New data on enantiomeric excess versus conversion during enantioselective hydrogenation of activated ketones on a platinum catalyst

Article abstract of DOI:10.1016/j.jcat.2004.03.020

The initial transient period (ITP) was studied in the enantioselective hydrogenation of activated ketones (ethyl pyruvate, pyruvaldehyde dimethyl acetal, methyl benzoyl formate) on Pt-alumina (E4759) catalyst modified by cinchonidine (CD) or dihydrocincho

Full text of DOI:10.1016/j.jcat.2004.03.020

Journal of Catalysis 224 (2004) 463–472
www.elsevier.com/locate/jcat
New data on enantiomeric excess versus conversion during
enantioselective hydrogenation of activated ketones on a platinum catalyst
Katalin Balázsik a and Mihály Bartók a,b,∗
a
Organic Catalysis Research Group of the Hungarian Academy of Sciences, Dóm tér 8, H-6720 Szeged, Hungary
b
Department of Organic Chemistry, University of Szeged, Dóm tér 8, H-6720 Szeged, Hungary
Received 26 January 2004; revised 8 March 2004; accepted 12 March 2004
Available online 20 April 2004
Abstract
The initial transient period (ITP) was studied in the enantioselective hydrogenation of activated ketones (ethyl pyruvate, pyruvaldehyde
dimethyl acetal, methyl benzoyl formate) on Pt-alumina (E4759) catalyst modified by cinchonidine (CD) or dihydrocinchonidine (DHCD)
under mild experimental conditions in toluene (hydrogen pressure: 1 bar, temperature: 253–298 K, modifier concentration: 0.001–1 mmol/L).
The effects of temperature, modifier concentration, different pretreatments of the catalyst, and structure of the reactants on the relationship
of the enantiomeric excess (ee) and conversion were studied. Our results suggest that the ITP is affected not only by impurities/contaminants
present in the catalytic system but also by the competitive adsorption of reactants, modifier, and solvent. Consequently it may not be excluded
that the ITP can be considered an intrinsic feature of this type of enantioselective hydrogenation reaction.
2004 Published by Elsevier Inc.
Keywords: Hydrogenation; Enantioselective; Pt-alumina; Cinchonidine; Activated ketones; Enantiomeric excess versus conversion; Initial transient period
1. Introduction
The theoretical and practical importance of the heteroge-
neous catalytic enantioselective hydrogenation of activated
ketones has been demonstrated by widely known experi-
mental results. These results have been summarized and
reviewed continuously: the first detailed overview was pub-
lished in 1997 [1] and the most recent one in 2003 [2]. The
Scheme 1.
most extensively studied reaction is the hydrogenation of
Experimental data prove that in the course of the hydro-
ethyl pyruvate (EtPy) to ethyl lactate (EtLt) on Pt-alumina
genation of EtPy, maximal ee can be attained only after a
catalyst modified by cinchona alkaloids (Scheme 1).
certain extent of conversion, i.e., in the initial period ee is
Despite the diverse research done over the years in this
low. The following statements were made about this phe-
field, numerous tasks await solution before satisfactory an-
nomenon in the two reviews cited above [1,2]: “At the mo-
swers can be given to the questions that have arisen. One
ment it is not clear why the catalyst is less selective at the
such problem is the reason for the rapid increase in enan-
very beginning of the reaction” (1997) and “maximum enan-
tioselectivity (ee) in the initial transient period (ITP) of the
tioselectivity and rate are reached only after a certain period
conversion. Details of previous studies in this field (Table 1
of time. Despite a spirited debate, the reasons for this effect
[3–17]) are reviewed in the Discussion of the present work.
are still controversial” (2003).
It was established at an early stage of research that the
efficiency of chiral hydrogenation is significantly affected
not only by the experimental parameters, which play a de-
*
cisive role, but also by the methods of execution [6–21]. In
general, four different methods and their variants [1,2] have
Corresponding author. Fax: +36 62 544 200.
E-mail address: bartok@chem.u-szeged.hu (M. Bartók).
0021-9517/$ – see front matter
doi:10.1016/j.jcat.2004.03.020
2004 Published by Elsevier Inc.

464
K. Balázsik, M. Bartók / Journal of Catalysis 224 (2004) 463–472
Table 1
Published experimental data on enantiomeric excess versus conversion during enantioselective hydrogenation of EtPy on Pt-alumina catalysts
a
Entry
Substrate (S)
Modifier (M)
Solvent
Temp. (K)
H
pressure (bar)
ee
Pt /M
s
S/M
1210
3500
400
Conclusion
Ref.
max
2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
EtPy
CD
EtOH
Tol.
343
363
343
293
296
293
343
293
293
263–323
298
296
295
295
293
283
20
20
20
10
70
10
70
20
20
5
77
85
77
77
80
60
73
85
75
60
58
76
80
50
84
60
100
380
100
30
34
15
100
1000
1000
800
300
43
480
620
230
130
A
A
A
E
[3]
[3]
[3]
[4]
[5]
[6]
[7]
[8]
EtPBA
MePy
DHCD
DHCD
CD
CD, DHCD
DHCD
CD
DHCD
DHCD
DHCD
DHCD
CD
MeOH
EtOH
EtOH
EtOH
EtOH
Tol.
EtOH
PrOH
MA
Tol., EtOH
AcOH, EtOH
DCB
Tol.
