Asymmetric hydrogenation catalyzed by a rhodium complex of (R)-(tert-butylmethylphosphino)(di-tert-butylphosphino)-methane: Scope of enantioselectivity and mechanistic study

Article abstract of DOI:10.1021/ja076542z

The rhodium complex of (R)-(tert-butylmethylphosphino)(di-tert- butylphosphino)methane used in Rh-catalyzed asymmetric hydrogenation of representative substrates 3-14 demonstrated high catalytic activity coupled with wide scope and nearly perfect enantioselectivity. Mechanistic studies (NMR and DFT computations) were carried out in order to investigate the mechanism of the enantioselection in the asymmetric hydrogenation of (Z)-α- acetamidocinnamate (3). Although catalyst-substrate complexes 15a,b with the double bond coordinated near the non-"chiral" phosphorus atom were formed as kinetic products upon the addition of 3 to solvate complex 2 at -100°C, they rapidly rearranged to more stable isomers 15c,d with the double bond coordinated near the "chiral" phosphorus atom. The thermodynamic and kinetic parameters of the interconversion between 15c and 15d were determined by NMR; mainly, the interconversion occurred intramolecularly via nonchelating catalyst-substrate complexes 16. The equilibrium between 15d and 16d was directly observed from NMR line shape changes at temperatures ranging from -100 to -40°C, whereas no such equilibrium was observed for 15c. This result was accounted for computationally by determining the corresponding transition states for the methanol insertion into 15c,d. Three sets of experiments of the low-temperature hydrogenation of different catalyst-substrate complexes gave the same order and sense of enantioselectivity (97% ee (R)) even in the case when 15c, having Re-coordinated double bond, was hydrogenated under the conditions precluding its isomerization to 15d. It was concluded that the hydrogenation of 15c,d does not occur directly, but is preceded by the dissociation of the double bond to result in the more reactive species 16. This indicates that enantioselection must occur at a later step of the catalytic cycle. DFT computations of association and migratory insertion steps suggest that enantioselection takes place during the association step when chelating dihydride 19d-MeOH is formed from nonchelating dihydride 18d.

Full text of DOI:10.1021/ja076542z

Published on Web 02/01/2008
Asymmetric Hydrogenation Catalyzed by a Rhodium Complex
of (R)-(tert-Butylmethylphosphino)(di-tert-butylphosphino)-
methane: Scope of Enantioselectivity and Mechanistic Study
Ilya D. Gridnev,*,†,‡ Tsuneo Imamoto,*,§ Garrett Hoge,*,| Mitsuhiro Kouchi,‡ and
Hidetoshi Takahashi§
Graduate School of Science and Engineering, Tokyo Institute of Technology, 2-12-1 Ookayama,
Meguro-ku, Tokyo 152-8552, Japan, Department of Chemistry, Graduate School of Science,
Tohoku UniVersity, Sendai 980-8578, Japan, Department of Chemistry, Graduate School of
Science, Chiba UniVersity, Chiba 263-8522, Japan, and High Pressure Laboratory, Chemical
R&D, Pfizer Global R&D, 2800 Plymouth Road, Ann Arbor, Michigan 48105
Received August 30, 2007; E-mail: gridnev.i.aa@m.titech.ac.jp
Abstract: The rhodium complex of (R)-(tert-butylmethylphosphino)(di-tert-butylphosphino)methane used
in Rh-catalyzed asymmetric hydrogenation of representative substrates 3-14 demonstrated high catalytic
activity coupled with wide scope and nearly perfect enantioselectivity. Mechanistic studies (NMR and DFT
computations) were carried out in order to investigate the mechanism of the enantioselection in the
asymmetric hydrogenation of (Z)-R-acetamidocinnamate (3). Although catalyst-substrate complexes 15a,b
with the double bond coordinated near the non-“chiral” phosphorus atom were formed as kinetic products
upon the addition of 3 to solvate complex 2 at -100 °C, they rapidly rearranged to more stable isomers
15c,d with the double bond coordinated near the “chiral” phosphorus atom. The thermodynamic and kinetic
parameters of the interconversion between 15c and 15d were determined by NMR; mainly, the
interconversion occurred intramolecularly via nonchelating catalyst-substrate complexes 16. The equilibrium
between 15d and 16d was directly observed from NMR line shape changes at temperatures ranging from
-100 to -40 °C, whereas no such equilibrium was observed for 15c. This result was accounted for
computationally by determining the corresponding transition states for the methanol insertion into 15c,d.
Three sets of experiments of the low-temperature hydrogenation of different catalyst-substrate complexes
gave the same order and sense of enantioselectivity (97% ee (R)) even in the case when 15c, having
Re-coordinated double bond, was hydrogenated under the conditions precluding its isomerization to 15d.
It was concluded that the hydrogenation of 15c,d does not occur directly, but is preceded by the dissociation
of the double bond to result in the more reactive species 16. This indicates that enantioselection must
occur at a later step of the catalytic cycle. DFT computations of association and migratory insertion steps
suggest that enantioselection takes place during the association step when chelating dihydride 19d‚MeOH
is formed from nonchelating dihydride 18d.
1. Introduction
abundance of experimental data. Much effort has been dedicated
to intercepting and characterizing reaction intermediates during
the process of asymmetric hydrogenation. This concentration
on asymmetric hydrogenation is unrivaled by mechanistic
investigations of other catalytic processes.13-15 Moreover, large-
scale ab initio computations that describe detailed reaction
pathways of the catalytic cycle are now available.16-20
The mechanism of Rh-catalyzed asymmetric hydrogenation
has been in the focus of active academic research in the past
three decades.1-12 This mechanistic investigation is fueled both
by the practical importance of the reaction itself and by the
† Tokyo Institute of Technology.
‡ Tohoku University.
§ Chiba University.
(8) Kagan, H. B.; Langois, N.; Dang, T. P. J. Organomet. Chem. 1975, 90,
353.
(9) Schrock, R. R.; Osborn, J. A. J. Am. Chem. Soc. 1976, 98, 134.
(10) Halpern, J.; Riley, D. P.; Chan, A. S. C.; Pluth, J. J. J. Am. Chem. Soc.
1977, 99, 8055.
(11) Brown, J. M.; Chaloner, P. A. J. Chem. Soc., Chem. Commun. 1978, 321.
(12) Chan, A. S. C; Halpern, J. J. Am. Chem. Soc. 1980, 102, 838.
(13) Gridnev, I. D.; Higashi, N.; Asakura, K.; Imamoto, T. J. Am. Chem. Soc.
2000, 122, 7183.
(14) Giernoth, R.; Heinrich, H.; Adams, N.; Deeth, R. J.; Bargon, J.; Brown, J.
M. J. Am. Chem. Soc. 2000, 122, 12381.
| Pfizer Global R&D.
(1) Knowles, W. S. Asymmetric Hydrogenations (Nobel Lecture). Angew.
Chem., Int. Ed. 2002, 41, 1998.
(2) Noyori, R. Asymmetric Catalysis: Science and Opportunities (Nobel
Lecture). Angew. Chem., Int. Ed. 2002, 41, 2008.
(3) Noyori, R. Asymmetric Catalysis in Organic Synthesis; John Wiley &
Sons: New York, 1994; pp 16-94.
(4) Ohkuma T.; Kitamura M.; Noyori R. In Catalytic Asymmetric Synthesis;
Ojima, I., Ed.; Wiley-VCH: Weinheim, 2000; pp 1-110.
(5) Brown, J. M. In ComprehensiVe Asymmetric Catalysis; Jacobsen, E. N.,
Pfaltz, A., Yamamoto, H., Eds.; Springer: Berlin, 1999; Vol. 1, pp 119-
182.
(15) Imamoto, T.; Yashio, K.; Cre´py, K. V. L.; Katagiri, K.; Takahashi, H.;
Kouchi, M.; Gridnev, I. D. Organometallics 2006, 25, 908.
(16) (a) Landis, C. R.; Feldgus, S. Angew. Chem., Int. Ed. 2000, 39, 2863. (b)
Feldgus, S.; Landis, C. R. J. Am. Chem. Soc. 2000, 122, 12714.
(6) Crepy, K. V. L.; Imamoto, T. AdV. Synth. Catal. 2003, 345, 79.
(7) Gridnev, I. D.; Imamoto, T. Acc. Chem. Res. 2004, 37, 633.
9
2560
J. AM. CHEM. SOC. 2008, 130, 2560-2572
10.1021/ja076542z CCC: $40.75 © 2008 American Chemical Society

