Chemoselective hydrogenation of nitrostyrene to aminostyrene over Pd- and Rh-based intermetallic compounds

Article abstract of DOI:10.1021/cs500082g

A noble catalytic system based on intermetallic compounds was developed for the chemoselective hydrogenation of p-nitrostyrene (NS) to p-aminostyrene (AS). The main concept of the catalyst design was to construct polar active sites consisting of two types of metals with different electronegativities. We prepared a series of Pd- and Rh-based intermetallics and investigated their catalytic properties in detail in the catalytic transfer hydrogenation of nitrobenzene and NS in the presence of 4-methyl-1-cyclohexene (MC) or methanol as a hydrogen donor. FT-IR studies of adsorbed CO confirmed that the number of Pd ensembles exposed on a surface was greatly decreased by the formation of an intermetallic phase; hence, the particle surface consisted of two metal elements being coadjacent at the atomic level. The product distribution achieved with Pd catalysts was dependent on the electronegativity of the second metal element: more electronegative metals gave higher AS selectivity and lower p-ethylnitrobenzene selectivity. Rh catalysts selectively gave AS, and their AS yields increased as the electronegativity of the second metal increased. The results revealed that an increase in the electronegativity of the second metal element provided polar sites and enhanced the activation of methanol as a hydrogen donor, which accelerated the hydrogenation of the nitro group of NS and, hence, improved the yield of AS. The high selectivity of Rh catalysts was due to the absence of MC activation ability, which caused the hydrogenation of the vinyl group of NS. Pd₁₃Pb₉ exhibited the highest chemoselectivity toward AS (92%) among the investigated Pd catalysts. Moreover, RhPb₂ exhibited not only high AS selectivity (93%) but also the highest NS conversion (94%) among the investigated catalysts. RhPb₂ also exhibited high selectivity toward AS (91%), even when H₂ was used as a hydrogen source. Thus, intermetallics that contain Pb, which was the most electronegative metal used in this study, afforded good catalytic performance and were observed to be good catalysts for the chemoselective hydrogenation of NS.

Full text of DOI:10.1021/cs500082g

Research Article
pubs.acs.org/acscatalysis
Chemoselective Hydrogenation of Nitrostyrene to Aminostyrene
over Pd- and Rh-Based Intermetallic Compounds
,†
†
,‡
Shinya Furukawa,* Yurika Yoshida, and Takayuki Komatsu*
†
‡
Department of Chemistry, Department of Chemistry and Materials Science, Tokyo Institute of Technology, 2-12-1-E1-10
Ookayama, Meguro-ku, Tokyo 152-8551, Japan
*
S Supporting Information
ABSTRACT: A noble catalytic system based on intermetallic compounds
was developed for the chemoselective hydrogenation of p-nitrostyrene
(
NS) to p-aminostyrene (AS). The main concept of the catalyst design
was to construct polar active sites consisting of two types of metals with
different electronegativities. We prepared a series of Pd- and Rh-based
intermetallics and investigated their catalytic properties in detail in the
catalytic transfer hydrogenation of nitrobenzene and NS in the presence of
4
-methyl-1-cyclohexene (MC) or methanol as a hydrogen donor. FT-IR
studies of adsorbed CO confirmed that the number of Pd ensembles
exposed on a surface was greatly decreased by the formation of an intermetallic phase; hence, the particle surface consisted of two
metal elements being coadjacent at the atomic level. The product distribution achieved with Pd catalysts was dependent on the
electronegativity of the second metal element: more electronegative metals gave higher AS selectivity and lower p-
ethylnitrobenzene selectivity. Rh catalysts selectively gave AS, and their AS yields increased as the electronegativity of the second
metal increased. The results revealed that an increase in the electronegativity of the second metal element provided polar sites
and enhanced the activation of methanol as a hydrogen donor, which accelerated the hydrogenation of the nitro group of NS
and, hence, improved the yield of AS. The high selectivity of Rh catalysts was due to the absence of MC activation ability, which
caused the hydrogenation of the vinyl group of NS. Pd Pb exhibited the highest chemoselectivity toward AS (92%) among the
13
9
investigated Pd catalysts. Moreover, RhPb exhibited not only high AS selectivity (93%) but also the highest NS conversion
2
(
94%) among the investigated catalysts. RhPb also exhibited high selectivity toward AS (91%), even when H was used as a
2
2
hydrogen source. Thus, intermetallics that contain Pb, which was the most electronegative metal used in this study, afforded good
catalytic performance and were observed to be good catalysts for the chemoselective hydrogenation of NS.
KEYWORDS: intermetallic, chemoselective, nitrostyrene, catalytic transfer hydrogenation, polarity
1
. INTRODUCTION
supported metal catalysts, however, such interfacial sites are
limited to only a small portion of the exposed surface metal
atoms. Mitsudome et al. designed a core−shell-structured Ag−
Selective hydrogenation of nitro groups in the presence of
other reducible functional groups is an important chemical
transformation to obtain functionalized anilines as important
intermediates for pharmaceuticals, polymers, herbicides, and
other fine chemicals. Selective hydrogenation has typically
been performed using a stoichiometric amount of reducing
agents, which produces a large amount of waste byproducts.
In view of the principles of green chemistry, the development
of a heterogeneous catalytic system for the chemoselective
hydrogenation of nitro compounds is desirable; however,
conventional platinum-group metal catalysts hydrogenate not
only the nitro group but also any olefinic or carbonyl functional
CeO nanocomposite with high chemoselectivity because of its
2
9
maximized fraction of interfacial sites.
1
,2
In this context, an ideal structure for chemoselective
hydrogenation would be an arrangement of interfacial sites
that exhibit high order at the atomic level. One of the promising
candidates for such a material is an intermetallic compound.
Intermetallic compounds are stoichiometric compounds of two
or more metal elements; they have specific crystal structures,
highly ordered surface atomic arrangements, and modified
2
−5
12
electronic structures compared with the parent metals.
6
Therefore, an appropriate combination of two metal elements
with different electronegativities would provide an ideal
arrangement of active sites to generate partially charged
hydrogen species and to exhibit high chemoselectivity. In the
case of hydrogenation of α,β-unsaturated aldehyde to the
groups. Currently, only a few catalytic systems that involve
nanoparticulate Au or Ag
7
,8
9,10
supported on specific oxides
have achieved the chemoselective hydrogenation of nitro
aromatic compounds, such as nitrostyrene into aminostyrene.
