Three-way catalytic reactions on Rh-based catalyst: Effect of Rh/ceria interfaces

Article abstract of DOI:10.1016/j.apcata.2017.07.003

Rh-based catalysts were prepared by various methods and it was found that preparation methods play an important role in metal-support interaction (MSI) control which affects the catalytic performance of catalyst. The results suggest that the catalytic reduction of NO is mainly achieved by C₃H₆ in exhaust and H₂ generated from the water-gas shift reaction as well as the steam reforming of C₃H₈ and CH₄. Concentration of water in reaction stream has a significant influence on the water-gas shift and steam reforming reactions. The removal of C₃H₆ is accomplished by oxygen-induced oxidation instead of steam reforming reaction. In addition, Rh/ceria interactions promote the formation of active oxygen species and surface oxygen vacancy that respectively favors CO oxidation and NO reduction with a high N₂ selectivity. Rh@CeO₂ system shows high thermal stability due to Rh/ceria interaction.

Full text of DOI:10.1016/j.apcata.2017.07.003

Accepted Manuscript
Title: Three-way catalytic reactions on Rh-based catalyst:
effect of Rh/ceria interfaces
Authors: Chengxiong Wang, Tingting Zheng, Jun Lu,
Xiaodong Wu, Harold Hochstadt, Yunkun Zhao
PII:
S0926-860X(17)30303-4
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.apcata.2017.07.003
APCATA 16308
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
23-2-2017
3-6-2017
3-7-2017
Please cite this article as: Chengxiong Wang, Tingting Zheng, Jun Lu,
Xiaodong Wu, Harold Hochstadt, Yunkun Zhao, Three-way catalytic reactions
on Rh-based catalyst: effect of Rh/ceria interfaces, Applied Catalysis A,
Generalhttp://dx.doi.org/10.1016/j.apcata.2017.07.003
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Three-way catalytic reactions on Rh-based catalyst: effect
of Rh/ceria interfaces
b
Chengxiong Wang a,b, Tingting Zheng b, Jun Lu , Xiaodong Wu c, Harold
Hochstadt b, Yunkun Zhao a,b*
a State Key Laboratory of Advanced Technologies for Comprehensive Utilization of Platinum Metals,
Kunming Institute of Precious Metals, Kunming 650106, China
b State-Local Joint Engineering Laboratory of Precious Metal Catalytic Technology and Application,
Kunming Sino-platinum Metals Catalysts Co. Ltd., Kunming 650106, China
c Key Laboratory of Advanced Materials of Ministry of Education, School of Materials Science and
Engineering, Tsinghua University, Beijing 100084, China
* Corresponding author at: State Key Laboratory of Advanced Technologies for Comprehensive Utilization of
Platinum Metals, Kunming Institute of Precious Metals, Kunming 650106, PR China.
** E-mail addresses: ykzhao@sohu.com
1

Graphical abstract:
Highlights:
 Rh/ceria interfaces and Rh-CeO2-δ interactions were investigated by steady-state
H2 chemisorption, XPS and TPR measurements.
 Rh/ceria interfacial interactions promote the formation of active oxygen species
and surface oxygen vacancy, which is advantageous for the increase of catalytic
activity.
 Ceria sintering was inhibited by Rh nanoparticles as a diffusion barrier in
Rh@CeO2 system due to Rh/ceria interaction.
ABSTRACT: Rh-based catalysts were prepared by various methods and it was found
that preparation methods play an important role in metal-support interaction (MSI)
control which affects the catalytic performance of catalyst. The results suggest that the
catalytic reduction of NO is mainly achieved by C3H6 in exhaust and H2 generated
from the water-gas shift reaction as well as the steam reforming of C3H8 and CH4.
2

Concentration of water in reaction stream has a significant influence on the water-gas
shift and steam reforming reactions. The removal of C3H6 is accomplished by
oxygen-induced oxidation instead of steam reforming reaction. In addition, Rh/ceria
interactions promote the formation of active oxygen species and surface oxygen
vacancy that respectively favors CO oxidation and NO reduction with a high N2
selectivity. Rh@CeO2 system shows high thermal stability due to Rh/ceria interaction.
Keywords: Rh/ceria interface; metal-support interaction; reaction relationship; N2
selectivity
1. Introduction
To meet increasingly demanding automobile emission standards, precious metals
(PM) catalyst has received more significant attention for the simultaneous treatment of
the two reducing pollutants such as CO and uncombusted HC and oxides of nitrogen
(NOx) as the oxidizing pollutant [1], owing to their higher activity and thermal
stability compared to base metal catalysts [2]. Nevertheless, there is few literatures to
understand detailed surface reaction relationship in a PM-containing three-way
catalyst [3]. Ceria, as a typical oxygen storage material, has been widely used in
three-way catalyst due to its known ability to undergo rapid CeO2 ⇌ CeO2-δ cycles [4,
5]. The change between Ce3+ and Ce4+ ions causes the formation of surface defects
that can promote the storage, transport and release of oxygen [6]. This makes
metal-CeO2-δ interfaces play a particularly significant role in the catalytic reactions
which require the participation of oxygen, such as CO oxidation [7], CH4 combustion
[8], C3H8 oxidation [9] in automobile exhaust purification.
Catalytic activity and thermal stability of catalyst are strongly affected by
metal-support interactions [10, 11]. The Pd-(Ce, Zr)O2 interfaces are favored for CO
oxidation while the Pd-Al2O3 interface results in high NO conversion and C3H8
oxidation activity [12]. The interactions of Rh with Al2O3 undergo rapid deactivation
3

during thermal aging [11], whereas the strong interactions of Rh with ceria-containing
component can effect on oxidative/reductive properties and improve high-temperature
stability [10]. An enhanced PM-support interaction can be achieved via an oxo (–O–)
bridge between the PM and support at metal-support interfaces, which is particularly
useful for activation of CO oxidation [13-15]. The conventional Rh-supported
catalysts are usually obtained by an impregnation method for catalytic applications,
which displays poor rhodium-ceria interfaces and weak MSI. Recently, a few research
studies have reported the relationship between Rh/ceria interactions and catalytic
performance. It should be noted that the particle-particle interaction between the PM
and the support greatly depends on the interfacial area [16, 17], indicating that
metal-support interfaces are particularly important for MSI. Although certain
construction methods of metal-support interfaces have been reported in catalysis
applications, such as ‘‘core-shell’’ [18] or ‘‘yolk-shell’’ [19] metal@oxide
nanostructures, the combination of metal nanoparticles (NPs) and oxides into hybrid
nanostructures is still rarely reported [20]. However, a facile preparation of the
Rh@CeO2 hybrid nanostructure catalyst assisted by the aggregation of cerium
hydroxide grains has not been previously reported.
In this work, the Rh@CeO2 hybrid nanostructure catalyst was prepared via a facile
method to investigate the influence of Rh/ceria interface interactions on the activity
and selectivity. To gain some differences of interface interactions, the results are
correlated to characterization by steady state H2 chemisorption, H2-TPR, XPS, Raman
and in situ CO-DRIFTS. Activity and selectivity tests under a lean simulated exhaust
are investigated to understand the possible reactions and Rh/ceria interface-promoted
effect on catalytic activity and selectivity in three-way catalysis. Detailed surface
reaction relationship during three-way catalysis on Rh-based catalyst and effect of
water on the reactions are also investigated.
4

