Fabrication of hybrid mesoporous TiO2-SiO2(Et) supported Ni nanoparticles: An efficient and air/water stable catalyst

Article abstract of DOI:10.1016/j.molcata.2015.10.023

We prepared a series of mesostructured Ni/TiO₂-SiO₂(Et) hybrid catalysts with highly dispersed Ni nanoparticles and incorporated ethane-bridged organosilica moieties. Ni/TiO₂-SiO₂(Et) showed high activity in the hydrogenation of nitrobenzene in water, and it could be recycled for several times with a constant activity and selectivity. It was confirmed that Ni/TiO₂-SiO₂(Et) catalyst is of hydrophobicity as the ethane-bridged organosilica fragments were incorporated into the mesoporous framework, and so the Ni active species was protected without contacting with water to form the inactive Ni species. In particularly, the Ni/TiO₂-SiO₂(Et) catalyst was air-stable, it could remain good activity after being exposed to air for a week. Accordingly, this work developed a kind of hydrophobic Ni catalyst with high stability to water and air, which is expected to have a wide application in the hydrogenation reactions.

Full text of DOI:10.1016/j.molcata.2015.10.023

Journal of Molecular Catalysis A: Chemical 411 (2016) 214–221
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Fabrication of hybrid mesoporous TiO –SiO (Et) supported Ni
2
2
nanoparticles: An efficient and air/water stable catalyst
Wei Lia,b,1, Haiyang Cheng
Fengyu Zhao
a
a,b,∗
, Weiwei Lina,b, Guanfeng Lianga,b, Chao Zhanga,b
,
a,b,∗
State Key Laboratory of Electroanalytical Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022, PR China
Laboratory of Green Chemistry and Process, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022, PR China
b
a r t i c l e i n f o
a b s t r a c t
Article history:
We prepared a series of mesostructured Ni/TiO –SiO (Et) hybrid catalysts with highly dispersed Ni
2
2
Received 12 September 2015
Received in revised form 14 October 2015
Accepted 26 October 2015
Available online 30 October 2015
nanoparticles and incorporated ethane-bridged organosilica moieties. Ni/TiO2–SiO2(Et) showed high
activity in the hydrogenation of nitrobenzene in water, and it could be recycled for several times with a
constant activity and selectivity. It was confirmed that Ni/TiO2–SiO2(Et) catalyst is of hydrophobicity as
the ethane-bridged organosilica fragments were incorporated into the mesoporous framework, and so
the Ni active species was protected without contacting with water to form the inactive Ni species. In par-
ticularly, the Ni/TiO2–SiO2(Et) catalyst was air-stable, it could remain good activity after being exposed
to air for a week. Accordingly, this work developed a kind of hydrophobic Ni catalyst with high stability
to water and air, which is expected to have a wide application in the hydrogenation reactions.
© 2015 Elsevier B.V. All rights reserved.
Keywords:
Ni-based catalyst
Stability
Mesoporous TiO2–SiO2
Hydrophobicity
1
. Introduction
More recently, the research on the nonprecious catalysts has
Ni/SiO2 and Ni/␥-Al O3 catalysts due to the interaction between
Ni and TiOx (x < 2) species [17].
2
More recently, water as a solvent or co-solvent has been well
investigated for the catalytic hydrogenation reactions. The addi-
tion of appropriate amounts of water into methanol or ethanol,
the reaction rate was enhanced dramatically for the selective
hydrogenation of CNB over a series of metal supported catalysts
[18–21], for example, when the molar ratio of water to methanol
increased from 0 to 1, the reaction rates of the transfer hydro-
genation of styrene and nitrobenzene (NB) over Pd-based catalyst
were increased more than 3 times [18]. Furthermore, it was also
reported that the increase in water content could raise the reac-
tion rate of the hydrogenation of p-nitrophenol in both ethanol
and dioxane solvents, and the highest reaction rate was obtained
in pure water [22]. Moreover, it was found that minor amount
of water (0.14–1 mL) could significantly promote hydrogenation
of 6-chloro-2-nitrotoluene over Pd/C catalyst [23]. In our previ-
ous work, we found water could accelerate the reaction rate and
improve the product selectivity largely in the hydrogenation of o-
CNB and NB [24–25], because the interaction between water and
N-phenyl-hydroxylamine (PHA) via OH O and OH N hydrogen
bonding which accelerates the rate-determining step of PHA to
aniline (AN), as the ␯(N O) of PHA presented a red-shift in the pres-
ence of water [25], besides, interfacial hydrogen bonding, OH ONO
attracted many interests, and the Ni-based catalysts have been paid
more attention due to its low cost and moderate activity with com-
paring to the precious catalysts. However, the Ni catalysts are still
far from the ideal application as they bear some intrinsic disad-
vantages [1–10]. For example, Raney Ni bears the severe corrosion
and the pollution originating from catalyst preparation as well as
the safety problem in storage and application [11,12]. To date, the
supported Ni catalyst like Ni/TiO was reported to be a promising
2
candidate for nitro-compounds hydrogenation because of its high
activity, easy availability, low price and safe in handling. TiO2 as
a reducible metal oxide support could be reduced at a high tem-
perature to form suboxide TiOx species (x < 2), and migrate to the
surface of metal particles [13–16]. And so, the adsorbed N O bond
was polarized and highly susceptible to the hydrogen attack, result-
ing in a high turnover frequency (TOF) in the hydrogenation of
nitro-compounds [15,16]. The similar results was also found in the
hydrogenation of o-chloronitrobenzene (o-CNB) to o-chloroaniline
(
o-CAN), in which Ni/TiO2 was much more active than Ni/ZrO2,
∗ Corresponding authors at: State Key Laboratory of Electroanalytical Chemistry,
Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun
between H O and NB molecules may weaken the N O bond of NB
2
1
30022, PR China. Fax: +86 431 8526 2410.
