In situ studies of the active sites for the water gas shift reaction over Cu-CeO2 catalysts: Complex interaction between metallic copper and oxygen vacancies of ceria

Article abstract of DOI:10.1021/jp055467g

New information about the active sites for the water gas shift (WGS) reaction over Cu-CeO₂ systems was obtained using in-situ, time-resolved X-ray diffraction (TR-XRD), X-ray absorption spectroscopy (TR-XAS, Cu K and Ce L₃ edges), and infrared spectroscopy (DRIFTS). Cu-CeO₂ nanoparticles prepared by a novel reversed microemulsion method (doped Ce_1-xCu_xO₂ sample) and an impregnation method (impregnated CuO_x/ CeO₂ sample) were studied. The results from all of the samples indicate that both metallic copper and oxygen vacancies in ceria were involved in the generation of active sites for the WGS reaction. Evidence was found for a synergistic Cu-O _vacancy interaction. This interaction enhances the chemical activity of Cu, and the presence of Cu facilitates the formation of O vacancies in ceria under reaction conditions. Water dissociation occurred on the O _vacancy sites or the Cu-O_vacancy interface. No significant amounts of formate were formed on the catalysts during the WGS reaction. The presence of strongly bound carbonates is an important factor for the deactivation of the catalysts at high temperatures. This work identifies for the first time the active sites for the WGS reaction on Cu-CeO₂ catalysts and illustrates the importance of in situ structural studies for heterogeneous catalytic reactions.

Full text of DOI:10.1021/jp055467g

428
J. Phys. Chem. B 2006, 110, 428-434
In Situ Studies of the Active Sites for the Water Gas Shift Reaction over Cu-CeO2
Catalysts: Complex Interaction between Metallic Copper and Oxygen Vacancies of Ceria
Xianqin Wang,† Jose´ A. Rodriguez,*,† Jonathan C. Hanson,† Daniel Gamarra,‡
Arturo Mart´ınez-Arias,*,‡ and Marcos Ferna´ndez-Garc´ıa*,‡
Chemistry Department, BrookhaVen National Laboratory, Upton, New York 11973, and Instituto de Cata´lisis y
Petroleoqu´ımica, CSIC, Campus Cantoblanco, 28049 Madrid, Spain
ReceiVed: September 26, 2005; In Final Form: October 27, 2005
New information about the active sites for the water gas shift (WGS) reaction over Cu-CeO2 systems was
obtained using in-situ, time-resolved X-ray diffraction (TR-XRD), X-ray absorption spectroscopy (TR-XAS,
Cu K and Ce L3 edges), and infrared spectroscopy (DRIFTS). Cu-CeO2 nanoparticles prepared by a novel
reversed microemulsion method (doped Ce1-xCuxO2 sample) and an impregnation method (impregnated CuOx/
CeO2 sample) were studied. The results from all of the samples indicate that both metallic copper and oxygen
vacancies in ceria were involved in the generation of active sites for the WGS reaction. Evidence was found
for a synergistic Cu-Ovacancy interaction. This interaction enhances the chemical activity of Cu, and the presence
of Cu facilitates the formation of O vacancies in ceria under reaction conditions. Water dissociation occurred
on the Ovacancy sites or the Cu-Ovacancy interface. No significant amounts of formate were formed on the
catalysts during the WGS reaction. The presence of strongly bound carbonates is an important factor for the
deactivation of the catalysts at high temperatures. This work identifies for the first time the active sites for
the WGS reaction on Cu-CeO2 catalysts and illustrates the importance of in situ structural studies for
heterogeneous catalytic reactions.
Introduction
CO conversions in the WGS than commercial Cu/ZnO
catalysts.13,17-19
Nowadays, we face new challenges in creating alternate fuels,
cleaning the environment, dealing with the causes of global
warming, and keeping us safe from toxic substances and
infectious agents.1 Hydrogen is a potential solution for satisfying
many of our energy needs while reducing (and eventually
eliminating) carbon dioxide and other greenhouse gas emis-
sions.2 At present, nearly 95% of the hydrogen supply is
produced from the reforming of crude oil, coal, natural gas,
wood, organic wastes, and biomass.3 The reformed fuel contains
1-10% CO, which degrades the performance of the Pt electrode
in fuel cell systems.4 To get clean hydrogen for fuel cells and
other industrial applications, the water gas shift (WGS) reaction
(CO + H2O f CO2 + H2) and CO oxidation (2CO + O2 f
2CO2) processes are critical.3 To accelerate them, heterogeneous
catalysts are frequently used.5 A fundamental understanding of
the configuration and properties of the active sites for the WGS
reaction is a prerequisite for designing catalysts with a high
activity or efficiency.5
In this work, the behavior of CuOx/CeO2 and Ce1-xCuxO2
catalysts for the WGS reaction is investigated. CeO2-based
catalysts have been reported to be very promising for the WGS
reaction owing to the peculiar redox properties of ceria and its
oxygen storage capacity.5 CuO-CeO2 mixed-metal oxides have
important applications as electrodes in fuel cells6 and gas
sensors7 and as efficient catalysts for various reactions, such as
the combustion of CO and methane,8-11 the WGS,11-13 the
reduction of SO2 by CO,14 methanol synthesis,15 and the wet
oxidation of phenol.16 It is anticipated that with proper develop-
ment, metal-promoted ceria catalysts should realize much higher
The roles played by the ceria and copper in the WGS over
Cu-CeO2 catalysts are a matter of debate.16-22 Either metallic
Cu or Cu1+ cations have been proposed as active sites for the
WGS reaction.22 Ceria may not be a simple spectator and may
play a direct role in the catalytic process.20,21 Concerning the
reaction mechanism, there is no agreement, either. A mechanism
with formate (CO + OH f HCOO) as an intermediate formed
on partially reduced ceria has been proposed13 in which the
breaking of the C-H bond in the formate was the rate-limiting
step for the WGS. On the other hand, a redox reaction (CO +
O f CO2) was alternatively proposed to be the rate-limiting
step by other authors.12 They also suggested that Cu was the
active site for the catalysis and that ceria had no effect on the
activity.12 In a similar way, metallic copper was proposed to
be the only active species for the steam reforming of dimethyl
ether,23 whereas the copper/ceria interface was reported to be
the active phase for the steam reforming of methanol or CO
oxidation reactions.9,24
In situ characterization of the Cu-CeO2 catalysts under
operational conditions is critical to clarify the roles of ceria and
Cu. In recent works, time-resolved X-ray diffraction (TR-XRD)
and time-resolved X-ray absorption fine structure (TR-XAFS)
were employed to study in situ the interaction of copper oxides
or ceria with H2 and CO at atmospheric pressures.25-28 These
powerful techniques, the product of combining the high intensity
of synchrotron radiation with rapid, new, parallel data-collection
devices, allowed a systematic study of the kinetics and mech-
anism for the reduction of CuO, Cu2O, and CeO2.26-29 In this
article, in situ TR-XRD, TR-XAFS, and infrared spectroscopy
(DRIFTS) are employed to examine the behavior of CuOx/CeO2
† Chemistry Department, Brookhaven National Laboratory.
