Correlating catalytic methanol oxidation with the structure and oxidation state of size-selected pt nanoparticles

Article abstract of DOI:10.1021/cs400234h

We have investigated the structure and chemical state of size-selected platinum nanoparticles (NPs) prepared by micelle encapsulation and supported on γ-Al₂O₃ during the oxidation of methanol under oxygen-rich reaction conditions following both oxidative and reductive pretreatments. X-ray absorption near-edge structure (XANES) and extended X-ray absorption fine-structure (EXAFS) spectroscopy measurements reveal that in both cases, the catalyst is substantially oxidized under reaction conditions at room temperature and becomes partially reduced when the reactor temperature is raised to 50 C. Reactivity tests show that at low temperatures, the preoxidized catalyst, in which a larger degree of oxidation was observed, is more active than the prereduced catalyst. We conclude that the differences in reactivity can be linked to the formation and stabilization of distinct active oxide species during the pretreatment.

Full text of DOI:10.1021/cs400234h

Research Article
pubs.acs.org/acscatalysis
Correlating Catalytic Methanol Oxidation with the Structure and
Oxidation State of Size-Selected Pt Nanoparticles
Lindsay R. Merte,† Mahdi Ahmadi,† Farzad Behafarid,† Luis K. Ono,† Estephania Lira,† Jeronimo Matos,†
Long Li,‡ Judith C. Yang,‡ and Beatriz Roldan Cuenya*,†
†Department of Physics, University of Central Florida, Orlando, Florida 32816, United States
‡Department of Chemical and Petroleum Engineering, Department of Physics, University of Pittsburgh, Pittsburgh, Pennsylvania
15261, United States
S
* Supporting Information
ABSTRACT: We have investigated the structure and
chemical state of size-selected platinum nanoparticles (NPs)
prepared by micelle encapsulation and supported on γ-Al2O3
during the oxidation of methanol under oxygen-rich reaction
conditions following both oxidative and reductive pretreat-
ments. X-ray absorption near-edge structure (XANES) and
extended X-ray absorption fine-structure (EXAFS) spectros-
copy measurements reveal that in both cases, the catalyst is
substantially oxidized under reaction conditions at room
temperature and becomes partially reduced when the reactor
temperature is raised to 50 °C. Reactivity tests show that at low temperatures, the preoxidized catalyst, in which a larger degree of
oxidation was observed, is more active than the prereduced catalyst. We conclude that the differences in reactivity can be linked
to the formation and stabilization of distinct active oxide species during the pretreatment.
KEYWORDS: platinum, methanol oxidation, operando, XAFS, EXAFS, XANES, alumina, nanoparticle, size-selected
the oxidation reaction but, in fact, enhance the activity or
modify the selectivity.14−17 Although the topic remains a matter
of debate,18,19 operando studies of nanoparticle (NP) samples
on high-surface-area supports have supported the idea that in
some cases, surface oxides might enhance the activity in
oxidation reactions.20−25
1. INTRODUCTION
The partial and total catalytic oxidation of methanol are
important chemical processes, the former being employed in
the synthesis of industrially relevant compounds such as
formaldehyde1,2 and methyl formate,3,4 and the latter in the
elimination of this toxic chemicals, along with other volatile
organic compounds, from exhaust gases.5−7 While partial
oxidation reactions are carried out using coinage metals or
various metal oxides,8 total oxidation is typically achieved with
platinum or palladium, which are able to convert methanol,
higher alcohols, hydrocarbons, and carbon monoxide to CO2
with high efficiencies at relatively low temperatures. Although
occurring in a rather different environment, the electrocatalytic
oxidation of methanol, which proceeds over Pt-based catalysts,
is also relevant. The rate of this process is a limiting factor in
the performance of direct methanol fuel cells, which produce
electricity from liquid fuel without the need for reforming.9,10
Despite having been studied for decades, some controversy
exists regarding the mechanisms of oxidation reactions over
platinum, particularly regarding the formation and participation
of oxide phases in the reaction process. On one hand,
formation of surface oxides has long been believed to cause
deactivation of Pt catalysts, for example, in the oxidation of
alcohols to aldehydes and ketones.11−13 On the other hand,
chiefly in the context of CO and CH4 oxidation, studies on
single crystal samples and real catalysts have suggested that
oxides may form under reaction conditions that do not hinder
One of the primary limitations in the study of supported
catalytic materials under realistic reaction conditions is the
complexity of such materials and the resulting difficulty in
characterizing them in detail. An emerging, successful approach
to reducing the complexity of such materials, enabling more
detailed characterization without sacrificing the direct applic-
ability of the results to commercially relevant chemical
processes, is to use NP synthesis techniques that produce
catalysts with well-defined sizes and shapes.26 Taking this
approach, we have used reverse micelle NP synthesis to study
methanol oxidation over a Pt/γ-Al2O3 catalyst, with particular
attention to the evolution of the structure and oxidation state of
Pt under reaction conditions. On the basis of the knowledge of
the chemical state of the catalyst at work that can be achieved
via X-ray absorption fine-structure spectroscopy (XAFS), the
effect of oxidative and reductive pretreatments on the catalyst’s
activity and selectivity will be discussed.
Received: March 27, 2013
Revised: May 16, 2013
© XXXX American Chemical Society
1460
dx.doi.org/10.1021/cs400234h | ACS Catal. 2013, 3, 1460−1468

ACS Catalysis
Research Article
alcohol, passed through a coil wrapped around the reactor tube,
and counter-heated with a heating tape powered by a variable-
voltage AC transformer. Measurements were made in
increasing temperature steps, and the conversion was allowed
to reach a steady state at each temperature. For testing the
reactivity of the diluted catalyst, the temperature was controlled
with a tube furnace and a PID controller (Watlow) and ramped
from 30 to 150 °C and then back at a rate of 10 °C/hour while
monitoring the reactant and product concentrations continu-
ously.
XAFS spectra were acquired at the Pt-L3 edge in transmission
mode at beamline X18B at the National Synchrotron Light
Source (NSLS I), Brookhaven National Laboratory (BNL).
Spectra were recorded from 150 eV below the absorption edge
to 1230 eV above the edge (kmax = 18 Å−1). The reactor used
for the operando XAS measurements consisted of a Kapton
tube into which 25 mg of undiluted catalyst powder was loaded
and held in place by quartz wool plugs, forming a catalyst bed
∼10 mm in length. The same space velocity (25 sccm total
flow), gas concentrations, and pretreatment conditions were
used at BNL and UCF for the reactivity measurements. During
the XAFS measurements, reactants and products were
monitored via an online mass spectrometer (SRS RGA100).
Details of these measurements are given in Supporting
Information Figure S1.
