Shape effect of ceria in Cu/ceria catalysts for preferential CO oxidation

Article abstract of DOI:10.1016/j.molcata.2010.11.017

Copper was deposited on different shapes of ceria supports (i.e., rods, cubes, and octahedra) and used as catalysts for preferential CO oxidation in excess amounts of hydrogen. When the same amount of copper was deposited, the copper content on the surface measured by X-ray photoelectron spectroscopy differed significantly, with more copper on the ceria octahedra. Copper seemed to migrate into the bulk ceria to a greater degree on the rods. The Cu/ceria-octahedra showed the highest activity of 95% at 140 °C among the three shapes, whereas the Cu/ceria-rods showed higher CO conversion than the Cu/ceria-octahedra at higher temperatures. The Cu/ceria-octahedra showed no activity degradation for CO conversion at 140 °C over 100 h, whereas the activity decreased by 13% for Cu/ceria-rod and 32% for Cu/ceria-cube at the same temperature. The metals Au and Pt were also deposited on the different shapes of ceria, and their activity and selectivity were evaluated.

Full text of DOI:10.1016/j.molcata.2010.11.017

Journal of Molecular Catalysis A: Chemical 335 (2011) 82–88
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Shape effect of ceria in Cu/ceria catalysts for preferential CO oxidation
a
b
b
a,∗
Jaeman Han , Hyung Jun Kim , Sangwoon Yoon , Hyunjoo Lee
a
Department of Chemical and Biomolecular Engineering, The Specialized Graduate School of Hydrogen & Fuel Cell, Yonsei University, Seoul 120-749, South Korea
Department of Chemistry, Dankook University, Gyeonggi 448-701, South Korea
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 3 September 2010
Received in revised form
Copper was deposited on different shapes of ceria supports (i.e., rods, cubes, and octahedra) and used as
catalysts for preferential CO oxidation in excess amounts of hydrogen. When the same amount of copper
was deposited, the copper content on the surface measured by X-ray photoelectron spectroscopy differed
significantly, with more copper on the ceria octahedra. Copper seemed to migrate into the bulk ceria to
a greater degree on the rods. The Cu/ceria-octahedra showed the highest activity of 95% at 140 C among
the three shapes, whereas the Cu/ceria-rods showed higher CO conversion than the Cu/ceria-octahedra
1
1 November 2010
Accepted 14 November 2010
Available online 21 November 2010
◦
at higher temperatures. The Cu/ceria-octahedra showed no activity degradation for CO conversion at
Keywords:
Ceria
Shape
◦
1
40 C over 100 h, whereas the activity decreased by 13% for Cu/ceria-rod and 32% for Cu/ceria-cube at
the same temperature. The metals Au and Pt were also deposited on the different shapes of ceria, and
their activity and selectivity were evaluated.
Copper
PROX
Long-term stability
© 2010 Elsevier B.V. All rights reserved.
1
. Introduction
Catalytic properties of activity and selectivity can be con-
and polyhedra [10]. The rod shape of a ceria nanocrystal has {1 1 0}
and {1 0 0} surfaces, the cubic shape has a {1 0 0} surface, and the
polyhedral shape mainly has a {1 1 1} surface. CO conversion was
◦
trolled by changing the shapes of the catalysts [1,2]. In the case
of metallic catalysts, different shapes of nanoparticles have dis-
tinct surface atomic arrangements, such as a square arrangement
for a {1 0 0} surface, a hexagonal arrangement for a {1 1 1} sur-
face, or a grooved arrangement for a {1 1 0} surface. The change
in surface atomic arrangement causes differences in the geom-
etry and binding strength of reactants adsorbed on the surface,
resulting in differences in activity and selectivity [3–5]. Metal
oxide catalysts with various surface crystalline structures also have
different formation energies for oxygen vacancy, leading to differ-
ences in oxidation activity. Ceria with different surface crystalline
structures, in particular, have been actively investigated [6,7]. The
formation and migration of oxygen vacancy and CO adsorption have
been estimated for ceria by theoretical simulation [8,9]. For exam-
ple, when CO is adsorbed on {1 1 0}, {1 0 0}, and {1 1 1} surfaces, a
CO molecule bridges two surface oxygen atoms on the surface of
evaluated within a range of 150–350 C. The rod with the smallest
oxygen vacancy formation energy showed the highest activity in
all ranges of temperature.
