Content: Quantum chemistry throughout the periodic table / Dennis R. Salahub and Michael C. Zerner -- Optimizations of t
535 50 7MB
English Pages 411 Year 1989
ACS SYMPOSIUM SERIES 394
The Challenge of d and f Electrons Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.fw001
Theory and Computation Dennis R. Salahub, EDITOR Universitéde Montréal
Michael C. Zerner, EDITOR University of Florida
Developed from a symposium sponsored by the Divisions of Inorganic Chemistry and of Physical Chemistry of the American Chemical Society and the Division of Physical and Theoretical Chemistry of the Canadian Society for Chemistry at the Third Chemical Congress of North America (195th National Meeting of the American Chemical Society), Toronto, Ontario, Canada, June 5-11, 1988
American Chemical Society, Washington, DC 1989
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.fw001
Library of Congress Cataloging-in-Publication Data The Challenge of d and f electrons. (ACS symposium series, ISSN 0065-6156; 394) "Developedfroma symposium sponsored by the Divisions of Inorganic Chemistry and of Physical Chemistry of the American Chemical Society and the Division of Physical and Theoretical Chemistry of the Canadian Society for Chemistry at the Third Chemical Congress of North America (195th National Meeting of the American Chemical Society), Toronto, Ontario, Canada, June 5-11, 1988." Includes bibliographies and indexes. 1. Molecular orbitals—Congresses. 2. Quantum chemistry—Congresses. I. Salahub, Dennis R., 1946- . II. Zerner, Michael C. III. American Chemical Society. Division of Inorganic Chemistry. IV. American Chemical Society. Division of Physical Chemistry. V. Canadian Society for Chemistry. Division of Physical and Theoretical Chemistry. VI. Chemical Congress of North America (3rd: 1988: Toronto, Ont.) VII. American Chemical Society. Meeting (195th: 1988: Toronto, Ont.) VIII. Series. GD461.C37 1989 541.2'2 89-6926 ISBN 0-8412-1628-2 CIP
Copyright ©1989 American Chemical Society All Rights Reserved. The appearance of the code at the bottom of the first page of each chapter in this volume indicates the copyright owner's consent that reprographic copies of the chapter may be made for personal or internal use or for the personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per-copy fee through the Copyright Clearance Center, Inc., 27 Congress Street, Salem, MA 01970, for copying beyond mat permitted by Sections 107 or 108 or the U.S. Copyright Law. This consent does not extend to copying or transmission by any means-graphic or electronic—for any other purpose, such as for general distribution, for advertising or promotional purposes, for creating a new collective work, for resale, or for information storage and retrieval systems. The copying fee for each chapter is indicated in the code at the bottom of thefirstpage of the chapter. The citation of trade names and/or names of manufacturers in this publication is not to be construed as an endorsement or as approval by ACS of the commercial products or services referenced herein; nor should the mere reference herein to any drawing, specification, chemical process, or other data be regarded as a license or as a conveyance of anyrightor permission to the holder, reader, or any other person or corporation, to manufacture, reproduce, use, or sell any patented invention or copyrighted work that may in any way be related thereto. Registered names, trademarks, etc., used in this publication, even without specific indication thereof, are not to be considered unprotectedbylaw. PRINTED IN THE UNITED STATES OF AMERICA
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
ACS Symposium Series M. Joan Comstock, Series Editor
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.fw001
1989 ACS Books Advisory Board Paul S. Anderson
Mary A. Kaiser
Merck Sharp & Dohme Research Laboratories
E. I. du Pont de Nemours and Company
Alexis T. Bell
Michael R. Ladisch
University of California—Berkeley
Harvey W. Blanch University of California—Berkeley
Malcolm H. Chisholm Indiana University
Purdue University
John L. Massingill Dow Chemical Company
Daniel M. Quinn University of Iowa
James C. Randall Alan Elzerman
Exxon Chemical Company
Clemson University
Elsa Reichmanis John W. Finley Nabisco Brands, Inc.
AT&T Bell Laboratories
C. M. Roland
Natalie Foster
U.S. Naval Research Laboratory
Lehigh University
Stephen A. Szabo
Marye Anne Fox The University of Texas—Austin
Conoco Inc.
Wendy A. Warr Imperial Chemical Industries
G. Wayne Ivie U.S. Department of Agriculture, Agricultural Research Service
Robert A. Weiss University of Connecticut
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.fw001
Foreword The ACS S Y M P O S I U M S E R I E S was founded in 1974 to provide a medium for publishing symposia quickly in book form. The format of the Series parallels that of the continuing A D V A N C E S IN C H E M I S T R Y S E R I E S except that, in order to save time, the papers are not typeset but are reproduced as they are submitted by the authors in camera-ready form. Papers are reviewed under the supervision of the Editors with the assistance of the Series Advisory Board and are selected to maintain the integrity of the symposia; however, verbatim reproductions of previously pub lished papers are not accepted. Both reviews and reports of research are acceptable, because symposia may embrace both types of presentation.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.pr001
Preface THE G O A L O F T H E O R E T I C A L C H E M I S T R Y is to predict chemistry, either with simple mathematical expressions or with the more complicated expressions that researchers must solve on a computer, rather than by working with test tube and beaker in the lab. Theorists have long dreamed of synthesizing and characterizing the properties of conceptualized materials and their stabilities and reactivities without early recourse to bench experiments. Computer experiments are already of considerable aid to the experimentalist in prescreening compounds of desired character. The day is fast approaching when organic chemists will routinely use computational procedures to examine targeted molecules for specific properties before attempting synthesis and characterization of these systems. Quantum chemical experiments already aid our understanding of chemical processes, and such computations are often included with experimental results in publications. The impact of such work is continually growing. Entire industries are being transformed at the most fundamental level, perhaps most strikingly so i n the pharmaceutical industry. Progress has been somewhat slower in the development of theoretical tools for transition metal systems. The localized nature of d and f electrons, for example, often not only makes molecular orbital calculations difficult but also makes the utility of such calculations uncertain. In addition, the chemistry of these systems requires consideration of large molecules, clusters, surfaces, and bulk systems. Phenomena as diverse as medicine, catalysis, and high-temperature superconductivity are complex, and they require the most modern techniques for their accurate study. Methods that are of proven value for the chemistry of hydrogen and the main group elements, methods that predict molecular structure with an accuracy rivaling experiment, are proving inadequate, even qualitatively, when applied to the more challenging metal-containing systems. The symposium on which this book is based presented recent advances in the theory and computation of systems containing d and f electrons and applications to a number of the complex systems that have recently been examined using these techniques. Leading experimentalists presented work that would greatly benefit from advanced theory. The latest developments in molecular orbital theories, in correlation ix In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.pr001
calculations, and in local and nonlocal spin density theories were discussed, as well as the effects that calculations using these techniques have had on our way of thinking about these systems. The foundations of future scientific, technological, and industrial revolutions are currently being laid in the laboratories and computer rooms dedicated to understanding d and f electrons. This book contains contributions from most of the leading scientists in this area. It presents a snapshot of the state of the art as it existed in June 1988. We take this opportunity to thank the sponsors of this symposium again, whose support helped make this meeting the dynamic forum it was. Financial support came from Eastman Kodak Company of Rochester, NY; IBM Corporation of Yorktown Heights, NY; Multiflow Computer Company of Branford, CT; and the Petroleum Research Fund of the American Chemical Society. We also take this opportunity to thank Tom Ellis in Montreal and Susanne Gaddy in Gainesville for their cheerful editorial assistance. D E N N I S R.
SALAHUB
Universitéde Montréal Montréal, Québec H3C 3J7, Canada MICHAEL
C.
ZERNER
University of Florida Gainesville, F L 32611 December 1, 1988
x In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Chapter 1 Quantum
Chemistry
Throughout
the
Periodic
Table 1
2
Dennis R. Salahub and Michael C. Zerner Département de Chimie, Universitéde Montréal, C.P. 6128, Succursale A, Montréal, Québec H3C 3J7, Canada Quantum Theory Project, Williamson Hall, University of Florida, Gainesville, FL 32611
1
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch001
2
An overview is presented of the state-of-the-art for quantum chemical calculations for d- and f- electron systems. The present role and the potential of ab initio, density functional and semi-empirical methods are discussed with reference to contemporary developments in related experimental disciplines. Progress towards a true computational chemistry including the transition metals, lanthanides, and actinides is outlined with emphasis both on achievements and on the remaining barriers. Imagine a chemist, a modern-day Rip Van Winkle, awake and refreshed following a twenty-year sleep and finding himself at the instrument exhibit at an ACS national meeting. Poor fellow! His search for rotating evaporators and simple bench-top IR's will not be an easy one. If there are any, they are lost in a sea of television screens; a dazzling display of rotating, vibrating, pulsating, dancing, reacting, technicolor molecules. A few well placed questions and Rip learns that, while he was sleeping the sleep of the just, there was quite a revolution going on, that now chemists and, it seems, in large numbers are using computer programs to simulate the molecules they will perhaps eventually make react. A bit more courage, and a few more questions and he learns that byand-large the molecules dancing on the screen are organic - carbon, hydrogen, nitrogen, oxygen, with an odd halogen, or sulfur or phosphorus. Finding the colorful transition metals that he liked so much before he went to sleep turns out to be another difficult task; and when he finally finds a demonstrator to explain how they are handled, after much hemming and hawing, he learns that the methodological basis for the computations is not the same as it is for the lighter, s-p, elements. The organic computational chemistry that so dazzled our allegorical eyes-wide-open chemist is based on the combined efforts of a large number of theoretical and experimental chemists over the last few decades. This combined experience has resulted, only 0097-6156/89/0394-0001$06.00A) c 1989 American Chemical Society
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch001
2
THE CHALLENGE
OF d AND f ELECTRONS
r e c e n t l y , i n a s e t of t o o l s t h a t can p r o v i d e the e n t i r e c h e m i s t r y community, s p e c i a l i s t and n o n - s p e c i a l i s t a l i k e , w i t h a manual and a s e t of "specs", i n much the same way as any o t h e r i n s t r u m e n t . The u s e r , once he has read, or has been p r o p e r l y i n f o r m e d , t h a t the i n s t r u m e n t i s a p p r o p r i a t e f o r h i s problem (an o r g a n i c system w i t h N atoms and a r e q u i r e d accuracy of x k c a l / m o l e ) can go ahead and use t h e s e enormously complex programs w i t h the g r e a t e s t of ease to h e l p him i n h i s work. That the same i s not t r u e i n g e n e r a l f o r problems i n v o l v i n g t r a n s i t i o n metals or l a n t h a n i d e s or a c t i n i d e s can be a t t r i b u t e d above a l l e l s e t o the d i f f i c u l t y of h a n d l i n g e l e c t r o n c o r r e l a t i o n f o r these systems. The d and f e l e c t r o n s are c o n f i n e d t o a s m a l l volume and t h e r e are many of them, p a r t i c u l a r l y i f m e t a l - m e t a l bonds are i n c l u d e d . T h i s c o r r e l a t i o n problem w i l l be a t the core of t h i s paper and indeed of the whole volume. But t h i s i s not the o n l y problem when i t comes to comparing d and f o r b i t a l systems w i t h systems t h a t do not c o n t a i n these t r o u b l e makers. Transition metal systems are most o f t e n o p e n - s h e l l systems, unlike their organic cousins. The t r e a t m e n t of o p e n - s h e l l systems i s more d i f f i c u l t both a t the H a r t r e e - F o c k l e v e l , and a t the p o s t H a r t r e e Fock l e v e l . Of importance i n t r a n s i t i o n metal complexes i s a comparison of the r e l a t i v e e n e r g i e s of complexes of d i f f e r i n g s p i n m u l t i p l i c i t i e s , and t h i s comparison i s made very d i f f i c u l t . The Hartree-Fock procedure greatly favors states of higher m u l t i p l i c i t y , and t h i s b i a s i s , a g a i n , o n l y c o r r e c t e d by a h i g h l e v e l c o r r e l a t e d t h e o r y (see DAVIDSON, f o r e x a m p l e ) . In a d d i t i o n , the g r e a t wealth of s t a t e s t h a t l i e near i n energy f o r t r a n s i t i o n metal systems o f t e n make the SCF s t e p r e q u i r e d i n most t r e a t m e n t s much more troublesome. F o r o r g a n i c systems, SCF i s u s u a l l y (but not a l w a y s ! ) not a p r o b l e m . There are i n g e n e r a l t h r e e common ways used today t o go about examining the e l e c t r o n i c s t r u c t u r e of t r a n s i t i o n metal systems. A l l of these are concerned i n one way o r another i n s o l v i n g (or getting around) the c o r r e l a t i o n problem and we w i l l g i v e an o v e r v i e w of a l l t h r e e , u n d e r l i n i n g t h e i r r e s p e c t i v e s t r e n g t h s and weaknesses: "standard" ab i n i t i o t e c h n i q u e s (the s u r e s t and o f t e n the b e s t way to go f o r cases where the c o m p u t a t i o n a l e f f o r t does not render them i m p r a c t i c a l or i m p o s s i b l e ) , D e n s i t y F u n c t i o n a l methods where the c o r r e l a t i o n i s handled through an e l e c t r o n gas model ( s t i l l an "ab i n i t i o " framework but p r a c t i c a b l e f o r much l a r g e r systems) and s e m i - e m p i r i c a l where " n a t u r e ' s c o r r e l a t i o n " , namely e x p e r i m e n t a l d a t a , are used t o a d j u s t p a r a m e t e r s , r e s u l t i n g i n very r a p i d methods t h a t , when used j u d i c i o u s l y , can p r o v i d e a c c u r a t e r e s u l t s f o r very l a r g e systems. I n t h i s b r i e f overview we can only touch on very few a s p e c t s of a v a s t and growing f i e l d . Our o n l y g o a l here i s t o p r o v i d e some s t r u c t u r e so t h a t newcomers, i n p a r t i c u l a r , may more r a p i d l y put the v a r i o u s q u e s t i o n s and the v a r i o u s techniques into proper p e r s p e c t i v e , and o b t a i n a f i r s t i m p r e s s i o n of the p r e s e n t s t a t e of the f i e l d and o f where i t i s g o i n g , as we see i t . Taken a l o n g w i t h the o t h e r chapters of the book ( r e f e r r e d to by the name of an a u t h o r i n c a p i t a l l e t t e r s ) and the o t h e r r e f e r e n c e s c i t e d , we b e l i e v e t h a t t h i s i m p r e s s i o n s h o u l d be r e a s o n a b l y u n b i a s e d .
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch001
1.
SALAHUB & ZERNER
Quantum Chemistry Throughout the Periodic Table
Before s t a r t i n g the m e t h o d o l o g i c a l o v e r v i e w , i t i s w o r t h w h i l e t o remind o u r s e l v e s b r i e f l y of the e x p e r i m e n t a l c o n t e x t . The presence of d and f e l e c t r o n s , whose wave f u n c t i o n s can be e i t h e r c o r e - l i k e , e s s e n t i a l l y a t o m i c , o r more d i f f u s e , and thus t r u e valence like, or a n y t h i n g i n between, does c o n f e r special, sometimes u n i q u e , p r o p e r t i e s . The f a s c i n a t i o n and the u t i l i t y of the t a r g e t systems f o r many of the c h a p t e r s of t h i s book are owed, i n the f i n a l a n a l y s i s , t o the range of b e h a v i o r t h a t e l e c t r o n s i n these wave f u n c t i o n s demonstrate. The f o u r " e x p e r i m e n t a l " c h a p t e r s of the book treat subjects ranging from ESR and other s p e c t r o s c o p i e s of c l u s t e r s i n m a t r i c e s (WELTNER and VAN Z E E ) , t o the thermodynamics and phase t r a n s i t i o n s of heavy fermions (OTT and F I S K ) , t o e l e c t r o n d e n s i t y maps (COPPENS), t o energy c o n v e r s i o n i n p h o t o s y n t h e s i s (SMITH and GRAY). These r e p r e s e n t o n l y a very s m a l l s a m p l i n g of the many c h a l l e n g e s b e i n g o f f e r e d to quantum c h e m i s t r y by r e c e n t e x p e r i m e n t a l advances. Other c h a p t e r s w i l l mention aspects of inorganic synthesis, of reaction mechanisms, of homogeneous, heterogeneous, and enzymatic c a t a l y s i s , of c l u s t e r beams, of s u r f a c e s c i e n c e , of s u p e r c o n d u c t i v i t y and magnetism, of NMR and Mossbaur s p e c t r o s c o p y , of b i o m o l e c u l e s and much, much more. In f a c t t h i s f i e l d p r o v i d e s a v e r i t a b l e smorgasbord of e x p e r i m e n t a l data waiting for interpretation and guidance from quantum chemistry. A good smorgasbord is stocked with individual d e l i c a c i e s but a l s o has a theme - ours has the d and f e l e c t r o n s t o p r o v i d e c o n t i n u i t y , coherence and, i f you w i l l , the f l a v o r t h a t each of the authors of t h i s volume so w e l l e n j o y . Ab-initio Calculations The term " a b - i n i t i o " i s o f t e n taken to i n c l u d e those methods which, g i v e n an i n i t i a l c h o i c e f o r the g e n e r a l form of the N - e l e c t r o n wave function, ( e . g . one or many d e t e r m i n a n t s ) a t t e m p t to s o l v e the Schrodinger Equation H¥ = E¥ H = K i n e t i c energy o p e r a t o r + P o t e n t i a l energy
operator
E= m o l e c u l a r energy ¥=Y(r1,r2,....rN) w i t h o u t i n t r o d u c i n g any e m p i r i c a l p a r a m e t e r s . T h i s does not mean t h a t no a p p r o x i m a t i o n s are made (indeed, the whole a r t i s i n the a p p r o x i m a t i o n s ) but only t h a t r e c o u r s e i s not t a k e n to experiment o t h e r than, p e r h a p s , a - p o s t e r i o r i , t h a t there r e s u l t s agreement w i t h experiment. Table I g i v e s a b r i e f summary of some of the main c l a s s e s of ab i n i t i o methods w i t h t h e i r c h a r a c t e r i z i n g f e a t u r e s . A good e n t r y i n t o the d e t a i l s of these methods i s the book by Szabo and O s t l u n d (J_). Recent i s s u e s of the J o u r n a l of C h e m i c a l P h y s i c s , C h e m i c a l P h y s i c s L e t t e r s , and the I n t e r n a t i o n a l J o u r n a l of Quantum C h e m i s t r y c o u l d a l s o be c o n s u l t e d . Examples of a p p l i c a t i o n s w i l l be found i n the chapters of WILLIAMSON and HALL (GVB,CASSCF), DEDIEU and BRANCHADELL (CASSCF,CI), DAVIDSON ( H F , M P P T , C I ) , NOVARRO
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
3
THE CHALLENGE OF d AND f ELECTRONS
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch001
Table I .
Some C l a s s e s of Ab I n i t i o Methods 4
Hartree-Fock HF (or SCF)
one ( s p i n adapted) determinant ¥ =U are e v a l u a t e d as the i n t e g r a l s h(i,j)
=
g = = - = / / d (1 ) d ( 2 ) ^ ( 1 ) ^ ( 2 ) T
T
[1-P(12)]
^(1)^(2)
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
1.
SALAHUB & ZERNER
Quantum Chemistry Throughout the Periodic Table
t h e n the method i s g e n e r a l l y c l a s s i f i e d as " a b - i n i t i o " . It is important to stress here that a l l successful semi-empirical e l e c t r o n i c s t r u c t u r e H a m i l t o n i a n s have t h i s same form, and abi n i t i o and s e m i - e m p i r i c a l THEORIES are the same. The Table i n the a b - i n i t i o section that describes theories, i s also appropriate here: SCF, C I , MCSCF, CPMET, e t c . Semi-empirical theories generally parameterize h and g. U s u a l l y e n t i r e c l a s s e s of i n t e g r a l s i n g are n e g l e c t e d , and those t h a t remain, a l o n g w i t h h , a r e p a r a m e t e r i z e d t o compensate f o r t h i s omission. The p a r a m e t e r i z a t i o n may be based on model a b - i n i t i o c a l c u l a t i o n s , o r d i r e c t l y on e x p e r i m e n t , and o f t e n the proponents of these two d i f f e r e n t p o i n t s of view d o n ' t t a l k t o one a n o t h e r ! A common g o a l , though, i s t h a t of r e d u c i n g the N g i n t e g r a l s , where N i s the s i z e of the b a s i s s e t , t o one of N o r N • F o r t r a n s i t i o n m e t a l systems, i t i s p r o b a b l y f a i r t o l i m i t our a t t e n t i o n t o t h r e e methods, o r r a t h e r t h r e e c l a s s e s of methods s i n c e a number of v a r i a n t s e x i s t f o r e a c h . These are summarized i n Table I V . 4
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch001
3
Table I V .
Some S e m i - E m p i r i c a l Methods Used f o r d and f E l e c t r o n Systems
Extended H u c k e l EH REX(Re1. EH)
h ( i , i ) r e l a t e d to i o n i z a t i o n potentials, h ( i , j ) empirical param. s e t p r o p o r t i o n a l t o overlap (S). S included i n secular Eq. (H-E(i)S)c(i)=0
p i and sigma elec. large systems not self-consist, relativistic vers.
I t e r a t i v e EH IEH, SCCEH ITEREX
As above, i t e r a t i o n t o charge c o n s i s t e n c y u s u a l l y based on e x t r a p . between I . P . ' s o f atomic i o n s .
As above•
Fenske-Hall
As above: h ( i , j ) w i t h terms p r o p o r t i o n a l t o S as w e l l as k i n e t i c energy t e r m s .
Highly e f f e c t i v e f o r complexes; iterative.
N e g l e c t of Differential Overlap. CNDO, INDO NDDO
one c e n t e r p a r t of h ( i , i ) from I P ' s . N u c l e a r a t t r a c t , i n c l u d e d . CNDO c o n t a i n s a l l < i , j / i , j > ; INDO i n add. a l l one-center < i , j / k , l > ; NDDO i n add. a l l r e m a i n i n g t w o - c e n t e r Coulomb t y p e . h ( i , j ) usually proportional to S. Two-elec. i n t s often parameterized.
Large complexes C I , MBBT, MCSCF versions. Useful f o r geometry e s t i m a t e and for e l e c t r o n i c spectroscopy.
Relativistic version.
The Extended Huckel method, as i t i s known today, was f i r s t implemented by Lohr and Lipscomb i n s t u d i e s of boron h y d r i d e s (23)
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
13
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch001
14
THE CHALLENGE OF d AND f ELECTRONS
and l a t e r extended t o o r g a n i c systems by Hoffmann ( 2 4 ) . I t has been enormously f r u i t f u l , p r o v i d i n g a framework f o r the a n a l y s i s o f o r b i t a l i n t e r a c t i o n s which pervades modern o r g a n i c c h e m i s t r y . The a p p l i c a t i o n of the p r i n c i p l e s b e h i n d the Extended H u c k e l method, however, actually stem from a p p l i c a t i o n on t r a n s i t i o n metal complexes, initially by W o l f s b e r g and Helmholz, and then by B a l l h a u s e n , and B a l l h a u s e n and Gray ( 2 5 ) . I t e r a t i v e schemes f o r t r a n s i t i o n - m e t a l s compounds were d e v i s e d by B a l l h a u s e n and Gray and by Z e r n e r and Gouterman ( 2 6 ) . More r e c e n t l y , p r i m a r i l y through the work of the s c h o o l o f Hoffmann, i t has been adapted t o s o l i d s and surfaces and a g a i n i s p r o v i d i n g i n s i g h t t h a t o f t e n cannot be o b t a i n e d n e a r l y so r e a d i l y w i t h o t h e r more e l a b o r a t e techniques (27). Indeed, i t c o u l d reasonably be argued t h a t i f one had t o choose one t e c h n i q u e t o "do" c h e m i s t r y ( l u c k i l y one d o e s n ' t have t o make t h i s c h o i c e ! ) then EH, o r i t s i t e r a t i v e v e r s i o n s , w i t h i t s easy t o i n t e r p r e t r e s u l t s , would p r o b a b l y be the b e s t c h o i c e f o r a g e n e r a l overview of a wide v a r i e t y of systems. I t can p r o v i d e i n v a l u a b l e guidance f o r q u e s t i o n s t h a t depend on o r b i t a l symmetry, o r b i t a l e n e r g i e s and o r b i t a l o v e r l a p . A good d e a l of u n d e r s t a n d i n g can be based o n l y on these s i m p l e c o n c e p t s ! The F e n s k e - H a l l method, which i s i n the s p i r i t of the Extended H u c k e l Methods (28) but i n c l u d e s k i n e t i c energy terms as w e l l as much of the electrostatics, p u r p o r t s to g r e a t e r a c c u r a c y , but q u a l i t a t i v e l y y i e l d s results s i m i l a r to simple IEH. F o r more q u a n t i t a t i v e , and more d e t a i l e d treatments o f , f o r example, m u l t i p l e t s and o t h e r a s p e c t s of s p e c t r o s c o p y , one must i n c l u d e most of the proper e l e c t r o s t a t i c s . The methods d e s c r i b e d g e n e r i c a l l y as Z e r o - D i f f e r e n t i a l Ovelap (ZDO) types do t h i s i n a h i e r a r c h y of t e c h n i q u e s ; CNDO, Complete N e g l e c t of D i f f e r e n t i a l O v e r l a p (29), INDO, I n t e r m e d i a t e N e g l e c t of D i f f e r e n t i a l O v e r l a p ( 3 0 ) , and NDDO, N e g l e c t o f D i f f e r e n t i a l D i a t o m i c O v e r l a p (_31_), a l l of which are g e n e r a l i z a t i o n s of the e a r l i e r P a r i s e r - P a r r - P o p l e (PPP) method d e v i s e d f o r p i e l e c t r o n s o n l y ( 3 2 ) . A t present there is o n l y one g e n e r a l l y a v a i l a b l e program of t h i s nature t h a t includes t r a n s i t i o n metals, dubbed f o r un-acknowledged reasons ZINDO ( 3 3 ) , although other interesting v a r i a t i o n s are being developed ( 3 4 J . An a p p l i c a t i o n of the INDO method i s found i n the c h a p t e r by LOEW, i n which s p i n s t a t e s of l a r g e p o r p h i n a t o F e ( I I I ) complexes are examined, a task very d i f f i c u l t by a b - i n i t i o t h e o r i e s , as d i s c u s s e d i n the c h a p t e r by DAVIDSON, and one s t i l l a w a i t i n g a t h e o r y of m u l t i p l e t s t r u c t u r e i n DF methods. V a r i o u s CNDO and INDO schemes have a l s o been proposed f o r t r a n s i t i o n - m e t a l c l u s t e r s and f o r c h e m i s o r p t i o n on them. Although t h e r e have been some s u c c e s s e s , i t remains t r u e t h a t both the l e v e l of theory b e i n g used and the p a r a m e t e r i z a t i o n are both very much experimental. The s e m i - e m p i r i c a l methods c o n t i n u e t o y i e l d guidance to e x p e r i m e n t a l i s t s and t h e o r i s t a l i k e , i n a wide v a r i e t y of f i e l d s . But i n a sense the a p p l i c a t i o n s of these methods have a moving target. As a b - i n i t i o and d e n s i t y f u n c t i o n a l methods improve, becoming more r a p i d and a p p l i c a b l e t o a w i d e r range of q u e s t i o n s , those a p p r o p r i a t e f o r s e m i - e m p i r i c a l s t u d i e s become more and more complex. But t h e r e i s a l s o a fundamental q u e s t i o n h e r e . Any c a l c u l a t i o n t h a t i s performed on any system can be, and w i l l be,
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
1. salahub & zerner
Quantum Chemistry Throughout the Periodic Table
performed at greater accuracy. At what level of theory and computation may the problem at hand be considered to be solved? Extended Huckel, the simplest of these theories, is apparently enough to describe a great deal of the actual chemistry of a complex; however, we s t i l l do not have ANY theory capable of reliably reproducing, for example, the ESR spectroscopy of these systems. Clearly the computational method of choice will depend on the problem at hand, and this will continue to be so.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch001
Concluding Remarks It has been impossible, in these few pages,to capture even a small fraction of the advances and of the excitement surrounding the treatment of d- and f- electron systems in contemporary quantum chemistry. The term quantum chemistry has been used advisedly, as opposed to either theoretical chemistry, with its connotations of dynamics and "things other than electronic structure calculations", or computational chemistry with its connotations of black-box programs and user-friendly graphics software. Very little has been done in either of these directions for the transition metals and even less for the lanthanides and actinides. Indeed, the excitement in the field is largely due to the fact that there is so much left to discover, and to the fact that the tools to allow pioneering discoveries are now just becoming available. At the Toronto meeting we saw the first signs of progress towards both theoretical and computational chemistry for the d- and f- electron systems. Dynamics are definitely on the agenda for the next few years as are the development of convenient computer interfaces and the extraction of force fields and the like that will allow more and more experimentalists in inorganic chemistry to spend some of their time fruitfully at the computer keyboard. Our friend Rip had better not go to sleep for another twenty years! Literature Cited 1. 2. 3.
4. 5. 6. 7. 8. 9.
Szabo, A.; Ostlund, N. S. Modern Quantum Chemistry: Introduction to Advanced Electronic Structure Theory MacMillan, New York, 1982. Salahub, D. R. Adv. Chem. Phys. 1987, 69, 447. e.g. Huzinaga, S.; Andzelm,J.;Klobukowski,M.;RadzioAndzelm, E.; Sakai, Y.; Tatewaki, H. Gaussian Basis Sets for Molecular Calculations; Elsevier, Amsterdam, 1984 and references therein. e.g. Huzinaga, S.; Klobukowski, M.; Sakai, Y. J. Phys. Chem. 1984, 88, 4880 and references therein. Pyykko, P. Chem. Rev. 1988, 88, 1. Hohenberg, P.; Kohn, W. Phys. Rev. 1964, 136, B864. Kohn, W.; Sham, L. J . Phys. Rev. 1965, 140, A1133. Lundqvist, S.; March, N. H., Eds. Theory of the Inhomogeneous Electron Gas; Plenum, New York, 1983. Dahl, J. P.; Avery, J., Eds. Local Density Approximations in Quantum Chemistry and Solid State Physics; Plenum, New York, 1984.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
15
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch001
16
the challenge of d and f electrons
10. Dreizler, R. M.; da Providencia, J., Eds. Density Functional Methods in Physics; Plenum, New York, 1984. 11. Slater, J. C. Adv. Quantum Chem. 1972, 6, 1; The SelfConsistent Field for Molecules and Solids; Vol. 4, McGraw-Hill, New York, 1974. 12. Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200. 13. Perdew, J. P.; Zunger, A. Phys. Rev. B 1981, 23, 5048. 14. Ceperley, D. M.; Alder, B. J . Phys. Rev. Lett. 1980, 45, 566. 15. Langreth, D. C.; Mehl, M. J. Phys. Rev. B 1983, 28, 1809; erratum 1984, 29, 2310 16. Becke, A. D. J. Chem. Phys. 1986, 84, 4524. 17. DePristo, A. E.; Kress, J . D. J. Chem. Phys. 1987, 86, 1425. 18. Perdew, J. P.; Yue, W. Phys. Rev. B 1986, 33, 8800; Perdew, J. P. Phys. Rev. B 1986, 33, 8822. 19. Johnson, K. H. Adv. Quantum Chem. 1973, 7, 143. 20. Muller, J . E.; Jones, R. O.; Harris, J . J. Chem. Phys. 1983, 79, 1874 and references therein. 21. Delley, B.; Ellis, D. E.; Freeman, A. J.; Baerends, E. J.; Post, D. Phys. Rev. 1983, 27, 2132 and references therein. 22. Dunlap, B. I.; Connolly, J. W. D.; Sabin, J. R. J. Chem. Phys. 1979, 71, 3386, 4993. 23. Lohr, L. L.; Lipscomb, W. N. J. Chem. Phys. 1963, 38, 1604. 24. Hoffmann, R. J . Chem. Phys. 1964, 39, 1397: ibid. 40, 2047; 40, 2474; 40, 2480; 40, 2745. 25. Ballhausen, C. J; Gray, H. B. Inorg. Chem. 1962, 1, 111: ibid. Molecular Orbital Theory, Benjamin Press, New York, 1964. 26. Zerner, M.; Gouterman, M. Theoret. Chim. Acta. 1966, 4, 44. 27. Hoffmann, R. Rev. Mod. Phys. 1988, 60, 601. 28. Fenske, R.; Hall, M. B. Inorg. Chem. 1972, 11, 768. 29. Pople, J. A.; Santry, D. P.; Segal, G. A. J. Chem. Phys. 1965, 43, S129; Pople, J. A.; Segal, G. A. J. Chem. Phys. 1965, 43, S136; ibid. 1966, 44, 3289. 30. Pople, J. A; Beveridge, D. L.; Dobosh J. Chem. Phys. 1967, 47, 2026. 31. Pople, J. A.; Beveridge, D. L. Approximate Molecular Orbital Theory, McGraw Hill, New York, 1970. 32. Parr, R. G. Quantum Theory of Molecular Electronic Structure, Benjamin Press, New York, 1963. 33. Ridley, J. E.; Zerner, M. C. Theoret. Chim. Acta, 1973, 32, 111; Bacon, A.; Zerner, M. C. Theoret. Chim. Acta. 1979, 53, 21; Zerner, M. C.; Loew, G. H.; Kirchner, R. F.; MuellerWesterhoff, U. T. J. Am. Chem. Soc. 1980, 102, 589. 34. Lipinski, J. Intern. J. Quantum Chem. 1988, 34, 423. RECEIVED
March 2, 1989
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Chapter Optimizations
of
the
Ti(IV)
2
Geometry
of
Tetrahedral
Complexes
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
A Basis Set and Correlation Study of Tetrachlorotitanium and Trichloromethyltitanium Rodney L. Williamson and Michael B. Hall Department of Chemistry, Texas A&M University, College Station, TX 77843 Optimization of the geometry of TiCl and TiCl CH at the SCF level results in Ti-Cl bond lengths longer than the experimental values, even when d- and f-type polarization functions are added to the basis set. The bond lengths remain too long even as the Hartree-Fock limit is approached because the SCF level of theory over-estimates the noble-gas-like Cl···Cl repulsions, which hinder close Ti-Cl approach. The Ti-C-H angle of T i C l C H is calculated to be close to tetrahedral geometry with little flattening of the hydrogen atoms, which apparently was observed in the electron diffraction. These same calculations do predict the anomalously low methyl-rocking frequency for thetitaniumcomplex in agreement with the experimental IR. This low methyl rocking frequency is due to stabilization of the Ti-C bond during the rocking motion by low lying empty d-orbitals on Ti. The large positive geminal hydrogen coupling constant observed in the NMR experiment is due primarily to the σ-donor and π-acceptor character of the TiCl moiety and not to any flattening of the methyl group. 4
3
3
3
3
3
In the past few years, geometry optimizations of transition metal complexes have seen increased attention and improvement. Two minimal basis set studies (1-2) optimized metal-ligand bond lengths with errors of 0.05-0.25 A. The errors decreased for most ligands in studies with moderate basis sets (3-8) with the exception of metalcyclopentadienyl (Cp) bond lengths which showed errors ranging from 0.16-0.24 A. Further work (9-12) has shown that the error in calculating the metal-Cp bond lengths with very large basis sets is primarily correlation error. Two recent studies with 3-21G type basis sets have predicted the equilibrium geometries of some transition metal complexes with reasonable accuracy. In one of these studies (13), metal-carbonyl and metal-Cp distances were predicted to be 0.03 A and 0.15 A longer, respectively, than experimental values. It was shown that these long metal-Cp distances could be significantly reduced by including electron correlation. In the other study Q4), the optimized metal-carbonyl distances averaged 0097-6156/89/0394-0017$06.00/0 c 1989 American Chemical Society
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
18
THE CHALLENGE OF d AND f ELECTRONS
0.11 A longer than the experimental values. The optimized geometries for a large number of tetrahedral metal-halide complexes gave M - F bond lengths shorter than experimental values and M - C l bond lengths longer than experimental values. In order to show how different basis sets and electron correlation affect bond lengths and angles we report here the results of self-consistent-field (SCF) and generalized-valence-bond (GVB) geometry optimizations of two tetrahedral titanium(IV) chloride complexes, T i C l and TiCl3CH . For TiCl3CH we also did complete-active-space-self-consistent-field (CASSCF) geometry optimizations. We chose TiCLt because of its high symmetry, which greatly simplifies the calculation and interpretation, and we chose TiCl3CH3 as a second Ti(IV) complex because of its "unique" geometry. This geometry, which was reported by Berry et al. (15) from electron diffraction (ED), appeared to have a flattened methyl group due to three agostic hydrogens. A n agostic hydrogen is defined as a hydrogen atom covalently bonded simultaneously to both a carbon atom and to a transition metal atom (16). Agostic hydrogens have been reported for several titanium alkyl complexes (16-19) such as [TiCl3(Me2PCH2CH PMe )R] (R=Et,Me). In these complexes, the geometry of the alkyl ligand containing the agostic hydrogen is distorted from the geometry it would have if it were bonded only to an organic substrate. In the methyl case this distortion is observed as a rocking of the methyl group such that one hydrogen atom moves toward the metal atom while the other two move away from the metal. The X-ray crystal structure of the above methyl complex (17) shows T i - C - H angles of 70(2)°,105(4)°, and 117(3)°. A later neutron diffraction study (12) on the same complex showed Ti-C-H angles of 93.5(2)°, 118.4(2)° and 112.9(2)°. If the complex did not have an agostic hydrogen, one would expect the Ti-C-H angles to be close to 109.5°. An ab initio molecular orbital study (2Q) of the agostic hydrogen interaction in Ti(CH3)(PH3)Cl3 reported direct interaction between the C-H a-bond and an unoccupied T i d-orbital. This type of interaction involving a single hydrogen bent towards the metal center is typical of most complexes with an agostic hydrogen. However, Berry et al. (15) reported all three methyl hydrogens "flattened" toward the titanium atom, the first such complex with three agostic hydrogens. In addition to the ED results, they reported a large difference between the methyl rocking vibrational frequencies of TiCl3CH3 and GeCl3CH3 and explained this difference as resulting from the flattening of the methyl hydrogens. They also did C and H N M R studies and found the H,H coupling constant to have a large positive value, which they suggested was further evidence of hydrogen flattening. We optimized the geometry of TiCl3CH3 to determine if the calculations would predict the symmetrical flattening of the methyl hydrogens. The model complexes TiH3CH3 and GeH3CH3 were used to calculate vibrational frequencies to determine if the calculation would predict the large difference between the rocking frequencies of the Ge and Ti complexes. Some of these results have been reported in a preliminary communication (21). And finally, we estimated the H,H coupling constant for the Is orbitals of two methyl hydrogens on TiCbCH3, CH4, and CH3CI using the method of Pople and Santry (22). Although this method is known to have serious problems in predicting the sign and absolute magnitude of the H,H coupling constant (22), it does correctly predict the direction and size of the change of the coupling constant resulting from changes in the nature of the methyl substituent or changes in geometry of the compound.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
4
2
3
3
2
1 3
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
l
2. WILLIAMSON & HALL
Geometric Optimizations ofTetrahedral Complexes 19
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
Computational Details We optimized the geometries of the Ti complexes using three different titanium basis sets. The first titanium basis set (I) is a (432-421-31) used in previous calculations (12), the second titanium basis set (H) is a Huzinaga (24) (5333-53-5) modified to a (533211-5211-3111) by splitting off the most diffuse s, p, and d functions and adding s, p, and d functions with exponents 1/3 the values of the split-off functions, and the third titanium basis set (III) is the titanium basis set II with an additional ftype polarization function (£=0.55) (25). For the main group atoms, basis set I is Cl(3321-321), C(321-21), H(21), basis set II is Cl(5321-521), C(421-31), H(31), basis set HI is Cl(531111-4211), C(721-41), H(31), basis set IV is Cl(5321-521-1), C(421-31-l), H(31-l), basis V is Cl(533-53), basis V I is Cl(53111-5111), basis VII is Cl(533-53-l), basis VIII is Cl(533-53-ll), and basis IX is Cl(5321-521-ll). Basis sets I, II, and VI-IX are Huzinaga (24) basis sets which were all modified (expect for basis set V) by singly or doubly splitting the most diffuse s and p functions. Basis sets IV, and VII-IX were further modified by adding polarization functions. Basis set IH is an unmodified Dunning-Hay-Huzinaga (26) basis set. The basis sets used for the geometry optimizations and force constant calculations of TiH3CH3 and GeH3CH3 are 3-21G type basis sets used in previous geometry optimization (13). The H,H coupling constant calculations were done using fully contracted titanium and chlorine basis sets I and carbon and hydrogen basis sets EL The geometry of TiCl3CH3 and TiCU were optimized in staggered C3 and Td symmetry, respectively. The G V B calculations (27) involve perfect-pairing for all seven sigma bonds for TiC^CTb and for the four Ti-Cl sigma bonds for TiCU. The first of two CAS SCF calculations on TiCbCH3 contains eight electrons in the eight orbitals (8/8) made up of four a-bonding and four a-antibonding T i - C and C - H orbitals. The second contains eight electrons in eleven orbitals (8/11) made up of the eight orbitals in the 8/8 calculation plus two T i - C rc-bonding orbitals and one additional Ti-C a-bonding orbital. A l l of the calculations were done with G A M E S S (Generalized Atomic and Molecular Electronic Structure Systems) except for the TiCU calculations with f-type polarization functions which were done with Q U E S T (QUantum Electronic STructure). These programs were run on a C R A Y X - M P at C R A Y Research in Mendota Heights, an FPS-264 at the Cornell National Supercomputer Facility, an I B M 3090/200 at Texas A & M University, and the Department of Chemistry's V A X 11/780 andFPS-164. V
Results and Discussion TiCl^ Geometry Optimization. The optimized Ti-Cl bond lengths (see Table I) for TiCU with a variety of basis sets are all longer than the experimental value and differ from that by 0.016 to 0.051 A. Splitting the Cl(533-53) basis set, which allows the orbitals freedom to expand or contract, only decreased the Ti-Cl bond length an average 0.005 A. However, when we add d-type polarization functions to the chlorine basis set the Ti-Cl bond distance decreases an average 0.026 A. The effect of adding polarization functions is also seen in deformation density maps of TiCU- The deformation density of TiCU without d-functions on CI (Figure la) shows a buildup of density in the Ti-Cl bonding region and the CI lone-pair regions. When a polarization function is added to CI (Figure lb) density in the bonding region increases and density in the chlorine lone-pair region decreases. The
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
ELECTRONS
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
THE CHALLENGE OF d AND f
Figure 1. Electron deformation density plots of TiCU in Cl-Ti-Cl plane: a) deformation density without d-functions b) deformation density with d-functions. Contours are geometric beginning at ±0.001 e~au- and incremented by doubling the previous contour value. 3
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
2. WILLIAMSON & HALL
Geometric Optimizations of Tetrahedral Complexes 21
Table I. Ti-Cl Bond Distances for TiCU
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
Tjtagiurn Pasis Sets Method I Chlorine Basis 1. SCF Cl(3321-321) (I) 2.197 2. SCF Cl(533-53) (V) 2.203 3. SCF 2.201 Cl(5321-521) (II) 4. SCF C1(531U-5111)(VI) 2.200 5. SCF 0(531111-4211) (HI) 2.202 6. SCF Cl(533-53-l) (VH) 7. SCF Cl(5321-521-1) (IV) 8. SCF Cl(533-53-ll)(Vni) 9. SCF C1(5321-521-11)(IX) GVB 10. Cl(5321-521) (II) Experimental 2.170(2)* Reference 28, the error is reported at a confidence level of 2.5
II 2.209 2.221 2.213 2.214 2.219 2.186 2.187 2.189 2.191 2.234
in
2.196
2.181
a.
origin of the shift in deformation density from the lone-pair region on CI to the Ti-Cl bonding region is also seen in orbital plots (Figures 2a-d) of both the Ti-Cl a-bonding andrc-bondingmolecular orbitals (MO). The value in the Ti-Cl bonding region of the ai M O with d-functions (Figure 2b) is greater than the value in the bonding region of the ai M O without d-functions (Figure 2a). The CI t ic-orbitals also show a similar increase in the Ti-Cl bonding region (Figures 2e,f). Molecular orbital plots of the other orbitals (Figures 2c-d,g-j) do not show any difference between the plots with dfunctions and the plots without d-functions. As we improve the TiCU wavefunction by adding f-type polarization functions to the titanium basis set the Ti-Cl bond distance shortens further. When a chlorine basis set without d-functions is used, the addition of an f-function on titanium shortens the Ti-Cl bond 0.017 A. However, when a chlorine basis set with d-functions is used, the Ti-Cl bond only shortens 0.006 A to 2.181 A. Although this Ti-Cl distance is the shortest of all the optimized geometry calculations, it is still 0.011 A longer than the experiment As the wavefunction approaches the Hartree-Fock limit one would expect the TiC l bond distance to be shorter than the experiment because of the lack of bond-pair correlation. The bond-pair correlation added by the G V B wavefunction lengthened the Ti-Cl bond 0.021 A, because the G V B wavefunction adds only limited left-right correlation and none of the dynamical correlation. For most A - B bonds, the calculated bond lengths at the SCF level are too short, and the correlation added by a G V B calculation accounts for a major portion of the non-dynamical correlation error in the SCF wavefunction. But for Ti-Cl bonds, both the SCF and G V B calculations predict too long a bond distance because they do not include necessary dynamical atomic correlation of the CI atoms. In TiCU, each CI satisfies the octet rule by sharing a pair of electrons with Ti. Thus, to each other, the CI atoms appear as noble gas atoms. As is well known (29i 30). the Hartree-Fock wavefunction does not adequately describe the attraction of two noble-gas-type atoms. SCF level calculations (29) of the H e , N e , and A r potentials show the atoms in these dimers to be too strongly repulsive at close 2
2
2
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
2
22
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
THE CHALLENGE OF d AND f ELECTRONS
Figure 2. Wavefunction plots of T i C l molecular orbitals in the C l - T i plane: a, ^ without d functions; b, ^ with d functions. Contours are geometric beginning at ±0.001 e~au~ and incremented by doubling the previous contour value. Continued on next page. 4
3
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
WILLIAMSON & HALL
Geometric Optimizations of Tetrahedral Complexes 23
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
2.
Figure 2. Continued. Wavefunction plots of T i C l molecular orbitals in the C l - T i plane: c, t a without d functions; d, ^ a with d functions. Contours are geometric beginning at ±0.001 e"au~ and incremented by doubling the previous contour value. Continued on next page. 4
2
3
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
24
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
THE CHALLENGE OF d AND f ELECTRONS
Figure 2. Continued Wavefunction plots of T i C l molecular orbitals in the C l - T i plane: e, t 7T without d functions; f, t * with d functions. Contours are geometric beginning at ±0.001 e~au~ and incremented by doubling the previous contour value. Continued on next page. 4
2
2
3
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Geometric Optimizations of Tetrahedral Complexes 25
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
WILLIAMSON & HALL
Figure 2. Continued Wavefunction plots of T i C l molecular orbitals in the C l - T i plane: g, e without d functions; h, e with d functions. Contours are geometric beginning at ±0.001 e"au~ and incremented by doubling the previous contour value. Continued on next page. 4
3
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
26
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
THE CHALLENGE OF d AND f ELECTRONS
Figure 2. Continued. Wavefunction plots of T i C l molecular orbitals in the C l - T i plane: i , tj without d functions; j , i with d functions. Contours are geometric beginning at ±0.001 e"au~ and incremented by doubling the previous contour value. 4
x
3
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
2. WILLIAMSON & HALL
Geometric Optimizations of Tetrahedral Complexes 27
distances and lacking a van der Waals minimum at long distances. Because the CI atoms are noble-gas-like, the C l - C l interactions in T i C U sterically hinder the CI atoms from bonding close to the T i atom. The accurate determination of the Ti-Cl bond distance will require a large configuration-interaction calculation. T i C h C H * Geometry Optimizations. The results of complete SCF geometry optimizations on T i C l C H using basis sets I-IV are shown in Table II. As the basis sets are improved, the Ti-C and C-H bond distances show the smallest changes of 3
3
Table II. SCF Geometries for C l T i C H
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
3
Basis Set M Ligand 1. I I 2. n II n 3. HI 4. n IV ED Experiment* Reference 15
3
Bond Distances (A) C-H Ti-C Ti-Cl 1.087 2.229 2.011 1.092 2.251 2.013 2.012 1.095 2.258 1.091 2.020 2.219 2.042 1.158 2.185
Ti-C-H 107.9 108.3 108.2 108.2 101.0
Angles O Cl-Ti-C 102.4 103.4 103.2 103.7 105.2
a
0.009 A and 0.008 A , respectively, and the Ti-Cl bond distance shows the largest change of 0.039 A. The T i - C - H and Cl-Ti-C angles vary only 0.4° and 1.3°, respectively. The geometry optimizations with T i and ligand basis sets II give Ti-C and C-H bond distances 0.029 A and 0.066 A shorter and a Ti-Cl bond length 0.066 A longer than the ED results. When polarization functions are added to the CI, C and H functions (basis set IV), the differences between the E D and the calculated Ti-Cl and T i - C bond lengths decrease to 0.034 A and 0.022 A, respectively, and the analogous difference for the C-H bond distance increases to 0.067 A. For basis set II, the Cl-Ti-C and Ti-C-H angles are 1.8° smaller and 7.2° larger, respectively, than the E D angle. When polarization functions are added to all atoms but T i , the differences decrease to 1.5° and 7.1°, respectively. Although most of these optimized geometric parameters are in good agreement (±0.03 A, ±1.5°) with the ED results, the optimized Ti-C-H angle and C-H distance do not agree with the ED result. In the G V B geometry (see Table III), the already long Ti-Cl distance increases 0.035 A, the Ti-C bond lengthens 0.168 A, and the Ti-C-H and Cl-Ti-C angles both decrease 4.4° and 3.9°, respectively. Previous work by Ditchfield and Seidman (21)
Table HI. G V B Geometries for C l T i C H 3
Basis Set M Ligand 1. IH in 2. in in in 3. in ED Experiment* Reference 15
Bond Distances (A) C-H Ti-Cl Ti-C 1.108 2.296 2.180 ED 1.111 ED ED 1.109 2.295 2.042 1.158 2.185
3
Angle? (°) Ti-C-H Cl-Ti-C 99.3 103.8 ED 106.5 100.5 105.5 105.2 101.0
a
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
28
THE CHALLENGE OF d AND f ELECTRONS
has shown that the addition of electron correlation to the SCF wavefunction in A H (A=C,N,0,F) molecules usually has a small effect on bond lengths and angles. Although an increase in the bond length is not surprising for a G V B calculation, the magnitude of the increase in the Ti-C bond length is surprising. Because the G V B wavefunction overemphasizes dissociation, the C ^ T i and CH3 moieties have too much radical-like C^Ti- and CH3 character. As the fragments become more radical like they flatten toward their equilibrium planar geometry. Even with the long Ti-C bond distance, the Ti-C-H angle is still 2.8° larger than the ED angle. If there were only steric repulsions between the hydrogens and titanium, the bond angles would increase as the fragments are forced back together. On the other hand, if there were attractive agostic hydrogen interactions, the hydrogens would bend towards the titanium atom as the fragments are brought closer, which would decrease the angle closer to the ED value. When we shorten the Ti-C bond distance to the E D bond length and optimize the remaining geometric parameters at the G V B level, the Ti-C-H and Cl-Ti-C angles increase 1.7° and 1.3°, respectively. When, in addition, we shorten the Ti-Cl bond length and the increase the Cl-Ti-C angle to the ED values, the Ti-C-H bond angle increases an additional 1.0°, to a value 5.5° larger than the reported ED angle. Thus, the bending of the hydrogen atoms away from the titanium atom as the Ti-C distance is shortened shows that the titanium-hydrogen interactions are repulsive and not attractive. We improved the calculation further by using a CASSCF wavefunction with an active space of 8 electrons in 8 orbitals (8/8). The results of the geometry optimization (see Table IV) give a Ti-C bond length of 2.119 A and a Ti-C-H angle of 105.4° (the Ti-Cl bond distance and Cl-Ti-C angle were frozen at the ED values). In
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
n
Table IV. CASSCF Geometries for C l T i C H 3
Basis Set M Ligand Method II II SCF II II 8/8 CAS II II 8/11 CAS ED Experiment* Reference 15
Bond Distances (A) C-H Ti-Cl Ti-C 1.092 2.251 2.013 ED 2.119 1.112 ED 2.106 1.112 2.042 1.158 2.185
3
Angles (°) Ti-C-H Cl-Ti-C 103.4 108.3 105.4 ED ED 106.2 105.2 101.0
a
contrast to the SCF geometries, but like the G V B geometry, the CASSCF Ti-C bond distance is longer than the E D value and the Ti-C-H angle shows a slightly larger decrease from the tetrahedral angle. However, when we increase the size of the active space to 11 orbitals (8/11), with additional Ti-C c and n bonding orbitals, the Ti-C bond shortens 0.013 A and the Ti-C-H angle increases 0.8°. This CASSCF includes those virtual T i drc-orbitals which are involved in the qualitative description of the agostic hydrogen interaction. The increase one observes in the Ti-C-H angle as the Ti-C bond distance shortens, which one also observed in the G V B geometry optimizations, shows that the Ti-C-H angle is very sensitive to changes in the Ti-C bond distance. If the wavefunction were improved by including even more correlation, one would expect the Ti-C bond distance to shorten further toward the ED value and as the Ti-C bond shortens, one would also expect the Ti-C-H angle to increase towards a tetrahedral angle.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
2. WILLIAMSON & HALL
Geometric Optimizations of Tetrahedral Complexes 29
Vibrational Frequencies Calculations. Berry et al. ( I D reported that the methyl rocking frequency of TiCbCH3 is much lower than the analogous frequency of GeCbCHs. They presumed this anomalously low frequency of TiCbCHs to result from the "flattening" of the methyl hydrogens. We calculated the vibration frequencies of the hypothetical model complexes TiH3CH3 and GeH3CH3 after first optimizing their geometries. The results of the geometry optimizations (see Table V) Table V . SCF Geometries for MH3CH3 (M = Ti, Ge) Angles (°) M-C-H H-M-C 109.9 108.3 110.7 110.3
Bond Distances (A)
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
1. 2.
