In situ synthesized Pd nanoparticles supported on B-MCM-41: An efficient catalyst for hydrogenation of nitroaromatics in supercritical carbon dioxide

Article abstract of DOI:10.1039/c2gc36160d

In situ synthesis of Pd nanoparticles supported on boron (B)-substituted MCM-41 (B-MCM-41) with Si/B ratio varying from 100 to 5 was carried out by hydrothermal method using H₃BO₃ as B source. The textural properties as well as thermal stability of the resultant material were investigated by XRD, TEM, FTIR and TG-DTA. Highly ordered materials were obtained depending on the Si/B ratio, which also influenced the particle size of Pd as well as dispersion. Pd/B-MCM-41 was a promising catalyst for the hydrogenation of nitrobenzene in supercritical carbon dioxide with exceptionally faster reaction rate [turnover frequency (TOF) = 5.2 × 10⁵ h^-1 (144 s^-1)] and high yield of aniline (100%). The observed reaction rate was strongly influenced by the Pd particle size related to Si/B ratio and physical properties of CO₂ such as pressure- and temperature-dependent solvent power. A comparison of catalytic activity with the Pd supported only on silica material of similar particle size inferred that the presence of even a small amount of B significantly changes the reaction rate from 70 (only Si) to 105 s^-1 (Si/B = 100). In addition, TOF of Pd/B-MCM-41 was high when compared with other Pd catalysts supported on Al-MCM-41 and Ga-MCM-41 obtained by a similar method, and follows the order: B (144 s^-1) > Ga (31.2 s^-1) > Al (10.2 s ^-1). The remarkable advantage of the present catalytic system involves low metal content (～1%), easy separation and it is successfully employed for the hydrogenation of substituted nitroaromatics, nitrile and phenol under mild reaction conditions. Furthermore, the catalyst was recyclable up to the 7th recycle without any loss of catalytic activity.

Full text of DOI:10.1039/c2gc36160d

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Green Chemistry
Cite this: Green Chem., 2012, 14, 3415
www.rsc.org/greenchem
PAPER
In situ synthesized Pd nanoparticles supported on B-MCM-41: an efficient
catalyst for hydrogenation of nitroaromatics in supercritical carbon dioxide†
Maya Chatterjee,* Takayuki Ishizaka, Toshishige Suzuki, Akira Suzuki and Hajime Kawanami*
Received 26th July 2012, Accepted 10th October 2012
DOI: 10.1039/c2gc36160d
In situ synthesis of Pd nanoparticles supported on boron (B)-substituted MCM-41 (B-MCM-41) with
Si/B ratio varying from 100 to 5 was carried out by hydrothermal method using H BO as B source. The
3
3
textural properties as well as thermal stability of the resultant material were investigated by XRD, TEM,
FTIR and TG-DTA. Highly ordered materials were obtained depending on the Si/B ratio, which also
influenced the particle size of Pd as well as dispersion. Pd/B-MCM-41 was a promising catalyst for the
hydrogenation of nitrobenzene in supercritical carbon dioxide with exceptionally faster reaction rate
5
−1
−1
[
turnover frequency (TOF) = 5.2 × 10 h (144 s )] and high yield of aniline (100%). The observed
reaction rate was strongly influenced by the Pd particle size related to Si/B ratio and physical properties of
CO such as pressure- and temperature-dependent solvent power. A comparison of catalytic activity with
2
the Pd supported only on silica material of similar particle size inferred that the presence of even a small
amount of B significantly changes the reaction rate from 70 (only Si) to 105 s−1 (Si/B = 100). In addition,
TOF of Pd/B-MCM-41 was high when compared with other Pd catalysts supported on Al-MCM-41 and
−
1
−1
Ga-MCM-41 obtained by a similar method, and follows the order: B (144 s ) > Ga (31.2 s ) > Al
(
10.2 s− ). The remarkable advantage of the present catalytic system involves low metal content (∼1%),
1
easy separation and it is successfully employed for the hydrogenation of substituted nitroaromatics, nitrile
and phenol under mild reaction conditions. Furthermore, the catalyst was recyclable up to the 7th recycle
without any loss of catalytic activity.
Introduction
as a very promising candidate as catalyst support. Generally, the
synthesis procedure of MCM-41 involves multidentate binding
of the silicate oligomers to the cationic surfactant, preferential
silicate polymerization in the interface region, and charge
A selective catalyst is a highly desirable candidate for the green-
ness of chemical transformation, because it improves the
efficiency of the reaction by lowering the energy input and by
avoiding the use of stoichiometric reagents. Thus, development
of an efficient catalyst for a heterogeneous system is one of the
challenging areas in catalysis.
Supported metal nanoparticles are potential heterogeneous
catalysts used for various types of reaction. Mesoporous
materials, which was discovered by Beck et al. in 1992, seem
to be ideal hosts for metal nanoparticles because of very high
surface areas to immobilize catalytically active species on, or the
ability to provide nano-size confinement inside the pore system.
The mesoporous molecular sieve MCM-41 (a member of the
M41S family) possesses a hexagonally arranged uniform pore
structure. The important characteristics of this novel material are
large BET surface area, high porosity, and controllable narrow
pore size distribution. Those characteristics manifest themselves
4
–7
density matching between the surfactant and the silicate.
1
Beginning with a relatively inert all-silica MCM-41, great
chemical and catalytic diversity may be generated by isomor-
phous substitution with trivalent cations in Si framework. The
substitution of Si by B is expected to modify structural properties
of MCM-41 and was first described by Oberhagemann et al.
using quaternary ammonium salts of different chain length as the
2
8
template. After that several authors described the synthesis and
characterization depending on B precursor (Na B O or H BO )
2
4
7
3
3
3
or using alkylamine as surfactant to obtain highly ordered
uniform hexagonal mesopores in the resultant material and stab-
9
–12
ility of B in the framework.
