M. Hatanaka et al. / Journal of Catalysis 266 (2009) 182–190
183
video images clearly captured the decrease in Pt particle size with
time under an oxidative atmosphere at 1093 K [31].
Samples of 6 g of the oxidized catalysts were heated from room
temperature to 873 K under a flow of 3% H /N (1000 ml/min) over
a period of 1 h and maintained under these conditions for 30 min,
then cooled to room temperature under the H /N flow to maintain
2
2
From these results it can be expected that oxidized Pt will exist
as a monolayer on the surface of CeO
2
-based oxide supports after
2
2
treatment in an oxidative atmosphere at high temperature. How-
ever, complete evidence of this phenomenon is not yet available.
In the present study, we focused on the oxidation state and nano-
the oxidation state of Pt. The catalysts treated under these reduc-
tive conditions are referred to as reduced catalysts.
Samples of 3 g of the reduced catalysts were further subjected
to the same treatment as for the oxidized catalysts to yield re-oxi-
dized catalysts.
2
structure of Pt on the surface of a CeO -based oxide support using
XPS and TEM. XPS is more surface-sensitive than XAFS, and ad-
vanced high-resolution TEM can capture direct evidence of the Pt
nanostructure that forms Pt–O–Ce bonds. We investigated whether
reversible changes in the Pt oxidation state and nanostructure oc-
2.3. Catalyst characterization
cur in sequentially oxidized, reduced, and re-oxidized Pt on a CeO
based support at high temperature.
2
-
Pt in the oxidized, reduced, and re-oxidized catalysts was char-
acterized using the following methods.
2.3.1. Pt particle diameter
2
. Experimental
The Pt particle diameter was quantified using CO pulse adsorp-
tion and XRD methods. CO adsorption measurements were per-
formed using an OHKURA RIKEN prototype low-temperature CO
adsorption apparatus [27]. A sample of 0.02 g of the catalyst was
placed in a sample cell and pretreated according to the procedure
established by the Catalysis Society of Japan. First, the catalyst was
heated to 673 K in a flow of pure O2 (20 ml/min) and maintained
under the same conditions for 15 min, then the gas flow was chan-
ged to 20 ml/min pure He for 10 min, followed by 20 ml/min pure
H for 15 min to convert PtO into metallic Pt. Finally, the gas flow
2
.1. Support and catalyst preparation
2 2
In previous studies [27,30,31], 55 wt.% CeO containing CeO –
2 2 3 2
ZrO –Y O (CZY) was used as the CeO -based support; however,
the composition of primary particles of this material varies some-
what. Use of such heterogeneous material as a support for XPS Pt
oxidation analysis would yield complicated results because the Pt
oxidation states on Ce-rich and Zr-rich domains must be different.
Thus, a support of simple composition is desirable to avoid ambi-
2
x
was changed back to 20 ml/min pure He for 10 min. The catalyst
was then cooled to 195 K under the pure He flow using a dry ice/
ethanol refrigerant to prevent CO adsorption on the CA support.
guity due to local heterogeneity in the support, and pure CeO
ideal for this purpose. However, in principle, primary particles of
pure CeO treated at high temperature agglomerate to larger parti-
cles because pure CeO is not thermally stable under these condi-
tions. Therefore, CeO –Al was used as the support. Under an
oxidative atmosphere, Al does not react with CeO in the solid
phase below 1273 K and is located at the grain boundaries of CeO
particles. Thus, Al acts as a diffusion barrier that restricts the
contact point of primary CeO particles and inhibits CeO particle
growth [24]. The support consisted of 89 wt.% CeO and 11 wt.%
Al , and hereafter the CeO –Al support is denoted as CA.
