Impact of the urea-matrix combustion method on the HDS performance of Ni-MoS2/γ-Al2O3 catalysts

Article abstract of DOI:10.1016/j.molcata.2005.06.055

A detailed study has been made of the different steps involved upon the preparation of γ-Al₂O₃-supported Ni-Mo HDS catalyst precursors by urea-matrix combustion (UMxC) method. Catalyst performance was evaluated using a tubular fixed-bed reactor and the hydrodesulfurization of thiophene under normal pressure as a model reaction. The oxidic and sulfurized states of the HDS catalysts were characterized by X-ray diffraction (XRD), laser Raman spectroscopy, thermogravimetric analysis (TGA), differential scanning calorimetry (DSC) and high resolution transmission electron microscopy (HRTEM) in order to correlate their oxidic and sulfurized properties with the catalytic behaviour. During the UMxC process several consecutive stages such as melting, dissolution and chemical reactions occurred. There was no evidence of residual carbon and well-dispersed Ni- and Mo-oxo-species supported on alumina were formed. Urea employed as fuel not only increases the combustion rate, but also undergoes a decomposition process (endothermic reaction) that could contribute to the reduction of the combustion temperature. The urea-matrix combustion method permit to synthesize highly active γ-Al₂O ₃-supported Ni-Mo HDS catalysts with a comparable promoter effect than that of corresponding catalyst prepared by impregnation method. In addition, an opposite relation between the activity and the hydrogenation properties was observed indicating that highly active HDS catalyst requires low consumption of hydrogen. Finally, both the ignition temperature and the urea-oxidizer ratio produce no significant changes in the HDS catalytic properties of Ni-Mo-based catalysts.

Full text of DOI:10.1016/j.molcata.2005.06.055

Journal of Molecular Catalysis A: Chemical 240 (2005) 214–225
Impact of the urea–matrix combustion method on the HDS
performance of Ni-MoS2/␥-Al2O3 catalysts
∗
´
´
Sergio L. Gonzalez-Cortes , Tian-Cun Xiao, Serbia M.A. Rodulfo-Baechler,
Malcolm L.H. Green
Wolfson Catalysis Centre, Inorganic Chemistry Laboratory, University of Oxford, South Parks Road, Oxford OX1 3QR, UK
Received 12 May 2005; received in revised form 8 June 2005; accepted 8 June 2005
Available online 8 August 2005
Abstract
A detailed study has been made of the different steps involved upon the preparation of ␥-Al2O3-supported Ni-Mo HDS catalyst precursors by
urea–matrix combustion (UMxC) method. Catalyst performance was evaluated using a tubular fixed-bed reactor and the hydrodesulfurization
of thiophene under normal pressure as a model reaction. The oxidic and sulfurized states of the HDS catalysts were characterized by X-ray
diffraction (XRD), laser Raman spectroscopy, thermogravimetric analysis (TGA), differential scanning calorimetry (DSC) and high resolution
transmission electron microscopy (HRTEM) in order to correlate their oxidic and sulfurized properties with the catalytic behaviour. During
the UMxC process several consecutive stages such as melting, dissolution and chemical reactions occurred. There was no evidence of residual
carbon and well-dispersed Ni- and Mo-oxo-species supported on alumina were formed.
Urea employed as fuel not only increases the combustion rate, but also undergoes a decomposition process (endothermic reaction) that
could contribute to the reduction of the combustion temperature. The urea–matrix combustion method permit to synthesize highly active
␥-Al2O3-supported Ni-Mo HDS catalysts with a comparable promoter effect than that of corresponding catalyst prepared by impregnation
method. In addition, an opposite relation between the activity and the hydrogenation properties was observed indicating that highly active
HDS catalyst requires low consumption of hydrogen. Finally, both the ignition temperature and the urea-oxidizer ratio produce no significant
changes in the HDS catalytic properties of Ni-Mo-based catalysts.
© 2005 Elsevier B.V. All rights reserved.
