Hydrogen pressure dependence in enantioselective hydrogenation of α,β-unsaturated acids with cinchonidine-modified Pd/TiO2 catalyst

Article abstract of DOI:10.1246/bcsj.74.1971

The enantioselectivity in the hydrogenation of (E)-2,3-diphenyl-2-propenoic acid with a cinchonidine-modified 5wt%Pd/TiO₂ catalyst decreases with increasing pressure of hydrogen up to 5 MPa contrary to the tendency observed in the hydrogenation of aliphatic acids, probably because of the difference in the adsorption strength of the substrates on both modified and unmodified sites.

Full text of DOI:10.1246/bcsj.74.1971

c. Jpn.
e Chemical
ciety of Ja-
n
Short Articles
Bull. Chem. Soc. Jpn., 74, 1971–1972 (2001) 1971
e Chemical
ciety of Ja-
n
Hydrogen Pressure Dependence in
Enantioselective Hydrogenation of
α,β-Unsaturated Acids with Cin-
chonidine-Modified Pd/TiO Cata-
lyst
2
0
1
7
7
1
2
Y. Nitta et al.
SJA8
9-2673
006
Yuriko Nitta
0
1
Niihama National College of Technology,
Yagumo-cho, Niihama, Ehime 792-8580
01
(
Received March 26, 2001)
Fig. 1. Hydrogen pressure dependence of the enantioselec-
tivity in the hydrogenations of (E)-2,3-diphenyl-2-prope-
noic acid (ꢀ) and (E)-2-methyl-2-butenoic acid (ꢁ).
0
1
2
The enantioselectivity in the hydrogenation of (E)-2,3-
diphenyl-2-propenoic acid with a cinchonidine-modified
wt%Pd/TiO catalyst decreases with increasing pressure of
.
5
2
of (E)-2-methyl-2-butenoic acid (tiglic acid, 3) as another ali-
phatic acid, by using a 5wt%Pd/TiO catalyst modified with
2
cinchonidine (CD).
hydrogen up to 5 MPa contrary to the tendency observed in
the hydrogenation of aliphatic acids, probably because of the
difference in the adsorption strength of the substrates on both
modified and unmodified sites.
As shown in Fig. 1, the enantioselectivity obtained in the
hydrogenation of 1 decreased monotonously with an increase
in the hydrogen pressure from 0.1 to 5 MPa, while the reac-
tions under lower partial pressures of hydrogen (0.02 and
0.035 MPa in helium) provided slightly lower selectivities
(54–56%ee). On the other hand, ee for the hydrogenation of 3
increased with increasing pressure of hydrogen up to 4 MPa,
Enantioselective hydrogenation of the C=C double bonds in
α, β-unsaturated acids is known to be catalyzed by palladium
catalysts modified with cinchona and vinca-type alkaloids.
1
,2
The best enantiomeric excesses (ee) of the products reported to
date are 72% and 66% for aromatic and aliphatic acids, respec-
7
similarly to the tendency reported for 2, although a slight de-
3
,4
tively. These ee values are still moderate, and the influence
of many factors, both in the catalyst preparation and reaction
conditions, remains to be clarified to give a satisfactory im-
provement of this catalytic system.
cline of ee was observed under higher H pressures. The posi-
2
tive hydrogen pressure dependence of ee, as observed for the
hydrogenations of 2 and 3, does not support our previous
3
presumption that the competitive adsorption of hydrogen and
After studying the influence of the reaction medium, we
have reported that polar solvents are preferable for the hydro-
genation of (E)-2,3-diphenyl-2-propenoic acid ((E)-α-phenyl-
CD on the catalyst surface leads to the decrease in ee under a
high pressure of hydrogen. It is well known that the enantiose-
lective hydrogenation of α, β-unsaturated acids with modified
Pd catalysts is much slower than the reactions with the unmod-
ified catalysts, in contrast to the hydrogenation of α-ketoesters
5
,6
cinnamic acid, 1), while Baiker et al. have found that nonpo-
lar solvents provide higher selectivities in the hydrogenation of
7
1,2
(
E)-2-methyl-2-pentenoic acid (2). They have proposed that
with cinchona-modified Pt catalysts. Actually, the reaction
the high hydrogen solubility in nonpolar solvents is advanta-
geous for enantiodifferentiation, based on the result that ee in
the hydrogenation of 2 in cyclohexane increases considerably
rates under an atmospheric pressure of hydrogen were de-
creased by the introduction of CD about one and two orders of
magnitude for the hydrogenations of 1 and 3, respectively.
Therefore, a slight change in the ratio of the modified and un-
modified sites will exert a considerable influence not only on
the product ee but also on the overall reaction rate. If the ratio
of the modified sites increased with increasing pressure of hy-
drogen, the overall reaction rates with the modified catalyst
must have decreased. This is in conflict with the observation
that the reaction rates with the modified catalyst, estimated
from the pressure change during the reaction, increased with
increasing initial pressure of hydrogen for both systems.
These findings indicate that the ratio of the selective sites to all
the reaction sites on the modified catalyst is kept constant un-
der different hydrogen pressures. Thus, the observed hydrogen
pressure dependence of the product ee should be explained by
the change in the relative reaction rates on modified and un-
7
with ascending pressure of hydrogen up to about 6 MPa.
However, no explanation has so far been given for the hydro-
gen pressure dependence. In the enantioselective hydrogena-
tion of 1, the highest ee was obtained in polar solvents, such as
N,N-dimethylformamide and 1,4-dioxane containing 2.5 vol%
of water, under an atmospheric pressure of hydrogen; a change
in the hydrogen pressure from 0.1 to 0.5 MPa resulted in a
3
slight decrease in ee. Therefore, it is necessary to confirm that
the two systems, i.e., aliphatic acid in nonpolar solvents and
aromatic acid in polar solvents, exhibit opposite tendencies in
the hydrogen pressure dependence, before discussing the
mechanism for this interesting reaction. Here we report on the
effects of hydrogen pressure, over a wider range, on the enanti-
oselective hydrogenation of 1 as well as on the hydrogenation