PrOH
b
MePy
77
400
77
EtPy
MePy
E
b
A
C
E
E
C
E
C
B
B
E
C
B,D
E
EtPy
EtPy
EtPy
EtPy
5880
15,200
15,200
2900
380
5000
1800
770
[8]
[9,10]
[11]
[12]
[13]
[13]
[14]
[15]
[16]
[17]
c
EtPy
1,10
50
1
EtPy
EtPy
TFAP
EtPy
EtPy
CD
CD
DHCD
CD
1
100
7
6700
10,000
No new experimental data
EtPy CD
CH Cl
293
50
80
2050
8000
2
2
Abbreviation used: MePy, methyl pyruvate; EtPy, ethyl pyruvate; TFAP, trifluoroacetophenone; EtPBA, 2-oxo-4-phenylbutyric acid ethyl ester; DCB, 1,2-
dichlorobenzene; MA, methyl acetate; CD, cinchonidine; DHCD, dihydrocinchonidine.
a
(A) ITP was not observed. (B) Emphasize role of impurities/contaminants. (C) Intrinsic feature of catalytic system is stressed. (D) Results of several
superimposed effects. (E) There is no definite conclusion.
b
EuPt1.
Pt-silica.
c
been adopted for performing liquid-phase chiral hydrogena-
tion:
2. Experimental
2.1. Materials
(i) The catalyst is previously modified with the cinchona
in the appropriate solvent;
Cinchonidine (CD), toluene, EtPy, and PADA were pur-
chased from Fluka, and EtPy and MBF from Aldrich. EtPy
(Fluka), PADA, and MBF were distilled before use and were
98% pure. In the case of EtPy the impurities are ∼ 1%
racemic EtLt, and ∼ 1% unknown. EtPy distilled on a Vi-
greaux column was used in the majority of the experiments.
At later stages of the work, measurements were also made
with EtPy purchased from Aldrich and distilled before use
to a purity of 99.9% (0.1% EtOH). As evidenced by GC and
HPLC analysis, this EtPy contained no EtLt.
The impurities in CD as determined by HPLC are 1.1%
quinine and 2.6% quinidine. DHCD was prepared by hy-
drogenation of CD (Pd/C, 1 N H2SO4/H2O, 1 bar, 298 K)
and used after crystallization. A 1 mmol/L DHCD solution
was prepared in toluene and the required volume was added
to the reactor. (The 1 mmol/L DHCD solution contained
∼ 1% AcOH, because the solubility of DHCD in toluene
is limited.) Because of the good solubility of CD in toluene,
solutions containing 1 mmol/L CD can be made up without
difficulty.
(ii) All the necessary reaction components (catalyst, sol-
vent, modifier, reactant) are added to the reactor and
hydrogenation is started after setting the appropriate
hydrogen pressure and temperature;
(iii) The reaction is performed in fixed-bed reactors with
continuous feeding of reactant stream with and with-
out modifier;
(iv) To the mixture of the catalyst and solvent present in
the reactor, a solution of the modifier is introduced in
the presence of hydrogen under constant stirring, and
a solution of the material to be hydrogenated is next
added within a very short time (a few minutes).
In the case of academic research, the last of these methods
has proven to be the most practicable. The measurements
described in this article also followed this method (see Sec-
tion 2).
The majority of the experiments addressing ITP were
conducted at hydrogen pressures exceeding 1 bar in vari-
ous solvents (Table 1), whereas those described in this ar-
ticle were carried out at 1 bar hydrogen pressure in toluene.
Here we present our new results concerning the hydrogena-
tion of mainly EtPy and also pyruvaldehyde dimethyl acetal
(PADA) and methyl benzoylformate (MBF) on Pt-alumina
(E4759) catalyst modified by CD or DHCD.
Based on data in the literature [1], among several cata-
lysts the one most often used is Engelhard 4759 (E4759).
E4759 was pretreated before use in a fixed-bed reactor by
flushing with 30 mL min−1 helium at 300–680 K for 30 min
and 30 mL min−1 hydrogen at 680 K for 100 min. After be-
ing cooled to room temperature in hydrogen, the catalyst was

K. Balázsik, M. Bartók / Journal of Catalysis 224 (2004) 463–472
465
flushed with helium for 30 min and stored under air before
use.
followed by a dispute [15,16]; however, it has not been de-
cided to date [17] whether the initial increase in ee with
conversion is a “reaction-driven” phenomenon or is due to
the presence of impurities in the system.
2.2. Hydrogenation
Consequently it seemed promising to initiate serial mea-
surements under well-defined mild experimental conditions
(as this complicated system behaves with extreme sensitiv-
ity, even to nearly imperceptible changes in experimental
conditions). In general, it is not easy to participate in a de-
bate and to take sides. It is new experimental data that inspire
the formulation of a clearly defined opinion. In fact, it is
the recently published statement of Studer et al. cited above
[2] that encourages scientists in this field to publish their
observations, even though there has been little progress in
interpreting their results.
Variable parameters in our experiments were: catalyst
pretreatment, the method of introducing the reaction part-
ners, the temperature of hydrogenation,the concentrations of
CD and DHCD, the purity of EtPy, modifications in hydro-
genation conditions and the structures of activated ketones
(different reactants). We tried to set up experimental condi-
tions that minimize the role of impurities (toluene as selected
solvent, various pretreatments, fresh catalyst, addition of re-
actant to reaction mixture as the last component, use of EtPy
free of EtLt).
Hydrogenation was performed in an atmospheric batch
glass reactor with a volume of 10 mL [22]. The agitator
speed was 1200 rpm to avoid the diffusion range. The cat-
alytic system including catalyst and 2 or 5 mL toluene was
purged three times with hydrogen, and after re-reduction
(30 min), the calculated amount of modifier, and after 1 min
0.1 mL of reactants, was introduced and stirred in the pres-
ence of hydrogen for the required reaction time. Standard
conditions and standard abbreviations are: 25 mg E4759, 2
or 5 mL solvent, 1 bar hydrogen, 297–298 K, 1200 rpm,
0.1 mL reactant, ee = cumulative enantiomeric excess, ꢀee
(dashed lines) = actual ee. The quantification of conversion
and ee is based on GC data [22]. The actual (or incremen-
tal) ee was calculated as ꢀee = (ee2y2 − ee1y1)/(y2 − y1),
where y = the yield to EtLt, index 2 = sample subsequent to
sample 1. The measurements were reproduced several times
(an effect was accepted as significant only after several con-
vincing reproductions).
