Fine-tunable Ni@porous silica core-shell nanocatalysts: Synthesis, characterization, and catalytic properties in partial oxidation of methane to syngas

Article abstract of DOI:10.1016/j.jcat.2012.01.004

Ni nanoparticles (NPs) of narrow size distribution encapsulated inside meso- and microporous silica were prepared through in situ reduction of NiO NPs coated with silica. By varying preparation parameters, the mean size of Ni NPs can be fine-tuned in the

Full text of DOI:10.1016/j.jcat.2012.01.004

Journal of Catalysis 288 (2012) 54–64
Contents lists available at SciVerse ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Fine-tunable Ni@porous silica core–shell nanocatalysts: Synthesis, characterization,
and catalytic properties in partial oxidation of methane to syngas
Lei Li , Shengchao He , Yanyan Song , Jing Zhao , Weijie Ji , Chak-Tong Au b,
a
a
a
a
a,
⇑
⇑
a
Key Laboratory of Mesoscopic Chemistry, Ministry of Education, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing 210093, China
Department of Chemistry, Hong Kong Baptist University, Hong Kong, China
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Ni nanoparticles (NPs) of narrow size distribution encapsulated inside meso- and microporous silica were
prepared through in situ reduction of NiO NPs coated with silica. By varying preparation parameters, the
mean size of Ni NPs can be fine-tuned in the range 6–45 nm. It was found that with variation in core size,
microcapsular cavity, and shell porosity, the as-obtained Ni@meso-SiO catalysts for the partial oxidation
2
of methane to synthesis gas are notably different in catalytic activity and durability. The catalyst activity
and durability are essentially determined by the size of the Ni cores, and also somewhat by the porosity
Received 8 October 2011
Revised 2 January 2012
Accepted 7 January 2012
Available online 16 February 2012
Keywords:
of SiO
2
shells, as well as the extent of core–shell interaction, which is influenced by the microcapsular
catalyst with Ni NPs of ca. 6 nm and SiO shells with 3–4 nm
Ni@porous silica
Ni nanoparticles
Silica
Core–shell structure
Partial oxidation of methane
cavity structure. The Ni-350@meso-SiO
2
2
mesopores is superior in both activity and durability, giving CH
4
conversion of ꢀ93%, H
2
selectivity of
À1 À1
À1
92–93% (750 °C and GHSV = 72,000 mL g
h
), and TOFCH4 of 37.9 s
.
Ó 2012 Elsevier Inc. All rights reserved.
1
. Introduction
temperature was found to decrease over the Ni-based catalysts
that were modified by Pd, Pt, and Ru. In these studies, however,
the effects of dopant–support interaction and metal particle size
were strongly interdependent, making it difficult to differentiate
the role of the support from the effect of particle size.
In the reported studies, the catalysts were usually prepared by
impregnation followed by calcination/reduction. It is observed that
the size distribution of metal clusters varies with the kind of sup-
Production of synthesis gas (syngas) through the partial oxida-
tion of methane (POM) or catalytic oxy–methane reforming
COMR) is of great interest [1]. Compared with the traditional
method of steam reforming, POM has the following advantages:
i) syngas with an H /CO ratio of approximately 2 can be obtained,
(
(
2
ideal for processes such as methanol [2] and Fischer–Tropsch
synthesis [3], and (ii) the POM process is energetically more favor-
able because it is slightly exothermic. Usually, Ni and noble metals
such as Pt and Rh are used as active components for this reaction
port. Lu et al. reported that variation in Ag particle size on CaCO
3
showed rather different effects on ethylene and propylene epoxi-
dation [15] and that very small Au particles (ꢀ1 nm) supported
on mesoporous Ti-containing silicate are the most active for pro-
pylene epoxidation [16]. Recently, encapsulation of metal nanopar-
ticles (NPs) with a porous shell such as silica or alumina was found
to stabilize the metallic NPs against sintering at high temperatures
[17–21]. The NPs isolated inside inert shells have a relatively uni-
form environment around the core surface. Joo et al. reported that
as a model catalyst, Pt cores enwrapped inside a mesoporous silica
shell were thermally stable for ethylene hydrogenation and CO
oxidation [20]. It was reported that even after pretreatment in
air at 700 °C, the silica-encapsulated Pd catalysts were stable for
CO oxidation and acetylene hydrogenation [17,21]. Compared to
[
1,4,5].
Because of the high cost of noble metals, various Ni catalysts
have been developed for the POM process [6–10]. Huszar et al.
11] first reported that Ni supported on alumina was an efficient
catalyst for POM, and the yields were close to the equilibrated val-
ues. Besides Al , other materials such as SiO and MgO were also
adopted as support. Choudhary et al. [12,13] studied NiO deposited
on refractory supports precoated with MgO, CaO, SrO, BaO, Sm
and Yb . Despite low surface area, the Ni catalyst in which the
support was precoated with MgO and the as-fabricated catalyst
calcined at 1173 K exhibited high CH conversion and syngas
selectivity. Wang et al. [14] studied the effect of doping Ce, La,
and Ca on Ni/ -Al and Ni/ -Al catalysts and found that there
was enhanced CH conversion and CO selectivity. The ignition
[
2
O
3
2
2 3
O ,
2 3
O
4
2 2
a traditional Pd/SiO catalyst, the core–shell structured Pd@SiO
c
2
O
3
a
2
O
3
counterpart showed higher resistance toward deactivation and
could easily be regenerated without sintering of core particles over
multiple reaction–regeneration cycles. Li et al. [22] prepared
4
⇑
Corresponding authors. Fax: +86 25 83317761.
E-mail addresses: jiwj@nju.edu.cn (W. Ji), pctau@hkbu.edu.hk (C.-T. Au).
Pd@SiO
2
structures via a two-step procedure, resulting in a non-
shells. The catalyst,
uniform distribution of Pd size and thick SiO
2
0
021-9517/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2012.01.004

L. Li et al. / Journal of Catalysis 288 (2012) 54–64
55
however, was found to be highly active for 4-carboxybenzaldehyde
hydrogenation to p-toluic acid.