[26], improve the hydrogenation of NB to PHA. Ni/TiO2 was an effec-
tive catalyst for the hydrogenation of nitro-compounds, however, it
E-mail addresses: zhaofy@ciac.ac.cn, hycyl@ciac.ac.cn (F. Zhao).
1
W. Li is currently with Triangle Tyre Co., Ltd., Weihai 264200, PR China.
http://dx.doi.org/10.1016/j.molcata.2015.10.023
381-1169/© 2015 Elsevier B.V. All rights reserved.
1

W. Li et al. / Journal of Molecular Catalysis A: Chemical 411 (2016) 214–221
215
significantly deactivated in water, while its stability was improved
significantly when coated with a layer of hydrophobic carbon, but
the stronger adsorption of organic products on the carbon layer
lowered the activity gradually during the recycles [27]. Therefore,
it is of great interest and a big challenge to design a stable Ni catalyst
for the aqueous phase hydrogenation.
tively coupled plasma atomic emission spectrometer (ICP-AES).
Ni/TiO was prepared according to our previous work [27].
2
2.4. Characterization of catalyst
Powder X-ray diffraction (XRD) was performed using a Bruker
Herein, a hybrid mesoporous TiO –SiO (Et) supported Ni
D8 Advance X-ray diffract meter with a Cu K␣ source at 40 kV and
2
2
◦ ◦ ◦
40 mA. The scans were performed from 0.6 to 5 at a 0.05 /min
nanoparticles (Ni/TiO –SiO (Et)) was prepared in scCO2-
2
2
◦
◦
◦
expanded ethanol by using co-condensation, hydrothermal
treatment technique. Ethane-bridged organosilica fragment
speed for low angle XRD (LXRD) and 10 to 90 at a 4 /min speed
for wide angel XRD (WXRD).
(
O1.5–Si–CH CH –Si–O1.5) was used as bridging component,
Nitrogen porosimetry measurement was performed on a
Micromeritics ASAP 2020 M instrument. The surface areas were
calculated using the BET equation. Pore size distributions were
calculated using the BJH model based on nitrogen desorption
2
2
which can improve the porosity and simultaneously increase
the surface hydrophobic of catalyst. The structural and textural
properties as well as the morphology of the Ni/TiO –SiO (Et)
2
2
◦
catalysts were well characterized and their catalytic performances
activity and stability) were evaluated and discussed with the
hydrogenation of nitrobenzene in water.
isotherms. The samples were treated under vacuum at 90 C for
◦
(
1 h and then 200 C for 12 h.
Transmission electron microscopy (TEM) study was carried out
with a JEOL JEM-2010 instrument at an accelerating voltage of
2
00 kV. The TEM samples were prepared by dispersing the cata-
2
. Experimental
lyst powder in ethanol under ultrasonic for 5–10 min and then the
resulted solution was dropped on a carbon film of copper grid.
2.1. Chemical and reagents
13
C CP-MAS NMR spectrum was recorded on a Bruker AVANCE
III 400WB spectrometer equipped with a 4 mm standard bore CP
Pluronic P123 (MW = 5800), and 1,2-bis-(triethoxysilyl) ethane
BTESE, 97%) were purchased from Sigma–Aldrich and used with-
out further purification. Tetrabutyl titanate (TBT, 98 %) and
MAS. Chemical shift for 13C CP-MAS NMR spectrum was referenced
(
to the signal of adamantane (C10H16) standard (␦CH = 38.5 ppm).
2
X-ray photoelectron spectroscopy (XPS, VG Microtech 3000
Multilab) was used to examine the electronic properties of Ni on the
surface of catalysts. The C 1 s peak at 284.6 eV arising from adven-
titious carbon was used as reference. This reference gives binding
energy values with a precision of ± 0.02 eV. The surface composition
of the samples was determined from the peak areas of the corre-
sponding lines using a Shirley type background and empirical cross
section factors of XPS. Generally, the sample for XPS measurement
Ni(NO ) ·6H O were obtained from Sinopharm Chemical Reagent
3
2
2
Co. Ltd., Beijing, China. All other chemicals were analytical reagents
and obtained from Beihua Fine Chemical Co., Beijing, China.
2
.2. Preparation of mesoporous TiO –SiO (Et-x)
2 2
A series of the TiO –SiO (Et-x) materials were synthesized by
2
2
the following process. Pluronic P123 (0.55 g) was dissolved with
H O (16.9 mL) to obtain a clear solution, and then HCl (37.0%,
is the fresh reduced Ni catalysts stored in ethanol with N protec-
2
2
tion. For etching the surface layers, bombardment by argon ions
with energy of 5000 eV was used.
