Enantioselective hydrogenation of ethyl pyruvate over Pt/alumina modified by (R)-1-(1-naphthyl)ethylamine derivatives

Article abstract of DOI:10.1006/jcat.1996.0144

A new chiral modifier, (R)-1-(1-naphthyl)ethylamine, has been tested in the enantioselective hydrogenation of ethyl pyruvate to ethyl lactate over 5 wt% Pt/Al₂O₃. The influence of catalyst (2-28 g liter^-1) and modifier concentration (0-20 mM), temperature (282-333 K), pressure (1-75 bar), and solvents was studied in a slurry reactor. The 82% enantiomeric excess (ee) at full conversion was achieved in acetic acid after optimizing the reaction parameters. Under mild conditions the new modifier affords ee's better than that achieved with cinchona alkaloids. A drop in ee at pressures higher than 10 bar and temperatures above 288 K is attributed to partial hydrogenation of the naphthalene ring, which hinders the adsorption of the modifier parallel to a flat Pt surface. Maximum rate acceleration by a factor of 12, compared to the racemic reaction, was observed after thermal treatment of the catalyst in flowing hydrogen at 673 K, followed by aerobic treatment at 273-298 K in acetic acid. It is shown that naphthylethylamine is only a precursor of the actual modifier, which is a secondary amine formed in situ from naphthylethylamine and ethyl pyruvate by condensation to the corresponding imine and subsequent reduction of the C=N bond. Several other derivatives of naphthylethylamine were prepared by reductive alkylation and tested as modifiers. The results indicate that the presence of an oxygen function such as a hydroxy or methoxy group, as in previously used modifiers, is not an indispensable requirement for obtaining high ee in the hydrogenation of α-ketoesters.

Full text of DOI:10.1006/jcat.1996.0144

JOURNAL OF CATALYSIS 160, 261–268 (1996)
ARTICLE NO. 0144
Enantioselective Hydrogenation of Ethyl Pyruvate over Pt/Alumina
Modified by (R)-1-(1-Naphthyl)ethylamine Derivatives
,1
ꢀ
ꢀ
ꢀ
ꢀ
B. Minder, M. Schu¨rch, T. Mallat, A. Baiker, T. Heinz,† and A. Pfaltz†
ꢀ
Department of Chemical Engineering and Industrial Chemistry, ETH Zentrum, CH-8092 Zu¨rich, Switzerland; and †Institute of Organic Chemistry,
University of Basel, St. Johanns-Ring 19, CH-4056 Basel, Switzerland
Received August 18, 1995; revised December 11, 1995; accepted January 22, 1996
line ring, or acylation of the OH group resulted in a substan-
tial decrease or a complete loss of enantiodifferentiation
(7). Other members of the cinchona family (cinchonine,
quinidine, quinine) (1) as well as various other nitrogen
compounds of natural origin (ephedrine, strychnine, pro-
line, codeine) were found to be less effective or hardly ef-
fective at all (3, 8, 9).
A new chiral modifier, (R)-1-(1-naphthyl)ethylamine, has been
tested in the enantioselective hydrogenation of ethyl pyruvate to
ethyl lactate over 5 wt% Pt/Al2O3. The influence of catalyst (2–28 g
literꢀ1) and modifier concentration (0–20 mM), temperature (282–
333 K), pressure (1–75 bar), and solvents was studied in a slurry
reactor. The 82% enantiomeric excess (ee) at full conversion was
achieved in acetic acid after optimizing the reaction parameters.
Under mild conditions the new modifier affords ee’s better than
that achieved with cinchona alkaloids. A drop in ee at pressures
higher than 10 bar and temperatures above 288 K is attributed to
partial hydrogenation of the naphthalene ring, which hinders the
adsorption of the modifier parallel to a flat Pt surface. Maximum
rate acceleration by a factor of 12, compared to the racemic reac-
tion, was observed after thermal treatment of the catalyst in flowing
hydrogen at 673 K, followed by aerobic treatment at 273–298 K in
acetic acid. It is shown that naphthylethylamine is only a precur-
sor of the actual modifier, which is a secondary amine formed in
The use of cinchonidine derivatives revealed the follow-
ing crucial structural elements of modifiers 1a–1c (7, 10–12):
(i) A flat aromatic (quinoline) ring system is neces-
sary for the adsorption at the surface Pt0 sites (“anchoring”
moiety).
(ii) The tertiary quinuclidine nitrogen is essential for the
interaction with the organic reactant.
