On the mechanism of cyclopropanation reactions catalyzed by a rhodium(I) catalyst bearing a chelating imine-functionalized NHC ligand: A computational study

Article abstract of DOI:10.1021/om200594q

A computational study at the DFT level using BP86 and dispersion-corrected D-BP86 methods has been performed on the mechanism of a highly cis-diastereoselective cyclopropanation reaction between ethyl diazoacetate and styrene, catalyzed by a Rh(I) complex bearing a chelating imine-functionalized NHC ligand. The key steps in the mechanism have been elucidated. The favored mechanistic pathway was found to be a stepwise mechanism involving the formation of a Rh metallacyclobutane intermediate. The results from the theoretical study indicate that the diastereoselectivity is determined in the step where styrene coordinates to the Rh(I) carbenoid and that the high cis-diastereoselectivity can be attributed to an unfavorable steric interaction between styrene and the substituents on the N-aryl ring in the ligand system, which disfavors the formation of the trans cyclopropanation product.

Full text of DOI:10.1021/om200594q

Article
pubs.acs.org/Organometallics
On the Mechanism of Cyclopropanation Reactions Catalyzed by a
Rhodium(I) Catalyst Bearing a Chelating Imine-Functionalized NHC
Ligand: A Computational Study
Marianne L. Rosenberg,† Andreas Krapp,‡ and Mats Tilset*,‡
†Department of Chemistry, University of Oslo, P.O. Box 1033 Blindern, N-0315 Oslo, Norway
‡Centre for Theoretical and Computational Chemistry (CTCC), Department of Chemistry, University of Oslo, P.O. Box 1033
Blindern, N-0315 Oslo, Norway
S
* Supporting Information
ABSTRACT: A computational study at the DFT level using BP86 and dispersion-corrected D-BP86 methods has been
performed on the mechanism of a highly cis-diastereoselective cyclopropanation reaction between ethyl diazoacetate and styrene,
catalyzed by a Rh(I) complex bearing a chelating imine-functionalized NHC ligand. The key steps in the mechanism have been
elucidated. The favored mechanistic pathway was found to be a stepwise mechanism involving the formation of a Rh
metallacyclobutane intermediate. The results from the theoretical study indicate that the diastereoselectivity is determined in the
step where styrene coordinates to the Rh(I) carbenoid and that the high cis-diastereoselectivity can be attributed to an
unfavorable steric interaction between styrene and the substituents on the N-aryl ring in the ligand system, which disfavors the
formation of the trans cyclopropanation product.
We have recently reported the use of new Rh(I) catalysts
with chelating imine-functionalized N-heterocyclic carbene
(NHC) ligands in cyclopropanation reactions.26−28 Among
the complexes tested by us, 1 (Scheme 1) has been best
studied. This Rh(I) complex displays a remarkably high
reactivity and cis-selectivity in the cyclopropanation reaction
between ethyl diazoacetate (EDA) and styrene. This is among
the highest cis-selectivities that have been reported in this
reaction and the highest cis-selectivity reported to date using a
Rh catalyst. The catalyst also gives similarly excellent to good
results with a range of different alkenes.
It was therefore of significant interest to learn more about the
mechanism of these Rh(I)-catalyzed cyclopropanation reactions
in order to improve the catalytic system by design. While
Rh(II) catalysts bearing carboxylate and carboxamidate ligands
are the most extensively studied and employed catalysts in
cyclopropanation reactions, there are only a few reports on
Rh(I)-catalyzed cyclopropanation reactions.7,8,29 Some of these
INTRODUCTION
■
Cyclopropanes are important substructures in many synthetic
and naturally occurring biologically active compounds.1,2
Among their reported biological activities are enzyme
inhibition, antimicrobial, antibiotic, antitumor, and antiviral
activities.2 Cyclopropanes are also interesting starting materials
and intermediates in organic synthesis, as they can undergo a
range of synthetically useful transformations.3,4 Metal-catalyzed
cyclopropanation reactions are well known, and the most
common way of obtaining cyclopropanes is the transfer of a
carbene moiety from a diazo carbonyl compound to an olefin.
In such a reaction, up to three new stereocenters can be
formed. One challenge in intermolecular cyclopropanation
reactions is the achievement of good diastereocontrol. As a
result, considerable efforts have been made to develop methods
and catalysts that are highly diastereoselective. Many catalysts
are known that are highly efficient and selective for formation
of the thermodynamically favored trans isomer.5−8 However,
there are only a few reports on highly cis-diastereoselective
catalysts for these types of reactions.8−25
Received: July 5, 2011
Published: December 1, 2011
© 2011 American Chemical Society
6562
dx.doi.org/10.1021/om200594q | Organometallics 2011, 30, 6562−6571

Organometallics
Article
the most stable form in each case will be discussed here. The energies
discussed in the text are Gibbs free energies calculated at 298.15 K,
unless otherwise stated. The geometry optimization and IRC
algorithms implemented in the Gaussian03 program43 were used. All
energy, gradient, and Hessian calculations were performed using the
Turbomole program package.44 This coupling of Gaussian and
Turbomole is possible due to the EXTERNAL option in Gaussian.
The cyclopropanation reactions were conducted experimentally in
CH2Cl2. The solvent effects were investigated by using the COSMO
scheme45 on important steps in the mechanism. No significant effects
were observed on the relative energies or selectivities. All structures
and energies discussed in the text are therefore without solvent
corrections.
Scheme 1. Highly cis-Selective Cyclopropanation Reaction
Catalyzed by the New Rh(I) Complex 1
The possible involvement of triplet transition states has not been
considered, although it is known that singlet−triplet crossing may
sometimes be involved in organorhodium chemistry.46,47
give promising results, but none of them give high
diastereoselectivities, and they have not been extensively
investigated.30−33 Hence, little is known about the mechanism
In order to investigate the importance of dispersive forces, the
empirical dispersion correction of Grimme48 was used in connection
with the BP86 functional using the def2-SVP basis. This level of theory
will be denoted D-BP86/def2-SVP. Single-point calculations on the
BP86/def2-SVP-optimized structures were performed using Møller−
Plesset perturbation theory of second order in combination with the
def2-TZVPP basis, making use of the resolution of the identity
(denoted MP2/def2-TZVPP).49−51
of Rh(I)-catalyzed cyclopropanation reactions.
