Electrospray ionization-mass spectrometry in the enantioselective hydrogenation of ethyl pyruvate catalyzed by dihydrocinchonidine modified Pt/Al2O3 in acetic acid

Article abstract of DOI:10.1006/jcat.2001.3423

The enantioselective hydrogenation of ethyl pyruvate (EtPy) on Pt-alumina (E 4759), Pt black, and Pt black + alumina (mixture) catalysts modified by dihydrocinchonidine (DHCD) in acetic acid was studied by electrospray ionization-mass spectrometry (ESI-MS). Application of the ESI-MS technique led to the recognition of a novel-type compound, O⁺[Al(OAc)₂]₃ (oxonium ions). The effect of the DHCD concentration, temperature, and oxonium cations on the reaction rate and the enantioselectivity has been studied. Using the Engelhard 4759 catalyst in acetic acid under mild experimental conditions (room temperature, hydrogen pressure 1 bar, DHCD concentration 0.01 mM/L) an optical yield of 92% can be achieved. The high enantioselectivity is accompanied by the following turnovers: EtPy/DHCD > 43,000, EtPy/Pt_surface > 1000, TOF = 1-2 s^-1, and DHCD/Pt_surface ratio = 0.0072. The enantioselectivity reducing factor is identified by ESI-MS as the gradual hydrogenation of the quinoline skeleton of DHCD that becomes more pronounced with increasing temperature and hydrogenation time. The discovery of oxonium cations, the extremely low DHCD/Pt_surface ratio, and the new data obtained by the Pt black + alumina mixture made possible an interpretation of the mechanism of the heterogeneous enantioselective hydrogenation of α-ketoesters.

Full text of DOI:10.1006/jcat.2001.3423

Journal of Catalysis 205, 168–176 (2002)
doi:10.1006/jcat.2001.3423, available online at http://www.idealibrary.com on
Electrospray Ionization–Mass Spectrometry in the Enantioselective
Hydrogenation of Ethyl Pyruvate Catalyzed by Dihydrocinchonidine
Modified Pt/Al2O3 in Acetic Acid
, ,1
Miha´ly Barto´k, † Katalin Bala´zsik, Gyo¨rgy Szo¨llo¨si, and Tibor Barto´k‡
Organic Catalysis Research Group of the Hungarian Academy of Sciences, Do´m te´r 8, H-6720 Szeged, Hungary; †Department
of Organic Chemistry, University of Szeged, Do´m te´r 8, H-6720 Szeged, Hungary; and ‡Analytical Laboratory
of the Cereal Research Institute, P.O.Box 391, H-6701 Szeged, Hungary
Received June 29, 2001; revised August 31, 2001; accepted August 31, 2001
enantioselective hydrogenation of ethyl pyruvate (EtPy)
resulting in the formation of (R)-ethyl lactate (EtLt)
with an enantioselectivity of 97% (7, 8) on an alumina-
supported platinum catalyst modified with dihydrocin-
chonidine (DHCD) has been studied in detail (Scheme 1).
Thedeterminationoftheoptimalparametersof(R)-EtLt
production and the interpretation of the reaction mecha-
nism is still an on-going process (see Refs. 9–15).
The following literature data are considered on the inter-
pretation of the chiral hydrogenation of pyruvates. (i) the
particle size and morphology of the Pt–Al2O3 catalyst (5, 6,
8, 13, 16–22), (ii) how DHCD is adsorbed and on which sur-
face sites (8, 16, 23–26), (iii) the adsorption of DHCD and
pyruvate (9, 23, 24, 26–30), (iv) the conformation of DHCD
(15, 31–33), (v) the stability of DHCD as a chiral modifier
under the conditions of the hydrogenation (15, 34), (vi) the
formation of the intermediate complex in solution and/or
on the surface of the catalyst (5, 6, 27, 35–37), (vii) the struc-
ture of the intermediate responsible for chirality (5, 9, 23,
27, 35, 36, 38–42), (viii) the roles of DHCD concentration
and of the DHCD/Pt and pyruvate/Pt ratios (6, 12, 43), and
The enantioselective hydrogenation of ethyl pyruvate (EtPy) on
Pt–alumina (E 4759), Pt black, and Pt black + alumina (mixture)
catalysts modified by dihydrocinchonidine (DHCD) in acetic acid
was studied by electrospray ionization–mass spectrometry (ESI–
MS). Application of the ESI–MS technique led to the recognition of
a novel-type compound, O+[Al(OAc)2]3 (oxonium ions). The effect
of the DHCD concentration, temperature, and oxonium cations
on the reaction rate and the enantioselectivity has been studied.
Using the Engelhard 4759 catalyst in acetic acid under mild exper-
imental conditions (room temperature, hydrogen pressure 1 bar,
DHCD concentration 0.01 mM/L) an optical yield of 92% can
be achieved. The high enantioselectivity is accompanied by the
following turnovers: EtPy/DHCD > 43,000, EtPy/Ptsurface > 1000,
TOF = 1–2 s 1, and DHCD/Ptsurface ratio = 0.0072. The enantios-
electivity reducing factor is identified by ESI–MS as the gradual
hydrogenation of the quinoline skeleton of DHCD that becomes
more pronounced with increasing temperature and hydrogena-
tion time. The discovery of oxonium cations, the extremely low
DHCD/Ptsurface ratio, and the new data obtained by the Pt black +
alumina mixture made possible an interpretation of the mech-
anism of the heterogeneous enantioselective hydrogenation of
-ketoesters.
c
2002 Elsevier Science
Key Words: hydrogenation; enantioselective; Pt/Al2O3; dihy- (ix) the results of kinetic reaction studies (35, 25, 43–45).
drocinchonidine; ethyl pyruvate; ESI–MS; intermediate; oxonium
cations.
