Partial oxidation of methane over bimetallic copper- and nickel-actinide oxides (Th, U)

Article abstract of DOI:10.1016/j.jallcom.2010.03.021

The study of partial oxidation of methane (POM) over bimetallic nickel- or copper-actinide oxides was undertaken. Binary intermetallic compounds of the type AnNi₂ (An = Th, U) and ThCu₂ were used as precursors and the products (2NiO·

Full text of DOI:10.1016/j.jallcom.2010.03.021

Journal of Alloys and Compounds 497 (2010) 249–258
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: www.elsevier.com/locate/jallcom
Partial oxidation of methane over bimetallic copper- and nickel-actinide
oxides (Th, U)
Ana C. Ferreiraa, A.P. Gonc¸ alvesa, T. Almeida Gaschea, A.M. Ferrariab, A.M. Botelho do Regob,
M.R. Correiac, A. Margarida Bolac, J.B. Brancoa,∗
a Instituto Tecnológico e Nuclear, Unidade de Ciências Químicas e Radiofarmacêuticas, Estrada Nacional 10, 2686-953 Sacavém, Portugal
b Universidade Técnica de Lisboa, IST, Centro de Química-Física Molecular and IN, Av. Rovisco Pais, 1049-001 Lisboa, Portugal
c I3N-Universidade de Aveiro, Department Física, Aveiro, Portugal
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 6 January 2010
Received in revised form 26 February 2010
Accepted 1 March 2010
Available online 6 March 2010
The study of partial oxidation of methane (POM) over bimetallic nickel- or copper-actinide oxides
was undertaken. Binary intermetallic compounds of the type AnNi2 (An = Th, U) and ThCu2 were used
as precursors and the products (2NiO·UO3, 2NiO·ThO2 and 2CuO·ThO2) characterized by means of X-
ray diffraction, X-ray photoelectron spectroscopy, Raman spectroscopy and temperature-programmed
reduction. The catalysts were active and selective for the conversion of methane to H2 and CO and sta-
ble for a period of time of ≈18 h on stream. The nickel catalysts were more active and selective than
the copper catalyst and, under the same conditions, show a catalytic behaviour comparable to that of
a platinum commercial catalyst, 5 wt% Pt/Al2O3. The catalytic activity increases when uranium replaces
thorium and the selectivity of this type of materials is clearly different from that of single metal oxides
and/or mechanical mixtures. The good catalytic behaviour of the bimetallic copper- and nickel-actinide
oxides was attributed to an unusual interaction between copper or nickel oxide and the actinide oxide
phase as showed by H2-TPR, XPS and Raman analysis of the catalysts before and after reaction.
© 2010 Elsevier B.V. All rights reserved.
Keywords:
Intermetallic
Bimetallic actinide oxides
Methane
Partial oxidation
Synthesis gas
1. Introduction
mixtures [11] and the behaviour of a number of Ni-containing cat-
alysts, such as NiO–MgO [12], NiO–CaO [13], NiO–rare earth oxides
The existence of large deposits of natural gas and the rising of
the crude oil price have stimulated numerous R&D efforts to con-
vert methane (main natural gas component) into more valuable
and easily transportable chemical products. Methane is also a by-
product formed in several industrial processes and a hydrocarbon
air pollutant that has a much larger detrimental greenhouse effect
than carbon dioxide [1]. Therefore, most studies on the utilization of
natural gas concentrate almost exclusively on methane activation.
For example, the partial oxidation of CH4 (POM) has been stud-
ied as an alternative to other processes such as steam reforming to
produce CO and H2 (synthesis gas) that can be directly used as feed-
stock for the synthesis of methanol or for Fischer–Tropsch process
[2–5].
The catalysts reported to be active for the POM reaction are
either noble metal (like Ir, Pt, Pd, Rh and particularly Ru) or
Ni-based compounds [6–10]. Despite the high activity of noble
metal-based catalysts, the high cost of such systems limits their
extensive industrial application. Ni-based catalysts exhibit good
activity and selectivity to synthesis gas formation from CH4/O2
[14–17], NiO/Al2O3 [18–25], Ni with noble metals (Ru, Rh, Pt, Ir)
[6,8–10,26–28], have been extensively studied. However, the major
drawbacks of this reaction are the deactivation caused by sintering,
changes in the oxidation state of the metal active phase and carbon
deposition [2,3,5,29].
To our knowledge, the behaviour of actinide oxide-based cat-
alysts in the partial oxidation of methane was only studied by
Choudhary et al. [30]. They reported high methane conversions
(>90 and >71% on NiO–ThO2 and NiO–UO2, respectively) and high
selectivity to H2 and CO on nickel oxide catalysts supported by
conventional impregnation techniques on actinide oxides at 700 ◦C.
In our laboratories, we have been using binary intermetal-
lic compounds LnCu2 (Ln = La, Ce, Pr, Nd) [31], ThCu2 and AnNi2
(An = U, Th) as catalytic precursors of bimetallic oxides, e.g.
3CuO·Ln2CuO4 or 2CuO·ThO2 [32] that exhibited selectivity for the
4-methylpentan-2-ol decomposition [31]. After reduction, these
bimetallic oxides were described as supported copper catalysts on
lanthanide oxides, 2Cu·CeO2 and 4Cu·Ln2O3, which were active and
selective for the mesityl oxide hydrogenation to 4-methylpentan-
2-one (methylisobutylketone, MIKB) [33]. Their catalytic activity
and selectivity were associated with the lanthanide-containing
phase that seems to play an important role in the formation of the
copper or nickel active sites.
∗
Corresponding author. Tel.: +351 219946116; fax: +351 219941455.
E-mail address: jbranco@itn.pt (J.B. Branco).
0925-8388/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2010.03.021

250
A.C. Ferreira et al. / Journal of Alloys and Compounds 497 (2010) 249–258
In this article, we report the catalytic performance and stability
of bimetallic nickel- and copper-actinide oxide catalysts for par-
tial oxidation of methane in the temperature range 350–850 ◦C at
atmospheric pressure. The catalysts were prepared by controlled
oxidation of the AnNi2 (An = Th, U) and ThCu2 and characterized by
powder X-ray diffraction (XRD), X-ray photoelectron spectroscopy
(XPS) and temperature-programmed reduction (H2-TPR).
2. Experimental
2.1. Catalysts preparation
The actinide intermetallic compounds, AnNi2 (An = Th, U) and ThCu2, were pre-
pared and characterized by powder X-ray diffraction as previously described [34].
The bimetallic actinides oxides (2CuO·ThO2, 2NiO·ThO2 and 2NiO·UO3) were pre-
pared by controlled oxidation of the intermetallic compounds under air (Air Liquide,
O2:N2 = 20:80 (vol%), purity 99.995%) at 10 ◦C/min heating rate up to 1000 ◦C, as
described elsewhere [31].
2.2. Catalysts characterization
Fig. 1. Influence of the temperature on the conversion of methane over nickel- or
copper-actinides oxides (An = Th, U).
Powder X-ray diffraction (XRD) patterns were obtained on a PANalitycal X’Pert
Pro diffractometer using monochromatized Cu k␣ radiation (ꢀ = 1.5406 Å). The oper-
ational settings for all scans were voltage = 45 kV; current = 40 mA; 2ꢁ scan range
5–80◦ using a step size of 0.03◦ at a scan speed of 0.02◦/min. For identification
purposes, the relative intensities (I/I0) and the d-spacing (Å) were compared with
standard JCPDS powder diffraction files [35]. The average particle size was deter-
mined by means of the Scherrer’s equation [36].
