First Enantioselective Hydrogenation of a Trifluoro-β-ketoester with Cinchona-Modified Platinum

Article abstract of DOI:10.1006/jcat.2000.2871

A series of trifluoromethyl ketones possessing different functional groups were hydrogenated over Pt/Al₂O₃ modified by cinchonidine and O-methylcinchonidine. The highest enantiomeric epcess of 90% was achieved in the hydrogenation of

Full text of DOI:10.1006/jcat.2000.2871

Journal of Catalysis 193, 161–164 (2000)
doi:10.1006/jcat.2000.2871, available online at http://www.idealibrary.com on
RESEARCH NOTE
First Enantioselective Hydrogenation of a Trifluoro-�-ketoester
with Cinchona-Modified Platinum
1
M. von Arx, T. Mallat, and A. Baiker
Laboratory of Technical Chemistry, Swiss Federal Institute of Technology, ETH-Zentrum, CH-8092 Z u¨ rich, Switzerland
Received December 15, 1999; revised March 21, 2000; accepted March 22, 2000
6
0% ee over Pt/Al2O3 modified with cinchonidine (CD).
A series of trifluoromethyl ketones possessing different func- But recently the hydrogenation of benzyltrifluoromethyl
tional groups were hydrogenated over Pt/Al2O3 modified by cin- ketone afforded rather poor enantioselectivity, only 18%
chonidine and O-methylcinchonidine. The highest enantiomeric
(
11). To clarify whether the Pt–cinchona system is a gener-
epcess of 90% was achieved in the hydrogenation of ethyl 4,4,4-
trifluoroacetoacetate. The remarkable variation of enantioselectiv-
ity in the synthesis of various �,�,�-trifluoromethyl alcohols is
interpreted byelectronic and steric effects. Future mechanistic stud-
ies have to clarify the unexpectedly big influence of acidic sol-
vent and the special role of the (C9)–OH group in cinchonidine.
allyusefulcatalyst for the synthesisof�,�,�-trifluoromethyl
alcohols, we have extended our investigations to a broader
range of substrates, using CD and some of its simple deriva-
tives as chiral modifiers for Pt. These studies provide further
insight into the crucial role of electronic and steric effects
in the asymmetric hydrogenation of activated ketones on
chirally modified Pt.
�
c 2000 Academic Press
Key Words: enantioselective hydrogenation; cinchonidine;
platinum–alumina; trifluoromethyl ketones.
2
. METHODS
1
. INTRODUCTION
All substrates, CD and CD hydrochloride (CD � HCl),
were used as received (Fluka, Aldrich). O-Methylcincho-
Chiral-fluorinated compounds have a great potential as nidine (MeOCD) was synthesized as described before (12).
antiinflammatory and cardiovascular drugs or anticancer A 5 wt% Pt/Al O catalyst (Engelhard 4759) was prere-
2
3
�
and antiviral agents (1). A possible synthetic route to �,�,�- duced in flowing hydrogen for 90 min at 400 C. After being
trifluoromethyl alcohols is the asymmetric hydrogenation cooled to room temperature in hydrogen, the catalyst was
of trifluoromethyl ketones. Various chiral boranes and alu- transferred to the reactor without exposure to air. Platinum
minium hydride reagents provided very good results in dispersion after heat treatment was 0.27, as determined by
the hydrogenation of aromatic and nonaromatic trifluo- TEM.
romethyl ketones. The chemical yields are generally above
0% and the enantiomeric excess (ee’s) are 90% or higher stirred 100-ml stainless steel autoclave equipped with a
Hydrogenations were carried out in a magnetically
8
under very mild conditions (2, 3). Enzymatic reduction by 50-ml glass liner and a PTFE cover. Under standard condi-
the use of bakers’ yeast or Geotrichum candidum acetone tions 42 � 2 mg of catalyst, 1.84 mmol of substrate, 6.8 �mol
powder afforded moderate to almost perfect enantioselec- of modifier, and 5 ml of solvent were stirred magnetically
tivities (4, 5). Homogeneous chiral transition metal cata- (1000 rpm) under a constant hydrogen pressure of 10 bar.
lysts, which perform excellently in the enantioselective hy- The reaction was followed by monitoring of the hydro-
drogenation of a variety of ketones (6), have been used only gen uptake (B u¨ chi BPC 9901). The experiments were car-
for the hydrogenation of 2,2,2-trifluoroacetophenone and ried out at room temperature except for the volatile com-
some of its aryl-substituted derivatives, affording up to 96% pound 7.
ee at 100% conversion (7, 8).
The products were isolated and identified by NMR
We have reported for the first time (9, 10) that 2,2,2- spectroscopy. The ee was determined by direct gas chro-
trifluoroacetophenone can be hydrogenated with around matographic analysis of the reaction mixture, using a
HP 6890 gas chromatograph and a Chirasil-DEX CB
1
(Chrompack) capillary column. In the case of 7 the
product was derivatized with acetic acid anhydride and
To whom correspondence should be addressed. E-mail: baiker@tech.
chem.ethz.ch. Fax: 0041 1 632 11 63.
161
0021-9517/00 $35.00
Copyright �c 2000 by Academic Press
All rights of reproduction in any form reserved.