‡ Instituto de Cata´lisis y Petroleoquimica.
10.1021/jp055467g CCC: $33.50 © 2006 American Chemical Society
Published on Web 12/08/2005

Metallic Cop per-Ceria Oxygen Vacancy Interactions
J. Phys. Chem. B, Vol. 110, No. 1, 2006 429
and Ce1-xCuxO2 nanoparticles under different WGS reaction
conditions and gas environments.
of the NSLS. The same cell was used for the XAFS experiments
as that for in situ XRD, except that the sample was loaded into
a Kapton capillary (DuPont) and heated with hot air. The X-ray
absorption spectra were taken repeatedly in the “fluorescence
yield mode” using a pips detector cooled with circulating water.
XAFS data have been analyzed using the Athena and Artimis
programs.34 They allow background removal by optimization
of the low-R portion of the EXAFS data (i.e., minimizing the
average value of the low-R Fourier transform, maintaining at
the same time the value for the first coordination shell at
maximum).35
The products from both TR-XRD and TR-XAFS experiments
were measured with a 0-100 amu quadrupole mass spectrom-
eter (QMS, Stanford Research Systems). A portion of the exit
gas flow passed through a leak valve and into the QMS vacuum
chamber. QMS signals at mass-to-charge ratios of 2(H2), 4(He),
12(C), 15(CH3), 17(OH), 18(H2O), 28(CO), 32(O2), and 44-
(CO2) were monitored during the experiments, and these were
recorded at the same time by an online computer.
In Situ DRIFTS. In situ DRIFTS analysis of the samples
under WGS reaction conditions was carried out using a Bruker
Equinox 55 FTIR spectrometer fitted with an MCT detector.
The DRIFTS cell (Harrick) was fitted with CaF2 windows and
a heating cartridge that allowed samples to be heated to 500
°C. Samples of ∼100 mg were exposed to 3% H2O and 5%
CO (He balance) at a flow corresponding to 1.5 × 105 h-1 space
velocity. The catalysts were heated under the reactant flow using
a ramp of 10 °C min-1. The heating ramp was stopped every
100 °C for recording a spectrum (average of 50 scans at 4 cm-1
resolution) after achieving steady state conditions based on the
product concentration measured simultaneously with a mass
spectrometer.
Materials and Characterization Methods
Materials Preparation. The reference CuO, Cu2O, and Cu
bulk samples used in this work were obtained from Alfa-Aesar
(>99.99% purity). Cu-doped ceria, labeled as Ce1-xCuxO2 (x
) 0.05 or 0.2) was prepared with a modified reversed micro-
emulsion method.25,30 The precursors were introduced in a
reverse microemulsion (water in oil) using n-heptane as the
organic phase, Triton X-100 (Aldrich) as surfactant, and hexanol
as cosurfactant. Then this suspension was mixed with another
similar suspension containing an alkali solution (tetramethyl-
ammonium hydroxide, Aldrich) as aqueous phase. All cations
were coprecipitated, and the resulting mixtures were stirred for
24 h, centrifuged, decanted, and rinsed with methanol. Finally,
the solid portion was dried overnight at 100 °C, and the resulting
powders were calcined under air at 500 °C for 2 h. Full details
of the preparation parameters employed can be found else-
where.30 ICP-AES chemical analysis of these samples confirmed
quantitative precipitation of both Cu and Ce cations. Surface
area values (SBET) obtained for these samples were observed to
increase with the copper content (130 and 151 m2 g-1 for Ce0.95
-
Cu0.05O2 and Ce0.8Cu0.2O2, respectively), in correlation with the
size decrease observed in the oxide nanoparticles.31
The sample of copper supported on CeO2 (5% weight in Cu,
denoted hereafter as 5Cu) was synthesized by incipient wetness
impregnation of a CeO2 support, prepared by microemulsion
in a similar manner as described above, with a copper nitrate
aqueous solution. Following impregnation, the sample was dried
overnight at 100 °C and, finally, calcined under air at 500 °C
for 2 h. The SBET value obtained for this sample prepared by
impregnation was 101 m2 g-1, close to that of the parent CeO2
support.