The X-ray photoelectron spectroscopy (XPS) measurements were performed
in a spectrometer XSAM800 (KRATOS). Operating conditions are described else-
where [37]. A Shirley background was subtracted in all regions and curve fitting for
component peaks was carried out using Gaussian–Lorentzian products. No charge
compensation (flood gun) was used. The binding energies were corrected using as
a reference the binding energy (BE) of sp3 carbon at 285 eV. X-ray source satellites
were subtracted. For quantification purposes, sensitivity factors were 0.66 for O 1s,
0.25 for C 1s, 4.16 for Cu 2p3/2, 3.53 for Ni 2p3/2, 5.76 for Th 4f7/2 and 11.23 for U 4f.
The Raman spectra (Stokes) were obtained at room temperature in back scatter-
ing configuration with a Jobin-Yvon LabRaman HR equipped with a Multichannel air
cooled (−70 ◦C) CCD detector. An objective of 40× magnification was used to focus
the surface of the sample pellet excited with a 325 nm He–Cd laser line.
defined as the number of mL of methane converted per g of catalyst and per hour
(mLCH4/g h). The conversion of methane, the selectivity and the yield of the products
were calculated as follows: Conv.CH4 (%) = {([CH4]i − [CH4]o)/[CH4]i} × 100; Sel.CO
(%) = {[CO]o/([CH4]i − [CH4]o)} × 100; Sel.CO2 (%) = {[CO2]o/([CH4]i − [CH4]o)} × 100
and Sel.H2 (%) = {[H2]o/2 × ([CH4]i − [CH4]o)} × 100, where [CH4]i are inlet flow rates,
and [CH4]o, [CO]o, [CO2]o, and [H2]o are outlet flow rates. The amount of reagents
and products was confirmed by an external standard method using reference mix-
tures of CH4, CO, CO2 and H2 (Air Liquide). The confidence level was better than 95%.
Unless stated otherwise, the values reported in this paper represent the initial activ-
ities of the catalysts after 1 h on stream. The catalysts activity was compared to that
of a commercial catalyst (5.0 wt% Pt/Al2O3, Aldrich), one standard in this research
field [41].
3. Results and discussion
3.1. Catalysts activity and stability
The temperature-programmed reduction (H2-TPR) experiments were per-
formed under pure hydrogen in an apparatus previously described [32,33]. The
samples were placed in a quartz U tube (6 mm diameter), then submitted to a flow
of hydrogen (50 cm3/min) under atmospheric pressure, and heated up to 950 ◦C at a
rate of 10 ◦C/min. The degree of reduction was determined by the quantitative anal-
ysis of water, the reduction product, as described elsewhere [38]. This procedure
allows to enhance the sensitivity of the method as well as to perform TPR exper-
iments under pure hydrogen [39]. Prior to H2-TPR, all samples were pretreated at
150 ◦C under He (F = 50 cm3/min) during 10 min to eliminate physisorbed water.
Optimized resolution was obtained by careful choice of the sample weight (≤30 mg)
and hydrogen flow (F = 50 cm3/min) taking into account the criteria established by
Ballivet-Tkatchenko and Delahay [40] for the temperature-programmed reduction
of V2O5 that we have also confirmed for the reduction of CuO. The TPR curves became
distorted, flat at their maximum and there is a broadening of the peaks for K > 20 s
Fig. 1 shows the effect of the temperature on the performance of
the copper- or nickel-actinide oxides as catalyst for POM in the tem-
perature range of 350–800 ◦C and under a molar ratio of CH4/O2 = 2.
For comparison purposes, the behaviour of a commercial platinum
catalyst, 5 wt% Pt/Al2O3, was also studied under the same experi-
mental conditions.
The conversion of CH4 increases with the reaction temperature.
The maximal activity at 800 ◦C, expressed as percentage of methane
converted, is as follows: NiO (46%), CuO (25%), 2CuO·ThO2 (20%), 5
wt% Pt/Al2O3 (90%), 2NiO·ThO2 oxide (93%) and 2NiO·UO3 oxide
(95%), which for the late two compounds was 2 times superior to
that of pure NiO. Moreover, the behaviour of the nickel-actinide
oxides was comparable to that of a commercial catalyst of platinum
supported on alumina, 5 wt% Pt/Al2O3.
(P > 3 ◦C) and the optimal conditions comprise between 10 and 20 s and 1.5 and 3 ◦
C
for K and P, respectively. Under our experimental conditions, the calculated K and
P values were ≈15 s and ≈3 ◦C. Water was analyzed on a Shimadzu 9A instrument
equipped with thermal conductivity detector (TCD), helium was used as carrier gas
(F = 50 cm3/min, TDetector = 105 ◦C, 150 mA), connected to a CR3 Shimadzu integrator.
Fig. 2 shows the evolution of the catalysts selectivity to hydrogen
and carbon dioxide with the temperature of reaction. The produc-
tion of H2 increases with the increase of temperature, whereas the
production of CO2 decreases significantly with the temperature
and becomes residual at T ≥ 750 ◦C. The other products detected
were carbon monoxide that increases with the temperature and
C2 hydrocarbons (ethylene and ethane) that were only detected
on 2CuO·ThO2 at T ≥ 650 ◦C and at low concentrations (≤6%). The
results obtained on 2NiO·ThO2 and 2NiO·UO3 were again very simi-
lar to those obtained on the platinum commercial catalyst and very
different from those obtained over pure metal oxides (CuO and NiO)
and 2CuO·ThO2 (please, see results at 800 ◦C).
2.3. Catalytic activity and stability
The catalytic partial oxidation of methane was carried out at atmospheric
pressure in a fixed-bed U-shaped cylindrical quartz reactor (10 mm i.d.) plug-
flow type reactor, with a quartz frit containing an quartz tube (45◦ angle, 2 mm
i.d., touching the quartz frit) equipped with a thermocouple for the measure-
ment of the reaction temperature. The cross-sectional area of the reactor was
0.7854 cm2, and the catalyst bed was <1 mm high. Mass flow controllers were used
to control CH4 (Air Liquide, purity 99.9995%), air (Air Liquide, purity 99.9995%)
and He (Air Liquide, purity 99.9995%) flows. Unless otherwise stated, a gaseous
mixture with a CH4/O2 molar ratio of 2 [CH4 (28%), air (O2 14%, N2 58%, total
flow = 3.6 l/h) was used and the reaction studied with an adequate Gas Hourly
Space Velocity (GHSV, mL of CH4/g of catalyst.h) of 8500, >12000 mL of CH4/mL
of catalyst.h, m = 100 mg. The initial empty reactor conversion was less than 3%,
typically 0.7% at 923 K furnace temperature. The outlet gas was cooled in an ice-
water trap and its composition analyzed on-line by gas chromatography using
a Restek ShinCarbon ST column (L = 2.0 m, ꢂ = 1/8 in., ID = 1 mm, 100/200 mesh)
and an Agilent 4890D GC equipped with a thermal conductivity detector (TCD)
and a 6-port gas sampling valve with a 0.250 ␮L loop. Catalyst activity (ri) was
Yaoqiang et al. [42] claimed that these results could be explained
by the adsorption characteristics of CO and H2. CO strongly adsorbs
on the catalysts surface where easily combines with adsorbed
oxygen species to produce CO2 at low temperature. Moreover,
desorption of CO increases with the temperature, which lead an

A.C. Ferreira et al. / Journal of Alloys and Compounds 497 (2010) 249–258
251
Fig. 3. Partial oxidation of methane over 2NiO·UO3 as a function of time on stream
at 800 ◦C.
capability for oxygen storage, which can help the oxidation of the
carbon deposited on the catalyst surface. There are also evidences
than the presence of rare earth oxides can stabilize and prevent the
catalysts from sintering during the POM high-temperature reac-
tion [5]. Consequently, the environment and synergism between
copper or nickel and the actinide oxides seems to be a main factor
that influences the bimetallic actinide oxides activity, selectivity
and stability.
