Asymmetric hydrogenation of cyclohexane-1,2-dione over cinchonidine-modified platinum

Article abstract of DOI:10.1016/S0021-9517(02)00144-6

The asymmetric hydrogenation of cyclohexane-1,2-dione over cinchonidine-modified platinum was investigated. Despite the fact that the first hydrogenation step is close to nonenantioselective, a high enantiomeric excess is obtained for the (R)-α-hydroxyketone due to kinetic resolution. In the second hydrogenation step one out of the four reactions of the network is substantially accelerated with respect to the others and with respect to the reaction in the absence of modifier, leading to an enantiomeric excess of (1R,2R)-trans-cyclohexane-1,2-diol of over 80%. Comparison with recently reported asymmetric hydrogenation of α-hydroxyethers indicates striking similarities, which hint at similar reactant-modifier interaction in both cases. The importance of cis versus trans conformation of the reactant for the reactant-modifier interaction emerges from a comparison of suggested reaction intermediates for cyclohexane-1,2-dione and butane-2,3-dione hydrogenation, respectively.

Full text of DOI:10.1016/S0021-9517(02)00144-6

Journal of Catalysis 215 (2003) 116–121
www.elsevier.com/locate/jcat
Asymmetric hydrogenation of cyclohexane-1,2-dione over
cinchonidine-modified platinum
Otmar J. Sonderegger, Thomas Bürgi, and Alfons Baiker ∗
Laboratory of Technical Chemistry, ETH Hönggerberg HCI, CH-8093 Zurich, Switzerland
Received 5 September 2002; revised 17 October 2002; accepted 18 October 2002
Abstract
The asymmetric hydrogenation of cyclohexane-1,2-dione over cinchonidine-modified platinum was investigated. Despite the fact that the
first hydrogenation step is close to nonenantioselective, a high enantiomeric excess is obtained for the (R)-α-hydroxyketone due to kinetic
resolution. In the second hydrogenation step one out of the four reactions of the network is substantially accelerated with respect to the
others and with respect to the reaction in the absence of modifier, leading to an enantiomeric excess of (1R,2R)-trans-cyclohexane-1,2-diol
of over 80%. Comparison with recently reported asymmetric hydrogenation of α-hydroxyethers indicates striking similarities, which hint at
similar reactant–modifier interaction in both cases. The importance of cis versus trans conformation of the reactant for the reactant–modifier
interaction emerges from a comparison of suggested reaction intermediates for cyclohexane-1,2-dione and butane-2,3-dione hydrogenation,
respectively.
2003 Elsevier Science (USA). All rights reserved.
Keywords: Cyclohexane-1,2-dione; Cyclohexane-1,2-diol; Asymmetric; Enantioselective; Hydrogenation; Cinchonidine; Platinum
1. Introduction
a helpful tool in deducing the stereochemical requirements
for fast reaction, since several reaction pathways compete,
as shown in the reaction network in Fig. 1. The vicinal dike-
tones investigated so far show considerable structural flexi-
bility. In particular they can exist in cis and trans conforma-
tion, which makes analysis of stereochemical requirement
difficult. We therefore decided to investigate cyclohexane-
1,2-dione (1), which has a fixed cis conformation of the keto
groups.
Several applications of the chiral hydrogenation prod-
uct cyclohexane-1,2-diol (3) have been reported. (R,R)-
cyclohexane-1,2-diol is used in organic synthesis as a chiral
auxiliary. The conjugate addition of organocuprate reagents
and Grignard reagents to α, β-unsatured esters of chiral
(R,R)-cyclohexane-1,2-diol proceed with high diastereose-
lectivity [11,12]. For instance, it is used as an auxiliary for
the enantio- and diastereoselective synthesis of β-substituted
five-, or six-membered cyclohexanecarboxylates [13]. It is
further used as a photocrosslinkable chiral doping agent in
liquid-crystal mixtures and in optical devices [14] and in the
preparation of carbohydrate mimetics having anti-adhesive
properties [15,16].
