Topics in organometallic chemistry vol 12 theorectical aspects of transition metal catalysis 2005 springer

12
Topics in Organometallic Chemistry

Editorial Board:
J.M. Brown · P.H. Dixneuf · A. Fürstner · L.S. Hegedus
P. Hofmann · P. Knochel · S. Murai · M. Reetz · G. van Koten


Topics in Organometallic Chemistry
Recently Published and Forthcoming Volumes

Palladium in Organic Synthesis
Volume Editor: J. Tsuji
Vol. 14, 2005
Metal Carbenes in Organic Synthesis
Volume Editor: K. H. Dötz
Vol. 13, 2004
Theoretical Aspects of Transition
Metal Catalysis
Volume Editor: G. Frenking
Vol. 12, 2005

Transition Metal Arene p -Complexes
in Organic Synthesis and Catalysis
Volume Editor: E.P. Kündig
Vol. 7, 2004
Organometallics in Process Chemistry
Volume Editor: R.D. Larsen
Vol. 6, 2004
Organolithiums in Enantioselective
Synthesis

Volume Editor: D.M. Hodgson
Vol. 5, 2003

Ruthenium Catalysts and
Fine Chemistry
Volume Editors:
C. Bruneau, P.H. Dixneuf
Vol. 11, 2004

Organometallic Bonding and
Reactivity: Fundamental Studies
Volume Editors: J.M. Brown, P. Hofmann
Vol. 4, 1999

New Aspects of Zirconium Containing
Organic Compounds
Volume Editor: I. Marek
Vol. 10, 2005

Activation of Unreactive Bonds
and Organic Synthesis
Volume Editor: S. Murai
Vol. 3, 1999

CVD Precursors
Volume Editor: R. Fischer
Vol. 9, 2005

Lanthanides: Chemistry and Use
in Organic Synthesis

Volume Editor: S. Kobayashi
Vol. 2, 1999

Metallocenes in Regio- and Stereoselective
Synthesis
Volume Editor: T. Takahashi
Vol. 8, 2005

Alkene Metathesis in Organic Synthesis
Volume Editor: A. Fürstner
Vol. 1, 1998


Theorectical Aspects
of Transition Metal Catalysis
Volume Editor : G. Frenking

With contributions by
D. V. Deubel · G. Drudis-Sole · G. Frenking · A. Lledos · C. Loschen
F. Maseras · A. Michalak · K. Morokuma · G. Musaev
S. Sakaki · V. Staemmler · S. Tobisch · G. Ujaque · T. Ziegler

2
3


The series Topics in Organometallic Chemistry presents critical overviews of research results in organometallic chemistry, where new developments are having a significant influence on such diverse areas as organic
synthesis, pharmaceutical research, biology, polymer research and materials science. Thus the scope of
coverage includes a broad range of topics of pure and applied organometallic chemistry. Coverage is designed
for a broad academic and industrial scientific readership starting at the graduate level, who want to be

informed about new developments of progress and trends in this increasinly interdisciplinary field. Where
appropriate, theoretical and mechanistic aspects are included in order to help the reader understand the
underlying principles involved.
The individual volumes are thematic and the contributions are invited by the volumes editors.
In references Topics in Organometallic Chemistry is abbreviated
Top. Organomet. Chem. and is cited as a journal
Springer WWW home page: springeronline.com
Visit the Top. Organomet. Chem. contents at springerlink.com

Library of Congress Control Number: 2004116518
ISSN 1436-6002
ISBN-10 3-540-23510-8 Springer Berlin Heidelberg New York
ISBN-13 978-3-540-23510-8
DOI 10.1007/b94252
Bibliographic information published by Die Deutsche Bibliothek
Die Deutsche Bibliothek lists this publication in the Deutsche Nationalbibliographie; detailed bibliographic
data is available in the Internet at <>.
This work is subject to copyright. All rights are reserved, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting,
reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication
or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965,
in its current version, and permission for use must always be obtained from Springer-Verlag. Violations are
liable to prosecution under the German Copyright Law.
Springer is a part of Springer Science+Business Media
springeronline.com
© Springer-Verlag Berlin Heidelberg 2005
Printed in Germany
The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a
specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.
Typesetting: Fotosatz-Service Köhler GmbH, Würzburg

Production editor: Christiane Messerschmidt, Rheinau
Cover: design & production GmbH, Heidelberg
Printed on acid-free paper 02/3141 me – 5 4 3 2 1 0


Volume Editor
Professor Dr. Gernot Frenking
Philipps-Universität Marburg
Fachbereich Chemie
Lahnberge
35032 Marburg, Germany


Editorial Board
Prof. John M. Brown

Prof. Pierre H. Dixneuf

Dyson Perrins Laboratory
South Parks Road
Oxford OX1 3QY


Campus de Beaulieu
Université de Rennes 1
Av. du Gl Leclerc
35042 Rennes Cedex, France


Prof. Alois Fürstner

Prof. Louis S. Hegedus

Max-Planck-Institut für Kohlenforschung
Kaiser-Wilhelm-Platz 1
45470 Mühlheim an der Ruhr, Germany


Department of Chemistry
Colorado State University
Fort Collins, Colorado 80523-1872, USA
hegedus@lamar. colostate.edu

Prof. Peter Hofmann

Prof. Paul Knochel

Organisch-Chemisches Institut
Universität Heidelberg
Im Neuenheimer Feld 270
69120 Heidelberg, Germany


Fachbereich Chemie
Ludwig-Maximilians-Universität
Butenandstr. 5–13
Gebäude F
81377 München, Germany

Prof. Gerard van Koten

Prof. Shinji Murai

Department of Metal-Mediated Synthesis
Debye Research Institute
Utrecht University
Padualaan 8
3584 CA Utrecht, The Netherlands


Faculty of Engineering
Department of Applied Chemistry
Osaka University
Yamadaoka 2-1, Suita-shi
Osaka 565, Japan


Prof. Manfred Reetz
Max-Planck-Institut für Kohlenforschung
Kaiser-Wilhelm-Platz 1
45470 Mülheim an der Ruhr, Germany



Topics in Organometallic Chemistry
is also Available Electronically

For all customers who have a standing order to Topics in Organometallic
Chemistry, we offer the electronic version via SpringerLink free of charge.

