Enantio-differentiating hydrogenation of alkyl 3-oxobutanoates over tartaric acid-modified Ni catalyst: Enthalpy-entropy compensation effect as a tool for elucidating mechanistic features

Article abstract of DOI:10.1016/j.mcat.2018.02.023

The enantio-differentiating hydrogenations of a series of alkyl 3-oxobutanoates were carried out at the temperatures ranging from 333 to 393 K over the (R,R)-tartaric acid-modified Ni catalyst prepared from commercially available Ni powder to achieve high enantiomeric excesses of 91-94%. It was demonstrated that the enantio-selectivity was not a simple function of the reaction temperature, being enhanced in the low temperature region to reach a maximum at 363–373 K and then decreased at higher temperatures. Nevertheless, all the differential enthalpies and entropies of activation calculated from the enantiomer ratios in the low and high temperature regions compensated with each other, indicating the same enantio-differentiation mechanism over the entire temperature range. A plausible enantio-differentiation mechanism explaining the effects of hydrogenation temperature on the enantio-selectivity is proposed.

Full text of DOI:10.1016/j.mcat.2018.02.023

MolecularCatalysis449(2018)131–136
Contents lists available at ScienceDirect
Molecular Catalysis
journal homepage: www.elsevier.com/locate/mcat
Enantio-differentiating hydrogenation of alkyl 3-oxobutanoates over tartaric
acid-modified Ni catalyst: Enthalpy-entropy compensation effect as a tool
for elucidating mechanistic features
Tsutomu Osawaa,⁎, Masahiro Wakasugia, Tomoko Kizawab, Victor Borovkovc,d, Yoshihisa Inouee
a
Graduate School of Science and Engineering for Research, University of Toyama, 3190 Gofuku, Toyama 930-8555, Japan
Metek Co. Ltd, Warada-cho, Kamitoba Minami-ku, Kyoto 601-8133, Japan
Department of Chemistry and Biotechnology, School of Science Tallinn University of Technology Akadeemia tee 15, 12618 Tallinn, Estonia
College of Chemistry and Materials Science South-central University for Nationalities 182# Minzu RD, Hongshan District, Wuhan City, Hubei province, 430074, China
b
c
d
e
Department of Applied Chemistry, Osaka University, Yamada-oka, Suita 565-0871, Japan
A R T I C L E I N F O
A B S T R A C T
Keywords:
Enantio-differentiating hydrogenation
Alkyl 3-oxobutanoate
Tartaric acid-modified nickel powder
Enthalpy-entropy compensation
The enantio-differentiating hydrogenations of a series of alkyl 3-oxobutanoates were carried out at the tem-
peratures ranging from 333 to 393 K over the (R,R)-tartaric acid-modified Ni catalyst prepared from commer-
cially available Ni powder to achieve high enantiomeric excesses of 91-94%. It was demonstrated that the
enantio-selectivity was not a simple function of the reaction temperature, being enhanced in the low tem-
perature region to reach a maximum at 363–373 K and then decreased at higher temperatures. Nevertheless, all
the differential enthalpies and entropies of activation calculated from the enantiomer ratios in the low and high
temperature regions compensated with each other, indicating the same enantio-differentiation mechanism over
the entire temperature range. A plausible enantio-differentiation mechanism explaining the effects of hydro-
genation temperature on the enantio-selectivity is proposed.
1. Introduction
immediately after the preparation of Raney Ni from Al-Ni alloy;
otherwise the nickel catalyst undergoes fast surface oxidation leading to
The production of optically active compounds is a key issue in
synthetic organic chemistry, especially for the synthesis of pharma-
ceuticals, flavor and aroma chemicals, and agrochemicals. A variety of
chiral catalysts, both homogeneous (organometallic complexes, en-
zymes, and organocatalysts) [1–5] and heterogeneous (modified solid
catalysts, heterogenized organometallic complexes, enzymes, and
metal-organic frameworks) [6–17], have hitherto been employed for
this purpose. Among them, chirally modified solid catalysts are parti-
cularly promising and advantageous for the large-scale production in
industry. This is because they are readily prepared at low cost, separ-
able from the reaction mixture, and reusable without any additional
treatments, all of which are attractive also from the ecological and
economical viewpoints [18].
Tartaric acid (TA)-modified Ni catalyst is one of the most successful
chiral solid catalysts for the enantio-differentiating hydrogenation of β-
keto-esters (KEs) [19–21] and 2-alkanones [22–24]. This catalyst is
usually prepared by adsorbing TA and NaBr on surface-activated Ni
catalyst, such as Raney Ni [25–27]. However, there is a serious draw-
back of this catalyst that the chiral modification has to be carried out
poor reproducibility [28]. Recently, we found that the TA-modified Ni
catalyst prepared from commercial Ni powder without any pretreat-
ment also functions as an equally effective and highly robust catalyst
for the enantio-differentiating hydrogenation of KEs to give the corre-
sponding hydroxyesters (HEs) with the enantio-selectivities higher than
those obtained with conventional TA-modified Raney Ni [29]. The
catalyst can be prepared simply by soaking Ni powder in an aqueous
solution of TA and NaBr without potentially hazardous pre-activation
by hydrogen gas, and is therefore more suitable for industrial use from
the safety aspect.