The most important conclusions of these investigations
on the enantioselective hydrogenation of EtPy can be sum-
marized as follows: (i) the highest enantioselectivities can
be reached on the Pt–alumina catalyst having its main par-
ticle size in the range of 1.5–4 nm; (ii) the best solvent was
INTRODUCTION
found to be AcOH; (iii) the best modifiers are the cinchona
Enantiomers play an extremely important role in the
alkaloids which have to be present in an optimal concentra-
synthesis of pharmaceuticals, agrochemicals, flavors, and
tion; (iv) the structure of the surface complex responsible
fragrances. Studies on heterogeneous asymmetric catalytic
for the chiral induction is the cinchona alkaloid “anti-open”
reactions involving enantiomers have therefore recently
conformer–pyruvate 1 : 1 complex; (v) the reaction rate of
gathered considerable momentum (1–4). The enantiose-
the enantioselective reaction is significantly higher than the
lective hydrogenation of -ketoesters is one of the best
rate of the racemic hydrogenation; the enantioselectivity
known examples of such reactions (5, 6). Of these, the
and the reaction rate have a maxima as a function of the
modifier concentration; (vi) under optimal reaction condi-
tions the decrease in ee can be ascribed to the hydrogena-
tion of the modifier.
1 To whom correspondence should be addressed. Fax: 36-62-544-200.
E-mail: bartok@chem.u-szeged.hu.
168
0021-9517/02 $35.00
c
2002 Elsevier Science
All rights reserved.

ENANTIOSELECTIVE HYDROGENATION OF ETHYL PYRUVATE
169
fication of conversion and ee were based on GC data.
The product identification and the enantiomeric excess
[ee% = ([R] [S]) 100/([R] + [S])]weremonitoredbygas
chromatography (HP 5890 GC-FID, 30-m-long Cyclodex-B
capillary column, uncertainty 2%).
SCHEME 1
ESI–MS Measurements
The major aim of this work is the presentation of our new
experimental results on the asymmetric hydrogenation of
EtPy on the Pt–alumina catalyst in acetic acid, in the pres-
ence of DHCD, made possible mainly by the application
of electospray ionization–mass spectrometry (ESI–MS and
ESI–MS/MS) (46–48). In this investigations ESI–MS is used
for the identification of the intermediates, different adducts,
and side products formed in the reaction. A great advantage
of ESI–MS as compared with other ionization methods is
that, due to the mildness of the ionization process, sample
fragmentation is less severe. In most cases intact molecular
ions are also detected in the sample.
A Hewlett–Packard HP 5989 B MS Engine quadrupole
mass spectrometer equipped with a high-energy dinode de-
tector and an atmospheric pressure ionization–electrospray
(API–ES) interface (HP 59987 A) were used. To perform
ES experiments, samples were dissolved separately in a
mixture of MeOH/AcOH or EtOH/AcOH (98/2 v/v) and
introduced into the ES ion source by using a Harward Type
22 syringe pump at a flow rate of 20 l min 1. A fine spray
was formed by applying nitrogen as nebulizer gas at a pres-
sure of 135 kPa. Nitrogen drying gas (heated to 563 K) was
1
used at a flow rate of 8 L min to facilitate solvent evap-
oration from the droplets. After tuning, mass spectra of
reaction mixtures were acquired by applying a capillary exit
voltage (CapEx) of 100 V. The instrument was run in SCAN
mode at above 0.1 mM/L DHCD concentration. When
the concentration of DHCD was in the range of 0.001–0.1
mM/L selected ion monitoring (SIM) was applied. Case
sampleswerepreparedforESI–MSinthefollowingway:af-
ter removing by rotary vacuum evaporation the acetic acid
contentofthehydrogenationsamplesobtainedafterincom-
plete and complete conversion, the samples were dissolved
separately in a mixture of 2 0.5 mL of MeOH/H2O/AcOH
(80/18/2 v/v). Product composition is unaffected by the
conditions of ESI below a CapEx of 100 V; above 100
V, fragmentation processes may be set off. Relative peak
intensities in the individual samples to be analyzed were
determined under completely identical ESI conditions, ren-
dering comparison of the compositions of samples hydro-
genated under different conditions. It must be noted that
peak intensities determined by ESI–MS were compared
only for DHCD and its hydrogenated derivatives.
EXPERIMENTAL
Materials
Cinchonidine(CD), AcOH, MeOH, andEtOHwerepur-
chased from Fluka. EtPy (Fluka) was distilled before use
to attain 99.5% purity. DHCD was prepared by hydrogena-
tion of CD (Pd/C, 1 N H2SO4/H2O, 1 bar, 298 K) and used
after crystallization.
Based on the data in the literature, among several cata-
lysts the one most often used is Engelhard 4759 (E 4759).