The high selectivities of these catalysts have been explained by
the formation of charged hydrogen species (H or H ) at the
metal−support interface; these species are active toward the
reduction of the polar nitro group, but are inactive toward the
+
−
Received: January 20, 2014
Revised: March 20, 2014
Published: March 24, 2014
8
−11
reduction of the nonpolar CC bond.
In the case of these
©
2014 American Chemical Society
1441
dx.doi.org/10.1021/cs500082g | ACS Catal. 2014, 4, 1441−1450

ACS Catalysis
Research Article
corresponding alcohol, it was reported that an increase in
polarity of the catalyst surface indeed enhanced the chemo-
selectivity. In addition to structural design, the choice of
reducing agent is also an important factor governing chemo-
and SnCl ·2H O. The composition ratio and the reduction
2
2
temperature of each intermetallic catalyst are as follows: Pd−Bi
(3:1), 1073 K; Pd−Cu (1:1), 1073 K; Pd−Fe (1:1), 773 K;
Pd−Ga (13:5), 873 K; Pd−Ga (5:3), 1073 K; Pd−Ga (1:1),
1073 K; Pd−Pb (1:1), 1073 K; Pd−Sn (3:1), 773 K; Pd−Zn
(1:1), 1073 K; Rh−Bi (1:1), 773 K; Rh−Fe (1:1), 1073 K;
Rh−Ni (1:1), 1073 K; Rh−Pb (3:2), 1073 K; Rh−Pb (1:1),
873 K; Rh−Pb (1:2), 1073 K; Rh−Sn (2:1), 1073 K; Rh−Ti
(3:1), 1073 K. We carried out the reduction treatment at high
1
3
selective hydrogenation. The use of gaseous hydrogen (H ) as a
2
reducing agent typically requires a high pressure of hydrogen
7
−10
and a special apparatus.
Moreover, the reduction of both
nitro and olefinic groups often occurs over platinum-group
6
metal catalysts. In contrast, catalytic transfer hydrogenation
(
CTH) using an appropriate hydrogen donor instead of H2
temperature with SiO support itself to investigate the effect of
2
makes handling easier and improves the selectivity of the
reaction.
have been reported to exhibit high catalytic activities in CTH
reactions of organic compounds.
In this study, we prepared a series of Pd- and Rh-based
intermetallic compounds supported on silica and investigated
their catalytic properties in CTH of p-nitrostyrene (NS). To
better understand the hydrogenation of the nitro group, we also
performed CTH of nitrobenzene in addition to that of NS.
Herein, we report our development of chemoselective
intermetallic catalysts and the dependence of the activity and
selectivity of these catalysts on their electronic and geometric
characteristics.
sintering of the support. N adsorption experiments before and
2
1
4−17
Among the platinum-group metals, Pd and Rh
after the treatment (at 1073 and 873 K) revealed that only a
2
−1
slight sintering occurred at 1073 K (SBET: 673 → 622 m ·g ),
1
7,18
whereas no sintering was observed at 873 K (SBET: 673 → 686
2
−1
m ·g ). Therefore, the treatment does not seem to result in a
considerable effect to the catalytic performances.
2.2. CTH Reaction. CTH reaction was carried out in a 100
mL, three-necked, round-bottom flask equipped with a reflux
condenser and a gas balloon. Prior to the reaction, catalyst (250
mg) was reduced in the reactor under flowing H (60 mL·
2
−1
min ) at 723 K for 0.5 h using a mantle heater. After the
reduction pretreatment, the reactor was cooled to room
temperature, and the atmosphere was completely replaced by
dry Ar. A methanol solution (5.0 mL) of nitrobenzene (0.50
mmol) or NS (0.25 mmol) was added into the reactor through
a silicone septum. The reaction was initiated by adding 4-
methyl-1-cyclohexene (0.75 mmol) as a hydrogen donor at 343
K. In the case of monometallic Pd/SiO and Rh/SiO , the
2
. EXPERIMENTAL SECTION
2
.1. Catalyst Preparation. Pd/SiO was prepared by a
2
pore-filling impregnation. An aqueous solution of Pd(NO )
(
3
2
4.8 × 10− g·mL , 0.64 mL, N.E. ChemCat) was diluted with
2
−1
2
2
catalyst amount and reduction temperature were 150 mg and
73 K, respectively. The products in the liquid phase were
ion-exchanged water and added to silica gel (5.0 g, Cariact G-6,
Fuji Silysia Co.) that had been previously dried at 403 K and
cooled in air to room temperature. The amount of palladium
solution was calculated to fill the pores of the silica gel and to
achieve the palladium loading of 3 wt %. The mixture was
sealed by a piece of plastic film overnight at room temperature.
It was then dried over a boiling water bath with stirring,
followed by calcination under dry at 403 K for 1 h and then at
5
identified by gas chromatograph mass spectroscopy (GC−MS,
JEOL, Automass System II) and quantitated by flame ionization
detection gas chromatograph (FID−GC, Shimadzu, GC-14B)
equipped with a capillary column (GL Sciences, TC-17, 0.25
mm ×30 m) using n-dodecane as an internal standard:
nitrostyrene, m/z = 149.0 (149.1 calcd); aminostyrene, m/z
= 119.0 (119.1 calcd); ethylnitrobenzene, m/z = 151.0 (151.1
calcd); ethylaniline, 121.0 (121.1 calcd); N-methylaminostyr-
ene, m/z = 133.0 (133.1 calcd); N-methylenestyrylamine, m/z
= 131.0 (131.1 calcd.). Gaseous products were analyzed by a
thermal conductivity detection gas chromatograph (TCD−GC,
Shimadzu, GC-14B) equipped with a packed column (GL
Sciences, Active Carbon). We confirmed that material balance
was almost equal to 1 (0.97−1.03) in the typical reaction
condition. Conversion of the reactant and selectivity to a
certain product were defined as follows: conversion (%) =
[(initial concentration of the reactant) − (concentration of the
reactant after reaction)]/(initial concentration of the reactant)
× 100, selectivity (%) = (concentration of the product)/(total
concentration of the all products) × 100.
