Enantioselective hydrogenation of α-phenylcinnamic acids over cinchonidine-modified Pd/C commercial catalysts

Article abstract of DOI:10.1007/s11164-015-2313-x

Enantioselective hydrogenation of α-phenylcinnamic acid (PCA) and p,p′-dimethoxyphenylcinnamic acid (DMPCA) was studied over a variety of commercial 5 % Pd/C catalysts to reveal catalyst properties suitable for obtaining high enantioselectivity. The catalysts were characterized by CO adsorption, X-ray photoelectron spectroscopy (XPS), and transmission electron microscopy (TEM). It is confirmed that pretreatment at 353 K under atmospheric pressure of H₂ before modification with cinchonidine is very effective for all the Pd/C catalysts used here to improve the selectivity and reaction rate. It is suggested that the distribution of Pd metal particles is crucial to attain high selectivity (ee% = 79 ± 1 for PCA, 89 ± 2 for DMPCA): a uniform or eggshell-type distribution of Pd is more suitable than an egg-white or egg-yolk-type distribution. It is also suggested that the dispersion of Pd metal particles controls the enantioselectivity over cinchonidine (CD)-modified Pd/C catalysts. XPS techniques are proposed to provide a convenient method to find desirable catalysts. The choice of such Pd/C catalysts could facilitate high-throughput guided study on highly enantioselective hydrogenation of α,β-unsaturated carboxylic acids.

Full text of DOI:10.1007/s11164-015-2313-x

Res Chem Intermed
DOI 10.1007/s11164-015-2313-x
Enantioselective hydrogenation of a-phenylcinnamic
acids over cinchonidine-modified Pd/C commercial
catalysts
1
1
1
•
Haruka Sato
Tomonori Misaki
•
Takuya Mameda
1
•
Kengo Nakai
2
•
•
Yuichi Haruyama
4
3
1
•
Seiji Sonobe
Takashi Sugimura
•
Takeshi Kubota
1
•
Yasuaki Okamoto
Received: 12 July 2015 / Accepted: 3 October 2015
Ó Springer Science+Business Media Dordrecht 2015
Abstract Enantioselective hydrogenation of a-phenylcinnamic acid (PCA) and
p,p -dimethoxyphenylcinnamic acid (DMPCA) was studied over a variety of com-
0
mercial 5 % Pd/C catalysts to reveal catalyst properties suitable for obtaining high
enantioselectivity. The catalysts were characterized by CO adsorption, X-ray pho-
toelectron spectroscopy (XPS), and transmission electron microscopy (TEM). It is
confirmed that pretreatment at 353 K under atmospheric pressure of H before
2
modification with cinchonidine is very effective for all the Pd/C catalysts used here
to improve the selectivity and reaction rate. It is suggested that the distribution of Pd
metal particles is crucial to attain high selectivity (ee% = 79 ± 1 for PCA, 89 ± 2
for DMPCA): a uniform or eggshell-type distribution of Pd is more suitable than an
egg-white or egg-yolk-type distribution. It is also suggested that the dispersion of Pd
metal particles controls the enantioselectivity over cinchonidine (CD)-modified Pd/
C catalysts. XPS techniques are proposed to provide a convenient method to find
Electronic supplementary material The online version of this article (doi:10.1007/s11164-015-2313-x)
contains supplementary material, which is available to authorized users.
&
Takashi Sugimura
sugimura@sci.u-hyogo.ac.jp
1
2
3
4
Graduate School of Material Science, University of Hyogo, 3-2-1 Kohto, Kamigori, Ako-Gun,
Hyogo 678-1297, Japan
Laboratory of Advanced Science and Technology for Industry, University of Hyogo, 3-2-1
Kohto, Kamigori, Ako-Gun, Hyogo 678-1297, Japan
Graduate School of Life Science, University of Hyogo, 3-2-1 Kohto, Kamigori, Ako-Gun,
Hyogo 678-1297, Japan
Interdisciplinary Graduate School of Science and Engineering, Shimane University,
Matsue 690-8504, Japan
123

