Chemical state study of palladium powder and ceria-supported palladium during low-temperature CO oxidation

Article abstract of DOI:10.1021/jp060435u

The chemical state of Pd at the surfaces of two sizes of Pd powders and ceria-supported Pd during low-temperature CO oxidation has been studied using X-ray photoelectron spectroscopy (XPS). During oxidation in O₂, metallic Pd powder is converted into PdO, and the thickness of the PdO layer increases with increasing reaction temperature. A similar Pd oxidation process occurs during the catalytic CO oxidation reaction, but the extent of the Pd oxidation is less due to the presence of CO which is a reducing agent. The reaction rate data indicate that the larger size Pd powder is about three times more active than the smaller size Pd powder on a surface area basis at 160 °C. Catalytic CO oxidation data obtained from 10 wt % Pd supported on nanocrystalline ceria powder indicate that there is a strong chemical interaction between the ceria and the supported Pd. The Pd is present as PdO on the fresh catalyst. During reaction, small amounts of Pd metal and PdO ₂: are formed at 50 °C-while less Pd metal and only a small amount of PdO₂: are present after running the reaction at 110 °C. At 50 °C, the catalytic activity decays rapidly due to accumulation of carbonate or bicarbonate species on the surface. This decay does not occur at 110 °C due to decomposition of the bicarbonate species.

Full text of DOI:10.1021/jp060435u

J. Phys. Chem. A 2006, 110, 7609-7613
7609
Chemical State Study of Palladium Powder and Ceria-Supported Palladium during
Low-Temperature CO Oxidation
Seung-Hoon Oh and Gar B. Hoflund*
Department of Chemical Engineering, UniVersity of Florida, GainesVille, Florida 32611
ReceiVed: January 20, 2006; In Final Form: April 12, 2006
The chemical state of Pd at the surfaces of two sizes of Pd powders and ceria-supported Pd during low-
temperature CO oxidation has been studied using X-ray photoelectron spectroscopy (XPS). During oxidation
in O2, metallic Pd powder is converted into PdO, and the thickness of the PdO layer increases with increasing
reaction temperature. A similar Pd oxidation process occurs during the catalytic CO oxidation reaction, but
the extent of the Pd oxidation is less due to the presence of CO which is a reducing agent. The reaction rate
data indicate that the larger size Pd powder is about three times more active than the smaller size Pd powder
on a surface area basis at 160 °C. Catalytic CO oxidation data obtained from 10 wt % Pd supported on
nanocrystalline ceria powder indicate that there is a strong chemical interaction between the ceria and the
supported Pd. The Pd is present as PdO on the fresh catalyst. During reaction, small amounts of Pd metal and
PdO2 are formed at 50 °C while less Pd metal and only a small amount of PdO2 are present after running the
reaction at 110 °C. At 50 °C, the catalytic activity decays rapidly due to accumulation of carbonate or
bicarbonate species on the surface. This decay does not occur at 110 °C due to decomposition of the bicarbonate
species.
Introduction
for Pd supported on different metal oxides (SiO2, TiO2, and
Al2O3) vary considerably and attribute this to varying densities
The development of low-temperature carbon monoxide
oxidation catalysts has been an active research area for many
years due to the many important environmental and technolog-
ical applications including closed-cycle CO2 pulsed lasers, CO
gas detection sensors, air-purification, gas masks, and polymer
electrolyte fuel cells.1-4 Many different types of catalysts have
been tested for low-temperature CO oxidation. Most of the
catalysts that exhibit high activity consist of one or more noble
metals dispersed on a reducible oxide support (NMRO cata-
lysts).1 There are many variables involved in the preparation,
pretreatment, and reaction which all significantly affect the
catalytic activity and decay behavior of the different catalysts.