Hydrogen, which is the main energy source for fuel cells, has
been produced from the water–gas shift reaction or the steam refor-
mation of fossil fuel. In this process, a small amount of CO inevitably
remains in the product gas. The residual CO severely poisons the
platinum catalyst inside the fuel cell assembly. Therefore, remov-
ing CO from hydrogen gas prior to fuel cell operation is essential,
and this is usually achieved by the preferential oxidation (PROX)
reaction. In the PROX reaction, CO is selectively oxidized to CO2 in
the presence of excessive amounts of hydrogen. Because the oxi-
dation of hydrogen into water is thermodynamically favorable at
higher temperatures, the selective CO oxidation can occur at low
temperatures [11]. The noble metal catalysts Au, Pt, and Ru have
been used to achieve this effect, but high costs hinder their practical
use [12–14]. Recently, Cu/ceria was reported to show good results
for PROX reactions, and the effects of dopant and nanostructure
have been actively investigated [15–20].
{
1 1 0} or {1 0 0}, pulling the atoms out of the lattice sites and form-
ing CO species. In this case, the surface structure is deformed due
3
to mild reduction. However, a CO molecule is weakly bound to the
{
1 1 1} surface, preserving the surface structure [9]. The effect of
In this study, copper was deposited on three different shapes of
ceria nanocrystals: rods, cubes, and octahedra. The activity, selec-
tivity and long-term stability of Cu/ceria catalysts with various ceria
shapes were evaluated for the PROX reaction. Copper was deposited
more on the surface for octahedra while copper seemed to migrate
more into lattice for rod. Although oxygen vacancies was formed
more on ceria rod, Cu/ceria-octahedra with more surface copper
ceria shape has been investigated for the water–gas shift reaction
(
H O + CO → H + CO ) by using gold-deposited ceria rods, cubes,
2
2
2
∗ Corresponding author. Tel.: +82 2 2123 5759; fax: +82 2 312 6401.
E-mail address: azhyun@yonsei.ac.kr (H. Lee).
1
381-1169/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2010.11.017

J. Han et al. / Journal of Molecular Catalysis A: Chemical 335 (2011) 82–88
83
3d 3/2
3d 5/2
a
4Cu/CeO2(oct)
4Cu/CeO2(cube)
4Cu/CeO2(rod)
920
910
900
890
880
870
Binding Energy(eV)
2p 1/2
2p 3/2
b
4
4
4
Cu/CeO2(oct)
Cu/CeO2(cube)
Cu/CeO2(rod)
920
970
960
950
940
930
Binding Energy(eV)
Fig. 2. X-ray photoelectron spectroscopy showing (a) Ce 3d peaks and (b) Cu 2p
peaks for 4 wt% Cu/ceria catalysts.
other shapes demonstrated significant degradation in catalytic
activity.
2. Experimental
2.1. Preparation of catalysts
Different shapes of ceria were synthesized by modifying a previ-
ously reported recipe [21]. 1.125 mmol of Ce(NO ) ·6H O (Kanto)
3
3
2
and 22.5 ml of NaOH (Duksan) solution (6 M for rods; 6 M for cubes;
.01 M for octahedra) were mixed and continuously stirred for
0
about 30 min. Next, a Teflon bottle containing this mixture was
held in a stainless steel autoclave, and the autoclave was sealed
tightly. The autoclave was subjected to hydrothermal treatment at
◦
◦
◦
different temperatures (100 C for rods; 180 C for cubes; 180 C
for octahedra) for 24 h. The obtained precipitates were separated
◦
by centrifugation, washed with DI water, and then dried at 80 C in a
vacuum oven overnight. The powders were calcined in air at 400 C
◦
for 4 h. To obtain the Cu/ceria, ceria powder (0.1 g) was slurried in
3
ml of deionized (DI) water. Then a 1, 4 or 10 wt% aqueous solu-
tion of Cu(NO ) ·3H O (Aldrich) was added to the above solution
3
2
2
dropwise and was continuously stirred for 1 h. Next, 1 M of Na CO
2
3
(
Aldrich) solution (molar ratio; Cu precursor:Na CO = 1:50) was
2 3
Fig. 1. TEM images of CeO2 with (a) rod shape, (b) cubic shape, and (c) octahedral
shape (the insets are HRTEM images).
added dropwise to the solution and stirred for 1 h. The pH value
was kept at 8–10. The obtained precipitates were then separated
by centrifugation, washed with DI water at 60–70 C, and then dried
at 80 C in a vacuum oven overnight. The powders were calcined
in air at 400 C for 4 h. The metal precursor and precipitating agent
◦
and oxygen vacancies similar to Cu/ceria-rod showed the best activ-
ity for PROX reaction performed at relatively low temperatures
◦
◦
◦
below 150 C. Additionally, the effect of ceria shape on a long-
were HAuCl ·6H O (Aldrich) and (NH ) CO (Aldrich) for Au/ceria,
term stability was evaluated for the first time. Cu/ceria-octahedra
showed no degradation over 100 h of PROX reaction while the
4
2
4
2
3
and (NH ) PtCl (Aldrich) and NaBH4 (Aldrich) for Pt/ceria.
