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