Results and DIscussion
Chemical State of Cu During the WGS Reaction. 5%
CuOx/CeO2 Sample. A typical set of time-resolved X-ray
diffraction patterns is shown in Figure 1a for the 5% CuOx/
CeO2 (denoted as 5Cu) impregnated sample during the WGS
reaction at different temperatures. A copper oxide phase was
clearly observed at the beginning of experiments at 25 °C and
transformed to metallic copper as soon as the temperature was
increased to 300 °C. By comparing the product curves (Figure
1b) with the XRD patterns, it is clear that the signals for H2
and CO increase concomitantly with the appearance of metallic
copper, indicating the important role of this species during the
water gas shift reaction. The concentration of products increased
when the temperature was raised to 400 °C, but it decreased at
500 °C, pointing to a deactivation of the 5Cu sample at high
temperature, probably due to the production of strongly bound
carbonate species (see below) that poison the active sites of the
catalyst.22,36
Similar experiments were carried out with in situ time-
resolved XAFS. Figure 2a displays the Fourier transforms of
the Cu K-edge EXAFS spectra collected over the 5Cu sample
cooled to room temperature after the WGS reaction at 25, 200,
300, and 400 °C. The position of the main peak located between
1.1 and 2.0 Å at room temperature, which corresponds to the
first Cu-O coordination shell, is similar to that of CuO.26 The
Cu-O peak decrease was combined with an increase in the Cu-
Cu peak in metallic copper and disappeared at high temperatures.
The product curves from in-situ XAFS experiments were not
shown here, since they were similar to those for in-situ TR-
XRD experiments. The results confirmed the important role of
metallic copper in the WGS reaction also observed in in situ
XRD results.
In Situ Time-Resolved X-ray Diffraction. In situ time-
resolved X-ray diffraction (TR-XRD) experiments were carried
out on beam line ×7B of the National Synchrotron Light Source
(NSLS) at Brookhaven National Laboratory. The sample was
loaded into a sapphire capillary cell that was attached to a flow
system.25-27A small resistance heater was wrapped around the
capillary, and the temperature was monitored with a 0.1-mm
chromel-alumel thermocouple that was placed straight into the
capillary near the sample. Two-dimensional powder patterns
were collected with a Mar345 image plate detector, and the
powder rings were integrated using the FIT2D code.32 The
instrument parameters (Thompson-Cox-Hastings profile coef-
ficients and Berar-Baldinozzi asymmetry correction) were
derived from the fit of a Si reference pattern. Lattice constants
were determined by a Rietveld analysis using the GSAS
(General Structure Analysis System) program.33 Diffraction
patterns were collected over the catalysts during the WGS
reaction and in different gas environments including 5% CO/
He, 5% H2O/He, and 5% O2/He (named as gas-switch experi-
ment). The WGS reaction was carried out isothermally at several
temperatures (200, 300, 400, and 500 °C) with a 5% CO and
95% He gas mixture at a flow rate of ∼10 mL/min [or a space
velocity, defined as (volumetric feed flow rate)/(catalyst vol-
ume), of ∼1.5 × 105/h]. This gas mixture passed through a water
tube before entering the reactor. The relative ratio of water vapor
pressure to CO in the gas mixture was ∼0.6.
In Situ Time-Resolved X-ray Absorption. Cu K-edge and
Ce L3-edge XAFS spectra were collected in air at room
temperature and in situ under different operation conditions,
similar to those for the TR-XRD experiments, at beamline ×18B

430 J. Phys. Chem. B, Vol. 110, No. 1, 2006
Wang et al.
Figure 1. (a) Time-resolved X-ray diffraction patterns for 5% CuOx/
CeO2 catalyst during the water gas shift reaction at different temper-
atures (λ ) 0.922 Å). (b) The relative concentration of H2 and CO2
products as a function of time at different temperatures during the water
gas shift reaction as shown in (a).
Figure 2. (a) Cu K-edge EXAFS spectra collected over 5% CuOx/
CeO2 cooled down to room temperature after the water gas shift reaction
at different temperatures. (b) Cu K-edge EXAFS spectra collected over
Ce0.8Cu0.2O2 cooled down to room temperature after the water gas shift
reaction at different temperatures.
Cu-Doped Ceria Samples. The WGS activity of the Ce1-x
-
CuxO2 compounds was tested with a methodology similar to
that described above for the 5% CuOx/CeO2 catalyst. Figures
2b and 3 show the results for the Ce0.8Cu0.2O2 sample. The Cu
K-edge EXAFS spectra were collected working at reaction
temperatures of 25, 100, 200, 300, and 400 °C. Under WGS
reaction conditions, Cu atoms in Ce0.8Cu0.2O2 were seen to
transform to Cu2O at 100 °C (as was confirmed by the line
shape of the FT trace and the peak at ∼3.1 Å characteristic of
Cu2O),31 then to metallic copper at elevated temperatures.
However, WGS activity was detected with the mass spectrom-
eter (data not shown) only after the formation of metallic copper.
The corresponding XRD patterns in Figure 3 indicate that no
copper or copper oxides were observed at low temperatures. It
must be taken into account in this respect that under these
conditions, the Cu atoms are more or less uniformly distributed
through the bulk and the surface of the ceria particles.31
However, clear features for metallic copper were seen at
temperatures above 150 °C, and their intensity increased when
raising the temperature. This reveals that copper sinters during
reduction under WGS conditions in this sample, as shown to
occur also under H2 in a previous work.31 The mass spectrometer
signals for H2 and CO2 in Figure 3b show a low WGS activity
at temperatures below 220 °C, and much higher activity at 300
and 400 °C with the presence of metallic copper. Very similar
In summary, the in-situ studies of TR-XRD and TR-XAFS
show that metallic copper plays a vital role in the WGS reaction
over both CuOx/CeO2 and Ce1-xCuxO2 catalysts. The copper
oxide species initially present in the catalysts must undergo a
complete reduction before any substantial WGS activity is seen.