1
.18 mL) was added. Subsequently, BTESE and TBT was added drop
wise to the above solution at hourly intervals, successively. The
◦
resulting mixture was stirred at 40 C for 24 h, and then heated
2.5. Catalytic performance tests
◦
◦
−1
up to 100 C with a heating rate of 2 C min and hydrothermal
treated for another 24 h. The molar ratio of P123/Ti/Si//HCl/H O in
2
Prior to reaction, the diffusion effect was checked by changing
the size of catalyst and the stirring speed. The calcined cata-
lysts were ground and separated by screen with a size from 140
≤109 ␮m) and 180 meshes (≤80 ␮m). The results show that the
the starting material is 0.06/1.00/(0.20–1.00)/8.90/691. The result-
ing white solid product was filtered, washed with deionized water
◦
and ethanol, and then air-dried at 100 C overnight. The final prod-
(
uct was denoted as TiO –SiO (Et-x), where x represents the molar
2
2
reaction rate kept at a constant value with a particle size smaller
than 140 meshes, indicating that the inner transfer resistance was
removed. In addition, it is confirmed when the stirring speed was up
to 800 rpm, the reaction rate did not increase further, indicating the
external diffusion has been removed. As a result, the catalysts with
a particle size smaller than 180 meshes (<80 ␮m) was used, and
1200 rpm was selected for the reactions to evaluate the catalytic
performances.
ratio of Si to Ti in the sample; herein, x = 0.2, 0.4, 0.6.
meso
The preparation of TiO2
was similar to the TiO –SiO (Et-
2 2
x) hybrid material as described above, except for without BTESE
involved.
2
.3. Preparation of Ni/TiO2–SiO (Et-x)
2
The as-prepared TiO /SiO (Et-x) materials (100 mg) were well
Before reaction, the catalyst (40 mg) was freshly reduced in a
quartz tube at appropriate temperature with H2 flow 30 mL/min
for 2 h; then, H was changed to N and the quartz tube was cooled
2
2
dispersed in ethanol solution (10 mL) of Ni(NO ) 6H O in a glass
3
2
2
bottle under ultrasonic treatment. The dense colloidal solution was
2
2
then transferred into an autoclave (50 mL), which subsequently was
down to room temperature. The reduced catalyst was transferred
into a 50-mL stainless steel autoclave reactor in N2 flow in which
◦
placed into an oil bath at 150 C and then pumped CO to form a
2
homogeneous expanded fluid under rapid stirring (12.0 MPa). And
then the reactor was heated to the reaction temperature of 200 C,
3 mL nitrobenzene and 3 mL H O was added. The reactor was then
sealed, and flushed with H2 more than three times to remove the
air and placed into a water bath preset to the reaction tempera-
2
◦
at which the pressure went up to ca. 22.0 MPa. After the reaction
was performed for 2 h, the autoclave was cooled down to room
temperature, and then CO2 was released slowly. After centrifuga-
tion, the composites were collected and dried. The samples were
ture for 15 min. Then, H (5.0 MPa) was introduced into the reactor,
2
◦
and the reaction was started at 80 C with an agitation speed of
1200 rpm. When the reaction was finished, the reactor was cooled
to room temperature in ice-water bath and then vented hydro-
gen to ambient pressure. The liquid product was extracted with
10 mL diethyl ether and analyzed with a gas chromatograph (Shi-
madzu GC-2010, Rtx-5 capillary column) using a flame ionization
◦
reduced under H2 flow at 350 C for 2 h as confirmed by TPR (Fig.
S1), the samples were collected and labelled as Ni/TiO –SiO (Et-x).
2
2
meso
Ni/TiO2
was prepared with similar procedure described above.
Ni loading in the catalyst was about 17 wt% as determined by induc-

2
16
W. Li et al. / Journal of Molecular Catalysis A: Chemical 411 (2016) 214–221
detector. Through GC–MS analysis, aniline (AN) was the main prod-
ucts besides several intermediates such as nitrosobenzene (NSB),
N-phenylhydroxylamine (PHA), azoxybenzene (AOB), azobenzene
(
AB) and hydrazobenzene (HAB).
The catalytic performance of the catalysts was characterized
quantitatively by the conversion of nitrobenzene and the selectivity
of aniline which was calculated as follows:
Conversion = (
[NB] − [NB] )
× 100%
× 100%
0
t
[
NB]
0
MAN
Selectivity = (MAN
+ Mby-products
)
where [NB]0 and [NB]t refers to the concentration of nitroben-
zene at 0 and t h, respectively; MAN and Mby-products represents the
amount of aniline and by-products (e.g.NSB, PHA and a series of
azo-compounds identified by GC–MS), respectively. The GC results
were collected with an internal standard method, and o-xylene was
used as internal standard. The carbon balance was near 100% on the
basis of the detected compounds of NB, NSB, PHA, AOB, AB and HAB.
For the catalyst recycling test, the reactions were operated with
the same procedure as described above. After the first run finished,
the catalyst was separated and used for the next run directly. In the
Fig. 1. 13C CP/MAS NMR spectrum of Ni/TiO2–SiO2(Et-0.6).
recycles, the H consumption was recorded by the pressure change
2
at the interval time, the end of reaction was judged by the cease of
pressure dropping.
