influence of zirconia crystal phase on the catalytic performance of au/zro2 catalysts for...
TRANSCRIPT
Influence of zirconia crystal phase on the catalytic performance of
Au/ZrO2 catalysts for low-temperature water gas shift reaction
Juan Li, Junli Chen, Wei Song, Junlong Liu, Wenjie Shen *
State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China
Received 11 July 2007; received in revised form 10 October 2007; accepted 22 October 2007
Available online 26 October 2007
www.elsevier.com/locate/apcata
Available online at www.sciencedirect.com
Applied Catalysis A: General 334 (2008) 321–329
Abstract
The influence of crystal phase of zirconia on the performance of Au/ZrO2 catalysts for low temperature water gas shift reaction was
investigated. Au/ZrO2 catalysts with pure tetragonal and monoclinic phases of ZrO2 were prepared by the deposition-precipitation method with
similar gold loading and dispersion. It was found that the Au/m-ZrO2 catalyst showed much higher activity than that of the Au/t-ZrO2 catalyst,
which could be attributed to the higher CO adsorption capacity of the Au/m-ZrO2 catalyst. The chemical state of gold that was strongly related to
the pretreatment atmosphere also played an essential role in determining the catalytic activity for water gas shift reaction. Hydrogen and/or helium
pretreated samples only contained Au0 species and exhibited higher activity than that of the sample pretreated with oxygen-containing atmosphere
which resulted in the co-existence of Au0 and Au+ species. FTIR study further revealed that the formate species formed by the reaction of the
adsorbed CO on gold nanoparticle with the hydroxyl groups on the surface of m-ZrO2 acted as the most important intermediates for the water gas
shift reaction.
# 2007 Elsevier B.V. All rights reserved.
Keywords: Au/ZrO2; Crystal phase; Low-temperature; Water gas shift; FTIR
1. Introduction
Recently, the water gas shift (WGS) reaction is receiving
renewed interest for CO cleanup during hydrogen production
from alcohols and/or hydrocarbons for fueling polymer
electrolyte fuel cells, which requires removing the traceable
amounts of CO in H2-rich stream for preventing the Pt electrode
catalysts from deactivation of by CO adsorption [1–3]. Due to
the thermodynamic equilibrium limit, low CO levels can only
be achieved at low temperature although it is not favorable
kinetically. The commercial low temperature WGS catalyst
(Cu/ZnO/Al2O3) is not suitable for this specific application,
because it is very sensitive to air and requires very careful
activation prior to use. Therefore, substantial efforts have been
paid to the development of novel WGS catalysts with
sufficiently high activity at low temperatures.
Au/ZrO2 catalysts have been reported to show exceptionally
high activities for WGS reaction [4,5]. In addition to the typical
* Corresponding author. Tel.: +86 411 84379085; fax: +86 411 84694447.
E-mail address: [email protected] (W. Shen).
0926-860X/$ – see front matter # 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2007.10.020
size-dependent effect as well as the chemical states of gold
nanoparticles [5,6], the nature of zirconia was also found to play
an essential role in determining the catalytic performance [7,8].
Au nanoparticles on well crystallized zirconia exhibited higher
WGS activity than that on amorphous one [4], and the high
activity was usually attributed to the perimeter interface between
Au nanoparticle and zirconia [5]. However, the influence of
zirconia crystal phase on the activity of Au/ZrO2 catalyst is rarely
concerned. Zirconia could present as tetragonal (t-ZrO2) and
monoclinic (m-ZrO2) phases, which have different coordination
environments of zirconium and oxygen [9]. Zr4+ cation is
octacoordinated and O2� anion is tetracoordinated in the
tetragonal phase, while Zr4+ cation is heptacoordinated and
O2� anion is either tri- or tetracoordinated in the monoclinic
phase. Because of this, the acidic/basic properties as well as the
concentrations of surface hydroxyl group might be quite different
between the two crystal phases [10–12], which would in turn
affect the interaction between gold nanoparticle and ZrO2. In
fact, the CO adsorption capacity of m-ZrO2 was found to be 5 to
10-fold higher than that of t-ZrO2 [11]. The Cu/m-ZrO2 catalyst
also showed 8-fold higher activity for methanol synthesis by CO
hydrogenation than that of the Cu/t-ZrO2 catalyst [13,14].
