Stereodifferentiation in heterogeneous catalytic hydrogenation. Kinetic resolution and asymmetric hydrogenation in the presence of (S)-proline: Catalyst-dependent processes

Article abstract of DOI:10.1016/j.jcat.2009.10.018

The kinetic resolution of 3,5,5-trimethyl cyclohexanone (TMCH) and asymmetric hydrogenation of isophorone (3,5,5-trimethyl cyclohex-2-enone, IP) were investigated on different Pd catalysts in the presence of (S)-proline (Pr). It could be proven that in isophorone hydrogenation the optically active TMCH was formed not only by kinetic resolution but also through asymmetric C{double bond, long}C hydrogenation. The activity and stereoselectivity of different Pd catalysts depended on the support material, preparation method, and reaction conditions as well, confirming our assumption that enantiodifferentiation takes also place on the catalyst surface and not only in the homogeneous liquid phase condensation reaction.

Full text of DOI:10.1016/j.jcat.2009.10.018

Journal of Catalysis 270 (2010) 2–8
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Stereodifferentiation in heterogeneous catalytic hydrogenation. Kinetic
resolution and asymmetric hydrogenation in the presence of (S)-proline:
Catalyst-dependent processes
Nóra Gy o} rffy , Antal Tungler , Mátyás Fodor b
a,*
a
a
Institute of Isotopes, Hungarian Academy of Sciences, P.O. Box 77, Budapest H-1525, Hungary
Budapest University of Technology and Economics, Department of Chemical and Environmental Process Engineering, Hungary
b
a r t i c l e i n f o
a b s t r a c t
Article history:
The kinetic resolution of 3,5,5-trimethyl cyclohexanone (TMCH) and asymmetric hydrogenation of isoph-
orone (3,5,5-trimethyl cyclohex-2-enone, IP) were investigated on different Pd catalysts in the presence
of (S)-proline (Pr). It could be proven that in isophorone hydrogenation the optically active TMCH was
formed not only by kinetic resolution but also through asymmetric C@C hydrogenation. The activity
and stereoselectivity of different Pd catalysts depended on the support material, preparation method,
and reaction conditions as well, confirming our assumption that enantiodifferentiation takes also place
on the catalyst surface and not only in the homogeneous liquid phase condensation reaction.
Ó 2010 Published by Elsevier Inc.
Received 28 August 2009
Revised 5 October 2009
Accepted 22 October 2009
Available online 22 January 2010
Keywords:
Kinetic resolution
3
,5,5-Trimethyl cyclohexanone
Asymmetric C@C hydrogenation
Isophorone
(S)-Proline
Palladium catalysts
1
. Introduction
different ketones, using O18-labeled water. Based on these results
they suggest a covalent, enamine intermediate (8).
Heterogeneous catalytic asymmetric hydrogenation of isopho-
rone (3,5,5-trimethyl cyclohex-2-enone, IP, 2) in the presence of
In IP hydrogenation [4], the optical purity of TMCH depended on
the catalytic metal. Pd and Rh gave higher ee, but only alkylated
proline was formed with Pt. The product ee changed also with
the solvent, methanol proved to be the best one. The effect of IP/
Pr molar ratio on ee was tested too, 1:1 ratio gave the best result.
Török and co-workers [18] initiated the revival of the Pd-med-
iated asymmetric hydrogenation of IP with Pr, they regarded Pr as
a catalyst modifier, not as a chiral auxiliary. Two years later Török
and co-workers [19] proposed another mechanism for IP/Pr hydro-
genation on different Pd catalysts. The participation of kinetic res-
olution in the formation of optically active TMCH was stressed.
They emphasized the role of Pr-modified catalyst surface in IP
hydrogenation, contradicting Lambert and co-workers [20], who
claimed that enantiodifferentiation takes place only in solution.
Lambert and co-workers tested the IP hydrogenation exten-
sively [20,21]. They claimed on the basis of spectroscopic and ki-
netic measurements that optically active TMCH was formed
solely in the kinetic resolution of this compound, namely with
reductive alkylation of Pr and in complete contrast to the case of
ketoester asymmetric hydrogenation, the metal surface was not in-
volved in the crucial enantiodifferentiation step.
(
[
S)-proline (Pr, 1) was first reported more than 20 years ago
1–3]. IP was the only substrate among several ,b-unsaturated ke-
a
tones which afforded significant ee (60% at 55% chemical yield of
trimethyl cyclohexanone, TMCH, 3) in this reaction [4]. Continuing
this work, asymmetric hydrogenation of acetophenone [5,6], dia-
stereoselective reductive alkylation of Pr with ethyl pyruvate
[
[
7,8], asymmetric hydrogenation of benzylidene benzosuberone
9], synthesis of chiral modifiers based on (S)-proline, and their
use in the enantioselective hydrogenation of IP [10–14] were also
investigated.