WGS Activity Comparison for Different Cu-CeO2
Samples. Figure 4 compares the production of H2 during the
WGS reaction over the same amount (∼5 mg) of CuO, Ce0.8
-
Cu0.2O2, Ce0.95Cu0.05O2, and 5% CuOx/CeO2 at different tem-
peratures. No activity was observed over pure nano ceria under
the present operating conditions. Figure 4 clearly indicates that
the impregnated 5% CuOx/CeO2 sample had better WGS activity
than those of doped samples. The catalyst based on pure CuO
had little activity. This catalyst certainly had the largest
concentration of copper sites among the studied samples. These
results indicate that interactions between copper and ceria
enhance the activity of the metal in Cu-CeO2 catalysts,
suggesting that the WGS reaction may take place preferentially
on the copper/ceria interface. In the next section, we will
examine these options and the behavior of ceria during the WGS
reaction.
Reduction/Oxidation of Ceria under Different Gas Envi-
ronments. It is well-known that oxygen vacancies are easily
formed in ceria nanoparticles.25 In Cu-CeO2 samples, there are
results were found when examining the behavior of a Ce0.95
-
Cu0.05O2 catalyst.

Metallic Cop per-Ceria Oxygen Vacancy Interactions
J. Phys. Chem. B, Vol. 110, No. 1, 2006 431
Figure 3. Time-resolved X-ray diffraction patterns for Ce0.8Cu0.2O2
catalyst during the water gas shift reaction at different temperatures (λ
) 0.992 Å). (b) The relative concentration of H2 and CO2 products as
a function of time at different temperatures during the water gas shift
reaction as shown in (a).
Figure 5. Ce L3-edge XANES spectra collected over Ce0.8Cu0.2O2
during the water gas shift reaction at different temperatures. (b) Ce
L3-edge XANES spectra collected over 5% CuOx/CeO2, Ce0.8Cu0.2O2,
nano ceria during the WGS reaction at 300 °C, and references of bulk
CeO2 and Ce(NO3)3‚6H2O.
tions (H2O/CO ∼ 0.6), H2O/He, or 5% O2/He at different
temperatures. The changes in the chemical state of Ce during
the WGS reaction were monitored by taking in situ Ce L3-edge
XANES spectra.
Ce L3-Edge TR-XANES Results. The Ce L3 edge is frequently
used as a “fingerprint” to characterize the electronic properties
of ceria-based materials.37 The electronic transitions behind these
XANES features are complex. In pure stoichiometric CeO2, the
Ce L3 edge exhibits two clear peaks, frequently labeled B1 and
C. A third peak, with lower photon energy than B1, can be
obtained by curve fitting.37
Figure 5a compares the Ce L3-edge XANES spectra from
the Ce0.8Cu0.2O2 sample in the water gas shift reaction at
different temperatures with the spectrum for a Ce(NO3)3‚6H2O
reference, in which the cerium atoms are trivalent. The two main
peaks in the spectrum of the sample at room temperature are
Figure 4. Relative concentrations of H2 produced from the WGS
reaction at different temperatures over 5% CuOx/CeO2, Ce0.8Cu0.2O2,
Ce0.95Cu0.05O2, and CuO using the same operating conditions (space
velocity ∼1.5 × 105/h).
two main processes to form oxygen vacancies: Ce4+ reduction
and Cu2+ replacement in the ceria lattice.31 In the reduction
process, two Ce4+ cations are reduced to two Ce3+ cations with
one oxygen vacancy formed (2Ce4+ + O2- f 2Ce3+ + Ovac).
Since Ce3+ is bigger with respect to Ce4+, a simple reduction
leads to a ceria lattice cell expansion.25 Different from the
reduction process, the ceria lattice cell will slightly contract if
one Cu2+ cation replaces one Ce4+ cation with one oxygen
vacancy (Ce4+ + O2- f Ce4+ + Cu2+ + Ovac).31 Thus, changes
in the lattice parameter can be directly related to the concentra-
tion of oxygen vacancies and Cu atoms in ceria. Variations in
the ceria lattice were studied exposing the 5% Cu/CeO2 and
Cu0.2Ce0.8O2 samples to He, 5% CO/He, WGS reaction condi-
separated by ∼7 eV, in agreement with previous results.25,26,37
.
On the basis of the comparison of the intensities of the spectra
near the Ce3+ position, we were able to confirm that oxygen
vacancies and Ce3+ were formed during the WGS reaction. The
amount of oxygen vacancies and Ce3+ were seen to increase
with the increase in reaction temperature up to 300 °C, but
decreased at higher temperatures, especially above 400 °C.
When the Ce0.8Cu0.2O2 catalyst was exposed to 5% CO in He,
the amount of Ce3+ formed was bigger than for the experiments
in Figure 5a and increased continuously with the temperature
of exposure. These results show that ceria was oxidized by H2O
under the WGS reaction, with oxygen vacancies and Ce3+ being
eliminated, especially at high temperature.

432 J. Phys. Chem. B, Vol. 110, No. 1, 2006
Wang et al.
Figure 7. Ceria lattice parameters in the gas-switch experiments over
5% Cu/CeO2 at 300 °C (a) ceria and (b) metallic copper.