2. Experimental
2.1 Sample preparation
For the synthesis of Rh NPs, polyvinylpyrrolidone (PVP) and citric acid (CA) were
added into 85 mL of an Rh(NO3)3 aqueous solution containing 0.1 g Rh with the molar
ratio of 5 (monomer units of PVP) : 4 (CA) : 1 (Rh) and then placed in a 250 mL
three-flask. Then, the mixture solution was pre-heated in nitrogen atmosphere under
magnetic stirring at 60 ºC for 5 min. Adjusting the pH value to 3.0 by addition of 10%
NH4OH solution was followed. The requisite tert butyl amine borane (TBAB), served
as reducing agent in the reaction, was dissolved in 15 mL of distilled water and added
into above mixture solution with a molar ratio of 5 (TBAB) : 1 (Rh). The reaction was
allowed to proceed at 60 ºC for 10 min.
Ceria was synthesized by the precipitation method, the slight excess of 25%
NH4OH solution was dropwise added into a requisite amount of aqueous solution of
Ce(NO3)3·5H2O under continuous stirring. The obtained gelatinous precipitate was
aged at 80 ºC for 3 h under vigorously stirring and then filtered, and washed with
distilled water for several times until no pH change was observed. The precipitate was
dried at 120 ºC overnight and then calcined in a 300 ºC static air for 1 h. Then, the
resulting powders were calcined at 550 ºC for 4 h in an air atmosphere.
The obtained gelatinous precipitate (containing 58.1 mmol Ce) was separated by
centrifugation at 3000 rpm for 5 min and then dispersed in 300 mL of deionized water
under vigorous stirring. When the precipitate became homodispersed again, Rh NPs
(~0.29 mmol) was dropwise added under stirring. After aging, dry and calcination by
the same procedure in ceria preparation, Rh@CeO2 catalyst powders were obtained. A
reference Rh/CeO2 catalyst was prepared by an ex-situ method to ensure the similar
size of Rh NPs to the Rh@CeO2 samples, adding requisite amount of ceria powder
into the acidified Rh nanoparticle colloid under continuous stirring at room
temperature for 2 h. Then, the mixture was slowly evaporated at 60 ºC with vigorously
5

stirring and the resulting sample was calcined in a 550 ºC static air for 4 h. Thermal
aging of the samples was accomplished in a 850 ºC static air atmosphere for 10 h.
2.2 XRD, XPS and HRTEM
X-ray diffraction (XRD) patterns were recorded on a Philips X’pert Pro
diffractometer in the 2θ range of 5-90º with an increment step of 0.02º, using a Ni
filtered Cu Kα radiation (λ= 0.15418 nm) source. The X-ray tube was operated at 36
kV and 30 mA. The mean grain sizes of ceria in the catalyst samples were calculated
by XRD line-width broadening measurements on the (111) plane by the Scherrer
equation [21]. The chemical state and chemical composition on the surface of the
catalyst samples were performed by X-ray photoelectron spectroscopy (XPS) using a
physical electronics PHI model 5500 spectrometer quipped with a monochromatic Al
Kα (1486.6 eV) X–ray source to excite photoelectrons in an ultrahigh vacuum
atmosphere lower than 2×10-9 Pa. The measurement conditions used were: an
electron emission angle of 45º, 23.5 eV pass energy and 0.1 eV per step. Calibration of
XPS data was accomplished using the binding energy of C 1s (284.6 eV) as an internal
standard. High resolution transmission electron microscopy (HRTEM) images were
taken on a TECNAI F-30 high-resolution transmission electron microscope with an
accelerating voltage of 300 kV.
2.3 BET surface area and pore size distribution
The textural properties of samples were measured by nitrogen physisorption at
about 196 ºC using a Micromeritics ASAP2020 analyzer. Before sorption analysis, the
powder samples were first degassed at 200 ºC under a vacuum condition for 2 h, and
then about 100 mg degassed samples were used for the measurements. The specific
BET surface area (SBET) was evaluated from the adsorption data in the P/P0 range of
0.05-0.2 using the Brunauer-Emmett-Teller (BET) equation. The pore size
distributions were analyzed by MIP using a Micromeritics Poresizer 9320. The total
pore volumes were calculated from the single nitrogen absorption amount at the P/P0
6

of about 0.97.
2.4 H2-TPR
H2-temperature programmed reduction (TPR) experiments were performed in an
automated chemisorption analyzer with a thermal conduction detector (TCD) and
H2-Ar mixture gas (9.91% of H2 by volume) as reductant. 100 mg of sample was
placed in a quartz U-tube reactor for each measurement. The sample was first
pretreated in a Ar stream at 350 ºC for 60 min and then switched to H2-Ar mixture gas
with the flow rate of 100 mL·min–1. Temperature programmed experiments were
started from ambient temperature to 800 ºC at a heating rate of 10 ºC·min–1. The H2
consumption was monitored by TCD.
2.5 Rh dispersion measurements
The rhodium dispersion on Rh-containing catalyst was determined by H2
chemisorption at 100 ºC in a ChemBET Pulsar. Prior to each H2 measurement, the
catalyst was heated to 350 ºC in Argon for 60 min. After cooling to 100 ºC, the H2
adsorption experiment was performed using H2 pulse method. A ratio of H2/Rh was
used for the amount calculation of Rh atoms on the surface [22]. Actual content of Rh
was analyzed on PerkinElmer Optima 8000 ICP-OES instrument. The total Rh atoms
were calculated from actual content of Rh in catalyst.
2.6 Raman and in situ DRIFTS
Raman spectra were recorded on a LabRAM HR 800 (HORIBAJobin Yvon, France)
spectrometer quipped with an argon ion laser for the exciting source with a
wavelength of 532 nm. Raman data were obtained through collecting the laser beam
onto an area 0.1×0.1 mm2 in size of the sample surface under normal temperature and
pressure condition.
The in situ diffuse reflectance infrared Fourier transformed spectroscopy (DRIFTS)
spectra were collected on a Nicolet 6700 FTIR spectrometer equipped with a MCT
7