1
4
62
VON ARX, MALLAT, AND BAIKER
-dimethylaminopyridine. The enantiomeric excess is ex-
pressed as ee(% ) = 100 � ꢀ(R� S)ꢀ/(R + S).
3
. RESULTS
The potential of the Pt–cinchonidine system in the enan-
tioselective hydrogenation of trifluoromethyl ketones is il-
lustrated in Table 1. A broad range of substrates have been
selected, compounds which possess alkyl, alkyaromatic,
aromatic, heteroaromatic, or ester functions. Although the
hydrogenation of 1 and 2 has already been reported, these
substrates are also included here to show the influence of
structural variations in the modifier. Most of the experi-
ments were carried out in toluene and acetic acid, solvents
which give generally the best results in the hydrogenation of
activated ketones over the Pt–CD system (13, 14). The only
exception is the hydrogenation of 1, a reaction in which not
toluene but dichlorobenzene is the best solvent.
FIG. 1. Pressure dependence of the hydrogenation of 5 under stan-
Variation of the structure of trifluoromethyl ketones dard conditions in acetic acid, using MeOCD as the modifier.
resulted in strikingly different reaction rates and enantiose-
in the reduction of 7. The reaction times in toluene, with or
without CD as the modifier, were 20 and 110 min, respec-
tively. Though many reactions were easy and the hydrogen
consumption ceased within 1–1.5 h, the thiophene ring in
lectivities (Table 1). In general, hydrogenations in the pres-
ence of cinchona alkaloid were faster than the correspond-
ing reactions over the unmodified catalyst. On the basis of
the reaction time necessary to achieve 50% conversion, the
highest rate acceleration (by a factor of 5.5) was measured
4
strongly poisoned the catalyst and high conversion could
not be achieved, even after prolonged reaction time and
at higher hydrogen pressure. A comparison of the results
obtained with 3 and 4 clearly demonstrates the detrimental
effect of the sulphur heteroatom.
TABLE 1
Enantioselective Hydrogenation of Substituted Trifluoromethyl
Ketones under Standard Conditions
Although in the experiments shown in Table 1 only two
solvents and two chiral modifiers were tested, 60–70% ee
was achieved in the hydrogenation of the substrates 3 and 5.
Optimization of reaction conditions and modifier structure
has not yet been attempted, but in two cases some more
reaction parameters were varied. In the hydrogenation of
Toluene + Toluene +
Acetic
CD � HCl acid + CD
�
N
Substrate
CD
62a
70a
ee (% )
conv. (% )
time (h)
6
43
1.5
1
2
3
4
5
6
7
100
1.5
100
1.5
ee (% )
conv. (% )
time (h)
10
98
3
18
83
2
5
100
2
1
the surface hydrogen concentration on Pt was reduced
by increasing the amount of catalyst (i.e., operating the re-
actor in the mass transport limited region), which resulted
in a small increase in ee from 70 to 74% in dichloroben-
zene. This value represents the highest ee achieved in this
reaction with a solid catalyst (9–11). In the hydrogenation
of 5, replacing CD by MeOCD enhanced the ee from 70%
to 90% in acetic acid. Further improvement by varying the
pressure or the amount of modifier failed, as illustrated in
Figs. 1 and 2. These additional experiments, together with
the results in Table 1, indicate that cinchona-modified Pt
is an efficient catalyst for the synthesis of trifluoromethyl
alcohols, albeit the best conditions vary for each substrate
and optimization may require considerable further effort.
ee (% )
conv. (% )
time (h)
56
93
2
60
81
3.5
30
43
3
0b
14
3
3b
2b
ee (% )
conv. (% )
time (h)
6
3
7
3
ee (% )
conv. (% )
time (h)
49
100
1
42
100
1
70
100
1.5
ee (% )
conv. (% )
time (h)
5
40
3
18c
—
—
—
18c
—
—
—
ee (% )
conv. (% )
time (h)
6
70
4
100
1
100
1.5
4
. DISCUSSION
a
Solvent: 1,2-dichlorobenzene, 3.6 mmol of substrate, 2 bar hydrogen
pressure.
It was proposed earlier (15) that the enantioselec-
tive hydrogenation over cinchona-modified Pt is limited
to trans-�-dicarbonyl compounds such as �-ketoesters or
b
5
0 bar hydrogen pressure.
c
�
Reaction at 0 C.