3
. Results
3
.1. Preparation of mesostructured Ni/TiO –SiO (Et) catalyst
2
2
TiO –SiO (Et) hybrid materials were prepared by co-
2
2
hydrolysis and condensation of TBT with bissilylated organic
precursor (BTESE) in the presence of triblock copolymer
surfactant P123, which was used as the structure directing
reagent. Generally, the hydrolysis rate of BTESE was obviously
slower than that of TBT. In order to successfully introduce
O1.5–Si–CH CH –Si–O as bridging component into TiO2 frame-
2
2
1.5
work through Ti O Si C C Si O linkages, prehydrolysis of
◦
BTESE was proceeded at 40 C for 1 h at the beginning of prepara-
tion [28,29]. Therefore, the mesostructured TiO -SiO (Et) hybrid
2
2
materials were obtained with ethane-bridged organosilica moi-
eties incorporating into the TiO2 framework as confirmed by 13
C
CP/MAS NMR spectrum in Fig. 1. The strong signal at ␦ of 5.57 ppm
was attributed to carbon atoms of the ethane-bridged organosilica
species, whereas the other three weak signals at ı = 17.06, 71.05,
and 76.05 ppm were originated from carbon species in the residual
surfactant P123 [30]. It also indicated that the ethane-moieties
Fig. 2. LXRD patterns of Ni/TiO2meso and Ni/TiO2–SiO2(Et-x).
the tested samples, except Ni/TiO , displayed type IV isotherm,
2
indicating the obtained Ni/TiO –SiO (Et-0.6) is of mesoporosity.
2
2
◦
meso
still existed in the samples after reduced at 350 C, which was
Ni/TiO –SiO (Et-0.6) and Ni/TiO
exhibited H2 hysteresis
2
2
2
also determined by the results of TGA (Fig. S2). The weight loss
increased with molar ratio increasing of Si to Ti (BTESE to TBT) at
the temperature range of 30–800 C, as the decomposition of the
loops with narrow BJH pore size distribution curves, suggesting
that these catalysts possessed uniform pore with interconnect-
ing channels. Additionally, the shape of hysteresis loop and the
amount of volume absorbed of TiO –SiO (Et-0.6) are similar to
◦
ethane-moieties in the matrix.
2
2
Meanwhile, the Ni nanoparticles were dispersed onto
TiO –SiO (Et) using scCO2-expanded ethanol technique as
the values of Ni/TiO –SiO (Et-0.6), implying that the dispersing
2 2
process of Ni particles in scCO -expanded ethanol did not change
2
2
2
described in previous work [31,32]. Consequently, a series of
mesostructured Ni/TiO –SiO (Et-x) catalysts were fabricated
the mesoporosity of TiO –SiO (Et-0.6). The similar results were
2
2
meso
meso
also observed for Ni/TiO2
and TiO2
. However, Ni/TiO2
2
2
with a Ni loading controlled and a highly dispersed Ni nanopar-
ticles on the surface and/or inside of pores. The structure and
physical properties were characterized and examined by LXRD,
TEM, and nitrogen porosimetry measurement. LXRD patterns
of Ni/TiO –SiO (Et-x) samples (x = Si/Ti molar ratio, 0.2–0.6)
exhibited a sole Bragg reflection at 2␪ = 1.17 (Fig. 2), indicating
that these materials have the uniformed mesopores. The nitro-
gen adsorption–desorption isotherms and pore size distribution
of the as-prepared Ni-based catalysts are shown in Fig. 3. All
showed H3 hysteresis loop with broad pore-size distribution. The
textural parameters of the prepared catalysts and supports are
summarized in Table 1, the BET surface area, the pore diameter,
2
and the pore volume of Ni/TiO –SiO (Et-0.6) was 404 m /g, 6.3 nm
2
2
3
meso
2
and 0.54 cm /g, but these values of Ni/TiO
3.8 nm and 0.23 cm /g, respectively. Compared with the support
TiO –SiO (Et-0.6), the BET surface area and the pore volume of the
was 203 m /g,
2
2
2
◦
3
2
2
supported Ni catalyst of Ni/TiO –SiO (Et-0.6) presented a slightly
2
2
meso
decrease. Moreover, the porosity of Ni/TiO2
prepared using the

W. Li et al. / Journal of Molecular Catalysis A: Chemical 411 (2016) 214–221
217
Fig. 3. Nitrogen adsorption–desorption isotherms (a) and pore size distribution profiles (b) of various catalysts and supports.
Table 1
Textural parameters of the supports and catalysts.
Entry
Catalyst
Ni particle size (nm)a
Textural property
2
Dp (nm)b
Vp (cm /g)c
3
SBET (m /g)
1
2
3
4
5
6
Ni/TiO2
TiO2
14
–
12
–
15
–
67
11.0
–
3.8
3.8
6.3
6.4
0.25
–
0.23
0.24
0.54
0.61
120
203
220
404
511
meso
Ni/TiO2
meso
TiO2
Ni/TiO2–SiO2(Et-0.6)
TiO2–SiO2(Et-0.6)
a
Ni particle size was determined by WXRD measurement.
Pore diameter was estimated from BJH desorption determination.
Pore volume was determined using the adsorption branch of the nitrogen isotherm curve at the P/P0 = 0.97 single point.
b
c
P123 surfactant-assisted co-condensation process was superior
quickly in the first run with an incomplete conversion as reported
in our previous work [27]. While as expected, Ni/TiO –SiO (Et-0.2)
to the Ni/TiO2 prepared using the commercial anatase TiO , the
2
2
2
meso
2
BET surface area of Ni/TiO2
(203 m /g) was triple of Ni/TiO2
and Ni/TiO –SiO (Et-0.6) exhibited improved stability, although
2 2
2
(
67 m /g). Additionally, as shown in the TEM images (Fig. 4a, c),
the reaction rate presented a little decrease in the first three runs,
meso
−1
Ni particles dispersed well on TiO –SiO (Et-0.6) and TiO
, and
but it could maintain a constant rate of 103 h in the following
2
2
2
the average size of Ni particles was approximately 15 and 12 nm,
respectively.
cycles for Ni/TiO –SiO (Et-0.6).