J. Li et al. / Applied Catalysis A: General 334 (2008) 321–329322
In this work, we investigated the effect of zirconia crystal
phase on the activity of Au/ZrO2 catalysts for low temperature
WGS reaction. Au/t-ZrO2 and Au/m-ZrO2 catalysts were
prepared with similar Au loading and dispersion. The structural
features of the catalysts were characterized with X-ray power
diffraction, transmission electron spectroscopy, temperature-
programmed reduction/desorption, and X-ray photoelectron
spectroscopy. The nature of the adsorption sites and the reaction
mechanism was studied by in situ FTIR.
2. Experimental
2.1. Preparation of ZrO2 supports and Au/ZrO2 catalysts
The tetragonal zirconia (t-ZrO2) was prepared by a
homogeneous precipitation process. 32 g ZrOCl2�8H2O and
60 g urea were dissolved in 500 mL deionized H2O, and the
solution was heated to 100 8C and kept at this temperature for
5 h. The precipitation was filtered, washed with hot water, and
then dried at 110 8C overnight. The solid obtained was calcined
at 450 8C for 4 h in air.
The monoclinic zirconia (m-ZrO2) was prepared using a
reflux method similar to that described by Jung and Bell [15].
80 g ZrOCl2�8H2O was dissolved in 500 mL deionized H2O,
and the solution was boiled under refluxed at 100 8C for 240 h.
After the suspension was cooled to room temperature, 25 wt.%
ammonia aqueous solution was slowly added to agglomerate
the fine particles to facilitate the filtration. The precipitate was
then dried at 110 8C overnight and calcined at 450 8C for 4 h in
air. For comparison, the dried solid was also calcined at 500 8Cfor 4 h in air and the obtained sample was nominated as
m-ZrO2-R.
The Au/ZrO2 catalysts were prepared by a deposition–
precipitation method. 3 g ZrO2 was suspended in 250 mL H2O
and heated to 60 8C. The pH value of the mixture was adjusted
to 8 by adding 0.25 M Na2CO3 solution. 50 mL aqueous
solution containing 0.2 g HAuCl4�4H2O was also adjusted to
pH = 8 with 0.25 M Na2CO3 solution, which was then added
dropwise to the slurry containing ZrO2. The resulting mixture
was further aged at 60 8C for 2 h under stirring. The solid was
filtrated and washed thoroughly with water, and then dried at
room temperature for 24 h.
2.2. Catalyst characterization
N2 adsorption–desorption isotherm was recorded at
�196 8C on a Micrometrics ASAP 2000 instrument. Prior to
measurement, the sample was degassed at 250 8C for 5 h. The
specific surface area was calculated from a multipoint
Braunauer–Emmett–Teller (BET) analysis of the nitrogen
adsorption isotherms.
The actual loading of gold was determined by inductively
coupled plasma atomic emission spectroscopy (ICP-AES) on a
Plasma-Spec-I spectrometer.
X-ray powder diffraction (XRD) pattern was recorded using
a D/Max-2500/PC powder diffractometer (Rigaku, Japan) with
Cu Ka radiation operated at 40 kV and 100 mA. The average
crystal size of ZrO2 was calculated by the Scherer equation
using the ð1 1 1Þ diffraction (2u = 28.2) for m-ZrO2 and the
(0 1 1) diffraction (2u = 30.38) for t-ZrO2, respectively.
High-resolution transmission electron spectroscopy
(HRTEM) image was obtained with a Philips Tecnai G220
instrument. The sample powders were ultrasonically dispersed
in anhydrous ethanol and drops of the suspensions were
deposited on a holey carbon film supported copper grid.
Temperature-programmed reduction (TPR) measurement
was conducted with a conventional setup equipped with a TCD
detector. 100 mg as-prepared samples were loaded and
pretreated with He (30 mL/min) at 250 8C for 30 min to
remove the adsorbed carbonates and hydrates. After cooling
down to room temperature and introducing the reduction agent
(5% H2/N2, 30 mL/min), the sample was then heated to 800 8Cat a ramp of 10 8C/min.
Temperature-programmed desorption (TPD) of CO was
conducted with a micro-reactor equipped with a quadrupole
mass spectrometer (Omnistar, Balzers). 50 mg as-prepared
samples were initially heated to 250 8C under He flowing
(40 mL/min) and held at this temperature for 30 min to remove
the adsorbed carbonates and hydrates. After cooling down to
room temperature, adsorption of CO was then performed by
passing 1% CO/He (40 mL/min) through the sample for 1 h.