This raises the question about the nature of the active sites for
the catalytic POM over copper- and nickel-actinide oxides and the
roles played by copper, nickel and the actinide oxide phase on their
formation/behaviour. Nickel oxide, copper oxide and the actinide
oxides (ThO2, UO3) are catalytic active for the partial oxidation of
methane [14–17,48,49], which was confirmed by our results. How-
ever, the catalytic behaviour of CuO and NiO is clearly different
from that of the corresponding copper- or nickel-actinide bimetal-
lic oxides. As an example, Fig. 4 shows the comparison of the activity
of NiO and UO3 with that of the nickel–uranium bimetallic oxide.
The behaviour of a mechanic mixture of 2NiO and UO3 was also
included for comparison purposes.
The evolution of the selectivities with the reaction temperature
confirms such differences (Fig. 5). Therefore, if copper and nickel
oxide, on one hand, and the actinide oxide, on the other, have a
role in the catalytic process, the enhanced behaviour of the copper-
and nickel-actinide bimetallic oxides can only be ascribed to the
existence of an unusual interaction between the d and f metal oxide
phases.
Fig. 2. Influence of the temperature on the selectivity to H2 (a) and CO2 (b) over the
nickel- or copper-actinides oxides (An = Th, U).
increase of the selectivity to CO and to a decrease of the selectivity to
CO2. On the other hand, hydrogen easily desorbs from the catalysts
surface and the increase of the H2 selectivity with the temperature
is mainly due to the increase of the conversion of methane.
To clarify the lower activity on the copper catalysts we must
take into account that a complete dissociation of CH4 is much
more difficult on Cu than on Ni, the chemisorption of methane
on metal surfaces, e.g. Rh(1 1 1), Ni(1 1 1) and Cu(1 1 1), is exother-
mic on nickel and endothermic on copper [43], which explains not
only the lower activity and also the relatively high selectivity to C2
hydrocarbons of copper-based catalysts in the partial oxidation of
methane [44].
The bimetallic copper- and nickel-actinide oxide catalysts were
also stable for at least 18 h of time on stream at each tempera-
ture studied (500, 600, 700, 750 and 800 ◦C), which correspond to
an overall time of 96 h. Other authors have used periods of 6–16 h
to study the evolution of the catalysts activity/selectivity [15,29].
As an example, Fig. 3 shows the conversion of methane and the
selectivities to CO, H2 and CO2 over 2NiO·UO3 as a function of the
time on stream at 800 ◦C. No deactivation was observed during
18 h on stream. The conversion of methane was very stable at 95%,
whereas the selectivities to CO and H2 remain stable at 97 and 98%,
respectively. The formation of CO2 was residual.
It is generally recognized that the catalyst deactivation during
POM is caused by the deposition of carbon on the catalyst surface
[4,45,46]. Moreover, it is known that the addition of rare earth metal
oxide or alkaline metal oxide to alumina or the use of rare earth
metal oxide as support can restrict carbon deposition [5,11,47].
The promotion of the rare earth oxide is usually attributed to its
Fig. 4. Temperature effect on the conversion of methane over NiO, UO3, 2NiO·UO3
and a mechanic mixture of 2NiO + UO3.

252
A.C. Ferreira et al. / Journal of Alloys and Compounds 497 (2010) 249–258
Fig. 6. XRD patterns for the copper- and nickel-actinide oxides after reaction.
phases that could correspond to the formation of new solid solu-
tions between CuO/NiO and the actinide oxide phase were by no
means detected but, the appearance of Ni, Cu and Cu2O due to
nickel and copper oxide total or partial reduction could be iden-
tified by XRD. Moreover, the diffraction patterns of NiO and CuO
were no longer present on the XRD spectra and UO3 (U6+) was
also partial reduced to UO2 (U4+). The average particle size deter-
mined by the Scherrer’s equation was comparable to those obtained
before reaction (Cu and Ni particles, 61 1; ThO2 particles, 77
4
and UO2/UO3 particles, 29 8 nm). Elemental analysis at the end
of reaction shows carbon contents that are variable: residual in the
case of NiO, CuO, ThO2 and UO3 (≈0.1 wt%; 0.01 g/g cat), low in
the case of 2CuO·ThO2, 2NiO·UO3 and the mechanical mixture of
NiO + UO3 (≈3 wt%; 0.30 g/g cat) and high in the case of 2NiO·ThO2
(≈18 wt%; 1.75 g/g cat).
The H2-TPR profiles of the copper and nickel-actinides oxides
exhibit only one reduction peak in the temperature range studied
(20–1000 ◦C); except for 2NiO·UO3 that shows two distinct reduc-
tion peaks (Fig. 7).
The temperature at maximum mass loss rate (Tm) shifts
to higher temperature when U substitutes Th on the nickel-
actinide oxides (found Tm, 320 and 347/381 ◦C, for 2NiO·ThO2 and
2NiO·UO3, respectively) and the Tms found are higher than that of
the reduction of pure NiO, especially in the case of the bimetallic
nickel–uranium oxide (347/381 vs. 290 ◦C) and within the experi-
mental error ( 5 ◦C). This decrease of oxygen lability is less evident
in the case of the bimetallic copper–thorium oxide (found Tm, 225
Fig. 5. Temperature effect on the selectivity to H2 (a) and CO2 (b) over NiO, UO3,
2NiO·UO3 and a mechanic mixture of 2NiO + UO3.
To our knowledge, only Choudhary et al. [30] have studied
the behaviour of nickel catalysts supported on thorium oxide
(NiO/ThO2) and uranium oxide (NiO/UO2) for the partial oxida-
tion of methane. They have found that the performance of the
nickel–thorium catalysts is superior to that of the nickel–uranium
catalyst. At 800 ◦C their results are comparable to those obtained
in this work. However, the conversion of methane is higher on the
nickel-actinide bimetallic oxides, e.g. 95 and 79% over 2NiO·UO3
and NiO/UO2, respectively. The catalysts studied by Choudhary are
also less stable, which stress the advantage of using binary inter-
metallic compounds of the type AnNi2 (An = Th, U) and ThCu2 has
bimetallic oxides precursors.
3.2. Catalysts characterization
The XRD patterns obtained for the nickel- or copper-actinide
oxides were consistent with those of standard NiO, CuO, ThO2
and UO3 cubic phases as reported on the standard JCPDS powder
diffraction files [35], with a nearly absence of differences in the
respective lattice parameters. Therefore, the stoichiometry of the
nickel- or copper-actinide oxides can be depicted has 2NiO·ThO2,
2CuO·ThO2 and 2NiO·UO3 as reported earlier by Branco et al. [32].