Rh-Catalyzed Asymmetric Hydrogenation
A R T I C L E S
Scheme 1. Mechanism of Asymmetric Hydrogenation Implying the
Chelate Coordination of the Substrate Prior to the Oxidative
Addition of Dihydrogen
those obtained in the catalytic reactions can be obtained by the
low-temperature reactions of solvate dihydrides derived from
the Rh complexes of electron-rich diphosphines with representa-
tive substrates.27 These observations opened the doors for the
possibility of the hydrogen activation prior to the substrate
coordination in these reactions.7 Hence, according to the recent
point of view, the mechanism of enantioselection can be
depending on the exact combination of the ligand and the
substrate.
The more reactive diastereomer of the catalyst-substrate
complex in early mechanistic experiments was less abundant
in the equilibrium mixture (MINOR in Scheme 1). That
experimental observation positioned the enantioselective Rh-
catalyzed hydrogenation apart from enzymatic reactions that are
believed to apply structural “Lock-and-Key” recognition be-
tween the substrate and the enzyme.28 However, in view of the
recent research, it became clear that there is no evident cause-
and-effect relationship between the selectivity of substrate
binding by the catalyst and the relative reactivity of the resulting
catalyst-substrate complexes toward hydrogen. Landis et al.
showed computationally how the definite asymmetric environ-
ment of the Rh atom can result simultaneously in decreasing
stability and increasing reactivity toward dihydrogen in the
catalyst-substrate complex of DuPHOS-Rh with R-formami-
doacrylonitrile.16 So far, however, there is no reason to believe
that this type of structure-reactivity correlation should be the
same for all other catalyst-substrate combinations. Evans et
al. reported a P/S ligand whose Rh chelate produced a single
isomer of the catalyst-substrate complex with the double bond
coordinated accordingly to the observed sense of induction.29
Similar observations were later reported by Heller et al. for the
DIPAMP complexes of â-acylaminoacrylates.30 Hence, the real
understanding of the correlation between the structure of a
catalyst-substrate complex and its reactivity toward hydrogen
has not been achieved so far.
The most important conclusion that can be derived from these
extensive mechanistic studies is a clear understanding of the
intrinsic mechanism of enantioselection. This knowledge is
important for design of chiral catalysts and the optimization of
reaction conditions. It has been widely accepted initially that
the stereoselection in Rh-catalyzed asymmetric hydrogenation
takes place via the difference in reactivity of diastereomeric
square planar catalyst-substrate complexes toward dihydrogen.
This conclusion originated from the early experiments of
Brown21 and Halpern,22-24 in which the structures of the
relatively reactive diastereomers of the catalyst-substrate
complexes have been in accord with the sense of enantioselec-
tion observed in these reactions (Scheme 1). Landis and Halpern
have studied the kinetics of measurable reaction steps and
equilibria for the case of asymmetric hydrogenation of methyl
(Z)-R-acetamidocinnamate catalyzed by Rh-DIPAMP, and
showed that they are in qualitative agreement with the expected
patterns.25
However, further research has demonstrated a considerable
diversification of the mechanistic details. Thus, a very late
agostic intermediate has been shown to be involved in a
reversible equilibrium with the catalyst in case of asymmetric
hydrogenation with the Rh-PHANEPHOS complex.26 More-
over, it has been shown that enantioselectivities comparable to
In this work, we investigated in detail the mechanism of
asymmetric hydrogenation catalyzed by a Rh-complex of (R)-
(tert-butylmethylphosphino)(di-tert-butylphosphino)methane
(27) (a) Gridnev, I. D.; Higashi, N.; Asakura, K.; Imamoto, T. J. Am. Chem.
Soc. 2000, 122, 7183. (b) Gridnev, I. D.; Higashi, N.; Imamoto, T. J. Am.
Chem. Soc. 2000, 122, 10486. (c) Gridnev, I. D.; Yasutake, M.; Higashi,
N.; Imamoto, T. J. Am. Chem. Soc. 2001, 123, 5268. (d) Gridnev, I. D.;
Higashi, N.; Imamoto, T. J. Am. Chem. Soc. 2001, 123, 4631. (e) Gridnev,
I. D.; Yamanoi, Y.; Higashi, N.; Tsuruta, H.; Yasutake, M.; Imamoto, T.
AdV. Synth. Catal. 2001, 343, 118. (f) Gridnev, I. D.; Higashi, N.; Imamoto,
T. Organometallics 2001, 20, 4542. (g) Yasutake, M.; Gridnev, I. D.;
Higashi, N.; Imamoto, T. Org. Lett. 2001, 3, 1701. (h) Gridnev, I. D.;
Yasutake, M.; Imamoto, T.; Beletskaya, I. P. Proc. Natl. Acad. Sci. U.S.A.
2004, 101, 5385.
(28) Fischer, E. Ber. Dtsch. Chem. Ges. 1894, 27, 2985.
(29) Evans, D. A.; Michael, F. E.; Tedrow, J. S.; Campos, K. R. J. Am. Chem.
Soc. 2003, 125, 3534.
(30) (a) Drexler, H.-J.; Baumann, W.; Schmidt, T.; Zhang, S.; Sun, A.;
Spannenberg, A.; Fischer, C.; Buschmann, H.; Heller, D. Angew. Chem.,
Int. Ed. 2005, 44, 1184. (b) Schmidt, T.; Baumann, W.; Drexler, H.-J.;
Arrieta, A.; Heller, D.; Buschmann, H. Organometallics 2005, 24, 3842.
(17) Feldgus, S.; Landis, C. R. Organometallics 2001, 20, 2374.
(18) Li, M.; Tang, D.; Luo, X.; Shen, W. Int. J. Quant. Chem. 2005, 102, 53.
(19) Mori, S.; Vreven, T.; Morokuma, K. Chem. Asian J. 2006, 1, 391.
(20) Donoghue, P. J.; Helquist, P.; Wiest, O. J. Org. Chem. 2007, 72, 839.
(21) Brown, J. M.; Chaloner, P. A. J. Chem. Soc., Chem. Commun. 1980, 344.
(22) Chua, P. S.; Roberts, N. K.; Bosnich, B.; Okrainski, S. J.; Halpern, J. J.
Chem. Soc., Chem. Commun. 1981, 1278.
(23) Halpern, J. Science 1982, 217, 401.
(24) Halpern, J. Asymmetric Catalytic Hydrogenation: Mechanism and Origin
of Enantioselection in Asymmetric Synthesis; Morrison, J. D., Ed.; Academic
Press: Orlando, 1985; Vol. 5, pp 41-69.
(25) Landis, C. R.; Halpern, J. J. Am. Chem. Soc. 1987, 109, 1746.
(26) Heinrich, H.; Giernoth, R.; Bargon, J.; Brown, J. M. Chem. Commun. 2001,
1296.
9
J. AM. CHEM. SOC. VOL. 130, NO. 8, 2008 2561