After purging with He, the sample was heated to 250 8C at a
ramp of 10 8C/min under He flowing (40 mL/min), and the
effluents from the reactor were measured by MS.
X-ray photoelectron spectroscopy (XPS) measurement was
performed with an ESCALAB MK-II spectrometer (VG
Scientific Ltd., UK) using Al Ka radiation (1486.6 eV)
operated at an acceleration voltage of 10 kV. The powder
samples were pressed into thin discs and mounted on a sample
rod placed in an analysis chamber, where the XP spectra of C
1s, O 1s, Zr 3d and Au 4f levels were recorded. Charge effects
were corrected by adjusting the binding energy of C 1s to
284.6 eV. The surface atomic ratio of Au/Zr was calculated
according to the proposed procedure [16,17].
2.3. Activity test
The WGS reaction was conducted with a continuous flow
fixed-bed quartz reactor under atmospheric pressure. 200 mg
samples (40–60 mesh) were placed between two layers of
quartz wool inside a quartz tube (i.d. = 6 mm). Prior to test, the
catalyst was heated to 250 8C under He flow (50 mL/min) for
2 h. The water was supplied by flowing He over a glass bubbler
containing distilled water kept at 51 8C. The H2O/He stream
was then mixed with the CO/He stream coming from a mass
flow controller. The typical feed gas composition was 3% CO/
9% H2O balanced with He. Effluents from the reactor were
analyzed by an on-line gas chromatograph equipped with TCD
and FID detectors.
2.4. FTIR
FTIR spectra were collected over a Biorad Vector 22
spectrometer equipped with a DTGS detector, using 32 scans
Fig. 1. XRD patterns of the zirconium oxides and the Au/ZrO2 catalysts.
J. Li et al. / Applied Catalysis A: General 334 (2008) 321–329 323
per spectrum in the region of 4000–1000 cm�1 with a
resolution of 4 cm�1. Au/ZrO2 sample was pressed into a thin
self-supporting pellet and introduced into a cell allowing
thermal treatments under controlled atmospheres. The pellet
was treated in a flow of He (40 mL/min) at 250 8C for 30 min to
remove the adsorbed H2O and carbonates. After cooling down
to given temperature under He flow, IR spectra were recorded
by exposing the sample to 1% CO/He or 1% CO/3% H2O/He
stream (40 mL/min). The reference of spectrum of the Au/ZrO2
wafer in He taken at the measurement temperature was
subtracted from each spectrum.
3. Results and discussion
3.1. Structural properties of the catalysts
Table 1 summarizes the structural and textural properties of
the ZrO2 supports and the Au/ZrO2 catalysts. The surface areas
of t-ZrO2 and m-ZrO2 were 96 and 136 m2/g, respectively.
Fig. 1 shows the XRD patterns of the zirconium oxides and the
Au/ZrO2 catalysts. The t-ZrO2 was indexed to pure tetragonal
phase with characteristic diffraction peak at 2u = 30.38 (JCPDS
50-1089) and crystalline size of 13 nm. The m-ZrO2 was well
described as monoclinic phase with characteristic diffraction
peaks at 2u= 28.28 and 31.38 (JCPDS 37-1484) and crystalline
size of 5 nm. The deposition of gold only slightly decreased the
surface areas of the supports, which were 94 m2/g for Au/t-
ZrO2 and 122 m2/g for Au/m-ZrO2. The actual loadings of gold
were 1.26 wt.% for Au/t-ZrO2 and 1.20 wt.% for Au/m-ZrO2,
respectively. Except a extra small peak at 2u = 28.28 attributed
to monoclinic phase appeared in the Au/t-ZrO2 after the
deposition of gold, the XRD patterns of the Au/ZrO2 catalysts
were similar to those of the corresponding ZrO2 supports, and
no distinct diffraction peaks of metallic gold could be observed,
indicating that the gold particles were too small to be detected
and/or were highly dispersed on ZrO2.
Fig. 2 shows the HRTEM images of the Au/ZrO2 catalysts. It
was observed that t-ZrO2 had a flocculent appearance and m-
ZrO2 had a discrete distribution with a large number of
wormhole-like channels. The particle sizes were about
15–20 nm for t-ZrO2 and 5–15 nm for m-ZrO2, respectively.