Supported platinum5 and gold6-8 catalysts were originally
proposed for low-temperature CO oxidation. Since then, other
noble metals including palladium, rhodium, and ruthenium have
received attention for CO oxidation.4 A review of platinum-
based and gold-containing NMRO catalysts was published by
Hoflund and Epling,9 and a review of gold-containing catalysts
was published by Haruta and Date.10
Palladium-based catalysts which contain Pt and Rh are widely
used for automotive three-way catalysts which can simulta-
neously remove nitric oxides, carbon monoxide, and hydrocar-
bons. These catalysts have better durability for the treatment
of automotive exhaust than platinum or rhodium. Automotive
catalysts which contain only Pd have been proposed recently
as an alternative for the industry. Most Pd-containing catalysts
used in CO oxidation studies and surface characterization studies
were supported on metal oxides such as Al2O3, SiO2, TiO2,
CoOx, CeO2, or zeolites. The influence of these different metal
oxides on low-temperature CO oxidation has been reported by
Pavlova and co-workers.11 They found that the CO activities
of surface defect centers which can stabilize CO and O2.
Hoffmann et al.12 have studied Pd size effects for CO oxidation
over Al2O3-supported Pd. They found that two different Pd
particle sizes (1.8 and 5.5 nm) exhibit different transient and
steady-state kinetics using molecular beam experiments.
Three catalysts which perform very well for low-temperature
CO oxidation are iron-oxide-supported Au,6 platinized tin
oxide,5,9 and manganese-oxide-supported gold.13,14 Au/MnOx is
an excellent catalyst when the CO2 concentration is low15,16
while Fe-stabilized Pt/SnOx performs well in a CO2-rich
environment.17 Au supported on titania has drawn attention
recently from a fundamental viewpoint,18-20 but this catalyst is
of little technological interest due to its high activity decay rate.21
In a more recent study by Oh and Hoflund,22 Pd supported on
MnOx, CeO2, TiO2, CoOx, and Al2O3 have been examined for
CO oxidation activity near stoichiometric conditions. Of these
catalysts, 10 wt % Pd/CeO2 performed the best. In this present
study, the chemical state of Pd has been determined using X-ray
photoelectron spectroscopy (XPS) for two sizes of Pd powders
and 10 wt % Pd supported on nanocrystalline ceria. Apparently,
CO oxidation has not been studied over Pd metal powder. This
eliminates the influence of the Pd-oxide support interaction so
that the chemistry of this simpler system can be examined.
Hopefully, this study will form a basis to understand the more
complex catalytic systems for CO oxidation.
Experimental Section
Several types of catalysts were examined in this study. Two
different sizes of Pd metal powder (99.95% purity on a metal
basis) were purchased from Alfa Aesar. One had a size range
from 0.25 to 0.55 µm with a BET surface area of 10.6 m2/g,
and the other had a size range from 1 to 1.5 µm with a BET
* To whom correspondence should be addressed. Fax: 352-392-9513.
E-mail: garho@hotmail.com.
10.1021/jp060435u CCC: $33.50 © 2006 American Chemical Society
Published on Web 06/01/2006

7610 J. Phys. Chem. A, Vol. 110, No. 24, 2006
Oh and Hoflund
surface area of 1.08 m2/g. An impregnated 10 wt % Pd/CeO2
catalyst (BET surface area of 43.2 m2/g) was prepared using
the incipient wetness method in which palladium nitrate (Pd-
(NO3)‚H2O) was added to nanocrystalline ceria (CeO2, Nano-
Phase, Inc., BET surface area of 50.7 m2/g) using purified water.
The catalyst was dried for 5 h in an oven at 100 °C and calcined
in air at 280 °C for 3 h. This temperature is high enough to
decompose the Pd precursor as evidenced by the absence of a
N 1s peak in XPS spectra obtained from the fresh catalyst. The
catalyst was granulated into fine particles using a mortar and
pestle. A 10 wt % Pd-CeO2 mixture was prepared by physically
combining Pd metal (0.25 to 0.55 µm) and nanocrystalline ceria
in a 1:9 weight ratio of Pd to ceria. This physical mixture has
a BET surface area of 46.7 m2/g.
