Effects of cerium incorporation into zirconia on the activity of Cu/ZrO2 for methanol synthesis via CO hydrogenation

Article abstract of DOI:10.1016/j.jcat.2005.09.002

The effects of Ce incorporation into ZrO₂ on the catalytic performance of Cu/ZrO₂ for the hydrogenation of CO were investigated. A Ce_0.3Zr_0.7O₂ solid solution was synthesized by forced hydrolysis at low pH. After calcination at 873 K, X-ray diffraction and Raman spectroscopy characterization indicated that the Ce _0.3Zr_0.7O₂ had a t″ crystal structure. It was found that 1.2 wt% Cu/Ce_0.3Zr_0.7O₂ exhibited H₂ consumption peaks at low temperature (<473 K) during H₂-TPR, indicating that a significant fraction (～70%) of Ce ⁴⁺ is reduced to Ce³⁺. The 1.2 wt% Cu/Ce _0.3Zr_0.7O₂ is 2.7 times more active for methanol synthesis than 1.2 wt% Cu/m-ZrO₂ at 3.0 MPa at temperatures between 473 and 523 K and exhibits a higher selectivity to methanol. In situ infrared spectroscopy shows that, analogous to Cu/m-ZrO₂, the primary surface species on Cu/Ce_0.3Zr_0.7O₂ during CO hydrogenation are formate and methoxide species. A shift in the band position of the bridged methoxide species indicated that some of these groups were bonded to both Zr⁴⁺ and Ce³⁺ cations. For both catalysts, the rate-limiting step for methanol synthesis is the reductive elimination of methoxide species. The higher rate of methanol synthesis on Cu/Ce _0.3Zr_0.7O₂ relative to Cu/m-ZrO₂ was due primarily to a ～2.4 times higher apparent rate constant, kapp, for methoxide hydrogenation, which is attributed to the higher surface concentration of H atoms on the former catalyst. The increased capacity of the Ce-containing catalyst is attributed to interactions of H atoms with CeO pairs present at the surface of the oxide phase.

Full text of DOI:10.1016/j.jcat.2005.09.002

Journal of Catalysis 235 (2005) 368–377
www.elsevier.com/locate/jcat
Effects of cerium incorporation into zirconia on the activity of Cu/ZrO2
for methanol synthesis via CO hydrogenation
∗
Konstantin A. Pokrovski, Michael D. Rhodes, Alexis T. Bell
Chemical Sciences Division, Lawrence Berkeley National Laboratory and Department of Chemical Engineering, University of California,
Berkeley, CA 94720-1462, USA
Received 18 May 2005; revised 25 August 2005; accepted 1 September 2005
Abstract
The effects of Ce incorporation into ZrO on the catalytic performance of Cu/ZrO for the hydrogenation of CO were investigated.
2
2
A Ce Zr O solid solution was synthesized by forced hydrolysis at low pH. After calcination at 873 K, X-ray diffraction and Raman spec-
0
.3 0.7 2
ꢀ
ꢀ
troscopy characterization indicated that the Ce Zr O had a t crystal structure. It was found that 1.2 wt% Cu/Ce Zr O exhibited H
0.3 0.7
2
0.3 0.7
2
2
4+
3+
consumption peaks at low temperature (<473 K) during H -TPR, indicating that a significant fraction (∼70%) of Ce is reduced to Ce . The
2
1
.2 wt% Cu/Ce Zr O is 2.7 times more active for methanol synthesis than 1.2 wt% Cu/m-ZrO at 3.0 MPa at temperatures between 473
0.3 0.7 2 2
and 523 K and exhibits a higher selectivity to methanol. In situ infrared spectroscopy shows that, analogous to Cu/m-ZrO , the primary surface
2
species on Cu/Ce Zr O during CO hydrogenation are formate and methoxide species. A shift in the band position of the bridged methoxide
0
.3 0.7 2
4+
3+
species indicated that some of these groups were bonded to both Zr and Ce cations. For both catalysts, the rate-limiting step for methanol
synthesis is the reductive elimination of methoxide species. The higher rate of methanol synthesis on Cu/Ce Zr O relative to Cu/m-ZrO
0.3 0.7
2
2
was due primarily to a ∼2.4 times higher apparent rate constant, kapp, for methoxide hydrogenation, which is attributed to the higher surface
concentration of H atoms on the former catalyst. The increased capacity of the Ce-containing catalyst is attributed to interactions of H atoms with
Ce–O pairs present at the surface of the oxide phase.

2005 Elsevier Inc. All rights reserved.
Keywords: Methanol synthesis; Cu; ZrO ; CO hydrogenation; Ce
2
1
. Introduction
ZrO2 (m-ZrO2, monoclinic zirconia) was shown to be nearly
an order of magnitude more active for methanol synthesis than
Cu/t-ZrO (t-ZrO , tetragonal zirconia) for equivalent zirconia
Several studies have indicated that support composition af-
2
2
fects the activity of Cu-based catalysts for CO hydrogenation
to methanol [1–6]. Zirconia has proven a particularly promis-
ing support for such applications, because Cu/ZrO2 catalysts
can be operated with or without the presence of CO2 [7–11].
Mechanistic studies have shown that for such catalysts, CO
adsorbs preferentially on the surface of zirconia to form for-
mate species, which then undergo hydrogenation to produce
methanol. The hydrogen atoms required for this process are
produced by dissociative adsorption of H2 on the surface of the
dispersed Cu and then spill over onto the zirconia surface. Re-
cent work by Bell et al. [12,13] has demonstrated that the phase
of zirconia can also influence catalyst activity. Thus Cu/m-
surface areas and surface concentrations of dispersed Cu. This
difference is attributed primarily to the presence of oxygen va-
cancies on the surface of m-ZrO , which facilitate the reaction
2
of CO with neighboring hydroxyl groups to generate formate
species and provide additional sites for hydrogen activation and
storage. These properties result in higher CO adsorption ca-
pacities and higher rates of reductive elimination of methoxide
species on Cu/m-ZrO2 relative to Cu/t-ZrO2. Because the avail-
ability of atomic H on the surface of zirconia is essential for
the hydrogenation of adsorbed CO, a motivation exists for ex-
ploring ways to increase the hydrogen adsorption capacity of
zirconia.
CeO2 has been used in automotive three-way catalysts to
serve as an oxygen-buffering component during rich/lean oscil-
lations of exhaust gases [14–18]. The ability of CeO2 to either
*
Corresponding author.
E-mail address: bell@cchem.berkeley.edu (A.T. Bell).