In the case of pretreatment in helium, the following
changes were made. After re-reduction, the catalytic system
(i.e., catalyst + toluene) was purged without mixing twice
with helium (99.996%, < 5 ppm oxygen) and after mixing in
helium (4 min) the calculated amount of modifier was intro-
duced and the system was stirred in the presence of helium
(1 min). The catalytic system was next purged with hydrogen
three times without stirring, 0.1 mL of reactant was added,
and the system was stirred in the presence of hydrogen for
the required reaction time.
3.1. Enantioselective hydrogenation of EtPy: effect of
temperature and DHCD concentration on the relationship
between ee and conversion
To study the ITP, of the two solvents (toluene, AcOH) en-
suring the highest ee (90–97%) in EtPy hydrogenation, the
practic choice was toluene because, on the one hand, the ITP
was less observable in AcOH [13] and, on the other hand,
our studies have suggested [23] that there is a far larger dif-
ference between the hydrogenation mechanisms in the two
solvents than was initially supposed [1,2]. Toluene was also
expected to serve as a more favorable solvent for these stud-
ies because it allows experiments on the ee/conversion re-
lationship under milder conditions (in AcOH, the mixture
freezes by 280–285 K, AcOH reacts with alumina). At a hy-
drogen pressure of 1 bar the reaction proceeds so fast that
the stage of 0–10% conversion is hardly measurable. Con-
sidering that ee is strongly dependent on the rate of hydrogen
transport [9,13], conditions and measurement techniques ex-
cluding the effect of mass transport processes on changes in
ee must be selected.
Earlier investigations have already demonstrated the ef-
fect of stirring rate, DHCD concentration, and solvent vol-
ume on conversion, which has been known to be closely
correlated with ee [24]: “a linear relationship should ex-
ist between the observed ee and the reciprocal observed
rate.” According to our preliminary experiments, measure-
ments within the kinetic range may be carried out under
the following experimental conditions: E4759 25 mg, hy-
drogen 1 bar, [DHCD] 0.001–0.1 mmol/L, EtPy ∼ 1 mM,
ESI-MS and ESI-MS/MS measurements were as described
in an earlier publication [22].
3. Results and discussion
The first experimental observations regarding the rela-
tionship between ee and conversion in the hydrogenation of
EtPy were published by Blaser and co-workers in 1989 [3].
As these authors had not studied the conversion range 0–
20%, they could not observe any changes. The phenomenon
was later investigated by several research groups and diverse
conclusions were drawn depending on the experimental con-
ditions (Table 1). After reviewing the experimental results
and completing various novel experiments, Blackmond and
co-workers [9,10] and Margitfalvi et al. [12], on the one
hand, and Baiker and co-workers [13] on the other, proposed
explanations for the ITP. According to the former authors
[9,10,12], a change in ee versus conversion is an intrinsic
feature of the catalytic system, whereas in the opinion of
the latter authors [13], the phenomenon is caused mainly by
impurities. Publication of the different interpretations was

466
K. Balázsik, M. Bartók / Journal of Catalysis 224 (2004) 463–472
Table 2
a
Experimental data on enantioselective hydrogenation of EtPy on Pt-alumina catalyst
Entry
Modifier (mmol/L)
Temp. (K)
Pretreatment
Time (min)
Conv. (%)
Rate
ee (%)
1
2
3
4
5
6
7
8
9
0.01
0.01
0.01
0.01
0.001
0.001
0.001
0.1
298
298
298
298
298
273
263
263
263
253
263
Hydrogen + DHCD 0 min
Hydrogen + DHCD 10 min
Hydrogen + DHCD 30 min
Hydrogen + DHCD 60 min
Hydrogen
Hydrogen
Hydrogen
Hydrogen
Hydrogen
16
20
25
25
20
50
60
30
31
75
60
99
99
99
98
98
97
51
98
97
100
51
3.6
3
2.4
2.4
3
1.2
0.4
1.8
1.6
0.5
0.4
66
40
31
24
9
18
28
85
86
86
29
0.01
0.01
0.001
10
11
Hydrogen
Hydrogen
a
Standard conditions, 5 mL toluene, Entries 1–4: 2 mL toluene, EtPy 98% purity, rate: mmol/(min g).
263–298 K, toluene 5 mL, mixing efficiency level 7–9 (level
9 = 1300 min−1). In this case ([DHCD] = 0.01 mmol/L),
Pts/DHCD = 35 and EtPy/DHCD = 20000.
Before discussion of the results it is necessary to men-
tion that in studies on the reaction mechanism it was found
that the modifier is also converted in the course of hydro-
genation [21,25–28], in the presence of EtPy, however, this
hydrogenation can be suppressed. The slowdown of hydro-
genation and the decrease in ee as a function of the duration
of catalyst pretreatment in the presence of DHCD are well
demonstrated in Table 2 (entries 1–4).