In this contribution we present a computational mechanistic
study on the cyclopropanation reaction between EDA and
styrene catalyzed by Rh(I) complex 1. This is, to the best of our
knowledge, the first report of a computational mechanistic
study on Rh(I)-catalyzed cyclopropanation reactions.
All computations were performed on the Stallo supercomputer
located at the University of Tromsø, Norway.
COMPUTATIONAL DETAILS
■
All structures (minima and transition states) were fully optimized at
the gradient-corrected density functional theory (DFT) level using the
exchange functional of Becke34 in conjunction with the correlation
functional of Perdew35 (denoted BP86), making use of the resolution
of the identity approximation.36 The basis sets that were used were the
Weigend−Ahlrichs basis sets, denoted def2-SVP and def2-TZVPP,37
and the corresponding fitting basis.38 The influence of the basis set size
on the structures and relative energies for significant steps in the
mechanism was tested, and it was found that the influence was minor.
Therefore all structures presented were obtained using the def2-SVP
basis set in combination with the BP86 functional (denoted as BP86/
def2-SVP). The applicability of gradient-corrected functionals as BP86
for the structural prediction of transition metal compounds is well
documented;39 it was also asserted that BP86 outperforms hybrid
functionals as B3LYP for second-row transition metal compounds. In
combination with the resolution of the identity (RI) technique and
medium-sized basis sets, this level of theory allows for an efficient
exploration of potential energy hypersurfaces. Such an efficient
method was needed in the present mechanistic study. As will be
discussed, BP86 and D-BP86 were tested against MP2, and good
performance was observed. Thus, we evaluate BP86 as being efficient
as well as sufficiently accurate for the purpose of the present
investigation.
RESULTS AND DISCUSSION
■
1. Formation of the Rh(I) Carbenoid. The mechanistic
study was initiated by a geometry optimization of the Rh(I)
complex 1 at BP86/def2-SVP. The calculated bond lengths and
angles were compared with the ones observed in an X-ray
crystal structure of complex 1.26 As shown in Table 1, the bond
distances and angles from the geometry optimization are in
good agreement with the data obtained from the crystal
structure.
Scheme 1 summarizes the experimental conditions for the
cyclopropanation reactions catalyzed by complex 1. The target
in this mechanistic study is the cyclopropanation reaction
between EDA and styrene. The first step in the reactions is the
activation of the catalyst, presumably by abstraction of the
chloride ligand from Rh, using an activating agent such as
AgOTf26 or, preferably,27 NaBArf (where Barf− = [B(3,5-
(CF3)2C6H3)4]−). It is assumed that a cationic Rh(I) species
with a vacant coordination site is then formed (2, Figure 2).
Dichloromethane was used as a solvent in the experiments;
thus it is likely that dichloromethane may coordinate to the
cationic Rh(I) metal center (3). The next step in the
experimental cyclopropanation procedure is the addition of
styrene, and this gives the possibility for formation of a Rh-
styrene complex (4). The styrene complex 4 was found to be
5.5 kcal/mol lower in energy than the dichloromethane adduct
3, indicating that 4 might be the preferred complex present in
Analytic Hessians were used to characterize the nature of the
stationary points on the potential energy surface and to obtain
thermodynamic contributions.40 The Hessian eigenvalues were scaled
by 0.9 in the calculation of the thermodynamic contributions. The
connectivities of minima and transition structures were verified by
intrinsic reaction coordinate (IRC) following calculations.41,42 The
possibility of different conformations was taken into account, but only
a
Table 1. Selected Bond Lengths and Angles for 1 from the X-ray Structure and Calculated Structure at the BP86/def2-SVP
Level
bond
X-ray (Å)
calculated (Å)
angle
X-ray (deg)
calculated (deg)
Rh−N(1)
Rh−Cl
Rh−C(1)
Rh−C(20)
C(5)−N(1)
C(5)−N(2)
2.131(2)
2.367(6)
1.953(2)
1.813(3)
1.293(3)
1.386(3)
2.159
2.351
1.942
1.821
1.311
1.397
C(1)−Rh−C(20)
C(20)−Rh−Cl
Cl−Rh−N(1)
97.39(11)
90.38(8)
93.40(6)
78.53(9)
97.46
91.75
92.93
77.91
N(1)−Rh−C(1)
a
From ref 26.
6563
dx.doi.org/10.1021/om200594q | Organometallics 2011, 30, 6562−6571

Organometallics
Article
part of the ligand system has been observed for neutral Pd
complexes that bear related chelating imine-functionalized
NHC ligands in a six-membered chelate ring,58 but so far this
behavior has not been observed for ligands that form five-
membered chelate rings such as that which is present here.59
A
four-coordinated complex in which styrene and EDA are both
coordinated, and not the imine, was actually found (complex 9,
Figure 3). This complex adopts a strongly distorted square-
planar geometry. The styrene and the CO ligands are trans to
each other with the CO ligand bent out of the plane by 25°.
Complex 9 was found to be as much as 24.6 kcal/mol higher in
energy than free styrene and complex 5, in which the imine and
EDA are coordinated. It is therefore likely that the imine does
not dissociate and, hence, that styrene is not coordinated to the
metal center at this stage in the mechanism.
It is generally accepted that metal-catalyzed cyclopropanation
reactions between diazo compounds and alkenes proceed
through the formation of metal carbenoid species.29,60
Supporting this notion, some Ru carbenoid complexes that
were formed from diazo compounds have been isolated and
characterized by X-ray crystallography, and their use in
Figure 1. Optimized structure of complex 1.
solution. The main contribution to the Rh−styrene interaction
is the donation of the alkene π-electrons to the Rh center;
however, there is some back-donation to the alkene from the
metal center, as indicated by the CC bond distance (1.397
Å), which is close to that of free styrene (1.351 Å).