Please contact your librarian who can receive a password for free access to the
full articles by registration at:
springerlink.com
If you do not have a subscription, you can still view the tables of contents of
the volumes and the abstract of each article by going to the SpringerLink
Homepage, clicking on “Browse by Online Libraries”, then “Chemical Sciences”,
and finally choose Topics in Organometallic Chemistry.
You will find information about the
– Editorial Board
– Aims and Scope
– Instructions for Authors
– Sample Contribution
at springeronline.com using the search function.


Preface

It has been stated in the past that the search for new catalysts has more the
character of an art than a science discipline. This is because there was usually
more speculation than true knowledge about the reaction mechanisms of
catalytic processes. Even the identity of the catalytically active species was
frequently not known, which is the reason that systematic testing of all possibly interesting compounds for catalytic reactions was carried out. This is
costly and time consuming. The situation has changed in the last decade
because much progress has been made in understanding the mechanisms of
many catalytic reactions. Besides sophisticated experimental tools, quantum
chemical calculations of transition states and reaction intermediates played
a prominent role in gaining much better insight into the fundamentals of
transition metal catalysis. Estimating solvent effects and the calculation of
spectroscopic data are now routinely included in many theoretical studies.
Although the design of new catalytically active species is still largely a trialand-error process, modern research is guided by theoretical calculations in

the search for new catalysts, which helps researchers to focus on more promising compounds. The progress in quantum chemical method development has
led to the present situation where theory and experiment are synergistically
used in an unprecedented manner. In particular, the calculation of transition
metal compounds is no longer a too-difficult task for quantum chemistry
because efficient methods are available for dealing with many-electron atoms
and with relativistic effects.
The seven articles in this volume do not provide a comprehensive view of
theoretical investigations of catalytic reactions, because the field has expanded
already beyond the scope that can be covered in one book. The contributions
written by experts in the field exemplarily demonstrate the strength but also
the present limitations of quantum chemical methods for giving insights into
the mechanism of transition-metal mediated reactions. Because the development of new theoretical methods is still a very active research area, much
progress can be expected in the coming years.
Marburg, Germany, February 2005

Gernot Frenking


Preface

Contents

Transition Metal Catalyzed s -Bond Activation and Formation Reactions
D. G. Musaev · K. Morokuma . . . . . . . . . . . . . . . . . . . . . . . . .

1

Theoretical Studies of C-H s -Bond Activation and Related Reactions
by Transition-Metal Complexes
S. Sakaki . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

Enantioselectivity in the Dihydroxylation of Alkenes
by Osmium Complexes
G. Drudis-Solé · G. Ujaque · F. Maseras · A. Lledós . . . . . . . . . . . . . 79
Organometallacycles as Intermediates in Oxygen-Transfer Reactions.
Reality or Fiction?
D. V. Deubel · C.Loschen · G. Frenking . . . . . . . . . . . . . . . . . . . . 109
Late Transition Metals as Homo- and Co-Polymerization Catalysts
A. Michalak · T. Ziegler . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Co-Oligomerization of 1,3-Butadiene and Ethylene Promoted
by Zerovalent ‘Bare’ Nickel Complexes
S. Tobisch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
The Cluster Approach for the Adsorption of Small Molecules
on Oxide Surfaces
V. Staemmler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Author Index Volume 1-14 . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263


Topics Organomet Chem (2005) 12: 1– 30
DOI 10.1007/b104397
© Springer-Verlag Berlin Heidelberg 2005

Transition Metal Catalyzed s -Bond Activation
and Formation Reactions
Djamaladdin G. Musaev (

) · Keiji Morokuma

Cherry L. Emerson Center for Scientific Computation and Department of Chemistry,
Emory University, 1515 Dickey Dr., Atlanta GA 30322, USA

,

1

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2

2

The Role of the Lower-Lying Electron States of Transition Metal Cations
in Oxidative Addition of the s -Bonds (such as H-H, C-H and C-C) . . . . . . .

2

3
3.1
3.2

Role of Cooperative Effects in the Transition Metal Clusters . . . . . . . . . .
Reaction of Pt and Pd Metal Atoms with H2/CH4 Molecules . . . . . . . . . . .
Reaction of Pd2 and Pt2 Dimers with H2/CH4 Molecules . . . . . . . . . . . . .

6
7
9

4

s -Bond Activation via Nucleophilic Mechanism: the Role of Redox Activity

of the Transition Metal Center – Hydrocarbon Hydroxylation
by Methanemonooxygenase (MMO) . . . . . . . . . . . . . . . . . . . . . . .

10

5

Vinyl-Vinyl Coupling on Late Transition Metals
Through C-C Reductive Elimination Mechanism . . . . . . . . . . . . . . .
5.1 Reductive Elimination from PtIV Halogen Complexes
[Pt(CH=CH2)2X4]2– (X=Cl, Br, I) . . . . . . . . . . . . . . . . . . . . . . . . .
5.2 Reductive Elimination from Mixed PtIV Complexes
[Pt{cis-/trans-(CH=CH2)2(PH3)2}Cl2] . . . . . . . . . . . . . . . . . . . . . .
5.3 Reductive Elimination from PtII Halogen Complexes
[Pt(CH=CH2)2X2]2– (X=Cl, Br, I) . . . . . . . . . . . . . . . . . . . . . . . . .
5.4 Reductive Elimination from PtII Complexes with Amine
and Phosphine Ligands [Pt(CH=CH2)2X2] (X=NH3, PH3) . . . . . . . . . . .
5.5 Reductive Elimination from PdIV Complexes [Pd(CH=CH2)2X4]2– (X=Cl, Br, I)
5.6 Reductive Elimination from Mixed PdIV Complex
[Pd{trans-(CH=CH2)2(PH3)2}Cl2] . . . . . . . . . . . . . . . . . . . . . . . .
5.7 Reductive Elimination from PdII Halogen Complexes
[Pd(CH=CH2)2X2]2– (X=Cl, Br, I) . . . . . . . . . . . . . . . . . . . . . . . .
5.8 Reductive Elimination from PdII Complexes with Nitrogen
and Phosphine Ligands [Pd(CH=CH2)2X2] (X=NH3, PH3) . . . . . . . . . . .
5.9 Reductive Elimination from RhIII, IrIII, RuII and OsII Complexes . . . . . . . .
5.10 General Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.11 Comparison of the Vinyl-Vinyl (Csp2-Csp2)
and Alkyl-Alkyl (Csp3-Csp3) Reductive Elimination . . . . . . . . . . . . . . .
6

.