Metal Ti Ge
Basis I I
M-H 1.710 1.533
M-C 2.035 1.959
C-H 1.092 1.086
show the expected differences in geometry between the T i and Ge complexes. The vibration frequencies were calculated by taking finite differences of energy gradients beginning at these optimized geometries. The calculated frequencies, when compared with the experimental frequencies (see Table VI), show errors expected of calculations at the SCF level. However, Table VI. MX3CH3 Vibrational Frequencies in cirr
Mode e C H rock i CH3 deform. e CH3 deform. a CH3 stretch C H stretch e Reference 15 3
a
3
Ge 978 1468 1611 3190 3270
Calcul. (X-Ti 577 1360 1564 3122 3208
Diff. 401 108 47 68 62
a
Ge 825 1246 1403 2940 3019
Exper. (X=C\) Ti Diff 580 245 1052 194 1375 28 2894 46 2980 39 a
when one compares the SCF differences between the T i and Ge model complexes with the experimental differences between the T i and Ge chloride complexes, the differences are in good agreement. The large difference in the methyl rocking frequency observed in the experiment is predicted by the SCF calculation, the same SCF calculation that predicts no flattening of the methyl hydrogen atoms. If the large difference between the T i and Ge methyl rocking frequency is not the result of hydrogen flattening, what then is the cause of the observed difference in the methyl rocking frequencies? To answer the above question we did point by point SCF calculations of the model complexes as the methyl group rocks. A plot of the resulting energies for each point calculated is shown in Figure 3 and, as expected, the titanium complex potential curve is flatter than the germanium complex potential curve. When we rock the methyl group on the model complexes +45°, the titanium complex is destabilized by
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
THE CHALLENGE OF d AND f ELECTRONS
• Ge • Ti
Rock Angle (degrees)
Figure 3. Relative energy of
TUH3CH3
and GeH3CH3 for the methyl rock
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
2. WILLIAMSON & HALL
Geometric Optimizations of Tetrahedral Complexes 31
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
-1
1
12.9 kcal m o l and the germanium complex is destabilized by 34.9 kcal m o l . Eisenstein and Jean (22) using extended Huckel calculations on the titanium model complex also found the rocking motion to be weakly destabilizing. At the equilibrium geometry the deformation density plot (Figure 4a) shows a large buildup of density aligned symmetrically along the Ti-C bond axis. As the methyl group rocks (Figures 4b and 4c) the density buildup is no longer symmetric about the Ti-C bond axis, but rather shifted to one side of the axis. Although the methyl group in Figures 4b and 4c has been tilted 45° from equilibrium the maximum density between the Ti and C atoms has only tilted an average 22°. As was seen in the titanium complex, the germanium complex at equilibrium geometry (Figure 5a) shows a similar symmetric buildup of density along the Ge-C bond axis. However, as the methyl group rocks 45° (Figures 5b and 5c) the density shifts to one side of die axis with an average tilt of 35°, 13° larger than the tilt in the T i complex. It is also interesting to note that the density in the C-H bond region decreases slightly for both the Ge and T i complexes as the hydrogen is tilted toward the metal. The density between the metal and the other hydrogen in the plot remains unchanged for the germanium complex but increases slightly for the titanium complex as the methyl group is rocked awayfromequilibrium. Goddard, Hoffmann, and Jemmis (33) in a study of alkyl tantalum complexes have suggested that rehybridization of the metal-carbon bond with empty d-orbitals on the metal stabilizes the rocked methyl group. Titanium, like tantalum, has low-lying empty d-orbitals that allow facile rehybridization of the metal-carbon bond. Germanium, on the other hand, has used its s- and p-orbitals and the empty d-orbitals are at very high energy, therefore no empty orbitals are available for rehybridization. The difference in the ease of rehybridization explains the much lower methyl rocking frequency for the titanium complex than for the germanium complex. The deformation density plots do not contradict this explanation. Furthermore, this explanation is supported by the total metal-carbon overlap populations which actually increase for the Ti-methyl rock but decrease for the Ge-methyl rock. A n increase in the titanium d^-carbon and p -carbon more than compensates for the decrease in the d 2-carbon overlap populations as the methyl group is rocked. The germanium complex however, shows no change in any germanium d-carbon overlap population and the very slight increase in the p -carbon overlap population does not compensate for the loss in the p -carbon overlap population as the methyl group is rocked. x
z
x
z
Geminal H.H Coupling Constant. Using the method of Pople and Santry (22) we calculated the difference between the H,H coupling constant of CICH3 and HCH3 to be 1.6 Hz, which is in good agreement with the experimentally determined value (34). The calculated H,H coupling constant for TiCbCH3 at the optimized geometry is 8.0 H z larger than the coupling constant for HCH3. Although this change is smaller than the 23.7 Hz change observed experimentally (15.34). this simple model does predict a change which is both large and positive. When the optimized geometry of TiCl3CH3 is flattened to the E D geometry the H,H coupling constant increases only 3.6 Hz. By comparing the increase in the coupling constant of X C H 3 as X changes from H to T i C l 3 with the increase in the coupling constant as the hydrogens in T i C l 3 C H 3 are flattened, we see that the flattening of the methyl hydrogens has a relatively small effect on the coupling constant when compared with the large effect of the T i C ^ substituent. However, Green and Payne (35) using extended Huckel calculations observed the effect of the substituent on the coupling constant to be only half of the
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
THE CHALLENGE OF d AND f ELECTRONS
Figure 4. Deformation density of TiH3CH in the H-C-Ti-H plane: a) equilibrium geometry b) methyl rocked 45° counterclockwise c) methyl rocked45° clockwise. Contours are geometric beginning at±0.001 e au and incremented by doubling the previous contour value. 3
_
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
3
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
WILLIAMSON & HALL
Geometric Optimizations of Tetrahedral Complexes
a
b
c Figure 5. Deformation density of GeHsCHsin the H-C-Ge-H plane: a) equilibrium geometry b) methyl rocked 45° counterclockwise c) methyl rocked 45° clockwise. Contours are geometric beginning at ±0.001 e~au and incremented by doubling the previous contour value. -3
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
34
THE CHALLENGE OF d AND f ELECTRONS
effect of the geometry change on the coupling constant, which is opposite to what we predict using ab initio calculations. We believe that ab initio calculations describe the true nature of this titanium complex more accurately than the semi-empirical extended Hiickel calculations. It is the strong a-donor and weak ic-acceptor character of the TiCl3 substituent which results in the observed positive shift in the coupling constant. Thus, one can explain the coupling constant without postulating any geometric distortions.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
Conclusions Optimization of the geometry of TiCL* at the SCF level results in a Ti-Cl bond length which is longer than the experiment, even when d- and f-type polarization functions are added to the basis set. For covalently bonded systems one expects a wavefunction at the Hartree-Fock limit to give bond lengths shorter than the experiment i f they are not sterically crowded. Because the Hartree-Fock wavefunction overestimates the CI—CI repulsions, the Ti-Cl bond distances remain long, even in large basis sets. Geometry optimizations of TiCl3CH3 also show long Ti-Cl bond distances. The T i - C - H angle is close to tetrahedral geometry with little, if any, flattening of the hydrogen atoms. Because of the known problem that electron diffraction has with determining the positions of hydrogen atoms, the large difference between the optimized value and the electron diffraction value for the T i - C - H angle is not surprising. Our calculations correctly predict the anomalously low methyl-rocking frequency for the titanium complex without hydrogen flattening. This low methylrocking frequency is due to stabilization of the rocking motion by low-lying empty dorbitals on Ti. The large positive geminal hydrogen coupling constant is primarily due to the a-donor and xc-acceptor character of the TiCb moiety and not the flattening of the methyl hydrogens. We do not believe agostic interactions are strong enough to symmetrically flatten a methyl group. The key to an agostic interaction is the low-energy rocking distortion. Thus, we would expect all agostic methyls to be rocked with the H-C-H angles close to those in similar organic compounds. Acknowledgments The authors would like to thank the National Science Foundation (Grant C H E 8619420) and C R A Y Research for support of the work. This research was conducted on an IBM-3090 and FPS-264 at the Cornell National Supercomputer Facility, a resource for the Center for Theory and Simulation in Science and Engineering at Cornell University, which is funded in part by the National Science Foundation, New York State, and the I B M Corporation, on a C R A Y X - M P at C R A Y Research, Mendota Heights, Minnesota, and on a V A X 11/780 and FPS-164 at the Chemistry Department of Texas A & M University. We would also like to thank Dr. Martyn F. Guest at SERC, Daresbury Laboratory, Warrington, U K . for making his version of GAMESS available and Drs. Thomas Dunning and Ron Shepard at Argonne National Laboratory for making QUEST available. Literature Cited 1. Pietro, W.J.; Hehre, W.J. J. Comp. Chem. 1983, 4, 241.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
2. WILLIAMSON & HALL
Geometric Optimizations of Tetrahedral Complexes 35
2. Seijo, L.; Barandiaran, Z.; Klobukowski, M.; Huzinaga, S. Chem. Phys. Lett. 1985, 117, 151. 3. Faegri Jr., K.; Almolf, J. Chem. Phys. Lett. 1984, 107, 121; 4. Luthi, H.P.; Ammeter, J.H.; Almolf, J.; Faegri Jr., K . J. Chem. Phys. 1982, 77, 2002. 5. Pitzer, R.M.; Goddard, J.D.; Schaefer, H.F. J. A m . Chem. Soc. 1981, 103, 5681. 6. Yates, J.H.; Pitzer, R . M . J. Chem. Phys. 1979, 70, 4049. 7. Kataura, K.; Sakaki, S.; Morokuma, K . Inorg. Chem. 1981, 20, 2292. 8. Spangler, D.; Wendoloski, J.J.; Dupuis, M.; Chen, M . M . L . ; Schaefer, H.F. J. Am. Chem. Soc. 1981, 103, 3985. 9. Guest, M.F.; Hillier, I.H.; Vincent, M.; Rosi, M. J. Chem. Soc. Chem.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
Commun. 1986, 438. 10. Luthi, H.P.; Siegbahn, P.E.M.; Almolf, J. J. Phys. Chem. 1985, 89, 2156. 11. Luthi, H.P.; Siegbahn, P.E.M.; Almolf, J.; Faegri Jr., K.; Heiberg, A . Chem. Phys. Lett. 1984, 111, 1. 12. Almolf, J.; Faegri Jr., K.; Schilling, B.E.R.; Luthi, J.P. Chem. Phys. Lett. 1984, 106, 266. 13. Williamson, R.L.; Hall, M.B. Int. J. Quantum Chem., Quantum Chem. Symp. 21, 1987, 503. 14. Dobbs, K.D.; Hehre, W.J. J. Comput. Chem. 1987, 8, 861. 15. Berry, A.; Dawoodi, Z.; Derome, A . E.; Dickinson, J. M.; Downs, A . J.; Green, J. C.; Green, M . L . H.; Hare, P. M . ; Payne, M . P.; Rankin, W. H . ; Robertson, H . E . J. J. Chem. Soc. Chem. Commun. 1986, 519. 16. Brookhart, M.; Green, M . L . H. J. Organomet. Chem. 1983, 250, 395. 17. Dawoodi, Z.; Green, M . L . H.; Mtetwa, V . S. B.; Prout, K . J. J. Chem. Soc. Chem. Commun. 1982, 1410. 18. Dawoodi, Z.; Green, M . L . H.; Mtetwa, V . S. B.; Prout, K . J. J. Chem. Soc. Chem. Commun. 1982, 802. 19. Dawoodi, Z.; Green, M . L . H.; Mtetwa, V . S. B.; Prout, K.; Schultz, A . J.; Williams, J. M.; Koetzle, T. F. J. Chem. Soc. Dalton Trans. 1986, 1629. 20. Obara, S.; Koga, N . Morokuma, K . J. Organomet. Chem. 1984, 270, C33. 21. Williamson, R.L.; Hall, M . B . J. Am. Chem. Soc. 1988, 110, 4428. 22. Pople, J.A.; Santry, D.P. Mol. Phys. 1963, 8, 1. 23. Maciel, G.E.; McIver Jr., J . W . ; Ostlund, N.S.; Pople, J.A. J. A m . Chem. Soc. 1970, 92, 4151. 24. Gaussian Basis Sets for Molecular Calculations: Huzinaga, S., Ed.; Amsterdam: Elsevier, 1984 25. Bauschlicher Jr., C.W.; Seigbahn, P.E.M. Chem. Phys. Lett. 1984, 104, 331. 26. Dunning Jr., T.H.; Hay, P.J. In Methods of Electronic Structure Theory; Schaefer, H.F., Ed.; Plenum Press: New York, 1977; Vol. 4 Chapter 1. 27. Bobrowicz, F.W.; Goddard, W.A. In Methods of Electronic Structure Theory: Schaefer, H.F., Ed.; Plenum Press: New York, 1977; Vol. 4 Chapter 4. 28. Morino, Y.; Uehara, H . J. Chem. Phys. 1966, 45, 4543. 29. Gilbert, T.L.; Wahl, A . C . J. Chem. Phys. 1967, 47, 3425. 30. Schaefer, H.F.; McLaughlin, D.R.; Harris, F.E.; Alder, B.J. Phys. Rev. Lett. 1970, 25, 988. 31. Ditchfield, R.; Seidman, K . Chem. Phys. Lett. 1978, 54, 57. 32. Eisenstein, O.; Jean, Y . J. Am. Chem. Soc. 1985, 107, 1177.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
THE CHALLENGE OF d AND f ELECTRONS
36
33. Goddard, R.J.; Hoffmann, R.; Jemmis, E.D. J. Am. Chem. Soc. 1980, 102, 7667. 34. Pople, J.A.; Bothner-By, A . A . J. Chem. Phys. 1965, 42, 1339. 35. Green, J. C.; Payne, M.P. Magn. Res. Chem. 1987, 25, 544.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch002
R E C E I V E D December 9, 1988
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Chapter 3
LCGTO—Xα
Study on the Agostic Interaction in Cl TiCH 3
3
N . Rösch and P. Knappe Lehrstuhl für Theoretische Chemie, Technische Universität München, 8046 Garching, Federal Republic of Germany The geometry and selected vibrational frequencies of Cl TiCH have been calculated using the LCGTO-Xα method. The methyl group is found essentially undistorted, thus providing no indication for an agostic interaction.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch003
3
3
We would like to comment on the findings of Williamson and Hall on tetrahedral T i f l V ) complexes (1,2) by reporting some results from our L C G T O - X a calculations on the geometry of C l T i C H and of related titanium complexes (Knappe, P.; Rosch, N . J . Organometal. Chem.. in press). Our study was prompted by the experimental finding (3) in C l T i C H of an unusual agostic interaction at an a carbon atom and by results of an Extended Hiickel analysis (4) of the agostic interaction in tetrahedral and octahedral titanium complexes. For T i C U (in TSL
L
H
W C I A
l
U„
2
^
c
1
H
I , *H %
A
l
C = C H
c-c c=c
Η Rh'
^Rrt
=
""Rh Cl^ >L
Rh Cl^ I >L S
-Ho
1
fast / - x c
Η
rate determing step
c
Η H
[RhCI(H)L (Ç-ÇH)]
'Rh
3
cr I s
L
n
ι H
dominant catalytic cycle Scheme 1
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch006
6.
KOGA & MOROKUMA
Potential Energy Surface of Olefin Hydrogénation
t i o n o f a l k a n e . The r a t e - d e t e r m i n i n g s t e p has been b e l i e v e d t o be the o l e f i n i n s e r t i o n s t e p . We u s e d PH^ i n s t e a d o f PR~ and C^H^ as a model o f o l e f i n . The model c a t a l y t i c c y c l e we s t u d i e d i s shown i n Scheme 2. A l l the e q u i l i b r i u m and t r a n s i t i o n s t a t e s t r u c t u r e s were o p t i m i z e d by the RHF energy g r a d i e n t method. W h i l e two p h o s p h i n e s a r e always t r a n s t o each o t h e r i n a l l the i n t e r m e d i a t e s i n the H a l p e m mechanism, Brown e t a l . have v e r y r e c e n t l y p r o p o s e d a d i f f e r e n t mechanism, i n which the c i s b i p h o s p h i n e i n t e r m e d i a t e s p l a y an e s s e n t i a l r o l e ( 4 ) . T h e i r m o l e c u l a r m o d e l i n g c a l c u l a t i o n s where the van der Waals energy i s c a l c u l a t e d between s u b s t i t u t e d o l e f i n s and the Rh fragment w i t h b u l k y t r a n s p h o s p h i n e s have s u g g e s t e d t h a t when the s u b s t i t u e n t s on the o l e f i n a r e b u l k y , the s t e r i c r e p u l s i o n i s too l a r g e f o r the o l e f i n t o c o o r d i n a t e . T h e i r NMR e x p e r i m e n t s have shown the e x i s t e n c e o f the f o l l o w i n g e q u i l i b r i u m ( E q u a t i o n 1) i n w h i c h an i n t e r m e d i a t e w i t h a p a i r o f c i s phosphines can be formed. I n f a c t , ^ R h C l C P R ^ ) ^ has been d e t e c t e d i n the c a t a l y t i c system.
H
H
Phi
Rh-
•pPh, b
PPh
I
.ΛΝ
ΡΡΠ
'
a
(1)
a
- Rh ' .
3
+ PPh
j ^ pph CI
3
b
Cl
3
Based on these r e s u l t s , t h e y have s u g g e s t e d t h a t the i n t e r m e d i a t e s o f the c a t a l y t i c system have two c i s p h o s p h i n e s . I n Scheme 3 i s shown the new mechanism. I n t h i s Scheme, the key s t e p i s i s o m e r i z a t i o n (Equation 2), t r a n s - H R h C l ( P R ) t o c i s - H R h C l ( P R > presum a b l y through p s e u d o r o t a t i o n . 2
PPh
Rri j PPh
2
2
I
Η
.
3
3
2
H
3
I .,Λ CI
3
Η —
,Λ ™3 Ρ
Rh. j ^PPh CI
(2) 3
Then, the o l e f i n i n s e r t i o n and the r e d u c t i v e e l i m i n a t i o n take p l a c e from the r e s u l t a n t c i s b i p h o s p h i n e complex. I n t h i s a r t i c l e , we w i l l compare the e n e r g e t i c s o f the 'conven t i o n a l ' H a l p e m mechanism w i t h t h a t o f the Brown mechanism. The b a s i s f u n c t i o n s u s e d a r e the 3-21G f o r e t h y l e n e and h y d r i d e s , the ST0-2G f o r ' s p e c t a t o r ' l i g a n d s , PH^ and CI, and v a l e n c e double z e t a b a s i s f u n c t i o n s f o r Rh w i t h e f f e c t i v e c o r e p o t e n t i a l r e p l a c i n g the c o r e e l e c t r o n s (up t o 4p) (5a.b.6). I n a d d i t i o n , we c a r r i e d out the MP2 c a l c u l a t i o n s a t s e l e c t e d , R H F - o p t i m i z e d s t r u c t u r e s w i t h a l a r g e r b a s i s s e t , which c o n s i s t s o f u n c o n t r a c t e d (3s,3p,4d) f u n c t i o n s from the above v a l e n c e DZ s e t f o r Rh, 4-31G f o r the e t h y l group, (10s,7p)/[3s2p] f o r Ρ and CI, and ( 4 s ) / [ 3 s ] f o r the h y d r i d e s (5ce ) . The b a s i s s e t u s e d i s l i m i t e d and the e l e c t r o n c o r r e l a t i o n t a k e n i n t o account f o r a few c r i t i c a l s t e p s i s m i n i m a l . T h e r e f o r e ,
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
79
THE CHALLENGE OF d AND f ELECTRONS
C H 2
6
A*-* Reductive e l i m i n a t i o n ^ / ^
J. Rh—Cl ^
H
2
I Oxidati
CH -Rh-Cl 2
H-CH
2
L
t
H 5b
l^-CH^CHz
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch006
Isomerization
Olefin coordination
g
L u Olefin insertion
Scheme 2
Scheme 3
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
6.
KOGA & MOROKUMA
Potential Energy Surjhce of Olefin Hydrogénation
the r e s u l t s p r e s e n t e d quantitative .
h e r e s h o u l d be
considered
t o be
semi-
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch006
H a l p e m mechanism S i n c e our c a l c u l a t i o n s on the H a l p e m mechanism have been p u b l i s h e d (2)» we w i l l g i v e a b r i e f summary f o r comparison i n a s u c c e e d i n g s e c t i o n . The p o t e n t i a l energy p r o f i l e shown i n F i g u r e 1 i s cons t r u c t e d from the e n e r g e t i c s o f the e l e m e n t a r y r e a c t i o n s i n v o l v e d i n the H a l p e m mechanism. The o p t i m i z e d s t r u c t u r e s a r e shown i n F i g u r e 2. The f i r s t s t e p o f the H^ o x i d a t i v e a d d i t i o n i s e x o t h e r m i c and l e a d s t o the d i h y d r i d e complex 3 . D u r i n g t h i s s t e p , t h e r e may be an H^ complex 2 from which o x i d a t i v e a d d i t i o n t a k e s p l a c e w i t h a l m o s t no a c t i v a t i o n b a r r i e r . The ethylene c o o r d i n a t i o n that follows r e q u i r e s no a c t i v a t i o n energy. The r e s u l t a n t e t h y l e n e d i h y d r i d e complex 4 i s i n the v a l l e y o f the p o t e n t i a l energy s u r f a c e o f the c a t a l y t i c c y c l e . E t h y l e n e i n s e r t i o n r e q u i r e s a much h i g h e r a c t i v a t i o n energy o f 18 k c a l / m o l and i s endothermic by 16 k c a l / m o l a t the RHF l e v e l . The t r a n s e t h y l h y d r i d e complex, the d i r e c t p r o d u c t o f e t h y l e n e i n s e r t i o n , i s u n s t a b l e due t o the c i s e f f e c t o f CI t o be mentioned below and the t r a n s e f f e c t o f H and o 5 ' T h e r e f o r e i s o m e r i z a t i o n takes p l a c e t o g i v e more s t a b l e e t h y l h y d r i d e comp l e x e s , w h i c h have e t h y l and h y d r i d e c i s t o each o t h e r and a r e the s t a r t i n g p o i n t o f the f i n a l reductive e l i m i n a t i o n step. This i s o m e r i z a t i o n p r o c e e d s t h r o u g h h y d r i d e and c h l o r i d e m i g r a t i o n . The f i n a l r e d u c t i v e e l i m i n a t i o n step r e q u i r e s a s u b s t a n t i a l energy b a r r i e r o f 15 k c a l / m o l . The p o t e n t i a l energy p r o f i l e i s smooth w i t h o u t e x c e s s i v e b a r r i e r s and too s t a b l e i n t e r m e d i a t e s which would b r e a k the sequence o f s t e p s . The r a t e - d e t e r m i n i n g s t e p i s found t o be o l e f i n i n s e r t i o n f o l l o w e d by i s o m e r i z a t i o n , s u p p o r t i n g the H a l p e m mechanism. I s o m e r i z a t i o n o f the e t h y l h y d r i d e complex i s an i m p o r t a n t p a r t o f the r a t e - d e t e r m i n i n g s t e p . These two r e a c t i o n s , e x o t h e r m i c o v e r a l l , has an o v e r a l l b a r r i e r h e i g h t o f about 20 k c a l / m o l . The t r a n s e t h y l h y d r i d e complex, the p r o d u c t o f e t h y l e n e i n s e r t i o n , may n o t be a l o c a l minimum (per MP2 c a l c u l a t i o n ) and t h e s e two s t e p s may w e l l be a combined s i n g l e s t e p . The a c t i v a t i o n b a r r i e r o f r e d u c t i v e e l i m i n a t i o n , though s u b s t a n t i a l , i s s m a l l e r t h a n t h a t o f the r e v e r s e o f the r a t e d e t e r m i n i n g s t e p ( i s o m e r i z a t i o n and 0-hydrogen e l i m i n a t i o n ) . T h i s i s a v e r y i m p o r t a n t r e q u i r e m e n t o f a good o l e f i n hydrogénation catalyst. I f t h i s r e v e r s e r e a c t i o n i s easy, i t would l e a d t o u n d e s i r a b l e o l e f i n i s o m e r i z a t i o n . F o r i n s t a n c e , i n the same hydrogén a t i o n c y c l e c a t a l y z e d by a Pt system, we f o u n d t h a t the r a t e determining step i s r e d u c t i v e e l i m i n a t i o n r a t h e r than o l e f i n i n s e r t i o n . T h i s p o t e n t i a l energy p r o f i l e i s e x p e c t e d t o g i v e o l e f i n i s o m e r i z a t i o n through s u c c e s s i v e o l e f i n i n s e r t i o n / ^ - h y d r o g e n e l i m i n a t i o n . Thus the Pt complex i s n o t a good c a t a l y s t f o r o l e f i n hydrogénation. On the o t h e r hand, the p o t e n t i a l p r o f i l e o f the W i l k i n s o n c a t a l y s t i n d i c a t e s an e f f i c i e n t hydrogénation w i t h o u t isomerization. c
H
I t i s i m p o r t a n t , as mentioned above, t h a t the o l e f i n i n s e r t i o n i s r a t e - d e t e r m i n i n g . T h e r e f o r e , we have compared the r e a c t i o n e n e r g e t i c s between H R h C l ( P H ) ( C H ) and H R h H ( P H ) ( C H ) , and 2
3
2
2
4
2
3
2
2
4
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
81
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
-40
-30
-20 -
-10
0
kcal/mole|
Olefin insertion
Isomerization
Reductive elimination
F i g u r e 1. P o t e n t i a l energy p r o f i l e o f the e n t i r e c a t a l y t i c c y c l e i n the H a l p e m mechanism f o r o l e f i n h y d r o g é n a t i o n , i n k c a l / m o l a t the RHF l e v e l , r e l a t i v e t o l+C^H^+H^. Numbers i n p a r e n t h e s e s a r e the MP2 energy a t the RHF o p t i m i z e d g e o m e t r i e s , r e l a t i v e t o 4 .
coordination
Oxidative addition Olefin
TS(5b+1)
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch006
ο
00
KOGA & MOROKUMA
Potential Energy Surface of Olefin Hydrogénation
H-? _. 2-280 I P s t-tuw 0-8631
H ;
A6
1
164-8
1.514 \ 2.302 ^Rhs—CI \-i ^— 145.5 M
H
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch006
Π
^ C = C ^ H
H ,
2.600 χ 1 2-575
n
2-282
5 7 2
^/.:R[>—Cl 160-8 TS(2-3)
W 2 C
/
« * ;
^ f L ^ R f t — CI 83.ΐΊ^925 2·3Α2 Η,1.504
/
^ t e o
Π
H1-587
TS(4-5)
H 1.081
1 1 8 1
Ha Hi
'l547
| 79-9\ 2.137
H
91>/W1-621
2 7 2
N
H
1026* 2.O6I 17^17^
' · / 115.3 H 751
H
H
2-247
V 2.095^-:
2
87-3 Α - Ο 0 2 Ό u^i-Rh^—a n
1.558
1.553/Η^λ®:- α VH
HU72 5
b
2.313
5a
1 H
Η
Λ
"7.2^\ .OOS Îoœ
TS(5+5a)
5
H
Γ1539
C 2-185 _165.1
4 ? ^ ^ —
CI
1-524 TS(5b+1)
F i g u r e 2. O p t i m i z e d s t r u c t u r e s ( i n À and deg) o f some i m p o r t a n t s p e c i e s . TS(2->3), f o r i n s t a n c e , denotes t h e t r a n s i t i o n s t a t e c o n n e c t i n g 2 and 3. Though p r a c t i c a l l y a l l t h e g e o m e t r i c a l p a r a m e t e r s were o p t i m i z e d , o n l y e s s e n t i a l v a l u e s a r e shown. Two PH3's, one above and one below the p l a n e o f paper, a r e o m i t t e d for c l a r i t y .
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
THE CHALLENGE OF d AND f ELECTRONS
84
f o u n d t h a t , i n the former complex, the weak t r a n s i n f l u e n c e o f CI makes the Rh-H bond t o be b r o k e n s t r o n g e r . I n a d d i t i o n the s t r o n g CI c i s e f f e c t makes the Rh-C bond t o be formed weaker, resulting in e n d o t h e r m i c e t h y l e n e i n s e r t i o n . These two e f f e c t s o f CI combined appears t o be e s s e n t i a l t o make e t h y l e n e i n s e r t i o n the r a t e determining step. firown, mechanism The f i r s t p o i n t o f d i f f e r e n c e o f the Brown mechanism from the H a l p e r n mechanism i s i s o m e r i z a t i o n o f H R h C l ( P H ~ ) . T h e r e f o r e , we have i n v e s t i g a t e d the s t a b i l i t y o f isomers o f R R h C l ( P H ) , 3. The o p t i m i z e d s t r u c t u r e s o f H R h C l ( P H ~ ) a r e shown i n F i g u r e 3. The most s t a b l e isomer i s found t o be 3, the t r a n s i n t e r m e d i a t e o f the H a l p e r n mechanism. A l l the o p t i m i z e d s t r u c t u r e s b u t 3 a r e n e a r l y square p y r a m i d a l (though o p t i m i z a t i o n was done w i t h o u t symmetry restriction). The most s t a b l e square p y r a m i d a l isomer i s 3a w i t h a p i c a l H and b a s a l c i s p h o s p h i n e s . 3b w i t h a p i c a l p h o s p h i n e and b a s a l c i s h y d r i d e s i s n e x t . The r e m a i n i n g t h r e e i s o m e r s , 3c, 3d, and 3e a r e much more u n s t a b l e ; the e n e r g i e s r e l a t i v e t o 3 a r e 33, 37 and 39 k c a l / m o l , r e s p e c t i v e l y . Two h y d r i d e s w i t h s t r o n g t r a n s i n f l u e n c e a r e l o c a t e d t r a n s t o each o t h e r i n 3c. T h i s makes 3c 12 k c a l / m o l l e s s s t a b l e t h a n 3b i n which two h y d r i d e s a r e c i s . The l e a s t s t a b l e i s o m e r s , 3d and 3e, have a p i c a l CI. Comparison o f the s t a b i l i t y among the isomers o f 3 l e a d s t o the o r d e r o f a p i c a l p r e f e r e n c e : H>PH^>C1. H w i t h the s t r o n g e s t t r a n s i n f l u e n c e p r e f e r s the a p i c a l p o s i t i o n t h a t i s t r a n s t o the v a c a n t s i t e , and the most weakly t r a n s - i n f l u e n c i n g CI a t the a p i c a l position g i v e s the most u n s t a b l e isomers o f 3d and 3e. 3b and 3c a r e inbetween, i n a c c o r d w i t h the s t r e n g t h o f PH~ t r a n s i n f l u e n c e . S i n c e 3a and 3b a r e low i n energy, we have i n v e s t i g a t e d i s o m e r i z a t i o n from 3 t o 3a and 3b. Brown e t a l . have p r o p o s e d t h a t the c i s i n t e r m e d i a t e o f the c a t a l y t i c c y c l e i s 3b i n w h i c h one o f the b u l k y p h o s p h i n e s i s t r a n s t o o l e f i n and thus the v a c a n t c o o r d i n a t i o n s i t e i s l e s s crowded. There a r e two p o s s i b l e pathways f o r i s o m e r i z a t i o n o f 3 as shown i n Scheme 4. The i n t e r m e d i a t e s o f the second pathway a r e u n s t a b l e 3d and 3c, and i t i s u n l i k e l y t h a t i s o m e r i z a t i o n t a k e s p l a c e t h r o u g h them. The e a s i e r i s o m e r i z a t i o n pathway i s through 3b t o 3a. The t r a n s i t i o n s t a t e f o r PH^ m i g r a t i o n c o n n e c t i n g 3 and 3b have been l o c a t e d , as shown i n F i g u r e 3, w i t h the a c t i v a t i o n b a r r i e r f o r i s o m e r i z a t i o n from 3 t o 3b o f 27 k c a l / m o l (See Scheme 4 ) . T h e r e f o r e , one can c o n c l u d e i s o m e r i z a t i o n from 3 t o 3b o r 3a i s r a t h e r d i f f i c u l t ( c f . 18 k c a l / m o l , the a c t i v a t i o n energy o f the r a t e - d e t e r m i n i n g s t e p i n the H a l p e r n mechanism a t the same l e v e l o f c a l c u l a t i o n ) . S e t t i n g a s i d e t h i s h i g h a c t i v a t i o n b a r r i e r f o r a moment, the r e m a i n i n g s t e p s o f the c a t a l y t i c c y c l e i n the Brown mechanism w i l l be d i s c u s s e d . As shown i n the energy p r o f i l e o f the H a l p e r n mechanism, the e l e m e n t a r y r e a c t i o n s i n v o l v e d h e r e a r e e x p e c t e d t o be v e r y easy, and thus we have j u s t d e t e r m i n e d the s t r u c t u r e s and energies of intermediates but not of t r a n s i t i o n s t a t e s (Figure 5). The r e l a t i v e e n e r g i e s o f the i n t e r m e d i a t e s , shown i n F i g u r e 4, would be a good i n d i c a t o r o f the b a r r i e r o f each e l e m e n t a r y r e a c t i o n c o n n e c t i n g them; an endothermic r e a c t i o n r e q u i r e s a l a r g e a c t i v a t i o n energy and the a c t i v a t i o n b a r r i e r o f an e x o t h e r m i c s t e p i s low. 2
2
2
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch006
2
3
2
2
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Potential Energy Surface of Olefin Hydrogénation
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch006
KOGA & MOROKUMA
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch006
THE CHALLENGE OF d AND f ELECTRONS
F i g u r e 3. O p t i m i z e d s t r u c t u r e s ( i n  and deg) o f isomers o f H ^ R h C l i P H ^ ) ^ and the 3-*3b t r a n s i t i o n s t a t e , and t h e i r e n e r g i e s ( i n k c a l / m o l ) r e l a t i v e t o 3.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
H
2
9
0
complex isomerization
Olefin coordination
Dihydride
olefin insertion
Isomerization
complex
Ethyl hydride
Reductive elimination
F i g u r e 4. P o t e n t i a l energy p r o f i l e o f t h e c a t a l y t i c c y c l e i n t h e Brown mechanism f o r o l e f i n hydrogénation, i n k c a l / m o l a t t h e RHF l e v e l , r e l a t i v e t o 1+C H.+H .
Oxidative addition
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch006
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch006
THE CHALLENGE OF d AND f ELECTRONS
F i g u r e 5. O p t i m i z e d s t r u c t u r e s ( i n  and deg) o f some i m p o r t a n t s p e c i e s i n the Brown mechanism.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
6. KOGA & MOROKUMA
Potential Energy Surface of Olefin Hydrogénation
E t h y l e n e c o o r d i n a t i o n t o 3b, t h e Brown's i n t e r m e d i a t e , g i v e s 4a w h i c h i s h i g h e r i n energy t h a n 4 by 6 k c a l / m o l . Olefin insertion o f 4a c a n l e a d t o 5c o r 5d. S i n c e 5c i s much more s t a b l e t h a n 5d, o l e f i n i n s e r t i o n g i v i n g 5c would take p l a c e e x c l u s i v e l y . The i n s t a b i l i t y o f 5d w i t h an a p i c a l CI i s s i m i l a r t o t h a t o f 3d and 3e d i s c u s s e d above. The e t h y l group and t h e h y d r i d e i n 5c a r e c i s t o each o t h e r and thus r e d u c t i v e e l i m i n a t i o n might take p l a c e d i r e c t l y w i t h o u t i s o m e r i z a t i o n . However, r e d u c t i v e e l i m i n a t i o n from a d fivec o o r d i n a t e complex would f a v o r a t r a n s i t i o n s t a t e where t h r e e l i g a n d s b u t C H,. and H a r e i n the same p l a n e , as shown below.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch006
9
\
L
The r e a s o n f o r t h i s p r e f e r e n c e i s t h a t t h e d o n a t i o n and back d o n a t i o n between a deformed a l k a n e and a m e t a l fragment shown below i s e x p e c t e d t o f a c i l i t a t e easy bond exchange.
C (
^DËO ® ®
c l
donation
back-donation
Therefore, p r i o r to reductive e l i m i n a t i o n , i s o m e r i z a t i o n should t a k e p l a c e from 5c t o e t h y l h y d r i d e complexes w h i c h have H o r C^H, as an a p i c a l group, as shown below.
Et
Rh-
H
Rh-
y Et E t h y l m i g r a t i o n from 5c l e a d s t o 5c i t s e l f , and h y d r i d e m i g r a t i o n g i v e s 5d, an u n s t a b l e i n t e r m e d i a t e ; n e i t h e r o f t h e s e g i v e s a p i c a l H o r C^H^. The two r e m a i n i n g m i g r a t i o n s g i v e a more s t a b l e e t h y l h y d r i d e complexes and they have e i t h e r an a p i c a l H o r C^H^; CI m i g r a t i o n l e a d s t o s t a b l e 5e and PH~ m i g r a t i o n r e s u l t s i n 5a w i t h t r a n s p h o s p h i n e s , the i n t e r m e d i a t e or t h e H a l p e r n mechanism. The t r a n s i t i o n s t a t e f o r r e d u c t i v e e l i m i n a t i o n o f 5e t o g i v e l a has been determined, as shown i n F i g u r e 5, and i t has an a c t i v a t i o n energy o f 14.8 k c a l / m o l . 5a i s more s t a b l e t h a n 5e by 5 k c a l / m o l and the a c t i v a t i o n energy f o r r e d u c t i v e e l i m i n a t i o n from 5a t o r e g e n e r a t e 1 i s c a l c u l a t e d t o be about 15 k c a l / m o l , comparable w i t h t h e 5e-+la b a r r i e r . T h e r e f o r e , the system i s e x p e c t e d t o r e t u r n t o t h e
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
89
90
THE CHALLENGE OF d AND f ELECTRONS
H a l p e r n mechanism, i f i t i s n o t p r e v e n t e d f o r some r e a s o n steric).
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch006
Comparison between two
(eg.
mechanisms
I n the Brown mechanism, s e t t i n g a s i d e the h i g h energy r e q u i r e d f o r i s o m e r i z a t i o n from 3 t o 3b, the f i n a l s t e p o f the r e d u c t i v e e l i m i n a t i o n would r e q u i r e the h i g h e s t a c t i v a t i o n energy. O l e f i n i n s e r t i o n , the r a t e - d e t e r m i n i n g s t e p i n the H a l p e r n mechanism, i s an easy p r o c e s s i n t h i s mechanism. The k i n e t i c s o f the Brown mechanism i s thus e x p e c t e d t o be c o m p l e t e l y d i f f e r e n t from t h a t o f t h e H a l p e r n mechanism. T h e r e f o r e , i n the c a s e s i n which o l e f i n i n s e r t i o n has been f o u n d t o be r a t e - d e t e r m i n i n g , the H a l p e r n mechanism i s c l e a r l y more c o n s i s t e n t and a c c e p t a b l e . The above f e a t u r e o f the Brown mechanism t h a t r e d u c t i v e e l i m i n a t i o n i s more d i f f i c u l t t h a n o l e f i n i n s e r t i o n may be r e l a t e d t o the n a t u r e o f c a t a l y s t h a v i n g a c h e l a t i n g b i d e n t a t e l i g a n d such as DIPHOS. H a l p e r n have a l s o i n v e s t i g a t e d the hydrogénation ( E q u a t i o n 3) ( 2 ) , where i s o m e r i z a t i o n from the t r a n s - t o c i s b i p h o s p h i n e complex i s n o t n e c e s s a r y .
( s )
Rh' *
S ( S )
ï + COOCH, > PnCH,CH
(3)
NHCOCHj
They have f o u n d t h a t a t the low temperature, r e d u c t i v e e l i m i n a t i o n i s r a t e - l i m i t i n g (ΔΗ -17.0 k c a l / m o l - 4 0 ° C ) . The p r e s e n t c a l c u l a t i o n s u s i n g PH^ as the phosphine and e t h y l e n e as the o l e f i n , however, does n o t e x c l u d e the p o s s i b i l i t y o f the c i s mechanism c o m p l e t e l y , s i n c e the s t e r i c f a c t o r has n o t been t a k e n i n t o a c c o u n t . The c i s mechanism might be a c c e s s i b l e i n the c a s e where o l e f i n i s too b u l k y t o c o o r d i n a t e t o the t r a n s phosphine complex. U s i n g the above c a l c u l a t i o n s as a g u i d e , h e r e we c o n s i d e r q u a l i t a t i v e l y what i s e x p e c t e d t o take p l a c e when a v e r y b u l k y o l e f i n i s h y d r o g e n a t e d . L e t us assume t h a t i n such a c a s e the e q u i l i b r i u m ( E q u a t i o n 1) g e n e r a t e s the c i s i n t e r m e d i a t e . Then, the r e a c t i o n r o u t e w i l l pass t h r o u g h 3b-+4a-»5c, each o f w h i c h i s s t e r i c a l l y n o t too crowded. I s o m e r i z a t i o n o f 5c i s n o t a l l o w e d t o l e a d t o 5a, w h i c h has the b u l k y a l k y l group c i s t o two p h o s p h i n e s and i s overcrowded. Thus i s o m e r i z a t i o n o f 5c has t o l e a d t o 5e, w h i c h i s i n t r i n s i c a l l y (without s t e r i c e f f e c t ) only s l i g h t l y l e s s s t a b l e but s t e r i c a l l y l e s s crowded t h a n 5a. R e d u c t i v e e l i m i n a t i o n o f 5e w i l l r e q u i r e an a c t i v a t i o n energy comparable t o t h a t o f 5a->l and g e n e r a t e cis-RhCl(PH ) , l a . 3
2
There a r e two p o s s i b i l i t i e s i n the r e a c t i o n s o f l a . The f i r s t i s t h a t l a i s o m e r i z e s t o 1 due t o the s t e r i c r e p u l s i o n and t h a t the same r e a c t i o n p a t h l-*2-*3->3b i s f o l l o w e d . The second p o s s i b i l i t y i s t h a t H^ o x i d a t i v e a d d i t i o n t o l a t a k e s p l a c e t o g i v e d i r e c t l y 3a and thus i n the subsequent c a t a l y t i c c y c l e s i n t e r m e d i a t e s always have c i s p h o s p h i n e s . O l e f i n c o o r d i n a t i o n t o 3a i s p r o h i b i t e d b e c a u s e o f the s t e r i c r e p u l s i o n between the b u l k y o l e f i n and two b u l k y phos p h i n e s c i s t o the o l e f i n . Thus 3a-»3b i s o m e r i z a t i o n has t o take p l a c e b e f o r e the c a t a l y t i c c y c l e p r o c e e d s .
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
6. KOGA & MOROKUMA
Potential Energy Surface of Olefin Hydrogénation
91
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch006
One can consider that the Halpern mechanism and the cis mech anism are two extremes. The Halpern mechanism is most widely accepted and our ab initio MO calculations support this from the point of view of intrinsic electronic energy. However, there may be cases where the steric effect overshadows the electronic effect. There may also be cases where both effects are important. It may be, as Collman et al. said, that "this multistep reaction is very complicated. Like a chameleon, the dominant reaction mechanism changes when the nature of the catalyst, the ligands, or the substrate is altered" (8). Conclusions In this work, we have compared the potential energy profiles of the model catalytic cycle of olefin hydrogénation by the Wilkinson catalyst between the Halpern and the Brown mechanisms. The former is a well-accepted mechanism in which all the intermediates have trans phosphines, while in the latter, proposed very recently, phosphines are located cis to each other to reduce the steric repulsion between bulky olefin and phosphines. Our ab initio calculations on a sterically unhindered model catalytic cycle have shown that the profile for the Halpern mechanism is smooth without too stable intermediates and too high activation barrier. On the other hand, the key cis dihydride intermediate in the cis mechanism is electron ically unstable and normally the sequence of elementary reactions would be broken. Possible sequences of reactions can be proposed from our calculation, if one assumes that steric effects of bulky olefin substituents prohibits some intermediates or reactions to be realized. Literature Cited. 1. Koga, N.; Morokuma, K. Tod.Phys.Organomet.Chem., in press. 2. (a) Koga, N.; Daniel, C.; Han, J.; Fu, X.Y.; Morokuma, K. J.Am.Chem.Soc., 1987, 109, 3455. (b) Daniel, C.; Koga, N.; Han, J.; Fu, X.Y.; Morokuma, K. J.Am.Chem.Soc., 1988, 110, 3773. 3. (a) Halpern, J.; Wong, C.S. J.Chem.Soc.Chem.Commun., 1973, 629. (b) Halpern, J. In Organotransition Metal Chemistry: Ishii, Y.; Tsutsui, M., Eds.; Plenum: New York, 1975; p109. (c) Halpern, J.; Okamoto, T.; Zakhariev, A. J.Mol.Catal., 1976, 2, 65. 4. Brown, J.M.; Evans, P.L.; Lucy, A.R. J . Chem. Soc. Perkin Trans. II, 1987, 1589. 5. (a) Binkley, J.S.; Pople, J.A.; Hehre, W.J. J.Am.Chem.Soc., 1980, 102, 939. (b) Hehre, W.J.; Stewart, R.F.; Pople, J.A. J.Chem.Phys., 1969, 51, 2657. (c) Ditchfield, R.; Hehre, W.J.; Pople, J.A. J.Chem. Phys., 1971, 54, 724. (d) Huzinaga, S.; Andzelm, J.; Kłobukowski, M.; Radzio-Andzelm, E.; Sakai, Y.; Tatewaki, H. Gaussian Basis Sets for Molecular Calculations: Elsevier: Amsterdam, 1984. 6. Hay, P.J.; Wadt, W.R. J . Chem. Phys., 1985, 82, 270. 7. Chan, A.S.C.; Halpern, J . J.Am.Chem.Soc.. 1980, 102, 838. 8. Collman, J.P.; Hegedus, L.S.; Norton J.R.; Finke, R.G. Principles and Applications of Organotransition Metal Chemistry: University Science Books: Mill Valley, 1987; p.535. RECEIVED
January 18, 1989
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Chapter 7
Ab Initio Studies of Transition Metal Dihydrogen Chemistry Edward M . Kober and P. Jeffrey Hay
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch007
Los Alamos National Laboratory, Los Alamos, NM 87545
Examples of transition metal complexes containing dihydrogen ligands are investigated using ab initio electronic structure calculations employing effective core potentials. Calculated geometrical structures and relative energies of various forms of WL (H ) complexes (L = CO, PR ) are reported, and the influence of the ligand on the relative stabilities of the dihydrogen and dihydride forms is studied. The possible intramolecular mechanisms for H - D scrambling involving Cr(CO) (H ) are investigated by examining the structures and energies of various polyhydride species. 5
2
3
4
2 2
The recent discoveries of a new class of metal complexes involving molecular hydrogen has spawned numerous experimental(l,2) and theoretical (2—5) investigations to understand the bonding and reactivity of these systems. In this paper we discuss recent theoretical calculations using ab initio electronic structure techniques. In the first part of this article we review calculations on W dihydrogen species of d W complexes with emphasis on predictions of structures and energies of various chemical forms and on comparisons with available experimental information. In the second part a mechanistic problem involving scrambling of H / D mixtures to HD by Cr dihydrogen complexes is addressed. In this section we hope to illustrate the role of for theory in helping to distinguish between various reaction mechanisms in transition metal chemistry. Before proceeding to the specific examples of dihydroge ?hfmistry, it is worthwhile to summarize the particular challenges transition metal and actinide compounds present to this type of approach. There is a striking contrast to most compounds of first— and second—row main—group elements where reasonably accurate bond lengths and bond ancles of stable species can be predicted at the SCF Hartree-Fock level with small basis sets and thermochemical quantities can be computed with reasonable accuracy by perturbative techniques to electron correlation (fi). Accurate studies of molecular properties or transition states of chemical reactions are feasible using multi-configuration SCF ( M C - S C F ) and configuration interaction (CI) techniques. In contrast, for transition metal compounds one encounters a e
2
2
0097-6156/89/0394-O092$06.00/0 c 1989 American Chemical Society
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
7. KOBER & HAY
93
Transition Metal Dihydrogen Chemistry
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch007
situation where metal ligand distances are predicted to be 0.05-0.25 Â too long at the SCF level of calculation even with very accurate basis sets (Ζ-1Ω) and such approaches as Moller—Plesset perturbation theory can be unreliable for treating electron correlation effects (11,12). The relatively poor description of the ground state by a single configuration and the presence of numerous low-lying electronic states with very different electron correlation energies often requires sophisticated M C - S C F and CI treatments to obtain reliable molecular geometries and bond energies (13). In addition there are additional challenges arising from the sheer number of electrons in transition metal compounds and the relativistic effects which become increasingly important in second— and third—transition series compounds. These latter difficulties have been largely circumvented by the development of relativistic effective core potentials(li) to replace the chemically inert core electrons and to incorporate the relativistic effects on the valence electrons into the effective potential.
Tungsten Dihydrogen Complexes In the molecular dihydrogen species W(CO)3(PR3)2(H2) first characterized by Kubas et al. in both X - r a y and neutron diffraction studies (1,2), the dihydrogen is bonded in τ/ sideways fashion to the d metal center. Over 50 compounds involving many of the transition metals have since been synthesized by various workers. In addition, reanalysis(15) of existing hydrides such as FeH4(PR ) have now been found to be formulated as molecular hydrogen complexes, i.e., as Fe(H )(H)2(PR3)3. In this section we review briefly our previous theoretical calculations(5) on two prototypical dihydrogen complexes W(CO)3(PH ) (H ) and W f P H s M ^ ) . The calculations employed a relativistic effective core potential ( E C P - 1 ) to replace the inner [Xe] (4f ) core on W and a nonrelativistic E C P on Ρ with a flexible gaussian basis to describe the valence electrons of the system. Details of the calculation are given in Ref. 5. Structures 1—3 in Fig. 1 exemplify the modes of H bonding to a W ( C O ) ( P H ) fragment: two sideways bonded (rçS-coordinated) forms (1 and 2) and the end-on bonded (^Coordinated) form (3). Using a rigid W ( C O ) ( P H ^ fragment, the geometries of these three forms of H coordination nave been optimized using Hartree—Fock wave functions. The sideways bonded species (Table I) are found to be stable with respect to the fragments 4 by 17 kcal/mol and more stable than the end-on form, which is bound by only 10 kcal/mol. Little difference in energy is observed between the two-sideways bonded form with the H axis parallel either to the P - W - P axis or to the C—W-C axis. The former orientation is slightly favored, leading to a rotational barrier about the midpoint of the W - H bond of 0.3 kcal/mol. The calculated structure of the lower energy η form (1) shows a slight lengthening (from 0.74 to 0.796 À) of the H - H bond from uncomplexed H with a W—Η distance of 2.15 A . Recent low—temperature neutron diffraction studies show two equal W—Η bonds (1.89 ± 0.01 Â) and with the H lying exactly parallel to the P—W-P axis as predicted by the present calculations and having a Η—Η separation of 0.82 ± 0.01 Â. Although the calculations have correctly described the preferred mode of H binding, there remain some quantitative differences (Table I) between the theoretical and observed bond lengths. These differences are reduced considerably when one employs an effective core potential ( E C P - 2 ) which also treats the outermost 5s and 5p core orbitals of W as valence orbitals. The W - H and Η—Η distances are now calculated to be 1.93 and 0.81 Â, respectively, in much better agreement with the neutron diffraction results. 2
6
3
3
2
3
2
2
14
2
3
3
2
3
3
2
2
2
2
2
2
2
2
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
94
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch007
THE CHALLENGE OF d AND f ELECTRONS
Fig. 1.