Keywords: Urea; HDS reactions; Ni-promoted MoS2 catalysts; Urea–matrix combustion method
1. Introduction
petroleum feedstocks to meet this requirement [2], whereas
others are based on the improvement of current hydrotreat-
ing technology. Nowadays, alumina-supported molybdenum
sulphide catalysts, promoted with cobalt or nickel, are the
commonly employed catalysts for HDS reactions of fossil
fuels due to their high activity and stability [3–7]. Sup-
ported nickel-molybdenum catalysts are usually prepared
by depositing molybdenum and nickel oxides on the sur-
face of ␥-alumina (Al2O3). The catalysts are used in the
active sulfide state and the nature and properties of the oxidic
precursors largely determine the performance of the cata-
lysts. Several studies focused on the promoter effect of Co
and Ni on the oxo-Mo species indicates that both elements
do not affect greatly the state of the Mo species over the
␥-Al2O3 [8,9]. However, other studies report that the inter-
The stringent environmental regulations in the US and
Europe are requiring significant improvement in the qual-
ity of transportation fuel. It is directed towards reducing the
amount of sulfur content in gasoline and diesel fuel to 10 ppm
by 2009 [1]. Many approaches based on novel processes such
as adsorption and selective oxidative desulfurization have
been proposed to remove sulfur-containing compounds from
∗
Corresponding author. Permanent address: Laboratorio de Cine´tica y
Cata´lisis, Universidad de Los Andes, Me´rida 5101, Venezuela.
Tel.: +44 1865 272660; fax: +44 1865 272690.
E-mail addresses: sergio.gonzalezcortes@chem.ox.ac.uk,
goncor@ula.ve (S.L. Gonza´lez-Corte´s).
1381-1169/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2005.06.055

S.L. Gonza´lez-Corte´s et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 214–225
215
action between the oxidic precursors after the calcination
process might favour either the formation of NiMoO4-type
phase [10] or the Ni-Mo-O species [11], which could be
optima precursors for the formation of the Ni-Mo-S active
phases.
support at ca. 50 ◦C for 2–3 h to obtain urea-based slurry
containing nickel and molybdenum fully loaded over the alu-
mina. The resulting paste was divided in two portions: (i) a
part was dried under high vacuum (ca. 10−4 Torr) for 24 h in
order to study the combustion process using TGA–DSC and
(ii) another part is ignited at 500 ◦C (furnace temperature)
in static air for 10 min to produce a yellow oxide catalyst
precursor material.
We recently developed a new approach called urea–matrix
combustion (UMxC) method to prepare highly active Co-
promoted alumina-supported MoS2 HDS catalysts [12]. This
method, based on a self-propagating combustion reaction
[13], requires low ignition temperature to induce a com-
bustion reaction between the catalyst precursor salts and
an organic matrix as the fuel. The final products of com-
bustion might contain a high concentration of structural
defects, which are beneficial as active centres in catalysis
[14]. The main characteristics are complete or partial elim-
ination of an external energy supply and very high rates
of heating and cooling, while control of process rate, tem-
perature, degree of conversion, composition and structure
of products can be obtained by varying the rates of heat
release and transfer [15,16]. This material synthesis method
has been used to prepare catalysts for deep oxidation of
methane [17], oxidative coupling of methane [18], oxidation
of carbon monoxide and reduction of NO [19,20] and more
recently Co-Mo HDS catalysts [12]. Since combustion syn-
thesis is based on the transformation of the precursor salts
through an exothermic reaction; it generates very high heat-
ing and cooling rates in very short period, which can lead
to the formation of new highly defective structures of nano-
scale meta-stable phases. This fact is particularly important
for the synthesis of promoted catalysts such as HDS cata-
lysts where the interaction between the oxidic precursors is
directly related to the catalytic performance of hydrotreating
catalysts [21].
2.2. Sulfurization and catalytic tests
The HDS reactions free of mass transport interference
were carried out in a fixed-bed tubular quartz reactor
(34 cm long, 4 mm i.d.). Oxidic catalyst precursors (200 mg,
<250 ␮m) were loaded between quartz wool plugs. The
activation (e.g., sulfurization) process was done using a
20 mL min−1 hydrogen (ultra high purity) flow rate, which
passedthroughathiophene(99+%, Aldrich)saturatoratroom
temperature (10.9 kPa). All lines located after the saturator
were heated at temperature high enough to avoid thiophene
condensation. The furnace temperature is increased linearly
from room temperature to 450 ◦C at 5 ◦C min−1 and held at
450 ◦C for 10 h. Then, the sample was treated with a pure
H2 flow at 450 ◦C for 1 h and cooled down to 350 ◦C. The
catalysts were tested at atmospheric pressure (0.1 MPa) and
the reactants consisted of a 20 mL min−1 H2 flow saturated
with thiophene vapour at 0 ◦C (2.96 kPa), resulting in a mix-
ture of ca. 3 mol% thiophene in H2. The reaction products
were analyzed using an on-line HP 5890 Series II chro-
matograph equipped with a Hayesep R packed column and
a flame ionization detector (FID). After completion of the
catalytic reaction, the catalyst sample was cooled down to
room temperature under flowing argon and then passivated
before exposure to the atmosphere and kept in a desiccator
thereupon.