1
972 Bull. Chem. Soc. Jpn., 74, No. 10 (2001)
© 2001 The Chemical Society of Japan
modified sites.
the unmodified sites in the hydrogenations of 1, while the se-
lective reaction on the modified sites is preferentially acceler-
ated in the case of 3. The different hydrogen pressure depen-
dences of the reaction rates are explained in terms of the rela-
tive adsorption strength of these substrates both on the modi-
fied and unmodified sites.
In the reaction of 1, the acid–base interaction between 1 and
the adsorbed modifier may be strong enough, and the rate of
8
the selective reaction will hardly be affected by a change in the
hydrogen pressure. Based on the effects of solvents and addi-
tives, it has been suggested that the rate-determining step for
the selective reaction on modified sites is the product desorp-
Although the use of different solvents in each hydrogena-
tions resulted in similar tendencies in the hydrogen pressure
dependence, the use of solvents having different polarity will
exert some influence on the adsorption-desorption behavior of
6
,9
tion step, which is in harmony with the idea that hydrogen
concentration has little effect on the selective reaction rate of
1. On the other hand, the nonselective reaction of 1 with un-
6
modified catalysts was accelerated almost in proportion to the
hydrogen pressure. Thus, it is clear that the higher hydrogen
pressure brings about the larger contribution of the nonselec-
tive reaction on unmodified sites to the overall reaction rate,
leading to a decrease in the product ee in the hydrogenation of
the reactants. Furthermore, the amount and the conformation
of the adsorbed modifier can be affected by the solvent em-
ployed, which also leads to a change in the apparent ratio of
the modified sites. Therefore, it seems difficult to explain the
solvent effect on the enantioselectivity, as observed for the re-
7
1.
action of 2, by means of the hydrogen solubility only. The
6
Although the hydrogenation rate of 3 with the unmodified
solvent effect on the reaction of 1 was described elsewhere.
catalyst was also increased with increasing pressure of hydro-
gen, the extent of the acceleration was only one tenth as com-
pared to the case of 1. The marked difference in the hydrogen
pressure dependence of the activities on the unmodified cata-
lyst between the two systems can be explained in terms of the
relative concentrations of adsorbed substrates and hydrogen by
assuming the Langmuir–Hinshelwood mechanism. The ad-
sorption of 3 on Pd catalyst is expected to be weaker than that
of 1, because 1 has phenyl groups with high affinity to Pd sur-
faces. In addition, the surface concentration of hydrogen is
higher in the nonpolar solvent, used for the reaction of 3, than
in polar solvents. Therefore, the additional increase in the hy-
drogen pressure will not cause such a remarkable increase in
the reaction rate of 3 as observed in the reaction of 1. On the
other hand, when the reaction on modified sites is controlled
by the surface reaction, the selective reaction can be directly
accelerated by the increase in the surface concentration of hy-
drogen; the adsorption of 3 on modified sites will be stronger
than that on unmodified sites. Preliminary experiments on
competitive adsorption of 1 and 3, carried out with the unmod-