2. EXPERIMENTAL METHODS
The Pt-loaded micellar precursor was synthesized by
dissolution of H2PtCl6 into a solution of polystyrene-b-
poly(2-vinylpyridine) [PS(27700)-b-P2VP(4300)] block co-
polymer (Polymer Source, Inc.) in toluene (50 mg/mL). This
results in reverse micelles with the Pt salt dissolved in the P2VP
core. The quantity of platinum salt added was chosen to give a
Pt/P2VP molar ratio of 0.2. The mixture was stirred for 2 days
to ensure complete dissolution and then filtered. The formation
of uniform micelles was checked by dip-coating a natively
oxidized Si wafer with the micellar solution and imaging the
wafer in an atomic force microscope (AFM). The solution was
then mixed with commercial γ-Al2O3 powder (Inframat
Advanced Materials, ∼150 m2/g) to yield a nominal Pt loading
of 1%. The toluene was evaporated under mechanical agitation
at 50 °C. The polymer was removed by calcination in 70% O2
(balanced with He) at 375 °C for 24 h. Complete removal of
the polymer was confirmed by checking the C-1s X-ray
photoelectron spectroscopy (XPS) signal, which showed only
small intensity attributable to adventitious carbon accumulated
during the brief exposure of the sample to air in the course of
its ex situ transfer to the XPS chamber.23,27
AFM measurements were conducted in tapping mode with a
Veeco Multimode microscope. XPS measurements were
conducted using a monochromatic Al Kα X-ray source
(1486.6 eV) and a Phoibos electron energy analyzer (SPECS,
GmbH). Binding energies were referenced to the Al-2p
photoemission peak of Al2O3 at 74.3 eV. Pt-4f photoemission
peaks were not detectable for this sample because of their
coincidence with the Al-2p peaks of the support, so the Pt-4d
region was used instead to characterize the Pt oxidation state.
High-angle annular dark field scanning transmission electron
microscopy (HAADF-STEM) measurements were conducted
after the EXAFS experiments using a JEOL JEM-2100F TEM/
STEM operated at 200 kV. The TEM samples were prepared
by making an ethanol suspension of the Pt/γ-Al2O3 powder and
placing a few drops of this liquid onto an ultrathin C-coated
TEM grid. The probe size of the STEM mode is about 0.2 nm.
The Pt NP diameters were determined by measuring the full
width at half-maximum of the HAADF intensity profile across
the individual Pt NPs.
Catalytic tests were performed at the University of Central
Florida (UCF) at atmospheric pressure in a vertically mounted
quartz tube continuous flow reactor with either 50 mg of pure
Pt/γ-Al2O3 catalyst or 1 mg Pt/γ-Al2O3 diluted with 200 mg α-
Al2O3, in both cases supported by a quartz wool plug. The
reactant mixture consisted of 4% O2 and 0.15% methanol by
volume, balanced with He. The methanol was introduced to the
mixture by passing He gas through a bubbler at ambient
temperature. Gas concentrations were varied using mass flow
controllers (MKS), and the total flow rate was set to 50 sccm.
Prior to catalytic testing, the catalyst was first reduced in H2
(50% in He) at 240 °C for 30 min. To test the influence of the
catalyst oxidation state on its catalytic activity, the reduction
step was followed by an oxidation step in O2 (70% in He) at
240 °C for 30 min. The methanol, carbon dioxide, and methyl
formate concentrations were measured with an online mass
spectrometer (Hiden HPR20) sampling the reactor outlet.
Conversion versus temperature measurements were carried out
in two different ways. For testing the reactivity of the pure
catalyst, the catalyst temperature was varied manually between
5 and 100 °C by cooling the quartz reactor cell with isopropyl
The Athena and Artemis programs28 were used for
normalization, background subtraction, and fitting of the
extended X-ray absorption fine-structure (EXAFS) spectra.
Theoretical scattering amplitudes and phase shifts used in the
fits were calculated with the FEFF8 ab initio code29,30 using fcc
Pt to simulate Pt−Pt scattering paths and Na2Pt(OH)6 to
simulate Pt−O scattering paths. To provide absolute energy
calibration, the EXAFS spectrum of a Pt foil placed after the
sample was measured simultaneously. Energies given in this
work are referenced to the first inflection point of the Pt foil
spectrum defined to be 11564 eV.
3. RESULTS
3.1. Sample Characterization. An AFM image of the Pt-
loaded PS-b-P2VP micelles used as the precursor to prepare the
γ-Al2O3-supported NP catalyst is shown in Figure 1a. The
strong quasi-hexagonal ordering and uniform interparticle
separation (∼27 nm) observed for the dip-coated sample
indicate the successful synthesis of monodisperse micelles,
consistent with previous reports utilizing this method.31−33
After impregnating the micelles onto γ-Al2O3 powder and
removing the polymer by heating in oxygen, the XPS spectrum,
showing a Pt 4d5/2 photoemission peak at 316.9 eV (Figure
1b), revealed that the sample contained oxidized platinum.34
STEM measurements of the sample after the operando XAFS
experiments revealed small Pt particles with an average
diameter of 0.7 nm, distributed with a standard deviation of
0.2 nm (Figures 1c,d). According to our previous work,27,35 no
sintering is expected after our in situ treatments during the
XAFS measurements which involved annealing treatments in
H2 and O2 up to only 240 °C and in a mix of methanol and O2
up to 50 °C.27,35
3.2. Catalytic Activity. Catalytic activity of the undiluted
Pt/γ-Al2O3 sample for methanol oxidation was tested at
temperatures between 5 and 100 °C following a reductive
and an oxidative pretreatment, respectively, and measured
conversions are plotted in Figure 2. Each experiment was
performed twice using untreated catalyst samples, and the
1461
dx.doi.org/10.1021/cs400234h | ACS Catal. 2013, 3, 1460−1468

ACS Catalysis
Research Article
Figure 1. (a) AFM image of the Pt-loaded PS-b-P2VP micellar
precursor used to prepare the Pt/γ-Al2O3 catalyst used in this study.
Micelles were dip-coated onto an SiO2/Si(111) wafer for imaging;
polymer ligands were not removed. (b) Pt-4d XPS spectrum of the as-
prepared (after calcination in oxygen) Pt/γ-Al2O3 catalyst. Binding
energies expected for Pt, PtO, and PtO2 are indicated. (c) HAADF-
STEM image of the Pt/γ-Al2O3 catalyst after ligand removal and after
operando XAFS measurements. (d) Particle size distribution extracted
from HAADF-STEM measurements.
Figure 2. Reactivity of the undiluted Pt/γ-Al2O3 catalyst for methanol
oxidation following (a) reductive (240 °C in 50% H2 for 30 min) and
(b) oxidative (240 °C in 70% O2 for 30 min.) pretreatments.