detector. 100 mg of sample was pre-treated at 500 ºC for 60 min in N2 stream and then
cooled to room temperature. Before starting the experiment, each background
spectrum was recorded at every 50 ºC from 50 ºC to 500 ºC in N2 stream at a heating
rate of 10 ºC·min–1. Each corresponding spectrum was measured in a reaction gas
stream with 1000 ppm CO for transient experiments. The flow rate of the gas stream in
these experiments was 100 mL·min-1.
2.7 Catalytic properties test
The catalytic activity and selectivity of the catalyst samples were studied with
three-way catalysis in a continuous flow fixed-bed microreactor. 500 mg catalyst
samples with an average diameter 40-60 mesh were placed in a stainless-steel tube
with an inner diameter of 5 mm. Before catalytic properties test, samples were
activated at 550 ºC in 21%O2/N2 stream for 30 min and then cooled down to ~100 ºC
in N2 stream. The total flow rate was 800 mL·min–1. The lean reaction mixture gas
(λ=1.007), included 1050 ppm CO, 560 ppm NO, 1800 ppm C3H6, 1000 ppm C3H8,
450 ppm CH4, 8% CO2, 5% H2O and 1.56% O2 in N2, was supplied at a flow rate of
800 mL·min–1 and the space velocity (Sv) was 96000 mL·g–1·h–1. Water was
introduced into the reactor by bubbling N2 gas through a stainless-steel vessel placed
liquid water, at 60 ºC. Temperature dependence of reaction was performed in a
temperature-programmed mode by heating from ambient temperature to 600 ºC at a
heating rate of 10 ºC·min–1. The composition of the effluent gas was analyzed online
using MKS instrument.
The conversion rate of CO, (CO,%) , was calculated according Eq. (1).
[CO]in [CO]
out 100%
(1)
(CO,%) 
[CO]in
where [CO]in and [CO]out represent the concentration of the injected and effluent CO
gas, respectively. The values of NO, C3H6, C3H8, and CH4 conversion rate were
calculated by the same method. The N2O, NH3, and N2 selectivity were calculated by
8

the Eqs. (2), (3) and (4) based on the concentration of N2O and NH3 products,
respectively.
2[N2O]
(2)
S(N2O,%) 
100%
[NO]in [NO]out
[NH3]
(3)
(4)
S(NH3,%) 
100%
[NO]in [NO]out
S(N2,%) 100  S(N2O,%)  S(NH3,%)
3. Results and discussion
3.1 Catalyst structures
3.1.1 XRD and chemical compositions
Fig. 1 shows the XRD patterns of catalyst samples after calcination. All XRD
patterns exhibits the characteristic peaks of well-crystallized fluorite-type phase of
ceria [23] and no Rh species are detected in all Rh-containing catalyst samples owing
to the concentration of Rh is lower than the detection limitation of XRD technique
[24]. The intensity of the characteristic peaks decreases when Rh NPs are supported,
especially Rh NPs are introduced through precipitation method.
To characterize the underlying interaction of Rh-CeO2-δ interface, XPS and Raman
measurements were employed to analyze the chemical state and electronic properties
of catalysts. Fig. 2 shows the detailed XPS spectra of Ce 3d, Rh 3d and O 1s and the
visible Raman spectra. It can be seen that each sample has a rather complex structure.
Ce3+ has been fitted with four peaks: Ce3+ 3d5/2(BE≈880.8 eV), Ce3+ 3d3/2(BE≈899 eV)
and two satellite peaks (s1, BE≈884.7, 902.6 eV) whereas Ce4+ has been fitted with six
peaks: Ce4+ 3d5/2(BE≈882.5 eV), Ce4+ 3d3/2(BE≈900.8 eV) and four satellite peaks (s2,
BE≈889.1, 898.3, 905.6, 916.8 eV) as shown in Table S1 [20, 25-27]. These results
illustrate the coexistence of Ce3+ and Ce4+ ions on the surface of all catalysts. There
exist Ce3+ ions in the ceria is closely related to the formation of oxygen vacancies
(VO-s) [6]. The calculated ratios of surface Ce3+/(Ce3+ + Ce4+) are listed in Table 1. The
9

ratio of surface Ce3+/(Ce3+ + Ce4+) of Rh@CeO2 sample is 0.300, which is higher than
CeO2 support and Rh/CeO2. That is to say, there is more concentration of VO-s on the
surface of Rh@CeO2. These small differences in amounts can be ascribed to the
interfacial interaction between Rh and CeO2-δ in order to maintain electrostatic
equilibrium.
In addition, the Rh 3d core level region of Rh metal and Rh2O3 in Rh-containing
catalysts are also shown in Fig. 2b. As shown in Table S2, the peaks at about 310 and
314.8 eV can be assigned to the 3 d5/2,3/2 doublet of Rh3+ final state whereas the other
two peaks centered at ~308.4 and ~313.3 eV are respectively related to Rh0 d5/2 and
Rh0 d3/2, which implies Rh3+ and metallic Rh coexist on the surface [27, 28]. As listed
in Table 1, the calculated ratio of surface Rh3+/(Rh0 + Rh3+) in Rh@CeO2 catalyst is
relatively high and the value is 0.345, which is much higher than that of Rh/CeO2. In
Fig. S1, two O 1s XPS peaks can be seen at binding energies of ~529.4 and ~531.6 eV
(Table S3), corresponding to surface lattice oxygen (OI) and non-lattice oxygen (OII),
respectively [6]. Table 1 shows the portion of non-lattice oxygen component on the
CeO2 surface is 36.1% and this oxygen species concentration significantly increases
when Rh NPs were introduced.
For investigating the bulk oxygen vacancies (VO-b) of catalyst, visible Raman
spectroscopy technique was used. Fig. 2c shows that all Raman spectra present a
strong peak in the range of 420–500 cm-1 with a weak shoulder at ∼600 cm-1, which
can be attributed to the F2g and defect-induced (D) modes of CeO2-x, respectively [29].
It can be clearly seen in Table 1 that Rh@CeO2 exhibits strongly higher intensity ratio
(ID/IF2g) than other samples, indicating that there is more concentration of Ce3+ ions in
Rh@CeO2 catalyst.
Table 1 reports the crystallize size of ceria calculated by the Scherrer equation based
on XRD line-width broadening of CeO2 (111) and displays the corresponding
chemical compositions. It can be seen that the pure CeO2 has an average size of 16.3
nm, and the addition of Rh NPs by precipitation results in a significant decrease in
10