17

.

18

.

19

.

21

.

21
23

.

23

.

23

.

.
.

23
24
24

.

26

Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

27

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

27

References


2

D. G. Musaev · K. Morokuma

Abstract The factors controlling the transition metal catalyzed s-bond (including H-H,
C-H and C-C) activation and formation, the fundamental steps of many chemical transformations, were analyzed. It was demonstrated that in the mono-nuclear transition metal
systems the (1) availability of the lower lying s1dn–1 and s0dn states of the transition metal
atoms, and (2) nature of the ligands facilitating the reduction of the energy gap between the

different oxidative states of the transition metal centers are very crucial. Meanwhile, in the
transition metal clusters the “cooperative” (or “cluster”) effects play important roles in the
catalytic activities of these clusters. Another important factor affecting the catalytic activity
of the transition metal systems shown to be their redox activity.
Keywords s-Bond activation and formation · Transition metal systems · Catalytic activity

1
Introduction
Sigma-bond (including H-H, C-H and C-C) activation and formation are
fundamental steps of many chemical transformations and have been subject of
numerous review articles [1]. It is well accepted that certain transition metal
complexes significantly facilitate the s-bond activation/formation steps, which
may occur via various mechanisms, including oxidative addition/reductive
elimination, metathesis and nucleophilic attack. However, the factors affecting
H-H, C-H and C-C activation/formation still need to be clarified in detail. In
this chapter we intend to analyze some factors that control the catalytic activity of transition metal complexes toward H-H, C-H and C-C bond activation/
formation. Namely, we elucidate the role of (a) lower-lying electronic states of
transition metal cations/atoms, (b) cooperative effects in transition metal clusters, (c) redox activity of the transition metal centers, and (d) the role of metal
and ligand effects in vinyl-vinyl coupling.

2
The Role of the Lower-Lying Electron States of Transition Metal Cations
in Oxidative Addition of the s -Bonds (such as H-H, C-H and C-C)
The study of gas-phase activation of H-H, C-H and C-C bonds of the hydrogen molecule and saturated hydrocarbons, respectively, by bare transition
metal atoms and cations is very attractive for getting insight to the mechanisms and factors (such as nature of metal atoms and their lower-lying
electronic states) controlling catalytic activities of transition metal complexes.
Such studies, which are free from the ligand and solvent effects, have been
subject of many experimental [2] and theoretical [3] papers in the past
10–15 years. Experimental studies indicate that reaction of some transition
metal cations (such as Fe+, Co+, and Rh+) with methane exclusively leads to

the ion-molecule complex M+(CH4), while others (such as Sc+ and Ir+) pro-


Fig. 1 Potential energy profile of the reaction Ir++CH4 calculated at the MR-SDCI-CASSCF level of theory

Transition Metal Catalyzed s-Bond Activation and Formation Reactions
3


4

D. G. Musaev · K. Morokuma

ceed further via oxidative addition mechanism and leads to hydrido-metalmethyl and/or MCH2++H2 products. In order to find some insight to the difference in the reactivity of early and late, as well as first-, second- and thirdrow transition metal cations (TMCs), we have studied the mechanism of
the reaction of M+ (M=Sc, Fe, Co, Rh and Ir) with CH4 at the CASSCF and
MR-SDCI levels of theory in conjunction with large basis sets. The results of
these studies have been published elsewhere [4]. Here we discuss general
trends, factors controlling reactivity of the transition metal cations toward
s-bonds, and predict the most favorable metal cations that can efficiently
insert into s-bonds.
As expected, the first step of the reaction M++CH4 is the formation of
ion-molecule complex M+(CH4) (see Fig. 1, which, as an example, includes the
potential energy surface of the reaction Ir++CH4 at the several lower-lying
electronic states of the Ir cation). Our calculations show that these complexes
are structurally non-rigid, where M+ can nearly freely rotate around the CH4
molecule by the pathways (C2v)´(C3v, TS)´(C2v)´… and/or (C3v)´(C2v,
TS)´(C3v)´…, depending on the nature of metal atom and the electronic
state of the complex M+(CH4). These complexes are stable by 21.9 (M=Sc), 15.5
(13.7±0.8) (M=Fe), 21.4 (22.9±0.7) (M=Co), 16.8 (M=Rh), and 20.7 (M=Ir)
kcal/mol relative to the ground state dissociation limit M++CH4 (experimental values are given in parentheses).