ified catalyst in methanol, indicated that the adsorption of 3 is
not as strong as 1, in agreement with the above expectation.
Subsequent introduction of CD into the mixture resulted in an
immediate adsorption of CD and, at the same time, in a com-
plete desorption of 3 once adsorbed on the catalyst surface, in-
dicating that the interaction of 3 with CD is much weaker than
that of 1. Furthermore, the selective reaction of 3 was scarcely
Experimental
The 5wt%Pd/TiO
2
catalyst was prepared by a precipitation-
deposition method according to the procedure described in our
11
previous paper. The modifier CD (Wako Pure Chemical, 99%),
the substrates 1 (Aldrich, 98%) and 3 (Tokyo Kasei, > 98%), and
the solvents (Wako Pure Chemical, Special grade) were used as
received. The solvents employed so as to give reasonable reaction
rates and ee values were 1,4-dioxane containing 2.5vol% of water
for 1 and hexane for 3. The reactions were performed at ambient
temperatures (ca. 293 K for 1 and ca. 303 K for 3) in a 30-mL
stainless steel autoclave equipped with a magnetic mixing system
(
1000 rpm), except for the reactions under atmospheric pressure
of hydrogen. The freshly reduced catalyst (0.02 g) was transferred
3
to a glass inlet with 10 cm of solvent, and the modifier (0.02
mmol) and the reactant (1 mmol) were added successively. After
the hydrogen uptake finished, the catalyst was filtered off and the
hydrogenation products were analyzed by HPLC on a chiral col-
umn (CHIRACEL OJ-R, DAICEL) and by GLC on a chiral capil-
lary column (Cyclodextrine-β-236M-19, Chrompack) for the hy-
drogenations of 1 and 3, respectively. The enantioselectivity is ex-
pressed as the enantiomeric excess (ee) at full conversion: ee (%)
=
100 × (S − R)/ (S + R). Detailed procedures for the catalyst
preparation and the product analyses are described elsewhere.
11
References
1
H.-U. Blaser, H.-P. Jallet, M. Muller, and M. Studer, Catal.
Today, 37, 441 (1997).
1
0
2
P. B. Wells and A. G. Wilkinson, Topics Catal., 5, 39
accelerated at all by the addition of amines or polar solvents,
(
1998).
3
4
suggesting that the rate-determining step for the selective reac-
tion of 3 is different from that for 1, i.e., not the product des-
orption step but perhaps the surface reaction step. Taking into
account these observations, we conclude that the reaction on
modified sites, rather than the reaction on unmodified sites, is
preferentially accelerated by the increase in the surface con-
centration of hydrogen, which leads to an increase in the prod-
uct ee in the hydrogenation of 3.
Thus, the different behavior in the hydrogen pressure depen-
dence of the product ee between the reactions of 1 and 3 is at-
tributed to the fact that the increase in the hydrogen pressure
brings about a fair acceleration of the nonselective reaction on
Y. Nitta and K. Kobiro, Chem. Lett., 1996, 897.
I. Kun, B. Torok, K. Felfordi, and M. Bartok, Appl. Catal.
A: General, 203, 71 (2000).
5
6
7
Y. Nitta and K. Kobiro, Chem. Lett., 1995, 165.
Y. Nitta, Topics Catal., 13, 179 (2000).
K. Borszeky, T. Mallat, and A. Baiker, Catal. Lett., 41, 199
(
1996).
8
9
Y. Nitta and A. Shibata, Chem. Lett., 1998, 161.
Y. Nitta, Chem. Lett., 1999, 635.
10 Unpublished results.
11 Y. Nitta, T. Kubota, and Y. Okamoto, Bull. Chem. Soc.
Jpn., 73, 2635 (2000).