A R T I C L E S
Gridnev et al.
Table 1. Asymmetric Hydrogenation of Prochiral Substrates
Catalyzed by Rh Complex 1
Figure 1. Evolution of the 31P spectrum of the sample obtaned by adding
1.5 equiv of MAC 3 to a sample containing the solution of the solvate
complex 2 precooled to -100 °C: a) immediately after addition; b) after
30 min at -100 °C; c) after 1 h at -100 °C; d) the temperature raised to
-80 °C and the spectrum taken immediately (the still broadened left
multiplet of 15b is under the much more intensive signal of 15a; signals of
15d are broad due to the fast equilibrium with 16d, see Figure 5 and further
discussion); e) the temperature raised to -40 °C and the spectrum taken
immediately.
pounds, such as phosphonate 9, â-dehydroamino acids 10 and
11, enamides 12 and 13, and itaconic ester 14, can be also
hydrogenated using 1 as a catalyst precursor with very high
ee’s. The very similar performance of (E)- and (Z)-â-dehy-
droamino acids enables use of isomeric mixtures as substrates.
The opposite sense of enantioselection for phenyl (12)- and tert-
butyl (13)-substituted enamides is in accord with previous
observations made for other Rh complexes of chiral diphos-
phines as catalysts.
2.2. Catalytic Cycle in the Asymmetric Hydrogenation of
Methyl (Z)-r-Acetamidocinnamate. 2.2.1. Reaction of Solvate
Complex 2 with Methyl (Z)-r-Acetamidocinnamate (3).
Hydrogenation of catalytic precursor 1 at ambient temperature
for 1.5 h yielded the corresponding solvate complex 2. We
investigated the formation of catalyst-substrate complexes by
mixing solutions of 2 and MAC (3, 1.5-fold excess) at -100
°C and putting the resulting reaction mixture in the probe of an
NMR spectrometer precooled to the same temperature. The
changes observed in the 31P NMR spectrum of the thus-prepared
sample are shown in Figure 1. First, two kinetic products are
formed (15a,b, Scheme 2). The signals of 15a,b in the 31P NMR
spectrum at -100 °C are notably broader than the signals of 2
at the same temperature (Figure 1a). At -80 °C, the signals of
15a,b became sharp, enabling the integration; the ratio 15a:
15b was 5:1 (Figure 1d). These line shape changes indicate that
both 15a and 15b are involved in some equilibria that are fast
at -80 °C and become slow in the NMR time scale at -100
°C. Unfortunately, the irreversible rearrangement of 15a,b to
(1).31,32 Using the results of extensive NMR investigations and
computational analysis of real systems, we present an explana-
tion for the different reactivities of diastereomeric catalyst-
substrate complexes as well as a new approach to rationalize
the enantioselection in Rh-catalyzed asymmetric hydrogenation.
2. Results and Discussion
2.1. Scope of Asymmetric Hydrogenation Catalyzed by
Rh Complex 1. The scope of asymmetric hydrogenation
catalyzed by complex 1 is shown in Table 1.
Analyzing the data of Table 1, one can conclude that Rh
complex 1 is a versatile catalyst: it gives almost perfect
stereoselectivity when hydrogenating dehydroamino acids 4 and
6, and their esters 3 and 5, including â,â-disubstituted com-
pounds 7 and 8 which are usually very difficult substrates for
asymmetric hydrogenation. Other classes of prochiral com-
(31) Hoge, G.; Wu, H.-P.; Kissel, W. S.; Pflum, D. A.; Greene, D. J.; Bao, J.
J. Am. Chem. Soc. 2004, 126, 5966.
(32) Hoge, G.; Wu, H.-P. Org. Lett. 2004, 6, 3645.
9
2562 J. AM. CHEM. SOC. VOL. 130, NO. 8, 2008

Rh-Catalyzed Asymmetric Hydrogenation
A R T I C L E S
Scheme 2. Reaction of Solvate Complex 2 with MAC 3. Red Numbers Indicate the Relative Energies Computed on the B3LYP/SDD Level
of Theory
15b,c (vide infra) occurred rapidly at temperatures above -100
°C, and hence, we were unable to analyze their spectra
separately. In the 13C NMR spectrum acquired at -100 °C, we
detected a broad signal at δ ) 78.0 belonging to the â-carbon
of the double bond in 15a,b (controlled by the experiment with
â-labeled 3), which indicates that the double bond is coordinated
to rhodium in 15a,b. Moreover, whereas the signal of the
carbonyl carbon from the carboxyl group was only slightly
shifted from the corresponding signal of 3 (∆δ ) 2.1), the amido
carbonyl signal demonstrated a significant downfield shift (∆δ
) 13.9). These data, are sufficient to conclude that kinetic
products 15a,b have structures of chelate catalyst-substrate
complexes. Further experimental and computational results (vide
supra) enabled us to conclude that the double bond in 15a,b is
coordinated to the rhodium from the side of the non-“chiral”
phosphorus atom.
°C gave the following parameters of equilibrium between 15c
and 15d: ∆H ) 0.72 kcal mol-1 and ∆S ) 2.0 cal mol K-1
.
At temperatures above 20 °C, dynamic effects caused by the
interconversion of 15c and 15d were clearly seen in the NMR
spectra of the equilibrium mixture (Figure 2). From Figure 2,
it is also clear that the intramolecular exchange between 15c
and 15d is significantly faster than the exchange pathway
involving complete dissociation. This implies that partially
dissociated species with coordinated carbonyl and dissociated
double bond are involved in the equilibrium. The activation
parameters of the interconversion (Scheme 2) were determined
1
from a series of phase-sensitive H-1H EXSY spectra.
1
The H NMR spectrum of the equilibrium mixture of 15c
and 15d in deuteriomethanol was completely resolved at 300
MHz, enabling assignment of the signals via routine 2D
correlation techniques. The chemical shifts of the double bond
carbons and of the amido carbonyl carbon confirmed that both
compounds are chelating catalyst-substrate complexes. From
31P-1H correlation spectrum, the signals of t-Bu groups and
methyl were assigned to each particular isomer. Moreover, the
t-Bu groups for each isomer could be clearly distinguished as
those belonging to the phosphorus atoms with either two t-Bu
groups or with t-Bu and Me group (“chiral” phosphorus).
Furthermore, in the phase-sensitive NOESY spectrum of the
equilibrium mixture, positive NOEs were observed between d
C-H proton and Me as well the t-Bu group from the same
Initially formed kinetic products 15a,b rearranged slowly at
-100 °C, yielding two further products 15c and 15d. Initially,
15d slightly predominated in the reaction mixture: the ratio
15c:15d was in the range of 1:1 to 1:1.4 in different experiments.
In each experiment, the ratio did not change significantly if the
sample temperature was kept below -40 °C (Figure 1); above
this temperature, the equilibrium ratio 15c:15d was established.
The equilibrium concentrations changed slightly from 1:0.85
at 25 °C to 1:0.70 at -40 °C; linearization of the ln K versus
inverse temperature in the temperature range of -40 °C to 25
9
J. AM. CHEM. SOC. VOL. 130, NO. 8, 2008 2563