From the resultant M+(CH4) complex the reaction proceeds via the C-H
bond activation transition state (TS) to give the hydrido-metal-methyl cation
complex, HMCH3+. In this step the C-H s-bond is broken and M-H and M-CH3
bonds are formed.Also, the oxidation number of the M-center increases by two.
In order to analyze the reactivity of TMCs toward C-H (as well as H-H and
C-C) bond, one has to elucidate the factors controlling thermodynamics and
kinetics of the reaction M+(CH4)ÆHMCH3+.
Our [4] and other [3] studies have shown that thermodynamics of the
reaction M+(CH4)ÆHMCH3+ is controlled by the two factors. The first factor is
the availability of the s1dn–1 state of the cation M+, which is expected to be the
dominating bonding state in the resultant HMCH3+ complex. The second factor
is the loss of exchange energy (the loss of high-spin coupling (exchange energy)
between valence electrons on the unsaturated transition-metal ion subsequent
to the formation of covalent metal-ligand bonds) upon formation of M-H and
M-CH3 bonds [5]. Upon formation of M-H and M-CH3 bonds, which stabilize
the system, the loss of exchange energy occurs and counteracts the stabilization.
Thus, if the s1dn–1 configuration of the cation M+ is the energetically most favorable one (or easily available, i.e. the promotion energy from the ground state to
the excited s1dn–1 state is small) and the loss of exchange energy for formation
of two, M-H and M-CH3, bonds in the s1dn–1 state is small, the reaction M+(CH4)
ÆHMCH3+ is thermodynamically favorable. Taking into account these factors,
one can easily explain the calculated trends in the exothermicity of the reaction
M+(CH4)ÆHMCH3+, and predict thermodynamically the most favorable reaction M+(CH4)ÆHMCH3+.


Transition Metal Catalyzed s-Bond Activation and Formation Reactions

5

Our studies show that the reaction M+(CH4)ÆHMCH3+ is endothermic by
20.3, 32.3, 37.7 and 40.3 kcal/mol for M=Sc, Fe, Co, and Rh, respectively, while

it is exothermic by 8.7 kcal/mol for M=Ir. This trend in the energy of the reaction M+(CH4)ÆHMCH3+ can be qualitatively explained in terms of the energy
gap between the lower lying s0dn and s1dn–1 states of the metal cations and the
necessary exchange energy loss for formation of two covalent bonds to the
s1dn–1 state. Indeed, the s1dn–1 state is a ground state for Ir+, and thus Ir+ can
easily form two M-L s-bonds. While the ground states of the Sc and Fe cations
are the s1dn–1 states, these cations need 3.7 and 41.4 kcal/mol energy (exchange
energy loss for formation of two covalent bonds to the s1dn–1 state) to form two
M-L bonds. Meanwhile the ground electronic configurations of Co and Rh are
the s0dn states, and the calculated exchange energy loss for formation of two
covalent bonds to the s1dn–1 state plus s0d8Æs1d7 promotion energy are 39.2 and
77.5 kcal/mol for Co+ and Rh+, respectively. Thus, the calculated trend in the
energy of the reaction M+(CH4)ÆHMCH3+, Ir (–6.6)(37.7)energy loss for formation of two covalent bonds plus the cost of promotion
to the s1dn–1 state: Irexchange plus promotion costs for formation of two bonds increases via Sc+
(3.7)(41.6)~Fe+ (41.4)>Co+ (39.2) (in kcal/mol) [5],we expect that the thermodynamic stability of insertion product to decrease via M=Sc+>Ti+>V+>Cr+ and
increase Mn+state becomes the most favorable, and the M-L bond strength significantly increases for the third-row transition metal cations. Therefore, one may expect
that exothermicity of the reaction M+(CH4)ÆHMCH3+ will significantly increase
upon going from the first- and second-row TMCs to the third-row.
Meanwhile, the kinetic stability (existence of the C-H bond activation TS and
the barrier height) of the HMCH3+ complexes is mainly controlled by: (1) the
endothermicity of reaction M+(CH4)ÆHMCH3+. The large endothermicity of
reaction reduces the barrier for the reverse reaction HMCH3+ÆM+(CH4), and
makes HMCH3+ unstable relative to M+(CH4).As noted above, our studies show
that the reaction M+(CH4)ÆHMCH3+ is endothermic by 20.3, 32.3, 37.7 and
40.3 kcal/mol for M=Sc, Fe, Co, and Rh respectively, while it is 8.7 kcal/mol
exothermic for M=Ir. (2) The availability of the s0dn electronic configuration of

the metal center. It is well known that upon oxidative addition of the C-H/H-H
s-bond to transition metal center a charge transfer from the doubly occupied
C-H/H-H s-orbitals to the s (s and ds) orbitals of metal center (called “donation”) and from metal p-orbitals to the s* antibonding orbital of the C-H/H-H
bond (called “back donation”) takes place (see Scheme 1).
These interactions are efficient when the metal center has empty (or partially
empty) s-type s and ds, orbitals and occupied dp orbitals. Since Fe+, Co+, Rh+,
and Ir+ (and all late transition metal atoms) have double occupied dp orbitals but
Sc+ (and all early transition metal atoms) has none, the “back donation” effect
is expected to be larger for late transition metals compared to the early ones.


6

D. G. Musaev · K. Morokuma

Scheme 1 Schematic presentation of “donation” and “back-donation” contributions on
M-(HX) interaction

In contrast, transition metal cations such as Co+ and Rh+, have the s0dn ground
state with empty s-orbital as opposed to Sc+, Ir+, and Fe+, and one may expect
the strong s(C-H)Æ(s,ds) (M) donation effects for the former cations. Thus,
availability of the s0dn state (with doubly occupied dp and empty s-orbitals) for
Co+ and Rh+, as well as Fe+, facilitates both “donation” and “back donation”
effects and makes the s-bond activation significantly easier for these cations
compared to Sc+ and Ir+. This statement is consistent with the calculated trends;
the reverse reaction HMCH3+ÆM+(CH4) occurs without barrier and is controlled by thermodynamic factors for M=Fe+, Co+ and Rh+. Meanwhile reaction
M+(CH4)ÆHMCH3+ occurs with energetic barrier of 38.5 and 2.1 kcal/mol, for
M=Sc and Ir, and is controlled by both thermodynamic and kinetic factors. The
reverse reaction HMCH3+ÆM+(CH4) occurs with 18.2 and 10.8 kcal/mol barriers for M=Sc and Ir, respectively.
On the basis of these discussions, we conclude that: (1) The s0dn state of the

TMC favors the formation of ion-molecule complexes, while the s1dn–1 state leads
to formation of oxidative addition product, (2) the availability of both s1dn–1 and
s0dn configurations of transition metal cations (and atoms) is absolutely necessary for their reactivity toward s-bonds (such as H-H, C-H, C-C, O-H, and N-H),
and (3) all early first-row (Sc+, Ti+ and V+) transition metals cations, having
empty s or d orbitals with a1 symmetry, as well as many third-row (Hf +, Ta+, W+,
Re+, Os+, Ir+ and Pt+) TMCs, can easily activate s-bonds in the gas phase, and
stabilize the oxidative addition product complexes.

