploited synthetically until Corey prepared the reagent by addition of Py to neutral chromium trioxide solutions and used it for the oxidation of alcohols to aldehydes, ketones, and acids.70 The reagent is not acidic and the neutral conditions required for the oxidation are superior for the oxidation of allylic alcohols. In dichloromethane, with a non-aqueous workup, oxidation of alcohols is similar to that of PCC. Addition of catalytic amounts of pyridinium trifluoroacetate in dichloromethane significantly increases the rate of oxidation. Allylic alcohols are oxidized faster than aliphatic alcohols, making PDC the reagent of choice for this transformation. Cyclohexenol, for example, is oxidized 10 times faster than cyclohexanol with PDC in dichloromethane at 25°C.70 Primary aliphatic alcohols are also oxidized under very mild conditions.71 This oxidation occurs under essentially neutral conditions.70 Even hindered alcohols can be oxidized with PDC, as in the synthesis of ansamycin by Kirschning and J€ urjens,72 which required the oxidation of 25 to 27. Many reagents were examined, but yields were poor in all cases. Heating for an extended time with PDC led to a 46% yield of 27.

63

Corey, E. J.; Boger, D. L. Tetrahedron Lett. 1978, 2461.

64

(a) Corey, E. J.; Ensley, H. E.; Suggs, J. W. J. Org. Chem. 1976, 41, 380; (b) Dauben, W. G.; Michno, D. M. J. Org. Chem. 1977, 42, 682. Also see, (c) Corey, E. J.; Ha, D. C. Tetrahedron Lett. 1988, 29, 3171; (d) Waddell, T. G.; Carter, A. D.; Miller, T. J.; Pagni, R. M. J. Org. Chem. 1992, 57, 381; (e) Luzzio, F. A.; Guziec, Jr., F. S. Org. Prep. Proceed. Int. 1988, 20, 533; (f ) Schlecht, M. F.; Kim, H. -J. J. Org. Chem. 1989, 54, 583.

65

Mehta, G.; Jayant D. Umarye, J. B. Org. Lett. 2002, 4, 1063.

66

For an example taken from a synthesis see Moreno-Dorado, F. J.; Guerra, F. M.; Aladro, F. J.; Bustamante, J. M.; Jorge, Z. D.; Massanet, G. M. J. Nat. Prod. 2000, 63, 934.

67

Maloney, J. R.; Lyle, R. E.; Saavedra, J. E.; Lyle, G. G. Synthesis 1978, 212.

68

Sonnet, P. E. Org. Prep. Proceed. Int. 1978, 10, 91.

69

Coates, W. M.; Corrigan, J. R. Chem. Ind. 1969, 1594.

70

Corey, E. J.; Schmidt, G. Tetrahedron Lett. 1979, 399.

71

Boeckman, Jr., R. K.; Perni, R. B. J. Org. Chem. 1986, 51, 5486.

72

J€ urjens, G.; Kirschning, A. Org. Lett. 2014, 16, 3000.

226

6. FUNCTIONAL GROUP EXCHANGE REACTIONS

Me

Me OTBS

TBSO

OTBS TBSO

OH Me

O

O

CH2Cl2, MS 3 Å , K2CO3 120°C , 20 d

H

OMe

NH Oi-Pr

OMe

NH

H

Oi-Pr

Cr2O72–

N+ 2

Me

i-PrO

O Me

Me

i-PrO

OMe

OMe

25

26

27 (46%)

In DMF, oxidation with PDC can lead to a significant difference in product formation in the oxidation of aliphatic primary alcohols versus that of allylic primary alcohols. Cyclohexenol is oxidized to cyclohexenone in 86% yield with PDC in DMF at 0°C.70 Similarly, geraniol was oxidized to geranial in 92% yield. However, treatment of nonconjugated primary alcohols with PDC in DMF (ambient temperature) typically gave carboxylic acids rather than an aldehyde. For example, in Tokuyama and coworker’s synthesis of (
)-mersicarpine,73 oxidation of 28 with PDC in DMF at 25°C gave carboxylic acid 29 in >60% yield. Aldehydes can also be oxidized to the carboxylic acid, as in Williams and Jain’s synthesis of (+)-negamycin,74 where the aldehyde unit in 30 was converted to carboxylic acid 31 in 97% yield. CO2Bn HO

CO2Bn PDC, DMF

HO

25°C

O 28

29 (>60%)

Ph

BnO2C

Ph Bn

Ph

O N

N

CO2Bn

PDC, DMF, 25°C

CHO

Bn

Ph

BnO2C

30

O

N

CO2Bn

N

CO2H 31 (97%)

6.2.2.3 Structurally Modified Chromium Reagents The literature is full of chromium trioxide complexes developed by adding amine or phosphine bases, usually with the goal of varying the specificity of the oxidation. In many cases, stabilized chromate esters have been prepared and found to be effective oxidizing agents. If the oxidizing power of these modified reagents can be generalized at all, it may be said they behave similarly to PCC or PDC. Some examples are 32,75 33,76 34,77 35,78 and 36.79 The reasons for developing modified reagents can vary with the application. In many instances, they were developed to alleviate problems associated with isolation of products, or the use of a large excess of the oxidizing reagent (as with Sarett and Collins oxidations). In other cases, the need for decreased acidity, for improved solubility in low-boiling aprotic

73

Iwama, Y.; Okano, K.; Sugimoto,K.; Tokuyama, H. Chem. Eur. J. 2013, 19, 9325.

74

Jain, R. P.; Williams, R. M. J. Org. Chem. 2002, 67, 6361.

75

Sharpless, K. B.; Akashi, K. J. Am. Chem. Soc. 1975, 97, 5927.

76

López, C.; González, A.; Cossío, F. P.; Palomo, C. Synth. Commun. 1985, 15, 1197.

77

Corey, E. J.; Barrette, E. P.; Magriotis, P. A. Tetrahedron Lett. 1985, 26, 5855.

78

(a) Davis, H. B.; Sheets, R. M.; Brannfors, J. M.; Paudler, W. W.; Gard, G. L. Heterocycles 1983, 20, 2029; (b) Firouzabadi, H.; Iranpoor, N.; Kiaeezadeh, F.; Toofan, J. Tetrahedron 1986, 42, 719. 79

Cristau, H. -J.; Torreilles, E.; Morand, P.; Christol, H. Tetrahedron Lett. 1986, 27, 1775.

227

6.2 ALCOHOLS TO CARBONYLS (CHdOH ! C]O)

organic solvents, or for increasing (or decreasing) the relative oxidizing power of the Cr reagent provided the driving force for developing the new reagent. For 32 and 34, chromyl chloride (CrO2Cl2) was the Cr precursor rather than chromium trioxide. HO2C

O

O

O + Cr2O72– NJH

Cr O

O Cr O

O

O

2

33

32 HJN

34 Ph3P

H

+N

H

ClCrO3 –

+ Cr2O72–

Ph3P+

35

2

36

Another modification used pyridine derivatives as bases, giving the PDC analogue 33, and the use of piperazine led to the PCC analogue, 35. Using a phosphine as the base, [e.g., bis(phosphine) 1,1-bis(triphenyl)phosphinomethane] generated the PDC analogue 36. Oxidation with modified Cr reagents is similar to PCC and PDC oxidations. Reagent 34 was developed by Corey et al.77 and is of particular interest in that it can be used as a catalytic oxidizing agent (only 2% of 34 was required for the oxidation of cyclooctanone from cyclooctanol) in the presence of peroxyacetic acid (in dichloromethane). The functional group transforms for the oxidations in this section follow: R

R O

R(H)

O OH

R(H)

R

OH R

OH

6.2.3 Oxidation With Dimethyl Sulfoxide-Based Reagents Reagents other than Cr(VI) can be used to oxidize alcohols. It has been shown by a number of workers that DMSO based reagents react with primary or secondary alcohols to form a sulfoxonium intermediate, where the β-hydrogen can be removed,80 and dimethyl sulfide (Me2S, abbreviated DMS) functions as the leaving group. This overall process leads to oxidation of alcohols and there are many variations of this reaction. The oxidation of alcohols using derivatives of the transition metal Cr(VI), which often required acidic conditions, was discussed in Section 6.2.2. However, many alcohols, especially those bearing sensitive functional groups, require neutral oxidizing conditions. The development of PDC and related reagents helped to alleviate this problem, but PCC is an acidic reagent. Either PDC or PCC sometimes give poor yields, especially with hindered alcohols. The key to Cr(VI) oxidations was formation of a chromate ester. In one sense, the chromate ester behaved as a leaving group and removal of the hydrogen atom that is β to the Cr makes the reaction conceptually analogous to an E2 reaction (Section 3.5.1). The problem with this analogy is, of course, that there is no direct evidence for removal of the hydrogen atom in 37 by an external base.25,26 Removal of the hydrogen atom via an intramolecular process is also viable, as in 38, which has a close analogy to the syn-elimination reaction discussed in Section 3.7. C C B:

C O CrO H 3 H 37

O

H C

C

O

C

O

Cr OH

38

Dimethyl sulfoxide functions as both a solvent and a reactant for a variety of alcohol substrates. The nucleophilic oxygen can react with electrophilic centers to form a sulfoxonium salt, which gives ketones or aldehydes under neutral conditions. Kornblum et al.81 first observed this process, but with halides rather than with alcohols. The reaction of 80

Moffatt, J. G. In Oxidation, Vol. 2, Augustine, R. L.; Trecker, D. J., Eds.; Marcel-Dekker Inc.: New York, NY, 1971; pp 1–64.

81

Kornblum, N.; Powers, J. W.; Anderson, G. J.; Jones, W. J.; Larson, H. O.; Levand, O.; Weaver, W. M. J. Am. Chem. Soc. 1957, 79, 6562.

228

6. FUNCTIONAL GROUP EXCHANGE REACTIONS

α-halo ketones with DMSO at elevated temperatures gave good yields of the corresponding glyoxal (an α-ketoaldehyde). Using this procedure, phenylacetyl bromide (2-bromo-1-phenylethan-1-one) was oxidized to 2-oxo-2-phenylacetaldehyde in 71% yield81 in what has been called the Kornblum aldehyde synthesis.82 The contact times were usually very short, and if the glyoxal could be removed from the reaction medium by distillation as it was formed, the reaction was very efficient. It was often difficult to isolate high-boiling glyoxals from DMSO, however. O

O DMSO

Br

Ph

Ph

> 100°C

2-Bromo-1-phenylethan-1-one

CHO

2-Oxo-2-phenylacetaldehyde (71%)

This transformation is not limited to α-halo ketones. Primary83 and secondary84 alkyl iodides or tosylates can be converted to aldehydes or ketones, although they are much less reactive than α-halo ketones. Reaction of 1-bromooctane with DMSO and NaHCO3 at 100°C, for example, gave octanal in 74% yield after 5 min.83 This outcome can be compared with the conversion of an alkyl tosylate to a ketone, which required temperatures of 150–170°C and the presence of an added base (e.g., sodium bicarbonate). The added base reacts with any acid as it is produced. Elimination is a serious side reaction, and attempts to oxidize secondary α-halo ketones often give poor yields.85 The mechanism probably involves nucleophilic displacement of halides by DMSO to form an alkoxysulfoxonium salt, 39. The dimethyl sulfide moiety is a good leaving group, and by analogy with 37, the α-hydrogen atom (Ha) is sufficiently acidic that DMSO will react in an acid-base reaction as shown. Removal of the α-proton, probably by DMSO, leads to the carbonyl product and DMS, which functions as the leaving group.86 This reaction can be used in synthetic applications, as in the synthesis of fragment A of brystatin 7 by Krische and coworkers,87 in which bromoketone 40 was converted to glyoxal 41 in 81% yield. R R

+

R

O Me

S Me

R

X

–O

Me

+

:B

R

R

Ha

S

+

Me

S

Me

+

+

Ha B

O

Me Dimethyl sulfoxide

Dimethyl sulfide

39

O

O Br

TBSO

1. 3 AgNO3, MeCN, 25°C 2. 4 NaOAc, DMSO, 25°C

H

TBSO O

40

41 (81%)

A similar type of oxidation was observed when alcohols reacted with DMSO. Traynelis and Hergenrother88 showed that benzylic and allylic alcohols were converted to the corresponding aldehyde in high yield by refluxing in DMSO, with air bubbling through the medium. Air was the oxidant in this reaction, DMSO was the solvent, and under these conditions cinnamyl alcohol was oxidized to cinnamaldehyde in 90% yield.88 It is known that oxidation can occur in the absence of air or oxygen, but a coreagent is usually required. There are several variations of this fundamental reaction involving DMSO. A mixture of DMSO and different coreagents react with alcohols to form a complex that leads to formation of aldehydes and ketones. In each case, a key reagent is added to activate the DMSO-alcohol oxidation reaction. Several variations will be presented in subsequent sections, and most are named reactions.

82

Mundy, B. P.; Ellerd, M. G.; Favaloro, Jr., F. G. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience, NJ, 2005; pp 376–377.

83

(a) Kornblum, N.; Jones, W. J.; Anderson, G. J. J. Am. Chem. Soc. 1959, 81, 4113; (b) Nace, H. R.; Monagle, J. J. J. Org. Chem. 1959, 24, 1792.

84

Baizer, M. M. J. Org. Chem. 1960, 25, 670.

85

(a) Jones, D. N.; Saeed, M. A. J. Chem. Soc. 1963, 4657; (b) Iacona, R. N.; Rowland, A. T.; Nace, H. R. J. Org. Chem. 1964, 29, 3495.

86

Hunsberger, I. M.; Tien, J. M. Chem. Ind. 1959, 88.

87

Lu, Y.; Woo, S. K.; Krische, M. J. J. Am. Chem. Soc. 2011, 133, 13876.

88

Traynelis, V. J.; Hergenrother, W. L. J. Am. Chem. Soc. 1964, 86, 298.

229

6.2 ALCOHOLS TO CARBONYLS (CHdOH ! C]O)

6.2.3.1 Swern Oxidation This variation is quite common in synthesis and was discovered by Mancuso and Swern,89 who found that DMSO could be activated for the oxidation of alcohols by the addition of trifluoroacetic anhydride. This reaction is usually done in dichloromethane at temperatures below
30°C. In the absence of a moderating solvent, the admixture of DMSO and (CF3CO)2O proceeds explosively.

The initially formed trifluoroacetyl derivative (42) is stable below
30°C and at these low temperatures 42 reacts with the alcohol to yield the requisite sulfoxonium intermediate, 39. Dimethyl sulfoxide reacts as a base with this intermediate to yield an aldehyde or a ketone and DMS. This mechanism is common for virtually all of the oxidations presented in this section, but the reagent used to generate 39 varies. Temperature control is important in this reaction, in part because the complex 42 is thermally unstable. It is known that at higher temperatures (>
30°C) 42 undergoes a Pummerer rearrangement to yield 43.90 > –30°C

O O Me

S

(CF3CO)2O

+S

Me

CH2Cl2

Me

Me Me

O

–O

S

CF3

CF3 O

43

2CCF3

42

O

OH R

H

R

R

R

R O

R

+ Me S Me + H

H

Me S +

:B

Me

B+

O

39

An important and interesting feature of the Swern oxidation is a low sensitivity to steric hindrance. Oxidation of the moderately hindered alcohol, 2,4-dimethylpentan-3-ol, yielded 2,4-dimethylpentan-3-one in 86%.88 Oxidation of a typically unhindered primary alcohol to the corresponding aldehyde is facile, however. Amines are often added to facilitate decomposition of the initially formed complex. Oxidation of decan-1-ol gave a 56% yield of decanal when triethylamine was added. Addition of diisopropylamine rather than triethylamine increased the yield to 81%.91 Swern oxidation can be used in highly functionalized molecules, and as mentioned it tolerates alcohols that are somewhat sterically hindered. Swern also found that oxalyl chloride activates DMSO for the oxidation of alcohols. The resulting reagent is superior to the DMSO-trifluoroacetic anhydride reagent.92 The reaction probably proceeds via intermediate complex 44, which is unstable above the preferred reaction temperature of
60°C. Oxalyl chloride and DMSO react violently and exothermically at ambient temperatures.

This reaction proceeds with loss of both CO2 and CO to yield chlorosulfonium salt 45.88 In the presence of an alcohol, 45 reacts to yield 39, and deprotonation yields the carbonyl and DMS. Cyclododecanol, for example, was oxidized to cyclododecanone in 97% yield with this reagent, and hex-2-en-1-ol was oxidized to the conjugated aldehyde hex-2-enal in good yield.88 Oxidation with DMSO and oxalyl chloride is called Swern oxidation,93 although this term can also be used with the DMSO-trifluoroacetic anhydride oxidation. An odorless Swern oxidation protocol has been developed using dodecyl methyl sulfide.94

89

Mancuso, A. J.; Swern, D. Synthesis 1981, 165.

90

(a) Sharma, A. K.; Swern, D. Tetrahedron Lett. 1974, 1503; (b) Sharma, A. K.; Ku, T.; Dawson, A. D.; Swern, D. J. Org. Chem. 1975, 40, 2758; (c) Pummerer, R. Berichte 1910, 43, 1401.

91

Huang, S. L.; Omura, K.; Swern, D. Synthesis 1978, 297.

92

(a) Omura, K.; Swern, D. Tetrahedron 1978, 34, 1651; (b) Mancuso, A. J.; Huang, S. -L.; Swern, D. J. Org. Chem. 1978, 43, 2480.

93

The Merck Index, 14th ed.; Merck & Co., Inc.; Whitehouse Station, NJ, 2006; p ONR-92.

94

Ohsugi, S. -i.; Nishide, K.; Oono, K.; Okuyama, K.; Fudesaka, M.; Kodama, S.; Node, M. Tetrahedron 2003, 59, 8393.

230

6. FUNCTIONAL GROUP EXCHANGE REACTIONS O

O

Me O

S

Me

O

Cl

Cl

+S

CH2Cl2 –60°C

Me

Me

Cl

O –Cl

Me

– CO2

+

S – CO

Cl

Me

O

44

45

A typical synthetic use involving oxalyl chloride and DMSO is the conversion of the two primary alcohol units in 46 to dialdehyde 47 in 73% yield, taken from Diethelm and Carreira’s95 synthesis of gelsemoxonine. In a synthesis of phorboxazole A by Williams et al.,96 trifluoroacetic anhydride and DMSO were used to oxidize the primary alcohol unit in 48 to aldehyde 49, in >90% yield. In many synthetic sequences, one will find the term Swern oxidation or simply Swern over a reaction arrow rather than giving the reagents. This practice is quite common, forcing the reader to recognize this named reaction. O

BocO

O

BocO OH

N

H

CHO

(COCl) 2 , DMSO

CHO

NEt3, CH2Cl2 –78°C to rt

OH

N

Boc

H

Boc 46

47 (73%) OSiPh2t-Bu

O2Ct-Bu

O

O

OH

(CF3CO)2O, DMSO

N

O2Ct-Bu

O

O

O

N

CH2Cl2, –78°C

O OSiPh2t-Bu

OSiPh2t-Bu

O OSiPh2t-Bu

48

49 (>90%)

The functional group transform for DMSO-type oxidations follows: R

R OH

O R

R

6.2.3.2 Moffatt Oxidation Another DMSO based oxidation predates the Swern oxidation, and uses a mixture of DMSO and 1,3-(dicyclohexylcarbodiimide) (DCC) in the presence of an acid catalyst to generate an intermediate (e.g., 39).97 This reaction is known as Moffatt oxidation,97 and an example is the conversion of 50 to 51, used in Crick and coworker’s98 synthesis of capuramycin. In this particular example, 51 was not isolated but rather reacted with TMSCN/Ti(Oi-Pr)4 and then aq acetic acid to give the cyanohydrin in 90% overall yield from 50. Moffatt and coworkers,97,99 provided a mechanistic rationale for this oxidation, where the driving force of the reaction is formation and separation of the highly insoluble dicyclohexylurea. The initial reaction of DMSO and DCC forms sulfoxonium intermediate 52, which now contains a urea leaving group. When the alcohol attacks the electrophilic sulfur atom in 52, dicyclohexylurea is displaced to generate the usual sulfoxonium salt 39. Deprotonation of the sulfoxonium salt in this case may generate a sulfur ylid (53; see Section 12.5.2) that is stabilized by the d-orbitals of sulfur.100 The carbanion center in 53 probably removes the α-hydrogen intramolecularly (as shown), although the intermolecular reaction alternative is also possible (analogous to 39). By either mechanism, DMS is lost and the 95

Diethelm, S.; Carreira, E. M. J. Am. Chem. Soc. 2013, 135, 8500.

96

Williams, D. R.; Kiryanov, A. A.; Emde, U.; Clark, M. P.; Berliner, M. A.; Reeves, J. T. Angew. Chem. Int. Ed. 2003, 42, 1258.

97

Pfitzner, K. E.; Moffatt, J. G. J. Am. Chem. Soc. 1965, 87, 5661.

98

Kurosu, M.; Li, K.; Crick, D. C. Org. Lett. 2009, 11, 2393.

99

Reference 80, p 12.

100

(a) Cilento, G. Chem. Rev. 1960, 60, 147; (b) Johnson, C. R.; Phillips, W. G. J. Am. Chem. Soc. 1969, 91, 682.

231

6.2 ALCOHOLS TO CARBONYLS (CHdOH ! C]O)

aldehyde or ketone is formed. Removal of the hydrogen as suggested by 53 is formally analogous to the synelimination mechanism discussed in Section 3.7. An alternative mechanism involving complex formation of the alcohol and DCC prior to its reaction with DMSO has not been observed.101 O

O N KC KN

N-BOM N

O

MeO

O

N-BOM

H

DCC

HO

O

N

O

DMSO, Cl2CHCO2H CH2Cl2

OAc

MeO

O

OAc

50

51 (90%)

The initial reaction with DMSO is usually fast, but overall oxidation is quite slow unless mineral acids or strong organic acids are added to the reaction.102 o-Phosphoric acid,103 dichloroacetic acid, and pyridinium trifluoroacetate are the most common additives used in this oxidation.104 Indeed, pyridinium salts of strong acids are good catalysts, particularly the pyridinium salts of orthophosphoric acid and trifluoroacetic acid. Reagents formed in the presence of these acids gave clean and rapid oxidation of primary and secondary alcohols in the Moffatt oxidation. H+ O N

N DMSO

+

N

H+

S R1

H3C

H

N H

Dicyclohexylurea

H

O

N K C KN

DCC

H

C

CH3 O

+

CH3

R1 R2

C

H

H

O R2

S H H

H 52

39 R1

CH3 O

R2

H

S C

53

H H

R1 C R2

CH3 O

+

S CH3 DMS

Moffatt oxidation has several problems associated with it. One is formation of dicyclohexylurea, which can be very difficult to separate from the other products. Treatment with oxalic acid is probably the most common method employed for the removal of the last vestiges of the dicyclohexylurea byproduct.105 In addition, the oxidation of homoallylic alcohols is sometimes accompanied by isomerization of the double bond into conjugation with the carbonyl group. This isomerization can be minimized or prevented by addition of pyridinium trifluoroacetate. A threefold excess of DCC and an excess of DMSO are usually required for Moffatt oxidation, and the use of excess reagents requires that they must be removed from the product. The DMSO need not be the solvent, however, and addition of a cosolvent (e.g., ethyl acetate) often leads to better results. Despite the drawbacks, Moffatt oxidation is useful in many synthetic applications. Tertiary alcohols are, of course, resistant to Moffatt oxidation, but dehydration

101

Reference 80, pp 13–15.

102

Reference 80, p 6.

103

Pfitzner, K. E.; Moffatt, J. G. J. Am. Chem. Soc. 1963, 85, 3027.

104

Reference 80, p 7.

105

Reference 80, p 44.

232

6. FUNCTIONAL GROUP EXCHANGE REACTIONS

has been observed when they are treated with Moffatt reagents,106 giving the alkene and regenerating DMSO.105 Some steroidal allylic alcohols dehydrate to a heteroannular diene under these conditions.107 6.2.3.3 Other DMSO Oxidations Albright and Goldman108 developed an oxidizing reagent using DMSO and acetic anhydride that formed an active sulfoxonium complex. Oxidation of alcohols with this reagent been called the Albright-Goldman oxidation. In the initial work, yohimbine was oxidized to the ketone (yohimbinone) in 85% yield, at ambient temperatures in 24 h.107 As with DCC, DMSO initially reacts with acetic anhydride to form sulfoxonium salt 54, which then reacts with the alcohol. Acetate is the leaving group attached to sulfur, and reaction with an alcohol generates 39. The acetate anion serves as the base in this case to remove the α-proton and generate the ketone or aldehyde. O Me

Me

S

O O

Me

R

O

Me

Me

O

S

– AcO-

Me

R2CHOH

O

CH3

S

– AcOH

54

R

H

Me

Me

O

R

O

R

39

A major advantage of this method is the production of water-soluble byproducts, and it is especially effective with hindered alcohols. An example, taken from Jeon and coworker’s109 synthesis of valienamine, is the oxidation of the secondary alcohol moiety in 55 to give a 94% yield of 56 in the presence of the dithioacetal unit. With unhindered alcohols the products are often acetate esters and thiomethoxymethyl esters. OH

OBn

SEt

O

OBn

SEt

DMSO

SEt OBn

OBn

SEt

Ac2O

OBn

OBn

OBn

OBn

56 (94%)

55

Onodera et al.110 developed a reagent for the oxidation of alcohols using DMSO and phosphorus pentoxide. This method is very efficient for the oxidation of the secondary alcohol units found in carbohydrates. The alcohol moiety in 57, for example, was converted to ketone 58 in 65% yield.109 The initial intermediate was probably the sulfoxonium derivative (59), which was attacked by the alcohol. Deprotonation of the alkoxy-sulfoxonium ion in the usual manner gave the ketone. Virtually all examples using this reagent involved carbohydrates. Me Me

Me

O O

Me

OH O

O O

O

DMSO

O

P4O10

O Me

Me

57

O

O O Me

Me

58 (65%)

The complex formed by pyridine-sulfur trioxide was shown to activate DMSO for oxidation, in the presence of a base (e.g., triethylamine).111 The presumed intermediate is 60, which reacts with the alcohol to yield the usual complex 39, and deprotonation by triethylamine yields the ketone or aldehyde. This reaction is sometimes called the Parikh-Doering oxidation.111 Alcohols with different stereochemical and conformational properties can exhibit great 106

Reference 80, p 38.

107

Reference 80, p 39.

108

Albright, J. D.; Goldman, L. J. Am. Chem. Soc. 1967, 89, 2416. Also see Albright, J. D.; Goldman, L. J. Am. Chem. Soc. 1965, 87, 4214.

109

Chang, Y. -K.; Lee, B. -Y.; Kim, D. J.; Lee, G. S.; Jeon, H. B.; Kim, K. S. J. Org. Chem. 2005, 70, 3299.

110

Onodera, K.; Hirano, S.; Kashimura, N. J. Am. Chem. Soc. 1965, 87, 4651.

111

Parikh, J. R.; von E. Doering, W. J. Am. Chem. Soc. 1967, 89, 5505.

233

6.2 ALCOHOLS TO CARBONYLS (CHdOH ! C]O)

differences in their rate of oxidation, which leads to differences in the selectivity for the oxidation of related alcohols. This modification is found in several syntheses of natural products. In the Shi et al.112 synthesis and structural verification of neofinaconitine, oxidation of the alcohol unit in N-benzyl-6-hydroxyhexanamide using the Parikh-Doering procedure proceeded smoothly to give the aldehyde, N-benzyl-6-oxohexanamide, in 93% yield. Me

Me S

OP4O10 –

O S

Me

Me

O–

60

O

H

SO3•Py

NHBn

S O

59 OH

O

O

O NHBn

DMSO, CH2Cl2, NEt3

N-Benzyl-6-oxohexanamide (93%)

N-Benzyl-6-hydroxyhexanamide

6.2.4 Dess-Martin Periodinane Oxidation A mild oxidizing reagent has been developed that contains a hypervalent iodine.113 Dess and Martin114 showed that 2-iodobenzoic acid reacted with KBrO3 in H2SO4 to give a 93% yield of 61. Subsequent heating to 100°C with acetic anhydride and acetic acid gave 62 in 93% yield, the so-called Dess-Martin periodinane [1,1,1-tris(acetyloxy)-1,1-dihydro-1,2beniodoxo-3-9(1H)-one]. The Dess-Martin reagent can be shock sensitive under some conditions, and it explodes >200°C.115 An improved procedure for its preparation is available.116

HO I KBrO3

CO2H

O

AcOH, Ac2O

I

OAc OAc O

100°C

H2SO 4

2-Iodobenzoic acid

AcO

O–

I

O 61 (93%)

O 62 (93%)

Alcohols are readily oxidized to ketones or aldehydes, as in the oxidation of cyclohexanol with 62 to cyclohexanone in 90% yield, in dichloromethane at 25°C.114 In this particular reaction, an intermediate was trapped and its structure shown to be 63.114 The reagent appears to have an indefinite shelf-life in a sealed container, but hydrolysis occurs upon long-term exposure to atmospheric moisture. Hypervalent iodine reagents are known to be oxidizing agents,117 and oxidation of the α-position of methyl ketones to give the corresponding α-hydroxyketone is an important application.118 An inert atmosphere is not required, but addition of catalytic amount of trifluoroacetic acid

112

Shi, Y. Wilmot, J. T.; Nordstrøm, L. U.; Tan, D. S.; Gin, D. Y. J. Am. Chem. Soc. 2013, 135, 14313.

113

For reactions of hypervalent iodine reagents in the synthesis of heterocyclic compounds, see Prakash, O.; Singh, S. P. Aldrichim. Acta 1994, 27, 15.

114

(a) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155. (b) See The Merck Index, 14th ed.; Merck & Co., Inc.; Whitehouse Station, NJ, 2006; p ONR-23.

115

Plumb, J. B.; Harper, D. J. Chem. Eng. News 1990, 68 (29), p 3.

116

(a) Ireland, R. E.; Liu, L. J. J. Org. Chem. 1993, 58, 2899. Also see, (b) Frigerio, M.; Santagostino, M.; Sputore, S. J. Org. Chem. 1999, 64, 4537.

117

Moriarty, R. M.; Prakash, O. Acc. Chem. Res. 1986, 19, 244.

118

(a) Moriarty, R. M.; John, L. S.; Du, P. C. J. Chem. Soc. Chem. Commun. 1981, 641; (b) Moriarty, R. M.; Gupta, S.; Hu, H.; Berenschot, D. R.; White, K. B. J. Am. Chem. Soc. 1981, 103, 686; (c) Moriarty, R. M.; Hu, H.; Gupta, S. C. Tetrahedron Lett. 1981, 22, 1283.

234

6. FUNCTIONAL GROUP EXCHANGE REACTIONS

accelerates the oxidation. The catalyst is not required, however. It has also been shown that a Dess-Martin intermediate can be generated in aqueous media and used for oxidation.119 Dess and Martin120 prepared several other periodinane reagents that also oxidizes alcohols, as well as a chiral hypervalent organoiodinane derived from a chiral binaphthyl.121 There are many synthetic applications in the recent literature.122 The oxidation of a primary alcohol unit in a substrate that can tolerate only a narrow range of reaction conditions, or is part of a highly functionalized molecule, is common.123 In Roush and Chen’s synthesis of the antifungal polyketide (
)-basiliskamide A,124 the primary alcohol unit in 64 was oxidized to an aldehyde (65), in >80% yield. Secondary alcohols are also oxidized under mild conditions. AcO O 62, 25°C

HO

I

CH2Cl2

O

O

OAc

O Cyclohexanol

Cyclohexanone (90%)

63 O

OH

O

O Ph

62 NaHCO3 CH2Cl2

CH3

CH3

O

O

Ph

H CH3

64

CH3 65 (>80%)

6.2.5 Oxidation With TEMPO and Oxammonium Salts Oxammonium salts (e.g., 66, Bobbitt’s reagent)125 are useful oxidizing agents for the selective oxidation of alcohols to aldehydes or ketones. Such salts can be generated catalytically from a nitroxide in the presence of a secondary oxidation procedure, either chemical or electrochemical,126 or with 2 equiv of acid and 2 equiv of a nitroxide.127 When 66 was mixed with 3-phenylprop-2-yn-1-ol, in dichloromethane, 3-phenylpropiolaldehyde was isolated in 93% yield.128 The reaction is readily monitored as the initial yellow slurry changes to a white slurry and the presence of unreacted oxidant can be checked with starch.127 It is not necessary to use anhydrous conditions, and it was discovered that the rate of reaction was enhanced by the presence of silica gel. This reagent is compatible for the mild oxidation of many alcohols, including aliphatic primary and secondary alcohols, as well as allylic and benzylic alcohols, although nitrogencontaining compounds are sometimes problematic unless Py is added to the reaction. Note that the stable free radical TEMPO also oxidizes alcohols in the presence of an acid catalyst. An example is the conversion of colletodiol to grahamimycin A, in 76% yield in a synthesis reported by O’Doherty and Hunter.129

119

Meyer, S. D.; Schreiber, S. L. J. Org. Chem. 1994, 59, 7549.

120

Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113, 7277.

121

Ochiai, M.; Takaoka, Y.; Masaki, Y.; Nagao, Y.; Shiro, M. J. Am. Chem. Soc. 1990, 112, 5677.

122

See (a) Zhdankin, V. V. J. Org. Chem. 2011, 76, 1185; (b) Stang, P. J. J. Org. Chem. 2003, 68, 2997; (c) Zhdankin, V. V.; Stang, P. J. Chem. Rev. 2008, 108, 5299.

123

Evans, D. A.; Sheppard, G. S. J. Org. Chem. 1990, 55, 5192.

124

Chen,M.; Roush, W. R. Org. Lett. 2012, 14, 1556.  G. Bull. Acad. Sci. (a) Bobbitt, J. M.; Flores, M. C. L. Heterocycles 1988, 27, 509. Also see, (b) Golubev, V. A.; Zhdanov, R. I.; Gida, V. M.; Rozantsev, E. USSR, Chem. Sci. 1971, 20, 768; (c) Golubev, V. A.; Miklyush, R. V. Russ. J. Org. Chem. 1972, 8, 1376. 125

126

(a) de Nooy, A. E. J.; Besemer, A. C.; van Bekkum, H. Synthesis 1996, 1153; (b) Anelli, P. L.; Montanari, F.; Quici, S. Org. Synth. Coll. 1993, VIII, p 367.

127

Ma, Z.; Bobbitt, J. M. J. Org. Chem. 1991, 56, 6110.

128

Bobbitt, J. M. J. Org. Chem. 1998, 63, 9367.

129

Hunter, T. J.; O’Doherty, G. A. Org. Lett. 2002, 4, 4447.

235

6.2 ALCOHOLS TO CARBONYLS (CHdOH ! C]O)

N KO ClO4

AcHN

CH2Cl2

Ph

Silica gel

Ph

CHO

3-Phenylpropiolaldehyde (93%)

3-Phenylprop-2-yn-1-ol

66 HO

OH

+

HO

OH

Me

O

Me

O

Me

O

+

p-TsOH

2 N

O

O

Me

O

O•

O

O

O Colletodiol

TEMPO

Grahamimycin A (76%)

6.2.6 Alternative Metal Compounds As Oxidizing Agents The oxidation of alcohols is not limited to Cr and DMSO derivatives or hypervalent iodine compounds. There are several metal-based reagents that are very effective, particularly with sensitive functionality. This section will examine several of the more important reagents. 6.2.6.1 Tetrapropylammonium Perruthenate: Ley Oxidation Ruthenium derivatives have been used for oxidation reactions. A particularly useful one is tetrapropylammonium perruthenate (TPAP), introduced by Ley and coworkers130 for the oxidation of alcohols and this reaction is now called the Ley oxidation. It is a very mild oxidizing agent that is compatible with many sensitive functional groups. An example is taken from a synthesis of caribenol A by Li and coworkers,131 in which the primary alcohol unit in 67 was oxidized with TPAP to give a 73% yield of ketone 68 (Pr ¼ n-propyl and NMO ¼ N-methylmorpholine N-oxide). HO

Me

OHC Me

(Pr) 4N+ RuO4–, CH2Cl2

TBSO Me 67

Me

MS 4 Å , NMO

TBSO Me

Me

68 (73%)

6.2.6.2 Oppenauer Oxidation A classical and alternative method for the oxidation of alcohols focuses on the reversible reaction between ketones and metal alkoxides, which is especially effective when the metal is aluminum.132 The aluminum alkoxide reaction is reversible, as first demonstrated by Verley133 and Ponndorf134 for the reaction of a ketone with an aluminum alkoxide, which led to formation of a new aluminum alkoxide and a new ketone. In 1937, Oppenauer135 oxidized unsaturated steroidal alcohols using aluminum triisopropoxide [Al(Oi-Pr)3] in acetone. The acetone acts as a hydrogen acceptor, and the presence of excess acetone drives the reaction toward the oxidation product. Oppenauer135 used this method to oxidize the alcohol unit of Δ5-3-hydroxy steroids (e.g., 69) to the Δ4-3 ketone (70). The observed shift of the double bond into conjugation is very common when the alkene moiety is in close proximity to the reactive center. This reaction has 130

(a) Griffith, W. P.; Ley, S. V.; Whitcombe, G. P.; White, A. D. J. Chem. Soc. Chem. Commun. 1987, 1625; (b) Griffith, W. P.; Ley, S. V. Aldrichim. Acta 1990, 23, 13–19; (c) Ley, S. V.; Norman, J.; Griffith, W. P.; Marsden, S. P. Synthesis 1994, 639.

131

Han, J. -C.; Liu, L. -Z.; Chang, Y. -Y.; Yue, G. -Z.; Guo, J.; Zhou, L. -Y.; Li, C. -C.; Yang, Z. J. Org. Chem. 2013, 78, 5492.

132

Djerassi, C. Org. React. 1951, 6, 207.

133

Verley, A. Bull. Chim. Soc. Fr. 1925, 37, 537.

134

Ponndorf, W. Angew. Chem. 1926, 39, 138.

135

Oppenauer, R. V. Rec. Trav. Chim. 1937, 56, 137.

236

6. FUNCTIONAL GROUP EXCHANGE REACTIONS

been applied to many steroids and simple alcohols, and has come to be called the Oppenauer oxidation.32,136 An uncatalyzed version of this reaction was reported to proceed in supercritical fluids.137 Note that this reaction is the reverse of the Meerwein-Ponndorf-Verley reduction138 (see Section 7.11.8). Me Me

O

R

H

C

Me

H

Me

Me

Al(Oi-Pr)3

H

R

Me H H

H

O

HO 69

70

Aluminum compound 71 is a Lewis acid and reacts with the carbonyl oxygen of the ketone to form an “ate” complex 72.132,139 Subsequent transfer of Ha from the alkoxide carbon of 72 to the alkoxy unit from the ketone carbon yields 73 via a six-center transition state. Transfer of Ha between 72 and 73 is a reversible process. Just as 71 and 72 are in equilibrium, so the new “ate” complex 73 is in equilibrium with a new alkoxide (74) and a new ketone. If an excess of the original ketone (RCOR1) is added, the equilibrium is shifted toward 74, and the new ketone (R2COR3). The original ketone (RCOR1) acts as a hydrogen acceptor in this equilibrium process. If the goal is oxidation of a secondary alcohol to a ketone, acetone is the solvent and aluminum triisopropoxide is the added reagent. R R2

R O 1R

+

R3

Oi-Pr O Al Ha

Oi-Pr

O

1R

R2

71 R1

O R3

3R

Oi-Pr

72

R O

Ha R2

Oi-Pr

Al

Ha

Al O

Oi-Pr

R

Oi-Pr

R1 Oi-Pr

O Al Ha

73

Oi-Pr

R2

+

O 3R

74

Oppenauer oxidation of saturated alcohols is often sluggish, but this problem can be overcome by changing the structure of the metal alcoholate, the hydrogen acceptor, or by changing the solvent. Aluminum tri-tert-butoxide, triisopropoxide, or tri-n-propoxide are the most commonly used alkoxides, but aluminum triphenoxide gave better yields with saturated hydroxyl-steroids.140 Acetone (often with benzene as a cosolvent) is the most commonly used hydrogen acceptor, although cyclohexanone in a solution of toluene or xylene is also used. The latter conditions allow higher reaction temperatures and shorter reaction times.140 Ideally, the ketone that serves as the hydrogen acceptor should have a high reduction potential, but a large excess of the ketone can compensate for a low reduction potential, as with acetone.141 Benzophenone is used occasionally because it is resistant to condensation reactions (Section 13.4) that can compete.

136

(a) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-68; (b) de Graauw, C. F.; Peters, J. A.; van Bekkum, H.; Huskens, J. Synthesis 1994, 1007.

137

Sominsky, L.; Rozental, E.; Gottlieb, H.; Gedanken, A.; Hoz, S. J. Org. Chem. 2004, 69, 1492.

138

The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-59.

139

(a) Woodward, R. B.; Wendler, N. L.; Brutschy, F. J. J. Am. Chem. Soc. 1945, 67, 1425; (b) Lutz, R. E.; Gillespie, Jr., J. S. J. Am. Chem. Soc. 1950, 72, 344; (c) Von E. Doering, W.; Young, R. W. J. Am. Chem. Soc. 1950, 72, 631.

140 141

(a) Reference 132, p 226; (b) Reich, H.; Reichstein, T. Arch. Inter. Pharmacodyn. 1941, 65, 415; (Chem. Abstr. 35: 55262, 1941).

(a) Adkins, H.; Cox, F. W. J. Am. Chem. Soc. 1938, 60, 1151; (b) Cox, F. W.; Adkins, H. J. Am. Chem. Soc. 1939, 61, 3364; (c) Baker, R. H.; Adkins, H. J. Am. Chem. Soc. 1940, 62, 3305; (d) Adkins, H.; Elofson, R. M.; Rossow, A. G.; Robinson, C. C. J. Am. Chem. Soc. 1949, 71, 3622.

237

6.2 ALCOHOLS TO CARBONYLS (CHdOH ! C]O)

Nonsteroidal alcohols can be oxidized by this method,142,143 but Oppenauer oxidation of primary alcohols tends to give a low yield of the aldehyde. Oxidation of benzyl alcohol gave 60% of the benzaldehyde, for example, but oxidation of geraniol gave only 35% of geranial.144 Schinz and coworkers145 found that aldehydes are effective hydrogen acceptors when used for the oxidation of primary alcohols to aldehydes. If the boiling point of that aldehyde was 50° C higher than the expected product, the aldehyde product could be distilled as it was formed.143b Cp2ZrH2 Benzophenone

OH

CHO

Geraniol

Geranial (92%)

There are two common side reactions, migration of a nonconjugated double bond into conjugation, and condensation of the aldehyde product with the carbonyl hydrogen acceptor (an acid-catalyzed aldol condensation, Section 13.4). Other Al catalysts have also been developed. Maruoka and coworkers, for example, reported that the modified Al catalyst 75 was highly effective for Oppenauer oxidation of alcohols under mild conditions.146 In recent years, metal catalysts other than Al have diminished the occurrence of these side reactions. An example of an alternative metal catalyst is the cyclopentadienyl Zr reagent (Cp2ZrH2) developed by Ishii et al.147 that was shown to be an effective catalyst in the Oppenauer oxidation using a 1:1 ratio of alcohol-to-hydrogen acceptor. The use of this catalyst led to excellent yields of aldehydes from primary alcohols, but heating in an autoclave without solvent was required despite being catalytic in the Zr reagent. Geraniol was oxidized to geranial in 92% yield by this method, using benzophenone as the hydrogen acceptor. Magnesium-catalyzed Oppenauer oxidations have been observed in Grignard alkylation reactions with aldehydes (Section 11.4.3.1), where direct oxidation of the magnesium alkoxide generated during the reaction prior to hydrolysis is possible.148 Me C8 F17 O2S

Al N

O

75

6.2.6.3 Oxidation With Manganese Dioxide Manganese dioxide (MnO2) is a common form of Mn(IV). Manganese dioxide is the usual end-product of permanganate oxidations in basic solution (Section 7.7), and it is rather insoluble in most organic solvents. Manganese dioxide is capable of oxidizing alcohols to ketones or aldehydes, and the reaction proceeds via a radical intermediate (see below), producing MnO, which is Mn2+ as the byproduct. Manganese dioxide oxidizes primary and secondary alcohols to the aldehyde or ketone, respectively, in neutral media.149 Ball et al.150 discovered this reaction, by precipitating MnO2 and then converting vitamin A to retinal in 80% yield.

142

(a) Reference 132, p 211; (b) Ungnade, H. E.; Ludutsky, A. J. Am. Chem. Soc. 1947, 69, 2629.

143

(a) Reference 132, p 219; (b) Adams, R.; Hamlin, Jr., K. E. J. Am. Chem. Soc. 1942, 64, 2597.

144

(a) Yamashita, M.; Matsumura, R. J. Chem. Soc. Jpn. 1943, 64, 506; (Chem. Abstr. 41: 3753g, 1947); (b) Reference 132, p 222.

145

(a) Schinz, H.; Lauchenauer, A.; Jeger, O.; R€ uegg, R. Helv. Chim. Acta 1948, 31, 2235; (b) R€ uegg, R.; Jeger, O. Helv. Chim. Acta 1948, 31, 1753; (c) Lauchenauer, A.; Schinz, H. Helv. Chim. Acta 1949, 32, 1265.

146

Ooi, T.; Otsuka, H.; Miuraa, T.; Ichikawa, H.; Maruoka, K. Org. Lett. 2002, 4, 2669.

147

Ishii, Y.; Nakano, T.; Inada, A.; Kishigami, Y.; Sakurai, K.; Ogawa, M. J. Org. Chem. 1986, 51, 240.

148

Byrne, B.; Karras, M. Tetrahedron Lett. 1987, 28, 769.

149

(a) Fatiadi, A. J. Synthesis 1976, 65, 133; (b) Evans, R. M. Q. Rev. Chem. Soc. (London) 1959, 13, 61.

150

Ball, S.; Goodwin, T. W.; Morton, R. A. Biochem. J. 1948, 42, 516.

238

6. FUNCTIONAL GROUP EXCHANGE REACTIONS

CHO

MnO2, 6 d

OH

Pet ether , 25°C

Vitamin A

Retinal (80%)

There are several ways to prepare this reagent, and its oxidizing power is strongly influenced by the method of preparation.149 One of the more common methods involves precipitation of MnO2 from a warm aqueous solution of MnSO4 and KMnO4 at a particular pH. The precipitated reagent is then activated by heating to 100–200°C or higher151 for several hours. The solvent used for the oxidation is important, since the reaction proceeds by coordination of both substrate and reagent to the MnO2. The solvent can influence the degree of adsorption and desorption of the alcohol on the manganese dioxide. If a primary or secondary alcohol is used as the solvent, competition for adsorption sites with the alcohol will diminish the yield of oxidation products. Many other solvents can be used, including alkanes (petroleum ether, cyclopentane), chlorinated hydrocarbons (CH2Cl2, CHCl3), ethers (diethyl ether, THF), acetone, acetonitrile, DMSO, DMF, and Py. Reaction times range from a few minutes for allylic alcohols to several hours for saturated alcohols. The oxidation of vitamin A required 6 days, an indication that activating the MnO2 prior to addition of an organic substrate is important. Separately, Goldman152 and Henbest and Stratford153 proposed a radical intermediate in oxidations using MnO2. Goldman’s mechanism is shown for reaction with benzyl alcohol, and it involves adsorption of the alcohol on MnO2, followed by formation of a coordination complex (76). Coordination in this manner allows electron transfer, which generates a radical (77) in a process that is accompanied by reduction of Mn(IV) to Mn(III). A second electron transfer generates the carbonyl product adsorbed on Mn(OH)2, as in 77. The product is desorbed, with loss of water, to complete the oxidation. Hall and Story154 proposed an alternative ionic mechanism for oxidation, which invoked formation of a manganate ester. Ph

MnO2

Ph

O

O

Mn

O H

Ph

O

H

O

O

OH 76 (Coordination)

(Adsorption) Ph

OH

•

Ph

MnIII

O

H O

OH 77

MnIV

OH MnII

PhCHO

+

MnO/H2O

OH (Desorption)

Reprinted with permission from Goldman, I.M. J. Org. Chem. 1969, 34, 3289. Copyright © 1969 American Chemical Society.

Manganese dioxide is probably most useful for the oxidation of allylic and benzylic alcohols. In a synthesis of polygalolide A, Adachi et al.155 oxidized the allylic alcohol unit in 78 to give conjugated ketone 79 in 64% yield using manganese dioxide. The rate of oxidation of alcohols by MnO2 is diminished by the steric hindrance around the carbon bearing the hydroxyl moiety. In general, there is no significant difference in the rate of oxidation of cis- and trans-isomeric alkenes, and there appears to be little cis-trans isomerization during the oxidation. Reaction of cis- or trans-2methylpent-2-en-1-ol with MnO2 led to either the cis- or the trans-aldehyde in 64 or 83% yield, respectively.156 Boehm et al.157 noted, however, that some cis-trans isomerization occurred in similar oxidation of pentadienols and pentenynols.

151

Pratt, E. F.; Van de Castle, J. F. J. Org. Chem. 1961, 26, 2973.

152

Goldman, I. M. J. Org. Chem. 1969, 34, 3289.

153

Henbest, H. B.; Stratford, M. J. W. Chem. Ind. (London) 1961, 1170.

154

Hall, T. K.; Story, P. R. J. Am. Chem. Soc. 1967, 89, 6759.

155

Adachi, M.; Yamada, H.; Isobe, M.; Nishikawa, T. Org. Lett. 2011, 13, 6532.

156

Chan, K. C.; Jewel, R. A.; Nutting, W. H.; Rapoport, H. J. Org. Chem. 1968, 33, 3382.

157

(a) Boehm, E. E.; Thaller, V.; Whiting, M. C. J. Chem. Soc. 1963, 2535; (b) Boehm, E. E.; Whiting, M. C. J. Chem. Soc. 1963, 2541.

239

6.2 ALCOHOLS TO CARBONYLS (CHdOH ! C]O)

O

O

O

O MnO2, CH2Cl2 rt

H

O

O

HO

H O

78

79b (64%)

Manganese dioxide oxidizes allylic and benzylic alcohols faster than primary saturated alcohols, but primary and secondary allylic alcohols react at about the same rate.158 Oxidation of an allylic or benzylic alcohol is generally faster than a saturated alcohol. In a synthesis of seimatopolide A by Prasad and Revu,159 the primary allylic alcohol group in 80 was oxidized to give aldehyde 81 in good yield, for example, and the secondary alcohol moiety was not oxidized. O

O

O

OH C9H19

MnO2, 2 h

O

O

CH2Cl2, Heat

OH

C9H19 H OH

80

81

Although the oxidation of allylic and benzylic alcohols is faster, saturated alcohols do react with MnO2. Their oxidation requires a neutral medium, freshly prepared and activated manganese dioxide, the proper solvent, and long reaction times. Simple examples are the oxidation of propan-2-ol to acetone and the oxidation of 2-methylpropan-1-ol to 2-methyl-propanal, each in 50% yield.160 The conditions are not too vigorous, however, and complex molecules are easily oxidized with MnOs. Even saturated secondary alcohols, which are relatively unreactive,160 can be oxidized if no allylic or benzylic alcohols are present elsewhere in the molecule.161 There is little difference in rate between secondary allylic alcohols and primary saturated alcohols, but steric and/or conjugative effects may have an influence on the reaction. Primary allylic and secondary allylic alcohols are often oxidized with essentially the same rate of reaction, leading to poor chemoselectivity.162 If a large excess of MnO2 is used, it is possible to oxidize more than one alcoholic position without affecting other labile functionality.163 Allylic alcohols can react with MnO2 under certain conditions to give the conjugated acid or ester (in alcoholic solvents). In an example taken from Weinreb and coworker’s164 synthesis of cylindrospermopsin, allylic alcohol 82 was treated with MnO2 in the presence of NaCN and methanol to give a 62% yield of methyl ester 83. The purpose of cyanide in this oxidation is to convert the conjugated aldehyde to a conjugated ester. In a related reaction using HCN, Corey et al.165 showed that reaction of a conjugated aldehyde with HCN/CN- generates a cyanohydrin (e.g., 84). Subsequent oxidation with MnO2 gave cyanoketone, (85), which reacted with the alcoholic solvent (methanol) to yield the corresponding methyl ester.

158

For a synthetic example involving primary allylic versus primary saturated alcohols, see Graham, S. H.; Jonas, D. A. J. Chem. Soc. C 1969, 188.

159

Prasad, K. R.; Revu, O. J. Org. Chem. 2014, 79, 1461.

160

Barakat, M. Z.; Abdel-Wahab, M. F.; El-Sadr, M. M. J. Chem. Soc. 1956, 4685.

161

Wada, K.; Enomoto, Y.; Matsui, K.; Munakata, K. Tetrahedron Lett. 1968, 4673.

162

Kienzle, F.; Mayer, H.; Minder, R. E.; Thommen, W. Helv. Chim. Acta 1978, 61, 2616.

163

(a) Meinwald, J.; Opheim, K.; Eisner, T. Tetrahedron Lett. 1973, 281; (b) Miller, C. H.; Katzenellenbogen, J. A.; Bowlus, S. B. Tetrahedron Lett. 1973, 285. 164

Heintzelman, G. R.; Fang, W. -K.; Keen, S. P.; Walalce, G. A.; Weinreb, S. M. J. Am. Chem. Soc. 2002, 124, 3939.

165

Corey, E. J.; Gilman, N. W.; Ganem, B. E. J. Am. Chem. Soc. 1968, 90, 5616.

240

6. FUNCTIONAL GROUP EXCHANGE REACTIONS

PhH2CO Me

H

OCH2OMe

H

MnO2 , NaCN MeOH, AcOH

N

NH

PhH2CO

OH

H

N

NH O

PhH2CO 82

83 (62%) R

R

R

HCN

CHO

OCH2OMe CO2Me

Me

O

PhH2CO

H

R

MnO2

CN

MeOH

OMe

CN

–CN

O

OH 84

O

85

The functional group transforms available from reactions of MnO2 follow: R R

R

R O

R

O

OH

R

R

CO2R

CHO

OH R

R

R

6.2.7 Oxidations With Silver (Silver Carbonate and Silver Oxide) Silver ion is an excellent oxidizing agent. The reduction potentials for several Ag reagents rank them among the more potent of reagents. The most common reagents are silver carbonate (Ag2CO3), silver (I) oxide (Ag2O), and silver (II) oxide (AgO). 6.2.7.1 Silver Carbonate Silver carbonate (Ag2CO3) is not a powerful oxidizing agent but it is useful in organic chemistry. Rapoport and Riest166 were probably the first to use silver carbonate for the oxidation of alcohols to carbonyl derivatives. Rapoport and Reist166 heated codeine to reflux with silver carbonate in benzene and obtained a 75% yield of codeinone. In later work, King et al.167 oxidized codeine with silver carbonate in refluxing toluene or xylene and obtained an 85% yield of codeinone with a much shorter reaction time. Me

Me N

N Ag2CO3

MeO

O

OH

Codeine

MeO

O

O

Codeinone (75%)

Fetizon and Golfier168 modified the reagent by condensing silver carbonate on Celite. The reaction of primary alcohols with an excess of silver carbonate on Celite gave excellent yields of the corresponding aldehyde, illustrated by the conversion of geraniol to geranial in 97% yield, using 4 equiv of Ag2CO3 on Celite, in refluxing benzene. This modified reagent is now known as Fetizon’s reagent.169 Saturated primary alcohols are converted to aldehydes in excellent yields 166

Rapoport, H.; Reist, H. N. J. Am. Chem. Soc. 1955, 77, 490.

167

King, W.; Penprase, W. G.; Kloetzel, M. C. J. Org. Chem. 1961, 26, 3558.

168

Fetizon, M.; Golfier, M. C.R. Acad. Sci. Ser. C 1968, 267, 900.

169

Mundy, B. P.; Ellerd, M. G.; Favaloro, Jr., F. G. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: New York, NJ, 2005; p 778.

241

6.2 ALCOHOLS TO CARBONYLS (CHdOH ! C]O)

and, as seen with Rapoport’s work, and secondary alcohols are also oxidized to ketones. Lactols are oxidized to the corresponding lactone using this reagent.170 Fetizon et al.171 proposed a mechanism by which adsorption of the alcohol on silver carbonate gave a species such as 86. One-electron transfer via the Ag ion in 87 liberated the protonated carbonyl and carbonic acid (see 88), which decomposed to CO2 and H2Owater. This mechanism involved four distinct processes172: 1. Reversible adsorption of alcohol on the surface of the oxidizing medium with the electron of oxygen forming a coordinated covalent bond with the Ag ions. 2. Oxidation of the CdH bond so the HCOH group is coplanar and perpendicular to the silver carbonate-Celite surface. 3. A concerted irreversible homolytic shift of electrons to generate reduced Ag atoms, hydrogen ions, and a protonated carbonyl. 4. Collapse to products with carbonate ion acting as a hydrogen ion acceptor, generating carbonic acid, which decomposes to CO2 and H2O.

H H

−

O

O

+O

Ag +

+

Ag

H

H O

−

H

Ag

Ag −

−

O

O Ag

O

O

86

87

− +

+O − + O Ag

O

O

+ CO2 + H2O + 2 Ag

O 88

Polar solvents inhibit the reaction, presumably by interfering with the adsorption process as noted in the mechanism proposed for MnO2 oxidations. Oxidation of heptan-1-ol to heptanal with Fetizon’s reagent was quantitative when the solvent was 35% hexanes. When benzene was used as a solvent, the yield of heptanal was 90%, but the yield was 10. The most common solvents are aqueous alcohol and aqueous acetone, and the reaction is usually buffered with phosphate or acetate. The reaction is rapid and may be monitored by disappearance of the purple color of the radical. Teuber and Rau207a (see also Ref. 207b) proposed a simple mechanism for the oxidation of phenol. Electron transfer from Fremy’s salt to phenol generates a radical species, the resonance contributor of cyclohexa-2,5-dien-1-one radical. This radical reacts with excess Fremy’s salt to yield an addition product, 96. Loss of the hydrogen α to the oxygen moiety in 96 leads to formation of the quinone (benzoquinone) and potassium aminodisulfonate [HN(SO3K)2]. O•

OH

Fremy's salt

Fremy's salt

• Phenol

O

O

O

Cyclohexa-2,5dien-1-one radical

Ha

ON(SO3K)2 96

O Benzoquinone

Fremy’s salt sometimes undergoes violent, spontaneous decomposition, but this appears to be due to the presence of catalytic amount of nitrite ion.

The steric demands on the reaction sometimes lead to formation of the less hindered p-quinone as the major product, with low yields of the ortho-quinones. The yield of o-quinone can be low even when the para-position is blocked. When both the ortho- and para-positions are not sterically hindered, the p-quinoid is usually the major product. When X]H in 97, oxidation follows path a to yield the 1,4-quinone. Formation of 98 is consistent with para-attack, and loss of the X group will yield benzoquinone. When X in 97 is alkoxy (X]OR) or alkyl (X]R), however, path b is preferred. Since the para-position is blocked the ortho-intermediate (99) is formed, and loss of the bis(disulfonate) in the usual manner (see above) gives the 1,2-quinone 100. When X]Cl in 100, the chlorine is usually lost with formation of the 1,4-quinone, benzoquinone.203

202

The Chemistry of the Quinoid Compounds, Parts 1 and 2, Patai, S., Ed.; John Wiley: New York, NY, 1974.

203

Thomson, R. H. In The Chemistry of the Quinoid Compounds, Part 1, Patai, S., Ed.; John Wiley: New York, NY, 1974; pp 111–161.

204

Zimmer, H.; Lankin, D. C.; Horgan, S. W. Chem. Rev. 1971, 71, 229.

205

Fremy, E. Ann. Chim. Phys. 1845, 15, 408.

206

Raschig, F. Schwefel und Stickstoff-Studien, Verlag-Chemie: Leipzig-Berlin, 1924 (from Ref. 2 in Ref. 204).

207

(a) Teuber, H. J.; Rau, W. Chem. Ber. 1953, 86, 1036. (b) Reference 203, p 112.

246

6. FUNCTIONAL GROUP EXCHANGE REACTIONS

OH

O

O

O

O

ON(SO3K)2

O

b

a

2 (SO3K)2N-O•

2 (SO3K)2N-O•

H

X

X

X

100

99

97

X

ON(SO3K)2

O Benzoquinone

98

Two important factors influence the oxidation. One is electronic stabilization of the incipient phenoxy radical, and the other is the steric requirement connected with formation of the cyclohexadiene intermediate. The presence of electron-withdrawing groups can cause the oxidation to fail, or at least lead to a poorer yield.203,204,208 Oxidation of phenols with Fremy’s salt can be applied to many phenolic derivatives, as in the synthesis of verapliquinone A by Moody and coworkers209 in which oxidation of 101 gave 1,4-quinone (102) in 65% yield. Oxidation of phenols with Fremy’s salt requires a buffered medium, since the salt is stable in only a limited pH range, as previously mentioned. Many o-quinones are unstable to acidic conditions, and may give other products if the pH > 7.210 However, o-quinones can be isolated. In the synthesis of thelephantin O by Fujiwara et al.,211 oxidation of 103 gave 104 in 80% yield under standard conditions. When the phenolic ring contains a p-chlorine203,212 or a tert-butyl group,213 elimination of these groups from the intermediate yield a quinone.212,213 Elimination of the amino moiety to yield a quinoid can accompany oxidation of aniline derivatives by Fremy’s salt. Oxidation of 3-ethoxy-5-methylaniline, for example, gave 3-ethoxy-5-methyl-1,4-benzoquinone in 93% yield.214 O

OH 2 (SO3K)2NJO•

MeO

MeO

Na2HPO4 aq Acetone

O

101 O

O

O

102 (65%)

O

2 (SO3K)2NJO•, H2O KHPO4, 0°C, 1 h

OH 103

O

O 104 (80%)

Many other oxidizing agents convert phenols to quinones, including the Cr(VI) reagents, but in general most give lower yields of quinone and higher percentages of dimerized products. Chromic acid is an effective oxidant for the conversion of phenols to quinones only when a para-substituent (e.g., halogen, hydroxyl, or amino) is present. The yields are often poor,215 but not always.216 Anodic oxidation of phenols to quinones217 can be very efficient. Bacon and Izzat218 described the oxidation of 2,6-xylenol (2,6-dimethylphenol) with a variety of oxidizing agents. In most

208

Maruyama, K.; Otsuki, T. Bull. Chem. Soc. Jpn. 1971, 44, 2873.

209

Davis, C. J.; Hurst, T. E.; Jacob, A. M.; Moody, C. J. J. Org. Chem. 2005, 70, 4414.

210

See Teuber, H. J.; Thaler, G. Chem. Ber. 1958, 91, 2253.

211

Fujiwara, K.; Sato, T.; Sano, Y.; Norikura, T.; Katoono, R.; Suzuki, T.; Matsue, H. J. Org. Chem. 2012, 77, 5161.

212

Teuber, H. J.; Thaler, G. Chem. Ber. 1959, 92, 667.

213

Magnusson, R. Acta Chem. Scand. 1964, 18, 759.

214

Teuber, H. J.; Hasselbach, M. Chem. Ber. 1959, 92, 674.

215

Reference 203, p 117.

216

Conant, J. B.; Fieser, L. F. J. Am. Chem. Soc. 1923, 45, 2194.

217

Reference 203, p 119.

218

Bacon, R. G. R.; Izzat, A. R. J. Chem. Soc. C 1966, 791.

247

6.3 FORMATION OF PHENOLS AND QUINONES

cases, oxidation gave not only the expected quinone (107) but also dimerized products [e.g., the coupled bis(phenol) (105) and the coupled quinone (106), as well as polymeric material].218 In this work, many oxidizing agents were examined,219 but Fremy’s salt was the superior reagent, yielding quinone 106 in 74% yield with virtually no 105 or 107. The coupled quinone predominated with most oxidizing agents, but polymerization was a problem. O

OH Me

Me

Me

Me

O

OH Me

[O]

Me

+ Me

+ Me

Me

O

Me O

OH 2,6-Dimethylphenol

Me

Me

105

106

107

Note that catechols (1,2-dihyroxybenzenes) are readily oxidized to o-quinones,220 but the products are often sensitive to any electrophilic or nucleophilic species in the reaction medium. Catechol itself is oxidized to cyclohexa-3,5diene-1,2-dione. This oxidation used silver carbonate, and silver salts are widely viewed as a classical oxidation reagent for such transformations.221 Dimerization is as much a problem with catechols as with monophenols.222 Other reagent have been used to oxidize catechol derivatives, including ceric sulfate, lead tetraacetate, DDQ (2,3dichloro-5,6-dicyano-1,4-benzoquinone),223 iodate, and periodate.224 There are many syntheses of molecules containing a quinone unit, and in most cases the quinone is oxidized late in the synthesis from a resorcinol or catechol derivative.225 As mentioned above, quinones can be formed from alkoxy derivatives of resorcinol or catechol, as well as from the parent hydroxyl compounds. In a synthesis of ()-streptonigrin, Donohoe et al.226 treated 108 with ceric ammonium nitrate (CAN) and obtained a 92% yield of quinone (109), where only the benzene ring that contains three methoxy groups is oxidized. OMe

O

MeO

MeO CO2Me OMe

CO2Me

aq CAN, MeCN

H2 N

Me

0°C

O

H2 N

Me

OBn

OBn

OMe OMe 108

OMe OMe 109 (92%)

219

Fatiadi, A. J. Synthesis 1976, 133.

220

References 203 and 204.

221

(a) Balogh, V.; Fetizon, M.; Golfier, M. J. Org. Chem. 1971, 36, 1339; (b) Cason, J. Org. React. 1948, 4, 305.

222

(a) Harley-Mason, J.; Laird, A. H. J. Chem. Soc. 1958, 1718; (b) Horner, L.; D€ urkheimer, W. Chem. Ber. 1958, 91, 2532.

223

Boldt, P. Chem. Ber. 1966, 99, 2322.

224

Reference 203, pp 124–125.

225

Vyvyan, J. R.; Loitz, C.; Looper, R. E.; Mattingly, C. S.; Peterson, E. A.; Staben, S. T. J. Org. Chem. 2004, 69, 2461.

226

Donohoe, T. J.; Jones, C. R.; Kornahrens, A. F.; Barbosa, L. C. A.; Walport, L. J.; Tatton, M. R.; O’Hagan, M.; Rathi, A. H.; Bake, D. B. J. Org. Chem. 2013, 78, 12338.

248

6. FUNCTIONAL GROUP EXCHANGE REACTIONS

Oxidation of phenol leads to several interesting functional group transforms, including the following: R

R

O

R X

O R

R OH

R

R

R R

OH

R

O R

R

R

O

R

X R

R

6.3.2 Phenols As discussed in Section 3.10.7, phenols are conveniently prepared by reaction of diazonium salts with aqueous acid, and also by SNAr reactions of aryl halides (Section 3.10.5). There are also several oxidative methods that convert aromatic derivatives to phenols.227 Direct hydroxylation can be accomplished by using free radical reagents [e.g., a mixture of ferrous ion (Fe2+) and H2O2 (Fenton’s reagent)].228 The yields are usually poor, in the 5–20% range, and formation of biphenyls (a coupling product) predominates. Oxidation of benzene, for example, gave a low yield of phenol, but a relatively large yield of biphenyl. In general, Fenton’s reagent is not very useful for the preparation of phenols. Udenfriend et al.229 introduced a modification of this procedure as a model for the biogenic hydroxylation of tyramine, using an oxygen-ferrous ion-ascorbic acid system in the presence of ethylenediaminetetraacetic acid (EDTA) (Udenfriend’s reagent).228a This reagent gave good yields of ortho- and para-phenolic derivatives from phenyl acetamide. An electrolytic version of this oxidation used anodic oxidation in the presence of Undenfriend’s reagent and oxygen to convert tyramine to a mixture of hydroxytyramines.230 HO

HO

NH2 Tyramine

HO

NH2

3-Hydroxytyramine (Dopamine)

Alkyl benzenes react with H2O2 hydrogen peroxide, Lewis acid catalysts (e.g., boron trifluoride),231 or strong Brønsted-Lowry acids (e.g., trifluoroperoxyacetic acid),228a,232 to yield the corresponding phenol. Chambers et al.232 used trifluoroperoxyacetic acid to oxidize mesitylene (1,3,5-trimethylbenzene) to 2,4,6-trimethylphenol in high yield based on consumed mesitylene (only a 17% conversion based on the peroxyacid). Similarly, McClure and Williams233 reported the oxidation of anisole to phenolic products in low yield (ortho derivative in 27% yield and para derivative in 7% yield), with a conversion of only 44%. Hart and Buchler231 found that the yield of 2,4,6-trimethylphenol from 1,3,5-trimethylbenzene was greatly improved (88% yield with good overall conversion) by using a mixture of boron trifluoride (BF3) and trifluoroperoxyacetic acid (in dichloromethane at
7°C). This modification is probably the best method for peroxyacid oxidation of aromatic hydrocarbons to the corresponding phenol. Aluminum chloride is also an effective catalyst in this transformation, giving primarily a mixture of ortho- and para-hydroxy derivatives.234

227

Reference 11, pp 977–978.

228

(a) Haines, A. H. Methods for the Oxidation of Organic Compounds; Academic Press: London, 1985, p 173; (b) Smith, J. R. L.; Norman, R. O. C. J. Chem. Soc. 1963, 2897.

229

Udenfriend, S.; Clark, C. T.; Axelrod, J.; Brodie, B. B. J. Biol. Chem. 1954, 208, 731.

230

(a) Blanchard, M.; Bouchoule, C.; Djaneye-Boundjou, G.; Canesson, P. Tetrahedron Lett. 1988, 29, 2177; (b) Maissant, J. M.; Bouchoule, C.; Canesson, P.; Blanchard, M. J. Mol. Catal. 1983, 18, 189; (c) Maissant, J. M.; Bouchoule, C.; Blanchard, M. J. Mol. Catal. 1982, 14, 333.

231

Hart, H.; Buehler, C. A. J. Org. Chem. 1964, 29, 2397.

232

Chambers, R. D.; Goggin, P.; Musgrave, W. K. R. J. Chem. Soc. 1959, 1804.

233

McClure, J. D.; Williams, P. H. J. Org. Chem. 1962, 27, 627.

234

Kurz, M. E.; Johnson, G. J. J. Org. Chem. 1971, 36, 3184.

249

6.4 OXIDATION OF ALKENES TO EPOXIDES

Phenol can be hydroxylated with potassium persulfate (K2S2O8) in alkaline media (the Elbs persulfate oxidation).235 In 1893, Elbs used ammonium persulfate to oxidize 2-nitrophenol to nitroquinol.236 Later workers used the K salt,237 which is now the reagent of choice, in order to obtain synthetically useful yields of para-hydroxy phenols. The initial oxidation product is the persulfate (potassium 4-hydroxyphenyl sulfate), which is hydrolyzed to hydroquinone as shown.238 The ortho-product (a pyrocatechol derivative) is formed if the para-position is blocked, but the reaction is slower and the yields are in the 30–50% range.239 The presence of electron-withdrawing groups on the aromatic ring improves the yield, but electron-releasing groups (e.g., methoxy, OMe) can also be tolerated.240 An alternative to direct hydroxylation is the thermal or photochemical decomposition of diacyl peroxides. Aryl derivatives react to yield acyloxylation.241 p-Xylene was converted to 2-(2,5-dimethylphenyl)-1-phenylethan-1-one in the presence of cupric chloride,242 although the yield was only 24%. The benzoate ester in 2-(2,5-dimethylphenyl)-1-phenylethan-1-one was easily hydrolyzed to the phenolic derivative. Lead tetraacetate [Pb(OAc)4, abbreviated LTA, see Section 6.7.3.1] gives a similar reaction, but this reagent yields the o-acetyl derivative from aryl derivatives. Polycyclic aromatics and aryl derivatives243,244 containing electron-releasing groups give the best yields of oxidation products. OH

KOH

H3O+

OSO3 –K+

HO

HO

OH

K2S2O8

Potassium 4-hydroxyphenyl sulfate

Phenol

Me

CuCl2, MeCN, 60°C

Me

Ph

p-Xylene

Me

Ph

O

O O

Hydroquinone

O

Me Ph

O

2-(2,5-Dimethylphenyl)-1- (24%) phenylethan-1-one

The functional group transform for this reaction follow: R

R

HO

6.4 OXIDATION OF ALKENES TO EPOXIDES The previous sections focused attention on the oxidation of a CdOH moiety to a C]O (carbonyl) group. Other functional groups are subject to oxidation, including the π-bond of alkenes. There are several different oxidative functional group transformations that involve alkenes, including incorporation of one oxygen (epoxidation), two oxygen atoms (dihydroxylation), or one oxygen and one nitrogen (aminohydroxylation), and finally, oxidative cleavage

235

(a) Reference 228a, p 174; (b) Sethna, S. M. Chem. Rev. 1951, 49, 91; (c) Behrman, E. J. Beilstein J. Org. Chem. 2006, 2, 22 (an online journal); (d) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-27.

236

Elbs, K. J. Prakt. Chem. 1893, 48, 179.

237

Chemische Fabrik auf Aktien vorm E. Schering: German Patents 81,068, 81,297 and 81,298: see Friedlander, Fortschritte der Teerfarbenfabrikation 1894–1897, 4th Part, pp 126, 127, 121, respectively, taken from Ref. 12 in Ref. 235b.

238

(a) Reference 228a, pp 180–181; (b) Baker, W.; Brown, N. C. J. Chem. Soc. 1948, 2303.

239

Reference 228a, p 180.

240

Baker, W.; Savage, R. I. J. Chem. Soc. 1938, 1602.

241

(a) Reference 228a, pp 177–178; (b) Rawlinson, D. J.; Sosnovsky, G. Synthesis 1972, 1.

242

Reid, C. G.; Kovacic, P. J. Org. Chem. 1969, 34, 3308.

243

Fieser, L. F.; Clapp, R. C.; Daudt, W. H. J. Am. Chem. Soc. 1947, 64, 2052.

244

Ritcher, H. J.; Dressler, R. L. J. Org. Chem. 1962, 27, 4066.

250

6. FUNCTIONAL GROUP EXCHANGE REACTIONS

(usually to carbonyl derivatives). Various methods for the conversion of alkenes to epoxides (oxiranes)245 will be discussed in Sections 6.4.1–6.4.4, and the other functional group transformations will be discussed in succeeding sections.

6.4.1 Epoxides from Halohydrins A simple method for producing epoxides uses the reaction of alkenes and hypohalous acids (generated by the reactions Cl2 + H2O ! HOCl or Br2 + H2O ! HOBr) to yield the trans-(or anti-) halohydrin as the major product. An example is the conversion of cyclohexene to 2-bromocyclohexanol (Section 2.5.3). Subsequent treatment with a base (e.g., sodium hydride) generates the alkoxide, and displacement of the adjacent bromine via an SN2 process (an intramolecular Williamson ether synthesis; Section 3.2.1.1) yields cyclohexene oxide. The SN2 nature of this base-induced cyclization restricts the reaction to halohydrins with a primary or secondary halide, and the halogen must be trans to the hydroxyl group in cyclic systems for epoxide formation. The halohydrin is often generated with aq NBS or NCS (see Section 3.2.1.1). In a synthesis of clividine by Banwell and coworkers,246 treatment of 110 with aq NBS gave bromohydrin 111 (one regioisomer is shown). Subsequent conversion to the dioxolane (in 92% yield from 110) to protect the diol unit (Section 5.3.3.1) and then reaction with NaH led to formation of the epoxide unit in 112, in good yield. Note that this epoxide-forming reaction is not restricted to halide leaving groups, and sulfonate esters (e.g., mesylate, tosylate,247 or triflate) can be used. 1,2-Diols can be converted to a mono-sulfonate ester, and subsequent reaction with KOH leads to an alkoxide that displaces the leaving group to yield the epoxide.248 Br OH

O

HO

OH

NBS, aq THF

OH

0–18°C, 4 h

OH

Br

2. NaH, THF, 0°C, 25 h

O Br

Br

110

O

1. 2,2-DMP, p-TsOH 18°C, 2 h

111 a (92%)

112

The functional group transform is simply R

R

OH

R

R

X

R

O R

6.4.2 Peroxide Induced Epoxidation The reaction of peroxides with alkenes is a common method for the preparation of epoxides (oxiranes), but the nature of the peroxide is very important to the success of the oxidation. Peroxides are a source of electrophilic oxygen when they react with the nucleophilic π-bond of an alkene. Hydrogen peroxide (H2O2) is an important oxidizing agent. While most peroxide reactions involve homolytic cleavage of the OdO bond, generating free radicals (Section 17.3), the reaction of H2O2 and its monosubstituted derivatives with an alkene can proceed via either a concerted or ionic mechanism.249 Three categories of peroxides are used for epoxidation, hydrogen peroxide, alkyl hydroperoxides, and peroxyacids.

245

Reference 11, pp 915–927.

246

White, L. V.; Schwartz, B. D.; Banwell, M. G.; Willis, A. C. J. Org. Chem. 2011, 76, 6250.

247

For an example taken from a synthesis of (
)-platensimycin, see Eey, S. T. -C.; Lear, M. J. Chem. Eur. J. 2014, 20, 11556.

248

For an example of epoxide formation from a bromohydrin, taken from a synthesis of (
)-coriolin, see Mizuno, H.; Domon, K.; Masuya, K.; Tanino, K.; Kuwajima, I. J. Org. Chem. 1999, 64, 2648. 249

Lewis, S. N. In Oxidation, Vol. 1, Augustine, R. L., Ed.; Marcel-Dekker: New York, NY, 1969; p 214.

251

6.4 OXIDATION OF ALKENES TO EPOXIDES

O HO

OH

Hydrogen peroxide

RO

OH

OH R O Peroxyacid

Alkyl hydroperoxide

H RO

OH

O

+

O

O

+

R—OH

R Alkyl hydroperoxide

Oxirane

113

Alcohol

The alkene can be viewed as a Lewis base in an initial reaction with the peroxide ROOH forming a coordination complex 113 that generates an oxygen atom with a positive dipole. Heterolytic cleavage transfers that oxygen to the alkene, and subsequent proton transfer liberates the byproduct (H2O from H2O2, an alcohol from an alkyl hydroperoxide, or a carboxylic acid from a peroxyacid). This sequence involves backside attack at oxygen in an SN2 like process (Section 3.2.1.1).250 The relative rate of oxidation with various peroxide reagents is largely determined by the nature of the OR group in 113, which behaves as a leaving group. The products are an oxirane and the alcohol, ROH. The oxidizing power of the peroxide is inversely related to the pKa of the conjugate acid generated by loss of the leaving group (ROH).251 It is difficult to compare different types of reagents, but a useful ranking of peroxides by their ability to convert alkenes to epoxides follows:250,251

t-BuOJOH
89% yield, as part of Sharma and coworkers124 synthesis of macrosphelide M. Reduction of sulfonate esters with LiAlH4 often gives yields 89%)

7.6.2.3 Epoxides Epoxides are easily generated from alkenes126 (Section 6.4), and also from ketones and aldehydes by reaction with sulfur ylids (see Section 12.5.2).127 Epoxides (oxiranes) are one of the most useful of all oxygenated functional groups 116

See Ref. 82, p. 779.

117

Rassat, A.; Ravet, J. P. Bull. Soc. Chim. Fr. 1968, 3679.

118

Son, J. B.; Kim, S. N.; Kim, N. Y.; Lee, D. H. Org. Lett. 2006, 8, 661.

119

Trevoy, L. W.; Brown, W. G. J. Am. Chem. Soc. 1949, 71, 1675.

120

Ref. 73, pp. 917–922.

121

Feutrill, G. I.; Mirrington, R. N.; Nichols, R. J. Aust. J. Chem. 1973, 26, 345.

122

Rossi, R.; Salvadori, P. A. Synthesis 1979, 209.

123

Fronza, G.; Fuganti, C.; Grasselli, P.; Marinoni, G. Tetrahedron Lett. 1979, 3883.

124

Gangavaram, V. M.; Sharma, G. V. M.; Reddy, P. S. Eur. J. Org. Chem. 2012, 2414.

125

For an example taken from a synthesis of chrysotricine, see Zhang, J.-X.; Wang, G.-X.; Xie, P.; Chen, S.-F.; Liang, X.-T. Tetrahedron Lett. 2000, 41, 2211.

126

Lewis, S. N. In Oxidation, Vol. 1, Augustine, R. L., ED.; Marcel–Dekker: New York, 1969, p. 223.

127

Trost, B. M.; Melvin, L. S. Jr. Sulfur Ylides; Academic Press: New York, 1975.

327

7.6 LITHIUM ALUMINUM HYDRIDE

due to the reactivity of the three-membered ring, and reduction of epoxides leads to alcohols.128 Lithium aluminum hydride usually delivers hydride to the less substituted carbon of an epoxide, with excellent selectivity.129 An example is taken from the Inoue and coworker’s130 synthesis of ()-4-hydroxyzinowol, where reduction of the epoxide unit in 64 gave alcohol 65 in >60% yield. When the epoxide oxygen can coordinate with the Al as an “ate” complex, epoxides can yield products where hydride is delivered to the more substituted carbon.131 Intramolecular delivery of hydride is to the most electropositive carbon, although intermolecular delivery is also possible.132 MeO HO

OMe

MeO HO

OMe

LiAlH4 , ether

O

O

O

HO

O

OMOM

O

64

OMOM

65 (>60%)

7.6.2.4 Alkyne-Alcohols Formation of an alcohol in the presence of another functional group via reduction can sometimes be accompanied by elimination to yield an alkene. This elimination process can be turned to an advantage when propargylic alcohols are reduced.133 The so-called Whiting reaction134 reduces an alkynyl diol (e.g., 2,7-dimethylocta-1,7-dien-4-yne-3,6-diol) with LiAlH4 to yield a diene moiety, in 2,7-dimethylocta-1,3,5,7-tetraene (cosmene), via addition of four hydrogen atoms followed by formal elimination of 2 equiv of water. In general, LiAlH4 does not reduce alkene or alkynyl moieties, but as noted for the Whiting reaction, propargylic alcohol derivatives (C^CCH2OH) are important exceptions.135,136 An alternative method for the reduction of alkynyl alcohols to trans-allylic alcohols employs hydrosilation, via treatment of a siloxy derivative [CdOSi(OMe)3] with TBAF and CuI.137 The LiAlH4 reduction of alkynyl alcohols, readily formed via condensation of alkyne anions with aldehydes and ketones (Section 11.3.3), generally gives the trans-allylic alcohol as seen in a synthesis of (+)-garvensintriol by Yadav et al.,138 in which propargyl alcohol 66 was reduced to 67 in 95% yield. HO

OH

1. LiAlH4 2. Hydrolysis

2,7-Dimethylocta-1,7-dien-4-yne-3,6-diol

2,7-Dimethylocta-1,3,5,7-tetraene

OBn LiAlH4 , THF , 3 h

Ph

Reflux

HO

OBn

Ph OH

66

67 (95%)

128

For reagents used to reduce epoxides and other cyclic ethers, see Ref. 82, p. 1019.

129

Ref. 73, p. 653.

130

Todoroki, H.; Iwatsu, M.; Urabe, D.; Inoue, M. J. Org. Chem. 2014 79, 8835.

131

Azuma, H.; Tamagaki, S.; Ogino, K. J. Org. Chem. 2000, 65, 3538.

132

Ref. 81, p. 104.

133

Karlsen, S.; Frøyen, P.; Skattebøl, L. Acta Chem. Scand. 1976, 30B, 664. (b) Brown, H. C.; McFarlin, R. F. J. Am. Chem. Soc. 1956, 78, 252.

134

(a) Nayler, P.; Whiting, M. C. J. Chem. Soc. 1954, 4006. (b) Isler, O.; Montavon, M.; R€ uegg, R.; Zeller, P. Helv. Chim. Acta 1956, 39, 454. (c) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006, p. ONR-100.

135

Ref. 73, pp. 968–975.

136

Trost, B. M.; Lee, D. C. J. Org. Chem. 1989, 54, 2271.

137

For an example used in a synthesis of bafilmycin A1, see Kleinbeck, F.; Fettes, G. J.; Fader, L. D.; Carreira, E. M. Chem. Eur. J. 2012, 18, 3598.

138

Yadav, J. S.; Reddy, U. V. S.; Anusha, B.; Subba Reddy, B. V. Tetrahedron Lett. 2010, 51, 5529.

328

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

This specialized reduction is not limited to C ^ CCHOH systems. Lithium aluminum hydride in diglyme that is heated to reflux also reduces homoallylic alkynyl alcohols (C ^ CCH2CH2OH). An example is the reduction of dec-3-yn-1-ol to (E)-dec-3-en-1-ol in 93% yield, as reported in Vyvyan et al.139 synthesis of gibbilimbol A. OH C6H13

1. LiAlH4 , Diglyme Reflux

C6H13

OH

2. Hydrolysis

Dec-3-yn-1-ol

(E)-Dec-3-en-1-ol (93%)

7.6.2.5 Nitriles Nitriles are synthetic precursors for both functional group transformation and carbon-carbon bond-forming reactions (Sections 11.2 and 13.4.2.4). They are usually formed by a SN2 substitution reaction in which the cyanide ion displaces a leaving group from an alkyl halide or sulfonate ester (Sections 3.2.1 and 11.2.1). Alternatively, nitriles can be formed by dehydration of amides. Once the nitrile has been formed, reduction of the cyano group with LiAlH4 will yield a primary amine,140 so nitriles function as protected (latent or surrogate) aminomethyl groups. This transformation is seen in a synthesis of ()-tubifoline by Mori et al.,141 in which the nitrile unit in 68 was reduced to an aminomethyl unit in 69 using LiAlH4. The amine product is sometimes trapped (protected, Section 5.3.4.2) by reaction with an acyl halide to yield an amide or a carbamate, since this allows further synthetic manipulation. CN

NH2 1. LiAlH4 2. Hydrolysis

N H 68

N H 69

Nitrile reduction by hydride is thought to proceed via initial formation of an iminium salt and subsequent conversion to a bis(iminoaluminate).142 Reduction with additional hydride and hydrolysis leads to the amine product. In some cases, removal of the α-proton occurs to yield the α-lithio nitrile (Section 13.4.2.4),142 especially when a hydride slurry is added to the nitrile solution (inverse addition). In such cases, a dimeric product is formed that is indicative of an enolate condensation. Addition of butanenitrile to LiAlH4 yielded butanamine in 79%, for example, but an inverse addition in ether gave only 34% of that amine. Inverse addition also led to 26% of 2-ethyl-1,3-hexanediamine, presumably via a Thorpe condensation143 (discussed in Section 13.4.2.4) of an intermediate nitrile carbanion.142

7.6.2.6 Azides Azides are important amine surrogates.144 The azide anion (N3) is an excellent nucleophile, and SN2 displacement of primary and secondary halides (or alkyl sulfonates) with sodium azide (Section 3.2.1.1) gives the corresponding alkyl azide. Reduction of the azido group with LiAlH4 yields a primary amino group (RdN3 ! RdNH2). An example is taken from Brown and coworker’s145 synthesis of (+)-allomatrine, in which reduction of the azide group in 70 gave amine 71 in good yield.

139

Vyvyan, J. R.; Holst, C. L.; Johnson, A. J.; Schwenk, C. M. J. Org. Chem. 2002, 67, 2263.

140

For reagents used to reduce nitriles, see Ref. 82, p. 1271.

141

Mori, M.; Nakanishi, M.; Kajishima, D.; Sato, Y. J. Am. Chem. Soc, 2003, 125, 9801.

142

Soffer, L. M.; Katz, M. J. Am. Chem. Soc. 1956, 78, 1705.

143

(a) Baron, H.; Remfry, F. G. P.; Thorpe, Y. F. J. Chem. Soc. 1904, 85, 1726. (b) Schaefer, J. P.; Bloomfield, J. J. Org. React. 1967, 15, 1.

144

See Ref. 82, p. 815.

145

Watkin, S. V.; Nicholas, P.; Camp, N. P.; Brown, R. C. D. Org. Lett. 2013, 15, 4596.

329

7.6 LITHIUM ALUMINUM HYDRIDE

N3 H

NH2

SiMe3

H

H

LiAlH4 , ether

SiMe3

H

0°C – rt , 1 h

N

N 70

71

7.6.2.7 Nitro Compounds Although they are less common than azides and nitriles, aliphatic nitro compounds can be reduced to amines,146 so they also function as amine surrogates. Reduction of nitriles or azides is usually the preferred method to introduce an amino group since reactions that introduce the nitro group often require strong acid. Aromatic nitration (Section 3.10.2) is relatively straightforward, however, and reduction to aniline derivatives would be a very attractive synthetic route to this important class of amines. However, the LiAlH4 reduction of nitrobenzene does not give aniline, but rather an azo compound. An example is the LiAlH4 reduction of nitrobenzene that yields 1,2-diphenyldiazene, in 84%.147 Catalytic hydrogenation (see Section 7.10.5) is the best method for the reduction of aromatic nitro compounds to aniline derivatives (e.g., nitrobenzene to aniline by catalytic hydrogenation). The difficulties observed with aromatic nitro compounds are not found with aliphatic nitro compounds, which are relatively easy to reduce with LiAlH4. The nitro group in 72 was reduced to give amine 73 in >70% yield as part of Zakarian and coworker’s148 synthesis of (+)-brevisamide. Note that in this case 73 was immediately treated with acetic anhydride to generate the N-acetyl derivative in 67% yield for both steps.

Ph

NO2

1. LiAlH4 , ether Reflux , 30 min

Nitrobenzene

NO2

Ph

N

N

Ph

2. H3O+

1,2-Diphenyldiazene (84%)

LiAlH4 , THF

NH2

–15 ·C - Reflux

O

O OH

72

OH 73 (>70%)

7.6.2.8 Sulfonamides Primary sulfonamides resist reductive cleavage with LiAlH4; secondary sulfonamides are cleaved, but the reaction requires a temperature of 120°C in dibutyl ether.149 Exceptions to this sluggish reactivity occur when the sulfonamide is conjugated to a carbonyl moiety. Oppolzer et al.150 developed a chiral sulfonamide auxiliary (Oppolzer’s sultam, 74) derived from camphor, and attachment of this auxiliary allows high asymmetric induction in a variety of reactions. In a synthesis of fumagillin by Perlmutter and coworkers,151 74 was acylated and converted to 75 in five steps. Reductive cleavage of the sulfonamide with LiAlH4 removed the auxiliary in low yield to give alcohol 76 along with the auxiliary 74, which could be recycled. Note that complete removal of this group is sometimes a problem, which can severely limit the utility of this auxiliary.

146

See Ref. 82, p. 821.

147

Hart, M. E.; Suchland, K. L.; Miyakawa, M.; Bunzow, J. R.; Grandy, D. K.; Scanlan, T. S. J. Med. Chem. 2006, 49, 1101.

148

Herrmann, A. T.; Martinez, S. R.; Zakarian, A. Org. Lett. 2011, 13, 3636.

149

(a) Searles, S.; Nukina, S. Chem. Rev. 1959, 59, 1077. (b) Klamann, D. Monatsh Chem. 1953, 84, 651.

150

Oppolzer, W.; Chapuis, C.; Bernardinelli, G. Helv. Chim. Acta 1984, 67, 1397.

151

Ciampini, M.; Perlmutter, P.; Watson, K. Tetrahedron Asymmetry 2007, 18, 243.

330

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

OBPS O

O

S O

N

Cl

1.

N

HO

OBPS

H

S 2. 5 Steps

O

O

LiAlH4

+

O O

74

O

O

O 76

75

74

7.6.2.9 Ozonides Ozonolysis is a powerful method for cleaving alkenes to carbonyl products (Section 6.7.2). Ozonolysis of alkenes initially generates an ozonide that can be reduced by a variety of reagents. Dimethyl sulfide or Zn in acetic acid are the most common reagents for the reduction of an ozonide to an aldehyde or ketone. Reduction of the ozonide with the more powerful LiAlH4, however, gives the alcohol directly. The functional group transforms in this section are summarized as follows: R

R

R

O

OH R

R

R

NH2

R

R

R

R1 R

X

C N

R

R1

X H

R1

R

R

X

R1 OSO2

H

R NH2

N3

OH

R1

R

R2

OH OH

R

7.7 HYDRIDE REDUCING AGENTS WITH ELECTRON-RELEASING GROUPS The important reagents borane, alane, sodium borohydride, and lithium aluminum hydride have been described in previous sections. If a hydrogen atom on each of these reagents is replaced with an electron-releasing group, the resulting reagent is generally a more powerful reducing agent. If a hydrogen atom is replaced with an electron-withdrawing substituent, the resulting reagent is generally a weaker reducing agent. Weaker reducing agents are generally more selective in reactions with various functional groups. This section will describe several of the more useful of hydride reagents that have an electron-releasing group attached to B or Al.

7.7.1 Borane and Alane Derivatives As described in Fig. 7.1, introducing electron-releasing groups (e.g., alkyl groups) or electron-withdrawing groups (e.g., halogen, cyano, or alkoxy groups) can modulate the reducing power of borane or alane. This section will discuss derivatives of borane and alane with electron-releasing groups.

7.7.1.1 Alkylboranes Simple alkylboranes [e.g., methylborane (MeBH2) and dimethyborane (Me2BH)] are known compounds,152 and diethylborane (Et2BH) is also known.153 There are few reports in the literature that use these compounds for reduction. Dicyclohexylborane is one example that has been used to reduce ketones to alcohols.154 Tertiary alkylboranes reduce aldehydes and ketones, however, but a discussion of such compounds will be delayed until Section 7.9 in connection with asymmetric reduction reactions.

152

Brown, H. C.; Cole, T. E.; Srebnik, M.; Kim, K. W. J. Org. Chem. 1986, 51, 4925.

153

See Midland, M. M.; Zolopa, A. R.; Halterman, R. L. J. Am. Chem. Soc. 1979, 101, 248.

154

(a) Brown, H. C.; Varma, V. J. Am. Chem. Soc. 1966, 88, 2871. (b) Brown, H. C.; Varma, V. J. Org. Chem. 1974, 39, 1631.

331

7.7 HYDRIDE REDUCING AGENTS WITH ELECTRON-RELEASING GROUPS

Me

Me B H

Me 2

77

More complicated secondary boranes have been explored as reducing agents. Di(1,2-dimethylpropyl)borane, or diisoamylborane [commonly known as disiamylborane ¼ (Sia)2BH], 77, is a useful secondary alkylborane that will be discussed in Section 9.2.1. The presence of the electron-releasing alkyl groups on B should make 77 a powerful oxidizing agent, but in some ways it is weaker than borane. The observed diminished reactivity is probably due to the increased steric hindrance of the alkyl groups on the B in 77, but is absent in borane, which inhibits formation of key intermediates.155 In the work of Brown et al.,155 it was shown that aldehydes and ketones are rapidly reduced, but the rate of reduction of benzophenone is only moderate and that of quinones is slow. It was also shown that aldehydes and ketones are easily reduced to an alcohol in the presence of halogen substituents, acid chlorides, or carboxylic acids. Borane readily reduces carboxylic acids to alcohols and lactones to diols, but reduction of carboxylic acids with 77 is less favorable due to formation of disiamylborinic acid. Lactones are reduced to the lactol. Borane reduces tertiary amides to amines, but other amides react with only 1 equiv of 77 and gives the aldehyde after hydrolysis.155 Interestingly, nitro compounds are slowly reduced to an amine with 77, whereas borane does not reduce nitro compounds. Epoxides are reduced with difficulty using either borane or 77.155 7.7.1.2 Diisobutylaluminum Hydride (Diisobutylalane) It is often difficult to generate pure AlH3, but aluminum hydride that is typically impure is routinely used although it contains other Al compounds. Such mixtures may cause problems in reactions with sensitive substrates, however. Alkylalanes have been developed that are analogous to the alkylboranes, and some have proven to be quite useful as reducing agents. This section will focus on what is arguably the most used dialkylalane, a commercially available and commonly used alkyl derivative is an alkylated alane known as diisobutylaluminum hydride ([(CH3)2CHCH2]2AlH, or dibal). Diisobutylaluminum hydride is prepared by heating triisobutylaluminum to reflux in the solvent heptane, although other solvents can also be used.156 It is a pyrophoric liquid when used neat, so it is commonly used as a heptane or ether solution. Diisobutylaluminum hydride is a strong reducing reagent, reducing most functional groups. Arguably, the synthetic utility lies in three areas: reduction of esters, reduction of lactones, and reduction of nitriles. Esters are commonly reduced to the alcohol analogous to LiAlH4.157 Note that lactams are not reduced under these conditions. An exception is the amide known as a Weinreb amide (see Section 4.2.3), which is cleanly reduced to an aldehyde, as in the reduction of 78–79 in 93% yield taken from Kato and coworker’s158 synthesis of (+)-gregatin B. O TBDMSO

Me TBDMSO

O N

OMe

dibal , THF –78 °C , 30 min

TBDMSO

Me

78

Me TBDMSO

H

79 (93%)

The reduction of esters to the corresponding alcohol with dibal is quite common, as in the reduction of 80–81 in 90% yield, in Lee and coworker’s159 synthesis of ()-yezo’otogirin C. In some cases, it is possible to limit the reduction of an ester to yield an aldehyde.160 Generally, reduction of an ester at low temperature leads to the aldehyde, whereas 155

Brown, H. C.; Bigley, D. B.; Arora, S. K.; Yoon, N. M. J. Am. Chem. Soc. 1970, 92, 7161.

156

(a) Ziegler, K.; Geller, H. G.; Lehmkuhl, H.; Pfohl, W.; Zosel, K. Annalen, 1960, 629, 1. (b) Ref. 15, p. 260. (c) Eisch, J. J.; Kaska, W. C. J. Am. Chem. Soc. 1966, 88, 2213. (d) Gensler, W. J.; Bruno, J. J. J. Org. Chem. 1963, 28, 1254. 157

Trost, B. M.; Cramer, N.; Bernsmann, H. J. Am. Chem. Soc. 2007, 129, 3086.

158

Kusakabe, T.; Kawai, Y.; Kato, K. Org. Lett. 2013, 15, 5102.

159

He, S.; Yang, W.; Zhu, L.; Du, G.; Lee, C.-S. Org. Lett. 2014, 16, 496.

160

(a) Boger, J.; Payne, L. S.; Perlow, D. S.; Lohr, N. S.; Poe, M.; Blaine, E. H.; Ulm, E. H.; Schorn, T. W.; LaMont, B. I.; Lin, T.-Y.; Kawai, M.; Rich, D. H.; Veber, D. H. J. Med. Chem. 1985, 28, 1779. (b) Luly, J. R.; Hsiao, C.-N.; BaMaung, N.; Plattner, J. J. J. Org. Chem. 1988, 53, 6109.

332

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

reduction at higher temperatures leads to the alcohol. An example is the reduction of the ester unit in 82 in toluene at 78°C to give aldehyde 83 in 87% yield as part of the Smith and coworker’s synthesis of ()-irciniastatin B.161 Both the solvent and the temperature play a role in this reduction. In many cases, however, significant amounts of alcohol are formed even at low temperatures. O

O dibal , –78°C

H

H

CO2Me

CH2OH

80

81 (90%)

DMBO

CO2Me

DMBO

dibal , CH2Cl2

CHO

–78°C

CO2Me

CO2Me DMB = 3,4-Dimethoxybenzyl ether

ODMB

ODMB

82

83 (87%)

Partial reduction of lactones to lactols rather than the diol is observed with dibal, but they are not always stable. In a synthesis of a late stage intermediate toward the herbicidins by Trauner and coworkers,162 lactone 84 was reduced to lactol 85 in 98% yield with dibal. It is also common for the lactone to be reduced to a diol, as in the conversion of 86–87 in 79% yield, taken from the Lovely and coworker’s synthesis of 70 -desmethylkealiiquinone.163 H

H O

H

O

H

dibal

O O

O

H H

O HO

O

O

84

H H

85 (98%) O

Me

OH O

N

O

dibal

Me

–78°C to rt

N

OH

N N

OMe 86

OMe 87 (79%)

There are useful specific modifications of dibal that are useful in synthesis. When mixed with alkylcopper reagents (MeCu or t-BuCu are the most common), dibal yields selective 1,4-reduction of conjugated carbonyl compounds. An example, taken from Corey et al.164 synthesis of desogestrel, is the conversion of 88–89 in 91% yield using HMPA as a solvent. The dibal also reduces nitriles to aldehydes. A synthetic example, taken from Schobert and Ostermeier’s165 synthesis of (+)-chloriolide, treated nitrile 90 with dibal to give an 86% yield of 91.

161

An, C.; Jurica, J. A.; Walsh, S. P.; Hoye, A. T.; Smith, III, A. B. J. Org. Chem. 2013, 78, 4278.

162

Hager, D.; Paulitz, C.; Tiebes, J.; Mayer, P.; Trauner, D. J. Org. Chem. 2013, 78, 10784.

163

Lima, H. M.; Sivappa, R.; Yousufuddin, M.; Lovely, C. J. Org. Lett. 2012, 14, 2274.

164

Corey, E. J.; Huang, A. X. J. Am. Chem. Soc. 1999, 121, 710.

165

Ostermeier, M.; Schobert, R. J. Org. Chem. 2014, 79, 4038.

7.7 HYDRIDE REDUCING AGENTS WITH ELECTRON-RELEASING GROUPS

333

OSiMe2t-Bu

OSiMe2t-Bu HMPA dibal , t-BuCu

O

O 88 CN O

O

H 89 (91%) CHO

dibal , CH2Cl2

O

O

–65 to –40°C , 1.5 h

90

O

O

91 (86%)

7.7.2 Borohydride Derivatives It is clear that borane and alane, along with their alkyl derivatives, offer significant advantages for the reduction of many functional groups. Alkylboranes and alkylalanes are important reducing agents. Sodium borohydride is a selective reducing agent, which means that it does not react with a wide range of functional groups. Incorporation of an electron-releasing alkyl group on B leads to alkylborohydrides, which have been shown to be more powerful reducing agents, and quite useful. In general, alkylborohydrides are more reactive than NaBH4. Two reagents in particular are important in organic synthesis, lithium triethylborohydride and lithium or potassium tri-sec-butylborohydride. The former reagent is commonly known as Super-Hydride and the latter is known as Selectride. Both terms suggest great utility. 7.7.2.1 Super-Hydride Alkyl groups are known to be electron releasing relative to boron. Therefore, when alkyl groups are attached to the B of a borohydride reagent the reducing power of that reagent is enhanced.10a,166 An extremely useful reducing agent using this concept is lithium triethylborohydride (LiBHEt3), which has the trade name Super-Hydride. In part, this term refers to the fact that it exhibits powerful reducing properties. It is prepared by reaction of lithium hydride with triethylborane in THF.167 Super-Hydride is a more powerful hydride nucleophile than LiBH4,168 giving reactions that mimic SN2 behavior. BEt3

Triethylborane

THF, 65∘ C, 15min + LiH ƒƒƒƒƒƒƒƒƒƒƒ!

LiBHEt3 Lithium triethylborohydride

Super-Hydride is commonly used for the reductive dehalogenation of alkyl halides. The reaction of 92 with LiEt3BH, taken from a synthesis of rapamycin by Ley et al.,169 selectively reduced the primary bromide to yield the corresponding methyl group in 93. Note that the vinyl bromide moiety was not reduced, nor was the dithiane unit. Deuterated Super-Hydride (LiEt3B–D) exhibits a nucleophilic nature, as shown in the conversion of 2-exo-bromobicyclo[2.2.1] heptane to the endo-d-derivative by deuteride displacement, which proceeded with inversion analogous to an SN2 process.168,170 Sulfonate esters (e.g., mesylates, tosylates, and triflates) are reduced with LiBHEt3 via CdO cleavage. In a synthesis of (+)-mupirocin H, Chakraborty and Udawant,171 methanesulfonate ester 94 was treated with LiBHEt3 to give the methyl group indicated in 95, in >86% overall yield.

166

Krishnamurthy, S. Aldrichimica Acta 1974, 7, 55.

167

Brown, H. C.; Krishnamurthy, S.; Hubbard, J. L. J. Am. Chem. Soc. 1978, 100, 3343.

168

Brown, H. C.; Krishnamurthy, S. J. Am. Chem. Soc. 1973, 95, 1669.

169

Ley, S. V.; Tackett, M. N.; Maddess, M. L.; Anderson, J. C.; Brennan, P. E.; Cappi, M. W.; Heer, J. P.; Helgen, C.; Kori, M.; Kouklovsky, C.; Marsden, S. P. Norman, J.; Osborn, D. P.; Palomero, M. A.; Pavey, J. B. J.; Pinel, C.; Robinson, L. A.; Schnaubelt, J.; Scott, J. S.; Spilling, C. D.; Watanabe, H.; Wesson, K. E.; Willis, M. C. Chemistry—A European Journal 2009, 15, 2874. 170

Coppi, L.; Ricci, A.; Taddei, M. J. Org. Chem. 1988, 53, 911.

171

Udawant, S. P.; Chakraborty, T. K. J. Org. Chem. 2011, 76, 6331.

334

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

Br S

LiEt3BH , THF

S

Br

0°C

Br S

S 92

93 (99%)

LiEt3BH , THF , 0°C , 5 h

OTBS

H3CO2SO

OTBS

OTIPS

OTIPS

94

95 (>86%)

Since lithium triethylborohydride is highly nucleophilic, epoxides can be opened regioselectively at the less sterically hindered carbon atom. An example is the reduction of 1-methyl-1,2-epoxycyclohexane to 1-methylcyclohexan-1-ol in 99% yield.172 In reactions with substrates that are sensitive to rearrangement, LiBHEt3 is superior to LiAlH4. Reduction of 96 with LiAlH4 gave 15% of 97 and 85% of the rearranged alcohol 98.172 Reduction with lithium triethylborohydride, however, gave 93% of 97 with 74%)

O

O

O

O

O

LiBHEt3 , THF –78 °C

– Et3O + BF4

N

OMe

OTBS

S

N

SEt

N

SEt

N

Me3Si

Me3Si 101

Me3Si 102

172

Krishnamurthy, S.; Schubert, R. M.; Brown, H. C. J. Am. Chem. Soc. 1973, 95, 8486.

173

Scully, S. S.; Porco, J. A., Jr. Angew. Chem. Int. Ed. 2011, 50, 9722.

174

Toyooka, N.; Yoshida, Y.; Yotsui, Y.; Momose, T. J. Org. Chem. 1999, 64, 4914.

175

Perron, F.; Albizati, K. F. J. Org. Chem. 1989, 54, 2044.

176

Heitz, M.-P.; Overman, L. E. J. Org. Chem. 1989, 54, 2591.

103 (84%)

335

7.7 HYDRIDE REDUCING AGENTS WITH ELECTRON-RELEASING GROUPS

7.7.2.2 The Selectrides It was established in Section 7.7.2.1 that adding alkyl groups to a borohydride unit increased the reducing power. However, as noted with substituted boranes in Section 7.3.1, this strength was modulated if the alkyl groups are sterically bulky since it is more difficult for the B–H unit to approach a substrate. The presence of large groups that provide significant steric hindrance can also lead to greater diastereoselectivity in the reduction of prochiral carbonyls (see Section 7.9.2). Two commonly available reagents that exploit this concept are prepared by the reaction of trisec-butylborane (104, see Section 9.2.1 for a discussion of this borane) with potassium hydride to yield potassium tri-sec-butyl-borohydride (105, known as K-Selectride) or lithium trimethoxyaluminum hydride to yield lithium trisec-butylborohydride (106, known as L-Selectride).10a,167,177,178, BH- K+

B

KH , THF

3

3

25 °C

105

BH- Li+

LiAlH(OMe)3 3

THF , 25 °C

104

106

Both the K and Li derivatives are effective for reductions although, by analogy to metal borohydrides (Section 7.5), the Li derivative may be somewhat stronger. Both reagents reduce aldehydes and ketones to the corresponding alcohol, even at temperatures as low as 78°C. In one example, L-Selectride reduced the ketone moiety of 107 selectively to alcohol 108 in good yield as part of Hatakeyama and coworker’s179 synthesis of oxazolomycin A. In a synthesis of 12,13-desepoxy ecklonialactone, Hiersemann and coworkers180 reduced the ketone moiety in 109 with K-Selectride to give alcohol 110 in 89% yield. Selectride is the typical choice rather than another hydride reducing agent when improved diastereoselectivity is an issue. NHFmoc

OMe

CO2CH2OTIPS Me N

NHFmoc

OMe

Li(sec-Bu)3BH ,THF

O O

O

O

Si i-Pr

–78 °C

O O

OH

Fmoc = Fluorenylmethyloxycarbonyl chloride

i-Pr

107

CO2CH2OTIPS Me N O

Si i-Pr

i-Pr

108 (Good yield)

CO2Me O

K(sec-Bu)3BH ,THF

CO2Me

–78 °C

BnO

OH BnO

109

110 (89%)

Alkyl borohydrides were prepared from trisiamylborane (siamyl ¼ Me2CHCHMe— ¼ sec-isoamyl, Section 9.2.1) with even greater steric hindrance, but with similar reactivity and selectivity to the Selectrides. Lithium trisiamylborohydride [LiBH(CHMeCHMe2)] and the potassium salt [KBH(CHMeCHMe2)], are similar to the L- and the K-Selectrides and are called LS-Selectride and KS-Selectride, respectively.181 In Ma and Zhu’s182 synthesis of (3S,4aS,6R,8S)-hyperaspine, LS-Selectride reduction of 111 gave 112 in 83% yield as a single isomer.

177

(a) Brown, C. A.; Krishnamurthy, S. J. Organomet. Chem. 1978, 156, 111. (b) Binger, P.; Benedikt, G.; Rotermund, G. W.; K€ oster, R. J. L. Ann. Chem. 1968, 717, 21. (c) Corey, E. J.; Albonico, S. M.; Koelliker, U.; Schaaf, T. K.; Varma, R. K. J. Am. Chem. Soc. 1971, 93, 1491.

178

Brown, C. A. J. Am. Chem. Soc. 1973, 95, 4100.

179

Eto, K.; Yoshino, M.; Takahashi, K.; Ishihara, J.; Hatakeyama, S. Org. Lett. 2011, 13, 5398.

180

Becker, J.; Butt, L.; von Kiedrowski, V.; Mischler, E.; Quentin, F.; Hiersemann, M. Org. Lett. 2013, 15, 5982.

181

Krishnamurthy, S.; Brown, H. C. J. Am. Chem. Soc. 1976, 98, 3383.

182

Zhu, W.; Ma, D. Org. Lett. 2003, 5, 5063.

336

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

O

H

Me

H

HO

Me

LS-Selectride , THF , –78 °C

N

O

N

C5H11

O

C5H11

111

112 (83%)

7.7.3 Alkylaluminum Hydrides Lithium aluminum hydride is a remarkably powerful reducing agent. Indeed, it will reduce almost any molecule that bears a heteroatom. Therefore, there is little need for a more powerful reducing agent, which would presumably be generated if alkylated reagents (e.g., LiAlHR3) were prepared. Nonetheless, some reagents of this type are known. Lithium triethylaluminum hydride (LiAlEt3H), for example, is known and has been used to reduce Mo complexes of B-chloroboropins (e.g., 113) to the crystalline boropin complex, 114.183 Lithium triethylaluminum hydride has also been used in reactions with alkenes to form alkylalanes.184

B

Cl

LiAlEt3H

B

Mo(CO)3 113

H

Mo(CO)3 114

These reagents do not appear to be used very often for the reduction of organic compounds, or at least their preparation and use has not been reported. For that reason, compounds of this type will not be discussed further.

7.8 HYDRIDE REDUCING AGENTS WITH ELECTRON-WITHDRAWING GROUPS Just as electron-releasing groups tend to increase the reducing power of B and Al compounds, electron-withdrawing groups will diminish the reducing power. The most common electron-withdrawing groups are alkoxy or acyloxy units, but cyano is also known. The following sections will discuss borohydride and aluminum hydride derivatives that possess these electron-withdrawing groups. In principle, borohydride or aluminum hydride reagents of this type will be less reactive and weaker reducing agents. This assumption is not always correct for reductions, however, because the reducing power often depends on the solvent and the coordinating ability of the reagent.

7.8.1 Alkoxyborohydrides Sodium borohydride is a relatively selective (weaker) reducing agent, so reducing agents that are even weaker will likely be used for rather specific applications. Alkoxy groups are known to be electron withdrawing when attached to B, so alkoxyborohydrides are a reasonable choice for such a reagent. Brown et al.185 prepared sodium trimethoxyborohydride by the reaction of sodium hydride (NaH) and trimethoxyborate, B(OMe)3, and in later work showed that sodium trimethoxyborohydride [NaB(OMe)3H] is a good reducing agent.186 Aldehydes, ketones, acid chlorides, and anhydrides are readily reduced, while esters and nitriles are reduced more slowly. It was observed that aldehydes are reduced slowly at 0°C, but the solubility of the reagent in the solvent plays a role in the rate of reduction.186 An excess of sodium trimethoxyborohydride selectively reduced the ester moiety of an acid-ester, in dimethoxyethane,187 whereas NaBH4 is known to reduce esters with difficulty in many cases (see Section 7.4). Sodium trimethoxyborohydride 183

Ashe, III, A. J.; Al-Taweel, S. M.; Drescher, C.; Kampf, J. W.; Klein, W. Organometallics 1997, 16, 1884. Also see Ashe, III, A. J.; Kampf, J. W.; Nakadaira, Y.; Pace, J. M. Angew. Chem. Int. Ed. 1992, 31, 1255.

184

Kobetz, P. U. S. Patent 3,098,862 1960, Ethyl Corp., New York.

185

Brown, H. C.; Schlesinger, H. I.; Sheft, I.; Ritter, D. M. J. Am. Chem. Soc. 1953, 75, 192.

186

Brown, H. C. Mead, E. J. J. Am. Chem. Soc. 1953, 75, 6263.

187

Bell, R. A.; Gravestock, M. B. Can. J. Chem. 1969, 47, 2099.

337

7.8 HYDRIDE REDUCING AGENTS WITH ELECTRON-WITHDRAWING GROUPS

was shown to be a superior reagent for the selective reduction 1-chloro-2-iodoperfluorocycloalkenes to yield 1-hydro-2-chloroperfluorocycloalkenes.188 Other reagents, including sodium or potassium triethoxyborohydride, tri(tert-butoxy)borohydride, and tri(isopropoxy)borohydride have been prepared.186,189 In that work, it was shown that sodium trimethoxyborohydride was a more powerful reducing agent when compared to NaBH4, in diglyme [1-methoxy-2-(2-methoxyethoxy)ethane] or isopropyl alcohol.189 Brown et al.189 showed that sodium tri(isopropoxy)borohydride reduced acetone significantly faster than NaBH4. This behavior is rationalized by the observation that alkylborates are weaker Lewis acid when compared to borane, and reduction requires transfer of hydride from the parent borohydride ion. This transfer requires separation from the Lewis acid, which is more facile from the weak Lewis acid, the alkyl borate.189 Some derivatives are unstable in ether solvents. Sodium alkoxyborohydrides when formed in THF are known to disproportionate to tetraalkoxyborohydrides, although sodium triphenoxyborohyride is rather stable.190 Potassium triisopropoxyborohydride was shown to be stable to disproportionation in THF, however, and was an excellent reducing agent.191 This reagent was reported to reduce only aldehydes and ketones, however, reacting with virtually no other functional group.192

7.8.2 Acyloxyborohydrides Replacement of hydride with an acyloxy group (dO2CR) leads to acyloxyborohydrides, which are significantly less reactive than NaBH4. Acyloxyborohydrides [NaBH4n(O2CR)n] are remarkably selective,193 as illustrated by the reduction of the aldehyde unit in (Z)-8-oxonon-4-enal in the presence of a ketone moiety.194 Reduction with potassium triacetoxyborohydride (generated by dissolving KBH4 in acetic acid) gave (Z)-9-hydroxynon-5-en-2-one in 60% yield. Good diastereoselectivity can be achieved, as in the Paquette et al.195 synthesis of ()-sanglifehrin A. Tetramethylammonium triacetoxyborohydride reduced the ketone unit in 115 to give 116 in 82% yield and with excellent selectivity for the diastereomer shown. Sodium triacetoxyborohydride has also been used for selective 1,4-reduction of conjugated esters.196 Acetoxyborohydride is also used for the reduction of imines and enamines, which is useful for the synthesis of alkaloid precursors. This reagent also reduces imines.197 O

O KBH(OHAc)3 PhH

OH

CHO

(Z)-9-Hydroxynon-5-en-2-one (60%)

(Z)-8-Oxonon-4-enal Et

Me

Me

Me

Et

Me

Me

Me

Me4NBH(OAc)3

OSiPh2t-Bu

AcOH , MeCN

OH

O

OPMB

OSiPh2t-Bu

115

OH

OH

116 (82%)

188

Natarajan, S.; Soulen, R. L. J. Fluorine Chem. 1981, 17, 485.

189

Brown, H. C. Mead, E. J.; Shoaf, C. J. J. Am. Chem. Soc. 1956, 78, 3616.

190

Golden, J. H.; Schreier, C.; Singaram, B.; Williamson, S. M Inorg. Chem. 1992, 31, 1533.

191

Brown, H,C.; Cha, J. S.; Nazer, B.; Kim, S. C.; Krishnamurthy, S.; Brown, C. A. J. Org. Chem. 1984, 49, 885.

192

Brown, C. A.; Krishnamurthy, S.; Kim, S. C. J. Chem. Soc. Chem. Commun. 1973, 391.

193

Gribble, G. W.; Nutaitis, C. F. Org. Prep. Proc. Int. 1985, 17, 317.

194

Tolstikov, G. A.; Odinokov, V. N.; Galeeva, R. I.; Bakeeva, R. S.; Akhunova, V. R. Tetrahedron Lett. 1979, 4851.

195

Paquette, L. A.; Duan, M.; Konetzki, I.; Kempmann, C. J. Am. Chem. Soc. 2002, 124, 4257.

196

For an example, see Hayahsi, Y.; Rohde, J. J.; Corey, E. J. J. Am. Chem. Soc. 1996, 118, 5502.

197

Smith, R. G.; Lucas, R. A.; Wasley, J. W. F. J. Med. Chem. 1980, 23, 952.

OPMB

338

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

Reductive alkylation is possible when an amine unit is present in the substrate (Section 7.8.3). The carbon chain of the carboxylic acid fragment is transferred to nitrogen, and an N-alkyl amine is the final product. An example is the conversion of 1,2,3,4-tetrahydroquinoline to the N-propyl derivative (1-propyl-1,2,3,4-tetrahydroquinoline) in 56% yield.198 When hydroxyl groups are present in the organic substrate, acetoxylation occurs if no other reducible functional group is present. An example is the reduction of 4-ethoxybenzyl alcohol to the O-acetyl derivative, in 95% yield.199 Although these latter reactions can compete with the reduction of other functional groups, acyloxyborohydrides are important chemoselective reducing agents. NaBH4 , THF CH3CH2CO2H

N

N

H 1,2,3,4-Tetrahydroquinoline

1-Propyl-1,2,3,4-tetrahydroquinoline (56%)

7.8.3 Sodium Cyanoborohydride Alkyl, alkoxy, or acyloxy groups are not the only substituents that can be attached to B in order to generate new borohydride derivatives. The reaction of NaBH4 with HCN gives sodium cyanoborohydride (NaBH3CN),200 for example, which is a remarkably stable reagent that it is significantly less reactive and very selective.201 This reagent does not decompose in acid solution (the pH should be less acidic than pH 3), for example. It is soluble in THF, MeOH, H2O, HMPA, DMF, and sulfolane and these solvents do not react. Sodium cyanoborohydride readily reduces iodides, bromides, or tosylates to the hydrocarbon in excellent yield when HMPA is the solvent,202 even in the presence of carbonyl groups. The reduction of the iodo group in 117 gave 118 in 62% yield without disturbing the lactone moiety was reported by Fukuyama and coworkers203 a synthesis of ()-anisatin. Alcohols are similarly reduced if zinc bromide (ZnBr2) is added to the reagent, as in the reduction of 119–120 in Johnson and coworker’s204 synthesis of longifolene. This reduction presumably proceeds via initial formation of a bromide, although a carbocation intermediate has also been suggested. O

O O

O OMOM

OMOM

NaBH3CN , HMPA THF-HMPA , 66°C

O

O

I

CH3 BnO

BnO 117 Me

118 (62%) Me OH

Me

Me H

ZnBr2 NaBH3CN

Me

Me 119

120

198

Gribble, G. W.; Heald, P. W. Synthesis 1975, 650.

199

Prashad, M.; Jigajinni, V. B.; Sharma, P. N. Indian J. Chem. 1980, 19B, 822.

200

(a) Wittig, G. J. L. Ann. Chem. 1951, 573, 195 (see p. 209). (b) Wade, R. C.; Sullivan, E. A.; Berchied, J. R., Jr.; Purcell, K. F. Inorg. Chem. 1970, 9, 2146.

201

Lane, C. F. Aldrichimica Acta 1975, 8, 3.

202

Hutchins, R. O.; Milewski, C. A.; Maryanoff, B. E. Org. Synth. Coll. Vol. 6 1988, 376.

203

Ogura, A.; Yamada, K.; Yokoshima, S.; Fukuyama, T. Org. Lett. 2012, 14, 1632.

204

Volkmann, R. A.; Andrews, G. C.; Johnson, W. A. J. Am. Chem. Soc. 1975, 97, 4777.

339

7.8 HYDRIDE REDUCING AGENTS WITH ELECTRON-WITHDRAWING GROUPS

Ketones and aldehydes are reduced in acidic media but not at neutral pH. Sodium cyanoborohydride is stable at pH 3–4, where a carbonyl is converted to the protonated form (C]OH+), which is then reduced.205 An example is the reduction of the aldehyde unit in 121 to yield 122 in 89% yield (at pH 4 in this case), in Varela and Di Nardo’s206 asymmetric synthesis of 4-hydroxypipecolic acid. Note that neither the lactone moiety nor the benzyl carbamate moieties were reduced under these conditions. O

O

HO

O

OHC

O

NaBH3CN

NHCO2CH2Ph

MeOH , pH 4

NHCO2CH2Ph 122 (89%)

121

The ability to reduce compounds under acidic conditions is ideal for the reduction of enamines and enamides. Iminium salts are readily reduced at neutral pH.207 When the reduction of an enamine is done in acidic media, protonation of nitrogen yields an iminium salt, which is reduced in situ with cyanoborohydride.205,208 In a synthesis of nominine, Muratake et al.209 reduced enamide 123 to the saturated derivative 124 in 91% yield using an acidic solution of sodium cyanoborohydride. This reagent is also useful for the reductive alkylation of amines.210 In a synthesis of (+)-lysergic acid by Jia and coworkers,211 the reaction of the amine unit in 125 with sodium cyanoborohydride and formaldehyde gave the methylamine, 126, in 81% yield. MeOH2COH2CH2CO

MeOH2COH2CH2CO OH

OH

NaBH3CN , MeOH

H

H

H

H

2.5% HCl–H2O

Me

Me

N

N

Boc

Boc 124 (91%)

123

MeO2C

MeO2C N H Cl

N CH3 NaBH3CN , HCHO

Cl

MeCN , rt

N

N Boc

125

Boc 126 (81%)

7.8.4 Alkoxyaluminum Hydrides Selective reduction of one functional group in the presence of others is important, because multifunctional molecules may contain several reducible functional groups (i.e., lactone + ester + aldehyde). In such a system, LiAlH4 could

205

Borch, R. F.; Bernstein, M. D.; Durst, H. D. J. Am. Chem. Soc. 1971, 93, 2897.

206

Di Nardo, C.; Varela, O. J. Org. Chem. 1999, 64, 6119. For reduction of an aryl aldehyde to a mixture of a benzylic alcohol and the hydrogenolysis product in a synthesis of herbertenediol, see Srikrishna, A.; Satyanarayana, G. Tetrahedron 2006, 62, 2892.

207

For the NaBH3CN reduction of an iminium salt in a synthesis of ()-nakadomarin A, see Bonazzi, S.; Cheng, B.; Wzorek, J. S.; Evans, D. A. J. Am. Chem. Soc. 2013, 135, 9338. Also see Nakagawa, Y.; Stevens, R. V. J. Org. Chem. 1988, 53, 1871. 208

Hanaoka, M.; Yoshida, S.; Mukai, C. J. Chem. Soc. Chem. Commun. 1984, 1703.

209

Muratake, H.; Natsume, M.; Nakai, H. Tetrahedron 2006, 62, 7093.

210

Borch, R. F.; Hassid, A. I. J. Org. Chem. 1972, 37, 1673.

211

Liu, Q.; Zhang, Y.-A.; Xu, P.; Jia, Y. J. Org. Chem. 2013, 78, 10885.

340

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

reduce all of them, as seen in the reduction of 127–128. In this molecule, the ketone, as well as the cyano and ester groups, were reduced at 78°C (95% yield) in Heathcock et al.212 synthesis of fawcettimine. A chemoselective (a term introduced by Trost) reduction will reduce a single functional group in high yield, in the presence of the other functional groups. This section will discuss derivatives of LiAlH4 that have an electron-withdrawing group attached to Al, with the goal of developing a less reactive reducing agent. The main focus of this section will be alkoxyaluminates. Alkoxyaluminum hydride derivatives will be shown to be good reducing agents that are less reactive and more selective than LiAlH4. H

Me

H

Me 1. LiAlH4 , –78°C 2. H3O+

OH

O

OH

CO2Me CN 127

NH2 128

The reaction of lithium aluminum hydride with a ketone or aldehyde leads to mono-di- and trialkoxyaluminates (Section 7.6.1), and each successive reaction with the carbonyl is slower. This observation is consistent with the fact that the alkoxyaluminates are increasingly less reactive as more OR units replace H. As long as an Al–H unit remains in the molecule (i.e., there is an active hydride), that alkoxyaluminate is a reducing agent. Alkoxyaluminates [e.g., LiAlHn(OR)4n] can be prepared by the direct reaction of LiAlH4 with an appropriate alcohol. Addition of n equiv of ROH leads to the alkoxyaluminate with 4  n H remaining on Al. Brown and Shoaf213 showed that many alcohols could be used to prepare alkoxyaluminum hydrides. Four equivalents of methanol, for example, reacted with LiAlH4 to form lithium tetramethoxyaluminum [LiAl(OMe)4], but 3 equiv (in ether, THF, or glyme) gave LiAlH(OMe)3 (lithium trimethoxyaluminum hydride).213 Similarly, 3 equiv of ethanol gave lithium triethoxyaluminum hydride, LiAlH(OEt)3. Brown and McFarlin214 first showed that treatment of LiAlH4 with 4 equiv of the hindered tert-butanol led to the release of only 3 equiv of hydrogen gas (i.e., only 3 equiv of hydride reacted), generating lithium tri-tert-butoxy aluminum hydride, LiAlH(Ot-Bu)3. This hydride failed to react with the excess tertbutyl alcohol at ambient temperatures (although it did react upon prolonged heating at elevated temperatures).83 This reagent is significantly less reactive than LiAlH4 or the other alkoxyaluminum hydrides. Brown and Yoon215 went further, and compared the reducing power of LiAlH4, lithium tri-tert-butoxyaluminum hydride, and lithium trimethoxyaluminum hydride with a variety of functional groups. Lithium trimethoxyaluminum hydride is about as strong a reducing agent as lithium aluminum hydride.216 Although the solid reagent is stable and sold commercially, ether solutions are not very stable. It is often preferable to prepare the reagent as needed, by addition of methanol to a THF slurry of LiAlH4.216 Reduction of epoxides, nitriles, and tosylates was slow, and reduction of nitriles with this reagent gave mixtures of amines and aldehydes, but in poor yield.217 Lithium triethoxyaluminum hydride is also unstable in solution and is usually prepared in situ by addition of 3 equiv of ethanol to LiAlH4 in ether at 0°C.216 This reducing agent showed good selectivity for reduction of nitriles and amides. Both aliphatic nitriles (e.g., isobutyronitrile), as well as aromatic nitriles, are reduced to the corresponding aldehyde (isobutryaldehyde) in good yield.217 Tertiary amides (e.g., N,N-dimethylbutanamide are reduced to the aldehyde (in this case butanal).218 As the steric bulk of the alkyl groups attached to the nitrogen of a tertiary amide increased, the yield of aldehyde was diminished. Reduction of N,N-diethylbutanamide, for example, gave only 47% of the aldehyde and reduction of N,N-diisopropylbutanamide did not give any aldehyde.218 Trimethoxyaluminum hydride and LiAlH4 gave significantly lower yields of aldehydes with both nitriles and tertiary amides.

212

Heathcock, C. H.; Blumenkopf, J. A.; Smith, K. A. J. Org. Chem. 1989, 54, 1548.

213

Brown, H. C.; Shoaf, C. J. J. Am. Chem. Soc. 1964, 86, 1079.

214

Brown, H. C.; McFarlin, R. F. J. Am. Chem. Soc. 1958, 80, 5372.

215

Brown, H. C.; Yoon, N. M. J. Am. Chem. Soc. 1966, 88, 1464.

216

Brown, H. C.; Weissman, P. M. J. Am. Chem. Soc. 1965, 87, 5614.

217

Brown, H. C.; Garg, C. P. J. Am. Chem. Soc. 1964, 86, 1085.

218

Brown, H. C.; Tsukamoto, A. J. Am. Chem. Soc. 1964, 86, 1089.

7.8 HYDRIDE REDUCING AGENTS WITH ELECTRON-WITHDRAWING GROUPS 1. LiAlH(OEt)3

C N

341

CHO

2. H3O+

Isobutyronitrile (2-Methylpropanenitrile)

NMe2

Isobutyraldehyde (2-Methylpropanal) H

1. LiAlH(OEt)3 2. H3O+

O

O

N,N-Dimethylbutanamide

Butanal

Alkoxyaluminates are relatively mild reducing agents that can reduce the carbonyl of a lactone to yield a lactol, an intermediate in Magnus and Hobbs’ synthesis of grandisol,219 which was isolated in 97% yield by reduction of 129 with lithium triethoxyaluminum hydride at 20°C. Trialkoxyaluminum hydrides reduce aldehydes, ketones, esters, and acid chlorides before reducing nitriles and amides. O O Me Me 129

Both lithium diethoxyaluminum hydride [LiAlH2(OEt)2]220 and [LiAlH3(OEt)],221 lithium ethoxyaluminum hydride have been used in syntheses, but LiAlH(Ot-Bu)3 is used more often. This latter reagent is stable in solution, gives useful concentrations in THF, diethyl ether, or glyme,214 and reacts with organic substrates faster in solutions, where it is in higher concentration. It is thermally stable and heating to 165°C for 5 h followed by titration with acid showed that 92% of its active hydrogen was retained.214Although lithium tri-tert-butoxyaluminum hydride is less reactive than the other alkoxyaluminum hydrides discussed above, it easily reduces aldehydes, ketones, some esters, anhydrides, or acid chlorides. Epoxides are reduced very slowly, and nitriles and dimethyl amides are not reduced at all. The use of this reagent is a preferred method for the selective reduction of acid chlorides to aldehydes. The best yield of the aldehyde is realized from aromatic acid chlorides, as in the reduction of p-nitrobenzoyl chloride to p-nitrobenzaldehyde in 80% yield.214 The yield of aldehyde from aliphatic acyl chlorides is generally in the 40–60% range. Alkyl esters do not react, allowing some selectivity for the reduction of ketones and aldehydes in the presence of esters and other functional groups. Depr
es et al.222 reduced the ketone moiety in 130 to alcohol 131 in 83% yield with excellent diastereoselectivity, and without reduction of the lactone unit, in a synthesis of homogynolide B. Although alkyl esters are generally unreactive with this reagent, many phenyl esters are reduced to the aldehyde, as in the conversion of methyl p-chlorobenzoate to p-chlorobenzaldehyde (22°C, 8 h) in 77% yield.223 The presence of the p-chloro unit undoubtedly enhanced the reactivity. O

Me

Me

O LiAlH(Ot-Bu)3, THF

O

Me

Me

O

0°C

H

O

130

219

H

OH

131 (83%)

(a) Hobbs, P. D.; Magnus, P. D. J. Chem. Soc. Chem. Commun. 1974, 856. (b) Hobbs, P. D.; Magnus, P. D. J. Am. Chem. Soc. 1976, 98, 4594.

220

Schwarz, M.; Oliver, J. E.; Sonnet, P. E. J. Org. Chem. 1975, 40, 2410.

221

(a) Morizur, J. P.; Bidan, G.; Kossanyi, J. Tetrahedron Lett. 1975, 4167. (b) Bidan, G.; Kossanyi, J.; Meyer, V.; Morizur, J.-P. Tetrahedron 1977, 33, 2193.

222

Brocksom, T. J.; Coelho, F.; Depr
es, J.-P.; Greene, A. E.; Freire de Lima, M. E.; Hamelin, O.; Hartmann, B.; Kanazawa, A. M.; Wang,Y. J. Am. Chem. Soc. 2002, 124, 15313.

223

Weissman, P. M. Brown, H. C. J. Org. Chem. 1966, 31, 283.

342

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

The reaction of conjugated carbonyls with LiAlH(Ot-Bu)3 yield primarily 1,2-reduction, as in the quantitative reduction of 132 to 133 in Marshall and Ruth’s224 synthesis of globulol. Both sulfonate esters (e.g., the mesylate group in 132, and halides)225 are resistant to reduction with LiAlH(Ot-Bu)3. Me

OMs

Me

1. LiAlH(Ot-Bu)3 THF, 0°C – rt 2. H3O+

O

OMs

HO Me

Me 132

133

Sodium bis(2-methoxyethoxy)aluminum hydride (Red-Al, Vitride) was prepared by Vit et al.226c in 1967. It was prepared in benzene, by the reaction of Na metal with 2-methoxyethan-1-ol, under a hydrogen atmosphere at temperatures >100°C.226 An important advantage of this reagent is its stability to air (it does not ignite in even moist air or oxygen), and it is thermally stable up to 200°C unlike LiAlH4, which can detonate at elevated temperatures. Perhaps the greatest utility of Red-Al is its solubility in aromatic hydrocarbon and ether solvents, which allows it to be conveniently used for applications that require inverse addition of hydrides.

2

MeO

H

Na, Al, H2

OH

MeO

PhH, >100°C

2-Methoxyethan-1-ol

O

H Al

O

Na+ OMe

Sodium bis(2-methoxyethoxy)aluminum hydride Red-Al

Despite the presence of the two alkoxy groups, the reducing power of Red-Al is close to that of LiAlH4, and it reduces aldehydes, ketones,227 carboxylic acids,228 and acid derivatives to alcohols. The strength of this reagent is reflected in the fact that it reduces ketones and aldehydes to the alcohol at 78°C.229 Esters are reduced to alcohols using Red-Al, in a manner that is identical with LiAlH4.230 The reduction of 134, for example, gave a 99% yield of 135 in a synthesis of carbazomadurin A by Choshi and coworkers.231 The reduction of conjugated carbonyls yields primarily 1,2-reduction to an allylic alcohol.232,233 The selectivity for 1,2-reduction can be changed to selectivity for 1,4-reduction by the addition of cuprous bromide (CuBr) to Red-Al, as in the conversion of 136 ! 137 in 72% yield.234 The active reducing agent is probably copper hydride. It is well known that addition of cuprous salts (Cu+) promotes conjugate addition (also see organocuprates in Section 12.3.2.2).235 Reaction of cuprous bromide (CuBr) and Red-Al gave a Na complex, whereas reaction of CuBr and LiAlH(OMe)3 gave a Li complex.236,237 224

Marshall, J. A.; Ruth, J. A. J. Org. Chem. 1974, 39, 1971.

225

For a synthetic example, see Daniewski, A. R.; Kiegel, J. J. Org. Chem. 1988, 53, 5534.

226

(a) Fieser, L. F.; Fieser, M. Reagents for Organic Synthesis, Vol. 2, Wiley: New York, 1969, p. 382. (b) Fieser, L. F.; Fieser, M. Reagents for Organic Synthesis, Vol. 3, Wiley: New York, 1972, p. 260. (c) Vit, J.; Cásensky, B.; Machácek, J. French Patent, 1,515,582 1968 (Chem. Abstr. 1967, 70: 115009x). 227  Capka, M.; Chvalovský, V.; Kochloefl, K.; Kraus, M. Collect. Czech. Chem. Commun. 1969, 34, 118. 228

Zurfl€ uh, R.; Dunham. L. L.; Spain, V. L.; Siddall, J. B. J. Am. Chem. Soc. 1970, 92, 425.

229

Stotter, P. L.; Friedman, M. D.; Minter, D. E. J. Org. Chem. 1985, 50, 29.

230

See (a) Avery, M. A.; Jennings-White, C.; Chong, W. K. M. J. Org. Chem. 1989, 54, 1789. (b) Pigza, J. A.; Han, J.-S.; Chandra, A.; Mutnick, D.; Pink, M.; Johnston, J. N. J. Org. Chem. 2013, 78, 822.

231

Hieda, Y.; Choshi, T.; Fujioka, H.; Hibino, S. Eur. J. Org. Chem. 2013, 7391.

232

(a) Markezich, R. L.; Willy, W. E.; McCarry, B. E.; Johnson, W. S. J. Am. Chem. Soc. 1973, 95, 4414. (b) McCarry, B. E.; Markezich, R. L.; Johnson, W. S. Ibid. 1973, 95, 4416. 233

For an example taken from a synthesis of azaspiracid-1, see Nicolaou, K. C.; Pihko, P. M.; Bernal, F.; Frederick, M. O.; Qian, W.; Uesaka, N.; Diedrichs, N.; Hinrichs, J.; Koftis, T. V.; Loizidou, E.; Petrovic, G.; Rodriquez, M.; Sarlah, D.; Zou, N. J. Am. Chem. Soc. 2006, 128, 2244. 234

Bestmann, H. J.; Schmidt, M. Tetrahedron Lett. 1986, 27, 1999.

235

Paquette, L. A.; Schaefer, A. G.; Springer, J. P. Tetrahedron 1987, 43, 5567.

236

Semmelhack, M. F.; Stauffer, R. D.; Yamashita, A. J. Org. Chem. 1977, 42, 3180.

237

(a) Vedejs, E.; Fedde, C. L.; Schwartz, C. E. J. Org. Chem. 1987, 52, 4269. (b) Negishi, E.; Akiyoshi, K. Chem. Lett. 1987, 1007.

343

7.8 HYDRIDE REDUCING AGENTS WITH ELECTRON-WITHDRAWING GROUPS

OMOM

OMOM Red-Al, Toluene

N H

MeO

CO2Et

OH

0°C, 3 h

N H

MeO

134

135 (99%)

O

Red-Al

O

OTHP

O O

OTHP

CuBr

O

O 136

137 (72%)

Other functional groups can be reduced (e.g., epoxides to alcohols238 and lactams to amines).239 Epoxides are reduced at the less hindered carbon,240 but the product distribution can vary with the solvent, especially when the steric hindrance is similar at each carbon of the epoxide. As with LiAlH4, Red-Al reduces aromatic nitriles to amines, although aliphatic nitriles are difficult to reduce and give poor yields.241 There are some interesting differences in selectivity, relative to LiAlH4. Cyanohydrin 1-hydroxycyclopentane-1-carbonitrile, for example, was reduced to α-hydroxyaldehyde, 1-hydroxycyclopentane-1-carbaldehyde,242 rather than to the amino-methyl compound obtained with LiAlH4. OH CN

OH

1. Red-Al 2. H3O+

1-Hydroxycyclopentane1-carbonitrile

CHO 1-Hydroxycyclopentane1-carbaldehyde

Hydrogenolysis of CdX bonds is usually faster than with LiAlH4, and Red-Al is preferred to LiAlH4 for reduction of aliphatic halides and many aromatic halides.243 Primary aliphatic monohalides are reduced to the hydrocarbon (1-bromoheptane was reduced to heptane, e.g., in 99% yield).243 Aromatic halides are also reduced, as in the conversion of bromobenzene to benzene in 99% yield.243 Under similar conditions, however, chlorobenzene gave only 15% of benzene. Selective mono-reduction is possible when more than one halide is present. Reduction of the gem-dibromide moiety in 7,7-dibromobicyclo[4.1.0]heptane, for example, gave the syn-monobromide (7-bromobicyclo[4.1.0]heptane) as the major product.244 Br

1. Red-Al

Br

Br

2. H3O+

H

7,7-Dibromobicyclo[4.1.0]heptane

7-Bromobicyclo[4.1.0]heptane

Sulfonate esters are reduced via CdO cleavage, as with LiAlH4. Primary tosylates and mesylates are reduced to the methyl derivative.245 Sterically hindered sulfonate esters often give significant amounts of SdO cleavage during the reduction. Although LiAlH4 does not readily reduce sulfonamides to the corresponding amine (Section 7.6.2.2),246 238

For an example taken from a synthesis of ()-8-O-methyltetrangomycin (MM 47755), see Kesenheimer, C.; Groth, U. Org. Lett. 2006, 8, 2507.

239

For an example taken from a synthesis of ()-ibogamine, see White, J. D.; Choi, Y. Org. Lett. 2000, 2, 2373.

(a) Gao, Y.; Sharpless, K. B. J. Org. Chem. 1988, 53, 4081. For an exanmple taken from a synthesis of δ-trans-tocotrienoloic acid, see (b) Maloney, D. J.; Hecht, S. M. Org. Lett. 2005, 7, 4297. 241   Cerny, M.; Málek, J.; Capka, M.; Chvalovský, V. Collect. Czech. Chem. Commun. 1969, 34, 1033. 240

242 243

Schlosser, M.; Brich, Z. Helv. Chim. Acta 1978, 61, 1903.  Capka, M.; Chvalovský, V. Collect. Czech. Chem. Commun. 1969, 34, 3110.

244

Sydnes, L.; Skattebøl, L. Tetrahedron Lett. 1974, 3703.

245

Zobácová, A.; Hermánková, V.; Jarý J. Collect. Czech. Chem. Commun. 1977, 42, 2540.

246

Gold, E. H.; Babad, E. J. Org. Chem. 1972, 37, 2208.

344

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

cleavage of the NdS bond with Red-Al is facile, as in the conversion of N-tosyl deoxyephedrine to deoxyephedrine in 64% yield.246 It was assumed that the sulfonate fragment lost from nitrogen in a sulfonamide was reduced to a thiol. Me Ph

Me

1. Red-Al, PhMe Reflux , 72 h

Me N

2.

Ts

Ph

H3O+

N

Me

H Deoxyephedrine (64%)

N-Tosyl deoxyephedrine

Benzylic alcohols are reduced to the hydrocarbon via hydrogenolysis using Red-Al at elevated temperatures. The reaction occurs more readily when the aromatic ring contains an electron-releasing group in the ortho- or para-position, but the temperatures required can be >140°C.247 An amino group at a benzylic position is subject to hydrogenolysis. Reduction of 4-(N,N-dimethylamino)benzaldehyde with Red-Al at 141°C gave a mixture of 4-hydroxymethyl-N,N-dimethylaniline (63% yield) and 4-methyl-N,N-dimethylaniline (8%).248 Similar reduction of 4-aminoacetophenone gave only toluidine, however, in 90% yield.248 As with LiAlH4, propargylic alcohols are reduced to allylic alcohols.249,250 Red-Al behaves as a base in certain applications, if a readily reducible functional group is not present in the molecule. Kametani et al.251 used Red-Al to induce a Steven’s rearrangement (Section 12.5.3.2) of berbine methiodide to spirobenzylisoquinoline (138) (9.4%) plus 139 (22.5%). The yield of rearranged products was poor, which is typical for this application of Red-Al. MeO

MeO MeO

Me

1. Red-Al , Reflux Dioxane, 24 h

N

MeO N Me

MeO

+

2. H3O

H

+

OMe

MeO

N

Me H

H

OMe

OMe

OMe

OMe OMe

Berbine methiodide

139 (22.5%)

138 (9.4%)

The functional group transforms most useful with Red-Al follows: R1 OH R

X

O R

R

R

O R

O R

R N

R N H

R1 SO2R2

R CH3

R CH2OSO2R1

7.9 STEREOSELECTIVITY IN REDUCTIONS Hydride reduction of prochiral ketones generally yields alcohol products with good diastereoselectivity, but each diastereomer is racemic. An example is the reduction of (S)-2-methoxy-1,2-diphenylethan-1-one with LiAlH4, dibal, KBH4, and K-Selectide.252 All of these reducing agents yield (1R,2S)-2-methoxy-1,2-diphenylethan-1-ol as the major product ( 70–90%) with 1–13% of (1S,2S)-2-methoxy-1,2-diphenylethan-1-ol. There is some difference in the diastereoselectivity (anti-syn selectivity) for reduction of ketones, which depends on the reducing agent, as well as the reaction conditions, but the major product remains the same. When diastereomers are 247 248

 Cerny, M.; Málek, J. Tetrahedron Lett. 1969, 1739.  Cerny, M.; Málek, J. Collect. Czech. Chem. Commun. 1970, 35, 1216.

249

(a) Mayer, H. J.; Rigassi, N.; Schwieter, U.; Weedon, B. C. L. Helv. Chim. Acta 1976, 59, 1424. (b) Kienzle, F.; Mayer, H.; Minder, R. E.; Thommen, W. Ibid. 1990, 524.

250

Jones, T. K.; Denmark, S. E. Org. Synth. Coll. Vol. 7 1985, 64, 182.

251

Kametani, T.; Huang, S-P.; Koseki, C.; Ihara, M.; Fukumoto, K. J. Org. Chem. 1977, 42, 3040.

252

Davis, F. A.; Haque, M. S.; Prezeslawski, R. M. J. Org. Chem. 1989, 54, 2021.

345

7.9 STEREOSELECTIVITY IN REDUCTIONS

formed by reduction of carbonyl groups, is it possible to predict the relative stereochemistry of such reductions and also the absolute stereochemistry? The following sections will attempt to answer these questions. O

HO

Ph H

H

H

Reduction

Ph

Ph

Ph H

OMe

(S)-2-Methoxy-1,2diphenylethan-1-one

+

OH Ph

Ph

OMe

H

(1R,2S)-2-Methoxy-1,2diphenylethan-1-ol

OMe

(1S,2S)-2-Methoxy-1,2diphenylethan-1-ol

7.9.1 Selectivity in the Reduction of Carbonyl Derivatives Containing a Stereogenic Carbon The reduction of (S)-2-methoxy-1,2-diphenylethan-1-one gave two diastereomers, but one was the major product. Clearly, the reduction proceeded with good-to-excellent stereoselectivity. In fact, the stereoselectivity was due, in large part, to the presence of the (S)-stereogenic carbon at C2. When the prochiral center is in a molecule that already contains a stereogenic center, reduction generates diastereomers. Two general cases are illustrated by the reduction of (S)-7-phenyloctan-2-one or (R)-3-methylpentan-2-one. In the former case, (S)-7-phenyloctan-2-one contains a (7S)-stereogenic center, but it is too far removed from the prochiral center to exert much influence. Reduction therefore generates a racemic mixture at C2 (the acyl carbon atom), so two diastereomeric products are formed. Note that the absolute configuration of the stereogenic center at C7 is unchanged in both (2S,7S)-7-phenyloctan-2-ol and (2R,7S)-7-phenyloctan-2-ol, and it is clear that C7 does not influence reduction of the carbonyl moiety. Reduction of this compound will be, more or less, the same as reduction of a ketone with no stereogenic center and the resulting mixture should be close to a 1:1 mixture of syn and anti diastereomers [(2S,7S)-7-phenyloctan-2-ol and (2R,7S)-7-phenyloctan-2-ol]. (R)-3-Methylpentan-2-one has a (3R)-stereogenic center that is adjacent to the carbonyl carbon at C2. The interaction of the reducing agent with this ketone will be influenced by the incipient chirality, which will influence the approach of the incoming reagent. Reagents generally approach from the face presenting the least amount of steric hindrance. (R)-3-methylpentan-2-one is drawn as two rotamers, A and B, and the re and si faces for the prochiral carbonyl carbon are labeled. Note that the re and si faces are on opposite sides of the molecule for each rotamer. In (R)-3-methylpentan2-one-A, approach over the hydrogen atom (the si face) will yield (2R,3R)-3-methylpentan-2-ol, but approach over the more hindered methyl group (the re face) will yield (2S,3R)-3-methylpentan-2-ol (these two alcohols are diastereomers). This analysis assumes that (R)-3-methylpentan-2-one-A is “frozen” into the rotamer shown, which is incorrect of course. This ketone and similar acyclic systems are not locked into a single conformation, but only two rotamers are considered, where the carbonyl either down [(R)-3-methylpentan-2-one-B in the first analysis] or up [(R)-3methylpentan-2-one-B] relative to the methyl group on the stereogenic carbon. In both cases, it is assumed that the carbonyl bisects a plane that includes H and Me, although many other rotamers are possible. O

OH 1. LiAlH4

+

2. H3O+

H

H

Ph

Me

H

Ph

(S)-7-Phenyloctan-2-one H

H

Me

+

re

OH

OH

(R)-3-Methylpentan-2-one-A Me

(2R,7S)-7-Phenyloctan-2-ol H

Me

Reduction

H

(2S,3R)-3-Methylpentan-2-ol H

si O

Ph

(2S,7S)-7-Phenyloctan-2-ol

si

O

OH

Reduction

(2R,3R)-3-Methylpentan-2-ol

Me

H OH

+

Me OH

re (R)-3-Methylpentan-2-one-B

(2R,3R)-3-Methylpentan-2-ol

(2S,3R)-3-Methylpentan-2-ol

346

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

Reduction from the si face in conformation (R)-3-methylpentan-2-one-B leads to (2S,3R)-3-methylpentan-2-ol, and reduction from the re face yields (2R,3R)-3-methylpentan-2-ol. In both rotamers of (R)-3-methylpentan-2-one, the major product of reduction is the one that occurs from the si face, over the less sterically demanding hydrogen atom. However, reduction of rotamer (R)-3-methylpentan-2-one-A leads to (2R,3R)-3-methylpentan-2-ol as the major product, whereas reduction of (R)-3-methylpentan-2-one-B from the si face yields (2S,3R)-3-methylpentan-2-ol as the major product. If there is free rotation about the C3dC2 bond, the reducing agent will react with both rotamers. If there is no control of the rotamer population at the time of reduction, both diastereomers should form in roughly equal amount, which indicates little or no stereoselectivity. Good stereoselectivity in a reaction demands that one face (re vs si) is preferred and also that one rotamer is preferred. Prediction of stereochemistry therefore demands a model that makes assumptions about both. In other words, does the reduction of (R)-3-methylpentan-2-one lead to a preference for one diastereomer over the other; Is it diastereoselective? It would be quite useful if the major diastereomer for the reduction of molecules (e.g., (R)-3-methylpentan-2-one) could be predicted. There are models that can assist in such predictions, but they require certain assumptions about the conformation of the carbonyl compound, and the angle of approach of the reducing agent to the carbonyl compound. In fact, the best models are used for ketones or aldehydes that have a stereogenic carbon α to the carbonyl carbon. The pertinent models for predicting asymmetric induction in systems containing an adjacent stereogenic center have been discussed by Morrison and Mosher.253 Cram et al.254 suggested a model for asymmetric induction in ketones [e.g., (R)-3-methylpentan-2-one] known as Cram’s open-chain model (Cram’s model), or simply Cram’s rule.255 This model uses a Newman projection (e.g., 140), and assumes a kinetically controlled reaction (nonequilibrating and noncatalytic) for asymmetric 1,2-addition to aldehydes and ketones. The three groups in 140 that are attached to the stereogenic center are labeled RS (small-sized substituent), RM (middle-sized substituent), and RL (large-sized substituent). Determining the relative size of the substituents is not always straightforward, since the largest group is defined as the most sterically demanding group relative to the other substituents and the incoming reagent. Ab initio calculations by Bienz and coworkers256 showed that anti-Cram selectivity was favored for compounds that possess very large groups, and that silicon containing compounds showed the same type of selectivity as the carbon analogue. Cram model 140 assumes that one predominant rotamer mimics the transition state for reduction, with a large-sized substituent (RL) syn- to the R1 group attached to the carbonyl. The hydride is delivered from the less sterically hindered face (over the smallest substituent RS) to yield 141 as the major diastereomer. The stereoselectivity of the reaction is determined by the extent of the O $ RS or O $ RM interaction with the metal hydride. As the steric bulk between RS and RM increases, selectivity improves. If RS and RM are close in size, this model predicts little or no selectivity. This model also assumes the RL$ R1 interaction is minimal, which is not entirely correct.257 As the steric bulk of R1 increases, the proportion of the anti-product (143) produced by LiAlH4 reduction of chiral ketone 142 increases (48% when R1 ¼ Me; 52% for R1 ¼ Et; 70% for R1 ¼ i-Pr; 96% for R1 ¼ t-Bu). This observation suggests that the RS  RM$ R1 rather than the RS  RM$ O interaction is important in this model. Al

Li

H

O RS

RM

•

•

R1 RL

R1 RL 140

141

O Me

H

Al O RM

RS H

LiAlH4

Me

H

HO

R1

R1 Ph

Ph 142

253

H

143

Morrison, J. D.; Mosher, H. S. Asymmetric Organic Reactions; American Chemical Society: Washington, DC, 1976, pp. 84–132.

254

(a) Cram, D. J.; AbdElhafez, F. A. J. Am. Chem. Soc. 1952, 74, 5828. (b) Cram, D. J.; Kopecky, K. R. Ibid. 1959, 81, 2748. (c) See also, Mengel, A. Reiser, O. Chem. Rev. 1999, 99, 1191.

255

Eliel, E. L.; Wilen, S. H.; Mander, L. N. Stereochemistry of Organic Compounds, Wiley: New York, 1994, p. 879.

256

Smith, R. J.; Trzoss, M.; B€ uhl, M.; Bienz, S. Eur. J. Org. Chem. 2002, 2770.

257

Ref. 253, p. 91.

347

7.9 STEREOSELECTIVITY IN REDUCTIONS

An example that allowed prediction of the selectivity in a reduction is taken from a synthesis of (+)-itomanallene A by Kim and coworkers,258 in which the L-Selectride reduction of 144 gave a >85% yield of a single diastereomer (145). The Cram model for 144 (see 146) must retain the absolute sterochemistry, of course, so the large group (the alkoxy unit, abbreviated for convenience as the OR chain) must be syn- to the allylic group on the carbonyl carbon, analogous to the model structure 140. Delivery of hydride over the small group (H) predicts the correct diastereomer, with the (S)-configuration at the newly formed alcohol.

O (S)

L-Selectride

OTIPS

(S) (S)

HO

(S) (S)

THF, –78°C

OPMB

O

O

OPMB

O

H R

OTIPS

(S) (S)

R

(S)

H O

(S) (S)

O

OH 145 (>85%)

144

B

Li

H

O

H H

O

H

B

•

•

H H3

H

(S)

O+

O HO R

O

R

R

(S)

O

146

145

The selectivity of a Selectride reagent is influenced by the addition of metal salts, just as the reactivity of NaBH4 was modified in Section 7.5. Tanis et al.259 showed the selectivity of reduction using L-Selectride was modulated by the addition of transition metals. Magnesium salts were the most effective, consistent with the chelating properties of that metal, and tetravalent titanium were also effective, but zinc iodide (ZnI2) was much less effective.259 When the relative difference in size of the groups is negligible, little or no diastereoselectivity is observed. Reduction of (S)-1,2-diphenyl-2-(4-methylphenyl)ethan-1-one, for example, gave a 1:1 mixture of syn- and anti-diastereomers 147 and 148.260 The RM/RL groups are phenyl and 4-methylphenyl and at the reactive center (the carbonyl), the methyl group of the p-tolyl moiety is rather far away. As the hydride reagent approaches, the phenyl and tolyl groups are virtually identical in size when analyzed by the Cram model. Ph 1. LiAlH4

O Ph H

Me

2. H3O+

Ph

H HO Ph

HO Me

+

H

Me

Ph H

H

(S)-1,2-Diphenyl-2-(4-methylphenyl)ethan-1-one

Ph

147

148

When an oxygen, sulfur, or nitrogen is attached to the carbon α to the carbonyl, the metal of the hydride can coordinate (chelate) to both the heteroatom and the carbonyl oxygen, which effectively locks the molecule into a single rotamer. The Cram designations of large, medium, and small no longer have the same meaning. With these substituent atoms, chelation leads to a cyclic structure and a different model is required, the so-called Cram cyclic model 258

Jeong, W.; Kim, M. J.; Kim, H.; Sanghee Kim, Kim, D.; Shin, K. J. Angew. Chem. Int. Ed. 2010, 49, 752.

259

Tanis, S. P.; Chuang, Y.-H.; Head, D. B. J. Org. Chem. 1988, 53, 4929.

260

Stocker, J. H.; Sidisunthorn, P.; Benjamin, B. M.; Collins, C. J. J. Am. Chem. Soc. 1960, 82, 3913.

348

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

(more commonly called the Cram chelation model).254c,260,261 When a hydroxyl group, an ether, amino, amide, thiol, or thioether group is in the α-position, chelation to that group makes it eclipse the carbonyl, as shown in complex 149. Delivery of hydride from the less hindered face (over RS) predicts the major diastereomer. O H O • RS

RM

R 149

In one example, taken from Sudhakar and Raghavaiah’s262 synthesis of (+)-varitriol, Super-Hydride reduced the ketone unit in 150 to yield the (R)-alcohol in 151 with a selectivity of 32.3:1. The dCH]CHAr unit in 150 is represented by “R,” and the hydroxyl group at the α-carbon forms a chelated complex (152) with Super-Hydride. Coordination with the hydroxyl unit in the cyclic complex 152, leads to delivery of hydride over the smaller hydrogen atom leads to the correct stereoselectivity in 153, and subsequent quenching with aqueous acid gave 151. In this work, SuperHydride was the only reagent that gave such good stereoselectivity. Both NaBH4/CeCl3 (1:1) and Zn(BH4)2 (1.3:1), and even K-Selectride (1.2:1) led to relatively poor selectivity. OMe OAc

OMe OAc

LiEt3BH , Ether (E)

OH

O

(R)

–78°C

OH

(E) (R)

(S) (R)

OH

150

H

AcO MeO

(R)

(R)

O

B

O

(R)

151 (82%, 32.3:1) Li O H O

OH OH H

= R

R

H

H H

(R)

OH (R) (E)

O

H O

(R)

H2O

•

• H

O

B

R

O

(R)

OAc OMe

152

153

151

The selectivity predicted by this model, sometimes called anti-Cram selectivity, will depend on the reagent, the chelating ability of the substituent, the coordinating ability of the solvent, as well as the relative size of the substituents (RS and RL).260,263 In derivatives where a normally coordinating hydroxyl is blocked (protected; see Section 5.3.1), the reduction proceeds with normal Cram selectivity.264 The distance between the chelating group and the carbonyl is an important consideration. Oishi and coworkers265 studied the diastereoselectivity in reduction reactions of α-alkoxy and α-hydroxy ketones (154) and they examined several reducing agents.265 When R1 ¼ H, there was a distinct preference for the anti (erythro) diastereomer 156 when Zn(BH4)2 was used for reduction. The anti-product was the major one with LiAlH4, but the reaction was less selective. The Cram chelation model for the anti-transition state (157), to yield 156, was favored over the model for syn-selectivity, where delivery over the methyl group would yield 155. When hydroxyl was replaced 261

Cram, D. J.; Wilson, D. R. J. Am. Chem. Soc. 1963, 85, 1245.

262

Sudhakar, G.; Raghavaiah, J. J. Org. Chem. 2013, 78, 8840.

263

Ref. 253, pp. 100–108.

264

Yamada, S.; Koga, K. Tetrahedron Lett. 1967, 1711.

265

Nakata, T.; Tanaka, T.; Oishi, T. Tetrahedron Lett. 1983, 26, 2653.

349

7.9 STEREOSELECTIVITY IN REDUCTIONS

with a bulky silyl ether (R1 ¼ SiPh2t-Bu), the selectivity changed from anti to syn. Chelation was diminished and Cram’s rule (158) must be used to predict the correct syn-selectivity. OR1 R

H4B

R2

H

R OH

154

155

Zn

OR1

+

R2

O

R2 OH 156 H

H

Li Al

O

O

H

Me

H

H

H

•

•

H

R

H

O

H B

OR1 [H]

Me

H

C5H11 OSiMe2t-Bu

C5H11 157

158

The presence of an α-alkoxy substituent does not always lead to anti-Cram selectivity. Reduction of 159 gave 82% of the syn-diastereomer, 160.266 Analysis of the Cram cyclic model for 159 gives an incorrect prediction of the anti-diastereomer as the major product. The normal Cram model predicts the correct syn-diastereomer (160) but, in reality, reduction requires that the alkoxy substituent at the C3 position be coordinated with the metal of the reducing agent. As the nature of the alkoxy substituent changes, it is clear that the model used to predict the major diastereomer may change.

O

CH2OEt Me O

1. LiAlH(Ot-Bu)3

OH

O

CH2OEt Me

2. H3O+

BnO

BnO Me Me

Me Me 160 (82%)

159

The reduction of 159 illustrated that chelating substituents at the carbon β to a carbonyl can influence the diastereoselectivity of hydride reductions, which is particularly important for the reduction of 1,3-diketones and β-ketoesters. When the α-substituent is a halogen, as in (S)-2-chloro-1-phenylpropan-1-one, the open-chain model does not predict the correct results nor does the chelation model. Cornforth et al.267 suggested that the electron pairs on the carbonyl oxygen and on the halogen repel, which leads to an anti-conformation, as in 161. In effect, the halogen becomes RL with delivery of hydride over the smallest group. This arrangement has come to be called the Cornforth model. Reduction of (S)-2chloro-1-phenylpropan-1-one gave a 75:25 mixture of (1S,2S)-2-chloro-1-phenylpropan-1-ol and (1R,2S)-2-chloro-1phenylpropan-1-ol, and the Cornforth model predicted (1S,2S)-2-chloro-1-phenylpropan-1-ol as the major product.268 Me O

•

H Me

O

Ph Cl

(S)-2-Chloro-1-phenylpropan-1-one

H H

LiAlH4 ; H3O+

Me

H

OH

H

Me HO

Ph

Cl

Ph

Cl

Cl 161

H

+

(1S, 2S)-2-Chloro-1-phenylpropan-1-ol

Ph

(1R, 2S)-2-Chloro-1-phenylpropan-1-ol

266

Hoagland, D.; Morita, Y.; Bai, D. L.; M€arki, H.-P.; Kees, K.; Brown, L.; Heathcock, C. H. J. Org. Chem. 1988, 53, 4730.

267

Cornforth, J. W.; Cornforth, R. H.; Mathew, K. K. J. Chem. Soc. 1959, 112.

268

Bodot, H.; Dieuzeide, E.; Jullien, J. Bull. Soc. Chim. Fr. 1960, 27, 1086.

350

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

The Cram open-chain model assumed the presence of a reactive species that led to a higher energy eclipsed conformation in the final product. The Cram open-chain model fails to predict the correct diastereoselectivity for many acyclic molecules that contain heteroatom substituents.269 The Cram model also fails to predict the correct diastereomer when the relative size of RS and RM is close. Karabatsos270 introduced a new model, in an attempt to correct the deficiencies of the Cram model. He assumed a reactant-like transition state that showed little bond breaking or making. The rotamer chosen for the model showed the groups on the stereogenic carbon to be aligned as in a normal sp3dsp2 bond, which led to three rotamers that were considered by the Karabatsos model (162–164). The incoming reagent was assumed to approach over the RS group.271 Note that reaction over RS in all three rotamers yields a product with an anti-conformation. Karabatsos271 showed that 164 was favored over 162 by 0.8 kcal (3.35 kJ) mol1 when RS ¼ H and RM ¼ Me. The selectivity of a reduction was assumed not to be dependent on the higher energy 164, but rather on rotamers 163 and 164. It was then assumed that the destabilizing O $ RL and O $ RM energy interactions in these two rotamers would determine the diastereoselectivity of the reaction. Although this model is inadequate when RS is relatively large, or when applied to cyclic molecules, it offers advantages over the Cram model in some applications. RL

RS O • RM

RL

RM

O

O

•

•

RM

RS

RS

R

RL R 164

R 163

162

Felkin and Anh and their coworkers272 proposed an alternative model that gave somewhat better results in these systems.273 The model was based on several assumptions: (1) the transition states are reactant-like, (2) the previous eclipsed models introduce significant torsion strain in the partial bonds of these transition states, (3) the important steric interaction in these transition states involves the group attached to the prochiral center and the incoming group, and (4) polar effects stabilize the transition states in which the separation of the incoming group and any electronegative substituent at the α-carbon is greatest and destabilize the others. Felkin and coworkers272 concluded that models 165 and 166 were the most important. The use of these structures to predict diastereoselectivity is called the Felkin-Anh model. RM

O

•

"H "

O RL

RS R1 165

RL

RS

• R1

"H " RM

166

The correctness of this model has been called into question, based on calculations done by Bientz and coworkers.274 “According to our calculations, the Felkin-Anh model does not correctly represent the modes of reaction on the course of nucleophilic attack to α-chiral carbonyl compounds. For typical reactions, the two transition states with the M and L groups arranged ‘inside’ and ‘anti’ are the key structures, and discrimination is due to the repulsive interactions of the incoming nucleophiles with the ‘inside’ groups. For compounds with very large groups L, the relevant transition structures are those with this L group placed ‘anti’ (Felkin-Anh conformations). For the discrimination, the interaction of the incoming nucleophile with the ‘inside’ rather than the ‘outside’ group is relevant. This leads to anti-Cram 269

Ref. 253, p. 118.

270

Karabatsos, G. J. J. Am. Chem. Soc. 1967, 89, 1367.

271

(a) Karabatsos, G. J.; Hsi, N. J. Am. Chem. Soc. 1965, 87, 2864. (b) Karabatsos, G. J.; Hsi, N. Tetrahedron 1967, 23, 1079. (c) Karabatsos, G. J.; Krumel, K. L. Ibid. 1967, 23, 1097. (d) Karabatsos, G. J.; Fenoglio, D. J.; Lande, S. S. J. Am. Chem. Soc. 1969, 91, 3572. (e) Karabatsos, G. J.; Fenoglio, D. J. Ibid. 1969, 91, 1124, 3577.

272

(a) Ch
erest, M.; Felkin, H.; Prudent, N. Tetrahedron Lett. 1968, 2199. (b) Ch
erest, M.; Felkin, H. Tetrahedron Lett. 1968, 2205. (c) Anh, N. T.; Eisenstein, O. Nov. J. Chem. 1977, 1, 61. 273

Ref. 255, p. 881.

274

Smith, R. J.; Trzoss, M.; B€ uhl, M.; Bienz, S. Eur. J. Org. Chem. 2002, 2770.

351

7.9 STEREOSELECTIVITY IN REDUCTIONS

selectivity for such compounds. Acylsilanes should show the same selectivity as the related ketones. The selectivity, however, is expected to be lower and the relevant transition structures might differ from those of the carba analogues.”274 When RM and RS are relatively close in size, the Felkin-Anh model predicts modest-to-poor selectivity because there is little difference in the destabilizing RM $ O, RS $ O or RS $ R1, RM $ R1 interactions. As RM or R1 increases in steric bulk, the increased RM$ R1 destabilizing interaction in 166 will favor 165.275 Reduction of (S)-3-phenylbutan-2-one with LiAlH4 was shown to give a 2:1 mixture of (2R,3S)-3-phenylbutan-2-ol and (2S,3S)-3-phenylbutan-2-ol.254c For an analysis of this reduction, the two models of interest are 167, which predicts (2R,3S)-3-phenylbutan-2-ol and 168, which predicts (2S,3S)-3-phenylbutan-2-ol. A molecular model is shown for each structure. The Me $ Me interaction in 168 is greater than the analogous H $ Me interaction in 167, making 167 more favorable, and delivery of hydride via 167 predicts (2R,3S)-3-phenylbutan-2-ol as the major product.

Me H Ph

O 1. LiAlH4 (S)

Me

(S)-3-Phenylbutan-2-one

Me

2. H3O+

Me H Ph

HO (S) (R)

H

+

Me

(2R, 3S)-3-Phenylbutan-2-ol

Me H Ph

H (S) (S)

OH Me

(2S, 3S)-3-Phenylbutan-2-ol

O

O

H

–

"H "

•

Ph

Ph

•

–

"H " Me

H Me

Me 167

168

The Felkin-Anh and Cram models are best applied to acyclic systems. Problems arise when any of these models are used to predict the products generated by the reduction of cyclic ketones. These problems will be analyzed and new models for predicting diastereoselectivity in the reduction of cyclic molecules will be discussed in Sections 7.9.4 and 7.9.5. Table 7.1 attempts to compare the Cram and Felkin-Anh model with the molecular modeling technique known as a LUMO map. Using the molecular modeling software Spartan, two models can be generated that are particularly useful: the local ionization potential map for electrophilic reactions and the LUMO map for nucleophilic reactions. The local ionization potential provides a measure of the relative ease of electron removal (ionization) at any location around a molecule. A LUMO map is the lowest unoccupied molecular orbital (the LUMO; see Sections 14.2 and 14.3) mapped upon the electron density for the molecule. LUMO map indicates the extent to which the LUMO “can be seen” at the “accessible surface” of a molecule. It is generated by displaying the (absolute) value of the LUMO, indicating the “most likely” regions for electrons to be added, on top of a surface of electron density.276 The region for electrons to be added corresponds to the face from which nucleophilic attack is most likely to occur. Electron density is the measure of the probability of an electron being present at a specific location. In the context of this discussion, this map can provide an indication of the most favorable approach of a nucleophile to the electrophilic carbonyl carbon. The LUMO map is shown for the reaction of (S)-3-phenylbutan-2-one with LiAlH4 in Fig. 7.2. In this calculation, the molecule is constrained into the conformation suggested by the Cram model prior to the calculation, and the calculation delivered the LUMO map shown. Two views of this map are shown for either face of the carbonyl, the pro-R and the pro-S (Section 1.4.5), indicating a preference for one face over the other (indicated by a greater “blue” color in the map). In this case, the most exposed LUMO leads to the prediction that (2R,3S)-3phenylbutan-2-ol is the major product, consistent with the observed results (see Table 7.1). All experimental data for the Cram and Felkin-Anh models in Table 7.1 are taken from the literature,254c and the major diastereomer expected from each substrate (A or B) is predicted using each model. It is clear that the Cram and

275 276

Ref. 253, p. 116.

Hehre, W. J., A Guide to Molecular Mechanics and Quantum Chemical Calculations; Wavefunction: Irvine, CA, 2003, pp. 85–100. Available at http:// www.wavefun.com.

352

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

TABLE 7.1 A Comparison of the Cram and Felkin-Anh Models With LUMO Maps, Against Experimentally Determined Ratios254 of Diastereomers Using Various Nucleophiles O

M X

• S

R R O

S

M

O– M •

• R L

X

Cram

R L

A

OH L R

R

Nucleophile = X

S X M R L OH

X

S M

S

Felkin–Anh

S M S

L A

•

X

L

O L

O–

M

L

X

OH

L

R

X

R

S M

S M

A

B

S X OH M R L

OH

S HO X M R L

In each case, the table shows a prediction of A or B as the major product, where Cram indicates the use of the Cram model shown, Felkin-Anh indicates the use of the Felkin-Anh model shown, and LUMO indicates the prediction is based on the LUMO map generated for each compound using MacSpartan. RS

RM

RL

R

Reagent

H H H H H H H H H H H H

Me Me Me Me Me Me Me Me C7H15 Me Me Me

Ph Ph Ph Ph Et c-Hex c-Hex Cl OBn NBn2 CO2Et OMe

H Me Me SiMe3 H H Me Ph Me H Ph Ph

MeLi LiAlH4 EtMgBr MeLi MeMgBr MeMgI LiAlH4 LiAlH4 BuMgI

H H

Me Me

OTBS OH

Et Me

MeMgI MeLi Me2Mg MeMgCl LiAlH4

A:B (exp)

Cram

80 : 20 74 : 26 75 : 25 98 : 02 60 : 40 82 : 18 62 : 38 75 : 25 95%)

BH3-THF , THF , –15 °C , 2 h

Acetophenone

(R)-1-Phenylethan-1-ol (88%) OC3H7 O

C3H7O O

NiB

OC3H7

Ph

O

O C3H7O

B N

Ph

O

C3H7O

H2 N

H

Ph OH Ph

O OC3H7 O C3H7O O

219

OC3H7

220

7.9.4 Selectivity in the Reduction of Monocyclic Molecules Predictions of stereochemistry for cyclic ketones using the Cram or Felkin-Anh models are generally unreliable. Reduction of 2-methylcyclopentanone,304,305 can be used to illustrate the difficulty. Model 221 is an approximation of the Cram model for 2-methylcyclopentanone, and it predicts cis-2-methylcyclopentan-1-ol to be the major product of hydride reduction. Analysis of cyclic ketones shows that fitting them to the Cram model requires distortion of the ring away from the low-energy conformations (chair or envelope for six- and five-membered rings, respectively). The Felkin-Anh model for 2-methylcyclopentanone (222) has similar problems. Although there is less distortion of the ring, this model also predicts cis-2-methylcyclopentan-1-ol as the product. Reductions of actual molecules show that trans-2methylcyclopentan-1-ol is favored except when very bulky reducing reagents are used. Both NaBH4 and LiAlH4 give a roughly 3:1 ratio of trans- to cis-2-methylcyclopentanols, as does lithium tri(tert-butoxy)aluminum hydride.304 Borane gives a slightly higher amount of the cis alcohol, but the ratio remains close to 3:1.304 L-Selectide yields 98% of cis-2methycyclopetnan-1-ol.304 Me

O

H

Me

O OH

Me

Me

OH

•

•

H 221

cis-2-Methylcyclopentan-1-ol

222

cis-2-Methylcyclopentan-1-ol

It is clear that both the Cram and Felkin-Anh models gave incorrect predictions. A different model must be used to provide an accurate prediction of stereoselectivity. The actual experimental results can be explained by using two envelope conformations available to 2-substituted cyclopentanone models, 223 and 224 (Section 1.5.2). In 223, approach of a reagent to the carbonyl at an angle of 110 degrees (the B€ urgi-Dunitz trajectory, also see Section 10.6)306 leads to an interaction with Ha (path b), making path a preferred and this leads to the trans-product. Remember that this angle assumes a late product-like transition state in which the approach angle approximately mimics the tetrahedral angles of the sp3 hybridized carbon found in the alcohol products. This result is apparently contradicted by 224, where approach via path a is inhibited by the R group. Since the R group will be pseudo-equatorial in the lowest energy conformation (as in 223), this model is used to predict the major diastereomer in reactions of cyclopentanone 304

Ref. 253, p. 119.

305

Also see (a) Umland, J. B.; Jefraim, M. I. J. Am. Chem. Soc. 1956, 78. 2788. (b) Brown, H. C.; Deck, H. R. J. Am. Chem. Soc. 1965, 87, 5620. (c) Ashby, E. C.; Sevenair, J. P.; Dobbs, F. R. J. Org. Chem. 1971, 36, 197. (d) Brown, H. C.; Bigley, D. B. J. Am. Chem. Soc. 1961, 83, 3166. (e) Caro, B.; Boyer, B.; Lamaty, G.; Jaouen, G. Bull. Soc. Chim. Fr. 1983, Pt. 2, 281.

306

B€ urgi, H. B.; Dunitz, J. D.; Lehn, J. M.; Wipff, G. Tetrahedron 1974, 30, 1563.

358

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

derivatives. Indeed, reduction via path a in 223 leads to the correct stereochemical predictions for reduction, except when bulky reducing agents are employed. Structure 225 is a molecular model of 224 (R ¼ Me), showing the steric influence of the group on approach to the acyl carbon. Inspection of LUMO map for 2-methylcyclopentanone shows a more exposed LUMO for the carbonyl carbon on the same face as the methyl group, suggesting a preference for the trans-alcohol, consistent with the observed selectivity of the reduction, although 223 yields an incorrect prediction when bulky reagents (e.g., Selectride) are used. a H R

O

H Ha

b

223

H

H

R

a

Ha b

O 224

225

As the α substituent in 2-substituted cyclopentanones increases in size from methyl to tert-butyl (using 223), path a becomes less favored and lower selectivity for the trans-2-substituted cyclopentanol is observed for reduction with several reducing agents.304 As discussed above for 2-methylcyclopentanone, reduction with most of the hydride reducing agents leads to a 3:1 mixture of trans- to cis-2-methylcyclopentanol. An analysis of 226, the tert-butyl derivative, shows that the methyl groups impose greater hindrance to path a. A LUMO map of 226 also indicates less selectivity. Even with this very bulky tert-butyl group, essentially a 1:1 mixture of stereoisomers is obtained upon reduction. For clarity, it must be emphasized that 223, 224, and 226 are distorted as drawn in a stylized envelope conformation (cf. 224). a H

O

Me Me

H

b

Me H 226

The Cram model is also a poor model for reduction of cyclohexanone derivatives. Antiperiplanar hyperconjugative effects have been invoked to explain the observed results,307 but Cieplak308 put forth an alternative proposal. The incipient bond (the one being formed) was found to be electron deficient. The Cieplak model led to some controversy, and a so-called exterior frontier orbital extension model has been proposed to explain the selectivity.309 The stereoselectivity associated with substituted cyclohexanones can be shown using the reaction of 4-tertbutylcyclohexan-1-one with LiAlH4 or NaBH4, which gave the trans-alcohol [(1r,4r)-4-(tert-butyl)cyclohexan-1-ol] as the major product (92% with LiAlH4 and 80% with NaBH4) via axial attack (path a).310 In general, axial approach to the carbonyl (path a) will lead to the major product. Note that the bulky tert-butyl group at C4 in 4-(tert-butyl)cyclohexan-1-one effectively locks the molecule into that chair conformation (Section 1.5.4). Close analysis of 4-(tert-butyl) cyclohexan-1-one suggests that approach via path b is encumbered by the axial hydrogen atoms on the bottom face of the molecule. The Cram model or the Felkin-Anh model predict path b as the major pathway to yield (1s,4s)-4-(tertbutyl)cyclohexan-1-ol, which is incorrect (see Section 1.4.2 for the lower case s, indicating a pseudochiral center). An important aspect of these models is the trajectory of approach for the hydride to the carbonyl, which will be close to 110 degrees (see 223 and 226).306 This angle of approach will exacerbate the steric problems posed by the axial hydrogen atoms at C2 and C6 for path b and the axial hydrogen atoms at C3 and C5 for path a.

307

(a) Burton, J. W.; Clark, J. S.; Derrer, S.; Stork, T. C.; Bendall, J. G.; Holmes, A. B. J. Am. Chem. Soc. 1997, 119, 7483. (b) Ahn, N. T.; Eisenstein, O.; Lefour, J.-M.; Tran Huu Dau, M. E. J. Am. Chem. Soc. 1976, 95, 6146. (c) Ahn, N. T.; Eisenstein, O. Nouv. J. Chim., 1976, 1, 61.

308

(a) Cieplak, A. S. J. Am. Chem. Soc. 1981, 103, 4540. (b) Ref. 255, pp. 882–883.

309

Tomoda, S.; Senju, T. Tetrahedron 1999, 55, 3871.

310

(a) Richer, J.-C. J. Org. Chem. 1965, 30, 324. (b) Richer, J.-C.; Perrault, G. Can. J. Chem. 1965, 43, 18.

359

7.9 STEREOSELECTIVITY IN REDUCTIONS

a

H

O 1.LiAlH 4

HH b 4-(tert-Butyl)cyclohexan-1-one

OH OH

H

+

2. H3O+

H

H

H

(1r, 4r)-4-(tert-Butyl)cyclohexan-1-ol

(1s, 4s)-4-(tert-Butyl)cyclohexan-1-ol

Structures 227 and 228 are molecular models for 2-methylcyclohexan-1-one, showing the steric impedance imposed by both an axial methyl group (228, blocking path b) and an equatorial methyl group (227). As the steric bulk at C2 is increased, approach from path a is somewhat inhibited, leading to increasing amounts of the cis-product (via path b). It is interesting to note that LiAlH(OMe)3 gives more cis-product than LiAlH(Ot-Bu)3. Generally, smaller reagents give more axial attack (path a) and larger reagents give more equatorial attack (path b). In their monograph, Morrison and Mosher311 arranged the relative size of hydride reagents as follows:

LiAlH4 < LiAlH(Ot-Bu)3 < NaBH4 < KBH4 < LiAlH(OMe)3 < Ip2BH (Section 9.2.1)

a

a

b

b

228

227

For simple 2-substituted cyclohexanone derivatives, model 229 can be used, and when R ¼ Me, 230 is favored over 231 by 2:1 with several hydride reducing agents.305e,c When R ¼ Et, 230 is favored by 1.8:1, but when R ¼ isopropyl 230 is favored by 1.8:1 with LiAlH4, but NaBH4 yields a 1:1 mixture of 230:231.305c When R ¼ tert-butyl, 231 is favored over 230 by 18:1.305c As marked in 229, the reagent will approach the carbonyl carbon from path a or b, using the B€ urgi-Dunitz trajectory. In general, path a is preferred and delivery of hydride from that face leads to 230. Delivery of hydride from path b leads to 231. When the α-carbon in 229 is substituted with relatively small groups, path a is preferred to path b, but when the α-carbon has relatively large groups, approach from path a or path b is relatively equal. Analysis of the intensities of the carbonyl carbon using a LUMO map of 2-methylcyclohexan-1-one also indicates that there is a preference for path a. a

H

O

R b H 229

OH OH

Reduction

H

+

R H 230

R H 231

Substituents at C3 or C5 also have a significant effect on the approach of reagents to the carbonyl of cyclohexanone derivatives. Using a model similar to the 2-substituted cyclohexan-1-ones, reduction of 3-methylcyclohexan-1-one can proceed via path a or path b (see 232). However, there are two equilibrating chair conformations (see Sections 1.5.2 and 1.5.3), 232 and 233. Since the axial methyl group imposes steric hindrance to path a, there is a preference for path b. It is cautioned that the lower energy chair conformation of 3-methylcyclohexanone is probably 233 rather than 232. Attack

311

Ref. 253, pp. 123–124 and Refs. 38, 55, and 56 cited therein.

360

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

via path b in 232 will give the cis-product, but note that attack along the preferred path a in 233 also yields the cisproduct. If 232 is used as the standard model, then path b is preferred. a

Me

O

Me

a

O

H

H

H b

b

Me

O 233

232

Substitution at C3 is a bit different because an axial C substituent effectively blocks approach of the reducing agent via path a. As shown for 3,3,5(R)-trimethylcyclohexan-1-one,305c,312 an axial methyl group at C3 of a cyclohexanone blocks path a [to yield (1R,5R)-3,3,5-trimethylcyclohexan-1-ol] and the major product is via path b to yield (1S,5R)-3,3,5trimethylcyclohexan-1-ol. The preferred mode of attack is path b to give the trans-alcohol, (1R,5R)-3,3,5trimethylcyclohexan-1-ol. Lithium aluminum anhydride gave roughly a 1:1 mixture, but as the size of the hydride reagent increases, path b is more highly favored to yield 3:1–4:1 mixture favoring (1S,5R)-3,3,5trimethylcyclohexan-1-ol.312 a Me

Me Me

H

b

3, 3, 5(R)-Trimethylcyclohexan-1-one

Me

Me

Reduction

H

H

OH

O

Me

+

H Me

Me

H OH Me

(1S, 5R)-3, 3, 5-Trimethylcyclohexan-1-ol

(1R, 5R)-3, 3, 5-Trimethylcyclohexan-1-ol

De Maio et al.313 studied the stereochemistry and relative rates of kax and keq for acyl addition reactions of 5-substituted adamantan-2-ones. This study has relevance to the reactions of the cyclohexanone derivative due to the presence of the chair conformations for the six-membered rings (see 5-methyladamantan-2-one). This kinetic data did not fit current theories of π-face diastereoselection. Their studies suggest that kax/keq measurements do not adequately indicate what is happening on the two sides of a stereogenic center. In addition, they found that axial reactivity always increases with increasing electronegativity of the substituted at C5 and that keq changes depend on the conformation of that group and on reaction conditions. In their words: “Simple theories, such as those based on dipole-dipole interactions, are unable to explain such behavior, nor can theories based on ground state MO calculations.”313

H3C

O

5-Methyladamantan-2-one

7.9.5 Stereoselectivity in the 1,2-Reduction of Cyclohexenone Derivatives The conformational problems of cyclohexene derivatives are somewhat different than those observed with cyclohexanone or cyclohexane derivatives. The presence of the alkene moiety effectively flattens four carbons of the ring, making the steric interactions of substituents with an incoming reagent different, relative to cyclohexanone derivatives. The most abundant conformations of cyclohexenone derivatives are the “half-chair” conformations 234 and 235. The first problem that arises in conjugated derivatives is the competition of 1,2- with 1,4-addition for many reducing agents, discussed in previous sections. In this section, only products arising from 1,2-reduction will be discussed.

312

Haubenstock, H.; Eliel, E. L. J. Am. Chem. Soc. 1962, 84, 2363, 2368.

313

Di Maio, G.; Solito, G.; Varì, M. R.; Vecchi, E. Tetrahedron 2000, 56, 7237.

361

7.9 STEREOSELECTIVITY IN REDUCTIONS

It appears that the C2 group (R1) offers little interference to either path a or b unless it is very large. The R2 group, however, can inhibit approach by path a if sufficiently large and a large R4 may inhibit approach by path b. R2 a H

H

R3

H

O H

H

R1 R4

O R2 R1 b R3

H b

234

a

R4

235 Et H

H a

HO

Et 6

5

4

O 1

2

3

236

O O

O O

H

237 (70%, 1 : 2.4)

NaBH4 , MeOH CeCl3•7 H2O

+

Et H

HO O O

H 238

The sodium borohydride-cerium chloride reagent (Luche reagent) gave selective 1,2-reduction of conjugated ketones in Section 7.5. The sensitivity of the cyclohexenyl system to steric blocking when substituents are present at C5 or C6 is shown by reduction of 236, which yielded a 2:1 mixture of 238 and 237.314 This preference illustrates that the steric encumbrance is less important in 237 than in cyclohexanone derivatives (see 232). A molecular model of 236 is provided to give a more realistic view of the stereochemical relationships.

7.9.6 Selectivity in the Reduction of Bicyclic and Polycyclic Derivatives As with monocyclic ketones, bicyclic ketones show excellent diastereoselectivity upon reduction with hydride reagents. This selectivity is clearly observed in the reduction of bicyclo[2.2.1]heptan-2-one (norbornanone) and 7,7-dimethylbicyclo[2.2.1]heptan-2-one. The exo-face (path a in 239) is usually preferred, leading to the endo-alcohol, bicyclo[2.2.1]heptan-2-ol. The endo hydrogen atoms (Ha in bicyclo[2.2.1]heptan-2-one) provide just enough steric encumbrance that path a is more accessible. This statement is supported by a LUMO map of bicyclo[2.2.1]heptan2-one, which is consistent with the observed exo-attack. Approach of a reagent to the carbonyl at an angle of 110 degrees brings the reagent into close proximity to this hydrogen. The exo-face appears somewhat hindered, but is actually less hindered than the endo-face. When a methyl group blocks the exo-face, as in 7,7-dimethylbicyclo [2.2.1]heptan-2-one (see path a and model 240), attack by path b is preferred and the exo-alcohol (7,7-dimethylbicyclo[2.2.1]heptan-2-ol) is the major product. Analysis of a LUMO map of 7,7-dimethylbicyclo[2.2.1]heptan-2-one reveals that path b is preferred. In all [2.2.1]-, [2.1.1]-, and [2.2.2]-bicyclic ketones of this type, the major product is predicted by the extent of steric hindrance at the exo-face (exo-attack with no substituents on the bridge, as in bicyclo[2.2.1]heptan-2one).305c,d,315 When the bridge has substituents (7,7-dimethylbicyclo[2.2.1]heptan-2-one), endo-attack is generally preferred.

314

Crimmins, M. T.; O’Mahoney, R. J. Org. Chem. 1989, 54, 1157.

315

Wheeler, O. H.; Mateos, J. L. Can. J. Chem. 1958, 36, 1431.

362

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

a

Me Hydride

H

H Ha

O

Me

Me

a Hydride

H

Reduction

Ha

Me

Reduction

OH

OH O

b

H

b

Bicyclo[2.2.1]heptan-2-one

7,7-Dimethylbicyclo[2.2.1]heptan-2-one

Bicyclo[2.2.1]heptan-2-ol

239

7,7-Dimethylbicyclo[2.2.1]heptan-2-ol

240

When “edge-fused” bicyclic systems (bicyclo[m.n.0] compounds) are reduced with NaBH4 or LiAlH4,305e approach to the carbonyl-bearing ring is very similar to that observed with monocyclic ketones. Reduction of decalone, indanone, and 241, for example, occurred primarily via axial attack (path a), similar to cyclohexanones.305e An axial alcohol results from equatorial attack by the hydride and the equatorial alcohol results from axial attack by the hydride. The presence of the bridging carbon inhibits path a somewhat in 242 and to a greater extent in 243.305e a

a

a

O

b

b

Decalone

a

a

O

O

O

b

Indanone

O

b

241

b

242

243

There is also a preference for path a in complex compounds (e.g., 3-cholestanone). Coprostanone is an isomer of 3-cholestanone and also gives axial attack as shown (via path a) although the conformation of the molecule is slightly different.316 The major diastereomeric product in bicyclic and polycyclic systems can be predicted with little or no modification of the simple cyclohexanone and cyclopentanone models presented above.

H

H

Me

a

H H

Me

O

H

H H

C8 H17

H

Me

Me

C8H17

3-Cholestanone

a O

H Coprostanone

7.9.7 Selectivity in the Reduction of Complex Molecules There are many synthetic examples where hydride reduction proceeds with high diastereoselectivity. Reduction of 244A to alcohol 245 with lithium tri-tert-butoxyaluminum hydride is taken from Donaldson and Greer’s317 synthesis of the C3–C15 bis(oxane) segment of phorboxazole. The conformational drawing (244B) suggests that path a is preferred in the more stable chair conformation, with the two substituents in the equatorial position. This model predicts the major product, which was isolated by conversion to the corresponding acetate.

316

Ref. 253, pp. 127–129 and references cited therein.

317

Greer, P. B.; Donaldson, W. A. Tetrahedron Lett. 2000, 41, 3801.

363

7.9 STEREOSELECTIVITY IN REDUCTIONS

1. LiAlH(Ot-Bu)3

PO

a

OMe

OMe O O

PO

O

PO

2. Ac2O/NaOAc

H O OAc

P = SiPh2t-Bu

O H O

Me 244A

245

b

244B

Another example involves a selective reduction of an acyclic ketone, albeit one that has a proximal dioxolane ring. Reduction of 246A gave 247 in 96% yield with L-Selectride, in Prasad and Gholap’s318 synthesis of (+)-7-epi-goniofufurone. The molecular model (246B) shows the large substituent is largely out-of-the way for path a or b. The conformation of the seven-membered ring, however, leads to some hindrance for path a, making path b preferred. This result is consistent with the stereochemistry of the final product, syn-diastereomer 247. Note that the substituent is positioned in such a way that a “pocket” is available for delivery via path b. a

t-BuMe2SiO

t-BuMe2SiO –78 °C

OH

O

Me O

O

L-Selectride THF

O

246A

O

247 (96%)

246B

Reduction of norbornane-like precursors to natural products is also predictable. Reduction of the ketone unit in 248, in Money and coworker’s319 synthesis of β-santalene for example, gave the exo-alcohol 249 via attack from the less hindered endo face. Conversely, removal of the steric hindrance at C7 in 250 gave attack via the exo-face to yield 251, in about a 4:1 ratio favoring the endo alcohol shown.320 Me

Me

Me 1. LiAlH(Ot-Bu)3

O

Me

Me

HO

Me

2. H3O+

H 248

249 H

H 1. LiAlH4

Ph O 250

2. H3O+

H

Ph

OH 251

318

Prasad, K. R.; Gholap, S. L. J. Org. Chem. 2007, 72, 2.

319

Hodgson, G. L.; MacSweeney, D. F.; Mills, R. W.; Money, T. J. Chem. Soc. Chem. Commun. 1973, 235.

320

(a) Benjamin, B. M.; Collins, C. J. J. Am. Chem. Soc. 1966, 88, 1556. (b) Kleinfelter, D. C.; Dye, T. E. Ibid. 1966, 88, 3174. (c) Kleinfelter, D. C.; Dye, T. E.; Mallory, J. E.; Trent, E. S. J. Org. Chem. 1967, 32, 1734.

364

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

There are many more examples of selectivity in reduction. These few illustrate that the simple principles discussed in this section, and the models used, can be applied to more complex systems with reasonable success.

7.10 CATALYTIC HYDROGENATION Alkenes do not react with hydrogen gas in the absence of any other additive. Since the hydrogen atoms in diatomic H2 are not polarized, they are not very reactive. If the HdH bond were ruptured to produce hydrogen atoms, however, reactivity would be greatly increased and alkenes could react. Indeed, when transition metal catalysts are mixed with hydrogen gas, the transition metal first reacts with the hydrogen gas to produce MdH units, which then react with an alkene. In this reaction, the π-bond is broken and a hydrogen atom is transferred to each sp2 hybridized carbon atom, forming new CdH bonds, where each carbon is sp3 hybridized. In other words, an alkene is reduced to an alkane. The catalyst is often mixed with an inert compound (carbon black, barium carbonate, calcium carbonate, etc.). Reduction of functional groups with hydrogen gas is a key reaction in organic chemistry, dating to the first hydrogenation of ethene to ethane by von Wilde in 1874.321 As stated by Hudlický,6 “true widespread use of catalytic hydrogenation did not start until 1897 when Sabatier and senderens322 and his coworker developed the reaction between hydrogen and organic compounds to be a universal reduction method (Nobel Prize, 1912).”323 Other monographs that describe synthetic applications of catalytic hydrogenation are available from Rylander324 and from Freifelder.325 Many transition metals or transition metal salts can be used as catalysts, including platinum black,326 PtC,326,327 platinum oxide,328 PdBaCO3,329 PdC,329 RhC,329 Pd hydroxideC,330 and nickel [one catalyst is Ni(R), which is Raney Ni].329,331 Applications of catalytic hydrogenation to organic chemistry are found in the series Catalysis in Organic Reactions.332 Catalytic hydrogenation is used to reduce alkenes,333 alkynes,334 ketones and aldehydes,335 nitriles,336 nitro compounds337, or aromatic rings.338

7.10.1 Catalytic Activity and Reactivity As mentioned, in the absence of a transition metal, alkenes or any other functional group do not react with hydrogen gas. When hydrogen gas is added to an organic molecule in the presence of a catalytic amount of a transition metal, the reaction proceeds by adsorption of both hydrogen and the substrate on the surface of the metal. The mechanistic 321

von Wilde, M. P. Deutsch. Chem. Ber. 1874, 7, 352.

322

Sabatier, P.; Senderens, J. B. Compt. Rend. 1897, 124, 1358.

323

Ref. 6, p. 3.

324

Rylander, P. N., Catalytic Hydrogenation in Organic Synthesis; Academic Press: New York, 1979.

325

Freifelder, M., Catalytic Hydrogenation in Organic Synthesis: Procedures and Commentary; Wiley: New York, 1978.

326

(a) Willst€ atter, R.; Waldschmidt-Leitz, E. Berichte 1921, 54, 113 (see p. 121). (b) Baltzly, R. J. Am. Chem. Soc. 1952, 74, 4586. (c) Theilacker, W.; Dr€ ossler, H. G. Chem. Ber. 1954, 87, 1676.

327

Baltzly, R. J. Org. Chem. 1976, 41, 920.

328

(a) Adams, R.; Shriner, R. L. J. Am. Chem. Soc. 1923, 45, 2171. (b) Carothers, W. H.; Adams, R. Ibid. 1923, 45, 1071. (c) Keenan, C. W.; Giesemann, B. W.; Smith, H. A. Ibid. 1954, 76, 229. (d) Frampton, V. L.; Edwards, J. D., Jr.; Henze, H. R. Ibid. 1951, 73, 4432.

329

Mozingo, R. Org. Synth. Collect. 1955, 3, 181.

330

Pearlman, W. M. Tetrahedron Lett. 1967, 1663.

331

(a) Pavlic, A. A.; Adkins, H. J. Am. Chem. Soc. 1946, 68, 1471. (b) Mozingo, R. Org. Synth. Coll. 1955, 3, 181.

332

(a) Moser, W. R., Ed. Catalysis of Organic Reactions; Marcel–Dekker: New York, 1981. (b) Delannay., F, Ed. Characterization of Heterogeneous Catalysts; Marcel–Dekker: New York, 1984. (c) Kosak, J. R., Ed. Catalysis of Organic Reactions: Marcel–Dekker: New York, 1984. (d) Augustine, R. L., Ed. Catalysis of Organic Reactions: Marcel–Dekker: New York, 1985. (e) Rylander, P. N.; Greenfield, H.; Augustine, R. L., Ed. Catalysis of Organic Reactions: Marcel Dekker: New York, 1988.

333

See Ref. 82, p. 7.

334

See Ref. 82, p. 28.

335

See Ref. 82, p. 1075.

336

See Ref. 82, p. 875.

337

See Ref. 82, p. 821.

338

See Ref. 82, p. 6.

7.10 CATALYTIC HYDROGENATION

365

rationale for this type of reduction will be discussed in Section 7.10.2. There are two major types of catalytic hydrogenation: heterogeneous, where the catalyst is insoluble in the reaction medium, and homogeneous, where the catalyst is soluble in the reaction medium. Catalytic heterogeneous hydrogenation can be supported (slurry and fixed gel operations)324 or unsupported (primarily solution reactions). The nature and amount of the catalyst and hydrogenation procedure employed influence which functional group is reduced, the extent of reduction, and the product distribution. Laboratory scale reductions of organic functional groups containing a π-bond usually involve a slurry process, which is heterogeneous catalysis. Industrial (kilogram or ton scale) applications often use fixed-bed heterogeneous catalysis, but homogeneous catalysis is also important. Birch and Williamson339 reviewed homogeneous hydrogenation in synthesis. The most commonly used heterogeneous catalysts are Pt, Pd, Ni, Rh, and Ru. Rylander gave references340 for the preparation of the most common catalysts. In some cases, salts of transition metals are used rather than the metal itself, although the pure metal adsorbed on a support (see above) is also commonly used. Hudlický341 presented an order of relative reactivity with propene for the Group 8–10 transition metal catalysts, based on the work of Mann and Lien.342 Rh > Ir > Ru > Pt > Pd > Ni > Fe > Co > Os 14:0 15:0 6:5 16:0 11:0 14:0 10:0 8:1 7:4 In this series, Rh is the most active catalyst. For the reaction of hydrogen with propene, it is noteworthy that the catalytic activity does not necessarily correlate with the activation energy for the hydrogenation reaction (numbers below the metal). Hudlický343 correlated several catalysts [5% Pd/C, PtO2; Ni(R); Pd/CaCO3; Pd/BaSO4; Lindlar catalyst; Rh/Al2O3; Rh/C] with the hydrogenation of many functional groups (alkenes, alkynes, carboxylic acids, aromatic compounds, aldehydes and ketones, halides, nitro compounds, azides, oximes, and nitriles). The reactivity of these catalysts will be elaborated in Section 7.10.2. No single catalyst gives excellent results for all functional groups, and there are significant differences in chemo- and stereoselectivity. Note that most hydrogenations are done at ambient temperature and pressure. In most cases, only a few catalysts require high pressures, and then only for selected functional groups. For the most part, 60% yield, an intermediate in Panek and Zhu’s602 convergent synthesis of (+)-SCH 351448.

O

O

O O

O

O

O O

p-TsNHNH2 , NaOAc DME , H2O , 100°C DME = Dimethoxyethane

OTBS O

OTBS O

OBn

OBn

BnO2C

BnO2C

404

403

(>60%)

Corey et al.603 showed that the course of a reduction with diimide could be altered by addition of cupric salts [e.g., cupric acetate, Cu(OAc)2]. This new reagent is highly selective for the reduction of conjugated alkenes, with specificity for the less substituted double bond of the diene.604 The functional group transform follows: R

R

R

R

7.12.3 Reduction With Silanes (Hydrosilylation) Alkyl silanes can be used for the reduction of carbonyls and alkenes. Methylcyclohexene was reduced to methylcyclohexane using a mixture of triethylsilane (Et3SiH) and trifluoroacetic acid (CF3CO2H), in 72% yield.605 Under the

596

Adam, W.; Eggelte, H. J. J. Org. Chem. 1977, 42, 3987.

597

(a) Ref. 6, pp. 33–34. (b) Ref. 27, pp. 257–258.

598

For an example taken from the synthesis of aplyronines, see Marshall, J. A.; Johns, B. A. J. Org. Chem. 2000, 65, 1501.

599

See Ref. 82, p. 12.

600

(a) H€ unig, S.; M€ uller, H.-R.; Thier, W. Tetrahedron Lett. 1961, 353. (b) van Tamelen, E. E.; Timmons, R. J. J. Am. Chem. Soc. 1962, 84, 1067.

601

van Tamelen, E. E.; Dewey, R. S.; Lease, M. F.; Pirkle, W. H. J. Am. Chem. Soc. 1961, 83, 4302.

602

Zhu, K.; Panek, J. S. Org. Lett. 2011, 13, 4652.

603

(a) Corey, E. J.; Mock, W. L.; Pasto, D. J. Tetrahedron Lett. 1961, 347. (b) Corey, E. J.; Hortmann, A. G. J. Am. Chem. Soc. 1965, 87, 5736.

604

For an example, see Corey, E. J.; Yamamoto, H. J. Am. Chem. Soc. 1970, 92, 6636.

605

Kursanov, D. N.; Parnes, Z. N.; Bassova, G. I.; Loim, N. M.; Zdanovich, V. I. Tetrahedron 1967, 23, 2235.

407

7.12 NONMETALLIC REDUCING AGENTS

same conditions, however, pent-1-ene was not reduced. This reagent is often used for the reduction of conjugated carbonyls, probably via formation of a silyl enolate (Sections 13.3.2 and 13.3.3 and see also Section 13.4.3), as in the reduction of cyclohexenone to cyclohexanone in 85% yield with Ph2SiH2.606 Transition metals (e.g., ZnCl2, Rh derivatives or Cu salts) can be added to the silane to facilitate the reduction.607 An example, taken from a synthesis of (+)-manoalide by Kocienski and coworkers, reduced the conjugated double bond in β-ionone to give 405 in 90% yield using the twostep procedure shown.608 Conjugated aldehydes tend to give 1,2-reduction, as in the reduction of 3-phenylprop-(2E)enal to 3-phenylprop-(2E)-en-1-ol in 95% yield.609 Another example of such selectivity is the reduction of the ketone unit in 2-bromo-1-phenylpropanone to give 2-bromo-1-phenylpropan-1-ol in 70% yield.562 Nonconjugated ketones and aldehydes are reduced by silanes in acid media, as in the reaction of cyclohexanone with Et3SiH and trifluoroacetic acid to give cyclohexanol in 74% yield.605 Ketones are usually reduced faster than epoxides.610,611 The use of a chiral additive leads to asymmetric reduction with silanes. Reduction of acetophenone gave (R)-phenethyl alcohol in 99% yield and 84.2 %ee, in the presence of 406.612 1. 1.5 Et3SiH , 80 °C 1% ClRh(PPh3

O

O

2. HF , aq MeCN , rt

405 (90%)

-Ionone

O

N

+

Ph

N

Me

H

Ph

HO

1. H2SiPh2 2. H2O

Me

H

Ph

Me

406

Reductive cleavage of carbonyls or alcohols, analogous to the Wolff-Kishner or Clemmensen reduction, is illustrated by conversion of cyclohexanone to cyclohexane in 90% yield,613 using a silane•BF3 complex. Other modifications (e.g., addition of cesium fluoride), give reduction of an ester moiety.614 The functional group transforms observed in this section follow: R

R

R

R

R

R

R

R OH

O R

O R

H H

R

O R

R

OH R

O R

O R

7.12.4 Reduction With Formic Acid Formic acid has been used for several rather specific reductions. Heating triphenylcarbinol in formic acid, for example, led to triphenylmethane.615 The most common application of formic acid reduction involves enamines, presumably via conversion to an iminium salt, which is the actual species reduced by formic acid.616

606

Sharf, V. Z.; Freidlin, L. Kh.; Shekoyan, I. S.; Krutii, V. N. Bull. Akad. USSR Chem. 1977, 26, 995.

607

Ojima, I.; Kogure, T. Organometallics 1982, 1, 1390.

608

Pommier, A.; Stepanenko, V.; Jarowicki, K.; Kocienski, P. J. J. Org. Chem. 2003, 68, 4008.

609

Boyer, J.; Corriu, R. J. P.; Perz, R.; R
ey
e, C. J. Chem. Soc. Chem. Commun. 1981, 121.

610

Mullholland, R. L., Jr.; Chamberlin, A. R. J. Org. Chem. 1988, 53, 1082.

611

Salomon, R. G.; Sachinvala, N. D.; Raychaudhuri, S. R.; Miller, D. B. J. Am. Chem. Soc. 1984, 106, 2211.

612

Brunner, H.; Riepl, G.; Weitzer, H. Angew. Chem. Int. Ed. 1983, 22, 331.

613

Fry, J. L.; Orfanopoulos, M.; Adlington, M. G.; Dittman, W. R., Jr.; Silverman, S. B. J. Org. Chem. 1978, 43, 374.

614

Boyer, J.; Corriu, R. J. P.; Perz, R.; Poirer, M.; R
ey
e, C. Synthesis 1981, 558.

615

(a) Kaufmann, H.; Pannwitz, P. Berichte 1912, 45, 766. (b) Kovache, A. Ann. Chim. 1918, 10, 184.

616

Paukstelis, J. V.; Kuehne, M. E. In Enamines: Synthesis Structure and Reactions; Cook, A. G., Ed.; Marcel–Dekker: New York, 1969, pp. 169–210 and 313–468.

408

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

Ph Ph C Ph

Ph Ph C Ph

HCO2H

OH

Heat

Triphenylmethanol

H

Triphenylmethane

Reduction of iminium salts proceeds by hydride transfer, with formation of carbon dioxide.617 Reduction of 3-()-erythro-2-(N-methyl)pyrrolidiniminiumbutanoic acid (407) to 409 illustrates that the reaction can be stereoselective. The hydride-transfer process is represented by 408, and the stereochemistry of 409 is predicted by analysis with Cram’s rule (Section 7.9.1).618 The hydride approaches from the more sterically accessible face. A proton-catalyzed tautomerization converts enamines to iminium salts, which are then reduced to the corresponding amine.619 Reduction of iminium salts is apparently responsible for the classical reductive methylation procedure of amines and formaldehyde in the presence of formic acid. The methylation protocol is illustrated by the N-methylation of the secondary amine moiety in 410 with formic acid and formaldehyde, to give a 99% yield of 411 in Tran and Kwon’s620 synthesis of alstonerine. CO2H CO2H HCO2H

N

N Me

Me H

H Me

407

H

– H+

H Me H

N

– CO2

Me O

H Me

O

408

409

H

CH3

N

O

CO2H

CO2Et

35% aq HCHO 88% aq HCOOH

N

O

CO2Et

Reflux

N

N Me

410

Me 411 (99%)

Several modifications of this reduction involve formates and are effective for reduction of various functional groups. A mixture of formic acid and ethyl magnesium bromide was used to reduce decanal to decanol in 70% yield.621 Decanal was also reduced to decanol in 69% yield by using sodium formate in N-methyl-2-pyrrolidinone as a solvent.622 Functional groups other than carbonyl derivatives can be reduced under relatively mild conditions with formate derivatives. Ammonium formate, in the presence of Pd/C was used to reduce an azide to a primary amine.623 Aliphatic nitro compounds are also converted to an amine with this reagent.624

617

(a) Lukeš, R.; Jizba, J. Chem. Listy 1953, 47, 1366 (Chem. Abstr. 1955, 49, 323g). (b) Idem Collect. Czech. Chem. Commun. 1954, 19, 941, 930. (c) Leonard, N. J.; Sauers, R. R. J. Am. Chem. Soc. 1957, 79, 6210. 618  Cervinka, O. Collect. Czech. Chem. Commun. 1959, 24, 1880. 619

(a) Leonard, N. J.; Thomas, P. D.; Gash, V. W. J. Am. Chem. Soc. 1955, 77, 1552. (b) Bonnett, R.; Clark, V. M.; Giddey, A.; Todd, A. J. Chem. Soc. 1959, 2087.

620

Tran, Y. S.; Kwon, O. Org. Lett. 2005, 7, 4289.

621

Babler, J. H.; Invergo, B. J. Tetrahedron Lett. 1981, 22, 621.

622

Babler, J. H.; Sarussi, S. J. J. Org. Chem. 1981, 46, 3367.

623

Gartiser, T.; Selve, C.; Delpeuch, J.-J. Tetrahedron Lett. 1983, 24, 1609.

624

(a) Ram, S.; Ehrenkaufer, R. E. Tetrahedron Lett. 1984, 25, 3415. For the use of this reaction in a synthesis of 8,14-dihydromorphinandienone Aalkaloids, see (b) Ghavimi, B.; Magnus, P. Org. Lett. 2014, 16, 1708.

