Enantioselective hydrogenation of ethyl pyruvate catalyzed by α- and β-isocinchonine-modified Pt/Al2O3 in acetic acid

Article abstract of DOI:10.1016/S0021-9517(03)00250-1

The enantioselective hydrogenation of ethyl pyruvate (EtPy) was studied on Pt-alumina catalysts modified by high-purity α-isocinchonine (α-ICN) and β-isocinchonine (β-ICN) in AcOH. The effect of the modifier concentration, temperature, and hydrogen pressure on the reaction rate and the enantioselectivity was examined. Using the Engelhard 4759 catalyst under medium experimental conditions (273 K, hydrogen pressure of 25 bar, α-ICN concentration of 1 mmol/L) an optical yield of 93-94% can be achieved. In the case of β-ICN, maximum ee is 70-72% (297 K, 20 bar, β-ICN concentration of 1 mmol/L); at the same time, the rate of EtPy hydrogenation is higher than in the case of dihydrocinchonine (DHCN) or α-ICN. Chiral modifiers themselves are converted under the conditions of hydrogenation. The results of ESI-MS, ESI-MS-MS, HPLC-MS, and desorption measurements strongly suggest that α-ICN and β-ICN compounds with a rigid structure are responsible for chiral induction, since DHCN and hydrogenated products of DHCN were not identifiable among the reaction products. These experimental results support the earlier assumption that it is cinchona alkaloids in the "open 3" (in another terminology "antiopen") conformation that participate in enantioselection. However, based on other experimental observations of this work, the participation of other, so far unknown factors in enantioselection may not be excluded either.

Full text of DOI:10.1016/S0021-9517(03)00250-1

Journal of Catalysis 220 (2003) 207–214
www.elsevier.com/locate/jcat
Enantioselective hydrogenation of ethyl pyruvate catalyzed by α- and
β-isocinchonine-modified Pt/Al2O3 in acetic acid
Mihály Bartók,a,b,∗ Mária Sutyinszki,b and Károly Felföldi 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 25 February 2003; revised 5 May 2003; accepted 23 May 2003
Abstract
The enantioselective hydrogenation of ethyl pyruvate (EtPy) was studied on Pt-alumina catalysts modified by high-purity α-isocinchonine
(α-ICN) and β-isocinchonine (β-ICN) in AcOH. The effect of the modifier concentration, temperature, and hydrogen pressure on the reaction
rate and the enantioselectivity was examined. Using the Engelhard 4759 catalyst under medium experimental conditions (273 K, hydrogen
pressure of 25 bar, α-ICN concentration of 1 mmol/L) an optical yield of 93–94% can be achieved. In the case of β-ICN, maximum ee is
70–72% (297 K, 20 bar, β-ICN concentration of 1 mmol/L); at the same time, the rate of EtPy hydrogenation is higher than in the case
of dihydrocinchonine (DHCN) or α-ICN. Chiral modifiers themselves are converted under the conditions of hydrogenation. The results of
ESI-MS, ESI-MS-MS, HPLC-MS, and desorption measurements strongly suggest that α-ICN and β-ICN compounds with a rigid structure
are responsible for chiral induction, since DHCN and hydrogenated products of DHCN were not identifiable among the reaction products.
These experimental results support the earlier assumption that it is cinchona alkaloids in the “open 3” (in another terminology “antiopen”)
conformation that participate in enantioselection. However, based on other experimental observations of this work, the participation of other,
so far unknown factors in enantioselection may not be excluded either.
2003 Elsevier Inc. All rights reserved.
Keywords: Hydrogenation; Enantioselective; Pt/Al O ; α- and β-isocinchonine; Ethyl pyruvate; ESI-MS; Intermediate; Conformation
2
3
1. Introduction
H2
H2
Pt-cinchonine
or Pt-quinidine
Pt-cinchonidine
or Pt-quinine
The significance of asymmetric syntheses, and especially
of heterogeneous catalytic reactions [1–3], has gained wide
recognition. In this field, the enantioselective hydrogena-
tion of activated ketones has received special attention [3,4].
Compounds most widely used for chiral transfer in asym-
metric syntheses are cinchona alkaloids [5,6] (Fig. 1).
Despite its 22-year-old history [7], the intensity of re-
search on the asymmetric hydrogenation of α-ketoesters
over cinchona–alkaloid-modified platinum catalysts still
keeps increasing. The preferred model compound of these
elaborate studies is ethyl pyruvate (EtPy), which could be
hydrogenated to ethyl lactate (EtLt) with an extremely high
enantioselectivity (96–97%) [8,9].
Successful applications with ee values usually higher
than 95% ee were recently reported including a cyclic ke-
toester, α-ketoesters, α-ketoacetals, α-ketodicarboxylic acid
esters, and 2,4-diketoesters [9–14]. Besides improving the
dihydrocinchonidine(DHCD)–platinum catalyst system, ex-
tensive efforts have been made in developing a reliable
mechanistic interpretation. Although this topic was exten-
sively reviewed in recent years [2–4], no significant new
details concerning the reaction mechanism have been pub-
lished since. The most important task in this respect would
be the elucidation of the origin of chiral induction, i.e., the
verification of the structure of the intermediate responsi-
ble for chiral induction. Numerous mechanistic details have
been described, including data (i) on the conformation of
cinchona alkaloids and other chiral modifiers [15–18], (ii)
how DHCD and EtPy are adsorbed and on which surface
site [8,19–29], and (iii) the structure of the intermediate
*
Corresponding author.