4
2
6

8
4
J. Han et al. / Journal of Molecular Catalysis A: Chemical 335 (2011) 82–88
6
6
6
6
6
5
5
b 1.0x10
a
3.0x10
CeO2 (rod)
CeO2 (cube)
CeO2 (oct)
CeO2 (rod)
CeO2 (cube)
CeO2 (oct)
4
4
4
4
2
2
1
1
5
.5x10
.0x10
.5x10
.0x10
.0x10
8
6
4
2
.0x10
.0x10
.0x10
.0x10
0
.0
0
.0
1
00
200
300
400
500
600
100
200
300
400
500
600
Temp(ºC)
Temp(ºC)
6
6
6
6
6
5
5
c 3.0x10
d 1.0x10
Cu 4wt%/CeO2(rod)
Cu 4wt%/CeO2(cube)
Cu 4wt%/CeO2(oct)
288.2
Cu 4wt%/CeO2(rod)
Cu 4wt%/CeO2(cube)
Cu 4wt%/CeO2(oct)
2
34.5
2
1
12.2
86.6
4
4
4
4
2
2
1
1
5
.5x10
.0x10
.5x10
.0x10
.0x10
8.0x10
6
4
2
.0x10
.0x10
.0x10
180.0
213.0
130.0
0.0
0
.0
1
00
200
300
Temp(ºC)
400
500
600
100
200
300
Temp(ºC)
400
500
600
Fig. 3. H2-TPR profiles of (a) CeO2 and (c) 4 wt% Cu/CeO2, and CO-TPR profiles of (b) CeO2 and (d) 4 wt% Cu/CeO2.
2.2. Measurements
follows:
CO conversion (%) = [
CO]in − [CO]out
× 100
Transmission electron microscopy (TEM) and high resolution
HR) TEM images were performed with a Tecnai 20 operated
[
CO]
in
(
at 200 kV. X-ray photoelectron spectroscopy (XPS) data were
acquired with a Thermo MultiLab 2000 system. All spectra were
collected using Al K␣ radiation (1486.6 eV) from a twin anode.
High-resolution spectra averaged over 20 scans with a dwell time
of 50 ms in steps of 0.1 eV were obtained. The binding energy
was referenced with C (1s) at 284.98 eV. H2 temperature pro-
grammed reduction (TPR) and CO-TPR were carried out by a
BELCAT-M instrument (BEL Japan, Inc.) using a thermal conductiv-
Selectivity (%) = 0
.5([CO]in
− [CO]out)
[
O ] − [O ]
2
in
2
out
3. Results and discussion
Fig. 1 shows rod, cube, and octahedral shapes of the ceria
nanocrystals. The shapes of each nanocrystal were seen more
clearly in the high resolution TEM images of the insets of Fig. 1.
These shaped nanocrystals provide distinct {1 1 0} and {1 0 0} sur-
face structures for rods, a {1 0 0} surface for cubes, and a {1 1 1}
surface for octahedra. The mean particle size was estimated as
8.7 nm for rods (width), 39.5 nm for cubes, and 15.3 nm for octa-
ity detector. Forty milligrams of the catalysts were pre-oxidized
◦
in a 20% O /He gas mixture at 400 C for 30 min, and the TPR
2
measurements were performed with a 5% H /Ar gas mixture or
2
a 5% CO/He gas mixture with a flow rate of 30 ml/min from
◦
room temperature to 600 C. The temperature increased with
◦
◦
a ramping rate of 10 C/min. Raman spectra were acquired by
hedra from peaks at 28.5 in X-ray diffraction patterns by using
excitation of the sample at 785 nm using a Raman microscope
system (Kaiser, Raman MicroProbe). The spectral resolution cor-
responded to 4 cm . The elemental analysis was performed with
inductively coupled plasma (ICP; IRIS Intrepid II, Thermo Elemen-
tal).
the Scherer equations. The lengths of the rods were estimated to
be 20–200 nm from the TEM images. When 4 wt% of copper was
deposited on the various shapes of ceria support, copper nanopar-
ticles were not observed in the TEM images or XRD patterns.