It was observed that, like Al, B
has been confirmed to occupy tetrahedral framework sites.
However, in comparison with Al-substituted MCM-41, less lit-
erature is available on the catalytic activity of B-containing
materials because of its small size and propensity for trigonal
coordination. Borosilicate molecular sieves are only mildly
acidic, which particularly make them more suitable for catalysts
Research Center for Compact Chemical System, AIST Tohoku, 4-2-1
Nigatake, Miyagino-ku 983-8551, Sendai, Japan.
E-mail: c-maya@aist.go.jp, h-kawanami@aist.go.jp;
Fax: +81 22 237 5388; Tel: +81 22 237 5213
13–17
demanding low acidity.
Supported Pd catalysts are widely used in a number of hydro-
genation reactions and in various organic syntheses due to their
†
1
Electronic supplementary information (ESI) available. See DOI:
0.1039/c2gc36160d
18
unique properties of absorbing large quantities of hydrogen.
This journal is © The Royal Society of Chemistry 2012
Green Chem., 2012, 14, 3415–3422 | 3415

View Article Online
Several methods have been described in the literature for the syn-
thesis of noble metal nanoparticles supported on mesoporous
1
9
materials, such as direct incorporation, incipient wetness
2
0–22
23
impregnation
photocatalytic reduction and gamma irradiation were also used
and ion exchange. In addition to those,
2
4
to prepare metal nanoparticles on MCM-41 support.
The present study describes the hydrothermal method of
in situ synthesis of Pd nanoparticles on B-MCM-41 using H BO3
3
as B source. The catalytic performance was investigated for the
hydrogenation of nitrobenzene in supercritical carbon dioxide
(
scCO ) under mild conditions and the results are compared with
2
Al- and Ga-supported Pd catalyst obtained by similar methods.
Results and discussion
Physicochemical characterization of the catalysts
Elemental analysis of the calcined product confirmed the Pd
loading of all the samples including Al and Ga-MCM-41 were
∼
1%. The colour of the as-synthesized B-MCM-41 was white
and changed to grey after incorporation of Pd (Pd/B-MCM-41),
which is taken as an initial indication of the presence of Pd in
the material. After calcination, the parent material (B-MCM-41)
remains white, while Pd/B-MCM-41 turns orange and
2
+
suggesting the migration of Pd to the surface, by reaction with
SiOH groups of defect sites. A similar result was reported by
Kawabata et al. on the synthesis of Pd-containing Al-MCM-41
25
2
+
and attributed to the migration of Pd ions from inside the
framework to the surface due to the insufficient substitution of
2
+
4+
3+
Pd ions for both Si and Al sites in the lattice. It has to be
mentioned that for Si/B ratio, the gel ratio has been considered
throughout the manuscript.
X-ray diffraction pattern of Pd/B-MCM-41
Fig. 1a exhibits the XRD pattern of Pd/B-MCM-41 with differ-
ent Si/B ratios. All materials show typical low angle (1 0 0)
reflections characteristic of MCM-41 along with the smaller
2
Bragg peaks. Beck et al. indexed those peaks for the hexagonal
unit cell and the cell parameter (a ) was calculated using the
0
formula a = 2d100/√3. It has to be mentioned that the quality of
0
the main peak varied significantly depending on the Si/B ratio.
For high Si/B ratios, the major peak was more sharp and appears
at 2θ = 2.120° along with three well-resolved peaks. However,
the scattering pattern of materials with low Si/B ratio exhibits a
major peak of decreased intensity with the absence of smaller
peaks originated from the ordered two-dimensional structures. It
Fig. 1 (a) Low-angle XRD pattern of Pd/B-MCM-41 with different Si/
B ratio. (b) Higher-angle XRD pattern of Pd/B-MCM-41 with different
Si/B ratio.
was observed that the unit cell parameter a slightly decreased
0
with increasing B content and corresponds well with the
Table 1 Physical properties of Pd/B-MCM-41
9
–12,26,27
literature
(Table 1). Furthermore, decrease in a0 of
Average Pd
particle size
(nm)
Weight loss of
template at
B-MCM-41 compared to the pure Si analogue is evidence for B
a
9
,10,28
in the silicate framework.
Si/B ratio
Only Si
d100 (nm)
A0 (nm)
120–320 °C
To detect the Pd nanoparticles, XRD measurement was also
performed at higher angle regions (2θ = 30° to 90°), and are dis-
played in Fig. 1b. Independent of the Si/B ratio of calcined
sample, five distinctive peaks at 2θ = 33.8, 42.0, 54.8, 60.7 and
4.6
4.2
4.1
4.0
3.8
5.31
4.85
4.73
4.62
4.38
—
38.6
35.1
33.4
30.2
28.5
1
2
00
5
10.9
19.5
21.5
22.6
10
5
7
(
1.4, corresponding to the diffraction lines of PdO (1 0 1),
1 1 0), (1 1 2), (1 0 3) and (2 1 1), respectively, appeared after
the incorporation of Pd in B-MCM-41. Particle sizes were
a
From XRD using Scherer’s equation.
3416 | Green Chem., 2012, 14, 3415–3422
This journal is © The Royal Society of Chemistry 2012

View Article Online
Fig. 2 TEM images along with particle size distribution of (a) Si/B =
00 and (b) Si/B = 5.
1
calculated from the XRD line broadening of the peak at 2θ =
2
9
3
3.8 using Scherer’s equation, which shows that there is a
change in the average particle size from 10.9 to 22.6 nm with
the change in Si/B ratio from 100 to 5 (Table 1). Moreover, after
hydrogenation the diffraction lines of PdO disappeared and a
Fig. 3 FTIR spectra of (a) only Si, (b) Si/B = 100, (c) Si/B = 25, (d)
Si/B = 10 and Si/B = 5.