E-mail address: bartok@chem.u-szeged.hu (M. Bartók).
0021-9517/$ – see front matter
2003 Elsevier Inc. All rights reserved.
doi:10.1016/S0021-9517(03)00250-1

208
M. Bartók et al. / Journal of Catalysis 220 (2003) 207–214
Fig. 3. The structure of α-isocinchonine (α-ICN) and β-isocinchonine
(β-ICN).
Fig. 1. The structure of the parent cinchona alkaloids (CD, cinchonidine;
CN, cinchonine; Q, quinine, QD, quinidine).
and calculations [3,4,8,18,20] indicate that it might be in the
antiopen conformation. Our own studies on isocinchona al-
kaloids [which, due to their rigid structure, may only exist in
the open conformation [33,34] (Fig. 3)] support the presence
of the open conformation [31].
The reader may ask why, and with reason, further stud-
ies on isocinchona alkaloids were necessary. The most im-
portant motive is a result we obtained in our latest experi-
ments [35], namely that DHCD is capable of ensuring enan-
tioselectivities over 90% under mild experimental conditions
(room temperature, hydrogen pressure of 1 bar) even at a
concentration as low as 0.004 mmol/L. More over ee is
80% even at 0.001 mmol/L DHCD. Considering that the α-
isocinchonine (α-ICN) and β-isocinchonine (β-ICN) modi-
fiers contained 3% impurities [including dihydrocinchonine
(DHCN)], the question arose whether the enantioselectivi-
ties attained [31,36] could be due to the DHCN present.
It was therefore necessary to purify isocinchona alkaloids
and the experiments had to be repeated with 100% pure
modifiers. By the evidence of ESI-MS measurements car-
ried out to date [31,36], the ion with maximal abundance in
the solution is the one with m/z = 299, an ion attributed
to the protonated form of tetrahydro-α-ICN (TH-α-ICN)
and tetrahydro-β-ICN (TH-β-ICN). In addition, optimiza-
tion measurements in order to determine maximal ee (α-ICN
and β-ICN concentration, temperature, pressure) as well as
further studies to elucidate the effect of experimental condi-
tions on the conversion of α-ICN and β-ICN in the course of
the hydrogenation were also necessary. These experiments
were also justified by our new observations demonstrating
an unexpected change in the sense of enantioselection with
β-ICN as modifier [37].
Fig. 2. The structure of the DHCD–EtPy intermediate complexes (A and
B, DHCD “antiopen;” C, “syn open;” D, “closed” conformation; DHCD,
dihydrocinchonidine; EtPy, ethyl pyruvate).
responsible for chirality [3,4,16,19,20,22,24,30–32]. How-
ever, no consensus has been reached over the structure of the
intermediate (a 1:1 complex of DHCD and pyruvate) respon-
sible for chiral induction.
The intermediate complexes published recently are sum-
marized in Fig. 2. DHCD in A (Augustine’s model [22]),
B (Wells’ and Baiker’s models [2,20]), and C (Blaser’s
model [32]) is in “open,” while in D (Margitfalvi’s model
[30]) is in “closed” conformation. The intermediates A–C
are anchored on the surface of the platinum by a multicenter
π-bond of the quinoline skeleton and the conjugated δπ sys-
tems of pyruvate. In contrast, structure D is already formed
in solution and this complex is hydrogenated.
According to NMR and XRD measurements on the parent
cinchona alkaloids as well as molecular mechanical calcu-
lations [15,17], from the four stable conformations possi-
ble, alkaloids are present in solution mostly in the so-called
“open 3” or “antiopen” conformation (B) (in “anti” confor-
mation, the benzene ring is turned away from the C9–OH
whereas “syn” denotes the conformation in which the ben-
zene ring is turned toward the C9–OH (C)). No direct ex-
perimental evidence is available to prove whether, in the
course of chiral hydrogenation, the cinchona alkaloid en-
gaged in the 1:1 intermediate on the surface of the catalyst is
in the open or in the closed conformation, as the intermedi-
ate has not been isolated. Indirect experimental observations
2. Experimental
2.1. Materials
Cinchonine (CN), AcOH, and solvents were purchased
from Fluka. EtPy (Fluka) was distilled before use to attain
99.5% purity. DHCN was prepared by hydrogenation of CN
(Pd/C, 1 N H2SO4/H2O, 1 bar H2, 298 K) and used after
crystallization. For the preparation of α- and β-ICN, 5 g
(17 mM) of CN (Fluka 27370) was dissolved in 30 ml of
70% H2SO4. The solution was stirred at 343–348 K for 24 h.