Heterogeneous enantioselective hydrogenation using chi-
rally modified metal catalysts is a promising route for
the synthesis of optically pure compounds using a hetero-
geneous process. One of the most investigated catalysts
for such reactions is the platinum cinchona alkaloid sys-
tem [1–7], and the mechanism of enantiodifferentiation is
the target of ongoing research in several laboratories. The
scope of the reaction is steadily increasing, although it is
still rather limited to α-keto acid derivatives and “activated”
ketones.
Here we investigate the heterogeneous enantioselective
hydrogenation of a cyclic vicinal diketone, cyclohexane-
1,2-dione. The enantioselective hydrogenation of vicinal
diketones on Pt with cinchonidine as modifier has been in
the focus of several studies. Investigated diketones include
butane-2,3-dione [8], hexane-3,4-dione [9], and 1-phenyl-
propane-1,2-dione [10]. The reduction of diketones can be
*
Corresponding author
E-mail address: baiker@tech.chem.ethz.ch (A. Baiker).
0021-9517/03/$ – see front matter
doi:10.1016/S0021-9517(02)00144-6
2003 Elsevier Science (USA). All rights reserved.

O.J. Sonderegger et al. / Journal of Catalysis 215 (2003) 116–121
117
◦
ee) and at 150 C it was negligible. The peaks in the chro-
matogram were identified by comparison with commercial
racemic compounds 2 (Aldrich) and cyclohexane-1,2-diol
(3) (Acros 98%). The absolute configuration of the prefer-
entially formed enantiomer was determined by comparison
with (R,R)-3 from Fluka (ꢀ 99%). The GC was calibrated
by injecting mixtures of 1, 2, and 3 of known composi-
tion. To follow the time dependence of the reaction 100-µl
samples were withdrawn from the reactor at defined times
using a syringe. Enantiomeric excess is expressed as ee
(%) = 100 × (|R − S|)/(R + S). Diastereomeric excess is
expressed as de (%) = 100 × (|cis − trans|)/(cis + trans).
The relative concentration of 1, 2 (sum of (R) and (S)) and 3
(sum of (R,R), (S,S) and (meso)), as well as the ratio of the
enantiomers of 2, were determined in one GC run and the
relative concentrations of (R,R)-3, (S,S)-3, and (meso)-3 in
a second run using a different temperature program.
3. Results
3.1. General features of the reaction
Fig. 1. Reaction network for the hydrogenation of cyclohexane-1,2-
The enantioselective hydrogenation of cyclohexane-1,2-
dione(1) resulted in an 81.3% enantiomeric excess (ee) in
favor of (1R, 2R)-trans-cyclohexane-1,2-diol ((R,R)-3) at
full conversion under standard conditions. The diastere-
omeric excess (de) of the cis-cyclohexane-1,2-diol ((R,S)-
3,(S,R)-3)) was 22.9% at complete conversion. Due to ki-
netic resolution the ee of the (R)-intermediate (R)-2 reached
very high values, close to 100%, toward the end of the reac-
tion. We found that the purity of the reactant and the water
content in the reaction mixture strongly affected the ee. Ad-
dition of 1❤ of water resulted in a decrease of ee to 29.2%.
Addition of 5❤ of water gave an ee of 18.4%. We also no-
ticed that distillation of the reactant was necessary to obtain
high ee and that the residue contained some polymerized re-
actant. Thus purification of the reactant seems to be a neces-
sary prerequisite to avoid catalyst deactivation.
The ee of the product depended on the cinchonidine con-
centration as shown in Fig. 2. The increase in ee was signif-
icant at low cinchonidine concentration. At 1 mg cinchoni-
dine, corresponding to a reactant/modifier ratio of 542, the
ee of the product (R,R)-3 amounted to 75.9% and increased
only slightly at higher cinchonidine concentration. Such be-
havior is quite common for the Pt–cinchona system. It is
likely due to the increasing adsorption of the modifier on
the Pt catalyst, which is saturated already at low concentra-
tion [17,18].
ꢀ
−1 −1
h
dione (1). The values indicate the rate constants k = k /V [g
]. Note
i
i
ꢀ
that k is used to be consistent with previously reported rate constants [26].
i
The values in parentheses are rate constants for the corresponding racemic
reaction. See text for details.