The reaction studies were performed in a tubular reactor
system operating at atmospheric pressure. A specified mass of
each catalyst was supported by glass wool in a quarter-inch
diameter quartz U-shaped tube. The tube was then inserted in
a Thermolyne 21100 furnace. A K-type thermocouple was
placed around the U-tube at the center of the catalyst bed and
connected to an auto-tuning PID controller. The reaction
temperature was maintained accurately inside the furnace within
(1 °C. Ultrahigh purity test gases obtained from Praxair were
mixed to yield a composition about 1% CO and 0.5% O2 in
helium. The feed stream was slightly oxygen rich by less than
150 ppm due to small fluctuations of flow rates. The total flow
rate was 45 cm3/min. Initially, the gas stream was bypassed
and analyzed in a Hewlett-Packard 5890A gas chromatograph
to determine the feed concentration. For product analysis, a 5A
molecular sieve column was connected to a thermal conductivity
detector (TCD). Their temperatures were kept essentially
constant at 80 and 150 °C, respectively. The reaction gas stream
and bypass gas line were connected to a Valco 6 port valve
with a 1 cm3 sample loop. The conversion of CO was calculated
using the peak area differences between the unreacted bypass
and the effluent from the catalyst bed. During reaction, care
was taken to allow the temperature to equilibrate before a sample
of the outlet gas was analyzed. Multiple samples were taken
and averaged to ensure that the catalytic system had reached
steady state.
Surface characterization studies were carried out on the
catalysts before and after oxidation or reaction at various
temperatures. The catalyst powders were pressed into aluminum
cups, inserted into the ultrahigh vacuum chamber and character-
ized using XPS. XPS was performed using a double-pass
cylindrical mirror analyzer (DPCMA, PHI Model 25-255AR)
and pulse counting detection23 by operating the DPCMA in the
retarding mode using a pass energy of 25 eV for the high-
resolution spectra and 50 eV for the survey spectra. The X-rays
were generated using a Mg KR X-ray source. Model experi-
ments were performed to ensure that the brief exposure to air
during the transfer from the catalytic reactor to the characteriza-
tion chamber does not alter the Pd chemical state.
Figure 1. CO conversion as a function of temperature for two different
size Pd metal powders. The feed stream was 1.0% CO and 0.5% O2 in
He at a total flow rate of 45 cm3/min.
Figure 2. High-resolution XPS Pd 3d spectra obtained from (a) fresh
Pd metal powder and Pd metal powder after exposure to 0.5% O2 for
60 min in the reactor at (b) 50, (c) 200, and (d) 300 °C. All Pd powders
were prereduced with 3.5% H2 in He at 100 °C for 60 min.
oxidation activity below 150 °C. The smaller size Pd powder
is more active above 150 °C and reaches complete conversion
at 180 °C while the larger size Pd powder oxidizes only 15%
of the CO at this temperature and does not reach complete
conversion until about 250 °C. The complete conversion of CO
corresponds to oxidation of 1.87 × 10-5 gmole/min of CO. The
average areal reaction rates in gmole/min‚m2 have been
calculated from the data in Figure 1 at 160 °C. On an areal
basis, the large Pd powder (1.08 m2/g) is 3.11 times more active
than the small Pd powder (10.6 m2/g) with average areal reaction
rates of 0.53 × 10-6 and 0.17 × 10-6 gmole/min‚m2, respec-
tively. This reaction is exothermic so 160 °C was chosen to
minimize heat effects. The reason for the difference in the areal-
based reaction rates is not known. It may be due to a size effect
or more likely to surface morphological differences. The oxygen
conversions over the whole temperature range (not shown) are
essentially identical to the carbon monoxide conversions.
The large Pd powder was then reduced in H2 as described
above to produce metallic Pd and then exposed to 0.5% O2 in
He for 1 h at 50, 200, and 300 °C. High-resolution XPS Pd 3d
spectra obtained after these treatments are shown in Figure 2,
and the corresponding O 1s and Pd 3p spectra are shown in
Results and Discussion
Pd Powders. The catalytic activities of the two Pd metal
powders for CO oxidation as a function of temperature are
shown in Figure 1. The Pd powders were initially given a
reductive pretreatment in 3.5% H2 at 100 °C for 1 h to ensure
that the Pd was metallic. This state is referred to as the fresh
catalyst for the Pd metal powders. As shown below, the Pd metal
powder develops Pd oxide at the surface after either oxidative
treatments or exposure to catalytic reaction conditions. Both
the 0.25-0.55 µm and the 1-1.5 µm powders exhibit poor CO

Study of Pd Metal during Catalytic CO Oxidation
J. Phys. Chem. A, Vol. 110, No. 24, 2006 7611
Figure 3. High-resolution XPS O 1s and Pd 3p spectra obtained from
(a) fresh Pd metal powder and Pd metal powder after exposure to 0.5%
O2 for 60 min in the reactor at (b) 50, (c) 200, and (d) 300 °C. All Pd
powders were prereduced with 3.5% H2 in He at 100 °C for 60 min.