In order to find out whether urea–matrix combustion
method is also suitable to prepare alumina-supported Ni-
promoted MoS2 HDS catalysts, we have synthesized and
studied the HDS activity of a series of catalysts with variable
Ni loadings. The molybdenum oxide content was maintained
constantat12 wt.%. Inaddition, wehavecarriedoutadetailed
thermo-chemical examination of the urea–matrix combustion
method in order to determine the influence of molybdenum
precursor (i.e., ammonium heptamolybdate) and the fuel-
oxidizer ratio on the combustion process.
2.3. Characterization of the precursors and HDS
catalysts
The crystalline phases of the oxidic precursors and the sul-
furized catalysts were identified by X-ray diffraction (XRD)
in a Philips PW1710 diffractometer equipped with an X-
˚
ray tube (Cu K␣ radiation λ = 1.5406 A; 40 kV, 30 mA). The
specimen were prepared by grinding a small amount of each
sample using an agate mortar and pestle and then loaded into
a flat sample holder. The data were collected in θ/2θ reflec-
tion mode, from 3 to 70◦ in 2θ, using steps of 0.05◦, with
time per step of 1.25 s.
2. Experimental
2.1. Preparation of the oxidic precursors
Bimetallic Ni-Mo catalyst precursors with 12 wt.% MoO3
and variable NiO composition supported on ␥-Al2O3
(164 m2/g) were prepared by mixing urea with ammonium
heptamolybdate-4-hydrate and Ni(II) nitrate employing vari-
able urea/(Mo + Ni) molar ratio and ca. 1 g/2 mL = (Ni- and
Mo-precursors + urea)/water ratio. This mixture was stirred
to form a homogeneous slurry, then mixed with the ␥-Al2O3
Thermogravimetric analysis (TGA) and differential scan-
ning calorimetry (DSC) were simultaneously carried out on a
Rheometric Scientific STA 1500 instrument. TGA–DSC pro-
files were recorded from room temperature to 750 ◦C, using
30 mL min−1 dry air flow and 10 ◦C min−1 heating rate. For
each analysis ca. 25–35 mg of sample was loaded into a small
alumina crucible, using alumina as reference.

216
S.L. Gonza´lez-Corte´s et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 214–225
The Raman spectra of the oxidic precursors and the sul-
3. Results and discussion
furized catalysts were recorded in an Yvon Jobin Labram
spectrometer with a 632 nm HeNe laser, run in a back-
scattered co-focal arrangement. The samples were pressed in
a microscope slide; with a 45 s scanning time and 2 cm−1 res-
olution. Several points of each catalyst surface were probed to
explore homogeneity of the sample and reproducibility of the
data.
High resolution transmission electron microscopy
(HRTEM) was carried out using a JEOL 4000FX electron
microscope with a 400 kV accelerating voltage. Ni-Mo
sulfurized catalyst was ground into a fine powder and
dispersed in AR-grade chloroform. Then, it was placed
in an ultrasonic bath for ca. 15 min, before a drop of the
suspension was put on a lacey carbon-coated copper grid
(Agar, 20 mesh) and subsequently analyzed.
3.1. Study of the combustion–decomposition processes
of the urea–matrix prepared catalysts
The TG curves and the differential TG profiles (inset)
of the urea-containing Ni-Mo precursor salts are given in
Fig. 1A. The major weight loss (WL) for the monometallic
(i.e., Ni and Mo) and bimetallic (Ni-Mo) precursor thermo-
grams takes place in the 130–300 ◦C range and it is attributed
to a complex series of overlapping chemical reactions such
as the partial dehydroxylation of alumina [22] and the partial
dehydration and decomposition of the hydrated nickel nitrate
and ammonium heptamolybdate salts [23–26]. In addition,
weight losses at temperatures greater than 300 ◦C for the
monometallic oxidic precursors are observed (Fig. 1Aa and
Fig. 1. (A) Thermogravimetric profiles of the ␥-alumina-supported Ni-Mo precursor salts with urea/metals mole ratio of 10 and dried at room temperature
in high vacuum (∼10−4 Torr). Inset is the differential TG profiles. (a) 12 wt.% MoO3/␥-Al2O3, (b) 3 wt.% NiO-12 wt.% MoO3/␥-Al2O3, (c) 12 wt.% NiO/␥-
Al2O3. (B) Differential scanning calorimetry curves of the ␥-alumina-supported Ni-Mo precursor salts with urea/metals mole ratio of 10 and dried at room
temperature in high vacuum. (a) 12 wt.% MoO3/␥-Al2O3, (b) 3 wt.% NiO-12 wt.% MoO3/␥-Al2O3, (c) 12 wt.% NiO/␥-Al2O3.

