Theoretical Aspects of Homogeneous Catalysis: Applications of Ab Initio Molecular Orbital Theory

218 Pages·1995·13.2 MB·English

by  P. H. M. Budzelaar

THEORETICAL ASPECTS OF HOMOGENEOUS CATALYSIS
Catalysis by Metal Complexes
VOLUME 18
Editors:
R. UoO, University 0/ Milan. Milan. Italy
B. R. JAMES, The University o/British Columbia. Vancouver. Canada
Advisory Board:
J. L. GARNETI, The University o/New South Wales. Kensington. N.S.W .• Australia
S. D. ITIEL, E. I. du Pont de Nemours Co .. Inc .. Wilmington. Del. . U.S.A.
P. W. N. M. VAN LEEUWEN. University 0/ Amsterdam. The Netherlands
L. MARKO, Hungarian Academy a/Sciences. Veszprem. Hungary
A. NAKAMURA, Osaka University. Osaka. Japan
W. H. ORME-JOHNSON, M.I.T. • Cambridge. Mass .• U.S.A.
R. L. RICHARDS, The University 0/ Sussex at Falmer. Brighton. U.K.
A. YAMAMOTO, Tokyo Institute o/Technology. Yokohama. Japan
The titles published in this series are listed at the end o/this volume.
THEORETICAL ASPECTS
OF HOMOGENEOUS CATALYSIS
Applications of Ah Initio Molecular Orbital Theory
Edited by
PIET W.N.M. VAN LEEUWEN
Van 't HofJ Research Institute. University of Amsterdam. The Netherlands
and
KEIJI MOROKUMA
CL. Emerson Center. Emory University. Atlanta. U.S.A.
and
JOOP H. VAN LENTHE
Theoretical Chemistry Group. University of Utrecht. The Netherlands
SPRINGER -SCIENCE+BUSINESS MEDIA, B.V.
Llbrary of Congress Cataloglng-In-Publlcatlon Data
Theoretical aspects of homogeneous catalysis : applications of ab
initia molecular orbital theory I edited by Plet W.N.M. van Leeuwen
and Keiji Morokuma and Joop H. van Lenthe.
p. cm. -- (Catalysis by metal complexes ; v. 18)
Inc l udes index.
ISBN 978-94-010-4212-3 ISBN 978-94-011-0475-3 (eBook)
DOI 10.1007/978-94-011-0475-3
1. Catalysis. 1. Leeuwen, P. W. N. M. van (Plet W. N. M.)
II. Morokuma, K. (KelJi), 1934- III. Lenthe, Joop H. van.
IV. Series.
OD505.T474 1995
547' .050595--dc20 94-31171
ISBN 978-94-010-4212-3
Printed on acid-free paper
AII Rights Reserved
© 1995 Springer Science+Business Media Dordrecht
Originally published by Kluwer Academic Publishers in 1995
Softcover reprint ofthe hardcover 1s t edition 1995
No pari of the material protected by this copyright notice may be reproduced or utilized in any
form or by any means, electronic or mechanical, including photocopying, recording or by any
information storage and retrieval system, without written permission from the copyright
owners.
Typeset in the United Kingdom
TABLE OF CONTENTS
P.W.N.M. VAN LEEUWEN: Preface
P.H.M. BUDZELAAR AND J.H. VAN LENTHE: An Introduction
to Quantumchemical Organometallic Chemistry 3
P.E.M. SIEGBAHN AND M.R.A. BLOMBERG: Oxidative
Addition Reactions 15
N. KOGA AND K. MOROKUMA: Alkene Migratory Insertions
and C-C Bond Formations 65
N. KOGA AND K. MOROKUMA: Carbonyl Migratory Insertions 93
E. FOLGA, T. WOO AND T. ZIEGLER: A Density Functional
Study on {2s+2sJ Addition Reactions in Organometallic Chemistry liS
A. DEDIEU: Wacker Reactions 167
R. ZWAANS, J.H. VAN LENTHE AND D.H.W. DEN BOER: A Study
on Possible Intermediates in the Epoxidation of Ethene Catalysed by
Manganese(Ill)-chloro-porphyrin 197
Index 215
v
P.W.N.M. VAN LEEUWEN
PREFACE
Homogeneous catalysis plays an important role both in the laboratory and in
the industry. Successful applications in industry involve new polymerisation
processes with complexes of zirconium and related metals, new carbonylation
processes employing palladium and rhodium, ring opening polymerisations,
and new enantioselective isomerisation catalysts as in the preparation of
menthol. Also in the synthesis of organic compounds in the laboratory highly
selective homogeneous catalysts represent an irreplaceable part of the toolbox
of the synthetic chemist. Examples of such reactions are cross-coupling (Ni,
Pd), nucleophilic substitution of allylpalladium complexes, Heck reactions
(Pd), asymmetric epoxidation, Wacker type reactions (Pd), asymmetric
hydrogenations (Rh, Ru), reactions of chromium complexes, enantioselective
reactions with Lewis acids, reactions with the McMurry reagent, etc. There
is hardly any multistep organic synthesis that does not involve one of these
metal catalysed reactions. Most of these catalysts have been developed by
empiricism.