A R T I C L E S
Gridnev et al.
(Scheme 1). The coordination mode of the double bond is very
similar in all the four isomers: R-carbon atom lies almost perfect-
ly in the P-Rh-P plane, whereas â-carbon atom is invariably
out of the plane. Furthermore, in all optimized structures, the
double bond is not perpendicular to the P-Rh-P plane but is
bent with its CHPh end toward the chelate cycle (the deviation
from perpendicular orientation is close to 20° in all the four
isomers). This bending of the double bond well explains the
significant difference in energies between 15a and 15b. Indeed,
the bending of the double bond results in close contact between
the phenyl ring and the t-Bu group in 15b, but is avoided in
15a since the phenyl group is located near the relatively small
methyl group. Nevertheless, 15a is still almost 3 kcal/mol less
stable than either 15c or 15d, demonstrating the general
disadvantages of the double bond coordination near the non-
“chiral” phosphorus atom having two t-Bu substituents. This is
a result of inevitable close contacts of the t-Bu groups even
with the non-substituted side of the double bond (the distance
between dCH proton and t-Bu group is less than 2 Å in 15a).
2.2.3. Low-Temperature Equilibrium of Catalyst-Sub-
strate Complex 15d. In the temperature range between 173
and 243 K, the NMR spectra of 15d exhibit characteristic line
shape dependence, indicating rapid equilibrium of two species
(the ratio is approximately 4:1 at 173 K). On the other hand,
the spectra of 15c remain unchanged in the same temperature
range (Figure 5). The equilibrium is quite fast; qualitative
estimation of line shape changes suggests that the activation
barrier is approximately 6 kcal/mol. We conclude that this
species is partially dissociated complex 16d (Scheme 3); the
existence of rapid equilibrium for such a species at ambient
temperature is evident from the fast intramolecular exchange
between 15c and 15d (vide supra). Since 15c can interconvert
with 15d at temperatures above -40 °C (equilibrium concentra-
tions of 15c and 15d are achieved after approximately 1 h at
-30 °C), 15c must also be capable of undergoing ligand
exchange reaction with methanol to yield partially dissociated
species 16d (or its easily available rotamer, e.g., 16c, see Scheme
3), so that the interconversion between 15c and 15d could
proceed while keeping the carbonyl group coordinated. How-
Figure 2. Section plot of phase sensitive 2D 1H-1H EXSY spectrum (300
MHz, CD3OD, 293K) of the equilibrium mixture containing 15c, 15d (in
fast equilibrium with 16d, vide infra) and free substrate 3 (concentration
of 3 was approximately equal to those of 15c and 15d). The exchange cross-
peaks between the olefinic protons of 15c and 15d and between the ortho-
aromatic protons of 15c and 15d are much more intensive than the exchange
cross-peaks of the olefinic protons of 15c and 15d with the olefinic proton
of 3. Hence, mainly the exchange between 15c and 15d occurs intramo-
lecularly.
phosphorus atom of 15a and only between dC-H proton and
the t-Bu group belonging to the chiral phosphorus atom of 15b.
These data are sufficient for assigning the structures of 15a,b,
as shown in the Figure 3; the NOEs observed for the CO2Me
groups give additional support for the assignment.
2.2.2. Computational Study of Complexes 15a-d. We have
optimized the geometries of four possible diastereomers 15a-d
on the B3LYP/SDD level of theory (Figure 4). The results are
in good accordance with the experimental data. The two most
stable isomers are 15c and 15d with almost equal stabilities;
15a and 15b are 6.0 and 2.8 kcal/mol less stable, respectively
Figure 3. Most important NOE’s (red arches) for the structural assignment of 15c and 15d. The geometries were optimized on the B3LYP/SDD level of
theory. The numbers show the distance in angstrom to the nearest protons in Me and t-Bu groups in the optimized structures.
9
2564 J. AM. CHEM. SOC. VOL. 130, NO. 8, 2008

Rh-Catalyzed Asymmetric Hydrogenation
A R T I C L E S
Scheme 3. Intramolecular Interconversion of 15c‚MeOH and
15d‚MeOH
Figure 4. Optimized geometries of the catalyst-substrate complexes 15a-d
presence of numerous readily available methanol molecules.
Thus, our computations showed that attaching the first, second,
and third methanol molecule to 15d is exothermic at ap-
proximately 10 kcal mol-1 each. Hence, one could not hope to
obtain realistic values of the activation barriers for any steps
with solvent participation without having to involve numerous
methanol molecules in computations. However, increasing the
number of computed methanol molecules attached to the Rh
complex leads to a dramatic increase in the number of possible
isomers that differ in position and binding mode of the methanol
molecules. This and other complications prevented previous
workers from considering solvent participation in theoretical
studies of asymmetric hydrogenation.16-20 Nevertheless, these
effects must be the same for different isomers of the Rh
complexes. Hence, by comparing the relative stabilities of the
computed transition states, one can elucidate the reasons for
the relative reactivities of the corresponding intermediates.
We have located the structures of 15d‚MeOH and 16d
connected through transition state TS1, (Figure 6). The two
intermediates with chelate coordination of the substrate 15d‚
MeOH and with only carbonyl group coordinated 16d have very
similar energies that is in accord with the experimental observa-
tion of the equilibrium between 15d and 16d. The computed
value of the activation energy (11.1 kcal mol-1) is approximately
twice as high as the experimentally observed barrier. This
deviation is probably a result of using only one methanol
molecule in computations.
Insertion of the methanol molecule occurs from the side of
the nonhindered quadrant. This transformation is facilitated by
the hydrogen bonding to the polar methoxycarbonyl group that
is maintained throughout the whole process (see structure of
TS1 in Figure 7). Use of the same reaction pathway for the
insertion of a methanol molecule into 15c‚MeOH is impossible
due to the large hindrance from the t-Bu group. In order to
realize the insertion, the methanol molecule must approach
sideways to minimize interference with the hindered quadrant
and to maintain contact with the methoxycarbonyl group (see
structure of TS2 in Figure 7). As a result, the activation barrier
of the methanol insertion into 15c‚MeOH is more than 2-fold
higher than that into 15d‚MeOH (Figure 6).
(B3LYP/SDD). Hydrogens are removed for clarity.
Figure 5. Temperature dependent 31P NMR spectrum of the sample
containing nonequilibrium concentrations of 15c and 15d (ratio 15c : 15d
was approximately 1 : 1 in this sample): a) at -100 °C; b) at -90 °C; c)
at -80 °C; d) at -60 °C; e) at -100 °C (recooled).
ever, the activation barrier of this ligand exchange reaction in
the case of 15c is apparently much higher than that of 15d.
2.2.4. Computational Study of Insertion of Methanol
Molecule into 15c,d. We attempted to support the above
conclusion computationally. Apparently, when the double bond
dissociates, the resulting free coordination space is immediately
occupied by a methanol molecule. Hence, computations were
based on the structures of 15c,d having one methanol molecule
bonded via the hydrogen bond. We also realized that in the real
system, the situation is much more complicated due to the
9
J. AM. CHEM. SOC. VOL. 130, NO. 8, 2008 2565