0
new peak corresponding to Pd appeared at 2θ = 40.1.
1
1
380 cm− and the increased intensity of the band might be cor-
TEM of Pd/B-MCM-41
related with the increased B content of the materials. Generally,
the tetrahedral form of B is stable only with Na counterbalan-
The pore structure of MCM-41 and the presence of Pd nanoparti-
cles were directly visible through TEM observation. Fig. 2a and
b shows TEM images of Pd/B-MCM-41 with Si/B = 100 and Si/
B = 5, respectively, through the pore axis and their correspond-
ing particle size distributions. The image corresponding to Si/B
cing, but is unstable in the H form (when H BO was used as B
3
3
27
precursor). However, the spectra of Si/B = 5 developed a small
−1
shoulder (Fig. 3e) at 940 cm representative of tetrahedral B,
which was absent in the other B-containing materials. In the
pure Si material, instead of the band at 1380 cm− , a 960 cm
1
−1
=
100 exhibits the regular ordered hexagonal channel, character-
band due to silanol group typical of calcined MCM-41 material
was observed.
istics of MCM-41. However, increasing B content damaged
regular order. In addition, TEM also focused on Pd particles sup-
ported on B-MCM-41. In each case, Pd particles were almost
spherical in shape and distributed throughout the support matrix.
Calculation of particle size distribution (Fig. 2) confirmed
average particle size of 12.1 nm for low B-containing material
Thermogravimetric analysis
The thermal analysis pattern of Pd/B-MCM-41 was very similar
to that of Pd-containing Si material (ESI; Fig. 1s†). Total weight
loss occurred in three distinctive steps. The first step was associ-
ated with the loss of physically adsorbed water in the tempera-
ture range of room temperature to 120 °C along with a weight
loss of ∼2–3%, and related to the hydrophobicity of the material.
The second stage of weight loss corresponding to desorption and
decomposition of the template from the pores occurred at
120–320 °C. There was a distinctive difference depending on the
Si/B ratio. For example, weight loss in the described region is
35.1 and 28.5% for Si/B = 100 and Si/B = 5, respectively, indi-
cating that with increasing B content a larger part of the template
is removed at high temperature as template cation bonded to
(
Si/B = 100), whereas larger particles of average size = 23.1 nm
found in the high B-containing material (Si/B = 5). These obser-
vations correspond well with the results of XRD measurement.
FTIR spectra of Pd/B-MCM-41
Fig. 3 represents FTIR spectra of Pd/B-MCM-41 of different
Si/B ratios along with the Si-only material. FTIR spectra of cal-
cined Pd/B-MCM-41 show a series of bands in the region of
00–1400 cm− characteristic of SiO4 tetrahedra and their
1
5
modification after B substitution. The presence of an IR band at
1
framework and representing the tri- and tetra-coordinated B,
respectively. All B-containing samples developed a band at
380 and 940 cm− confirmed B substitution in the silicate
1
2
the siloxy group decomposes at low temperature, which is
30
confirmed from the analysis of Si-only material (Table 1).
Green Chem., 2012, 14, 3415–3422 | 3417
This journal is © The Royal Society of Chemistry 2012

View Article Online
Scheme 1 Possible reaction pathway of nitrobenzene hydrogenation.
However, it is equally likely that less template was incorporated
into the composite of high B-containing material. If there was
more B, less electrostatic attraction between the metal precursor
and surfactant results in less incorporation of CTAB in the resul-
tant material. Finally, the weight loss of 7–8% from 320 to
Fig. 4 Variation of CO pressure; reaction conditions: catalyst = 0.1 g,
2
4
55 °C can be assigned to coke calcination, and the weight loss
H
substrate = 2.0 g, P 2 = 2 MPa, temperature = 50 °C, time = 5 min.
of 3–4% above 455 °C should be the loss of silanol groups
present in the framework (dehydroxylation). These results corres-
3
1
pond well with those reported previously. For all the materials,
a sharp DTA peak was observed and attributed to the strong
endothermic process due to the combustion and removal of the
surfactant from pores of MCM-41 matrix.
Catalytic activity
Scheme 1 depicts one of the possible reaction paths of hydrogen-
32
ation of nitrobenzene in scCO as described in the literature.
2
Selective hydrogenation of nitrobenzene in scCO2 was con-
ducted in a batch reactor using Pd/B-MCM-41 catalyst.
For optimization of the reaction conditions, maximum conver-
sion and highest selectivity of aniline were set as targeted
parameters. Despite of the low structural order, Pd/B-MCM-41
Fig. 5 Variation of H
substrate = 2.0 g, PCO2 = 12 MPa, temperature = 50 °C, time = 5 min.
2
pressure; reaction conditions: catalyst = 0.1 g,
(Si/B = 5) was chosen as the catalyst because of high B content.
The CO and H pressure, temperature, and the reaction time
2
2
were considered as parameters for optimization.
Effect of H
2
pressure
H pressure is one of the important parameters to optimize the
2
18d,e
2
Variation of CO pressure
reaction condition in scCO2.