2. Experimental
A 5 wt% Pt/Al2O3 catalyst (Engelhard 4759) was pre-
◦
reduced in flowing hydrogen for 90 min at 400 C. The
platinum dispersion after heat treatment was 0.27 as de-
termined by TEM measurements. Toluene and cinchoni-
dine were used as received (Baker, Fluka). The reactant
cyclohexane-1,2-dione (1) (Avocado > 98%) was distilled
before use. The hydrogenation reactions were carried out
in a 100-ml stainless-steel autoclave equipped with a 50-ml
glass liner and PTFE cover. The reactor was magnetically
stirred (n = 500 min−1). The pressure was held at a constant
value by computerized constant-volume–constant-pressure
equipment (Büchi BPC 9901), which allowed following the
hydrogen uptake. Under standard conditions, 42 2 mg pre-
reduced catalyst, 1.84 mmol substrate, 6.8 µmol modifier,
and 10 ml solvent were used and the reaction was carried
out at 30 bar and 15 ◦C.
Conversion, enantiomeric excess (ee), and diastereomeric
excess (de) were determined using an HP 6890 gas chro-
matograph (GC) and a Chirasil-DEX CB (ChromPack)
capillary column. The injector temperature was 150 ◦C.
We noticed that the measured enantiomeric excess of 2-
hydroxycyclohexanone (2) strongly depended on the injec-
tor temperature of the GC. At an injector temperature of
The highest ee reached for the product (R,R)-3 was
83.7% at full conversion (15 ◦C, 30 bar, 4 mg cinchonidine,
1/cinchonidine ratio 135). A decrease in the amount of cat-
alyst had a negative effect on the ee. Using 31 mg catalyst
instead of 42 mg led to a decrease of the ee to 70.6%. In-
creasing the pressure to 90 bar also had a significant nega-
tive effect on the ee of (R,R)-3, which decreased to 67.5%,
◦
250 C the ee was significantly lower (about 25%) than at
lower temperature, likely due to fast keto–enol tautomerism.
◦
At 200 C the effect was much smaller (about 2% lower

118
O.J. Sonderegger et al. / Journal of Catalysis 215 (2003) 116–121
Fig. 2. Enantiomeric excess of (R,R)-3 at full conversion as a function
of the reactant 1/cinchonidine ratio (mol/mol). The concentration of
cinchonidine was varied. Other parameters are the same as under standard
conditions (see text).
whereas decreasing the pressure to 10 bar lowered the ee
only slightly to 78.5%. The ee was also quite sensitive to
temperature in the following manner: 35 ◦C, 64% ee; 25 ◦C,
77.6% ee; 15 ◦C, 81.3% ee; 5 ◦C, 81.0% ee. Other solvents,
which have been tested are acetic acid, tetrahydrofurane, 2-
propanol, and acetonitrile, but the reaction rates and ee were
significantly lower than for toluene. We also noted that the
solubility of trans-3 in toluene is higher than that of cis-3.
Fig. 3. Kinetic behavior of the enantioselective hydrogenation of cyclo-
hexane-1,2-dione (1) over cinchonidine-modified Pt. Solid lines are ob-
tained from the fit of the model depicted in Fig. 1 to the experimental data.
Standard conditions as specified in the experimental part. Bottom: Frac-
tion of 1, 2 (sum of (R)-2 and (S)-2), and 3 (sum of (R,R)-3, (S,S)-3 and
(meso)-3) as a function of time. Top: Enantiomeric excesses of intermedi-
ate (ee-2) and final product (ee-3) and diastereomeric excess of final product
de-3.
3.2. Kinetics
The hydrogenation of 1 can give useful information on
the modified Pt catalyst, since several reactions compete.