In this study, to optimize the reaction conditions for better enantio-
selectivity and also to elucidate the factors and mechanism controlling
the enantio-differentiating hydrogenation, we expanded the range of
substrates (Scheme 1) and closely examined the effects of hydrogena-
tion temperature on the enantio-selectivity to reveal that the en-
antiomeric excess (e.e.) is not a simple function of the temperature to
reach a maximum in the middle of the temperature range examined.
The enantiomer ratio (e.r.) obtained at each temperature was subjected
to the Eyring analysis to afford the differential activation parameters,
⁎
Corresponding author.
E-mail address: osawa@sci.u-toyama.ac.jp (T. Osawa).
https://doi.org/10.1016/j.mcat.2018.02.023
Received 26 December 2017; Received in revised form 20 February 2018; Accepted 20 February 2018
2468-8231/©2018ElsevierB.V.Allrightsreserved.

T. Osawa et al.
MolecularCatalysis449(2018)131–136
Scheme 1. Enantio-differentiating hydrogenation of alkyl 3-oxobutanoates to hydroxyesters over nickel powder modified with (R,R)-tartaric acid and NaBr.
from which the origin of unconventional temperature-dependence of
the e.e. value is elucidated.
from the integrated areas of the enantiomer peaks. Repeated experi-
ments gave nearly the same e.e. values with high reproducibility; the
error of e.e. ≤1%.
2. Experimental Section
3. Results and discussion
2.1. Materials
3.1. Enantio-differentiating hydrogenation of alkyl 3-oxobutanoates (1a-
1f)
Methyl acetoacetate (> 95%), ethyl 3-oxobutanoate (> 98%),
acetic acid (> 99.7%), tetrahydrofuran (> 99.5%), tartaric acid
(> 99.5%), and sodium bromide (> 99.5%) were purchased from
Wako Pure Chemical Industries, Ltd. Propyl 3-oxobutanoate (> 98%),
butyl 3-oxobutanoate (> 98%), isopropyl 3-oxobutanoate (> 97%),
and 2-methylpropyl 3-oxobutanoate (> 98%) were purchased from
Tokyo Chemical Industry Co., Ltd. All the chemicals were used as re-
ceived. Ni powder of 5 μm diameter was purchased from Sigma-Aldrich
Japan K.K.
The enantio-differentiating hydrogenations of 1a-1f were carried
out at 333 − 393 K. The conversions and the product’s e.e. obtained
with the TA-modified Ni powder as well as the e.e. values reported for
the modified Raney Ni are summarized in Table 1.
The conversions were (quasi-)quantitative at temperatures ≥373 K
for all the substrates examined, but started to decrease at 363 K and
substantially diminished down to 3-8% at 333 K. No straightforward
relationship was found between the conversion and the alkyl (R') size
even at the lowest temperature examined. In contrast, the highest en-
antio-selectivities were consistently attained in the middle of tem-
perature range employed, i.e. 363 to 373 K, without any exception. This
enabled us to achieve the highest enantio-selectivity with a nearly
quantitative conversion. The highest e.e. values obtained for 1a-1f at
363–373 K are consistent with the previous data reported by us [34]. It
was shown that the enantio-selectivity was slightly smaller for the
methyl ester (1a) than for the longer/bulkier groups (1b-1f), although
the difference was only 1-3%.
Earlier, the enantio-differentiating hydrogenations of various alkyl
3-oxobutanoates with the TA-modified Raney Ni catalyst prepared si-
milarly have been investigated by Tai et al. [31,32]; the reported e.e.
values are also listed in Table 1 for a comparison purpose. The highest
enantio-selectivities obtained by using TA-modified Ni powder in the
present study are consistently higher than those attained with TA-
modified Raney Ni under comparable conditions. Another major dif-
ference was found in the temperature dependence of e.e. In the Raney
Ni case, the enantio-selectivity does not appreciably depend on the
hydrogenation temperature, whereas the e.e. value rapidly decreases
with lowering the temperature particularly below 363 K in the Ni
powder case.
2.2. Enantio-differentiating hydrogenation of alkyl 3-oxobutanoates (1a-
1f)
Commercially available nickel powder was directly subjected to the
chiral modification without any pre-activation by hydrogen gas, ac-
cording to the procedures described previously [30]. Briefly, nickel
powder (0.5 g) was immersed for 1 h in an aqueous solution (50 cm3)
containing (R,R)-TA (0.5 g) and NaBr (2.0 g) at 373 K, the pH of which
was adjusted to 3.2 by using a 1 mol dm−3 NaOH solution before
modification. After the modification, the Ni catalyst was successively
washed once with deionized water (10 cm3), twice with methanol
(25 cm3 each), and then twice with THF (10 cm3 each). The modified
catalyst was added to a mixture of alkyl 3-oxoalkanoate (21.5 mmol),
acetic acid (0.1 g), and THF (10 cm3) placed in an autoclave equipped
with a magnetic stirrer (OM Lab-Tech Co., Ltd., Tochigi, Japan). The
autoclave was pressurized to 9 MPa with hydrogen gas and heated to a
desired temperature. The mixture was stirred for 20 h at a constant rate
of 1370 rpm. Then, the pressure was released and the catalyst was re-
moved from the reaction mixture by decantation with the aid of
magnet. The conversion was determined by GLC analysis using a Hi-
tachi 263-30 instrument equipped with a 5% Thermon 1000-coated
Chromosorb W column (3 mm I.D. × 2 m).
