Synthesis and catalytic properties of nickel-silica composite hollow nanospheres

Article abstract of DOI:10.1021/jp049754g

A coating process was applied to prepare nickel-silica composite hollow nanospheres (650 nm) with controllable shell thickness. The nanospheres were characterized through X-ray diffraction, transmission electron microscopy, and field emission scanning electron microscopy. This material with large BET surface area (288 m²/g) exhibits good catalytic activity and high selectivity in acetone hydrogenation reaction, showing the potential application as a catalyst in industry.

Full text of DOI:10.1021/jp049754g

J. Phys. Chem. B 2004, 108, 6311-6314
6311
Synthesis and Catalytic Properties of Nickel-Silica Composite Hollow Nanospheres
Pu Jin,† Qianwang Chen,*,† Liqing Hao,‡ Ruifen Tian,‡ Lixin Zhang,† and Lin Wang†
Structure Research Laboratory and Department of Materials Science and Engineering, and Department of
Chemical Physics, UniVersity of Science and Technology of China, Hefei 230026, China
ReceiVed: January 16, 2004; In Final Form: March 14, 2004
A coating process was applied to prepare nickel-silica composite hollow nanospheres (650 nm) with
controllable shell thickness. The nanospheres were characterized through X-ray diffraction, transmission electron
microscopy, and field emission scanning electron microscopy. This material with large BET surface area
(288 m2/g) exhibits good catalytic activity and high selectivity in acetone hydrogenation reaction, showing
the potential application as a catalyst in industry.
Introduction
SCHEME 1: Schematic Illustration for the Preparation
of Nickel-Silica Composite Hollow Nanospheresa
Recently, monodisperse nanospheres, core-shell nanostruc-
tures, and hollow nanospheres with uniform physical and
chemical properties have received much attention because they
have potential applications, such as photonic materials and
microchip reactors.1-6 The common route for preparing hollow
nanospheres is to coat nanocrystals on the template spheres
followed by removing the templates through etching or calcina-
tions.7 Colvin et al. had successfully developed such a route to
produce macroporous metals such as three-dimensional nickel
films with pore diameters between 200 and 1000 nm. The porous
nickel formed in that way is over 70% airsresulting with large
surface area and sturdy structure that was suggested to be useful
in wide fields ranging from sensors to catalysis.8,9 It is also well-
known that the shape, size, and utility or properties are
inextricably linked. The architectures designed and fabricated
from appropriate nanoscale building blocks, including the use
of void space, show some new opportunities for improving the
performance of the catalyst.10
In this paper, we reported a process for synthesizing nickel-
silica composite hollow nanospheres with needlelike nickel
particles embedded in a silica shell. The mesoporous and
macroporous structures present in these spheres give the
potential applications in catalyst or an active carrier for catalyst.
This type of catalytic architecture differs from standard im-
pregnated catalysts that are described as a metal “flea” riding
on oxide “boulder” in some literature.10-12 In our process the
amount of metal particles loaded on the silica carriers is much
larger.
a The graphs are actual TEM images of the samples obtained at each
step.
28% ammonia as catalyst. The mixture was stirred mildly for 1
h, and the product weighed about 1.2 g after washing and drying.
The whole synthesis process is illustrated briefly in Scheme
1. First, 1.2 g of SiO2 spheres were dispersed in 100 mL of
solution containing 0.15 g of Ni(NO3)2‚6H2O and 0.3 g of CO-
(NH2)2. The suspension was stirred mildly at a constant
temperature of 95 °C for 4 h to precipitate completely, and the
system was about pH 8 at the end of the reaction. The process
for preparing uniform shells on silica cores by homogeneous
precipitation of slowly decomposing urea in nickel nitrate
solution was repeated several times,12 which controls the
thickness of the uniform layers. By immersing the core-shell
powder in 200 mL of 0.5 mol/L NaOH, the silica cores were
dissolved by alkaline and the remainder shells were washed by
distilled water until the pH of filtrate was lower than pH 7.0.