The XRD/XAFS results described above show that both
metallic copper and oxygen vacancies in ceria are involved in
the active sites for the WGS reaction. It is clear that neither
copper nor the O vacancies alone have the chemical properties
necessary for the high WGS activity seen over the Cu-CeO2
catalysts. From previous studies with single crystals and films,
it is well known that metallic copper is more active than pure
CuOx for the WGS.42 Recent studies performed in our laboratory
have shown that the O vacancies in ceria are very efficient for
the chemical activation of gold nanoparticles,43 and the same
probably occurs for copper nanoparticles.44 Which one plays
the leading role in the WGS, copper or the O vacancies? To
answer this question, gas-switch experiments were also carried
out over the 5% CuOx/CeO2 catalyst, and the ceria lattice
parameters in different gas environments are presented in Figure
7. A comparison shows that the variations in the ceria lattice
parameters for the Ce0.8Cu0.2O2 sample (Figure 6a) were much
larger than those for the 5% CuOx/CeO2 sample under WGS
reaction conditions, CO/He or H2O/He environments, indicating
the presence of a higher concentration of oxygen vacancies and
Ce3+ in the Ce0.8Cu0.2O2 sample. However, as shown in Figure
4, the activity for the 5% CuOx/CeO2 sample was significantly
higher than that for the Ce0.8Cu0.2O2 sample. Thus, the activity
of the Cu-CeO2 catalysts did not correlate directly with the
amount of oxygen vacancies or Ce3+ in the ceria support,
pointing to the leading role of metallic Cu. The lower activity
of the Ce0.8Cu0.2O2 sample could be due to the smaller amount
of surface metallic Cu atoms, since some Cu metal particles
can remain embedded in the ceria lattice and could not be easily
accessible to the reactants. On the other hand, carbonate species,
which may become poisons of the WGS reaction,22,36 could
affect to a greater extent the Ce0.8Cu0.2O2 catalyst, considering
the higher copper dispersion in it and the fact that formation of
such species is strongly promoted by the presence of copper-
ceria interfaces in this type of catalyst.45 This point will be
investigated in more detail in the next section.
Figure 6. Lattice parameters during the gas-switch experiments over
Ce0.8Cu0.2O2 at 300 °C: (a) ceria and (b) metallic copper.
Figure 5b presents the Ce L3-edge XANES spectra for Ce0.8
-
Cu0.2O2, 5% CuOx/CeO2 and nano ceria collected during the
WGS reaction at 300 °C plus reference spectra for bulk ceria
and Ce(NO3)3‚6H2O standards. The peak intensities at the Ce3+
position for the three samples followed the order Ce0.8Cu0.2O2
> 5% CuOx/CeO2 > nano ceria, pointing out the improved ceria
reducibility for the Cu-containing ceria samples, as reported in
the literature.9a,21a In accordance with this, the closer interaction
between copper and ceria in the mixed-metal oxide (or its higher
copper dispersion, compare Figures 1a and 3a) certainly explains
the higher ceria reducibility in it (Figure 5b).
Variations in Ceria Lattice during the WGS. Figure 6shows
the lattice parameters for ceria (Figure 6a) and metallic Cu
(Figure 6b) determined from (111) diffraction peaks of TR-
XRD patterns for Cu0.2Ce0.8O2 in different gases at 300 °C. The
sample was first heated to 300 °C in He, and no metallic Cu
was observed (Figure 6b). Metallic copper was formed as soon
as the gas was switched to 5% CO/He, stayed through the WGS
reaction and H2O exposure, and disappeared after O2 oxidation.
The disappearance of metallic Cu and the practical recovery of
the initial lattice parameter in an oxygen environment confirms
that the reduction and oxidation of Cu and Ce in Ce0.8Cu0.2O2
is reversible.31 The lattice of the formed metallic Cu did not
change under different gas environments of CO and H2O. In
contrast, the ceria lattice varied significantly (Figure 6a) with
an increase after exposure to CO and a decrease in H2O,
indicating that CO reduced ceria, and H2O oxidized it. In other
words, CO created oxygen vacancies, and H2O eliminated them.
The water-ceria interaction is a complex process and is closely
related to the surface termination and morphology of ceria.38
In this work, the H2O oxidation of CeO2-x is consistent with
studies over ceria powders39,40 and films of CeO2(111) grown
on Ru (0001).41 A comparison of the ceria lattices in Figure 6
for oxidation in water and O2 indicates that H2O does not
completely reoxidize the CeO2-x. The ceria lattice parameter
under the WGS reaction reflects a combination of the effects
of CO reduction and H2O oxidation, implying that oxygen
vacancies on the fluorite phase are involved in the chemistry
of the process.
Generation of OH and CO3 Groups on the Cu-CeO2
Catalysts: In Situ FTIR Studies. To understand what causes
the variations in activity seen in Figure 4, in situ DRIFTS
experiments were carried out over the Ce0.8Cu0.2O2 and 5%
CuOx/CeO2 samples. The corresponding spectra are shown in
Figure 8a and b, respectively; activity profiles obtained during
these runs (not shown) were similar to those presented in Figure
4. The spectra from both samples indicated the formation of a
Cu+ carbonyl species (showing a characteristic band in the
2120-2100 cm-1 range)9b at relatively low reaction tempera-
tures (displaying a maximum at 100 °C), in fair agreement with
XAFS results (Figure 2). Significant amounts of carbonate (or

Metallic Cop per-Ceria Oxygen Vacancy Interactions
J. Phys. Chem. B, Vol. 110, No. 1, 2006 433
related) species (whose most intense bands appear in the 1800-
1200 cm-1 range) were formed at a temperature as low as 100
°C, and their intensities increased with an elevation of the
reaction temperature. Monodentate carbonates (bands at ∼1480
and 1390 cm-1 46
and bidentate carbonates or carboxilates
)
(bands or shoulders at ∼1550 and 1330 cm-1) were formed at
100 °C, and a minor amount of formate species (bands or
shoulders at ∼2935, 2845, 1590, and 1360 cm-1 47
) appeared at
300 °C and displayed stronger intensity for Ce0.8Cu0.2O2. Other
bands observed in the spectra are those corresponding to
interacting hydroxyl species (broad band in the 3500-3000
cm-1 range), the intensities of which became stronger with
increasing reaction temperature. Perusal of the spectra show a
weak band overlapped on the R-branch of CO(g) (at ∼2175
cm-1) for the spectra recorded at 300 °C that may correspond
to an electronic transition, in particular, configurations of
Ce3+ 9b,48
.