Before discussing the activation parameters calculated from the e.e.
data obtained, we scrutinized the conceivable influences of reaction
parameters such as conversion, hydrogen pressure, and the desorption
of modifiers, as well as the potential racemization of hydrogenated
products.
The effect of conversion on e.e. in heterogeneous hydrogenation
over TA-modified nickels has been examined by several groups
[35–37]. One of the factors leading to conversion-dependent e.e. is the
difference in induction period of the enantio-differentiating (e.d.)
versus non-enantio-differentiating (non-e.d.) site; thus, the lowered e.e.
2.3. Determination of enantio-selectivity
The hydrogenated products were derived to the corresponding
acetyl esters by treating with acetyl chloride and their e.e. values were
determined by chiral GLC using a Shimadzu GC-18A instrument
equipped with a CP Chirasil DEX-CB column (0.25 mm x 25 m) [31].
The retention times of the enantiomers of each product are listed in
Table S1 in the Supporting Information. The e.e. value was calculated
132

T. Osawa et al.
MolecularCatalysis449(2018)131–136
Table 1
were desorbed from the nickel surface, no deterioration of the e.e. was
observed at least upon repeated use of TA-modified Ni catalyst prepared
from nickel oxide; for detailed analysis data for the amount of ad-
sorption species and e.e., see Table S3 in the Supporting Information.
The effect of hydrogen pressure has already been examined for the
hydrogenation of methyl acetoacetate (1a) over TA-NaBr-modified Ni
powder to reveal that the product e.e. is not affected by hydrogen
pressure to afford constant 90-91% at least in the pressure range of
9–11 MPa [29]. This result also indicates that the concentration of
hydrogen gas in the liquid phase does not affect the e.e. under the
present reaction conditions.
Enantio-differentiating hydrogenation of alkyl 3-oxobutanoates (CH3COCH2CO2R') over
TA-modified Ni powder and Raney Ni.
Substrate (R')
Hydrogenation
temperature/K
Ni powder
Raney Nia
e.e./%
Conversion/%
e.e./%
1a (Me)
393
383
373
363
353
343
333
393
383
373
363
353
343
333
393
383
373
363
353
343
333
393
383
373
363
353
343
333
393
383
373
363
353
343
333
393
383
373
363
353
343
333
100
100
100
98
71
29
8
100
96
97
90
42
23
4
100
100
100
74
39
21
3
100
100
87
74
43
22
4
98
100
98
90
58
21
6
100
100
97
77
49
31
7
81
89
89
90
87
85
61
88
92
94
92
90
84
61
88
91
92
92
91
89
38
88
91
93
92
91
88
39
86
91
92
94
92
83
54
88
92
94
92
90
86
46
86
86
88
1b (Et)
1c (Pr)
1d (Bu)
1e (iPr)
1f (iBu)
3.2. Differential activation parameters for enantio-differentiation:
Enthalpy-entropy compensation
To obtain a global picture of the enantio-differentiating hydro-
genation as well as insights into the origin and mechanism of the un-
precedented irregular temperature-dependence of product’s e.e. in the
catalytic hydrogenation of 1a-1f over TA-modified Ni powder, we
employed the Eyring approach [38] to this heterogeneous catalytic
hydrogenation system.
As widely accepted in the literature [6,39–42], the enantio-differ-
entiating hydrogenation of KE over TA-modified Ni catalyst (Ni-TA)
proceeds through a pair of diastereomeric transition states [Ni-TA···re-
KE]‡ and [Ni-TA···si-KE]‡ (where re-KE and si-KE denote the enantio-
topically differentiated substrate upon interaction with TA on the Ni
surface), which eventually afford an enantiomer pair of hydroxyesters,
(R)-HE and (S)-HE, upon subsequent hydrogenation. Fig. 1 illustrates
the energy profile for the catalytic hydrogenation of 1a-1f where (R)-
product is favored.
88
88
In the energy profile shown in Fig. 1, the net free energy of acti-
vation for the formation of (R)- or (S)-HE, i.e. ΔG‡re or ΔG‡si, is ex-
pressed by a sum of the apparent free energy for the adsorption of KE to
TA on the Ni surface with the enantiotopic re or si face exposed to the
85
87
bulk solution (ΔGad or ΔGadsi) and that for the addition of hydrogen
re
atom to the adsorbed KE (ΔG‡Hre or ΔGH‡ si) as follows.
ΔG‡re = ΔG‡Hre + ΔGad
re
ΔG‡si = ΔG‡Hsi + ΔGad
si
Experimentally, we can evaluate the apparent differential free en-
ergy of activation (ΔΔG‡ = ΔG‡re−ΔG‡si) from the enantiomer ratio,
e.r. = (R)-HE/(S)-HE, using Eq. (1).
a
Data from references [31], [32] and [33], where only the e.e. values are reported at
one or two temperatures without mentioning to the conversion.
observed in the early stage of reaction is explained by the longer in-
duction period of the e.d. site) [35]. Keane reported that the enantio-
selectivity is slightly reduced to 29% e.e. at a very low conversion (5%)
but recover to a constant value (31% e.e.) at conversions ≥20% [36].
More recently, Kukula and Cervený have shown that the induction
period becomes longer when the rate of reaction is slower [37].