The solution was next cooled and alkalized using ammonium

M. Bartók et al. / Journal of Catalysis 220 (2003) 207–214
209
hydroxide and the solid material precipitated was dissolved
in absolute EtOH. Insoluble inorganic salts were removed
by filtration, the filtrate was evaporated and the residue was
again dissolved in EtOH, filtrated, and evaporated. The pro-
cedure was repeated two more times in order to remove
inorganic salts. The complex reaction mixture was purified
by repeated column chromatography (Fluka 60741 silica gel
60; eluent: toluene/EtOAc/diethylamine 7/2/1). The frac-
tion composition was analyzed by thin-layer chromatogra-
phy (Fluka 60778 silica gel/TLC cards; eluent composition,
see above). α-ICN obtained were 0.7 g, m.p. = 398–400 K,
determined. The catalyst was finally kept at 473 K, 20 Hg
mm for 2 h, the escaping product was recovered at the tem-
perature of liquid nitrogen, and its composition was deter-
mined.
2.5. HPLC and HPLC-MS measurements
The HPLC/ESI-MS measurements were run on a Finni-
gan TSQ 7000 tandem mass spectrometer (Finnigan Ltd.,
San Jose, CA). An Applied Biosystems ABI 140C syringe
pump was used for HPLC separations using a 30-min linear
gradient from 5 to 95% solvent B (0.04% TFA in 80% aque-
ous acetonitrile) in solvent A (0.05% TFA in water). The
flow rate in the experiments was 150 µL/min. The injected
volume was 10 µL of sample. A Hypersil column (PEP 300
C18 5 µm, length 150 mm, i.d. 2.1 mm) was used for sepa-
ration.
[α]D = +63◦ (c = 0.5; EtOH), RF = 0.54; β-ICN, 0.4 g,
20
m.p. = 401–402 K, [α]D = −60◦ (c = 0.5; EtOH), RF =
20
0.43. By the evidence of HPLC, the compounds are 100%
pure. The 1H NMR and 13C NMR spectra of the compound
(Bruker DRX 500, CDCl3, 500 and 125.8 MHz) are identi-
cal with the published data [33,34].
2.2. Catalyst
3. Results and discussion
Based on the data in the literature, from several catalysts
the one most often used is Engelhard 4759 (E4759). E4759
was pretreated before use in a fixed-bed reactor by flush-
ing with 30 mL min−1 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 was stored under air before use.
In order to achieve the aim described in the Introduction,
the effect of the experimental parameters (α-ICN, β-ICN,
DHCN concentration, temperature, and pressure) on the rate
of the hydrogenation and on enantioselectivity was studied.
The conversion of the chiral modifiers in the course of the
hydrogenation reaction was followed by ESI-MS, ESI-MS-
MS, and HPLC-MS measurements. ESI-MS was also used
to determine the composition after desorption of cinchona
alkaloids adsorbed on the catalyst.
2.3. Hydrogenation
Hydrogenation was performed in an atmospheric batch
reactor or in a Berghof Bar 45 autoclave. The catalytic sys-
tem including catalyst (25 or 50 mg) and 2 or 4 mL of
solvent was purged three times with hydrogen. The catalyst
was stirred and prehydrogenated for 30 min. The calculated
amount of modifier was introduced and after several min
0.12–0.5 mL of EtPy was injected and stirred in the pres-
ence of hydrogen for the required reaction time (usually 10–
50 min). Standard conditions are 25 mg E4759, 2 mL AcOH,
0.1 mmol/L modifier, 1 bar H2, 293 K, 1200 rpm, 0.12 mL
EtPy. The product identification and the enantiomeric ex-
cess [ee% = ([S] − [R]) × 100/([S] + [R])] were moni-
tored by gas chromatography (HP 5890 GC-FID, 30-m long
Cyclodex-B capillary column, uncertainty 2%).
3.1. Effects of α- and β-ICN concentration
The relevant experimental results are summarized in
Figs. 4–6. Data measured at 293 and 273 K over a wide
range of modifier concentrations are presented in Figs. 4
and 5 and those obtained in the narrow concentration range
of 0.001–0.01 mmol/L at 273 K are summarized in Fig. 6.
The aim of the latter measurements was (just like in the study
on DHCD [35]) to determine the minimal modifier concen-
tration necessary for maximal enantioselectivity under mild
experimental conditions (low temperature, 1 bar hydrogen
pressure): this piece of information may contribute signifi-
cantly to the interpretation of the reaction mechanism which,
although intensively studied, still raises many unanswered
questions. These experimental results and their comparison
to the data obtained with DHCD allow the following impor-
tant conclusions to be drawn.
ESI-MS and ESI-MS/MS measurement methods were
described in an earlier publication [35].
2.4. Desorption studies
Since the configuration of carbon atoms C8 and C9,
responsible for the direction of the enantiodiscrimination,
is opposite to that of the corresponding carbon atoms of
DHCD, α-ICN and β-ICN direct the formation of excess
(S)-EtLt. In agreement with earlier observations [2–4], the
hydrogenation is completed faster on the modified catalyst
than on the unmodified one. Two explanations have been
proposed for this phenomenon in the case of hydrogenation
After enantioselective hydrogenation, the catalyst was
separated from the solution by centrifugation. The ratio of
hydrogenated cinchona alkaloids in the solution was deter-
mined by ESI-MS. The catalyst was washed with various
solutions under constant stirring (1200 rpm) and centrifuged
several times, alkaloids were isolated from the washing so-
lutions by the usual procedures, and their composition was

210
M. Bartók et al. / Journal of Catalysis 220 (2003) 207–214
Fig. 6. Enantioselective hydrogenation of EtPy in the narrow range
(0.001–0.01 mmol/L) of modifier concentration (25 mg E4759, 2 mL
AcOH, 1 bar H , 273 K, 1200 rpm, 0.12 mL EtPy).