To our best knowledge, the relationship between ee and
conversion has been studied mainly in alcohols, at rela-
tively high pressures and within a narrow range of modi-
fier concentrations (Table 1). Studies have also been con-
ducted in toluene and recently in dichloromethane; even in
this case, however, conditions were quite severe (relatively
high modifier concentration, temperatures: room tempera-
ture and above, hydrogen pressures of 20–100 bar). We have
found no published results obtained under mild experimen-
tal conditions (modifier concentrations 0.001–0.1 mmol/L,
hydrogen pressure 1 bar, room temperature and below). Be-
cause of the high hydrogenation rate, in some measurements
the concentration of DHCD was kept low (0.001 mmol/L)
to study the relationship between ee and conversion. Accord-
ing to the evidence presented in Table 2 (entries 5–11) and
Figs. 1 and 2, the ITP depends both on the temperature and
on the initial concentration of DHCD, namely, the lower the
initial concentration of DHCD, the more pronounced the in-
crease in ee. This is especially well shown by the change
in ꢀee.
The data in Table 2 illustrate the well-known fact that the
reaction rate increases with increasing temperature. Fig. 1
not only verifies the existence of the ITP, but also demon-
strates that, in the case of hydrogenation in toluene at a
DHCD concentration of 0.001 mmol/L, ee and ꢀee also
increase continuously above 20% conversion, except for
hydrogenation at 296 K. At 296 K, over ∼ 10% conver-
sion (time of reaction is 2–3 min) ꢀee decreases, probably
because—due to hydrogenation of the quinoline skeleton of
DHCD [21,25–28]; the concentration of DHCD responsi-
Fig. 1. Enantioselective hydrogenation of EtPy (98% purity): effect of tem-
perature and conversion on ee and ꢀee (dashed lines). Standard conditions,
[DHCD] = 0.001 mmol/L, toluene = 5 mL.
ble for enantioselection also decreases (Fig. 1). The same
is shown by the data in Fig. 2. The data obtained at 253 K
using EtPy of 99.9% purity are especially noteworthy. At
this temperature, 100% conversion can be attained only in
75 min. Despite this, ee and ꢀee did not decrease, because
hydrogenation of the quinoline skeleton does not take place
at 253 K. Within the ITP, the change in ꢀee is somewhat
faster in the case of 98% EtPy than when EtPy of 99.9% pu-
rity is used.
Changes in the composition of the hydrogenated cin-
chona derivatives desorbed from the catalyst are summa-
rized in Table 3. The relative abundances of hydrogenated
cinchonas increase with increasing conversion of EtPy. Be-
cause the adsorption strength of hydrogenated cinchonas is
considerably lower than that of DHCD, they are easily des-
orbed at high DHCD concentrations, whereas at low DHCD
concentrations (below 0.01 mmol/L) ꢀee will decrease, be-

K. Balázsik, M. Bartók / Journal of Catalysis 224 (2004) 463–472
467
Table 3
+
a
Relative abundances of [M + H] ions in the ESI mass spectra of cinchonas formed from DHCD in the enantioselective hydrogenation of EtPy
DHCD
Time
Conv.
m/z values (relative peak intensity, %)
+
+
+
+
(mmol/L)
(min)
(%)
[296 + H]
100
100
43
47
8
[298 + H]
[300 + H]
[306 + H]
ee (%)
0.001
0.001
0.01
0.1
0.1
1
15
30
10
9
15
10
25
30
60
43
77
99
−
−
–
14
10
52
79
79
79
69
−
−
–
–
–
28
–
37
100
100
100
24
18
35
34
70
65
100
100
100
1
40
–
a
Standard conditions, 2 mL toluene. m/z = 296 (DHCD: dihydrocinchonidine). m/z = 298 (THCD: tetrahydrocinchonidine). m/z = 300 (HHCD: hexahy-
drocinchonidine). m/z = 306 (DDHCD: dodecahydrocinchonidine).
ties was verified by the experimental data obtained by Baiker
and co-workers [13,31], it cannot be excluded that the phe-
nomenon is an intrinsic feature of the system [9,10,12]. To
elucidate this problem and also to obtain more information
for our suggestion [23] on the chemisorption mechanism of
the hydrogenation of EtPy in toluene, the methods of pre-
treatment of catalyst E4759 were studied. As attention had
earlier been called to the effect of purity [2,20] and origin
[32] of EtPy on ee, it also seemed practic to study EtPy of
different purities. The results obtained are summarized in
Figs. 3–5 and Table 4.
Fig. 3 and Table 4 (entries 1–4) demonstrate the effect of
the pretreatment conditions of the re-reducted catalyst on the
rate of enantioselective hydrogenation of EtPy (purity: 98%)
and on the relationship between ee/ꢀee and conversion. In
these experiments the catalyst was pretreated in toluene with
hydrogen or helium, as described under Section 2. The ex-
perimental data allow the following direct conclusions to be
drawn.
Fig. 2. Enantioselective hydrogenation of EtPy (98% purity): effect of
The most striking observation is that in the course of
DHCD treatment of the catalyst in the absence of hydrogen
(i.e., in helium), significantly more active chiral surface sites
are generated. We consider this to be both a significant and
surprising observation. Not only is hydrogenation faster on
these more active sites, but they also allow higher ee val-
ues to be attained, as has been reported. Reaction rate and
ee also increase with increasing DHCD concentration. The
initial increase in the ITP is quite conspicuous, especially in
ꢀee, which exceeds 90% despite the fact that the material to
be hydrogenated contains EtPy of only 98% purity.
Figs. 4a–c and Table 4 (entries 5–10) represent hydro-
genation on catalysts pretreated in the same way, with the
significant difference that EtPy of 99.9% purity, containing
no EtLt at all, was used. In this case the effect of temperature
was also studied under otherwise identical experimental con-
ditions. The data obtained are essentially identical to those
in Fig. 3. However, there is a significant difference: the data
measured at 298 K show that over a catalyst pretreated in
hydrogen, ꢀee significantly decreases, even after 20 min re-
action in the case of 0.01 mmol/L DHCD concentration.
