Physicochemical properties of nanostructured Pd/lanthanide-doped ceria spheres with high catalytic activity for CH4 combustion

Article abstract of DOI:10.1039/c8ta00203g

In this work, nanostructured Ce_0.9Gd_0.1O_2-δ (GDC) and Ce_0.9Pr_0.1O_2-δ (PrDC) spheres previously obtained by microwave assisted hydrothermal homogeneous co-precipitation were impregnated with 1 wt% Pd by the incipient wetness impregnation by an aqueous Pd²⁺ solution. Their properties were characterized by synchrotron radiation X-ray diffraction (SR-XRD), X-ray absorption near-edge spectroscopy (XANES) and scanning and high resolution electron microscopy with X-ray spectroscopy. Spherical particles with average diameters around 200 nm were found to consist of crystallites of average size, 10 nm, with small particles of PdO finely dispersed over the sphere surface. In situ XRD and XANES experiments were carried out under reducing and oxidizing conditions in order to investigate the redox behaviour of these materials and to evaluate the effect of Pd loading on the oxidation state of Ce. All of the lanthanide-doped ceria (LnDC) supports were found to have a cubic crystal structure (Fm3m space group). An increase in lattice parameters was observed under reducing conditions which was attributed to the reduction of Ce⁴⁺ ions to the larger Ce³⁺ ions, and to the associated increase in oxygen vacancy (V_O) concentration. Addition of Pd to the LnDC spheres increased their Ce³⁺ content. Finally, catalytic tests for CH₄ combustion were performed on the LnDC and Pd/LnDC nanocatalysts. The best performance was observed in samples with 1 wt% Pd loading, which exhibited T₅₀ values (temperature at which 50% of CH₄ conversion was reached) close to 310 °C. These values are 220 °C and 260 °C lower than those obtained for nanostructured PrDC and GDC spheres alone, respectively.

Full text of DOI:10.1039/c8ta00203g

Journal of
Materials Chemistry A
View Article Online
View Journal
PAPER
Physicochemical properties of nanostructured Pd/
lanthanide-doped ceria spheres with high catalytic
Cite this: DOI: 10.1039/c8ta00203g
activity for CH combustion†
4
Rodolfo O. Fuentes, *ab Leandro M. Acuna, A. Gabriela Leyva, Richard T. Baker,
bc
ad
e
˜
f
f
f
Huiyan Pan, Xiaowei Chen and Juan J. Delgado-Jaen
´
In this work, nanostructured Ce0.9Gd0.1O2ꢀd (GDC) and Ce0.9Pr0.1O2ꢀd (PrDC) spheres previously obtained
by microwave assisted hydrothermal homogeneous co-precipitation were impregnated with 1 wt% Pd by
2+
the incipient wetness impregnation by an aqueous Pd solution. Their properties were characterized by
synchrotron radiation X-ray diffraction (SR-XRD), X-ray absorption near-edge spectroscopy (XANES) and
scanning and high resolution electron microscopy with X-ray spectroscopy. Spherical particles with
average diameters around 200 nm were found to consist of crystallites of average size, 10 nm, with
small particles of PdO finely dispersed over the sphere surface. In situ XRD and XANES experiments were
carried out under reducing and oxidizing conditions in order to investigate the redox behaviour of these
materials and to evaluate the effect of Pd loading on the oxidation state of Ce. All of the lanthanide-
doped ceria (LnDC) supports were found to have a cubic crystal structure (Fm3m space group). An
increase in lattice parameters was observed under reducing conditions which was attributed to the
4+
3+
O
reduction of Ce ions to the larger Ce ions, and to the associated increase in oxygen vacancy (V )
3+
concentration. Addition of Pd to the LnDC spheres increased their Ce content. Finally, catalytic tests
for CH combustion were performed on the LnDC and Pd/LnDC nanocatalysts. The best performance
was observed in samples with 1 wt% Pd loading, which exhibited T50 values (temperature at which 50%
Received 7th January 2018
Accepted 29th March 2018
4
DOI: 10.1039/c8ta00203g
ꢁ
ꢁ
ꢁ
of CH conversion was reached) close to 310 C. These values are 220 C and 260 C lower than those
4
rsc.li/materials-a
obtained for nanostructured PrDC and GDC spheres alone, respectively.
Introduction
Their applications, such as in three-way catalysts, oxygen gas
sensors and electrodes and electrolytes for solid oxide fuel cells,
mean that ceria-based mixed oxides are materials with a large
a
Instituto de Nanociencia y Nanotecnolog ´ı a Departamento de F ´ı sica, Centro At ´o mico
Constituyentes, CNEA, Av. Gral. Paz 1499, (1650) San Mart ´ı n, Buenos Aires, Argentina
b
CONICET, Buenos Aires, Argentina. E-mail: rofuentes@conicet.gov.ar
c
DEINSO (Departamento de Investigaciones en S ´o lidos), UNIDEF-CITEDEF, J.B. de la technological impact, especially in the area of environmental
1–4
Salle 4397, (1603) Villa Martelli, Buenos Aires, Argentina
d
protection. A vast range of catalytic applications make use of
the reversible oxygen storage ability of ceria-based materials,
Escuela de Ciencia y Tecnolog ´ı a, Universidad Nacional de San Mart ´ı n, M de Yrigoyen
100, (1650) San Mart ´ı n, Buenos Aires, Argentina
3
4+
3+
a phenomenon which is closely connected to the Ce /Ce
e
EaStChem, School of Chemistry, University of St. Andrews, North Haugh, St. Andrews,
Fife, KY16 9ST, UK
5
redox couple.
f
Many of these applications rely on the high oxygen ion
Departamento de Ciencia de los Materiales e Ingenier ´ı a Metal u´ rgica y Qu ´ı mica
Inorg a´ nica, Universidad de C a´ diz, Campus R ´ı o San Pedro, Puerto Real, C a´ diz, conductivity and oxygen storage capacity which can easily be
Spain
achieved in ceria by doping with other aliovalent cations, such
†
Electronic supplementary information (ESI) available: Compositional as rare earth elements (e.g. Gd, Pr, Tb). This creates oxygen
proles and electron image (HAADF) and corresponding distribution maps
vacancies which enable migration of oxygen ions through the
of the elements in nanostructured 1 wt% Pd/PrDC spheres are presented in
6
–8
lattice.
ꢁ
Fig. S1 and S2, SR-XRD patterns recorded at 500
C under different
0
Under reducing atmospheres, both electronic ðCe CeÞ and
atmospheric conditions for nanostructured 1 wt% Pd/LnDC spheres are
ꢂꢂ
given in Fig. S3–S7 and their corresponding structural parameters and ionic ðV Þ charge carriers can be created through loss of oxygen
O
standard Rietveld agreement factors are presented in Tables S1 and S2. The ions from the lattice as molecular oxygen. The presence of both
4
performance (obtained using eqn (1)) of the catalysts for CH oxidation
types of charge carriers gives rise to mixed (ionic and electronic)
conducting behaviour in both undoped and doped cerias (e.g.
Lanthanide-doped cerias, LnDCs) under reducing conditions.