0021-9517/$ – see front matter  2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2005.09.002

K.A. Pokrovski et al. / Journal of Catalysis 235 (2005) 368–377
369
donate its oxygen or adsorb and store oxygen derived from
molecular sources stems from the facile, reversible redox cycle
was maintained at 573 K for 1 h before switching to a flow of
100% H for an additional 1 h.
2
4+
3+
Ce ꢀ Ce . Reduction of CeO2 to CeO2−x stoichiometries
results in the generation of oxygen vacancies throughout the ox-
ide lattice. Incorporation of Zr4+ into the CeO2 lattice enhances
2.2. Catalyst characterization
reducibility and increases the oxygen-storage capacity of the
solid solution [14,15,17,19]. Spectroscopic studies suggest that
the coordination number of the smaller Zr ion decreases in the
mixed oxide, increasing oxygen mobility and the creation of
vacancies [20].
The present study was undertaken to examine the influ-
ence of Ce incorporation into ZrO2 on the catalytic activ-
ity of Cu/CexZr1−xO2 for methanol synthesis from CO/H2.
ZrO2 and Ce0.3Zr0.7O2 were prepared by forced hydrolysis
at low pH and Cu was dispersed on the surface of these
materials by deposition precipitation. Catalysts were char-
acterized by X-ray diffraction (XRD), Raman spectroscopy,
and temperature-programmed reduction (TPR) in H2, and cat-
alyst H2 and CO adsorption capacities were measured by
temperature-programmed desorption (TPD). Steady-state cat-
alytic performance measurements were supplemented with dy-
namic surface studies using in situ infrared spectroscopy.
The crystallographic phase of each material was determined
by both XRD and Raman spectroscopy. XRD patterns were ob-
tained with a Siemens D5000 diffractometer, using Cu-Kα radi-
ation and a graphite monochrometer. Scans were made in the 2θ
◦
◦
◦
range of 10 to 120 with a step size of 0.02 and a time/step of
1 s. Rietveld refinement was performed using EXPGUI GSAS
software [21]. Raman spectra were recorded with a HoloLab
000 Raman spectrometer (Kaiser Optical) at room tempera-
1
5
−1
ture at a resolution of 2 cm . The stimulating light source
was a Nd:YAG laser, the output of which was frequency dou-
bled to 532 nm. Laser power at the sample was approximately
2
0 mW.
The BET surface area of each ZrO2 support was determined
using an Autosorb 1 (Quantachrome Instruments) gas adsorp-
tion system with nitrogen adsorption/desorption isotherms. Be-
fore each analysis, samples were dried at 393 K under vacuum
for >2 h. BET surface areas were calculated using a five-
2
. Experimental
point isotherm. After calcination, the surface areas of ZrO2 and
2
2
.1. Catalyst preparation
Ce0.3Zr0.7O were 143 and 120 m /g, respectively.
2
The concentration of exchangeable hydrogen on each sam-
ZrO2 was prepared by boiling a 0.5-M solution of zirconyl
ple was quantified using H/D exchange. Fully reduced samples
were purged with He at 298 K for 30 min, followed by temper-
ature ramping at 20 K/min from 298 to 623 K in 40 cm /min
chloride (ZrOCl2 · 8H2O, 99.99%, Aldrich) under reflux for
3
2
40 h. The final solution had a pH < 1. NH4OH was added
dropwise to agglomerate the resulting fine particles, to facili-
tate their filtration. The recovered precipitate was washed and
redispersed in deionized water several times to remove resid-
ual chlorine. AgNO3 was used to test the filtrate for any re-
maining Cl anions. The material was then dried in air at 383
K overnight before calcination. Ce0.3Zr0.7O2 was prepared by
boiling a 0.5-M solution (total metals basis) of zirconyl ni-
trate [ZrO(NO3)2 · 6H2O, 99.99%, Aldrich] and cerium(III) ni-
trate [Ce(NO3)3 · 6H2O, 99.999%, Aldrich] under reflux for
of D2. Both HD and H2 evolution were monitored using a mass
spectrometer (Minilab, Spectra Products), but only HD genera-
tion was observed. For each sample, exchange was complete by
∼
523 K.
H2-TPR studies were conducted using 0.15 g of a calcined
sample purged with He at 298 K for 30 min. The flow was
3
switched to 2% H2/He at a flow rate of 60 cm /min and then
ramped from 298 to 673 K while monitoring H2 consumption
using a mass spectrometer. Cu surface areas were quantified
using N2O titrations followed by H2-TPR [22]. A value of
2
40 h. The final solution had a pH < 1. NH4OH was added
dropwise to agglomerate the resulting fine particles to facili-
tate their filtration. The precipitated material was washed and
then dried overnight before calcination. Each sample was cal-
19
2
1
.46 × 10 Cu atoms/m and a stoichiometry of 2 Cu/H2 was
used [23].
Adsorption capacities for H2 and CO and were determined
3
cined in dry air flowing at 100 cm /min. The temperature was
using TPD. In the case of H2 the sample was calcined and re-
duced at 573 K. The sample was then cooled in H2 to 298 K
and purged in He. Desorption was conducted by ramping the
sample temperature at 20 K/min from 298 to 773 K in flowing
He while monitoring the desorbing gas using a mass spectrom-
eter. To determine the CO adsorption capacity, the sample was
calcined and reduced, then cooled to 523 K and flushed with
He for 30 min. Then 4.0% CO/He was flowed over the cata-
ramped from room temperature at a rate of 2 K/min to the fi-
nal temperature, which was maintained for 3 h. The ZrO2 and
Ce0.3Zr0.7O2 samples were calcined at 738 and 873 K, respec-
tively. Then 1.2 wt% Cu/ZrO2 catalysts were prepared with
each calcined support by deposition precipitation as described
previously [12].
Before testing or characterization, each catalyst sample was
3
calcined in a 10% O2/He mixture flowing at 60 cm /min.
3
lyst at a flow rate of 60 cm /min for 20 min. The sample was
The sample was heated from room temperature to 573 K at
0
.5 K/min and maintained at 573 K for 2 h. The sample was
then cooled to 298 K in 4.0% CO/He before being purged with
He to remove any weakly adsorbed species. Desorption of ad-
sorbed CO was carried out in a manner identical to that used for
adsorbed H2.
then cooled to 323 K, swept with He, and reduced in a 10%
H2/He mixture at a flow rate of 60 cm /min while increasing
the temperature by 2 K/min up to 573 K. The 10% H2/He flow
3

370
K.A. Pokrovski et al. / Journal of Catalysis 235 (2005) 368–377
2
.3. Catalyst testing
monoclinic ZrO2 was confirmed by Rietveld refinement in the
monoclinic symmetry, P21/c (Figs. 1d and e). The XRD pat-
Activity and selectivity measurements for CO hydrogenation
tern for Ce0.3Zr O exhibited a primary diffraction peak cen-
0.7
◦
2
were carried out in a continuous-flow, fixed-bed reactor as de-
scribed previously [12]. Reactions were carried out with 0.15 g
of catalyst at a pressure of 3.0 MPa. Total reactant gas flow was
tered at 2θ ∼ 30 , characteristic of both the tetragonal and cubic
phases of ZrO (t-ZrO and c-ZrO , respectively) (Fig. 1a).
