Kinetic analysis of the asymmetric hydrogenation of (: E)-2,3-diphenylpropenoic acid over cinchonidine derivative-modified Pd/C: Quinoline ring modification

Article abstract of DOI:10.1039/d0cy01380c

The effects of the quinoline ring modification of cinchonidine (CD) on the enantioselectivity of the asymmetric hydrogenation of (E)-2,3-diphenylpropenoic acid over chirally modified Pd/C were systematically analyzed from the kinetic points of view. The substitutions at the 2′- and/or 6′-positions of the quinoline ring of CD by a methyl, vinyl, n-butyl, or phenyl group decreased enantioselectivity over the whole range of the modifier concentration. Kinetic analysis allowed us to estimate the intrinsic enantioselectivity at modified sites and adsorption strength of the modifier. It is revealed that the substitutions reduce both the intrinsic enantioselectivity and adsorption strength of the parent modifier. The intrinsic enantioselectivity is correlated, most likely, to the modifier-substrate interaction strength. This journal is

Full text of DOI:10.1039/d0cy01380c

Catalysis
Science &
Technology
View Article Online
View Journal
PAPER
Kinetic analysis of the asymmetric hydrogenation
of (E)-2,3-diphenylpropenoic acid over
cinchonidine derivative-modified Pd/C: quinoline
ring modification†
Cite this: DOI: 10.1039/d0cy01380c
* *
and Takashi Sugimura
Makoto Nakatsuji, Morifumi Fujita,
Yasuaki Okamoto
The effects of the quinoline ring modification of cinchonidine (CD) on the enantioselectivity of the
asymmetric hydrogenation of (E)-2,3-diphenylpropenoic acid over chirally modified Pd/C were
systematically analyzed from the kinetic points of view. The substitutions at the 2′- and/or 6′-positions of
the quinoline ring of CD by a methyl, vinyl, n-butyl, or phenyl group decreased enantioselectivity over the
whole range of the modifier concentration. Kinetic analysis allowed us to estimate the intrinsic
enantioselectivity at modified sites and adsorption strength of the modifier. It is revealed that the
substitutions reduce both the intrinsic enantioselectivity and adsorption strength of the parent modifier.
The intrinsic enantioselectivity is correlated, most likely, to the modifier–substrate interaction strength.
Received 9th July 2020,
Accepted 8th August 2020
DOI: 10.1039/d0cy01380c
rsc.li/catalysis
enantiodifferentiating
modifier–substrate
interaction
Introduction
intermediates, thus inducing surface chiral recognition.1–10
The geometrical features of the adsorbed modifier define
the configuration of the chiral pocket and, thereby, the
energetics and dynamics of pro-R and pro-S modifier–
Enantioselective hydrogenations of α-ketoesters and α,β-
unsaturated carboxylic acids over cinchona alkaloid-modified
platinum and palladium metal heterogeneous catalysts,
respectively, have received extensive attentions because of
their pivotal significance in the molecular recognition
catalysis bestowed on a chirally modified metal surface as
well as in the feasibility of industrial applications.1–10
Spectroscopic and theoretical investigations, coupled with
molecular design approaches, have been extensively
conducted, in particular, with chirally modified platinum
substrate–metal
surface
intermediates,
consequently
elucidating intrinsic enantioselectivity at the modified
sites.1,2,5–7 The molecular structure of cinchona alkaloids is
composed of three parts: a quinuclidine base, quinoline ring,
and hydroxy-substituted stereogenic carbon linkage at C9
connecting them.1,2,5,6 It has been revealed from experimental
and theoretical studies that these molecular parts play crucial
roles as chiral modifiers to define the selectivity and reaction
rate in the enantioselective hydrogenations over platinum and
catalysts
to
obtain
deeper
insights
into
the
enantiodifferentiation mechanism11–20 and to extend the
scope of the substrates.1,6,21–24 Much less attention, however,
has been paid to their palladium counterparts.1,6,19,20,25–30
Despite the syntheses of artificial modifiers and cinchona
alkaloid derivatives by several research groups,1,2,6,31–37
naturally occurring cinchona alkaloids, cinchonidine (CD),
cinchonine (CN), quinine (QN), and quinidine (QD), have
been the most effective modifiers to yield even almost
complete enantioselectivity with specified modifier–substrate
combinations (Scheme 1).21–24,28–30 Extensive efforts have
been devoted to understand the nature of a chiral pocket
which is formed by a modifier adsorbed on a metal surface to
accommodate the substrate and thereby to form
Graduate School of Material Science, University of Hyogo, Kohto, Kamigori, Hyogo
678-1297, Japan. E-mail: yokamoto@sci.u-hyogo.ac.jp, sugimura@sci.u-hyogo.ac.jp
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
d0cy01380c
Scheme 1 Molecular structures of the modifiers frequently employed
for the enantioselective hydrogenations over platinum and palladium
metal catalysts. CD; cinchonidine, CN; cinchonine, QN; quinine, and
QD; quinidine.