3
Role of Cooperative Effects in the Transition Metal Clusters
In this section we expand the conclusions of the previous section to bare transition metal clusters in order to test them again and to identify another,“cooperative”(or “cluster”) effect that affects the reactivity of transition metals. In practice,
transition metals are important ingredients of heterogeneous and nano-catalysts,
therefore clear understanding of their reactivity at electronic level is essential to
unravel the secret of their catalytic activities. Diverse classes of experimental and
theoretical studies already have provided a wealth of information concerning the
electronic structure, spectroscopic as well as dynamic properties of variety types
of clusters, including Ptn [6], Pdn [7], Fen+ [8], Con+ [9], and Nbn+ [10].


Transition Metal Catalyzed s-Bond Activation and Formation Reactions

7

In particular, in the experimental work of Cox et al. [6, 7], the measured rate
constants of CH4 and H2 activation by unsupported Pt and Pd clusters (n=6~24)
show a large variation as functions of the cluster size. In the case of Pt clusters,
it was found that dimer through pentamer were the most reactive, while the
reactivity dropped significantly starting from Pt6. The single Pt atom was less
reactive compared to Pt2–5 by an order of magnitude, and the bulk was less
reactive by at least several orders of magnitude. In the case of Pd clusters, it was

found that the activation rate constants for both H-H and C-H bonds show significant oscillation in terms of the cluster size. The peak value of the measured
rate constant is around n=8 and 10, and the minimum rate constants have been
observed for n=3 and n=9. Understanding the size-dependence of reactivity of
clusters has become one of the most fascinating and intriguing issues in cluster
chemistry [11].
To unravel the reason behind the observed variation of reactivity as a function of cluster size [6, 7], we have chosen to study the detailed mechanism of
H2 and CH4 activation on small Ptn and Pdn (n=1–6) clusters. Results of these
studies have already been published [12]. Here, we intend to analyze the factors
controlling the reactivity of these clusters.
3.1
Reaction of Pt and Pd Metal Atoms with H2/CH4 Molecules
First of all, we recall briefly the electronic structure and reactivity of Pd and Pt
atoms, shown in the previous section, since they are the fundamental building
blocks of the clusters and their characteristics have a major influence on the
properties of clusters. According to a large number of theoretical as well as
experimental studies, Pd and Pt atoms have very different electronic structures
and consequently distinct reactivities. The ground state of Pd atom has a closedshell s0d10 configuration, where the open-shell s1d9(3D) state is 21.9 kcal/mol
above [13]. Therefore, based on the conclusions of the previous section, one
may expect that the Pd atom cannot break the H-H or C-H bonds in H2 and CH4
and rather forms molecular complexes Pd(H2) and Pd(CH4), respectively. The
calculated Pd-H2 bond energy is 16.2 kcal/mol. For the Pt atom, the s1d9(3D)
state is the ground electronic state, while the s0d10 configuration is 11.1 kcal/mol
higher in energy [13]. Consequently, it has been observed that Pt atom breaks
the H-H and C-H bond, in agreement with the conclusions of the previous section. Since the ground state of Pt atom is a triplet but resultant HPtH or HPtCH3
is a singlet, a curve crossing from the triplet to the singlet state is required, and
the minimum crossing point can be viewed as the transition state for the activation process starting from the ground electronic state atom (Fig. 2a). The
binding energy of H2 and CH4 to the Pt atoms is 47.4 kcal/mol and 34.3 kcal/mol,
respectively. Thus, the calculated results for the reactions Pd+H2/CH4 and
Pt+H2/CH4 once again confirm our conclusions from the previous section.

Fig. 2a–d Potential energy profile of the reactions: a Pt+H2/CH4; b Pd2+H2; c Pd2+CH4; d Pt2+CH4

8
D. G. Musaev · K. Morokuma


Transition Metal Catalyzed s-Bond Activation and Formation Reactions

9

3.2
Reaction of Pd2 and Pt2 Dimers with H2/CH4 Molecules
As seen in the potential energy profile of these reactions, shown in Fig. 2b–d,
both Pd2 and Pt2 activate H-H/C-H bonds with very small barrier. In reaction
Pt2+H2/CH4, H-H/C-H activation preferentially takes place on a single metal
atom.Afterwards, one of the H atoms migrates to the other Pt atom over a negligible isomerization barrier. On both the singlet and the triplet state, H-H
activation is expected to be barrierless, while C-H activation has a distinct
barrier on the triplet state for reaction starting from the ground triplet state Pt2.
Nevertheless, since the barrier for C-H activation on the triplet state is small
and lower than the expected minimum of seams of crossing (MSX) between
singlet and triplet in the Pt-CH4 system, it is expected that Pt2 activates the
C-H bond in CH4 with a faster rate than the Pt atom, which is in accord with the
experimental observation of Cox et al. [6, 7].
Calculations show that the ground state of the reactants Pd2+H2/CH4 is a
triplet state, while the product complexes have a singlet ground state. Therefore,
one may expect that the reaction proceeds either on the excited singlet state
surface or through the minimum of triplet-singlet seams of crossing. On the
singlet state potential energy surface of Pd2+H2/CH4 (see Fig. 2b,c) the reaction
is downhill without activation transition state. Meanwhile, the calculated