409

7.12 NONMETALLIC REDUCING AGENTS

7.12.5 Photoreduction Electron transfer to certain functional groups can be induced by irradiation with light. When combined with a sensitizer and a hydrogen-atom donor, reductions are possible. The basics of photochemical techniques and the use of sensitizers are discussed in Sections 15.2.2 and 17.4. In the presence of a hydrogen-atom donor (usually an alcohol or an amine), carbonyl derivatives or halides can be photochemically reduced.625 Photoreduction is often less stereoselective than other techniques. Both metal hydride reduction and catalytic hydrogenation of 412 gave exclusively the endo-alcohol 414, whereas dissolving metal reduction gave 47% of the exo-isomer 413.626 Photoreduction of 412 gave a 54% yield of a 4:1 mixture of 413 and 414.626

hv

OH

O 412

+

H OH

H 414

413

Competing photochemical pathways can be a problem when using photoreduction (also see Section 17.4), including coupling, isomerization, or migration of multiple bonds. Functional groups with larger quantum yields will react preferentially, but this preference is very dependent on the wavelength of the light used for irradiation (Section 15.2.2). Photoreduction of imines has been used in the synthesis of various alkaloids. Irradiation of Δ1,9-octahydroquinoline led to a 98% yield of trans-decahydroquinoline. Hydrogen is transferred to the nitrogen (propan-2-ol is the hydrogentransfer agent), leading to the thermodynamically more stable radical (415).627 A second hydrogen transfer delivers the hydrogen to the bottom face of 415 to yield trans-decahydroquinoline. H

H

OH

hv, 4 h

N

OH

N 9-Octahydroquinoline

H 415

H

N

H trans-Decahydroquinoline (98%)

7.12.6 Enzymatic Reductions Enzymes, and biocatalysts in general, are increasingly available and important tools for the reduction of organic compounds.628 Enzymatic reductions are often straightforward and highly stereoselective. There are now many enzymatic transformations that are compatible with the use of organic solvents,629 and the alcohol dehydrogenase obtained from Geotrichum candidum is active in supercritical CO2.630 Prelog631 studied the reduction of ketones with several enzymatic systems. Reduction of ketones with Curvularia fulcata, for example, gave predictable stereochemical induction based on identifying Large (L) and Small (S) groups around the carbonyl (this method is sometimes called Prelog’s rule). If the steric difference between (L) and (S) in 416 is large enough, the enzyme delivers hydrogen from the less hindered face (over S in 416) to yield 417. Much of the initial work in this area was done with two enzymatic systems, yeast alcohol dehydrogenase and horse liver alcohol dehydrogenase. This former reagent is usually obtained from yeast and is a protein with a molecular weight of 80,000, containing four subunits. The exact structure of the enzyme is not always correlated with that portion of the structure required for reduction (the active site). Such ambiguities of structure make exact mechanistic arguments difficult. An important part of the selectivity observed with these enzymes, however, is “determined by nonbonded interactions of substrate and enzyme in the hydrogen transfer-

625

Ito, Y.; Kawatsuki, N.; Matsuura, T. Tetrahedron Lett. 1984, 25, 2801.

626

Momose, T.; Muraoka, O.; Masuda, K. Chem. Pharm. Bull. 1984, 32, 3730.

627

Hornback, J. M.; Proehl, G. S.; Starner, I. J. J. Org. Chem. 1975, 40, 1077.

628

Nakamura, K.; Yamanaka, R.; Matsuda, T.; Harada, T. Tetrahedron Asymmetry 2003, 14, 2659.

629

Carrea, G.; Riva, S. Angew. Chem. Int. Ed., 2000, 39, 2226.

630

Matsuda, T.; Harada, T.; Nakamura, K. Chem. Commun. 2000, 1367.

631

Prelog, V. Pure Appl. Chem. 1964, 9, 119.

410

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

transition state.”631 Prelog argued that compounds capable of assuming a diamond lattice (see Section 1.5.2) gave higher selectivity. O

HO S

L

H

Enzyme

S

L

416

417

A common and often used enzymatic reagent is baker’s yeast (Saccharomyces cerevisiae), which gives selective reduction of β-keto esters and β-diketones. A dependence on the size of the ester group was discovered in the Sih and coworkers synthesis of L-carnitine,632 Reduction of ethyl acetoacetate (ethyl 3-oxobutanoate) with baker’s yeast gave the (S)-alcohol [ethyl (S)-3-hydroxybutanoate], but reduction of ethyl β-ketovalerate (ethyl 3-oxopentanoate) gave the (R)-alcohol [ethyl (R)-3-hydroxypentanoate].632 Sih and coauthors632 showed the selectivity of reduction changed from (S)-selectivity with small chain esters to (R)-selectivity with long-chain esters. The break point for selectivity occurred at a chain length of four to five carbons [RC(]O)dO(CHn)H, n ¼ 4, 5]. Baker’s yeast has been used in many synthetic applications. Ethyl (S)-3-hydroxybutanoate, for example, was used in Mori’s synthesis of (S)-(+)-sulcatol (6-methylhept-5-en-2-ol), an aggregation pheromone.633 Similar, and highly selective reduction of β-keto esters was observed with Rhizopus arrhizus mediated reactions.634 Carbonyl units in more complex systems can be reduced, as in the reduction of 418–419 (63% yield with 98 %ee) taken from the synthesis of the unnatural enantiomer of ()-bruceolline J by Gribble and coworkers.635 Baker’s yeast reduction was also used for the preparation of ()-ethyl (R)-2-hydroxyl-4phenylbutyrate, an important intermediate for the synthesis of angiotensin converting enzyme inhibitors.636 O

OH

Baker's yeast

CO2Et

CO2Et

Ethyl 3-oxobutanoate O

Ethyl (S)-3-hydroxybutanoate (67%) Baker's yeast

OH

CO2Et

CO2Et

Ethyl 3-oxopentanoate

Ethyl (R)-3-hydroxypentanoate

O

O O

OH

Baker's yeast , Sucrose H2O , rt , 14 d

N

N

H

H

418

419 (63%, 98 %ee)

1,3-Diketones are enzymatically reduced to the β-keto alcohol. Reduction of tert-butyl 6-chloro-3,5-dioxohexanoate with baker’s yeast, for example, gave tert-butyl (R)-6-chloro-5-hydroxy-3-oxohexanoate, in 50% yield, and 94 %ee, in a synthesis of ()-callystatin A by Enders and coworkers.637 Highly enantioselective reduction of prochiral ketones was also observed when Daucus carota root (carrot root) was used.638 Cultured cells of Marchantia polymorpha have a reductase that leads to the asymmetric reduction of 2-substituted 2-butenolides to the (R)-butenolide.639

632

Zhou, B.; Gopalan, A. S.; Van Middlesworth, F.; Shieh, W.-R.; Sih, C. J. J. Am. Chem. Soc. 1983, 105, 5925.

633

Mori, K. Tetrahedron 1981, 37, 1341.

634

Salvi, N. A.; Chattopadhyay, S. Tetrahedron Asymmetry 2004, 15, 3397.

635

Lopchuk, J. M.; Green, I. L.; Badenock, J. C.; Gribble, G. W. Org. Lett. 2013, 15, 4485.

636

Fadnavis, N. W.; Radhika, K. R. Tetrahedron Asymmetry 2004, 15, 3443.

637

Vicario, J.; Job, A.; Wolberg, M.; M€ uller, M.; Enders, D. Org. Lett. 2002, 4, 1023.

638

Yadav, J. S.; Nanda, S.; Reddy, P. T.; Rao,.B. J. Org. Chem. 2002, 67, 3900.

639

Shimoda, K.; Kubota, N. Tetrahedron Asymmetry 2004, 15, 3827.

411

7.12 NONMETALLIC REDUCING AGENTS

O

O

Baker's yeast

Cl

OH

CO2t-Bu

O

Cl

tert-Butyl 6-chloro-3,5-dioxohexanoate

CO2t-Bu

tert-Butyl (R)-6-chloro-5-hydroxy-3-oxohexanoate

Isolated ketone units can also be reduced with baker’s yeast.640 In a synthetic example, ketone moiety of the piperidone unit in 420 was reduced with baker’s yeast to give a 40% yield of 421 (98 %ee) in the Takeuchi et al.641 synthesis of isofebrifugine O

HO

Baker's yeast , Sucrose EtOH/H2O , rt , 1 d

N

N

CO2CH2Ph

CO2CH2Ph

420

421 (40%)

Problems were encountered in the reduction of aryl β-keto esters (e.g., the furan derivative 422),642 where treatment with baker’s yeast did not give reduction of the ketone moiety. Oishi and coworkers642 examined other selected yeasts including Kloeckera saturnus, which converted 422 to a 47:53 mixture of the syn- (423) and anti-(424) diastereomers, in 47% yield and 87 %ee and 53 %ee, respectively.642 Another yeast (Saccharomyces delbrueckii) gave 43% reduction with a syn/anti ratio of 39:61, and 99 %ee of 423 and 94 %ee of 424.642 Me

Me

Me

K. saturnus

CO2Me

O

CO2Me

O

O 422

+

OH 423

CO2Me

O OH 424

Oishi and coworkers643 examined the selectivity for reduction of 422 with a variety of organisms. In all cases, the organism produced an excess of either the syn- or the anti-diastereomer. The chemical yields were poor in some cases, and the reduction proceeded with modest-to-good diastereoselectivity in other cases. The enantioselectivity of the reaction was good to excellent, with 60–99 %ee observed for both the syn and the anti diastereomers produced. In other work, Oishi and coworkers643a found that selected yeasts showed different diastereoselectivity and enantioselectivity when the aryl group was changed from phenyl to thiophene. Other organisms have been used to reduce β-keto esters and either the (R)- or the (S)-alcohol can be obtained. Bernardi et al.644 showed that ethyl 3-oxobutanoate was reduced to ethyl (3R)-hydroxybutanoate in 96% yield, 7:93 (R/S) with G. candidum. The reaction medium is quite important. Nakamura et al.645 found, for example, that reduction of ketones with this enzyme in the presence of the hydrophobic polymer Amberlite XAD gave the (S)-alcohol in high ee, whereas low enantioselectivity was obtained in the absence of the polymer. Reduction of ethyl 3-oxobutanoate with Aspergillus niger gave 98% of a 75:25 mixture favoring the (R)-alcohol ethyl (R)-3-hydroxybutanoate.644 α-Keto esters (glyoxalates) are reduced with yeast, as illustrated by the conversion of methyl 4-Cbz-amino-2-oxobutanoate to methyl (S)-4-Cbz-amino-2-hydroxybutanoate in 47% yield (49 %ee).646 Aryl glyoxalates are usually reduced to the (R)-alcohol.646 It has been shown that using an acetone powder of a microorganism (e.g., G. candidum) led to improved stereoselectivity.647 Trifluoromethyl ketones are reduced with opposite stereochemistry relative to methyl ketones with this

640

Lieser, J. K. Synth. Commun. 1983, 13, 765.

641

Takeuchi, Y.; Azuma, K.; Takakura, K.; Abe, H.; Harayama, T. Chem. Commun. 2000, 1643.

642

Akita, H.; Furuichi, A.; Koshiji, H.; Horikoshi, K.; Oishi, T. Tetrahedron Lett. 1982, 23, 4051.

643

(a) Furiuchi, A.; Akita, H.; Koshiji, H.; Horikoshi, K.; Oishi, T. Chem. Pharm. Bull. 1984, 32, 1619. (b) Akita, H.; Furuichi, A.; Koshiji, H.; Horikoshi, K.; Oishi, T. Ibid. 1984, 32, 1342. (c) Idem Ibid. 1983, 31, 4376.

644

Bernardi, R.; Cardillo, R.; Ghiringhelli, D. J. Chem. Soc. Chem. Commun. 1984, 460.

645

Nakamura, K.; Fujii, M.; Ida, Y. J. Chem. Soc. Perkin Trans. 1 2000, 3205.

646

Iriuchijima, S.; Ogawa, M. Synthesis 1982, 41.

647

Matsuda, T.; Harada, T.; Nakajima, N.; Nakamura, K. Tetrahedron Lett. 2000, 41, 4135.

412

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

organism.648 Bioreduction using methanol as a cosolvent and the fungus Curvularia lunata CECT 2130 gave chemoselective reduction of aromatic β-keto nitriles to the corresponding (S)-β-hydroxy nitriles with high enantioselectivity.649 O

OH

A. niger

CO2Et

CO2Et

Ethyl 3-oxobutanoate

O

CbzHN

Ethyl (3R)-hydroxybutanoate (98%)

1. Baker's yeast

CO2Me

NH2

OH CO2Me

2. aq NaOH

Methyl (S)-4-Cbz-amino-2-hydroxybutanoate

Methyl 4-Cbz-amino-2-oxobutanoate

Veschambre and coworkers650 used Beauveria sulfurescens for the reduction of conjugated carbonyls. transCrotonaldehyde was reduced to but-(2E)-en-1-ol in 80% yield. (E)-2-Methylpent-2-enal, however, gave a mixture of 31% of the conjugated alcohol [(E)-2-methylpent-2-en-1-ol] and 69% of the completely reduced alcohol 2-methylpentan-1-ol.650 Later work showed this microbial reduction gave the (S)-alcohol by a stepwise process in which the conjugated double bond was reduced first, establishing the (S)-stereochemistry, followed by reduction of the aldehyde moiety to the alcohol. Veschambre and coworkers650b established a rule for this reduction, Veschambre’s rule. The hydrogen is delivered to the double bond α to the carbonyl from the front or back depending on the relative bulk of R1 and R2, which is simply a modification of Prelog’s rule. Reduction of geranial with baker’s yeast gave (R)-citronellol, consistent with the prediction. The (Z)-isomer neral was predicted to yield the (S)-isomer, but reduction with baker’s yeast gave a 6:4 (R/S) mixture, possibly due to isomerization of the double bond in neral prior to delivery of hydrogen.650b Shimoda et al.651 found that Synechococcus sp. PCC 7942, a cyanobacterium, reduced both the endocyclic double bond and s-trans-enones, as well as the exocyclic double bond of s-cis-enones with high enantioselectivity. B. sulfurescens

CHO

CH2OH

DMSO , 48 h , 20 °C

(E)-2-Methylpent-2-enal

+

(E)-2-Methylpent-2-en-1-ol (31%)

CH2OH 2-Methylpentan-1-ol (69%)

Another common organism for reduction of the carbonyl group is T. brockii, first mentioned in Section 7.9.3. Keinan et al.287 showed that small ketones (e.g., butan-2-one) were reduced by T. brockii to give (R)-butan-2-ol, in 12% yield and 48 %ee, but longer chain ketones (e.g., nonan-2-one) were reduced to (S)-hexan-2-ol in 85% yield and 96 %ee. An alcohol dehydrogenase found in Thermoanaerobacter ethanolicus reduces ethynyl ketoses [e.g., 4-methylpent-1-yn-3-one to (R)-4-methylpent-1-yn-3-ol] in good yield, and high enantioselectivity [in this case (R)-4-methylpent-1-yn-3-ol was isolated in 50% yield and >98 %ee, (S)].652 These examples illustrate typical enantioselectivity of the reduction and that the selectivity depends on the size and nature of the groups around the carbonyl. Changes in selectivity are explained by knowledge of the active site of the enzyme. It appears that various yeast preparations and T. brockii are the most commonly used reducing organisms. New organisms (e.g., lyophilized cells from Rhodococcus ruber DSM 44541), can be used to reduce ketones to alcohols in good yield, and with excellent enantioselectivity.653 The low cost and high selectivity of such enzymatic and microbial reagents foretell an increasingly important role in organic synthesis.

648

Matsuda, T.; Harada, T.; Nakajima, N.; Itoh, T.; Nakamura, K. J. Org. Chem. 2000, 65, 157.

649

Dehli, J. R.; Gotor, V. Tetrahedron Asymmetry 2000, 11, 3693.

650

(a) Desrut, M.; Kergomard, A.; Renard, M. F.; Veschambre, H. Tetrahedron 1981, 37, 3825. (b) Bostmembrun-Desrut, M.; Dauphin, G.; Kergomard, A.; Renard, M. F.; Veschambre, H. Ibid. 1985, 41, 3679. 651

Shimoda, K.; Kubota, N.; Hamada, H.; Kaji, M.; Hirata, T. Tetrahedron Asymmetry 2004, 15, 1677.

652

Heiss, C.; Phillips, R. S. J. Chem. Soc. Perkin Trans. 1 2000, 2821.

653

Stampfer, W.; Kosjek, B.; Faber, K.; Kroutil, W. J. Org. Chem. 2003, 68, 402.

413

7.13 CONCLUSION

O

OH

T. brockii

Butan-2-one

O

T. brockii

Hexan-2-one

(R)-Butan-2-ol

OH

(S)-Hexan-2-ol

(12% , 48 %ee, R)

(85% , 96 %ee, S)

OH

O T. ethanolicus

4-Methylpent-1-yn-3-one

(R)-4-Methylpent-1-yn-3-ol (50% , >98 %ee, S)

7.13 CONCLUSION It is obvious from this chapter that reductive processes are important in organic synthesis. The diastereoselectivity and enantioselectivity of many reducing agents are important for fixing stereogenic. An attempt has been made to show how reagents were developed, as well as their chemoselectivity. Other functional group transformations remain to be discussed. Reactions associated with hydroboration will be introduced in Chapter 9. This important synthetic methodology is a logical extension of the borohydride and borane reductions seen in this chapter. Chronologically, the reductive capabilities of boranes preceded the alkene addition reaction, which is the heart of hydroboration. In part, this is why hydroboration is presented after reduction. Also note that the stereochemical models introduced in this chapter will be used continuously in succeeding chapters. Addition of Grignard reagents and other organometallics to carbonyls (Chapters 11 and 12) generally follow Cram’s rule or the Felkin-Anh model. Chelation control will be important in many reactions including enolate alkylation reactions. This chapter has therefore been a vehicle for introducing not only reduction, but the stereochemical basis for all of organic chemistry.

HOMEWORK

1. Predict the correct diastereomeric product using (a) the appropriate Cram model and (b) the Felkin-Anh model. In each case, draw the model for that specific reaction. O 1. LiAlH4

(A)

O

(B)

2. dil H3O+

H

2. dil H3O+

O

(C)

H MeO

Ph O

1. Zn(BH4)2

Me

(D)

N

H

Me

H 1. Zn(BH4)2

Me 2. dil H O+ 3

2. dil H3O+

O

2. Predict the major product and explain the stereochemistry for that product. Me

1. LiAlH(Ot-Bu)3

O

NaBH4 , CeCl3

CO2H

414

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

3. In each case, predict the major diastereomer. Explain your answer. O NaBH4•CeCl3

(A)

Me

(B)

N

EtO

Me

Me

Me

Ts

OSiPh2t-Bu

Me

OH

(C)

0 °C

H H

OBn

O

Et H H 1. Zn(BH4)2 2. dil H3

H

LiAlH4 , THF

O 1. NaBH4

(D)

N

O+

2. dil H3O+

O

4. Explain why there is such a difference in selectivity for reduction of this conjugated ketone when the reducing agent is changed.

Me

H

OH

Me Me

1. L-Selectride , THF , –78°C

Me

O

H

H

2. H3O+

Me

H H

:

Me

H

OH

Me

97:3

:

1. dibal , THF , –78°C

R 93:7

Me

H

R

2. H3O+

THPO

5. Briefly discuss this transformation. H O TsO

CO2Et N CO2t-Bu

H

CHO

1. LiBEt3H 2. KOt-Bu

H Ph

Ph

H CO2Et N CO2t-Bu

6. Predict the major product of both reactions (A and B) and predict the stereochemistry of each product. Explain your choices. O OSiMe2Thex

O

mcpba Na2CO3

A

LiBEt3H , THF

B

H

7. Briefly discuss the following transformation, particularly focusing attention on the stereochemistry. 1. LiAlH4 , Ether 2. PPh3 , DEAD , PhCO2H

O

3. KOH , MeOH

OH

415

7.13 CONCLUSION

8. Explain each of the following observations or products: (a) The major process is 1,2-reduction. O

OH 1. AlH3 2. dil H3O+

(b) L-Selectride gives greater diastereoselectivity in this reduction than the sodium borohydride. Predict the major product. H H O H

(c) Treatment of benzene with Na/NH3/EtOH yields cyclohexa-1,4-diene, whereas similar treatment of cyclohexene with this reagent gives no reaction. (d) Birch reduction of anisole leads to 1-methoxycyclohexa-1,4-diene, but reduction of benzoic acid yields 3-carbalkoxy-1,4-cyclohexadiene. (e) Reduction with LiAlH4 yields a diol, but reduction with diborane yields a hydroxy ester. CO2H

CO2Et

(f ) Diimide reduces unconjugated alkenes faster than conjugated alkenes. (g) Reduction of A and B with zinc borohydride leads to opposite stereochemical results. O

O H

H

OH

OSiMe2t-Bu

A

B

(h) Birch reduction of A, and subsequent acid hydrolysis leads to the conjugated ketone B. O

OMe 1. Na , EtOH , Reflux

MeO

2. aq HCl , Reflux

MeO OMe

OMe

B

A

9. For each transformation, give two different reagents that will successfully complete the given conversion. O n-C5H11

O Cl

n-C5H11

n-C5H11 H

OTs

n-C5H11

CH3

416

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

10. In each case, predict the major product of the reaction shown. N2H4 , KOH Diethylene glycol

O

(A) O

1. LiAlH(Ot-Bu)3 , THF –10°C

Cl

(B)

O

2. H3O+

220°C

O H OH PivO(CH2)4

OPMB Me

Me

OPMB

O

1. Dess–Martin

(C) HO

MeO OPMB

PMB = p-Methoxybenzyl Piv = Pivaloyl

CO2H

MeO

NO2

H2 Ni

2. L-Selectride THF , –78°C

O

MeO

(D)

OTIPS OMe 1. Li , NH3

(E)

MeO2C

CO2Me

N

(F)

2. aq NH4Cl

CO2t-Bu

NH2 MeO

MeO

Br

MeO

CHO

NaBH(OAc)3 ClCH2CH2Cl OMe

O BH3•THF , THF

O

(G)

NH2

(H) N

t-BuMe2SiO

CO2H

CHO , THF

1.

2. NaBH4 , MeOH

Ph

CO2t-Bu CO2Me

(I)

AlH3 , THF

N CO Me 2

Br N

NHCO2t-Bu

(J)

rt

H

NaBH3CN , AcOH

N H

HO

H

Me Me H2

(K) t-BuMe2SiO

NaBH4 , MeOH

(L)

Lindlar catalyst

N

O

OH

H

t-Bu PhMe2Si

NHPhth N

(M)

O

CO2Me

CO2t-Bu

CO2Me

1. 6 equiv LiAlH4 2. H3O+

(N)

1. LiAlH4 , THF

O

2. Hydrolysis

N H

O

O

OH H

1. LiAlH4 2. Hydrolysis

(O) N

(P)

NH2NH2 , NaOH HOCH2CH2OH

C15 H31

Me

O O

CN

CO2Bn OH H

(Q) N O

1. PhOCSCl , Py 2. Bu3SnH , AIBN

OSiMe2t-Bu

OSiPh2t-Bu

(R)

CO2Me

1. dibal , CH2Cl2 –78°C 2. TPAP , NMO MeCN , MS 4 Å

417

7.13 CONCLUSION

(S) HO

O

OH

O

O

Bu3SnH , AIBN PhH

(T)

I

2. Hydrolysis

I

Me

O

Me3Si

1. Li , Liq NH3 THF

(U) Me

Me4NBH(OAc)3

Ph

(V)

2. H2O

O

O

OH

O

1. LiH2Al(OEt)2 , 0°C

NMe2

O

O

OH

MeCN , AcOH

O CN

OMe

MeO

CHO 1. NaBH4

(W)

(X)

O

2. PBr3 , Py

OMe

N

dibal , Toluene –78°C

NO2 CN

SMe

(Y)

Ni(R)

SMe Ni(R)

(Z )

SMe

SMe EtOH , Reflux

EtOH , Reflux

O

O TBSO

(AA)

H2 , Quinoline

O

O

Lindlar cat

O

Me2SiClH , CH2Cl2

(AB) Ph

Ph

5% InCl3

O CHO

(AC)

N

t-BuMe2SiO

Cl

(AD)

NaBH4 , EtOH

N

OH

Me4NBH(OAc)3 MeCN , AcOH

O

t-BuMe2SiO

–20°C

CO2Et CH3

11. In each case, provide a suitable synthesis showing all intermediate products and all reagents. Me

O O

OH

OPMB

Me

(A)

OH

OPMB

(C)

(B)

MeO HO

CO2Me

O

CO2Me

OHC Cl

HO

OH

OSiMe2t-Bu (CH 2)5

(D)

(E) CO2Me

O

(F)

Me O

CO2H

O

HO

O

O

(G)

SiMe3 O

O

OH

CO2Et O

O

OH

O Br

CO2H

(H)

n-C4H9

(I)

NH2

Cl O

N

O

O

H

(K)

EtO2C

(J)

OHC

N

N O

O MeO

SiMe3

OH MeO

H

OMe Cl

N

OMe

CHO

418

7. FUNCTIONAL GROUP EXCHANGE REACTIONS

12. Show a complete synthesis for each of the following: Use any starting material of your choosing, that contains six carbons or less. Show a retrosynthetic analysis and synthesis for each molecule. Br

O Ph

(A) Br

(C)

(B)

EtO2C

O

Me

(D) Ph

NHC4H9

13. Give reagents that will accomplish each of the following selective transformations:

O

OH

OMe

O

O O

Br

O

(D)

(C)

OH O

Br N CO2Bn

O

O

CHO

(E)

OMe

(B)

(A)

OCOPh

O N CO2Bn

OH

N3

(F) N Bn

O

NH2 N Bn

CHO

C H A P T E R

8 Synthetic Strategies 8.1 INTRODUCTION Throughout this book, functional group transforms for each new type of reaction are shown at the end of the discussion. In addition, disconnections for carbon-carbon bond-forming reactions will be given in Chapters 11–18. These transforms and disconnections form the basis of the modern approach to total synthesis. Complex molecules may require alternative strategies for their total synthesis rather than the simple ones presented earlier in this book. Professor E.J. Corey and his coworkers1a completed many syntheses spanning >30 years, which were summarized in a monograph. That monograph detailed several strategic approaches to total syntheses, where those principles were applied. Nicolaou et al.1b,c also published collections of total syntheses. The book by Hudlický and Reed2 discusses many syntheses as well as the planning and methodology behind those syntheses. The science of total synthesis dates to the 19th century, and Corey cited many examples that illustrate its history. In 1904 Perkin3 synthesized terpineol, in part to ascertain the actual structure of this natural product. In the synthesis of tropinone by Robinson in 1917,4 he envisioned that an “imaginary hydrolysis of the substance (tropinone) may be resolved into succinaldehyde, methylamine, and acetone.”4b If this analysis is examined, modern strategy terminology can be applied to see that Robinson was disconnecting tropinone into simple disconnect fragments. The synthesis of equilenin by Bachmann et al.5 in 1939 was the first multi-step synthesis of a steroid precursor to estrone. Fischer and Kirstahler6 synthesized heminand recognized that four similar pyrrole units composed the structure, requiring the synthesis of each fragment prior to assembling hemin. Me Me

Me O

Me

Me

N

N

N Fe

N

Me

OH

O

Me

N Me

Me

HO

CO2Et

EtO2C Terpineol

Tropinone

Equilenin

Hemin

1 (a) Corey, E. J.; Cheng, X.-M. The Logic of Chemical Synthesis; Wiley: New York, 1989. (b) Nicolaou, K. C.; Snyder, S. A. Classics in Total Synthesis I: More Targets, Strategies, Methods; Wiley: New York, 2003. (c) Nicolaou, K. C.; Sorenson, E. J. Classics in Total Synthesis; VCH: Weinheim, 1996. 2

Hudlický, T.; Reed, J. W. The Way of Synthesis; Wiley: Weinheim, 2007.

3

Perkin, W. H., Jr. J. Chem. Soc. 1904, 85, 654.

4

(a) Robinson, R. J. Chem. Soc. 1917, 111, 762. (b) Fleming, I. Selected Organic Syntheses; Wiley: London, 1973; p 18.

5

Bachmann, W. E.; Cole, W.; Wilds, A. L. J. Am. Chem. Soc. 1939, 61, 974.

6

(a) Fischer, H.; Kirstahler, A. Annalen 1928, 466, 178. (b) Fischer, H.; Zeile, K. Ibid. 1929, 468, 98.

Organic Synthesis http://dx.doi.org/10.1016/B978-0-12-800720-4.00008-8

419

Copyright © 2017 Michael Smith. Published by Elsevier Inc. All rights reserved.

420

8. SYNTHETIC STRATEGIES

The syntheses reported by Woodward,7 assumed a sophistication previously unknown, and targets (e.g., quinine, with von E. Doering),8 cortisone,9 strychnine,10 or reserpine11 were synthesized (along with many other molecules). Over the years, the structural and stereochemical features of the targets increased in complexity. As more complex structures were targeted, the newly acquired ability to control stereochemical features and stereogenic centers led to greater synthetic achievements. As part of many of the syntheses reported by Corey, the terminology for this approach to synthesis was formalized, making it an understandable process. This approach is called retrosynthesis, and it allows one to dissect a molecule and arrive at relatively simple starting materials in what is now called the disconnection (or synthon) approach.12 Most chemists now use the disconnection approach when they undertake a synthetic problem, but there are other strategies that can be applied to planning a synthesis.13 Corey included several strategies for synthesis in a computer program called Logic and Heuristics Applied to Synthetic Analysis (LHASA) (Section 8.4.2), with the goals of assisting chemists in synthetic planning and actually suggesting syntheses for complex molecules. This approach to synthesis will be presented in Section 8.4 this chapter, as will other techniques for assembling a retrosynthetic plan. OMe MeO OH H

MeO HO

H

H

H

H

H

H H

N

O

OH

Me

MeO

N

Me

O

N

O H

O

N OMe

N H

O

O

O

H

N

CO2Me

H Quinine

Cortisone

Strychnine

MeO

Reserpine

8.2 TARGET SELECTION 8.2.1 What Is the Rationale for Total Synthesis? The choice of a target is the obvious point of departure for a synthesis. The synthesis of an organic molecule usually begins with two issues: (1) why was this molecule chosen as a target? and (2) how can reactions for the synthesis, including the starting material and all reagents be determined? The answer to (1) often lies in the needs and interests of the synthetic chemist. It may be a challenging stereochemical problem or functional group combination. A molecule may possess unique chemical or biological properties. The answer to (2) is the basis for this chapter. Many reasons are cited for a synthesis, but the challenging problems encountered in a total synthesis that provide many opportunities to find new reactions, processes, or strategies that may be of value to other organic chemists is often the most satisfying reason. Corey and Wipke14 stated a few of the useful applications that may be found in a retrosynthetic analysis-synthesis problem: “(i) the selection of specific chemical reactions for transforming one synthetic intermediate to the next in the sequence, (ii) the selection of specific reagents, (iii) the design of experiments, and (iv) experimental execution and analysis. During these stages, observations, discoveries, inventions, or theorems of great importance may result.” Many of these points may be useful in solving problems for another target, or in applications to a completely different area of organic chemistry. 7

Woodward, R. B. In Perspectives in Organic Chemistry; Todd, A. R., Ed.; Interscience: New York, 1956, pp 159–184.

8

Woodward, R. B.; Doering, W. E. J. Am. Chem. Soc. 1945, 67, 860.

(a) Woodward, R. B.; Sondheimer, F.; Taub, D.; Heusler, K.; McLamore, W. M. J. Am. Chem. Soc. 1952, 74, 4223. (b) Woodward, R. B.; Sondheimer, F.; Taub, D. Ibid. 1951, 73, 4057, 3547. 9

10

Woodward, R. B.; Cava, M. P.; Ollis, W. D.; Hunger, A.; Daeniker, H. U.; Schenker, K. Tetrahedron 1963, 19, 247.

11

Woodward, R. B.; Bader, F. E.; Bickel, H.; Frey, A. J.; Kierstead, R. W. Tetrahedron 1958, 2, 1.

12

Reference 1, pp 81–91.

13

D’Angelo, J. D.; Smith, M. B. Hybrid Retrosynthesis: Organic Synthesis Using Reaxys and SciFinder; Elsevier: Oxford, 2015.

14

Corey, E. J.; Wipke, W. T. Science 1969, 166, 178.

421

8.2 TARGET SELECTION

A similar fundamental basis for total synthesis was described by Deslongchamps: “… since the researcher has put himself/herself in a situation that organic chemists have not faced before, the chances of discovering something new and original are then quite high, and … it is the chemistry that one discovers along the way that is the important parameter, not the fact that one succeeds in the synthesis of a given compound, natural or non-natural.”15 It is often true that synthesis of a particular target is of paramount importance to an individual or organization. The criterion for that importance is always obvious to those researcher(s). To one beginning a study of organic reactions and synthesis, however, it may be of value to examine several commonly quoted criteria for synthesis.

O

H N

O H N

O2S N

N

N Me

O

N

OH

N

OH

OH H N

N

N

N

OH OH

N

O

O

O NHt-Bu

Me (–)-Isatisine A

Viagra (Sildenafil)

Indinavir

In addition to the eight molecules shown in the Introduction, six molecules are shown as examples of the wide variety of organic molecules that have been synthetic targets. (
)-Isatisine A,16 was isolated from the “leaves and root of Isatis indigotica Fort, named ‘Da-Qing-Ye’ and ‘Ban-Lan-Gen,’ respectively, in Chinese, have been used in traditional Chinese medicine for the treatment of viral diseases for hundreds of years in China.”17 Identification of this structure, coupled with potential biological applications is a good incentive for total synthesis. Not all targets come from nature. Viagra (sildenafil)18 was developed by Pfizer Inc. to treat impotence, and Merck Inc. developed indinavir19 as an HIV-1 protease inhibitor. Epothilone is a compound with the same mechanism of action as the important anticancer drug taxol, but has a less complex structure that makes it easier to prepare.20 The macrocyclic depsipeptide kitastatin is a target for organic synthesis that is particularly interesting since it appears to have preferential cytotoxic activity against pancreatic tumor cells.21 The preparation of [4]-phenylene helped show that the delocalized bonds of a benzene ring were effectively localized in these compounds, generating the elusive cyclohexatriene unit.22 This discovery caused scientists to reexamine the underlying principles of aromaticity.

15

Deslongchamps, P. Aldrichim. Acta 1984, 17, 59.

16

Wu, W.; Xiao, M.; Wang, J.; Li, Y.; Xie, Z. Org. Lett. 2012, 14, 1624.

17

Zheng, H. Z.; Dong, Z. H.; Yu, Q. Modern Study of Traditional Chinese Medicine; Xueyuan Press: Beijing, 1997; Vol. 1, p 328.

(a) Bell, A. S.; Brown, D. T.; Nicholas, K. Eur. Pat. Appl. EP 463,756 (Chem. Abstr. 1992, 116, P255626q). For biological activity, see (b) Boolell, M.; Gepi-Attee, S.; Gingell, J. C.; Allen, M. J. Br. J. Urol. 1996, 78, 257. (c) Terrett, N. K.; Bell, A. S.; Brown, D.; Ellis, P. Bioorg. Med. Chem. Lett. 1996, 6, 1819.

18

(a) Vacca, J. P.; Holloway, M. K.; Dorsey, B. D.; Hungate, R. W.; Guare, J. P. Eur. Pat. Appl. 541,168 (Chem. Abstr. 1994, 120, 54552w). (b) Vacca, J. P.; Dorsey, B. D.; Guare, J. P.; Holloway, M. K.; Hungate, R. W.; Levin, R. B. US Patent 5,413,999 (Chem. Abstr. 1995, 123, 47889v). (c) Dorsey, B. D.; Levin, R. B.; McDaniel, S. L.; Vacca, J. P.; Guare, J. P.; Darke, P. L.; Zugay, J.; Emini, E. A.; Schleif, W. A.; Quintero, J. C.; Lin, J. H.; Chen, I.-W.; Holloway, M. K.; Fitzgerald, P. M. D.; Axel, M. G.; Ostovic, D.; Anderson, P. S.; Huff, J. R. J. Med. Chem. 1994, 37, 3443. (d) Vacca, J. P.; Dorsey, B. D.; Schleif, W. A.; Levin, R. B.; McDaniel, S. L.; Darke, P. L.; Zugay, J.; Quintero, J. C.; Blahy, O. M.; Roth, E.; Sardana, V. V.; Schlabach, A. J.; Graham, P. I.; Condra, J. H.; Gotlib, L.; Holloway, M. K.; Lin, J.; Chen, I.-W.; Vastag, K.; Ostovic, D.; Anderson, P. S.; Emini, E. A.; Huff, J. R. Proc. Natl. Acad. Sci. 1994, 91, 4096. 19

For a synthesis, see (a) Meng, D.; Bertinato, P.; Balog, A.; Su, D.-S.; Kamenecka, T.; Sorensen, E. J.; Danishefsky, S. J. J. Am. Chem. Soc. 1997, 119, 10073. For isolation and characterization, see (b) Gerth, K.; Bedorf, N.; H€ ofle, G. Irschik, H.; Reichenbach, H. J. Antibiot. 1996, 49, 560. (c) H€ ofle, G.; Bedorf, N.; Steinmetz, H.; Schomburg, D.; Gerth, K.; Reichenbach, H. Angew. Chem. Int. Ed. 1996, 35, 1567. For biological activity, see Bollag, D. M.; McQueney, P. A.; Zhu, J.; Hensens, O. Koupal, L.; Liesch, J. Goetz, M.; Lazarides, E.; Woods, C. M. Cancer Res. 1995, 55, 2325. 20

21 (a) Beveridge, R. E.; Batey, R. A. Org. Lett. 2014, 16, 2322. (b) Pettit, G. R.; Tan, R.; Pettit, R. K.; Smith, T. H.; Feng, S.; Doubek, D. L.; Richert, L.; Hamblin, J.; Weber, C.; Chapuis, J.-C. J. Nat. Prod. 2007, 70, 1069. (c) Izumikawa, M.; Ueda, J.-Y.; Chijiwa, S.; Takagi, M.; Shin-ya, K. J. Antibiot. 2007, 60, 640. (d) Pettit, G. R.; Smith, T. H.; Feng, S.; Knight, J. C.; Tan, R.; Pettit, R. K.; Hinrichs, P. A. J. Nat. Prod. 2007, 70, 1073.

The chemical name of [4]-phenylene is bis-benzo[3,4]cyclobuta[1,2-a:10 ,20 -c]biphenylene. See (a) Diercks, R.; Vollhardt, K. P. C. J. Am. Chem. Soc. 1986, 108, 3150. (b) Vollhardt, K. P. C.; Mohler, D. L. Adv. Strain Org. Chem. 1996, 5, 121. (c) Schulman, J. M.; Disch, R. L. J. Phys. Chem. A 1997, 101, 5596. 22

422

8. SYNTHETIC STRATEGIES

H3C Me

S N

O

O O H

OH

H O

CH3

O

H2 N

H O

O Me O

HO

O

O

NH

O

O

OH H3C

O

O O

CH3 CH3

Epothilone

Kitastatin

[4]-Phenylene

Differing structural features offer unique synthetic problems. When such syntheses are published, the step-by-step reaction sequence used to construct each molecule is reported and, occasionally, a summary of the rationale behind the synthesis. This summary sometimes provides insight into why a target was chosen, but not always. To a chemist searching for a target, inspection of previous work may not be of great value, although a close study of these syntheses will be of enormous value for understanding and devising strategies. The five criteria listed below are often cited as characteristic of important synthetic targets. These criteria may be of value to a chemist who is being introduced to synthesis for the first time. Much of the information gained from examining these criteria may be useful for disconnecting the target. (1) Structural verification; (2) important biological activity; (3) analog generation and studies; (4) structural or topological challenges and a fundamental understanding of the nature of bonding and molecules; (5) development of new reactions or reagents.

8.2.2 Structural Verification Single crystal X-ray crystallography has become a powerful method for determining the structures of complex organic molecules.23 An example where X-ray analysis was invaluable for the structure proof is taken from the Qian-Cutrone24 isolation of new antitumor alkaloids, stephacidin A and B, produced by Aspergillus ochraceus WC76466. Extensive analysis using nuclear magnetic resonance (NMR) suggested that stephacidin B was a dimeric structure related to stephacidin A. Although all structural fragments were determined, the point at which the monomeric units of stephacidin A were linked to form the dimer could not be established. Single-crystal X-ray analysis established the final structure as stephacidin B, with the relative stereochemistry shown, as well as the presence of the nitrone and hydroxylamine units.25 It is reasonable to ask why a lengthy synthetic verification might be necessary for a given target when such techniques are available. In many cases, X-ray analysis is not possible due to an unsuitable crystalline form, the unavailability of single crystals, poor morphology of the molecule or a derivative. Although crystalline derivatives can sometimes be prepared, in many cases the molecule does not form a crystalline derivative at all. As noted for volubilide, even high-field NMR techniques often fail to give the absolute stereochemistry of all chiral centers, although the relative stereochemistry may be discerned. Chemical degradation used in conjunction with spectroscopy is also an important identification protocol, but is not successful in every case. Total synthesis and comparison with an authentic sample therefore remains an important method for final structure determination. Despite the power of X-ray crystallography and high-field NMR, the structures of many complex natural products remain unknown, or were incorrectly reported. There are many examples, and some reported by Weinreb26 For example, see (a) Powell, H. M.; Prout, C. K.; Wallwork, S. C. Ann. Rept. Progr. Chem. 1964, 60, 593. (b) Seemann, V. AD 626597, U.S. Govt. Res. Develop. Rept. 1966, 41, 99 (Chem. Abstr. 1967, 67, R15845m). (c) Landoet-Boernster, Group III: Crystal and Solid State Physics. Pts A & B: Structure Data of Organic Crystals; Schredt, E.; Weitz, G., Eds.; Springer: New York, 1971; Vol. 5. (d) Koyama, H.; Okada, K.; Itoh, C. Acta Crystallogr. B 1970, 26, 444.

23

24

Qian-Cutrone, J.; Krampitz, K. D.; Shu, Y. Z.; Chang, L. P. US Patent 6,291,461, 2001.

Qian-Cutrone, J.; Huang, S.; Shu, Y.-Z.; Vyas, D.; Fairchild, C.; Menendez, A.; Krampitz, K.; Dalterio, R.; Klohr, S. E.; Gao, Q. J. Am. Chem. Soc. 2002, 124, 14556.

25

26

Weinreb, S. M. Acc. Chem. Res. 2003, 36, 59.

423

8.2 TARGET SELECTION

include cylindrospermopsin,27 the sclerophytins28 and batzelladine F,29 where the correct structures were determined by the total synthesis. Weinreb26 uses the example of lepadiformine to demonstrate the value of total synthesis for determination of the structure of a natural product where the original report gave an incorrect structure. Synthesis is used for proof of structure, or sometimes revision of an originally proposed structure as in Nagumo and coworker’s30 synthesis of sekothrixide, which was originally published by Seto and coworkers.31 The synthesis proved that the original stereochemistry at C4, C6, and C8 was incorrect, leading to a revision of the originally proposed structure. Synthesis of the compound allowed the correct structure to be defined, and revised. O

O

O N NH

OH

OH

OH

N

NH O

HN

H

O

O O

O

H O N O

Sekothrixide

O N O

N

O

N

HO Stephacidin A

Stephacidin B

8.2.3 Biological Activity An obvious reason for total synthesis is the importance of the target in medicine, agriculture, or other commercial and humanitarian ventures. An important or interesting molecule is usually isolated from a natural source in very small quantity. If the molecule is subsequently shown to possess significant biological activity, additional material may be required for further testing or determination of the structure. The small quantities isolated or the difficulties and expense inherent in gathering natural specimens often prevent collection of sufficient amounts of the natural product. A bioactive compound may be isolated from a marine sponge that is harvested in the ocean, but only in one small cove on the West Coast of one small island in the Arctic Ocean, and only at depths of >200 ft, for example. The example cited may be gratuitous, but if a molecule of interest is to undergo further biological screening, or be used clinically, total synthesis may be the only means to obtain sufficient material. An example is maeocrystal V, isolated from Isodon eriocalyx (Dunn.) Hara, found in southwestern China. This plant is important in Chinese folk medicine, used to treat sore throat, inflammation, influenza, hypertension, and dermatophytosis.32 Maeocrystal V showed strong inhibitory activity against HeLa cells at very low concentrations (IC50 ¼ 0.02 mg mL
1; cisplatin IC50 ¼ 0.99 mg mL
1) but was almost noncytotoxic toward the other four cell lines. This cytotoxicity is highly selective (only exhibiting potent cytotoxicity against gynecologic cancer cells). When coupled with an unusually complex structure, interest in its synthesis has been intense.

27

Heintzelman, G. R.; Fang, W.-K.; Keen, S. P.; Wallace, G. A.; Weinreb, S. M. J. Am. Chem. Soc. 2002, 124, 3939.

(a) Bernardelli, P.; Moradei, O. M.; Friedrich, D.; Yang, J.; Gallou, F.; Dyck, B. P.; Doskotch, R. W.; Lang, T.; Paquette, L. A. J. Am. Chem. Soc. 2001, 123, 9021. (b) Overman, L. E.; Pennington, L. D. Org. Lett. 2000, 2, 2683.

28

29

Cohen, F.; Overman, L. E. J. Am. Chem. Soc. 2001, 123, 10782.

30

Terayama, N.; Yasui, E.; Mizukami, M.; Miyashita, M.; Nagumo, S. Org. Lett. 2014, 16, 2794.

(a) Kim, Y. J.; Furihata, K.; Shimazu, A.; Furihata, K.; Seto, H. J. Antibiot. 1991, 44, 1280–1282. (b) Fujita, K.; Fujiwara, M.; Yamasaki, C.; Matsuura, T.; Furihata, K.; Seto, H. Stereochemistry Structure Determination Method Used by Computer. Proceedings of the 38th Symposium Chemical Natural Products, Sendai, Japan, Oct 14–16; 1996, pp 379–384.

31

32

Li, S.-H.; Wang, J.; Niu, X.-M.;Shen, Y-H.; Zhang, H.-J.; Sun, H.-D.; Li, M.-L.; Tian, Q.-E.; Lu, Y.; Cao, P.; Zheng, Q.-T. Org. Lett. 2004, 6, 4327.

424

8. SYNTHETIC STRATEGIES

O O

Me

O

Me Me

O O Maeocrystal V

The synthesis of a complex natural product by a multi-step synthetic route can be very expensive. To produce even 1 g of active product can require large quantities of both starting material and time. If the synthetic goal is to produce usable quantities of material at a reasonable cost, this must be factored into the retrosynthetic scheme. If one requires only milligram quantities, greater latitude in the choice of starting materials and reagents is possible.

8.2.4 Analog Studies A target that has interesting or commercially attractive properties may exhibit deleterious side effects, or be unstable to storage and handling. Subsequent studies often show that structural modification to the basic skeleton yields a molecule with different characteristics and the new molecule becomes the synthetic target.

Cl

Cl

N

N H2NO2S

O

S

N

H

O

Chlorothiazide

H2NO2S O Chlorexolone

Chlorothiazide is 2H,1,2,4-benzothiadiazine-7-sulfonamide, 6-chloro-1,1-dioxide.33 It is a diuretic also used to treat hypertension. Studies “determined that significant alterations in the pharmacological profile would follow changes in the heterocyclic ring rather than the ring bearing the sulfonamide group.”34 Structural modification led to the synthesis of chlorexolone.35 Conversion of the heterocyclic ring in chlorothiazide to the N-cyclohexyl lactam moiety in chlorexolone resulted in improved hypotensive36 and diuretic action, but did not increase urinary pH or urinary bicarbonate excretion relative to chlorothiazide.37 This example shows that synthetic modification of a known structure can identify a synthetic target.

(a) Novello, F. C.; Sprague, J. M. J. Am. Chem. Soc. 1957, 79, 2028. (b) Novello, F. C. US Patent 2,809,194 (Chem. Abstr. 1958, 52, 2939h). (c) Hinkley, D. F. US Patent 2,937,169 (Chem. Abstr. 1960, 54, 18565i). (d) Dupont, D.; Dideberg, O. Acta Crystallogr. B 1970, 26, 1884. (e) Wolff, F. W.; Basabe, J.; Grant, A.; Krees, S.; Lopez, N.; Vicktora, J. Med. Ann. DC, 1971, 40, 98 (Chem. Abstr. 1971, 75, 47093w). (f) Tubaro, E., Boll. Chim. Farm. 1963, 102, 505 (Chem. Abstr. 1964, 60, 2198b). (g) Ford, R. V. Ann. NY Acad. Sci. 1960, 88, 809 (Chem. Abstr. 1964, 61:12493d). (h) Sorice, F., Policlinico (Rome), Sez. Prat. 1960, 67, 625 (Chem. Abstr. 1961, 55, 14692cd).

33

34

Lednicer, D. L.; Mitscher, L. A. Organic Chemistry of Drug Design; Wiley: New York, 1977; Vol. 1, p 321.

35

May & Baker Ltd. Belg. Patent 620,654 (Chem. Abstr.1963, 59, P11436c).

(a) Maxwell, D. R.; McLusky, J. M. Nature (London) 1964, 202, 300. (b) Bayeli, P. F.; Montagnani, M.; Zampetti, L. P.; Antonelli, A. Boll. Soc. Ital. Biol. Sper. 1969, 45, 406 (Chem. Abstr. 1970, 73, 2550k). (c) Patterson, R. R.; Macaraeg, Jr., P. V. J.; Schrogie, J. J. Clin. Pharmacol. Ther. 1969, 10, 265 (Chem. Abstr. 1969, 70, 95348b). 36

37

Baba, W. L.; Lant, A. F.; Wilson, G. M. Clin. Pharmacol. Therap. 1966, 7, 212 (Chem. Abstr. 1966, 64, 18220h).

425

8.2 TARGET SELECTION

Another rationale for preparing analogs of a given target is to study the structure-biological activity profile of a molecule. Flavone 8-acetic acid was believed to have potential as an antitumor agent.38 Denny and coworkers39 showed that 9-oxo-9H-xanthene-4-acetic acid (a) is as active as flavone 8-acetic acid against colon-38 tumors in mice, and is more dose potent. It was also shown that small lipophilic substituents at the 5-position (b and c) enhanced dose potency.40 A systematic study41 of synthetic derivatives of 9-oxo-9H-xanthene-4-acetic acid showed that many related derivatives possessed enhanced antitumor activity, which is a typical case in which a known compound was structurally modified based on structure-activity properties of a related compound. The impetus for this synthesis was therefore a search for a more potent and efficacious drug. O

O

O

O CO2H

Flavone 8-acetic acid

X

CO2H

9-Oxo-9H-xanthene-4-acetic acid (a) X = H (b) X = Me (c) X = Cl

Dodecahedrane

8.2.5 Structural Challenges Occasionally, a molecule presents such a structural and chemical challenge that its potential as a target is irresistible. Such a synthesis often pushes back the limits of known chemistry for making carbon-carbon bonds, and gives insight into structure, bonding, or fundamental reaction properties of organic molecules. Paquette and coworkers42a synthesized dodecahedrane, and the topology (shape, structural features) is essentially that of a ball. Having the form of a dodecahedron, first described as one of five regular polyhedra in 400–350 B.C.E. in Plato’s Timaeus, dodecahedrane has a spherical superpolycyclopentanoid topology42b that possesses the highest known point group symmetry (In, icosahedral).42a It also shows a unique encapsulation of a cavity incapable of solvation, negligible angle strain, but great torsion strain,43 20 symmetry equivalent methine units, and an absence of structurally allied substances. Paquette and coworkers42 were the first to synthesize this molecule, and the creative and sometimes novel chemistry arising from the synthetic solution to targets (e.g., dodecahedrane) can usually be applied to other synthetic endeavors.

8.2.6 New Reactions and Reagents During a total synthesis, completion of a key transformation via known chemical reactions may be impossible. Development of new methodology or modification of existing reactions must be accomplished to give the desired transformation and complete the synthesis. Once developed, later work often shows the reagents developed for that synthesis can be utilized for a variety of other purposes. An example is the total synthesis of vitamin B12 reported by Woodward44a and Eschenmoser.44b Late in this synthesis, a key step required hydrolysis of an amide in an intermediate possessing six ester moieties. The reagent used in this transformation was an α-chloronitrone [(E)-2-chloro-Ncyclohexylethan-1-imine oxide], which gave a vinyl nitronium ion [(oxo(vinyl)ammonio)cyclohexane] when treated with silver ion.

(a) Smith, G. P.; Calveley, S. B.; Smith, M. J.; Baguley, B. C. Eur. J. Cancer Clin. Oncol. 1987, 23, 1209 (Chem. Abstr. 1988, 108, 15894s). (b) Ching, L.-M.; Baguley, B. C. Ibid. 1989, 25, 821 (Chem. Abstr. 1989, 111, 33243v). (c) Idem Ibid. 1989, 25, 1513 (Chem. Abstr. 1990, 112, 15992e). (d) Zwi, L. J.; Baguley, B. C.; Gavin, J. B.; Wilson, W. R. J. Natl. Cancer Inst. 1989, 81, 1005.

38

39

Rewcastle, G. W.; Atwell, G. J.; Baguley, B. C.; Calveley, S. B.; Denny, W. A. J. Med. Chem. 1989, 32, 793.

40

Atwell, G. L.; Rewcastle, G. W.; Baguley, B. C.; Denny, W. A. J. Med. Chem. 1990, 33, 1375.

41

Rewcastle, G. W.; Atwell, G. J.; Zhuang, L.; Baguley, B. C.; Denny, W. A. J. Med. Chem. 1991, 34, 217.

42

(a) Ternasky, R. J.; Balogh, D. W.; Paquette, L. A. J. Am. Chem. Soc. 1982, 104, 4503. (b) Mehta, G. J. Sci. Ind. Res. 1978, 37, 256.

43

Ermer, O. Angew. Chem. Int. Ed. Engl. 1977, 16, 411.

44

(a) Woodward, R. B. Pure Appl. Chem. 1973, 33, 145. (b) Eschenmoser, A. Naturwissenschaften 1974, 61, 513.

426

8. SYNTHETIC STRATEGIES

H2NOC

CONH2 Me

Me

H2NOC Me Me

NR

N

CONH2

Co N

N

HO

MeO2C Me

O

Me

Me

N H Me H

OH

Me

O

R =

O N

OH

P

CONH2

N

O O

N

Me

HO

O

N N NH2

N

O H HO

H

H

Me H

Vitamin B12

MeO2C O

N N

O

N

N

Me

H2N Cl (E)-2-Chloro-N-cyclohexylethan-1-imine oxide

MeO2C

+

N

H2N (Oxo(vinyl)ammonio)cyclohexane

MeO2C

Me

O

O Me N

HO

O 1

The reaction of the α-chloronitrone with the amide unit of vitamin B12 gave 1. Hydrolysis of 1 gave the corresponding acid, allowing the synthesis of vitamin B12 to be completed. Some years later, Eschenmoser and coworkers45 developed a series of transformations using derivatives of the α-chloronitrone, (E)-2-chloro-N-cyclohexylethan-1-imine oxide. This nitrone was used as a diene partner in a Diels-Alder reaction (Section 14.9), and the resultant cycloadduct (2) was transformed into several synthetically useful molecules. The secondary product that competed with the cycloaddition was the addition product (3), which was also converted to synthetically useful products. Note that the endproducts include furan and alkyne derivatives45 when the addition partner is changed from alkenes to alkynes to other derivatives. In the specific example shown, cycloaddition of 1-methylcyclohex-1-ene and (E)-2-chloro-N-cyclohexylpropan-1-imine oxide gave the addition product (3) in 59% yield, and subsequent hydrolysis led to a 90% yield of 2-(2-methylcyclohex-1en-1-yl)propanal. An additional product, cycloadduct (2), was formed in 30% yield. This sequence illustrates how a reagent developed for one purpose can be a powerful reagent in a completely different application. Indeed, the discovery of such new reagents and transformations is commonly the result of a total synthesis.

(a) Shatzmiller, S.; Gygax, P.; Hall, D.; Eschenmoser, A. Helv. Chim. Acta 1973, 56, 2961. (b) Kempe, U. M.; Das Gupta, T. K.; Blatt, K.; Gygax, P.; Felix, D.; Eschenmoser, A. Ibid. 1972, 55, 2187. (c) Gygax, P.; Das Gupta, T. K.; Eschenmoser, A. Ibid. 1972, 55, 2205. (d) Das Gupta, T. K.; Felix, D.; Kempe, U. M.; Eschenmoser, A. Ibid. 1972, 55, 2198. (e) Petrzilka, M.; Felix, D. Eschenmoser, A. Ibid. 1973, 56, 2950. (f) Shatzmiller, S. Eschenmoser, A. Ibid. 1973, 56, 2975.

45

427

8.3 RETROSYNTHESIS

Me N

O

+

N Me Cl

Cl

H

Me

Me AgBF4 SO2

1-Methylcyclohex1-ene

2 (30%) Me

Me

H3O+

(E)-2-Chloro-N-cyclohexylpropan-1-imine oxide

H

N Me

O

O

3 (59%)

Me 2-(2-Methylcyclohex- (90%) 1-en-1-yl)propanal

To summarize, the criteria discussed for target selection can be useful during a search for possible molecules to synthesize. They may also offer insights into possible applications to chemical problems that may arise. An understanding of the history of a target and how it is used often provides clues to its chemistry and physical properties. These are essential to execution of a synthesis, if not the actual retrosynthetic analysis.

8.3 RETROSYNTHESIS When a target is chosen, there are guidelines that allow its systematic disconnection to a starting material. The process that develops this roadmap of synthetic intermediates is called retrosynthesis. This section will describe several ways to analyze a molecule, with the goal of identifying key bonds for disconnection.

8.3.1 The Disconnection Approach Once a synthetic target has been chosen, reactions must be chosen for use in a total synthesis. The choice of a starting material (the molecule, purchased or readily prepared, that is used in the first reaction of the synthetic sequence) is critical. How does one analyze a target in order to determine the best starting material? The answer to that question is the essence of total synthesis, and its answer requires: (1) a detailed analysis of the structure of the target, (2) an excellent knowledge of chemical reactions, (3) a good understanding of stereochemistry, bonding, and reactivity, and (4) a well-developed chemical intuition. This information is then used to disconnect bonds in the target, simplifying the structure along reasonable chemical reaction pathways.1 Wipke and Howe46 defined the important bonds that are disconnected, as approached by the computer program LHASA (Section 8.4.2) developed by Corey et al.47,48 for analysis of retrosynthetic pathways: Their definition was “There are usually certain bonds in a molecule whose disconnection in the retrosynthetic direction leads to a significant simplification of the … structure. These bonds are termed strategic bonds.” If one of these bonds is to be disconnected, a reaction (or sequence of reactions) must be available that will chemically form that bond in the synthesis. Subsequent simplifications (disconnections) lead ultimately to a molecule that can be recognized as commercially available, available by simple chemical techniques, or already prepared by others. This process of structural simplification via disconnection leads to a series of molecular fragments that serve as key intermediates, and each is a synthetic target. Such intermediates allow the synthetic chemist to mentally bridge the starting material with the final target in a logical and sequential manner. This process allows the construction of a synthetic tree, for which chemical reactions must be provided to accomplish the planned transformations.

46

Wipke, W. T.; Howe, W. J. Computer Assisted Organic Synthesis; American Chemical Society: Washington, DC, 1971.

(a) Corey, E. J.; Howe, W. J.; Pensak, D. A. J. Am. Chem. Soc. 1974, 96, 7724. (b) Corey, E. J.; Wipke, W. T.; Cramer, R. D., III; Howe, W. J. J. Am. Chem. Soc. 1972, 94, 421. (c) Corey, E. J.; Jorgensen, W. L. Ibid. 1976, 98, 189.

47

48

Corey, E. J. Q. Rev. Chem. Soc. 1971, 25, 455.

428

8. SYNTHETIC STRATEGIES

OH

T o = Target O

OH Br

OH

OH

O

FIG. 8.1 A synthesis tree.

An online synthetic tree for a very simple molecule was presented by the Helsinki University of Technology, as shown in Fig. 8.1.49 The target (To) is disconnected to “a logically restricted set of structures that may be converted in a single synthetic operation (a chemical step) to the synthetic target.”14 This analysis yields the disconnection product in the first retro-reaction. The subtree represents further disconnection of one branch of the synthetic tree. After several retro-reactions, the synthetic tree will yield several starting materials from one or more of the first branches, which can then be used to construct the molecule. Working backwards toward the starting material in the manner illustrated in Fig. 8.1 is termed retrosynthesis.1 Each succeeding structure has been modified and always simplified. Bonds have been disconnected, and functional groups transformed until the relatively simple starting materials shown were obtained. The term disconnection14,50 therefore refers to “mentally breaking bonds to yield successively more simple precursor molecules, but always in a manner in which those bonds can be reformed by known or reasonable chemical reactions.” Warren50 has described this approach in great detail. Massanet and coworkers51 provided an example of the disconnection analysis as applied to a real synthetic problem, in a synthesis of thapsigargin (4). In Fig. 8.2, the retrosynthesis analysis begins with clipping the labile lactone ring to yield 5. The next disconnection leads to fragment 6, in which an isopropenyl unit is the precursor to the lactone-diol moiety. This fragment is, in turn, disconnected to yield fragment 7 indicating that no new carbon atoms are incorporated, but rather indicates a skeletal reorganization (in this case a rearrangement reaction). The final disconnection of 7 leads to two fragments (8 and 9), which are either commercially available or readily prepared. This analysis is outlined in the box and shows how a molecule is simplified into smaller, and more manageable fragments by a series of disconnections. OAc

H O

OH OH

O 4

OAc

H

OAc

H

O HO

O

OH OH

O

O

O OH

HO OH

O

5

7

6 9 7

6

5

4

8 51

FIG. 8.2 Massanet and coworker’s retrosynthetic analysis of thapsigargin (4). 49

Ottenheijm, H. C. J. Janssen Chim. Acta 1984, 2, 3.

50

Warren, S. Designing Organic Syntheses: A Programmed Introduction to the Synthon Approach; Wiley: New York, 1978.

51

Manzano, F. L.; Guerra, F. M.; Moreno-Dorado, F. J.; Jorge, Z. D.; Massanet, G. M. Org. Lett. 2006, 8, 2879.

8

9

429

8.3 RETROSYNTHESIS

Each fragment is prepared and the molecule assembled, following what is essentially the reverse of the retrosynthetic sequence shown. Analysis and simplification of a complex target using this approach is the essence of retrosynthesis. A decision must be made concerning the relative priority of simplification versus controlling the stereochemistry of the stereogenic center, which was not made clear in the above analysis. The analysis of Target 4 must include the reality of choosing a method to chemically form that bond in the context of the remaining functionality and stereochemistry. There is not necessarily one “correct” synthesis, but rather several possibilities, each with its strengths and weaknesses.

8.3.2 The Problem of Complex Targets The disconnection of truly complex targets raises the level of difficulty to a higher level of analyses. Rainier and coworker’s52 synthesis of brevenal makes the point, where the retrosynthesis shown in Scheme 8.1 was given in the paper. For the salient reactions of the synthesis, the reader referred to the paper, but several features are clear. First, the initial disconnection of ring C generates fragment 10. Refunctionalization of fragment 10 accompanied by disconnection of ring D gave 11. Disconnection of the ester unit in 11 led to the two key fragments, 12 and 13. The complexity of the target is reflected in the retrosynthetic analysis, which includes not only reactions and functionality, but also stereochemistry. OHC

Me

Me

Me H

H O

A

O H OH

H

OH

E

O D

C

B

O Me

H

H

H

Me

H O H

H

Me H

Me H

Brevenal A

O

O Me

B

PO H

H PO H

O

O

A O Me

O

B H

H

OP

O

OP

Me

H

H

H

10

Me H

O H

OP

OP

OP

E

D H

H

Me

H

H

OP

E O H

H

H

OP

H

OP 11

Me

H

O

O

O Me 12

H

OH

B H

H

OP

O

OP

OP

E H

H A

PO

HO

Me

H

OP

13

SCHEME 8.1

Rainier’s retrosynthetic analysis of brevenal.

Stereocenters are usually quite important in the total synthesis of complex targets, but not always. Indeed, some targets have no stereocenters, but construction of the molecule by making key bonds and incorporation of substituents and/or functional groups can be quite challenging. One example, in Scheme 8.2, is taken from Kelly and coworker’s53 synthesis of pterocellin A (14), and the first disconnection generates 15 by clipping the five-membered ring (an intramolecular coupling reaction for the synthesis). The next disconnection generates three fragments (16, 17, and 18), two pyridine fragments and a functionalized acyl fragment. Fragments 16 and 17 are combined to yield 18 52

Zhang, Y.; Rohanna, J.; Zhou, J.; Iyer, K.; Rainier, J. D. J. Am. Chem. Soc. 2011, 133, 3208.

53

O’Malley, M. M.; Damkaci, F.; Kelly, T. R. Org. Lett. 2006, 8, 2651.

430

8. SYNTHETIC STRATEGIES

as the next disconnection fragment, which can be prepared from the commercially available kojic acid (19). The fragment 20 is prepared from the commercially available 2-bromo-3-pyridinol, 21. Note that 14 has no stereocenters, but the synthesis must include making the key bonds, as well as positioning the heteroatoms and substituents at the proper positions.

16 14

15 20

O O

N

MeO

18 17

R1 O

19 21

O MeO

14

N

MeO

MeO

O O

N

16

O

N

CHO

R1 O

OH

HO 19 (Kojic acid)

18 O

N

Y

X

NH TfO

17

15

N

Br

OH

N

RO 20

21

53

SCHEME 8.2 Kelly and coworker’s retrosynthetic analysis of pterocellin A (14).

8.3.3 Consecutive Versus Convergent Syntheses Any synthetic strategy must describe how to proceed from starting material to product. One can work backwards linearly from the target to a starting material. Alternatively, disconnection can branch to several fragments, each with its own starting material. These fragments can be combined in a nonlinear fashion to prepare the target. A synthesis based on a linear pathway is called a consecutive synthesis, and a synthesis based on a branching pathway is called a convergent synthesis. In a convergent synthesis the several pieces of a molecule are synthesized individually, and the final target is sequentially assembled from the pieces. This approach differs from a consecutive synthesis, in which the target is assembled step by step from a single starting material until the final target is reached. These two approaches are illustrated in Fig. 8.3, which was used by Velluz et al.54 to describe their synthetic approach. It appears that the convergent synthesis should have fewer steps in the overall synthesis. If the largest sequence is split and divided into convergent solutions, there would be only four consecutive steps (A1 ! B1 ! C1 ! D1 ! Z) in a perfectly convergent pathway. It is difficult to achieve such a perfect convergence since functional group manipulation, steric considerations, and asymmetric structural features lead to “imperfections.”54

Consecutive synthesis

Target

Convergent synthesis

FIG. 8.3

54

Consecutive versus convergent syntheses.

Velluz, L.; Valls, J.; Mathieu, J. Angew. Chem. Int. Ed. Engl. 1967, 6, 778.

431

8.4 SYNTHETIC STRATEGIES

Massanet and coworker’s51 retrosynthetic analysis of thapsigargin in Fig. 8.2 is one example of a consecutive synthesis. Although it is not structurally complex, this consecutive synthesis illustrates the fundamental idea of beginning with a single compound and transforming it directly to the target. Both Ranier and coworker’s52 synthesis of brevenal (Scheme 8.1), and Kelly and coworker’s53 synthesis of pterocellin A (Scheme 8.2) illustrate convergent syntheses. Disconnection to separate fragments requires that each fragment be synthesized independently, and then those fragments are joined in one or more key steps to complete the synthesis. This fact is quite clear in the retrosynthetic analysis presented for 14. The reader is referred to the cited papers for more details of these examples. Both consecutive and convergent strategies can be attractive, depending on the target and only careful consideration of each will yield the best approach for an individual target. Obviously, once it is completed a given synthesis can be continually reexamined and, in principle, improved upon. In general, a convergent strategy is expected to produce larger amounts of the target in the fewest steps if a scheme can be devised that converges at a useful point. A symmetrical convergent strategy is often the most attractive55 (also see Section 8.11.1).

8.4 SYNTHETIC STRATEGIES 8.4.1 Defining Strategic Approaches Any total synthesis of a complex target requires planning and a workable strategy to determine what steps are required and what reagents must be used.56 Over many years, E.J. Corey et al.57 described an elegant approach to the synthesis of complex targets that was incorporated in the computer program LHASA (Section 8.4.2). The purpose of that program is to assist the analysis of complex synthetic targets.14,46,57 The monograph by Corey and Cheng1a describes approaches for constructing a synthetic tree, and it also evaluates the logic used for that construction. This approach includes:14 (1) interconversion, removal, or introduction of functional groups, (2) extension of the atomic chains or appendages, (3) generation of atomic rings, (4) rearrangement of chain or ring members, and (5) cleavage of chains or rings. Each chemical fragment in the synthetic tree must possess chemical properties that are predictable and allow selective combination in only one of many possible modes.1a,14 D

O B O

H

C

O E O

O

N O

HO2C

E

O

O

Isobenzofuran-1,3-dione

22

A

D C

A,B

O

O

H2N X

X

OH

2-Aminoethan-1-ol

HO Buta-1,3-diene

23

24

Corey and Wipke14 described three methodologies for synthesis: direct associative, intermediate associative, and logic centered. The direct associative approach disconnects the target (22)14 at bonds that generate a structure easily recognized by the chemist. Disconnection of bonds A and B suggested a Diels-Alder reaction (Section 14.5) from buta1,3-diene and 23. Hydrolysis of the ester (bond C) and the amide (bond D) led to fragment 24 and the fragment derived from the Diels-Alder reaction. Disconnection of bonds D and E in compound 22 led to isobenzofuran-1,3-dione (phthalic anhydride) and 2-aminoethan-1-ol. The intermediate associative approach recognizes a major subunit in the target that corresponds to a known, or potentially available starting material. The choice of this starting material directs the retrosynthetic analysis along a specific pathway.14 The scope and rigor of solutions to the problem are sometimes very limited, however.14 Sarett’s58 synthesis of 55

Bertz, S. A. J. Chem. Soc. Chem. Commun. 1984, 218.

56

See Tatsuta, K.; Hosokawa, S. Chem. Rev. 2005, 105, 4707.

57

Corey, E. J.; Long, A. K.; Rubenstein, S. D. Science 1985, 228, 408.

58

(a) Sarett, L. H. J. Biol. Chem. 1946, 162, 601. (b) Fieser, L. F. Fieser, M. Steroids; Van Nostrand Reinhold: New York, 1959; pp 640–650.

432

8. SYNTHETIC STRATEGIES

cortisone used deoxycholic acid as a starting material, taken from Kendall and coworker’s59 previous synthesis of 11-keto-etiolithocholic acid. This earlier synthesis required a total of 30 steps, but Sarett’s second synthesis was shorter, and introduced a general method for introduction of the 17α-hydroxyl group.60 HO Me

O

OH Me

O

CO2H

OH

Me

Me

H H

H H

H

H

HO

O Cortisone

Deoxycholic acid

The logic centered approach leads to a restricted set of structures that may be converted in a single step to the target. Each new structure can, in turn, be converted to a new structure. The purpose of the retrosynthetic analysis is to generate synthetic intermediates (the synthesis tree) that terminate with a number of starting materials.57 Corey et al. outlined this approach in 1964, in the synthesis of longifolene, which is shown in Scheme 8.3.61 This synthesis used the Wieland-Miescher ketone as the starting material.62 Analyses of targets (e.g., longifolene) begin with a search for key structural features that may be structurally significant. These key features are14 (1) individual molecular chains, rings, and appendages, (2) individual functional groups, (3) asymmetric centers and attached groups, and (4) chemical reactivity.

Me

O

O

O 1. Ethylene glycol , PhH, p-TsOH, Reflux

O

OsO4 Ether/Py

HO

Me O

2. Ph3PEt, BuLi Ether, 25°C 3. THF, Reflux

Me

O

1. p-TsCl, Py, CH2Cl2 2. Ice 3. LiClO4, THF, 2.5 d CaCO3, 50°C 4. 2N HCl, EtOH, 100°C

O

Me

Me OH

(63%)

Wieland-Miescher ketone

O

Me

NEt3, Ethylene glycol, Sealed tube 225°C, 24 h

O

Me

O

H

Me

Me (20%)

(48%) 1. Ph3CNa, MeI, Ether/Dioxane 2. BF3 3. LiAlH4, Ether 4. NH2NH2, Na, Ethylene glycol , 195°C 5. CrO3, AcOH

Me

Me 1. MeLi, Ether, Reflux, 3 d 2. H2O

O

Me

Me

3. Freon 11, Py SOCl2, 0°C

Me (39%)

Me Longifolene (88%)

SCHEME 8.3 Corey’s synthesis of longifolene.

An important strategy for a retrosynthesis is to reduce molecular complexity, which is accomplished by meeting the following goals:14 (1) simplification of internal connectivity by scission of rings, (2) reduction of molecular size by 59

McKenzie, B. F.; Mattox, V. R.; Engel, L. L.; Kendall, E. C. J. Biol. Chem. 1948, 173, 271.

60

(a) Sarett, L. H. J. Am. Chem. Soc. 1948, 70, 1454. (b) Sarett, L. H. Ibid. 1949, 71, 2443.

61

Corey, E. J.; Ohno, M.; Mitra, R. B.; Vatakencherry, P. A. J. Am. Chem. Soc. 1964, 86, 478. (b) Corey, E. J.; Mitra, R. B.; Uda, H. Ibid. 1964, 86, 485.

62

Wieland, P.; Miescher, K. Helv. Chim. Acta 1950, 33, 2215.

8.4 SYNTHETIC STRATEGIES

433

disconnection of chains or appendages, (3) removal of functionality, (4) modification or removal of sites of unusually high chemical reactivity of instability, and (5) simplification of stereochemistry. There are also certain transformations that are important adjuncts to these goals, but do not simplify the molecule. Examples of these transformations follow:14 (1) functional group interchanges, (2) introduction of functional groups, (3) modification of functional groups to control the level of chemical reactivity, (4) introduction of groups that permit stereochemical or positional control, and (5) internal rearrangement. Above all, the disconnections must lead to chemical reactions that will allow construction of the target. The basic mechanistic aspects of these reactions must be well understood, since the reaction pathway is the same in the forward and reverse directions (principle of microscopic reversibility).14 If a set of disconnections is found to correlate with a mechanistically sound set of chemical reactions in the forward directions, those steps are taken as part of the synthetic tree. In all steps, stereochemical and topological aspects must be considered. In an analysis of a target, the vast body of organic chemistry may lead to the generation of retrosynthetic trees that are too large to be manageable, and include numerous dead ends.57 It is essential that simplification techniques, search strategies, screening procedures, and logical analysis be employed as the tree is constructed. If subunits of the molecule that are potential synthons (the disconnection products often taken as the chemical building blocks) can be found, a strong measure of control over tree branching is possible.48,57 The retrosynthetic transform and synthetic reaction to accomplish that transform are intricately linked. Just as the power of a transform to simplify a structure is critically important, so the most powerful synthetic reactions are those that reliably increase molecular complexity.57 Transforms that decrease or increase the molecular complexity of a target are important in the retrosynthetic strategy. Bertz63 described the importance of molecular complexity in the context of synthetic analysis, particularly as applied to convergent syntheses, a topic also addressed by Merrifield and Simmons.63d Bertz63a suggested that “topological complexity, as well as other sources of molecular complexity, be examined in addition to the classical considerations of synthetic efficiency when evaluating synthetic routes.” From the preceding discussions it is clear that there are many approaches for a disconnection approach that constructs a synthesis tree. Corey et al.57 described five different strategies for retrosynthetic analysis: (1) Transform based strategies. Identify a powerful transform for a specific target that produces a reasonable line in the synthetic tree. Two or three powerful simplifying transforms can be applied successively to the target structure. Alternatively, the same simplifying transforms can be repeatedly applied.64 (2) Structure goal strategies. Identify a potential starting material, building block, retrosynthetic subunit, or initiating chiral element (see the chiral template approach in Section 8.9). (3) Topological strategies. Identify one or more bonds whose disconnection can lead to major molecular simplification, which is the essence of the strategic bond approach65 to be discussed in Section 8.5. These bonds may be in bridged- or fused-ring cyclic systems or appear as connectors to appendages at rings, functional groups, or stereocenters. (4) Stereochemical strategies. Heuristically derived procedures for reducing stereochemical complexity in the retrosynthetic direction, these procedures remove stereocenters and take advantage of steric screening or proximity to a functional group. The latter may allow simplification of the target by application of functional group removal transforms.66 Corey et al.47b defines heuristic as “a noun to mean heuristic principle, a ‘rule-of-thumb,’ which may lead by a shortcut to the solution of a problem or may lead to a blind alley.” (5) Functional group oriented strategies. When one or more functional groups are related to an interconnecting atom path, a simplifying transform can often be found. Interconversion, addition, or removal of functionality can pave the way for further retrosynthetic simplification. Both internal protection of functional groups and replacement of a highly reactive group by another less reactive equivalent can be useful.67

8.4.2 LHASA Corey and coworkers64–67 developed a computer program that uses the strategies described above in a way that emulates the most effective problem-solving approaches of a synthetic chemist, with an emphasis on complex rather (a) Bertz, S. H. J. Am. Chem. Soc. 1982, 104, 5801. (b) Idem Ibid, 1981, 103, 3599. (c) Idem J. Chem. Soc. Chem. Commun. 1981, 818. (d) Merrifield, R. E.; Simmons, H. E. Proc. Natl. Acad. Sci. USA 1981, 78, 1329, 692.

63

64

Corey, E. J.; Long, A. K. J. Org. Chem. 1978, 43, 2208.

65

Corey, E. J.; Howe, W. J.; Orf, H. W.; Pensak, D. A.; Petersson, G. J. Am. Chem. Soc. 1975, 97, 6116.

66

For a synthetic example, see Corey, E. J.; Arnett, J. F.; Widiger, G. N. J. Am. Chem. Soc. 1975, 97, 430.

67

Corey, E. J.; Orf, H. W.; Pensak, D. A. J. Am. Chem. Soc. 1976, 98, 210.

434

8. SYNTHETIC STRATEGIES

than routine synthetic problems. This computer program is called LHASA. It analyzes a molecule, develops retrosynthetic schemes, and supplies appropriate chemical reactions. Wipke and Howe46 provided a useful description of this program: One important aspect of the project has been the writing of a general purpose computer program which will aid the laboratory chemist and will employ both the basic and more complex techniques for synthetic design as elucidated by this study. The program (hereafter also called LHASA) is intended to propose a variety of synthetic routes to whatever molecule it is given. The responsibility for final evaluation of the merit of the routes lies with the chemist. The program is to be an adjunct to the laboratory chemist as much as any analytical tool.46 LHASA is an interactive program that displays the target, allowing the chemist to select a synthetic strategy (or strategies). The program then suggests synthetic routes that satisfy the goals of the selected strategy.57 The analysis generates retrosynthetic precursors and generates a synthetic tree (see Fig. 8.4).57 The chemist takes responsibility for choosing strategies and tactics and for deciding which precursors should be submitted to LHASA for further simplifications. The actual transforms are selected by LHASA, based on a database of known reactions. o-Hydroxylation of a phenol with an aldehyde

HO

HO

HO

HO

HO

HO

OH

O

Target 3

O

OH

Node 3

Prototype conditions: AlCl3 at 25°C

(A)

1

14

8

4

24 6

5

20 19

15

25

21

28

17 13

(C)

12

10

31 11

Br HO

HO

16 7 9

Organometallic addition to carbonyl

HO

27

23

O Node 7 Acetophenone

26

22

18

Hydride reduction

Node 5

(B) 3

OH

Electrophilic thallation then displacement

2

29

OH

30

(D)

Node 4

OH

Target 3

FIG. 8.4 Portions of a LHASA transform and retrosynthetic analysis. (A) Display of LHASA transform. (B) Retrosynthesis of Target 3 to acetophenone. (C) Part of the tree from the retrosynthetic analysis of target. (D) Retrosynthetic removal of the hydroxymethyl group from Target 3 by a route involving an organometallic addition transform. Reprinted from Johnson, A. P.; Marshall, C.; Judson, P. N. J. Chem. Inf. Comput. Sci. 1992, 32, 411. Copyright © 1992 American Chemical Society.

Portions of a typical LHASA analysis are shown in Fig. 8.4.68 A LHASA transform is shown in part (A), in this case based on the Fries rearrangement (see Section 16.4.4). In part (B), a target molecule is shown (Target 3), and a retrosynthesis is provided that is based toward acetophenone. Note that a description of the projected chemistry is shown over each disconnect arrow. In part (C), a retrosynthetic tree is shown as part of the analysis of a target, where the numbers represent chemical structures generated by LHASA. In part (D), a specific retrosynthetic analysis is shown in which one unit, the hydroxymethyl group, is removed from Target 3, with the projected reaction being an organometallic transform. The data provided by Fig. 8.4 is meant to illustrate the type of information that is available from the LHASA program. A sample retrosynthesis for aphidicolin is shown in Fig. 8.5, taken from the work of Corey et al.69

68 (a) Johnson, A. P.; Marshall, C.; Judson, P. N. J. Chem. Inf. Comput. Sci. 1992, 32, 411. Also see (b) Johnson, A. P.; Marshall, C. J. Chem. Inf. Comput. Sci. 1992, 32, 418. 69

Corey, E. J.; Long, A. K.; Lotto, G. I.; Rubenstein, S. D. Recl. Trav. Chim. Pays-Bas 1992, 111, 304.

435

8.5 THE STRATEGIC BOND APPROACH

HO

OH

HO

OH

RO

OR

H

O

H

O H

H

X

H

HO

H

HO H

H

H HO

Aphidocolin

O

X RO

H

RO H

H

H

TMSO

TMSO

RO H

HO

RO

H

H

H RO

RO

RO O

O

O

O H

RO

H

O

O

OR

O

RO

O

O

O

O

H

O

O

RO H

RO

H RO

H

H RO

FIG. 8.5

A LHASA generated retrosynthetic pathway for the synthesis of aphidicolin. Reproduced with permission from Corey, E. J.; Long, A. K.; Lotto, G. I.; Rubenstein, S. D. Recl. Trav. Chim. Pays-Bas 1992, 111, 304. Copyright © Wiley-VCH Verlag GmbH & Co. KGaA.

Note that a variation of this computer technique has been developed, with the goal of synthesis simplification. “Mathematics Applied to Synthetic Analysis (MASA) is a useful addition to LHASA, as it can be used to calculate the simplification afforded by alternative disconnections of a target molecule. One- and two-bond disconnections that are more efficient, as far as simplification is concerned, than those identified by the LHASA criteria have been found by using MASA.”70 This approach is based on identifying topological strategic bonds.

8.5 THE STRATEGIC BOND APPROACH Several available strategies can be applied to the specific task of breaking the first (and subsequent) bond(s) in the target. For the student beginning a study of synthesis, a more formalized approach would be useful. Several rules are available that tie together the strategies and goals of retrosynthetic analysis to generate a strategic bond analysis. These rules provide insight into the first disconnection and the complete synthetic tree. The analysis begins with an inspection of the target molecule to determine if the structure is compatible with simple solutions. Before the strategic bonds are determined, simplification may be possible if the molecule possesses symmetry, is structurally similar to another molecule that has been previously synthesized, or has repeating units within the structure. The following items are only guidelines. These guidelines are a reasonable place to begin planning a synthesis, but by no means the only way to approach a problem. If a disconnection is reasonable and there is either known or planned chemistry to reconstitute that bond in a synthesis, proceed.

8.5.1 The Preliminary Scan The analysis inspects the molecule to determine if it can be simplified. There are three criteria.71 8.5.1.1 Is There Symmetry or Near Symmetry in Two Parts of the Molecule? If so, the synthesis may be simplified by synthesizing and then joining together the two or more identical pieces. A good example is the molecule C-toxiferin I,72 which is essentially a dimer. In C-toxiferin, one-half of the molecule is identical to the other, and it has been shown that treatment with sulfuric acid cleaves the molecule into a single product that regenerates the dimer on treatment with hot sodium acetate.72 Synthesis of one piece and joining the two identical pieces together greatly simplifies an overall synthesis with a built-in convergence. 70

Bertz, S. H.; R€ ucker, C.; R€ ucker, G.; Sommer, T. J. Eur. J. Org. Chem. 2003, 4737.

71

Bersohn, M.; Esack, A. Chem. Rev. 1976, 76, 269.

(a) Bernauer, K.; Berlage, F.; von Philipsborn, W.; Schmid, H. Karrer, P. Helv. Chim. Acta 1958, 41, 2293. (b) Berlage, F. Bernauer, K.; von Philipsborn, W.; Waser, P.; Schmid, H. Karrer, P. Ibid. 1959, 42, 394.

72

436

8. SYNTHETIC STRATEGIES

Me

N

N

N

N

Me

C-toxiferin I

The symmetry may not be obvious and the target should be carefully analyzed. Usnic acid is an example of a molecule that possesses potential symmetry.73 The molecule may not be obviously symmetrical, but the symmetry appears upon disconnection of a key intermediate, and suggests a retro-phenolic oxidative coupling (see Section 17.8.2). The synthesis of usnic acid is shown in Scheme 8.4,74 with a phenolic oxidative coupling of a single phenol derivative [1-(2,4,6-trihydroxy-3-methylphenyl)ethan-1-one] to yield 25 via coupling of the initial oxidation product 2-acetyl-3,5dihydroxy-4-methylcyclohexa-2,4-dien-1-one. The phenolic oxidative coupling reaction (Section 17.8.2) is a radical process by which two phenol moieties are joined.75 The two phenolic moieties reacted via an acid-catalyzed Michael reaction to yield a dimer (25), which cyclized via the phenol hydroxyl group to yield 26. Treatment with concentrated sulfuric acid completed the two-step synthesis of usnic acid. Recognizing the potential symmetry greatly shortened the synthesis.

Ac Ac

Ac HO

OH

Potassium ferricyanide

HO

HO

OH

O

Me Me

Me

Me

OH HO

OH

OH 1-(2,4,6-Trihydroxy-3-methylphenyl)ethan-1-one

2-Acetyl-3,5-dihydroxy-4-methylcyclohexa-2,4-dien-1-one

O

Me

25

Ac

O

Ac

HO

O

HO

OH

O

conc H2SO 4

Me

O

Me OH

Me OH

HO

O

HO

Me 26

SCHEME 8.4

O

Me

O Me

Usnic acid

Phenolic oxidative coupling strategy for the synthesis of usnic acid.

73

Corey, E. J. Pure Appl. Chem. 1967, 14, 19.

74

Barton, D. H. R.; Deflorin, A. M.; Edwards, O. E. J. Chem. Soc. 1956, 530.

(a) Kotani, E.; Miyazaki, F.; Tobinaga, S. J. Chem. Soc. Chem. Commun. 1974, 300. (b) Barton, D. H. R.; James, R.; Kirby, G. W. Widdowson, D. A. Ibid. 1967, 266. (c) Kametani, T.; Fukumoto, K. Synthesis 1972, 657. (d) Kametani, T.; Yamaki, K. Terui, T.; Shibuya, S.; Kukumoto, K. J. Chem. Soc. Perkin 1972, I, 1513. 75

437

8.5 THE STRATEGIC BOND APPROACH

8.5.1.2 Is the Problem Like One Already Solved? The Woodward et al.11 synthesis of reserpine is shown in Scheme 8.5. Compound 27 was an important intermediate in that synthesis. Deserpidine76 is a natural product that is structurally similar to reserpine, but lacking the methoxy group on the indole moiety. It closely mimics the activity of reserpine,77 and a total synthesis of deserpidine was developed by Velluz et al.78 and later by Protiva.79b,c Intermediate 28, prepared from 27 from the Woodward synthesis, was Br O

HO

1. NaBH4 2. PhCO3H

H O HO2C

O

2. MeOH, H+

O

1. CrO3, AcOH 2. Zn, AcOH 3. CH2N2, Dioxane

O OMe

27

4. Ac2O, Py 5. OsO 4 6. KClO3

OH O

H

O

O

3. Al(Oi-Pr)3 i-PrOH Reflux

H

1. NBS, H2O, cat HOOH

H

O

OH H H

1. aq HIO4 2. CH2N2

OAc MeO2C

1. aq HIO4 2. CH2N2

OMe

3.

28

3.

MeO

NH2

N H

N

N

NH2

N H

MeO

N

N H MeO2C

H

OAc

N

MeO2C

29

H

OMe

N N H

30

OMe H

31 H MeO2C

O

OMe

Deserpidine MeO

H

1. NaOMe MeOH Heat

H

2.

OMe OMe H

N

OAc

H MeO2C

MeO2C

H

O

1. NaBH4 2. POCl3, Reflux; NaBH4, MeOH

MeO

OAc

N

MeO2C

O

MeO Cl

OMe

MeO

MeO O

MeO

N H

O

H H MeO2HC

Reseerpine

SCHEME 8.5

OMe

OMe MeO

OMe

Syntheses of reserpine and deserpidine.

(a) Wenkert, E.; Liu, L. H. Experientia 1955, 11, 302. (b) Schlittler, E.; Ulshafer, P. R.; Pandow, M. L.; Hunt, R. M.; Dorfman, L. Ibid. 1955, 11, 64. (c) Huebner, C. F.; MacPhillamy, H. B.; Schlittler, E.; St. Andre, A. F. Ibid. 1955, 11, 303. (d) MacPhillamy, H. B.; Dorfman, L.; Huebner, C. F.; Schlittler, E.; St. Andre, A. F. J. Am. Chem. Soc. 1955, 77, 1071. (e) Yamamoto, T.; Yamanaka, M. Shinyaku to Rinsho 1955, 4, 19, 175 (Chem. Abstr. 1957, 51, 14984h) and see the other articles cited in this abstract. 76

77

Lednicer, D. L.; Mitscher, L. A. Organic Chemistry of Drug Design; Wiley: New York, 1977; Vol. 1, pp 320–321.

Velluz, L.; Muller, G.; Joly, R.; Nomine, G.; Mathieu, J.; Allais, A. Warnant, J.; Valls, J.; Bucourt, R.; Jolly, J. Bull. Soc. Chim. Fr. 1958, 673.  (a) Weichet, H.; Pelz, K.; Bláha, L. Collect. Czech. Chem. Commun. 1961, 26, 1537. (b) Protiva, M.; Capek, A.; Jílek, J. O.; Kakác, B.; Tadra, M. Ibid. 1961, 26, 1537. (c) Ernest, I.; Protiva, M. Ibid. 1961, 26, 1137. (d) Protiva, M.; Rajšer, M.; Jilek, J. O. Monatsh. Chem. 1960, 91, 703.

78 79

438

8. SYNTHETIC STRATEGIES

used as the starting material to yield the alternative synthesis also shown in Scheme 8.5.79 Reaction of 28 with 6-methoxytryptamine (obtained with great difficulty) gave 29.79 Intermediate 29 was converted to 30, and then to reserpine. In the synthesis of deserpidine, 31 was prepared by the reaction of tryptamine with 28. This simple modification of Woodward’s synthesis obviously saved Velluz et al.78 the time and effort required to develop a new synthetic approach. This example suggests that thorough familiarization with the target, including knowledge of all previous syntheses of the target and also related compounds (i.e., thorough search of the literature) is always beneficial. One does not always want to mimic a literature approach, but when the target is structurally similar to one that is known, the use of a key intermediate or transformation sequence in designing a retrosynthesis is prudent and usually desirable. Obviously, if one has an improved method or route to a target, the retrosynthesis is biased to include this new method. It is common for a published total synthesis to be followed by several later syntheses that improve or expand upon it. Syntheses developed later are often useful for producing large quantities of the natural product, as the initial sequence may produce only small amounts (milligrams) of the target. 8.5.1.3 Is The Molecule a String of Available or at Least Simple Pieces? The synthesis80 of the macrolide antibiotic nonactin81 is a useful example. Nonactin consists of four units of nonactic acid,81 joined together in a tetrameric ring. A total synthesis would be greatly simplified by preparing nonactic acid, since the final synthesis of nonactin would require only macrocyclization (formation of a large lactone ring, Section 4.5.2).82 The Gerlach et al.80 synthesis of nonactin, which is shown in Scheme 8.6, utilized this approach. The synthetic precursor of nonactic acid was 32, produced in several steps from furanyl derivative 1-(furan-2-yl) O

O Cl

O

1.

, AgBF4

Me

1. CrO3

CH2Cl2 , –20°C

O

O

N

O

O

2. CH2N2

2. aq HCl

CO2Me

CHO

1-(Furan-2-yl)propan-2-one

Me

Me HO

O Rh, H2

1. NaBH4

H 2O

H O

2. aq NH4Cl

CO2Me H

H O H

H Me

32

Nonactic acid

1. 2 equiv. NaH , PhCH2Br Me

2.

H

Me

Me

Me SO2Cl

, Py

Me

O

H

O

O 3. CF3CO2H 4. Repeat 2 and 3

Me

H

O

H

O O

H

Me

Me

H

Me

O

Me

H O

H

O O

CO2Me H Me

O

H

O Me

Nonactin

SCHEME 8.6 The Gerlach et al.80 synthesis of nonactin.

80

(a) Gerlach, H.; Wetter, H. Helv. Chim. Acta 1974, 57, 2306. (b) Gerlach, H.; Oertle, K.; Thalmann, A.; Servi, S. Ibid. 1975, 58, 2036.

(a) Keller-Schierlein, W.; Gerlach, H. Fortschr. Chem. Org. Naturstoffe 1968, 26, 161. (b) Ando, K.; Oishi, H.; Hirano, S.; Okutomi, T. Suzuki, K.; Okazaki, H.; Sawada, M.; Sagawa, T. J. Antibiot. (Tokyo) 1971, 24, 347. (c) Dobler, M.; Dunitz, J. D. Kilbourn, B. T. Helv. Chim. Acta 1969, 52, 2573. (d) Dobler, M. Ibid. 1972, 55, 1371. (e) Iitaka, Y.; Sakamaki, T. Nawata, Y. Chem. Lett. 1972, 1225.

81

(a) Taub, D.; Girotra, N. N.: Hoffsommer, R. D.; Kuo, C. H.; Slates, H. L.; Weber, S.; Wendler, N. L. Tetrahedron 1968, 24, 2443. (b) Corey, E. J.; Nicolaou, K. C. J. Am. Chem. Soc. 1974, 96, 5614. (c) Masamune, S.; Kim, C. U.; Wilson, K. E.; Spessard, G. O. Georghiou, P. E.; Bates, G. S. Ibid. 1975, 97, 3512. (d) Masamune, S.; Yamamoto, H.; Kamata, S.; Fukuzawa, A. Ibid. 1975, 97, 3513.

82

439

8.5 THE STRATEGIC BOND APPROACH

propan-2-one. Note that when two remote, reactive ends of a molecule must be brought together so reaction can occur, intermolecular processes often dominate (Section 4.5) and the low yield of nonactin obtained by Gerlach et al.80 is a result of such problems, which does not diminish the simplicity of the synthetic approach that recognized nonactin as a string of repeating fragments.

8.5.2 Criteria for Disconnection of Strategic Bonds Once the preliminary scan is complete, the next step is to disconnect the molecule by breaking bonds that can be reconnected by known or reasonable chemical steps, to generate the synthetic tree. As discussed above, these important bonds are called strategic bonds. Bersohn and Esack71 restated Corey’s rules to tie together the strategies outlined in Section 8.4.47,71 One should not rigidly follow these statements, but rather treat them as excellent suggestions for how to begin planning a synthesis: 1. Labile groups should be removed first. 2. Transforms should remove as much functionality as possible and should remove stereochemical centers where possible. 3. Favor transforms that generate closely related intermediates. 4. If a certain substructure interferes with a key process, then use a transform to remove that substructure. 5. Where possible, use transforms to build bridges between the goal atoms. 6. Convert the molecule under consideration into the product of one of the powerful reactions. 7. Disconnect appendages attached to atoms bearing certain functional groups (e.g., OH or C]O). 8. Consider all pairs of functional groups in the molecule to ascertain whether known transforms can disconnect any of the intervening bonds. 9. In a ring system preferentially disconnect the strategic bonds. The examples that follow will illustrate an application of each of these rules: 1. Remove labile groups first. Removing a labile group is illustrated by a retrosynthetic step in Holmes and coworkers’83 synthesis of (
)-histrionicotoxin. The alcohol and conjugated eneyne units are quite reactive. Both units are removed in the initial disconnection and replaced with R and R0 groups, which represent functionality that can be elaborated to the conjugated system. In addition, the labile amine and alcohol units are converted to the isoxazolidine unit in 32. This disconnection allows one to focus on constructing the fundamental spirocyclic ring system, knowing that highly reactive groups will be added last.

N

N H HO (–)-Histrionicotoxin

R

R'

O 32

2. Remove stereochemical centers. The vast majority of natural products and important synthetic targets are characterized by several stereogenic centers that are often contiguous. A major task of a total synthesis is the generation of these stereogenic centers, in a controlled and predictable manner. It is therefore important to plan all reactions, and often the entire retrosynthesis around stereoselective reactions. Chemical reactions that involve formation of bonds at a chiral carbon allow a measure of control over stereochemical induction. The stereogenic center and the reaction center usually have a 1,2-, 1,3-, or occasionally a 1,4-relationship. It is also possible that a remote center is held in close proximity due to a particular conformation, and exerts an influence on the reaction. Molecular or computer modeling is usually necessary to observe the correct influence of remote centers on a disconnection. Disconnections are, in general, made at a bond connected to a stereogenic center (1,2-relationship) rather than to a bond containing a pendant group that contains a stereogenic center. In Paquette and coworker’s84 synthesis of magellanine, disconnection of the piperidine ring removed two stereogenic centers, although the allylic 83

Williams, G. M.; Roughley, S. D.; Davies, J. E.; Holmes, A. B. J. Am. Chem. Soc. 1999, 121, 4900.

84

Williams, J. P.; St. Laurent, D. R.; Friedrich, D.; Pinard, E.; Roden, B. A.; Paquette, L. A. J. Am. Chem. Soc. 1994, 116, 4689.

440

8. SYNTHETIC STRATEGIES

alcohol unit introduced another (see 33). The next disconnection removed the cyclohexenol ring along with two stereogenic centers, and the use of a protected carbonyl (the dithiane) effectively removed the stereogenic center at the secondary alcohol, giving 34. OH

Me

O

O O

S

H Me

N

H H

H H OH Magellanine

S

OH

33

34

3. Favor closely related intermediates. It is usually preferable to simplify the target molecule in small bites (i.e., prune small branches of the synthesis tree rather than large branches). Moving too far down the synthetic tree with large disconnections will mean that several carbon-carbon bond-forming reactions or functional group exchanges are required to convert one disconnect product to another. It may be difficult to find transforms where several bonds can be formed simultaneously in the synthesis, although such transformations do exist in some cases. For a first analysis, simplification should be done so that a minimum number of carbon-carbon bond-forming reactions and functional group exchanges are required for each disconnect product. In other words, disconnect one or two bonds at a time. In this way, it is easy to understand the simplification and the implications for the requisite chemical reactions, which is illustrated by the relatively simple disconnection sequence used by Kibayashi and coworkers85 in a synthesis of (
)-(3R,6S,9R)-decarestrictine C2, 35. The target was first converted to the seco-acid [(3R,6S,9R,E)-3,6,9-trihydroxydec-4-enoic acid], removing the labile ester unit, and disconnection to aldehyde [(4S,7R,E)-4,7-dihydroxyoct-2-enal] suggests a condensation reaction to make that bond. The last disconnection leads to (2S,5S)-hexane-1,2,5,6-tetraol, again suggesting a condensation reaction. In each case, one key bond is disconnected and the differences between target and disconnect product are structurally close and predictable. Generally, one disconnects one or two bonds per disconnection, but more may be disconnected if the pieces are closely related. Me

Me

O

OH

O

CO2H

OH

OH

OH

OH 35

(3R,6S,9 R, E)-3,6,9Trihydroxydec-4-enoic acid

OH CHO

OH

OH

OH

Me

OH (4S,7 R, E)-4,7Dihydroxyoct-2-enal

OH (2S,5S)-Hexane1,2,5,6-tetraol

4. Remove interfering substructures. An example provided by et al.67 is the disconnection of 3,4,4a,5,8,8a-hexahydro1,4-methanonaphthalen-2(1H)-one to 36 or to 37. Path a is a simple epoxidation (Section 6.4), and path b is a Baeyer-Villiger oxidation (Section 6.6). A cyclic cis-alkene moiety [e.g., as in 3,4,4a,5,8,8a-hexahydro-1,4methanonaphthalen-2(1H)-one] shows moderate reactivity with peroxyacids, but the ketone moiety is very reactive, since it is strained. Disconnection of 37 to 3,4,4a,5,8,8a-hexahydro-1,4-methanonaphthalen-2(1H)-one is viable, since the synthesis will proceed without interference from the alkene moiety. Disconnection of 36, however, requires that the interfering carbonyl be removed in any synthetic attempt. The best retrosynthesis of 3,4,4a,5,8,8a-hexahydro-1,4-methanonaphthalen-2(1H)-one will probably involve protection of the ketone (Section 5.3.3), as a ketal for example. In Fig. 8.4,68 the LHASA program was used to generate a retrosynthetic sequence with boxes around reactive or interfering functional groups that can be protected and dashed boxes around unstable (interfering) groups that are not easily protected. Quite often, this rule refers to protection of interfering structure.

85

Arai, M.; Morita, N.; Aoyagi, S.; Kibayashi, C. Tetrahedron Lett. 2000, 41, 1199.

441

8.5 THE STRATEGIC BOND APPROACH

O

O

a

O

b

O

O

3,4,4a,5,8,8a-Hexahydro-1,4methanonaphthalen-2(1H)-one

36

37

5. Use transforms to bridge goal atoms. If a disconnection results in cleavage of the target into separate pieces, chemical reconnection usually requires formation of one to four bonds. Two-bond disconnections are well known, although there are not many of them, and if three or four bonds must be formed, it is usually very difficult to rejoin all the bonds in one synthetic step. The Diels-Alder cycloaddition (Section 14.5.1) is an example in which two bonds are connected. If possible, it is useful to bridge the two pieces together by formation of one of the requisite bonds in the retrosynthesis, bringing the two pieces close enough to allow subsequent formation of the other bond(s). An illustration of this principle is Trost et al.86 retrosynthetic analysis of ibogamine in Scheme 8.7 (( represents a retrosynthetic disconnection and ! represents the synthetic step). The first disconnection of the nine-membered ring gave 41 in which two potential fragments are attached. The proximity of these fragments via the connecting bridge allows an usually difficult ninemembered ring to be formed in the synthesis. The next disconnection gave 40, which was the plan to generate the azabicyclic system. Disconnection to 40 led to two fragments, tryptamine and 39, recognized as precursors to the starting diene, 38. In the actual synthesis (also in Scheme 8.7), a Diels-Alder reaction of 38 and acrolein led to 39, and subsequent with tryptamine and reduction gave 40. This sequence bridged the indole and the azabicyclo[2.2.2]octane moieties, allowing formation of the azabicyclic moiety in 41. The organopalladium reagent used for this coupling will be discussed in Section 18.4. Final joining of the bridged pieces gave ibogamine, completing the total synthesis. O

O R

O

O R

O

O

CHO CHO

NH2

1.

N N

N H

H

PhMe, MgSO4, –10°C

BF3•OEt2

Et

H R

Et 38

2. NaBH4, MeOH, 0°C

Et 40

39 (90%)

N

N 5% (Ph3P)4Pd

Et

N

MeCN, 70°C

H

H 41 (45%)

SCHEME 8.7

The Trost et al.

86

1. AgBF4, MeCN, 1 h (MeCN)2PdCl2, rt 2. NaBH4, MeOH 75°C - 12 h 0°C - 1 h

Et

N H

H

Ibogamine (45%)

synthesis of ibogamine.

6. Convert the molecule into the product of one of the powerful reactions. Powerful reactions have been defined as those synthetic reactions that reliably increase molecular complexity. Disconnections that are based on powerful reactions will, therefore, greatly decrease the molecular complexity of the target, which is, of course, a major goal of a disconnection strategy. Corey et al.57 listed several examples of such reactions. Synthetic examples of these reactions are shown in Table 8.1.57 The intermolecular Diels-Alder reaction (Section 14.5.1) is shown first. Example 1 from Table 8.1 is a carba-DielsAlder reaction taken from the catalytic enantioselective synthesis of ()-przewalskin B by Xie and coworkers87 in 89% yield (94 %ee). Intermolecular and intramolecular versions of this important reaction are known. Example 2 is a

86

Trost, B. M.; Godleski, S. A.; Genet, J. P. J. Am. Chem. Soc. 1978, 100, 3930.

87

Xiao, M.; Wei, L.; Li, L.; Xie, Z. J. Org. Chem. 2014, 79, 2746.

442

8. SYNTHETIC STRATEGIES

quinone Diels-Alder reaction taken from Moody and coworker’s88 synthesis of hygrocin B in which Diels-alder cycloaddition gave the product shown in 82% yield. Example 3 is a hetero-Diels-Alder reaction (Section 14.7), taken from a synthesis of aphadilactones by Li and coworkers.89 A Lewis acid-catalyzed cyclization of an aldehyde, using a chiral Cr-salen catalyst, with the functionalized diene shown gave the dihydropyran product in 84% yield and 96 %ee. Example 4 is a Robinson annulation (Section 13.7.3) of 2-allylcyclohexan-1-3-dione and methyl vinyl ketone in a synthesis of (
)-anominine by Bonjoch and coworkers.90 TABLE 8.1 Powerful Reactions 1. Carbocyclic Diels-Alder

O

O

O

CHO

O

Neat, 40°C

Me

CHO

(89%)

2. Quinone and related Diels-Alder

O

OMe

O

O

H N

Me

Me

PhMe, SiO2

+

HO O Me

CO2Et

O Me (82%)

3. Heteroatom Diels-Alder Chiral Cr catal

+

O

MS 4Å, rt, 16 h

O OMe (84%, 96 %ee)

4. Robinson annulation

O

O

, 1% NEt3

1. O

O

2. Neat, Chiral organocatalyst 25% PhCO2H

O (91%, 94 %ee)

5. Position selective aromatic reduction

Li, NH3, EtOH

OH

OH (99%)

6. Cation π-cyclization

O

Air

TESO

OMe

H N

OMOM

OCH3 O

BF3•OEt 2

O

HO H3CO

88

Nawrat, C. C.; Kitson, R. R. A.; Moody, C. J. Org. Lett. 2014, 16, 1896.

89

Yin, J.-P.; Gu, M.; Li, Y.; Nan, F.-J. J. Org. Chem. 2014, 79, 6294.

90

Bradshaw, B.; Etxebarria-Jardí, G.; Bonjoch, J. J. Am. Chem. Soc. 2010, 132, 5966.

H (54%)

OCH3

CO2Et

443

8.5 THE STRATEGIC BOND APPROACH

TABLE 8.1 Powerful Reactions—cont’d 7. Radical π-cyclization

Br

H

H BEt3, (TMS) 2SiH PhH, 25°C

O

TBSO

O

O

TBSO

O

(68%) + (27%) of an isomer

8. Aldol condensation

O

O

OH

LDA, MeCHO

O

O

THF, –78°C

Si

Si

OTBS

OTBS (>90%)

9. Internal SN2 Cyclization

OH

O OMs

O

OH OH

Pb(OAc)4 PhH, Reflux

OMs

O

K2CO3 MeOH rt

O

O

O (80%)

10. Friedel-crafts type cyclization

OTBDPS

OTBDPS TFAA, CH2Cl2 0°C, 2 h

CO2H

O

O

O

11. Internal Ene reaction

(87%)

CO2Me BnO

Me

N

CO2Me

CO2Me CO2Me

5% In(OTf)3, Toluene

OBn

4.4% DBU, Reflux

N

O

Me

O (59%)

12. Cationic rearrangement

X–

CH(OMe)2 OPh

OMe

OMe

TiCl4, DCM –78 to –20°C

OPh

OPh

OTMS

OTMS

OTMS

OTBDPS

OTBDPS

OTBDPS OMe

OMe

OPh

OTMS

OPh

O OTBDPS

OTBDPS (70%)

Continued

444 TABLE 8.1

8. SYNTHETIC STRATEGIES

Powerful Reactions—cont’d

13. Photocyclization

Cl

O

Cl

O H N

H

hv, Li2CO3

O

H N

O

MeCN/H2O, rt

HO

O

O

O

N H

N H (40%)

14. Fischer indole synthesis

H O

Ns

H

N

N

PhNHNH2, TFA

Ns N

DCE, 40°C

H H

H

H

H

(74%)

15. Knorr and Paal-Knorr pyrrole synthesis

HN

O

CHO

Bu

NH4OAc

Bu (>70%)

16. [m,n]-Sigmatropic rearrangement (anionic Oxy-Cope)

KH, THF, Reflux 18-Crown-6, 12 h

OH (89%)

17. [m,n]-Sigmatropic rearrangement (Claisen)

CO2i-Pr

CO2i-Pr Chiral Cu-Salen cat

O

O

DCE rt, 16 h

(97%, 98 %ee)

18. Ring-closing metathesis

OTBS

OTBS OTES

OTES Grubbs’ cat II

O

Toluene, 110°C, 4 d

O

BnO

BnO C 5H11 O

C 5H11 O (Good yield)

19. Heck reaction and other palladium-catalyzed BnO BnO coupling reactions H3C

OAc O

OCH2OEt

O O

BnO BnO H3C

OAc O

O OCH2OEt

O

DMA, KOAc

I

OMe

cat P(PPh3)2Cl2

OMe OBn

OMe

OBn

OMe (64%)

The listing of reaction types 1–19 is adapted from Corey, E. J.; Long, A. K.; Rubenstein, S. D. Science 1985, 228, 408 by courtesy of Professor E. J. Corey. Reprinted with permission from AAAS. The specific examples are referenced.

8.5 THE STRATEGIC BOND APPROACH

445

Example 5 is taken from the Shvartsbart and Smith synthesis of (
)-calyciphylline N,91 in which Birch reduction (see Section 7.11.5) of the aromatic ring gave the cyclohexadiene in 99% yield. Example 6 is a cationic π-cyclization in a synthesis of (+)-schweinfurthin A reported by Wiemer and coworkers,92 in which an oxonium ion is generated by reaction with the Lewis acid, allowing attack by the alkene moiety to generate the tetrahydropyran unit via the cyclization reaction. In Example 7, a radical π-cyclization (Section 17.7) using triethylborane in the presence of di(trimethylsilyl)silane, reported by Zhai and coworkers93 in a synthesis of ()-sculponeatin N, generated the bicyclic product in 68% yield. Example 8 is an aldol condensation (Section 13.3.1) reported by Smith and Shvartsbart in a synthesis of (
)-calyciphylline.91 Fukuyama and coworkers94 synthesis of (
)-anisatin used an intramolecular SN2 displacement (Section 3.2.1) of the primary methanesulfonate by the phenolic oxygen nucleophile in Example 9. This sequence deprotected the dioxolane unit with lead tetraacetate to liberate the catechol intermediate shown, which immediately reacted with methanolic potassium carbonate to generate the phenolic anion, and the intramolecular SN2 reaction formed the pyran ring. Example 10 shows a Friedel-Crafts acylation reaction (Section 16.4.4) in which the carboxylic acid cyclized to the furan ring by treatment with trifluoroacetic anhydride to give ketone in 87% yield as part of Boukouvalas and Loach’s95 synthesis of (
)-auxofuran. Example 11 is an internal ene reaction (Section 15.6) with an alkyne taken from the Hatakeyama and coworker’s96 synthesis of oxazolomycin A, which gave the indicated product in 59% yield. In Example 12, Overman and coworker’s97 treated the dimethyl acetal-alkene starting material with several Lewis acids, but TiCl4 gave the best yield of the rearrangement product, formed via a semi-pinacol rearrangement, as part of a synthesis of (+)-sieboldine. Jia and coworker’s98 used a Witcop-type photocyclization99 (Sections 17.4 and 17.7) in a synthesis of ()-decursivine, as shown in Example 13. Garg and coworker’s100 synthesis of the Akuammiline alkaloid picrinine provides an illustration of the Fischer indole synthesis (Section 16.5.5), shown in Example 14. In Example 15, Thomson and coworker’s101 used a Paal-Knorr pyrrole synthesis102 procedure in the synthesis of streptorubin B. An anionic oxy-Cope rearrangement (Section 15.5.4.2) is shown in Example 16 to illustrate sigmatropic rearrangement processes (Section 15.5). This example is taken from the Maier and coworker’s103 synthesis of (
)-9-deoxyenglerin A. Example 17 illustrates another sigmatropic rearrangement, the Claisen rearrangement (Section 15.5.5), used by Hiersemann and coworkers104 in a synthesis of (+)-9,10-dihydroecklonialactone B. Increased interest in the ring-closing metathesis reaction (Section 18.6) is illustrated by Example 18 in the Goswami and coworker’s105 synthesis of cytospolide P. The ability to generate two alkene units in different parts of the molecule and then use them to form a ring certainly makes this reaction worthy of addition to the list of powerful reactions. Finally, Pd catalyzed coupling reactions have become extremely important in organic synthesis. The Heck reaction, Stille coupling, and Suzuki coupling are important variations that will be discussed in Section 18.4.1. Example 19 shows a Heck coupling, taken from a synthesis of polycarcin V by Minehan and coworkers.106 The reactions in Table 8.1 are classified as powerful due to the variety of transformations (formation of rings, molecular reorganization, generation of reactive functional groups from relatively unreactive 91

Shvartsbart, A.; Smith, III, A. B. J. Am. Chem. Soc. 2014, 136, 870.

92

Topczewski, J. J.; Kodet, J. G.; Wiemer, D. F. J. Org. Chem. 2011, 76, 909.

93

Pan, Z.; Zheng, C.; Wang, H.; Chen, Y.; Li, Y.; Cheng, B.; Zhai, H. Org. Lett. 2014, 16, 216.

94

Ogura, A.; Yamada, K.; Yokoshima, S.; Fukuyama, T. Org. Lett. 2012, 14, 1632.

95

Boukouvalas, J.; Loach, R. P. Org. Lett. 2013, 15, 4912.

96

Eto, K.; Yoshino, M.; Takahashi, K.; Ishihara, J.; Hatakeyama, S. Org. Lett. 2011, 13, 5398.

97

Canham, S. M.; France, D. J.; Overman, L. E. J. Org. Chem. 2013, 78, 9.

98

Qin, H.; Xu, Z.; Cui, Y.; Jia, Y. Angew. Chem. Int. Ed. 2011, 50, 4447.

(a) Yonemitsu, O.; Cerutti, P.; Witkop, B. J. Am. Chem. Soc. 1966, 88, 3941. (b) Theuns, H. G.; Lenting, H. B. M.; Salemink, C. A.; Tanaka, H.; Shibata, M. S.; Ito, K.; Lousberg, R. J. J. Heterocycles 1984, 22, 2007. 99

100

Smith, J. M.; Moreno, J.; Boal, B. W.; Garg, N. K. J. Am. Chem. Soc. 2014, 136, 4504.

101

Hu, D. X.; Clift, M. D.; Lazarski, K. E.; Thomson, R. J. J. Am. Chem. Soc. 2011, 133, 1799.

(a) Paal, C. Berichte 1884, 17, 2756. (b) Knorr, L. Berichte 1884, 17, 2863. (c) Amarnath, V.; Amarnath, K. J. Org. Chem. 1995, 60, 301. (d) Knorr, L. Berichte, 1884, 17, 1635. (e) Knorr, L. Annalen 1886, 236, 290. (f) Knorr, L.; Lange, H. Berichte, 1902, 35, 2998. 102

103

Ushakov, D. B.; Navickas, V.; Str€ obele, M.; Maichle-M€ ossmer, C.; Sasse, F.; Maier, M. E. Org. Lett. 2011, 13, 2090.

104

Becker, J.; Butt, L.; von Kiedrowski, V.; Mischler, E.; Quentin, F.; Hiersemann, M. Org. Lett. 2013, 15, 5982.

105

Chatterjee, S.; Guchhait, S.; Goswami, R. K. J. Org. Chem. 2014, 79, 7689.

106

Cai, X.; Ng, K.; Panesar, H.; Moon, S.-J.; Paredes, M.; Ishida, K.; Hertweck, C.; Minehan, T. G. Org. Lett. 2014, 16, 2962.

446

8. SYNTHETIC STRATEGIES

functionality, or for functional group insertion) that were achieved all in essentially one synthetic step. Other reactions could easily be termed powerful, but these are sufficient to illustrate that if a reaction induces extensive and useful structural modifications, the synthetic tree should be biased to take advantage of that powerful chemistry. 7. Disconnect from atoms bearing certain functional groups. Certain functional groups are tied together along a reaction pathway. In many cases, the functional group transforms (CHdOH ! C]O and C]O ! CHdOH, e.g.) involve ionic or polarized intermediates or reactants. Ionic functional group transformations are a useful illustration of this rule. Both anionic and cationic reactions involve a (positive) or (δ+) end, and a (negative) or (δ
) end. The functional groups corresponding to these reactive fragments have a special reactivity relationship, which can be exploited. When a bond is disconnected, identification of each end of the disconnect product as a donor or acceptor can lead to identification of possible reaction pathways to reconnect the bond (discussed in Section 1.2). The disconnection of 42 to 43, with the appropriate donor and acceptor sites, clearly suggests an aldol condensation (Section 13.4.1), where 42 can be synthesized from a synthetic equivalent of 43. a

a

d

d

O

O 43

42

The relationship between two or more atoms in a disconnection usually requires knowledge of a specific transformation. The disconnection 44 ) 45 suggests a relationship between the tertiary amine in 44 and the dienyl azide moieties in 45.107 When 45 was heated in toluene, a [3 + 2]-cycloaddition occurred to give a triazoline (46) (Section 15.4.8), but this unstable molecule decomposed (with loss of nitrogen) to form vinyl aziridine (47) as a mixture of stereoisomers.107 When 47 was heated to 480°C (using flash vacuum pyrolysis), the nitrogen analog of a vinylcyclopropane rearrangement (Section 15.5.2) occurred, and the resulting product was subjected to catalytic hydrogenation to yield 44. This reaction sequence (azide addition, vinylcyclopropane rearrangement) led to a bicyclic amine, linking the amine in 44 with the azido diene moieties in 45. Clearly, the special relationship of these groups is exploited by a combination of reactions. HO

OH

CO2Me

CO2Me N

N3

44 OH

OH

CO2Me

CO2Me 45

HO – N2

Heat

N3

45

N

N

CO2Me N

1. FVP (480°C) 2. H2, cat

HO

CO2Me

N

N 46

47

44

8. Consider Bonds Related by Known Transforms. Some functional groups are related by the reactions that transform one to the other. It is clear that the disconnection of 42 and 44 above relate to specific reactions or transforms. These transforms may be as obvious as alcohol ! ketone. There are other transformations that connect or disconnect several bonds and/or functional groups at one time. The relationship is not always obvious, but its recognition can lead to great simplification of a target in one disconnection. An important and useful example is the oxy-Cope rearrangement (see Example 16 in Table 8.1) in which the diene structure is related to the alkene-ketone functionality in the final product. Another example is the

107

Hudlický, T.; Seoane, G.; Lovelace, T. C. J. Org. Chem. 1988, 53, 2094.

447

8.6 STRATEGIC BONDS IN RINGS

well-known 1,3-elimination reaction known as the Grob fragmentation (Section 3.9).108 In a synthesis of periplanone C by Saicic and coworkers,109 treatment of 48 with methanesulfonyl chloride generated the primary mesylate. Subsequent treatment with KOH led to the alkoxide to yield alkoxide 49, and the mesylate leaving group has the proper relationship to the alkoxide to allow a Grob fragmentation (Section 3.9). Fragmentation gave 50 in 75% yield from 48. In this case, the alkoxide and the leaving group (mesylate) have a chemical reactivity relationship to the carbonyl and alkenyl unit via the 1,3-elimination. Recognition of the spatial and structural relationships of the Grob fragmentation allows one to plan the disconnection, 50 ) 48. OH

O–

1. MsCl, NEt3 CH2Cl2 2. KOH, PhH, rt 18-Crown-6

HO

O

MsO 50

49

48

(75%)

8.6 STRATEGIC BONDS IN RINGS 8.6.1 Carbocyclic Rings In addition to the eight criteria in Section 8.5.2, there is a ninth criterion71 that identifies strategic bonds in rings. Bonds are strategic because their disconnection maximizes the simplification of a polycyclic molecule. Corey and Howe65 devised a set of guidelines that accommodate the special problems encountered in monocyclic and polycyclic structures. 1. Strategic bonds must be in a primary ring. A primary ring cannot be expressed as the envelope of two or more smaller rings bridged or fused to one another.65,110 Five-, six-, and seven-membered rings are usually easy to form, whereas other rings (especially C8 and larger) are difficult to form by cyclization techniques (Sections 1.5.3 and 4.5.1). The two six-membered rings A and B in decahydronaphthalene, A and B are primary rings, but the 10-membered ring C that encloses A and B is not (i.e., bond k is not strategic). It is important to note that bonds a–i are all strategic by this rule since every one is a part of a primary ring. The presence of ring C as an envelope of rings A and B does not change this. j

i

A

h g

k f

b

a

e

c

B

C

d

Decahydronaphthalene

2. Strategic bonds must be directly attached to another ring. The strategic bond should be exo to another ring. A ring disconnection that produces two functionalized appendages leads to a more complex overall synthesis. Three disconnections are shown for 51. Paths I and II disconnect a bond exo to ring A to yield 53 and 54, respectively. Synthetically, this ring closure is straightforward, and may proceed with stereoselectivity. Disconnection III leads to 52, which is a more difficult intermediate to prepare. Although the ring closure for III is probably facile, the disconnections in I or II lead to greater simplification and have a greater potential for stereocontrol. There are very few ring-closure methods in which bonds are fused to preexisting three-membered rings and strategic bonds may not be exo to rings of that size.65,110

(a) Grob, C. A. Experientia 1957, 13, 126. (b) Cherbuliez, E. Baehler, Br.; Rabinowitz, J. Helv. Chim. Acta 1961, 44, 1820. (c) Zurfl€ uh, R.; Wall, E. N.; Siddall, J. B.; Edwards, J. A. J. Am. Chem. Soc. 1968, 90, 6224.

108

109

Ivkovic, A.; Matovic, R.; Saicic, R. N. Org. Lett. 2004, 6, 1221.

110

See Reference 65, footnotes 8–10 cited therein.

448

8. SYNTHETIC STRATEGIES

X A

I

Y

X A

III

53

Y

Y

A

II

52

51

X 54

3. Strategic bonds must be in rings with the greatest degree of bridging. Disconnection of a bond that is highly bridged (connected to several rings) leads to greater simplification of that structure. Disconnection A in the highly bridged four-membered ring of 54, which leads to 55 with a bicyclo[4.4.0]decane unit, whereas disconnection B leads to 56. The bicyclo[3.1.1]heptane ring system in 56 is not as easy to synthesize, and provides less simplification when compared to 55. B

•

A A

• •

•

B 56

55

54

A maximum bridging ring is defined as one that is bridged at the greatest number of sites. The maximum bridging ring(s) is(are) selected from the set of “synthetically significant rings that are defined as the set of all primary rings and all secondary rings less than eight membered.”65,110 Structure 57 has six significant rings: A, B, C, D, E, and F.110 The highlighted rings in A ! D are primary rings, but those in E and F are secondary. Ring structure C is bridged at four sites, more than any other, and is the maximal bridging ring.65 Ring B contains as many bridgehead sites as C, but it is not a maximal bridging ring, because it is bridged to other rings at only two of these sites (b and d). Therefore, the number of times a ring is bridged is not a valid criterion for determining maximal bridging character. Ring E is bridged to as many rings as ring A, but has one less bridgehead site than C.

d

a b

c

57

A

B

D

C

E

F

4. Bonds common to a pair of bridged primary rings are not strategic. Any bond common to a pair of bridged or fused primary rings whose envelope is eight membered or larger cannot be strategic.65 Such a disconnection would generate a ring of more than seven members. Disconnection of the central bond in decahydronaphthalene would result in a 10-membered ring compound (58), but rings of this size are difficult to synthesize by direct cyclization. For that reason, this type of disconnection should be avoided.110 • • Decahydronaphthalene

58

If the two fused or bridged rings are directly joined elsewhere by another bond (as in 59), the bond may be strategic. As shown in 59, disconnection of the indicated bond produces two fused six-membered rings (see 60) instead of a 10-membered ring. Similar disconnection in 61, however, yield a 10-membered ring (62), as mentioned above, and is not strategic.

449

8.6 STRATEGIC BONDS IN RINGS

•

•

• • 59

60 •

•

• 61

•

62

5. Bonds within aromatic rings are not strategic. Aromatic nuclei are common fragments in natural products, but they are usually incorporated as intact units of another fragment, although aromatic rings can be reduced and converted to other functionality. Although introduction of a phenyl ring and reduction (Birch reduction, see Section 7.11.5) to a cyclohexadiene or cyclohexenone is common, incorporation of a cyclohexene and later oxidation to a phenyl ring is rare. Notable exceptions are quinones (Section 6.3.1), which are commonly used as synthons and later converted to functionalized aromatic rings.111 Incorporation of a carbocyclic ring and subsequent aromatization often requires harsh conditions and several synthetic steps. For these reasons, it is rarely useful to create an aromatic ring in a disconnection step unless aromatization is facile and noninjurious to other functionality. 6. In a cyclic arc, cleavage of the bond should not leave stereocenters in the side chain. Reactions that join two fragments to generate a chiral center allow greater control than those reactions that join fragments containing a chiral center remote to the site of bond formation. Disconnection of bond B in 64 would lead to 65, requiring the synthesis of the alcohol with a stereogenic center on the chain prior to ring closure. It is not clear that there would be any control over the stereochemistry of that center as it is formed. The reaction resulting from disconnection of bond A (to 63), however, forms the ring and generates the stereogenic center simultaneously. Disconnection of 64 to 63 will lead to a shorter synthetic sequence, and there is the possibility that it can proceed with reasonable stereoselectivity. A special case arises when the bond is attached to a stereogenic center, but that center is not the only one on the arc linking the two common atoms.65 Such a bond may be strategic. In 66, bond a is strategic, since the arc in the disconnect product (67) does not contain an asymmetric center. In 68, the arc contains a stereogenic center, and bond a is not strategic, since that disconnection yields 69. CHO

X

B A

OH

B

X

H X

A

63

OH

64

65

NO2 O

NO2 c

b

a

OH Me

Me 66

67

NO2 O

NO2 c

a b

Me

111

OH Me

68

H

Me

Me 69

(a) Fu, P. P.; Harvey, R. G. Chem. Rev. 1978, 78, 317. (b) Walker, D.; Hiebert, J. D. Ibid. 1967, 67, 153. (c) Jackman, L. M. Adv. Org. Chem. 1960, 2, 329.

450

8. SYNTHETIC STRATEGIES

These rules can be applied to a carbocyclic molecule (e.g., 70). An analysis based on these six rules is shown in Table 8.2.65 Rule 1 shows that all bonds are part of a primary ring and rule 5 does not apply since there are no aromatic rings. Rule 2 shows that several bonds are exo to primary rings, indicated in A. The primary rings are highlighted in B–D. There are larger rings forming an envelope of these smaller rings, but bonds exo to those highlighted rings will take into account all possibilities. Rule 3 suggests that bonds i, j, l, f, and n are part of bridging rings. Disconnection of bond m is ruled out since the disconnection is away from the stereogenic center and does not involve an exo bond. Disconnection of bond k leads to a large ring and is ruled out. Bonds b–d, f, and g–i are not disconnected since it would violate rule 6.

TABLE 8.2 Strategic Bond Analysis for Carbocyclic Compounds l g d m f e c k h n i j a b

A

70

Rulea

B 1

2

3

C 4

5

Y

X

6

D Strategic

Bond a

Y

b

X

c

X

d

X

e

Y

f

Y

Y

X

Y

X

Y

Yes

g

X

h

X

i

Y

Y

X

Y

Yes

j

Y

Y

X

Y

Yes

k

Y

X

Y

l

Y

X

Y

Yes

Y

Yes

X Y

m n

X Y

Y

X

a

Y indicates the rule applies, and X indicates the rule is not applicable.

8.6.2 Heterocyclic Rings The six rules discussed in Section 8.6.1 apply to many carbocyclic ring systems. If a bond in the ring is connected to a heteroatom, that bond can be considered strategic if it satisfies rules 2, 4, 5, and 6. Corey and Howe65 generated the strategic bonds for lycopodine in Fig. 8.6, using these rules. Bonds 3, 12, 14, and 19 were found to be strategic for the first disconnection. Thereafter, these bonds may or may not be strategic in the disconnect product, which should be reexamined to determine new strategic bonds. This process is continued as needed to generate a synthetic tree. The process used to determine strategic bonds in lycopodine followed five steps. (1) Structure of lycopodine (arbitrary bond numbering). (2) Primary rings. There are four primary rings, highlighted in A–D. (3) No synthetically significant ( 6.29 When n ¼ 5, both 30 and 29 were formed, but 30 was not formed when n < 5. 28

Burk, L. A.; Soffer, M. D. Tetrahedron 1976, 32, 2083.

(a) Bredt, J.; Houben, J.; Levy, P. Berichte 1902, 35, 1286. (b) Bredt, J. Annalen 1924, 437, 1. (c) Fawcett, F. S. Chem. Rev. 1950, 47, 219. (d) K€ obrich, G. Angew. Chem. Int. Ed. Engl. 1973, 12, 464.

29

(a) Weiseman, J. R.; Chong, J. A. Ibid. 1969, 91, 7775. (b) Wiseman, J. R.; Pletcher, W. A. Ibid. 1970, 92, 956. (c) Quinn, C. B.; Wiseman, J. R. Ibid. 1973, 95, 1342, 6120. (d) Quinn, C. B.; Wiseman, J. R.; Calabrese, J. C. Ibid. 1973, 95, 6121. (e) Krabbenhoft, H. O.; Wiseman, J. R.; Quinn, C. B. Ibid. 1974, 96, 258. (f ) Marshall, J. A.; Fauble, H. Ibid. 1970, 92, 948.

30

(a) Prelog, V.; Ruzicka, L.; Barman, P.; Frenkiel, L. Helv. Chim. Acta 1948, 31, 92. (b) Prelog, V.; Barman, P.; Zimmermann, M. Ibid. 1949, 32, 1284. (c) Prelog, V. J. Chem. Soc. 1950, 420.

31

529

10.3 STEREOCONTROL IN CYCLIC SYSTEMS a

O n=5

(CH2)n

b

(CH2)n O

O

a

n>6

(CH 2)n O

b

CO2R

CO2R 28

29

30

Heat

Heat

O

NMe3 OH 33

O 31

CO2R

32

Fawcett29c attempted to correlate the lower limit of Bredt’s rule by calculating the sum of the numbers of the bridged atoms [m + n + o ¼ S] in systems (e.g., camphane and pinane) with three bridges. For camphane, S ¼ 2 + 2 + 1 ¼ 5, and for pinane, S ¼ 3 + 1 + 1 ¼ 5. In Prelog’s example (28), this [6.3.1] system has S ¼ 9 when n ¼ 5, and Bredt’s rule does not apply. When the values of n are < 5, however, Bredt’s rule applies for 28. Relatively small bridgehead alkenes with a [3.3.1] system (e.g., 32 S ¼ 7) can be prepared by pyrolysis of β-lactone 31 or ammonium salt 33.32 Several small bicyclic ring systems for S ¼ 2 to S ¼ 7 are shown in Fig. 10.1,29d and analysis of these alkenes suggests that the constraints upon Bredt alkenes are important when S is 7 or less.

S=7 _ [3.3.1]

_ [3.2.2]

_ [3.2.2]

_ [4.2.1]

_ [5.1.1]

_ [5.1.1]

S=6 _ [2.2.2]

_ [3.2.1]

_ [3.2.1]

_ [3.2.1]

_ [4.1.1]

_ [4.1.1]

_ [4.1.0]

_ [4.1.0]

_ [5.1.0]

S=5 _ [2.2.1]

_ [2.2.1]

_ [3.1.1]

_ [3.1.1]

S=4 _ [2.1.1]

_ [2.1.1]

_ [1.1.1]

_ [2.1.0]

_ [3.1.0]

_ [3.1.0]

_ [2.2.0]

S=3 _ [2.1.0]

S = Sum of the numbers of the bridged atoms

S=2 _ [1.1.0]

FIG. 10.1 Typical Bredt alkenes. K€obrich, G. Angew. Chem. Int. Ed. Engl. 1973, 12, 464. Copyright © 1973 Wiley-VCH Verlag GmbH & C KGaA. Scheme 1 on p 472 therein. Reproduced with permission.

32

(a) Marshall, J. A.; Faubl, H. J. Am. Chem. Soc. 1967, 89, 5965. (b) Chong, J. A.; Wiseman, J. R. Ibid. 1972, 94, 8627 and references cited therein.

530

10. FUNCTIONAL GROUP EXCHANGE REACTIONS: SELECTIVITY

K€ obrich established three guidelines29d that govern the application of Bredt’s rule. 1. For homologues with different S values, the ring strain varies inversely with S. 2. For a given S, the ring strain varies inversely with the size of the larger of the two rings with respect to which bridgehead bond is endocyclic. 3. For a given bicyclic ring skeleton, the ring strain varies inversely with the size of the bridge containing the bridgehead double bond. These rules predict that bicyclo[4.2.1]non-1-ene is more stable than bicyclo[4.2.1]non-6-ene.33 These three rules do not allow direct comparison in every case, but are generally applicable. K€ obrich29d presented a list of carbocyclic Bredt’s compounds and stated that many heteroatom and polycyclic analogs can be included in those compounds listed in Fig. 10.1.29d

Bicyclo[4.2.1]non-1-ene

Bicyclo[4.2.1]non-6-ene

10.3.3 Diastereoselectivity As mentioned previously, reactions of cyclic molecules that involve formation of stereogenic centers are similar to those of acyclic systems in that they can proceed with clean retention or inversion of configuration, or a mixture of the two. Reaction of sodium azide (NaN3) and cis-4-tert-butyl-1-bromocyclohexane is expected to give trans-4-tertbutylazidocyclohexane with complete inversion of configuration via an SN2 pathway. Many cyclic systems were shown to exist primarily in one or two low-energy conformations in Section 1.5.2, and this conformational bias can be used to control stereochemistry if other groups are present that sterically block approach of a reagent. A typical example is the reduction of a ketone moiety in a cyclic or polycyclic substrate, where the diastereoselectivity of the reduction is controlled by the conformational constraints of the system. In a synthesis of malayamycin A by Hanessian et al.34 the ketone unit in 34A was reduced to the alcohol unit in 34 with high diastereoselectivity. The conformation of the ring, shown in conformational drawing 34B allowed the diastereoselective reduction to proceed with sodium borohydride, primarily from face a rather than face b, and did not require one of the more expensive hydride reducing agents. N3 O

N3

MeO

H

N

O

HO

NaBH4 , MeOH

MeO

H

N

O

OMe O

OMe

N H

O

OPMB

b

34A O

H

N3

O H

H

OPMB 35

MeO H O

a

N

N OMe

H

N

34B

In many cases, reaction conditions and strategy combine to give a product with the incorrect stereochemistry at a key center, often an alcohol. In 36, an intermediate in Philips and Chamberlin’s35 synthesis of dysiherbaine, the stereochemistry of the secondary alcohol is incorrect for the synthetic target. In order to invert the stereochemistry, 36 was oxidized with TPAP and NMO (see Section 6.2.6.1) to give a 95% yield of ketone 37. Subsequent reduction with NaBH4 gave a quantitative yield of 38 with the correct stereochemistry. The observed inversion is possible because approach of 33

Wiseman, J. R.; Chan, H.-F.; Ahola, C. J. J. Am. Chem. Soc. 1969, 91, 2812.

34

Hanessian, S.; Marcotte, S.; Machaalani, R.; Huang, G. Org. Lett. 2003, 5, 4277.

35

Phillips, D.; Chamberlin, A. R. J. Org. Chem. 2002, 67, 3194.

531

10.3 STEREOCONTROL IN CYCLIC SYSTEMS

the hydride reagent to the carbonyl unit in 76 is restricted by the presence of the tricyclic system and other substituents, as well as the requisite B€ urgi-Dunitz trajectory common to hydride reduction of ketone units (see Section 7.9.4). Oxidation followed by reduction of a carbonyl is a simple, yet powerful tool for modifying stereochemistry. O

O O

HO

O

TPAP , NMO

t-BuO2CNH MeO2C

O O

O

t-BuO2CNH

CO2Me

MeO2C

O

t-BuO2CNH

CO2Me

36

O

HO

NaBH4

O CO2Me

MeO2C

37 (95%)

38

Sometimes, the natural selectivity of a reaction leads to a product with the incorrect stereochemistry required for a given target, but it may be possible to change it. There are at least two options: (1) change the stereocenter in the final product, as in 36 ! 38, or (2) change the synthetic pathway to generate a compatible intermediate that can produce the desired stereochemistry. O

O

OR = OSiMe2t-Bu

NH

O 5% H2O THF

NH

LiN(i-Pr)2 THF –78°C

–78°C

CO2Me

RO

NH

RO

LiO

39 (1:1.7 cis/trans)

CO2Me

RO

OMe

41

40

One type of stereocenter that can be corrected is a CHX moiety, where X is an electron-withdrawing group and the H can be removed by treatment with base to form a planar enolate anion (Section 13.2). In a synthesis of quinine and quinidine by Jacobsen and coworkers,36 lactam (39, as a 1:1.7 mixture of cis-trans isomers) was converted to the enolate anion (40) with lithium diisopropylamide (Section 13.2.2). It appears that there is only a small difference in steric hindrance of one face relative to the other, and this observation is consistent with the isolated product, a 3:1 mixture favoring 41. Exploiting such differences in product stability allows one to change the diastereomer population. The second approach noted above for correcting stereochemistry is illustrated by formation of epoxides 44 and 46 from conjugated ketone 42, by Grieco and coworkers.37 Epoxidation of 42 from the less hindered α-face gave 43, and reduction from that same face gave 44. The relative stereochemistry was different, however, when initial reduction of 42 (hydride was delivered from the β-face) gave alcohol 45. Subsequent epoxidation proceeded via coordination of the OH unit in 45 with the epoxidation agent (mcpba, Section 6.4.3), and the final product was 46 with the epoxide and hydroxyl unit cis (syn) to each other.37 Simply changing the order of reactions allowed Grieco to target one diastereomer or the other. Me H

Me H

Me H t-BuOOH

O BnO

1. NaBH4

O

0°C

42

BnO Me

BnO Me

Me [H]

O 43

Me H

O

44 Me H

OH

OH

mcpba

BnO Me 45

36

OH

2. H3O+

Me 46

O

Raheem, I. T.; Goodman, S. N.; Jacobsen, E. N. J. Am. Chem. Soc. 2004, 126, 706.

(a) Grieco, P. A.; Ohfune, Y.; Majetich, G. J. Org. Chem. 1979, 44, 3092. (b) Ohfune, Y.; Grieco, P. A.; Wang, C.-L. J.; Majetich, G. J. Am. Chem. Soc. 1978, 100, 5946.

37

532

10. FUNCTIONAL GROUP EXCHANGE REACTIONS: SELECTIVITY

One major difference between cyclic and acyclic systems is the inability of cyclic systems to undergo rotation about carbon-carbon bonds, but pseudorotation leads to a relatively well-defined low-energy conformation. If trans-4-tertbutyl-1-bromocyclohexane is used as an example, there is the usual equilibrating population of the two chair conformations (Section 1.5.4). The diequatorial conformer is significantly more stable and accounts for the major portion of the equilibrium. According to the Corey et al. protocol38 from Section 1.5.3, the equatorial form predominates (>99.99:99.99% of the diequatorial conformation. To predict the major product, nucleophilic backside attack on both conformations must be considered, and the less sterically hindered pathway is usually preferred. In this case, the diequatorial conformation is more reactive. H

Br H B A

Br H

47

48

Br H

When there is a large group attached to cyclohexane, there is a high preference for the conformation with that group in the equatorial position. If the group is sufficiently large, in effect the cyclohexane ring is conformationally “locked.” In other words, there is an extremely high percentage of one conformation. In a conformationally anchored system (47), SN2 displacement of the equatorial bromine may be extremely slow. If the reaction is too slow, the temperature can be raised and a more polar solvent (DMF, DMSO, etc.) can be used (Section 3.2.1.1). In some cases, the more vigorous reaction conditions will lead to an alternative reaction (e.g., elimination or thermal decomposition) rather than the desired SN2 displacement. Similarly, dibromide 48 is expected to undergo nucleophilic substitution in ring B since displacement of the bromide in ring A is very slow. OTs

O

Me

d

OTs

S H3C

Me

O

H

CH2Na

Me

Me

c

DMSO 75°C

a

Me

H

b

O

O

H 51A

50A

49

Me

a

a d

b

d

c

50B

b

c

Backside attack is not possible 51B

The relative stereochemistry of the leaving group is very important for intramolecular reactions, as seen in Corey and Watt’s39 synthesis of α-copaene and α-ylangene. Displacement of the tosyl group in 49 by the nucleophilic carbon of the enolate anion (50A, formed with dimsyl sodium, Sections 13.2.1 and 13.2.2) gave the bridged product (51A) characteristic of the ylangene natural products. Examination of the conformational drawing for 50A (see 50B) reveals that the enolate carbon is properly oriented to attack the tosyl carbon from the rear, so displacement of the OTs group is possible only when it is in the equatorial position and is syn to the bridgehead methyl group. Note the letters marking the key atoms. The isomeric axial tosylate (51B; the tosyl group is anti to the bridgehead methyl group) cannot be displaced via backside attack, since the enolate carbon cannot approach the tosyl carbon from the rear. Ketone 49 will react as shown to give 51A, but the isomeric ketone 51B will not. 38

Corey, E. J.; Feiner, N. F. J. Org. Chem. 1980, 45, 765.

39

Corey, E. J.; Watt, D. S. J. Am. Chem. Soc. 1973, 95, 2303.

533

10.3 STEREOCONTROL IN CYCLIC SYSTEMS

Marshall et al. 40 showed that alkenyl aldehydes can be cyclized on acidic silica gel, and the stereochemistry of the bicyclic alcohol product is controlled by the cis or trans geometry of the monocyclic precursor. This reaction is an ene reaction (Section 15.6), which is particularly effective when catalyzed by Lewis acids.40c The cis precursor 52 was cyclized to 53, whereas the trans-aldehyde 54 was cyclized to 55. Formation of a seven-membered ring in 53 requires a seven-membered chair-like ring for the ene reaction to occur (see 56). If the oxygen is in the axial position (see 57) rather than the equatorial position (see 59), then the alkenyl group is further removed from the acylium cationic center (an oxocarbenium ion), and transfer of the hydrogen and attack by the π-bond for an ene reaction is more difficult. Me

Me CHO

SiO2

OH

52

53 Me

Me CHO

PhH

OH

SnCl 4

H

H 54

55

Cyclization of 54 requires transition state 58. This transition state brings the alkenyl moiety close to the oxocarbenium ion, and the oxygen is in the axial position for efficient hydrogen transfer and attack by the alkene, leading to 55. The opposite conformation with the equatorial oxygen (see 59) has a greater transannular steric interaction, and is also in a higher energy boat-like transition state that disfavors the ene reaction in this particular system. When an alkene unit is the product of a reaction, there is the potential for both (E)-and (Z)-isomers. The examples previously cited for E2 and syn-elimination are illustrative. In both cases, the stereochemistry (E) or (Z) of the double bond is important. Stereochemistry may be controlled in acyclic systems by using resolved chiral precursors because the E2 elimination is stereospecific. Controlling cis-trans isomers for alkene derivatives is also possible in medium- and large-ring molecules, and other reactions are known that introduce the C]C unit.

H

Me

H

H H

H

H

Me

Me

H

H

O

H

H

H

OH

H

E

O

57

H 53

56 H

E

H

Me

H

H H

H

O

Me

E

H

58

H

Me

H

H

H OH

H

H

55

H

O 59

E

H

The Corey et al.41 synthesis of caryophyllene and isocaryophyllene illustrates this type of control. A Grob fragmentation42 (Section 3.9) in 60 led to the (Z)-double bond in 61, required for caryophyllene. Adjusting the stereochemistry of a carbon (a) Marshall, J. A.; Andersen, N. H.; Johnson, P. C. J. Org. Chem. 1970, 35, 186. (b) Marshall, J. A.; Andersen, N. H.; Schlicher, J. W. Ibid. 1970, 35, 858. (c) Snider, B. B. Acc. Chem. Res. 1980, 13, 426.

40

41

(a) Corey, E. J.; Mitra, R. B.; Uda, H. J. Am. Chem. Soc. 1963, 85, 362. (b) Idem, Ibid. 1964, 86, 485.

42

Grob, C. A. Angew. Chemie, Int. Ed. Engl. 1969, 8, 535.

534

10. FUNCTIONAL GROUP EXCHANGE REACTIONS: SELECTIVITY

that bears a leaving group in a cyclic precursor controlled the geometry of the double bond (see a discussion of this approach in Section 10.5). Once the (Z)-double bond had been incorporated, the α hydrogen was equilibrated (via enolization with potassium tert-butoxide) from the initially formed cis ring juncture (in 61) to the more stable trans-ring juncture in 62. To obtain the other geometric isomer (in isocaryophyllene) the relative stereochemistry of the OTs group had to be changed to that shown in 63. A Grob fragmentation of 63 gave the (E)-double bond in 64 that is characteristic of isocaryophyllene.42 As before, the cis-ring juncture in 64 was equilibrated to the more stable trans-ring juncture in 65 by enolization prior to conversion to isocaryophyllene. These examples show that the (E/Z)-stereochemistry of the alkene products was established by the relative stereochemistry of the tosylate precursors, once the reaction to generate the double bond in those alkenes was chosen. H

Me

OTs

Me

Na+ – CH2SOMe

H H

Me

O

Me

H 61

Na+– CH2SOMe

H

Me

DMSO

OH

t-BuOK

H

H

O Me

H Me

H Me Caryophyllene

62

Me

OTs

Me

H Me

Me 60

Me

H

H

t-BuOH

O

Me

Me

H

Me

t-BuOK

H

DMSO

OH

Me

H

Me

H

H

t-BuOH

Me

O

Me

H

H

Me

Me

Me 63

H

H

H Me

64

Isocaryophyllene

65

10.4 NEIGHBORING GROUP EFFECTS AND CHELATION EFFECTS The influence of heteroatom substituents to direct reactivity to one face or another has been noted in this, as well as several preceding chapters. A neighboring group effect influenced delivery of oxygen to the double bond of an allylic alcohol, described by Henbest and by Sharpless (Sections 6.4.3 and 6.4.4). The effect is identical to the effect described for acyclic control in Section 10.2. Peroxyacid epoxidation of cyclopent-2-en-1-ol proceeded via coordination of the peroxyacid to the alcohol (see 66) to deliver the electrophilic oxygen from that face to yield syn-6-oxabicyclo[3.1.0] hexan-2-ol.22 This result contrasts with epoxidation of allylic acetate cyclopent-2-en-1-yl acetate, which gave primarily anti-6-oxabicyclo[3.1.0]hexan-2-yl acetate via delivery from the less sterically hindered face. The acetate group inhibits coordination with the peroxyacid, and delivery of the electrophilic oxygen is from the less sterically hindered face, so the epoxide unit is on the opposite side of the ring. H H

H

H PhCO3H

O

H O

O

O

O

H

– PhCO2H

H

O

O H

H Ph Cyclopent-2-en-1-ol

syn-6-Oxabicyclo[3.1.0]hexan-2-ol

66 O

H CH3

O O

Cyclopent-2-en-1-yl acetate

PhCO3H

H

H H

CH3

O O

anti-6-Oxabicyclo[3.1.0]hexan-2-yl acetate

535

10.4 NEIGHBORING GROUP EFFECTS AND CHELATION EFFECTS

A synthetic example of this effect is taken from a synthesis of 1,6-dideoxynojirimycin by Park and coworkers,43 in which epoxidation of allylic alcohol 67 occurred from the same face as the hydroxyl group via a neighboring group effect, to give a 67% yield of 68. Note that this effect was seen in the conjugate epoxidation of ketone 42, versus the alcohol-directed epoxidation of cyclopent-2-en-1-ol.37 HO

O

HO mcpba , CH2Cl2

Me

rt

N

Me

Boc

Boc

67 O

O

N

68 (67%)

OH

O

Zn(BH4)2 , CH2Cl2 –78°C

t-BuO

OH

OH

t-BuO

70 (80%)

69

The oxygen atoms can chelate with the Zn in reductions that use zinc borohydride [Zn(BH4)2], as discussed in Section 7.5.24 In a synthesis of the C1dC22 fragment of leucascandrolide A, Panek and Dakin44 treated 69 with zinc borohydride. Coordination of the zinc reagent with the hydroxyl unit led to formation of 70 in 80% yield, with a dr of >15:1 favoring the diastereomer shown. Hydroboration of alkenes to yield alcohols typically follows an anti-Markovnikov orientation (Section 9.2.1). Chelation effects can sometimes control the product distribution, but the mere presence of a heteroatom does not always mean that this will be the case. Hydroboration of tabersonine is an example where the hydroboration-oxidation sequence gave 71 in 90% yield, as a mixture of diastereomeric secondary alcohols (4:1 β/α mixture favoring 71). This reaction was part of the Caron-Sigaut et al. 45 synthesis of 14-hydroxyvincadifformine, but they showed that chelation was not an important factor contrary to what might be suspected from a cursory examination of the starting materials. The enamine unit in tabersonine is part of a bicyclic system where the exo face is more accessible, leading to the stereochemistry in 71 (Sections 1.5.4 and 7.9.6). The regiochemistry of the alcohol unit in 71 probably results from steric interactions as borane approaches the C]C unit. A

••

••

B

N

OH

N 1. B2H6

N

CO2Me

H

2. NaOH , H 2O 2

N

CO2Me

H

Tabersonine

71 (90%)

Chelation is important in the dihydroxylation of 72 with OsO4 (Section 6.5.2), where reduction of the intermediate iminium ion gave 73, in a synthesis of Δ200 -200 -deoxyvinblastine.46b This transformation is compared with the reaction of 7246a with thallium acetate, where reduction of the intermediate iminium ion gave the epimeric alcohol 74.

43

Rengasamy, R.; Curtis-Long, M. J.; Seo, W. D.; Jeong, S. H.; Jeong, I.-Y.; Park, K. H. J. Org. Chem. 2008, 73, 2898.

44

Dakin, L. A.; Panek, J. S. Org. Lett. 2003, 5, 3995.

45

Caron-Sigaut, C.; Le Men-Olivier, L.; Hugel, G.; L
evy, J.; Le Men, J. Tetrahedron 1979, 35, 957.

(a) Mangeney, P.; Andriamialisoa, R. Z.; Langlois, N.; Langlois, Y.; Potier, P. J. Am. Chem. Soc. 1979, 101, 2243. (b) Langlois, N.; Potier, P. Tetrahedron Lett. 1976, 1099.

46

536

10. FUNCTIONAL GROUP EXCHANGE REACTIONS: SELECTIVITY

N

OH

N

N

1. OsO4 2.NaBH4

N

2.NaBH4

N

H R

OH

1. Tl(OAc)3

H R

CO2Me

73

N H R

CO2Me

72

CO2Me

74

The nitrogen lone pair is responsible for the reversal of diastereoselectivity. Osmium tetroxide does not coordinate well to the nitrogen and the osmate ester is formed on the less hindered α-face of 75, which leads to diol 76. Elimination of the α-hydroxylamine gave an iminium salt (77) that was subsequently reduced with NaBH4 to yield product 73. Thallium (III), however, coordinates quite well to the nitrogen lone pair (as in 78) to yield 79. Transfer of an oxygen from one acetate unit to the enamine moiety (see Section 13.6), from the β face, gives 80. Reduction leads to product 74.

N

N

NaBH4

N

N

OH Os O

OH

OH

O

O

OH

75

O

77

76 AcO

Tl

Tl(OAc)3

O

N

N

73

OAc O OAc

Me

OH

NaBH4

N

N

Tl(OAc)3

78

79

80

74

Neighboring group effects usually influence stereo- and regiochemistry of the functional group close to the chelating atom or group. Such effects can influence a reaction at sites distant from the coordination point.47 An example of a neighboring group effect with consequences at a remote site is the photochemical isomerization of 81 (n ¼ 1) to 82, in 55% yield (see Section 15.2.2 for a brief introduction to photochemistry).48 The β C3 hydroxyl group was converted to an ester, and the benzophenone moiety was positioned on the β-face of the molecule by a chain of methylene units. Irradiation generated the benzophenone triplet (see Section 15.2.2), and with an appropriate methylene spacer [(CH2)n] the hydrogen atom at C14 was selectively removed to yield the Δ14–15 alkene (82, n ¼ 1). With this particular spacer, the C8 hydrogen on the α-face of the molecule is inaccessible and not removed. Longer or shorter methylene spacers (n ¼ 0, n > 2) led to a diminished yield of 82. Me

Me

Me

H

Me

H

Me

Me

h H H

H H

O

H

H O

H

O

(CH 2)n

O

OH

(CH 2)n

O 81

82 (55%)

For examples, see (a) Sailes, H.; Whiting, A. J. Chem. Soc. Perkin Trans. 2000, 1, 1785. (b) Mitchell, H. J.; Nelson, A.; Warren, S. J. Chem. Soc. Perkin Trans. 1999, 1, 1899.

47

48

Breslow, R. Chem. Soc. Rev. 1972, 1, 553.

10.5 ACYCLIC STEREOCONTROL VIA CYCLIC PRECURSORS

537

Neighboring group effects can compete with steric effects, as illustrated by treatment of 2-butyl-3,3-dimethyloxetane with diethylaluminum-N-methylaniline. This elimination reaction gave a 99% yield of the (E)-alkene, (E)-2,2dimethylhept-3-en-1-ol, in a synthesis of humulene.49 The aluminum coordinated to the oxetane oxygen on the less hindered β-face in 83 and 84, the two key rotamers. Rotamer 83 is lower in energy due to the severe nonbonded interaction of the propyl group, and the oxetane methyl in 84. Removal of the indicated hydrogen atom by the basic nitrogen, in a six-center transition state as seen in 83, led to the (E)-isomer (E)-2,2-dimethylhept-3-en-1-ol. Removal of hydrogen in 84 would yield the (Z)-isomer, but the relatively high energy of that species relative to 83 makes it disfavored. C3H7

Et2Al•NMePh

C3H7

81 h

O 2-Butyl-3,3-dimethyloxetane

OH (E)-2,2-Dimethylhept-3-en-1-ol (99%)

Me H

H O

Me

H

H N

83

Me

H

H Al

O

Me

N

Al

84

In planning a reaction, not only the obvious effects of α- or β- hydroxyl, alkoxy, or amino groups must be considered, but also the more subtle effects of nearby heteroatom lone pairs. These chelating effects are apparent both in directing stereochemistry or regiochemistry and in slowing or stopping a desired reaction. These effects can be used synthetically to direct attack of a reagent, as shown above. If one wishes to prevent the chelating effect, the offending group must be blocked or removed. Alternatively, the planned synthetic sequence can be accomplished before the chelating atom is present. Careful analysis of all synthetic intermediates is required to use or avoid these effects.

10.5 ACYCLIC STEREOCONTROL VIA CYCLIC PRECURSORS It is apparent from preceding sections that stereocontrol in cyclic systems is much easier than in acyclic systems, which is due, of course, to the conformational bias inherent in cyclic systems. Synthetic chemists have exploited this fact for many years. A cyclic system can be used to position functional groups, often with control of regio- and stereochemistry. The ring is then opened to give an acyclic system, and the regio- and stereochemistry of the substituents has been fixed. There are many examples. The ozonolysis of cyclic alkenes was shown to generate α,ω-functionalized systems in Section 6.7.2. The utility of the process is demonstrated by the ozonolysis of (1Z,5Z)-cycloocta-1,5-diene to give diol (Z)-oct-4-ene-1,8-diol in 85% yield.50 Conversion of (Z)-oct-4-ene-1,8-diol to the racemic sex pheromone of the female face fly [2-decyl-3-(5-methylhexyl)oxirane] required four steps.50 Examining the overall sequence shows that the α- and ω-functional groups were incorporated by sequential alkylation reactions at the hydroxyl groups in (Z)-oct-4-ene-1,8-diol. Functionalization of each hydroxyl unit and then the alkene unit gave 2-decyl-3-(5-methylhexyl)oxirane.

49

Kitagawa, Y.; Itoh, A.; Hashimoto, S.; Yamamoto, H.; Nozaki, H. J. Am. Chem. Soc. 1977, 99, 3864.

(a) Tolstikov, G. A.; Odinokov, V. N.; Galeeva, P. J.; Bekeeva, R. S. Tetrahedron Lett. 1978, 1857. (b) Tolstikov, G. A.; Odinokov, V. N.; Galeeva, R. I.; Bakeeva, R. S.; Rafikov, S. R. Dokl. Akad. Nauk, SSSR 1978, 239, 1377 (Engl., p 174).

50

538

10. FUNCTIONAL GROUP EXCHANGE REACTIONS: SELECTIVITY

OH 1. O3

4 Steps

2. NaBH4

(1Z,5Z)-Cycloocta-1,5-diene

O C7H15

OH (Z )-Oct-4-ene-1,8-diol

2-Decyl-3-(5-methylhexyl)oxirane

(85%)

The synthesis of echinopine A by Chen and coworkers51 is another example of inserting and positioning functional groups, and it involves the conversion of 85 to 87. Treatment of ketone 85 with tosylhydrazine and trifluoroacetic acid to yield epoxy-hydrazone 86 in situ, and subsequent treatment with HCl led to fragmentation by what is known as Eschenmoser ring cleavage52 to keto-alkyne 87 in 95% yield. This named reaction transforms an epoxy-ketone to a ketoalkyne. OTBS

OTBS

OTBS

H

H

H TsNHNH2 , TFA

H

HCl

H

H

O

H

H

H

O

O O

NNHTs

85

86

87 (95%)

The diastereoselectivity obtained for reactions on a ring makes it possible to use a cyclic molecule to fix a stereocenter in an acyclic target, as seen in the formation of 90 from 88. In a synthesis of ceramides, Hung and coworkers53 converted D-glucosamine 88 to the azide 89 by a four-step sequence, with control of the stereocenters. Further elaboration of the functionality in seven steps led to ring opening to produce amino-tetraol 90, where the relative stereochemistry of this fragment was fixed by manipulation of the cyclic precursors. HO HO

O

OH OH

NH3 88

+Cl –

4 Steps

Ph

O

NH2

O

O

O2CPh OH

OH

7 Steps

N3

89

OH

OH

OH

C11 H23

90

10.6 BALDWIN’S RULES FOR RING CLOSURE Cyclic compounds play an important role in organic synthesis. The desired compound is not always commercially available, however, and must often be prepared by cyclization reactions from acyclic precursors. This finding is particularly true for large ring (macrocyclic) compounds and polycyclic molecules. In the latter case, a cyclic molecule acts as a template and the other rings are built onto the template (this is called annulation). This section will discuss the salient features of ring-forming reactions commonly encountered in synthesis. An introduction to cyclization reactions must include a discussion of Baldwin’s rules for ring closure, or simply Baldwin’s rules. Baldwin studied many nucleophilic, hemolytic, or cationic ring-closing processes and found a predictable pattern of reactivity. This approach is based on the stereochemical requirements of both reagent and substrate, as well as the angles of approach that are allowable when two reactive centers come together. To form a ring, the two reactive centers are connected by a tether of atoms, usually carbon atoms, but not always, and the length of the tether imposes constraints on the angles from which the reactive centers can approach one another, and on the stereochemistry of the product. If the length and nature of the chain (tether) linking terminal atoms X and Y allows this geometry to be attained, ring formation is possible (favored), and we make the predication that the reaction will succeed. If the proper geometry cannot be attained, ring formation is difficult (disfavored) and competitive processes often dominate. Baldwin and coworkers classified ring- closing reactions into two categories: exo 51

Peixoto, P. A.; Richard, J.-A.; Severin, R.; Chen, D. Y.-K. Org. Lett. 2011, 13, 5724.

52

M€ uller, R. K.; Felix, D.; Schreiber, J.; Eschenmoser, A. Helv. Chim. Acta 1970, 53, 1479.

53

Luo, S.-Y.; Kulkarni, S. S.; Chou, C.-H.; Liao, W. M.; Hung, S.-C. J. Org. Chem. 2006, 71, 1226.

539

10.6 BALDWIN’S RULES FOR RING CLOSURE

[the electron flow of the reaction is external to the ring being formed (92 from 91)], and endo [the electron flow is within the ring being formed 94 from 93)].54 Baldwin further classified reactions according to the hybridization of the atoms accepting the atom in the ring-closing process. If the atom being attacked is sp3 hybridized, as in 95, the reaction is termed tet, and an exo-tet reaction will generate a ring (e.g., 96). Attack at an sp2 atom (91) is termed trig (forming the ring 92 or 94). Attack at an sp hybridized atom (97) is dig, and an exo-dig reaction will generate ring 98. endo-trig

exo-trig

X

•

Y

X

•

X

Y

Y

X

• 91

92

93

•

94 exo-dig

exo-tet

X

Y •

Y

95

+

•

X

Y

X

96

Y

•

X

97

•

Y

98

It is therefore possible to describe ring-forming reactions by the number of atoms in the cyclic product, whether the reaction is exo or endo, and whether it involves tet, trig, or dig intermediates. A 5-exo-tet reaction represents formation of a five-membered ring by displacement at sp3 carbon by X, where Y is exo to the ring being formed. The cyclization reactions that form three- to seven-membered rings, along with all exo-endo, and tet-trig-dig possibilities are shown in Fig. 10.2.54

Y X-

Y

3-exo-tet

X-

Y

X-

4-exo-tet

X-

Y

5-exo-tet

Y

X-

X-

Y 7-exo-tet

6-exo-tet

5-endo-tet

Y X-

Y

X-

X3-exo-trig

6-endo-tet

Y

X-

X-

Y

X-

4-exo-trig

5-exo-trig

Y

Y

4-endo-trig

X-

X-

5-endo-trig

6-exo-trig

Y

X-

Y

6-endo-trig

7-endo-trig

3-exo-dig

Y

Y

Y

X-

X- Y

Y 4-exo-dig

Y

7-exo-trig

X-

X-

X-

X-

Y

Y

X-

3-endo-trig

Y

5-exo-dig

Y 7-exo-dig

6-exo-dig

X-

X-

3-endo-dig

4-endo-dig

Y X5-endo-dig

FIG. 10.2

X-

Y

6-endo-dig

X-

Y

7-endo-dig

Ring closures categorized by Baldwin’s rules.

(a) Baldwin, J. E. J. Chem. Soc. Chem. Commun. 1976, 734. (b) Baldwin, J. E.; Cutting, J.; Dupont, W.; Kruse, L.; Silberman, L.; Thomas, R. C. Ibid. 1976, 736.

54

540

10. FUNCTIONAL GROUP EXCHANGE REACTIONS: SELECTIVITY

When two reactive ends of a molecule come together, they can approach each other only from certain trajectories called the angle of attack. Before discussing Baldwin’s rules, it is first necessary to establish what angles of attack are most likely. Elliot and Graham-Richards55 described a method for predicting the preferred approach angles based only on the substrate. Note that these models assume a late transition state, so that the bond angles in the product are close to those in the important transition state (the Hammond postulate). Displacement at sp3 carbon generally requires backside attack (see 99) and the incoming group (X) must approach the Y-bearing carbon at an angle close to 180°. For exo processes, this is usually easy but, for endo processes it can be difficult. a X–

a

a = 180 o

Y

X

Y–

99

X–

X a

a = 109 o

a

Y

Y-

100

Reactions at double bonds (trig) are controlled by the planar nature of alkenes, imines, or carbonyls. The bond angles are 120°, but upon reaction the sp2 atom is converted to a sp3 atom. Since sp3 atoms are tetrahedral (with urgi-Dunitz trajecbond angles of 109°), the best trajectory for attack of sp2 atom (e.g., a carbonyl) is 109° (the B€ tory56; see 100 and Section 7.9.4). Treatment of 101 with AIBN and tributyltin hydride leads to a radical intermediate (102)57 (see Section 17.7 for a discussion of radical cyclization). Attack at the alkene unit easily allows approach at 109°, and cyclization yield 103. Hydrogen transfer to 103 from Bu3SnH completes the reaction, and generates the product 104. Formation of large rings is not a problem with respect to Baldwin’s rules, because there is sufficient flexibility in the tether to attain the required approach angle.58

ArO2S

n-Bu3SnH

N

AIBN

ArO2S

N

Br 102

101

CH2

CH2—H

X–H

ArO2S

N 103

ArO2S

N 104

Molecules containing a triple bond have a linear geometry, so the bond angles are 180°. Conversion of the sp atom to an sp2 atom with the 120° bond angle characteristic of alkenes. Using the same analogy as for the sp2-sp3 conversion, an incoming atom should approach the triple bond at an angle of 120° (see 105).59 Ring closure is difficult for reactions where the chain is small (a in 106). With longer chains (b and c in 107) the process may be more facile. For intramolecular reactions, attack at an angle of 120° or greater is possible in most cases (see 100).

55

Elliot, R. J.; Graham-Richards, W. J. Mol. Struct. 1982, 87, 247.

56

B€ urgi, H. B.; Dunitz, J. D.; Lehn, J. M.; Wipff, G. Tetrahedron 1974, 30, 1563.

(a) Padwa, A.; Nimmesgern, H.; Wong, G. S. K. J. Org. Chem. 1985, 50, 5620. (b) Idem Tetrahedron Lett. 1985, 26, 957. (c) Padwa, A.; Dent, W.; Nimmesgern, H.; Venkatramanan, M. K.; Wong, G. S. K. Chem. Ber. 1986, 119, 813. (d) Watanabe, Y.; Ueno, Y.; Tanaka, C.; Okawara, M.; Endo, T. Tetrahedron Lett. 1987, 28, 3953.

57

58

(a) Porter, N. A.; Magnin, D. R.; Wright, B. T. J. Am. Chem. Soc. 1986, 108, 2787. (b) Porter, N. A.; Chang, V. H.-T. Ibid. 1987, 109, 4976.

59

Baughman, R. H. J. Appl. Phys. 1972, 43, 4362.

541

10.6 BALDWIN’S RULES FOR RING CLOSURE

X–

a

C

a

C

120 o

X

a

C

a

Y+

a

C a

c

107

106

Y

105

b

These observations led Baldwin to establish several rules for ring closure, known as Baldwin’s rules:54 1. For tet systems 2. For trig systems

3. For dig systems

3–7 exo-tet are favored 5–6 endo-tet are disfavored 3–7 exo-trig are favored 6–7 endo-trig are favored 3–5 endo-trig are disfavored 5–7 exo-dig are favored 3–7 endo-dig are favored 3–4 exo-dig are disfavored

The disfavored reactions are not impossible, simply more difficult and usually slower than other competing reactions (inter- or intramolecular). Although some of the reactions show here will not be presented until Chapters 11–17, several examples of carbon-carbon bond-forming reactions will be given to illustrate these rules. Me

Me

OH Me

Br

Me

Me Me

0.2 M KOH

Me Me

aq MeCN

Me Me

Br

Me Me

Br

Me 110 Me

Me

Me OH

O

0.2 M KOH

Me Me 112

Me Me 111

O

+

109

108

Me

Me

O

O Me

aq MeCN

Me 113

When 108 reacted with KOH in aqueous acetonitrile, the 5-exo-trig product (109) was favored over formation of the 6-endo-trig product (110), by a ratio of 18:1.60 The alkynyl derivative (111) reacted 103 times slower than 108 and gave only the 5-exo-dig product (113). The 6-endo-dig product (112) was not isolated from the reaction, although the actual product distribution was probably >100:1 favoring 113 over 112. It was suggested that sp atoms generally prefer the exo mode of attack to the endo mode.60 There are reactions that seemingly contradict Baldwin’s rules. Presumably, these are reactions with an early transition state that will have different angles of attack. Fountain and Gerhardt61 showed that N-methyl-O-cinnamoyl hydroxylamine [O-cinnamoyl-N-methylhydroxylamine] gave 5-phenyl-3-pyrazolidinone [114, which exists as two valence tautomers that includes 3-(hydroxy(methyl)amino)-3-phenylpropanoic acid] under forcing conditions (15 h in dichlorobenzene heated at reflux), via the disfavored 5-endo-trig pathway. This reaction simply illustrates that the rules say favored or disfavored, not allowed or forbidden. It does point out, however, that factors other than geometry and stereochemistry must be considered. Fountain and Gerhardt61 suggested that electronic factors must be considered. Anselme62 showed that cyclization of 115 to 116 occurred via a 5-endo-trig process, but under more vigorous conditions than noted by Fountain and Gerhardt for the cyclization of 111. Presumably, the nitrogen lone pair electrons in 115 are more nucleophilic than in CO2NHMe precursor to 114.62

60

Evans, C. M.; Kirby, A. J. J. Chem. Soc. Perkin 1984, 2, 1269.

61

Fountain, K. R.; Gerhardt, G. Tetrahedron Lett. 1978, 3985.

62

Anselme, J. P. Tetrahedron Lett. 1977, 3615.

542

10. FUNCTIONAL GROUP EXCHANGE REACTIONS: SELECTIVITY

OH Me

Ph PhBr , reflux

ONHMe

N:

Ph

5h

H

OH2

O

N

Me

O-Cinnamoyl-N-methylhydroxylamine O

Ph

–H2O

Me

O

Cl Cl

N H

N

OH

O N N

Reflux , 15 h

Ph

CO2H

3-(Hydroxy(methyl)amino)3-phenylpropanoic acid

114

O

N

O

O

MeOH

O

Ph

O

115

116

Epoxides and other three-membered rings present special problems since they are in between tet and trig systems. Generally, three-membered rings prefer the exo mode of attack (see 117) to yield 118.63 The endo mode of attack generates 119. If 120 is used as an example, protonation of the epoxide oxygen with an acid (camphorsulfonic acid) followed by attack by the OH unit and ring opening gave the observed products.64 When R was CH2CH2CO2Me, the 5-exo-product 122 was the only one (94% yield), which is typical. When the R substituent in 120 was changed from alkyl to alkenyl (vinyl), however, the course of ring opening was changed to give only the 6-exo product (121) in 90% yield. In general, the 5-exo mode of attack is preferred. Y

exo

endo X endo

Y

X

exo

Y 119

117

118

HO

Camphorsulfonic acid

O

HO H

+

CH2Cl2 , –40°C

OH

R

X

R

120

R

O

H

121 (90%)

O 122

Three-membered rings other than an oxirane ring follow the same ring-opening pathways. An example of such an opening is the iodolactonization reaction discussed in Section 2.5.3. A synthetic example is taken from Imanishi and coworker’s65a synthesis of leustroducsin B, in which (R)-cyclohex-3-ene-1-carboxylic acid was treated with bicarbonate in the presence of iodine and KI, to yield carboxylate anion-iodonium intermediate 123. Intramolecular displacement of iodide by carboxylate (path a) is a 5-exo process, and is favored over the analogous 6-endo process (path b) to yield iodolactone 124 in >80% yield. Note that this reaction may not proceed by the discreet iodination addition product 123, in which case Baldwin’s rules do not apply. I HO2C

I

KI , I2 NaHCO3

H HO

CH2Cl2-H2O rt , 24 h

O

(R)-Cyclohex-3-ene-1-carboxylic acid

O b a O

H O

O

H O

I

O

123

63

Stork, G.; Cohen, J. F. J. Am. Chem. Soc. 1974, 96, 5270.

64

Nicolaou, K. C.; Duggan, M. E.; Hwang, C.-K.; Somers, P. K. J. Chem. Soc, Chem. Commun. 1985, 1359.

124 (>80%)

(a) Miyashita, M.; Tsunemi, T.; Hosokawa, T.; Ikejiri, M. Imanishi, T. Tetrahedron Lett. 2007, 48, 3829. For iodolactnization, see (b) Klein, J. J. Am. Chem. Soc. 1959, 81, 3611. (c) van Tamelen, E. E.; Shamma, M. Ibid. 1954, 76, 2315. 65

543

10.6 BALDWIN’S RULES FOR RING CLOSURE

Third-row elements are difficult to analyze by Baldwin’s rules because the atomic radii are larger and the bond lengths are large. Conformations can be attained that are unavailable to molecules that contain only second-row atoms. The constraints on approach angles are therefore less stringent. The thiol methyl 4-mercapto-2-methylenebutanoate, for example, was cyclized to methyl tetrahydrothiophene-3-carboxylate by a 5-endo-trig process (an internal Michael addition).54b,66 The sulfur atom has 3d-orbitals and can receive electrons via back donation from the occupied p-orbitals (see 125). This ability diminishes the requisite angle of attack from 109° to 90° or less, facilitating the endocyclic ring closure. CO2Me

1. NaOMe MeOH , 65°C

CO2Me

2. H3O+

SH

S

Methyl 4-mercapto-2methylenebutanoate

Methyl tetrahydrothiophene-3-carboxylate

Baldwin’s rules also apply to ketone enolate anions,67 but Baldwin and Lusch68 modified the rules to make the terminology more specific. The special angular requirements of enolate anions are shown for the exo-tet process (126), and the exo-trig process (127). For these two ring-closing reactions, the p-orbitals of the enolate anion must approach the reactive center at an angle of 180° or 109°, respectively. The orientation of the orbital and not just the nucleophilic atom (C or O) must be considered for determining the correct angle of approach (Section 13.4.1.2).

S

C

C C

C

C

C

C

C

H C X H

125

H

126

C X

127

There are two ways to view the ring closure. A kinetic enolate anion (e.g., 128), which was less substituted, was derived from removal of the most acidic proton (Section 13.2.5) and will close via exo displacement of Y to yield 129. A thermodynamic enolate anion (e.g., 130), which was more substituted resulted from equilibration to the most stable enolate anion product and also undergoes an exo displacement of Y, but yield 131. Conversion of 128 to 129 was termed enolendo-exo-tet, and conversion of 130 to 131 was termed enolexo-exo-tet. Similarly, exo-trig reactions can be classified in this manner. The kinetic enolate anion (132) yield 133 in an enolendo-exo-trig process. Cyclization of the thermodynamic enolate anion (134) yield 135 via an enolexo-exo-trig reaction. The rules are somewhat different for ring closure with enolate anions. 6–7 enolendo-exo-tet favored 3–7 enolexo-exo-tet favored 6–7 enolendo-exo-trig favored

−O

•

•

•

enolendo-exo-tet

128

CH

O

Y

•

O Me

131

CH2 133

•

Y

•

enolexo-exo-trig

•

Y−

O Me

Me 130

CH O

Y−

•

132

•

enolexo-exo-tet

•

enolendo-exo-trig

Y O

129

•

•

•

CH2

O

−

O

•

Y

3–5 enolendo-exo-tet disfavored 3–7 enolexo-exo-trig favored 3–5 enolendo-exo-trig disfavored

Me 134

135

See (a) Claeson, G.; Jonsson, H.-G. Arkiv. Kemi 1967, 28, 167 (Chem. Abstr. 1968, 68, 87083e). (b) Idem Ibid. 1966, 26, 247 (Chem. Abstr. 1967, 66, 85334x).

66

67

Baldwin, J. E.; Kruse, L. I. J. Chem. Soc. Chem. Commun. 1977, 233.

68

Baldwin, J. E.; Lusch, M. J. Tetrahedron 1982, 38, 2939.

544

10. FUNCTIONAL GROUP EXCHANGE REACTIONS: SELECTIVITY

Two competing intramolecular cyclizations (see the aldol condensation, Section 13.4.1) are illustrated for triketone 136.68 In this case, only the favored 6-enolendo-exo-trig product (137) was obtained, and not the disfavored 5-enol-endo-exo-trig products (138 and 139). Enolate anions generally show 5-exo-tet reactions at oxygen, and 6-enolendo-exo-tet reactions at carbon. O

5-enolendo exo-trig

O +

O

O

6-enolendo exo-trig

O

O

O

O

138

O

139

136

137

In order to form a five-membered ring from 140A, the oxygen lone pair must be at the correct angle for displacement of bromine (the product is a THE, 3,3-dimethyl-2-methylenetetrahydrofuran). Another view of 140A shows the positioning of the orbitals that will allow displacement (see 140B). In order to generate a six-membered ring, however, the orbitals of the enolate carbon in 141A must be at the correct angle for displacement (see 141B) and the product will be 2,2-dimethylcyclohexan-1-one. Br

Me Me

Br

O C

O

C

O 140A

3,3-Dimethyl-2-methylenetetrahydrofuran

140B

Br O Me

Me

C O

Br

O C

141A

141B

2,2-Dimethylcyclohexan-1-one

10.7 CONCLUSION The intent of this chapter was to tie together many of the principles introduced in earlier chapters. In addition, many of the principles used in the important chapters dealing with making carbon-carbon bonds have been discussed. If there is a theme to this chapter, it is that organic chemists can exercise a significant amount of control over synthetic reactions by understanding the mechanism of the trans-formations and the various interactions of heteroatoms and reagents. A thorough understanding of the conformational aspects of reactivity is essential. With knowledge of these principles, a useful plan can be assembled for the synthesis of even complex molecules. Without this understanding, syntheses beyond a few simple steps will be difficult. One last piece of information is required for synthesis in order to manipulate functional groups. In those cases where a heteroatom functional group interferes with a transformation and cannot be removed, modified, or inserted earlier or later in a synthesis, methods are available to temporarily block it. This method is termed protection and the moiety used to block the reactive fragment is called a protecting group, which was discussed in Chapter 5.

545

10.7 CONCLUSION

HOMEWORK

1. In each of the following, give the major product and classify each reaction according to Baldwin’s rules: I 1. BuLi , THF

(A)

SPh CN

2. H3O

H2 N

NaH

(B)

+

Heat

(C) O HO O Me

OH

CHO 1. LiN(i-Pr) , THF , –78°C 2

NaH

(D)

(E)

Me

Br

(G)

(F)

1. NaOEt , EtOH , Reflux

2. H3O + See Section 13.4.1

O

1. t-BuS– Tl

CO2H

HO

O

2. H3O+ See Section 13.4.1

+

2. Hg(OAc)2

2. Explain why the following sequence results in inversion of configuration of the hydroxyl group: O O

O Ph

O

1. PCC , CH2Cl2

N

Ph N

2. L-Selectride 3. H3O +

OH

OH

3. Explain this transformation by giving the three intermediate products. Me

Me

O 1. i. NaBH4 2. H3O +

ii. H3O+

3. mcpba 4. H3O +

MeO

O MeO

For related reactions see (a) Fieser, L. F.; Fieser, M. Advanced Organic Chemistry; Reinhold: NY, 1961, pp 897–898. (b) Technol. Repts. Osaka Univ. 1958, 8, 455. (Chem. Abstr. 1959, 53, 18920i.) 4. Draw the major product of this reaction and predict the correct stereochemistry. Briefly explain your answer.

H HO

1. NaH , DMF

H O

2. dil H3O +

5. Suggest a reason why the carboethoxy group epimerizes upon treatment with potassium carbonate. Why is the second step required? CO2t-Bu N

1. K2CO3 , MeOH , 80°C 2. SOCl 2 , EtOH

CO2t-Bu N CO2Et

CO2Et (95%)

546

10. FUNCTIONAL GROUP EXCHANGE REACTIONS: SELECTIVITY

6. Offer a mechanistic rationale for the following conversion that explains the stereoselectivity of the reaction: O

O 1. Bu2Cu(CN)Li2 , THF

OH

2. dil H3O+

O

n-C4H9 See Section 12.3.1.3

7. Show the structure of both A and B. Each of these reactions has a name associated with it. Give the name of each reaction and look up one recent review of each. TMS

Ti(Oi-Pr) 4 (+)-DIPT

Me O Si Me

t-BuOOH

O2N

CO2H

B

A EtO2C-N=N-CO2Et PPh3 , THF

TMS = SiMe 3

OH

8. Predict the major product of this reaction. Explain your answer in the context of other possible products. Br

KOH , EtOH Reflux

9. Give the major product of the following reaction O OH Me

OMe

O

OMe

1. TPAP, NMO, CH 2Cl2 2. NaBH4 , AcOH, THF

O Me

OH

10. In each case, provide a suitable synthesis. Show all intermediate products and all reagents. OCH2OMe

OH

NH2

OH

(B)

(A) OH OH

O

OH

O

(C) Me

(E)

OH

HO2C

OH

(D) Me

O

Me

CO2Me

C H A P T E R

11 Carbon-Carbon Bond-Forming Reactions: Cyanide, Alkyne Anions, Grignard Reagents, and Organolithium Reagents 11.1 INTRODUCTION The formation of new carbon-carbon bonds is a major component of any synthesis of a complex molecule. Introduction of new functional groups that are compatible with existing functional groups, and also with the reagents that are necessary to complete the transformation further complicates the problem. Finally, the problems of regioselectivity, stereoselectivity, chemoselectivity, and absolute configuration of the stereogenic centers must be addressed in a reasonable manner. In this book, the major reactions that build new carbon-carbon bonds have been classified into four major types: nucleophilic, electrophilic, pericyclic, and radical reactions. This chapter will focus on those reactions that employ a nucleophilic carbon. Referring back to Section 1.2, disconnection of a bond generated two fragments (RdCa and RdCd), where RdCd was a donor or nucleophilic fragment and RdCa was an acceptor or electrophilic fragment. The main purpose of this chapter is to discuss molecules that mostly contain Mg or Li, as well as the cyanide ion and alkyne anions. All of these compounds function as Cd.

11.2 CYANIDE 11.2.1 Formation of Nitriles Cyanide ion is a good nucleophile that reacts with suitable substrates to generate carbon-carbon bonds. When cyanide displaces a leaving group in a SN2 reaction, the carbon chain is extended by one carbon, and the resulting nitrile can be converted to a variety of other functional groups.1 Williamson reported the displacement reaction of ethyl ohler and Liebig3 halides or sulfonate esters by cyanide ion to yield propanenitrile (CH3CH2CH2CN)1 in 1854,2 but W€ had synthesized the first nitrile in 1832. Since the reaction with an alkyl halide is an SN2 process (Section 3.2.1.1), the best yields of nitrile are obtained with primary and secondary substrates, whereas tertiary halides (e.g., 2-chloro-2-methylbutane sometimes) react via elimination to yield the alkene (Section 3.5.1).4 Cyanide displacement of secondary halides sometimes gives poor yields, however. The use of polar, aprotic solvents (e.g., DMSO, DMF, or THF) lead to the best yields of nitriles (e.g., hexanenitrile, capronitrile, 97% yield after 20 min (Section 3.2.1.9).

1

Mowry, D. T. Chem. Rev. 1948, 42, 189.

2

Williamson, A. E. J. Prakt. Chem. 1854, 61, 60.

3

W€ ohler, F.; Liebig, J. Annalen 1832, 3, 249, 267.

4

Hass, H. B.; Marshall, J. R. Ind. Eng. Chem. 1931, 23, 352.

Organic Synthesis http://dx.doi.org/10.1016/B978-0-12-800720-4.00011-8

547

Copyright © 2017 Michael Smith. Published by Elsevier Inc. All rights reserved.

548

11. CARBON-CARBON BOND-FORMING REACTIONS DMSO

RdX + NaCN ƒƒƒƒƒƒ! RdC^N + NaX X ¼ Br, Cl, I, OSO2 R Sulfonate ester leaving groups can also be used, as in the conversion of alcohol 1 to the tosylate in 90% yield followed by reaction NaCN in DMF to give 2 in 87% yield, in a synthesis of (+)-chloriolide by Ostermeier and Schobert.5 This variation allows the alcohol ! sulfonate ester ! nitrile transformation. Another common reagent that generates a nitrile is tosylmethylisocyanide (TsCH2N^C, also called TosMIC),6 which reacts with ketones or aldehydes to generate a nitrile (R2C]O ! R2CH-C^N).6 HO

O

O

O

1. Tos-Cl, Py, 1d CH2Cl2, rt

(90%)

2. NaCN , DMF 60°C, 22 h

(87%)

NC

O

O

O

2 (78%)

1

Nitriles are versatile synthetic intermediates since they can be converted to several other functional groups. Three important reactions of nitriles are hydrolysis to acid derivatives,7 reduction to amines (Sections 7.6.2.5 and 7.10.5), and reduction to aldehydes8 (Sections 7.6.2.5 and 7.10.5). In this chapter, nitriles will be shown to react with both Grignard reagents (Section 11.4.3.3) and organolithium reagents (Section 11.6.6) to form ketones. Treatment of nitriles containing an α-proton with a strong base, (e.g., lithium diisopropylamide) generates the α-carbanion, which can react with alkyl halides, aldehydes, or ketones (Section 13.4.2.4) or another nitrile (see the Thorpe reaction in Section 13.4.2.4). Reactions of cyanide are not limited to alkylation of halides and sulfonate esters. Epoxides are also opened at the less sterically hindered carbon by cyanide via an SN2 like process.9 Disconnection of a CdCN bond is particularly attractive when the CN unit is attached to a primary or a secondary carbon, which reflects the relative ease of formation of this bond by the SN2 reaction, with clean inversion of configuration, and the variety of functional group transformations available. The disconnection for nitriles can be simplified as follows: R J CL N

RJX

11.2.2 Isonitriles: Formation and Reactions Both potassium and sodium cyanide react with alkyl halides to give excellent yields of the nitrile in the solvent DMSO. An isomeric product is often observed when the reaction is done in an alcohol solvent heated at reflux, or when certain metal cyanides are used. This isomer is known as an isonitrile, also called an isocyanide.10 Gautier and Hofmann were the first to report an isonitrile.11 Cyanide is a bidentate nucleophile, and based strictly on electronegativity, C is a better nucleophile than N. The reactivity of the carbon versus the nitrogen nucleophile depends on the metal counterion, however. If M+ in MCN is Na or K, which favor formation of ionic compounds, cyanide reacts primarily as a carbon nucleophile. If M+ is a metal such as Ag, which forms relatively covalent bonds, nucleophilic strength at carbon is diminished and cyanide reacts primarily as a nitrogen nucleophile. These competing pathways are illustrated by the reaction of MCN with 1-bromopentane to yield hexanenitrile or 1-isocyanopentane. Reaction with potassium cyanide (KCN) generates hexanenitrile, whereas reaction with silver cyanide (AgCN) yield the isomeric isonitrile, 1-isocyanopentane.

5

Ostermeier, M.; Schobert, R. J. Org. Chem. 2014, 79, 4038.

6

Oldenziel, O. H.; van Leusen, D.; van Leusen, A. M. J. Org. Chem. 1977, 42, 3114.

7

For an example taken from a synthesis of (R)-(+)-myrmicarin 217, see Sayah, B.; Pelloux-Leon, N.; Vallee, Y. J. Org. Chem. 2000, 65, 2824.

8

For an example taken from a synthesis of laulimalide, see Mulzer, J.; Hanbauer, M. Tetrahedron Lett. 2000, 41, 33.

9

Kergomard, A.; Veschambre, H. Tetrahedron Lett. 1976, 4069.

(a) Guillemard, H. Compt. Rend. 1906, 143, 1158; (b) Guillemard, H. Ibid. 1907, 144, 141; (c) Guillemard, H. Ibid. 1907, 144, 326; (d) Guillemard, H. Bull. Soc. Chim. Fr. 1907, 1, 269; (e) Guillemard, H. Ibid. 1907, 1, 530.

10

11

(a) Gautier, A. Ann. Chim. Paris 1869, 17, 103; (b) Gautier, A. Annalen 1867, 142, 289; (c) Hofmann, A. W. Ibid. 1867, 144, 114.

549

11.2 CYANIDE S N2 ⬙C⬙ M=K

C LN

M :CN: +

S N2 ⬙N⬙ M = Ag

Br

NL C:

1-Bromopentane

Hexanenitrile

1-Isocyanopentane

If the reaction is done in alcoholic solvents, which minimizes the nucleophilicity of ionic species due to solvation effects, the yield of isonitrile is improved. The nature of the metal in metal cyanides has a profound influence on the course of a reaction with alkyl halides.1 The reaction of alkyl halides with metal cyanides that have a counterion that forms more ionic bonds (e.g., NaCN or KCN) gives primarily the nitrile. However, silver and copper cyanide, which form highly covalent metal cyanides, give good yields of the corresponding isonitrile under the same conditions. Isonitrile formation is due, in part, to the formation and stability of intermediates (e.g., 3). Reaction of an alkyl halide with silver cyanide is a preparative route to isonitriles, but they can also be formed by dehydration of formamide derivatives.12 The hydrolysis of isonitriles with cold, dilute aqueous acid leads to an amine and formic acid. The reaction of cyanide with an alkyl halide may give a mixture of nitrile and isocyanide. An isocyanide is selectively hydrolyzed upon treatment of a mixture with aqueous acid, allowing the nitrile to be isolated. AgCN

RJX

Ag

R C

RJNL C:

N

X Isonitrile

3

Isocyanides can also be converted to the nitrile by thermal rearrangement, but the reaction requires temperatures of between 140 and 240°C.1 Heating 3-α-isocyanocholestane to 270°C, for example, gave 3-α-cyanocholestane. Interestingly, the reaction proceeded with >99% retention of configuration.13 If the isonitrile does not contain thermally sensitive functionality, this reaction sequence constitutes a good synthesis of nitriles. Isonitriles are also formed by formylation of a primary amine followed by dehydration, as in the conversion of 4 to 5 in a synthesis of the hapalindoles J and H by Williams and Rafferty.14 Initial formation of the formamide from 4 was followed by conversion to the isonitrile by reaction with Burgess reagent (see Section 3.7.5). Note that isocyanates15 are easily formed from isonitriles by reaction with halogens.15b Me Me

Me

C8 H17 Me

270°C

H

H H

H

H

H

C8 H17

NLC

CLN 3- -Isocyanocholestane

3- -Cyanocholestane

Me Me

Me

H

Me

H

NH2

1. HCO2H , CDMT 2. Burgess reagent

Me

H

Me

H

NC

CDMT = 2-chloro-4,6-dimethoxy1,3,5-triazine

N H 4

N H 5 (53%)

12

Ugi, I. Isonitrile Chemistry; Academic Press: New York, NY, 1971.

13

Horobin, R. W.; Khan, N. R.; McKenna, J.; Hutley, B. G. Tetrahedron Lett. 1966, 5087.

14

Rafferty, R. J.; Williams, R. M. J. Org. Chem. 2012, 77, 519.

15

(a) Nef, J. U. Annalen 1892, 270, 267; (b) Johnson, Jr., H. W.; Daughhetee, Jr., P. H. J. Org. Chem. 1964, 29, 246.

550

11. CARBON-CARBON BOND-FORMING REACTIONS

Isonitriles can be used for a few useful synthetic transformations.10,16 The Passerini reaction17 is a convenient route to amide esters, and it is categorized as a three-component coupling reaction.18a When propanoic acid was heated with acetone and tert-butylisonitrile, for example, the product was α-propanoyloxy amide [1-(tert-butylamino)-2methyl-1-oxopropan-2-yl propionate].17a A useful modification of the Passerini reaction used trifluoroacetic acid rather than a normal aliphatic carboxylic acid. When 2-phenoxyacetaldehyde reacted with tert-butylisonitrile and TFA, 6 was formed. Hydrolysis of this trifluoroacetyl ester with aqueous sodium carbonate led to isolation of the α-hydroxyamide, N-(tert-butyl)-2-hydroxy-3-phenoxypropanamide, in 69% yield.19 This modification has the advantage that hydrolysis conditions for trifluoroacetate esters are milder than for normal esters. Note that the four-component coupling of an isonitrile, an amine, an acid, and an aldehyde or ketone to generate an amido amide is called the Ugi reaction.18b–d O

O

O Me

OH

Me

O

NLC

t-Bu

O PhO

CHO

NLC

N Me

t-Bu

OH

O

F3C PhO

CF3CO2H , –50°C CHCl3

H

O 1-(tert-Butylamino)-2-methyl1-oxopropan-2-yl propionate

Propanoic acid

t-Bu

Me

aq Na2CO3

PhO

NHt-Bu

NHt-Bu O

O 2-Phenoxyacetaldehyde

N-(tert-Butyl)-2-hydroxy-3- (69%) phenoxypropanamide

6

Isonitriles are somewhat limited in their synthetic utility, but they have some interesting disconnections: R1

O R

O

NHR3 R2

RCO2H + + R3JNLC

R1

O

O R2

R3JNLC

R3JX

RJNKCKO

R3JNLC

11.2.3 Other Nitrile-Forming Reactions Aryl halides are inert to reaction with cyanide ion under normal SN2 conditions (Section 3.2.1.1), but they do react when heated with cuprous salts like cuprous cyanide (CuCN).20 This transformation is called the Rosenmund-von Braun reaction,21 can be applied to a wide range of aryl halides, and the products can subsequently be used in a variety of reactions.22 This transformation does not work when potassium or sodium cyanide is used. An amine base (e.g., N-methylpyrrolidine) is often added to facilitate the reaction. A synthetic example is taken from a synthesis of stachybotrylactam by Kende et al.,23 in which aryl bromide 7 was treated with CuCN in hot DMF to give a 92% yield of aryl nitrile 8.

16

(a) Ferosie, I. Aldrichim. Acta 1971, 4, 21; (b) D€ omling, A.; Ugi, I. Angew. Chem. Int. Ed. 2000, 39, 3168.

(a) Passerini, M. Gazz. Chim. Ital. 1921, 51 II, 126; (b) Passerini, M. Ibid. 1924, 54, 529; (c) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006, p ONR-69; (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, Jr., F. G. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; WileyInterscience: NJ, 2005, pp 482–483.

17

(a) For a review, of isocyanide based multi-component coupling reactions, see D€ omling, A. Chem. Rev. 2006, 106, 17. For information concerning the Ugi reaction, see (b) Lindhorst, T.; Bock, H.; Ugi, I. Tetrahedron 1999, 55, 7411; (c) D€ omling, A.; Ugi, I. Angew. Chem. Int. Ed. 2000, 39, 3168. See pp 3183–3184 and references therein. Also see (d) Tye, H.; Whittaker, M. Org. Biomol. Chem. 2004, 2, 813. 18

19

Lumma, Jr., W. C. J. Org. Chem. 1981, 46, 3668.

20

Newman, M. S.; Boden, H. J. Org. Chem. 1961, 26, 2525.

21

(a) Rosenmund, K. W.; Struck, E. Berichte 1916, 52, 1749; (b) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006, p ONR-81.

(a) Ito, T.; Watanabe, K. Bull. Chem. Soc. Jpn. 1968, 41, 419; (b) Akanuma, K.; Amamiya, H.; Hayashi, T.; Watanabe, K.; Hata, K. Nippon Kagakai Zashi 1960, 81, 333; (Chem. Abstr. 56: 406i, 1962); (c) Cohen, T.; Lewin, A. H. J. Am. Chem. Soc. 1966, 88, 4521.

22

23

Kende, A. S.; Deng, W. -P.; Zhong, M.; Guo, X. -C. Org. Lett. 2003, 5, 1785.

551

11.2 CYANIDE

BnO

HO

BnO

Br

CN

HO CuCN, DMF

CO2Me H

CO2Me

100°C

O

O

H

7

8 (92%)

Aliphatic or aromatic carboxylate salts are also nitrile precursors. For example, fusion (250–300°C) of sodium propionate with cyanogen bromide24 gave propanenitrile.,24b,25 Cyanide is a nucleophile that reacts with carbonyl derivatives via acyl addition, although this is usually a reversible reaction. The classical Strecker synthesis1,26 used to prepare racemic α-amino acids is based on controlled addition to yield a cyanohydrin [RCHCN(OH)]. When 2-phenylacetaldehyde was treated with NH3 and HCN, for example, the product was the amino nitrile, 2-amino-3phenylpropanenitrile. Hydrolysis of the nitrile generated the amino acid, phenylalanine (2-ammonio-3phenylpropanoate).26 Ph

NH3 , HCN

CHO

Ph

CN

H3O+

CO2

Ph

NH2 2-Phenylacetaldehyde

NH3

2-Amino-3-phenylpropanenitrile

Phenylalanine

Me

1. Swern oxidation 2. TMSCN, ZnCl2 CH2Cl2 , rt

AllylO H

Me

N H N

MeO

AllylO H

Me

N OH

H

N

MeO AllylO

AllylO

Me

MeO

MeO

OBn 9

NC

Alloc

BnO 10 (87% overall)

There are many modifications of the Strecker synthesis,1 including those by Erlenmeyer, Tiemann, Zelinsky, and Stadnikoff, or Knoevenagel and Bucherer. In all cases, the reaction produces racemic amino acids, but newer methodology introduces enantioselectivity into the process. One modification of the original Strecker procedure uses a chiral amine to produce amino acids with good asymmetric induction.27 Another example of a stereoselective Strecker reaction used (R)-phenylglycinol as chiral auxiliary for the diastereoselective synthesis of (R)-amino acids.28 In one example by Chen et al.,29 taken from a synthesis of ()-jorunnamycin A, an intramolecular Strecker reaction with the aldehyde formed by Swern oxidation of 9, catalyzed by zinc chloride, gave 10 in 87% yield. Catalytic, enantioselective Strecker reactions are also known.30

(a) Douglas, D. E.; Eccles, J.; Almond, A. E. Can. J. Chem. 1953, 31, 1127; (b) Douglas, D. E.; Burditt, A. M. Ibid. 1958, 36, 1256; (c) Barltrop, J. A.; Day, A. C.; Bigley, D. B. J. Chem. Soc. 1961, 3185.

24

25

Payot, P. H.; Dauben, W. G.; Replogle, L. J. Am. Chem. Soc. 1957, 79, 4136.

(a) Strecker, A. Annalen 1850, 75, 27; (b) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006, p ONR-90; (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, Jr., F. G. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005, pp 632–633.

26

27

Weinges, K.; Stemmle, B. Chem. Ber. 1973, 106, 2291.

(a) Chakraborty, T. K.; Hussain, K. A.; Reddy, G. V. Tetrahedron 1995, 51, 9179. Also see (b) Ma, D.; Tian, H.; Zou, G. J. Org. Chem. 1999, 64, 120; (c) Ferraris, D.; Young, B.; Cox, C.; Dudding, T.; Drury, III, W. J.; Ryzhkov, L.; Taggi, A. E.; Lectka, T. J. Am. Chem. Soc. 2002, 124, 67; (d) Kobayashi, S.; Ishitani, H. Chem. Rev. 1999, 99, 1069; (e) France, S.; Guerin, D. J.; Miller, S. J.; Lectka, T. Chem. Rev. 2003, 103, 2985.

28

29

Chen, R.; Liu, H.; Chen, X. J. Nat. Prod. 2013, 76, 1789.

30

For a review, see Groger, H. Chem. Rev. 2003, 103, 2795.

552

11. CARBON-CARBON BOND-FORMING REACTIONS

Other acid derivatives can also be prepared with HCN and carbonyl derivatives. Addition of HCN to a ketone or aldehyde gives a cyanohydrin in a reversible reaction, and hydrolysis gives the α-hydroxy acid.1,31 This process is an equilibrium process favoring the cyanohydrin for aliphatic aldehydes and ketones. The yields are usually poor with aryl and diaryl ketones, however. These reactions add the following disconnections to those previously shown for nitriles: R1

Ar JCLN

Ar JX

R

R1

O

R2

OH R

CO2H

R1

R1

NR2

O CLN

R2

CO2R

NR2

R2

R1

11.3 ALKYNE ANIONS (RdC^C:2) The conjugate base of an alkyne is an alkyne anion (older literature refers to them as acetylides), and alkyne anions are generated by reaction with a strong base and they are carbanions. Alkyne anions function as a nucleophile (a source of nucleophilic carbon) in SN2 reactions with alkyl halides and sulfonate esters. The product is a different alkyne. Alkyne anions also react with ketones or aldehydes via nucleophilic acyl addition, and with acid derivatives via nucleophilic acyl substitution to yield an alcohol. From the standpoint of retrosynthesis, acetylides, or other alkyne anions are important carbanion synthons for the formation of new carbon-carbon bonds. Some of the chemistry presented here will deal with the synthesis of alkynes from Section 3.5 in order to provide continuity with the discussion of acetylides.

11.3.1 Preparation of Alkynes Alkynes have been thoroughly studied over many years32 and many methods are available for their preparation.33 One of the most useful is dehydrohalogenation of dihalides. This method works best for aryl derivatives when transelimination is possible.34 Treatment of 1,2-bis(1,2-dibromoethyl)benzene with potassium tert-butoxide gave the usual E2 elimination (Section 3.5.1) to yield (E)-1-(1-bromovinyl)-2-(2-bromovinyl)benzene. Subsequent base induced dehydrohalogenation yield mono-alkyne (E)-1-(2-bromovinyl)-2-ethynylbenzene, since only one trans-elimination was possible.35 Treatment of (E)-1-(2-bromovinyl)-2-ethynylbenzene with a second equivalent of base gave bis(alkyne) 1,2-diethynylbenzene, but a cis-elimination pathway was required for this transformation. This second elimination reaction was more difficult, requiring excess base and more vigorous conditions. Br Br

Br

Br Dioxane

2 t-BuOK

t-BuOK , t-BuOH

Dioxane

PhH, Reflux

Br Br 1,2-Bis(1,2-dibromoethyl)benzene

CLCH t-BuOK

Br (E)-1-(1-Bromovinyl)-2(2-bromovinyl)benzene

(E)-1-(2-Bromovinyl)2-ethynylbenzene

CLCH 1,2-Diethynylbenzene

As the chain length of the vinyl halide increases, the yield of terminal alkyne product generally decreases.23,36 Vinyl bromide 11 (n ¼ 0) gave 61% of alkyne 12 (n ¼ 0), for example, but 1-bromohex-1-ene (11, n ¼ 4) gave only 18% of the corresponding alkyne (12, n ¼ 4).36 Alkynes can also be formed from nonterminal (internal) vinyl halides (e.g., 3-bromohept-3-ene), but when alkoxide bases are used, the vigorous conditions usually required for elimination of the halogen can cause the triple bond of the product to migrate to the terminal position.37 31

Reference 1, pp 231–246.

32

Viehe, H. R., Ed.; Chemistry of Acetylenes; Marcel-Dekker: New York, NY, 1969.

33

Larock, R. C. Comprehensive Organic Transformations, 2nd ed.; Wiley-VCH: New York, NY, 1999, pp 569–581.

34

(a) Reference 32, pp 100–168; (b) Fiesselmann, H.; Sasse, K. Chem. Ber. 1956, 89, 1775.

35

Gaudemar-Bardone, F. Ann. Chim. (Paris) 1958, 3, 52.

36

Loevenich, J.; Losen, J.; Dierichs, A. Berichte 1927, 60, 950.

37

Vaughn, T. H. J. Am. Chem. Soc. 1933, 55, 3453.

553

11.3 ALKYNE ANIONS (RdC^C:)

H (CH2)n

C H

NaOEt , EtOH

CHBr

H (CH2)n

Sealed tube 100°C

11

H

12

1,1-Dibromoalkenes are a useful source of terminal alkynes, which give an alkyne anion when treated with n-butyllithium. In a synthesis of (+)-demethoxycardinalin 3 by Fernandes and Mulay,38 the Corey-Fuchs procedure,39 which is related to the Wittig olefination reaction (see Section 12.5.1.1), was used to convert aldehyde 13 to the dibromoalkene with CBr4 and PPh3, and subsequent elimination induced by reaction with butyllithium gave 13, in 80% overall yield. 1. CBr4 , PPh3 , 0°C CH2Cl2 , 3.5 h

TBDMSO CHO

TBDMSO

2. BuLi , THF , –78°C, 2 h rt , 1 h

14 (80%)

13

11.3.2 Acidity of Terminal Alkynes Acetylene and other terminal alkynes have an acidic hydrogen atom (C^CdH), and they are weak acids. However, a strong base is required to remove that proton to yield an alkyne anion, but this carbanion is a carbon nucleophile that reacts with alkyl halides or sulfonate esters (R1dX, where R1 ¼ Me, 1°, 2° alkyl, and X ¼ halogen or OSO2R) via an SN2 sequence to yield disubstituted alkynes (e.g., 16). In a synthesis of ()-lycoposerramine-S by Fukuyama and coworkers,40 alkyne 15 was treated with butyllithium to generate the corresponding alkyne anion, which reacted with the primary alkyl iodide to give an 87% yield of 16. 1. BuLi , BuLi/DMPU

TBSO

2.

15

I

OTBS

TBSO

–78°C to rt

OTBS 16 (87%)

The relative acidity of various weak acids has been determined, including terminal alkynes, alkenes, ketones, and alcohols.41 The acidity of a proton is enhanced (smaller pKa) as the s character of the carbon to which it is attached increases.42 This observation is consistent with the fact that alkynes have a pKa of 25, whereas a terminal alkene has a pKa of 36. These values can be compared with alcohols, which have pKa values between 16 and 18 for the most part, and ketones (e.g., acetone) have a pKa of 19 (see Section 13.2.1). A variety of bases (sodium hydride, lithium, sodium or potassium amide, lithium dialkylamides, organolithium reagents, or Grignard reagents) can be used to generate an alkyne anion.43 Oct-1-yne reacted with methylmagnesium bromide, for example, to yield methane and the Mg salt of the acetylide.43 This acid-base reaction must be considered if a Grignard reaction (Sections 11.4.3–11.4.5) is one step of a planned synthesis that involves a molecule containing a terminal alkyne.44 The acidity of an alkynyl hydrogen is strongly influenced by the nature of the group on the other side of the triple bond. 45 Electron-releasing alkyl groups tend to lessen acidity, so acetylene (ethy-1-yne) is >12 times more acidic than hex-1-yne. Ethylmagnesium bromide is less basic so alkyne deprotonation is slower (30–90 min at reflux).45 The electron-withdrawing methoxy group45 makes methoxyacetyene (MeOC^CH) >2.5 times more acidic than acetylene.45 Conjugation enhances the acidity46 and n-PrC^CdC^CH reacts instantaneously with ethylmagnesium 38

Fernanders, R. A.; Mulay, S. V. J. Org. Chem. 2010, 75, 7029.

39

Corey, E. J.; Fuchs, P. L. Tetrahedron Lett. 1972, 13, 3769.

40

Shimada, N.; Abe, Y.; Yokoshima, S.; Fukuyama, T. Angew. Chem. Int. Ed. 2012, 51, 11824.

41

(a) Hopkinson, A. C. In The Chemistry of the Carbon-Carbon Triple Bond; Patai, S., Ed.; John Wiley & Sons, Chichester, NY, 1978, pp 75–136; (b) Matthews, W. S.; Bares, J. E.; Bartmess, J. E.; Bordwell, F. G.; Cornforth, F. J.; Drucker, G. E.; Margolin, Z.; McCallum, R. J.; McCollum, G. J.; Vanier, N. R. J. Am. Chem. Soc. 1975, 97, 7006; (c) McEwen, W. K. Ibid. 1936, 58, 1124; (d) Streitwieser, Jr., A.; Reuben, D. M. E. Ibid. 1971, 93, 1794.

(a) Cram. D. Fundamentals of Carbanion Chemistry; Academic Press: New York, NY, 1965; (b) Maksic, Z. B.; Eckert-Maksic, M. Tetrahedron 1969, 25, 5113; (c) Maksic, Z. B.; Randic, M. J. Am. Chem. Soc. 1973, 95, 6522; (d) Emsley, J. W.; Feenez, J.; Sutcliffe, C. H. High Resolution NMR; Pergamon Press: New York, NY, 1966.

42

43

Kríž, J.; Beneš, M. J.; Peška, J. Tetrahedron Lett. 1965, 2881.

44

(a) Prevost, C.; Gaudemar, M.; Honigberg, J. Compt. Rend. 1950, 230, 1186; (b) Wotiz, J. H.; Matthews, J. S.; Lieb, J. A. J. Am. Chem. Soc. 1951, 73, 5503.

45

(a) Reference 32, p 171; (b) Wotiz, J. H.; Hollinsworth, C. A.; Dessy, R. E. J. Org. Chem. 1955, 20, 1545.

46

Reference 32, p 15.

554

11. CARBON-CARBON BOND-FORMING REACTIONS

bromide in refluxing ether, in contrast to n-BuC^CH, which shows no reaction after 15 min under the same reaction conditions, although it reacts slowly when longer reaction times are used.45 At higher reaction temperatures, dilithiation (HC^CdH ! HC^CdLi ! LidC^CdLi) can be a problem with acetylene. Attempts to form HC^CdM with either n-butyllithium (M ¼ Li) or Grignard reagents (M ¼ MgBr) in ether led to MdC^CdM due to the instability of MdC^CdH. In THF, however, both HC^CMgBr and HC^CLi are stable enough 80% yield, as a mixture of diastereomers, in Doi and coworker’s102 synthesis of apratoxin C. In McMorris et al.103 synthesis of ()-irofulven, 42 reacted with methylmagnesium chloride, to give an 87% yield of 43. Note the selectivity for addition to the nonconjugated ketone, and delivery of methyl from the top face of 42, as the molecule is drawn. O

OPMB CH =CHMgBr 2

HO

OPMB

THF, 0°C

41 (>80%)

40

Me

O

Me

O Me

MeMgCl, THF

Cl

–78

0°C

Me Me

Cl

HO

H

O

H OAc

OAc 42

43 (87%)

The disconnection for this process follows: R1 R

O

OH R1

R1

+

R MgX

R X

R1

11.4.3.2 Acyl Substitution Although the reaction with aldehydes and ketones is the most common, Grignard reagents react with most other carbonyl derivatives.104 As observed in Section 4.2.3, acid derivatives contain a leaving group (e.g., Cl, O2R, OR, or NR2), and react with nucleophiles via acyl substitution. Initial addition of a Grignard reagent to the acyl carbon of an acid derivative generates an alkoxide intermediate (44), called a tetrahedral intermediate, and displacement of the leaving group (X) leads to a ketone. If the initially formed ketone product is more reactive than the acid derivative starting material, further reaction with the Grignard reagent can give tertiary alcohol (R3CdOH) via alkoxide 45.105 The competition between the ketone product and the acyl starting material can lead to mixtures of products, and there may be significant amounts of unreacted starting material. 102

Masuda, Y.; Suzuki, J.; Onda, Y.; Fujino, Y.; Yoshida, M.; Doi, T. J. Org. Chem. 2014, 79, 8000.

103

McMorris, T. C.; Staake, M. D.; Kelner, M. J. J. Org. Chem. 2004, 69, 619.

104

For reactions of this type, see Larock, R. C. Comprehensive Organic Transformations, 2nd ed.; Wiley-VCH: New York, NY, 1999, pp 1389–1411.

105

(a) Reference 78, pp 567–568; (b) Reference 79.