This is a difficulty for the application of LnDCs in SOFC
reported in the literature is presented in Table S3 and results of cyclic
catalytic stability tests performed on wt% Pd/LnDC spherical
nanocatalysts are presented in Fig. S8–S10 and Tables S4–S5. See DOI:
0.1039/c8ta00203g
1
1
This journal is © The Royal Society of Chemistry 2018
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
electrolytes – where electronic conductivity would give rise to an nanopowders prepared by two different methods (cation
internal short circuit of the electrolyte – but would be an complexation and incipient wetness impregnation of an
2
+
19
advantage for the use of these doped cerias in anodes for IT- aqueous Pd solution into Ce0.9Gd0.1
O
2ꢀd
nanopowders).
SOFCs. For example, the use of Gd-doped ceria in anode Catalysts with the higher Pd loading (5 wt%) exhibited the best
materials in intermediate temperature (IT)-SOFCs has been performance for CH combustion. The inuence of preparation
proposed in several studies in this and related areas, especially method was evident for catalysts with 1 wt% Pd, cation
4
9,10
on its use as a catalyst support. The increasing use of such complexation resulting in more active catalysts than the
materials as catalyst and fuel cell components has prompted wetness impregnation method, mainly due to the differences in
a renewed interest in their preparation in the form of nano- Pd distribution.
structured materials with controlled morphologies and high
surface areas.
The main objective of the present study is to focus on the
correlation between structural, redox and catalytic properties of
Hydrocarbons are proposed as fuels for SOFCs. Among the Pd supported on nanostructured ceria materials with
hydrocarbons, methane (CH ), being the main component of a controlled morphology and doped with either a univalent
natural gas, is of particular importance. Direct, electrochemical (Gd ) or a multivalent (Pr
conversion of methane in a fuel cell would represent a signi- nanostructured Ce0.9Gd0.1 2ꢀd (GDC) and Ce0.9Pr0.1
cant improvement in efficiency, and so a dramatic reduction in spheres previously obtained by microwave assisted hydro-
CO production, compared with conventional methane thermal homogeneous co-precipitation (HMW), as reported by
combustion, for electricity generation.
4
3
+
3+,4+
) cation. In the present work,
2ꢀd (PrDC)
O
O
2
11–14
20
Mu n˜ oz et al., were impregnated with 1 wt% Pd by incipient
2
+
In recent years, substantial progress in the controlled wetness impregnation (WI) by an aqueous Pd solution. The
synthesis and catalytic applications of ceria-based nanocatalysts resulting samples, Pd/LnDC (1 wt% Pd/GDC and 1 wt% Pd/
has been reported. Recent developments in ceria-based nano- PrDC), were characterized by synchrotron radiation X-ray
materials have paid particular attention to the general synthetic diffraction (SR-XRD), X-ray absorption near-edge spectroscopy
strategies and to the control of morphology, surface defects and (XANES), extended X-ray absorption ne structure (EXAFS),
composition. These investigations were focused on design and scanning and high resolution transmission electron microscopy
application of novel, highly efficient nanoceria catalysts for (SEM and HRTEM) and energy dispersive X-ray spectroscopy
practical applications such as CO oxidation, preferential (EDS). The redox properties of the materials were studied using
oxidation of CO (PROX) and the water–gas-shi (WGS) reaction in situ XAS (XANES) and XRD experiments carried out under
as well as other energy-related applications in SOFCs, proton reducing and oxidizing atmospheres at temperatures up to
ꢁ
exchange membrane fuel cells (PEMFCs), and photocatalytic 500 C. Finally, catalytic tests for CH
and catalysed organic reactions.
4
combustion were per-
15
formed on the Pd/LnDC spherical nanocatalysts.
In order to improve methane oxidation catalysts, a supra-
molecular approach has been reported in which single units
composed of a palladium (Pd) core and a ceria (CeO ) shell are
2
Experimental
preorganized in solution and then homogeneously deposited Nanostructured GDC and PrDC spheres were synthesized by
16
18
onto a modied hydrophobic alumina. Enhanced metal– HMW. Ce(NO ) $6H O (99.99%, Alfa Aesar), Gd(NO ) $6H O
3
3
2
3 3
2
support interactions led to exceptionally high methane oxida- (99.9%, Alfa Aesar) and Pr(NO ) $6H O (99.9%, Alfa Aesar) were
3
3
2
ꢁ
tion activity, with complete combustion below 400 C and employed as precursors. Each nitrate was dissolved in pure
outstanding thermal stability under demanding conditions. deionized H O separately and then the solutions were mixed to
2
Recently, S. Xie et al. reported the high activity of three- obtain a 0.1 M nitrate solution with molar ratios of Ce : Gd and
dimensionally ordered macroporous CeO -supported Pd@Co Ce : Pr appropriate for the preparation of GDC and PrDC,
2
nanoparticles for methane oxidation. The excellent catalytic respectively. Urea was added in order to give a molar ratio of
performance was associated with their ability to adsorb oxygen urea : nal oxide of 4 : 1 and 60 mL of the resulting solution was
and methane as well as with the unique core–shell structure of placed in a Teon-lined autoclave. A Milestone ETHOS 1
17
the Co@Pd nanoparticles.
Advanced Microwave Digestion system was employed. The
In a previous study, the authors reported on the synthesis sealed autoclave was placed in the oven and was heated to
ꢁ
ꢁ
ꢀ1
and catalytic properties of nanostructured 1 wt% Pd/Ce0.9
-
120 C with a ramp rate of 30 C min – with oven power set to
18
Gd0.1
O2ꢀd tubes. Results suggested that Pd cations, most likely 550 W – and then kept at this temperature for 1 h. Aer cooling,
the white powder product was collected by centrifugation
was stabilized against reduction in this phase. However, (2.5 min at 7830 rpm) and dried at 37 C for 12 h. Aer calci-
incorporation of the Pd (1 wt%) into the crystal lattice of the nation at 500 C in air for 1 h, the nanostructured LnDC spheres
nanostructured tubes also appeared to destabilize Ce , were obtained. These were impregnated with 1 wt% Pd by WI of
favoring reduction to Ce and giving rise to a signicant an aqueous Pd(NO ) solution. All resulting catalysts were dried
increase in the reducibility of this material. Preliminary cata- at 100 C and calcined at 500 C for 1 h in air, in order to
lytic studies showed a slightly improved catalytic activity toward decompose any unreacted nitrate precursor.