This journal is © The Royal Society of Chemistry 2020
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Table
1
Cinchonidine derivatives used as
a
modifier for the
palladium metal catalysts. The quinoline ring anchors the
cinchona molecules on the metal surface through adsorption
via π-electron and/or N-lone pair electron interactions with the
metal surface. The N-atom of the quinuclidine moiety interacts
with a substrate molecule through N–H–O hydrogen bond
interactions and constitutes the chiral pocket together with the
OH group attached to C9 of the linker part. The stereogenic
configurations of C8 and C9 define the sense of
enantioselectivity in the asymmetric hydrogenations over
cinchona-modified novel metal catalysts.1,2,5–7
enantioselective hydrogenation of PCA over Pd/C
Modifier
R2
R6
CD
H
H
H
H
H
CD–Me
CD–Vinyl
CD–n-Bu
CD–Ph
QN
Methyl
Vinyl
n-Butyl
Phenyl
H
H
Methoxy
Methoxy
OH
QN–Me
CD-6′OH
Methyl
H
Aiming at performance improvement of the modifier and
at better understanding of the reaction mechanisms,
molecular design approaches have been applied to find the
effects of the modifications of the quinuclidine moiety33,38,39
and C9-hydroxy group,12,38–47 combined with spectroscopic
and theoretical approaches. However, the effect of the
modification of the quinoline ring has rarely been
investigated with both platinum and palladium metal
catalysts for the asymmetric hydrogenations of CO and
CC bonds, respectively, in spite of its crucial influence on
the adsorption strength and geometry, which are expected to
result in conformational alterations of the chiral
pocket.40,41,48,49 Comparing CD and QN or CN and QD, Huck
et al. have shown that the substitution of the 6′-position of
the quinoline ring with a methoxy group weakens the
adsorption strength of the modifier on platinum41 and
palladium40 metal surfaces, accompanying a decrease in the
modified sites, modifier–substrate interaction mode and
interaction energy, and reaction rate constant at the modified
sites relative to that at unmodified sites. These characteristics
of the modifier are reflected on reaction kinetic parameters. In
consequence, kinetic approaches are expected to describe the
effects of the structural modification of the modifier on the
enantioselectivity as the changes in kinetic parameters.
Reaction kinetic approaches contribute to understanding
reaction dynamics and mechanisms, thereby providing
reasonable bases for improving the catalyst performance and
for industrial applications. However, kinetic approaches of
heterogeneous asymmetric catalyst systems are unfortunately
very limited.7,8,50–56 In addition, these kinetic studies were
usually too complicated with many undetermined parameters
or too simple to include the desired information. Recently, we
have developed a simple but informative reaction model and
formulation to evaluate the performance of cinchona alkaloid-
modified Pd/C for the enantioselective hydrogenations of
α-phenylcinnamic acids.57 We analyzed herein the
enantioselectivity of a Pd/C catalyst modified with a chiral
cinchonidine derivative by applying the kinetic model and
formulation for the hydrogenation of PCA as a representative
substrate to better understand the decisive reaction
parameter(s) for the catalyst performance. Cinchonidine (CD)
was substituted at the 2′ and/or 6′ positions of the quinoline
moiety to modify the features of the modifier.
enantioselectivity.1,6,38,40,41 Bürgi et al.48 have found
a
considerable decrease in enantiomeric excess (ee) by the
phenyl-substitution at the 2′-position of the quinoline ring of
10,11-dihydrocinchonidine (HCD) over a platinum catalyst
for the asymmetric hydrogenation of ethyl pyruvate in acetic
acid. We have previously demonstrated that the introduction
of OH at the 6′-position of the quinoline ring of CD does not
affect the performance of the modifier for the
enantioselective hydrogenation of (E)-2,3-diphenylpropenoic
acid (α-phenylcinnamic acid, PCA) over Pd/C.49 In the present
study, we systematically examined the effects of the
substitutions at the 2′- and/or 6′-positions of the quinoline
ring of CD (Scheme 2 and Table 1) on the enantioselectivity
of the asymmetric hydrogenation of PCA over cinchona-
modified Pd/C (Scheme 3).
It was shown in the present kinetic study that the
enantiomeric excess (ee) was decreased over the whole range
of modifier concentration by the substitution at the 2′
position of the quinoline moiety of CD with a methyl, vinyl,
n-butyl or phenyl group. It was found that the adsorption
strength of the modifier was decreased by the substitution,
while the adsorption strength of QN (CD substituted at the 6′
position by a methoxy group) was not affected further by
methyl-substitution at the 2′ position. The intrinsic
enantioselectivity at the modified sites was estimated by
It is conceivable with asymmetric hydrogenation over
chirally modified platinum and palladium metal catalysts that
the performance of the modifier is elucidated by the adsorption
strength of the modifier, intrinsic enantioselectivity at the
Scheme 2 Schematic illustration of the cinchonidine derivatives used
Scheme 3 Enantioselective hydrogenation of PCA over cinchonidine
as a modifier for the enantioselective hydrogenation of PCA over Pd/C.
derivative-modified Pd/C.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Catalysis Science & Technology
Paper
kinetic analysis. It is suggested that the decrease in the
intrinsic enantioselectivity by the substitution is most likely
correlated to the decrease in the interaction energy between
the substrate and modifier due to the change in the
adsorption geometry of the modifier.
together with those over CD–Pd/C49 for comparison. The
enantioselectivity of PCA sigmoidally increased as log CM
increased, regardless of the modifier, as reported previously
for cinchona alkaloid-modified palladium catalysts.49,57–59
Fig. 2 and 3 illustrate two sets of experimental results for
CD–vinyl and CD–Ph and CD-6′OH and QN–Me, respectively,
together with those for CD and QN49 for comparison. It is
likely that the vinyl-group of the quinoline ring is readily
hydrogenated to an ethyl group.
Results and discussion
The enantioselective hydrogenation of PCA was carried out
over Pd/C modified with a CD derivative by the use of a batch
reactor under atmospheric pressure of H2 as reported
previously.49,57,58 With the modifier used herein, the
S-enantiomer was the main product regardless of the
The maximum ee values (eemax) of the CD derivatives,
attained in the present study, are summarized in Table 2. It
is obvious that eemax is reduced by the substitution at the 2′-
and/or 6′-positions of the quinoline ring of CD, depending on
the substitution group in the order CD > CD–Me ∼ CD–vinyl
> CD–n-Bu > CD–Ph for the 2′-substitution and CD ∼ CD-6′
OH > QN > QN–Me for the 6′- and 2′-substitutions. The
observed decrease in the enantioselectivity by the phenyl-
substitution at the 2′-position of the quinoline ring of CD
agrees with the observation by Bürgi et al.48 for the
modifications. Fig.