After being dried in a vacuum stove, these powders were
reduced in H2 current (40 mL min-1) at 450 °C for 1 h in the
fixed-bed reactor and the products were kept in N2 current for
10 h after the temperature decreased to 60 °C, which would
prevent most of the product from forming oxide when exposed
to air.
Experimental Section
The chemical composition of the hollow capsules was
analyzed by X-ray fluorescence spectrum and dimethyl glyoxime
gravimetric analysis. The phases of the mediate and ultimate
product were also determined by X-ray diffraction (XRD). The
morphology and microstructure of the hollow spheres were
observed through transmission electron microscopy (TEM) and
scanning electron microscopy (SEM). BET surface area was
obtained by N2 adsorption/desorption isotherm data, while the
hydrogen adsorption on the samples was investigated by
pressure-composition-temperature (PCT) measurement at a
temperature of 25 °C and under hydrogen pressure ranging from
0.01 to 40 atm, which is the normal measurement for hydrogen
storage.14
All the reagents used in this investigation were analytical
grade and used without further purification. The monodisperse
silica nanospheres were prepared according to the process
developed by Sto¨ber et al.13 This method comprises two steps:
the hydrolysis and polycondensation of tetraethoxysilane (TEOS)
under alkaline conditions in ethanol. The starting reagents
contained 6 mL of TEOS, 62 mL of ethanol, and 12 mL of
* To whom correspondence should be addressed. E-mail:
cqw@ustc.edu.cn.
† Structure Research Laboratory and Department of Materials Science
and Engineering.
‡ Department of Chemical Physics.
10.1021/jp049754g CCC: $27.50 © 2004 American Chemical Society
Published on Web 04/27/2004

6312 J. Phys. Chem. B, Vol. 108, No. 20, 2004
Jin et al.
TABLE 1: X-ray Fluorescence Spectrum and X-ray
Photoelectron Spectroscopy Quantitative Analysis Results of
Nickel-Silica Composite Hollow Spheres
(a) X-ray Fluorescence Spectroscopy
element
at. %
element
at. %
Ni
Si
Na
Al
57.9
37.4
2.3
K
0.5
0.4
0.3
0.2
Ca
Fe
S
1.0
(b) X-ray Photoelectron Spectroscopy
Figure 1. XRD patterns of different nanospheres: (a) freshly coated
silica spheres, showing the formation of Ni3Si2O5(OH)4; (b) silica/nickel
core-shell structure, reduction at 550 °C before removing silica,
showing the presence of a large amount of silica with a broad diffraction
band peaking at 26.2°; (c) nickel-silica composite hollow spheres, with
reduction at 450 °C after removing silica, showing the decrease of the
amount of silica.
peak ID
at. %
peak ID
at. %
C 1s
O 1s
8.53
56.21
Ni 2p3
Si 2p
16.61
18.65
that the newly prepared silica would interact with the nickel
more easily than those after aging for several days. To decrease
this strong interaction, the silica cores were removed before
reduction, which made the shells more easily to be reduced even
at 450 °C. As shown in Figure 1c, the XRD pattern of the sample
depicts a significant decrease in the intensity of the diffraction
peak at 26.2°, indicating this process easily removes silica cores.
According to the Scherrer’s equation, the particle size of nickel
crystallites calculating from the peak Ni(111) is about 12.8 nm.
The X-ray fluorescence spectrum data show that the atomic ratio
of Ni to Si is 57.9 to 37.4. Furthermore, from the XPS peaks of
Ni 2p3 and Si 2p, it could be also known that Ni to Si is 16.61
to 18.65 on the surface of the shell. The impurities and its
contents were shown in Table 1a, which might be induced in
experiment or analysis process. The great amount of C and O
in XPS data (Table 1b) might be brought by the CO2 and O2
adsorbing on the active surface of nickel-silica composite
hollow nanospheres. All these data, especial those from dimethyl
glyoxime gravimetric analysis, prove that the weight of nickel
is more than 40% in the final product. It is therefore called the
product nickel-silica composite hollow nanospheres due to
certain amounts of silica still exist on the inner surface of the
hollow spheres as matrixes for Ni nanoparticles to form shells.