Inductively coupled plasma (ICP) elemental analysis showed that
the actual amount of deposited copper was 3.9 wt% for rods, 3.9 wt%
for cubes, and 4.2 wt% for octahedra. These results suggest that the
deposited copper was finely dispersed on the ceria.
−1
The PROX reaction was carried out in a quartz glass fixed-bed
reactor at atmospheric pressure. A total of 100 mg of catalyst and
4
00 mg of sea sand were mixed and deposited inside the reac-
tor. The inlet gas flow had a composition of 50% H , 1% O , 1%
CO and He in balance. The total flow rate was 60 ml/min. Before
The XPS results in Fig. 2 also show that most of the copper
occurred as a reduced copper species, not as CuO, due to the absence
of shake-up peaks that were expected for Cu2+ at 939–944 eV [22].
Highly dispersed CuOx species in proximity to Ce (Ce–O–Cu sites)
2
2
the PROX tests, the catalysts were pre-oxidized in a 10% O /He
2
◦
gas mixture at 400 C for 1 h. The compositions of the effluent
gases were measured with an on-line gas chromatograph (GC;
Younglin GC 6000 series) equipped with a thermal conductiv-
ity detector and a flame ionization detector with methanizer.
A 5A molecular sieve and Porapak N were used as a GC col-
umn. The CO conversion and selectivity values were calculated as
are usually considered active sites for CO oxidation, while H oxida-
2
tion is promoted by partially reduced copper oxide nanoparticles
(Cu–O–Cu sites) [23–25]. When the Ce 3d and Cu 2p peak posi-
tions in the XPS results were compared, the binding energy of the
Cu/ceria-cubes had the lowest value, whereas the Cu/ceria-rods

J. Han et al. / Journal of Molecular Catalysis A: Chemical 335 (2011) 82–88
85
a 100
b 100
5
4
3
2
CeO2(rod)
CeO2(cube)
CeO2(oct)
8
6
4
2
0
0
0
0
0
80
60
40
20
0
1
1
50 160 170 180 190 200
Temp (ºC)
CeO2(rod)
CeO2(cube)
CeO2(oct)
30
60
90 120 150 180 210
Temp (ºC)
30
60
90 120 150 180 210
Temp (ºC)
c 100
d 100
80
80
60
40
20
0
60
40
20
Cu 4wt%/CeO2(rod)
Cu 4wt%/CeO2(cube)
Cu 4wt%/CeO2(oct)
Cu 4wt%/CeO2(rod)
Cu 4wt%/CeO2(cube)
Cu 4wt%/CeO2(oct)
0
30
60 90 120 150 180 210
Temp (ºC)
30 60 90 120 150 180 210
Temp (ºC)
Fig. 4. CO conversion (a and c) and selectivity (b and d) of CeO2 and 4 wt% Cu/CeO2 in the PROX reaction (space velocity = 36,000 ml/g catalyst h).
or Cu/ceria-octahedra showed higher binding energies. This result
implies that ceria with rod or octahedral shapes have stronger
interactions with copper than ceria with a cubic shape. When the
Cu/(Cu + Ce) ratio was calculated from the surface atomic contents
measured by XPS, the Cu/ceria-rods had the lowest value of 0.21,
compared to 0.24 of the Cu/ceria-cubes and 0.35 of the Cu/ceria-
octahedra. The surface content of copper was much lower on the
rods than the octahedra. The change in surface copper content was
also reported for Cu/ceria reduced in different temperatures, which
was attributed to the migration of copper into the bulk ceria [25].
Based on the similar amount of copper measured by ICP elemental
analysis, the copper seemed to migrate inside the ceria nanocrystal
more for the rod than the other shapes. By controlling the shape
of the ceria support, the mode of copper deposition into the ceria
crystals can be changed, and these changes will affect the activity
and selectivity of the PROX reaction.