2
2
As we known, the active Ni species is very susceptive and easily
to be oxidized with air. Therefore, the Ni-based catalysts are usu-
ally pre-reduced and kept in the solvent without exposing to air.
Improvement of the air-stability of Ni catalyst is very important and
a challenge for the practical applications of Ni-based catalyst. To
3
.2. The catalytic performances of Ni/TiO –SiO (Et-x)
2 2
At first, the catalytic performance of Ni/TiO , Ni/TiO2meso
2
examine the air-stability, Ni/TiO –SiO (Et-0.6) was collected and
and Ni/TiO –SiO (Et-x) were compared for the hydrogenation of
2
2
2
2
stored in a beaker with exposing to air at room temperature for a
week, finally the activity and the stability were examined. As shown
in Table 3 and Fig. S3c, the initial reaction rate was slow in the 1st
run, and to reach complete conversion of NB needed 5 h, but in the
nitrobenzene in water. As the results displayed in Table 2, the order
of catalytic activity for the checked catalysts is Ni/TiO –SiO (Et-x),
2
2
meso
Ni/TiO2
to 0.6, the conversion of NB changed very less, but the selec-
tivity to aniline increased and simultaneously the selectivity of
NSB decreased. The highest selectivity over Ni/TiO –SiO (Et-0.6)
> Ni/TiO . With turning the molar ratio of Si/Ti from
2
0
2
nd run, the reaction rate increased largely in the first 1 h, and the
complete conversion of NB was achieved within 2.8 h. Most impor-
tantly, it could keep the similar activity in the following runs. It is
deduced that the surface Ni species was oxidized in somewhat as
exposed to air, but it could be reduced under the reaction condi-
tions with hydrogen in-situ, thus a higher activity was obtained in
2
2
with a Si/Ti of 0.6 reached 97%, which is much higher than that
meso
(
(
81.9%) of Ni/TiO2
without ethane-bridged organosilica linkage
entries 2–5). It indicates that the suitable amount of ethane-
bridged organosilica species could improve the transformation of
NSB to AN in H O. In addition, the reaction rates in H O were much
the following cycles. Therefore, Ni/TiO –SiO (Et-0.6) is an efficient
2
2
2
2
meso
non-noble metal catalyst with high stability to air and water. It is
much easier to storage and handle under the atmosphere compared
with the Raney nickel and conventional supported Ni catalysts.
higher than that in ethanol over Ni/TiO2
and Ni/TiO –SiO (Et-
2 2
0
.6) (ESI Table S1), indicating H O was an effective solvent for the
2
hydrogenation of NB as reported in literature [19–25].
Furthermore, the stability of Ni/TiO2meso, Ni/TiO –SiO (Et-0.2)
2
2
and Ni/TiO –SiO (Et-0.6) was examined and compared. During the
4. Discussion
2
2
recycling tests, the catalyst was reused directly without any treat-
ment after separation from the reaction solution, and the end of
reaction was judged by the cease of pressure dropping. As the
results shown in Table 3, Ni/TiO2meso deactivated clearly with a
4.1. Hydrophobicity of Ni/TiO –SiO (Et)
2
2
When we dispersed the catalysts into the reaction solution of
nitrobenzene and water, a biphasic solution was formed due to the
low solubility of nitrobenzene in water. Ni/TiO2
−1
−1
decrease in the reaction rate from 98 h (the 1st run) to 30 h
the 5th run), but it is better than Ni/TiO2 which deactivated very
meso
(
was dispersed

2
18
W. Li et al. / Journal of Molecular Catalysis A: Chemical 411 (2016) 214–221
Fig. 4. TEM images of fresh Ni/TiO2–SiO2(Et-0.6) (a), used Ni/TiO2–SiO2(Et-0.6) (b), fresh Ni/TiO2meso (c), and used Ni/TiO2meso (d), and Ni nanoparticle size distributions was
inserted into each TEM image.
Table 2
The catalytic performances of supported Ni catalysts in the hydrogenation of NB.
Sel. (%)a
NSB
Reaction rate (h )b
−1
Entry
Catalyst
Conv. (%)
AN
Others
1
2
3
4
5
Ni/TiO2
Ni/TiO2
Ni/TiO2–SiO2(Et-0.2)
Ni/TiO2–SiO2(Et-0.4)
Ni/TiO2–SiO2(Et-0.6)
49.0
73.5
72.7
72.8
74.0
5.6
17.0
12.5
5.9
93.2
81.9
85.7
93.8
97.0
1.2
1.1
1.8
0.3
0.4
125
188
186
186
189
meso
2.6
◦
Reaction conditions: NB: 3 mL, H2O: 3 mL, H2: 5 MPa, T: 80 C, NB/Ni = 256/1 (molar ratio), time: 1 h.
a
NSB and AN represents nitrosobenzene and aniline, respectively. Others include of N-phenylhydroxylamine (PHA), azoxybenzene (AOB), azobenzene (AB) and hydra-
zobenzene (HAB).
b
Reaction rate was calculated by the moles of NB converted per mol Ni per hour.

W. Li et al. / Journal of Molecular Catalysis A: Chemical 411 (2016) 214–221
219
Table 3
meso
Comparison of the reusability of Ni/TiO2
and Ni/TiO2–SiO2(Et-0.6).