Detailed circular dichroism, NMR and IR spectroscopy mea-
surements, and preparative experiments resulted in the conclu-
sion that asymmetric hydrogenation of IP can proceed through
carbonylamine (4) and oxazolidinone-type intermediate (5),
which was verified on the basis of the work of Joucla and Mortier
[
15]. Recently, Seebach et al. [16] studied these compounds and
their reactivity in detail, stressing their role in proline catalysis.
List et al. [17] tried to find out which intermediate compounds
play significant role in the proline-catalyzed aldol reaction of
2 3 2 3
Li et al. [22] tested Pd supported on Al O –K CO and concluded
that the main source of optically active TMCH was kinetic
resolution.
*
Corresponding author. Fax: +36 1 392 2533.
E-mail address: gyorffy@iki.kfki.hu (N. Gy o} rffy).
0
021-9517/$ - see front matter Ó 2010 Published by Elsevier Inc.
doi:10.1016/j.jcat.2009.10.018

N. Gy o} rffy et al. / Journal of Catalysis 270 (2010) 2–8
3
It was pointed out in our recent paper [23] that (i) on Pd/C even
at small conversions (0.12–0.2 mol hydrogen consumption) opti-
cally active TMCH was formed in significant amounts, due, likely
to the asymmetric hydrogenation of IP, rather than to kinetic res-
olution of the saturated ketone only; (ii) at high conversions a mix-
ture of the two alkylated proline products could be obtained which
shows that Pr reacts also with (S)-TMCH.
COO) in 2.5 molar excess with respect to palladate was added to
the boiling mixture. After 30 min, the suspension was cooled, and
the catalyst was filtered, washed with distilled water, and dried.
Catalysts, 5 wt%, Pd/BaCO
3
and Pd/MgO, were prepared as fol-
PdCl in
lows: the calculated amount of the catalyst precursor (H
2
4
3
3
5 cm water) was added to the 15 cm aqueous suspension of 2 g
support (nonporous powdered materials), after mixing for half an
3
An apparent analogy of our Pr chiral auxiliary-assisted asym-
metric hydrogenation could be the palladium-induced domino
reaction of benzyl b-ketoesters, where after the Pd-mediated
hydrogenolysis of the protecting benzyl group, asymmetric decar-
boxylation occurs catalyzed by a chiral amine, the result being
optically active ketones [24–27]. In the second step two reaction
routes are possible, the homogeneous route catalyzed by chiral
hour. Na(HCOO) in 10 cm aqueous solution in 2.5 molar excess
with respect to palladate was added to the boiling mixture. The
suspension was cooled after 30 min, and the catalyst was filtered,
washed with distilled water, and dried.
Methanol, n-hexanol, (S)-proline, and isophorone were supplied
by Sigma–Aldrich. The latter was distilled in vacuum before use.
TMCH was prepared in our laboratory by hydrogenating isopho-
rone without solvent, using Pd/C catalyst at ambient temperature
and 10 bar hydrogen pressure. TMCH content was >99%, deter-
mined with GC.
0
amino alcohol and the heterogeneous route, Pd metal surface plus
the adsorbed amino alcohol catalyzed asymmetric decarboxyl-
ation. Baiker and co-workers pointed out that the first, homoge-
neous reaction mechanism dominates, high enantioselectivity
needs at least a twofold chiral auxiliary:substrate molar ratio,
which can be ensured by slow deprotonation reaction, with small
intermediate ketoacid concentration [27]. The analogy is really vir-
tual in the IP/Pr and TMCH/Pr systems in contrast with the afore-
mentioned, the role of the Pd surface, the importance of the
heterogeneous reaction in stereocontrol should be proven.
Another debated issue is the role of Pr, whether it is a chiral
auxiliary [4,20,21] or a chiral modifier [18,19]. The auxiliary added
to the reaction mixture in commensurable amount to the sub-
strate, can react with it, the asymmetric induction takes place
within a usually covalent adduct of the auxiliary and the substrate,
which contains chiral and prochiral part alike. Finally the heteroge-
neous catalyst Pd distinguishes between the enantiomers of this
adduct, reacting faster with one isomer. Contrarily the chiral mod-
2.2. Catalysts characterizations
Adsorption measurements were made in an atmospheric flow
system [33] in order to determine the active surface of Pd catalysts
samples. O
other several times.
Prior to the first adsorption of O
1.2% H /Ar for 15 min and then in Ar gas to remove absorbed
2 2
titration and H titration were carried out after each
2
, the sample was treated in
2
hydrogen, to avoid the hydrogen absorption in the bulk phase of
the metal.
(Pd–H)
(0.1 ml each). Next (Pd–O)
tion of beta-PdH, O was adsorbed again.
The stoichiometry of the calculations was based on [34]:
s
was titrated with O
2
injections via a calibrated loop
s
2
was titrated with H . After decomposi-
2
ifier [28–32] is effective even in small ratio to the substrate (1:10–
5
1
:10 ). It adsorbs strongly on the metal surface (for example,
ꢀ for titration with O
2
cinchonidine), determining the adsorption geometry of the sub-
strate, which is bound to the modifier with second order interac-
tions, like H-bonding. The IP/Pr system from the point of view of
both quantitative and qualitative features belongs to the chiral
auxiliary-governed reactions. In this respect, we agree with Lam-
bert and co-workers [21], who pointed out the much stronger
adsorption of IP than that of Pr on metallic Pt surfaces.