563

11.4 GRIGNARD REAGENTS (CdMG)

O

R

R

O

O MgX

R1 MgX

X

R

X R1

O MgX R1

OH H3O+

R1

R1

44

R1

R

R1

R 45

Acid chlorides are the most reactive acid derivatives, where chlorine is the leaving group.106 Reaction of Grignard reagents with acid chlorides gives excellent yields of the tertiary alcohol when an excess of the Grignard reagent is used, especially under vigorous reaction conditions. When the reaction is done at low temperatures and/or when metal catalysts are added, isolation of the ketone product is often possible.107 In some cases, the ketone product can be isolated. Reaction of hexanoyl chloride and 46, taken from Dussalt and coworker’s108 synthesis of peroxyacarnoate A, gave a 78% yield of ketone 47. O

O

MgBr O

+

O O

Cl

O Hexanoyl chloride

46

47 (78%)

An excellent method for converting an acid chloride to a ketone employs transition metal catalysts (e.g., ferric chloride), in conjunction with low-reaction temperatures.109 Reaction of butylmagnesium bromide with capryl chloride (decanoyl chloride) at 60°C gave a mixture of 13% of nonadecan-10-one, 4% of 10-nonylnonadecan-10-ol (from the secondary reaction of butylmagnesium bromide and nonadecan-10-one, and 60% of capric acid (decanoic acid).109 When a catalytic amount (2 mol%) of ferric chloride (FeCl3) was added, however, a 76% yield of nonadecan-10-one was obtained along with 3% of the over alkylation product, 10-nonylnonadecan-10-ol.109 If the intermediate ketone is unreactive to nucleophilic substitution due to steric hindrance or peculiar electronic factors (diisopropyl ketone and phenyl-tert-butyl ketone are both sterically hindered), the ketone can usually be isolated (Section 11.4.7 for problems with hindered ketones). In many instances, however, even dry ice temperatures and reverse addition techniques give poor yields of the ketone.110 Changing the metal from Mg to one that gives a less reactive organometallic provides an alternative method for isolating the ketone. Gilman et al.111 discovered that a Grignard reagent reacts with cadmium chloride (CdCl2) to give a dialkylcadmium reagent (R2Cd), and this less reactive organometallic is more selective in its reactions with carbonyl derivatives.112 This reaction can be rather slow, however, and in many cases it is not clear that dialkylcadmium is the active reagent.113 Dialkylcadmium reagents are readily formed from primary Grignard reagents, but secondary and tertiary organocadmium reagents are relatively unstable.114 The likely decomposition pathway is dissociation to radicals, which disproportionate to yield alkane and alkene products. O C9H19

Cl 2. H3O+

Decanoyl chloride

OH

O

1. C4H9MgBr, cat

C9H19

C9H19

Nonadecan-10-one

cat = NONE , –60°C cat = 2% FeCl3 , –60°C

(13%) (76%)

+

C9H19

C9H19 C9H19

10-Nonylnonadecan-10-ol (4%) (3%)

O

+ C9H19

OH

Decanoic acid (60%) (15%)

106

For a review of reactions of acyl chlorides with organometallic reagents, see Dieter, R. K. Tetrahedron 1999, 55, 4177.

107

Stowell, J. C. J. Org. Chem. 1976, 41, 560.

108

Xu, C.; Raible, J. M.; Dussault, P. H. Org. Lett. 2005, 7, 2509.

109

Cason, J.; Kraus, K. W. J. Org. Chem. 1961, 26, 1768, 1772.

110

Newman, M. S.; Smith, A. S. J. Org. Chem. 1948, 13, 592.

(a) Gilman, H.; Nelson, J. F. Rec. Trav. Chim. 1936, 55, 518; (b) Cason, J. J. Org. Chem. 1948, 13, 227; (c) Le Guilly, L.; Tatibouët, F. Compt. Rend. Ser. C 1966, 262, 217; (d) LeGuilly, L.; Chenault, J.; Tatibouët, F. Ibid. 1965, 260, 6634.

111

112

(a) Cason, J. Chem. Rev. 1947, 40, 15; (b) Shirley, D. A. Org. React. 1954, 8, 28 (see pp 35–38).

113

Cason, J. J. Am. Chem. Soc. 1946, 68, 2078.

114

Cason, J.; Fessenden, R. J. J. Org. Chem. 1960, 25, 477.

564

11. CARBON-CARBON BOND-FORMING REACTIONS

Reagents (e.g., dipentylcadmium) are formed by reaction of a Grignard reagent (pentylmagnesium bromide) and anhydrous cadmium chloride (CdCl2). Such reagents react rapidly with acyl halides, but very slowly or not at all with ketones, aldehydes, esters, or amides. The yields of ketone products obtained by reaction with acid chlorides are moderate to good.115 In general, dialkylcadmium reagents do not react with the ketone products, but there are exceptions when highly reactive ketone products are formed.112 When dipentylcadmium reacted with α-chloroacetyl chloride, a 46% yield of 1-chloroheptan-2-one was obtained.115 O

CdCl2

MgBr

2

O

Cl

Cd

Cl

Cl

2

Dipentylcadmium

Pentylmagnesium bromide

1-Chloroheptan-2-one (46%)

Although esters are much less reactive than acid chlorides (alkoxy is a poorer leaving group than Cl), they are only slightly less reactive than the ketone products resulting from reaction with a Grignard reagent. A Grignard reaction with esters rarely gives the ketone in good yield. A mixture of ester, ketone, and tertiary alcohol is often observed when only 1 equiv. each of Grignard reagent and ester are used. Two or more molar equivalents of Grignard reagent lead to the alcohol as the only product.116 The tendency for overreaction of some acid derivatives can have useful synthetic applications. The reaction of 48 with an excess of methylmagnesium bromide117 illustrates the reaction at hand, but introduces an interesting feature. If the focus is on the benzyl ester, addition of 2 equiv. of Grignard reagent leads to the tertiary alcohol in 49. The acetate group in 48 is also an ester, and reaction with MeMgBr leads to the alcohol unit in 49 and tert-butanol via loss of acetone from initial addition of MeMgBr and subsequent reaction of acetone to yield the alcohol. Another application of this reaction is the previously discussed the Barbier-Wieland degradation (Section 6.7.1.4),118 used to decrease the chain length of carboxylic acid esters by one carbon. OAc

OH 6 MeMgBr, Ether, 0°C

OBn OH

O 48

49 (80%)

The usual disconnections for reaction with esters and acid chlorides follow: OH

O R

R1

R COX

+

2

R1 X

R1

R

R CO2H

+

2

R1 X

R1

Not all esters react in this standard way. Mesitoate esters (e.g., n-butyl mesitoate, butyl 2,4,6-trimethylbenzoate) resist the acyl addition reaction because the ortho methyl groups sterically hinder the acyl carbon. When butyl 2,4,6-trimethylbenzoate reacted with phenylmagnesium iodide, for example, the products were mesitoic acid (2,4,6-trimethylbenzoic acid) and iodobutane.119 Acyl addition is inhibited by the presence of the methyl groups, but iodide (assisted by Mg2+ complexation with the carbonyl) attacks the OdC group (OdBu in butyl 2,4,6-trimethylbenzoate) via an SN2 reaction that displaces the carboxyl moiety, which functions as a leaving group.

(a) Miyano, M.; Dorn, C. R.; Mueller, R. A. J. Org. Chem. 1972, 37, 1810; (b) Miyano, M.; Dorn, C. R. Tetrahedron Lett. 1969, 1615; (c) Archer, S.; Unser, M. J.; Froelich, E. J. Am. Chem. Soc. 1956, 78, 6182; (d) Bindra, J. S.; Bindra, R. Prostaglandin Synthesis; Academic Press: New York, NY, 1973, pp 52–53.

115

116

Fieser, L. F.; Heymann, H. J. Am. Chem. Soc. 1942, 64, 376.

117

Evans, D. A.; Barnes, D. M.; Johnson, J. S.; Lectka, T.; von Matt, P.; Miller, S. J.; Murry, J. A.; Norcross, R. D.; Shaughnessy, E. A.; Campos, K. R. J. Am. Chem. Soc. 1999, 121, 7582. 118

(a) Wieland, H. Berichte 1912, 45, 484; (b) Barbier, P.; Locquin, R. Compt. Rend. 1913, 156, 1443.

119

(a) Reference 79, p 567; (b) Fuson, R. C.; Bottorff, E. M.; Speck, S. B. J. Am. Chem. Soc. 1942, 64, 1450.

565

11.4 GRIGNARD REAGENTS (CdMG)

O

O 1. PhMgI

OBu

+

OH

2. H2O

Butyl 2,4,6-trimethylbenzoate

C4H9—I

2,4,6-Trimethylbenzoic acid

Grignard reactions with lactones are interesting in that ring opening accompanies acyl addition, giving a diol in which the two ends are differentiated. Valerolactone, for example, reacted with butylmagnesium iodide to yield 5-methylhexane-1,5-diol. Amides are less reactive than esters (NR2 is a poorer leaving group than OR) and the ketone product can often be isolated, although the yields are usually poor. Reaction of N-methylacetamide with 2 equiv. of phenylmagnesium bromide gave alkoxide 50.120 Hydrolysis led to loss of the amine (via 1-(methylamino)-1-phenylethan-1-ol) and formation of the ketone product. Since 50 cannot react with additional Grignard, the product is the ketone, acetophenone. N,N-Dialkyl amides also react with Grignard reagents to give a ketone, but as mentioned above, the yields are usually poor.121 An exception is a Weinreb amide, seen previously in Section 4.2.3. This activated amide dCONMe(OMe) is an excellent leaving group, and the reaction stops at the ketone. In a synthesis of amphilectolide by Trauner and coworkers, Weinreb amide 51 was converted to ketone 52 in >80% yield, by reaction with 2-methylprop-1-enylmagnesium bromide.122

H

H H3C

N

2 PhMgBr

Me

H3C

H

N Ph

O N-Methylacetamide

H3O+

Me

H3C

OMgBr

Ph

N Ph

O

OH

Acetophenone

1-(Methylamino)-1phenylethan-1-ol

50

Me

Me MeO

MgBr

H

N

Me

Me

H

THF, 0°C – rt

Me O

Me

O

O

51

Me

O

52 (>80%)

The reaction with lactams follows a somewhat different pathway than that just described for amides. The final product is an amine or an amino ketone. Pyrrolidin-2-one and piperidin-2-one derivatives give primarily the amine, with minor amounts of ring-cleavage products (amino ketones). In Schneider and coworker’s123 synthesis of (+)monomorine, (53) reacted with the butylmagnesium bromide to yield iminium salt 54, in 85% yield, and subsequent reduction gave 55 in 98% yield. Large-ring lactams give mainly the amino ketone upon reaction with Grignard reagents.124

C4H9

O N

PMP

2 BuMgBr, THF 5 h , –78°C

N

C4H9 0.3 B(ArF5)3 3.0 Ph3SiH

PMP

N

PMP

DCM 90 h –78°C to rt

CO2Et

CO2Et 53

54 (85%)

120

Heyns, K.; Pyrus, W. Chem. Ber. 1955, 88, 678.

121

Busch, M.; Fleischmann, M. Berichte 1910, 43, 2553.

122

Chen, I. T.; Baitinger, I.; Schreyer, L.; Trauner, D. Org. Lett. 2014, 16, 166.

123

Abels, F.; Lindemann, C.; Koch, E.; Schneider. C. Org. Lett. 2012, 14, 5972.

124

Lukeš, R.; Dudek, V.; Sedláková, O.; Korán, J. Coll. Czech. Chem. Commun. 1961, 26, 1105.

CO2Et 55 (98%)

566

11. CARBON-CARBON BOND-FORMING REACTIONS

11.4.3.3 Reaction With Nitriles Nitriles are carboxylic acid derivatives in Section 11.2.1 and react with Grignard reagents (also see Section 11.6.6).125 Initial attack generates an intermediate iminium salt that can be hydrolyzed (with loss of ammonia) to the corresponding ketone.126 The best yields are obtained with aryl nitriles, as illustrated by the reaction of 3-methoxybenzonitrile with the Grignard reagent derived from 4-bromobut-1-ene. The initial reaction generates iminium salt 56,127 and subsequent hydrolysis produced (E)-1-methoxyocta-2,7-dien-4-one. MgBr

CN

O

N

MgBr

H3O+

NEt3

MeO

OMe

OMe 56

3-Methoxybenzonitrile

(E)-1-Methoxyocta-2,7-dien-4-one

An example taken from the synthesis of the human chymase inhibitor SPF32629A by Boovanahalli and coworkers128 reacted 4-(benzyloxy)picolinonitrile with phenylmagnesium bromide to yield 77% of (4-(benzyloxy)pyridin-2-yl)(phenyl)methanone. The intermediate iminium salt is not usually isolated since hydrolysis to the ketone is rapid. When an alkyl nitrile reacts with a Grignard derived from a hindered alkyl halide, the yield of ketone can be poor, but this is not always the case. Hindered iminium salts can resist hydrolysis due to steric blocking of the imino carbon,129 however, and treatment with aqueous acid130 under vigorous conditions can be required to give even modest yields. Hindered nitriles often give no reaction at all when they are reacted with very hindered Grignard reagents. Reaction of tertbutylmagnesium chloride with cyanomethyl-2,4,6-trimethylbenzene, for example, gave 0% of the ketone.131 OBn

OBn PhMgBr, THF

Ph

–30°C to rt

N

N

CN

O (4-(Benzyloxy)pyridin-2-yl) (77%) (phenyl)methanone

4-(Benzyloxy)picolinonitrile

The disconnection for the nitrile-Grignard reagent reaction is: O RJCN + R

R1JX

R1

11.4.3.4 Reaction With Carbon Dioxide Carbon dioxide (CO2, O]C]O) is actually a carbonyl derivative, and reacts with a Grignard reagent132 to give a carboxylate salt, that gives the corresponding carboxylic acid upon hydrolysis.133 Conversion of benzylmagnesium bromide to phenylacetic acid134 is an example of this transformation. This functional group interchange of a halide 125

For reactions of this type, see Larock, R. C. Comprehensive Organic Transformations, 2nd ed.; Wiley-VCH: New York, NY, 1999, pp 1420–1422.

126

Reference 79, pp 767–845.

127

Taber, D. F.; Wang, Y.; Pahutski, Jr., T. F. J. Org. Chem. 2000, 65, 3861.

128

Vegi, S. R.; Boovanahalli, S. K.; Sharma, A. P.; Mukkanti, K. Tetrahedron Lett. 2008, 49, 6297.

(a) Hrubiec, R. T.; Smith, M. B. J. Org. Chem. 1984, 49, 431; (b) Hanack, M.; Ensslin, H. M. Annalen 1966, 697, 100; (c) Hanack, M.; Bocher, S.; Herterich, I.; Hummel, K.; V€ ott, V. J. L. Ann. Chem. 1970, 733, 5.

129

130

Citron J. D.; Becker, E. I. Can. J. Chem. 1963, 41, 1260.

131

Bruylants, A. Bull. Soc. Chem. Fr. 1958, 1291.

(a) Oppolzer, W.; K€ undig, E. P.; Bishop, P. M.; Perret, C. Tetrahedron Lett. 1982, 23, 3901; (b) Rowsell, D. G. Brit. Patent 1,392,907; (Chem. Abstr. 83: P114682t 1975).

132

133

(a) Sneeden, R. P. A. In The Chemistry of Carboxylic Acids and Esters; Patai, S., Ed.; Interscience: New York, NY, 1969, pp 137–175; (b) Reference 79, pp 913–948.

134

Eberson, L. Acta Chem. Scand. 1962, 16, 781.

567

11.4 GRIGNARD REAGENTS (CdMG)

to an acid proceeds with a one-carbon extension of the chain. An interesting example is taken from a synthesis of (+)-biotin by Seki et al.,135 in which nitrile 57 was generated in situ, and reacted with the bis(Grignard) reagent shown. Initial reaction with the nitrile led to formation of the ketone unit in 58 after hydrolysis (see Section 11.4.3.3), but the second CdMgBr unit reacted with CO2 and hydrolysis with aqueous citric acid generated the carboxylic unit. The yield of 58 was 79% in the solvent shown, but the yield was significantly lower in THF and other solvents. Boc Ph

OTMS

N S

Boc

MgBr

1. BrMg

Ph

N

Bu2O , Toluene 2. CO2 3. aq Citric acid

CN

OTMS

CO2H

S O

57

58 (79%)

The disconnection for the carboxylation reaction follows: R

CO2H

R

+

X

CO2

11.4.4 Conjugate Addition α,β-Unsaturated carbonyl derivatives have two reactive sites, and nucleophilic addition of a Grignard reagent can occur at either the acyl carbon (1,2-addition) or the alkenyl carbon (1,4-addition). When a Grignard reagent reacts with a conjugated carbonyl (e.g., 59), the 1,4-addition product is an enolate anion (60, see Section 13.2). The 1,2-addition product is the usual alkoxide 61. The course of this reaction varies with the steric bulk of R1 in the Grignard reagent, and R in the carbonyl compound. As the size of R increases, the amount of 1,4-addition increases, although α,βunsaturated aldehydes usually give only 1,2-addition.136 Conversely, as the R1 group of the Grignard increases in size, less conjugate addition is observed for 59 with a given R. In general, Grignard reagents undergo 1,2-addition with conjugated aldehydes and relatively unhindered conjugated ketones, although conjugate alkylation often competes in the latter case. In a synthesis of (+)-methynolide, Cossy et al.137 reacted methacrolein with ethylmagnesium bromide, and obtained a 68% yield of the 1,2-addition product, 2-methylpent-1-en-3-ol. O R1 MgX

R1

OMgX

OMgX

+

+ R

R

59

R

60

R1

61

1. EtMgBr, Ether

CHO

2. H3O+

Methacrylaldehyde

OH 2-Methylpent-1-en-3-ol (68%)

The influence of R2 (from the Grignard reagent) and R1 (attached to the carbonyl) on the percentage of conjugate addition have been reported for the reaction of alkyl Grignard reagents with 62.138 Note that the phenyl group in 62 (R3 ¼ Ph, R4 ¼ H) has a significant influence on the reaction, which proceeds by a six-center transition state (see 63) to give the conjugate addition product, 64. As R1 in 62 becomes larger, attack at the carbonyl (path a) in a four-center transition state induces a large and destabilizing R2$ R1 interaction leading to preferred attack at the end of the planar conjugated system (path b). When R1 is small, R2 is easily transferred to the carbonyl by the usual four-center transition state (path a).138 Increasing the size of R3 or R4 increases the magnitude of the R3(R4)$R2 interaction in the six-center

135

Seki, M.; Mori, Y.; Hatsuda, M. Yamada, S. J. Org. Chem. 2002, 67, 5527.

136

Hauser, F. M.; Hewawasam, P.; Rho, Y. S. J. Org. Chem. 1989, 54, 5110.

137

Cossy, J.; Bauer, D.; Bellosta, V. Tetrahedron 2002, 58, 5909.

138

Maroni-Barnaud, Y.; Maroni, P.; Fualdès, A. M. Compt. Rend. 1962, 254, 2360.

568

11. CARBON-CARBON BOND-FORMING REACTIONS

transition state (path b), diminishing the amount of 1,4-addition. Removing the Lewis acid byproduct (MgX2) facilitates formation of 63 with an increase in 1,4-addition product 64.138 With alkylidene malonates and related compounds, virtually no 1,2-addition is observed,139 although two electron-withdrawing groups attached to the alkene moiety make conjugate addition rather facile. R2 R3 R2 MgCl

O

+

2 R3 R

O R1

R4

R4

62

H3O+

R1

a

R3

R1

R4

X

Mg

b

O 63

64

A similar transition state can be drawn for the reaction of Grignard reagents and α,β-unsaturated esters, where steric hindrance in the ester moiety partially blocks the acyl carbon and leads to more 1,4-addition.140 An example is the reaction of butylmagnesium bromide with sec-butylcrotonate to yield sec-butyloctanoate.141 Asymmetric induction is possible in these reactions, as illustrated by the reaction of the ()-menthyl ester142 of crotonic acid 65 to give the (S)-(+) enantiomer of 3-phenyl derivative 66 in 46% yield.143 Interestingly, addition of Cu2Cl2 to the reaction gave the (R)-() enantiomer. O

H

Ph O

1. PhMgBr

O

O

2. H3O+

Me

Me 65

66 (46%)

In general, addition of cuprous [Cu(I)] salts to a Grignard facilitates conjugate addition, partly because a more highly reactive species (R2MgCu or RCu) is formed, and partly because Cu coordinates better to the carbonyl, which facilitates the six-center transition state. The organocuprates derived from organolithium reagents (R2CuLi, Section 12.3.1.6) are extremely useful in conjugate addition reactions.144 Organocopper reagents derived from the reaction of a Grignard reagent with a cuprous salt, however, are also very useful. In a synthesis of rippertenol by Snyder et al.,145 a functionalized Grignard reagent reacted with conjugated ketone 67 in the presence of CuBrSMe2 to give an 86% yield of the conjugate addition product 68. This example is typical in that cuprous salts generally lead to 1,4addition, despite the usual propensity of conjugated aldehyde to yield 1,2-addition with nucleophiles. A variety of conjugated carbonyl derivatives undergo 1,4-addition with Grignard reagents. Conjugate addition to α,β-unsaturated amides is known and proceeds with little variation from the results observed for conjugated ketones or esters.146 As mentioned above, the organocuprates derived from lithium will be discussed in greater detail in Section 12.3.1. O

O

O

2.5 O

Me

O

MgBr

O

CuBr•SMe2 , THF, 5.5 h –78 to –50°C

Me

TBDPSO

TBDPSO 67

139

Mane, R. B.; Krishna Rao, G. S. J. Chem. Soc. Perkin Trans. 1 1973, 1806.

140

Munch-Petersen, J. J. Org. Chem. 1957, 22, 170.

141

Munch-Petersen, J.; Jacobsen, S. Compt. Rend. 1962, 255, 1355.

142

68 (86%)

Seiji, S.; Yumiko, S.; Sawada, S.; Sejima, Y.; Ohi, S.; Shunsuke, O.; Inouye, Y. Bull Kyoto Univ. Educ. Ser. B, 1979, 55, 33; (Chem. Abstr. 92: 129098s 1980). 143

Inouye, Y.; Walborsky, H. M. J. Org. Chem. 1962, 27, 2706.

144

(a) Bernady, K. F.; Weiss, M. J. Prostaglandins 1973, 3, 505; (b) Reference 118d, p 111.

145

Snyder, S. A.; Wespe, D. A.; von Hof, J. M. J. Am. Chem. Soc. 2011, 133, 8850.

146

Gilbert, G.; Aycock, B. F. J. Org. Chem. 1957, 22, 1013.

569

11.4 GRIGNARD REAGENTS (CdMG)

11.4.5 Reaction With Epoxides Grignard reagents react with epoxides147 to form a new carbon-carbon bond via opening of the three-membered ring, and hydrolysis leads to an alcohol. There are two electrophilic carbon atoms, but a Grignard reagent usually attacks the less hindered carbon in an SN2 like reaction.148 If the carbon atoms of the epoxide are primary or secondary, attack can occur at either carbon leading to a mixture of alcohol products. In other words, when a Grignard reagent opens an epoxide, the major product usually has a new substituent on the β-carbon relative to the OH. When 2-ethyloxirane reacts with methylmagnesium bromide, attack at the primary carbon gave pentan-3-ol with only trace amounts of attack at the secondary carbon. When the epoxide has substituents that provide equal or close to equal steric hindrance at each carbon, as in 2-ethyl-3-propyloxirane, the reaction produces a mixture of regioisomeric alcohol products. In this case, reaction of this epoxide and methylmagnesium bromide gave a mixture of 4-methylheptan-3-ol and 3-methylheptan-4-ol. OH

O

Me

O

1.

MeMgBr

2. H3O+

+ Me

OH

1. 2. H3O+

OH

4-Methylheptan-3-ol

3-Methylheptan-4-ol

Pentan-3-ol

Metal salts (e.g., cuprous ion) with a Grignard reagent promotes the ring-opening reaction of epoxides, which is particularly useful for directing the incoming Grignard reagent to a particular carbon (to make the reaction more regioselective). An example is taken from Reddy and coworker’s149 synthesis of isofaregenedadiol in which the reaction of epoxide 69 with allylmagnesium bromide and cuprous iodide, gave 70 in 79% yield. Note that the Grignard reagent added to the less substituted carbon. Allylic epoxides can be opened in a conjugated addition manner using Grignard reagents, analogous to an SN20 reaction (Section 3.2.1.3). Isopropylmagnesium bromide reacted with 2-(1-chlorovinyl) oxirane, for example, attacking the alkene unit with concomitant opening of the epoxide to give a 76% yield of (Z)-3chloro-5-methylhex-2-en-1-ol with a 22.1:1.0 (Z-E) ratio.150 MgBr

O OPMB

CuI, THF, 0°C, 1 h

69

HO

OPMB 70 (79%)

O

OH

Me2CHMgBr, CuBr/SMe2

Cl 2-(1-Chlorovinyl)oxirane

THF, –20°C 1 h Addition

Cl (Z)-3-Chloro-5-methylhex-2-en-1-ol (76%)

This fundamental reaction leads to useful synthetic applications. Similar alkylation of aryl epoxides, followed by oxidation of the resulting alcohol, makes the Grignard-epoxide reaction an effective route for the synthesis of α-aryl ketones. Epoxidation of cyclohexene with m-chloroperoxybenzoic acid (Section 6.4.3) gave cyclohexene oxide. Subsequent reaction with phenylmagnesium bromide and hydrolysis led to 2-phenylcyclohexanol. Oxidation of the secondary alcohol moiety with chromium trioxide (or with another oxidizing agent as in Section 6.2) gave the targeted 2-phenylcyclohexanone.

(a) Taylor, S. K.; Haberkamp, W. C.; Brooks, D. W.; Whittern, D. N. J. Heterocyclic Chem. 1983, 20, 1745; (b) Linstrumelle, G.; Lorne, R.; Dang, H. P. Tetrahedron Lett. 1978, 4069; (c) Boireau, G.; Namy, J. L.; Abenhaïm, D. Bull. Soc. Chim. Fr. 1972, 1042.

147

148

Sano, M.; Kodama, H.; Matsuda, H.; Matsuda, S. Nippon Kagaku Kaishi 1974, 1716; (Chem. Abstr. 82: 42560f 1975).

149

Kurhade, S. E.; Sanchawala, A. I.; Ravikumar, V.; Bhuniya, D.; D. Reddy, S. Org. Lett. 2011, 13, 3690.

150

Taber, D. F.; Mitten, J. V. J. Org. Chem. 2002, 67, 3847.

570

11. CARBON-CARBON BOND-FORMING REACTIONS

This sequence generates the useful disconnections that follow: R

R

OH

R +

O R

R1

R

R-X

Ar

R R

R

R R

Ar = aryl

Looking back to the Schlenk equilibrium (Section 11.4.1), Grignard reagents contain the mild Lewis acid MgX2 even if no other reagent is added. This Lewis acid is part of the reagent, and it can induce acid-catalyzed rearrangements of acid-sensitive epoxides, complicating the normal product distribution. Me Me

C8H17

Me 1. MeMgI

H

Me

C8H17

H

2. H3O+

H

H HO

HO

O

Me

HO

5,6 -Epoxycholestane

H

H

H

4a -Methyl-A-homo- norcholestan-4a -ol

MgI2 MeMgI

Me Me

Me

H

O

Me H

H HO

C8 H17

H

H

C8 H17

H H

H

HO O

MgI 71

72

When 5,6β-epoxycholestane was treated with methylmagnesium iodide, the product was 4aα-methyl-A-homo-βnorcholestan-4aβ-ol,151 which is explained by reaction of the epoxide with MgI2 to yield 71. A 1,2-hydride shift (Sections 2.5.1 and 16.2.3) generated ketone derivative 72, which reacted with additional methylmagnesium iodide to give the alcohol, 4aα-methyl-A-homo-β-norcholestan-4aβ-ol. The involvement of MgI2 in the conversion of 5,6βepoxycholestane to 4aα-methyl-A-homo-β-norcholestan-4aβ-ol was later substantiated by similar reaction of a cholestan-3-one derivative with methylmagnesium iodide.152 The yield was low, since cationic rearrangement was slower than nucleophilic attack in the aprotic solvent used for this Grignard reaction. If the epoxide moiety is attached to a tertiary carbon, cationic rearrangements can lead to significant amounts of rearranged product. The cationic intermediate generated from the epoxide and MgX2 does not always rearrange to a ketone derivative, but can give SN1 type reactions if a tertiary center is present. The reaction of phenylmagnesium bromide with 2-methoxy-3,3-dimethyl-2-phenyloxirane, for example, led to a coordination complex of the magnesium and oxygen in 73. The oxonium salt opened to generate an oxygen-stabilized cation (74), which is particularly stable. When this intermediate reacted with the nucleophilic PhMgBr reagent, the product was 1-methoxy2-methyl-1,1-diphenylpropan-2-ol after hydrolysis.153

151

Frankel, J. J.; Julia, S.; Richard-Neuville, C. Bull. Soc. Chim. Fr. 1968, 4870.

(a) Rao, P. N.; Uroda, J. C. Tetrahedron Lett. 1964, 1117; (b) Hall, N. D.; Just, G. Steroids 1965, 6, 111; (c) Julia, S.; Lavaux, J. P.; Lorne, R.; Riz, J. -C. Bull. Soc. Chim. Fr. 1967, 3218.

152

153

Stevens, C. L.; Holland, W. J. Org. Chem. 1958, 23, 781.

571

11.4 GRIGNARD REAGENTS (CdMG)

O Ph MeO

MgBr Me Me

MgBr2

O

MeO

Ph

Me

MeO

Me

Ph 73

2-Methoxy-3,3-dimethyl2-phenyloxirane

Me OMgBr

PhMgBr, Ether, 0 °C

Ph OMe Me OH Ph Me

PhMgBr

Me 74

1-Methoxy-2-methyl-1,1diphenylpropan-2-ol

Reflux

11.4.6 Stereochemical Stability of Grignard Reagents In principle, the carbon atom bearing a MgX unit can be a stereogenic center, which has important stereochemical implications for any reaction. When the CdMg bond of a Grignard reagent is formed, however, the stereochemical integrity of that bond is poor. For this reason, the generation and use of chiral Grignard reagents is not a synthetically useful option in most systems. The NMR studies by Roberts and Whitesides154 and also Fraenkel et al.155 on Grignard reagents in ether, established that the CdMg bond has significant ionic character and undergoes rapid inversion at the magnesium-bearing carbon atom. Cyclohex-3-en-1-ylmagnesium bromide, for example, undergoes inversion of configuration at the CdMg position. MgBr MgBr Cyclohex-3-en-1-ylmagnesium bromide

H Me

Me H MgBr

sec-Butylmagnesium bromide

The rate of inversion decreased when the solvent was changed from ether to THF.156 Fraenkel and coworkers155,157 also showed that 2-butylmagnesium bromide (sec-butylmagnesium bromide) undergoes rapid inversion at ordinary temperatures, thereby making a Grignard reagent derived from chiral 2-bromobutane effectively racemic. However, Hoffmann and H€ olzer158 studied chiral Grignard reagent (S)-(1-phenylbutan-2-yl)magnesium chloride, which was configurationally stable at 78°C. In a nickel-catalyzed reaction of (S)-(1-phenylbutan-2-yl)magnesium chloride with vinyl bromide, the so-called Kumada-Corriu coupling,159 an 80% yield of (S)-(2-ethylbut-3-en-1-yl)benzene was obtained in 89 %ee, along with 11% of (E)-but-1-en-1-ylbenzene. The Grignard reagent (S)-(1-phenylbutan-2-yl)magnesium chloride was generated by the reaction of a chiral chlorosulfone with an excess of ethylmagnesium chloride at 78°C. Since the carbon-magnesium bond is stereochemically unstable, a Grignard reagent derived from a chiral halide equilibrates (75a ! 75b),160 so subsequent Grignard reaction with an aldehyde or ketone will give racemic products. There are a few exceptions to this configurational instability. Walborsky et al.161 reported that the Grignard reagent derived from (R)-1-bromo-1,2,2-trimethylcyclopropane reacted with CO2 to give the corresponding carboxylic acid with some retention of configuration (14 %ee) in 38% yield. The %ee can be as high as 80%, however, when the reactions are carried out in ether.161 Better selectivity was observed with menthol-based Grignard reagents.162 Cyclopropyl derivatives tend to give better selectivity than acyclic or larger ring Grignard reagents.

154

Whitesides, G. M.; Roberts, J. D. J. Am. Chem. Soc. 1965, 87, 4878.

155

(a) Fraenkel, G.; Dix, D. T. J. Am. Chem. Soc. 1966, 88, 979; (b) Fraenkel, G.; Dix, D. T.; Adams, D. G. Tetrahedron Lett. 1964, 3155.

156

Maercker, A.; Geuss, R. Angew. Chem. Int. Ed. 1971, 10, 270.

157

Fraenkel, G.; Dix, D. T.; Adams, D. G. Tetrahedron Lett. 1964, 3155.

(a) H€ olzer, B.; Hoffmann, R. W. Chem. Commun. 2003, 732. See also (b) Hoffmann, R. W.; H€ olzer, B.; Knopff, O.; Harms, K. Angew. Chem. Int. Ed. 2000, 39, 3072.

158

(a) Tamao, K.; Hiyama, T.; Negishi, E.-i. J. Organomet. Chem. 2002, 653, 1; (b) Corriu, R. J. P.; Masse, J. P. J. Chem. Soc. Chem. Commun. 1972, 144a; (c) Tamao, K.; Sumitani, K.; Kumada, M. J. Am. Chem. Soc. 1972, 94, 4374; (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, Jr., F. G. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005, pp 386–387.

159

160

Morrison, J. D.; Mosher, H. S. Asymmetric Organic Reactions; ACS: Washington, DC, 1976, pp 414–415.

161

Walborsky, H. M.; Impastato, F. J.; Young, A. E. J. Am. Chem. Soc. 1964, 86, 3283.

(a) Tanaka, M.; Ogata, I. Bull. Chem. Soc. Jpn. 1975, 48, 1094; (b) Schumann, H.; Wassermann, B. C.; Hahn, F. E. Organometallics 1992, 11, 2803; (c) Dakternieks, D.; Dunn, K.; Henry, D. J.; Schiesser, C. H.; Tiekink, E. R. Organometallics 1999, 18, 3342.

162

572

11. CARBON-CARBON BOND-FORMING REACTIONS

MgCl

CH2=CHBr , THF, –78°C, 5 d

Ph

Ph

+

Ph

cat NiCl2(-)DIOP

(S)-(2-Ethylbut-3-en-1yl)benzene (80%)

(S)-(1-Phenylbutan-2-yl)magnesium chloride R1

(E)-But-1-en-1ylbenzene (11%) Br

R1

R

MgX

XMg

R

H

Me Me Me (R)-1-Bromo-1,2,2-trimethylcyclopropane

H 75b

75a

Chiral Grignard reagents are known. α-Halo-Grignard reagents have been shown to be excellent precursor to chiral Grignard reagents. Hoffmann et al.163 showed that ethylmagnesium chloride reacted with α-halosulfoxide 76 to give (R)-(1-chloro-2-phenylethyl)magnesium chloride, along with 4-chloropenyl ethyl sulfoxide. When (R)-(1-chloro-2-phenylethyl)magnesium chloride was treated with more ethylmagnesium chloride between 50 and 30°C, Grignard reagent (S)-(1-phenylbutan-2-yl)magnesium chloride was formed in solution. Subsequent reaction with phenylisothiocyanate, thioamide (S)-2-benzyl-N-phenylbutanethioamide was isolated in 56% yield and 83 %ee. Hoffmann et al.164 observed that racemization of the Grignard reagent (S)-(1-phenylbutan-2-yl)magnesium chloride was slowest in the presence of chloride, an anion of low nucleophilicity. The use of ethylmagnesium chloride was important because the reaction was least complicated by formation of a rearranged Grignard product.165 Ph

Ph

Ph

S

••

O

EtMgCl , THF

S

EtMgCl, THF

ClMg

–78°C

+

Cl

Cl

Cl

Cl

PhHN

O S

••

Et

Et

Et

(S)-(1-Phenylbutan-2-yl)magnesium chloride

(R)-(1-Chloro-2-phenylethyl)magnesium chloride

76

Ph

PhNCS

ClMg

(S)-2-Benzyl-N-phenylbutanethioamide

When α-halo Grignard reagent is treated with another organometallic reagent (e,g,m another Grignard reagent), a new Grignard reagent is generated by a carbon-carbon bond-forming reaction.165 This new reagent appears to be a rearrangement product, but actually arises by formation of a carbenoid species (Section 17.9.5) that undergoes a CdH insertion reaction. The reaction of (2,2-diiodoethyl)benzene with isopropylmagnesium chloride, for example, generated the iodo-Grignard reagent (1-iodo-2-phenylethyl)magnesium chloride at 78°C. As the solution of (1-iodo-2phenylethyl)magnesium chloride warmed, it reacted with additional isopropylmagnesium chloride to yield (3methyl-1-phenylbutan-2-yl)magnesium chloride, and quenching with methanol gave the observed product 2-methyl-3-phenylbutane.165 For most of the reactions listed in this section, such carbenoid reactions should not be a major problem. I I

(2,2-Diiodoethyl)benzene

MgCl

MgCl

–78°C

Ph

MgCl

I

MgCl

–60°C

Ph

Ph (1-Iodo-2-phenylethyl)magnesium chloride

MeOH

(3-Methyl-1-phenylbutan-2-yl)magnesium chloride

Ph 2-Methyl-3-phenylbutane

A different stereochemical problem arises when vinylmagnesium halides are formed from alkenyl (vinyl) halides. Both (E)- and (Z)-isomers can be formed. Reaction of stereochemically pure (E)- or (Z)-alkenyl halides with Mg gave a mixture of (E)- and (Z)-vinyl Grignard reagents.166 Subsequent reaction with carbonyl derivatives gives an allylic alcohol product as an (E/Z) mixture. Martin et al.166 reported that reaction of pure (E)- or (Z)-1-bromo-1-alkenes with Mg

163

Hoffmann, R. W.; H€ olzer, B.; Knopff, O.; Harms, K. Angew. Chem. Int. Ed. 2000, 39, 3072.

164

Hoffmann, R. W.; Nell, P.; Leo, R.; Harms, K. Chem. Eur. J. 2000, 6, 3359.

165

Hoffmann, R. W.; Knopff, O.; Kusche, A. Angew. Chem. Int. Ed. 2000, 39, 1462.

166

Martin, G. J.; Mèchin, B.; Martin, M. L. C. R. Acad. Sci. Ser. C 1968, 267, 986.

573

11.4 GRIGNARD REAGENTS (CdMG)

metal and then with CO2 gave stereospecific conversion to the corresponding acid. Although (E)-halide gave mainly the (E)-product, this is misleading. In fact, severe loss of stereochemical integrity was observed and a mixture of both (E) and (Z) isomers was always obtained.167 It is important to note that there are significant differences between the stereochemical ratio of the Grignard reagent and that of the acid product. In the reaction of bromopropenes and bromohexenes, loss of stereochemical integrity in the sequence that converted bromo-alkenes to the corresponding Grignard reagent was more pronounced in longer chain alkenes.167 In addition, greater loss of stereochemical integrity is apparent upon reaction with the electrophile (CO2). A partial solution to this problem is reaction of the vinyl Grignard reagent with alkyl halides in the presence of cuprous iodide [CuII], which does little for the initial loss of stereochemistry upon formation of the Grignard, but in subsequent reactions, greater retention of stereochemistry is observed.168 An example is the reaction of (Z)-but-2-en-2-ylmagnesium bromide with iodopropane in the presence of cuprous iodide, giving a 97% yield of (Z)-3-methylhex-2-ene. The original 90:10 (Z/E) mixture in (Z)-but-2-en-2-ylmagnesium bromide was retained [88:12 (Z/E)] in the final product (Z)-3-methylhex-2-ene. 1. C3H7 —I , CuI, THF, –30°C

MgBr

2. H2O

(Z)-But-2-en-2-ylmagnesium bromide

(Z)-3-Methylhex-2-ene

([Z/E] = 90:10)

(97%, [ Z/E] = 88:12)

Another problem arises with allylic Grignard reagents. When the magnesium halides derived from 1-bromobut-(2E)-ene were quenched with water,169 the double bond had isomerized and a mixture of but-1-ene, but-(2Z)-ene and but-(2E)-ene was obtained. This allylic rearrangement is well known in reactions of Grignard reagents.170 Increasing the solvating power of the solvent, as in changing from diethyl ether to THF, increased the lability of the CdMg bond and gave more isomerization.

11.4.7 Stereoselectivity of Grignard Reactions The reaction of a Grignard reagent reacts with unsymmetrical ketones or aldehydes that contain a prochiral carbonyl carbon, leads to a new stereogenic center. Diastereomers result if another stereogenic center is present in either the Grignard reagent (but not at the carbon bearing the Mg) or the carbonyl substrate. If pentan-2-one reacts with the Grignard reagent derived from (1-phenylethyl)magnesium bromide, for example, the resulting alcohol will be a mixture of syn and anti diastereomers (3-methyl-2-phenylhexan-3-ol and 3-methyl-2-phenylhexan-3-ol, respectively). Each diastereomer will be racemic. In light of the previous discussion in Section 11.4.6, it may be possible for one diastereomer to predominate (the reaction would be diastereoselective), but it should not be enantioselective. O

HO

BrMg

1. Ether

+ Ph

Pentan-2-one

Me

+

2. Hydrolysis

(1-Phenylethyl)magnesium bromide

HO

Me

Ph 3-Methyl-2-phenylhexan-3-ol syn (racemic)

Ph 3-Methyl-2-phenylhexan-3-ol anti (racemic)

The products from such reactions can be organized into four categories that reflect simple combinations of chiral and achiral Grignard reagents reacting with chiral or achiral carbonyl derivatives. (1) Mixing an achiral Grignard reagent and a carbonyl derivative that does not possess a substituent attached to a stereogenic center generates one new stereogenic center. (2) Mixing a chiral Grignard reagent and a carbonyl derivative that does not possess a substituent attached to a stereogenic center generates diastereomers. (3) Mixing an achiral Grignard reagent and a carbonyl derivative that has a substituent attached to a stereogenic center leads to diastereomers, with possible diastereoselectivity. (4) Mixing a chiral Grignard reagent with a carbonyl derivative that has a substituent attached to a stereogenic center will lead to good diastereoselectivity. 167

Martin, G. J.; Martin, M. L. Bull. Soc. Chim. Fr. 1966, 1636.

168

Derguini-Boumechal, F.; Linstrumelle, G. Tetrahedron Lett. 1976, 3225.

169

Agami, C.; Andrac-Taussig, M.; Prevost, C. Bull. Soc. Chim. Fr. 1966, 1915.

170

Agami, C. Bull. Soc. Chim. Fr. 1967, 4031.

574

11. CARBON-CARBON BOND-FORMING REACTIONS

If there is steric bias in a molecule, and if the addition of a Grignard reagent generates diastereomers, the products may be formed with good diastereoselectivity. The rationale for diastereoselectivity discussed in Section 7.9 for reduction generally applies to reaction of Grignard reagents and carbonyl derivatives. In a synthesis of vinigrol, by Paquette et al.,171 ketone 77 reacted with vinylmagnesium chloride to give 54% of 78 and 24% of 79. The observed diastereoselectivity is typical of the exo-selectivity observed similar bicyclic systems (Sections 1.5.3 and 7.9.6).172 Me

Me MgBr

O

H

THF , –78

Me

H OH

0°C

H

+

H

TBSO

OH

H

H

TBSO

TBSO

77

78 (54%)

79 (24%)

Good diastereoselectivity can be observed in less sterically demanding systems for selected Grignard reagents. Hoffmann and coworkers173 reported good diastereoselectivity when α haloalkyl Grignard reagents reacted with benzaldehyde. The Grignard reagent (1-bromo-2-phenylethyl)magnesium bromide, for example, gave an 82% yield of syn-2-bromo-1,2-diphenylethan-1-ol + anti-2-bromo-1,2-diphenylethan-1-ol in a 92:8 ratio, showing the reaction clearly favored the syn diastereomer. Br

Br Ph

Br

PhCHO , THF

MgBr

(1-Bromo-2-phenylethyl)magnesium bromide

–78°C

Ph

Ph

+

OH syn-2-Bromo-1,2diphenylethan-1-ol

Ph

Ph OH

anti-2-Bromo-1,2diphenylethan-1-ol

(82%)

Assuming that Grignard reagent addition to a carbonyl is a diastereoselective reaction that produces racemic products, it is important to ask if it is possible for the reaction to be enantioselective. If the focus is on the carbon bearing the Mg atom, the answer is usually no. If a stereogenic center is incorporated elsewhere in the molecule, however, the relative merits of the four cases listed above must be discussed. Case 1 involves an achiral Grignard reagent reacting with a prochiral ketone to yield a racemic alcohol. Some asymmetric induction174 has been achieved by forming asymmetric complexes of the Grignard reagent with solvents [e.g., (2R, 3R)-(+)-dimethoxybutane].175 The two oxygen atoms of this chiral ether coordinate with the Grignard reagent to form a monomeric species (e.g., 80).160,175 In 80, coordination with the oxygen atoms renders the Mg asymmetric, and reaction with a prochiral ketone will generate a diastereomeric transition state for the acyl addition,160,175 so an excess of one enantiomer will result. Since R2 in 80 exists as an equilibrating mixture of stereoisomers (see 73a and 73b), complex 80 represents four stereochemically distinct solvated species, not necessarily present in equal amounts, and each can react with a prochiral substrate at different rates.160,175 The extent of induced stereoselectivity is generally low, often 2–3%,160,175 even in an asymmetric solvent. When phenylmagnesium bromide reacted with octan-2-one in the optically active solvent 2-methyltetrahydrofuran, for example, 2-phenyloctan-2-ol was formed in only 11–18 %ee.176 In some cases, it is possible to form a chiral, nonracemic alcohol from a prochiral ketone and a Grignard reagent by adding an asymmetric ligand (a chiral additive) to the reaction mixture. This case is differentiated from the reaction of Grignard reagent with a carbonyl species that has a stereogenic center present in the same molecule (case 3 above, a Cram’s rule type system: see Section 7.9.1 for a discussion of Cram’s rule). Presumably, the Grignard reagent forms a chiral complex with these additives, promoting facial selectivity in the addition reaction with the carbonyl. In work by 171

Paquette, L. A.; Guevel, R.; Sakamoto, S.; Kim, I. H.; Crawford, J. J. Org. Chem. 2003, 68, 6096.

172

(a) Jung, M. E.; Hudspeth, J. P. J. Am. Chem. Soc. 1980, 102, 2463; (b) Idem Ibid. 1978, 100, 4309.

173

Schulze, V.; Nell, P. G.; Burton, A.; Hoffmann, R. W. J. Org. Chem. 2003, 68, 4546.

174

Solladie, G. In Asymmetric Synthesis Vol. 2, Morrison, J. D., Ed.; Academic Press: New York, NY, 1983, pp 157–183.

175

(a) Cohen, H. L.; Wright, G. F. J. Org. Chem. 1953, 18, 432; (b) Allentoff, N.; Wright, G. F. Ibid. 1957, 22, 1.

176

Iffland, D. C.; Davis, J. E. J. Org. Chem. 1977, 42, 4150.

575

11.4 GRIGNARD REAGENTS (CdMG)

Nozaki et al.,177 addition of ()-sparteine to the reaction of ethylmagnesium bromide and benzaldehyde gave (+)-1phenylpropan-(1R)-ol in 15% yield and 22 %ee. Under the same conditions, reaction with acetophenone gave racemic 2-phenylbutan-2-ol in 11% yield. Seebach et al. 178 showed that addition of (2S,3S)-2,3-dimethoxy-N1,N1,N4,N4-tetramethylbutane-1,4-diamine to benzaldehyde gave 1-phenylpropan-1-ol in 6–8 %ee. Inch et al. added furanose 81179 to a mixture of methylmagnesium bromide and 1-phenylheptan-1-one, and obtained a 95% yield of (+)-2-phenyloctan(2R)-ol with 70 %ee. Under similar conditions, however, ethylmagnesium bromide reacted with acetophenone to give a 48% yield of ()-2-phenylbutan-(2S)-ol with 27 %ee.179 Chiral amino alcohols were used as additives by Battioni and Chodkiewicz, but gave optical yields of only 0–20%.180 Mukaiyama et al.181 obtained good results with the chiral ligand 82 in reactions of organolithium, dialkylmagnesium, and Grignard reagents with benzaldehyde. When 82 was added to the reaction, the optical yield of alcohol 83 depended on both the solvent and the temperature of the reaction. In case (2), the stereogenic center of the Grignard might be expected to give some selectivity for the synor anti-diastereomeric product. Configurational instability of the CdMg bond156,157,160 at the putative stereogenic carbon, however, leads to an equilibrating mixture that is effectively a chiral, racemic Grignard reagent, with poor syn-anti selectivity in the diastereomeric alcohol mixture. Me Me O R2 Mg X O

H

Me

H

N

Me

H

Me 80

Me

OMe

N

Me2N

H

O O

OH O

NMe2

O

OMe

(–)-Sparteine

O

(2S,3S)-2,3-Dimethoxy-N1, N1, N4, N4tetramethylbutane-1,4-diamine

81

Me

Me

OH H R M

+

Ph

CHO

N

+

R H Ph

N Me

82

OH

83

If the carbonyl partner contains a stereogenic carbon, as in case (3), the asymmetry of that center should exert an influence when the Grignard reagent reacts with the prochiral carbonyl (the faces of the carbonyl are diastereotopic), leading to some diastereoselectivity in the alcohol product. This system formed the basis of the Cram model,182 the Cornforth et al.183 model, the Felkin-Anh model,184 and the Karabatsos185 model, as discussed in Section 7.9.1. The Cram model, the Cram chelation model, or the Felkin-Anh model is generally used to predict selectivity in addition of Grignard reagents to a prochiral carbonyl with stereogenic α- or β-carbons. The presence of a stereogenic center in one of the reactive partners induces facial bias on approach of the two reagents, leading to moderate-to-good diastereoselection for either the syn- or the anti-isomer. Examples taken from Mosher and Morrison’s compilation186 show reactions of Grignard reagents with aldehydes with a stereogenic center at the α-carbon. In all case (3) reactions, Cram’s open-chain model was used to predict the diastereoselectivity. Ketone 84 reacted with various Grignard reagents (R2MgX) to yield a mixture of diastereomeric

177

(a) Nozaki, H.; Aratani, T.; Toraya, T.; Noyori, R. Tetrahedron 1971, 27, 905; (b) Nozaki, H.; Aratani, T.; Toraya, T. Tetrahedron Lett. 1968, 4097.

178

(a) Seebach, D.; Langer, W. Helv. Chim. Acta 1979, 62, 1701; (b) Seebach, D.; Crass, G.; Wilka, E. -M.; Hilvert, D.; Brunner, E. Ibid. 1979, 62, 2695.

179

Inch, T. D.; Lewis, G. J.; Swainsbury, G. L.; Sellers, D. J. Tetrahedron Lett. 1969, 3657.

180

Battioni, J. P.; Chodkiewicz, W. Bull. Soc. Chim. Fr. 1972, 2068.

(a) Mukaiyama, T.; Soai, K.; Sato, T.; Shimizu, H.; Suzuki, K. J. Am. Chem. Soc. 1979, 101, 1455; (b) Soai, K.; Mukaiyama, T. Chem. Lett. 1978, 491; (c) Sato, T.; Soai, K.; Suzuki, K.; Mukaiyama, T. Ibid. 1978, 601; (d) Mukaiyama, T.; Suzuki, K.; Soai, K.; Sato, T. Chem. Lett. 1979, 447.

181

182

(a) Cram, D. J.; Abd Elhafez, F.A J. Am. Chem. Soc. 1952, 74, 5828; (b) Cram, D. J.; Kopecky, K. R. Ibid. 1959, 81, 2748.

183

Cornforth, J. W.; Cornforth, R. H.; Mathew, K. K. J. Chem. Soc. 1959, 112.

184

Cherest, M.; Felkin, H.; Prudent, N. Tetrahedron Lett. 1968, 2199.

185

Karabatsos, G. J. J. Am. Chem. Soc. 1967, 89, 1367.

186

Reference 160, pp 92–93.

576

11. CARBON-CARBON BOND-FORMING REACTIONS

alcohols, 85 and 86. In all cases, 85 was the major product by a ratio of 2:1 to 4:1 85:86. Examination of a LUMO map (see Section 7.9.1) for Grignard reactions 84 (RS ¼ H, RM ¼ Me, RL ¼ Ph, R1 ¼ H) reveals a clear preference for 85. RM O RS

R2MgX

RS

H

OH R2

RM RL

RL 84

RM +

R2

OH

RS

H

H

RL

85

86

As discussed in Section 7.9.1, the Cram cyclic (chelation) model (87)187 is preferred for reactions of ketones that have heteroatoms in the α-position. The chelation model is very common in Grignard reactions, including reactions with acyclic systems, and it assumes that the Mg is coordinated to both the heteroatom and the carbonyl oxygen, as shown in 87. For this reason, coordination of an organometallic with carbonyls that have a heteroatom in the α- and β-positions has been termed chelation control when it results in formation of products with high diastereoselectivity.188 Part of this model assumes that the incoming group in 87 (R attached to Mg) is delivered from the less hindered face (over the smallest group, RS). In this model, the R1 group could be H or alkyl. An example of chelation control is the reaction of α-benzyloxy ketone [(R)-3-(benzyloxy)butan-2-one] with vinylmagnesium bromide, a key step in the Williams and White189 synthesis of ()-citreoviridin. In this case, diastereomer (3R,4R)-4-(benzyloxy)-3methylpent-1-en-3-ol was formed exclusively, and the chelated model (88) predicted the observed stereochemistry. X R

R1

Mg O O • RS

R1

RL

87 Br H BnO

Me

Me O

MgBr

THF, 0 °C

(R)-3-(Benzyloxy)butan-2-one

Mg BnO

H3O+

O

H Me

BnO H Me

H OH Me

Me OH Me

BnO

Me 88

(3R,4 R)-4-(Benzyloxy)-3-methylpent-1-en-3-ol

Case (4) substrates from above should give the best diastereoselectivity. The instability of the CdMg bond, however, leads to poor diastereoselectivity for the carbon bearing the Mg as observed in case (2). The stereogenic center on the carbonyl substrate remains intact, and it influences the reaction, leading to good diastereoselectivity in the alcohol product upon reaction with a prochiral carbonyl derivative. For all practical purposes this substrate is a case (3) situation, except that there is loss of diastereoselectivity due to the inability to control the configuration of the stereogenic center in the Grignard reagent. Reaction of an allylic magnesium halide with an electrophile can give a new bond at either carbon of the allylic carbanion, so two isomeric products are possible.190 Felkin and coworkers191 described the reaction with ketones as a noncyclic SE20 rearrangement, and it is sensitive to steric encumbrance at the carbonyl carbon. Allylic Grignard reagent 89, when generated from either the (E)- or the (Z)-bromide, formed an equilibrating mixture (89 and 92). When this mixture of Grignard reagents reacted with a ketone (via 90 and 93, respectively), isomeric alcohols 91 and 94 were formed. Increasing the steric bulk of R1 and R destabilizes 90 relative to 93, favoring 94 over 91. The but-2-enyl

187

(a) Reference 160, pp 94–98; (b) Cram, D. J.; Wilson, D. R. J. Am. Chem. Soc. 1963, 85, 1245.

188

Still, W. C.; Schneider, J. A. Tetrahedron Lett. 1980, 21, 1035.

189

Williams, D. R.; White, F. H. J. Org. Chem. 1987, 52, 5067.

190

Hoffmann, R. W. Angew. Chem. Int. Ed. 1982, 21, 555.

191

Cherest, M.; Felkin, H.; Frajerman, C. Tetrahedron Lett. 1971, 379.

577

11.5 GRIGNARD REAGENTS: REDUCTION, ORGANOCERIUM REAGENTS, AND ENOLIZATION

derivative (R1 ¼ Me) reacted primarily as the crotyl derivative (89),192 but 89 was in equilibrium with a small amount of 92.193 When 89/92 reacted with aldehydes194 or unhindered ketones,195 α-methallyl alcohol 91 was formed (94/91 ratio is 90% of the reduction product, alcohol 2-(tert-butyl)-2,4,4-trimethylpentan-1-ol with none of the carbonyl addition product.201 The reduction is thought to proceed by a coordination complex that leads to a transition state close to 95, which is similar to the transition state observed in the Meerwein-Ponndorf-Verley reduction202 (Section 7.11.8) and the Oppenauer oxidation (Section 6.2.6.2). Coordination between the carbonyl oxygen and magnesium leads to complex 95, where hydrogen transfer yield 2-(tert-butyl)-2,4,4-trimethylpentan-1-ol, accompanied by elimination of H and MgX from the Grignard to give an alkene (in this case isobutylene). Hamelin201b showed that even an unhindered aldehyde (e.g., propanal) can be reduced with a small Grignard reagent (e.g., ethylmagnesium bromide). A Grignard reaction between hindered ketones and aldehydes often leads to reduction as a synthetically useful process. Camphor is reduced to borneol and isoborneol upon reaction with Grignard reagents.203 The ratio of borneol/isoborneol was close to 1:1 and there was a slight preference for isoborneol upon reaction with ethylmagnesium bromide, isopropylmagnesium bromide, and n-propylmagnesium bromide.203 More hindered reagents gave a higher percentage of isoborneol, consistent with reduction via delivery of a hydrogen atom from the less hindered endo-face.

RMgX

H

+

OH

O H Isoborneol

OH Camphor

Borneol

The degree of asymmetric induction in a reduction process varies with the chiral Grignard reagent and asymmetric solvent.204 The %ee for reduction is usually small. Nasipuri et al.205 described the asymmetric reduction of phenylalkyl ketones.206 Both a cyclic model207 and an acyclic model208 were proposed to explain the selectivity. Asymmetric induction was higher with phenylalkyl ketones than with cyclohexylalkyl ketones or tert-butylalkyl ketones.209 Since the Grignard reduction can be used synthetically, the appropriate functional group transform is R

R O

R

H

O R

11.5.2 Organocerium Reagents Reduction in Grignard reactions of hindered ketones can severely limit a carbonyl alkylation reaction. This limitation can often be circumvented, however, by the conversion of the Grignard reagent to an organocerium reagent.

200

Reference 79, p 138.

201

(a) Whitmore, F. C.; Whitaker, J. S.; Mosher, W. A.; Breivik, O. N.; Wheeler, W. R.; Miner, Jr., C. S.; Sutherland, L. H.; Wagner, R. B.; Clapper, T. W.; Lewis, C. E.; Lux, A. R.; Popkin, A. H. J. Am. Chem. Soc. 1941, 63, 643; (b) Hamelin, A. Bull. Soc. Chim. Fr. 1961, 926. (a) Meerwein, H.; Schmidt, R. Ann. 1925, 444, 221; (b) Ponndorf, W. Angew. Chem. 1926, 39, 138; (c) Verley, A. Bull. Soc. Chim. Fr. 1925, 37, 537, 871; (d) Wilds, A. L. Org. React. 1944, 2, 178.

202

203

Malkonen, P. J. Suomen Kemistilehti 1965, 38B, 89; (Chem. Abstr. 63: 8411b 1965).

(a) Birtwistle, J. S.; Lee, K.; Morrison, J. D.; Sanderson, W. A.; Mosher, H. S. J. Org. Chem. 1964, 29, 37 and references cited therein; (b) Reference 160, pp 177–202. 204

205

Nasipuri, D.; Ghosh, C. K.; Mukherjee, P. R.; Venkataraman S. Tetrahedron Lett. 1971, 1587 and references cited therein.

206

Reference 160, pp 182–187.

207

Mathieu, J.; Weill-Raynal, J. Bull. Soc. Chim. Fr. 1968, 1211.

208

Cabaret, D.; Welvart, Z. Chem. Commun. 1970, 1064.

209

Morrison, J. D. Survey of Progress in Chemistry, Vol. 3, Academic Press: New York, NY, 1966, p 147.

579

11.5 GRIGNARD REAGENTS: REDUCTION, ORGANOCERIUM REAGENTS, AND ENOLIZATION

Imamoto et al.210 showed that methyl mesityl ketone (1-mesitylethan-1-one) gave only 10% of the acyl addition product [2-(2,4-dimethylphenyl)hexan-2-ol] upon reaction with butylmagnesium bromide. Butylcerium(III) chloride, formed by the reaction of butylmagnesium bromide and cerium chloride (CeCl3), reacted with butylcerium(III) chloride to give a 57% yield of 2-(2,4-dimethylphenyl)hexan-2-ol.210 CeCl3

+

CeCl2

MgBr Butylmagnesium bromide

MgBrCl

Butylcerium(III) chloride

Me O

OH

Butylcerium(III) chloride

Me

Me

THF, –78°C, 4 h

C4H9 Me

Me Me

Me 2-(2,4-Dimethylphenyl)hexan-2-ol

1-Mesitylethan-1-one

A synthetic example using an organocerium reagent is taken from Majetich and Zhang’s211 synthesis of perovskone, in which sterically hindered ketone 96 reacted with the organocerium reagent derived from vinylmagnesium bromide and CeCl3 to yield the 1,2-addition product 97. In this example, mild acid hydrolysis led to elimination of water and formation of the extended conjugated ketone 98 in an overall yield of 90%. MeO

OMe

OMe

MeO

OMe CeCl3

OMe

H+

MgBr

O

MeO

O

OMe

OH

OMe

OMe

OMe

96

97

98 (90%)

Organolithium reagents, which will be discussed in Section. 11.6, are also useful precursors. This variation is listed because many organocerium reagents are generated from the corresponding organolithium reagent. Vinylcerium reagent 100, for example, was prepared from vinyl bromide 99, in Ovaska and Roses’212 synthesis of fused polycyclic ring systems. When 100 reacted with the cyclopentanone derivative shown, alcohol 101 was isolated in 61% yield after hydrolysis. O

O

O

1. t-BuLi 2. CeCl3

O

O

O

1. 2. aq NH4Cl

O OH

CeCl2

Br 99

SiMe3

SiMe3

100

101 (61%)

11.5.3 Enolization Aldehydes and ketones have acidic protons on the α-carbon, which can be removed by the basic Grignard reagent in an acid-base reaction that generates an enolate anion (Section 13.2.1). If the acyl addition reaction is slow, the acid-base reaction can favorably compete with nucleophilic addition to the carbonyl. If there are no secondary reactions, hydrolysis of the enolate product regenerates the starting ketone and the net result is isolation of the starting material.

(a) Imamoto, T.; Sugiura, Y.; Takiyama, N. Tetrahedron Lett. 1984, 25, 4233; (b) Imamoto, T.; Sugiura, Y. J. Organomet. Chem. 1985, 285, C21; (c) Imamoto, T.; Takiyama, N.; Nakamura, K Tetrahedron Lett. 1985, 26, 4763.

210

211

Majetich, G.; Zhang, Y. J. Am. Chem. Soc. 1994, 116, 4979.

212

Ovaska, T. V.; Roses, J. B. Org. Lett. 2000, 2, 2361.

580

11. CARBON-CARBON BOND-FORMING REACTIONS

Nucleophilic addition to the carbonyl is usually faster than deprotonation, but small amounts of enolization can be observed, as shown in Table 11.1 for diisopropyl ketone (2,4-dimethylpentan-3-one).213 With unhindered Grignard reagents, insignificant amounts of enolate anion formed, and reduction is the major reaction that competes with nucleophilic addition. With the hindered isopropylmagnesium halide, steric interactions completely suppressed addition, and reduction accounted for 69% of products with only 2% enolization. Hamelin214 observed similar reactivity in reactions with ethylmagnesium bromide, in which the effect of solvent and temperature on reduction and enolization was probed. Increasing the size of the groups attached to the carbonyl increased the amount of reduction, but only slightly increased the amount of enolization. Increasing the size of the group attached to Mg, in the presence of a hindered ketone, significantly increased enolate formation.

TABLE 11.1

Competition for Nucleophilic Addition, Reduction, and Enolate Formation in Reactions of Diisopropyl Ketone With Grignard Reagents

O RMgX

O

OH

O

+

+ 2,4-Dimethylpentan-3-one

R

H2O Reduction

Enol

Addition

RMgX

RMgX/Ketone

EtMgBr

1.3

1

21

78

1.5

1

19

80

2.5

1

15

80

1.2

2

21

77

1.2

2

51

46

2.5

1

37

62

1.2

1

64

35

1.4

2

60

36

n-PrMgI

1.2

2

69

30

i-PrMgCl

1.2

28

72

0

i-PrMgBr

1.2

29

65

0

1.4

30

65

0

1.4

30

70

0

n-PrMgCl

n-PrMgBr

i-PrMgI

% Enolate

% Reduction

% Addition

Reprinted with permission from Cowan, D. O.; Mosher, H. S. J. Org. Chem. 1962, 27, 1. Copyright © 1962 American Chemical Society.

11.6 ORGANOLITHIUM REAGENTS (CdLI) Grignard reagents are clearly important in synthesis. If the MgX unit attached to carbon is replaced with Li, another class of organometallic reagents is available, the organolithium reagents. Organolithium reagents are also potent nucleophiles and they are more basic than Grignard reagents. Both features can be exploited in synthesis.215

213

Cowan, D. O.; Mosher, H. S. J. Org. Chem. 1962, 27, 1.

214

Hamelin, R. Bull Soc. Chim. Fr. 1961, 915 and references cited therein.

215

Organolithiums in Enantioselective Synthesis (Series: Topics in Organometallic Chemistry, Vol. 5), Hodgson, D. M., Ed.; Springer-Verlag: Heidelberg, 2003.

581

11.6 ORGANOLITHIUM REAGENTS (CdLI)

11.6.1 Preparation, Structure, and Stability Organolithium reagents, characterized by a CdLi bond, are as important in organic synthesis as the Grignard reagents. Lithium is less electronegative than carbon, so the carbon of the CdLi bond is polarized δ, as in Grignard reagents.216 Organolithium reagents are expected to behave both as a nucleophile and as a base. It is important to understand the chemical properties of organolithium reagents, and to note differences with Grignard reagents before discussing their reactions.217 Ethyllithium was discovered by Schlenk and Holtz218 in 1917, but it was not until 1930 that Ziegler and Colonius219 discovered that organolithium reagents could be prepared by direct reaction of Li metal and an alkyl halide,220 as in the reaction of Li and chlorobutane to yield butyllithium (CH3CH2CH2CH2Li or BuLi). This halogen-lithium exchange reaction is quite useful in synthesis, as seen by the conversion of bromide 102 to 103 by treatment with tert-butyllithium at low temperature, taken from a synthesis of okadaic acid by Forsyth and coworkers.221 In this synthesis, 103 was subsequently reacted with an aldehyde (see Section 11.6.4) to give a key intermediate in the preparation of one segment of okadaic acid. Two years prior to the work of Ziegler and Colonius,219 Schlenk and Bergmann222 had reported that ethyllithium (CH3CH2Li or EtLi) reacted with fluorene, with exchange of the acidic hydrogen with Li to yield fluorenyllithium. This transformation is an acid-base reaction that is commonly referred to as a lithium-hydrogen exchange reaction. O I

O

t-BuLi, Ether –78°C to rt

O

Li

O

102 H

103 Li

H

H

EtJLi

+ EtJH Fluorene

Fluorenyllithium

The lithium-halogen exchange and the lithium-hydrogen exchange reactions constitute the main methods for preparing organolithium reagents. The latter also describes the most useful synthetic application, which is removal of an acidic hydrogen to generate a new organolithium species. Organolithium reagents are usually written as RdLi, but this simple representation does not describe their actual structure.217,220,223 Some physical data220 suggest that the CdLi bond in an organolithium reagent is highly covalent, and it is known that organolithium reagents exist as associated aggregates, due in part to this covalent character. Both methyllithium (CH3Li)224 and ethyllithium (CH3CH2Li)225 have been obtained in crystalline form and their crystal structure determined by X-ray crystallography. The crystal structure of isopropyllithium as reported by Siemeling et al.226 is shown in Fig. 11.1, and it was shown to exist as a hexamer, (i-PrLi)6. The X-ray crystal structure of 9,9dilithiofluorene has also been reported,227 the first for a dilithiated hydrocarbon.

216

(a) Wakefield, B. J. The Chemistry of Organolithium Compounds; Pergamon Press: Oxford. UK, 1974; (b) Clayden, J. Organolithiums: Selectivity for Synthesis; Pergamon: Amsterdam, The Netherlands, 2002. 217

Mallan, J. M.; Bebb, R. L. Chem. Rev. 1969, 69, 693.

218

Schlenk, W.; Holtz, J. Berichte 1917, 50, 262.

219

Ziegler, K.; Colonius, H. Annalen 1930, 479, 135.

220

Deberitz, J. Janssen Chim. Acta 1984, 2, 3.

221

Fang, C.; Pang, Y.; Forsyth, C. J. Org. Lett. 2010, 12, 4528.

222

Schlenk, W.; Bergmann, E. Annalen 1928, 463, 1 (see p 98).

223

(a) Brown, T. L. Adv. Organomet. Chem. 1965, 3, 365; (b) Brown, J. M. Chem. Ind. (London) 1972, 454.

224

Weiss, E.; Lucken, E. A. C. J. Organomet. Chem. 1964, 2, 197.

225

Dietrich, H. Acta Crystallogr. 1963, 16, 681.

226

Siemeling, U.; Redecker, T.; Neumann, B.; Stammler, H. -G. J. Am. Chem. Soc. 1994, 116, 5507.

227

Linti, G.; Rodig,, A.; Pritzkow, H. Angew. Chem. Int. Ed. 2002, 41, 4503.

582

11. CARBON-CARBON BOND-FORMING REACTIONS

FIG. 11.1

X-ray crystal structure of isopropyllithium (H atoms are hidden for clarity). Reprinted with permission from Siemeling, U.; Redecker, T.; Neumann, B.; Stammler, H.-G. J. Am. Chem. Soc. 1994, 116, 5507. Copyright © 1994 American Chemical Society. Drawn with Spartan software, Wavefunction, Inc. Hydrogen atoms are omitted for clarity.

Methyllithium and ethyllithium are tetrameric in the solid state,225,228 and most organolithium reagents are highly associated in solution.216a,217 The degree of association is related to the solvent and structure of the organolithium reagent, but tends to be higher for straight chain when compared with branched (secondary and tertiary) organolithium reagents. The extent of association of the organolithium reagent is important since it can affect the rate of metal-halogen or metal-hydrogen exchange, as well as the product distribution. The degree of association of several common organolithium reagents in various solvents is shown Table 11.2,220,223a,229 and it was shown that the degree of association decreases as the coordinating ability of the solvent increases. This observation can loosely be compared with increasing solvent polarity. TABLE 11.2

Association of Organolithium Reagents in Common Solvents n

R

Solvent

n

Et

cyclohexane

6

hexane

6

ether

2

benzene

6,2

cyclohexane

6

benzene

6,2

ether

6

Bu (+ TMEDA)

hexane

1

t-Bu

hexane

4

benzene

4

ether

2

Bu

Ph

Reprinted with permission from Mallan, J. M.; Bebb, R. L. Chem. Rev., 1969, 69, 693. Copyright © 1969 American Chemical Society.

The most commonly used solvents in reactions of organolithium reagents are ether, THF, pentane, or hexane.230 Increased solvation diminishes the ability of the organolithium reagent to associate, and thereby form solvated 228

Dietrich, H. Z. Naturforsch. 1959, 14B, 739.

229

(a) Reference 217, p 696; (b) Brown, T. L. Acc. Chem. Res. 1968, 1, 23; (c) West, P.; Waack, R. J. Am. Chem. Soc. 1967, 89, 4395.

230

Miginiac-Groizeleau, L. Bull. Soc. Chim. Fr. 1963, 1449.

11.6 ORGANOLITHIUM REAGENTS (CdLI)

583

oligomers.230 Williard and Nichols231 determined the X-ray structures of the •TMEDA, •THF, and •DME complexes of butyllithium, which were tetramers. The ionic character of the relatively covalent CdLi bond is increased by complexation, lowering the energy requirements for the transition state leading to reactions.232 ()-Sparteine is often added to organolithium reagents to yield a reagent that leads to asymmetric induction in certain reactions. Strohmann et al. 233 determined the X-ray crystal structure of tert-butyllithium•()-sparteine, which has the monomeric structure 104. Although ethers are used as solvents for reactions of organolithium reagents, a chemical reaction is possible. Ethers are Lewis bases and coordinate to Li, stabilizing the organometallic reagent and participating in subsequent reactions. In most cases, an ether fragment does not appear in the final product.,216,234 However, a relatively slow reaction is possible in which the α-hydrogen atom of an ether is removed by the basic organolithium reagents, via LidH exchange. This exchange is an acid-base reaction, and the conjugate acid is an alkane, RLi ! RdH). Disproportionation of the lithio-ether derivative (105) generates an alkoxide and an alkene when THF is the solvent.235

Strohmann, C.; Seibel, T.; Strohfeldt, K. Angew. Chem. Int. Ed. 2003, 42, 4531. Copyright © 2003 Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission. Structure drawn with Spartan by Wavefunction, Inc. Hydrogen atoms are omitted for clarity.

The acidity of the α-proton varies with the ether and the organolithium reagent. It is not surprising, therefore, that organolithium reagents have different stabilities in different solvents. The half-life for the acid-base reaction of n-butyllithium with diethyl ether at 25°C is 153 h, whereas that of ethyllithium is 54 h, and that of cyclohexyllithium is 30 min.236 This stability to reaction of n-butyllithium can be used as an approximate reaction order for LidH exchange. Note that the relative stability of organolithium reagents can be determined, based on the tin-lithium exchange equilibrium reaction.237

231

Nichols, M. A.; Williard, P. G. J. Am. Chem. Soc. 1993, 115, 1568.

232

(a) Waack, R.; Doran, M. A. Chem. Ind. (London) 1962, 1290; (b) Zakharkin, L. I.; Okhlobystin, O. Yu.; Bilevitch, K. A. Tetrahedron 1965, 21, 881.

233

Strohmann, C.; Seibel, T.; Strohfeldt, K. Angew. Chem. Int. Ed. 2003, 42, 4531.

(a) Reference 217, p 696; (b) Applequist, D. E. O’Brien, D. F. J. Am. Chem. Soc. 1963, 85, 743; (c) Curtin, D. Y.; Koehl, Jr., W. J. Ibid. 1962, 84, 1967; (d) Eastham, J. F.; Gibson, G. W. J. Org. Chem. 1963, 28, 280; (e) Eastham, J. F.; Gibson, G. W. J. Am. Chem. Soc. 1963, 85, 2171; (f ) Mulvaney, J. E.; Gardlund, Z. G.; Gardlund, S. L. Ibid. 1963, 85, 3897; (g) Waack, R.; Doran, M. A.; rson, P. E. J. Organomet. Chem. 1965, 3, 481. 234

(a) Reference 217, p 697; (b) Bartlett, P. D.; Friedman, S.; Stiles, M. J. Am. Chem. Soc. 1953, 75, 1771; (c) Rembaum, A.; Siao, S. P.; Indictor, N. J. Polymer. Sci., 1962, 56, S17; (d) Gilman, H.; Gaj, B. J. J. Org. Chem. 1957, 22, 1165.

235

236

(a) Reference 217, pp 697, 716; (b) Seyferth, D.; Cohen, H. M. J. Organomet. Chem. 1963, 1, 15.

237

Graña, P.; Paleo, M. R.; Sardina, F. J. J. Am. Chem. Soc. 2002, 124, 12511.

584

11. CARBON-CARBON BOND-FORMING REACTIONS

Me

Me

H 2O

O– Li

EtLi

OH

O

O– LiH

THF

Li

O

H 2O

+ H2C CH2

O Li

Me

CHO

105

Primary organolithium reagents exhibit better stability (they are less reactive) when compared to secondary or tertiary reagents. It has also been determined that butyllithium decomposes after THF after 2 h at 25°C.238 The stability of butyllithium can be increased by the addition of a less acidic cosolvent (e.g., diethyl ether or THP). Solutions of butyllithium are stable for 24 h in an ether-THF mixture and up to 1 week in an ether-THP solution.238 Addition of a hydrocarbon cosolvent (e.g., pentane, hexane, or cyclohexane) to the ether also extends the lifetime of the organolithium reagent. In DME, the half-life of tert-butyllithium is 11 min at 70°C, but

Li

Li >

>

Li

Li >

Li >

>

Li

Li > Li

>

Li

Li >

>

>

Li > CH3Li

NMe2 > Li

The reactivity of organolithium reagents in ethers can be increased by the use of a Lewis base that is stronger than diethyl ether or THF. When DME is used as a solvent, for example, the reactivity is greatly increased.241 The two oxygen atoms in dimethoxyethane coordinate to the Li of the organolithium reagent, and generate what is essentially a monomeric organolithium species (RLi-DME). Such coordination may also increase the polarity of the CdLi bond, and thereby the carbanionic character. Similar effects are observed when stoichiometric amounts of nitrogen bases (e.g., DABCO242 and TMEDA)216,242 are added. Alkoxides (e.g., potassium tert-butoxide) have also been used,243 but they are less effective. The effects of the additives are apparent in the reaction of benzene with n-butyllithium.243 A refluxing solution of n-butyllithium in benzene is rather stable, giving Br > Cl. The % of racemization was 29–34%, when X in 113 is Cl, but increased to 54–58% when X is Br and to 64% when X is I.161 Further, the reaction of 113 [X ¼ I, 1% Na (25 μ)] gave 64% racemization, but decreasing the Na concentration to 0.002% gave 87% racemization. An increase in the particle size of the Na increased the amount of racemization. Both results are consistent with a complex heterogeneous surface reaction. The nature of the halide, the percentage of Na, and the particle size all contribute to the relative rate of formation of the organolithium, as well as the stereoselectivity of the reaction.274 Note that the choice of solvent can also influence the stereochemistry of organolithium reactions. There is some evidence that organolithium reagents have greater stereochemical stability in hydrocarbon solvents than they do in ether.275 Me X Ph

Ph

Me

1. Li , Ether 2. CO2

CO2H

3. H3O+

113

Ph

Ph 114

11.6.5.2 Diastereoselective Acyl Addition Reactions If the ketone or aldehyde contains a stereogenic center α to the carbonyl, addition of the organolithium reagent generates a new stereogenic center, and the reaction can proceed with good diastereoselectivity. The stereoselectivity of addition for Grignard reagents and organolithium reagents with aldehydes and ketones is predicted in a manner similar to that observed with reducing agents in Section 7.9. Acyl addition of organolithium reagents to ketones and aldehydes usually follows Cram’s rule (Section 7.9.A), for example, as seen in the reaction of 115 with organolithium reagents.276 Since a Li atom does not coordinate as readily as a Mg atom in reactions with carbonyls, chelation effects are somewhat diminished relative to Grignard reagents, and the Cram cyclic model is not as useful for predicting the selectivity. When complexing agents are added to the reaction, however, the cyclic model does predict the major diastereomer. A variety of organolithium reagents react with ketones or aldehydes (115) to yield two diastereomers, 116 and 117, and 116 was the major product in virtually all cases, as predicted by the Cram model.

(a) Witanowski, M.; Roberts, J. D. J. Am. Chem. Soc. 1966, 88, 737; (b) Lardicci, L.; Lucarini, L.; Palagi, P.; Pino, P. J. Organomet. Chem. 1965, 4, 341; (c) Buncel, E. Carbanions: Mechanistic and Isotopic Aspects; Elsevier: New York, NY, 1975.

273

274

Walborsky, H. M.; Aronoff, M. S. J. Organomet. Chem. 1965, 4, 418.

275

(a) Curtin, D. Y.; Koehl, Jr., W. J. J. Am. Chem. Soc. 1962, 84, 1967; (b) Reference 216a, pp 697–698.

276

Reference 160, pp 92–93.

590

11. CARBON-CARBON BOND-FORMING REACTIONS

O RM RS

1. R2Li

R1

2. H2O

HO RM RS

R2

+

R1

RL

RL

115

116

RM RS

R2 OH R1 RL 117

The reaction with cyclic ketones usually favors formation of the syn-diastereomer.277 As with Grignard reagents, attack at a carbonyl occurs at the less sterically hindered face of the molecule. Reaction of various organolithium and Grignard reagents with 2-methylcyclopentanone, for example, gave a mixture of cis (118) and trans (119) alcohols, generally favoring 118. OH

O R—M

R R Me

Me

OH

+

Me

118

119

11.6.6 Reactions of Organolithium Reagents With Other Functional Groups Organolithium reagents react with carbonyl compounds other than aldehydes and ketones, analogous to Grignard reagents.100 Esters usually react to yield a tertiary alcohol,278 and acid chlorides279 yield mixtures of unreacted starting material and tertiary alcohol. Adding an excess of the organolithium reagent yields the tertiary alcohol exclusively, which is formally analogous to the acyl substitution reactions noted for Grignard reagents (Section 11.4.3.2). If an ester is treated with excess methyllithium, for example, a tertiary dimethyl carbinol results. This reaction was used in the conversion of 120 to 121 (87% yield), in the Williams and Shah280 synthesis of (+)-ileabethoxazole (121). Careful selection of RLi and the reaction conditions allows the synthesis of ketones rather than tertiary alcohols in some cases.281 Me H

Me H

O

Me

O

Me 3 equiv MeLi

N

N

THF, –20°C

Me MeO

Me Me HO

O

Me

120

121 (87%)

The reaction of organolithium reagents with amides can yield a ketone,120,282 but removal of the proton on the α-carbon of amides or lactams by the basic organolithium to form an enolate anion is a common side reaction. For this reason, nonnucleophilic bases (e.g., lithium diisopropylamide) are commonly used (Section 13.2.2).283 Organolithium reagents react with Weinreb’s amide (see Section 4.2.3) to yield ketones, just as it was noted for Grignard reagents in Section 11.4.3.2.284 Primary amides react with excess organolithium to yield a nitrile, as in the conversion 277

Battioni, J. -P.; Capmau, M. -L.; Chodkiewicz, W. Bull. Soc. Chim. Fr. 1969, 976.

(a) Nelson, P. H.; Strosberg, A. M.; Untch, K. G. J. Med. Chem. 1980, 23, 180; (b) Kleemann, A.; Hesse, J.; Engel, J. Arzneim-Forsch. 1981, 31, 1178; (Chem. Abstr. 95: 186978q 1981).

278

279

Locksley, H. D.; Murray, I. G. J. Chem. Soc. C 1970, 392.

280

Williams, D. R.; Shah, A. A. J. Am. Chem. Soc. 2014, 136, 8829.

281

(a) Corey, E. J.; Kim, S.; Yoo, S.-e.; Nicolaou, K. C.; Melvin, Jr., L. S.; Brunelle, D. J.; Falck, J. R.; Trybulski, E. J.; Lett, R.; Sheldrake, P. W. J. Am. Chem. Soc. 1978, 100, 4620; (b) Corey, E. J.; Trybulski, E. J.; Melvin, Jr., L. S.; Nicolaou, K. C.; Secrist, J. A.; Lett. R.; Sheldrake, P. W.; Falck, J. R.; Brunelle, D. J.; Haslanger, M. F.; Kim, S.; Yoo, S.-e. Ibid. 1978, 100, 4618; (c) Corey, E. J.; Melvin, Jr., L. S.; Haslanger, M. F. Tetrahedron Lett. 1975, 3117. 282

(a) Evans, E. A. J. Chem. Soc. 1956, 4691; (b) Jones, E.; Moodie, I. M. J. Chem. Soc. C 1968, 1195.

(a) Hullot, P.; Cuvigny, T.; Larchev^eque, M.; Normant, H. Can. J. Chem. 1976, 54, 1098; (b) Durst, T.; Van Den Elzen, R.; Legault, R. Ibid. 1974, 52, 3206; (c) Crouse, D. N.; Seebach, D. Berichte 1968, 101, 3113.

283

284

For a synthetic example taken from a synthesis of (+)-frontalin, see Kanada, R. M.; Taniguchi, T.; Ogasawara, K. Tetrahedron Lett. 2000, 41, 3631.

591

11.6 ORGANOLITHIUM REAGENTS (CdLI)

of phenylacetamide (2-phenylacetamide) to benzonitrile in 72% yield.285 Reaction of 2-phenylacetamide with 3 equiv. of butyllithium gave trilithiated species 122, and subsequent fragmentation gave (cyano(phenyl)methyl)lithium and hydrolysis gave 2-phenylacetonitrile.285 Li

NH2

Ph

3 C4H9Li

N-Li

Ph

THF–Hexane

O

OLi 2-Phenylacetamide

122 Li Ph

H3O+

Ph

CN

CN

(Cyano(phenyl)methyl)lithium

2-Phenylacetonitrile

In Sections 11.6.6 and 11.6.7, the reaction of a Grignard reagent with DMF (an amide) gave an aldehyde. Organolithium reagents react similarly. An example is taken the synthesis of (+)-linoxepin by Lautens and coworkers,286 in which the reaction of 5-bromobenzo[d][1,3]dioxole with n-butyllithium gave an aryllithium reagent, which reacted with DMF in a second step to give an aryl aldehyde, benzo[d][1,3]dioxole-5-carbaldehyde, in 95% yield after hydrolysis. Br

CHO 1. BuLi , THF , –78°C 2. DMF, THF –78°C to rt

O

O

O

O

5-Bromobenzo[d][1,3]dioxole

Benzo[d][1,3]dioxole-5-carbaldehyde (95%)

Vinyllithium reagents, as well as vinyl Grignard reagents, react with DMF to produce conjugated aldehydes. In Kende and coworker’s287 synthesis of Stachybotrys spirolactams, vinyl iodide 123 was treated with tert-butyllithium to initiate I$Li exchange, and the resulting vinyllithium reagent reacted with an excess of DMF to give an 83% yield of conjugated aldehyde 124. CHO

I 1. t-BuLi , –78°C 2. DMF (5 equiv)

BnO

BnO

H 123

H 124 (83%)

Organolithium addition to acid derivatives yields the follows disconnections: OH R1

R

R

CO2X

+

R1

O X

R1

R R

CO2X

R1

285

Kaiser, E. M.; Vaulx, R. L.; Hauser, C. R. J. Org. Chem. 1967, 32, 3640.

286

Qureshi, Z.; Weinstabl, H.; Suhartono, M.; Liu, H.; Thesmar, P.; Lautens, M. Eur. J. Org. Chem. 2014, 4053.

287

Deng, W. -P.; Zhong, M.; Guo, X. -C.; Kende, A. S. J. Org. Chem. 2003, 68, 7422.

+

R1

X

592

11. CARBON-CARBON BOND-FORMING REACTIONS

Organolithium reagents are more reactive than the corresponding Grignard reagent. The reaction of a Grignard reagent and a carboxylic acid yields the carboxylate salt. Reaction of an organolithium reagent with a carboxylic acid yields the expected lithium carboxylate salt, but a second equivalent adds to the carboxylate to give a ketone. 2-Furyllithium (furan-2-yllithium) reacted with acetic acid, for example, to give the dianion (125).288 Hydrolysis of 125 gave the hydrate, which was converted to the ketone [1-(furan-2-yl)ethan-1-one] under the reaction conditions. The reaction of 125 with acetic acid in situ, however, generated some ketone product prior to the hydrolysis step. In the presence of an excess of the furyllithium reagent, further reaction gave 1,1-di(furan-2-yl)ethan-1-ol as a byproduct.288 Me 2 O

Li

O

Furan-2-yllithium

+ O

O–

Me

Me

H3O+

O–

MeCO2H

O

O

1-(Furan-2-yl)ethan-1-one

125

O OH

1,1-Di(furan-2-yl)ethan-1-ol

The disconnection for ketone formation from the acid follows: O R

RJ CO2H +

R1J X

R1

Organolithium reagents react vigorously with CO2 to yield the corresponding carboxylic acid after hydrolysis. Typically, the organolithium solution is carefully poured onto dry ice. An example is taken from Maier and K€ uhnert’s synthesis of apicularen A,289 in which carboxylic acid 126 was prepared from 3-tetrahydropyranyloxyanisole in 77% yield. The conversion of 113 to 114 in Section 11.6.5.1 is another example of this reaction. OMe

OMe 1. BuLi, Ether , 0°C 2. CO2 3. H3O+

OTHP 3-(Tetrahydropyranyloxy)anisole

CO2H OTHP 126 (77%)

Nitriles can be hydrolyzed to, or prepared from carboxylic acids, and nucleophiles290 can attack the electrophilic carbon, so they are considered to be acid derivatives. 3-Thienyllithium (thiophen-3-yllithium) reacted with 4-methylpyridine carbonitrile (4-methylpicolinonitrile), for example, to yield a N-lithioimine (127) analogous to the reaction of Grignard reagents. Subsequent acid hydrolysis converted the initially formed imine to the ketone 128.291 Addition of an organolithium reagent initially generates an imine, so reduction to an amine (Sections 7.6.2 and 7.10.5 and 7.10.7) is also possible. Reaction of butyllithium with nitrile 129, for example, gave an 84% yield of lithiated imine 130,292 and reduction with methanolic sodium borohydride gave a 90% yield of amine 131.

288

Heathcock, C. H.; Gulik, L. G.; Dehlinger, T. J. Heterocyclic Chem. 1969, 6, 141.

289

K€ uhnert, S. M.; Maier, M. E. Org. Lett. 2002, 4, 643.

290

(a) Reference 217, p 698; (b) O’Sullivan, W. I.; Swamer, F. W.; Humphlett, W. J.; Hauser, C. R. J. Org. Chem. 1961, 26, 2306.

(a) Granados Jarque, R.; Bosch Cartes, J.; Lopez Calahorra, F. Spanish Patent 453,484; (Chem. Abstr. 90: P152156b 1979); (c) Koshinaka, E.; Ogawa, N.; Yamagishi, K.; Kato, H.; Hanaoka, M. Yakugaku Zasshi 1980, 100, 88; (Chem. Abstr. 93: 71507b 1980). 291

292

(a) Zhu, J.; Quirion, J. -C.; Husson, H. -P. Tetrahedron Lett. 1989, 30, 6323; (b) Arseniyadis, S.; Huang, P. Q.; Husson, H. -P. Ibid.1988, 29, 1391.

593

11.6 ORGANOLITHIUM REAGENTS (CdLI)

Me

Me

N

Me

Li

CN

S

+

N

S Thiophen-3-yllithium

O

N

O

C4H9Li

Bu O

Ether , –70°C

128 Ph

N NC

O

Li

127

Li

Ph

Ph N

O

H Bu

1. NaBH4 , MeOH 2. aq NH4Cl

O

O

S

N N

4-Methylpicolinonitrile

H3O+

NH2 N

O

O O

129

130 (84%)

131 (90%)

The disconnections for these two reactions follow: RJ CO2H

O

RJ X + CO2 R

O

2 RJ X + CO2 R

R

R1

R1JX + RJC LN

RJX

11.6.7 Directed Ortho Metalation When an aromatic ring has a heteroatom or a heteroatom-containing substituent, reaction with a strong base (e.g., an organolithium reagent) usually leads to an ortho-lithiated species.293 Subsequent reaction with an electrophilic species gives the ortho substituted product. Gilman and Bebb294a and Wittig and Fuhrman294b discovered this ortho selectivity independently in 1939–40, when anisole was found to yield ortho deprotonation in the presence of butyllithium.294 Hauser and Puterbaugh295 greatly contributed to this reaction. Snieckus296 provides a table of relative directing abilities of various groups, and several categories of functional groups are capable of this ortho-directing effect. An example is taken from the synthesis of chlorocyclinone A, Mal and Karmakar297 reacted 132 with tert-butylithium in the presence of TMEDA to give the ortho-organolithium reagent, which reacted with DMF to give aldehyde 133 in 59% yield. When two directing metalation groups have a 1,3-relationship on the aromatic ring, they work cooperatively to enhance the ortho selectivity.298 An example is taken from the Moriarty et al.299 synthesis of treprostinil. Treatment of anisole derivative 134 with butyllithium and then allyl bromide, gave a 60% yield of 135. This regioselective approach to preparing functionalized aromatic rings is an important addition to synthetic methodology.

For a review of directed ortho-metalation, see (a) Snieckus, V. Chem. Rev. 1990, 90, 879; (b) Gschwend, H. W.; Rodriguez, H. R. Org. React. 1979, 26, 1. Also see (c) Green, L.; Chauder, B.; Snieckus, V. J. Heterocyclic Chem. 1999, 36, 1453.

293

294

(a) Gilman, H.; Bebb, R. L. J. Am. Chem. Soc. 1939, 61, 109; (b) Wittig, G.; Fuhrman, G. Chem. Ber. 1940, 73, 1197.

295

(a) Puterbaugh, W. H.; Hauser, C. R. J. Org. Chem. 1964, 29, 853; (b) Slocum, D. W.; Sugarman, D. I. Adv. Chem. Ser., 1974, No. 130, 227.

296

See Ref. 293a, and Scheme 7 cited therein (pp 884–885).

297

Karmakar, R.; Mal, D. J. Org. Chem. 2012, 77, 10235.

298

See Ref. 293a, and Table 3 cited therein (p 885) for cooperative effects of several 1,3-related directed metalation groups.

299

Moriarty, R. M.; Rani, N.; Enache, L. A.; Rao, M. S.; Batra, H.; Guo, L.; Penmasta, R. A.; Staszewski, J. P.; Tuladhar, S. M.; Prakash, O.; Crich, D.; Hirtopeanu, A.; Gilardi, R. J. Org. Chem. 2004, 69, 1890.

594

11. CARBON-CARBON BOND-FORMING REACTIONS

CHO

1. t-BuLi , TMEDA

NEt2 OMe

NEt2

2. DMF , Dry THF –78°C to rt

O

O

OMe

132

133 (59%)

OSiMe2t-Bu

OSiMe2t-Bu 1. BuLi 2.

Br

OMe

OMe

135 (60%)

134

Collum and coworker’s300 rate studies of the butyllithium-TMEDA mediated lithiation reaction of five arenes suggest that the lithiation reactions do not necessarily rely on a complex-induced proximity effect. A triple-ion based model was proposed300 that depends largely on inductive effects proposed by Schlosser and Maggi.301 As a rule, the powerful organolithium bases should be used with organic solvents in which they are highly soluble, due to their association into aggregates (see Section 11.6.1).293a Additives (e.g., TMEDA) can be important because they effectively break down the aggregates to generate monomers and dimers in solution.300 Many years ago, Roberts and Curtin302 proposed that ortho lithiated species are stabilized by coordination, and Snieckus293a summarized the early evidence pointing to the idea that the directed ortho-lithiation process is a three-step sequence. The three steps are coordination of the organolithium aggregate to the heteroatom-containing aromatic compound, deprotonation to give the coordinated ortho-lithiated species, and finally, reaction with an electrophile to give the final product as with any organolithium reagent. A disconnection for these reactions is generalized follows: X

X

Y

11.6.8 Addition to Epoxides, Alkenes, and Alkynes 11.6.8.1 Reactions With Epoxides Epoxides are opened to the alcohol as expected, untroubled by the MgX2 catalyzed rearrangements that sometimes plague Grignard reactions. In Xie and coworker’s303 synthesis of ()-walsuchochin B, aryl bromide 1-bromo-4-methoxy-2,3-dimethylbenzene reacted with butyllithium to yield the corresponding aryllithium reagent, which subsequently reacts with ethylene oxide to give 2-(4-methoxy-2,3-dimethylphenyl)ethan-1-ol in 82% yield. In general, an organolithium reagent attacks an epoxide at the less sterically hindered (less substituted) carbon, as was noted with Grignard reagents (Section 11.4.5).

(a) Chadwick, S. T.; Rennels, R. A.; Rutherford, J. L.; Collum, D. B. J. Am. Chem. Soc. 2000, 122, 8640. Also see (b) Collum, D. B. Acc. Chem. Res. 1992, 25, 448.

300

(a) Maggi, R.; Schlosser, M. Tetrahedron Lett. 1999, 40, 8797; (b) Schlosser, M. Angew. Chem. Int. Ed. 1998, 37, 1497; (c) B€ uker, H. H.; Nibbering, N. M. M.; Espinosa, D.; Mongin, F.; Schlosser, M. Tetrahedron Lett. 1997, 38, 8519.

301

(a) Roberts, J. D.; Curtin, D. Y. J. Am. Chem. Soc. 1946, 68, 1658; (b) Morton, A. A. J. Am. Chem. Soc. 1947, 69, 969; (c) Chatgilialoglu, C.; Snieckus, V. In Chemical Synthesis Gnosis to Prognosis; NATO ASI Series E. Applied Sciences, Vol. 320, p 191. Kluwer Academic Publishers, The Netherlands, 1994.

302

303

Xu, S.; Gu, J.; Li, H.; Ma, D.; Xie, X.; She, X. Org. Lett. 2014, 16, 1996.

595

11.6 ORGANOLITHIUM REAGENTS (CdLI)

Br

1.BuLi , THF , –78°C 2. Ethylene oxide –78°C to rt

MeO

OH

MeO

1-Bromo-4-methoxy-2,3dimethylbenzene

2-(4-Methoxy-2,3-dimethylphenyl)ethan-1-ol (82%)

The epoxide ring opening gives the disconnection: HO

O

R1 R1—X

R

+ R

R

R

R

R

11.6.8.2 Intramolecular Addition to Alkenes and Alkynes While organolithium reagents react with alkenes or alkynes in an intramolecular reaction, both are relatively inert to the analogous intermolecular reaction. Ward and Lawler304 treated alkynyl bromide 6-bromo-1-phenylhex-1-yne with n-butyllithium, which gave a 60% yield of the alkylidene cyclopentane, phenylmethylenecyclopentane. There were several other products produced in this reaction, and the mechanistic pathway for the cyclization was not well understood. Bailey et al.305 showed that the tert-butyllithium exchange reaction and the cyclization reaction are rapid at low temperatures. Bailey et al.305 also showed the initially formed organolithium product could be trapped with a variety of electrophiles, including aldehydes.305 Cooke et al.306 showed that alkyllithium and alkenyllithium derivatives containing an ester moiety can be cyclized to form ketone products by direct addition or by Michael addition. Bicyclic and spirocyclic molecules can be prepared by addition of TMEDA to the initially formed organolithium, because it induces a second cyclization reaction. The organolithium products were trapped with various electrophiles,307 including formaldehyde.308 Ph

1. C4H9Li , Hexane–Ether

Br

2. H2O

6-Bromo-1-phenylhex-1-yne

Phenylmethylenecyclopentane (60%)

There is good enantioselectivity in the presence of the additive ()-sparteine. The reaction of N-allyl-Nbenzyl-2-bromoaniline with tert-butyllithium and ()-sparteine gave 136, which cyclized to yield 137. Quenching with water gave (R)-1-benzyl-3-methylindoline in 60% yield (65 %ee), in ether, and 85% yield (87 %ee) in toluene.309 When the reaction was done in THF, however, an 80% yield of (R)-1-benzyl-3-methylindoline was obtained, but with 0 %ee. The extent of asymmetric induction is dependent on the structure of the added ligand.310

304

Ward, H. R.; Lawler, R. G. J. Am. Chem. Soc. 1967, 89, 5517.

305

Bailey, W. F.; Ovaska, T. V.; Leipert, T. K. Tetrahedron Lett. 1989, 30, 3901.

(a) Cooke, Jr., M. P. J. Org. Chem. 1992, 57, 1495; (b) Cooke, Jr., M. P.; Widener, R. K. Ibid. 1987, 52, 1381; (c) Cooke, Jr., M. P. Ibid. 1984, 49, 1144; (d) Cooke, Jr., M. P.; Houpis, I. N. Tetrahedron Lett. 1985, 26, 4987.

306

307

Bailey, W. F.; Khanolkar, A. D. J. Org. Chem. 1990, 55, 6058.

308

See Bailey, W. F.; Bakonyi, J. M. J. Org. Chem. 2013, 78, 3493.

(a) Gil, G. S.; Groth, U. M. J. Am. Chem. Soc. 2000, 122, 6789; (b) Bailey, W. F.; Mealy, M. J. J. Am. Chem. Soc. 2000, 122, 6787. Also see (c) Luderer, M. R.; Mealy, M. J.; Bailey, W. F. J. Org. Chem. 2014, 79, 10722.

309

310

Mealy, M. J.; Luderer, M. R.; Bailey, W. F.; Sommer, M. B. J. Org. Chem. 2004, 69, 6042.

596

11. CARBON-CARBON BOND-FORMING REACTIONS

Li

Br

Li H3O+

t-BuLi

N

(–)-Sparteine

N

N

N

Bn

–78 or –90°C

Bn

Bn

Bn

N-Allyl-N-benzyl2-bromoaniline

136

(R)-1-Benzyl-3methylindoline

137

11.6.9 Basicity: Metal-Hydrogen Exchange Organolithium reagents are stronger bases than Grignard reagents, so removal of the acidic α-hydrogen to form an enolate anion can be a problem in reactions with carbonyl derivatives. However, addition to the carbonyl of ketones and aldehydes is usually the faster process with common organolithium reagents. Specialized organolithium reagents can give different results. Lithium triphenylmethide, for example, often yield selective deprotonation with aldehydes and ketones. The extent of enolate formation with organolithium is dependent on many factors.311 Because organolithium reagents are powerful bases,312 the metal-hydrogen exchange reaction with functionalized substrates plays an important role in many syntheses.313 This exchange reaction is believed to proceed via a four-center transition state (e.g., 138), generating a new organolithium (RdLi) from the acid (RdH + R1dLi).216a,314 For this reaction to occur, RdH must be a stronger acid than R1dH, and R1dLi must be a stronger base than RdLi. The ratedetermining step315 in cleavage of the CdH bond is initial coordination of the Li atom with the atom bearing the most acidic hydrogen.315,316 In general, the more acidic the proton being removed in RdH, the more facile the reaction with R1dLi. Six factors influence the acidity of a CdH bond.317 R RJH

H

R

H

Li R1

Li

R1

+ R1JLi

RJ Li

+ R1JH

138

1. The first factor is the percentage of s-character.317a As the s-character of a bond increases, the attached proton becomes more acidic, because the 2s-orbital lies closer to the nucleus than the 2p-orbitals. When the proton is removed, increased electron density in the 2s-orbital gives greater stability to the carbanion (the conjugate base). This stability leads to enhanced acidity of the alkynyl CdH (pKa, 25) relative to the analogous alkenyl CdH (pKa, 36.5). Diminishing the CdCdC bond angle (i.e., cyclopropane) increases the % s-character, and a cyclopropane CdH is more acidic (pKa, 39) than a normal alkane hydrogen (pKa, 42). The percentage of s-character318 can be determined from 13C NMR nuclear spin-spin coupling constants. 2. Conjugative effects enhance the acidity of certain covalently bound hydrogen atoms.317a The presence of a substituent that can delocalize electron density toward a more electronegative atom makes that proton more acidic. In addition, the conjugate base (the carbanion) formed after its removal is more stable because the electrons can be

311

House, H. O.; Kramar, V. J. Org. Chem. 1963, 28, 3362.

312

Roberts, J. D.; Curtin, D. Y. J. Am. Chem. Soc. 1946, 68, 1658.

313

(a) Reference 217, pp 698–705; (b) Reference 216.

(a) Ingold, C. K. Helv. Chim. Acta 1964, 47, 1191; (b) Batalov, A. P.; Rostokin, G. A.; Korshunov, I. A. J. Gen. Chem. USSR 1965, 35, 2146 (Engl., p 2135).

314

(a) Shatenshtein, A. I.; Kamrad, A. G.; Shapiro, I. O.; Ranneva, Yu.I.; Zvyagintseva, E. N. Dokl. Akad. Nauk, SSSR 1966, 168, 364; (b) Barnes, R. A.; Nehmsmann, L. J. J. Org. Chem. 1962, 27, 1939; (c) Benkeser, R. A.; Trevillyan, A. E.; Hooz, J. J. Am. Chem. Soc. 1962, 84, 4971; (d) Bryce-Smith, D. J. Chem. Soc. 1963, 5983.

315

316

(a) Morton, A. A. J. Am. Chem. Soc. 1947, 69, 969; (b) Gilman, H.; Morton, Jr., J. W. Org. React. 1954, 8, 258.

317

(a) Reference 42a, pp 48–85; (b) Bent, H. A. Chem. Rev. 1961, 61, 275; (c) Walsh, A. D., Trans. Faraday Soc. 1949, 45, 179.

318

Closs, G. L.; Closs, L. E. J. Am. Chem. Soc. 1963, 85, 2022.

597

11.6 ORGANOLITHIUM REAGENTS (CdLI)

delocalized over several atoms. Carbonyl groups, for example provide delocalization in the resonance-stabilized enolate anion 139. The alkene bond linkage in enolate anion 139 is usually planar, but steric interactions, reduction of bond angles in small rings, and solvent interactions can inhibit formation of the planar enolate. The stabilizing effects of conjugation are diminished in an enolate anion as a result of these factors319 (Section 13.2). The proton Ha in dione 140, for example, is much less acidic when compared to Ha in dione 141.319 This effect has been attributed to poor overlap of the carbanion orbital with the π-orbitals of the carbonyl in 140, due to the steric constraints of the bicyclo[2.2.2]octane ring system. The diminished acidity is therefore not entirely due to the difficulty that arises when forming a planar enolate anion. Aryl and alkenyl groups also provide conjugative stabilization, as in the benzyl anion (142). Conjugating groups (e.g., nitro and cyano) are excellent stabilizing substituents, as shown in 143 and 144, respectively. O R

C

O C

R

CH2

O

CH2

O

Ha

139

H

O

140

H

H

C R

C

O R

N

H2 C

Ha

O

141 O

H2 C

O

N

H2 C

O

142

C

N

143

H2 C

C

N

144

3. Inductive effects317a involve both through-bond transmission of electrons, and field (through-space) effects (Section 2.2.1). As the electronegativity of a substituent increases, the central atom diverts increasing amounts of s-character to the orbital occupied by the unshared electron pair.317a Methoxyacetic acid (pKa, 3.57) is more acidic, for example, than acetic acid (pKa, 4.75).320 Most carbanion stabilizing groups act through a mixture of inductive and conjugative effects that are difficult to identify as separate phenomena. The fluoride anion and quaternary ammonium cations are the only common substituents that stabilize ions by pure inductive effects. 4. Homoconjugative effects317a occur when the α-carbon (the position adjacent to a group that induces acidity) is blocked, but the β-carbon bears a hydrogen atom. Removal of the β-hydrogen atom from certain ketones (called β-enolization) can yield a stabilized carbanion, when its p-orbital is held rigidly at the proper angle and distance to the p-orbital of the carbonyl. The effect321 is usually limited to a compound where an acidic proton is disposed β to the carbonyl, and the α-position does not possess hydrogen atoms. A homoenolate (first reported by Nickon et al.322) is believed to be resonance stabilized by species (e.g., 145A and 145B) derived from treatment of 3,3-dimethylbicyclo[2.2.1]heptan-2-one with base.323 The acidity of Ha in 3,3-dimethylbicyclo[2.2.1]heptan-2-one is thereby enhanced.324 Me Me H Me Ha

Me

Base – Ha

O

3,3-Dimethylbicyclo[2.2.1]heptan-2-one

Me Me O

O H

H 145A

145B

319

Bartlett, P. D.; Woods, G. F. J. Am. Chem. Soc. 1940, 62, 2933.

320

King, E. J. J. Am. Chem. Soc. 1960, 82, 3575.

321

Johnson, A. L.; Petersen, W. W.; Rampersad, M. B.; Stothers, J. B. Can. J. Chem. 1974, 52, 4143.

322

Nickon, A.; Kwasnik, H.; Swartz, T.; Williams, R. O.; DiGiorgio, J. B. J. Am. Chem. Soc. 1965, 87, 1615.

323

Werstiuk, N. H. Tetrahedron 1983, 39, 205.

324

(a) Nickon, A.; Lambert, J. L. J. Am. Chem. Soc. 1962, 84, 4604; (b) Idem Ibid. 1966, 88, 1905.

598

11. CARBON-CARBON BOND-FORMING REACTIONS

5. Aromatic character317a is an important consideration when comparing the conjugate bases of several weak acids. If one of the conjugate bases is a resonance-stabilized anion, the acid precursor will be more acidic when compared to an acid that generates a nonaromatic conjugate base. The pKa of cyclopentadiene is 15, due in part to removal of a proton that leads to the stable aromatic cyclopentadienyl anion (cyclopenta-2,4-dien-1-ide). This acidity sharply contrasts with cycloheptatriene (pKa of 36), where loss of the hydrogen would yield the antiaromatic 8 π-electron cycloheptatrienyl anion (cyclohepta-2,4,6-trien-1-ide).

Cyclopenta-2,4-dien-1-ide

Cyclohepta-2,4,6-trien-1-ide

6. d-Orbital effects317a are important only when atoms are present that are not in the second row of the periodic chart (e.g., S, P, or transition metals). Carbanions can be stabilized by overlap of the electron pair with 3d-orbitals of thirdrow elements (e.g., P and S).325 The dithioacetal, 1,1-diethylthophenylmethane, is more acidic than acetal 1,1diethoxyphenylmethane by 5–6 pKa units. The carbanionic p-orbital aligns itself to minimize electrostatic repulsion, as in 146.326 Forcing these orbitals into close proximity by confining them to a polycyclic system significantly decreases the acidity. Tris(sulfone) 147 is less acidic than tris(ethylsulfonyl)methane, but in this case solvation effects contribute to the difference in acidity.325,326 SEt Ph

OEt SEt

H 1,1-Diethylthophenylmethane

Ph

O D

O

S

OEt H

O2S 146

SO2Et H

C

A

1,1-Diethoxyphenylmethane

O2 S

E

147

SO2Et

EtO2S

SO2

H Tris(ethylsulfonyl)methane

In summary, all six factors described above contribute to the following order of acidity, based on the ability of that functional group to enhance the acidity of the following: NO2 > CO > SO2 > CO2H > CO2R

~ CN ~ C(=O)NH2 > X

> H >

R

The pKa values of several common functional groups are presented in Table 11.3 to show the effect on acidity of carbonyl substituents, in the context of other organic acids. The acidic proton from each molecule (underlined H) is removed to give the new organolithium reagent. In general, an organolithium reagent will exchange with an acidic hydrogen of another molecule if the conjugate acid of the newly formed organolithium reagent is a weaker acid than the original organolithium reagent. According to this rationale, methyllithium (pKa of the conjugate acid, CH4, is 40) will remove the acidic hydrogen in phenylacetylene (pKa, 21) (see Table 11.3). Metal-hydrogen exchange therefore generates a carbanion that can subsequently be used as a nucleophile in addition reactions, or for substitution reactions. Conversion of carbonyl derivatives to the corresponding enolate anion is usually accomplished with lithium dialkylamide bases rather than alkyllithium reagents, as discussed in Section 13.2.2. The acidities of several ketones have also been measured in DMSO.327 Metal-hydrogen exchange is useful for the preparation of allylic and alkyl338b, aryl,328a or heteroaryl lithio derivatives. Asymmetric deprotonation is possible using sec-butyllithium and ()-sparteine. The reaction of Boc pyrrolidine derivative 148, for example, gave 149 and subsequent reaction with the methylating agent DMS gave a 45% yield of 150 in a ratio of 93:7 syn-anti diastereomers.329 It is also possible to generate allyllithium derivatives from allylic hydrocarbons.330

325

von E. Doering, W.; Hoffmann, A. K. J. Am. Chem. Soc. 1955, 77, 521.

326

von E. Doering, W.; Levy, L. K. J. Am. Chem. Soc. 1955, 77, 509.

327

Bordwell, F. G.; Harrelson, Jr., J. A. Can. J. Chem. 1990, 68, 1714.

328

(a) Letsinger, R. L.; Schnizer, A. W. J. Org. Chem. 1951, 16, 869; (b) Breslow, R.; Grant, J. L. J. Am. Chem. Soc. 1977, 99, 7745.

329

Wu, S.; Lee, S.; Beak, P. J. Am. Chem. Soc. 1996, 118, 716.

330

For a synthetic example taken from a synthesis of β-bisabolene , see Crawford, R. J.; Erman, W. F.; Broaddus, C. D. J. Am. Chem. Soc. 1972, 94, 4298.

TABLE 11.3 ACID

The pKa Values of Organic Acids ACID

pKa

H

NO2

H

NO2

3.6a

H

CN

H

CN

O Ph-CO2H O

4.2a

O 4.7e SO2Me

H

MeCO2

H

4.8e

H

SO2Me

H H

O CHO

5.9a

H

SO2Me

5.85a

NO2 H H

H

8.6a

H3C

H

O

O Ph H

9.4a

H

9.8a

Me3

S

Ph

S

H

S

H

S H

H

H

H

13.3 a

21e

Ph3

14 a

Me H

H

Et H H

S

21.6b

H H H H

15 a

Ph2

H

23e

15.7 e

Et2

H

23e

15.3 e

MeSO2

O 9.9a,i

Ph H O

O2NCH2 H

10.2 a

O 10.5 k

O

O 11 a

H Me

Ph

34d

Ph

23a

18 d

HO2

2

H

H

18.6 e,b

Me H Me

19.5 b

H3C

H

20a

2

Ph H

O NH2

H

C

H H

C H

24h

S

Me

S

H

S

Et H

S

H H

25.6 j

25a

35e [41 g

H

25e

H

35e

Ph

23.5b,c

EtO2

H

Ph H

H

19.1 b

H H

H H Me

Ph

H

16.5 a

H

32.5e

H

H

t-Bu

31.3e, 36.5 f

H

S

21.3b

O

Cl

29.6d

O

H

H

O

H

pKa

O

H O

O

CO2Et

H

O

20.8b,c

O

Et

ACID

21e

14-15 e

H

H

H

12.7 a

H

EtO2C

pKa

H

CO2Et

5.0a

CHO

H

O

CO2Et

H

ACID

12 a

CO2Et H Et

CHO

H

pKa

H3

H

36e

36.5e [44] g

37e [44] g 38.3d

38.6d

30e

40e

H

H H

H2

27e

H

45e [51] g

a

Reference.331; bReference.332; cReference.333; dReference.334; eReference.335; fReference.336; gReference.337; hReference.276; iReference.338

331

Pearson, R. G.; Dillon, R. L. J. Am. Chem. Soc. 1953, 75, 2439.

332

Zook, H. D.; Kelly, W. L.; Posey, I. Y. J. Org. Chem. 1968, 33, 3477.

333

House, H. O. Modern Synthetic Reactions, 2nd ed.; W.A. Benjamin: Menlo Park, CA., 1972, p 494.

334

Streitweiser, Jr., A.; Ewing, S. P. J. Am. Chem. Soc. 1975, 97, 190.

335

Reference 42a, pp 4, 10, 13, 14, 43, 48.

336

For acidity of 1,3-dithiane in DMSO, see Xie, L.; Bors, D. A.; Streitwieser, A. J. Org. Chem. 1992, 57, 4986.

(a) Streitwieser, Jr., A.; Caldwell, R. A.; Granger, M. R. J. Am. Chem. Soc. 1964, 86, 3578; (b) Streitwieser, Jr., A.; Maskornick, M. J.; Ziegler, G. R. Tetrahedron Lett. 1971, 3927.

337

338

Liptak, M. D.; Gross, K. C.; Seybold, P. G.; Feldgus, S.; Shields, G. C. J. Am. Chem. Soc. 2002, 124, 6421.

600

11. CARBON-CARBON BOND-FORMING REACTIONS

Ph

s-BuLi , (–)-Sparteine

N

Ph

Eher, –78°C

N

Boc

(MeO) 2SO2

Li

Ph

Boc

148

Me

N Boc

149

150 (45%)

The disconnections for the aryl substitution reactions as follows: ArJCO2H Ar JCH2R

ArJH

Ar JH + R2CKO

ArCR2OH

Ar JH + RJCH2X

RCHKC(R)CH 2R

RCHKC(R)CH3

11.7 CONCLUSION This chapter has shown how incorporation of Mg or Li gives carbanion-like reagents that can be used to form new carbon-carbon bonds. The great variety of nucleophilic species, the range of electrophilic species with which they react, and the high degree of stereoselectivity that accompanies many of the reactions show why these are among the most powerful disconnections in organic synthesis. The useful carbonyl-stabilized carbanions, the enolate anions will be discussed in Chapter 13. HOMEWORK

1. Predict the major product and explain the stereochemistry of the following reaction: CO2Et t-BuMe2SiO

N Boc

BF3•OEt2 , MeMgBr CuBr•Me2SH

OMe

2. The Grignard reaction of compound A in the text gave alcohol B. Offer an explanation for this selectivity. Me

Me

O Me Cl

O

O Me

MeMgCl , THF –78

Me

0°C

HO

H OAc

Cl H OAc B

A

3. Given the reaction sequence, give a mechanistic rationale for formation of the cyclohexanone product. HO 1. KH, THF/HMPA, 2. aq AcOH

–11

+25°C

PhS

O

SPh

4. In each case, draw the transition state for the Cram, Karabatsos, and Felkin-Anh models of the reaction of each molecule with (1) MeMgBr; (2) PhMgCl; (3) CH3C^C:Na+; (4) EtLi. O Ph Me

(A)

O

O

H Me

(B)

MeO Me

Me Et

Me

(C)

Me

Me O

Me

(D)

Draw the major product(s) that result from each reaction, after hydrolysis.

Me

(E)

O

601

11.7 CONCLUSION

5. The following reaction gives a diastereomeric ratio of 3:1: Draw both of these products, and predict the major product. O OHC

BuLi

Et

O

Et

6. Give a mechanistic explanation for the formation of the product shown from the designated starting material. O

O Ph CO2Et

O

1. LiN(SiMe3)2

S O

S

Ph

2. AcOH

CO2Et

7. Give the major product for (a-e) when treated with (1) BuLi/THF/78°C; (2) MeI/78 ! 0°C; (3) aq NH4Cl. Me

Me N

O

(A)

(B)

H

O

(C)

8. For each of the following give a complete reaction that illustrates its use for the formation of a carbon-carbon bond: (a) CO; (b) N,N,N0 ,N0 -tetramethylethylenediamine; (c) FeCl3; (d) NaCN; (e) NaNH2; (f) NaH. 9. Rationalize formation of the indicated product in this reaction. O OMe N Me

Ph

O

1. CH2KCHMgBr , THF

Ph

2. H2O

N

OMe Me

10. Each of the following molecules can be used to standardize a solution of an organolithium reagent. Describe, with reactions, the acid-base chemistry involved in each reaction. MeO

OH

H N N Ph

2 EtLi

OMe

(A)

(B)

2 EtLi

OH

11. Provide a suitable synthetic sequence for each transformation (a-d). Ph

c

a

Ph

CHO

CO2Me

OH

d

CO2H

b

O

12. Give the major product of this reaction, with correct stereochemistry, and justify your choice. Me mcpba

A HO

Me

1. BuMgBr 2. H3O+

B

602

11. CARBON-CARBON BOND-FORMING REACTIONS

13. In each of the following reactions, predict the major product, with the correct stereochemistry where appropriate: N

O

(A)

LiNTTMS)2 , –78°C Toluene

H

t-BuO2C

O

SiMe3

1. H2O2

(B)

2. aq KOH

O O

O

MeO

1. t-BuLi, Ether –78 °C

(C) Br

N

O

2. CO2 , –78 3. H3O+

0°C

I

(D)

Me N

O

O

N

2.2 t-BuLi THF, –78°C

N

Boc

(E)

SePh

MeO

Me

TBDPSO O Me

O

OMe

H

N

Me

MgBr

OMe

THF, –78°C

(F)

O

O Me

O

(I)

1. BuMgBr

O

(G)

CN

Me H OMe

OMe

2. BCl3 , Bu4NI CH2Cl2 , –78°C

rt

O MgBr O

H

(J)

O

1. TPAP, NMO

HO

THF, 0°C

OTIPS

OAc

H

1. 2 sec-BuLi, TMEDA Allyl bromide, THF

CO2H

(H)

2. NaBH4

CHO

O H O

1. MeLi 2. SOCl 2 , Py

2.

OTIPS

MgBr

SO2Ph 1. BuLi, THF-HMPA 1. 2 BuLi, THF

(K)

N

N

Br

Br

(L)

2.

2. Excess CO2 3. Hydrolysis

Br

BnO2C

H OBn

1. LiAlH4 2. MeSO2Cl

(M)

N CO2Et

O

(N)

3. NaCN 4. aq NaOH/MeOH

O

MeO

MgBr

Cl

OMe MeO

1. NaH 2. n-BuLi, THF, –78°C

O

(O)

CO2Me

3. MOMOCH2Br

(Q)

N

(P) OMe O2 N S

O 1. Excess MeLi

O

(R)

2. H3O+

MeO

BuLi, DMF

(S)

THF –78°C

Br

S OTBDMS C3H7

OMe

CuI

Ph

O

O

MeMgBr, THF CuBr•SMe2 –40°C

O

1. BuLi, TMEDA-THF 2. Succinic anhdyride , THF

603

11.7 CONCLUSION

14. Provide a synthesis for each of the following transformations. Show all reagents and intermediate products.

(A)

C8 H17

(CH 2)7COOH

C18 H35

C8 H17

HO

(B) OPMP

O

O

OPiv

(C)

(D)

OH

OH

MeO2C OH O

HO

(E)

O H O

O O H O

H H

OAc

OH H H

O

SOPh O

C H A P T E R

12 Carbon-Carbon Bond-Forming Reactions: Stabilized Carbanions, Organocuprates, and Ylids 12.1 INTRODUCTION Reactions that make CdC bonds using Grignard reagents or organolithium reagents and related compounds were discussed in Chapter 11. Aliphatic substitution reactions were also discussed, using a cyanide ion as a reagent, as well as alkyne anions. Chapter 11 concluded with the so-called metal-hydrogen exchange reaction, which is simply an acid-base reaction. Organosulfur compounds and organophosphorus compounds can be converted to α-lithio derivatives using this reaction, and these new reagents react as carbanions. In addition, organocopper reagents, specifically organocuprates, react with alkyl halides, epoxides, acid chlorides, as well as aldehydes and ketones. Finally, phosphorus and sulfur carbanions known as ylids will be used to generate alkenes or epoxides, respectively. This chapter will therefore focus on acid-base reactions that generate carbanions or carbanion-like reagents, which in all cases exhibit nucleophilic reactivity.

12.2 SULFUR-STABILIZED CARBANIONS AND UMPOLUNG 12.2.1 Sulfur-Stabilized Organolithium Reagents Introducing a heteroatom (X) into an organic molecule can enhance the acidity of the proton adjacent to the heteroatom (HdCdX). A common example is a group that contains a sulfur atom, which makes the proton more acidic due to electron-withdrawing effects, as well as by stabilization of the resulting carbanion by the d-orbitals (see Sections 11.6.1 and 11.6.2).1,2 Several different sulfur-stabilized carbanions are known, and have been used extensively in organic synthesis.3,4 Sulfur-stabilized carbanions react with alkyl halides or with aldehydes and ketones in the same manner as other carbanions. Thioethers (sulfides)5 can be converted to their α-carbanion derivative (1, M ¼ Li, Na, K) by treatment with strong bases [e.g., n-butyllithium, sodium amide (NaNH2), or potassium amide (KNH2)]. It is fair to say that the most prevalent derivatives are the organolithium reagents. The hydrogen atom α to the S atom of sulfides is a weaker acid than the H atom α to a carbonyl, nitrile, or nitro group, and stronger bases are required for deprotonation. Sulfides are stronger acids than the allylic, vinyl, or aromatic hydrocarbons discussed above, however. The various oxides of sulfur, including sulfoxides (2), sulfones (3), or sulfonate esters (4) can be converted to α-carbanions. Increasing the number 1

von E. Doering, W.; Hoffmann, A. K. J. Am. Chem. Soc. 1955, 77, 521.

2

von E. Doering, W.; Levy, L. K. J. Am. Chem. Soc. 1955, 77, 509.

3

Stowell, J. C. Carbanions in Organic Synthesis, Wiley–Interscience: New York, 1979; pp 94–103.

4

Block, E. Aldrichimica Acta. 1978, 11, 51.

(a) Gilman, H.; Beaber, N. J. J. Am. Chem. Soc. 1925, 47, 1449. (b) Ipatieff, N. N.; Pines, H.; Friedman, B. S. Ibid. 1938, 60, 2731. (c) Hurd, C. D.; Gershbein, L. L. Ibid. 1947, 69, 2328. (d) Campbell, J. R. J. Org. Chem. 1964, 29, 1830. (e) Kharasch, M. S.; Fuchs, C. F. Ibid. 1948, 13, 97.

5

Organic Synthesis http://dx.doi.org/10.1016/B978-0-12-800720-4.00012-X

605

Copyright © 2017 Michael Smith. Published by Elsevier Inc. All rights reserved.

606

12. CARBON-CARBON BOND-FORMING REACTIONS

of oxygen atoms on the S atom generally increases the acidity of the α-proton. Bordwell and coworkers6 studied the acidity of several sulfur derivatives using DMSO as a solvent.5b Anders and coworkers7 studied the structure of sulfurstabilized allyllithium compounds in solution, and found that they were monomers in THF. O

M R R1

RS 1

O

S

M

R

O

R1

O

S

M

RO

O

R1

3

2

S

M R1

4

Oxidation of a sulfoxide to a sulfone places a second oxygen on sulfur (RSO2R, Section 6.9.1), which increases the relative acidity of the α-hydrogen. Dimethyl sulfone, for example, is more acidic than the corresponding sulfoxide (RSOR, DMSO) by 6.1 pKa units. The presence of an electron-releasing methyl group (PhSO2CH3 vs. PhSO2CH2CH3)8 causes a reduction of acidity by 2 pKa units. As the group attached to the sulfur becomes more electron withdrawing, the acidity increases (Me3CSO2CH3 ¼ 30.3, MeSO2CH3 ¼ 31.1, PhSO2CH3 ¼ 29.0, CF3SO2CH3 ¼ 18.8).6c,g A similar trend is observed for the thio-analogs of acetonitrile (Me3CSCH2CN ¼ 22.9, Me2CHSCH2CN ¼ 23.6, CH3CH2SCH2CN ¼ 24.0, CH3SCH2CN ¼ 24.3, PhSCH2CN ¼ 20.9).6g Two sulfur groups significantly increase the acidity of the proton (CH3SO2CH3 ¼ 31.1, CH3CH2SO2CH2SO2CH2CH3 ¼ 14.4, CH3SCH2SOCH3 ¼ 29.0),6g and another electron-withdrawing group leads to an increase in acidity (CH3SO2CH2CN ¼ 13.6, CH3SCH2CN ¼ 24.3).6g As mentioned, the proton on the α-carbon of a sulfide is a very weak acid and its removal requires a strong base.5 Organolithium reagents are commonly used for this purpose. Isopropylphenyl sulfide reacted with tert-butyllithium (THF, HMPA) to yield the α-lithio derivative 5. Alkyl sulfides often require addition of HMPA or TMEDA to facilitate LidH exchange. Organolithium reagent 5 reacted with allyl bromide to give 5-methyl-4-phenylthiohex-1-ene in 87% yield, or with benzaldehyde to give 3-methyl-1-phenyl-2-(phenylthio)butan-1-ol in 92% yield.9 Li t-BuLi

PhS

Br

PhS

PhS

THF–HMPA –78°C

Isobutylphenyl sulfide

5

5-Methyl-4-phenylthiohex-1-ene (87%) 1. PhCHO 2. H2O

HO

Ph

PhS

3-Methyl-1-phenyl-2-(phenylthio)butan-1-ol (92%)

There are many synthetic applications of sulfide carbanions. In an example taken from a synthesis of antroquinonol D by Chen and coworkers,10 treatment of sulfide 6 with butyllithium gave the α-lithiosulfide, which reacted with allylic bromide 7 to yield the coupling product. Subsequent hydrogenolysis of the sulfide moiety under dissolving metal conditions (see Section 7.11.4) gave 8 in 84% yield. The ability to remove the sulfur after activating a carbon for nucleophilic attack is one of the most attractive features of organosulfur chemistry. Coupling of allylic halides and allylic thiocarbanions has been called Biellmann alkylation.11 Intramolecular reactions occur readily and reaction with epoxides or alkyl halides is possible.

(a) Reference 3, p 9. (b) Bordwell, F. G.; Algrim, D. J. Org. Chem. 1976, 41, 2507. (c) Bordwell, F. G.; Bares, J. J.; Bartmess, J. E.; McCollum, G. J.; van der Puy, M.; Vanier, N. R.; Matthews, W. S. Ibid. 1977, 42, 321. (d) Bordwell, F. G.; Bares, J. E.; Bartmess, J. E.; Drucker, G. E.; Gerhold, J.; McCollum, G. J.; van der Puy, M.; Vanier, N. R.; Matthews, W. S. Ibid. 1977, 42, 326. (e) Bordwell, F. G.; Matthews, W. S.; Vanier, N. R. J. Am. Chem. Soc. 1975, 97, 442. (f) Bordwell, F. G.; Bartmess, J. E.; Drucker, G. E.; Margolin, Z.; Matthews, W. Ibid. 1975, 97, 3226.

6

7

Piffl, M.; Weston, J.; G€ unther, W.; Anders, E. J. Org. Chem. 2000, 65, 5942.

8

Benkeser, R. A.; Trevillyan, A. E.; Hooz, J. J. Am. Chem. Soc. 1962, 84, 4971.

9

(a) Dolak, T. M.; Bryson, T. A. Tetrahedron Lett. 1977, 1961. (b) Kondo, K.; Matsumoto, M. Ibid. 1976, 391.

10

(a) Sulake, R. S.; Jiang, Y.-F.; Lin, H,-H.; Chen, C. J. Org. Chem. 2014, 79, 10820. (b) Also see Sulake, R. S.; Chinpiao Chen, C. Org. Lett. 2015, 17, 1138.

(a) Biellmann, J. F.; Ducep, J. B. Tetrahedron 1971, 27, 5861. (b) Altman, L. J.; Ash, L.; Marson, S. Synthesis 1974, 129. (c) Grieco, P. A.; Masaki, Y. J. Org. Chem. 1974, 39, 2135.

11

607

12.2 SULFUR-STABILIZED CARBANIONS AND UMPOLUNG

OTBS MeO 1. BuLi, THF, –78°C

SPh

2.

OTBS 3. Li, NH3(liq), –60°C

MeO

MeO OMOM

MeO

6

Br

8 (84%)

OMOM

7

When a sulfide group is attached to the α-carbon of a molecule that also contains an electron-withdrawing group (e.g., a carbonyl, the sulfide is β to the carbonyl), the acidity of the α-proton is greatly enhanced (see Table 11.3). An example is found in the Dias and Ferreira12 synthesis of ()-goniotrinin, α-phenylthio lactone 9 was treated with lithium hexamethyldisilazide (see Section 13.2.2), the 2-lithio derivative was formed. Subsequent reaction with trifluoromethylsulfonate (10) gave a 62% yield of 11. In this synthesis, the sulfide was subsequently oxidized to the corresponding sulfoxide and heating induced syn-elimination to yield a conjugated lactone moiety (Section 3.7.4).12 The sulfide-alkylation sequence can be applied to many systems, and in some cases highly functionalized molecules can be prepared. Me

Me 3

O

1. LHMDS, THF, 10 HMPA –78 – 0°C TBDPSO

2.

O

PhS

O

TBDPSO OTf

O OTBDPS

9

0°C – rt, 2 h

10

O

SPh

OTBDPS

O

11 (62%)

Sulfoxide carbanions formed by deprotonation of sulfoxides are used in alkylation reactions.13 Chiral sulfoxides (Section 6.9.1.2) are possible, and reaction of the corresponding organolithium derivatives with aldehydes or ketones generates diastereomeric products. Once the organolithium derivative is generated, the stereochemical integrity of the CdLi bond is remarkably stable, possibly due to interactions with the adjacent functional groups. Conversion of the asymmetric keto-sulfoxide 12 to the carbanion, followed by condensation with propanal gave an alcohol-sulfoxide. The sulfoxide group was removed by treatment with aluminum amalgam to complete the asymmetric synthesis of (R)-4-hydroxyhexan-2-one in 67 % yield and 64 %ee.14 Note that tert-butylmagnesium bromide was used as the base. The reductive cleavage of the sulfoxide with aluminum amalgam (Section 7.11.8) is a common method for removing a sulfur moiety. O

O S

1. t-BuMgBr, THF⋅OEt2 –78°C, EtCHO

OH

O

2. Al/Hg, THF

••

12

(R)-4-Hydroxyhexan-2-one (67%)

Another variation in sulfoxide reactivity is shown after an initial conversion of 13 to 14 via reaction with LDA and alkylation with methyl iodide.15b In this case, removal of the sulfur moiety allowed formation of an allylic alcohol. Heating allylic sulfoxide 14 induced a facile [2,3]-sigmatropic rearrangement (Section 15.5.3) to yield 15, and hydrolysis of 15 gave a cyclohexenol derivative, (S)-3-methylcyclohex-2-en-1-ol. Trimethylphosphite [P(OMe)3] was added to trap a minor isomer of the equilibrium. This reaction is called the Mislow-Evans rearrangement.15,16

12

Dias, L. C.; Ferreira, M. A. B. J. Org. Chem. 2012, 77, 4046.

13

Reference 3, pp 94–97.

14

Schnedier, F.; Simon, R. Synthesis 1986, 582.

(a) Bickart, P.; Carson, F. W.; Jacobus, J.; Miller, E. G.; Mislow, K. J. Am. Chem. Soc. 1968, 90, 4869. (b) Evans, D. A.; Andrews, G. C. Acc. Chem. Res. 1974, 7, 147. (c) Hoffmann, R. W. Angew. Chemie. Int. Ed. Engl. 1979, 18, 563.

15

16

(a) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-62. (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; pp 430–431.

608

12. CARBON-CARBON BOND-FORMING REACTIONS

The [2,3]-sigmatropic rearrangement that occurs with allylic sulfoxides is useful and predictable, since it proceeds with high stereoselectivity.15c This rearrangement is useful for introducing functional groups other than allylic alcohols into a molecule, as in Otera and coworker’s17 use of this reaction to generate conjugated aldehydes. Selenoxides also give this reaction.18 O Ph S O

Ph

S

1. LDA 2. MeI

P(OMe)3 MeOH

13

Me P(OMe)3 PhS

O

HO

Me

14

Me

MeOH

(S)-3-Methylcyclohex-2-en-1-ol

15

Sulfones are useful organosulfur compounds.19 α-Lithio sulfones react as nucleophiles with alkyl halides to yield the corresponding alkylation product,20 and with an epoxide to yield the corresponding alcohol. An example of the latter transformation is taken from the Nagumo and coworker’s21 synthesis of sekothrixide, in which sulfone 16 was treated with lithium hexamethyldisilazide to generate the α-lithio sulfone, which reacted with 1,2-butene epoxide (2-ethyloxirane) to give a good yield of alcohol 17.

MOMO

O

Si

t-Bu

t-Bu

t-Bu

t-Bu

O

1. LHMDS, THF 2.

PhO2S

, THF

PhO2S MOMO

O

Si

O

HO

O

Me

Me

Me 16

Me

17

After sulfones are used in reactions, the sulfur moiety is usually removed from the molecule via reduction. Sodium amalgam or Ni(R) are common reagents that are used to remove sulfur from organic molecules (Sections 7.11.4 and 7.10.5).22 Reaction of a sulfone carbanion with an aldehyde, acetylation of the resulting alcohol, and subsequent elimination of the alcohol-sulfone moiety to form an alkene is called Julia olefination (sometimes the Julia-Lythgoe olefination), and there is a modified procedure known as Julia-Kocienski olefination.23 In a synthesis of (+)-SCH 351448, Panek and Zhu24 reacted aldehyde 18 with a tetrazole-sulfone25 and sodium hexamethyldisilazide to generate the vinyl group in 19, which was formed in >80% yield. Ph N

O

O

N

N SO2CH3,

NaHMDS

O

O

THF, –78°C

O

O

N

CHO

18

O O 19 (>80%)

In 1-lithio vinyl sulfides (Section 12.2) the S atom stabilizes the carbanionic center, allowing either alkylation or condensation reactions. Lithio vinylsulfones are similarly stabilized. 2-Phenyl-1-phenylsufonylethene, for example, 17

Sato, T.; Otera, J.; Nozaki, H. J. Org. Chem. 1989, 54, 2779.

18

For an example taken from a synthesis of pseudocodeine, see Kshirsagar, T. A.; Moe, S. T.; Portoghese, P. S. J. Org. Chem. 1998, 63, 1704.

19

(a) Magnus, P. D. Tetrahedron 1977, 33, 2019. (b) Reference 344, pp 97–103.

20

See Miyaoka, H.; Yamanishi, M.; Kajiwara, Y.; Yamada, Y. J. Org. Chem. 2003, 68, 3476.

21

Terayama, N.; Yasui, E.; Mizukami, M.; Miyashita, M.; Nagumo, S. Org. Lett. 2014, 16, 2794.

22

Welch, S. C.; Gruber, J. M. J. Org. Chem. 1982, 47, 385.

(a) Julia, M.; Paris, J. M. Tetrahedron Lett. 1973, 14, 4833. (b) Kocienski, P. J.; Lythgoe, B.; Ruston, S. J. Chem. Soc. Perkin Trans. 1 1978, 829. (c) Kocienski, P. J. Phosphorus Sulfur 1985, 24, 97. (d) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-50. (e) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; pp 356, 357.

23

24

Zhu, K.; Panek, J. S. Org. Lett. 2011, 13, 4652.

25

Leburn, M.-E.; Le Marquand, P.; Berthelette, C. J. Org. Chem. 2006, 71, 2009.

609

12.2 SULFUR-STABILIZED CARBANIONS AND UMPOLUNG

was converted to the 1-lithio derivative by reaction with methyllithium.26 This organometallic reagent reacted in the usual manner with alkyl halides (methyl iodide) to give the coupling product, 1-phenyl-2-phenylsufonylprop-1-ene. SO2Ph

SO2Ph

1. MeLi, LiBr, THF, –95°C

Ph

Ph

2. MeI

Me

1-Phenyl-2-phenylsufonylprop-1-ene

2-Phenyl-1-phenylsufonylethene

Vinyl sulfoxides undergo an unusual elimination reaction when a leaving group (e.g., chloride) is also present on the double bond. When 20 was treated with tert-butyllithium, an exchange reaction occurred to yield 21.27 In the presence of additional tert-butyllithium, chloride is lost and a lithio-alkyne (22) was formed in what is known as the FritschButtenberg-Wiechell rearrangement.28 Hydrolysis led to the final product, alkyne 23, in 74% yield. A similar reaction was observed when 20 was treated with Grignard reagents. O Ph

Li

S t-BuLi

Cl

H

Cl

t-BuLi

H

H 2O

Li

THF

H

20

OMe

OMe

OMe

OMe 21

23 (74%)

22

Formation of a carbanion from a α-halo sulfone is often accompanied by loss of the SO2 moiety via the so-called Ramberg-B€ acklund reaction.29 An example is the conversion of 1-bromoethyl ethyl sulfone to but-2-ene.30 Initial reaction with KOH generated the sulfone carbanion (24), which displaced bromide intramolecularly to give an episulfone, 2,3dimethylthiirane 1,1-dioxide. Under the reaction conditions, 2,3-dimethylthiirane 1,1-dioxide decomposed with loss of sulfur dioxide (SO2) to yield but-2-ene. Meyers et al.31 introduced a modification of this reaction in which the sulfone was converted directly to the alkene without isolating the α-halosulfone, using KOH in CCl4.32 Sulfuryl chloride and bromine can also be used.32 O Me

O

S O

Me

KOH

Me

Br

Bromoethyl ethyl sulfone

O

−O

Br

24

26

Eisch, J. J.; Galle, J. E. J. Org. Chem. 1979, 44, 3277, 3279.

27

Satoh, T.; Hayashi, Y.; Yamakawa, K. Bull. Chem. Soc. Jpn. 1993, 66, 1866.

O – SO2

S

Me

S

Me

Me

2,3-Dimethylthiirane 1,1-dioxide

Me Me But-2-ene

(a) Fritsch, P. Annalen 1894, 279, 319. (b) Buttenberg, W. P. Annalen 1894, 279, 327. (c) Wiechell, H. Annalen 1894, 279, 332. (d) Stang, P. J. Chem. Rev. 1978, 78, 383. (e) Stang, P. J.; Fox, D. P. J. Org. Chem. 1978, 43, 364. (f) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-35. (g) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; pp 262, 263.

28

(a) Ramberg, L.; B€ acklund, B. Arkiv. Kemi Mineral. Geol. 1940, 13A, 50 (Chem. Abstr. 1940, 34, 47255). (b) Paquette, L. A. Acc. Chem. Res. 1968, 1, 209. (c) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-77. (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; pp 540, 541.

29

(a) Block, E. Reactions of Organosulfur Compounds, Academic Press: New York, 1978. (b) Block, E. J. Chem. Educ. 1971, 48, 814. (c) Paquette, L. A., Org. React. 1977, 25, 1. (d) Bordwell, F. G.; Jarvis, B. B.; Corfield, P. W. R. J. Am. Chem. Soc. 1968, 90, 5298. (e) Paquette, L. A.; Philips, J. C. Ibid. 1969, 91, 3973. (f) Paquette, L. A.; Houser, K. W. Ibid. 1969, 91, 3870. (g) Corey, E. J.; Block, E. J. Org. Chem. 1969, 34, 1233.

30

31

Meyers, C. Y.; Malte, A. M.; Matthews, W. S. J. Am. Chem. Soc. 1969, 91, 7510.

(a) Pommelet, J.-C.; Nyns, C.; Lahousse, F.; Moreny, R.; Viehe, H. G. Angew. Chemie Int. Ed. 1981, 20, 585. (b) Martel, H. J. J.-B.; Rasmussen, M. Tetrahedron Lett. 1975, 947.

32

610

12. CARBON-CARBON BOND-FORMING REACTIONS

Sulfonate carbanions have been used less extensively than sulfone, sulfoxide, or sulfide carbanions. Sulfonates are good leaving groups and subject to nucleophilic displacement or elimination in the presence of the bases required to generate the carbanion. Cyclic sulfonates (sultones) have been alkylated33 via initial reaction with n-butyllithium and then an alkyl halide. Sultone 25 was used in the Wolinsky et al. synthesis of the terpene, β-santalene, via alkylation with 1-bromo-3-methylbut-2-ene to yield 26.34 The sulfur was removed by reduction with aluminum hydride (AlH3, Section 7.11.8), and subsequent treatment with POCl3 and Py generated the exo-methylene group (Section 3.5).34 Note that reductive cleavage of the sultone can give the thioalcohol via SdO cleavage, and/or the alcohol via CdO cleavage.33a Sultones and sulfonate esters are rarely used in synthesis, but sulfides and sulfones are common.

Me

O Me

S

O

1. C4H9Li, THF 2.

O

Br

Me

25

O Me

S

1. LiAlH4/AlCl3

O

2. POCl3, Py

Me CH2

O

β-Santalene

26

The disconnections for these reactions found in this section are as follows: O

O R

R R1

R

X

R

OH R

R1

R

R

R1

R—SO2R1

12.2.2 Umpolung A certain kind of reactivity is typically associated with a given functional group. In some cases, it is possible to reverse the reactivity, converting an electrophilic center into a nucleophilic one and vice versa, for example. This section will explore reactions that facilitate this reversal. Under normal conditions, a carbonyl group is polarized as shown in 27, with the carbonyl carbon electrophilic and the α carbon nucleophilic. Facile acyl addition of Grignard reagents, as well as facile formation of an enolate anion (see Section 13.2), demonstrate this bond polarization. In a hypothetical molecule 28, the acyl carbon would be nucleophilic. To attain this reversal, however, the carbonyl must be transformed into another functional group that allows the carbon to become nucleophilic. The ability to regenerate the carbonyl after completion of the desired reactions is another requirement. Seebach and coauthors35 termed this process umpolung. There is umpolung (reversal of reactivity) whenever one observes a 1,2-n relationship of the functional groups in a carbonyl compound, where n is the number of carbon atoms separating the functional groups. For umpolung to occur, an even number of carbons should separate the carbons bearing the functional groups (0, 2, 4, …).35a This relationship contrasts with normal reactivity, where an odd number of carbons separate the carbonyl bearing functional groups (1, 3, 5, …).35a Table 12.135c shows several umpolung equivalents, along with the reactivity of the normal fragment and typical reagents that give the umpolung reactivity.

33

(a) Smith, M. B.; Wolinsky, J. J. Org. Chem. 1981, 46, 101. (b) Durst, T.; Tin, K. C. Can. J. Chem. 1970, 48, 845.

(a) Wolinsky, J.; Marhenke, R. L.; Eustace, E. J. J. Org. Chem. 1973, 38, 1428. (b) Coffen, D. L.; Grant, B. D.; Williams, D. L. Int. J. Sulfur Chem. Part A 1971, 1, 13.

34

(a) Seebach, D. Angew. Chem. Int. Ed. Engl. 1979, 18, 239. (b) Seebach, D. Ibid. 1969, 8, 639. (c) Gr€ obel, B.-T.; Seebach, D. Synthesis 1977, 357. (d) Seebach, D.; Kolb, M. Chem. Ind. (London) 1974, 687.

35

611

12.2 SULFUR-STABILIZED CARBANIONS AND UMPOLUNG

TABLE 12.1

Typical Umpolung Equivalents and Reagents Normal

Desired umpolung O

O R

–O C +

C

-

H

a1

R 29

+ O R

Umpolung reagents

R

d1

S

S Ts

d2

R

SR

S

R

OR SR S R O

O

O C –

R

SR

27 "

S

SR

a2

R

R

S

N C

O

O

O R

C

+

O

O H

“ R

a3

SR

d3

S

R

RO2S

SR

28

O

O

d4

O

OR

SR

O

a5

SR

SR

a4

R

R

R

SR

S

d5

R

R

Reprinted from Gr€ obel, B.-T.; Seebach, D. Synthesis 1977, p. 357. Copyright © 1977 Georg Thieme Verlag KG.

12.2.2.1 Acyl Anion Equivalents The most common umpolung equivalent is the so-called acyl anion, 29, and dithianes derived from aldehydes (30) are the most common reagent used for this purpose. Dithianes are a common ketone protecting group (Section 5.3.3.2), prepared from ketones, but no acidic hydrogen atom is available for reaction with a base. Note that a nonthiolic, odorless equivalent has been reported.36 Propane-1,3-dithiol is the usual precursor for the preparation of dithianes (e.g., 30), by reaction with an aldehyde. The acidity of this hydrogen is largely due to greater polarizability of the sulfur, and the greater CdS bond length, rather than to the presence of the d-orbitals.37 S

H

S

R

30

R1 -Li

S

Li

S

R

31

The pKa of 1,3-dithiane (30, R ¼ H) is 31.1.38 2-Methyl-1,3-dithiane (30, R ¼ Me) has a pKa of 38.3, however, whereas 2-phenyl-1,3-dithiane (30, R ¼ Ph) has a pKa of 29.6.39 The acidity is dependent on the electronic effect of the substituent. There are many dithioacetal derivatives (remember that dithioketals do not posses an acidic hydrogen), including both acyclic and cyclic species. For reasons of stability, availability and ease of preparation, dithianes are used most

36

Liu, Q.; Che, G.; Yu, H.; Liu, Y.; Zhang, J.; Zhang, Q.; Dong, D. J. Org. Chem. 2003, 68, 9148.

37

Bernardi, F.; Csizmadia, I. G.; Mangini, A.; Schlegel, H. B.; Whangbo, M. H.; Wolfe, S. J. Am. Chem. Soc. 1975, 97, 2209.

(a) Streitwieser, A., Jr.; Caldwell, R. A.; Granger, M. R. J. Am. Chem. Soc. 1964, 86, 3578. (b) Streitwieser, A., Jr.; Maskornick, M. J.; Ziegler, G. R. Tetrahedron Lett. 1971, 3927. 38

39

House, H. O. Modern Synthetic Reactions, 2nd ed.; W. A. Benjamin: Menlo Park, CA, 1972; p 494.

612

12. CARBON-CARBON BOND-FORMING REACTIONS

often. Removal of the hydrogen atom adjacent to the sulfur in 30 requires a strong base (e.g., n-butyllithium), and the product is α-lithiodithiane (31). A typical application is the reaction of dithiane (32) with butyllithium to form the anion and subsequent reaction with an allylic bromide gave 33 in 92% yield, taken from the synthesis of ()-walsucochin B by Xie and coworkers.40 To complete the reversal of reactivity, the carbonyl must be unmasked, which is usually done by reaction with a Lewis acid that has a high affinity for sulfur (e.g., mercuric oxide with boron trifluoride in aq THF).41 In this particular example, treatment of 33 with iodine and calcium carbonate in THF gave an 82% yield of the ketone (34).40 Many different oxidants have been employed to unmask dithianes, including Cl2, Br2, t-BuOCl, NBS, Pb(OAc)4, Ce(NH3)6(NO3)4, Tl(O2CCF2)3, and so on.42 The choice of reagent depends on the facility of the hydrolysis and the sensitivity of the molecule to further reaction. Dithianes are widely used in natural product synthesis.43

O

S

O S

S 1. BuLi, THF, –78°C

I2, CaCO3

S 2. Br

MeO

O

0°C, 2 h

O

–78°C to rt

MeO

32

MeO 33 (92%)

34 (82%)

The disconnection for the acyl anion umpolung follows: O RJCHO R

R1

+

R1JX

Lithiated dithianes react best with primary and secondary alkyl halides,44 acyl halides, ketones and aldehydes,45 as well as epoxides.46 Nitriles and amides react only if there is no acidic proton associated with that molecule. An important synthetic use of this reaction is the preparation of highly functionalized aldehydes as a starting material.47 This route is important for the preparation of β-hydroxy and β-alkoxy aldehydes, and many substituents can be incorporated. Dithiane anions can be used with electrophilic species such as epoxides. In the construction of an advanced intermediate of spongistatin, Smith et al.48 reacted 2-silyl dithiane (38) with tert-butyllithium to form the 2-lithio species. Subsequent reaction with epoxide 36, and then epoxide 37, led to initial coupling, followed by in situ loss of the silyl unit to allow coupling of the second epoxide giving a 69% yield of 38. Smith et al.49 termed this sequence a "linchpin assembly," and the dithiane unit was converted to the ketone later in the synthesis.

40

Xu, S.; Gu, J.; Li, H.; Ma, D.; Xie, X.; She, X. Org. Lett. 2014, 16, 1996.

41

(a) Seebach, D.; Corey, E. J. J. Org. Chem. 1975, 40, 231. (b) Seebach, D. Synthesis 1969, 17.

42

Romanet, R. F.; Schlessinger, R. H. J. Am. Chem. Soc. 1974, 96, 3701.

43

Yus, M.; Nájera, C.; Foubelo, F. Tetrahedron 2003, 59, 6147.

44

Corey, E. J.; Andersen, N. H.; Carlson, R. M.; Paust, J.; Vedejs, E.; Vlattas, I.; Winter, R. E. K. J. Am. Chem. Soc. 1968, 90, 3245.

(a) Corey, E. J.; Weigel, L. O.; Chamberlin, A. R.; Lipshutz, B. J. Am. Chem. Soc. 1980, 102, 1439. (b) Corey, E. J.; Weigel, L. O.; Floyd, D.; Bock, M. G. Ibid. 1978, 100, 2916. 45

46

Corey, E. J.; Bock, M. G. Tetrahedron Lett. 1975, 2643.

47

Vedejs, E.; Fuchs, P. L. J. Org. Chem. 1971, 36, 366.

48

Smith, III, A. B.; Doughty, V. A.; Sfouggatakis, C.; Bennett, C. S.; Koyanagi, J.; Takeuchi, M. Org. Lett. 2002, 4, 783.

49

Smith, III, A. B.; Pitram, S. M.; Boldi, A. M.; Gaunt, M. J.; Sfouggatakis, C.; Moser, W. H. J. Am. Chem. Soc. 2003, 125, 14435.

613

12.2 SULFUR-STABILIZED CARBANIONS AND UMPOLUNG OR BnO

S

36

2.

1. t-BuLi, Ether

O

OR

SiMe2t-Bu 3.

S

BnO O

37

S

2.

S

OH

OR

38 (69%) O

Cl

1. BuLi, –10°C

S

R = SiMe2t-Bu

OR

35

S

OR

39

O

O

3. BuLi, HMPA, –10°C 4. 39

S

O

S

O

O 40 (70%)

This two-directional approach was used with halides in Stockman and Sinclair’s50 synthesis of perhydrohistrionicotoxin. 1,3-Dithiane was treated with butyllithium and then with halide 39. This sequence was followed by a second round of butyllithium and then 39, giving 40 in an overall yield of 70%. Later in this synthesis, the dithiane moiety was converted to the ketone by treatment with NCS and silver nitrate in aqueous acetonitrile.50 Although conjugate addition is known, dithiane anions most often give 1,2-addition with conjugated systems. Cyclohexenone reacted with 30(R ¼ Me) to give the 1,2-addition product 41 in 99% yield.51 In this case, treatment with aqueous acid led to a rearrangement that gave 42 in 83% yield. Unmasking the carbonyl with mercuric salts led to an 80% yield of 43.51 1,4-Addition products are observed in some reactions with stabilized carbanions of this type, but this product may be due to a kinetic–thermodynamic phenomenon in which the initially formed 1,2-addition product equilibrates to a 1,4adduct.52 This equilibration can be induced by treatment with base. For example, Wilson et al.53 showed that treatment of 41 with 1–1.5 equiv of KH in HMPA/THF led to a 23% yield of the 1,4-adduct (the 3-dithianylcyclohexanone, 42). The extent of 1,2- versus 1,4-addition observed is dependent on the reaction conditions employed. 1.

Me Li

S

HO

Me S aq HCl

S

O 2. H3O+

Cyclohex-2-en-1-one

HO Me S

S OH 41 (99%)

Hg2+

S 42 (83%)

Me

H2O

O 43 (80%)

As mentioned above, many reagents have been used for hydrolysis, which is due in part to variations in reactivity of the dithiane as the nature of the 2-alkyl group changes. Several problems can arise during hydrolysis. The hydrolysis reaction becomes acidic, so BaCO3, CaCO3, or HgO are often added to maintain neutrality. Hydrolysis of dithianes derived from aldehydes often give acetals in neutral media. When an alkenyl group is present, addition of mercuric ion can give a competitive oxymercuration reaction (Section 2.5.2). In general, HgO/BF3• OEt2 in aqueous organic solvents is used for hydrolysis of sensitive substrates.51 Acyl anion equivalents are not limited to dithianes, and virtually any dithioacetal can be used. Another common reagent is a dithiolane (44),54 prepared from an aldehyde and 1,2-ethanedithiol. A dithiolane (e.g., 44) reacts with organolithium reagents to yield 2-lithio-1,3-dithiolane derivative, and subsequent reaction with an alkyl halide generates the 2-substituted dithiolane. In some cases, lithio-dithiolanes can fragment, so dithiane derivatives are more commonly used as umpolung reagents. The same reagents that unmasked a dithiane to give back a carbonyl will convert a dithiolane to a carbonyl. Other reagents that function as acyl anion equivalents include 45,55 46,56 and 47.57 50

Stockman, R. A.; Sinclair, A.; Arini, L. G.; Szeto, P.; Hughes, D. L. J. Org. Chem. 2004, 69, 1598.

51

(a) Corey, E. J.; Seebach, D. Angew. Chem. Int. Ed. Engl. 1965, 4, 1075, 1077. (b) Corey, E. J.; Crouse, D. J. Org. Chem. 1968, 33, 298.

(a) Schultz, A. G.; Yee, Y. K. J. Org. Chem. 1976, 41, 4044. (b) Deschamps, B.; Anh, N. T.; Seyden-Penne, J. Tetrahedron Lett. 1973, 527. (c) Luchetti, J.; Krief, A. Ibid. 1978, 2697. (d) Still, W. C.; Mitra, A. Ibid. 1978, 2659.

52

53

(a) Wilson, S. R.; Misra, R. N.; Georgiadis, G. M. J. Org. Chem. 1980, 45, 2460. (b) Wilson, S. R.; Misra, R. N. Ibid. 1978, 43, 4903.

54

Herrmann, J. L.; Richman, J. E.; Schlessinger, R. H. Tetrahedron Lett. 1973, 2599.

55

(a) Hackett, S.; Livinghouse, T. J. Org. Chem. 1986, 51, 879, 1629. (b) Trost, B. M.; Miller, C. H. J. Am. Chem. Soc. 1975, 97, 7182.

56

Trost, B. M.; Salzmann, T. N. J. Am. Chem. Soc. 1973, 95, 6840.

57

(a) Iriuchijima, S.; Maniwa, K.; Tsuchihashi, G. J. Am. Chem. Soc. 1975, 97, 596. (b) Carey, F. A.; Hernandez, O. J. Org. Chem. 1973, 38, 2670.

614

12. CARBON-CARBON BOND-FORMING REACTIONS

S

S

R

Li

OMe

O

PhS

PhS MeS

Li

H 44

O

Ph

45

Li

O 46

SiMe3 47

Et

S

Et

Li

S 48

Schlessinger and coworkers used a mixed sulfide-sulfoxide derivative (48) to achieve an umpolung. The advantages of this reagent include ease of generation (the proton is more acidic and NaH, KH, LDA can be used), greater stability of the anion and lower cost. Since the anion is more stable it is less reactive, and the alkylation reactions generally require higher temperatures.58 The diethyl derivative is superior to the dimethyl.59 Acid hydrolysis is more facile, which is important since dithiane hydrolysis can be sluggish with highly substituted derivatives. The lithio derivative 48 gave primarily Michael addition (1,4- or conjugate addition) in conjugated systems rather than the 1,2-addition usually observed with dithiane anions.60 These examples are meant to show there are many possible acyl anion equivalents. Such exotic reagents are used to fine-tune an umpolung reaction when other reagents are unsatisfactory. In general, relatively simple dithiane and dithiolane derivatives are used. A major drawback to using dithiolane anions is their relative instability when compared with dithiane anions. Lithio-dithiolanes are stable only when R in 44 is an anion-stabilizing group (e.g., COR or CO2R), but when used in conjunction with such electron-withdrawing groups, the dithiolane becomes very useful. Enol ethers [e.g., methyl prop-1-enyl ether, (E)-1-methoxyprop-1-ene] constitute a second major class of acyl anion equivalents. When (E)-1-methoxyprop-1-ene was treated with tert-butyllithium (note the need for a stronger base with the less acidic vinyl hydrogen), and then condensed with benzaldehyde, the product was 2-methoxy-1-phenylbut-2-en-1-ol.61 Lithiation of vinyl derivatives was described in Section 11.6. Facile hydrolysis with aqueous acid liberated the corresponding ketone (1-hydroxy-1-phenylbutan-2-one), completing the acyl anion equivalency. Schlosser and coworkers62 found that a mixture of sec-butyllithium and potassium tert-butoxide could be used to generate the lithium anion of O-tetrahydropyranyl enol ethers. This modification generates a product that is more easily hydrolyzed to the ketone. MeO

1. t-BuLi, TMEDA THF, –70 → +20°C

OMe

Ph

O

Ph

H3O+

2. PhCHO, –20°C

OH (E)-1-Methoxyprop-1-ene

OH

2-Methoxy-1-phenylbut-2-en-1-ol

1-Hydroxy-1-phenylbutan-2-one

Vinyl sulfides can also function as acyl anion equivalents, but their preparation requires some chemistry that has not as yet been discussed.63,64 Vinyl sulfides can be converted to the corresponding vinyllithium reagent, and subsequent reaction with alkyl halides is usually facile. Hydrolysis with mercuric salts or another Lewis acid usually gives a ketone product. An example is the reaction of ethylvinyl sulfide with sec-butyllithium to yield alkenyllithium (49). Subsequent treatment with an alkyl halide (e.g., 1-bromooctane) led to 50, and hydrolysis with mercuric chloride65 gave decan-2one in 90% yield. S

THF–HMPA –78°C

Ethylvinyl sulfide

S

sec-BuLi

C8 H17 Br

S C8 H17

Li 49

HgCl2 aq MeCN

O

Reflux

50

Me

C8 H17

Decan-2-one (90%)

58

(a) Richmann, J. E.; Herrmann, J. L.; Schlessinger, R. H. Tetrahedron Lett. 1973, 3267. (b) Ogura, K.; Tsuchihashi, G. Ibid. 1971, 3151.

59

Carlson, R. M.; Helquist, P. M. J. Org. Chem. 1968, 33, 2596.

60

Herrmann, J. L.; Richman, J. E.; Schlessinger, R. H. Tetrahedron Lett. 1973, 3271.

61

Baldwin, J. E.; H€ ofle, G.; Lever, O. W., Jr. J. Am. Chem. Soc. 1974, 96, 7125.

62

Hartmann, J.; St€ ahle, M.; Schlosser, M. Synthesis 1974, 888.

63

(a) Corey, E. J.; Seebach, D. J. Org. Chem. 1966, 31, 4097. (b) Carey, F. A.; Court, A. S. Ibid. 1972, 37, 939.

64

For another synthetic example using stannane derivatives, see Hanessian, S.; Martin, M.; Desai, R. C. J. Chem. Soc. Chem. Commun. 1986, 926.

65

Oshima, K.; Shimoji, K.; Takahashi, H.; Yamamoto, H.; Nozaki, H. J. Am. Chem. Soc. 1973, 95, 2694.

615

12.2 SULFUR-STABILIZED CARBANIONS AND UMPOLUNG

The disconnection for vinyl sulfides and enol ethers follows: O R1

R

O

+

RJX

H

R1

12.2.2.2 Other Umpolung Equivalents Several umpolung equivalents other that acyl anion equivalents (29) are listed in Table 12.1. For most of these umpolung equivalents, the appropriate reagent is used only occasionally in synthesis, so only summaries of each type will be provided here. The R(CO)CH+2 umpolung equivalent is usually a ketene dithioacetal, often prepared by a Peterson olefination,66,67 or from dithiocarboxylic acid derivatives.68,69 Ketene dithioacetals can be prepared by reaction of an aldehyde or ketone with dithioalkyl phosphonate esters.70,71 The actual umpolung reaction involves conjugate addition of a nucleophile to a ketene dithioacetal, followed by hydrolysis.72 Ketene dithioacetals function as the umpolung equivalent for R(CO)CH2CH 2 , but the reactive center is more distant from the carbonyl, making it more difficult to generate the umpolung reagent. It appears that the umpolung is observed as the major process only with an electron-withdrawing group.35c,73 The dianion of an allylthiol (e.g., prop-2-ene-1-thiol), allows alkylation at carbon,74,75 but reaction as an umpolung appears to depend on the metal that is used.76 Allylic sulfides are used more often than allylic thiols since the monoanion yields α- and γ-substitution.15b,77 The umpolung equivalent R(C]O)CH2CH2CH+2 has a potentially positive center even further removed from the carbonyl. This umpolung equivalent is best achieved by reaction of a nucleophile with a vinylogous ketene dithioacetal.78 The ketene dithioacetal cannot have an acidic hydrogen atom present in the dithioacetal unit, since it would be selectively deprotonated. The R(C]O)CH2CH2CH2CH 2 equivalent has the reactive center separated from the carbonyl by three carbon atoms. This umpolung equivalent is relatively easy to obtain, however, by the use of an allyl vinyl sulfide, which yields a thio-Claisen rearrangement (Section 15.5.5.3) upon heating.79 A simple example is the reaction of allyl vinyl sulfide 51 with sec-butyllithium to give the reactive α-alphalithio sulfide. This compound reacted with benzyl bromide to yield 52. Heating in aqueous dimethoxyethane (DME) led to the thio-Claisen product (E)-6-phenylhex-4-enethial, and hydrolysis provided the aldehyde, (E)-6-phenylhex-4-enal, in 62% overall yield.80 The specificity for the trans-alkene moiety in (E)-6-phenylhex-4-enethial arises from the chair-like transition state (see 52; Section 15.5.5), in which the benzyl group is equatorial in the lowest energy conformation leading to product. S

1. sec-BuLi –78°C

H

Ph

Ph

H

2. PhCH2Br

51

Ph

DME–H2O

S

CaCO3, 11 h Reflux

52

S

H3O+

H (E)-6-Phenylhex4-enethial

O H (E)-6-Phenylhex(62%) 4-enal

(a) Peterson, D. J. J. Org. Chem. 1968, 33, 780. (b) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-71. (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; pp 496, 497.

66

67

(a) Carey, F. A.; Court, A. C. J. Org. Chem. 1972, 37, 1926. (b) Trost, B. M.; Kunz, R. A. J. Am. Chem. Soc. 1975, 97, 7152.

68

Beiner, J. M.; Thuillier, A. C. R. Acad. Sci. Ser. C 1972, 274, 642.

69

Ziegler, F. E.; Chan, C. M. J. Org. Chem. 1978, 43, 3065.

70

Mikolajczyk, M.; Grzejszczark, S.; Zatorski, A.; Mlotkowska, B.; Gross, H.; Costisella, B. Tetrahedron 1978, 34, 3081.

71

(a) Corey, E. J.; M€ arkl, G. Tetrahedron Lett. 1967, 3201. (b) Lemal, D. M.; Banitt, E. H. Ibid. 1964, 245.

72

Herrmann, J. L.; Kieczykowski, G. R.; Romanet, R. F.; Wepplo, P. J.; Schlessinger, R. H. Tetrahedron Lett. 1973, 4711.

73

Meyers, A. I.; Nolen, R. L.; Collington, E. W.; Narwid, T. A.; Strickland, R. C. J. Org. Chem. 1973, 38, 1974.

74

Geiss, K.-H.; Seuring, B.; Pieter, R.; Seebach, D. Angew. Chem. Int. Ed. Engl. 1974, 13, 479.

75

Hartmann, J.; Muthukrishnan, K.; Schlosser, M. Helv. Chim. Acta 1974, 57, 2261.

76

Seebach, D.; Geiss, K.-H.; Pohmakotr, M. Angew. Chem. Int. Ed. Engl. 1976, 15, 437.

77

Atlani, P. M.; Biellmann, J. F.; Dube, S.; Vicens, J. J. Tetrahedron Lett. 1974, 2665.

78

Seebach, D.; Kolb, M.; Gr€ obel, B.-T. Angew. Chem. Int. Ed. Engl. 1973, 12, 69.

79

Corey, E. J.; Shulman, J. I. J. Am. Chem. Soc. 1970, 92, 5522.

80

Oshima, K.; Takahashi, H.; Yamamoto, H.; Nozaki, H. J. Am. Chem. Soc. 1973, 95, 2693.

616

12. CARBON-CARBON BOND-FORMING REACTIONS

The ability to alter normal reactivity patterns of carbonyl derivatives allows a variety of carbon-carbon bondforming processes to proceed. The products derived from these reactions can often be obtained by no other route and provide great flexibility in synthetic planning.

12.3 ORGANOCOPPER REAGENTS (CdCU) In all previous sections, the reagent used as a nucleophilic source of carbon was an organometallic derived from Na, K, Mg or Li reagents. In this section, the focus will be on a class of reagents that contains a carbondcopper bond, the lithium dialkylcuprates (R0 2CuLi).

12.3.1 Organocuprates 12.3.1.1 Preparation of Gilman Reagents Gilman et al.81,82 first observed that reaction of cuprous salts with 2 equiv of an organolithium reagent generated an organocuprate (e.g., lithium dibutylcuprate, Bu2CuLi). The reagent formed is usually called a Gilman reagent, and it is drawn as R2CuLi.81,82 The mechanism83 of the reaction probably involves an intermediate and conversion of a Cu(I) species to a transient Cu(III) species, and it may proceed via one-electron transfer.84 The stability of such organocopper species varies considerably with their structure. Whitesides et al.85 showed that dimethylcuprate (Me2CuLi) was stable in ether for hours, at 0°C under nitrogen. Secondary and tertiary cuprates, however, rapidly decompose in ether above 20°C.85 Disproportionation is the primary decomposition route for organocuprates. The halide plays a role in the stability of the organocuprate. Cuprous iodide gives better results than cuprous bromide,85 but the dimethyl sulfide-cuprous bromide complex (Me2SCuBr) has been used with success.86 Ether

2 BuLi + CuI l



! ∘ 20 C

Bu2 CuLi Lithium dibutylcuprate

The two major uses of an organocuprate as a carbon nucleophile follow87: (1) reaction with alkyl halides and (2) conjugate addition with α,β-unsaturated ketones. With α,β-unsaturated ketones, conjugate addition is promoted when ether is used as a solvent.84 The substitution reaction is promoted by the use of THF or ether-HMPA as a solvent.84 As mentioned earlier, the mechanism of these reactions probably involves a one-electron transfer, although other mechanistic proposals are in the literature.88 The general reactivity of organocuprates with electrophiles follows the order:84 O

O R

>

>

RJCHO > RJOTs

Cl

O >

>

R

RJI

> RJBr > RJCl

R RJCO2R1

> RJCLN

>>

R

R

81

Gilman, H.; Jones, R. G.; Woods, L. A. J. Org. Chem. 1952, 17, 1630.

82

Gilman, H.; Straley, J. M. Recl. Trav. Chim. Pays-Bas 1936, 55, 821.

83

For an extensive discussion of the mechanism of reaction between organocuprates and alkyl haldies or epoxides, see Mori, S.; Nakamura, E.; Morokuma, K. J. Am. Chem. Soc. 2000, 122, 7294. 84

Posner, G. H. An Introduction to Synthesis Using Organocopper Reagents; Wiley: New York, 1980.

85

Whitesides, G. M.; Fischer, W. F., Jr.; San Filippo, J., Jr.; Bashe, R. W.; House, H. O. J. Am. Chem. Soc. 1969, 91, 4871.

86

House, H. O.; Wilkins, J. M. J. Org. Chem. 1978, 43, 2443.

(a) Ashby, E. C.; DePriest, R. N.; Tuncay, A.; Srivastava, S. Tetrahedron Lett. 1982, 23, 5251. (b) Posner, G. H.; Ting, J.-S.; Lentz, C. M. Tetrahedron 1976, 32, 2281. (c) Johnson, C. R.; Dutra, G. A. J. Amer. Chem. Soc. 1973, 95, 7777, 7783. (d) Smith, R. A. J.; Hannah, D. J. Tetrahedron 1979, 35, 1183. (e) Castro, C. E.; Havlin, R.; Honwad, V. K.; Malte, A.; Moje, S. J. Am. Chem. Soc. 1969, 91, 6464. (f) House, H. O.; Koespell, D. G.; Campbell, W. J. J. Org. Chem. 1972, 37, 1003.

87

88

Nakamura, E.; Mori, S. Angew. Chem. Int. Ed. 2000, 39, 3750.

12.3 ORGANOCOPPER REAGENTS (CdCU)

617

The R groups can be primary, secondary, tertiary alkyl, aryl, or heteroaryl and can bear remote functionality (e.g., ethers, acetals or ketals, sulfides, or other labile groups).84 12.3.1.2 Higher Order Organocuprates Gilman-type reagents are very useful, but there are a few problems associated with them, including reactivity and thermal instability. In an attempt to circumvent these problems, Lipshutz et al.89 developed the so-called higher order mixed cuprates, R2Cu(CN)Li2, prepared by reaction of 2 equiv of an organolithium reagent with cuprous cyanide (CuCN).90 Mixed cuprates react faster than Gilman reagents with alkyl halides, even secondary halides. Mixed organocuprates are generally the reagent of choice for this purpose. A survey of the reactions of alkyl halides with the higher order organocuprate derived from butyllithium and CuCN showed good-to-excellent yields of the coupling product, alkane, RdBu.91a In this study, secondary halides also gave excellent yields, the rate of alkylation was faster than observed with Gilman reagents, and even normally unreactive alkyl chlorides gave good yields of coupling products with reasonable reaction rates.91a 2RdLi + CuCN


! R2 CuðCNÞLi2 92

Bertz questioned the validity of the R2Cu(CN)Li2 structure given for organocuprates, citing 13C NMR evidence that the reagent actually existed as R2CuLiLiCN in THF. Lipshutz et al.93 contradicted this conclusion, when the 1H and 13 C NMR spectra of the higher order cuprate Me2Cu(CN)Li2 were compared with those of Me2CuLiLiBr and Me2CuLiLiI. He concluded that the spectra were not at all similar. Lipshutz concluded that the spectra proved the presence of a CudCN bond, and a higher order cuprate structure. Ab initio evidence,94a additional NMR evidence,95 as well as kinetic evidence provided by Bertz and coworkers94b,c again called these structures into question, and he suggested their structure to be cyano-Gilman reagents. A study by Bertz and coworkers94b,c concluded there is no higher order cyanocuprate, but that the active species is best described as R2CuLiLiCN. A theoretical study suggests that a minor species R(CN)CuLiLiR may play an important role due to its greatly enhanced reactivity.96 For reactions in this book, higher order cuprates are represented as R2Cu(CN)Li2. The disconnections for the higher order organocuprate are identical to those observed with Gilman reagents.

12.3.2 Reactions of Organocuprates 12.3.2.1 Alkyl Coupling Reactions As noted in Sections 11.4.2 and 11.6.2, Grignard reagents and organolithium reagents usually gave poor yields of coupling products in a reaction with an alkyl halide unless catalyzed by transition metals. Indeed, cuprous salts were used with Grignard reagents to catalyze coupling reactions (Section 11.4.2). Addition of cuprous salts to organolithium reagents gave organocuprates (R0 2CuLi), the active species in the coupling reaction, and this method is usually preferred for the formation of coupled products (RdR0 ) from alkyl halides, as indicated in Section 12.3.1.1.97 The reactivity of the cuprate is due in large part to the low ionic character of the CudC bond, the low oxidation potential (0.15 eV) separating Cu(I) and Cu(II) ions, the tendency for Cu to form polynuclear Cu clusters, and the formation of mixed-valence Cu compounds.84 Ether

R dX + R1 2 CuLi



! R dR1 o 20 C

The coupling reaction of organocuprates with alkyl halides, tosylates, or acetates is straightforward, but the mechanistic details have been a source of some controversy. Primary alkyl halides generally give good yields of the coupling 89

Lipshutz, B. H.; Wilhelm, R. S.; Kozlowski, J. A. Tetrahedron 1984, 40, 5005.

90

Lipshutz, B. H.; Wilhelm, R. S. J. Am. Chem. Soc. 1982, 104, 4696.

(a) Lipshutz, B. H.; Wilhelm, R. S.; Kozlowski, J. A.; Parker, D. J. Org. Chem. 1984, 49, 3928. (b) Yamamoto, K.; Iijima, M.; Ogimura, Y.; Tsuji, J. Tetrahedron Lett. 1984, 25, 2813. (c) Takahashi, T.; Okumoto, H.; Tsuji, J. Ibid. 1984, 25, 1925.

91

92

Bertz, S. H. J. Am. Chem. Soc. 1990, 112, 4031.

93

Lipshutz, B. H.; Sharma, S.; Ellsworth, E. L. J. Am. Chem. Soc. 1990, 112, 4032.

(a) Snyder, J. P.; Spangler, D. P.; Behling, J. R.; Rossiter, B. E. J. Org. Chem. 1994, 59, 2665. (b) Snyder, J. P.; Bertz, S. H. J. Org. Chem. 1995, 60, 4312. (c) Bertz, S. H.; Miao, G.; Eriksson, M. Chem. Commun. 1996, 815. Also see (d) Krause, N. Angew. Chem. Int. Ed. 1999, 38, 79. 95 € Snyder, J. P. Angew. Chem. Int. Ed. 1998, 37, 314. Bertz, S. H.; Nilsson, K.; Davidsson, O.; 94

96

Nakamura, E.; Yoshikai, N. Bull. Chem. Soc. Jpn. 2004, 77, 1.

97

For reactions of this type, see Larock, R. C. Comprehensive Organic Transformations, 2nd ed.; Wiley–VCH: New York, 1999; pp 118–122.

618

12. CARBON-CARBON BOND-FORMING REACTIONS

product (RdR1) in a Wurtz-type process (Section 11.6.3).98 The relative reactivity of dibutylcuprate with various organohalides to give the alkane RdR1 has been studied.85 The reaction proceeded faster and in higher yield in THF than in ether. From this study the order of reactivity of the halide substrate was determined to be RI > RBr > RCl. Secondary halides gave poorer yields, and tertiary halides gave virtually no reaction under conditions that produced large amounts of product from primary halides. Sulfonate esters (e.g., mesylates or tosylates) and also acetates react by coupling with organocuprates.99 The coupling reaction is extremely useful for joining alkyl fragments or alkyl-aryl, aryl-aryl, or heteroaryl fragments by reaction with alkyl halides or sulfonate esters.100 In a typical synthetic application, lithium dimethylcuprate reacted with tosylate 53 give a 91% yield of 54 in a synthesis o (+)-bourgeanic acid by Yadav and coworkers.101 Note that organocuprates react with secondary bromides with inversion, whereas the identical reaction with a secondary iodide proceeds with loss of enantiopurity, and sometimes racemization. In addition to the lithium dialkylcuprates, it is also possible to form magnesium cuprates that react similarly. An example is the reaction of tosylate 55 with the magnesium cuprate generated in situ by the reaction of allylmagnesium bromide with CuI, which gave >62% of 56 in a synthesis of virgineone by Yajima et al.102 Me

Me

Me

Me

TsO

Me2CuLi, Ether

OBn

O

O

Me

Me

Me

–30 to 0°C

OBn

53

O

O

54 (91%) OTs

Ph

Ph

MgBr, CuI, THF

O

O

0°C – rt

O

O 55

56 (>62%)

If there is excessive steric hindrance at an allylic carbonin a reaction with an organocuprate, the coupling may proceed by a SN20 like pathway (Section 3.2.1.3).103 A synthetic example is taken from a synthesis of paroxetine (PAXIL) by Krische and Koech104 in which allylic phosphonate 57 was displaced by the organocuprate derived from ClSiMe2(Oi-Pr) to give the SN20 product 58 in 96% yield. PhO PhO

P

O

F

F (i-PrO)Me2SiCl/Mg; CuCN THF

MeO2C

N

MeO2C 57

N

SiMe2(Oi-Pr) 58 (96%)

In addition to the expected displacement of alkyl halides, organocuprates react with vinyl halides and aryl halides to give good yields of the coupled product.105 In other words, organocuprates react with both sp3 and sp2 hybridized carbon atoms. This alkenyl halide coupling reaction proceeds with high stereoselectivity, in contrast to the Grignard

98

House, H. O. Acc. Chem. Res. 1976, 9, 59.

99

(a) Posner, G. H.; Ting, J. S.; Lentz, C. M. Tetrahedron 1976, 32, 2281. (b) Posner, G. H.; Ting, J. S. Tetrahedron Lett. 1974, 683.

100

Kojima, Y.; Wakita, S.; Kato, N. Tetrahedron Lett. 1979, 4577.

101

Yadav, J. S.; Raghavendra, K. V.; Ravindar, R. K.; Subba Reddy, B. V. Eur. J. Org. Chem. 2011, 58.

102

Yajima, A.; Ida, C.; Taniguchi, K.; Murata, S.; Katsuta, R.; Nukada, T. Tetrahedron Lett. 2013, 54, 2497.

(a) Magid, R. M. Tetrahedron 1980, 36, 1901. (b) DeWolfe, R. H.; Young, W. Chem. Rev. 1956, 56, 753. (c) Goering, H. L.; Singleton, V. D., Jr. J. Org. Chem. 1983, 48, 1531.

103

104

Koech, P. K.; Krische, M. J. Tetrahedron 2006, 62, 10594.

105

(a) Bowlus, S. B.; Katzenellenbogen, J. A. J. Org. Chem. 1973, 38, 2733. (b) Cooke, M. P., Jr. Tetrahedron Lett. 1973, 1983.

619

12.3 ORGANOCOPPER REAGENTS (CdCU)

and organolithium coupling reactions. (E)-2-Bromostyrene reacted with diphenylcuprate, for example, to give a 90:60%)

O

The presence of halogen atoms in a molecule is compatible with formation of an ylid and subsequent Wittig reactions usually proceed normally. A halogen atom may also be part of the ylid itself. Carbon tetrabromide and triphenylphosphine have been used to yield vinyl dibromide products, which are rapidly converted to the corresponding alkyne. This important synthetic route to alkynes is called the Corey–Fuchs procedure, previously mentioned in Section 3.5.1.193 Subsequent base-induced double elimination is possible to generate an alkyne. An illustration of this procedure is the reaction of 126 with triphenylphosphine and carbon tetrabromide to yield vinyl dibromide 127, as part of Kato and coworker’s194 synthesis of (+)-gregatin B. When 127 was treated with butyllithium, alkyne 128 was isolated in 89% yield over both steps. Br

O TBDMSO Me

H

5 PPh3 2.5 CBr4

TBDMSO Me

OTBDMS 126

Br BuLi, THF

H

TBDMSO Me OTBDMS

OTBDMS 127

128 (89%)

187

Marinier, A.; Deslongchamps, P. Tetrahedron Lett. 1988, 29, 6215.

188

Russell, M. G.; Warren, S. J. Chem. Soc. Perkin Trans. 1 2000, 505.

189

Murphy, P. J.; Lee, S. E. J. Chem. Soc. Perkin Trans. 1 1999, 3049.

190

For reactions of this type, see Larock, R. C. Comprehensive Organic Transformations, 2nd ed.; Wiley–VCH: New York, 1999; p 328.

191

Corey, E. J.; Carpino, P. J. Am. Chem. Soc. 1989, 111, 5472.

192

Nicolaou, K. C.; Ding, H.; Richard, J.-A.; Chen, D. Y.-K. J. Am. Chem. Soc. 2010, 132, 3815.

193

Corey, E. J.; Fuchs, P. L. Tetrahedron Lett. 1972, 3769.

194

Kusakabe, T.; Kawai, Y.; Kato, K. Org. Lett. 2013, 15, 5102.

633

12.5 YLIDS

Some substituents on a phosphonium salt are subject to spontaneous fragmentation when the carbanion is formed.195 Such fragmentation occurs most often when a leaving group is β to the P atom. Ylids with a heteroatom substituent can decompose to a carbene intermediate (Section 17.9). Butoxymethyl ylid 129, for example, spontaneously fragmented to triphenylphosphine and carbene 130. This carbene reacted with an additional molecule of the ylid (129) to yield 1,2-dibutoxyethene via the zwitterion 131.196 Ph3P

+

−

– PPh3

OBu

BuO

+ 129

CH:

45°C

PPh3

−

BuO

129

130

BuO

– PPh3

OBu

OBu

1,2-Dibutoxyethene

131

Phosphorus ylids rarely rearrange, but the presence of a reactive group (e.g., a halide197or ester)198 can lead to anomalous products. In one example, the carbanionic carbon of 132 displaced the bromide to give cyclobutyltriphenylphosphonium bromide, via an internal SN2 reaction.195 In another example, ester 133 suffered nucleophilic attack of the carbanionic carbon to generate α-keto ylid 134 via acyl substitution.196 +

+

−

Ph3P

Cyclobutyltriphenylphosphonium bromide

132

+

OEt +

Br–

Ph3P

Br

PPh3

−

−

O

Ph3P

O 133

134

12.5.1.2 (E/Z) Isomers in Wittig Reactions When a phosphorus ylid reacts with an aldehyde or ketone, the alkene is formed as a mixture of (E)- and (Z)-isomers.199 Initial postulations suggested that the stereochemistry of the alkene products was controlled by the stereochemistry of the betaine, which rapidly collapsed to an oxaphosphetane. Subsequent fragmentation to the alkene occurred via syn-elimination.165 This postulate is illustrated by the reaction of ylid 135 and butan-2-one to generate a mixture of betaines 136 and 137. Betaine 136 collapses to the anti-oxaphosphetane (138),165 and syn-elimination yields the (E)-alkene, (E)-3,4-dimethylhex-3-ene. Similarly, betaine (137) collapses to the syn-oxaphosphetane (139), which yields the (Z)-alkene, (Z)-3,4-dimethylhex-3-ene. +

−

Ph3P Me +

Ph3P

−

O Me

136

+

Me

Me Me

O

Me

Ph3P

O

Me (E)-3,4-Dimethylhex-3-ene

138

Me 135

+

−

Ph3P Me

O

Ph3P

O

Me

Me

Me

Me 137

Me

139

(Z)-3,4-Dimethylhex-3-ene

(a) Wittig, G.; Eggers, H.; Duffner, P. Annalen 1958, 619, 10. (b) Trippett, S. Chem. Ind. (London) 1956, 80. (c) Dicker, D. W.; Whiting, M. C. J. Chem. Soc. 1958, 1994.

195

196

(a) Wittig, G.; B€ oll, W. Chem. Ber. 1962, 95, 2526. (b) Trippett, S. Proc. Chem. Soc. 1963, 19.

197

(a) Mondon, A. Annalen 1957, 603, 115. (b) Scherer, K. V.; Lunt, III, R. S. J. Org. Chem. 1965, 30, 3215.

198

Bergel’son, L. D.; Vaver, V. A.; Barsukov, L. I.; Shemyakin, M. Izvest. Akad. Nauk. SSSR 1963, 1134.

199

Reucroft, J.; Sammes, P. G. Q. Rev. Chem. Soc. 1971, 25, 135.

634

12. CARBON-CARBON BOND-FORMING REACTIONS

When conjugated substituents (e.g., a carbonyl or an aromatic ring) are directly connected to the carbanionic center, the ylid is resonance stabilized. This resonance stability of ylids (e.g., such as Ph3P]CHCO2Et or Ph3P]CHPh) leads predominantly to the (E)-alkene, whereas ylids that cannot be stabilized by conjugating substituents, (e.g., Ph3P] CHEt), give predominantly the (Z)-alkene.200 These results are also rationalized in terms of a betaine intermediate. For resonance-stabilized ylids, formation of the betaine is reversible and equilibration favors formation of the thermodynamically more stable anti-betaine. The anti-betaine may also collapse faster than the syn-betaine.201 If elimination of triphenylphosphine oxide is slow, the possibility of equilibration to the anti-betaine is maximized, and there will be increased amounts of the (E)-alkene. Tricyclohexylphosphine ylids are less electrophilic than triphenylphosphine ylids, and they show greater selectivity for formation of the (E)-alkene.202 Addition of a protic solvent or a Lewis acid to coordinate the oxygen atom of the betaine leads to more (Z)-isomer.199,202 As indicated for the reaction of 135, the initial reaction of an aldehyde or a ketone with a ylid is not very selective, leading to a mixture of stereoisomeric betaines, which gives the isomeric alkenes. This mixture is subject to equilibration, which is attributed to the reversibility of the initial reaction. The initial reaction of an aldehyde with Ph3P] CHCO2Me, for example, leads to an equilibrium mixture of the aldehyde, Ph3P]CHCO2Me, 140, and 142. If rotation about the CdC bond occurs before collapse to an oxaphosphetane, 140 is in equilibrium with 141 and 142 with 143. The conjugating carbonyl unit stabilizes the betaine, which tends to slow collapse to the oxaphosphetane. R

+

−

M O

H CO Me 2

R

+

PPh3

H 143

CO2Me

H

H

R +

−

M O

CO2Me

PPh3 +

RCHO

+

Ph3P=CHCO2Me

H

R

CO2Me

H +

M O

142

−

PPh3 +

+

−

M O

H CO Me 2 +

H

PPh3

R

140

141

R CO2Me

Remember from Section 12.5.1.1 that the ylid is formed by reaction of a phosphonium salt (e.g., a triphenylphosphonium bromide) with a base (e.g., butyllithium), so in this case the salt LiBr is also present. In the presence of this salt, the metal can coordinate with the alkoxide moiety, which tends to stabilize 141 and 143, and slow down collapse to the betaine and then to formation of the alkene. In such a case, these coordinated species show some preference for the anticonformation 143, which gives a slight preference for the (Z)-isomer. These arguments refer only to stabilized phosphoranes and ignore the stereochemistry of the groups around the phosphorus atom, which are stable throughout the reaction.203 When the ylid is not stabilized, as when it possesses alkyl substituents rather than conjugating substituents, betaine, and oxaphosphetane formation is effectively irreversible. Reaction of unsaturated ylids with aromatic aldehydes or α,β-unsaturated aldehydes shows some reversibility, however.204 Since metal salts are usually present in the reaction medium, depending of the base use to generate the ylid. Under salt-free conditions, however, the reaction appears to be under kinetic control, leading to the syn-betaine and the (Z)-alkene.205 Indeed, Schlosser et al.206 found that manipulating the solvent and the salt concentration allows some control of the (E/Z) selectivity, and rationalized these observations in terms of a betaine intermediate. The betaine can react with unreacted n-butyllithium to form a dilithio species, which is in equilibrium with the epimeric dilithio species.206

200

Schlosser, M.; Christman, K. F. Annalen 1967, 708, 1.

201

Speziale, A. J.; Bissing, D. E. J. Am. Chem. Soc. 1963, 85, 1888, 3878.

202

Bestmann, H. J.; Kratzer, O. Berichte 1962, 95, 1894.

203

Blade-Font, A.; McEwen, W.; VanderWerf, C. A. J. Am. Chem. Soc. 1960, 82, 2646, 2396.

204

Schlosser, M.; Christmann, K. F. Angew. Chem. Int. Ed. Engl. 1965, 4, 689; Ibid. 1964, 3, 636.

205

(a) Bestmann, H. J. Angew. Chem. Int. Ed. Engl. 1965, 4, 583. (b) Wittig, G.; Eggers, H.; Duffner, P. Annalen 1958, 619, 10.

206

(a) Schlosser, M.; M€ uller, G.; Christmann, K. F. Angew. Chem. Int. Ed. Engl. 1966, 5, 667. (b) Schlosser, M.; Christmann, K. F. Ibid. 1966, 5, 126.

635

12.5 YLIDS

If Li salts are added to this equilibrium, oxaphosphetane formation and elimination from the kinetic syn-betaine is inhibited. Further addition of excess butyllithium favors the anti-species. Protonation leads to an oxaphosphetane, which collapses to the anti-alkene. This study suggests that excess base and excess salt promotes equilibration and formation of the anti (E)-alkene. As indicated previously in this section, salt-free conditions are required to obtain the (Z)-alkene. As the carbonyl reactant becomes more hindered, there is a greater preference for the (Z)-alkene.207 Schneider208 explained this effect by considering the importance of the substituent configuration around the phosphorus atom during reaction. The intermediate 144 (apical oxygen and equatorial ylid carbon) must rotate at the OdC bond to form the betaine (145 or 146). The Ph $ R2 steric interaction is maintained throughout, but the R1 $ Ph interaction in 155 inhibits formation of 146 and favors 145 (the syn-betaine), leading to the (Z)-alkene. With protic solvents or polar additives, the configuration of the phosphorane may change, and this model does not apply. For stabilized ylids, the pre-betaine intermediate is represented by 147 and is oriented so that the positive end is attracted to the nucleophilic acyl carbon, and forms the anti-betaine and the (E)-alkene.202 The R1 group rotates to form the betaine, and this rotation minimizes the R1 $ CO2Et interaction. O Ph

Ph

P Ph

O R1

1 Ph R

C − 144

H H

H Ph

H

Ph

P

C

+

R2

O C

Ph

Ph

C H

R2

Ph

Ph C

+

P Ph

C

C

146

H

−

O

C

R2

H

R2

145

R1

C

P

R1

O

Ph

147

OE

+

The intermediacy of the betaines in the Wittig reaction has been questioned, and an alternative has been proposed. The Wittig reaction is subject to solvent effects that indicate a nonpolar transition state for stabilized ylids.209 There appears to be no direct evidence for the presence of betaines, and none have been isolated. Alternatively, Vedejs and Snoble210 detected oxaphosphetanes as the only observable intermediates in several Wittig reactions of nonstabilized ylids, using 31P NMR.211 Vedejs and Marth212 devised a test for the betaine mechanism based on changes in phosphorus stereochemistry of betaines versus oxaphosphetanes. The results of this test suggested that “the conventional betaine mechanism,179e,213 can play at most a minor role in the Wittig reaction.”212 Vedejs pointed out that the “stereochemical test does not necessarily disprove mechanisms via intermediates with lifetimes that are short compared to the time scale of bond rotation.”212 R Ph Ph P

O

H

Ph

4

H 2

R1

1

H

Ph

3

148

Ph

1

2

P

3

O

Ph

R1 R H

4

149

Vedejs and Marth214 proposed two oxaphosphetane models to predict the stereochemistry in the Wittig reaction. For systems that give primarily cis-alkenes, a cis-oxaphosphetane is required if the reaction proceeds by an early transition state (like starting materials), and the oxaphosphetane has a puckered conformation (148). Similarly, a transoxaphosphetane is required for systems that give trans-alkenes as the major product. The reaction is assumed to proceed by a late transition state (product-like) and the oxaphosphetane has a planar structure (149). 207

For a synthetic example, see Axen, U.; Lincoln, F. H.; Thompson, J. L. J. Chem. Soc. D, Chem. Commun. 1969, 303.

208

Schneider, W. P. J. Chem. Soc. Chem. Commun. 1969, 785.

(a) Frøyen, P. Acta Chem. Scand. 1972, 26, 2163. (b) Aksnes, G.; Khalil, F. Y. Phosphorus 1972, 2, 105 (Chem. Abstr. 1973, 78, 70865v). (c) Idem Ibid. 1973, 3, 79 (Chem. Abstr. 1973, 79, 145749 s).

209

210

Vedejs, E.; Snoble, K. A. J. J. Am. Chem. Soc. 1973, 95, 5778.

211

Vedejs, E.; Meier, G. P.; Snoble, K. A. J. J. Am. Chem. Soc. 1981, 103, 2823.

212

Vedejs, E.; Marth, C. F. J. Am. Chem. Soc. 1990, 112, 3905.

213

(a) Maercker, A. Org. React. 1965, 14, 270. (b) Schlosser, M. Topics Stereochem. 1973, 5, 1.

214

Vedejs, E.; Marth, C. F. J. Am. Chem. Soc. 1988, 110, 3948.

636

12. CARBON-CARBON BOND-FORMING REACTIONS

A trans-oxaphosphetane is more stable than the cis-oxaphosphetane, and equilibrium conditions favors the transproduct. Kinetic control conditions appear to dominate Wittig reactions and lead to the cis-product using this analogy.210 Vedejs et al.,211 suggested that there is no single, dominant Wittig transition state geometry, but rather a continuum of related mechanistic variants. Examination of 148 and 149 suggests that the interplay of 1,2-steric interactions (interactions of substituents on P and C2 in 148 and also on C2dC3) and 1,3-steric interactions (interactions of substituents on P and C3 in 149) will determine cis or trans selectivity.212 Puckered transition states are more important when large groups at the α-carbon provide increasing 1,3-steric interactions. For stabilized ylids, 1,2-interactions dominate the transition state, leading to trans selectivity. Maryanoff et al.215 also studied this problem, and found a prevalence of (Z)-alkene and cis-oxaphosphetane in the saltfree Wittig reaction, which forms the alkene by syn-elimination of Ph3P]O. His work also suggested a very slow equilibration of cis- and trans-oxaphosphetanes relative to the rate of alkene formation. There is a significant concentration dependence on the stereochemistry of the reaction. Increasing concentration favored the trans-oxaphosphetane.215 At high dilution, alkenes were generally formed with (Z)-selectivity in THF with LiBr present. The concentration effect may be associated with sequestration of Li by THF. The ability of Li soluble salts to promote formation of (E)-alkenes from nonstabilized ylids is well known (see above), but it is concentration dependent in THF. Based on studies with many types of phosphorus ylids, Maryanoff et al.215 concluded that betaines may have a "meaningful, albeit transient, existence," despite the lack of direct evidence for their existence. Since kinetic experiments demonstrated that cisoxaphosphetanes undergo reversal to the ylid and an aldehyde faster than does the trans-oxaphosphetane (see above), the degree of stereochemical drift in Wittig reactions is probably due to the relative rates of oxaphosphetane reversal. A (Z)-selective Wittig reaction is seen in Stark and Adrian’s synthesis of muricadienin,216 in which aldehyde 150 was treated the phosphonium ylid generated in situ to give a 71% yield of 151 as a 5:95 mixture of (E)/(Z)-isomers. An example of an (E)-selective Wittig reaction is taken from a synthetic approach to the synthesis of cermizine D by Carter and coworkers,217 in which aldehyde 152 reacted with Ph3P]CHCO2Me to give the (E)-alkene (153) in 80% yield. +

MeO2CCH2(CH2)9CH2CHO

PPh3 Br−

C12 H25

NaHMDS THF, –20°C to rt

C12 H25

(CH2)11JCO2Me

150

151

(71%, >5:95, E/Z)

CO2Me CHO

N

Ph3P=CHCO2Me, rt

N

CH2Cl2

Boc

Boc

152

153 (80%)

12.5.1.3 Phosphine Oxides and Phosphonate Esters Many modifications to the Wittig reaction have been introduced that improve or modify the reactivity and/or stereoselectivity of the ylid.218 Horner et al.219 showed that α-lithiophosphine oxides such as that derived from butyldiphenylphosphine oxide (154), react with aldehydes or ketones to yield a β-hydroxy phosphine oxide (155) as an isolable species. Subsequent treatment with base liberates the alkene (1,1-diphenylpent-1-ene). O O Ph

1. PhLi, PhMe 2. Benzophenone

P Ph

3. H2O

154

Ph

P

t-BuOK

Ph Ph

Ph

Ph Ph

OH

155

1,1-Diphenylpent-1-ene

(a) Maryanoff, B. E.; Reitz, A. B.; Mutter, M. S.; Inners, R. R.; Almond, H. R., Jr.; Whittle, R. R.; Olofson, R. A. J. Am. Chem. Soc. 1986, 108, 7664. (b) Maryanoff, B. E.; Reitz, A. B.; Duhl-Emswiler, B. A. Ibid. 1985, 107, 217. 215

216

Adrian, J.; Stark, C. B. W. Org. Lett. 2014, 16, 5886.

217

Veerasamy, N.; Carlson, E. C.; Collett, N. D.; Saha, M.; Carter, R. G. J. Org. Chem.2013, 78, 4779.

218

For reactions of this type, see Larock, R. C. Comprehensive Organic Transformations, 2nd ed.; Wiley–VCH: New York, 1999; pp 332–335.

219

Horner, L.; Hoffmann, H.; Wippel, J. H. G.; Klahre, G. Berichte 1959, 92, 2499.

637

12.5 YLIDS

Wadsworth and Emmons220 modified the Horner reaction to use phosphonate ester derivatives (e.g., diethyl benzylphosphonate). Reaction with aldehydes or ketones (e.g., benzaldehyde) in the presence of base gave the olefination product, 1,2-diphenylethene, in 84% yield. These variations have come to be called the Horner-Wadsworth-Emmons modification of the Wittig reaction, or simply the Horner-Wadsworth-Emmons olefination.221 It is sometimes called HornerEmmons olefination. O

Ph

Ph

EtO P EtO Diethyl benzylphosphonate

NaH, PhCHO 60°C

Ph 1,2-Diphenylethene

The major product of olefination with phosphonate carbanions is usually the (E)-isomer.222 A (Z)-selective reaction has been developed that uses NaI and DBU as the base.223 Speziale and Ratts224 suggested that there was an increase in (Z)-alkene with increasing steric bulk of the L and L1 groups in 156 and 159, formed by reaction of the phosphonate ester ylid with a carbonyl compound. O

O

L

S

RO P RO L H

L1

H

O L1

S O−

159

+

−

O

161

O

RO RO

P

O

L

−

RO P RHO

L1

L

S

160

L1

L1 O

−

P

L1

L H

O−

RO RO

S H

RO P HS RO L H 156

S 158

O

L

L1 H 157

S

This equilibrium favors the anti-conformation 156 over 159, and elimination gives more (Z)-product, 158.124 This model assumes that steric hindrance is more important in 156 and 159 than in oxaphosphetanes 157 and 160, respectively, which are required for syn-elimination to the alkene. This equilibrium generally favors 158 over 161 due to the stereochemical preferences in the initially formed ylid products. Bases (e.g., butyllithium, potassium tert-butoxide, or sodium hydride) are usually required to generate the phosphonate carbanion.225 The nucleophilicity of that anion is influenced by the nature of the metal ion. Addition of a metal ligand to the reaction mixture also influences the relative ease of removal of the α-proton. With some metals, a weaker base can be used to generate the carbanion.226 Addition of metal salts to 162 (R ¼ Et) led to formation of carbanion 163, which was stabilized by chelation to the metal.227 Addition of complexing Lewis acids (e.g., LiBr or MgBr2) allowed the ylid to be formed with a base as weak as triethylamine. Magnesium bromide was the most efficient metal additive for producing this product.227 Subsequent reactions were done with cyclohexanone to yield ethyl 2-cyclohexylideneacetate(also with benzaldehyde to yield the ]CHPh derivative), and the best results were obtained in ether solvents or acetonitrile with Li halides, although magnesium bromide was also effective.

220

Wadsworth, W. S., Jr.; Emmons, W. D. J. Am. Chem. Soc. 1961, 83, 1733.

(a) Boutagy, J.; Thomas, R. Chem. Rev. 1974, 74, 87. (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; pp 334, 335.

221

222

Wadsworth, D. H.; Schupp, III, O. E.; Seus, E. J.; Ford, J. A., Jr. J. Org. Chem. 1965, 30, 680.

223

Ando, K.; Oishi, T.; Hirama, M.; Ohno, H.; Ibuka, T. J. Org. Chem. 2000, 65, 4745.

224

Speziale, A. J.; Ratts, K. W. O. J. Am. Chem. Soc. 1963, 85, 2790.

225

Wadsworth, W. S., Jr. Org. React. 1977, 25, 73.

(a) Bottin-Strzalko, T.; Corset, J.; Froment, F.; Ponet, M. J.; Seyden- Penne, J; Simmonin, M. J. J. Org. Chem. 1980, 45, 1270. (b) Blanchette, M. A.; Choy, W.; Davies, J. T.; Essenfeld, A. P.; Masamune, S.; Roush, W. R.; Sakai, T. Tetrahedron Lett. 1984, 25, 2183. 226

227

Rathke, M. W.; Nowak, M. J. Org. Chem. 1985, 50, 2624.

638

12. CARBON-CARBON BOND-FORMING REACTIONS

M

O O

M+

OEt

RO P RO

RO

Base

Cyclohexanone MX, NEt3, THF

O

O P

CO2Et

25°C, 3 h

OEt

RO 162

Ethyl 2-cyclohexylideneacetate

163

Still and Gennari228 showed that octanal reacted with the carbanion derived from 162 to give a mixture of methyl (Z)-dec-2-enoate and methyl (E)-dec-2-enoate. Using KN(TMS)2 (potassium hexamethyldisilazide) as a base (Section 13.2.2) in THF in the presence of 18-crown-6, however, gave an 8:1 mixture of methyl (Z)-dec-2-enoate/methyl (E)-dec-2-enoate in 81% yield when OR ¼ OMe in 162. Changing OR to OCH2CF3 led to isolation of a 12:1 mixture of methyl (Z)-dec-2-enoate/methyl (E)-dec-2-enoate in 90% yield.228 Reaction of 162 (R ¼ CH2CF3) with benzaldehyde, under the same conditions, gave a >50:1 mixture of (Z)/(E) alkenes in >95% yield.228 When Triton B (PhCH2NMe+3 OH) was used in THF with 162 (R ¼ CH2CF3) a 7:1 mixture of methyl (Z)-dec-2-enoate/methyl (E)-dec-2-enoate was obtained (84%), but the use of KOt-Bu in THF gave a 2:5 mixture of methyl (Z)-dec-2-enoate/methyl (E)-dec-2-enoate in 70% yield.128 These studies show that the preference for the (Z)-alkene product is strongly influenced by the reaction conditions and phosphonate carbanion chosen for the reaction. An asymmetric olefination using a benzopyranoisoxazolidine auxiliary was reported, generating alkylidene cyclohexane derivatives with good asymmetric induction.229 Base

C6H13 CHO

C6H13

CO2Me

162

+

Methyl (Z)-dec-2-enoate

Octanal

C6H13 CO2Me Methyl (E)-dec-2-enoate

There are many applications of phosphonate ester methodology in organic synthesis. In a synthesis of sculponeatin N, Zhai and coworkers230 reacted aldehyde 164 with the ylid formed from the phosphonate ester and butyllithium to give 165 in 75% yield. Intramolecular olefination reactions are possible using phosphonate carbanions. This type of cyclization is also possible with Wittig reagents, but it is more facile with the more stable phosphonate carbanions. An example is taken from a synthesis of palmerolide A by De Brabender and coworker’s,231 in which 166 was cyclized to macrolide 167 containing the conjugate ketone moiety in 70% yield from the primary alcohol precursor to 166. O

CHO

SiMe3

P EtO

SiMe3

, BuLi, HMPA

OEt

THF, –78 to 25°C

O

O

O

O 164

165 (75%) O

O P(O)(OMe)2

OHC

OTIPS

K2CO3, PhMe

OTIPS

18-Crown-6, 60°C

O

Me HO

O

MeO2C

O

O

Me HO

MeO2C Me

Me

Me 166

228

Still, W. C.; Gennari, C. Tetrahedron Lett. 1983, 24, 4405.

229

Abiko, A.; Masamune, S. Tetrahedron Lett. 1996, 37, 1077.

230

Pan, Z.; Zheng, C.; Wang, H.; Chen, Y.; Li, Y.; Cheng, B.; Zhai, H. Org. Lett. 2014, 16, 216.

231

Jiang, X.; Liu, B.; Lebreton, S.; De Brabander, J. K. J. Am. Chem. Soc. 2007, 129, 6386.

Me 167

639

12.5 YLIDS

Warren and Buss232 showed that olefination of ketones and aldehydes, or even esters with phosphine oxide carbanions, allowed some control of stereochemistry in the alkene products. Reaction with an aldehyde or ketone generates diastereomeric β-hydroxyphosphine oxide products, which can be isolated and then chromatographically separated. The reaction of ethyldiphenylphosphine oxide with n-butyllithium was followed by addition of 3,4methylenedioxybenzaldehyde (piperonal) to yield a 9:1 mixture of 168/169, the syn- and anti-alcohol, respectively.232 The pure syn-alcohol 168 was isolated in 75% yield by chromatography on silica gel. Subsequent treatment with sodium hydride in DMF gave the (Z)-alkene, α-isosafrole (cis-isosafrole). When ethyldiphenylphosphine oxide was first reacted with butyllithium and then with the ethyl ester of piperonylic acid, the keto-phosphine oxide 170 was produced. Reduction of the ketone with sodium borohydride gave predominantly (6:1) the anti-isomer 171, which gave the (E)-alkene trans-isosafrole upon reaction with sodium hydride. O Me

O 1. C4H9Li

P

Ph

Ph

+

O Me

H Ar

Me

1. NaBH4

Ar

Ph

H

Ph

O

OH

O

H Ar H

P

Ph

2. aq NH4Cl

α-Isosafrole (cis-isosafrole)

H P

169

Ethyl piperonylate

Ph

O

Ph

O

O Me Ph P

O

OH 168 (75%)

CO2Et

2. O

O

H Ar

O

Piperonal 1. C4H9Li

H

Ph

CHO

2. O

Ethyldiphenylphosphine oxide

P

Ph

O

OH 171

170

trans-Isosafrole

An interesting application of this methodology is to extend the chain of an aldehyde by one carbon (RdCHO ! RCH2CHO).233 Indeed, many functional groups can be incorporated into the phosphonate ester, including nitriles, esters, carboxylic acids, and so on. A variation was used by Sorensen and coworker’s234 in a synthesis of FR182877, in which Weinreb amide 172 was treated with LiCH2P(O)(OMe)2 (173) to yield phosphonate ester 174. Weinreb amides have been seen to react with Grignard reagents (Section 11.4), and have been reduced to aldehydes (Section 7.6.1). The reaction with phosphonate anions leads to highly functionalized phosphonate esters. Such phosphonate esters can, of course, be used in subsequent olefination reactions. Me3SiO

O N

OSiMe2t-Bu 172

Me3SiO

O OMe

+

Li

Me

P

OMe OMe

173

THF, –78°C

O

O P

OMe OMe

OSiMe2t-Bu

174

The use of phosphine oxide ylids and phosphonate ester carbanions just described is very similar to Corey’s use of α-lithiophosphonic acid bis(amides) for olefination reactions. Corey showed that the reaction of 175 with butyllithium gave the lithio derivative 176.235 Subsequent reaction with methyl 2,2-dimethylpropanoate (methyl pivaloate) gave the corresponding ketone, 177. Reduction with LiAlH4 and quenching with water led to isolation of alcohol 178. Heating in benzene with silica gel at reflux liberated the alkene (E)-2,2-dimethylundec-3-ene in 71% yield.233 Corey and Kwiatkowski236 also showed that the stereochemistry of the alkene could be controlled. Reaction of 175 with butyllithium and then pivalaldehyde, for example, gave 179. Thermal elimination gave a 3:1 (Z)/(E) mixture of (Z)-2,2-dimethylundec-3-ene and (E)-2,2-dimethylundec-3-ene favoring (Z)-2,2-dimethylundec-3-ene.236 232

Buss, A. D.; Warren, S. J. Chem. Soc. Chem. Commun. 1981, 100.

(a) Earnshaw, C.; Wallis, C. J.; Warren S. J. Chem. Soc. Chem. Commun. 1977, 314. (b) Earnshaw, C.; Wallis, C. J.; Warren, S. T. J. Chem. Soc. Perkin Trans. 1 1979, 3099. (c) Davidson, A. H.; Warren, S. Ibid. 1976, 639. 233

234

Vanderwal, C. D.; Vosburg, D. A.; Welle, S.; Sorensen, E. J. J. Am. Chem. Soc. 2003, 125, 5393.

235

Corey, E. J.; Kwiatkowski, G. T. J. Am. Chem. Soc. 1966, 88, 5652.

236

Corey, E. J.; Kwiatkowski, G. T. J. Am. Chem. Soc. 1966, 88, 5653.