Reaction rate (h 1)a
−
Catalyst
Runs
Time (h)
Conv. (%)
Sel. (%)
Ni/TiO2meso
1
2
3
4
5
2.5
3.5
4.5
7.0
8.5
95.3
99.9
99.9
97.0
99.9
92.3
99.9
99.9
98.8
99.8
98
73
57
35
30
Ni/TiO2–SiO2(Et-0.2)
Ni/TiO2–SiO2(Et-0.6)
1
2
3
4
5
2.2
2.5
3.0
3.0
3.0
99.6
98.4
99.9
99.2
99.9
93.2
95.1
98.9
98.0
98.0
116
101
85
85
85
1
2
3
4
5
2.0
99.9
99.6
98.4
99.0
98.9
99.6
98.3
96.1
97.9
98.0
128
113
103
103
103
2.25
2.45
2.45
2.45
Ni/TiO2–SiO2(Et-0.6) (exposed in air for 7 days)
1
2
3
4
5
6
5
2.8
2
2.3
2.8
2.8
99.9
99.9
99.9
99.9
99.9
99.9
95.8
98.4
99.0
99.8
99.2
98.3
51
91
128
111
91
91
◦
Reaction conditions: NB: 3 mL, H2O: 3 mL, H2: 5 MPa, T: 80 C, NB/Ni = 256/1 (molar ratio). The reaction was stopped when no more hydrogen consumed.
a
Reaction rate was calculated by the moles of NB converted per mol Ni per hour.
Fig. 5. Distribution of (a) Ni/TiO2meso
,
(b) Ni/TiO2–SiO2(Et-0.2) and (c)
Ni/TiO2–SiO2(Et-0.6) in the mixture of NB and H2O.
uniformly in the up water layer, while Ni/TiO –SiO (Et-0.2) and
2
2
Ni/TiO –SiO (Et-0.6) dispersed in the bottom nitrobenzene layer
2
2
as shown in the Fig. 5. It indicates that Ni/TiO –SiO (Et-x) cata-
2
2
lysts were hydrophobic due to incorporated with ethane-bridged
organosilica fragments. Ni/TiO –SiO (Et-x) showed similar reac-
2
2
meso
tion rate to Ni/TiO2
in the presence of water in Table 2 (entries
2
–5) as the reaction occurred on the interface of catalysts, the
interface area and environment are similar on these two catalysts
although they dispersed in different phases. While, the selectiv-
Fig. 6. WXRD patterns of the (a) fresh and (b) used Ni/TiO2, (c) fresh and (d) used
ity to AN was much higher over Ni/TiO –SiO (Et-x) than that on
meso
2
2
Ni/TiO2
, (e) fresh and (f) used Ni/TiO2–SiO2(Et-0.6). XRD patterns of the fresh
meso
Ni/TiO2
. The hydrophobic nature of the catalyst is favor to the
and used Ni/TiO2 reported in Ref. [27].
efficient adsorption and enrichment of organic substrates in the
nanopores of hydrophobic materix [33–35]. The NSB may be more
prone to adsorb on the hydrophobic Ni/TiO –SiO (Et-x), and then
◦
centred at 2␪ of 44.7, 52.1, and 76.6 presented for the fresh Ni/TiO ,
2
2
2
meso
convert to AN quickly.
Ni/TiO2
and Ni/TiO –SiO (Et-0.6) (Fig. 6a, c, e). For the used
2 2
Furthermore, the catalyst stability was improved signifi-
cantly after modified with ethane-bridged organosilica (Table 3).
Ni/TiO –SiO (Et-0.2) and Ni/TiO –SiO (Et-0.6) presented better
Ni/TiO , Ni(OH) diffraction peaks were detected (Fig. 6b) [27]. For
2 2
meso
the used Ni/TiO2
, the obvious diffraction peaks of NiOOH at 18.3
◦
◦
and 37.3 and the weak peaks of NiO at 37.2 and 43.3 were detected
2
2
2
2
meso
meso
stability compared to Ni/TiO2
. Ni/TiO2
showed a significant
(Fig. 6d). However, these diffraction peaks of Ni(OH) , NiOOH and
2
deactivation even though it has been improved largely compared to
NiO were not detected in the used Ni/TiO –SiO (Et-0.6) (Fig. 6f),
2
2
the Ni/TiO catalyst, with which the reaction cannot complete even
in the first run, as Ni(OH)2 was formed during the reaction [27].
From WXRD patterns in Fig. 6, the diffraction peaks of metallic Ni
indicating the hydrophobicity of catalyst could prohibit the con-
tacting of metallic Ni with water to form the Ni oxide/hydroxide
species, so the stability of Ni/TiO –SiO (Et-x) was enhanced. The
2
2
2

2
20
W. Li et al. / Journal of Molecular Catalysis A: Chemical 411 (2016) 214–221
0
to Ni , and 855.3 (Ni 2p3/2), 873.5 eV (Ni 2p1/2) were assigned
to Ni oxide/hydroxide, and the peaks at binding energies of 861.4
and 879.6 eV should be the satellite peaks of Ni oxide/hydroxide
[
36–38]. For the samples exposed to the air, the peaks of Ni
oxide/hydroxide was a little larger compared to the fresh ones as
0
Ni species were somewhat oxidized by air. However, after etch-
0
ing for 120 s (Fig. 7c, f), the peak of Ni at 851.6 eV became intense
and the peaks Ni oxide/hydroxide at 855.3 and 873.5 eV decreased
0
largely. That is to say, most of the Ni particles might be anchored
into the mesoporous channel and matrix, which could effectively
reduce the oxidation degree of Ni, and so Ni/TiO –SiO (Et-0.6) is
2
2
relative air-stable and could be stored in air and recycled suc-
cessfully. In addition, the mesoporous channels may favor the
meso
hydrogenation of NB, and the activity of Ni/TiO2
was much
higher than that of Ni/TiO (entries 1,2 in Table 2). The mesoporous
2
channels could somewhat protect the Ni particles from the corro-
meso
sion of water, Ni/TiO2
could be reused for five times (Table 3),
but Ni/TiO deactivated completely in the first run [27].