The next step is the addition of the diazo compound. Several
possible modes of coordination of EDA to the Rh(I) center are
conceivable.52−55 Complexes where the diazo compound is
coordinated to Rh via the carbon atom (5), the terminal diazo
nitrogen (6), the carbonyl oxygen (7), or both nitrogen and
oxygen (8) were computationally located (Figure 3). It is
complex 5 with the diazo compound coordinated via the
carbon atom that is presumed to be the product-forming
complex.
It is conceivable that the EDA-coordinated complexes 5−8
may alternatively form by styrene dissociation from the styrene
complex 4 via the cationic Rh(I) complex 2, which has a vacant
coordination site. Attempts were made to find a five-
coordinated structure in which both styrene and EDA were
coordinated to the metal center, but no such structure could be
located. It is therefore likely that such a process occurs through
a dissociative ligand exchange mechanism, as also found in
theoretical studies of some Cu and Ru complexes.56,57 The
possibility of dissociation of the imine moiety of the ligand
system was also explored, in order to examine whether the
imine might dissociate to furnish a vacant coordination, such
that styrene and EDA might both be coordinated to the Rh(I)
center at the same time. Such dynamic behavior of the imine
cyclopropanation reactions has been demonstrated.61−63
A
Cu carbenoid intermediate involved in cyclopropanation
reactions has also been detected by NMR spectroscopy.64
Rh(I) carbenoid 11 is presumably involved in the cyclo-
propanation reactions catalyzed by Rh(I) complex 1, and
calculated structures that have been located in the pathway for
formation of Rh(I) carbenoid 11 are presented in Scheme 2.
Formation of the Rh(I) carbenoid 11 starts with a
precoordination of the diazo compound via the carbon atom
(5) followed by loss of N2 (TS-10) to generate the Rh(I)
carbenoid 11 (Scheme 2).
In the transition state for loss of N2 (TS-10), the bond
between Rh(I) and the EDA carbon atom is shortened from
2.235 to 2.021 Å, and the carbon−nitrogen bond is elongated
from 1.364 to 1.870 Å compared to complex 5. The Rh−C
bond is somewhat bent out of the chelate ring plane by 5.4°. In
the Rh(I) carbenoid 11, the Rh−C(carbenoid) bond is bent
somewhat further out of the chelate ring plane by 6.7°, causing
a slight deviation from the square-planar geometry normally
expected for symmetrically substituted four-coordinate d8
complexes. Considerable distortions from the expected
square-planar geometry caused by electronic effects have been
reported even in simple metal carbonyl complexes.65,66 The
ester carbonyl group in EDA is almost perpendicular to the
Rh(I) carbenoid. The Rh−C−CO dihedral angle is 110°,
suggesting that the Rh(I) carbenoid and the ester carbonyl
Figure 2. Cationic complexes possibly formed after removal of the chloride. Energies in kcal/mol at the BP86/def2-SVP level. Energies are relative
to an assembly of isolated 2, styrene, and EDA.
6564
dx.doi.org/10.1021/om200594q | Organometallics 2011, 30, 6562−6571

Organometallics
Article
Figure 3. Possible coordination modes of EDA. Energies in kcal/mol at BP86/def2-SVP. Energies are relative to an assembly of isolated 2, styrene,
and EDA.
a
Scheme 2. Formation of the Rh(I) Carbenoid 11
a
Relative energies in kcal/mol at BP86/def2-SVP. Energies are relative to an assembly of isolated 2, styrene, and EDA.
group are not in conjugation. It has been suggested that this
alignment of the carbonyl group allows for favorable
interactions between the carbonyl π* and the Rh(I) carbenoid
π orbital or between the carbonyl π orbital and the Rh(I)
carbenoid σ orbital.52,54,67
It was observed that the transition state (TS-10), leading to
the Rh(I) carbenoid with the ester group pointed away from
the N-aryl ring, is favored by 5.8 kcal/mol over the alternative
transition state, where the ester group is pointed toward the N-
aryl ring (TS-12, Figure 4). This difference is presumably
caused by an unfavorable steric interaction between the ester
group and the N-aryl ring in transition state TS-12 when
compared to TS-10 (details of the structure of TS-12 are given
in the Supporting Information).
The energy diagram for formation of Rh(I) carbenoid 11 is
displayed in Figure 5. The reaction is exothermic and can be
considered to be irreversible. The barrier for loss of N2 from 5
is 12.7 kcal/mol (ΔH‡298 = 13.9 kcal/mol), and this is close to
Figure 4. The sterically less congested transition state TS-10 is favored
over TS-12. Relative energies are in kcal/mol at BP86/def2-SVP.
Energies are relative to an assembly of isolated 2, styrene, and EDA.
the calculated barrier of 14.8 kcal/mol (ΔH) calculated for
EDA with dirhodium(II) formate52 and the experimentally
observed barrier of 15 kcal/mol (ΔH) reported for EDA with
dirhodium(II) acetate.68
6565
dx.doi.org/10.1021/om200594q | Organometallics 2011, 30, 6562−6571

Organometallics
Article
propanation reaction. It is also well known that Rh metal-
lacyclobutanes can be prepared and isolated, and it has been
shown that they can undergo reductive eliminations to form
cyclopropanes.77,78
Despite an extensive computational search, no transition
states could be located for a concerted pathway starting from
Rh carbenoid 11. On the other hand, a well-defined pathway
was found involving the formation of a Rh metallacyclobutane
species. This general mechanism is depicted in Scheme 4. The
first step in this mechanism is coordination of styrene to the
Rh(I) carbenoid, giving a five-coordinate Rh(I) species a. Next,
the Rh metallacyclobutane c is formed via transition state b.
The metallacyclobutane c then undergoes a reductive
elimination via transition state d to furnish the cyclopropane.