Most of the gold particles were much smaller and hardly
observed. Only few gold particles with lattice fringes of
0.241 nm and particle size of about 3 nm were observed in the
Au/m-ZrO2 catalyst. Therefore, it can be assumed that there
should present almost same size distribution of gold
nanoparticles for the Au/ZrO2 catalysts due to the same
preparation route and gold loading. As a matter of fact, Zhang
Table 1
Structural and textural properties of the samples
Sample SBET (m2/g) XRD size (nm) TEM siz
t-ZrO2 96 13
m-ZrO2 136 5
1.26% Au/t-ZrO2 94 13 15–20 nm
1.20% Au/m-ZrO2 122 5 5–15 nm
et al. [7] also pointed out that gold particle size was independent
on the surface area for zirconia between 20 and 162 m2/g.
3.2. Temperature-programmed reduction/desorption
Fig. 3 shows the TPR profiles of the ZrO2 supports and the
Au/ZrO2 catalysts. There was a broad reduction peak at 600–
800 8C for t-ZrO2, which was probably associated with the
surface shell reduction [5]. Au addition shifted the major
surface reduction to about 650 8C, with a shoulder at about
530 8C. For m-ZrO2, no reduction took place even up to 800 8C,
but the loading of Au resulted in significant hydrogen
consumption at 540 8, indicating the partial reduction of
surface Zr4+ to Zr3+. Obviously, Au addition enhanced the
removal of surface oxygen, leading to oxygen deficiencies on
the surface of ZrO2. Pt was found to facilitate the generation of
bridging OH groups on the surface of zirconia by removing
oxygen from the OH groups or dissociating hydrogen directly
during hydrogen reduction [22]. This seems true for the Au/
ZrO2 catalysts. The easier surface reduction of the Au/m-ZrO2
catalyst indicated that the active OH groups formed in the
defect sites had higher concentrations than that of the Au/t-
ZrO2 catalyst. For both cases, however, there were no reduction
peaks which could be assigned to the reduction of gold species,
implying that the gold was mainly in metallic state, as
previously proposed [18].
Fig. 4 shows the CO-TPD profiles of the Au/ZrO2 catalysts.
Only desorption of CO2 was detected, which could be attributed
to the decomposition of formate and/or carbonate species on the
surface of ZrO2 [11]. CO2 desorption occurred at almost the
e (nm) TPR peak (8C) BE of Au 4f7/2 (eV) Au0 (%)
685 – –
– – –
650 83.8 100
540 83.5 100
Fig. 2. HRTEM images of the Au/ZrO2 catalysts: (a–b) Au/m-ZrO2 and (c–d) Au/t-ZrO2.
J. Li et al. / Applied Catalysis A: General 334 (2008) 321–329324
same temperature (about 80 8) for both catalysts, indicating a
similar bonding of CO on the surface of ZrO2. But the relatively
higher intensity of CO2 desorption over the Au/m-ZrO2 catalyst
indicated a higher CO adsorption capacity and reactivity with
the surface than that of the Au/t-ZrO2 catalyst. This
phenomenon further proved that higher concentration of anion
vacancies was presented on the surface of m-ZrO2, mainly due
to the surface atomic configuration [9]. A second CO-TPD
measurement following the initial experiment further con-
firmed the higher capacity of the Au/m-ZrO2 catalyst for CO
adsorption and activation.
3.3. XPS surface analysis
Fig. 5 shows the XPS spectra of Au4f in the Au/ZrO2
catalysts. The binding energy of Zr 3d5/2 in both catalysts was
182.2 eV, which is characteristic of Zr4+. The binding energies
of Au 4f7/2 and Au 4f5/2 are 83.9 and 87.7 eV for Au0, 84.7 and
88.2 eV for Au+, 86.3 and 89.6 eV for Au3+, respectively
[19,31]. The samples pretreated with hydrogen and helium
showed binding energies of 83.8 and 87.5 eV for Au/t-ZrO2,
and 83.5 and 87.2 eV for Au/m-ZrO2, which were character-
istics of metallic Au. As compared to bulk Au, the slightly
lower binding energies of Au 4f in these catalysts indicated a
modification of the electronic structure of Au nanoparticles due
to its interaction with zirconia. The XPS results were well
agreement with the TPR measurements, showing that gold
species were in metallic state after pretreatment with He.