640

12. CARBON-CARBON BOND-FORMING REACTIONS

By comparison, the triphenylphosphonium carbanion derived from 1-bromooctane reacted with pivalaldehyde to yield a 98.5:1.5 mixture of (Z)-2,2-dimethylundec-3-ene/(E)-2,2-dimethylundec-3-ene, which clearly favored the (Z) isomer.236 Good yields of (E)- or (Z)-alkenes were obtained by oxidation of the alcohol intermediate derived from reaction with aldehydes or ketones with manganese dioxide in chloroform. Sodium borohydride was used for reduction of this ketone, as shown above, allowing control of the relative stereochemistry of the intermediate product. O Me2N

C7H15

P

1. t-BuCO2Me 2. H2O

Li

Me2N

O

[H]

P

Me2N

Me2N

176

O

C7H15 Me2N

O

177

C7H15 PhH

H

Me2N

t-Bu

C7H15

P

Reflux

t-Bu

t-Bu

HO

(E )-2,2-Dimethylundec-3-ene

178

(71%) 1. t-BuCHO –78°C, THF

C4H9Li

O Me2N

O

2. H2O

C7H15

P

Me2N

175

C7H15

PhH

Me2N H

Me2N

C7H15

P

Reflux

OH t-Bu

t-Bu (Z )-2,2-Dimethylundec-3-ene

179

Corey and Kwiatkowski237 developed phosphonothioate derivative 180 and phosphonate derivative 181, and showed the corresponding ylids gave good yields of alkene products upon reaction with aldehydes and ketones when X ¼ Cl, C(]O)R, or Ar. When X ¼ H or alkyl in 181 (the ylid is not stabilized), formation of the lithio derivative was difficult and elimination to the alkene was also sluggish. The thio derivative (180), however, reacted with carbonyl derivatives and decomposed to alkene products at temperatures ranging from ambient to 65°C. An example is the reaction of O,O-dimethyl methylphosphonothioate with butyllithium and then 1-bromobutane to yield the alkylated product, 182. This result is unusual since Wittig-type reagents do not undergo alkylation reactions in good yield. Addition of more base to 182 gave a new ylid, and condensation with benzophenone at 25°C gave 1,1-diphenylpent-1-ene. Elimination to the alkene required higher temperatures (65°C) when the carbonyl component was a nonconjugated aldehyde or ketone. S P

MeO

O

X

MeO

MeO

180

181

S MeO

P

S CH3

X

P

MeO

1. BuLi 2. BuBr

MeO

MeO

C4H9

P

O,O-Dimethyl methylphosphonothioate

O

2.

MeO

Ph

1. BuLi

Ph

Ph Ph

1,1-Diphenylpent-1-ene

182

The disconnection for the Wittig- type reactions follow: R R

R2 R1

O R

+ R

X

X R1

Or R2

R

O

+ R

R1

R2

12.5.1.4 Alkenylphosphonium Salts Alkenylphosphonium salts are known, and they react with carbonyl compounds to form heterocyclic compounds. The phosphonium salt triphenyl(vinyl)phosphonium bromide was generated by reaction of 1-phenoxy-2-bromoethane with PPh3, followed by heating.238 Phosphonium salt 183 was formed from the 237

Corey, E. J.; Kwiatkowski, G. T. J. Am. Chem. Soc. 1966, 88, 5654.

238

Schweizer, E. E.; Bach, R. D. Org. Synth. 1968, 48, 129.

641

12.5 YLIDS

corresponding β-chloro-α,β-unsaturated ketone by reaction with PPh3 and triphenyl(vinyl)phosphonium bromide, followed by reaction with aq KBr.239 The most common use of these phosphonium salts is in a reaction with carbonyl compounds that possess a nucleophilic atom at the β- or γ-position. Treatment of 2-hydroxy-1,2-diphenylethan-1-one with sodium hydride generated an alkoxide, and addition to triphenyl(vinyl)phosphonium bromide gave ylid 184 and intramolecular olefination reaction gave 2,3-diphenyl-2,5-dihydrofuran in 50% yield.240 O Ph3P

+

Br

− +

Ph

183

Ph

O

O

2. Triphenyl(vinyl)phosphonium bromide

OH

Ph

PPh3

1. NaH

Ph

R

Ph3P

Triphenyl(vinyl)phosphonium bromide

Ph

2-Hydroxy-1,2-diphenylethan-1-one

O

Ph

O

2,3-Diphenyl-2,5(50%) dihydrofuran

184

α-Thioketones react with vinyl phosphonium salts to give dihydrothiophenes. When 3-mercapto-1;1-phenylpropan2-one reacted with Ph3P]CMe2, for example, the product was 2,2-dimethyl-4-phenyl-2,5-dihydrothiophene.241 In a similar manner, 1H-pyrrole-2-carbaldehyde reacted with vinylphosphonium salts to yield a pyrrolizine derivative (2,3-dihydro-1H-pyrrolizine).242 Ph Me S

N

H

Me

2,2-Dimethyl-4-phenyl2,5-dihydrothiophene

CHO

N

1 H-Pyrrole-2-carbaldehyde

2,3-Dihydro-1H-pyrrolizine

An interesting modification of this reaction generates a normal Wittig reagent by reaction of a vinyl phosphonium ylid with an organocuprate (Section 12.3.1). Addition of dibutylcuprate to triphenyl(vinyl)phosphonium bromide, for example, generated ylid 185. Subsequent reaction with hexanal gave the expected alkene, dodec-6-ene, as a 1:4 mixture of (Z)/(E) isomers.243 +

PPh3 Br–

(C 4H9)2CuLi

+

C4H9

THF, HMPA –50°C

Triphenyl(vinyl)phosphonium bromide

−

185

PPh3

1. C5H11 CHO –50 → –0°C

C4H9

C5H11

2. aq NH4Cl

Dodec-6-ene

12.5.2 Sulfur Ylids 12.5.2.1 Sulfonium and Sulfoxonium Ylids Just as phosphines react with alkyl halides to give the corresponding phosphonium salt, sulfides (thioethers) react with alkyl halides to give the corresponding sulfonium salt.166,244 Ingold and Jessop reported the preparation of 239

Zbiral, E.; Rasberger, M.; Hengstberger, H. Annalen 1965, 725, 22.

(a) Schweizer, E. E.; Liehr, J. C. J. Org. Chem. 1968, 33, 583. (b) Schweizer, E. E.; Creasy, W.; Liehr, J. C.; Jenkins, M. E.; Dalrymple, D. L. Ibid. 1970, 35, 601. (c) Schweizer, E. E. J. Am. Chem. Soc. 1964, 86, 2744.

240

241

McIntosh, J. M.; Goodbrand, H. B. Tetrahedron Lett. 1973, 3157.

242

Schweizer, E. E.; Light, K. K. J. Am. Chem. Soc. 1964, 86, 2963.

243

Just, G.; O’Connor, B. Tetrahedron Lett. 1985, 26, 1799.

244

Reference 165, pp 304–366.

642

12. CARBON-CARBON BOND-FORMING REACTIONS

sulfonium ylid 186 in 1930 (from the fluorene dimethylsulfonium salt 187), but this initially formed ylid rearranged to give 188245in a sulfur analog of the Sommelet-Hauser rearrangement246 (Section 12.5.3.2). It is now known that sulfur ylids are formed by reaction with a strong base247 (e.g., RO-/ROH, NaH/DMSO, RLi, or LiNR2 in THF or ether) with an appropriate sulfonium salt. The sulfonium salt is formed by reaction of a dialkyl sulfide or a diaryl sulfide (R2S) with primary allyl or benzyl halides.248 + CH 3 S

H3C

H3C

+ S

− CH2

H3C S

Base

186

188

187

The reaction of diphenylsulfide and iodoethane gave the corresponding sulfonium salt, ethyldiphenylsulfonium iodide.249 Subsequent reaction with tert-butyllithium generated the ylid (diphenylsulfonium ethylid), which contains the Ph2S+ moiety. This unit is a good leaving group for SN2 displacement by a nucleophile. The gegenion of the sulfonium salt is iodide, which can displace Ph2S. In other words, the reaction that generates the sulfonium salt is reversible.250 For this reason, the use of nonnucleophilic gegenions (e.g., tetrafluoroborate or perchlorate) is usually a necessity to maximize formation of the ylid. When highly electrophilic species (e.g., silver salts) are used, other complicating reactions are possible. Rearrangements can occur by reaction of AgBF4 with the alkyl halide to form the secondary cation (Section 16.2). The straight-chain sulfonium salt is usually the major product.251 Ph

Ph + – I S Ph

EtJI

S Ph

t-BuLi

Ph

THF

+

−

Ph Diphenylsulfonium ethylid

Ethyldiphenylsulfonium iodide

Diphenyl sulfide

S

Dimethyl sulfide is a precursor to sulfur ylids, reacting with iodomethane to yield trimethylsulfonium iodide. When this salt was treated with butyllithium, deprotonation of the hydrogen on the α-carbon led to dimethylsulfonium methylid. Dimethyl sulfoxide also reacts with alkyl halides to give a sulfoxonium salt. When DMSO reacted with iodomethane, trimethylsulfoxonium iodide was formed. Subsequent treatment with a strong base (e.g., butyllithium) generated the corresponding ylid, dimethylsulfoxonium methylid. Me

CH3I

S Me Dimethyl sulfide Me O

Me Dimethyl sulfoxide

+

S Me

I

−

CH3

Trimethylsulfonium iodide CH3I

S

Me

Me

BuLi

+

S

−

CH2

Me Dimethylsulfonium methylid

Me + BuLi H3C S O − Me I Trimethylsulfoxonium iodide

Me + − H2C S O Me Dimethylsulfoxonium methylid

245

(a) Ingold, C. K.; Jessop, J. A. J. Chem. Soc. 1930, 713. (b) Agami, C. Bull. Soc. Chim. Fr. 1965, 1021.

246

(a) Hilbert, G. E.; Pinck, L. A. J. Am. Chem. Soc. 1938, 60, 494; 1946, 68, 751. (b) Pine, S. H. Org. React. 1970, 18, 403.

247

Reference 166, pp 13–23, 29–33.

(a) Corey, E. J.; Chaykovsky, M. J. Am. Chem. Soc. 1965, 87, 1353. (b) Hatch, M. J. J. Org. Chem. 1969, 34, 2133. (c) LaRochelle, R. W.; Trost, B. M.; Kiepski, L. Ibid. 1971, 36, 1126. (d) Hauser, C. R.; Kantor, S. W.; Brasen, W. R. J. Am. Chem. Soc. 1953, 75, 2660. (e) Ratts, K. W.; Yao, A. N. J. Org. Chem. 1966, 31, 1185. (f) Speziale, A. J.; Tung, C. C.; Ratts, K. W.; Yao, A. N. J. Am. Chem. Soc. 1965, 87, 3460. 248

249

Corey, E. J.; Oppolzer, W. J. Am. Chem. Soc. 1964, 86, 1899.

250

(a) Ray, F. E.; Farmer, J. J. Org. Chem. 1943, 8, 391. (b) Ray, F. E.; Levine, I. Ibid. 1938, 2, 267.

251

(a) Trost, B. M.; Bogdanowicz, M. J. J. Am. Chem. Soc. 1973, 95, 5298. (b) Tang, C. S. F.; Rapoport, H. J. Org. Chem. 1973, 38, 2806.

643

12.5 YLIDS

In many reactions, this ylid behaves as a carbanion rather than a phosphonium ylid. Indeed, dimethylsulfonium methylid reacts with epoxides to yield an alcohol. A synthetic example, taken from a synthesis of spirasterellolide A methyl ester by F€ urstner and coworkers,252 reacted an epoxide, (R)-2-(2-((4-methoxybenzyl)oxy)ethyl)oxirane, with dimethylsulfonium methylid, generated in situ, to give an 88% yield of the allylic alcohol, (R)-5-((4-methoxybenzyl) oxy)pent-1-en-3-ol. When dimethylsulfonium methylid or dimethylsulfoxonium methylid reacts with carbonyl derivatives, initial reaction with the carbonyl leads to an alkoxide, and displacement of the sulfonium or sulfoxonium moiety leads to an epoxide. In a synthesis of aspinolide A by Kumar and coworkers,253 Swern oxidation of 5-((4-methoxybenzyl)oxy)pentan-1-ol gave the corresponding aldehyde, and subsequent reaction with dimethylsulfoxonium methylid gave 2-(4-((4-methoxybenzyl)oxy)butyl)oxirane in 70% yield. O

OH

Me3SI, BuLi, THF –10°C to rt

PMBO

PMBO

(R)-2-(2-((4-Methoxybenzyl)oxy)ethyl)oxirane

PMBO

OH

5-((4-Methoxybenzyl)oxy)pentan-1-ol

(R)-5-((4-Methoxybenzyl)(88%) oxy)pent-1-en-3-ol 1. (COCl) 2 , DMSO, CH2Cl2 –78 to –60°C, NEt3

O

2. Me3S=O, NaH DMSO, 60°C

PMBO 2-(4-((4-methoxybenzyl)(70%) oxy)butyl)oxirane

Sulfur ylids normally react with aldehydes or ketones to generate an epoxide, and not an alkene.254 Dimethylsulfoxonium methylid also reacts with carbonyl compounds to give an intermediate (189) that contains a leaving group, (DMSO) and displacement by the adjacent alkoxide also yields an oxirane (1-oxaspiro[2.5]octane).248 An intermediate (e.g., 189) was generated independently and shown to form an epoxide.255 O

O

−

Dimethylsulfoxonium methylid

Cyclohexanone

O

S Me +

O – DMSO

Me

1-Oxaspiro[2.5]octane

189

O

O

O +

4-(tert-Butyl)cyclohexan-1-one

+ Dimethylsulfonium methylid + Dimethylsulfoxonium methylid

190

191 (5 : 1) (0 : 100)

It is known that dimethylsulfonium methylid adds irreversibly to ketones and aldehydes. On the other hand, dimethylsulfoxonium methylid adds reversibly. This difference is reflected in their reactivity with cyclic ketones (e.g., 4-tert-butylcyclohexanone). Dimethylsulfonium methylid reacted with 4-tert-butylcyclohexanone to give the oxirane with an axial exocyclic CdC bond (191), which is less stable than the product with an axial CdO bond (dimethylsulfoxonium methylid).256 When dimethylsulfoxonium methylid reacted with 4-tert-butylcyclohexanone,257 however, the oxirane with the more stable equatorial exocyclic CdC bond (192) was formed. This selectivity is attributed to the Arlt, A.; Benson, S.; Schulthoff, S.; Gabor, B.; F€ urstner, A. Chem. Eur. J. 2013, 19, 3596. For an example taken from a synthesis of umuravumbolide, see Chowdhury, P. S.; Kumar, P. Eur. J. Org. Chem. 2013, 4586.

252

253

Chowdhury, P. S.; Gupta, P.; Kumar, P. Tetrahedron Lett. 2009, 50, 7018.

254

For reactions of this type, see Larock, R. C. Comprehensive Organic Transformations, 2nd ed.; Wiley–VCH: New York, 1999; pp 944–946.

255

Johnson, C. R.; Schroeck, C. W.; Shanklin, J. R. J. Am. Chem. Soc. 1973, 95, 7424.

(a) Carlson, R. G.; Behn, N. S. J. Org. Chem. 1967, 32, 1363. (b) Johnson, C. R.; Katekar, R. A. J. Am. Chem. Soc. 1970, 92, 5753. (c) Johnson, C. R. Acc. Chem. Res. 1973, 6, 341.

256

257

Reference 166, p 40.

644

12. CARBON-CARBON BOND-FORMING REACTIONS

reversibility of the addition of dimethylsulfoxonium methylid to the ketone leading to the more stable oxirane, whereas dimethylsulfonium methylid formed the intermediate salt irreversibly in what is essentially a kinetic process. Another difference in the reactivity of dimethylsulfonium methylid and dimethylsulfoxonium methylid is reflected in their reactions with α,β-unsaturated ketones. Carboethoxy dimethylsulfonium methylid reacted with cyclohex-2en-1-one to yield the usual alkoxy sulfonium salt 192. Intramolecular displacement of dimethyl sulfide led to the expected oxirane (ethyl 1-oxaspiro[2.5]oct-4-ene-2-carboxylate). This product represents a generalized reaction, and dialkylsulfonium ylids react via 1,2-addition with displacement of the sulfide to yield an allylic oxirane.255 When carboethoxy dimethylsulfoxonium ylid reacted with cyclohex-2-en-1-one, however, the 1,2-addition adduct was formed reversibly and 1,4-addition becomes competitive. The conjugate addition product 193 is an enolate anion (Sections 13.2 and 13.7) and intramolecular displacement of DMSO generated a cyclopropane ring, forming ethyl 2-oxobicyclo[4.1.0] heptane-7-carboxylate.257 −

O −

+

Me2S

+

O

SMe2 CO2Et

CHCO2Et

CO2Et

O –Me2S

Carboethoxy dimethylsulfonium methylid

Cyclohex-2-en-1-one

192 O−

O

O Me2S

+

Ethyl 1-oxaspiro[2.5]oct4-ene-2-carboxylate

O

−

CHCO2Et

SMe2

Carboethoxy dimethylsulfoxonium ylid

Cyclohex-2-en-1-one

+

O –DMSO

CO2Et CO2Et Ethyl 2-oxobicyclo[4.1.0]heptane-7-carboxylate

193

The latter reaction is more facile when the conjugate system has as structure that makes it more amenable to Michael addition. The acrylate derivative ethyl 3-methylbut-2-enoate, for example, gave only a 9% yield of ethyl 2,2dimethylcyclopropane-1-carboxylate after reaction for 1 h with dimethylsulfoxonium methylid, whereas diethyl 2-(propan-2-ylidene)malonate gave a 91% yield of diethyl 2,2-dimethylcyclopropane-1,1-dicarboxylate under the same reaction conditions.258 The reaction is synthetically useful despite the limitations, as shown by the conversion of 194 to 195 in >80% yield in the Zhou, and coworkers259 synthesis of α-kainic acid. Dimethylsulfoxonium methylid

CO2Et

Ethyl 3-methylbut-2-enoate

Ethyl 2,2-dimethylcyclopropane-1-carboxylate Dimethylsulfoxonium methylid

CO2Et

NHTs Ph

CO2Et

Diethyl 2,2-dimethylcyclopropane-1,1-dicarboxylate Me3SO+ I–, NaH

O

258

Landon, S. R.; Punja, N. J. Chem. Soc. C. 1967, 2495.

259

Luo, Z.; Zhou, B.; Li, Y. Org. Lett. 2012, 14, 2540.

(91%)

NHTs TBSO

Ph

DMSO, rt

194

(91%)

CO2Et

CO2Et Diethyl 2-(propan-2-ylidene)malonate

TBSO

CO2Et

195 (>80%)

O

645

12.5 YLIDS

Disconnections for reactions of these sulfur ylids follow: R1 O

R

O

R

+ R1

R2

R2

X

O

R1

R

+

R3

R3

R2

O

R1

X R

R2

12.5.2.2 Diphenylcyclopropylsulfonium Derivatives Trost and Bogdanowicz,260 showed that diphenylcyclopropylsulfonium ylids (e.g., 196) reacted with ketones and aldehydes in a manner similar to other sulfur ylids. Attack at the carbonyl generated the expected intermediate 197, and intramolecular displacement of diphenyl sulfide by alkoxide gave the oxirane 198. This oxirane, however, is an oxaspiropentane derivative and is susceptible to several additional synthetic manipulations.260,261a O

+

−

Cyclohexanone OBn

O

+

−

+

SPh2

SPh2

196

O

O –SPh2

HBF4

197

Spiro[3.5]nonan-1-one

198 OBn

OBn

OBn PhSeSePh NaBH4

196, KOH

N3

DMSO, rt

N3

199

200

OBn

OBn

OBn

mcpba , Py Hexane–CH2Cl2

N3 PhSe

EtOH, rt

O

O

OBn

OMe

OMe

OMe

OH

N3

–30°C to rt

OBn

+ OMe

O 201

OBn

202 (57%)

N3 O

OMe

203 (14%)

A useful transformation is the conversion of the oxaspiropentane moiety to a cyclobutanone derivative (spiro[3.5] nonan-1-one) by treatment with Lewis acids (e.g., HBF4, LiClO4, or Eu(fod)3, where fod ¼ tris(6,6,7,7,8,8,8)-heptafluoro-2,2-dimethyl-3,5-octanedionate).262 Such products arise via a cationic rearrangement of an intermediate oxonium ion (Section 16.2), formed by protonation of (or coordination to) the oxiranyl oxygen. Ring opening generated a cyclopropylcarbinyl carbocation, which rearranged to the cyclobutyl oxocarbenium ion (Section 16.2.4), and this intermediate lost a proton to yield the cyclobutanone product. When better coordinating Lewis acids are used [LiClO4 or Eu(fod)3], there is a chelation effect that further stabilizes the intermediate cation. An alternative conversion to the cyclobutanone moiety was reported by Tu and coworkers263 in a synthesis of ()-FR901483. Initial reaction of 199 with 196 gave spirocycle 200, and treatment with diphenyldiselenide and NaBH4 led to phenylselenyl-alcohol 201. Subsequent oxidation led to a mixture of isomeric cyclobutanones 202 (57% yield) and 203 (14% yield).263 An alternative reaction pathway is available when a carbon adjacent to the oxiranyl moiety possesses a hydrogen atom. Reaction of 2-hexyl-2-methyl-1-oxaspiro[2.2]pentane with lithium diisopropylamide (Section 13.2.2) removed the hydrogen on the carbon α to the oxiranyl carbon and formed 204 via opening of the oxaspiro moiety.

260

(a) Trost, B. M.; Bogdanowicz, M. J. J. Am. Chem. Soc. 1971, 93, 3773. (b) Trost, B. M.; LaRochelle, R.; Bogdanowicz, M. J. Tetrahedron Lett. 1970, 3449.

(a) Trost, B. M.; Mao, M. K.-T.; Balkovec, J. M.; Buhlmayer, P. J. Am. Chem. Soc. 1986, 108, 4965. (b) Trost, B. M.; Balkovec, J. M.; Mao, M. K.-T. Ibid. 1986, 108, 4974. 261

262

Trost, B. M.; Keeley, D.; Bogdanowicz, M. J. J. Am. Chem. Soc. 1973, 95, 3068.

263

Ma, A.-J.; Tu, Y.-Q.; Peng, J.-B.; Dou, Q.-Y.; Hou, S.-H.; Zhang, F.-M.; Wang, S.-H. Org. Lett. 2012, 14, 3604.

646

12. CARBON-CARBON BOND-FORMING REACTIONS

When this reaction was done in the presence of chlorotrimethylsilane, the alkoxide was trapped as the trimethylsilyl ether 205 (Section 5.3.1.2). Subsequent heating (flash vacuum pyrolysis at 300–500°C) led to a rearrangement that generated 206.264 −

O

OSiMe3

O Me3SiCl

LDA

C6H13

300–500°C

C6H13

2-Hexyl-2-methyl-1oxaspiro[2.2]pentane

C6H13

204

205

O

OH

OSiMe3 aq HCl

C6H13

C6H13

C6H13

2-Hexylcyclopentan-1-one

2-Hexylcyclopent-1-en-1-ol

206

This so-called vinyl cyclopropane rearrangement is a sigmatropic rearrangement that will be discussed in Section 15.5.2. Silyl enol ether (206) is simply a protected enolate (Sections 13.3.2 and 13.4.3) and aqueous hydrolysis liberated the unstable enol, 2-hexylcyclopent-1-en-1-ol (Sections 5.3.1.2 and 2.5.2). Tautomerization favored the final product, 2-hexylcyclopentan-1-one.264 Trost et al.261 applied this methodology to a synthesis of allamandin. The disconnections for the cyclopropyl sulfide reactions follow: O

O

O

+

R R

R

R

O

R

X

R

R

R

12.5.3 Nitrogen Ylids 12.5.3.1 Formation of Nitrogen Ylids Nitrogen ylids constitute an interesting class of reactive compounds.167,265 The first report of a nitrogen ylid was probably by Kr€ ohnke266 in 1935, who prepared the resonance stabilized pyridinium ylid 207 by reaction of pyridinium salt 208 with potassium carbonate (K2CO3). Wittig and Wetterling267 found that tetramethylammonium bromide reacted with phenyllithium to give the corresponding ylid (209). The ylid was shown to be the lithium bromide complex. Complexation with added or generated metal salts or other reagents is very important for the stability of nitrogen ylids. Ylid 209 decomposed to carbene (CH2:) (see Section 17.9) in solvents that strongly coordinate LiBr, and this carbene polymerized to polymethylene under the reaction conditions. In ether (ambient temperatures for 90 h), reaction of tetramethylammonium bromide and phenyllithium gave 20% trimethylamine and 12% polymethylene. In DME, 209 formed a strong complex with LiBr and polymethylene was formed in 74% yield. This result suggested that the free (uncomplexed) ylid was rather unstable. O N

Ph

+

O−

O

K2CO3

N

+ −

Ph

Ph

+

208

207 Me Br − + Me N CH3 Me

N

PhLi

Tetramethylammonium bromide

264

Trost, B. M.; Bogdanowicz, M. J. J. Am. Chem. Soc. 1973, 95, 289.

265

Reference 167, pp 251–283.

266

Kr€ ohnke, F. Berichte 1935, 68, 1177.

267

Wittig, G.; Wetterling, M. Annalen 1947, 557, 193.

Me + − Me N CH⋅LiBr Me 209

+

PhH

647

12.5 YLIDS

Wittig and Rieber268 showed that tetramethylammonium bromide reacted with 2 equiv of phenyllithium to form a species (210 and/or 211) that reacted with benzophenone to give 18% of monosubstituted product (212) and 20% of the bis-adduct (213) after hydrolysis. Although nitrogen ylids behave as carbanions, alkylation of 210 in THF is difficult because the carbanion is also basic, which can induce an elimination reaction when mixed with an alkyl halide via loss of trimethylamine. Reaction of 210 with bromocyclohexane, for example, gave a 92% yield of cyclohexene.269 Me + − Me N CH2⋅LiBr Me 210

O 1. Ph

PhLi

Me + Me N CH3 Me Br −

+ Me

Tetramethylammonium bromide

Me

2. H3O

+

Me + Me N Me

OH Ph Ph

212 (18%)

Ph +

Me

−

N CH2 Li

+

HO

N

Me

CH − 2

Ph Ph OH Ph Ph

213 (20%)

211

Ylid 210 is a trimethylammonium methylid, is not complexed, and behaves as a carbanion if the substrate does not possess a leaving group. Reaction with either benzonitrile or ethyl benzoate gave the ammonium ketone 214.269 Similar reaction with benzoyl chloride gave a mixture of the usual ketone product (214) and a secondary product resulting from O-alkylation of the enol form of the acid chloride. Nitrogen ylids behave as both a stronger base and a stronger carbanion than any of the phosphorus or sulfur ylids previously encountered. 1. PhCN, Ether

Me + − Me N CH2 Me

Me + Me N Me

2. HBr 1. PhCO2Et, Ether 2. HBr

Trimethylammonium methylid

O Ph

214

Pyridinium ylids [e.g., those discovered by Krohnke (see 207)] and simple N-alkyl derivatives generally undergo nucleophilic reactions as carbanions. Pyridinium ylids react with aldehydes in a Knoevenagel reaction (Section 13.4.2.3) rather than as an ylid. Carbanion 215, for example, reacted with 4-nitrobenzaldehyde to yield 216 after elimination of water from the initial alcohol product.267,268a,270 Pyridinium ylids271 undergo Michael addition on reaction with conjugated carbonyl systems. Another class of nitrogen ylids is azomethine ylids, which are useful synthetic intermediates and often formed by thermal ring opening of aziridine derivatives.272 The main synthetic use of azomethine ylids is in [3+2]-cycloadditions and this will be discussed in Section 15.4.6.273 CN

+

N

O2N

CHO

+

CN

N

−

Py

215

NO2 216

268

(a) Wittig, G.; Rieber, M. Annalen 1949, 562, 177. (b) Wittig, G.; Polster, R. Ibid. 1956, 599, 1.

269

(a) Weygand, F.; Daniel, H.; Schroll, A. Berichte 1964, 97, 1217. (b) Weygand, F.; Daniel, H. Ibid. 1961, 94, 3147.

270

Also see (a) Kr€ ohnke, F. Chem. Ber. 1951, 84, 388. (b) Idem Ibid. 1950, 83, 253.

(a) Zecher, W.; Kr€ ohnke, F. Berichte 1961, 94, 690. (b) Kr€ ohnke, F.; Zecher, W.; Curtze, J.; Drechsler, D.; Pfleghar, K.; Schnalke, K. E.; Weis, W. Angew. Chem. 1962, 74, 811.

271

272

Huisgen, R.; Scheer, W.; M€ader, H.; Brunn, E. Angew. Chem. Int. Ed. Engl. 1969, 8, 604.

(a) Lown, J. W. Recent Chem. Progr. 1971, 32, 51 (Chem. Abstr. 1972, 76, 3599 g). (b) Kellogg, R. M.; Tetrahedron 1976, 32, 2165. (c) Huisgen, R. J. Org. Chem. 1976, 41, 403. (d) Hermann, H.; Huisgen, R.; M€ader, H. J. Am. Chem. Soc. 1971, 93, 1779. (e) Fleming, I. Frontier Orbitals and Organic Chemical Reactions; Wiley–Interscience: London, 1976; pp 109, 110, 148–160. 273

648

12. CARBON-CARBON BOND-FORMING REACTIONS

12.5.3.2 The Stevens’ Rearrangement and the Sommelet Rearrangement The intermediacy of nitrogen ylids has been suggested for two classical reactions, the Stevens’ rearrangement,246c,274 and the Sommelet-Hauser rearrangement (sometimes called the Sommelet rearrangement).275 Stevens found that treatment of phenacylbenzyldimethylammonium bromide (217) with aqueous hydroxide gave an amino ketone,2-(dimethylamino)-1,3-diphenylpropan-1-one. The reaction probably proceeds via hydrogen abstraction from the ammonium salt (217) to give the ylid 218. A N!C 1,2-benzyl shift occurs via a carbanion-iminium salt complex (219) to yield the amino-ketone.274e The mechanism276involves a [2,3]-sigmatropic rearrangement of the ylid, followed by proton migration to restore aromaticity. O

aq NaOH

Me +

N Me

Ph

Ph

Ph

217

O

Me

O −

N+ Me

O Ph

Ph

−

H2C Ph

218

Me N+ Me 219

NMe2

Ph

Ph 2-(Dimethylamino)-1,3diphenylpropan-1-one

The preparation of complex alkaloids is an important synthetic application of the Stevens’ rearrangement. The reaction of bis(α,α-o-xylylene) ammonium bromide (220) with phenyllithium led to the α-lithio compound 221. This nitrogen ylid was unstable and rearranged with loss of LiBr, to 222.277 Li +

N

PhLi

Br − 220

+N

N

Br − 221

222

Formation of 1-(2-benzylphenyl)-N,N-dimethylmethanamine from benzhydryltrimethylammonium bromide (223),275 by heating to 180°C with concentrated hydroxide, illustrates the Sommelet rearrangement. Initial deprotonation probably occurred at the benzylic site, but equilibrium conditions generated ylid 224. Nucleophilic attack at the proximal benzene ring gave 225 via cleavage of the CdN bond, and subsequent aromatization gave the final product 1-(2benzylphenyl)-N,N-dimethylmethanamine. Evidence suggests the mechanism is a [1,2]-shift of the ylid via a caged radical pair intermediate (see Section 17.2).276 Steric effects appear to be less significant than electronic effects for ylid stability.278 The so-called ortho-substitution rearrangement279 involves reaction of the benzyltrimethylammonium salt with NaNH2 to yield 2-(dimethylaminomethyl)toluene. This rearrangement is either the same reaction as the Sommelet rearrangement, or it is closely related.279 Hauser and Jones280 showed that this rearrangement could be used for a ring expansion, where reaction of dimethylammonium salt 226 with sodium amide gave an 83% yield of 227. The Sommelet and Stevens rearrangements compete with each other in some cases.281

(a) Stevens, T. S.; Creighton, E. M.; Gordon, A. B.; MacNicol, M. J. Chem. Soc. 1928, 3193. (b) Stevens, T. S. Ibid. 1930, 2107. (c) Thomson, T.; Stevens, T. S. J. Chem. Soc. 1932, 55. (d) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-89. (e) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; pp 618, 619. 274

(a) Sommelet, M. Compt. Rend. 1937, 205, 56. (b) Hauser, C. R.; van Eenam, D. N. J. Am. Chem. Soc. 1956, 78, 5698. (c) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-88. (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; pp 608, 609. 275

276 (a) Tanaka, T.; Shirai, N.; Sugimori, J.; Sato, Y. J. Org. Chem. 1992, 57, 5034. (b) Lepley, A. R.; Giumanini, A. G. In Mechanism of Molecular Migrations; Thyagarajan, B. S., Ed., Wiley–Interscience: New York, 1971; Vol. 3, p 297. 277

Wittig, G.; Tenhaeff, H.; Schoch, W.; Koenig, G. Annalen 1951, 572, 1.

278

Heard, G. L.; Yates, B. F. Aust. J. Chem. 1994, 47, 1685.

279

(a) Kantor, S. W.; Hauser, C. R. J. Am. Chem. Soc. 1951, 73, 4122. (b) Puterbaugh, W. H.; Hauser, C. R. Ibid. 1964, 86, 1105.

280

Jones, G. C.; Hauser, C. R. J. Org. Chem. 1962, 27, 3572.

281

Pine, S. H.; Munemo, E. M.; Phillips, T. R.; Bartolini, G.; Cotton, W. D.; Andrews, G. C. J. Org. Chem. 1971, 36, 984.

649

12.6 TRANSITION METAL OLEFINATION REAGENTS

Br − +

− OH

CH3 N

Ph

CH3

+

180°C

Ph

CH3

N

CH3

H3C

CH3 CH3

−

NH3

+

N

Ph 225

1-(2-Benzylphenyl)- N, N-dimethylmethanamine

H3C + CH3 N − CH2

CH3 CH3

CH3

CH3

Ph

224

CH3

N

CH3

CH3

223

+

CH2 N

CH3

N

−

N(CH3)2

N(CH3)2 CH3

NaNH2

Benzyltrimethylammonium

2-(Dimethylaminomethyl)toluene

Note that a nitrogen ylid has been invoked in the Hofmann elimination reaction (Section 3.7.1).282 This ylid mechanism is highly questionable for many substrates283 that proceed via β-elimination with a coordinated hydroxide. In those cases, removal of the hydrogen directly by hydroxide appears more likely. The ylid mechanism probably operates when phenyllithium is used as a base284 and when there is steric inhibition of the usual β-elimination process.285 NaNH2, NH3

Me2N

+

N

X−

Me

226

227

The useful disconnections of nitrogen ylids follow:

R3N

+

O

+

R3N Me + EtO

R1

O

O R1

O

R1

R

R NMe2

Me N Me R1

+

+

NMe2 R

Me

N Me

R

12.6 TRANSITION METAL OLEFINATION REAGENTS Several organometallic derivatives give Wittig-type olefination reactions, and Ti derivatives are the most commonly used.286 Although most organometallic compounds are discussed in Chapter 18, the reagents that give olefination reactions will be discussed here so they can be compared with Wittig reactions. The initial reaction with the carbonyl gives a transient metalated alcohol that leads to an alkene.

282

Cope, A. C.; Ciganek, E.; LeBel, N. A. J. Am. Chem. Soc. 1959, 81, 2799.

283

Reference 167, pp 277–281.

284

Wittig, G.; Polster, R. Annalen 1956, 599, 13.

285

Cope, A. C.; Mehta, A. S. J. Am. Chem. Soc. 1963, 85, 1949.

286

For reactions of this type, see Larock, R. C. Comprehensive Organic Transformations, 2nd ed.; Wiley–VCH: New York, 1999; pp 341–350.

650

12. CARBON-CARBON BOND-FORMING REACTIONS

H

H

Cp

C

CH3

Ti

L

Al

Cp

CH3

Cl

(Cp2TiCH2⋅AlCl Me2)

228

An example is the Tebbe reagent, which exists as a bridged methylene species (see 228) where Cp is cyclopentadienyl,287 and the reaction with carbonyl compounds that leads to alkene derivatives is called Tebbe olefination. The aluminum species can be varied to include chloride (AlCl3), triperdeuteromethyl aluminum [Al(CD3)3], or trimethylsilyl [AlCl(CH2SiMe3)2]. The most common reagent is the aluminum dimethylaluminum chloride compound shown. The Tebbe reagent reacts with ketones or aldehydes to give an alkene, analogous to the Wittig reagent. In a synthesis of ()-nakadomarin A, Funk and Nilson288 converted aldehyde 229 to alkene 230 in >70% yield using the Tebbe reagent in THF. The Tebbe reagent is quite useful in that it reacts with the carbonyl of esters or lactones to give vinyl ethers, in contrast to common Wittig reagents. For example, in a synthesis of neovibsanin B, Imagawa, et al.289 converted lactone 231 to vinyl ether 232 in >90% yield. The Tebbe reagent also reacts with conjugated esters via 1,2addition to give the vinyl ether.287a O

O Cp2TiCH2⋅AlClMe2

H O H

H N Boc

O H

H H N O Boc 230 (>70%)

229

H TBSO

H O

O O

O Cp2TiCH2⋅AlClMe2

TBSO

O

Py, THF/Toluene rt, 1 h

231

232 (>90%)

Other Ti based olefination reagents have been developed. Eisch and Piotrowski290 and used a Zn derivative of the Tebbe reagent (see 233) in a reaction with benzophenone to yield 1,1-diphenylethene in 78% yield. Similarly, Clawson, et al291 used 234 in olefination reactions with ketones and aldehydes. Alkoxy-titanium reagents (e.g., 235) have been employed, as in the conversion of cyclohexanecarbaldehyde to 1-cyclohexylbuta-(1Z),3-diene in 86% yield [96:4 (Z)/(E)].292

(a) Tebbe, F. N.; Parshall, G. W.; Reddy, G. S. J. Am. Chem. Soc. 1978, 100, 3611. (b) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-93. (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; p 856. 287

288

Nilson, M. G.; Funk, R. L. Org. Lett. 2010, 12, 4912.

289

Imagawa, H.; Saijo, H.; Kurisaki, T.; Yamamoto, H.; Kubo, M.; Fukuyama, Y.; Nishizawa, M. Org. Lett. 2009, 11, 1253.

290

Eisch, J. J.; Piotrowski, A. Tetrahedron Lett. 1983, 24, 2043.

291

Clawson, L.; Buchwald, S. L.; Grubbs, R. H. Tetrahedron Lett. 1984, 25, 5733.

292

Ukai, J.; Ikeda, Y.; Ikeda, N.; Yamamoto, V. Tetrahedron Lett. 1983, 24, 4029.

651

12.6 TRANSITION METAL OLEFINATION REAGENTS

Cp Cp2TiCH2ZnI2

Ti Cp

233

234

Oi-Pr

+

CHO

Ti Oi-Pr Oi-Pr

Cyclohexanecarbaldehyde

1. –78 → 0°C 2. MeI, 0°C

235

1-Cyclohexylbuta-1(Z ),3-diene (86%)

Petasis and Bzowej293 developed Cp2TiMe2 (236), which is a very useful alternative to the Tebbe reagent and is called the Petasis reagent. The Petasis reagent is prepared by reaction of methyllithium with titanocene dichloride (Cp2TiCl2)], which avoids some of the difficulties of 228 (high cost, long preparation times, short shelf-lives, extreme sensitivity to air and water, and residual aluminum reagents). Aldehydes and ketones react with the Petasis reagent to give an alkene,294 and esters react with a vinyl ether, illustrated by the conversion of 237 to 238 in 92% yield, taken from Totah and Lam’s295 synthetic studies toward of spirastrellolide A. Petasis and Bzowej296 proposed that olefination proceeds via methyl transfer from an intermediate (e.g., 239).

+ O

Ti Me

Cp2TiMe2

O O

Me

O

O

O

TBSO 236

O

TBSO 238 (92%)

237

Hughes et al297 proposed that olefination of esters proceeds via a Ti carbene (e.g., 240), and an oxatitanacycle (241). This latter proposal has independent support, based on mass spectrometry evidence provided by Pilli and coworkers.298 O R1

Ti R2

Me L

Ti

Me Me O R2 1 R 239

Me O

236

TiKCH2

R1

R2

240

Ti CH2 O R2 1 R 241

In the absence of an aldehyde or ketone moiety, an ester299 or even an amide moiety can react. An example of this latter transformation is the conversion of the lactam unit in 242 to the enamine moiety in 243 (77% yield) in a synthesis of gelsemoxonine by Carareira and Diethelm.300 A Petasis-like reagent also has been reported that yields the same reaction: CH2(ZnI)2, TiCl2, and TMEDA in THF.301 Oshima and coworkers302 developed a TiCl4 reagent (Zn-CH2Br2-TiCl4) that gave olefination. It was electrophilic and 293

Petasis N. A.; Bzowej, E. I. J. Am. Chem. Soc. 1990, 112, 6392.

294

For an example taken from a synthesis of bryostatin and analogs, see Trost B. M.; Yang, H.; Dong, G. Chem. Eur. J. 2011, 17, 9789.

295

Lam, T.; Totah, N. I. Tetrahedron Lett. 2015, 56, 3349.

296

Petasis, N. A.; Bzowej, E. J. J. Org. Chem. 1992, 57, 1327.

297

Hughes, D. L.; Payack, J. F.; Cai, D.; Verhoeven, T. R.; Reider, P. J. Organometallics, 1996, 15, 663.

298

Meurer, E. C.; Santos, L. S.; Pilli, R. A.; Eberlin, M. N. Org. Lett. 2003, 5, 1391.

299

For an example in a synthesis of (–)-kendomycin, see Smith, III, A. B.; Mesaros, E. F.; Meyer, E. A. J. Am. Chem. Soc. 2006, 128, 5292.

300

Diethelm, S.; Carreira, E. M. J. Am. Chem. Soc. 2013, 135, 8500.

301

Matsubara, S.; Ukai, K.; Mizuno, T.; Utimoto, K. Chem. Lett. 1999, 825.

302

Takai, K.; Hotta, Y.; Oshima, K.; Nozaki, H. Tetrahedron Lett. 1978, 2417.

652

12. CARBON-CARBON BOND-FORMING REACTIONS

generally suppressed the tendency of a ketone substrate to enolize. Lombardo303 developed a more active form of this reagent by dropwise addition of TiCl4 to a stirred suspension of Zn dust in CH2Br2 and THF at 40°C, followed by warming the mixture to 5°C and stirring for 3 days. Lombardo303 used this reagent in synthetic studies of C20 gibberellins (GA38). Mehta and Islam304 used Lombardo’s reagent to convert 244 to 245, in 71% yield, in a synthesis of ottelione. Lombardo’s modified reagent is particularly important when the ketone or aldehyde to be used in the olefination reaction is sensitive to basic reagents. Another variation of this reagent uses CH2Br2, Zn, TiCl4, and a catalytic amount of PbCl2.305 O

O

BocO

CO2Et

BocO Py, 70°C, 8 h

O

N Boc

CO2Et

Cp2TiMe2, Toluene

N Boc

242

243 (77%) H H

OH

H H

ZnJTiCl4JCH2Br2

OH

CH2Cl2, 0°C

O 244

245 (71%)

The Nysted reagent (246)306 is commercially available, and it reacts with aldehydes or ketones in the presence of BF3• etherate to yield an alkene. Reaction with (S)-2-phenylpropanal, for example, gave an 82% yield of (R)but-3-en-2-ylbenzene.307 Other metal-ylid type reagents have been developed, including a chromium-based reagent. The Takai reaction308 uses an alkyl di- or triiodide in the presence of CrCl2 to generate an alkene moiety. An example is taken from Menche and coworker’s309 synthesis of rhizopodin, in which the alcohol moiety in 247 was oxidized to the corresponding aldehyde with the Dess-Martin periodinane (Section 6.2.4), and subsequent reaction with CrCl2 and CHI3 gave vinyl iodide 248 in 76% yield (for the two steps) as a 7:1 (E)/(Z) mixture.309 Br

O Zn

Zn

Br

+

Me

CHO

Zn

Ph

246

(S)-2-Phenylpropanal

OMe OTBS

2. CHI3, CrCl2

247

Me Ph

(R)-But-3-en-2-ylbenzene (82%)

OMe OTBS

1. Dess–Martin periodinane

CO2tBu

HO

BF3•OEt2, THF, 0°C

CO2tBu

I 248 (76%)

303

(a) Lombardo, L. Tetrahedron Lett. 1982, 23, 4293. (b) Lombardo, L. Org. Synth. Coll. 1993, 8, 386.

304

Mehta, G.; Islam, K. Angew. Chem. Int. Ed. 2002, 41, 2396.

(a) Takai, K.; Kakiuchi, T.; Kataoka, Y.; Utimoto, K. J. Org. Chem. 1994, 59, 2668. (b) Takai, K.; Kataoka, Y.; Miyai, J.; Okazoe, T.; Oshima, K.; Utimoto, K. Org. Synth. 1995, 73, 73. For a synthetic example taken from a synthesis of ()-periplanone B, see Hodgson, D. M.; Foley, A. M.; Boulton, L. T.; Lovell, P. J.; Maw, G. N. J. Chem. Soc. Perkin Trans. 1 1999, 2911. 305

306 (a) Nysted, L. N. US Patent 3 865 848, 1975 (Chem. Abstr. 1975, 83, 10406q). (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; p 826. 307

Matsubara, S.; Sugihara, M.; Utimoto, K. Synlett 1998, 313.

(a) Takai, K.; Nitta, K.; Utimoto, K. J. Am. Chem. Soc. 1986, 108, 7408. (b) Okazoe, T.; Takai, K.; Utimoto, K. J. Am. Chem. Soc. 1987, 109, 951. (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; pp 640, –641.

308

309

Kretschmer, M.; Dieckmann, M.; Li, P.; Rudolph, S.; Herkommer, D.; Troendlin, J.; Menche, D. Chem. Eur. J. 2013, 19, 15993.

653

12.7 SILANE REAGENTS

Other organometallic reagents have been reported to give Wittig-type olefination reactions with ketones and aldehydes. Examples are 249,310 250,311 251,312 252,313 and 253.314 This list is probably not representative, but it illustrates some of the reagent types that have appeared, and continue to appear. I

PPh3 (MeS) 2BKCH2

Cp Zr Cp

Cl3MoKCH2 250

249

O

W

C H 251

252

I

+

t-Bu

−

Me2TeJCHCO2Et

O t-Bu

253

The disconnections for these reagents are identical to those in Section 12.5.1 for the Wittig reagents.

12.7 SILANE REAGENTS 12.7.1 Allylsilane Reagents An aldehyde can be coupled with an allylsilane in the presence of a Lewis acid. Sakurai and Hosomi315 had previously observed the coupling of allylsilanes to aldehydes in the presence of TiCl4. Reetz et al.316 used TiCl4, SnCl4, BF3, or AlCl3 in reactions of allyltrimethylsilane, and observed good diastereoselectivity. In this study, the solubility of the Lewis acid, and the temperature of the reaction the Lewis acid employed were important factors for both reactivity and selectivity. A high degree of stereoselectivity is possible, if a chiral catalyst is used. In a synthesis of fostriecin, Shibasaki and coworkers317 reacted conjugated aldehyde 254 with allyltrimethoxysilane in the presence of AgF and (R)-p-tolylBINAP, and obtained an 80% yield of 255 (dr ¼ 28:1). OTIPS

20% AgF, MeOH 20% (R)- p-tolyl-BINAP

OHC

OTIPS HO

Me

Si(OMe)3

OMOM

Me

254

OMOM

255 (80%)

Heathcock and Kiyooka examined a similar coupling with aldehydes having a stereogenic α-carbon, as in (R)-2(benzyloxy)propanal, rather than the β-carbon. Upon reaction with allyltrimethylsilane, (R)-2-(benzyloxy)propanal gave a mixture of(2R,3S)-2-(benzyloxy)hex-5-en-3-ol and (2R,3R)-2-(benzyloxy)hex-5-en-3-ol.318 OBn

OBn Me

OBn

SiMe3

CHO

(R)-2-(Benzyloxy)propanal O

+

Me

Lewis acid

OH

(2R,3S)-2-(Benzyloxy)hex-5-en-3-ol

(2R,3 R)-2-(Benzyloxy)hex-5-en-3-ol O

OAc

AcO

SiMe3

O

O

Me

OH

0.3 TMSOTf 0°C

256

AcO O

O 257

310

Pelter, A.; Singaram, B.; Wilson, J. W. Tetrahedron Lett. 1983, 24, 635.

311

Kauffmann, T.; Ennen, B.; Sander, J.; Wieschollek, R. Angew. Chem. Int. Ed. 1983, 22, 244.

312

Clift, S. M.; Schwartz, J. J. Am. Chem. Soc. 1984, 106, 8300.

313

(a) Aguero, A.; Kress, J.; Osborn, J. A. J. Chem. Soc. Chem. Commun. 1986, 531. (b) Freudenberger, J. H.; Shrock, R. R. Organometallics 1986, 5, 398.

314

Osuka, A.; Mori, Y.; Shimizu, H.; Suzuki, H. Tetrahedron Lett. 1983, 24, 2599.

315

(a) Hosomi, A.; Sakurai, M. Tetrahedron Lett. 1976, 1295; also see (b) Trost, B. M.; Coppola, B. P. J. Am. Chem. Soc. 1982, 104, 6879.

316

Reetz, M. T.; Kesseler, K.; Jung, A. Tetrahedron Lett. 1984, 25, 729.

317

Fujii, K.; Maki, K.; Kanai, M.; Shibasaki, M. Org. Lett. 2003, 5, 733.

318

(a) Kiyooka, S.; Heathcock, C. H. Tetrahedron Lett. 1983, 24, 4765. (b) Heathcock, C. H.; Kiyooka, S.; Blumenkopf, T. A. J. Org. Chem. 1984, 49, 4214.

654

12. CARBON-CARBON BOND-FORMING REACTIONS

Using tin tetrachloride (SnCl4) as a catalyst provided better syn selectivity in the products [(2R,3S)-2-(benzyloxy) hex-5-en-3-ol and (2R,3R)-2-(benzyloxy)hex-5-en-3-ol] than did BF3 or TiCl4 (TiCl4 did not catalyze this coupling reaction).318 As with the allyltin complexes, these results are explained by chelation control. A synthetic example, taken from the F€ urstner et al.319 synthesis of herbarumin I, also illustrates an interesting variation. Furanose derivative 256 reacted with allyltrimethylsilane in the presence of TMSOTf to yield 257 (58% of the β anomer + 5% of the α anomer). The allylsilane reacted via the aldehyde generated in situ from the protected hemiacetal.

12.7.2 Silane Carbanions When a silane (e.g., 258) is treated with fluoride ion, fluoride attacks the silicon and displaces a carbanion, 259. If the anion is generated in the presence of an electrophile, typical carbanion reactions occur. It is unlikely that this carbanion (259) has a significant lifetime as a discrete entity. Indeed, it need not exist at all, since the carbanionic character of that carbon increases as the fluoride departs, forming a transient hypervalent silicon species. Allylic silanes are synthetically useful compounds320 that are usually prepared by reaction of chlorotrialkylsilanes and allylic Grignard reagents, as with the preparation of (E)-but-2-en-1-yl(tert-butyl)dimethylsilane by the reaction of (E)-but-2-en-1-ylmagnesium chloride and tert-butylchlorodimethylsilane.321 Sakurai and coworkers322 showed that allylic silanes react with ketones or aldehydes and tetra-n-butylammonium fluoride (TBAF) in THF. Allyl silanes (e.g., allyltrimethylsilane) reacted with butanal in the presence of TBAF to give an 83% yield of hept-1-en-4-ol. Similar reaction of allylic silanes with TBAF, in the presence of benzophenone gave the acyl substitution product.321 F−

RCH2SiR1 3

[ RCH2 ] −

258

F—SiR1 3

+

259

Me

OH

CHO

Si

Me Me Allyltrimethylsilane

Bu4NF, THF

Hept-1-en-4-ol

This type of coupling reaction of allylic silanes with aldehydes, catalyzed by Lewis acids, is known as the Sakurai reaction (or the Hosomi-Sakurai reaction).323 An example is taken from the synthesis of amphidinolide P by Williams et al.324 in which aldehyde (260) reacted with allylic silane (261) in the presence of boron trifluoride etherate to give a 64% yield of (262) (3.9:1 dr).

H

O Br

OHC

+

Me3Si

OPMB

H

O

OH

H

BF3⋅OEt2, –78°C CH2Cl2

Br

H 260

OPMB

261

262 (64%)

The disconnection for silane carbanion addition to carbonyls follow: R R

R1

R

R2

R

O

+

OH X

R1

R2

319

F€ urstner, A.; Radkowski, K.; Wirtz, C.; Goddard, R.; Lehmann, C. W.; Mynott, R. J. Am. Chem. Soc. 2002, 124, 7061.

320

Chabaud, L.; James,P.; Landais, Y. Eur. J. Org. Chem. 2004, 3173.

321

Hosomi, A.; Sakurai, H. Tetrahedron Lett. 1978, 2589.

322

(a) Hosomi, A.; Shirahata, A.; Sakurai, H. Tetrahedron Lett. 1978, 3043. (b) Sakurai, H. Pure Appl. Chem. 1982, 54, 1.

323

(a) Hosomi, A.; Sakurai, H. J. Am. Chem. Soc. 1977, 99, 1673. (b) Blumenkopf, T. A.; Heathcock, C. H. J. Am. Chem. Soc. 1983, 105, 2354.

324

Williams, D. R.; Myers, B. J.; Mi, L.; Binder, R. J. J. Org. Chem. 2013, 78, 4762.

655

12.8 CONCLUSION

Majetich et al325 showed that fluoride-generated carbanions undergo Michael-type additions. To achieve conjugate addition with carbanions generated from silanes usually requires the use of DMF as a solvent, and HMPA is often added to obtain satisfactory yields. In this example, 263 was treated with TBAF to generate the carbanion 264. Intramolecular Michael addition and hydrolysis led to the bicyclic product 265.325 Me

Me

Me TBAF, DMF 3 HMPA

O

O 263

O

−

SiMe3 264

H 265

Note that Vedejs et al326 used the fluoride/silane carbanion reaction to generate a novel imidate ylid that was used in a [3+2]-cycloaddition reaction with ethyl acrylate (Section 11.11.6). The disconnections possible with these latter reactions follow:

N

N O

O

O

X X

12.8 CONCLUSION This chapter has shown how incorporation of a metal or an appropriate electron-withdrawing group or atom allows generation of a carbanion that can be used to form new carbon-carbon bonds. The great variety of nucleophilic species, the range of electrophilic species with which they react, and the high degree of stereoselectivity, which accompanies many of the reactions show why these are among the most powerful disconnections in organic synthesis. The useful carbonyl stabilized carbanions known as enolate anions will be discussed in Chapter 13. HOMEWORK

1. Give a mechanistic explanation for the formation of the product shown from the designated starting material. O

O Ph CO2Et

S O

O

1. LiN(SiMe3)2

S

Ph

2. AcOH

CO2Et

2. Give the major product for each of the following when treated with (1) BuLi/THF/78°C; (2) MeI/78!0°C; (3) aq NH4Cl. SPh S

(A)

I

(B)

3. For each of the following give a complete reaction that illustrates its use for the formation of a carbon-carbon bond: (a) CdCl2; (b) Cp2TiMe2; (c) CuCN; (d) BF3; (e) PPh3; (f) NaH; (g) (MeO)2POMe; (h) DMSO. (a) Majetich, G.; Desmond, R. W., Jr.; Soria, J. J. J. Org. Chem. 1986, 51, 1753. (b) Majetich, G.; Casares, A.; Chapman, D.; Behnke, M. Ibid. 1986, 51, 1745.

325

(a) Vedejs, E.; Larsen, S.; West, F. G. J. Org. Chem. 1985, 50, 2170. (b) Vedejs, E.; Martinez, G. R. J. Am. Chem. Soc. 1979, 101, 6452; 1980, 102, 7993. (c) Vedejs, E.; West, F. G. J. Org. Chem. 1983, 48, 4773.

326

656

12. CARBON-CARBON BOND-FORMING REACTIONS

4. Give the major product of each reaction, with correct stereochemistry, and justify your choice. Me

Me

Me2CuLi

H

Br

Me2CuLi

H O

O

(A)

(B) 5. In each of the following reactions, predict the major product, with the correct stereochemistry where appropriate: O

HO

MeO

O

P

OMe

MeO

1. t-BuOK

TsO

KOt-Bu, THF 2. Me2CuLi

(A)

CHO

H

(B)

Me OTBS 1. BuLi, 1,3-Dithiane, THF

1. PPh3KCHOMe

(C)

OHC

CO2Me

2. Hg(OAc)2, aq THF

TsO

(D)

2. CH2KCHMgBr, CuI, THF

O

MeCLCMgBr, CeCl 3

SPh

Ph

LiN(ii-Pr)2, THF

O

OSiMe3

DABCO

(F)

THF

O

(E)

Me Me

OMe

Cp2Ti

Al Cl

N

(G)

Me2CuLi

Me

(H)

O

O

Cbz

S

PhLi +

(I)

N

Br−

O

(J) CsF, DMF

N

N +

(K)

2. n-BuLi, THF ≤ –78°C → rt

+

OTBS

N

H2CKCH(CH2)12MgBr CeCl3, THF, 0°C

N

N

(N)

Br

Me

O

O

(O)

CO2Me

S

2. Allyl bromide, DME rt, 15 min

(Q)

(S)

Ph

S OTBDMS

(R)

Zn, CH2Br2 TiCl4, THF

O

CHO

+ ClCH2CH2SHMe2 I −

t-BuOK , t-BuOH

(P)

1. Me2CuLi, Ether, 0°C

O

PPh3

O MeO P MeO

OTIPS

O

CH2Cl2, rt

1. BuLi, TMEDA–THF 2. Succinic anhdyride, THF

C3H7 Ph

O N

1.5 Cp2TiMe2, PhMe 80–110°C

(T) 1. CBr4, PPh3, Zn 2. BuLi, Ether

TBDPSO O HO

(U)

THF, Reflux

Ph

Me

1. dibal, CH2Cl2, –78°C

CN

+

Boc

2. LiN(TMS)2 , THF

Me

O

CH2OH

Ph

(M) O

1. CBr4 , PPh3 Zn, CH2Cl2

OSEM

OHC

(L)

SiMe3

O

2.5 equiv BnO

S

Ph Me

H

1. t-BuLi, THF–HMPA, –78°C 2. –78 → –45°C

t-BuMe2Si

1. Mesitoyl chloride Py

O OBn

2. HCLCLi, DMSO Ethylene diamine

CHO

(V)

3. BuLi, THF –78°C

657

12.8 CONCLUSION

6. Provide a synthesis for each of the following transformations. Show all reagents and intermediate products. HO

O

OHC

O

O OMe

MeO MeO

(A)

OMe OMe

OMe

OMe

MeO MeO

OMe

OTES

OH OBn

(B)

OMe OMe

OH OH

MeO2C

Et3Si

N

CO2Me OSiPh2t-Bu

(D)

MeO2C

(C)

N

EtO2C

O O

O

O

TrO CLCJEt

(E)

CHO

(F)

HO

(G)

CMe3

TBDMSO

O

EtO2C

Cl

I

CHO

OH OH

CO2Et N3

Cl

(H)

(I) OHC

CO2Et

I

I Me

Me

CO2Et

CO2Me

OTBS

OH

(K)

(J)

OH

OSiMe2t-Bu

Et

Me

Me

Me

Me

Me

OMe

7. Synthesize each of the following molecules from a starting material of no more than six carbons. That starting material must be commercially available from a chemical company. Show your retrosynthetic analysis and all reagents and intermediate products of the synthesis. O O C4H9

O

(A)

(B)

(C) O

O

(E)

(D)

(F)

Ph

(G)

C3H7

CH3

O

(H)

C H A P T E R

13 Nucleophilic Species That Form Carbon-Carbon Bonds: Enolate Anions 13.1 INTRODUCTION In Chapter 8, a hydrogen atom on a carbon α to an electron-withdrawing group was shown to be a weak acid. Removal of that proton by a base generated several synthetically useful reagents that contained a carbanion center, and they functioned as a Cd disconnect product (Section 1.2). When the electron-withdrawing group is a single carbonyl group (ketones, aldehydes, esters, etc.), the pKa of the adjacent protons are generally in the range pKa 19–24.1 As discussed in Section 11.6.9, the effects of the carbonyl group are due to (1) the inductive electronwithdrawing ability of the unsaturated substituent, but mainly due to (2) the ability of these substituents to delocalize the negative charge remaining after a proton has been removed.2 With the enhancement in acidity induced by the carbonyl, reagents that are weaker bases than organolithium reagents can be used for deprotonation.2 Deprotonation of a carbon α to a carbonyl leads to a resonance-stabilized enolate anion. The bases used in previous chapters are all rather nucleophilic, and when dealing with aldehydes, ketones, or carboxylic acid derivatives, acyl addition to the carbonyl may compete with deprotonation. Indeed, acyl addition must be suppressed or diminished if an enolate anion is to be generated and used in subsequent reactions. Acyl addition suppression can be achieved by decreasing the nucleophilicity of the base while maintaining or enhancing its basicity. This chapter will discuss enolate anions, and their chemical reactions.

13.2 FORMATION OF ENOLATE ANIONS 13.2.1 Preparation and Properties The attachment of an electron-withdrawing carbonyl on a CdH moiety enhances the acidity of that so-called α-hydrogen. Reaction of a carbonyl derivative (e.g., 3-methylbutan-2-one) and a suitable base leads to an acid-base reaction in which Ha is removed (as H+), making the α-carbon a carbanion (see 1). This resonance-stabilized carbanion from a simple ketone or aldehyde, called an enolate anion, has two resonance contributors and the electron density is delocalized on both the carbon and the oxygen. If the carbanion center is the site for any chemical reaction, an enolate anion will behave like any other nucleophilic center upon reaction with an electrophile. In such reactions, an enolate anion is a carbon nucleophile. In 3-methylbutan-2-one, there are two chemically different α-hydrogen atoms (Ha and Hb). There are three identical hydrogen atoms on the methyl group, which is C1 (labeled Ha), and one on C3 (labeled Hb). The presence of the electron-releasing methyl groups on the α-carbon diminishes the acidity of Hb because the CdHb bond is less polarized.

1

House, H. O. Modern Synthetic Reactions, 2nd ed.; W. A. Benjamin: Menlo Park, CA, 1972; p 494 and References 1, and 2b cited therein.

2

(a) d’Angelo, J. Tetrahedron 1976, 32, 2979. (b) Stowell, J. C. Carbanions in Organic Synthesis; Wiley-Interscience: New York, 1979; pp 127–216.

Organic Synthesis http://dx.doi.org/10.1016/B978-0-12-800720-4.00013-1

659

Copyright © 2017 Michael Smith. Published by Elsevier Inc. All rights reserved.

660

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

O

O

O

Base

Ha

Hb

– Ha

Hb 3-Methylbutan-2-one

Hb 1

If we compare acetone (pKa, 20) with 3,3-dimethylbutan-2-one (pKa 20.8), it is obvious that they are close in acidity. However, when 3,3-dimethylbutan-2-one is compared to 2,2-dimethylpentan-2-one, with a pKa 21.3, and 2,2,4trimethylpentan-2-one with a pKa 23.5, it is apparent that the presence of the alkyl substituents leads to a higher pKa (a less acidic proton). Since deprotonation is an acid-base reaction, and alkyl groups are electron releasing, the effect on the equilibrium is that the carbanionic center in the enolate anion is destabilized and that the CdH bond is stronger (see CdHb in 3-methylbutan-2-one). In other words, Hb is less acidic than Ha, and in general a hydrogen atom on a less substituted carbon is more acidic than a hydrogen atom on a more substituted carbon. The α-hydrogen of an aldehyde is more acidic than the α-hydrogen atom of a ketone, as seen by comparing acetone (pKa,
20) with acetaldehyde [pKa (16.5)].3 Presumably, the methyl group attached to the carbonyl group acetone is electron releasing relative to the hydrogen atom attached to the carbonyl group in acetaldehyde, weakening acetone as an acid relative to acetaldehyde. The presence of a second carbonyl is expected to enhance the acidity of an adjacent proton due to inductive effects and increased resonance. Dialdehydes are more acidic than diketones, as seen by comparing malonaldehyde (propane1,3-dial, pKa 5.9) with acetylacetone (pentane-2,4-dione, pKa 9.0). Enhancement in acidity is not limited to carbonyl compounds. The presence of two or more electron-withdrawing groups enhances the acidity of an α-proton, but the effect is not strictly additive, as shown in Table 13.1.4,5 Note that the pKa of trinitromethane and tricyanomethane is smaller (more acidic) than that of acetic acid (pKa, 4.78) and almost as small as that of nitric acid (pKa 1.3). Comparing the α-hydrogen of acetone to the α-hydrogen of acetic acid (pKa 20 vs. 24) shows that the α-hydrogen of the ketone is more acidic. If acetone is compared with acetylacetone (pentane-2,4-dione, pKa 9.07 in Table 13.26–8) it is clear that the presence of the second carbonyl group greatly enhances the acidity of Ha. The presence of an electronreleasing methyl group in 3-methylpentane-2,4-dione diminishes the acidity (pKa 11). An electron-withdrawing sulfonyl group [see 3-(methylsulfonyl)pentane-2,4-dione] enhances the acidity (pKa 3.6). TABLE 13.1

Enhancement of Acidity With Increasing Number of Electron-Withdrawing Groups

Compound

pKa

Compound

pKa

CH3NO2

11

CH3COCH3

20

CH2(NO2)2

4

CH2(COCH3)2

9

CH(NO2)3

0

CH(COCH3)3

6

CH3SO2CH3

23

CH3CN

25

CH2(SO2CH3)2

14

CH2(CN)2

12

CH(SO2CH3)3

0

CH(CN)3

0

Reprinted with permission from Pearson, R. G.; Dillon, R. C. J. Am. Chem. Soc. 1953, 75, 2439. Copyright © 1953 American Chemical Society.

The enol content of carbonyl derivatives generally correlates with the acidity of the α-proton. In ketone 2 (the keto form), the acidic proton is attached to the α-carbon atom, whereas in the enol (3) the acidic proton is attached to oxygen. Removal of Ha from 2 generates enolate anion 4. Removal of Ha from enol 3 simply generates the other resonance 3

Smith, M. B. March’s Advanced Organic Chemistry, 7th ed.; Wiley: Hoboken, NJ, 2013; p 318.

4

(a) Buncel, E. Carbanions: Mechanistic and Isotopic Aspects; Elsevier: Amsterdam, 1975; p 5. (b) Cram, D. J. Fundamentals of Carbanion Chemistry; Academic Press: New York, 1965; p 12.

5

Pearson, R. G.; Dillon, R. C. J. Am. Chem. Soc. 1953, 75, 2439.

6

Zook, H. D.; Kelly, W. L.; Posey, I. Y. J. Org. Chem. 1968, 33, 3477.

7

Reference 1, p 494.

8

Cram, D. J. Fundamentals of Carbanion Chemistry; Academic Press: New York, 1965; pp 4, 10, 13, 14, 43, 48.

661

13.2 FORMATION OF ENOLATE ANIONS

contributor of 4. In general, monoalkylated ketone derivatives (e.g., 2) exist primarily in this keto form, but 1,3dicarbonyl derivatives (e.g., diketones, dialdehydes, or diesters) have relatively high concentrations of enol, which correlate with the greater acidity associated with a proton on the α-carbon. O

O

Ha

Ha

R R1

R1

R1 3

Acid

TABLE 13.2 Carbon acid

4, Conjugate base

pKa

Carbon acid

25a

H

24a

H

O H

O Ph

H

20.8b,c 23.5b,c

O

21.3b 19.5b

H

O H

Ph

20a

19.1d,b

O

O

18.6b

Ph

H

H CO2Et H CO2Et Me Me

Cl

15a

11a

H

CO2Et CO2Et Me

5.9a

Me

13.3a

3.6a

9.0a

O H

NO2

16.5a

O H

CHO

NO2

16.5

H

O

CHO

H O

O H Me

a

H

pKa

O

HO

H

Conjugate acid

The pKa Values of Typical Carbon Acids

H

O

H

H–Base

R1

O N C

+

R

R

R 2

O

O Base

MeO2S

O Me Me

4.7a

O

Reference 5; b Reference 6; c Reference 7; d Reference 8.

The relative enol content of a molecule is strongly dependent on the solvent. Ethyl acetoacetate (ethyl 3-oxobutanoate), for example, has an enol content of 0.4% in water, 2.2% in 50% aq EtOH, 6.9% neat, 11% in EtOH, 27% in ether, and 46% in hexane.9 There is a clear relationship with hydrogen-bonding ability and polarity of the solvent. The extensive hydrogen bonding that is possible in water minimizes intramolecular stabilization of the enol, whereas aprotic solvents (e.g., ether or hexane) maximize the intramolecular hydrogen bonding, which stabilizes the enol and leads to a higher percentage of the enol. Removal of a proton connected to oxygen is expected to be more facile than similar removal from a carbon, so there is the expectation that as the enol content increases, the acidity should increase.

9

Stewart, R. The Investigation of Organic Compounds; Prentice Hall: Englewood Cliffs, NJ, 1966; p 12.

662

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

Inductive effects also play a role in the increased acidity of Ha for 1,3-dicarbonyl derivatives (e.g., ethyl acetoacetate, Section 2.2.1). The carbonyl group is a typical electron-withdrawing substituent, but there are many others. In Section 11.6.9, the ability to increase the acidity of a proton was shown to follow the order: NO2

>

O C

>

SO2R

O >

O

C OR

CLN >

C R

>

Ph

This order is reflected in the pKa values shown in Table 13.2. Note that recent studies of X-ray structures of enolate anions and amide bases suggest that the enolate anion may be formed by direct removal of the proton from the α-carbon via an intermediate tetrameric Li complex involving the base and the carbonyl starting material (Section 13.2.4). Whether the α-hydrogen is removed directly from carbon or from the OdH of an enol, the presence of an electron-withdrawing group enhances the acidity of that hydrogen.

13.2.2 Nonnucleophilic Bases Although the carbonyl group enhances the acidity of the α-proton, the carbonyl carbon can also react with a nucleophilic base used for the deprotonation, via acyl addition. Both organolithium or Grignard reagents are powerful bases, certainly strong enough to deprotonate a ketone or aldehyde, but these organometallic bases are also good nucleophiles, and they react with carbonyl derivatives to give an alkoxide product. An example is the reaction of hexan-3-one and butyllithium to yield 5. This reaction can be suppressed somewhat by lowering the reaction temperature to 78°C, 100°C, or even lower, but nucleophilic addition is usually quite facile. It is therefore important to have other bases available to remove the α-proton in the acid-base reaction. Ideally, the base would not be very nucleophilic, so that acyl addition will be slow relative to the acid-base reaction that removes an α-proton. However, the base must be sufficiently basic to remove a relatively weak acidic proton. O

BuLi

O Li Bu

Hexan-3-one

5

Hydroxide and alkoxides (formed from reaction of alcohols with NaH, NaOH, etc.) are strong enough bases to deprotonate many ketones, although the reaction is reversible and sometimes slow. Sodium hydroxide in water or alcohol can be used, as well as MeOH/MeO, EtOH/EtO, or t-BuO/t-BuOH. These are nucleophilic reagents, however, and their viability depends on the reversibility of the nucleophilic addition reaction that will allow the deprotonation to compete. The bases NaH and KH are poor nucleophiles and they are very useful for generating enolate anions, although the reaction is not sufficiently viable for a wide range of applications to simple aldehydes and ketones. However, NaH and KH are commonly used with malonate derivatives, acetoacetic esters, or with 1,3diketones. An acid-base reaction with these bases generates hydrogen gas (HdH) as the conjugate base, which escapes from the reaction medium and helps drive the reaction toward the enolate anion product. Amines are weak acids, with pKa values of 36–40, so they are deprotonated only in the presence of a very strong base. When a secondary amine is treated with a suitable base, a strong conjugate base (R2NH + Base ! R2N M+) is generated, the amide anion (R2N). This conjugate base is more basic than the alkoxide obtained from similar reaction with an alcohol, and it is easily capable of removing the α-proton of a ketone or aldehyde. Generation of NaNH2 from ammonia is well known, and reaction of a secondary amine with a basic molecule (e.g., butyllithium) yields the analogous lithium dialkylamide. Lithium diisopropylamide (6, LDA)10 is formed by reaction of diisopropylamine and butyllithium.11 At first glance, it appears that LDA has a nucleophilic nitrogen atom. In reactions with carbonyl derivatives, however, nucleophilic attack at the acyl carbon is sterically blocked. This poor nucleophilicity is illustrated in Fig. 13.1, which represents the approach of LDA from the top face to butan-2-one, with the carbonyl oxygen projected to the front. As the nitrogen approaches the carbonyl group at the required Bu€rgi-Dunitz angle of
110 degree (Sections 7.9.4, 10.3.3, and 10.6),12 the 10

Hamell, M.; Levine, R. J. Org. Chem. 1950, 15, 162.

11

(a) Albarella, J. P. J. Org. Chem. 1977, 42, 2009. (b) Sasson, I.; Labovitz, J. Ibid. 1975, 40, 3670.

12

(a) B€ urgi, H.-B.; Shefter, E.; Dunitz, J. D. Tetrahedron 1975, 31, 3089. Also see (b) Polt, R.; Seebach, D. J. Am. Chem. Soc. 1989, 110, 2622.

663

13.2 FORMATION OF ENOLATE ANIONS

Steric hindrance

Steric hindrance

FIG. 13.1

Steric hindrance that inhibits nucleophilic acyl addition of LDA (top) as it approaches butan-2-one (bottom). The NdLi is aligned with the C]O unit in the front of the diagram.

methyl groups of the isopropyl units are repelled by the methyl and ethyl groups, which inhibits nucleophilic attack by nitrogen at the carbonyl carbon. In other words, steric hindrance prevents close contact of the nitrogen and the acyl carbon. Therefore, the nitrogen atom is categorized as a poor nucleophile in this reaction. Since approach to an α-hydrogen is not as sterically hindered, LDA reacts as a base without any problem. In general, acyl addition of dialkylamide bases (e.g., LDA) is very poor due to steric hindrance, but deprotonation of ketones or aldehydes is much faster, so dialkylamides are categorized as nonnucleophilic bases. In other words, nucleophilic acyl addition is slow relative to the acid-base reaction that removes the α-proton. Virtually any secondary amine can be converted to the corresponding amide base (R2NLi), including lithium diethylamide (7), lithium tetramethylpiperidide (8)13 or lithium hexamethyldisilazide (9),14,15 but 6 is probably the most commonly used amide base. In 8, the four methyl groups, at C2 and C6, inhibit approach of nitrogen to an acyl carbon. In 9, a bis(silyl)amide, the bulky trimethylsilyl groups sterically block the nitrogen and in this case steric hindrance and also diminished base strength leads to slow deprotonation for many carbonyl species. The acid-base reaction of theses amide bases with unhindered ketones, aldehydes, or esters usually generates the corresponding enolate anion without major problems. The hindered chelating base 1,8-bis(dimethylamino)naphthalene (10) is another useful base that is a very poor nucleophile, but extremely efficient at removing protons.16 It is sold under the name Proton Sponge™. The amide anion can be used as a base, often in ammonia, and ammonia has a pKa of 9.24 (ammonium ion, NH3 ! NH+4 ),17 whereas amide (H2N) has a pKa of 38 (ammonia, H2N ! NH3),18 both measured in water BuLi , THF

N H

N Li

–78 → 0 → –78°C

6 Et Li N Et

Me Me

7

N Li

Me Me

8

Li N

SiMe3

Me2N

NMe2

SiMe3 9

10

Note that there are differences in base strength for the amide bases. It is known that pKa values can be used for a comparison of an amine with the conjugate acid of the reaction of an amide base and an acid (e.g., amine vs. ketone or aldehyde). This comparison allows an estimate of the relative position of the equilibrium in an acid-base reaction, allowing an estimate to the relative base strength for a reaction. Fraser and Mansour19 reported that the pKa of 13

Olofson, R. A.; Dougherty, C. M. J. Am. Chem. Soc. 1973, 95, 581.

14

Amonoo-Neizer, E. H.; Shaw, R. A.; Skovlin, D. O.; Smith, B. C. J. Chem. Soc. 1965, 2997.

15

For a discussion of the structure of lithium hexamethyldisilazide, see Lucht, B. L.; Collum, D. B. Acc. Chem. Res. 1999, 32, 1035.

16

(a) Ishida, A.; Mukaiyama, T. Chem. Lett. 1976, 1127. (b) Diem, M. J.; Burow, D. F.; Fry, J. L. J. Org. Chem. 1977, 42, 1801.

17

Bruckenstein, S.; Kolthoff, I. M. In Treatise on Analytical Chemistry; Kolthoff, I. M., Elving, P. J., Eds., Wiley: New York, 1959; Vol. 1, part. 1, pp 432, 433.

18

Bunbcel, E.; Menon, B. J. Am. Chem. Soc. 1977, 99, 4457.

19

Fraser, R. R.; Mansour, T. S. J. Org. Chem. 1984, 49, 3443.

664

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

(Me3Si)2NH is 29.5, diisopropylamine is 35.7, dicyclohexylamine is 35.7, and 2,2,6,6-tetramethylpiperidine is 37.3, all reported in THF. The stronger acid will generate the weaker conjugate base. Using this pKa data, 8 is more basic than 6, and 9 is the least basic in this series. If 6 reacts with acetophenone with a pKa of 19.1 (Table 13.2), the conjugate base is the enolate anion and the conjugate acid is diisopropylamine, with a pKa of 35.7. Clearly, the equilibrium lies on the side of the conjugate acid-base pair, so 6 is a good choice for generation of the enolate anion. If hydroxide is used, however, the conjugate acid is water with a pKa of 15.6. Now, the stronger acid is the conjugate acid when compared to the ketone, and the equilibrium lies to the left, and hydroxide is a relatively weak base in this system. The pKa data can be used to estimate the efficacy of other bases or solvents. Remember that if water is used as a solvent, it may be a stronger acid relative to the starting ketone or even the conjugate acid. The solvent will therefore play a role in determining the equilibrium constant for a given acid-base reaction. In water,20 the pKa of the conjugate acid of 10 is reported to be 12, but it is 7.47 in DMSO.21 The structure of LDA is not the monomeric species suggested by structure 6, but rather an aggregate. Jackman and De Brosse,22a and later Bauer and Seebach22b studied the colligative properties of LDA as a solution in THF, and concluded that dilute solutions exist as both dimers (11) and monomers (12). The structure of LDA in solution is assumed to be the THF solvated dimer shown by Williard and Salvino’s22c X-ray structure in Fig. 13.2, which shows the nitrogen may be even more sterically hindered than is suggested by the representation in Fig. 13.1. Note that 12 is the same representation as 11, where n ¼ 1. Solution structures of chiral lithium amides having internal sulfide coordination have also been determined, and mixed dimeric complexes are generally formed.23

N O

O Li

Li N

FIG. 13.2 X-ray crystal structure of LDA. Structure drawn using Spartan software by Wavefunction, Inc. Hydrogen atoms are omitted for clarity. Reprinted with permission from Williard, P. G.; Salvino, J. M. J. Org. Chem. 1993, 58, 1. Copyright © 1993 American Chemical Society.

Collum and Rutherford24 studied the solution structure of lithium diethylamide (7) in THF and in ether, where dimers and trimers, and also four-, five-, and six-rung ladder structures (e.g., 13) were detected. Collum and Galiano-Roth25 investigated the aggregation state of LDA in solution during the metalation of N,N-dimethylhydrazones (Section 13.4.6.2) using 6Li and 15N NMR. The NMR data also suggested a solvated, dimeric structure (e.g., 11). R R

N (thf) n

Li

Li

(thf) n

R

N

Li

N

Li

Li

N

Li

N

N(thf) n

N

R 11

20

Kresge, A. J. Pure. Appl. Chem. 1981, 53, 189.

21

Hess, A. S.; Yoder, R. A.; Johnston, J. N. Synlett 2006, 147.

R

12

22

R

R R

13

(a) Jackman, L. M.; De Brosse, C. W. J. Am. Chem. Soc. 1983, 105, 4177. (b) Bauer, W.; Seebach, D. Helv. Chim. Acta 1984, 67, 1972. (c) Williard, P. G.; Salvino, J. M. J. Org. Chem. 1993, 58, 1. Also see (d) Shobatake, K.; Nakamato, K. Inorg. Chem. Acta 1970, 4, 485.

23

Sott, R.; Faranander, J.; Diner, P.; Hilmersson, G. Tetrahedron Asymmetry 2004, 15, 267.

24

Rutherford, J. L.; Collum, D. B. J. Am. Chem. Soc. 1999, 121, 10198.

25

Galiano-Roth, A. S.; Collum, D. B. J. Am. Chem. Soc. 1989, 111, 6772.

665

13.2 FORMATION OF ENOLATE ANIONS

Collum and DePue26 found evidence the structure of lithium isopropylcyclohexylamine (LICA) was an equilibrating set of stereoisomeric dimers in solution, cis-14 and trans-14. Collum’s studies found no evidence for monomeric species in solution, and concluded that 12 was not a significant species in ether solutions of LDA.25,26 Collum and coworker’s27 showed that enolization of ketones with LiN(TMS)2, in the presence of diisopropyl ether, substituted THF derivatives, and cineole proceed via dimer-based transition structures. The hindered ethers accelerate the enolization by sterically destabilizing the reactants and stabilizing the transition structures. Pratt et al.28 found that lithium 2,2,6,6-tetramethylpiperidide (LTMP) in THF gave the best selectivity for formation of the (E)-enolate anion of ketones. Selectivity was improved further by using a LTMP-butyllithium mixed aggregate. Less polar solvents led to diminished selectivity. (thf) n Li N N Li (thf) n

(thf) n Li N N Li (thf) n 14

13.2.3 (E/Z) Geometry in Enolate Formation An enolate anion is usually formed from a ketone as a mixture of (E)- and (Z)-isomers. Treatment of 2-methylpentan-3-one with LDA (THF, 78°C), for example, gave a 60:40 mixture of the (Z)- and (E)-enolates (15 and 16).29 Subsequent acyl addition of 15 and 16 to a carbonyl compound (Section 13.4.1) could lead to different stereochemical consequences, even if the orientation and facial bias of the reaction is controlled. It is therefore important to understand the factors that control selectivity in enolate-forming reactions. Evans’ review30 includes a compilation of several studies of selective enolization. The influence of structure and base on enolate geometry for simple ketones has been studied.31 The reaction of pentane-3-one (17, R ¼ Et) with LTMP, LDA, or LICA yields a 3:1–4:1 mixture of (Z)-18-(E)-18, whereas reaction with LTMP/HMPA or LHDS yields a 2:1–10:1 mixture of (E)-18-(Z)-18.31 This result contracts with the reaction of 4-methylpentan-3-one (17, R ¼ CHMe2), all the bases studied gave a 2:1 preferences for (Z)-18 over (E)-18, although LTMP showed a 2:1 preference for (E)-18 over (Z)-18.31 Interestingly, 4,4-dimethylpentane-3-one (17, R ¼ CMe3) and 1-phenylpropane-1-one (17, R ¼ Ph) reacted with LDA to yield a >98:2 mixture of (Z)-18-(E)–18.31 LiN(i-Pr)2

+

O

O Li

O Li

15

16

2-Methylpentan-3-one

R1

Me O

LiNR2, THF

R1

+

–78 → 0°C

17

Me

R1

Me

O Li (Z)-18

O

Li

(E)-18

Only moderate variations in kinetic (E/Z) ratios are observed, except when LHDS was used as the base, which favored the thermodynamically more stable (Z)-enolate. The sterically demanding base LTMP favored formation of the kinetic (E)-enolate.32 This equilibration may involve an aldol addition-retro-aldol process as suggested by

26

DePue, J. S.; Collum, D. B. J. Am. Chem. Soc. 1988, 110, 5518, 5524.

27

Zhao, P.; Lucht, B. L.; Kenkre, S.; Collum, D. B. J. Org. Chem. 2004, 69, 242.

28

Pratt, L. M.; Newman, A.; St. Cyr, J.; Johnson, H.; Miles, B.; Lattier, A.; Austin, E.; Henderson, S.; Hershey, B.; Lin, M.; Balamraju, Y.; Sammonds, L.; Cheramie, J.; Karnes, J.; Hymel, E.; Woodford, B.; Carter, C. J. Org. Chem. 2003, 68, 6387. 29

Heathcock, C. H.; Buse, C. T.; Kleschick, W. A.; Pirrung, M. A.; Sohn, J. E.; Lampe, J. J. Org. Chem. 1980, 45, 1066.

30

Evans, D. A. in Asymmetric Syntheses; Morrison, J. D., Ed., Academic Press: New York, 1984; Vol. 3, pp 1–110.

31

Reference 30, p 16.

32

(a) Fataftah, A. Z.; Kopka, I. E.; Rathke, M. W. J. Am. Chem. Soc. 1980, 102, 3959. (b) also see Nakamura, E.; Hashimoto, K.; Kuwajima, I. Tetrahedron Lett. 1978, 2079.

666

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

Rathke and coworkers32 (via 19). The aldol condensation is discussed in Section 13.4.1. Addition of HMPT (hexamethylphosphorus triamide) destabilizes and disaggregates aldolate 19, thus promoting equilibration. O

R

O Li

O

Li O

R1

O

O Li

R1

R

Me

R

R1

Me

Me (E)

(Z)

19

The(E/Z) trends observed for ketones are also seen in enolate anions of esters and amides.33 The kinetic (E)-ester enolate usually predominates with esters unless HMPT is added,34 but amides35 give the thermodynamic (Z)-isomer. It is unlikely that the equilibration observed with ketones is responsible for reversal in enolate geometry in esters on addition of HMPT. Such equilibration would show a preference for formation the (Z)-enolate with both LDA and secbutyllithium, and would lead to significant amounts of Claisen condensation (Section 13.4.2.1). The influence of the dialkylamide base and the carbonyl structure on the (E/Z) enolate ratio can be rationalized for ketones, esters, and amides by the so-called Ireland model,34,36 where the two important steric factors are taken to be R1 (R1 ¼ OEt, i-Pr, Ph, OMe, OtBu, Net2) versus Me in 20 and Me versus L in 21.37 A related model known as the Zimmerman-Traxler model will be discussed in Section 13.5.1.3. The Ireland model was developed to predict stereochemistry in the anionic accelerated Claisen rearrangement (Section 15.5.5), but is used for predicting stereochemistry in ester enolates. Me

L O Li

O

H

R1

Li N

L

22

Me

H

O

Me

R1

Li

L N

H

R1 H

R1

Me

O Li

L

20

21

23

Enolate anions actually exist as aggregate dimers, tetramers, or hexamers (Section 13.2.4), so this model has some limitations although it generally gives reasonable predictions. This model is also a useful tool for predicting (E)- and (Z)-isomers for enolate anions of ketones, esters, and amides. When R1 is not sterically demanding (as in OR) the H $ L interaction in 20 is much smaller than the Me $ L interaction in 21 and more of the (E)-isomer (22) is produced.37 When R1 is large (e.g., tert-butyl) the large Me $ R1 interaction in 20 is larger than the H $ R1 interaction and usually larger than the Me $ L interaction in 21. In this case, 21 is favored, leading to an increased amount of the (Z)-isomer (23).37 It is possible that addition of HMPT disrupts the pericyclic transition states represented by 20 and 21.

13.2.4 Structure and Aggregation State of Enolate Anions The structure of the enolate anion is a very important factor in the stereochemical outcome, as is the solvent. Metal enolates exist as dimers38a or other aggregates in ether solvents.33 See Section 13.2.2 for the structure of LDA aggregates.38d Jackman and Szeverenyi39a suggested that the lithium enolate of isobutyrophenone exists as a tetramer (24) in THF solution, but as a dimer (25) in DME.40 House et al.41 proposed these aggregates, and found that ketone 33

Reference 30, p 17 and Refs. 39, 40, 45–47 cited therein.

34

Ireland, R. E.; Mueller, R. H.; Willard, A. K. J. Am. Chem. Soc. 1976, 98, 2868.

35

Reference 30, p 17.

36

Reference 30, pp 18–20.

37

(a) Reference 30, p 19; also see (b) Meyers, A. I.; Snyder, E. S.; Akerman, J. J. H. J. Am. Chem. Soc. 1978, 100, 8136. (c) Hoobler, M. A.; Bergbreiter, D. E.; Newcomb, M. Ibid. 1978, 100, 8182. (d) Davenport, K. G.; Eichenauer, H.; Enders, D.; Newcomb, M.; Bergbreiter, D. E. Ibid. 1979, 101, 5654. 38

(a) Bernstein, M. P.; Collum, D. B. J. Am. Chem. Soc. 1993, 115, 789. (b) Bernstein, M. P.; Romesberg, F. E.; Fuller, D. J.; Harrison, A. T.; Collum, D. B.; Liu, Q. Y.; Williard, P. G. Ibid. 1992, 114, 5100. (c) Collum, D. B. Acc. Chem. Res. 1992, 25, 448. (d) Reference 30, p 21 and Refs. 59–64 cited therein.

39

(a) Jackman, L. M.; Szeverenyi, N. M. J. Am. Chem. Soc. 1977, 99, 4954. (b) Jackman, L. M.; Lange, B. C. Ibid. 1981, 103, 4494.

40

Jackman, L. M.; Lange, B. C. Tetrahedron 1977, 33, 2737.

41

House, H. O.; Gall, M.; Olmstead, H. D. J. Org. Chem. 1971, 36, 2361.

667

13.2 FORMATION OF ENOLATE ANIONS

enolate anions of metals from groups 1, 2, and 3 exist primarily as the O-metal enolate.40,42 For lithium enolate anions, increasing the donor properties of the solvents from Ether ! THF ! DME increases the LidO bond polarization, enhancing the α-carbon π-electron density.43 An increase in π-electron density is also observed for the series: Li ! Na ! K. Increasing π-electron density is associated with greater reactivity of carbon in the enolate alkylation reactions.44 The solvent plays a significant role in the aggregation state of enolate anions, which directly influences their reactivity. Streitwieser et al.45 found that selected lithium enolate anions have lower aggregation states in dimethoxyethane than in THF, and that aggregation is much higher in methyl tert-butyl ether. He found that alkylation and acylation of the enolate anions was extremely slow in methyl tert-butyl ether. These reactions apparently require additional solvation of the lithium cation, which is ineffective in this solvent.

R S

Li

R

S Li

O O

S

R

S Li O R Li O R

S Li

O Li

O

S Li O R Li Cl

S

S 24

Me O

R

R = (Me 2CKC(Ph)—)

Li O Me S = solvent

Me O

R O

Li

O R

O Me

25

Reprinted with permission from Jackman, L. M.; Szeverenyi, N. M. J. Am. Chem. Soc. 1977, 99, 4954. Copyright © 1977 American Chemical Society.

Lithium enolate anions of ketones exist as aggregates in solution,39–41,44d,46 and mixed aggregates between the enolate anion and the amide base are also possible.47 In 1981, Seebach and coworkers48 used X-ray crystallography to confirm that the lithium enolates of pinacolone and cyclopentanone form a tetrameric aggregate in the solid state. It was assumed that a similar species exited in solution. A THF solvated tetramer of lithium pinacolonate is shown (see 26), as reported by Seebach and coworkers.48 Williard et al.49 reported the X-ray structure of the unsolvated lithium enolate of pinacolone to be a hexamer (27). The structure of the potassium enolate of pinacolone was a hexamer, although the sodium enolate was a tetramer. The tetramer was composed of a cubic cluster of four enolized ketone molecules, four sodium atoms, and four unenolized solvating ketones units, which contrasts with 27 in which each oxygen atom is coordinated to three Li atoms, and each enolate double bond is almost antiperiplanar to one Li atom. A variety of X-ray crystal structures for organolithium derivatives confirm the presence of aggregates in the solid state.50 There is also evidence that the aggregate structure is preserved in solution and is probably the actual reactive species.2b

42

(a) House, H. O.; Auerbach, R. A.; Gall, M.; Peet, N. P. J. Org. Chem. 1973, 38, 514. (b) Orsini, F.; Pelizzoni, F.; Ricca, G. Tetrahedron Lett. 1982, 23, 3945. (c) House, H. O.; Prabhu, A. V.; Phillips, W. V. J. Org. Chem. 1976, 41, 1209.

43

Reference 30, p 22.

44

(a) Zook, H. D.; Gumby, W. L. J. Am. Chem. Soc. 1960, 82, 1386. (b) Zook, H. D.; Russo, T. J. Ibid. 1960, 82, 1258. (c) Zook, H. D.; Russo, T. J.; Ferrand, E. F.; Stotz, D. S. J. Org. Chem. 1968, 33, 2222. (d) Zook, H. D.; Kelly, W. L.; Posey, I. Y. Ibid. 1968, 33, 3477. (e) Zook, H. D.; Miller, J. A. Ibid. 1971, 36, 1112.

45

Streitwieser, A.; Juaristi, E.; Kim, Y.-J.; Pugh, J. K. Org. Lett. 2000, 2, 3739.

46

Stork, G.; Hudrlik, P. F. J. Am. Chem. Soc. 1968, 90, 4464.

47

Sun, C.; Williard, P. G. J. Am. Chem. Soc. 2000, 122, 7829.

48

(a) Amstutz, R.; Schweizer, W. B.; Seebach, D.; Dunitz, J. D. Helv. Chim. Acta 1981, 64, 2617. (b) Seebach, D.; Amstutz, D.; Dunitz, J. D. Ibid. 1981, 64, 2622.

49 50

(a) Williard, P. G.; Carpenter, G. B. J. Am. Chem. Soc. 1986, 108, 462. (b) Williard, P. G.; Carpenter, G. B. Ibid. 1985, 107, 3345.

For example, see (a) Bauer, W.; Laube, T.; Seebach, D. Chem. Ber. 1985, 118, 764. (b) van Koten, G.; Jastrzebski, J. T. B. H. J. Am. Chem. Soc. 1988, 107, 697. (c) Polt, R. L.; Stork, G.; Carpenter, G. B.; Williard, P. G. Ibid. 1984, 106, 4276. (d) Amstutz, R.; Laube, T.; Schweizer, W. B.; Seebach, D.; Dunitz, J. D. Helv. Chim. Acta 1984, 67, 224.

668

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

Li OLi O

Li

O Li

• THF 4

O

26 Provided through the courtesy of Professor Dieter Seebach, E.T.H., Zurich, Switzerland. Structure drawn using Spartan software by Wavefunction, Inc. Hydrogen atoms are omitted for clarity.

Li

O

Li

O

27 Provided through the courtesy of Professor Paul Williard, Brown University, Providence, RI. Structure 27 is drawn using Spartan software by Wavefunction, Inc.

Seebach et al. 51 showed that lithium enolate anions derived from esters are aggregates, but the crystal contains two coordinated TMEDA molecules. The lithium enolate of (Z)-methyl 3,3-dimethylbutanoate was found to be a dimer containing four THF molecules. Focusing on colligative properties rather than X-ray analysis, Collum and coworkers52 determined that lithium 2-carbomethoxycyclohexanone dimethylhydrazone (Section 13.4.6.2) is a dimeric THF solvate in the solid state. Perhaps the most important implication of this work is that the reactivity of enolate anions in alkylation and condensation reactions (see below) will be influenced by the aggregate state of the enolate. If the enolate anion is attached to (or a part of ) a tetrameric or hexameric complex, then simple steric and electronic arguments for reactivity and selectivity are flawed if they are based only on examining monomeric species. The relative proportions of (E)- and (Z)-enolate anions will be influenced by the extent of solvation and the aggregation state.

51 52

Seebach, D.; Amstutz, R.; Laube, T.; Schweizer, W. B.; Dunitz, J. D. J. Am. Chem. Soc. 1985, 107, 5403.

(a) Wanat, R. A.; Collum, D. B.; Van Duyne, G.; Clardy, J.; De Pue, R. T. J. Am. Chem. Soc. 1986, 108, 3415. (b) Wanat, R. A.; Collum, D. B. Ibid. 1985, 107, 2078.

669

13.2 FORMATION OF ENOLATE ANIONS

Once formed, an enolate anion can react with a variety of electrophiles. The approach of the electrophile is usually assumed to be perpendicular (as in 28), in reactions with alkyl halides and epoxides, to maximize overlap of participating orbitals.53 This model also applies to the deprotonation reaction. The LUMO of the electrophile is important, and Dunitz and coworkers54 showed that a non vertical approach (represented by 29 or 30)55 is preferred when nucleophiles attack carbonyl centers. The angle of approach (as discussed in Sections 7.9.4, 10.3.3, and 10.6) is
110 degree and is called the Bu€rgi-Dunitz trajectory.12,54 Models 29 and 30 imply that as the enolate nucleophile approaches the carbonyl carbon, the approach angle is somewhat restricted, which enhances the steric effects between R and the approaching electrophile. Since the geometry of the enolate will influence the approach of the electrophile, geometric constrictions must be considered when examining stereoselectivity. It appears that enolate-alkyl halide transition states are essentially reactant-like, and that the product stereochemistry is determined largely by steric factors. Lithium enolate anions exist as large aggregates, and their approach to the electrophile is restricted by steric and electronic considerations, as well as by the relative geometry of the molecule. Nonetheless, good predictions for reactivity and diastereoselectivity can be made based on the steric requirements that would be present in a monomeric system. For example, in most reactions of enolate anions the electrophile will be delivered to the less hindered face of the enolate to give the major product. In all models used to describe reactivity (Section 13.5), a monomeric enolate anion will be shown, but the facial and orientation bias of the enolate anion is clearly influenced by the state of aggregation in solution. E+ R1

R1

H R2

M O

O

R2

M O

R

H

O

R2

28

R

C

E

R1

H

E+

R

R

R

R O

M

29

R

R

O

M

30

13.2.5 Kinetic Versus Thermodynamic Control The reaction of hexan-3-one with LDA should yield a mixture of 31 and 32, and each enolate anion is a mixture of (E)- and (Z)-isomers. In this case, a 1:1 mixture of 31 and 32 is expected. Hexan-3-one is an unsymmetrical ketone, and each α-carbon has one alkyl substituent (methyl and ethyl, respectively). The substitution pattern of the α-carbon for each alkyl group is the same, so the pKa of each proton is expected to be about the same, so deprotonation should occur at about the same rate for each proton. There are, therefore, two competing equilibrium reactions, leading to two different enolate anions. If the rate of deprotonation is about the same, these two equilibrium reactions should lead to a roughly 1:1 mixture of the two different product, which is the case for 31 and 32. An unsymmetrical ketone may have two α-carbon atoms, each with an acidic proton, but substituents may be present so the pKa values of the protons are different. The α- and α0 -carbon can be monosubstituted or disubstituted, for example. The α-protons at each carbon will have slightly different pKa values, with the less substituted position having the more acidic proton. As noted in Table 13.2, 3,3-dimethylbutan-2-one (O]CdCH3) has a pKa of 20.8 and 4,4dimethylpentan-3-one (O]CCH2CH3) has a pKa of 21.3. In general, the difference in pKa for monosubstituted versus disubstituted positions is typically only
1 pKa unit. When such a ketone is treated with an appropriate base, there may be two competing acid-base reactions and two different rates of deprotonation.

53

(a) Reference 30, p 25. (b) Velluz, L.; Valls, J.; Nomine, G. Angew. Chem. Int. Ed. Engl. 1965, 4, 181 (see p 189). (c) Valls, J.; Toromanoff, E. Bull. Soc. Chim. Fr. 1961, 758. (d) Toromanoff, E. Ibid. 1962, 708, 1190. (e) Bucourt, R. Ibid. 1964, 2080. (f ) Toromanoff, E.; Bucourt, R. Tetrahedron Lett. 1976, 3523.

54

(a) B€ urgi, H. B.; Dunitz, J. D.; Lehn, J. M.; Wipff, G. Tetrahedron 1974, 30, 1563. (b) B€ urgi, H. B.; Dunitz, J. D.; Shefter, E. J. Am. Chem. Soc. 1973, 95, 5065. (c) B€ urgi, H. B.; Lehn, J. M.; Wipff, G. Ibid. 1974, 96, 1956. (d) B€ urgi, H. B. Angew. Chem. Int. Ed. Engl. 1975, 14, 460.

55

(a) Agami, C. Tetrahedron Lett. 1977, 2801. (b) Agami, C.; Chauvin, M.; Levisalles, J. Ibid. 1979, 1855. (c) Agami, C.; Levisalles, J.; Lo Cicero, B. Tetrahedron 1979, 35, 961.

670

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

+

O

O

Hexan-3-one

O 32

31

O

O

Ha

O Hb

Ha 34

Hb

2-Methylpentan-3-one

33

2-Methylpentan-3-one has two acidic protons, Ha and Hb. There is a methyl group that is electron releasing to the carbon bearing Hb, reducing the acidity of Hb by 1
pKa unit relative to Ha. Removal of Ha leads to enolate 33, and loss of Hb leads to enolate 34. Enolate 33 results from loss of the most acidic proton, which is removed faster in an acid-base reaction to yield the kinetic enolate anion (less stable). Enolate 34 is the more thermodynamically stable since it has the more highly substituted C]C unit, but it is formed only after equilibration of the initially formed enolate anion and is known as the thermodynamic enolate anion. In an acid-base reaction, the more acidic proton should always be removed faster, by any base, to yield the kinetic enolate anion. How is it possible to remove the less acidic proton to generate the thermodynamic enolate anion? The enolate-forming process is an acid-base reaction, so it is an equilibrium reaction (Section 2.2.1). The key to regioselectively form one enolate or the other is to control the acid-base equilibrium. As with any acid-base reaction, the equilibrium is controlled by the relative strengths of the acid-conjugate acid and the base-conjugate base pairs. When a strong acid reacts with a strong base, the product is a weak conjugate acid and a weak conjugate base, which will shift the equilibrium to the right (Ka is large). When the conjugate acid and base are formed and are stronger than the acid and base starting materials, the equilibrium is shifted to the left (Ka is small).

Base

O Acid R

R

+

Base

k1

B:

R

R

k−1

H

Acid

O

35

+

B:H

36

Ka

=

k1 k–1

and

pKa

=

– log Ka

A typical equilibrium reaction of a ketone and a base is shown for the reaction of 35 to yield 36. Kinetic control (formation of the kinetic enolate as the major product) is achieved when Ka is large (k1 ≫ k1). If Ka is near unity, thermodynamic or equilibrium control (formation of the thermodynamic enolate as the major product) predominates.2 This equilibrium is controlled by the choice of solvent, base, cation, and temperature. Kinetic control is favored by several factors,2 including the following: 1. 2. 3. 4. 5.

The use of aprotic solvents. The use of strong bases that are weakly nucleophilic and generate a conjugate acid that is weaker than the ketone or aldehyde. The use of cations (e.g., Li) that form relatively covalent bonds to oxygen (not Na+ or K+). The use of low-reaction temperatures. The use of sort reaction times.

The conditions that favor thermodynamic control must be the opposite of those described for kinetic control.2 Thermodynamic control conditions include the following: 1. 2. 3. 4. 5.

The use of protic solvents. The use of bases that generate a conjugate acid that is stronger than the ketone or aldehyde (such bases are usually nucleophilic). The use of cations that form ionic MdO bonds. The use of higher reaction temperatures. The use of long reaction times that allow equilibration.

671

13.2 FORMATION OF ENOLATE ANIONS

These parameters can be explained by examining the three species involved in the equilibrium, 33 and 34. For kinetic control (formation of 33), the reaction should be essentially irreversible and the reaction conditions should suppress the equilibrium. Solvents can be grouped according to whether or not they have an acidic proton. Protic solvents (solvents containing an acidic hydrogen) usually involve structures with an OdH, SdH or NdH and they promote equilibrium conditions (small pKa) by acid-base reactions with the conjugate base. Remember that a Ka ¼ 1 represents a 50:50 equilibrium. Aprotic solvents have CdH bonds, but no proton is attached to a heteroatom so there is no acid-base reaction with the conjugate base. Aprotic solvents favor a large Ka (an equilibrium that is shifted to the right). In the reaction of 35, the ketone is the acid (pKa
21) and a protic solvent (e.g., ethanol, pKa
17) or water (pKa
15.8) will be a stronger acid. Once the conjugate base (the enolate anion, 36) is formed, it reacts with the acidic solvent to regenerate the ketone, driving the equilibrium to the left (small Ka). Aprotic solvents (e.g., THF) do not have an acidic proton and the enolate anion cannot react with them, meaning that once formed the enolate anion will be relatively stable (equilibrium shifted to the right, large Ka). In other words, aprotic solvents (e.g., ether or THF) favor kinetic control, and protic solvents (e.g., ethanol, water, methanol, or tert-butanol) favor thermodynamic control. The base plays two roles in this reaction. First, it must react with the acid (the ketone) quickly and efficiently (it must be a strong base). The base also generates a conjugate acid after deprotonation that plays a significant role in the position of the overall equilibrium. If the conjugate acid is more acidic than the ketone, an acid-base reaction will reprotonate the conjugate base (the enolate anion) and drive the equilibrium back toward the ketone (smaller pKa). When LDA reacts with the ketone to form the corresponding enolate anion in an aprotic solvent, the only acids in the equilibrium are the ketone (pKa
20) and diisopropylamine (pKa
25), the conjugate acid formed by reaction with LDA. Since diisopropylamine is a weaker acid than the ketone, reaction with the enolate anion (the conjugate base) is relatively slow, which shifts the overall equilibrium to the right (large Ka) and favors kinetic enolate formation. Note that diisopropylamine is a weak acid, and given sufficient time it will react with the enolate anion to regenerate the starting ketone and LDA, shifting the equilibrium back to the left (see below). If the reaction is done in THF with sodium ethoxide as the base, enolate 33 is also formed since Ha is the most acidic proton. The pKa of this ketone acid is
21 (as mentioned, see Table 13.2), but as ethoxide reacts with 2-methylpentan-3-one, ethanol is formed as the conjugate acid with a pKa
17. Ethanol easily reacts with the enolate anion, shifting the equilibrium to the left so that 2-methylpentan-3-one, ethoxide, ethanol, and some 33 are all in solution. Although Hb in 2-methylpentan-3-one is less acidic than Ha, it is acidic enough to be removed by ethoxide. N

O Me

Li

O

Me Ha

Me

Me

Hb

Me Hb

THF

+

N

H

Me

2-Methylpentan-3-one

33

Diisopropylamine

If the conversion of 2-methylpentan-3-one to 33 is slow and/or reversible, the less acidic Hb will be removed to generate enolate anion 34. The conjugate acid (ethanol) will react with the enolate anion, which reprotonates 34, establishing a second equilibrium between 33 and 34. If the solvent is ethanol, which is a stronger acid than 2-methylpentan-3-one, this equilibrium is favored and as mentioned, 2-methylpentan-3-one, 33, 34, ethanol, and NaOEt will all be in solution. The C]C unit of enolate anion 34 is more highly substituted and assumed to be thermodynamically more stable. Therefore, once it is established, the equilibrium will shift to favor the thermodynamically more stable enolate anion 34 (hence the term thermodynamic control). This discussion can be summarized by saying that strong bases (e.g., LiNH2 or LiNR2) favor kinetic control, since they generate weak conjugate acids, but strong bases (e.g., NaOH, NaOMe, NaOEt, or NaOt-Bu) will favor thermodynamic control because they generate conjugate acids that are stronger than the starting ketone. O Me

Me

Ha

Me 34

O

NaOEt, EtOH

EtOH

Ha

O

NaOEt, EtOH

Me

Me Me

Hb

2-Methylpentan-3-one

Me

Me

EtOH

Me 33

Hb

672

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

The counterion (M+) of the basic anion in M+NR2 will be the counterion of the enolate anion unless another metal ion is added. The covalent character of the MdO bond in the enolate anion plays a role in the kinetic versus thermodynamic equilibrium. If the MdO bond is relatively covalent (as with the LidO bond), the initially formed enolate (33) will form more easily and be less likely to react with an acid to yield 2-methylpentan-3-one, which favors the kinetic process. If the MdO bond is more ionic (M ¼ Na, K), the ionic character of the enolate increases and it becomes increasingly easier to reprotonate the anionic oxygen to yield 33. Using a metal that has significant ionic character promotes the equilibrium. This relationship is summarized and reinforced by noting that sodium and potassium enolates equilibrate readily, whereas Mg, Li, Zn, Cu, or Al enolates favor kinetic control.56 Reaction temperature is another important variable. If the reaction is kept cold, all reactions (including the forward and reverse reactions that constitute the equilibrium) are slowed because of less energy available to overcome the Eact barriers. If there is enough energy at low temperatures to convert 2-methylpentan-3-one to 33, the subsequent acid-base reactions (that will promote the equilibrium) will be slower, which favors kinetic control. Typical reaction temperatures are 78°C (CO2 in acetone or propan-2-ol) and 100°C (ether in CO2).57 Conversely, high-reaction temperatures promote the equilibrium process. In most cases, the temperatures observed with thermodynamically controlled reactions are the reflux temperatures of the solvent (refluxing ethanol, water, methanol, tert-butanol). The final parameter in this process is the length of time the acid and base are allowed to react. If the reaction time is short (typically minutes to hours) and the half-life of the reaction is longer, it is more difficult for the equilibration process to develop so kinetic control is favored. In other words, the acid-base reaction of the conjugate acid with the conjugate base is slow relative to the reaction of the ketone with the base. If the reaction is allowed to continue for a long time (typically hours to days, although minutes can be a long time in some reactions), equilibration and thermodynamic control is more likely. The time factor is linked to the relative strength of all acids in the system. When 33 is formed by reaction of LDA and 2-methylpentan-3-one, its conjugate acid is diisopropylamine. Diisopropylamine is a weak acid, and when given a sufficiently long contact time with a strong base (the enolate anion is a strong base) some acid-base equilibration can occur. Few reactions are quantitative, and virtually none are instantaneous. Remember that about six half-lives are required for a reaction to be >98% complete. Therefore, unreacted starting material will be available during the course of the reaction. Further, if the conversion of 2-methylpentan-3-one to 33 occurs in only 90%, then 10% of unreacted ketone remains. With a pKa of
21, 2-methylpentan-3-one is acidic enough to react with the enolate anion to reprotonate 33 and establish the equilibrium required for thermodynamic control, if given sufficient time. If the reaction itself is slow, enolate anion 33 is being formed in the presence of unreacted 2-methylpentan-3-one and there is the possibility of a reaction. The terms short and long are obviously relative and vague, and vary with the electrophile for a particular enolate anion. In general, short reaction times favor kinetic control and long reaction times favor thermodynamic control. The intramolecular cyclization reaction of bromoketone 6-bromo-3-phenylhexan-2-one58 illustrates how changing reaction conditions can influence formation of the kinetic or thermodynamic enolate. The reaction of 6-bromo-3-phenylhexan-2-one with LDA (ether, 0°C) generates the kinetic enolate (35). Subsequent intramolecular displacement of the bromide by the enolate anion yields a cyclohexanone, 2-phenylcyclohexan-1-one. Some elimination of 35 under the reaction conditions occurred, giving 3-phenylhex-5-en-2-one. A complicating factor in this example is that the oxygen in 35 is also a nucleophile, and O-alkylation (discussed in Section 13.3.2) gave the pyran product 6-methyl-5-phenyl-3,4-dihydro-2H-pyran, although 2-phenylcyclohexan-1-one was the major product. When 6-bromo-3-phenylhexan-2-one was treated with tert-butoxide in tert-butanol, the thermodynamic enolate (36) was formed and cyclized to the dihydropyran, 6-methyl-5-phenyl-3,4-dihydro-2H-pyran, via O-alkylation. A four-membered ring product [1-(1-phenylcyclobutyl)ethan-1-one] would be the normal enolate alkylation product that is analogous to 2-phenylcyclohexan-1-one, but its formation has a relatively high activation barrier and it was a minor product (Sections 1.5.2 and 4.5.1).

56

(a) Stork, G.; Rosen, P.; Goldman, N.; Coombs, R. V.; Tsuji, J. J. Am. Chem. Soc. 1965, 87, 275. (b) Patterson, J. W., Jr.; Fried, J. H. J. Org. Chem. 1974, 39, 2506.

57

Gordon, A. J.; Ford, R. A. The Chemist’s Companion; John: London, 1972; p 451.

58

(a) House, H. O.; Philips, W. V.; Sayer, T. S. B.; Yau, C.-C. J. Org. Chem. 1978, 43, 700. (b) Etheredge, S. J. Ibid. 1966, 31, 1990.

673

13.2 FORMATION OF ENOLATE ANIONS

CH3

Ph

Ph O

t-BuOK , t-BuOH

O

LDA , 2–4°C

CH2

Ether–Hexane

CH3

Br Br

Br

6-Bromo-3-phenylhexan-2-one

36

Me

O

Ph

Ph

O

Ph

Ph

Me

+

O

35

O

+

Me

+

Ph

O

Ph

O

6-Methyl-5-phenyl3,4-dihydro-2H-pyran

Me

1-(1-Phenylcyclobutyl)ethan-1-one

2-Phenylcyclohexan-1-one

3-Phenylhex-5en-2-one

6-Methyl-5-phenyl3,4-dihydro-2H-pyran

Carbonyl compounds other than ketones readily form enolate anions. Data in Table 13.2 show that the α-hydrogen of an aldehyde was more acidic than that of a ketone. When aldehydes are used, however, there is only one acidic proton, and in principle, either kinetic or thermodynamic control conditions can be used. Formation of (E)- or (Z)isomers of 37 remains a problem, however. It may be desirable to use kinetic control conditions to control geometry, or minimize subsequent reactions (e.g., nucleophilic addition of the conjugate base to the carbonyl, an aldol-type condensation), which is more facile with aldehydes than with ketones (Section 13.4.1). A nonnucleophilic base (e.g., LTMP) often maximizes formation of the enolate anion. Acid derivatives (e.g., esters and N,N-disubstituted amides) have only one acidic α-proton, and deprotonation yields a single regioisomeric enolate. Methyl butanoate reacted with LDA, for example, to give enolate 38 as a mixture of (E)- and (Z)-isomers. Self-condensation is a problem with esters (see Section 13.4.2.1), so kinetic is usually favored with these systems.

N Li

H

H

THF

O Butanal

O Li 37

OMe O

N Li

OMe O Li

THF

Methyl butanoate

38

When primary or secondary amides are treated with a base, there is a complicating reaction that was not possible with esters, ketones, or aldehydes. The NdH moiety is acidic enough to react with the bases used for deprotonation. Treatment of 39 with base gave the N-lithio derivative, but the α-lithio derivative (40) can be generated by addition of 2 equiv. of base.59 Enolate anion formation was straightforward with tertiary amides (e.g., dimethylisobutyramide, 39, R ¼ Me), and the resultant enolate anion (41) reacted with butanal to give amido-alcohol 42 in 68% yield59 (see Section 13.4.2).

N

Li

2 LDA

N R=H

40

59

R

LDA

N R = Me

Me

Me

Me 39

O

O Li

O

O Li

41

R

1. C3H7CHO 2. H3O+

HO n-C3H7

N

R

Me 42 (68%)

(a) Cuvigny, T.; Hullot, P.; Larchev^eque, M.; Normant, H. C. R. Hebd. Seances Acad. Sci. Ser. C 1974, 279, 569. (b) Hullot, P.; Cuvigny, T.; Larchev^eque, M.; Normant, H. Can. J. Chem. 1976, 54, 1098. (c) Crouse, D. N.; Seebach, D. Chem. Ber. 1968, 101, 3113.

674

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

When an ester has an acetyl at the β-position, as with methyl acetoacetate, it is possible to form a dianion. Subsequent reactions will occur first at the site of the second deprotonation, which is more nucleophilic. In the synthesis of (+)-gephyrotoxin by Vanucci-Bacque and coworkers,60 methyl acetoacetate (methyl 3-oxobutanoate) was treated with 2 equiv. of LDA to give dianion 61. The most acidic hydrogen between the 2 carbonyl groups was removed first, and the less acidic proton adjacent to the methyl group second to give dianion 43. This latter carbon is more nucleophilic, and reaction with the allylic bromide followed by quenching with an aqueous workup led to 44 in 79% yield. O

MeO

CH3 2 LDA O

O

MeO

CH2 O

Methyl 3-oxobutanoate

O

O

O

Br

MeO

O

O

O

43

44 (79%)

13.3 REACTIONS OF ENOLATE ANIONS WITH ELECTROPHILES There are two major reactions of enolates derived from aldehydes or ketones: (1) displacement reactions with alkyl halides or other suitable electrophiles and (2) nucleophilic acyl addition to carbonyl compounds. Reaction of enolate anion 43 with an allylic bromide to yield 44 is a simple example. Indeed, enolate anions are carbon nucleophiles, and react similarly to the alkyne anions discussed in Section 13.3.3. The reactions discussed in this section are solutionphase enolate anion reactions. Solid-phase enolate anion reactions61 have been investigated62 by high-resolution magic angle spinning NMR spectroscopy.

13.3.1 Enolate Alkylation An enolate anion reacts as a nucleophile, and both the oxygen and carbon are nucleophilic (Section 13.3.2), but the reaction at carbon usually predominates. Unlike Grignard reagents and organolithium reagents, enolate anions react easily with the electrophilic carbon of an alkyl halide.63 As with acetylides (Section 11.3), reaction of alkyl halide with an enolate anion forms a new carbon-carbon bond with concomitant formation of a metal halide (MdX). Many synthetic applications involve an enolate alkylation. In the Comins and Tsukanov64 synthesis of alkaloid 205B, 45 was treated with sodium hexamethyldisilazide to generate the enolate anion, and addition of an excess of iodomethane gave the alkylated ketone 46 in 85% yield, with 7–9:1 selectivity for the diastereomer shown. Note that HMPA was added to facilitate formation of the anion and also to assist the alkylation step. Another example is the reaction of 47 with LDA to yield enolate 48, where the Li is chelated with the nitrogen lone pair electron.65 The facial bias imposed by the stereogenic centers, and the relatively rigid conformation of 48 led to reaction from the bottom face (the α-face in steroid nomenclature). Addition of 1-bromocyclohex-2-ene from this less sterically hindered face led to 49 with high diastereoselectivity (>95:95% yield. In this study, the deuterated acid was the target, and hydrolysis of di-tertbutyl (E)-2-(hex-4-en-1-yl)malonate, to the diacid with trifluoroacetic acid was followed by α-deuteration via treatment of the diacid with NaOD in D2O, yielding (E)-2-(hex-4-en-1-yl)malonic-d acid-d2. Heating (E)-2-(hex-4-en-1-yl)malonic-d acid-d2 in the NaOD/D2O solution led to decarboxylation as well as incorporation of a second α-deuterium, and isolation of (E)-oct-6-enoic-2,2-d2 acid-d in 83% yield from di-tert-butyl (E)-2-(hex-4-en-1-yl)malonate. The second classical reaction mentioned above is the acetoacetic ester synthesis.70b,72 In this sequence, an ester of acetoacetic acid (ethyl 3-oxobutanoate), such as ethyl acetoacetate, is treated with base under thermodynamic control conditions and alkylated, as with the malonic ester synthesis. Reaction with sodium ethoxide in ethanol (an ethyl ester is being used) gave the enolate anion, and quenching with benzyl bromide led to ethyl 2-benzyl-3-oxobutanoate. Saponification and decarboxylation gave a substituted ketone (4-phenylbutan-2-one). Although the malonic ester synthesis and the acetoacetic ester synthesis are fundamentally similar, the different substrates lead to formation of either a highly substituted acid or a ketone. The reaction is not restricted to acetoacetate derivatives, and any β-keto ester can be used.73

67

Li, Y.-J.; Hou, C.-C.; Chang, K.-C.; Li, Y.-J.; Hou, C.-C.; Chang, K.-C. Eur. J. Org. Chem. 2015, 1659.

68

Granger, B. A.; Jewett, I. T.; Butler, J. D.; Martin, S. F. Tetrahedron 2014, 70, 4094.

69

(a) Meyer, K. H. Berichte 1912, 45, 2864. (b) Hauser, C. R.; Hudson, B. E., Jr. Org. React. 1942, 1, 266.

70

(a) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-57. (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 406, 407. 71

Sulikowski, G. A.; Agnelli, F.; Spencer, P.; Koomen, J. M.; Russell, D. H. Org. Lett. 2002, 4, 1447.

72

(a) Krauch, H.; Kunz, W. Organic Name Reactions; Wiley: New York, 1964; pp 1, 2. (b) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-1. (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 2, 3. 73

Cochrane, J. S.; Hanson, J. R. J. Chem. Soc. Perkin Trans. 1 1972, 361.

677

13.3 REACTIONS OF ENOLATE ANIONS WITH ELECTROPHILES

The key to all of these reactions is the enhanced acidity of the hydrogen atom on the carbon between the two carbonyls. As noted in Table 13.2, 1,3-dicarbonyl derivatives have a pKa in the range of 8–12 for that proton. The enolate anion formed from a 1,3-dicarbonyl compound is somewhat less reactive because the enolate is more stable due to resonance delocalization. For this reason, higher reaction temperatures are usually (but not always) required. A typical experiment will use the reflux temperature of the alcohol solvent being used. Another alternative is to use a complexing agent (e.g., HMPT or HMPA, see above), or DABCO (diazabicyclo[2.2.2]octane). CO2t-Bu NaH, THF

CO2t-Bu

CO2t-Bu I

CO2t-Bu Di-tert-butyl malonate

1. TFA 2. D2O MeCN cat NaOD

Di-tert-butyl (E)-2-(hex4-en-1-yl)malonate (95%)

1. NaOEt, EtOH

CO2Et

CO2D D D (E)-Oct-6-enoic(83%) 2,2-d2 acid-d

(E)-2-(Hex-4-en-1-yl)malonic-d acid-d2

O

O

Heat

CO2D D CO2D

O

1. aq NaOH 2. H3O+

CO2Et

3. Heat (– CO2)

2. BnBr

Ph

Ph Ethyl 2-benzyl-3-oxobutanoate

Ethyl 3-oxobutanoate

4-Phenylbutan-2-one

Variations of the malonic ester and acetoacetic ester sequences lead to many useful synthetic opportunities. In the examples quoted, the base-solvent pair used was ethanol-sodium ethoxide where the alkoxide is the conjugate base of the solvent. If NaOEt-EtOH were used with a methyl ester, transesterification74 would give a mixture of methyl and ethyl esters as products. For both malonic ester and acetoacetic ester removal of the most acidic proton (α to both carbonyls) also gives the more thermodynamically stable enolate. Either NaOEt-EtOH or LDA-THF will generate the desired enolate. The malonic ester synthesis is most useful for the synthesis of highly substituted monoacids, and the acetoacetic ester synthesis is used to prepare substituted methyl ketones. The disconnections for the malonic and acetoacetic ester syntheses follow: R

R X

CO2H

O

+ HO C 2

CO2H

O R1

R

R

+

CO2H

R1 X

13.3.2 O-Versus C-Alkylation of Enolate Anions As discussed in Section 13.2.5, cyclization of 6-bromo-3-phenylhexan-2-one gave 2-phenylcyclohexan-1-one via an intramolecular alkylation reaction of an enolate anion. However, vinyl ether 6-methyl-5-phenyl-3,4-dihydro-2H-pyran was also formed during this reaction, which illustrates an important competing reaction in the alkylation of enolates. These products were formed because an enolate anion is a bidentate nucleophile. In other words, it contains two nucleophilic centers, the carbanionic center and the oxygen anion as illustrated by resonance contributors 57. Reaction of the O-nucleophile in 57 with an alkyl halide leads to the vinyl ether 58. Reaction at the C-nucleophile of 57 generates the usual alkylation product 59. Since a carbanion is usually more nucleophilic than an alkoxy anion for common electrophiles, 59 tends to be the major product. R R1 O

O

R1 X

O

R

R 58

R 57

+

C −

O

O

R1 X

C −

R1

R 59

There are several factors that can influence the relative proportion of these products. When 59 is the major product, the process is referred to as C-alkylation and it is greatly preferred in ether solvents when lithium enolate anions are 74

(a) Smith, M. B. March’s Advanced Organic Chemistry, 7th ed.; Wiley: Hoboken, NJ, 2013; pp 1206–1208. (b) Koskikallio, J. In The Chemistry of Carboxylic Acids and Esters; Patai, S., Ed.; Wiley-Interscience: New York, 1969.

678

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

used. Indeed, ether solvents promote the nucleophilic nature of the enolate anion, favoring the carbanion. A lithium enolate anion forms a relatively covalent OdLi bond, again favoring C-alkylation. When very reactive halides (iodomethane, allyl bromide) react with sterically unhindered enolate anions (usually with a K or Na gegenion), O-alkylation (formation of product 58) is preferred. The highly ionic OdNa or OdK bond effectively increases the nucleophilic character of the oxygen. When the alkylation occurs intramolecularly, O-alkylation can be a much more serious problem. Approach of the carbon nucleophile to the electrophilic center is sometimes restricted due to the nature of the tether (see Section 10.6.1 for Baldwin’s rules). When the highly polar solvent DMSO is used, the enolate anion is largely monomeric, which increases reactivity and favors O-alkylation. The product distribution depends on the substitution pattern of the enolate anion and the nature of the counterion.75 In DMSO, the ratio of O- to C-alkylation is relatively insensitive to changes in the gegenion (the metal counterion), suggesting that the enolate anion is an unencumbered anion in this cation-solvating medium. Note that C-alkylation is preferred except when the steric bulk at the nucleophilic carbon becomes too great, O-alkylation is effectively the only process. Enolate anions react with acyl halides or anhydrides to yield the acylated product. Both C- and O-acylation are possible. In general, O-acylation predominates (Section 13.4.2).76 If C-acylation is desired, the acyl unit must be modified. One solution is to use acyl cyanides. In an example taken from the Nicolaou et al.77 synthesis of coleophomone B, acyl cyanide 61 was prepared from aldehyde 60 by treatment with Et2AlCN (Nagata’s reagent)78 followed by oxidation of the resulting alcohol with PCC (Section 6.2.2.2). When 61 was treated with 5-methylcyclohexane-1,3-dione and triethylamine, C-acylation proceeded smoothly to give 62 in 98% yield. In many reactions, O- versus C-acylation is very dependent on the local environment and electronic effects within the enolate anion.79 O

O

O

O

CHO

CN

1. Et2AlCN

O O

O

O NEt3, rt

2. PCC

O

O2C( p-BrC6H4)

O2C( p-BrC6H4) 60

O2C( p-BrC6H4)

61

62 (98%)

This type of reaction can sometimes be useful in synthesis, but the structural units on the substrate influence Cversus O- reactivity. The synthesis of (+)- and ()-saudin by Boeckman et al.80 employed an interesting variation that involved O-alkylation. Treatment of 63 with potassium hexamethyldisilazide removed a proton from the distal methyl group (it is acidic due to vinylogy that is made possible by the conjugated enone system), and subsequent reaction with the allylic triflate shown gave enol ether 64 in good yield. Apart from the use of a vinylogous enolate anion, this example illustrates that triflates also yield an O-alkylated product.

CO2Me

O

1. KHMDS, THF–HMPA, –78°C 2. –78°C ,

OTf

OSiPh2t-Bu O

OSiPh2t-Bu

63

CO2Me

64

To put this problem into perspective, when ether or THF are used as the solvent, O-alkylation is a much less serious problem in reactions with alkyl halides. With the possible exception of iodomethane, the use of lithium enolates in ether solvents leads to C-alkylation as the major product in virtually all cases. When the enolate carbanion center is sterically hindered, O-alkylation can be a problem even in THF or ether. Some reagents (e.g., silyl halides and anhydrides) show a preference for O-alkylation. Both of these O-alkylation reactions will be discussed in Section 13.3.2.

75

Zook, H. D.; Miller, J. A. J. Org. Chem. 1971, 36, 1112.

76

See Krapcho, A. P.; Diamanti, J.; Cayen, C.; Bingham, R. Org. Synth. Coll. Vol. 1973, 5, 198.

77

Nicolaou, K. C.; Vassilikogiannakis, G.; Montagnon, T. Angew. Chem. Int. Ed. 2002, 41, 3276.

78

Nagata, W.; Yoshioka, M.; Murakami, M. Org. Synth. Coll. Vol. 1988, 6, 307.

79

Honda, T.; Namiki, H.; Kudoh, M.; Watanabe, N.; Nagase, H.; Mizutani, H. Tetrahedron Lett. 2000, 41, 5927.

80

Boeckman, R. K., Jr.; Ferreira, M.d R. R.; Mitchell, L. H.; Shao, P. J. Am. Chem. Soc. 2002, 124, 190.

679

13.3 REACTIONS OF ENOLATE ANIONS WITH ELECTROPHILES

13.3.3 Enolate Anion Reactions With Noncarbonyl Electrophiles Enolate anions react with many electrophilic species other than alkyl halides, analogous to the reactions of other nucleophilic carbon species (Chapter 11). Nucleophilic ring opening of an epoxide to yield alcohols is very common,81 occurring at the less sterically hindered carbon. Grieco et al.82 reacted the dilithio salt of acetic acid (LiCH2CO2Li, Section 13.4.2.7) with epoxide 65 to give 66, in a synthesis of bigelovin. The regioselectivity in the ring opening is probably due to a combination of neighboring group effects with the adjacent hydroxyl group, and steric blocking of one carbon by the bridgehead methyl. Me

Me H

H 1. LiCH2CO2Li

OH BnO

Me

OH

2. H3O+

Me

BnO

O

OH

65

CO2H

66

The epoxide ring opening yields the interesting disconnection (see below). OH

O

O

O +

R

X

R

X

R

R

R

R

It is also possible to oxygenate an enolate anion or an enol ether to yield the corresponding acyloin (an α-hydroxy ketone) or alkoxy derivative. In Danishefsky and coworker’s83 synthesis of myrocin C, reaction of an enol ether (e.g., 67) with 3,3-dimethyldioxirane (DMDO) gave the protected α-hydroxy ketone (68). Reaction of an enolate anion with oxygen and trimethylphosphite [(MeO)3P] gives an α-hydroxy derivative.84 An effective oxygenation reagent (MoOPh) for enolate anions can be used with enolate anions of several different carbonyl compounds. In a synthesis of metasequirin-B tetramethyl ether diacetates, Gong and coworkers85 treated lactone 69 with lithium hexamethyldisilazide and then MoOPh to give 70 in >88% yield, >15:1 selectivity for the diastereomers shown. The actual structure of MoOPh is MoO5•Py•HMPA.86 H

O Me

O

Me

O

H TBSO O

TBSO

H

O

H

O

67

O

68

MeO

MeO OTBS

OTBS

MeO

MeO O

1. KHMDS, THF, –20°C

O

O

2. MoOPh, THF

MeO

O MeO

MeO

OH MeO

69

70 (>88%)

81

For a review of such reactions, see Taylor, S. K. Tetrahedron 2000, 56, 1149.

82

Grieco, P. A.; Ohfune, Y.; Majetich, G. J. Org. Chem. 1979, 44, 3092.

83

Chu-Moyer, M. Y.; Danishefsky, S. J.; Schulte, G. K. J. Am. Chem. Soc. 1994, 116, 11213.

84

For an example taken from a synthesis of brefeldin A, see Corey, E. J.; Wollenberg, R. H.; Williams, D. R. Tetrahedron Lett. 1977, 2243.

85

Wang, P.-S.; Zhou, X.-L.; Gong, L.-Z. Org. Lett. 2014, 16, 976.

86

(a) Mimoun, H.; Seree de Roch, L.; Sajus, L. Bull. Soc. Chim. Fr. 1969, 1481. (b) Regen, S. L.; Whitesides, G. M. J. Organomet. Chem. 1973, 59, 293.

680

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

Nozoe and coworkers87 showed that 2-phenyl-2-p-toluene-sulfonyloxazidine (Section 6.9.1.2) could convert the enolate anion derived from a 2-pyrrolidinone derivative to the 3-hydroxy derivative (α-hydroxylation) in good yield. Another reagent that reacts with enolate anions to generate α-hydroxy ketones is the Davis reagent (2-phenylsulfonyl3-phenyl-oxaziridine).88 A variety of enolate anions, including those derived from carboxylic acid derivatives, can be hydroxylated in this manner. In a synthesis of (+)-przewalskin B (72) by Zhang and coworkers,89 the lactone enolate anion generated from lactone 71 by reaction with LDA was treated with the Davis reagent to give a 59% yield of the α-hydroxy derivative 72 [(+)-przewalskin B]. O

O

O

O

OH

1. LDA, THF, –78°C

O

O

O

2. PhO2S

, THF, –78°C

N

Ph

71

72 (59%)

Enolate anions react with sulfenyl halides to yield an α-alkylthio ketone. As shown for cyclohexanone, reaction with LDA followed benzenesulfenyl chloride (PhSCl) produced phenylthio derivative 2-(phenylthio)cyclohexan-1-one. The α-proton adjacent to both the carbonyl and the sulfur was more acidic, and deprotonation with potassium hydride gave the enolate. Subsequent alkylation with iodobutane gave a 91% yield of 2-butyl-2-(phenylthio)cyclohexan-1one.90 As mentioned in Section 3.7.4, the sulfide moiety can be oxidized to a sulfoxide and thermally eliminated (syn elimination) to yield an alkene, 2-butylcyclohex-2-en-1-one.91 Similar methodology can be applied to formation and syn-elimination of the corresponding selenides, dCOdCH(R)dSedR1.92 O

O 1. LDA

SPh

1. KH, THF 2. Bu—I, 25°C

2. PhSCl THF/DMF

Cyclohexanone

O

O SPh

2-(Phenylthio)cyclohexan-1-one

Bu

[O]

Bu

Heat

2-Butyl-2-(phenylthio)cyclohexan-1-one

2-Butylcyclohex2-en-1-one

(91%)

In addition to the methods described for controlling the ratio of thermodynamic and kinetic products, other techniques have been developed to trap the enolate anion, based on reactions with reagents that prefer O-alkylation. Reaction of a ketone with acetic anhydride, usually in the presence of a catalytic amount of perchloric acid, generates the thermodynamic enol acetate 73.93 When 73 was treated with methyllithium, acyl addition to the acetyl group led to the thermodynamic enolate anion (74) in high yield. Once generated in situ, 74 was treated with an alkyl halide or a carbonyl derivative in the usual manner. Some isomerization occurs on treatment with methyllithium, but the major product of this process is that resulting from alkylation of the thermodynamic enolate anion. Ac2O

O

Trace HClO4

2-Methylcyclohexanone

87

Me

Me Me

Me

MeLi

O

Li O

O 73

74

Ohta, T.; Hosoi, A.; Nozoe, S. Tetrahedron Lett. 1988, 29, 329.

88

(a) Davis, F. A.; Chen, B.-C. Chem. Rev. 1992, 92, 919. (b) Also see Nagasaka, T.; Imai, T. Chem. Pharm. Bull. 1995, 43, 1081. (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis; 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005, p 755.

89

Zhuo, X.; Xiang, K.; Zhang, F.-M.; Tu, Y.-Q. J. Org. Chem. 2011, 76, 6918.

90

See Coates, R. M.; Pigott, H. D.; Ollinger, J. Tetrahedron Lett. 1974, 3955.

91

(a) Trost, B. M.; Salzmann, T. N. J. Org. Chem. 1975, 40, 148. (b) Grieco, P. A.; Reap, J. J. Tetrahedron Lett. 1974, 1097.

92

(a) Grieco, P. A.; Miyashita, M. J. Org. Chem. 1974, 39, 120. (b) Reich, H. J.; Renga, J. M.; Reich, I. L. J. Org. Chem. 1974, 39, 2133.

93

(a) House, H. O.; Kramar, V. J. Org. Chem. 1963, 28, 3362. (b) House, H. O.; Phillips, W. V.; VanDerveer, D. Ibid. 1979, 44, 2400.

681

13.4 ENOLATE CONDENSATION REACTIONS

Trialkylsilyl halides show a great propensity to react with the oxygen rather than the carbon of enolate anions. Stork and Hudrlik94 showed that enolate anions can be trapped as the trialkylsilyl enol ether via O-alkylation, which is most useful for kinetic enolates in which a lithium enolate (e.g., the kinetic enolate derived from 2-methylcyclohexanone and LDA) is reacted with trimethylsilyl chloride to give an isolable silyl enol ether, 75. The enolate anion is trapped with high efficiency, and conversion to the enolate anion is readily accomplished by treatment with methyllithium to generate kinetic enolate anion 76 and the volatile trimethylsilane (Me3SiH).95 This latter reaction is subject to some equilibration, but gives the kinetic alkylation product in good yield. Silyl enol ethers96 can be generated by other methods. To generate kinetic enolate anions, silyl enol ethers can be isolated and separated from amine by products and unreacted ketone. Unmasking and alkylation can be accomplished under controlled conditions that minimize equilibration, maximizing the yield of kinetic product. Me O

Me

Me 1. LDA DME, 0°C 2. Me3SiCl DME/NEt3, 0°C

2-Methylcyclohexanone

MeLi

Li O

Me3SiO 75

(74% yield , 99% pure)

76

13.4 ENOLATE CONDENSATION REACTIONS The condensation reaction of enolate anions and carbonyl derivatives is one of the most useful in organic chemistry. The condensation is nothing more than an acyl addition reaction of the nucleophilic enolate to an electrophilic carbonyl carbon.97 If an enolate anion of a ketone or aldehyde reacts with the carbonyl of an aldehyde or a ketone, an acyl addition reaction yield a β-hydroxy aldehyde or ketone (Section 13.4.2). There are many synthetic variations, and to further complicate matters the enolate may be generated under kinetic or thermodynamic conditions. The more common variations of this reaction are well known, and usually have a name associated with them (i.e., named reactions).98–101

13.4.1 The Aldol Condensation 13.4.1.1 Intermolecular Reactions The aldol condensation,102,103 is one of the more important classical reactions of organic chemistry, and was reported urtz103b and by Perkin.103c Prior to W€ urtz’s103b report, the composerby Chiozza in 1856103a and later expanded by W€ 104 chemist Alexander Borodin reported an acid-catalyzed variation of this reaction. In the most commonly used early versions, an aldehyde was mixed with a base and an aldehyde that did not contain an enolizable hydrogen atom (e.g., benzaldehyde), under what it now categorized as thermodynamic conditions. Ketones were also used. If acetophenone reacted with benzaldehyde in the presence of sodium ethoxide (in ethanol heated at reflux), the initial product was alkoxide 77. Hydrolysis provided the β-hydroxy ketone product (an aldol or aldolate) 3-hydroxy-1,3diphenylpropan-1-one. In general, the hydrolysis step provides the aldolate, but sometimes elimination of water (dehydration) accompanies the hydrolysis to yield a conjugated ketone [1,3-diphenylprop-2-en-1-one, otherwise 94

Stork, G.; Hudrlik, P. F. J. Am. Chem. Soc. 1968, 90, 4464, 4462.

95

House, H. O.; Czuba, L. J.; Gall, M.; Olmstead, H. D. J. Org. Chem. 1969, 34, 2324.

96

Rasmussen, J. K. Synthesis 1977, 91.

97

For reactions of this type, see Larock, R. C. Comprehensive Organic Transformations, 2nd ed.; Wiley-VCH: New York, 1999; pp 1317–1324.

98

The Merck Index, 14th ed.; Merck: Whitehouse, NJ, 2006; pp ONR-1-ONR-105.

99

Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005.

100

Krauch, H.; Kunz, W. Organic Name Reactions; Wiley: New York, 1964.

101

Surrey, A. R. Name Reactions in Organic Chemistry, 2nd ed.; Academic Press: New York, 1961.

102

(a) Modern Aldol Reactions; Mahrwald, R., Ed.; Wiley: New York, 2004. (b) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-1. (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 26, 27.

103

(a) Chiozza, L. Annalen 1856, 97, 350. (b) Wurtz, A. Compt. Rend. 1872, 74, 1361. (c) Perkin, W. Berichte 1882, 15, 2802. (d) Nielsen, A. T.; Houlihan, W. J. Org. React. 1968, 16, 1. (e) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; WileyInterscience: Hoboken, NJ, 2005; pp 8, 9. (f ) Reference 104, pp 6–10.

104

Borodin’s earliest results are reported in von Richter, V. Ber. Dtsch. Chem. Ges. 1869, 2, 552. See Borodin, A. Ber. Dtsch. Chem. Ges. 1873, 6, 982.

682

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

known as (E)-chalcone]. Note that heating is usually required to induce dehydration unless there is a conjugating substituent. When an aromatic aldehyde (e.g., benzaldehyde) is the reaction partner, the aqueous acid workup almost always leads to elimination of water from the aldol product. If an aliphatic aldehyde is used, the aldol product is usually easy to isolate. When an aromatic aldehyde is condensed with an enolate of an aliphatic aldehyde or ketone to yield the α-,β-unsaturated compound, the reaction is called the Claisen-Schmidt reaction.103d,105 Note that the aldol condensation can be done using high-intensity ultrasound,106 where good yields of aldol products were isolated that normally yield significant elimination under conditions that did not use sonication. O Ph

NaOEt, PhCHO EtOH, Reflux

CH3

O

O

Ph

O

H3O+

Ph

Ph

O

Heat – H 2O

Ph

Ph

3-Hydroxy-1, 3-diphenylpropan-1-one

77

Acetophenone

OH

Ph (E)-Chalcone

Self-condensation is a major drawback of the reaction as originally formulated. This reaction is a particular problem when the goal is a mixed-aldol condensation, which is the reaction of an enolate anion of one aldehyde or ketone with a different aldehyde or ketone. An example is the reaction of acetophenone with benzaldehyde, in which the enolate derived from acetophenone is in equilibrium with unreacted acetophenone when treated with sodium ethoxide in ethanol. At equilibrium, this enolate anion can react with either benzaldehyde or with another molecule of acetophenone, which is the self-condensation. Therefore, there are two possible aldol products, 3-hydroxy-1,3diphenylbutan-1-one and 3-hydroxy-1,3-diphenylpropan-1-one. For many years, these competing reactions limited the utility of the aldol condensation, and good yield of a mixed-aldol condensation was possible only when an aldehyde or a ketone with no α-hydrogen atom (e.g., benzaldehyde) was used as a reaction partner. A mixed-aldol, also known as a crossed-aldol condensation, is often a desirable target when the two reactive partners have an enolizable position, as between pentan-3-one with cyclopentanone. If pentan-3-one and benzaldehyde, which has no enolizable protons, can lead to two products, what will happen in this new case? If an aldol condensation occurs under thermodynamic conditions, pentan-3-one reacts with sodium ethoxide to yield enolate anion 78. Under these conditions, cyclopentanone reacts to yield enolate anion 79. Under equilibrium conditions, both enolate anions (78 and 79) will be formed, and will be available for reaction. In other words, both enolate anions are in equilibrium with unreacted pentane-3-one and cyclopentanone. O O

NaOEt EtOH

O

Me

2. H3O+

1.

O

O Me

OH

Me

O

1.

80

OH Pentan-3-one

78

+

2. H3O+

+

81 O

O

NaOEt EtOH

O

O

2. H3O+

1.

OH 82 O

Cyclopentanone

79

1.

O

2. H3

O+

HO 83

Enolate anion 78 can condense with either unenolized pentan-3-one (to produce 80) or with unenolized cyclopentanone (to produce 81). In addition to the formation of 80 and 81, enolate anion 79 can condense with either the

105

(a) Claisen, L.; Claparède, A. Berichte, 1881, 14, 2460. (b) Schmidt, J. G. Ibid. 1881, 14, 1459. (c) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-18.

106

Cravotto, G.; Demetri, A.; Nano, G. M.; Palmisano, G.; Penoni, A.; Tagliapietra, S. Eur. J. Org. Chem. 2003, 4438.

683

13.4 ENOLATE CONDENSATION REACTIONS

unenolized pentan-3-one (to produce 82) or with unenolized cyclopentanone (to produce 83), giving a total of four aldol products in this reaction. Diastereomers are possible for 80 and 81, increasing the actual number of products. The low yield of the desired aldolate, and problems associated with separation and isolation make this reaction unattractive. Even if one aldolate is produced with some selectivity, the problems of poor yields and isolation of the desired product remain. It is clear that this mixed-aldol condensation is not a synthetically useful reaction as described above. Both pentan-3-one and cyclopentanone are symmetrical, and there is an opportunity for only one enolate anion from each ketone, which prevents formation of additional products. If an unsymmetrical ketone is used in this mixed-aldol reaction, the problem of control is exacerbated. The reaction of 2-methylpentan-3-one with sodium ethoxide, under thermodynamic control conditions, generates two different enolate anions. Subsequent reaction with an aldehyde that has no α-hydrogen atoms (benzaldehyde) leads to two aldol products (84 and 85) if there is no self-condensation of 2-methylpentan-3-one. If 2-methylpentan-3-one were to react with sodium ethoxide under thermodynamic conditions in the presence of an unsymmetrical ketone (e.g., butan-2one), both the kinetic and thermodynamic enolate anions of both ketones can be formed. Therefore, four different enolate anions are formed and each one can react with two different ketones. An attempted mixed-aldol condensation of butan-2-one and 2-methylpentan-3-one should therefore produce eight different aldolate products. If elimination to the conjugated products occurs, even more products are possible. O

O

O

1. NaOEt, EtOH 2. PhCHO 3. Hydrolysis

OH

+ Ph

2-Methylpentan-3-one

Ph

OH 84

85

The possible formation of up to eight different aldol products is a significant disadvantage of the traditional aldol condensation, but there is the potential for a useful synthetic application. If selectivity can be controlled, the main advantage is the construction of a relatively complex carbon skeleton from simple precursors. In addition, this carbon-carbon bond-forming reaction generates new stereogenic centers, although racemic products will be formed. However, good-to-excellent diastereoselectivity is possible in many cases, as will be seen in Section 13.5.1.2. The reaction conditions used for the aldol condensation of pentan-3-one described above are rather harsh, which is a disadvantage if elimination of water from the aldol product occurs, or if self-condensation occurs under these equilibrating conditions. Aldehydes that do not contain an enolizable proton can be added to the enolate of a ketone to yield a cross-coupling product, but a large excess of the aldehyde is required to minimize self-condensation. One solution to this problem lies in changing from thermodynamic to kinetic control conditions. If pentan-3-one, for example, were treated with LDA in THF at 78°C, enolate 78 would be produced as the major enolate anion and in high yield. Under these conditions, enolate anion formation is largely irreversible if the reaction time is limited and a low-reaction temperature is maintained. In a subsequent step, cyclopentanone could be added to give a single aldol condensation product, 81. The advantage of kinetic control is the formation of primarily one enolate anion under nonequilibrating conditions, which minimizes self-condensation products. An example of a mixed-aldol condensation, under kinetic control conditions, is the reaction of pentan-2-one with LDA and subsequent condensation with butanal, which gave 6-hydroxynonan-4-one as the major product.107 If the thermodynamic aldolate is desired, the thermodynamic enolate must be trapped as the enol acetate (see Section 13.3.3) and then treated with a carbonyl compound after unmasking with methyllithium. 1. LiN(i-Pr)2, THF 2. Butanal

O Pentan-2-one

3. H3O+

O

OH

6-Hydroxynonan-4-one

There are countless synthetic examples of the aldol condensation. In Yu and coworker’s108 synthesis of broussoetine I, cyclopentanone was treated with LDA to generate the enolate anion, which reacted with aldehyde 86 to give a 98% yield of aldolate product 87. Enantioselective109 aldol condensation reactions (e.g., the transformation of cyclopentanone and 107

(a) Stork, G.; Kraus, G. A.; Garcia, G. A. J. Org. Chem. 1974, 39, 3459. (b) Gaudema, M. C. R. Acad. Sci. Ser. C 1974, 279, 961.

108

Zhao, H.; Kato, A.; Sato, K.; Jia, Y.-M.; Yu, C.-Y. J. Org. Chem. 2013, 78, 7896.

109

(a) Cordova, A.; Notz, W.; Barbas III, C. F. J. Org. Chem. 2002, 67, 301. (b) List, B.; Pojarliev, P.; Castello, C. Org. Lett. 2001, 3, 573. (c) Yoskikawa, N.; Kumagai, N.; Matsunaga, S.; Moll, G.; Oshima, T.; Suzuki, T.; Shibasaki, M. J. Am. Chem. Soc. 2001, 123, 2466. (d) Trost, B. M.; Silcoff, E. R.; Ito, H. Org. Lett. 2001, 3, 2497.

684

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

86 to 87) are rather common nowadays. Chiral bases can be used for enantioselective aldol condensation reactions. A simple example was reported by Doi and coworkers110 in a synthesis of apratoxin C. D-Proline was used as a chiral base to initiate enolate anion formation and the enantioselective cross-coupling of acetone with isobutyraldehyde, to give a 52% yield of 88. Stereoselectivity in aldol condensation reactions will be discussed in great detail in Section 13.5. OH

O

O

1. LDA, TMEDA, THF, –78°C

O CHO

2.

O

O O

86

CHO

87 (98%)

(20%)

+

O N H

OH

CO2H

DMSO, rt, 2 d

O

88 (52%)

Aldol condensation reactions can be done under acidic conditions. Borodin104 reported the self-condensation reaction of acetaldehyde to yield 3-hydroxybutanal when treated with acid. While the base-catalyzed reaction is more common, acid-catalyzed condensation reactions are well known. Lewis acids also catalyze the aldol condensation. In the presence of a catalytic amount of TiCl4,111 butan-2-one was condensed with benzaldehyde in toluene to give an 83% yield of the aldolate product. Under these conditions, the reaction was highly regioselective for the thermodynamic product. In another example, taken from the synthesis of siphonarin B by Patterson et al.112 aldehyde 89 was condensed with ketone 90 in the presence of Sn(OTf )2 to give a 92% yield of 91. The aldolate product formed diastereoselectively, generating a 2.7:1 mixture of syn,syn:(syn,anti + anti,anti)-diastereomers. BnO

OHC

i-PrEt2SiO

BnO

PMBO

O

+

TMSO

O

PMBO TMSO

CH2Cl2

i-PrEt2SiO 89

O HO

Sn(OTf)2, NEt3

O O

O

90

91 (92%)

Nevalainen and Simpura113 reported another variation of this classic reaction called the aldol-transfer reaction. In the presence of a suitable catalyst, usually an aluminum compound, an aldolate product reacts with an aldehyde to give a ketone and a new aldolate. An example is the reaction of benzaldehyde with aldolate 4-hydroxy-4-methylpentan-2-one in the presence of 5% of aluminum catalyst 92. In dichloromethane at ambient temperature, a 62% yield of aldolate 4-hydroxy-4-phenylbutan-2-one was obtained after a reaction time of 43 h. The other product of this reaction was acetone, which was readily removed. This transformation involves a retro-aldol reaction of 4-hydroxy-4-methylpentan-2-one (see Section 13.5.1.6) and the resultant enolate anion reacts with benzaldehyde. Several aldehydes were used and 4-hydroxy-4-methylpentan-2-one is particularly attractive (the aldol condensation product of acetone) because acetone is the second product.

HO Me

Me

O

O

+

O

Me

4-Hydroxy-4-methylpentan-2-one

Al2Me4

92

CH2Cl2, rt, 43 h PhCHO

Ph H

OH

O

O

+ Me

4-Hydroxy-4-phenylbutan-2-one

110

Masuda, Y.; Suzuki, J.; Onda, Y.; Fujino, Y.; Yoshida, M.; Doi, T. J. Org. Chem. 2014, 79, 8000.

111

Mahrwald, R.; G€ undogan, B. J. Am. Chem. Soc. 1998, 120, 413.

112

Paterson, I.; Chen, D. Y.-K.; Franklin, A. S. Org. Lett. 2002, 4, 391.

113

Simpura, I.; Nevalainen, V. Angew. Chem. Int. Ed. 2000, 39, 3422.

Me

Me

685

13.4 ENOLATE CONDENSATION REACTIONS

The aldol disconnection follows: HO

R2 R1

O

O

O +

R1

O

R

R2

R1

R

R

R2

13.4.1.2 Intramolecular Reactions The intramolecular aldol condensation is particularly useful, and it is illustrated by the simple reaction of hexane2,5-dione with LDA to first yield kinetic enolate 93. This enolate anion could condense with a second molecule of dione to yield an aldolate product, but intramolecular attack of the second carbonyl group is faster and leads to the cyclic aldolate product 94. This product yields 3-hydroxy-3-methylcyclopentan-1-one upon hydrolysis. A synthetic example is taken from Jacobsen and coworker’s114 synthesis of (+)-reserpine, in which intramolecular aldol condensation of 95 using piperidine as the base gave an 86% yield of 96, with a dr of >19:1. Note that enolate anion formation occurred at the carbon α to the aldehyde, consistent with an aldehyde α-proton being more acidic than that of a ketone. The course of the intramolecular aldol condensation can be influenced by steric and conformational factors, as well as the rate of competing cyclization reactions. O

O

O

LDA

O O

HO

H3O+

O

O

Hexane-2,5-dione

93

94

MeO

3-Hydroxy-3-methylcyclopentan-1-one

MeO N

N Ts

H

N

N

Piperidine, TsOH

H

Ts

Toluene

H

H

HO O

OPMB

OHC

OPMB

OHC OMe

OMe

95

96 (86%, >19:1 dr)

Acid-catalyzed, intramolecular aldol condensation reactions are known. In a synthesis of (+)-ent-spirocurcasone by Ito and coworkers,115 97 was stirred with concentrated H2SO4 acid in THF, which deprotected the acetal moiety to yield the aldehyde and the resulting intramolecular aldol condensation led to 98 in 76% yield. Note that water was eliminated from the initially formed aldolate during the course of the reaction.

O

H

conc H2SO4, THF

O (EtO)2HC

rt, 6 h

97

114

Rajapaksa, N. S.; McGowan, M. A.; Rienzo, M.; Jacobsen, E. N. Org. Lett. 2013, 15, 706.

115

Abe, H.; Sato, A.; Kobayashi, T.; Ito, H. Org. Lett. 2013, 15, 1298.

H

98 (76%)

686

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

Another example illustrates some of the problems inherent to the intramolecular reaction that do not arise with the intermolecular version. Deprotonation of 8-methylnon-8-ene-2,6-dione with sodium ethoxide under thermodynamic control conditions gave enolate anion 99 and isolated product 103 (water was lost from the initial aldolate product), which was used in Rouessac and Alexandre’s synthesis of isoshyobunone.116 Under these reaction conditions, four different enolate anions (99 and 104, 105, and 106) are possible from 8-methylnon-8-ene-2,6-dione. The intramolecular aldol condensation reaction with 104 and 106 would generate four-membered rings, which is energetically less favorable (Section 4.5.1) than formation of the six-membered rings that can be formed from the other enolate anions, 105 and 99. Dienone (103), the kinetic product relative to cyclization via 105, was also formed in the enolate condensation reaction of 8-methylnon-8-ene-2,6-dione. When enolate 99 closed to form 100, hydrolysis generated the aldolate 101. Dehydration gave the ketone with the C]C unit conjugated to the carbonyl group (102), but the isopropenyl doublebond remained out of conjugation. Under the acidic conditions of the workup, that double bond moved into conjugation to give the final product 103. Despite several possible reaction pathways, this intramolecular aldol cyclization proceeded with high selectivity for a single product. The cyclization is governed by the factors discussed in Section 4.5.1 for ring closures, which poses problems not encountered with the intermolecular aldol condensation. O

O

O

OH H3O+

Base

O 8-Methylnon-8-ene-2,6-dione

O

O 99

O 100

101

– H 2O

O 102 O

O

O

(4)

(6) O

103

O

(4)

O

104

O

(6) O

105

O

106

99

The disconnection for an intramolecular aldol condensation follows: O

O

O OH R

R

R O

13.4.2 Condensation Reactions of Acid Derivatives Just as ketone and aldehyde enolate anions react with ketones and aldehydes, they can also react with esters and other acid derivatives.117 Similarly, esters and other acid derivatives can be converted to their enolate anions, and then react with another acid derivative, or with an aldehyde or ketone. There are many variations of this reaction. When an acid derivative is attacked by a nucleophile, the alkoxide product contains a leaving group, which leads to acyl substitution as described for reactions of acid derivatives and Grignard reagents (Section 11.4.3.2) and organolithium reagents (Section 11.6.6).

116 117

(a) Alexandre, C.; Rouessac, F. J. Chem. Soc. Chem. Commun. 1975, 275. (b) Idem Bull. Chim. Soc. Fr. 1977, 117.

For reactions of this type, and related reactions, see Larock, R. C. Comprehensive Organic Transformations, 2nd ed.; Wiley-VCH: New York, 1999; pp 1528–1546.

687

13.4 ENOLATE CONDENSATION REACTIONS O

O

Base

Y

R

O

R

O

Y

R

O

O

X

X R1

R

R1

R Y

Y 107

108

109

110

O

O

N DMB

O

2. AcOH , THF –78°C → rt

N

N DMB

HO

1. LDA , THF , –78°C

Boc

O

N

DMB = 2,4-Dimethoxybenzyl

Boc

Br

111

Br

112 (92%)

This chemistry is summarized by the conversion of carbonyl derivative 107 to the enolate anion (108), and subsequent reaction with an acid derivative (acyl halides, anhydrides, or esters). Attack at the acyl carbon produces the usual tetrahedral intermediate (109), where the X group is a leaving group (Cl, O2CR, OR, NR2) and expulsion of X leads to a 1,3-dicarbonyl compound 110. This is, of course, the normal acyl substitution sequence. This section will examine those reactions where the carbonyl partner is an acid derivative and, in many cases, the enolate partner will also be an acid derivative. Esters, lactones, anhydrides, amides, or lactams can be used as the enolate anion precursor, reacting with aldehydes, ketones, or esters. Both intermolecular and intramolecular reactions are known. The remainder of this section will focus on a variety of carbonyl enolate precursors. To illustrate that virtually all acid derivatives can be used in this reaction, amide 111 was treated with LDA, and intramolecular condensation with the ketone unit gave a 92% yield of lactam-alcohol 112, in McWhorter and Liu’s118 synthesis of 8-desbromohinckdentine A. 13.4.2.1 The Claisen Condensation A classical reaction is the condensation of an ester enolate with an ester, illustrated by the self-condensation of ethyl butanoate when treated with sodium ethoxide to yield a β keto-ester,ethyl 2-ethyl-3-oxohexanoate. Initial reaction with the base under thermodynamic control in this case, generated the enolate anion (113). This anion attacked the carbonyl of a second molecule of ethyl butanoate to yield the tetrahedral intermediate 114. Displacement of ethoxide generated the ketone moiety in ethyl 2-ethyl-3-oxohexanoate. As shown here, this reaction is known as the Claisen condensation.119 NaOEt Reflux

CO2Et

CO2Et

O

CO2Et OEt

O EtO

Ethyl butanoate

113

CO2Et

– –OEt

O 114

Ethyl 2-ethyl-3-oxohexanoate

The reaction described for ethyl butanoate is a self-condensation, which simply means that one molecule of ethyl butanoate reacts with another molecule of ethyl butanoate. The reaction of two different esters is called the crossedClaisen (or a mixed-Claisen) condensation. As with the mixed-aldol condensation, a Claisen condensation of two different esters can result in a mixture of products under thermodynamic control conditions. A generalized reaction involving RCO2Et and R1CO2Et can lead to at least four different condensation products. One ester (RCO2Et) can condense with itself to yield 115 or with R1CO2Et to yield 116. Similarly, ester R1CO2Et can condense with itself to yield 117, or with RCO2Et to yield 118. As with the aldol condensation, kinetic control conditions are preferred if a particular crossed-Claisen product is desired. 118 119

Liu, Y.; McWhorter, W. W., Jr. J. Am. Chem. Soc. 2003, 125, 4240.

(a) Hellon, R.; Oppenheim, A. Berichte 1877, 10, 699. (b) Israel, A. Annalen 1855, 231, 197. (c) Claisen, L. Berichte 1887, 20, 655. (d) Beyer, C.; Claisen, L. Ibid. 1887, 20, 2178. (e) Claisen, L.; Stylos, N. Ibid. 1887, 20, 2188. (f ) Claisen, L. Annalen 1894, 281, 306. (g) Reference 100, pp 94–96. (h) Reference 101, pp 49–51. (i) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-18. (j) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 150, 151.

688

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

CO2Et

CO2Et R1

R

CO2Et R

R

R1

O

O 115

CO2Et R

O

116

R1

R1 O

117

118

A synthetic example that uses kinetic control conditions is taken from Panek and coworker’s120 synthesis of leucascandrolide A, in which the lithium enolate anion of tert-butyl acetate reacted with β-epoxy ester 119 in THF, to give a 70% yield of 120. Self-condensation of the lithium enolate with the parent ester is sometimes a problem when LDA is used as a base,121 but using the tert-butyl ester or LICA minimizes this competing process.122 O

O

–50 → 15°C

MeO 119

t-BuMe2SiO

O

CH3CO2t-Bu, LDA, THF

O

O

t-BuO 120 (70%)

CO2Me

CHO

1. LDA, THF 2. HCO2Et

121

t-BuMe2SiO

122

CO2Me

In general, forming the ester enolate under kinetic control conditions yields greater flexibility in the condensation, allowing the lithium enolate anion to be condensed with a variety of other esters (as in formation of 120). It is noted that a full equivalent of base is required to form the enolate anion. The Claisen condensation can be used to incorporate an aldehyde unit into a molecule. In Padwa and Danca’s123 synthesis of jamtine, ester 121 was treated with LDA to form the enolate anion, and subsequent reaction with ethyl formate gave aldehyde 122 via the acyl substitution reaction. The α-proton in a keto-ester Claisen product is more acidic than in the starting ester since it is adjacent to two carbonyls (see Table 13.2). Reaction of a keto-ester (e.g., 120) or an aldehyde-ester (e.g., 122) with base under thermodynamic conditions will yield the 1,3-dicarbonyl enolate anion. In other words, as the keto-ester product is formed, deprotonation can occur in the basic reaction medium. The resulting anion is relatively stable due to the increased resonance delocalization possible when two carbonyl groups are present (Section 13.2.1). The increase in stability means that the enolate anion product is less reactive than the enolate anion derived from the ester starting material, so there is little condensation product. This difference in reactivity means that the Claisen product will not compete as a reaction substrate, but this reaction does mean that an excess of base may be necessary. The thermodynamic conditions that employ an alkoxide base require that the alcohol solvent be the conjugate acid of that base (e.g., methanol with sodium methoxide), or transesterification will give two different esters for all products. There are two important variations of this condensation. In the first, an ester enolate is condensed with a ketone or aldehyde. This reaction has been called the Claisen reaction.124 In one example taken from Pons and coworker’s125 synthesis of goniothalamin, ethyl acetate reacted with cinnamaldehyde to give an 85% yield of ethyl (E)-3hydroxy-5-phenylpent-4-enoate. When the reaction is modified to use an aldehyde or ketone enolate rather than an ester enolate, but the reactive partner is an ester, it is still referred to as the Claisen reaction. In this variant, a ketone or aldehyde enolate anion reacts with an ester to yield a 1,3-diketone. In the Hua et al.126 synthesis of tetrahydro-1oxopyranobenzopyrans, the dianion of ethyl 3-oxobutanoate was condensed with ethyl pyridine-3-carboxylate (ethyl nicotinate) gave an 88% yield of ethyl (Z)-5-hydroxy-3-oxo-5-(pyridin-3-yl)pent-4-enoate, which exists largely in the enol form shown.

120

Su, Q.; Dakin, L. A.; Panek, J. S. J. Org. Chem. 2007, 72, 2.

121

(a) Rathke, M. W.; Sullivan, D. F. J. Am. Chem. Soc. 1973, 95, 3050. (b) Lochmann, L.; Lím, D. J. Organomet. Chem. 1973, 50, 9. (c) Sullivan, D. F.; Woodbury, R. P.; Rathke, M. W. J. Org. Chem. 1977, 42, 2038.

122

Rathke, M. W.; Lindert, A. J. Am. Chem. Soc. 1971, 93, 2318.

123

Padwa, A.; Danca, M. D. Org. Lett. 2002, 4, 715.

124

Claisen, L. Berichte 1890, 23, 976.

125

Fournier, L.; Kocienski, P.; Pons, J.-M. Tetrahedron 2004, 60, 1659.

126

Hua, D. H.; Chen, Y.; Sin, H.-S.; Maroto, M. J.; Robinson, P. D.; Newell, S. W.; Perchellet, E. M.; Ladesich, J. B.; Freeman, J. A.; Perchellet, J.-P.; Chiang, P. K. J. Org. Chem. 1997, 62, 6888.

689

13.4 ENOLATE CONDENSATION REACTIONS

O

Ph

1. LDA, THF

OEt Ethyl acetate

2. Ph

CO2Et OH Ethyl (E)-3-hydroxy-5phenylpent-4-enoate

CHO

CO2Et

2.5 LDA, Ether TMEDA, 0°C

O

(85%)

OH

O CO2Et

N

CO2Et N Ethyl 3-oxobutanoate

Ethyl (Z)-5-hydroxy-3-oxo5-(pyridin-3-yl)pent-4-enoate

Reaction of an ester enolate with an acid chloride will also generate a β-keto ester, and it is a useful alternative to the Claisen condensation. Ketone enolate anions can also be condensed with acid chlorides.127 An ester enolate anion can be trapped with trimethylsilyl chloride, as with aldehydes and ketones. An interesting variation of this condensation involved valerolactone (tetrahydro-2H-pyran-2-one), which was condensed with a lithio-acetate (R ¼ t-Bu, Et) in a modified Claisen condensation to yield 123.128 Rather than the normal loss of alkoxide to yield the β-diketone, 123 was hydrolyzed under mild conditions to yield alkoxy hydroxy-ester 124. Li

O O

LiCH2CO2R

O

O

O

–78°C → rt

OR

THF, –78°C

Tetrahydro-2H-pyran-2-one

O

O

H O OR

EtOH

123

124

The Claisen disconnection follows: O

O

R

O

O +

OH

R

R1

OH

R1

OH

A variation of the Claisen condensation involves the reaction of acid anhydrides with aldehydes, and it is called the Perkin reaction.129 Condensation of an aldehyde that has no enolizable protons with the enolate of an acid anhydride (e.g., acetic anhydride) leads to an acetoxy ester (e.g., 125).129a Internal acyl substitution by the alkoxide forms the O-acetyl ester and liberates the carboxylate anion (126). Subsequent reaction with another equivalent of acetic anhydride generates a new mixed anhydride, acetic 3-acetoxy-3-phenylpropanoic anhydride. Saponification leads to the β-hydroxy acid, which eliminates water in the acid hydrolysis step to yield the final product, the aryl acrylic acid derivative, 3-phenylacrylic acid (cinnamic acid). O O O O O

PhCHO

O

Me

O

O 2C

Ac2O

O

O

AcOK

O

Me O

O

Ph

125

126

Ph Acetic 3-acetoxy-3-phenylpropanoic anhydride

127

Beck, A. K.; Hoekstra, M. S.; Seebach, D. Tetrahedron Lett. 1977, 1187.

128

Duggan, A. J.; Adams, M. A.; Brynes, P. J.; Meinwald, J. Tetrahedron Lett. 1978, 4323.

129

O

O Ph

Acetic anhydride

Me

CO2H 1. aq KOH 2. H3O+

Ph 3-Phenylacrylic acid

(a) Perkin, W. H. J. Chem. Soc. 1869, 21, 53, 181. (b) Johnson, J. R. Org. React. 1942, 1, 210. (c) Reference 100, pp 344–346. (d) Reference 101, pp 184–186. (e) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-71. (f ) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 492, 493.

690

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

The Perkin disconnection follows: O CO2H

Ar

Ar

CHO

O

+ O

13.4.2.2 The Dieckmann Condensation Just as there is an intramolecular version of the aldol condensation (Section 13.4.1.2), there is an intramolecular version of the Claisen condensation, but it has been given a different name (the Dieckmann condensation).130,131 This reaction involves intramolecular cyclization of an α-,ω-diester (e.g., diethyl adipate). The reaction is usually done under equilibrating (thermodynamic) conditions, although kinetic control conditions can also be used. The initially formed enolate anion (127) attacked the ester group on the opposite side of the molecule to form a ring (tetrahedral intermediate 128). Loss of ethoxide in the usual manner gave the expected β-keto-ester,ethyl 2-oxocyclopentane-1-carboxylate. Hydrolysis and thermally induced decarboxylation (Section 3.8) gave cyclopentanone.134 A synthetic example is taken from the Park et al. 132 synthesis of (+)-isonitramine, in which diester 129 was treated with lithium hexamethyldisilazide in THF to yield the enolate anion, and subsequent cyclization gave 130 in 98% overall yield. The size of the ring being formed has a great influence on the course of the reaction (Sections 1.5.1 and 4.5.1). In general, formation of a three-membered ring is energetically less favorable, especially when compared with formation of a five- or sixmembered ring. The nature of the substituents and the reaction conditions influence the course of the reaction.133 O CO2Et CO2Et CO2Et

EtOH

Diethyl adipate

O

CO2Et 1.aq NaOH

OEt

OEt 127

Ph Ph

OEt

NaOEt

O+

2. H3 3. Heat (– CO2)

O

Cyclopentanone

Ethyl 2-oxocyclopentane-1-carboxylate

128

O

O CO2Me

N

O

O

CO2Et

LHMDS , THF

Ph

N

0°C, 0.5 h

CO2Et

Ph

O

129

130 (98%)

The Dieckmann disconnection follows: O HO2C

CO2H

13.4.2.3 The Knoevenagel Condensation Another classical reaction involves malonate-type enolate anions in a condensation reaction with aldehydes, usually a nonenolizable aldehyde. This reaction is known as the Knoevenagel condensation.129b,134 Malonate derivatives and 130

(a) Dieckmann, W. Berichte 1894, 27, 102, 965. (b) Idem., Ibid. 1900, 33, 2670. (c) Dieckmann, W.; Groenveld, A., Ibid. 1900, 33, 595. (d) Hauser, C. R.; Hudson, B. E. Org. React. 1942, 1, 266 (see p 274). (e) Schaefer, J. P.; Bloomfield, J. J. Ibid. 1967, 15, 1. (f ) Reference 100, pp 124, 125. (g) Reference 101, pp 75–77. (h) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-23. (i) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 204, 205.

131

For reactions of this type, see Larock, R. C. Comprehensive Organic Transformations, 2nd ed.; Wiley-VCH: New York, 1999; pp 1316, 1317.

132

Park, Y.; Lee, Y. J.; Hong, S.; Lee, M.; Park, H.-G. Org. Lett. 2012, 14, 852.

133

(a) Goldberg, M. W.; Hunziker, F.; Billeter, J. R.; Rosenberg, H. R. Helv. Chim. Acta 1947, 30, 200. (b) Banerjee, D. K. J. Indian Chem. Soc. 1940, 17, 453. (c) Chakravarty, N. K.; Banerjee, D. K. Ibid. 1946, 23, 377. (d) Dutta, J.; Biswas, R. N. Ibid. 1961, 38, 335.

134

(a) Japp, F. R.; Streatfeild, F. W. J. Chem. Soc. 1883, 43, 27. (b) Knoevenagel, F. Berichte 1896, 29, 172. (c) Idem Ibid. 1898, 31, 730. (d) Jones, G. Org. React. 1967, 15, 204. (e) Reference 104, pp 261, 262. (f ) Reference 101, pp 148, 149. (g) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-51. (h) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 364, 365.

691

13.4 ENOLATE CONDENSATION REACTIONS

compounds that have a methylene group dCH2d flanked by two carbonyl groups, but related derivatives are known that contain two cyano groups, two nitro groups, or two sulfonyl groups. Such compounds are known as active methylene compounds. Methylene compounds that have a mixture of two different electron-withdrawing groups are active methylene compounds. The salient feature of these compounds is the enhanced acidity of the methylene protons due to the proximity of the two electron-withdrawing groups, and pKa for such compounds is typically 8–13. Therefore, relatively weak bases can be used for the deprotonation. Active methylene compounds are the most common partners for a Knoevenagel condensation. An aldehyde or ketone can be condensed with an active methylene compound (H2CX2 or HRCX2), such as a malonic ester using a primary or secondary amine as the base. A competing aldol condensation can be avoided if the aldehyde has no α-protons. Diethyl malonate and benzaldehyde were condensed to form diethyl 2-(hydroxy(phenyl)methyl)malonate, for example. Spontaneous elimination of water gave the alkylidene product (diethyl 2-benzylidenemalonate), and saponification followed by decarboxylation gave the final Knoevenagel product, a conjugated acid (3-phenylacrylic acid). Contrary to the products observed in an aldol condensation that have reactants without conjugation in the products, elimination commonly occurs upon hydrolysis of the Knoevenagel products.

CO2Et CO2Et

1. PhCHO NHEt2

Ph

2. H3O+

HO

Diethyl malonate

CO2Et

– H 2O

CO2Et

Diethyl 2-(hydroxy(phenyl)methyl)malonate

Ph

CO2Et

1. aq NaOH 2. H3O+ 3. Heat (– CO2)

CO2Et Diethyl 2-benzylidenemalonate

Ph CO2H 3-Phenylacrylic acid

The Knoenevagel condensation provides a facile route to substituted alkylidene malonic acids [RCH]C (CO2H)2] and acrylic acid derivatives. When the base is Py with a trace of piperidine (rather than diethylamine) and malonic acid is the enolate precursor, the reaction is called the Doebner condensation or the Doebner modification.135 An example of this classical reaction is found in a synthesis of (+)-hapalindole Q by Kerr and Kinsman,136 malonic acid reacted with the aldehyde moiety in 1-tosyl-1H-indole-3-carbaldehyde, in pyridine heated at reflux, to give an 86% yield of (E)-3-(1-tosyl-1H-indol-3-yl)acrylic acid. Note that decarboxylation occurred under the reaction conditions. Kwon and coworkers137 showed that microwave irradiation facilitates the Knoevenagel condensation with aldehydes. CO2H

CHO CH2(CO2H)2, Py

N

20% Pyrrolidine Reflux, 12 h

Ts 1-Tosyl-1 H-indole3-carbaldehyde

N Ts (E)-3-(1-Tosyl-1H-indol(86%) 3-yl)acrylic acid

An interesting example is taken from the synthesis of ()-leuconinines A and B by Andrade and Sirasani,138 in which aldehyde 131 was heated with the half acid chloride-half methyl ester of malonic acid, in the presence of triethylamine. Formation of the enolate anion was followed by acyl addition to the aldehyde moiety to yield 132. The in situ reaction of the amine with the acid chloride moiety led to a lactam, and dehydration occurred under the reaction conditions to give an 82% yield of 133.

135 (a) Doebner, O. Berichte 1900, 33, 2140. (b) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-25. (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 212, 213. 136

Kinsman, A. C.; Kerr, M. A. J. Am. Chem. Soc. 2003, 125, 14120.

137

(a) Kim, S.-Y.; Kwon, P.-S.; Kwon, T.-W. Synth. Commun. 1997, 27, 533. (b) Kwon, P.-S.; Kim, Y.-M.; Kang, C.-J.; Kwon, T.-W. Synth. Commun. 1997, 27, 4091.

138

Sirasani, G.; Andrade, R. B. Org. Lett. 2011, 13, 4736.

692

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

N O

I H N H

N

I

N Cl

I

O

H

OMe

H

N H

NEt3, CH2Cl2, Heat

Cl

CHO

131

N

OH CO2Me

O

O

132

CO2Me 133 (82%)

Other derivatives containing two electron-withdrawing groups can be used in the reaction, including malononitrile. In work aimed at designing HIV-1 protease inhibitors, chiral building blocks were synthesized by Ramachary and Reddy.139 In one example, malononitrile was condensed with glyceraldehyde acetonide(134) using L-proline as a catalyst, generating Knoevenagel product 135 in situ. In this reaction, the Hantzsch ester was also added to reduce the initially formed 135 to 136, isolated in 99% yield.139 Cyanoacetic acid derivatives (e.g., ethyl 2-cyanoacetate) are also common partners in the Knoevenagel condensation.140 EtO2C

O

CHO

C L N 10% L-Proline

O

N H

+ MeCN, rt

CLN

O

CO2Et

CN

O

CN

O

134

O

135

CN CN 136 (99%)

As noted in Table 13.2, dinitromethane has a pKa of 3.6. Amines or even sodium carbonate are excellent bases for the deprotonation of dinitromethane to give the resonance-stabilized nitro-enolate 137. When this enolate anion reacted with piperonal in a Knoevenagel-type condensation, the dinitroalkylidene product 138 was formed in good yield. The Henry reaction141 (also known as the Kamlet reaction) is closely related to this condensation, and is often called a nitro aldol reaction. The Henry reaction involves condensation of nitromethane (or another nitroalkane) with aldehydes or ketones. An example using nitro enolate CH2NO2 is the condensation of nitromethane with benzaldehyde to give a 75% yield of 1-nitro-2-phenylethene.142 O NO2

Na2CO3

N

NO2

N

Dinitromethane

O

O O

N

O

N 137

O

O2N

O

O O

1. O

2. H3O+

NO2

CHO

O O

138

Mono-nitro compounds (e.g., nitrobutane) can also react with a suitable base to generate an enolate anion, and then react with aldehydes or with ketones.143 An example is the reaction of indole aldehyde 139 with nitromethane, using sodium acetate as the base, which gave alkenyl nitro compound 140 in good yield, in a synthesis of thiaplakortone A and derivatives by Quinn and coworkers.144 Several variations of this reaction include using microwave irradiation to assist the reaction145 or MgdAl hydrotalcites to induce condensation,146 and using powdered KOH without

139

Ramachary, D. B.; Reddy, Y. V. J. Org. Chem. 2010, 75, 74.

140

McElvain, S. M.; Lyle, R. E., Jr. J. Am. Chem. Soc. 1950, 72, 384.

141

(a) Henry, L. Compt. Rend. 1895, 120, 1265. (b) Kamlet, J. U. S. Patent 2,151,171, 1939 (Chem. Abstr. 1939, 33, 50039). (c) Hass, H. B.; Riley, E. F. Chem. Rev. 1943, 32, 373 (see p 406). (d) Lichtenthaler, F. W. Angew. Chem. Int. Ed. Engl. 1964, 3, 211. (e) For a review, see Luzzio, F. A. Tetrahedron 2001, 57, 915. (f ) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-43. (g) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 300, 301.

142

Knoevenagel, E.; Walter, L. Berichte 1904, 37, 4502.

143

Hass, H. B.; Susie, A. G.; Heider, R. L. J. Org. Chem. 1950, 15, 8.

144

Pouwer, R. H.; Deydier, S. M.; Le, P. V.; Schwartz, B. D.; Franken, N. C.; Davis, R. A.; Coster, M. J.; Charman, S. A.; Edstein, M. D.; Skinner-Adams, T. S.; Andrews, K. T.; Jenkins, I. D.; Quinn, R. J. ACS Med. Chem. Lett. 2014, 5, 178. 145

Varma, R. S.; Dahiya, R.; Kumar, S. Tetrahedron Lett. 1997, 38, 5131.

146

Bulbule, V. J.; Deshpande, V. H.; Velu, S.; Sudalai, A.; Sivasankar, S.; Sathe, V. T. Tetrahedron 1999, 55, 9325.

693

13.4 ENOLATE CONDENSATION REACTIONS

solvent147 or proazaphosphatranes [P(RNCH2CH2)3N] to promote the reaction.148 Asymmetric induction is possible in the Henry reaction.149 NO2 OBn

CHO

OBn CH3NO2, NaOAc Reflux, 1 h

N H

N H

139

140

Note that nitro enolates have other synthetic uses. When nitrobutane was treated with sodium hydroxide, nitroenolate 141 was formed. Rather than addition of an aldehyde or a ketone, 141 was treated with conc H2SO4 to form butanal with loss of N2O, in what is known as the Nef reaction.150 Modern versions of this reaction use bases (e.g., LDA) and less vigorous oxidizing agents (e.g., MoOPh [oxodiperoxymolybdenum(pyridine), often with hexamethylphosphoric triamide]).151

NO2

NaOH

N

O

H

– N2O

O Nitrobutane

O H2SO 4

Butanal

141

The Knoevenagel disconnection follows: R

CO2H

R CHO

+

HO2C—CH2—CO 2H

An interesting variation of the Knoevenagel condensation generates aldol condensation products (see Section 13.4.1). Condensation of β-keto esters (from a Claisen condensation) and aldehydes at pH 7.8 in aqueous media gives the aldolate. The condensation of ethyl 3-oxobutanoate and pentanal gave 4-hydroxyoctan-2-one in 80% yield.152 This new methodology constitutes a regioselective alternative to the aldol condensation. Note that heating the β-hydroxy-ketone at pH 1 liberates a conjugated ketone, prepared from the keto-ester and aldehyde in a one-pot procedure. 13.4.2.4 Nitrile Enolate Anions Nitrile enolates are formed by reaction of an alkyl nitrile with LDA or another suitable base. The reaction of 134 with malonitrile in Section 13.4.2.3 is one example, but mononitriles (RCH2CN) also generate carbanions, sometimes called nitrile enolates, that are useful in synthesis. Nitrile enolates exist and react as aggregates,153 with structures consistent with those found in the solid state.154 Both alkylation155 and condensation reactions with aldehydes156 or ketones are known.157 In addition to alkyl halides and carbonyl derivatives, condensation can occur with another nitrile. The base-catalyzed condensation of two nitriles to yield a cyano-ketone via an intermediate cyano-enolate anion,

147

Ballini, R.; Bosica, G.; Parrini, M. Chem. Lett. 1999, 1105.

148

Kisanga, P. B.; Verkade, J. G. J. Org. Chem. 1999, 64, 4298.

149

See Trost, B. M.; Yeh, V. S. C.; Ito, H.; Bremeyer, N. Org. Lett. 2002, 4, 2621.

150

(a) Nef, J. U. Annalen 1894, 280, 263. (b) Noland, W. E. Chem. Rev. 1955, 55, 137. (c) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-64. (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 452, 453. 151

Gabobardes, M.; Pinnick, H. Tetrahedron Lett. 1981, 22, 5235. See Vedejs, E.; Larsen, S. Org. Synth. 1986, 64, 127.

152

Kourouli, T.; Kefalas, P.; Ragoussis N.; Ragoussis V. J. Org. Chem. 2002, 67, 4615.

153

Fleming, F. F.; Shook, B. C. J. Org. Chem. 2002, 67, 2885, and Ref. 2b and 2c cited therein.

154

Carlier, P. R.; Lo, K. M. J. Org. Chem. 1994, 59, 4053.

155

For a synthetic example taken from a synthesis of (+)-neosymbioimine, see Varseev, G. N.; Maier, M. E. Org. Lett. 2007, 9, 1461.

156

For a synthetic example taken from a synthesis of AM6898D, see Fukuda, Y.; Sakurai, M.; Okamoto, Y. Tetrahedron Lett. 2000, 41, 4173.

157

For reactions of this type, see Larock, R. C., Comprehensive Organic Transformations, 2nd ed.; Wiley-VCH: New York, 1999; p 1317.

694

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

is known as the Thorpe reaction.130e,158 Reaction of butanenitrile with sodium ethoxide gave a nitrile enolate, which reacted with a second molecule of butanenitrile at the electrophilic cyano carbon to yield 142. Hydrolysis gave an intermediate imine-nitrile (2-ethyl-3-iminohexanenitrile), which is in equilibrium with the enamine form, (Z)-3amino-2-ethylhex-2-enenitrile (also see Section 13.6.1). Hydrolysis led to the final product of the Thorpe reaction, an α cyano-ketone 2-ethyl-3-oxohexanenitrile.158 Mixed-condensation reactions are possible when LDA and kinetic conditions are used to generate the α-lithionitrile (a mixed Thorpe reaction). When pentanenitrile was treated with LDA and condensed with benzonitrile, 2-cyano-1-phenylpentan-1-one was the isolated product after acid hydrolysis. Nitrile enolates can also be alkylated with a variety of alkyl halides.159 CN

EtCH2C N

CN

NaOEt

NH2

N-H Et 2-Ethyl-3-iminohexanenitrile

142

CN H3O+

N Et

Butanenitrile

CN

CN

H3O+

O

Et (Z)-3-Amino-2-ethylhex-2-enenitrile

Et 2-Ethyl-3-oxohexanenitrile

The Thorpe disconnection follows: R

CN

R1

O

R

R

+ CN R1

X R1

CN

X

There is an intramolecular version of the Thorpe reaction called the Thorpe-Ziegler reaction.160 When an α-,ω-dinitrile is treated with base, formation of the enolate is followed by cyclization. The reaction of adiponitrile with sodium ethoxide, for example, led to cyclization and formation of the usual imine-enamine mixture (143). Hydrolysis gave the cyano-ketone (2-cyanocyclopentanone, also known as 2-oxocyclopentane-1-carbonitrile).162b In Deslongchamp’s and coworkers161 synthesis of (+)-martimol, dinitrile 144 was treated with potassium tert-butoxide in tert-butanol at 85°C, and after heating to 115°C with acetic acid and phosphoric acid, ketone 145 was obtained in 68% yield. CN

CN

CN

CN H3O+

NaOEt

N-H

CN Adiponitrile

O

NH2

2-Oxocyclopentane1-carbonitrile

143

O

O

CN 1. t-BuOK , t-BuOH, 85°C

Me Me

Me

Me

CN

2. AcOH–H3PO4, 115°C

O Me

Me

144

145 (68%)

The Thorpe-Ziegler disconnection follows: O NLC

CLN

158

X

X

(a) Reference 100, p 449. (b) Baron H.; Remfry, F. G. P.; Thorpe, J. F. J. Chem. Soc. 1904, 85, 1726. (c) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-93. (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; WileyInterscience: Hoboken, NJ, 2005; pp 646, 647.

159

Larchev^eque, M.; Cuvigny, T. Tetrahedron Lett. 1975, 3851.

160

(a) Reference 100, pp 514–516. (b) Ziegler, K.; Eberle, H.; Ohlinger, H. Annalen 1933, 504, 94.

161

Toró, A.; Nowak, P.; Deslongchamps, P. J. Am. Chem. Soc. 2000, 122, 4526.

695

13.4 ENOLATE CONDENSATION REACTIONS

13.4.2.5 The Stobbe Condensation The condensation of succinic ester derivatives (e.g., diethyl succinate) with nonenolizable ketones or aldehydes (e.g., benzophenone) and a base yields alkoxide 146. However, the aldehyde or ketone substrate is not limited to nonenolizable derivatives. The alkoxide moiety reacts with the distal ester via acyl substitution to yield a lactone intermediate (147). In the original version of this reaction, saponification of 147 gave the α-alkylidene monoester 148. The reaction is limited to those α,ω-diesters for which the Dieckmann condensation is not a competitive reaction, and succinic acid derivatives are commonly used. This transformation is known as the Stobbe condensation.162 An example is the double-Stobbe condensation of 3,4-dimethoxybenzaldehyde with diethyl succinate (sodium ethoxide was used as a base) to yield 149, taken from Robinson and coworker’s163 synthesis of the dicaffeoyltartaric acid analogues.

O Ph

t-BuOK

+ CO2Et

Ph

Benzophenone

O

Ph

CO2Et

Ph

t-BuOH

Ph O

O

146

Ph

1. aq NaOH

EtO

Diethyl succinate

CO2Et

Ph Ph

CO2Et

HO2C

2. H3O+

CO2Et

O

147

148

MeO

MeO

EtO2C

CHO

1. Excess NaOEt,

CO2Et

MeO

CO2Et

2 2. H3O+

MeO

CO2H MeO MeO 149

3,4-Dimethoxybenzaldehyde

The Stobbe disconnection follows: R1 R2

CO2H

O

CO2R R1

CO2H

+

CO2H

2

R

13.4.2.6 The Darzens’ Glycidic Ester Condensation When an α-halo ester is treated with base, condensation of the resulting enolate anion with a carbonyl derivative leads to a halo-alkoxide product. This nucleophilic species can displace the halogen intramolecularly to produce an epoxide, which forms the basis of a classical reaction known as the Darzens’ glycidic ester condensation.164 The reaction of ethyl 2-chloroacetate and sodium ethoxide in the presence of benzaldehyde generated the usual alkoxide (150). Intramolecular displacement of chloride, however, gave the glycidic ester (ethyl 3-phenyloxirane-2-carboxylate). Saponification gave the epoxy acid (3-phenyloxirane-2-carboxylic acid), which was unstable and lost carbon dioxide to generate a substituted aldehyde or ketone (e.g., 2-phenylacetaldehyde). Cl

Cl

NaOEt

CO2Et

PhCHO

Ethyl 2-chloroacetate

Ph

Ph O 150

CO2Et

CO2Et O

Ethyl 3-phenyloxirane2-carboxylate

H3O+

Ph

CO2H

Heat

Ph

O

CHO

3-Phenyloxirane-2carboxylic acid

2-Phenylacetaldehyde

162

(a) Stobbe, H. Berichte 1893, 26, 2312. (b) Stobbe, H. Annalen 1894, 282, 280. (c) Johnson, W. S.; Daub, G. H. Org. React. 1951, 6, 1. (d) Reference 100, pp 439, 440. (e) Reference 101, pp 228–230. (f ) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-90. (g) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 626, 627. 163 164

Reinke, R. A.; King, P. J.; Victoria, J. G.; McDougall, B. R.; Ma, G.; Mao, Y.; Reinecke, M. G.; Robinson, W. E., Jr. J. Med. Chem. 2002, 45, 3669.

(a) Darzens, G. Compt. Rend. 1904, 139, 1214. (b) Idem Ibid. 1906, 142, 214. (c) Ballester, M. Chem. Rev. 1955, 55, 283. (d) Reference 100, pp 116–118. (e) Reference 101, pp 68–70. (f ) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-22. (g) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 198, 199.

696

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

This sequence is a useful chain extension reaction in which an aldehyde (PhCHO) is converted to a longer chain aldehyde homologue (PhCH2CHO). A ketone is converted to a longer chain aldehyde. Cyclohexanone, for example, reacted with ethyl α-chloroacetate in the presence of tert-butoxide to give cyclohexanecarbaldehyde via hydrolysis of the initially formed glycidic ester.165 Using an α-alkyl α-chloroester in this sequence leads to a ketone as the final product. Asymmetric versions of this reaction are known.166 In one example, taken from a synthesis of (+)-mefloquine by Coltart and coworkers,167 α-chlorohydrazone (151) was treated with LDA to generate the α-anion (see Section 13.4.6.2), and subsequent reaction with aldehyde 152 led to a chloro-alkoxide intermediate that gave 153 in 74% yield, with 98% of the trans-epoxide 153 and 2% of the isomeric cis-epoxide. An enantioselective synthesis of glycidic amides using camphor-derived sulfonium salts has been reported.168 A catalytic asymmetric Darzens condensation169 has been reported using a chiral phase-transfer agent. Note that a Wittig reaction that produces a vinyl ether (Ph3P]CHOR + ketone ! R2C]CHOR) that can be hydrolyzed to a chain-extended aldehyde (Section 12.5.1.1). It is an attractive alternative to the Darzens’ glycidic ester condensation.

O O

N

1. LDA, THF, –78°C

N

N

O

N O

2.

OHC

CF3

O CF3

Ph

N

Ph

N

CF3

Cl 151

152

153 (74%)

CF3

The Darzens’ disconnection follows: Ar (R1 )

R

X (R1 ) Ar

CHO

+

O

R CO2H

13.4.2.7 Acid Dianions The ability to produce kinetic enolate anions from acid derivatives allows another useful modification of the enolate reaction. Carboxylic acids have an acidic proton that is removed by 1 equiv. of base to first give a carboxylate anion (see 154). Addition of a second molar equivalent of a powerful base (e.g., a dialkylamide) leads to the dianion (155). Removal of the less acidic proton on carbon leads to the more nucleophilic site: The α-carbanionic site is more nucleophilic than the oxygen of the carboxylate anion. Subsequent reaction with an electrophilic species, in this case 1-bromobutane, occurred first at the more nucleophilic α-carbon to yield hexanoic acid.170 The carboxylate anion is usually generated with butyllithium and the enolate with LDA, although 2 equiv. of LDA can be used. As discussed in Chapter 11, treatment of a carboxylic acid with an organolithium reagent usually yields the ketone, via reaction of the carboxylate (154) with a second molar equivalent of an organolithium reagent (Section 11.6.6).171 For this reason, dialkylamide bases are commonly employed for the second deprotonation.172 An acid dianion yields the typical reactions of enolate anions derived from acid derivatives.173 When 3-methylbutanoic acid was treated first with NaH, and

165

Hunt, R. H.; Chinn, L. J.; Johnson, W. S. Org. Synth. Coll. Vol. 1963, 4, 459.

166

See Ohkata, K.; Kimura, J.; Shinohara, Y.; Takagi, R.; Hiraga, Y. Chem. Commun. 1996, 2411.

167

Knight, J. D.; Sauer, S. J.; Coltart, D. M. Org. Lett. 2011, 13, 3118.

168

Aggarwal, V. K.; Charmant, J. P. H.; Fuentes, D.; Harvey, J. N.; Hynd, G.; Ohara, D.; Picoul, W.; Robiette, R.; Smith, C.; Vasse, J.-L.; Winn, C. L. J. Am. Chem. Soc. 2006, 128, 2105.

169

Arai, S.; Shioiri, T. Tetrahedron Lett. 1998, 39, 2145.

170

(a) Reference 2b, p 158. (b) Pfeffer, P. E.; Silbert, L. S.; Chirinko, J. M., Jr. J. Org. Chem. 1972, 37, 451.

171

Heathcock, C. H.; Gulik, L. G.; Dehlinger, T. J. Heterocyclic Chem. 1969, 6, 141.

172

For a convenient preparation, see Parra, M.; Sotora, E.; Gil, S. Eur. J. Org. Chem. 2003, 1386.

173

Petragnani, N.; Yonashiro, M. Synthesis 1982, 521.

697

13.4 ENOLATE CONDENSATION REACTIONS

then with LDA, the dianion was formed.174 Subsequent reaction with 4-bromobut-1-ene occurred exclusively at the α-position, and hydrolysis gave an 95% yield of 2-isopropylhex-5-enoic acid, as reported in Koskinen and Nevalainen’s174 synthesis of nor-1,6-germacradien-5-ols. h  i 1: C H Br LDA, THF LDA 4 9 H3 CdCO2 H


! ½H3 CdCO2  






! H2 C dCO2 





! C4 H9 dH2 CdCO2 H THF HMPA, 50°C 2: aq HCl Acetic acid Hexanoic acid 154 155

175

176

177

Dianions of this type react with ketones, epoxides, or esters, as well as a wide variety of other electrophiles. The dilithio derivative of 2-methylpropionic acid was condensed with the epoxide moiety in 156 to form an hydroxy acid, which cyclized to form the lactone ring in 157.176 Since most of the enolates of acid derivatives contain a leaving group, the alkoxide resulting from reaction with an epoxide often displaces that leaving group to yield the lactone. Me

Me Me

O

O 1. Li

H H

Me

CO2 Li

Me

THF, –10°C

Me

2. H3O+

H

H H

H

HO

O

HO 156

157

The condensation of enolate anions with alkyl halides or other carbonyl derivatives allows a wide variety of synthetic and functional group transformations in the carbon-carbon bond-forming process. Enolate anions are, therefore, among the most useful synthetic intermediates known. In addition to generating a new carbon-carbon bond, the reaction proceeds with high diastereoselectivity in most cases, making it even more useful.

13.4.3 The Mukaiyama Aldol Reaction The acid-catalyzed aldol condensation was mentioned briefly in Section 13.4.1.1, including a reaction catalyzed by TiCl4. Using traditional Brønsted-Lowry acids generates the aldolate product, but the acidic conditions make the process reversible and poor yields can result due to deleterious cationic side reactions. For these reasons, this variation is not as widely used as the base-catalyzed aldol condensation. However, the base-catalyzed reaction can lead to dimers, polymers, self-condensation products or α-,β-unsaturated carbonyl derivatives, as described in Section 13.4.1. Mukaiyama and coworkers178 modified the acid-catalyzed reaction to use silyl enol ethers as the starting material. They found that a reaction with carbonyl compounds produced aldol-like products, in the presence of titanium tetrachloride (TiCl4) and other Lewis acids (e.g., BF3•OEt2, TMS-OTf ). O

Ph

OSiMe3 TiCl , CH Cl 4 2 2

Ph

Me

1. Base

Ph

Me

2. Me3SiCl

Ph

O Me

SiMe3 Cl

TiCl3 O

Ph

CHO

Ph

Ph

Ph

1,1-Diphenylpropan-2-one Cl O

– Me3SiCl

158

Cl Ti Cl O

159 O

H 2O

Me Me

Ph

Ph

Ph

OH Ph Ph

Ph 160

4-Hydroxy-3,3,6-triphenylhexan-2-one

174

Nevalainen, M.; Koskinen, A. M. P. J. Org. Chem. 2002, 67, 1554.

175

Moersch, G. W.; Burkett, A. R. J. Org. Chem. 1971, 36, 1149.

176

(a) Creger, P. L. J. Org. Chem. 1972, 37, 1907. (b) Creger, P. L. J. Am. Chem. Soc. 1967, 89, 2500.

177

(a) Kuo, Y. N.; Yahner, J. A.; Ainsworth, C. J. Am. Chem. Soc. 1971, 93, 6321. (b) Pfeffer, P. F.; Silbert, L. S. Tetrahedron Lett. 1970, 699.

178

Inomata, K.; Muraki, M.; Mukaiyama, T. Bull. Chem. Soc. Jpn. 1973, 46, 1807.

698

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

The requisite silyl-enol ether is prepared by treatment of a ketone [e.g., diphenylacetone (1,1-diphenylpropan2-one)] with base and trapping the thermodynamic enolate anion with chlorotrimethylsilane, to yield silyl enol ether 158. The concept of trapping enolate anions with chlorotrimethylsilane was briefly introduced in Section 13.3.2. When 158 was mixed with an aldehyde (e.g., 3-phenylpropanal), in the presence of TiCl4, a condensation occurred via a transition state (e.g., 159) to produce the titanium aldolate 160 via loss of chlorotrimethylsilane. Hydrolysis provided the aldol product, 4-hydroxy-3,3,6-triphenylhexan-2-one. The reaction proceeds without self-condensation and at 78°C. This reaction is now called the Mukaiyama aldol reaction.179 Other Lewis acids can initiate this coupling, including tin tetrachloride (SnCl4) and boron trifluoride etherate (BF3•OEt2). In a synthetic example by Brimble and coworkers,180 in a synthesis of virgatolide B, reacted aldehyde 161 with boron trifluoride etherate, and subsequent reaction with silyl enol ether 162 gave aldol 163 in >82%. The use of chiral substrates 161 and 162 led to excellent selectivity in this reaction, and isolation of 163 as a single diastereomer. OPMB OH OPMB

O

BOMO

1. BF3•OEt2 , CH2Cl2 , –78°C

CHO

OBOM

2. TMSO

BOMO BOMO

161

162

163 (>82%)

Mukaiyama et al.181 showed that a reaction of the trimethylsilyl enol ether of cyclohexanone [(cyclohex-1-en-1yloxy)trimethylsilane] with benzaldehyde and various Lewis acids generated the aldolate, 2-(hydroxy(phenyl) methyl)cyclohexan-1-one, as the major product, but the trimethylsilyl alkoxide and the alkylidene elimination product were observed with some Lewis acids. These results indicated that titanium tetrachloride (TiCl4) and tin tetrachloride (SnCl4) are the best catalysts, and they generally show good selectivity for reaction with aldehydes, even in the presence of a ketone moiety.175,182 Condensation does occur with ketones, however, if no aldehyde carbonyl is available and ketones react faster than esters.183 Denmark and Stavenger184 showed that trichlorosilyl enol ethers are very competent aldol reagents, even in the absence of additives. OSiMe3

1. PhCHO, MXn CH2Cl2 2. H2O

(Cyclohex-1-en-1-yloxy)trimethylsilane

O

OH Ph

2-(Hydroxy(phenyl)methyl)cyclohexan-1-one

In the initial reports of reactions that used the Mukaiyama protocol, diastereoselectivity was poor, often giving a mixture of syn-and anti-diastereomers in
1:1 ratio. The use of enantiopure aldehydes leads to good diastereoselectivity, however, as in the formation of 163. When complexing agents are added to the reaction medium, good diastereoselectivity can be obtained (see Section 13.5.1), and asymmetric induction is possible when chiral catalysts are used. The reaction of benzaldehyde with 164, for example, gave a 74% yield of 3-hydroxy-2-methyl-1,3-diphenylpropan-1-one(3.2:1 syn/anti),185 in the presence of cupric triflate and bis(oxazolidine) catalyst 165. Interestingly, this reaction was done in aqueous media. A water-accelerated reaction of ketene silyl acetals has been reported.186 179

(a) Mukaiyama, T. Angew. Chem. Int. Ed. Engl. 1977, 16, 817. (b) Mukaiyama, T. Org. React. 1982, 28, 203 (see pp 238–248). (c) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-63. (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 440, 441. For a discussion of the impact of this reaction in synthesis, see Kan, S. B. J.; Ng, K. K.-H.; Paterson, I. Angew. Chem. Int. Ed. 2013, 52, 9097.

180

Hume, P. A.; Furkert, D. P.; Brimble, M. A. Org. Lett. 2013, 15, 4588. boron trifluoride (BF3•OEt2) and zinc chloride (ZnCl2) also gave good results.

181

Mukaiyama, T.; Banno, K.; Narasaka, K. J. Am. Chem. Soc. 1974, 96, 7503.

182

see Mukaiyama, T.; Narasaka, K.; Banno, K. Chem. Lett. 1973, 1011.

183

Bohlmann, F.; Eickeler, E. Chem. Ber. 1979, 112, 2811.

184

Denmark, S. E.; Stavenger, R. A. J. Am. Chem. Soc. 2000, 122, 8837.

185

(a) Kobayashi, S.; Nagayama, S.; Busujima, T. Chem. Lett. 1999, 71. (b) Idem Tetrahedron 1999, 55, 8739.

186

Loh, T.-P.; Feng, L.-C.; Wei, L.-L. Tetrahedron 2000, 56, 7309.

699

13.4 ENOLATE CONDENSATION REACTIONS

OSiMe3

O

+ 20% Ph

OH

O N

N

i-Pr

O

Cu(OTf)2 , H2O–EtOH

Ph

PhCHO

Ph

i-Pr 3-Hydroxy-2-methyl-1,3diphenylpropan-1-one

165

164

(R)-BINOL-Titanium catalysts have been used in solvents (e.g., toluene or CH2Cl2), where BINOL is 1,10 -bi2-naphthol,187 supercritical fluoroform, and in supercritical carbon dioxide. Other reactions have been reported using catalysts closely related to bis(oxazolidine) catalyst 165, with dichloromethane as the solvent.188 Chiral oxazaborolidinone derivatives are effective catalysts for enantioselective Mukaiyama aldol reactions.189 In a synthesis of halipeptins A and D by Nicolaou and coworkers,190 (2R,5S)-5-methoxy-2-methyloctanal was condensed with Me2C]C(OMe) OTMS in the presence of 166 to yield methyl (3S,4R,7S)-3-hydroxy-7-methoxy-2,2,4-trimethyldecanoate in 87% yield and
95:5 diastereoselectivity. OSiMe3

O

OMe

O CHO

OMe

B H

+ N

OMe

, BH3•THF

OH CO2Me

CH2Cl2 , –78°C

Ts (2R,5S)-5-Methoxy-2-methyloctanal

166

Methyl (3S,4 R,7S)-3-hydroxy-7- (87%) methoxy-2,2,4-trimethyldecanoate

A variation of the Mukaiyama aldol reaction involves the reaction of silyl enol ethers with acetals,191 in the presence of TiCl4. An example is the reaction of silyl-enol ether 167 with 1,1-dimethoxycyclohexane (a protected cyclohexanone, Section 5.3.3.1) to give 168 in 91% yield.192–194 An intramolecular Mukaiyama reaction was used in a Hanessian et al.195 synthetic approach to pactamycin. The reaction of diketone 169 with chlorotrimethylsilane and H€ unig’s base generated a silyl enol ether, in situ, and in the presence of TiCl4 ring closure gave 170 in 85% yield. Ph

OSiMe3

1. TiCl4 , –78°C

OMe

2. H2O

+ Me

O

OMe

Me MeO

1,1-Dimethoxycyclohexane

167

168 (91%)

Me O

Me O

O PMP

(i-Pr)2NEt , TiCl4 , TMSCl

N

Ph

O

CH2Cl2 , 0°C

TESO

O

PMP N OH

TESO OTBDPS

169

187

Mikami, K.; Matsukawa, S.; Kayaki, Y.; Ikariya, T. Tetrahedron Lett. 2000, 41, 1931.

188

Evans, D. A.; Burgey, C. S.; Kozlowski, M. C.; Tregay, S. W. J. Am. Chem. Soc. 1999, 121, 686.

189

Ishihara, K.; Kondo, S.; Yamamoto, H. J. Org. Chem. 2000, 65, 9125.

OTBDPS 170 (85%)

190

Nicolaou, K. C.; Lizos, D. E.; Kim, D. W.; Schlawe, D.; de Noronha, R. G.; Longbottom, D. A.; Rodriquez, M.; Bucci, M.; Cirino, G. J. Am. Chem. Soc. 2006, 128, 4460.

191

Mukaiyama, T.; Murakami, M. Synthesis 1987, 1043.

192

Mukaiyama, T.; Hayashi, M. Chem. Lett. 1974, 15.

193

Ishihara, H.; Inomata, K.; Mukaiyama, T. Chem. Lett. 1975, 531.

194

Mukaiyama, T.; Ishihara, H.; Inomata, K. Chem. Lett. 1975, 527.

195

Hanessian, S.; Vakiti, R. R.; Dorich, S.; Banerjee, S.; Desch^enes-Simard, B. J. Org. Chem. 2012, 77, 9458.

700

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

Conjugated ketones react with silyl enol ethers under Mukaiyama conditions to give the 1,4-addition product in high yield. The reaction of (cyclohex-1-en-1-yloxy)trimethylsilane with (E)-chalcone, for example, gave a 95% yield of the 1,4-product, 2-(3-oxo-1,3-diphenylpropyl)cyclohexan-1-one.196 Aldol condensations with conjugated systems often lead to 1,4-addition of the enolate anion (Section 13.7.1). Since the aldol condensation is usually done under basic conditions, further reaction of the products including self-condensation of the starting materials is commonly observed. Mukaiyama conditions are relatively mild, and the 1,4-addition proceeds in good yield with few deleterious side reactions. A variation of this approach is the so-called vinylogous Mukaiyama aldol reaction, illustrated by the reaction of aldehyde 171 and vinylogous silyl enol ether 172, which reacted in the presence of boron trifluoride to give 173 in 62% yield. This transformation was incorporated into the Lhommet, and coworker’s197 synthesis of ()-quinolizidine 217A. As seen in the reaction of 172, the Mukaiyama reaction can be applied to ketene acetals, which are usually derived from the enolate anion of an ester.198 This variation is very useful since esters can be used as ketene acetal precursors, greatly expanding the utility of the reaction. O

O

+

OSiMe3

O Ph

Ph 2. H2O

Ph (Cyclohex-1-en-1-yloxy)trimethylsilane

Ph

1. TiCl4 , CH2Cl2 , –78°C

(E)-Chalcone

2-(3-Oxo-1,3-diphenylpropyl)cyclohexan-1-one

(94%)

OTBS CHO N

+

O BF3•OEt2 , CH2Cl2

O

O

171

N

–78°C

Cbz 172

OH Cbz

O

O

173 (62%)

The disconnections are identical to those shown for the aldol condensation.

13.4.4 Boronic Esters (Boron Enolates) Vinyl boronates (boronic esters, e.g., 175) are important intermediates in directed aldol condensations, particularly when derived from aldehydes. Trapping the enolate as the boronic ester, and separating the desired isomer can solve regiochemical problems in enolate formation. Subsequent condensation with an aldehyde proceeds with excellent diastereoselectivity (Section 13.5). Mukaiyama et al.200 showed that conjugate addition of tripropylborane (Section 9.2.1) with methyl vinyl ketone (MVK) gave 175, and reaction with benzaldehyde gave a 91% yield of 176:177 as a 33:1 mixture of diastereomers.178,199,200 Brown and coworkers first reported Mukaiyama’s conjugate addition method for the preparation of 175.201 Masamune et al.199 prepared 175 via reaction of diazoketone 174 with tributylborane (see Section 7.9.2.3 for the preparation and reactions of diazoketones). Other methods for the preparation of boronic esters include the reaction of borane derivatives with ketones, thioesters or ketone enolates. B(Bu)3

Me O

Me

Bu

Me

CHN2

OB(Bu)2

B(Bu)3

O

174

Bu 1. PhCHO

Me

Me

2. H2O2

175

Ph O

OH 176

Bu Ph

+ O

OH 177

196

(a) Narasaka, K.; Soai, K.; Mukaiyama, T. Chem. Lett. 1974, 1223. (b) Narasaka, K.; Soai, K.; Aikawa, Y.; Mukaiyama, T. Bull. Chem. Soc. Jpn. 1976, 49, 779.

197

Fellah, M.; Santarem, M.; Lhommet, G.; Mouriès-Mansuy, V. J. Org. Chem. 2010, 75, 7803.

198

For an example taken from a synthesis of ()-azaspirene, see Hayashi, Y.; Shoji, M.; Yamaguchi, J.; Sato, K.; Yamaguchi, S.; Mukaiyama, T.; Sakai, K.; Asami, Y.; Kakeya, H.; Osada, H. J. Am. Chem. Soc. 2002, 124, 12078.

199

Masamune, S.; Mori, S.; Van Horn, D.; Brooks, D. W. Tetrahedron Lett. 1979, 1665.

200

Mukaiyama, T.; Inomata, K.; Muraki, M. J. Am. Chem. Soc. 1973, 95, 967.

201

Suzuki, A.; Arase, A.; Matsumoto, H.; Itoh, M.; Brown, H. C.; Rogic, M. M.; Rathke, M. W. J. Am. Chem. Soc. 1967, 89, 5708.