(iii) The stereogenic centers at C-8 and C-9 are respon-
sible for the asymmetric induction. A further conclusion
–
–
–
situ from naphthylethylamine and ethyl pyruvate by condensation was that the N-1 C-8 C-9 O structural unit is important
==
to the corresponding imine and subsequent reduction of the C
N
for enantioselection, but this will be disputed in this paper.
bond. Several other derivatives of naphthylethylamine were pre-
pared by reductive alkylation and tested as modifiers. The results
indicate that the presence of an oxygen function such as a hydroxy
or methoxy group, as in previously used modifiers, is not an indis-
pensable requirement for obtaining high ee in the hydrogenation of
Instead of modifying the structure of the cinchona al-
kaloids, we recently synthesized various structurally sim-
ple amino-alcohols possessing the crucial structural ele-
ments mentioned above (13, 14). These studies led to
a new efficient modifier, (R)-2-(1-pyrrolidinyl)-1-(1-naph-
thyl)ethanol (2, Scheme 1), which induces enantiomeric ex-
cesses of up to 75% in the hydrogenation of EP (13, 15).
The structure of 2 can be easily modified in a variety of
ways. This facilitates a systematic study of the platinum–
modifier–reactant interaction and the mechanism of enan-
tioselection.
On the basis of these results, various commercially avail-
able nitrogen-containing chiral compounds were tested as
modifiers for the asymmetric hydrogenation of EP (16).
Preliminary screening showed that a great variety of aro-
matic and heteroaromatic ring systems can function as an-
choring moiety of the modifier. Moreover, not only tertiary
nitrogen compounds, but also primary amines and iso-
cyanates afforded 29–67% ee in the hydrogenation of EP. A
simple aromatic amine, (R)-1-(1-naphthyl)ethylamine (3,
Scheme 1) was found to be the most effective modifier.
ꢀ-ketoesters.
ꢁc 1996 Academic Press, Inc.
INTRODUCTION
Various chiral amino-alcohol-type modifiers of natural
origin have been tested in the enantioselective hydrogena-
tion of ꢀ-ketoesters to the corresponding ꢀ-hydroxyesters
(1–3). The most studied reaction is the hydrogenation of
ethyl pyruvate (EP) to (R)-ethyl lactate (Scheme 1) over
supported platinum catalysts (4–6). Cinchonidine and two
of its derivatives (1a–1c, Scheme 1) provided enantiomeric
excesses (ee) up to 95% in acetic acid at 100 bar (6). Fur-
ther modifications of cinchonidine, such as alkylation of the
quinuclidine nitrogen, partial hydrogenation of the quino-
1 To whom correspondence should be addressed.
261
0021-9517/96 $18.00
c
Copyright ꢁ 1996 by Academic Press, Inc.
All rights of reproduction in any form reserved.

262
MINDER ET AL.
TABLE 1
A Comparison of Chiral Modifiers in the Enantioselective Hy-
drogenation of Ethyl Pyruvate (Acetic Acid, Room Temperature,
1h Reaction Time, Reactant : Catalyst = 106 Weight Ratio, Reac-
tant : Modifier = 1500 Molar Ratio)
Modifier
Pressure (bar) Conversion (%) ee (%)
75–72
25–22
48
54
51 (R)
55 (R)
3
4
75–72
25–22
52
59
53 (S)
58 (S)
5
6
25–21
83
55 (R)
SCHEME 1
Enantioselectivities up to 67% and a rate acceleration by a
factor of 6, compared to the unmodified reaction, were ob-
served in acetic acid on Pt/alumina without optimization.
In other solvents, such as ethanol or toluene, enantioselec-
tivities were moderate or low.
Here we present a detailed analysis of the hydrogenation
of EP over Pt/alumina modified with 3. In addition, various
derivatives of 3 were prepared and tested as modifiers.
75–72
25–22
70
68
54 (S)
58 (S)
75–72
25–21
62
78
51 (S)
58 (S)
7
8
75–71
25–22
75
60
39 (S)
40 (S)
EXPERIMENTAL
Synthesis of Modifiers
75–73
25–23
41
39
23 (S)
36 (S)
General. All solvents were distilled before use. The re-
actions were carried out under argon atmosphere using
dried glassware. Flash column chromatography used silica
gel C 560, 35–70 ꢁm, Chemische Fabrik Uetikon, and TLC
used silica gel 60, Merck, 0.25 mm. Specific rotations were
measured on a Perkin Elmer 241 polarimeter at room tem-
perature, with estimated error ꢂ 5%.
9
75–73
25–22
33
47
34 (S)
45 (S)
10
Synthesized modifiers are designated with numbers and
their corresponding formulae are summarized in Table 1.
Analytical data of the synthesized materials are available
from the authors on request.
75–73
25–23
45
38
40 (R)
37 (R)
11a
11b
The methods used for the synthesis of the modifiers are
described here.