For all the samples, the average particle size was close to each other
≈30–60 nm (CuO, NiO and ThO2 particles, 62 1; UO3 particles,
30 4 nm).
After reaction, important modifications occur on the bimetal-
lic copper or nickel-actinide oxides surface as evidenced by the
XRD diffraction patterns reported in Fig. 6. The formation of oxide
Fig. 7. H2-TPR profiles under pure hydrogen.

A.C. Ferreira et al. / Journal of Alloys and Compounds 497 (2010) 249–258
253
and 238 ◦C for the reduction of pure CuO and 2CuO·ThO2, respec-
tively). The Tms for the reduction of copper and nickel oxides (225
and 290 ◦C, respectively) are in agreement with published values
[31–33,50].
The H2-TPR quantitative analysis of the water produced on each
mass loss gives a H2O:An molar ratio of 1.9 0.1 for the reduc-
tion of 2NiO·ThO2 and 2CuO·ThO2, whereas the quantification of
the two reduction peaks observed for 2NiO·UO3 gives H2O:U total
molar ratio of 2.9 0.1 (2.4 and 0.5 for the first and second reduc-
tion peak, respectively). After the H2-TPR studies, the XRD patterns
of the reduced samples shows pure crystalline phase of metallic
copper or nickel along with ThO2 and UO2, in agreement with the
data already published [32].
To understand such results we must take in account that the
standard Gibbs free energy for the reduction of copper or nickel
oxide is considerably more favourable than that of the actinide
oxides (e.g. ꢃfGo = −16.87 kJ/mol for the reduction of NiO and
+711.63 kJ/mol for the reduction of ThO2) [51]. Therefore, a com-
mon step can be assigned to the reduction of the CuO or NiO
phase on the bimetallic copper- or nickel-actinide oxides. On the
other hand, the Tm associated with the reduction of such CuO and,
particularly, NiO species were higher than those obtained for the
reduction of pure metal oxides. Such difference can be attributed
to the existence of unusual stable copper and nickel oxide particles
(oxygen less labile) that in the case of the reduction of the bimetal-
lic copper- or nickel-actinide oxides seems to be determined by a
strong interaction with the actinide oxide phase that hinders their
reduction. Gervasini et al. [52] proposed the same explanation for
the reduction of nano-sized CuO dispersed on synthesized silica
modified with alumina, titania and zirconia.
As previously reported, bulk diffusion of nickel atoms in
lanthanum oxide perovskite type structure catalysts, LaNiO3, is
inhibited as a consequence of the physical barriers established by
the La2O3 particles [53]. The metallic Ni crystallites maintain a good
dispersion on the La2O3 surface and the good stability of the cata-
lyst is mainly due to the ability to stabilize the Ni crystallites in a
high dispersion degree on a La2O3 matrix, thus limiting the extent
of metallic particles sintering.
The same argument can be used to explain the good stability
of the copper- and nickel-actinide oxides. The interaction of Cu, Ni
and Cu2O particles generated along the reaction with the actinide
oxide phase may slow down the sintering of metal particles at high
reaction temperatures and the repressed formation of aggregates
of copper or nickel particles inhibit the formation of coke residues
that would encapsulate or cover metallic Cu or Ni particles, and
hence the deactivation of the catalyst [54,55].
Therefore, it is clear that bimetallic copper- or nickel-actinide
oxides CuO and NiO were partial or totally reduced to metallic cop-
per or nickel at the end of reaction. Furthermore, the catalysts were
only active at temperatures superiors to 550 ◦C, which correlated
the formation of Cu, Ni and Cu2O, as expected from the H2-TPR
results, pointing for a major role of such species. A strong interac-
tion between copper or nickel and the actinide in the bimetallic
copper- or nickel-actinide oxides certainly explains their lower
reducibility and the good catalytic stability.
(Vertical lines in Fig. 8(a)) [56,57] cannot be completely excluded
despite not being expected. FWHM are 2.9 and 3.3 eV for BR and
AR, respectively.
The Th 4f region (Fig. 8(b)) before reaction is composed by one
doublet with spin-orbit separation of 9.3 eV and FWHM of 1.9 eV. Th
4f7/2 component (Th 4f5/2 is not shown), centred at 334.2 0.1 eV, is
assigned to Th4+ in ThO2. Its absolute BE is closer to the BE reported
for thorium nitrate compounds [58], but the difference between the
BEs of the O 1s (peak centred at 530.2 eV in Fig. 1(c)) and Th 4f and
the relative positions of the satellites in Th 4f (+ 7.1 0.1 eV, not
shown) are typical of Th in thoria.
After reaction (AR), one can find two doublets with spin-orbit
separations of 9.2 0.1 eV and 2.1 0.2 eV distant from each other.
These two components cannot correspond to two different oxida-
tion state since no oxidation state higher than +4 is possible in
thorium. Thence, they evince the existence of two different phases
having different conductivities [59]. Th 4f7/2 components, 2.4 eV
FWHM, centred at 335.1 eV (phase 1 slightly higher than the one
existing before reaction) and 337.2 eV (phase 2, more resistive) are
both attributed to ThO2. This effect is not noticed in the Cu 2p region
where the Cu 2p3/2 has the same binding energies as before the
reaction meaning that phase 1 is a mixed phase but phase 2 contains
only thoria.
Before reaction, the main peak in O 1s, centred at 530.2 0.2 eV,
is assigned to oxygen bound to copper and to thorium. After reac-
tion (both in AR and SS), a structure compatible with the existence
of two phases, as already noticed in Th 4f region, is seen. Peak 1,
centred at 531.1 eV, is assigned to oxygen bound to copper (Cu2+
)
and thorium (Th4+) in phase 1; peak 2, centred at 533.1 eV, corre-
sponds to oxygen bound to thorium in phase 2. Above 535 eV, some
oxygen in phase 2 hydroxyl groups can be found in all samples, as
expected from Cu(OH)2 detected in Cu 2p [60].
Neither copper nor thorium is detected in SS, mostly due to
the huge amount of carbon existing at the surface (see Table 1
and Raman results). The carbon existing before reaction is mostly
aliphatic carbon (component centred at 285 eV). However, after
reaction, in particular in SS, where no copper and/or thorium
are detected, the peak centred at 285 eV may include not only
the contribution of aliphatic carbons, but also, sp2 carbon from a
carbonised overlayer that covers both the d metal and actinide.
Centred at higher BE (from ∼287 to 292 eV) one can find other con-
tributions from carbon bound to oxygen and/or carbon in a highly
electronegative neighbourhood as, for instance, in carbonates [60].
Concerning the nickel–thorium oxide sample, the XPS results
show that Ni 2p region exhibits a mixture of different types of nickel
oxides and hydroxides. Before reaction, Ni 2p3/2 was fitted with
two components, 2.0 0.1 eV FWHM, centred at 854.5 and 856.4 eV
(Fig. 9(a)). Metallic nickel (Ni0), usually reported around 852.6 eV, is
not detected in BR. The main component in BR can be, as expected,
NiO, but also Ni (OH)2 (Ni2+). The component centred at higher BE
(856.4 eV) is due to the formation of nickel oxyhydroxides (NiOOH)
[61]. After reaction, a residual amount of nickel is detected, the
surface being mainly composed of a carbon overlayer (C 1s region,
not shown), as observed for the Cu–Th system. However, after a
long time on stream (SS), the Ni 2p region is clearly seen.