The H -TPR for bare ceria in Fig. 3(a) shows that the surface
2
◦
oxygen was reduced above 300 C, but the reducing tempera-
ture greatly decreased after the copper deposition, as shown in
Fig. 3(c). When 4 wt% of copper was deposited, the Cu/ceria were
◦
reduced by hydrogen at a temperature range of 185–210 C. The
peak area of the Cu/ceria-cubes was the smallest. The Cu/ceria-rods
and Cu/ceria-octahedra had similar peak areas. The Cu/ceria-rods
showed a relatively broad peak, implying that different kinds of
a
b 18000
4
000
000
0
1
2000
2
6
000
0
Cu 4wt%/CeO2(cube)
CeO2(cube)
Cu 4wt%/CeO2(rod)
CeO2(rod)
200
400
600
800
1000
200
400
600
800
1000
-
1
-1
Raman shift(cm )
Raman shift(cm )
0
.10
d
c
8
000
000
0
0.08
4
wt%
4wt%
Cu/CeO2
oct
0
0
0
0
.06
.04
.02
.00
Cu/CeO2
rod
4wt%
Cu/CeO2
cube
4
CeO2
cube
CeO2
oct
CeO2
rod
Cu 4wt%/CeO2(oct)
CeO2(oct)
200
400
600
800
1000
Catalysts
-
1
Raman shift(cm )
Fig. 5. Raman profiles for a (a) rod shape, (b) cubic shape, (c) octahedral shape, and (d) A595/A460 ratio of CeO2 and 4 wt% Cu/CeO2.

8
6
J. Han et al. / Journal of Molecular Catalysis A: Chemical 335 (2011) 82–88
Fig. 6. TEM images of (a) 4 wt% Cu/ceria-rods, (b) 4 wt% Cu/ceria-cubes, and (c) 4 wt% Cu/ceria-octahedra before PROX reaction and (d) 4 wt% Cu/ceria-rods, (e) 4 wt%
Cu/ceria-cubes, and (f) 4 wt% Cu/ceria-octahedra after PROX reaction.
reducing sites co-existed. Because a rod consisted of {1 1 0} and
Because the Cu/ceria-rods had a smaller amount of copper on the
surface although total amount of copper was similar, the Cu/ceria-
rods would have more doped copper inside the ceria crystal. These
doped coppers might move back to the surface, enhancing CO con-
sumption at high temperatures. However, the CO consumption
{
1 0 0} surfaces while the other shaped ceria consisted of either
1 1 1} or {1 0 0}, the interaction of copper on the rod surface would
{
be more diverse. Fig. 3(b) and (d) shows the CO-TPR results of the
bare ceria and 4 wt% Cu/ceria, respectively. While the bare ceria
showed very broad reduction peaks over a wide range of tem-
peratures with small intensities, the CO consumption was greatly
increased after copper deposition. It was recently reported that
CO is adsorbed strongly on Cu/CeO2 (1 1 1) surfaces while CO
is only weakly physisorbed on bare CeO2 (1 1 1) surfaces, based
on a theoretical simulation using density functional theory [26].
The amount of CO consumption varied for the different shapes of
ceria. The Cu/ceria-rods and Cu/ceria-octahedra showed 4.9 and
a
100
80
60
40
1
.8 times more CO consumption, respectively, than the Cu/ceria-
cubes. Although the Cu/ceria-rods and Cu/ceria-octahedra showed
similar peak areas for H -TPR, the Cu/ceria-rods had much larger
2
peak area than the Cu/ceria-octahedra, especially in high tempera-
◦
20
0
tures above 250 C. It has been suggested that copper deposited on
Cu 1wt%/CeO2(oct)
Cu 4wt%/CeO2(oct)
Cu 10wt%/CeO2(oct)
2+
ceria surfaces exists as CuO, doped Cu ions incorporated in the
+
2+
ceria lattice, or surface Cu ions, and that doped Cu ions facilitate
oxygen vacancy formation, thus enhancing CO oxidation [26,27].
30
60
90
120
150
180
210
Temp (ºC)
1
00
b
100
80
60
40
20
0
80
60
40
20
0
Cu 4wt%/ceria(oct)
Cu 4wt%/ceria(rod)
Cu 4wt%/ceria(cube)
Cu 1wt%/CeO2(oct)
Cu 4wt%/CeO2(oct)
Cu 10wt%/CeO2(oct)
0
20
40
60
80
100
30
60
90
120
150
180
210
Time (h)
Temp (ºC)
◦
Fig. 7. Long-term stability test for PROX reaction over 4 wt% Cu/CeO2 at 140 C.
Fig. 8. Effect of copper contents on PROX reaction: (a) activity and (b) selectivity.