After the coating process, the surface morphology changed
a lot. Figure 2a shows the TEM image of the bare silica
nanospheres with smooth surface and spherical shape. Figure
2b shows clearly the core-shell structure of silica/Ni3Si2O5-
(OH)4 with the shell thickness about 40 nm; the thickness could
be controlled by the times of coating, and each coating might
increase the thickness for 5 nm. The shells consist of needlelike
particles as revealed by TEM image (Figure 3b). Because of
the formation of a new compound of Ni3Si2O5(OH)4 and the
needlelike structure on the core-shell interface, the shells began
to separate from the cores with increasing thickness of the shells.
The TEM image reveals that the sheetlike phase did not contact
with the silica core closely (Figure 2b), which might cause this
separation. Just because of this separation, it has much prob-
ability to remove the core without destroying the shell. To keep
the spherical shape of the shell, the concentration of strong base
should be carefully controlled to avoid distortion of the shell.
Because of the support of the core, the shell did not collapse
when these powders were dispersed by ultrasound. After
reduction the morphology of the shells (Figure 3b,c) did not
change when compared to those before the reaction (Figure 2b).
In Figure 2c, the bright areas inside the spheres indicate that
the SiO2 cores were removed after treatment in alkaline solution,
while the surface morphology did not change significantly. The
SEM image reveals that the shells are composed of nanoparticles
with an average size of 15 nm integrating closely to form an
uniform layer (Figure 3a), and the broken shells in the inset of
The catalytic activity of this material and the activity as carrier
for noble metal catalysts were studied by acetone hydrogenation
reaction at 373, 423, and 473 K.16 A 1.0 wt % amount of Pd
was precipitated on the pure silica nanospheres and the
composite hollow spheres, respectively. At atmospheric pressure,
0.30 g of catalysts was reduced according to the above method
but without N2 treatment. The packed bed reactor was a
cylindrical tube with the diameter of 1 cm, and the length of
the part where the catalyst (0.30 g) lied was nearly 2.5 cm. Then
the hydrogenation reactions were carried out directly without
taking the catalysts out of the reactor to keep the clean surface
of the catalysts. When at reaction, the experiments were carried
out in the current of the mixture of acetone and H2 instead of
pure H2. The volume ratio of acetone to H2 was about 1:4, and
the H2 current was 120 mL min-1. The reaction time lasted for
70 min if the acetone was 10 mL. The products were collected
at 0 °C and analyzed by GC-MS.
Results and Discussion
Through the hydrolysis and condensation of TEOS under
alkaline condition, monodisperse silica nanospheres with an
average diameter of about 400 nm were obtained. These silica
nanospheres will be utilized as cores, the templates, to form
uniform shells by homogeneous precipitation of slowly decom-
posing urea in nickel nitrate solution. The diffraction peak from
amorphous SiO2 was around 2θ ) 26.2°, which can be clearly
observed in patterns a and b of Figure 1, but was not so obvious
in c because the silica cores were removed by alkaline solution.
In pattern a, the peaks for (200) and (060) are aroused from
monoclinic nickel hydrosilicate, a thermal stable substance at
500 °C.