Takenaka et al. [23,24] reported that silica-coated Ni catalysts
showed high activity and improved stability in the steam reform-
deionized water, and the solution was added dropwise into a solu-
tion that contained 100 mL of deionized water and 330 mg of poly-
ethylene glycol (PEG) (average MW = 20,000, Fluka) and 1.0 g
NaOH. The resulting solution was stirred for 1 h at room tempera-
ture (RT). The collected material was then washed several times
with deionized water and ethanol, dried at 50 °C for 24 h, and cal-
cined in air at 350, 500, and 600 °C for 2 h. The corresponding sam-
ples are denoted hereafter as NiO-350, NiO-500, and NiO-600,
respectively.
ing of propane and POM. Compared with conventional Ni/Al
2 3
O ,
Ni/MgO, and Ni/SiO catalysts, core–shell structured Ni catalysts
2
show less carbon deposition. It was thought that the strong inter-
action of the Ni cores with the silica shell prevents Ni particles
from sintering as well as hindering carbon deposition. Our previ-
ous studies on core–shell type catalysts suggested that, compared
to the core–shell structures of close contact, the ones with cavities
generated between the core surface and the inner wall of the shell
via in situ reduction of oxide cores can serve as a microcapsule-like
reactor and enhance surface adsorption and catalytic reactivity
2
The core–shell structured NiO@meso-SiO samples were pre-
pared by a modified Stöber method. Briefly, 0.2 g of NiO NPs was
added to a poly-(vinylpyrrolidone) (PVP, K30, 1.0 g) ethanol solu-
tion of 100 mL. After the solution was stirred for 12 h, NH
3
2
Á H O
(25 wt%) of 10 mL was added, and then the suspension was soni-
cated for 30 min in an ultrasound cleaner (KQ-100DE, 40 kHz,
100 W). Subsequently, an ethanol solution (5 mL) of tetraethyl
orthosilicate (TEOS, P99%, Aldrich) of 0.1 mL was injected into
the suspension. One hour later, the product was collected by cen-
trifugation, washed twice with distilled water and ethanol, and
dried at 80 °C in air for 6 h. The core–shell samples with NiO cores
subject to calcination at 350, 500, and 600 °C are denoted hereafter
[
25–27].
Porous silica not only allows access of reactant molecules to
core particles, but also prevents core NPs from aggregating during
POM at high temperatures. It is possible for us to design a core–
shell type catalyst in which the size of metallic core particles can
be fine-tuned (depending on the size of the corresponding oxide
precursor) and retain a rather uniform size distribution (Schemes
1
A and B). The size (and hence size distribution) of particles can
essentially be maintained with time on stream owing to the pro-
tection of a stable SiO shell. This allows us to clearly elaborate
the relationship between particle size and activity of Ni catalysts
in the POM process. On the other hand, the porosity of the SiO
shell can be changed by adopting a specific surfactant such as
PVP during SiO encapsulation (Scheme 1C). Based on such a strat-
egy, we prepared core–shell structured Ni@SiO catalysts. The
as NiO-350@meso-SiO
SiO , respectively. These precursors were further subjected to
in situ reduction in a 25% H /Ar flow at 600 °C for 3 h to generate
2 2
, NiO-500@meso-SiO , and NiO-600@meso-
2
2
2
0
core–shell samples with Ni cores, which are denoted hereafter
as Ni-350@meso-SiO Ni-500@meso-SiO and Ni-600@meso-
SiO , respectively. The procedure for synthesizing the core–shell
structured NiO-350@micro-SiO and Ni-350@micro-SiO was
essentially the same, except that there was no application of PVP.
2
2
,
2
,
2
2
2
2
2
as-obtained core–shell catalysts were applied for POM at various
temperatures. The effect of gas hourly space velocity (GHSV) on
activity with time on stream was also studied. Techniques such
as nitrogen sorption measurement, X-ray diffraction (XRD), trans-
mission electron microscopy (TEM), hydrogen temperature-
2
.2. Catalyst characterization
N
2
sorption measurement was performed on a NOVA-2020
material physical structure determinator. Before measurement,
the sample was degassed at 300 °C for 3 h. The BET surface area
was calculated from a multipoint BET analysis of the nitrogen
adsorption isotherms. The X-ray powder diffraction (XRD) patterns
were recorded on a Philips X’Pert MPD Pro X-ray diffractometer
2 2
programmed reduction (H TPR) and desorption (H TPD), and
temperature-programmed oxidation (TPO) were used to character-
ize the fresh and used catalysts. A structure–performance relation-
ship has been established based on the results of characterization
and evaluation.
with graphite-monochromatized Cu K
a radiation (k = 0.1541 nm).
The TEM images were taken over a JEOL JEM-2000EX instrument
operated at 100 kV. The HRTEM images were obtained with a Phi-
lips Tecnai G220 operated at 200 kV. To make the TEM observa-
2
. Experimental
2.1. Catalyst preparation
tions on the H
reduced in a continuous-flow fixed-bed reactor under H
2
-reduced catalysts, the precursors were first
at
2
The NiO NPs were prepared by a chemical precipitation method.
600 °C for 2 h. After the samples were cooled to RT under Ar flow,
Typically, 2.9 g of Ni(NO
3
)
2
2
Á6H O was dissolved in 40 mL of
anhydrous ethanol was quickly introduced by means of a syringe
A
B
PVP
Calcined
Reduction
H2 Ar
O
NH3 H 2O TEOS
550 C
C
Scheme 1. Graphical illustration of the strategy for the fabrication of Ni-based core–shell catalysts: (A) development of microcapsule-like structure of Ni@SiO
size of Ni(O) cores; and (C) tunable porosity of SiO shell (left: with PVP addition, right: without PVP addition).