S.L. Gonza´lez-Corte´s et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 214–225
217
Fig. 2. (A) Thermogravimetric profiles of the ␥-alumina-supported Ni-Mo precursor salts (i.e. 3 wt.% NiO-12 wt.% MoO3) containing different urea/(Ni + Mo)
mole ratio and dried at room temperature in high vacuum. Inset is the differential TG profiles. (a) 0.0, (b) 0.25, (c) 1.0, (d) 5.0, (e) 10.0. (B) Differential scanning
calorimetry of the ␥-alumina-supported Ni-Mo precursor salts (i.e. 3 wt.% NiO-12 wt.% MoO3) containing different urea/(Ni + Mo) mole ratio and dried at
room temperature in high vacuum. (a) 0.0, (b) 0.25, (c) 1.0, (d) 5.0, (e) 10.0.
Ac). The total weight losses coincided with the expected
values (assuming the formation of nickel and molybdenum
oxides), indicating that both the urea and the precursor salts
are completely decomposed without residue of carbon.
The DSC profiles of the urea-containing Ni-Mo precur-
sor salts are given in Fig. 1B. It shows several endothermic
processes at temperatures lower than 250 ◦C and differ-
ent exothermic reactions at higher temperatures. A narrow
endothermic peak at 135 ◦C associated with the melting pro-
cess of urea is shown in the DSC curve of 12 wt.% MoO3/␥-
Al2O3 oxidic precursor (see Fig. 1Ba) [25]. A similar stage
is also observed in the profiles of Ni-Mo and Ni precur-
sors (see Fig. 1Bb and 1Bc), however the peak position is
clearly shifted to lower temperature and becomes broader
with increasing of NiO loading. This trend can be due to the
presence of an urea ↔ nickel nitrate interaction and/or the
depressed effect of impurities on the melting point of urea
(i.e. colligative property).
On the other hand, a broad endothermic band occurs in
the same temperature range that the major weight loss takes
place and subsequently the sample underwent the combus-
tion process. This sequence of processes suggests that the
thermal decomposition of urea might play a very important
role as an attenuating process of the maximum combus-
tion temperature. These processes (i.e., decomposition and
combustion) can be overlapped as occurred in the Ni-Mo
oxidic precursor or can take place at temperature clearly dif-
ferent as occurred in the monometallic precursors. On the
other hand, not only the peak position of the heat flow lib-
erated during combustion reaction, but also the enthalpy of
the combustion processes depends markedly on the catalyst
composition.

218
S.L. Gonza´lez-Corte´s et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 214–225
The TG and differential TG profiles (inset) of the
is slightly shifted to higher temperatures and becomes nar-
row whereas the urea/(Ni + Mo) ratio increased up to 1.0.
A further rise of urea content leads to a widening and shift
of the stage to higher temperature and also produces a sec-
ond exothermic process at 346 ◦C when the urea/(Ni + Mo)
ratio is 10 due to further combustion of residual organic
matrix.
urea-containing alumina-supported Ni-Mo precursor salts,
corresponding to 3 wt.% NiO-12 wt.% MoO3/␥-Al2O3 oxi-
dic precursor, are given in Fig. 2A. The sample free of urea
presents two major weight losses of 3.5 and 7.8 wt.% max-
imized at ca. 100 and 233 ◦C, respectively (Fig. 2Aa). The
former is due to the desorption of physically adsorbed water
on alumina surface [27,28] and the latter is attributed to a
complex series of overlapping chemical reactions, such as
the partial dehydroxylation of alumina [22] and the partial
dehydration-decomposition and the subsequent total trans-
formation of the Ni and Mo precursor salts into the cat-
alyst oxidic precursor [23–26]. The presence of urea (i.e.,
urea/(Ni + Mo) of 0.25) produces a new weight loss step max-
imized at intermediate temperature (i.e., 181 ◦C) (Fig. 2Ab)
this step becomes greater and asymmetric with the increase
of urea content. Simultaneously, the weight loss step at higher
temperature undergoes both a slight increase of the maximum
temperature and a clear descent of its intensity compared
with the stage at intermediate temperature. This effect is
notable for the precursor containing an urea/(Ni + Mo) ratio
of 10 where one major weight loss at 183 ◦C is observed see
(Fig. 2Ae).
The DSC profiles of the alumina-supported Ni-Mo pre-
cursor salts are shown in Fig. 2B. DSC curve of the urea-free
sample displays two broad endothermic processes centred at
ca. 110 and 199 ◦C and one narrow exothermic at 243 ◦C
(Fig. 2Ba). The first stage is mainly due to water desorption
and the second one is associated with the partial dehydration-
decomposition of the Ni and Mo precursor salts, whereas the
third stage at 243 ◦C is assigned to an exothermic process
that leads to total transformation of the partially decom-
posed Ni and Mo precursors into the catalyst oxidic precursor.