2
Fig. 4. The effect of α-ICN (α) and β-ICN (β) concentration (mmol/L)
in enantioselective hydrogenation of EtPy at 293 K (25 mg E4759, 2 mL
AcOH, 1 bar H , 1200 rpm, 0.12 mL EtPy).
2
The rate of hydrogenation is lower in the case of the mod-
ifier α-ICN than in the presence of the DHCD. We have
explained this phenomenon for CN with the vicinity of the
Me group on C10 to the surface which inhibits rotation along
the C8–C9 axis. Due to this effect, the development of the
advantageous conformation of the 1:1 intermediate respon-
sible for enantioselection becomes somewhat inhibited.
There is a remarkable difference between the effects of
the chiral modifiers α-ICN and β-ICN on the rate of EtPy
hydrogenation and on enantioselectivity (Figs. 4–6). Hydro-
genation is faster in the presence of β-ICN and, at a concen-
tration as low as 0.001 mmol/L, high ee is obtained as com-
pared not only to α-ICN but also to DHCD [35]. The large
difference in reaction rate between α- and β-ICN prompted
us to perform the experiments under identical conditions
with DHCN (Fig. 7). The order of reaction rates obtained
is β-ICN > DHCN > α-ICN.
In contrast to DHCN, the reaction rate is increased in
proportion to increasing α-ICN and β-ICN concentrations,
while enantioselectivity already reaches its maximum at
0.1 mmol/L α-ICN (ee 82%) and at 0.01 mmol/L β-ICN
(ee 62%) (Table 1). In the case of isocinchonas, however,
this maximum does not reach the value of that obtained
with DHCN as modifier, which is 90–92% under identical
conditions. In the case of β-ICN it is quite striking that in-
creasing modifier concentration has an opposite effect on
reaction rate and ee: the former increases, whereas the latter
decreases with increasing concentration of the modifier. It is
well known that the hydrogenationrate of ketones may be in-
creased by bases [42,43]. For this reason the basicities of α-
and β-ICN were determined by potentiometric titration: for
α-ICN, pK1 = 8.23 0.02, pK2 = 2.96 0.04; and for β-
ICN, pK1 = 7.88 0.02, pK2 = 2.72 0.04 were obtained.
These data demonstrate that, from the two tertiary bases,
Fig. 5. The effect of α-ICN (α) and β-ICN (β) concentration (mmol/L)
in enantioselective hydrogenation of EtPy at 273 K (25 mg E4759, 2 mL
AcOH, 1 bar H , 1200 rpm, 0.12 mL EtPy).
2
in AcOH. One of these is the so-called ligand-accelerated
mechanism [38], whereas another interpretation emphasizes
the role of electrostatic acceleration [35]. At the same time, it
should be noted that in nonacidic solvents, when the hydro-
genation of EtPy proceeds in the presence of tertiary amines,
rate acceleration is observed [39,40]. The effect of cinchona
alkaloids, however, is much stronger than that of tertiary
amines. Based on theoretical calculations concerning the hy-
drogenation of acetophenones in toluene, a correlation has
been found between the orbital energy of the keto carbonyl
groups and hydrogenation rates [41].

M. Bartók et al. / Journal of Catalysis 220 (2003) 207–214
211
3.2. The effect of temperature
In the enantioselective hydrogenation of the EtPy the role
of temperature was first studied [47] under relatively se-
vere experimental conditions (high hydrogen pressure and
DHCD concentration). The results obtained under the mild
experimental conditions applied by us (hydrogen pressure
1 bar, DHCD concentration 0.001–1 mmol/L, temperature
263–293 K) in AcOH have been published recently [35].
Similar experiments in the case of α- and β-ICN have not
been reported. Our new experimental results obtained in the
presence of α- and β-ICN are summarized in Figs. 5, 6, and
8 and Table 1.
Naturally, it was to be expected that the rate of EtPy hy-
drogenation would decrease as a consequence of reducing
temperature. As temperature is decreased, ee increases to
273 K and reaches 86% (0.1 mmol/L α-ICN, 1 bar hydrogen
pressure, 100% conversion at 30 min in the case of α-ICN).
It seems that the increase in hydrogen concentration due to
temperature reduction enhances chiral hydrogenation more
than it does racemic hydrogenation. In the presence of β-
ICN an unusual effect was observed: reaction rate decreased
in parallel with increasing temperature (Fig. 8b), whereas ee
increased, quite unlike that with α-ICN. According to the ex-
perimental data in Table 1, in the case of β-ICN ee increases
with increasing temperature at all concentrations (0.01, 0.1,
1 mmol/L) with the exception of 0.001 mmol/L, in con-
trast to α-ICN, in the presence of which ee decreases. These
results indicate the conversion of isocinchonas during hy-
drogenation (see below) that brings about a more extensive
structural alteration in β-ICN than in α-ICN; alternatively,
they may participate in a different way/via a different mech-
anism in the development of chiral centers responsible for
enantioselectivity.