The properties of this contact catalyst are the following
(5): Pt content, 5% (w/w); Pt dispersion, as received, 38%;
after pretreatment, 27% (49); mean Pt particle size, 4.5 nm;
support, -Al2O3; specific surface area, 168 m2 g 1; mean
pore radius, 2 nm; specific pore volume, 0.27 mL g 1. E 4759
was pretreated before use in a fixed-bed reactor by flush-
1
ing with 30 mL min helium at 300–673 K for 30 min and
30 mL min 1 hydrogen at 673 K for 100 min. After cooling
to room temperature in hydrogen, the catalyst was flushed
with helium for 30 min and stored in air before use.
Platinum black (Aldrich 20,591-5), aluminum oxide
(Aldrich 19,997-4), and aluminum acetates (Aldrich 28,982-
5 and 29,485-3) were used as received.
ESI–MS/MS Measurements
ESI–MS/MS measurements were run on a FINNIGAN
TSQ7000 tandem mass spectrometer equipped with a
home-built nanospray source (50), where a pulled, gold-
coated capillary was used. The high voltage at the capillary
was set at 1100 V. The instrument was run in product ion
mode. The mass range of Q3 was set at 10–1500 u with a
SCAN time of 1.5 sec. Argon was used as collision gas at a
pressure of 0.33 Pa. The collision energy was manually set
in the range 20–35 eV.
Hydrogenation
Hydrogenationwasperformedinanatmosphericbathre-
actor or in a Berghof Bar 45 autoclave at room temperature
(298 K). The catalytic system including the catalyst and
2 mL of AcOH was purged three times with hydrogen after
prehydrogenation (30 min). The calculated amount of mod-
ifier and 0.1–2 mL of EtPy were introduced and stirred
(1200 rpm) in the presence of hydrogen for the required
reaction time (usually 10–50 min). The standard conditions
were: 25 mg E 4759, 2 mL AcOH, 0.01 mM/L DHCD,
RESULTS AND DISCUSSION
Since the effects of the reaction conditions of hydro-
1 bar H2, 296 K, 1200 rpm, and 0.1 mL EtPy. The quanti- genation on enantioselectivity have never been studied

´
BARTOK ET AL.
170
TABLE 1
by ESI–MS, we had to repeat several experiments which
had already been accomplished in the course of the
optimization of (R)-EtLt production.
Adducts Detected by Positive ESI–MS and ESI–MS/MS [m = 23
(Na), 39 (K), 32 (MeOH), 46 (EtOH), 116 (EtPy), 118 (EtLt), 296
DHCD), 71 (Py), 73 (Lt), 43 (Ac)]a
Unknown Adducts in the Reaction Mixture
m/z
Adducts
The experimental program included the determination
of the reaction rate and enantioselectivity as the func-
tions of DHCD concentration, temperature, and EtPy con-
centration. The ESI–MS spectrum of the raw product of
the hydrogenation reaction was recorded, and the peaks
with unknown m/z values were identified by ESI–MS/MS
(14).
The two representative spectra shown in Figs. 1a and
1b (recorded at different DHCD concentrations) were se-
lected from a great number of ESI–MS spectra; peaks so far
identified are listed in Table 1. The presence of MeOH and
EtOH in some adducts is due to their use as solvents in ESI.
Adducts containing solvent molecules are not generated
119
141
151
154
200
201
257
273
287
297
299
301
303
307
371
373
387
401
403
451
469
483
491
509
515
521
543
547
551
567
589
719
747
751
[EtLt + H]+
[EtLt + Na]+
[HHCD + 2H]2+
[DDHCD + 2H]2+
[THCD + 2MeOH + 2H2O + 2H]2+
[HHCD + 2MeOH + 2H2O + H]+
[EtPy + EtLt + Na]+
[EtPy + EtLt + K]+
[2EtPy + MeOH + Na]+
[DHCD + H]+
[THCD + H]+
[HHCD + H]+
[2EtPy + MeOH + K]+
[DDHCD + H]+
[3EtPy + Na]+
[2EtPy + EtLt + Na]+
[3EtPy + K]+
[HHCD + 2MeOH + 2H2O + H]+
[3EtPy + MeOH + Na]+
[O{Al(OAc)2}3]+
[451 + H2O]+
[451 + MeOH]+
[2EtPy + 2EtLt + Na]+
[O{Al(OAc)2 Al(OPy OLt)}]+
[451 + 2MeOH]+
[3EtPy + EtLt + MeOH + Na]+
[451 + 2EtOH]+
[451 + 3MeOH]+
[4EtPy + 2MeOH + Na]+
[O Al(OAc)2 Al(OPy)2 Al(OLt)2]+
[451 + 3EtOH]+
[6EtPy + Na]+
[451 + DHCD]+
[6EtPy + MeOH + Na]+
a Abbreviations used: EtPy, ethyl pyruvate; EtLt, ethyl lactate; DHCD,
dihydrocinchonidine; THCD, tetrahydrocinchonidine; HHCD, hexahy-
drocinchonidine; DDHCD, dodecahydrocinchonidine; Ac, acetyl; Py,
pyruvyl; Lt, lactyl.
undertheconditionsofESI:theyareformedinthecourseof
the dilution of the reaction product (with MeOH or EtOH),
during the preparation of the samples for the MS measure-
ments.
Inspection of the ESI–MS spectra, surprisingly, revealed
a large number of peaks and a high dependence of their
positions (m/z) on DHCD concentration. The adducts pre-
sentedinTable1canbeclassifiedintofourgroupsaccording
to their type: (i) DHCD and its hydrogenated derivatives,
(ii) the adducts of EtPy and its oligomers, (iii) EtLt and
its adducts, and (iv) oxonium compounds formed from
alumina.
