Enantioselective hydrogenation of isophorone over Pd catalysts in the presence of(-)-dihydroapovincaminic acid ethyl ester the effect of reduction method of Pd blacks on the enantiomeric excess

Article abstract of DOI:10.1016/S1381-1169(00)00440-4

The enantioselective heterogeneous catalytic hydrogenation of isophorone in the presence of apovincaminic acid ethyl ester was investigated. Various Pd black catalysts differing in their preparation method were studied. The catalysts were characterized by

Full text of DOI:10.1016/S1381-1169(00)00440-4

Journal of Molecular Catalysis A: Chemical 170 (2001) 101–107
Enantioselective hydrogenation of isophorone over Pd catalysts
in the presence of(−)-dihydroapovincaminic acid ethyl ester
The effect of reduction method of Pd blacks on the
enantiomeric excess
c
c
c,∗
a
b
G. Farkas , É. S ´ı pos , A. Tungler , A. Sárkány , J.L. Figueiredo
a
Institute of Isotopes, Chemical Research Center of the Hungarian Academy of Sciences, P.O. Box 77, H-1525 Budapest, Hungary
b
Laboratório de Catálise e Materiais, Chemical Engineering Department,
Faculty of Engineering, University of Porto, 4050-123 Porto, Portugal
Department of Chemical Technology, Technical University of Budapest, H-1521 Budapest, Hungary
c
Received 8 June 2000; accepted 22 September 2000
Abstract
The enantioselective heterogeneous catalytic hydrogenation of isophorone in the presence of apovincaminic acid ethyl
ester was investigated. Various Pd black catalysts differing in their preparation method were studied. The catalysts were
characterized by different methods such as physical adsorption and chemisorption, SEM and XPS. An explanation was
created for the different enantioselectivity of the catalysts being precipitated by different reducing agents. © 2001 Elsevier
Science B.V. All rights reserved.
Keywords: Pd black catalysts; Enantioselective hydrogenation; Isophorone
1
. Introduction
parts of the molecule are crucial for the selective re-
action [2]. The other aspect of the investigations was
how does the structure of the catalyst influence the
enantioselectivity of the reaction. The enantioselec-
tivity of supported Pd catalysts (Pd/C and Pd/TiO2)
depended on the support used, it turned out that lower
surface area of the support and smaller dispersion of
the active metal are advantageous [4].
Enantioselectivity is very sensitive to the method of
catalyst preparation, this is valid for the most effec-
tive modified catalytic systems: Pt/chinconidine and
Raney-Ni/tartaric acid [5,6].
In earlier publications we reported about the enan-
tioselective hydrogenation of the C=C double bond of
isophorone [1–4]. The best results were achieved in
the system that consists of (−)-dihydroapovincaminic
acid ethyl ester as chiral modifier and Pd black
catalyst reduced with sodium-formate [1]. This re-
action was studied in detail for understanding the
catalyst-modifier-substrate-interactions occurring dur-
ing the reaction. The effect of the structure of the
modifier on the enantioselectivity was studied, which
Previously we reported that the preparation method
of Pd black catalyst affects the enantioselectivity in
the hydrogenation of isophorone [4], but the corre-
lation between enantioselectivity and the structure or
∗
Corresponding author. Tel.: +36-1-463-1203;
fax: +36-1-463-1913.
E-mail address: tungler.ktt@chem.bme.hu (A. Tungler).
1381-1169/01/$ – see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S1381-1169(00)00440-4