The dependence of TOF on H2
pressure was studied at the fixed CO pressure (12 MPa), reac-
2
Fig. 4 exhibits variation of reaction rate (TOF) with CO pressure
under the studied reaction condition (PH2 = 2 MPa, temperature
2
tion time (5 min) and temperature of 50 °C (Fig. 5). The TOF of
nitrobenzene hydrogenation was found to be very sensitive to the
H2 pressure. When the pressure was changed from 0.5 to
=
50 °C and reaction time = 5 min). Notably, ∼7 times increase
in TOF was observed with the change in CO pressure from 6 to
2
−1
−1
2.0 MPa, the TOF increased from 46.3 to 144 s without affect-
1
4 MPa. The highest TOF (144 s ) was obtained at 12 MPa of
ing the product selectivity, probably due to the enhanced concen-
CO pressure. Visual inspection of the phase behaviour (ESI;
2
tration of H in the system. Further increase of pressure to 4 MPa
2
Fig. 2s†) conducted in a view-cell under similar conditions
did not change the conversion, however, the selectivity of aniline
decreased. Thus, an optimum pressure of 2 MPa was chosen to
achieve the targeted conversion, which is lower than the pressure
required for Si-only material indicating a possible effect of the
presence of B in MCM-41.
revealed that the mixture of nitrobenzene, H , and CO trans-
2
2
forms from a two-phase system (nitrobenzene and CO –H ) to a
2
2
single homogeneous phase (nitrobenzene–CO –H ) with the
2
2
change in the CO pressure. In the lower pressure region of 6–
2
1
0 MPa, a biphasic system prevailed, and a single homogeneous
phase was obtained at or above 12 MPa. As the pressure
changed from 6 to 12 MPa, the reaction rate (TOF) increased
from 23.3 to 144 s− . This is attributed to the transition from
Effect of reaction temperature
1
bi-phase to single phase of substrate–CO –H2 system. Thus,
Generally, solvent properties of CO are strongly influenced by
2
2
elimination of the gas–liquid interface was crucial to achieve
the pressure and temperature. Thus, fine tuning of temperature
highest performance of Pd/B-MCM-41 as hydrogenation cata-
can also change the density of CO and consequently the solvent
2
lyst. Independent to CO pressures, aniline was the only product
detected.
strength. To explore the effect of temperature, the reaction was
conducted at 35 to 70 °C using Pd/B-MCM-41 of Si/B ratio 5
2
3418 | Green Chem., 2012, 14, 3415–3422
This journal is © The Royal Society of Chemistry 2012

View Article Online
Fig. 6 Temperature vs. TOF over Pd/B-MCM-41; reaction conditions:
catalyst = 0.1 g, substrate = 2.0 g, PCO2 = 12 MPa, PH2 = 2 MPa, time =
5
min.
and 100 with the reaction time fixed at 5 min. In each case, the
conversion of nitrobenzene increased with temperature. For
example Si/B = 5 shows a conversion of 62% at 35 °C, which
increased to 100% when the temperature rose to 50 °C at a fixed
pressure of CO (12 MPa). Taking into the account of low con-
2
version data at 35 °C, it might be concluded that although a
single phase was formed between nitrobenzene, CO and H , the
2
2
kinetic effect was in action to enhance the reaction rate rather
than phase behaviour. Furthermore, Fig. 6 shows changes in
reaction rate along with temperature, depending on Si/B ratio. As
the temperature changes from 35 to 50 °C, the rate of the reac-
tion (TOF) also increased. In addition, with the decrease in Si/B
Fig. 7 (a) Time dependence of nitrobenzene conversion using Pd/
B-MCM-41; reaction conditions: catalyst = 0.1 g, substrate = 2.0 g,
P
CO2 = 12 MPa, P 2
H
= 2 MPa, temperature = 50 °C. (b) Effect of particle
−1
size on reaction rate.
ratio from 100 to 5, TOF changes from 43 (35 °C) to 105 s
−1
(
50 °C) and 89 (35 °C) to 144 s (50 °C), respectively. Hence,
it might be concluded that (i) independent of the B content and
other related parameters like particle size, structural order, etc.,
increase in temperature increased the reaction rate and (ii) low
B-containing material exhibited lower TOF even at higher
temperature of 70 °C. Thus, an optimum temperature of 50 °C
was chosen for all experiments as the highest conversion was
achieved at that temperature within the reaction time of 5 min
using Si/B = 5 catalyst.
dependence of nitrobenzene hydrogenation rate on particle size
is shown in Fig. 7b, which indicates that the initial TOF calcu-
lated per surface Pd atom increases with the particle size. The
changes in particle size from ∼10 to 20 nm increased the TOF
−1
from 59 to 144 s . Hence, larger Pd particles were considered
to be more active than smaller ones, assuming that all the par-
ticles took part in the reaction. Although there was a sharp
change of TOF with particle size, it is difficult to comment on
the structure sensitivity because the studied range of particle size
was limited. A similar phenomenon was also reported previously
in the liquid-phase hydrogenation of nitrobenzene in methanol
Effect of reaction time
33
The effect of reaction time on the conversion of nitrobenzene
was evaluated on two Pd/B-MCM-41 catalysts with Si/B ratios
of 100 and 5 (Fig. 7a). In each case, the reaction was compara-
tively fast, but complete conversion can be achieved within the
shortest reaction time of 5 min only when Si/B = 5 was used.
For these two catalysts, Pd particle sizes are different (Table 1).
Thus, to explain these observations there are two primary factors
to be considered: (i) effect of Pd particle size and (ii) B content,
which controlled the structural order. It has to be mentioned that
only support material was inactive and, hence, catalytic activity
of the metal-containing catalysts arises unambiguously due to
the presence of active metal species present on the support. Gen-
erally, the variation in catalytic activity depends on the dis-
persion of metal. This factor is further influenced by B content
of the support material. (i) Effect of Pd particle size: the
over Pd/C at 50 °C. (ii) B content: presence of B incorporates
weak acidity on the support surface, as described by many
researchers, and it might change catalyst morphologies such as
metal dispersion, etc. We had compared the results between low
B content (Si/B = 100) and Si-only material. The rate was com-
−1
paratively higher for Si/B = 100 (TOF = 105 s ) rather than Si-
only catalysts (TOF = 70 s− ) although both of them contain Pd
particles of similar size (∼10 nm) and highly ordered structures.
