124
J.-R. Kim et al. / Journal of Catalysis 263 (2009) 123–133
In this work, CeO2–ZrO2 mixed oxides were prepared by su-
BET surface area, pore volume, and pore size distribution of
Rh-loaded CeO2–ZrO2 mixed oxides were measured by N2 adsorp-
tion using ASAP2010 (Micromeritics Inc.). The samples were de-
percritical synthesis and co-precipitation method, respectively, and
they were used as support for Rh catalyst. The activities of Rh-
loaded CeO2–ZrO2 catalysts were investigated for catalytic reduc-
tion of NO by CO and their physicochemical properties were char-
acterized with TPR, N2 adsorption, O2-uptake, XRD, Raman, SEM,
AES, and H2/CO chemisorption. Discussions were made on the dif-
ferences in the catalytic performances between the two prepara-
tion methods of CeO2–ZrO2 supports in terms of reducibility, ho-
mogeneity, morphology, Rh dispersion, and thermal stability.
◦
gassed at 150 C for ca. 8 h, and N2 adsorption was carried out at
◦
−196 C.
The crystal structures of Rh-loaded CeO2–ZrO2 mixed oxides
were confirmed by powder X-ray diffraction (XRD) pattern using
monochromic CuK radiation (RIGAKU, D/MAX-2500) operating at
α
40 kV and 300 mA. The average crystallite size of CeO2–ZrO2
mixed oxide was also measured by the X-ray line broadening tech-
nique employing the Scherrer formula using the profiles of the
(1 1 1) peak.
2. Experimental
Vis-Raman spectra were obtained at room temperature using
a LabRAM HR UV/vis/NIR (Horiba Jobin Yvon) spectrometer with
an Ar ion laser of 514.5 nm excitation wavelength. Backscattering
2.1. Sample preparation
geometry was adopted for the measurement under the conditions
Ce0.65Zr0.35O2, which had the highest total OSC among CeO2–
ZrO2 mixed oxides [10], were prepared by supercritical synthe-
sis and co-precipitation method. In both preparation methods,
ZrO(NO3)2 and Ce(NO3)3 were used as the precursor of Zr and Ce,
while ammonia water as pH controller or precipitator. The detailed
processes for both preparation methods as well as the schematic
diagram of the apparatus for supercritical synthesis were described
elsewhere [10]. The final mixed oxides were obtained only with
drying in case of supercritical synthesis, while with further cal-
−1
of a laser power of 25 mW and a resolution of 2 cm
.
Scanning electron microscope (SEM) images of Rh-loaded CeO2–
ZrO2 mixed oxides were obtained by a field emission type—SEM
(HITACHI S4800) operating at an acceleration voltage of 1.0–2.0 kV.
The sample was prepared by sprinkling the powder onto viscous
graphite colloidal solution painted on a microscope stub, followed
by drying.
For the compositional depth profiles of Rh, auger electron spec-
troscope (AES) measurements were carried out with SAM4300
(Perkin–Elmer) operating at an acceleration voltage of 5 kV. The
cylindrical mirror analyzer (CMA) was used for the detection of
emitted auger electron. The powder sample was pressurized to a
pellet. The Ar ion beam used to sputter the pellet was operated at
an acceleration voltage of 3 kV.
◦
cination in air flow at 700 C for 5 h in case of co-precipitation
method. The fresh samples prepared by the supercritical synthesis
were denoted as “(S)”, and those by the co-precipitation method
as “(P)”.
Rh-loaded CeO2–ZrO2 mixed oxides with Rh content of 0.05,
0.1, 0.2, 0.5, and 1 wt% were prepared by incipient wetness im-
pregnation method. The aqueous solution of Rh(NO3)3 was added
ASAP2010 (Micromeritics Inc.) and Pulsechemisorb 2705 (Mi-
cromeritics Inc.) with TCD was used to determine the Rh dis-
persions. The amount of H2 adsorbed on Rh was determined on
◦
to CeO2–ZrO2 mixed oxide, followed by drying at 100 C for 12 h.