A R T I C L E S
Gridnev et al.
Figure 6. Schematic energy profile for the interconversion of solvated catalyst-substrate complexes 15c,d through partially dissociated species 16c,d
computed on the B3LYP/SDD level of theory.
Scheme 4. Equilibration of Partially Dissociated
Catalyst-substrate Complexes
Figure 7. Structures of TS1 and TS2 optimized on the B3LYP/SDD level
of theory.
Apparently, 15a,b can undergo methanol insertion via transi-
tion states resembling TS2. The low activation barriers for these
reactions (line shape changes in the 31P NMR spectrum suggest
that these equilibria are still faster than the case of 15d) can be
explained by the relative instability of these complexes.
Furthermore, the experimental data indicate the activation
barriers of the rearrangements of 15a,b to 15c,d to be low, since
the rearrangements proceed at a reasonable rate even at -100
°C. However, 15a,b evidently are not equilibrating with 2 in
the temperature range of -100 to -80 °C (see Figure 1). Hence,
the relatively rapid equilibrium of partially dissociated species
16a,b and 16c,d must be possible. That, in turn, requires
reorientation of the carbonyl group without complete dissocia-
tion of the substrate (Scheme 4).
Interesting information presents also selective formation of
the kinetic products 15a,b in the low-temperature experiments.
This fact demonstrates that the coordination of 3 is stepwise
and begins from the coordination of the carbonyl group which
9
2566 J. AM. CHEM. SOC. VOL. 130, NO. 8, 2008

Rh-Catalyzed Asymmetric Hydrogenation
A R T I C L E S
Scheme 5. Selective Hydrogenation of Different Catalyst-substrate Complexes
chooses initially the least hindered side of the molecule. Hence,
in the catalytic conditions when the catalyst 2 is released from
the catalytic cycle to bind the next molecule of the substrate,
the first intermediates are 16a,b; and if the hydrogenation of
these semi-dissociated species is fast enough it might (at least
partially) occur without prior chelating coordination.
In the third experiment, the hydrogenation was continued from
the point where it was stopped in the previous experiment.
Preparing the sample with the 31P NMR spectrum as in Figure
8c, we continued hydrogenation at -78 °C while monitoring
the progress of the reaction with 31P NMR. As can be seen from
Figure 8, 15c was completely hydrogenated after 5 h at this
temperature. The reaction product (conversion 100%) recovered
from this experiment had again 97% ee (R).
2.2.5. Low-Temperature Hydrogenation of Catalyst-
Substrate Complexes. Having established the details of the
equilibrium between various catalyst-substrate complexes
formed from 2 and 3, we investigated several regimes of low-
temperature hydrogenation to gain an insight into the relative
reactivity of isomeric catalyst-substrate complexes toward
hydrogen. Three experiments were designed (Scheme 5).
In the first experiment, we added an equivalent amount of 3
to a solution of 2 at -100 °C and immediately hydrogenated
the reaction mixture for 30 min while keeping the temperature
at -100 °C. According to Figure 1, under these conditions, the
reaction mixture contained only 15a,b together with nonreacted
2 and 3. After hydrogenation, hydrogen gas was replaced with
argon to eliminate the possibility of any hydrogenation at higher
temperatures. Then, the sample was quenched and the reaction
product was analyzed. NMR analysis showed that 40% of 3
was hydrogenated; the ee of the product was 97% (R).
In the second experiment, an equilibrium mixture of 15c and
15d (equilibrated with 16d) was prepared at 25 °C. Then, it
was cooled to -78 °C and hydrogenated at this temperature
with regular control by monitoring the 31P NMR spectrum
(Figure 8). After hydrogenation for 30 min, the equilibrium
mixture of 15d and 16d was selectively hydrogenated, whereas
15c remained unreacted (Figure 8c). Quenching the mixture at
this stage yielded a reaction product (conversion was again 40%)
with 97% ee (R).
That the same optical purity of the hydrogenation product
was obtained in all three experiments is quite remarkable.
Especially meaningful is the last experiment: since 15c has an
Re-coordinated double bond, it would be expected to produce
an S-product if it were hydrogenated directly. However, this is
clearly not the case. Moreover, the rate of equilibrium between
15c and 15d (vide supra) implies that at -78 °C, the temperature
at which both were hydrogenated, no appreciable interconversion
occurs. To check this point more carefully, we again prepared
a sample with 15d being selectively hydrogenated (as in Figure
8c, see Figure 9) and kept it for 6 h at -70 °C. As can be seen
from Figure 9, no traces of 15d could be found in the 31P NMR
spectrum. On the other hand, as can be expected from the
already known rate of interconversion, the formation of 15d
from 15c occurred quite rapidly at -30 °C.
Thus, we conclude that the formation of R-hydrogenation
product when 15c was hydrogenated cannot be explained by
its prior isomerization to 15d, since this isomerization does not
occur under the conditions of the experiment. Nevertheless, the
hydrogenation of 15c must be preceded by a process eliminating
the stereodiscriminating coordination of the prochiral double
bond in 15c to allow formation of the R-product. Apparently,
the hydrogenation of 15c occurs through the intermediacy of
nonchelating catalyst-substrate complex 16c (Scheme 6). It is
9
J. AM. CHEM. SOC. VOL. 130, NO. 8, 2008 2567

A R T I C L E S
Gridnev et al.
Figure 8. Evolution of the 31P NMR (62 MHz, CD3OD, -90 °C) of the
sample originally containing the equilibrium mixture of 15c and 15d: a)
H2 admitted to the sample instead of Ar at -78 °C, then the sample was
cooled to -100 °C and placed into the probe precooled to -100 °C, the
spectrum taken at -90 °C; b) after 10 min hydrogenation at -78 °C
(handling of the sample is same further on); c) after additional 20 min
hydrogenation at -78 °C; 15d is completely removed; d) after additional
1 h hydrogenation at -78 °C; e) after additional 1.5 h hydrogenation at
-78 °C; f) after additional 1 h hydrogenation at -78 °C; g) after additional
1 h hydrogenation at -78 °C, the total hydrogenation time was 5 h, the
product obtained from this sample was of 97% ee) (R).
Figure 9. Section plots of the 31P NMR spectra of the sample initially
containing the equilibrium mixture of 15c and 15d: a) starting spectrum at
-90 °C; b) after 30 min hydrogenation at -78 °C, spectrum taken at -90
°C; c) sample was kept for 5 h at -70 °C and the spectrum was taken at
this temperature; d) immediately after raising the temperature to -30 °C;
e) after 6 min at -30 °C; f) after 30 min at -30 °C.
quite reasonable to suggest that 16c would be more reactive
toward dihydrogen than much more crowded 15c. It has been
shown that the steric hindrance plays an important role in the
hydrogenation step of the catalytic cycle in Rh-catalyzed
asymmetric hydrogenation.16
Scheme 6. Hydrogenation of 15c
Moreover, quite remarkable is the straightforward correlation
between the ease of dissociation of the double bond and the
reactivity of the corresponding catalyst-substrate complex.
Thus, 15d is in rapid equilibrium with approximately 20% of
16d, and the corresponding equilibria of 15a,b are even faster,
indicating that the partially dissociated species generated from
these catalyst-substrate complexes are readily available, even
at -100 °C, and their hydrogenation can be rapidly ac-
complished at this temperature. Unfortunately, the equilibrium
is so fast that it was beyond our experimental techniques to
check if 16d can be selectively hydrogenated before its
equilibrium with 15d is re-established at -100 °C. Nevertheless,
the relative ease of hydrogenation of 15a-d is evidently
correlated with their partial dissociation rates, and the exactly
equal optical yields obtained in the three hydrogenation experi-
ments suggest that, in all three cases, hydrogenation proceeded
through the same semi-dissociated species 16 (the difference
between 16a and 16b (as well as between 16c and 16d) is purely
conformational, and these pairs can equilibrate either, see
Scheme 4). At a later stage of the catalytic cycle, the double
bond of the substrate must have become re-coordinated to ensure
the occurrence of enantioselective catalytic hydrogenation.
Unfortunately, no further intermediates were detected in low-
temperature hydrogenation experiments, and hence, we ad-
dressed this problem computationally.
It should be noted that the equilibration of the catalyst-
substrate complexes with the semi-dissociated species is not
unique feature of the system under study. Thus, Brown et al.
have shown the existence of similar species through intramo-
lecular excitation transfer in 31P NMR of the diastereomeric
DIPAMP-Rh complexes of R-dehydroamino acids.33,34 Impor-
(33) Brown, J. M.; Chaloner, P. A.; Morris, G. A. J. Chem. Soc., Chem. Commun.
1983, 664.
9
2568 J. AM. CHEM. SOC. VOL. 130, NO. 8, 2008