It is reported that only nickel hydrosilicate would be obtained
on the surface of silica spheres due to the interactions between
silica spheres and nickel particles, and this effect was more
obvious at high temperatures.12 This shell could not be reduced
completely even at 550 °C, though increasing temperature might
favor this process.12 To confirm the temperature of heat
treatment, nickel hydroxide without silica was synthesized by
the same method. TG analysis shows there is an obvious weight
loss at 320 °C, which might correspond to the decomposition
of hydroxide. The decomposed product was face-centered cubic
NiO. However, because nickel hydrosilicate was formed, this
transformation in the core-shell structure was not noticed even
though the powder was calcined at 500 °C for 5 h. The X-ray
diffraction patterns in Figure 1 a,b show that Ni3Si2O5(OH)4
and metallic nickel formed on the surface of silica nanospheres
after coating and further reducing, respectively. It was also found

Nickel-Silica Composite Hollow Nanospheres
J. Phys. Chem. B, Vol. 108, No. 20, 2004 6313
Figure 2. TEM images of (a) bare silica spheres and (b) silica/Ni3Si2O5(OH)4 core-shell structures; (c) Ni3Si2O5(OH)4 hollow spheres with an
average size of 650 nm. Nickel-silica composite hollow nanospheres with shell thickness about 40 nm.
Figure 3. SEM image of (a) Ni3Si2O5(OH)4 hollow nanospheres. Inset shows broken hollow nanospheres. TEM images of nickel-silica hollow
nanospheres prepared by reduction after removing silica cores (b) and removing silica cores after reduction (c).
Figure 4. N2 adsorption/desorption isotherm (a) and pore diameter
Figure 5. PCT measurement of samples: (a) normal nickel particles
distribution (b) of nickel-silica composite hollow spheres.
prepared by the same coprecipitation and reduction process without
silica cores; (b) nickel-silica composite hollow nanospheres without
Figure 3a also confirm that the spheres are hollow. The SEM
image also shows that these layers are of porous structure
(Figure 3a), which is much different from the smooth interface
of the bare silica nanospheres. It is expected that this porous
structure would be favorable for the transportation of H2 and
H2O during reduction.
It might be due to magnetic interaction between nickel
nanoparticles; most of the nickel-silica layers did not break
up after heating and cooling.
N2 treatment; (c) nickel-silica composite hollow nanospheres being
treated with N2 for 12 h after reduction.
and corresponds to that of other catalyst carriers such as Al2O3
and SiO2. The pore volume of the sample is 0.396 mL g-1, and
the mean pore diameter is 19.53 nm. A majority of the pores
have a diameter less than 4 nm, which might exist in silica
matrix; other pores with sizes from 10 to 60 nm might come
from the aggregates of Ni nanoparticles.
The surfaces of the nanoshells prepared by removing the cores
after the reduction process were smoother than those prepared
by reduction of hollow nickel hydrosilicate. Such a surface
change did not occur in the latter process so that the porous
surface was reserved. To reduce the shell more completely and
keep the porous surface that is believed to possess a larger
surface area, the latter approach is suitable.
The PCT measurement results (Figure 5) show that the normal
nickel particles have the lowest H2 adsorption capacity under
the same condition, while the nickel-silica composite hollow
spheres display better H2 adsorption properties. The difference
between the latter two samples is that sample b was taken out
of the reactor without the N2 treatment after cooling to room
temperature with part of the nickel being oxidized in air. The
N2 treatment at room temperature could remove the H2
adsorbing on the nickel surface and keep the high surface
activity so that sample c has the highest hydrogen storage
capacity in the PCT measurement.
The shape of the N2 adsorption/desorption isotherm (Figure
4) of nickel-silica hollow nanospheres further reveals the
characteristic of their porosity. The BET surface area is about
288 m2 g-1, which is larger than that of Raney Ni (95 m2 g-1 15
)

6314 J. Phys. Chem. B, Vol. 108, No. 20, 2004
Jin et al.
TABLE 2: Catalytic Effect on the Acetone Hydrogenation Conversion and Selectivity at 473 K over (a) 1% Pd/Silica, (b) 40%
Ni/Silica Hollow Spheres, and (c) 1% Pd, 40% Ni/Sicilca Hollow Sphere Catalysts (Acetone ) 7.5 mL h-1; Acetone:H2 (molar)
) 1:4; Weight of the Catalyst ) 0.30 g)
conversion (%)
423 K
selectivity at 473 K (%)
IBMC
catalyst
373 K
473 K
2-P
IBMK
DIPE
others
a
b
c
5.2
78
77
9.6
75
68
29.6
70
71
99.0
82.7
75.6
0
13.8
5.4
0
3.0
1.5
0
0
8.2
1.0
0.5
9.3
The catalytic activity of this material and the function as a
carrier for noble metal catalysts were evaluated by the acetone
hydrogenation reaction at 373, 423, and 473 K.16 The reactions
were carried out at atmospheric pressure in a fixed-bed reactor.