2
+
2+
Pd , formed a Pd–Ce–Gd oxide solid solution and that the Pd
ꢁ
ꢁ
4
+
3
+
3
2
ꢁ
ꢁ
CH
4
oxidation compared to pure Ce0.9Gd0.1
O
2ꢀd (either nano-
In order to verify the phase composition, conventional X-ray
tubes or nanopowders). In subsequent work, the authors re- diffraction (XRD) was performed in a PANalytical Empyrean 2
3
D
ported the catalytic properties of 1 and 5 wt% Pd/Gd0.1Ce0.9
O
2ꢀd with a PIXcell
detector employing Cu-K
a
radiation (of
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
wavelength 1.5418 A). Data in the angular range 2q ¼ 20–90
Journal of Materials Chemistry A
˚
ꢁ
D04B-XAFS1 beamline at LNLS in transmission mode using
were collected in step-scanning mode, with a step length of a Si(111) monochromator for the Ce L -edge (5726 eV) and Pr L -
3
3
ꢁ
0
.04 and a step-counting time of 4 s. The average crystallite edge (5960 eV). The nominal photon ux of the beamline is 3 ꢃ
9
size, DXRD, of the products was determined using the Scherrer 10 photons per (s mrad 100 mA)@6 keV. All spectra were
formula from the extent of peak broadening of the main XRD collected at energies in the range 5690–6100 eV (for the Ce L
3
-
19
reection, (111). Errors in crystallite size were derived by edge and Pr L
estimating the error in the FWHM (full-width at half-maximum) the Ce L -edge and Pr L
to be equal to the 2q step. (5989 eV). Two acquisitions were made on the same sample to
3
-edge) with E/DE ¼ 5000 to 10 000. Energies for
3
3
-edge were calibrated using a Cr foil
Nitrogen adsorption–desorption isotherms were obtained improve the signal to noise ratio. Samples were diluted with
using an Autosorb-1 instrument from Quantachrome. Out boron nitride and these mixtures were pressed into 15 mm
ꢁ
gassing was carried out at 200 C for 16 h prior to the diameter pellets (around 6 mg of sample and 70 mg of diluent
measurements. Analysis of the isotherms using the BET method were used). For the transmission measurements, the pellets
provided values of the specic surface area (SSA) of the samples. were placed in a tubular quartz furnace (diameter, 20 mm; X-ray
SEM images were obtained in secondary electron mode path length, 440 mm) sealed with refrigerated Kapton windows.
using a JEOL 6700F instrument with Field Emission Gun (FEG) Temperature was measured and controlled by a thermocouple
and the X-ray spectrometer attached to this microscope was passed down the sample holder and positioned close to the
used to record EDS spectra. Values of accelerating voltage (in surface of the pellet. Temperature-resolved XAS spectra at the
kV) and working distance (WD) are given in the images. Samples Ce L -edge and Pr L -edge were acquired during temperature
3
3
were imaged in the FEG-SEM uncoated.
programmed reduction (TPR) under 5% H /He (total ow: 20
2
ꢀ1
ꢁ
TEM images were recorded using a JEOL JEM 2011 instru- mL min ) at temperatures from 25 to 500 C at a heating rate of
ꢁ
ꢀ1
ment operating with a LaB
6
lament at an accelerating voltage 10 C min and with a total data acquisition time of 20 min per
ꢁ
of 200 kV and equipped with an Oxford Instruments EDS spectrum. Aer the data were collected at 500 C under 5% H
spectrometer. The images were captured using a Gatan CCD He, the system was purged with N
camera and analyzed using Digital Micrograph 3.4.4 soware. synthetic air (21% O /N ; total ow: 50 mL min ) was passed
2
/
ꢀ
1
2
(100 mL min ) and
ꢀ1
2
2
This soware was also used to calculate Digital Diffraction through the furnace. Aer 10 min, data were collected under
Patterns (DDPs) by fast Fourier transform of regions of interest these oxidizing conditions.
of the TEM images. The image analysis soware, ImageJ, was
Temperature-programmed reduction experiments in
used to obtain particle size information from the TEM images. A hydrogen (H -TPR) were carried out using a Quadruple Mass
2
Titan Themis 200 keV (FEI) transmission and scanning trans- Spectrometer (PrismaPlus® QMA 220 M). Before each TPR
mission electron microscope (S/TEM) equipped with an X-FEG measurement, ca. 0.1 g of the sample was loaded into a quartz
2
Schottky eld emission gun and a spherical aberration xed-bed U-shaped micro-reactor and pretreated in 5% O /Ar
ꢀ
1
ꢁ
corrector was employed to obtain electron images in High Angle ow of 60 mL min at 500 C for 1 h. Aer cooling in the same
ꢁ
Angular Dark Field (HAADF) mode. Its Super-X high sensitivity atmosphere to 150 C, the gas was switched to a ow of He and
windowless EDS detector was used to generate line scans and the reactor was cooled to RT. The pretreated sample was
ꢀ
1
two-dimensional maps of the elemental composition of the exposed to a ow (60 mL min ) of a 5% H /Ar gas mixture and
2
ꢁ
ꢁ
ꢀ1
samples. For TEM examination, the samples were suspended in heated from RT to 900 C at a ramp rate of 10 C min
.
acetone by ultrasonication and deposited onto holey carbon-
coated Cu grids.
Catalytic tests were performed in a quartz U-tube reactor
equipped with a furnace and temperature controller. All the
XRD patterns were recorded at high temperatures (HT) and experiments were carried out with 75 mg of catalyst diluted with
in controlled atmospheres using synchrotron radiation at the 150 mg of SiC. Before methane combustion tests, the samples
ꢀ
1
D10B-XPD beamline of the National Synchrotron Light Labo- were pretreated in 20% O
2
–Ar gas mixture (60 mL min ) at
ꢁ
ratory (LNLS, Campinas, Brazil). In these in situ XRD experi- 450 C for 1 h to remove the water and carbonates adsorbed on
ments, the sample was mounted on a ceramic sample-holder the sample surface. Aer cooling to room temperature, the
and placed in a furnace. The X-ray wavelength was set at reactor was purged with pure Ar for 30 minutes to eliminate
˚
ꢁ
1
.54892 A. Data in the angular range 2q ¼ 20–100 were weakly absorbed O
2
. The reaction mixture – of composition: 2.1
ꢁ
ꢀ1
ꢀ1
ꢀ1
collected in step-scanning mode, with a step length of 0.04 and mL min of CH
4 2
, 11.6 mL min of O and 61.3 mL min of
a step-counting time of 2 s. The data were collected at temper- Ar-was owed through the reactor at room temperature for
ꢁ
ꢁ
atures ranging from room temperature to 500 C. The sample 30 min aer which the reactor was heated to 900 C at a ramp
ꢁ
ꢀ1
ꢁ
ꢀ1
was heated at a rate of 10 C min , and a soak time of 10 min rate of 10 C min . A blank experiment was performed on
was employed at each temperature step before performing the 150 mg of SiC under the same reaction conditions as a blank
XRD scan. The thermal and redox behavior of the materials was experiment. A Quadruple Mass Spectrometer (PrismaPlus®
ꢀ
1
studied in 5% H /He (total ow: 20 mL min ) and in dry QMA 220 M) was used to collect data on reactant and product
2
ꢀ
1
synthetic air (total ow: 50 mL min ). NIST SRM 640c Si concentrations in the gas leaving the reactor. Cyclic catalytic
powder was used as the standard for the instrumental broad- stability tests were performed in the same experimental setup
ening correction.
and under the same conditions. Each cycle consisted of
ꢁ
ꢁ
In situ XAS (XANES) experiments under conditions of increasing the temperature from 125 C to 500 C at a ramp rate
controlled temperature and atmosphere were carried out at the of 10 C min and maintaining the system at 500 C for 2 h.
ꢁ
ꢀ1
ꢁ
This journal is © The Royal Society of Chemistry 2018
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
The reaction rates were calculated using the methane signal and magnication images indicate that these are composed of many
the following eqn (1):
much smaller particles. In the high resolution images these are
seen to be nanocrystals of around 5–10 nm diameter. For
example, the crystallite circled in Fig. 3d can be indexed to the
cubic uorite structure of ceria viewed along the [211] zone axis.
The nominal compositions of the GDC and PrDC supports were
conrmed by EDS in the SEM and TEM instruments and the
presence of the Pd at around 1 wt% was also conrmed.