2.3. Characterization. The crystal structure of supported
metal and intermetallic particles was examined by powder X-ray
diffraction (XRD) with a Rigaku RINT2400 using an X-ray
source of Cu Kα. Difference XRD patterns were obtained by
subtraction of the pattern of silica support from those of the
prepared catalysts so that the baseline was flattened. Trans-
mission electron microscopy (TEM) was conducted on a JEOL
JEM-2010F at the accelerating voltage of 200 kV. To prepare
the TEM specimen, all samples were sonicated in tetrachloro-
methane and then dispersed on a copper or molybdenum grid
supported by an ultrathin carbon film. Selected area electron
diffraction (SAED) patterns were acquired with a 20-μm-sized
area aperture and a camera distance of 30 cm. The camera
6
73 K for 1 h. After the calcination, the catalyst was reduced
−1
under flowing H (60 mL·min , 99.9995%, Taiyo Nippon
2
Sanso) at 403 K for 1 h and then at 673 K for 2 h. The catalyst
was cooled to room temperature with flowing helium and kept
in a drying desiccator.
In a similar manner, Rh/SiO (Rh 3 wt %) was also prepared
2
by using an aqueous solution of Rh(NO)(NO ) . Intermetallic
3
3
catalysts, Pd−M/SiO (M = Cu, Ga, Pb, and Zn) and Rh−M′/
2
SiO (M′ = Fe, Ni, Pb, Sb, Sn, and Ti), were prepared by a
2
successive impregnation with Pd/SiO and Rh/SiO , respec-
2
2
tively. In the case of Pd−Cu (1:1), an aqueous solution of
Cu(NO ) ·3H O (99%, Kanto) was added to Pd/SiO so that
the atomic ratio of Pd/Cu was adjusted to 1. The mixture was
sealed by a piece of plastic film overnight at room temperature.
It was then dried over a boiling water bath with stirring,
followed by reduction under flowing H (99.9995%, Taiyo
Nippon Sanso) at 1073 K for 1 h.
3
2
2
2
2
Other intermetallic catalysts were prepared in a similar way at
73, 873, or 1073 K of reduction temperature using a specific
7
amount of the second metal precursor: Ga(NO ) ·8H O,
3
3
2
Pb(NO ) Zn(NO ) ·6H O, Fe(NO ) ·9H O, Ni(NO ) ·
3
2
3
2
2
3
3
2
3 2
6
H O, SbCl , SnCl ·2H O, and TiCl (THF) . In the case of
2 3 2 2 4 2
TiCl (THF) , THF was used as the solvent instead of water.
Other intermetallic catalysts, Pd−M/SiO (M = Bi, Fe, and Sn)
and RhBi/SiO , were prepared by coimpregnation with mixed
aqueous solution of Pd(NO ) or Rh(NO)(NO ) and the
second metal precursor: Bi(NO ) ·5H O, Fe(NO ) ·9H O,
4
2
2
2
3
2
3 3
3
3
2
3 3
2
1
442
dx.doi.org/10.1021/cs500082g | ACS Catal. 2014, 4, 1441−1450

ACS Catalysis
Research Article
Figure 1. XRD patterns of the prepared (a) Pd- and (b) Rh-based intermetallic compounds supported on silica. Figure in parentheses indicates the
crystallite size estimated by Scherrer’s equation.
constant was calibrated by using an authentic Pt/Al O (Pt: 3
XRD. Only the preparation of Pd−Ga (13:5) resulted in the
2
3
wt %) for determination of reflection indices. Fourier
transformed infrared (FT-IR) spectra of adsorbed CO were
measured with a JASCO FT/IR-430 spectrometer in trans-
formation of a mixture of Pd Ga and monometallic Pd. The
13
5
crystallite sizes of the intermetallic particles in the catalysts were
estimated by the Scherrer equation using the most intense
diffraction peaks. The sizes were typically within 4−6 nm, with
the exception of Pd Pb (13 nm). Notably, most of the
−2
mission mode. A self-supporting wafer (∼20 mg·cm ) of
catalyst was placed in a quartz cell with CaF windows and
2
13
9
attached to a glass circulation system. The catalyst was reduced
at 723 K for 0.5 h with 20 kPa of circulating hydrogen through
a cold trap kept at 77 K. Then the catalyst was evacuated at the
same temperature for 0.5 h, followed by cooling to room
prepared intermetallic catalysts exhibited similar particle sizes.
This similarity is partly because most of the intermetallics were
prepared by successive impregnation, in which the resulting
particle size should depend on the size of the parent
monometallic particle. Figure 2 shows the TEM image and
temperature. In the case of monometallic Pd/SiO , the
2
temperature of the reduction and evacuation was kept at 573
K. CO (2.0 kPa) was introduced for 30 min and then
−1
evacuated. Spectra were recorded with 1 cm of resolution
before and during the evacuation at intervals of 5 min. X-ray
photoelectron spectra (XPS) of the intermetallics were
recorded with an ULVAC PHI 5000 VersaProbe spectrometer.
The catalyst was pressed into a pellet and placed into a quartz
reactor, where it was reduced under flowing hydrogen (50 mL
−1
min ) at 673 K for 0.5 h prior to the measurement. Spectra
were obtained with an Al Kα X-ray source using C 1s as a
reference for binding energy.
3. RESULTS AND DISCUSSION
3.1. Characterizations of Catalysts. Figure 1 shows the
XRD patterns of the prepared (a) Pd- and (b) Rh-based
catalysts. In the case of Pd−Bi (3:1), diffraction peaks assigned
to the intermetallic phase Pd Bi (2θ = 38.3°, 39.6°, and 41.0°)
3
were observed, and no residual monometallic Pd phase was
detected. Similarly, in most cases, the desired intermetallic or
alloy phases were formed as single phases with composition
ratios identical to the loaded ratios. The catalyst of Pd−Pb
(
1:1) exclusively gave Pd Pb phase, indicating that a portion
13 9
of Pb atoms did not participate in the intermetallic formation
and remained as small clusters that were not detectable by
Figure 2. TEM image of Pd Pb /SiO . Inset shows size distribution
13
9
2
of intermetallic nanoparticles.