H. Sato et al.
desirable catalysts. The choice of such Pd/C catalysts could facilitate high-
throughput guided study on highly enantioselective hydrogenation of a,b-unsatu-
rated carboxylic acids.
Keywords Enantioselective hydrogenation Á Heterogeneous catalysis Á Pd/C Á
Cinchonidine Á a,b-Unsaturated carboxylic acid
Introduction
Enantiodifferentiating hydrogenation of prochiral ketones and olefins over hetero-
geneous metal catalysts modified with adsorbed chiral molecules has attracted much
attention because of the prominent advantages of industrial application of solid
catalysts, which enable easier separation of reactants and catalysts and easier
handling and rejuvenation of catalysts, compared with corresponding homogeneous
catalyst systems [1–9]. Nevertheless, heterogeneous enantioselective hydrogenation
presents great opportunities for better understandings of two-dimensional molecular
recognition chemistry on metal surface and thereby sufficient improvements of the
selectivity and reaction rate of hydrogenation to meet industrial requirements. One
prospective heterogeneous catalyst system is represented by cinchonidine (CD)-
modified palladium catalysts for asymmetric hydrogenation of C=C double bonds of
0
a,b-unsaturated carboxylic acids, such as a-phenylcinnamic acid (PCA) and p,p -
dimethoxyphenylcinnamic acid (DMPCA) (Scheme 1) [1–3, 10–13]. Addition of
auxiliary achiral amine, for example, benzyl amine (BA), significantly improves the
enantioselectivity as well as the hydrogenation rate [14], although the roles of the
added amine are still under debate [10, 11, 15–20].
A variety of supported Pd catalysts have been used for enantiodifferentiating
hydrogenation of PCA. Perez et al. [21] showed for the first time that PCA was
enantioselectively hydrogenated over CD-modified 5 % Pd/C to produce (2S)-2,3-
diphenylpropionic acid with moderate enantioselectivity of 30.5 % enantiomeric
excess (ee). Later, Nitta and coworkers [10, 11, 22–24] extensively studied catalyst
preparation to improve the enantioselectivity of the reaction. They found that a CD-
modified 5 % Pd/TiO catalyst prepared by a precipitation–deposition method was
2
more active and selective (62 % ee) for hydrogenation of PCA (298 K) than SiO -,
2
Al O -, or ZrO - supported Pd catalysts, when PdCl was precipitated at 348 K by
2
3
2
2
using Na CO on nonporous TiO and then the catalyst was reduced in H flow at
2
3
2
2
COOH
COOH
>
COOH
HO
H
H
-
2
NH
2
X
CD Pd/C
X
X
N
H
+
BA
N
X
major product
minor product
cinchonidine (CD)
benzylamine (BA)
X
X
X = H. phenylcinnamic acid (PCA)
X = OMe.4,4'-dimethxoyphenylcinnamic acid (DMPCA)
Scheme 1 Enantioselective hydrogenation of a,b-unsaturated carboxylic acids over cinchonidine-
modified Pd catalysts
1
23

Enantioselective hydrogenation of a-phenylcinnamic acids…
473 K. More recently, Nitta and Kobiro [25] attained 72 % ee over homemade Pd/
TiO catalysts by optimizing PCA reaction conditions. Subsequently, Nitta [14]
2
made prominent findings that, for PCA hydrogenation over 5 % Pd/TiO , the
2
enantioselectivity could be greatly increased from 59 to 71 % ee by addition of BA
under standard reaction conditions. Using Pd/TiO prepared by an impregnation
2
method, Kubota et al. [26] showed that CD-modified 38 % Pd/TiO exhibited the
2
record highest ee of 91 % for enantioselective hydrogenation of PCA at 288 K at a
high CD/substrate ratio and in the presence of BA. CD-modified 5 % Pd/Al O and
Pd/CeO were less selective and less active than TiO -supported Pd catalysts under
2
3
2
2
the same reaction conditions, suggesting that strong Pd metal–support interactions
are unfavorable for enantioselective hydrogenation.
Pd/C is widely used as a typical Pd metal catalyst supported on almost inactive
support. Nitta et al. [22] showed that commercial and homemade Pd/C catalysts
modified with CD were unexpectedly less active and less selective (10–21 % ee
with PCA) than Pd catalysts supported on SiO , Al O , or TiO . However, more
2
2
3
2
recently, the selectivity was reported to increase up to 49 % ee when an eggshell-
type Pd/C catalyst was employed for hydrogenation of PCA [27]. The relatively low
enantioselectivity of Pd/C catalysts were overcome by the finding by Nitta et al.
[
28] that a 5 % Pd/C commercial catalyst exhibited very high selectivity of 79 % ee
for hydrogenation of PCA in the presence of BA only when the catalyst was
pretreated at elevated temperature (353 K) in solvent under atmospheric pressure of
H prior to modification with CD, in contrast to 43 % ee without pretreatment. This
2
pretreatment was found to be applicable to specially prepared homemade Pd/C
catalysts having uniform or eggshell-type distribution of Pd, rendering them highly
selective as observed with commercial Pd/C catalysts. On the other hand, the
pretreatment was not very effective for Pd/TiO or Pd/Al O . The establishment of
2
2 3
the in situ pretreatment method has enabled us to use conventional commercial Pd/
C for highly enantioselective hydrogenation of a,b-unsaturated carboxylic acids
with high reproducibility without using specially designed and prepared homemade
Pd metal catalysts such as Pd/TiO [29, 30]. Kim et al. [31] reported that high
2
enantioselectivity of PCA hydrogenation over CD-modified Pd/C catalysts could
also obtained by successive evacuation–solvent injection–evacuation pretreatment
before CD modification.
There are a variety of commercially available Pd/C catalysts, each of which is
specially designed and prepared for certain reactions under specific reaction
conditions. In our previous study [32], we proposed to employ commercial Pd/C
catalysts for high-throughput screening research on enantioselective hydrogenation
of a,b-unsaturated carboxylic acids. In the present study, we further extend this to
provide more information on commercial Pd/C catalysts for enantioselective
hydrogenation of PCA and DMPCA as representative substrates. It should be noted
that our purpose is not to recommend any specific commercial products but to reveal
the catalytic properties suitable for enantioselective hydrogenation from a scientific
point of view.
123