In order to further characterize the catalyst, XPS and Raman
studies were also performed. In what concerns the XPS studies,
in terms of the copper–Th oxide sample, the XPS Cu 2p regions
of copper–thorium oxides (Fig. 8(a)), before (BR) and after (AR)
reaction are composed by two doublets with spin-orbit separa-
Th 4f7/2 region, for the same samples, displays characteristics
very similar to the one found in copper-thorium oxides. Despite
of the presence of the carbonaceous overlayer in the after reaction
sample, the Th 4f7/2 is clearly seen and quantifiable. In both AR
and SS samples, a two-phase composition can be inferred peaks
are centred at 336.0 0.5 eV (phase 1) and 338.4 0.7 eV (phase 2)
with FWHM of 2.4 0.1 eV (Fig. 9(b)). The difference of energies
between Th 4f peaks assigned to different phases is 2.7 0.2 eV in
AR sample and 2.3 0.2 eV in SS sample. Nickel, given its binding
energies similar to the one found in BR sample, seems to exist only
in phase 1 (which is thence, a mixed phase of Ni–O and Th–O).
tions of 19.9 0.1 eV (only the fitting of 2p3/2 components is shown)
2+
and a multiplet structure characteristic of Cu
species, like CuO
or Cu(OH)2. Cu 2p3/2 components, centred at 934 0.1 eV and
936.2 0.1 eV, are assigned to Cu2+ bound to O and OH groups,
respectively. Although the presence of copper in its reduced forms
is not clear from spectra, the contribution of Cu+ (Cu2O) and Cu0

254
A.C. Ferreira et al. / Journal of Alloys and Compounds 497 (2010) 249–258
Fig. 8. XPS regions (a) Cu 2p3/2, (b) Th 4f7/2, (c) O 1s and (d) C 1s of copper–thorium oxides (top to bottom) BR, AR and SS. Th 4f and O 1s peaks attributed to different phases
are labelled 1 and 2. Full and dashed vertical lines indicate the position of species in different oxidation states as reported in the literature (see text).
The XPS results for nickel–uranium oxides system show that the
Ni 2p is similar to the one found in nickel–thorium oxides. XPS U 4f
region, in the sample before reaction, is composed by one doublet
with spin-orbit separation of 10.8 eV. U 4f7/2, centred at 381.4 eV,
is typical of U6+ species, like UO3. In AR, a second doublet, of lower
intensity is fitted in U 4f region. The U 4f7/2 component, centred at
383.5 eV, could be attributed to uranium with a very electroneg-
ative neighbourhood, as in halides. However, since no halide is
detected in the survey spectrum of this sample, the most proba-
ble origin for this high BE component is, again, differential charge
accumulation due to the coexistence of two phases as in copper–
or nickel–thorium samples. The behaviour of the O 1s peak (not
shown) reinforces this last assignment since it also spreads over
a larger range of BE. However, a clear fitting of the O 1s peak is
Fig. 9. XPS regions (a) Ni 2p3/2 and (b) Th 4f7/2 of nickel–thorium oxides (top to bottom) BR, AR and SS. Th 4f peaks attributed to different phases are labelled 1 and 2. Full
and dashed vertical lines indicate the position of species in different oxidation states as reported in the literature (see text).

A.C. Ferreira et al. / Journal of Alloys and Compounds 497 (2010) 249–258
255
Fig. 10. XPS regions (a) Ni 2p3/2 and (b) U 4f of nickel–uranium oxides (top to bottom) BR, AR and SS. Full and dashed vertical lines indicate the position of species in different
oxidation states as reported in the literature (see text).
not possible. The FWHM for all the peaks is 2.4 0.1 eV. Also the
increase of intensity of the peak Ni 2p, in opposition to a decrease
in the U 4f one, shows that nickel segregation at the surface occurs.
After 18 h on stream (SS), a huge amount of carbon completely hides
the nickel signal (Fig. 10(a) and Table 1). However, (U 4f), in spite
of its weak intensity, is still detected (Fig. 10(b)).The Ni 2p photo-
electrons, having BE lower than the U 4f ones are more effectively
attenuated.
Quantitative data is assembled in Table 1 where, besides the ele-
mental percentage for each sample, also a few atomic ratios helping
the discussion are included. From the quantitative point of view, the
atomic ratio An/(Cu or Ni) before reaction, at least at the surface,
was always higher than the expected one (=0.5) and compatible
with a positive superficial segregation of the actinide. Moreover,
for the Th oxide samples, the Th/(Cu or Ni) ratio at the surface, after
reaction (despite the large uncertainty of this ratio in the Th–Ni
oxide sample), increased whereas in the case of the U oxide sam-
ple, the U/ Ni ratio decreased. Since the Ni–U oxides sample is the
one which presents a higher catalytic activity in comparison with
the other two with Th, it can be proposed that this increase of d
metal at the surface is one of the possible reasons for this catalytic
behaviour.
segregate to the surface (ꢃHf ThO2 = 1226.4, UO3 = −1223.8;
NiO = −239.7; CuO = −157.3 kJ/mol). Therefore, the segregation of
actinide to the surface (quantifications are performed assum-
ing that a homogeneous elemental distribution in depth exists)
should be thermodynamically expected. Nevertheless, after reac-
tion everything changes, distinct phases appear and the An/(Ni
or Cu) become much more difficult to rationalize through ther-
modynamics arguments. Since the samples are powders, no
further exploitation of data (angular resolution, for instance) is
possible.
In Th samples the ratio Th/(Ni or Cu) increases after reaction
and decreases for SS samples. However for Ni–U oxide, an opposite
behaviour is displayed: it decreases to a stoichiometric ratio for the
AR sample and increases for SS depassing the ratio found for the BR
sample. This opposite behaviour of the samples with temperature
and reaction may be related to the dimensions of ions in pres-
ence: whereas in Ni or Cu–Th oxides the largest ion is the Th4+ one
(r (Ni2+) = 0.69 Å; r (Cu2+) = 0.73 Å; r (Th4+) = 0.97 Å), in U–Ni oxide
the largest ion is the Ni2+ one (r (U6+) = 0.52 Å). Thence, changes
of surface composition induced by temperature seem to be con-
trolled by kinetics rather than by thermodynamics. The following
tentative surface compositions can be written: (i) before reaction,
CuO·(ThO2)0.9, NiO·(ThO2)1.6 and NiO·(ThO2)1.1; (ii) after reaction,
It has been shown elsewhere [51,62–64] that, for a binary
alloy, the constituent element with the lower heat of subli-
mation, larger atomic radius, and higher reactivity with the
ambient gas (i.e. higher heat of formation of metal oxide) will
CuO·(ThO2)1.6, NiO·(ThO2)
and NiO·(ThO2)0.5; (iii) after stabil-
ꢀ1.6
ity studies, NiO·(ThO2)0.9; (iv) phase 1 after reaction, CuO·(ThO2)0.5
and NiO·(ThO2)0.4
.
Table 1
Relative atomic amounts and atomic ratios.