ðPd—HÞ þ 0:75O
ꢀ for titration with H
2
¼ ðPd—OÞ þ 0:5H
2
O
s
s
2
ðPd—OÞ þ 1:5H
2
¼ ðPd—HÞ þ H
2
O
s
s
In order to ascertain the details of these Pr-mediated reactions,
the kinetic resolution of TMCH and the asymmetric hydrogenation
of IP were studied with different Pd catalysts.
2.3. Hydrogenation
Hydrogenations were carried out at 25 °C, under hydrogen pres-
sures of 10–60 bar in a 250 cm stainless steel autoclave (Techno-
3
clave) equipped with a magnetic stirrer. The solvent was methanol,
with 3.5 vol% n-hexanol content, which served as internal standard
for GC measurements. Before hydrogenation, the reaction mixtures
were boiled for 5 min, then cooled, catalyst was added, and finally
stirred under nitrogen for 10 min in the reaction vessel.
2
. Experimental
2.1. Materials
Pd/C catalyst Selcat Q, 10% metal content, was purchased from
Fine Chemical Company. Its support is a high surface area-acti-
vated carbon (BET surface area 1200 m /g). Pd black catalyst was
2.4. Analysis
2
prepared according to the following procedure: 18 mmol (6.0 g)
K PdCl was dissolved in 50 ml water and reduced at boiling point
2 4
with 74 mmol (5.0 g) Na(HCOO) dissolved in 20 ml water. When
the reduction was complete, the pH of the suspension was basic
Reaction mixtures were analyzed with a Chrompack 9001 gas
chromatograph equipped with a b-cyclodextrine capillary column
(temperature-programed analysis: 90 °C (10 min) – 10 °C/min to
160 °C) and FID. Chromatograms were recorded and the peak area
was calculated with Chromatography Station for Windows V1.6
(DataApex Ltd., Prague). As internal standard, n-hexanol was used.
The peak area of TMCH enantiomers and isophorone (the FID
detector signals for same amount of TMCH and isophorone are
identical) was correlated with that of n-hexanol, in order to deter-
mine the amount converted to alkylated proline, which cannot be
detected with GC. Enantiomeric excess was defined as:
(
pH 9). The catalyst was filtered, washed several times with dis-
tilled water, and then dried in air at ambient temperature. Its
2
BET surface area is 8 m /g.
Catalysts, 5 wt%, Pd/TiO
lows: the calculated amount of the catalyst precursor (K
was added to the aqueous suspension of 10 g support (nonporous
2
and Pd/Al
2
O
3
, were prepared as fol-
2 4
PdCl )
2
2
TiO , BET surface area 40 m /g, or powdered alumina, surface area
2
5
0 m /g). The pH value of the solution was adjusted to 10–11 by
eeð%Þ ¼ ð½Rꢁ ꢂ ½SꢁÞ=ð½Rꢁ þ ½SꢁÞ ꢃ 100
adding KOH. The suspension was boiled for 1 h, and then Na(H-

4
N. Gy o} rffy et al. / Journal of Catalysis 270 (2010) 2–8
Table 1
Results of chemisorption measurements.
2
2
Catalyst
Nominal Pd
content (w%)
Final hydrogen adsorption
(mol/g)
Final oxygen adsorption
(mol/g)
Active surface area (m /g
catalyst)
Active surface area (m /g
Pd)
Hydrogen
Oxygen
Hydrogen
Oxygen
Pd/C Selcat Q
10
5
57
98
23
26
1.8
3.1
1.4
1.6
18
62
14
32
Pd/Al
2
O
3
Pd/TiO
2
5
89
36
2.8
2.3
56
46
Pd black
Pd/BaCO
Pd/MgO
100
5
5
760
39
66
350
13
26
23.9
1.2
2.1
22.0
0.8
1.6
23.9
24
42
22.0
16
32
3
Table 2
Results of TMCH + Pr hydrogenations.
Catalyst (mg)
p(H
2
) (bar)
Reaction time (h)
Conv. (%)
ee (%)
Missing S isomer (%)
Ratea
Activity in period I
Period I
Period II
Act.b
Spec. act.c
Pd/C
5
0
0
10
10
10
15
10
3.7
61
62
79
100
100
100
24
23
54
10
14
44
4.1
6.2
21
2
1.4
2.2
20
14
22
1
2
Pd/TiO
2
1
4
0
0
10
10
30
7.5
40
73
44
100
15
57
2
17
1.3
10
0.2
0.4
4
8
Pd black
5
5
0
60
10
10
24
140
80
60
47
51
89
68
60
21
12
20
5
0.6
6
2.5
0.34
0.4
1
0.12
0.6
1
0.12
0.6
1
2 3
Pd/Al O
2
1
0
0
10
10
20
116
53
63
100
92
7
31
6
5.4
2.5
0.48
0.3
0.54
6
10.8
Pd/BaCO
3
4
1
0
0
60
10
26
3
33
3
43
3
8
0
8
0
0.79
0
0.2
0
4
0
Pd/MgO
4
1
0
0
60
10
19
47
74
58
92
95
51
19
11
4
0.3
1.0
0.28
0.4
5.6
8
a
b
c
Ratio of conversion (%) and reaction time (h).