Conversion refers to the molar fraction of methanol in the feed
converted to a particular product. Plotted points are individual
measurements, and solid curves are guides to the eye. Carbon dioxide
(CO2) and methyl formate (CHOOCH3) were the only products
detected in significant quantities. All data shown correspond to steady-
state reaction conditions at each given temperature.
measurements are plotted together. The only products detected
were carbon dioxide (CO2) and methyl formate (CHOOCH3),
as expected from the following reaction pathways:
reduced catalyst: the former showed ∼45% conversion at 35
°C; the latter showed negligible conversion below that
temperature. At higher temperatures, the conversions of both
were similar, each exceeding 90% total conversion at ∼60 °C
and showing increased conversion to CO2 at the expense of
methyl formate with increasing temperature. Both the
preoxidized and prereduced catalysts showed high selectivity
to methyl formate at low temperature and complete conversion
to CO2 above ∼100 °C, although the transition point (where
conversions to CO2 and methyl formate are equal) occurs at
slightly higher temperature for the reduced catalyst (67 °C)
compared with the oxidized one (60 °C). Reactivity measure-
ments obtained during the subsequent downward temperature
sweep are shown in Figure 3c, with those from the oxidized
catalyst overlaid on those from the reduced. Differences in
reactivity in this case are almost negligible, indicating that the
effect of the pretreatment is lost following the initial
temperature ramp to 150 °C. The low-temperature reactivities
measured during the reverse temperature ramps are inter-
mediate between those obtained for the preoxidized and
prereduced catalysts during the initial upward temperature
ramp, each showing conversion of ∼20% at 35 °C.
2CH3OH + 3O2 → 2CO2 + 4H2O
(1)
2CH3OH + O2 → CHOOCH3 + 2H2O
(2)
At low temperatures, the total conversion over the
preoxidized catalyst was significantly higher than that observed
for the reduced catalyst. At 5 °C, the lowest temperature at
which reactivities were tested, the oxidized catalyst converted
∼80% of the methanol, wheras the reduced catalyst converted
only ∼40%. Product selectivities were also different for the two
pretreatments, even after accounting for the differences in
overall conversion. Over the prereduced catalyst at 80%
conversion (observed at about 22 °C), selectivity to methyl
formate was ∼40%, and over the oxidized catalyst at the same
total conversion (observed at 5 °C), the selectivity to methyl
formate was only ∼12%.
To obtain more accurate quantitative measurements of
methanol oxidation reaction rates above room temperature
(RT), we carried out similar reactivity measurements using a
diluted sample, consisting of 1 mg Pt/γ-Al2O3 mixed
thoroughly with 200 mg of α-Al2O3. In this experiment, the
temperature was ramped linearly from ∼30 to 150 °C and back
at a rate of 10 °C/min. Total conversions as well as CO2 and
methyl formate yields for the prereduced and preoxidized
catalysts during the upward sweep are shown in Figures 3a and
b, respectively. Measurements were constant above 120 °C and
were excluded from the plots. At low temperatures (<40 °C),
the oxidized catalyst was significantly more active than the
3.3. XANES and EXAFS. Operando XAFS measurements
were conducted under similar conditions as the catalytic tests
described above. The catalyst was first reduced in hydrogen at
240 °C and characterized at 25 °C in H2 in the reduced state.
After this, the sample was exposed to the reaction mixture
(MeOH and O2) and measured at 25 and 50 °C under steady-
1462
dx.doi.org/10.1021/cs400234h | ACS Catal. 2013, 3, 1460−1468

ACS Catalysis
Research Article
Figure 4. Pt-L3 XANES spectra of a Pt/γ-Al2O3 catalyst acquired
following oxidative and reductive pretreatments and under methanol
oxidation conditions after these pretreatments. (a) Comparison of
spectra acquired under reaction conditions at room temperature
beginning in the reduced state (treated in H2 at 240 °C) and an
oxidized state (treated in O2 at 240 °C), respectively. XANES spectra
acquired of the as-prepared (oxidized) catalyst and of a bulk Pt foil are
shown for reference. (b) Comparison of spectra acquired under
reaction conditions at different temperatures.
Figure 3. Reactivity of the diluted Pt/γ-Al2O3 catalyst for methanol
oxidation. Reactivity following (a) reductive and (b) oxidative
pretreatments, measured with increasing temperature. (c) Reactivity
following reductive and oxidative pretreatments, measured with
decreasing temperature following the measurements shown in a and b.
mixture at 25 °C led to a very slight decrease in WL intensity,
indicating only a very small change in the oxygen content of the
Pt NPs. However, the intensity of the WL of the prereduced
sample upon exposure to the reaction mixture at 25 °C was
slightly lower than that of the sample following oxidation at 240
°C.
Although the XANES spectra of the preoxidized and
prereduced catalysts under reaction conditions at 25 °C are
rather similar, the difference between them increased upon
heating to 50 °C, as shown in Figure 4b. In both cases, heating
to 50 °C led to a decrease in the WL intensity with no change
in peak position, but the difference in intensity between the 25
and 50 °C data was significantly larger for the prereduced
catalyst than for the preoxidized catalyst.
The EXAFS oscillations extracted from the postedge region
of the Pt-L3 absorption coefficient result from interference
effects due to the scattering of emitted photoelectrons by the
surrounding atoms and thus reflect the averaged local
coordination environment of the Pt atoms. Fourier-transformed
EXAFS spectra in Figure 5 exhibit characteristic peaks at
positions corresponding roughly (neglecting element- and
momentum-dependent phase shifts) to the nearest-neighbor
distances in the sample. The EXAFS spectra of the Pt catalyst in
the as-prepared, prereduced, and preoxidized initial states
(prereaction) as well as in the MeOH + O2 mix at 25 °C and at
50 °C are shown in Figure 5, together with the spectrum of a
bulk Pt foil. The spectrum of the prereduced catalyst shows
primarily a peak at ∼2.6 Å (phase uncorrected) corresponding
to Pt−Pt nearest-neighbor scattering, similar to that of the Pt
foil, but reduced in intensity because of the small size (and thus
reduced average coordination numbers) and enhanced static
state reaction conditions. After the measurements of the
catalyst following the reduction pretreatment, the same sample
was oxidized in O2 at 240 °C and again exposed to the reaction
mixture at 25 and at 50 °C.