In the present study, the effect of conversion on e.e. was negligible
within the experimental error ( 1%) upon hydrogenation of 1a and
1b at 373 K, affording constant 87-88% and 91-92% e.e., respectively,
over a wide range of conversion, i.e. 20-100%, although the e.e. for 1a
was appreciably decreased to 84% at 14% conversion, as shown in
Table S2. For this reason, all the e.e. data obtained at 333 K, where the
conversions were well below 10%, were not taken into consideration in
the Eyring analyses in the following section.
Racemization of the hydrogenated product was examined by
treating methyl 3-hydroxybutyrate of 83.6% e.e. at 393 K for 20 h under
exactly the same reaction conditions employed for the hydrogenation
over TA-modified Ni powder to recover the hydroxy-ester in essentially
the same 82.7% e.e.
Fig. 1. A schematic illustration of the energy profile for the enantio-differentiating hy-
drogenation of keto-ester (KE) to (R)- and (S)-hydroxyester (HE) over (R,R)-TA-modified
Ni catalyst via a pair of diastereomeric transition states [Ni-TA···re-KE]‡ and [Ni-TA···si-
KE]‡.
The effects of the desorption of modifiers during the reaction on e.e.
would also be negligible. Even when the small amounts of the modifiers
133

T. Osawa et al.
MolecularCatalysis449(2018)131–136
Fig. 2. Differential Eyring plots of the enantiomer ratios obtained in the enantio-differ-
entiating hydrogenation of 1a (R = Me) over TA-modified Ni powder in the high and low
temperature regions (green triangles and blue circles, respectively); the e.e. data at 333 K
(crossed out in the graph) is less reliable due to the very low conversion and hence not
taken into consideration in the subsequent quantitative treatments.
ΔΔG‡ = ΔG‡re−ΔG‡si = −RTln(e. r.)
(1)
Fig. 3. Enthalpy-entropy compensation plot of the differential activation parameters
obtained in the enantio-differentiating hydrogenation of various alkyl 3-oxoalkanoates
over TA-modified Ni powder catalyst (open and closed blue circles for the high and low
temperature regions) and Raney Ni catalyst [20] (yellow diamonds).
where R represents the gas constant. By combining the Eq. (1) with the
differential form of Gibbs-Helmholtz equation (ΔΔG‡ = ΔΔH‡
TΔΔS‡), we obtain Eq. (2).
−
ΔΔH‡
RT
ΔΔS‡
R
ΔH‡re − ΔH‡si
ΔS‡re − ΔS‡si
ln(e. r.) = −
+
= −
+
of the effective enantio-differentiation step involved. In this catalytic
reaction, the hydrogen addition is slowed down at lower temperatures
to dominate the overall rate of hydrogenation. While the amount of KE
on the TA-modified Ni surface decreases, the hydrogen addition to KE
from both the enantiotopic re- and si-faces becomes faster to approach
the diffusion limit of hydrogen (where no enantio-differentiation is
expected to occur) at higher temperatures and the overall hydrogena-
tion is governed by the desorption thermodynamics. Combining these
considerations with the activation parameters obtained experimentally,
we may assume that the hydrogen addition to adsorbed KE is en-
tropically controlled, while the adsorption/desorption process is en-
thalpically driven. However, gaining further evidence for this hypoth-
esis does not appear to be feasible due to the lack of tools to evaluate
separately the rates and equilibria of heterogeneous catalysis. More
detailed discussion of this topic is presented in Section 3.3.
(2)
RT
R
This equation allows us to evaluate the differential activation en-
thalpy (ΔΔH‡) and entropy (ΔΔS‡) for the enantio-differentiating hy-
drogenation over TA-modified Ni powder. Indeed, the plot of ln(e.r.)
calculated from the e.e. values (Table 1) as a function of the inverse
temperature gave a good straight line in each of the low and high
temperature regions, as exemplified for 1a in Fig. 2; for the plots for
other substrates, see Fig. S1 in the Supporting Information.
An apparently resembling bent Eyring plot has been reported for the
diastereo-differentiating Paternó-Büchi reaction of cyclohexanoates
with ketone [43], the origin of which was shown to be the switching of
diastereo-differentiation step by temperature.
From the slope and intercept of the regression line, the ΔΔH‡ and
ΔΔS‡ values were obtained for each region, as listed in Table 2.