1
443
dx.doi.org/10.1021/cs500082g | ACS Catal. 2014, 4, 1441−1450

ACS Catalysis
Research Article
size distribution of Pd Pb . The particles were mainly 8−13
Table 1. CTH Reaction of Nitrobenzene over Pd- and Rh-
Based Intermetallic Compounds Supported on SiO2
1
3
9
a
nm in size, and the particles of 13 nm exhibited the highest
population. This result agrees with the estimation based on the
Scherrer equation. No small particles with a diameter <3 nm
were observed. Similar results were observed for Rh-based
2
intermetallic, RhPb , as shown in Figure 3. The sizes of RhPb2
conversion (%)
selectivity (%)
b
entry
1
catalyst
D/Nm
nitrobenzene MC aniline toluene
Pd−M/SiO2
Pd
6
49
54
>99
>99
c
2
0
3
Pd Ga
5
14
27
35
90
99
1
3
d
4
c
Pd Ga3
5
6
13
>99
>99
5
<1
7
8
9
PdFe
PdZn
PdGa
PdCu
6
6
4
6
8
6
8
10
16
11
17
6
>99
>99
87
>99
>99
98
7
6
1
1
1
0
1
2
5
>99
65
>99
>99
Pd Sn
3
5
Pd Bi
3
0
0
Rh−M′/SiO2
1
1
1
1
1
1
1
3
4
5
6
7
8
9
Rh
4
6
0
63
58
92
94
15
13
13
0
<1
5
Rh Pb2
87
87
>99
>99
3
c
RhPb
RhPb2
RhSb
RhBi
6
6
5
6
0
0
5
0
92
>99
89
>99
>99
c
Figure 3. TEM image of RhPb /SiO . Left and right insets show size
20
21
22
2
2
c
,
e
distribution of intermetallic nanoparticles and SAED pattern of a single
particle, respectively.
Rh Sn
5
3
10
0
0
0
>99
2
2
3
Rh Ti
3
a
Reaction conditions: catalyst, 150 mg (monometallic Pd and Rh) or
particles were predominantly within 3−6 nm, which is roughly
consistent with the estimated values. The sample contained
very few particles that were hardly detectable by XRD (<2 nm).
2
50 mg (Pd−M and Rh−M′ intermetallics); solvent, 5 mL
b
(methanol); temp, 343 K; time, 1 h. D: crystallite size estimated by
c
d
Scherrer’s equation. Without MC. The catalyst contains mono-
metallic Pd phase. THF was used as a solvent instead of methanol.
e
In the case of RhPb , its XRD pattern appeared somewhat
2
broad; we therefore further characterized this sample. Selected-
area electron diffraction was performed on a single RhPb2
particle with a size of ∼6 nm (Figure 3, upper-right inset).
Moreover, the CTH reaction did not proceed in the absence of
MC (entries 2 and 5). These results indicate that the CTH
reaction from MC to nitrobenzene occurred; however, the
conversions of MC were higher than those of nitrobenzene in
all cases (entries 1−11). The observed higher conversions of
MC imply an involvement of the simple dehydrogenation of
Diffraction spots characteristic to RhPb (space group I4/mcm)
2
were observed, which clearly indicates that the particle was a
single crystal of intermetallic RhPb . The observed 211 and 310
2
diffractions correspond to the peaks at 33.7° and 42.8° in the
XRD pattern, respectively.
MC to form toluene and H . In some cases (entries 3, 9, and
2
3
.2. CTH Reaction of Nitrobenzene. The prepared
11), small amounts of N-methylaniline and methylcyclohexane
were formed as byproducts from nitrobenzene and MC,
respectively. The formation of N-methylaniline was not
observed when tetrahydrofuran (THF) was used as a solvent
(data not shown), suggesting that methanol was involved in the
methylation of aniline. The presence of methylcyclohexane
indicates that the self-CTH reaction of MC into toluene and
methylcyclohexane occurred.
catalysts were then applied to CTH reactions. First, several
hydrogen donors were examined in the CTH reaction of
nitrobenzene to aniline over Pd/SiO to identify an appropriate
hydrogen donor. The reaction did not proceed when methanol
or 2-propanol was used as a hydrogen donor. In contrast, a
moderate conversion (49%) without the evolution of H was
2
2
achieved when 4-methyl-1-cyclohexene (MC) was used.
Although formic acid (95%) and ammonium formate (70%)
Rh catalysts exhibited catalytic properties that differed from
gave higher conversions than MC, H was detected in the gas
those of Pd catalysts. Rh/SiO was not active toward the CTH
2
2
phase. The formation of H suggests that hydrogenation by H2
reaction (entry 13). This result is surprising because Rh has
2
17
also occurred in addition to the CTH reaction. Therefore, we
selected MC as an appropriate hydrogen donor in the present
study.
CTH reactions between nitrobenzene and MC using various
catalysts were performed to evaluate their catalytic abilities
toward the hydrogenation of a nitro group (Table 1). In the
been reported to be an effective catalyst for CTH reactions. A
possible reason for the low activity of Rh/SiO is the presence
2
of nitro groups, which can strongly bind to monometallic Rh
sites. To investigate the effect of nitro groups, we performed
the self-CTH reaction of MC in the presence and absence of
nitrobenzene (Supporting Information, Table S1). In the case
case of Pd catalysts (except Pd Bi), nitrobenzene and MC were
selectively converted into aniline and toluene, respectively.
of Pd/SiO , although MC conversion decreased by half, the
reaction proceeded even in the presence of nitrobenzene. Rh/
3
2
1
444
dx.doi.org/10.1021/cs500082g | ACS Catal. 2014, 4, 1441−1450

ACS Catalysis
Research Article
SiO gave a MC conversion similar to that of Pd/SiO in the
1.31 Å; Rh, 1.35 Å; Ga, 1.26 Å; Pb, 1.47 Å). The formation rate
of the product, here aniline, was used as the scale of the
catalytic activity. Shorter interatomic distances were correlated
with higher reaction rates, which indicates that the geometric
factor of the catalysts affects their catalytic activity. This
correlation is consistent with the common hypothesis that
longer Pd−Pd distances should increase the energy barrier of
C−H bond dissociation. A weak negative correlation was also
observed between the catalytic activity and the electronegativity
of the second metal, that is, electron-rich Pd affords greater
catalytic activity. A possible interpretation of this result is that
electron-rich Pd atoms enhance back-donation to antibonding
orbitals of C−H bond, which facilitates the dissociation of the
bond.
2
2
absence of nitrobenzene, consistent with the previously
1
7
reported results. However, in contrast, the reaction did not
occur at all in the presence of nitrobenzene. These results
indicate that the nitro group fatally poisons the monometallic
Rh sites and thereby inhibits the activation of methanol.