1 shows the enantioselectivity, as
expressed by enantiomeric excess (ee%), and the reaction
rate, as normalized by the rate without the modifier but in
the presence of benzylamine (BA), as a function of log CM
(CM: modifier concentration, mol L−1) for the PCA
hydrogenation over CD–Me and CD–n-Bu-modified Pd/C
Fig. 1 a) Enantioselectivity, as expressed by ee%, as a function of the
modifier concentration (CM, mol L−1) in a logarithm scale for the
asymmetric hydrogenation of PCA over Pd/C modified with cinchona
derivatives. Modifiers: CD ( ), CD–Me ( ), and CD–n-Bu ( ). The
smooth line for each modifier is the curve fitted by using the values of
ie (%) and log β in Table 2. b) Normalized initial reaction rate as a
function of the modifier concentration (CM, mol L−1) in a logarithm
Fig. 2 a) Enantioselectivity, as expressed by ee%, as a function of the
modifier concentration (CM, mol L−1) in a logarithm scale for the
asymmetric hydrogenation of PCA over Pd/C modified with cinchona
derivatives. Modifiers: CD ( ), CD–vinyl ( ), and CD–Ph ( ). The
smooth line for each modifier is the curve fitted by using the values of
ie (%) and log β in Table 2. b) Normalized initial reaction rate as a
function of the modifier concentration (CM, mol L−1) in a logarithm
scale. The vertical line indicates the concentration C* (mol L−1), at
M
scale. The vertical line indicates the concentration C* (mol L−1). The
which a peak or hump is observed in the reaction rate for each
modifier. The raw data for CD are taken from our previous study49 and
are shown for comparison.
M
raw data for CD are taken from our previous study49 and are shown
for comparison.
This journal is © The Royal Society of Chemistry 2020
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
As for the reaction rates of the PCA hydrogenation, an
activity peak (ligand acceleration)1,7,57–59 clearly observed
with CD-modified Pd/C was decreased by the substitution
and, in some cases, became ambiguous into a hump or
shoulder peak. The reaction rates in Fig. 1b–3b are
normalized by the reaction rate (140 mmol g−1 h−1) observed
in the absence of the modifier (CM = 0) but in the presence of
auxiliary additive BA (0.5 mmol). It is concluded that both
the enantioselectivity and reaction rate of CD–Pd/C are
decreased by the substitution at the 2′- and/or 6′-positions of
the quinoline moiety of CD for the asymmetric
hydrogenation of PCA. We analyzed the significant
enantioselectivity changes observed in Fig. 1a–3a from the
kinetic points of view to obtain information on the kinetic
parameters controlling such changes.
The enantioselectivity of the asymmetric hydrogenation of
PCA in Fig. 1a–3a was analyzed by the use of the kinetic
modeling and formulations previously reported elsewhere.57
We have assumed in the kinetic formulations that the
modifier, substrate, hydrogen (dissociatively), and auxiliary
additive BA are reversibly and competitively adsorbed on the
palladium metal surface and that in the presence of BA (BA/
PCA molar ratio = 1), the rate determining step of the reaction
is the initial hydrogen atom addition to the adsorbed
substrate at both modified and unmodified sites. The change
of the rate determining step of the reaction from the
desorption step of the products to a surface reaction was
proposed by Nitta27,60 and recently verified by Meemken
et al.25 using in situ ATR-IR. We have previously shown
syn-addition of hydrogen to PCA over CD–Pd/C.61 It is also
assumed that enantioselective hydrogenation proceeds at the
Fig. 3 a) Enantioselectivity, as expressed by ee%, as a function of the
modifier concentration (CM, mol L−1) in a logarithm scale for the
asymmetric hydrogenation of PCA over Pd/C modified with cinchona
derivatives. Modifiers: CD ( ), CD-6′OH ( ), QN ( ), and QN–Me ( ).
The smooth line for each modifier is the curve fitted by using the values
of ie (%) and log β in Table 2. b) Normalized initial reaction rate as a
function of the modifier concentration (CM, mol L−1) in a logarithm
modified sites where
a surface substrate–modifier 1 : 1
scale. The vertical line indicates the concentration C* (mol L−1). The
M
complex is formed.20,27,30,38,47 Vargas and coworkers16 have
suggested by theoretical calculations with CD-modified
platinum that the substrate likely approaches a chiral site
raw data for CD, CD-6′OH and QN are taken from our previous study49
and are shown for comparison.
either from solution or from
a physisorbed state. In
Table 2 Kinetic analysis of the enantioselective hydrogenation of PCA
over Pd/C modified with a cinchonidine derivative
accordance with their suggestions, the formation of the
complex intermediate has been formulated by the adsorption
of the substrate onto the preadsorbed modifier, rendering the
rate equations to be very simplified.57 Under these
assumptions and a Langmuir–Hinshelwood formalism, we
have obtained eqn (1) to correlate the observed ee to the
concentration of the modifier (CM) in the reaction solution.57
The detailed derivations of eqn (1) are presented in the ESI.†
log C* d
Modifier
eemaxa (%)
i
eb (%)
log βc
m
Δlog βe
CD
83
75
73
64
51
56
23
80
87
80
80
70
60
70
35
85
−4.1
−3.5
−3.3
−3.3
−3.2
−3.3sh
−2.1
−2.2
−3.6
0
CD–Me
CD–vinyl
CD–n-Bu
CD–Ph
QN
−3.8
0.08
0.08
0.19
0.60
0.36
0.64
0.0
−3.8
−3.6
−3.3
−2.3
QN–Me
CD-6′OH
5
−2.1 0.1
−4.2
ee = ierm/(rm + ru)
= ie/(1 + β/CM)
(1)
a
b
Maximum ee (%) observed for the modifier.