ENANTIOSELECTIVE HYDROGENATION
163
that the more flexible, nonplanar structure of 2 influences
both the adsorption on the Pt surface and the interaction
with the modifier. These examples demonstrate that
steric as well as electronic effects determine the enantio-
differentiation over the Pt–cinchona system.
The examples in Table 1 provide some information
concerning the substrate–modifier interaction during the
enantio-differentiating step. On the basis of the effect of
replacing toluene by acetic acid, the substrates may be di-
vided into two groups. In acetic acid the ee dropped dra-
matically in all reactions except in the hydrogenation of
5
, where the ee increased by 21% . It is not clear yet what
is the role of protonation of the quinuclidine N atom by
acetic acid in the substrate–modifier interactions. Another
intriguing observation is the unexpectedly high ee obtained
by replacing CD with MeOCD in the hydrogenation of 5.
For comparison, in the hydrogenation of �-ketoesters with
the same catalyst the difference between the ee’s obtained
in toluene or acetic acid, and with CD or MeOCD, was only
FIG. 2. Influence of the amount of modifier MeOCD on the enan-
tiomeric excess in the hydrogenation of 5 under standard conditions in
acetic acid.
2
–3% (14). Besides, ab initio calculations indicated that the
conformations of CD and MeOCD are similar (dominance
of “open 3” conformation (19)). It was proposed earlier
�
-diketones. The successful hydrogenation of 1 (9, 10)
and ketopantolactone (16) was the first piece of evidence
against this postulate, indicating that the real requirement
for the substrate is the presence of an electron-withdrawing
group in the � position. The recent highly enantioselec-
tive hydrogenation of �-ketoacetals (17, 18) provides ad-
ditional evidence in this direction.
The results in Table 1 allow further refinement con-
cerning the role of an electron-withdrawing group in the
substrate. Ethyl trifluoropyruvate 6 possesses two strong
electron-withdrawing groups on both sides of the carbonyl
group. Nevertheless, hydrogenation of this compound af-
forded very poor results. Hydrogen uptake ceased at mod-
erate conversion and the ee was very low. A possible reason
for this unexpected behavior is that the highly activated car-
bonyl compound reacts with the modifier, as evidenced by
(
9, 10, 13) that in the enantio-differentiating step the quin-
uclidine N atom of CD interacts with the O atom of the keto
carbonyl group of the substrate via a H bond (N–H–O-type
interaction), and the OH group of CD is not involved in this
interaction. The significant positive effect of replacing the
hydroxy group of CD by a methoxy group can be explained
by the importance of another type of substrate–modifier
interaction via the OH group which is competitive enough
to diminish the ee.
5
. CONCLUSIONS
The enantioselective hydrogenation of various trifluo-
romethyl ketones shown in this work demonstrates the po-
tential of cinchona-modified Pt for the synthesis of chiral-
fluorinated alcohols. Additional functional groups, such as
an ester function in the � position or amines, are not detri-
mental to enantio-differentiation. Hydrogenation of triflu-
oromethyl ketones represents a new example where up
to 90% ee can easily be achieved with a chirally modi-
fied metal hydrogenation catalyst. The poor ee’s obtained
with 3-phenyl-1,1,1-trifluoropropan-2-one (2) and 1,1,1-
trifluoroacetone (7) provide useful hints concerning the im-
portance of both electronic and steric effects in the enantio-
differentiating step. These observations, together with the
unexpectedly high effects of an acidic solvent and the pro-
tection of the OH group of CD, are good starting points for
future mechanistic studies.
1
13
NMR spectroscopy. The H-NMR and C-NMR spectra of
the mixture of 6 and CD could not be interpreted due to
19
overlapping of the signals, but the F signal of 6 at � 76 ppm
disappeared in favor of a new main peak at � 81.5 ppm.
When 2-propanol was used to mimick the secondary OH
function of CD, the rapid and complete conversion of 6 to
the corresponding hemiketal was confirmed by H-, C-
and 19F-NMR (with a � 76- to � 82-ppm shift in the latter
case). Apparently, the hemiketal of 6 and CD is a poor
modifier in the enantioselective hydrogenation of 6.
In case of 1,1,1-trifluoroacetone (7) the steric differ-
ence between the CH3 and the CF3 group is very small,
which can explain why the enantio-differentiation is so
moderate in this reaction. The structure of 3-phenyl-1,1,1-
trifluoropropan-2-one (2) is very similar to that of 1. The
only difference is an additional methylene group between
the carbonyl group and the aromatic ring in 2, which re-
sulted in a strong negative effect on the ee. It is assumed
1
13
REFERENCES
1. Soloshonok, V. A., Ed., “Enantiocontrolled Synthesis of Fluoro-
Organic Compounds.” Wiley, Chichester, 1999.