Figure 4. High-resolution XPS Pd 3d spectra obtained from (a) fresh
Pd metal powder and Pd metal powder after exposure to 1% CO and
0.5% O2 for 60 min in the catalytic reactor at (b) 50, (c) 190, (d) 240,
and (e) 300 °C. The Pd powder was reduced with 3.5% H2 in He at
100 °C for 60 min.
Figure 3. The Pd 3d and Pd 3p features are characteristic of
clean Pd with no Pd oxide present. After the O2 treatment at 50
°C (Figure 2b), all of the features are altered. The Pd 3d5/2 and
3d3/2 peaks are shifted to higher binding energies (BEs) by about
0.3 eV, and they are broadened on the high-BE side. The Pd
3p3/2 peak is broadened to both the low-BE and the high-BE
sides. The broadening on the low-BE side is due to the presence
of an O 1s feature from PdO, and the broadening on the high-
BE side is due to the formation of a PdO Pd 3p feature. These
assertions are consistent with the broadening on the high-BE
sides of the Pd 3d features and the shift toward PdO features in
Figure 2. The changes in these features occur to a greater extent
after the O2 exposure at 200 °C (Figure 2c). The shift in the Pd
3d peaks are still small after the 200 °C-O2 treatment so the
corresponding changes in the O 1s and Pd 3p features shown
in Figure 3c are also small. Nevertheless, a more well-defined
O 1s shoulder is apparent after this treatment. This shoulder is
particularly apparent in the same set of spectra which has not
been normalized (not shown). After the O2 exposure at 300 °C
(Figure 2d), the spectral changes are quite large due to nearly
complete conversion of Pd metal to PdO in the near-surface
region. In Figure 2d, the predominant features are PdO 3d peaks
and the small shoulders on the low-BE sides are small Pd metal
features. Two predominant features are now present in Figure
3d due to PdO Pd 3p and PdO O 1s peaks. These data indicate
that an oxide layer forms on the Pd particles and that the
thickness of this oxide layer depends on the temperature.
Oxidation of Pd occurs at lower temperatures, but the fact that
the oxide layers remain thin indicates that diffusion of oxygen
to the underlying Pd metal is the slow step in the oxidation
process.
Figure 5. High-resolution XPS O 1s and Pd 3p spectra obtained from
(a) fresh Pd metal powder and Pd metal powder after exposure to 1%
CO and 0.5% O2 for 60 min in the catalytic reactor at (b) 50, (c) 190,
(d) 240, and (e) 300 °C. The Pd powder was reduced with 3.5% H2 in
He at 100 °C for 60 min.
after the 300 °C-O2 treatment in Figure 2d and those obtained
after running the catalytic reaction at 300 °C in Figure 4e. The
O2 treatment results in much more complete oxidation of the
Pd compared with the catalytic reaction. This is due to the fact
the O2 treatment is purely oxidative while the catalytic reaction
is both oxidative and reductive due to the presence of both O2
and CO, respectively. A significant portion of O2 which would
go toward oxidizing Pd metal to PdO is instead reacted with
CO to produce CO2 resulting in a slower rate of Pd oxidation.
Electron energy loss spectroscopy (ELS) has proven to be a
useful technique for distinguishing between Pd metal and PdO.24
Advantages of ELS over XPS are that the primary beam can
be focused to a small size, depth sensitivity can easily be
XPS data were also obtained after running the catalytic
reaction over the large Pd powder at 50, 190, 240, and 300 °C.