The spatial distributions of the constituent elements of the
Pd/GDC material were obtained by recording EDS line scans
and maps such as those presented in Fig. 4 for a group of three
spherical particles. These show coincident distributions of Ce,
O and Gd, indicating the successful synthesis of the GDC mixed
oxide as a single phase. These distributions also match the
grayscale variations seen in the HAADF images, which are
directly related to sample density. The Pd signals also show that
this element is quite evenly distributed through the GDC
support material. However, the peaks seen in the Pd traces in
the line scans in Fig. 4b and d indicate that at least a proportion
of the Pd is present in the form of discreet Pd-rich nanoparticles
of a few nm in diameter. A similar set of maps and line scans are
given in Fig. 5 for the 1 wt% Pd/PrDC sample. The Ce, O and Pr
distributions again closely match with each other as well as with
the grayscale variations seen in the HAADF images, conrming
that the PrDC single phase has also been formed successfully.
The Pr La peak overlaps with some Ce peaks so in this gure the
Pr Ka peak, which does not overlap but which gives a weaker
response, was used to provide a map which is independent of
the Ce distribution (Fig. 5e). Again, Pd is present throughout the
support material but the localized concentrations of Pd
observed in the Pd elemental map (Fig. 5c) and implied in the
peaks seen in the line scans (Fig. 5g and h) indicate the pres-
ence of discreet Pd-rich nanoparticles. This is what would be
expected to result from successful impregnation by incipient
wetness of the Pd precursor onto previously prepared GDC or
PrDC spheres. Further HAADF images, line scans and elemental
maps are presented in Fig. S1 and S2 in ESI.†
F
CH4 ꢃ XCH4
rate ¼
(1)
1
00 ꢃ WPd
ꢀ1
where FCH4 is the molar ow rate of methane in mmols. min
Pd is the mass of Pd in grams and XCH4 is the methane
,
W
conversion in percent.
Results and discussion
The morphology and chemical compositions of the nano-
structured 1 wt% Pd/LnDC spheres were studied by SEM and
TEM. The SEM images in Fig. 1 conrmed that both samples
showed a very high yield of the expected uniform spherical
particles with a narrow size distribution. TEM images at a range
of magnications are presented for 1 wt% Pd/GDC and 1 wt%
Pd/PrDC in Fig. 2 and 3, respectively. Again there is very little
extraneous material, the large majority of the particles imaged
being regular spheres. These were measured from a number of
TEM images and found to have diameters of 205 ꢄ 36 (one
standard deviation) nm and 189 ꢄ 24 nm for the 1 wt% Pd/GDC
and 1 wt% Pd/PrDC samples, respectively. The contrast varia-
tions across the spherical particles in the medium
A crystallographic study was performed by Rietveld rene-
ment of the SR-XRD data employing the FullProf suite of so-
22
ware. For the cubic phase, the Fm3m space group was
3
+
3+
4+
2ꢀ
assumed, with (Gd , Pr , Ce ) cations and O anions in 4a
and 8c positions, respectively. The peak shape was assumed to
be a pseudo-Voigt function. The background of each prole was
n

tted using a six-parameter polynomial function in (2q) , n ¼ 0–
5
. The thermal parameters were assumed to be equal. A second
phase corresponding to PdO (space group P4 /mmc with Pd
2
and O in 2d and 2e positions, respectively, according to ICDS#
4692), was included. No diffraction lines corresponding to Pd
2
+
2
ꢀ
2
or PdO were observed in the SR-XRD pattern collected at room
temperature for 1 wt% Pd/GDC or 1 wt% Pd/PrDC (Fig. 6a and b,
respectively). In the same gure, the Rietveld tted pattern (line)
and the difference plot are included. Even a scan with a larger
counting time in the regions where the Pd(111) and PdO(111)
peaks would be expected did not show any evidence of metallic
Pd or PdO. However, a small difference between experimental
data and the tted line close to where the PdO(111) peak would
be expected was observed, when the PdO phase is not included
Fig. 1 FEG-SEM images of (a) 1 wt% Pd/GDC and (b) 1 wt% Pd/PrDC
samples.
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
Journal of Materials Chemistry A
Fig. 2 TEM images of 1 wt% Pd/GDC at increasing magnifications (a–d). The partial EDS spectrum inset in (d) is of a large area containing many
spherical particles and confirms the presence of Ce, Gd and a small concentration of Pd.
in the renement (see ESI, Fig. S3†). The results of Rietveld used for data obtained at room temperature. Moreover, for the
renement of the SR-XRD data are summarized in Table 1 and case of the reducing atmosphere, a second phase corresponding
0
are accompanied by reliability indices to judge the tting to Pd (space group Fm3m with Pd in the 4a position, according
quality. These indices are weighted R (Rwp), the reduced chi- to ICDS# 41517), was included. The results of Rietveld rene-
2
squared (c ), and R
e
, which are related just to the prole of ment of this SR-XRD data are summarized in Table 3 (and ESI,
, which is related to the crystal struc- Fig. S3 and S4†).
the XRD patterns, and R
ture. The reduced chi-squared is dened as (Rwp/R ) , where R
is the index that should be analyzed to verify if the renement is 500 C under reducing conditions (5% H /He), the lattice
converging and R is the expected statistical value for Rwp.
In Table 2, average crystallite size (DXRD), specic surface expansion of the unit cell and the partial reduction of Ce and Pr
area (SSA) and the primary particle size (dBET), calculated from in the solid solutions. On switching to oxidizing conditions
p
2
When the samples were heated from room temperature to
e
wp
ꢁ
2
23
parameters increased in all samples, both due to the thermal
e
21
ꢁ
the BET data, are summarized for all samples aer calcination (synthetic air) at 500 C, the lattice parameters were seen to
at 500 C. Both Pd-loaded nanocatalysts had similar SSA values, decrease for all samples, but to different extents. Clearly, the
ꢁ
which were slightly lower than those of pure GDC and PrDC unit cell parameter is strongly dependent on the concentration
2
ꢀ1
20
(
57.2 and 60.2 m g , respectively). The dBET/DXRD ratio in of the cation dopant as well as on the gas atmosphere and
both spheres was 1.6, indicating that the crystallites exhibited temperature. However, no substantial differences can be
a low degree of agglomeration. These results indicate that the observed between lattice parameters obtained under the same
thermal treatment used to impregnate the Pd did not strongly conditions for LnDC and 1 wt% Pd/LnDC (see ESI, Tables S1
20
affect the spherical morphology, the crystallite size or the SSA of and S2†).
these materials. Fig. 7 shows an expanded area of the SR-XRD patterns in
From SR-XRD data collected at 500 C under reducing (5% which the Pd(111) and PdO(111 and 200) reections are clear.
/He) and oxidizing (21% O /N ) conditions, the lattice Under reducing conditions, the main peak of metallic Pd is
ꢁ
H
2
2
2
parameters were rened assuming a cubic phase (Fm3m space clearly observed in 1 wt% Pd/GDC and 1 wt% Pd/PrDC samples
group) by the Rietveld method with the same assumptions as (Fig. 7a). On the other hand, under oxidizing conditions at
This journal is © The Royal Society of Chemistry 2018
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
Fig. 3 TEM images of 1 wt% Pd/PrDC at increasing magnifications (a–d). The partial EDS spectrum (i) is of a large area containing many spherical
particles and confirms the presence of Ce, Pr and a small concentration of Pd. The DDP inset (ii) was obtained from the crystallite circled and can
be indexed to ceria viewed along the [211] zone axis.