Reaction rate/catalyst amount.
Reaction rate/Pd amount.
3
. Results and discussion
.1. Properties of Pd catalysts
In order to prove the involvement of Pd surface in the enantio-
vs. time curves for 5 mg Pd/C catalyst are shown in Fig. 1. The
amount of consumed TMCH enantiomers vs. time curves for 5
and 20 mg Pd/C catalyst is depicted in Fig. 2. Increasing the amount
of Pd/C catalyst from 5 to 20 mg, the consumption of both TMCH
enantiomers becomes much faster, the selectivity becomes smal-
ler, corresponding to expectations.
3
differentiation process, Pd on different support materials were
prepared, characterized, and used, including the basic supports
used by Török and co-workers [18,19]. The supported catalysts
have similar active surface area; see Table 1, the small value of
the high surface area carbon-supported Pd can be attributed to
the smaller b-hydride formation. Pd black has been added to
the series of supported catalysts as it turned to be the best in
the enantioselective hydrogenation of IP with dihydroapovinc-
5 mg Pd/C
1
00
100
80
60
40
20
0
8
0
Period I
Period II
aminic acid [35], (S)-
ers [10–14], and of benzylidene benzosuberone with cinchonidine
modifier [36].
a,a
-diphenyl-2-pyrrolidinemethanol modifi-
60
40
ee%
conversion (%)
2
0
0
3.2. Results of hydrogenation of TMCH–proline
The amount of the catalysts was usually varied between 2 and
0
3
6
9
12 15 18
4
0 mg. The data in Table 2 represent reactions of measurable activ-
time (h)
ity. The reaction rate and stereoselectivity of reductive alkylation
of Pr, namely the kinetic resolution of TMCH, change substantially
with the properties of Pd catalyst, with the catalyst/substrate ratio
as well as with the hydrogen pressure. Value of ee and conversion
Fig. 1. ee and conversion vs. time. (Conditions: Pd/C catalyst, 5 mg, 20 cm3
methanol (hexanol) solvent, 700 mg TMCH, 575 mg (S)-proline, hydrogen pressure
10 bar, temperature 25 °C, mixing speed 700 rpm.)

N. Gy o} rffy et al. / Journal of Catalysis 270 (2010) 2–8
5
With a less selective catalyst and at high pressure (Fig. 4), the
consumption of both diastereoisomers is more evident from the
beginning of the hydrogenation. This confirms that both diastereo-
isomers of the proline–TMCH condensate are present in the reac-
tion mixture and they are hydrogenated with different relative
rates depending on the catalyst used and on the reaction
conditions. This proves that enantiodifferentiation occurs in the
Pd/C
1
00
R
(20 mg)
R
(5 mg)
8
6
4
2
0
0
0
0
0
S
(20 mg)
(S)-proline–TMCH reaction both in the homogeneous phase, i.e.
in the solution at adduct forming and on the Pd surface in hetero-
geneous hydrogenation. The explanation of the stereoselectivity
order of the different Pd catalysts needs further investigation.
S
(5 mg)
3.3. Hydrogenation of isophorone in the presence of proline
0
2
4
6
8
10 12 14 16 18
time (h)
The debate about the mechanism of IP/Pr asymmetric hydroge-
nation [19,20,22] can be summarized in one question: whether the
optically active ketone was formed (i) exclusively by kinetic reso-
lution of the saturated ketone or (ii) by the partial hydrogenation
of the IP–Pr condensate resulting also in enantiomeric excess of
TMCH. The possible reaction pathways (Scheme 1 and Scheme 2),
were reviewed. On Scheme 1 the oxazolidinone (5) is the key inter-
mediate [4,16], the formation of its precursor, a carbinolamine is
accompanied already with the formation of a new asymmetric car-
bon. In the other reaction route, the key intermediate is an enam-
ine (8) [3,17], which has a syn and anti isomer, depending on the
relative position of the proline carboxylate and the two methyl
groups on the cyclohexane ring. The adsorption geometry of the
stereo- and diastereoisomers of both key intermediates can signif-
icantly differ in the preliminary adsorption before C@C hydrogena-
tion, this explains the enantiodifferentiation in the hydrogenation.
The reaction routes which consume IP and the TMCH enantio-
Fig. 2. Consumed TMCH enantiomers vs. time, 5 mg Pd/C, 20 mg Pd/C. (Conditions:
same as before, with 5 mg catalyst.)