Figure 4a shows XANES spectra obtained for the catalyst in
the reduced state, in the oxidized state, and in the MeOH + O2
reaction mixture, all at 25 °C, following the two different
pretreatments. Similar data from a Pt foil are also included for
reference. Pt-L3 absorption spectra exhibit a prominent peak,
referred to as a “white line” (WL), just above the edge, resulting
from electronic transitions from 2p3/2 to 5d states. The
intensity of the WL is determined primarily by the unoccupied
5d density of states and, thus, reflects the average oxidation
state of the platinum atoms. The as-prepared sample, measured
following a 24-h calcination in O2 at 375 °C, exhibits an intense
WL, most likely indicating that a substantial fraction of the
platinum in the catalyst is in the form of PtO2.36−38 Compared
with a metallic platinum foil, the WL of the reduced nanosized
catalyst exhibits a shift to higher energy as well as some
broadening due to a combination of NP size/shape effects and
changes in the electronic structure resulting from hydrogen
chemisorption.39,40 Exposure of the reduced catalyst to the
oxygen-rich reaction mixture led to an increase in the white line
intensity and decrease in the peak energy, indicating the
presence of chemisorbed oxygen on Pt or PtOx species.38 In
the case of the preoxidized sample, exposure to the reaction
1463
dx.doi.org/10.1021/cs400234h | ACS Catal. 2013, 3, 1460−1468

ACS Catalysis
Research Article
Figure 6. (a) Fourier-transformed (k = 2.5−15.5 Å−1) EXAFS
spectrum of the initially reduced catalyst measured at 25 °C in MeOH
+ O2. The fitted spectrum as well as the magnitudes of the three
contributing scattering components corresponding to metallic Pt, a
Pt−O(1) bond at ∼2.0 Å, and Pt−O(2) at ∼2.5 Å are also shown. (b)
Inverse-Fourier-transformed (R = 1.2−3.2 Å) EXAFS spectrum
corresponding to that in part a.
space (Fourier-transformed) and k-space. From these fits, the
following distances (phase-corrected) were obtained: Pt−Pt
(2.76 Å), Pt−O(1) (2.0 Å), and Pt−O(2) (2.5 Å). The
remaining fits and relevant fit information are included in
Supporting Information Figures S2−S9 and Table S1.
Figure 5. Fourier-transformed k2-weighted EXAFS spectra of the Pt/γ-
Al2O3 catalyst after reductionand oxidation and under MeOH
oxidation reaction conditions following the two pretreatments leading
to the oxidation and reduction of the catalysts. Shown also are first-
shell fits to the experimental data using a combination of three
scattering components, as described in the text. A Hanning window
from 2.5 to 13 Å−1 with Δk = 1 Å−1 was used to compute the Fourier
transforms. The data are vertically displaced for clarity.
Best fit parameters from the analysis of the EXAFS data are
given in Table 1, and the Pt−Pt, Pt−O(1) and Pt−O(2)
coordination numbers (CNs) are plotted in Figure 7. The fitted
Pt−Pt CN for the reduced catalyst of ∼7 corresponds to
particles of ∼1 nm in diameter,41,42 which agrees well with the
mean particle diameter d measured by STEM, provided that the
volume-averaging nature of the XAFS measurement is
accounted for by weighting the histogram by d3. Although
the particles measured by STEM had a mean diameter of 0.7
nm, the greater contribution of larger particles results in an
apparent diameter (volume-weighted) of 0.9 nm. The Pt−O(1)
scattering component of the EXAFS spectra, with a bond length
of 2.0 Å, is typical of strong platinum−oxygen bonds, and its
appearance upon exposure of the reduced catalyst to the
reaction mixture and after oxidation treatments, concomitant
with a decrease in the Pt−Pt coordination number, indicates
that we do not have chemisorbed oxygen on Pt, but the
formation of platinum oxides.
Consistent with the XANES measurements, more extensive
oxidation is observed following the oxygen pretreatment at 240
°C than during exposure of the reduced catalyst to methanol
and oxygen at 25 °C, and little change in the oxidation state is
observed after exposure of the oxidized catalyst to the reactant
mixture. Determination of the extent of oxidation and of the
stoichiometric ratio of the oxides formed is not straightforward.
Since both the XANES and EXAFS spectra give averaged
information for the entire sample, it is difficult to distinguish
between a mixture of phases and a single phase of intermediate
atomic disorder of the NPs. The as-prepared sample (with the
highest PtOx content due to its extensive oxidation at 375 °C
for 24 h), in contrast, shows a dominant peak at ∼1.7 Å (phase-
uncorrected) due to Pt−O nearest-neighbor scattering. The
spectrum of the preoxidized catalyst (O2 treatment at 240 °C
for 30 min) and those of the catalyst obtained under reaction
conditions exhibit features corresponding to both Pt−Pt and
oxidic Pt−O scattering [designated as Pt−O(1) in Figure 5] as
well as a third feature at an intermediate scattering distance
(∼2.1 Å phase uncorrected) designated as Pt−O(2).
To separate and quantify the overlapping components of the
EXAFS spectra, least-squares fitting was carried out using
theoretically calculated scattering amplitudes and phase shifts
generated using FEFF. The best fits obtained in this way are
plotted together with the experimental data in Figure 5. The
combination of a Pt−Pt scattering component and two Pt−O
scattering components was found to adequately reproduce the
experimental spectra, as suggested by the observations above.
An example of one of the fits, corresponding to the initially
reduced catalyst under methanol oxidation reaction conditions
at 25 °C, is shown in Figure 6, with the contributions of the
individual scattering components plotted separately in both r-
1464
dx.doi.org/10.1021/cs400234h | ACS Catal. 2013, 3, 1460−1468

ACS Catalysis
Research Article
Table 1. Pt L3 EXAFS Fit Parameters for the Pt/γ-Al2O3 Catalyst in the As-Prepared, Reduced and Oxidized States As Well As
a
under Methanol Combustion Conditions Following Reducing and Oxidizing Pretreatments
nPt−Pt
dPt−Pt (Å)
nPt−O(1)
dPt−O(1) (Å)
nPt−O(2)
dPt−O(2) (Å)
as-prepared
4.1 0.3
7.3 0.7
4.6 1.0
5.5 1.0
3.1 1.7
3.7 0.6
4.7 0.9
12.0 0.4
2.763 0.010
2.760 0.004
2.760 0.006
2.752 0.005
2.760 0.010
2.761 0.006
2.759 0.006
2.765 0.001
3.7 0.6
1.98 0.02
1.7 0.6
0.5 0.3
1.2 0.8
0.6 0.4
1.0 0.9
1.2 0.5
1.1 0.6
2.51 0.05
2.51 0.02
2.52 0.03
2.54 0.03
2.55 0.04
2.52 0.02
2.51 0.02
reduced H2, 25 °C
reduced MeOH + O2 25°C
reduced MeOH + O2 50°C
oxidized He, 25°C
oxidized MeOH + O2 25°C
oxidized MeOH + O2 50°C
Pt Foil
2.1 0.7
2.1 0.7
2.9 0.8
2.9 0.6
3.2 1.1
2.00 0.02
1.97 0.02
1.98 0.03
1.99 0.01
1.98 0.02
a
Fit parameters for a Pt foil are shown for reference. nPt−Pt, nPt−O(1), and nPt−O(2) are coordination numbers, and dPt−Pt, dPt−O(1), and dPt−O(2) are bond
lengths for the three distinct scattering species.
adsorption of high coverages of hydrogen at the NP/support
interface.47 However, we consistently observe this long-bond
component more prominently under oxidizing conditions than
in the exclusive presence of hydrogen. We interpret its
consistent presence under oxidizing conditions as a sign of
significant contact area between the NPs and the Al2O3 support
surface, possibly due to a modification of the NP shape
(flattening) under reaction conditions. The latter is in
agreement with data from Markusse et al.12 on C-supported
NPs, which suggested spherical to hemispherical (or flattened)
shape transformations when reducing environments were
replaced by oxidizing reaction conditions.