A prerequisite for such an analysis is, however, the exam-
ination of the reaction network depicted in Fig. 1. In order
to perform this, the reaction was followed in time. The ki-
netic data shown in Figs. 3 and 4 for modified and unmod-
ified catalyst (presence and absence of cinchonidine) were
fit to a simple model corresponding to the reaction scheme
depicted in Fig. 1. An analogous reaction scheme has been
proposed for the hydrogenation of butane-2,3-dione [8]. The
simple kinetic model is applied to analyze the connectiv-
ity between the different steps and their relative reaction
rates. Each single reaction is considered as first order with
respect to reactant and catalyst. The reaction rates for the
different species i are defined as: ri [mol/hgcat] = 1/mcat
dNi/dt = 1/(mcat/V ) 1/V dNi/dt. Assuming constant vol-
ume, ri = V/mcat dci/dt, where Ni corresponds to moles of
species i, mcat is the mass of the catalyst, and V is the vol-
ume of the reaction mixture. The rate equations derived for
the reaction network (Fig. 1) are:
These equations were solved numerically. The experimental
data were fit using a least squares procedure, which is
a nonlinear χ2 fitting routine based on the Levenberg–
Marquardt method. The concentrations of all six involved
species, 1, (R)-2, (S)-2, (R,R)-3 meso-3 ((R,S)-3 plus
(S,R)-3), and (S,S)-3, at any time were calculated by
integrating the above rate laws taking into account the
known initial concentrations and initial guesses for the rate
constants k. The six rate constants kR, kS, kRR, kRS, kSR
,
and kSS were simultaneously fit to the following time-
dependent experimental values. The concentration of 1, the
concentration of intermediate product 2 (sum of (R)-2 and
(S)-2), the concentration of product 3 (sum of (R,R)-3,
meso-3, and (S,S)-3), the ee of 2, the ee of 3, and the de
of 3. A total of 48 data points (32 for the racemic reaction)
were used for the fit. For the racemic reaction (without
cinchonidine addition) the three rate constants kR = kS,
kRR = kSS, and kSR = kRS were fit to the concentration of
1, the concentration of intermediate product 2 (sum of (R)-2
and (S)-2), the concentration of product 3 (sum of (R,R)-
3, meso-3 and (S,S)-3), and the de of 3. The initial guess
for the rate constants was varied in order to search for local
minima, but all initial guesses converged to the same set
of rate constants. The fitting was implemented in a Fortran
program using standard numerical procedures [19].
r1 = −(kR + kS)[1],
ꢀ
ꢁ
r(R)−2 = kR[1] − (kRR + kRS) (R) − 2 ,
ꢀ
ꢁ
r(S)−2 = kS[1] − (kSS + kSR) (S) − 2 ,
ꢀ
ꢁ
r(R,R)−3 = kRR (R) − 2 ,
ꢀ
ꢁ
ꢀ
ꢁ
r(meso)−3 = kRS (R) − 2 + kSR (S) − 2 ,
ꢀ
ꢁ
r(S,S)−3 = kSS (S) − 2 .

O.J. Sonderegger et al. / Journal of Catalysis 215 (2003) 116–121
119
Fig. 5. Kinetic behavior of the hydrogenation of racemic α-2-hydroxycyclo-
hexanone (2) under standard conditions. Lines are only to guide the eyes.
for the reaction starting from the diketone 1 confirms the
kinetic analysis given above.
4. Discussion
Fig. 4. Kinetic behavior of the racemic hydrogenation of cyclo-
hexane-1,2-dione (1) over cinchonidine-modified Pt. Bottom: fraction of
1, 2, and 3 as a function of time. Solid lines are obtained from the fit of the
model depicted in Fig. 1 to the experimental data, considering that k = k ,
Rate constants for the enantioselective hydrogenation of
butane-2,3-dione, based on the same reaction network as
the one in Fig. 1, have previously been reported [8], which
allows a comparison of the reactions of the two diones. There
are similarities but also important differences. A striking
similarity is the finding that for the second reaction 2 → 3
the rate constant kSR is the largest, i.e., the (S)-2 → (S,R)-
3 reaction is the fastest. kSR is less than four times larger
than the other rate constants of the 2 → 3 reactions in
the case of butane-2,3-dione. In contrast, it is 15 times
larger for cyclohexane-1,2-dione. An important difference
concerns the relative rates of the 1 → 2 and 2 → 3 reactions.
The dominant pathway (S)-2 → (meso)-3 for the second
reaction 2 → 3 is about an order of magnitude faster than
the first reaction 1 → 2 in the case of cyclohexane-1,2-dione,
whereas it is about an order of magnitude slower for butane-
2,3-dione. Also, for butane-2,3-dione the first hydrogenation
step shows some enantioselectivity (42% as calculated from
the rate constants kR and kS) [8], whereas for cyclohexane-
1,2-dione the first hydrogenation step shows almost no
enantioselectivity (ee < 10%).