2
In summary, the enhanced hydrophobic property, high disper-
sion of the Ni particles into the mesoporous TiO –SiO (Et) as well as
2
2
their stronger interactions are benefit for the outstanding activity
and unexpected stability to air and water.
5. Conclusions
A novel mesostructured Ni/TiO –SiO (Et) catalyst with highly
2
2
dispersed Ni nanoparticles was successfully prepared. The
Ni/TiO2–SiO2(Et) catalysts exhibited high catalytic activity and sta-
bility toward the hydrogenation of nitrobenzene in water. These
unique catalytic performances of the Ni/TiO –SiO (Et) catalyst
Fig. 7. Ni 2p XPS spectra for the (a) fresh Ni/TiO2meso, (b) exposed Ni/TiO2meso
in air for 7 days, and (c) sample b was etched for 120 s, and for the (d) fresh
Ni/TiO2–SiO2(Et-0.6), (e) exposed Ni/TiO2–SiO2(Et-0.6) in air for 7 days, and (f)
sample e was etched for 120 s.
2
2
were mainly due to the incorporation of the hydrophobic group
ethane-bridged organosilica moieties) into TiO2 network, creat-
(
average sizes of Ni particles was about 15, 12 and 14 nm for the
ing a combination of increased hydrophobicity with well-defined
mesoporosity, which leads to a remarkable enhancement in the cat-
alytic activity and stability in water. The current studies confirmed
meso
fresh Ni/TiO –SiO (Et-0.6), Ni/TiO
and Ni/TiO , respectively,
2
2
2
2
which are in good agreement with the TEM results in Fig. 4. Ni par-
ticle size changed slightly after reaction (Fig. 4b, d), and the leaching
of active Ni species could be neglected based on the results of ICP
analysis.
that the as-prepared Ni/TiO –SiO (Et) is an excellent water-
2
2
tolerable and air-stable heterogeneous catalyst. It is expected to
have a wide application in the selective hydrogenation reactions in
water.
4
.2. Mesoporous structure of Ni/TiO –SiO (Et-x)
2 2
Acknowledgements
The enhanced catalytic activity should contribute to the perfect
textural properties of catalyst. As revealed by the characteriza-
tions, the mesoporous Ni/TiO –SiO (Et-x) had large BET surface
This work is supported by the Natural Science Fund Council
of China (NO. 21273222 and 21202159), and China postdoctoral
science foundation (NO. 2013M531000).
2
2
area, high pore volume as well as uniform pore size distribution.
Ethane-bridged organosilica fragments were introduced directly
and specifically into the TiO2 network, and participated in the
framework of TiO /SiO (Et) through Ti O Si C C Si covalent
Appendix A. Supplementary data
2
2
bonds, which clearly enlarged the pore size of the TiO framework
2
(
ESI Table S2). Moreover, the large pore diameter can minimize
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.molcata.2015.10.
023.
diffusion problem, whereas large BET surface area may give rise
to much more number of the available active sites. Thus, both the
larger pore structure and hydrophobicity ensure Ni/TiO –SiO (Et-
2
2
0
.6) to have higher activity and stability among all the tested
References
catalysts. As illustrated in Table 1, Ni/TiO –SiO (Et-0.6) exhibited
2
2
a decrease in pore volume but a similar pore size compared with
TiO –SiO (Et-0.6), suggesting that Ni nanoparticles not only dis-
[1] Y.F. Zheng, K. Ma, H.L. Wang, X. Sun, J. Jiang, C.F. Wang, R. Li, J.T. Ma, Catal.
Lett. 124 (2008) 268–276.
2
2
[
2] X.C. Meng, H.Y. Cheng, Y. Akiyama, Y.F. Hao, W.B. Qiao, Y.C. Yu, F.Y. Zhao, S.
Fujita, M. Arai, J. Catal. 264 (2009) 1–10.
persed onto the surface of TiO –SiO (Et-0.6) but also anchored
2
2
into the pore channels of TiO –SiO (Et-0.6), which was further
[3] Y.E. Wu, S.F. Cai, D.S. Wang, W. He, Y.D. Li, J. Am. Chem. Soc. 134 (2012)
975–8981.
4] S.A. Regenhardt, C.I. Meyer, T.F. Garetto, A.J. Marchi, Appl. Catal. A: Gen. 449
2012) 81–87.
[5] Y. Wang, X.Y. Chen, B. Yue, H.Y. He, Top. Catal. 55 (2012) 1022–1031.
2
2
8
confirmed by XPS analysis. The amount of Ni increased markedly
after etching (ESI Table S3) as the exposure of the part of Ni parti-
cles located in the pore channels, which should contribute to the
[
(
[
6] B. Legras, V.V. Ordomsky, C. Dujardin, M. Virginie, A.Y. Khodakov, ACS Catal. 4
(2014) 2785–2791.
air-stability of the catalysts. As the XPS spectra of Ni 2p shown
meso
in Fig. 7, Ni/TiO2
and Ni/TiO –SiO (Et-0.6) exposed in air for
2 2
[
7] C.M. Li, Y.D. Chen, S.T. Zhang, J.Y. Zhou, F. Wang, S. He, M. Wei, D.G. Evans, X.