701

13.4 ENOLATE CONDENSATION REACTIONS

There are several interesting features of the boronic ester aldol condensation. The condensation reaction that produced 176 and 177 proceeded with good diastereoselectivity, which is analogous to that observed in base-catalyzed enolate condensations (Section 13.5). It is important to note that the boron aldolate product was treated with an oxidizing agent to remove boron (just as boranes were oxidized in Section 9.4.1). The two most common oxidants are H2O2 and MoOPh,202 which did not lead to degradation of the diastereoselectivity of the initial condensation. Several factors help to control or at least influence the diastereoselectivity of this reaction. The first is the stereochemistry of the boron enolate. In most preparations, including 175, the (E)-isomer of boronic ester predominates. This isomer led to 176 with high diastereoselectivity. Masamune et al.199 found that treatment of 175 with lithium phenoxide led to isomerization of the (E)- to the (Z)-isomer, and subsequent condensation with benzaldehyde gave a 90% yield of 176 and 177 (>2:1 ratio). Masamune also found that the geometry of the boron enolate was greatly influenced by its method of preparation. Tributylborane reacted with MVK to give predominantly the (E)-isomer, as shown. The ketene, prop-1-en-1-one, reacted with sulfide Bu2BSt-Bu to give the (E)-enolate (178) as the major isomer. Condensation with isobutyraldehyde gave a 75% yield of 179 and 180, in a ratio of 5:>95 favoring 180.203

O

Me C

+

OB(C4H9)2

Bu2BSt-Bu

Me

O St-Bu

i-Pr2NEt

St-Bu

Me

S-(tert-Butyl)propanethioate

St-Bu Me

Me 179

OB(Bu)2

O

St-Bu +

178

Bu2BOTf

OH

O

2. aq H2O2

St-Bu

Prop-1-en-1-one

OH 1. i-PrCHO

OH

180 OH

O

1. i-PrCHO

St-Bu

2. aq H2O2

Me 179

181

O St-Bu

+ Me 180

Masamune also showed that boron enolates are generated by reaction of a thioester [S-(tert-butyl) propanethioate] with dibutylboryl triflate and diisopropylethylamine (H€ unigs base). The product of this reaction was boron enolate 181, and subsequent condensation with isobutyraldehyde gave 179 and 180, although the diastereomeric ratio was reversed to give a >95:5 ratio (79% yield) favoring 179.196 A similar reaction occurs with enolizable ketones (e.g., 1-cyclohexylpropan-1-one), where the syn/anti ratio of the aldolate products resulting from reaction of the boron enolate (182) varied as H€ unigs base or 2,6-lutidine was used as the base.204 The best selectivity for the syn-diastereomer (182) was observed with 9-BBN derivatives, and the best anti-diastereoselectivity (183>184) was with dicyclopentylboranes and H€ unigs base. A (Z)-boronic ester will generally react with an aldehyde to give the syn aldolate, where an (E)-boronic ester will give the anti-aldolate.202 The boron aldolates were converted to the aldolate by treatment with H2O2 or MoOPh.202 The diastereoselectivity of the aldol condensation varies with the substituents on the carbonyl precursor to the boron enolate. The best selectivity was obtained in an ether solvent with the enolate and product ratios a function of the base. The ligand on the boron was important and there were differences in (E/Z) selectivity for Bu versus cyclopentyl.202,205 R2BO

O

H

O

Me Amine

1-Cyclohexylpropan-1-one

O

OH

OH

1. PhCHO

R2BOTf

Ph

2. H2O2

182

+

Ph

Me 183

202

Evans, D. A.; Nelson, J. V.; Vogel, E.; Taber, T. R. J. Am. Chem. Soc. 1981, 103, 3099.

203

Hirama, M.; Masamune, S. Tetrahedron Lett. 1979, 2225.

204

Van Horn, D. E.; Masamune, S. Tetrahedron Lett. 1979, 2229.

205

Evans, D. A.; Bartroli, J.; Shih, T. L. J. Am. Chem. Soc. 1981, 103, 2127.

Me 184

702

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

Recent versions of this condensation react dialkylchloroboranes with ketones or aldehydes, in the presence of an amine base. A synthetic example of this reaction is taken from Paterson and coworker’s206 synthesis of aplyronine C. Ketone 186 was treated with the chiral B-chloro diisopinocampheylborane (see Sections 9.2 and 9.4.2) in the presence of triethylamine to yield the boronic ester in situ. When aldehyde 185 was added, a 61% % yield of 187 was obtained with good diastereoselectivity. O

O

OAc

Me N

TESO TESO

O

CHO

OTES O 186

H

B—Cl, NEt3, Ether, –78°C 2

MeO

OMe O

185 TESO TESO

O

OTES OH

O

OAc

Me N

MeO

CHO

OMe 187 (61%)

13.4.5 The Meyers Aldehyde Synthesis Groups other than a carbonyl can stabilize a carbanion. Meyers et al.207 described the preparation of α-lithio derivatives of dihydro-1,3-oxazines (188) and 2-oxazolines (190), and subsequent alkylation and condensation reactions via α-lithio derivatives (e.g., 189). Meyers et al.207 treated the commercially available 2,4,4,6-tetramethyl-5,6-dihydro-4H-1,3-oxazine with butyllithium to form the α-lithio derivative (191), stabilized by chelation with the nitrogen. When iodomethane was added, a carbanion displacement reaction occurred to give 2-ethyl-4,4,6-trimethyl-5,6-dihydro-4H-1,3-oxazine.207 Reduction of the imine unit in 2-ethyl-4,4,6-trimethyl-5,6-dihydro-4H-1,3-oxazine with sodium borohydride (Section 7.4) gave the amine, 2-ethyl-4,4,6-trimethyl-1,3-oxazinane, and hydrolysis produced the aldehyde (propanal). O

O

O

N

R

N R

N

R

Li

188

189

190 O

O N

O

BuLi

CH3

2, 4, 4, 6-Tetramethyl-5, 6dihydro-4H-1, 3-oxazine

N

O

MeI

N

NaBH4

O

CH3

N

Li 191

H3O+

Me

H Me

H 2-Ethyl-4, 4, 6-trimethyl-5, 6dihydro-4H-1, 3-oxazine

2-Ethyl-4, 4, 6-trimethyl1, 3-oxazinane

206

Paterson, I.; Fink, S. J.; Lee, L. Y. W.; Atkinson, S. J.; Blakey, S. B. Org. Lett. 2013, 15, 3118.

207

Meyers, A. I.; Nabeya, A.; Adickes, H. W.; Politzer, I. R. J. Am. Chem. Soc. 1969, 91, 763.

Propanal

703

13.4 ENOLATE CONDENSATION REACTIONS

These carbanions also react with aldehydes or ketones. When 191 was treated with benzophenone, the initial condensation product after hydrolysis was the alcohol (192). Reduction of the imine moiety provided 193,208 and hydrolysis led to conversion of the oxazolidine to 3-hydroxy-3,3-diphenylpropanal, but this aldehyde eliminated water under the hydrolysis conditions to yield the conjugated aldehyde (3,3-diphenylacrylaldehyde). This sequence is known as the Meyers’ aldehyde synthesis.207,209 The reaction of lithio derivatives (e.g., 191) with alkyl halides leads to aldehydes and reaction with aldehydes or ketones leads to conjugated carbonyl derivatives.

O

O

1. Ph 2. H3

Ph

O

N

N

Li 191

192

O

NaBH4

O+

OH Ph

Ph

OH

N Ph

H

Ph

193 O Ph

H3O+

H

Ph

O

–H2O

OH

Ph H

Ph

3,3-Diphenylacrylaldehyde

3-Hydroxy-3,3-diphenylpropanal

Meyers et al.208,209 and others210 showed that 2-oxazolines (e.g., 2,4,4-trimethyl-4,5-dihydrooxazole are) carbanionic synthons when they are converted to their α-lithio derivative. Reaction of 2,4,4-trimethyl-4,5-dihydrooxazole with butyllithium, for example, gave 194. Addition of benzyl bromide gave the alkylated product, 4,4-dimethyl-2phenethyl-4,5-dihydrooxazole, and heating that product in ethanolic sulfuric acid gave an 84% yield of the ethyl ester, ethyl 3-phenylpropanoate.211 Ketones react similarly to yield a conjugated acid derivative. O

O

O CH3 N

BuLi

H2SO 4

PhCH2Br

N

N

Ph

EtO2C

EtOH

Ph

Li 2, 4, 4-Trimethyl-4, 5dihydrooxazole

194

4, 4-Dimethyl-2-phenethyl4, 5-dihydrooxazole

Ethyl 3-phenylpropanoate

(84%)

These reactions are not limited to the simple oxazine and oxazolines described above. Other derivatives are prepared from aldehydes and amino alcohols. Meyers et al.212 used this methodology to prepare chiral precursors for use in asymmetric synthesis. When keto acid, 4-oxo-4-phenylbutanoic acid, was treated with valinol [(S)-2amino-3-methylbutan-1-ol], chiral bicyclic lactam 195 was formed.212 Sequential alkylation reactions with iodomethane, and then 4-methoxybenzyl bromide, gave 196 in 85% yield (30:1 selectivity). Upon hydrolysis, keto acid 197 was produced in 88% yield with >98 %ee.212 This methodology was used in the Meyers et al.213 asymmetric synthesis of (+)-mesembrine.

208

Meyers, A. I.; Nabeya, A.; Adickes, H. W.; Fitzpatrick, J. M.; Malone, G. R.; Politzer, I. R. J. Am. Chem. Soc. 1969, 91, 764.

209

(a) Meyers, A. I.; Nabeya, A.; Adickes, H. W.; Politzer, I. R.; Malone, G. R.; Kovelesky, A. C.; Nolen, R. L.; Portnoy, R. C. J. Org. Chem. 1973, 38, 36. (b) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-61. 210

For related work, see (a) Wehrmeister, H. L. J. Org. Chem. 1962, 27, 4418. (b) Allen, P.; Ginos, J. Ibid. 1963, 28, 2759.

211

Meyers, A. I.; Temple, D. L., Jr. J. Am. Chem. Soc. 1970, 92, 6644, 6646.

212

Meyers, A. I.; Harre, M.; Garland, R. J. Am. Chem. Soc. 1984, 106, 1146.

213

Meyers, A. I.; Hanreich, R.; Wanner, K. Th. J. Am. Chem. Soc. 1985, 107, 7776.

704

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

Ph

Ph

OH

O

O

H

+

1. BuLi 2.MeI

N NH2

3. BuLi

H

HO2C 4-Oxo-4-phenylbutanoic acid

Br

O

Valinol

4. MeO

195 Ph O

O

Ph

N

H

Me

HO2C O

Me

MeO MeO 196 (85%)

197 (>98%)

It is clear that the Meyers aldehyde synthesis provides a facile route not only to a variety of aldehydes, but also to esters, acids, and chiral derivatives. The disconnections available by this methodology include the following: R OHC

HO2C Me + R–X

R

R1

+ R

O

R1

O

R EtO2C

O HO2C Me

OHC

HO2C Me R1

+ R

R1

R

CO2H

R

X

X = CO2H or CHO

R1

O

R2 +

R1 –X

+

R2 –X

13.4.6 Imine and Hydrazone Carbanions Note that in Section 13.4.1 there were problems with aldol-type reactions, especially with the directed aldol condensation. In particular, aldehydes with an α-hydrogen atom have great difficulty adding to ketones due to their propensity for self-condensation. The ability to use kinetic control conditions in enolate reactions of ketones and aldehydes often solves this problem. There are also several alternative approaches that involve the use of carbanions derived from imines and hydrazones and these can be very useful.214 13.4.6.1 Imine Carbanions Wittig prepared the ethylamine imine of acetaldehyde [(Z)-N-ethylethanimine], and subsequent reaction with lithium diethylamide gave α-lithio derivative 198. When benzophenone was added, the expected carbanion condensation reaction occurred to give a 27% yield of (Z)-3-(ethylimino)-1,1-diphenylpropan-1-ol.215,216 With these experiments, Wittig et al.215,216 demonstrated several important features of imine carbanions. The imine moiety functioned as a protected carbonyl (see Section 5.3.3 for protecting groups used with ketones and aldehydes) in this reaction, and the carbonyl could be regenerated after the alkylation or condensation reaction. The imine clearly provides activation to the α-position to produce the carbanion on treatment with a suitably strong base. For Schiff bases (imines derived from aldehydes) deprotonation can occur only at the α-carbon, with no possibility of self-condensation or dimerization. The sequence shown above was one of the first examples of an aldehyde directed aldol condensation. N H

N

LiNEt2, Ether

CH3

H

(Z)-N-Ethylethanimine

CH2-Li 198

214

Wittig, G.; Reiff, H. Angew. Chem. Int. Ed. Engl. 1968, 7, 7.

215

Wittig, G.; Frommeld, H.-D. Chem. Ber. 1964, 97, 3541.

216

Wittig, G.; Frommeld, H.-D. Chem. Ber. 1964, 97, 3548.

N

1. Benzophenone 2. H3O+

H

Ph OH Ph

(Z)-3-(Ethylimino)-1,1diphenylpropan-1-ol

(27%)

705

13.4 ENOLATE CONDENSATION REACTIONS

In a synthesis of phomoidride A and phomoidride B by Nicolaou et al. 217 imine 199 was treated with LDA, and then the aldehyde to generate a substituted imine. Subsequent acid hydrolysis converted the imine to an aldehyde, with accompanying elimination of water to give a 50% overall yield of conjugated aldehyde 200. When imine anions of this type are condensed with an ester, the product is an enamino ketone.218 There is a potential problem when this approach is used. When imine carbanions react with ethyl chloroformate, two products are usually formed; C-acylation as the major product and N-acylation as a minor product.219 Condensation with a ketone and acid hydrolysis can be used to prepare conjugated aldehydes.220 TBSO

CHO

N

1. LDA, Ether, –78°C 2.

TBSO

CHO

O

O

C8 H15

3. aq Oxalic acid

O

199

O

200 (50% overall)

13.4.6.2 Hydrazone Carbanions Corey and Enders221 introduced an alternative to Wittig’s imine-carbanion methodology discussed in Section 13.4.6.1. Modification of the reaction to use a hydrazone rather than an imine made it more controllable and useful. Hydrazone carbanion 151 was used in Section 13.4.2.6 in an example of the Darzens’ glycidic ester synthesis. Condensation with aldehydes or ketones, or reaction with alkyl halides is also possible. In a synthesis of (+)-leucascandroide A by Paterson and Tudge,222 hydrazone (201, PMB ¼ p-methoxybenzyl; Section 5.3.1.1) was converted to the anion with LDA and subsequent reaction with the propargylic bromide gave 202. Subsequent treatment with tetrabutylammonium fluoride (TBAF) initiated conversion of the hydrazone to the ketone and the O-silyl group to the alcohol (Section 5.3.1.2), giving 1-hydroxy-7-((4-methoxybenzyl)oxy)hept-5-yn-2-one in 65% yield. These carbanions react normally with aldehydes or ketones. The reaction of the N,N-dimethylhydrazone derivative of cyclohexanone (2-cyclohexylidene-1,1-dimethylhydrazine) with LDA, and then with benzophenone, to give a 90% yield of 2-(hydroxydiphenylmethyl)cyclohexan-1-one.223 NNMe2

NNMe2

THF Bu4NF 20°C

OPMB

2.

OSiMe2t-Bu

O

OPMB

1. LDA, THF-HMPA

OSiMe2t-Bu

Br

201

202

OPMB

OH 1-Hydroxy-7-((4-methoxybenzyl)- (65%) oxy)hept-5-yn-2-one

O NMe2 N

1. LDA, THF 2. Benzophenone 3. H3O+

2-Cyclohexylidene-1,1dimethylhydrazine

Ph OH Ph 2-(Hydroxydiphenylmethyl)- (90%) cyclohexan-1-one

The hydrazone carbanion reacts like other carbanions, not only with alkyl halides, but with other electrophilic species (e.g., carbonyl derivatives or epoxides).224 Hydrazone carbanions generally show a great preference for 1,2addition to conjugated aldehydes (e.g., crotonaldehyde),223 which suggests that an α-lithiohydrazone is a highly

217

Nicolaou, K. C.; Jung, J.; Yoon, W. H.; Fong, K. C.; Choi, H.-S.; He, Y.; Zhong, Y.-L.; Baran, P. S. J. Am. Chem. Soc. 2002, 124, 2183.

218

Wittig, G.; Suchanek, P. Tetrahedron Suppl. 8, Part I 1966, 22, 347.

219

Reference 214, p 12 and Ref. 6 cited therein.

220

Corey, E. J.; Enders, D.; Bock, M. G. Tetrahedron Lett. 1976, 7.

221

(a) Corey, E. J.; Enders, D. Tetrahedron Lett. 1976, 3. (b) Idem Chem. Ber. 1978, 111, 1337.

222

Paterson, I.; Tudge, M. Angew. Chem. Int. Ed. 2003, 42, 343.

223

Corey, E. J.; Enders, D. Tetrahedron Lett. 1976, 11.

224

Corey, E. J.; Enders, D. Chem. Ber. 1978, 111, 1362.

706

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

carbanionic species, and that the reaction with aldehydes may be under chelation control. N-Tosyl hydrazone derivatives are also converted to a dianion and undergo the same type of reactions.225 Oximes are a useful source of carbanions in aldol-like condensations. Jung et al.226 showed that butan-2-one oxime reacted with 2 equiv. of n-butyllithium to yield the dianion 203, which reacted with 1-bromopropane.227 Subsequent treatment with water resulted in a 53% yield of 3-methylhexan-2-one oxime. O-Tetrahydropyranyl oximes (e.g., 204) can be used in this sequence, and the fact that the oxygen is blocked removes the necessity for generating a dianion, which may have solubility problems that diminish the yield of subsequent reactions. Conversion of 2-heptanone to the O-tetrahydropyranyl oxime produced a 61:39 mixture of (E/Z) isomers (204). Treatment with LDA followed by condensation with acetone led to a mixture of oximes (205/206) in 94% yield, maintaining the 61:39 (E/Z)ratio.228 Warming 204 to 50°C (from 78°C) led to isomerization, giving exclusively the (E)-isomer. When the condensation of (E)-204 with acetone was done at 50°C, a single product (205) was obtained.228 O– Li+

OH N

2 C4H9Li

OH N

N

1. 1-Bromopropane

Li 2. H3O+

Butan-2-one oxime

THPO

OTHP N

OTHP N

N

1. LDA, THF 2. Acetone 3. H3O+

C7H15

3-Methylhexan-2-one oxime (53%)

203

+ C7H15

204

OH 205

OH

C7H15 206

13.4.7 Nozaki-Hiyama-Kishi Coupling Another acyl addition reaction involves an organochromium intermediate. Although it is not formally an enolate anion condensation, it is related to the acyl addition reactions presented in this chapter, so it is discussed here rather than in Chapter 18 with other organometallic reactions. As originally formulated by Nozaki and coworkers,229 vinyl triflates (e.g., dodec-1-en-2-yl trifluoromethanesulfonate) were treated with an aldehyde and CrCl2 in the presence of a Ni catalyst. Initial formation of a vinyl nickel species was followed by in situ transmetalation to yield a vinyl Cr(III) species, which added to the aldehyde. Hydrolysis gave the alcohol product. When dodec-1-en-2-yl trifluoromethanesulfonate reacted with benzaldehyde, an 83% yield of 2-methylene-1-phenyldodecan-1-ol was obtained.229 Kishi and coworkers employed a similar reagent for coupling activated alkenes with aldehydes, during synthetic studies toward palytoxin,230 and later work involving the synthesis of taxanes231 expanded the reaction. Hiyama et al.232 studied reactions of bromobutenes with aldehydes in the presence of Cr(II) reagents. This transformation is now known as the urstener and Nozaki-Hiyama-Kishi reaction,233 although it also has been referred to as the Takai-Utimoto reaction. F€ Shi234 employed a Cr(II)-Mn(0) redox couple to give essentially the same coupling reaction. Boeckman and Hudack235 225

Lipton, M. F.; Shapiro, R. H. J. Org. Chem. 1978, 43, 1409.

226

Jung, M. E.; Blair, P. A.; Lowe, J. A. Tetrahedron Lett. 1976, 1439.

227

(a) Adlington, R. M.; Barrett, A. G. M. J. Chem. Soc. Chem. Commun. 1979, 1122. (b) Idem Ibid. 1978, 1071.

228

Ensley, H. E.; Lohr, R. Tetrahedron Lett. 1978, 1415.

229

Takai, K.; Tagashira, M.; Kuroda, T.; Oshima, K.; Utimoto, K.; Nozaki, H. J. Am. Chem. Soc. 1986, 108, 6048.

230

Jin, H.; Uenishi, J.; Christ, W. J.; Kishi, Y. J. Am. Chem. Soc. 1986, 108, 5644.

231

Kress, M. H.; Ruel, R.; Miller, L. W. H.; Kishi, Y. Tetrahedron Lett. 1993, 34, 5999.

232

(a) Hiyama, T.; Kimura, K.; Nozaki, H. Tetrahedron Lett. 1981, 22, 1037. (b) Hiyama, T.; Okude, Y.; Kimura, K.; Nozaki, H. Bull. Chem. Soc. Jpn. 1982, 55, 561.

233

(a) Cintas, P. Synthesis 1992, 248. (b) Wessohann, L. A.; Scheid, G. Synthesis 1999, 1 (see pp 16–18). (c) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-67. (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; WileyInterscience: Hoboken, NJ, 2005; pp 466, 467.

234

F€ urstner, A.; Shi, N. J. Am. Chem. Soc. 1996, 118, 12349.

235

Boeckman, R. K., Jr.; Hudack, R. A., Jr. J. Org. Chem. 1998, 63, 3524.

707

13.5 STEREOSELECTIVE ENOLATE REACTIONS

used this reagent to prepare anti-diols with high stereoselectivity. A slight modification of this fundamental reaction uses vinyl iodides.236 In a synthesis of ()-apicularen A, Palimkar and Uenishi, et al. 237 used the NiCl2 catalyzed CrCl2 reaction of vinyl iodide (207) with aldehyde 208 to give a mixture of diastereomers, 209 and 210 (32 and 42% yield, respectively). Intramolecular reactions are known.238 PhCHO, CrCl2, DMF

C10 H21

OTf

Ph

C10 H21

cat NiCl2, 25°C

OH 2-Methylene-1phenyldodecan-1-ol

Dodec-1-en-2-yl trifluoromethanesulfonate

(83%)

Ph

OBn CO2Me

+

O

0.1% NiCl2, CrCl2

O

DMSO

OHC

I

OBn 208

207 OBn

OBn

Ph CO2Me

OH

O

OBn +

O

Ph CO2Me

209 (32%)

OH

O

O

OBn

210 (42%)

13.5 STEREOSELECTIVE ENOLATE REACTIONS When an enolate anion reacts with an alkyl halide, one new stereogenic center is created as in 211 (when R 6¼ H). When a prochiral enolate reacts with an aldehyde or an unsymmetrical ketone to yield an aldolate product (212), two new stereogenic centers are created (when R 6¼ H). There have been many advances in the area of asymmetric enolate anion methodology.239 The purpose of this section is to discuss those factors that influence diastereoselectivity and enantioselectivity in the alkylation and condensation reactions of enolate anions. R2

O

Li R1

R

R2-X

R1

R

O

R2

R2 R3

R1

HO

O

211

O R3

R 212

Several factors influence both the stereoselectivity of hydrogen exchange and enolate anion formation in basepromoted reactions. Houk and coworkers240 found that differing conjugative stabilization by CH p-orbital overlap does not directly influence stereoselectivity. Steric effects dominate only in exceptionally crowded transition structures, but torsional strain involving vicinal bonds contributes significantly to the stereoselectivity of all cases studied. There are two essential factors that control diastereoselectivity in these reactions: the face from which the two reagents approach and the relative orientation of the two molecules.

236

Lee, K.-Y.; Oh, C.-Y.; Ham, W.-H. Org. Lett. 2002, 4, 4403.

237

Palimkar, S. S.; Uenishi, J.; Ii, H. J. Org. Chem. 2012, 77, 388.

238

See Roethle, P. A.; Trauner, D. Org. Lett. 2006, 8, 345. Also see Pospisil, J.; M€ uller, VC.; F€ urstner, A. Chem.–Eur. J. 2009, 15, 5956.

239

Arya, P.; Qin, H. Tetrahedron 2000, 56, 917.

240

Behnam, S. M.; Benham, S. E.; Ando, K.; Green, N. S.; Houk, K. N. J. Org. Chem. 2000, 65, 8970.

708

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

13.5.1 Simple Diastereoselection 13.5.1.1 Alkylation Enolate alkylation with simple aldehydes and ketones does not generally lend itself to enantioselective control due to the planar nature of the enolate π-system.241 Inspection of 213 shows that the si-re face (face a) has no more steric hindrance than the re-si face (face b). When this enolate anion reacts with iodomethane, therefore, no facial selectivity is anticipated and the methylated product will be racemic at the newly formed stereogenic center. In general, enolate alkylation reactions produce chiral, racemic products. The reaction can be diastereoselective, however, when substituents attached to the molecule provide facial bias. In most cases, enolate alkylation proceeds by approach of the enolate anion to the halide from the less sterically hindered face of the enolate anion. In the Molander and Haas242 synthesis of davanone, ketone 214 was treated with LDA and LiCl, then with iodomethane. Three products were formed, 58% of 216, 8% of 217, and 97:1. For cyclohexanone boronic ester 257 [locked into the (E) geometry], the reaction with benzaldehyde gave primarily the anti diastereomer (258, 33:67 syn/anti).256 A Zimmerman-Traxler like model also is useful for predicting the diastereoselection in boron enolates, as well as lithium enolates. O O

1. LDA

OBn-Bu2 Me

2. Bu2BOTf

Pentan-3-one

Cyclohexanone

Ph

2. H3O+

255

256 O

OBn-Bu2

O 1. LDA

1. PhCHO

2. Bu2BOTf

2. H3O+

257

OH

1. PhCHO

Me

OH Ph H

258

13.5.1.4 The Evans’ Model (Z)-Enolates are more stereoselective than (E)-enolates in condensation reactions, even when R1 is not large. The Zimmerman-Traxler model transition states 233–236 do not account for this observation. It has been suggested that the transition states are not chair-like, but skewed as in 259–262.258 In this representation, (Z)-enolate 259 leads to the syn-aldolate. Similarly, (Z)-enolate 260 gives the anti-aldolate, (E)-enolate 261 gives the anti-aldolate, and (E)-enolate 256

Evans, D. A.; Vogel, E.; Nelson, J. V. J. Am. Chem. Soc. 1979, 101, 6120.

257

Reference 243, p 151.

258

(a) Heathcock, C. H. Comprehensive Carbanion Chemistry; Durst, T.; Buncel, E., Eds., Elsevier: Amsterdam, 1985; Vol. 2. (b) Heathcock, C. H.; Buse, C. T.; Kleschick, W. A.; Pirrung, M. A.; Sohn, J. E.; Lampe, J. J. Org. Chem. 1980, 45, 1066. (c) Fellmann, P.; Dubois, J. E. Tetrahedron 1978, 34, 1349. (d) Dubois, J. E.; Fellmann, P. Tetrahedron Lett. 1975, 1225.

713

13.5 STEREOSELECTIVE ENOLATE REACTIONS

262 is the precursor to the syn aldolate. The major steric interactions in this model are those for R1 $ R3 and R2 $ R3. For both (Z)- and (E)-enolates, the R1 $ R3 interaction favors 259 and 261, respectively. The R2 $ R3 interaction is more important for the (E)-enolate and leads to 260 and 262 when R1 is large. The angle at which the reagents approach is obviously important for these transition states. It is assumed that the angle of attack of a nucleophile is 110 degrees as it attacks a carbonyl, the Bu€rgi-Dunitz trajectory (Sections 7.9.4 and 10.6).12 This angle of attack brings R2 and R3 in 261 into close proximity and relieves the R1 $ R3 interaction. Similar results are observed for 260 and 262, but to a lesser degree, which explains why (E)-enolates show comparable R2 $ R3 and R1 $ R3 interactions unless R1 is very large, and also accounts for the large effect of the R2 group on the diastereoselectivity. As R2 increases in size, transition states 259 and 261 become less important.

R3

O

M

[( Z) - anti] O M

R2

[( Z) - syn)]

R2

H

R3

R1 259

O

M O

R2

R2

H H

[(E) - syn]

H H

O

O

O H

[(E) - anti] O M

H R3

H

R1

R3

R1 261

260

R1

262

Evans et al. considered boat conformations 263–264259 as possible transition states, in addition to the chair conformations used in the Zimmerman-Traxler model. In these boat conformations, reaction of an aldehyde with (Z)-enolate via orientation 263 yields the syn-aldolate, but reaction via the opposite orientation of the aldehyde (see 264) yields the anti-aldolate. Likewise, reaction of an aldehyde with an (E)-enolate via the orientation shown in 265 gives the anti aldolate, whereas the opposite orientation (see 266) gives the syn aldolate. A given (Z)- or (E)-enolate, therefore, has one or two chair transition states, or one or two boat transition states that are energetically favorable for an aldol condensation. Only transition states 264 and 265 are considered important, because of the significant R2 $ R3 interactions in 263 and 266. As R2 becomes larger, however, the (Z)-enolate might favor 264 or the chair transition state. Boat transition states predict that increasing the steric bulk of R2 in an (E)-enolate will either have no effect on stereochemistry or increase selectivity for the anti-aldolate. There is some evidence for a modest increase in anti-selectivity with increasing the size of R2.260 R3 H R2

H

O

H

R2

M R1

O

O

R3

O

O

H

263

H

R3 H M R1 264

H

O

R2

O

R3 M

R1

H

O

M

R2

265

R1 266

13.5.1.5 The Noyori Open-Chain Model In the Mukaiyama reaction, the Zimmerman-Traxler and Evans’ models are not satisfactory for predicting diastereoselectivity. Several open (nonchelated) transition states have been considered as useful models. The condensation reaction of carboxylic acid dianions with aldehydes indicated that anti-selectivity increased with increasing dissociation of the gegenion (the cation, M+).261 When analyzing an aldol condensation that does not possess the bridging cation required for the Zimmerman-Traxler model, an aldehyde and enolate adapt an eclipsed orientation as they approach. Noyori et al.262 reported syn-selectivity for the reaction of a mixture of (Z)-silyl enol ether 267 and (E)-silyl enol ether 268 with benzaldehyde, in the presence of the cationic reagent tris(diethylamino) sulfonium (TAS). This reaction is clearly a variation of the Mukaiyama reaction, which usually proceeds with poor diastereoselectivity

259

Evans, D. A.; Nelson, J. V.; Taber, T. R. Topics Stereochem. 1982, 13, 1.

260

Reference 244, p 151.

261

Mulzer, J.; Zippel, M.; Br€ untrup, G.; Segner, J.; Finke, J. Liebigs Ann. Chem. 1980, 1108.

262

Noyori, R.; Yokoyama, K.; Sakata, J.; Kuwajima, I.; Nakamura, E.; Shimizu, M. J. Am. Chem. Soc. 1977, 99, 1265.

714

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

(Section 13.4.3). The reaction of the (Z)-enolate 267 and the (E)-enolate (268), led to formation of 269 as the major product. The 86:14 (269:270) preference263 for the syn-diastereomer was independent of the enolate geometry. OSiMe3

OSiMe3

+

C3H7

C4H9

C4H9

C3H7

267

Me3SiO

O

Me3SiO

PhCHO, TAS

Ph

Me3SiF, –70°C

+

C4H9

Ph

C4H9

C3H7

268

O

C3H7

269

270

(65%)

These results are predicted with a so-called open-chain model. The (Z)-enolate will form a transition state (e.g., 271), which leads to the anti-diastereomer. The (E)-enolate will react via 272 to give the syn-aldolate. In both cases, the opposite orientation will lead to a significant R2 $ R3 interaction that destabilizes that transition state. This observation does not explain why both (E)- and (Z)-enolates lead to the syn-diastereomer (269), and some equilibration or isomerization must occur during the course of the reaction, which will be discussed in Section 13.5.1.6. O–

R1 R3

R3

R1

R1

H

O–

R3

H

H

H

R2

H

O

–O

O

H

R2

R2

O–

271

–O

R3

R1

H

R3

R1

R1

O–

O

R3

H H

R2 –O

O

R2

H

R2

H

H O–

272

13.5.1.6 Isomerization in the Aldol Condensation It is clear from the previous discussion that steric effects control the orientation of the molecule as it approaches the enolate anion. In simple diastereoselection processes, one observes moderate to good syn-anti selection depending on the steric demands of the enolate and the carbonyl partner. However, other processes are at work in this condensation that lead to variations in the syn-anti ratio. Heathcock264 reported the selectivity for a variety of lithium enolate anions derived from 273 with benzaldehyde (see Table 13.3).29 The (E)- and (Z)-enolate anions (275 and 274, respectively) can be trapped with trimethylsilyl chloride to yield silyl enol ethers 277 and 276, respectively. The observed 276:277 ratio is taken to be the ratio of (Z)- and (E)-enolates originally generated from 273. When the enolate mixture reacted with benzaldehyde, the syn-aldolate 278 and the anti-aldolate 279 were formed.264 It is clear from Table 13.3 that the (E/Z) enolate ratio (taken to be 278:279) is not the same as observed in the aldolate products. One explanation is that the enolate anions equilibrate under the reaction conditions. Isomerization may occur during or after the aldolateforming step, and the enolate ratio may be the same, but the aldehyde adds from different faces of the enolate anion. Me3SiCl

R O

Base

R

Me

O– Li+

–78°C

273

274

R 275

+

1. PhCHO

OH

2. H3O+

278

Noyori, R.; Nishida, I.; Sakata, J. J. Am. Chem. Soc. 1981, 103, 2106. Reference 243, pp 122–132.

O

OH R

Me

264

R 277

276

Ph

263

OSiMe3

Me

OSiMe3

O– Li+

+ Me

R

Me

+

Ph

O R

Me 279

715

13.5 STEREOSELECTIVE ENOLATE REACTIONS

TABLE 13.3 Stereoselectivity in Enolate Formation and Diastereoselectivity in the Aldol Condensation R

Base

LiO

LDA

MeO

LDA

5 : 95

62 : 38

t-BuO

LDA

5 : 95

49 : 51

i-Pr2N

LDA

81 : 19

63 : 57

LTMP

52 : 48

68 : 32

MeLi

100 : 0

50 : 50

0 : 100

65 : 35

LDA

30 : 70

64 : 36

LICA

35 : 65

62 : 38

LHMDS

66 : 34

77 : 23

LTMP

20 : 80

66 : 34

LDA

60 : 40

82 : 18

LICA

59 : 41

75 : 25

LHMDS

>98 : 2

90 : 10

LTMP

32 : 68

58 : 42

LDA

>98 : 2

>98 : 2

LHMDS

>98 : 2

>98 : 2

LDA

>98 : 2

88 : 12

LICA

>98 : 2

87 : 13

LHMDS

>98 : 2

88 : 12

LTMP

>98 : 2

83 : 17

H

Et

i-Pr

t-Bu

Ph

276 : 277

278: 279 45 : 55

Reprinted with permission from Heathcock, C. H.; Buse, C. T.; Kleschick, W. A.; Pirrung, M. A.; Sohn, J. E.; Lampe, J. J. Org. Chem. 1980, 45, 1066. Copyright © 1980 American Chemical Society.

The nature of the amide base had a substantial effect on the (E/Z) ratio of the enolate anion. In general, LTMP gave mainly the (E)-enolate and LHMDS gave the (Z)-enolate anion, which is probably related to the aggregation state of the base. As the size of R increases, the use of LTMP and LDA increases the percentage of (Z)-enolate, which is related to the conformation of the ketone immediately prior to deprotonation, using the Felkin-Anh model (Section 7.9.2). Conformation 280 is more stable than 282 for propanal by 0.8 kcal (3.35 kJ) mol1.265 This difference is probably greater in ketones (R ¼ alkyl or aryl) and esters (R ¼ Odalkyl). This observation suggests that as R becomes larger the (Z)-enolate is preferred, regardless of the base. O–

O

Me

Me H

•

H

H

R

H

B

R

R 280

281 O–

O H

•

H Me

R 282

Karabatsos, G. J.; Hsi, N. J. Am. Chem. Soc. 1965, 87, 2864.

–O

H

R

Me

•

H

265

Me

O

•

H

Me

–

H R 283

B

716

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

The observation that LTMP gives more (E)-enolate can be explained by a steric interaction between the base and the methyl group (see 280). The methyl $ R interaction is greater, but the interaction of the methyl group and the base is minimized in 283. This model assumes approach of the base and the ketone is not along the axis of the CdH bond, but over the face of the incipient enolate plane.266 With benzaldehyde, (E)- and (Z)-enolates may exhibit different degrees of stereoselectivity. The results in Table 13.3 can be explained by the Zimmerman-Traxler like transition state (284). The R2 $ R5 and 1 R $ R4 interactions are very important. When R5 is large (e.g., adamantyl) the (Z)-enolate yields the syn-aldolate, and the (E)-enolate yields the anti-aldolate. When R5 is relatively small (Ph, Et), the (Z)-enolate gives primarily the synaldolate. When R5 ¼ H, reactions of the (Z)-enolate become stereorandom and lead to diastereomeric mixtures of products. When R5 is relatively small (Ph, Et), the (E)-enolate usually yields a 50:50 mixture of syn- and anti-aldolates, probably because the R1 $ R4 interaction balances the R2 $ R5 interaction. When R5 ¼ H or Me, the (E)-enolate is selective for the syn-diastereomer. R3 O–

O

M

R1

R5 R4 R2

284

A complicating factor is the observation that aldolates can undergo syn-anti equilibration by enolization or by reverse aldolization. Aldolates [e.g., ethyl (R)-3-hydroxypentanoate] can be deprotonated to the dianion (285), which undergoes alkylation with iodomethane to yield the anti-product, ethyl (2R,3R)-3-hydroxy-2-methylpentanoate, as shown.267 This equilibration is clearly the basis of the aldol-transfer reaction discussed in Section 13.4.1.1. If a new enolate anion is generated from ethyl (2R,3R)-3-hydroxy-2-methylpentanoate, equilibration can lead to a mixture of syn- and anti-products. The primary mechanism for syn-anti equilibration appears to be reverse aldolization.268 A retro-aldol condensation will convert the syn-diastereomer (286) into the aldehyde and enolate components, which can regenerate 286 or form the anti-diastereomer 287. Syn-anti equilibration can be much slower than reverse aldolization, as with the (Z)-enolate of 2,2-dimethylpentan-3-one.261 Aldolates derived from the more basic ketone enolates are more likely to suffer reverse aldolization than aldolates derived from the less basic enolates of esters, amides, or carboxylate salts. Steric crowding in an aldolate promotes reverse aldolization. The metal is very important, and some metals form stable chelates that generate aldolates resistant to reverse aldolization. Boron enolates, for example, do not undergo equilibration even at elevated temperatures. Lithium enolate anions equilibrate more slowly than other alkali metal enolates and potassium enolates equilibrate rapidly. OH

Li

O

O

2 LDA

Ethyl (R)-3-hydroxypentanoate M

O

Me Ethyl (2R,3 R)-3-hydroxy2-methylpentanoate M

R2 R

R2

OEt 285

O

R3

O

MeI

OEt

H

OEt

OH

Li

O

R3

O–

CHO + R1

286

O

O

R3

R R2 287

266

(a) Ireland, R. E.; Willard, A. K. Tetrahedron Lett. 1975, 3975. (b) Nakamura, E.; Hashimoto, K.; Kuwajima, I. Ibid. 1978, 2079.

267

(a) Fráter, G. Helv. Chim. Acta 1979, 62, 2825, 2829. (b) Seebach, D. Wasmuth, D. Ibid. 1980, 63, 197.

268

Reference 243, pp 161, 162.

717

13.5 STEREOSELECTIVE ENOLATE REACTIONS

13.5.2 Selectivity With Chiral, Nonracemic Reactants 13.5.2.1 Diastereoface Selectivity Heathcock and coworkers defined three kinds of stereoselection for the aldol condensation: simple diastereoselection, diastereoface selection,250,269 and double stereodifferentiation.250,261 Simple diastereoselection occurred when an achiral aldehyde and an achiral ketone reacted to yield a mixture of syn- and anti-aldol products, 288 and 289. This example illustrates the type of reaction discussed previously. OH

O R

+

CHO

O

R1

+

R1

R

2. H3O+

O

OH

1. Base

R1

R

Me

Me

288

289

If either the enolate or carbonyl partner has a stereogenic center near the reactive center, it will influence both orientation and facial selectivity.250,258 A reaction of an achiral enolate anion (e.g., 291) with an aldehyde or a ketone that possesses an (S) stereogenic center (290) will give the (S,R,S) diastereomer (292) and the (S,R,R) diastereomer (293) aldolate products. The (S) stereogenic center in 290 can induce high diastereoselectivity, with 292 as the major product. An aldehyde with an (R) stereocenter will similarly influence the reaction, giving 293 as the major aldolate product. Several factors influence the extent of diastereoselectivity. A second variation reacts an enolate possessing a stereogenic center (e.g., 294) with an achiral aldehyde or ketone. In (R)-294, the reaction with an aldehyde leads to the (R,S,R) diastereomer 295 and the (S,R,R) diastereomer, 296. R Me

OH

O–

CHO

H3O+

+

H

R

R1

Me

290 (S) O–

OH R1

OH

+

R1

1. R

O

R1

R1

R

2. H3O+

Me

Me 294 (R)

Me

+

Me 293

OH

CHO

O

R

Me Me 292 [(S,R,S) - syn, syn]

291

Me

O

[(S,R,R) - syn , anti]

O R1

R

Me

Me

295 [( R,S,R) - syn , syn]

296

Me [(S,R,R) - anti , syn]

Double stereodifferentiation270 results from the reaction of an enolate possessing a stereogenic center and an aldehyde or ketone that also has a stereogenic center. An example, taken from Calter and Liao’s271 synthesis of siphonarienedione, condensed chiral aldehyde 297 with enolate anion 298. In this reaction, stereoisomer 299 was produced in 65% yield as a result of double stereodifferentiation. O

Me

H

N

+

O O–

MeO

Li+

1. THF/CH2Cl2 Hexanes , –78°C 2. H3O+

C3H7 297

298

O Me

O

OH C3H7

N OMe 299 (65%)

Improvements in diastereoselection arise from factors that influence the orientation of the reaction, and the face from which the two reactants approach. For simple diastereoselectivity, steric factors influence the orientation of

269

Also see (a) White, C. T.; Heathcock, C. H. J. Org. Chem. 1981, 46, 191. (b) Heathcock, C. H.; Young, J. P.; Hagen, M. C.; Pirrung, M. C.; White, C. T.; VanDerveer, D. J. Org. Chem. 1980, 45, 3846.

270

(a) Horeau, A.; Kagan, H.-B.; Vigneron J.-P. Bull. Soc. Chim. Fr. 1968, 3795. (b) Izumi, Y.; Tai, A. Stereodifferentiating Reactions; Kodanshar/ Academic Press: Tokyo/New York, 1977.

271

Calter, M. A.; Liao, W. J. Am. Chem. Soc. 2002, 124, 13127.

718

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

the aldehyde and the enolate anion, leading to good syn-anti selectivity using the models previously described. When a stereogenic center is incorporated as a structural unit of the enolate anion or aldehyde, increased steric hindrance affects the approach and transition state energies, leading to greater diastereoselectivity. The presence of the stereogenic center makes one face of the molecule more hindered than another when comparing the possible transition states, and a preference for one over another is called facial selectivity. One study examined the condensation of an aldehyde possessing a stereogenic center, as in (R)-2-phenylpropanal with various enolate anions (300).272 In reactions with ester enolate anions, as well as ketone enolate anions, where M ¼ Li, ZnBr, and BR2, the syn-diastereomer (301) predominated over the anti-diastereomer (302) by 3:1 to 4:1. Me

O– M+

Me

+

Ph

R

Ph

2. H3O+

R

CHO

OH (R)-2-Phenylpropanal

Me (Anti)

(Syn)

1. THF

300

+

R

Ph

O

OH

301

O

302

In most enolate anion condensation reactions, Crams rule (Section 7.9.1) predicts the major isomer if the reaction partner (or partners) contain a stereogenic center.273 To understand how this rule applies to orientational and facial selectivity, the transition state of the reaction must be examined. The Zimmerman-Traxler model is used most often, and if it is applied to the reaction of (R)-2-phenylpropanal and 300, the model predicts the observed syn selectivity. The facial selectivity shown in 303 and 304 arises from the methyl group. In 304, the enolate approaches from the face opposite the methyl, leading to diminished steric interactions and syn product (305). If the enolate approaches via 303, the steric impedance of the methyl group destabilizes that transition state relative to 304. In both 303 and 304, a Cram orientation is assumed (see 306 and 307; Section 7.9.1), although reaction via other rotamers is possible. The appropriate rotamer for reaction is that where RL is anti to the carbonyl oxygen, as shown in 306 and 307. Since the phenyl group is RL, 303 and 304 are assumed to be the appropriate orientation for the aldehyde. If an aldehyde or ketone follows antiCram selectivity, this aldehyde orientation must be adjusted. H Me O

M

R

versus

Ph

R

O

O H

H

Me O

H

M

OH H

H

303

304 Enolate—O

305

Li O

RS

RM • R1 RL

306

O

Ph

Ph

R

Me

O RM

RS H

Li

• R1 RL

307

Cram selectivity results from approach of the enolate anion over the less hindered face of the stereogenic center of the aldehyde (over the H, RS) rather than over the larger methyl group (RM). The phenyl group is taken to be RL, and it is anti to the carbonyl oxygen. This model assumes a reasonable difference in size between RM or RS and the enolate anion, which is the source of steric repulsion, where 306 is preferred to 307. If the difference in size (steric hindrance) between the two groups is small, there is little facial selectivity in the transition state, and poor enantioselectivity is predicted in the final product. The presence of a stereogenic center in the enolate imparts facial selectivity, and reaction of (S)-3-methylpentan-2-one was treated with LDA and chiral enolate 308 was formed.274 Subsequent reaction with propanal gave the aldolates with a slight preference (57:43 to 63:36) for the anti (S,R)-isomer [(3S,6R)-6-hydroxy-3-methyloctan-4-one] over the syn 272

Reference 243, p 166.

273

(a) Cram, D. J.; AbdElhafez, F. A. J. Am. Chem. Soc. 1952, 74, 5828. (b) Cram, D. J.; Kopecky, K. R. Ibid. 1959, 81, 2748.

274

Seebach, D.; Ehrig, V.; Teschner, M. Liebigs Ann. Chem. 1976, 1357.

719

13.5 STEREOSELECTIVE ENOLATE REACTIONS

(S,S)-isomer [(3S,6S)-6-hydroxy-3-methyloctan-4-one]. Once again, the methyl group induced a facial bias in the approach of the reagent, leading to attack on the face opposite the methyl. The results were similar for both lithium and boron enolates. If the transition states for the boronic esters use Cram’s rule as a model,275 the observed facial bias favors approach of the reagent over the smaller group since it should generate the least hindered transition state. Me

Me

Me Me

Me

EtCHO

LDA THF

Et

O– M+

O (S)-3-Methylpentan-2-one

O

Et

+

OH

O

(3S,6 R)-6-Hydroxy-3methyloctan-4-one

308

OH

(3S,6S)-6-Hydroxy-3methyloctan-4-one

Diastereoface selection is observed when a chiral auxiliary is attached to either the enolates or the carbonyl substrate. A chiral auxiliary is an asymmetric group attached to an achiral substrate. The chirality of the auxiliary influences the course of the reaction (an aldol condensation) by inducing a chiral transition state, which influences the chirality of the final product. The auxiliary is cleaved to give the normal aldolate product. Oppolzer et al.276 developed a sulfonamide alcohol auxiliary based on camphor (Oppolzer’s sultam) in which an acid derivative is attached as a chiral ester. The auxiliary was removed by reduction of the ester group with LiAlH4. In Mori’s synthesis of (1S,3S,7R)-3methyl-α-himachalene,277 the sex pheromone of the sand fly Lutzomyia longipalpis, amide 309 was prepared from the appropriate acid chloride and Oppolzer’s sultam.278 Treatment with butyllithium and iodomethane gave an 81% yield of 310, and subsequent removal of the auxiliary by basic hydrolysis gave 311 in 67% yield and >99 %ee. Helmchen and coworkers279 achieved good asymmetric induction in the alkylation of ester enolate anions using a camphor-based sulfonamide auxiliary Other chiral auxiliaries are commonly used. In Nagumo and coworker’s280 synthesis of sekothrixide, a so-called Evans’ auxiliary (the chiral, substituted oxazolidinone unit) was used in 312. Generation of the enolate anion with sodium hexamethyldisilazide was followed by reaction with iodomethane to give an 83% yield of 313. Me N O

N

BuLi

O2S

Me

O

LiOH, H2O2

O2S

O

MeI, THF–HMPA

311 (>99%)

310 (81%)

309

OH

THF–H2O

Bn

Bn 1. LiN(SiMe3)2, THF –70 to –10°C

N

N

2. CH3I

OMPM

O

O

OMPM

312

O

O

313 (83%)

Using a chiral additive often generates a reactive species that functions as a transient auxiliary, and constitutes another method for inducing enantioselectivity in aldol-like reactions. Koga and coworkers281 reacted the achiral ketone acetophenone with a chiral, nonracemic base (314) in the presence of TMEDA, to form an enolate-coordinated 275

Evans, D. A.; Taber, T. R. Tetrahedron Lett. 1980, 21, 4675.

276

Oppolzer, W.; Dudfield, P.; Stevenson, T.; Godel, T. Helv. Chim. Acta 1985, 68, 212.

277

Tashiro, T.; Bando, M.; Mori, K. Synthesis 2000, 1852.

278

Oppolzer, W.; Chapuis, C.; Dupuis, D.; Guo, M. Helv. Chim. Acta 1985, 68, 2100.

279

Schmierer, R.; Grotemeier, G.; Helmchen, G.; Selim, A. Angew. Chem. Int. Ed. 1981, 20, 207.

280

Terayama, N.; Yasui, E.; Mizukami, M.; Miyashita, M.; Nagumo, S. Org. Lett. 2014, 16, 2794.

281

Muraoka, M.; Kawasaki, H.; Koga, K. Tetrahedron Lett. 1988, 29, 337.

720

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

chiral complex. Reaction of this complex with benzaldehyde at 100°C led to asymmetric induction in the final aldolate product, (R)-3-hydroxy-1,3-diphenylpropan-1-one, which was produced in 73% yield with high enantioselectivity for the (R)-enantiomer.281 List and Notz282 showed that the addition of 30 mol% of L-proline induced an asymmetric aldol condensation between hydroxyacetone (1-hydroxypropan-2-one) and cyclohexanecarbaldehyde to give (3S,4S)4-cyclohexyl-3,4-dihydroxybutan-2-one in 60% yield, with >20:1 regioselectivity, >20:1 diastereoselectivity and >100:1 enantioselectivity. The use of chiral Lewis bases in the aldol reaction has been reviewed,283 as have catalytic aldol reactions.284 Ph

H

O Me , TMEDA

1. Ph

N

N Li

Ph

1. PhCHO, –100°C

2. C4H9Li

2. H2O

Ph H

O

314

OH

(R)-3-Hydroxy-1,3-diphenylpropan-1-one O

O

O

(73%)

OH

30% L-Proline, DMSO

+

H rt, 2 d

OH

OH

1-Hydroxypropan-2-one

(3S,4S)-4-Cyclohexyl-3,4(60%) dihydroxybutan-2-one

Cyclohexanecarbaldehyde

The presence of a chiral auxiliary can influence the stereochemistry of products formed in a Mukaiyama aldol reaction. A stereogenic center derived from a chiral amino alcohol was incorporated in the enol ether moiety (see 315). When reacted with benzaldehyde and TiCl4, the alcoholate products were 316 and 317 (R* ¼ the chiral auxiliary). The diastereoselectivity was quite good (95:5 favoring 316), and each product was formed with high enantioselectivity.285 It is also possible to incorporate the stereogenic center into the starting material (a chiral template, as discussed in Section 8.9).286 Ph Me H

Me

H

NMe2

O

Me PhCHO, TiCl4

OSiMe3

Me

*RO

CH2Cl2

315

Ph O

OH 316

+

*RO

Ph O

OH 317

13.5.2.2 Double Stereodifferentiation Double stereodifferentiation is used for a reaction that employs a stereogenic center on both the enolate anion and the aldehyde or ketone (see 299 above). The incipient chirality of each reactant will influence the reaction, but there is a lower energy transition state during the reaction with compatible, or matched stereochemistry partners (consonant double stereodifferentiation). Alternatively, the reaction with incompatible, or mismatched stereochemistry partners will lead to a higher energy transition state (dissonant double stereodifferentiation).260,287 This concept of matched and mismatched was introduced in the discussion of the Sharpless asymmetric epoxidation in Section 6.4.4.1. The reaction of chiral ketone 318 with LDA generates the corresponding enolate anion, and subsequent reaction with benzaldehyde

282

Notz, W.; List, B. J. Am. Chem. Soc. 2000, 122, 7386.

283

Denmark, S. E.; Stavenger, R. A. Acc. Chem. Res. 2000, 33, 432.

284

Machajewski, T. D.; Wong, C.-H. Angew. Chem. Int. Ed. 2000, 39, 1352.

285

Gennari, C.; Molinari, F.; Cozza, P.; Oliva, A. Tetrahedron Lett. 1989, 30, 5163.

286

Woodward, R. B.; Logusch, E.; Nambiar, K. P.; Sakan, K.; Ward, D. E.; Au-Yeung, B.-W.; Balaram, P.; Browne, L. J.; Card, P. J.; Chen, C. H.; Ch^enevert, R. B.; Fliri, A.; Frobel, K.; Gais, H.-J.; Garratt, D. G.; Hayakawa, K.; Heggie, W.; Hesson, D. P.; Hoppe, D.; Hoppe, I.; Hyatt, J. A.; Ikeda, D.; Jacobi, P. A.; Kim, K. S.; Kobuke, Y.; Kojima, K.; Krowicki, K.; Lee, V. J.; Leutert, T.; Malchenko, S.; Martens, J.; Matthews, R. S.; Ong, B. S.; Press, J. B.; Rajan Babu, T. V.; Rousseau, G.; Sauter, H. M.; Suzuki, M.; Tatsuta, K.; Tolbert, L. M.; Truesdale, E. A.; Uchida, I.; Ueda, Y.; Uyehara, T.; Vasella, A. T.; Vladuchick, W. C.; Wade, P. A.; Williams, R. M.; Wong, H. N.-C. J. Am. Chem. Soc. 1981, 103, 3210. 287

Heathcock, C. H.; White, C. T. J. Am. Chem. Soc. 1979, 101, 7076.

721

13.5 STEREOSELECTIVE ENOLATE REACTIONS

occurs with high selectivity for the syn-isomer (319 rather than 319). This reaction demonstrates the incipient selectivity of an aldol condensation with chiral ketone 319. HO

O

O O

O 1. LDA

O

O

O

O

O

+

Me

O

O

O

318

Ph O

O

2. PhCHO

O

HO

O

Ph

Me

O

O

319

320

If the reaction of 318 with the achiral benzaldehyde is replaced with a reaction with a chiral aldehyde, matched versus mismatched reactions in an aldol condensation can be compared. The condensation of the enolate anion of 318 with the (R)-aldehyde (321) or the (S)-aldehyde (324) is illustrative.250,287 When 318 reacted with the mismatched aldehyde 321, the selectivity was poor due to dissonance in the transition state, with formation of a 62:38 mixture of 322 and 323. Some anti-diastereomers were also produced, which is a quite different result when 318 reacted with 324, the matched aldehyde. Consonant stereodifferentiation in this latter case led to excellent selectivity for a single isomer (325), produced in > 97% yield.

O O

2.

O

CHO O

O

O 318

+

Me O

O O

O O

321

322

Me O

O 323

OH

O

O

O

O

O

O O

O

1. LDA

O O

O

O

O

OH

O

O

1. LDA

O

OH

O

O

2.

O

O

CHO

O

O

O

318

O

O

Me O

O

324

325

The steric effects of the consonance versus the dissonance approach can be described in terms of the generalized Zimmerman-Traxler models 326 and 327. In 326, the R and R2 groups are oriented away from each other (matched), minimizing the steric interaction. In 327, however, the R $ R2 interaction is maximized (mismatched), destabilizing that transition state. O–

O O–

R2 (Matched , consonant)

H

R1

H R1

R3 H

Versus

H

R

O

(Mismatched , dissonant)

R2 H H

R R3

326

327

Good selectivity is observed in the condensation of chiral aldehyde (S)-2-phenylpropanal with the chiral enolate anion 328. The matched transition state has the methyl groups oriented away from each other, which is lower in energy than the opposite case in which the methyl groups interfere with each other (both options assume a Cram orientation). The matched orientation predicts formation of the anti-diastereomer 329 for this reaction, as shown. When applied to the Mukaiyama reaction, stereodifferentiation leads to excellent diastereoselectivity and enantioselectivity. When silyl ketene acetal 330 reacted with chiral aldehyde (S)-3-(benzyloxy)-2-methylpropanal in the presence of TiCl4, a 75% yield of 331 and 332 was obtained as a 98:2 mixture favoring 331.288 288

Shirai, F.; Nakai, T. Tetrahedron Lett. 1988, 29, 6461.

722

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

Me

O Ph

+ Ph

H

Et3SiO

OSiEt3

O

Me

OH

OH

O

Me MeO2C

OH

OBn

OBn H

(S)-3-(Benzyloxy)2-methylpropanal

330

Me 329

TiCl4

H

Me

Ph

328 Me

+

O

Ph

–

Me (S)-2-Phenylpropanal

OMe

OH

H3O+

+

Me MeO2C

Me 331

OH OBn H

Me

332

13.5.3 Chelation Control Chelating substituents influence the stereoselectivity of enolate condensation and alkylation reactions, usually by generating a complex that is conformationally rigid. When β-hydroxyl ester 333 was treated with 2 equiv. of LDA, the dilithio derivative was generated. Lithium was chelated to both oxygen atoms of the dianion, and this type of intermediate was seen in previous sections (see 48 in Section 13.3.1 and 314 in Section 13.5.2.1).30,289 Reaction of the (Z)-enolate (333) with an electrophile leads to its delivery from the less hindered face, giving the anti-product 335. In the (E)-enolate (334), chelation again leads to delivery of the electrophile from the less hindered face (away from R1), and 335 is again the major product.30,289 If the hydroxyl group is blocked (protected), the chelating effect is greatly diminished or removed. Reaction of OSiR3 derivatives of 333, where the chelating effect is removed because there is no OdH unit, with LDA/THF and an electrophile gave the opposite aldolate ratio of stereoisomers with an (E/Z) ratio of > 90:10.30,289 Kraus and Taschner289 used this analysis (334 and 335) to predict the stereoselectivity of an epoxidation reaction that gave glycidic esters (336). The model was later applied to enolate alkylation and condensation reactions, as shown.30 E+

Li OH

O

R1

R2

O

Li

O

2 LDA

+

Li

OR2

H

H

O

OH

Li O

O

E+

O

R1

R1

OR2

OR2 E

R1 334

333

336

335

Evans and Takacs290 prepared several chiral auxiliaries derived from amino alcohols (e.g., valinol or prolinol), know known as Evans’ auxiliaries (see 312). Prolinol amides (e.g., 337) preferentially form the (Z)-enolate (338) over the (E)-enolate (339). Alkylation proceeds with chelation control and good diastereoselectivity (from the si face) to give the alkylated products 340 and 341, favoring 340.30,290 The (Z)-enolate (338) is preferred over 339, since 338 is chelated and 339 is not. Addition of the electrophile from the less hindered face of each enolate anion yields the observed products. Evans et al.291 also prepared asymmetric carbamates that showed opposite diastereoface selectivity. In both cases, small alkyl halides (e.g., methyl iodide) showed less stereoselectivity than the more sterically demanding benzyl bromide. Diastereoselectivity is observed in reactions of carbanions derived from imines and hydrazones when those species contain a stereogenic center or a chiral auxiliary (Section 13.4.6). Asymmetric imines can be used, and chiral oxazoline derivatives have also been prepared and used in the alkylation sequence (Section 13.3.1). Li

OH

O

O O

2 LDA

N

Li

+

R1

R1

N

R1

Li 337

338

HO

Li

O

HO

O

O

+

N O 339

E

R1

N E

+ R1

N E

340

341

289

Kraus, G. A.; Taschner, M. J. Tetrahedron Lett. 1977, 4575.

290

(a) Evans, D. A.; Takacs, J. M. Tetrahedron Lett. 1980, 21, 4233. (b) Sonnett, P. E.; Heath, R. R. J. Org. Chem. 1980, 45, 3138.

291

Evans, D. A.; Ennis, M. D.; Mathre, D. J. J. Am. Chem. Soc. 1982, 104, 1737.

723

13.5 STEREOSELECTIVE ENOLATE REACTIONS

Meyers et al.292 showed that chiral oxazoline (342) could be alkylated to yield the ethyl derivative 343. A second alkylation generated the diastereomeric product 344, and hydrolysis provided the chiral lactone (345) in 58% yield and with a selectivity of 70 %ee for the (R)-enantiomer.289 In a synthesis of integric acid, Rutjes and coworker’s293 prepared the chiral hydrazone 347 by reaction of pentanal with the hydrazone known as N-amino-(2R)-(methoxymethyl) pyrrolidine (346).294 Subsequent treatment with LDA, and then iodomethane, gave the alkylated product 347 in 75% yield. Ozonolysis cleaved the RAMP group to give methylated aldehyde (R)-2-methylpentanal in >60% yield. The enantiomer of 346 is known as SAMP, N-amino-(2S)-(methoxymethyl)pyrrolidine and both enantiomers have been applied in organic synthesis.295 H

Me

O

O Ph

N

N

2. MeI

OMe

OSiMe3

OMe 344

+

1. DCM, MS 4 Å

O3, –78°C CH2Cl2

N N

2. LDA; MeI THF

Pentnanal

345 (58%)

OMe

OMe NH2 N

O

OSiMe3

343

CHO

H

N

OMe

342

H3O+

Ph 2. I

O

Me

O

1. LDA

Ph

1. LDA

Me

346

CHO

(R)-2-Methylpentanal ( >60%)

347 (75%)

13.5.4 Diastereoselectivity in Alkylidene Enolates Cyclic compounds cyclohexane carboxylic acid or cyclohexanecarbaldehyde generate enolate anions that are unique. These enolate anions have an exocyclic double bond that can exist as (E)- and (Z)-isomers. The facial and orientational bias in alkylation and condensation reactions of such enolate anions is influenced by the conformation of the ring it is attached to. Alkylidene cyclohexane enolate anions show a preference for equatorial attack, as observed in addition reactions of cyclohexanone derivatives (Section 7.9.4). OLi CO2Me R

a

OMe

1. LDA

MeO2C

CO2Me R

Me

R

b

MeO

H b

349A

Me +

H Me

2. MeI

348

Me

a

OLi

Major 349B

350

351

The position and stereochemistry of substituents on the ring influences the selectivity. In 1,2-chiral systems (e.g., 348) treatment with LDA leads to an equilibrating mixture of enolate anions (349A and 349B). The destabilizing Me $ OMe interaction in 349A is absent in 349B, and reaction with CH3I proceeds from the less sterically hindered path b to yield 350 rather than 351.30,296 When the R in 348 was hydrogen, a 70:30 mixture of 350/351 was obtained, but when R was methyl, an 80:20 mixture was obtained, favoring 351. This result is opposite to results observed in the reduction of cyclohexanone derivatives. One explanation for this difference is that enolate anions (e.g., 349A) exist as aggregates (dimers or tetramers) that increases the blockage to attack via path a. A 1,3-disubstituted system is shown with ester 352, where treatment with LDA generates the equilibrating enolate anions 353A and 353B, favoring 353B. When the enolate precursor was the acid (R ¼ H in 352), alkylation with iodomethane showed a slight preference for formation of 353B over 353A. Note that reaction of the carboxylic acid 292

(a) Meyers, A. I.; Yamamoto, Y.; Mihelich, E. D.; Bell, R. A. J. Org. Chem. 1980, 45, 2792. (b) Hoobler, M. A.; Bergbreiter, D. Z.; Newcomb, M. J. Am. Chem. Soc. 1978, 100, 8182. 293

Waalboer, D. C. J.; van Kalkeren, H. A.; Schaapman, M. C.; van Delft, F. L.; Rutjes, F. P. J. T. J. Org. Chem. 2009, 74, 8878.

294

(a) Enders, D.; Eichenauer, H. Chem. Ber. 1979, 112, 2933. (b) Enders, D.; Frey, P.; Kipphardt, H. Org. Synth. 1987, 65, 173.

295

See (a) Job, A.; Janeck, C. F.; Bettray, W.; Peters, R.; Enders, D. Tetrahedron 2002, 58, 2253. (b) Enders, D.; Lenzen, A.; Backes, M.; Janeck, C.; Catlin, K.; Lannou, M.-I.; Runsink, J.; Raabe, G. J. Org. Chem. 2005, 70, 10538.

296

(a) Krapcho, A. P.; Dundulis, E. A. J. Org. Chem. 1980, 45, 3236. (b) Johnson, F.; Malhotra, S. K. J. Am. Chem. Soc. 1965, 87, 5492, 5493.

724

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

proceeded via formation of a dianion. When R1 was methyl (R ¼ H in 352), however, a 52:48 mixture of 353A/353B was obtained. The methyl ester of 352 (R ¼ methyl) showed a distinct preference for 353A, when R1 was OMe, where a 78:22 mixture of 353A/353B was obtained. When R1 ¼ methyl and R ¼ OMe, a 90:10 mixture of 353A/353B was obtained. This mixture is explained by delivery of iodomethane from path a in 353B (equatorial attack).295b Presumably, steric interactions with the OR group of the enolate anion moiety are minimal (or equal) in both conformations, making the conformation with the equatorial R1 group (353B) preferred. CO2R

OLi

a 1. LDA

R1 b

R1

H

352

Me

Me

353A

R1

OLi

RO

b

CO2R

+

a

H

2. MeI

RO2C

Major

R1

OR

353B

R1

354

355

In the 1,4-disubstituted system (356), reaction with LDA generated the alkylidene enolate 357. Alkylation with iodomethane proceeded via path b (interaction with the 1,3-diaxial hydrogen atoms in 357 destabilized attack along path a) to yield the major product (359).295b,297 An 84:16 mixture of 358/359 was obtained when R was either OMe or t-Bu. Addition of the electrophile favored equatorial alkylation (path b to yield 358), but a mixture of 358 and 359 was obtained. CO2Me

MeO2C

a

Me

Me

CO2Me

OLi

1. LDA

+

OMe

R

2. MeI

b

357

R

R

358

359

R 356

Five- and six-membered ring enolate anions derived from cyclopentanone and cyclohexanone derivatives have an endocyclic double bond, and reactions with an electrophile occur from the sterically less hindered face. Enolate anion 361 was obtained by treatment of 360 with lithium amide. Subsequent reaction with an alkyl halide led to delivery of the halide from the face opposite the alkenyl group (path a), to give the trans product shown (363) in 60% yield.298 Approach via path b would have serious steric consequences, and that transition state is destabilized. Similar effects are observed with 3-alkyl-cyclohexanone derivatives.297–299 C6H13

TMSO

O b X

NH3

CO2H

CO2H

O-Li C6H13

361

360

a

362 O–

O t-BuOK

Me

Me

Me

Me

O

O–

+ H

H

Br

Br

Br 363

C6H13

H

364

365

H 366

Delivery of an electrophile to the less hindered face of an enolate also occurs in intramolecular alkylation reactions. When 363 was treated with potassium tert-butoxide, a mixture of (E)- and (Z)-enolates (364 and 365, respectively) was obtained. Intramolecular displacement of bromide generated a single isomer (366).300 In this case, the electrophile can 297

House, H. O.; Bare, T. M. J. Org. Chem. 1968, 33, 943.

298

Patterson, J. W., Jr.; Fried, J. H. J. Org. Chem. 1974, 39, 2506.

299

Posner, G. H.; Sterling, J. J.; Whitten, C. E.; Lentz, C. M.; Brunelle, D. J. J. Am. Chem. Soc. 1975, 97, 107.

300

House, H. O.; Sayer, T. S. B.; Yau, C.-C. J. Org. Chem. 1978, 43, 2153.

725

13.6 ENAMINES

approach the enolate from only one face (the bottom or α-face). Because of this conformational constraint, both (E)- and (Z)-enolate anions lead to the same product. In cyclopentanone and cyclohexanone enolates, an increase in the size of a facial blocking group increases selectivity. When that group was small, the selectivity decreased. The preference for the lowest energy conformation of the enolate anion is seen in larger ring systems as well (Sections 1.5.2 and 1.5.3), leading to good selectivity in alkylation and condensation reactions. The methyl group provides only small steric encumbrance to approach of the electrophile in enolate 367 [derived from the reaction of (S)-10-methyloxecan-2-one and LDA]. The preferred mode of attack for this relatively stable conformation was from the top face (path a, pseudoequatorial attack) and gave the syn-diastereomer, (3R,10S)-3,10-dimethyloxecan-2-one, with >99:1 selectivity.301 CH3 LDA

O H3C

H

H3CJI

a

O Li– O

O b

O

(S)-10-Methyloxecan-2-one

367

Me

O H3CJI H3C (3R,10S)-3, 10-Dimethyloxecan-2-one

13.6 ENAMINES Enamines are a class of compounds that have an amine group attached to a C]C unit. The reaction of a ketone with a primary amine generates an imine, as shown for the reaction of cyclohexanone with methanamine to yield N-methylcyclohexanimine. Alternatively, the reaction of acetone with a secondary amine generates an enamine, as shown for the reaction of cyclohexanone with dimethylamine (N-methylmethanamine) to give N,Ndimethylcyclohex-1-en-1-amine (see Section 4.2.2). Imines and enamines have a tautomeric relationship, but it is possible to prepare and isolate enamines. This section will discuss reactions of enamines, in the context of enolate anion reactions. O

CH3NH2 cat H+

Cyclohexanone

O

N-CH3 An imine N-Methylcyclohexanimine

(CH3)2NH cat H

+

Cyclohexanone

N(CH3)2 An enamine N, N-Dimethylcyclohex-1-en-1-amine

13.6.1 The Stork Enamine Synthesis Before the ability to control enolate anion reactions by using kinetic or thermodynamic reactions conditions was available, regioselectivity in the aldol condensation or enolate alkylation was an important problem. A method was developed that is effective in controlling the kinetic-thermodynamic product. In 1954, Stork et al.302 reported that cyclohexanone reacted with pyrrolidine to yield the corresponding enamine [1-(cyclohex-1-en-1-yl)pyrrolidine],303 a term introduced by Wittig and Blumenthal.304 An enamine is essentially a nitrogen enolate, and it can react with alkyl halides (e.g., iodomethane) to give an equilibrium mixture of the iminium salt (368) and the alkylated enamine (369).302 Hydrolysis with aqueous acid led to loss of pyrrolidine and formation of 2-methylcyclohexanone.

301

Still, W. C.; Galynker, I. Tetrahedron 1981, 37, 3981.

302

Stork, G.; Terrell, R.; Szmuszkovicz, J. J. Am. Chem. Soc. 1954, 76, 2029.

303

For an older review of the chemistry of enamines see Cook, A. G., Ed., Enamines: Synthesis Structure and Reactions; Marcel-Dekker: New York, 1969.

304

Wittig, G.; Blumenthal, H. Berichte 1927, 60, 1085.

726

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

O N

N H

I–

N

MeI

1-(Cyclohex-1-en-1yl)pyrrolidine

H3O+

Me

Me

Cyclohexanone

O

+ HI

N

368

Me

2-Methylcyclohexanone

369

A more general sequence shows the behavior of a generic enamine (370) reacting with a halide via the β-carbon to yield an alkyl iminium salt, via an SN2 process. The intermediate iminium salt (371) can isomerize under the reaction conditions to the less substituted enamine 372 in an equilibrium process, and hydrolysis yields the corresponding α-alkyl carbonyl compound. Alkylation gives the best yields with reactive primary halides, since it is essentially an SN2 reaction. This sequence is referred to as the Stork enamine synthesis.302,305 Enamines are usually formed by reaction of a secondary amine with a ketone, in the presence of an acid catalyst.306 R

N

R

R

R1

+

R

R

N

H

R

H

R1

R1

R1

370

O

H3O+

N

X

X 371

372

Enamines (e.g., 370) are bidentate nucleophiles, with both nitrogen and carbon functioning as nucleophiles.307 Alkylation at nitrogen is a problem with reactive halides (e.g., methyl iodide or allyl bromide), particularly when the groups on nitrogen are small. Nucleophilicity usually parallels that of the unsubstituted amine. In general, the less substituted enamine can be viewed as the kinetic nitrogen enolate, due to reduced steric hindrance.308 An enamine exists primarily as a planar species (373 or 374) in which the lone-pair electrons on nitrogen are parallel with the π bond. If the steric interaction (R $ R1) in 373 is greater than the R $ Me interaction in 374, there is a preference for that kinetic nitrogen enolate. When diethylamino or dimethylamino moieties are used, these interactions are small, and a mixture of 373 and 374 is obtained. As R and R1 increase in size, the steric hindrance increases, favoring 374. Pyrrolidine, piperidine, morpholine, or diethylamine are the most common amine partners used to form enamines. Note that when aldehydes (e.g., 375) react with secondary amines, initial reaction yields an iminium salt. Reaction with excess amine leads to aminal 376 rather than to the enamine. Treatment with a base (e.g., sodium carbonate) is required to induce elimination of the amine (pyrrolidine in this case), to yield enamine 377.309 H

Me H Me R

C N

C

R1 H

Me

H

C

C

Me

R1 R

N 374

373

N

R

R

R

CHO

R Na2CO3, Heat

H

N

-

H 375

R

N

N 376

N

H

H

377

305 (a) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-90. (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience, NJ, 2005, pp 628, 629. 306

(a) Hayne, L. W. in Ref. 300, pp 55–100. (b) Herr, M. E.; Heyl, F. W. J. Am. Chem. Soc. 1952, 74, 3627. (c) Heyl, F. W.; Herr, M. E. Ibid. 1953, 75, 1918. (d) Herr, M. E.; Heyl, F. W. Ibid. 1953, 75, 5927.

307

For a review of enamine chemistry, as well as that of imines and oximes, see Adams, J. P. J. Chem. Soc. Perkin Trans. 2000, 1, 125.

308

(a) Stork, G.; Brizzolara, A.; Landesman, H.; Szmuszkovicz, J.; Terrell, R. J. Am. Chem. Soc. 1963, 85, 207. (b) Kuehne, M. Ibid. 1959, 81, 5400. (c) Kuehne, M. Ibid. 1962, 84, 837.

309

Mannich, C.; Davidsen, H. Berichte 1936, 69, 2106.

727

13.6 ENAMINES

13.6.2 Reactions of Enamines As noted in Section 13.6.1, enamines behave as nitrogen enolates in their reactions with alkyl halides, generating an iminium salt (see 371) that can isomerize to the less substituted enamine (see 372). Although this new enamine can react with additional halide, it is present in very small quantities in most cases, and overalkylation is not a serious side reaction. If an excess of the halide is used, however, the second alkylation can be used to advantage. A typical enamine synthesis is taken from the work of Carpenter and Davis, in which ketone 378 reacted with pyrrolidine to yield 379.310 Reaction with ethyl bromoacetate, followed by hydrolysis, gave 380. Enamines can react with aryl halides in addition to alkyl halides, but the aryl halide must be activated by the presence of electron-withdrawing groups, since this coupling reaction is formally a nucleophilic aromatic substitution (Section 3.10.4).308c,311 As noted above, pyrrolidine enamines are usually too reactive at carbon for N-acylation to compete. O

O

O

1. BrCH2CO2Et

Pyrrolidine

O O

O

N

PhH, Reflux

O

2. H3O+

O 379

378

CO2Et

380

Enamines react with α-,β-unsaturated compounds almost exclusively by conjugate addition to yield the corresponding substituted ester, ketone, or nitrile. Reaction of 1-(cyclohex-1-en-1-yl)pyrrolidine with acrylonitrile, for example, gave a new enamine (381). Subsequent hydrolysis liberated the substituted ketone (382) in 80% overall yield from 1-(cyclohex-1-en-1-yl)pyrrolidine.308a,312 CN

H3O+

N

N Dioxane

O CN

CN

1-(Cyclohex-1-en-1-yl)pyrrolidine

381

382 (80%)

With the exception of pyrrolidine enamines, as noted above, enamines react very efficiently with acyl halides and the product after hydrolysis is a dicarbonyl compound. The morpholine enamine of isobutryaldehyde [4-(2-methylprop-1-en-1-yl)morpholine] reacted with acetyl chloride to yield keto-iminium salt 383. Hydrolysis gave ketoaldehyde 2,2-dimethyl-3-oxobutanal in 66% yield.313 Enamines react with acid chlorides in the presence of a base to yield a ketene, and this reacts via a thermal [2+2]-cycloaddition to give a cyclobutanone derivative (see Section 15.2.1). Ketene formation is usually not a significant problem in acylation reactions, since the enamine is the only base in the system.314 Morpholino enamines of cyclic ketones usually give the best yield of ketone product on reaction with acyl halides. A side reaction that is sometimes important is an aldol-like reaction between enamines and aldehydes, which are both in solution,315 but it is not a serious problem when the enamine is derived from a ketone. O

O

Me

Me

O

Me

Cl

N

N Me

Cl–

4-(2-Methylprop-1-en1-yl)morpholine

Me

Me Me

H3O+

O 383

Me

H O

Me

O

2,2-Dimethyl-3(66%) oxobutanal

310

Davis, K. M.; Carpenter, B. K. J. Org. Chem. 1996, 61, 4617.

311

Alt, G. H. in Ref. 303, p 131.

312

Williamson, W. R. N. Tetrahedron 1958, 3, 314.

313

(a) Alt, G. H. in Ref. 303, pp 135, 136. (b) Inukai, T.; Yoshizawa, R. J. Org. Chem. 1967, 32, 404.

314

H€ unig, S.; Benzing, E.; L€ ucke, E. Berichte 1957, 90, 2833.

315

(a) Alt, G. H. in Ref. 303, pp 156–165. (b) Birkofer, L.; Kim, S. M.; Engels, H. D. Berichte 1962, 95, 1495.

728

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

Prior to the development of kinetic control in enolate anion condensation reactions, enamines were used extensively in synthesis.316,317 An example is the synthesis of lupinine using the optically active chlorocarbonate 384 and octahydroquinolizine (1,3,4,6,7,8-hexahydro-2H-quinolizine).318 The enamine product of the initial reaction (385) was isolated in this case and selectively reduced to lupinine. Other reactions are possible with enamines. Cossy and Belotti319 reported that cyclohexene-enamines can be converted to aromatic amines. In their work, cyclohexanone and pyrrolidine gave 1-pyrrolidinecyclohexene [1-(cyclohex-1-en-1-yl)pyrrolidine] and this was converted to 1-phenylpyrrolidine in 83% yield by heating with Pd/C in the presence of nitrobenzene. O

O N

+ Cl

H

O

OH

N

O

N 1, 3, 4, 6, 7, 8-Hexahydro-2H-quinolizine

384

Lupinine

385

N

10% Pd–C, PhNO2, Toluene

N MS 4 Å, Reflux

1-(Cyclohex-1-en-1-yl)pyrrolidine

1-Phenylpyrrolidine (83%)

13.6.3 Asymmetric Enamine Syntheses Asymmetric enamine derivatives generally yield moderate-to-good diastereoselectivity.320 The synthesis of lupinine provided one example where an acid chloride substrate possessed a chiral auxiliary. An alternative approach is to prepare the enamine from a chiral amine. Yamada et al.321 prepared proline ester enamines (e.g., 386). In this case, the chiral enamine was used in a Michael addition with acrylonitrile or methyl acrylate to yield 387. The yield of the ketone product was rather poor, and the asymmetric induction (%ee) was in the range 15–59% with acrylonitrile and methyl acrylate.321 The chiral amine precursor to the enamine was synthesized from proline derivatives. Distillation of this chiral amine led to extensive racemization, which is a major drawback to a practical use of this procedure. In addition, some racemization occurred during isolation of the enamine product (386). Failure to resolve these problems led to the use of crude amines and enamine products in all subsequent reactions.322 It is possible to obtain good enantioselectivity with chiral enamines, if the chiral amine precursor is carefully chosen. Whitesell and Felman323 found that enamine 388 reacts with alkyl halides to give chiral ketones (e.g., 389) with good optical yields (+MeI ¼ 83 %ee; + n-PrI ¼ 93 %ee; +CH2]CHCH2Br ¼ 87 %ee).

316

Kuehne, M. E. in Reference 303, pp 313–468.

317

Lansbury, P. T.; Wang, N. Y.; Rhodes, J. E. Tetrahedron Lett. 1972, 2053.

318

Goldberg, S. I.; Ragade, I. J. Org. Chem. 1967, 32, 1046.

319

Cossy, J.; Belotti, D. Org. Lett. 2002, 4, 2557.

320

Seebach, D.; Imwinkelried, R.; Weber, T. In Modern Synthetic Methods 1986; Scheffold, R., Ed.; Springer-Verlag: Berlin, 1986; pp 217–246.

321

(a) Yamada, S.; Hiroi, K.; Achiwa, K. Tetrahedron Lett. 1969, 4233. (b) Yamada, S.; Otani, G. Ibid. 1969, 4237.

322

(a) Bergbreiter, D. E.; Newcomb, M. In Asymmetric Synthesis; Morrison, J. D. Ed.; Academic Press: New York, 1980; Vol. 2, p 220. (b) Hiroi, K.; Yamada, S. Chem. Pharm. Bull. 1973, 21, 47. (c) Otani, G.; Yamada, S. Ibid. 1973, 21, 2112.

323

Whitesell, J. K.; Felman, S. W. J. Org. Chem. 1977, 42, 1663.

729

13.7 MICHAEL ADDITION AND RELATED REACTIONS

N

X

1.

O

+

H CO R 2

2. H3O

H X

386

387 Me O

1. R—X

N

+

2. H3 O

H

R

Me 388

389

The disconnections for enamine reactions are the same as those shown for enolates.

13.7 MICHAEL ADDITION AND RELATED REACTIONS 13.7.1 Michael Addition The 1,4- (conjugate) addition324e of a carbon nucleophile to an α-,β-unsaturated carbonyl system is usually reversible, and referred to as Michael addition.324 The term Michael addition initially applied only to the conjugate addition of amines to unsaturated carbonyl compounds. The addition of dimethylamine to 2-acetamido-N,N-diethylacrylamide in water,325 for example, generated 2-acetamido-3-(dimethylamino)-N,N-diethylpropanamide, and illustrates the Michael addition of an amine. The rate acceleration effect observed when the reaction is done in water makes this particular reaction less common. O

O H N

H N

Me2NH, H2O, rt

NEt2

36 h

O

O 2-Acetamido-N, Ndiethylacrylamide

H3O+

Ph Ph

O

O

CO2Et

O

NMe2

2-Acetamido-3-(dimethylamino)N, N-diethylpropanamide

O O

DBU

1-Phenylbutane-1,3-dione

NEt2

O

Ph O

O 390

3-Benzoylheptane-2,6-dione (90%)

Conjugate addition reactions with other nucleophiles is common, and all are categorized under the term Michael addition. Indeed, enolates and carbanions are common partners in Michael additions.326 The anion of 1-phenylbutane-1,3-dione generated in situ by reaction of the diketone with DBU, reacted with ethyl acrylate to yield the Michael addition product, enolate anion 390. Quenching the enolate anion with aqueous acid generated diketoester 3-benzoylheptane-2,6-dione in 90% yield, in Taber and Teng’s327 synthesis of the ethyl ester of the major urinary metabolite of prostaglandin E2. The conjugate addition of enolate anions is often reversible, and the relative stability of 324 (a) Michael, A. J. Prakt. Chem. 1887, 35, 379. (b) Bergmann, E. D.; Gingberg, D.; Pappo, R. Org. React. 1959, 10, 179. (c) Reference 100, pp 315, 316. (d) Reference 101, pp 173, 174. (e) Perlmutter, P. Conjugative Addition Reactions in Organic Synthesis; Pergamon Press: Oxford, 1992. (f ) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-61. (g) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 428, 429. 325

Naidu, B. N.; Sorenson, M. E.; Connolly, T. P.; Ueda, Y. J. Org. Chem. 2003, 68, 10098.

326

For reactions of this type, see Larock, R. C. Comprehensive Organic Transformations, 2nd ed.; Wiley-VCH: New York, 1999; p 1317.

327

Taber, D. F.; Teng, D. J. Org. Chem. 2002, 67, 1607.

730

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

the enolate starting material (1-phenylbutane-1,3-dione) versus the enolate product (390) will determine the position of the equilibrium. Solvent is important, and alcoholic solvents usually promote the equilibrium. Enolate anions add to conjugated systems in a Michael fashion. Enantioselective Michael additions are well known,328a and a lanthanide-linked BINOL complex as been used effectively for the addition of malonate anions to cyclohexenone.328b One synthetic example is taken from the Takasu et al.329 synthesis of culmorin, in which treatment of 391 with LHDMS led to a double Michael addition to yield 393 in 87% yield. Initial formation of enolate anion 392 allows a Michael addition to the conjugated ester to form a seven-membered ring in 393. This newly formed ester enolate reacts with the conjugated ketone by a second Michael addition to yield enolate anion 394, which is drawn a second time to show the proper relationship of the functional groups. Hydrolysis gave the final product 395. O

O

CO2Me

O

LHMDS, THF –78°C

CO2Me

CO2Me 391

392 O

393 (87%)

CO2Me O

Hydrolysis

O

CO2Me

CO2Me 395

394

13.7.2 Baylis-Hillman Reaction An interesting variation of the Michael addition has been used to prepare highly functionalized acylate derivatives. In its fundamental form, an acrylate ester reacts with an aldehyde in the presence of an amine or phosphine (e.g., tributylphosphine) catalyst. Presumably, Michael addition generates an enolate anion (e.g., 396), which condenses with the aldehyde and then loses the phosphine or amine to yield the final product 397. CO2Me

OH

CO2Me R3P

RCHO

PR3

CO2Me

R

– PR3

396

397

CHO

O

PPh3, 1, 1'-Naphthol THF, rt, 96 h

O (R)-5-Methylcyclohex-2-en-1-one

OH 398 (83%)

This transformation is known as the Baylis-Hillman reaction330 (also known as the Morita-Baylis-Hillman reaction).331 Aldol condensation, induced by the amine catalyst, can be a competing side reaction. Coupling is often slow, but Leahy 328

(a) Sibi, M. P.; Manyem, S. Tetrahedron 2000, 56, 8033. (b) Kim, Y. S.; Matsunaga, S.; Das, J.; Sekine, A.; Ohshima, T.; Shibasaki, M. J. Am. Chem. Soc. 2000, 122, 6506.

329

Takasu, K.; Mizutani, S.; Noguchi, M.; Makita, K.; Ihara, M. J. Org. Chem. 2000, 65, 4112.

330

(a) Ciganek, E. Org. React. 1997, 51, 201. (b) Basavaiah, D.; Rao, P. D.; Hyma, R. S. Tetrahedron 1996, 52, 8001. (c) Drewes, S. E.; Roos, G. H. P. Tetrahedron 1988, 44, 4653. (d) Shi, M.; Li, C.-Q.; Jiang, J.-K. Tetrahedron 2003, 59, 1181. (e) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-7. (f ) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 74, 75.

331

(a) Morita, K.; Suzuki, Z.; Hirose, H. Bull. Chem. Soc. Jpn. 1968, 41, 2815. (b) Baylis, A. B.; Hillman, M E. D. German Patent 2155113, 1972 (Chem. Abstr. 1972, 77, 34174q). For a recent review, see (c) Basavaiah, D.; Rao, A. J.: Satyanarayana, T. Chem. Rev. 2003, 103, 811.

731

13.7 MICHAEL ADDITION AND RELATED REACTIONS

and Rafel332 found that using DABCO as the catalyst at 0°C led to faster formation of Baylis-Hillman products, and in good yield. This coupling reaction can be applied to conjugated ketones or aldehydes, as well as conjugated esters. An example is the reaction of (5R)-methylcyclohex-2-en-1-one with pent-4-enal in the presence of triphenylphosphine, giving an 83% yield of 398 in the Yao and coworker’s333 synthesis of ()-lannotinidine B. Chloromethyl derivatives formed from TiCl4 and triethylamine in CH2Cl2 at 78°C are treated with triethylamine or DBU to yield a BaylisHillman product.334 Samarium iodide (SmI2) mediates the Baylis-Hillman reaction.335 Asymmetric Baylis-Hillman reactions are known.336 Using Oppolzer’s sultam337 as an auxiliary, Leahy and coworkers338 reacted amide 399 with propanal to give an 85% yield of 400 in >99 %ee.337 Treatment of 400 with camphorsulfonic acid (CSA) in methanol, in a second step, gave an 85% yield of (+)-methyl (R)-3hydroxy-2-methylenepentanoate. Another camphor-based auxiliary has also been used with success.339 A chiral acetylenic ester titanium alkoxide complex has been used to generate chiral Baylis-Hillman products,340 and the use of lanthanide triflates complexed with chiral diamine ligands derived from camphor gave Baylis-Hillman products with high enantioselectivity.341 Intramolecular Baylis-Hillman reactions have also been reported.342 O

O EtCHO, DABCO

N S

CSA, MeOH

O

CH2Cl2, 0°C

O

Et

O 399

Et

MeO2C O

OH

Et

(+)-Methyl (R)-3-hydroxy-2(85%) methylenepentanoate

400 (85%)

13.7.3 Robinson Annulation A major synthetic challenge in the 1930s through the 1950s was construction of various steroid molecules. One important route that emerged from this effort used a Michael addition protocol to produce a bicyclic ketone. This sequence soon came to be known as the Robinson annulation.324d,343 The reaction can be illustrated by treatment of pentan-3-one with sodium ethoxide (equilibrating conditions) in the presence of MVK. O

O

O

NaOEt, EtOH

+ 5

4

2 3

1

5

O

+

H3O

6

O – HO 2

2

6 6

OH

1

4

5

1

O

2

4

3

3

401

3-Ethyl-3-hydroxy-4-methylcyclohexan-1-one

3-Ethyl-4-methylcyclohex-2-en-1-one

332

Rafel, S.; Leahy, J. W. J. Org. Chem. 1997, 62, 1521. For the use of ethyl acrylate in a synthesis of (+)-tubelactomicin A, see Motozaki, T.; Sawamura, K.; Suzuki, A.; Yoshida, K.; Ueki, T.; Ohara, A.; Munakata, R.; Takao, K.-I.; Tadano, K.-I. Org. Lett. 2005, 7, 2265. 333

Ge, H. M.; Zhang, L.-D.; R. X.; Yao, Z.-J. J. Am. Chem. Soc. 2012, 134, 12323.

334

Shi, M.; Jiang, J.-K.; Feng, Y.-S. Org. Lett. 2000, 2, 2397.

335

Youn, S. W.; Park, H. S.; Kim, Y. H. Chem. Commun. 2000, 2005.

336

Langer, P. Angew. Chem. Int. Ed. 2000, 39, 3049.

337

(a) Oppolzer, W. Pure Appl. Chem. 1990, 62, 1241. (b) Kim, B. H.; Curran, D. P. Tetrahedron 1993, 49, 293. (c) Oppolzer, W. Tetrahedron 1987, 43, 1969. (d) Oppolzer, W. Pure Appl. Chem. 1988, 60, 39.

338

Brezinski, L. J.; Rafel, S.; Leahy, J. W. J. Am. Chem. Soc. 1997, 119, 4317.

339

Yang, K.-S.; Chen, K. Org. Lett. 2000, 2, 729.

340

Suzuki, D.; Urabe, H.; Sato, F. Angew. Chem. Int. Ed. 2000, 39, 3290.

341

Yang, K.-S.; Lee, W.-D.; Pan, J.-F.; Chen, K. J. Org. Chem. 2003, 68, 915.

342

Keck, G. E.; Welch, D. S. Org. Lett. 2002, 4, 3687. Also see Krishna, P. R.; Kannan, V.; Sharma, G. V. M. J. Org. Chem. 2004, 69, 6467.

343

(a) Annulation is the act or process of forming a ring; the formation of rings or ring-shaped parts. Annelation is an alternative, but less desirable term, which means to anneal one ring onto another. The term annelation refers to the phenomenon whereby some rings in fused systems give up part of their aromaticity to adjacent rings, as on p 57 in Smith, M. B. March’s Advanced Organic Chemistry, 7th ed.; Wiley: Hoboken, NJ, 2013. (b) DuFeu, E. C.; McQuillin, F. J.; Robinson, R. J. Chem. Soc. 1937, 53. (c) Balasubramanian, K.; John, J. P.; Swamnathan, S. Synthesis 1974, 51. (d) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-80. (e) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 556, 557.

732

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

Other bases have been used, including lanthanoid triisopropoxides that behave as catalysts in this reaction.344 The initial product is a cyclic alkoxy ketone (401), which gave 3-ethyl-3-hydroxy-4-methylcyclohexan-1-one upon hydrolysis. This alcohol usually loses water with heating to give the enone, 3-ethyl-4-methylcyclohex-2-en-1-one, as the final Robinson annulation product. The reaction conditions shown for this reaction are classified as thermodynamic control conditions, which favor formation of the more highly substituted enolate anion. In this case, pentan-3-one is a symmetrical ketone, but the thermodynamic conditions are important for this reaction. Conversion of pentan-3-one to its enolate anion (pent-2en-3-olate) was followed by Michael addition to MVK to yield a new enolate product (402). This enolate anion could react with the carbonyl at C6 to give a four-membered ring (path a), but as discussed in Section 4.5.1, the energy requirements for this reaction are relatively high and the process is energetically unfavorable. Under the equilibrium conditions of the reaction, a small portion of kinetic enolate 403 can be formed. An intramolecular aldol condensation at C6 is possible (path b) that generates a six-membered ring (401), and this is energetically more favorable than forming a four-membered ring from 402. When a cyclic ketone (e.g., cyclohexanone) reacts with MVK, Robinson annulation generates a bicyclic ketone.343b–e 1

O

O 6

–

5

O

4

3

2

1

4

5

O

O

Pent-2-en-3-olate

1

6

5 5

O–

b

a

6

EtOH

–O

2 3

O–

OEt

–

O 2

4

3

6

402

403

401

A synthetic example is taken from a synthesis of ()-anominine by Bradshaw, Bonjoch, et al.345 in which cyclohexanedione 404 reacted with MVK and triethylamine to give 405 in 91% yield. Note the syn of the chiral catalyst, which led to formation of 405 in 94 %ee. There are many variations of the Robinson annulation that can be used to prepare synthetically useful products,346,347 including some using a conjugated ketone surrogate known as a Mannich base, which is a β-amino ketone. O O

O , NEt3, 2.5% PhCO2H

O 1%

404

NHTs H O N

O NH

405 (91%, 94 %ee)

Under the basic conditions of the Robinson annulation, the β-amino ketone eliminates diethylamine to form the conjugated ketone in situ, which may avoid the deleterious side reactions caused by the equilibrium conditions. A Mannich base is produced by a Mannich reaction,348 where dimethylamine, for example, reacts with formaldehyde and acetone to yield 4-(dimethylamino)butan-2-one.348a,349 Acid-catalyzed Robinson annulation reactions are also known. An example is the conversion of 406 to 407 in good yield.350

344

Okano, T.; Satou, Y.; Tamura, M.; Kiji, J. Bull. Chem. Soc. Jpn. 1997, 70, 1879.

345

Bradshaw, B.; Etxebarria-Jardí, G.; Bonjoch, J. J. Am. Chem. Soc. 2010, 132, 5966.

346

Scott, W. L.; Evans, D. A. J. Am. Chem. Soc. 1972, 94, 4779.

347

Hackett, S.; Livinghouse, T. J. Org. Chem. 1986, 51, 1629.

348

(a) Mannich, C.; Kr€ osche, W. Arch. Pharm. 1912, 250, 647. (b) Blicke, F. F. Org. React. 1942, 1, 303. (c) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-57. (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; WileyInterscience: Hoboken, NJ, 2005; pp 408, 409.

349

For a synthetic example using a Mannich reaction for saframycin alkaloids, see Zhou, B.; Guo, J.; Danishefsky, S. J. Tetrahedron Lett. 2000, 41, 2043.

350

Busch-Petersen, J.; Corey, E. J. Tetrahedron Lett. 2000, 41, 6941.

733

13.7 MICHAEL ADDITION AND RELATED REACTIONS

Me

Me

HCHO

O Acetone t-Bu

Me2N

Me2NH

O 4-(Dimethylamino)butan-2-one O

O

p-TsOH PhH, 80°C

MeO2C

Me

t-Bu CO2Me

O 406

407

The Robinson annulation disconnection follows: O

R

O + O

R

The reversible nature of the initial Michael addition is a problem with the Robinson annulation. A solution is to use a conjugated system that is particularly prone to Michael addition and forms the product, essentially irreversibly. α-Silyl vinyl ketones have been shown to be powerful Michael acceptors.351 The lithium enolate of cyclohexanone (lithium cyclohex-1-en-1-olate) reacted with conjugated ketone 3-(triethylsilyl)but-3-en-2-one to produce the Michael product, 408.351a In this case, the initially formed Michael adduct was stabilized by the presence of the silyl group at the α-position, driving the reaction toward the product. Hydrolysis produced 2-(3-oxobutyl)cyclohexan-1-one, which was converted to the Robinson product [4,4a,5,6,7,8-hexahydronaphthalen-2(3H)-one] in 80% overall yield by treatment with NaOMe/MeOH under the requisite thermodynamic conditions.351a When compared to this sequential process, normal treatment of cyclohexanone enolate with MVK under Robinson conditions, which gave 4,4a,5,6,7,8-hexahydronaphthalen-2(3H)-one in < 5% yield. O O–

Li+

SiEt3

O

+

O

SiEt3

–78°C

1. NaOMe, MeOH

H 2O

THF

2. H3O+

rt

O Lithium cyclo- 3-(Triethylsilyl)buthex-1-en-1-olate 3-en-2-one

O

O 408

2-(3-Oxobutyl)cyclohexan-1-one

4, 4a, 5, 6, 7, 8-Hexahydro- (80% overall) naphthalen-2(3H)-one

13.7.4 Selectivity in the Robinson Annulation Many modifications of the fundamental Robinson annulation sequence have appeared over the years, and one uses silyl derivatives to generate a silyl enol ether. An important modification uses chiral precursors or chiral additives to yield asymmetric induction either during the Michael addition, during the aldol cyclization or in both reactions. Excellent asymmetric induction has been obtained by using an asymmetric base [e.g., (S)-()-proline shown in the reaction] to generate the enolate. Hajos and Parrish352 obtained bicyclic ketone 409 (Hajos diketone) by treating 2-methylcyclopentane-1,3-dione with MVK and proline. The Robinson product (409) was isolated with an optical purity of 93.4%. Proline acts as a catalyst in an asymmetric Robinson annulation.353 Similarly, Wynberg, and Helder showed that addition of asymmetric catalysts (e.g., Cinchona alkaloids) gave optical yields of 5–25 %ee for Michael additions of cyclohexanone derivatives,354 but up to 71 %ee in some cases.355 Langstrom and Bergson356 first employed 351

(a) Stork, G.; Ganem, B. J. Am. Chem. Soc. 1973, 95, 6152. (b) Stork, G.; Singh, J. Ibid. 1974, 96, 6181.

352

Hajos, Z. G.; Parrish, D. R. J. Org. Chem. 1974, 39, 1612, 1615.

353

Bui, T.; Barbas, III, C. F. Tetrahedron Lett. 2000, 41, 6951.

354

Hermann, K.; Wynberg, H. J. Org. Chem. 1979, 44, 2238.

355

Wynberg, H.; Helder, R. Tetrahedron Lett. 1975, 4057.

356

Langstrom, B.; Bergson, G. Acta Chem. Scand. 1973, 27, 3118.

734

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

asymmetric catalysts in the Michael addition. An antibody-catalyzed enantioselective Robinson annulation has also been reported.357 H

O

Me

N

,

1.

O

O

H

H

Me

CO2–

O

2. H3O+

O 2-Methylcyclopentane-1, 3-dione

409

Solvent influences the stereochemistry of the Robinson annulation.358 The condensation of methylcyclohexanone and (E)-pent-3-en-2-one with sodium hydride in dioxane led to >95% of 410 with a syn relationship for the methyl groups. Similar reaction in DMSO, however, led to >95% of 411 with an anti-relationship for those methyl groups. The change in stereochemistry was explained by a hydrogen transfer, which is facile in the highly polar DMSO, but not in dioxane.353 In general, polar protic solvents tend to give the trans-product, whereas less polar aprotic solvents favor the cis-product. Me

Me

Me

Me

1. NaH, Dioxane, rt 100 h

Me 1. NaH, DMSO, rt, 3 h

+

2. H3O+

+

2. H3O

O

O

Me

2-Methylcyclohexanone

410 (>95%)

Me

O

O

(E)-pent-3-en-2-one

411 (>95%)

The metal counterion also plays a role. The presence of a highly covalent OdLi bond favors the trans-product, but a more ionic OdNa or OdK bond in alcoholic solvents favors the cis-bond.358 The Li species is less bulky, due to poorer solvation in the aprotic solvent. A protic solvent will hydrogen bond with the ionic sodium species, leading to a large increase in the relative “size” of the alkoxide moiety. This bulky group influences the geometry of the intermediate and leads to an increase in the cis-product.358

13.8 ENOLATE REACTIONS OF α-HALO CARBONYL DERIVATIVES 13.8.1 α-Halogenation Several reactions formally involve enolates, but the course of the reaction is altered by the presence of reactive functionality. The Darzens’ glycidic ester synthesis discussed in Section 13.4.2.6 is one example. Another involves the formation of α-halogenated ketones. Treatment of a ketone with X2 or NXS (X ¼ Cl, Br, I) usually yields the α-halo ketone, as in the conversion of cyclopentanone to 2-bromocyclopentanone.359 When there is the possibility of forming kinetic and thermodynamic products, the thermodynamic product usually predominates. In the synthesis of this α-bromoketone, KClO3 oxidized the HBr byproduct back to Br2. A serious problem in the halogenation is further reaction of the product with the halogen to yield an α,α-dihalo compound. The presence of the first halogen promotes further enolization and polyhalogenation is often a serious problem. O

O Br2, AcOH KClO3

Cyclopentanone

O 1. PhMgBr

Br

2. H3O+

2-Bromocyclopentanone

Ph

2-Phenylcyclopentanone

357

Zhong, G.; Hoffmann, T.; Lerner, R. A.; Danishefsky, S.; Barbas, III, C. F. J. Am. Chem. Soc. 1997, 119, 8131.

358

(a) Marshall, J. A.; Warne, T. M., Jr. J. Org. Chem. 1971, 36, 178. (b) Scanio, C. J. V.; Starrett, R. M. J. Am. Chem. Soc. 1971, 93, 1539.

359

(a) Reed, S. F., Jr. J. Org. Chem. 1965, 30, 2195. (b) Rappe, C.; Kumar, R. Arkiv. Kemi. 1965, 23, 475 (Chem. Abstr. 1965, 63, 5521c). (c) Zbiral, E.; Rasberger, M. Tetrahedron 1969, 25, 1871. (d) Wyman, D. P.; Kaufman, P. R. J. Org. Chem. 1964, 29, 1956.

735

13.8 ENOLATE REACTIONS OF α-HALO CARBONYL DERIVATIVES

Halogenation of silyl enol ethers gives a good yield of monohaloketone.360 Cupric (Cu2+) salts (e.g., CuCl2 and CuBr2) are often added to convert a ketone to α-chloro or α-bromo ketones.361 The presence of the halogen modifies the bond polarity of the molecule so that displacement of halogen by nucleophiles is faster than acyl addition to the carbonyl. The reaction of 2-bromocyclopentanone with phenylmagnesium bromide gave 2-phenylcyclopentanone,362 for example. This variation of the Grignard reaction is an excellent route to α-aryl ketones.362 This leads to the following interesting disconnection: O

O

+

Ar (R)

Ar

where Ar = Aryl

X R X

Polyhalogenation can be a severe side reaction in the formation of α-halo ketones, but can be used to synthetic advantage in the haloform reaction (this has been called the Lieben iodoform reaction).363 Reaction of butan-2-one with sodium hydroxide gave the enolate anion, and in the presence of bromine the enolate displaced bromide from BrdBr to form 1-bromobutan-2-one. Incorporation of the bromine made the α-hydrogen more acidic, and a second bromination occurred to yield 1,1-dibromobutan-2-one, followed by a third bromination to yield 1,1,1-tribromobutan-2-one. Nucleophilic acyl addition of hydroxide to all carbonyl derivatives (e.g., butan-2-one, 1-bromobutan-2-one, and 1,1dibromobutan-2-one) was reversible up to this point. The Br3C group is a good leaving group, however, so nucleophilic attack at the carbonyl leads to cleavage of the O]CdCBr3 bond (via 412) to produce butanoic acid and Br3C. This carbanion rapidly reacted with butanoic acid, and deprotonation produced the two final products, butanoic acid and a haloform (bromoform, CHBr3). O

O–

NaOH

O

O

Br2

NaOH

Br

Br

NaOH Br2

Br2

Br 1,1-Dibromobutan-2-one

1-Bromobutan-2-one O–

O Br Br

NaOH

Br HO

Br 1,1,1-Tribromobutan-2-one

Br

Br

412

CO2–

CO2H

+

+

–CBr

H—CBr3

3

This sequence is a synthetic method for oxidative cleavage of a methyl group attached to the carbonyl, complementing the known cleavage of carbonyls by Cr(VI) (Section 6.7.1.4). This reaction also cleaves the H3CdC bond of methyl carbinols [RCH(OH)Me] to yield the corresponding acid and iodoform. When iodine is used rather than bromine, the cleavage of methyl ketones or methyl carbinols is referred to as the iodoform reaction. The iodoform reaction constitutes a classical test for the presence of a methyl ketone or a methyl carbinol moiety in an unknown molecule. The haloform disconnection follows: O R CO2H

+

X3C H R

CH3

360

(a) Reuss, R. H.; Hassner, A. J. Org. Chem. 1974, 39, 1785. (b) Blanco, L.; Amice, P.; Conia, J. M. Synthesis 1976, 194, 196.

361

Smith, M. B. March’s Advanced Organic Chemistry, 7th ed.; Wiley: Hoboken, NJ, 2013; pp 672–678.

362

Ando, T. Yuki Gosei Kagaku Kyokaishi 1959, 17, 777 (Chem. Abstr. 1960, 54, 4492b).

363

(a) Fuson, R. C.; Bull, B. A. Chem. Rev. 1934, 15, 275. (b) Seelye, R. N.; Turney, T. A.; J. Chem. Educ. 1959, 36, 572. (c) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-56. (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 392, 393.

736

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

13.8.2 The Reformatsky Reaction Another enolate-like reaction is the Reformatsky reaction,364 which employs a nucleophilic organozinc intermediate, generated from an α-halo carbonyl and Zn metal. This condensation reaction is widely used in synthesis.365 The organozinc reagent (413), in this case derived from ethyl 2-bromopropanoate, attacks the carbonyl of an aldehyde or ketone (e.g., acetophenone) to yield 414. Hydrolysis yields the condensation product, β-hydroxyester, ethyl 3-hydroxy-2-methyl-3-phenylbutanoate. Preparation of the organozinc complex in the Reformatsky reaction can be a problem, and it often requires special preparation of the Zn (activated Zn). Activated Zn has been prepared by various procedures.366 The use of ultrasound techniques produces a finely dispersed zinc that also facilitates the Reformatsky reaction.367 As noted above, hydrolysis gives the hydroxy ester to complete this two-carbon chain extension process. Br

Ph

EtOH

CO2Et

O

ZnBr

Zn

Ph OZnBr

Me Me

CO2Et

Ethyl 2-bromopropanoate

Me

H3O+

CO2Et

413

Ph OH CO2Et

Ethyl 3-hydroxy-2-methyl3-phenylbutanoate

414

The organozinc reagent is generally less reactive than a Grignard or organolithium reagent, and condensation reactions proceed well with aldehydes and ketones, but are sluggish with esters. A synthetic example is taken from a synthesis of rugulovasine A by Jia and coworkers,368 in which ketone 415 was treated with Zn and an allylic bromide to yield hydroxy-ester 416, which cyclized to lactone 417 in 90% yield. A SmI3 reaction has been reported that gives the Reformatsky product with high diastereoselectivity in aqueous media.369 O

CO2Me OH NMeBoc

NMeBoc

O

CO2Me

NMeBoc Br

Zn, I2, THF, 50°C

N

N

N

Boc

Boc

Boc 415

416

417 (90%) SiMe3

SiMe3 H

H

Ph

Ph THF

Bz

(75%)

O

N

SmI2, CF3CO2H

O

Cl

CHO

H

H

Zn (Rieke Zn)

(46%)

418

O

N Bz

O 419

OH

364

(a) Reformatsky, S. Berichte 1887, 20, 1210. (b) Diaper, D. G. M.; Kuksis A. Chem. Rev. 1959, 59, 89. (c) Rathke, M. W. Org. React. 1975, 22, 423. (d) Reference 100, pp 370, 371. (e) Reference 101, pp 200–202. (f ) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-78. (g) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 544, 545.

365

Ocampo, R.; Dolbier, W. R., Jr. Tetrahedron 2004, 60, 9325.

366

(a) Rieke, R. D.; Li, P. T.-J.; Burns, T. P.; Uhm, S. T. J. Org. Chem. 1981, 46, 4323. (b) Rieke, R. D.; Uhm, S. J. Synthesis 1975, 452.

367

Han, B. H.; Boudjouk, P. J. Org. Chem. 1982, 47, 5030.

368

Zhang, Y.-A.; Liu, Q.; Wang, C.; Jia, Y. Org. Lett. 2013, 15, 3662.

369

Hayakawa, R.; Shimizu, M. Chem. Lett. 1999, 591.

737

13.8 ENOLATE REACTIONS OF α-HALO CARBONYL DERIVATIVES

Vedejs and Ahmad370 used an intramolecular version of this reaction aimed at the cytochalasin ring system. In this reaction, 418, was cyclized to yield 419 as a 1:1 mixture of epimeric alcohols. The use of highly reactive Rieke zinc371 led to a 75% yield of 419. When samarium iodide (SmI2) in trifluoroacetic acid was used for the cyclization,372 a 46% yield of 419 was obtained as one diastereomeric alcohol. The Reformatsky disconnection follows: R1

O

R3

+

R2 OH

R1

CO2R

R3 CO2R

R2

Heathcock and coworkers,373,374 showed that the Reformatsky reaction exhibits mixed selectivity. The syn-anti selectivity for the hydroxy ester products (anti-product 420 and syn-product 421) differs in reactions with α-haloesters with ketones and aldehydes.374 Aldehydes generally show poorer selectivity than do ketones for formation of the antiproduct (420). The bromozinc aldolate products from ketones were shown to equilibrate under reaction conditions, but those from aldehydes did not. In general, increasing the size of R2 in the α-halo-carbonyl derivative led to greater antiselectivity. R2

O Ph

R1

+

R1 OH

CO2Me Zn, Benzene Br

Reflux

R1 OH CO2Me

Ph

+

CO2Me

Ph

R2 420

R2 421

In more recent work, In catalysts have been used,375 Ge catalysts (e.g., GeCl4) gave a highly diastereoselective Reformatsky reaction favoring the syn-diastereomer.376 Another syn selective reaction was reported using TiCl2 and Cu.377 Enantioselective Reformatsky reactions have also used chiral amino alcohols as additives.378

13.8.3 The Favorskii Rearrangement There is a classical reaction of enolate anions derived from α-halo ketones that involves a rearrangement. In the Favorskii rearrangement,379 also called the Wallach degradation, an α-chloroketone (e.g., 2-chlorocyclohexanone) is treated with NaOH under thermodynamic conditions to form the enolate anion 422. Intramolecular displacement of chlorine leads to 423, which is attacked by hydroxide to yield the acyl addition product 424. Cleavage of the adjacent CdC bond leads to enolate 425, which is protonated by the adjacent carboxylic acid to yield carboxylate 426. Hydrolysis gives the final product, cyclopentanecarboxylic acid. This reaction can be useful synthetically, as in Ley and

370

Vedejs, E.; Ahmad, S. Tetrahedron Lett. 1988, 29, 2291.

371

Arnold, R. T.; Kulenovic, S. T. Synth. Commun. 1977, 7, 223.

372

Tabuchi, T.; Kawamura, K.; Inanaga, J.; Yamaguchi, M. Tetrahedron Lett. 1986, 27, 3889.

373

Reference 243, pp 144–152.

374

(a) Reference 243, p 146. (b) Canceill, J.; Basselier, J. J.; Jacques, J. Bull. Soc. Chim. Fr. 1967, 1024. (c) Canceill, J.; Jacques, J. Ibid. 1970, 2180.

375

See (a) Ishihara, J.; Tsuru, H.; Hatakeyama, S. J. Org. Chem. 2014, 79, 5908. (b) Yahata, K.; Minami, M.; Watanabe, K.; Fujioka, H. Org. Lett. 2014, 16, 3680.

376

Kagoshima, H.; Hashimoto, Y.; Oguro, D.; Saigo, K. J. Org. Chem. 1998, 63, 691.

377

Mukaiyama, T.; Kagayama, A.; Igarashi, K.; Shiina, I. Chem. Lett. 1999, 1157.

378

(a) Andres, J. M.; Martín, Y.; Pedrosa, R.; Perez-Encabo, A. Tetrahedron 1997, 53, 3787. (b) Mi, A.; Wang, Z.; Zhang, J.; Jiang, Y. Synth. Commun. 1997, 27, 1469.

379

(a) Favorskii, A. E. J. Prakt. Chem. 1913, 88, 658. (b) Wallach, O. Annalen 1918, 414, 296. (c) Kende, A. S. Org. React. 1960, 11, 261. (d) Turro, N. J. Acc. Chem. Res. 1969, 2, 25. (e) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-30. (f ) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 238, 239.

738

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

coworker’s380 synthesis of trilobolide. Treatment of 427 with sodium methoxide gave a 95% yield of the Favorskii rearrangement product (428), obtained as the methyl ester since the reaction was done in methanol. O

O

O Cl

Cl NaOH

2-Chlorocyclohexanone

–

422

O

HO

O–

–

O

O

H3O+

OH

423

H

THPO

HO

– Cl–

424

425

O

Cyclopentane carboxylic acid

426

Me

HO

Me H NaOMe, MeOH, 0°C

THPO Me Cl

H

Me

O 427

CO2Me

428 (95%)

13.9 CONCLUSION In conclusion, the reaction of enolate anions with alkyl halides and carbonyl compounds provides one of the most powerful and useful of all carbon-carbon bond-forming processes. Not only are a variety of synthetic products available, but the reaction can be made to proceed with excellent diastereoselectivity. In conjunction with the reactions of carbanions, nucleophilic methods constitute the largest single type of methodology for making new carboncarbon bonds. HOMEWORK

1. Predict the major product of this reaction, and explain the stereochemistry. H N

O

H

O

1. LDA, HMPA 2. MeI

H

2. Explain formation of the product shown given the reaction conditions. PhO2S N

N SO2Ph

Br K2CO3, Toluene

O

Bu4NBr, Reflux

MeO

MeO

OMe CHO

O

OMe 92%

3. Give a mechanistic rationale for the following reaction: O 1. t-BuOK

CHO

H Me

H

2. H+ Ion exchange

H

380

OH Me

O

(a) Oliver, S. F.; H€ ogenauer, K.; Simic, O.; Antonello, A.; Smith, M. D.; Ley, S. V. Angew. Chem. 2003, 42, 5996. (b) Also see Lee, E.; Yoon, C. H. J. Chem. Soc. Chem. Commun. 1994, 479.

739

13.9 CONCLUSION

4. In the following sequence, pulegone is transformed into 2,3-dimethylcyclohexanone. Explain this transformation. O

O

1. LDA, LiCl MeI 2. KOH, Reflux

5. For the reactions shown, draw the syn- an anti-products. The anti-product predominates with PhCHO, but there is essentially a 1:1 mixture with cyclohexanone. Explain why there is a difference in selectivity. Me

PhCHO

O

Me

>98:2 anti/syn 48:52 anti/syn

O

6. For the reaction of 3-pentanone and lithium diisopropylamide (78°C, THF): (a) Draw both (E)- and (Z)-enolate anions. (b) Give the major product of the (Z)-enolate when PhCHO approaches from the re face; from the si face. Show the Zimmerman-Traxler model for both approaches. (c) Give the major product when the (E)-enolate reacts with A, assuming A is chiral and nonracemic. H CHO A

(d) Give the major product when the (Z)-enolate anion reacts with chiral, nonracemic A. 7. Predict the major product (A or B) for each reaction. Discuss formation of each product and why each reaction might follow a different pathway. MeO H MeO

O

N

MeO

NaH, THF Reflux

H A

H

MeO

SMe

I

N

MeO2C

CO2Me

O

NaH, THF Reflux

B

SMe Br

8. Enolate formation is an acid-base equilibrium process. For reactions A and B, discuss the relative acidity of all hydrogen atoms in 3-methylbutan-2-one. Discuss the role of solvent, base, temperature, and reaction time for reactions A and B. O

LiN(i-Pr)2, THF

B

NaOEt, EtOH, 20 h

Me

–78°C, 30 min

A

Reflux

Write the acid-base equations for both A and B using the standard reaction shown. Ka

Acid + Base
!
Conjugate acid + Conjugate base 9. Explain the following transformation: O

NH2 MeO

1.

Cl

2. H3O+

O

Cl

Cl

N O

740

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

10. In each of the following reactions, predict the major diastereomer and show the Zimmerman-Traxler transition state: 1. LDA, THF, –78°C

(A)

Ph

CO2Me

O

OLi

Ph +

(C) H

O

H 2O

O

OLi

O

+

H 2O

Me

O

H 2O

Me

O

4. H2O

OLi O +

(D) H

Me

2. LiN(i-P)2

CO2H 3. PhCHO

Me H

(E) H

1. BuLi

Ph

(B)

3. H3O+

2. i-PrCHO

Me3SiO H

11. Use models to predict the major diastereomer formed during this methylation reaction. OTBDPS O

O N

Et

1. NaN(SiMe3)2

O

2. MeI

Ph

12. Provide a mechanism for the following transformation: O

O

O

CHO

O

O

O

1. NaOH +

2. H

13. Provide a mechanistic rationale for the following transformation: O

O

H

Me H H

3 equiv PhSH 5 equiv K2CO3

H

MeOH, Overnight

H

PhS Me O

O

14. For each of the following, give all products. What is the strategic purpose of the first condensation with benzaldehyde? Compare and contrast routes (a) and (b): O

(A)

Me Me OAc

PhCHO

p-TsCl

I

Py

KOH, EtOH

O

(B)

II

NaH, DME Heat

NaH, DMF

V

KOH, HMPA Ethylene glycol

III

+

4-Aminobutanoic acid, 195°C

IV

VI

TsO Me Me

15. Show a full synthetic sequence, with all intermediate products and reagents, for conversion of 1-methylcyclohex-1-ene into each of the molecules shown. O CHO Ph

(C)

(A) Ph O

(B) Bn

741

13.9 CONCLUSION

16. In each case, give the major product, with correct stereochemistry where appropriate.

(A)

EtO2C

PhCHO , Microwaves P2O5, Piperidine

CN

O

(C)

CO2H

O MeO

(E)

(B) t-BuPh2SiO

PhCl 1. 2 equiv LDA , THF 2. Allyl bromide

Me

1. 3 equiv LHMDS , THF –78°C

O

2.

Me

O

(F)

(G)

KOH, PhH, rt Dibenzo-18-crown-6

CHO

CO2Et

(I)

NaOEt, EtOH Reflux

NHPf

(H)

CHO Zn, BrCH2CO2Et

(J)

O+

Bz2O2, aq THF

Cl

BF3•OEt2

H 1. Pyrrolidine, cat H+

(K)

O

(L)

2. PhCH2CHO 3. Hydrolysis

OTBDPS CN

(N)

2

2. Me2CHCHO

(P)

N O O Me O

1. LiNEt2, THF 2. Allyl bromide

CN

Ph

20%

O

N

N

O

Me 20% aq HCl

O

(T)

O

OH

Me

NaH, Toluene 18-crown-6

(V)

N

–78°C

Me

O

O O

Bu2BOTf, NEt3, –78°C CH2Cl2

O

(X)

N (MeO)2HC

CH2=CHCHO

Ph

C6H11 1. LDA, THF, DMPU CH3

Toluene, Reflux

(Z)

2. MeI

1. KN(TMS)2 THF, –50°C

O O

O Ph

HO

(AD)

(AB)

O

2.

C3H7O

MeO2C

N

H O

O

(AC) Ph

CHO , 0°C

Ph cat

SO2Ph

O Ph

N , CH2Cl2 N

t-BuOK PhH

t-Bu

1. DBU, Toluene, 60°C, 2 h

O

N Bn CO2Me

N Me

C11 H23 CO2H

AA

(CH 2) 4OTs

N

CO2Et NaH, MeOH

1. 2 equiv NaHMDS, THF

O

2. 3. H2O

O

(Y)

THF, 50°C

LDA , THF–HMPA

I

O

(W)

H

O OAc

OTBS

BnO

Me OH

O

t-Bu

t-Bu

O Piperidine, AcOH

CHO

HO

PhCHO, Cu(OTf)2 aq EtOH , 0°C O

O

(R)

3. dibal, Pentane –78°C

OSiMe3

OH 1. 2 equiv LDA, THF

H

2

3. LDA, DMPU

MeO2C

t-BuOK, THF

N Bn

CN Et

LDA, DMPU CO2Me 1, 2. ClCH CH Br

(S)

N

2. Allyl bromide

(O)

(Q)

1. LDA, THF, –78°C 2. 1-Cyclohexenylmethyl bromide

O

1. MeLi, THF

(M) Me3SiO

(U)

2. H3O+ Pf = 9-Phenyl-9-fluorenyl

1. LDA, THF, –78°C

Ph CH2CLN 2. EtCN 3. H3

OMe

CO2Me 1. LTMP , MeOH/THF

MeO2C

OSiMe2t-Bu

O

Cl

CO2Et

Cl O

2. NaOH, H2O2

O

H

1. CH2(CO 2Me)2 , NaOMe , DMF 23°C, 2 d 2. NaOMe, MeOH, 65°C, 3.5 h

CO2Me

Me

(D)

1. (–)-Ipc 2B(allyl) , Ether

CHO

2. 1 N HCl wash

(AE)

OSiMe3

Ph

OEt

SmI3, THF

N

Bn

O O

O KOH, EtOH

(AF)

Reflux

O

(AG)

O

1. O3, CH2Cl2, –78°C 2. Me2S

O O

3. DBU, CH2Cl2, 1 d 4. Ac2O, DMAP

742

13. NUCLEOPHILIC SPECIES THAT FORM CARBON-CARBON BONDS: ENOLATE ANIONS

17. In each case, provide a suitable synthesis. Provide all reagents and show all intermediate products. O

OBz O

MeO

OMe

(B)

MeO 2C

O

MeO 2C

(C)

EtO 2C

EtO 2C

Et

(A) O

O

(D)

TBSO

Et HO

O CO 2Et

Me

O O

CO 2Me

(E)

NH2

HO

(F)

O

Et

Ph

N

OTIPS

N H

H

Et

O H

O H

O

MeO 2C Bu Me3Si

O

(G)

(H) BnO

O BnO

Br

O

NHBn OMe O

OH

CH(OMe) 2

OH

OSiMe 2t-Bu

OMe O

OH

O

O

CHO

(I)

(J) HO Me

Me

OMe O O Ph

O

OH

OMe O O

O Ph

CHO

(K) Ph

Me

Me

Me

(L)

CO 2Et

Ph

O OBn OBz

OBn

OH

O

(O)

O

O

(M)

O

(N) Ph

O

(P)

CO 2Et

S

O

HO

S

Me

18. For each of the following, choose a commercially available starting of six carbons or less. Convert that starting material into the target shown, showing all intermediate products and all reagents. Show your retrosynthetic analysis: OMe

(A)

(E)

(B)

C N O

O Me

O

(F)

O

O Ph

(C)

(D) O

OH Ph

O

(G)

(H)

C H A P T E R

14 Pericyclic Reactions: The Diels-Alder Reaction 14.1 INTRODUCTION Reactions involving ionic or highly polarized compounds comprise the majority of organic reactions, as seen in previous chapters. Chapters 11–13 discussed methods for making carbon bonds via nucleophilic carbon species. Chapter 16 will discuss carbon-carbon bond formation with reagents involving electrophilic carbon, and Chapter 17 will discuss making carbon-carbon bonds via radical, carbine, and related processes. Most reactions in previous chapters involve one-carbon disconnections. There are reactions involving two-bond disconnections, however, that lead to useful disconnect products. Many of these reactions generate carbon-carbon bonds by electronic reorganization of molecules, rather than via ionic or radical species in what are generically called pericyclic reactions. The Diels-Alder reaction, Claisen rearrangement, and the Cope rearrangement fall into this category. Without question, these reactions are among the most useful in synthesis. This chapter will focus on methods that use pericyclic reactions, with a focus on [4+2]-cycloaddition reactions. Other pericyclic reactions will be discussed in Chapter 15. As in previous chapters, however, the emphasis will be on the synthetic utility.

14.2 FRONTIER MOLECULAR ORBITAL THEORY This chapter deals with molecules containing π-bonds: alkenes, dienes, carbonyls, imines, nitriles, and so on. In a typical alkene, bonding is described in terms of the σ- and π-orbitals. Electronic interactions are represented by the molecular orbitals, using σ-orbitals (sp2 hybridized for alkenes, 1) and π-orbitals (2). These orbitals take on a directional character when they are in close proximity to another atom, orbital, or polarized species, as in a covalent bond. +

–

+

–

1

2

There are two possibilities when two sp2 hybridized orbitals come together to share electron density in a covalent bond. Orbitals of the same sign can be directed toward each other as in 3, or orbitals of opposite sign can be directed toward each other as in 4. In 3, the electron density between the two nuclei is maximized (electronic attraction for the positive nucleus), but in 4, it is minimized. A strong covalent bond (mutual sharing of electron density between nuclei) is associated with significant electron density between the nuclei. Therefore orbital diagram 3 represents a bonding interaction, and 4 is usually viewed as an antibonding interaction for a σ covalent bond. These terms have meaning only when considering the outermost orbital containing electrons, the valence electrons (valence orbitals). Similar arguments can be made for the overlap of the p-orbitals that constitute a π-bond. In 5, the overlap of orbitals with the same sign represents the maximum bonding interaction for the π-bond of ethylene (a bonding molecular orbital), and 6 represents an antibonding molecular orbital. As with a σ-bond, the net energy of the bonding interaction in a π-bond is lower than that of the antibonding interaction, as shown in Fig. 14.1, and it is more stable. The energy of each molecular orbital can be measured, and this energy difference (ΔE) will vary with substituents on the alkene. This observation has important implications for the reactivity of these compounds, which will be discussed later in this chapter.

Organic Synthesis http://dx.doi.org/10.1016/B978-0-12-800720-4.00014-3

743

Copyright © 2017 Michael Smith. Published by Elsevier Inc. All rights reserved.

744

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

+1.5 eV +

–

+

–

E

4

E –

+

E +

6

CH2KCH2 E

– –10.52 eV

3 5

FIG. 14.1 Bonding and antibonding σ- and π-orbitals.

The π-orbitals are the key to understanding pericyclic reactions, and the focus of this section will be on a π-bond of alkenes and dienes. The difference in energy between the two π molecular orbitals (5 and 6) can be represented in terms of their relative energies, as shown for ethene in Fig. 14.1 from Houk and coworker’s1 studies. The lowest energy orbital containing bonding electrons is called the highest occupied molecular orbital, the HOMO. Experimentally, the energy of the HOMO is closely associated with the negative of the ionization potential (IP) of the molecule.2 In terms of chemical reactivity, the HOMO is the orbital that donates electrons and the IP is a reasonable experimental measure of its energy. The higher energy orbital does not contain electrons, but it is the next available energy level if electrons are available. This orbital is called the lowest unoccupied molecular orbital, the LUMO. The energy that describes accepting electrons is the electron affinity (EA), and experimentally, the energy of the LUMO is closely associated with the negative of the EA of the alkene.1,2 In a simplistic reaction model, the LUMO would be expected to accept electrons from the HOMO of another π-bond. The difference in energy between the HOMO and LUMO (ΔE) is determined by the difference of the IP (
10.52 eV for ethane),2 where 1 eV ¼ electron volt ¼ 23.06 kcal (96.53 kJ) mol
1, and the EA is +1.5 eV [for a ΔE of 12.02 eV ¼ 277.2 kcal (1160.4 kJ) mol
1]. The HOMO and LUMO of ethene (or of any other molecule) are referred to as frontier molecular orbitals (FMO), after the pioneering work of Fukui et al.3 By analyzing several different types of reactions with FMO theory, the HOMO was believed to deliver electrons to the LUMO of a reactive center that could accept the charge. The FMO theory has now been shown to be “a first approximation to a perturbation treatment of chemical reactivity.”4 Perturbation theory,5 first described by Coulson and Longuet-Higgins,6 treats the molecular orbital of two interacting components as a perturbation of the product of their individual orbitals.7 The theory fails to explain large perturbations, and does not predict the transition state, but gives “an estimation of the slope of an early part of the path along the reaction coordinate leading to the transition state.”7 This theory usually allows one to identify the higher and the lower energy transition state, and this information is useful for predicting reactivity. The transition state energies can be described as reactant-like (early transition state) or product-like (late transition state). The well-known Hammond postulate states “that transition states for exothermic reactions are reactant-like, and for endothermic reactions are product-like.”8 Frontier orbital effects are particularly important in exothermic reactions.9 Analysis of a reaction by frontier orbital theory has additional benefits, particularly for predicting reactivity and stereochemistry. The concept of orbital symmetry is important for all pericyclic reactions. Of particular importance is the difference in energy between the HOMO of one π-system, and the LUMO of a second π-system, because this energy difference (ΔE) will be used to predict reactivity (see Section 14.5.2). Woodward and Hoffman10 pointed 1

Houk, K. N.; Sims, J.; Watts, C. R.; Luskus, L. J. J. Am. Chem. Soc. 1973, 95, 7301.

2

Houk, K. N. In Pericyclic Reactions; Marchand, A. P.; Lehr, R. E., Eds.; Academic Press: NY, 1977; Vol. II; p 203.

(a) Fukui, K.; Yonezawa, T.; Nagata, C.; Shingu, H. J. Chem. Phys. 1954, 22, 1433. (b) Fukui, K. In Molecular Orbitals in Chemistry, Physics and Biology; L€ owdin, P.-O.; Pullman, B., Eds.; Academic Press: NY, 1964; p 513.

3

4

Reference 2, p 183.

5

(a) Dewar, M. J. S.; Dougherty, R. C. The P.M.O. Theory of Organic Chemistry; Plenum Press: NY, 1975. (b) Dewar, M. J. S. Molecular Orbital Theory for Organic Chemists; McGraw-Hill: NY, 1969. 6

Coulson, C. A.; Longuet-Higgins, H. C. Proc. R. Soc. Lond. Ser. A 1947, 192, 16.

7

Fleming, I. Frontier Molecular Orbitals and Organic Chemical Reactions; Wiley: London, UK, 1976; p 23.

8

Hammond, G. S. J. Am. Chem. Soc. 1955, 77, 334.

9

Reference 7, p 24.

(a) Woodward, R. B.; Hoffman, R. Angew. Chem. Int. Ed. Engl. 1969, 8, 781. (b) Idem The Conservation of Orbital Symmetry; Verlag-Chemie GmbH/ Academic Press: Weinheim, 1971.

10

745

14.3 HOMO, LUMO ENERGIES, AND ORBITAL COEFFICIENTS

out “that electrocyclic reactions followed the stereochemistry dictated by the symmetry, or nodal properties of the HOMO of the polyene.”11 A cautionary note was raised by Spino et al.,12 who showed that the HOMOdiene-LUMOalkene values should be used with caution when attempting to predict inverse electron demand Diels-Alder reactions (see Section 14.5), where the asynchronicity of the transition state should be taken into account.

14.3 HOMO, LUMO ENERGIES, AND ORBITAL COEFFICIENTS Fig. 14.1 showed the molecular orbitals for ethene, and these orbitals are shown again in Fig. 14.2. The magnitude of each orbital is called the orbital coefficient. The + and
signs correlate with the symmetry of the orbital,2 which is somewhat arbitrary, but useful for examining the directionality of two reactive orbitals. When two π-systems differ in their substitution pattern, there will be differences in the magnitude of orbital densities (larger or smaller orbitals). Predictions concerning reactivity must, therefore, indicate the magnitude of electron density in each orbital. Orbital coefficients can be used to understand both reactivity and regiochemistry.13 The orbital coefficients in the HOMO of a symmetrical molecule (e.g., ethene) are identical in magnitude and sign. The LUMO coefficients are equal in magnitude, but opposite in sign, also a result of this symmetry, which leads to no node in 5 and one node in 6 (a node is a point of zero electron density relative to the orbital lobes). The importance of orbital coefficients will be discussed in Section 14.5.3.2. Low-energy orbitals are associated with a minimal number of nodes, and higher energy orbitals with increasing numbers of nodes. If the structure of ethene is changed by replacing a hydrogen atom with an electron-withdrawing group, as is the case for methyl acrylate, electron density is distorted away from the π-bond of the alkene toward the carbonyl. This distortion will influence the energies of the HOMO and LUMO, as well as the magnitude of the orbital coefficients. The energy of the HOMO of methyl acrylate is
10.72 eV and the LUMO energy is 0 eV.13 These energy values are lower than the frontier orbitals of ethene, which is shown in Fig. 14.2,2 and the dashed line marks the energy of the LUMO of ethene as the reference point. The LUMO energy is affected to a greater extent than the HOMO energy. The electron-withdrawing group diminishes the size (orbital coefficient) of the orbitals closest to the δ+ carbonyl carbon, leaving the most distant orbital with the largest coefficient.

CH2KCHCO2Me

CH2KCH2 +0.71

0 eV

CH2KCHOMe +2.0 eV +0.66

+1.5 eV

+0.69

6 –0.47

+0.71 +0.71

–10.72 eV

+0.43 +0.33

FIG. 14.2

–0.71

–9.05 eV

–0.72

+0.61 +0.39

–10.52 eV

5

Changes in alkene HOMO and LUMO energies upon introduction of substituents.

When a hydrogen atom of ethene is replaced with an electron-releasing group (e.g., OMe, methyl vinyl ether in Fig. 14.2), the opposite effect is expected. The oxygen atom releases electrons toward the π-bond, raising the energy of the orbitals relative to ethene (to
9.05 eV for the HOMO and +2.0 eV for the LUMO).2 In Fig. 14.2,2 the HOMO coefficients of the carbon proximal to the methoxy are diminished (0.39 vs. 0.61), whereas the orbital density of the distal carbon is increased. In the LUMO, the orbital on the proximal carbon is slightly larger than that on the distal carbon. 11

Reference 2, pp 183–184.

12

Spino, C.; Rezaei, H.; Dory, Y. L. J. Org. Chem. 2004, 69, 757.

13

Reference 2, pp 203–205.

746

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

TABLE 14.1 The HOMO and LUMO Energies and Orbital Coefficients of Common Alkenes c2

c1

c3

c4 LUMO

HOMO

X

X Alkene

HOMO (eV)

c1 a

c2a

LUMO (eV)

c3a

c 4a

H2CKCH2b

–10.52

0.71 [0.63]

0.71 [0.63]

1.5

0.71 [0.86]

–0.71 [–0.86]

–10.15

0.44

0.30

0.5

0.67

–0.54

CKCHClb

H2

H2CHMeb

0.67

0.6

1.8b,c

0.67

–0.65

MeHCKCHMe

d

–9.13

[–0.61]

[–0.61]

2.22e

[–0.78]

[0.78]

EtHCKCH2

–9.63d

[–0.64]

[–0.59]

2.01f

[–0.80]

[0.80]

Cyclohexene

–8.94g,h

[–0.60]

[–0.60]

2.1h

[–0.79]

[0.79]

H2CKCHPh

–8.48

0.49 [–0.43]

0.32 [–0.29]

0.8

0.48 [–0.51]

–0.33 [0.35]

H2CKCHCH2OHb,e,i

–10.1

–0.31

–0.36

–0.64

–0.56

H2CKCHOMeb,j

–9.05 [–8.93]d

0.61 [–0.60]

0.39 [–0.42]

0.66 [–0.79]

–0.72 [0.83]

H2CKCHOEt

–9.03b,d

5.53b,d

H2CKC(OEt)2

–8.68b,d

6.28b,d

H2CKCHSMeb

–8.45

0.34

0.17

0.63

–0.48

H2CKCHNMe2b

–9.0

0.50

0.20

2.5

0.62

–0.69

H2CKCHCO2Hc

–10.93

[–0.49]

[–0.47]

2.91

–0.63 [–0.65]

0.46 [0.47]

0.69

–0.47

0.60c

–0.58c [–0.62]

–0.40c [0.43]

0.43

–0.53

0.25

[–0.61]

[0.44]

–9.88

Mek

H2CKCHCO2

–10.72

[–10.64]b,d

0.33

0.43

MeO2CHCKCHCO2Meb,e –10.95b,d

H2CKC(CO2Et)2

–10.89c c

2.0

1.0

0

[3.12]b,d

1.94b,d

–10.96b,d

H2CKCHCHO

1.1

2.43b,d 0.58b [–0.59]

0.48b [–0.55]

[–0.10]

[–0.18]

0.56b,c

0.70b,c

–0.64m

–0.74b,c

0.52b,c

H2CKCHCHO•BF3

–12.49

Methyl vinyl ketone

–10.16

l

Cyclopentenone

–9.34d,m

p-Benzoquinone

–10.29n

[–0.34]

[–0.34]

–1.91o

[–0.33]

[0.33]

Maleic anhdyride

–11.95n [–12.13]b,d

[–0.34]

[–0.34]

–0.57p [–1.03]b,d

[–0.51]

[0.51]

N-Phenylmaleimide

–10.64q

0[.10]

–0.49o

[0.46]

0.49 [0.6]

0b,c

0.68 [0.55]

–0.54[–0.67]b,c

–1.54

0.66

–0.49

0.54 [0.50]

–0.32 [0.30]

[–0.86]

[0.86]

[–0.10]

H2CKCHCN

–10.92

0.60

H2CKC(CN)2r

–11.38

0.61

0.45

0.62 [0.005]

0.60 [–0.002]

[–0.62]

[–0.62]

(NC)2CKC(CN)2

–11.8s

H2CKCHNO2b

–11.4

HCLCH MeCLCMe

–11.4 –9.9

d

h

[0.64]b,c

[–0.46] b,c

–1.80t[–2.03]u 0.7 h

2.6

v

3.43

747

14.3 HOMO, LUMO ENERGIES, AND ORBITAL COEFFICIENTS

TABLE 14.1 The HOMO and LUMO Energies and Orbital Coefficients of Common Alkenes—cont’d Alkene

HOMO (eV)

c1 a

c2a

LUMO (eV)

c3a

c 4a

HCLCCNr MeO2CCLCCO2Mew

–11.81

0.56 [0.37]

0.43 [0.37]

0

–11.5q

–0.60

0.57 [0.42]

–0.41 [–0.42]

Acetaldehydex

–11.49

0.44y

4.21

[0.44]

[–0.44]

Thioacetaldehydex

–9.25

1.81

0.64

–0.62 [.09]z

Phenanthrene (C9-C10)

–8.1d,m

0.62y [–0.37]

[0.37]

–0.31m

0.65

–0.66 [0.07]z

H2CKCKCH2aa

–10.14

–0.47

–0.56 [0.20)z

2.4

0.67

–0.63 [0.10]z

–0.66

z

0.78

0.54

–0.65 [0.05]z

z

1.01

0.48

–0.63 [0.04]z

z

–0.01

0.22

0.57 [0.75]ab

z

–0.07

–0.51

0.71 [–0.32]ab

–0.16

0.45 [–0.23]ab

aa

H2CKCKCMe2 H2CKCKCHOMeaa H2CKCKCHCNaa

–9.67 –9.33 –10.45

–0.53 –0.63

g

–0.63 [0.09] –0.63 [0.12] –0.55 [0.22]

H2CKCKCHCO2Meaa

–10.62

–0.67

–0.53 [0.23]

H2CKCKOac

–12.55 [–12.7]ad

–0.73

MeHCKCKOac

–11.52 [–8.95]ae

–0.67

–0.27 [0.61]ac 2.55 –0.33 [0.55]ac

PhHCKCKO Me2CKCKO

[–10.61]

–0.53

–0.30 [0.43]

ClHCKCKO Cl2CKCKO

–9.24ae

NCHCKCKO

–10.07ae

–8.45ae –9.15ae

a Values in [ ] calculated using MacSpartan, v2.0.2, Hartree-Fock calculations at STO-3G level; bAll data from Reference 13 unless otherwise noted; cFrom Reference14; dFrom Reference15; eFrom Reference16; fFrom Reference17; gFrom Reference18; hFrom Reference19; iFrom Reference20; jFrom Reference21; k From Reference22; lFrom Reference23; mFrom Reference24; nFrom Reference25; oFrom Reference26; pFrom Reference27; qFrom Reference28; rAll data from Reference29; sFrom Reference30; tFrom Reference31; uFrom Reference32; vFrom Reference33; wFrom Reference34; xFrom Reference35; yFor O]C, where c1 is for O and c2 is for C; zThird coefficient (c1 for CR2; c2 for ]C]; c3 for CH2]); aaAll data from Reference36; abThird coefficient (c1 for RCH; c2 for ]C]; c3 for ]O); acAll data from Reference37 unless otherwise noted; adFrom Reference38; aeFrom Reference39.

14

Kahn, S. D.; Pau, C. F.; Overman, L. E.; Hehre, W. J. J. Am. Chem. Soc. 1986, 108, 7381.

15

Handbook of Chemistry and Physics, 87th ed.; CRC/Taylor and Francis: Boca Raton, FL, 2006; pp 10-203–10-223.

16

Jordan, K. D.; Michejda, J. A.; Burrow, P. D. Chem. Phys. Lett. 1976, 42, 227.

17

Kadifachi, S. Chem. Phys. Lett. 1984, 108, 233.

18

Sauer, J.; Wiest, H.; Mielert, A. Chem. Ber. 1964, 97, 3183.

19

Pearson, R. G. J. Org. Chem. 1989, 54, 1423.

20

Yoshitake, Y.; Yamaguchi, K.; Kai, C.; Akiyama, T.; Handa, C.; Jikyo, T.; Harano, K. J. Org. Chem. 2001, 66, 8902.

21

Spino, C.; Rezaei, H.; Dory, Y. L. J. Org. Chem. 2004, 69, 757.

22

Suishu, T.; Shimo, T.; Somekawa, K. Tetrahedron 1997, 53, 3545.

23

Hentrich, G.; Gunkel, E.; Klessinger, M. J. Mol. Struct. 1974, 21, 231.

24

Michl, J.; Becker, R. J. J. Chem. Phys. 1976, 46, 3889.

25

Dougherty, D.; Brint, P.; McGlynn, S. P. J. Am. Chem. Soc. 1978, 100, 5597.

26

Heinis, T.; Chowdhury, S.; Scott, S. L. Kebarle, P. J. Am. Chem. Soc. 1988, 110, 400.

27

Samuilov, Ya.D.; Uryadov, V. G.; Uryadova, L. F.; Konovalov, A. I. Zh. Org. Khim. 1985, 21, 1249 (Engl. 1137).

28

El-Basil, S.; Said, M. Ind. J. Chem. 1980, 19B, 1071.

29

Houk, K. N. Acc. Chem. Res. 1975, 8, 361.

30

Dinur, U.; Honig, B. J. Am. Chem. Soc. 1979, 101, 4453.

31

Konovalov, A. I.; Kiselev, V. D.; Vigdorovich, O. A. Zh. Org. Khim. 1967, 3, 2085 (Engl. 2034).

32

Dewar, M. J. S.; Rzepa, H. S. J. Am. Chem. Soc. 1978, 100, 784.

33

Ng, L.; Jordan, K. D.; Krebs, A.; R€ uger, W. J. Am. Chem. Soc. 1982, 104, 7414.

34

Bihlmaier, W.; Huisgen, R.; Reissig, H.-U.; Voss, S. Tetrahedron Lett. 1979, 2621.

35

Yu, Z.-X.; Wu, Y.-D. J. Org. Chem. 2003, 68, 412.

36

Padwa, A.; Bullock, W. H.; Kline, D. N.; Perumattam, J. J. Org. Chem. 1989, 54, 2862.

37

Meslin, J. C.; N’Guessan, Y. T.; Quiniou, H.; Tonnard, F. Tetrahedron 1975, 31, 2679.

38

Kuzuya, M.; Miyake, F.; Okuda, T. J. Chem. Soc. Perkin Trans. 2 1984, 1471.

39

Bock, H.; Hirabayashi, T.; Mohmand, S. Chem. Ber. 1981, 114, 2595.

748

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

In general, the HOMO coefficients for methyl vinyl ether are typical of other electron-releasing substituents, but the LUMO coefficients are often closer to each other in magnitude. Table 14.1 shows the HOMO and LUMO energies for several common alkenes, along with the HOMO and LUMO coefficients where available. The data in Table 14.1 came from a variety of sources, including calculations40 and experimentally determined IP and EA data. These numbers vary with the experimental method, and the sophistication level of the calculation. They should be viewed as useful first estimates when comparing ΔE values using values from two different sources (and/or methods). The purpose of this table is to use the HOMO-LUMO energies to predict chemical reactivity (Section 14.5.2). The orbital coefficients will be used to predict regioselectivity (Section 14.5.3.2). When two or more electronwithdrawing groups are present (tetracyanoethylene, benzoquinone, maleic anhydride, etc.), greater lowering of the HOMO is apparent and the LUMO energy may fall TiCl4 > BF3, see Section 2.3) have a greater effect on this reaction. Increased selectivity and rate of reaction are also observed with acyclic dienes. Additional examples of Lewis-acid catalysis will be presented throughout this chapter. There is also a report that amidinium ions catalyze the Diels-Alder reaction.125 There are many synthetic applications of Lewis-acid catalyzed Diels-Alder reactions. Formation of 56 in Section 14.5.3.2 is one example. Another example is taken from the synthesis of (
)-lyconadin C by Waters and Cheng,126 in which 5(R)-methylcyclohex-2-en-1-one reacted with 2-tert-butyldimethylsilyloxybuta-1,3-diene 78, in the presence of the diethylaluminum chloride, to give a 96% yield of 79, with a selectivity of 20:1 cis/trans). –0.475

+

Uncatalyzed

+ Ph

Ph

Ph 0.625

Catalyzed

–0.39

H

H

–0.475

CHO

O

O H

–0.09

0.625 0.59

0.69

77 O

O

+ OTBS

0°C to rt

Me (5R)-Methylcyclohex2-en-1-one

H

Et2AlCl, PhCH3, 4.5 h

78

Me

H

OTBS

79 (96%, 20:1 cis/trans)

14.6.2 Rate Enhancement in Aqueous Media For a long time, solvent polarity was believed to have little or no effect on the course of a Diels-Alder reaction. Berson et al.,127 however, showed a clear relationship between the endo-exo-product ratio and solvent polarity in the Diels-Alder reaction of cyclopenta-1,3-diene and acrylates. A patent by Hopff and Rautenstrauch128 in 1939 reported what appears to be the first water accelerated reaction when they showed that yields in the Diels-Alder reaction were 124

Sauer, J.; Kredel, J. Tetrahedron Lett. 1966, 731.

125

Schuster, T.; Kurz, M.; G€ obel, M. W. J. Org. Chem. 2000, 65, 1697.

126

Cheng, X.; Waters, S. P. Org. Lett. 2013, 15, 4226.

127

Berson, J. A.; Hamlet, Z.; Mueller, W. A. J. Am. Chem. Soc. 1962, 84, 297.

128

Hopff, H.; Rautenstrauch, C. W. U. S. Patent 2,262,002, 1939 (Chem. Abstr. 1942, 36, 10469).

768

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

enhanced in aqueous detergent solutions. In 1980, Breslow and Rideout129 showed there was a hydrophobic acceleration for an intermolecular Diels-Alder reaction in which cyclopenta-1,3-diene reacted with MVK to yield 80. When nonpolar compounds are suspended in water their relative insolubility causes them to associate, diminishing the water-hydrocarbon interfacial area (a hydrophobic effect). This association is greater in water than in methanol, which brings the reactive partners into close proximity, increasing the rate of reaction. Any additive that increases the hydrophobic effect will increase the rate, as shown in Table 14.4,129 for the preparation of 80. Lithium chloride (LiCl) increases the hydrophobic effect by salting-out nonpolar material, but guanidinium chloride decreases hydrophobic interactions. β-Cyclodextrin (cycloheptaamylose) possesses a hydrophobic cavity and if the reactive intermediates can fit within this cavity there is a significant rate enhancement. The smaller α-cyclodextrin cannot accommodate the reactive species, and the rate is significantly diminished. Microwave-assisted Diels-Alder reactions have been reported in both water and ionic liquid media using Lewis-acid catalysts.130

TABLE 14.4

Hydrophobic Effects in the Diels-Alder Reaction of Cyclopentadiene and MVK in Aqueous Media O

+

Additive Me

O

20°C Me

80 Solvent

k2 105 M–1 s–1

Additive

Isooctane

5.94

Methanol

75.5

Water

4410 10,800

LiCl +

–

(NH2)3C Cl

4300

-Cyclodextrin

10,900

-Cyclodextrin

2610

Reprinted with permission from Rideout, D.; Breslow, R. J. Am. Chem. Soc. 1980, 102, 7816. Copyright © 1980 American Chemical Society.

There are many examples of Diels-Alder reactions that exploit this rate-enhancement effect. Grieco and Garner’s131 synthesis of quassinoids led to the reaction of 81 and 82, in benzene at 25°C, which gave a 1:1.18 mixture of 83 and 84 in 52% yield, but required a reaction time of 288 h. When the same reaction was done in water, an 82% yield was obtained (1.3:1 83/84) after 168 h. Diels-Alder cyclization with the free acid rather than the ester, in water, gave an 85% yield (1.5:1 83/84) in 17 h. Similar reaction in 1:1 water/methanol gave a quantitative yield after 97 h (1:1.25 83/84). When the sodium salt of the acid was used, a quantitative yield was obtained in 5 h (3:1 83/ 84).131 In all cases, the reactions were carried out at room temperature with a fivefold excess of diene over dienophile. Breslow et al.132 showed that addition of guanidinium perchlorate to a Diels-Alder reaction slowed the reaction,

129

Rideout, D.; Breslow, R. J. Am. Chem. Soc. 1980, 102, 7816.

130

Chen, I.-H.; Young, J.-N.; Yu, S. J. Tetrahedron 2004, 60, 11903.

131

Grieco, P. A.; Garner, P.; He, Z. Tetrahedron Lett. 1983, 24, 1897.

132

Breslow, R.; Maitra, U.; Rideout, D. Tetrahedron Lett. 1983, 24, 1901.

769

14.6 RATE ENHANCEMENT IN DIELS-ALDER REACTIONS

which was taken as evidence that the special ability of water to enhance the rate of the Diels-Alder reaction was due to a hydrophobic effect. MeO

Me

CHO

CHO

288 h

CO2Et

MeO

CO2Et

H

Me

O

H

Me

PhH, rt

+ Me

Me

Me MeO

CHO

84

83

82

O

H Me

Me 81

CO2Et

H

+

O

H

Me

An aqueous environment creates a medium where the “reactants orient themselves within a micelle, thereby increasing their effective molarity and, consequently, the rate.”133 Liotta and coworkers133, however, point out that reaction of 2,6-dimethylbenzoquinone (2,6-dimethylcyclohexa-2,5-diene-1,4-dione) and a diene carbonate,(E)-buta1,3-dien-1-yl methyl carbonate, showed no rate enhancement in water relative to the same reaction in benzene. The carbonate was essentially insoluble in water and rate enhancement requires at least partial solubility of the diene and/or dienophile in the aqueous medium. Other solvents with the ability to enhance molecular aggregation while maintaining the solubility of the reactants showed an increase in rate. The reaction of (E)-buta-1,3-dien-1-yl methyl carbonate and 2,6-dimethylcyclohexa-2,5-diene-1,4-dione showed a large rate enhancement in ethylene glycol (relative to methanol), for example, to give a quantitative yield of 85. The influence of several solvents (relative dielectric constants and relative rates with reaction in benzene as the standard) was examined for several substrates.133 Liotta and coworkers133 explained this effect by comparing the relative positions of diene and dienophile in a micelle, where a π-stacked arrangement in the micelle assumes a smaller volume and is probably preferred. Grieco and Larsen134 also showed that iminium salts (e.g., 86), formed by reaction of an amine and formaldehyde (Section 7.4) reacted with cyclopenta-1,3-diene in aqueous media to give a near-quantitative yield of the Diels-Alder adduct, 87.134 An internal Diels-Alder reaction was also possible (Section 14.7). MeO2CO

Me MeO2CO

+

O

O Me

O Me

(E)-Buta-1,3-dien-1-yl methyl carbonate

Me

2,6-Dimethylcyclohexa2,5-diene-1,4-dione

H

O

85

The use of highly polar media (e.g., water) clearly assists the Diels-Alder reaction, and similar rate enhancement has been observed in other reactions. Grieco et al.135 showed that 5.0 M lithium perchlorate in diethyl ether is an effective medium for these reactions. In a synthesis of (
)-elisapterosin B, Rychnovsky and Kim136 used the reagent to facilitate the Diels-Alder reaction of diene 88 and quinone 89. The two products were formed in 75% yield, as a 1.7:1 mixture of 90 and 91.

133

Dunams, T.; Hoekstra, W.; Pentaleri, M.; Liotta, D. Tetrahedron Lett. 1988, 29, 3745.

134

Larsen, S. D.; Grieco, P. A. J. Am. Chem. Soc. 1985, 107, 1768.

For examples, see (a) Grieco, P. A.; Nunes, J. J.; Gaul, M. D. J. Am. Chem. Soc. 1990, 112, 4595. (b) Grieco, P. A.; Moher, E. D. Tetrahdron Lett. 1993, 34, 5567. (c) Grieco, P. A.; Beck, J. P. Tetrahdron Lett. 1993, 34, 7367. (d) Grieco, P. A.; Beck, J. P.; Handy, S. T.; Saito, N.; Daeuble, J. F. Tetrahdron Lett. 1994, 35, 6783. (e) Grieco, P. A.; Piñeiro-Nuñez, M. M. J. Am. Chem. Soc. 1994, 116, 7606. (f) Grieco, P. A.; Dai, Y. J. Am. Chem. Soc. 1998, 120, 5128. (g) Grieco, P. A.; Kaufman, M. D. Tetrahedron Lett. 1999, 40, 1265.

135

136

Kim, A. I.; Rychnovsky, S. D. Angew. Chem. Int. Ed. 2003, 42, 1267.

770

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

Ph

HCHO

Ph

N

NH2•HCl

H 2O

CH

H

N

86

87

OAc

AcO O OMe

+ Me

H

O

AcO OMe

H

O OMe

5 M LiClO4, Ether 1d

H

Me

H

Me

O OTIPS 88

Ph

Cl

Me O

+ H

H

Me

OTIPS

O

OTIPS

90

89

Me

91

When highly polar media such as this are used, weak acids become strong acids, and protonation of dienes has been observed in intramolecular reactions, that can lead to diene isomerization. Such isomerization may be competitive with cycloaddition.137 Grieco and Kaufman137 found that for such reactions, water is the polar medium of choice. Auge et al.138 found that lithium trifluoromethane-sulfonate in ether or acetonitrile is a useful substitute for lithium perchlorate in cycloaddition reactions.

14.6.3 Rate Enhancement Under High-Pressure Conditions The rate of a Diels-Alder reaction can be increased by the application of high pressure. An early example by Dauben and Kozikowski139 demonstrated that methyl (E)-penta-2,4-dienoate reacted with 92 to give 93 in 88% yield. Matsumoto et al.140 reviewed applications of high-pressure techniques to organic synthesis, and also published a monograph.141 The second part of the review140b deals with high-pressure pericyclic reactions, including Diels-Alder reactions. High pressure is but one of a number of methods for activating reactions, including microwave and ultrasound enhancement.142 Microwave-assisted Diels-Alder reactions have been shown to be more effective when pressurized reaction vessels are used.143 CO2Me MeO2C

10 kbar, 18 h

+

N

N Me

Methyl (E)-penta-2,4-dienoate

92

Me 93 (88%)

The thermodynamic properties of solutions are well known, and the rate of a reaction can be expressed in terms of the activation volume, ΔV{ δ ln k ΔV { ¼ RT δp

(14.4)

“The activation volume is the difference in partial molal volume between the transition state and the initial state. From a synthetic point of view, this could be approximated by the molar volume.”142 Eq. (14.4) suggests that the rate of the 137

Grieco, P. A.; Kaufman, M. D. J. Org. Chem. 1999, 64, 6041.

138

Auge, J.; Gil, R.; Kalsey, S.; Lubin-Germain, N. Synlett 2000, 877.

139

Dauben, W. G.; Kozikowski, A. P. J. Am. Chem. Soc. 1974, 96, 3664.

140

(a) Matsumoto, K; Sera, A.; Uchida, T. Synthesis 1985, 1. (b) Matsumoto, K.; Sera, A. Synthesis 1985, 999.

141

Matsumoto, K.; Morris, A. R. Organic Synthesis at High Pressure; John Wiley and Sons, Inc.: NY, 1991.

142

Jenner, G. Tetrahedron 2002, 58, 5185.

143

Kaval, N.; Dehaen, W.; Kappe, C. O.; Van der Eycken, E. Org. Biomol. Chem. 2004, 2, 154.

771

14.6 RATE ENHANCEMENT IN DIELS-ALDER REACTIONS

reaction will be accelerated with increasing pressure if the volume of activation is negative. As the pressure increases, the value of ΔV{ decreases. When the pressure is >10 kbar (1 bar ¼ 0.986924 atm ¼ 1.1019716 kg cm
2), the system does not strictly obey Eq. (14.4). Considerable information has been collected concerning ΔV{, principally from work by le Noble.144 If the transition state of a reaction involves bond formation, concentration of charge, or ionization, a negative volume of activation often results. Cleavage of a bond, dispersal of charge, or neutralization of the transition state and diffusion control lead to a positive volume of activation. Matsumoto and Morris141 summarizes the reactions for which rate enhancement is expected at high pressure. (1) Reactions in which the molecularity number (number of molecules) decreases when starting materials are converted to products: cycloaddition reactions, condensation reactions. (2) Reactions that proceed via cyclic transition states: Claisen and Cope rearrangements (Sections 15.5.5 and 15.5.4, respectively). (3) Reactions with dipolar transition states: Menschutkin reaction145 (tertiary amines with alkyl halides to produce quaternary ammonium halides), electrophilic aromatic substitution. (4) Reactions with steric hindrance. Table 14.5142 shows the variation of rate constants with pressure. Many high-pressure reactions are done neat, but if a solvent is used, the influence of pressure on that solvent is important. The melting point generally increases at elevated pressures, and this influences the viscosity of the medium (the viscosity of liquids increases approximately two times per kilobar increase in pressure). Controlling the rate of diffusion of reactants in the medium is also important, leading to another influence of high pressure on reactivity.142,146 In most reactions, pressure is applied (5–20 kbar) at room temperature and then the temperature is increased until reaction takes place. The temperature is lowered and the pressure is reduced to isolate the products.

TABLE 14.5

V‡ (cm3 mol–1) –10

–20

–30

Variations of Rate Constants With Pressure

Temperature (°C) 5 kbar

15 kbar

20 kbar

30 kbar 105

50 kbar 7.5 108

25

7.5

57

430

3200

1.8

50

6.4

41

270

1700

7.1 104

1.5

108

100

5.0

25

130

630

104

1.2

107

(–5)

(–10)

(–15)

(–20)

57

3200

25

41

1700

100

25

630

(–10)

(–20) 105

1.8

50

270

7.1 104

130

4

(–15)

(–30)

1.6

104

1.0

10

3.3

10

(–50)

10

5.6 1017

9

2.4 1016

2.9 10

5.0 10

4.0 105

2.5 108

(–40)

7.7 107 1.9

1.6

(–30) 7 6

(–30)

430

10

10

7.1 10

25

1.6

1.8

5

4

50

100

a

kp/k1a [ GG‡ (kJ mol–1)] 10 kbar

3.3

(–60)

1010

107

5.0 109

6

8

2.0 10

2.5 10

(–45)

(–60)

6.0 1015 3.6

1.5

1014

(–100) 4.2 1026

1014

3.7

1024

12

1.9

1021

4.0 10

(–90)

(–150)

Rate constant under high pressure = kp; Rate constant at atmospheric pressure = k1.

Reprinted with permission from Matsumoto, K.; Sera, A.; Uchida, T. Synthesis 1985, 1. Copyright © 1985 Georg Thieme Verlag KG.

(a) le Noble, W. J. Progr. Phys. Org. Chem. 1967, 5, 207. (b) Isaacs, N. S. Liquid Phase High Pressure Chemistry; John Wiley: Chichester, UK, 1981. (c) Asano, T.; le Noble, W. J. Chem. Rev. 1978, 78, 407.

144

(a) Menschutkin, N. Z. Physik. Chem. 1890, 5, 589. (b) Menschutkin, N. Z. Physik. Chem. 1890, 6, 41. (c) Ingold, C. K. Structure and Mechanism in Organic Chemistry, 2nd ed.; Cornell University Press: NY, 1969; p 435. (d) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-60.

145

146

Firestone, R. A.; Vitale, M. A. J. Org. Chem. 1981, 46, 2160.

772

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

O R2

R3

R3

+ O

R1

R

R1

R2

R 94

O C6H13 O

H O

N

C6H13

CH2Cl2, 12 kbar, 1 d

+

O

O

Tf

Tf

Bn 95

96

N CHO Bn

97 (93%)

Diels-Alder reactions with furans, as in the reaction that forms 94, can be very problematic due to the thermal sensitivity of the cycloadduct, and their tendency for cycloreversion (see Section 14.8.2). These problems are due to the strain inherent to the 7-oxabicycloheptane cycloadduct (94), as well as the aromatic character of the furan. Raising the pressure of the reaction up to 13.5–15 kbar in dichloromethane as a solvent leads to fast reactions and yields approaching those obtained with Lewis-acid catalysts and long reaction times.142 2-Methylfuran is more reactive than furan, but substituents on the alkene (dienophile) lower the yield of the cycloadduct, even under high pressure. One synthetic example is taken from Funk and Maeng’s147 synthesis of fasicularin, in which diene 95 reacted with 96 in dichloromethane at 12 kbar to give a 93% yield of 97. Furan reacted with benzoquinone at 20 kbar to give a 14% yield of the endo-adduct and a 15% yield of the exoadduct, along with 71% of unreacted benzoquinone.148 Better results were obtained with functionalized furans (e.g., 98), which reacted with 4-methoxycyclopent-2-en-1-one at 15 kbar to give a 1:2 mixture of 99/100 in 95% yield (after 2 d), in Ghosez and coworker’s149 synthesis of polyfunctionalized cis-hydrindanones. Thiophenes react with maleic anhydride under high-pressure conditions, where it was observed that solvent-free conditions led to a significant lowering of the requisite reaction pressure, improvement in yield of products, and allowed less reactive dienophiles (e.g., methyl acrylate) to be used.150 2-Pyrones are another candidate for high-pressure reactions since the normal Diels-Alder reaction typically requires high temperatures, and the products often lose carbon dioxide in a retro-Diels-Alder reaction.151 TBSO

O TBSO

O

Me

H Toluene, 15 kbar

+

98

4-Methoxycyclopent2-en-1-one

H

+

O

40°C, 2 d

OMe

Me

O

H Me

OMe

O

O H TBSO

99

OMe

100

Pyridones react at high pressure to yield the cycloadduct,152 whereas heating at atmospheric pressure usually gives no reaction at all.152 A variety of relatively unreactive dienes and/or dienophiles give an enhanced yield of cycloadducts under pressure. Naphthalene reacted with maleic anhydride at 9.5 kbar and 100°C.153 Acetaldehyde reacted with 1-methoxybuta-1,3-diene (sealed tube, 160–180°C, 6 h), but gave only 2% of the dihydropyran cycloadduct.154 When the reaction was done neat at 14 kbar and 80°C for 6 h, a 50% yield of pyran was obtained154 and reaction

147

Maeng, J.-H.; Funk, R. L. Org. Lett. 2002, 4, 331.

148

Jurczak, J.; Kozluk, T.; Filipek, S.; Eugster, C. H. Helv. Chim. Acta 1983, 66, 222.

149

Trembleau, L.; Patiny, L.; Ghosez, L. Tetrahedron Lett. 2000, 41, 6377.

150

Kumamoto, K.; Fukada, I.; Kotsuki, H. Angew. Chem. Int. Ed. 2004, 43, 2015.

151

Pfaff, E.; Plieninger, H. Chem. Ber. 1982, 115, 1967.

152

Gisby, G. P.; Royall, S. E.; Sammes, P. G. J. Chem. Soc. Chem. Commun. 1979, 501.

153

(a) Jones, W. H.; Mangold, D.; Plieninger, H. Tetrahedron 1962, 18, 267. (b) Plieninger, H.; Wild, D.; Westphal, J. Tetrahedron 1969, 25, 5561.

154

Makin, S. M.; El’yanov, B. S.; Raifel’d, Yu. E. Izv. Akad Nauk. SSSR Ser. Khim. 1976, 831 (Engl. p 810).

773

14.7 INTRAMOLECULAR DIELS-ALDER REACTIONS

in ether at 20 kbar (65°C, 5 h)155 gave a 62% yield of pyran. Dienamines reacted with conjugated esters at 13.9 kbar to give a 70% yield of an enamine cycloadduct.156,141 Intermolecular Diels-Alder reactions have a large negative volume of activation157 (c.
25 to
45 cm3 mol
1) and a large negative volume of reaction. The activation volume for intramolecular Diels-Alder reactions was measured as
25 cm3 mol
1.158 The effects on the intermolecular Diels-Alder can be dramatic, but intramolecular Diels-Alder reactions do not always respond to high pressure.141,156 There are exceptions.159 In one example, 101 could be cyclized to give a 1:1 mixture of 102/103 in refluxing toluene, at ambient pressure.160 In order to suppress side reactions, a small amount of the radical scavenger BHT (2,6-di-tert-butyl-4-phenylphenol) was used. When done at 13 kbar in dichloromethane at room temperature, an 88% yield of a 1:2.3 mixture of 102/103 was obtained without the need for added BHT. Note that attempts to cyclize 101 using several different Lewis acids were not successful. Intramolecular DielsAlder reactions will be discussed in Section 14.7. A variation in this technique has recently been reported, although not necessarily using the high pressures described above. Diels-Alder reactions can be done under pressure in supercritical CO2, although the synthetic application may be limited by slow reaction rates.161 Addition of a Lewis acid can compensate, as illustrated by Kobayashi’s use of scandium triflate162a in the reaction of MVK with 2-methylbuta-1,3-diene to yield a 93:7 mixture of 104/105. Rayner and coworkers162b had previously used scandium triflate in supercritical CO2, and reported that the maximum selectivity was obtained at a density of 1.12 g mL
1. O2S

O

H

O2S

O

See text

H

O2S

+ H

H 102

101 O

+

O

103 O

5% Sc(OTf)3 50°C Supercritical CO2

O

+

150 atm, 1 d (41%)

104

105

14.7 INTRAMOLECULAR DIELS-ALDER REACTIONS The Diels-Alder reaction can occur intramolecularly if the diene and alkene are connected by an intervening chain of atoms,163 illustrated by the conversion of 106–107. This version of the reaction has become one of the most powerful synthetic methods (Section 8.5.2) for constructing cyclic compounds with high regio- and 155

Jurczak, J.; Chmielewski, M.; Filipek, S. Synthesis 1979, 41.

156

Darling, S. D.; Subramanian, N. J. Org. Chem. 1975, 40, 2851.

(a) vor der Br€ uck, D.; B€ uhler, R.; H€ uck, C.-Ch.; Plieninger, H.; Weale, K. E.; Westphal, J.; Wild, D. Chem.-Ztg. 1970, 94, 183 [Chem. Abstr. 1970, 73, 3128r]. (b) McCabe, Jr.; Eckert, C. A. Acc. Chem. Res. 1974, 7, 251. 157

158

Isaacs, N. S.; der Beeke, P. V. Tetrahedron Lett. 1982, 23, 2147.

159

(a) Kl€ arner, F.-G.; Diedrich, M. K.; Wigger, A. E. Chemistry Under Extreme or Non-Classical Conditions, van Eldik, R.; Hubbard, C. D., Eds.; John Wiley & Sons, Inc.: NY, 1997; pp 103–161. (b) Jurczak, J.; Gryko, D. T. Chemistry Under Extreme or Non-Classical Conditions, van Eldik, R.; Hubbard, C. D., Eds.; John Wiley & Sons, Inc.: NY, 1997; pp 163–188. (c) Ibata, T. Organic Synthesis at High Pressures, Matsumoto, K.; Archeson, R. M., Eds.; John Wiley & Sons, Inc.: NY, 1991; pp 409–422. (e) Diedrich, M. K.; Kl€arner, F.-G. J. Am. Chem. Soc. 1998, 120, 6212. (f) Ciobanu, M.; Matsumoto, K. Liebigs Ann.-Recl. 1997, 623.

160

Plietker, B.; Seng, D.; Fr€ olich, R.; Metz, P. Tetrahedron 2000, 56, 873.

(a) Renslo, A. R.; Weinstein, R. D.; Tester, J. W.; Danheiser, R. L. J. Org. Chem. 1997, 62, 4530. (b) Weinstein, R. D.; Renso, A. R.; Danheiser, R. L.; Harris, J. G.; Tester, J. W. J. Phys. Chem. 1996, 100, 12337. (c) Hyatt, J. A. J. Org. Chem. 1984, 49, 5097. (d) Paulaitis, M. E.; Alexander, G. C. Pure Appl. Chem. 1987, 59, 61. (e) Ikushima, Y.; Saito, N.; Arai, M. Bull. Chem. Soc. Jpn. 1991, 64, 282. (f) Isaacs, N. S.; Keating, N. J. Chem. Soc. Chem. Commun. 1992, 876.

161

(a) Matsuo, J.-i.; Tsuchiya, T.; Odashima, K.; Kobayashi, S. Chem. Lett. 2000, 178. (b) Oakes, R. S.; Heppenstall, T. J.; Shezad, N.; Clifford, A. A.; Rayner, C. M. Chem. Commun. 1999, 1459.

162

163

See Takao, K.-i.; Munakata, R.; Tadano, K.-i. Chem. Rev. 2005, 105, 4779.

774

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

stereoselectivity.164 There is much interest in intramolecular Diels-Alder reactions and many reviews, including those by Carlson,165 Oppolzer,166 Brieger and Bennett,167 Fallis,168 and Smith.169 More recent reviews include those by Juhl and Tanner,170 Brocksom et al.,171 Taber,172 Heravi and Vavsari,173 and that by Hua and coworkers.174 Virtually all of the intermolecular reactions shown in previous sections have intramolecular analogues. The atoms that link the diene and alkene (the tether) can sometimes inhibit close approach of those moieties. Indeed, the ground-state conformation of a molecule can greatly influence the reaction.

106

107

Padwa and coworker’s175 studied the intramolecular Diels-Alder reaction of N-alkenyl amidofurans, and found that placement of a carbonyl in the tether can cause the molecule to adopt a conformation that is closer to or further from the reactive conformation. The geometry of the diene and alkene (E) or (Z) also plays a role, which can be illustrated by comparing the intramolecular Diels-Alder reactions of 108A and 108B. trans-Diene 108A readily attains the proper geometry for reaction (see 108B) with the methylene groups in the tether assuming a relatively normal conformation for a small ring. cis-Diene 109A, however, requires significant distortion of the methylene groups in the tether to even approach the proper orientation for the cycloaddition. In model 109B, severe distortion is apparent in the tether, where one or more bonds must be elongated to achieve the conformation shown. Such bond elongation and distortion of the tether is energetically unfavorable for this molecule.

H H

108A

108B

108A

108B

The length of the tether is very important.176,177 Boeckman and Demko178 showed that Diels-Alder cyclization of 109 gave a 50% yield of a 70:30 mixture (βH:αH ¼ cis/trans) of 110. Attempted cyclization of 111, which has fewer carbons in the tether, gave no cyclized product. The diene in this latter example shows a preference for the s-trans conformation, which has a higher energy of activation for interconversion of the two rotamers than 109.178,179a 164

Fallis, A. G. Acc. Chem. Res. 1999, 32, 464.

165

Carlson, R. G. Ann. Rep. Med. Chem. 1974, 9, 270.

166

Oppolzer, W. Angew. Chem. Int. Ed. Engl. 1977, 16, 10 (see pp 10–18).

167

Brieger, G.; Bennett, J. N. Chem. Rev. 1980, 80, 63 (see p 67).

168

Fallis, A. G. Can. J. Chem. 1984, 62, 183.

169

Smith, M. B. Org. Prep. Proc. Int. 1990, 22, 315.

170

Juhl, M.; Tanner, D. Chem. Soc. Rev. 2009, 38, 2983.

171

Brocksom, T. J.; Nakamura, J.; Ferreira, M. L.; Brocksom, U. J. Braz. Chem. Soc. 2001, 12, 597.

172

Taber, D. F. Org. Chem. Highlights 2007, August 13. Available at: http://www.organic-chemistry.org/Highlights/2007/13August.shtm.

173

Heravi, M..; Vavsari, V. F. RSC Adv. 2015, 5, 50890.

174

ChunYun, W.; Jun, D.; Hua, L.; Ming, B.; Ang, L. Science China Chem. 2014 57, 926.

175

Bur, S. K.; Lynch, S. M.; Padwa, A. Org. Lett. 2002, 4, 473.

176

Kraus, G. A.; Bougie, D.; Jacobson, R. A.; Su, Y. J. Org. Chem. 1989, 54, 2425.

177

Tagmazyan, K.Ts.; Mkrtchyan, R. S.; Babayan, A. T. Zh. Org., Khim. 1974, 10, 1642. (Engl. p 1657).

178

Boeckman, Jr., R. K.; Demko D. M. J. Org. Chem. 1982, 47, 1789.

179

(a) For a related reaction with furan derivatives, see Parker, K. A.; Adamchuck, M. R. Tetrahedron Lett. 1978, 1689. (b) Reference 81b, p 167.

775

14.7 INTRAMOLECULAR DIELS-ALDER REACTIONS

CO2Me

170°C 22 h

CO2Me

H

O 109

O 110 (50%)

240°C

CO2Me

No reaction

O 111

The influence of the tether length is particularly striking in the substituted cyclopenta-1,3-dienes examined by Stille and Grubbs.180 Each cyclopenta-1,3-diene derivative is in equilibrium with three isomers, the 1-substituted, 2substituted, and 3-substituted cyclopenta-1,3-dienes. A two-carbon tether (n ¼ 1) in 112 gave a 53% yield of 115 when heated in benzene.180 The 2-substituted derivative (113) and 3-substituted derivative (114) gave no cycloadduct (116 and 117, respectively) where n ¼ 1. Presumably, this result is due to the difficulty in forming the four-membered ring in 116, or attaining the proper geometry for the bridgehead π-bond in 117 (see Bredt’s rule in Section 10.3.2). The endotransition state required to generate 115 is relatively easy to achieve. When one carbon is added to the tether (112, n ¼ 2), however, heating led to a 6.4:1 mixture of 116 and 117 (n ¼ 2).180 The cyclization of 112 ! 115 (n ¼ 2) was not observed. The 2-substituted derivative gave the highest yield of cyclized product when n ¼ 2.180 Adding a total of four carbons to the tether gave results similar to when n ¼ 2 in 112–114.180 These examples clearly show that the length and nature of the tether is important, and inspection of all possible transition states is essential for predicting the cycloadducts arising from intramolecular Diels-Alder reactions.

(CH 2)n

Ot-Bu 112

O

O

O

Ot-Bu

(CH 2)n

113

Ot-Bu (CH 2)n

114

0% when n = 2 0% when n = 1

(CH 2)n

0% when n = 1

(CH 2)n

CO2t-Bu

CO2t-Bu

115 (53%)

116

CO2t-Bu

(CH 2)n

117

Reprinted with permission from Stille, J. R.; Grubbs, R. H. J. Org. Chem. 1989, 54, 434. Copyright © 1989 American Chemical Society.

In addition to the problems concerning reactivity and the nature of the tether, at least two different regioisomers are possible for an internal Diels-Alder reaction of a triene (e.g., (E)-octadeca-1,3,17-triene, 118). Cyclization of 118 can yield either regioisomer, 119 or 120, or both can form depending on the orientation of the alkene as it approaches the diene. If the tether is short, only 119 is possible, but with longer tethers, 120 can be a side product or even the major product. 180

Stille, J. R.; Grubbs, R. H. J. Org. Chem. 1989, 54, 434.

776

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

•

• •

•

• •

•

•

• 119

•

•

•

118

120

In the Thomas and coworker’s studies181 on the synthesis of cytochalasin, an intramolecular Diels-Alder reaction of 121 gave a 27:5% mixture of 122 and 123. In general, (E)-dienes do not give products (e.g., 119), although (Z)-dienes may do so. As the reactive termini approach, the alkene arm can assume an exo- or an endo-mode. The exo-transition state leads to the trans-product, and the endo-transition state leads to the cis-product. The final outcome depends on the conformation assumed in the transition state, based on the structures of the reactive species (Section 1.5.4).167 In general, the exo-transition state is preferred. O

O O

O

O

O

O

O

O

+

O O

O

O

Me

Me

O O

Me 121

122

123

The length of the tether, reaction conditions, and catalyst, if any, will influence the cis/trans ratio. Houk and coworkers182 compared the cyclization of trienes (124) with several tether lengths to give the bicyclic cycloadduct 125. The cis-adduct was preferred except when n ¼ 5 (to form a nine-membered ring). Gras and Bertrand183 initially reported that when n ¼ 2 the trans-product was preferred (later corrected), but Houk found that the cis-product was preferred. When n ¼ 4–7, the product yields were decreased, reflecting the difficulty in generating the transition state for an 8- to 11-membered ring in 125 (Sections 1.5.2 and 1.5.3 and 4.5.2). O

Ha

O (CH 2)n (CH 2)n

H 124

125

Many applications of the intramolecular Diels-Alder reaction to natural product synthesis are known. In a synthesis of (
)-calyciphylline N, Smith and Shvartsbart184 stirred 126 at 0°C with the Lewis-acid catalyst Et2AlCl (see Section 14.6.1), and obtained 127 in >50% yield, with a 9:1 dr. Using an intramolecular Diels-Alder strategy that involved a transannular cyclization, Roush and coworkers185 converted 128–129, in >33% yield, in a synthesis of (+)-superstolide A.

181

Bailey, S. J.; Thomas, E. J.; Turner, W. B.; Jarvis, J. A. J. Chem. Soc. Chem. Commun. 1978, 474.

182

Smith, D. A.; Sakan, K.; Houk, K. N. Tetrahedron Lett. 1986, 27, 4877.

183

Gras, J.-L.; Bertrand, M. Tetrahedron Lett. 1979, 4549.

184

Shvartsbart, A.; Smith, III, A. B. J. Am. Chem. Soc. 2014, 136, 870.

185

Tortosa, M.; Yakelis, N. A.; Roush, W. R. J. Org. Chem. 2008, 73, 9657.

777

14.7 INTRAMOLECULAR DIELS-ALDER REACTIONS

Me Me

Si

O Et2AlCl, Toluene

O

0°C

Si

Me

CO2Et 126

CO2Et

127 ( >50%, 9:1 dr)

OTBDPS

Me

Me

OTBDPS H Me Me

MeO

O

O

Me

Me

H

O N

Boc

O Me

MeO

2h

N

Me

Me

80°C, Toluene

O

Me

Me

O

128

Me

Boc

Me

129 (>33%)

Even when a single substrate is used, the ratio of diastereomeric products can vary with reaction conditions. In the Hart et al.186 synthesis of (+)-himbeline and (+)-himbacine, 130 (R ¼ CO2Et) was cyclized to a mixture of 131 and 132. At higher temperatures, the reaction favored 132, but with an acid promoter at lower temperatures, 131 was favored. When the ester group in 130 was changed to the reduced form (a TBS protected alcohol, R ¼ CH2OTBS), heating to 110°C gave a 1:1 mixture, whereas the acid promoted lower temperature reaction gave a 20:1 mixture favoring 131. Me

H

R

R

H

PhMe

Me O

O

H

R

H

+

Me O

Heat

130

H

O

131

H

O

O

132

In the cyclization of dienyl amides, Oppolzer et al.187 noted a difference in regiochemistry between dienes having the carbonyl external to the ring being formed (133 ! 134),188 and dienes having the carbonyl in the ring being formed (135 ! 136 + 137).187a This difference was rationalized by a preference for the transition state that maximized overlap of the amide π-orbitals, and the diene π-orbitals. This secondary orbital interaction is analogous to the endo-effect noted in Section 14.5.3.3. CO2Me MeO2C PhMe, 210°C, 14 h Sealed tube

N

N

CO2Me

CO2Me 133

134 (60%)

190°C, 24 h Sealed tube

N

N

O

+

N

C3H7

O

C3H7

O 135

136 (62%)

137

186

Hart, D. J.; Li, J.; Wu, W.-L.; Kozikowski, A. P. J. Org. Chem. 1997, 62, 5023.

187

(a) Oppolzer, W.; Fr€ ostl, W. Helv. Chim. Acta 1975, 58, 590. (b) Oppolzer, W.; Fr€ ostl, W.; Weber, H. P. Helv. Chim. Acta 1975, 58, 593.

188

Witiak, D. T.; Tomita, K.; Patch, R. J. J. Med. Chem. 1981, 24, 788.

778

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

The concept of tethered intramolecular reactions has become popular. The basis of this concept is that the diene and alkene units are linked together by a species that allows an intramolecular Diels-Alder, but can be chemically removed later. In the work of Batey et al.,189a reaction of (E)-styrylboronic acid with sorbic alcohol [(2E,4E)-hexa2,4-dien-1-ol] gave 138, and heating in a sealed tube at 190°C in the presence of the radical inhibitor BHT (2,6-ditert-butyl-4-phenylphenol) gave the cycloadduct 139. Oxidation with trimethylamine-N-oxide and hydrolysis gave a 90:10 mixture of 140/141 in 84% yield. The boronic ester linkage in 138 was a tether leading to the regio- and stereoselectivity of an intramolecular process. Treatment with the amine oxide and hydrolysis removed the tether to give the cycloadducts. In a synthesis of (+)-tetronolide by Boeckman et al.,189b an acetal tether was used to anchor an intramolecular Diels-Alder reaction. Ph Ph

OH B(OH)2

+

Me

B OH

THF

O

MS 4 Å

Me (E)-Styrylboronic acid

(2E,4 E)-Hexa-2,4dien-1-ol

138

HO Ph

B

190°C PhMe, 5% BHT

Ph O

+

Me

2. H2O, 60°C

Me

Ph

OH

1. Me3NO, PhH, 80°C

OH

139

140

OH

Me

OH 141

Benzyne derivatives can undergo Diels-Alder reactions. In a synthesis of arnottin I, Madan and Cheng190 treated o-dibromide 142 with n-butyllithium in THF at
78°C and then with furan and obtained cycloadduct 145 in 52% yield. Aryl bromine-lithium exchange yields 143, which loses LiBr to give the benzyne, 144, which subsequently traps furan in a Diels-Alder reaction to yield 145. TBDMSO

Br

TBDMSO

Br

BuLi THF

TBDMSO

Li

TBDMSO

Br

–78°C rt

142

143

TBDMSO

TBDMSO O

TBDMSO

O TBDMSO

144

145 (52%)

Another example that utilized a benzyne intermediate is taken from a synthesis of isokidamycin, reported by Martin and coworkers.191 Treatment of aryl dibromide 146 with n-butyllithium generated the benzyne intermediate. A subsequent intramolecular Diels-Alder reaction with the furan moiety led to 147 in 92% yield. In this synthesis, the silicon tether was cleaved in a second step by heating with TBAF in DMF. Note that a silicon tether was also used for the conversion of 126–129.

(a) Batey, R. A.; Thadani, A. N.; Lough, A. J. J. Am. Chem. Soc. 1999, 121, 450. (b) Boeckman, R. K., Jr.; Shao, P.; Wrobleski, S. T.; Boehmler, D. J.; Heintzelman, G. R.; Barbosa, A. J. J. Am. Chem. Soc. 2006, 128, 10572.

189

190

Madan, S.; Cheng, C.-H. J. Org. Chem 2006, 71, 8312.

191

O’Keefe, B. M.; Mans, D. M.; Kaelin, Jr., D. E.; Martin, S. F. J. Am. Chem. Soc. 2010, 132, 15528.

779

14.8 INVERSE ELECTRON DEMAND AND THE RETRO-DIELS-ALDER REACTIONS

Me Me Si

Me

Me

O

OBn

Si O Me

O

O

OBn

O BuLi, THF, –25°C

Br

BnO

Br

Me

NMeBoc

OMe

OMe

O

Me Me

NMeBoc OBn

146

147 (92%)

There are many synthetic examples of carbocyclic, intramolecular Diels-Alder reactions that involve a heteroatom unit that are part of the diene or part of the alkene (also see Section 14.9). An example is taken from Boger and coworker’s192 synthesis of vindorosine. Amide (148) contains an indole moiety, as well as an oxadiazole, and heating in xylene gave a 72% yield of 149. OMOM Et

O

OMOM

O OBn

N

N Xylene, 150°C

N O N CO2Me

Me 148

Et

O

10 h

N

N Me

H

OBn CO2Me

149 (72%)

14.8 INVERSE ELECTRON DEMAND AND THE RETRO-DIELS-ALDER REACTIONS There is useful variation for the Diels-Alder reaction that manipulates heteroatoms and/or substituents on the diene and/or alkene units to modify the HOMO-LUMO interactions that drive the reaction. In this variation, the reaction is driven by the LUMOdiene-HOMOalkene interaction rather than the HOMOdiene-LUMOalkene interaction. A second variation uses the reversible nature of the Diels-Alder reaction. Manipulation of the cycloadducts followed by a retroDiels-Alder reaction gives a new diene or alkene. Both variations will be examined in this section.

14.8.1 Inverse Electron Demand Diels-Alder Reactions In a “normal” Diels-Alder reaction of buta-1,3-diene and acrolein, the HOMOdiene-LUMOalkene interaction is lower than the comparable LUMOdiene-HOMOalkene interaction. In other words, the activation energy for the former interaction is significantly lower than that for the latter reaction. However, the LUMOdiene-HOMOalkene interaction for buta1,3-diene and the electron-rich methyl vinyl ether is lower than the comparable HOMOdiene-LUMOalkene interaction. Again, the activation energy for the former reaction is lower than that for the latter reaction and should lead to the transition state for the major product. When a Diels-Alder reaction occurs by the LUMOdiene-HOMOalkene interaction, it is usually referred to as an inverse electron demand Diels-Alder reaction. An example taken from a synthesis of vindoline by Boger and coworkers193 heated oxadiazole 150, which gave a 72% yield of 151 via an intramolecular Diels-Alder reaction. In general, an electron-rich alkene will react faster than an electron-poor alkene, the opposite of what is normally observed (see reactions in Section 14.5).

192

Sasaki, Y.; Kato, D.; Boger, D. L. J. Am. Chem. Soc. 2010, 132, 13533.

193

Kato, D.; Sasaki, Y.; Boger, D. L. J. Am. Chem. Soc. 2010, 132, 3685.

780

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

OMOM

O

O OBn

N N

N Xylene, 150°C 10 h

N

O

O N

N

Me

OMOM

OBn

Me

MeO2C

CO2Me

150

151 (72%)

In another example, Boger and Kasper194 prepared 1-azadiene (152, azadienes are discussed in Section 14.9.3.5) and showed that the LUMO of the diene controlled the subsequent Diels-Alder reaction. Reaction of 152 with butyl vinyl ether (n-BuOCH ¼ CH2) gave a 74% yield of 153 (>20:1 endo/exo).194 This reaction was sluggish, and high pressure (Section 14.6.3) was required to force the reaction to product with a reasonable reaction time and in good yield. Boger and coworkers195 showed that tetrazines react with alkynes with inverse electron demand to yield 1,2-diazines, after loss of nitrogen from the original Diels-Alder cycloadduct. It is apparent from these few examples that the structures of most reverse electron demand Diels-Alder reactions involve heteroatoms, or another unique feature. Me

N

SO2Ph SO2Ph

Me

OBn

N

Neat , 12 kbar , 70 h (5 equiv) OBn

H 153 (74%)

152

Inverse electron demand Diels-Alder reactions are less prevalent than the normal electron demand Diels-Alder reactions discussed elsewhere in this chapter. Nonetheless, these reactions are useful in organic synthesis. The preparation of 151 is one example. Another example is taken from Boger and Oakdale’s196 synthesis of lycogalic acid in which 1,2bis(tributylstanyl)acetylene was heated with 1,2,4,5-tetrazinedicarboxylic ester 154, a useful cycloaddition partner,197 to yield cycloadduct 155, which lost nitrogen (N^N) to give 156 in 97% yield. The stannyl moieties were converted to an indole moiety in subsequent steps. N N N

Dioxane, 45°C

MeO2C

Bu3Sn

CO2Me Bu3Sn

N N

N

SnBu3

Bu3SnMeO C 2

154

N

CO2Me

N

– N2

N

SnBu3 156 (97%)

O

OMe

O MgBr2 , THF, rt

+

157

CO2Me

Bu3Sn

155

TBSO

N

MeO2C

H OBn (S)-2-(Benzyloxy)propanal

BnO

O 158 (75%)

Enantioselective inverse-electron demand hetero-Diels-Alder reactions have been reported. Webb and coworkers,198 in a synthesis of herboxidiene, reported that diene 157 reacted with chiral aldehyde (S)-2-(benzyloxy)propanal, catalyzed by magnesium bromide, to yield the Diels-Alder adduct 158 in 75% overall yield. Jacobsen 194

Boger, D. L.; Kasper, A. M. J. Am. Chem. Soc. 1989, 111, 1517.

195

Soenen, D. R.; Zimpleman, J. M.; Boger, D. L. J. Org. Chem. 2003, 68, 3593.

196

Oakdale, J. S.; Boger, D. L. Org. Lett. 2010, 12, 1132.

197

see (a) Fu, L.; Gribble, G. W. Tetrahedron Lett. 2010, 51, 537;(b) Boger, D. L.; Panek, J. S.; Patel, M. Org. Synth. 1992, 70, 79.

198

Lagisetti, C.; Yermolina, M. V.; Sharma, L. K.; Palacios, G.; Prigaro, B. J.; Webb, T. R. ACS Chem. Biol. 2014, 9, 643.

781

14.8 INVERSE ELECTRON DEMAND AND THE RETRO-DIELS-ALDER REACTIONS

and coworkers199 reported that Cr complex 159 catalyzed the inverse electron Diels-Alder reaction of 10 equiv. of ethyl vinyl ether and (E)-but-2-enal, to give pyran 160, in 75% yield and 94 %ee.

Neat, MS 4 Å, 1 d

+ H

N

5%

(10 equiv)

Cr O O Cl

O

(E)-But-2-enal

OEt

O

159

OEt 160 (75%)

A typical inverse electron demand disconnection follows: X

X R1

R

R1

+

R

Y

Y

14.8.2 Retro-Diels-Alder Reactions

25°C

+

H H

160–240°C

161

The Diels-Alder reaction is reversible under certain circumstances.200 A classical example of a retro Diels-Alder reaction is the self-condensation of cyclopenta-1,3-diene to yield cyclopenta-1,3-dienyl dimer 163. This reaction occurs quickly at temperatures >25°C, but is slow at low temperatures (e.g.,
78°C. Many Diels-Alder reactions involving cyclopenta-1,3-diene are done at low temperatures in the presence of Lewis-acid catalysts (Section 14.6.1) to suppress cyclopenta-1,3-diene dimerization. When heated to 160–240°C, a retro-Diels-Alder reaction of 161 gives 2 equiv. of monomeric cyclopenta-1,3-diene.201 The reversibility of the Diels-Alder reaction can be exploited synthetically with the correct choice of diene, dienophile, and reaction conditions because the equilibrium usually favors the thermodynamically more stable cycloadduct as the major product. Woodward and Baer202 showed that maleic anhydride (furan-2,5-dione) reacted with fulvalene (cyclopenta-2,4-dien-1-ylidenecyclohexane) at ambient temperatures to yield 163, the endo-product. When heated to 50°C for 10 min, however, 162 was the major product with only 20.5% of 163, and this yield dropped to 3.1% upon heating at 50°C for 60 min.201 This result is consistent with a reversible reaction pathway favoring the more stable exoproduct.201 This reverse reaction (cycloadduct ! diene + dienophile) is called a retro Diels-Alder reaction.203

O

H

O O H

O

rt

Reflux

+

H

O – PhH

– PhH

O

O

H 162

O

Fulvalene

Furan-2,5-dione

163

O

199

Gademann, K.; Chavez, D. E.; Jacobsen, E. N. Angew. Chem. Int. Ed. 2002, 41, 3059.

200

(a) Kwart, H.; King, K. Chem. Rev. 1968, 68, 415. (b) Smith, G. G.; Kelly, F. W. Progr. Phys.Org. Chem. 1971, 8, 75 (see p 201).

201

(a) Moffett, R. B. Org. Synth. Coll. Vol. 4 1963, 238. (b) Korach, M.; Nielsen, D. R.; Rideout, W. H. Org. Synth. Coll. Vol. 5 1973, 414.

202

Woodward, R. B.; Baer, H. J. Am. Chem. Soc. 1944, 66, 645.

(a) Ripoll, J. L.; Rouessac, A.; Rouessac, F. Tetrahedron 1978, 34, 19. (b) Lasne, M.-C.; Ripoll, J. L. Synthesis 1985, 121. (c) Ichihara, A. Synthesis 1987, 207. (c) Rickborn, B. Org. React. 1998, 52, 1. (d) Rickborn, B. Org. React. 1998, 53, 223. (e) Klunder, A. J. H.; Zhu, J.; Zwanenburg, B. Chem. Rev. 1999, 99, 1163. (f) Stajer, G.; Csende, F.; Fueloep, F. Curr. Org. Chem. 2003, 7, 1423.

203

782

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

Most of the synthetic applications of retro-Diels-Alder reactions involve cyclopentadiene or furan diene partners, but alkyne dienophiles are common and there are many potential substrates. An example is Beaudry and Zhao’s204 synthesis of cavicularin in which pyrone (164) was heated under microwave irradiation to yield cycloadduct 165, which lost PhOSOH to generate alkene 166. Subsequent heating allowed loss of CO2 via a retro-Diels-Alder reaction to give an 83% yield of 167. O

O MeO

O MeO

O

O

O

PhO2S O

O

– PhOSOH

Microwaves 240°C, 8 h

MeO

MeO MeO

MeO O

MeO

MeO

MeO 164

165

166

MeO O

– CO2

MeO MeO 167 (83%)

The retro-Diels-Alder reaction usually requires higher temperatures than the “forward” Diels-Alder reaction, so the “forward” Diels-Alder product is usually obtained without competition from the retro-reaction. When the retro-Diels-Alder reaction is desired, flash vacuum pyrolysis is a common technique used in synthesis.205 Lewis acids also catalyze retro-Diels-Alder reactions.206 In a synthesis of pentenocin B, Sugahara et al.207 utilized a retro-DielsAlder reaction beginning with the 3a,4,7,7a-tetrahydro-1H-4,7-methanoinden-1-one ketodicyclopentadiene. Compound 3a,4,7,7a-tetrahydro-1H-4,7-methanoinden-1-one was converted, in 10 steps, to the highly functionalized compound 168. When 168 was heated to 280°C, the retro-Diels-Alder reaction liberated cyclopenta-1,3-diene and the targeted conjugated ketone 169 in 93% yield. H H

H

H

O

O O

10 Steps

O

O

O

Ph2O, 15 min 280°C

O

OMOM 3a,4,7,7a-Tetrahydro-1H-4,7methanoinden-1-one

168

204

Zhao, P.; Beaudry, C. M. Org. Lett. 2013, 15, 402.

205

Stork, G.; Nelson, G. L.; Rouessac, F.; Gringore, O. J. Am. Chem. Soc. 1971, 93, 3091.

206

Grieco, P. A.; Abood, N. J. Org. Chem. 1989, 54, 6008.

207

Sugahara, T.; Fukuda, H.; Iwabuchi, Y. J. Org. Chem. 2004, 69, 1744.

OMOM 169

(93%)

783

14.9 HETEROATOM DIELS-ALDER REACTIONS

A variety of interesting and useful products are available by using the Diels-Alder reaction to insert functionality and the retro-Diels-Alder to transform that product into another. The retro-Diels-Alder reaction therefore leads to several interesting disconnections, including the following: R

R1

R1

R R1

R R

R1

R

+

O

O R1

O

R

O

O

R1

O

+

O

O

+ O

O O

O

14.9 HETEROATOM DIELS-ALDER REACTIONS The Diels-Alder reaction is quite remarkable in its scope. One of the more attractive features is the ability to incorporate heteroatoms into either the diene, or the alkene unit, generating heterocyclic compounds. This section will examine several of the more useful variations.

14.9.1 Heteroatom Substituents In previous sections, several examples were shown using alkoxydienes or alkoxyalkenes (vinyl ethers). Alkoxy dienes are particularly useful in the synthesis of pyran derivatives. One example is Li and coworker’s208 synthesis of aphadilactone A-D, which reacted a methoxydiene,(E)-1-methoxy-3-methylbuta-1,3-diene, with but-2-ynal in the presence of a chiral chromium catalyst to give 170 in 84% yield and 98 %ee. Me Me

Me

+

cat Chiral Cr complex

H

O

MS 4 Å, rt, 16 h

OMe (E)-1-Methoxy-3-methylbuta-1,3-diene

O OMe But-2-ynal

170 (84%, 98 %ee)

Danishefsky209 used a variety of silyloxybuta-1,3-dienes in Diels-Alder reactions209 in place of alkoxy and acetoxy derivatives. The preparation of silyl enol ethers is usually easier than the preparation of the corresponding enol ether.210 Treatment of the resulting silyloxy cycloadducts with fluoride ion (Section 5.3.1.2) or simply with aqueous acid will generate an enol, which tautomerizes to the carbonyl under very mild conditions. In Poisson and coworker’s211 synthesis of (
)-kainic acid A, proline derivative 171 reacted with diene 172 to yield 173, under high-pressure conditions, as a 3:1 endo/exo mixture. Subsequent reaction with aqueous potassium hydrogen sulfate led to hydrolysis of the silyl enol ether to yield the ketone moiety, and induced elimination of the methoxy group to give a 62% overall yield of 175, along with a 23% yield of the partial hydrolysis product 174. Treatment of 174 with DBU induced elimination to give an 81% yield of 175. 208

Yin, J.-P.; Gu, M.; Li, Y.; Nan, F.-J. J. Org. Chem. 2014, 79, 6294.

209

(a) Danishefsky, S. Acc. Chem. Res. 1981, 14, 400. (b) Danishefsky, S. Aldrichimica Acta 1986, 19, 59.

(a) Stork, G.; Hudrlik, P. J. Am. Chem. Soc. 1968, 90, 4462. (b) House, H. O.; Czuba, L.; Gall, M.; Olmstead, H. D. J. Org. Chem. 1969, 34, 2324. (c) House, H. O.; Gall, M.; Olmstead, H. D. J. Org. Chem. 1971, 36, 2361.

210

211

Orellana, A.; Pandey, S. K.; Carret, S.; Greene, A. E.; Poisson, J.-F. J. Org. Chem. 2012, 77, 5286.

784

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

Diene 172 is known as Danishefsky’s diene.212 This diene, or analogues, have been very useful for a variety of natural product syntheses.212a As shown, an attractive feature of dienes that contain a vinyl ether or vinyl ester moiety is the facile hydrolysis of the cycloadduct (the hydrolysis of vinyl ethers was previously shown with the Birch reduction (Section 7.11.5). OTMS

OTMS

MeO2C

MeO

172 N

CO2Me DCM, 15 kbar

MeO MeO2C

H N

Boc

CO2Me

Boc 171

173

DBU, Toluene

O

O

(81%) aq KHSO4 THF

MeO MeO2C

H

MeO2C

H

CO2Me

N

CO2Me

N

Boc

Boc 175 (62%, 90 %ee)

174 (23%)

Many structural variations are possible for the alkoxybuta-1,3-diene partner in the Diels-Alder reaction. Chiral hetero-substituted buta-1,3-dienes213 are known in [4+2]-cycloaddition reactions and chiral dienophiles are used. Avenoza et al.214 used an asymmetric hetero-Diels-Alder reaction in a synthesis of carbacephams. Chiral nonracemic imine 176215 reacted with 172 to yield a 17:83 mixture of 177:178, in 65% yield. Chiral Diels-Alder reactions will be discussed in Section 14.10. OSiMe3

Boc

Boc 1. MeO

N O

172

N

O

CH2Cl2, Et2AlCl –40°C, 17 h 2. HCl

NBn

Boc

N O

O

O

N

Bn

176

+

Bn

177

N 178

(65%, 17:83)

It is clear from these examples that oxygenated dienes are convenient and useful precursors to cyclohexenones and 7-oxabicycloheptane derivatives. Typical disconnections for oxygenated dienes follows: O

R1

R2O

R1

O

R2

+

R R

R2

OR2

R

R

R1

R2 R

R1 +

R2

O

R

14.9.2 Amino and Amido dienes Azadienes (amino or amido dienes) constitute another important class of hetero-substituted dienes. There are four major types, amino dienes (e.g., 179), amido dienes (e.g., 180), carbamoyl dienes (e.g., 181), and pyrrole derivatives (a) Danishefsky, S.; Kitahara, T.; Schuda, P. F.; Etheredge, S. J. J. Am. Chem. Soc. 1976, 98, 3028. (b) Danishefsky, S.; Kitahara, T.; McKee, R.; Schuda, P. F. J. Am. Chem. Soc. 1976, 98, 6715. (c) Danishefsky, S.; Schuda, P. F.; Kitahara, T. Etheredge, S. J. J. Am. Chem. Soc. 1977, 99, 6066. (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, Jr., F. G. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; p 753.

212

213

Barluenga, J.; Suárez-Sobrino, A.; López, L. A. Aldrichimica Acta 1999, 32, 4.

214

Avenoza, A.; Busto, J. H.; Cativiela, C.; Corzana, F.; Peregrina, J. M.; Zurbano, M. M. J. Org. Chem. 2002, 67, 598.

(a) Langenbeck. W.; G€ odde, O.; Weschky, L.; Schaller, R. Chem. Ber. 1942, 75, 232; also see (b) H€ unig, S.; Kahanek, H. C. Chem. Ber. 1957, 90, 238. (c) Farmer, M. L.; Billups, W. E.; Greenlee, R. E.; Kurtz, A. N. J. Org. Chem. 1966, 31, 2885. (d) Mannich, C.; Handke, K.; Roth, K. Chem. Ber. 1936, 69, 2112. 215

785

14.9 HETEROATOM DIELS-ALDER REACTIONS

(e.g., 182). Both 1- and 2-substituted amino dienes are possible, and they function as diene partners in Diels-Alder reactions. RO

R R

O

N

N

R

N

N R1

179

R O

R1 181

180

182

There are a few examples of Diels-Alder reactions with 1-(dialkylamino)buta-1,3-dienes.215 Note that a density functional study of the Diels-Alder reaction of 1,2-diazabuta-1,3-diene with alkenes showed the transition state to be concerted, but asynchronous.216 The allylic amine cycloadducts of this reaction show a propensity for elimination, which has led to the use of dienyl amides (180) and carbamates (181), whose cycloadducts are significantly more stable and easier to isolate.215 An example that used a chiral amido diene was reported by Takayama and coworkers217 in a synthesis of voacangalactone, where 183 reacted with the alkylidene malonate dimethyl 2-methylenemalonate to give an 81% yield of 184 as the single diastereomer shown. Amido diene 183 was prepared by a Cu(I) catalyzed coupling with a vinyl iodide, but amido dienes can also be prepared from conjugated aldehydes with an amine, a base, and an acid chloride.218 The synthesis of 149 in Section 14.7 showed the use of furan as an enophile. Pyrrole derivatives can be used also. In a synthesis of asparagamine A, Overman and coworker’s219 showed that pyrrole 185 reacted with a nitro-acrylate, ethyl (E)-3-nitroacrylate, to yield 186, which was hydrogenated to give 187 in 73% yield. OTBS

O N

Ph

MeO2C

O

Ph OTBS

CO2Me

O

DCM, rt, 60 h

N

Ph O MeO C CO2Me 2

Ph 183

184 (81%)

Boc Boc N OH OMe 185

CO2Et

O2N

Boc

O2N N

H2, Pd/C

O2N N

EtOAc, rt

rt

OMe

EtO2C

EtO2C

186

OMe 187 (73%)

Overman prepared a dienyl carbamate, benzyl (E)-buta-1,3-dien-1-ylcarbamate, via a Curtius rearrangement220 (a thermal rearrangement of a diazoketone to an isocyanate; dienyl azido-ketone (E)-penta-2,4-dienoyl azide ! (E)1-isocyanatobuta-1,3-diene; see Section 4.4.3). Heating with benzyl alcohol gave benzyl (E)-buta-1,3-dien-1-ylcarbamate.221 Overman et al.222,223 also found thermolysis of trichloroacetimidic esters of propargyl alcohols generated trichloromethyl dienyl carbamates. Dienyl carbamate reacted with styrene to give a quantitative yield of a mixture

216

Avalos, M.; Babiano, R.; Clemente, F. R.; Cintas, P.; Gordillo, R.; Jimenez, J. L.; Palacios, J. C. J. Org. Chem. 2000, 65, 8251.

217

Harada, M.; Asaba, K. N.; Iwai, M.; Kogure, N.; Kitajima, M.; Takayama, H. Org. Lett. 2012, 14, 5800.

218

(a) Oppolzer, W.; Fr€ ostl, W. Helv. Chim. Acta 1975, 58, 587. (b) Oppolzer, W.; Bieber, L.; Francotte, E. Tetrahedron Lett. 1979, 981.

219

Br€ uggemann, M.; McDonald, A. I.; Overman, L. E.; Rosen, M. D.; Schwink, L. Scott, J. P. J. Am. Chem. Soc. 2003, 125, 15284.

220

(a) Curtius, T. J. Prakt. Chem. 1894, 50, 275. (b) Smith, P. A. S. Org. React. 1946, 3, 337. (c) Saunders, J. H.; Slocombe, R. J. Chem. Rev. 1948, 43, 203.

(a) Overman, L. E.; Taylor, G. F.; Jessup, P. J. Tetrahedron Lett. 1976, 3089. (b) Overman, L. E.; Taylor, G. F.; Petty, C. B.; Jessup, P. J. J. Org. Chem. 1978, 43, 2164. 221

222

(a) Overman, L. E.; Clizbe, L. A. J. Am. Chem. Soc. 1976, 98, 2352. (b) Overman, L. E. J. Am. Chem. Soc. 1976, 98, 2901.

223

Overman, L. E.; Clizbe, L. A.; Freerks, R. L.; Marlowe, C. K. J. Am. Chem. Soc. 1981, 103, 2807.

786

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

of 188 and 189 in a 93:7 ratio.224 1-Amino-3-siloxybuta-1,3-dienes also reacted with unactivated aldehydes under mild conditions, at room temperature and in the absence of Lewis acids.225

Ph

BnOH 110°C

Heat

N O

H

C

N3

N

Ph

OBn O

O

OBn

N

H

O

(E)-1-Isocyanato- Benzyl (E)-buta-1,3buta-1,3-diene dien-1-ylcarbamate

(E)-Penta-2,4dienoyl azide

N

H

Ph

+

OBn O

188

189

Dienyl lactams (e.g., 190) also have been prepared and shown to undergo Diels-Alder reactions.226 The same orthoand endo-selectivity that was observed with dienyl amides was observed when N-(buta-1,3-dienyl)pyrrolidin-2-one reacted with ethyl acrylate and other alkenes.226c When chiral lactam 191 reacted with ethyl acrylate, Smith and coworkers226d showed that the cycloadduct was formed with excellent enantioselectivity. Rawal and coworkers227 prepared 191, and showed Diels-Alder reactions proceed in good yield and with reasonable stereoselectivity. There are two other classes of dienyl lactams that have been used synthetically, N-alkyl-2-pyridones (192), and Nacyl-1,2-dihydropyridines (193). Pyridones tend to be relatively unreactive, requiring high temperatures, often giving low yields of the cycloadduct.228 N-Acyl pyrrole derivatives229 also function as diene partners in the Diels-Alder reaction. O

O O

N

N

CO2Et 190

N R

OTBS

Ph

N

O

191

O

192

R OEt

193

Nitrogen containing dienes lead to the following useful synthetic disconnections: O R1

O

R N

R R

N

O

R1

+

R1

R N

R

R R

+

N O

R R1

14.9.3 Heteroatom Dienes and Heteroatom Alkene Dienophiles Heteroatoms can be incorporated into the π framework of either the alkene or the diene partner of Diels-Alder reactions. This section will focus on two structural types that function as dienophiles: ketones or aldehydes (194), and imines (195). The discussion will also focus on three structural types that function as dienes: conjugated carbonyls (196), 1-azadienes (197), and 2-azadienes (198). The use of heteroatom dienes230 and alkenes in organic synthesis was thoroughly reviewed by Boger and Weinreb,231 and by Fringuelli and Taticchi.232 (a) Overman, L. E.; Freerks, R. L.; Petty, C. B.; Clizbe, L. A.; Ono, R. K.; Taylor, G. F.; Jessup, P. J. J. Am. Chem. Soc. 1981, 103, 2816. (b) Overman, L. E.; Taylor, G. F.; Houk, K. N.; Domelsmith, L. N. J. Am. Chem. Soc. 1978, 100, 3182.

224

225

(a) Huang, Y.; Rawal, V. H. Org. Lett. 2000, 2, 3321. (b) Huang, Y.; Iwama, T.; Rawal, V. H. J. Am. Chem. Soc. 2000, 122, 7843.

(a) Murata, K.; Terada, A. Bull Chem. Soc. Jpn. 1967, 40, 414. (b) Terada, A.; Murata, K. Bull Chem. Soc. Jpn. 1967, 40, 1644. (c) Zezza, C. A.; Smith, M. B. J. Org. Chem. 1988, 53, 1161. (d) Menezes, R. F.; Zezza, C. A.; Sheu, J. L.; Smith, M. B. Tetrahedron Lett. 1989, 30, 3295.

226

227

Janey, J. M.; Iwama, T.; Kozmin, S. A.; Rawal, V. H. J. Org. Chem. 2000, 65, 9059.