FIG. 1. Positive ion ESI–MS (CapEx 100V) of the EtPy hydrogena-
tion product (25 mg E 4759, 2 mL AcOH, 1 bar H2, 298 K, 0.1 mL EtPy).
(a) 3 mM/L DHCD and (b) 0.014 mM/L DHCD. For abbreviations, see
Table 1.
Identification of the peak with m/z = 451 was quite prob-
lematic and finally led to the recognition of a novel-type

ENANTIOSELECTIVE HYDROGENATION OF ETHYL PYRUVATE
171
FIG. 2. The structure of m/z = 451.
compound (51) (Fig. 2). As shown in Table 1, the oxo-
nium cation with m/z = 451 was identifiable on the basis
of several adducts, the most important of these being the
DHCD adduct of the oxonium cation, m/z = 747. This
compound probably plays a role in the development of
the chiral environment on the catalyst surface. Na+ and
K+ adducts were not generated by the intentional addi-
tion of Na+ and K+ ions but are contaminants present in
all components of the system (catalyst, solvents, glass re-
actor, etc.). Studies on the effects of the various adducts
on enantioselectivity may constitute the subject of further
experiments.
FIG. 3. Hydrogenation of EtPy on DHCD-modified platinum (25 mg
E 4759, 1 bar H2, 2 mL AcOH, 0.5 mL EtPy, 296 K, 1200 rpm). For abbre-
viations, see Table 1.
relationship between hydrogenation time, conversion,
DHCD concentration, and enantioselectivity is shown in
Fig. 4b and in Table 2.
The Effect of DHCD Concentration
Althoughtherehavebeenindicationsrelatedtothecause
of reduced enantioselectivity when the reaction time and
conversion were increased (5, 6, 12), the progress of the
formation of hydrogenated DHCD derivatives is experi-
mentally confirmed by the ESI–MS data shown in Table 3.
According to the data presented in Table 3, in the
course of the hydrogenation of EtPy the quinoline skele-
ton of DHCD hydrogenates also. The ratio of hydro-
genated cinchonas increases with the increasing conversion
of EtPy. Since the adsorption strength of hydrogenated
cinchonas is considerably lower than that of DHCD,
they are easily desorbed at high DHCD concentrations,
whereas at low DHCD concentrations (below 0.1 mM/L)
enantioselectivity will decrease because there is not suffi-
cient DHCD present to produce optimal surface coverage.
The same is confirmed by the results obtained when the
The basic experimental data, revealing for the first time
that the DHCD concentration has a significant influence on
the rate and selectivity of the enantioselective hydrogena-
tionofEtPy, werepublishedbyBlaserandco-workersinthe
early 1990s (43, 52). However, there are no data published
on the effect of the DHCD concentration on the reaction
rate and enantioselectivity in AcOH at 1 bar H2 pressure
on the E 4759 catalyst. This study had to be carried out, as
such experimental data may be essential in studying the re-
action mechanism. Our pertinent experimental results are
summarized in Figs. 3 and 4a and 4b and Table 2.
Theachievementofhighopticalyieldsisdependentonan
optimal DHCD concentration adjusted to the experimental
conditions. The function describing the dependence on
DHCD concentration of the reaction rate has maxima. For
the interpretation of these maxima there are two expla-
nations in the literature (12, 43). Blaser et al., on the ba-
sis of their kinetic study, explained this phenomenon by
the so-called “three-site model” (43), while LeBlond et al.
(12) interpreted it by the change of the optimal adsorp-
tion geometry of DHCD at higher concentrations of the
modifier.
TABLE 2
Time of Reaction, Conversion, and Enantioselectivity as Functions
of DHCD Concentration for Hydrogenation of EtPya
DHCD (mM/L)
Time (min)
Conversion (%)
ee (%)
The experimental data (Fig. 4a) indicate that, under the
standard conditions described in the Experimental sec-
tion, the minimal amount of DHCD necessary for the
achievement of maximal enantioselectivity is in the range of
0.001–0.01 mM/L. For this reason additional investigations
were necessary in this DHCD concentration range. The
0.001
0.002
0.006
0.01
90
90
40
25
15
64
70
90
96
82
39
51
85
88
91
0.01
a For conditions, see Fig. 4.

´
BARTOK ET AL.
172
TABLE 3
Relative Abundances of [M +H]+ Ions in the ESI Mass Spectra
of Cinchonas Formed from DHCD in the Enantioselective Hydro-
genation of EtPy under Standard Conditionsa
m/z Values (relative peak intensity %)
DHCD
Time Conversion
(mM/L) (min)
(%)
297
299
301
307
ee (%)
0.001a
0.01a
0.1b
15
40
15
30
10
20
15
25
19
40
68
97
49
99
80
100
46
36
81
36
100
10
100
63
34
56
47
20
20
55
53
64
60
56
45
50
15
30
57
100
100
100
23
100
5
10
63
51
92
91
91
92
92
92
1.0b
100
100
a SIM—selective ion monitoring.
b SCAN.
us (hydrogen pressure 1 bar, DHCD concentration 0.001–
0.1 mM/L, temperature 263 K, 273 K) have not been pub-
lished previously. Our experimental results are summarized
inFigs. 6and7. Reactiontimesemployedintheexperiments
are presented in Fig. 6, and the corresponding conversions
are listed in Table 4.