2
2
2
◦
The position of the main peak, Ce0.3Zr0.7O , occurred at 29.5 ,
as opposed to 30.2 for t-ZrO and c-ZrO , consistent with
2
3
◦
6
0 cm /min, with a H2/CO ratio of 3/1. Product gas mixtures
2 2
were analyzed after 2 h on stream at a given temperature. Tem-
perature was then ramped to the successive temperature at a rate
of 2 K/min. Conversion and selectivity were referenced to CO,
the limiting reactant.
an increase in the lattice cell parameter with the substitution
of Zr4 (radius 0.86 Å) by Ce cations (radius 0.97 Å). The
symmetry of the diffraction peaks and the absence of features
associated with CeO2 suggest the formation of a CexZr1−xO2
solid solution [25,26]. The thermodynamically stable phase of
CexZr1−xO2 solutions is monoclinic for Zr-rich compositions
+
4+
2
.4. Infrared spectroscopy studies
(
<10 at% Ce) and cubic for Ce-rich compositions (>80 at%
In situ transmission infrared spectroscopy experiments were
Ce) [27,28]. Intermediate compositions are reported to exhibit
a number of stable and metastable phases of tetragonal sym-
metry [29–31]. The phase of Ce0.3Zr0.7O2 was analyzed by
Rietveld refinement of the measured diffraction pattern. Re-
finement was performed in both cubic (Fm3m, a = 5.20 Å)
conducted using a low-dead volume infrared cell with CaF2
windows [24]. Infrared spectra were recorded with Magna 750
−1
Series II IR spectrometer (Nicolet) at a resolution of 2 cm . In
an effort to remove any residual surface species before testing,
each sample was calcined in a 10% O2/He mixture flowing at
and tetragonal (P42/nmc, a = 3.68, c = 5.19, af/c = 1.00(3),
3
√
6
5
0 cm /min. The sample was heated from room temperature to
23 K at 2 K/min and maintained at 523 K for 8 h. The sample
where af = a 2 ) symmetries (Figs. 1b and c, respectively).
The broadness of the diffraction peaks made it impossible to
identify the displacement of oxygen ions from their ideal flu-
orite positions by Rietveld refinement; hence a more definitive
assessment of the crystalline structure of Ce0.3Zr0.7O2 was at-
tempted using Raman spectroscopy [32].
was then cooled to 323 K, swept with He, and reduced in a 10%
3
H2/He mixture at a flow rate of 60 cm /min while increasing
the temperature at the rate of 2 K/min up to 523 K. The flow
of 10% H2/He was maintained at 523 K for 1 h before switch-
ing to a flow of 100% H2 for an additional 1–3 h. The sample
was then flushed with He for 1 h before sample testing.
The Raman spectra of ZrO2 and Ce0.3Zr0.7O2 are presented
in Fig. 2. ZrO2 exhibited peaks at 182, 333, 377, 475, 559,
−1
and 623 cm , characteristic of m-ZrO2 [33–35]. Thus both
XRD and Raman spectroscopy confirmed that the ZrO2 sam-
ple was m-ZrO2. The Ce0.3Zr0.7O2 sample exhibited a prin-
3
. Results and discussion
3
.1. Material characterization of ZrO2 supports
XRD patterns of the ZrO2 and Ce0.3Zr0.7O2 are shown in
−
1
ciple peak at 474 cm and two smaller peaks at 316 and
6
−1
−1
21 cm . The peak at 474 cm was characteristic of a cu-
bic fluorite structure [36,37]; however, the presence of peaks at
Fig. 1. The diffraction pattern for ZrO2 exhibited peaks charac-
teristic of monoclinic ZrO2 (m-ZrO2). The occurrence of pure
−
1
3
16 and 621 cm and the absence of the characteristic tetrag-
−1
onal peak at ∼275 cm indicated that Ce0.3Zr0.7O2 was in the
ꢀ
ꢀ
t phase [29]. This phase is intermediate between the tetrago-
Fig. 1. XRD patterns for ZrO and Ce Zr O . (a) Observed profile for
2
0.3 0.7 2
Ce Zr O , (b) and (c) difference between observed and calculated profiles
0.3 0.7 2
in cubic symmetry, Fm3m space group, and tetragonal symmetry, P4 /nmc
2
space group, respectively; (d) observed profile for ZrO ; and (e) difference be-
2
tween observed and calculated profiles in monoclinic symmetry P2 /c space
1
group).
Fig. 2. Raman spectra for ZrO and Ce Zr O at room temperature.
2 0.3 0.7 2

K.A. Pokrovski et al. / Journal of Catalysis 235 (2005) 368–377
371
Table 1
Cu surface area and exchangeable H for 1.2 wt% Cu/m-ZrO and 1.2 wt%
2
Cu/Ce Zr
O
2
0
.3 0.7
Sample
Cu surface area Cu dispersion Exchangeable
2
(%)
2
(
m /g)
H (µmol/m )
1
.2 wt% Cu/m-ZrO
0.87
0.73
11.1
9.3
10.9
7.8
2
1.2 wt% Cu/Ce Zr
O
2
0.3 0.7
nal and cubic phases, in which the oxygen atoms are displaced
from the positions of these atoms in an ideal fluorite structure.
ꢀ
ꢀ
Whereas the t phase is commonly reported to exist at Ce com-
positions between 65 and 80 at% [29], others have reported the
generation of this phase at lower Ce content for small particles
[
38,39].
3
.2. Material characterization of Cu/m-ZrO2 and
ꢀ
ꢀ
Cu/t -Ce0.3Zr0.7O2
Fig. 4. H -TPR spectra for 1.2 wt% Cu/m-ZrO and 1.2 wt% Cu/Ce Zr O .
2
2
0.3 0.7 2
A loading of 1.2 wt% Cu was deposited onto ZrO2 and
Ce0.3Zr0.7O2. The Cu surface area and associated dispersion for
each sample are given in Table 1. The Cu particles dispersed on
m-ZrO2 had a slightly higher overall Cu surface area, but each
catalyst had a very similar Cu surface area to support surface
area ratio. The surface concentration of exchangeable hydrogen
for each sample, quantified by H/D exchange, is also listed in
Table 1. In situ infrared studies have shown that this quantity
indicates the concentration of hydroxyl groups on the cata-
lyst surface [40,41]. The surface concentration of OH groups
was ∼40% higher on 1.2 wt% Cu/m-ZrO2 than on 1.2 wt%
Cu/Ce0.3Zr0.7O2. The infrared spectra of the O–H stretching
region after calcination and reduction for 1.2 wt% Cu/m-ZrO2
and 1.2 wt% Cu/Ce0.3Zr0.7O2 are shown in Fig. 3. The spec-
tra were referenced to the empty cell in He. Each sample ex-
3
Heating rate = 20 K/min; 2% H /He flow rate = 60 cm /min.