In agreement with the XANES measurements, which showed
a decrease in WL intensity upon raising the reactor temperature
to 50 °C, we observe an increase in the Pt−Pt coordination
number in EXAFS spectra of both the initially reduced and
initially oxidized catalysts, as would be expected upon either
partial reduction of Pt oxides to Pt metal or upon changing the
NP shape from a 2D to 3D morphology. Such a transformation
would also be expected to lead to a decrease in Pt−O(1) and
Pt−O(2) coordination numbers, which is not clearly observed,
although the latter changes may be within the uncertainty of
the measurements.
Figure 7. Coordination numbers of three different Pt species extracted
from first-shell analysis of EXAFS spectra. The sample was exposed to
two different pretreatments under reductive (left) and oxidative (right)
conditions. Spectra were acquired under the conditions indicated, in
order from left to right.
stoichiometry. The persistent presence of a metallic Pt−Pt
EXAFS component following oxidative treatments appears to
indicate only partial oxidation of the NPs, yielding a structure
with an oxide shell around a metallic core. Nevertheless, we
cannot rule out potential contributions from large grains of Pt
in the sample that could be only partially oxidized, whereas the
smaller NPs might be completely oxidized. Although not
detected in STEM measurements of this sample, even an
extremely small number of grains of only modest diameter
(10−20 nm) can potentially make a detectable contribution to
the XAFS spectra owing to their much greater volume.24 We
note, however, that due to the very low surface-to-volume ratio
of such large grains, their contribution to the catalytic activity
should in any case be negligible.
4. DISCUSSION
Our experiments reveal that Pt NP/γ-Al2O3 catalysts exhibit a
complex behavior in the oxidation of methanol that is
substantially different from that of bulk Pt. Over pure Pt wire
catalysts, McCabe and McCready48 found that methanol
oxidation exhibited reaction kinetics similar to CO oxidation,
which is characterized by CO-saturated surfaces and low
activity at low temperatures, versus oxygen-saturated surfaces
and high activity at high temperature. It was therefore proposed
that methanol oxidation over Pt proceeds through a strongly
bound intermediate, such as CO, which prevents adsorption of
O2 and thus limits the rate at low temperature. This is similar to
what has been observed for electrochemical methanol oxidation
over Pt/C catalysts, where poisoning of the Pt anode surfaces
with CO has been directly observed and found to cause large
overpotentials in the reaction.9 However, such poisoning does
not seem to occur in our experiments, since according to
Safonova et al.,49 a significant shift and broadening to higher
energy of the Pt-L3 white line is expected upon adsorption of
CO, which was not observed in our data. In our experiments,
XANES measurements indicate that under reaction conditions
at 25 °C, the platinum catalyst is in an oxygen-rich state
qualitatively similar to that of the preoxidized sample, regardless
of the pretreatment.
The Pt−O(2) scattering component, modeled as a
platinum−oxygen bond with a length of ∼2.5 Å, could not
be quantified with great accuracy because of its strong overlap
with the Pt−Pt and Pt−O(1) components, but exclusion of this
component led to significantly poorer fits to the EXAFS
spectra. The need to include a long Pt−O component in
EXAFS spectra of noble metals on various supports has been
noted in several previous studies.43−46 Koningsberger and
Gates observed such a component in spectra of various
catalysts, and assigned it to bonds between platinum atoms and
oxide ions of the Al2O3 support, with the unusual length
attributed to the presence of hydrogen at the interface.12,43
A
theoretical study investigating hydrogenated platinum clusters
on γ-Al2O3 appears to support this, showing also drastic NP
shape changes and breakage of Pt-substrate bonds upon
1465
dx.doi.org/10.1021/cs400234h | ACS Catal. 2013, 3, 1460−1468

ACS Catalysis
Research Article
In addition, the observation of methyl formate as the main
partial oxidation product is consistent with a different
mechanism for MeOH oxidation over highly dispersed Pt
NPs/Al2O3, as compared with pure Pt.50,51 In particular, over
pure Pt catalysts, the main partial oxidation product was
formaldehyde and was formed in only small yields.48,52 For Pd/
Al2O3 catalysts, methyl formate has been observed to be the
only product formed over PdO surfaces.53 Methyl formate
production could occur through different reaction paths
involving surface formate species (HCHO).53 Our high
selectivity toward CO2 at moderate temperatures suggests
that the oxidation of formate species is faster than the
condensations of absorbed methoxide with HCHO, dimeriza-
tion of HCHO, or esterification of formic acid (HCOOH)
intermediates.53
Furthermore, the oxidation state effect seems to be complex,
and system-dependent. For example, in methanol oxidation
reactions over high surface area palladium catalysts (Pd/Al2O3),
the reverse trend to what we observed was reported, with a
higher catalytic activity53,54 (up to 40 times) and a lower
selectivity toward methyl formate53 being detected for the
reduced Pd catalysts as compared with PdO. Moreover, the
methanol oxidation turnover frequency was shown to be
significantly lower for smaller Pd NPs53 because of stronger
binding of chemisorbed oxygen, resulting in higher O coverages
and the blockage of active sites.55 A similar size-dependent
binding was also reported for the O−Pt system,56−58 but no
poisoning due to chemisorbed oxygen was observed in our
study, since both the initially reduced and initially oxidized
samples were found to become/remain oxidized in the presence
of the reactants (evident from the decrease in Pt−Pt
coordination number and increase in Pt−O), and the sample
with the highest content of oxygen (preoxidized) was also
found to be the most active.