228

Raucher, S.; Lawrence, R. F. Tetrahedron Lett. 1983, 24, 2927.

229

Groves, J. K.; Cundasawmy, N. E.; Anderson, H. J. Can. J. Chem. 1973, 51, 1089.

230

See Ding, X.; Nguyen, S. T.; Wiliams, J. D.; Peet, N. P. Tetrahedron Lett. 2014, 55, 7002.

231

Boger, D. L.; Weinreb, S. M. Hetero Diels-Alder Methodology in Organic Synthesis; Academic Press: San Diego, CA, 1987.

232

Fringuelli, F.; Taticchi, A. Dienes in the Diels-Alder Reaction; John Wiley & Sons, Inc.: NY, 1990.

787

14.9 HETEROATOM DIELS-ALDER REACTIONS

R1

R1 (R,R1 = alkyl, H)

O

R1

O

R2

R 194

195

R1

R

R

R

N

N

N

R

196

197

198

14.9.3.1 Aldehydes and Ketones Most reactions that use carbonyl compounds as dienophiles involve aldehydes, which are more reactive than ketones. Hetero-Diels-Alder reactions of ketones are known, however.233 Simple aldehydes and ketones do not react very well in uncatalyzed reactions,234 although electron-deficient aldehydes and ketones are good partners. Chloral reacted with cyclohexa-1,3-diene to yield 185.235 Glyoxalates (RO2CCHO)236 also add to dienes. The carbonyl moiety of electron-deficient ketones functions as a dienophile,237a but ketone dienophiles may require elevated temperatures and/or high pressure to obtain good reactivity.237 Unactivated ketones react in hydrogen bond promoted reactions238 with activated conjugated dienes.239 As mentioned, relatively simple aliphatic and aromatic aldehydes are unreactive except with very reactive dienes, and the reaction usually requires Lewis-acid catalysis.240 As shown in Section 14.9.1, Danishefsky’s diene (172) reacts with aldehydes to yield pyrones.241 In Jurczak and coworker’s242 synthesis of the (
)-centrolobine ring system, diene 172 reacted with 4-methoxybenzaldehyde in the presence of chromium-salen catalyst to give pyrone 200 in 91% yield and 93 %ee. Pyrones and pyrans are particularly useful for the preparation of acyclic precursors to macrocyclic antibiotics. Exploiting the selectivity inherent in the cyclic pyran (Section 10.5) controls the stereochemistry of the highly oxygenated acyclic fragment.243,244 O Cl3C

H

O

125°C, 26 h

CCl3 Cyclohexa-1,3-diene

199 (30%) OMe

MeO

TMSO H

2% Chiral Cr–Salen complex

+

O

MTBE

O 4-Methoxybenzaldehyde

233

OMe 172

O

200 (91%, 93 %ee)

Jørgensen, K. A. Eur. J. Org. Chem. 2004, 2093.

(a) Tietze, L. F.; Kettschau, G. Top. Curr. Chem. 1997, 189, 1. (b) Kumar, A. Chem. Rev. 2001, 101, 1. (c) Jørgensen, K. A. Angew. Chem. Int. Ed. 2001, 39, 3558.

234

(a) Begley, M. J.; Benner, J. P.; Gill, G. B. J. Chem. Soc. Perkin Trans. 1 1981, 1112. (b) Smushkevich, Y. I.; Belov, V. N.; Kleev, B. V.; Akimova, A. Y. Zh. Org. Khim. 1967, 3, 1036 (Engl. p 997). (c) Dale, W. J.; Sisti, A. J. J. Org. Chem. 1957, 22, 449.

235

(a) Kanowal, A.; Jurczak, J.; Zamojski, A. Rocz. Chem. 1968, 42, 2045 (Chem. Abstr. 1969, 71, 3209b). (b) Zamojski, A.; Konowal, A.; Jurczak, J. Rocz. Chem. 1970, 44, 1981 (Chem. Abstr. 1971, 75, 35008j). (c) Yablonovskaya, S. D.; Shekhtman, N. M.; Antonova, N. D.; Bogatkov, S. V.; Makin, S. M.; Zefirov, N. S. Zh. Org. Khim. 1970, 6, 871 (Engl. p 871).

236

237

(a) Daniewski, W. M.; Kubak, E. I.; Jurczak, J. J. Org. Chem. 1985, 50, 3963. (b) Bonjouklian, R.; Ruden, R. A. J. Org. Chem. 1977, 42, 4095.

238

Huang, Y.; Rawal, V. H. J. Am. Chem. Soc. 2002, 124, 9662.

239

Jørgensen, K. A. Eur. J. Org. Chem. 2004, 2093.

240

Jurczak, J.; Gołe˛biowski, A.; Rahm, A. Tetrahedron Lett. 1986, 27, 853.

(a) Reference 231, p 108. (b) Danishefsky, S.; Kerwin, Jr., J. F.; Kobayashi, S. J. Am.Chem. Soc. 1982, 104, 358. (c) Danishefsky, S.; Kerwin Jr., J. F. J. Org. Chem. 1982, 47, 3183. (d) Danishefsky, S.; Webb, II, R. R. J. Org. Chem. 1984, 49, 1955.

241

242

Chaładaj, W.; Kowalczyk, R.; Jurczak, J. J. Org. Chem. 2010, 75, 1740.

243

(a) Danishefsky, S.; Pearson, W. H.; Harvey, D. F. J. Am. Chem. Soc. 1984, 106, 2456. (b) Kishi, Y. Pure Appl. Chem. 1981, 53, 1163.

244

Danishefsky, S. J. Aldrichimica Acta 1986, 19, 59 (see p 62).

788

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

The disconnection follows: R1

R R1

R2

O

O R2

+ R1

H

R2

R2

R

R1

14.9.3.2 Imines Imines are used less often as dienophiles, but there are examples. Iminium salts are also known to react as dienophiles. Simple imines are usually generated in situ, although imines can be stabilized by the presence of electronwithdrawing groups on the nitrogen. N-Tosyl imines (e.g., 202) can be isolated, for example. When 202 reacted with diene 201, a 58% yield of tetrahydropyridine (203) was obtained as part of Weinreb and Heintzelman’s245 synthesis of cylindrospermopsin. N-Acyl imines react similarly via the corresponding iminium salt. Bis(carbamates) are precursors ohme et al.247 generated iminium salt 204 from chloromethyto iminium salts that undergo Diels-Alder reactions.246 B€ lamine [N-(chloromethyl)-N-ethylethanamine], and subsequent reaction with 2,3-dimethylbuta-1,3-diene gave 205 in 69% yield.

N

CO2Et

Ts ZnCl2, PhMe, rt

+

N

CO2Et

OBn

OBn 201

Ts

203 (58%)

202

Me NKCH2

N Cl N-(Chloromethyl)N-ethylethanamine

Me

N Cl

Cl 204

205 (69%)

Formation of nitrogen-containing cycloadducts via Diels-Alder reactions is an attractive route for the synthesis of alkaloids. Even simple imines can function as dienophiles.248 The Diels-Alder reaction of imines can be applied to more complex systems. The reaction of conjugated ketone 206 with 207, in Jacobsen and coworker’s249 synthesis of (+)reserpine, generated chiral enamine 208 in situ, which reacted with imine 209 to yield 210. In this case, other diastereomers were formed as minor products, but the chiral amine led to formation of the chiral enamine, which induced selectivity in formation of the cycloadduct. A large variety of stable imines are easily prepared, making this technique valuable for the preparation of nitrogen-containing cyclic and polycyclic molecules. Catalytic asymmetric heteroDiels-Alder reactions using imines are known.250

245

Heintzelman, G. R.; Weinreb, S. M. J. Org. Chem. 1996, 61, 4594.

246

Merten, R.; Muller, G. Angew. Chem. 1962, 74, 866.

247

B€ ohme, H.; Hartke, K.; M€ uller, A. Chem. Ber. 1963, 96, 607.

248

See Danishefsky, S.; Vogel, C. J. Org. Chem. 1986, 51, 3915.

249

Rajapaksa, N. S.; McGowan, M. A.; Rienzo, M.; Jacobsen, E. N. Org. Lett. 2013, 15, 706.

250

For a review see Jørgensen, K. A. Angew. Chem. Int. Ed. 2000, 39, 3558.

789

14.9 HETEROATOM DIELS-ALDER REACTIONS

Me Ph Ph

O

OPMB O

MeO

CMe3 S

N

N H

N H

N

MeO

207

OPMB

RHN

Ts

AcOH, Toluene, 23°C

OMe

Ts

H

H

209

OMe

TBSO

N

N

N

NH2

O

OPMB

TBSO OMe

TBSO 206

210

208

Two important disconnections for this type of Diels-Alder reactions follows: R

N

R1

R2

R

R2

R1

N

R2

R2

N

+ R2

N

R2 + R2

R2

14.9.3.3 Nitroso-Type Compounds Nitroso compounds (R-N ¼ O) are dienophiles in Diels-Alder reactions, giving heterocyclic products. In Kibayashi and coworker’s251a synthesis of fasicularin, hydroxamic acid 211 was treated with tetrapropylammonium periodate and 9,10-dimethylanthracene to yield transient acylnitroso compound 212, and the resultant Diels-Alder product 213 was formed in 84% yield. In this particular example, the Diels-Alder adduct was essentially a “protected” acyl nitroso unit, which was used in a subsequent reaction. OBn Me

OBn

Br

O Pr4NIO4

Br

OBn

CHCl3, rt

N O

O

Br

Me

N O

Me

O

NHOH

Me 211

212

213

(84%)

14.9.3.4 Conjugated Aldehydes and Ketones Conjugated aldehydes and ketones function as dienes, in addition to their better known capabilities as dienophiles, as observed by Sherlin et al.252 MVK reacted with acrolein, but gave only 15% of a pyran cycloadduct in the Mundy et al.253 synthesis of brevicomin. However, pyran formation by this reaction is well known,254 and 2-substituted 3,4dihydropyrans are the major products in almost all cases. This cycloaddition is obviously related to the reaction of aldehydes with dienes, but provides pyran products with regiochemical differences relative to those prepared in Section 14.9.3.1. Both Lewis-acid catalysis and high-pressure techniques facilitate the cycloaddition. Electron-deficient alkenes react as dienophiles with conjugated carbonyl derivatives,255 but electron-rich alkenes (e.g., methyl vinyl ether) also react. An example is the reaction of the conjugated aldehyde 5-methylcyclopent-1-ene-1-carbaldehyde with (a) Abe, H.; Aoyagi, S.; Kibayashi, C. J. Am. Chem. Soc. 2000, 122, 4583. For the use of nitroso dienophiles for the synthesis of mannosidase and fucosidase inhibitors, see (b) Joubert, M.; Defoin, A.; Tarnus, C.; Streith, J. Synlett 2000, 1366. Also see (c) Bach, P.; Bols, M. Tetrahedron Lett. 1999, 40, 3461.

251

252

Sherlin, S. M.; Berlin, A. Y.; Serebrennikova, T. A.; Rabinovich, R. F. J. Gen. Chem. USSR 1938, 8, 7, 22.

(a) Mundy, B. P.; Otzenberger, R. D.; DeBernardis, A. R. J. Org. Chem. 1971, 36, 2390, 3830. (b) Lipkowitz, K. B.; Mundy, B. P.; Geeseman, D. Synth. Commun. 1973, 3, 453; also see (c) Bhupathy, M.; Cohen, T. Tetrahedron Lett. 1985, 26, 2619. 253

254

Desimoni, G.; Tacconi, G. Chem. Rev. 1975, 75, 651.

255

Childers Jr., W. E.; Pinnick, H. W. J. Org. Chem. 1984, 49, 5276.

790

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

the C]C unit of vinyl ether (Z)-1-isopropoxyprop-1-ene, to give a 42% yield of 214 in the Korte et al.256 synthesis of dl, iso-iridomyrmecin. Boger and Weinreb.257 show several reactions of this type. H

O O

Me

+

O

Me

170°C, Autoclave

O

Me cat HO

OH

H

48 h

(Z)-1-Isopropoxyprop-1-ene

5-Methylcyclopent1-ene-1-carbaldehyde

Me

214 (42%)

The disconnection for this heteroatom Diels-Alder reaction follows: R2

R2 + R

O

R

X

O

X

14.9.3.5 Azadienes Azadienes (e.g., 197 and 198)258 are useful diene partners that can be used for the preparation of many heterocycles and alkaloids. The utility of 1-azadienes in the Diels-Alder reaction is limited by the fact that isomerization occurs to give a conjugated imine, probably via a proton transfer. An example is the equilibration of 215 and 216. Addition of maleic anhydride to this mixture gave the Diels-Alder adduct 217, derived from aminobutadiene 215.259 Incorporation of tertiary alkyl substituents are used to stabilize the 1-azadiene. Even when isomerization is not possible, the aminobutadiene is relatively unreactive and the yields can be poor. An important side reaction is a competitive [2+2]cycloaddition (Section 15.2), further contributing to poor yields of the [4+2]-cycloadduct.260,261 Et

Et

Et O

O

O

O O

Et

Et

Et

NHPh 215

N—Ph 216

NHPh

O

217

2-Azadienes (e.g., 218) were prepared by Ghosez and coworkers262 and then reacted with conjugated alkynes (e.g., the methyl propiolate) to give an initial cycloadduct 219, but facile loss of dimethylamine led to aromatization and formation of pyridine derivative methyl 6-(dimethylamino)nicotinate as the final product. The previously noted azadiene-aminobutadiene rearrangement is impossible in 218, and the dimethylamino moieties further stabilize the azadiene structure. Ghosez and coworkers262 also used 2-azadienes that were stabilized by the presence of silyloxy groups.

256

Korte, F.; B€ uchel, K. H.; Zschocke, A. Chem. Ber. 1961, 94, 1952.

257

Reference 231, pp 167–213.

258

For a review, see Behforouz, M.; Ahmadian, M. Tetrahedron 2000, 56, 5259.

259

(a) Snyder, H. R.; Robinson, Jr., J. C. J. Am. Chem. Soc. 1941, 63, 3279. (b) Snyder, H. R.; Cohen, H.; Tapp, W. J. J. Am. Chem. Soc. 1939, 61, 3560.

260

(a) Reference 231, p 242 and Reference 18a-c cited therein. (b) Taylor, E. C.; Eckroth, D. R.; Bartulin, J. J. Org. Chem. 1967, 32, 1899.

261

Garashchenko, Z. M.; Skvortsova, G. G.; Shestova, L. A. USSR Patent 370,208, 1973 (Chem. Abstr. 1973, 79, 31900d).

262

Sainte, F.; Serckx-Poncin, B.; Hesbain-Frisque, A.-M.; Ghosez, L. J. Am. Chem. Soc. 1982, 104, 1428.

791

14.10 ENANTIOSELECTIVE DIELS-ALDER REACTIONS

Me2N

Me2N

H

+

N

Me2N N

NMe2

CO2Me

CO2Me

NMe2

Methyl propiolate

218

N

CO2Me

Methyl 6-(dimethylamino)nicotinate

219

H N HN

O

N

N

221

N

OEt

N H

O N

+

OEt

OEt O

O

220

N N

N

O

N

N

CHCl3, rt, 1 d

O

222 (55%)

232 (18%)

Boger and Weinreb263 discuss several examples of 1- and 2-azadienes in Diels-Alder reactions. Kende et al. used a less obvious azadiene in a synthesis of alantrypinone.264 When 6H-pyrazino[2.1-b]quinazoline-6-one (220) was heated with 3-methyleneoxindole (221), a mixture of 222 and 223 was formed in 55% yield and 18% yield, respectively. The synthetic utility of azadienes Diels-Alder reactions lies in the synthesis of heterocycles and alkaloids. Three disconnections related to this section follow: R1

R

R1 R2

N R

R

R1

R1

R2 N

R2

+ R

N

+ N

R1

R1

R2

R2

N

N + R2

R

R

14.10 ENANTIOSELECTIVE DIELS-ALDER REACTIONS The diastereoselectivity inherent to the Diels-Alder reaction is apparent in several examples in preceding reactions. The reaction is not inherently enantioselective since there is no facial control to bias approach of alkene to diene from a single face for intermolecular reactions. Some facial control is available for intramolecular reactions. The ortho rule, the endo rule (secondary orbital interactions), and steric interactions provide some orientational control, but facial control is also required for enantioselectivity. When ethyl acrylate reacts with 2-methylpenta-1,3-diene, the endo approach can occur from the bottom as in 233A, or from the top as in 233B. Clearly, the two products resulting from attack at opposite faces (234A and 234B) are mirror images and enantiomers. This lack of facial selectivity leads to racemic mixtures in all Diels-Alder cyclization reactions discussed to this point. As in previous chapters, the key to enantioselectivity is to provide both facial and orientational control in the transition state of the reaction. The orientational control (regioselection and endo-selectivity) for the diene and alkene is provided by orbital interactions, including secondary orbital effects. As in previous chapters, facial selectivity will be provided by the presence of stereogenic centers in the molecule that usually make one face in the transition state more sterically hindered than the other. The diene can be constructed from a chiral precursor (a chiral template, Section 8.9). Alternatively, a chiral auxiliary (Xc) can be attached to the diene or alkene. A third approach is related to using a chiral auxiliary in that a chiral material (usually a chiral catalyst) is added to the reaction, usually providing a transient chiral auxiliary (Section 7.9). Catalytic, enantioselective Diels-Alder reactions are now well known, and many catalysts and 263

Reference 231, pp 239–358.

264

Kende, A. S.; Fan, J.; Chen, Z. Org. Lett. 2003, 5, 3205.

792

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

auxiliaries are available.265 The rare earth(III) salt-catalyzed asymmetric Diels-Alder reaction with a chiral dienophile in supercritical CO2 has been reported, and the use of alternative solvents is an on-going area of study.266 Me

Me

Me

H Me

L

Me O

H

EtO

CO2Et

Me

EtO2C 234A

233A

O

Me

EtO2C

EtO

H

Me

H Me

Me

L

Me

CO2Et

Me

234B

233B

14.10.1 Chiral Auxiliaries A chiral auxiliary is a chiral molecule, usually derived from Nature, which is attached to the reaction substrate (in high yield and under mild conditions), provides facial selectivity and, hopefully, orientation control as well in a reaction.267 After the reaction of interest, the auxiliary must be removed in high yield and under mild conditions to regenerate the original functional group. For a Diels-Alder reaction, the chiral auxiliary must be attached to either an achiral diene (as in 235) or an achiral dienophile (as in 238) to generate a chiral molecule. A limitation, of course, is the availability of a functional group in the alkene or diene that can react with a chiral auxiliary. A Diels-Alder reaction with 235, for example, would generate 236 as a mixture of diastereomers, each in high enantiomeric purity (%ee, Section 1.4.6). The auxiliary is then removed to yield the chiral target 237. A similar sequence with 238 generated 239, and removal of Xc gave 240. The auxiliary provides facial selectivity in the Diels-Alder reaction, transferring that chirality to the cycloadduct. The Diels-Alder reactions of both 235 and 238 may proceed with excellent enantioselectivity, but they do proceed with moderate diastereoselectivity. This problem of diastereoselectivity and enantioselectivity must be addressed in all asymmetric reactions, and was seen in reductions (Chapter 7), as well as nucleophilic carbon bond-forming reactions (Chapters 11 and 12), and especially in enolate reactions (Chapter 13). Y

Xc

Xc

R

Xc – Xc

O

O

R Y

Xc

Xc

O

Y

O

235 R

R

R

O

236

238

R

R

R1

O

237

Xc R1

O 239

– Xc

Y R1

O 240

Morrison and Mosher268 showed that menthyl was an early auxiliary, and when an acrylic acid derivative reacts with menthol to form a menthyl ester (e.g., 241), asymmetric induction is possible upon reaction with cyclopentadiene

265

Corey, E. J. Angew. Chem. Int. Ed. 2002, 41, 1651.

266

Fukuzawa, S.; Metoki, K.; Esumi, S. Tetrahedron 2003, 59, 10445.

267

For a review of bicyclic lactam chiral auxiliaries, see Groaning, M. D.; Meyers, A. I. Tetrahedron 2000, 56, 9843.

268

Morrison, J. D.; Mosher, H. S. Asymmetric Organic Reactions; American Chemical Society: Washington, DC, 1976; p 256.

793

14.10 ENANTIOSELECTIVE DIELS-ALDER REACTIONS

to yield diastereomers 242 and 243.269 In the absence of a Lewis acid, the %ee is rather poor.269 The use of boron trifluoride etherate and lower temperatures produced the highest observed enantiomeric excess of 74–85 %ee. O O

1. LiAlH4

+ Me

+

2. H2O

HO

HO

241

243

242

One of the first examples of an asymmetric Diels-Alder reaction is the work of Walborsky and Barash270 in the reaction of (
)-dimenthyl fumarate and buta-1,3-diene to yield 244 and 245, after reduction of the ester products with LiAlH4.271 Reaction in toluene using a Lewis-acid catalyst provided the best asymmetric induction, in good yield. Hydride reduction of esters (Section 7.6.1) is a common method for removing ester auxiliaries. In general, menthol-type auxiliaries provide only moderate selectivity.272 CO2menthyl 1. Hydroquinone, 6 h PhH, Sealed tube, Heat

OH

2. LiAlH4 3. H3O+

menthylO2C

OH

+

OH

(–)-Dimenthyl fumarate

OH

244

245

In Oppolzer’s review,273 several chiral auxiliaries related to menthol and other chiral alcohols were presented, giving the diastereomeric endo-adducts 246 and 247, as shown in Table 14.6.274 In all cases, the chiral ester in Table 14.6 TABLE 14.6

Enantioselectivity in the Diels-Alder Reaction of Cyclopentadiene With Conjugated Esters Derived From Chiral Alcohols O

Xc Xc

Xc

+

246

Ph

Me

O

88% (R)

63% (R) Me

Me

Me Me

O

OBn O

Me Ph

Me

85% (R)

Ph Me

%ee

Xc Ph

O

Ph

Me

247

%ee

Xc

O

O

TiCl4

Me

Ph

Ph Ph

88% (S)

81.5% (R)

Me O

O Me

Ph Me Me

82.7% (S)

269  (a) Sauer, J.; Kredel, J. Tetrahedron Lett. 1966, 6359. (b) Farmer, R. F.; Hamer, J. J. Org. Chem. 1966, 31, 2418. (c) Cervinka O.; Kríž, O. Collect. Czech. Chem. Commun. 1968, 33, 2342. 270

Walborsky, H. M.; Barash, L.; Davis, T. C. Tetrahedron 1963, 19, 2333.

271

Reference 231, p 253.

272

Roush, W. R.; Gillis, H. R.; Ko, A. I. J. Am. Chem. Soc. 1982, 104, 2269.

273

Oppolzer, W. Angew. Chem. Int. Ed. 1984, 23, 876.

274

Oppolzer, W.; Kurth, M.; Reichlin, D.; Chapuis, C.; Mohnhaupt, A.; Moffatt, F. Helv. Chim.Acta 1981, 64, 2802.

794

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

provided good-to-excellent selectivity. Oppolzer et al.274 discussed the source of such selectivity by invoking Fig. 14.12 (from 5-benzyloxymethyl cyclopenta-1,3-diene) in which the derivatized menthyloxy chiral auxiliary 248 provided facial selectivity to the acrylate, for attack from either the re or the si face. The si face is less hindered and 249 [the (R) cycloadduct] is preferred to 250, consistent with the results in Table 14.6. Fig. 14.12 can be used to illustrate the source of facial selectivity for other chiral auxiliaries. The poor-to-moderate asymmetric induction with menthol and related auxiliaries contrasts with Trost’s use of 251 with napthoquinone 252, which gave 98% of 253 with >97 %ee.275

H Ph

si Si

O

O

Ph

H Xc

OBn O

O

249 Re

O

re

Me

H

248 O

Ph

Xc

H 250

FIG. 14.12 Facial selectivity of chiral acrylates in the Diels-Alder reaction. Reproduced with permission from Oppolzer, W. Angew. Chem. Int. Ed. 1984, 23, 876. Copyright © 1984 Wiley-VCH Verlag GmbH & Co. KGaA.

O

O

H

O O

B(OAc)3

+

CHCl3, 0°C

Ph OMe

OH

251

OH

O 252

H

H O

O

O

253

Ph OMe

(98%)

Evans developed a new generation of chiral auxiliaries, based on conversion of amino alcohols to oxazolidinone derivatives. The use of these auxiliaries to yield the corresponding conjugated amide 254 led to excellent selectivity in the intramolecular Diels-Alder reaction, which generated mixtures of 255 and 256.276 Coordination with the Lewis-acid catalyst generated a complex in which facial selectivity is enhanced relative to the uncomplexed species. When dimethylaluminum chloride was used as a Lewis-acid catalyst with 257, for example, chelated intermediate 258 was formed. This provided orientational control for the dienophile, and the chiral nature of the auxiliary provided the requisite facial control, with approach of the diene from the less sterically hindered face.

O Xc

Xc

O

+

Xc K

CH2Cl2

255

Ph

O N

H

H 254

O

O N

H

Me2AlCl

O

Xc

O

H 256

(95:5 255:256) O

O

Trost, B. M.; O’Krongly, D.; Belletire, J. L. J. Am. Chem. Soc. 1980, 102, 7595.

276

Evans, D. A.; Chapman, K. T.; Bisaha, J. Tetrahedron Lett. 1984, 25, 4071.

O

Ph N

N Me

(15:85 255:256)

275

(83:17 255:256) O

C6H11 (3:97 255:256)

795

14.10 ENANTIOSELECTIVE DIELS-ALDER REACTIONS

Good selectivity was observed with the intermolecular Diels-Alder reaction as well.277 In a synthesis of (+)eutipoxide B by Okamura and coworkers,278 the Cinchonine-catalyzed reaction of pyrone 3-hydroxy-2H-pyran-2one and acrylamide 259 (containing the Evan’s auxiliary shown) proceeding good yield, with >95% selectivity for 260. Me O N

Me

Me

si

O

Al

O

O

O

2 Me2AlCl

Me2AlCl2 N

Me

O

Ph Ph re

257

O

O

O

258

O 0°C

O

N

+

OH

O

Ph 3-Hydroxy-2H-pyran-2-one

O

O

O

0.2 equiv Cinchonine i-PrOH, H2O

259

N OH 260

O

Ph

(>95%)

Oppolzer et al.279 developed a different set of auxiliaries based on a camphor precursor, typified by sultam 261 (Oppolzer’s sultam) and used in a chiral synthesis of the aglycone of loganin.280 Sultam (261) was also the precursor to acrylamide (262), which reacted with diene 263 to give a 66% yield of 264 (7:1 exo/endo), in the Roush et al.281 synthesis of the spirotetronate subunits of quartromicin. Selectivity in this Diels-Alder reaction is due to chelation of the sultam unit with the catalyst (MeAlCl2). t-BuMe2SiO Me

N

H

O

O

S O

261

O

N

CH2Cl2

+

N

S

SO2 Me 1.8 MeAlCl 2 –78

Me

0°C

O

t-BuPh2SiO

O

Me t-BuMe2SiO

262

263

264

OSiPh2t-Bu

(66%)

Many auxiliaries yield excellent asymmetric induction in the Diels-Alder reaction. Helmchen et al.282 developed an auxiliary based on camphor that is quite useful. Helmchen and coworkers283 also prepared the lactone auxiliary 265 and the imide auxiliary 266, and showed that they give opposite enantioface selectivity in the Diels-Alder reaction. In the presence of a Lewis acid (e.g., TiCl4), differences in coordination site led to a different chiral face in the transition state of the cycloaddition and opposite stereochemistry. Helmchen and coworkers284 used an auxiliary based on 265 in a synthesis of cyclosarkomycin. Other auxiliaries include 267, developed by Marchand-

277

Evans, D. A. Chapman, K. T.; Bisaha, J. J. Am. Chem. Soc. 1984, 106, 4261.

278

Shimizu, H.; Okamura, H.; Iwagawa, T.; Nakatani, M. Tetrahedron 2001, 57, 1903.

279

Oppolzer, W.; Chapuis, C.; Dupuis, D.; Guo, M. Helv. Chim. Acta 1985, 68, 2100.

280

Vandewalle, M.; van der Eycken, J.; Oppolzer, W.; Vullioud, C. Tetrahedron 1986, 42, 4035.

281

Roush, W. R.; Limberakis, C.; Kunz, R. K.; Barda, D. A. Org. Lett. 2002, 4, 1543.

(a) Helmchen, G.; Schmierer, R. Angew. Chem. Int. Ed. 1981, 20, 205. (b) Schmierer, R.; Grotemeier, G. Helmchen, G.; Selim, A. Angew. Chem. Int. Ed. 1981, 20, 207. 282

283

Poll, T.; Abdel Hady, A. F.; Karge, R.; Linz, G.; Weetman, J.; Helmchen, G. Tetrahedron Lett. 1989, 30, 5595.

284

Linz, G.; Weetman, J.; Abdel Hady, A. F.; Helmchen, G. Tetrahedron Lett. 1989, 30, 5599.

796

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

Brynaert and coworkers285 for 1-amino dienes, and the exo-selective auxiliary 268 developed by Kudo and Kawamura.286 Lamy-Schelkens and Ghosez287 showed that changing the catalyst in Diels-Alder reactions using a chiral auxiliary can influence selectivity. A major drawback with auxiliaries is that they can sometimes be difficult to remove (e.g., vigorous aqueous acid or base or treatment with LiAlH4 may be required). Dienes and dienophiles must have heteroatom substituents for these auxiliaries or other methodology must be developed.

O

O

O

O O

O

O

265

N

N

O

Me

O

N

O

O

N

N

O

O

Ph

266

267

268

14.10.2 Chiral Additives and Chiral Catalysts Chiral additives are clearly related to chiral auxiliaries in that they typically react to form a transient complex (auxiliary), which provides the requisite facial selectivity. In Section 7.9.3, naturally occurring chiral amines were added to reducing agents to generate chiral additives, chiral catalysts, or chiral auxiliaries in order to achieve asymmetric induction. Protected proline and abrine esters (L-abrine is N-methyl-L-tryptophan) and related compounds have been used as chiral catalysts.288 These amines can also be used in Diels-Alder reactions, with variable results. O O

N

O

Me

H

H

N Me O

Additive

OH HO Anthracen-9-ol

269

One example is the reaction of anthracen-9-ol with N-methylmaleimide to yield 269,289 in the presence of a chiral additive. The use of various chiral amine additives [quinine, quinidine, cinchonine, cinchonidine, (S)-prolinol, and (1R,2S)-N-methylephedrine], gave enantiomeric excesses between 16% and 61%) in the cycloadduct. In this particular example, quinidine proved the most effective additive with a 61 %ee. Presumably, coordination between the chiral amine and anthracen-9-ol via hydrogen bonding (or a dipole-dipole interaction) provides diastereoselection in the subsequent reaction with N-methylmaleimide. Preparation of a chiral catalyst will provide facial selectivity when it coordinates to the substrate (usually the dienophile), forming a chiral transition state containing the transient auxiliary. Roush et al.272 used menthoxyl and bornyloxyaluminum catalysts, providing an example in which traditional chiral materials are attached to aluminum chloride. Significantly more effective catalysts have been developed based on coordination of chiral molecules [usually bis(amines) or diols] with Al or Ti reagents. Corey et al.290 prepared chiral catalyst 271, which reacted with 270 to form chiral complex 272. This complex was formed in situ, and reacted with cyclopenta-1,3-diene to give an 88% yield of 273, with 94 %ee (in a ratio of 96:4 endo/exo).

285

Robiette, R.; Cheboub-Benchaba, K.; Peeters, D.; Marchand-Brynaert, J. J. Org. Chem. 2003, 68, 9809.

286

Kawamura, M.; Kudo, K. Chirality 2002, 14, 727.

287

(a) Lamy-Schelkens, H.; Ghosez, L. Tetrahedron Lett. 1989, 30, 5891. (b) Lamy-Schelkens, H.; Giomi, D.; Ghosez, L. Tetrahedron Lett. 1989, 30, 5887.

288

Ahrendt, K. A.; Borths, C. J.; MacMillan, D. W. C. J. Am. Chem. Soc. 2000, 122, 4243.

289

Riant, O.; Kagan, H. B. Tetrahedron Lett. 1989, 30, 7403.

290

Corey, E. J.; Imwinkelried, R.; Pikul, S.; Xiang, Y. B. J. Am. Chem. Soc. 1989, 111, 5493.

797

14.10 ENANTIOSELECTIVE DIELS-ALDER REACTIONS

Ph 0.5 M, CH2Cl2

O

O

N

–78°C, 16 h

N

O

O

N O

F3CO2S

F3CO2S N Al N SO2CF3

Me H

O

Al

Ph Ph

Ph

SO2CF3 Me

O

H

N

N

O

O

Me

270

271

272

273

(88%)

Corey et al.291 developed other catalysts for enantioselective Diels-Alder reactions. One is a C2-symmetric chiral bis (oxazoline)-Fe(III) complex (274). Similar Cu coordinated catalysts have been developed and yield good-to-excellent enantioselectivity in Diels-Alder reactions.292 Analogues of 274 reacted with cyclopenta-1,3-diene in the presence of a 274FeCl2I catalyst to yield cycloadducts with good enantioselectivity (93:7, with 99:1 endo/exo selectivity) at
50°C in dichloromethane (15, 10 mol% of the catalyst).291 The FeI3 complex that is shown with 274, as well as the FeCl3 complex, gave slightly poorer selectivity than the FeCl2I complex. The (S)-tryptophan derived oxazoborolidine (275) catalyst was designed to facilitate intramolecular interactions in highly enantioselective Diels-Alder reactions.293 In a synthesis of estrone,294 Corey and coworkers295 reacted 277 with ethyl (E)-3-methyl-4-oxobut-2-enoate, in the presence of chiral catalyst 276 (an oxazaborolidinium salt), to give a 92% yield of 278, in 94 %ee. Catalysts related to 276 have been shown to be broad-spectrum catalysts for enantioselective Diels-Alder reactions. H Me Me

H

Me O

O N

O • FeI3

Ph

Ph

O

N

Bu

B

274

N

H

S

N

Tf2N

O

O

H

Me

N

H

B

O

H

275

276 Me

Me

CHO

+

CHO

0.2 equiv 276

CO2Et

CH2Cl2, –78°C , 8 h

EtO2C MeO

H

H

MeO 277

Ethyl (E)-3-methyl-4oxobut-2-enoate

278 (92%, 94 %ee)

Yamamoto and coworkers296 reported an asymmetric hetero-Diels-Alder reaction that was catalyzed by a chiral organoaluminum reagent (280). When diene 279 reacted with benzaldehyde in the presence of 10 mol% of 280, a mixture of cis-dihydropyrone (281) and trans-dihydropyrone (282) was obtained in 77 and 7% yield, respectively. The optical activity of 281 was 95 %ee, and that of 282 was 53 %ee.287 In the complex that mediates this reaction, the diene approaches benzaldehyde with an endo-alignment that minimizes steric repulsion between the incoming diene and the triphenylsilyl moiety on that face.

291

Corey, E. J.; Imai, N.; Zhang, H.-Y. J. Am. Chem. Soc. 1991, 113, 728.

292

For copper-coordinated catalysts, see (a) Evans, D. A.; Barnes, D. M.; Johnson, J. S.; Lectka, T.; von Matt, P.; Miller, S. J.; Murry, J. A.; Norcross, R. D.; Shaughnessy, E. A.; Campos, K. R. J. Am. Chem. Soc. 1999, 121, 7582. For copper- and zinc-coordinated catalysts see (b) Yao, S.; Roberson, M.; Reichel, F.; Hazell, R. G.; Jørgensen, K. A. J. Org. Chem. 1999, 64, 6677.

293

Corey, E. J.; Loh, T.-P. J. Am. Chem. Soc. 1991, 113, 8966.

294

Hu, Q.-Y.; Rege, P. D.; Corey, E. J. J. Am. Chem. Soc. 2004, 126, 5984.

295

Ryu, D. H.; Lee, T. W.; Corey, E. J. J. Am. Chem. Soc. 2002, 124, 9992.

296

Maruoka, K.; Itoh, T.; Shirasaka, T.; Yamamoto, H. J. Am. Chem. Soc. 1988, 110, 310.

798

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

Narasaka et al.297 developed a Ti catalyst generated by complexation with chiral diols. The dienophile must contain functionality that will coordinate with the metal catalyst to form a chiral complex, and these catalysts are less effective with dienes and dienophiles that do not contain heteroatoms. A related Ti-BINOL complex has been used to catalyze Diels-Alder reactions.298 Kelly et al.299 prepared a transient boron catalyst, prepared in situ with borane and a chiral diol. It is clear that chiral additives, particularly chiral catalysts can be effective. The asymmetric induction is usually good-to-excellent with properly designed dienophiles. Effective chiral catalysts continue to be reported and will undoubtedly be a major force in chiral synthesis involving Diels-Alder reactions. SiPh3

OMe

+

O

1. PhCHO, PhMe –20°C, 2 h

Al Me

O

2. CF3CO2H

O

Me3SiO

Me

Me

O

Me

O Ph

+ O

Ph

Me

Me 279

280

SiPh3

Me

281

282

14.10.3 Chiral Templates The last approach to chiral synthesis begins with a chiral precursor whose components will be integrated into the final product, which is the chiral template approach (Section 8.9). The examples of asymmetric syntheses involving DielsAlder templates are as varied as the targets. One is by Oppolzer and Flaskamp,300 who reported an asymmetric synthesis of pumiliotoxin C via Diels-Alder cyclization of 283, prepared from (R)-norvaline, to give 284. Me Me N

Heat

H

N

H

O O 283

284

Carbohydrates are efficient chiral templates that have been used in the Diels-Alder reaction. Sherburn and Lilly301 used L-ascorbic acid as a template to synthesize 285. Heating this compound in refluxing toluene led to a 96:4 mixture of 286 and 287 (68% yield). The chirality inherent in the carbohydrate precursor provides the needed facial selectivity that is transferred to the cycloadduct product. The use of chiral templates for preparing Diels-Alder precursors will undoubtedly increase in importance. As the need for enantiomerically pure material increases, the chiral template driven Diels-Alder reactions will be increasingly effective. O

O O TIPSO

O

MeO2C

O H

PhMe

O

Reflux

TIPSO

H

H

O

H

+ MeO2C

H

TIPSO MeO2C

286

O

H

O

O 285

O

O 287

(68%)

(a) Narasaka, K.; Iwasawa, N.; Inoue, M.; Yamada, T.; Nakashima, M.; Sugimori, J. J. Am. Chem. Soc. 1989, 111, 5340. (b) Narasaka, K.; Inoue, M.; Yamada, T.; Sugimori, J.; Iwasawa, N. Chem. Lett. 1987, 2409.

297

298

Mikami, K.; Motoyama, Y.; Terada, M. J. Am. Chem. Soc. 1994, 116, 2812.

299

Kelly, T. R.; Whiting, A.; Chandrakumar, N. S. J. Am. Chem. Soc. 1986, 108, 3510.

300

Oppolzer, W.; Flaskamp, E. Helv. Chim. Acta 1977, 60, 204.

301

Lilly, M. J.; Sherburn, M. S. Chem. Commun. 1997, 967.

799

14.11 CONCLUSION

14.11 CONCLUSION It is clear from the preceding sections that a wide range of cyclic and polycyclic molecules can be prepared by pericyclic reactions. In addition, stereocontrolled syntheses of acyclic molecules are possible via initial cycloaddition followed by cleavage of the ring. The power of these reactions lies in the ability to make carbon bonds with high regioselectivity and stereoselectivity. If the proper chiron or chiral auxiliary is used, high enantioselectivity can be obtained as well. Few synthetic methods allow access to such a large number of natural products, with compatibility with such a large number of functional groups. The next chapter will introduce pericyclic reactions that involve other πsystems, including [2+2]- and [3+2]-cycloadditions, sigmatropic rearrangements, electrocyclic reactions, and the ene reaction. HOMEWORK

1. Using the tables in this chapter (where possible), particularly Tables 14.1 and 14.2, evaluate the relative rate of reaction for each of the following: For the faster reaction, predict the structure, and where data is available predict the regiochemistry and stereochemistry of the major product(s). O

Ph Me-C C-Me

(A)

(B)

MeO

O

O

O

Me

MeO

Cl

(C)

OHC

O2N

NC

(D)

N H

2. In Section 14.8.1, Boger and coworker’s195 conversion of tetrazines to diazines via inverse electron demand DielsAlder reactions was mentioned. Give a mechanistic rationale for the transformation shown, based on this paper, that accounts for formation of the indicated product. O

N N

O

Me Me

N N MeO

OMe

MeO

NEt2

O

NEt2

Dioxane, 23°C, 2 h

O N N

MeO

OMe

OMe

MeO

OMe

3. The macrocyclic compound shown, generated in situ, spontaneously reacted to yield the indicated product in 63% yield. Explain the formation of this product. TBSO TBSO

CO2Et Me Me

O

Me OTBS

TBSO H OTBS EtO2C O

TBSO

Me H

Me Br

Br

Me

4. In a synthesis of merrilactone A (see J. Am. Chem. Soc. 2002, 124, 2080), Danishefsky reports the Diels-Alder reaction yields the cycloadduct shown. This product results from an exo approach of the anhydride to the diene, assuming a disrotatory motion of the OTBS group. Although this result is not discussed in the cited reference, offer a suggestion why the Alder endo rule does not predict the major diastereomer. t-BuMe2SiO

O

t-BuMe2SiO

+

O

Mesitylene, 165°C

O

O

Methylene blue, Collidine

O

O (74%)

800

14. PERICYCLIC REACTIONS: THE DIELS-ALDER REACTION

5. For each of the following reactions give the major product, with correct regiochemistry and stereochemistry: OSiMe2t-Bu

(A)

EtO2C

O

CO2Et

(B)

PhH, 20°C

NMe2

CHCl3, –20°C 120 h

Ph 1. MOMOCH2CHO BF3⋅OEt2, CH2Cl2, –78°C

OSiMe2t-Bu

(C)

CO2Et

O2N

(D)

OMe 2. NaBH4, EtOH, CeCl3⋅7 H2O

PPh3, DEAD, THF 2. 150°C, 10 d

3. Ac2O, Py, DMAP

O BnO

OMOM

Me NaIO4

Heat

(F)

N

H2O–DMF

O

Me OTMS

O EtO2C

(G)

CF3

O NHOH

(E)

CO2H

1.

HO

CO2Et

(H)

Cat–Cu(OTf)2 CH2Cl2

MeO2C

C

1.

CO2Me

Neat, 25°C 2. NHEt3F, EtOH

MeO OTMS

Me N Br CN

(I)

(J)

OTMS

O

O

1.

TMSO OMe

,

NEt3

PhBr, Reflux

O

Me

MeO

OMe

2. Hydrolysis

MeO BnO

Cl

EtO CHCl3, 60°C MS 4 Å, 6 d

O

(K)

1. t-BuLi, THF, –78°C 2. Furan

(L)

N

N

OMe

MOMO

Me

OH Me

(M)

1. SOCl2, Py 2. MeO2CCLCCO2Me 110 °C

Me

6. In each case, provide a suitable synthesis. Show all intermediate products and all reagents.

Me

(A)

(B)

O

O

O

O Br

O O

O Cl

MeS

Cl

MeS

(C) O

OTBS

O O NH

O

Me

Me

OH

C H A P T E R

15 Pericyclic Reactions: [m + n]-Cycloadditions, Sigmatropic Rearrangements, Electrocyclic, and Ene Reactions 15.1 INTRODUCTION The Diels-Alder reaction was the main focus of Chapter 14, because this reaction is an important method for generating carbon-carbon bond forming reactions. Without question, this reaction is among the most useful in synthesis, but there are other pericyclic reactions that are also commonly used. This chapter will focus on those reactions, including other cycloaddition reactions, sigmatropic rearrangement reactions, and the ene reaction. This chapter will present a brief overview of the mechanistic basis for each reaction. As in previous chapters, however, the emphasis will be on the synthetic utility. As with the Diels-Alder reaction in Chapter 14, π-orbitals are the key to understanding the pericyclic reactions presented in this chapter. The discussion will focus on the π-bonds of alkenes and dienes, the so-called frontier molecular orbitals (FMO).1 As described in Chapter 14, the lowest energy orbital containing bonding electrons is called the highest occupied molecular orbital, the HOMO.2 The higher energy orbital does not contain electrons, but is the next available energy level if electrons are available and it is called the lowest unoccupied molecular orbital, the LUMO.

15.2 [2 + 2]-CYCLOADDITION REACTIONS The Diels-Alder reaction is described as a [4 + 2]-cycloaddition because the 4π-electrons of a diene react with the 2π-electrons of an alkene. If a reaction occurs between the 2π-electrons of an alkene and the 2π-electrons of another alkene, the reaction is identified as a [2 + 2]-cycloaddition. A similar [2 + 2]-cycloaddition reaction can occur between an alkene and a carbonyl or an imine.3 Cyclobutanes and oxetanes are prepared, respectively, by this reaction, and β-lactams and β-lactones have been prepared using this methodology. This section will focus on both thermal and photochemical [2 + 2]-cycloaddition reactions.

15.2.1 Thermal [2 + 2]-Cycloadditions In Section 14.4, pericyclic thermal [2 + 2]-cycloaddition reactions that form cyclobutane derivatives (e.g., 3)4 were identified as forbidden due to symmetry considerations, but photochemically allowed. There are, however, several examples of thermal [2 + 2]-cycloadditions that probably occur by a different mechanism. This alternative mechanism 1 (a) Fukui, K.; Yonezawa, T.; Nagata, C.; Shingu, H. J. Chem. Phys. 1954, 22, 1433; (b) Fukui, K. In Molecular Orbitals in Chemistry, Physics and Biology; L€ owdin, P. -O.; Pullman, B., Eds.; Academic Press: New York, NY, 1964; p 513. 2

Houk, K. N.; Sims, J.; Watts, C. R.; Luskus, L. J. J. Am. Chem. Soc. 1973, 95, 7301.

3

(a) Larock, R. C. Comprehensive Organic Transformations, 2nd ed.; Wiley-VCH, New York, NY, 1999; pp 161–165; (b) Lee-Ruff, E.; Mladenova, G. Chem. Rev. 2003, 103, 1449.

4

(a) Desimoni, G.; Tacconi, G.; Barco, A.; Pollini, G. P. Natural Product Synthesis Through Pericyclic Reactions; American Chemical Society: Washington, DC, 1983; pp 33–63; (a) Woodward, R. B.; Hoffman, R. Angew. Chem. Int. Ed. Engl. 1969, 8, 781. Organic Synthesis http://dx.doi.org/10.1016/B978-0-12-800720-4.00015-5

801

Copyright © 2017 Michael Smith. Published by Elsevier Inc. All rights reserved.

802

15. PERICYCLIC REACTIONS

probably involves a radical or dipolar intermediate (1 or 2, respectively), or it occurs via HOMO-HOMO or LUMOLUMO interactions.5 Lewis acid catalyzed reactions of alkynes with alkenes or conjugated carbonyl compounds are also known,6 proceeding via a metal-stabilized cationic intermediate.6b The reaction of dimethylaminoisobutene (N,N,2-trimethylprop-1-en-1-amine) with methyl acrylate gave cyclobutane (5) by what is believed to be a zwitterionic intermediate (4).7 Ionic intermediates need not be invoked in all cases, since some reactions proceed via diradical intermediates. The HOMO of N,N,2-trimethylprop-1-en-1-amine and LUMO of methyl acrylate can be used to predict the regiochemistry in 5.8 In general, the presence of an electron-withdrawing group on a ketene, (e.g., a chlorine), an alkoxy group, an aryl group, or a vinyl group greatly enhances the rate of the [2 + 2]-cycloaddition reaction with alkenes, particularly intramolecular cycloadditions.9 It is also known that cis-alkenes are more reactive than trans-alkenes in reactions with ketenes.10 Cycloaddition reactions with cis-alkenes are stereospecific, but some loss of stereochemistry is observed in reactions with-trans alkenes.4a,9,11 R

R

R1

R

R1

Heat or

+ R

R

R R1

R

R1

or R

R1

hν

R1

R

R1

R

+

−

R

R1 1

R

R1 R1

R R1

R1

R1 R1

R

R1

R 3

2 Me

Me Me

R R1

Me

CO2Me

NMe2

Me

Me + N

Me

−

CO2Me

Me2N

CO2Me

Me

N, N,2-Trimethylprop1-en-1-amine

5

4

The C]C unit of ketenes (R2C]C]O) react with alkenes via thermal [2 + 2]-cycloaddition reactions to yield cyclobutanone derivatives. Ketenes are highly electrophilic due to a low-lying LUMO (π*-C]O), and the HOMO is relatively high in energy (see Table 14.1). Ketene itself has a HOMO at
12.55 eV and a LUMO at +2.55 eV.11,12 Those values, along with those of the orbital coefficients, are shown in Fig. 15.1. Polarization of the C]C unit in ketene (the combination of electrophilic and nucleophilic character) is responsible for the facile [2 + 2]-cycloaddition reactions.

–0.22

R R

d−

R R

d+

CK CKO d+

d−

CKCKO

0.74

R R

–0.57

0.27

CKCKO

0.75

+2.52 eV LUMO –0.61

R = H –12.55 eV

HOMO

FIG. 15.1 Orbital diagram for ketene.

The presence of an electron-withdrawing group is known to enhance the rate of the [2 + 2]-cycloaddition reaction. Inspection of Table 14.1 shows that the energy of the HOMO for ketene is
12.55 eV, the energy of the HOMO for PhCH]C]O is
10.61 eV, that for ClCH]C]O is
9.24 eV, and that for Cl2C]C]O is
9.15 eV. Clearly, the 5

Reference 4a, p 35.

(a) Dopper, J. H.; Greijdanus, B.; Oudman, D.; Wynberg, H. J. Chem. Soc. Chem. Commun. 1975, 972; (b) Meyers, A. I.; Tschantz, M. A.; Brengel, G. P. J. Org. Chem. 1995, 60, 4359; (c) Groaning, M. D.; Brengel, G. P.; Meyers, A. I. Tetrahedron 2001, 57, 2635. 6

7

Brannock, K. C.; Bell, A.; Burpitt, E. D.; Kelly, C. A. J. Org. Chem. 1961, 26, 625.

8

Reference 4a, p 33.

9

(a) Snider, B. B. Chem. Rev. 1988, 88, 793; (b) Snider, B. B.; Walner, M. Tetrahedron 1989, 45, 3171.

10

(a) Ghosez, L.; O'Donnell, M. J. In Pericyclic Reactions, Vol. II, Marchand, A. P.; Lehr, R. E., Eds.; Academic Press: New York, NY, 1977, pp 79–140; (b) Frey, H. M.; Isaacs, N. S. J. Chem. Soc. B 1970, 830.

11

Houk, K. N. J. Am. Chem. Soc. 1973, 95, 4092.

12

(a) Fleming, I. Frontier Molecular Orbitals and Organic Chemical Reactions; John Wiley, London, 1976; p 143; (b) Reference 4a, p 33–42.

803

15.2 [2 + 2]-CYCLOADDITION REACTIONS

presence of these groups raises the energy of the HOMO, and would be expected to increase the rate of a cycloaddition driven by the LUMO of an alkene. This effect has been noted in many intermolecular reactions of ketenes and alkenes.13 The attempted intramolecular cyclization of 6 (X ¼ H) failed (also see formation of 17). When the α-chloro analog (6, X ¼ Cl) was prepared, however, treatment with triethylamine led to 7, which cyclized to give a 55% yield of the [2 + 2]-cycloadduct (8) along with 19% of the ene adduct (9, see Section 11.13), where X ¼ Cl. Me

Me

Me Me Me

X

Me

NEt3 , PhH Reflux

Me

X

X

C

O

O

Cl 6

Me

Me

7

+

Me Me

Me

O

O 8 (55%)

9 (19%)

X

In a typical reaction, dimethylketene and ethene were heated to yield 2,2-dimethylcyclobutan-1-one in a reaction controlled by the HOMOalkene-LUMOketene interaction.14 Inspection of the orbital picture for this reaction (see 10) shows that [2 + 2]-cycloadditions are allowed when one component is suprafacial and the other is antarafacial. In 10, the HOMO of the alkene is of the correct symmetry to interact with the low-lying LUMO of the ketene.15a Reaction of an alkene and a ketene (see 10) yields a cyclobutanone (11).16 O CH2 CH2

O

Heat

C

+

Me Me 2,2-Dimethylcyclobutan-1-one

Me Me Dimethylketene

R O

R R

O R 10

11

Reprinted with permission from Tidwell, T. T. Ketenes; Wiley: New York, NY, 1995; p 487. Copyright © 1995 by John Wiley and Sons, Inc.

Ketenes (e.g., dimethylketene) are readily generated by the reaction of an acid chloride with an amine,16 or by treatment with ZndCu in the case of α-chloro acid chlorides. Since carboxylic acids and acid chlorides are readily available, ketenes are convenient partners for reaction with both alkenes and dienes.17 In Danishefsky and coworker’s18 synthesis of tricholomalide A, trichloroacetyl chloride reacted with Zn, with sonication, to yield dichloroketene in situ. Reaction with 12 gave dichlorocyclobutanone (13). Subsequent reduction of the chlorine moieties in 13 with zinc/acetic acid gave cyclobutanone (14) in >85% yield. It is also known that alkynes react with ketenes to form cyclobutene derivatives.19 O Me

Sonication, Ether CCl3COCl/Zn

Me C

Me AcOH, Zn 100°C

Me

O

Cl

12

Cl Cl Me

Cl

TIPSO

O Me

TIPSO

Dichloroketene

TIPSO 13

14 (>85%)

(a) Krepski, L. R.; Hassner, A. J. Org. Chem. 1978, 43, 2879; (b) Bak, D. A.; Brady, W. T. Ibid. 1979, 44, 107; (c) Martin, P.; Greuter, H.; Belluš, D. Helv. Chim. Acta. 1984, 64, 64; (d) Brady, W. T. Synthesis 1971, 415.

13

14

(a) Reference 4a, pp 38–40; (b) Sustmann, R.; Ansmann, A.; Vahrenholt, F. J. Am. Chem. Soc. 1972, 94, 8099.

15

(a) Tidwell, T. T. Ketenes; John Wiley & Sons, Inc.: New York, NY, 1995; p 487; (b) Also see Ref. 4a, p 39.

16

Brady, W. T. Tetrahedron 1981, 37, 2949.

17

For a reaction with cyclopentadiene, see Grieco, P. A. J. Org. Chem. 1972, 37, 2363.

18

Wang, Z.; Min, S. -J.; Danishefsky, S. J. J. Am. Chem. Soc. 2009, 131, 10848.

19

For a synthetic example, see Fex, T.; Froborg, J.; Magnusson, G.; Thoren, S. J. Org. Chem. 1976, 41, 3518.

804

15. PERICYCLIC REACTIONS

The intramolecular version of this reaction is usually more efficient than the intermolecular version. When 15 (R ¼ Me) was treated with triethylamine (0.2% solution of acid chloride in refluxing dichloromethane), ketene 16 (R ¼ Me) was formed and an intramolecular [2 + 2]-cycloaddition gave an 80% yield of 17.20 Ghosez and coworkers21 reported, however, that reaction of 15 (R ¼ H) under the same conditions gave only 3% of 17 (R ¼ H). Many unactivated ketones show similar poor reactivity, but Ghosez and coworkers21 developed a ketene iminium salt cyclization procedure that led to enhanced reactivity. In one example, the pyrrolidine amide 1-(pyrrolidin-1-yl)hept-6-en-1-one was treated with triflic anhydride [(CF3SO2)2O] and collidine to generate the ketene iminium salt (18). Subsequent internal [2 + 2]-cycloaddition gave 19, and after hydrolysis of the iminium salt product 17 was obtained in 75% yield. H R

Cl

R

NEt3

O

C O

O

R 16

15

17

C

Collidine

N

(CF3SO2)2O

H

N

N

H 2O CCl4

O

17 (75%)

H 1-(Pyrrolidin-1-yl)hept-6-en-1-one

18

19

Ghosez and Marchand-Brynaert22 showed that ketene iminium salts were more electrophilic than ketenes and did not dimerize. Another example of this approach is the conversion of 20 to 21 in 65% yield.23 There are many examples of internal [2 + 2]-cycloaddition of ketenes, primarily from the laboratory of Ghosez and coworkers,20 by Snider et al. (as in the conversion of 22 to 23),24 and by Brady and Giang.25 In both Snider’s and Brady’s examples, the presence of an α-alkoxy substituent led to a more facile cycloaddition. Vinyl ketenes also react well in the cycloaddition. For intramolecular [2 + 2]-cycloadditions involving vinyl ketenes, Snider defined three different reaction types: Type I (the ketene carbon has a side chain)26; Type II (a side chain is attached α to the ketene carbon)27; and Type III (a side chain is attached β to the ketene carbon).24a O Ts

N

H N

1. CH2Cl2, Reflux 1d

Ts

O

N

2. Hydrolysis

H OBn

BnO 20

21 (~65%) O O

O Cl

NEt3

H O

22

23 (58%)

An example of a Type I cyclization is taken from the synthesis of debromoflustramine B by Shishido and coworkers,28 in which carboxylic acid 24 was converted to ketene 25 by initial conversion to the acid chloride followed by treatment with triethylamine. Generated in situ, [2 + 2]-cyclization of the ketene gave 26 in 85% yield. 20

Markó, I.; Ronsmans, B.; Hesbain-Frisque, A. -M.; Dumas, S.; Ghosez, L. J. Am. Chem. Soc. 1985, 107, 2192.

21

Falmagne, J. B.; Escudero, J.; Taleb-Sahraoui, S.; Ghosez, L. Angew. Chem. Int. Ed. 1981, 20, 879.

22

Ghosez, L.; Marchand-Brynaert, J. Iminium Salts in Organic Chemistry, Part I; B€ ohme, J.; Viehe, H. G., Eds., John Wiley: New York, NY, 1976; pp 421–532.

23

Monache, G. D.; Misiti, D.; Salvatore, P.; Zappia, G.; Pierini, M. Tetrahedron: Asymmetry 2000, 11, 2653.

24

(a) Snider, B. B.; Hui, R. A. H. F.; Kulkarni, Y. S. J. Am. Chem. Soc. 1985, 107, 2194; (b) Snider, B. B.; Hui, R. A. H. F. J. Org. Chem. 1985, 50, 5167.

25

Brady, W. T.; Giang, Y. F. J. Org. Chem. 1985, 50, 5177.

26

Kulkarni, Y. S.; Burbaum, B. W.; Snider, B. B. Tetrahedron Lett. 1985, 26, 5619.

27

Kulkarni, Y. S.; Snider, B. B. J. Org. Chem. 1985, 50, 2809.

28

Ozawa, T.; Kanematsu, M.; Yokoe, H.; Yoshida, M.; Shishido, K. J. Org. Chem. 2012, 77, 9240.

805

15.2 [2 + 2]-CYCLOADDITION REACTIONS

O

OH 1.

Cl

OH

Cl O

O

PhH , Reflux

N

O

C O

N

2. NEt3 , PhH Reflux

CO2Me OH

H

N

CO2Me

CO2Me

24

26 (85%)

25

With vinyl ketenes, [1,5]-sigmatropic hydrogen shifts are sometimes a problem29 (Section 15.5.1). Snider used a Type III cyclization of this type in a synthesis of isocomene.30 Funk et al.31 synthesized clovene using this methodology, and Ernst and coworkers32 showed that this method was effective for the synthesis of triquinane derivatives. In another extension of this methodology, Halcomb and McCaleb33 reported an intramolecular ketene-allene cycloaddition. Ketenes react with imines via [2 + 2]-cycloaddition to produce β-lactams.34 An example is taken from a study of the synthesis and evaluation of cholesterol absorption inhibitors by Huang and coworkers,35 in which the reaction of methyl 5-chloro-5-oxopentanoate with tributylamine generated ketene 27 in situ, which subsequently reacted with imine 28 to yield β-lactam 29. Asymmetric synthesis is possible, and chiral ammonium salts derived from chiral quinidine compounds and Cinchona alkaloids have been used to catalyze the [2 + 2] cycloaddition of ketenes and imines, producing β-lactams with excellent enantioselectivity.36 Ketenes are useful in many asymmetric syntheses.37 N-Substituted isocyanates also undergo thermal [2 + 2]-cycloaddition reactions with alkenes, generating β-lactams.38 Aldehydes react with silylketenes to yield β-lactones, as in the conversion of 30 to 31.39 OMe

MeO

MeO2C

MeO2C

N

NBu3

Cl

MeO2C

28

Toluene Reflux

C

O

27

Methyl 5-chloro-5-oxopentanoate

Ph

H TBSO

N

O

O

29

SiMe3

Ph EtAlCl2, Ether •

O

TBSO

O

O O

Me3Si

30

31

Thermal [2 + 2]-cycloadditions lead to the following disconnections: R1

O

+ R2

R

R1

R1

O

R1

O

CO2H

C R

R

R1

+

N SO2R

O

R1

R1

O

C O

C N SO2R

R

R

29

Lee, S. Y.; Kulkarni, Y. S.; Burbaum, B. W.; Johnston, M. I.; Snider, B. B. J. Org. Chem. 1988, 53, 1848.

30

Snider, B. B.; Beal, R. B. J. Org. Chem. 1988, 53, 4508.

31

Funk, R. L.; Novak, P. M.; Abelman, M. M. Tetrahedron Lett. 1988, 29, 1493.

32

Veenstra, S. J.; De Mesmaeker, A.; Ernst, B. Tetrahedron Lett. 1988, 29, 2303.

33

McCaleb, K. L.; Halcomb, R. L. Org. Lett. 2000, 2, 2631.

R HO2C

For reviews of the formation of β-lactams, see (a) Brown, M. J. Heterocycles 1989, 29, 2225; (b) Isaacs, N. S. Chem. Soc. Rev. 1976, 5, 181; (c) Mukerjee, A. K.; Srivastava, R. C. Synthesis 1973, 327. For a review of cycloaddition reactions of imines, see (d) Sandhu, J. S.; Sain, B. Heterocycles 1987, 26, 777. 34

35

Wang, Y.; Zhang, H.; Huang, W.; Kong, J.; Zhou, J.; Zhang, B. Eur. J. Med. Chem. 2009, 44, 1638.

36

Taggi, A. E.; Hafez, A. M.; Wack, H.; Young, B.; Ferraris, D.; Lecktka, T. J. Am. Chem. Soc. 2002, 124, 6626.

37

Orr, R. K.; Calter, M. A. Tetrahedron 2003, 59, 3545.

(a) Johnston, D. B. R.; Schmitt, S. M.; Bouffard, A. F.; Christensen, B. G. J. Am. Chem. Soc. 1978, 100, 313; (b) Bouffard, A. F.; Johnston, D. B. R.; Christensen, B. G. J. Org. Chem. 1980, 45, 1130.

38

39

Fournier, L.; Kocienski, P.; Pons, J. -M. Tetrahedron 2004, 60, 1659.

806

15. PERICYCLIC REACTIONS

An intramolecular ketene-alkene cycloaddition has also been used to prepare highly substituted phenols by the conversion of a cyclobutenone to a ketene, which reacted with an alkyne40 to yield a new cyclobutenone. A subsequent four-electron electrocyclic cleavage reaction followed by a six-electron electrocyclic ring closure (and tautomerization) gave a phenol. This sequence is now known as Danheiser annulation.41 Electrocyclic reactions are discussed in Section 15.3. In a synthetic application taken from Danheiser and coworker’s42 formal synthesis of (+)-FR900482, alkyne 32 was heated in the presence of 33 to give 34 in 88–94% yield. Resorcinols43 have been prepared by this method. Liebeskind et al. reported a related method for the synthesis of highly substituted quinones.44 OBn

OBn

O

OH OTBDMS

OTBDMS

Toluene, 80–110°C

+

3h

N

N

OH

CO2t-Bu

OH 33

32

CO2t-Bu 34 (88–94%)

15.2.2 General Principles of Photochemistry It is known that a normal [2 + 2]-cycloaddition of two alkene moieties does not occur readily under thermal reaction conditions. In one sense, the thermal [2 + 2]-cycloaddition reactions noted in Section 15.2.1 can be considered highly useful special cases. In contrast to the thermal reaction, photochemically induced cycloaddition occurs with great ease. In the photochemical [2 + 2]-cycloaddition, promotion of an electron from the populated HOMO of the alkene to the unpopulated LUMO will generate an excited state. Orbital diagrams45 can be used to illustrate this process. The HOMO and LUMO for two ethane molecules are shown in Fig. 15.2. CH2=CH2 +0.71

CH2=CH2 –0.71 +0.71

+0.71

FIG. 15.2

+1.5 eV

+1.5 eV

–10.52 eV

–10.52 eV

–0.71

+0.71 +0.71

+0.71

The HOMO and LUMO for two molecules of ethane.

For a reaction to occur between two alkenes, the HOMO of one alkene must react with the LUMO of the second, but the HOMOalkene-LUMOalkene interaction is symmetry forbidden. When a photon of light is absorbed by an alkene, (a) Danheiser, R. L.; Gee, S. K. J. Org. Chem. 1984, 49, 1672; (b) Danheiser, R. L.; Brisbois, R. G.; Kowalczyk, J. J.; Miller, R. F. J. Am. Chem. Soc. 1990, 112, 3093.

40

41

Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 196–197. 42

Mak, X. Y.; Crombie, A. L.; Danheiser, R. L. J. Org. Chem. 2011, 76,1852.

43

Danheiser, R. L.; Nishida, A.; Savariar, S.; Trova, M. P. Tetrahedron Lett. 1988, 29, 4917.

44

Liebeskind, L. S.; Iyer, S.; Jewell, C. F., Jr. J. Org. Chem. 1986, 51, 3065.

45

Reference 12a, pp 128, 219.

807

15.2 [2 + 2]-CYCLOADDITION REACTIONS

an electron is promoted from the filled HOMO to the empty LUMO. The higher energy orbital is now a singly occupied molecular orbital (an SOMO, see Section 2.4.2), and is of the correct symmetry to interact with the LUMO of another alkene. In other words, the cycloaddition occurs by the SOMOalkene-LUMOalkene interaction, which is now symmetry allowed (the SOMO represents the excited state). Before discussing photochemical reactions, some general principles of photochemistry will be introduced. Photochemistry is “concerned with the chemical change that may be brought about by the absorption of light.”46 Chemical change is not a requirement, however, since fluorescence (light emitted from a species that has absorbed radiation) and chemiluminescence (light emitted as a product of a chemical reaction) can occur.46 If a molecule absorbs a photon of light, the photoactivated species can undergo a [2 + 2]-cycloaddition reaction or a variety of other reactions (see Sections 17.4.5). The application of light increases the energy of a molecule according to E ¼ hv where v ¼

c λ

(15.1)

where h is Planck’s constant (6.63  10
34 J s), c is the speed of light (3  108 m s
1), ν is its frequency in s
1, and λ is the wavelength of the light in nm, which is inversely proportional to frequency. As E is usually expressed in kcal mol
1, the molecular value of E calculated from Eq. (15.1) must be multiplied by Avogadro’s number, NA ¼ 6.02  1023 mol
1. This expression gives E ¼ (2.86  104/λ(nm)) kcal mol
1, including conversion factors. An energy of 1 kcal (4.18 kJ) mol
1 corresponds to a wavelength (λ) of 2.86  105 Å, and in general47: 1kcal mol
1 ¼

2:86  105 kcal mol
1 ˚   and λ ¼ 2:86  105 A ˚ λ A

(15.2)

This wavelength corresponds to a wavenumber (ν) given by ˚ cm
1 c 108 A v¼ ¼ ¼ 353 cm
1 and ¼ 1kcal mol
1 ˚ λ 2:86  105 A

(15.3)

With respect to energy, 1 kcal mol
1 ¼ 2.86  105 Å ¼ 353 cm
1, which is an absorption in the IR region of the electromagnetic spectrum. An energy of 10 kcal (41.8 kJ) mol
1 corresponds to 3530 cm
1, also in the IR region. Light of wavelength 2000 Å is at the lower end of the UV region (see Fig. 15.3) and corresponds to 143 kcal (598.3 kJ) mol
1. Light in the UV and visible region can excite molecules to higher electronic states. Wavelengths in the UV are usually expressed in nanometers (1 nm ¼ 1 mμ ¼ 1  10
7 cm), and the UV region is 200–400 nm. Red

Violet

UV 10 2860 11,972 100 1000 × 103

FIG. 15.3

Near UV 200 143 598.6 2000 50 × 10 3

IR

Vis 400 71.5 299.3 4000 25 × 10 3

800 35.75 149.7 8000 12.5 × 10 3

2860 10 41.86 2.86 × 104 3530

28,600 1 4.186 2.86 × 105 353

n (mμ = nm) E (kcal mol–1) E (kJ mol–1) l (Å) n (cm–1)

A portion of the electromagnetic spectrum.

The wavelength of light that must be used for a particular molecule must be determined. Each functional group absorbs light according to its individual structure. If that absorption is in the visible, UV or the IR, a visible, a UV, or an IR spectrum will provide the wavelength of light. Ultraviolet light is probably used most often, and the wavelength of maximum absorption (λmax) is a good guide as to which wavelength of light that should be chosen for

46

Wayne, R. P. Principles and Applications of Photochemistry; Oxford University Press: Oxford, 1988; p 1.

47

DePuy, C. H.; Chapman, O. L. Molecular Reactions and Photochemistry; Prentice-Hall: Englewood Cliffs, NJ, 1972; p 5.

808

15. PERICYCLIC REACTIONS

irradiation. Table 15.148 shows the energy of absorption in the UV (λmax) for several common types of organic molecules, along with the energy in kcal mol
1 and kJ mol
1. TABLE 15.1 Absorption Maxima for Representative Molecules and Functional Groups Molecule

C 4H9I

Transition

max (nm)

E (kcal mol )

E (kJ mol )

*

224

127.7

534.6

CH2=CH 2

*

165

173.3

725.4

HC CH

*

173

165.3

691.9

Acetone

*

150

190.7

798.3

n

*

188

152.1

636.7

n

*

279

102.5

429.1

CH2=CHCH=CH 2

*

217

131.8

551.7

CH2=CHCHO

*

210

136.2

570.1

*

315

90.8

380.1

*

180

158.9

665.2

200

143.0

598.6

255

112.2

469.7

n

n Benzene

Functional Group

RCH=CHR

RC CR

RCHO

max (nm)

E (kcal mol )

E (kJ mol )

165

173.3

725.4

193

148.2

620.4

173

165.3

691.9

188

152.1

636.7

279

102.5

429.1

290

98.6

412.7

RCOOH

99:1 favoring 329).319a,320 R 400°C

R = H, 490°C R = CO2Me, 400°C

CH2JR

No reaction R=H

Me

H

Me 326 Me

327 Me

Me

Me

180°C, 7 h ZnBr2

H MeO2C

(75%)

+ MeO2C

CO2Me 328

318

Me CO2Me 329

Me

MeO2C CO2Me 330

Hunstman, W. D.; Lang, P. C.; Madison, N. L.; Uhrick, D. A. J. Org. Chem. 1962, 27, 1983.

(a) Tietze, L. F.; Beifuss, U. Angew. Chem. Int. Ed. 1985, 24, 1042; (b) Idem Liebigs Ann. Chem. 1988, 321; (c) Nakatani, Y.; Kawashima, K. Synthesis 1978, 147; (d) Sakane, S.; Maruoka, K.; Yamamoto, H. Tetrahedron 1986, 42, 2203.

319

320

Tietze, L. F.; Beifuss, U.; Ruther, M. J. Org. Chem. 1989, 54, 3120.

856

15. PERICYCLIC REACTIONS

15.6.3 Carbonyl and Imino Ene Reactions When a carbonyl moiety is the ene component of a 1,6- or 1,7-diene system, the ene reaction is more favorable. Heating 331 to 110°C, for example, led to a 94% yield of 332 in White and Somers’321 synthesis of 2-desoxystemodinone. In 331, the 1,6-diene is the π-bond of the aldehyde carbonyl and the π-bond of the cyclohexene unit, which is more easily seen in the conformational drawing. The reaction occurred in refluxing toluene in good yield, and with excellent stereoselectivity due to the rigid nature of the tricyclic system. H

Me O H L Me

H

Toluene 110°C, 16 h

Me

Me OH H O

OH

OH

O H

H

H

H

L

OH Me

OH H

H

331

332 (94%)

The carbonyl ene reaction is catalyzed by Lewis acids.322 Isobutylene reacted with methyl propiolate at 220°C to give 47% of 333 and 3% of 334. The presence of AlCl3, however, allowed reaction to occur at 25°C and gave a 61% yield of 333.323 Coordination of the Lewis acid and enophile lowers the LUMO, accelerating the reaction due to the diminished ΔE, just as in the Diels-Alder reaction. The C]O of an aldehyde reacts as an ene. The Me2AlCl catalyzed ene reaction of 2-OTBS acetaldehyde with 335 gave 336 in 68% yield as part of Schmalz and coworkers324 synthesis of helioporin C. Alkenyl ketones also undergo ene reactions. As with the conversion of 335 to 336, dimethylaluminum chloride (Me2AlCl) and diethylaluminum chloride (Et2AlCl) are excellent catalysts for the carbonyl ene reaction and are often the catalyst of choice. When catalyzed by chiral Cr(III) complexes, high asymmetric induction is possible in hetero-ene reactions of this type.325 CO2Me

+ H

MeO2C

+

H CO2Me

Methyl propiolate

333

334

(47%) (61%)

230 °C AlCl3, 25°C

(3%) –

Me

Me OTBS H

Me

H

H O

+

O

O Me 2-OTBS acetaldehyde

O

cat Me2AlCl

335

O

HO Me OTBS

336 (68%)

The choice of catalyst is important for the reaction. The reaction of 337 with formaldehyde gave 338, but when ZnCl2, Me3Al, AlCl3, EtAlCl2, and Et2AlCl were examined, only Et2AlCl gave good yields (61%).326 The selectivity in the reaction is due to the more favorable interaction of the aluminum and aryl moieties in the transition state. Good selectivity was also observed in the reaction of but-(2E)-ene with methyl 2-oxoacetate. Antiselectivity to give 339 was observed when tin chloride (SnCl4) catalyzed the reaction with either cis- or trans-but-2-ene, but syn selectivity (to 340) was observed when alkyl aluminum catalysts were used (Me2AlCl, Me2AlOTf, MeAl(OTf )2).327

321

White, J. D.; Somers, T. C. J. Am. Chem. Soc. 1994, 116, 9912.

322

Snider, B. B. Acc. Chem. Res. 1980, 13, 426.

(a) Snider, B. B. J. Org. Chem. 1974, 39, 255; (b) Greuter, H.; Bellus, D. Synth. Commun. 1976, 6, 409; (c) Åkermark, B.; Ljungqvist, A. J. Org. Chem. 1978, 43, 4387; (d) Snider, B. B. J. Org. Chem. 1976, 41, 3061; (e) Snider, B. B.; Rodini, D. J.; Conn, R. S. E.; Sealfon, S. J. Am. Chem. Soc. 1979, 101, 5283. 323

324

L€ olsberg, W.; Werle, S.; Neud€ orfl, J. -M.; Schmalz, H. -G. Org. Lett. 2012, 14, 5996.

325

Ruck, R. T.; Jacobsen, E. N. J. Am. Chem. Soc. 2002, 124, 2882.

326

Prashard, M.; Tomesch, J. C.; Shapiro, M. J. Tetrahedron Lett. 1989, 30, 4757.

327

Mikami, K.; Loh, T. -P.; Nakai, T. Tetrahedron Lett. 1988, 29, 6305.

857

15.6 THE ENE REACTION

MeO

MeO

OMe

OMe (CH 2O)n , rt

MeO

Lewis acid CH2Cl2

OMe MeO

MeO MeO

OMe

HO

OMe

OMe

337

338 OH

O Me

+ Me

OMe

OMe

H

Me

O But-(2E)-ene

OH

Methyl 2-oxoacetate

OMe

+

O

Me

339

O

340

Imino-ene reactions are known.328 In one report, lanthanum derivatives catalyzed the reaction. When imine 341 reacted with α-methyl styrene in the presence of Yb(OTf )3, for example, a 90% yield of the ene product (342) was obtained when chlorotrimethylsilane was used as an additive.329 Catalytic, enantioselective imino-ene reactions of α-imino esters have also been reported.330 Me

SO2Tol

N Ph

Ph

, 25% Yb(OTf)3

TolSO2HN

Me3SiCl, CH2Cl2 –THF , rt

H

Ph

Ph 342 (90%)

341

15.6.4 Chiral Ene Reactions The fundamental principles of asymmetric synthesis employed in previous chapters and to reactions in this chapter also apply to ene reactions. Chiral additives, chiral auxiliaries, or the preparation of chiral templates all lead to good enantioselectivity when applied to the ene reactions. Most chiral-ene reactions seem to involve addition of a chiral catalyst or the use of a chiral auxiliary. Oppolzer et al.331 utilized a chiral auxiliary built into 343 to prepare 344 (90 %de) in a synthesis of (+)-α-allokainic acid. Me Me

Ph

H

O O

Me

Me N

AlEt2Cl

O

EtO2C 343

Me Me

O

Ph

H Me

O Me

CO2Et X c

H EtO2C 344

N CO2Et

O Xc

Yamamoto and coworkers332 used the chiral Al catalyst 345 for the intermolecular ene reaction of 1,6dichlorobenzaldehyde and 2-phenylthioprop-1-ene to give 346 in 96% yield (65 %ee). These catalysts are similar to those used in Section 14.6.2. A similar Ti catalyst (347) facilitated the coupling of methylenecyclohexane and methyl glyoxalate to give an 89% yield of 348 (98 %ee). Catalyst 347333 was prepared in situ by reaction of (R)-binaphthyl and bis(diisopropoxytitanium) dibromide [(i-PrO)2TiBr2]. These few examples are sufficient to illustrate that chiral induction is possible in the ene reaction, using the same techniques that were successful in other pericyclic reactions. For examples, see (a) Achmatowicz, O., Jr.; Pietraszkiewwicz, M. J. Chem. Soc. Chem. Commun. 1976, 484; (b) Borzilleri, R. M.; Weinreb, S. M. Synthesis 1995, 347; (c) Laschat, S.; Grehl, M. Angew. Chem. Int. Ed. 1994, 33, 458; (d) Mikami, K.; Kaneko, M.; Yajima, T. Tetrahedron Lett. 1993, 34, 4841. 328

329

Yamanaka, M.; Nishida, A.; Nakagawa, M. Org. Lett. 2000, 2, 159.

(a) Drury, W. J., III; Ferraris, D.; Cox, C.; Young, B.; Lectka, T. J. Am. Chem. Soc. 1998, 120, 11006; (b) Yao, S.; Fang, X.; Jørgensen, K. A. Chem. Commun. 1998, 2547.

330

(a) Oppolzer, W.; Robbiani, C.; B€attig, K. Helv. Chim. Acta 1980, 63, 2015; (b) Idem Tetrahedron 1984, 40, 1391; (c) Oppolzer, W.; Mirza, S. Helv. Chim. Acta 1984, 67, 730.

331

332

Maruoka, K.; Hoshino, Y.; Shirasaka, T.; Yamamoto, H. Tetrahedron Lett. 1988, 29, 3967.

333

Mikami, K.; Terada, M.; Nakai, T. J. Am. Chem. Soc. 1989, 111, 1940.

858

15. PERICYCLIC REACTIONS Cl

SiPh3

CHO

O

Cl , MS 4 Å

OH

Cl

Al Me SPh

O

, CH2Cl2 , –78°C, 2 h

SPh

Cl

Me

346 (96%, 65 %ee)

345

O

Br

OH

, MeO2CJCHO

Ti CO2Me

MS 4 Å CH2Cl2 , 3 h

Br

O

348 (89%, 98 %ee)

347

15.7 CONCLUSION It is clear from the preceding sections that a wide range of cyclic and polycyclic molecules can be prepared by pericyclic reactions. In addition, stereocontrolled syntheses of acyclic molecules are possible via initial cycloaddition followed by cleavage of the ring. The power of these reactions lies in the ability to make carbon bonds with high regioselectivity and stereoselectivity. If the proper chiron or chiral auxiliary is used, high enantioselectivity can be obtained as well. Few synthetic methods allow access to such a large number of natural products, with compatibility with such a large number of functional groups. In succeeding chapters, methods for generating carbon bonds via electrophilic or radical intermediates will be presented. HOMEWORK

1. Give a mechanistic explanation for the following transformation: O O

O

LiN(TMS)2 Me2t-BuSiCl

H 2O

HMPA, THF

THF

O

O MeO

O

HO2C

MeO

2. Give a mechanistic rationale for the transformation that accounts for one product being formed preferentially and also accounts for the stereochemistry of both products. D D

220°C

Ph H

H

D

+ H

Ph

(86%)

(14%)

3. Explain the different stereochemical outcomes of these reactions. H H CH2KCKCH2 hv

O

O H H CH2KCKCH2

O

hv

Ph

O

859

15.7 CONCLUSION

4. Explain the following transformation: Me

Me MeO2C

Heat

CO2Me Me

Me

5. Explain the following transformation: OTBS

Ph NaIO4, MeOH

O

PhMe2Si

NaHCO3, H2O

O

Ph

OMe DBU, Xylene Sealed tube 185°C

SePh

O

PhMe2Si O

6. Draw a reasonable transition state for the following reaction: MOMO BnO

MOMO

Neutral Al2O3, 60°C

O

O

O

O

OBn

7. Give the product for this reaction, and rationalize its formation. AcO

O Me

Me

OTES hv

O O AcO BzO

8. Reaction of N-tosylimines and allylic sulfonium salts under basic conditions leads to aziridines in this particular work. When the imine shown was treated with the sulfonium salt given in the reaction, a mixture of aziridine and a dihydroazepine were formed. Draw both products and explain formation of the dihydroazepine. N Ts Ph

+

Me2S

Ph

KOH, MeCN

9. Use the tables in this chapter (where possible) to estimate the correct regiochemistry for the products in the following reactions:

(A) (B)

+

C5H11JCLNJO− +

MeJCLNJO− −O OMe

(C) +

N Me

+ +

HJCLCJ(CH2)12Me S

S CHO

+

CO2Me

860

15. PERICYCLIC REACTIONS

10. For each of the following reactions give the major product, with correct regiochemistry and stereochemistry:

(A)

Ph

BnO

OMOM

NaIO4

t-Bu

H2O-DMF

O

Ph

(C)

O−

N

NHOH

SiMe3

Ph BnO

HO

(G)

BnO

O

140°C, 3 h

O

1. SOCl2

CO2H

5% EtCO2H/EtC(OEt)3

(F)

Me

H

N3

HO

300°C, 1 h

(E)

(I)

CO2Et

O

N

Na2SO3

OBz O

Br

ClCO2NKCKO

SiPh2t-Bu

(H)

2. CH2KCHOt-Bu NEt3, MeCN, 80°C

t-BuO

1. Toluene, 110°C 2. hv, THF

OH

(D)

Me Me

OBn

OBn

Ot-Bu

Neat, 50°C

H

PhH , Reflux

N

(B)

O Benzophenone, hv

(J)

O

Toluene, rt

OBn O PhOPh, Reflux

(K) Bn

hv, Acetone

(L)

H

N

HCLCH

CO2Bn

t-BuMe2SiO

i-Pr

OMe O

(M) O

hv, Acetone

N

(CHO) n CSA

(N)

Na2SO 4

NHMe EtO2C

Me

H Toluene, Heat

(O)

OTBS

(P)

Bn

N

5% Cp2Ti(CO)2 Toluene, Reflux

(Low yield)

Me

O Ts NaCl-AlCl3 Melt 180–190°C

(Q) O

O O

5 min

H 520°C, 0.01 mmHg

(S) O

CO2Et

(R)

Bn

N

SmI2, THF-HMPA PhCLCPh

Ts

Hexane, 6 h

861

15.7 CONCLUSION

11. In each case, provide a suitable synthesis. Show all intermediate products and all reagents. OMe

O O

(A)

OH

O

H H

(B)

MeO

O

OMe

OMe

OMe

MeO MeO MeO

MeO

MeO

(C)

CHO

(D) MeO

N

MeO

MeO

O N

O

H

MeO O

OBn

(E)

Me

HO

(F)

O OBn HO

O

Br MeO

(G)

O O

MeO

OTMS MeO MeO

O

12. In each case, devise a retrosynthetic strategy that will lead to a suitable starting material that costs $1.00–5.00/g. Show the complete synthesis, with all intermediate products and all reagents. O

OH

O

Me

N

Me Me OH

(A) OMe

Me

OH O H

(E)

(B)

(C) Me

OH

Me Me

Me Me

Me

O

(G)

(F) O

Me

OHC

Me

H

Me Me

N

(D)

O

Me Me

Et

Me

OH OHC

Me Me CHO

O

Me

(H)

OHC

Me Me CO2Me

C H A P T E R

16 Carbon-Carbon Bond-Forming Reactions: Carbocation and Oxocarbenium Ion Intermediates 16.1 INTRODUCTION Many reactions used to make carbon-carbon bonds were introduced in previous chapters. Molecules possessing a polarized electrophilic carbon (δ+) were common in previous chapters, as with the reaction of an alkyl halide or a carbonyl compound with nucleophiles or Grignard reagents in Chapter 11, or with nucleophilic enolate anions in Chapter 13. In Section 1.2, disconnection of carbon-carbon bonds generated Cd and Ca fragments and Chapters 11–13 focused on Cd fragments. This chapter will focus on reagents and reactions that effectively generate equivalents for a Ca fragment, including substrates possessing a Cδ+ unit. Carbocations obviously correspond to a Ca fragment, but they are reactive products (intermediates) rather than reagents. However, if carbocations can be stabilized, they may be used in ways similar to reagents. This concept will be elaborated later in this chapter. The structure and nature of carbocations in the context of SN1 reactions was introduced in Section 3.2.2. It was clear that the nucleophilic partners reacting with a carbocation were limited to alkenes, alkynes, aromatic rings or compounds containing heteroatoms (e.g., oxygen, sulfur, halogen, or nitrogen). Reactions of carbocations with these species will also be discussed in this chapter, with an emphasis on their synthetic utility for making carbon-carbon bonds. This chapter will begin with a discussion of free carbocations and exploitation of their molecular rearrangements to generate carbon-carbon bonds and different structures.

16.2 CARBOCATIONS A carbocation is perhaps best viewed as an sp2 hybridized carbon bearing three substituents, with an empty p-orbital, perpendicular to the plane of the other atoms (see 1), with a formal charge of +1. In other words, it is an electrophilic species. The sp2 hybridization dictates a trigonal-planar geometry for 1. This intermediate has been termed a carbenium ion by Olah.1 Note that the old term carbonium ion has been applied to 1, but carbocation is the more descriptive term and will be used throughout, in lieu of carbenium or carbonium ion.

16.2.1 Cation Stability From the standpoint of reactivity, a carbocation is an intermediate (a transient product), and it is high in energy relative to starting materials and products, relatively unstable, and very reactive. Indeed, a carbocation reacts with reagents that will provide a pair of electrons to form a new covalent bond to the electrophilic carbon. In solvolysis or other ionization reactions, formation of a carbocation is usually the slowest (the rate-determining step) when compared to subsequent reactions with a nucleophilic species, which are fast. The classical SN1 reaction of tert-butyl bromide and aqueous potassium iodide (KI) to yield tert-butyl iodide is an example where the slowest step 1

Olah, G. A. J. Am. Chem. Soc. 1972, 94, 808.

Organic Synthesis http://dx.doi.org/10.1016/B978-0-12-800720-4.00016-7

863

Copyright © 2017 Michael Smith. Published by Elsevier Inc. All rights reserved.

864

16. CARBON-CARBON BOND-FORMING REACTIONS

(k1, Section 3.2.2) is ionization of the bromide to the planar carbocation 1. This reaction is slow because the bromide atom must be “pulled” from the alkyl halide, which is formally an ionization reaction. This ionization step is facilitated by the water, which can hydrogen bond to the bromine, effectively pulling on it, which is assisted by the ability of water to solvate ions after ionization occurs (see Section 3.2.2). The second reaction, trapping the cation, is the fastest step (k2), so k2 >>>>> k1. Heterolytic bond cleavage of a halide or sulfonate ester (CdX, X ¼ Br, Cl, I, OMs, OTs, etc.) in aqueous solvents (solvolysis) often requires heating. Note that 1 is planar with respect to C+, so there is no facial bias when the nucleophilic iodide ion reacts by donating two electrons to form a new CdI bond. Note also that the iodide product formed will have a new stereogenic center. If there is no facial bias, the iodide ion will approach from either face, leading to a 1:1 mixture of the stereoisomeric iodides. In other words, the iodide product will be racemic. There are many different methods for generating carbocations in addition to the classical ionization of an alkyl halide associated with SN1 reactions, and the rate-determining step can vary with each method. One of the most common methods to generate a carbocation is to treat an alkene with a Brønsted-Lowry acid (e.g., HCl). In this reaction, the π-bond of an alkene (e.g., 2-methylbut-1-ene) reacts as a Brønsted-Lowry base in the presence of the acid (H+) to generate carbocation 2, as identified in Section 2.5.1. The subsequent reaction with the nucleophilic chloride ion is similar to the sequence shown for carbocation 1, and the overall reaction leads to the final product, 2-chloro-2-methylbutane. Traditionally, this reaction is called an addition reaction, but in fact it is two chemical reactions: An acid-base reaction followed by reaction of a nucleophile with a carbocation intermediate. Alkynes also have π-bonds, and some similarity in reactivity is expected. Indeed, the reaction of an acidic reagent (HX) to an alkyne generates a vinyl carbocation intermediate (Section 2.5.1) and a vinyl halide (XdC]C) as the final product. -

Br

R2

-

R3

R1

O

+

R1

H

H

R1

I– k2

k1

R2

R1 I

R2

+

R3

R3

I

R2 R3

1 Cl–

H—Cl

Cl 2-Methylbut-1-ene

2

2-Chloro-2methylbutane

3

The reaction of 2-methylbut-1-ene with HCl to yield 2 raises the issue of regioselectivity in formation of a carbocation. When the π-bond shown attacks HCl, the proton can attach to either carbon of the C]C unit to form a primary carbocation (3) or the tertiary carbocation. 2. A methyl carbocation has three hydrogen atoms attached to C+. A primary carbocation abbreviated 1°, has one carbon group, an R group) and two hydrogen atoms attached; a secondary carbocation, abbreviated 2°, has two carbon groups and one hydrogen atom attached; a tertiary carbocation, abbreviated 3°, has three carbon groups and no hydrogen atoms attached. The reaction with 2-methylbut-1-ene yields the tertiary carbocation rather than the primary. How does one know this statement is true? Experimentally, the product is 2-chloro-2-methylbutane, which must arise from the tertiary carbocation 2, but 1-chloro-2-methybutane must arise from a primary carbocation, 3. In general, tertiary carbocations are formed easily, secondary carbocations with difficulty, and primary carbocations with extreme difficulty. There are two possible reactions when 2-methylbut-1-ene reacts with HCl, one pathway to generate a 1° carbocation and another to a 3° carbocation. If the major product arises from the 3° carbocation, then the transition state leading to formation of carbocations 3 must be lower in energy than that leading to 3. The lower energy pathway to the tertiary carbocation ensures that it will be formed with high selectivity. H

H H

Methyl carbocation

H

R H

Primary carbocation

H

R R

Secondary carbocation

R

R R

Tertiary carbocation

Reactions of Brønsted-Lowry acids (HX) with alkenes and alkynes will always give the more stable cation, which will lead to the major product. The order of stability is 3° > 2° > 1°. Dipolar effects (inductive and field) largely explain this order, although hyperconjugation (see 4) is also invoked. If the focus is on dipolar effects, carbocations 1 and 2 can be viewed as carbon atoms with an empty p-orbital. When a carbon group is attached to C+, redistribution of electron

865

16.2 CARBOCATIONS

density in the CdC+ bond will be toward the positive center, which is an inductive effect. In other words, electron density from bonds adjacent to the electron deficient C+ is distorted toward C+. This inductive effect will diminish the net charge, which stabilizes the cationic center. For carbocation 1, when R1, R2, and R3 in 1 are alkyl groups, the carbon atoms of the alkyl substituents release electrons to the positive center. The inductive effect is greater with increasing numbers of alkyl groups, leading to greater stability. Therefore, a 3° cation (R1 ¼ R2 ¼ R3 ¼ alkyl in 1) will be more stable than a 2° cation (R1 ¼ R2 ¼ alkyl, R3 ¼ H), which is more stable than a 1° cation (R1 ¼ R2 ¼ H, R3 ¼ alkyl). The least stable cation in this alkyl series will be the methyl cation (R1 ¼ R2 ¼ R3 ¼ H in 1). The energy difference between a 1° and a 2°, or a 2° and a 3° carbocation is 11–15 kcal (46.1–62.8 kJ) mol1.2 It is possible to rationalize the stabilization of a carbocation by hyperconjugation.3 The overlap of the end p-orbitals of the two ethylene units that are to be attached leads to conjugation, which in turn leads to a lowering of the energy of the system. If a σ-orbital overlaps with a π-orbital, then one has “hyperconjugation.”4 This effect can be illustrated by a typical case (e.g., carbocation 4), where hyperconjugation is invoked to explain relative stability as attached groups are varied. H+

H R H

R R

H H

R

H

4

In hyperconjugation, the σ-orbital lies at a lower energy than the π-orbital, and the electrons delocalize out of the σ-orbital to a lesser extent than from a π-orbital in conjugation. Therefore, hyperconjugative forms make small but definite contributions to the ground state of a molecule.5 Hyperconjugation is probably important for carbocations, for free radicals,6 and for excited states of molecules. It has been stated that hyperconjugation is most important in stabilizing carbocations when the relevant CdC bond(s) have >75% p-character.7 Note that a σ-bond overlaps a π-bond in 4, and an alkene and a closely bound proton constitute a canonical form that helps stabilize the carbocation. Indeed, if the orbitals of an adjacent CdH bond align with the empty orbital of the positive center; a canonical form can be drawn by electron donation to C+ that is formally an alkene and H+, as shown. It is important to understand that resonance contributors involving the CdH bonds are used to represent hyperconjugation, in which those bonds are elongated, which is a consequence of hyperconjugation. Hyperconjugation is actually a bond-stretching effect,8 and bond elongation can be viewed as a hyperconjugation signature for stabilization of a carbocation.3 Bond elongation is found in propene, and is due to hyperconjugation and represented as the canonical forms shown for 5, where the charge separation illustrates bond elongation. There is evidence in favor of hyperconjugation in the ground states of neutral molecules,9 and hyperconjugation appears to operate for both carbon and hydrogen in various systems.10 In principle, the more carbon atoms attached to C+ in a carbocation, the greater the hyperconjugation effects that will contribute to greater stability H

H C C

H

C

H H

C C

H

C

5

(a) Lossing, F. D.; Semeluk, G. P. Can. J. Chem. 1970, 48, 955; (b) Radom, L.; Pople, J. A.; Schleyer, P. v. R. J. Am. Chem. Soc. 1972, 94, 5935; (c) Arnett, E. M.; Petro, C. Ibid. 1978, 100, 5408.

2

3

See Radom, L.; Poppinger, D.; Haddon, R. C. In Carbonium Ions; Olah, G. A.; Schleyer, P. v. R., John Wiley & Sons, Inc.: NY, 1976; Vol. 5, pp 2303–2426. 4

Smith, M. B. March’s Advanced Organic Chemistry, 7th ed.; John Wiley & Sons, Inc.: NJ, 2013; pp 85–89.

5

Pauling, L.; Springall, H. D.; Palmer, K. J. J. Am. Chem. Soc. 1939, 61, 927; Wheland, G. W. J. Chem. Phys. 1934, 2, 474.

6

Symons, M. C. R. Tetrahedron 1962, 18, 333.

7

Jensen, F. R.; Smart, B. E. J. Am. Chem. Soc. 1969, 91, 5686.

8

Allinger, N. L. J. Comput. Aided Mol. Des. 2011, 25, 295.

9

See Laube, T.; Ha, T. J. Am. Chem. Soc. 1988, 110, 5511.

10

Allinger, N. L. Molecular Structure: Understanding Steric and Electronic Effect from Molecular Mechanics; John Wiley & Sons, Inc.: Hoboken, NJ, 2010.

866

16. CARBON-CARBON BOND-FORMING REACTIONS

Another useful method to generate a carbocation is the reaction of a secondary or tertiary alcohol with an acid catalyst, which initially yields an onium salt (oxonium ion 6), but loses water (a good leaving group) to give the carbocation. Not surprisingly, 3° alcohols react faster than 2° alcohols. Reaction of amines with nitrous acid (HONO) initially gives an alkyl diazoalkane 7, which readily decomposes to a carbocation.11 When generated under different conditions, a carbene can be generated from 7 (see Section 17.9.2). R

H

R

O

R

H+

H

R

R

R

– H 2O

O

R

H

R

R

6 R R R

R R R

HONO

NH2

R

–N2

R

N2

R 7

The discussion that concerned carbocation stability only addressed intermediates with carbon groups attached to the C+. It is possible to attach substituents that will stabilize or destabilize the carbocation via inductive effects. Electron-releasing substituents (e.g., dOH, dOR, dSR (there is a negative or δ charge, i.e., high electron density) will stabilize the charge when attached to the positive carbon of a carbocation, leading to a more stable ion. Both an oxygen-stabilized cation (8; known as an oxocarbenium ion) and a sulfur-stabilized cation (see 9) are known. The reaction of aldehydes and ketones with an acid catalyst (Brønsted-Lowry or Lewis), for example, gives the oxocarbenium ions 10 or 11, respectively (see Section 4.2). R1 C

R2

R1

O

R2

R 8

O

R 9

O—H

H+

R1

R

S

C

O—H

R1

R

R1

R 10

R

O

BF3

O R1

R

O

BF3 R1

R

BF3 R1

11

The stability provided by an adjacent oxygen atom is apparent in the solvolysis data for substituted primary chloroalkanes.12 When compared to formation of the primary cation derived from 1-chlorobutane (CH3CH2CH2CH2+), solvolysis of EtOCH2Cl to give oxocarbenium ion EtOCH2+ is much faster (109 vs 1).12 Stabilization by the proximal oxygen atom is a significant contributor to the greater rate of solvolysis for chloromethyl ethyl ether. On the other hand, moving the EtO group one atom further from the carbon bearing the chlorine has the opposite effect. The electron-withdrawing effect at the α-carbon in EtOCH2CH2Cl destabilizes the resulting carbocation and lowers the rate (0.2 vs 1 for 1-chlorobutane). A nitrogen atom to the C+ of a cation provides substantial stabilization (see putative resonance form 13). Indeed, the actual structure is called an iminium salt (12) which is a discreet, usually isolable, and well-characterized species with a formal π-bond between the carbon and nitrogen. In other words, it is an unstable (highly reactive) product and not necessarily a transient product, depending on the substituents that are attached. In terms of chemical reactivity, it is useful to consider iminium salts (e.g., carbonyl-like, C]O vs C]N) since they react with nucleophiles at the electrophilic carbon. Structures, such as 13 are formally “amino cations” (considered to be a resonance contributor of 12) and are not usually invoked in a mechanistic pathway. (a) Collins, C. J. Acc. Chem. Res. 1971, 4, 315; (b) Friedman, L. In Carbonium Ions; Olah, G. A.; Schleyer, P. v. R., Eds.; John Wiley: NY, 1970; Vol. II, p 655; (c) Kirmse, W. Angew. Chem. Int. Ed. Engl. 1976, 15, 251.

11

12

See Streitweiser, A., Jr. Solvolytic Displacement Reactions, McGraw-Hill: NY, 1962; pp 102–103.

867

16.2 CARBOCATIONS

R1

R1

R

R

N R2

N R2

R 12

R 13

In general, a heteroatom attached to the positively charged carbon stabilizes a carbocation. If the cationic center bears an electron-withdrawing group, where the (+) and δ+ centers repel, the carbocation is destabilized, as in 14. Destabilization of a carbocation is usually associated with difficulty in formation of the carbocation. When a cyclopropyl group is conjugated to a carbocation (see cyclopropylcarbinyl cation 15), it is more stable.13 The cyclopropane bonds have a relatively high degree of s character (Sections 11.3.2 and 11.6.9) that provides stabilization to a cationic center only when conjugated to that center. The molecular model of 15 indicates the relationship of the cyclopropane bonds relative to the positive center.

O

R1 R2

R CH2

C

C

O 14

15

If a π system (not bearing a δ+ center) is conjugated to the cationic center, C+, the positive charge will be dispersed via resonance, diminishing the net charge on the initially formed cationic carbon, and leading to overall stabilization of the carbocation. In other words, a charge is dispersed over several atoms in a resonance-stabilized carbocation, which means that the charge is dispersed over a larger surface area. Therefore, the resonance-stabilized carbocation is less reactive with a nucleophile. In general, a carbocation will be more reactive as the charge is more concentrated, and less reactive as the charge is more dispersed. An allylic carbocation disperses the charge over three atoms (see 16). The molecular model of the allyl carbocation (16, R1]R2]H) shows blue and blue-green areas over all three-carbon atoms, indicative of the dispersed charge (resonance). If the cationic carbon is further conjugated to a diene (as in 17) the charge is dispersed over five atoms, and 17 is expected to be more stable (less reactive) than 16. In benzylic carbocations see (18) the charge is dispersed over seven atoms because the π-bonds of the benzene ring are conjugated to C+. In principle, 18 should be more stable than 17, which is more stable than 16. In the absence of any experimental data, this assumption is reasonable. Inspection of the molecular models for 17 and for 18, respectively, show the blue and blue-green color dispersed over five atoms in 17 and over the seven atoms in 18, which is due to resonance.

R1 R2

C H

R1

C CH 2

R2

C

C

CH2

16

R1

R1 R2

R1 R2

R2 17

R1

R1 R2

R1 R2

R1 R2

18

R2

868

16. CARBON-CARBON BOND-FORMING REACTIONS

The relative order of carbocation stability is related to the ionization potential of that carbocation, which is experimentally determined by electron bombardment in a mass spectrometer. The ionization potential and relative energies (determined by mass spectrometry) can be used as a measure of the relative order of cation stability.14 The lower energy ionization (7.42 vs 8.64 eV as measured for tert-butyl vs n-butyl) indicates that the tertiary carbocation is more stable. Indeed, a 3° carbocation is lower in energy than a 2° carbocation, which is lower in energy than a 1° carbocation. The mass spectral data obviously exclude solvent effects, which can have a profound influence on the relative stability of the cation. The data is consistent, and based on ΔH{calc data for alkyl carbocations 3° is more stable than 2°, which is more stable than 1°, and it is clear that the primary benzyl cation is less stable than the tert-butyl cation, but more stable than the secondary isopropyl cation.15 Electron-releasing groups on the aromatic ring make the benzylic cation more stable, and electron-withdrawing groups make it less stable due to the usual inductive effects (Section 2.2.2). The ΔH{calc data suggest that the primary allyl cation is more stable than a primary alkyl cation, but less stable than methyl. Interestingly, the allylic cation is taken to be only slightly less stable than the benzylic cation, and the propargyl cation is significantly less stable than the allylic cation. The stability of secondary cations derived from cyclic molecules is influenced by the size of the ring, with a stability order C6 > C5 > C4 > C3, and the cyclopentyl cation is about equal in energy to the isopropyl cation.15 The relative stability of carbocations can also be measured by determining the ΔHrxn for hydrolysis of alkyl halides, and the relative stability of various carbocations formed as intermediates from the appropriate chloroalkane.16 An advantage of this measurement is that solvent effects, which are clearly important in most organic transformations, play a significant role. Cyclic 2° and 3° alkyl carbocations are more stable than their acyclic counterparts, and tertiary benzylic are more stable than the tertiary alkyl carbocations. An important conclusion is that the order 3° > 2° > 1° applies to all carbocations, and 3° benzylic > 2° benzylic > 1° benzylic. It is not surprising that a 3° benzylic carbocation is more stable than a 3° alkyl carbocation since the former is resonance stabilized. The tertiary alkyl cation is more stable than the primary benzylic cation, however, which provides an important benchmark for the following order of carbocation stability:15 3° benzylic > 2° benzylic > 3° alkyl > 1° benzylic  1° allylic  2° alkyl > 1° alkyl > methyl

The relative rates of solvolysis for several chloroalkanes provide another practical measure of carbocation stability.17 The rate of ionization to carbocation 19 shows that the more stable carbocations are formed fastest (R1 ¼ R2 ¼ Ph >> R1 ¼ H, R2 ¼ Ph > R1 ¼ R2 ¼ Me), leading to a faster reaction since this is the rate-determining step.16 The presence of a second phenyl group in chlorodiphenylmethane led to greater conjugation, greater stability, and a significantly higher rate of ionization.

R1

Me C Cl

R1

C

Me R2

R2 19

Vinyl carbocations can be generated from a variety of sources, but they are generally less stable than alkyl carbocations.18 Vinyl carbocations were first detected by Grob et al.19 in solvolysis reactions of α aryl vinyl halides, but they have been observed in solvolysis reactions of other alkenyl halides20 and in electrophilic reactions of alkynes.21 14

Deno, N. C. In Progress in Physical Organic Chemistry; Cohen, S. G.; Streitwieser, A., Jr.; Taft, R. W., Eds.; Interscience: NY, 1964; Vol. 2, p 135.

15

Isaacs, N. S. Reactive Intermediates in Organic Chemistry; John Wiley & Sons: London/NY, 1974; p 151.

16

Arnett, E. M.; Hofelich, T. C. J. Am. Chem. Soc. 1983, 105, 2889.

17

(a) Lowry, T. H.; Richardson, K. S. Mechanism and Theory in Organic Chemistry, 3rd ed.; Harper and Row: NY, 1987; p 385; (b) Brown, H. C.; Rei, M. J. Am. Chem. Soc. 1964, 86, 5008. 18

In Dicoordinated Carbocations; Rappoport, Z; Stang, P. J., Eds.; John Wiley & Sons, Inc.: NY, 1997.

19

Grob, C. A.; Csapilla, J.; Csch, G. Helv. Chim. Acta 1964, 47, 1590.

20

Hanack, M. Acc. Chem. Res. 1976, 9, 364.

21

Lucchini, V.; Modena, G.; Pasquato, L. In Dicoordinated Carbocations; Rappoport, Z; Stang, P. J., Eds.; John Wiley & Sons, Inc.: NY, 1997; p 321.

869

16.2 CARBOCATIONS

As with other carbocations, the stability of this intermediate depends on the groups attached to the positive center. The X-ray structure of a bis(silyl)-stabilized vinyl carbocation was reported, for example,22 indicating the special stability of that cation. Br–

H—Br

20

3-Methylbut-1-yne

Br

2-Bromo-3-methylbut-1-ene

Vinyl carbocations can be generated by the reaction of an alkene (e.g., 3-methylbut-1-yne) with a Brønsted-Lowry acid (Section 2.5.1). The reaction of 3-methylbut-1-yne with HBr, for example, generates the secondary vinyl carbocation 20, which is more stable than a possible primary vinyl carbocation. Subsequent reaction of the highly reactive 20 with the nucleophilic bromide ion gives the product, 2-bromo-3-methylbut-1-ene. Note that 20 does not rearrange prior to reaction with the bromide ion, and in general, the rate of nucleophilic reaction is faster than the rate of rearrangement for vinyl carbocations. In other words, vinyl carbocations do not rearrange in this reaction. With the data presented thus far on carbocation stability, the goal is to estimate the relative stability of common carbocations in order to make a prediction concerning product distribution when there are competitive processes. In a simple example, when alkene 21 is treated with HX, several products are possible. Ether 22 can be formed via carbocation 23, or ether 25 via carbocation 24. The major product will logically be determined by formation of the more stable carbocation intermediate. Which is more stable, the primary benzylic cation 24 or the oxygen-stabilized carbocation 23?

H+

X–

X–

H+

X X OMe 22

OMe 23

OMe 21

OMe 24

OMe 25

The relative energies for various carbocations have been compiled,23 and Fig. 16.1 was constructed to show the order of stability based on D(R+-H-), which is equivalent to the ΔHf (RH) for the reaction RdH ! R+ + H. The common carbocations H3C+, MeH2C+, Me2HC+, and Me3C+ are shown to have an energy of 313, 271, 252, and 233 kcal (1308, 1132, 1053, 974 kJ) mol1, respectively.23 Note that the vinyl cation [287 kcal (1199 kJ) mol1] is less stable than the ethyl carbocation, but more stable than the methyl carbocation. Replacing a hydrogen atom in +CH3 with an electronwithdrawing group (e.g., cyano [+CH2CN, 305 kcal (1274 kJ) mol1]) is destabilizing, relative to replacing it with a methyl group (the ethyl cation). Replacing hydrogen with an OH [+CH2OH, 252 kcal (1053 kJ) mol1] leads to an ion that is as stable as the secondary isopropyl cation. The primary allylic carbocation [256 kcal (1070 kJ) mol1] is almost as stable as a secondary alkyl such as isopropyl cation [252 kcal (1053 kJ) mol1]. The propargyl cation [271 kcal (1132 kJ) mol1)] is less stable than the primary allylic. The benzyl carbocation [238 kcal (994 kJ) mol1] is almost as stable as the tertiary tert-butyl cation [233 kcal (974 kJ) mol1], and more stable than the primary allylic and slightly more stable than the secondary allylic carbocation [241 kcal (1007 kJ) mol1]. Placing the positive charge in a five-membered ring [263 kcal (1099 kJ) mol1] is less stable than the cyclohexyl cation [243 kcal (1015 kJ) mol1], with the acyclic isopropyl cation in between. Acylium ions are quite stable. The formyl cation [257 kcal (1074 kJ) mol1] is about as stable as the primary allyl, and the acetyl cation [230 kcal (961 kJ) mol1] is more stable than the tertiary alkyl cation. Note that the amino cation (iminium ion) is very stable, and the tropylium ion shows the greatest stability in this series. Conversely, the alkyne carbocation is extremely unstable [386 kcal (1613 kJ) mol1]. Returning to the question posed for 24 versus 23, a primary benzylic cation [238 kcal (994 kJ) mol1] is expected to be less stable than the oxygen-stabilized cation [220 kcal (919 kJ) mol1], using Fig. 16.1 as a guide. Therefore, 22

22

M€ uller, T.; Juhasz, M.; Reed, C. A. Angew. Chem. Int. Ed. 2004, 43, 1543.

23

Lossing, F. P.; Holmes, J. L. J. Am. Chem. Soc. 1984, 106, 6917.

870

16. CARBON-CARBON BOND-FORMING REACTIONS H

H O

O

C

C

C

Me

203

C

C OH

230

238

220

200

205

218

225

H

H

H

C

C

NHMe

NH2

243

258

252

241

233

256-257

C

C

Me

Me

H

O

C Me

Me C

H

C H

H

Me3C

H C

287

H

280

265

H

N C

H

H

Me 271

263

260

240

Me H

H

C

OH

H

OH 220

212

H H

H

Me

C

NH2

H C

Me

Me Me

H C

Me

H

198

Me

C

H

H

305

386

300

320

294 H

C H

313 H

H C H

Me

Hf (RH) FIG. 16.1 Relative stability of common carbocations based on D(R+-H-), in kcal mol1.17. [The data, but not the figure, is used and reprinted with permission from Lossing, F.P.; Holmes, J.L. J. Am. Chem. Soc. 1984, 106, 6917. Copyright © 1984, American Chemical Society.]

is predicted to be the major product. It is likely that a mixture of products will be observed, but the data in Fig. 16.1 allows a practical comparison of carbocations that allows one to make reasonable predictions of major and minor products in many cases.

16.2.2 Configurational Instability As briefly discussed in Sections 2.5, 3.2.2, and 16.2.1, carbocations are planar about C+. Therefore, a nucleophile can approach C+ from either face in an SN1 type reaction, which is expected to proceed with racemization. In other words, there is little or no facial selectivity in the simple cases. Retention or partial retention and even inversion can occur “if the structure contains a configuration-holding group, a nucleophilic group that can interact with the carbonium center”.24 If a tight ion pair is generated, the SN1 reaction will show some memory of the departing group since it is in close proximity and partial inversion can occur, which will clearly be solvent dependent. In water or aqueous alcohol, a series of secondary halides undergo solvolysis reactions with varying digress of inversion.25 Heating in 60% aq EtOH, for example, gave the alcohol with 83% inversion and (R)-(1-bromoethyl)benzene reacted with 80% aq EtOH to yield the alcohol with 60% inversion.25 These results contrast with the solvolysis reaction of the carboxylate anion of (R)-2-bromopropanoic acid in water, which proceeded with racemization (0% inversion).25 In this work, the addition of silver ion increased the percentage inversion, presumably via coordination with the halide such that a free cation will not be generated, and the departing bromide is close to the cationic center at the time of the substitution. As a practical matter, reactions of carbocations do not necessarily proceed with complete inversion or retention, but many proceed with partial inversion. It is reasonable, however, to assume that racemization or partial racemization will be the stereochemical outcome of cationic reactions unless there is experimental evidence or literature precedent to support inversion or retention.

16.2.3 Carbocation Rearrangements A second property of carbocations that influences their synthetic utility is their propensity to rearrange to a more stable cation. When 1-phenylpropan-2-ol was treated with HCl, two products are possible. The first is the direct SN1 24

Ref. 15, pp 118–119.

25

Ref. 15, p 119.

871

16.2 CARBOCATIONS

product (2-chloropropyl)benzene, see Section 3.2.2) that results from trapping the initially formed cation (26) with chloride ion (rate constant ¼ k4). The second product is the rearranged benzylic chloride, (1-chloropropyl)benzene. Rearrangement via a 1,2-hydride shift can occur (rate constant ¼ k2) because the rearrangement product, benzylic cation 27, is more stable than the initially formed 26 (see Fig. 16.1), and subsequent trapping with chloride ion (rate constant ¼ k3) yields (1-chloropropyl)benzene. In the limiting cases, if k2 ≫ k4 only (1-chloropropyl)benzene will be observed, whereas only (2-chloropropyl)benzene will be observed if k4 ≫ k2. In both cases, the initial ionization step (rate constant ¼ k1) is the slowest (rate-determining) step, and k3  k4. If the two cation intermediates are close in energy, then k2  k4, and a mixture of products will result or only a modest amount of rearrangement will occur. In many cases, the rearrangement (k2) proceeds faster than the direct substitution (k4) and the rearranged product predominates. If the cation and nucleophile are particularly reactive, however, k4 may be much faster than k2. A simplifying assumption is that k2 > k4 in most cases, which is not always true, but provides a good working model for relatively simple substrates. Ph Ph

+ HCl – H 2O

Me H

Ph

k1

H 27

H

OH 1-Phenylpropan-2-ol

+ Cl –

26

Me Cl

k3

k2

Me

Ph

+ Cl –

Me

(1-Chloropropyl)benzene

Ph

Me

k4

H Cl (2-Chloropropyl)benzene

A carbocation intermediate is best viewed as an electron-deficient p-orbital (the positive center), and rotation about the adjacent single σ-bond allows the σ-hybrid orbital and the empty p-orbital to become parallel. When the adjacent σ-bond (CdH) and the p-orbital are parallel (as in 28) the σ- and p-orbitals can overlap, which allows electron density (i.e., the bond) to migrate toward the positive center. The transition state for this process is usually represented as a bridged electrophilic species, 29. The transformation of 28 to 30 via 29 requires transport of the hydrogen atom as the electrons in the bond move, and this migration is described as a 1,2-hydride shift of the hydrogen atom to the adjacent carbon to yield 30. This rearrangement occurs when the reaction (28 ! 30) is exothermic, and if the bond bearing the migrating group can become parallel with the p-orbital of the cationic center. In this case, the more stable 3° carbocation is favored over the 2° cation, and the 1,2-hydrogen shift is an exothermic reaction. In general, secondary ! tertiary, primary ! secondary, and primary ! tertiary rearrangements are exothermic reactions, and the rearrangement is facile. H

H

H

Me

Me Me

H

Me Me 28

Me Me

H 29

Me

Me H 30

Why did the hydrogen atom in 28 migrate and not the methyl group? First, migration of the methyl would lead to a significantly less stable 1° cation as the product, so the rearrangement reaction would be endothermic. Second, migration of the larger methyl group may require more energy than the smaller hydrogen atom. For the most part, a smaller atom migrates in preference to a group, and a smaller group in preference to a larger group. There are some exceptions to this statement, and a methyl group can certainly migrate as part of a carbocation rearrangement. Other groups can migrate, including phenyl. Indeed, a relatively low-energy phenonium ion (e.g., 31) has been invoked in many aryl migrations.26 In a transition state (e.g., 29), considerable positive charge can be present and charge dispersal (as in 31) can facilitate the rearrangement. A lesson from this observation is that careful analysis of a system is required to determine a priori which atom will preferentially migrate.

26

(a) Lancelot, C. J.; Cram, D. J.; Schleyer, P. v. R. In Carbonium Ions; Olah, G.A.; Schleyer, P. v. R., Eds.; John Wiley: NY, 1972; Vol. III; (b) Ref. 17a, p 367.

872

16. CARBON-CARBON BOND-FORMING REACTIONS

H R

H R 31

As mentioned above, and in the absence of special electronic effects, alkyl groups show a clear dependence on the size of the migrating group. In general, smaller groups migrate before larger ones: H > Me > CHMe2 > CMe3 Note that it is difficult to give an absolute scale for migratory aptitude since migratory aptitude is inevitably linked to the stability of the cation being formed and the relative energy of the transition state. An example is cation 32, where the larger ethyl group can migrate preferentially (to yield 34) rather than the smaller methyl group to yield 33. The p-methoxyphenyl group will stabilize cation 34 to a greater extent than the p-nitrophenyl group stabilizes 33, however. The electron-releasing methoxy diminishes the net charge and stabilizes 34, but the electron-withdrawing nitro group destabilizes the cationic center in 33. Therefore, the larger ethyl group will migrate rather than the smaller methyl group due to formation of a more stable carbocation. Any analysis for determining which group migrates must take into account both the stability of the carbocation being formed (relative to that initially available), and the size of the migrating group. If the rearrangement of either hydrogen or methyl will generate a cation of similar stability, the relative size of the migrating group is usually the determining factor and the hydrogen will migrate preferentially. NO2

NO2

NO2

1,2-Ethyl shift

1,2-Methyl shift

Me Me MeO

Et

Me Me

Me

Me

Et

MeO

33

Et MeO

32

34

In some cases, atoms or groups attached to or close to C+ can assist migration, which increases the rate of rearrangement. Such participation is another example of a neighboring-group effect.27 A through-space interaction (Section 2.2.1) of an empty p-orbital with the electrons in a neighboring bond is believed to be important for 1,2-shits. The term σ-participation has been used to describe this interaction. Transition state 35 is shown in Fig. 16.2 for a 1,2-shift that assumes a concerted migration, with ionization of the leaving group (this is effectively a symmetrical bridging transition state).28 Structures 36 represents σ-conjugation, which is usually the same as hyperconjugation, and “implies sideways interaction of orbitals. In Fig. 16.2, 36a represents a hyperconjugative interaction of Z with Ca, whereas 36b represents

C C

C C

C C

X

X

X

35

FIG. 16.2 The σ conjugation in 1,2-bond migrations.

27

Ref. 4, pp 429–439.

28

Ref. 4, p 431.

Z

Z

Z

36a

36b

873

16.2 CARBOCATIONS

the case where there is a small amount of σ participation.28 Such neighboring-group effects can both accelerate the rate of rearrangement and stabilize the carbocation. Carbocation formation and rearrangements are found rather often in organic syntheses. In the synthesis of (+)-strongylin A by Katoh and coworkers,29 alcohol 37 was treated with boron trifluoride etherate to give (+)strongylin A in 84% yield. This conversion constitutes a cascade reaction in which initial conversion of 38 to tertiary carbocation 38 was followed by rearrangement to 39 by a 1,2-methyl shift. A subsequent 1,2-hydride shift generated 40, which reacted with a proximal phenolic hydroxyl group to yield oxonium ion 41. Loss of a proton in an acid-base reaction yields the final product. HO

HO

HO

HO

HO

OMe

OMe

HO

OMe

BF3•OEt2

H

H

H

CH2Cl2 –78–0°C

HO 37

38

39

HO

HO

HO

H

OMe

H

HO

O

OMe

O

H 40

OMe

H (+)-Strongylin A (84%)

41

Cyclopropylcarbinyl systems, where the cationic carbon is connected directly to a cyclopropyl ring, undergo 1,2-alkyl shifts that lead to ring expansion. In Section 16.2.1, it was noted that the cyclopropyl group imparts special stability to an adjacent carbocation, and migration of the strained cyclopropyl bonds in cyclopropylcarbinyl cations (e.g., 15) leads to the less strained homoallylic cation 42.30 Participation of the cyclopropane ring is apparent in the solvolysis of cyclopropylcarbinyl tosylate, which occurs 106 times faster than solvolysis of the tosylate of 2-methyl-1-propanol (isobutyl 4-methylbenzenesulfonate).31 A cyclobutyl cation (43) can also form under these conditions via a ring-expansion reaction. The cyclobutylcarbinyl system (cyclobutylmethyl 4-methylbenzenesulfonate) is less strained than cyclopropylmethyl 4-methylbenzenesulfonate, but also undergoes ring expansion via cyclobutylcarbinyl cation, 44. Solvolysis of cyclobutylmethyl 4-methylbenzenesulfonate leads to the cyclopentyl cation (45), via ring expansion of 44. OTs

+

CH2

Cyclopropylmethyl 4-methylbenzenesulfonate

15

42

43

OTs

Cyclobutylmethyl 4-methylbenzenesulfonate

29

44

45

Kamishima, T.; Kikuchi, T.; Katoh, T. Eur. J. Org. Chem. 2013, 4558.

(a) Roberts, J. D.; Mazur, R. H. J. Am. Chem. Soc. 1951, 73, 2509; (b) Mazur, R. H.; White, W. N.; Semenow, D. A.; Lee, C. C.; Silver, M. S.; Roberts, J. D. Ibid. 1959, 81, 4390; (c) Brady, S. F.; Ilton, M. A.; Johnson, W. S. Ibid. 1968, 90, 2882; (d) Julia, M.; Mouzin, G.; Descoins, C. C.R. Hebd. Seances Acad. Sci. 1967, 264, 330. 30

31

Roberts, D. D. J. Org. Chem. 1964, 29, 294; (b) Idem Ibid. 1965, 30, 23.

874

16. CARBON-CARBON BOND-FORMING REACTIONS

Cationic rearrangement can be used in other ring-expansion reactions, such as the transformation of cyclooctylmethyl 4-methylbenzenesulfonate to (Z)-cyclononene upon treatment with acid, albeit in only 9% yield.32 Solvolysis of the tosyl group in cyclooctylmethyl 4-methylbenzenesulfonate produced the unstable primary carbocation 46. Rearrangement led to ring expansion to the more stable secondary carbocation 47, which lost a proton to give the alkene product, (Z)-cyclononene. The overall result was expansion of the eight-membered to a nine-membered ring. If bond migration leads to a medium-size ring (8–13), the high energy inherent in these rings may inhibit the bond migration (Section 4.5.1) and give poor yields. Indeed, despite the fact that a secondary carbocation is much more stable than a primary carbocation, the yield of (Z)-cyclononene was rather poor in large part because the nine-membered ring is higher in energy than the eight-membered ring precursor. OTs

–H +

H+

Cyclooctylmethyl 4-methylbenzenesulfonate

46

(Z)-Cyclononene

47

Cationic rearrangements can be highly selective, and several named reactions involve this reaction. The Mundy et al.33 monograph is an excellent source that presents many examples of these named reactions. Carbocations are available from precursors other than halides or sulfonate esters. Treatment of an amine with nitrous acid (HONO) generates an unstable diazonium salt, which decomposes to a transient primary carbocation. The initially formed carbocation rearranged to the larger ring carbocation in what is now known as the Demjanov rearrangement.34 In the initially reported reaction, water was present, which trapped the ring-expanded cation as the alcohol.35,11a,c Synthetic applications of this ring expansion generally use the milder reaction conditions associated with alcohol or sulfonate ester precursors. In Fitjer and Mandelt’s36 synthesis of laurene, alcohol 48 was heated with p-toluenesulfonic acid in benzene to generate carbocation 49, with the positive charge adjacent to a strained four-membered ring. Opening the strained four-membered ring to the lower energy five-membered ring product occurred by a 1,2-alkyl shift to yield 50. Under the reaction conditions, elimination occurred to give a 90% yield of 51 along with 10% of 52. Note that the stereochemistry of the groups adjacent to the cation can have a significant effect on the migratory aptitude of the groups adjacent to the cationic center.37

p-TsOH

HO

+

+H+ ; – H 2O

48

49

50

51 (90%)

OH 52 (10%)

A relatively simple variation of this reaction is the Tiffeneau-Demjanov ring expansion,38,34c which involves treatment of amino alcohols with nitrous acid to yield ring-expanded ketones. The conversion of 53 to a mixture of isomeric ketones in 81% yield is an example.39 This example illustrates a common problem in ring-expansion reactions in particular and in rearrangements in general, that of regioselectivity. Treatment of 53 with nitrous acid produced cation 54.

32

Huisgen, R.; Seidl, G. Tetrahedron 1964, 20, 231.

33

Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; John Wiley & Sons, Inc.: NY, 2005.

(a) Demjanov, N. Y.; Lushnikov, M. J. Russ. Phys. Chem. 1903, 35, 26; (b) Idem Chem. Zentr. 1903, I, 1468; (c) Smith, P. A. S.; Baer, D. R. Org. React. 1960, 11, 157; (d) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-23. 34

35

Friedman, L. In Carbonium Ions; Olah, G. A.; Schleyer, P. v. R., Eds.; John Wiley: NY, 1970; Vol. II, p 655.

36

Mandelt, K.; Fitjer, L. Synthesis 1998, 1523.

37

Nikishova, N. G.; Bundel, Yu. G.; Vestn. Mosk. Univ., Ser. 2: Khim 1985, 26, 486 (Chem. Abstr. 104: 167800 p, 1986).

(a) Tiffeneau, M.; Weill, P.; Tchoubar, B. Compt. Rend. 1937, 205, 54; (b) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-93; (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 648–649. 38

39

Parham, W. E.; Roosevelt, C. S. J. Org. Chem. 1972, 37, 1975.

875

16.2 CARBOCATIONS

Two ring expansion pathways are possible, paths a and b. The major pathway is path a, first giving ring expansion to oxocarbenium ion 55, which lost a proton to give a 55% yield of 56. This pathway also produced 0.7% of the ketone with a trans ring juncture. Path b led to oxocarbenium ion 57, which produced 25.6% of ketone 58. Incorporation of silicon and tin substituents into the molecule facilitated ring expansion.40

Path b OH 55

b HO

a Path a

54

53

56 (55%)

CH2

HO

NH3 Cl

O

OH

O

57

58 (25.6%)

This ring-expansion process is quite useful when applied to polycyclic systems. Corey and Nozie41 treated 59 with 40% sulfuric acid, and isolated α-caryophyllene alcohol. The initially formed cyclobutylcarbinyl cation (60) rearranged to the cyclopentyl cation (61), which was trapped by water to yield α-caryophyllene alcohol. Trost et al.42,43 used ring expansion to generate oxaspiropentane derivatives that were converted to cyclobutane derivatives. In Section 12.2.2.2, oxaspiropentanes (e.g., 62, 9-oxadispiro[2.0.44.13]nonane) were prepared by the reaction of ketones and sulfur-stabilized cyclopropyl ylids. Treatment of 62 with aqueous tetrafluoroboric acid (HBF4) gave carbocation 63, and ring-expanding rearrangement yielded 64. Subsequent reaction with water and loss of a proton gave the final product, spiro[3.4]octan-1-one.35,34 Another variation of this ring expansion used sulfur-stabilized cations).44,45 Me

Me

Me

H

H

Me

H

40% H2SO 4 THF –5

25°C

Me

Me

H

H

Me

H

OH Me

Me

Me

H 2O

H

H

Me

Me

H

H

Me

OH

Me

Me

59

60

61

OH

OH

O

62

Me

63

64

Caryophyllene alcohol

O

Spiro[3.4]octan-1-one

40

Chow, L.; McClure, M.; White, J. Org. Biomol. Chem. 2004, 2, 648.

41

(a) Corey, E. J.; Nozoe, S. J. Am. Chem. Soc. 1964, 86, 1652; (b) Idem Ibid. 1965, 87, 5733.

42

(a) Trost, B. M. Acc. Chem. Res. 1974, 7, 85; (b) Trost, B. M.; LaRochelle, R.; Bogdanowicz, M. J. Tetrahedron Lett. 1970, 3449.

43

Trost, B. M.; Bogdanowicz, M. J. J. Am. Chem. Soc. 1973, 95, 5311.

44

Tanino, K.; Sato, K.; Kumajima, I. Tetrahedron Lett. 1989, 30, 6551.

45

Ranu, B. C.; Jana, U. J. Org. Chem. 1999, 64, 6380.

876

16. CARBON-CARBON BOND-FORMING REACTIONS

Epoxides other than oxaspiropentanes (e.g., 62) will react with Brønsted-Lowry or Lewis acids to generate carbocations that are subject to rearrangement. In Tanino et al.46 synthesis of ingenol, treatment of 65 with trimethylaluminum gave 68 in >80% yield. The epoxide unit reacted with a Lewis acid to generate oxonium ion 66, which rearranged by a 1,2-alkyl shift to yield oxocarbenium ion 67. Loss of a proton and protonation of the OdAl unit gave ketone 68. HO O

Me3Al

OH

OH

O

Me3Al

Me3Al

CH2Cl2

O

1,2-Alkyl shift

OTIPS

OMe

O

OH

OTIPS

OMe

65

66

MeO

OTIPS

OMe

OTIPS

67

68 (>80%)

A specialized carbocation rearrangement first observed with bicyclo[2.2.1] systems was the acid catalyzed reaction of camphene hydrochloride to isobornyl chloride.47 The reaction, which proceeded by a 1,2-alkyl shift called the Wagner-Meerwein rearrangement,48 initially generated bornyl cation 69, but rearrangement to 70 occurred prior to trapping of chloride ion. The term Wagner-Meerwein is often applied to the acid-catalyzed rearrangement of [m.n.o]-bicyclo and [m.n.o]-tricyclo species, although it applies formally to bicyclo[2.2.1]heptane derivatives. 7

Me Me Cl

4

Aqueous solvent H

Me

5 3

6

+

1

Me

2

Me

Me 69

Camphene hydrochloride 7 5

4 1

3 2

6

Me

Me

3

2

6

Me

Me

Me Me

+ Cl

–

Me Me Cl

1

Me

5

70

4

Isobornyl chloride

This reaction has been controversial concerning the structure of norbornyl carbocations (e.g., 69 and 70).49,50 Norbornyl carbocations have been described in terms of the resonance structures shown in 71, or as the nonclassical ion 72.51 This view was questioned, primarily by Brown,49 because he did not believe that the evidence proved the existence of a bridged intermediate.52 Brown suggested that the facile rearrangement associated with 71 is explained by simple steric effects that lead to a series of rapidly equilibrating ions.53 However, there is good evidence that the norbornyl cation rearranges with considerable σ participation.54 Evidence provided by NMR for the nonclassical

46

Tanino, K.; Onuki, K.; Asano, K.; Miyashita, M.; Nakamura, T.; Takahashi, Y.; Kuwajima, I. J. Am. Chem. Soc. 2003, 125, 1498.

47

(a) Wagner, G. J. Russ. Phys. Chem. Soc. 1899, 31, 690; (b) Meerwein, H. Annalen 1914, 405, 129; (c) Birladeanu, L. J. Chem. Educ. 2000, 77, 858.

48

(a) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-98; (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 678–679. 49

Brown, H. C. The Non-Classical Ion Problem, Plenum: NY, 1977.

50

Story, R. R.; Clark, B. C. In Carbonium Ions; Olah, G. A.; Schleyer, P. v. R., Eds.; John Wiley: NY, 1972; Vol. III, p 1007.

(a) Berson, J. A. In Molecular Rearrangements; Mayo, P., Ed., NY; Vol. I, p 111; (b) Sargent, G. D. Q. Rev. Chem. Soc. 1966, 20, 301; (c) Olah, G. A. Acc. Chem. Res. 1976, 9, 41; (d) Scheppelle, S. E. Chem. Rev. 1972, 72, 511. 51

52

(a) Brown, H. C. Tetrahedron 1976, 32, 179; (b) Brown, H. C.; Kawakami, J. H. J. Am. Chem. Soc. 1970, 92, 1990.

53

Brown, H. C.; Ravindranathan, M. J. Am. Chem. Soc. 1978, 100, 1865.

54

(a) Coates, R. M.; Fretz, E. R. J. Am. Chem. Soc. 1977, 99, 297; (b) Brown, H. C.; Ravindranathan, M. Ibid. 1977, 99, 299.

877

16.2 CARBOCATIONS

ion, in super acids at low temperatures, is strong.55 Olah55 was more forceful, stating “it can be concluded that the original views of Winstein56 on the nonclassical nature of the norbornyl cation, based on kinetic and stereochemical reactivity, have been substantiated by direct spectroscopic studies of the long-lived ion”.57 7

7 4

5 6

1

5 3

2

7 4

1

3

1

2

6

4

5

3

2

6 71

72

The mechanistic controversies have not diminished the synthetically interesting transformations possible with controlled cation rearrangements.58 An example is taken from a synthesis of the longipinane skeleton by Joseph-Nathan and coworkers,59 in which 73 was treated with sulfuric acid to give a 64% yield of the rearranged product 75. The reaction involved rearrangement of the secondary carbocation 74, with migration of bond C10–C11.

HO

11

10

11

11

OAc 4

10

1

1

1

OAc

OAc

OAc 10

OAc

OAc 4

73

4

75 (64%)

74

Although they are difficult to generalize, several interesting disconnections are possible utilizing cationic rearrangements. Some general examples follow: OH (CH 2)n

R1

NH2 (CH 2)n

R2

OH

R1 R X R

X

O

R2

O

A highly specialized rearrangement is the shift observed during liver microsome-mediated oxidations of aromatic substrates, and during solvolysis of labeled arene oxides,60 now called the NIH shift.61 “In most enzymatic hydroxylations of aromatic substrates the substituent (2H, 3H, or halogen) present at the position of the entering oxygen migrates to either one of the adjacent ring positions.”62 As observed with an ipso nitration reaction, this shift probably involves an intramolecular hydrogen shift.63 A specific example involves enzymatic hydroxylation of chlorobenzene to yield 76. The NIH shift produces 77 and deprotonation, with aromatization, gives the substitution product, 4-chlorophenol.61

55

Olah, G. A. Carbocations and Electrophilic Reactions; Verlag Chemie/Wiley: NY, 1974; pp 80–89.

56

(a) Winstein, S. Q. Rev. Chem. Soc. 1969, 23, 141; (b) Winstein, S.; Trifan, D. S. J. Am. Chem. Soc. 1949, 71, 2953; (c) Idem Ibid. 1952, 74, 1154.

57

(a) Ref. 55, p 89; (b) Olah, G. A.; White, A. M.; DeMember, J. R.; Commeyras, A.; Lui, C. Y. J. Am. Chem. Soc. 1970, 92, 4627.

58

Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd Ed.; John Wiley & Sons, Inc.: NY, 2005; pp 678–679. 59

Cerda-García-Rojas, C. M.; Flores-Sandoval, C. A.; Román, L. U.; Hernández, J. D.; Joseph-Nathan, P. Tetrahedron 2002, 58, 1061.

60

(a) Daley, J. W.; Hernia, D. M.; Witkop, B. Experentia 1972, 28, 1129; (b) Bruice, T. C.; Bruice, P. Y. Acc. Chem. Res. 1976, 9, 378.

61

Guroff, G.; Daly, J.; Jerina, D.; Renson, J.; Witkop, B.; Udenfriend, S. Science 1967, 157, 1524.

62

Jerina, D. M.; Daly, J. W.; Witkop, B. J. Am. Chem. Soc. 1967, 89, 5488.

63

Feldman, K. S.; McDermott, A.; Myhre, P. C. J. Am. Chem. Soc. 1979, 101, 505.

878

16. CARBON-CARBON BOND-FORMING REACTIONS

H

Enzymatic hydroxylation

Cl

H

H NIH shift

H

H

– H+

H

Cl

Cl

OH

Cl

OH

OH 77

76

16.2.4 The Pinacol Rearrangement The Pinacol rearrangement is an important cationic rearrangement reaction widely used for synthesis.64 Fittig65,64a was the first to discover that treatment of 2,3-dimethylbutane-2,3-diol with sulfuric acid generated 3,3-dimethylbutan-2one (pinacolone). This transformation gave rise to the name of the reaction, and was shown to be applicable to the acid-catalyzed rearrangement of most 1,2-diols. Rearrangement of pinacol (2,3-dimethylbutane-2,3-diol, for which the reaction is named) to pinacolone (3,3-dimethylbutan-2-one) is the classical example, but the reaction can be done with a variety of 1,2-diols (cyclic and acyclic), utilizing either a Brønsted-Lowry and a Lewis acid HO

O

OH Me

Me

Me

H2SO 4

Me

Me Me 2,3-Dimethylbutane-2,3-diol

HO

OH R1

R R

R1 78

H+

3,3-Dimethylbutan-2-one

OH R

R1 OH R1

R1

R

Me Me

R1 O—H R1

R

79

R1

R

R

O

R1

R 80

R

R1 R 81

The reaction can be generalized to show that 1,2-diols (78) are converted first to a β-hydroxy carbocation (79). Rearrangement to the more stable 80 (an oxocarbenium ion rather than a tertiary alkyl cation, Section 16.2.1) is followed by loss of a proton to yield 81.66 Two important factors are the acid used to initiate the reaction, and the concentration at which the reaction is done.67 With increasing dilution of the acid, as well as the conjugate base, a higher percentage of elimination products are formed rather than the rearrangement product.67 When Lewis acids are used rather than Brønsted-Lowry acids, differences in the reaction pathway are often observed. The reaction of 82 with aqueous perchloric acid gave an 81% yield of 2,2-dimethylcycloheptan-1-one, but reaction with boron trifluoride in dichloromethane gave an 88% yield of 1-(1-methylcyclohexyl)ethan-1-one.68 In aqueous media with the Brønsted-Lowry acid, tertiary carbocation 83 was formed from 82. Ring expansion to cycloheptyl oxocarbenium ion 84 followed by loss of a proton gave 2,2-dimethylcycloheptan-1-one. Reaction with boron trifluoride in aprotic media, however, proceeded by coordination with the cyclohexanol moiety to form cation 85. Subsequent 1,2-methyl shift to oxocarbenium ion 86 occurred rather than ring expansion, and loss of a proton gave 1-(1-methylcyclohexyl)ethan-1-one. Cation formation can be influenced by rather subtle factors. When the reaction was examined with an analogue that possessed a double bond in the ring of 82,69 ring expansion occurred, but no products [e.g., 1-(1-methylcyclohexyl)ethan-1-one] were observed.

(a) Fittig, R. Annalen 1860, 114, 54; (b) Collins, C. J. Q. Rev. Chem. Soc. 1960, 14, 357; (c) Mundy, B. P.; Otzenberger, R. D. J. Chem. Educ. 1971, 48, 431; (d) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-74; (e) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 514–516.

64

65

Fittig, R. Annalen 1859, 110, 17.

66

For reactions of this type, see Larock, R. C. Comprehensive Organic Transformations, 2nd ed.; Wiley-VCH: NY, 1999; pp 1281–1282.

67

De Lezaeta, M.; Sattar, W.; Svoronos, P.; Karimi, S.; Subramaniam, G. Tetrahedron Lett. 2002, 43, 9307.

68

Corey, E. J.; Danheiser, R. L.; Chandrasekaran, S. J. Org. Chem. 1976, 41, 260.

69

Bhushan, V.; Chandrasekaran, S. Chem. Lett. 1982, 1537.

879

16.2 CARBOCATIONS

Me –20°C

OH Me OH

82

Me

Me OH

OH

O

70% aq HClO4

83 OH Me

0.5 equiv BF3•OEt2

2,2-Dimethylcycloheptan-1-one (81%)

84

O

OH

Me

Me

CH2Cl2 0°C

85

1-(1-Methylcyclohexyl)ethan-1-one (88%)

86

This reaction is quite useful. An example is the conversion of 87 to 88 in 97% yield upon treatment with the acidic resin Amberlyst A-15,70 taken from Alvarez-Manzaneda et al. synthesis of ligphagal. Pinacol rearrangements are useful for the conversion of cyclic diols to spirocyclic ketones. Many spirocyclic compounds,71,72 are formed by this type of rearrangement. The pinacol rearrangement has been used to construct stereogenic centers adjacent to heterocycles.73 O

O

O O

HO

OBn

Amberlyst A-15, rt

OBn

Me

CH2Cl2 , 50 min

O

Me

Me

H

Me

OH H

H 87

88 (97%)

The disconnection for these reactions follow: O

R

R

R

R HO

R

R

R OH R

O R

R

Pinacol-like rearrangements are possible from precursors other than diols, especially oxiranes,74,75 and allylic alcohols.76 The term semipinacol rearrangement is applied to the reaction of many substrates, including α-hydroxy

70

Alvarez-Manzaneda, E.; Chahboun, R.; Alvarez, E.; Cano, M. J.; Ali Haidour, A.; Alvarez-Manzaneda, R. Org. Lett. 2010, 12, 4450.

71

Mundy, B. P.; Kim, Y.; Warnet, R. J. Heterocycles 1983, 20, 1727.

72

Krapcho, A. P. Synthesis 1976, 425.

73

Shinohara, T.; Suzuki, K. Tetrahedron Lett. 2002, 43, 6937.

74

Alarcón, P.; Pardo, M.; Sato, J. L. J. Heterocyclic Chem. 1985, 22, 273.

(a) Yandovskii, V. N.; Ershov, B. A. Russ. Chem. Rev. 1972, 41, 403, 410; (b) Also see Sudha, R.; Narashimhan, K. M.; Saraswathy, V. G.; Sankararaman, S. J. Org. Chem. 1996, 61, 1877; (c) Ranu, B. C.; Jana, U. J. Org. Chem. 1998, 63, 8212. 75

76

(a) Suzuki, K.; Katayama, E.; Tsuchihashi, G. Tetrahedron Lett. 1984, 25, 1817; (b) Idem Ibid. 1983, 24, 4997.

880

16. CARBON-CARBON BOND-FORMING REACTIONS

aldehydes,77 epoxy alcohols,78 and epoxy ethers,79 although the term pinacol-like rearrangement is also applied to rearrangement of these substrates. A highly stereoselective rearrangement of 2,3-epoxy alcohols was reported, using the initiator bis(iodozincio)methane.80 The Meinwald rearrangement81 is a related rearrangement induced by the enzyme pig liver esterase.82 In principle, any molecule that generates an intermediate with a positive charge on a carbon α to one bearing an OH group can undergo rearrangement. β-Amino alcohols, for example, rearrange after treatment with nitrous acid via the so-called the semipinacol rearrangement.83 Similarly, iodohydrins84 and allylic alcohols rearrange upon treatment with a strong acid that protonates the double bond. An example of the semipinacol rearrangement is taken from Tu and coworker’s85 synthesis of ()-FR901483 in which cyclopropanol-selenide (89) was treated with mcpba in the presence of Py to give >59% yield of cyclobutanone (90). OBn

OBn

OBn

OBn

mcpba , Py

N3 PhSe

OMe

OH

Hexane-CH2Cl2 –30°C to rt

N3

OMe

O 89

90 (>59%)

Another example is a synthesis of (+)-asteltoxin by Cha and coworker’s86 in which epoxide 91 was treated with titanium tetrachloride and gave a 96% yield of aldehyde 92. In another variation, taken from a synthesis of 2-thiocyanaotoneopupukeanane by Uyehara et al.,87 hydroxy ether 93 was treated with tosic acid to give a 96% yield of ketone 94. TMSO

OH

O

TiCl4

OHC

OTIPS 91

Me OTIPS 92 (96%)

HO TsOH, PhH Reflux

MeO 93

O 94 (96%)

For example, see (a) Miller, T. C. J. Org. Chem. 1969, 34, 3829; (b) Schor, L.; Gros, E. G.; Seldes, A. M. J. Chem. Soc. Perkin Trans. 1992, 1, 453; (c) Joshi, A. P.; Nayak, U. R.; Dev, S. Tetrahedron 1976, 32, 1423; (d) Benjamin, L. J.; Adamson, G.; Mander, L. N. Heterocycles 1999, 50, 365; (e) Marson, C. M.; Oare, C. A.; McGregor, J.; Walsgrove, T.; Grinter, T. J.; Adams, H. Tetrahedron Lett. 2003, 44, 141.

77

78

Marson, C. M.; Khan, A.; Porter, R. A.; Cobb, A. J. A., Tetrahedron Lett. 2002, 43, 6637.

For a synthesis of (+)-grindelic acid using a pinacol ring expansion involving hydroxy ethers, see Paquette, L. A.; Wang, H.-L. Tetrahedron Lett. 1995, 36, 6005.

79

80

Matsubara, S.; Yamamoto, H.; Oshima, K. Angew. Chem. Int. Ed. 2002, 41, 2837.

(a) Meinwald, J.; Labana, S. S.; Chadha, M. S. J. Am. Chem. Soc. 1963, 85, 582; (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 420–421.

81

82

Niwayama, S.; Noguchi, H.; Ohno, M.; Kobayashi, S. Tetrahedron Lett. 1993, 34, 665.

(a) Pollack, P. I.; Curtin, D. Y. J. Amer. Chem. Soc. 1950, 72, 961; (b) Curtin, D. J.; Pollack, P. I. J. Am. Chem. Soc. 1951, 73, 992; (c) Curtin, D. Y.; Harris, E. E.; Pollack, P. I. J. Am. Chem. Soc. 1951, 73, 3453; (d) Curtin, D. Y.; Meislich, E. K. J. Am. Chem. Soc. 1952, 74, 5905; (e) Curtin, D. Y.; Crew, M. C. J. Am. Chem. Soc. 1954, 76, 3719. Also see (f) Spivak, C. E.; Harris, F. L. J. Org. Chem. 1972, 37, 2494. 83

84

Krief, A.; Laboureur, J. L.; Dumont, W.; Labar, D. Bull. Soc. Chim. Fr. 1990, 681.

85

Ma, A.-J.; Tu, Y.-Q.; Peng, J.-B.; Dou, Q.-Y.; Hou, S.-H.; Zhang, F.-M.; Wang, S.-H. Org. Lett. 2012, 14, 3604.

86

Eom, K. D.; Raman, J. V.; Kim, H.; Cha, J. K. J. Am. Chem. Soc. 2003, 125, 5415.

87

Uyehara, T.; Onda, K.; Nozaki, N.; Karikomi, M.; Ueno, M.; Sato, T. Tetrahedron Lett. 2001, 42, 699.

881

16.3 CARBOCATIONS AND CARBON-CARBON BOND-FORMING REACTIONS

16.3 CARBOCATIONS AND CARBON-CARBON BOND-FORMING REACTIONS There are several synthetically useful and selective carbocation reactions and some classical named reactions that involve cationic intermediates. Variations in functional groups, and the reaction conditions used to generate the carbocation, lead to chemoselective transformations that are often highly stereoselective. The intent of this section is to focus attention on the variety of reaction types, and on synthetic applications.

16.3.1 Carbocation Reactions With Alkene Nucleophiles Alkenes react as two-electron donors to a carbon atom (nucleophiles) in reactions with carbocations. Two electrons from the π-bond are used to form a new CdC bond to one carbon of the C]C, generating a new carbocation (see 96). Both intermolecular (as in the formation of 96) and intramolecular (as in the formation of 99) versions are known. The intermolecular process is often plagued by cationic polymerization and/or competitive elimination reactions (an E1 reaction, Section 3.5.2), depending of the solvent and relative basicity of the counterion. The intramolecular reaction is subject to competing inter- and intramolecular reactions (Section 4.5), but is generally less prone to polymerization, although elimination remains a problem. In an intermolecular example, 2-methylbutan-2-ol was treated with acid, leading to protonation of the oxygen atom. Loss of water from this oxonium salt led to tertiary carbocation 95. R H+

OH

OH2

– H 2O

R

R

R 95

R

OH

96

R

H+ – H 2O

R

97

98

99

In the presence of a second molecule (the alkene), the π-bond attacked the cationic center in an intermolecular reaction to form a new carbocation (96). In a similar manner, alcohol 97 was the precursor to carbocation 98 when treated with acid. Subsequent intramolecular attack by the alkene moiety on the other side of the same molecule led to carbocation 99. Treatment of 100 with 90% sulfuric acid gave 101 as a reactive intermediate that was trapped intramolecularly by the alkene to yield the chlorine-stabilized carbocation 102. Subsequent reaction with water led to a chlorohydrin, and elimination of HCl gave an enol that tautomerized to ketone 103 (obtained in 15% yield).88 Me

Me

Me

Me H 2O

90% H2SO 4

OH

Cl

0°C

O

Cl Cl 100

101

102

103 (15%)

Alkynes are useful partners in cationic cyclization reactions. Initial reaction of the allylic alcohol moiety in 104 with formic acid gave allyl cation 105. Subsequent attack by the alkyne moiety across the molecule generated vinyl cation 106. Trapping the formate anion generated a formate enol ester (C]CdOCHO) and hydrolysis liberated the final ketone product 107, which Lansbury and Serelis89 converted to damsinic acid.

88

Lansbury, P. T.; Nienhouse, E. J. J. Am. Chem. Soc. 1966, 88, 4290.

89

Lansbury, P. T.; Serelis, A. K. Tetrahedron Lett. 1978, 1909.

882

16. CARBON-CARBON BOND-FORMING REACTIONS

Me

OH Me

Me

Me

H

90% HCO2H

H 2O HCO2 –

Me H

O

Me

O

O 104

105

Me

O

106

Me

O

107

The Johnson polyene cyclization (the Stork-Eschenmoser postulate),90 described in Section 8.8.1, is an example of a cascade cationic cyclization. Polyenes (e.g., squalene) are expected to assume a steroid-like conformation in the lowest energy conformation (Section 1.5.5), based on the biogenetic preparation of cholesterol from squalene.91 In practice, initial reactions of a polyene with acid led to a very low yield of tri- or tetracyclic products, giving significant amounts of polymeric material. Diligent work over many years prevailed, however, and Johnson et al. 92 solved the many problems, as described in Section 8.8.1. This reaction is a useful and often efficient synthetic route to di-, tri-, and tetracyclic molecules. One of the later examples of polyene cyclization uses an allylic silane to quench the cyclization process.92 Initial reaction of the dioxolane unit in 108 with tin tetrachloride (SnCl4) led to a 23% yield of tetracyclic 109a (2:3 α:β 17-vinyl) and 11% of 109b (1:1 α:β 17-vinyl). In this case, the Lewis acid partially cleaved the ketal-protecting group (Section 5.3.3.1) during the cation initiation process. The cyclization reaction generated the stereochemical centers at the ring junctures with high selectivity, but provided little stereocontrol for the final reaction, which gave a mixture of 17α- and 17β-vinyl groups. Better results were obtained by use of a cyclopentenyl moiety rather than a dioxolane unit to conformationally anchor the cationic cyclization process and incorporating an alkyne to terminate that process (with a vinyl carbocation that was converted to a ketone).93 Me

Me

Me SiMe3

Me

0.2 M SnCl4 Pentane , 0 → 15°C

O

H

H

H

2

(a) R1 = –OCH2CH2OH , R = H , 23% (b) R1 = H , R2 = –OCH2CH2OH , 11%

O

R1

R2

108

109

Cyclization can be accomplished when there are heteroatom substituents. Reaction of amide-diene 110 with paraformaldehyde and formic acid generated iminium salt 111. Cationic cyclization, where the formate anion is the nucleophilic trapping agent, generated the perhydroisoquinoline 112.94 A related cyclization procedure that involves N-acyl iminium intermediates was developed by Speckamp and Hiemstra,95 and has been used for the preparation of alkaloids.96 NHCO2Et Ph

(HCHO) n , HCO2H

O-CHO

N 110

OHCO

Ph

Ph

111

CO2Et

N CO Et 2 H 112

(a) Johnson, W. S. Acc. Chem. Res. 1968, 1, 1; (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 348–349.

90

91

Clayton, R. B. Q. Rev. Chem. Soc. 1965, 19, 168.

92

Johnson, W. S.; Chen, Y.-Q., Kellogg, M. S. J. Am. Chem. Soc. 1983, 105, 6653.

93

Gravestock, M. B.; Johnson, W. S.; McCarry, B. E.; Parry, R. J.; Ratcliffe, B. E. J. Am. Chem. Soc. 1978, 100, 4274.

94

Kano, S.; Yokomatsu, T.; Yuasa, Y.; Shibuya, S. Tetrahedron Lett. 1983, 24, 1813.

95

Speckamp, W. N.; Hiemstra, H. Tetrahedron 1985, 41, 4367.

(a) Speckamp, W. N.; Hiemstra, H. Tetrahedron 1985, 41, 4367; (b) Speckamp, W. N.; DeBoer, J. J. J. J. R. Neth. Chem. Soc. 1983, 102, 405 (Chem. Abstr. 100: 51412b, 1984).

96

883

16.3 CARBOCATIONS AND CARBON-CARBON BOND-FORMING REACTIONS

This route has been used to prepare chiral heterocycles.97 In this reaction, an N-alkenyl, ω-hydroxy (or alkoxy) lactam (e.g., 113) is treated with formic acid, generating the acyl iminium salt 114. Selective reduction of an imide precursor gave the ethoxy lactam (113) required for generating the iminium salt (Section 7.4). Subsequent alkene cyclization led to 115,98 and trapping with the formate anion (an SN1 reaction) gave the final indolizidine product 116. The reaction appears to be under kinetic control and proceeds via a chair-like transition state, where the incoming nucleophile (HCO 2 ) and the carbon being attacked are antiperiplanar. Acyl iminium ions have become useful synthetic intermediates.99 Speckamp and Wijnberg applied this method to the synthesis of mesembrine.100 Overman and Robichaud used a variation of this reaction in a synthesis of geissoschizine.101 Initial reaction of 117 with paraformaldehyde and camphorsulfonic acid generated iminium salt 118 in situ. Reaction of the vinyl silane moiety with NaF led to addition of the C]C to the iminium salt. The alkylidene product (119) was produced in 80% yield as a 9.1:1 mixture of (Z/E) isomers. O O

N

HCO2H

OEt

N

O

HCO2

N

O

N OCHO

113

114

N

H SiMe3

N

115

N

Camphorsulfonic acid

N 30 equiv (HCHO) n H2O , 50°C , NaF

H Me

N

SiMe3

H

H

116

H

MeO2C

H

N H

H MeO2C

117

H

H

Me

Me

MeO2C

118

119 (80%)

A similar cyclization procedure used C-acylnitrilium ions for the synthesis of heterocyclic compounds.102 A typical example treated isonitrile 120 with trimethylacetyl chloride to yield 121.103 Subsequent treatment with CF3SO3Ag (silver trifluoromethanesulfonate) gave an intermediate C-acylnitrilium ion (122), which cyclized to give dihydroisoquinoline derivative 123 in 82% yield. O

N MeO

N C

Cl

CMe3

MeO Cl

CH2Cl2 , 25°C

MeO

O

CF3SO3Ag

CMe3

MeO 120

121

N MeO

N

MeO C CMe3

MeO

MeO

O 122

97

Royer, J.; Bonin, M.; Micouin, L. Chem. Rev. 2004, 104, 2311.

98

Speckamp, W. N. Rec. Trav. Chim. Pay-Bas 1981, 100, 345.

99

Maryanoff, B. E.; Zhang, H.-C.; Cohen, J. H.; Turchi, I. H.; Maryanoff, C. A. Chem. Rev. 2004, 104, 1431.

100

Wijnberg, J. B. P. A.; Speckamp, W. N. Tetrahedron Lett. 1975, 3963.

101

Overman, L. E.; Robichaud, A. J. J. Am. Chem. Soc. 1989, 111, 300.

102

Livinghouse, T. Tetrahedron 1999, 55, 9947.

103

Westling, M.; Livinghouse, T. Tetrahedron Lett. 1985, 26, 5389.

O 123 (82%)

CMe3

884

16. CARBON-CARBON BOND-FORMING REACTIONS

Livinghouse and coworkers104 used this strategy to synthesize dendrobine, where a key step was the coupling of 124 with 125 in the presence of molecular sieves. Subsequent treatment with silver tetrafluoroborate gave an 88% yield of 126. Stevens and Kenney105 had previously reported a Hg(II) nitrate-mediated Ritter coupling that involved formation of an organomercury compound. Addition to an isonitrile gave an iminium salt that could be reduced. Although this method did not involve an acylnitrilium ion, it is related to this chemistry.

+ Me

C N

O

CO2Me

OTBDMS

H

1. MS 4 Å , 40°C

Cl

Me

N

Me

124

CO2Me Me

–20°C

2. AgBF4 , –78

O

Me

O

125

Me

126 (88%)

Other cationic precursors are available. Aldehydes and ketones are protonated under various conditions to form an oxocarbenium ion (+CdOH), which is then attacked by an alkene to form a new carbon-carbon bond. In Section 6.2.2.2, it was shown that oxidation of acid-sensitive alcohols with PCC could be accompanied by secondary reactions. The oxidation of citronellol [(S)-3,7-dimethyloct-6-en-1-ol] with PCC not only gave the expected aldehyde, but the aldehyde oxygen was protonated to yield 127.106 Once formed, this oxocarbenium ion was attacked by the adjacent π-bond of the alkene to form cation 128. Elimination to an alkene (E1, Section 3.5.2), and oxidation to the ketone under the reaction conditions leads to the final product, isopulegone [(S)-5-methyl-2-(propan-2-ylidene)cyclohexan-1-one)]. H

OH Me Me Me

O

PCC

Me CH2Cl2

Me Me 127

Citronellol OH

O Me

Me

PCC

Me Me

– H+

Me

Me

128

Isopulegone

A variety of disconnections are possible utilizing intramolecular cationic reactions. A few of the more useful reactions follow: R1

R1

R R

Me

Me

Me R

Me R

R RO R

O

O

N

N

O

OR OR

16.3.2 Koch-Haaf Carbonylation When an alcohol is heated in strong acid, in the presence of carbon monoxide (CO), the initially generated carbocation intermediate traps CO to yield a carboxylic acid in what is known as the Koch-Haaf carbonylation.107 The reaction of 2,3,3-trimethylbutan-2-ol with aqueous sulfuric acid initially gave cation 129. In the presence of CO (generated in situ from formic acid and sulfuric acid), acylium ion 130 was formed and reaction with water gave 104

Lee, C. H.; Westling, M.; Livinghouse, T.; Williams, A. C. J. Am. Chem. Soc. 1992, 114, 4089.

105

Stevens, R. V.; Kenney, P. M. J. Chem. Soc. Chem. Commun. 1983, 384.

106

(a) Corey, E. J.; Boger, D. L. Tetrahedron Lett. 1978, 2461; (b) Corey, E. J.; Ensley, H. E.; Suggs, J. W. J. Org. Chem. 1976, 41, 380.

(a) Koch, H. Brennst. Chem. 1955, 36, 321 (Chem. Abstr. 50:6019 g 1956); (b) Koch, H.; Haaf, W. Annalen 1958, 618, 251; (c) M€ oller, K. E. Brennst. Chem. 1966, 47, 10 (Chem. Abstr. 64:12563b 1966); (d) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-52; (e) Mundy, B. P.; Ellerd, M. G.; Favaloro F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 368–369.

107

885

16.3 CARBOCATIONS AND CARBON-CARBON BOND-FORMING REACTIONS

2,2,3,3-tetramethylbutanoic acid in 90% yield. Similar treatment of alkenes with these reagents also gave carboxylic acids, but the yields were usually poorer. As with any other carbocation, rearrangement can occur, usually prior to reaction with CO. In most cases, using a sulfuric-formic acid mixture to generate CO is more efficient than bubbling in CO gas. Lactones are formed when diols are treated with CO and acid.108 C O

aq H2SO 4

OH

C O

H 2O

CO2H

HCO2H

2,3,3-Trimethylbutan-2-ol

129

2,2,3,3-Tetramethylbutanoic acid (90%)

130

The Koch-Haaf disconnection follow: R2HC CO2H

R2HC OH

16.3.3 Nazarov Cyclization Divinyl ketones are useful precursors for the synthesis of cyclopentenones.109 In a typical example, reaction of a divinyl alkyne (e.g. 131) with aqueous acid generated a conjugated ketone, (2E,6Z)-3,6-dimethylocta-2,6-dien-4-one, via hydration of an intermediate vinyl cation.109b Subsequent treatment with a mixture of phosphoric and formic acids led to a new oxocarbenium ion (132) that reacted with the adjacent alkene (a cyclization reaction) to form allylic carbocation 133. Elimination led to the conjugated ketone, 4-ethyl-2,3,4-trimethylcyclopent-2-en-1-one.109b This cationic ring-closing reaction is called the Nazarov cyclization.109,110 Several methods are available for the preparation of the bis(conjugated) ketones required for this reaction. Most involve coupling a vinyl carbanion equivalent with a conjugated carbonyl. O

Me

OH

+

aq H

Me 131

H3PO4

Me

Me

HCO2H

Me

(2E,6 Z)-3,6-Dimethylocta2,6-dien-4-one

132 O

OH

Me

Me

133

4-Ethyl-2,3,4-trimethylcyclopent-2-en-1-one

The strongly acidic conditions used in the reaction that gave 4-ethyl-2,3,4-trimethylcyclopent-2-en-1-one are not required for the cyclization,111 and Lewis acids [e.g., tin chloride (SnCl4) or iodotrimethylsilane] can be used in a Nazarov cyclization. A synthetic example, taken from Smith III and Shvartsbart’s synthesis of ()-calyciphylline N, involved treatment of 134 with tetrafluoroboric acid•etherate to yield the cyclopentenone unit in 135, in 82% yield.112 108

Takahashi, Y.; Yoneda, N.; Nagai, H. Chem. Lett. 1982, 1187.

(a) Nazarov, I. N.; Torgov, I. B.; Terekhova, L. N. Izv. Akad. Nauk. SSSR otd. Khim. Nauk 1942, 200; (b) Braude, E. A.; Forbes, W. F. J. Chem. Soc. 1953, 2208.

109

(a) Pellissier, H. Tetrahedron 2005, 61, 6479; (b) Frontier, A. J.; Collison, C. Tetrahedron 2005, 61, 7577; (c) Tius, M. A. Eur. J. Org. Chem. 2005, 2193. For reactions of this type, see (d) ref. 79, p 1308. Also see (e) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-64; (f) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 448–449.

110

111

Marino, J. P.; Linderman, R. J. J. Org. Chem. 1981, 46, 3696.

112

Shvartsbart, A.; Smith, A. B., III. J. Am. Chem. Soc. 2014, 136, 870.

886

16. CARBON-CARBON BOND-FORMING REACTIONS

OTBS PhthN

OTBS PhthN

O

Me

O

Me HBF4•OEt2 , CH2Cl2

Me Me

O

O

Me

Si

Me

Si

Me

Me 134

135 (82%)

The cyclization can be accomplished under even milder conditions by incorporating a stabilizing group on the alkene. Both silane and stannane derivatives have been prepared. Cyclization of 136 gave 137, for example, but both the temperature and the nature of the Lewis acid had a significant effect on the reaction.113 Ferric chloride (FeCl3) gave the best yield of the Nazarov product (137) in this study. Vinyl silane cyclization can proceed with modest diastereoselectivity.114 A variation in this approach was reported by West and Giese,115 who showed that formation of a reduced product accompanied formation of a normal product in a Lewis acid catalyzed Nazarov cyclization using triethylsilane. When 138 was treated with BF3•OEt2 and 2 equiv of triethylsilane, followed by 1 N HCl in a second step, a 71% yield of 139 was obtained along with 14% of 140.115 Other variations are known. Vinyl stannanes are good Nazarov precursors, for example, generating cyclopentenone derivatives.116 Harmata and Lee117 reported a retro-Nazarov reaction. This reaction appears to be a good route to dienones. O

O FeCl3

SiMe3 136

O

1. 1.1 BF3•OEt2 2 Et3SiH

137

H

O

O

H

+

2. 1 N HCl

H 138

H 139 (71%)

140 (14%)

Another carbon bond-forming reaction is related to the Nazarov cyclization, but it involves the reaction of diiron nonacarbonyl (see Section 18.9) with halocarbonyl compounds. Noyori et al. 118 found that α,α’-dibromoketones (e.g., 2,4-dibromopentan-3-one) react with diiron nonacarbonyl [Fe2(CO)9] to yield an iron-stabilized alkoxy zwitterion (141). The intermediate π-allyl iron species reacts with alkenes in a stepwise manner, initially producing 142 and then a cyclic ketone (e.g., 143). This product is equivalent to the product of a [3+2]-cycloaddition with an alkene (see Section 15.4). This cyclization method is now known as Noyori annulation.119 This strategy was used for a synthesis of nezukone.120 113

Denmark, S. E.; Jones, T. K. J. Am. Chem. Soc. 1982, 104, 2642.

114

Jones, T. K.; Denmark, S.E. Helv. Chim. Acta 1983, 66, 2397.

115

Giese, S.; West, F. G. Tetrahedron 2000, 56, 10221.

116

Peel, M. R.; Johnson, C. R. Tetrahedron Lett. 1986, 27, 5947.

117

Harmata, M.; Lee, D. R. J. Am. Chem. Soc. 2002, 124, 14328.

118

(a) Noyori, R.; Yokoyama, K; Hayakawa, Y. J. Am. Chem. Soc. 1973, 95, 2722; (b) Idem Ibid. 1978, 100, 1791; (c) Noyori, R. Acc. Chem. Res. 1979, 12, 61.

119

Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 464–465.

120

(a) Hayakawa, Y.; Sakai, M.; Noyori, R. Chem. Lett. 1975, 509; (b) Hayakawa, Y.; Baba, Y.; Makino, S.; Noyori, R. J. Am. Chem. Soc. 1978, 100, 1786.

887

16.3 CARBOCATIONS AND CARBON-CARBON BOND-FORMING REACTIONS

O-Fe(CO)4

O Me

O Me Fe2(CO)9

Me

Me

O

Me

Me

Me

Br

Br

R

Fe(CO)4

R

R 2,4-Dibromopentan-3-one

141

Me

142

143

The Nazarov disconnection follow: O

O R1

R

R2

R

R

R1

R

R2

R

O

X +

R1

X

R

R2

16.3.4 The Prins Reaction When aldehydes react with a Brønsted-Lowry acid, an oxocarbenium ion is formed (see Section 16.2.1). In the presence of an alkene, oxocarbenium ions will react with the alkene π-bond. Formaldehyde can be coupled to an alkene in the presence of an acid, for example, to yield a diol (144) or a 1,3-dioxane derivative (146) in what is known as the Prins reaction.121 Allylic alcohols (e.g., 145) can also be produced in this reaction. The use of alternative acids, including Lewis acids, may lead to other products. In a synthesis of FD-891 by Yadav et al.,122 alkene-alcohol (S,E)-2methylhex-4-en-1-ol was reacted with (R)-3-(benzyloxy)-2-methylpropanal, in the presence of trifluoroacetic acid, to give 147 in 50% overall yield. R

OH

CH2O

R

aq H+

+

OH

R

R

OH

+ O

144

145

O 146 OH

OH

OBn

+

1. TFA , CH2Cl2 , 0°C – rt

OHC

HO

2. MeOH , K2CO3 , rt

(S, E)-2-methylhex-4-en-1-ol

(R)-3-(Benzyloxy)-2methylpropanal

O

OBn

147 (50%)

The Prins disconnection follow: OH

OH

R

R

The examples in this section demonstrate that stable carbocations can be generated and trapped by reaction with alkenes, alcohols, or carbonyl derivatives. The resulting reaction products contain new carbon-carbon bonds, often produced with controllable and predictable stereo- and regiochemistry.

(a) Prins, H. J. Chem. Weekblad 1919, 16, 64, 1072, 1510 (Chem. Abstr. 13: 3155, 1919) and (Chem. Abstr. 14: 1119, 1920); (b) Arundale, R.; Mikeska, L. A. Chem. Rev. 1952, 51, 505; (c) Adams, D. R.; Bhatnagar, S. P. Synthesis 1977, 661; (d) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-76; (e) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 532–533.

121

122

Yadav, J. S.; Das, S. K.; Sabitha, G. J. Org. Chem. 2012, 77, 11109.

888

16. CARBON-CARBON BOND-FORMING REACTIONS

16.4 FRIEDEL-CRAFTS REACTIONS This section discusses a specialized reaction of carbocations and a nucleophilic species. In this section, however, the nucleophilic species is an aromatic ring and reaction with a carbocation generates an arenium ion intermediate (Section 3.10.2) that leads to a substituted aromatic system. Known as Friedel-Crafts reactions, they are among the most important reactions in organic chemistry.

16.4.1 Electrophilic Aromatic Substitution In Sections 3.10.1,B, electrophilic reagents (e.g., the bromonium ion [Br+]), generated by the reaction of diatomic bromine (Br2) and aluminum bromide (AlBr3)], were shown to react with benzene to yield a resonance-stabilized cation called an arenium ion. Loss of a proton from the arenium ion intermediate and concomitant aromatization by loss of a proton gave bromobenzene. In a similar manner, the formation of several substituted benzene derivatives from an arenium ion was discussed in Section 3.10. This section will discuss electrophilic substitution reactions of aromatic derivatives that form carbon-carbon bonds.123 In Section 3.10.2, the presence of electron-releasing groups on the aromatic ring (alkyl, oxygen, nitrogen, sulfur, aryl) led to ortho- and para-substituted derivatives as the major products. All of these substituents are electron rich, and most contain a δ dipole or have a () charge adjacent to the aromatic ring. Such groups are known as activating substituents because they increase the rate of substitution relative to benzene.123 Groups that release electron density towards the benzene ring increase the base strength (electron-donating ability) of the benzene ring, which increases the rate of reaction with an electrophilic species. The rate of nitration of anisole was 9.7 x 106 times faster than the rate of nitration of benzene, for example. The selectivity for ortho and para products for aromatic rings with an electronreleasing substituent is explained by examining the arenium ion intermediate for reaction at the ortho, meta, and para positions. In all cases, the arenium ion will have a positive charge in the ring. With an electron-releasing substituent, the positive charge of the arenium ion for reaction at the ortho and para positions is adjacent to a substituent that has a () or δ center (including an electron- releasing alkyl group). Such an intermediate is more stable, and for electron-releasing substituents they will favor formation of the ortho- and para-substituted products. Therefore, activating substituents are commonly called ortho-para directors. Reaction of anisole with a nitric acid-sulfuric acid mixture, for example, gave 44% of the ortho product (2-nitroanisole), 56% of the para product (4-nitroanisole) and

>

O

S

N

16.4.2 Friedel-Crafts Alkylation 16.4.2.1 Alkyl Carbocations and Regioselectivity This chapter deals with reactions that form carbon-carbon bonds via carbocation intermediates, which includes coupling reactions with aromatic rings. An aromatic compound will react with a carbocation intermediate (also see Section 16.4.1) to form an arenium ion intermediate (148). Subsequent loss of a proton, with concomitant aromatization, completes the substitution reaction to form arene 149 via electrophilic aromatic substitution. The reaction with an alkyl carbocation is called Friedel-Crafts alkylation,127 first reported by Friedel and Crafts in 1877.128 The carbocation required for this reaction can be generated from an alkene, an alcohol, or an alkyl halide. H

CR3

–H+

+ R3C

CR3 149

148

Solvolysis reactions of alkyl halides in aqueous or highly protic media were shown to form carbocations that typically lead to SN1 or E1 products (see Sections 3.2.2 and 3.5.2, respectively). Such reactions were limited to tertiary and secondary halides, due to the energy demands for the initial ionization. Alkyl halides also react with Lewis acids, but tertiary, secondary, or primary halides react to form tertiary, secondary, or primary carbocations (150). Formation of a primary cation (150, R1 ¼ R2 ¼ H, R3 ¼ alkyl) is slower relative to the others and is, of course, subject to rearrangement. As a result, primary alkyl derivatives are rarely observed under these conditions. The extent of rearrangement from primary and secondary carbocations depends on the reaction conditions, the strength of the nucleophile, and the nature of the carbocation. Reaction of benzene and 1-phenyl-2-chloropropane, for example, gave 60% of 1,2-diphenylpropane,129 with little rearrangement to the benzylic carbocation. This result suggests that the initially formed cation can be trapped very rapidly, possibly via a tight ion pair or via an intermediate complex. In this particular case, the solvent (carbon disulfide, CS2) played a major role in minimizing solvent separation of the ions. R1

R1 R2 C

Cl

AlCl3

R2 C

AlCl4

R3

R3

150 Cl

AlCl3 , CS2 , 0 °C

Me

PhH, 1 h

2-Chloro-1-phenylpropane

126

Ph Me 1,2-Diphenylpropane (60%)

Olah, G. A. Friedel-Crafts Chemistry; John Wiley: NY, 1973; p 42.

127

For reactions of this type, see Ref. 126, pp 129–133. Also see (a) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR34; (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 256–257. (a) Friedel, C.; Crafts, J. M. Compt. Rend. 1877, 84, 1392, 1450; (b) Price, C. C. Org. React. 1946, 3, 1; (c) Gore, P. Chem. Rev. 1955, 55, 229; (d) Baddeley, G. Q. Rev. Chem. Soc. 1954, 8, 355.

128

129

Masuda, S.; Nakajima, T.; Suga, S. Bull. Chem. Soc. Jpn. 1983, 56, 1089.

890

16. CARBON-CARBON BOND-FORMING REACTIONS

Alkyl fluorides are the most reactive of the alkyl halides. All the boron trihalides catalyze the reaction of alkyl fluorides and benzene, although they are ineffective catalysts for the analogous reaction with alkyl chlorides and bromides.130 In a reaction with mixed halides, the CdF bond reacts faster than the CdCl, CdBr, or CdI bonds.131 This order is consistent with the observed order of reactivity for alkyl halides132 with aluminum chloride (AlCl3): RdF > RdCl > RdBr > RdI As with other electrophilic aromatic substitution reactions, Friedel-Crafts alkylation of anisole, toluene, or aromatic derivatives bearing an electron-releasing group is faster than benzene, giving primarily ortho- and para- substituted products. The reaction is slowed by the presence of electron-withdrawing substituents (e.g., NO2 in nitrobenzene and C]O in aryl ketones) on the aromatic ring. Indeed, Friedel-Crafts alkylation is very difficult with deactivated aromatic nuclei. The product distribution varies with the halide, the catalyst, the solvent, and the aromatic substrate, but is generally 2:1 to 3:1 favoring ortho over para. Ortho-substituted products are generally preferred except with highly hindered alkyl halides (e.g., tert-butyl).133 This preference for the ortho product has been called the ortho effect, which is defined as additional stabilization of cation intermediates by an electron-releasing group to stabilize the intermediate generated by ortho-substitution relative to the intermediate generated by para-substitution.133 The ortho effect is more pronounced when the aromatic ring contains the stronger electron-releasing methoxy group (anisole) when compared to a methyl group (toluene). This effect is usually explained by the linear coordination of a heteroatom substituent with the incoming reagent. In the case of oxygen substituents, an O-alkyl product can be formed, followed by an O!C migration, where migration to the ortho carbon is faster than migration to the para carbon.134 Increasing the electron-releasing ability of the activating group also correlates with a decrease in the yield of meta product. Separation of the isomeric arene products can be a problem. When very active catalysts (e.g., aluminum chloride) are used for Friedel-Crafts alkylation with alkyl halides, significant amounts of meta products are often observed, even when the aromatic ring contains activating substituents.130a,b,131 Such products are due to the thermodynamic nature of the reaction with aluminum chloride, which shows a preference for the meta product. Excess catalyst and high reaction temperatures also favor more meta product. The use of BF3, sulfuric acid, ferric chloride (FeCl3), or zinc chloride (ZnCl2) in lower concentrations and at lower temperatures, are preferred if primarily ortho-para products are the desired targets.135 Friedel-Crafts alkylation with alkyl halides proceeds via a carbocation, and chiral alkyl halides are expected to give racemic arene products. The extent of racemization is dependent on the Lewis acid and the reaction conditions. Suga et al.136 found some asymmetric induction during the aromatic substitution, if the reaction temperature was maintained < 0 °C and the reaction time was kept to a minimum. Reaction of (2S)-chlorobutane and benzene gave 2-phenylbutane in up to 24 %ee (R) using short contact times at low temperature. Some asymmetric induction was observed when mild Lewis acids were used, even at ambient temperatures. The use of strong Lewis acids, higher temperatures, and longer contact times, however, led to the expected racemization. Presumably, at lower temperatures the CdCldM+X species is generated as a tight ion pair, where some asymmetric character is retained in the transition state of the reaction. The typical Friedel-Crafts alkylation disconnection follows: R +

R

X

16.4.2.2 Isomerization, Polyalkylation, and Deactivation Several problems are often encountered with Friedel-Crafts alkylation reactions. The first occurs when primary and secondary halides are used, because the cations that are formed are subject to rearrangement. Despite the reaction 130

(a) Ref. 126, p 40; (b) Oláh, G. A.; Kuhn, S.; Oláh, J. J. Chem. Soc. 1957, 2174.

131

Ref. 126, p 41 (Ref. 471 in Chapter II).

132

Calloway, N. O. J. Am. Chem. Soc. 1937, 59, 1474.

133

Olah, G. A.; Olah, J. A.; Ohyama, T. J. Am. Chem. Soc. 1984, 106, 5284.

134

Kovacic, P.; Hiller, J. H., Jr. J. Org. Chem. 1965, 30, 1581.

135

(a) Cullinane, N. M.; Leyshon, D. M. J. Chem. Soc. 1954, 2942; (b) Slanina, S. J.; Sowa, F. J.; Nieuwland, J. A. J. Am. Chem. Soc. 1935, 57, 1547.

136

Suga, S.; Segi, M.; Kitano, K.; Masuda, S.; Nakajima, T. Bull. Chem. Soc. Jpn. 1981, 54, 3611.

891

16.4 FRIEDEL-CRAFTS REACTIONS

of 1-phenyl-2-chloropropane in Section 16.4.2.1, the reaction of benzene with 1-bromopropane and aluminum bromide (AlBr3) generated primary carbocation 151. Rearrangement to the more stable secondary carbocation (152) was followed by reaction with benzene to give the final product, isopropylbenzene (cumene).137 Due to this propensity for rearrangement, preparation of primary arenes (straight-chain alkyl benzenes) is very difficult via Friedel-Crafts alkylation. On the other hand, the preparation of tertiary arenes (e.g., tert-butylbenzene), is straightforward because the easily formed tertiary carbocations usually do not rearrange. Br

AlBr3

1-Bromobutane

151

Cumene

152

Rearrangement can be accompanied by isomerization of the initially formed product by a secondary reaction with the Lewis acid (e.g., AlCl3), which is required for the initial reaction.138 Isomerization of groups can occur139 via 1,2shifts or via dissociation to a cation and readdition. 1,1-Dimethylpropylbenzene (tert-pentylbenzene) dissociated in the presence of AlCl3 to yield 153, for example. Under these conditions, the reaction was reversible and addition of the cation to the aromatic ring generated two new carbocations, 154, yielding (3-methylbutan-2-yl)benzene, or 155, giving neopentylbenzene. The extent to which neopentylbenzene is formed is related to the length of exposure of tert-pentylbenzene to AlCl3. This latter reaction is a thermodynamic process that favors the most stable isomer, and polyalkylation is also possible. Exposure of sec-butylbenzene to the mixed-acid HF-BF3 led to a mixture of 36.7% of benzene, 10.9% of n-butylbenzene, 21.8% of sec-butylbenzene, and 30.6% of di-sec-butylbenzene.140 Similarly, when n-butylbenzene was heated with aluminum chloride (100 °C, 3 h), 45.2% of butylbenzene (99.4% n-butylbenzene and 0.6% sec-butyl-benzene), 15.2% of benzene, and 27.7% of dibutylbenzene (>90% meta) along with 11.9% of polyalkylated benzene products were obtained.141 Lewis acid induced isomerization of dialkylbenzenes usually leads to an increase in the relative percentage of meta isomer, as seen with n-butylbenzene, based on the greater thermodynamic stability of the meta product relative to the ortho and para products.142 Me

Me

Me

Me

Me

AlCl3 , 80 °C

tert-Pentylbenzene

Me

Me

1h

(3-Methylbutan-2-yl)benzene

154

153 Me Me Me

Me

Me

Me

+

(36.7%) sec-Butylbenzene

+

(10.9%) Butylbenzene

Me Me

Neopentylbenzene

155

137

Me

Me

Me

Me

+

(21.8%)

Me

(30.6%)

sec-Butylbenzene

Ref. 126, p 68.

See (a) Nenitzesco, C. D.; Avram, M.; Sliam, E. Bull. Chim. Soc. Fr. 1955, 1266; (b) Nenitzescu, C. D.; Necsc¸oiu, I.; Glatz, A.; Zalman, M. Berichte 1959, 92, 10.

138

139

Ref. 126, p 70.

140

McCaulay, D. A.; Lien, A. P. J. Am. Chem. Soc. 1953, 75, 2411.

141

Kinney, R. E.; Hamilton, L. A. J. Am. Chem. Soc. 1954, 76, 786

142

Taylor, W. J.; Wagmen, D. D.; Williams, M. G.; Pitzer, K. S.; Rossini, F. D. J. Res. Na. Bur. Stand. Natl. (U.S.) 1946, 37, 95.

892

16. CARBON-CARBON BOND-FORMING REACTIONS

In Friedel-Crafts alkylation reactions, the alkyl substituents in the products are weakly electron releasing (and therefore activating), making the arene product more reactive than the benzene starting material. This increased reactivity leads to the second observed problem with Friedel-Crafts alkylation, polyalkylation. Alkylation of the arene leads to disubstituted products, which are often contaminants or occasionally, the major product in Friedel-Crafts alkylation reactions. If benzene, for example, reacts with 2-chloropropane, the initial product is isopropylbenzene (cumene). Since cumene is more reactive than benzene, it reacts with the carbocation intermediate in competition with benzene, leading to both 1,2-diisopropylbenzene and 1,4-diisopropylbenzene as secondary products. The steric hindrance inherent to formation of the ortho-disubstituted product can lead to a large proportion of the para product, if the steric interaction is severe. In a large excess of benzene, polyalkylation143,128b can be suppressed. Francis144 reported alkyl groups on benzene have only a small effect on the rate of Friedel-Crafts alkylation. Francis144 showed that “alkylation occurred in a heterogeneous reaction system, specifically in the catalyst layer and that the reason for polysubstitution is the preferential extraction of the early reaction product by this catalyst layer”.145 Using a solvent that will solubilize both hydrocarbon and catalyst will minimize polyalkylation, as does efficient (high speed) stirring and higher reaction temperatures, which solubilizes the aluminum chloride.139 Cl

+

+

AlCl3

Cumene

1,2-Diisopropylbenzene

1,4-Diisopropylbenzene

When an aromatic ring contains several alkyl groups or complex alkyl groups, treatment with strong acids can lead to formation of isomeric polyalkylated benzene derivatives at relatively low temperatures (10:1 regioselectivity after elimination of the initially formed hydroxy compound. Other cation precursors can be used in these reactions. Ts

H

H

N

AlCl3 , CH2Cl2

Me

MeO

N

Ts

rt , 12 h

MeO

OH

Me

161

162 (88%)

O

O H2O-AcOH-Acetone

O

N

40°C

O

OMe

t-BuO

N

t-BuO

OMe 163

164

Typical Friedel-Crafts disconnections involving alcohol and alkene precursors follow: R + X

R R

R

R

R R R

+ R1 X

R

R1 R

16.4.4 Friedel-Crafts Acylation The reaction of a Lewis acid with an acyl halide or an anhydride generates a resonance-stabilized acylium ion (165). Once formed, the π-bond of a benzene ring attacks 165 to produce the usual arenium ion intermediate 166, and loss of a proton accompanied by aromatization yields an aryl ketone (167). This reaction is known as Friedel-Crafts acylation.164,128a The carbonyl group in the ketone product is polarized δ+ and attached directly to the aromatic ring, so it is deactivated relative to benzene. For this reason, 167 is less reactive than benzene, and further reaction to give a polyacylated derivative is not a problem. In other words, Friedel-Crafts acylation does not lead to polyacylation. The resonance- stabilized acylium ion (165) is not subject to skeletal rearrangement prior to reaction with the aromatic ring. Isomerization of the ketone product is not a problem. H O

O

O

AlCl3

R

Cl

AlCl4 R

163 164

R – H+

O

O

R 165

R

166

167

Movassaghi, M.; Ondrus, A. E. Org. Lett. 2005, 7, 4423.

For reactions of this type of reaction, see Ref. 126, pp 1422–1433. Also see (a) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-34; (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 254–255.

896

16. CARBON-CARBON BOND-FORMING REACTIONS

Acylium ions can be produced from acyl bromides and iodides, but the more common acid chlorides are typically used. Several catalysts can be used for the reaction of benzoyl chloride and anisole.165 In all cases, the para product (169) predominated over the ortho product (168) by 31:1.165 Mild Lewis acids can be used to facilitate Friedel-Crafts cyclization, which is particularly important for synthetic targets bearing other functional groups. OMe O

O

Ph

Ph

Cl

MeO

Ph

MeO

+

O

Lewis acid

168

169

Friedel-Crafts acylation is not limited to reactions with benzene, of course, but deactivated aromatic rings either do not undergo Friedel-Crafts reactions at all, or do so with difficulty. Benzamide derivatives, for example, typically give no reaction in Friedel-Crafts acylation reactions. Activated aromatic rings react quite well in most cases. In a synthesis of chlorocyclinone A by Karmakar and Mal,166 acetyl chloride reacted with phenol 170 in the presence of titanium tetrachloride to give a 75% yield of 171. Other carboxylic acid derivatives can be used in Friedel-Crafts acylation reactions, including anhydrides and esters. Buccini and Piggott167 reacted 172 with carboxylic acid 173 and trifluoroacetic anhydride to give ketone 174 in 85% yield, presumably by formation of an intermediate anhydride in situ, in an unsuccessful route to the synthesis of raderachol. Lactones are useful substrates in Friedel-Crafts cyclization reactions, generating ketones.168 MeO

CH3

MeO

CH3

CH3COCl , TiCl4

CO2Me

CO2Me

CH2Cl2 , 0°C – rt

OH

O

170

OH 171 (75%)

OMe O

OMe O

CO2Me

CO2Me

TFA , TFAA

+ HO

Reflux

OMe

OMe

172

173

174 (85%)

Intramolecular Friedel-Crafts acylations are extremely valuable for the synthesis of polycyclic compounds. An example is taken from Baati and coworker’s169 synthesis of rhein and diacerhein, treated acid 175 with trifluoroacetic acid to give ketone 176 in 96% yield. OMe

OMe

OMe

OMe

TFA CH2Cl2

Et2N CO2H

O

Et2N O

175

O 176 (96%)

For intramolecular reactions, the length of the tether that links the aromatic ring with the acylium ion directly influences the site of attack on the aromatic nucleus, as well as the facility of the cyclization. The aromatic nucleus can be benzene, another simple aromatic compound, a polynuclear aromatic hydrocarbon, or a heteroaromatic compound (e.g., pyrrole), furan, or thiophene. An example of this latter type of substrate is the reaction of pyrrole-diester 177 165

(a) Tsukervanik, I. P.; Veber, N. V. Dokl. Akad. Nauk. SSSR 1968, 180, 892; (b) Ref. 12, p 195.

166

Karmakar, R.; Mal, D. J. Org. Chem. 2012, 77, 10235.

167

Buccini, M.; Piggott, M. J. Org. Lett. 2014, 16, 2490.

168

Inouye, Y.; Uchida, Y.; Kakisawa, H. Chem. Lett. 1975, 1317.

169

Gonnot, V.; Tisserand, S.; Nicolas, M.; Baati, R.; Mioskowski, C. Tetrahedron Lett. 2007, 48, 7117.

897

16.4 FRIEDEL-CRAFTS REACTIONS

with BBr3, which led to ketone 178 in Vallee and coworker’s170 synthesis of (R)-(+)-myrmicarin. In intermolecular Friedel-Crafts acylation reactions, pyrroles usually give the α-acyl derivative with the soft acylium electrophile (Section 2.4), as with the formation of 178.

BBr3

N EtO2C

CO2Et

O

N EtO2C

177

178

When pyrrole was treated with acetic anhydride (at 150–200 °C), a mixture of 2-acetylpyrrole and 2,5diacetylpyrrole was formed by reaction with the soft acylium ion.171 Using a hard electrophile (see Section 2.4, e.g., trimethylsilyl triflate, Me3SiOTf), however, led to the β-derivative.172 Acylation of an N-silyl pyrrole led predominantly to the β-isomer rather than the α-isomer (an 83:17 mixture in 46% yield), when reacted with either anhydrides or acyl chlorides.173 The bulky SiMe3 group presumably blocked the C2 position. For benzene derivatives, both ortho positions are subject to acylation if they do not bear substituents. Friedel-Crafts acylation of 179 gave a mixture of coumarins 180 and 181.174 Sulfuric acid gave a mixture favoring 181, whereas a mixture of phosphorus pentabromide and aluminum chloride (PBr3/AlCl3) was highly regioselective for the formation of 180. It is clear that the ortho selectivity depends on the catalyst used, as well as on the nature of the substrate. Cl

CO2H

Cl

Me

O

O PBr3

Cl

Cl

+

O

Cl

AlCl3

Cl

O

Me

O

180

179

Me

181

A related acylation procedure uses iminium salts rather than acyl halides. The Vilsmeier-Haack reaction175 is a wellknown process, illustrated by reaction of pyrrole with the POCl3 complex of N,N-dimethylacetamide (182, which can decompose to a chloroiminium salt). The acylation reaction initially gave 183, which was converted to 2-acetylpyrrole [1-(1H-pyrrol-2-yl)ethan-1-one] by hydrolysis with aqueous sodium acetate.176 A synthetic example is taken from a synthesis of zeaenol by Jana and Nanda,177 in which 1,3-dimethoxy-5-vinylbenzene reacted with POCl3 and DMF, and then aq sodium acetate to yield 2,4-dimethoxy-6-vinylbenzaldehyde in 87%. Me N H

Me

Cl

+

Cl

Me

N P

Me

O

aq NaOAc

N H

H

Me

O 182

Me

N

N Me

O

2-Acetylpyrrole

183

OMe

OMe CHO

POCl3 , DMF

MeO 1,3-Dimethoxy-5-vinylbenzene

MeO 2,4-Dimethoxy-6-vinylbenzaldehyde (87%)

170

Sayah, B.; Pelloux-Leon, N.; Vallee, Y. J. Org. Chem. 2000, 65, 2824.

171

Katritzky, A. R. Handbook of Heterocyclic Chemistry; Pergamon Press: Oxford, UK, 1985, p 254.

172

Simchen, G.; Majchrzak, M. W. Tetrahedron Lett. 1985, 26, 5035.

173

McCombie, S. W.; Shankar, B. B.; Ganguly, A. K. Tetrahedron Lett. 1987, 28, 4123.

174

Provided through the courtesy of Dr. Frank Urban, Pfizer Central Research, Groton, CT, from (a) Lipinski, C. A. U.S. Patent 4,853,410, 1989; EP 230379A2, 1988 [Chem. Abstr. 108: 75224 h, 1988]; (b) Moore, B. S. Unpublished results. (a) Vilsmeier, A.; Haack, A. Berichte 1927, 60, 119; (b) deMaheas, M. R. Bull. Chim. Soc. Fr. 1962, 1989; (c) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-96; (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 558–559.

175

176

Ref. 171, p 254.

177

Jana, N.; Nanda, S. Eur. J. Org. Chem. 2012, 4313.

898

16. CARBON-CARBON BOND-FORMING REACTIONS

A variation of the Vilsmeier-Haack reaction uses a formamide (e.g., N-methyl-N-phenylformamide) with benzene (or another aromatic compound) to yield two products, the aromatic aldehyde (benzaldehyde) and the amine precursor to the formamide (N-methylaniline in this case).175a A solvent-free Vilsmeier reaction has been reported using microwave irradiation on silica gel.178 Me CHO

POCl3

N

+

NHMe

+

CHO

N-Methyl-N-phenylformamide

16.4.5 Synthesis of Polycyclic Aromatics That Do Not Contain Nitrogen The Haworth phenanthrene synthesis179 is a classical reaction that exploits the power inherent to Friedel-Crafts techniques, employing anhydrides in a Friedel-Crafts reaction with aromatic compounds to prepare polynuclear aromatic derivatives in a stepwise manner. When 1-methylnaphthalene was heated with succinic anhydride in the presence of AlCl3, keto acid 184 was formed. Clemmensen reduction (Section 7.11.6) of the ketone gave 185, and a second FriedelCrafts cyclization in the presence of sulfuric acid led to tricyclic ketone 186. Reduction of the carbonyl and aromatization by heating with selenium led to 1-methylphenanthrene. Note that a naphthalene derivative would be formed if benzene were used as a starting material in this sequence. Intermediate products produced by this sequence can be used for other purposes, making this sequence even more useful. A modified Haworth synthesis was used in a synthesis of hamigeran B by Clive and Wang,180 in which m-cresol was converted to 187, and then methylated to yield 188 (see Section 13.3.1). Subsequent Friedel-Crafts cyclization gave 189 in 83% yield. Note the use of Zn/HCl/HgCl2 for reduction of the ketone after acylation with succinic anhydride, and POCl3 to initiate the Friedel-Crafts acylation with the carboxylic acid. O O

O

AlCl3 PhNO2

Me

OH

m-Cresol

1. Zn , HCl 2. Se , Heat

Me 185

OH

1. Me2SO 4 , NaOH 2. LDA , THF-HMPA Then MeI

186

1-Methylphenanthrene OMe

OMe POCl3 , Heat

CO2H

HO2C 187

Me

Me

184

1. Succinic anhydride AlCl3 2. Zn , HCl HgCl2 , Heat

O

H2SO 4

Zn , HCl

Me

1-Methylnaphthalene

CO2H

CO2H

O

188

O 189 (83%)

Another useful sequence that produces polynuclear aromatic compounds is the Bally-Scholl synthesis.181 This variation involves addition of a glycerol derivative (a propane-1,2,3-triol) to an aromatic ring. An example is the reaction of anthraquinone with pentane-1,2,3-triol to form an alkenyl diol. Under acidic conditions, elimination gave a diene, which reacts with the acid to give a carbocation in situ. Cyclization to the aromatic ring gave 1-ethylmesobenzanthrone, or 3-ethyl-7H-benzo[de]anthracen-7-one, but only in 5% yield.182 The Skraup reaction183 178

Paul, S.; Gupta, M.; Gupta, R. Synlett 2000, 1115.

(a) Haworth, R. D. J. Chem. Soc. 1932, 1125; (b) Haworth, R. D.; Mavin, C. R. Ibid. 1932, 2720; (c) Haworth, R. D. Ibid. 1932, 2717; (d) Haworth, R. D.; Letsky, B. M.; Mavin, C. R., Ibid. 1932, 1784; (e) Haworth, R. D.; Bolam, F. M. Ibid. 1932, 2248; (f) Haworth, R. D.; Mavin, C. R.; Sheldrick, G. Ibid. 1934, 454; (g) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-41.

179

180

Clive, D. L. J.; Wang, J. J. Org. Chem. 2004, 69, 2773.

181

(a) Bally, O. Berichte 1905, 38, 194; (b) Bally, O.; Scholl, R. Ibid. 1917, 44, 1656; (c) Meerwein, H.; Klinz, J. J. Prakt. Chem. 1918, 97, 235.

182

Baddar, F. G.; Warren, F. L. J. Chem. Soc. 1938, 401.

(a) Skraup, Z. H. Berichte 1880, 13, 2086; (b) Manske, R. H.; Kulka, M. Org. React. 1953, 7, 59; (c) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-87; (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed., Wiley-Interscience: NJ, 2005; pp 602–603.

183

899

16.4 FRIEDEL-CRAFTS REACTIONS

presents a variation in which an aniline derivative reacts with glycerol in the presence of acid, to generate a quinoline derivative. While this reaction is perhaps better presented in the section relating to the synthesis of heteroaromatic compounds (Section 16.5), it is shown here because of the similarity to the Bally-Scholl synthesis. In an example of this reaction, Rapoport reacted 3-aminopyridine with glycerol to produce 1,5-naphthyridine in 31% yield.184 The reaction is regioselective185 rather than regiospecific, since Rapoport reported about 4% of 3-methyl-1,5-naphthyridine and 3-ethyl-1,5-naphthyridine. Another Friedel-Crafts route to phenanthrenes is the Bardhan-Sengupta phenanthrene synthesis.186 An example is the cyclization of phenethylcyclohexanol by reaction with phosphorus pentoxide at 140 °C, followed by aromatization by heating with selenium to yield a phenanthrene derivative, 1-methyl-7-isopropylphenanthrene (otherwise known as retene).186 An improvement in the reaction used hydrogen fluoride (HF) to induce cyclization.187 Both phenanthrenes and hydronapthalenes can be synthesized via the Bogert-Cook synthesis.188,186 When the Grignard reagent derived from 1-(2-bromoethyl)-3-isopropylbenzene reacted with 2,5-dimethylcyclohexanone, alcohol 190 was produced in 43% yield. Treatment with 85% sulfuric acid effected dehydration of the alcohol to a cyclohexene derivative, which reacted with the acid to give a carbocation in situ, and reaction with the aromatic ring gave 191 in 81% yield. Dehydrogenation with selenium at 300 °C gave retene (7-isopropyl-1-methylphenanthrene), whereas hydrogenation with Ni(R) (Section 7.10.7) gave abietane in 39% yield.189 The use of 1-bromonaphthalene derivatives as a starting material will lead to larger polycyclic rings.188b An alkene unit can also be used to generate a cation, and subsequent cyclization will generate a polycyclic aromatic compound. The reaction of phenethylmagnesium bromide with acetaldehyde gave but-3-en-1-ylbenzene, which was cyclized by treatment with sulfuric acid to yield 1,2,3,4tetrahydronaphthalene. This modification does not incorporate the additional ring of the cyclic ketones used in the Bogert-Cook reaction; it is sometimes called the Bogert synthesis. Me

Me HO

Br

Me

Se , 300°C , 6 h

Me

1. Mg

Me Me

2.

Me

Me

Me

85% H2SO 4

Retene

Me Me

Me

O

Me

Me

Me

1-(2-Bromoethyl)-3-isopropylbenzene

190 (43%)

3 H2

Me

Ni(R) 225°C

191 Me Me Me Abietane

1. MeCHO

MgBr

H2SO 4

2. KHSO4

Phenethylmagnesium bromide

But-3-en-1-ylbenzene

1,2,3,4-Tetrahydronaphthalene

A related modification is called the Darzens tetralin synthesis.190 In this process, the alkenyl group is used directly as a carbocation precursor. An example is treatment of 2-benzylpent-4-enoic acid with concentrated sulfuric acid, followed

184

Rapoport, H.; Batcho, A. D. J. Org. Chem. 1963, 28, 1753.

(a) Perche, J.-C.; Saint-Ruf, G.; Buu-Hoï, N. P. J. Chem. Soc. Perkin Trans. 1972, 1, 260; (b) Buu-Hoï, N. P.; Jacquignon, P.; Thang, D. C.; Bartnik, T. Ibid. 1972, 263. 185

186

Bardhan, J. C.; Sengupta, S. C. J. Chem. Soc. 1932, 2520, 2798.

187

Renfrow, W. B.; Renfrow, A.; Shoun, E.; Sears, C. A. J. Am. Chem. Soc. 1951, 73, 317.

188

(a) Bogert, M. T. Science 1933, 77, 289; (b) Cook, J. W.; Hewett, C. L. J. Chem. Soc. 1933, 1098.

189

Sterling, E. D.; Bogert, M. T. J. Org. Chem. 1939, 4, 20.

(a) Darzens, G. Compt. Rend. 1926, 183, 748; (b) Idem Ibid. 1935, 201, 730; (c) Darzens, G.; Levy, A. Ibid. 1936, 202, 427; (d) Idem Ibid. 1936, 203, 669; (e) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-22.

190

900

16. CARBON-CARBON BOND-FORMING REACTIONS

by gentle heating to yield the tetralin (tetrahydronaphthalene, 192).191 The difference between this reaction and the Bogert synthesis is the method used to form the alkenyl arene. In general, tetralin formation from an alkenylbenzene derivative is most often referred to as the Darzens tetralin synthesis rather than the Bogert synthesis. CO2H

CO2H 80% H2SO 4 < 45°C

Me 2-Benzylpent-4-enoic acid

192

The Bradsher reaction192 converts aryl ketones (193) to polycyclic compounds (e.g., 194, X ¼ CR2). This reaction can also be applied to heterocyclic derivatives of 193 (X ¼ O, S, Se). A common route to ketones (e.g., 193) is shown as part of a synthesis of 9-ethylanthracene. Two aryl groups were incorporated by a Grignard reaction of 2-chlorobenzaldehyde with phenylmagnesium bromide, producing 1-benzyl-2-chlorobenzene in 81% yield. Subsequent treatment with cuprous cyanide (also see Section 3.10.5) led to 2-benzylbenzonitrile, and reaction with ethylmagnesium bromide gave the ketone, 1-(2-benzylphenyl)propan-1-one. Subsequent reaction with HBr (heated at reflux for 4 d) generated 9-ethylanthracene in 69% yield.192a X

X O

R

R 193 CHO

a

194 c

b Cl

Cl 2-Chlorobenzaldehyde

d O

NC

1-Benzyl-2-chloro(81%) benzene

2-Benzylbenzonitrile

1-(2-Benzylphenyl)propan-1-one

9-Ethylanthracene

(69%)

(a) i. PhMgBr ii. I2 , P , AcOH , Reflux , 29 h (b) CuCN , 250°C , 23 h (c) i. EtMgBr , PhH , Reflux, 20 h ii. H3O+ (d) 30% HBr , AcOH , 4 d

Note that the alkene reactions given in this section suggest that Friedel-Crafts cyclization reactions are sensitive to minor changes in substrate and reaction conditions. The disconnections found in this section follow: R

X

O CHO

R R R

R1

R1

O

R

X

16.5 FRIEDEL-CRAFTS REACTIONS: FORMATION OF HETEROATOM-CONTAINING DERIVATIVES Friedel-Crafts alkylation and acylation reactions are quite versatile. They form the basis of several named reactions that have been used to synthesize heteroatom-containing compounds (e.g., quinolines, isoquinolines, and others). This section examines several of the more common transformations that will show the scope and utility of these processes. It will also introduce a modest amount of heterocyclic chemistry into our synthetic discussion. 191

Darzens, G.; Levy, A. Compt. Rend. 1935, 200, 469.

192

(a) Bradsher, C. K. J. Am. Chem. Soc. 1970, 62, 486; (b) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-14.

901

16.5 FRIEDEL-CRAFTS REACTIONS: FORMATION OF HETEROATOM-CONTAINING DERIVATIVES

16.5.1 Synthesis of Quinoline Derivatives The quinoline unit is an essential feature of many natural products, and there are several reactions that generate quinolines. One of the older, yet still useful methods, is the Knorr quinoline synthesis.193 This reaction couples an aromatic amine (usually aniline or a derivative) with a β-keto ester. This particular reaction is subject to formation of regioisomeric products, depending on the temperature and reaction conditions. Heating aniline and ethyl acetoacetate to 140 °C, for example, gave 50% of 4-methylquinolin-2-ol. However, if aniline and ethyl acetoacetate were reacted at room temperature for 5 d (or for 3 d with a catalytic amount of HCl), 60% of the isomeric hydroxyquinoline 2-methylquinolin-4-ol was isolated. A 70% yield of 2-methylquinolin-4-ol was also obtained when aniline was treated with ethyl acetoacetate in ethanol with a catalytic amount of acetic acid in the presence of Drierite®. In an example taken from a synthesis of lavaendamycin, by Nissen and Detert,194 initial cyclization of aniline derivative 195 via treatment with sulfuric acid led to a Knorr quinoline reaction that gave pyridone 196 in 97% yield. Further modification was required for the synthesis, and heating the pyridone derivative with POCl3 in DMF gave a 95% yield of the 2-chloroquinoline derivative, 197. OH

O Me

N

Me

or

CO2Et

Me

rt, 5d EtOH , Drierite , 4 h cat AcOH , Steam bath

140°C, 4 h

NH2

N

2-Methylquinolin-4-ol (60%) OMe

OMe

N H

OH

4-Methylquinolin-2-ol (50%) OMe

OMe O2N

Me

O

CO2Et

O

OMe

H2SO 4

POCl3, DMF

O2N

N H

OMe

O

O2N

N

OMe 195

Cl

OMe

196 (97%)

197 (95%)

The lower temperature variation of this reaction initially forms an imine or an enamine. Friedel-Crafts cyclization leads to the 4-hydroxyquinoline in what is called the Conrad-Limpach reaction.195 This reaction generally gives the opposite regioisomeric product to that obtained by the Knorr quinoline synthesis. The initially formed product is usually the enamine (as in the formation of 198 from aniline and ethyl acetoacetate).196 Under acidic conditions the iminium salt was formed and cyclized with the aromatic ring. A more efficient method simply heated 198 to 250 °C in mineral oil, which gave a 90% yield of 2-methylquinolin-4-ol. A variety of other functional groups can be tolerated in the molecule when this procedure is used. OH CO2Et

O Me

CO2Et

250°C Mineral oil

N

NH2

Me

N

Me

H 198

2-Methylquinolin-4-ol (90%)

(a) Knorr, L. Annalen 1886, 236, 69; (b) Idem Ibid. 1888, 245, 357; (c) Bergstr€ om, F. W. Chem. Rev. 1944, 35, 77 (see p 157). Also see (c) Hodgkinson, A. J.; Staskun, B. J. Org. Chem. 1969, 34, 1709; (d) Hauser, C. R.; Reynolds, G. A. J. Am. Chem. Soc. 1948, 70, 2402; (e) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-52;.

193

194

Nissen, F.; Detert, H. Eur. J. Org. Chem. 2011, 2845.

(a) Conrad, M.; Limpach, L. Berichte 1887, 20, 944; (b) Idem Ibid. 1891, 24, 2990; (c) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-19; (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 166–167.

195

196

(a) Manske, R. H. Chem. Rev. 1942, 30, 113 (see p 121); (b) Reitsema, R. H. Ibid. 1948, 43, 43 (see p 47).

902

16. CARBON-CARBON BOND-FORMING REACTIONS

The Gould-Jacobs reaction197 is related to the reactions just described, in that an aromatic amine reacted with EtOCH]C(CO2Et)2 to generate an enamino ester. Subsequent heating to 250 °C induced cyclization to give 4-hydroxyquinoline ester in 83% yield.197a,198 This sequence can be used with many aromatic substrates, but when secondary amines are used as precursors the reaction takes a slightly different course. The Knorr and Conrad-Limpach disconnections follow: R

OH

O

O +

+ R NH2

O

N H

CO2R1

NH2

R

N

R

CO2R1

If a β-diketone is used rather than a β-keto ester, the result is a 4-alkylquinoline in what is known as the Combes quinoline synthesis.199 Reaction of aniline with acetyl acetonate (pentane-2,5-dione), for example, generated enamine 199 (the tautomer of imine 200). Enolization to 201 in the presence of HF was followed by cyclization to give 2,4dimethylquinoline in 96% yield.200,199a The Doebner-Miller reaction201 is the reaction of a primary aromatic amine (e.g., aniline) with a carbonyl compound (e.g., acetaldehyde) in the presence of acid to yield a 2-alkylquinoline. Reaction of aniline with acetaldehyde gave (E)-N-phenylethanimine, but a subsequent aldol condensation with the imine derived from acetaldehyde (Section 13.4.6) led to 202.202 Friedel-Crafts cyclization was followed by loss of aniline to yield 2-methyl-1,2-dihydroquinoline, and aromatization gave 2-methylquinoline. NH2

O

H N

O

Me

Me

N

N

O

Me 199

HO

Me

O 200

Me HF

Me

N

Me

Me

201

2,4-Dimethylquinoline (96%)

Ph N

H+

Aldol

2 MeCHO HCl , Reflux

N

Me

N

NH2

H (E)-N-Phenylethanimine

–PhNH2

N

Me

Me

N

Me

H 2-Methyl-1,2dihydroquinoline

202

2-Methylquinoline

The Combes and Doebner-Miller disconnection follow: + N

R

RCHO

NH2

The Doebner reaction203,199b condensed aromatic amines and aldehydes with pyruvic acid derivatives. Reaction of p-toluidine with pyruvic acid (2-oxopropanoic acid) gave the γ-aminopyruvic acid (203) in situ, and subsequent (a) Gould, R. G., Jr.; Jacobs, W. A. J. Am. Chem. Soc. 1939, 61, 2890; (b) Ref. 196b, see p 53; (c) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-37; (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 278–279.

197

198

(a) Markees, D. G.; Schwab, L. S. Helv. Chim. Acta 1972, 55, 1319; (b) Albrecht, R.; Hoyer, G.-A. Berichte 1972, 105, 3118.

(a) Combes, A. Bull. Chim. Soc. Fr. 1888, 49, 89; (b) Ref. 193c, see p 156; (c) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-19; (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 164–165.

199

200

Born, J. L. J. Org. Chem. 1972, 37, 3952.

201

(a) Doebner, O.; Miller, W. Berichte 1883, 16, 2464; (b) Ref. 193c, see p 154.

202

(a) Ogata, Y.; Kawasaki, A.; Suyama, S. J. Chem. Soc. B 1969, 805; (b) Forrest, T. P.; Dauphinee, G. A.; Miles, W.F. Can. J. Chem. 1969, 47, 2121.