ꢁ
5
00 C, the main peaks of PdO alone were observed in both
Data analysis was conducted by least-squares tting four
samples (Fig. 7b, and ESI, Fig. S6 and S7†) showing clearly that Gaussian proles and one arctangent function to the experi-
the metallic Pd was reoxidised. These results indicate that, at mental XANES data in the range between 5710 and 5750 eV. The
3
+
this temperature and in these samples, palladium changes ratio between the area of peak C (associated with Ce ) and the
0
II
easily from Pd under reducing conditions to Pd under sum of the areas of peaks A, B and C (A and B are associated with
4
+
oxidizing conditions.
Ce ) gives direct information about the fraction of Ce present
3+
19
Normalized Ce L -edge XANES spectra for the nano- as Ce . All values were normalized to the fraction of Ce
structured 1 wt% Pd/GDC and 1 wt% Pd/PrDC spheres obtained present as Ce in pure CeO2 at room temperature. In both
at 500 C under reducing and oxidizing conditions, are pre- samples (Table 4), in air, no large differences were observed in
sented in Fig. 8. These exhibit two clear peaks frequently the fractions of Ce /(Ce + Ce ) between room temperature
labelled A and B. Peak A is assigned as a Ce peak with the nal and 500 C. In contrast, under reducing conditions at 500 C,
state 2p 4f 5d , which denotes that an electron is excited from peak C (Fig. 8) is clearly larger – and the percentage of Ce
the Ce 2p shell to its 5d shell, with no electron in the 4f shell. calculated is dramatically higher – than in air at the same
Peak B is also a Ce peak, with the nal state 2p 4f 5d v, which temperature.
denotes that in addition to an electron excited from the Ce 2p The negative values in Table 4 simply indicate that the Ce
shell to the 5d shell, another electron is also excited from the content of the 1 wt% Pd/PrDC samples is lower than in the
valence band (O 2p shell) to the Ce 4f shell, leaving a hole (v) in reference material, pure, undoped CeO . As expected, in both
3
3
+
ꢁ
3
+
4+
3+
4
+
ꢁ
ꢁ
5
0
1
3+
4
+
5
1
1
3
+
2
3
+
24
3+
the valence band. Some authors refer to peak C as a Ce peak.
Gd- and Pr-doped samples, only small amounts of Ce are seen
An additional small peak (D) is present at pre-edge and likely in the oxidizing environment (both at room temperature and
ꢁ
arises from transitions to the bottom of the conduction band. 500 C) and these levels increase considerably on exposing the
ꢁ
The Ce L
3
-absorption edges for all samples are close to samples to H
2
at 500 C. In a previous study, an increase in
5
725.7 eV (values were determined from the rst and second reducibility on doping pure ceria with Gd or Pr (so introducing
3
+
derivatives of the Ce L -edge XANES spectra for each sample).
oxygen ion vacancies) was observed, with a Ce content of 7.3
3
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
Journal of Materials Chemistry A
Fig. 4 Electron images (HAADF) and corresponding elemental profiles
along the lines indicated for two (a, b) and one (c, d) 1 wt% Pd/GDC Fig. 5 Electron (HAADF) image (a) and corresponding distribution
spheres; distribution maps of the elements Gd (e), Ce (f), Pd (g) and O maps of the elements Ce (b), Pd (c), O (d) and Pr (e) in a single Pd/PrDC
(
h) for the same group of spheres.
sphere; HAADF image of two 1 wt% Pd/PrDC spheres (f) with elemental
profiles along the lines indicated for each (g, h).
20
and 8.4% for GDC and PrDC, respectively. A signicant
increase in reducibility caused by the incorporation of Pd was
observed in the present work for 1 wt% Pd/GDC (17.3%) and
of hydrogen from Pd nanoparticles is known to occur easily,
25–27
even at low temperatures.
Therefore, reduction via spillover
from the (reduced) Pd nanoparticles is likely to dominate over
direct reduction of the oxide. This explains the increase in
reducibility on going from LnDC to 1 wt% Pd/LnDC materials.
1
wt% Pd/PrDC (15.0%). Two mechanisms for reduction of the
Ce are possible. The rst one would involve the direct adsorp-
tion of H on the oxide, followed by reaction and release of
2
Fig. 9 presents the normalized XANES spectra at the Pr L
3
-
water. The second would involve reduction of PdO to Pd with
subsequent adsorption of H on the Pd and spillover of H
2
species onto the oxide surface, where these could reduce the
oxide and again form water. Evidence of facile reduction of PdO
to Pd is given in the SR-XRD data in Fig. 7a and b and spillover
ꢁ
edge obtained at 500 C under reducing (5% H
2
/He, 20
ꢀ
1
mL min ) and subsequent oxidizing (synthetic air: 21% O /N ,
2
2
ꢀ1
5
0 mL min ) conditions. At rst sight, it could be assumed that
3
+
4+
the extent of re-oxidation (seen as the Pr switching to Pr in
This journal is © The Royal Society of Chemistry 2018
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
Table
3
Structural parameters and standard Rietveld agreement
factors for SR-XRD patterns of 1 wt% Pd/GDC and 1 wt% Pd/PrDC
ꢁ
samples taken at 500 C under reducing and oxidizing conditions
2
˚
Sample
Pd/GDC
Pd/PrDC
Atmosphere
a (A)
R
p
R
wp
R
e
c
5% H
2
/He
1% O /N
/He
/N
5.4514(5)
5.4416(7)
5.4667(5)
5.4457(8)
3.18
3.03
3.27
3.32
3.88
3.68
4.09
4.13
2.44
2.49
2.34
2.4
2.53
2.19
3.05
2.95
2
2
2
5% H
2
21% O
2
2
content, but they are related to the maximum number of oxygen
2
0,29
vacancies allowed for the structure.
The proles of H
2
O formation (m/z ¼ 18) in the H
2
-TPR
experiments are shown in Fig. 10. Two broad peaks are observed
ꢁ
at 540–563 and 757–778 C for CeO and LnDC (Fig. 10a–c). These
can be attributed to the reduction of Ce to Ce on the surface
and in the bulk of these materials, respectively.
2
4
+
3+
30–32
In the traces
for 1 wt% Pd/GDC (Fig. 10d) and 1 wt% Pd/PrDC (Fig. 10e), these
two peaks were also observed. However, at 215–238 C a further
Fig. 6 Synchrotron XRD patterns recorded at room temperature
ꢁ
(
empty circles) with the Rietveld-fitted pattern (red line) and the
peak is clearly observed. Since the reduction of the support
materials, in the absence of Pd, starts at much higher tempera-
tures (above 380 C for GDC and PrDC), this reduction peak can
difference plot for (a) 1 wt% Pd/GDC and (b) Pd/PrDC spheres.