2
were very similar to those for 1.2 wt% Cu/m-ZrO2 (3668
−1
and 3729 cm ). Analogous to 1.2 wt% Cu/m-ZrO2, 1.2 wt%
ꢀꢀ
Cu/t -Ce0.3Zr0.7O2 exhibited a higher relative concentration of
low-frequency hydroxyl groups. The positions of these features
were similar to those reported previously for OH bands on
t-ZrO2 and m-ZrO2 [39–41]. Although the exact positions of
such bands can depend on the degree of surface hydroxylation,
the higher-frequency band is commonly assigned to terminal
hydroxyl groups, and the lower-frequency band is commonly
assigned to either bi- or tri-bridging groups [40,42]. Recent the-
oretical studies of the surfaces of both ZrO2 polymorphs also
lead to the conclusion that a combination of terminal and both
bi- and tri-bridged OH groups should be present [43–46].
The H2-TPR profiles for the catalysts are shown in Fig. 4.
The 1.2 wt% Cu/m-ZrO2 exhibited one principle peak with a
maximum at 473 K, attributed to the reduction of highly dis-
−1
hibited two types of isolated hydroxyl groups (>3600 cm ).
Despite a pseudo-tetragonal/cubic crystal structure, the peak
−1
positions of 1.2 wt% Cu/Ce0.3Zr0.7O2 (3670 and 3731 cm
)
persed CuO or Cu2 ions in an octahedral environment [42,43].
The amount of H2 consumed was slightly greater than the
value corresponding to the complete reduction of CuO species
−
ꢀ
ꢀ
(
H2/CuO ≈ 1.1). The 1.2 wt% Cu/t -Ce0.3Zr0.7O2 exhibited
significantly greater H2 consumption, due to the reduction of
4+
3+
Ce to Ce . If the H2/CuO ratio is assumed to be ∼1, then
the additional H2 consumption corresponds to a reduction of
∼
70% of the Ce atoms in the support. Reduction occurs with
two primary peaks at 403 and 443 K. The presence of ei-
ther noble or base metals facilitates reduction of Ce because
of spillover of H atoms from the metal surface [44–46]. Sim-
ilar reduction peaks have been reported during H2-TPR with
CuO/Ce0.44Zr0.56O2 [47].
Table 2 lists the H2 and CO adsorption capacities of 1.2 wt%
ꢀ
ꢀ
Cu/m-ZrO2 and 1.2 wt% Cu/t -Ce0.3Zr0.7O2 determined from
TPD spectra. The amount of H2 adsorbed per unit area was
ꢀ
ꢀ
3
1
.3 times higher for 1.2 wt% Cu/t -Ce0.3Zr0.7O2 than for
.2 wt% Cu/m-ZrO2. Although the total amount of CO ad-
Fig. 3. Infrared spectra of the hydroxyl group stretching region taken for
.2 wt% Cu/ZrO and 1.2 wt% Cu/Ce Zr following calcination and re-
duction. Spectra referenced to empty cell in He.
1
O
0.3 0.7 2
sorbed was significantly higher than the amount of H2 adsorbed
on both catalysts, the ratio of CO adsorbed on 1.2 wt% Cu/t -
2
ꢀ
ꢀ

372
K.A. Pokrovski et al. / Journal of Catalysis 235 (2005) 368–377
Table 2
Effect of Ce incorporation on the adsorption capacity and binding strength of CO at 523 K
Sample
CO desorbed
2
Peak max. T
(K)
CO desorbed
2
Peak max. T
(K)
Total COx desorbed
H desorbed
2
2
2
2
(
µmol/m )
(µmol/m )
(µmol/m )
(µmol/m )
1
.2 wt% Cu/m-ZrO
0.51
0.94
620
575, 615
0.77
0.60
590, 675
635
1.28
1.54
0.26
0.87
2
1.2 wt% Cu/Ce Zr
O
0.3 0.7 2
(a)
(b)
Fig. 5. Effect of temperature on the selective conversion (a) and overall selectivity (b) of CO to methanol during CO hydrogenation: catalyst mass, 0.15 g;
3
p = 3.0 MPa; H /CO = 3; total flow rate, 60 cm /min.
2
Ce0.3Zr0.7O2 relative to that on 1.2 wt% Cu/m-ZrO2 was only
a factor of 1.2. Thus it is apparent that the dominant effect of
incorporating Ce into the zirconia lattice is to increase the H2
adsorption capacity of the oxide. Although the physical reason
for this effect is not known, we propose that it may be due to
the increased ability of Ce–O bonds in the oxide to stabilize H
frared spectroscopy. Fig. 6 shows spectra obtained during
CO adsorption on 1.2 wt% Cu/m-ZrO2 at 523 K. The cat-
alyst was exposed to a flow of 15% CO/He at a total pres-
sure of 0.50 MPa. The bands observed at 1563, 1386, and
−1
1366 cm
can be attributed to the νas(OCO), δ(CH), and
νs(OCO) modes of b-HCOO–Zr, respectively [2,5,50–56]. Ac-
companying features for b-HCOO–Zr in the CH stretching
δ−
δ+
atoms as H and H species [48,49].
−
1
−1
region occurred at 2969, 2883 cm , and 2748 cm , charac-
teristic of [νas(OCO) + δ(CH)] [49,50,53], ν(CH) [2,5,46–51],
and [νs(OCO) + δ(CH)] [54,57], respectively. C–H stretching
3
.3. Catalytic performance of Cu/m-ZrO2 and
ꢀ
ꢀ
Cu/t -Ce0.3Zr0.7O2
−
1
features attributed to CH3O–Zr (2934 and 2830 cm ) [4,37,
50,51,53,55,58,59] appeared almost immediately, even in the
The effects of reaction temperature on the activity and
ꢀ
ꢀ
selectivity of the 1.2 wt% Cu/m-ZrO2 and 1.2 wt% Cu/t -
Ce0.3Zr0.7O2 catalysts are shown in Fig. 5. Conversion to
methanol increased over the temperature range 473–523 K and
was accompanied by a decrease in selectivity. The only major
byproduct observed was methane. These conversion levels were
far below the equilibrium values for the given temperatures, in-
dicating that the observed rate of methanol formation was not
absence of gas phase H . These features continued to increase
2
for the duration of the CO adsorption experiment (∼1 h). At
−1
the same time, bands at 1039 and 1142 cm appeared and
increased in intensity at a rate similar to that for the bands
−1
at 2934 and 2830 cm . These peaks were assigned to C–O
stretching vibrations of terminal (t-OCH3) and bridged (b-
OCH3) methoxide species on ZrO2, respectively [58,60,61].