However, for the pre-oxidized samples, in addition to Au0, the
XPS spectra also showed bonding energies of 84.8 and 88.5 eV
for Au/t-ZrO2, and 84.7 and 88.3 eV for Au/m-ZrO2,
representing Au+ species. As a result, the pre-oxidized Au/t-
ZrO2 catalyst contained 37% Au+ and 63% Au0 and the pre-
oxidized Au/m-ZrO2 sample contained 59% Au+ and 41% Au0.
For all the samples pretreated with He, H2 and O2, the surface
Au/Zr ratios were 0.012–0.013 for Au/t-ZrO2 and 0.012–0.015
for Au/m-ZrO2, respectively. This suggested that the surface
concentrations of Au and Zr were similar and almost
independent on the nature of the support and the pretreatment
atmosphere.
3.4. Water-gas shift reaction
Fig. 6 shows the CO conversion as a function of time-on-
stream (TOS) at 120 8C over the Au/ZrO2 catalysts. Clearly, the
Au/m-ZrO2 catalyst showed higher CO conversion than that of
the Au/t-ZrO2 catalyst. The conversion of CO reached 52%
after 10 h TOS over the Au/m-ZrO2 catalyst, however, the
corresponding conversion of CO was only 41% over the Au/t-
ZrO2 catalyst. To exclude the influence of the surface area of
zirconia in the catalytic activity, the Au/m-ZrO2-R which had
similar surface area (91 m2/g) to the Au/t-ZrO2 catalyst was
Fig. 3. H2-TPR profiles of the ZrO2 supports and the Au/ZrO2 catalysts.
J. Li et al. / Applied Catalysis A: General 334 (2008) 321–329 325
tested for WGS reaction under the same reaction conditions. It
revealed that this reference catalyst could show almost the same
activity as that of the Au/m-ZrO2 catalyst, although the later
had higher surface area. This indicated that the higher catalytic
activity of Au/m-ZrO2 catalyst than that of the Au/t-ZrO2
catalyst could not be simply attributed to the surface area of
zirconia. Fig. 7 further shows the effect of space velocity on CO
Fig. 4. CO-TPD profiles of the Au/ZrO2 catalysts:
conversion at 200 8C over the Au/ZrO2 catalysts. The Au/m-
ZrO2 catalyst exhibited only a slight decrease in CO conversion
(from 96 to 90%), whereas the Au/t-ZrO2 catalyst showed a
remarkable decrease in CO conversion (from 92 to 78%) when
the space velocity was increased from 20; 000 mLg�1cat h�1 to
40; 000 mL g�1cat h�1.
For supported gold catalysts, the activity is usually
controlled by the contact structure of Au nanoparticles with
the support, that is, the size of Au particles and the nature of the
support [20,21]. Since the dispersion and the chemical state of
gold were almost the same for the Au/ZrO2 catalysts, the higher
activity of the Au/m-ZrO2 catalyst was mainly related to the
monoclinic crystal phase of zirconia. As demonstrated by the
CO-TPD measurement, the Au/m-ZrO2 catalyst had a higher
CO adsorption capacity than that of the Au/t-ZrO2 catalyst,
indicating that higher concentration and more nucleophilic
hydroxyl groups were present on the surface of m-ZrO2 [11].
This will contribute to the strong adsorption of CO as HCOO-Zr
groups, which would act as the important intermediates for the
WGS reaction. Chenu et al. [22] also noted that the higher
activity of the Pt/m-ZrO2 catalyst for WGS reaction than that of
the Pt/t-ZrO2 catalyst could be well correlated with the
intensities of the adsorbed formate formed by CO adsorption.
Significantly higher surface concentration of formate was
observed on the Pt/m-ZrO2 catalyst due to the strong interaction
of CO and the bridging OH groups. Therefore, the CO
adsorption capacity gave an important indication of the number
of active sites, and consequently, the WGS reaction activity.
Another possibility might be that the m-ZrO2 was composed of
very small particles/crystallites which could make much more
Au-ZrO2 contact boundary [7]. This kind of nanocomposite
structure in the Au/m-ZrO2 catalyst would lead to easier CO
migration from Au to ZrO2 surface, facilitating the WGS
reaction.