In the racemic reaction (Fig. 4) the hydrogenation rate
constants kRS and kSR for the reactions leading to the
cis product are larger than kRR and kSS for the reactions
resulting in trans product. This can be rationalized if one
assumes that (i) the hydrogen approaches the carbonyl bond
from the surface side, and that (ii) 2 prefers an adsorption
on the enantioface that is opposite to the O–H group
such that the O–H group points away from rather than
towards the surface. Note that for the hydrogenation of
the corresponding 2-hydroxy-3-butanone the opposite was
found (kRS = kSR < kRR = kSS).
R
S
k
= k and k = k
for the racemic reaction. Top: diastereomeric
RS
RR
SS
SR
excess de-3. The experimental data were measured under standard condi-
tions except that no cinchonidine was added to the reaction mixture.
Figs. 3 and 4 show that the simple kinetic model based
on the scheme depicted in Fig. 1 fits the data well. The
resulting rate constants for the individual reactions are
given in Fig. 1 In the presence of cinchonidine the first
step 1 → 2 is only slightly enantioselective. From the
values of kR and kS an enantiomeric excess ee of less
than 10% is obtained. The reaction (S)-2 → (meso)-3 is an
order of magnitude faster than any of the other individual
reactions, which has the following consequences: (i) (R)-
2 is much more abundant than (S)-2 (except for very low
conversion), leading to high ee of 2 even though both
enantiomers are formed at similar rates (kinetic resolution),
(ii) 2 is never present at high concentration (about 20%
at most), (iii) (R,R)-3 is much more abundant than (S,S)-
3 (high ee-3), although the rate constants kRR and kSS
for their formation are very similar, and (iv) the meso
form of 3 is formed preferentially. A comparison with the
racemic reaction reveals that the first reaction, 1 → 2, slows
down in the presence of cinchonidine, whereas the second,
2 → 3, is accelerated. As a consequence the maximum
concentration of the intermediate is relatively low. Fig. 5
shows experimental data obtained for the enantioselective
hydrogenation of racemic 2 under standard conditions. The
observation that similar values of ee and de are obtained as

120
O.J. Sonderegger et al. / Journal of Catalysis 215 (2003) 116–121
The most important structural difference between cyclo-
hexane-1,2-dione (1) and butane-2,3-dione is the greater
rigidity of the former. Considering this fact, the finding is
not surprising that for the former the first hydrogenation
step is close to racemic, whereas for the latter moderate
enantioselectivity was observed. For butane-2,3-dione the
trans is much more stable than the cis conformation. In
fact, ab initio calculations (Hartree–Fock, 6-31G(d,p) ba-
sis set) [20] reveal that trans is more stable than cis by
about 8 kcal/mol and that the cis conformer is a saddle
point rather than a minimum on the potential energy sur-
face. Mechanistic models for the enantiodifferentiation in
the ethyl pyruvate reaction propose a hydrogen-bonding in-
teraction between the quinuclidine N of the modifier and
the carbonyl group being hydrogenated. This hydrogen bond
can be formed either between the protonated quinuclidine
N and the keto oxygen [21] or between the quinuclidine N
and the half-hydrogenated state of the keto group, in case
the quinuclidine N is not protonated [22,23]. It was further-
more proposed that the quinoline part of the modifier plays
an important role for enantiodifferentiation because of repul-
sive interactions with the reactant [24,25]. Due to this repul-
sion one would expect that in the presence of cinchonidine,
adsorption on the one enantioface is favored if the butane-
2,3-dione adopts trans conformation. This is schematically
shown in Fig. 6. Note that this view of enantiodifferentiation
predicts the correct absolute configuration of the major reac-
tion product (R)-2-hydroxy-3-butanone. Adsorption, which
would lead to the (S)-product, is less favored due to repulsive
interactions. For the cis conformation adsorption on the two
enantiofaces is identical also in the presence of cinchonidine
(see Fig. 6c). Cyclohexane-1,2-dione (1) is fixed in the cis
conformation. Hence, according to the view of enantiodif-
ferentiation outlined above, one would not expect enantios-
election for the reaction yielding 2, in good agreement with
observation. We should, however, note that for 1 the two car-
bonyl groups are not exactly oriented coplanar due to the
six-membered ring.