triplet-singlet MSX lies lower in energy than the triplet state transition state.
Therefore, the H2/CH4 activation by Pd2 starting from the triplet ground state
dimer is expected to proceed via an intersystem crossing mechanism with very
small barrier. Interestingly, the activation of s-bonds occurs only upon perpendicular approach of H-H/C-H bonds to the Pd-Pd bond.
Thus, in contrast to the single atom case where Pd and H2/CH4 form only a
molecular complex and no H-H/C-H bond cleavage occur, two Pd atoms work
“cooperatively” and readily break H2/CH4. This “cooperative” mechanism for
H2/CH4 activation on Pd2 is different from the case of Pt2+H2/CH4. In Pt2 dimer
H2/CH4 activation takes place preferentially on a single atom, while in Pd2 dimer
it occurs on the Pd-Pd bond. Moreover, in the final activation products, H/CH3
groups prefer the bridged sites of Pds, but are localized on metal sites in Pt2.
Those results can be rationalized as the following, as illustrated in Scheme 2. The
singlet state Pd2 consists of mainly two s0d10 Pd atoms, and the LUMO sg has a
correct symmetry to accept electron density effectively from the H2/CH4 s orbital
upon perpendicular approach.As a result, the activation takes place preferentially
in this approach. In the case of Pt2+H2/CH4 reaction, the metal HOMO and LUMO
are of localized metal d character (as established in a number of studies of metal
clusters, the s-s contribution in the metal-metal bonding is dominant, while d-d
interaction is weak). Therefore, the HOMO and LUMO of triplet Pt2 are all of
localized d characters, while the s and s* orbitals that contain large s characters
are much lower and higher in energy, respectively, and therefore activation takes
place preferentially on a single atom (rather than on the Pt-Pt bond, where the
strongest HOMO/LUMO interaction between the metal and H2 is expected.


10

D. G. Musaev · K. Morokuma

Scheme 2 Orbital diagram of triplet Pd2 and Pt2, MO energies (in hartree) are calculated at

the full valence (20el./12orb.) CASSCF level

In the final singlet products, Pt2(H)2/Pt2(H)(CH3) and Pd2(H)2/Pd2(H)(CH3),
the metal dimers actually are in their triplet configurations. In the case of Pt2,
the metal-metal s bonding orbital is low in energy, and the Pt-H/CH3 bond has
large d character. Therefore, the H/CH3 groups in Pt2(H)2/Pt2(H)(CH3) do not
like the bridged sites, but rather localize on each Pt atom. In the case of Pd2, in
its triplet electronic state, the metal-metal s bonding orbital is the HOMO.
Therefore, both the CH3/H-Pd-Pd bonding and antibonding orbitals have much
metal s component. As a result, the H/CH3 ligands prefer the bridged sites
rather than the localized metal sites.
Thus, our studies of the reactivity of Pd/Pt clusters with H2/CH4 molecules
clearly show a “cooperative” effect that could play a significant role in the reactivity of the transition metal clusters. Thus, the catalytically inactive metal
atoms could form very active clusters !

4
s -Bond Activation via Nucleophilic Mechanism: the Role of Redox Activity
of the Transition Metal Center – Hydrocarbon Hydroxylation
by Methanemonooxygenase (MMO)
Another important factor controlling the reactivity of transition metal centers toward s-bond is their redox activity. Indeed, it is well established that
transition metal centers with low redox potential can be active catalysts [14].
For example, let us discuss the reactivity in hydrocarbon hydroxylation by
Methanemonooxygenase (MMO).


Transition Metal Catalyzed s-Bond Activation and Formation Reactions

11

MMO, one of the members of diiron-containing metalloenzyme family, is

an enzyme that catalyzes methane oxidation reaction, i.e. conversion of the
inert methane molecule to methanol [15]. During this reaction two reducing
equivalents from HAD(P)H are utilized to split the O-O bond of O2. One O atom
is reduced to water by 2-electron reduction, while the second is incorporated
into the substrate to yield methanol:
CH4 + O2 + NAD(P)H + H+ Æ CH3OH + NAD(P)+ + H2O
Experimental studies [16] show that the best-characterized forms of the soluble MMO (sMMO) contain three protein components: hydroxylase (MMOH),
so-called B component (MMOB) and reductase (MMOR), each of which is required for efficient substrate hydroxylation coupled to NADH oxidation. The
hydroxylase, MMOH, which binds O2 and substrate and catalyzes the oxidation,
is a hydroxyl-bridged binuclear iron cluster. In the resting state of MMOH
(MMOHox), the diiron cluster is in the diferric state [FeIII-FeIII], and can accept
one or two electrons to generate the mixed-valence [FeIII-FeII] or diferrous state
[FeII-FeII], respectively. The diferrous state of hydroxylase (MMOHred) is the
only one capable of reacting with dioxygen and initiating the catalytic cycle.
X-ray crystallographic studies of the enzyme from Methylococcus capsulatus (Bath) [17] and Methylosinus trichosporium OB3b [18] have unveiled
the coordination environment of the Fe centers of MMOHox and MMOHred.
According to these studies, in MMOHox each Fe center has six-coordinate
octahedral environment (see Scheme 3). Fe ions are bridged by a hydroxy ion,
a bidentate Glu g-carboxylate and a water molecule (or another carboxylate).
In addition, each Fe ion is coordinated by one His nitrogen ligand and one
monodentate Glu carboxylate. The two Fe centers are different from each other
in that one of them (Fe2) has an additional monodentate glutamate carboxylate,
while the other Fe (Fe1) has one additional water molecule. Upon reduction, one
of carboxylate ligands undergoes a so-called “1,2-carboxylate shift” from being
a terminal, monodentate ligand bound to Fe2 to a monodentate, bridging
ligand between the two irons, with the second oxygen of this carboxylate also
coordinated to Fe2. In addition, the hydroxyl bridge is lost, and the other
hydroxyl/water bridge shifts from serving as a bridge to being terminally
bound to Fe1. Also, the terminal water bound to Fe1 in the oxidized form of
MMOH seems to move out upon reduction of the cluster. Thus, in reduced form

of MMOHred the ligand environment of Fe ions becomes effectively five coordinated, which is reasonable since this is the form of the cluster that activates
dioxygen.
It was established that MMOHred reacts very fast with O2 and forms a
metastable, so-called compound O, which spontaneously converts to another
compound called P (see Scheme 4). Spectroscopic studies [19] indicate that
compound P is a peroxide species, where both oxygens are bound symmetrically to the irons. Compound P spontaneously converts to compound Q, which
was proposed to contain two antiferromagnetically coupled high-spin Fe(IV)