1
64
VON ARX, MALLAT, AND BAIKER
2
3
4
5
6
7
. Ramachandran, P. V., Teodorovic, A. V., and Brown, H. C., Tetrahe-
dron 49, 1725 (1993).
. Ramachandran, P. V., Chen, G. M., Lu, Z. H., and Brown, H. C., Tetra-
hedron Lett. 37, 3795 (1996).
. Nakamura, K., Matsuda, T., Itoh, T., and Ohno, A., Tetrahedron Lett.
11. T o¨ r o¨ k, B., Balazsik, K., Sz o¨ ll o¨ si, G., Felf o¨ ldi, K., and Bartok, M., Chi-
rality 11, 470 (1999).
12. Borszeky, K., B u¨ rgi, T., Zhaouhui, Z., Mallat, T., and Baiker, A.,
J. Catal. 187, 160 (1999).
13. Baiker, A., J. Mol. Catal. A: Chem. 115, 473 (1997).
14. Blaser, H.-U., Jallet, H. P., M u¨ ller, M., and Studer, M., Catal. Today
37, 441 (1997).
15. Simons, K. E., Meheux, P. A., Griffiths, S. P., Sutherland, I. M.,
Johnston, P., Wells, P. B., Carley, A. F., Rajumon, M. K., Roberts,
M. W., and Ibbotson, A., Rec. Trav. Chim. Pays-Bas 113, 465 (1994).
3
7, 5727 (1996).
. Fujisawa, T., Onogawa, Y., Sato, A., Mitsuya, T., and Shimizu, M.,
Tetrahedron 54, 4267 (1998).
. Noyori, R., “Asymmetric Catalysis in Organic Synthesis.” Wiley, New
York, 1994.
. Ohkuma, T., Koizumi, M., Doucet, H., Pham, T., Kozawa, M., Murata, 16. Sch u¨ rch, M., Schwalm, O., Mallat, T., Weber, J., and Baiker, A.,
K., Katayama, E., Yokozawa, T., Ikariya, T., and Noyori, R., J. Am.
Chem. Soc. 120, 13529 (1998).
J. Catal. 169, 275 (1997).
17. T o¨ r o¨ k, B., Felf o¨ ldi, K., Balazsik, K., and Bartok, M., Chem. Commun.
8
9
. Ohkuma, T., J. Syn. Org. Chem. Jpn. 57, 667 (1999).
. Mallat, T., Bodmer, M., and Baiker, A., Catal. Lett. 44, 95 (1997).
1725 (1999).
18. Studer, M., Burkhardt, S., and Blaser, H.-U., Chem. Commun. 1727
(1999).
1
0. Bodmer, M., Mallat, T., and Baiker, A., in “Catalysis of Organic Re-
actions” (F. E. Herkes, Ed.), p. 75. Marcel Dekker, New York, 1998.
19. B u¨ rgi, T., and Baiker, A., unpublished results.