Intrinsic
enantioselectivity ie (%) defined by eqn (1) and (2). Kinetic
c
where β = (ku/km)(KS/KM)(1 + KMSCS)/KMS. rm and km denote
the reaction rate and rate constant at the modified sites,
respectively, while ru and ku are those at the unmodified
sites. K and C represent the equilibrium adsorption constant
and concentration, respectively. The subscripts M, S and MS
are the modifier, substrate and modifier–substrate complex,
respectively. It is noted that KMS is elucidated by the
interactions between the substrate and surface modifier as
parameter β = (ku/km)(KS/KM)(1 + KMSCS)/KMS in a logarithm scale.
d
Concentration of the modifier (C* , mol L−1) giving a reaction rate
M
e
peak at the modified sites. Log β relative to that of cinchonidine
after the correction for the KM of the modifier.
asymmetric hydrogenation of ethyl pyruvate over a platinum
catalyst modified with HCD.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Catalysis Science & Technology
Paper
well as by the interactions of the substrate with surface
palladium atoms. The intrinsic enantioselectivity ie (%) in
eqn (1) is defined by the ee of the hydrogenation at the
modified sites.
are obtained, validating the kinetic modeling and
formulations used herein. The values of ie and log β thus
estimated are listed in Table 2. The simulation curves
generated by the use of the values of ie and log β in Table 2
are shown in Fig. 1a–3a and are compared with the observed
ee for the modifiers. Satisfactory fittings are obtained,
confirming the appropriateness of these values to describe
the enantioselectivity as a function of log CM.
ie = 100 × |[S]m − [R]m|/([S]m + [R]m)
(2)
where [S]m and [R]m represent the amounts of the S and R
enantiomers formed via pro-S and pro-R intermediate
complexes at the modified sites, respectively. eqn (1) is
transformed into eqn (3).
It is revealed in Table 2 that the intrinsic enantioselectivity
of CD is decreased by the substitution at the 2′- and/or 6′-
positions of the quinoline ring; ie decreases in the order CD
> CD–Me ∼ CD–vinyl > CD–n-Bu > CD–Ph for the 2′-
substitution and CD ∼ CD-6′OH > QN ≫ QN–Me for the 6′-
and 2′-substitutions. This order is roughly parallel to that of
eemax, although ie is always larger than eemax by definition. It
is noted that the concentration to attain eemax depends on
the modifier. Both eemax and the modifier concentration at
eemax for the asymmetric hydrogenation of the substrate are
defined by kinetic factors such as the adsorption strength of
the modifier relative to that of the substrate, reaction rate
constant of the hydrogenation at the modified sites relative
to that at unmodified racemic sites, interaction strength
between the modifier and substrate and even the solubility of
the modifier as well as the intrinsic enantioselectivity of the
modifier–substrate interaction complexes. With the present
series of the modifiers, the parallel correlation between
eemax and ie leads us to conclude that the intrinsic
enantioselectivity of the modifier is a dominant parameter to
determine the observed maximum enantioselectivity.
log[ee/(ie − ee)] = log CM − log β
(3)
eqn (3) predicts a linear correlation with a slope of unity
between log[ee/(ie − ee)] and log CM. The value of ie cannot be
directly determined by experiments in spite of its extreme
importance
in
understanding
the
enantioselective
hydrogenation mechanism. However, we can use eqn (3) to
estimate it from the dependency of ee on the concentration
of the modifier, CM, by choosing the appropriate value of ie
so as to provide a theoretical linear correlation with a slope
of unity over the widest CM region.57 In addition, we can
obtain the value of log β from eqn (3). We applied eqn (3) to
the analysis of the observed enantioselectivity in Fig. 1a–3a
for the asymmetric hydrogenation of PCA over CD derivative-
modified Pd/C to estimate the intrinsic enantioselectivity ie
and the kinetic parameter β in a logarithm scale.
Fig. 4 illustrates log ee/(ie − ee) against log CM for some
representative modifiers after choosing the appropriate value
of ie. With the other modifiers used here, the best plots are
shown in the ESI† (Fig. S1). Satisfactory linear correlations
The performance of the modifier in the asymmetric CO
and CC hydrogenations on platinum and palladium,
respectively, is sometimes evaluated at a fixed CM. Fig. 1a–3a,
Fig. 4 Correlation between log[ee/(ie − ee)] and log CM for CD, CD–Me, CD–n-Bu, and CD–vinyl. The linear line in the plots is a hypothetical line
with a slope of unity according to eqn (3). The values of ie and log β thus obtained are listed in Table 2.
This journal is © The Royal Society of Chemistry 2020
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
however, show that the modifier concentration to achieve
eemax considerably depends on the modifier. The
enantioselectivity is determined by the kinetic parameters
such as km/ku, KM/KS, and KMS as well as the intrinsic
enantioselectivity, as presented by eqn (1). Thus, it is
necessary to examine the enantioselectivity over as wide a
concentration range of the modifier as possible for evaluating
the performance of the modifier on the chirally modified
novel metal catalysts.
is considered that the weakening of the adsorption strength
of CD by the substitution at the 2′-position of the quinoline
moiety is caused by tilting of the quinoline ring around the
short axis due to steric effects from the flat lying adsorption
mode via π-electrons on the palladium metal surface. The
methyl-, vinyl-, n-butyl- and phenyl-groups exhibit almost the
same effects. Tilting of the quinoline ring around the long
axis due to the methoxy-substitution at the 6′-position
exhibits much stronger effects on the adsorption strength.
Huck et al.40 have shown from nonlinear effects observed
with a modifier mixture of CD and QN or CN and QD over
Pd/TiO2 that QN/QD is adsorbed more weakly than CD/CN
due to the tilting of the quinoline ring relative to the
palladium metal surface by the introduction of the methoxy-
group at the 6′-position. It is proposed herein that the tilting
of the quinoline ring of the CD-derivative modifier around its
short and/or long axes results in the change in the
conformation of the chiral pocket constituted by the
adsorbed modifier and, thereby, in the alteration of the
intrinsic enantioselectivity. Kubota and Zaera62 have
demonstrated by in situ RA-IR that the flat-lying geometry of
the modifier is responsible for the enantioselective
hydrogenation of ethyl pyruvate over an HCD-modified
platinum catalyst. Thus, as far as a series of the CD-
derivatives is concerned, the adsorption strength of the
modifier may be related to the intrinsic enantioselectivity,
though qualitatively and indirectly.