(a) D€ obner, O. Annalen 1887, 242, 265; (b) Idem Berichte 1887, 20, 277; (c) Idem Ibid. 1894, 27, 352, 2020; (c) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-25; (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 212–213. 203

903

16.5 FRIEDEL-CRAFTS REACTIONS: FORMATION OF HETEROATOM-CONTAINING DERIVATIVES

cyclization and air oxidation under Friedel-Crafts conditions generated the 4-carboalkoxyquinoline (2,6dimethylquinoline-4-carboxylic acid).204 Presumably, the reaction proceeds via initial imine formation (via reaction with benzaldehyde) followed by condensation with pyruvate in its enol form.193c,d Similar cyclization occurs with imines and pyruvate derivatives.205 CO2H

Me

+

Me

PhCHO

O

EtOH

HO2C

N

Pyruvic acid

N

Me

H

NH2 p-Toluidine

CO2H

O

Me

Me

Me

2,6-Dimethylquinoline-4-carboxylic acid

203

The Doebner disconnection follows: CO2H

O +

R

NH2 N

CO2H

R

16.5.2 Synthesis of Isoquinoline Derivatives Isoquinolines are as important as quinolines in the chemistry of natural products, and there are also many syntheses of these compounds. One of the most important Friedel-Crafts routes is the Bischler-Napieralski reaction.206 This process is the reaction of an N-acyl amine (e.g., N-phenethylacetamide) derived from phenethylamine (2-phenylethan-1amine), with a reagent (e.g., PPA or P2O5) to yield a dehydroisoquinoline derivative (1-methyl-3,4-dihydroisoquinoline). For example, in a synthesis of ()-(S)-stepholidine, Yang and Cheng207 reacted amide 204 with POCl3, and obtained 205, but 205 was found to be unstable and immediately reduced to the amine by hydrogenation to yield 205 in 84% overall yield. Ac2O

H

2-Phenylethan-1-amine

Me

N

NH2

P 2O 5

O

N-Phenethylacetamide

MeO

BnO

N

Me 1-Methyl-3,4-dihydroisoquinoline

MeO HN

O

OAc 1. POCl3 , MeCN

OMe

BnO

N

2. H2 , chiral catalyst

OMe

OBn 204

OAc

OBn 205 (84% overall)

A useful variation first converts the amino group to a formamide derivative, usually by treatment of a ketone (e.g., 206) with formamide and formic acid. Treatment of 206 with these reagents led to formamide 207. When treated with POCl3, cyclization occurred via the N-acyl group to give the isoquinoline, 208, in 66% yield. Zee-Cheng and Cheng208 used this sequence in a synthesis of nitidine. 204

D€ obner, O.; Gieseke, M. Annalen 1887, 242, 290.

205

Cuisa, R.; Musajo, L. Gazz. Chim. Ital. 1929, 59, 796.

(a) Bischler, A.; Napieralski, B. Berichte 1893, 26, 1903; (b) Bischler, A. Ibid. 1893, 26, 189, 1891; (c) Whaley, W. M.; Govindachari, T. R. Org. React. 1951, 6, 74; (d) Fodor, G.; Gal, J.; Phillips, B. A. Angew. Chem. Int. Ed. Engl. 1972, 11, 919; (e) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-11; (f) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; WileyInterscience: NJ, 2005; pp 96–97.

206

207

Cheng, J.-J.; Yang, Y.-S. J. Org. Chem. 2009, 74, 9225.

208

Zee-Cheng, K. Y.; Cheng, C. C. J. Heterocyclic Chem. 1973, 10, 85.

904

16. CARBON-CARBON BOND-FORMING REACTIONS

O

O HCONH2

O

OMe

POCl3

O

HCO2H , 6 h (NH 4)2SO 4 180 – 182°C

O

MeO

O

NHCHO

MeO

O N

MeO

OMe

OMe

206

207

208 (66%)

Bischler-Napieralski disconnections follow: R

R

N NH2

O

N R1

The Pictet-Spengler isoquinoline synthesis is another classical reaction.209,206c This variation generates an iminium salt from an amine and an aldehyde (a Schiff base), which cyclizes with an aromatic ring to complete the reaction in the presence of the acid.210 A synthetic example is taken from a synthesis of (+)-spegatrine by Cook and coworkers.211 The reaction of amino ester 209 with methyl 3-formylpropanoate gave a 93% yield of 210. CO2Et

MeO

CO2Et

MeO 1. H

CO2Me O

NHBn

N Bn

AcOH , CH2Cl2 , rt , 12 h 2. TFA , CH2Cl2 , 4 d

N H

N H Ph

209

210 (93%)

Modifying groups on the aromatic ring can control the regioselectivity of the reaction. A typical Pictet-Spengler reaction with formaldehyde converted 2-(3-methoxyphenyl)ethan-1-amine to 6-methoxy-2-methyl-1,2,3,4tetrahydroisoquinoline in 22% yield. The silyl directed reaction, however, converted 2-(3-methoxy-2-(trimethylsilyl) phenyl)ethan-1-amine to 8-methoxy-2-methyl-1,2,3,4-tetrahydroisoquinoline in 72% yield.212 Jacobsen and Taylor213 reported a Pictet-Spengler reaction, catalyzed by a chiral thiourea derivative that gave the product in good yield, with excellent enantioselectivity. A microwave-accelerated Pictet-Spengler reaction has also been used to prepare indole alkaloids.214 MeO NH2

1. HCHO , pH 6 2. HCO2H

OMe 2-(3-Methoxyphenyl)ethan-1-amine

N

6-Methoxy-2-methyl-1,2,3,4tetrahydroisoquinoline

NH2 1. HCHO , pH 6 SiMe3

Me

N

2. HCO2H

OMe 2-(3-Methoxy-2-(trimethylsilyl)phenyl)ethan-1-amine

(22%)

Me

OMe 8-Methoxy-2-methyl-1,2,3,4tetrahydroisoquinoline

(72%)

(a) Pictet, A.; Spengler, T. Berichte, 1911, 44, 2030; (b) Cox, E. D.; Cook, J. M. Chem. Rev. 1995, 95, 1797; (c) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-73; (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 508–509.

209

210

Ong, H. H.; May, E. L. J. Heterocyclic Chem. 1971, 8, 1007.

211

Edwankar, C. R.; Edwankar, R. V.; Namjoshi, O. A.; Liao, X.; Cook, J. M. J. Org. Chem. 2013, 78, 6471.

212

Miller, R. B.; Tsang, T. Tetrahedron Lett. 1988, 29, 6715.

213

Taylor, M. S.; Jacobsen, E. N. J. Am. Chem. Soc. 2004, 126, 10558.

214

Kuo, F.-M.; Tseng, M.-C.; Yen, Y.-H.; Chu, Y.-H. Tetrahedron 2004, 60, 12075.

905

16.5 FRIEDEL-CRAFTS REACTIONS: FORMATION OF HETEROATOM-CONTAINING DERIVATIVES

The Pictet-Spengler disconnection follows:

N H

+

NH2

R

RCHO

Another method for the synthesis of isoquinoline derivatives is the Pomeranz-Fritsch reaction.215 Simple aniline derivatives can be used, but the reaction is most useful for acid sensitive substrates (e.g., thiophene or pyrrole derivatives). In those cases, it is a good alternative to the Pictet-Spengler reaction. An example of an acid-sensitive substrate is 1-methyl-1H-pyrrole-2-carbaldehyde, which reacted with glycinal diethyl acetal (2,2-diethoxyethan-1-amine) to generate Schiff base 211 in 91% yield. Subsequent treatment with PPA/POCl3 gave a 21% yield of pyrrolo[1,2a]pyrazine (also known as 1-methyl-1H-pyrrolo[2,3-c]pyridine).216 When the reactive partners were 1-phenylethan-1-amine and glyoxal semiacetal, Schiff base 212 was formed. Treatment with acid led directly to 1-methylisoquinoline.217 The yield of 1-methylisoquinoline was only 15% by the Pomeranz-Fritsch method, using 2,2-diethoxyethan-1-amine and acetophenone, but 40% with this variation using 1-phenylethan-1-amine, known as the Schlittler-M€ uller modification.217 OEt H2NCH2CH(OEt)2

CHO

N

N

POCl3

OEt

N

PPA , 100°C

N

Me

N

Me

1-Methyl-1 H-pyrrole2-carbaldehyde

NH2

EtO

+

OEt

PhMe

CHO

N EtO

125–135°C

EtO

Me

Me 1-Methyl-1 H-pyrrolo[2,3-c]pyridine (21%)

211 (91%)

Me

1-Phenylethan-1-amine

N

Me 1-Methylisoquinoline (40%)

212 (72%)

Glyoxal semiacetal

H2SO 4 160°C, 2 min

The Pomeranz-Fritsch and Schlittler-M€ uller disconnections follow:

Ar

N

Ar

NH2

CHO

N R

R

16.5.3 Synthesis of Acridines, Carbazoles, and Phenanthridines Acridines are useful and medicinally important compounds. Acridines can be produced in a straightforward manner by a Friedel-Crafts reaction known as the Bernthsen acridine synthesis,218 which couples diarylamines (e.g., diphenylamine) with a carboxylic acid (e.g., benzoic acid) in the presence of a Lewis acid (zinc chloride is a typical reagent). Heating diphenylamine to 260 °C (10 h) with benzoic acid and zinc chloride gave a 48% yield of 9-phenylacridine.218b

(a) Pomeranz, C. Monatsh 1893, 14, 116; (b) Fritsch, P. Berichte 1893, 26, 419; (c) Gensler, W. J. Org. React. 1951, 6, 191; (d) Popp, F. D.; McEwen, W. E. Chem. Rev. 1958, 58, 321 (see p 328); (e) The Merck Index, 14th ed., Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-75; (f) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 526–527.

215

216

Herz, W.; Tocker, S. J. Am. Chem. Soc. 1955, 77, 6355.

217

Schlittler, E.; M€ uller, J. Helv. Chim. Acta 1948, 31, 914, 1119.

(a) Bernthsen, A. Annalen 1878, 192, 1; (b) Idem Ibid. 1884, 224, 1; (c) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-10. 218

906

16. CARBON-CARBON BOND-FORMING REACTIONS

The carbazole nucleus (the ‘parent’ carbazole is 9H-carbazole) is an important structural feature of many alkaloids. The Borsche-Drechel cyclization219 is a Friedel-Crafts route to these compounds, related to the Fischer indole synthesis discussed in Section 16.5.5. A typical reaction used phenylhydrazone derivatives of ketones (e.g., cyclohexanone phenylhydrazone), which was cyclized upon treatment with sulfuric acid to give 2,3,4,4a,9,9a-hexahydro-1H-carbazole in good yield. Aromatization with lead oxide (Pb3O4) gave 9H-carbazole in low yield. Glacial acetic acid gave cleaner products in the dehydrogenation step,220 but treatment with tetrachlorobenzoquinone (also known as chloranil) led to the best yield of 9H-carbazole 221 N

H PhCO2H, ZnCl2

N

260°C , 10 h

Ph Diphenylamine

9-Phenylacridine (48%) Cl

NNHPh

Cl

O

H2SO 4

O

Cl

N

Cl

N

(chloranil)

H 2,3,4,4a,9,9a-Hexahydro1 H-carbazole

Cyclohexanone phenylhydrazone

H 9H-Carbazole

Phenanthridines can be prepared by modification of the Friedel-Crafts reaction. The Pictet-Hubert reaction222 reacted an acyl o-aminobiphenyl (N-([1,1’-biphenyl]-2-yl)propionamide) with zinc chloride at 300 °C,223 or more commonly with POCl3 in refluxing nitrobenzene,224 to give the Friedel-Crafts cyclization product, a phenanthridine, 6-ethylphenanthridine. Zinc chloride required long heating and wasteful purification, and gave poor results with reactive substrates. The use of both POCl3 and a high-boiling solvent improved the yields of phenanthridine products. O

Et

H N

Et

N

POCl3, 1 h

N-([1,1'-biphenyl]-2-yl)propionamide

6-Ethylphenanthridine

The disconnections discussed in this section follow: N Ph2NH + RCO2H R

R N

O

+ PhNHNH2

N H H2 N + RCO2H

(a) Dreschsel, E. J. Prakt. Chem. 1888, 38, 69; (b) Borsche, W.; Feise, M. Berichte 1907, 40, 378; (c) Campbell, N.; Barclay, B. M. Chem. Rev. 1947, 40, 359 (see p 361); (d) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-13; (e) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 108–109.

219

220

Perkin, W. H., Jr.; Plant, S. G. P. J. Chem. Soc. 1921, 119, 1825.

221

Barclay, B. M.; Campbell, N. J. Chem. Soc. 1945, 530.

(a) Pictet, A.; Hubert, A. Berichte 1896, 29, 1182; (b) Morgan, C. T.; Walls, L. P. J. Chem. Soc. 1931, 2447; (c) Idem Ibid. 1932, 2225; (d) Eisch, J.; Gilman, H. Chem. Rev. 1957, 57, 525; (e) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-73; (f) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 506–507.

222

223

Theobald, R. S.; Schofield, K. Chem. Rev. 1950, 46, 171 (see p 175).

224

Morgan, G. T.; Walls, L. P. J. Chem. Soc. 1931, 2447.

16.5 FRIEDEL-CRAFTS REACTIONS: FORMATION OF HETEROATOM-CONTAINING DERIVATIVES

907

16.5.4 Synthesis of Oxygenated Aromatic Derivatives There are many phenol derivatives, which are characterized by an oxygen atom that is appended to a simple aromatic ring or to a polycyclic aromatic compound. An interesting route to phenol derivatives exploits the propensity of aryls to rearrange under Friedel-Crafts conditions. When a phenolic ester (e.g., phenyl acetate) is heated with a Lewis acid (e.g., AlCl3), a rearrangement occurs to generate a para-ketone, 1-(4-hydroxyphenyl)ethan-1-one, and an orthoketone, 1-(2-hydroxyphenyl)ethan-1-one, in what is known as the Fries rearrangement.225 The reaction generates an acylium ion (213), which fragments to an aluminum alkoxide (214) ion paired with the acylium ion MeCO+. Migration and Friedel-Crafts acylation generates both ortho and para-products. O

O

AlCl3 O

Me

AlCl3

O

O

Me

Me

AlCl3

OH

O

MeC O

+

Me

OH Phenyl acetate

213

214

1-(4-Hydroxyphenyl)ethan-1-one

1-(2-Hydroxyphenyl)ethan-1-one

This reaction is probably used most often to prepare methyl ketone derivatives from acetates,226 as in the conversion of 4-methoxy-2,3,5-trimethylphenyl acetate to 215 in 84% yield, in a synthesis of (2RS)-α-tocopherol by Schmalz and coworkers.227 The Fries rearrangement is not restricted to an acetate unit, however, and other acyl units have been transferred to an aromatic ring.228 In many cases, there is a regiochemical preference for the para-product due to steric effects that hinder reaction at an available ortho position.229 When the aromatic ring contains electron-withdrawing groups, the reaction does not work well. O MeO 2 TiCl4 ,CH 2Cl2

MeO

40°C

OAc

4-Methoxy-2,3,5-trimethylphenyl acetate

OH

215 (84%)

Anionic Fries rearrangements are known.230 A thia-Fries rearrangement of aryl sulfonate esters (e.g., phenyl p-toluenesulfonate) has been reported.231 Irradiation of this tosylate with microwaves on silica gel, for 10 min gave an 87% yield of o- and p-toluenesulfonyl phenol (80:20). There is also a photo-Fries rearrangement, which proceeds via radical intermediates formed by the initial photolysis (Section 15.2.2, for a brief overview of photochemical theory). The ketone is formed from the bis(acyl) derivative under photolytic conditions.232

(a) Fries, K.; Finck, G. Berichte 1908, 41, 4271; (b) Fries, K.; Pfaffendorf, W. Ibid. 1910, 43, 212; (c) Blatt, A. H. Org. React. 1942, 1, 342; (d) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-35; (e) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 260–261.

225

226 For an example using BF3 as a catalyst with hydroquinone diester, see Boyer, J. L.; Krum, J. E.; Myers, M. C.; Fazal, A. N.; Wigal, C. T. J. Org. Chem. 2000, 65, 4712. 227

Termath, A. O.; Velder, J.; Stemmler, R. T.; Netscher, T.; Bonrath, W.; Schmalz, H.-G. Eur. J. Org. Chem. 2014, 3337.

228

For an example see Cairns, N.; Harwood, L. M.; Astles, D. P. Tetrahedron Lett. 1988, 29, 1311.

229

Jefferson, A.; Wangchareontrakul, S. Aust. J. Chem. 1985, 38, 605.

230

See (a) Nicolaou, K. C.; Koide, K.; Bunnage, M. E. Chem. Eur. J. 1995, 1, 454; (b) Wright, N. E.; Snyder, S. A. Angew. Chem. Int. Ed. 2014, 53, 3409.

231

Moghaddam, F. M.; Dakamin, M. G. Tetrahedron Lett. 2000, 41, 3479.

232

Anderson, J. C.; Reese, C. B. Proc. Chem. Soc. 1960, 217.

908

16. CARBON-CARBON BOND-FORMING REACTIONS

The Fries rearrangement disconnection follow: OH O +

OH

R

OH

RCO2H

R O

Coumarin derivatives (see 4-methyl-2H-chromen-2-one) can be formed via Friedel-Crafts techniques. The Pechmann condensation233 is illustrated by esterification of phenol with a β-keto-ester (ethyl 3-oxobutanoate) by a transesterification reaction, which gave phenyl 3-oxobutanoate. Subsequent cyclization with AlCl3 generated a coumarin, 4-methyl-2H-chromen-2-one. Isolation of esters (e.g., phenyl 3-oxobutanoate is not always necessary and BrønstedLowry acids can be used. In a synthesis of ocimarin 1 by Yang et al.,234 resorcinol was condensed with ethyl 3-oxobutanoate in the presence of cerium iodide to give a 57% yield of the coumarin, ocimarin 1. The Pechmann condensation is facilitated by the presence of hydroxyl (OH), dimethylamino (NMe2), and alkyl groups (R) meta to the hydroxyl of the phenol.235,236 Pechmann condensation in an ionic liquid (1-butyl-3-methylimidazolium chloroaluminate, [bmim]Cl•2AlCl3]) using ethyl acetate has also been reported.237 O

O

Me

Me

CO2Et

AlCl3

OH

O

O

O Phenyl 3-oxobutanoate OH

O

4-Methyl-2H-chromen-2-one Me

O CO2Et

CeCl3•7 H2O , 90 °C

OH

HO

O

O

OH

Ocimarin I (57%)

Resorcinol

The Pechmann disconnection follows: R

O OH

O

+

O

R

CO2R1

The Bradsher reaction (see 194 in Section 16.4.5, where X ¼ O) can generate aromatic derivatives containing an oxygen atom in a polycyclic framework

16.5.5 Synthesis of Indoles This section has been devoted to variations of electrophilic aromatic substitution to make carbon-carbon bonds. There are other alkaloid-forming reactions that do not readily fit into this category, but since they involve acidcatalyzed processes are placed here. In 1883, Fischer reported a straightforward synthesis of indoles known as the (a) von Pechmann, H.; Duisberg, C. Berichte 1883, 16, 2119; (b) Sethna, S.; Shah, N. M. Chem. Rev. 1945, 36, 1 (see p 10); (c) Sethna, S.; Phadke, R. Org. React., 1953, 7, 1; (d) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-70; (e) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 490–491. 233

234

Yang, J. H.; Li, Y. F.; Ji, C. B.; Jiang, S. Z.; Liu, W. Y. Chinese Chem. Lett. 2010, 21, 1165.

235

Shah, M. M.; Shah, R. C. Berichte 1938, 71, 2075.

236

Miyano, M.; Dorn, C. R. J. Org. Chem. 1972, 37, 259.

237

In bmim chloroaluminate, 1-butyl-3-methylimidazolium chloroaluminate, [bmim]Cl•2AlCl3: Potdar, M. K.; Mohile, S. S.; Salunkhe, M. M. Tetrahedron Lett. 2001, 42, 9285.

909

16.5 FRIEDEL-CRAFTS REACTIONS: FORMATION OF HETEROATOM-CONTAINING DERIVATIVES

Fischer indole synthesis238 that is today considered a powerful reaction (Section 8.4.2). The indole nucleus (see 219) is a component of many natural products and medicinal compounds.239 Although many total syntheses of indole alkaloids begin with indole, some syntheses prepare the indole nucleus as part of the synthesis and the Fischer indole synthesis remains an important method for the preparation of indoles.240 A synthesis begins with conversion of an aldehyde or ketone to a phenylhydrazone (ArNHN]CR2), by treatment with phenylhydrazine. Subsequent treatment with a Lewis acid (ZnCl2 is a good choice) or a Brønsted-Lowry acid induces a reaction cascade that results in indole (219). The mechanism of this transformation involves formation of a hydrazone that reacts via the N-amino enamine form of the hydrazone (see 216), in what is probably an electrocyclic process analogous to a Claisen rearrangement (Section 15.5.5) to form 217. The imine nitrogen of 217 attacks an activated C]N. The NH reacts with the Lewis acid to form a complex with concomitant rearomatization of the benzene ring and formation of the five-membered ring in 218, that has a NHX group. When a Brønsted-Lowry acid is used, the ]NH moiety is converted to ]NH2 which gives dNH2 in 218, whereas reaction with ZnCl2 would give dNHZnCl2 in 218.241 Loss of an ammonium species by what is effectively an E1 process from 218 completes the synthesis of the indole ring in 219. An application of the Fischer indole reaction is taken from Smith III and coworker’s242 synthesis of (+)-scholarisine A. In this synthesis, ketone 220 was heated with N-benzylphenylhydrazine with pyridine•HCl to give 221 in >70% yield. R1

R1

R1

R1 R

PhNHNH2 ZnCl2

O

R1 NHX

R

R N N

N H

H

H

Bn

O O O

N H

218

H 219

O

N Ph

R

N

H 217

216

N

R

N

NH2 Py/HCl 110°C

Bz 220

O N

N Bn

Bz 221 (>70%)

The cyclization reaction that forms the five-membered ring of the indole proceeds with reasonable regioselectivity. The ketone moiety in 222 has two α-carbons, and two different enamine structures could be formed, as well as two bis(imine) intermediates.243 The final indole products were 223 and 224. As the concentration of acid catalyst was increased, the yield decreases, but the regioselectivity is about the same (2:1 223/224).243,244 The stereochemical course of the reaction also depends on the stereochemistry of the hydrazone substrate. In another example, hydrazone 225 (with a cis ring juncture) was treated with HCl•ether, in dichloromethane at room temperature to give a 61% yield of 226.245 Reaction of 225 with 85% sulfuric acid, however, gave a 42% yield of the isomeric 227 and the course of the reaction depended on the acid strength of the medium. When the trans derivative was treated with ethereal HCl under the same conditions, 38% of a single product was isolated.245 Treatment with 85% sulfuric acid gave no

(a) Fischer, E.; Jourdan, F. Berichte 1883, 16, 2241; (b) Fischer, E.; Hess, O. Ibid. 1884, 17, 559; (c) The Merck Index, 14th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2006; p ONR-32; (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: NJ, 2005; pp 244–245.

238

239

For a review of methods for indole ring synthesis, see Gribble, G. W. J. Chem. Soc. Perkin Trans. 2000, 1, 1045.

240

For a list of other named reactions that produce indoles, see Li, J.-J.; Corey, E. J. Name Reactions in Heterocyclic Chemistry, Wiley-Interscience: NJ, 2005; pp 99–158. 241

Robinson, B. Chem. Rev. 1969, 69, 227, and references cited therein.

242

Adams, G. L.; Carroll, P. J.; Smith, A. B., III. J. Am. Chem. Soc. 2012, 134, 4037.

243

Bryan Reed, G. W.; Cheng, P. T. W.; McLean, S. Can. J. Chem. 1982, 60, 419.

Rashidyan, L. G.; Asratyan, S. N.; Karagezyan, K. S.; Mkrtchyan, A. R.; Sedrakyan, R. O.; Tatevosyan, G. T. Arm. Khim. Zh. 1968, 21, 793 [Chem. Abstr. 71: 21972z, 1969].

244

245

Freter, K.; Fuchs, V.; Pitner, T. P. J. Org. Chem. 1983, 48, 4593.

910

16. CARBON-CARBON BOND-FORMING REACTIONS

cyclization, and only unreacted starting material. Treatment with more concentrated acid led to decomposition of the starting material but no cyclization. O

Me

O

O

N PhNHNH2

O

N Me

+

80°C , 6 h

O

N O

H

222

N H

Bz

rt

N H

N

227 (42%)

Bz HCl•Ether

N

Me

H

N

225

Bz

H

CH2Cl2 , rt 2h

H

NHPh

Me

224

N

85% H2SO 4

H

O

H 223

H

Me

N Me N

226 (61%)

The initially formed imine product may be subject to other reactions, including rearrangement under the strongly acidic reaction conditions. The reaction of phenylhydrazone and cyclopentane carbaldehyde gave the corresponding phenylhydrazone, and subsequent reaction with acetic acid gave a 90% yield of 2,3,4,9-tetrahydro-1H-carbazole via the acid-catalyzed rearrangement of imine 228.246 Cyclopentane carbaldehyde derivatives gave only fused ring indoles, but larger ring carbaldehydes led to spiro derivatives or mixtures of fused-ring and spirocyclic products.246 N

CHO PhNHNH2

NHPh

H+

N 228

N N

N

H

H

N

H

H 2,3,4,9-Tetrahydro-1H-carbazole (90%)

The Fisher indole disconnection follows: R1 N H

R

R1

R

+ PhHN NH2

O

A related route to indolones is the Stolle synthesis,247 where a secondary aryl amine (e.g., diphenylamine) reacts with α-chloro acid chloride or α-bromo acid bromide, to yield α-bromoamide (e.g., 229) or α-chloroamide. Oxalyl chloride can be used, but a more typical example is the reaction of 2-bromopropanoyl bromide with 4-ethoxy-Nmethylaniline to yield 229. Friedel-Crafts cyclization with AlCl3 led to the indolone, 5-hydroxy-1,3dimethylindolin-2-one.248 246

Rodríquez, J. G.; Benito, Y.; Temprano, E. J. Heterocyclic Chem. 1985, 22, 1207.

(a) Stolle, R. Berichte 1913 ,46, 3915; (b) Idem Ibid. 1914, 47, 2120; (c) Idem J. Prakt. Chem. 1923, 105, 137; (d) Stolle, R.; Bergdoll, R.; Luther, M.; Auerhahn, A.; Wacker, W. Ibid. 1930, 128, 1; (e) Sumpter, W.C. Chem. Rev. 1945, 37, 443 (see p 446).

247

248

Julian, P. L.; Pikl, J. J. Am. Chem. Soc. 1935, 57, 563.

911

16.5 FRIEDEL-CRAFTS REACTIONS: FORMATION OF HETEROATOM-CONTAINING DERIVATIVES Br

EtO

Br

Me

EtO

HO

Me

Br

Me AlCl3 , Neat

O

PhH , Water bath

N

NHMe

185°C

O

4-Ethoxy-N-methylaniline

O

N

Me

Me 5-Hydroxy-1,3-dimethylindolin-2-one

229

The Stolle disconnection follows: R1

R1 R R

N H

+

PhHN NH2

O

There are, of course, other methods to prepare the indole nucleus.249,240 A few methods are presented here to illustrate some of the new methodology. In one approach, Barluenga et al. 250 treated 2-bromo-N-(2-bromoallyl)-Nmethylaniline with tert-butyllithium to yield 230. Cyclization and elimination gave lithiated indole, ((1-methyl-1Hindol-3-yl)methyl)lithium, which reacted with electrophiles (e.g., 4-chlorobenzaldehyde) to give 231 (71% yield). Br

Li

Br

Li

Li

4 equiv t-BuLi

CuCN

–78°C

or TMEDA

N

N

Me

N

Me

2-bromo-N-(2-bromoallyl)N-methylaniline

Me ((1-Methyl-1H-indol-3-yl)methyl)lithium

230 OH Cl

CHO

N

Cl

Me 231 (71%)

NH2 Ph

CO2Et

N2 EtO2C

O

P(O)(OEt)2

1% Rh2(OAc) 4 Toluene

2-Aminobenzophenone

HN P

O O

OEt OEt

Ph 232 (86%)

DBU

H N CO2Et

Toluene

Ph Ethyl 3-phenyl-1H-indole2-carboxylate

Another approach used α-diazophosphonates. O’Shea and Coleman251 reported an organolithium addition strategy. Nakamura and Ukita252 treated 2-aminobenzophenone with triethyl diazophosphonoacetate, in the presence of the rhodium acetate catalyst (see Section 17.9.5), and obtained an 86% yield of 232. Subsequent treatment with DBU (Sections 2.2.2 and 3.5.1) led to a 99% yield of indole ethyl 3-phenyl-1H-indole-2-carboxylate. Hiroya et al.253a reported a method for the synthesis of indole-2-carboxylates, based on a palladium-catalyzed coupling strategy (see Section 18.4), in a synthesis of duocarmycin SA. Coupling methyl propiolate and 249

For a review, see Humphrey, G. R.; Kuethe, J. T. Chem. Rev. 2006, 106, 2875.

250

Barluenga, J.; Sanz, R.; Granados, A.; Fañanás, F. J. J. Am. Chem. Soc. 1998, 120, 4865.

251

Coleman, C. M.; O’Shea, D. F. J. Am. Chem. Soc. 2003, 125, 4054.

252

Nakamura, Y.; Ukita, T. Org. Lett. 2002, 4, 2317.

(a) Hiroya, K.; Matsumoto, S.; Sakamoto, T. Org. Lett. 2004, 6, 2953; (b) Negishi, E.; Okukado, N.; King, A. O.; Van Horn, D. E.; Spiegel, B. I. J. Am. Chem. Soc. 1978, 100, 2254.

253

912

16. CARBON-CARBON BOND-FORMING REACTIONS

N-(4-cyano-2-iodophenyl)methanesulfonamide under what is called Negishi coupling253b conditions led to a 69% yield of the indole, methyl 5-cyano-1-(methylsulfonyl)-1H-indole-2-carboxylate. NC

NC

2 equiv HC C–CO2Me , THF , Reflux 3% Pd(PPh3)4 , 6 equiv i-Pr2NEt

I

CO2Me N

3 equiv ZnBr2 , 17 h

NHMs

Ms Methyl 5-cyano-1-(methylsulfonyl)1 H-indole-2-carboxylate

N-(4-Cyano-2-iodophenyl)methanesulfonamide

(69%)

16.6 CONCLUSION This chapter discussed chemical equivalents for Ca, primarily involving carbocation intermediates. Although carbocations are subject to racemization and rearrangement, they can be stabilized so they can react with carbon nucleophiles ranging from alkenes and alkynes, to aromatic compounds, to Grignard reagents, or to enolate anions. Carbocations can be used synthetically for intramolecular coupling and/or if a tertiary or oxygen-stabilized cation is generated. Friedel-Crafts reactions are among the more powerful and useful of all organic reactions for constructing carbon-carbon bonds when aromatic nuclei are involved. In Chapter 17, methods for generating carbon-carbon bonds via radical intermediates and carbenes will be discussed, complimenting the collection of chapters for generating carbon-carbon bonds in organic synthesis. HOMEWORK

1. Friedel-Crafts alkylation is not a useful synthetic method for the synthesis of primary arenes (e.g., 1-phenylhexane). Explain why not. 2. Briefly explain the regioselectivity for this reaction. O

AcCl , AlCl3 PhNO2 , 100 °C

HOOC NH2

HOOC NH2

HO

HO

3. Provide a mechanistic rationale for the following transformation: OMe

OH

OMe

TsOH , aq Acetone

H MeO

OMe

+

H MeO

4. Offer a mechanistic rationale for the following transformation: Me

OH CH2 Me

Br

Me

5. Give a mechanistic rationale for the following transformation: Cl MeO

Me

NBS , CH2Cl2

H+

MeO

O

913

16.6 CONCLUSION

6. Give a mechanistic rationale for this transformation:

HO O

O

Me3Al , CH2Cl2

OH

MeO

OSi(i-Pr)3

OMe

OSi(i-Pr)3

7. Give the complete mechanism for the following transformation: MeO

MeO dil HCl

HO

H N Me

O

MeHN

MeO

OH Thebaine

Thebenine

8. What are the mechanistic implications of the following experiment? Of what possible value is this experiment to synthesis if the RdNH2 ! RdOH transformation is planned? NH2 H3C

OH HONO

14

H3C

CH3

Ph

CH3

Ph

OH

14

14

+

H3C

CH3 Ph

(1 : 1)

9. Give the complete mechanism for the following transformation: BnO

O

N

HCOOH , 0 °C

OBn

Toluene/THF , 2 h

NHBoc

HCO2

C6H13

Boc H

C6H13

10. Explain the fact that Wolff-Kishner reduction of A gave B but Clemmensen reduction of A gave C. O

Me Clemmenson

Me

N

Me

Wolff–Kishner

N

N

C

B

A

11. Give a reasonable mechanism for the following transformation: OTs

OTs Ph O

BF3•OEt2 , CH2Cl2

Ph

0 °C

O

12. Provide a mechanism for the following transformation: H Br N

Br HCOOH , Toluene Reflux

HO

(65%)

H N

O

914

16. CARBON-CARBON BOND-FORMING REACTIONS

13. For each of the following, give the major product. Show the correct stereochemistry where appropriate. OMOM

O

HO

(A)

OH

O

Ph

N

TFA , CH2Cl2 –10°C

O

CSA , CH2Cl2

O

(B)

O SiMe3 AcO +

(C)

O

CO2Et

BF3•OEt2 , –78 °C CH2Cl2 , 0.1 M

Me OTBS

(D)

MeO

2. 1% TFA , CH2Cl2

NHBn N Me

OMOM

OMs

OMe HO2C

MeO

1. EtO2CH2CH2CHO

NHAc 60% H2SO 4

(E)

Acetone

(F)

MeSO3H CH2Cl2

OMe MeO

Br AlBr3

OAc

(H)

(G)

AlCl3 , PhNO2 60°C , 4 h

H Me H H

1. AlCl3 , o-Xylene 2. LiAlH4 , Ether

(I)

(J)

3. H2SO 4 , Cyclohexane

Me

Ac2O , DMAP Py , CH2Cl2

O OH

OH

TFA

N

1. MeO

(N)

NaNO2 , 0.25 M H2SO 4

NH2

H

O OMe N

(O)

OMe H

NHNH2•HCl

Na2CO3 , EtOH Reflux 2. AcOH , 95°C

Et

O MeO2C

O

O

O

(M)

O

(L)

OH

MeO2CHN

OH

O

OH

(K) MeO

Ether , 0°C

H BzO

BF3•OEt2

O

H

O O

Me

1. NaBH4 , EtOH 2. Camphorsulfonic acid

OH

(P)

2. aq NaOH; aq HCl

O

Me

(Q) MeO

1. DMF , POCl3

Toluene , 80°C

NH2

1. NaNO2 , aq HCl , 0°C , KOH Ethyl -ethylacetoacetate

OH CO2H

(R) NH2

2. EtOH , HCl , heat

O Et PPA , 90°C

N

(S) N H

(T)

EtO2C

N H Ph N

EtO2C CO2H

HCOOH Ac2O , HCl

O

Me

1. ZnBr2 2. NEt3

915

16.6 CONCLUSION

14. In each case, give a complete synthesis of the target from the designated starting material. Show all intermediate products and give all reagents. H

CO2Me

(A)

N O

N

N

H

H

O CH2Ph

NH2

(B) N

N H

Me

H OH

H

(C) O

H

H

(D)

O NH2

O

CO2t-Bu

H O

CHO

O

N

OMe

(E)

H

MeO

(F)

OH

Ar

OMe OMOM

OMe

Ar = Me

MeO

Me N

Ar OMe Me

C H A P T E R

17 Formation of Carbon-Carbon Bonds Via Radicals and Carbenes 17.1 INTRODUCTION A carbanion can be viewed as a tetrahedral species containing a pair of electrons in an orbital (1). A carbocation (carbenium ion), described in Chapter 16 as sp2 hybridized, is essentially a trigonal-planar carbon with an empty p-orbital with attached hydrogen atoms and/or alkyl groups (2). A carbon radical can be viewed as a trivalent species containing a single electron in a p-orbital. A radical, which contains one electron in an orbital, can be tetrahedral, planar, or “in between” with properties of both a carbanion and a carbocation. As shown in 3, a reasonable “in between” structure is a flattened tetrahedron. In terms of its reactivity, radical 3 could be considered electron rich or electron poor. In most of its reactions, the electron-deficient characterization is useful for predicting products.

R R

C R

C

R

R C R

R

R 1

R

3

2

Gomberg1 was the first to characterize a free radical when, in 1900, he generated triphenylmethyl radical 4 by reacting chlorotriphenylmethane with Zn metal. Triphenylmethyl radical is unusual in that it is quite stable, and its formation is probably the first experimental verification of a free radical. Frankland,2 however, may have been the first to generate transient methyl and ethyl radicals in the reaction of iodomethane and iodoethane with Zn in 1849. A good deal of attention is now focused on the reactivity and applications of radicals to organic synthesis. Excellent monographs by Davies and Parrott,3 Lazár et al.,4 Hay,5 Giese,6 Kochi,7 Togo,8 Zard,9 Renaud and Sibi,10 and by Pryor11 describe radical

1

(a) Gomberg, M. J. Am. Chem. Soc. 1900, 22, 757. (b) Idem Chem. Ber. 1900, 33, 3150.

2

Frankland, E. Annalen 1849, 71, 171, 213.

3

Davies, D. I.; Parrott, M. J. Free Radicals in Organic Synthesis; Springer-Verlag: Berlin, 1978.

4

Lazár, M.; Rychl
y, J.; Klimo, V.; Pelikán, P.; Valko, L. Free Radicals in Chemistry and Biology; CRC Press: Boca Raton, FL, 1989.

5

Hay, J. M. Reactive Free Radicals; Academic Press: London, 1974.

6

Giese, B. Radicals in Organic Synthesis: Formation of Carbon-Carbon Bonds; Pergamon Press: Oxford, 1986.

7

Kochi, J. K., Ed. Free Radicals; John Wiley: New York, 1973; Vol. 1 and 2.

8

Togo, H. Advanced Free Radical Reactions for Organic Synthesis; Elsevier: Amsterdam, 2004.

9

Zard, S. Z. Radical Reactions in Organic Synthesis; Oxford University Press: New York, 2003.

10

Renaud, P.; Sibi, M. P., Eds., Radicals in Organic Synthesis, Volume 1. Basic Principles, Volume 2. Applications; Wiley-VCH: Weinheim, 2001.

11

Pryor, W. A. Frontiers of Free Radical Chemistry; Elsevier: New York, 2012.

Organic Synthesis http://dx.doi.org/10.1016/B978-0-12-800720-4.00017-9

917

Copyright © 2017 Michael Smith. Published by Elsevier Inc. All rights reserved.

918

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

chemistry in great detail, as well as the many applications to synthesis. A review by Curran12 placed an emphasis on synthetic applications, and books by Togo13 and by Zard14 emphasize synthesis. Ph Ph

2

Ph

Zn

C

Cl

Ph

+

C

ZnCl2

Ph

Ph Chlorotriphenylmethane

4

17.2 STRUCTURE OF RADICALS As mentioned in the introduction, there are several possible structures for a radical. The two most common are pyramidal (see 3) and planar (see 7). Formation of pyramidal radical (3) “involves a transition state that differs from the original molecules only in the length of the central bond,” represented by 5.15 Dissociation to a planar radical (7) is also possible, but requires transition state 6 where both fragments are flattened.5.15 Transition state 6 is generally favored over 5, where 5 represents an “instantaneous bond dissociation.” R

R C

R

C

C

R

R

R

R

R

R

R

R

C

2 R

R

C

R R

5 R

R C

R R

R

R

R

C R

3

R

C

C

R

R

R 2 R

R R

C

7

6

The lower energy dissociation of the molecule, in the transition state prior to cleavage, is graphically presented in Fig. 17.1.16 For path 1b, the CdC bond is changing from Csp3dCsp3 to CπdCπ, which is facile. Formation of the CdR bond is bond strengthening, and energy is released as the radical changes from a pyramidal structure to the planar arrangement. Path 1a is the instantaneous bond dissociation, and does not possess as much stabilizing energy. Note that if flattening in 6 is hindered by large steric interactions, then Eact increases and in those cases 3 may be the intermediate. R C R R

1a

R

1b

R R

Reaction coordinate

FIG. 17.1

C

Energy diagram for dissociation to planar and pyramidal radicals.

12

Curran, D. P. Synthesis 1988, 417, 489.

13

Togo, H. Advanced Free Radical Reactions for Organic Synthesis; Elsevier: New York, 2004.

14

Zard, S. Z. Radical Reactions in Organic Synthesis; Oxford University Press: New York, 2003.

15

Reference 5, p 42.

16

Reference 5, p 43.

919

17.3 FORMATION OF RADICALS BY THERMOLYSIS

FIG. 17.2 Walsh diagram for inversion of the pyramidal methyl radical. H–H interaction Total energy X

X

E

kcal mol–1

C–H bond strength

Electron energy

H

C

H

H C

H H

H

C

H

H H

The methyl radical (H3C•) can be used to begin a discussion of the properties and structure of free radicals. Structure 3 (R ¼ H) represents the ground state of the radical, experimentally determined to be planar, or nearly so.17 The inversion energy for the methyl radical has been estimated to be only 4.2 kcal (17.6 kJ) mol1.18 As the temperature increases, the pyramidal model (7, R ¼ H) explains the 1H and 13C NMR, as well as the electron spin resonance spectra.19 Three factors control the configuration of a radical: “(a) the energy of the unpaired electrons, (b) the energies of the bond in the radical, and (c) the HdH interaction,20 which is illustrated by a Walsh diagram,21 shown in Fig. 17.2 for the methyl radical.”20 The total energy for inversion supports a double barrier for inversion between the two pyramidal forms, proceeding through the planar form (see the total energy curve in Fig. 17.2). The energy of the most stable configuration of the methyl radical was calculated to be 5 kcal (20.9 kJ) mol1,17 and the energy of the pyramidal methyl was calculated to be 10 kcal (41.9 kJ) mol1 higher than the ground state.20,22

17.3 FORMATION OF RADICALS BY THERMOLYSIS Radicals can be formed in several ways. Many involve dissociative homolytic cleavage (one electron is transferred to each adjacent atom from the bond) as a key step, as depicted by the fragmentation of XdY to yield two radical products. Another major route to radical intermediates involves the reaction of a radical (X•) and a neutral molecule (XdY), producing a new radical (Y•) and a new neutral molecule (XdX). This latter pathway will be discussed under radical reactions in Section 17.5. The equilibrium constant for these processes depends on both the relative bond strength of XdY and also on the relative stabilities of X• and Y•. For homolytic cleavage, raising the temperature of the reaction will generally shift the equilibrium toward a higher concentration of free radicals.23 This equilibrium makes it convenient to correlate homolytic cleavage with bond dissociation energy. 17

(a) Fessenden, R. W.; Schuler, R. H. J. Chem. Phys. 1963, 39, 2147. (b) Herzberg, G. Proc. R. Soc. London A. 1961, 262, 291.

18

Andrews, L.; Pimentel, G. C. J. Chem. Phys. 1967, 47, 3637.

19

Chang, S. Y.; Davidson, E. R.; Vincow, G. J. Chem. Phys. 1970, 52, 5596.

20

Reference 5, p 45.

21

Walsh, A. D. J. Chem. Soc. 1953, 2288, 2306, 2325.

22

Kibby, C. L.; Weston, R. E., Jr. J. Am. Chem. Soc. 1968, 90, 1084.

23

Reference 4, p 8.

920

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

X X

X

+

X

Y

X

Y

Y

+

X

Y

Lazár et al.24 published a table of bond energies23,25 that correspond to the instantaneous bond dissociation energies (D). In principle, lower bond dissociation energies should correlate with an increased propensity for homolytic cleavage. Dissociation of CdC [85 kcal (357 kJ) mol1],24 CdO [84 kcal (351 kJ) mol1],21 HdO [111 kcal (464 kJ) mol1],24 and Cdhalogen bonds [55 kcal (230 kJ) mol1 for CdI to 79.8 kcal (351 kJ) mol1 for CdCl]24 all require high reaction temperatures for homolytic cleavage. Alkali metals (KdK, NadNa, LidLi) with bond dissociation energies of 55 kcal (229.9 kJ) mol1, 69 kcal (188.4 kJ) mol1, and 103 kcal (430.5 kJ) mol1, respectively,24 undergo homolytic cleavage under relatively mild reaction conditions. The potent electron-transfer capabilities of these metals for dissolving metal reductions were discussed in Section 7.11.1. Particularly weak bonds that are expected to yield homolytic cleavage are the peroxide bond [OdO, 140 kcal (1801 kJ) mol1]24 and those of the dihalogens, which is reflected in the chemical methods used to generate organic radicals. The methods usually begin with thermolysis of peroxides (H2O2, ROOR, etc.), diazo compounds (RN2), or dihalogens. HO  OH ƒƒƒ! 2  OH RO  OH ƒƒƒ!  OR +  OH RO  OR ƒƒƒ! 2  OR Peroxides are potentially dangerous compounds because many of them can detonate upon exposure to heat or light, or under conditions where there is friction or shock. Peroxides are known to oxidize alkenes to oxiranes (especially in the presence of transition metals, see Section 6.4.2), but if heated they decompose to hydroxy radicals (HO•) or alkoxy radicals (RO•). Lazár et al.26 published a table of approximate decomposition temperatures for several common peroxides. Hydrogen peroxide requires the rather high temperature of 380°C for a decomposition half-life of 1 h,26 and it is not very soluble in organic solvents. Peroxide 8, however, decomposes at such a low temperature (16°C)26 that controlling the reaction is difficult in most instances. Benzoyl peroxide cleaves at 95°C,26 a temperature that is compatible with many organic reactions, allowing some control of the reaction, and is reasonably soluble in organic solvents. Di-tert-butyl peroxide decomposes at 150°C26 to produce Me3CdO•. Although this peroxide is a useful radical initiator, the higher temperature required for decomposition occasionally causes problems. Me Ph Me

O

O

O

Me Ph Me

Ph

Me Me Me

2

OH

tert-Butyl hydroperoxide

Me Me

Me Me

O

Me

Me O

Ph

Me

Me Me Me

Me O O Me Di-tert-butyl peroxide

2

O

O Benzoyl peroxide

8 Me

O

O

+

Me Me

O

O

+

H2O

9

tert-Butyl hydroperoxide is another useful peroxide that decomposes to yield Me3CdO• (9) and H2O. The presence of an α-hydrogen atom, as in diisopropyl peroxide, increases the facility of additional decomposition reactions. Initial

24

Reference 4, p 9.

25

(a) Sanderson, R. T. Chemical Bonds and Bond Energies; Academic Press: New York, 1976. (b) Idem J. Am. Chem. Soc. 1983, 105, 2259.

26

Reference 4, p. 12.

921

17.3 FORMATION OF RADICALS BY THERMOLYSIS

homolytic cleavage gives the expected Me2HCdO•, but it reacts with diisopropyl peroxide to remove the α-hydrogen to yield 10 and propan-2-ol. The carbon radical (10) fragments to acetone and Me2HCdO•.27 Me

Me H Me

Me

Me H Me

H O O Me Diisopropyl peroxide

2 H Me

O

O

Me H Me

O

OH

Me

Me

Me O

O

Me

+ Me

H Me

C

Me

O

+

O

Me

Me H Me

10

Acyl peroxides, (e.g., the commonly used benzoyl peroxide), react to yield phenacyl radicals (see 11) that decarboxylate under the reaction conditions to yield an aryl radical (e.g., 12). Decomposition of alkyl radical from an alkyl acyl precursor [(RCO2)2] leads to an alkyl radical. Decarboxylation of primary alkyl acyl peroxides (e.g., acetyl peroxide) is very facile, leading to 13 and decarboxylation generates the methyl radical •CH3.27 Decarboxylation of an aryl acyl radical is slower. If a peroxide decomposes, it has been determined that the oxygen radical remains in a “cage” for 1011 s before diffusing away. It can recombine (dimerize) or react further (see Section 17.5). O Ph

O

Ph O

2 11

O O

O

– CO2

O

O Benzoyl peroxide H3C

O

2

Ph

12

H3C

O

2

CH3

– CO2

CH3

2

O

O Acetyl peroxide

13

Azo compounds, characterized by a dN]Nd bond, are free radical precursors that liberate stable nitrogen gas (NN) upon decomposition. One of the most used azo compounds is (azobisisobutyronitrile) (AIBN), which decomposes much faster than benzoyl peroxide to yield nitrogen and the cyano-stabilized radical 14.28 Just as with carbanion and carbocation intermediates, the presence of the conjugating π-bond of the cyano unit will delocalize the single electron of the radical via resonance, increasing the stability of that intermediate. Me Me

C

Me N

N

CN

C Me CN

Me

Heat or hn – N2

2

AIBN

Me

Me

C

Me

C

C

N

C N 14

Homolytic dissociation of symmetrical azo compounds may be stepwise,29 as in the decomposition of azobistriphenylmethane to 15. Loss of nitrogen gas yields the triphenylmethyl radical (16).30 The decomposition temperature for diazo compounds is dependent on the groups attached to the α-carbon.31 Homolytic cleavage of AIBN occurs at 67°C,27 whereas MeN]NMe decomposes at 275°C, and azobis(2,2-dimethylethane) (Me3CN ¼ NCMe3) decomposes at 172°C.27 The stability imparted by the electron-withdrawing cyano group leads to greater stability of the radical and more facile decomposition.

27

Reference 4, p 13.

28

(a) Yoshino, K.; Ohkatsu, J.; Tsuruta, T. Polym. J. 1977, 9, 275. (b) Hinz, J.; Oberlinner, A.; R€ uchardt, C. Tetrahedron Lett. 1973, 1975.

29

Dannenberg, J. J.; Rocklin, D. J. Org. Chem. 1982, 47, 4529.

30

Newman, R. C., Jr.; Lockyer, G. D., Jr. J. Am. Chem Soc. 1983, 105, 3982.

31

Reference 4, p 15.

922

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

Ph Ph

C

Ph

Ph N

N

Ph

C

Ph

– Ph3C•

Ph

C

Ph N

– N2

N

Ph

Ph

Ph

Azobis(triphenylmethane)

C Ph

15

16

17.4 PHOTOCHEMICAL FORMATION OF RADICALS The general principles of photochemistry were introduced in Section 15.2.2. A molecule can absorb a photon of light and the resultant photoactivated species can, in the context of this chapter, undergo homolytic cleavage to form radicals. The energy required for bond cleavage can be provided by various wavelengths of light. An energy of 10 kcal (42 kJ) mol1 corresponds to 3500 cm1, which is in the IR region of the electromagnetic spectrum. Light of wavelength 2000 Å corresponds to 143 kcal (598 kJ) mol1, and is at the lower end of the UV region. Light in the UV or visible region can excite molecules to higher electronic states. Wavelengths in the UV are usually expressed in nanometers (1 nm ¼ 1 mμ ¼ 1  107 cm) and a wavelength of 200–400 nm is used most often for inducing photolysis in the UV region32 for common functional groups (Section 15.2.2). If a molecule were irradiated with a light source generating several intense wavelengths of light, it is possible for several different functional groups to absorb the light and react, including the product that is the target of a synthesis. A convenient method for limiting the wavelength of light (and thereby the energy of irradiation) is to filter out higher energy wavelengths (and lower energy wavelengths if necessary). A filter is typically imposed between the light source and the reactants. Sometimes the filter is simply the reaction vessel, made of Pyrex or quartz glass,33,34 but more often the filter material is attached to the light source. When a molecule absorbs a photon of light, an electron is promoted from the bonding to an antibonding molecular orbital (n ! σ*, n ! π*, or π ! π*) to generate the singly occupied molecular orbital (a SOMO) pertinent to radical reactions (Sections 2.4.2 and 15.2.2).35 If the excited state is expressed as AB*, there are several pathways available for dissipation of this energy, as shown by Wayne (see Fig. 17.3).36 The pathway that is pertinent to formation of radicals is dissociation of AB* to A• and B•: hv

AB

[AB]*

A•

+

FIG. 17.3 Routes available for loss of elec-

AB (Physical quenching)

AB + h (Luminescence)

AB+ (Intramolecular energy transfer) [Radiationless transition]

A• + B• (Dissociation) AB* +E + CD

AB + CD2+ (Intermolecular energy transfer)

tronic excitation.

AB+ + e– (Ionization)

+M

B•

BA (Isomerization)

AE + B or ABE (Direct reaction) AB•+ + E•– or AB•– + (Charge transfer)

E•+

32

Johnson, E. In The Physiology and Pathophysiology of the Skin, Vol. 8 of The Photobiology of the Skin: Lasers and the Skin; Jarrett, A., Ed., Academic Press: New York, 1984; p 2401. 33

Murov, S. L. Handbook of Photochemistry, Marcel-Dekker, New York, 1973; pp 107–108.

34

Calvert, J. G.; Pitts, J. N., Jr. Photochemistry; Wiley: New York, 1966; pp 474, 490–491.

35

DePuy, C. H.; Chapman, O. L. Molecular Reactions and Photochemistry; Prentice Hall: Englewood Cliffs, NJ, 1972; p 6.

36

Wayne, R. P. Principles and Applications of Photochemistry; Oxford University Press: Oxford, 1988; p 6.

923

17.4 PHOTOCHEMICAL FORMATION OF RADICALS

Photochemical homolysis of the carbon-carbon bond of alkanes is difficult. The carbon-carbon bond of ethane has a bond strength of 83 (347.4 kJ) mol1, and it fragments to yield methyl radicals (•CH3) only at temperatures approaching 600°C. Absorption of a photon of light is also difficult since alkanes absorb only in the “vacuum” UV [methane absorbs at 144 nm ¼ 198.6 kcal (831.4 kJ) mol1].37 Loss of molecular hydrogen (H2) usually accompanies conversion of an alkane (R2CH2) to the alkyl radical (17).38 If a long alkyl chain is present, loss of H• can occur at many points, leading to a large number of different radicals. H H

H

C

C

H

H

H 600°C

H

H

H R

C

H

C

C

H

H

H

H hv

H

R

C

R

+

1 2 H2

R 17

Alkenes show a π ! π* absorption for the π-bond at  180 nm [158.9 kcal (665.1 kJ) mol1].38 Conjugated alkenes show a shift in absorption toward the visible spectrum (lower energy). Isomerization (of ethene to ethyne) and fragmentation are observed.38 The yield of alkenyl radical decreases as the size of the alkyl portion of 1-alkenes increases.39 Photolysis of a terminal alkene generates radical 18 in a typical photolysis reaction. R

hn

R 18

O H3C

C

CH3

O

O

hn

H3C

C

CH3

+

CH3

C

CH3

+

CLO

CH3

19

Carbonyl compounds show a symmetry forbidden n ! π* transition at 290 nm [98.6 kcal (412.8 kJ) mol1] for aliphatic aldehydes, and at 280 nm [102.1 kcal (427.6 kJ) mol1] for ketones. Conjugation to an aromatic ring (as in benzophenone) shifts the absorption to longer wavelength [340 nm, 84.1 kcal (2.1 kJ) mol1]. Acid derivatives absorb at a shorter wavelength [76%)

17.5.4 Reduction by Atom Transfer When a radical is generated in the presence of a hydrogen-transfer agent, it is possible to reduce various functional groups. When a hydrogen atom is transferred in this manner, the reaction is formally a reduction. Since an X group is replaced with H, this is a form of the substitution reaction described in Section 17.5.3, but it is separated into a different section because of the synthetic importance. The best example may be the one presented in Section 17.5.4 in which an alkyl halide is reduced by treatment with tributyltin hydride, illustrated here by the reduction of the iodide moiety 3-iodo-2,5-diphenyl-1-tosylpyrrolidine in the presence of AIBN and Bu3SnH to give a 98% yield of 2,5-diphenyl-1tosylpyrrolidine, in an example taken from the Davis et al.74 synthesis of pyrrolidine ()-197B. Azobisisobutyronitrile is a radical initiator, and reaction leads to loss of iodide to yield the radical. The intermediate radical reacted with Bu3SnH via hydrogen transfer to yield 2,5-diphenyl-1-tosylpyrrolidine and Bu3Sn•, which generated Bu3SnI. This reaction is a very effective method for the controlled reduction of halides. I Bu3SnH, AIBN

Ph

N

Ph

Ph

N

Ph

PhMe, 80°C

Ts

Ts 3-Iodo-2,5-diphenyl1-tosylpyrrolidine

2,5-Diphenyl-1tosylpyrrolidine

(98%)

A useful radical-based reaction has been developed that can be applied to alcohols. As seen in Section 7.11.7, conversion of an alcohol to a thionocarbonate followed by treatment with tributyltin hydride under radical conditions gives cleavage to the CdO bond to give the reduction product. This transformation is called Barton deoxygenation (the Barton-McCombie reaction).75 In a synthesis of plakortone L, Sugimura et al.76 treated 52 with PhOC(S)Cl and DMAP to give xanthate 53 in 84% yield. Subsequent reduction with tributyltin hydride and AIBN gave a 97% yield of 54. S HO

OPh

OTBDPS

H O

O

DMAP, MeCN, rt, 3 h

O

O

S

Me

O H

Cl

OPh

O O

PhMe, Reflux, 3 h

O Me

O

OTBDPS

H OTBDPS Bu SnH, AIBN 3

H

Me

O H

H 52

53 (84%)

54 (97%)

17.5.5 Fragmentation A radical can fragment to form a new radical and a neutral molecule. Typically, the neutral molecule is an alkene, making the overall process an elimination.77 Addition of trichloromethyl radical (Cl3C•), generated by thermolysis of 74

Davis, F. A.; Song, M.; Augustine, A. J. Org. Chem. 2006, 71, 2779.

(a) Barton, D. H. R.; McCombie, S. W. J. Chem. Soc. Perkin Trans. 1975, 1, 1574. (b) Lopez, R. M.; Hays, D. S.; Fu, G. C. J. Am. Chem. Soc. 1997, 119, 6949. Phenyl thionocarbonates were first used by (c) Robins, M. J.; Wilson, J. S.; Hansske, F. J. Am. Chem. Soc. 1983, 105, 4059. (d) Robins, M. J.; Wilson, J. S. J. Am. Chem. Soc. 1981, 103, 932. See (e) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-6. (f ) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 68–69.

75

76

Sugimura, H.; Sato, S.; Tokudome, K.; Yamada, T. Org. Lett. 2014, 16, 3384.

77

Wilt, J. W. Chapter 8 in Free Radicals; Kochi, J. K. Ed., Wiley: New York, 1973; p 333, Vol. 1.

932

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

the acyl peroxide 55 to 2,3-dimethylbut-2-ene, gave 56. Abstraction of a hydrogen atom78 by •CCl3 gave trichloromethane and 4,4,4-trichloro-2,3,3-trimethylbut-1-ene. Transfer of a hydrogen atom to 56 (from solvent, the alkane, or a hydrogen atom), however, led to 1,1,1-trichloro-2,2,3-trimethylbutane. Both processes occurred in this reaction, and typically accompany radical additions. O Cl3C

O

O

CCl3

CCl4 Heat

CCl3

H

Cl3C

Cl3C

H

O 55

56

1,1,1-Trichloro-2,2,3trimethylbutane • CCl3

Cl3C

– CHCl3

4,4,4-Trichloro-2,3,3trimethylbut-1-ene

Alkoxy radicals [e.g., 58 formed via loss of a Cl atom from hypochlorite (57)] often lose a hydrogen atom (to yield butan-2-one when R ¼ H) or an alkyl radical (to yield 59, when R ¼ Me) via cleavage of the bond β to the radical center.79 Thermolysis of tert-pentyl hypochlorite, for example, gave acetone and chloroethane via coupling of the intermediate ethyl radical with alkoxy radical 60.80 Hypochlorites are excellent sources of alkoxy radicals, and can be used to initiate other radical reactions. O – H• (R = H)

R

R

Heat

OCl

O•

– Cl•

O

– Et• (R = Me)

Me

Me

R

57

58 Me

Heat

59 Me

OCl Me

Cl O•

– Cl •

tert-Pentyl hypochlorite

– CH3CH2 •

Me 60

+

Me

O Me

Cl•

17.5.6 Rearrangement and Hydrogen-Atom Abstraction There are a few rearrangement processes of free radicals.67,81 When chlorine, phenyl, acetoxy, and acyl groups are in the β-position with respect to the radical center, 1,2-shifts have been observed. An example is the rearrangement of 61 to 63 via 62.82 Skeletal reorganization occurs in radical intermediates primarily via hydrogen-atom shifts that are probably hydrogen-abstraction reactions. An example is the Barton reaction,83 in which an alkoxy radical (64), generated from the hypochlorite (2-methylhexan-2-yl hypochlorite), abstracts a hydrogen atom via a six-center transition state to yield 65. Subsequently, 65 abstracts a Cl atom from 2-methylhexan-2-yl hypochlorite, via an atom-transfer reaction, to yield 5-chloro-2-methylhexan-2-ol and regenerate reactive radical 65.84 78

Loken, H. Y.; Lawler, R. G.; Ward, H. R. J. Org. Chem. 1973, 38, 106.

79

Reference 3, p 9.

80

Chattaway, F. D.; Baekeberg, O. G. J. Chem. Soc. 1923, 123, 2999.

81

(a) Walling, C. Free Radical Rearrangements in Molecular Rearrangements; de Mayo, P., Ed., Wiley: New York, 1963; p 407, Vol. 1, Chapter 7. (b) Freidlina, R. Kh. Adv. Free Radical Chem. 1965, 1, 211.

82

Kharasch, M. S.; Poshkus, A. C.; Fono, A.; Nudenberg, W. J. Org. Chem. 1951, 16, 1458.

(a) Barton, D. H. R.; Beaton, J. M.; Geller, L. E.; Pechet, M. M. J. Am. Chem. Soc. 1960, 82, 2640. (b) Idem Ibid. 1961, 83, 4076. (c) Barton, D. H. R. Pure Appl. Chem. 1968, 16, 1. (d) Hesse, R. H. Adv. Free Radical Chem. 1969, 3, 83. (e) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-6. (f ) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 66–67.

83

84

Walling, C.; Padwa, A. J. Am. Chem. Soc. 1963, 85, 1593, 1597.

933

17.5 REACTIONS OF FREE RADICALS

Ph O• Ph

X

X O

X

Ph

H

63

HO

O•H Me

+ 2-Methylhexan-2-yl hypochlorite

Me

Me

2-Methylhexan-2-yl hypochlorite

64

Ph

O

62

61 ClO

Ph

Ph

HO

Cl Me +

65

5-Chloro-2-methylhexan-2-ol

65

In a synthetic example, Barton et al.83a converted 3β-acetoxy-5α-pregnan-20β-ol to the nitrite derivative (66) via reaction with nitrosyl chloride (NOCl). Photochemical cleavage (Hg lamp and a Pyrex filter) of 66 generated alkoxy radical 67, which abstracted the hydrogen atom from the neighboring methyl that was properly positioned for reaction via a six-center transition state by virtue of the rigid steroidal structure. The newly formed radical (68) was trapped by the nitrosyl radical to yield 69 and treatment with water gave the oxime (70) in 34% yield. Hydrolysis with dilute acid gave lactol 71. The net result of this transformation is an oxidation (dCH3 ! dCHO; Section 6.8). HO

O=N O

Me

CH3

CH3

H3C

NOCl

O

HO

N CH 2 • NO

HO Me

N

H 2O

H3C

H H

H 67

66

HO CH

68

O

HO

Me 2% aq HCl

H3C

Me

CH3

AcO

H

3 -Acetoxy-5 -pregnan-20 -ol

CH2

H AcO

AcO

H O Me

H3C

H

H

O

CH2

hn

H3C

H AcO

H Me

Me

H3C

Acetone

H AcO

AcO

H

H

H AcO

H

H

69

70 (34%)

71

The reaction is not restricted to steroids, and simply requires that the abstracted hydrogen atom is positioned close enough to react with the alkoxy radical. Reaction of alcohol 72 with bromine and mercuric bromide gave the corresponding hypobromite (OH ! OBr). Subsequent photolysis gave a 43% yield of 73.85a,b Oxidation with chromium trioxide in acetic acid gave the lactone (74), which was used by Hobbs and Magnus85d in a synthesis of grandisol. Gibson and Erman85c used similar intermediates and methodology for the synthesis of cis-bergamotenes. Me

O OH

Me

O 1. HgO, Br2, CCl4, 0°C

Me

2. hn (W lamp), 0°C

Me 72

O

AcOH, CrO3

Me

80°C

Me 73 (43%)

Me 74

(a) Bosworth, N.; Magnus, P. D. J. Chem. Soc. Perkin Trans. 1972, 1, 943. (b) Hortmann, A. G.; Youngstrom, R. E. J. Org. Chem. 1969, 34, 3392. (c) Gibson, T. W.; Erman, W. F. J. Am. Chem. Soc. 1969, 91, 4771. (d) Hobbs, P. D.; Magnus, P. D. J. Am. Chem. Soc. 1976, 98, 4594.

85

934

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

There are potential side reactions. Once the oxygen radical is generated, the hydrogen can be abstracted via a fourcenter transition state if the six-center transition state is not readily accessible. A Barton-like reaction involved photolysis of nitrite 75 to yield the oxygen radical 76.86 Removal of Ha in 76 via a five-center transition state (•OdCdCdCdH) gave the corresponding carbon radical (77), which gave two products via either cyclization in the presence of the alkene and trapping by NO [to give 2-(2-(nitrosomethyl)cyclohexyl)ethan-1-ol in 25% yield] or by direct trapping with NO to give 3-nitrosonon-8-en-1-ol in 14% yield.

hn

Ha

PhH

NO

•

O•

O-N=O 75

OH

NO OH

OH

77

76

+

2-(2-(Nitrosomethyl)cyclohexyl)ethan-1-ol

3-Nitrosonon8-en-1-ol

(25%)

(14%)

17.6 INTERMOLECULAR RADICAL REACTIONS In 1849 Kolbe found87 that electrolysis of potassium acetate generated CO2 and ethane, presumably via acetoxy radical 13 that lost CO2 to give the methyl radical (•CH3). The methyl radical (•CH3) coupled with another methyl radical to yield ethane. The reaction works very well with long, straight-chain alkanoic acids. This reaction is known as the Kolbe electrolytic synthesis87,88 and it was used to convert hexadecanoic acid to n-triacontane (n-C30H62) in 88% yield.89 Kolbe electrosynthesis has been used to prepare rigid, rod-shaped hydrocarbons based on oligo-bicyclo[2.2.2]octane derivatives.90 Much larger and more complex molecules can be synthesized by this method. Stork et al.91 used the Kolbe electrosynthesis to dimerize 78 to 79 (in 40% yield) in a synthesis of α-onocerin. As illustrated by the synthesis of 79, the reaction is very efficient when two identical acids are coupled together. The Kolbe synthesis of two different acids (RCOOH + R0 COOH) generally leads to a mixture of products arising from statistical coupling (RdR, RdR0 , R0 dR0 ). Using an excess of one acid can lead to the desired mixed-coupling product (RdR0 ), as in the reaction of oleic acid with an excess of heptanoic acid to give (Z)-tricos-9-ene in 80% yield.92 There are several variations.92 O H3C

O

e–

O – K+

H3C

• CH3

O•

+

CO2

13 2 • CH3

CH3CH3

CO2H Me

O

H

O

AcO

AcO Me Me

Me

Me O

H

CO2H Oleic acid

86

 Cekovi c, Z.; Ilijev, D. Tetrahedron Lett. 1988, 29, 1441.

87

Kolbe, H. Annalen 1849, 69, 257.

H OAc

Me

78

C8 H17

Me Me

Me

e–

79 (40%)

+

Me(CH2)5CO2H Excess

e–

(CH 2)5Me C8H17 (Z)-Tricos-9-ene (80%)

(a) Crum Brown, A.; Walker, J. Annalen 1891, 261, 107. (b) Lindsey, A. S.; Jeskey, H. Chem. Rev. 1957, 57, 583. (c) Vijh, A. K.; Conway, B. E. Ibid. 1967, 67, 623. (d) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-53. (e) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 372–373.

88

89

Petersen, J. B. Z. Electrochem. 1906, 12, 141.

90

Nuding, G.; V€ ogtle, F.; Danielmeier, K.; Steckhan, E. Synthesis 1996, 71.

91

Stork, G.; Meisels, A.; Davies, J. E. J. Am. Chem. Soc. 1963, 85, 3419.

(a) Seidel, W.; Knolle, J.; Sch€afer, H. J. Chem. Ber. 1977, 110, 3544; also see (b) Knolle, J.; Sch€afer, H. J. Angew. Chem. Int. Ed. Engl. 1975, 14, 758. (c) Kl€ unenberg, H.; Sch€afer, H. J. Ibid. 1978, 17, 47.

92

935

17.6 INTERMOLECULAR RADICAL REACTIONS

Aldehydes can also be a source of radicals via reaction with transition metal salts [e.g., Mn(III) acetate]. Aldehydebased radical addition to alkenes generates ketones in good yield.93 An acyl radical is initially generated via reaction with the transition metal [Mn(III) and Fe(II) are the most common].94 Reaction of butanal and hept-1-ene, in the presence of manganese acetate [Mn(OAc)3], began with addition of the initially formed acyl radical (80) to both carbons of the π-bond, giving regioisomers undecan-4-one (48% yield) and 5-methyldecan-4-one (3% yield), as well as a 27% yield of 2-ethylnonanal and a 22% yield of 3-formyldecan-5-yl acetate. Removal of the hydrogen atom α to the carbonyl led to 81, and addition to the alkene gave 2-ethylnonanal.95 This reaction also occurs with ketones, but usually requires catalysis (with di-tert-butyl peroxide, e.g.).96 A simple application is seen in the coupling reaction of cyclopentanone with dec-1-ene to give 2-decylcyclopentanone in 71% yield.96b Mn(OAc)3 AcOH

O H

Et

80

Et

81

Me

5-Methyldecan-4-one

(48%)

(3%)

O

+

H

Et

C5H11

Undecan-4-one

O

C5H11

H

Et

O

+

Et C5H11

Et

70°C , 1.5 h

O

O

C5H11

O

H

Et

OAc

C5H11

2-Ethylnonanal

C5H11 3-Formyldecan-5-yl acetate

(27%)

(22%)

Giese6,97 describes many synthetic applications of radical coupling with alkenes. Reactions with simple alkenes are not very efficient, but reactions with alkenes bearing an electron-withdrawing group are usually quite facile. Indeed, conjugate addition of radicals is well known.98 Perhaps the most common method for generating the radicals for reaction with alkenes is to heat an alkyl halide with AIBN, in the presence of tri-n-butyltin hydride (Bu3SndH). In a typical reaction, AIBN is thermally decomposed (refluxing benzene) or decomposed by exposure to light, and the resulting radical intermediate reacts with iodocyclohexane to yield 82. This radical reacts with ethyl acrylate to yield 83. Tributyltin hydride is an efficient intermolecular hydrogen donor, reacting with 83 to form the product (ethyl 3-cyclohexylpropanoate), and generate a chain-carrying radical Bu3Sn•, which can react with additional halide to produce 82. Giese and Gerth99 used this technique to couple 84 with methyl methacrylate, giving 85 in 70% yield. Removal of the benzylidene acetal (Section 5.3.3.1) by hydrogenation and subsequent lactonization gave malyngolide.100 CO2Et

I

HSnBu3

AIBN

+ Bu3Sn • CO2Et

CO2Et

+ Me2C(I)CN

82

83

Ph CO2Me

I

O

, EtOH

O

C9H19

O

Ph

Me MeO2C

0.2 equiv. Bu3SnCl Excess NaBH4

84

93

Ethyl 3-cyclohexylpropanoate

O

O

C9H19

85 (70%)

H2–Pd

Me

O

OH

C9H19 Malyngolide

Reference 3, p 69.

94

(a) Sosnovsky, G. Free Radical Reactions in Preparative Organic Chemistry; MacMillan: New York, 1964. (b) Vinogradov, M. G.; Nikishin, G. I. Usp. Khim. 1971, 40, 1960. (c) Nikishin, G. I.; Vinogradov, M. G.; Il’ina, G. P. Synthesis 1972, 376.

95

Nikishin, G. I.; Vinogradov, M. G.; Verenchikov, S. P.; Kostyukov, I. N.; Kereselidze, R. V. J. Org. Chem. USSR 1972, 8, 539 (Engl. p 544).

(a) Nikishin, G. I.; Somov, G. V.; Petrov, A. D. Dokl. Akad. Nauk. 1961, 136, 1099. (b) Idem Izvest. Akad. Nauk. 1961, 2065 (Engl. p 1924). (c) Glukhovtsev, V. G.; Spektor, S. S.; Golubev, I. N.; Nikishin, G. I. J. Org. Chem. USSR 1973, 9, 316 (Engl. p 317).

96

97

(a) Giese, B. Angew. Chem. Int. Ed. 1983, 22, 753. (b) Idem Ibid. 1985, 24, 553.

98

See Srikanth, G. S. C.; Castle, S. L. Tetrahedron 2005, 61, 10377.

99

Gerth, D. B.; Giese, B. J. Org. Chem. 1986, 51, 3726.

100

Reference 6, p 58.

936

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

Conjugate addition of radicals is facile. The radical addition of alkyl halides with conjugated ketones, esters, and nitriles (CH2]CHY) to give a coupling product (RCH2CH2Y), occurs using both photochemical and thermal conditions.100 Primary, secondary, and tertiary halides react under photochemical conditions, or by treatment with AIBN, to give good-to-excellent yields of product. Styrene reacts under similar conditions. Competing hydrogen transfer to directly reduce the alkyl halide (dCH2I ! dCH2H), as described in Sections 7.11.7 and 17.5.4, is a problem with radical coupling in the presence of a hydrogen-donating species (e.g., tin hydride). If the initially formed radical is long lived and/or if the radical coupling reaction is slow, that radical can be reduced by hydrogen abstraction from the tin hydride. In some cases, the reduction process can be the major reaction, but 20–25% yields are typical. α-Halo esters add to alkenes to form either lactones or substituted ester derivatives.101,102 An alternative route has been used for radical coupling, involving alkyl halides and allyl tin derivatives (e.g., allyltributyltin, Bu3SnCH2CH]CH2).103,104 Migata and coworkers105 and also Pereyre and coworkers,106 introduced allyltin compounds used in this context, but Keck and et al.103,104 contributed significantly to the synthetic development of these reagents. Allyl tin compounds react with alkyl halides to give the corresponding coupling product (see Section 18.8.1). An electron-withdrawing substituent at C2 of the allyl stannane increased the reactivity, illustrated by the coupling of 1-bromoadamantane and 86 to give 87 in 70% yield.107 EtO2C Br

SnPh3

+

AIBN

CO2Et

Heat

1-Bromoadamantane

86

87 (70%)

Allyl stannanes bearing substituents at C1or C3 cannot generally be used in these reactions,108 but there are exceptions.109 When there is a substituent at the C3 position, addition of the radical is slow. When C1 is substituted, the stannane can rearrange, which can be competitive with addition.110 This latter problem was solved, at least for prenylation, by the use of an allyl sulfide precursor (e.g., 88). Keck and Byers111 showed that 89 reacted with 88 in the presence of tributyltin dimer (Bu3SndSnBu3) to give 90 in 76% yield. Photolysis of tributyltin dimer generates 2 equiv. of Bu3Sn•, which initiates the reaction. Good diastereoselectivity may be achieved in radical addition reactions. When 91 was treated with tert-butyl iodide, Bu3SnH, and Sc(OTf )3, a 58% yield of 92 was obtained as a 1:99 anti/syn mixture.112 Me Me

O

Ph

+

SPh

Ph Bu3SnSnBu3

O

Ph

Ph

O O Me

hn

Br

88

89 Me

MeO2C

90 (76%) Me

t-BuI, Bu3SnH, Sc(OTf)3

CO2Me

CH2Cl2, –78°C

MeO2C

91

+ Bu3SnBr + Bu3SnSPh

Me t-Bu

CO2Me

92

101

Kharasch, M. S.; Skell, P. S.; Fisher, P. J. Am. Chem. Soc. 1948, 70, 1055.

102

(a) Reference 12, p 501; (b) Giese, B.; Horler, H.; Leising, M. Chem. Ber. 1986, 119, 444.

103

Keck, G. E.; Yates, J. B. J. Am. Chem. Soc. 1982, 104, 5829.

104

Keck, G. E.; Enholm, E. J.; Yates, J. B.; Wiley: M. R. Tetrahedron 1985, 41, 4079.

(a) Kosugi, M.; Kurino, K.; Takayama, K.; Migata, T. J. Organomet. Chem. 1973, 56, C11. (b) Migata, T.; Nagai, K.; Kosugi, M. Bull. Chem. Soc. Jpn. 1983, 56, 2480.

105

106

(a) Grignon, J.; Pereyre, M. J. Organomet. Chem. 1973, 61, C33. (b) Grignon, J.; Servens, C.; Pereyre, M. Ibid. 1975, 96, 225.

107

Baldwin, J. E.; Adlington, R. M.; Birch, D. J.; Crawford, J. A.; Sweeney, J. B. J. Chem. Soc. Chem. Commun. 1986, 1339.

108

Reference 12, p 491.

109

Fliri, H.; Mak, C.-P. J. Org. Chem. 1985, 50, 3438.

110

(a) Keck, G. E.; Yates, J. B. J. Organomet. Chem. 1983, 248, C21. (b) Baldwin, J. E.; Adlington, R. M.; Basak, A. J. Chem. Soc. Chem. Commun. 1984, 1284.

111

Keck, G. E.; Byers, J. H. J. Org. Chem. 1985, 50, 5442.

(a) Hayen, A.; Koch, R.; Saak, W.; Haase, D.; Metzger, J. O. J. Am. Chem. Soc. 2000, 122, 12458. Also see (b) Hayen, A.; Koch, R.; Metzger, J. O. Angew. Chem. Int. Ed. 2000, 39, 2758.

112

937

17.6 INTERMOLECULAR RADICAL REACTIONS

When heteroatom-containing substrates react with peroxides or other radical initiators, hydrogen-atom transfer can occur as in the transfer from an acetal to the radical to give an alkane and the α-alkoxy radical 93. The presence of the heteroatom α to the carbon bearing the radical center leads to enhanced stability. Such radicals add to alkenes, usually with anti-Markovnikov orientation, as in the radical induced addition of HBr to alkenes (Section 2.5.1). The reaction usually involves slow addition of a large excess (5–10 equiv.) of the heteroatom-containing substrate. Peroxides and peroxyesters are the usual radical initiators. The hydrogen-atom transfer reaction is often very slow, and an excess of the hydrogen-atom donor (the heteroatom substrate) and higher reaction temperatures are essential.113 Curran114 showed several general examples of this type of reaction involving the reaction of alcohols with alkenes to yield 94, amines with alkenes to yield 95, esters with alkenes to yield 96, or aldehydes with alkenes to yield 97. H

H

RO

• R1

+

RO

OR 93

C8 H17

OH Me

R1—H

OR Me

C8 H17

HO

Me

Me

94

CH2OH

N

N

H

H OBu

C8 H17

95

C8H17

CO2Me 96 C8 H17

Me

OH

OBu CO2Me

O

CHO Me

C8 H17

97

Alkynes are generally less reactive than alkenes in radical coupling reactions since the LUMO of the alkyne is energetically higher than the LUMO of an alkene (Table 14.1). Formation of the requisite SOMO is therefore more difficult.115 The intermediate vinyl radical (98) undergoes coupling to yield a mixture of cis and trans-alkenes, as shown, but when R is bulky the cis compound, (Z)-1-cyclohexyl-2-phenylethene, is favored via trans-attack.115 Addition to methyl propiolate showed better reactivity, but poorer selectivity.115 As the size of R increases, syn approach of the hydrogen donor is increasingly hindered.115 Nonradical nucleophiles usually react faster with alkynes than with alkenes, however.116

H

• R1

R

R1

R1

X—Y

R1

R

X

+

R

X

R

98

The disconnections that are pertinent to intermolecular radical coupling reactions include the following: O

O

R R

R

CO2H

R1

R R2

O R

R1

RJCHO +

+

R

R1

R2 X

R1 R

RJX +

Z

113

Reference 12, p 497.

114

Reference 12, p 498.

115

(a) Giese, B.; Lachhein, S. Angew. Chem. Int. Ed. 1982, 21, 768. (b) Giese, B.; Meixner, J. Ibid. 1979, 18, 154.

116

Dickstein, J. I.; Miller, G. I. In The Chemistry of Carbon Carbon Triple Bonds; Patai, S., Ed., Wiley: New York, 1978; Vol. 2.

938

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

17.7 INTRAMOLECULAR RADICAL REACTIONS (RADICAL CYCLIZATION) 17.7.1 Carbocyclic Ring Systems There are many examples of radical reactions that produce rings.12,117 The fundamentals of the reaction can be illustrated by the AIBN induced reaction of tributyltin hydride and 99 (X ¼ I or Br), which generates radical 47. Once formed, at least three pathways are possible under these conditions.118 The first is path a, a 5-exo-trig-cyclization to yield 48. Path b is a 6-endo-trig-cyclization to yield 82. These descriptors are based on Baldwin’s rules (Section 10.6) and both pathways are favored. The last pathway is the usual hydrogen-transfer process where tin hydride reacts with 47 to yield hex-1-ene (path c), as discussed in Section 17.5.4. In general, the 5-exo-trig-pathway is lower in energy than path b and faster than path c, leading to 48 as the major product. This outcome is subject to steric effects or electronic effects that arise from interaction of substituents on the alkenyl moiety. At high concentrations (>5 M), hydrogen transfer to give hex-1-ene usually predominates so radical cyclization reactions are typically done at low concentrations (95:5. That ratio is 93:7 when R is CO2R1, and 75:25 when R is COR or COPh.122 Another potential problem with this reaction is formation of cis- or trans-isomers. Methyl 2-iodohept-6-enoate gave essentially a 1:1 mixture of cis (105) and trans (106) products under a variety of conditions,122 but in all cases the 6-endotrig product (107) is formed as a minor constituent. In the absence of special controlling factors (steric, electronic, neighboring group effects, etc.) the cyclization proceeds with poor cis-trans selectivity. R

+ R

R

102

103

104 I

I

I

0.3 M, PhH

I

+ CO2Me

CO2Me Methyl 2-iodohept-6-enoate

CO2Me

105

+

MeO2C

106

107

In the formation of polycyclic systems, the cyclization process can lead to more than one regioisomer or stereoisomer, and the ratio of (E)- and (Z)-isomers is often 1:1. In most cyclizations, cis-ring fusion predominates when the final product is a bicyclo[3.3.0]- or bicyclo[4.3.0]-system, but formation of a bicyclo[5.3.0]-system usually leads to significant amounts of trans-fused products.123 In the cyclization of 108, a mixture of (E)- and (Z)-isomers is possible, and there are two different alkene moieties in 108 that lead to different regioisomeric products.123 Abstraction of the bromine atom in 108 with Bu3Sn• generated a radical that cyclized, and was quenched by reaction with tin hydride, via hydrogen-atom transfer, to yield a 1:1 mixture of 109/110 in 89% yield. Note that there are diastereomers resulting from the presence of OMe and the ratio of α/β-methoxy isomers in 110 was 3:1.124 All products were cis-fused. The reaction was controlled by frontier molecular orbitals (SOMO-LUMO) rather than stabilities of radical intermediates. OMe

H

Br Bu3SnH

H

+

AIBN

MeO

CO2Me 108

MeO

CO2Me 109

CO2Me 110

There are many synthetic examples that use radical cyclization as a key step, and the radical precursor is not limited to unfunctionalized alkyl iodides or bromides. In Zhai and coworker’s125 synthesis of sculponeatin N, the vinyl bromide unit in 111 was converted to the corresponding radical, and intramolecular cyclization yielded 112 in 68%. Note the use of triethylborane as the radical initiator. Radical cascade reactions can be used for the synthesis of polycyclic ring systems. In these reactions, polyenes are subjected to radical cyclization, generating tricyclic or even tetracyclic ring systems.126 Chiral auxiliaries have been used effectively in radical cyclization reactions.127

122

Curran, D. P.; Chang, C.-T. J. Org. Chem. 1989, 54, 3140.

123

(a) Reference 12, p 421. (b) Clive, D. L. J.; Cheshire, D. R.; Set, L. J. Chem. Soc. Chem. Commun. 1987, 353.

124

Beckwith, A. L. J.; O’Shea, D. M.; Roberts, D. H. J. Chem. Soc. Chem. Commun. 1983, 1445.

125

Pan, Z.; Zheng, C.; Wang, H.; Chen, Y.; Li, Y.; Cheng, B.; Zhai, H. Org. Lett. 2014, 16, 216.

For examples, see (a) Boehm, H. M.; Handa, S.; Pattenden, G.; Roberts, L.; Blake, A. J.; Li, W.-S. J. Chem. Soc. Perkin Trans. 2000, 1, 3522. (b) Fensterbank, L.; Mainetti, E.; Devin, P.; Malacria, M. Synlett 2000, 1342.

126

127

For an example taken from a synthesis of (+)-triptocallol, see Yang, D.; Xu, M.; Bian, M.-Y. Org. Lett. 2001, 3, 111.

940

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

H

H

Br BEt3, (TMS)3SiH PhH, 25°C

O

O

O

O

TBSO

TBSO 111

112 (68%)

Alkynes are common precursors to exo-methylene compounds, as in Sha and coworker’s128 synthesis of ()-bakkenolide. Treatment of iodide 113 with tributyltin dimer and then AIBN/Bu3SnH led to the radical intermediate and cyclization. The second step was required to reduce the iodide formed in the first reaction of the radical by tributyltin hydride to give a 93% yield of 114. The trimethylsilyl group attached to the alkyne unit is important to sterically inhibit a 6-endo-dig reaction that can complete with the 5-exo-dig process. The silyl group in 114 was removed by treatment with trifluoroacetic acid in dichloromethane, giving 115 in 80% yield along with 10% of the conjugated ketone in which the C]C unit migrated to the endocyclic position. O

O SiMe3

I

SiMe3

H

O

H

CF3CO2H

1. (BuSn) 2 h

Benzene

2. Bu3SnH, AIBN

113

114 (93%)

115 (80%)

Vinyl radicals are useful in radical coupling and addition reactions.129 Stork and Baine130 used the cyclization of vinyl radical 117 (derived from vinyl bromide 116) to complete a synthesis of norseychellanone. The cyclization step proceeded in 70% yield. Coupling one aromatic ring to another is also possible using radical techniques. In a synthesis of goniopedaline, Couture and coworkers131 heated 118 at reflux in benzene, with AIBN and Bu3SnH, and obtained a 75% yield of 119. Me O

Br

Me O

Bu3SnH

•

116

BnO

N

OMe

Bu3SnH, AIBN PhH, Reflux

118

128

Me

Norseychellanone

OMe O

O

Pd

117

MeO I

H2

Me

AIBN, PhH, Reflux

Me

BnO

Me

OMe

MeO

O OMe N

119 (75%)

Jiang, C.-H.; Bhatacharyya, A.; Sha, C.-K. Org. Lett. 2007, 9, 3241.

(a) Stork, G.; Baine, N. H. Am. Chem. Soc. 1982, 104, 2321. (b) Stork, G.; Mook, R., Jr. Ibid. 1983, 105, 3720. (c) Nozaki, K.; Oshima, K.; Utimoto, K. Ibid. 1987, 109, 2547. (d) Stork, G.; Mook, R., Jr. Ibid. 1987, 109, 2829. (e) Beckwith, A. L. J.; O’Shea, D. M. Tetrahedron Lett. 1986, 27, 4525. (f ) Stork, G.; Mook, R., Jr. Ibid. 1986, 27, 4529.

129

130

Stork, G.; Baine, N. H. Tetrahedron Lett. 1985, 26, 5927.

131

Rys, V.; Couture, A.; Deniau, E.; Grandclaudon, P. Eur. J. Org. Chem. 2003, 1231.

941

17.7 INTRAMOLECULAR RADICAL REACTIONS (RADICAL CYCLIZATION)

The radical cyclization disconnections follow: R

R X

(CH 2)n

(CH 2)n

X (CH 2)n

(CH 2)n

R

R

X R

R

X

R X

X

R

X

17.7.2 Heteroatom Ring Systems Many compounds that have heterocyclic rings can be prepared by radical cyclization.132 Indeed, heteroatoms can be part of the chain that links the radical precursor and the alkene. The standard reagents of AIBN and tributyltin hydride, heated at reflux in benzene, are common, but other solvents can also be used. For example, radical cyclization of 120 in aqueous ethanol using a combination of phosphoric acid and a base led to 121 in 83% yield.133 Stork et al.134 prepared protected lactones via radical cyclization of alkenyl ethers. Tang et al.135 used this reaction in a synthesis of aplykurodinone-1, in which iodoester 122 was converted to a radical that gave 123 in 55% yield. aq H3PO2, AIBN, 5 h NaHCO3, EtOH, Reflux

I

O

O 120

121 (83%)

Heterocyclic rings can be prepared by simply incorporating an oxygen atom into the chain, as in the cyclization of 120 to 121 or 122 to 123. This reaction is not limited to oxygen, and a nitrogen-containing substrate was reported by Koreeda et al.136 in a synthesis of ()-sibirine. The reaction of selenide 124 with AIBN/Bu3SnH, heated at reflux in toluene, led to a close to 1:1 mixture of the 5-exo-trig product 125 and the 7-endo-trig product, 126. Based on Baldwin’s rules (Section 10.6), it is clear that there is little difference in the transition state energies leading to the two products.

H

BEt3, Bu3SnH

O O

O

Toluene

O

H

I

O 122

123 (55%) CO2Me

OTBS

O

SePh

TBSO Bu3SnH, AIBN

N

TBSO

N

H

+

Toluene, Reflux

N

CO2Me 124

CO2Me 125

(2:1)

126

(a) Majumdar, K. C.; Basu, P. K.; Mukhopadhyay, P. P. Tetrahedron 2004, 60, 6239. (b) Bowman, W. R.; Fletcher, A. J.; Potts, G. B. S. J. Chem. Soc. Perkin Trans. 2002, 1, 2747. 132

133

Yorimitsu, H.; Shinokubo, H.; Oshima, K. Chem. Lett. 2000, 104.

(a) Stork, G.; La Clair, J. J.; Spargo, P.; Nargund, R. P.; Totah, N. J. Am. Chem. Soc. 1996, 118, 5304. Also see (b) Stork, G.; Mook, R., Jr.; Biller, S. A.; Rychnovsky, S. D. J. Am. Chem. Soc. 1983, 105, 3741.

134

135

Tang, Y.; Liu, J.-T.; Chen, P.; Lv, M.-C.; Wang, Z.-Z.; Huang, Y.-K. J. Org. Chem. 2014, 79, 11729.

136

Koreeda, M.; Wang, Y.; Zhang, L. Org. Lett. 2002, 4, 3329.

942

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

Both phenylthio (dSPh) and phenylseleno (dSePh) groups can be used as leaving groups for radical cyclization. In these cases, sulfur or selenium group transfer allows formation of the radical. Hart and Tsai137c used this technique to cyclize 127 in a synthesis of isoretronecanol. The cyclization provided primarily racemic diastereomer 128. Smith and Keusenkothen,138 however, used a chiral template approach (Section 8.9) and prepared alkenyl iodide 129 from the chiral precursor pyroglutamate. Radical cyclization gave pyrrolizidinone 130 in a synthesis of pseudoheliotridane, with high asymmetric induction.138 CO2t-Bu

t-BuO2C

O

H

Bu3SnH, AIBN

N

PhH, Reflux

N O

SPh 127

128

N

Bu3SnH, AIBN

O

N

PhH, Reflux

I

O

130

129

Knapp et al.139 also used pyroglutamate derivatives for asymmetric radical cyclizations. Other heterocycles or reduced heterocyclic rings can be used in radical cyclizations,140 as in the treatment of pyrrole derivative 131 with AIBN and tributyltin hydride gave a 60% yield of 132. The sulfonyl group on the pyrrole ring was required for good yields of cyclized product.141 Radical cyclization can proceed with high diastereoselectivity and high asymmetric induction when chiral precursors are used (as with 129 ! 130 from pyroglutamate). Livinghouse and Jolly142 reported a chiral synthesis of ()trachelanthamidine via cyclization of 133, prepared from prolinal, which gave a 30:1 mixture of 134 and 135 in 55% yield. In this reaction, initiation with AIBN in the presence of tributyltin hydride led to significant amounts of the hydrogentransfer product (2-ethenyl-N-acetyl-2-pyrrolidinone). Hydrogen transfer is a common problem, and was solved in this work by changing the method by which the radical is generated. Radical cyclization in the presence of tributyltin dimer (Bu3SnSnBu3) gave the desired products (as shown), but required photochemical initiation with a sunlamp.

MeO2S

Bu3SnH, AIBN, PhH, Reflux

N

MeO2S

N

Br 131

132 (60%)

I

H 0.55 equiv. (Bu3Sn)2

N

I

EtJI, hn

I

H

+ N

N

O 133

O

O 134

135

(a) Choi, J.-K.; Hart, D. J.; Tsai, Y.-M. Tetrahedron Lett. 1982, 23, 4765. (b) Burnett, D. A.; Choi, J.-K.; Hart, D. J.; Tsai, Y.-M. J. Am. Chem. Soc. 1984, 106, 8201. (c) Hart, D. J.; Tsai, Y.-M. Ibid. 1982, 104, 1430. (d) Idem Ibid. 1984, 106, 8209. (e) Choi, J.-K.; Hart, D. J. Tetrahedron 1985, 41, 3959.

137

138

(a) Keusenkothen, P. F.; Smith, M. B. J. Chem. Soc. Perkin Trans. 1994, 1, 2485. (b) Keusenkothen, P. F.; Smith, M. B. Tetrahedron Lett. 1989, 30, 3369.

139

Knapp, S.; Gibson, F. S.; Choe, Y. H. Tetrahedron Lett. 1990, 31, 5397.

140

Bowman, W. R.; Bridge, C. F.; Brookes, P. J. Chem. Soc. Perkin Trans. 2000, 1, 1.

141

Antonio, Y.; de la Cruz, E.; Galeazzi, E.; Guzman, A.; Bray, B. L.; Greenhouse, R.; Kurz, L. J.; Lustig, D. A.; Maddox, M. L.; Muchowski, J. M. Can. J. Chem. 1994, 72, 15. 142

Jolly, R. S.; Livinghouse, T. J. Am. Chem. Soc. 1988, 110, 7536.

943

17.7 INTRAMOLECULAR RADICAL REACTIONS (RADICAL CYCLIZATION)

The use of dibutyltin dimer under photochemical methods relies on atom transfer, and it is most useful when the cyclization step is slow, as in the reaction of 133. When cyclization is slow relative to hydrogen transfer, the dominant reaction is usually hydrogen-atom transfer to yield reduced products. Photolysis of the tin dimer generates the radical intermediate without a hydrogen-atom transfer agent present (other than the substrate and the solvent). Another example illustrating the utility of this reagent in cyclization reactions is Curran and Chang’s143 attempt to cyclize 136, which gave only 137 on treatment with tin hydride/AIBN, but no 138. Cyclization with dibutyltin dimer and photochemical induction gave a radical intermediate that was quenched by reaction with the iodine radical byproduct to yield 138. The iodine could be reduced by reaction with tin hydride (see Section 17.5.4).143 In radical cyclization reactions that involve formation of a six-membered ring, allylic hydrogen abstraction can be the preferred reaction. Attempted cyclization of 139 gave only 142 via hydrogen abstraction (from 141 to 142), and no cyclization was observed.144 A six-center transition state is possible in 140 for removal of the allylic hydrogen atom, and this process is faster than the 6-exo-trig transition state required for ring formation. The methyl groups in 140 presumably prevent close approach of the requisite hydrogen to the radical center. If an activating group is attached to the alkene receptor, formation of the six-membered ring is accelerated relative to hydrogen transfer. O

O

Bu3SnH 0.02 M, PhH

O

O

137

I

OH

H

O

O

O

H

Bu3SnH

H

OH

Bu3SnH

AIBN, PhMe Reflux, 16 h

SePh

O 138

OH

OH

Me

O

0.3 M, PhH

136

H

O

(Bu3Sn)2, hn

I

Me

Me

Me

Me 140

139

Me

Me

Me

141

142

Bach and coworkers145 reported an asymmetric radical cyclization, in which alkenyl-iodide 143 cyclized under radical conditions (initiated by BEt3) in the presence of the chiral complexing agent 144. The selectivity is believed to rise from hydrogen-bond mediation of the radical cyclization in the presence of 144. Under these conditions, cyclization of 143 gave a 71% yield of 145 in 79 %ee. Radical conjugate additions of this type are very facile, and many internal cyclization reactions have been reported.146 Relatively large rings can be prepared, as illustrated by the cyclization of 146 to 147 and 148 (63% and 22% yield, respectively).147 The cyclization process is faster than the hydrogen transfer under high dilution conditions. The two remote reactive fragments of the long chain must be in close proximity for the cyclization in the transition state. O

I H

N

+ O

NH

N

H Bu3SnH, BEt3

H

Toluene, –78°C

O

O 143

144

O

N

145 (71%)

Bu3SnH, AIBN Reflux

O

O

+

0.005 M, PhH

I 146 143

Curran, D. P.; Chang, C.-T. Tetrahedron Lett. 1987, 28, 2477.

144

Leonard, W. R.; Livinghouse, T. Tetrahedron Lett. 1985, 26, 6431.

145

Aechtner, T.; Dressel, M.; Bach, T. Angew. Chem. Int. Ed. 2004, 43, 5849.

146

For example, see Winkler, J. D.; Sridar, V. Tetrahedron Lett. 1988, 29, 6219.

147

Porter, N. A.; Chang, V. H.-T. J. Am. Chem. Soc. 1987, 109, 4976.

147 (63%)

148 (22%)

944

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

The alkenyl and alkynyl radical disconnections follow: R

(CH 2)n

R

R

X

(CH 2)n

R

(CH 2)n

(CH 2)n

R

O

O

R

(CH 2)n

X

X (CH 2)n

(CH 2)n

(CH 2)n

Samarium(II) iodide mediates radical-type cyclization reactions, and has been used in the synthesis of natural products.148 Two examples illustrate the utility, and similarity to the transformations described above. In a synthesis of sclerophytin F, Clark et al.149 showed that ketone 149 reacted with SmI2 to yield the cyclized product, (150) in 89% yield, via intramolecular conjugate addition of the initially formed HOdC• radical. Ohta and coworkers150 formed the radical from the primary alkyl bromide in a synthesis of pyrrolam A. Alkynyl lactam 151 was treated with excess SmI2, leading to a 90% yield of cyclized product indolizidinone (152). O OH

SmI2, MeOH

CO2Et

O

THF, rt

H

149

O

H

CO2Et

150 (89%)

Br

Ph

N

H Ph

0°C, 1 h

N

3 equiv. SmI2 THF–HMPA

O

O

151

152 (90%)

17.7.3 Cyclization of •CdX Radicals and Cyclization of Heteroatom Radicals Radical cyclization is not limited to alkene and alkyne substrates. Carbonyls and imines can also be radical receptors. As shown in the cyclization of 149, carbon radicals that contain oxygen can be formed from carbonyls. Enholm and et al.151 showed that when 153 was treated with AIBN and tributyltin hydride, the major product was a mixture of alcohols 154 and 155. Fu and Hays152 showed that this cyclization could be made catalytic in tributyltin hydride. When 153 was treated with 0.1 Bu3SnH, 0.5 PhSiH, 2 equiv. of EtOH and AIBN, an 85% yield was obtained as a 1.1:1 154/155 mixture. Ph

1.5 Bu3SnH, AIBN

CHO

Ph

+

PhH, 80°C

153

Ph

OH 154

OH 155

Radical cyclization to the C]N unit of an imine generates amine derivatives. In the Keck et al.153 synthesis of 7-deoxypancratistatin, radical precursor 156 was cyclized to the oximine ether unit under radical conditions to yield a 3:1 mixture of 157 and 158 in 90% yield. Another variation involves generation, cyclization, and trapping of amidyl radicals, as in a synthesis of tylophorine shown by Wang and coworkers.154 When thiosemicarbazone (159) was treated with dilauroyl peroxide, the amidyl radical produced in this reaction cyclized to give lactam 160 in 55% yield. 148

Edmonds, D. J.; Johnston, D.; Procter, D. J. Chem. Rev. 2004, 104, 3371.

149

Clark, J. S.; Delion, L.; Farrugia, L. J. Org. Lett. 2014, 16, 4300.

Aoyagi, Y.; Manabe, T.; Ohta, A.; Kurihara, T.; Pang, G.-L.; Yuhara, T. Tetrahedron 1996, 52, 869. Also see Curran, D. P.; Fevig, T. L.; Totleben, M. J. Synlett 1990, 773.

150

(a) Enholm, E. J.; Prasad, G. Tetrahedron Lett. 1989, 30, 4939. (b) Enholm, E. J.; Burroff, J. A. Tetrahedron Lett. 1992, 33, 1835. (c) Enholm, E. J.; Burroff, J. A. Tetrahedron 1997, 53, 13583. (d) Enholm, E. J.; Kinter, K. S. J. Org. Chem. 1995, 60, 4850.

151

152

Hays, D. S.; Fu, G. C. Tetrahedron 1999, 55, 8815.

153

Keck, G. E.; McHardy, S. F.; Murry, J. A. J. Org. Chem. 1999, 64, 4465.

154

Han, G.; Liu, Y.; Wang, Q. Org. Lett. 2013, 15, 5334.

945

17.7 INTRAMOLECULAR RADICAL REACTIONS (RADICAL CYCLIZATION)

MOMO O

OMOM

MOMO O

MOMO

O

N

MOMO

O

Bu3SnH

O

O

OMOM

N–OBn

+

O

AIBN

O

S

NHOBn

O

O

O

O NHOBn

O

N 156

157

MeO

158

MeO

SMe Ph

N

S Dilauroyl peroxide

N

N

O

MeO

O

MeO OMe

OMe

159

160 (55%)

The C]X radical cyclization disconnection follows: Y

HJX

(CH 2)n

X

(CH 2)n

A classical named reaction provides another example of a hydrogen-atom transfer reaction that produces a heterocyclic compound. When N-halo-amines (e.g., 1-bromo-2-propylpiperidine)155 are treated with concentrated sulfuric acid, the amino radical cation 161 is produced. Hydrogen-atom transfer via a six-center transition state generates the radical ammonium salt (162), which reacts with the protonated form of 1-bromo-2-propylpiperidine (1-bromo2-propylpiperidin-1-ium) to yield 2-(3-bromopropyl)piperidin-1-ium. When 2-(3-bromopropyl)piperidin-1-ium is neutralized, the amine displaces the primary halide to yield octahydroindolizine.156 The conversion of N-haloamines to cyclic amines is called the Hofmann-L€ offler-Freytag reaction.157

H2SO 4

N Br

N Heat or hn

1-Bromo-2-propylpiperidine

N

H

H

H

H 161

162

N Br

NaOH

H

N H H Br 2-(3-Bromopropyl)piperidin-1-ium

155

Kovacic, P.; Lowery, M. K.; Field, K. W. Chem. Rev. 1970, 70, 639.

156

Corey, E. J.; Hertler, W. R. J. Am. Chem. Soc. 1960, 82, 1657.

N

Octahydroindolizine

(a) Hofmann, A. W. Berichte 1883, 16, 558. (b) Idem Ibid. 1885, 18, 5. (c) L€ offler, K.; Freytag, C. Ibid. 1909, 42, 3427. (d) Wolff, M. E. Chem. Rev. 1963, 63, 55. (e) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-45. (f ) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 328–329.

157

946

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

In its original form, this reaction generated pyrrolidine derivatives and other cyclic amines, as in the conversion of 163 to 164. Treatment with a base generated the free amine, and subsequent nucleophilic displacement of chloride (cyclization) led to 165 as the final product. This acid-mediated reaction can be initiated photochemically,158 as in the conversion of ethyl L-isoleucinate to 166 in 99  53% overall yield, taken from Sanz-Cervera and Williams’159 synthesis of ()-VM55599. Initial N-chlorination of the primary amine in ethyl L-isoleucinate gave a 99% yield of the chloroamine, ethyl N-chloro-L-isoleucinate. This reaction was followed by irradiation in sulfuric acid to yield ethyl (2S,3S)-2-amino-5-chloro-3-methylpentanoate. The final step was basification to generate the pyrrolidine (in this case a proline derivative), and protection of the nitrogen as a tert-butylcarbamate (N-Boc, see Section 5.3.4.3), 166. The high temperatures required in older versions of this conversion can lead to poor stereoselectivity. One solution to this problem is to use a mixture of cuprous chloride and cupric chloride (CuCldCuCl2) in acidic media to effect cyclization.160 R1 R

N (CH 2)n

H+, Heat

Cl

hn

R

163

(CH 2)n

N H H

NaOH

R

t-BuOCl

CO2Et

R1

165 CO2t-Bu

Cl

NHCl

N (CH 2)n

164

NH2

Ethyl L-isoleucinate

R1

Cl

or

hn H2SO 4

NH2

CO2Et

1. NaOH, pH 7 2. (t-BuOCO)2O

N

CO2Et

CO2Et

Ethyl N-chloroL-isoleucinate

Ethyl (2S,3S)-2-amino-5chloro-3-methylpentanoate

166

The Hofmann-L€ offler-Freytag disconnection follows: R1

R1 (CH 2)n

N R

(CH 2)n

N R Cl

17.7.4 Bergman Cyclization Ene-diynes undergo a cycloaromatization reaction that is formally a thermally allowed electrocyclic reaction. In principle, this reaction belongs in Section 15.3, but it is known that the reaction proceeds by a diradical. For that reason, and since a ring is formed, it is placed in this section. However, it is not formally a radical cyclization. In a generic example, ene-diyne 167 cyclizes to yield diradical 168, which loses two hydrogen atoms via aromatization to yield the benzene ring in 169. This conversion of an ene-diyne to a benzene derivative is called the Bergman cyclization.161 The activation barrier for this process can be greatly influenced by electronic-substituent effects.162 A metal-accelerated Bergman cycloaromatization has been reported, using an arene complex [(η5-C5Me5)Ru-η6-3,4-benzocyclodec-3-ene1,5-diyne][OTf].163 This reaction has taken on great importance with the discovery that the biological action of several anticancer antibiotics (calicheamicin, dynemicin, esperamicin, and kedarcidin) is linked to Bergman cyclization of an ene-diyne core structure accompanying cleavage of double-stranded DNA164

158

Furstoss, R.; Teissier, P.; Waegell, B. Tetrahedron Lett. 1970, 1263.

159

Sanz-Cervera, J.; Williams, R. M. J. Am. Chem. Soc. 2002, 124, 2556.

160

Broka, C. A.; Eng, K. K. J. Org. Chem. 1986, 51, 5043.

(a) Jones, R. R.; Bergman, R. G. J. Am. Chem. Soc. 1972, 94, 660. (b) Bergman, R. G. Acc. Chem. Res. 1973, 6, 25. (c) Darby, N.; Kim, C. U.; Shelton, K. W.; Takada, S.; Masamune, S. J. Chem. Soc. (D) 1971, 23, 1516. (d) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-9. (e) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005, pp 90–91.

161

162

Schmittel, M.; Kiau, S. Chem. Lett. 1995, 953.

163

O’Connor, J. M.; Lee, L. I.; Gantzel, P. J. Am. Chem. Soc. 2000, 122, 12057.

(a) Nicolaou, K. C.; Stabila, P.; Esmaeli-Azad, B.; Wrasidlo, W.; Hiatt, A. Proc. Natl. Acad. Sci. USA 1993, 90, 3142. (b) Zein, N.; Solomon, W.; Casazza, A. M.; Kadow, J. F.; Krishnan, B. S.; Tun, M. M.; Vyas, D. M.; Doyle, T. W. Bioorg. Med. Chem. Lett. 1993, 3, 1351.

164

947

17.8 METAL-INDUCED RADICAL REACTIONS

R1

•

R2

•

167

R1

R1

R2

R2

168

169

From a purely synthetic viewpoint, the radicals generated by the Bergman cyclization have been used to initiate further cyclization, including the preparation of aromatic compounds. In one example, Jones and Plourde165 prepared a series of chlorinated ene-diynes and used the Bergman cyclization to convert them to chlorinated aromatic compounds. Heating (E)-3-chlorocyclodeca-3-en-1,5-diyne (170), for example, gave a 40% yield of 5,6,8-trichloro-1,2,3,4-tetrahydronaphthalene.165 Interestingly, heating (E)-3-chlorocyclodeca-3-en-1,5-diyne with tetramethylpiperidinyloxy free radical (TEMPO; also see Section 6.2.3) led to a 30% yield of the chlorinated benzoquinone 2-chloro-5,6,7,8-tetrahydronaphthalene-1,4-dione. The additional chorine atoms in 5,6,8-trichloro-1,2,3,4tetrahydronaphthalene are attributed to atom-transfer chemistry from CCl4 to the diradical intermediate. This reaction is not limited to carbocyclic compounds, and an aza-Bergman rearrangement reaction has been reported.166 O

Cl Cl

Cl

CCl4, 60°C

TEMPO

Cl

PhH, 60°C

Cl 5,6,8-Trichloro-1,2,3,4- (40%) tetrahydronaphthalene

170

O 2-Chloro-5,6,7,8-tetrahydronaphthalene-1,4-dione

(30%)

17.8 METAL-INDUCED RADICAL REACTIONS 17.8.1 General Principles Metal-induced radical reactions can take at least two forms: (1) electron transfer of a metal to a peroxide, leading to reactive free radicals, and (2) direct electron transfer from a metal to an alkene, carbonyl, or aromatic substrate. An example of the first type of reaction is oxidation with Fenton’s reagent (Section 6.3.2),167 which initiates oxidation of aryls to phenols, although the yields are usually only 5–20%. In this reaction, ferrous ion (Fe2+) is oxidized to ferric ion (Fe3+) by H2O2, which gives hydroxide and hydroxyl radicals (HO•).167 The presence of the radicals initiates a chain process, which can be terminated by reaction of HO• with Fe2+. Oxidation of benzene with Fenton’s reagent gave only a small amount of phenol, but large amounts of biphenyl arising from coupling of the aryl radicals. Another metal system is Ti3+dH2O2 (at pH 2), which generates HO• and HOO• radicals.168 Transition metals (e.g., Ti, Co, or Mn) are often used in conjunction with hydroperoxides, and decompose under these conditions to alkoxy radicals (RO•) and alkyl peroxyls (ROO•).167a,169 In all cases, reaction with the metal generates a radical, which then reacts with the carbon fragment to induce the desired radical reaction. The latter metal-peroxide reaction was discussed previously in Section 6.4.2 in connection with the transition metal-peroxide oxidation of alkenes to oxiranes.170

17.8.2 Phenolic Oxidative Coupling The phenolic oxidative coupling is a classical oxidation reaction that involves radicals,171 and it was introduced in Section 8.5.1.1. In this reaction, electron transfer from a metal salt to a bis(phenol) leads to intramolecular coupling 165

Jones, G. B.; Plourde, II, G. W. Org. Lett. 2000, 2, 1757.

166

Feng, L.; Kumar, D.; Kerwin, S. M. J. Org. Chem. 2003, 68, 2234.

167

(a) Haines, A. H. Methods for the Oxidation of Organic Compounds; Academic Press: London, 1985; p 173. (b) Smith, J. R. L.; Norman, R. O. C. J. Chem. Soc. 1963, 2897. 168

(a) Reference 4, p 31. (b) Irvine, M. J.; Wilson, I. R. Aust. J. Chem. 1979, 32, 2131, 2283.

169

Black, J. F. J. Am. Chem. Soc. 1978, 100, 527.

170

Hiatt, R. In Oxidation; Augustine, R. L.; Trecher, D. J., Eds., Marcel-Dekker: New York, 1971; pp 117–124, Vol. 2.

171

(a) Barton, D. H. R.; Deflorin, A. M.; Edward, O. E. J. Chem. Soc. 1956, 530. (b) Scott, A. I. Q. Rev. 1965, 19, 1.

948

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

and a quinone product. In early experiments, yields were poor. For example, Barton and Kirby reacted 171 with potassium ferricyanide [K3Fe(CN)6] and the initial product was aryl radical 172. This radical reacted with the second phenolic ring in an intramolecular process that gave 173. Loss of a hydrogen atom led to the quinone-containing narwedine, as part of a synthesis of galanthamine.172 The yield of narwedine, however, was only 1.4% under optimal conditions. Kametani et al.173 improved the yield of the oxidation only slightly (5% yield) by using an amide derivative of 171, in a synthesis of galanthamine. OH

HO

K3[Fe(CN) 6]

MeO

H

HCO3

•

NaHCO3

N

•

O

MeO

O

OH

OH

N

HO N

Me 171

172

Me

Me

O

Me

N

MeO

MeO 173

Narwedine

A similar oxidation procedure was used by Gervay et al.174 to generate erythrinadienone in 35% yield, which constituted a great improvement. Yields remained relatively low, however, due to the shortness of the tether between the two aryl moieties, or due to the steric hindrance provided by the hydroxyl and alkoxyl groups.175 Indeed, success for phenolic oxidative coupling is dependent on the nature of the substituents on the aromatic rings, as well as the length of the tether.176 Other oxidative methods can be used, including anodic oxidation under electrolysis conditions. Good yields can be obtained in such systems, as in the anodic oxidation of tetramethoxy derivative 174 to give 175 in 80% yield, where 175 was a key intermediate in Tobinaga coworkers175 synthesis of colchicine. MeO

MeO e– (SCE), MeCN

OMe

MeO OMe

OH

MeO MeO

Anodic oxidation HBF4, 20 min SCE = Saturated calomel electrode

174

OMe O 175 (80%)

In another variation, Tang and coworkers177 used VOF3 to convert 176 to 177 in 32% yield, which was deprotected using hydrogenation to yield decinine in 80%. Phenyliodine(III)bis(trifluoroacetate) has also been used to effect phenolic oxidative coupling,178 as have other hypervalent iodine(II) reagents.179 In a synthesis of (+)-puupehenone, Quideau, et al.180 treated 178 with [bis(trifuoroacetoxy)iodo]benzene and obtained a 67% yield of 179.

172

Barton, D. H. R.; Kirby, G. W. J. Chem. Soc. 1962, 806.

(a) Kametani, T.; Shishido, K.; Hayashi, E.; Seino, C.; Kohno, T.; Shibuya, S.; Fukumoto, K. J. Org. Chem. 1971, 36, 1295. (b) Kametani, T.; Seino, C.; Yamaki, K.; Shibuya, S.; Fukumoto, K.; Kigasawa, K.; Satoh, F.; Hiiragi, M.; Hayasaka, T. J. Chem. Soc. C 1971, 1043. (c) Kametani, T.; Yamaki, K.; Yagi, H.; Fukumoto, K. Ibid. 1969, 2602.

173

(a) Gervay, J. E.; McCapra, F.; Money, T.; Sharma, G. M.; Scott, A. I. J. Chem. Soc. Chem. Commun. 1966, 142. (b) Mondon, A.; Ehrhardt, M. Tetrahedron Lett. 1966, 2557.

174

175

(a) Kotani, E.; Miyazaki, F.; Tobinaga, S. J. Chem. Soc. Chem. Commun. 1974, 300. (b) Tobinaga, S.; Kotani, E. J. Am. Chem. Soc. 1972, 94, 309.

176

See, for example, Krauss, A. S.; Taylor, W. C. Aust. J. Chem. 1992, 45, 925, 935.

177

Shan, Z.-H.; Liu, J.; Xu, L.-M.; Tang, Y.-F.; Chen, J.-H.; Yang, Z. Org. Lett. 2012, 14, 3712.

178

Kita, Y.; Gyoten, M.; Ohtsubo, M.; Tohma, H.; Takada, T. Chem. Commun. 1996, 1481.

179

(a) Quideau, S.; Pouysegu, L. Org. Prep. Proc. Int. 1999, 31, 617. (b) Quideau, S.; Pouysegu, L.; Oxoby, M.; Looney, M. A. Tetrahedron 2001, 57, 319.

180

Quideau, S.; Lebon, M.; Lamidey, A.-M. Org. Lett. 2002, 4, 3975.

949

17.8 METAL-INDUCED RADICAL REACTIONS

O

O

O

O 10 VOF3

N

100 TFA CH2Cl2, 0°C

PNBO MeO

N PNBO

OMe

OMe

MeO

176

177 (32%) HO

OH

O

OH PhI(O2CCF3)2

O

CH2Cl2, –25°C

OH

H

H

179 (67%)

178

Note that many phenolic coupling reactions may proceed via radical cation intermediates and/or phenoxonium ions, especially when the coupling is induced by electrochemical methods. In such cases, it is not always clear that the mechanism involves a radical-radical coupling reaction. The phenolic oxidative coupling disconnection follows: (CH 2)n

(CH 2)n

HO MeO

O

OH

17.8.3 Pinacol Coupling A classical metal-transfer reaction is the pinacol reaction or pinacol coupling,181 in which alkali metals react with a ketone (e.g., pentan-3-one) to produce a radical anion (ketyl 180) via electron transfer (Sections 7.11.1.2).182 Other reagents have been developed that yield pinacol coupling,183 including samarium iodide,184 titanium(IV) iodide,185 or TiCldBu4NI.186 In another study, lanthanide metals were shown to be effective for promoting pinacol coupling, in the presence of chlorotrimethylsilane.187 If a monovalent metal (e.g., Na or K) is used in the presence of an alcohol (ethanol is the most common), the carbonyl is reduced to the alcohol (pentan-3-one ! pentan-3-ol, e.g., Section 7.11.2). If a bivalent metal (e.g., Mg) is used, however, the ketyl (the radical anion) is stabilized to a greater extent, and reduction is slow. In such a case, the ketyl behaves more like a radical, leading to the coupling product (181). Hydrolysis leads to the diol, 3,4-diethylhexane-3,4-diol.181a,188 This reaction with ketones or aldehydes is a preferred method for the synthesis of 1,2-diols.188,189

(a) Schreibman, A. A. P. Tetrahedron Lett. 1970, 4271. (b) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-74. (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 512–513. 181

182

For reactions of this type, see Larock, R. C., Comprehensive Organic Transformations; Wiley-VCH: New York, 1999; pp 1111–1114.

183

Wirth, T. Angew. Chem. Int. Ed. 1996, 35, 61.

184

Ghatak, A.; Becker, F. F.; Banik, B. K. Tetrahedron Lett. 2000, 41, 3793.

185

Hayakawa, R.; Shimizu, M. Chem. Lett. 2000, 724.

186

Tsuritani, T.; Ito, S.; Shinokubo, H.; Oshima, K. J. Org. Chem. 2000, 65, 5066.

187

Ogawa, A.; Takeuchi, H.; Hirao, T. Tetrahedron Lett. 1999, 40, 7113.

188

Popp, F. D.; Schultz, H. P. Chem. Rev. 1962, 62, 19 (see pp 27–30).

189

Weber, J. E.; Boggs, A. D. J. Chem. Educ. 1952, 29, 363.

950

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

O

O

Pentan-3-one

180

H3O+

Mg

•

M

O

O

HO

OH

Na EtOH

181

H

3,4-Diethylhexane-3,4-diol

OH Pentan-3-ol

An asymmetric pinacol coupling of aromatic aldehydes was reported by Enders and Ullrich,190 using TiCl2 and an enantiopure amine or hydrazine reagent. When benzaldehyde was treated with TiCl2 and a chiral amine, (S)-2-(methoxymethyl)pyrrolidine, a 31% yield of 1,2-diphenylethane-1,2-diol was obtained as an 81:19 mixture of dl/mesodiastereomers, and the dl-diastereomer was obtained in 65 %ee (S,S).188 When the reaction time was increased to 48 h, the yield improved to near quantitative, but the diastereomeric ratio was 56:44 dl/meso and the enantioselectivity was only 37 %ee. N H

PhCHO

OH

OMe

Ph

Ph TiCl2, THF, 23 h, –78°C

OH 1,2-Diphenylethane-1,2-diol (31%)

The preferred method for reductive coupling uses a mixture of magnesium and magnesium iodide (Mg + MgI2), although magnesium amalgam (Mg/Hg) can also be used.191 A classical example is reductive condensation of cyclopentanone to the 1,2-glycol,[1,10 -bi(cyclopentane)]-1,10 -diol.181a 2,3-Dimethylbutane-2,3-diol (also known as pinacol) is formed by pinacol reduction of acetone and gives the reaction its name.181a,188,189 Many other reducing agents can be used for this condensation reaction in addition to Mg or Na. Samarium iodide (SmI2), mentioned above, was used by Nicolaou and et al.192 for the intramolecular pinacol coupling of keto-aldehyde 182 to give diol 183 in 50% yield, in a synthesis of echinopine A. OH

Mg, EtOH

O HO [1,1⬘-Bi(cyclopentane)]-1,1⬘-diol

H

H

SmI2, HMPA

O OTBS OHC 182

HO OTBS HO H

183 (50%)

Another lanthanide, ytterbium (Yb), has been used as the free metal for the reduction of ketones, but extensive coupling was observed. The extent of coupling depended on the ratio of Yb to ketone.193 A related ketyl-coupling reaction is the Molander et al.194 conversion of 184 to the samarium-stabilized ketyl 185. Conjugate radical addition to the acrylate moiety gave 186, and hydrolysis led to an 88% yield of 187 (200:1 diastereoselectivity). A catalytic and enantioselective pinacol coupling was reported that used chiral Ti(III) complexes, where the chiral ligand was a salen [bis (salicylidine)ethylenediamine] compound.195

190

Enders, D.; Ullrich, E. C. Tetrahedron: Asymmetry 2000, 11, 3861.

191

Adams, R.; Adams, E. W. Org. Synth. Coll. 1941, 1, 459.

192

Nicolaou, K. C.; Ding, H.; Richard, J.-A.; Chen, D. Y.-K. J. Am. Chem. Soc. 2010, 132, 3815.

193

Hou, Z.; Takamine, K.; Aoki, D.; Shiraishi, H.; Fujiwara, U.; Taniguchi, H. J. Org. Chem. 1988, 53, 6077.

194

(a) Molander, G. A.; Kenny, C. J. Am. Chem. Soc. 1989, 111, 8236. (b) Molander, G. A.; Etter, J. B.; Zinke, P. W. Ibid. 1987, 109, 453.

195

Chatterjee, A.; Bennur, T. H.; Joshi, N. N. J. Org. Chem. 2003, 68, 5668.

951

17.8 METAL-INDUCED RADICAL REACTIONS

EtO2C O

Me SmI2

Me

EtO Me

EtO Me

O

O

SmO

CO2Et 184

Me EtO2C

Me SmI2

Sm O Me

185

EtO2C HO

H3O+

Me

CO2Et

186

CO2Et

187 (88%)

The pinacol coupling reaction can be applied to most ketones and aldehydes when two identical molecules are being coupled. An improvement in the reaction involves trapping the initial coupling product as the bis(trimethylsilyl) derivative. Coupling of 40 -methylacetophenone with Zn196 in the presence of chlorotrimethylsilane gave 76% of the bis(O-silyl) ether (188). Treatment with tetrabutylammonium fluoride removed the silyl protecting group (Section 5.3.1.2) to yield the pinacol product 2,3-di-p-tolylbutane-2,3-diol. Coupling of two different ketones may lead to several different products in relative proportions close to those expected from statistical coupling. One product may predominate, but the yields can be low. When cyclopentanone was coupled to cycloheptanone in the presence of aluminum and mercuric chloride (HgCl2), for example, only 23% of the cross-coupled diol, 1-(1-hydroxycyclopentyl)cycloheptan-1-ol, was formed.197 O

Me3SiCl, Zn

Me

OSiMe3 Me

Me

aq Bu4NF

Me

Dioxane, 2 h

Me

OH Me

Me

Me HO

Me3SiO Me

4⬘-Methylacetophenone

2,3-Di-p-tolylbutane-2,3-diol

188 (76%)

+

O

OH

HgCl2, Al, 2 h

O PhH, Reflux

Cyclopentanone

Me

HO 1-(1-Hydroxycyclopentyl)cycloheptan-1-ol

Cycloheptanone

The pinacol disconnection follows: R2C(OH)JC(OH)R2

R2CHKO

Photoreductive coupling of aldimines leads to the synthesis of C2 symmetrical diamines. An example is the photocoupling of aldimine (189) to a 52:48 mixture of 191/192 in 85% yield.198

NH

hv (Pyrex)

N

i-PrOH/Acetone

NH

HN

HN

+ N

N

189

N

N

N 190

191

17.8.4 The Acyloin Condensation A variation of the pinacol reaction has been applied to esters for the preparation of α-hydroxy ketones (acyloins). A simple example is the reaction of dimethyl heptanedioate with sodium metal (Na) in refluxing xylene,199 which gave a good yield of 2-hydroxycycloheptan-1-one after hydrolysis. In 1905, Bouveault and Loquin200 reported the 196

So, J.-H.; Park, M.-K.; Boudjou, P. J. Org. Chem. 1988, 53, 5871.

197

(a) Sands, R. D. Tetrahedron, 1965, 21, 887. (b) Sands, R. D.; Botteron, D. G. J. Org. Chem. 1963, 28, 2690.

198

Campos, P. J.; Arranz, J.; Rodríguez, M. A. Tetrahedron 2000, 56, 7285.

(a) McElvain, S. M. Org. React. 1948, 4, 256. (b) Finley, K. T. Chem. Rev. 1964, 64, 573. (c) House, H. O. Modern Synthetic Reactions, 2nd ed.; W. A. Benjamin: Menlo Park, CA, 1972, pp 169–172.

199

200

Bouveault, L.; Loquin, R. Compt. Rend. 1905, 140, 1593.

952

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

condensation of a α,ω-diester to an α-hydroxy ketone,201 the acyloin condensation.199,202 The reaction works well with relatively high molecular weight esters in an intermolecular reaction. Methyl laurate (methyl dodecanoate), for example, was treated with Na (in xylene at 110°C) to give a 90% yield of lauroin (13-hydroxytetracosan-12-one).203 The reaction conditions shown are typical, where high-dilution techniques (Section 6.6.2) were not required, but effective stirring (high-speed stirrer at 2000–2500 rpm) was necessary. O

O

MeO

OMe

2. H3O+

OH 2-Hydroxycycloheptan-1-one

Dimethyl heptanedioate

C11 H23

O

1. Na, Xylene, Reflux

1. Na, Xylene, 110°C Mechanical stirrer (2000–2500 rpm)

CO2Me

O C11 H23

C11 H23

2. H3O+

OH Lauroin (90%) (13-hydroxytetracosan-12-one)

Methyl dodecanoate

The acyloin condensation proceeds204 via initial transfer of an electron to form a ketyl (193), which is in equilibrium with the coupling product (192) and the diketone (194). Transfer of a second electron to this mixture led to 195, and protonation generated a dihydroxy intermediate (196), which gave acyloin 197. Radical coupling of the ketyl intermediate is most efficient when the ketyl is bound to, or closely associated with the surface of the metal. The high-speed stirrer generates a fine “sodium sand,” which maximizes the surface area for effective electron transfer and radical coupling of the ketyl. Under these conditions, both intermolecular coupling and intramolecular cyclization reactions are possible. The preparation of 197 is one example of an intermolecular acyloin condensation. The intramolecular cyclization of α,ω-diesters (198) to form cyclic acyloins (199) appears to be the most efficient version of this reaction. This reaction does not require high dilution, and because of the template effect arising from reaction on the surface of sodium, good yields can be obtained even with medium-sized rings. As shown in Table 17.1,205 the yields of acyloin (199) from simple diesters199a or diesters containing heteroatoms or aromatic “spacers”199a are quite good, even when n + m in 198 is large. Introduction of double and triple bonds is also possible if the ring size of the final product is large.199b Amino and aryl functionality can be also incorporated (as in the paracyclophanes).199a O Na O

Na

2 R

OR1

R • OR1

R

O Na 192

193

+ 2 NaOR1

O Na 2

R

O 194

O Na

O

2

OR1

R

R

O Na

OR1

R

Na

+

OR1

R Na R

O Na

R

OH

R

R

O Na

R

OH

R

O

2 NaOR1

195

196

OH 197

Reprinted with permission from McElvain, S. M. Org. React. 1948, 4, 256. Copyright © 1948 by John Wiley & Sons, Inc.

201

For transformations of this type, see Ref. 182, pp 1313–1315.

(a) Bloomfield, J. J.; Owsley, D. C.; Nelke, J. M. Org. React. 1976, 23, 259; (b) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-1. (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience, NJ, 2005; pp 4–5.

202

203

(a) Hansley, V. L. J. Am. Chem. Soc. 1935, 57, 2303. (b) Ruzicka, L.; Plattner, P. A.; Widmer, W. Helv. Chim. Acta 1942, 25, 604, 1086.

204

Reference 199a, pp 257 and 259.

205

Finley, K. T. Chem. Rev. 1964, 64, 573.

953

17.8 METAL-INDUCED RADICAL REACTIONS

TABLE 17.1

X

Intramolecular Acyloin Condensations of 198

(CH2)m CH2CO2Me

1. Na, Xylene

(CH2)n CH2CO2Me

2. H3O+

X

(CH2)m

O

(CH2)n

OH

198 X

m

n

CH2

2 2 2 3 4 4 5 5 6 6 7 13 2 3 3 2 2 1 2 2 3 3 2 2 6 5 4 3 2 1

1 2 3 4 4c 5c 5c 6c 6c 7c 8c 14 c 2 3 4 2 2 1 2 2 3 3 2 2 6 5 4 4 5 6

MeCH CLC cis-CHKCH trans-CHKCH PhJN MeJN EtJN C6H11JN p-MeJPhJN 1,4-C6H4

a b c

199

Reference 199b. Reference 199a. Diethylester.

Ring Size

% 199

8 9 10 12 13 14 15 16 17 18 20 34 9 12 11 10 10 7 9 9 11 11 9 9 20 18 16 15 15 15

47 a 9b 45b 76b,c 72a 79b 77b 84 b 85 b 96b 96b 20a 60a 73a 0a 80 a 51a 10 a 3a 75a 64a 85 a 50a 38 a 0a 70a 76a 36 a 0a 75a

Reprinted with permission from Finley, K. T. Chem. Rev. 1964, 64, 573. Copyright © 1964 American Chemical Society.

For simple diesters, the yield was reasonable for medium- and large-ring compounds. Alkenyl derivatives gave good yields if the ring size was nine or greater, where the large ring size can accommodate flattening of the molecule due to the presence of the double bond (see Section 1.5.4). The cis-alkene derivatives could be formed with even small rings, but an eight-membered ring or larger was usually necessary for a trans-alkene. Even larger rings are required for derivatives containing an alkyne moiety. The two examples in Table 17.1 suggest that only 12-membered rings and larger can be formed if 199 contains an alkynyl group. The presence of an amino group is compatible with good yields for a variety of ring sizes, although the substituent on nitrogen plays a role. As expected, the paracyclophanes are more rigid, and only relatively large rings are formed in good yield (15–18 membered).206 The 20-membered ring (m ¼ n ¼ 6), however, gave no acyloin under these conditions. A modification of the acyloin condensation adds chlorotrimethylsilane to trap the alkoxide intermediate as a bis(silyl enol ether). This modification has become the standard version of the acyloin condensation. Schr€ apler and (a) Cram, D. J.; Antar, M. F. J. Am. Chem. Soc. 1958, 80, 3109. (b) Cram, D. J.; Daeniker, H. U. Ibid. 1954, 76, 2743. (c) Cram, D. J.; Cordon, M. Ibid. 1955, 77, 4090.

206

954

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

R€ uhlmann207 showed that ethyl butanoate reacted with excess Na to generate the expected bis(alkoxide) (200). Addition of chlorotrimethylsilane led to a 92% yield of the bis(trimethylsilyl)derivative, (201), and hydrolysis with 1 N HCl in THF gave 96% of the acyloin (5-hydroxyoctan-4-one). This modification serves two purposes. First, trapping the enolate product allows its isolation and minimizes side reactions. Second, the O-TMS group protects the acyloin (see Section 3.1.2), allowing further manipulation prior to unmasking the relatively sensitive α-hydroxyketone moiety. When there are no enolizable hydrogen atoms in the O-TMS derivative, treatment with bromine generates a 1,2-diketone.208 When enolizable hydrogen atoms are present the normal acyloin product is observed. A synthetic example is taken from Sih and coworker’s209 synthesis of castanospermine, in which diester 202 derived from hydroxy-proline was refluxed with Na metal and chlorotrimethylsilane in toluene. This reaction led to formation of 203, which could be isolated by chromatography, and converted to the corresponding acyloin by mild acid hydrolysis.

2

4 Na

CO2Et

4h

C3H7

O

C3H7

O

C3H7 C3H7

MeO2C

THF

OSiMe3

TMSO TMSO Na, PhMe

N

O

C3H7

OH

5-Hydroxyoctan-4-one (96%)

201 (92%)

H OTBDMS

C3H7 1 N HCl, 1 h

200

MeO2C

OSiMe3

2 Me3SiCl

H

OTBDMS

TMS-Cl

N

202

203

McMurry and Rico210 generated an acyloin-like reaction using a titanium trichloride (TiCl3)-Zn/Cu couple. α,ωDialdehydes are coupled with this reagent to yield 1,2-diols. This procedure generates mixtures of cis- and trans-diols with the cis predominating for smaller rings and the trans predominating for larger rings.211 Cyclization of hexane-1,6dial [OHCd(CH2)4dCHO] under these conditions gave only cis-cyclohexane-1,2-diol in 85% yield, but cyclization of dodecane-1,12-dial [OHCd(CH2)10dCHO] gave a 25:75 cis-trans mixture of 1,2-cyclododecanediol211 in 75% yield. As with the acyloin condensation, good yields are realized even with medium size (8–13 membered) rings. An example, taken from Williams and Heidebrecht’s synthesis of (+)-4,5-deoxyneodolabelline,211 cyclized 204 with TiCl3 and Zn/Cu (see Section 13.8.6) to give an 8:2:1:1 (R,R; S,R; R,S; S,S) mixture of 205 in 85% yield.

H

TiCl3, 80°C Zn/Cu, DME

H

O

O CHO

O

OH 204

OH

205 (85%)

The acyloin disconnections follow: O

OH R

R

R CO2

R1

O

CO2R OH CO2R

207

(a) Schr€ apler, U.; R€ uhlmann, K. Chem. Ber. 1963, 96, 2780. (b) R€ uhlmann, K.; Poredda, S. J. Prakt. Chem. 1960, 12, 18.

208

Strating, J.; Reiffers, S.; Wynberg, H. Synthesis 1971, 209.

209

Bhide, R.; Mortezaei, R.; Scilimati, A.; Sih, C. J. Tetrahedron Lett. 1990, 31, 4827.

210

McMurry, J. E.; Rico, J. G. Tetrahedron Lett. 1989, 30, 1169.

211

Williams, D. R.; Heidebrecht, R. W., Jr. J. Am. Chem. Soc. 2003, 125, 1843.

955

17.8 METAL-INDUCED RADICAL REACTIONS

17.8.5 McMurry Olefination A useful relative of the acyloin condensation is the McMurry olefination reaction.212 In this reaction, ketones or aldehydes are treated with Ti(0) (TiCl3 + LiAlH4; TiCl3 + K; TiCl3 + Li)213 to yield alkenes.214 Reductive coupling is possible with a wide variety of ketones and aldehydes.213a,215 Retinal, for example, was coupled with this reagent to give β-carotene in 85% yield.215 The TiCl3-K mixture is a commonly used reagent, and a variety of carbonyl compounds can be coupled to give the corresponding alkene.213a,216 Treatment of decanal with TiCl3 and K led to a 60% yield of the C20 alkene (icos-10ene).216 Cross-coupling is also possible as illustrated by the treatment of a mixture of cycloheptanone and acetone with TiCl3-Li to give a 50% yield of propan-2-ylidenecycloheptane, along with 26% of 1,10 -bi(cycloheptylidene).216

CHO

LiAlH4•TiCl3

O

2

β-Carotene

Retinal

O

Me

+

Me

TiCl3, 3 equiv. Li

Propan-2-ylidenecycloheptane (50%)

1,1⬘-bi(cycloheptylidene)

The mechanistic studies of this coupling suggest it occurs “in a heterogeneous process on the surface of an active titanium particle,” as outlined in Fig. 17.6.214 The byproduct is titanium dioxide (TiO2) and electron transfer to the carbonyl generates the dianion (206), presumably via a ketyl-like intermediate. It has been suggested that the metallopinacol is not the only precursor, and that carbenoid intermediates and/or nucleophilic intermediates can be important.216 In the commonly accepted mechanism, the initially formed dianion binds to the titanium particle as shown, and sequential homolytic cleavage (via 207) liberates the alkene and TiO2. McMurry’s experiments discredited alternative mechanisms (e.g., formation of a discrete cyclic or acyclic) titanium dialkoxy species.213a This reagent also led to elimination of 1,2-diols to alkenes, as in the conversion of trans-cycloheptane-1,2-diol to cycloheptene in 55%.213a,217

FIG. 17.6 Mechanism of the McMurry coupling reaction. Reprinted with permission from McMurry, J. E.; Fleming, M. P.; Kees, K. L.; Krepski, L. R. J. Org. Chem. 1978, 43, 3255. Copyright © 1978 American Chemical Society.

O R

Ti

R

R

R

O– O– R

R

Ti

R

O O R

R R

206 R O Ti

O•

R

R • R

Ti

O• O•

R

R

R

R

+

207

(a) McMurry, J. E. Acc. Chem. Res. 1974, 7, 281. (b) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-58. (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 416–417. 212

213

McMurry, J. E.; Fleming, M. P.; Kees, K. L.; Krepski, L. R. J. Org. Chem. 1978, 43, 3255.

214

For transformations of this type, see Ref. 182, pp 305–308.

215

McMurry, J. E.; Fleming, M. P. J. Am. Chem. Soc. 1974, 96, 4708.

For discussions of the mechanism of this reaction, see (a) Stahl, M.; Pidun, U.; Frenking, G. Angew. Chem. Int. Ed. 1997, 36, 2234. (b) Villiers, C.; Ephritikhine, M. Angew. Chem. Int. Ed. 1997, 36, 2380. 216

217

McMurry, J. E.; Fleming, M. P. J. Org. Chem. 1976, 41, 896.

956

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

This reaction can be done intramolecularly and Zn/Cu, TiCl3 is a common catalyst, although Zn and TiCl4 are also used.214 The final product of the intramolecular reaction is a cyclic alkene, and rings ranging from 4 to 22 members were formed, with different alkyl substituents, in 65–90% yield.214 Since the reaction proceeds via a Ti bound ketyl species, the two reactive ends are in close proximity, allowing efficient coupling of even medium and large rings analogous to the acyloin condensation.218 Medium- and large-ring alkenes are readily prepared from both diketones and dialdehydes.214 In a synthetic example, Corey and Hu219 converted diketone 208 to δ-araneosene in 90% yield using McMurry coupling. Note that the McMurry coupling of esters leads to enol ethers.

O

TiCl3, Zn/Cu

O

DME, Reflux

d-Araneosene (90%)

208

The disconnections for McMurry coupling follows:

R

R1

R O +

R

R1

R

OH

R1

O

R

R1

O R1

OH

R1

R O

17.9 CARBENES AND CARBENOIDS Carbenes constitute another class of molecules that in many ways can be considered to be radical-like. Carbenes have been known since the work of Curtius and Buchner220 in the late 19th century and Staudinger and Kupfer221 in the early 20th century, and other work222 is known. Carbenes are important in several synthetic methods and are growing in importance, especially the intramolecular versions. N-Heterocyclic carbenes are important as universal ligands in organometallic and inorganic coordination chemistry.223 “N-Heterocyclic carbenes both stabilize and activate metal centers in quite different key catalytic steps of organic syntheses.”224 The remaining sections of this chapter discuss the preparation and reactions of carbenes and related compounds.

17.9.1 Definition of a Carbene A carbene is a divalent carbon species linked to two adjacent groups by covalent bonds,224 possessing two nonbonded electrons and six valence electrons (as in 209).225 The two nonbonded electrons can have anti-parallel spins in a single orbital (see 210, a singlet carbene), or parallel spins in different orbitals (see 211, a triplet carbene). Substituted carbenes (212) are generated in the singlet state from most precursors, but “intersystem crossing to the triplet (Section 15.2.2) may or may not occur before the individual carbene reacts with a suitable precursor.”226 “Singlet carbenes are electron-deficient species comparable to carbocations,” but “they possess a nonbonding pair of electrons 218

McMurry, J. E.; Kees, K. L. J. Org. Chem. 1977, 42, 2655.

219

Hu, T.; Corey, E. J. Org. Lett. 2002, 4, 2441.

220

Buchner, E.; Curtius, T. Ber. Dtsch. Chem. Ges. 1885, 8, 2377.

221

Staudinger, H.; Kupfer, O. Ber. Dtsch. Chem. Ges. 1912, 45, 501.

222

Arduengo, III, A. J.; Krafczyk, R. Chem. Unserer Z . 1998, 32, 6.

223

Herrmann, W. A. Angew. Chem. Int. Ed. 2002, 41, 1291.

224

For a discussion of stable carbenes, see Arduengo, III, A. J. Acc. Chem. Res. 1999, 32, 913.

(a) Kirmse, W. Carbene Chemistry; Academic Press: New York, 1971; pp. 5–7. (b) Chinoporos, E. Chem. Rev. 1963, 63, 235. (c) Closs, G. L. Top. Stereochem. 1968, 3, 193. (d) Parham, W. E.; Schweizer, E. E. Org. React. 1964, 13, 55. (e) Brinker, U. H., Ed., Advances in Carbene Chemistry; Jai Press: Greenwich and Stamford, CT, 1994 and 1998; Vol. 1 and Vol. 2. (f ) Regitz, M. Angew. Chem. Int. Ed. 1996, 35, 725.

225

226

Reference 225a, p 5.

957

17.9 CARBENES AND CARBENOIDS

comparable to that of carbanions.”226a If an adjacent group on the carbene withdraws electrons, the carbene will have electrophilic character. If that group supplies electrons, the carbene will have nucleophilic character. “Triplet carbenes may be considered as diradicals, although the interaction of two unpaired electrons in orbitals on the same carbon atom gives rise to some peculiarities.”227 “The term carbenoid was suggested” by Closs and Moss “for the description of intermediates that exhibit reactions qualitatively similar to those of carbenes without necessarily being free divalent carbon species.”227 Although generally considered to be reactive intermediates, stable carbenes are known, particularly persistent triplet diarylcarbenes and heteroatom-substituted singlet carbenes.228 H

H C

H

R

C

H

H 209

C

C R

H 210

211

212

17.9.2 Preparation of Carbenes There are several methods for the preparation of carbenes. A simple approach is to photolyze ketenes (ketene itself, known as ethenone) at 270–370 nm [100.5–76.5 kcal (438.8–320.2 kJ) mol1] in the presence of oxygen.229 Irradiation of ketene at 270 nm gave CO (with a quantum yield ¼ 2) and the yield of ethene was independent of the pressure of oxygen. At 370 nm, the quantum yield of CO varies with the pressure of applied oxygen and also with the temperature. The presence of additional oxygen diminished the yield of ethene by about two-thirds.229 The reaction proceeded via formation of excited ketene (ethenone), which fragments to methylene (209) and CO. The reactive carbene reacts with additional ketene to yield ethene and CO. Note that a possible intermediate to 213 is an excited cyclopropanone species.230 Photolysis of alkyl and aryl ketenes generates alkyl and aryl carbenes. Methylketene and dimethylketene led to methylcarbene231 and dimethylcarbene,232 respectively, upon irradiation. The quantum yields were close to unity at 270 nm.233 Methylcarbene coupled with itself to form an activated ethene, which gave both ethene and ethyne.232 But-2-ene was also formed by reaction with methyl carbene and the starting ketene. Similarly, photolysis of dimethylketene led to cyclopropanone (via dimethylcarbene and the ketene), which lost CO under the photolytic conditions to yield 2,3-dimethylbut-2-ene.233 H CKCKO

hv

H

*

H

H

209

213

H CKCKO

209

CLO

H

H Ketene

+

C

CKCKO

CH2KCH2

+

CLO

H Ketene

17.9.2.1 Diazomethane Diazoalkanes are important precursors for the preparation of carbenes and related molecules. Although diazomethane is not used in carbene preparations, it has other synthetic applications, and is presented here for comparative purposes. There are four convenient preparations of diazomethane, which is a highly reactive gas. Since diazomethane can detonate on contact with ground glass, its preparation requires the use of specialized glassware. Diazomethane is also toxic.

227

Reference 225a, p 6.

228

Bourissou, D.; Guerret, O.; Gabbai, F. P.; Bertrand, G. Chem. Rev. 2000, 100, 39.

229

Reference 225a, p 9.

(a) Knox, K.; Norrish, R. G. W.; Porter, G. J. Chem. Soc. 1952, 1477. (b) Kistiakowsky, G. B.; Sauer, K. J. Am. Chem. Soc. 1956, 78, 5699. (c) Idem Ibid. 1958, 80, 1066.

230

231

(a) Kistiakowsky, G. B.; Mahan, B. H. J. Am. Chem. Soc. 1957, 79, 2412. (b) Chong, D. P.; Kistiakowsky, G. B. J. Phys. Chem. 1964, 68, 1793.

232

Halroyd, R. A.; Blacet, F. E. J. Am. Chem. Soc. 1957, 79, 4830.

233

Reference 225a, p 14.

958

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

All preparative methods for diazomethane (CH2N2) involve N-nitroso compounds. Treatment of commercially available N-methyl-N-nitroso-p-toluenesulfonamide (Diazald, 214) with hydroxide liberates diazomethane and a sulfonate ester, ethyl 4-methylbenzenesulfonate.234 Similar treatment of N,N0 -dimethyl-N,N0 -dinitrosoterephthalamide (215),235 N-methyl-N0 -nitro-N-nitrosoguanidine (MNNG, 216),236 or N-methylnitrosourea (216)237 give diazomethane. Note that MNNG is allergenic and is potentially mutagenic. These reagents and synthetic uses of diazomethane are outlined in an excellent review by Hopps.238 Note that at least two other structural isomers of diazomethane are known: diazirine239 and isodiazomethane.240 O Me

S

NO N

O

O

KOH, EtOH

+

CH2N2

Ether

Me

N

S

OEt

O Ethyl 4-methylbenzenesulfonate

214 ON

Me

O

O

C

C

NO N Me

Me 215

NH

H N O2N

C

O

NO H2 N

N

C

Me 216

NO N Me

217

N H2 C

N

HCKNKN-H

Diazirine

HJCLNJN-H

HCKNJN-H

Isodiazomethane

17.9.2.2 Diazoalkanes Diazoalkanes are one of the most convenient and useful sources of reactive carbenes. Diazoalkanes (RN2) have been known and used synthetically for many years.241 The simplest diazoalkane is diazomethane, whose structure is represented by several resonance hybrid structures that are zwitterionic, although it can also be written as a nitrene. Note that nitrenes are quite useful reagents, particularly for generating aziridines, although it will not be discussed here.242 The preparation and reactions of other alkyl diazoalkanes are described in a review by Cowell and Ledwith.241b,243 Hydrazones (e.g., that from acetophenone) are convenient sources of carbenes via their reaction with p-toluenesulfonyl azide in THF to yield the corresponding diazoalkane (218) along with p-toluenesulfonamide.244 The first aliphatic diazo compound known was ethyl diazoacetate, (219), which was prepared by Curtius in 1883.245 Indeed, several types of amines can be converted to diazo compounds by treatment with potassium nitrite or nitrous acid. Treatment of ethyl glycinate hydrochloride with potassium nitrite (KNO2) gave ethyl diazoacetate. This technique is most useful when an electron-withdrawing group is adjacent to the amino group.

234

(a) de Boer, T. J.; Backer, H. J. Org. Synth., Coll. 1963, 4, 250. (b) Idem Rec. Trav. Chim 1954, 73, 229.

235

Moore, J. A.; Reed, D. E. Org. Synth. Coll. 1973, 5, 351.

(a) McKay, A. F. J. Am. Chem. Soc. 1948, 70, 1974. (b) McKay, A. F.; Ott, W. L.; Taylor, G. W.; Buchanan, M. N.; Crooker, J. F. Can. J. Res. 1950, 28B, 683.

236

237

Arndt, F. Org. Synth. Coll. 1943, 2, 165.

238

Hopps, H. B. Aldrichimica Acta 1970, 3, 9.

(a) Schmitz, E. Angew. Chem. Int. Ed. Engl. 1964, 3, 333. (b) Idem Adv. Heterocyclic Chem. 1963, 2, 83 (see p 122); (c) Schmitz, E.; Ohme, R. Tetrahedron Lett. 1961, 612. (d) Paulsen, S. R. Angew. Chem. 1960, 72, 781. 239

240

(a) Anselme, J. P. J. Chem. Educ. 1966, 43, 596. (b) M€ uller, E.; Ludsteck, D. Chem. Ber. 1954, 87, 1887.

241

(a) Huisgen, R. Angew. Chem. 1955, 67, 439. (b) Cowell, G. W.; Ledwith, A. Q. Rev. Chem. Soc. 1970, 24, 119.

242

M€ uller, P.; Fruit, C. Chem. Rev. 2003, 103, 2905.

(a) Jones, W. M.; Muck, D. L. J. Am. Chem. Soc. 1966, 88, 3798. (b) Jones, W. M.; Muck, D. L.; Tandy, T. K., Jr. Ibid. 1966, 88, 68. (c) Applequist, D. E.; McGreer, D. E. Ibid. 1960, 82, 1965. 243

244

Fischer, W.; Anselme, J. P. Tetrahedron Lett. 1968, 877.

245

Curtius, T. Chem. Ber. 1883, 16, 2230.

959

17.9 CARBENES AND CARBENOIDS

H2CJNKN

H2CKNKN

H2CJNKN

H2CJNLN

Diazomethane Ph N

NH2

Me

Ph

SO2N3

+ N2 + Me

N2

Me

SO2NH2

Me

Acetophenone hydrazone

218 KNO2

EtO2CH2NH2•HCl

EtO2CC=N2 +

Ethyl glycinate hydrochloride

+

KCl

2 H2O

Ethyl diazoacetate

Photolysis of diazo compounds (219) generates nitrogen and the carbene 220. Photolysis of diazomethane (219, R ¼ H), for example, generated methylene (H2C:).246 Benzophenone is often added as a sensitizer. Under these conditions, triplet methylene is formed via intersystem crossing (S1 ! T1)247 (Section 15.2.2). Energy transfer from triplet benzophenone to triplet diazomethane followed, and triplet diazomethane decomposed to triplet methylene.248 A similar thermal reaction gives a carbine which reacts normally.249 Diazoalkanes (221) can undergo valence isomerization to diazirines (222) upon photolysis, but reversal to the diazoalkane occurs both photochemically and thermally.239c,d,250 Diazirines can be photolyzed back to the diazoalkane or directly to the carbene.251 R R2CN2

+

L

–

NKN

hv

+

N2

R

R

R 219

R

+

NKN

–

220 hv

Heat or hv

R 221

R

N N

R 222

The diazoalkane disconnections follow: OR RN2

RNH2

O

NH2

17.9.2.3 α-Diazocarbonyl Compounds Another reaction that generates carbenes is a rearrangement of α-diazocarbonyl compounds (called the Wolff rearrangement),252 in which a diazocarbonyl compound (223) loses nitrogen (N2) to give an acyl carbene (224). This reaction occurs under both thermal and photochemical conditions, and the resulting carbene rearranges to a ketene (225) (completing the Wolff rearrangement of diazoketone to ketene). In this case, the Wolff rearrangement to ketene 225 was followed by an internal [2+2]-cycloaddition (Section 15.2) to give 226 in 79% yield.253 246

Herzberg, G.; Shoosmith, J. Nature (London) 1959, 183, 1801.

(a) B€ ackstr€ om, H. L. J.; Sandros, K. Acta Chem. Scand. 1958, 12, 823. (b) Idem Ibid. 1960, 14, 48. (c) Hammond, G. S.; Moore, W. M. J. Am. Chem. Soc. 1959, 81, 6334.

247

248

(a) Kopecky, K. R.; Hammond, G. S.; Leermakers, P. A. J. Am. Chem. Soc. 1961, 83, 2397. (b) Idem Ibid. 1962, 84, 1015.

249

Staudinger, H.; Kupfer, O. Berichte 1912, 45, 501.

250

(a) Schmitz, E.; Oehme, R. Angew. Chem. 1961, 73, 115. (b) Idem Chem. Ber. 1961, 94, 2166.

251

(a) Frey, H. M. Pure Appl. Chem. 1964, 9, 527. (b) Idem Advan. Photochem. 1966, 4, 225. (b) Overberger, C. G.; Anselme, J. P. Tetrahedron Lett. 1963, 1405.

(a) Wolff, L. Annalen 1912, 394, 23. (b) Reference 253a, pp. 475–492. (c) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-103. (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 702–703.

252

253

Yates, P.; Fallis, A. G. Tetrahedron Lett. 1968, 2493.

960

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

O CHN2

hv

C C

CH:

O

O

223

H

224

O

225

226 (79%)

OMe

OMe

MeO

MeO

Me

Me

hv, MeOH

OMe

Me

H Me N2

OMe

O Me

H Me

CO2Et

228 (30%)

227

Note that hydrolysis of a ketene will generate an acid (or an ester in the presence of an alcohol).254 In a synthesis of taiwaniaquinones A and F, Li and coworkers255 used this rearrangement to “shrink” a six-membered ring to a fivemembered ring. Irradiation of 227 in methanol gave a 30% yield of 228 via initial formation of a carbene and rearrangement to the ketene. In this work, Li discovered that heating 227 to 160°C in 2,4,6-collidine and benzyl alcohol provided the benzyl ester of 228 in 56% yield. Diazoketones (e.g., 223) can be formed from acid chlorides by reaction with diazomethane. Subsequent treatment with aq Ag2O leads to the Wolff rearrangement and formation of a carboxylic acid of one carbon more than the starting acid chloride. This sequence is known as the Arndt-Eistert synthesis256 and the ketene can be trapped as an ester or an amide. An example is taken from Reisman and coworker’s257 synthesis of (+)-salvileucalin B, in which acid 229 was first converted to the acid chloride and subsequent reaction with diazomethane generated the diazoketone. Reaction with silver triflate yielded 69% (overall) of 230 via a Wolff rearrangement. O

O 1. (COCl)2, cat DMF 2. CH2N2, THF

O

3. AgTFA, MeOH NEt3, THF –30 to 22°C

HO2C

O MeO2C

229

230 (69% overall)

The disconnections pertinent to these α-diazocarbonyl compounds follow: R

R

R CO2H

R1

O

R CO2H

R1 R1

CO2H

CO2H

O

O

17.9.2.4 Photolysis of Cyclopropanes Photolysis of cyclopropane derivatives258 will generate a carbene via elimination of an alkene. Photolysis of 1,1-dimethylcyclopropane in the gas phase gave dimethylmethylene (Me2C:) and ethene,259 and irradiation of 254

Bergmann, E. D.; Hoffmann, E. J. Org. Chem. 1961, 26, 3555.

255

Deng, J.; Li, R.; Luo, Y.; Li, J.; Zhou, S.; Li, Y.; Hu, J.; Li, A. Org. Lett. 2013, 15, 2022.

(a) Arndt, F.; Eistert, B. Berichte 1935, 68, 200. (b) Bachmann, W. E.; Struve, W. S. Org. React. 1942, I, 38. (c) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006, p ONR-3. (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 44–45.

256

257

Levin, S.; Nani, R. R.; Reisman, S. E. J. Am. Chem. Soc. 2011, 133, 774.

258

Griffen, G. W.; Bertoniere, N. R. In Carbenes; Jones, M., Jr., Moss, R. A., Eds., John Wiley, New York, 1977; pp. 305–31, 7, Vol 1.

259

Dhingra, A. K.; Koob, R. D. J. Phys. Chem. 1970, 74, 4490.

961

17.9 CARBENES AND CARBENOIDS

9,10-dihydro-9,10-methanophenanthrene [1a,9b-dihydro-1H-cyclopropa[l]phenanthrene] gave carbene and phenanthrene.260 Oxiranes can be photolyzed to carbenes as well.261 Me

hv

Me2C :

CH2KCH2

+

Me 1,1-Dimethylcyclopropane

hv

+ 1a,9b-Dihydro-1H-cyclopropa[l]phenanthrene

: CH2

Phenanthrene

The cyclopropane-alkene disconnection is

R

R

R

R

17.9.2.5 The Bamford-Stevens Reaction Aldehydes and ketones can be converted to a tosylhydrazone (e.g., 231) by reaction with tosylhydrazine (see Section 4.2.2). Subsequent treatment with base followed by heating results in the loss of p-toluenesulfinate to yield the corresponding diazo compound (232), which generates a carbene by loss of nitrogen to form an alkene. The formation of alkenes from tosyl hydrazones when they are treated with base is called the Bamford-Stevens reaction.262,263a The carbene nature of the intermediate was detected by examination of camphor tosylhydrazone,263 which on heating with hydroxide give a mixture of camphene and tricyclene. When the reaction was done in acidic alcohols (e.g., ethylene glycol), camphene was the main product (via rearrangement). In aprotic solvents, the carbene insertion products were observed (Section 17.9.5.4). A synthetic example of the Bamford-Stevens reaction is taken from a synthesis of (+)lasonolide A by Shishido and coworkers264 in which ketone 233 was converted to the tosylhydrazone (234) in 97% yield, and subsequent treatment with base gave alkene 235 in 86% yield. O O

Me

S

H N

O

R1 Me

R

SO2NHNH2

– Me

R1

BnO

OTBS

OTBS

p-TsNHNH2

233

232

O

THF. MeOH

O

R

R

R

OTBS

O

R1 R1

N

231

O

N2

SO2 –

Ethylene glycol

O BnO

O

HOCH2CH2O – Na+

NNHTs 234 (97%)

O BnO 235 (86%)

260

Richardson, D. B.; Durrett, L. H.; Martin, J. M., Jr.; Putnam, W. E.; Slaymaker, S. C.; Dvoretzky, I. J. Am. Chem. Soc. 1965, 87, 2763.

261

Reference 258, pp 318-331.

(a) Bamford, W. R.; Stevens, T. S. J. Chem. Soc. 1952, 4735. (b) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-4. (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 56–57. 262

263 (a) Bayless, J. H.; Friedman, L.; Cook, F. B.; Schechter, H. J. Am. Chem. Soc. 1968, 90, 531. (b) Davies, H. W.; Schwarz, M. J. Org. Chem. 1965, 30, 1242. (d) Powell, J. W.; Whiting, M. C. Tetrahedron 1959, 7, 305. (e) Clarke, P.; Whiting, M. C.; Papenmeier, G.; Reusch, W. J. Org. Chem. 1962, 27, 3356. 264

Yoshimura, T.; Yakushiji, F.; Kondo, S.; Wu, X.; Shindo, M.; Shishido, K. Org. Lett. 2006, 8, 475.

962

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

Although not a carbene mechanism, it involves extrusion of nitrogen via a carbanion and it is included here for comparison with the reactions just cited. Grieco et al.265 used this method to convert 236 to diene 237 (>98% yield) in a synthesis of ()-ivangulin. When an organolithium reagent is used as the base, the alkene-forming reaction is sometimes called the Shapiro reaction.266 Two molar equivalents of an organolithium reagent are required, and the reaction generates the less substituted alkene. In the Sha et al.267 synthesis of ()-pinguisenol, tosylhydrazone 238 (generated from the ketone by reaction with tosylhydrazone in 91% yield) was treated with butyllithium and TMEDA to give a 69% yield of 239. The Bamford-Stevens reaction generates the more substituted alkene. This reaction proceeds by an intermediate vinyllithium compound, and it is possible to intercept this reactive intermediate with appropriate electrophilic reagents. In a synthesis of vinigrol by Baran and coworkers,268 ketone 240 was converted to the tosylhydrazone and subsequent reaction with sec-butyllithium generated vinylithium reagent 241, which reacted with the DMF in the reaction medium to give aldehyde 242 in 63% overall yield. PhO

PhO

Me

Me LiN(i-Pr)2

N Me

NHTs

Me 237 (>98%)

236 NNHTs BuLi, TMEDA Hexane

239 (69%)

238

The relative amount of base has a significant effect on the product distribution. When the tosylhydrazone of pinacolone was treated with >1.8 M equiv. of n-butyllithium, 3,3-dimethylbut-1-ene was formed. When only 1 equiv. of n-butyllithium was used, however, a 57% yield of 3,3-dimethylbut-1-ene, 40% of 1,1,2-trimethylcyclopropane, and 3% of the rearranged alkene (2,3-dimethylbut-1-ene) were isolated.269 The type of base employed is also important. The product distribution from the hydrazone of cyclopentanone, for example, gave a mixture of cyclopentene/2-pent-(2E)ene/pent-(2Z)-ene.270 When sodium methoxide was used, a 2:84:14 mixture of these products was obtained, whereas sodium hydride led to a 83:14:3 mixture and sodium amide a 60:35:3 mixture.270 O

Li

1. TrisNHNH2, cat TsOH MeCN (73%)

CHO DMF

2. sec-BuLI, TMEDA Hexanes, DMF (86%)

240

242 (63% overall)

241

The Bamford-Stevens disconnection follows: R R1

265

R N2

O R1

R

R1

R

O R1

Grieco, P. A.; Oguri, T.; Wang, C.-L, J.; Williams, E. J. Org. Chem. 1977, 42, 4113.

(a) Shapiro, R. H. Tetrahedron Lett. 1968, 345. (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley: New York, 2005; pp 590–591.

266

267

Sha, C.-K.; Liao, H.-W.; Cheng, P.-C.; Yen, S.-C. J. Org. Chem. 2003, 68, 8704.

268

Maimone, T. J.; Shi, J.; Ashida, S.; Baran, P. S. J. Am. Chem. Soc.2009, 131, 17066.

269

(a) Kaufman, G.; Cook, F.; Shechter, H.; Bayless, J.; Friedman, L. J. Am. Chem. Soc. 1967, 89, 5736. (b) Shapiro, R. H.; Heath, M. J. Ibid. 1967, 89, 5734.

270

Kirmse, W.; von B€ ulow, B. G.; Schepp, H. J. L. Ann. Chem. 1966, 691, 41.

963

17.9 CARBENES AND CARBENOIDS

17.9.2.6 Photolysis of Carbonyls Under certain conditions, carbonyl compounds can be photolyzed to diradicals, which can rearrange to carbenes.271 Photolysis of cyclobutanone 243 generated diradical 244, and in one pathway elimination gave an alkene (246), as well as ketene 245. Decarbonylation led to formation of cyclopropane 247 via an alternative pathway. Yet another pathway involved ring expansion to yield carbene 248, which reacted with methanol to yield 249.272 When R ¼ R1 ¼ H, 48% of 245 and 246 was observed, with 8% of 249 and a trace of 247.272a When R ¼ R1 ¼ Me, however, 68% of 249 was obtained, along with 13% of 245 + 246 and 11% of 247.272a R1

R O

R R1

R R

hv

•

R

R

– CO

R1

R1 243

CKO

O

R1

R

R1

R

• R1

+

245

246

R1 247

244

OMe

R

••

R

O

MeOH

R1

R R1 248

R

O R1 R1 249

The carbonyl disconnection follows: R

O

R

R1

OH

R

O

R

17.9.2.7 Halocarbenes Polyhalomethanes can be converted to halocarbenes via photolysis, thermolysis, or treatment with base. When cyclohexene was treated with a mixture of chloromethane and phenylsodium, methylene (:CH2) was formed, and it added to the alkene to give norcarane (bicyclo[4.1.0]heptane), but in only 3.2% yield.273 Obviously, chloromethane is not the best source of carbenes. Polyhalomethanes give much better results, and the most common method for generating halocarbenes is treatment of dihalomethanes and trihalomethanes with Na metal or with another base.274 Chloroform is a particularly useful carbene precursor. Reaction of chloroform with hydroxide begins with an acid-base reaction to generate the anion (Cl3C:) and water. Loss of chloride ion generates dichlorocarbene (Cl2C:). Geuther275 suggested this type of carbene in 1862. Potassium tert-butoxide is the base most commonly used to generate dichlorocarbene, after its initial introduction by von E. Doering and Hoffmann.276 A typical reaction is that of cyclohexene and chloroform with potassium tert-butoxide to yield 7,7-dichlorobicyclo[4.1.0]heptane. In general, reactions with alkenes yield dichlorocyclopropanes [dibromocyclopropanes with bromoform (CHBr3) and diiodocyclopropanes with iodoform (CHI3)]. Tri-n-butyltin hydride (Bu3SnH) can be used to reduce the product dihalide to a monohalide (Section 7.11.7). MeCl, PhNa Hexadecane

Norcarane (Bicyclo[4.1.0]heptane) CHCl3

Cl

t-BuOK

Cl

7,7-Dichlorobicyclo[4.1.0]heptane

271

Reference 225a, pp 47–51.

272

(a) Turro, N. J.; Southam, R. M. Tetrahedron Lett. 1967, 545. (b) Morton, D. R.; Lee-Ruff, E.; Southam, R. M.; Turro, N. J. J. Am. Chem. Soc. 1970, 92, 4349.

273

Friedman, L.; Berger, J. G. J. Am. Chem. Soc. 1960, 82, 5758.

274

Reference 225a, pp. 129–150.

275

Geuther, A. Annalen 1862, 123, 121.

276

von E. Doering, W.; Hoffmann, A. K. J. Am. Chem. Soc. 1954, 76, 6162.

964

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

Typical halocarbene disconnections follow: R

X

R

X

R

R

17.9.3 N-Heterocyclic Carbenes N-Heterocyclic carbenes, (abb NHC) are singlet carbenes with a divalent carbine center connected directly to a heteroatom. There are several synthetic routes to NHCs.277 N-Heterocyclic carbenes are classified as electron-rich, nucleophilic species, and form strong bonds to metal centers.277 Metals include Ru, Os, Fe, Mn,278 Tc, Au, or Cu,279 Pd,280 and Rh.281 Many of the resulting complexes are rather stable, resistant to decomposition, and useful as precatalysts and organocatalysts.282 The chemistry of FedNHCs has been reviewed, and includes CdC coupling reactions,283 alkoxyalkylation, aziridination of alkenes, sulfonylations, reductions, dehydration reactions, and cyclization reactions.284 Another class of important metal carbenes are the Fischer carbine complexes.285 +

R

N

N

X–

+

R N

R1

Cyclic formamidinium salt

– O X

Oxazolium salt

+

R

N

– S X

Thiazolium salt

RO +

R N

R2 X

M(CO)5

–

R1 Pyrrolinium salt

R Fischer carbene complex

There are a variety of NHC precursors. Common structural motifs include cyclic formamidinium salts, oxazolium salts,286 thiazolium salts, and pyrrolium salts. Common counterions include halide ions PF6  , and BF4  . One of the most important uses of N-heterocyclic carbenes in as a component in ruthenium-based olefin metathesis catalysts.287 Metathesis reactions will be described in detail in Section 18.6. Carbene precatalysts include imadazoline dimesylate (250), and imadazolin-2-ylidene dimesylate (251), as well as 252 and 253. Note that there are several other synthetic applications for the metathesis catalysts.288

277

Benhamou, L.; Chardon, E.; Lavigne, G.; Bellemin-Laponnaz, S.; Cesar, V. Chem. Rev. 2011, 111, 2705.

278

See Cahiez, G.; Duplais, C.; Buendia, J. Chem. Rev. 2009, 109, 1434.

(s) With Au and Cu, see Gaillard, S.; Cazin, C. S. J.; Nolan, S. P. Acc. Chem. Res. 2012, 45, 778. (b) With Au, see Nolan, S. P. Acc. Chem. Res. 2011, 44, 91. 279

280

Marion, N.; Nolan, S. P. Acc. Chem. Res. 2008, 41, 1440.

(a) Díez-González, S.; Marion, N.; Nolan, S. P. Chem. Rev. 2009, 109, 3612. (b) See Schuster, O.; Yang, L.; Raubenheimer, H. G.; Albrecht, M. Chem. Rev. 2009, 109, 3445. (c) Colby, D. A.; Bergman, R. G.; Ellman, J. A. Chem. Rev. 2010, 110, 624. (d) Lewis, J. C.; Bergman, R. G.; Ellman, J. A. Acc. Chem. Res. 2008, 41, 1013. 281

282

(a) Glorius F., Ed., N-Heterocyclic Carbenes in Transition Metal Catalysis; Topics in Organimetallic Chemistry; Springer-Verlag: Berlin/Heidelberg, 2007; Vol. 21. (b) Nolan, S. R., Ed., N-Heterocyclic Carbenes in Synthesis; Wiley-VCH: New York, 2006. (c) Nair, V.; Vellalath, S.; Babu, B. P. Chem. Soc. Rev. 2008, 37, 2691. (d) Enders, D.; Nemeier, O.; Henseler, A. Chem. Rev. 2007, 107, 5606. (e) Marion, N.; Diez-Gonxalez, S.; Nolan, S. P. Angew. Chem. Int. Ed. 2007, 46, 2988.

283

For coupling of aldehyde with various partners, see Biju, A. T.; Kuhl, N.; Glorius, F. Acc. Chem. Res. 2011, 44, 1182.

284

Riener, K.; Haslinger, S.; Raba, A.; H€ ogerl, M. P.; Cokoja, M.; Herrmann, W. A.; and K€ uhn, F. E. Chem. Rev. 2014, 114, 5215.

285

D€ otz, K. H.; Stendel, Jr., J. Chem. Rev. 2009, 109, 3227.

For a review of N-Heterocyclic Carbene-Catalyzed Reactions Involving Acyl Azoliums, see Mahatthananchai, J.; Bode, J. W. Acc. Chem. Res. 2014, 47, 696. 286

287

(a) Vougioukalakis, G. C.; Grubbs, R. H. Chem. Rev. 2010, 110, 1746. (b) Samojłowicz, C.; Bieniek, M.; Grela, K. Chem. Rev. 2009, 109, 3708.

288

Alcaide, B.; Almendros, P.; Luna, A. Chem. Rev. 2009, 109, 3817.

965

17.9 CARBENES AND CARBENOIDS

Mes

N

N Mes

Mes

N

250

N Mes

i-Pr

N

251

N

i-Pr

Cy

N

252

N Cy 253

17.9.4 Reactions of Diazomethane As with other diazoalkanes, diazomethane reacts with alkenes to form cyclopropane derivatives (Section 17.9.5.1).289 Reaction with aromatic derivatives leads to ring expansion to cycloheptatriene derivatives.290 Both of these reactions (addition to an alkene or arene insertion) involve generation of an intermediate carbene and addition to a π-bond. Both reactions will be discussed below. Many of the reactions of diazomethane tend to be ionic in nature and are, therefore, set aside from the other diazoalkane chemistry discussed in this section. One of the most prevalent uses of diazomethane itself is for esterification of small quantities of carboxylic acids, especially acids that are obtained in very small quantities, usually at the end of a multistep synthesis, or cannot tolerate acidic or basic conditions. The reaction proceeds in near quantitative yield, producing a single product, as in Petasis and coworker’s291 conversion of 254 to methyl ester 255, in a synthesis of Lipid Mediator Resolvin D3. Note that the esterification reaction is faster than reaction with the C]C units. HO

HO

CO2H

CO2Me

HO

HO CH2N2

OH

OH

254

255

Diazomethane is used in other useful reactions. As noted in Section 17.9.2.3, diazomethane is used to convert acid chlorides to diazoketones. Diazoalkanes undergo [3 + 2]-cycloadditions in the presence of alkenes to yield pyrazolines, as discussed in Section 17.9.2.3. Diazomethane reacted with methacrylonitrile, for example, to give the pyrazoline (3methyl-4,5-dihydro-3H-pyrazole-3-carbonitrile) in 76% yield.292 In another reaction, diazomethane reacts with aldehydes to yield methyl ketones,293 as in the conversion of benzaldehyde to acetophenone (97% yield), which is a useful reaction for extending carbon chains. Diazomethane also reacts with cyclic ketones to yield ring-expanded ketones, along with epoxides as minor products. This method was used to prepare cycloheptanone from cyclohexanone in 65% yield by reaction with diazomethane, along with 15% of oxaspirooctane.234a,294 Me

Me CH2N2

CN Methacrylonitrile

Ether, 0°C

CN N N 3-Methyl-4,5-dihydro-3Hpyrazole-3-carbonitrile

(76%)

289

Muck, D. L.; Wilson, E. R. J. Org. Chem. 1968, 33, 419.

290

M€ uller, E.; Kessler, H.; Fricke, H.; Kiedaisch, W. Annalen 1964, 675, 63.

291

Winkler, J. W.; Uddin, J.; Serhan, C. N.; Petasis, N. A. Org. Lett. 2013, 15, 1424.

292

Gotkis, D.; Cloke, J. B. J. Am. Chem. Soc. 1934, 56, 2710.

293

(a) Schlotterbeck, F. Chem. Ber. 1907, 40, 479. (b) Gutsche, C. D. Org. React. 1954, 8, 364.

294

(a) House, H. O.; Grubbs, E. J.; Cannon, W. F. J. Am. Chem. Soc. 1960, 82, 4099. (b) Cram, D. J.; Helgeson, R. C. Ibid. 1966, 88, 3515.

966

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

Some of these specialized diazoalkane disconnections follow: O RCO2Me

RCO2H

R

R CHO CH3

17.9.5 Carbene Reactions of Diazoalkanes Carbenes are relatively easy to generate and have several synthetic uses. Diazoalkanes are converted to the corresponding carbene photolytically, or in the presence of Cu derivatives [e.g., Cu(acac)2]. Dirhodium tetraacetate [Rh2(OAc)4] was found to be a better catalyst in some cases, initiating reaction with electron-deficient alkenes.295 Metal carbenes are important in synthesis.296 Carbenes can undergo 1,2-hydrogen shifts to generate alkene products. When 3-benzyl-3-chloro-3H-diazirine was photolyzed [355 nm, 103.7 kcal (433.9 kJ) mol1], chlorocarbene 256 was formed. In the presence of 2,3-dimethylbut-2-ene, a 73% yield of the cyclopropane product (257) was produced.297 In addition to 257, 1-chloro-1-phenylethene was formed by a 1,2-hydrogen shift during the course of the reaction and was competitive with the addition. The energy barrier for this rearrangement was calculated to be 6.4  2 kcal (26.79  8.37 kJ) mol1 and proceeded at a rate of 4.9–6.7  107 s1.297a Although the addition reaction is usually faster, the 1,2-shift should be considered an important side reaction in many carbene reactions. Ph 1,2-Hydrogen shift

N

Cl

Ph

Cl

hv

1-Chloro-1-phenylethene

• •

N Ph

Cl

Me2CKCMe2

Me Ph

Me Me

Cl 3-Benzyl-3-chloro3H-diazirine

256

Me 257 (73%)

17.9.5.1 Addition to Alkenes Perhaps the most common reaction of carbenes is their addition to alkenes to generate cyclopropanes,298 as shown in the formation of 257 above. A generalized example is the preparation of 259 from carbene 220 and alkene 258.299 Carbenes can be generated by any of the methods discussed in Section 17.9.2. The addition is usually exothermic.300 A typical value is 91 kcal (380.9 kJ) mol1, with an activation energy of  64 kcal (267.9 kJ) mol1. Reaction of cis- and trans-but-2-ene with diazomethane (under photolytic conditions) showed that the yield of cyclopropanes increased with increasing pressure. Insertion (see Section 17.9.5.4) into allylic and vinyl bonds was competitive in many cases. The reaction is stereoselective for retention of the geometry of the starting alkene, when diazomethane is used.301 For other methylene precursors, trans-1,2-dimethylcyclopropane is favored for both cis- and trans-but-2-ene.302

295

Aggarwal, V. K.; Smith, H. W.; Hynd, G.; Jones, R. V.; Fieldhouse, R.; Spey, S. E. J. Chem. Soc. Perkin Trans. 2000, 1, 3267.

296

D€ orwald, F. Z. Metal Carbenes in Organic Synthesis; Wiley-VCH: New York, 1999.

(a) Jackson, J. E.; Soundararajan, N.; White, W.; Liu, M. T. H.; Bonneau, R.; Platz, M. S. J. Am. Chem. Soc. 1989, 111, 6874. (b) LaVilla, J. A.; Goodman, J. L. Ibid. 1989, 111, 6877.

297

298

For transformations of this type, see Ref. 182, pp 135–152. Also see Lebel, H.; Marcoux, J.-F.; Molinaro, C.; Charette, A. B. Chem. Rev. 2003, 103, 977.

(a) Reference 225a, Chapter 8, pp 267–362. (b) Moss, R. A. In Carbenes; Jones, M., Jr.; Moss, R. A., Eds.; John Wiley: New York, 1973, pp 153–304, Vol. 1.

299

300

Reference 225a, p 268.

(a) Reference 225a, p 271. (b) McKnight, C.; Lee, P. S. T.; Rowland, F. S. J. Am. Chem. Soc. 1967, 89, 6802. (c) Eder, T. W.; Carr, R. W., Jr. J. Phys. Chem. 1969, 73, 2074. (d) Ring, D. F.; Rabinovitch, B. S. Ibid. 1968, 72, 191. (e) Duncan, F. J.; Cvetanovi_c, R. J. J. Am Chem. Soc. 1962,84, 3593. (f ) Montague, D. C.; Rowland, F. S. J. Phys. Chem. 1968, 72, 3705. (g) Frey, H. M. J. Am. Chem. Soc. 1960, 82, 5947. 301

302

(a) Frey, H. M. Proc. R. Soc. Ser. A 1959, 250, 409. (b) Idem Ibid. 1959, 251, 575.

967

17.9 CARBENES AND CARBENOIDS

R

R R

R

R1

R2

R1

R2

+

220

R1 R1

258

R2 R2 259

The ratio of cis- to trans-isomers is dependent on the wavelength of light used, and on the nature of any additive. The reaction was more stereoselective at higher pressures, and when long wavelengths of light were used.301 Molecular oxygen appears to be the best additive, and irradiation at 313 nm, 91.4 kcal ¼ (382.6 kJ) mol1 is optimal for methylene generation from ketenes. The yield of cyclopropane products from but-2-ene was 87% at 366 nm, 78.1 kcal ¼ (326.9 kJ) mol1, but only 29% at 313 nm.301c Oxygen scavenges triplet methylene, which gives diminished stereoselectivity relative to singlet methylene, although isomerization of the active intermediates is also a factor.303 When the reaction occurs in the liquid phase, which is more common, the reaction is stereoselective unless the carbene has a readily accessible triplet state.304 The reaction of various diazo compounds (and the carbenes they produce) with but-2-ene has been studied.305 Photolysis of diazomethane in the presence of cis-but-2-ene, for example, gave cis-1,2-dimethylcyclopropane with complete retention of configuration.306 Other diazoalkanes led to a mixture of trans-1,2-dimethylcyclopropane and cis-1,2-dimethylcyclopropane, but the ratio of cis/trans often changed with structural features of the diazo compound. Aryl diazo compounds tended to give more cis-product, whereas diazoketones gave more trans-cyclopropanes. A decrease in stereoselectivity was observed when triplet methylene was produced (by photolysis of diazomethane) in the presence of a large excess of perfluoropropane (CF3CF2CF3).301d There was a decrease in stereoselectivity for the sensitized reaction in comparison with direct photolysis. It is known that the rate of carbene addition also varies with the type of alkene.307 Me

: CH2

+ Me

Me

Me

Me

Me

cis-1,2-Dimethylcyclopropane

trans-1,2-Dimethylcyclopropane

The reaction of triplet methylene with a cis-alkene (260) proceeds by a two-step mechanism. The initially generated diradical (261) is in equilibrium with rotamer 262. Both triplet diradicals are slowly converted to singlet diradical 263 and 264, which leads to the cis-cyclopropane 265 and the trans-cyclopropane 266, respectively.308,321 Singlet carbenes are produced from singlet diradicals, and ring closure proceeds with high selectivity. Bond rotation in the triplet diradical is faster than T1 ! S1 conversion or ring closure. Ring closure of the singlet radical is faster than bond rotation, and there is an equilibrium between 263 and 264 and unfavorable relative to conversion to 265 or 266. H

•

• R

H

R

R

H

H

• R

H

H R

R

R 263

H

•

• H

262

Reference 225a, p 275.

R H

261

•

Reference 225a, p 272.

R H

265

: CH2

H

304

•

• H

260

303

H

R H

R

H R

R

R

H 264

266

(a) Reference 225a, p 279. (b) Moritani, I.; Yamamoto, Y.; Murahashi, S. Tetrahedron Lett. 1968, 5697. (c) Cowan, D. O.; Couch, M. M.; Kopecky, K. R.; Hammond, G. S. J. Org. Chem. 1964, 29, 1922. (d) Jones, M., Jr.; Ando, W. J. Am. Chem. Soc. 1968, 90, 2200. (e) Jones, M., Jr.; Ando, W.; Kulczycki, A., Jr. Tetrahedron Lett. 1967, 1391.

305

(a) Skell, P. S.; Woodworth, R. C. J. Am. Chem. Soc. 1956, 78, 4496, 6427. (b) Idem Ibid. 1959, 81, 3383. (c) von E. Doering, W.; LaFlamme, P. Ibid. 1956, 78, 5447.

306

307

Reference 225a, pp. 294–297.

308

Reference 225a, p 283.

968

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

When more than one alkene moiety is present, there is the potential for formation of isomers via addition to one or both of the double bonds. In general, carbenes add to the more electron-rich alkene. Reaction of 2,3,4,7-tetrahydro-1Hindene with dichlorocarbene (Cl2C:) gave only 267, but 1,2,3,4,5,8-hexahydronaphthalene gave an 80% yield of 268, with 20% of 269.309 Where it is appropriate, there seems to be a preference for reaction with trans double bonds. An example is the reaction of methylene with cis,trans,trans-cyclodeca-1,5,9-triene, which gave a 97:3 mixture of bicyclo [12.1.0]pentadeca-4,9-diene/bicyclo[12.1.0]pentadeca-5,9-diene.310 Note that the trans double bond may be more sterically accessible, which can account for the stereochemical preference. Similar reaction with dichlorocarbene and also with the Simmons-Smith reagent (Section 17.9.6) showed the same preference for reaction with the (E)-alkene (this preference ranged from 2:1 to 9:1). Cl

Cl

: CCl2

2,3,4,7-Tetrahydro-1H-indene

267 Cl

: CCl2

Cl

Cl

+ Cl

1,2,3,4,5,8-Hexahydronaphthalene

268 (80%)

269 (20%)

CH2N2, Cu(II)

Cyclotetradeca1,5,10-triene

+

Bicyclo[12.1.0]pentadeca-4,9-diene

Bicyclo[12.1.0]pentadeca-5,9-diene

A wide range of cyclopropane derivatives can be prepared by carbene addition to alkenes.311 There is a general preference for the less congested exo-product in reactions with cyclic alkylidene compounds. Initial addition of the carbene can be accompanied by rearrangement of the product, as discussed in Section 17.9.5.1.312 In many cases, however, the cyclopropanation proceeds without rearrangement. Diazoalkanes add to alkenes to give cyclopropane derivatives in the presence of transition metal reagents. An example is the reaction of alkene 270 with ethyl diazoacetate in the presence of cuprous iodide to give 271 (in >80% yield and 7:1 dr), taken from the Wicha and Michalak313 synthesis of 17-epi-calcitriol derivatives. Me

Me

H

CO2Et

N2CH2CO2Et

H

10% CuI•P(OMe)3

H OBz 270

309

H OBz 271 (>80%, 7:1 dr)

Sims, J. J.; Honwad, V. K. J. Org. Chem. 1969, 34, 496.

(a) Nozaki, H.; Kawanisi, M.; Noyori, R. J. Org. Chem. 1965, 30, 2216. (b) Nozaki, H.; Kat^ o, S.; Noyori, R. Can. J. Chem. 1966, 44, 1021. (c) Locke, J. M.; Duck, E. W. Chem. Ind. (London) 1965, 1727. (d) M€ uhlst€adt, M.; Graefe, J. Chem. Ber. 1966, 99, 1192. (e) Idem Z. Chem. 1969, 9, 303 (Chem. Abstr. 71:90918y 1969).

310

311

Reference 225a, pp 304–320.

312

Reference 225a, pp 321–328.

313

Michalak, K.; Wicha, J. J. Org. Chem. 2011, 76, 6906.

969

17.9 CARBENES AND CARBENOIDS

Most carbenes react similarly. In Trost and Oslob’s314 synthesis of ()-anatoxin, the alkene unit in 272 reacted with dibromocarbene, generated from bromoform and base, to give an 85% yield of 273 and 274 as a 3.5:1 mixture of exo and endo isomers. Carbenes react with enamines in the normal way, but in some cases the reaction is accompanied by a ring expansion to give a ketone via rearrangement of the initially formed aminocyclopropane. The cyclopropane product can be very stable and is usually isolated as the major product. Dichlorocarbene reacted with 1-piperidinocyclopentene [1-(cyclopent-1-en-1-yl)piperidine], derived from piperidine and cyclopentanone (see Section 13.6.1) to yield 275. Loss of chloride ion generated the conjugated iminium chloride (276), and hydrolysis gave 2-chlorocyclohex-2-en-1-one.315

Ts N

Br

N

Pentane, –15°C

Boc

H

Ts

CHBr3, KOt-Bu

273

Br

H

Boc

Cl

••

Br

Cl H3O+

: CCl2

N

H

274

Cl

Cl

Br

N

+

Boc

272

H

Ts

N

••

N

1-(Cyclopent-1-en1-yl)piperidine

275

O 2-Chlorocyclohex2-en-1-one

276

The disconnections available from carbene additions follow: R

R

R

R

R

R1 R

X X

R R

17.9.5.2 Addition to Aromatic Derivatives Aromatic compounds react with carbenes, but ring expansion usually follows the initial cyclopropanation. In a typical example, 2-methoxynaphthalene reacted with dichlorocarbene to yield 277, but in situ ring expansion gave 278.316 Enol ethers react to give either unsaturated acetals or unsaturated carbonyl compounds.317 Rearrangement or ring opening of the cyclopropane adducts derived from aromatic compounds are possible, depending on the carbene. Addition of carbene to benzene gave norcarane 279 (R ¼ H) and cycloheptatriene (280, R ¼ H) in equilibrium. With functionalized norcarane products, ring opening usually occurs to yield the substituted benzene, (281, R ¼ R1 ¼ CN, e.g.). Product 280 is favored when 279 is generated by reaction of benzene with diazomethane or ester carbenes (e.g., :CHCO2R).318 Photolysis of diazomethane in benzene, for example, gave a 4.8:1 mixture of cycloheptatriene and toluene.319

314

Trost, B. M.; Oslob, J. D. J. Am. Chem. Soc. 1999, 121, 3057.

315

(a) Wolinsky, J.; Chan, D.; Novak, R. Chem. Ind. (London) 1965, 720. (b) Ohno, M. Tetrahedron Lett. 1963, 1753.

316

Parham, W. E.; Bolon, D. A.; Schweizer, E. E. J. Am. Chem. Soc. 1961, 83, 603.

(a) Buddrus, J.; Nerdel, F.; Hentschel, P.; Klamann, D. Tetrahedron Lett. 1966, 5379. (b) Nerdel, F.; Buddrus, J.; Brodowski, W.; Hentschell, P.; Klamann, D.; Weyerstahl, P. Annalen 1968, 710, 36.

317

318

Reference 225a, p 381.

(a) von E. Doering, W.; Knox, L. H. J. Am. Chem. Soc. 1950, 72, 2305. (b) Meerwein, H.; Disselnk€ otter, H.; Rappen, F.; von Rintelen, H.; van de Vloed, H. Annalen, 1957, 604, 151. (c) Lemmon, R. M.; Strohmeier, W. J. Am. Chem. Soc. 1959, 81, 106. (d) Russell, G. A.; Hendry, D. G. J. Org. Chem. 1963, 28, 1933. 319

970

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

Cl OMe

Cl

Cl

: CCl2

OMe

2-Methoxynaphthalene

O

277

278 R

R C:

R

R

R R

280

R 279

R 281

This reaction is not restricted to simple benzene derivatives, and other aromatic rings can participate. In synthetic studies directed toward the guanacastepenes by Trauner and coworkers,320 for example, diazo derivative 282 was treated with a Rh to give 283 by cyclopropanation of the aromatic furan ring (Section 17.9.5.2). Ring opening led to a 50% yield of 284 by ring expansion and formation of an enol that tautomerized to the aldehyde unit. O

Me

Me CO2Et

Me N2 TBDPSO O

O

O

Rh2(OAc) 4 CH2Cl2, rt 0.002

TBDPSO

CO2Et

O

TBDPSO O

CO2Et

H

282

283

284 (50%)

Other carbenes react similarly, and the reaction can be applied to much more sophisticated substrates. In a synthetic example taken from the Feldman et al.321 synthesis of pareitropone, alkynyl-tin biaryl 285 was first converted to hypervalent iodine derivative 286. Subsequent treatment with base led to a cyclization reaction that formed the six-membered nitrogen-containing ring, and a vinyl carbene (287). Carbene insertion to the phenyl ring and ring expansion in the expected manner led to formation of the seven-membered ring found in pareitropone, (288), in 64% overall yield.322 The course of the reaction is dictated by the inherent stability of the final product.322 OSi(i-Pr)3

OSi(i-Pr)3

IPhOTf

SnBu3 PhI(CN)OTf

MeO

MeO

NHTs

LiN(SiMe3)2

MeO

MeO

NHTs

285

286 OSi(i-Pr)3

OSi(i-Pr)3

MeO

MeO

NTs

NTs

MeO

MeO

287 320

Hughes, C. C.; Kennedy-Smith, J. J.; Truaner, D. Org. Lett. 2003, 5, 4113.

321

Feldman, K. S.; Cutarelli, T. D.; Di Florio, R. J. Org. Chem. 2002, 67, 8528.

288 (64%)

(a) Sonnenberg, J.; Winstein, S. J. Org. Chem. 1962, 27, 748. (b) Skell, P. S.; Sandler, S. R. J. Am. Chem. Soc. 1958, 80, 2024. (c) Gatlin, L.; Glick, R. E.; Skell, P. S. Tetrahedron 1965, 21, 1315.

322

971

17.9 CARBENES AND CARBENOIDS

The carbene ring expansion disconnection follows: R

17.9.5.3 Cyclization Reactions of α-Diazoketones An attractive application of the cyclopropanation reaction is the intramolecular trapping of carbenes to form bicyclic compounds, with a three-membered ring appended to another ring. An example of this reaction is the previously discussed conversion of 270 to 271 in Section 17.9.5.1. When diazoalkanes are treated with transition metals in the presence of an alkene, particularly Cu or Rh derivatives, cyclopropanation occurs although the reactive intermediate may not be a free carbene. With Rh compounds, in particular, a C]Rh unit is the reactive species rather than a free carbene. A synthetic example, taken from Srikrishna and Dethe’s323 synthesis of ()-cucumin H, treated acid chloride 289 with diazomethane to yield diazoketone 290. Subsequent reaction with CudCuSO4 gave cyclopropane derivative 291 in >70% overall yield. The mixture just cited is not required, and heating with cupric sulfate or copper metal324 can also be used to initiate this cyclization, as in the von E. Doering et al.325 synthesis of barbaralone by heating 292 with Cu. O

O

O

CH2N2, rt

Cl

Ether

Cu–CuSO4, W-lamp Cyclohexane, Reflux

N2

289

290

291 (>70%) O

O

Cu

N2

Heat

292

Barbaralone

This carbene insertion reaction has been used in a variety of syntheses and coupled with other synthetic techniques. Taber et al.326 used carbene cyclopropanation in several syntheses, including a copper-catalyzed synthesis of (+)-isoneonepatelactone. Other diazo-carbonyl compounds can be used. In a synthesis of ()-spirotryprostatin B, Carreira and Meyers327 reacted diazolactam 293 (prepared from isatin)328 with penta-1,3-diene, in the presence of the Rh catalyst, to give a 71% yield of cyclopropyl derivative 294. In another example, the reaction of diazoketone (295) and rhodium octanoate [Rh2(oct)4] produced a 52% yield of 296, along with 28% of 297. These compounds were used in further synthetic transformations in the Taber and Teng329 synthesis of the ethyl ester of the major urinary metabolite of prostaglandin E2.

323

Srikrishna, A.; Dethe, D. H. Org. Lett. 2003, 5, 2295.

324

Julia, S.; Linstrumelle, G. Bull. Soc. Chim. Fr. 1966, 3490.

325

von E. Doering, W.; Ferrier, B. M.; Fossel, E. T.; Hartenstein, J. H.; Jones, M., Jr.; Klumpp, G.; Rubin, R. M.; Saunders, M. Tetrahedron 1967, 23, 3943.

326

Taber, D. F.; Amedio, J. C., Jr.; Raman, K. J. Org. Chem. 1988, 53, 2984.

327

Meyers, C.; Carreira, E. M. Angew. Chem. Int. Ed. 2003, 42, 694.

328

Cava, M. P.; Little, R. L.; Napier, D. R. J. Am. Chem. Soc. 1957, 80, 2257.

329

Taber, D. F.; Teng, D. J. Org. Chem., 2002, 67, 1607.

972

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

N2 1% [Rh(OAc)2]2, PhH Reflux, CH2Cl2

O

O

N H

N

Slow addition of 293

H 294 (71%)

293 O O TBSO

O

Rh2(oct)4, CH2Cl2

N2

rt, 8 h

CO2Et

TBSO

CO2Et

+

CO2Et

TBSO

(CH2)4CO2Et

(CH2)4CO2Et

CO2Et 296 (52%)

295

297 (28%)

The intramolecular carbene addition disconnections follow: O

O R (CH2)n

R

R

R R

(CH2)n R

R R1

R R1 NH2

O

O

17.9.5.4 The CdH Insertion Reactions of α-Diazoketones Taber and coworkers330b,331 showed that diazocarbonyls undergo CdH insertion reactions in the presence of rhodium acetate [Rh2(OAc)4], as illustrated in a syntheses of ()-pentalenolactone E methyl ester332 and α-cuparenone.333 Other Rh catalysts can be used. Wakimoto, et al.334 applied this reaction to a synthesis of aperidine. When 298 was treated with a catalytic amount of a chiral Rh catalyst, CdH insertion led to 299 in 61% yield. Several dirodium catalysts have been developed, including the Davies catalyst Rh2(S-DOSP)2.335 Note that Doyle and coworkers336 developed several important Rh catalysts, including the Doyle catalysts,337 that are widely used in cyclopropanation and CdH insertion reactions. In a different example with a different catalyst, taken from a synthesis of ()-tetrodotoxin, Du Bois and Hinman338 used Rh2(HNCOCPh2)4 to convert 300 to 301.

330

(a) Wenkert, E.; Davis, L. L.; Mylari, B. L.; Solomon, M. F.; da Silva, R. R.; Shulman, S.; Warnet, R. J.; Ceccherelli, P.; Curini, M.; Pellicciari, R. J. Org. Chem. 1982, 47, 3242. (b) Taber, D. F.; Petty, E. H. Ibid. 1982, 47, 4808.

331

Taber, D. F.; Ruckle, R. E., Jr. J. Am. Chem. Soc. 1986, 108, 7686.

332

Taber, D. F.; Schuchardt, J. L. J. Am. Chem. Soc. 1985, 107, 5289.

333

Taber, D. F.; Petty, E. H.; Raman, K. J. Am. Chem. Soc. 1985, 107, 196.

334

Wakimoto, T.; Miyata, K.; Ohuchi, H.; Asakawa, T.; Nukaya, H.; Suwa, Y.; Kan, T. Org. Lett. 2011, 13, 2789.

(a) Davies, H. M. L.; Hansen, T.; Churchill, M. R. J. Am. Chem. Soc. 2000, 122, 3063; (b) Davies, H. M. L.; Hansen, T. J. Am. Chem. Soc. 1997, 119, 9075. Also see (c) Davies, H. M. L. Aldrichimica Acta 1997, 30, 107. 335

(a) Doyle, M. P. Pure Appl. Chem. 1998, 70, 1123. (b) Doyle, M. P.; Protopopova, M. N. Tetrahedron, 1998, 54, 7919. (c) Martin, S. F.; Spaller, M. R.; Liras, L.; Hartman, B. J. Am. Chem. Soc. 1994, 116, 4493. (d) Doyle, M. P.; Kalinin, A. V. J. Org. Chem. 1996, 61, 2179. (e) Doyle, M. P.; Dyatkin, A. B.; Roos, G. H. P.; Cañas, F., Pierson, D. A.; van Basten, A.; Mueller, P.; Polleux, P. J. Am. Chem. Soc. 1994, 116, 4507. Also see Doyle, M. P.; Hu, W.; Valenzuela, M. V. J. Org. Chem. 2002, 67, 2954 for a synthesis of (S)-(+)-imperanene using these catalysts.

336

337

Colacot, T. J. Proc. Indian Acad. Sci. (Chem. Sci.) 2000, 112, 197.

338

Hinman, A.; Du Bois, J. J. Am. Chem. Soc. 2003, 125, 11510.

973

17.9 CARBENES AND CARBENOIDS

OBn

t-Bu

N

O

O

CO2Me

298

CO2Me

L Rh2(S-PTTL) 4

O

299 (61%)

Rh Rh

O

N2

O

O

1.5% Rh2(HNCOCPh3)4

PivO O

O

PivO

CCl4

O O

OBn MeO2C

O

N2

MeO2C

O

CH2Cl2, MS 4 Å, –78 to 0°C, 11 h

O

O

OSiMe2t-Bu

O

O

OSiMe2t-Bu

301

300

The carbene CdH insertion disconnection follows: O

O CO2R1

CO2R1 R

R

There is an interesting variation of this reaction that involves intramolecular NdH insertion of α-diazocarbonyls. When diazoketone (302) was treated with cupric bis(acetylacetonate), for example, a 74% yield of an azetidinone, 1-tosylazetidin-3-one, was obtained.339 It was shown that competitive intramolecular CdH insertion by the carbenoid was not a problem, illustrated by the presence of 10 equiv. of styrene in the reaction. The styrene actually seemed to enhance the NdH insertion reaction. CHN2

TsHN

O

0.1 Cu(acac)2 , PhH 10 equiv PhCH2KCH2

N

Reflux

O

Ts 1-Tosylazetidin-3-one (74%)

302

17.9.6 Carbenoids This section involves a class of compounds called carbenoids, which are reactive intermediates that react similarly to a carbene, but carbenes are not formed. The most commonly used carbenoid is generated by reaction of diiodomethane and a Zn-Cu couple. When the carbenoid formed in this manner adds to alkenes, it is called the Simmons-Smith reaction.340 I I CH2I2

ZnI ZnI 303 R ICH2ZnR'

R

H C

R

Norcarane

304

R

H

R1 Zn I

– ZnIR1

R

R

305 339

Wang, J.; Hou, Y.; Wu, P. J. Chem. Soc. Perkin Trans. 1 1999, 2277.

(a) Simmons, H. E.; Smith, R. D. J. Am. Chem. Soc. 1958, 80, 5323. (b) Idem Ibid. 1959, 81, 4256. (c) Denis, J. M.; Girard, J. M.; Conia, J. M. Synthesis 1972, 549. (d) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-87. (e) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp. 600–601. 340

974

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

A simple example is the conversion of cyclohexene to bicyclo[4.1.0]heptane (norcarane). Initial reaction of diiodomethane with Zn gave an iodozinc compound (303, see Section 13.8.2 for a brief discussion of organozinc compounds), which reacted with the alkene to yield 304. Loss of zinc iodide (ZnI2) gave the cyclopropane derivative (304 in this case).341 A one-step mechanism has also been proposed that involves 305,340a,342 and generates the cyclopropane derivative. The iodomethylene zinc intermediate is represented as 306,343 based in part on the structure of known organozinc compounds.344 As with carbenes, more highly substituted alkenes react somewhat faster than less substituted alkenes.342d Even highly substituted alkenes [e.g., 1,10 -bi(cyclohexylidene)] gave an 87% yield of the Simmons-Smith product (dispiro [5.0.57.16]tridecane).345 It is known that 1-substituted cyclic alkenes react faster than unsubstituted alkenes.346 1-Methylcyclohexene (relative rate, 21.4) and methylenecyclohexane (relative rate, 3.84) both react faster than cyclohexene (relative rate, 1.0). 1,2-Dimethylcyclohexene, however, reacts slightly slower, with relative rate of 0.94. Cyclohexene (relative rate, 1.0) reacts slower than cyclopentene (relative rate, 1.60) or norbornene (relative rate, 1.70).346 In addition, 1-methylcyclopentene (relative rate, 5.14) reacts faster than cyclopentene (relative rate, 1.60).346

I

Zn

306 Reprinted with permission from McElvain, S. M. Org. React. 1948, 4, 256. Copyright © 1948 by John Wiley & Sons, Inc.

Preparation of the Zn is critical to good yields in the Simmons-Smith reaction. Both ZndCu and Zn-Ag couples have been used with both I and Li.347 The Zn must be activated or the yield of cyclopropane product is poor. A useful development is the use of ultrasound to activate the Zn. Treatment of ()-α-pinene with the usual Simmons-Smith reagent gave only 12% of 307.348 When the Zn was activated by ultrasound, however, a 67% yield of 307 was obtained.349 CH2I2 Zn–Cu

1,1'-Bi(cyclohexylidene)

Dispiro[5.0.57.1 6]tridecane

Zn-Cu, CH2I2 , Ether 12 h

(12%)

Zn (ultrasound), CH2I2

(-)-α-Pinene 341

Ether

(67%)

307

Sawada, S.; Inouye, Y. Bull. Chem. Soc. Jpn. 1969, 42, 2669.

(a) Wittig, G.; Schwarzenbach, K. Annalen 1961, 650, 1. (b) Simmons, H. E.; Blanchard, E. P.; Smith, R. D. J. Am. Chem. Soc, 1964, 86, 1347. (c) Wittig, G.; Wingler, F. Chem. Ber. 1964, 97, 2146. (d) Blanchard, E.-P.; Simmons, H. E. J. Am. Chem. Soc. 1964, 86, 1337.

342

343

Simmons, H. E.; Cairns, T. L.; Vladuchick, S. A.; Hoiness, C. M. Org. React. 1973, 20, 1.

344

Shearer, H. M. M.; Spencer, C. B. J. Chem. Soc. Chem. Commun. 1966, 194.

345

Koch, S. D.; Kliss, R. M.; Lopiekes, D. V.; Wineman, R. J. J. Org. Chem. 1961, 26, 3122.

(a) Reference 225a, p 299. (b) Rickborn, B.; Chan, J. H. H. J. Org Chem. 1967, 32, 3576. (c) Balenkova, E. S.; Kochnova, G. P.; Khromov, S. I. Neftekhimiya 1969, 9, 29 (Chem. Abstr. 70: 106025x, 1969).

346

347

Rieke, R. D.; Li, P. T.-J., Burns, T. P.; Uhm, S. T. J. Org. Chem. 1981, 46, 4323.

348

Filliatre, C.; Gueraud, C. C. R. Acad. Sci. Paris, Ser. C 1971, 273, 1186.

349

Repic, O.; Vogt, S. Tetrahedron Lett. 1982, 23, 2729.

975

17.9 CARBENES AND CARBENOIDS

An example of the Simmons-Smith cyclopropanation is taken from a synthesis of solandelactone A by White et al.,350 in which 308 was converted to 309 in 97% yield. Note that an alternative preparative method was used to generate the carbenoid that avoided the use of the Zn-Cu couple, using diethyl zinc (Et2Zn).351 It is known that electron-rich alkenes (e.g., vinyl ethers and silyl enol ethers) usually react faster than other alkenes.352 There is a pronounced neighboring-group effect when a hydrogen-bonding oxygen atom is present as a substituent. The oxygen will “coordinate with the reagent, increasing the rate and control the stereochemistry of the addition.”346a,b,353 The reaction of cyclopent-3-en-1-ol, for example, gave bicyclo[3.1.0]hexan-2-ol via coordination of the oxygen in 310. In a synthetic example, Casares and Maldonado354 used this neighboring group effect to convert 311 to 312 in a synthesis of β-cuparenone. Me TBDPSO

N

TBDPSO

OMe OH

Me

H

Et2Zn, CH2I2 CH2Cl2

N

OMe H

O

308

OH

O

309 (97%)

Asymmetric induction is possible in the Simmons-Smith reaction if a suitable chiral additive or auxiliary is used, or if the chirality is built in via a chiral template precursor (Section 10.9). An example using a chiral additive is taken from the synthesis of coibacin A, by Pilli and coworkers.355 In this work, the reaction of (E)-but-2-en-1-ol with chiral additive 313 in the presence of the Simmons-Smith reagent gave a good yield of 314 with an enantiomeric ratio (er) of 94:6. In this case, the product was rather volatile and was used immediately without purification. I H2 C

CH2I2

OH

Zn

OH

O H

Zn/Cu

Cyclopent-3-en-1-ol

Bicyclo[3.1.0]hexan-2-ol

310 OH

H

OH

CH2I2

Me

Zn/Cu

Me

Me

311

Me Me

Me

312

When 315 was prepared from diethyl tartrate, subsequent treatment with diethylzinc gave a 99% yield of 316. Subsequent hydrolysis gave 317 (74% yield and 90 %ee), which allowed Yamamoto and coworkers356 to complete an asymmetric synthesis of 5,6-methanoleukotriene A4. Imai et al.357 has also reported a chiral boronic acid auxiliary that leads to good enantioselectivity in the Simmons-Smith reaction. The ability to produce the highly useful cyclopropane derivatives with good-to-excellent enantiomeric purity greatly enhances the utility of an already classical reaction.

White, J. D.; Lincoln, C. M.; Yang, J.; Martin, W. H. C.; Chan, D. B. J. Org. Chem. 2008, 73, 4139. Also see White, J. D.; Martin, W. H. C.; Lincoln, C.; Yang, J. Org. Lett. 2007, 9, 3481. 350

351

Furukawa, J.; Kawabata, N.; Nishimura, J. Tetrahedron 1968, 24, 53.

(a) Salim, H.; Piva, O. Tetrahedron Lett. 2007, 48, 2059. (b) Ragauskas, A. J.; Stothers, J. B. Can. J. Chem. 1985, 63, 2969. (c) Hoyano, Y.; Patel, V.; Stothers, J. B. Ibid. 1980, 58, 2730. (d) Jurlina, J. L.; Patel, H. A.; Stothers, J. B. Ibid. 1984, 62, 1159.

352

(a) Reference 225a, p 95. (b) Winstein, S.; Sonnenberg, J. J. Am. Chem. Soc. 1961, 83, 3235. (c) Dauben, W. G.; Berezin, G. H. Isbid. 1963, 85, 468. (d) Chan, J. H.; Rickborn, B. Ibid. 1968, 90, 6406.

353

354

(a) Casares, A.; Maldonado, L. A. Synth. Commun. 1976, 6, 11; also see (b) Ando, M.; Sayama, S.; Takase, K. Chem. Lett. 1979, 191.

355

Carneiro, V. M. T.; Avila, C. M.; Balunas, M. J.; Gerwick, W. H.; Pilli, R. A. J. Org. Chem. 2014, 79, 630.

(a) Arai, I.; Mori, A.; Yamamoto, H. J. Am. Chem. Soc. 1985, 107, 8254. (b) Mori, A.; Arai, I.; Yamamoto, H. Tetrahedron, 1986, 42, 6447. (c) Mash, E. A.; Nelson, K. A. Ibid. 1987, 43, 679. (d) Mash, E. A.; Nelson, K. A. Tetrahedron Lett. 1986, 27, 1441.

356

357

Imai, T.; Mineta, H.; Nishida, S. J. Org. Chem. 1990, 55, 4986.

976

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES Me2NOC

CONMe2

O

O B Bu

H

313 HO

H

Zn(CH2I)2 , DME

HO

(E)-But-2-en-1-ol

i-PrO2C i-PrO2C

314

i-PrO2C

O

CH2I2 , Et2Zn

O

i-PrO2C

H

O

OHC

TsOH MeOH

O

H CO2Me

CO2Me

315

CO2Me

316 (99%)

317 (74%, 90 %ee)

The Simmons-Smith disconnection follows:

R

R

R

R

An interesting extension of reactivity that involves the Simmons-Smith reagent chain extends carbonyl compounds, as reported by Ronsheim and Zercher.358 In a simple example, methyl 4,4-dimethyl-3-oxopentanoate was treated with diethylzinc and diiodomethane followed by iodine. Destruction of excess iodine with sodium thiosulfate, and elimination with DBU (Section 2.2.2.1.) gave an 86% yield of methyl (E)-5,5-dimethyl-4-oxohex-2-enoate. This sequence is a carbene-mediated chain extension. O

O

O

1. CH2I2, Et2Zn 2. I2

OMe

OMe

3. Satd Na2S2O3 4. DBU

O

Methyl 4,4-dimethyl-3oxopentanoate

Methyl (E)-5,5-dimethyl4-oxohex-2-enoate

(86%)

The carbene-mediated chain extension disconnection follows: O R

O

OR1 O

O

R

OR1

17.10 CONCLUSION This chapter has shown the great control and selectivity of modern radical reactions, especially intramolecular radical coupling reactions. Carbenes, especially those derived from diazoalkenes, have also played a prominent role in organic synthesis. The chemical reactions illustrated in this chapter are excellent additions to the list of nucleophilic, pericyclic, and electrophilic reactions presented previously. This chapter concludes the methodology for generating carbon-carbon bonds. With the functional group reactions in the first part of the book, all the tools for pursuing a synthesis are at hand. Chapter 18 will present reaction types that are generally categorized as organometallic reactions. The focus will be on organometallic reagents used to generate CdC bond.

358

Ronsheim, M. D. Zercher, C. K. J. Org. Chem. 2003, 68, 4535.

977

17.10 CONCLUSION

HOMEWORK

1. Predict the major product of the following reaction and briefly discuss why it is formed and other possible products are not (or are minor products). BrCCl3, hn O

2. Treatment of this molecule with Hg produced a radical. This radical product literally sat on the shelf for 25 years before the work was published. Give the structure of the radical and discuss its stability.

Cl

Hg

3. The structure of Fremy’s salt is shown. It is a rather stable free radical. Discuss the stability of this salt and also discuss its synthetic utility.