J. Han et al. / Journal of Molecular Catalysis A: Chemical 335 (2011) 82–88
87
a 100
b 100
80
80
60
40
20
0
60
40
Au 1wt%/CeO2(rod)
Au 1wt%/CeO2(cube)
Au 1wt%/CeO2(oct)
20
0
Au 1wt%/CeO2(rod)
Au 1wt%/CeO2(cube)
Au 1wt%/CeO2(oct)
30
60 90 120 150 180 210
Temp (ºC)
30 60 90 120 150 180 210
Temp (ºC)
d 100
c 100
Pt 1wt%/CeO2(rod)
Pt 1wt%/CeO2(cube)
Pt 1wt%/CeO2(oct)
Pt 1wt%/CeO2(rod)
Pt 1wt%/CeO2(cube)
Pt 1wt%/CeO2(oct)
80
60
40
20
0
80
60
40
20
0
30
60 90 120 150 180 210
Temp (ºC)
30 60 90 120 150 180 210
Temp (ºC)
Fig. 9. (a) Activity and (b) selectivity for 1 wt% Au/CeO2 and (c) activity and (d) selectivity for 1 wt% Pt/CeO2 in PROX reaction.
◦
in low temperatures below 250 C was higher for the Cu/ceria-
octahedra than the Cu/ceria-rods, probably due to more copper on
the surface.
and the cubes had the least oxygen vacancy. However, when the
copper was deposited, the Cu/ceria-rods and Cu/ceria-octahedra
had comparable oxygen vacancies, although the bare octahedra
had an oxygen vacancy as low as the bare cubes. The Cu/ceria-
octahedra had an oxygen vacancy comparable to the Cu/ceria-rods,
and more copper adsorbed on the surface, resulting in the highest
Fig. 4 shows the activity and selectivity for the PROX reac-
tion. The PROX reaction was performed in a gas flow containing
0% H , 1% O , 1% CO, and He in balance with a space veloc-
ity of 36,000 ml/g catalyst h. For the bare ceria, the activity was
as low as <5%. The rods showed the highest activity among the
shapes, but the difference was minimal. The selectivity, which is a
ratio of the oxygen used for CO conversion to the total consumed
oxygen, also varied, with the highest selectivity for octahedra
and the lowest selectivity for cubes. For the 4 wt% Cu/ceria cat-
alyst, the Cu/ceria-octahedra showed the greatest CO conversion
5
2 2
◦
CO conversion at 140 C.
The shape change of the 4 wt% Cu/ceria was checked after the
PROX measurements. Fig. 6(a)–(c) shows TEM images of Cu/ceria
before the PROX test and Fig. 6(d)–(f) shows the TEM images after
the PROX test. No shape change was observed, nor were distinct
copper oxide nanoparticles observed. The removal of surface oxy-
gen by the oxidation reaction and the re-filling of the oxygen
vacancy did not change the shape of the ceria nanoparticles signifi-
cantly. The long-term stability of the PROX reaction was tested over
◦
of 95% at 140 C. The conversion was 87% for the Cu/ceria-rods
and 67% for the Cu/ceria-cubes at the same temperature. The
reaction rate at 140 C was calculated as 4.7 ␮mol/g catalyst s for
◦
◦
100 h at 140 C. As shown in Fig. 7, the Cu/ceria-octahedra showed
the Cu/ceria-octahedra, 4.3 ␮mol/g catalyst s for the Cu/ceria-rods,
and 3.3 ␮mol/g catalyst s for the Cu/ceria-cubes. However, as the
temperature increased above 140 C, the Cu/ceria-rods showed
the highest CO conversion while the conversion for the Cu/ceria-
octahedra rapidly decreased.
no degradation over 100 h while the CO conversion decreased
from 84% to 73% for the Cu/ceria-rods, from 67% to 46% for the
Cu/ceria-cubes, respectively. The stability of the low index plane
was estimated to be {1 1 1} > {1 1 0} > {1 0 0} from the theoretical
simulation [6]. As expected, the Cu/ceria-octahedra with the {1 1 1}
surface showed the best long-term stability. The shape of the cata-
lyst affects not only the activity and selectivity, but also the stability
of a long-term reaction.