Duan, ChemCatChem 6 (2014) 824–831.
7
days (Fig. 7b, e) exhibited the similar results to the fresh ones
(Fig. 7a, d), the peak at binding energy of 851.6 eV was assigned
[8] J.X. Chen, T. Guo, K.L. Li, L.M. Sun, Catal. Sci. Technol. 5 (2015) 2670–2680.

W. Li et al. / Journal of Molecular Catalysis A: Chemical 411 (2016) 214–221
221
[9] P.S. Rathore, R. Patidar, T. Shripathi, S. Thakore, Catal. Sci. Technol. 5 (2015)
[26] M. Jorge, M.N.D.S. Cordeiro, J. Phys.Chem. C 111 (2007) 17612–17626.
2
86–295.
[27] W.W. Lin, H.Y. Cheng, J. Ming, Y.C. Yu, F.Y. Zhao, J. Catal. 291 (2012) 149–154.
[28] W. Li, F.Y. Ma, F. Su, L. Ma, S.Q. Zhang, Y.H. Guo, ChemSusChem 4 (2011)
744–756.
[29] W. Li, F.Y. Ma, F. Su, L. Ma, S.Q. Zhang, Y.H. Guo, ChemCatChem 4 (2012)
1798–1807.
[
[
[
10] D.H. Kim, H.O. Seo, M.G. Jeong, Y.D. Kim, Catal. Lett. 144 (2014) 56–61.
11] Z. Yang, Y.B. Huang, Q.X. Guo, Y. Fu, Chem. Commun. 49 (2013) 5328–5330.
12] Z.M. Rong, Z.H. Sun, L. Wang, J.K. Lv, Y. Wang, Y. Wang, Catal. Lett. 144 (2014)
1
766–1771.
[
[
13] S.J. Tauster, Acc. Chem. Res. 20 (1987) 389–394.
14] A. Corma, P. Serna, P. Concepcion, J.J. Calvino, J. Am. Chem. Soc. 130 (2008)
[30] L. Zhao, G.S. Zhu, D.L. Zhang, Y. Di, Y. Chen, O. Terasaki, S.L. Qiu, J. Phys. Chem.
B 109 (2005) 764–768.
8
748–8753.
[31] J. Ming, C.Y. Wu, H.Y. Cheng, Y.C. Yu, F.Y. Zhao, J. Supercrit. Fluids 57 (2011)
137–142.
[32] J. Ming, H.Y. Cheng, Y.C. Yu, Y.Q. Wu, F.Y. Zhao, J. Mater. Chem. 21 (2011)
6654–6659.
[
[
[
[
15] B. Coq, A. Tijani, R. Dutartre, F. Figueras, J. Mol. Catal. 79 (1993) 253–264.
16] X.X. Han, R.X. Zhou, G.H. Lai, X.M. Zheng, Catal. Today 93–95 (2004) 433–437.
17] J. Xiong, J.X. Chen, J.Y. Zhang, Catal. Commun. 8 (2007) 345–350.
18] Y.Z. Xiang, X.N. Li, C.S. Lu, L. Ma, Q.F. Zhang, Appl. Catal. A: Gen. 375 (2010)
[33] F.J. Liu, L. Wang, Q. Sun, L.F. Zhu, X.J. Meng, F.S. Xiao, J. Am. Chem. Soc. 134
(2012) 16948–16950.
2
89–294.
[
19] J.B. Ning, J. Xu, J. Liu, H. Miao, H. Ma, C. Chen, X.Q. Li, L.P. Zhou, W.Q. Yu, Catal.
[34] L. Wang, J. Zhang, J. Sun, L.F. Zhu, H.Y. Zhang, F.J. Liu, D.F. Zheng, X.J. Meng,
X.D. Shi, F.S. Xiao, ChemCatChem 5 (2013) 1606–1613.
[35] L. Wang, F.S. Xiao, ChemCatChem 6 (2014) 3048–3052.
[36] Z.Y. Hou, O. Yokota, T. Tanaka, T. Yashima, Appl. Catal. A: Gen. 253 (2003)
381–387.
Commun. 8 (2007) 1763–1766.
[20] M. Pietrowski, M. Zielinski, M. Wojciechowska, Catal. Lett. 128 (2009) 31–35.
21] G.Y. Fan, L. Zhang, H.Y. Fu, M.L. Yuan, R.X. Li, H. Chen, X.J. Li, Catal. Commun.
[
1
1 (2010) 451–455.
[
[
22] K. Taya, Sci. Papers I.P.C.R. 56 (1962) 285–289.
23] J.H. Lyu, X.B. He, C.S. Lu, L. Ma, Q.F. Zhang, F. Feng, X.N. Li, J.G. Wang, Chin.
Chem. Lett. 25 (2014) 205–208.
[37] N. Mahata, A.F. Cunha, J.J.M. Orfao, J.L. Figueiredo, Appl. Catal. A: Gen. 351
(2008) 204–209.
[38] N. Mahata, A.F. Cunha, J.J.M. Orfao, J.L. Figueiredo, ChemCatChem 2 (2010)
330–335.
[
[
24] Y.F. Hao, R.X. Liu, X.C. Meng, H.Y. Cheng, F.Y. Zhao, J. Mol. Catal. A: Chem. 335
(
2011) 183–188.
25] X.C. Meng, H.Y. Cheng, S. Fujita, Y.C. Yu, F.Y. Zhao, M. Arai, Green Chem. 13
2011) 570–572.
(