ꢁ
be assigned to the reduction of PdO to Pd. However, it is well
Table
1 Structural parameters and standard Rietveld agreement
factors for SR-XRD patterns of 1 wt% Pd/GDC and 1 wt% Pd/PrDC known that noble metals such as Pd can promote (via hydrogen
samples taken at room temperature in air
spillover) the reduction of the support. Therefore, some contri-
bution to these peaks from such Pd-promoted reduction of the
support – once enough metallic Pd is present – cannot be dis-
2
˚
a (A)
Sample
R
p
R
wp
R
e
c
counted. Indeed, the “bridges” seen in the TPR spectra between
Pd/GDC
Pd/PrDC
5.4192(6)
5.4197(6)
3.36
2.77
4.18
3.59
2.56
2.40
2.73
ꢁ
2.22 the initial peaks at 215/238 C and those expected for the support
ꢁ
materials (from 380 C upwards) are likely to be largely due to this
process because the PdO particles, being so small, would be ex-
15
pected to be reduced quickly.
oxidizing atmospheres) of these samples is rather low, since the
In order to compare their catalytic activity, several tests were
performed on the LnDC and 1 wt% Pd/LnDC spherical nano-
catalysts, as explained in the Experimental section. For
comparison, a blank test was carried out using SiC only. In
4
+
white line assigned to Pr is rather small. However, there are
studies that show that even in XANES spectra of samples con-
taining (nominally) 100% of the Pr in the +4 state, the white line
4
+
for Pr is not so pronounced, and that the presence of any peak
4
Fig. 11, CH conversion versus temperature for the different
4
+
28
should be taken as indicative of signicant Pr content.
catalyst samples is plotted. In addition, T , T and T90 values
1
0
50
3
+
4+
Hence, our samples show signicant levels of both Pr and Pr
at 500 C in oxidizing conditions. On the other hand, the spectra
obtained under reducing conditions show that the Pr is
completely reduced to Pr . This is in agreement with the in situ
SR-XRD results, where the moderate increase in lattice param-
eter can be ascribed to the increase in concentration of both
ꢁ
3
+
3
+
4+
Pr and the related oxygen vacancies. The reduction of Ce
also contributes to a lesser extent to this unit cell expansion. As
the authors previously showed in a study of another multivalent
3
+
3+
rare earth oxide, Ce content is not independent of Pr
Table 2 Average crystallite size (DXRD), specific surface area (SSA) and
calculated primary particle size (dBET) for nanostructured 1 wt% Pd/
GDC and 1 wt% Pd/PrDC spheres
2
ꢀ1
Sample
D
XRD/nm
SSA/m g
dBET/nm
dBET/DXRD
Fig. 7 Synchrotron HT-XRD patterns of 1 wt% Pd/GDC and 1 wt% Pd/
Pd/GDC
Pd/PrDC
9.7
10.4
54.9
53.7
15.5
16.3
1.6
1.6
PrDC samples in the vicinity of the Pd(111), PdO(111) and PdO(200)
peaks at 500 C under (a) reducing and (b) oxidizing conditions.
ꢁ
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
Journal of Materials Chemistry A
ꢁ
Fig. 9 Normalized Pr-L
3
edge XANES spectra obtained at 500
C
under reducing (red line) and oxidizing (blue line) conditions. Spectra
3+
3 3
at room temperature for Pr(NO ) (Pr ; dotted line) is included for
comparison in this figure.
caused by impregnation of the LnDC by 1 wt% Pd is remarkable,
ꢁ
the decrease in T50 being about 250 C.
Fig. 8 Normalized Ce L
obtained at 500 C in (a) reducing and (b) oxidizing conditions; and
3
-edge XANES spectra for 1 wt% Pd/GDC
ꢁ
The reaction rates at 300 C was estimated from eqn (1) for the
ꢁ
1
wt% Pd/GDC and 1 wt% Pd/PrDC samples, as well as for some
ꢁ
1
wt% Pd/PrDC obtained at 500 C in (c) reducing and (d) oxidizing
previous results reported by the authors in the literature, per-
conditions, showing the experimental data (empty circles), four
19
Gaussian peaks (A–D), and one arctangent function obtained by least- formed under the same conditions of the present work. The
squares fitting and the sum of all five functions (continuous black lines). comparison indicates that the samples studied in the present
work show the highest reaction rate per palladium atom (about
ꢀ
1
ꢀ1
2
9 mmols gPd min ) in comparison with the other palladium-
were estimated from the experimental curves and are given in
Table 5.
ꢀ1
ꢀ1
based catalysts (about 7 mmols gPd min ) (see ESI, Table S3
and Fig. S8†). Here it is important to note that it is not possible to
compare directly with other palladium-based catalysts because
The catalytic activity of the bare supports was signicantly
better than that obtained in the blank test, and PrDC performed
slightly better than GDC. The addition of Pd dramatically
increased the activity of both samples, although the effect of the
support dopant is very small. The temperature at which 50%
16,17
the experiments were carried out under different conditions.
4
conversion of CH was reached (T50) was used as an index of
catalytic activity. The blank test exhibited the highest value of
ꢁ
T
50 (800 C). The 1 wt% Pd/GDC and 1 wt% Pd/PrDC samples
exhibited very similar behaviour, and gave the lowest values of
ꢁ
T50 (about 310 C) and, therefore, were by far the most active
catalysts under these conditions. The improvement in T50
3+
Table 4 Fraction of Ce present as Ce in 1 wt% Pd/GDC and 1 wt%
ꢁ
Pd/PrDC samples at 25 C in air and under reducing and oxidizing
ꢁ
conditions at 500 C, estimated from fittings to the Ce L
3
-edge XANES
spectra
ꢁ
3+
4+
3+
Sample
Pd/GDC
Temperature ( C)
Atmosphere
Ce /(Ce + Ce ) %
25
21% O
2
/N
2
0.5
5
5
25
00
00
5% H /He
2
17.3
0.4
0.6
21% O
21% O
2
2
/N
/N
2
Pd/PrDC
2
5
5
00
00
5% H
21% O
2
/He
/N
15.0
ꢀ0.7
Fig. 10
GDC and (e) 1 wt% Pd/PrDC.
H
2
-TPR profiles of (a) CeO
2
, (b) GDC, (c) PrDC, (d) 1 wt% Pd/
2
2
This journal is © The Royal Society of Chemistry 2018
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
4
Table 5 Temperature at which 10, 50 and 90% conversion of CH was
reached (T10, T50 and T90, respectively), estimated from the experi-
mental curves (Fig. 11). The values for 1 wt% Pd/GDC and 1 wt% Pd/
st
th
PrDC correspond to for the 1 and 4 cycle
ꢁ
ꢁ
ꢁ
T90 ( C)
Sample
T
10 ( C)
T
50 ( C)
Blank
GDC
PrDC
772
490
478
282
310
282
320
800
570
530
310
355
310
367
801
659
604
324
446
336
437
st
1
1
1
1
wt% Pd/GDC (1 )
wt% Pd/GDC (4 )
wt% Pd/PrDC (1 )
wt% Pd/PrDC (4 )
th
st
th
Fig. 11 Catalytic activities of CH
4
combustion for 1 wt% Pd/GDC (red)
The most likely causes of catalyst deactivation are obstruc-
and 1 wt% Pd/PrDC (blue) spherical nanocatalysts as a function of tion or modication of active sites because of coke deposition,
temperature. Nanostructured GDC (dotted red) and PrDC (dotted
blue) spheres and non-catalytic sample (blank) are included for
reaction conditions used and the thermal sintering – and loss of
comparison.
modication of the oxidation state of the Pd under the specic
specic surface area – of the Pd-rich particles. Because of the
large oxygen excess used in these experiments, it is very unlikely
that the catalysts suffer deactivation or poisoning by carbon
The fact that the spherical support particles are composed of
4
+
3+
deposits. In addition, a temperature programmed oxidation
nanoparticles of doped cerias containing one (Ce /Ce ) or two
th
4
+
3+
4+
3+
experiment was carried out aer the 4 cycle and a negligible
(
Ce /Ce , Pr /Pr ) accessible redox couples and a high
2
amount of CO was produced. Reduction of palladium oxide
oxygen vacancy concentration is very likely to signicantly
increase the rate of transport of oxygen to the active sites in
these catalysts and so improve their activity for oxidation reac-
tions, as is seen in the data described above. The addition of the
Pd function further dramatically increases catalytic activity.