At. Conc. %
Cu–Th oxides
BR
Ni–Th oxides
Ni–U oxides
BR
AR
SS
BR
AR
SS
AR
SS
C
30.9
47.8
10.2
73.2
19.1
1.5
96.6
28.4
48.7
13.9
76.6
18.0
1.2
84.1
11.9
0.7
36.0
51.2
62.2
27.6
95.2
4.2
O
3.4
Tha
b
3.2
2.9
4.3
1.3
b
Cu
Ni
Ua
11.1
b
b
9.0
2.1
6.2
6.7
6.8
2.7
0.5
0.5
Actinide/Cu or Ni
Phase 1c
Totald
0.9
0.9
0.5
1.6
–
–
1.5
1.5
–
–
0.3
0.9
1.1
1.1
0.4
0.5
–
–
n = 2 and 3 for Th and U, respectively.
a
Upper line concerns phase 1 and lower line phase 2. Samples with a single phase display a single value.
Not detected or residual.
Ratio in phase 1, computed from the area of the d metal peak and the area of the actinide peak assigned to phase 1.
Total ratio computed from the total area of each XPS peak (the ratio in phase 2 is always 0 since there is no Cu or Ni).
b
c
d

256
A.C. Ferreira et al. / Journal of Alloys and Compounds 497 (2010) 249–258
Fig. 11. Raman spectra at RT for 2NiO·UO3 when excited with the HeCd 325 nm
laser line.
Fig. 11 shows the Raman spectra in the 200–2000 cm−1 range
for each sample of the bimetallic Ni–U oxide system, after and
before reaction, obtained under a 325 nm laser line excitation. For
comparison purposes the Raman spectra of NiO and UO3 are also
included. For the ␦-UO3 no Raman activity is expected [65] because
it belongs to the cubic symmetry with the Pm3m space group. The
spectrum recorded for the UO3 oxide has a lower ratio signal/noise,
which makes it difficult to unambiguously assign the weak broad
band observed at ∼450 cm−1 and 786 cm−1, which is most proba-
bly related with the oxidation of surface leading to more complex
uranium oxides structures [66,67].
In what concerns the Raman measurements, the NiO spectrum
presents a Raman broad band at ∼1090 cm−1, which is attributed
to the 2LO overtone. Since the k = 0 phonons in the paramagnetic
(cubic) NiO phase transform as ꢄ4, no first-order scattering is
expected, however in defect-rich samples the first phonon scat-
tering intensity increases significantly. Usually the 2LO overtone
is accompanied by first-order peaks near TO (400 cm−1) and LO
(580 cm−1) frequencies originated from the parity-breaking imper-
fections [68], not detected in our NiO spectrum. Most probably the
relative enhancement observed for the 2LO is resonantly activated
under these excitation energy conditions.
The spectrum recorded before reaction for each sample of the
bimetallic Ni–U oxide system shows three large bands that, within
the experimental error, can be considered the same frequency
as the related ones observed for the individual oxides NiO and
UO3. After reaction and accordingly with XPS results, the spec-
trum is dominated by the D (1420 cm−1; FWHM = 60 cm−1) and G
(1580 cm−1; FWHM = 28 cm−1) bands, associated to graphite mate-
rial [69]. It should be mentioned that the D band is dispersive [69],
the reason why it appears located above the value usually observed
when visible light is used as excitation energy. Contrary to XRD
results the UO2 phase is not detected by UV-Raman spectroscopy,
probably due to the small penetration depth of the UV light as result
of the presence of large amounts of graphite at the surface. After
18 h on stream (SS), it is only observed an increase on the relative
intensity between G and D bands (IG/ID).
Fig. 12. Raman spectra at RT for (a) 2NiO·ThO2 and (b) 2NiO + ThO2 mechanic mix-
ture when excited with the HeCd 325 nm laser line.
The two most intense Raman features are located at
449 cm−1(FWHM = 38 cm−1) and 1131 cm−1 (FWHM = 92 cm−1),
respectively. In the bimetallic oxide, we observe a widening and a
red shift of the F2g phonon mode of ThO2, indicating a strong inter-
action between the NiO and ThO2 crystal lattices that confirm the
existence of an unusual interaction between these two phases as
evidenced by the H2-TPR results. The broad band located at higher
frequency presents the same spectral shape of the NiO 2LO phonon;
however it is shifted to higher frequencies. From the XRD results it
is found that the lattice constant in the bimetallic has the same
values of the expected results for standard ThO2 and NiO cubic
phases. Consequently it is not probably that the shifts observed on
the Raman bands could be related with a new mixed oxide phase.
Thus the structural changes affected locally the environment of the
atoms involved in those vibrations. Moreover, after reaction the
Raman bands practically get back to the frequency and band width
values observed in the oxides ThO2 and NiO, suggesting that the
catalytic reaction allowed the ThO2 and NiO structural recovery.
Taking into account the symmetry of the ThO2 Raman peak and the
fact that Ni vacancies are appointed as one of the cases of Raman
scattering activation [72] in NiO oxide, it is most probable that the
catalytic active sites are related with the chemical environment of
the anion atom belonging to both ThO2 and NiO unit cells.
In terms of the 2NiO·ThO2 bimetallic oxide and the respective
oxides, the Raman spectra measured are presented in Fig. 12(a).
The spectrum measured for thorium oxide exhibits an intense and
narrow (7 cm−1) Raman line at 466 cm−1, which corresponds to the
Raman active optical phonon of F2g symmetry expected in the cubic
fluorite-type lattices as ThO2 [70]. The additional Raman features
located at higher frequencies are related with the second-order
Raman spectrum associated to the two-phonon dispersion curves
of fluorite-type lattices [71].
For comparison purposes, the Raman spectrum of a mechanic
mixture of 2NiO + ThO2 oxides was also collected (Fig. 12(b)). For
this mixture and before reaction, the behaviour observed in the
main Raman bands, relatively to the frequency values measured
for the individual oxides, is similar to that observed for bimetal-

A.C. Ferreira et al. / Journal of Alloys and Compounds 497 (2010) 249–258
257
work is required in order to explain the differences between the
nickel compounds when Th is replaced by U. To our knowledge,
this is the first time that such results are reported for bimetallic
copper- or nickel-actinide oxides, which stresses the advantage of
using binary intermetallic compounds of the type AnNi2 (An = Th,
U) and ThCu2 as bimetallic oxide precursors.
4. Conclusions
The results obtained in the present work show that the bimetal-
lic copper- and nickel-actinide oxide catalysts obtained by an
intermetallic route via oxidation of AnNi2 (An = Th, U) and ThCu2
were very active and selective for the partial oxidation of methane
and production of synthesis gas. The poorest catalytic performance
was obtained over the bimetallic copper–thorium oxide. In terms
of the nickel-actinide oxides the activity increases when uranium
substitutes thorium. However, this increase is not substantial. In
comparison with nickel oxide, copper oxide or the actinide oxides
(ThO2, UO3) which are catalytic active for the partial oxidation of
methane, the copper- and nickel-actinide bimetallic oxides have
an improved catalytic behaviour that lies on the existence of an
unusual interaction between the d metal and the actinide oxide
phases, particularly in the case of the nickel-actinide bimetallic
oxides. The analysis of the catalysts before and after reaction by
H2-TPR, XPS and RAMAN confirm such hypothesis. The activity,
selectivity and stability of the nickel-actinide oxides were, under
the same conditions, comparable to that of a platinum commercial
catalyst, 5 wt% Pt/Al2O3, which stresses the advantage of (i) this
new type of compounds and (ii) the synthetic method used in this
work.