The most important result of the pretreatment studies in
helium is that ee values 15–20% higher than those without
DHCD concentrations (0.001–0.1 mmol/L) and conversion on ee and ꢀee.
Standard conditions, toluene = 5 mL. EtPy 99.9% purity.
*
cause there is insufficient DHCD present to produce optimal
surface coverage [23]. At 0.001 mmol/L DHCD, when the
amount of DHCD present is sufficient to cover only 10% of
the Pt surface [22], hydrogenated cinchonas do not appear in
the solution, probably due to the detection limit of the ana-
lytical method used and irreversible adsorption [29,30]. At
DHCD concentrations above 0.01 mmol/L, desorbing hy-
drogenation products of the quinoline skeleton of DHCD
(THCD, HHCD, DDHCD) are readily identified in the liquid
phase.
3.2. Enantioselective hydrogenation of EtPy: effect of
catalyst pretreatment on the relationship between ee and
conversion
The studies described above verified that despite a great
variation in experimental conditions, the ITP was detectable
in every case. These studies, however, cannot explain the
cause of this transient stage. Although the role of impuri-

468
K. Balázsik, M. Bartók / Journal of Catalysis 224 (2004) 463–472
Fig. 3. Enantioselective hydrogenation of EtPy (98% purity): pretreatment
of re-reduced catalyst in toluene with hydrogen or helium: effect of pretreat-
ment conditions on relationship between conversion and ee/ꢀee. Standard
conditions, [DHCD] = 0.001 and 0.01 mmol/L, 263 K, toluene = 5 mL.
Table 4
Experimental data on enantioselective hydrogenation of EtPy on Pt-alumina
a
catalyst
Entry Modifier
Temp. Pretreatment Time Conv. Rate ee
(min) (%) (%)
(mmol/L) (K)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
0.01
0.01
0.001
0.001
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
263
263
263
263
298
298
273
273
263
263
298
298
298
298
Hydrogen
Helium
Hydrogen
Helium
Hydrogen
Helium
Hydrogen
Helium
Hydrogen
Helium
Air
Hydrogen
Hydrogen
Hydrogen
32
20
90
60
20
10
30
25
35
22
13
16
19
16
99
99
60
50
99
100
99
99
100
99
99
1.4
2.1
0.3
0.4
4
6.7
1.7
1.9
1.2
1.8
4.9
5.3
3.4
4.0
80
90
28
48
56
89
79
89
88
92
75
64
75
82
98
97
96
a
Standard conditions, 5 mL toluene, entries 1–4: EtPy 98% purity, en-
tries 5–14: EtPy 99.9% purity, rate: mmol/(min g).
pretreatment in helium can be attained. In our opinion the
∼ 4-min treatment of the catalyst with helium removes the
surface impurities (e.g., CO) retained on the catalyst surface
during the 30-min re-reduction. On the other hand, it may
not be excluded that the oxygen present in helium oxidizes
the hydrogen atoms chemisorbed on the Pt surface, thereby
generating new Pt clusters, which in turn can form chiral sur-
face sites or complexes of higher activity with DHCD or CD
via surface interaction or chemisorption and increase hydro-
genation rate, ee, and ꢀee.
Fig. 4. Enantioselective hydrogenation of EtPy (99.9% purity): effects of
pretreated conditions of catalyst and temperature on ee/ꢀee versus con-
version. Standard conditions, [CD] = 0.01 mmol/L, toluene = 5 mL:
(a) 298 K, (b) 273 K, (c) 263 K.
Thus, the explanation of the role of helium pretreatment
most probably lies in the presence of 5 ppm oxygen in he-

K. Balázsik, M. Bartók / Journal of Catalysis 224 (2004) 463–472
469
lium. When helium was replaced by air, pretreatment gave
results similar to those obtained with helium (Table 4, en-
try 11). The effect of oxygen on the adsorption of CD and
DHCD and on ee was studied in several laboratories under
various conditions [17,33–39]. When interpreting the effect,
opposite conclusions were drawn, probably due to differ-
ences in experimental conditions and measurement tech-
niques. As the measurement procedures described are not
available to us, we had no opportunity to analyze their main
conclusions.
Further experimental results yielding new information
about the ITP are presented in Fig. 5 and Table 4 (entries 12–
14). If the intrinsic feature of the catalytic system, or, in other
words, the “reaction-driven equilibration” [9,10], had no ef-
fect on the ITP, there should be no great difference between
the results obtained on catalysts on which racemic hydro-
genation of EtPy and, after removal of the reaction mixtures,
chiral hydrogenation of EtPy (Fig. 5b) have been performed,
on the one hand, and the results of the original chiral hydro-
genation (Fig. 5a), on the other hand. In a similar fashion,
a large difference cannot be expected in hydrogenation over
a catalyst that was previously used in the enantioselective
hydrogenation of another substrate (e.g., PADA) (Fig. 5c).
The data in Fig. 5, however, call attention to significant dif-
ferences exceeding experimental error, especially in the case
of curves representing the e ꢀee, i.e., chiral hydrogenations
following racemic hydrogenation (Fig. 5b) and PADA hy-
drogenation (Fig. 5c): (i) the ITP is clearly recorded in all
three cases; (ii) ee and ꢀee increased considerably in both
cases (b,c) as compared with the control (a). In the case of
(c), the decrease in ꢀee with progressing conversion may
be explained by deactivation of the catalyst during the pro-
longed reaction (90 min).
Fig. 5. Enantioselective hydrogenation of EtPy (99.9% purity) over cat-
alyst under standard conditions (a), after racemic hydrogenation of EtPy
(b), and after enantioselective hydrogenation of PADA (c): effect of conver-
sion on ee/ꢀee. Standard conditions, [CD] = 0.01 mmol/L, toluene 5 mL,
PADA = 0.1 mL.