◦
A previous report showed that when gold was deposited on rod,
cubic, or polyhedral shapes of ceria, the Au/ceria-rods had the best
◦
activity for the water–gas shift reaction at 150–350 C [10]. The
difference in activity for the various shapes was explained by oxy-
gen vacancy formation. Because the formation energy for oxygen
vacancy has an order of {1 1 0} < {1 0 0} < {1 1 1} for ceria [8], the
rods with a {1 1 0} surface would be the most reactive. However, the
ceria {1 0 0} surface is polar, with a positively charged Ce layer and
negatively charged O layers, and is unstable upon heating, relaxing
to a low-energy surface. Therefore, the {1 0 0} surface may undergo
surface reconstruction with poorer activity, although the {1 0 0}
surface itself would have a strong adsorption with CO. The forma-
tion of oxygen vacancy was estimated by Raman spectroscopy as
shown in Fig. 5. The relative concentration of oxygen vacancies was
The effect of the amount of deposited copper was evaluated
as shown in Fig. 8. The PROX activity and selectivity were com-
pared for the 1, 4, and 10 wt% Cu/ceria-octahedra. As the copper
was deposited on the ceria, the oxygen vacancy and CO adsorp-
tion increased, resulting in higher PROX activity. The 4 wt% Cu/ceria
showed the best activity. The 1 wt% Cu/ceria seemed to have an
insufficient number of active sites for CO oxidation. However, the
10 wt% Cu/ceria showed poorer activity than the 4 wt% Cu/ceria.
According to the XRD of the 10 wt% Cu/ceria, separate CuO peaks
were observed. Too much CuO on the surface promotes H2 oxi-
dation, as evidenced by the lower selectivity at low temperatures
[31].
−
1
represented by the area ratio of peaks at 595 and 460 cm of the
Raman spectroscopy [28,29]. As demonstrated in Fig. 5(d), the oxy-
gen vacancy increased greatly upon copper deposition [30]. When
bare ceria were compared, the rods had the most oxygen vacancy,
Gold and platinum were also deposited on the various shapes of
ceria, and their PROX activities were measured as shown in Fig. 9.
A 1 wt% of metals was deposited. When a 4 wt% of metals was

8
8
J. Han et al. / Journal of Molecular Catalysis A: Chemical 335 (2011) 82–88
deposited, the activity was similar to or lower than the 1 wt% cases.
For both Au and Pt cases, the rods showed the highest activity at low
temperatures while the activity was similar for rods and octahedra
at high temperatures. These noble metals showed higher activity
than Cu/ceria at the lower temperature region, but their maximum
References
[
[
1] A.R. Tao, S. Habas, P. Yang, Small 4 (2008) 310.
2] Y. Xia, Y.J. Xiong, B. Lim, S.E. Skrabalak, Angew. Chem., Int. Ed. 48 (2009)
60.
[3] K.M. Bratlie, H. Lee, K. Komvopoulos, P. Yang, G.A. Somorjai, Nano Lett. 7 (2007)
3097.
◦
activity was lower than Cu/ceria, being 84% at 40 C for Au/ceria
[
4] S.E. Habas, H. Lee, V. Radmilovic, G.A. Somorjai, P. Yang, Nat. Mater. 6 (2007)
92.
[5] C. Kim, H. Lee, Catal. Commun. 11 (2009) 7.
◦
◦
and 74% at 80 C for Pt/ceria compared to 95% at 140 C for Cu/ceria.
The surface defects of ceria play a role as anchoring sites of met-
als [32], and the interactions of the ceria surface defects and the
anchored metals would vary for different kinds of metal. Conse-
quently, the PROX properties were greatly affected by the kind of
deposited metal.
6
[
[
[
6] E. Aneggi, J. Llorca, M. Boaro, A. Trovarelli, J. Catal. 234 (2005) 88.
7] K.B. Zhou, X. Wang, X.M. Sun, Q. Peng, Y.D. Li, J. Catal. 229 (2005) 206.
8] M. Nolan, J.E. Fearon, G.W. Watson, Solid State Ionics 177 (2006) 3069.
[9] M. Nolan, G.W. Watson, J. Phys. Chem. B 110 (2006) 16600.
[10] R. Si, M. Flytzani-Stephanopoulos, Angew. Chem., Int. Ed. 47 (2008) 2884.
[
11] P. Ratnasamy, D. Srinivas, C.V.V. Satyanarayana, P. Manikandan, R.S.S. Kumaran,
M. Sachin, V.N. Shetti, J. Catal. 221 (2004) 455.
4
. Conclusion
[12] W.B. Kim, T. Voitl, G.J. Rodriguez-Rivera, S.T. Evans, J.A. Dumesic, Angew. Chem.,
Int. Ed. 44 (2005) 778.
[
13] C. Rossignol, S. Arrii, F. Morfin, L. Piccolo, V. Caps, J.L. Rousset, J. Catal. 230 (2005)
76.
14] S. Alayoglu, A.U. Nilekar, M. Mavrikakis, B. Eichhorn, Nat. Mater. 7 (2008)
333.