Structural forms of W(CO) (PR3)2(H ) species. 3
2
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
7. KOBER & HAY
95
Transition Metal Dihydrogen Chemistry
Table I. Calculated and Experimental W - H and H - H Bond Lengths and Rotational Barriers for W(CO)3(PR3)2(H ) Species
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch007
2
Method
R
R(W-H),À
R(H-H),A
Rotational barrier, kcal/mole
SCFcalc. ( E C P - 1 on W)
H
2.153
0.796
0.3
SCFcalc ( E C P - 2 on W)
H
1.932
0.806
1.3
Low-temp X - r a y diffraction
i-Pr
1.95*0.23
0.75*0.16
Low—temp i-Pr neutron diffraction
1.89*0.01
0.82*0.01
2.4
Examination of the Mulliken population analyses for the fragment and the η complex reveals an overall increase of 0.12 e on the W atom upon complexation and the total charge on each hydrogen has decreased slightly from 1.00 to 0.98 e. The σ-bonding orbitals of the W atom (6s, 6p , and 5 d | show a net increase of 0.13 e, while the τ—bonding orbit sus (6p and 5d ) undergo a net loss of 0.03 e. Although the other five ligands also influence the amount of charge on the metal, the above trends are consistent with a mechanism involving some σ-donation from the H2 ligand and a lesser degree of π—back—donation from the metal. Of the possible seven-coordinate dihydrides let us consider the least motion reaction in which the two W—L bonds originally parallel to the H—H axis bend back as two W - H bonds are formed. The energies of these species are compared with the r/S-dihydrogen forms and the fragments in Fig. 2. Both seven-coordinate dihydride species lie higher in energy than the dihydrogen from (17 and 11 kcal/mol, respectively) and are only slightly bound compared to W L + H . For the case of having all P H ligands in the W ( P H ) ( H ) complexes, a much different situation prevails concerning the oxidative addition reaction. In contrast to W ( C O W P H ) ( H ) , the seven-coordinate dihydride W ( P H ) ( H ) lies 3 kcal/mol below the η complex! Replacing the CO ligands by PR3 groups favors the oxidative addition reaction proceeding to completion rather than being arrested in the ^dihydrogen stage. This preference for ^Coordination in W(CO)3(PH ) (H) correlates with the overall stabilization of the 5d orbitals, and the 5 d orbital in particular, by the back—bonding CO ligands. When these ligands are replaced by the less stabilizing PH3 groups, the dihydride is the most favored form. This is consistent with experimental observations that the dihydrogen-dihydride equilibrium can be shifted depending on the basicity of the ligands. For example, in a series of complexes, M o ( C O ) ( R P C H P R ) H , the coordination mode changes from dihydrogen from R = F h to dihydride for the more basic R=Et(2). 2
2
z
y
5
2
3
3
2
3
5
2
2
2
3
5
2
3
2
2
xz
2
2
4
2
2
2
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
z
yz
THE CHALLENGE OF d AND f ELECTRONS
V
I/'
o OC
XJ
I
°0. oc υ
H
oc
w
o
20
W + H 1
y
0°
L
ù
(3 (ô) FRAGMENTS L=PR
(0)
LU
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch007
(-7)
O υ LU CE
3
(-17)
7] -DIHYDROGEN COMPLEX 2
OCo o DIHYDRIDE COMPLEXES
1
I/
-W + H, υ
20
H
ce
o
(0)
LU Ζ LU
S-
20
(-19)
(-21)
(-33) L
FRAGMENTS L-PRo
η DIHYDROGEN COMPLEX 2
DIHYDRIDE COMPLEXES Fig. 2
Relative energies in kcal/mole of W ( C O ) ( P H ) 2 ( H 2 ) species (above) compared to W ( P H ) ( H ) species (below). 3
3
5
3
2
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
7.
KOBER & HAY
97
Transition Metal Dihydrogen Chemistry
H - D Exchange Involving CrfCO^Dihydrogen Species Background. The gas phase reaction H
+ D -ι 2HD
2
2
occurs only under severe conditions such as in shock tubes with an activation energy (~100 kcal/mole) comparable to the H—H bond energy. In fact, the kinetics have been interpreted in terms of a free radical mechanism involving H atoms rather than the bimolecular process indicated in the above equation. By contrast, several cases of facile H2—D exchange have been observed under thermal or low temperature conditions involving dihydrogen complexes 2
L M H + D -» L MHD + HD η ζ ζ η Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch007
0
0
Upmacis, Poliakoff and Turner(l£) observed HD exchange for mixtures of Cr(CO)4(H )2 and D but interestingly nol for mixtures ot Cr(CO)5(H ) and D . Kubas et al observe a similar phenomenon where HD is produced from reacting W(CO)3(PR )2(H ) with D either in solution or in the solid state. In addition other cases of H - D scrambling occur readily with metal hydride complexes(U) as in the case of Cp*ScH or Cp2ZrH where Cp* = 2
2
2
2
3
2
2
CsMe . Since the work of Upmacis et al. on Cr(CC>4)(H2)2 complexes is suggestive of an intramolecular mechanism, Burdett et α/.(1£) have examined various polyhydride structures as possible intermediates in this process using extended Huckel theory. These studies have led us to pursue ab initio electronic structure calculations of Cr(CO)4(H )2 species and possible mechanisms leading to H2—D exchange. Implicit in these studies is the assumption that there is rapid equilibrium between 5
2
2
Cr(CO) (H ) + D - Cr(CO) (H )(D ) + H 4
2
2
2
4
2
2
2
which subsequently undergoes intramolecular exchange C r ( C O ) ( H ) ( D ) -, C r ( C O ) ( H D ) 4
2
2
4
2
although this is only inferred from the experimental studies. What is actually observed is Cr(CO) (HD) formation in a mixture of Cr(CO) (D ) and Cr(CO) (D )2 when reacted with H . 5
4
5
2
2
2
Results of ab initio calculations. The calculations have been carried out on stable structures of Cr(CO)4(H )2 or its fragments at the Hartree—Fock level, where the structures have been optimized using gradient techniques with the modified GAUSSIAN82(I£) or the MESA(2Q) electronic structure codes. An effective core potential was used(H) to replace the [Ne] core of Cr with a [3s 3p 2d] contracted Gaussian basis to describe the 3s, 4s, 3p, 4p and 3d orbitals. A flexible [3s lp] bais was used for hydrogen and an S T O - 3 G basis was employed for C and 0 . (Some results are presented using an unpolarized (3s) hydrogen basis.) Of the possible forms for the parent molecule cis—Cr(CO)4(H2)2 the lowest geometrical structure is found to have the H2 molecules oriented in an 2
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
98
THE CHALLENGE OF d AND f ELECTRONS
upright position relative to the equatorial plane (Fig. 3). The structure with both H lieands lying in the equatorial plane is 3.1 kcal/mole higher in energy, corresponding to a rotational barrier of 1.5 kcal/mole for the rotation of one H about the metal-H2 bond. Removal of one H 2 to form Cr(CO)4(H ) requires 15 kcal/mole and removal of the second H requires another 14 kcal/mole. At this level of calculation the dihydrogen-dihydride energies 2
2
2
2
Cr(CO) (H ) - Cr(CO) (H) 4
2
4
2
are comparable with the dihydride lower by 1 kcal/mole. The calculated C r - H and H - H bond lengths are 1.787 and 0.77 A respectively for the upright form. (Fig. 4). Some of the possible polyhydride forms—having either a square H 4 or a H 3 — H " species coordinated to Cr(CO)4—are found to be very high in energy (at least 50 kcal/mole) and hence are unlikely intermediates in the H 2 - D exchange reaction. Another pathway is shown schematically in Fig. 5, where the conversion of the bis—dihydrogen complex to a dihydrogen-dihydride complex. The process is symmetry allowed in the sense that the three relevant orbitals in the equatorial plane, the Cr d . and the two H σ orbitals of the bis dihydrogen species transform into the C r - H σ and H σ bonds of the dihydrogen dihydride species. The diagram is oversimplified in the sense that the orbital characters change qualitatively through the course of the reaction. The d - orbital is empty in the dihydrogen dihydride complex and the H bonding orbital actually descends from higher orbitals in the bis—dihydrogen complex. The dihydride form is calculated to lie 10 kcal/mole higher in energy with a C r - H and C r - H bond lengths of 1.729 and 1.923 Â, respectively. The transition state for the reaction 2
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch007
4
2
2
x
2
y
2
2
2
x
2
y
2
2
Cr(CO) (H ) HCr(CO) 4
2
2
4
H (H) 2
2
was located at the SCF level assuming C symmetry and treating the H - C r - H bond angle between the two central Η atoms as the reaction coordinate. A n activation barrier of 24 kcal/mole (see Fig. 6) was found for this process—a relatively low barrier compared to some of the other polyhydride species. A non—C2V—pathway was also investigated where a similar barrier was also found. Because of the electron correlation effects can have a significant effect on calculated energies for chemical reactions, calculations on several of the above C r f C O ) ( H 2 ) species were carried out using configuration interaction (CI) techniques. These particular calculations consisted of single and double excitations (SDCI) with respect to the single Hartree—Fock configuration employed in the above studies. The optimized geometries from the SCF calculations were used for the respective species. No excitations were allowed from the inner 16 orbitals corresponding to the carbon Is, oxygen Is and 2s, and chromium 3s and 3p core orbitals. This resulted in 140,642 spin eigenfunctions in C 2 symmetry for the SDCI calculations. 2 v
4
2
V
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
KOBER & HAY
Transition Metal Dihydrogen Chemistry
Cr (CO) (H ) 4
2
2
STRUCTURES AND FRAGMENTS E(kcal/mole)
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch007
....
Λ
I
C rr" '
0.0
-..jA
+3.1
Cr
Cr
+H
+15.1
2
Η Cr^.
+H
+14.4
2
Η
Cr
Fig. 3.
+2H
2
+29.4
Relative energies in kcal/mole of Cr(CO)4(H )2 species. 2
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
THE CHALLENGE OF d AND f ELECTRONS
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch007
OPTIMIZED GEOMETRICAL
PARAMETERS
Bond Lengths (A) Cr - H
2
H - H
1.787
1.772
1.923
0.772
0.779
0.764
Cr - H,
1.729
Cr - C,
1.997
1.980
2.084
Cr - C
1.971
1.992
2.077
Ci - 0 ,
1.147
1.148
1.143
ci - o
1.150
1.146
1.142
2
2
2
Bond Anales idea) 46.9
44.9
42.1
2
90.8
93.6
94.2
«3
90.3
88.9
67.4
i
a
tt
. 4.
Calculated structural parameters for Cr(CO)4(H )2 species. 2
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
7. KOBER & HAY
Transition Metal Dihydrogen Chemistry
CO
oc
H /
\
101
H
•
Cr Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch007
/ I V CO
Fig. 5.
OC
1
CO
Η
Correlation diagram for bis—dihydrogen to dihydrogen-dihydride forms of Cr(CO) (H2)2 species. 4
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch007
102
Fig. 6.
THE CHALLENGE OF d AND f ELECTRONS
Calculated energies from Hartree-Fock (SCF) and configuration interaction (CI) calculations for one possible path for H - D exchange involving Cr(CO)4(H )2. 2
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
7. KOBER & HAY
103
Transition Metal Dihydrogen Chemistry Table II. Relative Energies (kcal/mole) for Cr(CO) (H )2 and Related Species 2
4
SCE CrfCOVH )2—upright Cr(CO) (H } —coplanar Cr(CO) (H )—coplanar Cr(CO) H (H ) Cr(CO) H + H 2
4
2
4
4
2
4
2
0.0 3.1b 22.5 29.7 43.3
0.0 3.1* 22.5 11.8 14.4
2
4
2
2
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch007
*E = -532.823307 a.u. bE = -533.416149 a.u.
The CI results are contrasted with the SCF calculations in Fig. 6 and in Table II. The bis-dihydrogen complex C r ( C O ) ( H ) is 27 kcal/mole and 40 kcal/mole more stable, respectively, than the dihydrogen-dihydride species C r ( C O ) H ( H ) and the C r ( C O ) H + H fragment, respectively. These energies compare to 9 and 11 kcal/mole for the respective SCF calculations. The potential energy surface has also changed in that the C r ( C O ) ( H ) intermediate appeared to be a transition state at the SCF level, but actually lies below the energy of the C r ( C O ) ( H ) product at the CI level. Finally we compare the results ot the above ab initio calculations with the earlier extended Huckel theory (EHT) calculations of Burdett et al (IS). In this work, which helped to stimulate our own calculations, the authors emphasized that they were probing general trends and that reliable thermodynamic stabilities of M H species could not be obtained using this method. With these points in mind we compare the relative energies of tetrahydrogen species of C r ( C O ) in Table III. 4
4
2
2
4
2
2
2
2
4
4
2
2
n
4
Table III. Comparison of extended Huckel theory (EHT) results (lfi) with SCF ab initio calculations on C r ( C O ) tetranydrogen species 4
Rel. energy (kcal/mole) EHT SCF
C r ( C O ) fragment 4
ds-dihydrogen dihydrogen-ndihydride dihydride + H dihydrogen + H planar H linear H 3 triangular H 3 tetrahydride square H tetrahedral H
0
2
2
4
4
4
17 22 35 36 86 160
0 12 14 15 23 22 76 65 54
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
4
104
THE CHALLENGE OF d AND f ELECTRONS
While the earlier study did not investigate the H ( H ) species, the other intermediates involving "open " planar H3 and H4 moieties are actually placed at similar energies, while the intermediates involving closed species do not correspond with our findings. In particular, the (H )(H) species previously described as having two—electron triangular H3* and H ' ligands lies at considerably higher energy. There has, in fact, been considerable experimental activity to identify and isolate complexes containing H and H " ligands, but we find no evidence to support these forms for this particular class of complexes. 2
2
3
+
3
3
Discussion of Mechanisms. In summary, the above calculations have identified relatively low-energy (i.e., less than 30 kcal/mole) pathways for the conversion of H 2 - D 2 into HD as exemplified by the dihydrogen—dihydride species originally formed according to the process C r ( C O ) ( H ) ( D ) 1 Cr(CO) (H)(HD)(D) Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch007
4
2
2
4
as shown in Figs. 5 and 6. Once the incipient HD bond has begun to form, the reaction could proceed further by (a) rotation of HD about the C r - H D bond followed by collapse to Cr(CO) (HD) , (b) dissociation of HD to form Cr(CO)4(H)(D) followed by insertion of H or D , or (c) dissociation of HD followed by collapse to Cr(CO)4(HD), to mention some of the possibilities. Despite the above low-energy pathways discussed above, it is not clear whether they are actually involved in the liquid xenon experiments of Upmacis et al since relatively small barriers must be involved for any process at these temperatures. We would point out that the possibility of radical processes involving the presence of H atoms should also be examined thoroughly before these mechanisms are definitively understood. We observe that scrambling involving coordinated H and H atoms is much more facile than processes involving two H ligands. In cases such as 4
2
2
2
2
2
Cr(H )(H)(H) - C r ( H — H — H ) ( H ) -+ Cr(H)(H )(H) 2
2
and V ( C O ) ( H ) ( H ) -, V ( C O ) ( H — H — H ) -, V ( C O ) ( H ) ( H ) 5
2
5
5
2
the barriers involved in open H intermediates are only 10-13 kcal/mole. A second possibility to be considered is that the Cr(CO)4H dihydride species actually possesses a triplet ground state much lower in energy than the singlet species discussed in Fig. 5 and Table II. Such a species could react with H to form H—atom containing species. In summary, the nature of some likely reaction intermediates involved in HD formation from H and D complexes of Cr(CO)4 have been identified here. Unraveling the further details of the mechanisms involved in these fascinating complexes will require more extensive experimental and theoretical studies. 3
2
2
2
2
Acknowledgment This work was carried out under the auspices of the U.S. Department of Energy.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
7. K O B E R & I I A Y
Transition Metal Dihydrogen Chemistry
105
Literature Cited
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch007
1.
Kubas, G.J; Ryan, R.R.; Swanson, B.I.; Vergamini, P.J.; Wasserman, H. J. Am. Chem. Soc. 1984, 106, 451. (b) Kubas, G.J.; Ryan, R.R.; Wrobleski, D. ibid. 1986, 108, 1339. (c) Kubas, G.J.; Unkefer, G.J.; Swanson, B.I.; Fukishima, E. ibid. 1986, 108, 7000. 2. Kubas, G.J. Acc. Chem. Res. 1988, 21, 120 and references therein. 3. Saillard, J.-Y.; Hoffmann, R. J. A m . Chem. Soc. 1984, 106, 2006. 4. Jean, Y . ; Eisenstein, O., Volatron, F.; Maouche, B.; Sefta, F. J . Am. Chem. Soc. 1986, 108, 6587. 5. Hay, P . J . J . A m . Chem. Soc. 1987, 109, 705. 6. Hehre, W . J . ; Radom, L.; Schleyer, Paul v.R.; Pople, J . Α. Ab Initio Molecular Orbital Theory. Wiley: New York, 1986. 7. Spangler, D.; Wendoloski, J.L.; Dupuis, M . ; Chen M . M . L . ; Schaefer III, H.F. J. A m . Chem. Soc. 1981, 103, 3985. 8. Faegri, K . ; Almlof, J . Chem. Phys. Lett. 1984, 107 121. 9. Dobbs, K . D . ; Hehre, W . J . J . Computational Chem. 1987, 8, 861. 10. Williamson, R . L . ; Hall, M . B . Int. J. Quantum Chem. Symp. 1987, 21, 503. 11. Rohlfing, C . M . ; Martin, R.L. Chem. Phys. Lett. 1985, 115,104. 12. Rohfling, C . M . ; Hay, P . J . J . Chem. Phys. 1985, 83 4641. 13. Bauschlicher, Jr., C.W.; Walch, S.P.; Langhoff, S.R. Quantum Chemistry: The Challenge of Transition Metals and Coordination Chemistry; Veillard, Α., Ed.; Reidel: Dordrecht, Holland, 1985; p. 15. 14. (a) Hay, P.J.; Wadt, W.R. J. Chem. Phys. 1985, 82, 270. (b) Wadt, W. R.; Hay, P . J . ibid. 1985, 82, 274. (c) Hay P. J.; Wadt, W . R . ibid. 1985, 82, 299. 15. Crabtree, R.H.; Hamilton, D . G . J . A m . Chem. Soc. 1986, 108, 3124. 16. Upmacis, R . K . ; Poliakoff, M . ; Turner, J.J. J . A m . Chem. Soc. 1986, 108, 3645. 17. Thompson, M.E.; Baxter, S.M.; Bulls, A.R.; Burger, B . J . ; Nolan, M . C . ; Santarsiero, B.D.; Schaefer, W.P.; Bercaw, J.E. J . A m . Chem. Soc. 1987, 109, 203. 18. Burdett, J . K . ; Phillips, J.R.; Pourian, M . ; Upmacis, R. Inorg. Chem. 1987, 26, 3061. 19. Modified GAUSSIAN82 Program: J.S. Binkley and R.L. Martin. 20. M E S A program: P.W. Saxe and R.L. Martin. RECEIVED December 9, 1988
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Chapter 8
Activation of Small Molecules by Transition Metal Atoms Theoretical Interpretation of Low-Temperature Experiments with Cu, Pd, and Pt Atoms
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch008
O. A. Novaro Instituto de Física, Universidad Nacional Autónoma de México, 01000 México D.F., Mexico Theoretical-experimental results on transition metal atom-small mole cules systems are reported. The theoretical studies employ the PSHONDO-CIPSI sequence of programs that allow for variational and perturbational configuration-interaction calculations including up to 10 configurations for each ground and excited-states of the system. These theoretical results are contrasted with data from low-temperature ma trix isolation experiments on these same systems supported by infrared, visible-ultraviolet, epr and other spectroscopic techniques. Interesting correlations between theory and experiment are found, including the fol lowing: the photoactivation of H and methane by Cu atoms at low tem peratures are rationalized from a theoretical standpoint; the theoretical prediction of H activation by ground-state Palladium is verified exper imentally and the preference of insertion over abstraction reactions and the formation of Cu(N ) complexes serve to explain some extraordinary isotopic effects found in experiment. 6
2
2
2 n
Matrix isolation experimental techniques [1-10] stand out among many other mod ern chemical research methods with regard to their ability to provide direct comparisons with quantum mechanical calculations. The use of photoexcitation methods to induce reactions [7-9] as well as the applications of multiple spectroscopic techniques to study such photochemical reactions allows for close control of the reaction parameters. Most of the high temperature and entropy effects, otherwise very large in thermochemical re actions, are therefore not present here and thus some of the limitations associated with applications of precise quantum mechanical calculations to kinetic processes disappear. Specifically the low temperature studies which concern elementary interactions of small molecules and transition metal clusters or atoms isolated over "inert" solid matri ces [5-10] are of high interest, especially now that the Schrodinger equation representing such interactions can be solved to relatively high precision using ab-initio configurationinteraction methods. Among such methods we could mention the CASSCF and CCI, GVB-CI, Monstergauss, PSHONDO and CIPSI [11-15] among other methods and pro grams, many of them mentioned and described in this book. The fact is that theoretical physicists and chemists in the recent past have developed very accurate methods for the study of d- and /-electron systems. Therefore, while several low temperature exper iments concerning transition metal atoms or clusters and their interactions with small molecules have appeared in the literature, simultaneously many quantum mechanical calculations are appearing on the same type of systems. Rarely in the history of quan0097-6156/89/0394-O106$06.00/0 c 1989 American Chemical Society
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch008
8. NOVARO
Activation of Small Molecules by Transition Metal Atoms 107
turn chemistry has a situation so tempting been found where theory can be compared with experiments concerning systems of great practical chemical interest and yet small enough and with such strict control of variables as to make direct comparisons feasible. However, actual theoretical experimental collaborations are not at all common. Some joint papers exist [16-18] and sometimes theoretical and a experimental papers on the same system are published back to back [19-20]. An example is that of Weltner and coworkers, who performed esr and endor studies of MnH supported over solid Argon at 4K [21] in order to compare with previous calculations on the same MnH molecules [22], among other similar efforts. The alternative situation of theoreticians directing their calculations of previous or current matrix isolation experiments is of course common also [22-31] but considering the potential benefits of following the experiments closely, we feel that in reality much more work should be done in this direction. In particular we believe that in order to be really relevant to the understanding of the kinetics, calcula tions should not only aim at obtaining the structure of a molecule but very importantly at determining reaction coordinates and activation barriers. This allows for predictions about feasibility and selectivity that may be contrasted with the experiments. In this chapter we shall review a few cases of such theoretical-experimental collaborations with out pretending to be exhaustive but rather as examples of the mutual influence of theory and experiment. Method The method used in our calculations is the ab-initio pseudopotential method of Durand et al. [32-34]. We apply it for the Cu, Pd and Pt metal atoms whose pseudopotentials are also given in the literature: that of Cu in [2£], that of Pd in [M], that of Pt in [3fi]. In every case all of the valence electrons as well as all the electrons of the outermost d-subshell are always treated explicitely and without restrictions. The basis sets used are always of double-zeta quality at least, those for Cu and H are given in [3Z], that of Pt in [3£], that of C by Pacchioni et al. [2£], that of Pd in [24], that of Ν by Daudey (Daudey, J.P. Preprint of the Laboratoire de Physique Quantique, Université Paul Sabatier, Toulouse, France, 1986). The convergence criterion of the SCF iteration energies was set at 10~ . Basis set superposition errors were systematically tested and corrected for by following the counterpoise correction of Kolos [22] when necessary. The basis sets were selected by thoroughly testing their accuracy in the reproduction of the energy splittings between the ground and lowest excited states (eg. the P- S and D- S splittings in Cu, the D- S splitting in Pt, etc.) but this is better seen by reading the original papers [36-37]. One limitation of these basis sets is the lack of /-polarization functions. In some instances it has been shown that their use introduces only marginal improvements in some calculations involving Cu [4Ω] or Pt [H]. The use of /-functions is however an open subject of current interest as is exemplified by several chapters of this book. We use the CIPSI algorithm [1£] which introduces configuration interaction by perturbation with multiconfigurational wave functions selected by an iterative process using the M0ller-Plesset barycentric values as proposed by Malrieu [42]. By this ap proach we then include many more configurations (typically of the order of a million or more) that interact effectively with the original reference states. These reference states correspond to more than one metal atom state, generally we take three or more states of each metal considered, say the ground and the lowest lying excited states. The im portance of this will be evident when discussing the comparison of the specific results with the experimental data. In every case a careful analysis of the configurations that are included in this large CI scheme is carried out trying to determine how their role during the process is to be understood. Such configurations represent very often a polarization of the d-subshell, which in many cases is closed or nearly closed, so that its relaxation substantially lowers the interaction energy between a transition metal atom and a small closed-shell molecule 6
2
3
2
l
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
2
2
108
THE CHALLENGE OF d AND f ELECTRONS
(eg. H2, C H or N2). As interesting as this and also many other trends shown by the configurations participating in the CI schemes are, it is not adequate to describe them within the overall fashion in which we shall discuss the results on several different systems in the present paper. The reader is therefore referred to the original articles to review this important aspect [36-37 40.43-44]. Comparisons with other theoretical methods are important. Our Cu calculations are based on the pseudopotential of Péllisier [35], who applied it to the C u system. A controversy arose when CASSCF-CCI calculations on the same system seemed to imply [4Q] that the CIPSI calculations matched the experimental energy too well for a non-relativistic method. Péllisier [45] replied by showing his pseudopotential included relativistic effects. On the other hand a recent CASSCF-CCI calculation on the Pt+H reaction was published [4£]. The author apparently was not aware of our previous work on the same system [25] and yet he obtained potential energy surfaces that are virtually identical to those obtained using CIPSI as is commented elsewhere [47]. He did make some comparisons with GVB-CI calculations [4£] on the same P t H systems concluding that both this method and his agreed well. From all this we must conclude that sometimes CIPSI, CASSCF-CCI and GVB calculations can lead to the same results and quite similar chemical pictures. Other methods also match well in their predictions. In one of our papers both the CIPSI and Monstergauss programs were used in the CuH-f H thermal reaction giving coinciding results in most aspects of the process [18]. 4
T
2
2
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch008
2
Activation of Small Molecules by Closed-Shell Transition Metal Atom States The matrix isolation experiments using epr, ir, uv-visible and other spectroscopic tech niques on transition metal-olefin complexes [8,49] have naturally attracted the attention of theoretical chemists and calculations on the Ni-C H4 system were reported in one of the first theoretical-experimental papers mentioned in the introduction [15]. These results were later supplemented with a larger (double-zeta) basis set [5Q] and also [51] extended for a N i ( C H ) system. The main conclusions are that a net charge transfer of almost 1/5 of an electron from the metal to the ethylene is evident and that a dona tion and back donation mechanism consistent with a classical Dewar-Chatt-Duncanson model exists. The Ni-ethylene binding energy is 12.8 kcal/mol. Another system that has been studied theoretically is Cu-C H4 where, in con trast [54], it was suggested that a weak charge transfer from the olefin to the metal (0.164e) without the participation of the carbons and the unpaired electron remaining in the 45 - 4p hybridized orbital exists. This is indeed very far from the Dewar-ChattDuncanson model [52-53]. We shall now report our results on Pd-C H4 which are much more in coincidence with those of Ozin et al. [Ifi] and Siegbahn and coworkers [50-51]. 2
2
4
2
2
2
Palladium-Ethvlene Interaction The case of the interaction of Pd with C H is interesting because the Palladium ground state has a closed shell, d , configuration. This notwithstanding, the existence of a stable P d C H complex was established experimentally [&] and a net charge transfer from Palladium to the olefin carbons was reported. They also showed that the πbonded Pd-C H4 complex had very similar stretching modes to those observed for ethylene adsorption on Palladium surfaces [55], thus concluding that the complex is an acceptable model for this adsorbed species. This system is then interesting enough to justify a theoretical study and this was done by us using the methods described above [56-57]. The main conclusions of the matrix isolation experiments were confirmed by us as depicted in Fig. 1 where the Pd-fC H interaction energy curve as well as the geometrical and charge transfer properties of the complex are given. The binding energy of our complex, 47 Κ Joules/mole, was close to the value (54 Κ Joules/mole) of the desorption energy on Pd surfaces reported in [55], fulfilling the expectations of Huber, 2
4
10
2
4
2
2
4
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
8. NOVARO
Activation of Small Molecules by Transition Metal Atoms 109
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch008
f Energy (Kcal/mol) 9,0 \-
Figure 1. Potential energy curve and geometrical and charge transfer parameters of the Pd-ethylene complex.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
110
THE CHALLENGE OF d AND f ELECTRONS
Ozin and Power [g] that the P d - C 2 H complex is an acceptable model of the adsorbed olefin species. We also see from Fig. 1 that the C-C distance is lengthened and the C-H2 planes are rotated by 15° to the original plane of the ethylene, thus explaining the red shifts in the respective vibrational modes reported in [fi]. Also, their ultraviolet results show a charge transfer from the metal to the olefin, which is also evident in Fig. 1. The value of this donation is similar to that reported for the N 1 - C 2 H 4 complex [52] and we also find that a reasonable consistency of the Dewar-Chatt-Duncanson model exists for Palladium as for Nickel [52] but apparently not for C u C H [54] although the fact that the latter study included the cf-electron subshell in the pseudopotential may completely falsify this aspect of their results. At least our own studies of Cu reactions, to be described later on, systematically showed the need of having a flexible and explicit description of the d-subshell in order to obtain the very important avoided crossings and activation barriers and in general the multiple-well potential energy surfaces that will be discussed later on. 4
2
4
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch008
Ho Capture by Palladium Atoms Several calculations on the P d H 2 system exist [58-60] showing that a weakly bound A i ( C 2 v ) Palladium dihydride can be stable. However no attempts to depart from this C2V symmetry were carried out except for some analyses of the H2 positions around a Palladium dimer [fil]. We have studied both the side-on ( C 2 ) and the headon approaches of Pd to H2 showing that both present attractive curves without any activation barriers [12]. More recently we have studied a substantial part of the potential energy surface of the Pd(4 Cu + H
(2)
2
which produced high yields of ground state Copper atoms and hydrogen. Theoretical calculations using both the Monstergauss and CIPSI programs showed an energetically downhill reaction coordinate for the H+CuH addition reaction [IS]. The addition re action implies a linear approach of the H atom towards the Cu moiety of CuH. Also an abstraction reaction was studied, which implied a linear approach of H now towards the H moiety of HCu. The latter process does have an activation barrier of less than 7 kcal/mol [IS]. The CuH complex formed by either of these reactions spontaneously dissociates into the final products (H and ground state Copper atoms) thus explaining the experimental results of [64]. 2
2
Photoexcited Cu Activation of Ho In their study of the activation of H by photoexcitation of Cu atoms deposited in rare gas matrices, Ozing and coworkers [65-66] irradiated with 320nm photons to produce the 3d 4p ( P) «- 3d 4s ( S) transition. This was sufficient and necessary for the reaction of eq. (1) to take place. The efficiency of this photochemical process was originally attributed by them [££] to the radiationless transition of the Ρ state to the lower excited D state which hypothetical^ was the one that reacted with H . Our first calculations showed that the Ρ state itself is also capable [37] of capturing H effectively. Furthermore it was established that while the photoactivation by the 320nm was a sine qua non condition for the process to be triggered (without it Cu simply does not activate H ) right after it the process of eq. (1) proceeds regardless of whether Cu* suffers a radiationless transition to the lower excited state D or eventually to the ground state S of Cu. This was demonstrated [4£] from the relatively moderate (~ 20 kcal/mol) activation barriers that the potential energy curves of Cu^S), Cu( D) and Cu( P) present to H capture. These barriers are easily overcome by the great energy gain from the transition from the Ρ state (whose own potential curve is initially downhill in energy) to the lower D and S states. This process implies a very interesting mechanism involving Landau-Zener-Stuckelberg transitions and Herzberg-Teller couplings between the main A\ and P symmetry potential energy surfaces of the C Cu*+H system. These potential energy surfaces [44] are reproduced in Fig. 6 but the details of the reaction mechanisms envolving the three S, P and D states of Copper in activating H and eventually leading from the CuH energy minima of Fig. 6 towards the final products of eq. (1) cannot, for reasons of space, be reproduced here. For this the reader is referred to the original papers [37 43-44]. It is worth mentioning here however that the restriction to C symmetry in Fig. 6 is not a limitation for the actual chemical process of eq. (1) because, as was shown in [43], any deviation from C symmetry would only enhance the scission of H and the liberation of the CuH-|-H products because the activation barrier (~ 20 kcal/mol) would necessarily be lowered even more considering that both the Αχ and the B surfaces belong to a single representation A' of the C group when C symmetry is broken [21]. In conclusion: the activation of H by Cu is only feasible if the Ρ fc(abstraction). This of course would be disregarding any tunneling effects of an H moiety into the abstraction pathway barrier. However, such tunneling would not be present if D is used in the reaction. At this stage in conclusion we would necessarily infer from our theoretical results (M.E. Ruiz; G.A. Ozin and 0. Novaro, work in progress) that this reaction would indeed lead to no products of the type CuD or D, in complete agreement with the observed facts. 2
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Activation of Small Molecules by Transition Metal Atoms 119
8. NOVARO
Nitrogen Matrix Hindrance of Ho Activation by Copper An amazing discovery (Gracie, C , M.Sc. Thesis, University of Toronto, 1985) was made when the photoexcitation reaction 2
Cu( P) + H + D ^ CuH + CuD + H + D 2
(6)
2
previously mentioned and which had a reaction rate ratio of about kjj/kj) ~ 1.5 [55] was carried out replacing the rare gas matrix by a N molecular matrix, with all other conditions (temperature, reactants, etc.) kept equal. The new results showed the reaction products 2
2
Cu( P) + H + D ^ CuH + H + D 2
2
(7)
2
with the CuD and D subproducts being much scarcer, by a factor of about one thousand (in effect, kjj/kj) « 10 ). The suggestion was made based on infrared, optical and epr spectroscopical observations (Ozin, G.A., Gracie, C. and Mattar, S.M., Toronto University, unpublished data, 1984) that a CuN complex existed of the type
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch008
3
2
Cu -
(8)
( Ν ΞΞ N ) „ 2
2
When a Copper atomic resonance line Ρ 1, i s a large z e r o - f i e l d s p l i t t i n g parameter (D) relative to X-band quanta; this usually occurs for molecules with large spin-orbit constants and/or through coupling to low-lying electronic states. (V , with a Σ ground state but D = 75 cm , i s therefore undetectable i n matrix ESR. ) These two conditions for not observing an ESR spectrum, even though there i s good reason to believe that a magnetic molecule has been matrix isolated (29), w i l l be used as circumstantial evidence supporting some of the ground states to be suggested here. q
2
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch015
2
3
2
-1
Transition-Metal Diatomic Molecules Figure 1 shows the array of a l l possible diatomics derived from the first-row t r a n s i t i o n metals (except Zn). Ground states are not r e l i a b l y derived from Mossbauer data (30, 31 ) so that those given for the seven molecules containing Fe must be considered as suspect. To an e a r l i e r version of this Figure (32 ) we have added a designation of the number of valence 3d + 4s electrons i n the isoelectronic molecules occurring i n blocks perpendicular to the p r i n c i p a l diagonal of homonuclear molecules. Thus, C r , VMn, TiFe, and ScCo a l l have 12 valence electrons. C r i s known experimentally (as a l l the bold borders indicate) to have a Σg ground state. Can we then presume that the other three isoelectronic diatomics also have that ground state? (TiFe and ScCo in that series have been "inferred from experiment" to be Χ Σ; that w i l l be discussed below.) Similar groups of isoelectronic molecules are l i s t e d i n Table I, including those involving Zn. 2
2
1
1
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
WELTNER & VAN ZEE
Diatomic and Monocarbonyl Molecules
215
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch015
TRANSITION METAL DIATOMICS
11 GAS PHASE ES ESR/MATRIX 0
MOSSBAUER/MATRIX 1 θ '
H INFERRED FROM EXPT. THEORY
/ \
VALENCE ELECTRONS
Figure 1. The ground states of possible diatomics formed from the first-row transition metals, excluding Zn. Those i n bold borders are d e f i n i t e l y established, the others are derived as indicated. "Inferred from experiment" and "valence electrons" are explained i n the text.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
216
THE CHALLENGE OF d AND f ELECTRONS Table I.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch015
Valence Electrons 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
First-Row Transit!on-Metal Diatomics
Experi mentally Studied Molecules Sc
Experi mental Ground State
Isoelectronic Molecules
0
T i
2 TiV
1 4
*î Σ 9
g Cr TiCo, ScNi Mn VNi 0
2
z
g
4
E
g
2
ScTi ScV ScCr TiCr,, ScMn VCr, TiMn, ScFe VMn, TiFet, ScCot CrMn , VFe CrFe \ VCot, T i N i t , ScCu MnFe , CrCo, TiCu, ScZn 2 ' MnCo*, CrNi*, VCu, TiZn FeCo MnNi*, VZn * Co , FeNi , MnCu CoNi. * , FeCu*, MnZn . * CoCu , FeZn 2' CoZn NiZn c
F e
CrCu CrZn
Λ
2
N i
NiCu Cu CuZn Zn 2
1r+
0
J Ground state inferred from experiment - see text. Not detected i n ESR but believed to have been prepared.
Table I I . Mixed-Row Transition-Metal Diatomics
Valence Electrons 13 15 17 18 19 20 21 22 23
Experi mentally Studied Molecules YNi, VPd, CrAg, MnAg
Expéri menta 1 Ground State
Isovalent Molecules ScPt, YPt VPt, NbPd, NbPt MnAu FeAg* CoAg NiAg*
CrCd
CuAg, CuCd, AgZn, AuZn,
In Table II our d e f i n i t i o n of "isoelectronic" has been broadened to just requiring the same number of d + s electrons (regardless of
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
15.
WELTNER & VAN ZEE
217
Diatomic and Monocarbonyl Molecules
p r i n c i p a l quantum numbers), so that mixed-row diatomic molecules can be included, and a more appropriate designation i s "isovalent". Table II l i s t s only experimentally-studied molecules and their "isovalent" counterparts, but there are, of course, a large number (465!) of possible transition-metal diatomics. [In these Tables an asterisk (*) indicates a molecule not observed i n the ESR but believed to have been prepared (29), while a dagger (t) indicates that the ground state of the molecule was "inferred from experi ment", as discussed below.] Four molecules i n column 4 of Table I, TiFe, ScCo with 12 valence electrons and VCo, TiNi with 14 valence electrons (see Figure 1) have been inferred to have Σ ground state (33 ). The reasoning i s that the addition or subtraction of one electron i n the dïr d6 da sa sa configuration of TiCo, ScNi would probably lead to closed s h e l l molecules. ScCu, although also i n that category, was considered more doubtful because of the reluctance of Cu to form multiple bonds (however, see below). As Figure 1 indicates, there are no experimental, or t h e o r e t i c a l , data on any of these putative Σ molecules, but the "inferences" are i n accord with the known ground states of C r and Mn . 1
4
4
2
2
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch015
2
1
2
2
Pursuing the "Isovalent" P r i n c i p l e . The question, admittedly naive, i s whether the "isoelectronic" p r i n c i p l e of simple MO theory applies among these electronically-complex transition-metal molecules. It seems that exact adherrence to such a rule might be surprising i n these cases, but i f there i s a commonality among the lowest-lying states, i t i s worth pursuing. We have already considered two first-row metal series contain ing 12 and 14 valence electrons where there are hints of adherence to the p r i n c i p l e . However, there i s a more d e f i n i t e example of "isoelectronic" behavior where the ground states of four 13-electron molecules have been established to be Σ · These molecules are also i n the class of what have been referred to, after Gingerich, et a l . (34-36), as Brewer-Engel molecules (37); each involves two elements from opposite ends of the periodic table, i n this case Groups IIIB and VIII. Such molecules are expected to form strong multiple bonds and therefore to be of low spin. ScNi, S c P d , YNi, Y Pd not only a l l have S = 1/2 but the unpaired spin has similar charac t e r i s t i c s in the four molecules, as derived from hyperfine i n t e r action with the indicated nuclei (38). The spin is largely (~70%) on the lighter atom and has about 30% s character throughout. Shim and Gingerich (36 ) have made an a l l - e l e c t r o n Hartree-Fock c a l c u l a tion for YPd and find a Δ ground state, not i n agreement with the ESR result, but Σ and Π states l i e only about 0.2 eV higher i n energy. It i s l i k e l y that improvement i n the calculation could change the ordering of these levels. The resulting bonding i s unusual (and reminiscent of that i n metal carbonyls) in that the dr i c h Pd donates electrons to the d-poor Y and Y back donates s electrons. Their s i m i l a r i t i e s indicate that this type of bonding prevails in a l l four molecules. Extension of the experimental studies to La and Pt would be interesting, and a more extensive ab i n i t i o calculation, perhaps on ScNi, would be worthwhile. More recently i n our laboratory, Cheeseman has extended the e a r l i e r study of VNi (39) with 15 electrons to VPd, VPt, and NbNi to find that they also have Σ ground states (40). 2
45
4 5
1 0 5
89
8 9
2
2
+
2
4
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
1 0 5
218
THE CHALLENGE OF d AND f ELECTRONS
"Isoelectronic" S c (Χ Σ") and Y are bothersome, since only the former has an X-band ESR spectrum (41, 42) and theory finds the same ground state for both (43, 44). Walch and Bauschlicher (44) suggest that a more extensive calculation may lower the stronglybonded Σ + state below the Σ ~ i n the case of Y , but ESR cannot distinguish between that choice and the p o s s i b i l i t y of large zfs i n a Σ ~ ground state. Surprisingly, the binding of Cu has been puzzling i n the two diatomics CrCu and MnCu. CrCu does not appear to have a Σ ground state, as do CrAg and CrAu, and after considerable agonizing, the spectrum has been interpreted as Σ (45). Thus i t i s probably t r i p l y bonded with properties intermediate between C r and Cu * An exceptionally large e l e c t r i c f i e l d gradient at the Cu nucleus i n CrCu supports i t s anomalous ground state. The situation i n the 18 valence electron series MnCu, MnAg, MnAu i s similar but d i f f e r e n t . Whereas MnAg i s easily formed and characterized as a Σ molecule, MnCu (and MnAu) remains undetected i n the ESR (29). Other evidence for a preference for Σ ground states among this isoelectronic class i s that CrZn also has Χ Σ (29); however Co , even after many t r i a l s , was not observed v i a ESR. In summarizing this section, one can say that the present meager experimental data are i n encouraging support of the applica tion of the "isoelectronic" p r i n c i p l e to these diatomics. Both first-row and/or mixed-row diatomics containing 12, 13, 14, 15, and 18 valence electrons show indications of having Σ , Σ , Σ, Σ , and 2
1
2
5
2
5
4
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch015
2
2
7
7
7
2
1
2
1
4
7
Σ ground states, respectively. At present there are no definite discrepancies except i n the CrCu, CrAg, CrAu s e r i e s . It i s clear that theory has much to contribute here. Transition-Metal Monocarbonyls,
M-CO
Molecules
The experimental ground states of most of these molecules are un known. ESR has established VCO as Σ and CuCO as Σ , and recently also ScCO as Σ , but i t has f a i l e d to detect FeCO, which theory i n dicates has a Σ ~ (or Σ~) ground state. CrCO(^) has also appar ently been detected via ESR. The present state of a f f a i r s i s sum marized simply i n Figure 2, where the o r i g i n of a given ground state i s indicated as experimental (E), theoretical (T), or suggested (?), usually from ESR evidence. The general scheme of bonding i s shown in Figure 3 for VCO (46). The c l a s s i c Dewar-Chatt-Duncanson model (47,48) involves donation by the CO 3σ o r b i t a l into the metal 3da + 4sa + 4ρσ o r b i t a l to form the 3σ VCO o r b i t a l and back donation by the metal an o r b i t a l s into the antibonding 2π o r b i t a l to form the 2π MO, weakening the CO bond. Experimentally, the effects of the metal-CO interaction are evidenced by lowering of the C-0 stretching frequency, v , below 2143 cm" , the value i n the free CO mole cule. The trend i n the CO stretching frequency varies as shown i n Figure 4 for first-row t r a n s i t i o n metals (49). An unknown frequency i n this Figure i s that of MnCO (and perhaps ScCO) and i t i s proposed to be high, i . e . , the bonding of Mn to CO to be weak. (This remains to be established by some other experiment.) The variation i n v , taken equivalent here to CO bond strength, may also be considered as inversely proportional to the M-C bond strength. Qualitatively, one can account for the bonding i n these MCO molecules by considering the 4 s 3 d ~ ·• 4 s 3 d ~ promotion energy of 6
2
4
5
3
1
c o
c o
2
n
2
1
n
1
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
15.
WELTNER & VAN Z E E
219
Diatomic and Monocarbonyl Molecules
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch015
ScCO TiCO VCO CrCO MnCQFeCOCoCO NiCO CuCO ~4s
5σ
-ff:
4σ
-3d
16
2π
4
Σ(Τ,Ε)
'Φ(τ)
5
Σ(Ε)
7
Σ(Ε)
5
Σ(?)
5
Σ(τ)
»Δ(?)
•Σ(τ)
2
Σ(Ε,Τ)
Figure 2. Molecular o r b i t a l schemes and ground states of MCO molecules (M = first-row transition metal). Boxes here emphasize the only molecules with experimentally established ground states. Ground states are indicated as: theoretically calculated (T); experimentally determined (E); or suggested (?), based on absence of an ESR spectrum.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
220
THE CHALLENGE OF d AND f ELECTRONS Vatom
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch015
4—H
\
-H Figure (46).
3.
Molecular
orbital
scheme f o r the °Σ VCO molecule
ο ο 3
Se Ti V Cr Mn Fe Co Ni Cu Figure 4. Plot of the CO stretching frequencies, v , i n the first-row transition-metal monocarbonyl molecules MCO ( c i r c l e d points are tentative) (49). Ρ gives the variation of the atomic energy of promotion corresponding to 4s 3d ~ 4 s 3 d ~ , where η i s the number of valence electrons. Curve Β i s a crude indication of the expected trend i n CO bonding strength due to decreasing di\ backbonding from Sc to Cu. c o
2
1
n
1
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
n
2
15.
WELTNER & VAN ZEE
221
Diatomic and Monocarbonyl Molecules
the metal atom, the απ backbonding involved i n the DCD theory, and hybridization. The former yields the extreme variations (trace P) shown i n Figure 4 and provides a barrier to σ-bond formation. Pro motion and/or hybridization of the 4s electrons relieves the a n t i bonding between these metal electrons and the CO 3σ o r b i t a l , r e s u l t ing i n a shorter M-C bond and stronger Mdïï back donation. Back donation i s expected to generally decrease as one proceeds from l e f t to right because the 3d o r b i t a l shrinks i n size (44, 50) · This e f f e c t tends to strengthen the CO bond as one moves across the tran s i t i o n series from l e f t to right, as depicted roughly (trace B) i n Figure 4. Then the sum of Β + Ρ should have the contours of v , which i s approximately true. What about the ground states? We w i l l consider each of the carbonyls as the 3d s h e l l i s f i l l e d up, beginning with ScCO. Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch015
c o
1 "λ ScCO. In the IR ScCO was presumably formed since a proper CO s h i f t was observed, but the bands were uncomfortably broad (49) · Then, although not firm, v = 1950 cm" , as given i n Figure 4, indicating a bound MCO molecule. Jeung and Kouteck^ (51) have made a MRD SDCI pseudopotential calculation and found a strongly bound Σ ~ lowest state corresponding to the configuration 2π 4σ shown i n Figure 2. A repulsive Π state i s next highest and then a strongly bound Π state (2Π 1θ 4σ ) about 0.2 eV higher. In these two bound states the Sc 4s electron i s polarized away from the CO i n the 4σ orbital. Also i n both states there i s a net electron transfer from Sc to CO. ESR experiments, after several t r i a l s , were successful i n detecting ScCO i n an argon matrix at 4 Κ (52). The ground state was found to be Σ , as predicted by theory, with the z e r o - f i e l d s p l i t ting parameter | D | > 1 cm" . Hyperfine interaction with the S c nucleus was observed but substitution of C 0 produced no s p l i t tings, indicating only a small spin density on the C atom of that ligand. The s character at the Sc atom i s estimated to be about 60% from the hyperfine parameters, and the very large g s h i f t i s reasonably accounted for by the low lying Π state founa by Jeung and Kouteck? (51). J
1
C Q
4
2
4
1
1
1
1
4
1
4 5
13
4
TiCO. GVB theory (perhaps the oldest theoretical treatment of an MCO molecule) finds the c l a s s i c a l σ donation of the CO nonbonding o r b i t a l to Ti da and d e r e a l i z a t i o n of T i απ to acquire CO π* character (53). Nonbonding T i electrons are polarized away from the CO i n 4s-Xdp o r b i t a l s . In our symbolism the ground state i s Φ (4σ 2π 1δ ) where 2π i s mainly 4ρπ rather than 3άπ. (It i s stated that this inducing of 4p character into the valence orbitals should be enhanced i n Sc and decrease rapidly i n proceeding past V. ) Ex perimentally, v has not been established; i n Figure 4 i t has been assumed to l i e along the line joining v for ScCO and VCO. An ESR spectrum for TiCO was not detected which i s i n accord with a Φ ground state. 3
2
1
1
c o
c o
3
5 1
1 3
VCO. V and C hyperfine structure (hfs) i n the ESR spectrum established the molecule as VCO and yielded a complete electronic and magnetic picture of a S = 5/2 ground state molecule including g tensor, hf parameters and z e r o - f i e l d - s p l i t t i n g parameter |D| (46). The d i s t r i b u t i o n of the five unpaired electrons i s i n approximate
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
THE CHALLENGE OF d AND f ELECTRONS
222 Ί
accord with the (2π) (1δ) (4σ) configuration i n Figure 3. The σ electron has about 30% s character, the remainder being 3da and 4ρσ. The spin density on the CO i s quite small. There i s the interesting observation of two conformations of VCO; both were observed i n argon matrices, one i n neon, and the other i n krypton. It i s reasoned that one of these i s bent and that i t i s most l i k e l y the gas-phase form. E a r l i e r infrared studies by Hanlan, e t a l . (54), had suggested that VCO might be bent, and extended Htlckel theory (assuming low spin) indicated that there was a monotonie decrease i n energy of the molecule as the angle decreased. Thus, an ab i n i t i o calculation would be of real interest. A strong M-CO bond i s indicated by a decrease of 240 cm" (in s o l i d argon) i n v ( ) (see Figure 4). 1
5 4
c o
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch015
1
CrCO. The CO stretching frequency was observed at 1977 cnT (49), but there appear to be no other data on CrCO. Unpublished ESR spectra i n our laboratory showed uncharacteristic very broad lines when chromium metal was trapped i n an argon matrix containing CO i n various concentrations. A tentative analysis indicated a Σ mole cule, the reason for the extraordinary breadth of the lines was not understood. In Figure 2, the ground state i s tentatively given as Σ (2π 1δ 4σ 5σ ), i n agreement with the addition of one electron to the known VCO configuration. 7
7
2
2
1
1
MnCO. Two sets of authors agreed on the infrared spectrum of Mn + CO i n matrices (55,49), but the firm i d e n t i f i c a t i o n of a MnCO signal was i n doubt. I t was suggested by us that the molecule i s essen t i a l l y nonbonded and therefore that v = cm" . This was rationalized, as indicated i n Figure 4, by the high promotion energy of the Mn atom, implying that the 4s electrons provide a repulsive interaction with CO. There i s evidence of the existence of MnCO i n the gas phase, at least with a lifetime long enough to allow i o n i z a tion to form MnCO , i f the suggested mechanism for i t s preparation i s correct. That ion and Mn , Mn were observed during photofrag mentation of Mn (CO) and believed to be produced by multiphoton ionization of the neutral MnCO (and Mn, Mn ) photoproducts (56). The signal shows f i r s t - o r d e r dependence upon the sample pressure. (Note that Mn i s proposed to be formed from Mn , which i s a van der Waals molecule as proposed for MnCO.) 2
1
4
0
1
c o
+
+
2
2
10
2
2
2
1
8
9
8
1
FeCO. Experimentally, IR i n matrices finds v cm" (57,58), indicating a relatively strong Fe-C bond, but EngeIking and Lineberger (59) estimate the bond energy as 1.0 ± 0.3 eV. Also the Mossbauer isomer s h i f t of -0.60 mm/s i s close to that of the free atom, -0.75 mm/s (59). Theory has not quite decided whether the ground state i s Σ ~ or Σ ~ and i s finding i t d i f f i c u l t to account for the isomer s h i f t (50,60-65). Thus, Guenzberger, et a l . (62), employing the discrete variational method with the Χα l o c a l approxi mation, place the 6 o r b i t a l below the σ and, on an aufbau basis, obtain a dir d6 4a , Σ ~ ground state. The calculated isomer s h i f t i s -0.12 mm/s, i n disagreement with experiment. A recent c a l c u l a tion by Marathe, et a l . (65 ) (using SCF + MP4SDTQ) derived a spinquintet (d-?rdo 4σ 5σ , Σ " ) ground state rather than a spintriplet. Π states (such as π δ 4 σ 5 σ ) are calculated to l i e at least 1.4 eV higher i n energy. However, with the Σ " ground state c o
3
4
4
5
2
5
2
1
3
1
5
3
2
2
5
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
15.