crystallize size of ceria as well as a slight shift toward lower angles caused by the
formation of Ce3+ ions according to XPS and Raman results. However, Rh/CeO2 has
no obvious decrease in crystallize size and any angle shifts, compared to the results of
the pure CeO2 support. The XPS and visible Raman investigations show that
Rh@CeO2 catalyst has the highest Ce3+ concentration which agrees well with the XRD
results. The additional Ce3+ ions in the catalyst surface maintains electrostatic
equilibrium on Rh/ceria interfaces [30]. This indicates that the precipitate process
facilitates the interaction between Rh and CeO2-x. The experimental compositions of
Rh-containing catalysts determined by ICP-OES are similar to the nominal ones and
Rh/CeO2 catalyst shows the higher Rh dispersion compared to Rh@CeO2, reaching
16.5%.
3.1.2 BET and pore structures
As can be seen from Fig. S2a, the N2 adsorption-desorption isotherms of the three
samples are typical isotherm type IV according to the IUPAC classification, which
indicates mesoporous materials. All samples exhibit an obvious type H2 hysteresis
loop at the part of higher relative pressure than 0.6. Fig. S2b shows the corresponding
results of the pore size distribution measurements. It can be observed that there is a
pore size distribution centered at 13 nm for CeO2. However, the BJH pore size
distribution of Rh-containing catalysts shows two peaks centered at pore size of 4 nm
and above. Of course, the main pore sizes of Rh/CeO2 and Rh@CeO2 are centered at
13 and 8 nm, respectively. The textual properties of the catalyst samples are
summarized in Table S4. The BET surface area of Rh@CeO2 is much larger than those
of CeO2 (41.5 m2/g) and Rh/CeO2 (36.1 m2/g), reaching 83 m2/g. Moreover,
Rh@CeO2 has a 0.136 cm3·g−1 pore volume that is higher than those of other two
samples.
As above mentioned, a smaller crystallite size in Rh@CeO2 sample results in a
significant increase in Sexter and VT values as well as decrease in Dp value. The H2 type
hysteresis loop in the P/P0 range of 0.4–0.95 illustrates that the sample is characterized
11

with mesoporous structure [6]. The results are consistent with pore size distribution.
Nevertheless, supporting Rh NPs by ex-situ method leads to the smaller hysteresis
loop and the reduction of SBET and VT than CeO2 sample, implying that partial pores
are occupied by Rh NPs. But more noteworthy is shift of the CeO2(111) diffraction
peak in Rh@CeO2 sample toward lower angle, suggesting that the lattice parameters
of CeO2 have been changed and more Ce3+ cations form due to a larger ionic radius of
the Ce3+ than Ce4+ ions.
3.1.3 HRTEM
The results of electron microscope observations of catalyst samples are shown in Fig.
3. The average crystallize sizes of ceria in CeO2, Rh/CeO2 and Rh@CeO2 samples
determined from TEM images (Fig. 3a, c, e) are about 19 nm, 18 and 9 nm,
respectively, which well agrees with the XRD results. However, clear Rh NPs are not
observed from TEM image with a low magnification because the high electron density
of ceria support makes the observation and discrimination of Rh NPs by TEM
technique particularly difficult [16]. Fig. 3b, Fig. 3d and Fig. 3f show the HRTEM
images of CeO2, Rh/CeO2 and Rh@CeO2 samples respectively and lattices of ceria
crystal and Rh (111) plane can be clearly observed. Rh NPs with a size of about 5 nm
and the typical Rh@CeO2 hybrid nanostructure and the Rh/ceria interface can be
clearly seen from Fig. 3f.
3.2 Temperature-programmed reduction
The reducibility of catalyst samples was investigated by H2-TPR measurement (Fig.
4). In the case of the pure CeO2 support, the H2 consumption values of peaks at 133,
361 and 552 ºC are 12, 64 and 45 μmol/gcat. as listed in Table S5, respectively. The first
peak is assigned to the formation of CeO2Hx due to H2 adsorption and the reduction of
surface capping oxygen (O2–) whereas the additional peaks are related to surface layers
reduction (2CeO2 + H2 → Ce2O3 + H2O) [21, 29, 31].
On the Rh-containing samples, the process was more complex because it includes
12

the reduction of Rh2O3 (Rh2O3 + 3H2 → 2Rh0 + 3H2O) and CeO2, as well as the H2
chemisorption and desorption equilibrium on the metallic Rh (2Rh0 + xH2 → 2RhHx)
[31]. The H2 consumption in the temperature range of 100–240 ºC is mainly associated
with the reduction of surface O2- species because the H2 uptake value on the metallic
Rh to form RhHx species is no more than 2.5 μmol/gcat. according to H2 pulse
experiment results [29]. Most of the O2−-derived H2 consumption is attributable to the
reduction of CeO2 that is in contact with Rh metal and another is due to the reduction
of Rh2O3 to Rh0 [6, 27]. Thus, the first two peaks are closely relevant to Rh/ceria
interfaces. The related H2 consumption value on Rh/CeO2 is 189 μmol/gcat., which is
larger than the 98 μmol/gcat. required to reduce O2- species on the surface of Rh@CeO2.
The H2 uptake peaks at higher temperature can be assigned to CeO2 surface layers
reduction and the corresponding H2 consumption values on Rh/CeO2 and Rh@CeO2
are 60 and 314 μmol/gcat., respectively.
DRIFTS spectra recorded for the catalyst samples during CO-TPR experiments are
shown in Fig. 5. The spectra are dominated by abutting shoulders at 2240–1960 cm-1,
which can be attributable to the formation of carbonyl species due to CO adsorption
on catalyst surface [32, 33]. As for the CeO2 catalyst, the gas-phase CO (v = 2194 cm-1)
and Ceδ+-carbonyl (bands at 2150 and 2140 cm-1) peaks [32, 34] can not be observed
until raising the temperature to 250 ºC, illustrating that a clear CO adsorption on CeO2
requires relatively high temperature. By contrast, a band at ~2170 cm-1, ascribed to CO
adsorbed on unsaturated Ceδ+ cations, and three bands at 2078, 2012 and 1976 cm-1,
respectively assigned to gem-dicarbonyl, linear and bridged carbonyl species formed
by CO adsorbed on Rh sites [33, 35], start to be observed at 100 ºC from DRIFTS
spectrum for the Rh/CeO2 catalyst. The bridged carbonyl species disappears after
increasing the temperature to 150 ºC in CO/N2 stream while the band at 2078 cm-1
shows a shift to higher wavenumber indicating an increase in the CO surface coverage
on Rh sites [36]. A further increase in temperature results in an increase in the
intensity of the carbonyl bands. The peaks associated with Rh-(CO)x species become
13