In general, two main reaction mechanisms have been
proposed for the WGS over metal-promoted ceria catalysts.12,13,20
In one of them, H2O dissociates on the surface of the catalyst,
forming OH groups that then react with CO to yield formate
intermediates.13 The formates react with water, producing H2
and unidentate carbonates (rate-limiting step), which in the final
step decompose to generate CO2.13a This reaction mechanism
is probably not dominating the chemistry on the Cu-CeO2
catalysts, since the intensity of surface formate peaks seen in
the DRIFT data of Figure 8 is very small or negligible, and
furthermore, no correlation is found between this observable
and the respective catalytic activities (Figures 4 and 8). In a
second proposed reaction mechanism,12,20 CO bonds to metal
sites and then reacts with ceria to form CO2 and O vacancies
in the support. The water undergoes decomposition on the O
vacancies to generate H2.20 The data in Figures 6, 7, and 8 are
clear evidence of the feasibility of a redox-type reaction
mechanism for the Cu-CeO2 catalysts.
Comparing the data in Figures 4 and 8, the highest catalytic
activity of 5% CuOx/CeO2 could be a consequence of a large
concentration of copper on the surface that facilitates the
dissociation of water (larger IR signal for OH groups). In
addition, the Ce0.8Cu0.2O2 exhibits a stronger signal for carbon-
ates, and this could prevent the Ce3+/Ce4+ redox interplay at
the metal-promoter interface and poison active sites of the
catalyst.22,36 An indication that this type of phenomenon could
be occurring in this catalyst at high temperature was obtained
from the Ce L3 results discussed above. Moreover, the possibility
of ceria over-reduction has been discussed recently as being a
factor that would increase carbonate stability, resulting in surface
poisoning.49 So a careful balance between the redox state of
Cu and CeO2 phases could be needed for optimum catalytic
activity, and this may not only concern interface zones of both
constituents but also suffer from a medium to large range
influence. For 5% CuOx/CeO2, the catalytic activity decrease
at temperatures above 400 °C can be related to the significant
increase in the coverage of carbonates.
Figure 8. In situ DRIFTS spectra recorded under 3% H2O and 5%
CO at the indicated temperatures over 5% CuOx/CeO2. (b) In situ
DRIFTS spectra recorded under 3% H2O and 5% CO at the indicated
temperatures over Ce0.8Cu0.2O2.
presence of Cu facilitates the formation of O vacancies in ceria.
H2O dissociation occurred on the Ovacancy sites or the Cu-
Ovacancy interface. In situ DRIFTS indicates that the amount of
formate formed during the WGS reaction on the Cu-CeO2
catalysts is very small or negligible and does not correlate with
the activity results. The presence of strongly bound carbonates
appears to be an important factor for explaining the catalytic
properties of these systems, including deactivation at high
temperatures. This work identified for the first time the active
sites for the WGS reaction on Cu-CeO2 catalysts and illustrates
the importance of in situ structural studies for heterogeneous
catalytic reactions.
Acknowledgment. The research carried out at the Chemistry
Department of Brookhaven National Laboratory was financed
through Contract No. DE-AC02-98CH10086 with the U.S.
Department of Energy, Division of Chemical Sciences. The
NSLS is supported by the Divisions of Materials and Chemical
Sciences of DOE. Work at the ‘‘Instituto de Cata´lisis (CSIC)”
was done with financial support from CICYT (Contracts No.
CTQ2004-03409/BQU and MAT2003-03925). D. G. thanks the
FPI program of the Ministerio de Educacio´n y Ciencia for a
PhD grant. Thanks are due to Syed Khalid and Wolfgang
Caliebe for their help and facilities given in the operation of
Conclusions
In situ DRIFTS, time-resolved X-ray diffraction, and (Cu
K-edge and Ce L3-edge) X-ray absorption spectroscopy were
used to identify the active sites for the WGS reaction on CuOx/
CeO2 and Ce1-xCuxO2 catalysts. The results from all the samples
indicate that both metallic copper and oxygen vacancies in ceria
are involved in the generation of the catalytically active sites.
Evidence was found for a synergistic Cu-Ovacancy interaction.
This interaction enhances the chemical activity of Cu, and the

434 J. Phys. Chem. B, Vol. 110, No. 1, 2006
Wang et al.
(23) Matsumoto, T.; Nishiguchi, T.; Kanai, H.; Utani, K.; Matsumura,
Y.; Imamura, S. Appl. Catal., A 2004, 276, 267.
(24) Men, Y.; Gnaser, H.; Zapf, R.; Hessel, V.; Ziegler, C.; Kolb, G.
beamlines ×18B and ×19A and Zhong Zhong for his help in
the operation of beamline ×17B1 of the NSLS.
Appl. Catal., A 2004, 277, 83.
References and Notes
(25) (a) Wang, X. Q.; Hanson, J. C.; Liu, G.; Rodriguez, J. A.; Iglesias-
Juez, A.; Ferna´ndez-Garc´ıa, M. J. Chem. Phys. 2004, 121, 5434. (b) Wang,
X.; Hanson, J. C.; Rodriguez, J. A.; Belver, C.; Ferna´ndez-Garc´ıa, M. J.
Chem. Phys. 2005, 122, 154711.
(26) Wang, X.; Hanson, J. C.; Frenkel, A. I.; Kim, J.-Y.; Rodriguez, J.
A. J. Phys. Chem. B 2004, 108, 13667.
(1) Energy Policy Act (EPAct), 1992.
(2) (a) President’s Hydrogen Initiative, 2003. (b) Jacobson, M. Z.;
Colella, W. G.; Golden, D. M. Science 2005, 308, 1901.