On the basis of these results it might be concluded that even in
the presence of a small amount of B the reaction rate increased,
which implies that weak acid sites or hydrophobicity (related to
the B species) participate in the reaction process under the
studied reaction conditions. To check the role of support contain-
ing acid sites, the hydrogenation of nitrobenzene was conducted
on ∼1% Pd-containing Al and Ga-MCM-41 under similar
1
This journal is © The Royal Society of Chemistry 2012
Green Chem., 2012, 14, 3415–3422 | 3419

View Article Online
reaction conditions as described for Pd/B-MCM-41. Generally,
Al-MCM-41 possesses stronger surface acidity than its B
Table 2 Hydrogenation of different substrates on Pd/B-MCM-41 in
a
scCO
2
34,35
36
counterpart
and the order of acid strength is Al > Ga ≫ B.
Product
Following the comparison of reaction rate (TOF) between Pd
supported on B-MCM-41 (Si/B = 10; particle size = 21.5 nm),
Al-MCM-41 (Si/Al = 10; particle size = 18.9 nm) and
Ga-MCM-41 (Si/Ga = 10; particle size 21.2 nm), the order is B
Entry Substrate
1
Conversion (%) (selectivity %)
68.2
−1
−1
−1
(140 s ) > Ga (31.2 s ) > Al (10.2 s ). In each case, Pd par-
ticle size was chosen as ∼20 nm (ESI Fig. 3s:† low angle and
higher angle XRD pattern of Pd-containing Al and Ga-
MCM-41). Thus, acidity of support could not be the responsible
factor for high activity of the catalyst. Substitution of a trivalent
cation like B, Al and Ga modified the hydrophobic properties of
Si-MCM-41 and the hydrophobicity order for substituted Si sur-
2
82.2
3
4
100
3
7,38
faces is B > Ga > Al.
Primarily, the order of hydrophobicity
53.5
(
prevents limited access of the substrate due to the stronger
water adsorption) seems to be a possible deciding factor for
higher reaction rate of Pd/B-MCM-41. However, information
gathered from the TG analysis of Si-only and B-MCM-41 (Si/B
5
6
66.9
=
(
100) reveals a different scenario as similar hydrophobicity
almost the same amount of water loss) was detected for both of
the materials. As mentioned before, only Si and B-MCM-41
Si/B = 100) possess significant differences in the reaction rate.
89.2
60.2
(
Therefore, higher reaction rate of Pd/B-MCM-41 cannot be
directly related to the hydrophobic nature of the support. It might
be speculated that the presence of B in the support material
b
7
3
9a
causes an electronic interaction with Pd. This brings about an
increase in the metal’s electron density, causing it to interact
40
more strongly with H2 adsorbed on the metal surface and
remarkably increases the reaction rate. A similar result was also
observed during the liquid-phase hydrogenation of nitrobenzene
8
98.5
3
9b
over Pt supported on borate.
a
Reaction conditions: entries 1–6: catalyst = 0.1 g, substrate = 2.0 g,
CO2 = 12 MPa, P 2 = 2 MPa, temperature = 50 °C; time = 10 min.
Entry 7 catalyst = 0.1 g, substrate = 1.0 g, PCO2 = 10 MPa, P 2 = 2 MPa,
temperature = 50 °C; time = 2 h. Entry 8: catalyst = 0.1 g, substrate =
.0 g, PCO2 = 12 MPa, P 2 = 4 MPa, temperature = 50 °C; time = 4 h.
P
H
H
Hydrogenation of other substrates
1
H
The optimized reaction conditions (50 °C, 5 min, PCO2 = 12
MPa and PH2 = 2 MPa) of nitrobenzene have been applied for
the hydrogenation of different nitroaromatics with electron-with-
drawing group (–Cl) and electron-donating group (–OMe) and
the results are shown in Table 2. The hydrogenation of o, m and
p-chloronitrobenzene results in the corresponding amine with a
very high selectivity of >99% and conversion follows the order
of p > m > o (entries 1–3). A similar reaction over 5% Pd/C cata-
b
Dibenzylamine = 1%.
primary amine with a selectivity of 99%. The hydrogenation of
phenol (Table 2; entry 8), where ring hydrogenation is more pre-
ferable because of presence of the –OH group (electron-donating
group), results in cyclohexanone as the only product. The ring
43
hydrogenation was activated by acidity of the support due to
the presence of B in the support material, as confirmed by com-
parison with Pd/MCM-41 (only Si) catalyst under similar reac-
lyst was also described by Ichikawa et al. in scCO and obtained
2
1
4
1
00% conversion with 82.1% selectivity of chloroaniline at
0 °C, reaction time = 150 min, P = 1.1 MPa and PCO2
=
H2
41
tion conditions (temperature = 50 °C, time = 4 h, P
MPa, PH2 = 4 MPa) as described previously.
= 10
CO
2
0 MPa. Other compounds such as nitroanisoles also behaved
44
in a similar manner (entries 4–6) though conversion was lower
than halonitroaromatics. In each case, only the nitro group was
hydrogenated, which might be attributed to the capability of the
nitro group attached to the ring to pull electrons more strongly
compared to other functional groups. As a result it could be
easily attracted to the catalyst surface and be hydrogenated
Comparison with organic solvents
For comparison, hydrogenation of nitrobenzene was conducted
in neat organic solvents like hexane and ethanol, and the results
are shown in Table 3. In each case the amount of catalyst, sub-
4
2
instead of the other groups. The potential of the catalyst is
once again described by the hydrogenation of nitrile (–CN; elec-
tron-withdrawing group) and phenol (–OH; electron-donating
group) (Table 2; entries 6 and 7). It has to be mentioned that,
similar to nitro compounds, only –CN was hydrogenated to
strate, H pressure, temperature and reaction time were similar to
2
those in scCO . In ethanol the reaction was also fast (Table 3:
2
entry 2) and comparable with scCO2 (Table 3; entry 1).