less intense when the temperature is raised to 250 ºC, and at the same time, the bands
are shifted to lower wavenumbers. Nevertheless, there is no difference in intensity of
Ceδ+-carbonyl with temperature increase, which shows that it is difficult to activate
Ceδ+-carbonyl species. The spectrum features of CO(ads) on the Rh@CeO2 catalyst
appear at lower wavenumbers at every same temperature, compared to that of the
Rh/CeO2 catalyst. The results can be explained by a lower CO surface coverage [36],
which reveals that the Rh@CeO2 catalyst has a stronger Rh/ceria interaction and a
better activity for CO oxidation.
Rh@CeO2 catalyst shows the lower Rh dispersion (Table 1) and O2−-derived H2
uptake values (Table 3) as well as the higher reduction temperature of H2 (Fig. 5) by
contrast to Rh/CeO2, which implies less exposed Rh atoms and enhanced Rh/ceria
interactions. As displayed in Fig. 6, the Rh-(CO)x bands from Rh@CeO2 also show a
lower wavenumbers than that of Rh/CeO2 at every same temperature, which indicates
the lower coverage of CO(ads) species. These results suggest that Rh/ceria interface
interactions increase in the following order: Rh/CeO2 < Rh@CeO2.
3.3 Catalytic activity and selectivity
3.3.1 Catalytic activity
Three-way catalysis reactions are of great importance in automobile exhaust
purification. In the work, three-way catalysis is carried out using a typical fixed-bed
flow reactor under a stoichiometric simulated exhaust. The typical CO and C3H6
conversion curves as a function of temperature in three-way catalysis reactions are
shown in Fig. 6 and the corresponding characteristic conversion temperatures,
including the light-off temperatures at 50% conversion (T50) and 90% conversion (T90)
are displayed in Table 2. The comparison of the catalytic activity is confirmed by the
characteristic temperatures, T50 and T90.
It can be seen that, in Fig. 6a, CO pollutant starts to be clearly oxidized on the pure
CeO2 support at 250 ºC. Unlike C3H6 conversion, the CO conversion rate decreases
increasing temperature to 390 ºC with a CO conversion of 85% until up to 495 ºC,
14

illustrating that CO forms due to partial oxidation of CH4 and C3H8 in the temperature
range of 390–495 ºC according to the results shown in Fig. S3. The purification
efficiency of CO pollutant on the pure CeO2 support can reach 90%, above 500 ºC.
When the Rh NPs are introduced, similar phenomena for CO conversion curves are
also observed but the purification efficiency of CO pollutant decreases with a
temperature increment to ~350 ºC although the activity for CO oxidation improves in
the low temperature domain. It’s also observed that CO is almost completely oxidized
on the Rh@CeO2 and CeO2 catalysts in the temperature ranges of 310–370 ºC and
520–590 ºC, respectively, whereas the CO complete conversion can’t be achieved on
the conventional Rh/CeO2 supported catalyst. Table 2 presents that the T50 values of
CO and C3H6 oxidation reactions on the pure CeO2 support are 312 and 416 ºC, and
the corresponding T90 values are 504 and 474 ºC, respectively. Moreover, the
characteristic temperatures of CO (T50= 248 ºC, T90=322 ºC) and C3H6 (T50= 327 ºC,
T90=399 ºC) conversions on the Rh/CeO2 catalyst are further higher than that of the
Rh@CeO2 catalyst. When Rh NPs was supported, the light-off temperature shows a
significant reduction. But the CO conversion begins to degrade above 345 ºC with a
slow increase in conversion of C3H6 likely due to the lack of active oxygen species for
CO oxidation, which indicates that C3H6 competes with CO for O2 [3]. A significant
activity enhancement for CO and C3H6 oxidation can also be observed along with the
increasing interactions of Rh/ceria interfaces, since the high values of surface
Ce3+/(Ce3+ + Ce4+) and Rh3+/(Rh0 + Rh3+) are is beneficial to the activity increase of
catalytic oxidation [6,27].
Fig. S3 shows that the Rh@CeO2 catalyst presents highest C3H8 and CH4
conversions whereas Rh/CeO2 catalyst shows marginally lower conversions and CeO2
catalyst presents significantly lower C3H8 (below 500 ºC) and CH4 conversions. The
C3H8 oxidation activity of Rh@CeO2 and Rh/CeO2 catalysts is comparable with a
maximum C3H8 conversion of 73% at 430 ºC and 57% at 444 ºC, and the light-off
temperatures of 389 and 417 ºC, respectively. However increasing C3H8 and CH4
15

conversions for CeO2 catalyst are showed with increase in the reaction temperature. In
all catalysts, the CH4 oxidation activity can be comparable with CH4 conversion rate
comparison at the same temperature. Therefore, Fig. S2 depict that the activity of C3H8
and CH4 oxidation increases in the following order: CeO2 < Rh/CeO2 < Rh@CeO2.
The complete removal of C3H8 and CH4 pollutants can’t be achieved under a lean
simulated exhaust, which indicates that only part of C3H8 and CH4 are oxidized
although there is excess oxygen in the mixture stream. Moreover, conversion of C3H8
and CH4 is primarily achieved on active Rh and Rh/ceria sites due to the low
conversion rate on the ceria surfaces. The Rh@CeO2 catalyst shows the higher
catalytic activity for C3H8 and CH4 oxidation than that of Rh/CeO2 at every reaction
temperature, which can be attributed to Rh/ceria interface-promoted activity
enhancement.
3.3.2 N2O, NH3 and N2 Selectivity
Fig. 7 shows the NO conversion and corresponding selectivity curves for the
reduction of NO to N2 in the presence of excess O2 on different catalysts. In the case
of CeO2, the NO conversion reaches 17% at 485 ºC and the light-off temperature is
580 ºC. It also can be observed that the N2O and NH3 are mainly detected below 500
ºC. The N2O selectivity raise to 11% at 345 ºC but then it starts to drop with further
temperature increase, yet Rh-containing catalysts exhibit a similar change regulation
of N2O selectivity curves. Nevertheless, the NH3 selectivity gradually decreases with
increase in temperature and is significantly higher below ~400 ºC but marginally
lower with further temperature increments than that of the Rh-containing catalysts. N2
selectivity increases slowly from 64% at 250 ºC to 99% when the temperature is raised
to 590 ºC.
When Rh NPs were introduced, the light-off temperature is significantly lower than
that of NO reduction on CeO2, indicating that Rh sites play a significant role in the
reduction of NO. The Rh@CeO2 sample shows a T50 of 314 ºC and T90 of 333 ºC for
NO conversion, by comparison, while there are the T50 and T90 increments of 33 and
16