(3) (a) Spivey, J. J. Catal. Today 2005, 100 (1-2), 171. (b) Tonkovich,
A. Y.; Zilka, J. L.; LaMont, M. J.; Wang, Y.; Wegeng, R. S. Chem. Eng.
Sci. 1999, 54, 2947.
(4) (a) Liu, Y.; Fu, Q.; Flytzani-Stephanopoulos, M. Catal. Today 2004,
93-95, 241. (b) Suh, D. J.; Kwak, C.; Kim, J. H.; Kwon, S. M.; Park, T.
J. J. Power Sources 2005, 142 (1-2), 70.
(5) Li, Y.; Fu, Q.; Flytzani-Stephanopoulos, M. Appl. Catal., B 2000,
27, 179.
(6) (a) Inoue, T.; Setoguchi, T.; Eguchi, K.; Arai, H. Solid State Ionics
1989, 34, 285. (b) Park, S.; Vohs; J. M.; Gorte, R. J. Nature 2000, 404,
265. (c) Gorte, R. J.; Vohs, J. M.; McIntosch, S. Solid State Ionics 2004,
175, 1.
(7) Lampe, U.; Gerblinger, J.; Meixner, H. Sens. Actuators 1992, B7,
787.
(8) (a) Kau, L.-S.; Solomon, D. J.; Penner-Haln, J. E.; Hodgson, K.
O.; Solomon, E. I. J. Am. Chem. Soc. 1987, 109, 6433. (b) Skårman, B.;
Grandjean, D.; Benfield, R. E.; Hinz, A.; Andersson, A.; Wallenberg, L.
R.; J. Catal. 2002, 211, 119, and references therein.
(9) (a) Mart´ınez-Arias, A.; Hungr´ıa, A. B.; Ferna´ndez-Garc´ıa, M.;
Conesa, J. C.; Munuera, G. J. Phys. Chem. B 2004, 108, 17983, and
references therein. (b) Mart´ınez-Arias, A.; Ferna´ndez-Garc´ıa, M.; Soria,
J.; Conesa, J. C. J. Catal. 1999, 182, 367, and references therein. (c)
Mart´ınez-Arias, A.; Ferna´ndez-Garc´ıa, M.; Ga´lvez, O.; Coronado, J. M.;
Anderson, J. A.; Conesa, J. C.; Soria, J.; Munuera, G. J. Catal. 2000, 195,
207.
(27) Kim, J. Y.; Rodriguez, J. A.; Hanson, J. C.; Frenkel, A. I.; Lee, P.
L. J. Am. Chem. Soc. 2003, 125, 10684.
(28) Rodriguez, J. A.; Hanson, J. C.; Kim, J.-Y.; Liu, G.; Iglesias-Juez,
A.; Ferna´ndez-Garc´ıa, M. J. Phys. Chem. B 2003, 107, 3535.
(29) Rodriguez, J. A.; Wang, X. Q.; Liu, G.; Hanson, J. C.; Hrbek, H.;
Peden, C. H. F.; Iglesias-Juez, A.; Ferna´ndez-Garc´ıa, M. J. Mol. Catal. A:
Chem. 2005, 228, 11.
(30) (a) Ferna´ndez-Garc´ıa, M.; Mart´ınez-Arias, A.; Guerrero-Ruiz, A.;
Conesa, J. C.; Soria, J. J. Catal. 2002, 211, 326. (b) Iglesias-Juez, A.;
Hungr´ıa, A. B.; Ga´lvez, O.; Mart´ınez-Arias, A.; Ferna´ndez-Garc´ıa, M.;
Conesa, J. C.; Soria, J. Stud. Surf. Sci. Catal. 2001, 138, 347. (c) Ferna´ndez-
Garc´ıa, M.; Mart´ınez-Arias, A.; Hungr´ıa, A. B.; Iglesias-Juez, A.; Conesa,
J. C.; Soria, J. Phys. Chem. Chem. Phys. 2002, 4, 2473. (d) Hungria, A. B.;
Mart´ınez-Arias, A.; Ferna´ndez-Garc´ıa, M.; Iglesias-Juez, A.; Guerrero-Ruiz,
A.; Calvino, J. J.; Conesa, J. C.; Soria, J. Chem. Mater. 2003, 15, 4309. (e)
Mart´ınez-Arias, A.; Ferna´ndez-Garc´ıa, M.; Ballesteros, V.; Salamanca, L.
N.; Conesa, J. C.; Otero, C.; Soria, J. Langmuir 1999, 15, 4796.
(31) Wang, X.; Rodriguez, J. A.; Hanson, J. C.; Gamarra, D.; Fernandez-
Garcia, M.; Martinez-Arias, A. J. Phys. Chem. B 2005, 109, 19595.
(32) Hammersley, A. P. FIT2D: An Introduction and OVerView; ESRF
Internal Report, ESRF97HA02T, 1997.
(33) Larson, A. C.; Von Dreele, R. B. GSAS General Structure Analysis
System; Report LAUR 86-748; Los Alamos National Laboratory: Los
Alamos, NM, 1995. (b) A. M. Reitveld, J. Appl. Crystallogr. 1969, 2, 65.
(c) Toby, B. H. J. Appl. Crystallogr. 2001, 34, 210.
(10) Harrison, P. G.; Ball, I. K.; Azelee, W.; Daniell, W.; Goldfarb, D.
Chem. Mater. 2000, 12, 3715.
(11) (a) Liu, W.; Flytzani-Stephanopoulos, M. J. Catal. 1995, 153, 304.
(b) Liu, W.; Flytzani-Stephanopoulos, M. J. Catal. 1995, 153, 317. (c) Qi.