The intrinsic enantioselectivity ie of CD is significantly
reduced by the substitution of the quinoline ring at the 2′-
and/or 6′-positions, as presented in Table 2. It is presumed
that the change in the ie value is caused by the alteration in
the adsorption geometry of the modifier by the substitution,
which results in a conformational change in the chiral pocket
formed by the adsorbed modifier on the palladium metal
surface and thereby the change in the geometry of the
modifier–substrate interaction complexes. From the kinetic
points of view, these changes are reflected on the adsorption
strength of the modifier and the interaction energy between
the substrate and modifier. It is noted that these kinetic
factors are included in the parameter β.
The relative adsorption strength of the modifier can be
estimated from the modifier concentration, C* (mol L−1), at
M
which the reaction rate shows a peak or hump (ligand
acceleration).57 By solving drm/dCM = 0, eqn (4) can be
derived (ESI†).
To seek the other kinetic factors more directly correlated
to the intrinsic enantioselectivity, we tried to separate the
effects of the adsorption strength of the modifier from the
parameter β. Taking into consideration the weak dependency
of ee on the substrate concentration, CS,57 we can estimate
more relevant parameters, as follows.
C* ¼ α=K M
(4)
M
where α is a constant so long as the same substrate is
concerned. KM is inversely proportional to C* . From Fig. 1b–
3b, the values of log C* are elucidated and listed in Table 2. It
M
M
is evident that the adsorption strength of the modifier, as
expressed by KM, is reduced by the substitution and decreases
in the order: CD > CD–Me ∼ CD–vinyl ∼ CD–Ph ∼ CD–n-Bu
for the 2′-substitution and CD-6′OH ∼ CD ≫ QN ∼ QN–Me for
the 6′- and 2′-substitutions. The decrease in the adsorption
strength of CD by the methoxy-substitution at the 6′-position
agrees with the observations by Huck et al.39,40 The
substitution of the quinoline ring of CD at the 2′-position by
the methyl-, vinyl-, phenyl-, or n-butyl group induces a 1.6–2
log β = log(kuKS) − log KM − log km + log[(1 + KMSCS)/KMS
∼ log(kuKS) − log KM – log(kmKMS
]
)
(5)
Combining eqn (5) with eqn (4),
log β ∼ logðkuKS=αÞ − logðkmKMSÞ þ log C*
(6)
M
Since the term kuKS/α is common for the same substrate
concerned, eqn (6) predicts a linear correlation with a slope of
À
Á
times decrease of the KM value Δlog C* ¼ 0:2–0:3 . The
M
unity between log β and log C* for the substrates exhibiting
methoxy-substitution at the 6′-position of CD to form QN
shows much stronger effects, a 20–25 times decrease in KM. On
the other hand, the hydroxyl-group at the 6′-position exhibits
no notable or even beneficial effects on the adsorption strength
of CD (CD-6′OH). The substitution at the 2′-position of QN by a
methyl-group does not cause any further decrease in the
M
the same kmKMS. Fig. 5 presents log β against log C* together
M
with a hypothetical linear line having a slope of unity,
expected for a modifier having the same kmKMS as that of CD
but with varying KM. The difference in log β for each modifier,
Δlog β, from the linear line at its value of log C* shows the
M
adsorption strength within the accuracy of log C* ( 0.1).
difference in log(kmKMS) from that of CD. Thus, Fig. 5 can be
used to estimate log(kmKMS) of a modifier, relative to that of
CD, from log β after the correction of the effect of the
adsorption strength of the modifier. The values of Δlog β thus
obtained from Fig. 5 are summarized in Table 2.
M
There is a vague tendency as shown in Table 2 that the
intrinsic enantioselectivity decreases as the adsorption
strength of the modifier decreases. It is generally accepted
that CD is adsorbed on a geometry parallel to a platinum or
palladium metal surface via π-electrons of the quinoline ring
of CD in the energetically most stable adsorption mode.1–10 It
Fig.
6
shows
a
correlation between the intrinsic
enantioselectivity and Δlog β. The intrinsic enantioselectivity
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Catalysis Science & Technology
Paper
similar structure. Actually, according to the calculations by
Schmidt et al.,20 ΔE for CD (2.2 kcal mol−1) is larger than that
for CN (1.0 kcal mol−1) and the magnitude of interaction
energy averaged over the pro-R and pro-S complexes for CD
(30.0 kcal) is larger than that for CN (28.9 kcal mol−1). The
simple correlation in Fig. 6 prompts us to speculate that
stronger interactions between the substrate and surface
modifier result in a larger energy difference between the pro-
R and pro-S intermediate surface complexes and, thereby, are
beneficial for achieving a higher intrinsic selectivity in the
asymmetric hydrogenation of PCA over the present series of
CD derivative-modified Pd/C. As discussed above, it is
considered that the substitutions at the 2′- and 6′-positions of
the quinoline ring of CD result in the tilting of the modifier
molecule around the short and long axes of the ring,
respectively, relative to the palladium metal surface,
accompanying the decrease in the adsorption energy (as
expressed by KM). It is argued that the modification of the
adsorption geometry of cinchona molecules alters the
conformation of the chiral pocket formed by the stereogenic
quinuclidine moiety and C9–OH on the linker, and thereby
the intrinsic enantioselectivity. Schmidt et al.20 have, on the
other hand, pointed out from the theoretical calculations on
the asymmetric hydrogenation on 9-epi-CD-modified
platinum that the concept of the thermodynamic control is
not generally applicable to predict the stereochemical
outcomes. It is actually considered that km and KMS are in
some cases strongly coupled with each other7 (for instance,
compensation effects64,65); an increase of KMS may
accompany a decrease of km due to enhanced activation
energy and vice versa.