The most important observation is, however, that at 263 K
and 0.001 mmol/L α-ICN and β-ICN concentrations (Ta-
ble 2), the formation of (S)-EtLt proceeded 44–53%: this
is important because at 263 K, hydrogenolysis of the C–O
bond may be excluded. Comparison of the experimental re-
sults obtained with DHCD and isocinchonas at a modifier
concentration of 0.001 mmol/L reveals a significant dif-
ference (Table 2). The higher ee may be explained by the
higher stability of DHCD under the conditions of hydro-
genation.
Fig. 7. The effect of dihydrocinchonine (DHCN) concentration (mmol/L)
in enantioselective hydrogenation of EtPy (25 mg E4759, 2 mL AcOH,
1 bar H , 293 K, 1200 rpm, 0.12 mL EtPy).
2
β-ICN may be converted to a stronger base. The stronger
base formed from β-ICN increases the hydrogenation rate
whereas it is less effective in chiral induction, in the devel-
opment of chiral catalytic centers or in the formation of the
intermediate responsible for enantioselectivity.
In the case of DHCD the curve describing the depen-
dence of reaction rate on DHCD concentration has a max-
imum. There are two interpretations for this maximum in
the literature [44,45]. Based on their kinetic study, Blaser
et al. explained the phenomenon by the so-called “three-
site model” [44], whereas LeBlond et al. [45] attributed it
to the change in optimal adsorption geometry of DHCD
at higher modifier concentrations in AcOH. The latter was
verified by attenuated total infrared spectroscopy [27,46].
Given the absence of conclusive experimental evidence, the
reason why no maximum appears in the case of α-ICN
and β-ICN can only be guessed (rigid conformation, de-
velopment of chiral surface sites in a way different from
that in the presence of DHCD, possible gradual hydrogena-
tive ring opening of isocinchonas, etc.). It may be estab-
lished, however, that the modifier concentration necessary
for maximal ee is higher for α-ICN than for DHCN and
β-ICN.
Table 1
The effect of temperature in enantioselective hydrogenation of EtPy (25 mg E4759, 2 mL AcOH, 1 bar H , 1200 rpm, 0.12 mL EtPy)
2
Temperature (K)
Modifier (mmol/L)
273
0.001
293
0.001
273
0.01
293
0.01
273
0.1
293
0.1
273
1
293
1
303
1
333
1
a
ee (%)
38
55
15
44
68
52
64
62
82
20
74
50
79
21
73
52
73
56
10
54
b
ee (%)
a
In the case of α-ICN.
In the case of β-ICN.
b

212
M. Bartók et al. / Journal of Catalysis 220 (2003) 207–214
hydrogen pressure, due to increasing hydrogen concentra-
tion in the solution, and eventually reached saturation [47,
48]. Similar experiments using α- and β-ICN as modifiers
have not been reported. Some characteristic data are pre-
sented in Table 3.
In measurements performed at a hydrogen pressure of
25 bar (at the temperature and α-ICN concentration op-
timized for 1 bar hydrogen pressure) (Table 3) ee values
exceeding 90% were unexpectedly observed. When hydro-
gen pressure is further increased under identical conditions,
ee is reduced, most probably due to hydrogenation of the
quinoline skeleton of α-ICN. In contrast to α-ICN, increas-
ing hydrogen pressure had no significant effect on ee when
a β-ICN modifier was used.
(a)
3.4. Desorption studies
The aim of these studies was—using a relatively simple
method—to identify hydrogenated cinchonas formed on the
catalyst. The target compound was DHCN which might pos-
sibly be formed from α-ICN. The results obtained by the
method described in the experimental section are summa-
rized in Table 4.
It is demonstrated by the data in Table 4 that DHCN
([M+H]+ = 297) was not formed. Desorbed cinchona alka-
loids are α-ICN, TH-α-ICN, and DCH-α-ICN [M + H]+ =
295, 299, and 305, respectively; for abbreviations, see Fig. 9.
Thus, only cinchonas of rigid structure are identified; in
other words, cinchona(s) in which rotation along the C8–C9
bond is free could not be identified in the course of desorp-
tion either.
(b)
Fig. 8. The effect of temperature in enantioselective hydrogenation of EtPy
Table 3
The effect of hydrogen pressure in enantioselective hydrogenation of EtPy
on α-ICN modified catalyst (50 mg E4759, 4 mL AcOH at 297 K, 4 mL
AcOH:toluene (3:1) at 273 K, 1200 rpm, 0.24 mL EtPy)
(25 mg E4759, 2 mL AcOH, 1 bar H , 293 K, 1200 rpm, 0.12 mL EtPy, (a)
2
α-ICN concentration 1 mmol/L, (b) β-ICN concentration 0.01 mmol/L).