75–73
25–23
32
48
31 (R)
49 (R)
Method A. A mixture of 5.8 mg (34 ꢁmol) of (S)-1-
(1-naphthyl)ethylamine 4, 50 mg of 5 wt% Pt/alumina cata-
lyst, and 1.2 g (10.3 mmol) of EP in 2 ml acetic acid
was stirred at 25 bar H2 for 1 h. After filtration, the sol-
vent was removed by distillation (100ꢃC, 20 mbar). The
residue was dissolved in ether, K2CO3 was added, and
the mixture was stirred for 1 h. Filtration, removal of
the solvent in vacuo followed by flash chromatography
with hexane/EtOAc (5 : 1) afforded 3.7 mg (40%) of (2S,
10S)-N-[10-(1-naphthyl)ethyl]-2-amino propionic acid ethyl
75–74
25–22
19
19 (S)
9 (S)
12
53a
a Reaction time, 3 h.

ENANTIOSELECTIVE HYDROGENATION OF ETHYL PYRUVATE
263
ester 6. Method A was also used for the preparation of removed in vacuo. Purification by flash chromatography
(S)-N-benzyl-1-(1-naphthyl)ethylamine10(yellowoil, 34% with EtOAc afforded 363.3 mg (84%) of the carbamate as
yield) and (10R)-N-[10-(1-naphthyl)ethyl]-3-amino butyric a yellow solid.
acid methyl ester 11 (diastereomer 11a as yellow oil, 131.6 The carbamate (297.6 mg, 1.22 mmol) in 3.5 ml THF was
mg, 28% yield; diastereomer 11b as yellow oil, 30.0 mg, 6% added dropwise to a stirred suspension of 294.4 mg (7.76
yield).
mmol) of LiAlH4 in 7 ml THF at 0ꢃC. The reaction mixture
was warmed and refluxed for 14 h. After cooling to room
temperature excess LiAlH4 was quenched with 0.5 ml wa-
ter, and then NaOH solution (15%, 0.5 ml) and 0.16 ml
water were added. The precipitate was filtered and washed
with ether. The filtrate was concentrated in vacuo and chro-
matographed with ethanol/NEt3 (100 : 1) to give (S)-[1-(1-
naphthyl)ethyl]methylamine 8 (203.7 mg, 90%) as a pale
yellow oil.
Method B (17). A mixture of 617 ꢁl (5.57 mmol) of EP,
954 mg (5.57 mmol) of amine 4, and 1.68 g (7.35 mmol) of
titanium (IV) ethoxide was stirred at room temperature for
1 h. The viscous solution was diluted with 5.5 ml ethanol.
Sodium cyanoborohydride (274 mg, 3.7 mmol) was added,
and the red solution was stirred for 16 h. After the addi-
tion of 1.1 ml water, the resulting precipitate was filtered
and washed with ethanol. The filtrate was concentrated in
vacuo, dissolved in ethyl acetate, filtered to remove the re-
Method D (19). To a solution of 186 ꢁl (1.17 mmol)
maining inorganic solids, and concentrated in vacuo again. of amine 4 in 5 ml hexamethylphosphorous triamide,
The two diastereomers 6 and 7 were separated by flash chro- 196.1 mg (2.34 mmol) of sodium bicarbonate and 465.4
matography (hexane/EtOAc 5 : 1). Each diastereomer was mg (1.17 mmol) of 1,4-butanediol ditosylate were added.
rechromatographed with CH2Cl2 as eluent. (2S,10S)-N-[10- The reaction mixture was then stirred at 130ꢃC for 30 h.
(1-naphthyl)ethyl]-2-amino propionic acid ethyl ester 6 was After cooling to room temperature EtOAc (20 ml) was
obtained as pale yellow oil (387 mg, 26% yield); (2R, 10S)- added. The organic phase was washed five times with
N-[10-(1-naphthyl)ethyl]-2-amino propionic acid ethyl ester water, dried over K2CO3, and concentrated in vacuo. After
7 as yellow oil (324 mg, 22% yield).
flash chromatography with EtOAc/NEt3 (100 : 1) (S)-N-[1-
The absolute configuration of the newly formed stere- (1-naphthyl)ethyl]pyrrolidine 12 (138.1 mg, 52%) was ob-
ogenic center was determined by GC analysis of the corre- tained as a yellow oil.