The DSC profiles of the urea-containing precursors display
a similar thermal behaviour to that of the urea-free sample.
However, the major endothermic band becomes stronger and
asymmetric, due to the decomposition of urea, leading to the
formation of two well defined peaks when urea/(Ni + Mo)
ratio is 5.0 and a broad band for an urea/(Ni + Mo) ratio
of 10 (Fig. 2Bd and Be) is observed. Also, these profiles
displayed a small endothermic peak at temperatures clearly
lower than the melting point of urea (i.e., 135 ◦C [25]),
probably due to the presence of ionic species within urea
crystal structure that reduce its melting temperature (col-
ligative property). On the other hand, the exothermic feature
The experimental and theoretical total weight losses and
combustion rates of alumina-supported 3Ni-12Mo catalyst
precursors are listed in Table 1. The experimental total weigh
losses are very close to the theoretical total weight losses,
especially at high urea compositions. This suggests that dur-
ing the combustion process the Ni and Mo oxidic precursors
arenotlostbyvolatilizationandalsothatcarbonmaterialdoes
not remain in the catalyst precursor. Therefore the extent of
combustion reaction based on the carbon content is com-
plete. On the other hand, the weight loss by combustion
against total weight loss (i.e., WLComb/WLTEoxtpal) for urea-
free precursor is clearly greater than those for urea-containing
precursors. This finding is attributed to the notable contribu-
tion of the urea thermal decomposition over the total weight
loss even at a composition close to the stoichiometric com-
position of the redox mixture (i.e., urea/(Ni + Mo) = 0.27).
Under this optimal composition the total oxidizing and reduc-
ing valences of the oxidizer–fuel mixture is unity, according
to the propellant chemistry and hence the energy released
during the combustion is maximum [29–31]. Indeed, the
combustion enthalpy of the urea-containing precursor (i.e.,
urea/(Ni + Mo) = 0.25) is ca. 32% greater than that produced
by urea-free precursor, a quite similar trend is also observed
for the combustion rates, Table 1. A global analysis of
these results indicates that the increase of the urea load-
ing facilitates the urea partial thermal decomposition before
starting the combustion reaction at the ignition tempera-
ture of ∼200 ◦C and both processes are overlapped when
high urea composition is employed. In consequence, the
combustion temperature could be markedly lower than that
of the sample prepared from the urea/(Ni + Mo) ratio of
0.25.
The combustion reactions can reach extremely high tem-
peratures within very short time. It is therefore reasonable to
assume that a thermally isolated system exists because there
is very little time for the heat to disperse to its surroundings.
Therefore, the maximum temperature achieved by the reac-
tion products is assumed to be adiabatic temperature (Tad),
Table 1
Weight losses and combustion rates of 3% NiO-12% MoO3/␥-Al2O3 prepared by urea–matrix combustion method
Total
Theor.
Total
Exp.
Urea/(Ni + Mo) mole ratio
WL
(%)
WL
(%)
WLComb./WLTEoxtpa.l (%)
Combustion ratea (WL%/◦C) × 104
0.0
0.25
1.0
5.0
10.0
10.4
11.9
16.1
32.9
46.4
11.3
13.2
16.3
33.0
45.5
69.0
56.8
39.3
37.6
nd
1009.7 5.3
1389.7 5.2
1463.9 6.5
1429.1 1.9
nd
WL, weight loss; nd, value not determined because of overlapping of combustion and decomposition processes.
a
Slope of TGA curve in the 200–250 ◦C region with a correlation coefficient greater than 0.999.

S.L. Gonza´lez-Corte´s et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 214–225
219
which can be calculated using the Eq. (1) [32]:
acts as fuel upon combustion process. It is supported not
only by the exothermic stage displayed in the DSC profile of
the urea-free precursor (see Fig. 2Ba), but also by the clear
increase of the adiabatic temperature whereas Mo composi-
tion is increased despite the lower urea composition present
in the redox mixture.
Nevertheless, the presence of urea is of paramount impor-
tance for combustion synthesis since it increases notably the
combustion rate (see Table 1). Urea, even at stoichiomet-
ric composition, also undergoes a decomposition process
in addition to the combustion process. Hence, when great
amount of urea (greater than the optimal redox mixture) is
used longer reaction time is required (i.e. 10 min) but also
very likely urea contributes to the diminution of the combus-
tion temperature since higher amount of gases are generated
upon decomposition–combustion process. This fact could
affect the growth and/or formation of hard agglomerates as
was determined for the synthesis of Ni-Zn ferrite powders
[36].