WELTNER & VAN ZEE
223
Diatomic and Monocarbonyl Molecules
the calculated isomer s h i f t i s -0.11 mm/s, s t i l l i n poor agreement with experiment, although the calculated C-0 stretching force constant i s i n good agreement with that derived from the experi mental frequency. I t i s suggested that perhaps the matrix effects are large, which i s not consistent with the vibrational results. However, another recent calculation by Daoudi, et a l . (66) [using SCF with (CIPSI)] finds the Σ ~ state to be lower than the" !*" state by 0.4 eV, the Fe-C bond energy to be 1.34 eV and v ~ · I t i s disappointing that an ESR spectrum for FeCO was not observed since i t could resolve the m u l t i p l i c i t y problem i n the ground state, but i f Σ states are lowest, a large z e r o - f i e l d - s p l i t t i n g i s implied. 3
5
=
1
9
8
6 c m
c o
1
CoCO. v i n this molecule has been determined to be 1953 cm" i n s o l i d argon, but i t s ESR spectrum was not detected a t the time even though there was no question that Co atoms and CO were present i n the matrix (67 ). I t i s probable that the molecule contains at least one unpaired electron whether one reasons from FeCO or NiCO, so that i t s undetectability i n powder ESR spectra i s due either to o r b i t a l angular momentum or large zfs i n the ground state. The former seems more l i k e l y so that the Δ state suggested i n Figure 2 is reasonable.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch015
c o
4
NiCO. The dissociation energy has been determined but with rather large uncertainty, 29 ± 15 kcal/mole (68, see also 69), and the CO stretching frequency found to be 1996 cm"" (70,68). The electron a f f i n i t y has been measured (68). That i s e s s e n t i a l l y the extent of the experimental data except for observation of NiCO over the nickel surface under special conditions (71,72). Early theory predicted a Δ ground state, (73,74,75-81) but Rives and Fenske (82) were the f i r s t to show, using a many configuration wavefunction, that the Σ state i s s l i g h t l y lower (0.15 eV) than the Δ , but the bond distance i s much shorter i n the s i n g l e t (1.70 Â) and the binding energy (2.7 eV) and Ni-C v i b r a t i o n a l frequency (505 cm*" ) much larger. The most recent theoretical studies (83,84,50,85,86-96) have indicated that π bonding i s much more important than σ bonding, which i s i n fact repulsive. This i s "softened" by sda hybridization. The n i c k e l atom i s close to d rather than d , expected for zero valent Ni because the promotion to the S ( d ^ ) state i s energetic whereas the s d state i s almost degenerate with the ground state. 1
3
1
3
1
9
1 0
1
1
9
2
CuCO. Kasai and Jones (97) proved that CuCO has a Σ ground state and observed Cu and C hfs. The CO stretching frequency i s 2010 cm"* i n an argon matrix (98-100). Theory does not predict a bound complex (50,73,74,83-85,101-103 ) or at least only a "possible weak van der Waals interaction for the Σ state" (104). However, the ESR hf data indicate that the spin density d i s t r i b u t i o n i s p ( 4 S ) = +0.67, p ( 4 p a ) = +0.08, p(2sa) = +0.05 with the appreciable f r a c t i o n remaining probably i n the C(2pa) o r b i t a l . Thus, although weakly bound, the spin on Cu i s polarized, and more important, perhaps 20% of the spin i s on the CO ligand. 1 3
1
2
+
Cu
Cu
c
Acknowledgments The the
authors wish to thank their co-workers who have contributed to topics discussed here: S. B. H. Bach, C. A. Baumann, M.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
224
THE CHALLENGE OF d AND f ELECTRONS
Cheeseman, C. A. Taylor, R. L. DeKock, L. B. Knight, J r . , and M. T. Vala. This research was supported by National Science Foundation Grant CHE 8514585. Literature Cited 1. 2. 3. 4. 5. 6.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch015
7. 8. 9. 10. 11. 12.
13. 14.
15. 16. 17.
18. 19. 20. 21. 22. 23. 24. 25. 26.
Weltner, W., J r . ; Van Zee, R. J . Ann. Rev. Phys. Chem. 1984, 35, 291. Shim, I. Mat.-Fyx. Meddr. Danske Vidensk. Salsk. (16 Res. Rep. of Niels Bohr Fellows) 1985, 41, 147. Morse, M. D. Chem. Rev. 1986, 86, 1049-1109. Koutecky, J.; Fantucci, P. Chem. Rev. 1986, 86, 539. Salahub, D. R. Adv. Chem. Phys. 1987, 69, 447-520. Efremov, Y. M.; Samoilova, A. N.; Kozhikhovsky, V. B.; Gurvich, L. V. J . Mol. Spectrosc. 1978, 73, 430-40. Michalopoulos, D. L.; Geusic, M. E.; Hansen, S. G.; Powers, D. E., Smalley, R. E. J . Phys. Chem. 1982, 86, 3914-16. Riley, S. J.; Parks, Ε. K.; Pobo, L. G.; Wexler, S. J. Chem. Phys. 1983, 79, 2577-82. Bondybey, V. E.; English, J . H. Chem. Phys. Lett. 1983, 94, 443-47. Efremov, Y. M.; Samoilova, A. N.; Gurvich, L. V. Chem. Phys. Lett. 1976, 44, 108. Hopkins, J . B.; Langridge-Smith, P. R. R.; Morse, M. D.; Smalley, R. E. J . Chem. Phys. 1983, 78, 1627-37. Langridge-Smith, P. R. R.; Morse, M. D.; Hansen, G. P.; Smalley, R. E.; Merer, A. J . J . Chem. Phys. 1984, 80, 593600. Leopold, D. G.; Almlöf, J.; Lineberger, W. C., Taylor, P. R. J. Chem. Phys. 1988, 88, 3780. Morse, M. D.; Hansen, G. P.; Langridge-Smith, P. R. R.; Zheng, L. S.; Geusic, M. E.; Michalopoulos, D. L.; Smalley, R. E. J . Chem. Phys. 1984, 80, 5400. Åslund, N.; Barrow, R. F.; Richards, W. G.; Travis, D. N. Ark. Fys. 1965, 30, 171-85. Lochet, J . J . Phys. B. 1978, 11, L55-L57. Ruamps, J . C. Spectrochim. Acta, 1957, Suppl. 11, 329. Joshi, K. C.; Majumdar, K. Proc. Phys. Soc. London 1961, 78, 197. Ruamps, J. C. C. R. Hebd. Seances Acad. S c i . 1954, 239, 1200. Numerous studies, see reference 3. Fu, Z.; Morse, M. D. J . Chem. Phys. (submitted). Taylor, S.; Lemire, G. W.; Hamrick, Y. M.; Fu, Z.; Morse, M. D. J. Chem. Phys. (submitted). Femenias, J . L.; Cheval, G.; Merer, A. J.; Sassenberg, U. J . Mol. Spectrosc. 1987, 124, 348-68. See, for example, Jacox, M. E. J . Mol. Spectrosc. 1985, 113, 286-301. Weltner, W., J r . Magnetic Atoms and Molecules; Van Nostrand Reinhold: New York, 1983. Kasai, P. H.; McLeod, D., J r . Faraday Symp. Chem. Soc. 1980, 14, 65-78. McNab, T. K.; M i c k l i t z , H.; Barrett, P. H. Phys. Rev. Β 1971, 4, 3787-97.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
WELTNER & VAN Z E E 27. 28. 29. 30. 31. 32.
33.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch015
34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44.
45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56.
Diatomic and Monocarbonyl Molecules
225
Montano, P. A. Faraday Symp. Chem. Soc. 1980, 14, 79-86. Curl, R. F. Mol. Phys. 1965, 9, 585. Baumann, C. Α.; Van Zee, R. J.; Weltner, W., J r . J . Phys. Chem. 1984, 88, 1815-1820. Guenzburger, D.; Saitovich, Ε. M. B. Phys. Rev. Β 1981, 24, 2368-79. Nagarthna, Η. M.; Montano, P. Α.; Naik., V. M. J . Am. Chem. Soc. 1983, 105, 2938-43. Weltner, W., J r . In Comparison of Ab I n i t i o Quantum Chemistry with Experiment; B a r t l e t t , R. J., Ed.; Reidel, Dordrecht, 1985; pages 1-16. Van Zee, R. J.; Weltner, W., J r . High Temp. S c i . 1984, 17, 181-191. Cocke, D. L.; Gingerich, K. A. J . Chem. Phys. 1974, 60, 1958. Gingerich, K. A. Faraday Symp. Chem. Soc. 1980, 14, 109. Shim, I.; Gingerich, K. A. Chem. Phys. Lett. 1983, 101, 528. Brewer, L. Science 1968, 161, 115. Van Zee, R. J.; Weltner, W., J r . Chem. Phys. Lett. (in press). Van Zee, R. J.; Weltner, W. J r . Chem. Phys. Lett. 1984, 107, 173. Cheeseman, M.; Van Zee, R. J.; Weltner, W., J r . (to be published). Knight, L. B., Jr.; Van Zee, R. J.; Weltner, W., J r . Chem. Phys. Lett. 1983, 94, 296-99. Knight, L. B. J r . ; Woodward, R. W.; Van Zee, R. J.; Weltner, W., J r . J . Chem. Phys. 1983, 79, 5820. Walch, S. P.; Bauschlicher, C. W., J r . J . Chem. Phys. 1983, 79, 3590-91. Walch, S. P.; Bauschlicher, C. W., J r . In Comparison of Ab I n i t i o Quantum Chemistry with Experment; Bartlett, R. J . Ed.; Reidel, Dordrecht, 1985; pages 17-51. Baumann, C. Α.; Van Zee, R. J.; Weltner, W., J r . J . Chem. Phys. 1983, 79, 5272-79. Van Zee, R. J.; Bach, S. Β. H.; Weltner, W., J r . J . Phys. Chem. 1986, 90, 583-88. Dewar, M. J . S. B u l l . Soc. Chem. F r . 1951, 18C, 79. Chatt, J; Duncanson, L. A. J . Chem. Soc. 1953, 2939. Bach, S. Β. H.; Taylor, C. Α.; Van Zee, R. J.; Vala, M. T.; Weltner, W., J r . J . Am. Chem. Soc. 1986, 108, 7104. Bauschlicher, C. W., J r . ; Bagus, P. S.; Nelin, C. J.; Roos, Β. O. J. Chem. Phys. 1986, 85, 354-364. Jeung, G. H.; Koutecký, J . Chem. Phys. Lett. 1986, 129, 56976. Van Zee, R. J.; Cheeseman, M.; Weltner, W., J r . (to be published). Mortola, A. P.; Goddard, W. Α., III J . Am. Chem. Soc. 1974, 96, 1-10. Hanlan, L.; Huber, H.; Ozin, G. A. Inorg. Chem. 1976, 15, 2592-97. Huber, H.; Kundig, E. P.; Ozin, G. Α.; Poe, A. J . J . Am. Chem. Soc. 1975, 97, 308-14. L i c h t i n , D. Α.; Bernstein, R. B.; Vaida, V. J . Am. Chem. Soc. 1982, 104, 1830-34.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
226 57.
THE CHALLENGE OF d AND f ELECTRONS
59.
Moskovits, M.; Ozin, G. A. Cryochemistry (eds. Moskovits and Ozin, Wiley, New York, 1976) pages 261-439. Peden, C. H. F.; Parker, S. F.; Barrett, P. H.; Pearson, R. G. J. Phys. Chem. 1983, 87, 2329-36. Engelking, P. C.; Lineberger, W. C. J. Am. Chem. Soc. 1979,
60.
Bursten, B. D.; Freier,
58.
101,
5569-73.
1980,
61.
19,
D. G.; Fenske, R. F.
Inorg. Chem.
1810-11.
Barbier, C.; Del Re, G. Folia
Chim. Theor. Lat.
1984, 12,
27-44.
62. 63.
Guenzburger, D.; Saitovitch, E. M. B.; De Paoli, Μ. Α.; Manela, H. J. Chem. Phys., 1984, 80, 735-744. Daoudi, Α.; Suard, M.; Barbier, C. C. R. Acad. Sc. Paris, Ser.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch015
64.
II
1985,
301,
911-914.
65.
Sawaryn, Α.; Aldridge, L. P.; Blaes, R.; Marathe, V. R.; Trautwein, Α. X. Hyperfine Interact. 1986, 29, 1303-6. Marathe, V. R.; Sawaryn, Α.; Trautwein, Α. X.; Dolg., M.; Igel-Mann, G.;-Stoll, H. Hyperfine Interact. 1987, 36, 39-
66.
Daoudi, Α.; Suard, M.; Barbier, C.
58.
J. chim. phys. 1987, 84,
795-798.
67. 68. 69. 70.
71.
Hanlan, L. Α.; Huber, H.; Kuendig, E. P.; McGarvey, B. R.; Ozin, G. A. J. Am. Chem. Soc. 1975, 97, 7054-68. Stevens, A. E.; Feigerle, C. S.; Lineberger, W. C. J. Am. Chem. Soc. 1982, 104, 5026-31. Blomberg, M.; Brandemark, U.; Johansson, J.; Siegbahn, P.; Wennerberg, J. J. Chem. Phys. 1988, 88, 4324-33. DeKock, R. L. Inorg. Chem. 1971, 10, 1205-11. Garrison, B. J.; Winograd, N.; Harrison, D. E., Jr. J. Vac. Sci.
72.
Sci., 73.
74.
76.
78. 79.
80.
789-92.
1983,
126,
392-6.
3,
Z. Naturforsch. A 1979, 34, 114-116. Fundament. Res. Homogeneous Catal.
73-82.
Cederbaum, L. S.; Domcke, W.; Von Niessen, W.; Brenig, W. Z. Phys. Β 1975, 21, 381-88. Walch, S. P.; Goddard, W. Α., III J. Am. Chem. Soc. 1976, 98,
77.
16,
Itoh, H.; Kunz, Α. Β. Itoh, H.; Kunz, A. B. 1979,
75.
1979,
Technol.
Liang, D. B.; Abend, G.; Block, J. H.; Kruse, N. Surf.
7908-17.
Clark, D. T.; Cromarty, B. J.; Sgamellotti, A Chem. Phys. Lett. 1978, 55, 482-87. Bullett, D. W.; O'Reilly, E. P. Surf. Sci. 1979, 89, 274-81. Bagus, P. S.; Hermann, K. Surf. Sci. 1979, 89, 588-95. Rosen, Α.; Grundevik, P.; Morovic, T. Surf. Sci. 1980, 95, 477-95.
81. 82.
Rives, A. B.; Weinhold, F. Int. J. Quantum Chem., Quantum Chem. Symp. 1980, 14, 201-209. Rives, A. B.; Fenske, R. F. J. Chem. Phys. 1981, 75, 1293-
83.
McIntosh, D. F.; Ozin, G. Α.; Messmer, R. P.
1302. 1981,
84.
Inorg. Chem.
3640-50.
Ha., T.-K.; Nguyen, M. T. 109,
85.
20,
J. Mol. Struct.
(Theochem) 1984,
331-338.
Bauschlicher,
C. W., Jr.
J. Chem. Phys.
1986,
84,
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
260-267.
15.
WELTNER & VAN Z E E
86. 87. 88. 89. 90. 91. 92. 93. 94.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch015
95. 96. 97. 98. 99. 100. 101. 102. 103. 104.
Diatomic and Monocarbonyl Molecules
in
Saddei, D.; Freund, H. J.; Hohlmeicher, G. Chem. Phys. 1981, 55, 339-54. Howard, I. A.; Pratt, G. W.; Johnson, K. H.; Dresselhaus, G. J. Chem. Phys. 1981, 74, 3415-19. Bagus, P. S.; Roos, B. O. J . Chem. Phys. 1981, 75, 5961-62. Dunlap, Β. I.; Yu, H. L.; Antoniewicz, P. R. Phys. Rev. A 1982, 25, 7-13. You, X. Jiegou Huaxue 1983, 2, 183-188. Huzinaga, S.; Klobukowski, M.; Sakai, Y. J . Phys. Chem. 1984, 88, 4880-86. Kao, C. M.; Messmer, R. P. Phys. Rev. Β 1985, 31, 4835-47. Bauschlicher, C. W. Chem. Phys. Lett. 1985, 115, 387-391. Rohlfing, C. M.; Hay, P. J . J . Chem. Phys. 1985, 83, 464149. Madhavan, P. V.; Whitten, J . L. Chem. Phys. Lett. 1986, 127, 354-359. Carsky, P.; Dedieu, A. Chem. Phys. 1986, 103, 265-75. Kasai, P. H.; Jones, P. M. J . Am. Chem. Soc. 1985, 107, 81318. Huber, H.; Kuendig, E. P.; Moskovits, M.; Ozin, G. A. J . Am. Chem. Soc. 1975, 97, 2097-2106. Ozin, G. Α.; VanderVoet, A. Prog. Inorg. Chem. 1975, 19, 105-172. Moskovits, M.; Ozin, G. A. Vibrational Spectra and Structure (Durig, J., Ed. Elsevier, Amsterdam 1975). Bagus, P. S.; Hermann, K.; Seel, M. J . Vac. S c i . Technol. 1981, 18, 435-452. Bagus, P. S.; Nelin, C. J.; Bauschlicher, C. W., J r . J . Vac. S c i . Technol. 1984, A2, 905-909. Kuźminskii, M. Β.; Bagatuŕyants, Α. Α.; Kazanskii, V. Β. Izv. Akad. Nauk. SSSR, Ser. Khim. 1986 , 284-8. Merchán, M.; Nebot-Gil, I.; González-Luque, R.; O r t i , Ε. J. Chem. Phys. 1987, 87, 1690-1700.
R E C E I V E D December 9, 1988
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Chapter 16
Spin Density Functional Approach to the Chemistry of Transition Metal Clusters Gaussian-Type Orbital Implementation 1
1
2
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch016
J. Andzelm , E. Wimmer , and Dennis R. Salahub 1
Cray Research, Inc., 1333 Northland Drive, Mendota Heights, M N 55120 Département de Chimie, Université de Montréal, Montréal, Québec H3C 3J7, Canada 2
In this contribution we review the accuracy and computational efficiency of the local spin density functional (LSDF) approach using linear combinations of Gaussian-type orbitals (LCGTO) for the calculation of the electronic structures, ground-state geometries, and vibrational properties of transition metal compounds and clusters. Specifically this is demonstrated for (1) bis(π-allyl) nickel where this approach gives an excellent qualitative and quantitative interpretation of the observed photoemission spectrum; (2) chemisorption of C atoms on a Ni(100) surface, where the present com putational approach determines the adsorption site of C as the four fold-hollow position above the surface with a calculated C - N i bond length of 1.79 Å(exp.: 1.75 ± 0.05 Å) and a vibrational frequency of 407 cm , (exp.: 410 cm ); (3) vibrational frequencies of C O on Pd: the calculations reveal that inclusion of surface/subsurface Pd-Pd motions couple significantly to the CO-Pd vibration leading to a reduction of the vibrational frequency by about 20% compared with a rigid substrate model. Inclusion of an external electrical field shows a stiffening of the C-O vibration with increased positive potential of the electrode; and (4) the electronic structure of Zn clusters where it is found that properties such as the first (s-electron) ionization potential converge rather slowly towards the value of the extended system requiring at least 20 transition metal atoms for an accurate description of the surface electronic structure. It is demonstrated that the computation of threeindex two-electron integrals can be achieved with a highly efficient vector/parallel algorithm based on recursive integral formulas recently published by Obara and Saika. Furthermore, we present the theoretical framework for L C G T O - L S D F gradient calculations. -1
-1
0097-6156/89/0394-0228$06.00/0 © 1989 American Chemical Society
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch016
16. A N D Z E L M ET A L
Spin Density Functional Approach
229
There is growing evidence that local spin density functional (LSDF) theory (1) provides a unified theoretical framework for the study of electronic, geometric, and vibrational structures of solids, surfaces, interfaces, clusters and molecular systems (2,3) encompassing metallic as well as covalent bonding. Although widely used in solid state physics, including semiconductors, transition metals, lanthanides and actinides, density functional methods are still rather rarely applied to problems in chemistry. One of the reasons appears to be the lack of experience with this method i n addressing typical chemical questions such as molecular conformations, vibrations, and reactivity. Such investigations require the capability to calculate accurate analytical derivatives of the total energy with respect to displacements of the atomic nuclei. Evidently, the potential of L S D F gradient calculations has not yet been fully developed. However, i f such a goal could be achieved for molecules and large clusters including transition metals, one would not only have an additional theoretical/computational tool to investigate large molecules with first and second row atoms (which is the current domain of Hartree-Fock calculations) but one could also study organometallic compounds, investigate reactions on metallic surfaces, and simulate large and complex systems such as zeolites and enzyme catalysts at an unprecedented level of detail. In this paper, we present results obtained with a particular molecular orbital implementation of local spin density functional theory using a linear combination of Gaussian-type orbitals (LCGTO's). The results, derived from single-point total energy calculations, illustrate the applicability of the L C G T O - L S D F approach to questions of electronic structure, vibrational properties, and geometries of transition metal clusters and compounds and, in addition, shed light on the computational issues encountered in this kind of large-scale molecular simulation. Furthermore, we demonstrate that the L C G T O implementation allows a compact analytical formulation of energy gradients thus setting the stage for future exploitations of this chemically important feature. The paper is organized in the following way. First the key features of the L C G T O L S D F method are reviewed (3); for a more detailed description of L S D F calculations, the reader is referred to recent reviews (2,3). Four examples, discussed next, highlight the performance of the L C G T O - L S D F approach to predict (1) the electronic structure (photoelectron spectrum) of a transition metal complex, bis(TC-allyl) nickel; (2) chemisorption geometries of carbon atoms on the Ni(100) surface; (3) the vibrational properties o f C O on Pd including the influence of an external electric field; and (4) electronic properties of Z n clusters as a function of cluster size. The last example also illustrates the dependence of computational requirements on the size of the system.
The Linear Combination of Gaussian-Type Orbitals (LCGTO) Implementation Basis Sets. Since the suggestion of Boys (4), Gaussian-type basis functions have become the standard i n quantum chemical ab initio methods. In the Hartree-Fock theory the occurence of four-center two-electron integrals makes this choice a computational necessity. In local density functional theory, on the other hand, a conceptually simpler Hamiltonian gives greater freedom in selecting the variational basis set. For example, in many solid state calculations it has become common practice to use plane waves and augmented plane waves (5) with numerically generated radial functions i n a linearized form (6). A n elegant alternative consists in the use of numerically generated atomic orbitals as basis for molecular orbitals (7). Both approaches using numerical basis functions, the solid state linearized augmented plane wave (LAPW) method (6,8) and the molecular/cluster approach (7,9) have proven extremely useful in carrying out highprecision local spin density functional calculations for solids, surfaces, clusters, and
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
230
THE CHALLENGE OF d AND f ELECTRONS
molecules (2). On the other hand, the accurate evaluation of analytic energy derivatives within these implementations turns out to be a considerable challenge (10).
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch016
In the present studies, we use Gaussian-type basis functions for the molecular orbitals to solve the local density functional equations (11). This choice offers several advantages: (i) there is a wealth of experience from numerous ab initio Hartree-Fock calculations i n using G T O ' s for molecular calculations (12), (ii) computationally, this approach can be implemented in a highly efficient way, as will be discussed below, (iii) the analytic nature of the basis functions opens the possibility for accurate analytic calculations of energy gradients for geometry optimizations and density gradients for non-local corrections (13), and (iv) effective core potentials or model potentials can be readily incorporated (14). Besides the Gaussian basis set for the wavefunctions, there are two other sets of Gaussian expansions used in the present approach, one for the electron density and one for the exchange-correlation potential. The expansion of the electron density is used in the evaluation of Coulomb integrals. Hence the expansion coefficients of the electron density are chosen (lib) such as to minimize the error in the Coulomb energy arising from the difference between the "exact" electron density (i.e. the density originating directly from the wavefunctions) and the fitted electron density. A l l necessary steps to obtain the expansion coefficients of the electron density can be carried out analytically. On the other hand, the expansion coefficients for the exchange-correlation potential have to be obtained numerically by generating the values of the exchange-correlation potential on a grid, which are then used to fit a Gaussian expansioa After this numerical step, the matrix elements of the exchange-correlation potential operator are calculated analytically.
Integral Evaluation. In contrast to Hartree-Fock methods, the L C G T O - L S D F approach requires evaluation of only three-index integrals, thus representing an N algorithm (with Ν being the number of basis functions). A l l examples discussed below, except the Z n clusters, were calculated using the Hermite Gaussian basis originally implemented by Sambe and Felton (11a) and further developed by Dunlap, Connolly and Sabin (lib). In contrast, Cartesian, not Hermite, Gaussians are the most widely used choice i n ab initio quantum chemistry. The scheme of recursive computation of four-index cartesian Gaussian integrals, originally developed by Obara and Saika (15) for the Hartree Fock method, has now been reformulated for the three-index integrals needed i n the present method. A s shown below, a computationally highly efficient scheme results from this approach. 3
Two kinds of three-index integrals are needed, Coulomb integrals of the form (adopting the notation of Obara and Saika (15)) I = [a(l)b(l)llc(2)] c
(1)
where "II" refers to the l / r operator, and overlap-like integrals to calculate exchangecorrelation potentials and energies of the form 1 2
I
xc
= [a(l)b(l)c(l)]
(2)
Here a and b stand for orbital basis functions and c denotes Gaussian functions used in the fitting of the electron density or the exchange-correlation potential. Most of the time is spent in the computation of I . We can rewrite the original Obara and Saika formula (15) in a form suitable for computation of three-index integrals: c
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
16.
A N D Z E L M ET AL.
231
Spin Density Functional Approach
[àblICc+lj)]^ =
[abllc]
(m+1)
(m+1
+ Vi η N^c) {[ablKc-lj)]^ - (ρ/η) [abll(c-l.)] >} + ^ ( ζ + η ) ) {N.(a) [(a-lj)bllc]< > + N.(b) [a(b-ipilc]^ >) m
n+1
(3)
Here, l is a short-hand notation for a p , D or p function and function c+lj has an angular momentum one order higher than c. The recursive nature of Equation 3 allows to build, for example, integrals with d-functions from integrals with only s- and p- functions. The superscript (m) refers to the order of the incomplete Γ function. W. and Q are related to geometries and values of exponents of Gaussian functions; η , ζ and ρ depend on these exponents and N (c) is a generalized Kronecker delta. For details the reader is referred to the original paper by Obara and Saika (15), Equation 39. i
x
z
{
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch016
{
Expansion of the integral with a single function c, instead of a (or b) already cuts the number of arithmetic operations by as much as 30% for d-type integrals. Substantial savings in the time of integral evaluation occur i f we calculate entire groups of integrals with shared exponents in various symmetries. Particularly, we can design basis sets for the electron density and exchange-correlation potential with shared exponents without loosing accuracy in the fitting process (16,17). Hegarty and van der Velde (18) analyzed the number of arithmetic operations necessary to calculate 4-index integrals. The best algorithm (18) requires about 22 operations per integral with d functions. In contrast, the present method requires about 8 operations per 3-index integral. The new formulas for the integral calculations can be efficiently programmed on a vector computer and the algorithm is amenable to parallelization as will be shown below. As for Hartree-Fock calculations, storage of integrals becomes the computational bottleneck for systems with a large number of basis functions. In this case, a "direct S C F ' (19) scheme can be adopted (20). For the L C G T O - L S D F method we deal with a smaller number of integrals (of the order N rather than N as i n Hartree-Fock calculations) and therefore we may use the standard approach for up to about N=1000, as discussed below for the case of large Zn clusters. There is an additional advantage in the direct scheme as we have to calculate the full Hamiltonian matrix only once and then, in each iteration, add the changes to the matrix which correspond to modifications i n the density and the exchange-correlation potential, but only for those matrix elements where this change is greater then a threshhold. During the course of iterations towards selfconsistency, a smaller and smaller number of matrix elements needs to be updated. 3
4
Gradients. The calculation of energy gradients within the L S D F method using localized basis sets has been investigated by a number of researchers (10,21). However, there has been no published formulation for the case of the L C G T O - L S D F method. We will now give an outline of the formulas that have recently been implemented (22). Full details will be published i n due course. In this method, both Coulomb and exchange-correlation energies are calculated analytically once the fitting coefficients for the electron density, p with ρ = e , and exchange-correlation energy-density, e , and potential, μ , are obtained (11). The total energy ( E ^ p ) can then be expressed as r
s
E
LSDF =
Σ
P Ν
h
pq ί p ,
+
*r Pr ^
+
Σ
£
, , ^
>"*
Σ
8
U
« Pr P,
+ »
4
Here Ρ denotes the density matrix, h contains kinetic energy and electron-nuclear attraction operators; [pqllr] and [rllt] are Coulomb repulsion integrals with 3 and 2 indices, respectively, [pqs] denote one-electron 3-index integrals and U is the nuclear-nuclear repulsion term. The form of Equation 4 ensures (lib) that the Coulomb energies are accurate up to second order in the difference between the fitted density and the "exact" density obtained directly from the wavefunction. n
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
THE CHALLENGE OF d AND f ELECTRONS
232
In order to obtain the energy gradient of E ^ p by differentiation of Equation 4 with respect to a nuclear coordinate, x, one can, in the first step, closely follow the procedure used in Hartree-Fock theory (23). In addition, derivatives of the density fitting coefficients occur in the present approach. These terms can be eliminated using the equation for the density fitting together with the normalization condition of the total density. At this point, the intermediate gradient formula is given by a E
LSDp/
=ς ρ - vt ς
a x
μ
α
a +
Z
ν
P r
Pq
{atyax + ς a[pqiir]/ax+ ς e aipqsi/ax j atriiti/ax + a u y a x - ς ^ w atpqi/ax γP r
8
P t
V
s
m
a x
{
Σ
(e
* s - V ) [pqs]} + Σ s
P
Ν
M
Σ,
Βεβχ
[pqs]
(5)
Here, W is an energy-weighted density matrix element as in the Hartree-Fock gradient formula {23). Equation 5 contains two "difficult" terms (the last two), the derivative of density matrix elements and the fitting coefficients e . It turns out that these terms can be eliminated by using the relationship
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch016
s
ap/ax^
-e
x c
x c
) = p ae /ax
(6)
xc
which can be obtained by differentiation of the fundamental L S D F formula (1) =
d
£
a
£
< xc Ρ > / P = xc
+
Ρ
a
E
xc /
d
7
Ρ
after multiplying both sides by p. In practice, μ and e are obtained through a fitting procedure. The fitted quantities do not correspond strictly to the original density and eq. (6) no longer holds exactly. However, with the fitting basis sets currently used, this approximation appears to be reasonable (22). Assuming therefore the validity of Equation 6, we obtain the following L S D F energy gradient formula which is valid also for the spin-polarized case and in the presence of non-local corrections to the Hamiltonian (20)) xc
a E
LSDF^
X
=
F
HFB
+
F
8
D
()
with F
S
P
+
HFB = pq pq {
Σ
γ Pr U
F is the Hellman-Feynman force plus the correction for the orbital basis set dependence on the nuclear coordinate, x. The term F is specific to the present L C G T O L S D F implementation. Equation 10 is equivalent to R F B
D
F = D
f
^p [a(r)/axll(p-p )] r
f
(11)
with ρ being the "exact" density and p the fitted density. Qearly, in the case of a perfect fit F vanishes. It is important to realize that within the L C G T O implementation, evaluation of L S D F gradients boils down to computations of 2 and 3 index integrals D
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
16.
ANDZELMETAL.
233
Spin Density Functional Approach
which can be accomplished by the same efficient technique of integral calculation as described above.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch016
Illustrative Examples Electronic Structure and Photoelectron Spectrum of bis(TC-allyl) nickel. The photoelectron (PE) spectrum of bis(TC-allyl) nickel has attracted considerable attention since it was first measured in 1972 by LLoyd and Lynaugh (24). Later experiments (25) established an accurate basis for comparisons with theory. The first theoretical investigation by Veillard et al. (26) using the Hartree Fock approach (cf. Table I) revealed that the interaction between d orbitals of N i and π* orbitals of the allyl group is responsible for most of the bonding. Koopmans* theorem was found to be invalid in this case and total energy A S C F calculations are required. However, even large-scale CI calculations (27) could not correctly identify the first band as ionization from the ligand π-type orbital (7a orbital assuming symmetry of the molecule). A semi-empirical Green function approach (28) (cf. Table I) provided an efficient calculation of all of the ionization potentials (IP's) and gave satisfactory assignment for most of the P E bands. Hancock et al. (29) performed scattered-wave (SW-Xot) calculations of IP's applying the transition state method. Compared with experiment, the calculated spectra are shifted considerably towards lower energies and especially ionizations out of π-type orbitals seem to be in error. u
The present calculations were performed with an all electron version of the L C G T O L S D F method. Triple zeta basis sets for nickel and carbon atoms including polarization functions were employed. Details of basis sets and the method of their optimization are given in Ref. 16. A recent neutron diffraction study (30) revealed a pronounced bending of the anti-hydrogen atoms (by 30°), as a result of their strong repulsion from the nickel atom. First we discuss results obtained assuming planar geometry of the allyl groups as this allows a direct comparisons with the other theoretical calculations of the P E spectra performed so far. The interaction diagram between the 3d and 4s orbitals of N i and the ( C H p fragment together with the resulting orbital levels of bis(TC-allyl) nickel are shown in Fig. 1. The basis set superposition error (BSSE) (17) is corrected for by using the same basis set in all calculations except for the isolated N i atom. This accounts for the splitting in the N i ( C ) case. 3
2
2h
Bonding is caused by the interaction between d-electrons of N i and the π* electrons of the ( C H ) fragment. By symmetry, bonding is allowed within the b and a manifolds. The donation of electrons from occupied 2b orbitals of the metal atom to the unoccupied 5b orbital of the ( C H ) fragment results in the bonding molecular orbital, 5b . Another bonding orbital, 10a is formed as a combination of 7a and 6a orbitals of Dis(TC-allyl) and nickel, respectively. This bonding effect is, however, largely cancelled by the antibonding partner, the 13a orbital. Close to the H O M O (6b ) there are levels of mainly 3
5
2
g
g
g
g
3
g
5
2
g
g
Mulliken charges reveals that the corresponding molecular orbitals (7a and l l b ) have some admixture of metal 4p states. The coupling with 4p states causes a transfer of the ligand*s charge to originally unoccupied 4p orbitals of the metal. This "back donation" mechanism was first discovered in SW-Χα calculations (29) and is confirmed by the present L C G T O - L S D F study. u
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
u
THE CHALLENGE OF d AND f ELECTRONS
234
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch016
Table I. Experimental and theoretical results for the photoelectron spectrum of bis(TC-allyl) nickel (in eV) The principal orbital character of the ionized electron is added in parenthesis
band exp. (a)
ASCF (b)
1 2
7.8 (π) 8.2 (d)
3
8.6 (d)
4
9.4 (d, π)
5 6 7
10.4 (π) 11.6 (π) 12.7
8
14.2
5.5 5.6 6.8
11.0
ASCF-CI (c)
GF (d)
SW-Xcc (e)
6.4 6.6 6.7
8.7 8.9 9.2 9.2 9.5 10.0
2.5 4.5
10.8
10.9 12.2 13.013.2 15.4 15.6
5.0 5.1 5.5 5.6 6.6 7.9-8.2 9.09.2
LCGTO-LSDF(f) planar bent
7.8 8.0 8.2 8.4
8.1 8.1 8.4 8.4
(π, (d, (d, (d,
9.2 9.5 10.7 11.9 12.112.5 13.6 13.8
9.4 (d, l i a ) 9.9 (d, π, 10.3 (π, lib*) 11.2 (π A 10a ) 11.8- (σ) 13.3 14.3 (σ) 14.6
(a) experimental data : Batich, Réf. 25 (b) Veillard; Rohmer et al., Réf. 26 (c) Moncrieft et al.; Hillier, Réf. 27 (d) semi-empirical Green Function calculation: Bohm and Gleiter, Réf. 28 (e) scattered-wave Χ α calculation: Hancock et al., Réf. 29 (f) present calculations: allyl groups are planar or hydrogen atoms are bent
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
7a ) 13aJ 12ap u
6bJ
Aj
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Figure 1. Orbital diagram of the L S D F one-particle energy levels of isolated nickel, nickeKC^), bis(TC-allyl) nickeKC^) and the ( C j H ^ C L ) fragment. • and Δ indicate spin up and spin down electrons. The three-dimensional structure of bis(7c-allyl)nickel is shown on the left-hand side.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch016
236
THE CHALLENGE O F d A N D f ELECTRONS
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch016
Using a planar geometry for the H atoms, the calculated L S D F A S C F values for the P E agree rather well (within 0.2 eV) with experiment (cf. Table I). Using a bent geometry (30), the total energy of the transition metal complex is lowered by 0.9 eV. The corresponding calculated photoelectron spectrum agrees less convincingly with experiment. The main bonding mechanism, i.e. donation from N i 3d to allyl π*, and back donation from allyl π to Ni-4p, remains, however unchanged. The d orbitals of N i are barely affected by this geometry change, and the important shifts are found in the ionizations from ligand π and σ orbitals. It should be noted that the experimental P E was taken in the gas phase while the "bent" structure was deduced from measurements on the crystalline solid (30). Clearly, a full geometry optimization of the isolated complex would settle the question of crystal packing effects and clarify the details of the photoelectron spectrum.
Chemisorption of Carbon Atoms on the Ni(100) Surface. The next example demonstrates the capability of the L C G T O - L S D F method to predict the geometrical structures and vibrational frequencies of carbon atoms chemisorbed on the Ni(100) surface. The C / N i system is of fundamental importance in catalytic petrochemical processes and thus has been the subject of many experimental and theoretical studies. Despite these efforts, many aspects such as the equilibrium position of the C atoms (above or below the surface N i atoms) remained unsettled. Recently, a joint experimental and theoretical study of the chemisorption of carbon on Ni(100) led to a clearer understanding of this system (31). It was found that in the ground state, C is adsorbed in four-fold hollow sites above the surface with a N i - C distance of 1.79 Â, in agreement with the experimental value of 1.75 ± 0.05 Â, obtained from surface extended energy loss fine structure (SEELFS) measurements. Furthermore, the calculated vibrational frequency of the perpendicular mode of the adsorbed C atom, 407 cm is in excellent agreement with the experimental value of 410 c m . _ 1
_ 1
Vibrational Frequency of CO Adsorbed on Pd Clusters. In the next example, we present results for the vibrational properties of a C O molecule on a Pd surface, modelled by a cluster of 14 transition metal atoms (32,33). Furthermore, the influence of an external electric field is investigated using a smaller cluster, P d C O (33,34). In both cases, C O is assumed to be bonded in the bridge position. The chemically inactive Pd core electrons are described by a relativistic model potential as described in detail in Ref. 14. 2
Two coupling modes are considered: for the P d C O cluster the first mode (denoted as h) represents vibration of the rigid C O molecule with respect to the transition metal surface. The second mode is either the Pd-Pd vibration within the plane of Pd surface atoms (r) or out-of-plane stretch of the surface/sub-surface Pd-Pd bond (z). The total energy surfaces (h,r) and (h,z) are calculated for discrete points and then fitted to a fourth order polynomial. Variational and Quantum Monte Carlo (QMC) methods were subsequently applied to calculate the ground and first excited vibrational states of each two-dimensional potential surfaces. The results of the vibrational frequences ω using both the variational and Q M C approach are displayed in Table II. 14
If one assumes a rigid substrate which, at first, seems reasonable because of the high mass of Pd compared with the C O molecule, a frequency of almost 500 cm" is obtained for the vibration of the entire (rigid) C O molecule perpendicular to the surface. This value is in significant disagreement with the experimental value of 340 cm" (35). As can be seen from Table II, a substantial lowering of this "beating" mode occurs due to anharmonic effects in the coupling of the ζ and h modes. In other words, the vibration of the surface/sub-surface Pd-Pd bond stretching couples to the C O beating mode and lowers 1
1
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
16.
ANDZELMETAL.
237
Spin Density Functional Approach 1
Table II. Vibrational frequencies (cm' ) of the two coupled modes (h,r) and (h,z) obtained from a harmonic (Har), variational (Var) and Quantum Monte Carlo ( Q M Q approach
mode
Har
Var
QMC
240 498
192 521
192 521
196 498
65 402
66 397
(h,r): ω
Γ , usually a determinant with η single f-states. The actual wave function for a p a r t i c u l a r experimental s i t u a t i o n h, for example an i n i t i a l state h = i or a f i n a l state h = f, i s
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch017
n
n
|h>=EC |f >
(8)
n h
t h i s being the so called configuration i n t e r a c t i o n (CI) or multiconfiguration (MC) schemes (see, for example (8-10), where the t o t a l number of f electrons n w i l l not, i n general, be an integer value. The configuration wave function |f > i s usually represented by a determinant for the " a c t i v e " space where only the ν valence e l e c trons contribute to one column each: η columns for the f electrons, v-n-1 columns for the d electrons and one column for the (s-p) conduction electrons for each RE atom η - nf. In t h i s case the amount of atomic-like f-character: f
n
atomic n^ = ZnC, f hn η v
n 2
2
a
, < n η — f
/\ (9) 0
£
t o m i c
There should be some additional contributions to n ^ in Equation 9 , from the cross terms i n which can be important, mostly i f several C are of the same order of magnitude and at least two of these terms have η > 0. For a given h the parameter n | i s a suitable quantity to measure the amount of very l o c a l i z e d , atomic-like f-character, and i t s r e l a t i o n to the properties of the system. atomic i unique because i t depends on the choice of and ψ£» , but a consistent scheme w i l l lead to avoid t h i s , well known, shortcoming of population analysis. The configuration i n t e r a c t i o n ( i n the MC scheme) wave function |h> i s needed i n the condensed matter problem because the reduced symmetry of the c r y s t a l and, mainly, the e f f e c t of the scattering wave boundary conditions for energies above the i n t e r s t i t i a l po t e n t i a l or the existence of bonding states between the anions and the heavy metal ions, w i l l allow several types of coupling or cor r e l a t i o n between the f l e v e l and the valence or conduction e l e c trons. This coupling, usually denoted by Δ , i s i n fact the sum of several contributions which are responsible for either MV, Kondo or other e f f e c t s . In general there w i l l be a dominant configuration n n
t o m : L C
n
s
n o t
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
252 η C
THE CHALLENGE OF d AND f ELECTRONS contribution
n+1
(°
r
C
(C Δ
n-P
n
~
1.0)
Ε
and
Ε
then a second with lower amplitude E
- ^ η + 1 "~ ίΐ ) » î}
b e i n
t h e
8
total
energy
computed
n
for a fixed f configuration. For example i n the d i f f e r e n t phases of the Ce problem the ground state i s usually described as i n a f configuration, but i f a core hole i s present or for the case of BIS spectroscopy the f configuration w i l l then take a larger weight. In the ground state of α-Ce the increased i n t e r a c t i o n between the f electrons and the conduction band w i l l bring i n the f° configuration because i n t h i s case the energy difference E ^ i c ^ a ^ g i - ^ f ^ ^ o s i s small; t h i s being one of the reasons for the complicated behaviour of Ce metal and Ce compounds. For these materials the CI method has been extensively used by Gunnarson and Schonhammer (13) (and re ferences therein). The c o e f f i c i e n t s Cjj w i l l i n general depend on the hybridization of the f electron wave functions which i s the o r i g i n of the forma tion of an f band. But, as already pointed out by Haldane (14), these materials tend to behave as "low density" or impurity l i k e systems at intermediate temperature. This consideration i s import ant because i n a material such as CeAl3 , at intermediate tempera tures, transport properties correspond to incoherent scattering by each RE atom at the maximum resonance scattering cross section (15), as expected from an impurity l i k e Kondo-resonance, and the coherence between the RE atoms, although fundamental for the under standing of the low temperature regime, can be introduced a poste riori. We can describe the process of pressure induced valence tran s i t i o n s (as i n F i g . 1 of Ref. (3)) as a three step phenomenon. The f i r s t , i n the pressure range Ρ 95
49
-
005
0
The s u b s c r i p t c i n d i c a t e s t h a t o n l y c o n n e c t e d diagrams have to be taken. The last equation follows after applying a linked cluster theorem. Furthermore the abbreviation < . . . · > = < Φ Ι . . . ιΦ > has been used. The energy i s evaluated a f t e r expanding 0
2 Σ i j l
Ε = -
+ Σ ijmn
Σ 11·
0
Dij
c
+ (
Dij . The c i r c l e s s y m b o l i z e atoms and t h e five segments the different d-orbitals. The d-electron occupancy p e r atom i s chosen t o be 2 . 5 . 0
Figure 2. Charge f l u c t u a t i o n s as f u n c t i o n o f d-band f i l l i n g n U . E l e c t r o n c o r r e l a t i o n s influence strongly the energy difference between nonmagnetic and magnetic states, l e a d i n g t o d r a s t i c changes o f the Stoner-Wohlfarth crit e r i o n for the onset of ferromagnetic order. The r e a s o n i s t h a t e l e c t r o n i c charge f l u c t u a t i o n s are smaller i n a ferromagnetically ordered than i n a nonmagnetic state. Therefore e l e c t r o n c o r r e l a t i o n s decrease the energy by a l a r g e r amount o f a s t a t e w h i c h i s n o n m a g n e t i c , t h a n o f a f e r r o m a g n e t i c s t a t e . F o r example, i n Fe t h e energy g a i n d u e t o f e r r o m a g n e t i c o r d e r i s 0 . 5 6 e V / a t o m when t h e HF a p p r o x i m a t i o n i s made a n d a r a t i o U / W = 0 . 4 4 i s assumed. When d e n s i t y and in addition spin correlations are i n c l u d e d , t h i s energy reduces t o 0.22 eV/atom and 0.15 e V / a t o m , r e s p e c t i v e l y (10) . I n t h e LDA t o t h e density functional theory, Hund's rule correlations are not taken into account, because t h e y are n o t p r e s e n t i n an unpolarized homogeneous electron gas from which the exchange-correlation potential i s taken. When t h e L S D approximation is applied instead, they are partially included. Spin correlations, however, modify the generalized Stoner parameter strongly ( 1 3 ) . The latter can be related to the exchange correlation energy E ( M ) f o r f i x e d m a g n e t i z a t i o n M by w r i t i n g
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch020
2
c
C
X C
Q
0
Mo E
X C
(M )
= E
0
x c
(0)
+
\
J dM I 2
x c
(M)
(13)
ο where I = I ( 0 ) i s the o r i g i n a l 4 d i s p l a y s the magnetic f i e l d parameter as obtained within mations for the case of Co. x c
Stoner parameter. Figure dependence o f the Stoner three different approxi I i i s the result of a
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
20.
STOLLHOFF & FULDE
285
Correlations in d and f Electron Systems
Ό
5
10
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch020
2
F i g u r e 3. Atomic spin correlations S as f u n c t i o n o f d-band filling n. U/W=0.5. (Adapted from r e f . 9.) 0
1.6
ι—ι—ι—ι—ι—ι—ι—r
F i g u r e 4. S t o n e r parameter I(M) and l o s s o f k i n e t i c energy D(M) (dashed line) for Co as functions of magnetization. 1^ - f u l l c o r r e l a t i o n c a l c u l a t i o n , I2 - neglecting spin correlations, I3 - n e g l e c t i n g a l l c o r r e l a t i o n s , r e s u l t i n g from J ^ j . (Adapted from r e f .