Fig. 8 shows the effect of the pretreatment atmosphere of the
Au/ZrO2 catalysts on CO conversion at 200 8C. H2 and He-
treated catalysts gave similar CO conversion, whereas O2-
(a) the first and (b) the second measurement.
Fig. 5. XPS spectra of Au 4f in the Au/ZrO2 catalysts pretreated with H2, He and O2 atmospheres at 250 8C.
J. Li et al. / Applied Catalysis A: General 334 (2008) 321–329326
treated catalysts led to lower CO conversion. This is in good
accordance with the XPS observation. The Au/ZrO2 catalysts
pretreated with H2 and He only contained Au0, however, the
catalysts pretreated with O2 contained Au0 and Au+ species
simultaneously. The positively charged gold sites have stronger
Au–CO bonding [23], and thus resulted in a decline in CO
conversion. This phenomenon further suggests that metallic
gold nanoparticles are much more active for WGS reaction than
cationic gold species. Kim and Thompson [24] treated the Au/
CeO2 catalyst with NaCN aqueous solution and found that the
reaction rates of WGS for the unleached catalyst were
significantly higher than those for the leached catalyst, which
only contained highly dispersed cationic gold species. The
perimeter interface between Au nanoparticles and the oxide
support was regarded as a unique reaction site, where CO was
adsorbed on Au nanoparticles and H2O was activated on the
surfaces of oxide [21].
Fig. 6. CO conversion as a function of time-on-stream over the Au/ZrO2
catalysts. Reaction conditions: 120 8C; 3% CO/9% H2O/He; 20; 000 mL g�1cat h�1.
3.5. Reaction mechanism
Fig. 9 shows the FTIR spectra of CO adsorption on the m-
ZrO2 support and the Au/m-ZrO2 catalyst, which were recorded
after exposure to 1% CO/He stream for 15 min at 25 8C. On the
Au/m-ZrO2 catalyst, the characteristic band representing CO
adsorption on metallic gold nanoparticles appeared at
2117 cm�1. This typical band was previously attributed CO
co-adsorbed with oxygen species on Au nanoparticles [25,26].
For the current Au/m-ZrO2 catalyst, the co-adsorbed oxygen
could originate only from the spillover of oxygen in the OH
group on m-ZrO2 to gold nanoparticles. More interestingly, this
band displayed a broadening towards lower frequencies,
suggesting that CO was adsorbed on very small gold clusters
which strongly interacted with ZrO2 [27,28]. This result further
evidenced the presence of highly dispersed gold clusters,
although they were not detected by XRD and HRTEM. The
weak band at 2177 cm�1 was characteristic for CO adsorption
on m-ZrO2 surface [29].
Fig. 7. Effect of space velocity on CO conversion over the Au/ZrO2 catalysts.
Reaction conditions: 200 8C; 3% CO/9% H2O/He; (&) 20; 000 mL g�1cat h�1;
(~) 40; 000 mL g�1cat h�1.
Fig. 8. Effect of pretreatment atmosphere on CO conversion over the Au/ZrO2
catalysts. Reaction conditions: 200 8C; 3% CO/9% H2O/He; 40; 000 mL g�1cat h�1
Fig. 9. FTIR spectra of CO adsorption on the m-ZrO2 support and the Au/m-
ZrO2 catalyst recorded after exposure to 1% CO/He stream for 15 min at 25 8C.
J. Li et al. / Applied Catalysis A: General 334 (2008) 321–329 327
The bands at 1630, 1427 and 1221 cm�1 were attributed to
bicarbonate species, and the weak bands at 1555 and 1328 cm�1
could be assigned to bidentate carbonate species [11,27]. The
band at 3613 cm�1 was likely corresponding to the m-OH
vibration in the hydrogen-carbonate (i.e. bicarbonate) species.
For m-ZrO2 alone, after CO adsorption, the formation of surface
bicarbonate and bidentate carbonate species with COO vibra-
tions in 1200–1700 cm�1 region were also observed, but with
much lower intensities compared to the Au/m–ZrO2 catalyst.