There are eight different orientations by which the olefinic
CC bond of styrene can coordinate at the Rh(I) carbenoid
complex 11. All eight possibilities (and the pathways leading
therefrom) were investigated, and all these structures are
described in the Supporting Information. However, only the
lowest energy pathways that lead to each of the cis and trans
isomers will be discussed here. The calculated structures that
are involved in these two lowest energy pathways are displayed
in Scheme 5. The styrene complexes 13a and 13b adopt a
slightly distorted square-pyramidal geometry. The styrene
ligand is coordinated so that the olefinic CC bond is
approximately parallel with the Rh(I)−C(carbenoid) bond. In
styrene complex 13a, styrene is oriented such that the styrene
phenyl ring and the ester group on the carbenoid are on the
same side of the plane that is defined by the styrene olefinic
CC bond and the Rh(I)−C(carbenoid) bond. This
orientation will lead to the cis cyclopropane. In styrene
complex 13b, the styrene phenyl ring and the carbenoid ester
group are on the opposite sides of this plane, leading to the
trans cyclopropane.
Figure 5. Energy diagram for the formation of Rh(I) carbenoid 11.
Relative energies are in kcal/mol at BP86/def2-SVP.
Even though it is only 5 that is the product formed by N2
expulsion, it may be argued that the other EDA-coordinated
complexes 6−8 should also be taken into consideration.52,54
Pirrung et al.69 argued in the case of Rh(II) carboxylates that
reversible, nonproductive binding of the substrate to the
catalyst can affect the reaction rate by lowering the overall
initial energy of the system. If these non-product-forming
complexes were to be taken into account for our system, the
barrier for loss of N2 starting from the species of lowest possible
energy (8) would be the energy difference between this most
stable complex (8) and TS-10, and the activation barrier would
thereby increase to 21.4 kcal/mol.
2. Cyclopropanation. The next step in the catalytic cycle
is the reaction between the metal carbenoid and the alkene to
form the cyclopropane. Two possible mechanisms for this
reaction have been proposed (Scheme 3). One is a concerted
Scheme 3. Two Possible Mechanistic Pathways for
Cyclopropanation of Alkenes with Metal Carbenoids
The possible involvement of imine side arm dissociation was
also explored for these styrene complexes in order to examine
whether a four-coordinate complex with a noncoordinated
imine might be more stable than the five-coordinated styrene
complexes 13a and 13b. Dissociation of the imine side arm did,
however, result in complexes that were significantly less stable
(21.0−23.4 kcal/mol) than the five-coordinated complexes
(calculated structures can be found in the Supporting
Information). This strongly suggests that the imine is
coordinated to the metal center throughout the entire catalytic
cycle; the outcome of the reactions would then be expected to
strongly depend on substituent patterns at the N-aryl group,
which would always be in close proximity to the Rh center.
The Rh−styrene bonding is primarily made up by donation
of the alkene π electrons to the Rh(I) center with some back-
donation from the metal center to the alkene (CC distances,
1.403−1.410 Å). Similarly to the situation in Rh(I) carbenoid
11, the Rh(I)−C(carbenoid) bond in both styrene complexes
13a and 13b are also bent out of the chelate ring plane by 19−
21°. The Rh(I)−C(carbenoid) bond is bent out of this plane to
the opposite side of where styrene is coordinated. The Rh−C−
CO dihedral angles in the carbenoid unit are smaller than
those observed for Rh(I) carbenoid 11. In complex 13a the
Rh−C−CO dihedral angle is 15°, and in complex 13b the
ester carbonyl is nearly in the same plane as the Rh(I)
carbenoid, as the dihedral angle is only 1°. Complex 13a, which
leads to the cis isomer, is 2.4 kcal/mol lower in energy than
complex 13b, which leads to the trans isomer. An unfavorable
steric interaction between the styrene phenyl ring and the
mechanism, where the alkene directly attacks the carbenoid
carbon and the cyclopropane is formed through a three-
membered ring transition state. This mechanism has been
shown by calculations to be the favored pathway for the dimeric
Rh(II) carboxylates52,54,70 and a number of Cu,57,71,72 Ru,56 and
Fe73 catalysts.
The other possibility is a stepwise mechanism involving the
formation of a metallacyclobutane in which the metal has been
formally oxidized. Further, the metallacyclobutane undergoes a
reductive elimination and the cyclopropane is formed. This
mechanism has been proposed to be the favored pathway for
some Pd-74,75 and Fe73-catalyzed cyclopropanations. Hoffmann
and co-workers76 prepared and isolated a palladacyclobutane,
which was demonstrated to be involved in the cyclo-
6566
dx.doi.org/10.1021/om200594q | Organometallics 2011, 30, 6562−6571

Organometallics
Article
a
Scheme 4. Metallacyclobutane Mechanism Found for the Cyclopropanation Reaction Catalyzed by Rh(I) Complex 1
a
[Rh] = (N−C)Rh(CO)+.
methyl groups in the ortho positions on the N-aryl ring in
complex 13b (Figure 6) might explain the energy difference
between these two complexes. In complex 13a the styrene
phenyl ring is not in close proximity to any of the groups in the
iminocarbene ligand system. The distance between Rh and the
α olefinic carbon is somewhat elongated in complex 13b (2.645
Å) compared to 13a (2.481 Å). This difference in bond lengths
might also be enforced by steric interactions between the
styrene phenyl ring and the methyl groups in the ortho
positions on the N-aryl ring.
cyclopropane is lower in energy than the transition state leading
to the trans cyclopropane, by 4.8 kcal/mol. The energy profile
for reaction of the Rh(I) carbenoid 11 with styrene is depicted
in Figure 7.