102
G. Farkas et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 101–107
the state of the catalyst surface itself has not been re-
ported yet. The objective of this study was to explore
with the characterization of the catalysts the relation
between the surface properties of the Pd black cata-
lysts and their enantioselectivity, how to prepare the
best Pd black catalyst by different reduction methods,
and to determine the effect of drying on air.
(Pd–H)S was titrated with O2 injections via a cali-
brated loop (0.1 ml each). Next (Pd–O)S was titrated
with H2. After decomposition of ␤-PdH O2 was ad-
sorbed again. Finally, Ar stream was replaced with
H2 and CO adsorption was recorded. The number of
surface sites was calculated considering the accepted
stoichiometry of these surface reactions.
The stoichiometry calculations were based on the
following:
for titration with O2
2
. Experimental
(
(
(
Pd–H)S + 0.75O2 = (Pd–O) + H2O
for titration with H2
2
.1. Materials
Pd black catalysts were prepared according to the
Pd–O)S + 1.5H2 = (Pd–H) + H2O
for CO adsorption
following procedure: 18 mmol (6.0 g) K2PdCl4 was
dissolved in 100 ml water and reduced at boiling point
with different reducing agents dissolved in 20 ml wa-
ter. In the case of catalyst 1, the pH of the solution dur-
ing the preparation was basic, and the whole amount
of the reducing agent (HCOONa) was added at the be-
ginning of the reaction. For type 2 catalyst, an equimo-
lar amount of HCOONa was used, and at the end of
the reduction the pH was acidic. For type 3 catalyst,
the solution was basic like type 1, but the HCOONa
was added drop by drop. Type 4 catalyst was reduced
with NaBH4, type 5 with hydrazine. The other method
comprised, after similar preparation steps, the reduc-
tion of Pd(OH)2 precursor with H2 in water on 5 bar
in a stainless autoclave with magnetic stirrer (type
Pd–H)S + CO = (Pd–CO) + 0.5H2
XPS tests were conducted in order to determine the
surface chemical composition of Pd black catalysts.
The tests were performed with a VG Scien-
tific ESCALAB 200 A spectrometer utilizing a
non-monochromatized Mg K␣ radiation (1253.6 eV).
The vacuum in the analysis chamber was always
<
−7
1 × 10 Pa. The catalysts were observed by SEM
(JEOL JSM 6301 F) to determine the particle size
and the morphology of the catalysts.
The BET surface areas were determined by N2 ad-
sorption at 77 K with an equipment Coulter Omnisorp
6
). All six catalysts were filtered and washed several
1
00 CX by continuous flow techniques.
times with distilled water until the water became neu-
tral, and divided into two parts: A type catalysts were
dried in air, B type catalysts were stored under water,
and added wet to the reaction mixture. Apovincaminic
acid ethyl ester was supplied by Richter Gedeon Co.
2
.3. Hydrogenation
The hydrogenation of isophorone was carried out at
◦
2
5 C and under 50 bar hydrogen pressure in a Büchi
Bep 280 autoclave equipped with a magnetically
driven turbine stirrer and a gas-flow controlling and
measuring unit. Before hydrogenation, the reaction
mixtures were stirred under nitrogen for 15 min in the
reaction vessel. The water content of methanol was
optimized because in the absence of water isophorone
diacetal formation was observed, the reaction rate
decreased and the ee moderated. The optimal water
content of the methanol for this reaction was found
to be between 3 and 10 vol.%.
2
.2. Catalysts characterizations
Adsorption measurements were made in an atmo-
spheric flow system [10,11] in order to determine the
number of surface sites on the Pd black samples. Four
types of measurements were made: O2 titration, H2
titration, O2 titration again and direct CO adsorption
in H2 carrier gas. Prior to the first adsorption of O2,
the sample was treated in 1.2% H2/Ar for 15 min and
then in Ar gas to remove ␤-PdH, in order to avoid
hydrogen absorption into the bulk phase of the metal.
The reaction mixtures were analyzed with a gas
chromatograph equipped with a ␤-cyclodextrin capil-