Scheme 3 Structural representation of the binuclear Fe center of diferric MMOHox and diferrous MMOHred (see [15])

12
D. G. Musaev · K. Morokuma


Transition Metal Catalyzed s-Bond Activation and Formation Reactions

13

Scheme 4 Experimentally proposed catalytic cycle of MMO (see [15])

centers. EXAFS and spectroscopic studies [20, 21] of compound Q, trapped
from M. trichosporium OB3b and M. capsulatus, have demonstrated that compound Q has diamond core, (FeIV)2(m-O)2 structure with one short (1.77 Å) and
one long (2.05 Å) Fe-O bond per Fe atom and a short Fe-Fe distance of 2.46 Å.
Compound Q has been proposed to be the key oxidizing species for MMO.
In the literature there have been several computational attempts [22–25] to
elucidate mechanism of methane oxidation by intermediate Q. Our results [25]
show that reaction proceeds via the mechanism presented in Fig. 3. Later, this
mechanism was validated by several times and currently is well accepted.
As seen in Fig. 3, reaction of compound Q (modeled as structure I) with

methane starts with coordination of CH4. In general, the CH4 molecule could
coordinate to I via two distinct pathways: O-side and N-side. The O-side pathway corresponds to the coordination of the methane molecule from the side
where the two Glu (carboxylate) located, while the N-side pathway corresponds
to the coordination of CH4 from the two His (imidazole) side. Our calculations
show that both pathways proceed via very similar transition states and intermediates, and the N-side pathway is thermodynamically and kinetically more
favorable than O-side. In spite of this, in this paper we base our discussions only
on the O-side mechanism because it is believe to correspond to the process
occurring in the protein. The coordination of CH4 to complex I leads to the
methane-Q complex, structure II. The interaction between methane and structure I (compound Q) is extremely weak; the complexation energy is calculated
(relative to the corresponding reactants) to be 0.7 and 0.3 kcal/mol for the
9A and 11A state, respectively. Because of unfavorable zero point energy and
entropy factors, it is most likely that the complex II does not exist in reality, and
therefore we will not discuss it.


14

D. G. Musaev · K. Morokuma

Fig. 3 The potential energy profile (in kcal/mol) for both the 9A state and the 11A state of
the methane activation reaction via O-site pathways: (NH2)(H2O)Fe(m-O)2(h2-COO)2Fe(H2O)(NH2)+CH4Æ(NH2)(H2O)Fe(m-O)(HOCH3)(h2-HCOO)2Fe(H2O)(NH2)

Results in Table 1 show that the 9A state of structure I very qualitatively is the
Fe(IV)-Fe(IV) complex. On the other hand, in the 11A state it is the Fe(IV)Fe(III) mixed valence species, where Fe2 is in the Fe(III) state with five spins that
are heavily delocalized onto O2. The calculated fact that I_9 (here and below in
A_B, A is the structure, while B is the electronic state) is lower in energy than
I_11 suggests that Fe(IV)-Fe(IV) is the preferred state for complex I (and compound Q). This is consistent with the experimental conclusions [21].
The activation of the methane C-H bond takes place on the diamond oxygen
O1. At the TS1, structure III, the C-H bond to be broken is elongated from
1.089 Å in II to 1.271 Å and 1.296 Å in transition states III_9 and III_11. Furthermore, the O-H bond is nearly formed, with distance of 1.250 Å and 1.241 Å

at the TSs, compared to 0.983 Å and 0.978 Å in products IV_9 and IV_11,
respectively. These geometrical changes indicate clearly that III_9 and III_11
are the TSs corresponding to the H abstraction process. The H-abstraction
barriers are calculated to be 23.2 and 19.0 kcal/mol for the 9A and 11A states,
respectively, relative to the corresponding CH4 complexes II_9 and II_11, respectively. These values of the barrier are in reasonable agreement with available experimental estimates, 14–18 kcal/mol [23]. The spin densities for TS1, III,
and product IV are found to be similar to each other within their respective 9A
and 11A states (see Scheme 5 and Table 1). Furthermore, the spin densities are
nearly identical between 9A and 11A, except for those on the O2..H..CH3 frag-


Transition Metal Catalyzed s-Bond Activation and Formation Reactions

15

Scheme 5 The spin recoupling scheme in the intermediates of the reaction
Table 1 Relative (in kcal/mol, relative to the 9A reactants) energies, and Mulliken atomic spin
densities (in e) of various intermediates and transition states, for 9A and 11A states, for the
reaction of the complex I with molecule of methanea. The numbers after slash are relative
to the 11A reactants

Structures

DE (in
kcal/mol)

Atomic spin densities (in e)
LnFe2

LnFe1

O2

O1

Hb

CH3c

0.0
–1.5
23.2
11.3
20.6
–41.8
–34.3
23.7

3.44
3.43
4.59
4.64
4.58
4.54
4.49
4.52

3.55
3.52
3.54
3.51

3.23
2.92
2.95
1.69

0.44
0.44
0.40
0.43
0.38
0.35
0.38
0.56

0.30
0.31
–0.37
0.08
0.20
0.00
0.00
0.07

–
–
0.06
0.00
0.00
0.00
0.00

0.00

–
–
–0.55
–0.99
–0.73
0.00
0.00
0.99

5.4/0.0
4.2/–1.2
24.4/19.0
–11.3/5.9
18.6/13.2
–53.9/–59.3
–46.7/–52.1
–15.5/10.1