Fig. 5 log β against log CM* (mol L−1) for the asymmetric hydrogenation
of PCA over CD derivative-modified Pd/C. The dotted line through the
points for CD is a hypothetical line with a slope of unity, expected for
a modifier exhibiting the same kmKMS value as that of CD but with
varying KM, according to eqn (6).
ie decreases as Δlog β increases or kmKMS decreases regardless
of the substitutions at the 2′- and 6′-positions. A satisfactory
and simple relationship in Fig. 6 implies that ie is defined by
kmKMS from the kinetic points of view. Unfortunately, km and
KMS are not decoupled in the present study. In our previous
study on the enantioselective hydrogenation of ketones over
a tartaric acid-modified RANEY®-type Ni catalyst, we suggested
by combining kinetic analysis and DFT calculations that the
enantioselectivity increased as the interaction energy between
the modifier–substrate or KMS in the present description
increased.63 Theoretical calculations by Schmidt et al.20 have
shown with the enantioselective hydrogenation of
trifluoroacetophenone over CD- and CN-modified platinum
catalysts that the interaction energy difference, ΔE, between
the pro-S and pro-R substrate–modifier surface complexes is
a controlling factor over the (intrinsic) enantioselectivity
(thermodynamic control). It is conceivable that the energy
difference between the pro-S and pro-R modifier–substrate
interaction complexes increases as the magnitude of KMS
increases for a series of interaction complexes possessing a
Rodríguez-García et al.26 have demonstrated with CN–Pd/
TiO2 by in situ ATR-IR combined with UV-vis that the
adsorption geometry of CN depends strongly on the surface
coverage of the modifier. Combining the ATR-IR results with
the selectivity and activity of the hydrogenation of 4-methoxy-
6-methyl-2-pyrone on CN–Pd/TiO2, they suggested that less
stable CN species with a tilted aromatic ring anchor provided
a stronger enantioselective control. Meemken et al.25 have
shown by a time resolved ATR-IR spectroscopic study that
two types of adsorbed CD are detected on Pd/TiO2, π-bonded
CD and N-lone pair bonded CD, in the absence of BA, but
that the coexistence of CD and BA leads to a dynamic
transformation of N-lone pair bonded CD to π-bonded CD. In
consequence, we are inclined to tentatively suggest from the
present kinetic study that in the presence of BA, a (almost)
flat-lying adsorption geometry of CD on the palladium metal
surface is beneficial for achieving a high enantioselectivity
and that the introduction of the substituent(s) at the
quinoline ring alters the geometry to a tilted one due to steric
hindrance, accompanying decreases in the adsorption
strength of the modifier, the substrate–modifier interaction
energy, and thereby the intrinsic enantioselectivity. Further
studies on the microkinetics coupled with molecular
modeling and theoretical calculations are needed for better
understanding of the controlling parameters of the intrinsic
Fig. 6 Correlation between the intrinsic enantioselectivity ie (%) and
Δlog β.
This journal is © The Royal Society of Chemistry 2020
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
enantioselectivity in the asymmetric hydrogenations on
chirally modified palladium and platinum metal catalysts.
literature.66–68 The yields of CD–n-Bu and CD–Ph were 28%
and 35%, respectively. QN–Me was synthesized and
characterized similarly (yield, 18%).66,67 CD–vinyl was
Experimental
Reaction procedures
synthesized by a reaction of CD with vinyl-MgBr and
characterized by NMR and MS (yield, 14%) according to the
synthesis procedures and characterization techniques by
Yardley et al.66 The chemical data for CD–vinyl are presented
in the ESI.† CD-6′OH was prepared by demethylation of QN
and characterized accordingly (yield, 65%).49,69
All chemicals were purchased from commercial sources and
purified via distillation when necessary. The enantioselective
hydrogenation reaction of (E)-2,3-diphenylpropenoic acid (α-
phenylcinnamic acid, PCA) was carried out over chiral
cinchonidine derivative-modified Pd/C.30,49,57,58 In brief, a
stirred suspension of 5% Pd/C (43 mg as a 50% content, wet
form STD-type supplied by N.E. Chemcat) and 5 mL of 2.5%
H2O-containing dioxane (wet dioxane) as a solvent was
heated under 0.1 MPa of hydrogen at 353 K for 30 min. After
cooling to 296 K, a solution of the modifier (5.1 × 10−6 − 41 ×
10−3 mmol) in the solvent (1 mL) was added. After 30 min,
0.5 mmol PCA in the solvent (4 mL) and then benzylamine
(BA, 0.5 mmol) was added, while vigorously stirring (1200
rpm). The reaction temperature and pressure were 296 K and
0.1 MPa of hydrogen, respectively. The initial reaction rate
was calculated from the rate of hydrogen consumption at
25% conversion. The hydrogen consumption continued for
1–3 h before completion. After additional 2 h, 2 M HCl aq. (1
mL) was added to the solution and the mixture was filtered
to remove the catalyst. The filtrate was extracted with ethyl
acetate (2 mL) and washed with water (2 mL). The extract was
analyzed by HPLC with a chiral column (Daicel OJ-3). The
enantioselectivity is expressed as the enantiomeric excess (ee)
of the S or R enantiomer,
Conclusions
Asymmetric hydrogenations of α,β-unsaturated carboxylic
acids over cinchona alkaloid-modified palladium metal
heterogeneous catalysts have received considerable attention
because of their scientific importance in molecular
recognition catalysis and feasibility of industrial applications.
In the present study, the effect of the quinoline ring
modification of cinchonidine (CD) on the enantioselective
hydrogenation of α-phenylcinnamic acid (PCA) was
systematically analyzed from the kinetic points of view. The
substitution at the 2′- and/or 6′-positions of the quinoline
moiety by methyl, vinyl, n-butyl or phenyl decreased the
observed enantioselectivity over the whole range of the
modifier concentration. The present kinetic analysis of the
enantioselectivity allowed us to estimate the intrinsic
enantioselectivity at the modified sites and the adsorption
strength of the modifier. It is revealed that the substitution
at the quinoline ring of CD reduces the adsorption strength
of the modifier, probably due to the tilting of the quinoline
ring from a flat-lying adsorption geometry, accompanying a
decrease in the intrinsic enantioselectivity of CD. The
ee = 100 × |[S] − [R]|/([S] + [R]),
intrinsic enantioselectivity is correlated to
a
kinetic
where [S] and [R] represent the concentrations of the S and R
enantiomers, respectively. The experimental reproducibility
of ee was usually 2% under the present reaction conditions.