H. Sato et al.
Experimental
Catalysts
The catalysts employed in the present study were wet-form 5 % Pd/C commercial
catalysts. Table 1 presents a list of the catalysts, some of which were already used in
our previous study [32, 33]. The catalysts were donated by N.E. Chemcat Co. and
Evonik Industries AG.
Hydrogenation procedures
The substrate a-phenylcinnamic acid (PCA, Aldrich; 98 %) was purified by
recrystallization from acetone solution. p,p -Dimethoxyphenylcinnamic acid
0
(
DMPCA) was synthesized from p-anisaldehyde and p-methoxyphenylacetic acid
by the Perkin reaction and characterized by nuclear magnetic resonance (NMR), as
reported previously [34]. Solvents (Wako Pure Chemicals; special grade),
cinchonidine (CD, Wako Pure Chemicals; 99 %), and benzylamine (BA, Wako
Pure Chemicals, 98 %) were used as received.
Enantioselective hydrogenation of PCA or DMPCA was conducted as reported in
our previous study [28]. Typically, wet 5 % Pd/C (23 mg as dry form, 0.01 mmol of
Pd) was pretreated in 5 ml wet dioxane (dioxane containing 2.5 % water) at 353 K
under atmospheric pressure of H for 30 min. After the catalyst had cooled to
2
2
suspension was then stirred for 30 min, while the H atmosphere was maintained.
96 K, CD (0 or 0.02 mmol) in wet dioxane was added to the catalyst. The resulting
2
Next, the substrate (PCA or DMPCA, 0.5 mmol) and then BA (0.3 mmol) was
injected into the reactor. Hydrogen uptake was measured, while the reaction
solution was maintained under atmospheric pressure of H in order to monitor the
2
progress of the reaction. Hydrogenation rates were determined at 20–25 %
conversion of the substrate. In another hydrogenation procedure, the catalyst and
CD were prereduced at 296 K by stirring for 30 min in 5 ml of the solvent under H2
Table 1 Dispersion of Pd in the
commercial 5 % Pd/C catalysts
used in the present study
Catalyst
Dispersion of Pd (%)
E-1
E-2
E-3
E-4
E-5
E-6
E-7
E-8
STD
53
49
41
45
47
45
41
39
a
Ref. [33]
b
a
Nominal dispersion
59, 57
c
b
76 (2.2 nm)
c
Particle size from extended
X-ray absorption fine-structure
b
67
AER
(EXAFS) analysis, Ref. [32]
1
23

Enantioselective hydrogenation of a-phenylcinnamic acids…
atmosphere prior to substrate injection, followed by BA addition. This latter
procedure is called ‘‘without the pretreatment’’ here.
The product ee% was determined via high-performance liquid chromatography
(
retention times: 15.4 min (R) and 24.4 min (S) at a flow rate of 1 mL/min], and the
HPLC) [Daicel Chiralpak OJ-H; 25 cm 9 4.6 mm; Hex/IPA/TFA = 95/5/0.1,
1
final reaction conversions were determined via H NMR (600 MHz) spectroscopy.
After washing each portion with 1 N HCl, the reaction mixture was found to contain
only the substrate and the corresponding hydrogenation products. The enantiomeric
excess (ee%) was calculated according to the following equation:
ee% ¼ 100  ð½SŠÀ½RŠÞ=ð½SŠ þ ½RŠÞ;
where [S] and [R] represent the concentrations of S- and R-enantiomers,
respectively.
Catalyst characterizations
We used CO adsorption, XPS, and TEM to characterize the 5 % Pd/C catalysts. The
dispersion of Pd metal particles on Pd/C was evaluated on the basis of CO
adsorption capacity using a pulse technique [33]. Wet Pd/C was dried in air at
373 K for 24 h. A certain amount of dried Pd/C (20–50 mg) was placed in a
U-shaped tube and purged with He for 10 min. The sample was then heated to
3
1
13 K in flowing He, then the He stream was replaced with a stream of H for
2
0 min, followed by flushing with He for 10 min. CO pulses (4.1 lmol/pulse) were
repeatedly injected at 313 K into the carrier gas (He) until no adsorption was
detected (3–6 pulses). The Pd dispersion was estimated from the amounts of
chemisorbed CO and Pd, assuming a ratio of 1:2 for chemisorbed CO/surface Pd
atoms.
X-ray photoelectron spectra of the Pd/C catalysts as received were measured on a
photoelectron spectrometer (CL 150, VSW Ltd.) using Mg K radiation (1253.6 eV,
a
4
to a measurement chamber. The background pressure of the spectrometer was better
50 W) [33]. The catalyst sample was pressed onto indium foil in air and transferred
-
7
than 2 9 10 Pa during the measurements. The binding energies were referenced
to the C 1s level at 248.6 eV due to the carbon supports. The background of the
spectra was removed by assuming a Shirley function.
Transmission electron microscopy (TEM) images were obtained using a
Hitachi 8000 [32]. The sample was prepared by mounting the catalyst suspension
onto carbon grids, followed by drying in ambient atmosphere. TEM images of the
sample were observed at 200 keV to reduce any electron beam-induced damage.
Results and discussion
Table 1 lists the commercial Pd/C catalysts used in the present study. The Pd metal
dispersions of the Pd/C catalysts are presented in Table 1, together with previous
results for STD and AER [32, 33]. The dispersion of Pd metal in STD (59 %) is
123

H. Sato et al.
consistent with the previous value (57 %) [33], but slightly smaller than the nominal
value (76 %). The dispersions of Pd metal in the commercial Pd/C catalysts
examined here were considerably high, ranging between 30 and 60 % after H2
treatment at 313 K.
TEM images of 5 % Pd/C were observed to provide qualitative information on
the Pd metal particles supported on carbon. Representative images are presented in
Fig. 1. For E-1, one can observe mainly well-dispersed particles with diameter of
Fig. 1 TEM images of E-1,
E-2, E-5, and E-8
E-1
E-2
x2
1
00nm
x2
1
00nm
x2
E-5
5
0nm
E-8
x2
1
00nm
1
23