sponding N-Boc-alanine derivative. A mixture of 68.4 mg
Catalytic Hydrogenation
(0.4 mmol) of (S)-1-(1-naphthyl)ethylamine 4, 100 mg of
5 wt% Pt/alumina catalyst, and 1.2 g (10.3 mmol) of ethyl
pyruvate in 4 ml acetic acid was stirred at 25 bar H2 for
The commercial modifiers (R)-1-(1-naphthyl)ethylamine
(3, Fluka) and (S)-1-(1-naphthyl)ethylamine (4, Fluka)
1 h. After filtration, the solvent was removed by distilla-
were used without further purification. Ethyl pyruvate
tion (100ꢃC, 20 mbar). The residue was dissolved in ether,
(Aldrich) was freshly distilled under vacuum before each
K2CO3 was added, and the mixture was stirred for 1 h. After
reaction. A 5 wt% Pt/alumina catalyst (Engelhard 4759)
filtration and removal of the solvent in vacuo the crude sec-
with a metal dispersion of 22% (determined by CO
chemisorption (20)) was used in all experiments.
The hydrogenation was carried out in a 100-ml stainless-
steel autoclave (Baskerville) with a 50-ml glass liner and
ondary amine 6 was obtained (95% diastereomeric excess
by HPLC). A mixture of the unpurified amine 6 and 200
mg of palladium hydroxide on charcoal in 3 ml ethanol was
stirred at 25 bar H2 for 16 h. After filtration, 0.3 ml of NEt3
PTFE cover. Under standard conditions, 100 mg catalyst,
and 143.5 mg (0.66 mmol) of di-t-butyl dicarbonate were
6.2 ꢁmol modifier, 10 ml (0.09 mol) EP, and 20 ml acetic acid
added. Removal of the solvent in vacuo followed by flash
were used. The reaction mixture was magnetically stirred
at 1250 rpm under 10 bar hydrogen at 298 K. Before use
chromatography with hexane/EtOAc (2 : 1) afforded 27.2
mg (31%) of 6b as a colorless oil. The absolute configura-
the catalyst was prereduced at 673 K for 1.5 h in a hydrogen
flow of 30 ml minꢄ1 and transferred into the reactor under
solvent with the exclusion of oxygen. The catalyst was first
contacted with the solvent containing the proper amount of
modifier. Reaction temperatures represent the controlled
temperature of the bath into which the autoclave was im-
tion of 6b was determined by GC analysis with a chiral col-
umn (heptakis-(2,3,6-tri-O-methyl)-ꢂ-cyclodextrin in OV
1701-vinyl).
Method B was also used for the preparation of (S)-[1-(1-
naphthyl)ethyl]isopropylamine 9 (colorless oil, 35% yield).
Method C (18). A solution of 280 ꢁl (1.77 mmol) of mersed.
amine 4 in 11 ml toluene and 0.65 ml pyridine was cooled to
The highest initial rate was achieved after the follow-
0ꢃC. Ethyl chloroformate (210 ꢁl, 2.2 mmol) in 1 ml toluene ing oxidative pretreatment: 100 mg prereduced catalyst was
was added dropwise over 15 min. The reaction mixture was mixed in 20 ml acetic acid in air for 1 h in an ice bath (by the
warmed to room temperature and stirred for 4 h. After end of this period the solvent was solid!); 12.4 ꢁmol modi-
quenching with 10 ml of 2 N KOH, the aqueous layer was fier 6 was added and the slurry was mixed at 298 K for 2 h
extracted three times with toluene. The combined organic in air, then EP was added and the hydrogenation reaction
extracts were dried over MgSO4 and the solvent was was performed at 10 bar.

264
MINDER ET AL.
The initial reaction rate was calculated from the hydro-
gen consumption measured by a gas flow controller (Bu¨chi
BPC 9901). The enantiomeric excess and conversion were
determined by a HP 5890A gas chromatography, using
a chiral WCOT Cyclodextrin-ꢂ-2,3,6-M-19 (Chrompack)
capillary column. The enantioselectivity is expressed as ee
ꢀ
ꢀ
ꢀ
ꢀ
(%) = 100 ꢅ ( [R]ꢄ[S] )/([R] + [S]).
It was found that under standard conditions ee was within
a range of ꢂ1%, independent of the conversion (between
32 and 100%). This fluctuation is within the range of the
standard deviation of the analytical method.
RESULTS
Influence of Reaction Conditions
A 5 wt% Pt/alumina catalyst modified with (R)-1-(1-
naphthyl)ethylamine (3, Scheme 1) was used in acetic acid
in the enantioselective hydrogenation of ethyl pyruvate to
(R)-ethyl lactate. The influence of catalyst concentration
is shown in Fig. 1. The modifier : catalyst weight ratio was
kept constant at 0.011, which corresponds to 1.1 molecule
of modifier per surface platinum atom (Pts). There exists a
linear relationship between reaction rate and catalyst load-
ing below about 14 g literꢄ1. The deviation from linearity
above this value is due to mass transport limitations in the
slurry reactor. Accordingly, a catalyst loading of 3.6 g literꢄ1
was used in further experiments which ensured the reactor
was working in the kinetic region. The enantioselectivity
FIG. 2. Initial rate (᭹) and enantiomeric excess (᭺) as a function of
modifier concentration in acetic acid under otherwise standard conditions.
reached its broad maximum above 7 g literꢄ1. There seems
to be no clear correlation between ee and the hydrogen
supply. Both high hydrogen supply (low catalyst loading)
and low hydrogen supply (high catalyst loading) result in
lower ee’s.