Although the hydrogenation of nitrobenzene to aniline
proceeded over Rh-based intermetallic catalysts, the dehydro-
genation of MC to toluene scarcely occurred (entries 14, 16−
1
9, 22, and 23). Moreover, the hydrogenation of nitrobenzene
occurred even in the absence of MC (entries 15 and 20), with
the nitrobenzene conversion almost maintained. However, no
reaction occurred when THF was used as a solvent instead of
methanol (entry 21). These results suggest that methanol
functions as a hydrogen donor over Rh-based intermetallic
catalysts, whereas MC does not.
A similar study was also performed on Rh-based intermetallic
catalysts, as shown in Figure 5. In contrast to Pd series, no
The conversion of nitrobenzene on Pd- and Rh-based
intermetallic catalysts differed, although the crystallite sizes of
these catalysts were similar. We subsequently investigated the
difference in their catalytic activities in terms of geometric and
electronic effects. Figure 4 shows the dependences of their
Figure 5. Relationship between the formation rate of aniline and (a)
the interatomic distance of the nearest Rh−Rh or Rh−M′ and (b) the
electronegativity of M′.
Figure 4. Dependences of the formation rate of aniline on (a) the
interatomic distance of the nearest Pd−Pd or Pd−M and on (b) the
electronegativity of M.
apparent relationship was observed between the catalytic
activity and interatomic distance (Figure 5a); however, a
positive correlation was observed with respect to the electro-
negativity of the second metal (Figure 5b). As previously
discussed, Rh catalysts require O−H activation of methanol to
catalytic activities on (a) the interatomic distance between the
nearest Pd−Pd or Pd−M and on (b) the electronegativity of
the second metal. In the present study, we employed Allred−
22,23
trigger the reaction. According to the literature,
O−H
dissociation of methanol in a homolytic fashion on nonpolar
metallic sites is highly endothermic and therefore kinetically
unfavorable; however, this dissociation is essentially thermo-
1
9,20
Rochow electronegativity,
which has been regarded as a
21
valid scale to discuss electron distribution in a compound.
24
25,26
Allred−Rochow electronegativity depends on the surface
neutral or exothermic
on polar sites, such as oxide ions in
2
electric field on an atom (eZ /r , where e is the elementary
a heterolytic fashion. Actually, in the case of a reported CTH
eff
27
charge, Z is the effective nuclear charge, and r is the covalent
system of styrene and 2-propanol over Cu/Al O , the authors
2 3
eff
2
0,21
radius)
and, hence, considers the difference in covalent
proposed that O−H activation of 2-propanol occurs on the
radii. Especially in the case of intermetallic compounds, it is
very important to take the difference in covalent radii into
account because the difference is not so small (for example: Pd,
basic site of Al O but not on metallic Cu. In our catalytic
2
3
system, a positive kinetic isotope effect (k /k = 7.2) was
H
D
observed when deuterated methanol (CH OD) was used
3
1
445
dx.doi.org/10.1021/cs500082g | ACS Catal. 2014, 4, 1441−1450

ACS Catalysis
Research Article
instead of nondeuterated methanol (CH OH) in CTH reaction
reported for monometallic Pd and Pt catalysts. The conversions
of Pd-based intermetallic catalysts were lower than those of Pd/
3
over Rh Pb , which indicates that the rate-determining step is
3
2
O−H dissociation of methanol. Therefore, the increase in the
catalytic activity with increasing electronegativity is attributed
to the enhancement of the activation rate of methanol, which is
probably due to the formation of polar sites via the
incorporation of electronegative atoms. Thus, the significant
factor to determine the catalytic activity was different between
the Pd and the Rh catalysts.
SiO catalyst. Pd-based intermetallic catalysts were divided into
2
two categories in terms of their product distribution: Pd Ga ,
1
3
5
Pd Ga , PdGa, and PdZn selectively gave EN (entries 2, 3, 8,
5
3
and 9), whereas Pd Pb , Pd Bi, and PdFe dominantly gave AS
13
9
3
(entries 5, 7, and 10). Notably, Pd Pb exhibited high
13
9
chemoselectivity to AS (93%, entry 5). In the absence of
MC, the CTH reaction of NS did not occur over the EN-
selective catalyst Pd Ga (entry 4), similar to the previously
3.3. CTH Reaction of NS. We subsequently performed
5
3
CTH reactions of NS using the prepared catalysts (Table 2).
discussed case of nitrobenzene hydrogenation. Surprisingly,
however, the AS-selective catalyst Pd Pb exhibited a similar
13
9
Table 2. CTH Reaction of p-Nitrostyrene (NS) over Pd- and
Rh-Based Intermetallic Compounds Supported on SiO2
catalytic performance in the absence of MC (entry 6). The high
selectivity exhibited by Pd Pb was maintained even when the
a
13
9
conversion reached 99% (Supporting Information, Figure S1).
Rh/SiO exhibited very low catalytic activity and yielded EN
2
selectively (entry 12). Rh-based intermetallic catalysts, in
contrast, gave AS with high selectivities (entries 13, 17−21).
The intermetallic catalysts with Fe and Ti did not catalyze the
CTH reaction (entries 22 and 23). The catalytic activity of Rh
catalysts depended on the second metal element, as was the
case with nitrobenzene hydrogenation. Notably, RhPb2
exhibited high NS conversion (94%) and chemoselectivity
toward AS (93%) and was the most effective intermetallic
catalyst for the chemoselective hydrogenation of NS to AS
selectivity (%)
b
entry
catalyst
D/Nm
conv. of NS (%)
AS
EN
EA
Pd−M/SiO2
1
2
Pd
Pd Ga
6
6
100
22
14
<1
10
10
10
7
0
63
>99
82
26
0
b
0
0
1
3
5
(entry 13). The addition of MC to the reaction with RhPb did
2
c
3
Pd Ga3
18
0
5
not affect its catalytic performance (entry 14). Moreover, the
self-CTH reaction of MC in the absence of NS did not occur
(data not shown). These results indicate that MC is a spectator
c,d
4
0
>99
5
5
Pd Pb
9
13
93
93
43
0
2
13
d
6
4
3
and is not activated by RhPb . Furthermore, the CTH reaction
2
7
8
9
0
1
Pd Bi
3
6
6
4
6
8
27
25
0
of styrene with methanol did not proceed over RhPb (entry
2
PdZn
PdGa
PdFe
>99
>99
45
1
5), which revealed that the vinyl group was not hydrogenated
5
0
0
by hydrogen species derived from methanol. On the basis of the
results, we conclude that the difference in the actual hydrogen
donor determines which functional group of NS is hydro-
genated: the CTH reaction with MC gave EN, and that with
1
1
2
55
0
Pd Sn
3
<1
d
Rh−M′/SiO2
1
1
2
Rh
4
6
<1
94
95
0
0
93
95
>99
6
0
1
1
methanol gave AS. The catalytic performance of RhPb was also
2
3
e
RhPb2
examined in the simple hydrogenation of NS with H . High
2
14
0
selectivity toward AS (91%) was obtained even with H (entry
2
f
15
1
6), which revealed the potential applicability of the
g
16
36
38
15
14
6
91
94
5
0
0
0
0
5
4
0
0
0
0
0
intermetallic compound for chemoselective hydrogenation
using H2.