Considering that, on reuse of the catalyst, ee increased
significantly despite the increase in contamination, the ITP
cannot be accounted for solely by the effect of various im-
purities/contaminants. Consequently, in the case of the enan-
tioselective hydrogenation of EtPy in toluene, the ITP of the
ee/conversion relationship can also be considered an intrin-
sic feature of the catalytic system. As the above-described
results concerning the ee/conversion relationship (Fig. 5)
strongly support the effect of modified catalyst, they serve as
further evidence for the so-called adsorption model of chiral
induction (reviewed in Refs. [1,2,29]).
Fig. 6. Enantioselective hydrogenation of PADA: effects of temperature and
CD concentrations (0.01 and 0.1 mmol/L) on relationship between conver-
sion and ee/ꢀee. Standard conditions, 2 mL toluene.
3.3. Relationship between ee and conversion: effect of the
structure of reactants
To test whether the phenomenon observed during chiral
hydrogenation of EtPy (i.e., the significant increase in ee
with increasing conversion during the ITP) is also detectable
in the case of other activated ketones, serial measurements
were carried out using PADA and MBF. Enantioselective
hydrogenation of PADA and ethyl benzoyl formate under
identical conditions gave outstandingly high ee (96–98%)
[40,41]. MBF was selected as model compound for the
experiments presented here, because it can be prepared at
higher purity than the ethyl ester. The reaction rate increases
in the order PADA < MBF < EtPy. The results are summa-
rized in Figs. 6 and 7 and Table 5. The main conclusions are
as follows.
Chiral hydrogenation of PADA (Fig. 6; Table 5, entries
1–3) proceeds considerably more slowly than that of EtPy
and MBF [41]. Sixty to seventy percent conversion of PADA

470
K. Balázsik, M. Bartók / Journal of Catalysis 224 (2004) 463–472
Table 5
Based on the above results, it appears that an ITP with a
rapid increase in ee up to ca. 20% conversion cannot be ob-
served in the case of slow chiral hydrogenations probably
because of various competitions and strong adsorption of re-
actant.
Experimental data on enantioselective hydrogenation of PADA and MBF
on Pt-alumina catalyst
a
Entry Modifier
Temp. Pretreatment Time Conv. Rate ee
(min) (%) (%)
(mmol/L) (K)
1
2
3
4
5
6
0.1
0.1
0.1
0.1
1
298
298
263
298
298
263
Hydrogen
Hydrogen
Hydrogen
Hydrogen
Hydrogen
Hydrogen
120
90
90
12
12
30
84
76
66
100
92
74
0.7
1.2
0.4
3.3
2.9
1.1
21
65
87
80
89
90
4. Interpretation of results and conclusion
1
The role of the presence of contaminants in the ITP is
quite obvious and has been convincingly verified by ap-
propriate experiments [13]. It is, however, more difficult to
prove that the phenomenon “is an intrinsic feature of the
catalytic system.” Based on the results published in the lit-
erature concerning the interpretation of the initial transient
behavior observed during chiral hydrogenation of EtPy as
well as the present results, the cleaning effect of hydrogen
cannot be excluded, but neither can the assumption that the
ITP is an intrinsic feature of the catalytic system. The dom-
inance of one or the other standpoint is profoundly affected
or, in some cases, even determined by the experimental con-
ditions, because the system reacts very sensitively even to
minimal, hardly perceivable changes.
a
Standard conditions, entries 1–3 (PADA): 2 mL toluene, entries 4–6
(MBF): 5 mL toluene, rate: mmol/(min g).
The following experimental results described in the
present work, concerning the relationship between ee and
conversion, indicate the role of factors that play a decisive
role in the development and operation of the chiral catalytic
system:
(i) New experimental data on the ITP, obtained in toluene,
under mild conditions never studied before;
(ii) The data of experiments using EtPy of 98% purity as
well as with 99.9% EtPy containing not even traces of
EtLt;
(iii) The increase in reaction rate and ee brought about by
the in situ helium pretreatment of the catalyst pretreated
at 673 K;
Fig. 7. Enantioselective hydrogenation of MBF: effects of temperature and
CD concentrations (0.1 and 1 mmol/L) on relationship between conversion
and ee/ꢀee. Standard conditions, 5 mL toluene.
takes nearly 90 min, whereas hydrogenation of EtPy or MBF
is completed within 15–30 min, depending on the temper-
ature. In the case of this slow hydrogenation the ITP is
essentially nondetectable; ꢀee slowly increases at a rate
depending on modifier concentration and temperature and
later slowly decreases due to hydrogenation of the modi-
fier. When PADA is hydrogenated at a CD concentration
of 0.01 mmol/L—a value below the optimal modifier con-
centration necessary for chiral hydrogenation—a slight ITP
is already observed: ꢀee continuously increases up to 10%
conversion (Fig. 6). A similar phenomenon can be detected
in the case of hydrogenation at 263 K. The probable reason
for the latter two phenomena is a slower development of the
optimal chiral environment on the Pt surface.
(iv) The results of the second hydrogenation that followed
racemic and chiral hydrogenation;
(v) New data on the ITP of the chiral hydrogenation of
other reactants (PADA, MBF).
These results suggest that the ITP is an intrinsic feature of
this type of reaction.