Enantioselective hydrogenation of a-phenylcinnamic acids…
3
–4 nm, accompanying a small number of larger particles. Unfortunately, we failed
to observe smaller particles (\3 nm) using the present TEM settings. Similar TEM
images were observed for E-2, E-3, and E-4. As shown in Fig. 1 for E-2, most of the
Pd particles in view are present as 20–30 nm particles, with fewer smaller particles
of \4 nm. In the images of E-5, Pd metal particles are well dispersed (\3–4 nm)
with almost no larger particles being observed. E-6 and E-7 showed TEM images
similar to E-5. On the other hand, in the images of E-8 one can observe a
homogeneous distribution of relatively large Pd metal particles of 20–30 nm in
diameter. It should be noted that Pd metal particles of \3 nm escaped from the
present TEM observations due to the limitation of resolution. With STD and AER,
we previously reported TEM images showing that Pd metal particles of \2.5 nm
were well dispersed on STD and AER, while some larger Pd metal particles were
also observed to be present in the latter catalyst [32].
Table 2 Enantioselective hydrogenation of PCA on 5 % Pd/C: reaction rate and selectivity (ee)
Catalyst
Amount of CD
added (mmol)
Without the pretreatment
With the pretreatment
r (mmol/g-h)
ee (%)
r (mmol/g-h)
ee (%)
E-1
E-2
E-3
E-4
E-5
E-6
E-7
E-8
STD
AER
a
0
132
55
–
82
–
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.02
.02
.02
.02
.02
.02
.02
.02
.02
.02
53
–
51
70
–
136
64
120
74
49
–
75
–
114
62
102
86
60
–
79
–
90
82
40
47
–
47
72
–
136
72
128
93
56
–
78
–
144
64
102
77
49
–
78
–
108
66
111
111
93
56
–
80
–
96
53
58
–
84
80
–
c,b
79 , 81
120
–
a,b
c
58, 43 , 42
a,b
c,b
93 , 93
a,b
a,b
47, 48
–
–
–
–
a,b
66
a,b
41 , 59
d
c,b
c,b
102
81
Ref. [33]
b
Reaction conditions are the same except the amount of BA (0.5 mmol instead of 0.3 mmol in the
present study) added to the reaction mixture
c
Ref. [28]
d
Ref. [32]
123

H. Sato et al.
Hydrogenation reactions of PCA and DMPCA were conducted at 296 K over
pristine (without CD modification) and CD-modified 5 % Pd/C catalysts with or
without the pretreatment at 353 K. The reaction rate (r) and the enantioselectivity
(ee) are summarized in Tables 2 and 3 for PCA and DMPCA, respectively. First, we
compare the activities of pristine 5 % Pd/C (CD = 0 in Table 2) differently
pretreated before the PCA hydrogenation to examine the effect of the pretreatment
at 353 K under atmospheric H on the reaction rate. It is noted that this reaction
2
forms just racemic products, since this is a simple hydrogenation of the olefin on
5
% Pd/C. It is obvious that the reaction rate is reduced by 5–10 % by the
pretreatment at 353 K, except for E-1 which shows a larger decrease and E-7 which
exhibits even a 10 % increase. It is considered that the activity decrease can be
ascribed to slight sintering of Pd metal particles during the pretreatment at elevated
temperature. In fact, Kim et al. [32] noted with STD by use of EXAFS that the
pretreatment induced a slight increase in the Pd–Pd coordination number. It is
expected that Pd metal particles in E-1 experience more extensive sintering than the
other 5 % Pd/C. The increase in the activity of E-7 may be ascribed to removal of a
relatively higher level of impurities on the Pd metal surface under atmospheric
pressure of H at 353 K to compensate the possible sintering effect. However,
2
further study is required to rationalize the specific behaviors of E-1 and E-7.
Table 3 Enantioselective
Catalyst Amount of CD added (mmol) With the pretreatment
hydrogenation of DMPCA on
5 % Pd/C: reaction rate and
selectivity (ee)
r (mmol/g-h) ee (%)
E-1
E-2
E-3
E-4
E-5
E-6
E-7
E-8
STD
AER
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
20
27
26
28
30
40
18
14
37
33
30
36
27
53
23
41
–
–
.02
.02
.02
.02
.02
.02
.02
.02
.02
.02
84
–
86
–
87
–
85
–
88
–
88
–
90
–
91
a
Ref. [28]
–
b
Reaction conditions are the
same except the amount of BA
a
32 , 43
a,b
a
89 , 92
a,b
(0.5 mmol instead of 0.3 mmol
–
–
in the present study) added to
the reaction mixture
a,b
60
a,b
92
1
23