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
286
THE CHALLENGE OF d A N Df ELECTRONS
complete correlation calculation. I2 is the Stoner p a r a m e t e r when s p i n c o r r e l a t i o n s a r e n e g l e c t e d . Since spin correlations play no role for fully magnetic s t a t e s , b o t h c u r v e s become e q u a l a t M =1.6. W h i l e I2 does not d i s p l a y s i z e a b l e magnetic f i e l d dependencies and compares in this respect with results of LSD c o m p u t a t i o n s , 1^ i n c r e a s e s b y 20% f r o m M=0 t o M . 13 finally is the curve obtained when a l l exchange contributions (~Jij) to the interaction part of the H a m i l t o n i a n ( E q . 1) a r e t r e a t e d i n H F a p p r o x i m a t i o n . I2 a n d I3 m a y b e c o n s i d e r e d a s l o w e r a n d u p p l e r l i m i t o f t h e d e f i c i e n c i e s o f L S D . A l t h o u g h a c h a n g e o f I b y 20% may s e e m s m a l l , i t h a s t h e e f f e c t o f c h a n g i n g t h e C u r i e temperature T by approximately a f a c t o r o f two, because T ~ ( I N ( O ) - l ) / * a n d I N ( 0 ) * 1, However, even t h a t is not s u f f i c i e n t i n order to b r i n g the l a r g e r c a l c u l a t e d values for T i n agreement w i t h experiments. This is due t o the fact that Stoner theory does not contain fluctuations of the order parameter. For improved calculations of T see e . g . Ref. (14). Another point of considerable importance i s the n o n l o c a l character of the exchange. The l a t t e r always favours non-uniform distributions of electrons (or holes) a m o n g t h e d i f f e r e n t d o r b i t a l s , e . g . e g a n d t2g orbitals, when t h e system is cubic. Direct Coulomb interactions as well as correlations favour uniform occupations of the different atomic orbitals and therefore counteract the effect of nonlocality of the exchange. Despite this, the anisotropies caused by exchange are important, i n p a r t i c u l a r f o r b u l k N i (10) as w e l l as f o r i t s s u r f a c e (15). Finally it is of interest to compute spin c o r r e l a t i o n s between n e i g h b o r i n g s i t e s , i.e. m a x
m
a
x
c
1
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch020
c
c
c
2
S (o)
= , o
(14)
0
where δ denotes a n e a r e s t neighbor o f s i t e 0. IΨ> is c a l c u l a t e d by s t a r t i n g from the nonmagnetic S C F - s t a t e as b e f o r e , b u t b y i n c l u d i n g i n S ( s e e E q . 6) a l s o operators of the form O i j ( 1 , l + a ) = £ > i ( 1 ) S j (1+©) . The effect of these operators i s that additional ferromagnetic corre lations betweeen electrons on neighboring sites are b u i l t i n t o I V » > , e x c e p t f o r b a n d f i l l i n g s c l o s e t o ncf. The i n d e x 1 s t a n d s for l i g a n d o r b i t a l a n d we a s s u m e t h a t i t i s r a t h e r extended s o t h a t we m a y n e g l e c t C o u l o m b r e p u l s i o n s w i t h i n that orbital. The H a m i l t o n i a n t h e n r e a d s Η = ε,Σ σ
1σ1
χ
+ σ
c
f Σ φ σ
+ V Σ (φσ+ΐσ^σ) σ
σ
+
Un£nf * τ
(15) + with η = f f and U v e r y l a r g e . We w a n t t o d i s c u s s t h e s o l u t i o n s o f t h e e i g e n v a l u e problem f o r two e l e c t r o n s . F i r s t we s e t V = 0 . In that case, because ci>tf the ground s t a t e i s a quartet w i t h e n e r g y E = c i + C f , i . e . one e l e c t r o n i s i n t h e f o r b i t a l and the o t h e r i s i n the 1 o r b i t a l . The e x c i t e d s t a t e i s a s i n g l e t with E = 2 c i , i . e . both electrons are i n the 1 orbital. W h e n VfO i s t a k e n i n t o a c c o u n t , the grounds t a t e quartet s p l i t s i n t o a low l y i n g s i n g l e t f
σ
a
a
Q
s
72
•*o> =
(φΐ-φΐ)
U-(V/Ac)2)
I0>
- ^
φ|ΐΟ> (16)
with
energy
l*cl>
=
2
Ε =εχ+ε£-2ν /Δε 0
(1-(V/Ac)2)
and a
φ|ΐΟ>+
triplet
j2 ( φ ί - φ τ ) Ό > (17)
Φΐ
,0>
'*c2> = ? ε 3 > = ΦΪ»0> w i t h energy Et=ci+Cf. We h a v e s e t ε χ - ε £ = Δ ε 4 ) . The f o r m a t i o n o f t h e s i n g l e t ιΨ> w i t h triplet state of e x c i t a t i o n energy Ε =2ν characteristic feature of strongly correlated systems. When t h e f o r b i t a l i s e m b e d d e d i c o n d u c t i o n e l e c t r o n s , t h e energy g a i n due t o f o r m a t i o n becomes , ψ
0
β χ
ΔΕ= Dexp[
2N(Ô)V^
2
(see figure an e x c i t e d /Δε, is a f-electron n a sea of the singlet
]
2
(
1
8
)
i n s t e a d o f ΔΕ=2ν /Δε, as i n the case o f two o r b i t a l s . H e r e D i s t h e c o n d u c t i o n - e l e c t r o n band w i d t h and 2N(0) is their density of states. We h a v e set the Fermi energy equal to zero. The energy gain is usually identified with a characteristic temperature kgTi^E,
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch020
288
THE CHALLENGE OF d AND
f ELECTRONS
t h e Kondo t e m p e r a t u r e . The low l y i n g e x c i t a t i o n s l e a d t o h e a v y - f e r m i o n b e h a v i o u r when t h e i o n s w i t h f e l e c t r o n form a l a t t i c e . The above calculation suggests that a singlet f o r m a t i o n due to strong correlations with a triplet e x c i t e d s t a t e s h o u l d be f o u n d i n a p p r o p r i a t e m o l e c u l e s . The effect r e q u i r e s an e v e n t o t a l number o f v a l e n c e electrons. I n o r d e r t o d e t e c t i t one s h o u l d s e a r c h e.g. f o r m o l e c u l e s c o n t a i n i n g Ce, w h i c h a r e d i a m a g n e t i c , b u t w h i c h show a f - e l e c t r o n c o u n t c l o s e t o 1, when p h o t o emission experiments are performed. The f o r m a t i o n o f a s i n g l e t s t a t e due t o s t r o n g c o r r e l a t i o n s i m p l i e s a l s o a new k i n d o f e l e c t r o n - p h o n o n coupling. The e n e r g y g a i n Δ Ε due t o s i n g l e t f o r m a t i o n d e p e n d s on t h e h y b r i d i z a t i o n V, w h i c h i n t u r n d e p e n d s on p r e s s u r e Ρ o r volume Ω · In p a r t i c u l a r i n a s o l i d t h i s dependence ΔΕ(V) is very strong (see Eg. (18)), r e s u l t i n g i n a s t r o n g e l e c t r o n phonon c o u p l i n g . Its s t r e n g t h c a n be c h a r a c t e r i z e d by an e l e c t r o n i c G r u n e i s e n parameter
= η
-
din T din Ω
K ±y
< '
M e a s u r e d v a l u e s o f n a r e a s l a r g e a s 100-200 i n h e a v y f e r m i o n systems (17). One important problem, which i s presently under i n t e n s e i n v e s t i g a t i o n s i s t h a t of the Fermi s u r f a c e of s t r o n g l y c o r r e l a t e d f - e l e c t r o n systems. I t was a s u r prise, at least to the present authors, that the m e a s u r e d F e r m i s u r f a c e o f t h e h e a v y - f e r m i o n s y s t e m UPt3 (18) i s v e r y much i n a c c o r d w i t h t h e one computed w i t h i n LDA ( 1 9 ) . T h e r e i s no a p r i o r i r e a s o n why t h e t o p o l o g y o f t h e F e r m i s u r f a c e s h o u l d come o u t c o r r e c t l y when e l e c t r o n c o r r e l a t i o n s a r e s t r o n g and a LDA i s made. But f o r UPt3 i t d o e s come o u t s u r p r i s i n g l y w e l l , a l t h o u g h t h e m e a s u r e d e f f e c t i v e masses a r e o f f b y a f a c t o r o f o r d e r 20 a s compared w i t h t h e c a l c u l a t e d o n e s . Detailed i n v e s t i g a t i o n s h a v e shown (20) t h a t t h e g o o d a g r e e m e n t in the case o f UPt3 i s due to a large spin-orbit s p l i t t i n g and a c r y s t a l - f i e l d (CEF) s p l i t t i n g , w h i c h i s much l e s s t h a n kgT^, i . e . t h e e n e r g y g a i n due t o s i n g l e t formation. In that case, the theory becomes a one-parameter ( w h i c h i s t h e e f f e c t i v e mass) t h e o r y , and t h e t o p o l o g y o f t h e F e r m i s u r f a c e due t o the heavy q u a s i p a r t i c l e s i s c o m p l e t e l y d e t e r m i n e d by t h e g e o m e t r y of the unit c e l l . I n c a s e s i n w h i c h t h e CEF s p l i t t i n g i s l a r g e r t h a n kgTR, one e x p e c t s d i f f e r e n c e s b e t w e e n t h e m e a s u r e d F e r m i s u r f a c e and t h e one w h i c h f o l l o w s from a p p l y i n g t h e LDA. I n o r d e r t o improve t h e computation o f t h e F e r m i s u r f a c e one c a n p r o c e e d a s f o l l o w s , at l e a s t f o r Ce compounds. One a p p l i e s t h e LDA t o t h e d e n s i t y f u n c t i o n a l theory f o r a l l e l e c t r o n s , except the f-electrons. The potential a c t i n g on t h e l a t t e r i s
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
20. STOLLHOFF & FULDE
289
Correlations in d andf Electron Systems
d e s c r i b e d by an energy dependent phase s h i f t $1=3(ε), for which a simple, phenomenological ansatz is made. Only these channels w i t h i n t h e 1=3 m a n i f o l d o b t a i n a phase s h i f t d i f f e r e n t from z e r o , w h i c h have t h e symmetry Γ o f the c r y s t a l - f i e l d ground state. The l a t t e r is usually known from inelastic neutron scattering experiments. T h e s l o p e 70.
suggest
the
of
to
by be
of
bodies
light)
as
two
known has
structure
it
been by
is
non-
does not
obey
assumes),
it
is
safe
to
(NRQM)
the
simplest
These for
simple
velocity
Bohr model
that
this
will
(NR)
(which
mechanics
conclude
to be
study
of
to
of
if
that
one-electron for
(which the
use
behaviour
appropriate
happen
understanding
that
the
comparable
considerations would
a proper
special
significantly
Newtonian
and m o l e c u l e s would not
Ζ predicts
a c c o r d i n g to
is
non-relativistic infinite')
systems moved a t
that
the
well
Schrôdinger equation
fast-moving
velocity
quantum m e c h a n i c s
even
charge
and m o l e c u l a r
Lorentz-invariant
predicted
atoms
these
not
also
Schrôdinger equation
atomic
However,
constitute is
relativity.
equation
of
light.
is
finite
of
non-relativistic
with
of
that
of
in
it
behaviour
(with
velocity
electrons
investigate
and c h e m i s t s .
relativistic; the
to
It
an
the of
atom atomic
therefore the
electronic
0097-6156/89A)394-0291$06.00/0 ο 1989 American Chemical Society
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
THE CHALLENGE OF d AND f ELECTRONS
292 structure (Z>75)
of
the
mechanics
atoms
which
relativity
is
is
in
involving
quantum m e c h a n i c s
conformity
mandatory.
with
Various
and
special
theory
Dirac
who
in
d i s c o v e r e d the
(4)
1928
(linear
in
only
Lorentz-invariant
of
is
momentum
electron
predictions Lamb and
shift forms
atomic that for
agree
the
(and H
basis
(6)
+
be
cannot
made
remain
theory
with
the
is
being
for of
the
for
decades because
not
for
to
(in
the
electrons
of
first
in
calculation
on the
relativity
and
electrons
of
shown b y
compared to
it
known
(7)
that
a t o m was
out
due
to
of
that an
came
the
the
electrons, total
angular
the
notation
i
(7)
and i
the
electrons
case
behaviour
who
of
i-1/2
Swirles
no
valence
that
of
valence
that
whereas
increased were
and
5d)
effect,
the (8).
-
which
d and
relativistic
and j
i+1/2,
A few
Hg
years
weakening
orbitals
s electrons, for
the
more
(i.e., for
Mayers
in
contracts thereby
important
Since
atomic
charge
indirect
mean
electron,
5d e l e c t r o n s
these
strongly
values.
5d e l e c t r o n s
relativity
the
the
less
larger
the
the
by
one-electron
the
unimportant
designate
introduced by
all
nuclear
This
-
was
of
relativistic
found
was
energy
viz.
b e i n g most
momentum j
dynamics
there
H E A V Y ATOMS
occupied by
in
the
primarily
their
the
orbitals
binding
shielding of
and
chemists
decreased s i g n i f i c a n t l y ;
effect,
although
sp>d>d.
level
the
systems of
EFFECTS
reliable
systems.
so
face
the
very
effect
and the
very
to
appreciable
correlation
calculations
(10-13),
(RIP),
is
electron
which happens
RELATIVISTIC
using
elements
due
very
and 5f
an i>0
are
is
these
i n v o l v i n g heavy
correlation
effect
affected
the
increase
calculation of
electrons
of
valence
near
order
6d)
destabilization
time
momentum o f
elements
c h e m i s t s must
the
effect.
electrons)
2
s and ρ e l e c t r o n s
(and
elements
also
of
especially valence
the
systems
quantum
the
splits
heavy
heavy
addition
amount
effects,
chemistry
(and
/
relativistic
in
These e f f e c t s
electrons
for
5d
interaction
±1/2.
in
the
1
relativistic
the
electrons
atom
indirect
and p
angular
relativistic
destabilized while
(s
therefore
direct
smaller
the
velocity
and are
increasing total
uranium
to
occupying penetrating
momenta
spend an a p p r e c i a b l e
that
for
the
due
electrons
angular
s t a b i l i z e d by they
293
Chemistry of Third-Row Transition Elements and Actinides
termed
to
one
are
the DFAO
small.
the
core,
accommodated i n
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
the
THE CHALLENGE OF d AND f ELECTRONS
294 valence
RMOs c o n s t i t u t i n g
general, The b a s i s four
thevalence
a r e composed o f s e v e r a l s e tused t o express
different (I)
types
DFAOs
wavefunction
DFAOs
which,
i n
o f theconstituent
thevalence
atoms.
RMOs c a n c o n s i s t o f u p t o
o f functions (14):
that
aren o tcompletely
filled
i n the isolated
atoms ; (II)
DFAOs be
completely
filled
significantly
i n the isolated
affected
atoms
by the formation
that
might
o f the
molecule ; (III)
E x c i t e d DFAOs u n o c c u p i e d i n t h e i s o l a t e d might
contribute
significantly
atoms
b u t which
to the formation
of the
molecule ; (IV)
Functions, needed
called
augmenting
to describe
valence
charge
small
functions
residual
distribution
(AF),
which a r e
distortions
onformation
of the
o f the
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch021
molecule. Although (I), the
s e t must
needed
same a t o m , basis,
A F s make features
called
selected
functions
atom
lost
o f type
thecoefficients
predicting
have
incorrect RCI
energy, The
techniques
theresults
well
importance
of
calculations significant
This
by using the
i s achieved by using the to construct
wavefunctions
properties
other
feature
as the computational
(14) and because
However,
o f space
we p r e s e n t
relativistic
a summary
D i r a c - F o c k SCF a s basis set
f o r AuH t o d e m o n s t r a t e t h e
as well
as electron
transition
correlation
elements. RELATIVISTIC
CALCULATIONS
binding
t h e 6s-6p
FOR_AuH
hybridization
o f thebonding so that set.
energy
and
distributions, etc.
t h e I s DFAO o f t h e h y d r o g e n a t o m .
( 1 4 ) show t h a t
more
can
than the
s e t f o r AuH c o n s i s t s o f t h e 5 d , 5 d a n d
the chemical basis
experimental
ionicities and
a r e remedied
D F S C F E B S AND
(RCI)
energy
INTERACTION
o u tu s i n g a n extended
functions
FULLY RELATIVISTIC
t h e g o l d atom p l u s
enter
carried
to the
o f thebinding
molecular
as well
fully
o f third-row
CONFIGURATION INTERACTION chemical basis
contribute
charge
here.
ab i n i t i o
of therelativistic
i n diatomics
INITIO
moments,
(III), f o r
i s deduced b y
a n RMO c a l c u l a t i o n
methodology
basis
importance o f
CONFIGURATION
(14).
molecular
i n bonding
i n the gold
set.
described elsewhere
o f our
The
t h e RMO a n d R C I
be repeated
o f 27 v a l e n c e
effects
The
from Both
as RCIc a l c u l a t i o n s
(EBS)
AB
cannot
they
o f overestimating
dipole
o f our
arefully
limitations
to the Hence t h e
played
5 d DFAOs
t o type
thefraction
calculations
e.g.,
details
(14).
d i s s o c i a t i o n products
wavefunctions.
role
i n t h e g o l d atom,
thebasis
been used to calculate
total
t o t h e RMOs.
_The
belonging
THE R E L A T I V I S T I C
u n o c c u p i e d RMOs r e s u l t i n g accurate
Since
t o DFAOs b e l o n g i n g
t h e 5d and
with which
from
o f RMO t h e o r y
t o perform
o f type
work.
consisting o f functions
(III).
elsewhere
orbitals
6p DFAOs
them
(CB),
and
o f which
i n detail
involving
by excluding
accurate
contributions
RMOs a n d b y c a l c u l a t i n g
defects
RIP
(II)
(II),
CORRELATION E F F E C T S : The
small
(I),
t h e 6p a n d
examining valence
only
types
arediscussed
example,
a l l the functions
o f thebonding can be understood b y u s i n g a
the chemical basis
from
hybridization
of
include
f o rquantitatively
AFs a r e constructed t o be orthogonal
essential
by
thebasis
AFs a r e only
However,
6 s DFAOs Our
6p a n d 6p DFAOs
since
DFSCF
i s n o ta do n o t
1.0 eV o u t _ o f t h e
o f 3 . 3 6 eV a r e l o s t
i f t h e 5 d and 5d
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
21.
MALLI
DFAOs two in
Chemistry of Third-Row Transition Elements and Actinides
and t h e i r
DFAOs which
energy
value.
The
5 d DFAOs
in
ten of
electrons 0.663
of
The
interaction
were
relativistic (the bond
shows
(R )
(a> ).
2102
for
R
the
non-relativistic
are
in
2.879
serious au and
1
improved wavefunctions 2.963
au and Our
2102
results
molecular
cm"
of
even
are
the
au and
the
R
features
chemical basis
of
1745
;
the
is
most
readily
from
the
a RIP
computation.
three
hybrid
orbital
in
interaction is,
which
is
with
however,
with α
the
responsible the
found
erroneously
interaction 5d-6s
whilst energy
the
direct
gap
of
polarity
AuH.
for
two
from p r e v i o u s
an approximate play
conclusion
is
incorrect
be
left
in
SCF c a l c u l a t i o n our
relativistic
the
5d-6s
a the
hybridization
hybridization calculation
solely
Is
greater
in
configuration
the
core
of
potential
relativity
role of
AuH. of
in
the
our Our
6s
to
interaction
(ECP)
that
eV
for
5d
5d 1 0
basis the
The
5d-6s in
the
greater
results Hay
et
al.
calculation
the
that
extended
1.682
the
e.g.,
chemical bond
result
DFAO. the
predicts
contrary
the
DFAOs
hybridization
calculations;
of
greater
because 5d
non-
degree
The
arises
the
The
the
significantly
5d-6s
from
DFAOs.
5d and
NR w a v e f u n c t i o n
rigorous
a value
the
stabilizes
reduces
the
i n view
core
formed
orbital 5d-6s
its
nonbonding o r b i t a l s .
effect
the
dominant
predicts
a π gold
important
hydrogen
destabilizes
treatment
to
to a
from
which
non-relativistic
and
an e f f e c t i v e
appear
the
conclusions are
less
of
bonding through
bond i s
effects to
two
reducing
This
revealed
understood
demonstrated
show t h a t
the
the
Moreover,
concluded from
cannot
of
the
the
relativity.
be
seriously underestimates
of
effect
above-mentioned
reported using
one
leading
case.
Our
also
relativistic
these
thus
6s
third
limit,
DFAO. in
that
gold
importance
relativistic
not
in
relativistic
combination
(11)
the
Is
absent
predicts
hybridization
indirect
be
calculation
relativistic
the
for
hydrogen
to
between
relativistic
s p i n and
of
RMOs o b t a i n e d
l o c a l i z e d AuH RMOs,
non-relativistic
are
bonding,
cannot
RMOs c a l c u l a t e d
orbitals
of the
β
localized valence
spin
latter
ω .
non-relativistic
σ
This
and
using
changed by
of
au
(18)
predictions
sets,
the
with
the
values
and
e
of
velocity
2.993
cm
using
with β
the
compared
using
become
theory.
of
contracts
calculations
for
basis
vibration
be
substantially
qualitative
The
large
calculations
The p r e d i c t i o n s
from RCI
both
A comparison
experimental
quantitative
AuH a r e
with
to
and
and
value
the
5d
vibration
predictions
respectively,
of
the
b y RMO c a l c u l a t i o n s
increases
3.431
with
resulting ,
the
AFs on
using a
significantly
respectively.
show t h a t
properties
Furthermore,
1
orbital
the
a
various
energies.
increasing
relativistic
values
cm" ,
with
at
only
experimental
several
and fundamental
respectively,
disagreement
2305
of
these
calculation
predicts
the
calculated
these
core,
RCI
wavefunction
addition
relativity
the e
core of
molecular
substantially
a n d u? ,
e
the 20%
was
from
simply by
Thus
e
in
AuH p r e d i c t e d
the
the
corresponding n o n - r e l a t i v i s t i c
that
and
e
frequency cm
predicted
performed
length
of
in
about
the
The bond l e n g t h
then
used)
is
methods
and e x a c t l y
latter
light
placed
energy
placed
Indeed our
an accurate
distances by both
including AFs.
frequency
are
are
set.
which
requires
valence
Au and H atoms.
set
eV,
the
internuclear
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch021
chemical basis
computation
configuration
of
the
these
binding
the
associated electrons
enter
295
D
1 0
core
does
in
AuH;
this
electrons set
e
of
calculation with
(EBS), AuH the
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
and
DF
296
THE CHALLENGE OF d AND f ELECTRONS
above-mentioned including 2.014
and of
|/?>).
the
We
the
express
|f m >
NRL the
molecular
components
with
terms
the au;
relativistic
total
through
order large
and
3.877
(58916eV!).
defined
|d m >
functions
component
two-component RMOs c a n b e
m >,
than
2165.19
Pyper
m
relativistic
bound at
relativistic
and |p
the
be
RMOs i n where
like
(the
au
3.5
calculated
shall
s
ThO
although
3.877
m
for
molecule
R -
|i m >
eV
molecule
the about
However,
3.5d)
R -
angstrom.
that
at
calculated
experimental
9.0
indicates
predict
NRL D
prediction.
The
0.529171
atoms.
c a n be
calculations
b)
1 au -
experimental
(negative)
au.
non-
separations
Relativistic
)
Relativistic
(ref.
the
18
doubly
(RMO) as
was
and
III.
various
b
ThO
appropriate
DF S C F a n d
various
for
Th atom
Thus,
the of
-0.437,
energies
(D )
dissociation
(au)
a
spinors as
calculated
-0.938,
set
the
orbitals
'cores'
9
Non-Relativistic
R
of
SCF method
taken
relativistic
dissociation
Table
the
atomic
the
The
-1.055, total
basis
thereby
molecular
ThO was
products
-1.246,
in
including
(numerical) for
valence 6d DFAOs
DF S C F w a v e f u n c t i o n .
elsewhere(14).
calculated
and
constructed via
wavefunction
antisymmetrized
the
7s
relativistic
were of
the
thus
6p,
valence
occupied valence spinors
and
6s,
299
equations the
of
the
functions number
momentum
m and correspond
Therefore,
|ρσα>,
have
|ρπβ>,
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
the |daa>,
THE CHALLENGE OF d AND f ELECTRONS
300 |άττ0>,
|faa>,
bonding in
|fc70> f u n c t i o n s ,
characteristics
question.
the
double
The v a l e n c e
group
representation has
the
It
-
of
|7el/2>
«
|6el/2>
-
|5el/2>
and Th
the
0
(using
the
with
calculated the
D
below
the
small
of
the
|2ρπα>
the
0
claimed
(lying
and
|6dπα>
below
the
8|e3/2>
value
of
D
due
e
au,
energy
about
molecular
corresponding -1.531,
-0.165
au
0.51 D
eV
to
the
2165
energies
at
neglect D
at
R -
-0.837, these
MOs;
NRL c a l c u l a t e d
orbital
energy
for
than
the
ThO a r e
-0.326,
between the
given
lowest
above.
relativistic
lying
valence
Thus
the
IP
of
low
by
-1.50
eV
compared to
the
experimental
u n d e r s t o o d as
the
calculated valence
N R L HOMO ( a t
pure
on the
au)
whereas hybrid with
consists
in of
the
contrast |7sa>
of
an
the
(-0.936)
coefficients
eV p r e d i c t e d
almost
calculated and
given
|βάσ> in
by
the
IP
RHOMO ( g i v e n
parenthesis.
HOMOs i s
NRL c a l c u l a t i o n of
|6d5> o r b i t a l
(-0.350)
the
-0.299
higher
eV.
4.49
for
much
the
The
whereas
3.877
corresponding difference
at
calculated
-0.315,
lie
energies
eV)
NRL
separations.
values
orbital
(16.33
for
eV
total
higher
yield
correlation
and 0.73 the
eV
(18). to
orbitals
easily
the
au
eV
expected
1.81
au
-0.948,-0.887,
0.6
9.00
electron
9 valence of
and
internuclear 3.877
in
whereas
of
eV!)
and
-6.00
au)
not
-1.140,
difference
of
is
for
the
is
IP
8.7
2ρπ 0
ionization eV
of
e
the the
6.14
However,
(58912
these
of
of
reported
R-3.877
between
predict au
involving
orbitals
calculated
(at
lies
e
correponding r e l a t i v i s t i c a
the
respectively.
DF S C F e n e r g y
NRL o r b i t a l
bonding
valence
βάσ
experimentally
NRL c a l c u l a t i o n s is
-0.379|6ρπ£1/2>
DF S C F w a v e f u n c t i o n
and 4.277
can be
and
valence
viz;
Koopman's theorem) is
e
The
3.877
the
just of
substantial
the
effects.
and
RMOs
Moreover,
R -
is
au
forms:
a n d 2s,
experimentally,
There
|8e3/2>
| βάσ>
6daal/2>-0.253|6paal/2>
The
be
Th atom w i t h
12s a n d
2saal/2>-0.123
(IP)
to
(using
irreducible
R =
-0.899 2ρσα1/2>+0.363
agreement
and
at
-0.215
excellent
valence
|9el/2>
-0.Ill|2sal/2>
-0.734|6paal/2>+0.629|2saal/2>
atoms,
the
the
RMO
following
2ρσ and
single
σ - h y b r i d on the
from
a bonding combination of
potential
A
diatomic)
-0.350|6daal/2>
calculated valence
the
valence
(18).
heteronuclear
as
additional
the
species
O.813|2p^l/2>+O.418|6d7r01/2>,
-
6d7r,
the
into
molecular
-0.815|2ρπα3/2>-0.417^πα3/2>+0.095|6ρπα3/2>. three
are
These
is
orbitals
next
RMO)
7s-6d
The v a l e n c e
|8e3/2> The
a
however
valence
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch021
of
for
insight
the
ThO d e s i g n a t e d
notation
the
can give
RMOs o f
form:
contribution
atoms.
RHOMO,
of
which
valence
RHOMO o f
-0.936|7scral/2>
consists
bonding Th
(AIR)
etc,
the
theoretical
following
|9el/2>
of
6.00
above)
orbitals Since
of
the
eV. Th
is
This
7s
Th
DFAO
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
R
=
atom,
consists the
1.50 too
of atom is
a
21.
MALLI
stabilized is
(due
differences the
in
easily
pure
the
(due the
|2s
the
Th
(due
to
in
the
it
can be
orbital
and
the
considerably
and
that
effects
due
to
atom
the
bonding
in
species
(>2000°C)
in
The
at
5f
electrons in
geometry
(linear)
of
the to
which
For 5f
the
i.e.,
6d a
DFAOs
total
its
core.
was
found
the
of
The at
to
a
the
Th0
is
the
the
the
U atom
U:
the
lowest
au;
diatomic. R -
3.55
the
2346
molecular
total
to
be
2s
the
at
(1
with 3
of
2
the
in
to
R -
3.55
UO p r e d i c t s
relativistic
calculation,
it
=
atom
a
D
Thus, e
of
should not
R -
for
at
of
R
»
6p, 0
7s,
atom;
atomic
[core]
contains
also
was
3.05,
kept
that
a
as
single for
the
calculation
only the
and
and
binding
limit
in
au
s t i l l 4.05
-0.11971
above
twice be
diatomic
-28145.039469
although
almost
This the
the
were
predict
eV)
2 +
2
valence
the
unbound by
27.211
au.
of
mentioned
energy
of
set
the
-0.14393,
fails
the
difference
and
following
where
at
in
the
performed
UO m o l e c u l e
s h o u l d be
UO m o l e c u l e
for
paramount
6d
basis
energy
while
(18). (U)
significant.
DFAOs
0
1
(U0 )
the
the
unbound by
hartree
for
the
were 2p
cm"
orbitals
structure
included
eV)
It
of
and
argon
5f
very
and
eV
in
isoelectronic
a valence
non-relativistic
hartrees energy
Th
of
corresponding non-relativistic
wavefunction the
the
7.8
0
2.5
T h O a n d UO b o n d s .
SCF w a v e f u n c t i o n
au p r e d i c t s
however, about
The
the
actinide
the
however,
be
is 5f
1 6
in
6
to
of
-
e
an
explanation
molecular
respectively.
spinors
6p 5f 6d7s
was
found
e
electronic
electrons
total
R — 3.55
DF
of
of U
roles
were
Γcore]
Is
a n d c*> x
role
and
of
the
wavefunction
4.55
au,
DFAO's
of
and
(-1.68
-0.17596
differ
temperatures
involving
calculations
electrons
viz
the
au u s i n g
au
determinant
6p
high
1
cm"
the
expected
-0.061739
it
energies
the
relativistic
spectrum
atomic
(bent)
2
terms
and U atoms
18
A n IR
825.0
valence
in
at
diatomic
unbound by au
and
of
the
bonding
ThO u s i n g
theories
6d
(-0.969)
DFAO
MO.
d i s s o c i a t i o n energy
eV.
-
β
exist a
Moreover,
molecular and
for
6s
explains
IPs,
significant the
|6sa>
effect)
is
the
the
|lel/2>
orbital
of
of
above
|lel/2>
whereas
predicted
very
investigate
and 4.55 of
configurations, electrons
ω
investigate
4.05
the
RMO
of
DFAO
the
diatomic.
5.6
of
UO m o l e c u l e ,
relativistic 80
to
to
found Th
ThO
of
relativity
3.55,
and
in
between
us
for
predicted
are
claimed
bond.
may b e
orbitals
3.05,
a hybrid
phase w i t h
led
order
prompted in
the
structure
actinide-oxygen
UO
be
the
6d
FOR UO
gas
15°K has
importance in
of
potential
electromic
with
to
relativistic
the
mentioned
valence
stabilization
NRL m o l e c u l a r
UO i s
the
ionization
matrix
atom;
participation
SCF CALCULATIONS The
an
the
orbitals
Th
that
there
the
the
bonding characteristics
relativistic
for
of
energies
stated
and
effect),
lying
The
direct
effect)
relativistic
lowest
calculated
the
Hence
DF
energies The
orbitals.
difference
the
relativistic
indirect
orbital
N R L MO i s
(-0.124)
atom
the
orbital
|6sa> v a l e n c e
and
and
direct
to
understood.
corresponding
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch021
to
destabilized
can be
301
Chemistry of Third-Row Transition Elements and Actinides
-0.973
eV;
UO m o l e c u l e the
the
DF S C F
is
total
NRL
much as
regarded
U0 at
as
the
better
than
DF SCF c a l c u l a t i o n . The
9|el/2>
relativistic consists
from
\5ΐπβ>
Malli
and
and Pyper
of
a
highest
occupied molecular
7sc7-6d m
functions.
s
(where
'e'
representation
ω -
1/2,3/2,5/2),
the
one-electron 9|el/2> -
T h e RHOMO 9 | e l / 2 >
denotes of
the
the
the
total
UO h a s
the
additional
RMO a n d 1 / 2 , 3 / 2 , 5 / 2
where ω i n d i c a t e s RMO),
of
two-dimensional
corresponds
angular
momentum
to
of
viz.
-0.873|7saal/2>
-0.3981 6 o V a l / 2 > + 0 . 1 8 3 | 5 ί π 0 1 / 2 >
- 0 . 1 7 6 | 5 f a a l / 2 > + 0 . 1 2 9 | 6d7r01/2> - 0 . 0 9 3 | 2 s a l / 2 > -0.069|2paal/2> It
is
an almost
contributions the
0
atom,
However,
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch021
are
where
the
following
from
n o n b o n d i n g RMO s i n c e
the the
|2saa>
coefficients
RMO 8 | e 3 / 2 >
form which
definitely
lying
clearly
involved
8|e3/2> ~
(-0.093)
in
are
just
shows
the
it
and
very
small
|2ρσα> (-0.069)
contains
DFAOs
given
below that
in
the
RHOMO h a s
|5fcS> a n d
bonding of
UO,
of
parentheses. the
16dn> D F A O s
of
U
viz,
0.727|2ρπα3/2> -0.465|5f*03/2> +0.344|6dπα3/2> -0.111|6ρπα3/2>
The v a l e n c e significant and, has
in
fact,
the
|5ίπα>
and
electrons contain
5f
in
the
DFAOs
the
We g i v e interaction
and
for
ground state
below
and
can o n l y be
destabilizes
these
DFAOs a r e
The
substantial
orbital
roles
the
expressions indicate
2ρπ DFAOs o f
0
clearly
their
for
the
energies 5f
to
in the
(electrons
p u s h e d up
the
associated (which
configuration)
u n d e r s t o o d due
which
spinors.
au)
\5£δβ>,
significantly
effect
very
atom
-0.502
actinides
electronic involved
U
the
results
so t h a t
clearly
the
in
contain the
of
for
(DFAOs)
bonding
6 d DFAOs a r e
energy
-0.38 These
spinors
UO a n d T h O d i a t o m i c s ,
also of
a n d 6 d DFAOs o f
respectively.
results
atomic
indicate
for
8 | e 3 / 2 > RMO a l s o
5f
+0.49
atomic
their
relativistic
UO w h i c h
atom,
-0.74,
5f
these
as v a l e n c e
the
the
significant
ThO,
Both
RMOs a l s o
of
very
6d and 5f
bonding
of
the
DFAO i n
that
indirect act
that
below
from
5 | e 3 / 2 > RMO ( w i t h o r b i t a l
|2ρπα> DFAOs, are
bonding. the)
the
coefficients
demonstrate
except
RMOs l y i n g
contributions
of
in
in
energy
the
and
valence
a n d 6 d DFAOs
in
respectively.
for
the
7|el/2>
and
π-bonding arising and the
due
6dπ a n d / o r
| 5 e 3 / 2 > RMOs to
the
5 ί π DFAOs
of
U
viz, 7|el/2> «
Ο . 7 8 8 | 2 ρ π 0 1 / 2 > +0.34616άπβ1/2>
-0.225 |
5ίπβ1/2>
-0.098|6ρπ01/2>, 5|e3/2> Although,
the
-0.74|5f*j33/2> +0.4915f*a3/2> 6 ρ π DFAO w a s n o t
these
valence
RMOs ( e x c e p t
noted
above);
the
|5f D F A O s
6 ρ σ DFAO o f
contribute
to
found
a very the
to
minor
be
involved
contribution
U contributes 61 e l / 2 )
-0.38 | 2ρπα3/2>. significantly to
a s much a s
RMO w h i c h h a s
the
8|el/2> the
|6da>
following
form: 6|el/2> «
-0.87|2ρσα1/2> +0.30|6daal/2> -0.29|6ρσα1/2> -0.29|5faal/2>
-0.2312s
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
in
as and
21. M A L L I The
calculated
relativistic
orbital
of the
7|el/2>,
8 | e 3 / 2 > RMOs a n d t h e RHOMO 9 | e l / 2 >
-0.5017,
-0.4551,
-0.4444,
-0.4281 and -0.2266
and u s i n g Koopman's theorem,
the calculated
e V f o r UO i s i n e x c e l l e n t
experimental molecular
value
calculated
agreement
o f 6 ± 0 . 5 eV ( 1 8 ) .
energies,
UO a r e c o l l e c t e d
e t c . , at various
i n Table
IV.
It
IV.
Calculated total
Non-Relativistic
(E
N R
(E
D F
+28144)
total
t h e UO m o l e c u l e
( E ^ ) and
i n au ( l a u -
( i n e V ) f o r U0 a t (R) i n a u ( l a u -
27.211 eV) various
0.529171
angstrom)
NON-RELATIVISTIC
D (eV)
(E
a
e
M R
+25379)
D (eV) e
-0.95728
-3.9165
-0.02107
-5.0213
3.447
-1.0394
-1.6800
-0.14147
-1.7451
3.877
-0.98150
-3.2573
-0.12528
-2.1855
4.277
-0.92535
-4.7852
-0.07869
-3.5142
-0.06620
-3.7930
-
a)
-
The experimental
(negative) respect
value
D
i s reported
e
indicates
t o b e 7.8 eV ( 1 8 ) .
the molecule
to be bound
A positive
(unbound)
Since
the lowest
total
molecular
e n e r g y was c a l c u l a t e d
at R —
a u , t h e c o r r e s p o n d i n g NRL c a l c u l a t i o n was a l s o p e r f o r m e d
3.55
au i n order
3.55
arising
to gain
insight
into
due t o r e l a t i v i t y .
It
eV w h i c h
i s about
energy
i s about
energy
at R -
ionization
2406 h a r t r e e s
3.55 a u .
potential
Moreover,
has
with
the following 9|el/2>NR -
than
that
the t o t a l
at R a D
value.
of
molecular
1.6 eV lower
o f 6 . 1 4 eV w h i c h ,
the experimental
e
NRL m o l e c u l a r
t h e NRHOMO 9 | e l / 2 > p r e d i c t s i s about
at R =
in
predicted by the
t h e DF SCFt o t a l
o f 4.42 eV, which
c o r r e s p o n d i n g DF SCF v a l u e
excellently
was f o u n d t h a t
however,
above
differences
f o r t h e UO m o l e c u l e
0 . 7 0 eV g r e a t e r
c o r r e s p o n d i n g DF SCF c a l c u l a t i o n ;
the
the major
a u , t h e NRL c a l c u l a t i o n p r e d i c t s
-0.97
with
t o t h e two a t o m s .
3.55
bonding
however,
the than
agrees
T h e 9 | e l / 2 > NRHOMO o f UO
expression, v i z . -0.539|5f*01/2>
-0.483|5faal/2>
-0.449|6daal/2>
-0.348|7saal/2>
-0.288|6άπβ1/2>
-0.213|2ρσα1/2>
+0.144|6paal/2> T h e NRHOMO h a s a m u c h l a r g e r 6dπ DFAOs to
contribution
from
the 5 ί π , 5fa,
o f t h e U a t o m a n d t h e 2 ρ σ DFAO o f t h e 0 a t o m
t h e RHOMO; h o w e v e r ,
atom
is
separations.
3.05
4.677
of
separations f o r
internuclear
RELATIVISTIC R (au)
potential
the corresponding
can be seen that
Relativistic
seperations
au, respectively;
The c a l c u l a t e d
) energies
and d i s s o c i a t i o n e n e r g i e s internuclear
with
5|e3/2>,
o f UO a r e
ionization
internuclear
t o be unbound a t a l l these
Table
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch021
eigenvalues
6|el/2>,
6.17
303
Chemistry of Third-Row Transition Elements and Actinides
the contribution
t o t h e NRHOMO i s m u c h s m a l l e r .
o f t h e 7 sa
Moreover,
6d DFAO the
consists mostly
(with
RMO 8 | e l / 2 >
|5f5>
(-0.465)
the
7|el/2>
the
NR 7 | e l / 2 >
(0.381),
has
6d
and
6|el/2>
a coefficient
from
contributions
(0.344)
from
o f t h e U atom.
contributions
from
t h e 2ρπ
DFAOs
above
(0.788),
|5f53/2>
(0.20)
t h e |2ρπ>
Similarly,
o f U; while,
(0.346)
e.g.,
(0.435), 6ρσ
6fa
the
i s a π - t y p e MO w i t h 6dπ
DFAO),
(0.727),
substantially;
o f 2pa (0.746),
( a n d \5fnfil/2>)
RMO a s d i s c u s s e d
and
major
5ίπ
(-0.225)
o f t h e U atom. The
7|el/2>
orbital
about
eigenvalues
a r e -0.4281
respectively, for
DFAOs
a n d |2ρπ>
RMOs a n d NRMOs d i f f e r
consists mostly
(-0.288)
o f 0.90) o f
(0.338)
\5£π>
corresponding DFAOs
and
the 8|e3/2>
( t h e NRMOs)
a n d-0.4444
i s clear
than
and
results
o f t h e RMO
(-0.2642)
i t
4 . 4 eV lower Our
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch021
(with
contributions
that
t h e NRL o r b i t a l
7|el/2>
therefore
clearly
demonstrate
predictions
o f t h e NRL a n d D F S C F c a l c u l a t i o n s
etc.,
as quantitative
energies,
f o r the diatomics
energies
MOs.
as well
total
eigenvalues are
t h e c o r r e s p o n d i n g DF SCF o r b i t a l
qualitative bonding,
8|e3/2> and
(-0.2989) a . u .
orbital
energies,
involving
that
differences
there
f o r the nature
dissociation
actinides
a r e marked
between t h e
duet o v e r y
relativistic
effects
i n such
RELATIVISTIC
EFFECTS
FOR D I P O L E MOMENTS O F D I A T O M I C S
of
energies significant
systems. O F HEAVY
ELEMENTS The
RIPhas
for
the diatomic
using
been
ab i n i t i o
chemical
basis
internuclear value
basis and
a u was
dipole
functions,
Moreover,
dipole
relativistic Pyper are
(14).
predicted
dipole
experimental
and
PbTe,
about
values
respectively,
turns
40%
out that
au differ
and
0.019
au,predicted with
TiH,
+
a n d NRL
and
27 ( E B 2 7 ) (STO)
f o r AuH ( 2 1 ) .
EB27) a n d
reported
and B i H ;
a u and
by Malli and dipole
moments
however, t h e
very
1.0623
dipole
well
with
au f o r T i l
moment
f o r
the
+
moment
calculated
wavefunction s e tused
the predicted
(at R
e
=
i n the calculation
dipole
s e twavefunctions) from
thevalues
b y t h e NRL w a v e f u n c t i o n s , of TiH, (CB
f o r AuH
i s smaller (by
b y t h e c o r r e s p o n d i n g NRL
thebasis
In the case +
a
In
asA B " .
predicted
considerably
v i z Ti H";
f o r which
f o r AuH u s i n g t h e and
au agree
a positive
the relativistic
polarity
+
orbital
reported
experimental
PbH
1.2655
chemical basis
respectively.
expected
where
PbH
distance.
20 (EB20)
s e t (EB20
(NRL)
t h e CB s e t r e l a t i v i s t i c
However,
0.371
been
BiH,
limit
a t the experimental
except
calculated
(18) o f 1.8137
that
and
calculated
(with a u and
depending upon
relativistic
BiH,
+
the dipole
the
have
moments and
the relativistic with
at present, HgH ,
PbTe
b y a 6p S l a t e r - t y p e
were
the relativistic
the wavefunction.
from sets
basis
i t s polarity
t o 50%) t h a n
wavefunction, of
moments
au) using
species
(CI) wavefunctions
f o r AuH,
AB i n d i c a t e s
It 2.8794
curves
extended
o f 1.9078
the
basis
b y a 6p DFAO,
Unfortunately,
wavefunctions)
species
calculated
interaction
notavailable
PbH ,
(WF) c a l c u l a t e d
e
extended
dipole +
T i l ,
as n o n - r e l a t i v i s t i c
( R ) o f each
moment
CB s e t ,
configuration
as well
CB s e t a u g m e n t e d
CB s e t a u g m e n t e d
TiH,
used f o r the internuclear
moments
using
(21) t o evaluate +
HgH ,
s e twavefunctions
separation
obtained
b y Ramos
s p e c i e s AuH, relativistic
o f 3.5884
addition, WFs,
adapted
although
moments
o f 1.372, f o r AuH,
of dipole
0.323 -0.120
T i H and
the dipole
set)wavefunction
the value
(using
o f 0.976,
moment
moment
predicts the (-0.12 au)
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
21.
MALLI
obtained
from
polarity
for
moments AuH, (/i
),
>
A*CB
EB27
Ε
>
^CB
is
at
au,
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch021
(μ
Ε Β
R
=
e
2.8794
by
the
moments
(21)
electron
Table
R
C
I
n
c
the *
internuclear W
^ R C I ^
that
at μ
ε
RCI
that
and
the
diatomic in
study
systems
Tables
effects
than of
Ε
Β
2
it
Ε
in β
2
fairly
set
A*
C B
and
lesser μ
polarity
the >
7
0.967
order: μ
Ζ Β
and
a
that
predicted
lesser
polarity both
by
chemical basis effect
can be
that
>
7
for CB
order:
that
the
dipole
the
of
i n v o l v i n g heavy
and
V-VI
are
au,
relativistic
first
μ
a
are
R C I
thereby
the
in
0.967
that
3.3794
wavefunctions
au,
are
R C I
the
relativistic
^ F S decrease
and / *
β
opposite
separations
decrease
thereby R «
ε
from
3.3794
(MRCI) μ
s
and / z
at
the
F
R =
β
the
A comparison of
predicted
MS c a l c u l a t i o n s ,
presented such
of
indicating
is
correlation of
*
limit
of
DF S C F LCAS
moments
(
au,
RCI
This
set
indicating a
respectively,
dipole
results
N
au,
27)
non-relativistic
using
A
^ R C I> except
wavefunctions. on
2.8794
set
predicted
the
=
e
>
^EB27
0.846
(MEB2 7^
a
Ti"H .
values
μ
respectively,
EB27
whereas
for
the
+
viz
^RCI » except
R
au,
) ,
Β
at
that
set >
^EB27
0.846
c o r r e s p o n d i n g N R L WF i n d i c a t e s molecule
calculated
(μ),
whereas ( μ
the this
indicates
C B
305
Chemistry of Third-Row Transition Elements and Actinides
set
relativity
atoms
or
ions
concluded from
relativistic
significant
our
and
for
dipole
systems.
V.
Dipole
Calculated
Moments
for
(μ)
LiH,
Til
by U s i n g Chemical Basis
b
and
PbTe
Wavefunctions
8
d
c
A-B
it ( a u )
LiH
2.575
Til
1.908
1.814
PbTe
1.266
1.062
a
Reproduced with
+
A B"
c
polarity.
reference
(18).
l e
EXP (2.367)
p e r m i s s i o n from au -
2.542
d
D.
U s i n g extended
Ref.
2.314
e
b
21.
A l l
Experimental
basis
values
values
function
of
indicate
from
Malli
and
Pyper
(14).
CONCLUDING We h a v e LCAS
REMARKS
c o n c l u s i v e l y shown f r o m
MS
calculations
quantitative diatomics
features
that of
properly
understood using
based
on
the
that,
for
Schrôdinger
the
chemistry
the
core
the
6p
to are
the
in
and
sixth of
gold
heavier 6p
row
DFAOs for
and v e r y the
not
In the
relativistic
well
as
(Z
>
Moreover, in
of
90)
in
cannot
non-relativistic and
it
it
is
theory
heavier
has
5d DFAOs
gold
safe
DF S C F
the
and bonding
atoms
compounds, whereas
involved
chemistry
as
addition, 5d
non-relativistic
the
fully
structure heavy
equation.
and mercury
are
initio
traditional
elements,
elements.
semi-empirical
significant
ab
qualitative
electronic
i n v o l v i n g heavy
the
the
be
theory
been
shown
participate they
to
chemistry
in
belong
state (in
predictions),
to
that contrast but
elements.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
they
306
THE CHALLENGE OF d AND f ELECTRONS
Table
VI.
AuH R e l a t i v i s t i c
Dipole
Non-Relativistic Values
at
Moment
Dipole
R ,
c
in
e
Curves,
(μ)
Moment
au
8
and
b
d
Wave f u n e t i o n s CB R(au)
+ CB
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch021
2. 6294 2.. 8 7 9 4
+
6DSL
f
6pDF
-
0.968 g
CB e
-
RCI
EB20
EB27
0.905
0.829
0.738 0.802
0.976
1.065
1.271
0.984
0.901
(1.373)
(1.443)
(1.925)
(1.434)
(1.363)
3. . 1 2 9 4
0.967
-
-
1.046
0.956
0.846
3. . 3 7 9 4
0.968
1.132
1.045
0.909
8
-
C a l c u l a t e d by u s i n g the
Chemical
Basis
functions,
(CB),
EB20
functions),
-
wavefunctions
extended b a s i s
(same
as
EB27 b u t
and r e l a t i v i s t i c b
wavefunction.
Non-relativistic
without values
polarization
6 p DFAO a s
(see
reference
We a l s o 5d,
structure
DFAOs
involved
relativistic
significant
in
hoped that
would fully
our
(and to
in
the
of
a u -
(RCI)
parentheses.
2.542
D.
centered
effects) the
the
prediction
of
for
the
the
knotty
gold
in
e.g.,
the
electronic
third-row the
formidable
for
dipole of
a
bottlenecks
systems. be
moment,
fairly and
properties
relativistic of
DF S C F c a l c u l a t i o n s h a v e b e e n b r o k e n ,
dual
of
such
shown t o
non-energetic
quality
that
are
calculation
effects
CB atom
18).
heavy
have been
f
on g o l d .
for
present
accurate
effects
All
Slater-type
DF SCF c a l c u l a t i o n s
and they
the
properties,
Thus,
l
in
associated electrons)
correlation
criterion
future.
d
6p'
(reference
e
initio
and a c t i n i d e s ,
accurate
R
diatomics
relativistic
relativistic
ab
of
given
2.75
relativistic
and e l e c t r o n
the
-
polarization
interaction
21.
ζ
(14);
polarization
MCDF c a l c u l a t i o n
their
the
non-energetic
supplement
wavefunction
the
quantum c h e m i s t s
Furthermore, is
(due
elements to
exponent
from
conclude from
5 f
are
CB plus
Experimental
and bonding o f
transition challenge
9
14).
e
polarity.
obtained
6d and 5f
definitely
the
Au H"
function with
plus
the
+
indicate
reference
including
configuration
R e p r o d u c e d w i t h p e r m i s s i o n from Ref. values
from
EB27
ab and
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
initio it
is
it
21.
MALLI
Chemistry of Third-Row Transition Elements and Actinides
gratifying for
that
performing
calculations
the computational reliable
for diatomics
It
i s hoped that
ab
initio
with
machinery fully
fully
c o n t a i n i n g heavy
is currently
relativistic
c o n t a i n i n g heavy
the a v a i l a b i l i t y
(all-electron)
polyatomics near
ab i n i t i o
and very
of faster
relativistic
atoms w i l l
hand
DF SCF heavy
atoms.
supercomputers
calculations
become
at
307
feasible
for i n the
future.
ACKNOWLEDGMENTS I
sincerely
inviting to
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch021
Yu, My for
Professors Dennis symposium.
Salahub
T h i s work
and Mike
Zerner
for
h a s b e e n made p o s s i b l e d u e
t h e c o o p e r a t i o n a n d t h e e n t h u s i a s m o f my c o l l e a g u e s a n d c o w o r k e r s
over to
thank
me t o t h i s
many y e a r s ;
in particular,
Dr. N.C. Pyper, for their thanks their
Research financial
also
D r . R.
contributions go t o
cordial
would
cooperation. through
like
t o a c k n o w l e d g e my d e b t
M e s s r s A . F . Ramos a n d D.
to the research reported
the operations
C o u n c i l o f Canada support
I
Arratia-Perez,
no.
is
in this
paper.
o f our Computing S e r v i c e s
The N a t u r a l
(NSERC) grant
staff
Sciences and Engineering
thanked
for their
continuous
A3598.
LITERATURE CITED 1. Schrödinger, Ε. Ann. Physik. 1926, 81, 109. 2. Klein, O. Z. Physik. 1926, 37, 895. 3. Gordon, W. Z. Physik. 1926, 40, 117. 4. Dirac, P. A. M. Proc. Roy. Soc. Lond. 1928, A117. 610. 5. Burrau, O. Kgl. Danske. Videnskab. Mat. Fys. 1927, 7, 14. 6. Dirac, P. A. M. Proc. Roy. Soc. Lond. 1928, A123. 714-33. 7. Mayers, D. F. Proc. Roy. Soc. Lond. 1957, A241. 93. 8. Swirles, B. Proc. Roy. Soc. Lond. 1935, A152. 625-49. 9. Boyd, R. G.; Larson, A. C.; Waber, J. T. Phys. Rev. 1963, 129, 1629-30. 10. Mingos, D. M. P. Phil. Trans. Roy. Soc. Lond. 1982, A308, 7583. 11. Hay, P. J.; Wadt, W. R.; Kahn, L. R.; Bobrowicz, F. W. J. Chem. Phys. 1978, 69, 984. 12. Ziegler, T.; Snijders, J. G.; Baerends, E. J. J. Chem. Phys. 1981, 74, 1271. 13. Jiang, Y.; Alarez, S.; Hoffmann, R. Inorg. Chem. 1985, 24, 749-57. 14. Malli, G. L.; Pyper, N. C. Proc. Roy. Soc. Lond. 1986, A407. 377404. 15. Grant, I. P.; Mckenzie, B. J.; Norrington, P. H.; Mayers, D. F.; Pyper, N. C. Comput. Phys. Commun. 1980, 21, 207. 16. Brown, G. E.; Ravenhall, D. G. Proc. Roy. Soc. Lond. 1951, A208. 552-9. 17. Sucher, J. Phys. Rev. 1980, A22, 348-62. 18. Huber, K. P.; Herzberg, G. Molecular Spectra and Molecular Structure IV. Constants of Diatomic Molecules: Van Nostrand Reinhold, New York, 1979. 19. Pyykkö, P. Chem. Rev. 1988, 88, 563. 20. Lee, Y. S.; McLean, A. D. J. Chem. Phys. 1982, 76, 735.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
THE C H A L L E N G E OF d AND f ELECTRONS
308
21. 22. 23. 24. 25.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch021
26. 27. 28. 29. 30.