The influence of the modifier concentration on ee and ini-
tial rate is illustrated in Fig. 2. Both rate and ee increase with
ascending modifier concentration and reach a broad maxi-
mum between 0.2 and 20 mM. For the lower value, the cor-
responding modifier : reactant molar ratio is 1 : 15000 and
the modifier : Pts molar ratio is 1 : 1. For comparison, when
cinchona alkaloids are used as modifiers for Pt/alumina in
the same reaction, the modifier : Pts ratio at the maximum
in ee is less than 0.1 in acetic acid and 0.25 in toluene (6, 21).
The remarkably strong effect of temperature is shown
in Fig. 3. The initial rate increases with ascending temper-
ature but reaches a plateau at 318 K, whereas selectivity
decreases rapidly above 288 K. Similarly, a hydrogen pres-
sure higher than 10 bar has a detrimental effect on ee, with-
out a considerable influence on the reaction rate (Fig. 4).
We propose that the negative effect at pressures higher than
10 bar and temperatures higher than 288 K is due to the par-
tial hydrogenation of the naphthalene ring of modifier 3. A
similar phenomenon was observed in the hydrogenation of
EP when Pt/alumina was modified with 2 (15). The partial
saturation of the unsubstituted naphthalene ring (as identi-
fied by NMR analysis (15)) weakens the adsorption of the
modifier on Pt and results in a loss in ee. The higher the tem-
perature and/or the hydrogen pressure during reaction, the
higher is the extent of this undesired side reaction. Under
FIG. 1. The influence of catalyst concentration on the initial rate (᭹)
and enantiomeric excess (᭺) under otherwise standard conditions.

ENANTIOSELECTIVE HYDROGENATION OF ETHYL PYRUVATE
265
FIG. 4. Initial rate (᭹) and enantiomeric excess (᭺) as a function of
hydrogen pressure under otherwise standard conditions.
FIG. 3. Influence of reaction temperature on the enantiomeric excess
(᭺) and initial rate (᭹) under otherwise standard conditions.
no data could be found in the literature, according to which
75% or higher ee can be achieved with cinchonidine or its
derivatives at pressures of 10 bar or below (25). The best
enantioselectivity reported to date for the Pt–1b system is
92% ee (95% ee for 1c) at 100 bar (6).
mild conditions the stability of the modifier is high enough
to provide good ee. Note that Pt is one of the best noble
metal catalysts for the hydrogenation of an aromatic ring
under relatively mild conditions (22).
The choice of solvent is of crucial importance, as illus-
trated in Table 2. In contrast to pyruvate hydrogenations
over Pt modified with 1 or 2 (15, 23), modifier 3 is only ef-
fective in acetic acid. Other polar solvents, such as ethanol
or water, and apolar solvents like toluene resulted in low
ee and moderate reaction rate.
Modifier–Reactant Interaction
An interesting feature of this enantioselective hydro-
genation reaction is that the primary amino group of naph-
thylethylamine 3 or 4 and the keto group of EP, which is
activated by the neighboring ester group, react with each
other under reaction conditions (16, 26). The resulting
It is seen in Figs. 1–4 and Table 2 that the initial rate
of pyruvate hydrogenation is accelerated in the presence of
modifier3byafactorofupto6, comparedtotheunmodified
reaction (120 mmol gꢄ1 ꢄ1). This effect is comparable to
h
TABLE 2
the rate acceleration observed in the presence of cinchona
alkaloids or 2 as modifiers (7, 15, 24).