Fig. 6. Possible interaction of trans and cis diketones with cinchonidine
modifier. In the presence of cinchonidine, adsorption on the two enantio-
faces is different for the trans dione (a), (b), whereas this is not the case for
the cis dione (c).
requirements of the chiral site. Very recently Studer et al. re-
ported on the dynamic kinetic resolution of α-keto ethers,
using cinchonidine modified Pt catalysts [26]. They found
that the (S)-α-keto ether reacts faster than the (R)-α-keto
ether and that the major product is the (R,S)-hydroxy ether.
Depending on conditions the pseudo-first-order kinetic con-
stant kSR and kRS differed by one to two orders of magni-
tude. This striking similarity between the behavior of α-keto
ethers and α-hydroxy-ketone (2) strongly supports a simi-
lar mechanism for enantiodifferentiation for the two, which
indicates that the second oxygen besides the keto O may
act as a second hydrogen-bond acceptor, as has been pro-
posed also for α-ketoester hydrogenation [25]. Furthermore
the similarity shows that the O–H group of 2 is not involved
in enantiodifferentiation and that it does not hinder the cru-
cial modifier–reactant interaction.
As concerns the second reaction 2 → 3, the relative
arrangement of O–H and keto group may be similar for the
2-hydroxy-3-butanoneand the 2-hydroxycyclohexanone(2),
since here the former is more likely to adopt a cis con-
formation, due to intramolecular hydrogen-bonding. This is
also confirmed by ab initio calculations, which show that
the cis conformation is more stable by about 2 kcal/mol
(Hartree–Fock, 6-31G(d,p) basis set). In this light it is not
surprising that for both the 2-hydroxy-3-butanone and the
2-hydroxycyclohexanone (2) the same reaction among the
four possible (see Fig. 1) dominates. Due to the presence of
cinchonidine, kSR is larger than kRS. For 2-hydroxycyclohe-
5. Conclusions
The racemic and asymmetric hydrogenation of cyclo-
hexane-1,2-dione (1) was investigated and the reaction
network analyzed. In the absence of modifier the first
hydrogenation step is considerably faster than the sec-
ond, leading to a high intermediate concentration of the
2-hydroxycyclohexanone (2). In the presence of cinchoni-
dine the latter is never present at high concentration due
to rate acceleration of the second hydrogenation step in-
xanone (2) the determined ratio of the rate constants kSR
/
kRS is 53, whereas it is 3–4 for 2-hydroxy-3-butanone.
The lower value for the latter may indicate that structural
rigidity is favorable for discrimination between (R)-2 and
(S)-2. Since (R)-2 and (S)-2 are electronically identical,
their vastly different reaction rate can be traced to the steric

O.J. Sonderegger et al. / Journal of Catalysis 215 (2003) 116–121
121
duced by the modifier. Kinetic analysis of the reaction
network shows that one of the four possible reactions of
2-hydroxycyclohexanone (2) to cyclohexane-1,2-diol (3) is
drastically accelerated. Comparison with the related hydro-
genation of butane-2,3-dione shows that the same reaction is
favored in the second hydrogenation step; however, discrim-
ination is more pronounced for cyclohexane-1,2-dione (1),
possibly due to its greater structural rigidity. In contrast
to the asymmetric hydrogenation of butane-2,3-dione the
hydrogenation of cyclohexane-1,2-dione (1) to 2-hydroxy-
cyclohexanone (2) in the presence of cinchonidine shows
almost no enantioselectivity. This finding can be explained
by the fixed cis conformation of cyclohexane-1,2-dione (1).
When interacting with the cinchonidine modifier butane-2,3-
dione may be present partly in trans conformation, leading
to different stability for adsorption on the two enantiofaces,
as predicted by proposed modifier–reactant interaction mod-
els.
[9] W.A.H. Vermeer, A. Fulford, P. Johnston, P.B. Wells, J. Chem. Soc.
Chem. Commun. (1993) 1053.