It is obvious that both TMCH enantiomers participate in reduc-
tive alkylation as early as the start of the reaction. The stereoselec-
tivity of the process depends largely on the catalyst/substrate ratio.
With 5 mg catalyst the consumption of the (S)-TMCH, (the enantio-
mer in excess), is only 24% until reaching 100% ee. With 20 mg cat-
alyst, however, the corresponding value is 54%.
In all hydrogenations with different catalysts there are two
periods with respect to reaction rate: initially approximately until
4
0% conversion it is faster, followed by a significantly slower sec-
tion (see Fig. 1). The reaction parameters and results together with
the calculated reaction rate and specific activity values are summa-
rized in Table 2. The specific activities were calculated for the first
period of the hydrogenations, they can be handled as initial rates.
Pd/BaCO catalyst could only be tested at 60 bar hydrogen pres-
3
sure. It had negligible activity at 10 bar.
mers are the alkylated proline producing reaction routes: r
2 4
+ r ,
r
1
3
+ r . If optically active TMCH is formed only through kinetic res-
olution, the molar amount of consumed or missing IP + TMCH
would be always greater than the molar amount of the excess
TMCH enantiomer (S). That is why hydrogenations with different
catalysts, under different conditions were followed; the molar
amounts of the ketones were measured as a function of time and
conversion, by means of the internal standard n-hexanol. If our
The activity order of the catalysts follows the sequence of active
surface area, exception is Pd/C:
Pd=C > Pd=Al
2
O
3
ffi Pd=TiO
2
ffi Pd=MgO > Pd=BaCO ffi Pd black
3
With respect to stereoselectivity, Pd on alumina is superior, at
1
00% ee the amount of missing S enantiomer is only 7%, second
2
assumption is correct, r , the chemoselective C@C hydrogenation
is Pd/C, with 23% loss of S-TMCH. But even on the best catalyst
the hydrogenation of the minor diastereoisomer condensate takes
place, especially after the consumption of the major one (see
Fig. 3).
The observed stereoselectivity order of the catalysts, based on
the missing S isomer values at high ee, is:
of the oxazolidinones (5), or of enamines (8) produces (after hydro-
lysis) optically active TMCH. In this case, the value of molar
amount of TMCH ee could be greater than the moles of missing ke-
tones obviously only in the first period of the reaction, when the
kinetic resolution is marginal.
The results with Pd/C catalyst, the IP conversions, and amounts
of TMCH enantiomers vs. time are represented in Fig. 5. In Fig. 6,
3
the missing IP + TMCH vs. conversion for Pd/C and Pd/BaCO cata-
Pd=Al
2
O
3
> Pd=C > Pd=MgO > Pd black > Pd=BaCO
3
> Pd=TiO
2
1
00
2 3
20 mg Pd/Al O
100
8
6
4
2
0
0
0
0
0
8
6
0
0
4
0 mg Pd/MgO
(60 bar)
S
R
40
20
0
S
R
0
5
10
15
20
25
0
4
8
12 16 20 24 28
time (h)
time (h)
Fig. 3. Consumed TMCH enantiomers vs. time, 20 mg Pd/Al
Al O
2 3
(
2
O
3
. (Conditions: Pd/
catalyst, 20 mg, 20 cm methanol (hexanol) solvent, 700 mg TMCH, 575 mg
S)-proline, hydrogen pressure 10 bar, temperature 25 °C, mixing speed 700 rpm.)
Fig. 4. Consumed TMCH enantiomers vs. time, 40 mg Pd/MgO. (Conditions: Pd/
MgO catalyst, 40 mg, 20 cm3 methanol (hexanol) solvent, 700 mg TMCH, 575 mg
(S)-proline, hydrogen pressure 60 bar, temperature 25 °C, mixing speed 700 rpm.)
3

6
N. Gy o} rffy et al. / Journal of Catalysis 270 (2010) 2–8
O
COOH
COOH
COOH
CO
CO
N
H
+
N
N
-H2O
(
1)
+
N
N
OH
OH
O
O
(2)
(5)
(4)
r
1
Pd +H2
COOH
racemic
O
asymmetric hydrogenation
of isophorone
O
Pd
2S
+H2
+
O
N
H
r
r
(3)
2R
COOH
N
H
+H2O
CO
CO
-
H2O
N
N
>>
O
kinetic resolution
O
(6)
COOH
CO
Pd +H
2
r
3R
N
N
O
Pd
(
7)
>>
CO
COOH
r
4
+H2
>>
N
N
O
(6)
Pd +H2
r
3S
Scheme 1. The reaction scheme of the hydrogenation of the IP–Pr 1:1 mixture, key intermediate is oxazolidinone. (1) S-proline, (2) isophorone, (3) trimethyl cyclohexanone,
(4) carbinolamine, (5) unsaturated oxazolidinone, (6) saturated oxazolidinone, and (7) N-(trimethyl cyclohexyl) proline.