51 ºC in the case of the Rh/CeO2 sample. At low temperature, by contrast to the CeO2
support, an increase in selectivity of N2O is discernible and likely due to adsorption of
NO on Rh surface nitrosyl species which favor to form N2O [37]. Furthermore,
additional NH3 formation can be observed with a rise in temperature to 400 ºC or
above. Those observations emphasize the fact that there is selective reduction of NO to
NH3 by hydrogen. At 445 ºC, NH3 selectivity of 15% is obtained on the Rh/CeO2
catalyst whereas less NH3 (only 5.5% of selectivity) forms with a more accentuated N2
selectivity on Rh@CeO2.
As far as the reduction of NO to N2 on Rh-containing catalysts is concerned, it is
difficult to completely remove NO at lean conditions without water feed unless the
NO/H2 reaction [3, 37]. Interestingly, NO conversion was reached to 100% at
300–350 °C and then kept 100% in higher temperature range during three-way
catalysis on Rh-containing catalysts. This is principally because of the reduction of
NO by hydrogen.
3.3.3 Three-way reactions on the Rh@CeO2 catalyst
Although many researches concern to the mechanism of three-way reactions, there
is as yet no agreement as to how different components react with each other in real
situations [3, 12]. Thus, as shown in Figure 8a, the consumption of CO, NO, C3H6,
C3H8 and CH4 as well as the formation of CO2, H2O, N2O and NH3 were
simultaneously analyzed to understand the reaction relationship. At 230 ºC, reducing
NO concentration to 405 ppm with N2O formation is observed. A further increase in
the temperature to 270 ºC results in a decrease in the concentration of NO with
increased N2O formation. Near full conversion of NO is observed at 336 ºC, but at the
same time 190 ppm C3H6 can still be detected. The results suggests that NO reduction
is primarily attributed to CO-derived reaction below 230 ºC while NO mainly reacts
with C3H6 to release N2O increasing temperature to 336 ºC. When the temperature is
raised to 278 ºC or above, C3H8 starts significant conversion, which leads to an
increase in the CO concentration. It also can be seen that CO conversion increases as
17

rise in the C3H8 concentration, implying that CO is produced by the partial oxidation
of propane. In the temperature range of 406–450 ºC, produced CO2 clearly reduces in
spite of C3H8 consumption, which indicates partial C3H8 conversion via a reforming
reaction. Then, hydrogen produced by reforming of propane reacts with NO to form
NH3 as can be seen from Figure 8a.
From Figure 8b, H2-TPR profiles of used Rh@CeO2 catalysts show the evolution of
surface active oxygen species during reactions. No clear H2 uptake at around 228 and
245 °C can be observed when the reaction temperature was raised to 250 °C. The main
peaks at 119 and 182 °C correspond to reduction of different Rh sites in the CeO2
matrix and the hydrogen uptake values are 18 and 9.6 μmol/gcat., respectively, and the
broad H2 consumption peak centered at ~515 °C can be attributed to the reduction of
ceria [27]. These results indicate that O2- species play an important role in the catalytic
oxidation of CO. Consequently, a high active O2- species concentration, which is
related to Rh/ceria interactions according to the results listed in Table S5, might be
advantageous for CO oxidation. The negative hydrogen uptake at 343 °C is closely
related to H2 desorption, as is seen in the H2-TPD result. There is hardly H2
consumption associated with the reduction of surface O2– species increasing
temperature up to 600 °C.
In three-way catalytic reactions, water forms via oxidation reaction of HC and can
also convert CO and HC by the water-gas shift and steam reforming reactions, which
makes investigating surface reaction mechanism more difficult. The activity of the
Rh@CeO2 catalyst for CO, C3H6, C3H8 and CH4 oxidation as well as NO reduction as
a function of water concentration can be observed in Figure S4a. The addition of water
to the feed stream has no significant effect on C3H6 oxidation and N2O formation
which is well consistent with the literature [38], and decreasing water concentration
results in enhanced C3H6 and NO conversions due to the competitive adsorption of
H2O, C3H6 and NO molecules on the same sites. Unlike C3H6 and NO, the
enhancement effect of water on CO, C3H8 and CH4 conversions can be observed,
18

illustrating that the roles of the water-gas shift and steam reforming reactions should
be considered. It also clearly shows that increasing water addition leads to increase in
CO2 and NH3 formation. Concerning the hydrogen formation in three-way reactions, it
is obvious from Figure S4a that three are water-induced reactions including the
water-gas shift and steam reforming of propane and methane to forms hydrogen, as
follows [39].
CO + H2O → H2 + CO2
C3H8 + 3H2O → 7H2 + 3CO
CH4 + H2O → 3H2 + CO
(5)
(6)
(7)
Figure S4b depicts the steady-state sequence of the stoichiometric/rich and
stoichiometric/lean fuel cycles at 440 °C, which further reveals the reaction
relationships between different components. In the initial step, the Rh@CeO2 catalyst
was exposed to a stoichiometric mixture for 5 min and the values of all reactant
consumption were kept constant. When the reaction mixture was switched to rich
condition (λ=0.8), both CO and HC conversions drop suddenly with increase in NO
conversion, which results in the reduced CO2, H2O and increased NH3 formations. On
the contrary, the activities of CO and HC oxidation are dramatically increased whereas
NO reduction activity is depressed when excess O2 (λ=1.2) was added in the feed
mixture. At this state, increasing N2O forms as the reduction of NO consumption and
NH3 formation. Interestingly, switching the reaction mixture from lean to
stoichiometric conditions leads to more C3H8 and CH4 consumption in spite of less
C3H6 conversion and H2O formation, at the same time CO concentration is more than
the initial state, illustrating that C3H8 and CH4 are oxidized by water via steam
reforming reaction to release hydrogen besides partially oxidized to form CO.
As the concern about reduction of NO to N2, the results from Fig. 7 indicates that
Rh/ceria interfacial interactions can promote NO dissociation and improve N2
selectivity, which can be ascribed to surface oxygen vacancies [40]. Surprisingly, NO
conversion was reached to 100% at 300-350 °C and then kept 100% in higher
19