G. Farkas et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 101–107
103
Table 1
Characterization of the catalyst by XPS, SEM and N2 adsorption
2
Pd black catalysts
Reducing agent
Atomic ratio (%)
Particle size (nm)
SBET (m /g)
Pd
PdO
PdCl2
1
2
3
4
5
6
HCOONa in excess
HCOONa stochiometric
HCOONa drop by drop
NaBH4
NH2NH2
H2
50.0
50.0
50.1
51.0
54.8
78.1
34.5
30.0
32.9
34.9
33.5
14.5
15.5
20.0
16.2
14.0
11.6
7.3
50–100
100–150
≤50
5.7
6.0
14.4
6.0
7.0
1.7
≤50
70–100
230–450
lary column (analysis temperature: dihydroisophorone
at 110 C) and FID. The chromatograms were recorded
tive reaction. The results of adsorption measurements
are summarized in Table 2.
◦
and peak areas were calculated with Chromatography
Station for Windows V1.6 (DataApex Ltd., Prague).
The enantiomeric excess was defined as
The dispersion data were calculated from the CO
gas adsorption measurements. As expected, the values
were rather small and uniform. Only the dispersion
of catalyst 3 was twice that of the others, and the
dispersion of catalyst 6 was much smaller.
The amounts of different gases adsorbed on the cat-
alysts were similar, and there was no important dif-
ference in the number of active sites on the cata-
lysts surfaces (see Table 2) determined by these small
molecules.
ꢀ
ꢁ
[
R] − [S]
R] + [S]
ee (%) =
× 100
[
3
. Results and discussion
3
.1. Characterization of Pd blacks
The Sads surfaces were calculated from the titration
data and are in good agreement with the SBET mea-
surements.
The reduction method of precious metal precursors
The number of active sites of a used catalyst was
measured in order to compare the active surface area,
and there was no significant difference from the orig-
inal catalyst.
influence the surface area, the particle size and the
chemical composition of the catalysts. The results of
XPS, SEM and N2 adsorption measurements are sum-
marized in Table 1.
As the XPS measurements indicate, catalyst 6 (re-
duced with H2) has the highest ratio of metallic Pd
phase on its surface (Table 1). It means that most of the
active sites were formed during the catalyst prepara-
tion, not during in situ reduction, in the enantioselec-
3.2. SEM measurements
Figs. 2–5 show the SEM pictures of some of the Pd
black catalysts. The structure of the different catalysts
Table 2
Number of the active sites on the catalysts surface and the calculated surface area
2
Sample
(Pd–H)S O2 (T)
(Pd–O)S H2 (T)
mol (Pd-X)/g
(Pd–H)S O2 (T)
Pd black
(Pd–CO)S CO (A)
n.m.
D
Sads (m /g)
1.36 × 10−
4
1.37 × 10
−4
−4
−4
−4
−4
−5
1.30 × 10
−4
−4
−4
−4
−4
−5
0.014
0.009
0.022
0.013
0.014
0.0009
6.7
6.0
11.3
6.0
7.0
0.6
1
2
3
4
5
6
−
−
−
−
6
4
4
4
4
−5
1.07 × 10
2.24 × 10
1.08 × 10
1.41 × 10
1.23 × 10
1.11 × 10
8.57 × 10
−4
−4
−4
1.97 × 10
2.24 × 10
2.15 × 10
1.23 × 10
1.19 × 10
1.23 × 10
1.53 × 10
1.41 × 10
1.32 × 10
8.3 × 10−
1.56 × 10
1.50 × 10
8.5 × 10
−6
1.11 × 10−
4
1.20 × 10
−4
1.05 × 10
−4
1.01 × 10
−4
0.011
–
After reaction

104
G. Farkas et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 101–107
is rather uniform. The elementary palladium par-
ticles are stuck together to a big grain forming a
structure like a sponge. The elementary particle
sizes are different, as can be seen in the figures,
in the case of catalyst 1 the particle size diameter
is approx. 50 nm, in the case of catalyst 5 approx.
Fig. 1. Hydrogenation of isophorone.
270 nm.
Fig. 2. The SEM picture of catalyst 3.
Fig. 3. The SEM picture of catalyst 4.

G. Farkas et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 101–107
105
Fig. 4. The SEM picture of catalyst 6.
3
.3. Catalytic tests
The enantioselectivities of the catalyst prepared
with different methods are summarized in Table 3.
The reaction conditions were optimized earlier [1].
Catalyst 1 gives the best enantioselectivity, 50% of
R-(−)-dihydroisophorone. Therefore, it seems that the
The Pd black catalysts prepared with different re-
duction methods were tested in the enantioselective
hydrogenation of isophorone (Fig. 1).
Fig. 5. The SEM picture of catalyst 1.