parameter, most likely, to the modifier–substrate interaction
energy. It is proposed that the tilting of the quinoline ring of
the CD-derivative modifier around its short and/or long axes
results in the change in the conformation of the chiral pocket
constituted by the adsorbed modifier and, thereby, in the
alteration of the intrinsic enantioselectivity. It is suggested
that the present kinetic model and formulation can be
generally applied to describe the catalytic behaviors and
performance of cinchona-modified palladium and platinum
catalysts for the asymmetric hydrogenations because of their
identical reaction schemes from the kinetic viewpoints:
reversible adsorption of the substrate and modifier,
Synthesis of cinchonidine derivatives
According to the reported procedures,66–68 the substitution of
the quinoline ring of cinchonidine (CD) was carried out by
the use of alkyl- or phenyl-Li to yield 2′-substituted
derivatives, CD–Me, CD–n-butyl or CD–Ph. In brief, for
instance, CD was treated with MeLi (3 eq.) at 195 K in THF
under N2 atmosphere. The solution turned yellow in color
after 12 h. Then, the reaction was quenched by adding an
aqueous solution of NaHCO3, followed by extraction of the
product with ethyl ether. After confirming the complete loss
of 2′-H of CD by NMR, the product was treated with I2 in THF
for 2 h. The reaction mixture was extracted with ethyl ether
in the presence of a Na2SO3 aqueous solution, followed by
purification with SiO2 column chromatography (eluent:
MeOH/CHCl3) and with recrystallization from ethyl acetate to
yield CD–Me in 14% yield as a white solid. The formations of
CD–Me, CD–n-Bu, and CD–Ph were confirmed by 1H- and
13C-NMR, IR and MS techniques.66–68 The NMR data of these
compounds agreed very well with those reported in
formation of
a
surface substrate–modifier interaction
complex, and subsequent hydrogen addition, followed by
desorption of products.
Conflicts of interest
There are no conflicts to declare.
Notes and references
1 F. Meemken and A. Baiker, Chem. Rev., 2017, 117,
11522–11569.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Catalysis Science & Technology
Paper
2 A. Baiker, Chem. Soc. Rev., 2015, 44, 7449–7464.
3 E. Zhan, C. Chen, Y. Li and W. Shen, Catal. Sci. Technol.,
2015, 5, 650–659.
4 M. Bartók, Chem. Rev., 2010, 110, 1663–1705.
5 F. Zaera, J. Phys. Chem. C, 2008, 112, 16196–16203.
6 T. Mallat, E. Orglmeister and A. Baiker, Chem. Rev.,
2007, 107, 4863–4890.
7 H.-U. Blaser and M. Studer, Acc. Chem. Res., 2007, 40,
1348–1356.
8 D. Y. Murzin, P. Mäki-Arvela, E. Toukoniitty and T. Salmi,
Catal. Rev.: Sci. Eng., 2005, 47, 175–256.
34 S. Diezi, M. Hess, E. Olgmeister, T. Mallat and A. Baiker,
Catal. Lett., 2005, 102, 121–125.
35 E. Orgmeister, T. Mallat and A. Baiker, Adv. Synth. Catal.,
2005, 347, 78–86.
36 C. Mondelli, A. Vargas, G. Santarossa and A. Baiker, J. Phys.
Chem. C, 2009, 113, 15246–15259.
37 M. C. Holland, F. Meemken, A. Baiker and R. Gilmour,
J. Mol. Catal. A: Chem., 2015, 396, 335–345.
38 Y. Nitta and A. Shibata, Chem. Lett., 1998, 27, 161–162.
39 W. R. Huck, T. Bürgi, T. Mallat and A. Baiker, J. Catal.,
2003, 219, 41–51.
9 T. Bürgi and A. Baiker, Acc. Chem. Res., 2004, 37, 909–917.
10 M. Studer, H.-U. Blaser and C. Exner, Adv. Synth. Catal.,
2003, 345, 45–65.
11 A. Vargas, T. Bürgi and A. Baiker, J. Catal., 2004, 226, 69–82.
12 N. Bonalumi, A. Vargas, D. Ferri, T. Bürgi, T. Mallat and A.
Baiker, J. Am. Chem. Soc., 2005, 127, 8467–8477.
13 Z. Ma and F. Zaera, J. Am. Chem. Soc., 2006, 128,
16414–16415.
14 S. Lavoie, M.-A. Laliberte, I. Temperano and P. McBreen,
J. Am. Chem. Soc., 2006, 128, 7588–7593.
15 N. Bonalumi, A. Vargas, D. Ferri and A. Baiker, J. Phys.
Chem. C, 2007, 111, 9349–9358.
16 A. Vargas, G. Santarossa, M. Iannuzzi and A. Baiker, J. Phys.
Chem. C, 2008, 112, 10200–10208.
17 K. R. Hahn, A. P. Seitsonen and A. Baiker, Phys. Chem. Chem.
Phys., 2015, 17, 27615–27629.
18 F. Meemken, N. Maeda, K. Hungerbühler and A. Baiker,
Angew. Chem., Int. Ed., 2012, 51, 8212–8216.
40 W. R. Huck, T. Mallat and A. Baiker, Catal. Lett., 2003, 87,
241–247.
41 W. R. Huck, T. Bürgi, T. Mallat and A. Baiker, J. Catal.,
2003, 216, 276–287.
42 S. Cserényi, K. Felföldi, K. Balázsik, G. Szöllösi, I. Bucsi
and M. Bartók, J. Mol. Catal. A: Chem., 2006, 247,
108–115.
43 N. Bonalumi, A. Vargas, D. Ferri and A. Baiker, J. Phys.
Chem. C, 2007, 111, 9349–9358.
44 A. Vargas, D. Ferri, N. Bonalumi, T. Mallat and A. Baiker,
Angew. Chem., Int. Ed., 2007, 46, 3905–3908.