temperature range although the excess amount of O2 was present in the reaction gas.
This might due to the reduction of NO by formed H2 and NH3 according to following
reactions [37, 41]:
2NO + 2H2 → N2 + 2H2O
(8)
2NO + 5H2 → 2NH3 + 2H2O
6NO + 4NH3 → 5N2 + 6H2O
4NO + 4NH3 + O2 → 4N2 + 6H2O
(9)
(10)
(11)
Of course, NO conversion is closely related to air/fuel ratio fluctuation as shown in
Figure S4b. Fluctuating stoichiometric to rich conditions results in more NH3
formation, which is advantageous for the catalytic reduction of NO [41]. However,
increasing O2 concentration makes against NO reduction because less NH3 forms.
3.4 Thermal stability
Thermal-induced sintering of ceria in catalysts was investigated by means of XRD
and specific surface area analysis. XRD patterns for aged samples are shown in Figure
S5 and the corresponding crystallite sizes of ceria for fresh and aged samples are listed
in Table 3. It can be observed that aged Rh/CeO2 sample has much stronger
characteristic diffraction peaks and bigger ceria crystallite, compared to that of aged
Rh@CeO2 sample. Thermal aging leads to significant SBET loss as can be seen in Table
3. The measured SBET value for the Rh/CeO2 catalyst reduces from 36.1 m2/g to 4.3
m2/g while Rh@CeO2 system still has a SBET value of 13.7 m2/g after high-temperature
thermal aging. In aim to compare the sintering rate of samples, (Da–Df)/Df and
(Sf–Sa)/Df values were introduced, where Df, Da, Sf, Sa present crystallite sizes of ceria
in fresh and aged samples, SBET values of fresh and aged samples, respectively. A high
value in either (Da–Df)/Df or (Sf–Sa)/Sf implies a high sintering rate. From XRD and
specific surface area results, it can be concluded that Rh@CeO2 system shows a better
structural stability and lower sintering rate with comparison to conventional Rh/CeO2
supported catalyst. This is to a large extent on Rh nanoparticles as a diffusion barrier
in Rh@CeO2 system inhibit ceria sintering due to Rh/ceria interaction.
20

Figure S6 shows light-off curves of three-way catalysis for aged samples and
comparison in temperature of 50% conversion (T50) for fresh and aged samples are
summarized in Table 3. The results indicate that a noticeable decrease in catalytic
activity occurs during the thermal treatment. For instance, there is no catalytic activity
for CH4 oxidation in studied range of test temperature undergoing thermal aging at
high temperatures. Thermal aging procedure of samples also results in significant
increase of T50 values. However, 50% conversions of CO, NO, C3H6 and C3H8 were
respectively accomplished at 250, 348, 353 and 580 ºC on the surface of Rh@CeO2
catalyst, which is much lower than ones for Rh/CeO2. The results illustrate that
Rh@CeO2 system exhibits high thermal stability.
4 Conclusions
Effective construction of Rh/ceria interfaces was achieved via a hybrid
aggregation-mediated crystallization method. Rh/ceria interfacial interactions have a
significant promoting effect on the formation of active oxygen species, which is
beneficial to the enhancement of catalytic performance for CO oxidation. More
surface oxygen vacancies were formed with increase in Rh/ceria interactions to
promote NO dissociation. MSI also plays a significant role in improvement of
catalytic activity of HC oxidation and N2 selectivity. In three-way catalytic reactions,
C3H6 is mainly oxidized by O2 and NO whereas the removal of C3H8 and CH4 can be
achieved by the steam reforming reaction to release H2. Hydrogen, also generated
from the water gas shift reaction at high temperatures, further reacts with NO to form
NH3. These formed H2 and NH3 as the reductants will promote the catalytic reduction
of NO. The water gas shift and steam reforming reactions significantly depend on
concentration of water in reaction stream. Rh nanoparticles as a diffusion barrier in
Rh@CeO2 system inhibit ceria sintering due to Rh/ceria interaction, making that
Rh@CeO2 system shows high thermal stability and might be a promising candidate
catalyst to three-way catalysis.
21

Acknowledgement
We are grateful for the financial supports of the National Natural Science
Foundation of China (Nos. 21463015) and the Provincial Natural Science Foundation
of Yunnan (Nos. 2014FA045).
References
[1] J.J. He, W.C. Wang, T.T. Zheng, Y.K. Zhao, Johnson Matthey Technol. Rev., 60
(2016) 196-203.
[2] H. Gandhi, J. Catal., 216 (2003) 433-442.
[3] D. Chatterjee, O. Deutschmann, J.r. Warnatz, Faraday Discuss., 119 (2001)
371-384.
[4] J. Guo, Z. Shi, D. Wu, H. Yin, M. Gong, Y. Chen, J. Alloy. Compd., 621 (2015)
104-115.
[5] L. Ilieva, G. Pantaleo, N. Velinov, T. Tabakova, P. Petrova, I. Ivanov, G. Avdeev,
D. Paneva, A.M. Venezia, Appl. Catal. B: Environ., 174-175 (2015) 176-184.
[6] Y. Su, Z. Tang, W. Han, P. Zhang, Y. Song, G. Lu, CrystEngComm, 16 (2014)
5189-5197.
[7] P. Venkataswamy, K.N. Rao, D. Jampaiah, B.M. Reddy, Appl. Catal. B: Environ.,
162 (2015) 122-132.
[8] M. Cargnello, J.J. Delgado Jaen, J.C. Hernandez Garrido, K. Bakhmutsky, T.
Montini, J.J. Calvino Gamez, R.J. Gorte, P. Fornasiero, Science, 337 (2012)
713-717.
[9] J.A. Enterkin, W. Setthapun, J.W. Elam, S.T. Christensen, F.A. Rabuffetti, L.D.
Marks, P.C. Stair, K.R. Poeppelmeier, C.L. Marshall, ACS Catal., 1 (2011)
629-635.
[10] E.A. Alikin, A.A. Vedyagin, Top. Catal., 59 (2016) 1033-1038.
[11] V.O. Stoyanovskii, A.A. Vedyagin, G.I. Aleshina, A.M. Volodin, A.S. Noskov,
Appl. Catal. B: Environ., 90 (2009) 141-146.
22

[12] M. Shen, M. Yang, J. Wang, J. Wen, M. Zhao, W. Wang, J. Phys. Chem. C, 113
(2009) 3212–3221.
[13] S. Hinokuma, H. Fujii, Y. Katsuhara, K. Ikeue, M. Machida, Catal. Sci. Technol.,
4 (2014) 2990-3017.
[14] S. Hinokuma, H. Fujii, M. Okamoto, K. Ikeue, M. Machida, Chem. Mater., 22
(2010) 6183-6190.
[15] Y. Nagai, K. Dohmae, Y. Ikeda, N. Takagi, N. Hara, T. Tanabe, G. Guilera, S.
Pascarelli, M.A. Newton, N. Takahashi, H. Shinjoh, S.i. Matsumoto, Catal. Today,
175 (2011) 133-140.
[16] M. Cargnello, V.V. Doan-Nguyen, T.R. Gordon, R.E. Diaz, E.A. Stach, R.J. Gorte,
P. Fornasiero, C.B. Murray, Science, 341 (2013) 771-773.
[17] R. Kou, Y. Shao, D. Mei, Z. Nie, D. Wang, C. Wang, V.V. Viswanathan, S. Park,
I.A. Aksay, Y. Lin, Y. Wang, J. Liu, J. Am. Chem. Soc., 133 (2011) 2541-2547.
[18] L. Adijanto, D.A. Bennett, C. Chen, A.S. Yu, M. Cargnello, P. Fornasiero, R.J.
Gorte, J.M. Vohs, Nano Lett., 13 (2013) 2252-2257.
[19] C. Chen, X. Fang, B. Wu, L. Huang, N. Zheng, ChemCatChem, 4 (2012)
1578-1586.
[20] Q. Tan, C. Du, Y. Sun, G. Yin, Y. Gao, J. Mater. Chem. A, 2 (2014) 1429-1435.
[21] L. Liu, B. Liu, L. Dong, J. Zhu, H. Wan, K. Sun, B. Zhao, H. Zhu, L. Dong, Y.
Chen, Appl. Catal. B: Environ., 90 (2009) 578-586.
[22] C. Force, A. Ruiz Paniego, J. M. Guil, J. M. Gatica, C. López-Cartes, S. Bernal, J.
Sanz, Langmuir, 17 (2001) 2720-2726.
[23] J.A. Rodriguez, J.C. Hanson, J.-Y. Kim, G. Liu, A. Iglesias-Juez, M.
Fernández-García, J. Phys. Chem. B, 107 (2003) 3535-3543.
[24] F.-Y. Chang, M.-Y. Wey, J.-C. Chen, J. Hazard. Mater., 156 (2008) 348-355.
[25] F. Larachi, J. Pierre, A. Adnot, A. Bernis, Appl. Surf. Sci., 195 (2002) 236-250.
[26] W. Xiao, Q. Guo, E.G. Wang, Chem. Phys. Lett., 268 (2003) 527-531.
[27] A. Gayen, K.R. Priolkar, P.R. Sarode, V. Jayaram, M.S. Hegde, G.N. Subbanna, S.
23