Nevertheless, although XANES and EXAFS measurements
indicate that in both cases the samples are oxidized under
reaction conditions, our experiments show clearly that the
initial oxidation state of the Pt NPs has a strong effect on the
catalysts’ low-temperature reactivity and that a greater degree of
oxidation leads to larger reaction rates. The results of the
EXAFS analysis displayed in Figure 7 reveal a higher degree of
oxidation (larger contribution of the Pt−O(1) bond) for the
preoxidized NPs under reactant exposure at 25 and 50 °C. The
latter finding is in accord with previous results by Markusse et
al.,12 revealing more extensive catalyst oxidation under aqueous
alcohol oxidation reaction conditions after an oxidative catalyst
start-up versus a reductive start-up.
It should be mentioned that in our case, the preoxidized
catalyst cannot be considered a true oxygen reservoir, since the
amount of oxygen stored in the form of Pt oxide is insignificant
as compared with the oxygen consumed during the reaction.
Therefore, if any constituents of the catalysts’ lattice (in this
case oxygen) leave the NP surface as part of the reaction
products, they should be subsequently replenished by the
oxygen reactant in the gas stream following a Mars−van-
Krevelen (M−v-K) process.59 On the other hand, a Langmuir−
Hinshelwood (L−H) mechanism in which both of the reactants
adsorb on the NP surface and then react has also been
observed, mainly on reduced metal surfaces,60,61 but also on
some metal oxide surfaces (e.g., RuO2).62,63 Furthermore, in
numerous previous studies, the drastic differences in reactivity
observed for oxidized versus reduced catalyst have been
attributed to the onset of one of the former distinct reaction
mechanisms. For instance, in CO oxidation reactions over Pd
catalysts, it has been shown that the Langmuir−Hinshelwood
mechanism takes place over reduced Pd surfaces, but with a
lower activity as compared with oxidized surfaces following
Mars−van-Krevelen processes.60,64 Similar behavior has also
been reported for Pt, showing higher activity for the oxidized
catalyst during CO oxidation.14,15,65 Nevertheless, the NP/
support interface and oxygen or hydroxyl exchange between the
metal NP phase and the oxide support might also need to be
considered.66,67
In our case, the difference in reactivity between the
preoxidized and prereduced catalysts cannot be attributed to
a distinct reaction mechanism taking place over reduced versus
oxidized Pt surfaces (e.g., M−v-K vs L−H), since our operando
spectroscopic characterization (XANES and EXAFS) revealed
that both catalysts are oxidized under reaction conditions. In
contrast, our operando data indicate that the catalytic activity of
the NPs is influenced not only by the presence of oxide
overlayers but also by their precise structure, chemical
composition, and possibly also thickness, which likely differ
due to the difference in temperature during the initial oxygen
exposure (240 °C in the case of the preoxidized catalyst, 25 °C
in the case of the prereduced catalyst). This is in line with
previous studies showing that the catalytic activities of different
PtOx phases vary substantially. For example, theoretical
calculations by Seriani et al.68 showed that although PtO and
PtO2 layers on Pt surfaces are inert toward CO and CH4
oxidation, a Pt3O4 surface can be highly active for these
reactions. Other studies have shown that Pt oxides with higher
oxidation states such as PtO2 or Pt(OH)3 could act as
poisonous species for primary alcohol oxidation (e.g.,
methanol), and lower oxidation states such as PtO and
Pt(OH) were identified as active species.69 Unfortunately,
EXAFS cannot be used to distinguish the different PtOx phases,
since they are all characterized by a nearly identical Pt−O
distance of ∼2 Å.
The differences in the XANES peak intensity and position
observed between the as-prepared NPs (PtO2-rich) and those
measured under reaction conditions (Figure 4a) can serve as
guidance to distinguish different chemical states of Pt in PtOx
compounds. In particular, the XANES spectra of our catalyst
under reaction conditions and after the oxidative pretreatment
exhibit white lines which are significantly less intense than that
of the as-prepared sample, measured following a 24-h
calcination at 375 °C. The drastic difference in intensity,
despite substantial Pt oxide formation under reaction
conditions at 25 °C and in oxygen at 240 °C, is an indication
that the oxides formed under these conditions consist primarily
of Pt in a 2+ oxidation state, whereas the prolonged higher-
temperature treatment (375 °C) likely produces Pt in the 4+
oxidation state. Such self-limiting oxidation of Pt NPs at low
temperatures (and also at low potentials under electrochemical
conditions38) has been observed in a number of studies to date.
Considering that the thermodynamically favored, well-ordered
PtO2 phase has been found to be catalytically inert for a
number of reactions, the existence of partially oxidized species
(e.g., PtO) appears to be an important element in the high
oxidation activities observed for certain catalytic reactions.20
Furthermore, XANES data acquired at 50 °C on the
prereduced sample (Figure 4b) revealed a significant decrease
in the WL intensity, not observed on the preoxidized sample,
which might be assigned either to a more facile reduction of
PtOx species or to the loss of chemisorbed oxygen species.
1466
dx.doi.org/10.1021/cs400234h | ACS Catal. 2013, 3, 1460−1468

ACS Catalysis
Research Article
Since no significant changes are observed in the Pt−Pt and Pt−
O(1) EXAFS components upon heating the sample to 50 °C, a
removal of chemisorbed species (which should have a weaker
overall effect on the EXAFS coordination numbers) appears
more likely. Nevertheless, it should be noted that the measured
catalytic activity of such sample is still lower than that of the
preoxidized counterpart.
Recent theoretical and experimental studies of methanol
oxidation over Pd catalysts assigned the most active Pd species
to either metallic Pd or a thick and well-ordered PdO structure,
with catalysts containing only one layer of oxide having higher
reaction barriers and lower reactivity.70 According to our
operando spectroscopic data, it is plausible that a better-
ordered oxide layer (likely PtO) might be present in the
preoxidized sample, and a thinner surface oxide layer might
exist under reactant exposure on the reduced sample. This
might be responsible for the initially higher activity observed for
our oxidized sample as compared with the reduced sample.
In summary, our study highlights the importance of the
specific chemical structure (Pt coordination environment and
degree of disorder) and thickness of the oxidic species formed
on NPs after different sample pretreatments as well as directly
under operando reaction conditions for oxidative catalytic
reactions.
support. This work has been made possible thanks to the
financial support of the Office of Basic Energy Sciences of the
U.S. Department of Energy under Grants DE-FG02-
08ER15995 (B.R.C.) and DE-FG02-03ER15476 (J.C.Y.).
Support to beamline X18B at NSLS-BNL, where the XAFS
experiments were conducted, was provided by the DOE’s
Synchrotron Catalysis Consortium (DE-FG02-05ER15688)
and DOE-BES (DE-AC02-98CH10866). NFCF at University
of Pittsburgh is acknowledged for the use of JEOL JEM-2100F.
REFERENCES
■
(1) Lefferts, L.; van Ommen, J. G.; Ross, J. R. H. Appl. Catal. 1986,
23, 385.
(2) Nagy, A.; Mestl, G. Appl. Catal., A 1999, 188, 337.