The metal catalysed processes consist of a series of elementary steps which
often have been studied in isolation in organometallic chemistry. The knowl
edge of such elementary steps - effect ofligands, anions, coordination number,
valence states - has greatly contributed to the development of improved cata
lysts for the reactions mentioned above. In addition to the empirical approach
theoretical methods have given support and guidance to the development of
improved processes. Often the key steps of a cycle escape from a direct ob
servation and then theoretical contributions are even more wanted. Thus, we
have seen in recent years a growth of both molecular mechanics and
quantumchemical studies in homogeneous catalysis. Until a decade ago the
quantumchemical studies mostly concerned extended Hilckel methods which
gave a clear qualitative insight in the symmetry of the orbitals involved in
the elementary steps of the organometallic catalysts. Especially for those
instances where empirical methods failed the need for more quantitative data
on the elementary steps grew in order to guide the experimental work. In the
last decade ab initio MO methods and Density Functional Theory methods
have been successfully introduced for the study of elementary steps of cata-
P. W N.M. van Leeuwen et al. (eds) , Theoretical aspects of homogeneous catalysis, 1-2.
© 1995 Kluwer Academic Publishers.
2 P.W.N.M. VAN LEEUWEN
lytic cycles. The aim of this book is to familiarise the people who work on
the development of homogeneous catalysts with the recent advances in this
field. It is not the aim of this work to consider the methods in detail. The
prominent elementary steps of homogeneous catalysis are dealt with and the
reader will learn about the most up-to-date treatment of these steps.
We will see that quantitative predictions can be made for a variety of el
ementary steps. For a certain reaction a trend for ligand effects can be pre
dicted as well. Most authors claim, under the best circumstances, an accuracy
of the calculated activation energies of 3 kcal per mol. In terms of predicting
"rates and selectivities of catalysts" this is still rather disappointing; both
rates and selectivities may be off by two orders of magnitude. One would
like to know selectivity enhancements of an even smaller magnitude. Fur
thermore, in a calculation we can consider only steps that are very similar. In
catalysis often competitive reactions occur for which the pathways are not
suited for a quantitative comparison, e.g. ionic species, solvent molecules or
different substrates may be involved. More importantly, the actual catalysts
contain far more atoms than can be handled by our computers in spite of the
enormous progress made in the last two decades. The introductory Chapter I
outlines some of the possibilities and limitations.
The study of oxidative addition reactions (Chapter 2) shows the importance
of ligands and the geometry of the complexes formed in this reaction as
compared to the gaseous metal atom. Chapters 3 and 4 deal with migratory
insertion reactions which have received a great deal of attention in literature.
In this instance the experimental knowledge on e.g. the effects of ligands on
the rate of this reaction is much less developed although this reaction is of
great industrial importance. This is clearly an example of a reaction for which
a direct experimental study is difficult when a very fast catalyst is involved.
Organometallic 2+2 reactions (Chapter 5) are important in metathesis and
polymerisation reactions. Nucleophilic attack of coordinated alkenes is dis
cussed in Chapter 6 (Wacker reactions). In the last chapter (7, epoxidation
with manganese porphyrin catalysts) the reader is confronted with the occur
rence of paramagnetic states which may well be typical of epoxidation/
porphyrin reactions. Thus, a broad spectrum of organometallic catalytic re
actions or their elementary steps is discussed employing also a broad spec
trum ab initio methods thus giving an impression of the state of affairs. It is
hoped that the experimental chemists will find useful concepts in this vol
ume that may help in the development of better catalysts.
Piet WN.M. van Leeuwen
P.H.M. BUDZELAARAND J.H. VAN LENTHE
AN INTRODUCTION TO QUANTUMCHEMICAL
ORGANOMETALLIC CHEMISTRY
I. Introduction
During the last two decades, computational organic chemistry has earned its
place as a discipline complementary to experimental chemistry. Starting out
as a somewhat esoteric occupation of sometimes doubtful relevance, it is
now recognised as a valuable research tool used by many researchers in both
industry and academia. The calculations have become accurate enough to be
relevant for experimental chemists and they have become quick enough not
to be a full-time occupation.
The continuing expansion of computer technology and associated improve
ment of computer software has made this possible. The improvements are
nicely illustrated by a table from the GAUSSIAN brochure, showing the tim
ing for Gaussian92 test job number 178[1]:
TABLE I
RHF/6-3I G** Single point energy for tri-amino-trinitro-benzene (300 basis functions)
Program Computer system Approx. CPU time
Polyatom (c. 1967) CDC 1604 200 years
Gaussian 80 VAX 111780 I week·
Gaussian 88 CrayY-MP I hour
IBM RS/6000 Model 550 4.5 hours
Gaussian 92 CrayY-MP 9 minutes
Cray C90 4.5 minutes
IBM RS/6000 Model 550 I hour
486 DX2/50 20 hours
• Ignoring memory and disk limitations.