Before reaction, the catalysts were reduced in 3.5% H2 in He
for 1 h at 100 °C to ensure the presence of Pd metal. High-
resolution Pd 3d spectra and Pd 3p/O 1s spectra are shown in
Figures 4 and 5, respectively. Similar to the results for the O2
treatment, Pd metal is oxidized during the catalytic reaction,
but it is not oxidized to the extent as during the oxidation. This
is most readily apparent by comparing the Pd 3d peaks obtained

7612 J. Phys. Chem. A, Vol. 110, No. 24, 2006
Oh and Hoflund
Figure 7. Conversion decay curves obtained from 10 wt % Pd/n-
Figure 6. CO conversion as a function of temperature for (a) 10 wt
% Pd/n-CeO2 (impregnated), (b) a physical mixture of 0.010 g of Pd
metal powder with sizes ranging from 0.25 to 0.55 µm and 0.090 g of
nanocrystalline ceria, and (c) nanocrystalline ceria. The feed stream
was 1.0% CO and 0.5% O2 in He at a total flow rate of 45 cm3/min.
CeO2 (impregnated) while operating the catalytic reactor at 50 and 110
°C.
changed by varying the primary beam energy, and chemical
state differences are often more apparent in ELS data. In a
related study, ELS has been used to study H2-reduced Pd powder
after use as a catalyst for methane oxidation.25 The results
demonstrate that a shell of PdO develops on the Pd particles
and that metallic Pd lies only beneath this shell. This implies
that methane oxidation occurs on PdO. The fact that no Pd metal
is present in the PdO layer indicates that an oxidative-reductive
cycle (Pd0 T Pd2+) is not responsible for low-temperature (<300
°C) methane oxidation. A proposed mechanism is that a methane
C-H bond is broken, the C bonds to an O in PdO to form a
methoxy species, and the H bonds to an O in PdO to form a
hydroxyl species. The methoxy species then loses the other H
atoms to form hydroxyl groups which combine and desorb as
water molecules. The remaining C combines with 2O from PdO
to form CO2, which desorbs, and oxygen vacancies in the PdO
layer. The last step is for these vacancies to adsorb O2 to form
PdO.
These are several important similarities between catalytic
methane oxidation and catalytic CO oxidation including (1) O2
is a reactant in both cases, (2) CO2 is a product in both cases,
and (3) both reactions occur over many of the same catalysts.
The differences are that (1) the CtO bond is not broken while
all the C-H bonds are broken and (2) water is a product in
methane oxidation. Since both reactants are usually run in an
oxygen-rich environment, the presence of PdO on Pd powder
formed during CO oxidation most likely is a shell over Pd metal
as has been shown to be the case for methane oxidation.
Ceria-Supported Pd. CO conversion vs temperature curves
are shown in Figure 6 for nanocrystalline ceria, a physical
mixture of nanocrystalline ceria and Pd metal, and nanocrys-
talline ceria impregnated with Pd (10 wt %). Nanocrystalline
ceria exhibits the lowest catalytic activity, but it does exhibit
significant activity above 200 °C. Bare oxides generally are not
very catalytically active at low temperatures, and their activities
can be enhanced by the addition of metal powder or supported
metal. When Pd metal powder is physically mixed with
nanocrystalline ceria (0.010 g of Pd metal powder with size
ranging from 0.25 to 0.55 µm and 0.090 g of nanocrystalline
ceria), the catalytic activity is increased considerably achieving
100% conversion below 250 °C. The third curve in Figure 6
indicates that 10 wt % impregnated Pd/n-CeO2 is much more
active than either the bare ceria or the physical mixture of Pd
powder and ceria. It exhibits activity at room temperature and
reaches 100% conversion below 100 °C, which is before the
Figure 8. High-resolution XPS Pd 3d spectra obtained from (a) the
fresh 10 wt % Pd/n-CeO2, (b) the 10 wt % Pd/n-CeO2 after 90 min
reaction at 50 °C, and (c) after 90 min reaction at 110 °C. The feed
stream was 1.0% CO and 0.5% O2 in He at a total flow rate of 45
cm3/min.
other two catalysts exhibit any activity. The data shown in Figure
6 indicate that there is a chemical interaction between ceria and
the supported Pd which has a very large effect on the catalytic
properties. This increased catalytic activity may be due to a Pd
size effect or a chemical interaction between the ceria and Pd
or a combination of these effects.