Dual-site mechanisms, where the nanoparticle–support inter-
face plays a crucial role in activating the reactants, are
commonly accepted to be important in the oxidation of
methane over catalysts containing noble metals supported on
can be achieved even at high concentrations of oxygen, but at
16
higher temperatures than those employed here. Even though
it was no possible to completely characterize the size and
morphology of the Pd-rich nanoparticles, it is reasonable to
assume that the activity loss is related to the sintering of the Pd-
rich nanoparticles and possibly also to changes in their
morphology that would affect the metal–support interaction
and the chemical environment of the supercial Pd atoms.
It is well known that ceria-based mixed oxides act as oxygen
storage materials, or oxygen buffers, in oxidation reactions
33
reducible oxides. In the catalysts of the present work, the
interactions between the Pd-rich phase and the support are
likely to be important because of the intimate relationship
between the well-dispersed nanoparticles of the Pd-rich phase
and the nanoparticulate support. This may explain the high
catalytic activity observed. Unfortunately, the thickness and
porosity of the spherical support particles make it impossible to
carry out a full analysis of the size of the Pd-rich nanoparticles
or of the nature of the metal–support interface. However, from
the microscopy and other data presented here, it is expected
that the microstructure of the support is such that it facilitates
the dispersion of the Pd-rich phase in the form of nanoparticles,
and therefore increases the total area of the interface between
the active phase and the support, as well as the total length of
the boundary between Pd-rich phase, ceria-based support and
the reactants in the gas phase (i.e. the nanoparticle edges).
Either or both of these phenomena may be important in
explaining the excellent catalytic activity observed here.
3
+
4+
because of the accessible Ce /Ce redox couple. Reduction of
4
+
3+
Ce to Ce is accompanied by loss of oxygen from the oxide,
which can be used in the reaction. It is commonly accepted that
ceria-based materials follow the Mars Van Krevelen mechanism
in which the catalyst acts as the oxygen donor towards the
reactant and this oxygen is replenished from the oxygen present
in the gas phase and which thus acts as an indirect reactant, so
34
keeping the catalyst in its oxidised, active form. In the present
work, the catalytic tests were carried out in excess oxygen, so
this mechanism was favoured.
3
+
In pure GDC, Gd ions introduce extrinsic oxygen vacancies,
3
+
which in turn diminish the Ce content in the material from
35
a simple equilibrium point of view. The GDC then shows
4
+
a higher Ce content than pure ceria (under the same external
conditions), and, it is proposed that, as a consequence, it displays
higher catalytic activity. The higher oxygen vacancy concentration
would also tend to allow faster loss and uptake of oxygen from the
gas phase, so improving the redox kinetics. T50 data for pure ceria
Results of cyclic catalytic stability experiments performed on
wt% Pd/LnDC spherical nanocatalysts show a deactivation
aer four cycles (see ESI, Fig. S9 and S10†). From the rst to the
1
19
and pure GDC were reported in a previous study.
ꢁ
ꢁ
In Pd-containing materials, the exact catalytic mechanism
for hydrocarbon oxidation is not fully understood, and neither
are the species that are involved. Nevertheless, there is a general
fourth cycles, the values of T50 increased from 310 C to 355 C
ꢁ
ꢁ
for 1 wt% Pd/GDC and from 310 C to 367 C for 1 wt% Pd/PrDC
Table 5 and ESI Tables S5 and S6†).
(
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
Journal of Materials Chemistry A
consensus that the active phase is probably PdO or an equiva- Tecnol ´o gica (Argentina, PICT 2012-1506). Electron Microscopy
lent intermediate, but that metallic Pd shows worse or no was performed by RTB at the Advanced Functional Materials
36–38
catalytic activity at all.
There is also evidence for an Research Centre (funded through EPSRCs Capital for Great
enhancement effect conferred by ceria-related materials on the Technologies grant scheme) at the University of St Andrews. The
39,40
PdO active phase and hence on the catalytic activity.
Even in authors are grateful to Fernando Mu n˜ oz, Anna Paula da Silva
the case of catalytic tests using unsupported Pd, it has been Sotero Levinsky, Cristiane Rodella, F ´a bio Zambello, Tamiris
shown that the active phase for hydrocarbon oxidation is PdO Bouças Piva and Simone Ba u´ Betim for their invaluable exper-
41
and not metallic Pd.
imental assistance at the LNLS and Gabriela Numma for gures
In order to further understand the physicochemical properties and graphics design. Dr L. M. Acu n˜ a and Dr R. O. Fuentes are
of these catalysts and the catalytic mechanism, a future detailed members of CIC-CONICET, Argentina. Dr J. J. Delgado thanks
study using theoretical ab initio density functional theory the “Ram ´o n y Cajal” program of the Ministry of Economy,
(DFT+U) calculations is planned by the authors. The chemical Industry and Competitiveness of Spain. Dr R. O. Fuentes
and structural properties of active sites on the Pd–ceria surface gratefully acknowledges the UCA-International fellowship
are difficult to characterize, especially under reaction conditions. (UCA/R82REC/2016), Universidad de Cadiz, Spain.
Strong interactions between palladium and the ceria support may
d+
stabilize oxidized Pd species, which may contribute to the
signicant activity of Pd/ceria based materials for methane
oxidation. Of special interest is the question of whether calcula-
References
tions from rst principles predict the presence of oxygen vacan-
cies under the reaction conditions employed in this study and, if
so, where they are located. This is a very important issue since
oxygen vacancies play a central role in the transport properties
within these materials during redox reactions.
1 R. Di Monte and J. Kaspar, Top. Catal., 2004, 28, 47–57.
2 S. D. Park, J. M. Vohs and R. J. Gorte, Nature, 2000, 404, 265–
267.
3 S. M. Haile, Mater. Today, 2003, 6, 24–29.
4 B. H. C. Steele, Nature, 1999, 400, 619–621.
5
J. Kaspar and P. Fornasiero, in Catalysis by Ceria and Related
Materials, ed. A. Trovarelli, Imperial College Press, London,
Conclusions
2002, pp. 217–236.
In the present work, nanostructured Ce0.9Gd0.1
O
(GDC) and
6 J. A. Kilner, Solid State Ionics, 2000, 123, 13–23.
7 B. C. H. Steele, Solid State Ionics, 2000, 129, 95–110.
8 H. L. Tuller, S. R. Bishop, D. Chen, Y. Kuru, J.-J. Kim and
T. S. Stefanik, Solid State Ionics, 2012, 225, 194–197.
9 A. A. Yaremchenko, A. A. Valente, V. V. Kharton,
I. A. Bashmakov, J. Rocha and F. M. B. Marques, Catal.
Commun., 2003, 4, 477–483.