Enantioselective hydrogenation of a-phenylcinnamic acids…
As shown in Table 2, the hydrogenation activity of pristine Pd/C is considerably
decreased by CD modification (CD = 0.02 mmol), obviously due to poisoning
effects of the Pd metal surface by adsorbed CD molecules that suppress adsorption
and subsequent hydrogenation of PCA. In contrast to the pristine Pd/C without CD
modification, however, the pretreatment at 353 K under atmospheric pressure of H2
increased the reaction rate of CD-modified Pd/C as well as the enantioselectivity, as
reported previously [28]. Taking into account both the slight sintering of Pd metal
particles and the great increase in selectivity by the pretreatment, it is considered
that the simultaneous increases in the rate and selectivity are correlated to an
increase in the number of favorable surface enantioselective ensemble sites created
by adsorbed CD molecules, where the substrate (PCA) can be adsorbed through
substrate–surface CD interactions and subsequently hydrogenated enantioselec-
tively. It is considered that the H pretreatment at 353 K causes removal of surface
2
impurities and/or exposure of specific Pd metal facets to provide more abundant
surface ensembles, allowing increased amounts of CD molecules to be adsorbed and
thus an increased number of substrate–surface CD reaction intermediates. These
surface changes by the pretreatment may also allow adsorption of more abundant
CD molecules in a favorable configuration. It was well envisaged in our previous
study [34, 35] that stronger interactions between the surface CD and a substrate
induce higher ‘‘ligand acceleration’’ effects, accompanying enhanced hydrogenation
rate and enantioselectivity.
Table 3 presents the reaction rate and selectivity of the enantioselective
hydrogenation of DMPCA over CD-modified Pd/C pretreated at 353 K under
atmospheric pressure of H . In contrast to the enantioselective hydrogenation of
2
PCA, the hydrogenation rate is increased by CD modification. As revealed by
Fig. 2 Pd 3d XP spectra for a
series of 5 % Pd/C commercial
catalysts as received
123

H. Sato et al.
Mameda et al. [34], the substrate DMPCA shows much stronger ‘‘ligand acceler-
ation’’ effects (370 %) than PCA (130 %). It is obvious that the activity increase by
CD modification in Table 3 results from the strong interactions between adsorbed
CD molecules and DMPCA, thus reaching higher selectivity of up to 91–92 % ee
than that of PCA (80–81 % ee). Sz o¨ ll o¨ si et al. [36, 37] reported as high a selectivity
as shown in Tables 2 and 3 for enantioselective hydrogenation of PCA (80 %) and
DMPCA (90 %) over CD-modified 5 % Pd/Al O at 294 K.
2
3
The chemical state and dispersion/distribution of Pd species of the Pd/C catalysts
were estimated using XPS. The X-ray photoelectron (XP) spectra of the Pd 3d level
are presented in Fig. 2 for the series of as-received 5 % Pd/C catalysts listed in
Table 1. The XP spectra for STD and AER have already been reported elsewhere
[
33]. It is noted that all the XP spectra were measured under ex situ conditions. The
wide-scan spectra of the Pd/C catalysts showed no detectable impurities on the
catalysts except oxygen. The Pd 3d spectra apparently show the presence of more
than one Pd species. We found that the spectral envelopes were best fit by assuming
two Pd components whose Pd 3d5/2 binding energies (BEs) are summarized in
Table 4. The Pd species with Pd 3d5/2 BE of 335.3 eV is assigned to Pd metal, while
2
?
the one at 336.9 eV is assigned to Pd species, possibly PdO-like species [38],
indicating that part of the Pd in the commercial 5 % Pd/C catalysts is oxidized as
received or during storage. The fraction of Pd metal estimated by deconvolution is
summarized in Table 4. Palladium is heavily oxidized in E-1 through E-6, while Pd
metal prevails in E-7 and E-8. It is considered that the chemical states of Pd in the
catalyst depend on the preparation procedures and history of sample handling and
storage. We could not, however, find any correlation between the catalytic activity
of 5 % Pd/C in Table 2 or 3 and the oxidation state of Pd in Table 4, indicating that
the initial states of Pd are not very important in the present hydrogenations. Our
Table 4 XPS analysis of 5 % Pd/C commercial catalysts as received
Catalyst
BE of Pd 3d5/2 level (eV)
Fraction of
Pd metal (%)
Pd 3d/C 1s XPS
intensity ratio
Metal
Oxides
E-1
E-2
E-3
E-4
E-5
E-6
E-7
E-8
335.3
335.2
335.2
335.2
335.2
335.2
335.3
335.3
335.3
335.3
337.0
336.9
336.9
336.8
336.9
336.9
336.9
336.9
336.9
336.9
32
17
29
25
35
5
0.47
0.45
1.25
0.33
1.22
1.76
0.68
0.65
0.92
1.42
77
85
40
40
a
STD
a
AER
a
Ref. [33]. The XPS intensity Pd 3d5/2/C 1s ratio in the reference is converted to the Pd 3d/C 1s ratio by
multiplying the value by 10/6
1
23