Influence of Solvent on Initial Rate (r0) and Enantiomeric Excess
(ee) in the Enantioselective Hydrogenation of Ethyl Pyruvate on
Pt/Alumina Catalyst Modified with 3 or 6 (Standard Conditions)
Optimization of a limited number of parameters (pres-
sure, catalyst : modifier and catalyst : reactant ratios) has
been attempted at a fixed temperature of 282 K, using fac-
torial design and gradient method. At the optimum, 82%
ee was achieved at full conversion. The optimum condi-
tions were : 8 bar, 13 g literꢄ1 catalyst concentration, cat-
alyst : reactant weight ratio = 0.038, Pts : modifier 3 molar
ratio = 8.1. At the temperature used, the influence of pres-
sure below 10 bar was negligible. For comparison, the same
Pt/alumina catalyst afforded 73% ee at 1 bar and 87% ee
at 75 bar in the presence of 10,11-dihydrocinchonidine 1b
under similar conditions (13). It is worth mentioning that
ee (%)
r0 (mmol gꢄ1 hꢄ1
)
Solvent
3
6
3
6
Acetic acid
Ethanol
67
29
22
10
13
10
28
68
50
—
9
—
—
—
730
125
215
160
127
260
204
770
330
—
310
—
Ethanol + acetic acida
Toluene
Toluene + acetic acida
Water
—
—
Acetic acid : water = 3 : 1(v/v)
a 35 ꢁmol acetic acid.

266
MINDER ET AL.
clear.) The pyrrolidine derivative 12 was prepared by alky-
lation with 1,4-butanediol ditosylate. None of these modi-
fiers can compete with 6 or 7 (Table 1).
Although reductive amination of 3 with EP is fast un-
der the hydrogenation conditions in acetic acid, this is
not necessarily the case in other solvents. We compared
the efficiency of modifiers 3 and 6 in different solvents
(Table 2). As mentioned above, both compounds provide
essentially the same ee’s and initial rates in acetic acid at
medium pressures. The ee’s in toluene are very low with
both compounds; however, in ethanol 3 and 6 provide dis-
tinctly different ee’s (29 vs 50% ee), indicating the reduc-
tive amination leading to 6 is slower than hydrogenation of
EP in this solvent. The reason for the low ee measured in
toluene is not yet clear. Interestingly, the enantioselectivi-
ties observed with modifiers 1a, 1b, or 2 in acetic acid and
toluene were comparable (within a 10% range) (6, 15).
SCHEME 2
Catalyst Pretreatment
Orito et al. (29) reported first that catalyst pretreatment
can substantially improve the ee in the hydrogenation of
ꢀ-ketoesters. On the basis of this observation, we also tested
the effect of various reductive and oxidative catalyst pre-
treatment procedures at 273–673 K. Using modifier 6, none
of these procedures provided higher ee than 68% (stan-
dard hydrogenation conditions, Table 2). However, oxida-
tive treatment of the prereduced catalyst in acetic acid at
273–298K, inthepresenceorabsenceofmodifier, increased
the initial rate from 770 to 1060–1550 mmol gꢄ1hꢄ1 (see Ex-
perimental section for descriptions of the best conditions).
To examine the effect of pretreatment on the catalyst
structure, an in situ FTIR study of 5 wt% Pt/alumina
was carried out. This study, completed with temperature-
programmed oxidation experiments, indicated that there is
a considerable cleaning of the catalyst surface during pre-
hydrogenation at 673 K. The removal of organic species
(presumably residues from the Pt-precursor) is accompa-
nied by CO generation resulting from a decarbonylation
reaction catalyzed by Pt at this elevated temperature. An
oxidative treatment at 273–298 K (after prehydrogenation
at673K)completesthepurificationoftheactivesites, which
is shown by the increase of the initial rate of EP hydrogena-
tion by 50–100%. More details of the FTIR analysis will be
published elsewhere.
imine is subsequently reduced on the Pt surface, affording
a secondary amine (Scheme 2). This secondary amine is the
actual active modifier in the hydrogenation of EP. Further
reductive alkylation with ethyl pyruvate to a tertiary amine
would require elevated temperature (27) and does not take
place under reaction conditions.
NMR analysis of the reaction mixture showed that naph-
thylethylamine 3 is quantitatively consumed during the hy-
drogenation reaction and converted to the secondary amine
5 (Table 1) and minor amounts of unidentified products
(presumably di- or tetrahydronaphthalene derivatives).
When modifier 5 or the corresponding (S)-enantiomer 6
was synthesized, purified, and used as a modifier in the hy-
drogenation of EP, the enantioselectivity was essentially the
same in acetic acid and toluene as those observed with 3 and
4 (Table 1). Interestingly, the formation of 5 and 6 from 3
and 4, respectively, is a highly diastereoselective reaction
(95% diastereomeric excess; Scheme 2) (26, 28). The ab-
solute configuration of the newly formed stereogenic cen-
ter was determined by transformation to the (S)-alanine
derivative 6b.