In the area of organometallic chemistry, it is taking computational chemis
try much longer to achieve similar importance. There are a number of rea
sons for this, which we will discuss in more detail below. Nevertheless even
in this field "computer chemistry" is growing in importance. It has passed
the stage of qualitative orbital diagrams and is now making at least semi
quantitative predictions that can help experimental chemists.
3
P.WN.M. van Leeuwen et al. (eds). Theoretical aspects o{homogeneous catalysis, 3-\3.
© 1995 Kluwer Academic Publishers.
4 P.H.M. BUDZELAARAND J.H. VAN LENTHE
The purpose of this book is to highlight some of the insights obtained from
theoretical studies and make them available to researchers in organometallic
chemistry and homogeneous catalysis. Therefore, attention in this volume is
focused on results and interpretation, rather than on methodological aspects.
Each of the following chapters focuses on a particular reaction type. Subjects
are Oxidative Addition, Alkene Migratory Insertions and C-C Bond Forma
tion, 2+2 reactions, Wacker-type reactions and Epoxidation.
The treatment is not uniform; different authors treat their subjects at dif
ferent levels of sophistication, which is brought about by a combination of
the requirements of the systems studied and hardware and software limita
tions, as well as by the personal preference of the authors. The chapters con
tain a significant amount of theoretical background themselves.
In this introductory chapter we summarise the history ofc omputational chem
istry in organic and organometallic chemistry. We consider briefly the theoreti
cal methods useful in the study of organometallic compounds and the choices
one has to make when using them. Finally, we discuss the kind of results one
may obtain from a computational study and list some books, that may be of use
to the reader, who is interested in the quantumchemical methods.
2. A history of theoretical organic chemistry
The earliest example of a "theoretical model" used in organic chemistry is
probably the development by Le Bel and Van 't Hoff of the "tetrahedral car
bon" model [2, 3]. This model, while purely descriptive, could be used to
explain or at least bring some order into a large part of organic chemistry.
When quantum mechanics was postulated in the beginning of this century, it
was immediately applied to the covalent bonds of organic compounds, lead
ing to qualitative understanding of the nature of the chemical bond [4].
The calculation of n-orbital energies using the Huckel method [5] in the
1930's is probably the first example ofa "computational chemistry" approach.
Soon thereafter the treatment was generalised to include a-bonding, leading
to the Extended Huckel method [6-9] in the 1950's. These qualitative meth
ods provided a framework for discussing the electronic structure and chemi
cal behaviour of organic molecules. They produced concepts like "forbidden"
and "allowed" reactions, "frontier orbitals" and the famous Woodward
Hoffmann rules [IOJ.
The advent of electronic computers provided the opportunity for quantita
tive calculations. Three different classes of methods, each with their own
INTRODUCTION TO QUANTUMCHEMICAL ORGANOMETALLIC CHEMISTRY 5
area of application, made their entry into organic chemistry, approximately
simultaneously in the period of 1970-1980:
- Ab initio methods attempt to solve the Schrodinger equation in a fairly
rigorous way. They are used to study the electronic structure in "small"
molecules. Such calculations nearly always employ model systems, i.e.
simplified versions of the molecules that are studied experimentally, to
investigate some effects in isolation.
- Semi-empirical methods use most of the fonnalisms of ab initio methods,
but replace the parts of the energy expression, that are difficult or time
consuming to calculate, by approximations fitted to give the best results
for a set of reference molecules. These methods can handle larger systems
than the ab initio approach and are used for "medium sized" molecules.
They may fail for molecules with unusual bonding characteristics, which
were not present in the "reference set", but they can be at least as accurate
as the ab initio methods for standard molecules.
- Force-field calculations treat molecules as "ball and spring" classical sys
tems, ignoring electronic structure and quantum mechanics altogether.
Again parameters (bond strengths, steric repulsion, etc.) are fitted to re
produce experimental data. Since organic chemistry is dominated by lo
calised, covalent bonds, force field methods can be very accurate in this
area (better than either ab initio or semi-empirical methods), provided
that the molecule studied is very similar to the ones used in the reference
set from which the parameters were determined. Because of the very sim
ple nature of the energy expressions force-field calculations can handle
very large systems (up to 106 atoms). A serious disadvantage is that the
user has to assign atom and bond types a priori, based on chemical knowl
edge or intuition, which degrades the predictive value of the theory.
All three approaches owe much of their success and acceptance to the
development by dedicated research groups of widely distributed general pur
pose computer codes. For ab initio programs the premier example is the
GAUSSIAN series of programs by Pople et al. [II, 12] and many others are
now available like GAMESS, CADPAC, MOLPRO, TURBOMOL, SPAR
TAN, etc. For Semi-empirical programs the scene was set by the MINDO
[13-15]-MNDO [16]-AM [17] series of programs by Dewar and others,
and for Force-field calculations we may mention Allinger's MM [18-20]
programs. Although there is a wealth of programs to choose from, consid
erable standardisation has been obtained in the field of computational or
ganic chemistry.