The main products of 2-propanol (2-P), isobutyl methyl ketone
(IBMK), and isobutyl methyl carbinol (IBMC) were collected
at 0 °C and analyzed by GC-MS. At 473 K, all the catalysts
showed a certain activity at this temperature; therefore the
selectivities are listed and compared in Table 2. With a small
amount of catalyst, a minor H2/acetone ratio, and high velocity
of acetone, this material exhibits relatively high conversion and
good selectivity. It was also found that some disopropyl ether
(DIPE) appeared in the products with 1% Pd and 40% Ni/SiO2
as the catalyst, but it was not detected in other products.
Acknowledgment. This work was supported by the National
Natural Science Foundation of China (Grants 20125103 and
90206034).
References and Notes
(1) Polman, A.; Wiltzius, P. MRS Bull. 2001, 26, 608.
(2) Jiang, P.; Cizeron, J.; Bertone, J. F.; Colvin, V. L. J. Am. Chem.
Soc. 1999, 121, 7957.
(3) Gau, H.; Herminghaus, S.; Lenz, P.; Lipowsky, R. Science 1999,
283, 46.
(4) Caruso, F.; Caruso, R. A.; Mohwald, H. Science 1998, 282, 1111.
(5) Hu, Y.; Chen, J. F.; Chen, W. M. Lin, X. H.; Li, X. L. AdV. Mater.
2003, 15, 726.
(6) Dong, A. G.; Ren, N.; Tang, Y.; Wang, Y. J.; Zhang, Y. H.; Hua,
W. M.; Gao, Z. J. Am. Chem. Soc. 2003, 125, 4976.
(7) Jiang, P.; Bertone, J. F.; Colvin, V. L. Science 2001, 291, 453.
(8) Kulinowski, K. M.; Jiang, P.; Vaswani, H.; Colvin, V. L. AdV.
Mater. 2000, 12, 833.
Conclusions
(9) Yan, H. W.; Blanford, C. F.; Holland, B. T.; Parent, M.; Smyrl,
W. H.; Stein, A. AdV. Mater. 1999, 11, 1003.
(10) Rolison, D. R. Science 2003, 299, 1698.
(11) Takahashi, R.; Sato, S.; Sodesawa, T.; Kato, M.; Takenaka, S.;
Yoshida, S. J. Catal. 2001, 204, 259.
(12) James, T. R.; Regis, J. D. J. Catal. 1978, 54, 207.
(13) Sto¨ber, W.; Fink, A.; Bohn, E. J. Colloid Interface Sci. 1968, 26,
62.
(14) Chen, J.; Li, S. L.; Tao, Z. L.; Shen, Y. T.; Cui, C. X. J. Am. Chem.
Soc. 2003, 125, 5284.
(15) Shim, J.; Lee, H. K. Mater. Chem. Phys. 2001, 69, 72.
(16) Narayanan, S.; Unnikrishnan, R. J. Chem. Soc., Faraday Trans.
1998, 94, 1123.
The above studies illustrate a route for synthesizing nickel-
silica composite hollow nanospheres. The nanospheres are 650
nm in outer diameter, with a thickness of 40 nm. The shells
were found to be composed of nickel nanoparticles with an
average size of 15 nm. The as-prepared hollow nanospheres
exhibit high catalytic activity and good selectivity in acetone
hydrogenation reaction, showing potential applications as nano-
catalysts in the future. It is expected that further work could
lead to the as-prepared nickel-silica composite hollow nano-
spheres being used in industry.