2
; (B) tunable
2

5
6
L. Li et al. / Journal of Catalysis 288 (2012) 54–64
pump, with Ar as a carrier gas, until the reduced samples were
completely immersed. H TPR was performed in a conventional
setup equipped with a thermoconductive detector. In typical runs,
0 mg of NiO@SiO was heated to 120 °C at 10 °C/min under Ar
the (111), (200), and (220) planes at 2h = 44°, 52°, and 76° (JCPDS
No. 04-0850) (Fig. 1B). Note that the crystallinity of Ni is also af-
0
2
fected by that of NiO: the higher calcination temperature of NiO re-
0
3
2
sults in higher crystallinity of Ni . XRD patterns of the used
flow of 50 mL/min and kept at this temperature for 1 h to remove
catalysts suggest the presence of fibrous carbon species, most obvi-
adsorbed water. After the sample cooled down to RT and was
2
ously in the case of Ni-500@meso-SiO (Fig. 2). Claridge et al. re-
switched to a 25% H
temperature was programmed to 700 °C at 10 °C/min. H
was carried out to estimate surface exposure of Ni cores. The sam-
ple (0.1 g) was heated to 550 °C in a flow of 25% H /Ar at a rate of
0 °C/min, and kept at this temperature for 2 h. After cooling to
0 °C, the sample was purged with He gas for 2 h and then heated
TPD curve was re-
corded. TPO experiments were conducted using a TGA analyzer
SDT Q 600) with an online mass spectrometer to examine the
amount of carbon deposited on the catalysts. Approximately
0 mg of used catalyst was heated in a pure O flow (40 mL/min)
2
/N
2
(v/v, 50 mL/min) mixture, the sample
ported that there was generally carbon deposition on supported
Ni catalysts during the POM reaction [28].
2
TPD
2
3.1.2. (HR)TEM
1
6
Fig. 3 illustrates the TEM images and size distribution of NiO-
3
50, NiO-500, and NiO-600. Clearly, the size of NiO NPs is obvi-
to 500 °C at a rate of 10 °C/min while the H
2
ously affected by the calcination temperature. Based on the results
of measuring 50–100 randomly distributed NiO particles, the par-
ticle size distributions, the mean sizes, and the standard deviations
can be derived. The mean size is 6.3 nm with a standard deviation
of 2.1 nm for NiO-350, while mean sizes of 21.0 and 48.7 nm with
standard deviations of 3.8 and 8.1 nm, respectively, are determined
for NiO-500 and NiO-600.
(
1
2
from RT to 800 °C at a heating rate of 10 °C/min, while the CO
2
signal was recorded.
2
To achieve uniform SiO encapsulation and avoid formation of
2.3. Catalyst evaluation
individual silica particles, the conditions for encapsulation have
to be properly controlled. Graf et al. [29] reported a method for
coating colloidal particles with a silica layer using PVP as a cou-
pling agent. The amphiphilic and nonionic PVP polymer can be ad-
sorbed onto the colloidal particles. The stabilized colloidal particles
are then transferred into an ammonia solution of ethanol and di-
rectly coated with homogeneous silica shells of variable thickness,
using TEOS as the Si source. Similarly, we have the NiO NPs of
Catalyst evaluation was conducted in a continuous-flow fixed-
bed reactor operated at atmospheric pressure. The catalyst
0.05 g) was placed in a quartz tube reactor. Before being tested,
the catalyst was reduced in situ at 600 °C in a flow of 25% H /Ar
40 mL/min) for 3 h. The feed mixture (CH /O /N = 2:1:3) was
then introduced at a total flow rate of 60 mL/min, giving a corre-
(
2
(
4
2
2
3
À1 À1
sponding gas hourly space velocity (GHSV) of 72,000 cm g
h .
different sizes enwrapped rather homogeneously in SiO
Fig. 4A–C). The statistical results on shell thickness were obtained
over 50 randomly distributed core–shell structures (Fig. 4a–c). The
shell thickness of NiO-350@meso-SiO , NiO-500@meso-SiO , and
NiO-600@meso-SiO is found to be 14.7 ± 1.8, 24.4 ± 3.8, and
6.0 ± 5.0 nm, respectively. It is observed that the shell thickness
2
shells
The reaction products were measured online by a gas chromato-
graph (GC122). The catalysts were also evaluated at different tem-
peratures and hourly space velocities.
(
2
2
2
3
. Results and discussion
4
increases in parallel with increasing core particle size.
Naked metal NPs are normally air-sensitive, and the introduc-
tion of protective layers such as silica onto them can increase their
3
3
.1. Characterization
.1.1. XRD
chemical stability. H
with silica layers will generate M@SiO
core entities [26,30]. In our study, through a simple reduction
treatment of NiO@SiO at 600 °C for 3 h under H , a corresponding
Ni@SiO sample can be obtained. Park et al. [31] prepared Co@SiO
2
reduction of metal oxide (MO
x
) NPs covered
Shown in Fig. 1A are the XRD patterns of the NiO NPs obtained
2
NPs without sintering of
by calcining the nickel hydroxide precipitate at different tempera-
tures. After calcination, the samples show only the fcc-NiO phase
2
2
(
JCPDS No. 78-0429), with typical reflections of the (111), (200),
2
2
and (220) planes at 2h = 37°, 43°, and 64°, respectively. With high-
er calcination temperatures, the crystallinity of NiO particles in-
creases. After in situ reduction, the NiO entities were completely
yolk-shell nanocatalysts and found contraction in volume and
decrease in diameter due to the removal of lattice oxygen from
0
CoO to generate Co . Such phenomena are quite common when
0
transformed into elemental Ni , as revealed by the reflections of
x 2 2
MO @SiO is transformed to M@SiO (M = Fe, Co, Ni, Ru) under
H
2
reduction [25–27]. TEM images of Ni-T@meso-SiO
2
are shown
a
NiO
(200)
NiO-350
NiO-500
(111)
(220)
(311)
∇− C
− Νi
NiO-600
a
∇
b
(111)
Ni
(220)
b
c
Ni-350@meso-SiO
2
(200)
Ni-500@meso-SiO
2
Ni-600@meso-SiO
2
d
1
0
20
30
40
50
60
70
80
10
20
30
40
50
60
70
80
2 °
θθ /
Fig. 1. XRD patterns of (a) NiO NPs calcined at different temperatures and (b) core–
shell Ni@meso-SiO after reduction.