ꢀ
ꢀ
−1
influenced by methanol decomposition. The 1.2 wt% Cu/t -
Ce0.3Zr0.7O2 catalyst exhibited higher conversion to methanol
than the m-ZrO2 supported catalysts over the temperature range
studied; for example, the methanol productivity of 1.2 wt%
The shoulder located at approximately 1320 cm was assigned
to b-CO32 –Zr species [41,56].
−
Fig. 7 shows spectra obtained during CO adsorption on
1.2 wt% Cu/Ce0.3Zr0.7O2 at 523 K. The catalyst was exposed
to a flow of 15% CO/He at a total pressure of 0.50 MPa. As
for the 1.2 wt% Cu/m-ZrO2, here the primary bands can be
attributed to bidentate formate species with features at 1576,
ꢀ
ꢀ
Cu/t -Ce0.3Zr0.7O2 was approximately double that of 1.2 wt%
Cu/m-ZrO2 at 523 K. The higher activity of the Ce0.3Zr0.7O2-
supported catalyst was also accompanied by a greater selectiv-
ity to methanol.
−1
1379, 1370, 2966, and 2878 cm . Each band shifted relative
to the position on pure ZrO2, but bands reported for formate
species on CeO2 were noticeably absent [62–66]. A compar-
ison with Fig. 6 reveals that formate species were formed at a
3
.4. Infrared spectroscopy studies
The nature of surface species and the dynamics of CO
adsorption and hydrogenation were studied using in situ in-
2
faster rate on 1.2 wt% Cu/Ce0.3Zr0.7O . Methoxide species fea-
tures at 1143, 1040, 2934, and 2828 cm were nearly coinci-
−1

K.A. Pokrovski et al. / Journal of Catalysis 235 (2005) 368–377
373
Fig. 6. Infrared spectra taken for 1.2 wt% Cu/m-ZrO at 523 K after switching feed from 0.50 MPa He to 0.05 MPa CO and 0.45 MPa He flowing at a total rate of
2
3
6
0 cm /min. Spectra referenced to 1.2 wt% Cu/m-ZrO under 0.50 MPa He flow at 523 K.
2
Fig. 7. Infrared spectra taken for 1.2 wt% Cu/Ce Zr
rate of 60 cm /min. Spectra referenced to 1.2 wt% Cu/Ce Zr O under 0.50 MPa He flow at 523 K.
O at 523 K after switching feed from 0.50 MPa He to 0.05 MPa CO and 0.45 MPa He flowing at a total
2
0.3 0.7
3
0.3 0.7 2
−
dent with those observed on 1.2 wt% Cu/m-ZrO2 and increased
over the duration of the transient. In addition to formate and
methoxide species, the appearance of broad features at ∼1550–
approximately 3–8 cm 1, presumably due to interactions be-
tween surface species with increasing surface coverage. The
−1
δ(CH) feature for CH3O–Zr (1446 cm ) [5,54,58] was evident
initially but remained relatively small throughout the transient.
Fig. 9 shows transient response spectra obtained during CO
hydrogenation on 1.2 wt% Cu/t -Ce0.3Zr0.7O2 at 523 K. Spec-
tra were obtained by introducing H2 into the flowing 15%
CO/He (total pressure = 0.50 MPa) to achieve a H2/CO ratio
of 3/1. After the introduction of H2, the broad bands ascribable
to carbonate and carboxylate species decreased rapidly. The
features for b-HCOO at 1576, 1379, and 1370 cm generally
decreased over the course of the transient, whereas CH3O fea-
tures at 1148, 1061, 2923, and 2816 cm increased rapidly and
then approached steady-state levels. The ν(CO) mode of ter-
minal CH3O species was nearly coincident with that observed
with m-ZrO2, but the ν(CO) mode of bridged CH3O species
−1
1
450 and 1350–1250 cm is indicative of various carbonate
and carboxylate species [54,67], suggesting that the presence of
Ce favors the formation of these species. Similar broad bands
have been observed after CO adsorption at high temperature on
CeO2 [67].
ꢀ
ꢀ
After CO adsorption for 1 h, transient response spectra were
obtained during CO hydrogenation by introducing H2 into the
flowing 15% CO/He (total pressure = 0.50 MPa) to achieve a
H2/CO ratio of 3/1. Fig. 8 shows spectra obtained on 1.2 wt%
Cu/m-ZrO2 after the switch at 523 K. The features of b-HCOO–
−1
−1
−1
Zr at 1564, 1385, and 1367 cm decreased rapidly in inten-
sity as soon as H2 was introduced and simultaneously peaked
−1
for CH3O–Zr at 1147 and 1047 cm . In the C–H stretching
region, bands for CH3O–Zr increased until reaching a steady-
state level after approximately 80 min, analogous to the pat-
tern observed for the corresponding C–O stretching vibrations.
During CO hydrogenation, each C–H stretch red shifted ap-
−1
occurred at a higher wave number (1061 vs. 1047 cm ). Binet
et al. [68,69] reported C–O stretches of bridged CH3O species
on Ce3 sites at ∼1080 cm . Because H2-TPR demonstrates
+
−1
4+
3+
that most surface Ce cations are reduced to Ce , the upward
−1
proximately 1015 cm whereas each C–O stretch blue shifted
shift of the bridged ν(CO) mode likely reflects the presence

374
K.A. Pokrovski et al. / Journal of Catalysis 235 (2005) 368–377
Fig. 8. Infrared spectra taken for 1.2 wt% Cu/m-ZrO at 523 K after switching feed from 0.05 MPa CO and 0.45 MPa He to 0.05 MPa CO, 0.15 MPa H , and
2
2
3
0
.30 MPa He flowing at a total rate of 60 cm /min. Spectra referenced to 1.2 wt% Cu/m-ZrO under 0.50 MPa He flow at 523 K.
2
Fig. 9. Infrared spectra taken for 1.2 wt% Cu/Ce Zr O at 523 K after switching feed from 0.05 MPa CO and 0.45 MPa He to 0.05 MPa CO, 0.15 MPa H , and
0
.3 0.7
2
2
3
0
.30 MPa He flowing at a total rate of 60 cm /min. Spectra referenced to 1.2 wt% Cu/Ce Zr O under 0.50 MPa He flow at 523 K.