Fig. 13. Raman spectra at RT for (a) 2NiO·UO3, (b) 2NiO·ThO2, (c) 2NiO + ThO2
mechanic mixture and (d) 2CuO·ThO2 when excited with the HeCd 325 nm laser
line.
lic 2NiO·ThO2. Nevertheless, after reaction the Raman bands do
not return to the frequency and widths values measured for each
of the individual oxides. Moreover, the amount of carbon phases
deposited at the surface is residual. It is interesting to notice that the
Raman band at ∼560 cm−1, also observed in the bimetallic, becomes
now more pronounced. The spectral position and shape corre-
late well with the disorder activated LO one-phonon reported for
NiO crystal [68,73]. Thus, we conclude that the mechanic mixture
promotes more the existence of NiO defect-rich or antiferromag-
netically ordered than the process undertaken to synthesize the
binary intermetallic 2NiO·ThO2. However, the condition to an effec-
tive interaction between both lattices is not equivalent and thus the
lower catalytic activity.
Acknowledgment
This work was supported by the Portuguese “Fundac¸ ão
para a Ciência e a Tecnologia” under the contract PTDC/EQU-
EQU/65126/2006.
Fig. 13 shows the Raman spectra observed for 2CuO·ThO2. When
Cu substitutes Ni in the thorium systems, the shift to lower fre-
quency and the width observed for the related ThO2 principal band
decrease, suggesting a lower interaction between both crystal lat-
tices that explains the lower catalytic activity of the bimetallic
Cu–Th oxides when copper is substitute by nickel. Two additional
bands are observed, located at 287 and 627 cm−1, and are attributed
to CuO optical phonon modes of symmetry Ag and Bg, respec-
tively [74]. The fact that these wave numbers are lower than those
reported in the literature for bulk CuO may be attributed to size
effects [75].
After reaction and contrary to what is observed for 2NiO·ThO2,
the related ThO2 Raman main peak shifts even more towards lower
frequency and broadens. This behaviour is mainly related with the
thermal expansion of the ThO2 lattice and not due to a strong inter-
action with the CuO lattice. We also note that only the graphite zone
center E2g mode (G peak) is present at 1780 cm−1, and it is residual
compared with results obtained for the other samples, in the same
conditions. Another interesting aspect is the instability of graphite
phase after 18 h on stream (SS). The D band increases dramatically
relatively to the G band indicating the presence of highly disorder
graphite phase or of new carbon structures [69].
References
[1] T.V. Choudhary, S. Banerjee, V.R. Choudhary, Appl. Catal. A 234 (2002) 1–23.
[2] J.M. Li, F.Y. Huang, W.Z. Weng, X.Q. Pei, C.R. Luo, H.Q. Lin, C.J. Huang, H.L. Wan,
Catal. Today 131 (2008) 179–187.
[3] C. Mateos-Pedrero, C. Cellier, P. Eloy, P. Ruiz, Catal. Today 128 (2007) 216–222.
[4] H. Morioka, Y. Shimizu, M. Sukenobu, K. Ito, E. Tanabe, T. Shishido, K. Takehira,
Appl. Catal. A 215 (2001) 11–19.
[5] A.P.E. York, T.C. Xiao, M.L.H. Green, Top. Catal. 22 (2003) 345–358.
[6] E.P.J. Mallens, J.H.B.J. Hoebink, G.B. Marin, J. Catal. 167 (1997) 43–56.
[7] M. Fathi, K.H. Hofstad, T. Sperle, O.A. Rokstad, A. Holmen, Catal. Today 42 (1998)
205–209.
[8] K. Sekizawa, H. Widjaja, S. Maeda, Y. Ozawa, K. Eguchi, Catal. Today 59 (2000)
69–74.
[9] S. Rabe, M. Nachtegaal, F. Vogel, Phys. Chem. Chem. Phys. 9 (2007) 1461–1468.
[10] A. Donazzi, A. Beretta, G. Groppi, P. Forzatti, J. Catal. 255 (2008) 241–258.
[11] T.L. Zhu, M. Flytzani-Stephanopoulos, Appl. Catal. A 208 (2001) 403–417.
[12] V.R. Choudhary, A.S. Mamman, S.D. Sansare, Angew. Chem. Int. Ed. 31 (1992)
1189–1190.
[13] V.R. Choudhary, A.M. Rajput, B. Prabhakar, Catal. Lett. 15 (1992) 363–370.
[14] S. Xu, X.B. Yan, X.L. Wang, Fuel 85 (2006) 2243–2247.
[15] V.R. Choudhary, V.H. Rane, A.M. Rajput, Catal. Lett. 22 (1993) 289–297.
[16] A. Djaidja, A. Barama, M.M. Bettahar, Catal. Today 61 (2000) 303–307.
[17] V.R. Choudhary, A.M. Rajput, V.H. Rane, Catal. Lett. 16 (1992) 269–272.
[18] A. Slagtern, H.M. Swaan, U. Olsbye, I.M. Dahl, C. Mirodatos, Catal. Today 46
(1998) 107–115.
[19] R.C. Jin, Y.X. Chen, W.Z. Li, W. Cui, Y.Y. Ji, C.Y. Yu, Y. Jiang, Appl. Catal. A 201
(2000) 71–80.
[20] J. Barbero, M.A. Pena, J.M. Campos-Martin, J.L.G. Fierro, P.L. Arias, Catal. Lett. 87
(2003) 211–218.
[21] V.R. Choudhary, V.H. Rane, A.M. Rajput, Appl. Catal. A 162 (1997) 235–238.
[22] V.R. Choudhary, A.M. Rajput, B. Prabhakar, J. Catal. 139 (1993) 326–328.
[23] Q. Miao, G.X. Xiong, S.S. Sheng, W. Cui, L. Xu, X.X. Guo, Appl. Catal. A 154 (1997)
17–27.
[24] L. Cao, Y. Chen, W. Li, Nat. Gas Convers. IV 107 (1997) 467–471.
[25] Q.G. Yan, W. Chu, L.Z. Gao, Z.L. Yu, S.Y. Yuan, Nat. Gas Convers. V 119 (1998)
855–860.
Therefore, after analysis of the catalysts before and after reac-
tion it can be said that the true nature of the active species that
explain the performance of the bimetallic Ni- or Cu-actinide oxides
is mainly connected to the existence and strength of an unusual
interaction between the catalysts crystal lattices as evidenced by
the H2-TPR, XPS and Raman results. The strength of this interaction
certainly explains the lower catalytic activity of the bimetallic tho-
rium oxide catalysts when copper is substitute by nickel but, further

258
A.C. Ferreira et al. / Journal of Alloys and Compounds 497 (2010) 249–258
[26] R. Horn, K.A. Williams, N.J. Degenstein, A. Bitsch-Larsen, D.D. Nogare, S.A. Tupy,
L.D. Schmidt, J. Catal. 249 (2007) 380–393.
[52] S. Bennici, P. Carniti, A. Gervasini, Catal. Lett. 98 (2004) 187–194.
[53] M.E. Rivas, J.L.G. Fierro, R. Guil-Lopez, M.A. Pena, V. La Parola, M.R. Goldwasser,
Catal. Today 133 (2008) 367–373.