Fig. 2. XRD patterns of catalysts used: (a) Ni-350@meso-SiO
SiO , (c) Ni-500@meso-SiO , and (d) Ni-600@meso-SiO
2
, (b) Ni-350@micro-
2
2
2
2
.

L. Li et al. / Journal of Catalysis 288 (2012) 54–64
57
4
3
2
1
0
0
0
0
0
NiO-350
a
2
4
6
8
10
12
Diameter/nm
6
5
4
3
2
1
0
0
0
0
0
0
NiO-500
b
0
1
2
16
20
24
28
Diameter/nm
5
4
3
2
0
c
NiO-600
0
0
0
1
0
0
3
0
35
40
45
50
55
60
65
70
Diameter/nm
Fig. 3. (A–C) TEM images and (a–c) size distributions of the NiO NPs calcined at different temperatures: (A, a) NiO-350, (B, b) NiO-500, and (C, c) NiO-600.
in Fig. 5A–C. A comparatively larger cavity between the core sur-
face and the inner wall of the shell is evidenced (Fig. 5C), owing
to the reduction of larger NiO-600 particles. The statistics of Ni par-
ticle size for core–shell catalysts have been measured over 50 ran-
domly distributed Ni particles (Fig. 5a–c). The mean particle size of
reaction temperatures. However, it is very difficult to realize
100% single-particle encapsulation; there is a certain degree of
multiparticles encapsulation for the Ni@SiO
catalysts. Within capsules containing multiple particles, the Ni
particles are incompletely isolated by SiO walls, and thus could
aggregate to some extent at high reaction temperatures. Statistical
analysis of Ni particle size in the used Ni-350@meso-SiO and
Ni-500@meso-SiO catalysts (Fig. 6a and b) indicated that the in-
2
core–shell structured
2
Ni in Ni-350@meso-SiO
SiO is 6.0 ± 2.0, 20.0 ± 3.7, and 45.1 ± 8.6 nm, respectively (Table
). The standard deviation increases with increasing mean particle
2 2
, Ni-500@meso-SiO , and Ni-600@meso-
2
2
1
2
size. The difference in Ni particle size of three representative
catalysts, however, is still statistically significant.
crease in particle size is minor (from 6.0 ± 2.0 to 7.9 ± 2.4 nm) for
the former and nonsignificant (from 20.0 ± 3.7 to 26.4 ± 5.2 nm)
Fig. 6A and B shows the TEM images of the core–shell struc-
2
for the latter. Therefore, most SiO -encapsulated Ni particles
tured catalysts after the POM reaction at 700 °C for approximately
essentially maintain their particle size, and a small fraction of
encapsulated multiparticles could experience slight aggregation
or sintering.
5
00 min. Ideally, the presence of a SiO
2
shell can isolate the Ni
particles, thus preventing them from sintering, especially at high

5
8
L. Li et al. / Journal of Catalysis 288 (2012) 54–64
80
70
60
50
40
30
20
10
0
NiO-350@meso-SiO2
a
1
1
12
13
14
15
16
17
18
19
Diameter/nm
6
5
4
3
2
1
0
0
0
0
0
0
0
b
NiO-500@meso-SiO2
1
8
20
22
24
26
28
30
32
34
Diameter/nm
7
6
5
4
3
2
1
0
NiO-600@meso-SiO2
c
0
0
0
0
0
0
0
3
6
42
48
54
60
Diameter/nm
Fig. 4. (A–C) TEM images and (a–c) distributions of shell thickness of the NiO-T@SiO
00@meso-SiO
2 2 2
samples: (A, a) NiO-350@meso-SiO , (B, b) NiO-500@meso-SiO , and (C, c) NiO-
6
2
.
3
.1.3. N
There is generally a certain amount of PVP surfactant remaining
inside the NiO@SiO precursor. If this is the case, it will be unfavor-
able for diffusion of reactant and product molecules across the SiO
2
sorption
addition, there are mesopores developed in the SiO
pore size distributions of NiO-350@meso-SiO and NiO-350@mi-
cro-SiO are shown in Fig. 7B. Clearly, the use of PVP greatly
enhances the generation of mesopores in the ranges 3–4 and
40–50 nm. Chou et al. [32] suggested that the mesopores of
3–4 nm are essentially generated due to the template effect of
PVP, while those of 30–50 nm are due to interstices between the
primary particles. Included in Table 1 are the BET surface area
2
shells. The
2
2
2
2
shells. To remove PVP surfactant, the precursor was calcined in air
at 550 °C for 3 h to make sure that the meso- and/or micropores
throughout the SiO
of the calcined NiO@SiO
The adsorption–desorption isotherms (Fig. 7A) indicate that the
NiO-350@meso-SiO (with PVP addition) and NiO-350@micro-
SiO (without PVP addition) show type IV and type I capillary con-
densation, respectively. The observation confirmed that with PVP
2
shells are not blocked by PVP. The porosity
2
was measured by means of N sorption.
2
2
and pore volume of the NiO@SiO samples. Based on the statistics
2
on core particle size and shell thickness, it can be deduced that the
BET surface area of the sample is co-influenced by core particle size
and shell thickness. With increasing Ni particle size, the
2

L. Li et al. / Journal of Catalysis 288 (2012) 54–64
59
6
5
4
3
2
1
0
0
0
0
0
0
0
a
Ni-350@meso-SiO2
2
4
6
8
10
12
Diameter/nm
60
50
40
30
20
10
0
b
Ni-500@meso-SiO2
1
2
16
20
24
28
Diameter/nm
5
4
3
2
1
0
0
0
0
0
0
c
Ni-600@meso-SiO2
2
5
30
35
40
45
50
55
60
Diameter/nm
Fig. 5. (A–C) TEM images and (a–c) size distribution of Ni NPs in the Ni-T@SiO
2
samples: (A, a) Ni-350@meso-SiO
2
2 2
, (B, b) Ni-500@meso-SiO , and (C, c) Ni-600@meso-SiO .