1
7
8
9
0
1
2
3
RhPb
6
6
6
5
5
5
3
1
1
2
2
2
2
Rh Pb2
3
>99
>99
>99
95
To investigate the further applicability of the catalysts, we
performed the chemoselective hydrogenation of various nitro
aromatic compounds containing olefinic bonds using the most
selective Pd Pb and RhPb catalysts, as shown in Table 3. The
RhBi
Rh Sn
2
RhSb
RhFe
3
13
9
2
0
reaction involving Pd Pb catalyst was performed at 363 K to
13
9
Rh Ti
3
0
increase the reaction rate. As shown in Supporting Information
Figure S1, NS was converted into AS almost quantitatively in 6
a
Reaction conditions: catalyst, 150 mg (monometallic Pd and Rh) or
50 mg (Pd−M and Rh−M′ intermetallics); solvent, 5 mL
methanol); temperature, 343 K; time, 1 h. D: crystallite size
estimated by Scherrer’s equation. The catalyst contains monometallic
Pd phase. Without MC. With MC. Styrene was used as a substrate.
H (60 mL min ) and THF were used as a hydrogen donor and a
2
(
h (entry 1, 96% yield). RhPb exhibited much higher catalytic
b
2
activity (100% conversion within 1 h), whereas the selectivity
was slightly lower than that of Pd Pb (entry 2). Similar results
c
d
e
f
13
9
g
−1
were obtained when 3-nitrostyrene was used as a substrate
entries 3 and 5). These results are consistent with the result
2
(
solvent, respectively, with 0.5 h of reaction time.
that styrene was not hydrogenated under these experimental
conditions, as previously discussed. Since the hydrogen species
formed on these catalysts do not hydrogenate the vinyl group,
its hydrogenation should be suppressed, irrespective of the
vinyl group’s position. After the reaction, these catalysts were
separated from the reaction mixture via simple decantation and
were reused without any loss of catalytic activity or selectivity
(entries 4 and 6). Other nitroaromatic compounds, such as 4-
nitrostilbene and nitroindole isomers, were also tested using
MC was used as a hydrogen donor only with Pd catalysts. In
this reaction, NS undergoes the hydrogenation of its nitro or
vinyl group to form p-aminostyrene (AS) or p-ethylnitroben-
zene (EN), respectively. Overhydrogenation of these products
also occurs to yield p-ethylaniline (EA). Pd/SiO gave EN with
2
moderate selectivity, and overhydrogenation to EA also
occurred (entry 1). These results are consistent with those
1
446
dx.doi.org/10.1021/cs500082g | ACS Catal. 2014, 4, 1441−1450

ACS Catalysis
Research Article
Table 3. Chemoselective Hydrogenation of Various Nitroaromatic Compounds Having Olefinic Bonds into Amino Derivatives
a
Using Pd Pb /SiO and RhPb /SiO
13
9
2
2
2
a
b
Reaction conditions: catalyst, 250 mg; solvent, 5 mL (methanol); atmosphere, 1 atm Ar. MC was not used in these reactions. Solvent, 15 mL
methanol).
(
RhPb as a catalyst and were selectively converted into the
2
corresponding amino derivatives within 1 h (entries 7−10). As
was the case of nitrostyrenes, no difference in selectivity was
observed in the case of 6-, 5-, and 4-nitroindoles. Thus, these
catalysts have broad applicability for the chemoselective
hydrogenation of nitroaromatic compounds.
We subsequently studied the geometric and electronic effects
on the catalytic performance to understand the factors that
govern the catalytic activity and product distribution. Figure 6
shows the dependence of the formation rate of the product on
the interatomic distance of (a) Pd−Pd in Pd catalysts and (b)
Rh−Rh in Rh catalysts. Negative correlations with interatomic
distance were observed among the EN-selective catalysts
(
Figure 6a, circle). This trend is consistent with the results of
the hydrogenation of nitrobenzene over Pd catalysts (Figure
a). Because MC functions as a hydrogen donor over EN
4
selective catalysts, the activation of MC appears to determine
the reaction rate in a manner similar to the previously discussed
hydrogenation of nitrobenzene. In contrast, no apparent
dependence was observed with AS-dominant Pd catalysts
(
Figure 6a, triangles) and Rh catalysts (Figure 6b).
Figure 7 shows the relationship between the formation rates
of the products of (a) Pd and (b) Rh catalysts and the
electronegativity of the second metal. A weak negative
correlation was observed between the electronegativity of the
second metal and the formation rate of EN (Figure 7a, circles),
whereas that of AS increased as the electronegativity increased
in both catalysts (Figures 7a and b, triangles). For Pd catalysts,
the electronegativity of the second metal appears to play a
significant role to determine the product distribution: a higher
electronegativity makes the hydrogenation of the nitro group,
which corresponds to the methanol-mediated CTH reaction,
Figure 6. Dependences of the formation rate of the product on
interatomic distance of (a) Pd−Pd of the Pd catalysts and (b) Rh−Rh
of the Rh ones.
more favorable than that of the vinyl group. As discussed in
section 3.2, the incorporation of the electronegative atoms
suppresses C−H activation of MC but enhances O−H
1
447
dx.doi.org/10.1021/cs500082g | ACS Catal. 2014, 4, 1441−1450

ACS Catalysis
Research Article
prominently observed at 38.8° and 39.9° in the XRD pattern
of Pd Pb (Figure 1a). Large Pd ensembles disappeared via the
13
9
incorporation of Pb atoms, and these atoms are coadjacent at
the atomic level. Thus, the intermetallic compound appears to
provide the ideal platform for the chemoselective hydro-
genation of NS.