4.62
4.62
4.63
4.67
4.57
4.48
4.54
4.65

3.47

3.47
3.52
3.58
3.97
4.59
4.54
4.54

1.03
1.03
0.43
0.43
0.50
0.66
0.67
0.42

0.43
0.43
0.55
0.10
0.01
0.00
0.00
0.08

–
–
–0.07
0.00

0.01
0.00
0.00
0.00

–
–
0.59
0.98
0.81
0.00
0.00
0.99

9A

state
I+CH4
II
III
IV
V
VI
VII
IVd
11

A state
I
II

III
IV
V
VI
VII
IVd

a

Here, LnFe stands for the (H2O)(NH2)Fe-fragment. This table does not include the portion
of spin densities located on the bridging carboxylate ligands, each of which may have about
0.10–0.15e spin.
b H atom located between O2 and CH fragments.
3
c The number for the entire CH fragment.
3


16

D. G. Musaev · K. Morokuma

ment. For the CH3 groupitself, the total Mulliken charge (not given in Table 1)
is at most +0.03e for both the 9A and 11A states and the spin densities on this
group for 9A and 11A are of same magnitude but of opposite sign. One can
interpret all these values in the following way. In both TSs, III_9 and III_11, a
radical center begins to develop on the CH3 group, with spin densities of –0.46e
and +0.52e, respectively, and in both intermediates, IV_9 and IV_11, the CH3
group is now a real radical with spin densities of –0.98e and 1.00e, respectively.
The spin densities on Fe1 and Fe2 in IV_9 and IV_11, which formally can be

written as L4Fe(m-O)(m-OH)FeL4 with the methyl radical only weakly interacting via a C...HO interaction, can qualitatively be considered to correspond to
Fe(IV) with four spins and Fe(III) with five spins, respectively. In going from
II_9 to III_9, the very qualitative formal oxidation state of Fe2 changes from
Fe(IV) to Fe(III), while from II_11 (which is already Fe(III)) to III_11, no such
change is required. Since the two Fe centers are coupled ferromagnetically in
both 9A and 11A states, the spin of the CH3 radical in both III and IV has to
couple antiferromagnetically (with negative spin) and ferromagnetically (with
position spin) to make the total spin 2S+1 equal to 9 and 11, respectively. In the
radical complexes IV_9 and IV_11 the interaction of the CH3 radical with the
two iron atoms is very weak and, therefore, their total energies are nearly identical. It is quite interesting that we find that a mixed valence state is responsible
for the methane oxidation reaction. The present spin density analysis clearly
demonstrates that (1) the methane oxidation proceeds via a bound-radical
mechanism, and (2) the first electron transfer from substrate to Fe-centers occurs
through the TS1 and is completed at the resultant bound-radical complex IV.
The second electron transfer from the substrate to Fe-centers starts with the
recombination of methyl radical with the bridging hydroxyl ligand at the transition state, TS2, structure V. Indeed, our results show that in the 11A state upon
going from IV_11 to the methanol complex VI_11, Fe1 changes its formal oxidation state from Fe(IV) with four spins to Fe(III) with five spins, while the spin
density on the methyl radical is completely annihilated upon forming a covalent
bond between CH3 and OH. The transition state V_11 has a spin distribution
between that of IV_11 and VI_11. On the other hand, in the 9A state upon going
from IV_9 to the methanol complex VI_9, the spin density on Fe1 is reduced by
about 0.5, corresponding to the disappearance of roughly one unpaired electron.
Since Fe(V) is not a stable species, it is most likely Fe1 changed its formal oxidation state from Fe(IV) with four spins to Fe(III) with five formal d electrons.
Because of the restriction 2S+1=9, i.e., the total number of unpaired electrons
must be 8 within the Fe(III)-Fe(III) core, Fe1 in VI_9 chose to form one d-lone
pair with only three spins remaining. This complex VI_9 is thus higher in energy than the corresponding complex VI_11 in violation of the Hund rule.
The barrier heights for the CH3 addition to the hydroxyl ligand calculated
relative to the intermediate IV_9 and IV_11 are 9.3 and 7.3 kcal/mol for the 9A
and 11A states, respectively. Obviously, this step of the reaction is not rate-determining, and can occur rather fast. Overcoming the barriers at TS2 leads to
the complexes methanol-complexes VI, L4Fe(OHCH3)(m-O)FeL4, and completes

Transition Metal Catalyzed s-Bond Activation and Formation Reactions

17

the second electron transfer process. The overall reaction I + CH4ÆVI is
calculated to be exothermic by 34.3 and 46.8 kcal/mol for the 9Aand 11A states,
respectively. The final step of elimination of the methanol molecule and regeneration of the enzyme could be a complex process, and possibilities of different
mechanisms exist but have not yet been studied computationally.
Thus, these results clearly show that the Fe-centers are not directly involved
in methane C-H bond cleavage. However, they play a crucial role in the methane
oxidation process, by accepting two electrons from the substrate. Having low
oxidation potential for Fe-centers is definitely an important factor in this
process and significantly facilitates it. Thus, during the hydrocarbon hydroxylation by MMO, the Fe centers undergo multiple reduction and oxidation
steps; at first, MMOHox [ Fe(III)-Fe(III) state] should be reduced (2-electron
reduction) to MMOHred [Fe(II)-Fe(II) state], then it should be oxidized (4-electron oxidation) by O2 molecule to [Fe(IV)-Fe(IV) state], after which it again is
reduced by substrate to Fe(IV)-Fe(III) state and Fe(III)-Fe(III) states (see
Scheme 6). Thus facile oxidation and reduction of Fe-center plays a crucial role
in the hydrocarbon hydroxylation by MMO.

Scheme 6 The redox cycle of Fe-centers during hydrocarbon hydroxylation by MMO

5
Vinyl-Vinyl Coupling on Late Transition Metals Through C-C Reductive
Elimination Mechanism
Another important s-bond activation/formation process discussed in this
article is vinyl-vinyl coupling, shown in Scheme 7. Vinyl-vinyl coupling opens
a convenient route to conjugated 1,3-dienes and is widely employed in many
catalytic coupling reactions. The great potential of the field is still under

continuous development [26, 27] and, therefore, elucidation of the C-C bond
formation mechanism and the factors controlling it are very crucial. In literature, numerous mechanistic studies on C-C reductive elimination and reverse
process, oxidative addition (C-C bond activation), have been reported for di-