106
G. Farkas et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 101–107
Table 3
The enantiomeric excess on different Pd black catalysts
a
Pd black catalyst (reducing agent)
Type
Reaction time (h)
Conversion (%)
Enantiomeric excess (%)
1
2
3
4
5
6
(HCOONa) low SBET
(HCOONa) acidic
(HCOONa) high SBET
(NaBH4)
A
B
A
B
A
B
A
B
A
B
A
6
7
7.5
8
8
6
5
6
6.5
7
100
100
100
100
81
–
100
75
50
23
25
20
24
–
30
23
40
28
24
(NH2NH2)
100
33
80
(H2)
a
7
◦
Conditions: 0.05 mol isophorone, 0.02 g (−)-DHVIN, 0.2 g AcOH, 50 ml methanol, 50 bar, 25 C.
best reducing agent is the HCOONa. With respect to
the other reduction methods, the catalyst made with
hydrazine gave also a proper ee, but the catalyst pre-
pared with hydrogen in the liquid phase showed re-
duced activity and enantioselectivity.
alyst in two ways: dried in air atmosphere, and stored
under water to exclude air. The results indicate that
dried catalysts give higher enantioselectivity and have
better activity. This observation is significant for all
types of catalyst. In this case, when the catalyst is
dried in air, the surface is oxidized and the catalyst
become active during the reaction by the presence of
hydrogen. As the air was excluded from the catalysts
type B, there was no oxidation and reduction step in
the presence of the reactants. In our earlier paper we
reported that the in situ reduction of the catalysts in
the presence of the substrate and the modifier is ad-
vantageous for the ee.
However, the reducing agent is not the only factor,
since the catalysts types 2 and 3 were prepared also
with Na-formate and the optical purity in this case was
only half of the former catalyst. Types 1 and 3 cat-
alysts were prepared with the same stoichiometry of
the reactants, but according to the preparation method
catalyst type 1 has much lower surface area and disper-
sion, so the enantioselectivity depends on the surface
area, and in this reaction surface area and dispersion
do not influence much the catalyst activity. In the case
of the catalysts 1 and 3, the catalyst with low surface
gave better enantioselectivity. It seems that a surface
4. Conclusion
The methods used for the preparation of Pd black
2
area around 6–7 m /g is advantageous for enantios-
catalysts have strong influence on the enantiomeric
excess. Among the catalysts prepared with the same
reducing agent, the catalyst with lower surface area
was more effective. The surface chemistry affected the
optical yield also. The results show unequivocally that
the in situ reduction of the catalysts by the presence
of the substrate and the modifier is crucial for the
enantiodifferentation process.
electivity. Much higher and much lower surface ar-
eas do not influence the enantioselectivity in the right
way. Our earlier investigations with the Pd/C catalyst
in isophorone hydrogenation gave the same results,
namely the lower dispersion is advantageous for the
ee, or the dispersion has an optimal extent [4].
In the often studied enantioselective hydrogenation
of ␣-ketoesters using cinchona modified platinum cat-
alyst, Wells and co-workers found that an elaborate
aerobic modification procedure of the catalyst with the
modifier molecule in ethanol led to much better enan-
tioselectivity than the anaerobic one [7]. Additional
explanations were offered for the role of oxygen [8,9].
In our experiments we used the prepared Pd black cat-
Acknowledgements
The authors gratefully acknowledge the financial
support Hungarian OTKA Foundation under the con-
tract number T 029557 and T 025041, as well as

G. Farkas et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 101–107
107
that of the Ministry of Education, FKFP 0017/1999.
They are grateful to Gedeon Richter Co. for sup-
plying apovincaminic acid ethyl ester. Support for
mobility was provided by the joint Programme ICCTI
[3] A. Tungler, Y. Nitta, K. Fodor, G. Farkas, T. Máthé, J. Mol.
Catal. 149 (1999) 135.
[
4] G. Farkas, L. Hegedüs, A. Tungler, T. Máthé, J.L. Figueiredo,
M. Freitas, J. Mol. Catal. 153 (2000) 215.
[
5] Y. Nitta, F. Sekine, T. Imanaka, S. Teranishi, J. Catal. 74
(Portugal)/OMFB (Hungary). The assistance of Dr.
(
1982) 382.
[6] H.U. Blaser, H.P. Jalett, M. Müller, M. Struder, Catal. Today
7 (1997) 441.
M.M.A. Freitas on catalyst characterizations is also
acknowledged.
3
[
[
7] G. Webb, P.B. Wells, Catal. Today 12 (1992) 319.
8] P. Johnston, P.B. Wells, J. Catal. 156 (1995) 180.
References
[9] B. Minder, T. Mallat, P. Scrabble, A. Baiker, Catal. Lett. 29
1994) 115.
(
[
[
1] T. Tarnai, A. Tungler, T. Máthé, J. Petró, R.A. Sheldon, G.
Tóth, J. Mol. Catal. 102 (1995) 41.
2] G. Farkas, K. Fodor, A. Tungler, T. Máthé, G. Tóth, R.A.
Sheldon, J. Mol. Catal. 138 (1999) 123.
[10] A. Sarkany, Z. Zsoldos, F. Furlong, J.W. Hightower, L. Guczi,
J. Catal. 141 (1993) 566.
[11] A. Sarkany, Gy. Stefler, J.W. Hightower, Appl. Catal. 127
(1995) 77.