Rh-Catalyzed Asymmetric Hydrogenation
A R T I C L E S
Scheme 7. Two Alternative Pathways for the Association and
tantly, it has been also recognized that the semi-dissociated
species have the potential to capture dihydrogen, thus affecting
the particular details of the catalytic cycle.33 The same conclu-
sions were later made from the 2D 31P-31P EXSY studies of
the DIPAMP,35 CHIRAPHOS,35 and 1-((2-methoxyphenyl)-
phenylphosphino)-2-diphenylphosphinoethane36 Rh complexes
of 3, as well as of Rh-BisP* complexes of enamides.27c
Similarly, the dramatic line shape changes in the 31P NMR of
the seven-membered ring chelates of Rh-bis(phosphinites) are
also explained by fast intramolecular interconversion of the
corresponding catalyst-substrate complexes that involves the
intermediacy of the semi-dissociated species.37 Thus, the
pathway involving the hydrogenation of the species with a non-
coordinated double bond might be operative in other catalytic
systems helping to avoid the high activation barriers of oxidative
addition in square planar catalyst-substrate complexes. It is
important therefore to understand how and when the stereose-
lection can occur if the stereodiscriminating coordination of the
double bond is lost during the oxidative addition step.
Migratory Insertion
2.2.6. Computational Study of Association and Migratory
Insertion Steps. From previous computational studies of Rh-
catalyzed asymmetric hydrogenation, it is known that the
oxidative addition of dihydrogen to an Rh complex is preceded
by the formation of molecular hydrogen complexes.16-20
Moreover, in some reaction pathways, this step has a very high
activation barrier due to the steric hindrance provided by the
coordinated substrate. However, these complications are avoided
in our system as the hydrogenation proceeds through partially
dissociated species 16 that is much less hindered, and different
modes of H2 addition become equally possible. Besides,
intermediates with non-coordinated double bonds can easily
interconvert, as is seen in the experimentally observable
transformation of 16a,b into 16b,c that occurs rapidly at -100
°C (Scheme 4). Hence, we have skipped in our analysis the
stages preceding double bond recoordination followed by
migratory insertion, since they cannot possibly contribute to the
generation of enantioselectivity.
to such configurations from the semi-coordinated complexes are
evidently unrealistic.
There are numerous possible isomers of partially coordinated
dihydride intermediates that theoretically can lead to multiple
reaction pathways. However, using the results of previous
theoretical studies and our own conclusions, we can reduce the
problem of the enantioselectivity to the difference between only
two pathways. Since we start the analysis from easily intercon-
vertible intermediates with non-coordinated double bonds, any
configuration of the chelate dihydride can be achieved through
the double bond coordination preceding the migratory insertion
step that allows the system to choose the lowest available
barriers. Furthermore, although previous studies of other
catalytic systems considered the possibility of migratory inser-
tion occurring through dihydride intermediates with apically
coordinated double bonds, we found it absolutely impossible
to accommodate the molecule of 3 in such a way with our
catalyst. The large spherical t-Bu groups inevitably collide with
either the phenyl or the methoxycarbonyl group, and approaches
This leaves four isomers 18a-d for consideration (Scheme
7). However, we did not think about the possibility of the
reaction taking place through 18a,b as that would imply
coordination of the double bond near the non-“chiral” phos-
phorus atom that is evidently disfavored (e.g., see above the
relative stabilities of corresponding catalyst-substrate com-
plexes). Hence, the problem of the origin of enantioselectivity
can be reduced in the present case to the analysis of the double
bond recoordination and the subsequent migratory insertion
starting from two semi-coordinated dihydrides 18c,d, which are
precursors of S- and R-products, respectively (Scheme 7).
Similarly to the computations performed for the catalyst-
substrate complexes, we used for calculations of association and
migratory insertion steps real compounds with one methanol
molecule either coordinated to the free coordination space at
rhodium or hydrogen-bonded to the carboxyl group of the
substrate. According to previous research, the R-configuration
of the hydrogenation product together with the catalyst config-
uration implies that the first hydrogen is transferred to the
â-carbon atom of the double bond, leading to the formation of
R-monohydride intermediate. Hence, we looked for reaction
pathways converting nonchelated dihydride intermediates 18c,d
(34) Brown, J. M.; Chaloner, P. A.; Morris, G. A. J. Chem. Soc., Perkin Trans.
II 1987, 1583.
(35) Bircher, H.; Bender, B. R.; von Philipsborn, W. Magn. Reson. Chem. 1993,
31, 293.
(36) Ramsden, J. A.; Claridge, T. D. W.; Brown, J. M. J. Chem. Soc., Chem.
Commun. 1995, 2469.
(37) Kadyrov, R.; Freier, T.; Heller, D.; Michalik, M.; Selke, R. J. Chem. Soc.,
Chem. Commun. 1995, 1745.
9
J. AM. CHEM. SOC. VOL. 130, NO. 8, 2008 2569

A R T I C L E S
Gridnev et al.
Figure 10. Alternative pathways for the association and migratory insertion steps computed on the B3LYP/SDD level of theory.
9
2570 J. AM. CHEM. SOC. VOL. 130, NO. 8, 2008