However, the selectivity of aniline dropped (73.2%) due to the
3420 | Green Chem., 2012, 14, 3415–3422
This journal is © The Royal Society of Chemistry 2012

View Article Online
Table 3 Nitrobenzene hydrogenation in different solvent and catalyst
recycling
it is an excellent catalyst for the hydrogenation reaction of nitro-
benzene under very mild conditions. The reaction was extremely
fast due to the presence of B in the support and aniline was the
only product formed. Furthermore, the effect of B can then be
understood from the comparison with silica-only material. When
the hydrogenation of nitrobenzene was conducted over Pd/
Al-MCM-41 and Pd/Ga-MCM-41, the reaction rate followed the
order of B > Ga > Al. The described reaction system is free from
any by-product formation and long-term catalytic activity can be
achieved, as observed from recycling of the catalyst. This simple
low metal-containing catalyst was also adaptable for successful
chemo-selective hydrogenation of substituted nitroaromatics as
well as other aromatic compounds and might be a potential can-
didate for further development of other clean chemical
processes.
Selectivity (%)
Entry
1
Solvent
scCO
Ethanol
Hexane
Conversion (%)
Aniline
Others
0.0
26.8
0.0
2
100
100
33.6
100
73.2
100
a
b
2
a
3
c
4
5
6
1st recycle
4th recycle
8th recycle
84.0
84.1
80.2
100
100
100
0.0
0.0
0.0
a
b
Solvent = 5 ml, only H
2
of 2 MPa. Nitrosobenzene (3.8%),
azoxybenzene (12.7%) and azobenzene (10.3%); reaction conditions:
catalyst = 0.1 g, substrate = 2.0 g, PCO2 = 12 MPa, PH2 = 2 MPa,
c
temperature = 50 °C; time = 5 min. Recycling after shorter reaction
time of 2 min (entries 4–6).
Experimental
formation of by-products such as nitrosobenzene (3.8%), azoxy-
benzene (12.7%) and azobenzene (10.3%). It is well-known that
Synthesis, characterization techniques and catalytic activity
measurements
scCO has a low dielectric constant (ε = 1.2–1.6) under all con-
ditions and it is believed that there are some similarity between
2
Materials. Tetraethyl orthosilicate (TEOS) as silica source,
boric acid (H BO ), sodium aluminate and gallium nitrate were
3
3
the behaviour of CO and hexane. Hydrogenation of nitroben-
2
taken as B, Al and Ga sources, respectively. All of the above-
mentioned chemicals, sodium hydroxide and nitrobenzene were
purchased from Wako Pure Chemicals. Cetyltrimethylammonium
zene in hexane under the studied reaction conditions exhibited
very low conversion (33.1%) (Table 3; entry 3), but aniline was
the only product detected. This is excellent support to suggest
bromide (CTAB) and PdCl were from Aldrich Chemical Co.
2
that scCO is a unique medium to achieve high activity and
2
and used as received. CO (>99.99%) has been provided by the
2
selectivity due to better solubility of H2.
Again, a comparison with our previous work (catalyst = 5%
Pd/C, conversion = 52%, selectivity = 100% reaction time =
Nippon Sanso Co. Ltd.
Synthesis. Hydrothermal synthesis of Pd/B-MCM-41 has
been carried out as follows: Typically, sodium hydroxide and
CTAB were added to de-ionized water and stirred until dissolved.
After that, the required amount of H BO was introduced slowly
1
0 min, PH2 = 4 MPa and PCO2 = 14 MPa; temperature = 35 °C)
related to the hydrogenation of nitrobenzene using Pd catalyst in
scCO revealed that the reaction was exceptionally faster over
2
3
3
3
2a
Pd/B-MCM-41 catalyst.
under stirring and stirring was continued for another 1 h, fol-
lowed by the addition of 1 wt% solution of Pd salt and again
stirred for another 1 h. Finally, TEOS was added to the solution
with stirring for another 1 h and the gel mixture was autoclaved
at 140 °C for 48 h. Final gel composition was: 1 TEOS, x B,
Catalyst recycling
Recycling of catalyst is one of the most important criteria for
B-containing catalyst. Potential recyclability of the catalyst was
conducted with a reaction time of 2 min under the similar reac-
tion conditions as described before. After the designated reaction
time, the catalyst was filtered and then transferred to another
reactor containing fresh nitrobenzene. The process was repeated
for eight successive runs. The conversion was consistent until
the 7th run and then dropped to 80.2% without any change in
the selectivity (Table 3; entries 4–6). TEM observations of the
catalyst after the 4th recycle confirmed its stability, whereas
some changes of particle size in comparison to the parent
material (Fig. 2) were observed after the 8th recycle (see ESI,
Fig. 4s†).
0
.45 Na O, 0.12 CTAB, 118 H O, where x = 0.01–0.2. The solid
2 2
product was then filtered, washed thoroughly with de-ionized
water followed by oven drying at 60 °C. The synthesized
material was calcined at 550 °C for 8 h in air to remove the
template. For comparison, Pd supported on Al and Ga-
MCM-41 materials were also synthesized in a similar manner.
The resultant product was characterized by X-ray diffraction
XRD), transmission electron microscopy (TEM), Fourier trans-
form infrared spectroscopy (FTIR) and thermogravimetric-differ-
ential thermal analysis (TG-DTA).