Enantioselective hydrogenation of a-phenylcinnamic acids…
previous EXAFS results showed that oxidized Pd species were readily reduced
during the pretreatment at 353 K or even during the CD modification at the reaction
temperature [32]. The TEM images of as-received Pd/C in Fig. 1 were discussed
above by assuming the formation of ‘‘Pd metal particles’’ instead of the presence of
PdO-like species; It is rational to assume that PdO-like particles in the Pd/C
catalysts are readily reduced to Pd metal particles under electron beam flooding
during TEM observation.
Together with the information on the chemical states of Pd, XPS can provide
clear information on the dispersion and/or distribution of Pd in the supported
catalysts by analyzing signal intensities, given that the supported amount of Pd is
the same [39, 40]. With 5 % Pd catalysts supported on carbon, the XPS intensity
ratio of the Pd 3d level due to the active species to that of the C 1s level due to the
support, that is, the Pd 3d/C 1s XPS intensity ratio, reflects both the dispersion of Pd
on the surface (particle size) and the distribution of Pd inside the catalyst particles
(
depth profile). Table 4 summarizes the Pd 3d/C 1s intensity ratio for the 5 % Pd/C
catalysts used here together with those for STD and AER. It is noted that STD is a
commercial catalyst with a uniform distribution of Pd, while AER has an eggshell-
type distribution of Pd [33]. In agreement with the Pd distribution, AER shows a
considerably higher Pd 3d/C 1s ratio than STD. Figure 3 shows a correlation of the
Pd 3d/C 1s XPS intensity ratio and the dispersion of Pd in Table 1. The dotted line
in Fig. 3 shows a correlation estimated for a uniform dispersion [39, 40], assuming
that Pd is uniformly distributed in STD. The intensity ratios for E-7 and E-8 lie on
the line, indicating that Pd metal particles are uniformly distributed in the catalyst
particles in these catalysts, although the particle size of Pd metal is smaller in STD
than in E-7 or E-8 (Table 1).
As shown in Fig. 3, the intensity ratios for E-3, E-5, and E-6 are far above the
line, while the ratios for E-1, E-2, and E-4 are apparently below the line. It is
concluded that E-3, E-5, and E-6 have an eggshell-type distribution of Pd, whereas
2
1
1
1
1
.8
.6
.4
.2
1
E-6
eggshell
E-5
E-3
uniform
0
0
0
0
.8
.6
.4
.2
0
E-8
STD
E-7
E-1
E-2
50
Egg white/yolk
E-4
30
40
60
70
Dispersion of Pd/%
Fig. 3 Correlation between Pd 3d/C 1s XPS intensity ratio and Pd dispersion for 5 % Pd/C catalysts. The
dotted line shows a correlation estimated for a uniform distribution of Pd, assuming a uniform distribution
of Pd in STD
123

H. Sato et al.
E-1, E-2, and E-4 have an egg-white- or egg-yolk-type distribution inside these
catalyst particles. Considering that the catalyst particles are provided in fine powder
form for the present catalysts, it is estimated that the Pd metal particles in E-1, E-2,
and E-4 are embedded relatively deep in the pores of the support.
We now discuss the catalyst properties of 5 % Pd/C suitable for enantioselective
hydrogenation of PCA or DMPCA. As presented in Tables 2 and 3, the
enantioselectivity or ee% considerably depends on the catalyst and the pretreatment
before the reaction. We confirmed in the present study the importance of the
pretreatment of Pd/C at 353 K under atmospheric pressure of H2 before
modification with CD to obtain high selectivity as well as enhanced reaction rate
[
28]. Thus, we focus here on the enantioselectivity of the catalysts pretreated at
53 K.
It has been revealed that the selectivity of the enantiodifferentiating hydrogena-
3
tion of PCA over CD-modified Pd catalysts depends on a variety of catalyst
properties as well as the reaction conditions and pretreatments. Regarding catalyst
properties, it is suggested that the selectivity and reaction rate are affected by the
size and distribution of Pd metal particles [10, 11, 23, 41], the pore size of the
support where Pd metal particles are embedded [27], and the support material,
which may modify the Pd metal particle–support interactions and thereby
adsorption properties of the substrate and modifier molecules involved [10, 11,
2
3, 24, 42, 43]. Nitta et al. [23, 41] revealed that, with CD-modified Pd/TiO , the
2
highest selectivity for PCA hydrogenation was attained at moderate Pd dispersion of
0–25 %, with the selectivity gradually decreasing for higher or lower dispersion.
However, as presented in Table 2, STD shows a very high ee at Pd dispersion of
9 %. Thus, the dependence of ee on the dispersion of Pd or the particle size of Pd
2
5
metal is different between TiO - and carbon-supported Pd catalysts, as previously
2
suggested by Kubota and coworkers [33]. A comparison between the selectivity
ee%) on CD-modified Pd/C pretreated at 353 K (Table 2 or 3) and the dispersion of
(
100
DMPCA
90
80
70
60
PCA
1.5
0
0.5
1
2
Pd3d/C1s XPS intensity raꢀo
Fig. 4 Correlation between enantioselectivity of hydrogenation of PCA or DMPCA over CD-modified
% Pd/C and Pd 3d/C 1s XPS intensity ratio
5
1
23