Fig. 4 compares the laser Raman spectra of urea and
different formulations of the Ni-Mo catalyst precursors.
The ␥-alumina-supported Ni-Mo precursor salts display a
major peak at 1012 cm−1 and small bands at 1500–1700 and
1178 cm−1 assigned to the C N, C O bond stretch modes
and coupling between the νs(CN) and ρs(NH2), respectively,
[37,38] of the urea matrix. The Raman peak at 548 cm−1 is
attributed to the C O and/or CN deformation vibrations [37].
In addition, a strong peak at 944 and a shoulder at 900 cm−1
assigned to the symmetric and asymmetric stretching and a
weak band at 353 cm−1 attributed to the bending mode of
the terminal Mo O for the Anderson-type heteropolyanion
(i.e. (NH4)3[Al(OH)6Mo6O18])[39]areobserved. Thedefor-
mation of the Mo O Mo bridges at 221 cm−1 and a weak
band at 573 cm−1 attributed to the Al O stretching mode
are as well detected [40,41] see (Fig. 4b and c). This entity
results from the partial dissolution of alumina and its subse-
quent complexation with the heptamolybdate anion. A small
ꢀ
Tad
◦
−ꢀHr =
Cp(products) dT
(1)
T0
where ꢀH◦ is the enthalpy of reaction at T0 (298 K) and Cp
r
is the heat capacity of the products at T0. The optimal com-
position of urea was worked out using the total oxidizing and
reducing valencies of the components, so that the equivalence
ratio is unity [30]. The reducing valences of ammonium hep-
tamolybdate and urea are +18 and +6, respectively, whereas
the oxidizing valence of nickel (II) nitrate is −10, consider-
ing the corresponding valencies of the metals, −2 for oxygen
and zero (0) for nitrogen.
Considering the reducing/oxidizing valencies of the start-
ing materials and the required catalyst composition (i.e.,
Ni/Mo molar ratio), it is possible to determine the amount
of urea necessary to obtain an optimal redox mixture. For
instance, for the synthesis of NiMoO4 (Ni/Mo = 1) and
assuming that under stoichiometric composition urea decom-
posed to CO2, H2O and N2 the overall reaction is:
21Ni(NO3)2·6H2O + 3(NH4)6Mo7O24·4H2O + 26CO(NH2)2
→ 21NiMoO4 + 226H2O + 26CO2 + 56N2
(2)
Using the relevant thermodynamic data [25,33–35], we cal-
culated the enthalpy change involved in the chemical reaction
(2) and subsequently the adiabatic temperature employing the
Eq. (1).
The adiabatic temperatures and the redox mixture com-
positions against the catalyst molar compositions are repre-
sented graphically in Fig. 3. This reveals that small increases
of the Mo/Ni molar ratio produce strong rises of the adia-
batic temperature until the system achieves temperatures that
induce phase transitions and likely Mo-species volatilization
when Mo/Ni ratio is greater than 1. This fact attenuates the
increase of the adiabatic temperature. On the other hand, the
urea-oxidizer mixture composition continuously decreases
as the Mo/Ni mole ratio is raised. Considering this oppo-
site behaviour it is inferred that ammonium heptamolybdate
Fig. 4. Laser Raman spectra of the ␥-alumina-supported Ni-Mo precursor
salts with different formulations and a constant urea/(Ni + Mo) mole ratio of
10. (a) Urea, (b) 12 wt.% MoO3/␥-Al2O3, (c) 3 wt.% NiO-12 wt.% MoO3/␥-
Al2O3, (d) 12 wt.% NiO/␥-Al2O3.
Fig. 3. Influence of the Mo/Ni molar ratio on the adiabatic temperature and
the oxidizer–fuel mixture.

220
S.L. Gonza´lez-Corte´s et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 214–225
the catalyst precursors is proposed in Fig. 6. The impregna-
tion or deposition of the precursor salts and urea on alumina at
room temperature leads to the formation of poorly-organized
6-molybdoaluminate species as major phase. When the tem-
perature reaches the melting point of urea (i.e. 135 ◦C), it
is spread out over alumina surface and the Ni and Mo salts
are partially dissolved and dehydrated, a further increase of
the temperature up to 160–175 ◦C produces a partial decom-
position of the precursor salts and of the fuel. It is envis-
aged that this step facilitates the precipitation of the partially
decomposed Ni and Mo precursor salts since the content of
urea is significantly reduced. When the sample achieves the
ignition temperature (ca. 200 ◦C) a violent exothermic reac-
tion occurs (simultaneously with the urea thermolysis when
a high urea amount is used), releasing heat and different
gaseous molecules as a consequence of the total transfor-
mation of urea and the precursors to nickel and molybdenum
oxides.