(
Catalytic activity. All the reactions were conducted in a 50 ml
44
stainless steel batch reactor and the details are given elsewhere.
Briefly, 0.1 g of catalyst and 2 g of the reactant were introduced
into the reactor and placed in an oven with fan heater to maintain
Conclusions
the desired temperature. At first H of required pressure is intro-
2
In conclusion, Pd nanoparticles supported on B-MCM-41 were
synthesized successfully by a hydrothermal method. Different
characterization techniques revealed that, depending on Si/B
ratio, structural order as well as Pd particle size varied. Investi-
gation of the catalytic activity of Pd/B-MCM-41 suggested that
duced into the reactor. After that liquid CO was charged using a
2
high-pressure liquid pump (JASCO) and then compressed to the
desired pressure. The reaction mixture was stirred during the
reaction. The liquid product was separated from the catalyst and
identified by GC-MS (Varian Saturn 2200), followed by
This journal is © The Royal Society of Chemistry 2012
Green Chem., 2012, 14, 3415–3422 | 3421

View Article Online
quantitative analysis using a GC (HP 6890) equipped with capil-
lary column and flame ionization detector. For all results
reported, the selectivity is defined as follows:
11 A. Tuel and S. Gontier, Chem. Mater., 1996, 8, 114–122.
1
1
1
1
2 Z. Y. Yuan, Q. Luo, J. Q. Liu, T. H. Chen, J. Z. Wang and H. X. Li,
Microporous Mesoporous Mater., 2001, 42, 289–297.
3 V. Sundaramurthy and N. Lingappan, Microporous Mesoporous Mater.,
2003, 65, 243–255.
%
selectivity ¼ conc: of the product=total conc: of products
100
4 T. D. Conesa, J. M. Hidalgo, R. Luque, J. M. Campelo and
A. A. Romero, Appl. Catal., A, 2006, 299, 224–234.
5 M. G. Dekamin, Z. Mokhtari and Z. Karimi, Sci. Iran. C, 2011, 18,
ꢀ
1356–1364.
1
1
6 L. Dai, K. Koyama and T. Tatsumi, Catal. Lett., 1998, 53, 211–214.
7 Usman, T. Kubota, Y. Araki, K. Ishida and Y. Okamoto, J. Catal., 2004,
For organic solvent in place of CO , 5 ml of organic solvent
has been used.
2
2
27, 523–529.
8 (a) M. Chatterjee, A. Chatterjee and Y. Ikushima, Green Chem., 2004, 6,
14–118; (b) M. Chatterjee, A. Chatterjee, T. Ishizaka, T. Suzuki,
A. Suzuki and H. Kawanam, Adv. Synth. Catal., 2012, 354, 2009–2018;
c) E. Bogel-Łukasik, R. Bogel-Łukasik and M. Nunes da Ponte, Ind.
1
1
Catalyst characterization techniques
(
In order to have an approximation of the elemental composition
and metal loading in Pd/B-MCM-41 samples, energy dispersive
spectroscopy (EDS) was carried out with XL 30 s Philips Co.
equipment coupled Scanning Electron Microscope. The analyses
were randomly taken in several sample zones to have a represen-
tative value of the elemental composition.
Eng. Chem. Res., 2009, 48, 7060–7064; (d) C. I. Melo, R. Bogel-
Łukasik, M. G. da Silva and E. Bogel-Łukasik, Green Chem., 2011, 13,
2
825–2830; (e) E. Bogel-Łukasik, M. G. da Silva, L. D. Nogueira,
R. Bogel-Łukasik and M. Nunes da Ponte, Green Chem., 2009, 11,
847–1856.
19 M. Chatterjee, T. Iwasaki, Y. Onodera and T. Nagase, Catal. Lett., 1999,
1, 199–202.
1
6
2
2
0 C. A. Koh, R. Nooney and S. Tahir, Catal. Lett., 1991, 47, 199–203.
1 A. Fukuoka, H. Araki, Y. Sakamoto, S. Inagaki, Y. Fukushima and
M. Ichikawa, Inorg. Chim. Acta, 2003, 350, 371–378.
Powder X-ray diffraction pattern was recorded on a Rigaku-
RAD-X system using monochromatized Cu Kα radiation (λ =
1
.542 Å). In general, the diffraction data were collected using
22 J. Panpranot, K. Pattamakomsan, J. G. Goodwin Jr. and P. Prasertham,
−1
Catal. Commun., 2004, 5, 583–590.
continuous scan mode with a scan speed of 2° min over the
scan range 2θ = 1.5–15° and for wide-angle (2θ = 30–70°)
measurements.
Transmission electron microscopy (TEM) of the material was
recorded on a Philips Technai operating at 200 kV. The powder
was suspended in ethanol by ultrasonic method. A drop of this
solution was placed on a grid with a holey carbon copper film
and then allowed to dry, covered by a watch glass.
Fourier-transform infrared spectra (FT-IR) were taken at room
temperature on a Horiba spectrometer FT-720. Before measure-
ment, samples were ground with KBr and pressed into thin
wafers.
23 I. Yuranova, P. Moeckli, E. Suvorova, P. Buffat, L. Kiwi-Minsker and
A. Renken, J. Mol. Catal. A: Chem., 2003, 192, 239–251.
24 S. Zheng and L. Gao, Mater. Chem. Phys., 2003, 78, 512–517.
25 T. Kawabata, I. Atake, Y. Ohishi, T. Shishido, Y. Tian, K. Takakiand and
K. Takehira, Appl. Catal., B, 2006, 66, 151–160.
26 U. Oberhagemann, M. Jeschke and H. Paap, Microporous Mesoporous
Mater., 1999, 33, 165–172.