2ꢀd
Ce0.9Pr0.1
O
(PrDC) spheres previously obtained by microwave
2ꢀd
assisted hydrothermal homogeneous co-precipitation were
impregnated with 1 wt% Pd by the incipient wetness impreg-
2
+
nation of an aqueous Pd solution. Their properties were
characterized by synchrotron radiation X-ray diffraction (SR-
XRD), X-ray absorption near-edge spectroscopy (XANES) and
scanning and high resolution electron microscopy with X-ray 10 S. P. Jiang, X. J. Chen, S. H. Chan and J. T. Kwok, J.
spectroscopy.
Electrochem. Soc., 2006, 153(5), A850–A856.
The addition of 1 wt% Pd to the LnDC dramatically increased 11 R. T. Baker, I. S. Metcalfe, P. H. Middleton and
the reducibility of the Ce in the mixed oxide. This was demon-
B. C. H. Steele, Solid State Ionics, 1994, 72, 328–333.
strated by analysis of in situ XANES spectra obtained under 12 R. T. Baker and I. S. Metcalfe, Appl. Catal., A, 1995, 126, 297–
3
+
4+
reducing conditions. Clearly, the Pd improves the Ce /Ce
redox couple in LnDC materials.
317.
13 R. T. Baker and I. S. Metcalfe, Appl. Catal., A, 1995, 126, 319–
High activity for CH
4
oxidation was observed in nano-
332.
structured 1 wt% Pd/LnDC spheres with total conversion of CH
attained at a temperature below 400 C.
4
14 S. Tao, J. T. S. Irvine and S. M. Plint, J. Phys. Chem. B, 2006,
110, 21771–21776.
ꢁ
These ndings open up an interesting avenue for future work 15 K. Wu, L.-D. Sun and C.-H. Yan, Adv. Energy Mater., 2016, 6,
in this area, and indicate great promise for application of these
particular catalyst systems.
1–7.
16 M. Cargnello, J. J. Delgado Ja ´e n, J. C. Hern ´a ndez Garrido,
K. Bakhmutsky, T. Montini, J. J. Calvino G ´a mez, R. J. Gorte
and P. Fornasiero, Science, 2012, 337, 713–717.
7 S. Xie, Y. Liu, J. Deng, X. Zhao, J. Yang, K. Zhang, Z. Han and
H. Dai, J. Catal., 2016, 342, 17–26.
Conflicts of interest
1
There were no conicts of interest in this work.
1
8 F. F. Mu n˜ oz, M. D. Cabezas, L. M. Acu n˜ a, A. G. Leyva,
R. T. Baker and R. O. Fuentes, J. Phys. Chem. C, 2011, 115,
Acknowledgements
19687–19696.
This work has been supported by: the Brazilian Synchrotron 19 F. F. Mu n˜ oz, R. T. Baker, A. G. Leyva and R. O. Fuentes, Appl.
Light Laboratory (LNLS, Brazil), under proposals D04B-XAFS1- Catal., B, 2013, 136–137, 122–132.
3435 and D12A-XRD1-13437; Chinese Scholarship Council 20 F. F. Mu n˜ oz, L. M. Acu n˜ a, C. Albornoz, A. G. Leyva,
1
(
CSC, China); and Agencia Nacional de Promoci ´o n Cient ´ı ca y
R. T. Baker and R. O. Fuentes, Nanoscale, 2015, 7, 271–281.
This journal is © The Royal Society of Chemistry 2018
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
2
1 H. Klug and L. Alexander, in X-ray Diffraction Procedures for 32 L. Jiang, S. Fern ´a ndez-Garc ´ı a, M. Tinoco, Z. Yan, Q. Xue,
Polycrystalline and Amorphous Materials, John Wiley, New
York, 1974, p. 618.
G. Blanco, J. J. Calvino, A. B. Hungr ´ı a and X. Chen, ACS
Appl. Mater. Interfaces, 2017, 9(22), 18595–18608.
2
2
2
2
2
2
2 J. Rodr ´ı guez-Carvajal, FullProf Suite Program, Version 2.05. 33 I. X. Green, W. Tang, M. Neurock and J. T. Yates Jr, Science,
Laboratoire Le ´o n Brillouin, CEA-CNRS, Saclay, France, 2011.
2011, 333, 736–739.
3 R. A. Young, in The Rietveld Method, ed. R. A. Young, Oxford 34 J. L. G. Fierro, Metal Oxides Chemistry, 2006, ch. 8, p. 217.
University Press, Oxford, 1993, ch. 1, pp. 21–24.
4 F. Zhang, P. Wang, J. Koberstein, S. Khalid and S.-W. Chan,
Surf. Sci., 2004, 563, 74–82.
5 J. M. Gatica, R. T. Baker, P. Fornasiero, S. Bernal and
J. Ka ˇs par, J. Phys. Chem. B, 2001, 105, 1191–1199.
6 A. Norman and V. Perrichon, Phys. Chem. Chem. Phys., 2003,
35 U. Hennings and R. Reimert, Appl. Catal., B, 2007, 70, 498–
508.
36 S. Colussi, A. Trovarelli, E. Vesselli, A. Baraldic, G. Comelli,
G. Groppi and J. Llorca, Appl. Catal., A, 2010, 390, 1–10.
37 S. Colussi, A. Trovarelli, G. Groppi and J. Llorca, Catal.
Commun., 2007, 8, 1263–1266.
5
, 3557–3564.
38 A. Hellman, A. Resta, N. M. Martin, J. Gustafson, A. Trin-
chero, P. A. Carlsson, O. Balmes, R. Felici, R. van Rijn,
J. W. M. Frenken, J. N. Andersen, E. Lundgren and
H. Gr ¨o nbeck, J. Phys. Chem. Lett., 2012, 3, 678–682.
7 F. C. Gennari, A. Carbajal Ramos, A. Cond ´o , T. Montini,
S. Bengi ´o , A. Cortesi, J. J. Andrade Gamboa and
P. Fornasiero, Appl. Catal., A, 2011, 398, 123–133.
2
8 I. A. Sluchinskaya, A. I. Lebedev and A. Erko, Phys. Solid 39 L. M. T. Simpl ´ı cio, S. T. Brand ˜a o, D. Domingos, F. Bozon-
State, 2012, 54(5), 975–979. Verduraz and E. A. Sales, Appl. Catal., A, 2009, 360, 2–7.
9 L. M. Acu n˜ a, F. F. Mu n˜ oz, C. Albornoz, A. G. Leyva, 40 L. Meng, J.-J. Lin, Z.-Y. Pu, L.-F. Luo, A.-P. Jia, W.-X. Huang,
2
R. T. Baker and R. O. Fuentes, J. Mater. Chem. A, 2015, 3,
6120–16131.
M.-F. Luo and J.-Q. Lu, Appl. Catal., B, 2012, 119–120, 117–
121.
1
30 J. Kaspar, P. Fornasiero and M. Graziani, Catal. Today, 1999, 41 B. G. Hound, H. A. E. Hagelin, J. F. Weaver and
5
0, 285–298.
1 Y. Zuo, X. Huang, L. Li and G. Li, J. Mater. Chem. A, 2013, 1,
74–380.
G. N. Salaita, Appl. Surf. Sci., 2003, 205, 102–112.
3
3
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2018