Fig. 5. Laser Raman spectra of the ␥-alumina-supported 3Ni-12Mo precur-
sor salts with different urea/(Ni + Mo) mole ratio. (a) 0.0, (b) 0.25, (c) 1.0,
(d) 5.0, (e) 10.
Raman peak at 1051 cm−1 characteristic of the NO3 anion
−
The calculated adiabatic temperatures for the synthesis of
single and mixed oxides from the nickel nitrate and ammo-
nium heptamolybdate is in the same magnitude order as those
determined for synthesis of different single and mixed oxides
(i.e. from 700 to 1600 ◦C) [13,30,31]. The nature of combus-
tion differs from flaming to non-flaming (smouldering) and
also from the final product characteristics and its microstruc-
ture [13,30]. For the preparation of Al2O3-supported oxides
this temperature is likely lowered not only by the stoichio-
metric excess of urea, but also the lower content of the Ni
and Mo precursor salts and the combustion-energy dissipa-
tion through the support. Taking into account these factors
and the low alumina reactivity under combustion condition
[15], it is thought that the transformation of the Mo and Ni
precursor salts proceeds under a reaction of solid state com-
bustion, hence avoiding the molybdenum oxide loss by sub-
limation or vaporization. Even though, the UMxC-prepared
oxidic precursors was obtained from a combustion process at
the ignition temperature of 500 ◦C instead of a temperature-
programmed combustion, one can envisage that in a very
short period of time (∼30 s) the sample undergoes water-
removal and melting of urea followed by the combustion
reaction to produce the oxidic catalyst precursor as was men-
tioned earlier.
fromthepromoterprecursorsaltisillustrated[42,43]whereas
Raman peaks characteristic of Ni-oxo species are not clearly
observed (Fig. 4d) owing to its weak Raman scatterers.
The influence of the urea on the initial aggregated state
of 3Ni-12Mo precursor salts supported on ␥-alumina is
illustrated in Fig. 5. The spectra of the samples show the
main Raman features of 6-molybdoaluminate entities (i.e.,
Anderson-type heteropolyanion) [39–41], dispersed on alu-
mina (see Fig. 5a). On the other hand, the spectra of the
urea-containing catalyst precursors suggest that urea partially
hinders the formation of the Anderson-type heteropolyan-
ion. In addition, the main Raman band of the urea matrix at
1012 cm−1, characteristic of the symmetrical C N stretch-
ing vibrational [37], is slightly shifted to lower wavenumbers
and becomes broader with urea composition decreasing. This
indicates the presence of a weak interaction between urea and
the Lewis-type acidic species, probably Ni2+ or even cationic
species of partially-dissolved alumina.
Several consecutive stages such as melting, dissolution
and chemical reactions take place before and upon com-
bustion process to transform the urea and the Ni and Mo
precursor salts to well-dispersed oxidic species over alumina
surface. A schematic representation of the different states of
Fig. 6. Schematic representation of the combustion process occurring over the alumina surface upon preparation of Ni-Mo catalyst precursor.

S.L. Gonza´lez-Corte´s et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 214–225
221
Fig. 7. X-ray diffraction patterns of the ␥-Al2O3-supported Ni-Mo oxi-
dic precursors prepared by urea–matrix combustion. (a) 12 wt.% MoO3/␥-
Al2O3, (b) 1 wt.% NiO-12 wt.% MoO3/␥-Al2O3, (c) 3 wt.% NiO-12 wt.%
MoO3/␥-Al2O3, (d) 6 wt.% NiO-12 wt.% MoO3/␥-Al2O3, (e) 12 wt.% NiO-
12 wt.% MoO3/␥-Al2O3, (f) 12 wt.% NiO/␥-Al2O3.
Fig. 8. Laser Raman spectra of 3 wt.% NiO-12 wt.% MoO3/␥-Al2O3 cata-
lyst under different stages of preparation. (a) Dried in high vacuum at room
temperature, (b) combusted at 500 ◦C for 10 min, (c) sulfurized at 500 ◦C
using thiophene as sulfurizing agent.
3.2. Characterization and HDS performance of the
γ-Al2O3-supported Ni-Mo catalysts
new broad peak at ∼37 and 43.3◦ in 2θ characteristic of NiO
(PDF No. 55-1159) are clearly observed (see Fig. 7e and
f). These findings indicate that nickel loadings lower than
6.0 wt.% lead to the formation of highly dispersed and/or
poorly-ordered Ni-oxo species on alumina surface whereas
higher promoter loadings cause the formation and segrega-
tion of NiO. The XRD patterns of the sulfurized samples (not
shown) suggest the formation of poorly crystalline and/or
highly dispersed Mo and Ni sulfide on ␥-alumina since their
XRD peaks were not observed.