Naturally, it was to be expected that the hydrogenation
rate of EtPy would decrease as a consequence of reducing
temperature. The most important observation is, however,
that the hydrogenation rate of the quinoline ring of DHCD
decreased to an even higher extent: the hydrogenation of
DHCD was not detectable at all (DHCD 0.001 mM/L, EtPy
conversion32%, reactiontime150min)whiletheformation
of (R)-EtLt still proceeded with high (80%) enantioselec-
tivity under these conditions.
FIG. 4. (a, b) Hydrogenation of EtPy on DHCD-modified platinum
(25 mg E 4759, 1 bar H2, 2 mL AcOH, 0.1 mL EtPy, 296 K, 1200 rpm). For
abbreviations, see Table 1.
amount of EtPy was increased under identical experimen-
tal conditions, as shown in Fig. 5.
The data in Table 3 also confirm that enantioselectivity
of 92–93% can be achieved under very mild experimental
conditions (hydrogen pressure: 1 bar, room temperature,
short reaction time) if an optimal chiral environment is
created.
As previously mentioned, these investigations permit-
ted the determination of the minimal amount of DHCD
required for obtaining the maximal enantioselectivity.
The Effect of Temperature
The role of temperature was first studied under relatively
severe experimental conditions (high hydrogen pressure
and DHCD concentration) (35, 53), and important con-
clusions were drawn. As far as we know, the results ob-
FIG. 5. Conversion and enantioselectivity as function of EtPy con-
tained under the mild experimental conditions applied by centration on the hydrogenation of EtPy under standard conditions.

ENANTIOSELECTIVE HYDROGENATION OF ETHYL PYRUVATE
TABLE 4
173
Reaction Time and Conversion as Functions of DHCD
Concentration and Temperaturea
Reaction time (min)/conversion (%)
DHCD (mM/L)
263 K
273 K
297 K
0.001
0.01
0.1
150/32
130/64
120/86
150/35
60/66
60/98
150/45
50/97
37/98
a For conditions see Fig. 6.
aims of this study. For comparison, experimental data ob-
tained on catalyst E 4759 on mixtures of Pt+HOAl(OAc)2,
Pt + (HO)2Al(OAc), and DHCD-free platinum (racemic
hydrogenation) are also included.
It is clearly seen that hydrogenation on Pt is significantly
slower than that on a Pt + Al2O3 mixture and on mixtures
of Pt + HOAl(OAc)2 and Pt + (HO)2Al(OAc) containing
an identical amount of Pt.
FIG. 6. Effect of the DHCD concentration and temperature for the
hydrogenation of EtPy on optical yield (25 mg E 4759, 1 mL AcOH +
1 mL toluene in order to keep AcOH dissolved, 1 bar H2, 0.5 mL EtPy).
Our dry-box studies on the filtrate containing oxonium
cations have so far yielded no appreciable results. It is
possible that oxonium cations, being moisture sensitive,
disappear during hydrogenation. In the presence of alu-
mina, however, they are formed continuously.
The Effect of Alumina
Because of the discovery of the new type of aluminum
compounds in AcOH, it was necessary to investigate
the role of the alumina support on the hydrogenation of
EtPy.
The experiments were carried out on support-free plat-
inum (platinum black) as well as on a mechanical mixture of
platinum and aluminum oxide, under standard conditions.
The results are summarized in Figs. 8 and 9 and Table 5.
It has to be noted that optimization was not among the
Interpretation of the Enantioselective Hydrogenation
of -Ketoesters
Our new experimental data make it possible to complete
the mechanism of chiral hydrogenation; for example, (i) the
extremely low DHCD/Ptsurface ratio for obtaining the
FIG. 7. Effect of the temperature for the hydrogenation of EtPy on
the conversion (25 mg E 4759, 1 mL AcOH + 1 mL toluene, 1 bar H2,
0.01 mM/L DHCD, 0.5 mL EtPy).
FIG. 8. Enantioselective hydrogenation of ethyl pyruvate on pla-
tinum black in acetic acid at room temperature (1.5 mg Pt, 0.01 mM/L
dihydrocinchonidine). For abbreviations and conditions see Table 5.

´
BARTOK ET AL.
174
According to earlier investigations the DHCD/Ptsurface
ratio = 0.1–1, depending on the solvent (5, 6). In AcOH
at 303 K under 5.6 bar of hydrogen pressure this ratio
was found to be 0.038 (12). According to Fig. 10 this
value is 0.0072 (conditions: 25 mg E 4759, 2 mL AcOH,
1 bar hydrogen, 296 K, 0.1 mL EtPy, 0.006 mM/L DHCD,
Ptsurface = 0.00173 mM, DHCD = 0.0000125 mM). Ac-
cordingly, under our experimental conditions the minimal
DHCD amount for obtaining the maximal ee is five times
less than that determined in Ref. 12. We attempted to repro-
duce the experimental data reported by LeBlond et al. (12),
using the 1% Pt–alumina catalyst purchased from Aldrich
(i.e., the same as that used by LeBlond et al.); however, we
did not succeed. One probable reason for our failure may be
that the two 1% Pt–alumina catalysts come from different
batches, and no information is available about their char-
acteristics. On the other hand, experimental results from
different laboratories using catalyst E 4759 are fully repro-
ducible.
FIG. 9. Enantioselective hydrogenation of ethyl pyruvate over pla-
Suppose that adsorption of one DHCD molecule de-
mands an area of 15 Pt atoms on the catalyst surface (5, 6).