O
-
COO
COOH
COOH
OH
+
COOH
-
COO
N
+
N
N
2
-H O
+
N
H
(1)
+
N
OH
(2)
(8)
(4)
r
1
Pd +H
2
racemic
O
asymmetric hydrogenation
of isophorone
Pd
2S
+H
r
O
2
COOH
+
O
N
H
r
(3)
2R
COOH
N
H
-
-
+
H O
2
COO
COO
(3) S
+
+
-
H2O
N
N
>
>
kinetic resolution
(9)
-
COO
Pd +H
2
COOH
r
+
3R
N
N
Pd
(
7)
-
>>
COO
COOH
r
4
+H2
+
>>
N
N
(9)
Pd +H
r
3S
2
Scheme 2. The reaction scheme of the hydrogenation of the IP–Pr 1:1 mixture, key intermediate is enamine. (1) S-proline, (2) isophorone, (3) trimethyl cyclohexanone, (4)
carbinolamine, (7) N-(trimethyl cyclohexyl) proline, (8) unsaturated enamine, and (9) saturated enamine.
lysts is shown. The values are negative in the conversion range up
to 60%, meaning that during this period the TMCH ee did not arise
exclusively from kinetic resolution.
IP hydrogenation was tested with all catalysts in order to find
out their activity and enantiodifferentiating ability. The results
are summarized in Table 3. The activity order in IP hydrogenation
was similar to that of the TMCH reductive alkylation:
catalysts are depicted, giving answer to the debated question about
the sources of TMCH ee. The negative values of [missing
IP + TMCH] ꢂ [S excess] on some catalysts indicate that, up to a par-
ticular conversion, the asymmetric C@C hydrogenation of the IP–Pr
condensate takes place. The ratio of the two processes producing
excess (S)-TMCH, namely kinetic resolution and asymmetric C@C
hydrogenation cannot be established on the basis of the data in Ta-
ble 3. The consecutive hydrogenation (C@C saturation + reductive
Pd=C > Pd=Al
2
O
3
> Pd=TiO
2
ꢄ Pd=MgO > Pd=BaCO
3
> Pd black
2 4
alkylation without desorption, r + r ) cannot be excluded either.
Among supported catalysts, the activity of Pd on basic supports was
smaller. In Fig. 6, data characterizing the enantioselectivity of two
This consumes the ketone, but does not result in optically active
TMCH.

N. Gy o} rffy et al. / Journal of Catalysis 270 (2010) 2–8
7
Period I
5 mg Pd/C
10 mg Pd/C(2 bar H2)
0 mg Pd/BaCO3 (10 bar H2)
1
0
5
0
5
1
00
50
40
30
20
10
0
4
8
6
4
2
0
0
0
0
0
IP (%)
S/IP
o
R/IP
o
0
20
40
60
80
100
-
0
2
4
6
8 10 12 14 16 18 20
time (h)
-10
conversion (%)
Fig. 5. The amount of residual IP and TMCH enantiomers vs. time, 10 mg Pd/C.
Conditions: Pd/C catalyst, 10 mg, 20 cm methanol (hexanol) solvent, 700 mg IP,
75 mg Pr, hydrogen pressure 2 bar, temperature 25 °C, mixing speed 700 rpm.)
3
(
5
Fig. 6. The amount of [missing IP + TMCH] ꢂ [S excess] correlated with molar
amount of starting IP vs. conversion, 10 mg Pd/C, 40 mg Pd/BaCO . (Conditions: Pd/
C catalyst, 10 mg, 20 cm methanol (hexanol) solvent, 700 mg IP, 575 mg Pr,
o
3
3
hydrogen pressure 2 bar, temperature 25 °C, mixing speed 700 rpm. Pd/BaCO
3
Data in the last column of Table 3 can be used also for charac-
catalyst, 40 mg, 20 cm3 methanol (hexanol) solvent, 700 mg IP, 575 mg Pr,
terizing the stereoselectivity of the catalyst. The less is the relative
amount of missing ketones minus TMCH ee at the highest enantio-
meric excess in the given reaction, the better is the catalyst stere-
oselectivity with respect to both enantiomeric excess producing
processes, kinetic resolution + C@C saturation.
hydrogen pressure 10 bar, temperature 25 °C, mixing speed 700 rpm.)
carbon support was most active, when applied in small catalyst
substrate ratio (5/700). Even its selectivity was acceptable: at
1
00% ee, the amount of missing (S)-TMCH was only 24%. With re-
2 3
Pd/C at low hydrogen pressure, Pd/MgO, and Pd/Al O have the
spect to both activity and stereoselectivity, Pd on alumina catalyst
proved to be the best, at 100% ee the amount of missing (S)-TMCH
was only 7%. The catalysts on basic supports (BaCO , MgO) had no
3
best stereoselectivity with ee values of ꢅ90%. The explanation of
the chemo- and stereoselectivity differences of the Pd catalysts
on different supports needs further investigation.
special advantages. Their basicity and greater proline adsorption
ability, supposed by Török and co-workers [18] did not act in the
reductive alkylation.