Rh-Catalyzed Asymmetric Hydrogenation
A R T I C L E S
Scheme 8. Direct Experimental Observation of the Formation of
Nonchelating “dihydride” Complexes in the Reaction of the
Dirhodium Tetraphosphine Solvate Dihydride with 315
Figure 11. Structures of the nonchelating dihydride 18d (left) and the
transition state TS3 (right) optimized on the B3LYP/SDD level of theory.
The arrows indicate relatively close contacts between the methyl groups
that would make impossible the same association pathway in the case of
18c (with the P-But group instead of the P-Me).
into chelated species 19c,d, which are direct precursors of
corresponding R-monohydride intermediates 20c,d (Scheme 7).
We found several transition states for the association step
that converts various conformations of open dihydride 18d into
chelate dihydride intermediate 19d. One of them, viz. TS3
(Figure 10), had significantly lower energy (7-12 kcal/mol)
than the others. The transformation of 18d into 19d‚MeOH
implies changing the lobe of the oxygen atom used to form the
coordination bond with rhodium with simultaneous pushing of
the methanol molecule out of the coordination sphere of rhodium
through transition state TS3.
This process requires much activity in the single nonhindered
quadrant and cannot be therefore paralleled in the association
of the double bond leading to 19c‚MeOH (Figure 11). In this
case, the lowest activation barrier was found for the process in
which the methanol molecule is pushed out through the
nonhindered quadrant via TS5 that is 6.9 kcal/mol higher in
energy than TS3. Apparently, this difference is the origin of
the enantioselectivity, since the activation barriers for the
migratory insertion step in both pathways are very low and
transition states TS4 and TS6 are almost equally stable.
It is appropriate to note that, recently, we directly observed
the operation of that kind of enantioselection on another three-
hindered-quadrant catalyst, dirhodium complex of tetraphosphine
ligand.15 In that case, it was possible to observe the initial
formation of nonchelating tetrahydride 22 when tetrahydride
21 was reacted with 3 at -100 °C. Although 22 had the catalyst,
activated hydrogen, and the substrate combined in one molecule,
it did not undergo migratory insertion unless it rearranged to
isomer 23 that immediately gave the corresponding “monohy-
dride” 24 (Scheme 8). That is the direct analogy to the
transformation 18c f 18d f 19d‚MeOH. It illustrates that,
although the intermediate 22 with incorrect configuration of Rh
is generated quantitatively, it does not undergo migratory
insertion due to its inability to coordinate the double bond
effectively. Migratory insertion occurs only when the isomer-
ization to 23 becomes possible, providing the configuration
appropriate for the effective double bond coordination.
mechanistic study demonstrated and emphasized the productivity
of this kind of ligand design. The non-chiral half of the catalyst
does not contribute to the flux of catalysis because the approach
of the prochiral double bond from that side is effectively
blocked. We have confirmed that this conclusion is valid for
representative substrates 3-14 and will report details of these
studies elsewhere.
Further, we showed that in the catalytic system containing 2
as a catalyst and 3 as a substrate, the reactivity of the catalyst-
substrate complex toward dihydrogen is correlated with the ease
of its partial dissociation. Moreover, the absolutely equal ee’s
obtained from low-temperature hydrogenations of different
catalyst-substrate complexes suggest that the enantioselection
occurs at the later stage of the reaction. DFT calculations
performed for a real catalytic system containing one methanol
molecule give ample proof that the enantioselectivity is gener-
ated during the formation of an octahedral chelate complex,
which precedes the irreversible migratory insertion step.
The relative ease of solvent insertion into the catalyst-
substrate complexes is determined by the configuration of the
Rh complex and of the double bond coordination mode:
methanol easily displaces the double bond having a polar
methoxycarbonyl group near the nonhindered quadrant. In turn,
the relative availability of chelate dihydrides 19c‚MeOH and
19d‚MeOH from nonchelating species 18 is also evidently
regulated by the configuration of the catalyst. Hence, the two
phenomena, viz. relative reactivity of catalyst-substrate com-
plexes and enantioselection, are correlated, but they do not stand
in the cause-and-effect relationship. Any chiral information
3. Conclusions
The Rh complex of (R)-(tert-butylmethylphosphino)(di-tert-
butylphosphino)methane (1) was the first reported catalyst for
the Rh-catalyzed asymmetric hydrogenation that must be
considered as having three hindered quadrants. The present
9
J. AM. CHEM. SOC. VOL. 130, NO. 8, 2008 2571