Fig. 10. FTIR spectra of the Au/m-ZrO2 catalyst recorded after 1% CO/He w
Fig. 10 shows the FTIR spectra in carbonyl and carbonate
regions by switching the gas stream from CO/He to He at 25 8Cfor the Au/m-ZrO2 catalyst. The intensity of the band at
2177 cm�1 representing CO adsorption on Zr4+ rapidly
decreased and almost disappeared after 3 min. But the band
at 2117 cm�1 corresponding to CO adsorption on Au
nanoparticles decreased slowly and practically vanished after
10 min. However, the bands related to carbonate species with
COO vibrations in 1200–1700 cm�1 region only decreased
slightly, indicating that these species were strongly and
irreversibly adsorbed. Coupled with the results in Fig. 9, it
can be assumed that CO was initially adsorbed on gold
nanoparticles, and then spilled over to the surface of m-ZrO2,
where it further reacted with the OH groups on ZrO2 surface to
form stable bicarbonate and/or carbonate species.
as switched to He at 25 8C: (a) carbonyl region and (b) carbonate region.
Fig. 12. FTIR spectra of the Au/m-ZrO2 catalyst recorded after 1% CO/3%
H2O/He was switched to He at 200 8C.
J. Li et al. / Applied Catalysis A: General 334 (2008) 321–329328
Fig. 11 shows the FTIR spectra of the Au/m-ZrO2 catalyst by
switching the stream of 1% CO/He to 1% CO/3% H2O/He. At
25 8C, the inlet of water caused a broad absorption band at
3600–2800 cm�1, representing the OH stretching modes of
water molecules interacting with the OH groups on the surface
of m-ZrO2 through hydrogen bonds. Simultaneously, the
intensities of the bands at 3737 and 3637 cm�1 attributed to the
surface OH groups decreased, as evidenced by the appearance
of negative peaks in this region. Moreover, a band at
1636 cm�1, typical for non-dissociated water molecule bending
mode [30], appeared. Meanwhile, the band at 2117 cm�1 for
CO adsorption on Au nanoparticles reduced its intensity. The
weak bands at 2360 and 2335 cm�1 due to the formation of CO2
were also observed. The bands at 1630, 1427 and 1221 cm�1
belonging to bicarbonate species disappeared, and the bands at
1555 and 1327 cm�1 for bidentate carbonate species increased.
This indicates that the addition of water promoted the
decomposition of the bicarbonate species, and the produced
CO2 was re-adsorbed on the surface of m-ZrO2, leading to the
formation of surface bidentate carbonate species.
With elevating of the temperature to 150 8C, the broad band
at 3600–2800 cm�1 related to adsorbed water molecular
reduced its intensity gradually. The bidentate carbonate species
became unstable and decomposed slowly. Whereas the bands at
Fig. 11. FTIR spectra of the Au/m-ZrO2 catalyst. (1) 15 min after 1% CO/He
flow at 25 8C; (2) 15 min after 1% CO/3% H2O/He flow 25 8C; and 5 min at (3)
50 8C; (4) 75 8C; (5) 100 8C; (6) 150 8C; (7) 200 8C; and (8) 250 8C.
1573, 1383, 1357, 2967 and 2870 cm�1 associated with formate
species gradually built up (curves 2–6) due to the interaction
between the adsorbed CO and the surface OH groups of m-
ZrO2. Further increase in the intensity of formate species was
observed when the temperature was raised up to 200 8C (curve
7). Meanwhile, the bands related to surface carbonate species
disappeared completely and the bands related to CO2 increased
remarkably. Hence, it is most likely that the formate species
acted as the important intermediates in the WGS reaction over
the Au/m-ZrO2 catalyst.
Fig. 12 also shows FTIR spectra of the Au/m-ZrO2 catalyst
recorded after CO/H2O/He was switched to He at 200 8C. The
bands at 1573, 1383, 1357, 2967 and 2870 cm�1 associated
with formate species gradually reduced their intensities. After
10 min of desorption, formate species have decomposed
completely and a small quantity of carbonate species still
retained on the surface. Therefore, it can be assumed that the
buildup of carbonate species could be one reason for the slow
deactivation of the Au/ZrO2 catalyst with time-on-stream.
4. Conclusion
Au nanoparticles supported on m-ZrO2 showed significantly
higher activity for WGS reaction than that on t-ZrO2. The
chemical state of gold was also proven to be important for the
WGS reaction. The active sites associated with metallic Au
nanoparticles are more active than the ones associated with
cationic Au+ species. FTIR measurements further identified
that the interaction between CO adsorbed on gold nanoparticles
and OH groups on the surface of m-ZrO2 produced formate
species, which acted as the most important intermediates and
finally decomposed into CO2 and H2 in the presence of water.