Table 1
Characteristics of NiO@SiO
2
samples determined by N
2
sorption, XRD, and TEM measurements.
Metal oxide loading (wt%)a
2
3
Ni (nm)b
D
Shell (nm)c
Samples
Si/Ni molar ratio (nominal)
Surface area (m /g)
Pore volume (cm /g)
D
NiO-350@meso-SiO
NiO-350@micro-SiO
2
0.67
0.67
0.67
0.67
54.8
55.2
56.1
55.8
117
71
64
0.125
–
0.132
0.128
6.0 ± 2.0
–
20.0 ± 3.7
45.1 ± 8.6
14.7 ± 1.8
–
24.4 ± 3.8
46.0 ± 5.0
2
NiO-500@meso-SiO
NiO-600@meso-SiO
2
2
63
a
b
c
Measured by XRF.
Measured over 50–100 randomly distributed Ni particles.
Measured over 50 randomly distributed core–shell structures.
contribution of core particles to surface area declines, while that of
SiO shells increases (due to parallel increase in shell thickness).
Moreover, the porosity of SiO
area of samples; for instance, the particle size of NiO cores is
2
shells can also affect the surface
2

6
0
L. Li et al. / Journal of Catalysis 288 (2012) 54–64
6
5
4
3
2
1
0
0
0
0
0
0
0
Ni-350@meso-SiO2
a
5
6
7
8
9
10 11 12 13 14 15
Diameter/nm
4
3
2
1
0
0
0
0
0
Ni-500@meso-SiO2
b
1
2
16
20
24
28
32
36
40
Diameter/nm
2 2
Fig. 6. (A and B) TEM images and (a and b) size distribution of Ni particles in the used catalysts (700 °C for 500 min): (A, a) Ni-350@meso-SiO and (B, b) Ni-500@meso-SiO .
identical for NiO-350@meso-SiO
2
and NiO-350@micro-SiO
2
, while
2 4
there was the formation of Ni–Mg–O solid solution and NiAl O
2
the former has much greater surface area (117 m /g) than the
spine structure, which in turn affected the reduction behavior of
2
2+
latter (71 m /g), owing to the higher porosity of the SiO
NiO-350@meso-SiO (Table 1).
2
shell of
Ni species. For the calculation of TOF value, it is necessary to
2
know the number of surface Ni sites per unit mass of catalyst,
which can be determined by means of H
chiometry of H/Ni = 1:1 [26,39,40]. Fig. 8B shows the typical H
TPD profile of Ni-350@meso-SiO . The peak area of H TPD was
normalized by that of H TPR of a standard CuO sample. The
amount of hydrogen atoms adsorbed onto Ni-350@meso-SiO
was found to be 7.3 mol/gcat (Table 2). Based on the number of
2
TPD with adsorption stoi-
2
3
.1.4. H
Fig. 8A illustrates the H
2
TPR/TPD
2
2
2
2
TPR profiles of the NiO@SiO samples.
2
The broad hydrogen consumption peaks in the range 200–500 °C
can be assigned to the reduction of NiO species [33]. It is apparent
that the reduction of NiO cores becomes difficult with increasing
NiO particle size. Compared to the conventional supported Ni cat-
2
l
Ni sites and activity data, one can figure out the TOFCH4 of
À1
3
7.9 s over the Ni-350@meso-SiO
2
catalyst for the POM reaction
alyst, the core–shell structured catalyst showed a broader H
peak [34]. Boudjahem et al. [35] reported that H -treatment at
00 °C can cause the generation of defects in silica shells. The
2
TPR
(Table 2).
2
4
hydrogen molecules or atoms can adsorb onto these defects and
possibly migrate onto the NiO core particles, a spillover-like phe-
nomenon usually observed over supported metal systems. With a
rise in the size of Ni NPs, the peak temperature shifts toward the
higher end. As reported by Takenaka et al. [24,36], compared to
the supported Ni catalysts, the core–shell structured Ni counter-
parts are difficult to reduce, as a result of the stronger core–shell
interaction. It was observed by Wang et al. [37] that the reduction
temperature of NiO increased with increasing crystal size in the
range 6–17 nm, which may also explain the rise of reduction tem-
perature, along with the increment of NiO core particle size in the
present study. On the other hand, we observed that the shell thick-
ness also increases with increasing core particle size, and the thick-
3.1.5. TPO
Coke deposition is an important issue for the POM reaction over
the Ni-based catalysts. Thus, we studied the used core–shell cata-
lysts by means of TPO to gain insight into this matter, and the
results are depicted in Fig. 9. From the TPO profiles, one can see
2
that there are CO peaks in the range 450–700 °C. Chen et al.
[41] reported that the temperature for the burning of carbon
species increases with increasing degree of graphitization. The
amorphous carbon species were found to be burnt out beyond
300 °C in air, while the graphitized carbon species would be burnt
at ca. 700 °C. In view of these observations, the peaks/shoulders
observed at low and high temperatures can be ascribed to amor-
phous and graphitized carbon species, respectively [42]. Our
er SiO
reduction. The TPR studies on Ni/MgO and Ni/Al
culty in reduction of dispersed NiO [6,13,38], and in those cases,
2
shells may cause difficulty in hydrogen diffusion and NiO
2
results indicated that Ni-500@meso-SiO showed comparatively
more pronounced coke deposition. Li et al. reported that Ni particle
size played an important role in methane decomposition [33]. It is
2
O
3
showed diffi-

L. Li et al. / Journal of Catalysis 288 (2012) 54–64
61
A 100
A
NiO-350@meso-SiO2
NiO-350@micro-SiO2
8
6
4
2
0
0
0
0
0
d
c
b
a
2
00
300
400
500
600
700
0
.0
0.2
0.4
0.6
0.8
1.0
Temperature/°C
Relative Pressure (P/P0)
1
1
.2
.0
B
B
0.30
Ni-350@meso-SiO2
NiO-350@micro-SiO2
NiO-350@meso-SiO2
0
0
0
0
0
0
.25
.20
.15
.10
.05
.00
0.8
.6
0
0.4
0.2
0
.0
-0.2
1
00
200
300
400
500
1
10
100
Temperature/°C
Pore width/nm
Fig. 8. (A) H
NiO-500@meso-SiO
350@meso-SiO
2
TPR profiles of (a) NiO-350@meso-SiO
2
, (b) NiO-350@micro-SiO
2
, (c)
Fig. 7. (a) Adsorption–desorption isotherms and (b) pore size distribution of
NiO@SiO nanostructures after calcination in air at 550 °C for 3 h.