The addition of styrene to the Rh(I) carbenoid 11 is
endothermic according to this energy diagram. Despite an
extensive search, no transition state could be located for the
coordination of styrene to the Rh(I) carbenoid 11 to give 13a
and 13b. The barriers for the formation of the metal-
lacyclobutanes are as low as 2.5 kcal/mol from 13a leading to
the cis isomer and 4.5 kcal/mol from 13b leading to the trans
isomer. The formation of the metallacyclobutanes is highly
exothermic, and the process can be considered to be
irreversible. Metallacyclobutanes 15a and 15b are substantially
stabilized compared to the Rh(I) carbenoid 11 and free styrene
by 22.5 and 18.1 kcal/mol, respectively. The barriers for the
reductive elimination (TS-16a, 10.8 kcal/mol and TS-16b, 11.2
kcal/mol) are considerably higher than for the formation of the
metallacyclobutanes. This might reflect the higher stability of
the metallacyclobutanes (15a and 15b) compared to the
styrene complexes (13a and 13b). As mentioned, many
metallacyclobutanes are sufficiently stable to be isolated. The
extrusion of N2 in the formation of Rh(I) carbenoid 11
presents the highest barrier in the catalytic cycle (TS-10) and
therefore represents the rate-determining step of the reaction.
This has also been reported to be the case in theoretical studies
of Rh(II)-catalyzed cyclopropanation reactions.52,54
We found it somewhat surprising that the coordination of
styrene at the Rh(I) carbenoid 11 appeared to be endothermic.
Considering that 11 is a 16-electron cationic species, and
despite the fact that it is a four-coordinate square planar
complex, one might envision that the coordination of styrene
would have a stabilizing effect, rendering the process
exothermic. It was observed that there were only very small
changes in bond distances and angles around the metal center
upon coordination of styrene, also indicating that the
coordination of styrene would not be destabilizing. It is well
known that standard DFT methods like the ones used in this
study are not able to account for dispersion forces.48 As the
imine-functionalized NHC ligand system is quite large, one
might expect the dispersion forces to be of considerable
importance. In order to probe this, reoptimizations of the
stationary structures of the key steps in the mechanism were
In the transition states leading from the styrene complexes to
the metallacyclobutanes, the distances between the β olefinic
carbon of styrene and the Rh-carbenoid carbon are shortened
in both transition state TS-14a (2.994 → 2.434 Å) and TS-14b
(3.237 → 2.393 Å), and the Rh−C(carbenoid) bonds are
elongated (TS-14a (1.912 → 1.965 Å) and TS-14b (1.900 →
1.965 Å)). The double-bond character of the olefinic CC
bond in styrene is reduced, as indicated by increasing C−C
bond distances of 0.018 Å (TS-14a) and 0.025 Å (TS-14b).
The Rh(I)−C(carbenoid) bond is still bent out of chelate ring
plane (9−14°) but to a lesser extent than in the styrene
complexes 13a and 13b. The transition state leading to the cis
isomer (TS-14a) is also in this step lower in energy than
transition state TS-14b, leading to the trans isomer, by 4.4 kcal/
mol. In the metallacyclobutanes the bond lengths of the C−C
bonds that originate from styrene have increased further to
1.525 Å for 15a and 1.516 Å for 15b; these bond lengths
correspond to single bonds. It is observed that in the
metallacyclobutane 15b that leads to the trans cyclopropane,
the four-membered ring is more twisted than in complex 15a,
leading to the cis cyclopropane, and the bond length between
Rh and the olefinic α carbon of styrene is greater in 15b than in
15a by 0.028 Å. This may again be attributed to the unfavorable
steric interactions between the styrene phenyl ring and the
methyl groups in the ortho positions on the N-aryl ring in
metallacyclobutane 15b. Metallacyclobutane 15a is 4.4
kcal/mol lower in energy than metallacyclobutane 15b. In the
final transition states (TS-16a and TS-16b) for the reductive
elimination leading to the cyclopropanes, the Rh−C bonds that
undergo cleavage are elongated and the C−C distance is
shortened. The transition states have a strong product-like
character with the cyclopropanes almost completely formed.
Also in this case, the transition state leading to the cis
6567
dx.doi.org/10.1021/om200594q | Organometallics 2011, 30, 6562−6571

Organometallics
Article
a
Scheme 5. Calculated Structures for the Pathways Leading to the cis and trans Cyclopropanes (17a and 17b)
a
Relative energies in kcal/mol at BP86/def2-SVP. Energies are relative to an assembly of isolated 2, styrene, and EDA.
performed using a dispersion-corrected method (D-BP86/def2-
SVP). All reoptimized structures can be found in the
Supporting Information. Only small changes in the structural
parameters were found with this new method. The reoptimized
structures of the Rh(I) carbenoid (11′) and styrene (13a′ and
13b′) complexes are shown in Scheme 6 (all new, or
reoptimized, structures are given the same number as
previously, but are denoted with an slanted prime ′).
In the reoptimized styrene complexes, the Rh−C(carbenoid)
bonds were bent out of the chelate ring plane by 4° less in 13a′
than in 13a and by 4° more in 13b′ than in 13b. Complex 13a′,
which leads to the cis cyclopropane, was found to be 3.9 kcal/
mol lower in energy than 13b′, which leads to the trans
cyclopropane. This compares to the 2.4 kcal/mol energy
difference at BP86/def2-SVP. It was found for both complexes
(13a′ and 13b′) that the Rh−styrene distances were shorter
than what was observed without dispersion correction. This
indicates that the dispersion-corrected method predicts styrene
to be more firmly bound to the metal center. The energy profile
of the reaction between the Rh(I) carbenoid 11′ and styrene at
the D-BP86/def2-SVP level is displayed in Figure 8. The most
important and significant difference between the two methods
is that the dispersion-corrected method in fact predicts the
coordination of styrene to the Rh(I) carbenoid 11′ to be
exothermic. Apart from this, there were only small changes in
the overall energy profile. For all the steps in the mechanism, it
6568
dx.doi.org/10.1021/om200594q | Organometallics 2011, 30, 6562−6571

Organometallics
Article
Figure 8. Energy profile for the cyclopropanation reaction between
Rh(I) carbenoid 11′ and styrene at the D-BP86/def2-SVP level.