Ramos, A. F.; Pyper, N. C.; Malli, G. L. Phys. Rev. 1988, A 38 2729-2739. Katz, J. J.; Seaborg, G. T.; Morss, L. R. The Chemistry of the Actinide Elements. Chapman and Hall: London, 1986. Oetting, F. L.; Rand, M. H.; Ackermann, R. J. The Chemical Thermodynamics of Actinide Elements and Compounds Part 1: International Atomic Energy Agency: Vienna, 1976. Oetting, F. L.; Fuger, J. The Chemical Thermodynamics of Actinide Elements and Compounds Part 2: International Atomic Energy Agency: Vienna, 1976. Erdos, P.; Robinson, J. M. The Physics of Actinide Compounds: Plenum Press: New York, 1983. Handbook on the Physics and Chemistry of the Actinides Vols. 15; Freeman, A. J.; Lander, G. H. Eds.; North Holland: Amsterdam, 1987. Grant, I. P.; McKenzie, B. J.; Norrington, P. H.; Mayers, D. F.; Pyper, N. C. Computer Phys. Commun. 1980, 21, 218. Ackermann, R. J.; Rauh, E. G. Higher Temp. Sci. 1973, 5, 463; J. Chem. Phys. 1974, 60, 2266. Hildenbrand, D. L.; Murad, E. J. Chem. Phys. 1974, 61, 1232. Malli, G. L.; Oreg, J. J. Chem. Phys. 1975, 63, 830-841.
RECEIVED March 21, 1989
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Chapter 22
Relativistic Effective Potentials in Quantum Monte Carlo Studies Phillip A. Christiansen
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch022
Department of Chemistry, Clarkson University, Potsdam, NY 13676
An overview of quantum Monte Carlo electronic structure studies i n the context of recent effective potential implementations i s given. New results for three electron systems are presented. As long as care i s taken i n the selection of t r i a l wavefunctions, and appropriate frozen core corrections are included, agreement with experiment i s excellent (errors less than 0.1 eV). This approach offers promise as a means of avoiding the excessive configuration expansions that have plagued more conventional transition metal studies.
In the l a s t ten years considerable e f f o r t has gone i n t o the study of small metal clusters. Several reviews on the subject have appeared i n the l i t e r a t u r e . Volume 156 of Surface Science, Volume 86 of Chemical Reviews and a portion of Volume 91 (especially No. 10) o f the Journal of Physical Chemistry are devoted t o t h i s topic. The small t r a n s i t i o n metal c l u s t e r s are p a r t i c u l a r l y i n t r i g u i n g as a r e s u l t of t h e i r unique structures (multiple d bonding, etc.) and as possible models f o r c a t a l y t i c processes. Furthermore, as can be seen from the compendium by Huber and Herzberg (1) and a l s o from the encyclopedic reviews by Weltner and Van Zee (2) and more r e c e n t l y by Morse Ç3) and by Salahub (4), r e l a t i v e l y l i t t l e i s known about the d e t a i l e d structures of even the simplest c l u s t e r s (diatomics) of the elements beyond the f i r s t t r a n s i t i o n row. The f i e l d would appear t o be wide open f o r computational chemists. For c l u s t e r s of only a few atoms one would expect rigorous e l e c t r o n i c structure studies (SCF plus large CI, etc.) t o y i e l d useful imformation regarding molecular geometries, d i s s o c i a t i o n energies, v i b r a t i o n a l frequencies, etc. Unfortunately, i n contrast t o recent l i g h t element work, e a r l y t r a n s i t i o n element studies proved somewhat disappointing. The chromium diatom (5-9) i s probably the best known example. However a more d i s t u r b i n g case i s SC2, the simplest t r a n s i t i o n metal diatomic. The SC2 d i s s o c i a t i o n energy (1.65 eV) i s known from the mass spectrometric work of Verhaegen e t a l . (10) although there may be some error due t o the use of rather imprecise molecular p a r t i t i o n functions (11,12). (A value of 1.22 eV was o r i g i n a l l y given but i t 0097-6156/89/0394-0309$06.00/0 ο 1989 American Chemical Society
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch022
310
THE CHALLENGE OF d AND f ELECTRONS
has since been reported i n various references (1,3) as 1.65.) Resonance Raman (13), ESR (14) and MCD (15) matrix i s o l a t i o n studies suggest a 5 Σ ground state with a fundamental v i b r a t i o n a l frequency of about 239cnf 1, which i s consistent with the assignment given by the best ab i n i t i o calculation. Nevertheless, i n t h e i r f a i r l y extensive study, Walch and Bauschlicher (12) were able t o account f o r only a f r a c t i o n of the experimentally determined d i s s o c i a t i o n energy. As with diatomic chromium the d i f f i c u l t y involves c o r r e l a t i o n i n a multiple d bonded system. These discouraging r e s u l t s have prompted Morse Ç3) t o suggest that f o r t r a n s i t i o n element problems the density functional approaches (4,17-19) might be more appropriate. Of course the t r a n s i t i o n metal electron c o r r e l a t i o n problems do not necessarily begin a t the molecular l e v e l . Ab i n i t i o studies (2024) t y p i c a l l y show errors i n atomic e x c i t a t i o n energies of about 0.3 eV or more f o r t r a n s i t i o n s i n v o l v i n g the outer s and d electrons. The errors seen i n the above examples are of course the r e s u l t of the necessary incompleteness of o r b i t a l and configuration basis sets. The power of these expansion approaches i s that i f one works hard enough (uses a s u f f i c i e n t l y complete, or a t l e a s t appropriate, basis) one should get the r i g h t answer. The recent extensive t r a n s i t i o n metal hydride studies i n d i c a t e the p o s s i b i l i t i e s (25-30). Nevertheless, heavy atom electron c o r r e l a t i o n involving d and even f subshells i s such an enormous problem that every a l t e r n a t i v e should be explored. E f f e c t i v e P o t e n t i a l Quantum Monte C a r l o As configuration expansions approach the m u l t i - m i l l i o n range, alternatives such as quantum Monte Carlo (QMC) techniques begin t o appear a t t r a c t i v e . A useful overview of QMC has been given by Ceperly and Alder (31). Pioneering work i n t h i s f i e l d was done i n the mid 70 s by Anderson (32,33) as w e l l as by Kalos and coworkers (34,35). This has been followed by considerable development work as w e l l as molecular and atomic applications (36-56). The advantage of QMC i s that i t does not depend on the exhaustive configuration and o r b i t a l basis set expansions that have plagued conventional studies. As one moves down the p e r i o d i c table t o the t r a n s i t i o n elements with occupied d s h e l l s and t o the Lanthanides and Actinides with f s h e l l s , the QMC advantage becomes more apparent. Unfortunately, t o a considerable extent, what one gains i n the e l i m i n a t i o n of i n f i n i t e basis set expansions, one looses t o s t a t i s t i c a l sampling error. And furthermore the sampling e r r o r increases r a p i d l y as a function of the nuclear charge. D o l l (57), Ceperly (58), and most recently Hammond e t a l . (59) have given arguments i n d i c a t i n g that the QMC computer requirements increase with about the s i x t h power of the nuclear charge. As a r e s u l t QMC has, t o date, offered l i t t l e competition f o r conventional calculations, and we are aware of no a l l - e l e c t r o n QMC studies i n v o l v i n g elements beyond the f i r s t row. Although a l l - e l e c t r o n heavy element QMC studies are a t the present time out of the question, we have recently shown that by replacing the core electrons (and the corresponding f r a c t i o n of the nuclear charge) with an appropriate r e l a t i v i s t i c e f f e c t i v e potential (REP) the QMC domain can be quite r e a d i l y extended t o the lower portion of the p e r i o d i c table (60-62). To our knowledge, reference (60) i s the f i r s t QMC study i n v o l v i n g an element from below the f i r s t
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
22.
CHRISTIANSEN
311
Quantum Monte Carlo Studies
row and i s a l s o the f i r s t t o include r e l a t i v i t y . This work was followed c l o s e l y by a study by Hammond e t a l . (59) who used an almost i d e n t i c a l approach i n a l k a l i and a l k a l i n e earth atomic and molecular studies. As pointed out i n the review by Ceperly and Alder (_31) the d i f f u s i o n interpretation of the Schroedinger equation has an extensive history. The d i f f u s i o n analogy becomes apparent i f one writes the time-dependent Schroedinger Equation (one electron f o r s i m p l i c i t y ) i n terms o f imaginary time, t , | f = |ν2ψ - (ν - Ε )ψ
(1)
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch022
τ
Ψ then corresponds t o a concentration and the equation i s simulated by a combination o f random p a r t i c l e movement ( f i r s t term on the right) as w e l l as p a r t i c l e b i r t h and death according t o the f i r s t order rate constant, V-Bp (32,33). The a r b i t r a r y reference energy, Erp, can be adjusted t o maintain normalization. However a f a r more e f f i c i e n t approach involves importance sampling (34,36,37,41,42). By defining the function ί=ψΨτ# the product o f Ψ with a time-independent t r i a l wavefunction, Ψφ, Equation 1 becomes df
^
,
Η
ο
= lv2f - V · (f-νΐηψτ) -
Ψτ - E )f.
(2)
T
The middle term on the r i g h t adds a d r i f t v e l o c i t y (νψ /ψ ) t o the simulation which greatly reduces sampling i n regions of low electron density. In addition the nodes i n (resulting from e i t h e r o r b i t a l nodes or antisymmetry) can be used t o define sampling region boundaries, an assumption of the f i x e d node approximation (32,33). E l e c t r o n i c energies are u l t i m a t e l y obtained from averages of the l o c a l energies, Η Ψ / ψ . Detailed descriptions of algorithms based on Equation 2 can be found i n the l i t e r a t u r e (31,39,42,52). For QMC simulations involving atomic o r molecular systems with more than a small number of electrons, p o t e n t i a l sources of d i f f i c u l t y are f a i r l y obvious (58-60). In regions of high electron density (such as near nuclei) one sees the corresponding high density of s i n g u l a r i t i e s i n the hamiltonian r e s u l t i n g from the two-electron e l e c t r o s t a t i c interaction. A t the same time, wavefunction antisymmetry causes a high nodal density. The dense nodal structure forces one t o employ short time steps, thereby g r e a t l y increasing the computational requirements. And although the e f f e c t s of electronnucleus and two-electron s i n g u l a r i t i e s i n the p o t e n t i a l can be c o n t r o l l e d t o considerable extent using p a i r - c o r r e l a t i o n functions [see reference (37) f o r instance], unless ψ τ i s a good approximation to ψ, c o r r e l a t i o n e r r o r w i l l become overwhelmingly apparent i n the l o c a l energies, leading t o large s t a t i s t i c a l errors i n the average. The more densely packed the electrons become (this w i l l be most serious i n the core region) the more acute the d i f f i c u l t i e s w i l l be. Furthermore, with the exception of the work by Vrbik e t a l . (63) i t i s not clear how r e l a t i v i t y (essential f o r heavy element studies) would be included i n a l l - e l e c t r o n QMC work. In t h i s context the advantages i n the use o f e f f e c t i v e potentials are quite clear. The potentials eliminate the high electron density (and associated nodes) near the n u c l e i , thereby τ
τ
τ
τ
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
312
THE
CHALLENGE O F d AND f ELECTRONS
reducing sampling i n a region i n which both the wavefunction and the potential are r a p i d l y varying. This eliminates a d i f f i c u l t f r a c t i o n of the multi-electron p o t e n t i a l from the wave equation and a t the same time makes the use o f much longer time steps appropriate. Perhaps equally important/ r e l a t i v i s t i c e f f e c t i v e potentials allow one t o introduce r e l a t i v i t y i n a p a r t i c u l a r l y convenient form. The key t o the use o f conventional semi-local REPs i n QMC involves the transformation o f the PEP t o l o c a l form. In reference (60) Hurley e t a l . proposed the many-electron l o c a l p o t e n t i a l , V ^ , 1
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch022
V
REP
s
υΚΕΡ
ψ τ /
ψ
Τ β
(
3
)
\βΕΡ i s the conventional e f f e c t i v e potential and i s the t r i a l wavefunction (at least the determinant portion) from Equation 2. To carry out REP-QMC calculations one simply adds t o the valenceelectron hamiltonian, H, i n the importance sampling algorithm. In l i g h t o f the l o c a l energy expression, Ηφτ/ψ^/ i n Equation 2, t h i s d e f i n i t i o n o f the l o c a l potential i s rather obvious and has a l s o been used i n the work by Hammond e t a l . (59) but without r e l a t i v i t y . Equation 3 obviously adds approximations. These include the usual e f f e c t i v e potential assumptions (frozen core, etc.) i n addition to the l o c a l i z a t i o n shown i n the equation. However i n one sense i t i s a trade-off i n that the l o c a l potential e f f e c t i v e l y eliminates the f i x e d node approximation i n the core region. Atomic Studies In Table I electron a f f i n i t i e s f o r L i , Na and Κ computed using Equation 3 with e i t h e r r e l a t i v i s t i c (60) o r n o n r e l a t i v i s t i c (59) e f f e c t i v e potentials are compared with the respective experimental values (64-66). Only i n the r e l a t i v i s t i c L i c a l c u l a t i o n do we see a s i g n i f i c a n t discrepancy, and even then the e r r o r i s w e l l below 0.1 eV. In a l l o f these calculations s i n g l e determinant t r i a l wavefunctions were employed. While t h i s i s no approximation f o r the one-electron neutral atoms we might see minor problems f o r the anions, and L i could be a case i n point. Table I. A l k a l i Electron A f f i n i t i e s ( i n eV) obtained from R e l a t i v i s t i c and N o n r e l a t i v i s t i c E f f e c t i v e Potential QMC Simulations Atom Li Li Na Κ
reference 60 59 II
60
e f f e c t i v e p o t e n t i a l QMC 0.56(2) 0.61(2) 0.56(2) 0.52(1)
Expt. 0.62 0.55 0.50
Conventional shape-consistent e f f e c t i v e potentials (67-70), whether r e l a t i v i s t i c o r not, are t y p i c a l l y formulated as expansions of l o c a l potentials, Ui,(r), m u l t i p l i e d by angular projection operators. The expansions are truncated a f t e r the lowest angular function not contained i n the core. The l a s t (residual) term i n the expansion t y p i c a l l y represents l i t t l e more than the simple coulombic i n t e r a c t i o n between a valence electron and the core (electrons and corresponding f r a c t i o n o f the nuclear charge) and i s predominantly a t t r a c t i v e . The lower & terms, on the other hand., include strongly
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
22.
CHRISTIANSEN
313
Quantum Monte Carlo Studies
repulsive "Pauli" contributions. This i s i l l u s t r a t e d i n Figure 1 where we have plotted the two terms i n the L i REP (68). Note that the difference between the two curves goes r a p i d l y t o zero f o r large values of r . A s i n g l e determinant s t r i a l wavefunction f o r L i would result i n a that included only the repulsive curve. However the i n c l u s i o n of s t o ρ promotions i n the t r i a l wavefunction would introduce a small contribution from the a t t r a c t i v e term and would tend t o lower the e l e c t r o n i c energy. Fortunately, f o r L i the anion electron density i s r e l a t i v e l y d i f f u s e and the correction quite small. In some cases however serious errors can r e s u l t from the use of such a simple t r i a l wavefunction i n Equation 3. The terms i n the Be REP are quite s i m i l a r t o those of L i , but the ground state electron d i s t r i b u t i o n i s considerably more compact and the c o r r e l a t i o n correction from the p2 configuration f a r more important. In Table I I we have l i s t e d SCF and REP-QMC energies f o r the lowest ^S, P, and D states of Be along with experimental values (71) f o r comparison. Numbers i n square brackets include core p o l a r i z a t i o n corrections 2
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch022
3
3
(22). 3
3
For the Be P and D excited states the c o r r e l a t i o n corrections are r e l a t i v e l y small and as can be seen from the table, the s i n g l e determinant approximation i n Equation 3 i s quite good. In contrast the *S state i s too high by 0.3 eV. However, as Equation 3 would suggest, a simple two-configuration ( 2 s + 2p2) wavefunction brings i n the a t t r a c t i v e "p" contribution and we get the value l i s t e d i n column four. Curiously, i n an a l l - e l e c t r o n QMC study Harrison e t a l . (74) observed a s i m i l a r d i f f i c u l t y with the Be ground state. They found that a single determinant t r i a l wavefunction gave the energy about 0.3 eV too high due t o the f i x e d node approximation. The use of a multiconfiguration t r i a l wavefunction eliminated the error. 2
Table I I . E f f e c t i v e Potential QMC Energies ( i n eV) f o r various states of Be and Mg State Be+ S Be D Be P Be i-S
reference
Z
3
3
II
Mg+ Mg Mg Mq
2
61 II II
SCF 0.0 -1.54 -6.39 -8.06
Single 0.0 -1.63(2) -6.56(2) -9.04(3)
0.0 -6.59
0.0 -7.58(1)
II
S
is
62
II
II
" "
59 73
Multi. 0.0
-9.32(1) [-9.34(1)] 0.0 -7.55(1) [-7.66(1)]
Expt. 0.0 -1.63 -6.59 -9.32 0.0 -7.65
-7.64(3) -7.57(3)
In the e f f e c t i v e p o t e n t i a l approximation Mg i s i s o e l e c t r o n i c with Be. But, as can be seen i n Figure 2, the Mg REP i s composed of t h r e e terms (s, ρ and d) w i t h the s and ρ both r e p u l s i v e . As a r e s u l t , even though the c o r r e l a t i o n correction i s almost as large as i n Be the multi-determinant correction r e s u l t i n g from Equation 3 i s only a tenth as b i g (see Table II). The discrepancy between values from references (62) and (59) i s due t o large s t a t i s t i c a l o r extrapolation error. Note that unlike Be one cannot make comparisons with experimental r e s u l t s without f i r s t taking core-valence
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch022
314
THE CHALLENGE OF d AND f ELECTRONS
r (a.u.) Figure
1.
Radial
consistent
plots
effective
of
the
s and ρ
potential.
(Data
terms are
of
from
a Li
shape
ref.
68.)
r(a.u.) Figure
2.
consistent
Radial
plots
effective
of
the
s,
potential.
p,
and d terms
(Data
are
from
of ref.
a Mg s h a p e 70.)
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
22.
CHRISTIANSEN
315
Quantum Monte Carlo Studies
c o r r e l a t i o n (72) i n t o account. For the heavier a l k a l i s and a l k a l i n e earths such corrections can amount t o several tenths o f an eV. The e f f e c t i v e p o t e n t i a l QMC studies discussed above have a l l employed more or less conventional shape consistent r e l a t i v i s t i c (60 61) o r n o n r e l a t i v i s t i c (59,62) e f f e c t i v e p o t e n t i a l s l o c a l i z e d according t o Equation 3. Yoshida and Iguchi (73) on the other hand have recently published Mg, Ca and Sr studies employing model potentials o f the type developed by Huzinaga e t a l . (75). By comparison with the nodeless o r b i t a l s i n the preceeding studies the model p o t e n t i a l approach employs representations o f normal HartreeFock valence o r b i t a l s . In reference (73) the p o t e n t i a l was j u s t the coulombic i n t e r a c t i o n between normal core and valence o r b i t a l s which i s e s s e n t i a l l y equivalent t o the residual term i n the shape consistent REPs. The advantage t o t h i s approach i s that the potential i s already i n simple l o c a l form and there i s no need f o r Equation 3. The dissadvantage i s that the t r i a l wavefunctions include numerous nodes i n the core region (which forces one t o employ shorter time steps) and a l s o may include a s i z a b l e amplitude near the nucleus which f o r Gaussian basis sets might require the use of nuclear cusp functions. In the QMC simulation t h i s approach almost looks more l i k e a frozen core study rather than e f f e c t i v e p o t e n t i a l . An additional (not nearly so w e l l understood) complication i s that i n the simple shielded nucleus potential a t r i a l wavefunction formed from the valence Hartree-Fock o r b i t a l s looks l i k e an excited state. One must therefore choose the t r i a l wavefunction c a r e f u l l y t o ensure orthogonality t o the f a l s e lower energy solutions. The above studies a l l involved only one and two-electron systems. And with the exception of the Be high spin excited states (61) none required the use of "Fermi s t a t i s t i c s " (wavefunction antisymmetry) i n the Monte Carlo simulations. This i s o f course a prerequisite f o r multi-electron systems. We have recently c a r r i e d out REP-QMC simulations on some three-electron systems. Aluminum i s probably the simplest. In Table I I I we show energies f o r two states o f A l and a l s o f o r A l .
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch022
f
+
Table I I I . Comparison of SCF, REP-QMC and Experimental Energies ( i n eV) f o r various states o f A l State A1+ J-S Al P Al P 4
2
SCF 0.0 -3.11 -5.48
Single 0.00(2) -2.42(3) -5.92(4)
Multi. 0.00(1) [-2.42(2)] -5.92(2)
Expt. 0.0 -2.38 -5.98
For the A l ground state and a l s o f o r the cation we c a r r i e d out simulations using both s i n g l e and multiple determinant t r i a l wavefunctions. (The brackets indicate a s i n g l e determinant t r i a l function f o r the P state but multiple f o r the S.) As opposed t o e i t h e r Be or Mg we see no s i g n i f i c a n t adjustment i n the energies r e s u l t i n g from the use o f the more accurate t r i a l functions. We do however see a reduction i n the s t a t i s t i c a l error. The agreement with experiment i s excellent except f o r the 0.06 eV systematic error i n the ground state. This could be due t o a s t i l l inadequate t r i a l wavefunction i n Equation 3, but based on Muller's Mg work (72) we a l s o suspect core p o l a r i z a t i o n . This remains t o be determined. As indicated i n the introduction a major motivation i n the 4
l
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
THE CHALLENGE OF d AND f ELECTRONS
316
development of REP-QMC involves applications t o t r a n s i t i o n element problems. Although t o our knowledge there are as yet no such applications i n the l i t e r a t u r e we have recently been running preliminary three-electron simulations involving the D ground and F excited states o f Sc and Y. Our r e s u l t s so f a r have been somewhat disappointing. For both elements we see e x c i t a t i o n energy errors of about 0.5(2) eV which are two t o three times what we would have anticipated f o r the three-electron approximation. The poor r e s u l t s could be due t o the s i n g l e determinant t r i a l functions but i t might a l s o be the r e s u l t o f core p o l a r i z a t i o n or some other aspect o f the frozen-core approximation. We may be forced t o use 11-electron REPs t o achieve adequate results. I n t h i s event, a recently developed "frozen core" approach (76) could prove invaluable i n combination with the 11-electron REP. Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch022
2
4
Molecules To date the only molecular e f f e c t i v e p o t e n t i a l QMC r e s u l t s that we are aware o f are those o f Hammond e t a l . (59). The d i f f i c u l t y i n extending t h i s work t o diatomic or polyatomic systems involves the e f f i c i e n t evaluation o f the l o c a l p o t e n t i a l , V * ^ , i n conjunction with the projection operators i n the p o t e n t i a l and basis functions centered on d i f f e r e n t n u c l e i . In the NaH and Ν&2 work o f reference (59) the angular projection integrations were c a r r i e d out by more o r l e s s conventional means. In preliminary work however we have found a simple a l t e r n a t i v e useful. Since the r a d i a l functions i n ifiEP decay very r a p i d l y with increasing r , the product o f the Ufl,(r) with a function centered on another nucleus can be approximated quite accurately by short one-center expansions. Conveniently, the angular part o f the expansion need not go beyond the highest £ quantum number i n the core. For f i r s t row atoms ( L i , Be, C, etc.) only a small number o f s functions are required. Κ o r Mg studies would require only s and ρ functions. In preliminary studies we have found that f o r K2 the use o f s functions alone r e s u l t s i n an e r r o r o f only about a tenth o f an eV due t o the neglect o f ρ functions i n the one-center expansion. For l i n e a r molecules a p a r t i c u l a r l y e f f i c i e n t scheme (in terms of the QMC sampling) might be t o tabulate the product o f the ifi^ with each o r b i t a l on a coarse two-dimensional grid. In t h i s sense one can see that an i n t e r e s t i n g approach t o diatomic REP-QMC could be based on numerically determined SCF o r MCSCF t r i a l wavefunctions (77,78). Discussion Quantum Monte Carlo techniques have considerable p o t e n t i a l f o r a p p l i c a t i o n t o problems involving open d o r f s h e l l s where the treatment o f electron c o r r e l a t i o n has proven p a r t i c u l a r l y d i f f i c u l t . However i f QMC i s t o be a v i a b l e a l t e r n a t i v e one must be able t o l i m i t the simulations t o small numbers o f electrons and i n addition r e l a t i v e i t y must be included. R e l a t i v i s t i c e f f e c t i v e potentials o f f e r one avenue (at the present time the only avenue) f o r achieving these conditions. However, as we have indicated, REPs do introduce complications. Because o f Equation 3 one must be somewhat more c a r e f u l i n the s e l e c t i o n o f t r i a l wavefunctions and multiconfiguration algorithms are e s s e n t i a l , unfortunately the additional configurations
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
22.
CHRISTIANSEN
Quantum Monte Carlo Studies
necessarily increase the computer requirements. For the Be ground state f o r instance the addition of the p configuration j u s t about doubles the amount of processing per QMC time step. On the other hand as one can see from the Tables the sampling error i s t y p i c a l l y reduced by more than a factor of two. Noting that the sampling error varies inversely with the square root of the number of samples we see that the use of multiconfiguration t r i a l wavefunctions a c t u a l l y reduces the o v e r a l l computer requirements (for a given error level) by more than a factor of two. This i s consistent with the arguments given i n references (58,59) and with the r e s u l t s of Moskowitz e t a l . (48,49). Though not discussed above, i n a l l the studies mentioned the t r i a l wavefunctions included p a i r c o r r e l a t i o n functions, J j i , as prescribed by Reynolds e t a l . (42). Moskowitz e t a l . (48,49) have shown that the product of a r e l a t i v e l y simple multiconfiguration wavefunction with p a i r c o r r e l a t i o n functions can provide a rather accurate approximation to the exact wavefunction. In our calculations and i n those of Hammond e t a l . (59) the many-electron l o c a l p o t e n t i a l , V ^ , has been obtained by allowing the REP t o operate only on the determinantal portions of the t r i a l wavefunction. The e f f e c t s of the p a i r c o r r e l a t i o n functions have been ignored. As pointed out i n (61) the e f f e c t s of the p a i r c o r r e l a t i o n functions on the l o c a l potentials could be included by means of zeta-function expansions (79). However i n our multiconfiguration c a l c u l a t i o n s the J i j were parametrized t o correct f o r short range d i f f i c u l t i e s only, (We assumed that the configuration expansions properly accounted f o r long-range and near-degeneracy e f f e c t s ) , and we would therefore expect the J ^ j to have only a n e g l i g i b l e e f f e c t on the V^^p. A l l of the e f f e c t i v e p o t e n t i a l QMC studies that we are aware of have employed r e l a t i v e l y simple fixed-node d i f f u s i o n Monte Carlo algorithms. This i s not t o suggest that these are preferable, but rather easy t o program. One should not underestimate the advantages of Green's Function [see references (50,51) f o r instance] or other more recently developed approaches (80). E s s e n t i a l l y a l l of the e f f e c t i v e p o t e n t i a l QMC work i n the l i t e r a t u r e t o date has been, or could be, c a r r i e d out on small mini or microcomputers. For instance, using QMC algorithms s i m i l a r t o those of references (39) and (52), our Be, Mg and A l (two and three electron) studies required around 50 t o 200 MicroVAX hours f o r s i n g l e atomic state energies with standard errors of about 0.01 eV. However work on larger systems w i l l obviously require more powerful computing machinery. Although we would not a n t i c i p a t e dramatic increases i n speed due t o vectorization, (at l e a s t not f o r problems involving smaller numbers of electrons) QMC algorithms are almost t r i v i a l l y adaptable t o "massively p a r a l l e l " computing environments (81-83). By simply d i s t r i b u t i n g configurations (particles) evenly among processors one should be able t o obtain near peak p a r a l l e l e f f i c i e n c y . The only complication we forsee would be due t o the occasional renormalization required by the p a r t i c l e m u l t i p l i c a t i o n and destruction events. This would necessitate the communication of p a r t i c l e coordinates, etc. between processors t o maintain an even configuration d i s t r i b u t i o n . Fortunately the time required t o t r a n s f e r configurations between processors w i l l be proportional t o the number of electrons per configuration whereas the processing time (per time step) i s proportional to the number of electrons squared or 2
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch022
317
1
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch022
318
THE CHALLENGE OF d AND f ELECTRONS
cubed (42). As a r e s u l t one would expect i n e f f i c i e n c i e s due t o processor imbalance t o become less important f o r bigger problems. Furthermore f o r "hypercube" machines (81-83) with a thousand o r fewer processors the transfer o f configurations could probably be l i m i t e d to adjacent nodes, thereby minimizing the transfer time. Massively p a r a l l e l (multiple instruction, m u l t i p l e data) computers with tens o r hundreds of processors are not r e a d i l y accessible t o the majority o f quantum chemists a t the present time. However the cost o f currently available hypercube machines with tens of processors (each with about the power of a VAX) i s comparable t o that o f supermini s but with up t o a hundred times the power. F o r applications of the type discussed above the performance of a machine with as few as 32 o r 64 processors would be comparable t o (or perhaps even exceed) that of a s i n g l e processor supercomputer. Although computer requirements currently l i m i t QMC applications (even with e f f e c t i v e potentials) the p r o l i f e r a t i o n of inexpensive massively p a r a l l e l machines could conceivably make the application of r e l a t i v i s t i c e f f e c t i v e potentials with QMC quite competitive with more conventional e l e c t r o n i c structure techniques. Our REP-QMC work (60-62) along with the studies by Hammond e t a l . (59) provide evidence that with the proper precautions the combination of r e l a t i v i s t i c e f f e c t i v e potentials with quantum Monte Carlo procedures may provide an a l t e r n a t i v e f o r obtaining accurate e l e c t r o n i c structure information. The possible elimination o f excessive basis set and configuration expansions f o r the t r a n s i t i o n and heavier elements i s e s p e c i a l l y appealing. And a t the same time the t r a n s i t i o n t o p a r a l l e l computing i s p a r t i c u l a r l y simple. The p o s s i b i l i t y o f carrying out d e f i n i t i v e calculations on small t r a n s i t i o n metal c l u s t e r s should not be dismissed offhand. Although we would not expect QMC (or REP-QMC) t o replace conventional approaches (for one thing an accurate conventional t r i a l wavefunction appears t o be an essential prerequisite) the electron c o r r e l a t i o n problem f o r elements containing occupied d and f subshells i s p o t e n t i a l l y so enormous that a l l possible avenues should be thoroughly researched. Acknowledgments This work has been supported i n part by the National Science Foundation under Grant No. CHE-8214665 and also by the Research Corporation.
Literature Cited 1. 2. 3. 4. 5. 6.
K.P. Huber and G. Herzberg, Molecular Spectra and Molecular Structure IV. Constants of Diatomic Molecules, Van Nostrand Reinhold Company: New York, 1979. W. Weltner and R.J. Van Zee, Ann. Rev. Phys. Chem. 1984, 35, 291. M.D. Morse, Chem. Rev. 1986, 86, 1049. D.R. Salahub, Adv. Chem. Phys. 1987, 69, 447. M.M. Goodgame and W.A. Goddard, I I I , J. Phys. Chem. 1981, 86, 215. S.P. Walch, C.W. Bauschlicher, J r . , B.O. Roos and C.J. Nelin, Chem. Phys. Lett. 1983, 103, 175.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
22. CHRISTIANSEN
7. 8. 9. 10. 11. 12. 13. 14.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch022
15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.
Quantum Monte Carlo Studies
319
G.P. Das and R.L. Jaffe, Chem. Phys. Lett. 1984, 109, 206. A.D. McLean and B. Liu, Chem. Phys. Lett. 1983, 101, 144. M.M. Goodgame and W.A. Goddard, III, Phys. Rev. Lett. 1985, 54, 661. G. Verhaegen, S. Smoes and J. Drowart, J. Chem. Phys. 1964, 40, 239. G. Das, Chem. Phys. Lett. 1982, 86, 482. S.P. Walch and C.W. Bauschlicher, Jr., J. Chem. Phys. 1983, 79, 3590. M. Moskovits, D.P. DiLella and W. Limm, J. Chem. Phys. 1984, 80, 626. L.B. Knight, Jr., R.J. Van Zee and W. Weltner, Jr., Chem. Phys. Lett. 1983, 94, 296. L.B. Knight, Jr., R.W. Woodward, R.J. Van Zee and W. Weltner, Jr., J. Chem. Phys. 1983, 79, 5820. R.J. Singer and R. Grinter, Chem. Phys. 1987, 113, 99. N.A. Baykara, B.N. McMaster and D.R. Salahub, Mol. Phys. 1984, 52, 891. B. Delley, A.J. Freeman and D.E. Ellis, Phys. Rev. Lett. 1983, 50, 488. J. Bernholc and N.A.W. Holzwarth, Phys. Rev. Lett. 1983, 50, 1451. C.W. Bauschlicher, Jr., S.P. Walch and H. Partridge, J. Chem. Phys. 1982, 76, 1033. K.K. Sunil and K.D. Jordan, J. Chem. Phys. 1985, 82, 873. C.M. Rohlfing and R.L. Martin, Chem. Phys. Lett. 1985, 115, 104. S.R. Langhoff and C.W. Bauschlicher, Jr., J. Chem. Phys. 1986, 84, 4485. C.W. Bauschlicher, Jr., J. Chem. Phys. 1987, 86, 5591. D.P. Chong, S.R. Langhoff, C.W. Bauschlicher, Jr., S.P. Walch and H. Partridge, J. Chem. Phys. 1986, 85, 2850. S.R. Langhoff, L.G.M. Pettersson, C.W. Bauschlicher, Jr. and H. Partridge, J. Chem. Phys. 1987, 86, 268. L.G.M. Pettersson, C.W. Bauschlicher, Jr., S.R. Langhoff and H. Partridge, J. Chem. Phys. 1987, 87, 481. A.E. Alvarado-Swaisgood, J. Allison and J.F. Harrison, J. Phys. Chem. 1985, 89, 2517. J.B. Schilling, W.A. Goddard III and J.L. Beauchamp, J. Am. Chem. Soc. 1986, 108, 582. C.W. Bauschlicher, Jr. and S.P. Walch, J. Chem. Phys. 1982, 76, 4560. D.M. Ceperly and B.J. Alder, Science 1986, 231, 555. J.B. Anderson, J. Chem. Phys. 1975, 63, 1499. J.B. Anderson, J. Chem. Phys. 1976, 65, 4121. M.H. Kalos, D. Levesque and L. Verlet, Phys. Rev. A 1974, 9, 2178. D.M. Ceperly, G.V. Chester and M.H. Kalos, Phys. Rev. Β 1977, 16, 3081. J. B. Anderson, J. Chem. Phys. 1980, 73, 3897. F. Mentch and J.B. Anderson, J. Chem. Phys. 1981, 74, 6307. F. Mentch and J.B. Anderson, J. Chem. Phys. 1984, 80, 2675. J.B. Anderson, J. Chem. Phys. 1985, 82, 2662. D.R. Garmer and J.B. Anderson, J. Chem. Phys. 1987, 86, 4025. D.M. Ceperley and B.J. Alder, Phys. Rev. Lett. 1980, 45, 566.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
320 42. 43. 44. 45. 46. 47. 48.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch022
49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71.
72. 73. 74.
THE CHALLENGE OF d AND f ELECTRONS P.J. Reynolds, D.M. Ceperley, B.J. A l d e r and W.A. L e s t e r , J r . , J . Chem. Phys. 1982, 77, 5593. P.J. Reynolds, M. Dupuis and W.A. L e s t e r , J r . , J . Chem. Phys. 1983, 82, 1983. R.N. B a r n e t t , P.J. Reynolds and W.A. L e s t e r , J r . , J . Chem. Phys. 1985, 82, 2700. R.N. B a r n e t t , P.J. Reynolds and W.A. L e s t e r , J r . , J . Chem. Phys. 1986, 84, 4992. R.N. B a r n e t t , P.J. Reynolds and W.A. L e s t e r , J r . , J . Chem. Phys. 1987, 91, 2004. P.J. Reynolds, R.N. B a r n e t t , B.L. Hammond and W.A. L e s t e r , J r . , J . S t a t . Phys. 1986, 43, 1017. J.W. Moskowitz, K.E. Schmidt, M.A. Lee and M.H. K a l o s , J . Chem. Phys. 1982, 76, 1064. J.W. Moskowitz, K.E. Schmidt, M.A. Lee and M.H. K a l o s , J . Chem. Phys. 1982, 77, 349. D.W. Skinner, J.W. Moskowitz, M.A. Lee, P.A. W h i t l o c k and K.E. Schmidt, J . Chem. Phys. 1985, 83, 4668. K.E. Schmidt and J.W. Moskowitz, J . S t a t . Phys. 1986, 43, 1027. J . V r b i k and S.M. R o t h s t e i n , J . Comput. Phys. 1986, 63, 130. J . Vrbik, J . Phys. A 1985, 18, 1327. J . V r b i k and S.M. R o t h s t e i n . I n t . J . Quant. Chem. 1986, 29, 461. S.M.Rothstein, N. Patil and J . V r b i k , J . Comput. Chem. 1987, 8, 412. S.M. R o t h s t e i n and J . V r b i k , J . Comput. Phys. 1988, 74, 127. J.D. D o l l , Chem. Phys. L e t t . 1981, 81, 335. D.M. Ceperley, J . Stat. Phys. 1986, 43, 815. B.L. Hammond, P.J. Reynolds and W.A. L e s t e r , J r . , J . Chem. Phys. 1987, 87, 1130. M.M. H u r l e y and P.A. C h r i s t i a n s e n , J . Chem. Phys. 1987, 86, 1069. P.A. C h r i s t i a n s e n , J . Chem. Phys. 1988, 88, 4867. P.A. C h r i s t i a n s e n and L.A. LaJohn, Chem. Phys. L e t t . 1988, 146, 162. J . V r b i k , M.F. DePasquale and S.M. R o t h s t e i n , J . Chem. Phys. 1988, 88, 3784. D. Feldmann, Phys. A 1976, 277, 19. R.D. Mead, P.A. Schulz and W.C. Lineberger, Phys. Rev. A. J . S l a t e r , F.H. Read, S.E. Novick and W.C. L i n e b e r g e r , Phys. Rev. A 1978, 17, 201. P.A. C h r i s t i a n s e n , Y.S. L e e and K.S. P i t z e r , J . Chem. Phys. 1979, 71, 4445. L.F. P a c i o s and P.A. C h r i s t i a n s e n , J . Chem. Phys. 1985, 82, 2664. P.J. Hay and W.R. Wadt, J. Chem. Phys. 1985, 82, 270. W.J. Stevens, H. Basch and M. Krauss, J . Chem. Phys. 1984, 81, 6026. C.E. Moore, Atomic Energy L e v e l s , N a t l . Bur. Stand. C i r c . 467, Vols. I and I I , U.S. Government P r i n t O f f i c e : Washington, DC, 1949, 1952. W. M u l l e r , J . F l e s c h and W. Meyer, J . Chem. Phys. 1984, 80, 3297. T. Yoshida and K. I g u c h i , J . Chem. Phys. 1988, 88, 1032. R.J. H a r r i s o n and N.C. Handy, Chem. Phys. L e t t . 1985, 113, 257.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
22.
75. 76. 77. 78. 79. 80. 81. 82. 83.
CHRISTIANSEN
Quantum Monte Carlo Studies
321
S. Huzinaga, L. Seijo, Z. Barandiaran and M. Klobukowski, J. Chem. Phys. 1987, 86, 2132. W.A. Lester, Jr., private communication. E.A. McCullough, Jr., J. Chem. Phys. 1975, 62, 3991. P. Pyykko, G.H.F. Diercksen, F. Muller-Plathe and L. Laaksonen, Chem. Phys. Lett. 1987, 134, 575. M.P. Barnett, Methods Comput. Phys. 1963, 2, 95. J.B. Anderson, J. Chem. Phys. 1987, 86, 2839. R.A. Whiteside, J.S. Binkley, M.E. Colvin and H.F. Schaefer III, J. Chem. Phys. 1987, 86, 2185. Science News 1988, 133. J.L. Gustafson, G.R. Montry and R.E. Benner, SIAM J. Sci. and Stat. Computing 1988, 9.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch022
RECEIVED February 16, 1989
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Chapter 23
Relativistic Effects on Compounds Containing Heavy Elements The Influence of Kinetic Energy on Chemical Bonds 1
2
2
T. Ziegler , J . G. Snijders , and E. J . Baerends
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch023
1
Department of Chemistry, University of Calgary, Calgary, Alberta T2N 1N4, Canada Department of Theoretical Chemistry, The Free University, Amsterdam, Netherlands
2
It
i s shown that
relativity
will
reduce the
k i n e t i c energy of the e l e c t r o n s i n a number of compounds c o n t a i n i n g heavy elements. The
reduction o f the k i n e t i c energy leads t o
bond s t a b i l i z a t i o n and bond c o n t r a c t i o n and
i n f l u e n c e s s i g n i f i c a n t l y the chemistry
of t h i r d row t r a n s i t i o n
finite nuclei
Valence
e l e c t r o n s i n atoms
(albeit
small)
and they
instantaneous valence in
metals.
probability
c a n as
velocities.In
a
and molecules of being
this
acquire
fact,the velocities
e l e c t r o n s can approach t h a t o f l i g h t
reason
become
close to the
consequence
close proximity t o heavier n u c l e i with not too surprising
o f importance
compounds
series
5d-block
or 5f-block
as t h e y
relativistic
pass
i sfor effects
properties of
elements elements
high
for the
Ζ >72.It
f o r the chemical
containing
transition
that
have a
i n the third
i n the actinide
series. We s h a l l
here
d i s c u s s how r e l a t i v i s t i c
, r e l a t e d t o t h ehigh instantaneous v e l o c i t i e s near
heavy
involving
nuclei,
will
influence
5d- and 5f-elements.In
effects
of electrons
the chemical
particular,we
0097-6156/89/0394-0322$06.00/0 c 1989 American Chemical Society
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
bond shall
23.
ZIEGLER ET AL.
Relativistic Effects on Compounds of Heavy Elements
demonstrate t h a t the in
bonds the
and
relativity
contract
k i n e t i c energy of
illustrated
how
p e r i o d i c trends The
i n many c a s e s w i l l
the
bond
the
relativistic
within
by
Pitzer
a triad
(1)
and
Slater in
method
to
some c a s e s by
method
by
studies Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch023
due
(£)
due
It
the
to
extend
hydrogen
retardation
atom as
1
written
as
- \
V i=l
H
D
(Zr2L) · The
relativistic and
the
will
been
results
Hartree-Fock-
Baerends
analysis
e f f e c t s . D i r a c s Hamiltonian be
be
(4.) , augmented functional
build
on
previous
Baerends.
Theories
the
neglecting
influence
r e c e n t l y proposed density .Our
is possible for
reduction further
quantum c h e m i s t r y has
t o Z i e g l e r , S n i j d e r s and
Relativistic
theory
effects
Pyykko
Snijders
the
Becke
a
of t r a n s i t i o n metals.
by
p r e s e n t e d h e r e a r e b a s e d on
strengthen
by
electrons.lt will
f i e l d of r e l a t i v i s t i c
reviewed
distances
323
1
h (i)
(7_)
relativistic
η-electron
well
as
systems
certain
by
magnetic
f o r a many e l e c t r o n s y s t e m
V (rt)+f
+
D
Dirac's to
N
i=l
Xl/|fïi*j
rj|
can
(1)
where
h (i) and
VN(r~i) In
i s the
p(i)
matrices,
electron-nuclear
a?(i),
Eq. (1)
i s the
2
c($>(±) . p ( i ) +
=
D
j = l,2,3
c p
(2)
attraction potential.
and
β
are
the
4x4
momentum o p e r a t o r , a n d c t h e v e l o c i t y
of l i g h t . T ^ e corresponding Η ° Ψ ° = E F°
D i r a c wave e q u a t i o n
reads
DV
where
the
wave
combination
of
function
solve
due
to
proposed
can
be
S l a t e r determinants
Dirac
components.
(3) Ψ°
the
presence and
as
constructed
a
from
,
linear four-
{a }.
wave e q u a t i o n
Foldy
expressed
1
component D i r a c - s p i n o r s The
Dirac
of
is
four,in
Wouthuysen
a transformation
somewhat (£_)
which a l l o w s
cumbersome
general have one
to
to
complex,
fortunately approximate
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
324
THE CHALLENGE OF d AND f ELECTRONS
the four-component Dirac equation,Eq.(3),by a twocomponent S c h r O d i n g e r e q u a t i o n i n t h e f a m i l i a r 2x2 P a u l i r e p r e s e n t a t i o n t o any g i v e n o r d e r i n t h e f i n e s t r u c t u r e constant a.The Hamiltonian H o f E q . (1) t a k e s a f t e r t h e A
D
Foldy-Wouthuysen t r a n s f o r m a t i o n t h e form sum o f t e r m s i n i n c r e a s i n g o r d e r s o f (X
o f an
infinite
2
A
H
FW
=
where
H° + a Hi + a H 2
H°
is
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch023
combination
the
*F
eigenfunctions
6
+ 0(a )
4
FW
(4),
non-relativistic
of
of Slater
2
F W
H
c a n be
determinants
Hamiltonian.The
written
as
a
linear
c o n s t r u c t e d from
o r b i t a l s as i n t h e n o n - r e l a t i v i s t i c c a s e . The method o f £ n i j d e r s and B a e r e n d s (4) in E q . (4) up t o OC Hi ,where H^ i s g i v e n by
spin
retains
terms
2
Hi The the
=
H
M V
+
H
first
order
2
(in a )
non-relativistic kinetic T
is
Dar
w + H
S 0
= - \
N R
(5) .
Λ
H M V Iwhich
s o - c a l l e d m a s s - v e l o c i t y term
relativistic energy
Σ V? i
represents
correction
t o the
operator
=
j
Σ i
P (i) 2
(6)
g i v e n by HMV
whereas
= - | Σ i
Vf
t h e Darwin
electrons
=
4
term,from
(ϋ_) , a f t e r
insignificant
~ 3" Σ P ( i ) i the Zitterbewegung
neglecting
two-electron
(7a)
operators
some
of the
numerically
(£a_) , t a k e s
on t h e
form Hoarw =
| Σ i
V?
(7b)
(V (ii)) N
Λ The
spin-orbit
total
energy
operator
Hso
does
of the closed s h e l l
t h e f o l l o w i n g and need n o t t h u s The energy ,with
relativistic
,is
obtained
wave
be s p e c i f i e d
considered i n here.
f u n c t i o n , and r e l a t e d
by f i r s t
the n o n - r e l a t i v i s t i c
not c o n t r i b u t e t o the
molecules
order
wave
total
perturbation theory
function
as
zero-order
solution.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
23.
Relativistic Effects on Compounds of Heavy Elements
ZIEGLER ET AL.
R e l a t i v i s t i c C a l c u l a t i o n s on M e t a l Dimers and Hydrides .Relativistic Bond Contraction Relativistic Bond Stabilization We
present
bond e n e r g i e s for
the
simple
M=Cu,Ag,and 'with
Au
MH
(4_)
Results
from M
Hg.Table
a similar
have
a
distances,as compounds
frequencies
coinage
isoelectronic I
contains
triad
series
MH
+
experimental (JUL) and
s e t o f c a l c u l a t i o n s on t h e as w e l l
are presented
follows
effects
the
H a r t r e e - F o c k - S l a t e r (HFS) c a l c u l a t i o n s .
(M=Zn,Cd,and Hg) It
of
Metal and
c a l c u l a t i o n s on
from n o n - r e l a t i v i s t i c
(M=Cu,Ag,and Au)
2
from
and v i b r a t i o n a l
as t h e
as w e l l as r e s u l t s
relativistic dimers
I results
hydrides
as w e l l
M=Zn,Cd,and
d a t a (10)
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch023
i n Table
,bond d i s t a n c e s
325
from
Tables
sizable
well
as
containing
M2
+ 2
i n Table I I . I and
II that
influence
on
vibrational the
metal-
as t h e d i c a t i o n s
bond
relativistic energies,bond
frequencies
heavier
5d-member
f o r the
(Au
or
Hg)
w i t h i n the t r i a d . In t h e n o n - r e l a t i v i s t i c the
wrong
order(compared
c a s e bond e n e r g i e s
with
experiment)
row>second
row>third-row.Relativistic
for
Hg
Au
and
,provide
stabilize
on t h e o t h e r
row>first
row>second
row
and
in
contract
distances proximity
the
of
by
and
0.3
third
Â
- 1
third limit,
rowsecond
compounds
.An
i s c a u s e d b y an
dfirst analysis increase
orbital
as
we
decend t h e t r i a d . c. S y n e r g i c
Metal-Ligand
Bonds.
A number o f l i g a n d s Cp" ( c y c l o p e n t a d i e n y l ) acceptors.That
is,they
s u c h as CO,O2,C2H2, C2H4, CôHg, a n d
bind have
to empty
metal
centers
π-orbitals
as
π-
capable
of
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
334
THE CHALLENGE OF d AND f ELECTRONS
accepting orbitals
electron density ,as
addition ability
from
σ-type
occupied
occupied
in la
i t i s shown f o r CO
π-type
or
metal
.The
based
d-
l i g a n d s have i n
orbitals
with
the
metal-orbitals(lb).
t o d o n a t e d e n s i t y t o empty
o ®
ô
V
υ 1c
1b
1a A number o f d e n s i t y Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch023
Four-electron two-orbital repulsion
Ligand metal donation
Metal-ligand (back)donatlon
f u n c t i o n a l c a l c u l a t i o n s on
carbonyls (lia)
, c o m p l e x e s of 02,C2H2, and
as
2
X ,CX,CX ,H CX 2
have
2
shown
bonds from
the
primarily the
metal
(X=0,S,Se,Te)
that
metal
is
due
center
carbonyls)
reached
strength to
in studies
the
and
same
conclusion
based
on
other
well (17d)
6
metal-ligand
(back) d o n a t i o n ligands (la
as
C6H ,Cp~
synergic
the
to
.The
(17c) of
metal-
C2H4 (17b)
of
i n the has
methods
charge case
also (12)
,see
of
been for
references. In an
attempt
found(17e) allowed
that
to
to q u a n t i f y the
operate
significantly the
acts
sum
of
Our
they
bond e n e r g y i n c r e a s e s by
investigations in
homologous row~third Figure
the
and
4
do
was are
indeed
interaction
50
% compared
π-back-bonding
the
by
from their
the
the 4d
energy and
case
M(CO)s
that
and
consequence (17a)
to
5d
when
the
of
of
the
each
the
(17b)
donor
metal
π-acceptor effects to
higher
are
first
found
row>third
as
they
on
the
in
M (CO) 6 which
ligand
orbitals
of t h i s
shown
trend,
of
orbitals
a
row>second
carbonyls to
the
,for
,as
3d
orbitals
for a qualifier
stability
first
(M=Fe,Ru,Os).This from
that
bonds
limit
metal
3 d - o r b i t a l s are
better
further
order
c o u n t e r p a r t s . The
Relativistic order
the
non-relativistic
(back) d o n a t i o n
fact
indicate
metal-ligand
,follow
i n the
in
(17.)
synergic
series row
(M=Cr,Mo,W) and
in
simultaneously
σ-bonding
the
it
π-back-bonding
alone.
strengths
set
synergic effect and
r e i n f o r c e each other.The o r b i t a l
c o n t r i b u t i o n t o the to
σ-bonding
i f the
,
energy are are
is
stems than as
a
closer
ligands(see
rationalization). (H)
to
change
row>second
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
row
the by
23.
ZIEGLER ET AL.
stabilizing
the
transition in
bonds i n v o l v i n g
series
Figure
also
Relativistic Effects on Compounds ofHeavy Elements
4.
A
,as
similar
illustrated
Figure
5.The
not
involved
in
electron
interactions
but
also
,which
rise
to
energy
particular
We
shall
effects
compounds
ligand
occupied
are
constitute
related
finally
compounds type the
for
of
of
involving
the
to
actinides
,by
Table I I I . C a l c u l a t e d and (R=CH H)
Bond
Energies
1 8
four-
in
the
metals,in electron
orbitals.