45 K. Balazsik, I. Bucsi, S. Cserényi, G. Szöllösi and M. Bartók,
J. Mol. Catal. A: Chem., 2008, 280, 87–95.
46 G. Szöllösi, B. Hermán, F. Fülöp and M. Bartók, J. Catal.,
2010, 276, 259–267.
47 G. Szöllösi, I. Busygin, B. Hermán, R. Leino, I. Bucsi, D. Y.
Murzin, F. Fülöp and M. Bartók, ACS Catal., 2011, 1,
1316–1326.
19 D. Ferri, T. Bürgi and A. Baiker, J. Catal., 2002, 210, 160–170.
20 E. Schmidt, C. Bucher, G. Santarossa, T. Mallat, R. Gilmour
and A. Baiker, J. Catal., 2012, 289, 238–248.
21 X. Zuo, H. Liu and M. Liu, Tetrahedron Lett., 1998, 39,
1941–1944.
22 B. Török, K. Felföldi, G. Szakonyi, K. Balazsik and M. Bartók,
Catal. Lett., 1998, 52, 81–84.
23 B. Török, K. Balázsik, M. Török, G. Szöllösi and M. Bartók,
Ultrason. Sonochem., 2000, 7, 151–155.
48 T. Bürgi, Z. Zhou, N. Künzle, T. Mallat and A. Baiker,
J. Catal., 1999, 183, 405–408.
49 M. Nakatsuji, T. Misaki, Y. Okamoto and T. Sugimura, Bull.
Chem. Soc. Jpn., 2016, 89, 1187–1191.
50 M. Garland and H.-U. Blaser, J. Am. Chem. Soc., 1990, 112,
7048–7050.
51 H.-U. Blaser, M. Garland and H. P. Jalett, J. Catal., 1993, 144,
569–578.
52 H.-U. Blaser, H.-P. Jalett, M. Garland, M. Studer, H. Thies
and A. Wirth-Tijani, J. Catal., 1998, 173, 282–294.
24 M. Sutyinszki, K. Szöri, K. Felföldi and M. Bartók, Catal.
Commun., 2002, 3, 125–127.
˘
53 E. Toukoniitty, B. Ševeíková, P. Mäki-Arvela, J. Wärmá, T.
25 F. Meemken, N. Maeda, K. Hungerbühler and A. Baiker, ACS
Catal., 2012, 2, 464–467.
26 L. Rodríguez-García, K. Hungerbühler, A. Baiker and F.
Meemken, ACS Catal., 2017, 7, 3799–3809.
Salmi and D. Y. Murzin, J. Catal., 2003, 213, 7–16.
54 D. Y. Murzin and E. Toukoniitty, Catal. Lett., 2006, 109,
125–131.
55 D. Y. Murzin, J. Mol. Catal. A: Chem., 2008, 289, 91–94.
56 D. Y. Murzin and G. Szöllösi, React. Kinet., Mech. Catal.,
2011, 103, 1–9.
57 B. Kim, M. Nakatsuji, T. Mameda, T. Kubota, M. Fujita, T.
Sugimura and Y. Okamoto, Bull. Chem. Soc. Jpn., 2020, 93,
163–175.
58 T. Mameda, K. Nakai, T. Misaki, Y. Okamoto and T.
Sugimura, Catal. Today, 2015, 245, 129–133.
59 H. Ogawa, T. Misaki, Y. Okamoto and T. Sugimura, Chem.
Lett., 2013, 42, 813–815.
27 Y. Nitta, Top. Catal., 2000, 13, 179–185.
28 W. R. Huck, T. Mallat and A. Baiker, New J. Chem., 2002, 26, 6–8.
29 G. Szöllösi, B. Hermán, K. Felföldi, F. Fulöp and M. Bartók,
Adv. Synth. Catal., 2008, 350, 2804–2814.
30 T. Sugimura, T. Uchida, J. Watanabe, T. Kubota, Y. Okamoto,
T. Misaki and T. Okuyama, J. Catal., 2009, 262, 57–64.
31 M. Schürch, T. Heinz, R. Aeschimann, T. Mallat, A. Pfaltz
and A. Baiker, J. Catal., 1998, 173, 187–195.
32 H.-U. Blaser, H. P. Jalett, W. Lottenbach and M. Studer,
J. Am. Chem. Soc., 2000, 122, 199–203.
33 C. Exner, A. Pfaltz, M. Studer and H.-U. Blaser, Adv. Synth.
Catal., 2003, 345, 1253–1260.
60 Y. Nitta, Chem. Lett., 1999, 28, 635–636.
61 T. Sugimura, S. Tomatsuri, M. Fujita and Y. Okamoto, Bull.
Chem. Soc. Jpn., 2019, 92, 1737–1742.
This journal is © The Royal Society of Chemistry 2020
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
62 J. Kubota and F. Zaera, J. Am. Chem. Soc., 2001, 123,
11115–11116.
66 J. P. Yardley, R. E. Bright, L. Rane, R. W. Rees, P. B. Russell
and H. Smith, J. Med. Chem., 1971, 14, 62–65.
63 A. A. Choliq, E. Murakami, S. Yamamoto, T. Misaki, M.
Fujita, Y. Okamoto and T. Sugimura, Bull. Chem. Soc. Jpn.,
2018, 91, 1325–1332.
64 A. K. Galway, Adv. Catal., 1977, 26, 247–322.
65 T. Osawa, M. Wakasugi, T. Kizawa, V. Borovkov and Y. Inoue,
Mol. Catal., 2018, 449, 131–136.
67 D. Worgull, L. Öhler, J. P. Strache, T. Friedrichs and P.
Ullrich, Eur. J. Org. Chem., 2017, 2017, 6077–6080.
68 L. Hintermann, M. Schmitz and U. Englert, Angew. Chem.,
Int. Ed., 2007, 46, 5164–5167.
69 K. Muto, J. Yamaguchi and K. Itami, J. Am. Chem. Soc.,
2012, 134, 169–172.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2020