To characterize the actual geometric and electronic states of
the catalysts, we performed an FT-IR study of adsorbed CO
molecules. Figure 9 shows the FT-IR spectra of CO adsorbed
Figure 9. FT-IR spectra of CO adsorbed on Pd/SiO and Pd Pb /
2
13
9
SiO . Assignments of the peaks are illustrated: linear, bridged, and 3-
Figure 7. Relationship between the formation rate of the product and
the electronegativity of the second metal elements of the (a) Pd and
2
fold coordination. Inset shows time course of the peak position
assigned to the vibration of linear-type CO during evacuation at room
temperature.
(
b) Rh catalsysts.
activation of methanol. Therefore, the results obtained using Pd
catalysts (Figure 7a) are attributed to a change in the activation
rate of each hydrogen donor. This concept can also be applied
to the change in the catalytic activity of Rh catalysts. Notably,
because the crystallite size of Pd Pb is larger than that of other
onto Pd and Pd Pb . The spectrum of monometallic Pd
13
9
showed several peaks (at 2077, 1990−1930, and 1920−1870
−1
cm ), which are assigned to CO stretching vibrations of CO
in linear, bridged, and 3-fold coordination modes, respec-
1
3
9
28
tively. A higher coordination number resulted in a lower
intermetallics, its true catalytic performance (specifically, the
turnover frequency based on the number of exposed Pd atoms)
appears to be much greater than those of other Pd catalysts.
However, if so, the relationship between the catalytic activity
and electronegativity is not changed because Pd Pb is the
wavenumber because of the enhancement of π back-donation
to antibonding orbitals of CO bond. A control experiment
revealed that no peak was observed when CO was introduced
onto Pb/SiO (data not shown), thereby indicating that CO
2
1
3
9
was not adsorbed onto Pb atoms. Bridged and 3-fold
adsorptions were dominant on monometallic Pd, whereas
linear adsorption was observed mainly on Pd Pb . The drastic
most active investigated catalyst toward AS formation and is
almost inactive toward EN formation.
13
9
Figure 8a describes the crystal structure of Pd Pb , where Pd
1
3
9
decrease in the multicoordinated species can be interpreted to
result from the disappearance of large Pd ensembles as a
consequence of the formation of the intermetallic phase with
and Pb atoms are arranged homogeneously and regularly.
Typical facets of Pd Pb , (−223) and (623), are also illustrated
1
3
9
in Figure 8b; lattice diffractions of these facets were
29
Pb atoms. Therefore, the obtained results are consistent with
the intermetallic phase being exposed on the catalyst surface.
The inset in Figure 9 shows the time course of the peak
positions assigned to the linear-type CO during evacuation at
room temperature. The peak positions changed and plateaued
after 20 min of evacuation. This change is attributed to
diminished dipole−dipole coupling between adsorbed CO
30
molecules at high CO coverage. Therefore, the final peak
positions at sufficiently low CO coverage should reflect the true
electronic states of Pd. The peak position of linearly adsorbed
CO in the spectrum of Pd Pb appeared at higher wave-
13
9
numbers compared with that in the spectrum of Pd after
saturation, indicating a lower electronic density of Pd atoms in
Pd Pb compared with those in monometallic Pd. This result
13
9
is consistent with the results reported in the literature, in which
an XPS analysis of the electronic state of Pd in Pd−Pb
Figure 8. (a) Crystal structure of intermetallic Pd Pb with the single
unit cell in a monoclinic C2/c space group (PDF: #04-007-1453) and
13
9
intermetallics supported on SiO revealed that Pd became
2
(
b) its (−223) and (623) planes.
electron-deficient upon the formation of Pd−Pb intermetallic
1
448
dx.doi.org/10.1021/cs500082g | ACS Catal. 2014, 4, 1441−1450

ACS Catalysis
Research Article
3
1
phases. We also performed similar XPS measurements in the
present study on monometallic Pd and Pd−Pb intermetallics.
The peaks of Pd 3d and 3d emissions were shifted by 0.6
ASSOCIATED CONTENT
Supporting Information
Results of the self-CTH reaction of MC, time course of
conversion and selectivity in CTH reaction of NS, XPS. This
material is available free of charge via the Internet at http://
pubs.acs.org.
■
*
S
5
/2
3/2
and 0.5 eV to higher binding energy, respectively, by the
formation of Pd−Pb intermetallics (Supporting Information,
Figure S2), which confirms that the formation of an
intermetallic phase makes Pd electron-deficient. These results
support the hypothesis that a charge transfer from Pd to Pb
occurs via the formation of the intermetallics; hence, the polar
sites are, indeed, constructed.
In this study, Pb-containing catalysts exhibited good catalytic
performance. Pd−Pb catalysts that are so-called Lindlar
catalysts are well-known to be effective for the semi-
hydrogenation of alkynes to alkenes. The enhanced selectivity
induced by Pb modification is considered to arise from the
poisoning of the strong hydrogenation ability of Pd to be
milder so that overhydrogenation is suppressed. In our case, in
contrast, the role of Pb atoms is to provide new active Pd−Pb
sites that are chemically different from the parent Pd. Pd−Pb
catalysts are also used as effective catalysts in various aerobic
oxidation systems. The direct oxidative esterification of
AUTHOR INFORMATION
■
*
Corresponding Authors
Phone: +81-3-5734-2602. Fax: +81-3-5734-2758. E-mail:
furukawa.s.af@m.titech.ac.jp.
Phone: +81-3-5734-3532. Fax: +81-3-5734-2758. E-mail:
komatsu.t.ad@m.titech.ac.jp.
*
32
Notes
Notes. The authors declare no competing financial interest.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by JSPS KAKENHI Grant No.
23360353. We thank the Center for Advanced Materials
Analysis Tokyo Institute of Technology for their assistance in
TEM analysis.
methacrolein to methyl methacrylate (MMA) using Pd Pb
3
33
has been commercialized as an industrial process. In this
system, Pb inhibits undesired side reactions, such as the
decarbonylation of methacrolein to enhance the product
selectivity. We recently discovered that Pd Pb exhibits much
higher catalytic activity than monometallic Pd in the oxidative
dehydrogenation of an amine to an imine. A kinetic study
revealed that Pd Pb phase accelerated the desorption of imine,
which was the late-determining step. Thus, although Pd−Pb
catalysts exhibit unique catalytic properties in various catalytic
systems, the roles of Pb in the reported Pd−Pb systems differ
among the systems and from that in the present study.
REFERENCES
■
3
(1) Downing, R. S.; Kunkeler, P. J.; vanBekkum, H. Catal. Today
1997, 37, 121−136.