27 D. T. On, P. N. Joshi, G. Lemay and S. Kaliaguin, Stud. Surf. Sci. Catal.,
995, 97, 543.
1
28 S. X. Liu, H. Y. He, Z. H. Luan and J. Klinowski, J. Chem. Soc.,
Faraday Trans., 1996, 92, 2011–2015.
29 P. Scherrer and N. G. W. Gottingen, Math-Phys. Kl, 1918, 2, 96–100.
30 (a) R. D. Ruiter, K. Famine, A. P. M. Kentgens, J. C. Jansen and H. van-
Bekkum, Zeolites, 1993, 13, 611–621; (b) R. D. Ruiter,
A. P. M. Kentgens, J. Grootendost, J. C. Jansen and H. van-Bekkum,
Zeolites, 1993, 13, 128–138.
Thermogravimetric analysis was carried out by using a
Rigaku Thermoflex TAS200 unit. About 10 mg of the sample
was placed in a platinum pan and was heated at a heating rate
31 C. Y. Chen, H. X. Li and M. E. Davis, Microporous Mater., 1993, 2,
17–26.
of 10 °C min−1
32 (a) F. Zhou, R. Zhang, M. Chatterjee, Y. Ikushima and M. Arai, Adv.
Synth. Catal., 2004, 346, 661–668; (b) F. Zhou, Y. Ikushima and M. Arai,
J. Catal., 2004, 224, 479–483.
.
3
3
3
3
3
3
3
4
4
4
3 E. A. Gelder, S. D. Jackson and C. M. Lok, Catal. Lett., 2002, 84,
05–208.
4 C. T.-W. Chu, G. H. Kuehl, R. M. Lago and C. D. Change, J. Catal.,
985, 93, 451–458.
5 M. Adjdir, T. Ali-Dahmane and P. G. Weidler, C. R. Chim., 2009, 12,
93–800.
6 H. Kosslick, G. Lische, H. Landmesser, B. Parlitz, W. Storek and
R. Fricke, J. Catal., 1998, 176, 102–114.
7 D. P. Serrano, G. Calleja, J. A. Botas and F. J. Gutierrez, Ind. Eng. Chem.
Res., 2004, 43, 7010–7018.
Notes and references
2
1
(a) C. del Pozo, A. Corma, M. Iglesias and F. Sanchez, Green Chem.,
011, 13, 2471–2481; (b) E. Bogel-Łukasik, S. Santos, R. Bogel-Łukasik
and M. Nunes da Ponte, J. Supercrit. Fluids, 2010, 54, 210–217;
c) J. Arras, M. Steffan, Y. Shayeghi and P. Claus, Chem. Commun.,
008, 4058–4060.
1
2
7
(
2
2
3
J. S. Beck, J. C. Vartuli, W. J. Roth, M. E. Leonowicz, C. T. Kresge,
K. D. Schmitt, C. T.-W. Chu, D. H. Olson, E. W. Sheppard,
S. C. McCullen, J. B. Higgins and J. L. Schlenker, J. Am. Chem. Soc.,
8 L. P. Zhang, P. Wu and M. B. Sullivan, J. Phys. Chem. C, 2011, 115,
1
992, 114, 10834–10843.
(a) R. Köhn and M. Förba, Catal. Today, 2001, 68, 227–236;
b) M. Morey, A. Davidson, H. Eckert and G. D. Stucky, Chem. Mater.,
996, 8, 486–492.
4289–4296.
9 (a) Y. W. Chen and C. Li, Catal. Lett., 1992, 13, 359–361; (b) C. Li,
Y. W. Chen and W. J. Wang, Appl. Catal., A, 1994, 119, 185–194.
0 G.-X. Chen, J.-M. Zhang, D.-D. Wang and K.-W. Xu, Phys. B, 2009,
(
1
4
5
C. G. Sonwane and S. K. Bhatia, Langmuir, 1999, 15, 2809–2816.
D. M. Antonelli and J. Y. Ying, Curr. Opin. Colloid Interface Sci., 1996,
4
04, 4173–4177.
1 S. Ichikawa, M. Tada, Y. Iwasawa and T. Ikariya, Chem. Commun., 2005,
24–926.
1, 523–529.
9
6
7
8
A. Sayari and Y. Yang, J. Phys. Chem. B, 2000, 104, 4835–4839.
A. Corma, Top. Catal., 1997, 4, 249–260.
U. Oberhagemann, I. Kinski, I. Dierdorf, B. Marier and H. Gies, J. Non-
Cryst. Solids, 1996, 197, 145–153.
2 (a) M. Bhagiyalakshmi, S. Vishnupriya, J. Herbert Mabel,
M. Palanichamy and V. Murugesan, Catal. Commun., 2008, 9, 2007–
2012; (b) S. K. Mohapatra, S. U. Sonavane, R. V. Jayaram and P. Selvam,
Appl. Catal., B, 2003, 46, 155–163.
9
D. T. On, P. N. Joshi and S. Kaliaguine, J. Phys. Chem., 1996, 100,
43 H. Z. Liu, T. Jiang, B. X. Han, S. G. Liang and Y. X. Zhou, Science,
6
743–6748.
2009, 326, 1250–1252.
1
0 A. Sayari, A. I. Moudrakovski, C. Danumah, C. I. Ratcliffe,
44 M. Chatterjee, H. Kawanami, M. Sato, A. Chatterjee, T. Yokoyama and
J. A. Ripmeester and K. F. Preston, J. Phys. Chem., 1995, 99, 16373–
T. Suzuki, Adv. Synth. Catal., 2009, 351, 1912–1924.
1
6379.
3422 | Green Chem., 2012, 14, 3415–3422
This journal is © The Royal Society of Chemistry 2012