Temperature (K)
273 297 273 297 273 297 273 297
25 25 25 25
Modifier (mmol/L) 0.01 0.01 0.01 0.01 0.1 0.1 0.1 0.1
Table 2
H
pressure (bar)
1
1
1
1
2
Effect of temperature for the hydrogenation of EtPy on optical yield (25 mg
E4759; 0.001 mmol/L modifier; solvent: 1 mL AcOH + 1 mL toluene in
ee (%)
72
64
89
78
86
74
94
88
order to keep AcOH dissolved; 1 bar H ; 0.1 mL EtPy for α-ICN and β-
2
ICN; 0.5 mL EtPy for DHCD)
Modifier
Temperature
(K)
Time
(min)
Conversion
(%)
ee
(%)
Ref.
Table 4
Desorption studies after EtPy hydrogenation on α-ICN-modified catalyst
using ESI-MS (100 mg E4759, 1 mg α-ICN, 273 K, 25 bar H , 0.4 mL
EtPy, 4 mL AcOH: toluene (1:1), reaction time 60 min)
DHCD
α-ICN
β-ICN
DHCD
α-ICN
β-ICN
DHCD
α-ICN
β-ICN
293
293
293
273
273
273
263
263
263
150
60
30
150
30
30
150
30
30
40
20
48
35
5
35
32
6
30
15
44
70
38
55
83
53
44
[35]
2
This work
This work
[35]
This work
This work
[35]
Method
m/z values (relative peak intensity %)
295
297
299
305
A
B
C
1
3
13
0
0
0
0
0
10
5
16
0
100
100
100
0
This work
This work
a
27
D
A, in the solution after EtPy hydrogenation; B, in the two pooled washes
of the catalyst with AcOH + T (1:1); C, in the three pooled washes of the
catalyst with concentrated ammonium hydroxide; D, in the distillate of the
catalyst after A, B, C at 20 Hg mm, 473 K.
3.3. Effect of hydrogen pressure
a
The ee value attained in EtPy hydrogenation in the pres-
m/z signals of many polycarbonates of unknown structure between
375 and 991 which differ from each other by m/z = 44.
ence of CD and DHCD usually increased with increasing

M. Bartók et al. / Journal of Catalysis 220 (2003) 207–214
213
3.5. Results of HPLC, HPLC-MS, and ESI-MS
Table 5
Relative abundances of the cinchona alkaloids obtained from hydrogenation
of α-ICN and β-ICN over Pt catalysts in AcOH (50 mg E4759, 293 K, 2 mg
It has been established by HPLC measurements that nei-
ther α-ICN nor β-ICN are isomerized to α- and β-ICNN,
respectively, at room temperature and below, that is, at a tem-
perature of hydrogenation of EtPy. The minimal temperature
at which this isomerization takes place at a measurable rate
is 343 K. At 343 K the conversions of β-ICN and α-ICN
at 30 min are 5 and 3%, respectively (these conversions at
120 min are 36 and 17%). These experiments make it neces-
sary to correct former conclusions drawn from ESI-MS mea-
surements, i.e., that β-ICN is not converted in glacial acetic
acid [37]. The reason why the inconsistency developed is
that ring opening of β-ICN in AcOH was not detected by
ESI-MS, since the molecular ion [M + H]+ (m/z = 295)
was not changed. Isomerization under the given experimen-
tal conditions was not considered at that time.
HPLC-MS studies were done only in a few cases, be-
cause the large number of cinchona alkaloid isomers and
diastereoisomers would have made chromatographic sepa-
ration extremely difficult. It was, however, possible to estab-
lish that the hydrogenation products of α-ICN and β-ICN
obtained on platinum catalysts in the presence of AcOH, for
each m/z value of 295, 299, and 305, two ions are identified
by MS. Because of extremely low concentrations, isolation
of these isomers and verification of their structure by NMR
were not possible. For this reason, the compounds shown
in Fig. 9 were identified with the help of characteristic frag-
ment ions of the ESI-MS-MS spectra of the product mixture.
Under the conditions of EtPy hydrogenation, i.e., at room
temperature in the presence of Pt–alumina and hydrogen,
however, both isocinchona alkaloids undergo significant
conversion, as indicated by the ESI-MS data in Table 5.
In agreement with data of Table 5, hydrogenation of the
quinoline skeleton of α- and β-ICN in the absence of EtPy
is rapidly completed, yielding tetrahydro- and decahydro-
derivatives. It is important to note, however, that the forma-
tion of [DHCN + H]+ (m/z = 297) and [DDHCN + H]+
(m/z = 307) is not observed, indicating that under the given
mild experimental conditions, hydrogenolysis of the cyclic
ether structure fails to occur; i.e., the structure is stable. As
shown above (Table 1), in the case of α- and β-ICN ee val-
modifier, 5 mL AcOH, 1 bar H pressure)
2
+
Modifier
Time
hydrogenation
m/z values of [M + H]
(relative peak intensity %)
(min)
295
297
a
299
305
307
α-ICN
α-ICN
α-ICN
β-ICN
β-ICN
β-ICN
30
60
120
30
60
120
3
0
0
37
0
0
63
30
a
100
100
100
6
18
100
0
0
0
0
0
0
0
0
a
100
100
18
0
0
a
“Noise level” formation.
ues of over 80 and 60%, respectively, were attained under
these conditions.