Conversion decay curves are shown in Figure 7 for the
impregnated 10 wt % Pd/n-CeO2 catalyst (170 mg) while
operating the reaction at 50 and 110 °C. At 50 °C, the activity
decays rapidly from 100 to about 40% conversion after 23 min
and then stabilizes near this value. Previous studies9 have shown
that this decay in activity is due to accumulation of bicarbonate
species on the surface. Elimination of these species by heating
or subjecting the catalyst to vacuum or an inert gas flow results
in restoration of its high catalytic activity. However, at 110 °C,
a high conversion (100%) is maintained over the period tested.
This fact indicates that at 110 °C the bicarbonate species
decompose to yield CO2 thereby maintaining a catalytically
active surface.
XPS data have been obtained from the 10 wt % Pd/n-CeO2
catalyst before and after running the CO oxidation reaction at

Study of Pd Metal during Catalytic CO Oxidation
J. Phys. Chem. A, Vol. 110, No. 24, 2006 7613
50 and 110 °C. High-resolution XPS Pd 3d spectra are shown
in Figure 8. The Pd 3d feature obtained from the fresh catalyst
(Figure 8a) is narrow and well-defined with a BE characteristic
of PdO. No other species including metallic Pd or PdO2 are
observable. The spectrum obtained after running the catalytic
reaction at 50 °C (Figure 8b) is much broader due to a shoulder
on the low-BE side due to metallic Pd and a shoulder on the
high-BE side due to PdO2. A distinct O 1s peak (not shown)
due to bicarbonate is present. This is consistent with the decay
in activity shown in Figure 7. The Pd 3d spectrum obtained
from the catalyst after running the reaction at 110 °C (Figure
8c) indicates that less Pd metal is present. The peak due to
bicarbonate is not present in the O 1s spectrum (not shown)
which is consistent with the fact that the activity does not decay
at 110 °C. Only a very small amount of PdO2 is present. Most
of the Pd is present as PdO which is believed to be the active
catalytic species. The reactant gas mixture contains O2 which
is an oxidizing agent and CO which is a reducing agent. The
XPS data indicate that a small amount of PdO is oxidized to
PdO2 and that a small amount of PdO is reduced to metallic
Pd.
Reaction studies indicate that bare ceria exhibits catalytic
activity for CO oxidation above 200 °C. If impregnated with
10 wt % Pd, it becomes highly active between 0 and 100 °C
due to a chemical interaction between ceria and Pd. There is
no decay in activity at 110 °C while there is significant decay
at 50 °C due to accumulation of carbonate or bicarbonate species
on the surface. XPS data indicate that PdO is the only Pd species
on the surface of the fresh catalyst. During reaction at 50 °C,
both PdO2 and Pd metal form. Carbonate or bicarbonate species
also accumulate on the surface resulting in a rapid decay in
activity. At 110 °C, the activity does not decay and the carbonate
or bicarbonate species does not accumulate on the surface. A
smaller amount of Pd metal is present on the surface as is a
much smaller amount of PdO2.
References and Notes
(1) Schryer, D. R., Hoflund, G. B., Eds. Low-Temperature CO oxidation
Catalysts for Long-Life CO2 Lasers; NASA Conference Publication 3076;
NASA: Washington, DC, 1990.
(2) Srinivason, B.; Gardner, S. D. Surf. Interface Anal. 1998, 26, 1035.
(3) T-Sanchez, R. M.; Udea, A.; Tanaka, K.; Haruta, M. J. Catal. 1997,
168, 125.
(4) Oh, S.; Sinkevitch, M. J. Catal. 1993, 142, 254.
(5) Stark, D. S.; Crocker, A.; Stewart, G. J. J. Phys. E: Sci. Instrum.
1983, 16, 158.
Summary
(6) Haruta, M.; Yamada, N.; Kobayashi, T.; Iijima, S. J. Catal. 1989,
115, 301.
(7) Haruta, M.; Tsubota, S.; Kobayashi, T.; Kageyama, H.; Genet, M.;
Delmon, B. J. Catal. 1993, 144, 175.
(8) Gardner, S. D.; Hoflund, G. B.; Upchurch, B. T.; Schryer, D. R.;
Kielin, E. J.; Schryer, J. J. Catal. 1991, 129, 114.