[10] E. Toukoniitty, P. Mäki-Arvela, M. Kuzma, A. Villela, A.K. Neyes-
tanaki, T. Salmi, R. Sjöholm, R. Leino, E. Laine, D.Y. Murzin, J.
Catal. 204 (2001) 281.
[11] F. Cheng-Lin, S. Hiroshi, S. Kiyoshi, Tetrahedron Letters 31 (1990)
4751.
[12] F. Cheng-Lin, S. Tomoyuki, S. Hiroshi, S. Kiyoshi, Tetrahedron
Asymmetry 2 (1991) 389.
[13] F. Cheng-Lin, S. Hiroshi, S. Kiyoshi, J. Org. Chem. 57 (1991) 4300.
[14] S. Kelly, in: EP 755915 A2, Rolic AG, Switzerland, 1997.
[15] A. Toepfer, G. Kretzschmar, E. Bartnik, D. Seiffge, in: HEP 671408
A2, Hoechst AG, Germany, 1995.
[16] A. Toepfer, G. Kretzschmar, E. Bartnik, D. Seiffge, in: EP 671409 A2,
Hoechst AG, Germany, 1995.
[17] D. Ferri, T. Bürgi, A. Baiker, J. Chem. Soc. Chem. Commun. (2001)
1172.
[18] D. Ferri, T. Bürgi, J. Am. Chem. Soc. 123 (2001) 12074.
[19] W.H. Press, Numerical Recipes in Fortran, Cambridge Univ. Press,
Cambridge, UK, 1992.
[20] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb,
J.R. Cheeseman, V.G. Zakrzewski, J.A. Montgomery, R.E. Stratmann,
J.C. Burant, S. Dapprich, J.M. Millam, A.D. Daniels, K.N. Kudin,
M.C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi,
B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G.A.
Petersson, P.Y. Ayala, Q. Cui, K. Morokuma, D.K. Malick, A.D.
Rabuck, K. Raghavachari, J.B. Foresman, J. Cioslowski, J.V. Ortiz,
A.G. Baboul, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.
Komaromi, R. Gomperts, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-
Laham, C.Y. Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe,
P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, J.L. Andres, C.
Gonzalez, M. Head-Gordon, E.S. Replogle, J.A. Pople, Gaussian Inc.,
Pittsburgh, PA, 1998.
[21] O. Schwalm, B. Minder, J. Weber, A. Baiker, Catal. Lett. 23 (1994)
271.
[22] K.E. Simons, P.A. Meheux, S.P. Griffiths, I.M. Sutherland, P. Johnston,
P.B. Wells, A.F. Carley, M.K. Rajumon, M.W. Roberts, A. Ibbotson,
Recl. Trav. Chim. Pays-Bas 113 (1994) 465.
[23] M. Schürch, O. Schwalm, T. Mallat, J. Weber, A. Baiker, J. Catal. 169
(1997) 275.
[24] A. Vargas, T. Bürgi, A. Baiker, J. Catal. 197 (2001) 378.
[25] T. Bürgi, A. Baiker, J. Catal. 194 (2000) 445.
Acknowledgment
Financial support by the Swiss National Science Founda-
tion is kindly acknowledged.
References
[1] Y. Orito, S. Imai, S. Niwa, J. Chem. Soc. Jpn. (1979) 1118.
[2] P.B. Wells, A.G. Wilkinson, Top. Catal. 5 (1998) 39.
[3] H.U. Blaser, H.P. Jalett, M. Müller, M. Studer, Catal. Today 37 (1997)
441.
[4] A. Baiker, J. Mol. Catal. A Chem. 115 (1997) 473.
[5] A. Baiker, J. Mol. Catal. A Chem. 163 (2000) 205.
[6] A. Baiker, H.U. Blaser, in: G. Ertl, H. Knözinger, J. Weitkamp (Eds.),
Handbook of Heterogeneous Catalysis, Vol. 5, VCH, Weinheim, 1997,
p. 2422.
[7] G.J. Hutchings, D.J. Willock, Top. Catal. 5 (1998) 177.
[8] M. Studer, V. Okafor, H.U. Blaser, J. Chem. Soc. Chem. Commun.
(1998) 1053.
[26] M. Studer, H.-U. Blaser, S. Burkhardt, Adv. Synth. Catal. 344 (2002)
511.