Since there is no C8–C9 rotation in α- and β-ICN, these
cinchonas behave as conformationally rigid chiral ligands on
the surface. β-ICN with “oxazatwistane” structures are more
rigid than the α-ICN in which a certain extent of confor-
mational movement is possible due to the “homotwistane”
structure. The ee expected with β-ICN is therefore lower
than with the α-ICN: the more rigid structure is less capa-
ble of sterically fitting into the 1:1 intermediate responsible
for the induction of chirality.
4. Conclusion
It is assumed highly probable that in the course of EtPy
hydrogenation on Pt catalysts modified with high-purity α-
ICN and β-ICN, these chiral compounds of rigid struc-
ture which have antiopen conformation are responsible for
enantioselectivity. Although a portion of α-ICN and β-ICN
undergoes conversions during hydrogenation, formation of
DHCN could not be verified. Consequently, it may be de-
clare that the “closed” conformation of cinchona alkaloid
is not a precondition for chiral induction in enantioselective
hydrogenation of EtPy on Pt–alumina–cinchona system.
In order to achieve ee values over 90%, higher cinchona
alkaloid concentration and hydrogen pressure are needed
with α-ICN than with DHCD [35], pointing to the role so
Fig. 9. Isomers and hydrogenated isocinchonas formed from α-ICN and β-ICN.

214
M. Bartók et al. / Journal of Catalysis 220 (2003) 207–214
far of unidentified factor(s). The latter statement is also sup-
ported by studies on β-ICN. When using β-ICN, maximal
enantioselectivity was only 70–72%, despite the fact that hy-
drogenation rate exceeded not only that with α-ICN but also
that with DHCN.
Since, according to studies published so far, enantiose-
lection happens on the surface of the Pt catalyst (also see
in the most recent review [4]), the results presented in this
manuscript can offer no answer to the questions raised by
the chemistry on the catalyst surface. On the basis of the
recent experimental data by in situ STM on Cu(111), an
ordered adsorption of CD in “closed 1” conformation was
observed [28]. These data, however, do not make it possi-
ble to interpret the enantioselection, since the measurements
were not performed on Pt(111).
It is to be hoped that further development of IR or other
techniques will soon make them applicable to studies on
the adsorption of cinchona alkaloids with rigid skeletons on
Pt surfaces. Investigations using isocinchona alkaloids and
their isomers may also give new information to help under-
stand this complicated process, bringing evidence for the for-
mation of chirally active surface sites of the metal–organic
type [49] by irreversible adsorption of cinchonas [29] and
the participation of these sites in enantioselection.
[13] K. Balázsik, K. Szöri, K. Felföldi, B. Török, M. Bartók, J. Chem. Soc.,
Chem. Commun. (2000) 555.
[14] M. Studer, S. Burkhardt, A.F. Indolese, H.-U. Blaser, J. Chem. Soc.,
Chem. Commun. (2000) 1327.
[15] G.D.H. Dijkstra, R.M. Kellogg, H. Wynberg, J.S. Svendsen, I. Marko,
K.B. Sharpless, J. Am. Chem. Soc. 111 (1989) 8069.
[16] A. Pfaltz, T. Heinz, Topics Catal. 4 (1997) 229.
[17] T. Bürgi, A. Baiker, J. Am. Chem. Soc. 120 (1998) 12920.
[18] H.-U. Blaser, H.-P. Jalett, W. Lottenbach, M. Studer, J. Am. Chem.
Soc. 122 (2000) 12675.
[19] I.M. Sutherland, A. Ibbotson, R.B. Moyes, P.B. Wells, J. Catal. 125
(1990) 77.
[20] 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.
[21] A.F. Carley, M.K. Rajumon, M.W. Roberts, P.B. Wells, J. Chem. Soc.,
Faraday Trans. 91 (1995) 2167.
[22] R.L. Augustine, S.K. Tanielyan, J. Mol. Catal. A 112 (1996) 93.
[23] T. Evans, A.P. Woodhead, A. Gutiérrez-Sosa, G. Thornton, T.J. Hall,
A.A. Davis, N.A. Young, P.B. Wells, R.J. Oldman, O. Plashkevych, O.
Vahtras, H. Ågren, V. Carravetta, Surf. Sci. 436 (1999) L691.
[24] T. Bürgi, A. Baiker, J. Catal. 194 (2000) 445.
[25] J.M. Bonello, R.M. Lambert, N. Künzle, A. Baiker, J. Am. Chem.
Soc. 122 (2000) 9864.
[26] D. Ferri, T. Bürgi, A. Baiker, J. Chem. Soc., Chem. Commun. (2001)
1172.
[27] J. Kubota, F. Zaera, J. Am. Chem. Soc. 123 (2001) 11115.
[28] Q.-M. Xu, D. Wang, L.-J. Wan, C.-L. Bai, Y. Wang, J. Am. Chem.
Soc. 124 (2002) 14300.
[29] I. Bakos, S. Szabó, M. Bartók, E. Kálmán, J. Electroanal. Chem. 532
(2002) 113.