Over the entire range of temperature examined, (R,R)-TA-modified
Ni catalyst consistently affords (R)-HE in good-to-high enantio-se-
lectivities (Table 1). However, its origin turned out to be totally dif-
ferent in the low and high temperature regions. The activation para-
meters obtained (Table 2) revealed that the high enantio-selectivity is
exclusively driven by the positive differential entropy of activation
(ΔΔS‡ > 0) in the low temperature region, but by the negative differ-
ential enthalpy of activation (ΔΔH‡ < 0) in the high temperature re-
gion. One of the plausible explanations for this alteration in the driving
force, as well as for the obviously bent plot (Fig. 2), would be switching
Of particular interest, all of the ΔΔH‡ and ΔΔS‡ values obtained in
both the low and high temperature regions (Table 2) perfectly com-
pensate with each other to give a single straight line shown in Fig. 3
(blue circles), despite the temperature switching of enantio-differ-
entiating step (vide supra) [44]. Furthermore, the activation parameters
calculated from the data reported for the catalytic hydrogenation of a
series of alkyl 3-oxoalkanoates (R = Me, Et, Pr, Bu, hexyl, cPr, cyclo-
pentyl, cyclohexyl, neopentyl, R’ = Me and R = Pr, R’ = Me) over TA-
modified Raney Ni [20] are plotted very closely to the same line (Fig. 3,
yellow circles). The regression analysis of all the plotted data affords an
empirical equation: ΔΔH‡ = 365ΔΔS‡ − 9.89 (correlation coefficient
r = 0.9979). The slope of regression line has a dimension of tempera-
ture and is called isokinetic temperature [45], which is 365 K in the
present system. By incorporating the above empirical equation into the
differential Gibbs-Helmholtz equation, ΔΔG‡ = ΔΔH‡ − TΔΔS‡, we ob-
tained the following equations:
Table 2
Differential activation parameters for the enantio-differentiating hydrogenation of alkyl
3-oxolkanoates 1-4 with TA-modified Ni powder in the temperature ranges 343–363 K
and 373–393 K.a
Substrate 343–363 K
373–393 K
ΔΔH‡/kJ
ΔΔS‡/J
ΔΔH‡/kJ
ΔΔS‡/J
ΔΔG‡ = (365–T) ΔΔS‡ − 9.89
mol−1
mol−1 K−1
mol−1
mol−1 K−1
ΔΔG‡ = (1–T/365) ΔΔH‡ − 9.89(T/365)
1a
1b
1c
1d
1e
1f
23
38
21
22
51
31
89
132
84
86
169
112
-35
-22
-25
-30
-29
-38
-70
-32
-41
-54
-52
-74
where ΔΔG‡ is expressed solely by ΔΔS‡ or ΔΔH‡. These equations in-
dicate that the enantio-selectivity becomes identical at 365 K for all the
substrates to afford the same e.e. of 96% (corresponding to the ΔΔG‡
value of −9.89 kJ mol−1), irrespective of the R/R’ substituents and the
Ni catalyst (Ni powder or Raney Ni).
ΔΔH‡ = ΔHR‡ – ΔH‡S; ΔΔS‡ = ΔS‡R – ΔS‡S, The ΔΔH‡ and ΔΔS‡ values were obtained by
a
Similar compensatory enthalpy-entropy relationship has been
the analysis of the e.e. values at different temperatures (Table 1) using Eq. (2).
134

T. Osawa et al.
MolecularCatalysis449(2018)131–136
Scheme 2. Reaction mechanism for the enantio-dif-
ferentiating hydrogenation of KE to (R)- and (S)-HE
over TA-modified Ni catalyst (Ni-TA), which pro-
ceeds through a pair of diastereomeric complexes
[Ni-TA···re-KE] and [Ni-TA···si-KE] followed by the
attack of hydrogen generated on Ni surface (Ni-H)
and subsequent spontaneous desorption of produced
HE from the surface.
observed for the activation parameters derived from the enantiomeric
ratios obtained in the photosensitized enantio-differentiating geome-
trical isomerization of (Z)-cyclooctene to chiral (E)-isomer via diaster-
eomeric excited-state complexes of the substrate with a chiral sensi-
tizer, and taken as evidence in support of the same enantio-
differentiation mechanism operative in all the solvents employed
[46–48]. In the present case, the effective enantio-differentiating step is
likely to be switched in the middle of temperature range employed (the
origin of which will be discussed in the sections to follow). Never-
theless, no obvious deviation or kink is seen in the regression line
(Fig. 3), which is hard to rationalize if the enantio-differentiation me-
chanism suffers significant changes. Therefore, it is reasonable to con-
clude that the same enantio-differentiation mechanism, involving the
interaction of prochiral KE with Ni-TA as a major driving force, oper-
ates for all the substrates and over the entire temperature range upon
catalytic hydrogenation with TA-modified Ni powder and with TA-
modified Raney Ni as well.
It is interesting to further discuss the origin of the different driving
forces for determining the enantio-selectivity upon hydrogenation over
modified Ni powder versus Raney Ni. Thus, the high enantio-selectivity
is driven by negative ΔΔH‡ in the modified Raney Ni case, but by po-
sitive ΔΔS‡ in the low temperature region and by negative ΔΔH‡ in the
high temperature region in the modified Ni powder case. One of the
plausible explanations for this apparent switching of the determinant
activation parameter is that the optimum temperature for giving the
highest e.e. is in the middle of the experimental temperature range in
the Ni power case, but is much lower for the RNi catalyst due to its
higher hydrogenation activity.
In the initial stage, prochiral KE is adsorbed and activated on the Ni
surface by forming a pair of the diastereomeric hydrogen-bonding
complexes with TA immobilized on Ni, i.e., [Ni-TA···re-KE] and [Ni-
TA···si-KE], which in turn expose the enantiotopic re- and si-face to the
bulk solution. Subsequently, the attack of hydrogen on nickel surface
(Ni-H) occurs from the opposite face to give (R)- and (S)-HE, respec-
tively (Scheme 2). Similar reaction scheme has already been proposed
for the hydrogenation over heterogeneous [49] and homogeneous cat-
alysts [50]. The diastereomeric complexes are in equilibrium with free
KE at the forward rate constant k1re or k1si, and backward rate constant
re
k-1 or k-1si, while the hydrogenation process proceeds irreversibly at
re
si
rate constant kH or kH
.