4
. Conclusions
The consumption of the TMCH enantiomers and their rate of
hydrogenation in the reductive alkylation depended on the catalyst
and on the reaction conditions (hydrogen pressure and catalyst/
substrate ratio). The second enantiodifferentiating step, the hydro-
genation of the condensated products of (S)-proline and the TMCH
The kinetic resolution of TMCH with (S)-proline gave (S)-TMCH
in excess, the final ee approached 100%, but the rate of the reduc-
tive alkylation and the yield of the optically active ketone de-
pended substantially on the Pd catalyst used. Pd on activated
Table 3
IP hydrogenation results with different Pd catalysts.
Catalyst (mg)
p(H
2
) (bar)
Reaction time (h)
Conv. (%)
ee (%)
Reaction ratea
Activity in period I
à (%)
ÃÃ (%)
ÃÃÃ (%)
Period I
Period II
Act.b
Spec. act.c
Pd/C
5
0
10
2
18.1
20
12.7
4.3
99
100
99
81
87
93
92
11.6
14.9
24.4
77.4
2.4
2.5
3.7
6.8
2.3
1.5
2.4
3.9
23
15
24
39
0–80
0–60
–
ꢂ13
ꢂ4.3
–
ꢂ3.3
18.7
9
29.9
31
1
10
2
0
10
100
0–80
Pd/TiO
2
2
0
10
19.7
78
51
9.1
3.0
0.35
7
–
–
5.3
Pd black
5
0
50
10
16
80
93
20
76
10
13.7
1
2.2
0.14
2.7
0.05
2.7
0.05
0–40
–
ꢂ6.7
14.7
–
2
–
2 3
Pd/Al O
2
0
10
50
30
100
91
12.7
1.8
0.64
12.7
–
–
25.8
Pd/BaCO
4
3
0
51
49
95
64
64
25
43
22
25
4.6
2.5
1.1
1.0
1.0
0.1
0.12
0.06
0.06
2.4
1.2
1.2
0–50
0–60
0–60
ꢂ8.7
ꢂ15
ꢂ12
4.7
10
ꢂ7.5
2
0
10
ꢂ10
Pd/MgO
4
0
10
49
27.5
100
98
100
70
92
100
65
12.9
14.6
3
1.2
2.3
0.4
0.32
0.37
0.15
6.4
7.4
3.0
0–40
0–45
0–40
ꢂ3.1
ꢂ5
19.9
47.6
20
50
2
0
10
ꢂ3
ÃConversion range of IP asymmetric hydrogenation. ÃÃThe biggest negative value of [missing (IP + TMCH)] ꢂ [S excess]/IP
at the highest ee value in the reaction.
o
. ÃÃÃ[missing IP + TMCH] ꢂ [S excess]/IP
o
measured
a
Conversion (%)/reaction time (h).
Reaction rate/catalyst amount.
b
c
Reaction rate/Pd amount.

8
N. Gy o} rffy et al. / Journal of Catalysis 270 (2010) 2–8
[7] A. Tungler, T. Tarnai, T. Máthé, J. Petró, J. Mol. Catal. 70 (1991) L5.
enantiomers takes place at the metal surface of the Pd catalysts
with different rates, depending largely on the catalyst properties,
resulting in different optically active TMCH yields at 100% ee.
Asymmetric hydrogenation of IP in the presence of Pr was
investigated also with different Pd catalysts. The exact amount of
IP and TMCH enantiomers during the reaction was followed, so it
could be proven that the optically active TMCH was formed not
only by kinetic resolution but also through asymmetric C@C hydro-
genation. The activity and stereoselectivity of the different Pd cat-
alysts depended on the support material, preparation method, and
reaction conditions as well, confirming our assumption that enan-
tiodifferentiation takes also place on the catalyst surface, not only
in the homogeneous liquid phase reaction between TMCH and Pr.
The key intermediate in this reaction route can be an oxazolidi-
none (5) or an enamine (8), the literature and the present data
do not allow to select between them. The explanation of the activ-
ity and selectivity differences of the tested Pd catalysts needs also
further study.
[
8] G. Tóth, A. Kovács, T. Tarnai, A. Tungler, Tetrahedron Asymmetr. 4 (3) (1993)
31.
9] G. Fogassy, A. Tungler, A. Lévai, G. Tóth, J. Mol. Catal. A. 179 (2002) 101.
3
[
[10] É. Sipos, A. Tungler, I. Bitter, M. Kubinyi, J. Mol. Catal. A 186 (1–2) (2002) 187.
[
[
[
11] É. Sipos, A. Tungler, I. Bitter, J. Mol. Catal. A 198 (2003) 167.
12] É. Sípos, A. Tungler, I. Bitter, React. Kinet. Catal. Lett. 79 (1) (2003) 101.
13] É. Sípos, A. Tungler, React. Kinet. Catal. Lett. 80 (2) (2003) 365.