[30] J.L. Margitfalvi, M. Hegedüs, E. Tfirst, Tetrahedron: Asymmetry 7
(1996) 571.
Acknowledgments
[31] M. Bartók, K. Felföldi, B. Török, T. Bartók, J. Chem. Soc., Chem.
Commun. (1998) 2605.
[32] H.-U. Blaser, H.-P. Jalett, M. Garland, M. Studer, H. Thies, A. Wirth-
Tijani, J. Catal. 173 (1998) 282.
[33] J. Thiel, P. Fiedorow, J. Mol. Struct. 405 (1997) 219.
[34] W. Braje, J. Frackenpohl, P. Langer, H.M.R. Hoffmann, Tetrahe-
dron 54 (1998) 3495.
[35] M. Bartók, K. Balázsik, Gy. Szöllösi, T. Bartók, J. Catal. 205 (2002)
168.
Financial support by the Hungarian National Science
Foundation (OTKA Grant T031707 and TS 044690) is
highly appreciated. The authors are grateful to Professor T.
Kiss, Professor B. Penke, Z. Kele (University of Szeged)
and Dr. Gy. Szendrei (Richter Gedeon RT, Budapest) for
the possibilities of measurements of potentiometric titration,
HPLC, ESI-MS, and HPLC-MS. We also thank Professor F.
Notheisz (University of Szeged) for valuable discussions.
[36] Gy. Szöllösi, K. Felföldi, T. Bartók, M. Bartók, React. Kinet. Catal.
Lett. 71 (2000) 99.
[37] M. Bartók, M. Sutyinszki, K. Felföldi, Gy. Szöllösi, J. Chem. Soc.,
Chem. Commun. (2002) 1130.
[38] M. Garland, H.-U. Blaser, J. Am. Chem. Soc. 112 (1990) 7048.
[39] H.-U. Blaser, H.-P. Jalett, D.M. Monti, J.F. Reber, J.T. Wehrly, Stud.
Surf. Sci. Catal. 41 (1988) 153.
[40] G. Bond, P.A. Meheux, A. Ibbotson, P.B. Wells, Catal. Today 10
(1991) 371.
[41] A. Vargas, T. Bürgi, M. von Arx, R. Hess, A. Baiker, J. Catal. 209
(2002) 489.
References
[1] G. Jannes, V. Dubois (Eds.), Chiral Reactions in Heterogeneous Cata-
lysis, Plenum, New York, 1995.
[2] A. Baiker, H.-U. Blaser, in: G. Ertl, H. Knözinger, J. Weitkamp (Eds.),
Handbook of Heterogeneous Catalysis, Vol. 5, VCH, Weinheim, 1997,
p. 2422.
[3] D.E. De Vos, I.F.J. Vankelecom, P.A. Jacobs (Eds.), Chiral Catalyst
Immobilization and Recycling, Wiley–VCH, Weinheim, 2000.
[4] M. Studer, H.-U. Blaser, C. Exner, Adv. Synth. Catal. 345 (2003) 45.
[5] H. Wynberg, Top. Stereochem. 16 (1986) 87.
[42] M. Freifelder, Practical Catalytic Hydrogenation. Techniques and Ap-
plications, Wiley–Interscience, New York, 1971.
[43] P.N. Rylander, Catalytic Hydrogenation in Organic Syntheses, Acad-
emic Press, New York, 1979.
[6] K. Kacprzak, J. Gawronski, Synthesis (2001) 961.
[7] Y. Orito, S. Imai, S. Niwa, J. Chem. Soc. Jpn. (1979) 1118, 670 (1980).
[8] X. Zuo, H. Liu, D. Guo, X. Yang, Tetrahedron 55 (1999) 7787.
[9] K. Szöri, M. Sutyinszki, K. Felföldi, M. Bartók, Appl. Catal. A 237
(2002) 275.
[10] M. Schürch, N. Kunzle, T. Mallat, A. Baiker, J. Catal. 176 (1998) 569.
[11] B. Török, K. Felföldi, K. Balázsik, M. Bartók, J. Chem. Soc., Chem.
Commun. (1999) 1725.
[44] H.-U. Blaser, M. Garland, H.-P. Jalett, J. Catal. 144 (1993) 569.
[45] C. LeBlond, J. Wang, J. Liu, A.T. Andrews, Y.-K. Sun, J. Am. Chem.
Soc. 121 (1999) 4920; Top. Catal. 13 (2000) 169.
[46] D. Ferri, T. Bürgi, J. Am. Chem. Soc. 123 (2001) 12074.
[47] H.-U. Blaser, H.-P. Jalett, M. Müller, M. Studer, Catal. Today 37
(1997) 441.
[48] Y. Sun, R.N. Landau, J. Wang, C. LeBlond, D.G. Blackmond, J. Am.
Chem. Soc. 118 (1996) 1348.
[49] M. Bartók, K. Balázsik, F. Notheisz, React. Kinet. Catal. Lett. 77
(2002) 363.
[12] M. Studer, S. Burkhardt, H.-U. Blaser, J. Chem. Soc., Chem. Com-
mun. (1999) 1727.