The CA support was prepared using the following precipitation
method. Certain amounts of Ce(NO O and Al(NO ꢀ9H
ꢀ6H
Wako Pure Chemicals) were dissolved in ion-exchanged water
2
is
2
Then, 0.75 lmol/pulse CO was injected into the catalyst under
2
the He flow until the CO content in the outlet gas reached a con-
stant value. The amount of CO adsorbed was calculated as the dif-
ference between the total amount injected and the sum of all CO
detected in the outlet gas. The Pt particle diameter was estimated
assuming that one CO molecule adsorbed onto each Pt atom on the
surface of hemispherical Pt particles [27,32].
XRD measurements of the catalysts after the CO adsorptions
were performed on a Bruker AXS D8 Advanced powder X-ray dif-
fractometer using Cu Ka radiation (k = 0.1541 nm) operated at
35 kV and 40 mA with a super-speed VANTEC-1 detector. Data
were recorded in the 2h range from 35° to 45° with an angular step
size of 0.014° and a counting time of 1 s/step. The Pt particle diam-
eter (D nm) was calculated from the width at half-height (B) of the
Pt(111) diffraction line using the Scherrer equation D = Kk/Bcosh
(K = 0.9).
2
2 3
O
2
O
3
2
2
2 3
O
2
2
2
2
O
3
2
2 3
O
3
)
3
2
3
)
3
2
O
(
and ammonia solution was added. The precipitated material was
dried at 673 K for 5 h, calcined at 873 K for 5 h, and finally fired
at 1073 K for 5 h in air in an ADVANTEC KM-420 muffle furnace.
For comparison, a pure Al
method, except no Ce(NO
face area of the CA and pure Al
2
O
3
support was prepared using a similar
3
)
3
ꢀ6H O was included. The specific sur-
2
2
2
O
3
supports was 55 and 101 m /
2.3.2. Analysis of the Pt oxidation state
g, respectively, as determined by the BET one-point method using
a Microdata Microsorp 4232II instrument.
XPS analysis of the oxidation state of supported Pt was carried
out in two independent experiments. In the first experiment, the
Pt oxidation state was measured for pulverized samples of the oxi-
dized, reduced, and re-oxidized Pt/CA and Pt/Al O catalysts with-
Pt/CA and Pt/Al
2
O
3
catalysts were prepared by conventional wet
(NO nitric
impregnation of the support powders with Pt(NH
3
)
2
2 2
)
2
3
acid solution (Tanaka Kikinzoku Kogyo K.K.). The Pt loading on
the catalysts was controlled at 1.5 wt.%. The Pt-impregnated pow-
ders were dried at 393 K overnight and calcined at 773 K for 2 h in
a muffle furnace. Then the powders were pressed into disks,
crushed, and sieved to yield particles of 0.5–1.0 mm in diameter.
These samples are referred to as fresh catalysts.
out further treatment. An ULVAC-PHI PHI-Quantera SXM XPS
system was used for this analysis with monochromatic Al Ka X-
rays (1486.6 eV) at ambient temperature [33]. The electron escape
depth and diameter of the analysis spot were approximately 2 nm
and 100 lm, respectively. In the second experiment, changes in the
Pt oxidation state were observed for the pre-reduced Pt/CA catalyst
after sequential in-situ re-oxidizing treatment at 673, 873, and
1073 K for 10 min under 20% O /N with an O2 pressure of
2.2. Catalyst treatments
2
2
0
.05 MPa using the treatment chamber adjacent to an ULVAC-PHI
Samples of 9 g of fresh catalysts were heated from room tem-
PHI-5500MC XPS system. This experimental set-up allows direct
transfer of treated catalysts from the treatment chamber to the
measurement chamber, preventing catalyst exposure to room air.
perature to 1073 K under a flow of air (1000 ml/min) over a period
of 1 h and maintained under these conditions for 5 h, then cooled
to room temperature in flowing air to maintain the oxidation state
of Pt. The catalysts treated under these oxidative conditions are re-
ferred to as oxidized catalysts.
In this case, Mg K (1253.6 eV) radiation was used, and the electron
a
escape depth and diameter of the analysis spot were approxi-
mately 2 nm and 800 lm, respectively. In both analyses, the Pt