15] T. Caputo, L. Lisi, R. Pirone, G. Russo, Ind. Eng. Chem. Res. 46 (2007) 6793.
16] D. Gamarra, A. Hornes, Z. Koppany, Z. Schay, G. Munuera, J. Soria, A. Martinez-
Arias, J. Power Sources 169 (2007) 110.
Copper was deposited on different shapes of ceria nanocrystals
rods, cubes, and octahedra). The activity, selectivity and long-term
4
(
[
stability of the Cu/ceria catalysts were evaluated for the PROX
reaction. When the copper was deposited on the ceria support,
the surface content of the copper significantly differed depending
on the ceria shapes, with more copper on the octahedra. Cop-
per seemed to incorporate inside the ceria lattice more in ceria
rods. The CO adsorption increased greatly upon copper deposition.
The Cu/ceria-rods showed greater CO adsorption in high tem-
[
[
[17] A. Gomez-Cortes, Y. Marquez, J. Arenas-Alatorre, G. Diaz, Catal. Today 133
2008) 743.
18] C.R. Jung, J. Han, S.W. Nam, T.H. Lim, S.A. Hong, H.I. Lee, Catal. Today 93–95
2004) 183.
(
[
(
[19] F. Marino, C. Descorme, D. Duprez, Appl. Catal. B: Environ. 58 (2005)
175.
[
◦
peratures above 250 C, whereas the Cu/ceria-octahedra showed
20] A. Martinez-Arias, A.B. Hungria, M. Fernandez-Garcia, J.C. Conesa, G. Munuera,
J. Power Sources 151 (2005) 32.
slightly more CO adsorption at low temperatures. Copper incorpo-
rated inside the ceria lattice would move back to the ceria surface
more at high temperatures, resulting in more CO oxidation. The
oxygen vacancy increased significantly after copper deposition. The
Cu/ceria catalysts showed no apparent shape degradation after
the PROX reaction. The high activity of the Cu/ceria-octahedra,
especially, remained unchanged over 100 h while the other shapes
showed degradation. Au or Pt was also deposited on the different
ceria shapes, and their PROX activity and selectivity were evaluated.
[21] H.X. Mai, L.D. Sun, Y.W. Zhang, R. Si, W. Feng, H.P. Zhang, H.C. Liu, C.H. Yan, J.
Phys. Chem. B 109 (2005) 24380.
[
[
22] G. Avgouropoulos, T. Ioannides, Appl. Catal. B: Environ. 67 (2006) 1.
23] D. Gamarra, C. Belver, M. Fernandez-Garcia, A. Martinez-Arias, J. Am. Chem.
Soc. 129 (2007) 12064.
[24] C.S. Polster, H. Nair, C.D. Baertsch, J. Catal. 266 (2009) 308.
[
[
[
25] X.L. Tang, B.C. Zhang, Y. Li, Y.D. Xu, Q. Xin, W.J. Shen, Appl. Catal. A: Gen. 288
2005) 116.
(
26] Z.X. Yang, B.L. He, Z.S. Lu, K. Hermansson, J. Phys. Chem. C 114 (2010)
4486.
27] D. Gamarra, G. Munuera, A.B. Hungria, M. Fernandez-Garcia, J.C. Conesa, P.A.
Midgley, X.Q. Wang, J.C. Hanson, J.A. Rodriguez, A. Martinez-Arias, J. Phys.
Chem. C 111 (2007) 11026.
Acknowledgements
[
28] J. Li, P.F. Zhu, S.F. Zuo, Q.Q. Huang, R.X. Zhou, Appl. Catal. A: Gen. 381 (2010)
2
61.
This work was supported by the DAPA/ADD of Korea, the
National Research Foundation of Korea Grant Funded by the
Korean Government (MEST) (NRF-2009-C1AAA001-0092926) and
the Korean Ministry of Environment as Converging technology
project (202-091-001).
[
[
29] Z.Y. Pu, J.Q. Lu, M.F. Luo, Y.L. Me, J. Phys. Chem. C 111 (2007) 18695.
30] L. Dong, Y. Hu, M. Shen, T. Jin, J. Wang, W. Ding, Y. Chen, Chem. Mater. 13 (2001)
4227.
31] T. Caputo, L. Lisi, R. Pirone, G. Russo, Appl. Catal. A: Gen. 348 (2008) 42.
32] Q. Fu, H. Saltsburg, M. Flytzani-Stephanopoulos, Science 301 (2003) 935.
[
[