Using a different method for reductive alkylation of 4
with EP allowed the preparation of both diastereoisomers
(S,S)-6 and (S,R)-7 (see Experimental section). Unexpect-
edly, both (S, S)-6 and (S,R)-7 induced the same enantiose-
lectivity in the hydrogenation of EP. A control experiment
showed that less than 3% isomerization of 7 to 6 occurs dur-
ing the reaction. This implies that the results obtained with 7
are significant and that the configuration at the stereogenic
DISCUSSION
The kinetic analysis of EP hydrogenation over Pt/
center in ꢀ position to the ester group has no influence on alumina, modified with (R)-1-(1-naphthyl)ethylamine (3,
the enantioselection of modifiers 6 and 7. Scheme 1) and the limited optimization of reaction condi-
Reductive alkylation of 4 with different aldehydes or ke- tions revealed that up to 82% ee can be achieved at 282 K
tones provides easy access to a variety of related modifers and at hydrogen pressures below 10 bar. At low pressure
such as 8–11. (Note that the absolute configuration of 11a this new modifier is more efficient than cinchonidine 1a or
and 11b is not known; only their relative configuration is 10,11-dihydrocinchonidine 1b.

ENANTIOSELECTIVE HYDROGENATION OF ETHYL PYRUVATE
267
Separation and identification of the modifier after EP ifier (secondary amine) is formed by reductive alkylation
hydrogenation supplied evidence for the formation of a of 3 with the reactant (EP). This reaction offers the op-
secondary amine 5 by reductive alkylation of the primary portunity of changing the modifier structure by apply-
amine 3 with EP. Formation of 5 on the prereduced Pt ing other carbonyl compounds instead of EP. In the low-
is assumed to precede the hydrogenation of EP, at least pressure range (10 bar) the ee reached with this new mod-
in acetic acid. Modifier 5 and various other derivatives of ifier (82%) is better than those reported with cinchona al-
3 have been synthesized and tested in EP hydrogenation kaloids and their derivatives. For use at high pressure the
(Table 1). From the structural analysis we can conclude that resistance of the anchoring group (naphthalene) against hy-
the presence of an N–C–C–O structural unit, characteristic drogenationneedstobe improved. The studies alsoindicate
of cinchona alkaloids and their derivatives, is not a neces- that the earlier proposal concerning the importance of the
sary requirement for enantiodifferentiation in the hydro- N–C–C–O structural unit (7) does not hold for the (R)-
genation of ꢀ-ketoesters. The distance between the O and 1-(1-naphthyl)ethylamine derived modifiers presented in
N atoms can be shorter or even much longer without los- this work.
ing enantioselection. Moreover, the ee can be substantial
even in the absence of an oxygen function in the modifier
(see, e.g., Table 1). These observations provide strong ev-
ACKNOWLEDGMENT
idence against a possible modifier–reactant interaction, in
which a nucleophilic attraction between the O-atom in the
modifier and the ester carbon atom in EP plays a crucial
role (30).
FinancialsupportofthisworkbytheSwissNationalScienceFoundation
(Program Chiral 2) is gratefully acknowledged.
REFERENCES
The results are in good agreement with our former pro-
posal concerning the nature of enantiodifferentiation in
the hydrogenation of EP in acetic acid (12, 31). The in-
teraction of the protonated N-base modifier with the car-
bonyl O-atom of EP via H-bonding and the sufficiently
strong (“fixed”) adsorption of the complex on the Pt-
surface provide the steric requirements of enantiodiffer-
entiation. Molecular modeling studies of methyl pyruvate
hydrogenation over Pt catalysts modified by cinchonidine
1a or (R)-2-(1-pyrrolidinyl)-1-(1-naphthyl)ethanol 2 can be
found elsewhere (12, 14).
The strong adsorption of the flat aromatic ring (“anchor-
ing” moiety) on a flat Pt surface is a basic requirement in
our model (there is substantial evidence for the adsorption
of naphthalene parallel to Pt(111) (32–34)). Any deviation
from this ideal structure is expected to diminish the enan-
1. Orito, Y., Imai, S., and Niwa, S., J. Chem. Soc. Jpn., 670 (1980).
2. Blaser, H. U., Tetrahedron: Asymmetry 2, 843 (1991).
3. Blaser, H. U., Chem. Rev. 92, 935 (1992).
4. Webb, G., and Wells, P. B., Catal. Today 12, 319 (1992).
5. Baiker, A., DGMK Tagungsber., No. 9305, p. 119, 1993.
6. Blaser, H. U., Jalett, H. P., and Wiehl, J., J. Mol. Catal. 68, 215
(1991).
7. Blaser, H. U., Jalett, H. P., Monti, D. M., Baiker, A., and Wehrli, J. T.,
Stud. Surf. Sci. Catal. 67, 147 (1991).
8. Tungler, A., Tarnai, T., Ma`the´, T., and Petro`, J., J. Mol. Catal. 70, L5
(1991).