, and (d) NiO-600@meso-SiO ; and (B) H TPD profile of Ni-
2
2
2
2
2
.
speculated that the cleavage of C–H bonds and polymerization of C
species on the Ni surfaces are structure-sensitive (i.e., dependent
on the variation of Ni particle size). In the present study, it seems
Compared with Ni-350@meso-SiO , Ni-350@micro-SiO2 shows
similar methane conversion (ca. 90%), but notably lower selectivity
2
to H . This may be ascribed to the difference in the local environ-
2
that there is more notable coke deposition on Ni-500@meso-SiO
2
ment of Ni sites resulting from the difference in core–shell interac-
with Ni particle size 20.0 ± 3.7 nm. On the other hand, Takenaka
et al. reported that when the coated Ni strongly interacted with
the silica shell, there was no sufficient room to grow the fibrous
carbon [24]. This may also explain the observations in our case that
tion as well as in molecule diffusion (mesoporosity in the shell of
Ni-350@meso-SiO2 vs. microporosity in the shell of Ni-350@mi-
cro-SiO ). Similar observations were made over the core–shell
2
structured Fe catalysts for NH3 decomposition [25,43]. It is also
the cramped cavity in Ni-350@meso-SiO
the roomy cavity in Ni-500@meso-SiO
growth. Nevertheless, the influence of cavities of microcapsule-like
nanostructure on carbon deposition needs further study.
2
suppressed coking while
could result in carbon
observed that the catalysts show the following sequence in
2
terms of durability: Ni-350@meso-SiO ꢁ Ni-350@micro-SiO >
2
2
2 2
Ni-500@meso-SiO > Ni-600@meso-SiO .
À1 À1
Based on the average H2 formation rates (mol h
g
) within
cat
5
00 min reaction over different core–shell catalysts (Fig. 10D), a
clear sequence in catalyst activity can be obtained: Ni-
350@meso-SiO > Ni-350@micro-SiO > Ni-500@meso-SiO > Ni-
3
3
.2. POM catalytic activity
2
2
2
6
2
00@meso-SiO . Compared to a conventional supported Ni catalyst
.2.1. Size effects of Ni cores and porosity effects of SiO
Illustrated in Fig. 10A–C are CH
2
shells
with identical Ni loading evaluated under the same reaction condi-
4
conversion and CO and H
2
tions (Fig. 11), the Ni-350@meso-SiO catalyst also showed notably
2
selectivities with time on stream at 700 °C in POM over the core–
shell catalysts. Repeated activity measurements were made and
the error bars were determined and added to the reaction data.
Most of the data points derived on various core–shell catalysts
show reasonable deviations within a reaction period of 500 min.
better activity and stability.
3.2.2. Effect of GHSV
The effect of GHSV on the catalytic activity of Ni-350@meso-
The averaged values of CH
00-min reaction were also calculated and are presented in the in-
sets of Fig. 10A–C. Both H and CO selectivities showed similar
trends with time on stream, and the H selectivity was found to
4
conversion and CO (H
2
) selectivity in
SiO
2
was investigated and the results are presented in Fig. 12. Note
À1 À1
5
that with a rise in GHSV from 36,000 to 72,000 mL g
only nonsignificant reduction in CH
average), while a further rise in GHSV to 108,000 mL g
h , there is
2
4
conversion (<2.5% on the
À1 À1
2
h
re-
be only slightly lower than that of CO. It is observed that the cata-
lytic activity is obviously dependent on the mean size of Ni cores.
With rise in the mean size of Ni NPs from 6 to 45 nm, there is a
sults in a notable decline in CH conversion (>10%). Therefore,
4
0
the optimized GHSV for POM over Ni-350@meso-SiO
2
is
À1 À1
72,000 mL g
4 2
h , at which CH conversion of ca. 90% and H
gradual decline in CH
4
conversion and selectivities to CO and H
2
.
selectivity of ꢀ88% can be obtained at 700 °C. Such performance

6
2
L. Li et al. / Journal of Catalysis 288 (2012) 54–64
Table 2
Catalytic activity, amount of H atoms adsorbed, and TOFCH4 over Ni-350@meso-SiO
.a
2
À1
Sample
CH
4
conversion (%)
R
CH4 (mol/gcat h)
Amount of H atoms
Surface Ni exposure
(atoms/gcat
TOFCH4 (s
37.9
)
adsorbed (
l
mol/g)
)
18
Ni-350@meso-SiO
2
93.0
1.00
7.3
4.4 Â 10
a
À1 À1
Catalytic activity was measured at 750 °C and GHSV = 72,000 mL h
TPD result.
g ; amount of H atoms adsorbed and surface Ni exposure are calculated based on the H
2
is better than that of the supported Ni-based catalysts, and the
superiority is likely due to the unique microcapsular structure
[
27,44]. Pelletier et al. [45] reported CH
4
conversion of 80.0% in
À1
POM at 750 °C and GHSV of 5400 h over Ni catalysts supported
on
Slightly lower CH
POM over Li O- and La
c
-Al
2
O
3
, as well as on Al
conversion was observed by Shao et al. [46] in
-doped Ni/Al catalysts under similar
2 3
O modified by aluminum phosphate.