Figure 6. Unfavorable steric interaction in styrene complex 13b.
without dispersion correction, the transition state for the loss of
N2 (TS-10′) represents the highest barrier in the catalytic cycle
(12.9 kcal/mol at D-BP86/def2-SVP) and is therefore the rate-
determining step.
Interestingly, the energetic trends at the dispersion-corrected
DFT level of theory, especially the exothermicity of the styrene
addition, reproduce very well the behavior that was observed
when single-point calculations at the MP2/def2-TZVPP level of
theory were performed on BP86/def2-SVP structures (energies
can be found in the Supporting Information). This indicates
that the overall reaction profile is well described at the DFT
level of theory and that dispersion-including methods are
needed in order to describe the styrene addition step more
correctly. The dispersion effects introduced at the D-BP86 level
are quite large. Especially, and as might be anticipated, this is
the case for the first step in Figures 7 and 8, the addition of
styrene, which introduces numerous additional dispersive
interactions, which serve to stabilize the addition product.
From the results discussed in this contribution it appears that
the cis/trans diastereoselectivity of the cyclopropanation
reaction is determined by the orientation of the incoming
styrene upon coordination at Rh. The diastereoselectivity is
determined in the steps where the styrene complexes (13a′ and
13b′) are formed or in the following transition states (TS-14a′
and TS-14b′). The difference between the activation barriers
for the two reactions 13b′ → 15b′ via 14b′ and 13a′ → 15a′ via
14a′ (Δ(ΔErel) = 1.4 kcal/mol) is not large enough to
rationalize the very high cis-diastereoselectivity obtained in the
laboratory; the observed >99.5:0.5 selectivity corresponds to a
greater than 3.2 kcal/mol barrier difference at 298 K. (At the
Figure 7. Energy diagram for the cyclopropanation reaction between
Rh(I) carbenoid 11 and styrene. Relative energies in kcal/mol at
BP86/def2-SVP.
was found that the complexes leading to the cis cyclopropane
were lower in energy than the complexes leading to the trans
cyclopropane, also with the dispersion-corrected method.
Again, no transition state for the addition of styrene could be
identified. The barriers for the formation of the metal-
lacyclobutanes 15a′ and 15b′ from the styrene complexes 13a′
and 13b′ were slightly lower at D-BP86/def2-SVP (3.3 and 1.9
kcal/mol) than at BP86/def2-SVP (4.5 and 2.5 kcal/mol,
respectively). The barriers for the reductive elimination were
also predicted to be somewhat lower (11.2 and 10.8 kcal/mol at
BP86/def2-SVP and 10.6 and 9.7 at D-BP86/def2-SVP). As
Scheme 6. D-BP86/def2-SVP Structures of the Rh(I) Carbenoid (11′) and the Styrene-Coordinated Complexes (13a′ and
a
13b′)
a
Energies are relative to an assembly of isolated 2, styrene, and EDA.
6569
dx.doi.org/10.1021/om200594q | Organometallics 2011, 30, 6562−6571

Organometallics
Article
(4) Wong, H. N. C.; Hon, M.-Y.; Tse, C.-W.; Yip, Y.-C.; Tanko, J.;
Hudlicky, T. Chem. Rev. 1989, 89, 165−198.
MP2/def2-TZVPP//BP86/def2-SVP level, the computationally
predicted energy difference was even lower, at Δ(ΔErel) = 0.2
kcal/mol.) This suggests that the coordination of styrene to the
Rh(I) carbenoid 11′ might be the diastereoselectivity-
determining step. No transition states for the coordination of
styrene to Rh(I) carbenoid 11′ could be located. If this step is
indeed barrierless, the reactivity must be controlled by reaction
dynamics, e.g., different collision frequencies.79 Realistic
molecular dynamic simulations, however, lie far outside the
scope of this work, and hence the details of this step are not
clear at this point in time.
(5) Doyle, M. P. Angew. Chem., Int. Ed. 2009, 48, 850−852.
(6) Langlotz, B. K.; Wadepohl, H.; Gade, L. H. Angew. Chem., Int. Ed.
2008, 47, 4670−4674.
(7) Lebel, H.; Marcoux, J.-F.; Molinaro, C.; Charette, A. B. Chem.
Rev. 2003, 103, 977−1050.
́
(8) Caballero, A.; Prieto, A.; Díaz-Requejo, M. M.; Perez, P. J. Eur. J.
Inorg. Chem. 2009, 1137−1144.
(9) Suematsu, H.; Kanchiku, S.; Uchida, T.; Katsuki, T. J. Am. Chem.
Soc. 2008, 130, 10327−10337.
(10) Alexander, K.; Cook, S.; Gibson, C. L. Tetrahedron Lett. 2000,
41, 7135−7138.
(11) Uchida, T.; Irie, R.; Katsuki, T. Synlett 1999, 1793−1795.
(12) Uchida, T.; Irie, R.; Katsuki, T. Synlett 1999, 1163−1165.
(13) Uchida, T.; Irie, R.; Katsuki, T. Tetrahedron 2000, 56, 3501−
3509.
(14) Uchida, T.; Katsuki, T. Synthesis 2006, 1715−1723.
(15) Bachmann, S.; Mezzetti, A. Helv. Chim. Acta 2001, 84, 3063−
3074.