Enantioselective hydrogenation of a-phenylcinnamic acids…
Pd in Table 1 showed no simple correlation between the dispersions of 30 and
6
0 %, as presented in Fig. S1 (supplementary material) for PCA, suggesting that
other factors may also be important in controlling the selectivity of the Pd/C
catalysts in the enantioselective hydrogenation of PCA and DMPCA.
Figure 4 presents the enantioselectivities of the hydrogenation of PCA and
DMPCA on CD-modified 5 % Pd/C as a function of the Pd 3d/C 1s XPS intensity
ratio. It is highly significant that the selectivity is nicely correlated to the Pd 3d/C
1
s ratio. The selectivity of PCA hydrogenation is high and remained almost constant
at ee% = 79 ± 1 at Pd 3d/C 1s ratio [0.65, above which ratio Pd metal particles
are uniformly distributed over the catalyst particles or have an eggshell-type
distribution inside the catalyst particles, as presented in Fig. 3. On the other hand,
the selectivity decreases to a moderate value at Pd 3d/C 1s ratio \0.65, where Pd
metal particles are estimated to have an egg-white/yolk-type distribution inside the
catalyst particles. For DMPCA hydrogenation, an identical correlation is obtained
between the enantioselectivity and the Pd 3d/C 1s ratio, although the ee values are
higher (ee% = 89 ± 2 at Pd 3d/C 1s ratio [0.65) than those for PCA hydrogena-
tion. The high selectivities of the PCA and DMPCA hydrogenations on CD-
modified STD and AER can also be ascribed to preferable distributions of Pd metal
particles, that is, uniform and eggshell-type distributions, respectively. Thus, it can
be concluded that the distribution of Pd metal in the catalyst particles is one of
major catalyst properties determining the selectivity in the enantioselective
hydrogenation over CD-modified Pd/C. Nitta et al. [23, 24] found that Pd/TiO2
prepared using a nonporous TiO support showed excellent selectivity, suggesting
2
the effects of Pd location and/or distribution of Pd inside catalyst particles on the
selectivity. In fact, Nitta and Okamoto [27] showed that Pd metal particles
embedded into the mesopores of MCM-41 (pore diameter 2.2 nm) or KIT-1
(
2.8 nm) exhibited considerably low selectivity in the hydrogenation of PCA,
possibly due to strong steric hindrance to CD modification and/or to effective
formation of surface intermediates involving CD and PCA. These findings are
correlated to a moderate selectivity over the catalysts E-1, E-2, and E-4 with egg-
white/yolk-type distribution of Pd metal particles, where these metal particles are
expected to be located relatively deep in the small pores of the supports.
Scrutinizing the correlations in Fig. 4, however, it is likely that the ee values
attained around Pd 3d/C 1s = 0.65 very slightly decrease as the intensity ratio
increases, suggesting that, with uniform distribution catalysts, Pd metal particles
with dispersion of 40 % (E-7, E-8) are more favorable for the hydrogenations than
those with 59 % dispersion (STD). It is concluded that the enantioselectivity of CD-
modified 5 % Pd/C depends on the particle size of Pd metal, but differently from the
dependence for Pd/TiO [23, 41]. Consequently, we propose that 5 % Pd/C catalysts
2
with uniform or eggshell-type distribution of Pd metal particles, possibly with Pd
dispersion of 40 %, can be selected for highly selective enantiodifferentiating
hydrogenation of a,b-unsaturated carboxylic acids. These Pd/C catalysts could be
used for high-throughput directed study in the development of enantioselective
hydrogenation. In addition, the uniform correlations in Fig. 4 suggest that XPS
techniques provide a very convenient method to select Pd/C catalysts suitable for
enantioselective hydrogenation of a,b-unsaturated carboxylic acids.
123

H. Sato et al.
Conclusions
We conducted enantioselective hydrogenation of a-phenylcinnamic acid (PCA) and
p,p -dimethoxyphenylcinnamic acid (DMPCA) over a variety of commercial 5 %
0
Pd/C catalysts to find catalyst properties suitable for obtaining high enantioselec-
tivity. It was confirmed that the pretreatment of as-received Pd/C at 353 K under
atmospheric pressure of H before modification with cinchonidine (CD) was very
2
effective for all the Pd/C catalysts used here to improve the selectivity and reaction
rate. On the basis of a correlation between the selectivity and the Pd 3d/C 1s XPS
intensity ratio, we suggest that the distribution of Pd metal particles is a crucial
catalyst property to attain high selectivity (ee% = 79 ± 1 for PCA, 89 ± 2 for
DMPCA): a uniform or eggshell-type distribution of Pd is more desirable than an
egg-white- or egg-yolk-type distribution. It is also suggested that the enantioselec-
tivity of CD-modified 5 % Pd/C depends on the dispersion of Pd. We suggest that
5
% Pd/C catalysts with uniform or eggshell-type distribution and Pd dispersion of
around 40 % could be used to facilitate high-throughput guided study on
enantioselective hydrogenation of a,b-unsaturated carboxylic acids. We propose
that XPS techniques provide a very convenient method to select Pd/C catalysts
suitable for enantioselective hydrogenation of a,b-unsaturated carboxylic acids.
Acknowledgments We would like to express our sincere thanks to Evonik Industries AG and N.E.
Chemcat Co. for donating the Pd/C catalysts used in the present study.
References
1
. T. Mallat, E. Orglmeister, A. Baiker, Chem. Rev. 107, 4863 (2007)
. A. Baiker, H. U. Blaser, in Handbook of Heterogeneous Catalysis, Eds. G. Ertl, H. Kn o¨ zinger, J.
Weitkamp, VCH, Weinheim, Vol.5, p. 2422 (1997)
2
3
4
5
6
7
8
9
. E. Zhan, C. Chen, Y. Li, W. Shen, Catal. Sci. Technol. 5, 650 (2015)
. M. Studer, H.U. Blaser, C. Exner, Adv. Synth. Catal. 345, 45 (2003)
. H.U. Blaser, F. Spindler, M. Studer, Appl. Catal. A 221, 119 (2001)
. D.Y. Murzin, P. Maki-Arvela, E. Toukoniitty, T. Salmi, Catal. Rev. Sci. Eng. 47, 175 (2005)
. Y. Izumi, Adv. Catal. 32, 215 (1983)
. H.U. Blaser, Tetrahedron Asymmetry 2, 843 (1991)
. T. Sugimura, Catal. Surv. Jpn. 3, 37 (1999)
1
1
1
1
0. Y. Nitta, Top. Catal. 13, 179 (2000)
1. Y. Nitta, J. Synth. Org. Chem. Jpn. 64, 827 (2006)
2. A. Tungler, E. Sipos, H. Hada, Curr. Org. Chem. 10, 1569 (2006)
3. T. Sugimura in Handbook of Asymmetric Heterogeneous Catalysis Eds. K. Ding, Y. Uozumi, Wiley-
VCH, Verlag, Weinheim, p. 357 (2008)
1
1
1
1
1
1
2
4. Y. Nitta, Chem. Lett. 28, 635 (1999)
5. T.Y. Kim, T. Sugimura, J. Mol. Catal. A 327, 58 (2010)
6. G. Sz o¨ ll o¨ si, B. Herman, F. Fulop, M. Bartok, J. Catal. 276, 259 (2010)
7. G. Szollosi, T. Hanaoka, S. Niwa, F. Mizukami, M. Bartok, J. Catal. 231, 480 (2005)
8. Z. Makra, G. Sz o¨ ll o¨ si, M. Bartok, Catal. Today 181, 56 (2012)
9. F. Meemken, N. Maeda, K. Hungerbuhler, A. Baiker, ACS Catal. 2, 464 (2012)
0. K. Nakai, T. Misaki, Y. Okamoto, T. Sugimura, Bull. Chem. Soc. Jpn. 88, 300 (2015)
1
23