0.3 0.7 2
of methoxide species bonded to both Zr4 and Ce cations.
+
3+
∼1550–1450 and 1350–1250 cm . Hydrogenation of carbon-
ate species to formate species was previously observed during
studies of CO2 hydrogenation on Cu/ZrO2 [70]. On both cata-
lysts, the concentration of methoxide species increased rapidly
from the levels observed during CO adsorption, reaching values
within 95% of the steady-state concentration in <1 h.
−1
Consistent with a reduction of most of the Ce atoms in the
Ce0.3Zr0.7O2 lattice, no terminal CH3O was observed on Ce
4+
−1
cations at ∼1105 cm [68,69].
Fig. 10 compares the evolution of the principle surface
ꢀ
ꢀ
species on 1.2 wt% Cu/m-ZrO2 and 1.2 wt% Cu/t -Ce0.3Zr0.7O2
during the CO hydrogenation experiments presented in Figs. 8
and 9. The ν(CH3) and νas(OCO) bands were used to track the
development of CH3O–Zr and b-HCOO–Zr species, respec-
tively. Peak areas for both b-HCOO–Zr and CH3O–Zr were
normalized to the observed values after approximately 12 h.
The concentration of formate species on 1.2 wt% Cu/m-ZrO2
decreased rapidly from the level observed in the absence of
gas phase H2 during the first 15 min, before declining more
slowly over the remainder of the transient. In contrast, the con-
centration of formate species reached a maximum at ∼3 min
The relative rates of consumption of formate and methox-
ide species were examined by switching from a CO/H2 mixture
to only H2. Fig. 11 compares the dynamics of consumption of
b-HCOO–Zr and CH3O–Zr species on 1.2 wt% Cu/m-ZrO2
and 1.2 wt% Cu/t -Ce0.3Zr0.7O2. Transient response spectra
were obtained by replacing the 15% CO/He in the 3/1 H2/CO
flow with He while maintaining the total pressure at 0.50 MPa.
Peak areas for both b-HCOO–Zr and CH3O–Zr were nor-
malized to the value observed at the beginning of the tran-
sient. The b-HCOO species decreased more rapidly initially on
1.2 wt% Cu/m-ZrO2, al though the overall decrease in inten-
sity was similar for both materials at the end of the transient.
The concentration of CH3O species decreased more rapidly
ꢀ
ꢀ
ꢀ
ꢀ
on 1.2 wt% Cu/t -Ce0.3Zr0.7O2 before decreasing more slowly
over the course of the experiment. The maximum formate con-
centration is attributed to the rapid hydrogenation of surface
carbonate and/or carboxylate species to produce new formate
groups, as evidenced by the rapid decline in the broad bands at
ꢀ
ꢀ
on 1.2 wt% Cu/t -Ce0.3Zr0.7O2 than on 1.2 wt% Cu/m-ZrO2,
and there was a significant disparity in the relative concentra-

K.A. Pokrovski et al. / Journal of Catalysis 235 (2005) 368–377
375
Fig. 10. Peak areas of (a) b-HCOO–Zr and (b) CH O–Zr features for 1.2 wt% Cu/m-ZrO and 1.2 wt% Cu/Ce Zr
O
during the experiments in Figs. 5.9
3
2
0.3 0.7
2
and 5.10. Areas normalized to the values observed at the end of the transient.
Fig. 11. Peak areas of (a) b-HCOO–Zr, (b) CH O–Zr, and (c) CH O–Zr features for 1.2 wt% Cu/m-ZrO and 1.2 wt% Cu/Ce Zr
O at 523 K after switching
2
3
3
2
0.3 0.7
3
feed from 0.05 MPa CO, 0.15 MPa H , and 0.30 MPa He to 0.15 MPa H and 0.35 MPa He flowing at a total rate of 60 cm /min. Areas normalized to the values
2
2
observed at the beginning of the transient.
tion of these species at the conclusion of the 6-h experiment.
The initial decrease in methoxide species intensity also oc-
curred more rapidly on 1.2 wt% Cu/t -Ce0.3Zr0.7O2 (∼2.5 min)
than on 1.2 wt% Cu/m-ZrO2 (∼4 min); see Fig. 11c. The ap-
parent first-order rate constants for the removal of methoxide
species, kapp, determined from the initial portion of the tran-
ꢀ
ꢀ
−3
−1
sient were 4.1 × 10 min for 1.2 wt% Cu/m-ZrO and
2
−2
−1
ꢀꢀ
1.0 × 10 min for 1.2 wt% Cu/t -Ce0.3Zr0.7O2.

376
K.A. Pokrovski et al. / Journal of Catalysis 235 (2005) 368–377
Fig. 12. Infrared spectra taken for (a) 1.2 wt% Cu/m-ZrO and (b) 1.2 wt% Cu/Ce Zr
O at 523 K after switching feed from 0.05 MPa CO, 0.15 MPa H , and
2 2
2
0.3 0.7
3
0
.45 MPa He to 0.15 MPa H and 0.35 MPa He flowing at a total rate of 60 cm /min. Spectra referenced to 1.2 wt% Cu/m-ZrO and 1.2 wt% Cu/Ce Zr
O
2
2
0.3 0.7 2
under 0.50 MPa He flow at 523 K, respectively.
Fig. 12 shows the evolution of the O–H stretching region for
Cu/m-ZrO2. The higher value of kapp for the former catalyst is
ꢀ
ꢀ
ꢀꢀ
both 1.2 wt% Cu/m-ZrO2 and 1.2 wt% Cu/t -Ce0.3Zr0.7O2 dur-
ing the H2 reduction experiments presented in Fig. 10. Each ma-
terial exhibited hydroxyl group consumption bands at ∼3728
quite possibly due to the greater H adsorption capacity of t -
2
Ce0.3Zr0.7O , because k
2
app
is the product of the intrinsic rate
coefficient for reductive elimination and the coverage of the
oxide surface by H atoms. If the intrinsic rate coefficients on
both catalysts were assumed to be the same, then the observed
ratio of apparent coefficients would imply a 2.9 (corrected for
the ratios of oxide surface areas) higher H-atom coverage on
−1
and ∼3670 cm at the conclusion of CO hydrogenation (t = 0
in Fig. 12). After the switch to an H2-only feed, the bands
decreased in negative absorbance over time, indicating a net
recovery of hydroxyl groups over the course of the experi-
ment. This is consistent with removal of formate and methoxide
groups adsorbed at these sites during reaction through hydro-
genation to methanol [5]. Notably, a greater portion of the hy-
ꢀꢀ
t -Ce0.3Zr0.7O2, which is consistent with the measured ratio
in H2 adsorption capacities (3.3). Based on the peak areas of
the terminal and bridged methoxide C–O stretches, the ratio of
methoxide species adsorbed on Ce0.3Zr0.7O2 relative to ZrO2
is 1.3 (corrected for the ratios of oxide surface areas) under
ꢀ
ꢀ
droxyl groups on 1.2 wt% Cu/t -Ce0.3Zr0.7O2 was recovered
during the course of the transient. Based on the changes in
peak intensity, the recovery rates for hydroxyl groups after 6 h
were ∼35% on 1.2 wt% Cu/m-ZrO2 and ∼75% on 1.2 wt%
steady-state reaction conditions. Hence, the ratio of the prod-
ꢀꢀ
uct kappθCH O is ∼3.8 in favor of 1.2 wt% Cu/t -Ce0.3Zr0.7O2.