X.; Flytzani-Stephanopoulos, M. Ind. Eng. Chem. Res. 2004, 43, 3055. (d)
Shan, W.; Shen, W.; Li, C. Chem. Mater. 2003, 15, 4761.
(12) (a) Bunluesin, T.; Gorte, R. J.; Graham, G. W. Appl. Catal., B 1998,
15, 107. (b) Koryabkina, N. A.; Phatak, A. A.; Ruettinger, W. F.; Farrauto,
R. J.; Ribeiro, F. H. J. Catal. 2003, 217, 233.
(13) (a) Shido, T.; Iwasawa, Y. J. Catal. 1993, 141, 71. (b) Jacobs, G.;
Chenu, E.; Patterson, P. M.; Williams, L.; Sparks, D.; Thomas, G.; Davis,
B. H. Appl. Catal., A 2004, 258, 203.
(14) Zhu, T.; Kundakovic, L.; Dreher, A.; Flytzani-Stephanopoulos, M.
Catal. Today 1999, 50, 381.
(15) Shaw, E. A.; Rayment, T.; Walker, A. P.; Lambert, R. M.; Gauntlett,
T.; Oldman, R. J.; Dent, A. Catal. Today 1991, 9, 197.
(16) (a) Hocˇevar, S.; Batista, J.; Levec, J. J. Catal. 1999, 184, 39. (b)
Hocˇevar, S.; Krasovec, U.O.; Orel, B.; Arico, A. S.; Kim, H. Appl. Catal.,
B 2000, 28, 113.
(17) Liu, Y.; Hayakama, T.; Suzuki, K.; Hamakawa, S. Catal. Commun.
2001, 2, 195.
(18) Men, Y.; Gnaser, H.; Zapf, R.; Hessel, V.; Ziegler, C.; Kolb, G.
Appl. Catal., A 2004, 277, 83.
(19) Papavasiliou, J.; Avgouropoulos, G.; Ioannides, T. Catal. Commun.
2004, 5, 231.
(20) (a) Bunluesin, T.; Gorte, R.; Graham, G. Appl. Catal., B 1998, 15,
107. (b) Hilaire, S.; Wang, X.; Luo, T.; Gorte, R. J.; Wagner, J. Appl. Catal.,
B 2001, 215, 271.
(34) (a) Newville, M.; Ravel, B.; Haskel, D.; Rehr, J. J.; Stern, E. A.;
Yacoby, Y. Physica B 1995, 208-209, 154. (b) Ravel, B.; Newville, M. J.
Synchrotron Radiat. 2005, 12 (4), 537.
(35) Sayers, D. E.; Bunker, B. A. In X-ray Absorption: Principles,
Applications, Techniques of EXAFS, SEXAFS, and XANES; Koningsberger,
D. C., Prins, R., Eds.; John Wiley & Sons: New York, 1988; p 211.
(36) Kim, C. H.; Thompson, L. T. J. Catal. 2005, 230, 66.
(37) Ferna´ndez-Garc´ıa, M. Catal. ReV.-Sci. Eng. 2002, 44, 59.
(38) Henderson, M. A.; Perkins, C. L.; Engelhard, M. H.; Thevuthasan,
S.; Peden, C. H. F. Surf. Sci. 2003, 526, 1.
(39) Padeste, C.; Cant, N. W.; Trimm, D. L. Catal. Lett. 1993, 18, 305.
(40) Otsuka, K.; Hatano, M.; Morikawa, A. J. Catal. 1983, 79, 493.
(41) Kundakovic, L.; Mullins, D. R.; Overbury, S. H. Surf. Sci. 2000,
457, 51.
(42) (a) Campbell, C. T.; Daube, K. J. Catal. 1987, 104, 109. (b)
Nakamura, J.; Campbell, J. M.; Campbell, C. T. J. Chem. Soc., Faraday
Trans. 1990, 86, 2725.
(43) Rodriguez, J. A.; Perez, M.; Evans, J.; Liu, G.; Hrbek, J. J. Chem.
Phys. 2005, 122, 241101.
(44) Zhao, X.; Hrbek, J.; Rodriguez, J. A.; Perez, M. To be published.
(45) Mart´ınez-Arias, A.; Hungr´ıa, A. B.; Ferna´ndez-Garc´ıa, M.; Conesa,
J. C.; Munuera, G. J. Power Sources 2005, 151, 32.
(46) Li, C.; Sakata, Y.; Arai, T.; Domen, K.; Maruya, K.-i.; Onishi, T.
J. Chem. Soc., Faraday Trans 1 1989, 85, 929.
(47) Li, C.; Sakata, Y.; Arai, T.; Domen, K.; Maruya, K.-i.; Onishi, T.
J. Chem. Soc., Faraday Trans 1 1989, 85, 1451.
(21) (a) Li, Y.; Fu, Q.; Flytzani-Stephanopoulos, M. Appl. Catal., B
2000, 27, 179. (b) Fu, Q.; Weber, A.; Flytzani-Stephanopoulos, M. Catal.
Lett. 2001, 77 (1-3), 87.
(22) (a) Tscho¨pe, A.; Schaadt, D.; Birringer, R.; Ying, J. Y. Nanostruct.
Mater. 1997, 9, 423. (b) Ovensen, C. V.; Stolze, P.; Nørskov, J. K.;
Campbell, C. T. J. Catal. 1992, 134, 445.
(48) (a) Bensalem, A.; Muller, J.-C.; Tessier, D.; Bozon-Verduraz, F.
J. Chem. Soc., Faraday Trans. 1996, 92, 3233. (b) Dai, D.; Li, L.; Ren, J.;
Whangbo, M.-H. J. Chem. Phys. 1998, 108, 3479.
(49) Meunier, F. C.; Tibiletti, D.; Goguet, A.; Reid, D.; Burch, R. Appl.
Catal.. A 2005, 289, 104.