In the case of monolayer adsorption only 10% of the
0.00173 mM Ptsurface will be covered by DHCD under
our experimental conditions, while the ee obtained on
this catalyst surpasses 90%. Thus on the remaining 90%
Ptsurface which is not covered by DHCD the hydrogenation
does not occur. This may be interpreted by the following
explanation: the chiral active sites are formed on well-
defined surface Pt atoms having a special coordination
(corners, vertices, kinks) (4, 16, 54–56) by the adsorbed
DHCD molecule, and the hydrogenation will proceed with
a high reaction rate only on these centers. Accordingly, it
tinum black in acetic acid at room temperature (1.5 mg Pt, a, 5 mg Pt). For
abbreviations and conditions see Table 5.
maximal ee; (ii) the new data obtained by platinum
black + alumina mixture; (iii) in the course of the analy-
sis of the products of the enantioselective hydrogenation
of EtPy by ESI-MS, an unknown substance was detected in
the mass spectrum at m/z = 451 with maximal abundance.
TABLE 5
EnantioselectiveHydrogenationofEthylPyruvateoverPlatinum
in Acetic Acid at Room Temperature (2 ml of AcOH, hydrogen
pressure: 1 bar, 0.1 mL ethyl pyruvate, stirring: 1200 rpm)
DHCD
(mM/L)
Conversion
(%)
ee
(%)
Catalysta
1.5 mg Pt
Time
150 min
66 hb
0.01
0.01
0
0.01
0.01
0
40
84
36
90
99
60
77
45
38
0
68
64
0
150 minc
85 min
66 hb
1.5 mg Pt +
20 mg Al2O3
3 hc
1.5 mg Pt +
120 min
0.01
61
20 mg HOAl(OAc)2
1.5 mg Pt +
150 min
0.01
78
62
20 mg AcOAl(OH)2
1.5 mg Pt +
45 min
45 min
60 min
60 min
60 min
25 min
25 min
25 min
0.1
0.2
0.4
1
10
0.1
1
98
98
98
97
60
100
100
100
83
85
81
75
67
74
72
90
20 mg Al2O3
5 mg Pt +
20 mg Al2O3
25 mg Pt/Al2O3(E 4759)
0.01
a Pt = Platinum black.
FIG. 10. Initial rate and enantioselectivity as functions of DHCD/
Ptsurface for hydrogenation of EtPy. For conditions and abbreviations, see
Fig. 4.
b Without stirring.
c Without dihydrocinchonidine.

ENANTIOSELECTIVE HYDROGENATION OF ETHYL PYRUVATE
175
seems that the sizes of the Pt clusters have no particular dency to be adsorbed on the surface of both platinum and
importance, as was concluded in previous studies (reviewed support.
in Ref. 5). If the enantioselective hydrogenation were to
proceed on edge and terrace atoms, then a significantly
higher DHCD concentration would be necessary for
CONCLUSION
obtaining high ee-s. In this context there are noteworthy
studies by Augustine et al. (16) on the mechanism of the
enantioselective hydrogenation. It is regrettable that in
AcOH no such study has been carried out.
By studying the optimization of the enantioselective hy-
drogenation of EtPy as a function of the DHCD concen-
tration it could be established that on the catalyst under
optimal conditions only 10% of the Ptsurface is covered by
DHCD. DHCD is adsorbed only at sites of higher coor-
dination with a well-defined morphology (edges, corners),
and hydrogenation may take place only at these locations.
The rest of the Pt surface remains intact.
Asregardsthemechanismofthehydrogenationreaction,
we emphasize the role of electrostatic acceleration rather
than that of ligand acceleration, because the former allows
the ionic character of the hydrogenation to be taken into
account.
Furthermore, the present results suggest an explanation,
entirely different from the currently accepted one, as to
why acetic acid has proven to be the optimal solvent in
the asymmetric hydrogenation of EtPy to EtLt, and why
aluminum oxide is the optimal support for platinum.
Our future plan is the experimental verification of the
presumed role of oxonium cations by the extensive investi-
gation of their role in the hydrogenation of EtPy on mixed
catalysts comprising platinum black and carefully charac-
terized Al2O3 and SiO2. A similarly important program is
the isolation and preparation of the aluminum-containing
oxonium cations.
In the course of studies on the mechanisms of liquid-
phase heterogeneous catalytic reactions, hydrogen defi-
ciency at the catalytic sites must be taken into account (4).
The dissociation of hydrogen on platinum proceeds at a
significantly higher rate on edge sites and corner sites of
the surface, which is where chiral active centers develop in
the course of DHCD adsorption. Surface hydrogen rapidly
reacts with polarized ethyl pyruvate at the site of genera-
tion, prior to migration to the terrace atoms. This means
that even if adsorbed EtPy were available on the terrace
atoms of platinum, no reaction would take place due to the
hydrogen-deficient environment.
What is the role of the oxonium cation? Owing to its high
reactivity, this electrophilic compound present both in the
solution and on the surface of the catalyst may affect the
chemical reaction. To the best of our knowledge, oxonium
ions of this structure have not been described in the perti-
nent literature (see, e.g., the recently published monograph
by Olah et al. (57)).
In acetic acid the oxonium cations may affect the chemi-
cal reaction in a hitherto unknown but most probably multi-
ple fashion: oxonium cation removes impurities; it polarizes
ethyl pyruvate; it stabilizes transition-state structures on
platinum; it may have a role in the development of electro-
static potential on the surface. (It must also be kept in mind
that the support itself (58) and/or the unknown substances
present therein may also affect (increase or decrease) the
reaction rate.)