ꢀ
ꢀ
3
Cu/t -Ce0.3Zr0.7O2. These values agree well with the fraction
of methoxide species removed during the same experiment (see
Fig. 12) (∼42% on 1.2 wt% Cu/m-ZrO2 and ∼78% on 1.2 wt%
This figure is in reasonable agreement with the observed ratio
of methanol synthesis activities (∼2.7) once the measured rates
are corrected for differences in the surface areas of the oxide
and the dispersed Cu.
ꢀ
ꢀ
Cu/t -Ce0.3Zr0.7O2), and suggest that most of the recovery in
OH band intensity is associated with the removal of methoxide
groups. Such a result would be expected from a consideration
of the relative intensities of the band intensities for methoxide
and formate species (see Figs. 8 and 9).
4
. Conclusions
In their studies of methanol synthesis over Cu/t-ZrO2 and
Cu/m-ZrO2, Bell et al. [13] established that the rate-limiting
step is reductive elimination of Zr–OCH3 species. As a re-
sult, the observed rate of methanol synthesis is proportional
to the product of the apparent first-order rate coefficient for
methoxide reductive elimination, kapp, and the surface con-
The activity for methanol synthesis from CO and H2 of
ꢀ
ꢀ
Cu/t -Ce0.3Zr0.7O2 is approximately two-fold higher than that
for Cu/m-ZrO , assuming the same surface areas of the ox-
2
4+
4+
ides and the dispersed Cu. Although Ce cations replace Zr
cations isomorphously in the zirconia lattice during synthesis,
the substituted cations are readily reducible to Ce3 . The pres-
ence of Ce increases the adsorption capacity of the oxide for
H atoms produced by dissociative adsorption of H2 on the dis-
persed particles of Cu and spillover onto the oxide surface. The
concentration of adsorbed methoxide species is also somewhat
higher when Ce is substituted into the zirconium lattice. These
+
centration of methoxide groups, θCH O. The same reasoning
3
can be used to interpret the higher rate of methanol synthesis
ꢀ
ꢀ
on Cu/t -Ce0.3Zr0.7O2 relative to Cu/m-ZrO2 observed in the
present study. The values of kapp based on the initial data of
−
3
−1
the transient shown in Figs. 11b and c were 4.1 × 10 min
−2
−1
for 1.2 wt% Cu/m-ZrO2 and 1.0 × 10 min for 1.2 wt%
ꢀ
ꢀ
ꢀ
ꢀ
Cu/t -Ce0.3Zr0.7O2. Hence the apparent rate coefficient was
two factors appear to account for the higher activity of Cu/t -
ꢀꢀ
.4 higher for 1.2 wt% Cu/t -Ce0.3Zr0.7O2 than for 1.2 wt%
2
Ce0.3Zr0.7O2.

K.A. Pokrovski et al. / Journal of Catalysis 235 (2005) 368–377
377
Acknowledgments
[35] C. Li, M. Li, J. Raman Spectrosc. 33 (2002) 301.
[
[
[
[
36] A. Trovarelli, F. Zamar, J. Lorca, C. Leitenburg, G. Dolcetti, J.T. Kiss,
J. Catal. 169 (1997) 490.
37] V.S. Escribano, E.F. López, M. Panizza, C. Resini, J.M.G. Amores,
G. Busca, Solid State Sci. 5 (2003) 1369.
38] P. Fornasiero, G. Balducci, R. Di Monte, J. Kašpar, V. Sergo, G. Gubitosa,
A. Ferrero, M. Graziani, J. Catal. 164 (1996) 173.
39] S. Enzo, F. Delogu, R. Frattini, A. Primavera, A. Trovarelli, J. Mater.
Res. 15 (2000) 1538.
This work was supported by the Director, Office of Basic
Energy Sciences, Chemical Sciences Division, US Department
of Energy under contract DE-AC03-76SF00098.
References
[
[
[
[
1] J. Saussey, J.C. Lavalley, J. Mol. Catal. 50 (1989) 343.
[40] D. Martin, D. Duprez, J. Phys. Chem. B 101 (1997) 4428.
[41] K.D. Jung, A.T. Bell, J. Catal. 193 (2000) 207.
[42] M. Shimokawabe, H. Asakawa, N. Takezawa, Appl. Catal. 59 (1990) 45.
[43] R. Zhou, T. Yu, X. Jiang, F. Chen, X. Zheng, Appl. Surf. Sci. 148 (1999)
263.
2] C. Schild, A. Wokaun, A. Baiker, J. Mol. Catal. 63 (1990) 243.
3] S. Fujita, M. Usui, H. Ito, N. Takezawa, J. Catal. 157 (1995) 403.
4] J. Weigel, R.A. Koeppel, A. Baiker, A. Wokaun, Langmuir 12 (1996)
5319.
[
[
[
[
[
5] I.A. Fisher, A.T. Bell, J. Catal. 178 (1998) 153.
[44] A. Trovarelli, G. Dolcetti, C. de Leitenburg, J. Kašpar, P. Finetti, A. San-
toni, J. Chem. Soc., Faraday Trans. 88 (1992) 1311.
[45] S. Bernal, J.J. Calvino, G.A. Cifredo, A. Laachir, V. Perrichon, J.M. Her-
rmann, Langmuir 10 (1994) 717.
[46] C. de Leitenburg, A. Trovarelli, J. Kašpar, J. Catal. 166 (1997) 98.
[47] M. Boaro, M. Vicaro, C. de Leitenburg, G. Dolcetti, A. Trovarelli, Catal.
Today 77 (2003) 407.
[48] C. Lamonier, A. Ponchel, A. D’Huysser, L. Jalowiecki-Duhamel, Catal.
Today 50 (1999) 247.
[49] J.L.G. Fierro, J. Soria, J. Sanz, J.M. Rojo, J. Solid State Chem. 66 (1987)
154.
6] G. Meitzner, E. Iglesia, Catal. Today 53 (1999) 433.