In our opinion the assumption that the initial transient
behavior of the chiral hydrogenation of EtPy is an intrin-
sic feature of the system is also supported rather than ex-
cluded by the results of several very important studies pub-
lished recently [17,28,39]: the ITP has been observed with
premodified catalysts [17], and, furthermore, the different
adsorption strengths of CD and quinidine, which result in
enrichment of CD on the Pt surface [28], both prove the
development of the chiral environment responsible for enan-
tioselectivity. An intrinsic character may also be attributed to
the adsorbate-induced restructuring of the platinum surface
[36,39], which was verified at the solid/liquid interface at
The hydrogenation of MBF (Fig. 7; Table 5, entries 4–
6) is a fast reaction, and there is strong competition be-
tween CD and MBF. The ITP can therefore be studied only
at higher modifier concentrations. The increase in ꢀee up
to ∼ 10% conversion can also be observed in these cases.

K. Balázsik, M. Bartók / Journal of Catalysis 224 (2004) 463–472
471
Fig. 8. Proposed adsorbed complexes on platinum.
room temperature [42,43], and the same holds for the results
of sonochemical pretreatment of Pt-alumina in the presence
of CD [44].
Of course, it also cannot be excluded that the irreversibly ad-
sorbed modifier does not participate in the enantioselection
but acts only as a spectator.
Differences in mechanisms are also indicated by new re-
sults [40,52–54] of studies on the widely accepted relation-
ship between chiral hydrogenation rate and enantioselectiv-
ity [20]. According to these new results, high ee may also
be realized at low reaction rates. This indicates that ligand
acceleration is a welcome effect but not a necessary prereq-
uisite for achieving high enantioselectivity.
At the same time, the results presented in this work also
indicate that the change in the ee/conversion relationship can
be minimized by the appropriate choice of reaction condi-
tions. With respect to the ITP, the role of the competitive
adsorption of the chiral modifier, the reactant, the solvent,
and hydrogen must naturally also be taken into account.
Considering that the chemical system studied is extremely
complicated and also that not all the details of the origin of
chiral induction are fully known, the observed phenomenon
may also be the result of several sumperimposed effects and
its explanation necessitates further studies and new experi-
mental evidence.
According to the reaction mechanism of enantioselective
hydrogenation of activated ketones in toluene [1,2], half-
hydrogenated EtPy is bonded to the nitrogen of the quin-
uclidine, also via hydrogen bond (Fig. 8A). The hydrogen
source of the latter is the hydrogen dissociatively adsorbed
on the platinum. New suggestions have recently been pub-
lished, which can be supplemented and, as soon as further
studies yield new experimental evidence, may modify the
presently accepted picture [23,45,46]. According to an as-
sumption DHCD binds the ketone, as a nucleophile (Fig. 8B)
[45,46].
We proposed [23] that in toluene the modifier molecules
participate in the formation of the surface complex as ligands
of the surface Pt atoms (Figs. 8C and 8D). The formation of
a surface complex C is supported by the following data:
(i) Metal complexes containing cinchona ligands of sim-
ilar structure have been prepared and their structures
determined by X-ray diffraction [47];
(ii) The mechanism of homogeneous catalytic hydrogena-
tion was confirmed with the help of this type of metal
complexe [48];
(iii) Extremely high ee values (97%) were achieved using
Pt colloid with 1.2-nm particle size [49] that are not
accounted for by the intermediates proposed to date;
(iv) Only the highly reactive and coordinatively unsatu-
rated corner and adatoms participate in chiral induc-
tion [20];
Acknowledgments
Financial support by the Hungarian National Science
Foundation (OTKA Grants TS 044690 and T 042466) is
highly appreciated. The authors are grateful to Professor F.
Notheisz (University of Szeged) for valuable discussions.
References
(v) It may be verified by IR [50,51], and by electrochem-
ical [30] and surface science [34] measurements that
DHCD is irreversibly adsorbed on platinum.
[1] H.-U. Blaser, H.-P. Jalett, M. Müller, M. Studer, Catal. Today 37
(1997) 441.

472
K. Balázsik, M. Bartók / Journal of Catalysis 224 (2004) 463–472
[2] M. Studer, H.-U. Blaser, C. Exner, Adv. Synth. Catal. 345 (2003) 45.
[3] J.T. Wehrly, A. Baiker, D.M. Monti, H.-U. Blaser, H.-P. Jalett, J. Mol.
Catal. 57 (1989) 245.
[4] I.M. Sutherland, A. Ibbotson, R.B. Moyes, P.B. Wells, J. Catal. 125
(1990) 77.
[5] J.L. Margitfalvi, P. Marti, A. Baiker, L. Botz, O. Sticher, Catal. Lett. 6
(1990) 281.
[6] P.A. Meheux, A. Ibbotson, P.B. Wells, J. Catal. 128 (1991) 387.
[7] J.L. Margitfalvi, B. Minder, E. Tálas, L. Botz, A. Baiker, in: L.
Guczi, F. Solymosi, P. Tétényi (Eds.), New Frontiers in Catalysis: Pro-
ceedings 10th International Congress on Catalysis, Budapest, 1992,
Akadémiai Kiadó, Budapest, 1993, p. 2471.
[8] H.-U. Blaser, M. Garland, H.-P. Jalett, J. Catal. 144 (1993) 569.
[9] U.K. Singh, R.N. Landau, Y. Sun, C. LeBlond, D.G. Blackmond, S.K.
Tanielyan, R.L. Augustine, J. Catal. 154 (1995) 91.
[10] J. Wang, Y. Sun, C. LeBlond, R.N. Landau, D.G. Blackmond, J. Ca-
tal. 161 (1996) 752.
[11] R.L. Augustine, S.K. Tanielyan, J. Mol. Catal. A 112 (1996) 93.
[12] J.L. Margitfalvi, M. Hegedu˝s, E. Tfirst, Tetrahedron Asymmetry 7
(1996) 571;
J.L. Margitfalvi, E. Tálas, E. Tfirst, C.V. Kumar, A. Gergely, Appl.
Catal. A 191 (2000) 177.