(3) Guerreiro, E. D.; Gorriz, O. F.; Larsen, G.; Arrua, L. A. Appl.
́
Catal., A 2000, 204, 33.
(4) Wittstock, A.; Zielasek, V.; Biener, J.; Friend, C. M.; Baumer, M.
̈
Science 2010, 327, 319.
(5) Spivey, J. J. Ind. Eng. Chem. Res. 1987, 26, 2165.
(6) Sharma, R. K.; Zhou, B.; Tong, S.; Chuane, K. T.; Tg, A. Ind. Eng.
Chem. Res. 1995, 34, 4310.
(7) Gandhi, H. S.; Graham, G. W.; McCabe, R. W. J. Catal. 2003,
216, 433.
(8) Tatibouet, J. M. Appl. Catal., A 1997, 148, 213.
̈
(9) Iwasita, T. Electrochim. Acta 2002, 47, 3663.
(10) Lamy, C.; Lima, A.; LeRhun, V.; Delime, F.; Coutanceau, C.;
CONCLUSIONS
́
Leger, J.-M. J. Power Sources 2002, 105, 283.
■
(11) Mallat, T.; Baiker, A. Chem. Rev. 2004, 104, 3037.
(12) Markusse, A. P.; Kuster, B. F. M.; Koningsberger, D.; Marin, G.
B. Catal. Lett. 1998, 55, 141.
We have investigated the chemical state of micelle-synthesized
Pt NPs supported on γ-Al2O3 under methanol oxidation
reaction conditions. Reactivity measurements following oxida-
tive and reductive pretreatments show clear differences at low
temperature, with an enhanced reactivity of the oxygen
pretreated NPs. The most highly active NPs are shown by
XANES and EXAFS measurements to consist predominantly of
platinum oxides. Similar measurements of the prereduced
catalyst show that metallic NPs are also partially oxidized upon
exposure to the reactants at room temperature, but with a
different structure from those formed in the oxidative
pretreatment at 240 °C, which makes them less active and
less stable. Our study supports the notion that platinum oxides
enhance the activity of oxidation catalysts at low temperature
but reveals some subtleties regarding the precise nature of the
oxides formed that are key for the understanding of the
reactivity trends observed. In particular, our work illustrates the
importance of operando studies to gain fundamental insight
into structure, chemical state, and reactivity correlations of
nanoscale catalysts.
(13) Nicoletti, J. W.; Whitesides, G. M. J. Phys. Chem. 1989, 93, 759.
(14) Ackermann, M. D.; Pedersen, T. M.; Hendriksen, B. L. M.;
Robach, O.; Bobaru, S. C.; Popa, I.; Quiros, C.; Kim, H.; Hammer, B.;
Ferrer, S.; Frenken, J. W. M. Phys. Rev. Lett. 2005, 95, 255505.
(15) Hendriksen, B. L. M.; Frenken, J. W. M. Phys. Rev. Lett. 2002,
89, 2.
(16) Li, W.-X. J. Phys.: Condens. Matter 2008, 20, 184022.
(17) Mallens, E. P. J.; Hoebink, J. H. B. J.; Marin, G. B. Catal. Lett.
1995, 33, 291.
(18) Gao, F.; Wang, Y.; Cai, Y.; Goodman, D. W. J. Phys. Chem. C
2009, 113, 174.
(19) McClure, S. M.; Goodman, D. W. Chem. Phys. Lett. 2009, 469,
1.
(20) Alayon, E. M. C.; Singh, J.; Nachtegaal, M.; Harfouche, M.; van
Bokhoven, J. A. J. Catal. 2009, 263, 228.
(21) Singh, J.; van Bokhoven, J. A. Catal. Today 2010, 155, 199.
(22) Croy, J. R.; Mostafa, S.; Heinrich, H.; Roldan Cuenya, B. Catal.
Lett. 2009, 131, 21.
(23) Mostafa, S.; Behafarid, F.; Croy, J. R.; Ono, L. K.; Li, L.; Yang, J.
C.; Frenkel, A. I.; Roldan Cuenya, B. J. Am. Chem. Soc. 2010, 132,
15714.
(24) Paredis, K.; Ono, L. K.; Mostafa, S.; Li, L.; Zhang, Z.; Yang, J.
C.; Barrio, L.; Frenkel, A. I.; Roldan Cuenya, B. J. Am. Chem. Soc.
2011, 133, 6728.
(25) Singh, J.; Nachtegaal, M.; Alayon, E. M. C.; Stotzel, J.; van
Bokhoven, J. A. ChemCatChem 2010, 2, 653.
ASSOCIATED CONTENT
* Supporting Information
Additional information as noted in text. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
S
(26) Roldan Cuenya, B. Thin Solid Films 2010, 518, 3127.
(27) Matos, J.; Ono, L. K.; Behafarid, F.; Croy, J. R.; Mostafa, S.;
DeLaRiva, A. T.; Datye, A. K.; Frenkel, A. I.; Roldan Cuenya, B. Phys.
Chem. Chem. Phys. 2012, 14, 11457.
AUTHOR INFORMATION
Corresponding Author
*E-mail: roldan@ucf.edu.
■
(28) Ravel, B.; Newville, M. J. Synchrotron Radiat. 2005, 12, 537.
(29) Rehr, J. J.; Albers, R. C. Rev. Mod. Phys. 2000, 72, 621.
(30) Ankudinov, A. L.; Bouldin, C. E.; Rehr, J. J.; Sims, J.; Hung, H.
Phys. Rev. B 2002, 65, 104107.
(31) Croy, J. R.; Mostafa, S.; Liu, J.; Sohn, Y.-h.; Roldan Cuenya, B.
Catal. Lett. 2007, 118, 1.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors acknowledge Anatoly I. Frenkel (Yeshiva
University) for assistance with the evaluation of the XAFS
data and Nebojsa Marinkovic (BNL) for his excellent beamline
(32) Kastle, G.; Boyen, H.-G.; Weigl, F.; Lengl, G.; Herzog, T.;
̈
Ziemann, P.; Riethmuller, S.; Mayer, O.; Hartmann, C.; Spatz, J. P.;
̈
1467
dx.doi.org/10.1021/cs400234h | ACS Catal. 2013, 3, 1460−1468

ACS Catalysis
Research Article
Moller, M.; Ozawa, M.; Banhart, F.; Garnier, M. G.; Oelhafen, P. Adv.
Funct. Mater. 2003, 13, 853.
(66) Schwartz, W. R.; Pfefferle, L. D. J. Phys. Chem. C 2012, 116,
8571.
̈
(67) Muller, C. A.; Maciejewski, M.; Koeppel, R. A.; Tschan, R.;
(33) Roldan Cuenya, B.; Baeck, S.-H.; Jaramillo, T. F.; McFarland, E.