Enantioselective hydrogenation of a-phenylcinnamic acids…
2
2
2
2
2
2
1. J.R.G. Perez, J. Malthete, J. Jacques, C. R. Acad. Sci. Paris 300, 169 (1985)
2. Y. Nitta, Y. Ueda, T. Imanaka, Chem. Lett. 23, 1095 (1994)
3. Y. Nitta, T. Kubota, Y. Okamoto, Bull. Chem. Soc. Jpn. 73, 2635 (2000)
4. Y. Nitta, K. Kobiro, Y. Okamoto, Stud. Surf. Sci. Catal. 108, 191 (1997)
5. Y. Nitta, K. Kobiro, Chem. Lett. 25, 897 (1996)
6. T. Kubota, H. Kubota, T. Kubota, E. Moriyasu, T. Uchida, Y. Nitta, T. Sugimura, Y. Okamoto, Catal.
Lett. 129, 387 (2009)
27. Y. Nitta, Y. Okamoto, Chem. Lett. 27, 1115 (1998)
28. Y. Nitta, J. Watanabe, T. Okuyama, T. Sugimura, J. Catal. 236, 164 (2005)
29. T. Sugimura, T. Uchida, J. Watanabe, T. Kubota, Y. Okamoto, T. Misaki, T. Okuyama, J. Catal. 262,
5
7 (2009)
0. T. Misaki, H. Otsuka, T. Uchida, T. Kubota, Y. Okamoto, T. Sugimura, J. Mol. Catal. A 312, 48
2009)
3
(
3
1. T.Y. Kim, M. Yokota, T. Uchida, T. Sugimura, Catal. Lett. 131, 279 (2009)
2. T.Y. Kim, T. Uchida, H. Ogawa, Y. Nitta, T. Okuyama, T. Sugimura, S. Hirayama, T. Honma, M.
Sugiura, T. Kubota, Y. Okamoto, Top. Catal. 53, 116 (2010)
3
33. T. Kubota, H. Ogawa, Y. Okamoto, T. Misaki, T. Sugimura, Appl. Catal. A 437/438, 18 (2012)
34. T. Mameda, K. Nakai, T. Misaki, Y. Okamoto, T. Sugimura, Catal. Today 245, 129 (2015)
35. H. Ogawa, T. Mameda, T. Misaki, Y. Okamoto, T. Sugimura, Chem. Lett. 42, 813 (2013)
3
6. G. Sz o¨ ll o¨ si, B. Herman, K. Felfoldi, F. Fulop, M. Bartok, Adv. Synth. Catal. 359, 2804 (2008)
3
7. G. Sz o¨ ll o¨ si, B. Herman, K. Felfoldi, F. Fulop, M. Bartok, J. Mol. Catal. A 290, 54 (2008)
38. C.D. Wagner, W.M. Riggs, L.E. Davis, J.E. Moulder, G.E. Mullenberg, Handbook of X-Ray Pho-
toelectron Spectroscopy (PerkinElmer Corp., Minnesota, 1979)
3
4
4
4
4
9. J. W. Niemantsverdriet, Spectroscopy in Catalysis, Chap. 3, VCH, Weinheim, p. 52 (1995)
0. F.P.J.M. Kerkhof, J.A. Moulijn, J. Phys. Chem. 83, 1612 (1979)
1. Y. Nitta, T. Kubota, Y. Okamoto, Bull. Chem. Soc. Jpn 74, 2161 (2001)
2. T.S.B. Trung, Y. Kim, S. Kang, H. Lee, S. Kim, Catal. Commun. 66, 21 (2015)
3. K. Szori, R. Puskas, G. Sz o¨ ll o¨ si, L. Bertoti, J. Szepvolgyi, M. Bartok, Catal. Lett. 14, 539 (2013)
123