The steady-state approximation applied to the intermediate complex
of prochiral KE with TA on Ni surface leads to the general Eq. (3) for the
e.r. value of HE produced.
vR
vS
[R]
[S]
k1re (k si + kHsi [Ni-H]) kHre kdR
−1
e. r. =
=
=
⋅
⋅
⋅
(k re + kHre [Ni-H])
k1si
kHsi kdS
(3)
−1
where the dissociation rate constants of both enantiomers from the
catalyst surface, kdR and kdS, are also taken into consideration. The last
term is practically close to unity (kdR/kdS ≈ 1) and hence negligible, as
the enantioselectivity is not dependent on the conversion and no pro-
duct inhibition was observed in all the examined cases. This general
equation can be further simplified in different ways in the low and high
temperature regions as detailed in the following sections for better
understanding of the factors and mechanism that control the en-
antioselectivity in these two regions.
3.3.1. Enantioselectivity in the high temperature region (373–393 K)
At higher temperatures, KE bound to TA-Ni is more rapidly con-
sumed by the attack of hydrogen on the Ni surface and the dissociation
constant of KE-TA complex becomes negligible in the rate equation; i.e.,
k-1re < < kHre[Ni-H] and k-1si < < kHsi[Ni-H]. Under such a condition,
the Eq. (3) is reduced to Eq. (4), which indicates that the enantios-
electivity is determined solely by the relative rate of complex forma-
tion.
3.3. Enantio-differentiation mechanism
In general, the enantio-differentiation mechanism operative in
heterogeneous catalysis is apparently more complicated than that in
homogeneous catalysis. However, the essential process(es) that de-
termine the product’s e.e. do not essentially differ. Thus, in both the
homogeneous and heterogeneous catalyses, the intermediate associated
with the enantio-differentiation is a pair of diastereomeric catalyst-
substrate complexes formed either in isotropic solution or on solid
surface. Thus, the difference in activation free energy (ΔΔG‡) between
the diastereomeric complex pair, precursor to the (R)- and (S)-products
(Fig. 1), is the only determinant in both the cases.
Although a general mechanism for the enantio-differentiating hy-
drogenation over TA-modified Raney Ni has already been proposed by
Tai et al. [6] and also by our research group [42] (Fig. S2), these models
do not fully explain the effect of hydrogenation temperature on the
enantio-selectivity. The present catalytic system, which employs ap-
parently similar TA-modified Ni powder as catalyst but exhibits the
unprecedented temperature-dependence behavior of enantio-se-
lectivity, enabled us to examine this mechanism more closely based on
the effect of the hydrogenation temperature on e.e. Indeed, as discussed
above, the same enantio-differentiation mechanism is considered to
operate in the high and low temperature regions despite the opposite
temperature-dependence of e.e. observed in these two regions (vide
supra). We now propose a mechanism compatible with all the experi-
mental results obtained in the past and present studies on the enantio-
differentiating hydrogenation with TA-modified Ni catalysts.
k1re
e. r. =
k1si
(4)
The validity of this equation has been experimentally confirmed for
the enantio-differentiating hydrogenation of methyl acetoacetate over
Raney Ni catalyst modified with chiral amino acids and hydroxyalk-
anoic acids in a kinetic study [49], where the apparent hydrogenation
rate constants for the conversion of adsorbed KE to the both HE en-
antiomers were demonstrated to be comparable. If this is also the case
with the hydrogenation over TA-modified Ni powder in the high tem-
perature region, the differential activation parameters obtained ex-
perimentally (Table 2) should belong to the initial complex formation
process (Eq. (4)). Thus, the (R)-preference observed is kinetic in origin
si
(k1re/k1 > 1) and ascribed to the large negative (favorable) ΔΔH‡
values despite the entropic drawback (ΔΔS‡ < 0). This is the cause that
lowers the enantio-selectivity when the hydrogenation temperature is
elevated in the high temperature region (Fig. 2). The kinetic (R)-pre-
ference is likely to originate from the thermodynamic stability of in-
termediate complex [Ni-TA···re-KE] compared to that of [Ni-TA···si-KE],
as depicted in Fig. 1.
135

T. Osawa et al.
MolecularCatalysis449(2018)131–136
3.3.2. Enantioselectivity in the low temperature region (343–363 K)
At lower temperatures, the hydrogen addition to adsorbed KE slows
down and the pre-equilibrium between free and adsorbed KE is estab-
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the
online version, at http://dx.doi.org/10.1016/j.mcat.2018.02.023.
re
si
lished; i.e., k-1 > > kHre[Ni-H] and k-1 > > kHsi[Ni-H]. Then, the
Eq. (3) is reduced to Eq. (5) as follows.
References
k1re k si kHre
kHre
kHsi
Kre
Ksi
−1
e. r. =
·
·
=
·
k re
k1si
kHsi
(5)
−1
[1] I.F.J. Vankelecom, P.A. Jacobs, D.E.D. Vos, I.F.J. Vankelecom, P.A. Jacobs (Eds.), Chiral
Catalyst Immobilization and Recycling, Wiley-VCH, 2000, p. 19.
[2] M. Beller, H.U. Blaser, Organometallics as Catalysts in the Fine Chemical Industry,
Springer GmbH, 2012.
[3] M.M. Heravi, S. Asadi, Tetrahedron: Asymmetry 23 (2012) 1431–1465.
[4] Y. Li, P.C. Cirino, Biotechnol. Bioeng. 111 (2014) 1273–1287.