9. Griffiths, S. P., Johnston, P., Vermeer, W. A. H., and Wells, P. B., J.
Chem. Soc., Chem. Comm., 2431 (1994).
10. Saus, A., Zimmermann, K., and Gu¨rtler, O., Chem.-Ztg. 115, 252
(1991).
11. Baiker, A., Mallat, T., Minder, B., Schwalm, O., Simons, K. E., and
Weber, J., in “Chiral Reactions in Heterogeneous Catalysis” (G.
Jannes and V. Dubois. Eds.), p. 95. Plenum, New York, 1995.
tioselectivity. The partial hydrogenation of the naphthalene 12. Schwalm, O., Minder, B., Weber, J., and Baiker, A., Catal. Lett. 23,
271 (1994).
ring at pressures higher than 10 bar, resulting in the for-
mation of (5,6,7,8)-tetrahydronaphthalene derivatives (15),
distorts the planar structure of the anchoring moiety and
lowers the ee. For future improvement of our modifiers, it
13. Wang, G., Heinz, T., Pfaltz, A., Minder, B., Mallat, T., and Baiker, A.,
J. Chem. Soc., Chem. Commun., 2047 (1994).
14. Simons, K. E., Wang, G., Heinz, T., Giger, T., Mallat, T., Pfaltz, A., and
Baiker, A., Tetrahedron: Asymmetry 6, 505 (1995).
will be crucial to improve the resistance of the anchoring 15. Minder, B., Mallat, T., Baiker, A., Wang, G., Heinz, T., and Pfaltz, A.,
J. Catal. 154, 371 (1995).
16. Minder, B., Schu¨rch, M., Mallat, T., and Baiker, A., Catal. Lett. 31, 143
aromatic ring against hydrogenation. In this way it should
be possible to achieve high enantioselectivity also at higher
pressure, as with cinchona alkaloids and their most effective
derivatives.
(1995).
17. Mattson, R. J., Pham, K. M., Leuck, D. J., and Cowen, K. A., J. Org.
Chem. 55, 2552 (1990).
18. Benson, S. C., Cai, P., Colon, M., Haiza, M. A., Tokles, M., and Snyder,
J. K., J. Org. Chem. 53, 5335 (1988).
CONCLUSION
19. Juaristi, E., and Madrigal, D., Tetrahedron 45, 629 (1989).
20. Wehrli, J. T., Baiker, A., Monti, D. M., and Blaser, H. U., J. Mol. Catal.
61, 207 (1990).
21. Garland, M., and Blaser, H. U., J. Am. Chem. Soc. 112, 7048 (1990).
22. Freifelder, M., “Practical Catalytic Hydrogenation.” Wiley–Inter-
science, New York, 1971.
(R)-1-(1-naphthyl)ethylamine 3 is a versatile and com-
mercially available precursor for the in situ preparation
of effective modifiers for the platinum-catalyzed enantio-
selective hydrogenation of ethyl pyruvate. In contrast to
the previously used systems, in this case the actual mod-
23. Wehrli, J. T., Baiker, A., Monti, D. M., Blaser, H. U., and Jalett, H. P.,
J. Mol. Catal. 57, 245 (1989).

268
MINDER ET AL.
29. Orito, Y., Imai, S., and Niwa, S., J. Chem. Soc. Jpn., 1118 (1979).
24. Wehrli, J. T., Ph.D. thesis, ETH No. 8833, Zu¨rich, 1989.
25. Blaser, H. U., Jalett, H. P., Monti, D. M., Reber, J. F., and Wehrli, J. T., 30. Augustine, R. L., Tanielyan, S. K., and Doyle, L. K., Tetrahedron:
Stud. Surf. Sci. Catal. 41, 153 (1988). Asymmetry 4, 1803 (1993).
26. Heinz, T., Wang, G., Pfaltz, A., Minder, B., Schu¨rch, M., Mallat, T., 31. Minder, B., Mallat, T., Skrabal, P., and Baiker, A., Catal. Lett. 29, 115
and Baiker, A., J. Chem. Soc., Chem. Commun., 1421 (1995). (1994).
27. Frahm, A. W., in “Methoden der Organischen Chemie,” Thieme, 32. Gland, J. L., and Somorjai, G. A., Surf. Sci. 38, 157 (1973).
Stuttgart, 1980. 33. Hallmark, V. M., and Chiang, S., Surf. Sci. 286, 190 (1993).
28. Harada, K., in “Asymmetric Synthesis, Vol. 5, Chiral Catalysis,” 34. Carley, A. F., Rajumon, M. K., Roberts, M. W., andWells, P. B., J. Chem.
Academic Press, San Diego, 1985. Soc. Faraday Trans. 91, 2167 (1995).