CONCLUSION
■
A computational study of the mechanism of a cyclopropanation
reaction between styrene and EDA catalyzed by Rh(I) catalyst
1 that bears a chelating imine-functionalized NHC ligand has
been presented. This is, to the best of our knowledge, the first
report of a computational study of a Rh(I) catalyzed
cyclopropanation reaction. Density functional theory at the
BP86/def2-SVP level and dispersion-corrected DFT (D-BP86/
def2-SVP) as well as ab initio theory (MP2/def2-TZVPP) were
used in the investigations. The rate-determining step of the
catalytic cycle was found to be extrusion of N2 in the formation
of a Rh(I) carbenoid, representing a barrier of 12.9 kcal/mol
(D-BP86/def2-SVP). The reaction was shown to proceed via a
stepwise mechanism, which involves the formation of Rh
metallacyclobutanes. The origin of the high cis-diastereoselec-
tivity probably lies in an unfavorable steric interaction between
the styrene phenyl ring and the methyl groups in the ortho
positions on the ligand N-aryl ring in the complexes leading to
the trans isomer. The results provide valuable information
about the origin of the high cis-diastereoselectivity, which might
prove useful in further developments of the Rh(I)-catalyzed
cyclopropanations. In this respect, it is interesting to note that
manipulation of the N-aryl ortho substituent of the
iminocarbene ligand strongly influences catalyst performance
and diastereoselectivities.28
(16) Bachmann, S.; Furler, M.; Mezzetti, A. Organometallics 2001, 20,
2102−2108.
(17) Bonaccorsi, C.; Mezzetti, A. Organometallics 2005, 24, 4953−
4960.
(18) Bonaccorsi, C.; Santoro, F.; Gischig, S.; Mezzetti, A.
Organometallics 2006, 25, 2002−2010.
(19) Bonaccorsi, C.; Bachmann, S.; Mezzetti, A. Tetrahedron:
Asymmetry 2003, 14, 845−854.
(20) Casey, C. P.; Polichnowski, S. W.; Shusterman, A. J.; Jones, C.
R. J. Am. Chem. Soc. 1979, 101, 7282−7292.
(21) Hoang, V. D. M.; Reddy, P. A. N.; Kim, T.-J. Tetrahedron Lett.
2007, 48, 8014−8017.
(22) Díaz-Requejo, M. M.; Caballero, A.; Belderrain, T. R.; Nicasio,
M. C.; Trofimenko, S.; Per
983.
́
ez, P. J. J. Am. Chem. Soc. 2002, 124, 978−
(23) Luning, U.; Fahrenkrug, F. Eur. J. Org. Chem. 2006, 916−923.
̈
(24) Kanchiku, S.; Suematsu, H.; Matsumoto, K.; Uchida, T.;
Katsuki, T. Angew. Chem., Int. Ed. 2007, 46, 3889−3891.
(25) Shitama, H.; Katsuki, T. Chem.Eur. J. 2007, 13, 4849−4858.
(26) Rosenberg, M. L.; Krivokapic, A.; Tilset, M. Org. Lett. 2009, 11,
547−550.
̌ ́
(27) Rosenberg, M. L.; Vlasana, K.; Gupta, N. S.; Wragg, D.; Tilset,
M. J. Org. Chem. 2011, 76, 2465−2470.
(28) Rosenberg, M. L.; Langseth, E.; Krivokapic, A.; Gupta, N. S.;
Tilset, M. New J. Chem. 2011, 35, 2306−2313.
(29) Doyle, M. P.; McKervey, M. A.; Ye, T. Modern Catalytic Methods
for Organic Synthesis with Diazo Compounds: From Cyclopropanes to
Ylides; Wiley: New York, 1998.
ASSOCIATED CONTENT
* Supporting Information
Optimized geometries, energies, and Cartesian coordinates for
all the calculated species. This material is available free of
charge via the Internet at http://pubs.acs.org.
■
S
(30) Dias, E. L.; Brookhart, M.; White, P. S. Chem. Commun. 2001,
AUTHOR INFORMATION
Corresponding Author
*E-mail: mats.tilset@kjemi.uio.no.
423−424.
■
(31) Teixidor, F.; Nunez, R.; Flores, M. A.; Demonceau, A.; Vinas, C.
́
̃
̃
J. Organomet. Chem. 2000, 614−615, 48−56.
̈
(32) Çetinkaya, B.; Ozdemir, I.; Dixneuf, P. H. J. Organomet. Chem.
1997, 534, 153−158.
ACKNOWLEDGMENTS
■
́ ́
(33) Carmona, A.; Corma, A.; Iglesias, M.; San Jose, A.; Sanchez, F. J.
The authors gratefully acknowledge financial support from the
Norwegian Research Council (NFR) through the Centre of
Excellence program (grant no. 179568/V30 to the Centre of
Theoretical and Computational Chemistry) and through the
KOSK program (grant no. 177325/V30, stipend to M.L.R.).
We also thank NOTUR for providing generous computational
resources and Dr. Ole Swang, SINTEF, for helpful discussions.
Organomet. Chem. 1995, 492, 11−21.
(34) Becke, A. D. Phys. Rev. A 1988, 38, 3098−3100.
(35) Perdew, J. P. Phys. Rev. B 1986, 33, 8822−8824.
̈
(36) Eichkorn, K.; Treutler, O.; Ohm, H.; Haser, M.; Ahlrichs, R.
̈
Chem. Phys. Lett. 1995, 242, 652−660.
(37) Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 7,
3297−3305.
(38) Weigend, F. Phys. Chem. Chem. Phys. 2006, 8, 1057−1065.
(39) Waller, M. P.; Braun, H.; Hojdis, N.; Buhl, M. J. Chem. Theory
̈
REFERENCES
Comput. 2007, 3, 2234−2242, , and references therein..
(40) Deglmann, P.; Furche, F.; Ahlrichs, R. Chem. Phys. Lett. 2002,
362, 511−518.
■
(1) Donaldson, W. A. Tetrahedron 2001, 57, 8589−8627.
(2) Salaun, J. Top. Curr. Chem. 2000, 207, 1−67.
̈
(41) Gonzalez, C.; Schlegel, H. B. J. Phys. Chem. 1990, 94, 5523−
5527.
(3) Rubin, M.; Rubina, M.; Gevorgyan, V. Chem. Rev. 2007, 107,
3117−3179.
6570
dx.doi.org/10.1021/om200594q | Organometallics 2011, 30, 6562−6571

Organometallics
Article
(42) Gonzalez, C.; Schlegel, H. B. J. Chem. Phys. 1989, 90, 2154−
2161.