Energies
in
which
relativistic
bond
energies
representing
(Kcal
are
metal
,reduce
5p
Bond
for
a
kinetic
four
RMCI3 w i t h M=Th,U and
on
the
row
and
degree to
in
the
carbonyls
in
two-orbital 5s
is
orbitals.These
third
importance
from c a l c u l a t i o n s
empty
discussed
of
core-like
probe the
be
to
orbitals
to
increase
effects will,as
the
O2
(back)donation
metal
illustrated
Corrections Actinides.
might
stability
not
donations
substantial
involving
Relativistic Compounds o f
do
corrections
with
in
for
interactions
of
in destabilizing two-orbital
are
a
energy.Relativistic kinetic
to
favorable
(lb)
lc,give
metal
third
C2H2,C2H4,and
π-type donor o r b i t a l s . T h e s e
orbitals
interactions
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch023
i n the
the
metal-carbonyls
order
of
corrections
relativistic
σ-type or
only
complexes
relativistic
.Instead,the ligand
f o r the
relativistic
for
change (enhancement)
5d-elements of
illustrated
335
in
results
R=H,CH3.
1
mol" )
in
RMCI3
(M=Th,U)
3/
D(M-R) HThCl
N R
D(M-R)
30.1
3
R
a
76.0
Exp.
~ 80
CH3THCI3
35.8
79.8
~80
HUCI3
10.5
70.1
76
16.8
CH3UCI3
72.2
72
a
D(M-R) bond energies from CpM(Cl)R of Ref. It
follows
corrections
to
in
to
order
relativistic some
50-60
the
Table
actinide-R
reproduce
bond Kcal
from
energies 1
mol" .The
III
bond
19 that
energies
experimental are
seen
to
importance
be of
relativistic are
necessary
results.Non too
small
by
relativistic
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
336
THE CHALLENGE OF d AND f ELECTRONS
b
a 211
211 178 174
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch023
Cr MoW Non-Rel Figure
4 . (a)
216
216
210 178
162
162 130
Fe Ru Os
Cr MoW
Fe Ru Os
Non-Rel
Rel
R e l a t i v i s t i c and n o n - r e l a t i v i s t i c
176
Rel averaged
i n t r i n s i c bond energies f17a) i n M(CO)q (m=Cr,Mo,W) . ( b )
intrinsic 1
bond energies (17a) i n M(CO)s (m=Fe Ru Os).Energies i n kJ mol" . f
Figure
5 . (a)
f
R e l a t i v i s t i c bond energies (12h) i n (PH ) MX ,with 3
2
M=Ni,Pd, Pt and X= C2H2, C 2 H 4 , 0 . (a) R e l a t i v i s t i c bond energies (Uh) 2
i n (PH ) MX ,with M=Co,Rh,Ir and X=C2H , C2H4, O2 .Energies i n kJ mol" 3
4
2
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
1
23. ZIEGLER ET AL. corrections
Relativistic Effects on Compounds ofHeavy Elements
f o r the chemistry
of actinides
i s currently
(13.,21) .
under i n v e s t i g a t i o n
Acknowledgment All
calculations
carried
out a t t h e Cyber-205
i n Calgary(ACS) and Amsterdam(SARA).This
investigation
was s u p p o r t e d
Engineering Research
by t h e N a t u r a l
Research Council
the Netherlands
Organization
S c i e n c e and
o f Canada(NSERC)
as w e l l as
f o r t h e Advancement
o f Pure
(ZWO).
Cited
Literature
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch023
were
installations
1.
Pitzer,
2.
Pyykkö,Ρ.;Desclaux,J.Ρ.Acc.Chem.Res
K.S. Acc.Chem.Res.
3.
Pyykkö,Ρ.Chem.Rev.,in
4
1979,12,271. 1979,12,276.
press.
(a)Snijders,J.G.:Baerends,J.G.Mol.Phys.1978,36,1789. (b)Snijders,J.G.;Baerends,J.G.:Ros,P.Mol.Phys.1979,38, 1909.
5. 6
Becke,A.
J.Chem.Phys.
(a)
1986,84,4524.
Ziegler,Τ.;Snijders,J.G.;Baerends,Ε.J.J.Chem.Phys.
1981,74,1271. (b)Ziegler,T.;Snijders,J.G.;Baerends,E.J. Chem.Phys.Lett.
1980,75,1.
7.
Dirac,P.A.M.
Proc.R.Soc.London
8.
Foldy,
9.
Moss,R.E. Advanced
L.L.;Wouthuysen,S.A.
,Chapman and 10.
Molecular
1928 ,117,610. 1950,78,29.
Quantum
Mechanics
Hall,London,1973.
Krasnov,K.S.;Timoshinin,V.S.;Danilova,Τ.G.;Khandozhko ,S.V.Handbook of Molecular Compounds
11.
Ser.A Phys.Rev.
,Jerusalem
Baerends,E.J.;
Constants
of
Inorganic
1970.
Ellis,D.Ε.;Ros,Ρ.Chem.Phys.
1973,2,71. 12.
(a)Lee,Y.S.;Ermler,W.C.;Pitzer,K.S.;McLean,A.D. J.Chem.Phys.
1979,70,293.
(b)Hay,P.J.;Wadt,W.R;Kahn,L.R.;Bobrowicz,F.W. J.Chem.Phys. 13. Pilar
1978,69,984.
F.L. Elementary
Hill,New
Quantum Chemistry
14. Ziegler,T.;Tschinke,V.;Becke,A. 15.
Polyhedron
1987,6,685.
Baykara,N.A.;McMaster,B.N.;Salahub,D.R. Mol.Phys.1984,52
16.
,McGraw
York 1968.
(a)Ziegler,T.
891. Oragnometallics
1985,4,5675.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
337
338
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch023
17
18.
19. 20. 21.
THE CHALLENGE OF d AND f ELECTRONS
(b) Ziegler,T.;Tschinke, V.;Becke,A.J.Am.Chem.Soc. 1987,109,1351. (c) Ziegler,Τ.; Cheng,W.;Baerends,Ε.J.;Ravenek,W. Inorg.Chem.,1988,accepted. (d) Z i e g l e r , Τ . ; T s c h i n k e , V . ; V e r s l u i s , L ; B a e r e n d s , Ε . J . ; Ravenek,W.Polyhedron,in press. ( a ) Z i e g l e r , Τ . ; T s c h i n k e , V . ; U r s e n b a c h , C . J.Am.Chem.Soc. 1987,102,4825. (b) Ziegler,T. Inorg.Chem. 1985,24,1547. (c) Ziegler,T. Inorg.Chem. 1986,25,2721. (d) Ziegler,T.; Cheng,W.,unpublished work. (e) Baerends,Ε.J.;A.Rozendaal NATO AST 1986,C176, 159. Ziegler,T.;Baerends,Ε.J.;Snijders,J.G.;Ravenek,W. J.Chem.Phys,submitted.These calculations are based on a quasi-relativistic approach in which the valence density is determined variationally ,rather than by f i r s t order perturbation theory. Bruno,J.W.;Stecher,H.A.;Morss,L.R.;Sonnenberg,D.C. ;Marks,T.J. J.Am Chem.Soc. 1986,108,7275. Schwarz,W.H.F.;Chu,S.Y.;Mark,F. Mol.Phys. 1983,50,603 (a)Boerrigter,P.M.;Baerends,Ε.J.;Snijders,J.G. C h e m . P h y s . 1988,122,357. (b)Boerrigter,Ρ.Μ.,thesis,Vrije Universiteit,1987.
RECEIVED October 24, 1988
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Chapter 24
Ground-State Properties of Heme Complexes in Model Compounds and Intact Proteins 1
1
1
2
Frank U. Axe , Lek Chantranupong , Ahmad Waleh , Jack Collins , and Gilda H. Loew 2
1
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch024
2
The Rockefeller University, 701 Welch Road, Suite 213, Palo Alto, CA 94304 SRI International, 333 Ravenswood Avenue, Menlo Park, CA 94025 The ground and low-lying spin states of f e r r i c porphy r i n (heme) complexes found in model compounds and intact proteins have been studied using an INDO-RHF-SCF method parameterized to include t r a n s i t i o n metals. The results for the model compounds using known c r y s t a l structure geometries are consistent with and help explain the o r i g i n of their observed electromagnetic properties. These studies demonstrate the a b i l i t y of this method to determine with a high degree of reliability the r e l a t i v e energies of the manifold of spin states of f e r r i c heme complexes. The same method has been used to address unresolved questions regarding the resting states of four heme proteins, cytochrome P450 which is a monofunctional oxidase, cytochrome c peroxidase (CCP) and catalase (CAT) both with peroxi dase oxidizing a c t i v i t y , and metmyoglobin (MMB) which lacks s i g n i f i c a n t peroxidase oxidizing a c t i v i t y . The characterization of the P450 resting state leads to a possible explanation of the o r i g i n of i t s low-spin (S = 1/2) form. This result helps resolve the apparent contradiction between the presence of water as an a x i a l ligand as determined by the x-ray structure and the absence of hyperfine s p l i t t i n g s in ESR spectra with 65% O enriched water. Comparisons of the resting state calculations of CCP and MMB provides an understanding of the origins of the differences in observed e l e c t r o magnetic properties for MMB and CCP in spite of the s i m i l a r i t y of their active s i t e s . Differences in function between MMB and CCP could not, however, be understood from properties of the active s i t e alone. Changing the imidazole ligand to an imidazolate in CCP makes i t s active s i t e more similar to CAT. Thus, the anionic form of the imidazole ligand of CCP, thought to be p a r t i a l l y formed by Η-bonding to a nearby Asp residue, could account at least in part for the similar a c t i v i t i e s of CCP and CAT as oxidizing enzymes. cam
17
0097-6156/89/0394-0339$06.00/0 ο 1989 American Chemical Society In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch024
340
THE CHALLENGE OF d AND f ELECTRONS
Heme proteins a l l share a common active s i t e or prosthetic group consisting of an iron-porphyrin (heme) unit, a nearly planar e n t i t y embedded i n the globular protein and connected to i t by one or at most two nearby amino acids which serve as a x i a l ligands. This important family of proteins performs three basic b i o l o g i c a l functions (1-5): reversible oxygen transport (globins), electron transfer (cytochromes), and metabolic oxidation of small organic molecules and peroxides (peroxidases, catalases, and cytochrome P450s). In a l l heme proteins, the b i o l o g i c a l function i s centered on the heme unit and primarily on the iron i t s e l f ( 1-5). Thus the oxidation and spin state of the iron, the nature of the a x i a l ligands, and the protein environment of the heme unit serve as subtle modulators of b i o l o g i c a l behavior. The heme group i s also the p r i n c i p a l o r i g i n of spectroscopic features of these proteins. Both electronic spectra (6-9) and ground-state electromagnetic properties such as quadrupole s p l i t t i n g s i n MOssbauer resonance spectra (10-14), anisotropic g values and hyperfine s p l i t t i n g s i n electron and nuclear spin resonance spectra (15-23) and temperature-dependent magnetic moments (24-29) originate almost e n t i r e l y on the heme u n i t . Consequently, a large f i e l d dedicated to the study o f model heme complexes has emerged i n an e f f o r t to understand the e f f e c t of changes i n the heme unit i t s e l f on these observed properties. These studies are useful i n understanding the properties o f intact heme proteins since isolated heme complexes have electromagnetic proper t i e s very similar to heme units embedded i n proteins. Some of these model heme systems have also been shown to mimic the b i o l o g i c a l a c t i v i t y of intact heme proteins. For instance, model oxo-iron compounds have been found to epoxidize o l e f i n s much l i k e the cytochrome P450s (34, 35). The r e l a t i v e s i m p l i c i t y o f model heme complexes makes i t possible to study the important role of the a x i a l ligands (30-33) i n modulating electronic structure and geometries without the e f f e c t of the nearby amino acid residues present i n the proteins. The insights gained from such studies can help to separately assess the r e l a t i v e importance of the heme unit i t s e l f and of i t s protein environment on the function of intact heme proteins. Up to now most quantum mechanical studies of the ground and excited states of model heme complexes have focused primarily on diamagnetic systems (36), with less frequent treatment of heme systems with unpaired spins (37-42). With the inclusion of a r e s t r i c t e d Hartree-Fock treatment (37, 38) within an INDO formalism parameterized f o r t r a n s i t i o n metals (39, 40, 42), i t i s now possible to calculate the r e l a t i v e energies of d i f f e r e n t spin states o f f e r r i c heme complexes i n an evenhanded fashion at a semiempirical l e v e l . In the work reported here we have used t h i s method i n two types of studies. The f i r s t study i s a systematic investigation of the e f f e c t of changes i n geometry and ligand type on the r e l a t i v e energies of the low-lying spin states and observable properties of eight model f e r r i c heme complexes. This study also represents a test of the c a p a b i l i t i e s of the INDO-RHF method to characterize the lowest lying doublet (S = 1/2), quartet (S = 3/2) and sextet (S = 5/2) spin states of these model f e r r i c heme complexes. The eight complexes chosen a l l have known c r y s t a l structures and include those with varying a x i a l ligands and high-, intermediate-, and low-spin ground
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
24.
A X E ET AL.
Ground-State Properties of Heme Complexes
341
states as inferred from magnetic s u s c e p t i b i l i t y measurements (43-50) and g values in electron spin resonance spectra (47-48). They also have known quadrupole s p l i t t i n g s ( A E Q ) from Mossbauer resonance spectra (44, 45, 49, 51-53), a quantity which we d i r e c t l y calculate. In the second type of study, using insights gained from the model compound studies, the active s i t e of the resting state ( i . e . , state of the enzyme when not involved in i t s biochemical cycle) of four heme proteins, cytochrome P450 , metmyoglobin (MMB), cytochrome c peroxidase (CCP), and caualase (CAT) have been characterized. These four proteins belong to d i f f e r e n t classes of heme proteins. P450, CCP, and CAT are oxidative metabolizing enzymes thought to share a similar highly oxidized b i o l o g i c a l l y active state, and MMB i s the oxidized form of an oxygen transport protein with l i t t l e or no peroxidase or monofunctional oxidase a c t i v i t y Q ) . Each of these proteins have f e r r i c resting states which have been characterized by x-ray structure determinations (54-57). Paradoxically, while a number of long-standing questions have been resolved by these structure determinations, new ones are emerging. This study addresses f i v e such s p e c i f i c questions.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch024
c
The f i r s t two questions involve properties of the resting state of cytochrome P450 , the only P450 with a known structure (54). The camphor-free resting state i s mostly in a low-spin (S = 1/2) form while the camphor-bound state i s a high-spin (S = 5/2) ferric complex. The x-ray structure (58) reveals, that as previously deduced, the camphor-bound state i s 5-coordinated with a cysteine residue as the single a x i a l ligand and the iron s i g n i f i c a n t l y out of the porphyrin plane. The camphor-free state retains the cysteine ligand, but s u r p r i s i n g l y , a water, and not a second amino acid as previously thought (5), binds to the iron in the d i s t a l ligand binding s i t e . There i s also evidence that t h i s water i s part of an Η-bonded network involving four more water molecules. c
With the insight gained from the x-ray structure, two puzzling aspects of the camphor-free resting state have emerged. One i s the o r i g i n of the low-spin form deduced from observed electromagnetic properties (5). This r e s u l t i s surprising since other f e r r i c heme proteins with H 0 as a s i x t h ligand such as MMB have primarily highspin (S = 5/2) ground states. The other question raised i s : If water i s an a x i a l ligand, as reported in the x-ray structure, why i s no broadening of the ESR spectra from hyperfine interactions observed in 65% enriched 0 HpO (H. Beinert, private communeiation), as i t i s in MMB (59). Since the magnitude of the hyperfine s p l i t t i n g depends d i r e c t l y on the amount of unpaired spin density on the water oxygen atom, the p o s s i b i l i t y that the negative results obtained for P450 could be a consequence of reduced spin density on the water has been investigated. 2
17
A l l the remaining questions focus on comparisons of CCP, MMB, and CAT. The f i r s t question addressed i s : Can the differences in the active s i t e of CCP, MMB, and CAT account for the differences in their observed electromagnetic properties? MMB and CCP (55, 56) have been found to have the same heme unit, a f e r r i c protoporphyrin-IX with a water and an imidazole as a x i a l ligands. CAT has a single
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
342
THE CHALLENGE OF d AND f ELECTRONS
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch024
phenolate group from a nearby tyrosine residue as an a x i a l ligand (57). In addition to the c r y s t a l structures, the nature of the resting states of CCP, MMB, and CAT have been probed by Mossbauer (11-13), temperature-dependent magnetic s u s c e p t i b i l i t y (27-29) and electron paramagnetic resonance (21-23) experiments. The observed properties of both CAT and MMB have been interpreted in terms of a d e f i n i t e high-spin resting state, while the properties of CCP have been interpreted in terms of a thermal d i s t r i b u t i o n of high- and lowspin states (_Π, 29). The determination of the r e l a t i v e energies of the low-lying sextet (S = 5/2), quartet (S = 3/2) and doublet (S = 1/2) states of the active s i t e s of these proteins should lead to a better understanding of the o r i g i n of these properties. The f i n a l two questions raised are the extent to which the active s i t e i t s e l f controls the function of CCP, CAT, and MMB. S p e c i f i c a l l y , we have asked: To what extent can s i m i l a r i t i e s in function between CAT and CCP be understood in l i g h t of their d i f f e r ent active s i t e s ? F i n a l l y , we have asked: To what extent can differences in the function of MMB and CCP be understood in terms of their active s i t e c h a r a c t e r i s t i c s alone? Knowledge of the extent to which the active s i t e can account for function should help to under stand the r e l a t i v e importance of the protein environment around the heme in determining the function of each protein. Methods All calculations were carried out within the approximation of intermediate neglect of d i f f e r e n t i a l overlap (37-42) (INDO-RHF-SCF) which includes parameterization for t r a n s i t i o n metals. A restricted open-shell formalism, developed by Zerner et a l . (37,38), was employed to prevent spin contamination and to make the quantitative evaluation of the r e l a t i v e spin state energies possible. This method has been used successfully to study simple t r a n s i t i o n metal complexes l i k e [ F e C l J * (42), [CUC1J2- (42), and ferrocene (4Ί) as well as larger and more complicated systems l i k e model oxyheme (6M) and carbonylheme (6l_) and model oxyhorseradish peroxidase (62) complexes. Energies of the lowest l y i n g sextet, quartet and doublet states were calculated for each of the heme units studied. The geometries of the complexes were taken from c r y s t a l structures and s i m p l i f i e d to unsubstituted porphyrins. The orientations of the porphyrin macrocycles were such that the pyrrole nitrogens were on the x- and y-axes. The choice of the lowest energy configurations for each state was as follows: Doublet state: Quartet state: Sextet state:
(d
and d
)
3
The sextet state configuration i s unique. The choice of the lowest energy configuration for the doublet and quartet states was confirmed by comparisons of the r e l a t i v e energies of various quartet and doublet configurations, obtained by assignment of unpaired electron(s) to d i f f e r e n t iron d o r b i t a l s , in some representative complexes. It i s also corroborated by a recent detailed study of
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
24.
A X E ET AL.
343
Ground-State Properties of Heme Complexes
hemin u s i n g the INDO/RHF ( J . P h y s . Chem., i n p r e s s ) .
method
by
Edwards,
Weiner
and
Zerner
The quadrupole s p l i t t i n g ( A E Q ) o b s e r v e d i n Mossbauer resonance o f heme compounds was c a l c u l a t e * from the INDO-RHF e i g e n v e c t o r s . T h i s q u a n t i t y was determined by f i r s t c a l c u l a t i n g the n i n e components (VjJ o f the e l e c t r i c f i e l d g r a d i e n t t e n s o r , u s i n g the a p p r o p r i a t e o n e - e l e c t r o n o p e r a t o r , and c o n s i d e r i n g o n l y the c o n t r i b u t i o n o f the i r o n from a l l i t s f i l l e d o r b i t a l s . The 3> IV j j j > | V j J . These v a l u e s were then used i n the e x p r e s s i o n : k
AE
= 8 ( 1 -
R)Qq[1
+
2
η /3]*
.
( 1 )
where q = V , η = ( V ^ V , J / V * (0 < η < 1 ) , ( 1 - R ) = S t e r n h e i m e r S h i e l d i n g c o n s t a n t , ana Q = n u c l e a r quadrupole moment. The s i g n o f A E Q i s the s i g n o f the l a r g e s t component V . . . V a l u e s o f Q and ( 1 - R ) used i n these c a l c u l a t i o n s a r e 0 . 1 8 7 and O . b o , r e s p e c t i v e l y . i
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch024
Q
i
i
R e s u l t s and D i s c u s s i o n Model Heme Complexes. Presented i n T a b l e I a r e the c a l c u l a t e d r e l a t i v e energy d i f f e r e n c e s i n k c a l / m o l f o r the s e x t e t , q u a r t e t , and d o u b l e t s t a t e s o f the model f e r r i c heme complexes i n c l u d e d i n the present study. Also included i n Table I are the calculated quadrupole s p l i t t i n g s U E Q ) f o r the r e l e v a n t s p i n s t a t e , a l o n g w i t h the e x p e r i m e n t a l l y observed v a l u e s o f A E Q and the measured e f f e c t i v e magnetic moments. The r e s u l t s c l e a r l y demonstrate t h a t the ground s p i n s t a t e c a l c u l a t e d f o r each model complex a g r e e s w i t h the one i n f e r r e d from measured e f f e c t i v e magnetic moments. Moreover the energy s e p a r a t i o n s between these ground s t a t e s and the o t h e r two s p i n s t a t e s a r e c l e a r l y consistent with observable electromagnetic properties and help explain their origins. The observed e f f e c t i v e magnetic moments ( 4 3 - 4 6 ) f o r the 5 - and 6 - c o o r d i n a t e d complexes found t o have s e x t e t ground s t a t e s a r e a l l i n the range o f 5 . 9 - 6 y t y p i c a l o f h i g h - s p i n complexes. The c a l c u l a t e d AEQS f o r the h i g h - s p i n s t a t e o f t h e s e complexes a r e a l s o i n good agreement w i t h the e x p e r i m e n t a l v a l u e s known f o r t h r e e o f them ( 4 4 b
46,
5 0 ) .
Both 5 - and 6 - c o o r d i n a t e d h i g h - s p i n complexes have s i g n i f i c a n t s p i n d e n s i t y on the p o r p h y r i n r i n g , 6 0 ? o f which i s on the p y r r o l e nitrogens. T h i s s h o u l d be m a n i f e s t i n h y p e r f i n e s p l i t t i n g s o b s e r v a b l e i n ESR o r ENDOR e x p e r i m e n t s . The u n p a i r e d s p i n d e n s i t y on the a x i a l l i g a n d s i s much l e s s than on the p o r p h y r i n r i n g and g r e a t e r on a n i o n i c than n e u t r a l l i g a n d s . The c a l c u l a t e d r e s u l t s f o r the 5 - c o o r d i n a t e d h i g h - s p i n complexes i n d i c a t e t h a t i t i s d e f i n i t i v e l y more s t a b l e than the d o u b l e t and q u a r t e t s t a t e s by - 3 0 k c a l / m o l and - 1 2 k c a l / m o l , r e s p e c t i v e l y . The 6 - c o o r d i n a t e d h i g h - s p i n complexes e x h i b i t a s i g n i f i c a n t r e d u c t i o n o f the energy s e p a r a t i o n between the s e x t e t and the q u a r t e t s t a t e s ,
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
2
5.9
5.90
5.92
experimental
6.05
1.12 1.22
0 5 18
TMS0 TMS0
4.5-5.3
3.20 3.5
2 0 25
ClO^
2.09
2.26 3.7-4.7
0 12 12
2.19 2.25
3
a
N" pyridine
0.35
2.31
0 22 19
CN" CN"
Low Spin
3.14 2.7
0 17
3
3-Clpyridine 3-Clpyridine
Intermediate Spin
value of azide complex of MMB and CCP (Reference 53, page 3)
Exp. ( y )
y
e f f
0.12
0 5 18
(NCS)" pyridine
1.01 0.76
0 12 28
2
(SpN0 r
0.44 0.46
A E Q (mm/sec)
0 12 33
Cl""
Cale. Exp.
b
(kcal/mol)
S = 5/2 S = 3/2 S = 1/2
ΔΕ
L
High Spin
"-1
1-2
Table I. Relative Energies of Different Spin States of Model Ferric-Heme Complexes
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch024
24.
A X E ET A K
Ground-State Properties of Heme Complexes
345
attributable mainly to the presence of the second a x i a l ligand which increases the tendency of the iron atom to move into the plane of the porphyrin.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch024
In both 5- and 6-coordinated complexes the energy of the doublet state i s predicted to be too high to play a s i g n i f i c a n t role in determining the observed electromagnetic properties. However, because of the smaller separation of the sextet and quartet states in the 6-coordinated high-spin compounds compared with the 5-coordinated compounds, spin mixing (\f>S) between them should be enhanced. Therefore larger zero f i e l d s p l i t t i n g s and more aniso tropic g values (63) in the ESR spectra should be observed. Intermediate-spin (63) heme complexes are rare and two complexes inferred to have quartet ground states have been included in our studies. As shown in Table I, the predicted ground state of each complex i s a S = 3/2 state in agreement with the spin state assign ment deduced from observed properties. The calculated r e l a t i v e energy of the doublet spin state i s - 20 kcal/mol above the quartet spin state, while the sextet states of [Fe(TPP)(ClO^)] and [Fe(0EP)(3-Clpy)p] are only ~ 1.8 and 2.8 kcal/mol above their respective quartet state. These results strongly suggest that observable properties can best be understood in terms of s i g n i f i c a n t spin-orbit coupling of these two low-lying states, together with the p o s s i b i l i t y of a thermal equilibrium of such spin-mixed states. +
Effective magnetic moment measurements of [Fe(TPP) (ClOi,) ] (47) have yielded values in the range of 4.5-5.3 y at 77-300°K. This temperature dependence and range of values i s consistent with c o n t r i butions from sextet and quartet states with the quartet lower in energy. ESR data (47) for this complex yielded values for g and g| of 4.75 and 2.03, respectively. These results are a t y p i c a l for a high-spin complex and lend further support to the conclusion that the ground state i s a S = 3/2 or a 3/2,5/2 mixture with predominant S = 3/2 character. b
(
Magnetic susceptibility measurements (48) for [Fe(0EP)(3C l p y ) ] , the other intermediate complex studied, y i e l d a range of y between 3.7-4.7 y for the temperature range of 77-294°K. This range of magnetic moments i s also consistent with an intermediatespin or spin-mixed ground state. The EPR spectrum (48) for t h i s complex yielded values for g and g» of 4.92 and 1.97 respectively which are similar to values obtained for [Fe(TPP)(ClO^)]. +
2
e f f
b
The A E Q values calculated for the quartet state of these complexes also agree very well with the experimentally observed values (Table I) for the same complexes. A l l these r e s u l t s taken together are highly suggestive that the S = 3/2 spin state i s the p r i n c i p a l contributor to the ground electronic state of these complexes. For the dicyano and azide-pyridine complexes, our calculated results indicate that in each case a S = 1/2 spin state i s the lowest energy state with the quartet and sextet states much higher in energy. The observed e f f e c t i v e magnetic moments (Table I) of both
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
346
THE CHALLENGE OF d AND f ELECTRONS
the dicyano (49) and the azide-pyridine (50) complexes indicate that these two complexes are indeed essentially low-spin at a l l temperatures. The only experimental A E Q measured f o r either of the two lowspin complexes i s f o r the dicyano complex. Our calculated value of 2.31 mm/sec for the S = 1/2 spin state i s i n poor agreement with the experimental value (53) of 0.35 mm/sec. However, the reported experimental A E Q seems to be anomolously low for what i s considered to be a low-spin complex. There i s apparently no experimental A E Q measured for the [Fe(TPP)(N^)(py)] complex. The calculated value or 2.19 mm/sec for the doublet state i s , however, i n good agreement with the experimental values (53) of 2.45 and 2.25 mm/sec f o r CCP-N and MMB-N, respectively, which d i f f e r only by one a x i a l ligand being an imidazole rather than pyridine. This provides further evidence that [Fe(TPP)(N^)(py)] has a doublet ground state. 3
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch024
3
An important conclusion from t h i s study of model compounds i s the additional evidence obtained for the key role o f the S = 3/2 spin state i n the chemistry of f e r r i c heme complexes. There are no complexes for which S = 5/2 and 1/2 spin states are close enough to interact without an even greater contribution of the S = 3/2 spin state. Thus, the widely used assumption (64) of high-spin/low-spin thermal contributions to explain observable properties o f heme complexes appears to be incorrect. Explanations involving highspin/intermediate-spin interaction are much more plausible, since small energy separations between these states were found. In general, the r e l a t i v e spin state energies calculated for a l l the model heme complexes studied are consistent with and help explain their observed electromagnetic properties. Thus the INDO-RHF method used appears to be sensitive to the effect of the varying a x i a l ligands and predicts the correct energy order of spin states produced by each of them. The a b i l i t y of the method to predict the patterns of spin state behavior i n these model complexes lends credence to the use made of i t i n the second part of these studies, to further characterize the heme units i n the resting state of four heme proteins. Comparative Studies of Resting State Active Sites of Four Heme Proteins. In this second type of study reported, we have used the x-ray structure of the active s i t e of four heme proteins: cytochrome P450 (54), CCP (55), MMB (56), and CAT (57) s i m p l i f i e d to the f e r r i c porphyrin complexes, shown i n Table I I , to calculate the r e l a t i v e energies and electron and spin d i s t r i b u t i o n s i n their lowlying sextet, quartet and doublet states. çam
As shown in Table I I I , a high-spin ground state i s d e f i n i t i v e l y obtained for camphor-bound 4 5 0 i n which the single a x i a l ligand i s a mercaptide. For camphor-free 4 5 0 , with water and mercaptide as a x i a l ligands at their x-ray structure values, the sextet state i s s t i l l the lowest energy, but the energy separation to the low-spin (S = 1/2) state i s greatly diminished. In the x-ray structure deter mination of the resting state, the value of the Fe-water distance was p
c a m
p
c a m
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
24.
A X E ET AL.
347
Ground-State Properties of Heme Complexes
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch024
Table I I . Ligand Distances (Â, X-Ray) Used for Oxidized Resting State of Four Model Heme Proteins
[MMB] L
[CCP]
H0
2
2
L
1 Fe-0 Fe-L 7 a *Fe
H0
--
Imidazole
Phenolate
2
Imidazole
[CAT]
[P450] H0 2
(SCH3)"
2.40
--
2.02
1.93
1.76
2.32
0.25
0.13
0.13
0.24
1.90
2.24
Extent of out-of-planarity of the iron atom from the mean porphyrin ring plane.
constrained and the exact position of the water was not e x p l i c i t l y optimized. Thus we have considered the consequences of a shorter iron-water distance (2.0 Â) and movement of the iron into the heme plane. In this geometry, a low-spin state i s calculated to predominate. An alternative o r i g i n of the s t a b i l i z a t i o n of the lowspin state comes from the p o s s i b i l i t y that some anionic character i s imparted to the water ligand by i t s postulated interaction with the network of Η-bonded water molecules, seen in the x-ray structure (54). This e f f e c t was simulated by using an OH" as an a x i a l ligand with an Fe-0 bond length of 1.75 A. As seen in Table I I I , in t h i s model of the resting state, the low-spin (S = 1/2) state i s favored by 16 kcal/mol over the high-spin state. While this i s an extreme model for the e f f e c t of Η-bonding, i t does demonstrate that p a r t i a l anionic character of the water ligand could account for the predominant low-spin ground state observed. Table IV gives the spin densities calculated on the water oxygen for P 4 5 0 and for comparison, in MMB. Experimental values of y f f and our calculated results (Table V) indicate MMB i s in a high-spin cam
e
American Chemical Society Library 1155 16th St, N.W. Washington, D.C. Salahub, 20036 D., el al.; In The Challenge of d and f Electrons; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
348
THE CHALLENGE OF d AND f ELECTRONS Table I I I . Origin of the Low-Spin Form of the Resting State of Cytochrome P450
Models for Resting State a
1
(SCH )~
L
H0
H0
S = 1/2
6
S = 5/2
0
L
3
2
2
(SCH )"
Substrate Bound State (SCH )~
b
C
3
3
(SCH )"
(0ΗΓ
—
0
0
15
1.5
16
0
2
d
3
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch024
ΔΕ (kcal/mol)
a
Geometry from x-ray structure as shown in Table II (Ref. 54)
b
X-ray structure with Fe-Water distance of 2.00 Â and iron moved into the porphyrin plane
c
X-ray structure with Fe-0H~ distance at 1.75 Â as i t i s in model Fe-0 complexes.
d
Geometry form x-ray structure as in Reference 58.
2
ground state. In i t s c r y s t a l structure geometry, the water oxygen of the high-spin f e r r i c MMB i s calculated to have 0.057e or about 1.1? of the t o t a l spin. For this protein, a barely detectible amount of broadening of the g=2 signal was observed in the ESR spectra in the presence of 0 enriched H 0 (59). By contrast, in both the highspin state and the low-spin state of P450, in i t s c r y s t a l structure geometry, the spin density on the a x i a l water ligand i s much lower than in the corresponding state of MMB. Allowing the Fe-0 distance of the water ligand to decrease, or simulating i t s ionic character by an 0H~, both of which favor a low-spin ground state, somewhat increases the spin density on the oxygen. However i t remains at most about 1/6 that of MMB. Since the broadening in MMB was barely detectable, no measurable broadening of the ESR spectra in 0 enriched water would be expected for either low-spin model of 4 5 0 currently proposed here. These results then account for the absence of such broadening in a manner consistent with the presence of water as an a x i a l ligand in the resting state of 4 5 0 as observed in i t s x-ray structure. 17
2
17
p
c a m
p
c a m
Turning now to comparisons of CCP, MMB, and CAT, the r e l a t i v e energies of the germane doublet, quartet, and sextet spin states have been calculated using the same INDO-RHF-SCF method as for the model complexes and the results are presented in Tables V and VI. The geometries for the resting states of CCP, MMB, and used here were takem dire'ctly from their respective x-ray c r y s t a l
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
CAT
24.
A X E ET AL.
349
Ground-State Properties of Heme Complexes
Table IV. Calculated Spin Densities of Oxygen of H 0 Ligands in P 4 5 0 and MMB
a
a
2
cam
MMB
P450 OH"
H0 2
S = 5/2
s
0.01
= 3/2
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch024
B
0.16
0.057
0.01
0.030
0.01
0.004
(0.03)
s = 1/2 a
(0.03)
0.0005 (0.004)
Values underlined are for lowest lying state.
^Values in parenthesis calculated for H 0 at 2.00 Â and Fe in the porphyrin plane. 2
Table V.
Effect of Geometry on Resting States of CCP, MMB, and CAT
ΔΕ
D
4h
X-Ray
4h
X-Ray
D
0
0
0
1.5
2.1
2.8
5.7
5.8
6.3
D
4h
(kcal/mol)
S = 5/2
0 (0)
S = 3/2
-1.2 (-2.3)
S = 1/2 AEQ
e f f
a
8.0 (13.0)
b
0
0 (0)
-1.1
2.7 (5.6)
7.7
7.5 (11.4)
(mm/sec)
Cale. Exp. y
CAT
MMB
CCP X-Ray
Exp. ( y ) b
3.20
0.76 1.33
0.72 0.84
4.86
6.00
5.92
a
Values
b
Values in parenthesis with iron moved 0.1 Â further out of mean plane of porphyrin ring.
in parenthesis without d i s t a l water
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
350
THE CHALLENGE OF d AND f ELECTRONS Table VI. E f f e c t of Ionization State of Axial Ligand i n CCP and Catalase
CCP L L
1
Catalase
H0 Im
H0 Im"
2
2
2
Phenolate
Phenol
ΔΕ (kcal/mol)
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch024
S = 5/2 S = 3/2 S = 1/2
1.2 0.0 9.2
0 2.4 9.5
Net Charge Fe 1.25 0.11 1 0.18 L •0.55 Porph. Spin Fe 2.79 L1 0.02 L2 0.07 0.12 Porph. a
t
2
2
0 2.1 5.8
b
b
1.35 0.10 -0.63 -0.83
4.26 0.02 0.16 0.56
a
F o r S = 3/2 spin state
b
F o r S = 5/2 spin state
1.37 —
b
4.7 0 12.2
1.28* —
-0.55 -0.82
0.14 -0.42
4.24
2.74
—
0.25 0.51
—
0.11 0.15
coordinates (55-57). However, i n order to examine various geometric e f f e c t s on the spin states of each heme unit, calculations were also carried out at several step-wise regularized geometries, s t a r t i n g from the c r y s t a l geometry of each protein. The e f f e c t s of porphyrin r u f f l i n g and doming were examined by regularizing the porphyrin c r y s t a l geometry to symmetry f o r CCP, MMB, and CAT, while leaving the a x i a l ligands at the same geometry as i n their c r y s t a l structure. Further differences i n the geometries of CCP and MMB were examined by removal of the a x i a l water in CCP and by increasing the out of plane distance of the iron in MMB by 0.1 Â. The calculated r e l a t i v e energies of CCP (Table V) indicate that the S = 3/2 state i s the lowest energy spin state in CCP, with the S = 5/2 and S = 1/2 spin states being -1 kcal/mol and ~9 kcal/mol higher i n energy. Furthermore, the energy ordering and separation of the spin states are rather insensitive to regularization of the porphyrin and a x i a l ligand geometries. The predominance of the quartet state i n CCP appears to be due to a combination of near planarity of the iron and a weak a x i a l ligand. The Fe-water distance in CCP at 2.4 Â i s considerably longer than that i n MMB. Indeed, calculations at the c r y s t a l geometry i n which the water ligand i s
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
24. A X E ET AL.
351
Ground-State Properties of Heme Complexes
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch024
removed r e s u l t s i n the enhanced s t a b i l i t y o f the q u a r t e t s t a t e o f the complex i n s p i t e o f the 5 - c o o r d i n a t e n a t u r e o f the i r o n ( T a b l e V ) . For MMB ( T a b l e V ) , the s e x t e t s t a t e i s the l o w e s t energy s t a t e f o r the n e u t r a l i m i d a z o l e l i g a n d w i t h the q u a r t e t and d o u b l e t s t a t e s ~3 and ~8 k c a l / m o l h i g h e r i n energy. In c o n t r a s t t o CCP where r e g u l a r i z a t i o n o f the p o r p h y r i n r i n g t o symmetry had l i t t l e e f f e c t ; f o r MMB i t l o w e r s the energy o f b o t h the q u a r t e t and d o u b l e t s t a t e s r e l a t i v e t o the s e x t e t . These r e s u l t s suggest t h a t the enhanced doming o f the p o r p h y r i n r i n g observed i n MMB r e l a t i v e t o CCP i s a f a c t o r i n s t a b i l i z i n g the s e x t e t s p i n s t a t e i n MMB. However, s i n c e the s e x t e t s t a t e i s s t i l l l o w e s t i n energy even when the p o r p h y r i n i s r e g u l a r i z e d t o D j ^ symmetry, the enhanced o u t - o f - p l a n e d i s t a n c e o f the i r o n must be the main c o n t r i b u t o r t o the s t a b i l i z a t i o n o f the s e x t e t s t a t e . T h i s e f f e c t i s v e r i f i e d by the f u r t h e r s t a b i l i z a t i o n o f the s e x t e t r e l a t i v e t o the q u a r t e t s t a t e when the i r o n atom i n MMB i s moved by an a d d i t i o n a l 0.1 Â out o f the mean porphyrin plane (Table V ) . The c a l c u l a t e d r e l a t i v e s p i n s t a t e e n e r g i e s f o r CAT ( T a b l e V) a t the c r y s t a l geometry shows t h a t the the s e x t e t s t a t e i s the most s t a b l e s t a t e w i t h the S = 3/2 and S = 1/2 s t a t e s , r e s p e c t i v e l y , ~2 k c a l / m o l and ~6 k c a l / m o l h i g h e r i n e n e r g y . The c l o s e e n e r g e t i c p r o x i m i t y o f the S = 3/2 s p i n s t a t e i s a r e s u l t o f the s m a l l d i s p l a c e m e n t o f the i r o n atom from the p y r r o l e n i t r o g e n plane. Changing the h i g h l y r u f f l e d p o r p h y r i n m a c r o c y c l e o f CAT t o one o f pure Djj symmetry l e a d s o n l y t o a v e r y s m a l l s t a b i l i z a t i o n o f the s e x t e t s p i n s t a t e o f - 0 . 5 k c a l / m o l r e l a t i v e t o the q u a r t e t and doublet s p i n s t a t e s . Thus, the h i g h l y i r r e g u l a r p o r p h y r i n m a c r o c y c l e i n the c r y s t a l s t r u c t u r e has v e r y l i t t l e e f f e c t upon the r e l a t i v e s p i n s t a t e o r d e r i n g s i n t h i s system. n
In a d d i t i o n t o geometry v a r i a t i o n s , the e f f e c t s o f hydrogen bonding and the r e s u l t i n g i o n i c i t y o f the p r o x i m a l l i g a n d s i n CCP and CAT were s i m u l a t e d by d e p r o t o n a t i o n o f the i m i d a z o l e Ν i n CCP and p r o t o n a t i o n o f the t y r o s i n e oxygen i n CAT. D e p r o t o n a t i o n t o form an Im" l i g a n d i n CCP r e v e r s e s the o r d e r o f the s e x t e t and q u a r t e t s t a t e e n e r g i e s ( T a b l e V I ) . S i n c e t h i s i s an extreme model f o r the p a r t i a l p r o t o n t r a n s f e r t h a t c o u l d o c c u r as a r e s u l t o f the i m i d a z o l e b i n d i n g t o a nearby a s p a r t a t e r e s i d u e i n CCP, the p a r t i a l a n i o n i c n a t u r e c o u l d r e s u l t i n near degeneracy o f the q u a r t e t and s e x t e t s t a t e s . δ
Both the q u a l i t a t i v e and q u a n t i t a t i v e r e s u l t s o b t a i n e d f o r the a c t i v e s i t e s o f the t h r e e p r o t e i n s p r o v i d e an improved b a s i s f o r u n d e r s t a n d i n g the observed e l e c t r o m a g n e t i c p r o p e r t i e s o f the r e s t i n g s t a t e s o f CCP, MMB, and CAT. An i m p o r t a n t a s p e c t o f the p r e s e n t results i s t h a t f o r t h e s e p r o t e i n a c t i v e s i t e s , the s e x t e t and q u a r t e t s t a t e s a r e c l o s e i n energy and the d o u b l e t s t a t e i s s i g n i f i cantly higher. Thus, the dominant contributions to observed p r o p e r t i e s i n these p r o t e i n s a r e expected t o come from the S = 5/2 and S = 3/2 s p i n s t a t e s , which can mix by s p i n - o r b i t c o u p l i n g (63) as w e l l as be i n thermal e q u i l i b r i u m . These r e s u l t s provide a consistent e x p l a n a t i o n o f the e l e c t r o m a g n e t i c p r o p e r t i e s o f the r e s t i n g s t a t e s o f CCP, MMB, and CAT. The a l t e r n a t i v e e x p l a n a t i o n , a thermal e q u i l i b r i u m between s e x t e t and d o u b l e t s t a t e s , w i t h o u t a
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
352
THE CHALLENGE OF d AND f ELECTRONS
contribution from the quartet proteins.
state, does not seem possible i n these
These results are p a r t i c u l a r l y important for a correct q u a l i t a t i v e understanding of the the observed properties of CCP. For CCP the experimentally observed magnetic s u s c e p t i b i l i t y ( 2 9 ) and MOssbauer spectra (JM) have been interpreted in terms of a thermal mixing between high- and low-spin states, ignoring any contribution from the intermediate-spin state. This explanation i s contrary to our findings of E ç ^ ~ E3/2 1 / 2 * Measured values of y f f ( 2 9 ) for CCP that are i n the range of 3 . 7 to 4 . 0 y over a temperature range of 7 7 - 2 5 0 ° K can more correctly be understood in terms of a thermal contribution from heavily spin-mixed sextet and quartet spin states. These r e s u l t s also strongly indicate that a re-analysis of the Mossbauer resonance spectra of CCP (JM) as a mixture of quartet and sextet states would also be more appropriate.
-1.0 V (SSCE). Reduction potentials for alkyl halides of interest are generally more negative than -1.5 V (SSCE) (17). Alkyl halide photoreduction is observed for binuclear d& complexes whose excited-state reduction potentials are more positive than -1.0 V (SSCE) in CH3CN. An alternative pathway to outer-sphere electron transfer, which yields similar photoredox products with alkyl halides, is excited-state atom transfer (Figure 3b). Data obtained for Pt2(P20sH2)4 ~ indicate that alkyl Z
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch025
357
+
Î
0
+
2
2
+
2
2
4
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
z
THE CHALLENGE OF d AND f ELECTRONS
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch025
358
Figure 2. P i c t o r i a l representation of the M2-localized hole in a 3(da*pa) state.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch025
25.
8
SMITH & GRAY
Atom-Transfer Reactivity ofBinuclear d Complexes
b.
RX + M-M
°M-M
•
+ RX
•M-M-X + R.
RX + . M - M . ·
•M-M-X + R. R. =
/
X-M-M-X + C H C H 2
2
\
R. = .CR'X (R* = alkyl, aryl) R-M-M-X
Figure 3. a. S R ^ I mechanistic scheme for halocarbon photooxida tive addition to binuclear d^ complexes, b. Atom-transfer mechanism for halocarbon photooxidative addition.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
359
360
THE CHALLENGE OF d AND f ELECTRONS
and aryl halides react with the 3(0) and t h e c l u s t e r s
Fe(II)
make
spins
5/2
identifiable
that
c a n be
break
describe
2
magnitudes
transparently
that
iron
b y Gibson and
approximated as
Β i s a
splitting
i n the
to
2
and S
1
with
These
Ε -
(S'+h)
e
i s favored
to
models
application
S ' = 9 / 2 , and i n which
(4).
by just
closely
Here
S
l
by recognizing
derealization more
Here
i n which
strong
indistinguishable rationalized given
JS
andfour-iron
occurs,
few y e a r s
Hamiltonian
cluster.
character,
a
not
theclassic
a n d have
coupling
valence
i n the past
or spin Hamiltonian
H -
Fe(III)/Fe(II)
antiferromagnetic sites.
follow
clusters.
2
iron
mixed-valence trapped
coupling"
(3) o f t h e H e i s e n b e r g
describe
with
p r o g r e s s h a s b e e n made
o f "vector
co-workers
367
Spin Coupling and Electron Derealization
calculations
picture
included an alternate derivation
of
work,
(6,8)and had
many
can be used of
their
o f t h e (S'+H)
factor
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
368
THE C H A L L E N G E OF d AND f ELECTRONS
relating this
resonance e n e r g i e s t o c l u s t e r s p i n (6) .
a n a l y s i s t o t h e o x i d i z e d forms
linear
a n dc u b a n e - l i k e
forms o f t h e three reduced The
systems
.
(11) the
.
these
iron
to create
of three-iron
Broken
since
i t
crux
o f our computational
broken
metal
sites
symmetry have
compute a n d i n t e r p r e t , used.
cobalt
interest
because i t
approach a r i s e s
the
from
( i n which
broken
symmetry
the recognition
the
pure-spin
ground
three
S=5/2
extend for in
locations
We
Heisenberg
the
pure
We h a v e
spin
earlier
applied
H e n c e , we
computed
from
including
this
clusters
to the reduced and doubly
are
that are
parameters
states,
three-iron
t i m e we c o n s i d e r t h e e f f e c t s
assume
spins
andunderstand. to energies
methods
(which
states
anduse the resulting
as i n oxidized
the analysis
polynuclear
iron
state.
spins,
the f i r s t
to approximate
easy to
functional
wavefunctions
spin Hamiltonian of
equivalent
are relatively
density
"correct"
wavefunctions,
estimate
can be systems
stabilizes
otherwise
2
t o f i t a n (assumed)
four-iron
mixed-metal
eigenfunctions of S ) aregenerally multiconfiguration choose
to
a n dz i n c
Cluster.
especially i f local
c o n s i d e r a b l y more d i f f i c u l t
procedure to ( £ ) .
H e r e we
reduced species,
of electron
where
derealization
clusters. that
the true
together
electrostatic
can be
replaced
interactions
by
an
that
interaction
couple of the
type:
2
1 2
that
the off-diagonal
electron
derealization
discussed then
above.
consists
matrix
The essence of
choosing
eigenvalues this
important may
be We
we w i l l
basis
spin matrices,
for
first
state.
In
on each
a
three
where
Hamiltonian to
be
model
included,
andcomparing the r e s u l t i n g
i n order
to estimate
interpretations
the high site
three
iron spin
Β and J .
of
models
a l l
We f o r m
basis
three
configurations, in
a
configurations
worked
with
the
equivalent
Fe(II)/Fe(III)/Fe(III)
are aligned
(which must b e spin-down) We h a v e
cluster
reduced",
spin-up.
sites.
states
more c o m p l e x s p i n H a m i l t o n i a n
quantitative
d-electron three
states
c o n s i d e r o n l y the s i m p l e s t models t h a t have t h e
i n the "singly
d-electrons
connecting
be o f the form B(S'+H), as
data.
consider
sites
oxidation
will
of choosing a spin the
physical interactions; necessary
experimental metal
elements
t o t h e computed Χα e n e r g i e s
paper,
(3)
1
i s "allowed"
diagonalizing theresultant In
to the
H — J SiS +Ji3S 'S3+J23S2S3 and
of
certain
i s diamagnetic.
spin populations)
By c o n t r a s t ,
in
c l u s t e r s b u t does n o t a f f e c t t h e
wavefunctions
different
i nt h e
andi n aconitase
converted
novel
Symmetry A n a l y s i s f o r a T h r e e - I r o n
that
sites
vinlandii)
shown t h a t
TheZn complex i s o f p a r t i c u l a r coupling problem
are
been
clusters,
3
a r e themselves
the active
c a n be r e a d i l y
of the fourth
more r e d u c e d forms
The
for
a n d Azobacter
sites
both
states.
chosen f o r i l l u s t r a t i o n
andi t has recently
i n place
spin
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch026
of
extended
(with
we c o n s i d e r t h e r e d u c e d
and analogous ZnFe
as models
( f r o m D. gigas
Many
clusters,
we h a v e
interest
ferredoxins
used
iron clusters,
We r e c e n t l y
iron clusters
( 9 ) . ) Here
a n dd o u b l y - r e d u c e d o x i d a t i o n
considerable (10)
geometries
o f three
to reside
out the
parallel
five
fashion, say
by allowing i n turn
matrix
formal
the f i r s t the
final
on each o f the
elements
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
of the
26. SONTUM ET AL· Heisenberg
portion
delocalization Hence,
of
the Hamiltonian
terms
characterizes
we
resonance
assume matrix
and below,
between
each
the diagonal
5 for parallel
equivalent
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch026
(65/4)J these
+
elements
clusters, symmetry
atoms
(which
equivalent
we
atoms
configurations, the
2
3
find
state,
we
J =
2
represents
(65/4)J
"a") the
dimers. -
5B'
five of
pair
d-electron.
Following
the
λ
opposite
Papaefthymiou
only between
t h e two i r o n s
x
lowest.
For
For the
are
to
that
both
three
basis of the
et
s t i l l
iron
of
possible locations
that
(7),
al.
we w i l l
adopt
resonance i n t e r a c t i o n
o f t h e same s p i n ,
pair
"b" .
is
Here
spin matrix i s :
Here
we
have
symmetry results the
allowed
state,
0
0
•(15/4)J
0
0
•(25/4)J
the
The
symmetry
from t h a t
case
are E
1
J ' s and B ' s c a n , thus,
differences
arising
from
state
value)
energies
are
then
Hamiltonian. various
orbital
-
the model
o f Β and J are given
the
broken the Χα
Eigenvalues f o r 3
=
-(15/4)J.
by comparing those
state
the
computed
and estimates
resulting used
in
+ 5B a n d E
with
approach,
(5)
spin state;
indeed happens.