A R T I C L E S
Gridnev et al.
acquired before the irreversible step38 is lost due to the reversible
dissociation of the double bond.
Studies of other related systems in line with the conclusions
obtained in this work are underway in our laboratories.
4.4. Hydrogenation Experiments. Due to the importance of
temperature control for the conclusions of the present research, utmost
care was given to maintain the temperature of the hydrogenation
reaction. The temperature of the cooling bath was maintained at -78
°C (dry ice + acetone), and the sample connected to the rubber balloon
with dihydrogen was manually kept inside the bath throughout the whole
hydrogenation process. To enable the necessary mass transfer, the
sample was intensely shaken manually. This technique was previously
proven to be faultless; 100% conversion of a catalyst-substrate complex
into a monohydride intermediate unstable below -85 °C was achieved
via hydrogenation with the bath temperature set at -90 °C (see Gridnev,
I, D.; et al. J. Am. Chem. Soc. 2001, 123, 5268). To account for the
casual rise in temperature in the process of manual shaking, we carried
out a control experiment that demonstrated the lack of interconversion
between 4a and 4b at -70 °C.
4. Experimental Section
4.1. General Procedures. All reactions and manipulations were
performed under dry argon atmosphere using standard Schlenk-type
techniques. NMR experiments were carried out on a Jeol AL-300
spectrometer. Methanol-d4 of grade “100%” (99.6% D) packed in sealed
ampules was purchased from Cambridge Isotope Laboratories, Inc.
Hydrogen (99.9999%, Tanuma Sanso) was used for mechanistic studies.
4.2. Solvate Complex 2. A solution of catalytic precursor 1 (20-
40 mg) in 0.5 mL of CD3OD was prepared in a 5 mm NMR tube under
argon. Then, the sample was degassed by three cycles of freezing,
pumping, and warming. The sample was cooled to -20 °C, 2 atm H2
was admitted, and the temperature was raised to ambient. The sample
was intensely shaken manually during the hydrogenation at ambient
5. Computational Details
Geometries of all stationary points were optimized using
analytical energy gradients of self-consistent-field39 and density
functional theory (DFT).40 The latter utilized Becke’s three-
parameter exchange-correlation functional41 including the non-
local gradient corrections described by Lee-Yang-Parr (LYP),42
as implemented in the Gaussian 03 program package.43 All
geometry optimizations were performed using the SDD basis
set.44 This approach is widely used for the recent computational
studies of transition metal complexes and was shown to yield
results conforming with experimental data in the works of our
own45 and of others.46
temperature. The progress of the hydrogenation was monitored by 31
P
NMR; complete hydrogenation of the coordinated COD required 60-
90 min at room temperature. After completion of the reaction, excess
hydrogen was removed by several cycles of freezing, pumping, and
warming. The thus-prepared solution of solvate complex 2 in deute-
riomethanol was stable at ambient temperature.
4.3. Catalyst-Substrate Complexes 15. A solution of 1.5 equiv
of methyl (Z)-R-acetamidocinnamate (3) in CD3OD was added to the
above prepared solution of 2 in CD3OD either at ambient temperature
or at -100 °C.
1
15c: H NMR (300 MHz, CD3OD, 20 °C): δ 1.14 (d, 9H, 3CH3,
Acknowledgment. This work was financially supported by
the Global COE Program of Tokyo Institute of Technology.
Computational results in this research were obtained using
supercomputing resources at Information Synergy Center, Tokyo
Institute of Technology.
3JHP ) 15 Hz), 1.23 (d, 9H, 3CH3, 3JHP ) 16 Hz), 1.47 (d, 9H, 3CH3,
3JHP ) 15 Hz), 1.39 (dd, 3H, CH3, 2JHP ) 9 Hz, 4JHP ) 2 Hz,), 2.28 (s,
3H, CH3), 3.2-3.3 (m, 2H, CH2), 3.79 (s, 3H, CH3O), 6.13 (dd, 1H,
1
2
CHd, JHRh, JHP ) 7 and 2 Hz), 7.3-7.6 (m, 5H, C6H5); 13C NMR
(75 MHz, CD3OD, 20 °C): δ 9.43 (d, CH3, 1JCP ) 25 Hz), 19.87 (dd,
CH3CON, 4JCRh, 5JCP ) 3 and 1 Hz), 27.38 (d, But, 2JCP ) 4 Hz), 29.04
Supporting Information Available: NMR Charts, cartesian
coordinates of the optimized structures, and complete ref 43.
This material is available free of charge via the Internet at
http://pubs.acs.org.
2
2
(d, But, JCP ) 6 Hz), 29.4 (m, CH2), 30.30 (d, But, JCP ) 6 Hz),
34.49 (dd, C tert, 1JCP ) 21 Hz, 2JCRh ) 3 Hz,), 35.28 (dd, C tert, 2JCRh
) 2 Hz, 1JCP ) 14 Hz), 39.05 (dd, C tert, 1JCP ) 9 Hz, 2JCRh ) 2 Hz),
53.27 (OMe), 83.78 (d, CHd, 3JCP ) 10 Hz), 90.25 (dd, 3JCP ) 10 Hz,
2JCRh ) 6 Hz), 130.10, 130.19, 130.73 (5CH, C6H5), 137.83 (C tert),
JA076542Z
4
3
4
(39) Pulay, P. In Modern Theoretical Chemistry; Schaefer, H. F., Ed.; Plenum:
New York, 1977; Vol. 4, p 153.
168.6 (d, CdO, JCP ) 4 Hz), 188.4 (dd, NCO, JCP ) 6 Hz, JCRh
)
)
2 Hz); 31P NMR (122 MHz, CD3OD, 20 °C): δ -20.1 (dd, JPRh
1
(40) (a) Parr, R. G.; Yang, W. Density Functional Theory of Atoms and
Molecules; Oxford University Press: New York, 1989. (b) Cheeseman, J.
R.; Frisch, M. J.; Devlin, F. J.; Stephens, P. J. Chem. Phys. Lett. 1996, 52,
211.
2
1
2
136 Hz, JPP ) 89 Hz), 0.9 (dd, JPRh ) 136 Hz, JPP ) 89 Hz).
1
15d: H NMR (300 MHz, CD3OD, 20 °C): δ 1.18 (d, 9H, 3CH3,
3JHP ) 15 Hz), 1.29 (d, 9H, 3CH3, 3JHP ) 15 Hz), 1.46 (d, 9H, 3CH3,
3JHP ) 15 Hz), 1.62 (dd, 3H, CH3, 2JHP ) 9 Hz, 4JHP ) 1 Hz,), 2.20 (s,
3H, CH3), 3.2-3.3 (m, 2H, CH2), 3.80 (s, 3H, CH3O), 6.43 (dd, 1H,
(41) (a) Becke, A. D. Phys. ReV. A 1988, 38, 3098. (b) Becke, A. D. J. Chem.
Phys. 1993, 98, 5648.
(42) (a) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785. (b) Miehlich,
B.; Savin, A.; Stoll, H.; Preuss, H. Chem. Phys. Lett. 1989, 157, 200.
(43) Frisch, M. J.; et al. Gaussian 03, Revision B.05; Gaussian, Inc.: Pittsburgh
PA, 2003.
CHd, JHRh, JHP ) 5 and 2 Hz), 7.3-7.6 (m, 5H, C6H5); 13C NMR
(75 MHz, CD3OD, 20 °C): δ 11.06 (d, CH3, 1JCP ) 25 Hz), 19.74 (dd,
CH3CON, 4JCRh, 5JCP ) 3 and 1 Hz), 26.99 (d, But, 2JCP ) 5 Hz), 29.4
1
2
(44) (a) Dunning, T. H., Jr.; Hay, P. J. In Modern Theoretical Chemistry;
Schaefer, H. F., III, Ed.; Plenum,: New York, 1976; Vol. 3, p 1. (b)
Leininger, T.; Nicklass, A.; Stoll, H.; Dolg, M.; Schwerdtfeger, P. J. Chem.
Phys. 1996, 105, 1052.
(m, CH2), 29.70 (d, But, JCP ) 5 Hz), 30.15 (d, But, JCP ) 6 Hz),
2
2
33.68 (dd, C tert, 2JCRh ) 3 Hz, 1JCP ) 23 Hz), 36.31 (dd, C tert, JCP
1
(45) (a) Nakamura, I.; Bajracharya, G. B.; Wu, H.; Oishi, K.; Mizushima, Y.;
Gridnev, I. D.; Yamamoto, Y. J. Am. Chem. Soc. 2004, 126, 15423. (b)
Nakamura, I.; Mizushima, Y.; Gridnev, I. D.; Yamamoto, Y. J. Am. Chem.
Soc. 2005, 127, 9844. (c) Nishikata, T.; Yamamoto, Y.; Gridnev, I. D.;
Miyaura, N. Organometallics 2005, 25, 5025. (d) Gridnev, I. D.; del
Rosario, M. K. C.; Zhanpeisov, N. U. J. Phys. Chem. B 2005, 109, 12498.
(e) Patil, N. T.; Lutete, L. M.; Wu, H.; Pahadi, N.; Gridnev, I. D.;
Yamamoto, Y. J. Org. Chem. 2006, 71, 4270. (f) Gridnev, I. D.; Kikuchi,
S.; Touchy, A. S.; Kadota, I.; Yamamoto, Y. J. Org. Chem. 2007, 72, 8371.
(46) (a) Bittner, M.; Koppel, H. J. Phys. Chem. A. 2004, 108, 11116. (b) McKee,
M. L.; Swart, M. Inorg. Chem. 2005, 44, 6975. (c) Kondo, T.; Tsunawaki,
F.; Ura, Y.; Sadaoka, K.; Iwasa, T.; Wada, K.; Mitsudo, T. Organometallics
2005, 24, 905. (d) Nowroozi-Isfahani, T.; Musaev, D. G.; McDonald, F.
E.; Morokuma, K. Organometallics 2005, 24, 2921.
2
1
2
) 11 Hz, JCRh ) 2 Hz), 37.56 (dd, C tert, JCP ) 12 Hz, JCRh ) 2
3
3
Hz,), 52.97 (OMe), 83.05 (d, CHd, JCP ) 10 Hz), 89.68 (dd, JCP
)
17 Hz, 2JCRh ) 5 Hz), 129.75, 130.60 (5CH, C6H5), 130.84, 138.08 (C
4
3
tert), 168.8 (d, CdO, JCP ) 2 Hz), 189.2 (dd, NCO, JCP ) 6 Hz,
4JCRh ) 3 Hz); 31P NMR (122 MHz, CD3OD, 20 °C): δ -16.4 (dd,
1JPRh ) 136 Hz, 2JPP ) 81 Hz), -1.50 (dd, 1JPRh ) 137 Hz, 2JPP ) 81
Hz).
(38) In the present case, it is probably reductive elimination, since monohydride
20d is quite unstable and could not be detected.
9
2572 J. AM. CHEM. SOC. VOL. 130, NO. 8, 2008