References
[1] D.L. Trimm, Z.I. Onsan, Catal. Rev. Sci. Eng. 43 (2001) 31.
[2] A.N. Fatsikostas, D.I. Kondarides, X.E. Verykios, Catal. Today 75 (2002)
145.
J. Li et al. / Applied Catalysis A: General 334 (2008) 321–329 329
[3] C.C. Elam, C.E.G. Padro, G. Sandrock, A. Luzzi, P. Lindblad, E.F. Hagen,
Int. J. Hydrogen Energy 28 (2003) 601.
[4] T. Tabakova, V. Idakiev, D. Andreeva, I. Mitov, Appl. Catal. A 202 (2000)
91.
[5] V. Idakiev, T. Tabakova, A. Naydenov, Z.Y. Yuan, B.L. Su, Appl. Catal. B
63 (2005) 178.
[6] X. Zhang, H. Shi, B.Q. Xu, Catal. Today 122 (2007) 330.
[7] X. Zhang, H. Wang, B.Q. Xu, J. Phys. Chem. B 109 (2005) 9678.
[8] W. Stichert, F. Schuth, S. Kuba, H. Knozinger, J. Catal. 198 (2001) 277.
[9] K.-H. Jacob, E. Knozinger, S. Benfer, J. Mater. Chem. 3 (1993) 651.
[10] W. Hertl, Langmuir 5 (1989) 96.
[11] K. Pokrovski, K.T. Jung, A.T. Bell, Langmuir 17 (2001) 4297.
[12] Y. Zhao, W. Li, M. Zhang, K. Tao, Catal. Commun. 3 (2002) 239.
[13] M.D. Rhodes, A.T. Bell, J. Catal. 233 (2005) 198.
[14] K.T. Jung, A.T. Bell, Catal. Lett. 80 (2002) 63.
[15] K.T. Jung, A.T. Bell, J. Mol. Catal. A 163 (2000) 27.
[16] Y. Boudeville, F. Figueras, M. Forissier, J.-L. Portefaix, J.C. Vedrine, J.
Catal. 58 (1979) 52.
[17] J.H. Scofield, J. Elec. Spec. 8 (1976) 129.
[18] X. Zhang, H. Shi, B.Q. Xu, Angew. Chem. Int. Ed. 44 (2005) 7132.
[19] E.D. Park, J.S. Lee, J. Catal. 186 (1999) 1.
[20] A.I. Kozlov, A.P. Kozlova, H. Liu, Y. Iwasawa, Appl. Catal. A 182 (1999)
9.
[21] M. Haruta, M. Date, Appl. Catal. A 222 (2001) 427.
[22] E. Chenu, G. Jacobs, A.C. Crawford, R.A. Keogh, P.M. Patterson, D.E.
Sparks, B.H. Davis, Appl. Catal. B 59 (2005) 45.
[23] F. Boccuxxi, G. Cerrato, F. Pinna, G. Strukul, J. Phys. Chem. B 102 (1998)
5733.
[24] C.H. Kim, L.T. Thompson, J. Catal. 244 (2006) 248.
[25] T. Tabakova, F. Boccuzzi, M. Manzoli, D. Andreeva, Appl. Catal. A 252
(2003) 385.
[26] F. Boccuzzi, A. Chiorino, M. Manzoli, P. Lu, T. Akita, S. Ichikawa, M.
Haruta, J. Catal. 202 (2001) 256.
[27] F. Boccuzzi, A. Chiorino, M. Manzoli, D. Andreeva, T. Tabakova, J. Catal.
188 (1999) 176.
[28] T. Tabakova, V. Idakiev, K. Tenchev, F. Boccuzzi, M. Manzoli, A.
Chiorino, Appl. Catal. B 63 (2005) 94.
[29] M. Manzoli, A. Chiorino, F. Boccuzzi, Surf. Sci. 532–532 (2003) 377.
[30] C. Binet, M. Daturi, J.-C. Lavalley, Catal. Today 50 (1999) 207.
[31] U.R. Pillai, S. Deevi, Appl. Catal. A 299 (2006) 266.