[5] M. Kataoka, T. Miyakawa, S. Shimizu, M. Tanokura, Appl. Microbiol. Biotechnol. 100
(2016) 5747–5757.
This indicates that the enantioselectivity is governed both thermo-
dynamically and kinetically by the relative complex stability (Kre/Ksi)
and the relative rate of hydrogen addition (kHre/kHsi).
In this region also, the catalytic hydrogenation over TA-modified Ni
powder leads to the predominant formation of (R)-HE, which is how-
ever driven by the positive ΔΔS‡ (Table 2). Hence, the enantioselectivity
is enhanced by increasing the hydrogenation temperature (Table 1), as
ΔΔG‡ = ΔΔH‡ − TΔΔS‡, where consistently ΔΔH‡ < TΔΔS‡ (ΔΔH‡ ≥ 0,
ΔΔS‡ > 0). However, the origin of the positive ΔΔS‡ is not immediately
assignable to a single cause, as the enantioselectivity is a function of
two factors, i.e., Kre/Ksi and kHre/kHsi. On the basis of the accepted
complexation models for (R,R)-TA-modified Raney Ni catalysis [6,42]
and the fact that (R)-HE is predominantly produced in the present
catalytic hydrogenation, the [Ni-TA···re-KE] complex precursor to (R)-
HE is considered to be more stable than the [Ni-TA···si-KE] complex due
to its favorable conformation to build stronger hydrogen-bonding and/
or ion-dipole interactions of KE with Ni-TA (where TA is adsorbed as
sodium salt of tartaric acid [51,52]). The more stable complex is con-
formationally better defined (more rigid) with lesser freedoms and
hence its formation, though driven by the enthalpic gain from the
above-mentioned interactions, should be inherently less favored by
entropy. Therefore, the complexation step is not likely to be the origin
of positive ΔΔS‡ observed for the present system. Instead, the con-
formationally more rigid [Ni-TA···re-KE] intermediate, rather than the
more flexible [Ni-TA···si-KE] complex, will gain additional freedoms in
the transition state upon subsequent hydrogen attack, affording a po-
sitive ΔΔS‡ value, as observed experimentally (Table 2).
[6] A. Tai, T. Sugimura, I.F.J. Vos (Ed.), Chiral Catalyst Immobilization and Recycling, Wiley-
VCH2000, 2018, pp. 173–209.
[7] U. Kragl, T. Dwars, Trends Biotechnol. 19 (2001) 442–449.
[8] M. Studer, H.U. Blaser, C. Exner, Adv. Synth. Catal. 345 (2003) 45–65.
[9] D.Y. Murzin, P. Mäki-Arvela, E. Toukoniitty, T. Salmi, Catal. Rev. 47 (2005) 175–256.
[10] T. Osawa, T. Harada, O. Takayasu, Curr. Org. Chem. 10 (2006) 1513–1531.
[11] T. Mallat, E. Orglmeister, A. Baiker, Chem. Rev. 107 (2007) 4863–4890.
[12] Y. Liu, W. Xuan, Y. Cui, Adv. Mater. 22 (2010) 4112–4135.
[13] G. Nickerl, A. Henschel, R. Gruenker, K. Gedrich, S. Kaskel, Chem. Ing. Tech. 83 (2011)
90–103.
[14] C.J. Baddeley, T.E. Jones, A.G. Trant, K.E. Wilson, Top. Catal. 54 (2011) 1348–1356.
[15] M. Yoon, R. Srirambalaji, K. Kim, Chem. Rev. 112 (2012) 1196–1231.
[16] E. Zhan, C. Chen, Y. Li, W. Shen, Catal. Sci. Technol. 5 (2015) 650–659.
[17] C.M. Kisukuri, L.H. Andrade, Org. Biomol. Chem. 13 (2015) 10086–10107.
[18] H.U. Blaser, H.J. Federsel (Eds.), Asymmetric Catalysis On Industrial Scale: Challenges,
Approaches, And Solutions, 2nd Edition, Wiley-VCH Verlag GmbH & Co. KGaA, 2010.
[19] Y. Izumi, Adv. Catal. 32 (1983) 215–271.
[20] T. Sugimura, S. Nakagawa, A. Tai, Bull. Chem. Soc. Jpn. 75 (2002) 355–363.
[21] C. Ding, W. Wei, H. Sun, J. Ding, J. Ren, X. Qu, Carbon 79 (2014) 615–622.
[22] A. López-Martínez, M.A. Keane, J. Mol. Catal. A: Chem. 153 (2000) 257–266.
[23] T. Osawa, Y. Nakagawa, M. Ando, T. Harada, O. Takayasu, J. Mol. Catal. A: Chem. 302
(2009) 43–47.
[24] Z. Lou, X. Chen, L. Tian, M. Qiao, K. Fan, H. He, X. Zhang, B. Zong, J. Mol. Catal. A: Chem.
326 (2010) 113–120.
[25] J. Masson, P. Cividino, J. Court, J. Mol. Catal. A: Chem. 111 (1996) 289–295.
[26] P. Kukula, L. Cerveny, Appl. Catal., A 210 (2001) 237–246.
[27] S. Cai, K. Ma, H. Chen, Y. Zheng, J. Jiang, R. Li, Catal. Lett 128 (2009) 227–234.
[28] T. Osawa, T. Kizawa, I.-Y.S. Lee, S. Ikeda, T. Kitamura, Y. Inoue, V. Borovkov, Catal.