Emura, Chem. Mater., 16 (2004) 2317-2328.
[28] D.R. Mullins, S.H. Overbury, Surf. Sci., 511 (2002) L293-L297.
[29] Y. Gao, A. Duan, S. Liu, X. Wu, W. Liu, M. Li, S. Chen, X. Wang, D. Weng,
Appl. Catal. B: Environ., 203 (2017) 116-126.
[30] Y. Su, Z. Tang, W. Han, P. Zhang, Y. Song, G. Lu, CrystEngComm, (2014).
[31] J.E. Fallah, S. Boujana, H. Dexpert, A. Kiennemann, J. Majerus, O. Touret, F.
Villain, F.L. Normand, J. Phys. Chem, 98 (1994) 5522-5533.
[32] A. Hornés, P. Bera, A.L. Cámara, D. Gamarra, G. Munuera, A. Martínez-Arias, J.
Catal., 268 (2009) 367-375.
[33] P. LÖÖf, B. Kasemo, L. BjÖrnkvist, S. Andersson, A. Frestad, Stud. Surf. Sci.
Catal., 71 (1991) 253-274.
[34] C.D. Zeinalipour-Yazdi, A.L. Cooksy, A.M. Efstathiou, J. Phys. Chem. C, 111
(2007) 13872-13878.
[35] V. Schwartz, A. Campos, A. Egbebi, J.J. Spivey, S.H. Overbury, ACS Catal., 1
(2011) 1298-1306.
[36] A. Karelovic, P. Ruiz, ACS Catal., 3 (2013) 2799-2812.
[37] P. Granger, S. Pietrzyk, CR Chim., 17 (2014) 656-671.
[38] H.J. Kwon, J.H. Baik, Y.T. Kwon, I.-S. Nam, S.H. Oh, Chem. Eng. J., 141 (2008)
194-203.
[39] J. Wang, M. Shen, J. Wang, J. Gao, J. Ma, S. Liu, J. Rare Earth., 30 (2012)
748-752.
[40] W. Ding, W. Li, Chinese J. Catal., 35 (2014) 1937-1943.
[41] P. Bera, M.S. Hegde, J. Indian I. Sci., 90 (2010) 299-325.
24

Figures
Rh@CeO2
Rh/CeO2
CeO2
20
30
40
50
2 / ()
60
70
80
90
Fig. 1. XRD patterns of the catalyst samples
4+
(a)
Ce (3d3/2
)
Ce 3d
Ce4+(3d5/2
Ce3+(3d5/2
)
Ce3+(3d3/2
)
s2
s
2
)
Rh@CeO2
s
1
s
2
s1
s2
Rh/CeO2
CeO2
925 920 915 910 905 900 895 890 885 880 875
Binding Energy / eV
25

(b)
Rh 3d
Rh0(3d5/2
)
Rh3+(3d5/2
)
Rh0(3d3/2
)
Rh3+(3d3/2
)
Rh@CeO2
Rh/CeO2
318
316
314
312
310
308
306
Binding Energy / eV
(c)
CeO2
Rh/CeO2
Rh@CeO2
591
605
600
560
580
600
620
Raman Shift / cm-1
300 350 400 450 500 550 600 650 700
Raman Shift / cm-1
Fig. 2. (a-b) XPS and (c) visible Raman spectra of the catalysts
(a)
(b)
26

(c)
(d)
(f)
(e)
Fig. 3. TEM images of (a-b) the pure CeO2 support, (c-d) the Rh/CeO2 and (e-f) the
Rh@CeO2 catalyst samples
245 C
228 C
189 C
186 C
Rh@CeO2
Rh/CeO2
426 C
579 C
158 C
271 C
518 C
552 C
CeO2
361 C
133 C
0
100 200 300 400 500 600 700 800
Temperature / C
Fig. 4. H2-TPR profiles of catalyst samples.
27

100 C
150 C
200 C
250 C
2012
2078
2150
2170 2140
1976
2194
Rh@CeO2
Rh/CeO2
CeO2
2280 2240 2200 2160 2120 2080 2040 2000 1960
Wavenumber / cm-1
Fig. 5. The DRIFTS results of CO adsorption on different catalyst samples as a
function of temperature in transient mode.
100
(a)
90
80
70
60
50
40
30
CeO2
20
Rh/CeO2
10
Rh@CeO2
0
100 150 200 250 300 350 400 450 500 550 600
Reaction temperature / C
28

100
90
80
70
60
50
40
30
20
10
0
(b)
CeO2
Rh/CeO2
Rh@CeO2
200 250 300 350 400 450 500 550 600
Reaction temperature / C
Fig. 6. Conversion curves of (a) CO and (b) C3H6 in three-way catalysis on different
catalysts as a function of temperature. Reaction conditions: 1050 ppm CO, 560 ppm
NO, 1800 ppm C3H6, 1000 ppm C3H8, 450 ppm CH4, 8% CO2, 5% H2O and 1.56% O2
in N2; λ=1.007; Sv = 96000 mL·g–1·h–1.
100
90
80
70
60
50
40
30
20
10
0
100
90
80
70
60
50
40
30
20
10
0
(a)
N2
NO
550
NH3
N2O
250
300
350
400
450
500
600
Reaction temperature / C
29