(2) Suchy, M.; Winternitz, P.; Zeller, M. World Patent 91/00278,
1991.
3) Burawoy, A.; Critchley, J. P. Tetrahedron 1959, 5, 340−351.
4) Kovar, F.; Armond, F. E. U.S. Patent 3,975,444, 1976.
5) Butera, J.; Bagli, J. World Patent 91/09023, 1991.
34
3
(
(
(
(
6) Rylander, P. N. Catalytic Hydrogenation in Organic Synthesis;
Academic Press: New York, 1979.
(
(
7) Corma, A.; Serna, P. Science 2006, 313, 332−334.
8) Shimizu, K.; Miyamoto, Y.; Kawasaki, T.; Tanji, T.; Tai, Y. J. Phys.
CONCLUSION
■
Chem. C 2009, 113, 17803−17810.
In this study, a series of Pd- and Rh-based intermetallic
compounds supported on silica were prepared, and their
catalytic properties in the CTH reactions of nitrobenzene and
NS were investigated. The formation of the intermetallic
compounds makes an ideal platform for the chemoselective
hydrogenation of NS: two metal elements are coadjacent at the
atomic level on the surface of the nanoparticles. An increase in
the electronegativity of the second metal element provides
polar sites and enhances the activation of methanol as a
hydrogen donor, which accelerates the hydrogenation of the
nitro group of NS and hence improves the yield of AS. Over Rh
catalysts, AS is selectively formed as a result of the absence of
MC activation ability. Pd Pb exhibits the highest chemo-
(9) Mitsudome, T.; Mikami, Y.; Matoba, M.; Mizugaki, T.; Jitsukawa,
K.; Kaneda, K. Angew. Chem., Int. Ed. 2012, 51, 136−139.
(10) Shimizu, K.; Miyamoto, Y.; Satsuma, A. J. Catal. 2010, 270, 86−
94.
(
(
11) Bullock, R. M. Chem.−Eur. J. 2004, 10, 2366−2374.
12) Westbrook, J. H.; Fleischer, R. L. Magnetic, Electrical and Optical
Properties and Applications of Intermetallic Compounds; Wiley; New
York, 2000.
(13) Galloway, E.; Armbruster, M.; Kovnir, K.; Tikhov, M. S.;
Lambert, R. M. J. Catal. 2009, 261, 60−65.
(14) Johnstone, R. A. W.; Wilby, A. H. Tetrahedron 1981, 37, 3667−
3670.
(15) Sala, R.; Doria, G.; Passarotti, C. Tetrahedron Lett. 1984, 25,
4
(
565−4568.
1
3
9
16) Tike, M. A.; Mahajani, V. V. Chem. Eng. J. 2006, 123, 31−41.
17) Cristina, M.; Oliveira, F. Appl. Catal., A 2007, 329, 7−15.
18) Johnstone, R. A. W.; Wilby, A. H.; Entwistle, I. D. Chem. Rev.
selectivity in the conversion of NS to AS (92%) among Pd
catalysts. In addition to NS, various nitroaromatic compounds
having olefinic bonds were selectively converted into the
(
(
1
985, 85, 129−170.
corresponding amino derivatives within 1 or 1.5 h using PdRh2.
(19) Allred, A. L. J. Inorg. Nuclear Chem. 1961, 17, 215−221.
(20) Allred, A. L.; Rochow, E. G. J. Inorg. Nuclear Chem. 1958, 5,
264−268.
Moreover, RhPb not only exhibits high selectivity toward AS
2
(
93%) but also exhibits the highest NS conversion (94%). This
catalyst also gives high selectivity toward AS (91%), even when
(21) Shriver, D. F.; Atkins, P. W. Inorganic Chemistry; 3rd ed.; Oxford
University Press: Oxford, 1999.
22) Greeley, J.; Mavrikakis, M. J. Am. Chem. Soc. 2002, 124, 7193−
201.
23) Desai, S. K.; Neurock, M.; Kourtakis, K. J. Phys. Chem. B 2002,
06, 2559−2568.
(24) Beste, A.; Mullins, D. R.; Overbury, S. H.; Harrison, R. J. Surf.
Sci. 2008, 602, 162−175.
H is used as a hydrogen source. Thus, the present study
2
(
7
(
1
demonstrates the excellence of these catalysts in chemoselective
hydrogenation and the validity of the strategy of employing
these intermetallic compounds. The obtained insights open the
opportunity to develop more efficient and well-structured
catalytic systems for selective conversion reactions.
1
449
dx.doi.org/10.1021/cs500082g | ACS Catal. 2014, 4, 1441−1450

ACS Catalysis
Research Article
(
25) Feng, G.; Huo, C. F.; Deng, C. M.; Huang, L.; Li, Y. W.; Wang,
J. G.; Jiao, H. J. J. Mol. Catal. A: Chem. 2009, 304, 58−64.
26) Branda, M. M.; Collins, S. E.; Castellani, N. J.; Baltanas, M. A.;
Bonivardi, A. L. J. Phys. Chem. B 2006, 110, 11847−11853.
27) Huang, L.; Zhu, Y. L.; Huo, C. F.; Zheng, H. Y.; Feng, G.;
Zhang, C. H.; Li, Y. W. J. Mol. Catal. A: Chem. 2008, 288, 109−115.
28) Kim, W. J.; Shin, E. W.; Kang, J. H.; Moon, S. H. Appl. Catal., A
003, 251, 305−313.
29) Wei, T.; Wang, J.; Goodman, D. W. J. Phys. Chem. C 2007, 111,
781−8788.
30) Primet, M.; Demenorval, L. C.; Fraissard, J.; Ito, T. J. Chem. Soc.
Faraday Trans. 1 1985, 81, 2867−−2874.
31) Hirano, T.; Kazahaya, Y.; Nakamura, A.; Miyao, T.; Naito, S.
(
(
(
2
(
8
(
(
Catal. Lett. 2007, 117, 73−78.
(
(
32) Lindlar, H. Helv. Chim. Acta 1952, 35, 446−456.
33) Yamamatsu, S.; Yamaguchi, T.; Yokota, K.; Nagano, O.; Chono,
M.; Aoshima, A. Catal. Surv. Asia 2010, 14, 124−131.
34) Furukawa, S.; Suga, A.; Komatsu, T. Chem. Commun. 2014, 50,
277−3280.
(
3
1
450
dx.doi.org/10.1021/cs500082g | ACS Catal. 2014, 4, 1441−1450