[14] É. Sípos, G. Fogassy, A. Tungler, P.V. Samant, J.L. Figueiredo, J. Mol. Catal. A 212
1–2) (2004) 245.
15] M. Joucla, J. Mortier, Bull. Soc. Chim. France 3 (1988) 579.
16] D. Seebach, A.K. Beck, D.M. Badine, M. Limbach, A. Eschenmoser, A.M.
Treasurywala, R. Hobi, W. Prikoszovich, B. Linder, Helv. Chim. Acta 90 (2007)
25.
17] B. List, L. Hoang, H.J. Martin, Proc. Natl. Acad. Sci. USA 101 (16) (2004) 5839–
842.
[18] S.C. Mhadgut, I. Bucsi, M. Török, B. Török, Chem. Commun. 8 (2004) 984.
19] S.C. Mhadgut, M. Török, J. Esquibel, B. Török, J. Catal. 238 (2006) 441.
20] A.I. McIntosh, D.J. Watson, J.W. Burton, R.M. Lambert, JACS 128 (2006) 7329.
21] A.I. McIntosh, D.J. Watson, R.M. Lambert, Langmuir 23 (2007) 6113.
[22] Shuang Li, Ensheng Zhan, Yong Li, Yide Xu, Wenjie Shen, Catal. Today 131 (1–
4) (2007) 347.
23] M. Fodor, A. Tungler, L. Vida, Catal. Today 140 (1–2) (2009) 58.
(
[
[
4
[
5
[
[
[
[
[
24] O. Roy, F. Loiseau, A. Riahi, F. Hénin, J. Muzart, Tetrahedron 59 (2003) 9641–
9
648.
[25] J.-F. Detalle, A. Riahi, V. Steinmetz, F. Hénin, J. Muzart, J. Org. Chem. 69 (2004)
528–6532.
26] P. Kukula, V. Matousek, T. Mallat, A. Baiker, Tetrahedron Asymmetr. 18 (2007)
859–2868.
[27] P. Kukula, V. Matousek, T. Mallat, A. Baiker, Chem. Eur. J. 14 (2008) 2699–2708.
Acknowledgments
6
[
The authors are grateful to Zsuzsa Koppány for the adsorption
measurements, and to Prof. Zoltán Paál for valuable discussion.
2
[
[
[
28] T. Mallat, E. Orglmeister, A. Baiker, Chem. Rev. 107 (2007) 4863–4890.
29] M. Bartók, Curr. Org. Chem. 10 (2006) 1533–1567.
30] M. Heitbaum, F. Glorius, I. Escher, Angew. Chem., Int. Ed. 45 (2006) 4732–
References
4
762.
[
[
[
1] A. Tungler, T. Máthé, J. Petró, in: Proceedings of the Second Czechoslovakian
Conf. on Het. Cat., Bechyne, 1985.
2] A. Tungler, M. Kajtár, T. Máthé, J. Petró, in: Proceedings of the Fourth Belgian-
Hungarian Conf. on Catalysis, Herbeaumont, 1987.
[31] H.U. Blaser, B. Pugin, F. Spindler, J. Mol. Catal. A: Chem. 231 (2005) 1–20.
[32] T. Mallat, S. Diezi, A. Baiker, in: G. Ertl, H. Knözinger, F. Schüth, J. Weitkamp
(Eds.), Handbook of Heterogeneous Catalysis, vol. 7, Wiley–VCH, 2008, pp.
3603–3626.
[33] G. Lu, Z. Zsoldos, Z. Koppány, L. Guczi, Catal. Lett. 24 (1–2) (1994) 15.
[34] J.R. Anderson, Structure of Metallic Catalysts, AP, London, 1975, p. 309.
[35] A. Tungler, T. Máthé, T. Tarnai, K. Fodor, G. Tóth, J. Kajtár, I. Kolossváry, B.
Herényi, R.A. Sheldon, Tetrahedron Asymmetr. 6 (9) (1995) 2395–2402.
[36] G. Fogassy, A. Tungler, A. Lévai, J. Mol. Catal. A: Chem. 192 (2003) 189–194.
3] A. Tungler, J. Kajtár, T. Máthé, T. Tóth, E. Fogassy, J. Petró, Catal. Today 5 (1989)
1
59.
[
[
[
4] A. Tungler, T. Máthé, J. Petró, T. Tarnai, J. Mol. Catal. A 61 (1990) 259.
5] A. Tungler, T. Tarnai, T. Máthé, J. Petró, J. Mol. Catal. 67 (1991) 277.
6] A. Tungler, T. Tarnai, A. Deák, S. Kemény, A. Gy o} ry, T. Máthé, J. Petró, in: M.
Guisnet, J. Barbier, J. Barrault, C. Bouchoule, D. Duprez, G. Pérot, C. Montassier
(
Eds.), Heterogeneous Catalysis and Fine Chemicals, vol. III, Elsevier Science BV,
Poitiers, 1993, p. 99.