(69) Pirrung, M. C.; Liu, H.; Morehead, A. T. Jr. J. Am. Chem. Soc.
2002, 124, 1014−1023.
(70) Hansen, J.; Autschbach, J.; Davies, H. M. L. J. Org. Chem. 2009,
74, 6555−6563.
(43) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.;
Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.;
Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.;
Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.;
Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J.
B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R.
E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.;
Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J.
J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.;
Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman,
J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, Rev. E.01;
Gaussian Inc.: Wallingford, CT, 2004.
(71) Fraile, J. M.; García, J. I.; Martínez-Merino, V.; Mayoral, J. A.;
Salvatella, L. J. Am. Chem. Soc. 2001, 123, 7616−7625.
(72) Suenobu, K.; Itagaki, M.; Nakamura, E. J. Am. Chem. Soc. 2004,
126, 7271−7280.
(73) Wang, F.; Meng, Q.; Li, M. Int. J. Quantum Chem. 2008, 108,
945−953.
(74) Berthon-Gelloz, G.; Marchant, M.; Straub, B. F.; Marko, I. E.
Chem.Eur. J. 2009, 15, 2923−2931.
(75) Bernardi, F.; Bottoni, A.; Miscione, G. P. Organometallics 2001,
20, 2751−2758.
(76) Hoffmann, H. M. R.; Otte, A. R.; Wilde, A.; Menzer, S.;
Williams, D. J. Angew. Chem., Int. Ed. Engl. 1995, 34, 100−102.
(77) Jennings, P. W.; Johnson, L. L. Chem. Rev. 1994, 94, 2241−
2290.
(78) Tjaden, E. B.; Stryker, J. M. J. Am. Chem. Soc. 1990, 112, 6420−
6422.
(79) Swiegers, G. F.; Huang, J.; Brimblecombe, R.; Chen, J.;
Dismukes, G. C.; Mueller-Westerhoff, U. T.; Spiccia, L.; Wallace, G. G.
Chem.Eur. J. 2009, 15, 4746−4759.
(44) TURBOMOLE, V6.2, a development of University of Karlsruhe
and Forschungszentrum Karlsruhe GmbH, 1989−2007, TURBO-
MOLE GmbH, since 2007; available from http://www.turbomole.
com.
(45) Klamt, A.; Schuurmann, G. J. Chem. Soc., Perkin Trans. 2 1993,
̈
̈
799−805.
(46) Petit, A.; Richard, P.; Cacelli, I.; Poli, R. Chem.Eur. J. 2006,
12, 813−823.
(47) Blake, A. J.; George, M. W.; Hall, M. B.; McMaster, J.; Portius,
́
P.; Sun, X. Z.; Towrie, M.; Webster, C. E.; Wilson, C.; Zaric, S. D.
Organometallics 2008, 27, 189−201.
(48) Grimme, S. J. Comput. Chem. 2004, 25, 1463−1473.
(49) Hellweg, A.; Hattig, C.; Hofener, S.; Klopper, W. Theor. Chem.
̈
̈
Acc. 2007, 117, 587−597.
(50) Weigend, F.; Haser, M.; Patzelt, H.; Ahlrichs, R. Chem. Phys.
̈
Lett. 1998, 294, 143−152.
(51) Weigend, F.; Haser, M. Theor. Chem. Acc. 1997, 97, 331−340.
̈
(52) Bonge, H. T.; Hansen, T. J. Org. Chem. 2010, 75, 2309−2320.
̈
(53) Ozen, C.; Tuzun, N. Ş. Organometallics 2008, 27, 4600−4610.
̈
̈
(54) Nowlan, D. T. III; Gregg, T. M.; Davies, H. M. L.; Singleton, D.
A. J. Am. Chem. Soc. 2003, 125, 15902−15911.
(55) Demonceau, A.; Noels, A. F.; Hubert, A. J. Aspects Homogeneous
Catal. 1988, 6, 199−232.
(56) Cornejo, A.; Fraile, J. M.; García, J. I.; Gil, M. J.; Martínez-
Merino, V.; Mayoral, J. A.; Salvatella, L. Organometallics 2005, 24,
3448−3457.
(57) Fraile, J. M.; García, J. I.; Gil, M. J.; Martínez-Merino, V.;
Mayoral, J. A.; Salvatella, L. Chem.Eur. J. 2004, 10, 758−765.
(58) Frøseth, M.; Dhindsa, A.; Røise, H.; Tilset, M. Dalton Trans.
2003, 4516−4524.
(59) Frøseth, M.; Netland, K. A.; Rømming, C.; Tilset, M. J.
Organomet. Chem. 2005, 690, 6125−6132.
(60) Yates, P. J. Am. Chem. Soc. 1952, 74, 5376−5381.
(61) Galardon, E.; Maux, P. L.; Toupet, L.; Simonneaux, G.
Organometallics 1998, 17, 565−569.
́
(62) Galardon, E.; Roue, S.; Le Maux, P.; Simonneaux, G.
Tetrahedron Lett. 1998, 39, 2333−2334.
(63) Park, S.-B.; Sakata, N.; Nishiyama, H. Chem.Eur. J. 1996, 2,
303−306.
(64) Straub, B. F.; Hofmann, P. Angew. Chem., Int. Ed. 2001, 40,
1288−1290.
(65) Ogasawara, M.; Macgregor, S. A.; Streib, W. E.; Folting, K.;
Eisenstein, O.; Caulton, K. G. J. Am. Chem. Soc. 1995, 117, 8869−
8870.
(66) Elian, M.; Hoffmann, R. Inorg. Chem. 1975, 14, 1058−1076.
(67) Nakamura, E.; Yoshikai, N.; Yamanaka, M. J. Am. Chem. Soc.
2002, 124, 7181−7192.
(68) Anciaux, A. J.; Hubert, A. J.; Noels, A. F.; Petiniot, N.; Teyssie,
́
P. J. Org. Chem. 1980, 45, 695−702.
6571
dx.doi.org/10.1021/om200594q | Organometallics 2011, 30, 6562−6571