◦
The samples were calcined in air flow at 600 C for 3 h.
◦
ASAP2010 by H2 chemisorption at −80 C [26,27]. The amount of
CO adsorbed on Rh was also determined on Pulsechemisorb 2705
by O2–CO2–H2–CO pulse method proposed by Takeguchi et al. [28].
The details of each procedure are explained in the following:
2.2. Sample characterization
Temperature-programmed reduction (TPR) for Rh-loaded CeO2–
ZrO2 mixed oxide was carried out in a conventional flow apparatus
(Pulsechemisorb 2705, Micromeritics Inc.). A 0.05 g of sample was
◦
– H2 chemisorption at −80 C: The sample (0.5 g) was pre-
◦
◦
treated in O2 flow at 550 C for 1 h, then evacuated at 550 C
for 1 h. The sample was reduced in H2 flow at 500 C for 1 h,
then evacuated at 400 C for 4 h. After the sample was cooled
◦
◦
pretreated in synthetic air flow at 550 C for 1 h, and cooled in He
◦
◦
flow from 150 C to room temperature for O2 purge. The sample
◦
was then reduced in H2(5%)/Ar flow with the temperature increas-
down to −80 C using the cryogen (liquid nitrogen mixed with
◦
ing from room temperature to 1000 C at a constant heating rate
isopropyl alcohol), H2 chemisorption was carried out over a
range of H2 pressure of 2–20 Torr.
◦
of 10 C/min. The 4A molecular sieve trap was used to remove the
produced H2O during reduction, and the amount of consumed H2
was detected using thermal conductivity detector (TCD).
– O2–CO2–H2–CO pulse method: The catalyst (0.1 g) was de-
◦
gassed in O2 flow at 300 C for 1 h, then cooled down to
◦
Total OSC of Rh-loaded CeO2–ZrO2 mixed oxide, the total
amount of O2 which may be extracted from the sample at a pre-
established temperature and partial pressure of the reducing agent
(H2 and CO), was determined by the O2-uptake. The O2-uptake of
Rh-loaded CeO2–ZrO2 mixed oxide was measured in the same ap-
paratus as that of TPR. The sample was reduced in H2(5%)/Ar flow
30 C and flushed with He for 10 min. TPR was carried out
◦
in H2(5%)/Ar flow at a heating rate of 10 C/min and was
stopped as soon as the end of the first reduction peak. After
◦
the sample was cooled down to 30 C, it was flushed with He
for 10 min and was exposed to O2 flow for 10 min. CO2 was
fed to the sample for 10 min, and then purged with He for
10 min. H2(5%)/Ar gas was fed to the sample for 10 min, and
then purged with He for 1 h. 0.1 mL of CO was pulsed every
3 min until the intensity of the peak was a constant value.
◦
◦
at 1000 C for 30 min, and cooled to 427 C. An oxygen pulse was
injected every 3 min to the sample in the main stream of He at
◦
427 C to obtain the breakthrough curve, from which the total OSC
was determined.
The thermal stability of Rh-loaded CeO2–ZrO2 mixed oxide
could be evaluated by comparing the properties of the sample
redox-aged at high temperature with those of the fresh sample.
In a conventional flow apparatus, Rh-loaded CeO2–ZrO2 mixed ox-
A chemisorption stoichiometry ratio of CO:Rh and H:Rh was as-
sumed to be 1:1 for both cases.
2.3. Catalytic activity measurement
◦
ide was reduced in H2(5%)/He flow at 1000 C for 6 h, purged with
◦
He for 1 h, and oxidized in O2(5%)/He flow at 1000 C for 6 h. The
The catalytic activity was measured using a fixed-bed flow re-
actor at atmospheric pressure. Prior to each reaction, the sam-
redox-aged samples prepared by the supercritical synthesis were
denoted as “(S)_r”, and those by the co-precipitation method as
“(P)_r”.
◦
ples (0.5 g) were pretreated in O2(5%)/He flow at 550 C for 2 h.
A feed gas mixture of CO(1%), NO(0.1%), and O2(0.45%) passed