4
2
2
O
3
2 3
O
conditions. The supported Ni catalysts generally deactivate quickly
with time on stream. The effect of doping certain element(s) to our
core–shell structured catalysts is currently under investigation.
c
b
a
3.2.3. Effect of temperature
Fig. 13 shows the effect of reaction temperature on catalyst
3
00
400
500
600
700
800
performance. With a rise in temperature from 650 to 750 °C, both
CH conversion and H selectivity increase steadily. CH conversion
of ꢀ93% and H
Ni-350@meso-SiO
Temperature/°C
4
2
4
Fig. 9. TPO profiles of the Ni@SiO
2
catalysts used in POM reaction at 700 °C for
selectivity of 92–93% were obtained over
at 750 °C and GHSV = 72,000 mL g
2
À1 À1
À1 À1
5
SiO
00 min (catalyst amount = 50 mg, GHSV = 72,000 mL h
, (b) Ni-350@micro-SiO , and (c) Ni-500@meso-SiO .
2 2
g ): (a) Ni-350@meso-
2
h . The
2
catalyst is stable even at a higher temperature of 750 °C. The
A 100
B
100
9
8
7
6
5
4
0
0
0
0
0
0
9
8
7
0
0
0
9
9
8
8
7
7
5
0
5
0
5
0
9
5
90
85
80
75
a
b
c
d
a
b
c
d
60
50
40
7
0
a
b
c
d
a
b
c
d
0
100
200
300
400
500
0
100
200
300
400
500
Time on stream (min)
Time on stream (min)
1
00
C
D
16
9
8
7
6
5
4
0
0
0
0
0
0
14
1
2
0
8
6
4
2
0
1
9
9
5
0
a
b
c
d
85
8
0
75
7
0
a
b
c
d
0
100
200
300
400
500
a
b
c
d
Time on stream (min)
Fig. 10. (A) CH
4
conversion, (B) CO selectivity, (C) H
2
selectivity with time on stream in POM over core–shell catalysts at 700 °C, and (D) averaged H
2
formation rates within
5
N
00 min reaction: (a) Ni-350@meso-SiO
2
, (b) Ni-350@micro-SiO
g .
2
, (c) Ni-500@meso-SiO , and (d) Ni-600@meso-SiO , respectively. Catalyst amount = 0.050 g, CH /O /
2
2
4
2
À1 À1
2
= 2:1:3 (v/v/v), GHSV = 72,000 mL h

L. Li et al. / Journal of Catalysis 288 (2012) 54–64
63
1
00
a
100
9
8
7
6
5
0
0
0
0
0
9
8
0
0
Ni-350@meso-SiO2
Ni/SiO2
70
1
00
200
300
400
500
6
7
7
50°C
00°C
50°C
6
5
4
0
0
0
1
00
9
8
7
6
5
0
0
0
0
0
0
100
200
300
400
500
Ni-350@meso-SiO2
Time on stream (min)
Ni/SiO2
b 100
1
00
200
300
400
500
Time on stream/min
90
2 2
Fig. 11. Comparison of catalyst performance of Ni-350@meso-SiO and Ni/SiO
8
7
6
5
4
0
0
0
0
0
with identical Ni loadings under the same reaction conditions: T = 750 °C,
À1 À1
GHSV = 72,000 mL h
g
.
650°C
00°C
50°C
a
100
7
7
9
8
7
6
5
0
0
0
0
0
0
100
200
300
400
500
Time on stream (min)
Fig. 13. (a) CH
Ni-350@meso-SiO
4
conversion and (b) H
2
selectivity with time on stream in POM over
À1 À1
2
at different temperatures for GHSV = 72,000 mL h
g .
perature. Once again, the results demonstrate that the core–shell
structured Ni-350@meso-SiO catalyst with Ni NPs of 6 nm is
promising in terms of activity and stability for POM.
2
0
50
100
150
200
250
300
4
. Concluding remarks
Time on stream (min)
Silica-encapsulated Ni NPs with controllable size distribution
(6–45 nm) were synthesized and applied for POM to syngas. The
core–shell structured catalysts exhibit a narrow size distribution
b 100
2
of Ni NPs and meso- or microporosity of SiO shells. With increas-
ing core size, the precursor NiO NPs are difficult to reduce. It is
clear that the catalyst activity is dependent on the size of the Ni
cores, the porosity of the SiO shells, and the extent of core–shell
2
interaction. The microcapsular structure of the core–shell catalysts
may be favorable for the reaction. The optimized Ni-350@meso-
9
8
7
6
5
0
0
0
0
0
9
9
8
8
7
7
5
0
5
0
5
0
SiO
2
catalyst is more active, giving CH
selectivity of 92–93% at GHSV = 72,000 mL g
4
conversion of ꢀ93% and
À1 À1
H
2
h
(750 °C). It
is also more durable, showing no decline in CH
4
conversion and
H
2
selectivity at 750 °C within a test period of 500 min.
Acknowledgments
3
6,000
72,000
108,000
0
50
100
150
200
250
300
The financial support from NSFC (21173118), NSFJS
(BK2011439), and the Ministry of Education for a special discipline
Time on stream (min)
grant of
a doctoral supervisor (20110091110023) is greatly
Fig. 12. Effect of GHSV on (a) CH
4
conversion and (b) H
2
selectivity in POM at 700 °C
appreciated.
over Ni-350@meso-SiO : GHSV = 36,000 (j), 72,000 (d), and 108,000 (N)
2
À1 À1
mL h
g .
References
2
compositions of H and CO in the effluent gas at 750 °C were found
[
1] T.V. Choudhary, V.R. Choudhary, Angew. Chem. Int. Ed. 47 (2008) 1828.
to be similar to the thermodynamic equilibrium values at this tem-
[2] P. Tijm, F. Waller, D. Brown, Appl. Catal. A 221 (2001) 275.

6
4
L. Li et al. / Journal of Catalysis 288 (2012) 54–64
[
3] G.P. Van Der Laan, A. Beenackers, Catal. Rev. 41 (1999) 255. [24] S. Takenaka, H. Umebayashi, E. Tanabe, H. Matsune, M. Kishida, J. Catal. 245
[
[
4] B.C. Enger, R. Lodeng, A. Holmen, Appl. Catal. A 346 (2008) 1.
5] P. Prieto, A. Ferreira, P. Haddad, D. Zanchet, J. Bueno, J. Catal. 276 (2010) 351–
(2007) 392.