7] B. Denise, R.P.A. Sneeden, Appl. Catal. 28 (1986) 235.
8] Y. Sun, P.A. Sermon, J. Chem. Soc. Commun. (1993) 1242.
9] Y. Sun, P.A. Sermon, Catal. Lett. 29 (1994) 361.
[
[
[
[
[
[
[
10] I.A. Fisher, H.C. Woo, A.T. Bell, Catal. Lett. 44 (1997) 11.
11] Y.W. Suh, S.H. Moon, H.K. Rhee, Catal. Today 63 (2000) 447.
12] M.D. Rhodes, A.T. Bell, J. Catal. 233 (2005) 198.
13] M.D. Rhodes, K.A. Pokrovski, A.T. Bell, J. Catal. 233 (2005) 210.
14] M. Ozawa, J. Alloys Compd. 275–277 (1998) 886.
15] J. Kašpar, P. Fornasiero, M. Graziani, Catal. Today 50 (1999) 285.
16] T. Masui, T. Ozaki, K. Machida, G. Adachi, J. Alloys Compd. 303–304
[50] M.Y. He, J.G. Ekerdt, J. Catal. 87 (1984) 381.
[51] J. Kondo, H. Abe, Y. Sakata, K. Maruya, K. Domen, T. Onishi, J. Chem.
Soc., Faraday Trans. I 84 (1988) 511.
(2000) 49.
[
17] E. Mamontov, T. Egani, R. Brezny, M. Korrane, S. Tyagi, J. Phys. Chem.
B 104 (2000) 11110.
[52] W. Hertl, Langmuir 5 (1989) 96.
[
[
18] J. Kašpar, P. Fornasiero, J. Solid State Chem. 171 (2003) 19.
19] M. Daturi, E. Finocchio, C. Binet, J.C. Lavalley, F. Fally, U. Perrichon,
H. Vidal, N. Hickey, J. Kašpar, J. Phys. Chem. B 104 (2000) 9186.
20] G. Vilac, R. Di Monte, P. Fornasiero, E. Fonda, J. Kašpar, M. Graziani,
Catalysis and Automotive Pollution Control IV, Studies in Surface Sci-
ence and Catalysis, vol. 116, Elsevier, Amsterdam, 1998.
[53] E. Guglielminotti, Langmuir 6 (1990) 1455.
[54] D. Bianchi, T. Chafik, M. Khalfallah, S.J. Teichner, Appl. Catal. A 105
(1993) 223.
[55] H. Kalies, N. Pinto, G.M. Pajonk, D. Bianchi, Appl. Catal. A 202 (2000)
197.
[56] K. Pokrovski, K.T. Jung, A.T. Bell, Langmuir 17 (2001) 4297.
[57] J. Edwards, G. Schrader, J. Phys. Chem. 89 (1985) 782.
[58] M. Bensitel, V. Moravek, J. Lamonte, O. Saur, J.C. Lavalley, Spectrochim.
Acta 43a (1987) 1487.
[
[21] B.H. Toby, J. Appl. Crystallogr. 34 (2001) 210–213.
[22] S. Sato, R. Takahashi, T. Sodesawa, K. Yuma, Y. Obata, J. Catal. 196
(2000) 195.
[
[
23] K.J. Sorenson, N.W. Cant, Catal. Lett. 33 (1995) 117.
24] R.F. Hicks, C.S. Kellner, B.J. Savatsky, W.C. Hecker, A.T. Bell, J. Ca-
tal. 71 (1981) 216.
[59] D. Bianchi, T. Chafik, M. Khalfallah, S.J. Teichner, Appl. Catal. A 123
(1995) 89.
[60] F. Ouyang, J.N. Kondo, K. Maruya, K. Domen, J. Phys. Chem. B 101
(1997) 4867.
[25] C.E. Hori, H. Permana, K.Y. Simon Ng, A. Brenner, K. More, K.M. Rah-
moeller, D.N. Belton, Appl. Catal. B 16 (1998) 105.
[61] F. Ouyang, J.N. Kondo, K. Maruya, K. Domen, Catal. Lett. 50 (1998) 179.
[62] C. Li, Y. Sakata, T. Arai, K. Domen, K. Maruya, T. Onishi, J. Chem. Soc.,
Faraday Trans. I 85 (1989) 929.
[63] T. Shido, Y. Iwasawa, J. Catal. 141 (1993) 71.
[64] E. Finocchio, M. Daturi, C. Binet, J.C. Lavalley, G. Blanchard, Catal. To-
day 52 (1999) 53.
[
[
[
26] C. Bozo, F. Gaillard, N. Guilhaume, Appl. Catal. A 220 (2001) 69.
27] E. Tani, M. Yoshimura, S. Smiya, J. Am. Ceram. Soc. 66 (1983) 506.
28] P. Duran, M. Gonzales, C. Moure, J.R. Jurdo, C. Pascal, J. Mater. Sci. 25
(1990) 5001.
[
[
[
[
29] M. Yashima, H. Arashi, M. Kakihana, M. Yoshimura, J. Am. Ceram.
Soc. 77 (1994) 1067.
30] M. Yashima, K. Morimoto, N. Ishizawa, M. Yoshimura, J. Am. Ceram.
Soc. 76 (1993) 1745.
31] M. Yashima, K. Morimoto, N. Ishizawa, M. Yoshimura, J. Am. Ceram.
Soc. 77 (1993) 2865.
32] J.R. Ferraro, K. Nakamoto, Introductory Raman Spectroscopy, Academic
Press, New York, 1994.
[65] G. Jacobs, L. Williams, U. Graham, G.A. Thomas, D.E. Sparks, B.H.
Davis, Appl. Catal. A 252 (2003) 107.
[66] G. Jacobs, P.M. Patterson, L. Williams, E. Chenu, D. Sparks, G. Thomas,
B.H. Davis, Appl. Catal. A 262 (2004) 177.
[67] C. Li, Y. Sakata, T. Arai, K. Domen, K. Maruya, T. Onishi, J. Chem. Soc.,
Faraday Trans. I 85 (1989) 1451.
[68] A. Badri, C. Binet, J.C. Lavalley, J. Chem. Soc., Faraday Trans. 93 (1997)
1159.
[33] C. Schild, A. Wokaun, R.A. Koeppel, A. Baiker, J. Catal. 130 (1991) 657.
34] M. Li, Z. Feng, G. Xiong, P. Ying, Q. Xin, C. Li, J. Phys. Chem. B 105
[
[69] C. Binet, M. Daturi, Catal. Today 70 (2001) 155.
[70] I.A. Fisher, A.T. Bell, J. Catal. 172 (1997) 222.
(2001) 8107.