Commun. 15 (2011) 15–17.
[29] T. Osawa, T. Kizawa, F. Takano, S. Ikeda, T. Kitamura, Y. Inoue, V. Borovkov,
ChemCatChem 6 (2014) 170–178.
[30] T. Osawa, I.-Y.S. Lee, S. Ikeda, T. Kitamura, Y. Inoue, V. Borovkov, Appl. Catal., A: Gen.
445–446 (2012) 269–273.
[31] T. Sugimura, T. Osawa, S. Nakagawa, T. Harada, A. Tai, Stud. Surf. Sci. Catal. 101 (1996)
231–240.
4. Conclusions
[32] A. Tai, T. Harada, Y. Hiraki, S. Murakami, Bull. Chem. Soc. Jpn. 56 (1983) 1414–1419.
[33] A. Tai, D.G. in: Morrell (Ed.), Catalysis of Organic Reactions, Marcel Dekker, Inc., 2003,
pp. 191–223.
[34] T. Osawa, T. Kizawa, S. Ikeda, T. Kitamura, Y. Inoue, V. Borovkov, Tetrahedron:
Asymmetry 25 (2014) 1630–1633.
In the present study, the enantio-differentiating hydrogenation of a
series of alkyl 3-oxobutanoates over TA-modified Ni powder catalyst
was examined at the temperatures ranging from 333 to 393 K. The
enantioselectivity of alkyl 3-hydroxybutanoates produced was not a
simple function of the hydrogenation temperature to afford a bent
Eyring plot with the highest (ceiling) enantioselectivity of 91-94% e.e.
attained at 363–373 K for all the substrates. Nevertheless, the differ-
ential activation parameters derived from the enantioselectivities ob-
tained in both the low and high temperature regions for TA-modified Ni
powder and Raney Ni catalysts exhibited a nearly perfect enthalpy-
entropy compensation. The compensation effect observed indicates that
a common enantio-differentiation mechanism operates for all the sub-
strates and Ni catalysts over the entire temperature range employed.
The slope and intercept of the regression line enabled us to gain in-
depth insights into the factors and mechanism that control the catalytic
hydrogenation over the chirally modified Ni catalysts. The results and
concept obtained in this study allow us further rational design of het-
erogeneous catalytic systems.
[35] Y. Nitta, F. Sekine, J. Sasaki, T. Imanaka, S. Teranishi, J. Catal. 79 (1983) 211–214.
[36] M.A. Keane, Langmuir 13 (1997) 41–50.
[37] P. Kukula, L. Cerveny, J. Mol. Catal. A: Chem. 185 (2002) 195–202.
[38] Y. Inoue, N. Yamasaki, T. Yokoyama, A. Tai, J. Org. Chem. 57 (1992) 1332–1345.
[39] A. Tai, H. Watanabe, T. Harada, Bull. Chem. Soc. Jpn. 52 (1979) 1468–1472.
[40] I. Yasumori, M. Yokozeki, Y. Inoue, Faraday Discuss. Chem. Soc. 72 (1981) 385–396.
[41] E. Leclercq, A. Rives, E. Payen, R. Hubaut, Appl. Catal. A: Gen. 168 (1998) 279–288.
[42] T. Osawa, T. Harada, O. Takayasu, Top. Catal. 13 (2000) 155–168.
[43] H. Buschmann, H.D. Scharf, N. Hoffmann, M.W. Plath, J. Runsink, J. Am. Chem. Soc. 111
(1989) 5367–5373.
[44] A main difference between homogeneous and heterogeneous catalysis would be the en-
tropic frustration found in proteins. However, differing from conformationally more
flexible proteins, the lattice structure of metallic nickel surface used as a catalyst scaffold
in this study is robust and does not suffer any changes during the reaction. The chiral
discrimination, for which we obtained the differential enthalpy and entropy values, arises
solely from the interaction between the tartaric acid adsorbed on nickel surface and the
substrate. Therefore, the entropic frustration is not expected to affect the differential
enthalpy and entropy values in this study.
[45] E.G. John, E. Leffler, Rates and Equilibria of Organic Reactions, Dover Publishing, 1989.
[46] Y. Inoue, E. Matsushima, T. Wada, J. Am. Chem. Soc. 120 (1998) 10687–10696.
[47] M.V. Rekharsky, Y. Inoue, Chem. Rev. 98 (1998) 1875–1917.
[48] C. Yang, A. Nakamura, G. Fukuhara, Y. Origane, T. Mori, T. Wada, Y. Inoue, J. Org. Chem.
71 (2006) 3126–3136.
Funding Sources
[49] H. Ozaki, A. Tai, S. Kobatake, H. Watanabe, Y. Izumi, Bull. Chem. Soc. Jpn. 51 (1978)
3559–3563.
[50] C.R. Landis, J. Halpern, J. Am. Chem. Soc. 109 (1987) 1746–1754.
[51] T. Harada, Bull. Chem. Soc. Jpn. 53 (1980) 1019–1022.
[52] T. Harada, M. Yamamoto, S. Onaka, M. Imaida, H. Ozaki, A. Tai, Y. Izumi, Bull. Chem.
Soc. Jpn. 54 (1981) 2323–2329.
V.B. acknowledges funding from the European Union’s Seventh
Framework Program for research, Technological Development and
Demonstration under Grant Agreement No. 621364 (TUTIC-Green).
136