+

–

K O3S

O N

SO3– K +

Chem. Ind. 1953, 244 Ann. Chim. Phys. 1845, 15, 459

A

4. The two products of this reaction show stereochemical variation in the five-membered ring. Explain why the relative stereochemistry of the methyl group and two ring-juncture hydrogen atoms are the same in both products. H CO2Me

Ph

Me

Bu3SnH–AIBN

I

CO2Me PhH, Reflux

N Ts

Me Ts

Ph +

N

H CO2Me

Me Ts

H (20%)

Ph

N H (39%)

5. Explain the following results: H

H

CH :

CH :

Major product

Only this product H

H

6. Explain the following rate data for the reaction shown:  R (relative rate): Ph (17.0); n-BuO (3.9); MeO2C (2.0); n-C6H 13 (1.0); ClCH2 (0.7). R

HS

CH3 (CH 2)10

R

S

(CH 2)10

CH3

t-BuOOt-Bu, 60°C

7. Why is benzophenone added to the following reaction? Give a mechanistic rationale for this transformation. O CHO

CO2Me hn, Benzophenone

CO2Me Me

978

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

8. Give the mechanism for this transformation. O Me N N Me

t-Bu Benzene, 170°C

O CN

N

Bu3SnH, AIBN Syringe pump

Br

N Me

CN

9. Explain the formation of the indicated product from β-pinene and CCl4. CCl4, Bz2O2

Cl CCl3

10. Explain the following transformation: H

O

Bu3SnH, AIBN

H H

PhSe

O H

11. Explain the following transformation: O

EtO2C

hn, –10°C

N2 CO2Et

12. Two major cyclized products are produced in this reaction. Draw their structures and discuss the mechanism of their formation. H O

AIBN, Ph3SnH

O

SMe Me

13. Explain the following transformation: N2 O

1. Rh2(OOct)4, CH2Cl2 2. Excess LDA, TBSCl, THF 3. TBAF, THF

O

O

O

14. Three products are formed in this reaction: two minor and one major. Show all three products and indicate which one is the major product. Explain your choice. CuBr, CH2N2

979

17.10 CONCLUSION

15. In each case, show the major product. Indicate stereochemistry where appropriate. CHO CHO TiCl3 DME

CuCl, CuCl2

n-C3H7

N Cl

(A)

Zn(Cu)

O

(B)

O

O

O

Tl(OCOCF3)3

O

(C)

Bu3SnH

O

OH

O

CHO

(D) O OTBS O

COCF3

1. [Cp 2TiCl2], THF AlMe3

OPMB

SPh

2. Toluene, 110°C PMB = p-Methoxybenzyl

(E)

OBn OBn

O

Br EtO C3H7

(I)

SPh

O

(H) Bu3SnH, AIBN

CO2Me

I

O

Heat

PhH, 80°C

2 Ph

2. NH2NHSO2C6H4Me MeOH, Trace HCl 3. MeLi, Ether

O O

N2

CO2Et

O MeO Me N

O Ph

(L)

Ph HO Me

Rh2(Oct) 4, Toluene

(N)

TBDPSO

H

AcO Me

(O)

H

Me

n-C8 H17

H

1. NOCl 2. hn, PhMe

H

Me

Br Bu3SnH, AIBN PhH, 80°C Syringe pump

O

CHBr3, NaOH

OAc

CH2N2 THF, 0°C

O

CH2Cl2, –20°C

C5H11

60°C

(J)

1. Me2CuLi, Ether

(K)

CuCl, DMF

SiMe3

H

(M)

Br

Bu3SnH BEt3, O2

O

(G)

N

(F)

O

AIBN, PhH Bu3SnH, Reflux

N

(P)

EtO2C BnO

OBn

OH 1. CH2I2, PhH, ZnEt2, rt 2. 17 atm H2, MeOH PtO2, 10%HCl

(Q) CHO

(R)

TiCl4

O

Me CO2H

MeO2C

Bu4NI

(S)

hn, Acetone

e–, MeOH

(T)

N Me FeCl3, H2O

MeO

MeO

N

O

(V) OH

Ph

(W)

BnO

1. RuCl3, NaIO4 NaHCO3, aq MeCN

O

O Me

Br MeO

O

2. CH2N2, Ether

(X)

Ph

0.3% Rh2(S-DOSP)4 CH2Cl2

N

N2 O

HO

(U)

O

MOMO

N2

1. hn, (Me3Si)2I THF, 0 °C 2. H3O+

HO O HN N S O

O O

(Y)

1. BuLi , MeI 2. BuLi

PhI(O2CCF3)2

3. (HCHO) n

MeO

(Z)

BnO

N COCF 3

CF3CH2OH –25 °C, 30 min

980

17. FORMATION OF CARBON-CARBON BONDS VIA RADICALS AND CARBENES

16. In each case, provide a suitable synthesis from the designated SM. Show all intermediate products and all reagents. OEt

H O O

(A)

(B)

OTBS OH

TBSO

CN

TBSO O

O

H

H

O O

O H

(C)

OH

O

(D)

O

O O

O O

C H A P T E R

18 Metal-Mediated, Carbon-Carbon Bond-Forming Reactions 18.1 INTRODUCTION Among the many carbon-carbon bond-forming reactions presented in previous chapters, the acyl addition reactions of Grignard reagents in Section 11.4 and organolithium reagents in Section 11.6 are among the most important. Organolithium reagents were also shown to be remarkable reagents for a variety of synthetic applications. Organocuprate reagents (Section 12.3) are important reagents for coupling reactions with alkyl and acyl halides. All three of these reagents are characterized by the presence of a metal: Mg, Li, or Cu. Reagents bearing metals are known as organometallic reagents. Of the many advances in synthetic organic chemistry in the last 60 years or so, the use of organometallic reagents certainly rises to the top or near the top of any list. For the most part, the metal reagents serve as catalysts, but the variety of reactions that can be classified as organometallic reactions is truly remarkable. There are many books and reviews describing these reagents and their use in synthesis.1 This chapter will present selected organometallic reactions that have become mainstay reactions in organic synthesis, and many have acquired names.

18.2 COPPER-CATALYZED COUPLING REACTIONS 18.2.1 Aryl Coupling Reactions Many methods have been developed that form an aryl-aryl bond,2 including the metal-catalyzed radical reaction known as Meerwein arylation.3 It is also possible to make aryl-alkene bonds, and an example is the cupric chloride (CuCl2) catalyzed reaction of an aryl diazonium salt (e.g., benzenediazonium chloride) with an electron-deficient alkene (e.g., ethyl acrylate) to yield a phenyl substituted acrylate derivative (ethyl cinnamate). Under these conditions, CuCl2 induced loss of nitrogen (N2) from benzenediazonium chloride to yield the aryl radical (1), and formation of this radical was followed by addition to the alkene to yield 2. Subsequent reaction of this reactive intermediate with CuCl2 generated the alkene (ethyl cinnamate) via loss of HCl, generating CuCl at the same time. This reaction is similar to the

1

Selected examples include (a) Crabtree, R. H. The Organometallic Chemistry of the Transition Metals; Wiley: Hoboken, NJ, 2014. (b) Hartwig, J. Organotransition Metal Chemistry: From Bonding to Catalysis; University Science Books: South Orange, NJ, 2009. (c) Hegedus, L.; Soderberg, B. Transition Metals in the Synthesis of Complex Organic Molecules; University Science Books: South Orange, NJ, 2009. (d) Spessard, G. O.; Miessler, G. L. Organometallic Chemistry; Oxford University Press: Oxford, 2015. (e) Gupta, B. D.; Elias, A. J. Basic Organometallic Chemistry: Concepts, Syntheses and Applications; Universities Press: Hyderabad, 2013. (f) Jenkins, P. R. Organometallic Reagents in Synthesis (Oxford Chemistry Primers); Oxford University Press: Oxford, 1992.

2

Hassan, J.; Savignon, M.; Gozzi, C.; Schulz, E.; Lemaire, M. Chem. Rev. 2002, 102, 1359.

3

(a) Meerwein, H.; B€ uchner, E.; van Emster, K. J. Prakt. Chem. 1939, 152, 237. (b) Rondestvedt, C. S., Jr. Org. React. 1960, 11, 189. (c) Idem Ibid. 1976, 24, 225. (d) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-59.

Organic Synthesis http://dx.doi.org/10.1016/B978-0-12-800720-4.00018-0

981

Copyright © 2017 Michael Smith. Published by Elsevier Inc. All rights reserved.

982

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

Heck reaction that will be discussed in Section 18.4.1. Titanium trichloride (TiCl3) has been used to initiate 1,4-addition to conjugated carbonyl derivatives.4 There are many variations, and a wide range of aromatic diazonium salts can be used. One method for generating the diazonium salt is treatment of an aromatic amine with nitrous acid (HONO) via reaction of HNO3/HCl or NaNO2/ 2 HCl (see Section 3.10.6).5 The coupling of diazonium salts to other aromatic compounds6 in the presence of base is commonly called the Gomberg-Bachmann reaction. Treatment of 4-bromoaniline with nitrous acid, for example, gave the diazonium salt (4-bromobenzenediazonium chloride), which reacted with benzene to give the biphenyl derivative (4-bromo-1,10 -biphenyl) in 46% yield.7 The aromatic amine precursors are usually obtained by hydrogenation of nitrobenzene derivatives (Section 7.10.5).8

CuCl2

CO2Et

CuCl2

•

+

NLN Cl−

CO2Et

CO2Et

•

Benzenediazonium chloride

1 + N2 + CuCl + Cl2

2 equiv HCl

Br

NH2 4-Bromoaniline

Br

Ethyl cinnamate

2

N2 Cl−

NaNO2

4-Bromobenzenediazonium chloride

Br

40% aq NaOH

4-bromo-1,1 -biphenyl (46%)

A related process that involves diazonium salts is the Pschorr reaction,9 which also couples aryl diazonium compounds to other aromatic rings. This diazonium salt coupling can be done under acidic conditions, but addition of Cu powder usually promotes a radical process. Aryl amines generate aryl diazonium salts upon treatment with nitrous acid.10 An example is the reaction of 3 to yield an aryl diazonium salt, which cyclized in the presence of Cu to give thaliporphine in 43% yield. Kupchan et al.11called this transformation an improved Pschorr reaction. The Ullmann reaction12,13 is very similar in that aryl halides are coupled to form biaryls (e.g., biphenyl) by heating with Cu. Once again, the Cu facilitates formation of an aryl radical (in this case 1), which reacts with Cu+ to form an arylcopper (I) species (4). Subsequent reaction of the aryl copper with iodobenzene leads to a coupling reaction that yields 1,10 -biphenyl. This coupling reaction is related to the organocuprate coupling reactions discussed in Section 12.3.2.1. This reaction is similar to the Ni(0) coupling reactions discussed in Section 18.5,14,15 and see the Suzuki-Miyaura reaction in Section 18.4.3.

4

Citterio, A.; Cominelli, A.; Bonavoglia, F. Synthesis 1986, 308.

5

(a) Ridd, J. H. Q. Rev. Chem. Soc. 1961, 15, 418. (b) Hegarty, A. F. In The Chemistry of Diazonium and Diazo Groups, Part 2; Patai, S., Ed.; Wiley: New York, 1978; pp 511–591.

6

(a) Bachmann, W. E.; Hoffman, R. A. Org. React. 1944, 2, 224 (see pp 225–226). (b) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-37.

7

Gomberg, M.; Bachmann, W. E. J. Am. Chem. Soc. 1924, 46, 2339.

8

Rylander, P. N. Catalytic Hydrogenation Over Platinum Metal; Academic Press: New York, 1967; pp 168–202.

9

(a) Pschorr, R. Berichte 1896, 29, 496. (b) Leake, P. H. Chem. Rev. 1956, 56, 27. (c) DeTar, D. F. Org. React. 1957, 9, 409 (see pp 411–416). (d) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-76. (e) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; pp 534–535.

10

(a) Kametani, T.; Fukumoto, K.; Satoh, F.; Yagi, H. J. Chem. Soc. Chem. Commun. 1968, 1398. (b) Idem J. Chem. Soc. C. 1969, 520. (c) Kametani, T.; Ihara, M.; Fukumoto, K.; Yagi, H. Ibid. 1969, 2030. 11

Kupchan, S. M.; Kameswaran, V.; Findlay, J. W. A. J. Org. Chem. 1973, 38, 405.

12

(a) Ullmann, F.; Meyer, G. M.; Loewenthal, O.; Gilli, E. Annalen 1904, 332, 38. (b) Ullmann, F.; Sponagel, P. Berichte 1905, 38, 2211. (c) Fanta, P. E. Chem. Rev. 1946, 38, 139. (d) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-96. (e) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; pp 662–663.

13

Fanta, P. E. Chem. Rev. 1964, 64, 613.

14

Semmelhack, M.; Helquist, P.; Jones, L. D.; Keller, L.; Mendelson, L.; Ryono, L. S.; Smith, J. G.; Stauffer, R. D. J. Am. Chem. Soc. 1981, 103, 6460.

15

See Tsou, T. T.; Kochi, J. K. J. Am. Chem. Soc. 1979, 101, 7547.

983

18.2 COPPER-CATALYZED COUPLING REACTIONS

Me

Me

N

OMe

N

1. NaNO2 , 20% H2SO 4 AcOH 2. Cu powder, Heat

OMe

OMe

H2 N MeO

OH

MeO

OH

OMe

Thaliporphine (43%)

3

Aryl halide reactivity depends on the halide: I > Br > Cl.13 An electronegative group in the ortho position activates the reaction. Nitro groups on the aromatic ring are activating in this type of coupling16: ortho > para > meta. 2-Iodonitrobenzene is very reactive, but the 3- and 4-iodo derivatives are about as reactive as iodobenzene. Aromatic rings containing substituents readily undergo a coupling reaction in the presence of Cu metal or Cu salts.13 In a synthesis of (+)-isokotanin A, Bringmann et al.17 coupled methyl 2-bromo-3,5-dimethoxybenzoate with Cu in DMF to give symmetrical biaryl 5 in 89% yield. Functionalized aryls (e.g., ferrocene) can be coupled using Cu salts.18

I

Cu

I

CuI

•

Cu

1

1,1 -biphenyl

4

CO2Me

MeO2C CO2Me

Cu, DMF

MeO

Br

MeO

OMe OMe

OMe

MeO 5 (89%)

Methyl 2-bromo-3,5dimethoxybenzoate

For unsymmetrical coupling (two different aryl halides) the best yields are observed when one aryl halide is very reactive and the other is relatively unreactive. Reactive aryl halides have electronegative groups, (e.g., nitro or carbomethoxy), ortho to the halogen. Aryl bromides and chlorides are best for unsymmetrical coupling, while iodo compounds tend to yield symmetrical coupling (ArdAr rather than ArdAr0 ).19 Unreactive halides lack an ortho electronegative group and are usually iodo derivatives. When unsymmetrical coupling is attempted with two unreactive aryls, three biaryls are usually produced in approximately equal amounts.13 Ultrasound is effective for the synthesis of diaryl ethers via the Ullmann reaction.20 The disconnections for Cu promoted coupling follow: NH2

R1 R +

R1

R

NH2 +

R

R R

R1

R

X+ X

R1

16

Davey, W.; Latter, R. W. J. Chem. Soc. 1948, 264.

17

Bringmann, G.; Hinrichs, J.; Henschel, P.; Kraus, J.; Peters, K.; Peters, E.-M. Eur. J. Org. Chem. 2002, 1096.

18

Rausch, M. D. J. Org. Chem. 1961, 26, 1802.

19

Forrest, J. J. Chem. Soc. 1960, 566, 574, 581, 589, 592, 594.

20

Smith, K.; Jones, D. J. Chem. Soc. Perkin Trans. 1992, 1, 407.

984

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

18.2.2 Alkyne Coupling Reactions Coupling of alkynes is an important reaction that leads to diynes. Two classical alkyne-coupling reactions involve Cu derivatives, but a variety of conditions can be used.21 In the Glaser reaction,22 an alkyne [e.g., phenylacetylene (ethynylbenzene)] reacts with basic cupric chloride (CuCl2), and subsequent air oxidation yields a diyne. In this case, the product is 1,4-diphenylbuta-1,3-diyne, in 90% yield. This reaction also has been reported to give a quantitative yield of 1,4-diphenylbuta-1,3-diyne in the presence of CuCl2 and NaOAc, using supercritical CO2 as the solvent.23 The mechanism of this reaction has been studied, and Cu(II) has been proposed as an oxidant in the reaction.24 It has been shown that molecular oxygen forms adducts with Cu(I) supported by tertiary amines,25 which might be an intermediate, and mechanistic considerations26 for this variation have also been reported. The coupling can be done both inter- and intramolecularly.27 An example of the latter is the coupling of the two terminal alkyne units in 6 to give diyne 7 (65% yield), in Myers et al.28 synthesis of kedarcidin. This example clearly shows that the Glaser reaction is compatible with molecules bearing a vast array of functionality and stereochemistry. Glaser-type coupling of alkynes can be accomplished with Cu derivatives and bases other than hydroxide. Cupric acetate and pyridine are also an effective reagent.29 Note the similarity to the Sonogashira coupling to be discussed in Section 18.2.3 that uses a Pd catalyst. Ph

1. CuCl2 , NH4OH

H

Ph

2. Air

Ethynylbenzene

MeOH2CO

Ph

1,4-Diphenylbuta-1,3-diyne (90%)

O OSi(i-Pr)3

O

NH

MeOH2CO

OSi(i-Pr)3

O O

H N

OH

OH

O N

O

60 equiv Cu(OAc)2

OMe i-PrO

N Cl

OMe

O

Br

THF, 15 equiv CuI Py, 45°C

OMe

i-PrO OMe

Cl

Br O

t-BuMe2SiO 6

7 (65%)

OSiMe2t-Bu

The second important reaction is the Cadiot-Chodkiewicz coupling,30 in which a bromoalkyne reacts with a monoalkyne in the presence of cupric chloride and an amine to give the diyne.30a In a synthesis of panaxytriol by Danishefsky and Yun,31 Cadiot-Chodkiewicz cross-coupling of bromo-alkyne (R)-5-bromopent-1-en-4-yn-3-ol and alkyne (4R,5R)dodec-1-yne-4,5-diol gave a 63% yield of diyne (3R,9R,10R)-heptadeca-1-en-4,6-diyne-3,9,10-triol. It is also possible to

21

Cadiot, P.; Chodkiewicz, W. In Chemistry of Acetylenes; Viehe, H. G., Ed.; Marcel-Dekker: New York, 1969; pp 597–648.

22

(a) Glaser, C. Berichte 1869, 2, 422. (b) Idem Annalen 1870, 154, 137. (c) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-37. (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; pp 276–277. 23

Li, J.; Jiang, H. Chem. Commun. 1999, 2369.

24

Bohlmann, F.; Schoenowsky, H.; Inhoffen, E.; Grau, G. Chem. Ber. 1964, 97, 794.

25

Wieghardt, K.; Chaudhury, P. Prog. Inorg. Chem. 1988, 35, 329.

26

Fomina, L.; Vazquez, B.; Tkatchouk, E.; Fomine, S. Tetrahedron 2002, 58, 6741.

27

For an example taken from a synthesis of the kedarcidin core structure, see Myers, A. G.; Goldberg, S. D. Angew. Chem. Int. Ed. 2000, 39, 2732.

28

Myers, A. G.; Hogan, P. C.; Hurd, A. R.; Goldberg, S. D. Angew. Chem. Int. Ed. 2002, 41, 1062.

29

Brown, A. B.; Whitlock, Jr., H. W. J. Am. Chem. Soc. 1989, 111, 3640.

30

(a) Chodkiewicz, W.; Cadiot, P. Compt. Rend. 1955, 241, 1055. (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; pp 128–129.

31

Yun, H.; Danishefsky, S. J. J. Org. Chem. 2003, 68, 4519.

985

18.2 COPPER-CATALYZED COUPLING REACTIONS

couple alkynyl copper derivatives with aryl halides32,33 or other aromatic substrates.34 This methodology can also be applied in an Ullmann-type reaction, analogous to that described for aryl halides where a vinyl iodide is coupled in the presence of Cu.35 Alkynylsilanes can also be coupled using Cu(I) salts.36 Internal coupling reactions are also possible. Other metals can be used, and In metal37 was used to couple alkynes with allyl bromide with an anti-Markovnikov regiochemistry. C7H15

HO Br

+ OH

HO

HO CuCl, EtNH2, 0°C

HO

(4R,5R)-Dodec-1-yne4,5-diol

(R)-5-Bromopent-1-en4-yn-3-ol

C7H15

NH2OH•HCl, 1 h

OH

(3R,9R,10R)-Heptadeca-1-en(63%) 4,6-diyne-3,9,10-triol

The disconnections for Cu promoted coupling of alkynes and alkenes follow: R

R1

R

H + H

R1

R

R1

R

X +

X

R1

18.2.3 Stephens-Castro and Sonogashira Coupling Under certain conditions, alkynes can be coupled to aryl halides.38 When aryl halides react with copper acetylides to yield 1-aryl alkynes (e.g., 8), the reaction is known as Castro-Stephens coupling.39 Both aliphatic and aromatic substituents can be attached to the alkyne unit, and a variety of aryl iodides have been used. A Pd catalyzed variation is also known in which an aryl halide or a vinyl derivative reacts with a terminal alkyne to yield coupling products. In this variation, called the Sonogashira coupling,40 the reaction is catalyzed by a Pd-Cu complex at or near ambient temperatures. In a synthesis of rubriflordilactone A, Li and coworker’s41 reacted vinyl iodide 9 and trimethylsilylethyne in the presence of Pd(0) and cuprous iodide to give 10 in 99% yield. Several new reaction variations have brought renewed attention to the coupling. A Cu free Sonogashira coupling has been reported in an ionic liquid medium.42 A combination of Pd(PhCN)2Cl2 and (t-Bu)3P is very effective for Sonogashira coupling at room temperature,43 and the reaction has been done without solvent, on alumina, using microwave irradiation.44

32

(a) Posner, G. H. An Introduction to Synthesis Using Organocopper Reagents; Wiley: New York, 1980; p 41. (b) House, H. O.; Umen, M. J. J. Org. Chem. 1973, 38, 3893.

33

(a) Hatchard, W. R. J. Org. Chem. 1963, 28, 2163. (b) Stevens, R. D.; Castro, C. E. Ibid. 1963, 28, 3313.

34

(a) Bohlmann, F.; Kleine, K.-M.; Arndt, C. Berichte 1966, 99, 1642. (b) Bohlmann, F.; Bonnet, P.-H.; Hofmeister, H. Ibid. 1967, 100, 1200. (c) Bohlmann, F.; Zdero, C.; Gordon, W. Ibid. 1967, 100, 1193.

35

Cohen, T.; Poeth, T. J. Am. Chem. Soc. 1972, 94, 4363.

36

Nishihara, Y.; Ikegashira, K.; Hirabayashi, K.; Ando, J.-I.; Mori, A.; Hiyama, T. J. Org. Chem. 2000, 65, 1780.

37

(a) Araki, S.; Imai, A.; Shimizu, M.; Yamada, A.; Mori, Y.; Butsugan, Y. J. Org. Chem. 1995, 60, 1841. (b) Ranu, B. C. Eur. J. Org. Chem. 2000, 2347.

38

For reactions of this type, see Larock, R. C. Conmprehensive Organic Transformations, 2nd ed.; Wiley–VCH: New York, 1999; pp 596–599.

39

(a) Castro, C. E.; Stephens, R. D. J. Org. Chem. 1963, 28, 2163. (b) Stephens, R. D.; Castro, C. E. J. Org. Chem. 1963, 28, 3313. (c) Sladkov, A. M.; Ukhin, L. Yu.; Korshak, V. V. Bull. Acad. Sci. USSR. Div. Chem. Sci. 1963, 2043. (d) Sladkov, A. M.; Gol’ding, I. R. Russ. Chem. Rev. 1979, 48, 868. (e) Bumagin, N. A.; Kalinovskii, I. O.; Ponomarov, A. B.; Beletskaya, I. P. Doklad. Chem. 1982, 265, 262. (f) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-16. (g) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 136–137. 40

(a) Sonogashira, K.; Tohda, Y.; Hagihara, N. Tetrahedron Lett. 1975, 4467. (b) Sonogashira, K. In Comprehensive Organic Synthesis; Trost, B. M.; Fleming, I., Eds.; Pergamon Press: New York, 1999; Vol. 3, Chapter 2.4. (c) Rossi, R.; Carpita, A.; Bellina, F. Org. Prep. Proceed. Int. 1995, 27, 127. (d) Sonogashira, K. In Metal-Catalyzed Cross-Coupling Reactions; Diederich, F.; Stang, P. J., Eds.: Wiley–VCH: New York, 1998; Chapter 5. (e) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-88. (f) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; pp 612–613. 41

Li, J.; Yang, P.; Yao, M.; Deng, J.; Li, A. J. Am. Chem. Soc. 2014, 136, 16477.

42

Fukuyama, T.; Shinmen, M.; Nishitani, S.; Sato, M.; Ryu, I. Org. Lett. 2002, 4, 1691.

43

Hundertmark, T.; Littke, A. F.; Buchwald, S. L.; Fu, G. C. Org. Lett. 2000, 2, 1729.

44

Kabalka, G. W.; Wang, L.; Namboodiri, V.; Pagni, R. M. Tetrahedron Lett. 2000, 41, 5151.

986

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

RJCLCJCu

R

I Iodobenzene

8 SiMe3

I Me3Si

OH

, CuI

PdCl2(PPh3)2

OH

9

10 (99%)

18.3 π-ALLYL PALLADIUM COMPLEXES While Cu reagents have been widely used for coupling reactions, modern variations use Pd compounds. The Sonogashira coupling is one example. In general, allyl palladium salts react with allylic hydrocarbons and with functionalized allylic derivatives to produce π-allyl complexes. These complexes can react with nucleophilic species, including enolate anions, to produce new carbon-carbon bonds.45 In such reactions, π-allyl Pd complexes function as Ca equivalents (see Section 1.2).

18.3.1 The Wacker Process The first reaction to be discussed is a traditional one, but it probably does not proceed via a π-complex. The Wacker process (or the Wacker oxidation) is used in industry to convert ethylene to acetaldehyde using soluble Pd catalysts.46 An example is the reaction of ethylene with cupric chloride (CuCl2) and palladium chloride (PdCl2) to yield acetaldehyde. Other alkenes can be oxidized to carbonyls as well, as in the conversion of an alkene to a ketone.47 Both aldehydes and ketones can be formed, usually in moderate-to-excellent yields. Conjugated dienes (e.g., buta-1,3-diene) yield conjugated aldehydes under these conditions. Cyclic alkenes give the expected cyclic ketones. Vinyl and allylic halides are converted to the corresponding carbonyl compound.48,49 Wacker oxidation has been reported using molecular oxygen. A catalytic amount of palladium(II) acetate, pyridine, and oxygen converted dodec-1-ene to dodecan-2-one, in up to 70% yield.50 Similarly, cyclopentene was oxidized to cyclopentanone under an oxygen atmosphere using palladium acetate and molybdovanadophosphate on activated carbon.51 H2CKCH2

O

CuCl2 , PdCl2 H 2O

Me

H

The Wacker oxidation can be used with more complex substrates than those shown, or to construct relatively simple fragments as part of a longer synthesis. An example of the former application is taken from Mehta and Shinde’s52 synthesis of merrillianone, in which the terminal alkene unit in 11 was converted to methyl ketone (12) in 80% yield.

45

Poli, G.; Giambastiani, G.; Heumann, A. Tetrahedron 2000, 56, 5959.

46

(a) Smidt, J.; Hafner, W.; Jira, R.; Sieber, R.; Sedlmeier, J.; Sabel, A. Angew. Chem. Int. Ed. Engl. 1962, 1, 80. (b) Jira, R.; Freiesleben W. Organomet. React. 1972, 3, 1. (c) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-98. (d) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; pp 676–677. 47

Smidt, J.; Hafner, W.; Jira, R.; Sedlmeier, J.; Sieber, R.; R€ uttinger, R.; Kojer, H. Angew. Chem. 1959, 71, 176.

48

Byrom, N. T.; Grigg, R.; Kongkathip, B. J. Chem. Soc. Chem. Commun. 1976, 216.

49

Clement, W. H.; Selwitz, C. M. J. Org. Chem. 1964, 29, 241.

50

Nishimura, T.; Kakiuchi, N.; Onoue, T.; Ohe, K.; Uemura, S. J. Chem. Soc. Perkin. Trans. 2000, 1, 1915.

51

Kishi, A.; Higashino, T.; Sakaguchi, S.; Ishii, Y. Tetrahedron Lett. 2000, 41, 99.

52

Mehta, G.; Shinde, H. M. J. Org. Chem. 2012, 77, 8056.

987

18.3 π-ALLYL PALLADIUM COMPLEXES

OTBS

O

Me

O

OTBS

O

Me

O PdCl2, CuCl, O2

Me

Me

rt, 8 h

OMOM

OMOM

O

O O 11

12 (80%)

18.3.2 Formation of π-Allyl Palladium Complexes The Wacker process involved alkenes, but when the alkene has an allylic position, a new type of organometallic can be formed. With Pd reagents, prop-1-ene reacts to form a π-allyl metal complex (e.g., 13), stabilized by back-donation from the metal atom. A substitution reaction of this π-allyl complex with a suitable nucleophile will generate the allylic species, (14), where X is the nucleophile. Early work by H€ uttel and Christ,53 and also by Volger54 established that π-allyl complexes could be prepared, but often in very poor yield.53c This low yield was probably due to preferential formation and attempted oxidation of a hydropalladium species (e.g., 15, derived from cyclohexene), which readily decomposed. Heck55 showed that arylpalladium chlorides added to conjugated dienes. It was later discovered that oxidation of 15 converts it to the more stable halo-bridged dimer (16).56 This observation allowed π-allyl palladium complexes to be examined for their synthetic utility. Trost et al.56 reacted allylic compounds with disodium tetrachloropalladate (Na2PdCl4) in the presence of sodium acetate (NaOAc) and acetic acid (AcOH), for example, and found that halogen dimers were formed. The alkene was used in excess, and a ratio of 2:1 alkene/Pd was optimal. A variety of alkenes can be converted to their Pd complex using Trost’s conditions, including pent-(2Z)-ene, which generated an 87:13 mixture of 17/18 in 83% yield. M

+

H

X−

M

X 13 X Pd H

PdX2

14 X

Oxidation

Pd

15

16

NaCl, PdCl2, AcOH

+

NaOAc, CuCl 2 (83%)

Pd X

X

17

PdCl/2

PdCl/2 18

This Pd mediated coupling reaction follows a Markovnikov orientation (see Sections 2.5.1 and 10.2.1), with the removed hydrogen atom being allylic to the most substituted end, as in the removal of Ha in the conversion (E)-4methylhept-3-ene to 19. Addition of cupric chloride (CuCl2) influences the ratio of the Markovnikov-like to the anti-Markovnikov-like product.56 Palladation of 2-methylbut-2-ene without CuCl2 removed Ha to yield 20 (71% 20 + 29% 21). Addition of CuCl2 led to removal of Hb to yield 74% of 21 and 26% of 20.57 The general order56 for removal of allylic protons appears to be CH3 > CH2 ≫ CH. Conformational effects can alter this order or reactivity, however. The syn-complex usually predominates, as in the reaction of 4-methylhept-3-ene to yield 19. Both (E)- and (Z)-isomers of the alkene led to the syn-isomer 19, without a trace of the anti-isomer (22).

53

(a) H€ uttel, R.; Christ, H. Chem. Ber. 1963, 96, 3101. (b) Idem Ibid. 1964, 97, 1439. (c) H€ uttel, R.; Dietl, H.; Christ, H. Ibid. 1964, 97, 2037. (d) H€ uttel, R.; Christ, H.; Herzig, K. Ibid. 1964, 97, 2710. (e) H€ uttel, R.; Dietl, H. Ibid. 1965, 98, 1753. (f) H€ uttel, R.; Schmid, H. Ibid. 1968, 101, 252.

54

Volger, H. C. Recl. Trav. Chim. Pays-Bas, 1969, 88, 225.

55

Heck, R. F. J. Am. Chem. Soc. 1968, 90, 5542.

56

Trost, B. M.; Strege, P. E.; Weber, L.; Fullerton, T. J.; Dietsche, T. J. J. Am. Chem. Soc. 1978, 100, 3407.

57

H€ uttel, R.; McNiff, M. Chem. Ber. 1973, 106, 1789.

988

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

Me

Hb

Me

Hb

– Ha

PdCl/2

Ha (E)-4-Methylhept-3-ene

19

Ha

PdCl/2

Ha PdCl2, CuCl2

PdCl/2

– Hb

21

PdCl2 , no CuCl2 – Ha

Hb 2-Methylbut-2-ene

Hb 20 Me

Me

Me

PdCl2

+ PdCl/2 (E)-4-Methylhept-3-ene

PdCl/2

19

22

More highly substituted alkenes react faster (trisubstituted > disubstituted > monosubstituted),57 as illustrated by the reaction of 4,8-dimethylnona-1,7-diene to yield 23. Allylic systems conjugated to a carbonyl react slower than unconjugated alkenes. Conjugated alkenes do react if there is no competition.56,58 PdCl2

PdCl/2 4,8-Dimethylnona-1,7-diene

23

Trost et al.56 proposed the mechanism shown with 24-26 to explain formation of π-allyl complexes from alkenes. A mechanism for the Pd mediated arylation of alkenes by arylmercury compounds had been proposed earlier by Heck.59 Initial reaction with PdCl4 generates a π-alkene complex (24), which is in equilibrium with the π-allyl palladium hydride complex 25. The presence of a base (e.g., an amine or a phosphine) or CuCl2 led to removal of the hydrogen from PddH to yield the π-allyl palladium dimer (2). The CuCl2 oxidative procedure is more efficient than addition of a base. PdCl42−

H

PdCl4

– Cl− + Cl−

H

Cl

Pd Cl

Cl

– Cl− + Cl−

24 Base or CuCl2

PdCl/2

HCl

25

Cl

Pd H Cl 26

18.3.3 Reactions of π-Allyl Palladium Complexes Initial reactions with π-allyl Pd complexes involved treatment with nucleophiles. Heck60 reported the arylation, methylation, and carboxyalkylation of alkenes by organomercury compounds, mediated by Pd, but the use of other transition metal derivatives was also reported. In later work, Trost et al.61 found that organopalladium reagents (e.g., 27) reacted with diethyl sodiomalonate in the presence of triphenylphosphine (PPh3). The phosphine acted as a basic ligand, and coordinated with the Pd. The major product from reaction with sodium malonate was 28, formed in 68% yield. This reaction is a practical method to form a new carbon bond by reacting what is effectively a carbocation 58

(a) Jones, D. N.; Knox, S. D. J. Chem. Soc. Chem. Commun. 1975, 165. (b) Harrison, I. T.; Kimura, E.; Bohme, E.; Fried, J. H. Tetrahedron Lett. 1969, 1589. (c) Susuki, T.; Tsuji, J. Bull. Chem. Soc. Jpn. 1973, 46, 655.

59

Heck, R. F. J. Am. Chem. Soc. 1969, 91, 6707.

60

Heck, R. F. J. Am. Chem. Soc. 1968, 90, 5518.

61

Trost, B. M.; Weber, L.; Strege, P. E.; Fullerton, T. J.; Dietsche, T. J. J. Am. Chem. Soc. 1978, 100, 3416.

989

18.3 π-ALLYL PALLADIUM COMPLEXES

surrogate with a carbon nucleophile. The distribution of products shows a preference for attack at the less substituted carbon [the ratio (28 + 29)/30 was 8:1, and the ratio 28/29 was 1.6:1]. When 27 reacted with methyl methanesulfonyl sodioacetate (sodium salt of methyl methanesulfonate), only one product was observed (31) in 80% yield.61 (EtO2C)2HC PdCl/2

CH(CO2Et)2

[NaH, CH2(CO2Et)2]

+

PPh3, THF

+ CH(CO2Et)2

27

28 (68%)

29

30

MeO2S

PdCl/2

SO2Me

NaH, PPh3, THF

+

MeO2C

rt, 45 min

CO2Me

31 (80%)

27

The extent of attack at the less substituted carbon was dependent on the nature and amount of the added phosphine ligand.61 Many common ligands can be used for the conversion of 32 to the π-allyl complex 33 using various ligands, which then reacted with methyl methanesulfonyl sodioacetate to form a mixture of 34 and 35. For most phosphine ligands in DMSO or THF, 34 was the major product in a 2:1-to-10:1 ratio. A variety of carbanion nucleophiles can be reacted with π-allylpalladium complexes.62 SO2Me

Ligand

PdCl/2

NaH

Ligand

SO2Me

+

CO2Me

+ Methyl methanesulfonyl sodioacetate

Solvent

32

CO2Me

33

34

35

When tri-o-tolylphosphine (TOT) was used in place of HMPT, the regioselectivity was reversed.61 The coordinating ligand is critically important, but there does not seem to be a strong solvent effect on the stereochemistry of the reaction. Addition of bis(phosphine) ligands [e.g., bis(diphenylphosphinoethane)], dppe, generates a complex (35) rather than the dimeric palladium chloride complex observed previously. Reaction of 32 with methyl methanesulfonyl sodioacetate in the presence of dppe gave a 4:1 mixture of 34 and 35. A synthetic application is the Marchand et al.63 reaction of 36 and the allylic sulfone anion (from 37) gave a 52% yield of 38. Elimination of the sulfone moiety generated vitamin A alcohol 39.

Me OAc

PdCl/2 36

Me

Me

SO2Ph

Me

Me

Me

SO2Ph Me

OAc

NaH, PPh3

+

DMF

Me

Me 37

38 (52%) Me

Me

Me

NaOEt NHEt2

Me 39

62

Trost, B. M.; Weber, L.; Strege, P. E.; Fullerton, T. J.; Dietsche, T. J. J. Am. Chem. Soc. 1978, 100, 3426.

63

Manchand, P. S.; Wong, H. S.; Blount, J. F. J. Org. Chem. 1978, 43, 4769.

Me

OH

990

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

The π-allyl palladium coupling disconnections follow: R

R2 R1

R1

R

R2

+

R3

R3

18.3.4 Catalytic π-Allyl Palladium Reactions The allyl-palladium alkylation process discussed in Sections 18.3.2 and 18.3.3 can be made catalytic in Pd. In early work, Heck64 showed that the Pd catalyzed arylation of enol esters gave aryl-substituted alkenes in modest yield. Melplder and Heck also reported the Pd catalyzed arylation of allylic alcohols with aryl halides.65 The basis for this catalytic activity involves the use of an allylic substituted species (allylic acetate, allylic chloride, etc.) rather than an allylic hydrocarbon. A tetrakis-palladium(0) complex (L4Pd) reacts with the allylic substrate to form a π-allylpalladium complex (40). Substitution with a nucleophile generates 41 with liberation of Pd species L4Pd, which is recycled to react with additional allylic starting material.66

X

+

Pd

L4Pd L

Nuc

Nuc

L

40

+

L4Pd

41

Three of the common Pd(0) catalysts are tetrakis-(triphenylphosphino)palladium (42) and bis[1,2-bis-(diphenylphosphino)ethane] palladium [43, i.e., Pd(dppe)2], prepared by reduction of palladium acetate [Pd(OAc)2] in the presence of triphenylphosphine or dppe.66c,d The third is tris(dibenzylideneacetone)dipalladium(0), Pd2(dba)3, which has structure 44. Ph

[ Ph3P ] 4 Pd L

Ph3P

Pd

PPh3 PPh3

Ph3P

Ph Ph P Ph

Ph P Ph Pd

P

P

Ph

Ph Ph

Pd Pd2(dba)3 L

O Pd Ph

42

43

3

44

18.3.4.1 Reactions With Nucleophiles Previously, a π-allylic Pd complex was generated by reaction of Pd reagents with allylic hydrocarbons prior to reaction with nucleophiles. In the catalytic version of this reaction, an allylic halide or an allylic acetate is used with a Pd(0) reagent. Why use a Pd complex when enolate alkylation is a well-known process (see Section 13.3.1)? A typical enolate coupling reaction is the conversion of 2-methylcyclopentane-1,3-dione to the enolate anion by reaction with NaOH, allowing reaction with allyl bromide. Under these conditions only 34% of 2-allyl-2-methylcyclopentane-1,3-dione was obtained. When allyl acetate was used in place of allyl bromide in this reaction and tetrakis(triphenylphosphino)palladium was used as a catalyst, a 94% yield of 2-allyl-2-methylcyclopentane-1,3-dione was obtained.67 In this reaction, formation of the π-allyl Pd complex facilitated coupling with the nucleophilic enolate derived from 2-methylcyclopentane-1,3-dione, which exhibited poor reactivity in the normal enolate alkylation sequence. This version of the π-allyl Pd reaction uses an allylic acetate or chloride. The use of the acetate is more common, in part because the acetate ion is a weaker nucleophile than chloride ion. More recently, methyl carbonate groups have 64

Heck, R. F. J. Am. Chem. Soc. 1968, 90, 5535.

65

Melplder, J. B.; Heck, R. F. J. Org. Chem. 1976, 41, 265.

66

(a) Trost, B. M.; Verhoeven, T. R. J. Am. Chem. Soc. 1976, 98, 630. (b) Idem Ibid. 1978, 100, 3435. (c) Takahashi, K.; Miyake, A.; Hata, G. Bull. Chem. Soc. Jpn. 1970, 45, 230, 1183. (d) Trost, B. M.; Verhoeven, T. R. J. Org. Chem. 1976, 41, 3215. (e) Idem J. Am. Chem. Soc. 1980, 102, 4730. 67

(a) Newman, M. S.; Manhart, J. H. J. Org. Chem. 1961, 26, 2113. (b) Unpublished observations of Trost, B. M.; Curran, D. reported in Reference 31 of Trost, B. M. Acc. Chem. Res. 1980, 13, 385.

991

18.3 π-ALLYL PALLADIUM COMPLEXES

been shown to be good leaving groups in this reaction.68 When it involves a substrate where diastereomeric products can result, the stereochemistry of the nucleophilic displacement is an important issue. The addition of chiral ligands and chiral additives lead to asymmetric induction.69,70 Palladium assisted alkylation proceeds with net retention of configuration of the acetate or chloride. In the case of 45, alkylation gave 46 in 91% yield in Wood and coworker’s71 synthesis of actinophyllic acid. A variety of active methylene compounds can be used as nucleophiles,72 and enolate anions as well.73 Intramolecular cyclization is possible when the active methylene compound and an allylic acetate or carbonate is incorporated into the same molecule.74 This reaction has been called Trost-allylation, although it is also known as the Tsuji-Trost reaction.75 Br, NaOH

Me O

O

rt, 4 h

(34%)

Me O

O

OAc

2-Methylcyclopentane-1,3-dione

(94%)

2-Allyl-2-methylcyclopentane-1,3-dione

O

O Bn

Pd(PPh3)4

Excess MeOCO2allyl 10% Pd(PPh3)4

N

THF, rt

O

O Bn

N

OTBDPS H 45

OTBDPS H 46 (91%)

Trost et al.76 proposed the following mechanism to account for these catalytic transformations. Reaction of the Pd catalyst with (R,Z)-pent-3-en-2-yl acetate generates π-alkene-palladium complex (47). Palladium removes the allylic hydrogen atom, and loss of the acetate moiety yields π-allyl palladium complex 48. Nucleophilic attack at Ca leads to 49, with expulsion of the PdL2 species, whereas attack at Cb leads to 50. Palladium coordinates on the face of the alkene distal to the acetate (distant from the acetate: Ca rather than Cb). Palladium displaces acetate with inversion (47 ! 48). When the nucleophile displaces the Pd, a second inversion occurs at Ca or Cb, whichever is less sterically hindered, to give a net retention of configuration for the conversion (R,Z)-pent-3-en-2-yl acetate ! 49 and/or 50. Organopalladium catalysts are now common for the coupling reactions of alkenes, alkynes, or aryls.77

68

For a —OCO2Ph leaving group, see Ito, K.; Kashiwagi, R.; Hayashi, S.; Uchida, T.; Katsuki, T. Synlett 2001, 284.

69

(a) Kuwano, R.; Kondo, Y.; Matsuyama, Y. J. Am. Chem. Soc. 2003, 125, 12104. (b) Faller, J. W.; Wilt, J. C. Tetrahedron Lett. 2004, 45, 7613.

70

(a) Evans, P. A.; Uraguchi, D. J. Am. Chem. Soc. 2003, 125, 7158. (b) Evans, P. A.; Leahy, D. K.; Slieker, L. M. Tetrahedron Asymm. 2003, 14, 3613. (c) Evans, P. A.; Leahy, D. K. J. Am. Chem. Soc. 2000, 122, 5012. (d) López, F.; Ohmura, T.; Hartwig, J. F. J. Am. Chem. Soc. 2003, 125, 3426. (e) Matsushima, Y.; Onitsuka, K.; Kondo, T.; Mitsudo, T.-A.; Takahashi, S. J. Am. Chem. Soc. 2001, 123, 10405. (f) Ohmura, T.; Hartwig, J. F. J. Am. Chem. Soc. 2002, 124, 15164. (g) Evans, P. A.; Robinson, J. E.; Baum, E. W.; Fazal, A. N. J. Am. Chem. Soc. 2002, 124, 8782. (h) Bandini, M.; Melloni, A.; Umani-Ronchi, A. Org. Lett. 2004, 6, 3199.

71

Vaswani, R. G.; Day, J. J.; Wood, J. L. Org. Lett. 2009, 11, 4532.

72

Kazmaier, U.; Zumpe, F. L. Angew. Chem. Int. Ed. 1999, 38, 1468.

73

(a) You, S.-L.; Hou, X.-L.; Dai, L.-X.; Zhu, X.-Z. Org. Lett. 2001, 3, 149. (b) Evans, P. A.; Leahy, D. K. J. Am. Chem. Soc. 2003, 125, 8974. (c) Evans, P. A.; Lawler, M. J. J. Am. Chem. Soc. 2004, 126, 8642.

74

(a) Castaño, A. M.; Mendez, M.; Ruano, M.; Echavarren, A. M. J. Org. Chem. 2001, 66, 589. (b) Zhang, Q.; Lu, X.; Han, X. J. Org. Chem. 2001, 66, 7676.

75

The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-94.

76

Trost, B. M.; Crawley, M. L. Chem. Rev. 2003, 103, 2921.

77

(a) Sonogashira, K.; Tohda, Y.; Hagihara, N. Tetrahedron Lett. 1975, 4467. (b) Dieck, H. A.; Heck, R. F. J. Organomet. Chem. 1975, 93, 259. (c) Edo, K.; Yamanaka, H.; Sakamoto, T. Heterocycles 1978, 9, 271. (d) Ohsawa, A.; Abe, Y.; Igeta, H. Chem. Lett. 1979, 241.

992

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

b

AcO H

AcO

L4Pd

−

–

H

OAc

Nuc

50 Nuc

PdL2 +

−

OAc

H

a

48

47

(R, Z )-Pent-3-en2-yl acetate

Nuc

PdL2

H

49

Nuc

When chiral ligands are used with the Pd catalyst, good enantioselectivity can be achieved in the alkylation reaction.76 Two common ligands78 are the bis(phosphine) complex from chlorodiphenylphosphine, and optically active amino alcohols (e.g., 51). Phosphino-ester binaphthyls (e.g., 52) have also been used.79 The enantioselectivity can be quite good, as when the allylic acetate (E)-1,3-diphenylallyl acetate, reacted with diethyl malonate in the presence of 52 and N,O-bis(trimethylsilyl)acetamide (BSA) and LiOAc as additives, to give an 89% yield of dimethyl (R,E)-2(1,3-diphenylallyl)malonate with 99 %ee.79 The identical reaction with (E)-1,3-diphenylallyl acetate, but using 51 as a ligand gave a 93% yield of dimethyl (R,E)-2-(1,3-diphenylallyl)malonate, but in only 56 %ee.78 Ph

O

Ph

NiPr2

O Me N O PPh2 PPh2

PPh2

51 OAc Ph

52

[Pd(

Ph

CO2Me

MeO2C

CH2(CO 2Me)2 , BSA, LiOAc 3-C H )Cl] 3 5 2 52, rt, 1 d

Ph

(E)-1,3-Diphenylallyl acetate

Ph

Dimethyl (R, E)-2-(1,3diphenylallyl)malonate

18.3.4.2 Palladium-Mediated Cyclization Reactions It is possible to use organopalladium chemistry for intramolecular coupling (cyclization) reactions. Treatment of 1-methoxy-5,5-dimethyloct-7-en-2-yn-4-ol with the Pd catalyst ([dba]3Pd2, where dba ¼ dibenzylideneacetone, 44) led to cyclization, and generated the five-membered ring in 53, in 78% yield.80 There is a template effect in the π-allyl palladium reactions81 that can be exploited to form large rings (Section 4.5). Ester 54 was converted to lactone 55 in 70% yield by this procedure, for example, but the reported yields were poor in some cases.81 HO

OH

OMe (dba)3Pd2 - CHCl3

Me Me

Me

3

P , 5% AcOH

53

O

CO2Me O

Me

48 h

1-Methoxy-5,5-dimethyloct-7-en-2-yn-4-ol O

OMe

Me

SO2Ph

NaH, Pd(PPh3)4

O

THF, Reflux

OAc

PhO2S 54

CO2Me

55 (70%)

78

Gong, L.; Chen, G.; Mi, A.; Jiang, Y.; Fu, F.; Cui, X.; Chan, A. S. C. Tetrahedron: Asymm. 2000, 11, 4297.

79

Kodama, H.; Taiji, T.; Ohta, T.; Furukawa, I. Tetrahedron: Asymm. 2000, 11, 4009.

80

Trost, B. M.; Lee, D. C.; Rise, F. Tetrahedron Lett. 1989, 30, 651.

81

(a) Trost, B. M.; Verhoeven, T. R. J. Am. Chem. Soc. 1980, 102, 4743. (b) Idem Ibid. 1977, 99, 3867. (c) Idem Tetrahedron Lett. 1978, 2275.

993

18.3 π-ALLYL PALLADIUM COMPLEXES

18.3.4.3 Palladium-Mediated Arylation Reactions In the presence of Pd catalysts, enolate anions of malonic esters react with aryl halides to yield 2-aryl malonates.82 The reaction of diethyl malonate with sodium hydride and bromobenzene, with the Pd(0) catalyst generated in situ, gave an 89% yield of diethyl 2-phenylmalonate. Similar coupling reactions had been reported earlier, but aryl iodides were used as substrates and stoichiometric amounts of Cu derivatives were required.83 The arylation of malonates and cyanoesters was later reported using catalytic amounts of copper derivatives.84 CO2Et CO2Et

CO2Et

NaH, PhBr, THF, 70°C, 1 h

Ph CO2Et

2% Pd(dba)2, 4% P(t-Bu)3

Diethyl malonate

Diethyl 2-phenylmalonate (89%)

18.3.4.4 Trimethylenemethane Equivalents Palladium catalysts can be used to convert trimethylsilyl allylic acetate[2-((trimethylsilyl)methyl)allyl acetate] to a trimethylenemethane equivalent (TMM, 56). A [3+2]-cycloaddition reaction (Section 15.4) with alkenes generates cyclopentanes in what is called quinane annulation.85 In this reaction, the trimethylsilyl unit is a carbanion equivalent and acetate is a carbocation equivalent. In an example taken from a synthesis of (+)-brefeldin A, Trost and Crawly86 reacted 57 and 2-((trimethylsilyl)methyl)allyl acetate in the presence of 2.5 mol% palladium acetate and triisopropyl phosphite [P(Oi-Pr)3] (toluene,100°C) to generate 58 in 93% yield (>93:2 dr). The reaction is characterized by retention of the geometry of the dipolarophile in the cycloadduct, and proceeds with high diastereoselectivity and good regioselectivity. This reaction can also be applied to the formation of larger rings if a diene is used as the dipolarophile.87 Trost later showed that (Z)-alkenes show greater selectivity in this cyclization process.88 Reagents related to 2-((trimethylsilyl)methyl)allyl acetate react with aldehydes, as well as alkenes, to produce THF derivatives.89 Alkynes also react with the TMM equiv to form methylenecyclopentene derivatives.90

OAc

SiMe3

2-((Trimethylsilyl)methyl)allyl acetate

56

20% P(Oi-Pr)3, PhMe 2.5% Pd(OAc)2, 12 h

O

H O

2-((Trimethylsilyl)methyl)allyl acetate 100°C

O

H 57

O 58 (93%)

When TMM equivalents are tethered to alkenes, intramolecular cyclization generates bicyclic methylenecyclopentane derivatives. Cyclization of 59, for example, led to a 51% yield of 60 along with 18% of uncyclized triene 61.91 The cyclization reaction proceeded with excellent diastereoselectivity, as shown. Trost and MacPherson87 has also

82

Beare, N. A.; Hartwig, J. F. J. Org. Chem. 2002, 67, 541.

83

(a) Osuka, A.; Kobayashi, T.; Suzuki, H. Synthesis 1983, 67. (b) Setsune, J.-I.; Matsukawa, K.; Wakemoto, H.; Kitao, T. Chem. Lett. 1981, 367.

84

Okuro, K.; Furuune, M.; Miura, M.; Nomura, M. J. Org. Chem. 1993, 58, 7606.

85

Trost, B. M. Angew. Chem. Int. Ed. 1986, 25, 1.

86

Trost, B. M.; Crawley, M. L. J. Am. Chem. Soc. 2002, 124, 9328.

87

Trost, B. M.; MacPherson, D. T. J. Am. Chem. Soc. 1987, 109, 3483.

88

Trost, B. M.; Yang, B.; Miller, M. L. J. Am. Chem. Soc. 1989, 111, 6482.

89

(a) Trost, B. M.; King, S. A. J. Am. Chem. Soc. 1990, 112, 408. (b) Idem Tetrahedron Lett. 1986, 27, 5971.

90

Trost, B. M.; Balkovec, J. M.; Angle, S. R. Tetrahedron Lett. 1986, 27, 1445.

91

Trost, B. M.; Grese, T. A.; Chan, D. M. T. J. Am. Chem. Soc. 1991, 113, 7350.

994

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

examined the attachment of chiral auxiliaries to the alkene partner to yield asymmetric induction in reactions with TMM equivalents. H SiMe3

1. BSA, THF, rt, 1.5 h

+

2. dppe, Pd(PPh3)4 Reflux, 42 h

OMe CO2Me

H

59

CO2Me

CO2Me

60 (51%)

61 (18%)

The disconnections for catalytic allyl palladium reactions are essentially the same as for the uncatalyzed version. A few representative examples follow: R2

OH R3

R R1

+ R

R1

R

R

R2 X

R3 X1

+ OH

R1

OAc SiMe3

R1

18.4 NAMED PALLADIUM-CATALYZED COUPLING REACTIONS As organometallic chemistry in general, and organopalladium chemistry in particular, have become more popular, many new reactions have been developed. Several are so useful that they have joined the list of named reactions. This section will explore several named reactions that involve Pd catalysts.

18.4.1 The Heck Reaction There are several variations of the organopalladium reaction, some involving alkenyl palladium and others σ-Pd complexes. One variation generates the Pd species from a heterocyclic alkene. As mentioned in Sections 18.3.2–18.3.4, Heck reported several Pd mediated reactions, including the coupling of aryl compounds with alkenes. Heck and Nolley92 later showed that substituted alkenes could be prepared by the Pd catalyzed coupling of alkenes and aryl halides, and this transformation is now known as the Heck reaction.93 The Heck reaction usually involves the coupling of aryl halides or aryl sulfonate esters with alkenes.94 As Heck reported, the reaction was discovered independently at about the same time by Mizoroki et al.93d,e so the reaction is often called the Mizoroki-Heck reaction. Aryl Pd complexes that are formed via oxidative coupling of aryl halides with Pd(0) can undergo the Heck reaction.95,96 The reaction proceeds with initial formation of aryl-Pd complex 62, which reacts with the alkene to form a Pd species (e.g., 63), with elimination of Pd to yield arylated alkene derivatives (e.g., styrene).97 Aryl halides differ greatly in their reactivity, with ArI being the most reactive followed by aryl bromides. In general, aryl chlorides are very unreactive in the Heck reaction unless the chlorine is attached to a heteroaromatic ring.

92

Heck, R. F.; Nolley, J. P., Jr. J. Org. Chem. 1972, 37, 2320.

93

(a) See Heck, R. F. Synlett 2006, 2855, and the articles in special issue No. 18 of Synlett, 2006; 2855–3184. (b) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-42. (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed; Wiley-Interscience: Hoboken, NJ, 2005; pp 296–297. (d) Mizoroki, T.; Mori, K.; Ozaki, A. Bull. Chem. Soc. Jpn. 1972, 44, 581. (e) Mizoroki, T.; Mori, K.; Ozaki, A. Bull. Chem. Soc. Jpn. 1973, 46, 1505.

94

For other reactions of this type, see Reference 38, pp 386–392.

95

Heck, R. F. Org. React. 1982, 27, 345.

96

(a) Patel, P. A.; Ziegler, C. B.; Cortese, N. A.; Plevyak, J. E.; Zebovitz, T. C.; Terpko, M.; Heck, R. F. J. Org. Chem. 1977, 42, 3903. (b) Cortese, N. A.; Ziegler, C. B., Jr.; Hrnjez, B. J.; Heck, R. F. Ibid. 1978, 43, 2952. (c) Mori, M.; Kudo, S.; Ban, Y. J. Chem. Soc. Perkin Trans. 1979, 1, 771.

97

For a review of mechanistic considerations for the Heck reaction, see Amatore, C.; Jutland, A. Acc. Chem. Res. 2000, 33, 314.

995

18.4 NAMED PALLADIUM-CATALYZED COUPLING REACTIONS

PdX

X

Pd

CH2KCH2

PdX 62

Ph

Ph

Ph3P Pd

+

63

Ph3P

NEt3, 80°C

Ph

Ph

Cl Ph

+ (Ph P) Pd 3 2

trans-Stilbene

64

65

Aryl triflates are partners in this reaction. Coupling occurs with a rather wide variety of alkenes and aryl derivatives. Styrene reacted with phenyl bis(triphenylphosphino)palladium chloride (64), in the presence of triethylamine, to give trans-stilbene with loss of the Pd complex 65.98 The Pd reagent (64) was generated in situ by reaction of palladium acetate [Pd(OAc)2] and triphenylphosphine in the presence of iodobenzene. The reaction is thought to proceed by syn addition of the organopalladium, followed by syn elimination of palladium hydride.98 Heterocyclic substrates can be used in the Heck reaction. In Pettus and coworker’s99 synthesis of γ-rubromycin, aryl iodide 66 reacted with methyl acrylate to give 67 in 93% yield. New catalysts have been developed that allow Heck coupling with deactivated aryl chlorides.100 Beller and coworkers100 showed that chlorobenzene reacted with styrene, for example, to give trans-stilbene in 63% yield in the presence of K2CO3 and a Pd catalyst having a di(1-adamantyl)-nbutylphosphine ligand. A ligand-free version of the Heck reaction (does not require the use of phosphine ligands) was recently reported, using palladium acetate and potassium phosphate in N,N-dimethylacetamide.101 Aryl chlorides undergo the Heck reaction at ambient temperatures, with high selectivity using a Pd/P(t-Bu)3/Cy2NMe catalyst.102 OHC

OHC

MeO2C

1.9 CH2KCHCO2Me

I

MeO

Pd(OAc)2, PPh3 . LiCl 1.8 NEt3, DMF, 80°C

OH

CO2Me MeO2C MeO

66

OH 67 (93%)

A Heck-like reaction is possible when vinyl iodides react with conjugated carbonyl compounds in the presence of a Pd catalyst. An example is the reaction of vinyl iodide (68) with methyl acrylate, in the presence of a Pd catalyst, to give 69 in 92% yield as part of a synthesis of kopsiyunnanine E by Takayama and coworkers.103 CO2Me

BocHN

I

CH2KCHCO2Me, NEt3

OTBS

BocHN

Pd(OAc)2, PPh3 DMF, 100°C

68

OTBS 69 (92%)

The intramolecular version of this reaction is well known104,105 and quite useful in natural product synthesis, including asymmetric syntheses.106 An example of this variation is taken from Jia and coworker’s107 synthesis of (+)-lysergic 98

Dieck, H. A.; Heck, R. F. J. Am. Chem. Soc. 1974, 96, 1133.

99

Wu, K.-L.; Mercado, E. V.; Pettus, T. R. R. J. Am. Chem. Soc. 2011, 133, 6114.

100

Ehrentraut, A.; Zapf, A.; Beller, M. Synlett 2000, 1589.

101

Yao, Q.; Kinney, E. P.; Yang, Z. J. Org. Chem. 2003, 68, 7528.

102

Littke, A. F.; Fu, G. C. J. Am. Chem. Soc. 2001, 123, 6989.

103

Kitajima, M.; Murakami, Y.; Takahashi, N.; Wu, Y.; Kogure, N.; Zhang, R.-P.; Takayama, H. Org. Lett. 2014, 16, 5000.

104

Zhang, Y.; Negishi, E. J. Am. Chem. Soc. 1989, 111, 3454.

105

Abelman, M. M.; Oh, T.; Overman, L. E. J. Org. Chem. 1987, 52, 4130.

106

Dounay, A. B.; Overman, L. E. Chem. Rev. 2003, 103, 2945.

107

Liu, Q.; Zhang, Y.-A.; Xu, P.; Jia, Y. J. Org. Chem. 2013, 78, 10885.

996

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

acid, where the Pd catalyzed reaction of 70 gave an 84% yield of 71, based on recovered starting material. Other highly enantioselective intramolecular Heck reactions have been reported using a monodentate ligand.108 MeO2C

MeO2C N Me

Me H

H

I

N

Pd(OAc)2, PPh3 Ag2CO3, MeCN, 80°C (84%, Based on recovered starting material)

N

N Boc

Boc 70

71

There are several variations that involve heteroatom substituents, including biaryl ethers, amines, or ketones, which can be coupled to alkenes using palladium acetate.109 Medium-ring heterocycles can be prepared via an intramolecular Heck reaction.110 Silyl enol ethers can also be coupled to another alkene upon treatment with palladium acetate to form conjugated ketones.111 Solvents can change with the reactive partners. In one variation, aryldiazonium salts are used rather than aryl halides for coupling to alkenes, using Pd catalysts.112a Carbohydrate-based vinyl bromides have been coupled to methyl acrylate using Pd(dba)2 in aq DMF.113 Alkynes have been coupled to pyrrole using Heck-like conditions.114 Recyclable Pd(II) catalysts with bis(imidazole) ligands are used with Heck reactions in ionic liquids.115 A noncatalytic Heck reaction was reported using supercritical water as the medium.116 The use of chiral ligands allows asymmetric induction in some Heck reactions.117 Conjugate addition is usually preferred when the alkene contains an electron-withdrawing group. Without such a group, addition occurs at the less substituted carbon.98 The Heck and related disconnections follow: R

R2

R + R1

X

+

R R

R

X

R1

R1

R2

Y

R +

X R1

X

18.4.2 Stille Coupling Vinyl triflates (C]CdOSO2CF3) react with vinyltin derivatives to yield dienes, in what is known as the Stille coupling.118,119 Preparation of vinyl triflate (72) is illustrated by trapping the enolate of 2-methylcyclohexanone (also see Sections 13.3.2 and 13.4.3) with N-phenyl triflimide,120 and reaction with a vinyltin compound in the presence of 108

Imbos, R.; Minnaard, A. J.; Feringa, B. L. J. Am. Chem. Soc. 2002, 124, 184.

109

(a) Åkermark, B.; Eberson, L.; Jonsson, E.; Pettersson, E. J. Org. Chem. 1971, 40, 1365. (b) Itahara, T.; Sakakibara, T. Synthesis 1978, 607. (c) Itahara, T. Ibid. 1979, 151. 110

Arnold, L. A.; Luo, W.; Guy, R. K. Org. Lett. 2004, 6, 3005.

111

Ito, Y.; Aoyama, H.; Hirao, T.; Mochizuki, A.; Saegusa, T. J. Am. Chem. Soc. 1979, 101, 494.

112

(a) Sengupta, S.; Bhattacharya, S. J. Chem. Soc. Perkin Trans. 1993, 1, 1943. (b) For a Heck coupling with an aryl diazonium salt used in a synthesis of (–)-codonopsinine see Severino, E. A.; Correia, C. R. D. Org. Lett. 2000, 2, 3039; and in a synthesis of the defucogilvocarcin M chromophore, see Patra, A.; Pahari, P.; Ray, S.; Mal, D. J. Org. Chem. 2005, 70, 9017.

113

Hayashi, M.; Amano, K.; Tsukada, K.; Lamberth, C. J. Chem. Soc. Perkin Trans. 1999, 1, 239.

114

Lu, W.; Jia, C.; Kitamura, T.; Fujiwara, Y. Org. Lett. 2000, 2, 2927.

115

Park, S. B.; Alper, H. Org. Lett. 2003, 5, 3209.

116

Zhang, R.; Zhao, F.; Sato, M.; Ikushima, Y. Chem. Commun. 2003, 1548.

117

Deng, W.-P.; Hou, X.-L.; Dai, L.-X.; Dong, X.-W. Chem. Commun. 2000, 1483.

118

(a) Scott, W. J.; Crisp, G. T.; Stille, J. K. J. Am. Chem. Soc. 1984, 106, 4630. (b) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-89. (c) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; pp 620–621.

119

For reactions involving organotin compounds, see Reference 79, pp 434–437.

120

McMurry, J. E.; Scott, W. J. Tetrahedron Lett. 1983, 24, 979.

997

18.4 NAMED PALLADIUM-CATALYZED COUPLING REACTIONS

tetrakis-triphenylphosphino palladium(0) gave diene 73, quantitatively, as a 95:5 (E,E)/(E,Z) mixture.120 Coupling two different alkene units in this way has been referred to as Stille cross-coupling, and vinyl iodides are commonly used.121 The reaction can also be applied to coupling aryl units with vinyl or allylic units. In the Kuwahara and Nagasawa122 synthesis of lactimidomycin A1, vinyltin 74 was coupled to vinyl iodide 75, in the presence of Pd2(dba)3 to give an 89% yield of lactimidomycin A1. Note that aryl bromides have been coupled to tributylvinyltin compounds.123 Me

OTf

O Me

Me

1. LDA

Pd(PPh3)4, LiCl SnBu3

2. Tf2NPh

Me3Si

Me3Si

72 (91%)

73

I Bu3Sn

Pd2(dba)3•CHCl3

OTES

+

PhSe

SePh

OTES

EtO2C

LiCl, CuCl

OTMS

OTMS

CO2Et

Lactimidomycin A1 (89%)

75

74

Mechanistic rationales for the Stille reaction have been reviewed,124 and the Stille coupling has been shown to be catalytic in tin.125,126 Alkynes are important partners in the Stille reaction. The reaction of alkyne 76 with the vinyl bromide (E)-(2-bromovinyl)benzene, in the presence of a Pd catalyst and 6% Me3SnCl, for example, gave a 90% yield of 77.125 This reaction required syringe pump addition of the vinyl bromide over 15 h to the reaction medium, but the yields were quite good. 6% Me3SnCl, Ether 1% PdCl2(PPh3)2 aq Na2CO3, PMHS

+

Br

Ph

OH

Ph

1% Pd2(dba)3, 4% (2-furyl)3P 37°C, 15 h Syringe pump addition

OH 77 (90%)

(E)-(2-Bromovinyl)benzene

76

SnBu3 I O

Pd2(dba)3, LiCl

O

O

O

DMF, rt

O O

O

O

OMe

OMe PMP

PMP 78

79 (>53%)

121 For the use of a vinyl iodide in a Stille coupling, in a synthesis of elysiapyrones A and B, see Barbarow, J. E.; Miller, A. K.; Trauner, D. Org. Lett. 2005, 7, 2901. 122

Nagasawa, T.; Kuwahara, S. Org. Lett. 2013, 15, 3002.

123

For an example taken from a synthesis of ageladine A, see Meketa, M. L.; Weinreb, S. M. Org. Lett. 2006, 8, 1443.

124

Espinet, P.; Echavarren, A. M. Angew. Chem. Int. Ed. 2004, 43, 4704.

125

Maleczka, R. E., Jr.; Gallagher, W. P.; Terstiege, I. J. Am. Chem. Soc. 2000, 122, 384.

126

The mechanism of the Stille coupling has been studied. See Casado, A. L.; Espinet, P.; Gallego, A. M. J. Am. Chem. Soc. 2000, 122, 11771.

998

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

A method has been developed that allows coupling with aryl chlorides, which are generally less reactive.127 There has been increasing interest in intramolecular Stille reactions in organic synthesis.128 An example is taken from Suenaga and coworker’s129 synthesis of biselyngbyolide A, in which 78 was converted to 79 in >53% yield. The Stille coupling disconnections follow: R

R1

R

O

R1

O

(CH 2)m

+ SnR23

R

R

(CH 2)m

(CH 2)n

I

O

(CH 2)n

Bu3Sn

18.4.3 Suzuki-Miyaura Coupling The reaction of an organoboronic acid [RB(OH)2] with an aryl halide leads to formation of biaryls. The use of organoboronic acids has a connection with the chemistry presented in Sections 9.2.2 and 9.5.1 and 9.5.2 for organoboranes. This particular transformation involves organopalladium catalysts, so it is related to the Heck reaction and Stille coupling and is presented as a synthetic route to biaryls.130 Suzuki and coworkers131 found that aryl triflates react with arylboronic acids [ArB(OH)2], in the presence of a Pd catalyst, to give biaryls in a reaction that is now known as Suzuki coupling132 or, more commonly, Suzuki-Miyaura coupling. In an example taken from a synthesis of pareitropone by Kim and coworkers133, boronic acid (4-((triisopropylsilyl)oxy)phenyl)boronic acid reacted with 2-bromo-3,4-dimethoxybenzaldehyde in the presence of Pd(PPh3)4 to give 80 in near quantitative yield.134 Even hindered boronic acids give good yields of the coupled product. The reaction is not restricted to simple aromatic compounds. Heteroaromatic compounds can be used, as illustrated by the reaction of a borylated indole, 81, with 5-bromo-2-iodo-3-methoxypyrazine to give a 76% yield of 82 in Stoltz and coworker’s135 synthesis of dragmacidin D. In addition to boronic acid derivatives, organoboranes react with aryl triflates under these conditions to form biaryls.136 A microwave-promoted Suzuki-type coupling was reported that did not involve the use of transition metal compounds.137 MeO

OMe MeO

B(OH)2

Br

+

OMe

Pd(PPh3)4, Na2CO3

OTIPS

TIPSO

CHO

CHO

2-Bromo-3,4-dimethoxybenzaldehyde

(4-((Triisopropylsilyl)oxy)phenyl)boronic acid

80 (quant) Ts

Br

Ts N

N

I

+ Br

N

OMe

N

Pd(PPh3)4, MeOH PhH, 80°C, 72 h

Br N

Na2CO3, H2O

(HO) 2B Br

5-Bromo-2-iodo-3methoxypyrazine

N

81

OMe 82 (76%)

127

Littke, A. F.; Fu, G. C. Angew. Chem. Int. Ed. 1999, 38, 2411.

128

For a review, see Duncton, M. A. J.; Pattenden, G. J. Chem. Soc. Perkin Trans. 1999, 1, 1235.

129

Tanabe, Y.; Sato, E.; Nakajima, N.; Ohkubo, A.; Ohno, O.; Suenaga, K. Org. Lett. 2014, 16, 2858.

130

For reactions of this type, see Reference 38, pp 362–364.

131

(a) Miyaura, N.; Yanagi, T.; Suzuki, A. Synth. Commun. 1981, 11, 513. (b) Cheng, W.; Snieckus, V. Tetrahedron Lett. 1987, 28, 5097. (c) Badone, D.; Baroni, M.; Cardomone, R.; Ielmini, A.; Guzzi, U. J. Org. Chem. 1997, 62, 7170.

132

(a) Miyaura, N.; Suzuki, A. Chem. Rev. 1995, 95, 2457. (b) Farinola, G. M.; Fiandanese, V.; Mazzone, L.; Naso, F. J. Chem. Soc. Chem. Commun. 1995, 2523. (c) Gen^et, J. P.; Linquist, A.; Blart, E.; Mouriès, V.; Savignac, M.; Vaultier, M. Tetrahedron Lett. 1995, 36, 1443. (d) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-92. (e) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; pp 636–637. 133

(a) Watanabe, T.; Miyaura, N.; Suzuki, A. Synlett 1992, 207. (b) Chaumeil, H.; Signorella, S.; Le Drian, C. Tetrahedron 2000, 56, 9655.

134

Hong, S.-K.; Kim, H.; Seo, Y.; Lee, S. H.; Cha, J. K.; Kim, Y. G. Org. Lett. 2010, 12, 3954.

135

Garg, N. K.; Sarpong, R.; Stoltz, B. M. J. Am. Chem. Soc. 2002, 124, 13179.

136

F€ urstner, A.; Seidel, G. Synlett 1998, 161.

137

Leadbeaater, N. E.; Marco, M. J. Org. Chem. 2003, 68, 5660.

999

18.4 NAMED PALLADIUM-CATALYZED COUPLING REACTIONS

The reaction is not restricted to the formation of biaryls, and there are other synthetic applications.138 Aryl boronic acids can also react with vinyl halides and vinyl triflates in the presence of Pd(0). Another modification uses alkynes as coupling partners. In a synthesis of fiscuseptine, Bracher and Daab139 coupled aryl iodide 83 with propargyl alcohol, in the presence of the Pd catalyst and CuI, to give an 81% yield of 84. MeO

MeO

OMe

OMe

HCLCCH2OH, CuI cat PdCl2(PPh3)2 , NEt3

N

OH N

I

83

84 (81%)

Thallium(I) ethoxide140 is known to promote Suzuki-Miyaura cross coupling reactions for vinyl or arylboronic acids with vinyl and aryl halide partners. A mixture of Pd2(dba)3 and tri-tert-butylphosphine has proven to be very effective.141 Nickel on charcoal,142 Rh catalysts,143 and silver oxide-mediated Pd catalysts144 have also been used. The latter catalyst was effective for the preparation of 4-substituted 2(5H)-furanones.145 Palladium-on-carbon was shown to be a reusable catalyst in aqueous media.146 Another variation of this coupling reaction modified the boronic acid moiety. Molander and Rivero147 showed the use of potassium trifluoroborates and a Pd catalyst leads to a cross-coupling reaction with aryl and vinyl halides. In a synthesis of cortistatin J, Funk and Nilson148 coupled potassium 2-trimethylsilyltrifluorovinylborate with vinyl triflate 85 to give an 84% yield of 86. Potassium organotrifluoroborates (RBF3K) are prepared by the addition of inexpensive KHF2 to organoboron intermediates,149 and they can be used in several of the applications where boronic acids or esters are used.150

TMS TfO

N

TMS

N

BF3K

O

O Cs2CO3, Pd(PPh3)4

H OH

H 85

OH

138

Kotha, S.; Lahiri, K.; Kashinath, D. Tetrahedron 2002, 58, 9633.

139

Bracher, F.; Daab, J. Eur. J. Org. Chem. 2002, 2288.

140

Frank, S. A.; Chen, H.; Kunz, R. K.; Schnaderbeck, M. J.; Roush, W. R. Org. Lett. 2000, 2, 2691.

141

Littke, A. F.; Dai, C.; Fu, G. C. J. Am. Chem. Soc. 2000, 122, 4020.

142

Lipshutz, B. H.; Sclafani, J. A.; Blomgren, P. A. Tetrahedron 2000, 56, 2139.

143

Ueda, M.; Miyaura, N. J. Org. Chem. 2000, 65, 4450.

144

Chen, H.; Deng, M.-Z. J. Org. Chem. 2000, 65, 4444.

145

Yao, M.-L.; Deng, M.-Z. J. Org. Chem. 2000, 65, 5034.

146

Sakurai, H.; Tsukuda, T.; Hirao, T. J. Org. Chem. 2002, 67, 2721.

147

Molander, G. A.; Rivero, M. R. Org. Lett. 2002, 4, 107.

148

Nilson, M. G.; Funk, R. L. J. Am. Chem. Soc. 2011, 133, 12451.

86 (84%)

149

(a) Vedejs, E.; Chapman, R. W.; Fields, S. C.; Lin, S.; Schrimpf, M. R. J. Org. Chem. 1995, 60, 3020. (b) Vedejs, E.; Fields, S. C.; Hayashi, R.; Hitchcock, S. R.; Powell, D. R.; Schrimpf, M. R. J. Am. Chem. Soc. 1999, 121, 2460. 150

(a) Molander, G. A.; Ito, T. Org. Lett. 2001, 3, 393. (b) Molander, G. A.; Biolatto, B. Org. Lett. 2002, 4, 1867. (c) Molander, G. A.; Biolatto, B. J. Org. Chem. 2003, 68, 4302. (d) Molander, G. A.; Katona, B. W.; Machrouhi, F. J. Org. Chem. 2002, 67, 8416. (e) Molander, G. A.; Yun, C.; Ribagorda, M.; Biolatto, B. J. Org. Chem. 2003, 68, 5534. (f) Molander, G. A.; Ribagorda, M. J. Am. Chem. Soc. 2003, 125, 11148.

1000

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

18.5 π-ALLYL NICKEL COMPLEXES Nickel can stabilize allylic cations in a manner similar to Pd.151 Indeed, the use of organonickel complexes for generating carbon-carbon bonds152 is older than the analogous organopalladium chemistry. Nickel chloride (NiCl2) is the Ni analogue of PdCl2 and reacts similarly to form zero-valent complexes [Ni(0)], which are the reactive species in π-allylnickel complexes (e.g., 87). A common Ni(0) complex is nickel tetracarbonyl [Ni(CO)4], but it is a volatile, extremely toxic, and dangerous material. Nickel tetracarbonyl should be used with great caution and only in a specially ventilated hood.

Tetrakis(triphenoxyphosphino)nickel is also a useful organonickel reagent. The bis-complex of nickel and 1,5-cyclooctadiene (cod), 88 [Ni(cod)2], bis(diphenylphosphinoethane)nickel chloride (89), and tetraphenylnickel (90) are widely used Ni(0) complexes. L

Li4Ni

Ni

Nuc

Nuc

L

87 Ph Ph P

Ni

Ni

Cl 88

Ph Ph

P

Ph

Ph Ni Ph

Ph

Cl 89

90

A typical reaction involving Ni(0) treats an allylic halide (e.g., 3-chloro-2-methylprop-1-ene) with Ni(CO)4 to form 91,153 a Ni dimer analogous to the palladium-chloride dimer discussed in Section 18.4. This complex reacts with an alkene (e.g., norbornene) to form a new complex (92), and subsequent reaction with a variety of electrophilic reagents removes Ni. In this example, the chlorine was converted to an acetate ligand (in 93) and reacted with methanolic CO to yield the ester (94).153 Cl

Ni(0) PhH , rt

Cl Ni

Ni

Ni

Cl

Cl 3-Chloro-2-methylprop-1-ene

91

92 CO , MeOH

NaOAc

MgO

Ni 93

OAc

CO2Me 94

π-Nickel complexes facilitate the coupling of vinyl ethers and vinyl sulfides with Grignard reagents. Wenkert et al.154 showed that the vinyl sulfide (E)-methyl(oct-1-en-1-yl)sulfane was converted to (E)-oct-1-en-1-ylbenzene in 80% yield (as a 4:1 trans/cis mixture) by reaction with phenylmagnesium bromide and bis(diphenylphosphinoethane)nickel chloride. Zero-valent nickel can also catalyze the coupling of aryl halides and alkenes.154a,b,155

151

For reactions of this type, see Reference 38, pp 116–117.

152

For a review, see Montgomery, J. Acc. Chem. Res. 2000, 33, 467.

153

(a) Gallazzi, M. C.; Hanlon, T. L.; Vitulli, G.; Porri, L. J. Organomet. Chem. 1971, 33, C45. (b) Hughes, R. P.; Powell, J. Ibid. 1971, 30, C45. (c) Guerrieri, F.; Chiusoli, G. P. Ibid. 1968, 15, 209. 154

(a) Wenkert, E.; Ferreira, T. W.; Michelotti, E. L. J. Chem. Soc. Chem. Commun. 1979, 637. (b) Wenkert, E.; Michelotti, E. L.; Swindell, C. S. J. Am. Chem. Soc. 1979, 101, 2246.

155

Mori, M; Ban, Y. Tetrahedron Lett. 1976, 1803, 1807.

1001

18.5 π-ALLYL NICKEL COMPLEXES

A Ni catalyzed ene cyclization (see Section 15.6) has been reported that uses Ni(cod)2. The reaction proceeds by initial formation of a π-allylnickel complex, which facilitates the intramolecular ene reaction with an allylic amine unit.156 π-Allylnickel complexes can be used in coupling reactions with both aryl and alkyl halides. Enolate anions react with Ni(0) reagents to form a complex that subsequently couples to aryl iodides. Semmelhack et al.157 final step in the synthesis of cephalotaxinone treated 95 with Ni(cod)2 to produce the target in 94% yield. C6H13

C6H13

0.1 (diphos)NiCl 2 PhMgBr

MeS

Ph (E)-Oct-1-en-1- (80%) ylbenzene

(E)-Methyl(oct-1-en1-yl)sulfane

O

N

N

O

1. LiN(i-Pr)2

I

O

O

2. Ni(cod)2

O

O OMe

OMe Cephalotaxinone(94%)

95

The Ni coupling reaction is not limited to aryl halides. Alkyl halides also react, as seen in Marshall and Wuts’158 coupling of 96 with aπ-allyl nickel dimer to yield dictyolene. Me

Me H Me

Me I

H

H Me 96

H

DMF

H H OAc Me Dictyolene

NiBr/2

OAc

The nickel(0) coupling disconnections follow: R R

R

R1

R1

Nuc

R

R3

R1

R2

X

R3 +

R1

X

R2

Y

Br

PhSSPh h (> 350 nm)

4 Ni(CO)4 , 50°C

Br

25°C, 2.5 h N

97

O

98

Me

Humulene

O

O O

O

Ni(cod)2 , PPh3

H Me Me

Me

H

H

90°C

Me Me

99

Me H 100 (67%)

156

Oppolzer, W.; Bedoya-Zurita, M.; Switzer, C. Y. Tetrahedron Lett. 1988, 29, 6433.

157

Semmelhack, M. F.; Chong, B. P.; Stauffer, R. D.; Rogerson, T. D.; Chong, A.; Jones, L. D. J. Am. Chem. Soc. 1975, 97, 2507.

158

Marshall, J. A.; Wuts, P. G. M. J. Am. Chem. Soc. 1978, 100, 1627.

1002

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

Zero-valent Ni is especially effective for the cyclization of bis(allyl) bromides.159,160 Corey et al.159 used this technique to cyclize 97 to 98. Subsequent photolysis in the presence of diphenyl disulfide (PhSSPh) gave humulene. Wender et al.161 reported a novel [4 + 4]-cycloaddition reaction in which two diene moieties are coupled with a Ni(0) catalyst to form eight-membered rings. In a synthesis of asteriscanolide, 99 was converted to 100 in 67% yield using the nickel-cyclooctadiene catalyst, cod, with RCH2CH] > R2CHCH] > R2C].168 Metathesis reactions169 are well known, using both homogeneous170 and heterogeneous171 catalysts. Although other catalysts have been used,172 Ru complexes173 are the most common in modern applications, including the important Grubbs I catalyst (107).174 With some of the early applications, alkenes containing functional groups175 did not work very well with the Grubbs I catalyst. This limitation prompted the development of alternative metathesis catalysts including 108 (the so-called second generation Grubbs catalyst, or Grubbs II catalyst, where Mes ¼ mesitoate and Cy ¼ cyclohexyl) and the Mo-based Schrock catalyst (109).176 With these new catalysts, the reaction has taken on increased synthetic utility. The most common application is formation of cyclic alkenes [ring-closing metathesis (RCM) reactions].

166

(a) Kroll, W. R.; Doyle, G. Chem. Commun. 1971, 839. (b) Grubbs, R. H.; Miller, S. J.; Fu, G. C. Acc. Chem. Res. 1995, 28, 446.

167

Calderon, N.; Ofstead, E. A.; Ward, J. P.; Judy, W. A.; Scott, K. W. J. Am. Chem. Soc. 1968, 90, 4133.

168

(a) McGinnis, J.; Katz, T. J.; Hurwitz, S. J. Am. Chem. Soc. 1976, 98, 605. (b) Casey, C. J.; Tuinstra, H. E.; Saeman, M. C. J. Am. Chem. Soc. 1976, 98, 608.

169

Grubbs, R. H.; Miller, S. J.; Fu, G. C. Acc. Chem. Res. 1995, 28, 446.

170

(a) Calderon, N.; Chen, H. Y.; Scott, K. W. Tetrahedron Lett. 1967, 3327. (b) Hughes, W. B. Organomet. Chem. Synth. 1972, 1, 362. (c) Toreki, R.; Schrock, R. R. J. Am. Chem. Soc. 1990, 112, 2448.

171

Banks, R. L.; Bailey, G. C. Ind. Eng. Chem. Prod. Res. Dev. 1964, 3, 170.

172

Banks, R. L. Fortschr. Chem. Forsch. 1972, 25, 41.

173

(a) Gilbertson, S. R.; Hoge, G. S.; Genov, D. G. J. Org. Chem. 1998, 63, 10077. (b) Maier, M. E.; Bugl, M. Synlett, 1998, 1390. (c) Stefinovic, M.; Snieckus, V. J. Org. Chem. 1998, 63, 2808.

174

(a) Schwab, P.; Grubbs, R. H.; Ziller, J. W. J. Am. Chem. Soc. 1996, 118, 100. (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2005; p 783. 175

(a) Mol, J. C. Chemtech 1983, 250–255. (b) Bosma, R. H. A.; van den Aardweg, G. C. N.; Mol, J. C. J. Organomet. Chem. 1983, 255, 159. (c) Idem Ibid. 1985, 280, 115. (d) Xiaoding, X.; Mol, J. C. J. Chem. Soc. Chem. Commun. 1985, 631. (e) Crisp, C. T.; Collis, M. P. Aust. J. Chem. 1988, 41, 935.

176

(a) Bazan, G. C.; Oskam, J. H.; Cho, H.-N.; Park, L. Y.; Schrock, R. R. J. Am. Chem. Soc. 1991, 113, 6899, and references cited therein. (b) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; Wiley–Interscience: Hoboken, NJ, 2005; p 845.

1004

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

PCy3

Cl

Ru Cl

N

Mes Cl

PCy3 Ph

N

F3C F3C

Mes

Cl

PCy3

F3C

Ph

N

O

Me

Ru

Mo Ph

O

F3C

Me Me

107

108

109

Me

The Grubbs catalysts are stable to Lewis acid conditions.177 Catalysts have been developed that allow the reaction to take place in aqueous and alcohol solutions,178 and polymer-bound catalysts have been used.179 Microwave-assisted ringclosing metastasis reactions have also been reported.180 There are some problems in handling the Grubbs’ I catalyst (107). Specifically, prolonged exposure to air and moisture deactivates the complex. Therefore, the catalyst is usually stored under an inert atmosphere. Attaching the active Ru complex to a polymer support improves the stability of the catalyst while retaining its reactivity in ring-closing metathesis.181 Taber and Frankowski182 reported that Grubbs catalyst dispersed in paraffin is easily handled, and retains its activity indefinitely with no special storage precautions. Catalyst 108 is more stable and more active than 107, and there are many synthetic examples that employ 108 rather than 107. Intramolecular metathesis reactions dominate synthetic applications, and terminal alkene moieties containing a]CH2 unit are used so that one product is ethylene. Ethylene escapes from the reaction medium and drives the equilibrium toward the cyclized product. A representative example is taken from Ohba and Nakata’s183 synthesis of paecilomycin B, in which diene 110 was treated with Grubbs II catalyst 108 to yield the cyclized product 111 in 81% yield, along with the volatile ethylene.

O

OH

O O

OMe OTBS

110

OH

O O

16 h

MOMO BnO

O

108, Reflux CH2Cl2 (5 mM)

+ CH2KCH2 OMe

MOMO BnO

OTBS 111 (81%)

As observed with 111, compounds with heteroatoms in the ring being formed are easily prepared.184 Another example is taken from Nishida and coworker’s185 synthesis of nakadomarin A, where the two vinyl units in 112 were cyclized to form the eight-membered lactam ring in 113 using 108.186 One of the more attractive features of this methodology is the ability to form macrocyclic lactones, as seen in the synthesis of 111 and also 114 from 115 in >55% yield,

177

Bentz, D.; Laschat, S. Synthesis 2000, 1766.

178

(a) Kirkland, T. A.; Lynn, D. M.; Grubbs, R. H. J. Org. Chem. 1998, 63, 9904. (b) Lynn, D. M.; Mohr, B.; Grubbs, R. H.; Henling, L. M.; Day, M. W. J. Am. Chem. Soc. 2000, 122, 6601.

179

(a) Yao, Q. Angew. Chem. Int. Ed. 2000, 39, 3896. (b) Sch€ urer, S. C.; Gessler, S.; Buschmann, N.; Blechert, S. Angew. Chem. Int. Ed. 2000, 39, 3898.

180

(a) Mayo, K. G.; Nearhoof, E. H.; Kiddle, J. J. Org. Lett. 2002, 4, 1567. (b) Garbacia, S.; Desai, B.; Lavastre, O.; Kappe, C. O. J. Org. Chem. 2003, 68, 9136.

181

(a) Dowden, J.; Savovic, J. Chem. Commun. 2001, 37. (b) Jafarpour, L.; Nolan, S. P. Org. Lett. 2000, 2, 4075. (c) Ahmed, M.; Barrett, A. G. M.; Braddock, D. C.; Cramp, S. M.; Procopiou, P. A. Tetrahedron Lett. 1999, 40, 8657. (d) Ahmed, M.; Arnauld, T.; Barrett, A. M. G.; Braddock, D. C.; Procopiou, P. A. Synlett 2000, 1007.

182

Taber, D. F.; Frankowski, K. J. J. Org. Chem. 2003, 68, 6047.

183

Ohba, K.; Nakata, M. Org. Lett. 2015, 17, 2890.

184

Phillips, A. J.; Abell, A. D. Aldrichimica Acta 1999, 32, 75.

185

Nagata, T.; Nakagawa, M.; Nishida, A. J. Am. Chem. Soc. 2003, 125, 7484.

186

(a) Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999, 1, 953. Also see (b) Love, J. A.; Sanford, M. S.; Day, M. W.; Grubbs, R. H. J. Am. Chem. Soc. 2003, 125, 10103.

1005

18.6 METATHESIS REACTIONS

using 108 as the catalyst in a synthesis of sekothrixide by Nagumo and coworkers.187 Other metal catalysts are known, and in a variety of applications the metathesis reaction has been shown to proceed with good stereoselectivity.188 H

H

O

O

OAc

N

Bn

N

20% 108, CH2Cl2

Bn

OAc

N O

N

2 mM , 50°C

112

113

MPM MPM EEO

108, CH2Cl2

EEO

O

O

O

O

O

Reflux

O

O O

OTBS 114

OH

115 (>55%)

O Si O

O

O

109

O

PMBO

O

O

Si O

PhH, rt, 2 d

PMBO 116

117 (78%)

The Schrock catalyst (109) is also widely used. An example is taken from Vilarrasa and coworker’s189 synthesis of amphidinolide X, in which 116 was converted to 117 in 78% yield. As a synthetic method, cross-metathesis reactions190 were initially limited by the lack of predictability of product selectively and stereoselectivity. Grubbs and coworkers191 advanced a general model that is useful for predicting selectivity. As stated by Grubbs and coworkers191, a general ranking of alkene reactivity is achieved by categorizing alkenes by their relative ability to undergo homodimerization via cross-metathesis, and the susceptibility of their homodimers toward secondary metathesis reactions. Suppressing the rate of homodimerization of one component and controlling the rate of secondary metathesis on the desired cross-product gives product selectivity.191 Grubbs provided a table (see Table 18.1) to assist ranking alkenes as to their reactivity in cross-metathesis reactions.191 A synthetic example of a cross-metathesis reaction is taken from a synthesis of ()-tetrahydrolipstatin by Coates and coworkers192 in which (S)-pentadec-1-en-4-ol was coupled with oct-1-ene using the Grubbs’ II catalyst (108) to give (S,E)-henicos-7-en-10-ol in 59% yield. This particular reaction was used to prepare an authentic sample of (S,E)henicos-7-en-10-ol to verify the identity of a key synthetic intermediate.

187

Terayama, N.; Yasui, E.; Mizukami, M.; Miyashita, M.; Nagumo, S. Org. Lett. 2014, 16, 2794.

188

Lee, C. W.; Grubbs, R. H. Org. Lett. 2000, 2, 2145.

189

Rodríguez-Escrich, C.; Urpí, F.; Vilarrasa, J. Org. Lett. 2008, 10, 5191.

190

Connon, S. J.; Blechert, S. Angew. Chem. Int. Ed. 2003, 42, 1900.

191

Chatterjee, A. K.; Choi, T.-L.; Sanders, D. P.; Grubbs, R. H. J. Am. Chem. Soc. 2003, 125, 11360.

192

Mulzer, M.; Tiegs, B. J.; Wang, Y.; Coates, G. W.; O’Doherty, G. A. J. Am. Chem. Soc. 2014, 136, 10814.

1006

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

OH

OH

Oct-1-ene, 108

C11 H23

C11 H23

(S)-Pentadec-1-en-4-ol

C6H13

(S, E )-Henicos-7-en-10-ol (59%)

TABLE 18.1 Alkene Categories for Selective Cross Metathesis Metathesis Catalyst Alkene Type ( )

108

107

109

Type I (fast homodimerization)

Terminal alkenes, 1 degree allylic alcohols, esters, allyl boronate esters, allyl halides, styrenes (no large ortho substitutents), allyl silanes, allyl phosphontes, allyl sulfides, allyl phosphine oxides, protected allyl amines

Terminal alkenes, allyl silanes, 1 degree allylic alcohols, ethers, esters, allyl boronate esters, allyl halides

Terminal alkenes, allyl silanes

Type II (slow homodimerization)

Styrenes (large ortho substituents), acylates, acrylamides, acrylic acid, acrolein, vinyl ketones, unprotected, 3 degree allylic alcohols, vinyl epoxides, 2 degree allylic alcohols, perfluorinated alkanes, alkenes

Styrene, 2 degree allylic alcohols, vinyl dioxolanes, vinyl boronates

Styrene, allyl stannanes

Type III (no homodimerization)

1,1-Disubstituted alkenes, non-bulky trisubstituted alkenes, vinyl phosphonates, phenyl vinyl sulfone, 4 degree allylic carbons (all alkyl substitutents, 3 degree allylic alcohols (protected)

Vinyl siloxanes

3 degree Allyl amines, acrylonitrile

Type IV (spectators to cross-metathesis)

Vinyl nitro alkenes trisubstituted allylic alcohols (protected)

1,1-Disubstituted alkenes, disubstituted α,β-unsaturated carbonyls, 4 degree allylic carbon alkenes, perfluorinated alkanes, alkenes, 3 degree allyl amines (protected)

1,1-Disubstituted alkenes

Reprinted with permission from Chatterjee, A.K.; Choi, T.-L.; Sanders, D.P.; Grubbs, R.H. J. Am. Chem. Soc. 2003, 125, 11360. Copyright © 2003, American Chemical Society.

Metathesis with alkynes is also quite useful in synthesis,193 particularly for reactions of internal alkynes194 although terminal alkynes are not good partners in this reaction.195 Internal metathesis reactions with alkynes are known,196 including the conversion of 118 to 119(in 75% yield) in Barrett and coworker’s197 synthesis of cruentaren A. A Mo metathesis catalyst was used in this reaction.

193

Bunz, U. H. F.; Kloppenburg, L. Angew. Chem. Int. Ed. 1999, 38, 478.

194

(a) Pennella, F.; Banks, R. L.; Bailey, G. C. Chem. Commun. 1968, 1548. (b) Villemin, D.; Cadiot, P. Tetrahedron Lett. 1982, 23, 5139. (c) McCullough, L. G.; Schrock, R. R. J. Am. Chem. Soc. 1984, 106, 4067. (d) Tamao, K.; Kobayashi, K.; Ito, Y. Synlett 1992, 539.

195

McCullough, L. G.; Listemann, M. L.; Schrock, R. R.; Churchill, M. R.; Ziller, J. W. J. Am. Chem. Soc. 1983, 105, 6729.

196

(a) Trost, B. M.; Trost, M. K. J. Am. Chem. Soc. 1991, 113, 1850. (b) Gilbertson, S. R.; Hoge, G. S. Tetrahedron Lett. 1998, 39, 2075.

197

Fouche, M.; Rooney, L.; Barrett, A. G. M. J. Org. Chem. 2012, 77, 3060.

1007

18.6 METATHESIS REACTIONS

OPMB

OPMB N

TBSO 40%

OMe

OSiPh3

Mo

OMe

N

O PhMe, 110°C Sealed tube

O

MeO

Ph3SiO Ph3SiO

TBSO O O

MeO TIPSO

TIPSO

118

119 (75%) Me

Me Ts

N

Ph

H

Me

H

5% RuCl2(=CHPh)(PCy3)2 CH2Cl2 , 40°C, Sealed tube

Ts

Ph

+

N

Ts

Ph 4-Methyl-6-phenyl2-tosylisoindoline

120

N 4-Methyl-5-phenyl2-tosylisoindoline

Diynes also react with alkynes in an intermolecular reaction to form aromatic rings. An example is the conversion of 120 to a 6:1 mixture of 4-methyl-6-phenyl-2-tosylisoindoline/4-methyl-5-phenyl-2-tosylisoindoline in 82% yield.198a A similar, Pd catalyzed cycloaromatization is also known.199 Shmidt et al.200 reported cautionary note, in that radical cyclization processes (see Section 17.7) compete effectively with difficult-to-cyclize substrates. The first- and secondgeneration catalysts, 107 and 108, exhibit different activities when competing with radical cyclization. Grubbs and coworkers201 reported another useful variation of the metathesis reaction, for the synthesis of macrocyclic compounds. When 121 reacted with cyclooctene in the presence of 108, metathesis led to the ring expanded diester (122) in 45% yield. Identical ring expansion reactions were reported for diketones that generated macrocyclic diketones.201 O O , CH2Cl2

O

O

O

O

108

O

O 121

122 (45%)

The ring-closing metathesis disconnections follows: R (CH 2)n

(CH 2)n

(CH 2)n

(CH 2)n

X

X

H

X R

198

R

X +

R

H

(a) Witulski, B.; Stengel, T.; Fernández-Hernández, J. M. Chem. Commun. 2000, 1965. Also see (b) Yamamoto, Y.; Ogawa, R.; Itoh, K. Chem. Commun. 2000, 549.

199

Yamamoto, Y.; Nagata, A.; Itoh, K. Tetrahedron Lett. 1999, 40, 5035.

200

Schmidt, B.; Pohler, M.; Costisella, B. J. Org. Chem. 2004, 69, 1421.

201

Lee, C. W.; Choi, T.-L.; Grubbs, R. H. J. Am. Chem. Soc. 2002, 124, 3224.

1008

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

Another important application of metathesis reactions is the reverse of the ring-closing metathesis reactions, where C]C bonds are broken to form dienes in what is called ring-opening metathesis (ROM).202 Both 107 and 108 have been used, as well as other catalysts. This methodology has been used extensively in polymerization reactions called ringopening metathesis polymerization (ROMP),203 but only recently in total synthesis. An example of ring opening metathesis to produce dienes is taken from a synthesis of F2-isoprostanes, Schrader and Snapper204 reacted 123 with 108 to give a 67% yield of 124, as a 1.4:1 mixture of (Z/E) isomers. TBSO

TBSO

H

C5H11

TBSO

H

CH2Cl2

108

OH OH

H 123

TBSO

H

C5H11 124

The ring-opening metathesis disconnection follows:

(CH 2)n

Mes Cl

N

(CH 2)n

N

N Mes

Mes

Ru

N

N Mes Ru

Cl

Cl Cl

MeO

O

125

Cl

O i-Pr

O Ph

126

N

Mo N

t-Bu

i-PrO

Me

MeO

N

Ru

Cl

t-Bu t-Bu

N Mes

Et

Mes

127

128

The synthetic importance of ring-closing and -opening metathesis reactions has led to the development of several new catalysts.205 Air stability, reactivity, robust character, and improved reactivity are the goals of such new catalysts. A sample of newer catalysts includes 125 by Hoveyda and coworkers,206 126 by Grela and Kim,207 127 by Wakamatsu and Blechert,208 and 128 by Moore and coworkers.209 The latter catalyst (128) was particularly useful for the metathesis homocoupling of alkynes. Hoveyda and coworkers210 developed a polymer-supported, recyclable Mo catalyst.

202

(a) Bespalova, N. B.; Bovina, M. A.; Sergeeva, M. B.; Oppengeim, V. D.; Zaikin, V. G. J. Mol. Catal. 1994, 90, 21. (b) Zuercher, W. J.; Scholl, M.; Grubbs, R. H. J. Org. Chem. 1998, 63, 4291. (c) Randl, S.; Connon, S. J.; Blechert, S. Chem. Commun. 2001, 1796. (d) Morgan, J. P.; Morrill, C.; Grubbs, R. H. Org. Lett. 2002, 4, 67. (a) Novak, B. M.; Grubbs, R. H. Encycl. Polym. Sci. Eng. 1990, 420. (b) Amass, A. J. New Methods Polym. Synth. 1991, 76. (c) Grubbs, R. H.; Khosravi, E. Mater. Sci. Technol. 1999, 20 (Synthesis of Polymers) 65.

203

204

Schrader, T. O.; Snapper, M. L. J. Am. Chem. Soc. 2002, 124, 10998.

205

Schrock, R. R.; Hoveyda, A. H. Angew. Chem. Int. Ed. 2003, 42, 4592.

206

(a) Kingsbury, J. S.; Harrity, J. P. A.; Bonitatebus, P. J.; Hoveyda, A. H. J. Am. Chem. Soc. 1999, 121, 791. (b) Garber, S. B.; Kingsbury, J. S.; Gray, B. L.; Hoveyda, A. H. Ibid, 2000, 122, 8168. Also see Sattely, E. S.; Meek, S. J.; Malcolmson, S. J.; Schrock, R. R.; Hoveyda, A. H. J. Am. Chem. Soc. 2009, 131, 943.

207

(a) Grela, K.; Kim, M. Eur. J. Org. Chem. 2003, 963. (b) Grela, K.; Harutyunyan, S.; Michrowska, A. Angew. Chem. Int. Ed. 2002, 41, 4038.

208

(a) Wakamatsu, H.; Blechert, S. Angew. Chem. Int. Ed. 2002, 41, 794. (b) Conon, S. J.; Dunne, A. M.; Blechert, S. Angew. Chem. Int. Ed. 2002, 41, 3835.

209

Zhang, W.; Kraft, S.; Moore, J. S. J. Am. Chem. Soc. 2004, 126, 329. Also see Heppekausen, J.; Stade, R.; Goddard, R.; F€ urstner, A. J. Am. Chem. Soc. 2010, 132, 11045.

210

Hutlzsch, K. C.; Jernelius, J. A.; Hoveyda, A. H.; Schrock, R. R. Angew. Chem. Int. Ed. 2002, 41, 589.

1009

18.7 PAUSON-KHAND REACTION

18.7 PAUSON-KHAND REACTION As seen in previous sections, Cu, Ni, Pd, and Ru are important catalysts for organic reactions. Another useful metal for organic transformations is Co, which is used in an interesting cyclization reaction that involved the reaction of dienes, diynes, or ene-ynes to form cyclopentenone derivatives in the presence of CO.211 An example is taken from a synthesis of (+)-fusarisetin A by Li, Yang, and coworkers212 in which ene-yne 129 was heated with dicobalt octacarbonyl to give an 82% yield of 130. This transformation has become an important synthetic tool known as the Pauson-Khand reaction.213 Me Me H

H

Co2(CO)8, PhMe

OTBS

Me

120°C

OTBS

OH

OH

H

Me

Me 129

130 (82%)

The mechanism probably involves insertion of the alkene (or alkyne) into the transition metal bond, which is why it is presented in this section. Formally, it is a [2 + 2 + 1]-cycloaddition, but the accepted mechanism is the one proposed by Magnus and coworkers,214,215 shown in Fig. 18.1,which shows the conversion of (R)-8-(trimethylsilyll)oct-1-en-7-yn-4-ol to (5S,6aR)-5-hydroxy-3-(trimethylsilyl)-4,5,6,6a-tetrahydropentalen-2(1H)-one.214a It has been stated that further study is required to firmly establish the mechanism, although Krafft’s216 work supports the one reported by Magnus.214 Gimbert et al217 showed that CO is lost from the Pauson-Khand complex prior to alkene coordination and insertion. Milet, and coworker’s218 performed calculations that concluded the LUMO of the coordinated alkene plays a crucial role in alkene reactivity, by largely determining the degree of back-donation in the complex. Formation of the CdC bond is strongly

Co

Co Co

SiMe

Heat

CO2(CO)8

Co

SiMe3

HO HO

HO (R)-8-(Trimethylsilyl)oct1-en-7-yn-4-ol

Co

Co

Me3Si

SiMe3

Co

Co O

O HO

Me3Si

O

HO

Me3Si

HO (5S,6a R)-5-Hydroxy-3-(trimethylsilyl)4,5,6,6a-tetrahydropentalen-2(1H)-one

FIG. 18.1 Magnus mechanism for the Pauson-Khand reaction with (R)-8-(trimethylsilyll)oct-1-en-7-yn-4-ol. Adapted from Scheme II and reprinted with permission from Magnus, P.; Becker, D.P. J. Am. Chem. Soc. 1987, 109, 7495. Copyright © 1987, American Chemical Society.

211

For transformations of this type, see Reference 38, pp 1369, 1371–1372.

212

Huang, J.; Fang, L.; Long, R.; Shi, L.-L.; Shen, H.-J.; Li, C.-C.; Yang, Z. Org. Lett. 2013, 15, 4018.

213

(a) Khand, I. U.; Knox, G. R.; Pauson, P. L.; Watts, W. E.; Foreman, M. I. J. Chem. Soc. Perkin Trans. 1973, 1, 977. (b) Khand, I. U.; Pauson, P. L.; Habib, M. J. J. Chem. Res. (S) 1978, 348. (c) Khand, I. U; Pauson, P. L. J. Chem. Soc. Perkin Trans. 1976, 1, 30. (d) The Merck Index, 14th ed.; Merck: Whitehouse Station, NJ, 2006; p ONR-69. (e) Mundy, B. P.; Ellerd, M. G.; Favaloro, F. G., Jr. Name Reactions and Reagents in Organic Synthesis, 2nd ed.; WileyInterscience: Hoboken, NJ, 2005; pp 486–487.

214

(a) Magnus, P.; Becker, D. P. J. Am. Chem. Soc. 1987, 109, 7495; Also see (b) Magnus, P.; Principe, L. M. Tetrahedron Lett. 1985, 26, 4851.

215

For a review, see Brummond, K. M.; Kent, J. L. Tetrahedron 2000, 56, 3263.

216

Krafft, M. E. Tetrahedron Lett. 1988, 29, 999.

217

Gimbert, Y.; Lesage, D.; Milet, A.; Fournier, F.; Greene, A. E.; Tabet, J.-C. Org. Lett. 2003, 5, 4073.

218

de Bruin, T. J. M.; Milet, A.; Greene, A. E.; Gimbert, Y. J. Org. Chem. 2004, 69, 1075.

1010

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

influenced by the effectiveness of the LUMO overlap with the HOMO of the CO2(CO)x complex, and such overlap is impacted by the degree of back-donation.218 Other metal complexes have been used,219 and the addition of primary amines has been shown to enhance the rate of the reaction.220 Aldehydes have been used as a source of CO in some cases.221 Gibson et al.222 reported the use of heptacarbonyl(triphenylphosphine)cobalt(0) as a robust, and stable Pauson-Khand catalyst. This reaction has found its way into organic synthesis,215 and is attractive because of the generality and selectivity. Very sophisticated ring systems can be produced by this technique. An example is Cook and coworker’s223 synthesis of [5.5.5.5]-tetracyclic systems (pentalenes) in which 131 was treated with Mo(CO)6 to give an 89% yield of 132. O

O

2.4 Mo(CO)6 , 10 DMSO Toluene, 100°C, 4.5 h

O O O

O 131

132 (89%)

Although this reaction usually requires stoichiometric amounts of the metal carbonyl, Kraft and Bañaga224 showed that sub-stoichiometric amounts can be used (35–50% of CO2(CO)8), nitrogen atmosphere, in dimethoxyethane with 3 equiv of cyclohexylamine. Kraft and Boñaga225 developed dodecacarbonyltetracobalt as a viable catalyst, and catalytic Pauson-Khand reactions are now well established.226 Molecular sieves promote the Pauson-Khand reaction,227 and high intensity ultrasound is also effective.228 Polymer-supported promoters of this reaction are known.229 In the presence of Rh catalysts under solvent-free conditions, aldehydes serve as a CO source.230 Manipulating the reaction conditions, the catalyst, solvent and the length of time allowed for the reaction can give bicyclooctanones, as well as the usual bicyclooctenone products.231 Enantioselective versions of the Pauson-Khand reaction have been reported.232 Brucine-N-oxide promotes an asymmetric Pauson-Khand reaction, for example.233 An enantiospecific variation234 has been reported by using a menthol chiral auxiliary and by replacing one of the Co centers with an isoelectronic Mo fragment. Chiral phosphine derivatives have also proven to be effective,235 as have catalytic amounts of chiral aryl biphosphites.236 An aza-Pauson-Khand reaction was used in a synthesis of physostigmine by Mukai et al.237 the reaction of alkyne-diimide 133 and the Co catalyst gave a 55% yield of 134 with a pyrrolo[2,3-b]indol-2-one unit.

219

(a) Koga, Y.; Kobayashi, T.; Narasaka, K. Chem. Lett. 1998, 249. (b) Hoye, T. R.; Suriano, J. A. J. Am. Chem. Soc. 1993, 115, 1154.

220

Sugihara, T.; Yamada, M.; Ban, H.; Yamaguchi, M.; Kaneko, C. Angew. Chem. Int. Ed. 1997, 36, 2801.

221

Shibata, T.; Toshida, N.; Takagi, K. Org. Lett. 2002, 4, 1619.

222

Gibson, S. E.; Johnston, C.; Stevenazzi, A. Tetrahedron 2002, 58, 4937.

223

Cao, H.; Van Ornum, S. G.; Cook, J. M. Tetrahedron Lett. 2000, 41, 5313.

224

Krafft, M. E.; Bañaga, L. V. R. Synlett, 2000, 959.

225

Krafft, M. E.; Boñaga, L. V. R. Angew. Chem. Int. Ed. 2000, 39, 3676.

226

(a) Gibson, S. E.; Johnstone, C.; Stevenazzi, A. Tetrahedron 2002, 58, 4937. (b) Gibson, S. E.; Stevenazzi, A., Angew. Chem. Int. Ed. 2003, 42, 1800.

227

Perez-Serrano, L.; Blanco-Urgoiti, J.; Casarrubios, L.; Domínguez, G.; Perez-Castells, J. J. Org. Chem. 2000, 65, 3513.

228

Ford, J. G.; Kerr, W. J.; Kirk, G. G.; Lindsay, D. M.; Middlemiss, D. Synlett 2000, 1415.

229

(a) Kerr, W. J.; Lindsay, D. M.; McLaughlin, M.; Pauson, P. L. Chem. Commun. 2000, 1467. (b) Brown, D. S.; Campbell, E.; Kerr, W. J.; Lindsay, D. M.; Morrison, A. J.; Pike, K. G.; Watson, S. P. Synlett, 2000, 1573. 230

Shibata, T.; Tshida, N.; Takagi, K. J. Org. Chem. 2002, 67, 7446.

231

Krafft, M. E.; Boñaga, L. V. R.; Wright, J. A.; Hirosawa, C. J. Org. Chem. 2002, 67, 1233.

232

Ingate, S. T.; Marco-Contelles, J. Org. Prep. Proceed. Int. 1998, 30, 121.

233

Kerr, W. J.; Lindsay, D. M.; Rankin, E. M.; Scott, J. S.; Watson, S. P. Tetrahedron Lett. 2000, 41, 3229.

234

Fletcher, A. J.; Rutherford, D. T.; Christie, S. D. R. Synlett 2000, 1040.

235

Verdaguer, X.; Moyano, A.; Pericàs, M. A.; Riera, A.; Maestro, M. A.; Mahía, J. J. Am. Chem. Soc. 2000, 122, 10242.

236

Sturla, S. J.; Buchwald, S. L. J. Org. Chem. 2002, 67, 3398.

237

Mukai, C.; Yoshida, T.; Sorimachi, M.; Odani, A. Org. Lett. 2006, 8, 83.

1011

18.8 ORGANOMETALLIC COMPOUNDS AS CARBANIONIC REAGENTS

TMS

MeO 0.2 Co2(CO)8, PhH, CO

TMS

O

MeO

1.2 TMTU, 70°C

N

TMTU = Tetramethylthiourea

NKCKNMe

Me

N

133

134 (55%)

The Pauson-Khand disconnection follows: R (CH 2)n

(CH 2)n

O

R

18.8 ORGANOMETALLIC COMPOUNDS AS CARBANIONIC REAGENTS Other metal derivatives are useful in synthesis, including several organometallic reagents that react more-or-less as carbanions. Among the more popular reagents are allyl- or alkyl-metallic compounds of Sn and Ti. Organoiron compounds can also serve as nucleophilic reagents.

18.8.1 Allyltin Reagents Tetravalent Sn complexes add to aldehydes and ketones in the presence of a Lewis acid. Allyltin complexes are, by far, the most widely used of these compounds.238 A typical example is taken from the work of Keck and Boden,239 in which a chiral aldehyde (135) was treated with allyltributyltin and various Lewis acids, and a mixture of syn (136) and anti (137) products was obtained.239 The ratio of 136/137 was dependent on the structure of the R group in 135, the solvent, and the Lewis acid.239 The anti-product (137) was obtained by using the tert-butyldimethylsilyloxy derivative (Section 5.3.1.1) of 135 with 2 equiv of boron trifluoride in dichloromethane. The stereoselectivity depends on the nature of the Lewis acid to some extent. The syn-product (136) was obtained preferentially when the benzyloxy derivative of 135 was used with titanium tetrachloride in dichloromethane.239 OR

Sn(Bu)3

CHO

OR

OR

+ Lewis acid

135

OH 136

OH 137

Allyltin and crotyltin coupling is highly diastereoselective. In a synthesis of (+)-roxaticin, Evans and Connell240 reacted (S)-3-(benzyloxy)-2-methylpropanal with allyltributytin in the presence of SnCl4, and obtained a 90% yield of (2S,3R)-1-(benzyloxy)-2-methylhex-5-en-3-ol, with a 35:1 diastereoselectivity. As the size of the group attached to the α-carbon of the aldehyde increased, the diastereoface selectivity increased for the syn diastereomer relative to allyltin compounds. In the example cited, the chelation product exceeded the product predicted by the Felkin-Anh model (Section 7.9.1), being formed in a ratio of >200:1. In general, the Felkin-Anh model (predicting the syn-diastereomer) gave the greatest success when BF3 was used as a catalyst in the presence of diphenyl-tert-butylsilyloxy aldehydes.241 Chelation control with silyloxy derivatives did not contribute significantly to any condensation product,241a although MgBr2 gave 91:9 diastereofacial selectivity and 89:11 syn-selectivity.241a Enantioselective addition of allyltributytin to aldehydes is

238

For reactions of this type, see Larock, R. C. Conmprehensive Organic Transformations, 2nd ed.; Wiley-VCH: New York, 1999; pp 373–378.

239

Keck, G. E.; Boden, E. P. Tetrahedron Lett. 1984, 25, 265. Allyltributyl tin can undergo conjugate addition, as in a synthesis of ent-Sch 47554 by Morton, G. E.; Barrett, A. G. M. Org. Lett. 2006, 8, 2859.

240

Evans, D. A.; Connell, C. T. J. Am. Chem. Soc. 2003, 125, 10899.

241

(a) Keck, G. E.; Abbott, D. E. Tetrahedron Lett. 1984, 1883. (b) Keck, G. E.; Abbott, D. E.; Boden, E. P.; Enholm, E. J. Ibid. 1984, 25, 3927.

1012

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

also possible, using Ti(Oi-Pr)4 and a chiral binaphthol derivative (e.g., (S)-BINOL [(S)-()-1,10 -bi-2-naphthol]). In a synthesis of mucocin, Takahashi et al.242 reacted 4-((4-methoxybenzyl)oxy)butanal with allyltributytin in the presence of (S)-BINOL, and obtained a 76% yield of (R)-7-((4-methoxybenzyl)oxy)hept-1-en-4-ol (>98 %ee). O

OBn

OH

SnCl4, –78°C, CH2Cl2

H

OBn

SnBu3

Me

Me (2S,3R)-1-(Benzyloxy)-2-methylhex-5-en-3-ol

(S)-3-(Benzyloxy)-2methylpropanal

(90%) SnBu3

OH

CHO

MPMO

MPMO

0.2 Ti(Oi-Pr)4, 0.2 (S)-BINOL –25°C (5 d) –78°C (1 h)

4-((4-Methoxybenzyl)oxy)butanal

(R)-7-((4-Methoxybenzyl)oxy)hept-1-en-4-ol (76%, >98 %ee)

Homoallylic alcohols can be prepared. When benzaldehyde reacted with 2-methylbut-3-en-1-ol,243a a 78% yield of (E)-1-phenylpent-3-en-1-ol was obtained as a 49/1 (E/Z) mixture. Tin-catalyzed addition of the alcohol substrate to the aldehyde, followed by a [3,3]-sigmatropic rearrangement, led to the products. Clearly, this is not the same reaction, but it is useful to compare it with the other reactions. OH

PhCHO, 10% Sn(OTf)2

HO

CH2Cl2, 0°C, 2 h

Me 2-Methylbut-3-en-1-ol

Me

Ph

(E)-1-Phenylpent-3-en-1-ol (78%)

The allyltin disconnections follow:

R

X

OH

R

R + R X1

Ph—X Ph

+ X1

R1

R1

CHO

+

R23Sn R

R

Other allyl metal complexes can be used to generate carbon-carbon bonds, including allylzinc complexes. Stille showed that 3-methyl-1-bromobut-2-ene reacted with zinc chloride (ZnCl2) to generate the π-allylzinc species.243b Coupling required the use of an organotin species (e.g., α-trimethyltin isoprene [trimethyl(2-methylenebut-3-en-1-yl)stannane]). Reaction in refluxing THF led to a 94% yield of myrcene (7-methyl-3-methyleneocta-1,6-diene).243b Me

10% ZnCl2, THF

SnMe3 Trimethyl(2-methylenebut-3en-1-yl)stannane

Br

Me Me

Me Myrcene

18.8.2 Alkyltitanium Reagents Another organometallic coupling reaction is the condensation of aldehydes with alkyl trichlorotitanium compounds (RTiCl3). The reaction of methyltrichlorotitanium (MeTiCl3) with (S)-3-(benzyloxy)heptanal in dichloromethane at 78 °C yielded a 91:9 mixture of (2S,4S)-4-(benzyloxy)octan-2-ol/(2R,4S)-4-(benzyloxy)octan-2-ol

242

Takahashi, S.; Kubota, A.; Nakata, T. Angew. Chem. Int. Ed. 2002, 41, 4751.

243

(a) Sumida, S.-i.; Ohga, M.; Mitani, J.; Nokami, J. J. Am. Chem. Soc. 2000, 122, 1310. (b) Godschalx, J. P.; Stille, J. K. Tetrahedron Lett. 1983, 24, 1905.

1013

18.8 ORGANOMETALLIC COMPOUNDS AS CARBANIONIC REAGENTS

via the chelated complex 138.244 Allyltitanium compounds also couple to aldehydes. In a synthesis of FR66979, Ciufolini and Ducray245 reacted aldehyde 139 with allyltitanium anion 140 and obtained alcohol 141 in good yield. Ph

H C4H9

CHO O

C4H9

O H

Ph O

(S)-3-(Benzyloxy)heptanal OBn

Me Ti Cl Cl Cl

H C4H9

138

OH

H

+

H OBn

(2S,4S)-4-(Benzyloxy)octan-2-ol

OBn

(2R,4S)-4-(Benzyloxy)octan-2-ol OBn

OBn −Ti(Oi-Pr)

4

Li−

OH H

C4H9

OBn CHO

Me

Me

H

OH

THF, 78°C

SiMe3

+ Me3Si

N3

N3

OBn

OBn 139

140

141

18.8.3 Organoaluminum Compounds Organoaluminum compounds are mostly electrophilic in nature, and typically react as Lewis acids. Zweifel and Whitney,246 however, showed that conversion of an organoaluminum to its “ate” complex changed it into a nucleophilic species. Examples of this reaction have been largely confined to the vinylalanates, which react with electrophiles to give the (E)- or (Z)-isomer, depending on the reaction conditions. Treatment of an alkyne (e.g., hex-3-yne) with diisobutylaluminum hydride (dibal; see Section 7.7.1.2) gave vinyl alanate (142). When 142 reacted with methyllithium, aluminate 143 formed in 73% yield. Subsequent reaction with an electrophile (e.g., CO2) led to a transfer of the vinyl group to the carbonyl, forming the (E)-acrylic acid derivative (E)-2-ethylpent-2-enoic acid. The geometry of the intermediate alanate and the final product was changed by first reacting methyllithium with dibal to give the aluminate 144. When this “ate” complex reacted with the alkyne (hex-3-yne), a new “ate” complex (145) was formed. Subsequent reaction with CO2 gave (Z)-2-ethylpent-2-enoic acid. Et

Et

dibal

Et

Et MeLi, Ether

Al

Et Heptane, 50°C

MeLi

H Al Diisobutylaluminum hydride

H Me

(E)-2-Ethylpent2-enoic acid Et

−

−

100°C

Al Et

144

Al

Et Et

CO2H

2. H3O+

143

Li+

Et

1. CO2

Al

142

Et

Li

Me

–20°C

Hex-3-yne

Et

Me 145

Li+ Et

1. CO2 2. H3O+

Et

CO2H

(Z)-2-Ethylpent2-enoic acid

Aluminum “ate” complexes also react with alkynes, and hydrolysis converts the intermediate into substituted buta1,3-dienes.247 Reaction of the “ate complex” with halogens gave the corresponding vinyl halide, with retention of the geometry observed in the complex.246 Similar reaction with alkyl halides gave the substitution product.248 Vinyl alanates generated in this way can be quenched with other electrophiles. In a synthesis of (+)-testudinariol A,

244

Reetz, M. T.; Jung, A. J. Am. Chem. Soc. 1983, 105, 4833.

245

Ducray, R.; Ciufolini, M. A. Angew. Chem. Int. Ed. 2002, 41, 4688.

246

(a) Zweifel, G.; Whitney, C. C. J. Am. Chem. Soc. 1967, 89, 2753. (b) Zweifel, G.; Steele, R. R. Ibid. 1967, 89, 2754, 5085.

247

(a) Zweifel, G.; Polston, N. L.; Whitney, C. C. J. Am. Chem. Soc. 1968, 90, 6243. (b) Zweifel, G.; Miller, R. L. Ibid. 1970, 92, 6678.

248

Yamamoto, Y.; Yatagai, H.; Maruyama, K. J. Org. Chem. 1980, 45, 195.

1014

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

Amarasinghe and Montgomery249 converted a silyl-alkyne to a vinyl bromide by reacting the alkyne unit with dibal followed by bromination with Br2. Vinyl alanes are also generated by the transition metal catalyzed addition of trialkylaluminum reagents to alkynes. In the Hoye et al.250 synthesis of elenic acid, dodec-1-yne reacted with Me3Al and Cp2ZrCl2, and then MeLi. The resulting “ate” complex, (146), was subsequently treated with the triflate of methyl (S)-2-hydroxypropanoate to give a 96% yield of the alkene, methyl (S,E)-2,4-dimethyltetradec-3-enoate. − 1. Me3Al Cp2ZrCl2

Me

AlMe3

C10 H21 2. MeLi

C10 H21

Dodec-1-yne

Me

Li+

1. TfO

Me Me

CO2Me

2. aq NaHCO3

H

CO2Me

C10 H21

H

Methyl (S, E )-2,4-dimethyltetradec-3-enoate

146

(96%)

The disconnection for the alkyne-alanate reaction follows: X R R

+

R

X

R

18.8.4 Organochromium Reagents The Takai reaction was introduced in Section 12.6 in which CrCl2 reacted with CH3I to generate an intermediate organochromium reagent that reacted with aldehydes to yield alkene derivatives. Other organochromium reagents have been reported. Benzene and its derivatives react with chromium hexacarbonyl [Cr(CO)6]251 to give an arylchromium complex (e.g., the benzene derivative, 147). Chromium resides in a position perpendicular to the plane of the ring, as shown. Semmelhack et al.252 found that the Cr activated the benzene ring to metal-hydrogen exchange with organolithium reagents (Section 11.6.9), and these new organolithium reagents reacted with alkyl halides to yield the alkyl substitution product. The reaction of 147 with butyllithium generated 148 via lithium-hydrogen exchange. Subsequent reaction with an electrophile gave 149 in good yield. The reaction with CO2 in methanol gave the methyl ester in 72% yield; reaction with acetone gave the corresponding alcohol in 29% yield although reaction with benzaldehyde gave no alcohol at all.252

Li

C4H9Li

L Cr(CO)3 147

Cr(CO)3 148

E+

E Cr(CO)3 149

Arylchromium complex 147 reacts with carbon nucleophiles via nucleophilic aromatic substitution. As shown in Table 18.2,253 organolithium reagents and enolates add to 147 to generate a carbanionic complex (150). The nucleophilic addition can be reversible if complex 150 is heated. The complex can be decomposed by reaction with iodine to give 151 or by protonolysis (with trifluoroacetic acid, TFA) to give cyclohexadiene derivative 152. The reagents used to remove the metal are similar to those used with organoiron complexes (Section 18.8.5).

249

Amarasinghe, K. K. D.; Montgomery, J. J. Am. Chem. Soc. 2002, 124, 9366.

250

Hoye, R. C.; Baigorria, A. S.; Danielson, M. E.; Pragman, A. A.; Rajapakse, H. A. J. Org. Chem. 1999, 64, 2450.

251

(a) Strohmeier, W. Chem. Ber. 1961, 94, 2490. (b) Rausch, M. D. J. Org. Chem. 1974, 39, 1787.

252

Semmelhack, M. F.; Bisaha, J.; Czarny, M. J. Am. Chem. Soc. 1979, 101, 768.

253

(a) Semmelhack, M. F.; Hall, H. T.; Yoshifuji, M.; Clark, G. J. Am. Chem. Soc. 1975, 97, 1247. (b) Semmelhack, M. F.; Hall, H. T., Jr.; Yoshifuji, M. Ibid. 1976, 98, 6387.

1015

18.8 ORGANOMETALLIC COMPOUNDS AS CARBANIONIC REAGENTS

TABLE 18.2 Nucleophilic Aromatic Substitution of 147 by Organolithium Reagents and Enolate Anions R

R−

H

I2, –25°C

Heat

Cr(CO)3

R

151

CF3CO2H Cr(CO)3

R

–78°C

150

147

152

R-

% 151

LiCH2CN

68

LiCMe2CN

94

2-Lithiodithiane

93

LiCMe3

97

LiCH2CO2CMe3

87

Reprinted with permission from Semmelhack, M.F.; Hall, H.T.; Yoshifuji, M.; Clark, G. J. Am. Chem. Soc. 1975, 97, 1247 and Semmelhack, M.F.; Hall Jr., H.T.; Yoshifuji, M. J. Am. Chem. Soc. 1976, 98, 6387. Copyrights © 1976 and 1975, American Chemical Society.

When the benzene ring is functionalized, the Cr complex can react with nucleophiles to give ortho, meta, or para isomers. In general, the meta-substitution product predominates,254,255 but all three isomeric products are formed. These transformations have been used in synthesis. Semmelhack et al.,254,256 used an arylchromium complex in a synthesis of acorenone and also in a synthesis of ()-frenolicin that exploited the lithium-hydrogen exchange reaction.257 Chromium complexes of polynuclear aromatic molecules undergo nucleophilic substitution, but not on the ring complexed to Cr.258 Asymmetric induction is possible when chiral ligands are used in conjunction with the organolithium addition reactions. The reaction of 153 with phenyllithium in toluene, in the presence of ()-sparteine gave complex 154, which was treated with propargyl bromide to give a 72% yield of 155 in 54 %ee.259 −

N Ph

O

H

Br

N

PhLi, PhMe

O

(−)-Sparteine – 78°C

HMPA –78 → 20°C

N

O Ph

Cr(CO)3 Cr(CO)3 153

154

155 (72%, 54 %ee)

A different asymmetric application allows formation of CH2Li units from ortho-methyl groups in Cr-arene complexes, which then react as carbon nucleophiles. Amide 156 reacted with butyllithium260 in the presence of a chiral amine, and subsequent reaction with benzyl bromide gave 157 in 63% yield and 67 %ee. The amide unit clearly activates the substituent at the ortho-position to substitution, and the presence of the chiral amine led to asymmetric induction.

254

Semmelhack, M. F.; Harrison, J. J.; Thebtaranonth, Y. J. Org. Chem. 1979, 44, 3275.

255

(a) Semmelhack, M. F.; Clark, G. R.; Farina, R.; Saeman, M. J. Am. Chem. Soc. 1979, 101, 217. (b) Semmelhack, M. F.; Hall, H. T., Jr.; Farina, R.; Yoshifuji, M.; Clark, G.; Bargar, T.; Hirotsu, K.; Clardy, J. Ibid. 1979, 101, 3535.

256

Semmelhack, M. F.; Yamashita, A. J. Am. Chem. Soc. 1980, 102, 5924

257

Semmelhack, M. F.; Zask, A. J. Am. Chem. Soc. 1983, 105, 2034.

258

Semmelhack, M. F.; Seufert, W.; Keller, L. J. Am. Chem. Soc. 1980, 102, 6584.

259

Amurrio, D.; Khan, K.; K€ undig, E. P. J. Org. Chem. 1996, 61, 2258.

260

Koide, H.; Hata, T.; Uemura, M. J. Org. Chem. 2002, 67, 1929.

1016

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

Me

Me

Ph N

O Me

Me

Me

1. BuLi, Chiral amine THF, –78 → −30°C

Ph N

O CH2CH2Ph

2. PhCH2Br, –78°C

Cr(CO)3

Cr(CO)3

156

157 (63%, 67 %ee)

The disconnections available from these organochromium reactions follow: R

X

+ RJX X X

R N O

+

H2CJX R

Ar

Ar

N O R1

In general, nucleophilic aromatic substitution reactions are rather difficult with unsubstituted aryl derivatives, or when the aromatic ring contains a strongly electron-releasing group. Formation of the Cr complex activates such aromatic compounds to nucleophilic substitution. Since the nucleophiles are carbon nucleophiles, this technique offers a route to carbon bonds that would be very difficult to form by other methods.

18.8.5 Organoiron Compounds 18.8.5.1 Formation and Stability A few organoiron compounds have found their way into organic synthesis.261 In general, FedC bonds are sensitive to homolytic cleavage, producing organic radicals and the metal in a lower oxidation state. This instability is due to small energy differences between the filled d-orbitals, and the valence s and p anti-bonding orbitals of the FedC bond. This bond can be stabilized in at least two ways. The first involves addition of a ligand possessing acceptor properties (CO, cyclopentadienyl, phosphines, amines, etc.). The second involves alteration of the effective electronegativity of the carbon (induce a different hybridization state), or attachment of strongly electronegative groups (e.g., fluorine to the carbon). The iron must eventually be removed from the molecule to be useful in organic synthesis. Reaction with a proton source is one way to break the FedC bond.262 Reaction with either water or an alcohol is usually sufficient, although a catalytic amount of acid is sometimes required.262 This process simply protonates the alkyl fragment in an organoiron compound (e.g., 158) to give methane and 159.262 Some organoiron complexes are thermally labile, generating an alkene via β-elimination.263 Thermal disproportionation is also observed and the thermal stability of simple alkyl groups in RFe(CO)2Cp was reported to be R ¼ Me  Ph ≫ Et > i-Pr. An example of the disproportionation reaction is the fragmentation of 160 to the alkene. Reaction of the organoiron complex with a halogen262 also cleaves the CdFe bond. When 161 was treated with iodine, an alkyl halide (RdI) and the iron iodide (62) were formed. The two most common methods for converting an organoiron to the alkyl fragment are protonolysis and halogenation. Both of these techniques will be used in the following sections:

261

Davies, S. G. Organotransition Metal Derivatives: Applications to Organic Synthesis; Pergamon Press: Oxford, 1982.

262

Piper, T. S.; Wilkinson, G. J. Inorg. Nucl. Chem. 1956, 3, 104.

263

Braterman, P. S.; Cross, R. J. J. Chem. Soc. Dalton Trans. 1972, 657.

1017

18.8 ORGANOMETALLIC COMPOUNDS AS CARBANIONIC REAGENTS H+

MeFe(CO)2Cp

+

+

CH4

[CpFe(CO) 2(H2O)]

158

159 Heat

LnMJCH2CH2R 160 I2

RFe(CO)2Cp

LnMJH

+

R—I

+

CH2KCHR

IJFe(CO)2Cp 162

161

18.8.5.2 Cyclopentadienylirondicarbonyl (Fp) Compounds An anionic Fe reagent is prepared by reaction of iron pentacarbonyl with the dimer of cyclopenta-1,3-diene (163, Section 14.8.2) at 200 °C to give the dimeric species, (164). The anionic derivative 165 can be converted to an alkyl derivative (166, where R ¼ alkyl) by reaction with an alkyl halide, or to the protio derivative (166, R ¼ H) by reaction with an acid.264 Treatment of 164 with base leads to loss of a cyclopenta-1,3-diene unit and formation of the anionic species 165. A byproduct of this process is ferrocene, formed by extrusion of CO2 from 164.264 −

Fe(CO)5

+

Fe(CO)2 163

2

164

CpFe(CO)2 Na+ 165

RX

CpFe(CO)2R 166

Fe Ferrocene

Note that cyclopentadienylbis(carbonyliron), better known as ferrocene (abb as Fp), is a common organoiron reagent.265 The protonated form of ferrocene can add to conjugated carbonyl derivatives (e.g., acrylonitrile)266 and also with conjugated dienes.267 Both SO3268 and CO can be inserted into the FedC bond of FpdR.268 Protonolysis provides a synthetic route to substituted nitriles or alkenes. In each case, the Fp unit can also be removed photochemically, thermally, or by treatment with triphenylphosphine. Hydride extraction to give an alkene or a diene is also known.269 18.8.5.3 Sodium Tetracarbonyl Ferrate An extremely useful organoiron reagent is sodium tetracarbonyl ferrate [Na2Fe(CO)4],270 usually prepared by the reduction of iron pentacarbonyl. The synthetic utility of this reagent lies in its ability to react with alkyl halides in a stepwise manner, including the reaction with two different alkyl halides as reported independently by Cooke and by Collman and coworkers.270,271 FeðCOÞ5

1% NaðHgÞ ƒƒƒƒƒƒƒƒ!

Iron pentacarbonyl

Na2 FeðCOÞ4 Sodium tetracarbonyl ferrate

When sodium tetracarbonyl ferrate reacted with 1-bromooctane, a trigonal-bipyramidal complex (167) was formed by insertion of the alkyl fragment. Subsequent treatment with iodoethane led to a second insertion and expansion of the coordination shell to an octahedral species (168). When the reaction was pressurized with carbon monoxide, CO was inserted in the CdFe bond to give an acyl derivative (169). Protonolysis with acetic acid cleaved the FedC bond, and 264

Cotton, F. A.; Wilkinson, G.; Murillo, C. A.; Bochmann, M. Advanced Inorganic Chemistry, 6th ed.; Wiley: New York, 1999; pp 808–812.

265

Eisch, J. J.; King, R. B. Organomet. Synth. 1965, 1, 114, 152.

266

Green, M. L. H.; Nagy, P. L. I. J. Chem. Soc. 1963, 189.

267

Ariyaratne, J. K. P.; Green, M. L. H. J. Chem. Soc. 1963, 2976.

268

Bibler, J. P.; Wojcicki, A. J. Am. Chem. Soc. 1964, 86, 5051.

269

Green, M. L. H.; Smith, M. J. J. Chem. Soc. A, 1971, 3220.

270

Cooke, Jr., M. P. J. Am. Chem. Soc. 1970, 92, 6080.

271

(a) Collman, J. P.; Winter, S. R.; Clark, D. R. J. Am. Chem. Soc. 1972, 94, 1788. (b) Johnson, B. F. G.; Lewis, J.; Thompson, D. J. Tetrahedron Lett. 1974, 3789. (c) Cooke, Jr., M. P.; Parlman, R. M. J. Am. Chem. Soc. 1975, 97, 6863.

1018

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

concomitant migration led to coupling of the alkyl and acyl fragments to produce undecan-3-one, along with iron tricarbonyl [Fe(CO)3]. Overall when 167 was treated with triphenylphosphine, CO insertion occurred via an intramolecular migration and protonolysis led to undecan-3-one. C8 H17 C8 H17 Br

Na2Fe(CO)4

OC

Sodium tetracabonyl ferrate

CO

Fe

C8 H17

−

OC

Et-I

CO CO

OC

CO

167

CO

168 C8 H17

O CO

Et

Fe

OC

AcOH

Fe OC

O

- Fe(CO)3

Et

Et

C8H17

CO CO Undecan-3-one

169

The sequence shown generates an unsymmetrical ketone. A modification of this reaction involves addition of only one alkyl halide to sodium tetracarbonyl ferrate, followed by pressurization with CO. Reaction with 1-bromohexane gave the expected complex, 170. When treated with CO, carbonyl insertion gave 171 and subsequent oxidation with oxygen in the presence of a base led to heptanoic acid.272 Iodination in the presence of water also gave the acid, but when the reaction was done in ethanol ethyl, heptanoate was formed. Similarly, N,N-diethylheptanamide was generated by reaction with iodine and diethylamine.272 In general, the yields of ketone were higher with primary aliphatic halides and tosylates. Secondary substrates gave significant amounts of elimination, although this product could be minimized if THF was used as a solvent. O Et2N

I2 , NHEt2

O

C6H13 Na2Fe(CO)4

C6H13 Br

OC

Fe CO

C6H13

CO

C6H13 O

I2 , EtOH

CO

OC

Fe

CO

CO I 2 , H 2O

CO

1. O3, NaOH

CO

170

171

EtO

C6H13 O

HO

2. H2O

C6H13 O

HO

C6H13

The disconnection for this process follows: O R

R–X

+

R1–X

+

CO

R1

18.8.5.4 Chiral Organoiron Species In work reported independently by Davies and by Liebeskind, iron reagents (e.g., 172) were prepared, shown to be chiral, and resolved into the (R) and (S) antipodes. One of the phenyl groups in 172 effectively blocks one face of the carbonyl moiety, inducing high diastereoface selection in reactions at that carbonyl or at the adjacent carbon. The carbonyl partner of the Fe species can undergo enolate anion reactions similar to those of an acid derivative (Sections 13.4.1 and 13.4.2). Oxidation with iodine or bromine, as shown in Section 18.9.1,272 cleaves the CdFe bond

272

(a) Collman, J. P.; Winter, S. R.; Komoto, R. G. J. Am. Chem. Soc. 1973, 95, 249. (b) Watanabe, Y.; Yamashita, M.; Mitsudo, T.; Tanaka, M.; Yakegami, Y. Tetrahedron Lett. 1973, 3535. (c) Yamashita, M.; Watanabe, Y.; Mitsudo, T.; Takegami, Y. Ibid. 1976, 1585.

1019

18.9 ELECTROPHILIC IRON COMPLEXES

and generates the acid derivative. Davies et al.273 showed that the asymmetric acetyl derivative 172 can be deprotonated to form an enolate and condensed with an aldehyde (e.g., propanal) or alkylated (as with iodomethane). Reaction of 172 with butyllithium and then propanal led to a second reaction with excess butyllithium, followed by alkylation with iodomethane and oxidation of the CdFe bond with bromine gave (2R,3R)-3hydroxy-2-methylpentanoic acid with high asymmetric induction.273,274 O Cp

L

Me

Fe

Me HO2C

4. 2 C4H9Li 5. MeI 6. Br2 (oxidation)

CO

Ph3P

1. C4H9Li 2. Et2AlCl 3. EtCHO

172

OH (2R,3 R)-3-Hydroxy-2-methylpentanoic acid

Liebeskind and Welkers275 reacted 172 with lithium diisopropylamide (Section 13.2.2) and propanal, and showed that the enantioselectivity in the final β-hydroxy acids, (R)-3-hydroxypentanoic acid and (S)-3-hydroxypentanoic acid, was dependent on the conditions used to form the enolate. If diisobutylaluminum chloride (i-Bu2AlCl) was used, a 5.2:1 mixture of (R)-3-hydroxypentanoic acid and (S)-3-hydroxypentanoic acid was reversed (to 1:11.6), in 66% yield. The Fe group functions as a chiral auxiliary,276 where the acyl iron derivative behaves essentially as a protected acid. The chiral iron moieties are useful variation of enolate condensation chemistry (Section 13.4). In addition to the formation of the condensation product, the high asymmetric induction will prove valuable. O Cp Fe

Me CO

Ph3P

1. LDA 2. MX 3. EtCHO 4. Br2

HO HO2C

+ H

(R)-3-Hydroxypentanoic acid

172

OH HO2C

H

(S)-3-Hydroxypentanoic acid

The disconnections possible with these organoiron compounds follow: R

HO2C R1

HO2C

R +

R R1 –X HO2C

R2 OH R1

O HO2C

R

+ R1

R2

18.9 ELECTROPHILIC IRON COMPLEXES 18.9.1 Alkene and Diene Iron Complexes In Fe reactions where the reagent was equivalent to Cd, as described in Section 18.8.5, the iron moiety was used as an auxiliary. Iron can also stabilize cations, which then react with nucleophiles to generate new carbon-carbon bonds.277 These cations are formed as iron-alkene complexes, usually by reaction of cyclopentadienyl dicarbonyl ferrate anion (173) with an allylic halide (e.g., 3-chloro-2-methylprop-1-ene). The initial product in this case was the σ iron adduct (174). Subsequent reaction with an acid generated the π iron complex (175), where the double bond was a two-πelectron donor.278 273

(a) Davies, S. G.; Dordor-Hedgecock, I. M.; Warner, P.; Tetrahedron Lett. 1985, 26, 2125. (b) Ambler, P. W.; Davies, S. G. Ibid. 1985, 26, 2129. (c) Davies, S. G.; Dordor, I. M.; Walker, J. C.; Warner, P. Ibid. 1984, 25, 2709. (d) Davies, S. G.; Dordor, I. M.; Warner, P. J. Chem. Soc. Chem. Commun. 1984, 956. 274

Davies, S. G.; Walker, J. C. J. Chem. Soc. Chem. Commun. 1985, 209.

275

Liebeskind, L. S.; Welker, M. E. Tetrahedron Lett. 1984, 25, 4341.

276

(a) Liebeskind, L. S.; Welker, M. E. Tetrahedron Lett. 1985, 26, 3079. (b) Davies, S. G.; Easton, R. J. C.; Gonzalez, A.; Preston, S. C.; Sutton K. H.; Walker, J. C. Tetrahedron 1986, 42, 3987.

277

For reactions of this type, see Reference 38, p 445.

278

Green, M. L. H.; Nagy, P. L. I. J. Chem. Soc. 1963, 189.

1020

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

Cl

CO

H+

Fe Fe

CO 173

Fe

CO CO

CO CO

174

175

An alternative preparation treated an alkyl cyclopentadienyl dicarbonyl ferrate complex (e.g., 176) with trityl cation (Ph3C+) to form 177.279 Diene complexes can also be formed by reactions of iron pentacarbonyl [Fe(CO)5].280 The presence and nature of the substituents on the diene 178 will influence the reaction outcome with iron pentacarbonyl. Under thermal conditions, for example, a 4:1 mixture of 179 and 180 was obtained with greater selectivity for reaction at the unhindered face. Photolysis, however, led to a 1:1 mixture of 179 and 180.281 Treatment of an unconjugated diene with iron pentacarbonyl led to conjugated dienyl complexes, which were converted to the isomeric π-complexes with trityl carbocation.282 Cl

CO

H+

Fe Fe

CO 173

Fe

CO CO

174 Me

Me

Fe(CO)5

CO CO

175 Me

Me Me

Me

Fe(CO)3

Fe(CO)3

Me 178

Me

+ Me

179

180

18.9.2 Reactions of Iron Complexes With Nucleophiles Alkyliron complexes behave as electrophiles in the presence of nucleophiles. Rosenblum and Rosan283 showed that the ethene complex of cyclopentadienyl dicarbonyl iron (177) reacted with dimethyl lithiomalonate to yield 181. Conjugated Fe complexes can also be prepared.282 These alkene complexes react with a variety of other nucleophiles,284 including water, alcohols, amines,285 phosphines, or thiols, as well as carbon nucleophiles (enamines, organocuprates, enolates),286,287 and dialkylcadmium reagents (Section 11.4.3.2).288

Fe

Fe

LiCH(CO2Me)2

CO 177 279

CO2Me

OC

OC

CO2Me

CO 181

(a) Green, M. L. H.; Nagy, P. L. I. J. Organomet. Chem. 1963, 1, 58. (b) Laycock, D. E.; Hartgerink, J.; Baird, M. C. J. Org. Chem. 1980, 45, 291.

280

(a) Whitesides, T. M.; Arhart, R. W. Inorg. Chem. 1975, 14, 209. (b) Brookhart, M.; Whitesides, T. M.; Crockett, J. M. Ibid. 1976, 15, 1550. (c) Whitesides, T. M.; Arhart, R. W.; Slaven, R. W. J. Am. Chem. Soc. 1973, 95, 5792. (d) Impastato, F. J.; Ihrman, K. G. Ibid. 1961, 83, 3726.

281

(a) McArdle, P.; Higgins, T. Inorg. Chim. Acta 1978, 30, L303. (b) Davies, S. G. Organotransition Metal Chemistry: Applications to Organic Synthesis; Pergamon Press: Oxford, 1982; p 54.

282

(a) Birch, A. J.; Chamberlain, K. B. Org. Synth. Coll. 1988, 6, 996. (b) Birch, A. J.; Cross, P. E.; Lewis, J.; White, D. A.; Wild, S. B. J. Chem. Soc. A 1968, 332. (c) Birch, A. J.; Williamson, D. H. J. Chem. Soc. Perkin Trans. 1973, 1, 1892.

283

Rosan, A.; Rosenblum, M. J. Org. Chem. 1975, 40, 3621.

284

Davies, S. G. Organotransition Metal Chemistry: Applications to Organic Synthesis; Pergamon Press: Oxford, 1982.

285

Berryhill, S. R.; Rosenblum, M. J. Org. Chem. 1980, 45, 1984.

286

(a) Reference 284, p 132. (b) Busetto, L.; Palazzi, A.; Ros, R.; Belluco, U. J. Organomet. Chem. 1970, 25, 207; (c) Lennon, P.; Madhavarao, M.; Rosan, A.; Rosenblum, M. Ibid. 1976, 108, 93.

287

Reference 284, p 142.

288

Bayoud, R. S.; Biehl, E. R.; Reeves, P. C. J. Organomet. Chem. 1979, 174, 297.

1021

18.10 CONCLUSION

Diene complexes (e.g., 182) were converted to the corresponding cationic complexes (183) and reaction with the malonate anion gave 184. The Fe complex was removed by reaction with trimethylamine N-oxide (Me3NdO) to yield 185.289 CO2Me

CO2Me Ph3C+

CO2Me

MeO 182

MeO

Fe(CO)3 183

MeO2C

CO2Me

NaCH(CO2Me)2

Fe(CO)3

MeO

MeO2C

CO2Me CO2Me

Me3N+ –O−

MeO

Fe(CO)3 184

185

The iron carbonyl unit is a protecting group for dienes. Iron diene complex 186,290 for example, reacted with 187 in a Mukaiyama aldol reaction (Section 13.4.3) to give an 81% yield of 188 and 189 in an 84:16 ratio. In a synthesis of heptitol derivatives, Pearson and Katiyan291 protected tropone (cyclohepta-2,4,6-trien-1-one) as the iron tricarbonyl derivative (190), and prepared 191 using a multistep sequence before deprotection of the diene to yield 192. O CHO

1. BF3•OEt2, CH2Cl2 –78°C

OTMS (CO)3Fe

+

Ph

HO

2. TBAF, THF

Ph

HO

+

(CO)3Fe

Ph

O

(CO)3Fe

Bu Bu 186

187

188

Fe(CO)3

Fe2(CO) 9 PhH, 55°C

189

Fe(CO)3

O

O

O

Me3NO

O

O

Acetone

O Tropone

Bu

OTBS

OTBS 190 (96%)

191

192 (97%)

The disconnections for this section follow: R

O

O

+

+ R

R R

18.10 CONCLUSION This chapter discussed organometallic reagents that are useful in organic synthesis. The older copper-based reagents have been superseded by the more recent π-allylpalladium, π-allylnickel, π-allyliron, and other organometallic reagents.

289

Pearson, A. J. J. Chem. Soc. Perkin Trans. 1979, 1, 1255.

290

Harvey, D. F.; Selchau, V. B. J. Org. Chem. 2000, 65, 2282.

291

Pearson, A. J.; Katiyar, S. Tetrahedron 2000, 56, 2297.

1022

18. METAL-MEDIATED, CARBON-CARBON BOND-FORMING REACTIONS

HOMEWORK

1. Give a mechanistic rationale for this transformation:

HO O

O

Me3Al, CH2Cl2

OH

MeO

OSi(i-Pr)3

OMe

OSi(i-Pr)3

2. Give the major product for each of the following reactions:

(A)

S I

Me

Pd(OAc)2, PPh3 Ag2CO3, MeCN

O

B(OH)2

O

(B)

Tol

Cl

0.5% Pd2(dba)3, rt 1% P(t-Bu)3, THF 3.3 KF

CO2Me SnCl 4

(C)

O

Me AcO

(D)

CH2Cl2-H2O

OTBS OSiMe2t-Bu

I

O

NHCbz

Bn

Cbz

Me

2.5 equiv NaHCO3 Bu4NCl, MS, 60°C

Br

Br MeO

N

PhB(OH)2, H2O 3 eq K2CO3, rt

(H)

N

(G)

2 eq 5% Pd(OAc)2, DMF

I

(F)

Pd(OAc)2, NEt3 P(o-tolyl)3, 90°C MeCN

N

10% Pd(OAc)2, THF 25% dppp, 65°C

Ts

CO2Me

(E) BnO

N

HO

I 0.3% Pd/C, 12 h

cat Pd(OAc)2, rt cat Chiral phosphine K3PO4•THF

O B O

O

H

(I)

SiMe3

O O2, PdCl2 CuCl, 45°C

(J)

O

SnCl4, –78°C CH2Cl2, MeNO2

HO N

H aq DMF

OTf

(K)

Bu3Sn

O

OTHP

cat Pd(PPh3)4 LiCl, THF, Heat

THPO

3. In each case, give a complete synthesis of the target from the designated starting material. Show all intermediate products and give all reagents.

N

O O

(A) O

O

O

O

H OH

(B)

O

N CO2Et

O CO2Et O

1023

18.10 CONCLUSION

4. Explain the following transformation:

Mes

H H Me

Me

Cl Cl

5%

N

N Mes Ru PCy3

H

Me H

CH2KCH2 50°C, 10 h, PhH Then Reflux (10 h)

Me O

Me

H Ph

Me O

5. Give the major product for each of the following reactions:

CuCl, CuCl2

(A)

n-C3H7

CuCl, DMF

(B)

N

2

Ph

SiMe3

60°C

Cl CHO

(C)

CO2Me

N

Cl2(PCy3)2Ru=CHPh CH2Cl2

BF3•OEt2

O

TBDPSO

H

(E)

TiCl3

CHO

Me

SnBu3

2

(D)

1. Co2(CO)8, Ether, rt

(F)

N

THF

OBn

2. Me3NO, –10°C MS 4 Å, Toluene

O O Et

6. Provide a synthesis for each of the following: H

MeO2C H

N

(A)

Boc

O

MeO2C H

OAc

(B)

N

O

O

MeO

H N

H

OSiMe3

O

O

OTBDPS

Bn

N

O O

Bn N3

(C) OH

(D) Me

BocJN

O

CO2Me

H

H

OH

O

(G) MeO

H

(F) Ph O

OH HO

OAc

(E)

OH

H Me

O

H O

OAc

O

Ph

O O O

7. For each of the following give a complete reaction that illustrates its use for the formation of a carbon-carbon bond: (a) CdCl2 (b) Cp2TiMe2 (c) Na2Fe(CO)4 (d) i-Bu2AlH

C H A P T E R

19 Combinatorial and Process Chemistry Methodology and strategies for organic synthesis have been developed that transcend the typical bench chemistry utilized for the reactions presented in this book. In combinatorial chemistry, hundreds and sometimes thousands of products resulting from multi-step operations are produced as a mixture, but under conditions where individual synthetic products can be identified. This technique is typically used to screen reactions, reaction conditions, and even target molecules to produce compounds with certain characteristics and/or biological activity. When a suitable target molecule is identified, the methods used to develop the compound use virtually any of the reactions and reagents presented in this book, and more. However, to synthesize thousands or tens of thousands of kilograms of that compound in a commercial production, it is usually necessary to completely change the synthetic route. This new route must pay attention to the toxicity and stability of reagents, reactants, intermediate products, and each reaction itself. These scale-up procedures are referred to as process chemistry in the pharmaceutical industry. Finally, the ability to combine several reactions in a sequence, without formally isolating each product has been studied and automated reactors can produce puree products of multi-step reactions in a process known as continuous flow synthesis. Each of these important topics will be introduced in this chapter.

19.1 COMBINATORIAL CHEMISTRY Combinatorial chemistry is a term used to describe various microscale methods of solid-state synthesis and testing. It involves the synthesis of large numbers of compounds (called libraries), by doing reactions in a manner that produces large combinations of products, usually as mixtures.1 This approach has been called irrational drug design, since early approaches involved making a large vat of all possible chemical combinations of several reactants. Compare this approach with parallel synthesis, where the same reactions are repeated separately to produce many individual, but related products. In other words, a parallel synthesis means that a compound library is constructed by synthesizing many compounds in parallel, keeping each compound in a separate reaction vessel. When the final compounds are kept separate in this manner, methods are available to discover their identity. The pin-method of combinatorial synthesis, described below, is one such parallel synthesis method. Pooling strategies can also generate libraries of compounds, and the teabag method described below is a simple example of this type of approach. Representative methods of combinatorial synthesis will be described below, but there are several strategies for efficient synthesis of a target.1c Combinatorial syntheses can be performed both in solution and on a solid support.

19.1.1 Combinatorial Techniques for Synthesis In 1963, Merrifield2 introduced solid-state synthesis for the synthesis of peptides. This technique involves chemical functionalization of a polystyrene bead (or another polymeric bead) that reacts with the carboxylic acid portion of a 1

(a) Czarnik, A. W., DeWitt, S. H., Eds. A Practical Guide to Combinatorial Chemistry; American Chemical Society: Washington, DC, 1997. (b) Chaiken, I. N., Janda, K. D., Eds. Molecular Diversity and Combinatorial Chemistry: Libraries and Drug Discovery; American Chemical Society: Washington, DC, 1996. (c) Balkenhol, F.; von dem Bussche-H€ unnefeld, C.; Lansky, A.; Zechel, C. Angew. Chem. Int. Ed. 1996, 35, 2289. (d) Thompson, L. A.; Ellman, J. A. Chem. Rev. 1996, 96, 555. (e) Pavia, M. R., Sawyer, T. K., Moos, W. H., Eds. Bioorg. Med. Chem. Lett. Symposia-in-print no. 4, 1993, 3, 381. (f) Pirrung, M. C. Chem. Rev. 1997, 97, 473.

2

Merrifield, R. B. J. Am. Chem. Soc. 1963, 85, 2149.

Organic Synthesis http://dx.doi.org/10.1016/B978-0-12-800720-4.00019-2

1025

Copyright © 2017 Michael Smith. Published by Elsevier Inc. All rights reserved.

1026

19. COMBINATORIAL AND PROCESS CHEMISTRY

N-protected amino acid to give a polymer-bound amino ester (e.g., 1). When 1 is treated with a reagent to deprotect the amine, it can react with another N-protected amino acid, activated at the carbonyl, to yield a dipeptide. This procedure can be repeated to generate the desired polypeptide, and when the target has been attained a reagent is added to cleave the polypeptide from the bead (usually hydrolysis). This solid-state synthesis can be applied to other types of chemical transformations.3 R

O

N O H

O

= Polystyrene bead Ot-Bu

1

The fundamental idea of the Merrifield solid-state synthesis can be extended to use several beads, where each is attached to a different amino acid (V, I, S, etc., in Fig. 19.1).4 Subsequent reaction can generate a “pool” of peptides. If this mixture of beads was treated with three activated amino acids (V, I, S), a total of nine products would be generated, each one attached to a bead, as shown in Fig. 19.1. This approach can be expanded to generate a large number of peptides. If the bead containing V were made to react with 20 different amino acids, a mixture of 20 dipeptides would be generated. If each of these 20 dipeptides were to subsequently react with 20 different amino acids, 20
20 or 400 tripeptides would be generated as a mixture. If this 400-member tripeptide library were to react subsequently with 20 amino acids, a mixture of 20
20
20, or 8000 tetrapeptides, would be generated. If this process is continued, 204 (160,000) pentapeptides, 205 (320,000) hexapeptides, 206 (1,280,000,000) heptapeptides, and 207 (25,600,000,000) octapeptides would be generated; a huge library of compounds all mixed together.5

O O O

FIG. 19.1

VJV

O

V O

O

SJV

O SJI

O

IJS

O

O

O S

O IJI

O

VJS

O

O IJV

O

O

VJI

O

O

V+I+ S I

O

O

O

O

O

SJS

Generation of a nine-component dipeptide library.

It is easy to imagine variations that would build other libraries. A mixture of the three amino acids shown in Fig. 19.1, for example, could be mixed with 20 amino acids to generate 60 dipeptides. If this new library of dipeptides were reacted with 20 amino acids, 6020 different tetrapeptides (3.656
1035) compounds would be generated, which is a new library. O O-(CH2)6-CH2-X

O N H

(CH 2)5

N H 3

2

OH X

OMe X

O MeO

OR

OMe 4

3

NH2

5

(a) Crowley, J. I.; Rapoport, H. Acc. Chem. Res. 1976, 9, 135. (b) Leznoff, C. C. Acc. Chem. Res. 1978, 11, 327.

4

The 3-letter codes and the appropriate 1-letter code for the essential amino acids are given here. glycine (gly, G), alanine (ala, A), valine (val, V), leucine (leu, L), isoleucine (ile, I), phenylalanine (phe, F), serine (ser, S), threonine (thr, T), tyrosine (tyr, Y), cysteine (cys, C), methionine (met, M), asparagine (asn, N), glutamine (gln, Q), aspartic acid (asp, D), glutamic acid (glu, E), lysine (lys, K), tryptophan (trp, W), histidine (his, H), arginine (arg, R), proline (pro, P). (a) Nielsen, J. Chem. Ind. 1994, 902. (b) Sepetov, N. F.; Krchnˇ ák, V.; Stanková, M.; Wade, S.; Lam, K. S.; Lebl, M. Proc. Natl. Acad. Sci. USA 1995, 92, 5426.

5

1027

19.1 COMBINATORIAL CHEMISTRY

Early in the development of combinatorial chemistry, it was discovered that when the starting material was bound close to the polymer bead, chemical reactions were problematic, both in terms of chemical reactivity and for the chemical reaction required to release the final product from the bead. An important variation added a spacer between the bead and the starting material attached to the bead. Hydrocarbon chains connected to the ether linkage (as in 2) diamide groups, and polyamide chains (e.g., 3) are commonly used nowadays. The group is usually attached to the bead by an acid-labile unit, or one that is susceptible to enzymatic cleavage. Two common, and highly acid-labile anchors, are the aryl derivatives 4-(20 ,40 -dimethoxyphenylhydroxylmethyl)phenoxy (4) or 2-methoxy-4-alkoxybenzyl (5).6 Side view

Front view

Polyethylene bottle Seal Cleavage of the FMocprotecting group

Polypropylene mesh Tea bag Resin

Piperidine/dimethylformamide Washing steps

Label

(A)

Solvent: methanol and dimethylformamide DMF Methanol

Sorting of bags according to the coupling amino acids

Coupling Fmoc-amino acids TBTU/1-Hydroxybenzotriazine Diisopropylethylamine in dimethylformamide

Waste Computer

Washing steps

Tea bags in bottle

(B)

Solvent" methanol dimethylformamide

Shaker

(C) FIG. 19.2 The teabag approach to combinatorial peptide synthesis. Jung, G.; Beck-Sickinger, A. G. Angew. Chem. Int. Ed. 1992, 31, 367. Copyright © 1992 Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.

It is reasonable to question how the combinatorial theory can be translated to an actual experiment. Two methods will be presented to illustrate the fundamental techniques. Houghten7 described the first method, a teabag method, that could prepare as many as 150 different peptides with yields of up to 50 mg.8 A polypropylene bag (see Fig. 19.2)8 is charged with  100 mg of a polymeric support (usually polystyrene-1% divinylbenzene). Each bag contains a label and 6

(a) Rink, H. Tetrahedron Lett. 1990, 28, 3787. (b) Barlos, K.; Gratos, D.; Kallitsis, J.; Papaphotiu, G.; Sotiriu, P.; Wenqing, Y.; Sch€afer, W. Tetrahedron Lett. 1989, 30, 3943.

7

Houghten, R. A. Proc. Natl. Acad. Sci. USA 1985, 82, 5131.

8

(a) Pirrung, M. C. Chem. Rev. 1997, 97, 473. Also see (b) Jung, G.; Beck-Sickinger, A. G. Angew. Chem. Int. Ed. 1992, 31, 367.

1028

19. COMBINATORIAL AND PROCESS CHEMISTRY

is immersed in a polyethylene bottle containing piperidine-DMF. A computer-automated device adds amino acids, and the peptide is generated within each teabag. Treatment of the bags with a deprotection reagent and subsequent washing liberates the peptides.

Deprotect

Couple O N H

H N

H N

H N

Fmoc

O O Fmoc = Fluorenylmethyloxycarbonyl

FIG. 19.3 Peptides-on-pins synthesis with parallel coupling and deprotection, and structure of the linker on a pin. Reprinted with permission from Pirrung, M. C. Chem. Rev. 1997, 97, 473. Copyright © 1997 American Chemical Society.

A second technique accomplishes peptide synthesis on polyethylene rods (called pins), as developed by Geysen et al.9 A typical array consists of 96 pins (typically 4 mm in diameter and 40 mm in length),8 arranged in 8 rows of 12 pins (see Fig. 19.3). The pins fit into the wells of plates (see Fig. 19.3),8 which were initially developed for the enzyme-linked immunosorbent assay (ELISA). The polyethylene pins are functionalized with acrylic acid, and Nβ-Fmoc-β-alanyl-1,6-diaminohexane is attached as a spacer (see Fig. 19.3). The idea is to attach an amino acid to each pin, and add a solution containing the reactants into each well. The array of pins is lowered so that each pin is inserted into a well containing the reagents, and allowed to react. The pin array is removed from the plate containing the reactants after a predetermined time and thoroughly rinsed. To deprotect the terminal position, deprotection reagent is added to a new plate of wells, and the pin array is lowered such that each pin is immersed in a well containing the deprotection reagent. The “polyethylene rods were chosen because they do not swell or shrink, and any adsorbed molecules are only removed with difficulty. It is important to reiterate that thorough washing after each reaction step of the synthesis cycle is of utmost importance.”8 Recent work has shown that soft, but mechanically strong, polymer pins derived from styrene-divinylbenzene copolymers prepared by adjusting the reaction conditions10 can be cut into disks with good swelling characteristic in various solvents, and they are resistant to osmotic shock. Such disks provide an alternative support for solid-phase synthesis. Chemical reactivity is an important consideration. If a bead reacts with many different reactants (n coupling partners), there will be n products, which is also a problem in solution combinatorial synthesis. Individual substances can be obtained by fast parallel syntheses, that is, n substances are synthesized in n reaction vessels.1c It is reasonable to assume that some reactants will react faster, and in higher yields than others, so there may be different amounts of the n products. In other words, some coupling reactions occur at faster rates than others. This problem of different reaction rates can be avoided by the so-called split method.1d,11 As shown in Fig. 19.4,1d the solid support is divided into three equal parts, in this case, and each part is treated with a component A. The resulting products are mixed and divided again, 9

Geysen, H. M.; Meloen, R. H.; Barteling, S. J. Proc. Natl. Acad. Sci. USA 1984, 81, 3998. Also see Kerr, J. M.; Banville, S. C.; Zukermann, R. N. Bioorg. Med. Chem. Lett. 1993, 3, 463.

10

Hird, N.; Hughes, I.; Hunter, D.; Morrison, M. G. J. T.; Sherrington, D. C.; Stevenson, L. Tetrahedron 1999, 55, 9575.

11

Sebestyen, F.; Dibó, G.; Kovács, A.; Furkua, A. Bioorg. Med. Chem. Lett. 1993, 3, 413.

1029

19.1 COMBINATORIAL CHEMISTRY

giving three mixtures in which the resin-linked components (A1, A2, A3) are present in equimolar amounts. When these are treated with new reagents (B1, B2, B3), nine defined products are produced. Since there is only one reactant in solution in each reaction vessel at any one time, all reactions proceed to completion, despite possible differences in kinetics.1c Repetition of the cycle of division, reaction, and mixing gives large compound libraries in which all compounds are present in equimolar amounts. Ley and coworkers12 developed an approach for the rapid automated optimization of polymersupported reagents in synthesis using an optimized set of reaction conditions with an array of 80 compounds. C C

F A

F B

F C

E A

E B

E C

B

D A

D B

D C

A

Resin

I F C

H H H H H D D D E E A B C A B

H E C

H H F F A B

H F C

G G G G G D D D E E A B C A B

G E C

G G F F A B

G F C

G

D L

I I F F A B

H

E

A

I E C

I

F

B

I I I I I D D D E E A B C A B

L Recombine and mix

FIG. 19.4 Split synthesis method. Reprinted with permission from Thompson, L. A.; Ellman, J. A. Chem. Rev. 1996, 96, 555. Copyright © 1996 American Chemical Society.

When the library is a mixture of peptides, several methods have been developed to improve the overall process.13 One is the ladder peptide synthesis.14 For each coupling cycle of the peptide synthesis, a capping reagent is added to 10% of the product to effect partial termination. At the end of the synthetic sequence, each resin bead contains small amounts of sequence-specific termination products. The sequence of a given product is determined with matrix-assisted laser desorption ionization (MALDI) mass spectrometry, which measures mass differences between adjacent members of the termination series, making it possible to identify the corresponding monomers. In this work, the capping agent was N-acetyl-D,L-alanine. The concept of peptide arrays in combinatorial chemistry essentially involves making a microarray to generate very small amounts of each product.15 Microarray technologies include the use of photolithographic peptide synthesis on glass surfaces,16 and the so-called SPOT synthesis of peptides on membrane supports.17 In this latter method, cellulose paper sheets are used as absorptive membranes. Peptides are assembled by manual or automated spotting of small aliquots of solutions containing the activated amino acid derivatives onto marked positions on the sheets. The use of microarrays in combinatorial-type chemistry is an important area of research.18 Schreiber and coworkers19 developed a microarray system on glass slides. A glass microscope slide was treated with 3-(triethoxysilyl)propan-1-amine followed by 2,5-dioxopyrrolidin-1-yl 3-(2,5-dioxo-2,5-dihydro-1H-pyrrol-1-yl) 12

Jamieson, C.; Congreve, M. S.; Emiabata-Smith, D. F.; Ley, S. V. Synlett 2000, 1603.

13

Shin, D.-S.; Ki, D.-H.; Chung, W.-J.; Lee,Y.-S. J. Biochem. Mol. Biol. 2005, 38, 517.

14

(a) Youngquist, R. S.; Fuentes, G. R.; Lacey, M. P.; Keough, T. J. Am. Chem. Soc. 1995, 117, 3900. Also see Wang, X.; Peng, L.; Liu, R.; Gill, S. S.; Lam, K. S. J. Comb. Chem. 2005, 7, 197. 15

Southern, E. M. Great Britain Patent Application GB 8810400.5, 1988.

16

Fodor, S. P. A.; Read, J. L.; Pirrung, M. C.; Stryer, L.; Lu, A. T.; Solas, D. Science 1991, 251, 767.

17

Frank, R. Tetrahedron 1992, 48, 9217.

18

(a) Trabocchi, A.; Schreiber, S. L. Diversity-Oriented Synthesis: Basics and Applications in Organic Synthesis, Drug Discovery, and Chemical Biology; John Wiley & Sons, Inc.: New York, NY, 2013. (b) Jensen, K. J.; Shelton, P. T. Peptide Synthesis and Applications (Methods in Molecular Biology); Humana Press: ToTowa, NJ, 2013. (c) Uttamchandani, M.; Yao, S.Q. Small Molecule Microarrays: Methods and Protocols (Methods in Molecular Biology); Humana Press: ToTowa, NJ, 2010. (d) Cretich, M.; Chiari, M. Peptide Microarrays: Methods and Protocols (Methods in Molecular Biology); Humana Press: ToTowa, NJ, 2009. 19

MacBeath, G.; Koehler, A. N.; Schreiber, S. L. J. Am. Chem. Soc. 1999, 121, 7967.

1030

19. COMBINATORIAL AND PROCESS CHEMISTRY

propanoate (6) to yield a slide coated with maleimide units (see 7 in Fig. 19.5). This array was prepared to allow a method that was developed in the Schreiber laboratory, known as small molecule printing (SMP). As described by Schreiber and coworkers,19 synthesis beads were distributed into polypropylene microtiter plates at a density of one bead/well. The attached compounds were then released from their beads and dissolved in a small volume of a suitable solvent. A high-precision robot was used to pick up a small volume of dissolved compound from each well and repetitively deliver 1 nL of solution to defined locations on a series of chemically derivatized glass microscope slides. Each compound prepared contained a common functional group that mediated covalent attachment to the slide surface. In this way, compounds were prepared in an arrayed manner and immobilized on glass slides. Each slide was probed with a different tagged protein and any binding event was detected by a fluorescence-linked assay.19 O

O N

N O

OO

O OH Si O

OH Si O

OH Si

1.

H2 N O

Si(OEt)3

NH

NH

H2 N

O

O

2. N O

N

O

O 6

HO Si O Si

O Si

O

O Si OH

O Si

O

O Si

7

FIG. 19.5 Microarrays of synthesis beads. Reprinted with permission from MacBeath, G.; Koehler, A. N.; Schreiber, S. L. J. Am. Chem. Soc. 1999, 121, 7967. Copyright © 1999 American Chemical Society.

Schreiber and coworkers20 found that standard glass slides can be activated for the selective reaction with alcohols using high capacity 500–560-mm polystyrene beads. When equipped with an all hydrocarbon and silicon linker for the temporary attachment and eventual fluoride-mediated release of synthetic, alcohol-containing compounds, this microarray method can be used to prepare libraries of alcohols.21 Small molecules resulting from diversity-oriented syntheses can contain a wide array of functional groups, including secondary and phenolic hydroxyls. This alcoholarraying technique is compatible with split-pool synthesis. This microarray approach has been applied to the reparation of an inhibitor of a transcription factor,22 indole-like molecules,23 macrocyclic Hedgehog-pathway inhibitors,24 and spirooxiindoles.25

19.1.2 Deconvolution The split method is common to many modern combinatorial protocols. The library formed by these methods will be a mixture, and when the mixture must be tested in solution, structure determination can be very difficult. An important component of the split method is, therefore, deconvolution. An example of deconvolution is shown in Fig. 19.61c for a molecule with four variable substituents (A–D), where five different possibilities for each substituent lead to a total library of 54 ¼ 625 compounds. In the first deconvolution, 25 sublibraries are synthesized, each containing 25 compounds, where the substituents A and B are defined. Using a predetermined measure of activity, which compounds are being targeted, the most active sublibrary contains the optimal combination of substituents A and B, A2B1 20

Hergenrother, P. J.; Depew, K. M.; Schreiber, S. L. J. Am. Chem. Soc. 2000, 122, 7849.

21

Ref. 20,-Ref. 7 cited therein.

22

Koehler, A. N.; Shamji, A. F.; Schreiber, S. L. J. Am. Chem. Soc. 2003, 125, 8420.

23

Oguri, H.; Schreiber, S. L. Org. Lett. 2005, 7, 47.

24

Dockendorff, C.; Nagiec, M. M.; Weïwer, M.; Buhrlage, S.; Ting, A.; Nag, P. P.; Germain, A.; Kim, H.-J.; Youngsaye, W.; Scherer, C.; Bennion, M.; Xue, L.; Stanton, B. Z.; Lewis, T. A.; MacPherson, L.; Palmer, M.; Foley, M. A.; Perez, J. R.; Schreiber, S. L. ACS Med. Chem. Lett. 2012, 3, 808. 25

Lo, M. M.-C.; Neumann, C. S.; Nagayama, S.; Perlstein, E. O.; Schreiber, S. L. J. Am. Chem. Soc. 2004, 126, 16077.

1031

19.1 COMBINATORIAL CHEMISTRY

in Fig. 19.6. Based on this, five further sublibraries in the form A2B1CnC1–5 are prepared,1c each containing five components in which C is now defined. The optimal residue D is determined in the last step by synthesis of the five individual compounds in the form A2B1C3Dn, and in Fig. 19.6, the most active compound was A2B1C3D5.1c To get around the need for a new synthesis of another sublibrary for each assay of the total library, Han et al.26 used a procedure called recursive deconvolution. An aliquot is held back after each cycle of the split synthesis. From active mixtures, similarly in one or more steps, the single active compound is identified.1c No new split syntheses are involved since only the intermediates retained during the first split synthesis need to be elaborated further.1c A1-5 B1-5

D1-5

C1-5 Total number of compounds = 54 = 625 A1 B1

A2 D1-5

A2

B1

D1-5

C1-5

B2

C1-5

A5

A4 D1-5

B4 .........

C1-5

A5

B4

D1-5

D1-5

B5

C1-5

C1-5

D1-5 C1-5

25 Mixtures (25 compounds each)

A2 B1

D1-5 C1

B1

D1-5

B1

B1

D1-5

B1

D1-5

D1-5 C5

C4

C3

C2

A2

A2

A2

A2

25 Mixtures (5 compounds each)

B1

D1 C3

B1

D2 C3

B1

D3 C3

A2

A2

A2

A2

A2

B1

D4 C3

B1

D5

5 Single compounds

C3

FIG. 19.6 Identification of the most active library component by deconvolution. Reprinted with permission from Balkenhol, F.; von dem Bussche-H€ unnefeld, C.; Lansky, A.; Zechel, C. Angew. Chem. Int. Ed. 1996, 35, 2289. Copyright © 1996 Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.

19.1.3 Product Identification Is it reasonable to ask how each product of a massive library can be identified as a unique entity and isolated? There are several techniques by which this can be accomplished. All products can be detached from the beads and isolated as a mixture. This mixture can be analyzed using high-performance liquid chromatography (HPLC) and gas chromatography (GC) techniques, particularly when combined with mass spectrometry (HPLC-mass spectrometry and GC-mass spectrometry). In principle, each component can be separated and identified.11,27 Direct monitoring of solid-phase reactions, on the bead itself without prior cleavage from the resin is possible using soft-laser desorption 26

(a) Han, H.; Wolfe, M.; Brenner, S.; Janda, K. D. Proc. Natl. Acad. Sci. USA 1995, 92, 6419. (b) Erb, E.; Janda, K.D.; Brenner, S. Proc. Natl. Acad. Sci. USA 1994, 91, 11422.

27

For examples of peptide analysis using these techniques, see (a) Hean, M. T. W., Regnies, F. E., Wehr, C. T, Eds., High Performance Liquid Chromatography of Proteins and Peptides; Academic Press: New York, NY, 1983. (b) Hean, M. T. W. Adv. Chromatogr. 1982, 20, 1. (c) Bruins, A. P.; Covey, T. R.; Henion, J. D. Anal. Chem. 1987, 59, 2642. (d) Bruins, A. P.; Weidolf, O. G.; Henion, J. D.; Buddle, W. L. Anal. Chem. 1989, 59, 2647. (e) Grolo, M.; Takeuchi, T.; Ishii, D. Adv. Chromatogr. 1989, 30, 167.

1032

19. COMBINATORIAL AND PROCESS CHEMISTRY

time-of-flight mass spectrometry, if photocleavable phenacyl ester or o-nitroveratryl linker groups are used.28 Lam and coworkers29 reported a “one-bead-one-compound” encoding approach, based on mass spectrometry. As stated by Lam, prior to library synthesis the inner core of each bead is derivatized with three to four different coding blocks on a cleavable linker.30 An individual coding block that contains a functional group with the same chemical reactivity encodes each functional group on the scaffold.30 During the library synthesis, the same chemical reactions take place on the scaffold (outer layer of the bead) and coding blocks (inner core of the bead) concurrently.30 After screening, the coding tags in the positive beads are released, followed by mass determination of the molecules using MALDI Fourier transform mass spectrometry.30 It would be useful to know the bead to which each product is attached. A solution to this problem is the use of a molecular tag, work pioneered by Brenner and Lerner31 in 1992, and by Nielsen et al.9,32 The idea is to give a binary chemical tag for each reagent. In the work of Still and coworkers,33 a constant segment of a library is attached to a polymeric bead using standard solid-phase methods. A tag for the first reagent is attached, and then the amino acid is attached. After processing, a second tag is attached, followed by the second amino acid in the sequence. Still and coworker’s33 coding strategy used haloaromatic tags (e.g., 8).1d When attached as an amide, the chain length (n) can be varied along with the haloaryl ether unit, to create a binary synthesis code. The reagents can be designated in binary as 001 (reagent 1), 010 (reagent 2), 011 (reagent 3), to 111 (reagent 7). If reagent 3 were used in the first step and reagent 1 in the second and reagent 6 in the third, the description would be 001 011 011.33 In 8, nine different haloaryl units could be attached to create nine tags, T1-T9 and each could be linked to one reagent. Therefore, aryl halide A might be 001 (T1), B 002 (T2), and C 011 (T6), and so on. Two possibilities are pentachlorophenyl and 2,4,6-trichlorophenyl. T, T1, T2, etc. O O

Cl O

(CH 2)n

Cl

OAr

Cl

Cl

Cl Cl

F

Ar =

HO

NO2

Cl

O Linker

Electrophoric tag

Cl

Cl

A

B

Cl C

8

The use of different haloaromatic units will result in a large number of unique tags. As the peptide chain grows, each tag gives the exact sequence of that chain. In this case, any attached tag can be detached by UV and decoded by electron capture GC. Still and coworkers33 related the tagging molecules to the binary bits of the synthesis code by arranging them (T1-sT9) in order of their GC elution order, where T1 is retained the longest on the GC column used and designates the rightmost bit of the binary synthesis code. In this way, the tags are easily identified and the product can be identified. An interesting solution to product identification is to color code the beads and the vessel in which the reaction is done.30 If there were eight subunits, for example, each one could be partitioned into different containers with different color caps. If each subunit were attached to a different color bead, one bead of each color could be added to each color-coded vessel. When the next subunit is attached, the compounds formed can be sorted individually by cap and by color. This process can be contained as each new subunit is attached. Another color-coded approach can be used to identify products that are susceptible to a particular chemical reaction. The approach couples the acceptor molecule to an enzyme (e.g., alkaline phosphatase) or to a fluorescent, and then adds these in soluble form to generate the peptide-bead library.34 Those beads that contain compounds active for this analysis will be stained and are usually visible to the naked eye or under a lowpower microscope. The stained beads can then be removed for analysis. In one reported study, a monoclonal antibody 28

Gerdes, J. M.; Waldmann, H. J. Comb. Chem. 2003, 5, 814.

29

Song, A.; Zhang, J.; Lebrilla, C. B.; Lam, K. S. J. Am. Chem. Soc. 2003, 125, 6180.

30

Guiles, J. W.; Lanter, C. L.; Rivero, R. A. Angew. Chem. Int. Ed. 1998, 37, 926.

31

Brenner, S.; Lerner, R. A. Proc. Natl. Acad. Sci. USA 1992, 89, 5381.

32

Nielsen, J.; Brenner, S.; Janda, K. D. J. Am. Chem. Soc. 1993, 115, 9812.

33

Ohlmeyer, M. H. J.; Swanson, R. N.; Dillard, L. W.; Reader, J. C.; Asouline, G.; Kobayahsi, R.; Wigler, M.; Still, W. C. Proc. Natl. Acad. Sci. USA 1993, 90, 10922.

34

Lam, K. S.; Salmon, S. E.; Hersh, E. M.; Hruby, V. J.; Kazmierski, W. M.; Knapp, R. J. Nature (London), 1991,354, 82.

19.1 COMBINATORIAL CHEMISTRY

1033

active against β-endorphin was of interest because it had a high affinity for the epitope sequence UCCFL,30 and six reactive beads were obtained from 2 million screened beads from a pentapeptide library. A powerful method to identify individual components of a combinatorial library is known as encoded combinatorial chemistry. Brenner and Lerner31 designed a method where each chemical sequence is labeled with a so-called genetic tag (nucleotide sequences), and each tag is constructed by chemical synthesis. A monomeric chemical unit is attached to a polymeric structure and this is followed by addition of an oligonucleotide sequence. The sequence of that oligonucleotide constitutes a tag. Active products from the library are selected by binding to a receptor. After the chemical entity is bound to a target, the genetic tag can be amplified by replication, and utilized for enrichment of the bound molecules by serial hybridization to a subset of the library.31 The nature of the chemical structure bound to the receptor is decoded by sequencing the nucleotide tag.31 In this way, so-called DNA Encoded chemical libraries are prepared and characterized by individual small organic chemical moieties on DNA fragments that function as amplifiable identification barcodes, allowing the simultaneous screening of very large sets of compounds. Any potential hit compounds can be identified and quantified by polymerase chain-reaction-amplification of the DNA barcode followed by DNA sequencing.35 An encoding technique is illustrated by synthesizing a peptide and a nucleotide chain in an alternating, bidirectional manner.5 Sequencing the nucleotide chain determines the sequence of the peptide chain. Another technique is dynamic combinatorial chemistry, “in which simple molecular units (building blocks) are held together by noncovalent or reversible covalent bonds, generating a complex mixture of products that continuously interconvert: The composition of the mixture at equilibrium is thermodynamically controlled and is referred to as a dynamic combinatorial library.”36 Several different reversible reactions can be used in this approach.37

19.1.4 Synthesis Using Combinatorial Techniques An overriding theme for this entire book is synthesis. How is combinatorial chemistry pertinent to this discussion? Many small molecules can be synthesized using combinatorial techniques and evaluated for biological activity. One example is the synthesis of substituted benzodiazepine precursors (e.g., 9) on a bead, and release from the bead yields 10, as shown in Scheme 19.1.38 This approach has become important to the pharmaceutical industry for drug discovery.39 Originally used to develop the Merrifield process (see 1), peptides are an important target for combinatorial chemistry,40 as described in Sections 19.1.1–19.1.3. Millions or billions of peptides (peptide libraries) can be prepared by combinatorial chemistry. Peptide libraries can subsequently be screened as enzymatic substrates, enzymatic inhibitors, or for cell-binding peptides.41 Cyclic peptides are important targets, but ring-closure methods for small-to-medium size cyclic peptides are commonly slow, and cyclodimerization is an annoying side reaction.42 Chiba and coworkers43 reported a soluble tagassisted method that can be applied to head-to-tail cyclization reactions, using mahafacyclin B as a model. The tagged glycine derivative 11 was used in this study.

35

Franzini, R. M.; Neri, D.; Scheuermann, J. Acc. Chem. Res. 2014, 47, 1247.

36

Cougnon, F. B. L.; Sanders, J. K. M. Acc. Chem. Res. 2012, 45, 2211.

37

(a) Corbett, P. T.; Leclaire, J.; Vial, L.; West, K. R.; Wietor, J.-L.; Sanders, J. K. M.; Otto, S. Chem. Rev. 2006, 106, 3652. (b) Herrmann, A. Org. Biomol. Chem. 2009, 7, 3195. (c) Ladame, S. Org. Biomol. Chem. 2008, 6, 219. (d) Lehn, J.-M. Chem. Soc. Rev. 2007, 36, 151. (e) de Bruin, B.; Hauwert, P.; Reek, J. N. H. Angew. Chem. Int. Ed. 2006, 45, 2660. 38

(a) Bunin, B. A.; Plunkett, M. J.; Ellman, J. A. Proc. Natl. Acad. Sci. USA 1994, 91, 4708. Also see (b) DeWitt, S. H.; Kiely, J. S.; Stankovic, C. J.; Schroeder, M. C.; Cody, D. M. R.; Pavia, M. R. Proc. Natl. Acad. Sci. USA 1993, 90, 6909.

39 (a) Gallop, M. A.; Barrett, R. W.; Dower, W. J.; Fodor, S. P. A.; Fordon, E. M. J. Med. Chem. 1994, 37, 1233. (b) Gordon, E. M.; Barrett, R. W.; Dower, W. J.; Fodor, S. P.; Gallop, M. A. J. Med. Chem. 1994, 37, 1385. (c) Alper, J. Science, 1994, 264, 1399. (d) Houghten, R. A.; Pinilla, C.; Blondelle, S. E.; Appel, J. R.; Dooley, C. T.; Cuervo, J. H. Nature (London) 1991, 354, 84. 40

See Shin, D.-S.; Kim, D.-H.; Chung, W.-J.; Lee, Y.-S. J. Biochem. Mol. Biol. 2005, 38, 517.

41

Gray, B. P.; Brown, K. C. Chem. Rev. 2014, 114, 1020.

42

White, C. J.; Yudin, A. K. Nat. Chem. 2011, 3, 509.

43

(a) Fujita, Y.; Fujita, S.; Okada, Y.; Chiba, K. Org. Lett. 2013, 15, 1155. Also see (b) Masuda, Y.; Tanaka, R.; Kai, K.; Ganesan, A.; Doi, T. J. Org. Chem. 2014, 79, 7844. (c) Mashiach, R.; Meijler, M. M. Org. Lett. 2013, 15, 1702.

1034

19. COMBINATORIAL AND PROCESS CHEMISTRY

SCHEME 19.1 Combinatorial synthesis of benzodiazepines.

Ph O

O Ph

N H O

HN

NH

O HN

C22H45O

OC22H45

Tag

Ph NH O Ph

H N

HN

O

HN

O HO Mahafacyclin B

CO2Me

11

Many combinatorial-based syntheses have been reported.44 Schreiber and coworkers45 described a synthesis strategy that yielded diverse small molecules by combinatorial techniques. Andrus et al.,46 with the goal of multidrug-resistance reversal, have prepared a solution-phase indexed combinatorial library of non-natural polyenes (e.g., 12). Varying R and R1 in 12 generated this library. Ellman and coworkers47 reported a combinatorial library of synthetic receptors targeting vancomycin-resistant bacteria, and Paterson et al.48 prepared polyketide-type libraries by iterative asymmetric aldol reactions on a solid support. Rieser and coworkers49 used combinatorial liquid-phase synthesis to prepare [1,4]-oxazepine-7-ones by the Baylis-Hillman reaction (see Section 13.7.2). Schreiber and coworkers50 reported the synthesis and evaluation of a library of polycyclic small molecules for use in chemical genetic assays. Bauer et al.51 reported a library of N-substituted 2-pyrazoline compounds (e.g., 13), by parallel solution-phase synthesis, for screening as therapeutic agents as antibacterials, antivirals, and anti-inflammatory compounds. Libraries of trisaccharides have been prepared by combinatorial methods.52 Han et al.53 reported highly substituted thiophene derivatives (e.g., 14) as novel phosphodiesterase-4 (PDE-4) inhibitors, and Ley et al.54 reported the solution-phase synthesis of functionalized bicyclo[2.2.2]octanes. Libraries of heterocyclic compounds have been prepared from peptides 44

For reviews, see Dolle, R. E.; Le Bourdonnec, B.; Morales, G. A.; Moriarty K. J.; Salvino, J. M. J. Comb. Chem. 2006, 8, 597; Dolle, R.E. J. Comb. Chem.; Idem 2005, 7, 623; 2004, Idem 6, 623; 2003, 5, 693; Idem 2002, 4, 369; Idem 2001, 3, 477; Idem 2000, 2, 383.

45

Burke, M. D.; Berger, E. M.; Schreiber, S. L. J. Am. Chem. Soc. 2004, 126, 14095.

46

Andrus, M. B.; Turner, T. M.; Asgari, D.; Li, W. J. Org. Chem. 1999, 64, 2978.

47

Xu, R.; Greiveldinger, G.; Marenus, L. E.; Cooper, A.; Ellman, J. A. J. Am. Chem. Soc. 1999, 121, 4898.

48

Paterson, I.; Donghi, M.; Gerlach, K. Angew. Chem. Int. Ed. 2000, 39, 3315.

49

R€ acker, R.; D€ oring, K.; Reiser, O. J. Org. Chem. 2000, 65, 6932.

50

Tan, D. S.; Foley, M. A.; Stockwell, B. R.; Shair, M. D.; Schreiber, S. L. J. Am. Chem. Soc. 1999, 121, 9073.

51

Bauer, U.; Egner, B. J.; Nilsson, I.; Berghult, M. Tetrahedron Lett. 2000, 41, 2713.

52

Takahshi, T.; Adachi, M.; Matsuda, A.; Doi, T. Tetrahedron Lett. 2000, 41, 2599.

53

Han, Y.; Giroux, A.; Lepine, C.; Laliberte, F.; Huang, Z.; Perrier, H.; Bayly, C. I.; Young, R. N. Tetrahedron 1999, 55, 11669.

54

Ley, S. V.; Massi, A. J. Chem. Soc. Perkin Trans. 1 2000, 3645.

1035

19.2 PROCESS CHEMISTRY

and polyamides.55 Wolf and Hawes56 also showed that combinatorial techniques are useful for high-throughput screening of enantioselective catalysts, allowing rapid evaluation of their utility. Me

R1

N

N

O O

Me Me

N

N

S

HN OBn

N

NO2 S

OH R

N 12

HO2C 13

14

19.2 PROCESS CHEMISTRY Most of this book has been devoted to understanding and using chemical reactions in organic synthesis. For the most part, the reactions reported in the chemical literature citations involve transformations of a few grams to a few milligrams. In the pharmaceutical industry, drug discovery involves discovering a “lead compound” and devising a workable synthesis of that compound. Sufficient material must be synthesized to allow the proper identification and characterization of the compound, and sufficient material for preliminary biological evaluation. All byproducts and secondary products must be identified and characterized, and hazards with all compounds produced by the synthesis identified. If the synthesized lead compound has important biological activity, but there appears to be toxicity issues or other problems, the initially synthetized target is structurally modified to improve efficacy and diminish deleterious side effects. Once a promising compound is found, the synthesis is scaled up to provide sufficient material for clinical trials. If the compound survives clinical trials and is identified for commercial production, large amounts of the compound must be prepared, often on an industrial scale. Many times completely new routes are required for safety and cost considerations. In many cases, the reactions and reagents and solvents used in the initial drug discovery process must be changed to accommodate large-scale reactions. In practice, many of these issues must be addressed during the initial scale-up procedures used to provide material for clinical trials. The process of modifying a drug discovery route to a large-scale preparative route is known as Process Chemistry. As noted, when chemical reactions and transformations are done on a multi-gram, multi-kilogram or industrial scale, the conditions of the reaction and the workup often require drastic modification. In the pharmaceutical industry, process chemistry is a critically important component if a drug or other chemical products are brought to the market.57

19.2.1 Principles of Green Chemistry For a large-scale chemical process, several criteria are important. The process “not only needs to be robust and predictable; it should also be operationally simple, safe, and straightforward. Ideally, reactions should use inexpensive, environmentally benign starting materials, reagents, and solvents and produce the target compound not only in high yield but also in very high quality as well, with a minimum of impurities that are easily removed, preferably by crystallization. If the process is catalytic, turnover numbers and turnover frequencies must be high and the product must be free of trace contaminants (e.g., heavy metal salts or complexes).”58 Green chemistry, or environmentally friendly chemistry, is very important in today’s world, and critically important. The Environmental Protection Agency (EPA) defines green chemistry as follows: the design of chemical products and processes that reduce or eliminate the use or generation of hazardous substances.59 Green chemistry has also been called 55

Nefzi, A.; Ostresh, J. M.; Yu, J.; Houghten, R. A. J. Org. Chem. 2004, 69, 3603.

56

Wolf, C.; Hawes, P. A. J. Org. Chem. 2002, 67, 2727.

57

See Yasuda, N., Ed. The Art of Process Chemistry; Wiley-VCH: Weinheim, 2010. Also see, Zhang, T. Y. Chem. Rev. 2006, 106, 2583.

58

Lipton, M. F.; Barrett, A. G. M. Chem. Rev. 2006, 106, 2581.

59

Available at http://www2.epa.gov/green-chemistry/basics-green-chemistry.

1036

19. COMBINATORIAL AND PROCESS CHEMISTRY

sustainable chemistry. Green chemistry differs from cleaning up pollution in that “green chemistry reduces pollution at its source by minimizing or eliminating the hazards of chemical feedstocks, reagents, solvents, and products.”59 If a reaction or multi-step synthetic sequence is to be scaled up to commercial and industrial levels, waste production and waste disposal, toxicity of chemicals and by-products, hazards to health and the environment, not to mention legal issues and the cost of production are important issues that must be addressed. In other words, the principles of green chemistry should be incorporated into process chemistry. It has been stated, “green chemistry is good process chemistry.”60 Choosing the proper solvent for a process, for example, is important, but not always a simple matter. Developing and modifying the chemical synthesis of a molecule with multiple functional groups, and perhaps sensitive functional groups, and sophisticated structural features is clearly complex. How can green chemical principles be incorporated into a complex process chemistry problem? Perhaps the place to begin is to state the “Twelve Principles of Green Chemistry,”59,61 which are taken as the main framework of sustainable chemistry.62

1. Prevent waste: Design chemical syntheses to prevent waste, leaving no waste to treat or clean up. 2. Design safer chemicals and products: Design chemical products to be fully effective, yet have little or no toxicity. 3. Design less hazardous chemical syntheses: Design syntheses to use and generate substances with little or no toxicity to humans and the environment. 4. Use renewable feedstocks: Use raw materials and feedstocks that are renewable rather than depleting. Renewable feedstocks are often made from agricultural products or are the wastes of other processes; depleting feedstocks are made from fossil fuels (petroleum, natural gas, or coal) or are mined. 5. Use catalysts, not stoichiometric reagents: Minimize waste by using catalytic reactions. Catalysts are used in small amounts and can carry out a single reaction many times. They are preferable to stoichiometric reagents, which are used in excess and work only once. 6. Avoid chemical derivatives: Avoid using blocking or protecting groups or any temporary modifications if possible. Derivatives use additional reagents and generate waste. 7. Maximize atom economy: Design syntheses so that the final product contains the maximum proportion of the starting materials. There should be few, if any, wasted atoms. 8. Use safer solvents and reaction conditions: Avoid using solvents, separation agents, or other auxiliary chemicals. If these chemicals are necessary, use innocuous chemicals. 9. Increase energy efficiency: Run chemical reactions at ambient temperature and pressure whenever possible. 10. Design chemicals and products to degrade after use: Design chemical products to break down to innocuous substances after use so that they do not accumulate in the environment. 11. Analyze in real time to prevent pollution: Include in-process real-time monitoring and control during syntheses to minimize or eliminate the formation of byproducts. 12. Minimize the potential for accidents: Design chemicals and their forms (solid, liquid, or gas) to minimize the potential for chemical accidents including explosions, fires, and releases to the environment.

19.2.2 Scale-Up Problems and Safety Issues In most drug discovery programs a synthetic route is developed using almost any available reagent, at least in many cases. This initial route is probably not amenable to scale-up into a commercial process, and process chemists must change the synthetic route, sometimes several times. The main types of issues associated with process and worker safety are as follows: (1) thermal runaway, (2) gas evolution, (3) potentially explosive, shock-sensitive materials, (4) highly corrosive materials, (5) acute toxicity, (6) chronic toxicity, (7) genotoxicity, (8) pyrophoric and highly flammable materials.63

60

Laird, T. Org. Process Res. Dev. 2012, 16, 1.

61

Anastas, P. T.; Warner, J. C. Green Chemistry: Theory and Practice; Oxford University Press: New York, NY 1998, p 30.

62

Appell, R.; Gala, D.; Sanghvi, Y. S. Org. Process Res. Dev. 2011, 15, 898. Also see Ritter, S. K. Chem. Eng. News Archive 2010, 88 (43), 45.

63

Butters, M.; Catterick, D.; Craig, A.; Curzons, A.; Dale, D.; Gillmore, A.; Green, S. P.; Marziano, I.; Sherlock, J.-P.; White, W. Chem. Rev. 2006, 106, 3002.

19.2 PROCESS CHEMISTRY

1037

Thermal and reactive hazards constitute one issue and toxic hazards another. Safety data for commercial compounds is available from the material safety data sheet or other sources (e.g., Sax’s Dangerous Properties of Industrial Materials).64 For new or otherwise unknown compounds, the toxicity should be measured, and an Ames test should be done. The Ames test is a biological assay that will assess the potential of chemical compounds to be mutagens.65 Since many cancers are linked to DNA damage, the test also serves as a quick assay to estimate the carcinogenic potential of a compound. Any chemical process, but certainly large-scale processes, will generate waste.66 Any and all chemical waste and unwanted byproducts that cannot be recycled must be properly disposed of, which can be quite expensive. Information is publically available for toxic waste and waste management laws.67 For protection of the environment, protection of the people involved in the commercial production, and to minimize costs, the goal is to keep waste to a minimum. This goal leads to a consideration of green chemistry principles for process chemistry.68 Of all the energy used in a pharmaceutical process, solvent and post-treatment green house gas emissions are quite important.69 Therefore, solvent selection and potential recovery is an important consideration. Once potential problem solvents are identified, it is possible to design a process that will minimize the type and amount of solvent. Note that if only one solvent can be used, it is probably easier to recycle or dispose of. It goes without saying, that when designing a scale-up process, toxic compounds and toxic solvents should be avoided, and all chemical components for all reactions must be examined. Some standard is required to assess the toxicity of a given component. A standard measure of toxicity as it relates to humans is the LD50, stated in milligrams (mg) of pesticide per kilogram (kg) of body weight. The LD50 is the individual dose required to kill 50% of a population of test animals. The LD50 values in Table 19.1 can be used to evaluate compounds to be used in process research.70 TABLE 19.1

Toxicity Levels Associated with LD50 Values

LD50(mg kg-1)

Toxicity

15,000

Low toxicity

Reprinted with permission from From Bench to Pilot Plant. Process Research in the Pharmaceutical Industry, Nafissi, M.; Ragan, J. A.; DeVries, K. M. ACS, Washington, DC, 2002, Table 1, p. 9. Copyright © 2002 American Chemical Society.

There are compilations that correlate the toxicity of various solvents.71 Solvents that are useful for sustainable chemistry have also been listed as “Sanofi’s Solvent Selection Guide,” and categorized as “banned,” “recommended,” or “substitution advisable.”72 As an example, ethyl acetate is recommended, but methyl acetate is categorized as substitution advisable. Diethyl ether is listed as banned, THF as substitution advisable, and anisole is recommended. 64

Lewis, R. J., Sr. Sax’s Dangerous Properties of Industrial Materials—3 Volume Set;, Wiley-Interscience: Hoboken, NJ, 2000.

65

Breslow, L., Ed. Ames Test, Encyclopedia of Public Health; Gale Cengage: Farmington Hills, MI, 2002. Ames, B. N.; Gurney, E. G.; Miller, J. A.; Bartsch, H. Proc. Natl. Acad. Sci.USA 1972, 69, 3128.

66

See The Toxics Release Inventory (TRI) available at http://www.epa.gov/tri/.

67

The US Environmental Protection Agency (EPA) maintains this information in a national database called the Toxics Release Inventorys, which is available to the public via the Internet, available at www.epa.gov/tr 68

Laird, T. Org. Process Res. Dev. 2012, 16, 1. Also see Ager, D. J. Managing Hazardous Reactions and Compounds in Process Chemistry; ACS Symposium Series, 2014; Vol. 1181; Chapter 12, pp 285–351.

69

Constable, D. J. C.; Curzons, A. D.; Cunningham, V. L.; Mortimer, D. N. Green Chem. 2001, 3, 1.

70

Taken from Nafissi, M., Ragan, J. A., DeVries, K. M., Eds. From Bench to Pilot Plant. Process Research in the Pharmaceutical Industry; ACS: Washington, DC, 2002, Table 1, p 9. 71

Taken from Nafissi, M., Ragan, J. A., DeVries, K. M. Eds. From Bench to Pilot Plant. Process Research in the Pharmaceutical Industry; ACS: Washington, DC, 2002, Table III, pp 11–12, Table IV, p 13.

72

Prat, D.; Pardigon, O.; Flemming, H,-W.; Letestu, S.; Ducandas, V.; Isnard, P.; Guntrum, E.; Senac, T.; Ruisseau, S.; Cruciani, P.; Hosek, P. Org. Process Res. Dev. 2013, 17, 1517.

1038

19. COMBINATORIAL AND PROCESS CHEMISTRY

All chlorinated solvents are listed as either banned, or substitution is advisable, and while pentane is listed as banned, most other common alkane solvents are listed as substitution advisable.72 As a practical matter, very toxic, seriously toxic, and especially dangerously toxic material should be avoided, and carcinogenic materials should not be used. When applied to the choice of a solvent, carcinogenic solvents include benzene, chloroform, carbon tetrachloride, dioxane, and hexamethylphosphoric triamide (HMPT). A casual inspection of the many reactions discussed in Chapters 2–18 show that while some, if not most, of these solvents are common, but they should be avoided in scale-up and process reactions. When modifying a drug discovery synthesis, such solvents are usually replaced with a less toxic or nontoxic alternative.73 Ethyl acetate can be used to replace chloroform or carbon tetrachloride, for example, toluene or xylene replaces benzene, and THF replaces dioxane.74 aq NaOH (10 equiv) 20 mM (5:2) Dioxane:Water

Br2HC

CHBr2 CHBr2

Br2HC

110°C, 30 min

Br

Br

(83%)

aq NaOH (9.5 equiv)

Br

Br

0.11 M i-PrOH 80°C , 1 h (79%)

15

16

A kilogram-scale synthesis of corannulene required the preparation of 1,2,7,8-tetrabromocorannulene (16) as a key intermediate, prepared by the reaction of 15 with an excess of NaOH in a dilute solution of dioxane. To scale-up this reaction the cost and toxicity of dioxane as a solvent was a problem. An improved synthesis showed that the reaction of 15 with an excess of NaOH in isopropyl alcohol gave better yields, 79% of 16, and avoided the cost and toxicity issues associated with dioxane.75 It is also important to avoid volatile and flammable solvents (e.g., pentane or hexane), or diethyl ether. In general, solvents with a flash point of g d1 s)/daily dose (g d1). If the dose is >10 g d1, other options are used to determine the concentration. Typical concentrations permitted for oral and parenteral administration (not via the intestinal tract; commonly, intravenously) are shown in Table 19.2.86

82

Lizarraga, E.; Zabaleta, C.; Palop, J. A. J. Therm. Anal. Calorim. 2007, 89, 783.

83

Naka, T.; Ozawa, H.; Toyama, K.-I.; Shirasaka, T. Org. Process Res. Dev. 2015, 19, ASAP.

84

Hida, T.; Mitsumori, S.; Honma, T.; Hiramatsu, Y.; Hasizume, H.; Okada, T.; Kakinuma, M.; Kawata, K.; Oda, K.; Hasegawa, A.; Masui, T.; Nogusa, H. Org. Process Res. Dev. 2009, 13, 1413. 85

Veedhi, S.; Sawant, A. J. Therm. Anal. Calorim. 2013, 111, 1093.

86

Thayer, A. Chem. Eng. News, Sept. 5, 2005 83 (36), 40–58 (see p 58).

1041

19.2 PROCESS CHEMISTRY

TABLE 19.2

Metal Residue Levels in Drug Products Concentration (ppm)

Metal

Oral

Parenteral

Pt, Pd, Ir,

5

0.5

Mo, V, Ni, Cr

10

1

Cu, Mn

15

1.5

Zn, Fe

20

Rh, Ru, Os

2 1

Note: All values calculated based on PDE, 10 g d and 60-g body weight. Reprinted with permission from Thayer, A. Chem. Eng. News, Sept. 5, 2005, 83 (36), 40–58 (see p. 58). Copyright © 2005 American Chemical Society.

If a metal, metal complex, or metal catalyst is used in a reaction, it must be removed from the final product. However, removal of metals to the required levels can be problematic. In principle, the metal(s) can be removed using an adsorbent if there is a filtration step, but it may be necessary to switch solvents, which can be a problem.87 Palladium catalysts are widely used (see Sections 18.3 and 18.3.4), for example, but when the product is to be purified, the state of the Pd is not easily predicted. Therefore one method may not remove all of the metal. Palladium, as well as other metals, are expensive, so recovery rather than removal is often desirable if not preferred. If the metal remains bound, but can be removed during isolation of the product, the use of heterogeneous or immobilized catalysts bound to supports may reduce metal contamination.86 In practice, several methods are used to remove metals from a pharmaceutical product.88 Metal removal strategies include extraction, treatment of a metal-rich organic solution with aq acid or base, or extraction of the metal as a complex from the aqueous-rich phase into the organic-rich phase, or vice versa.89 Using Pd as an example, acid extraction of Pd(II) is effective for Pd that is leached from a support (e.g., carbon), as in catalytic hydrogenation using Pd/C (Section 7.10.1). Extraction or precipitation with a specialized reagent may be necessary if Pd is both leached and bound by the product. For Pd/C particles, from catalytic hydrogenation (Section 7.10) or use in Heck-type reactions (Section 18.4.1), particles may pass through the filter, which may be due to particle attrition or a poor filtration technique or an improper filter.90 O

S RHN

H

SR

P

OH

H

H

Dithiocarbamate

S

O P

O-R

H

H

Phosphinic acid (hypophosphorus acid)

Phosphinate ester

SH

Dithiophosphinic acid

S

O O

RO

P H

SR

Xanthate

Calix[4]arene, n = 1 Calix[6]arene, n = 3 Calix[8]arene, n = 5

OH

OH HO OH

O

O O O Crown ether

n

87

Prasad, K.; Garrett, C. E. Adv. Synth. Catal. 2004, 346, 889.

88

Larsen, R. D. Ed. Organometallics in Process Chemistry; Springer: Berlin, 2004, p 264.

89

Larsen, R. D. Ed. Organometallics in Process Chemistry; Springer: Berlin, 2004, pp 266–267.

90

Larsen, R. D. Ed. Organometallics in Process Chemistry; Springer: Berlin, 2004, p 265.

O

CH3

1042

19. COMBINATORIAL AND PROCESS CHEMISTRY

If extraction is required to remove the metal, there are many putative extraction reagents. These include amines, quaternary ammonium salts, calixarenes, and carboxylic acids (e.g., citric, lactic, or tartaric acid). In addition, crowns and lariats, hydroxyaromatics [e.g., 1-(2-pyridylazo)-2-napthol, known as PAN], hydroxyamine, oximes, phosphorus compounds [phosphine oxides, phosphonium salts, tributyl- or triphenylphosphine], sulfur compounds [dioctyl sulfide, dithiocarbamates (Aquamet), thiourea, cysteine, xanthates, CYANEX 30, which is bis(2,3,3-trimethylphenyl) dithiophosphinic acid], are used.91 If a crown ether is used to remove a residual metal, each crown ether binds different ions, depending on the size of the cavity. For example, 22 binds Li+, but not K+, while 23 binds K+, but not Li+. Similarly, 24 binds Hg2+, but not Cd2+ or Zn2+, and Sr2+, but not Ca2+.91 18-Crown-5 binds alkali and ammonium cations >1000 times weaker than 18-crown-6, presumably because the larger 18-crown-6 cavity involves more hydrogen bonds. Certain derivatives of 14-crown-4 and 12-crown-3 show very high selectivity for Li+ compared to the other alkali metal ions. O O

O

O

O

O

O

O O

O

O

O

O 22

O

O O

23

24

If crystallization is used as the purification method, the metal species should have a much larger solubility than the metal-free product. If the metal cocrystalizes, adsorbs on a surface or has poor solubility, crystallization is an ineffective method. Addition of a stabilizer (e.g., a phosphine) may enhance solubility characteristics of the metal.91 There are other potential problems (e.g., precipitation of the metal as fine particles or an amorphous solid, or the metal may simply precipitate rather than crystallize). In one example, “a zinc chelate was precipitated from an ethyl acetate product solution by adding aq potassium carbonate, which generated zinc carbonate, which precipitated from the solution.”91 Note that decantation (also called siphoning) can be used in some cases, in lieu of filtration, but it takes time for metal particles to settle below the suction inlet (the siphon). Adsorption is another method used to remove metals. Adsorbents include activated carbon, functionalized polymer resins, alumina, zeolites, and clays.92 An example is the use of SMOPEX isothiouronium fibers to reduce homogenous Pd catalysts levels from 395 to 3 ppm. Using SiliCycle thiol silica to reduce Pd(II) levels from 1000 to 2500 ppm) and also Fe (>250 ppm). After exploring several modifications, a revised procedure removed Pd after the Suzuki-Miyaura coupling. The reaction mixture was first diluted with

91

Larsen, R. D., Ed. Organometallics in Process Chemistry; Springer: Berlin, 2004, pp 268–270.

92

Larsen, R. D., Ed. Organometallics in Process Chemistry; Springer: Berlin, 2004, pp 270–274.

93

Bullock, K. M.; Mitchell, M. B.; Toczko, J. F. Org. Process Res. Dev. 2008, 12, 896.

1043

19.2 PROCESS CHEMISTRY

toluene-heptane, followed by addition of a 20% solution of sodium bisulfite (NaHSO3). This slurry was heated to 60°C for 1 h, and the precipitate removed by filtration, allowing 25 to be isolated via crystallization. Scale-up to 20 L equipment gave an 82% yield of 25 in >99.8% purity (