W. J. Am. Chem. Soc. 2003, 125, 12928.
(34) Damyanova, S.; Bueno, J. M. C. Appl. Catal., A 2003, 253, 135.
(35) Roldan Cuenya, B.; Croy, J. R.; Mostafa, S.; Behafarid, F.; Li, L.;
Zhang, Z.; Yang, J. C.; Wang, Q.; Frenkel, A. I. J. Am. Chem. Soc. 2010,
132, 8747.
(36) Mansour, A. N.; Sayers, D. E.; Cook, J. W.; Short, D. R.;
Shannon, R. D.; Katzer, J. R. J. Phys. Chem. 1984, 88, 1778.
(37) Friebel, D.; Miller, D. J.; O’Grady, C. P.; Anniyev, T.; Bargar, J.;
Bergmann, U.; Ogasawara, H.; Wikfeldt, K. T.; Pettersson, L. G. M.;
Nilsson, A. Phys. Chem. Chem. Phys. 2011, 13, 262.
(38) Merte, L. R.; Behafarid, F.; Miller, D. J.; Friebel, D.; Cho, S.;
Mbuga, F.; Sokaras, D.; Alonso-Mori, R.; Weng, T.-C.; Nordlund, D.;
Nilsson, A.; Roldan Cuenya, B. ACS Catal. 2012, 2, 2371.
(39) Behafarid, F.; Ono, L. K.; Mostafa, S.; Croy, J. R.; Shafai, G.;
Hong, S.; Rahman, T. S.; Bare, S.; Roldan Cuenya, B. Phys. Chem.
Chem. Phys. 2012, 14, 11766.
̈
Baiker, A. J. Phys. Chem. 1996, 100, 20006.
(68) Seriani, N.; Pompe, W.; Ciacchi, L. C. J. Phys. Chem. B 2006,
110, 14860.
(69) Li, N.-H.; Sun, S.-G.; Chen, S.-P. J. Electroanal. Chem. 1997, 430,
57.
(70) Hellman, A.; Resta, A.; Martin, N. M.; Gustafson, J.; Trinchero,
A.; Carlsson, P. A.; Balmes, O.; Felici, R.; van Rijn, R.; Frenken, J. W.
M.; Andersen, J. N.; Lundgren, E.; Gronbeck, H. J. Phys. Chem. Lett.
̈
2012, 3, 678.
(40) Lei, Y.; Jelic, J.; Nitsche, L. C.; Meyer, R.; Miller, J. Top. Catal.
2011, 54, 334.
(41) Frenkel, A. I.; Yevick, A.; Cooper, C.; Vasic, R. Annu. Rev. Anal.
Chem. 2011, 4, 23.
(42) Jentys, A. Phys. Chem. Chem. Phys. 1999, 1, 4059.
(43) Koningsberger, D. C.; Gates, B. C. Catal. Lett. 1992, 14, 271.
(44) Paredis, K.; Ono, L. K.; Behafarid, F.; Zhang, Z.; Yang, J. C.;
Frenkel, A. I.; Roldan Cuenya, B. J. Am. Chem. Soc. 2011, 133, 13455.
(45) Vaarkamp, M.; Miller, J. T.; Modica, F. S.; Koningsberger, D. C.
J. Catal. 1996, 163, 294.
(46) Zhang, Y.; Toebes, M. L.; van der Eerden, A.; O’Grady, W. E.;
de Jong, K. P.; Koningsberger, D. C. J. Phys. Chem. B 2004, 108,
18509.
(47) Mager-Maury, C.; Bonnard, G.; Chizallet, C.; Sautet, P.;
Raybaud, P. ChemCatChem 2012, 3, 200.
(48) McCabe, R. W.; McCready, D. F. J. Phys. Chem. 1986, 90, 1428.
(49) Safonova, O. V.; Tromp, M.; van Bokhoven, J. A.; de Groot, F.
M. F.; Evans, J.; Glatzel, P. J. Phys. Chem. B 2006, 110, 16162.
(50) Chantaravitoon, P.; Chavadej, S.; Schwank, J. Chem. Eng. J 2004,
97, 161.
(51) McCabe, R. W.; Mitchell, P. J. Appl. Catal. 1986, 27, 83.
(52) Gentry, J.; Jones, A.; Walsh, T.; Sheffield, S. J. Chem. Soc.,
Faraday Trans. I 1980, 76, 2084.
(53) Lichtenberger, J.; Lee, D.; Iglesia, E. Phys. Chem. Chem. Phys.
2007, 9, 4902.
(54) Brewer, T. F.; Abraham, M. A.; Silver, R. G. Ind. Eng. Chem. Res.
1994, 33, 526.
(55) Chou, P.; Vannice, M. A. J. Catal. 1987, 105, 342.
(56) Xu, Y.; Shelton, W. A.; Schneider, W. F. J. Phys. Chem. B 2006,
110, 16591.
(57) Xu, Y.; Shelton, W. A.; Schneider, W. F. J. Phys. Chem. A 2006,
110, 5839.
(58) Ono, L. K.; Croy, J. R.; Heinrich, H.; Roldan Cuenya, B. J. Phys.
Chem. C 2011, 115, 16856.
(59) Mars, P.; van Krevelen, D. W. Chem. Eng. Sci. 1954, 3, 41.
(60) Hendriksen, B. L. M.; Ackermann, M. D.; van Rijn, R.; Stoltz,
D.; Popa, I.; Balmes, O.; Resta, A.; Wermeille, D.; Felici, R.; Ferrer, S.;
Frenken, J. W. M. Nat. Chem. 2010, 2, 730.
(61) Endo, M.; Matsumoto, T.; Kubota, J.; Domen, K.; Hirose, C.
Surf. Sci. 1999, 441, L931.
(62) Gong, X.-Q.; Liu, Z.-P.; Raval, R.; Hu, P. J. Am. Chem. Soc. 2004,
126, 8.
(63) Reuter, K.; Frenkel, D.; Scheffler, M. Phys. Rev. Lett. 2004, 93,
249701.
(64) Lashina, E. A.; Slavinskaya, E. M.; Chumakova, N. A.; Stonkus,
O. A.; Gulyaev, R. V.; Stadnichenko, A. I.; Chumakov, G. A.; Boronin,
A. I.; Demidenko, G. V. Chem. Eng. Sci. 2012, 83, 149.
(65) Jensen, R.; Andersen, T.; Nierhoff, A.; Pedersen, T.; Hansen, O.;
Dahl, S.; Chorkendorff, I. Phys. Chem. Chem. Phys. 2013, 15, 2698.
1468
dx.doi.org/10.1021/cs400234h | ACS Catal. 2013, 3, 1460−1468