(9) Hoflund, G. B.; Epling, W. S. Recent Res. DeV. Catal. 1996, 1,
31.
(10) Haruta, M.; Date, M. Appl. Catal., A 2001, 222, 427.
(11) Pavlova, S. N.; Sadykov, V. A.; Bulgakov, N. N.; Bredikhin, M.
N. J. Catal. 1996, 161, 517.
In this study, the catalytic behavior for low-temperature CO
oxidation of two Pd metal powders and ceria-supported Pd have
been examined as a function of temperature. XPS has been
performed on these catalysts to determine the Pd chemical state
before and after running catalytic CO oxidation at different
temperatures. For comparison, oxidation experiments also have
been carried out on the larger Pd powder. Model studies have
been performed to ensure that the brief exposure to air during
transfer from the catalytic reactor to the UHV characterization
experiment do not alter the Pd chemical state.
(12) Hoffmann, J.; Meusel, I.; Hartmann, J.; Libuda, J.; Freund, H.-J.
J. Catal. 2001, 204, 378.
In each case, the Pd powder was reduced in hydrogen to form
metallic Pd as the starting point. The oxidation studies were
carried out by exposing the Pd powder to O2 in He at varying
temperatures. This treatment forms a layer of PdO on the Pd
particles, and the thickness of this layer depends on the reactor
temperature. At 300 °C, the PdO layer is so thick that the
underlying Pd metal signal is quite small.
Catalytic CO oxidation was carried out over the small and
large Pd powders. The areal activity of the large Pd powder is
about three times higher than that of the small Pd powder at
160 °C. Similar chemical state changes occur during reaction
as in the oxidation case. Metallic Pd is converted to a layer of
PdO but at a slower rate than for oxidation because CO is a
reducing agent. A proposed mechanism for catalytic CO
oxidation is that CO adsorbs to form carbonate or bicarbonate
species which decompose to form CO2. Then oxygen chemisorbs
at the vacancies. In this mechanism, the active surface is a PdO
surface which may have adsorbed carbonate or bicarbonate
species and oxygen vacancies but no Pd metal.
(13) Gardner, S. D.; Hoflund, G. B.; Schryer, D. R.; Schryer, J.;
Upchurch, B. T.; Kielin, E. J. Langmuir 1991, 7, 2135.
(14) Gardner, S. D.; Hoflund, G. B.; Davidson, M. R.; Laitinen, H. A.;
Schryer, D. R.; Upchurch, B. T. Langmuir 1991, 7, 2140.
(15) Hoflund, G. B.; Gardner, S. D.; Schryer, D. R.; Upchurch, B. T.;
Kielin, E. J. Langmuir 1995, 11, 3431.
(16) Hoflund, G. B.; Gardner, S. D.; Schryer, D. R.; Upchurch, B. T.;
Kielin, E. J. React. Kinet. Catal. Lett. 1996, 58, 19.
(17) Hoflund, G. B.; Upchurch, B. T.; Kielin, E. J.; Schryer, D. R. Catal.
Lett. 1995, 31, 133.
(18) Soares, J. M. C.; Morrall, P.; Crossley, A.; Harris, P.; Bowder, M.
J. Catal. 2005, 219, 17.
(19) Park, E. D.; Lee, J. S. J. Catal. 1999, 186, 1.
(20) Maeda, Y.; Fujitani, T.; Tsubota, S.; Haruta, M. Surf. Sci. 2004,
562, 1.
(21) Konova, P.; Naydenov, A.; Venkov, Cv.; Mehandjiev, D.; Andreeva,
D.; Tabakova, T. J. Mol. Catal. A: Chem. 2004, 213, 235.
(22) Oh, S.-H.; Hoflund, G. B. To be submitted for publication.
(23) Gilbert, R. E.; Cox, D. F.; Hoflund, G. B. ReV. Sci. Instrum. 1982,
53, 1281.
(24) Hagelin, H. A. E.; Weaver, J. F.; Hoflund, G. B.; Salaita, G. N. J.
Electron Spectrosc. Relat. Phenom. 2002, 124, 1.
(25) Hoflund, G. B.; Hagelin, H. A. E.; Weaver, J. F.; Salaita, G. N.
Appl. Surf. Sci. 2003, 205, 102.