[13] T. Mallat, Z. Bodnar, B. Minder, K. Borszeky, A. Baiker, J. Catal. 168
(1997) 183.
[14] H.-U. Blaser, D. Imhof, M. Studer, Stud. Surf. Sci. Catal. 108 (1997)
175.
[15] D.G. Blackmond, J. Catal. 176 (1998) 267.
[16] T. Mallat, A. Baiker, J. Catal. 176 (1998) 271.
[17] X. Li, R.P.K. Wells, P.B. Wells, G.J. Hutchings, Catal. Lett. 89 (2003)
163.
[18] Y. Orito, S. Imai, S. Niwa, J. Chem. Soc. Jpn. (1979) 1118.
[19] Y. Orito, S. Imai, S. Niwa, J. Chem. Soc. Jpn. (1980) 670.
[20] R.L. Augustine, S.K. Tanielyan, L.K. Doyle, Tetrahedron Asymme-
try 4 (1993) 1803.
[27] C. LeBlond, J. Wang, J. Liu, A.T. Andrews, Y.-K. Sun, J. Am. Chem.
Soc. 121 (1999) 4920.
[28] W.-R. Huck, T. Bürgi, T. Mallat, A. Baiker, J. Catal. 216 (2003) 276.
[29] P.B. Wells, R.P.K. Wells, in: D.E. De Vos, I.F.J. Vankelecom, P.A.
Jacobs (Eds.), Chiral Catalyst Immobilization and Recycling, Wiley–
VCH, Weinheim, 2000, p. 130, (6.3.1. The adsorption model).
[30] I. Bakos, S. Szabó, M. Bartók, E. Kálmán, J. Electroanal. Chem. 532
(2002) 113.
[31] N. Künzle, R. Hess, T. Mallat, A. Baiker, J. Catal. 186 (1999) 239.
[32] H.-U. Blaser, H.-P. Jalett, F. Spindler, J. Mol. Catal. A 107 (1996) 85.
[33] 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.
[34] A.F. Carley, M.K. Rajumon, M.W. Roberts, P.B. Wells, J. Chem. Soc.,
Faraday Trans. 91 (1995) 2167.
[35] R.L. Augustine, S.K. Tanielyan, J. Mol. Catal. A 118 (1997) 79.
[36] T. Mallat, S. Frauchiger, P.J. Kooyman, M. Schürch, A. Baiker, Catal.
Lett. 63 (1999) 121.
[37] S.P. Griffiths, P. Johnston, P.B. Wells, Appl. Catal. A 191 (2000) 193.
[38] E. Toukoniitty, P. Mäki-Arvela, A.N. Villela, A.K. Neyestanaki, T.
Salmi, R. Leino, R. Sjöholm, E. Laine, J. Väyrynen, T. Ollonqvist,
P.J. Kooyman, Catal. Today 60 (2000) 175.
[39] Z. Ma, J. Kubota, F. Zaera, J. Catal. 219 (2003) 404.
[40] M. Sutyinszki, K. Szöri, K. Felföldi, M. Bartók, Catal. Commun. 3
(2002) 125.
[41] K. Felföldi, K. Balázsik, M. Bartók, J. Mol. Catal. A 202 (2003) 163.
[42] D. Ferri, T. Bürgi, A. Baiker, J. Phys. Chem. B 105 (2001) 3187.
[43] R. Hess, F. Krumeich, T. Mallat, A. Baiker, Catal. Lett. 92 (2004) 141.
[44] B. Török, K. Balázsik, M. Török, Gy. Szöllösi, M. Bartók, Ultrason.
Sonochem. 7 (2000) 151.
[45] M. Bartók, M. Sutyinszki, K. Felföldi, Gy. Szöllösi, Chem. Commun.
(2002) 1130.
[21] M. Bartók, Gy. Szöllösi, K. Balázsik, T. Bartók, J. Mol. Catal. A 177
(2002) 299.
[22] M. Bartók, K. Balázsik, Gy. Szöllo˝si, T. Bartók, J. Catal. 205 (2002)
168.
[23] M. Bartók, K. Balázsik, F. Notheisz, React. Kinet. Catal. Lett. 77
(2002) 363;
M. Bartók, K. Balázsik, T. Bartók, Z. Kele, Catal. Lett. 87 (2003) 235.
[24] M. Garland, H.-U. Blaser, J. Am. Chem. Soc. 112 (1990) 7048.
[25] V. Morawsky, U. Prüsse, L. Witte, K.D. Vorlop, Catal. Commun. 1
(2000) 15.
[46] G. Vayner, K.N. Houk, Y.-K. Sun, J. Am. Chem. Soc. 126 (2004) 199.
[47] R. Hubel, K. Polborn, W. Beck, Eur. J. Chem. (1999) 471.
[48] J. Halpern, Science 217 (1982) 401.
[49] X. Zuo, H. Liu, D. Guo, X. Yang, Tetrahedron 55 (1999) 7787.
[50] D. Ferri, T. Bürgi, J. Am. Chem. Soc. 123 (2001) 12074.
[51] J. Kubota, F. Zaera, J. Am. Chem. Soc. 123 (2001) 11115.
[52] J.U. Köhler, J.S. Bradley, Catal. Lett. 45 (1997) 203.
[53] M. von Arx, N. Dummer, D.J. Willock, S.H. Taylor, R.P.K. Wells, P.B.
Wells, G.J. Hutchings, Chem. Commun. (2003) 1926.
[54] E. Toukoniitty, B. Sevciková, P. Maki-Arvela, J. Warna, T. Salmi,
D.Yu. Murzin, J. Catal. 213 (2003) 7.
[26] U. Böhmer, F. Franke, K. Morgenschweis, T. Bieber, W. Resche-
tilowski, Catal. Today 60 (2000) 167.