-(25/4)J
the ground
from
(For the simple
values
2
formulas
(including made
>
parameter
i n the high
be e s t i m a t e d
these
b r o k e n symmetry m o l e c u l a r spin
delocalization
B, to d i f f e r
r e p o r t e d b e l o w show t h a t t h i s
broken
5B
(25/4)J 5B
energy
carry
out an approximate
spin
and i t s
parametrized
here,
i n R e f . 7.)
projection
energy from
spin spin
the eigenstates In
a
o f the pure
for
the language
o u r p r e v i o u s p a p e r s , we a r e u s i n g t h e H e i s e n b e r g H a m i l t o n i a n to
three
are E =
o f one o f t h e
spin
There
simplest delocalization hypothesis,
important
lies
Ε
corresponding to the three
give
(S'+^s)
degenerate).
d-electrons
"b".
energy
elements
We a r e a s s u m i n g
(doubly
and, hence
is
the system
The e i g e n v a l u e s
2
B'2.0
3 shows t h e m o l e c u l a r
orbitals
h a s some
the Fermi
corresponds to antiferromagnetic and a
observed irons is
S-2 Fe(II)
(11).
monomer,
are reduced to Fe(II) low-energy
electrons. there
yielding
net
for
and the other 55a').
This
Fe(II)/Fe(III)
spin
"a" iron 1
of
5/2,
remains
Fe(III) several
a r e now p o s s i b l e a c c e p t e r s low-lying
orbitals
"a" site
orbitals
o f many m o r e
the existence
as
However,
o f these
to the u combination o f the "b" pair
a r e now a d d i t i o n a l
this
i n which both o f the "b"
and the unique
and 37a") and on the u n i q u e , imply
57a'),
a
occupation,
features,
level
orbital
g
i n energy by 1,300 cm" ).
orbitals
In a d d i t i o n
the
parameters.
w i t h one e x t r a
coupling o f an S ' - 9 / 2
(An a l t e r n a t i v e
computed to be h i g h e r
other
(in orbital (in
the
knowledge
by the zinc i o n .
l o c a l i z e d on the " a " i r o n "b" pair
worth
of
special
d-orbitals near
is
much
electron
dimer
cubane-like It
f o r o u r l o w e s t - e n e r g y b r o k e n symmetry s t a t e , the
likewise
geometries
n o r d o we h a v e
of the iron
(for
since,
i s predicted
(7).
cluster,
over
we
the ground s t a t e
cluster
3
o f the f o r the
t h a n t h o s e shown i n t h e
o f J and Β t o changes i n s t r u c t u r a l
reduced
stabilization
delocalized
than
discussed below);
experimental
a r e n o t known,
As with
| B | / J a r e around 1.5
|B|/J to be l a r g e r
with a l l J ' s equal,
though,
doubly
of
the
the doubly Zn complex,
about the s e n s i t i v i t y The
for
combination. t o the geometry
the computed J ' s t o be t o o l a r g e ,
cluster,
remembering,
values
Computed magnitudes
except
we e x p e c t
three-iron
to be r e l a t e d
absolute
the true values of
Table, to
below the antisymmetric
o f Β appear
both
there
(38a" and 6 0 a ' ) . states
than
extra
(36a"),
(e.g.,
56a'
These
extra
i n our model
o f E q . (7) , a n d s u g g e s t t h a t more c o m p l i c a t e d s p i n H a m i l t o n i a n m o d e l s may
be necessary
to
Hence,
the estimate
rather
approximate;
much s m a l l e r The
nevertheless,
reported
that
gas often
electron (19).
species
correlation
the q u a l i t a t i v e use
include
than
effects
(9).
effects
of
the values
past
found
in
a
spin-dependent
generally
studies
a
approximation,
effects
(20),
on the reduced
exchange-correlation potentials
of
reliable.
exchange-correlation
account
states
to increase J over
Calculations
Χα
density
a better
high-spin
conclusion
should be
the c o r r e l a t i o n
give
was i n d e e d f o u n d i n o u r e a r l i e r
clusters
calculations.
and c a n be compared t o
local
Since correlation
more
the
interest
the
uniform
this
here
Within
parametrizations properties
orbital
f o r J i n the doubly reduced complex i s
which has h i s t o r i c a l
calculations.
spin
the molecular
1
J f o r the doubly reduced complex
results
function,
describe
o f 45 cm"
stabilize
one would
low
expect
reported here, and
of oxidized species
three-iron
using
improved
give
serious
are i n progress.
Conclusion The
calculations
present
here
among
c o n s i d e r a t i o n to the competing e f f e c t s delocalization
in
polynuclear
qualitative
features
to
experimental
a
describe particular
solely
support
mechanism
from a n a l y s i s
transition
the "double
spectra in
the
these
(7),
first
to
o f s p i n coupling and e l e c t r o n metal exchange"
although
we d o n o t
calculations--the
o f computed t o t a l
energies
clusters. model
put
The forth
postulate
results
for various
arise states.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
26.
SONTUM E T A L .
-0.
80
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch026
-0.
Iron Β
Iron A
05k
-0.10
375
Spin Coupling and Electron Delocalization
α'
38
a"
57
α',
58%
85%
1
86%
39
α "_
36%
59
α*
25%
5b
α'
33%
37
a".
37%
58
α'
39%
38
α"
45%
15
CO
55 a'...! — 44%
cn 0) JO
>^ - 0 .
20
54
c LU
.22%
α ' 3 9 %
53
-0.
ο*.
.26%
25
Figure 3. As i n Figure 1, for the lowest energy broken symmetry state of the doubly-reduced ZnFe cluster. Valence o r b i t a l s are numbered within each symmetry and have been placed i n three columns depicting the location of their primary charge d i s t r i b u t i o n . V e r t i c a l bars indicate occupied o r b i t a l s . 3
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
376
THE CHALLENGE OF d AND f ELECTRONS
Even
though
general
the magnitudes
trends
qualitative elsewhere
information (21)
that
spin coupling i n Fe S 4
at the
first
sight
+ 3
delocalization others
lines
(23).
of
Although there
wave m o d e l s , to
they
obtaining
ideas
Furthermore, computers)
are clear
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch026
for
a wide
range
can
a clear
to
relatively
that
incorporate
and
electron
a n d t h e method
i s being
to
t o ab
local
exchange
the method, initio
t h e scheme o u t l i n e d h e r e localized
are
interactions
the
models
shows
explain
results
and
the such
limitations
and
along
should enable
and d e l o c a l i z e d
adapted
muffin-tin
and one c a n
studies
t o b e made b e t w e e n s p i n H a m i l t o n i a n m o d e l s a n d p r a c t i c a l metal
Even
one o f u s ( L . N . ) these
offer
useful
exchange
are not essential
Thus,
too high,
c l u s t e r s , where t h e e x p e r i m e n t a l
of
(with faster
are often
to be r e l i a b l e .
useful:
extension
i n complex c l u s t e r s ,
(22).
approximations forward
path
are l i k e l y
c a n be quite
A
physics
I
simple
perplexing.
straightforward
by
a
Χα and scattered
essential
o f J we p r e d i c t
seen i n Table
look
t h e same
connections calculations
polynuclear
transition
complexes.
Acknowledgments We t h a n k
E c k a r d Munck
Institutes
of Health
f o r many u s e f u l (GM39914)
discussions,
for financial
and The N a t i o n a l
support.
Literature Cited 1.
Spiro. T.G., ed. "Iron-Sulfur Proteins", Vol. 4; New York, John Wiley, 1982. 2. a. Anderson, P.W. Phys. Rev. 1959, 115; 2. b. Hay, P.J.; Thibeault, J.C.; Hoffman, R. J. Am. Chem. Soc. 1975, 97, 4884; c. Ballhausen, C.J. "Molecular Electronic Structures of Transition Metal Complexes", New York, McGraw-Hill, 1979; Section 3-6. 3. Gibson, J.F.; Hall, D.O.; Thornley, J.H.M.; Whatley, F.R. Proc. Natl. Acad. Sci. USA 1966, 56, 987. 4. For a review, see Munck, E.; Kent, T.A. Hyp. Int. 1986, 27, 161. 5. Anderson, P.W.; Hasegawa, H. Phys. Rev. 1955, 100, 675. See also Borshch, S.A.; Kotov, I.N.; Bersuker, I.B. Sov. J. Chem. Phys. 1985, 3, 1009. Borshch, S.A. Sov. Phys. Solid State 1984, 26, 1142. 6. Noodleman, L.; Baerends, E.J. J. Am. Chem. Soc. 1984, 106, 2316. 7. Papaefthymiou, V.; Girerd, J.-J.; Moura, I.; Moura, J.J.G.; Münck, E. J. Am. Chem. Soc. 1987, 109, 4703; Münck, E.; Papaefthymiou, V; Surerus, K.K.; Girerd, J.-J., ACS Symposium Series, L. Que, ed. (in press). 8. a. Noodleman, L. J. Chem. Phys. 1981, 74, 5737; b. Norman, J.G., Jr.; Ryan, P.B.; Noodleman, L. J. Am. Chem. Soc. 1980, 102, 4279; c. Aizman, Α.; Case, D.A. ibid. 1982, 104, 3269. d. Noodleman, L.; NormanJ.G., Jr.; Osborne, J.H.; Aizman, Α.; Case, D.A. ibid. 1985, 107, 3418; e. Noodleman, L.; Davidson, E.R. Chem. Phys. 1986, 109, 131. 9. Noodleman, L.; Case, D.A.; Aizman, A. J. Am. Chem. Soc. 1988, 110, 1001. 10. a. Beinert, H.; Thomsom, A.J. Arch. Biochem. Biophys. 1983, 222. 333, and references therein. See also Ref. 4. 11. Moura, I.; Moura, J.J.G.; Münck, E.; Papaefthymiou, V.; LeGall,
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
26.
12. 13. 14.
15.
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch026
16. 17. 18. 19. 20. 21. 22. 23.
SONTUMETAL.
Spin Coupling and Electron Delocalization
377
J. J. Am. Chem. Soc. 1986, 108, 349; Surerus, K.K.; Münck, E.; Moura, I.; Moura, J.J.G.; LeGall, J. ibid. 1987, 109, 3805. a. Johnson, K.H. Annu. Rev. Phys. Chem. 1975, 26, 39. Case, D.A. ibid. 1982, 33.151. Cook, M.; Karplus, M. J. Chem. Phys. 1985, 83, 6344. Antonio, M.R.; Averill, B.A.; Moura, I.; Moura, J.J.G.; Orme-Johnson, W.H.; Teo, B.-K.; Xavier, A.V. J. Biol. Chem. 1982, 257, 6646. Beinert, H.; Emptage, M.H.; Dreyer, J.-L.; Scott, R.A.; Hahn, J.E.; Hodgson, K.O.; Thomson, A.J. Proc. Natl. Acad. Sci. USA 1983, 80, 393. Stephens, P.J.; Morgan, T.V.; Devlin, F.; Penner-Hahn, J.E.; Hodgson, K.O.; Scott, R.A.; Stout, C.D.; Burgess, B.K. ibid. 1985, 82, 5661. Robbins, A.H.; Stout, C.D. "Iron-Sulfur Cluster in Aconitase at 3.0 ÅResolution", (submitted for publication). Hagen, K.S.; Holm, R.H. J. Am. Chem. Soc. 1982, 104, 5496. Cook, M.; Case, D.A. Quantum Chemistry Program Exchange #465, Bloomington, Indiana. Girerd, J.J.; Papaefthymiou, G.C.; Watson, A.D.; Gamp, E.; Hagen, K.S.; Edelstein, N; Frankel, R.B.; Holm, R.H. J. Am. Chem. Soc. 1984, 106, 5941. Vosko, SN.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58. 1200. Salahub, D.R. In: other ab initio Methods in Quantum ChemistryII. K.P. Lawley, ed. (John Wiley, 1987), p. 447. Noodleman, L., Inorg. Chem. (in press). Bencini, Α.; Gatteschi, D. J. Am. Chem. Soc. 1988, 108, 5763; S. Mattar, personal communication. Yamaguchi, K.; Tsunekawa, T.; Toyoda, Y.; Fueno, T. Chem. Phys. Lett. 1988, 143, 371. For a recent overview of Ab initio calculations, see de Loth, P; Karafiloglou, P.; Daudey, J.-P.; Kahn, O. J. Am. Chem. Soc. 1988, 110, 5676. See also Ref. 6.
RECEIVED October 24, 1988
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
Chapter 27
Role of High- and Low-Spin Electronic States in the Co(NH )6 Exchange Reaction 2+/3+
3
Marshall D . Newton
1
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch027
Institute for Molecular Science, Myodaiji, Okazaki 444, Japan
Ab initio SCF and Möller Plesset calculations with flexible valence basis sets including 4f orbitals are carried out for the ground and first excited spin states of the Co(NH )6 and Co(NH3)6 complexes. The results of the calculations in conjunction with a first-order spin-orbit coupling model yield an estimate of 10 for the electronic transmission factor in the Co(NH )6 exchange reaction using an apex-to-apex approach of reactants, thus providing a mechanism characterized by only a modest degree of non-adiabaticity, consistent with the experimental kinetic data. 2+
3+
3
-2
2+/3+
3
The mechanistic analysis of the kinetics of electron transfer processes involving transition metal complexes in solution continues to stimulate intense theoretical activity (1-17). In terms of the conventional transition state expression for the rate constant for activated electron transfer,
it is of particular importance to assess the various contributions to the activation energy (Ε ) and to the electronic transmission factor (κ ), which is a measure of the probability of successful reaction once the reactants have been activated (v and Γ are, respectively, the effective harmonic frequency associated with the reaction coordinate, and the nuclear tunnelling factor). In the context of the present volume, it is to be emphasized that the techniques of computational quantum chemistry have proven to be valuable tools for estimating and analyzing the relevant activation energies and electronic transmission factor (2-11). Since electron transfer between transition metal complexes generally involves the nominal exchange of valence d-electrons, it is seen that the particular challenge to quantum chemistry is that of treating "d-electrons" -- including the energetics of the various possible electronic states associated with d-electron manifolds, and the †
el
n
1
η
Permanent address: Department of Chemistry, Brookhaven National Laboratory, Upton, NY 11973
0097-6156/89/0394-0378$06.00/0 c 1989 American Chemical Society
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
27.
NEWTON
nature
o f the
orbitals not
states
always
either
from
for
be h i g h o f view
electronic
latter
quantity
since
the
reliable
complexes Electron required states
i n general
with
pairing
schemes
(via
available
d-orbitals)
can y i e l d
reliable
(18-25)»
excitation
Among t h e
arising
energy o r treatment
determining (7-10,16).
valuable
various
as w e l l
information,
types
metal
correlation.
t o obtain
i s certainly
energies,
especially
for
electron
of electron
metal
asradial
of d-electrons
even
and
the
intermediate
ordifferent
intransition
4f orbitals)
correlation
activation for
i n some c a s e s
and i t —
A balanced
inclusion of electron
lengths
state
of transition
spin multiplicities
( 1_1 ) .
effects
complexes,
by v a r i o u s
properties
the
bond
for
different
correlation
o f the
i s necessary
metal-ligand
the
dominated
valence
dominant
i s important
Fock model
necessitates
correlation
reasonable
states
the
of low-lying
super-exchange mechanisms
Hartree
calculation
often
metal
kinetically
of minimizing
i t i s often
through
the
and
density
transmission factor.
of different
states
The
transition
i s the
energies
Although
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch027
may
the
"virtual"
d-orbitals
ligands.
point
the
metal
coordinated
obvious which
the
maximizing of
c o u p l i n g between
of their
electronic is
379
High-and Low-Spin Electronic States
complexes,
(via
angular
relatively
i s expected
diffuse
t o be important
(26-33). With
this
interest,
b a c k g r o u n d we
the
Co(NH )£
Co(NH ) 2+ 3
r
[t
5 2
(the
e
g
/
,
m
T
lowest
species
+
6
2. 4 g
l
reaction
g
r
]
6 [ t
formally
provide
2 + 6
[t2 e /
2
g
2 +
3
4
[t2ge / g
the
two
Co(NH ) 3
6 2
g
T
e
i
low
2 +
Ί
-
3 +
3
g
-
T
l
g
]
3
t
^
3 +
3
l
g
l
g
]
each the
nonadiabatic as written (a t
2
*+ e
g
g
inter-complex e low
( 1_6).
g
probability
Accordingly,
states
s oa s t o
2 +
Co(NH ) 3
+
]
2
E ] g
2 +
Co(NH ) 3
4
2
[t\ e / g
(3)
6
g
6
/ T
for
[tf e{/
ig
Co(NH )
[
+
A]
2g
,
g
(11,17) :
l
6
[t /
6
3 + 6
2 , 4 e / T
While
excited
exchange
3
ig
[ti e{/
g
g
a very
process
(2)
6
m
2
state
process
t o have
Co(NH )
A]
3
g
(4)
6
T
l
g
]
reactions",
+
1 , 2^ / E]
g
e
X
Co(NH )
"cross
6
3 + 6
and
as the
ligand-field for
5 [ t
r
]
reaction
"1-electron"
pathways
g
3
a strongly
as w e l l
be expected
lying
g
]
g
l
< 1 ) , the
e
would
l4 /
+
6
g
2 +
Co(NH )
i n brackets).
indicate
K ^
3
g
4
2
configuration
do n o t
+
6
6 , 1 , [ t / A
i s indicated
Co(NH )
E ]
g
Co(NH )
[t
thus
+
r
]
g
3 +
3
to a "three-electron"
"1-electron"
3
r
,
reaction:
Co(NH )
each reactant,
t o invoke
Co(NH )
-
l
t o a conventional
seeks
and
. 1 / A
g
(34)
within
relative
3 + 6
one where
interchange
one
now c o n s i d e r a s p e c i f i c p r o c e s s o f exchange
electronic
corresponds
and
+
A
2
data
(i.e.,
3
/
3
energy
exchange),
+
Co(NH )
at equilibrium
experimental
2
3
Co(NH ) 3
r
[t
5 2
g
e
3 +
-
6
1, 3 / T
ι
m
g
l
g
]
Co(NH ) 3
r
6 [ t
3 +
, 1 / A A
2
g
+
6
Co(NH ) 3
-ι ] «[ t. 5 e2. / 4T_
2 +
r
l
g
2
g
g
(5)
6
Ί 2
g
]
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
THE CHALLENGE OF d AND f ELECTRONS
380 2 +
Co(NH ) 3
4
[t'geg/
T
These
l
separations
through
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch027
3
+
l g
where
2 +
2
c
E
1
degree
first-order
where
equilibrium states due
T
g
mixing
theory),
ligand-field
experiment
are available
e t a l . (~ 3 3 0 0 - 3 8 0 0
7000
1
(17).
a modified
INDO m o d e l ,
associated
with
than
In orbital
theory
perturbation
"1-electron o f H^f i f
v i a the coefficients
thus
f o r these be known
vary
an i n v e r s e underscoring
quantities.
at the transition smaller
arises
than f o r
because
low-spin
C o - N bond
A variety
ranging
from
ligand-field
reactants
lengths
i s known
of
estimates
i n t h e 2+
the recent
model
estimate
estimate
yields
an estimate
out CI c a l c u l a t i o n s the transition
may b e n o h i g h e r
of
of using
state i n energy
state.
we e v a l u a t e
C o - N bond
the high-spin/low-spin
lengths,
T h e known e n e r g y
geometry
sides of
β
separation
(RHF a n d U H F ) , i n c l u d i n g
theory.
i t s equilibrium
calculated
1
work,
matrix
f o r t h e 3+ c o m p l e x cm"*.
et a l . carried
g
f o r various
f o r t h e 2+ a n d
κ ^ t o have
at larger
and concluded that
2
(7) ]
g
separations ( v i a
-
transition
the present
separations
(e.g.,
c m * ) (1_1) t o a n e a r l i e r
the E / A ^ g
t h e ^"T^g/^A^g
c E
(17)
strength.
(11.17.36).
Larsson
2 + 6
2
+
t o the square
must
favorable
A conventional
cm~T~(36).
case.
the spin-orbit
situation
* A ^ g •+ ^ T j g e n e r g y
complex
l g
hand
may be s u b s t a n t i a l l y
(35) t o be ~ 13,700
Larsson
[ T
separations,
t h e a n a l o g o u s " ~ ^ T i g •+ E g v e r t i c a l
9000 c m "
with
estimates
This
less
4
]
g
the spin-orbit
separations
become r e l a t i v e l y
3
and r i g h t
we e x p e c t
reliable
( U ) .
Co(NH )
b e a sum o f t h e v a r i o u s
and s i n c e
magnitudes
l
spin-orbit
which couples t h e i n i t i a l and
i s not too great
their
+
T
of
2-6, to yield
coefficients
e
geometries
states. energy
of the
electron-transfer
to the left
energy
either
Boltzmann
i n the latter
3 +
3
C
quadratically
(.16),
i n previous
estimation
3
ig
will
]
g
in the reaction
theory
lA +
]
E
by appropriate
Co(NH )
2
for
at
l
2
g
(6)
6
the ground e l e c t r o n i c
l
(1-17).
of having
[tf e{/
analyzed
the h i g h - s p i n / l o w - s p i n energy
The v e r t i c a l
of
->
these
t o reduced
from
3 +
dependence on t h e energy
importance
Furthermore,
]
3
and f o r estimates
The o v e r a l l
perturbation
fourth-order
g
2 +
Co(NH )
of the high-spin/low-spin
K ^ i s proportional
model)
with
into
case,
6
of adiabaticity
Landau-Zener inversely
Since
been
perturbation
(corresponding
coefficients.
state,
mixing
3
C
vary
l
may p a r t i c i p a t e
are the spin-orbit
and w i l l
T
(as determined
7, r e s p e c t i v e l y ) ,
pathways"
3
+
we m a y c o m b i n e E q u a t i o n s
3
ig
6
i . e . , for direct
Co(NH )
[A +
3 +
3
g
estimates
respectively.
states
[t| .J/
w h i c h have
coupling,
l
Co(NH )
]
(11.17).
H^f = J $ i H $ f
Equation
the
states,
first-order
]
g
c and c
element,
the
l
-
i n the former
+
6
3+ s t a t e s , final
i g
/
are required:
spin-orbit
4
A
g
reliable
factors
Co(NH ) [ T
[t|
population
case,
Boltzmann With
6
or v i a spin-orbit
either
mixing
3
studies
thermal
factors) In
]
g
3 +
Co(NH )
excited
theoretical through
+
6
using
electron
separation
i s used t o c a l i b r a t e
ab i n i t i o
energy
molecular
correlation v i a f o r t h e 3+ c o m p l e x and adjust
t h e raw
results.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
27. NEWTON
High-and Low-Spin Electronic States
Computational
Details
The
ab
initio
calculations
second-order Gaussian the
82
computer
Institute
basis
(.39)
Molecular
(14s,
9p,
and
p-GTO's
4p
orbital
with
orbital
GTO
(after
Hay)
(40),
2.58
a^ ,
level,
consists
of
no.
Table
an
a
in
Wachters VI
instead
(8/5/3/1)
1
of
of
The
original
of
basis.
[3/2]),
For
the
the
full
39,
Co
a
2
+
with fit
additional,
and C o
that
as
ions
two
to
a n STO 3d
exponent
at
for
the
MP2
Co
(contraction
employed above,
4-31G
at
diffuse
basis
we
noted the
+
basis
contracted
except
3
the
Wachters
supplemented an
of
computers
least-squares
ligands,
3
Co a t o m s ,
contracted
extended
NH
HITAC
GTO ( w i t h GTO o r b i t a l
(8/4/2)
Reference
a version
the
aQ ),
4f
SCF and
the
1
1.0
the using
for
was
a 2-GTO
single
u s i n g UHF o r b i t a l s ) .
1
as
exponent
and
For
primitives)
(contracted
at
(37,),
adapted
Science.
5d
out
(MP2)
(38)
based on o p t i m i z a t i o n
contraction
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch027
for
carried
level
program
additional
2
were
Môller-Plesset
381
a
[4/1]
thus
basis
3d
yielding
(41)
was
employed. The were using The
parameters
optimized the
computer
parameters
parameters. the
outer
above
functions
for
Co(NH )$
of
For *Ag
Cr
subject
2
^Tjg to
(2+),
the
is
the
ascertained cautions
the
the
for
SCF
(42). in
all
Co s - f u n c t i o n s
employed
and
scaled
of
Co
level,
original
were
two
were
the
a total
contracted
state
hole
in
the
xy
hole
was
the
on the at
for
results
3
+
2
E
by
2.0
in
SCF energy
of
45
the for
contracted
functions
of
also
the
g
in a
for
the
core,
for were
and
2
z
2
for
43,).
Representation
t geg/^Ti 2
to
g
ls-2p(Co) to
yield the
the eg
provide and
Extension of
some o f
the
state,
within
2
small
octahedral
10~ of
5
au
the
use
of
an
orbital.
An
state.
associated with
found
quite
the
energies
required
the
reflect
e
total
intended the
is
t§g g/^Eg
or
the
departure
calculations the
of
C0N5
in
Accordingly,
2
the
energy
was
2
the
the
protons
3
an o x c u p i e d x - y
calculations.
shell
actual
determinant and
Accordingly,
ls-2p this
yield
Reference
correlation
3s/3p
the
lengths
for
NH
au).
a
exponent
(26).
employed
the 3
x -y
employed
MP2
10""
employed
a 4f
optimized
symmetry,
choices of
single
al.
C o - N bond
symmetry
3 (
section
with
D j
For
calculations
the
to
symmetry.
manifold
g
the
structures
i n d u c e d by
orbital
next
Botch et
octahedral 3
n
the
Co a t o m ,
complexes were
Co(NH )$
though 0
of
(2+)
g
in
the
complexes,
x
even
as
t|
configurations. in
reported
levels
atom f2 very
these
for
find
~
20%
the
the
similarity
is
(positive) angular
calculated
2
2
•+ f )
correlation
MP c a l c u l a t i o n energies
in
3s/3p rather
(MP2
small,
is
I
do
II
were
with
only
differences in
by
the
MP
correspond
So a s
not
MP4).
typically
can
a frozen
diffuse
and
included
they
10-20%).
employed
-
energy
can change
shell
of
energies
Tables
are
(both
with
calculations
correlation
in al.
energies
importance
unimportance
Additional
et
series.
increasing (d
UHF
increase
correlation
relative
that
of
the the
seen to
correlation
the
of
o b s e r v e d by J a n k o w s k i
the
of
associated
increasing
isoelectronic
the
indications
and
magnitude
with
calculated
have
Cr
of
account
the
(26)»
the
all
in
to
employing
In
in
MP-level
a n d d2
configuration),
(typically we
entry
those
diffuse
similar
of
in
the
been
the
eV)
as
to
wavefunction
The
(3.4
found
configurations of
more
UHF
calculations
has
indicate
3d
(similar
the
in
the
viewed
study
f-orbital
UHF c o r e .
be
for
the
correlation
d-electronic these
while
of
in
"tight"
energy.
ion,
decrease
The
and
can
d orbital
d-orbitals
angular
electrons
final
(manifested
and
correlation
energy.
Co3+
again,
charge,
on the
is
in
the
calculations
five
expansion.
electronic
the
of
variational
limit
orbital
diffuse
behavior
II.
appreciable source
the
a
to
the
Cr(d^)
orbital
energy
The
correlation
observed.
frozen
ls-2p
of
seen to
relative)
valence a
is
positive are
3d
previous set.
correlation of
to
correlation
the
comparisons of
strongly
The based
2
a given
studies
Table
present
RHF r e f e r e n c e ,
isoelectronic
Similar
provided
in
basis
the
β spin which
a correlation
associated with
their
the
an o r t h o g o n a l
Furthermore,
(for
Additional
of
as
high-spin
their
of
a tight
the
correlation
energy
charge
and to
a natural -» f
the
a n d RHF r e s u l t s ) . correlation
in
atom
and
for
one
b a s e d on v a r i a t i o n a l
importance
d-electron
of
1 eV o f
calculated
MCSCF
the
has
energy
UHF a n d RHF e n e r g i e s ,
orbital
estimate
and d
least
the
configurations of
Co
demonstrated
accounting
at the
mixture
similar
angular
total
in
in
one so
correlation
corresponds roughly
eV o b t a i n e d an
in
in
utility
manifested
one
yields
sets
the
is
nearly
energy,
studies,
the
consider
radial
elements
place
relative
large of
to
electrons)
correlation basis
and
we
the
However,
previously
for
calculations
50-50
offers
have
latter
state.
calculations,
row
and
a
ground
d-orbitals
type,
study
the
(2£)
which
captures
present
in
other al.
6.5,
Thus
which in
case
spin
α-spin
character. the
of
difference
a
states
in
the
the
various
configuration
configuration
d-orbital
of
of
d^
course,
CI
appreciable
The
nominal
tendency
a diffuse
orbitals
2
electronic
I).
Of
transition
3 d
3d^/^D
(Table
consider
examining
the
diffuse
first
the
species
we
by
is
Botch et
multi-reference
the
it
+
complexes,
correlation
complex.
comparison with
neutral
in
compare 3
Co
d^ c o n f i g u r a t i o n
states.
n + 1
and
1
importance 3d
Co(NH3)5 +/3+
on e l e c t r o n
We f i r s t
Cr^
CoiN^)^" "
low-spin
2
the
charge
interest
allow
to
Ions
to
to
eliminate
ls-2p
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
core
for
27. NEWTON
383
High- and Low-Spin Electronic States
Table
I.
Valence
Shell
6
5
3d / D
Species
Correlation
States
Correlated Wave
Energy
o f C r a n d Co3+
/
(eV)
f o r the
a
SCF
Basis
/Reference
Set
a
(8/5/3/1)
(8/5/3)
2.6
1.9
3.4
2.7
Function
Cr
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch027
present r e s u l t s /
UHF 2
(=6.5)
UMP3 previous results
m
7
+
l
+
/ R H F
2
RHF+1+2 Co3
present results
+
/
RHF
d
3.7
3.1
/
RHF
6
4.6
3.5
2.0
0.9
2.2
1.1
^ /
UHF 2
(=6.0)
XJMP3
a)
The c o r r e l a t i o n
wavefunction wavefunction. energy
In
relative
constraints
of
calculations. those
b)
ls-3p
section,
The e x p e c t a t i o n compared w i t h
d)
Variational
(26),
using
multi
reference
CI
correlated
this
c o r r e s p o n d s t o t h e UMP3
were
Both
Table
the spin
relaxed
(to within
II).
and
spatial
i n t h e UHF 0.2
eV)
to
T h e MP c a l c u l a t i o n s
core.
results,
the basis
with
values
the exact
singles
a basis
sets
or without
singles
energy
calculation
Sekiya,
the
value
are as d e f i n e d
the 4f
using
operator of
and d o u b l e s
function
i n the (see
also
and w i t h
CI
relative
CI, a 4f
eV ( r e l a t i v e
a very
at
t h e UHF l e v e l
are to
6.0.
to that
and double
of 4.2
private
of
set similar
a MCSFC c a l c u l a t i o n ,
correlation
(M.
the
d and e ) .
be
e)
of
the r e f e r e n c e SCF
are s i m i l a r
(see also
either
c)
from
of
work,
results
a n d UMP4
F o r the present
footnotes
i s the energy
t h e UHF o r RHF e n e r g y .
T h e UMP3
C
RHF
t h e RHF w a v e f u n c t i o n
a frozen
previous
/
to the energy
the present
to
f r o m UMP2
employed
energy
relative
C
large
to
a RHF r e f e r e n c e
used
i n the present
using
configurations
virtual,
yielded
work.
a
t o t h e RHF e n e r g y ) .
atomic
orbital
basis
set
communication).
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
A
obtained
384 Co
THE CHALLENGE OF d AND f ELECTRONS in
the
orbitals the
Co(NH3>6 were
Ν atom
Is
the
excluded
orbitals,
From t h i s conclude
calculations.
also
that
brief for
set
model.
of
This
zero-valent reduced the
4f is
MP2
to
be
s-level
multi-configuration
ls-2p
of
with
a
provides with
transition
systems
nucleus
increase
the
importance
diffuse
state
did
not
and
the of
and more
appear
Co d - o r b i t a l
(40)
with
energies,
interest basis,
cases
formal
diffuse be
in
the
a
structure
of
(26-32)
where
t
the
occupation
including
to
we
here,
and
electronic
some n o t a b l e
the
it
of
d-level
suitable
metal
along
correlation
complexes
from
reference
unoccupied
space.
split a
five
orbitals),
atomic
charged
contrasted
Even though
additional
level,
state
highest
MP a c t i v e
positively
Coulombic f i e l d
orbitals.
the
primitives,
ground
valence
the
from
consideration
the
single-reference
single
The
("virtual"
d
of
a
and
crucial,
f
we
included
Co(NH )$ 3
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch027
calculations. Properties As
of
a partial
compare
in
for
the
the
3+
the
bond
Co(NH^)^ measure
Table
*Aig and
of
III
of
equilibrium
bond
gravity,
as
noted
expected
(18-25)
the in
has
been
lengths
noted for
agreement In
the
the
oxidized
and give
experimental
data.
geometry
length
rt
for
the
*Ai
arithmetic
constant
species
(.14),
equilibrium thus
activation The
is and
total
symmetric (2+)
a value
~
shift
of
0.10
(14.15).
the
features
(the Â),
CoN bond
the
one
of
bond
lengths
of
find
that
the
(0.22 the
both
simply
Â)
about
of
2.15
twice  for
as
by
that
CoN bond
reaction (.16).
constant, be
that
(SCF)
value
Since for
and ~
16
energy
Thus
from
aside give
potential
the
the
the value
(MP2).
estimated uniform
encouraging
energy
surface
2+
calculated
kcal/mole is
the
the
r t , b a s e d on of
If
then
the
corresponding experimental the
the
transition
species
force
activation
calculations
of
reactions,
Co-N distances.
is
with
kcal/mole
inner-shell
kcal/mole Â,
2.05
of
better
in
of
it
bond
parents.
reduced
breathing
equilibrium
lengths 15
energy
Perhaps states,
significantly
defined
system would
estimate
17
in
g
the
each
major
oxidation
quantity
reaction
As the
 for
of
metal-ligand
component
oxidized
species
energy
this
( 1_1 ) . than
(0.10
neutral
a n d we
exchange
3+
~
in the
both
the
about
variations
for
"inner-shell"
al.)
(the
of
complexes.
difference
(.16),
for
be
metal
are
center
cases
positive
for
the
species
the
to
obtained
complex
longer
uniform
charge-transfer
is
two
to
is
are
calculated
ions
the
we
for
those
of
experimental
important
that
than
same
SCF bond
estimated
compared
activated
(3+)/^Ti
g
mean
bias
results
(2V)
which minimizes the
this
same v a l u e
the
et
when
positive
The
for
s p e c i e s had
force
of
the
Larsson
for
have
(2+)
this
some t r a n s i t i o n
quantities
( r t ) , common t o
partners, each
for
reduced
calculations state
modest
E
about
lengths
experiment
important
2
we
we
lengths
states
addition,
of
present
context
most
C o - N bond
electronic
In
bond
metallocene
the
ground
work
previously
with
SCF c a l c u l a t i o n s ,
experimental
disproportionate
However,
observed
the
SCF L e v e l
earlier
relatively the
the
the
calculated
values.
with
of
and
respectively.
the
the
species)
consistent
quality
at
octahedrally-constrained
experimental exaggeration
Complexes
species,
lengths
t
3 +
the
complexes,
length
and
/
calculated
and ^ T j g
2+
2 +
to
be
horizontal
account
of
pertaining
length.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
to
is
27. NEWTON
Table
II.
Valence
Ion
Shell Correlation
Level
Energy
a
2 +
7
3 +
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch027
a)
2
+
and
Co
3
+
Ions
Basis (8/5/3)
4
MP 2
3.24
3.19
1.86
MP 3
3.26
3.22
1.84
6
b
3.26-3.35
5
(3d / D) MP 2
1.91
1.90
0.83
MP 3
2.01
2.00
0.87
MP4
b
2.01-2.04
B a s e d on UHF o r b i t a l s .
the
Co
(3d / F)
MP4 Co
for
(8/5/2/1)
(8/5/3/1) Co
385
High-and Low-Spin Electronic States
fourth
significant
UHF s p i n
figure
contamination 2
the
calculations
employed
b)
corresponds to presence
The range
frozen
of
ls-3p
values.
shows up o n l y
in
T h e MP
UHF c o r e . or
absence
of
single
and
triple
excitations.
Table
III.
Equilibrium
C o - N Bond L e n g t h s
(Â) Results
Species
Present
Experiment
Results*
N e 4
T
l
2
E
g
< g
3 +
>
(2+) (2+)
ab 1.97
2.29
2.198
2.23
[2.13]
a)
Based on an o c t a h e d r a l RHF o r U H F .
c)
Taken
d)
Calculated
e)
Reference
f)
References
14 a n d 4 5 .
g)
References
14 a n d 4 4 .
h)
Estimate
all
three
f
2.07
crystal minus
Larsson
d
5
b)
from
Ar
0
C0N5
a l .
e
INDO
2.03
2.02
0.10
2.27
2.14
2.20
2.09
h
framework.
and aqueous
experimental
of
0.10 [0.10]
h
initio
et
solution
diffraction
data.
value.
11.
based
on the
assumption
that
the
same
Ar value
applies
species.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
to
386
THE CHALLENGE OF d AND f ELECTRONS
Energy The
Splittings
salient IV.
state
separations.
The
charge
t|g
pair
(MP2)
state
is that
the of
it
which
calculated
known v e r t i c a l
for
brings
T
•* ^ l g
T
E
"* ^ g
amount.
excitation
A similar
calculations
the
raw has
been
energy
Â,
and
2.15
to
the
length
is
state,
differing
 and
2.25
Â,
that
the
length.
In
fact,
expected
to
be
due
to
of
state
the
force
linear
factor,
section,
splittings
interest. states (ΔΕ the
As
slightly 2 +
4
for
~ 4200
All
MP2
T i
2
g
/ E g
splitting 4f
by
virtual
splitting The are
initial
by
the lg/^
final
lg
spin
spin
is
state,
electrons. useful the
detail of
estimates
electronic
in
the
next
greatest was
estimated
find
the
two
rt to
be v e r y
close
in
larger
rt, although
still
to
spin
energy
state
appreciably
IV
990
include and
= 2.15
the
splitting
at
(by well
au,
Co 3 s / 3 p
of
the
reduces
Â.
below
the
of
the the
^Aig/^Tjg 4
the
T^g 4
on t h e
reduction
in
increases
excitations
both
effect
a net
shell
UHF c o r e
the
While
magnitude
(e.g.,
the
as p a r t
1
cm"
0
2-7),
by
the
exchange
aside
above
from
non-interacting.
we in
discussed
which,
states note the
that
the
If
of
to
and
the 2
Eg
2
Tjg/ Eg
only
initial
While
(Equation
H^f 3)
for is
rt are
bond
we
440
cnT*
exchange
energies
at
final
matrix the
on
constraint
state
full
reaction
of
transition-state and
based
length
now c o n s i d e r t h e
the
relative
bimolecular in
for
the
associated with
electron-transfer
introduction. T
higher
state
shell
rc jj
splittings
by d i f f e r e n c e s
A
of
Â).
affected
l
spin bond
(17).
by ~ 0.06
states
in
this
at
low-spin
represented
E
procedure
each
lower
and
C o - N bond
r t ) , we
at
Table
much s m a l l e r
otherwise
(Equations
- 1
for
are
Â
at
is
cm"^
increase
reactants
and
in
splitting
state
individual
found
Freezing cm
at
2
the
2.15
equilibrium
corresponding calculated
9000
results
510
(where
lengths
for with
the
the
2+
linear
state
are
and
This
(eg)
1
cm"" )
same
energy
for
more
Â)
,
C o - N bond
estimating
in
2.15
state
geometry
that
energies
is
= 2.25
(rt),
The
of
space
correlation rCo-N
than
space.
11 g
complexes
1
estimate
active 4
2+
Tj^
with
r t (~
Reference
cm" ).
cm"*)
at 4
the
that
with
(13,700
11.
and
obtained
the
anti-bonding
dealt
e
the
3+
constant)
for
than
1
the
expression should provide
K ^,
equilibrium
earlier
the
the
= 2200
Tj[g
in
below
constant
-
cm
Reference
of
of
For purposes of
transmission
lie
variation
—
vertical
6000
of
term
at
coincidence
complex
the
results
magnitude
of
in
for
a
calculated
separation
the
respectively.
smaller
population
the
separations. the
the
MP2
which
the
into 3+
for
a range
of
a common f o r c e
in
somewhat
a higher
Nevertheless,
(with
only
Â)
the
adopted
in
same c o n s t a n t
energy
estimating
over
extent
in
been d e c r e a s e d by
p r o c e d u r e was
harmonic
T^g
i n c r e a s e d by
has
Level
corrected
correct
by the
splitting
interpolation
exact
to
(2.29
expressions for
linear
3
MP2
displayed the
eg p r o c e s s
2
for
MP2
the
high-spin transition
length
splittings
is
a tg
are are
to
energy
provided 2.05
interest
* A ^ g •+
high-spin/low-spin by
Complexes at
each process
transition
excitation
Convenient
3 +
low-spin
*A^g C o - N bond
A c c o r d i n g l y , the
^ l g
^
seems r e a s o n a b l e
energy
the
^ lg
MP2
primary
Since the
broken,
2 +
Co(NH3)^
corresponds to
the (35).
of
results
transition
namely,
the
results
Table each
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch027
of
high-spin
c o m p l e x may coupling,
element,
H^f,
and be
as
discussed
"three-electron"
expected
to
be
quite
small,
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
as
27. NEWTON
Table
IV.
High-Spin/Low-Spin Energy for
Δ Ε
Co(NH 3+
ΔΕ A)
Results
at
Ξ
2 +
E
3
(
) 3
2
+
/
3
+
l
)
g
-
2
Ξ
E( E )
l
E( A
(2.15 Â)
=
E( T
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch027
2 +
(2.15 Â)
C)
Dependence
A
E
Results
i
3
2
ΔΕ
shift
2 +
=
on C o - N Bond
+
+
Length
r
(
r
at
< C0N>
"
CoN>
"
3 +
(2.07 Â)
=
ΔΕ
2 +
(2.29 Â)
=
is shifted
b) X
The t o t a l A
l
g
value
separation CoN bond
(uncorrected
2.2
(corrected
8.1
4
-0.079181,
-1.840157
(
c)
interpolation
Linear
2.05 d)
 (3+)
Based
e)
T
l
g
(2+));
or 2.25 Â
on c a l c u l a t e d
splittings
have
Reference
3+
MP2 r e s u l t ) < »
D
j
MP2
(uncorrected
a
result)
MP2
3
result)^ J
0
1
0
9
6
-
3
· 1
"
4
6+
·
6
2
9
,
·
2
"CON
7
r
CoN
13.7 (present 6.4
) upward
(present
been
results)
(IND0)
amount
cm" ) at
MP2 ( l s - 3 p
-1.241193
(^T
results
1
cm"" )
the calculated
(2.07 Â ) , as d e s c r i b e d
and - 0 . 6 4 4 1 7 9 ,
e
(5900
core)
to -1717 a u ) . -0.068346,
of
results)
needed
so as t o match the
1
(13,700
length
(relative
(3+));
result)
(SCF r e s u l t )
b y t h e same
(ΔΕ
SCF and u n c o r r e c t e d
respectively (
downward
the ^ T i g - ^ ^ g
experimental
MP2
Equilibrium^
ΔΕ
equilibrium
1
cm"" )
)
(SCF r e s u l t )
3.3-4.7 a)
g
2.4
^13.8
I
l
10.0 (corrected
r
A
3
)
i g
4
-
g
r 4.0
ΔΕ
(10
Complexes
6
T
Separations
rt
3 +
ΔΕ
B)
387
High-and Low-Spin Electronic States
l g
(3+));
text.
energies are,
-1.259428 -0.707071,
-1.803265
based
^ i g
i n the
(
2
E
g
(2+)).
on 2.15 Â and
either
(2+). (SCF) e q u i l i b r i u m
corrected
r£ jj 0
(see footnote
values. a and t h e
T h e MP2 text).
11.
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
to
388
THE CHALLENGE OF d AND f ELECTRONS
noted
above,
(Equation
4)
orientation (16)
that
crossing l
H
ifl
the
coupling
may
be
of
the
reactants
the
diabatic
the
2-state
is
avoided).
reactants
to
attain
octahedral provided
an
contact of
detailed
suppression expected energy ~
of
of
the
1
2
is
less
Â),
at
in
the
evaluate
at
K p, e
Electronic cases
of
expression Equation
^Ajg/^Tig
electron 2-state
for
the
E^ i s
effective state in v
409
n
«
1
E^ in
(the
the
e
the
for
are -
each
relevant
clear
require
that
as w e l l
as
an
below
that
energetics
/ T i g couple.
g
required 16
to
kcal/mole the
the Further
ascertain
appropriate
in
of
4
*Ai
the
of
state
the
on t h e
of
rt
(see
next
if
the
value) previous
section
to
state.
( κ ^ ) between
relatively leads
weakly
coupled
to
the
following
transmission factor,
K p,
(as
to
( ^ / E
f
-
E^
n
better
+ E
o
2
+
u
t
3
/
to
1
2
/
energy
/ h v
e
in
we m u s t
the
-
-
97.7
a
by
7.
e
we
effect 87.6
To complete H^f
or
final
is
implicit less
take
(for
small
replaced
evaluate
K ^
reaction
kcal/mol
include
the
factors
10% w h e n
exchange
by E q u a t i o n
is
n
initial
exponential
than
(6.15).
(8)
n
and v
+
s h o u l d be
E q u a t i o n 8)
and c *
R T )
K
associated with
E^ ~ 4Et;
kcal/mol
c
i
expansion of
represented
express
2
2 H
Co(NH3)é
room t e m p e r a t u r e w
is
may p o s s i b l y p l a c e
that
reorganization
accurate
a n d E\
97.7
schematically (iZ)»
is
*
frequency
reactions,
evaluating at
H^f
weight
the
|H^f|,
Landau-Zener model
efc
total
For
cm"
tunnelling,
e
8
0.2).
exchange
K g,
the
harmonic
Equation -
have
other
(II),
transition
electronic
system (15.16)
than
to
1
1): K
where
due
cm"
two
their
of
nevertheless
proceed
transfer
the
1000
complex would
(based
Transmission Factor
reactants,
rms v a l u e
encounter
is
underway)
we
-
the
fact
4
for
by
the
the
of
recall
the
reduced
of
somewhat
g
axis
assuming equal
favors
meantime,
the
the
4
energy
of
which
magnitudes
energy
A x / T g value
the
reduction in
be
(i.e.,
determination it
the
will
solved
*Aig/ Tig transition of
reactants
1
the
In
H^f
barrier
couple
(currently
than
in
relaxation
spite
activation
section).
For
2
A^g/ Eg
^Aig/^Eg
calculations A^g/ Eg
accurate
We
s e m i e m p i r i c a l methods
(2),
simulations,
the
couple
2.15
value
orientations
"non-interacting" 1
an
Jahn-Teller
^ig/^Tig (r*
of
smaller
the
(6.7.11).
states
is
a common 4 - f o l d
suggesting that
While
computer
final
Various
the
(7.11).
distribution
and
process
on
associated with
configuration
along
(2).
apex-to-apex
orientation.
state
energy
equation
*Aig/ Eg
depending
T h e maximum b a r r i e r
frameworks
the
activation
initial
orientations of
cm"" ,
transition
apex-to-apex
estimates
1
1000
the
2
"1-electron"
-
secular
-
25%
the
as
in
zeroth-order
of
corresponds
for
high
when
crossing
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch027
as
for
the the
Following
from
kcal/mol
when
evaluation
of
process Buhks et
al.
as,
c
-
- ( J f
C
-
-(J
4
T
H
l
**3τ l
H l
g
g
Q
*
* SO
2
l
)/AE
E
2 +
)/AE
A
A
l
(rt)
3 +
(rt)
g
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
(9)
(10)
27.
NEWTON In
mixes 4
Tig
the
the
0^
which mix
perturbation to
coupling
usual
E
, and
g
both
one
and
these
for
two
the
rate
two
initial and
2
where
+
3d
and
one
model
the
/ 2
o
The
states
Γ3
two
distinct
2
and 4 ξ
+
/ 3
effective
with
spin-orbit
(17). of
must
for
the
activated
state)
s
first-order
available
states
for
final
+
ξ
is
+
are
final
2
at
2
orbital
levels
of
H
*A|g.
9 yields
level
lie ξ
with
g
degenerate
the
states
Γ3
constant
initial
2
coupling via ^Tj
Equation
At
c.
Co
of
four-fold
state,
low-lying the
spin orbit
component
accordingly,
c^
Γ$ ground
integral in
(46),
different
theory,
2-state
(i.e.,
two
denoted
the
Because species
2
with
group
symmetric
yields
coefficients, respect
double
totally
state
389
High- and Low-Spin Electronic States
the
electron
be
Co(NH3)^
reaction,
2 +
the
transfer
generalized
as
follows : 2
Publication Date: June 8, 1989 | doi: 10.1021/bk-1989-0394.ch027
, act
Σ
et
Y
(11)
k
J ik
j,k-l where
the
normalized
Boltzmann
factor
yj
is
given
by
3 y
-
g.
exp
-
(βΕ
)/(
J
J
In
Equation
Epir^) -
12,
doubly-degenerate state
Γ3 the
(through
degeneracy
first
with
Equation
the
11
four
v ,
and E t
T,
n
n
"paths"
or
in
and
of
e
«
i
e f i f
We h a v e the
Co(NH3)5 with
Υ,
j - i
J
a
-
7
= E^/4 in
ν
ell
are
l
c'
AE
3 +
employed
(rt)
are
1
1/5 /
2
(
For using
C J L
),
1,
we
second
gj
are
. find
it
convenient
to
(13)
r
to
be
common t o
and where
all
(using
2
ΥΛΗ|ϊ) ](π3/Ι J
j k-i transfer
assuming and
l
f
Equation
K
RT)
1
/
2
/hv
x
8),
(14) n
in
ξ
2
+
»
Reference
H^f
(where
and
final
initial
an e f f e c t i v e (2+)
1_7).
2
simplicity,
a one-electron model
numerators (c ),
and
2
we
of
3d
ξ
3
+
the -
cm"
1
(c») the
in
wavefunctions
AE
1
9 and
10 of
many-electron
2 +
the h,
c^,
c, 2
same
(rt)
and 1
cm" ,
have
magnitudes
ξ. wavefunctions
on e a c h Co
contain
of
appropriate
(the
10,000
units
nominally
by
cm""
for
spin-orbitals
denote
coefficients
600 and
Equations 2
construct
basis
,
(2200
in 1
i
The v a l u e s
IV 6 '
-
j,k
states
Hamiltonian,
represented
and w i t h
cm
the Γ3
1-electron
states
515
from Table
The 1
Γ3
integrals, of
wavefunctions,
using
3/5 '
Although these
1
the
The
βχρ(-βΙ*)
n
11,
Σ
combinations
taken
respectively).
2[
j
(3+)
rj
evaluated
values
«
the
multi-determinantal and
Γ
η
taken
Equation
κ
complex)
relevant
coupling).
f
possible 2 +
the with
2
Σ
evaluated
four
since
^ β Τ ) "
Equation
2
κ
β Ξ
(12)
degenerate
spin-orbit
4),
) K
follows:
^ et
involved
accidentally
form
as
k
where
(2
βχρ(-βΕ
K
E$(Tj) « £ 2 ^ 3 )
and is
order
factors
Consistent reexpress
0,
state
Γ7
g
k«0
only
site. the
In The Challenge of d and f Electrons; Salahub, D., el al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1989.
390
THE CHALLENGE OF d AND f ELECTRONS
thirteen 6
d /d
7
electrons
final
states),
the
influence
the
various
also
in
the
L7,
To
along
z-axis,
matrix
element
n3d
z
23d 2»
*
"pathways"
as
between t
l
i
e
s
a
m
all
other
[. j Equation
14
which
an e
in
This spin-orbit
3d
by
with
calculated
for
limit,
to The
this
two
axis
Co
possible
matrix
achieve be
atoms,
1-electron
of
1000
cm"
(6,7),
and
elements
We f i n a l l y
to
Hamiltonian
calculations
elements.
the
especially
4-fold
on the
four
off-diagonal
of
obtain
h,
1
to
as
for
well
the
14,
VjKÏ) ]
ev, ft
= (0.030 h
is
-
1.0
which
at
least
recent
3 d z 2 3 d 2 2
2
to
the
in
)
2
(15)
degree
K ^ e
mean v a l u e
of of
an o r d e r