ACKNOWLEDGMENTS
Financial support by the Hungarian National Science Foundation
(OTKA Grant T031707) and the Hungarian Academy of Sciences (AKP
2000-20 2,4) is highly appreciated. We also thank Professor F. Notheisz for
valuable discussions.
In our opinion these cations make the so-called electro-
static catalysis (59) based on electrostatic acceleration (60)
possible. The chiral sites and the electrostatic potential may
be responsible for the increased rate of the chiral hydro-
genation for the decrease of the activation energy, that is,
for the catalysis. Molecular mechanics calculations, on the
binding of the reacting molecules on crystal surfaces, should
take place due to formation of a strong electrostatic field
on the surface (59). If the polarity of the transition com-
plex in a reaction is higher than that in the ground state, the
electrostatic stabilization effect of the dipoles of the envi-
ronment (in our case these are the cations) will increase the
rate by electrostatic acceleration (60).
REFERENCES
1. Blaser, H.-U., and Mu¨ller, M., Stud. Surf. Sci. Catal. 59, 73 (1991).
2. Jannes, G., and Dubois, V., Eds., “Chiral Reactions in Heterogeneous
Catalysis.” Plenum Press, New York, 1995.
3. Baiker, A., and Blaser, H.-U., in, “Handbook of Heterogeneous Cata-
lysis” (G. Ertl, H. Kno¨zinger, and J. Weitkamp, Eds.), Vol. 5. p. 2422.
VCH, Weinheim, 1997.
4. Smith, G. V., and Notheisz, F., “Heterogeneous Catalysis in Organic
Chemistry.” Academic Press, New York, 1999.
5. Baiker, A., J. Mol. Catal. A Chem. 115, 473 (1997).
6. Blaser, H.-U., Jalett, H.-P., Mu¨ller, M., and Studer, M., Catal. Today
37, 441 (1997).
7. To¨ro¨k, B., Felfo¨ldi, K., Szakonyi, G., Bala´zsik, K., and Barto´k, M.,
Catal. Lett. 52, 81 (1998).
8. Zuo, X., Liu, H., Guo, D., and Yang, X., Tetrahedron 55, 7787
(1999).
9. Bu¨rgi, T., and Baiker, A., J. Catal. 194, 445 (2000).
10. Bhaduri, S., Lahiri, G. K., Munshi, P., and Mukesh, D., Catal. Lett. 65,
61 (2000).
This possibility is confirmed by the presence of the molec-
ular ion identifiable at m/z = 747 in the ESI–MS spectrum
(Table 1), small amounts of which appear in the solution at
high DHCD concentrations (above 1 mM/L). We think it
is justified to assume that these species have a strong ten-

´
BARTOK ET AL.
176
11. Hall, T. J., Halder, J. E., Hutchings, G. J., Jenkins, R. L., Johnston, P., 32. Pfaltz, A., and Heinz, T., Topics Catal. 4, 229 (1997).
McMorn, P., Wells, P. B., and Wells, R. P. K., Topics Catal. 11/12, 351 33. Bu¨rgi, T., and Baiker, A., J. Am. Chem. Soc. 120, 12920 (1998).
(2000).
34. Morawsky, V., Pru¨ße, U., Witte, L., and. Vorlop, K.-D., Catal. Commun.
1, 15 (2000).
35. Blaser, H.-U., Jalett, H.-P., Garland, M., Studer, M., Thies, H.,
and Wirth-Tijani, A., J. Catal. 173, 282 (1998).
36. Ta´las, E., Botz, L., Margitfalvi, J. L., Sticher, O., and Baiker, A.,
J. Planar Chromatogr. 5, 28 (1992).
37. Margitfalvi, J. L., and Hegedu¨s, M., J. Mol. Catal. A Chem. 107, 281
(1996).
38. Augustine, R. L., and Tanielyan, S. K., J. Mol. Catal. A Chem. 112, 93
(1996).
39. Margitfalvi, J. L., Hegedu¨s, M., and Tfirst, E., Tetrahedron Asymmetry
7, 571 (1996).
40. Barto´k, M., Felfo¨ldi, K., To¨ro¨k, B., and Barto´k, T., Chem. Commun.
2605 (1998).
41. Barto´k, M., Felfo¨ldi, K., Szo¨llo¨si, Gy., and Barto´k, T., Catal. Lett. 61,
1 (1999).
42. Szo¨llo¨si, Gy., Felfo¨ldi, K., Barto´k, T., and Barto´k, M., React. Kinet.
Catal. Lett. 71, 99 (2000).
43. Blaser, H.-U., Garland, M., and Jalett, H.-P., J. Catal. 144, 569 (1993).
44. Sun, Y., Landau, R. N., Wang, J., LeBlond, C., and Blackmond, D. G.,
J. Am. Chem. Soc. 118, 1348 (1996).
45. Margitfalvi, J. L., Ta´las, E., Tfirst, E., Kumar, C. V., and Gergely, A.,
Appl. Catal. A Gen. 191, 177 (2000).
46. Fenn, J. B., Mann, M., Meng, C. K., Wong, S. F., and Whitehouse,
C. M., Science 246, 64 (1989).
47. Gaskell, S. J., J. Mass Spectrom. 32, 677 (1997);
48. Kebarle, P., J. Mass. Spectrom. 32, 922 (1997).
49. von Arx, M., Mallat, T., and Baiker, A., J. Catal. 193, 161 (2000).