[54] G.H. Li, L.J. Hu, J.M. Hill, Appl. Catal. A 301 (2006) 16–24.
[55] J.I. Villacampa, C. Royo, E. Romeo, J.A. Montoya, P. Del Angel, A. Monzon, Appl.
Catal. A 252 (2003) 363–383.
[27] A. Donazzi, A. Beretta, G. Groppi, P. Forzatti, J. Catal. 255 (2008) 259–268.
[28] P. Gelin, M. Primet, Appl. Catal. B—Environ. 39 (2002) 1–37.
[29] B.C. Enger, R. Lodeng, A. Holmen, Appl. Catal. A 346 (2008) 1–27.
[30] V.R. Choudhary, A.M. Rajput, B. Prabhakar, A.S. Mamman, Fuel 77 (1998)
1803–1807.
[31] D. Ballivet-Tkatchenko, J. Branco, A.P. Dematos, J. Phys. Chem. 99 (1995)
5481–5484.
[32] J. Branco, C. de Jesus Dias, A.P. Goncalves, T.A. Gasche, A.P. de Matos, Ther-
mochim. Acta 420 (2004) 169–173.
[56] J. Ghijsen, L.H. Tjeng, J. Vanelp, H. Eskes, J. Westerink, G.A. Sawatzky, M.T.
Czyzyk, Phys. Rev. B 38 (1988) 11322–11330.
[57] C.D. Wagner, A.V. Naumkin, A. Kraut-Vass, J.W. Allison, C.J. Powell, J.R.R. Jr., NIST
X-ray Photoelectron Spectroscopy Database, Standard Reference Database 20,
Version 3.5, National Institute of Standards and Technology (NIST), 2007.
[58] S. Dash, M. Kamruddin, P.K. Ajikumar, A.K. Tyagi, B. Raj, S. Bera, S.V. Narasimhan,
J. Nucl. Mater. 278 (2000) 173–185.
[33] J.B. Branco, D. Ballivet-Tkatchenko, A.P. de Matos, J. Phys. Chem. C 111 (2007)
15084–15088.
[34] J.B. Branco, T.A. Gasche, A.P. Goncalves, A.P. de Matos, J. Alloys Compd. 323
(2001) 610–613.
[59] A.I. Melato, L.M. Abrantes, A.M. Botelho do Rego, Thin Solid Films (2009),
doi:10.1016/j.tsf.2009.1007.1157.
[35] JCPDS, The Powder Diffraction File, JCPDS, 1601 Park Avenue, Swarthmore, PA,
1981.
[60] G. Beamson, D. Briggs, High Resolution XPS of Organic Polymers. The Scienta
ESCA300 Database, John Wiley & Sons, New York, 1992.
[61] A.P. Grosvenor, M.C. Biesinger, R.S. Smart, N.S. McIntyre, Surf. Sci. 600 (2006)
1771–1779.
[62] M. Houalla, C.L. Kibby, L. Petrakis, D.M. Hercules, J. Phys. Chem. 87 (1983)
3689–3693.
[36] V. Uvarov, I. Popov, Mater. Charact. 58 (2007) 883–891.
[37] A.C. Ferreira, A.M. Ferraria, A.M.B.d. Rego, A.P. Gonc¸ alves, R. Correia, T.A. Gasche,
J.B. Branco, J. Alloys Compd. (2009), doi:10.1016/j.jallcom.2009.1009.1082.
[38] D. Ballivet-Tkatchenko, G. Delahay, J. Therm. Anal. 41 (1994) 1141.
[39] J.B. Branco, D. Ballivet-Tkatchenko, A.P. de Matos, J. Alloys Compd. 464 (2008)
399–406.
[63] C.S. Huang, M. Houalla, D.M. Hercules, C.L. Kibby, L. Petrakis, J. Phys. Chem. 94
(1990) 6749–6753.
[40] D. Ballivet-Tkatchenko, G. Delahay, J. Therm. Anal. 41 (1994) 1141–1151.
[41] D.A. Hickman, L.D. Schmidt, Science 259 (1993) 343–346.
[42] D. Ma, D.J. Mei, X. Li, M.C. Gong, Y.Q. Chen, J. Rare Earths 24 (2006) 451–455.
[43] M.S. Liao, C.T. Au, C.F. Ng, Chem. Phys. Lett. 272 (1997) 445–452.
[44] Y.Z. Lin, J. Sun, J. Yi, J.D. Lin, H.B. Chen, D.W. Liao, J. Mol. Struct.—THEOCHEM
587 (2002) 63–71.
[45] P. Chen, H.B. Zhang, G.D. Lin, K.R. Tsai, Appl. Catal. A 166 (1998) 343–350.
[46] A.M. DeGroote, G.F. Froment, Catal. Today 37 (1997) 309–329.
[47] W. Chu, W.G. Yan, X. Liu, Q. Li, Z.L. Yu, G.X. Xiong, Nat. Gas Convers. V 119 (1998)
849–854.
[48] A.G. Dedov, A.S. Loktev, I.I. Moiseev, A. Aboukais, J.F. Lamonier, I.N. Filimonov,
Appl. Catal. A 245 (2003) 209–220.
[49] S. Kus, M. Otremba, M. Taniewski, Fuel 82 (2003) 1331–1338.
[50] D. Monti, A. Baiker, J. Catal. 83 (1983) 323–335.
[51] D.D. Wagman, W.H. Evans, V.B. Parker, R.H. Schumm, I. Halow, S.M. Bailey, K.L.
Churney, R.L. Nuttall, J. Phys. Chem. (Suppl. 2) (1982), Ref. Data 11.
[64] Vansante, W.M. Ra, Sachtler, J. Catal. 33 (1974) 202–209.
[65] F.M. Jiang, S. Kojima, Jpn. J. Appl. Phys., Part 1-Reg. Papers Short Notes & Review
Papers 38 (1999) 5128–5132.
[66] D. Manara, B. Renker, J. Nucl. Mater. 321 (2003) 233–237.
[67] M.L. Palacios, S.H. Taylor, Appl. Spectrosc. 54 (2000) 1372–1378.
[68] R.E. Dietz, G.I. Parisot, A.E. Meixner, Phys. Rev. B 4 (1971) 2302-&.
[69] C. Thomsen, S. Reich, Phys. Rev. Lett. 85 (2000) 5214–5217.
[70] Vg. Keramida, W.B. White, J. Chem. Phys. 59 (1973) 1561–1562.
[71] A. Jayaraman, G.A. Kourouklis, L.G. Vanuitert, Pramana 30 (1988) 225–231.
[72] W.Z. Wang, Y.K. Liu, C.K. Xu, C.L. Zheng, G.H. Wang, Chem. Phys. Lett. 362 (2002)
119–122.
[73] N. Mironova-Ulmane, A. Kuzmin, I. Steins, I.S.J. Grabis, M. Pärs, J. Phys.: Conf.
Ser. 93 (2007) 012039.
[74] J. Chrzanowski, J.C. Irwin, Solid State Commun. 70 (1989) 11–14.
[75] J.F. Xu, W. Ji, Z.X. Shen, W.S. Li, S.H. Tang, X.R. Ye, D.Z. Jia, X.Q. Xin, J. Raman
Spectrosc. 30 (1999) 413–415.