[25] Y. Li, L. Yao, S. Liu, J. Zhao, W. Ji, C.T. Au, Catal. Today 160 (2010) 79.
[26] L. Yao, Y. Li, J. Zhao, W. Ji, C. Au, Catal. Today 158 (2010) 401.
[27] Y. Li, L. Yao, Y. Song, S. Liu, J. Zhao, W. Ji, C.T. Au, Chem. Commun. 46 (2010)
5298.
[28] J.B. Claridge, M.L.H. Green, S.C. Tsang, A.P.E. York, A.T. Ashcroft, P.D. Battle,
Catal. Lett. 22 (1993) 299.
3
59.
[
[
[
6] K. Takehira, T. Shishido, P. Wang, T. Kosaka, K. Takaki, J. Catal. 221 (2004) 43.
7] C. Berger-Karin, J. Radnik, E.V. Kondratenko, J. Catal. 280 (2011) 116.
8] T. Shishido, M. Sukenobu, H. Morioka, M. Kondo, Y. Wang, K. Takaki, K.
Takehira, Appl. Catal. A 223 (2002) 35.
[
9] Z. Hou, J. Gao, J. Guo, D. Liang, H. Lou, X. Zheng, J. Catal. 250 (2007) 331.
[29] C. Graf, D.L.J. Vossen, A. Imhof, A. van Blaaderen, Langmuir 19 (2003) 6693.
[30] A.H. Lu, E.L. Salabas, F. Schuth, Angew. Chem. Int. Ed. 46 (2007) 1222.
[31] J.C. Park, H.J. Lee, H.S. Jung, M. Kim, H.J. Kim, K.H. Park, H. Song, Chem. Cat.
Chem. 3 (2010) 755.
[32] K.S. Chou, C.C. Chen, Micropor. Mesopor. Mater. 98 (2007) 208.
[33] Y. Li, B.C. Zhang, X.W. Xie, J.L. Liu, Y.D. Xu, W.J. Shen, J. Catal. 238 (2006) 412.
[34] R.Y. Xie, D.B. Li, B. Hou, J.A. Wang, L.T. Jia, Y.H. Sun, Catal. Commun. 12 (2011)
380.
[35] A. Boudjahem, S. Monteverdi, M. Mercy, M. Bettahar, J. Catal. 221 (2004) 325.
[36] S. Takenaka, E. Kato, Y. Tomikubo, K. Otsuka, J. Catal. 219 (2003) 176.
[37] C. Wang, G. Gau, S. Gau, C. Tang, J. Bi, Catal. Lett. 101 (2005).
[38] J. Requies, M. Cabrero, V. Barrio, M. Guemez, J. Cambra, P. Arias, F. Pérez-
Alonso, M. Ojeda, M. Pena, J. Fierro, Appl. Catal. A 289 (2005) 214.
[39] T. Choudhary, C. Sivadinarayana, D. Goodman, Catal. Lett. 72 (2001) 197.
[40] X.K. Li, W.J. Ji, J. Zhao, S.J. Wang, C.T. Au, J. Catal. 236 (2005) 181.
[41] L. Chen, Y. Lu, Q. Hong, J. Lin, F. Dautzenberg, Appl. Catal. A 292 (2005) 295.
[42] H. Zdemir, M. Faruk ksüz mer, M. Ali Gürkaynak, Int. J. Hydrogen Energy 35
(2010) 12147.
[
10] Y. Kobayashi, J. Horiguchi, S. Kobayashi, Y. Yamazaki, K. Omata, D. Nagao, M.
Konno, M. Yamada, Appl. Catal. A 395 (2011) 129.
11] K. Huszar, G. Racz, G. Szekely, Acta. Chim. Acad. Sci. 70 (1971) 287.
12] V. Choudhary, B. Uphade, A. Mamman, Catal. Lett. 32 (1995) 387.
13] V. Choudhary, B. Uphade, A. Mamman, J. Catal. 172 (1997) 281.
14] H. Wang, Z. Li, S. Tian, React. Kinet. Catal. 83 (2004) 245.
[
[
[
[
[
[
15] J. Lu, J.J. Bravo-Suarez, A. Takahashi, M. Haruta, S.T. Oyama, J. Catal. 232 (2005) 85.
16] J. Lu, X. Zhang, J.J. Bravo-Suárez, K.K. Bando, T. Fujitani, S.T. Oyama, J. Catal. 250
(
2007) 350.
[
17] A.J. Forman, J.N. Park, W. Tang, Y.S. Hu, G.D. Stucky, E.W. McFarland, Chem. Cat.
Chem. 2 (2010) 1318.
18] J.C. Park, J.U. Bang, J. Lee, C.H. Ko, H. Song, J. Mater. Chem. 20 (2009) 1239.
19] J.N. Park, P. Zhang, Y.S. Hu, E.W. McFarland, Nanotechnology 21 (2009)
25708.
20] S.H. Joo, J.Y. Park, C.K. Tsung, Y. Yamada, P. Yang, G.A. Somorjai, Nat. Mater. 8
2008) 126.
21] J.N. Park, A.J. Forman, W. Tang, J.H. Cheng, Y.S. Hu, H.F. Lin, E.W. McFarland,
Small 4 (2008) 1694.
[
[
2
[
[
(
[43] Y. Li, S. Liu, L. Yao, W. Ji, C.T. Au, Catal. Commun. 11 (2010) 368.
[44] J. Lee, J.C. Park, H. Song, Adv. Mater. 20 (2008) 1523.
[45] L. Pelletier, D.D.S. Liu, Appl. Catal. A 317 (2007) 293.
[46] Y. Chen, W. Zhou, Z. Shao, N. Xu, Catal. Commun. 9 (2008) 1418.
[
[
22] K.T. Li, M.H. Hsu, I. Wang, Catal. Commun. 9 (2008) 2257.
23] S. Takenaka, Y. Orita, H. Umebayashi, H. Matsune, M. Kishida, Appl. Catal. A
3
51 (2008) 189.