50. Kele, Z., Jana´ky, T., Me´sza´ros, T., Fehe´r, A., Dudits, D., and Szabo´,
P. T. , Rapid Commun. Mass. Spectrom. 12, 1564 (1998).
51. Barto´k, M., Szabo´, P. T., Barto´k, T., Szo¨llo¨si, Gy., and Bala´zsik, K.,
Rapid Commun. Mass Spectrom. 15, 65 (2001).
12. LeBlond, C., Wang, J., Liu, J., Andrews, A. T., and Sun, Y.-K., J. Am.
Chem. Soc. 121, 4920 (1999); Topics Catal. 13, 169 (2000).
13. Bo¨hmer, U., Franke, F., Morgenschweis, K., Bieber, T., and
Reschetilowski, W., Catal. Today 60, 167 (2000).
14. Barto´k, M., Szabo´, P. T., Barto´k, T., and Szo¨llo¨si, Gy., Rapid
Commun. Mass Spectrom. 14, 509 (2000).
15. Blaser, H.-U., Jalett, H.-P., Lottenbach, W., and Studer, M., J. Am.
Chem. Soc. 122, 12675 (2000).
16. Augustine, R. L., Tanielyan, S. K., andDoyle, L. K., TetrahedronAsym-
metry 4, 1803 (1993).
17. Blaser, H.-U., Jalett, H.-P., Monti, D. M., Reber, J. F., and Wehrli, J. T.,
Stud. Surf. Sci. Catal. 41, 153 (1988).
18. Wehrli, J. T., Baiker, A., Monti, D. M., and Blaser, H.-U., J. Mol. Catal.
61, 207 (1990).
19. Blaser, H.-U., Jalett, H.-P., Monti, D. M., Baiker, A., and Wehrli, J. T.,
Stud. Surf. Sci. Catal. 67, 147 (1991).
20. Jackson, S. D., Willis, J., McLellan, G. D., Webb, G., Keegan, M. B. T.,
Moyes, R. B., Simpson, S., Wells, P. B., and Whyman, R., J. Catal. 139,
191 (1993).
21. Mallat, T., Frauchiger, S., Kooyman, P. J., Schu¨rch, M., and Baiker, A.,
Catal. Lett. 63, 121 (1999).
22. Collier, P. J., Iggo, J. A., and Whyman, R., J. Mol. Catal. A Chem. 146,
149 (1999).
23. 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., Recl. Trav. Chim. Pays-Bas 113, 465
(1994).
24. Carley, A. F., Rajumon, M. K., Roberts, M. W., and Wells, P. B.,
J. Chem. Soc., Faraday Trans. 91, 2167 (1995).
25. Blaser, H.-U., Imhof, D., and Studer, M., Stud. Surf. Sci. Catal. 108,
175 (1997).
26. Evans, T., Woodhead, A. P., Gutie´rrez-Sosa, A., Thornton, G., Hall,
T. J., Davis, A. A., Young, N. A., Wells, P. B., Oldman, R. J.,
52. Blaser, H.-U., and Garland, M., J. Am. Chem. Soc. 112, 7048 (1990).
53. Meheux, P. A., Ibbotson, A., and Wells, P. B., J. Catal. 128, 387 (1991).
54. Siegel, S., Outlaw, J., and Garti, N., J. Catal. 52, 102 (1978).
55. Somorjai, G. A., “Chemistry Two Dimensions: Surfaces.” Cornell
Univ. Press, Ithaca, New York and London, 1981.
56. Augustine, R. L., and Warner, R. W., J. Catal. 80, 358 (1983).
57. Olah, G. A., Laali, K. K., Wang Q., and Prakash, G. K. S., “Onium
Ions.” Wiley, New York, 1998.
58. Kropp, P. J., Daus, K. A., Tubergen, M. W., Kepler, K. D., Wilson, V. P.,
Craig, S. L., Baillargeon, M. M., and Breton, G. W., J. Am. Chem. Soc.
115, 3071 (1993).
59. Na´ray-Szabo´, G., in “Encyclopedia of Computational Chemistry”
(P. v. R. Schleyer, Ed.), Vol. 1. Wiley, New York, 1998.
60. Jiao, H., and Schleyer, P. v. R., J. Chem. Soc., Faraday Trans. 90, 1559
(1994).
˚
Plashkevych, O., Vahtras, O., Agren, H., and Carravetta, V., Surf. Sci.
Lett. 436, L691–L696 (1999).
27. Sutherland, I. M., Ibbotson, A., Moyes, R. B., and Wells, P. B., J. Catal.
125, 77 (1990).
28. Bu¨rgi, T., Atamny, F., Knop-Gericke, A., Ha¨vecker, M., Schedel-
Niedrig, T., Schlo¨gl, R., and Baiker, A., Catal. Lett. 66, 109 (2000).
29. Bu¨rgi, T., Atamny, F., Schlo¨gl, R., and Baiker, A., J. Phys. Chem. B
104, 5953 (2000).
30. Bonello, J. M., Lambert, R. M., Ku¨nzle, N., and Baiker, A., J. Am.
Chem. Soc. 122, 9864 (2000).
31. Dijkstra, G. D. H., Kellogg, R. M., Wynberg, H., Svendsen, J. S.,
Marko, I., and Sharpless, K. B., J. Am. Chem. Soc. 111, 8069
(1989).