3.2.1. X-ray diffraction and laser Raman spectroscopy
The powder X-ray diffraction patterns of the monometal-
lic (Ni or Mo) and bimetallic (Ni and Mo) catalyst oxi-
dic precursors are shown in Fig. 7. The XRD pattern of
12 wt.% MoO3/␥-Al2O3 shows the main diffraction peaks
of ␥-Al2O3 support (PDF No. 48-367), taken from the PDF-
ICDD database crystalline materials [44]. Typical features of
molybdenum oxides are not observed owing to a high disper-
sion on alumina and/or poor crystalline structure, Fig. 7a. On
the other hand, the bimetallic oxidic precursors with variable
nickel loading display no defined additional XRD peaks up to
6.0 wt.% however for greater nickel loading (i.e., 12 wt.%) a
Fig. 8 compares the effect of the different stages (i.e.,
drying, combustion and sulfurization) on a typical sample.
Again the dried sample displayed the major peaks charac-
teristic of urea matrix [37,38] (Fig. 8a). The Raman bands
Fig. 9. HRTEM images of the Ni-MoS2/␥-Al2O3 HDS catalysts prepared by urea–matrix combustion method.

222
S.L. Gonza´lez-Corte´s et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 214–225
associated with the Mo-oxo species are also observed. After
combustion, the spectrum displayed a dominant asymmet-
symmetry) and the latter is attributed to A1g disorder mode
(D peak) because of the small crystallite size of graphite
[52–54]. In addition, this spectrum shows two weak peaks
at 369 and 397 cm−1 attributed to strongly distorted-MoS2
along the basal plane since the Mo S stretching mode (E21g)
along the basal plane and the S Mo S bond mode along the
c-axis (A1g) is observed at 381 and 405 cm−1 for a well-
ordered MoS2 structure [12]. From these findings one can
conclude that poorly-ordered Mo-oxo entities over alumina
surface form polymolybdate species after the combustion
process. These species lead to the formation of strongly
distorted-MoS2 and graphitic carbon when thiophene is used
as sulfurizing agent.
ric feature at 948 cm−1 and a small band at 344 cm−1
,
attributed to the symmetric and bending mode of the ter-
minal Mo O bond of octahedrally coordinated Mo species
6−
for hydrated Mo7O24
anion, respectively (Fig. 8b). In
addition, the Raman bands at ∼556 and 222 cm−1 are asso-
ciated with the symmetric and deformation modes of the
Mo O Mo bridge for the polymolybdate species [45–49],
respectively. The characteristic Raman features of NiO were
not observed, likely due to the incorporation of the pro-
moter into the structure of the Mo-oxo anion. In fact, it
has been claimed that isopolymolybdate of nickel [50] and
Ni heteropolymolybdates [51] might lead to the formation
of the HDS active phase. The Raman peaks of urea and
residual carbon were not detected either, indicating a total
extension of the combustion reaction in line with the TGA
results.
Representative HRTEM images in different nano-scale
of 3Ni-12MoS2/␥-Al2O3 catalyst are shown in Fig. 9. The
micrographs display mainly the edge or prism planes ori-
ented along or roughly parallel to the electron beam direction
˚
[55,56] with layer stacking spacing of ca. 6 A characteristic
On the other hand, the Raman spectrum of the sulfurized
catalyst reveals two major Raman features: a strong peak at
1589 and a broad band around 1359 cm−1 characteristic of
carbon material deposited during sulfurizing process when
thiophene is employed as sulfurizing agent (Fig. 8c). The
former is associated with the G peak of graphitic carbon (E2g
of bulky MoS2 [57]. However, other orientations such as the
basal planes and the (1 1 0) orientation were also observed but
in less extent [58]. In addition, dislocations where the slabs
are partially intercalated by another slab and bent on a longer
scale to maximize the interaction with the surface are also
observed, see image Fig. 9b. On the other hand, the degree of
Fig. 10. (A) Catalytic behaviour and (B) and (C) product distribution vs. time on stream of ␥-Al2O3-supported Ni-Mo HDS catalysts at 350 ◦C, p = 0.1 MPa
and υ = 100 mL (gcat min)−1
.

S.L. Gonza´lez-Corte´s et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 214–225
223
stacking is around 3–4 layer and the slab length distribution
are fairly broad with an average slab length of ca. 3 nm, owing
to the relatively high sulfurizing temperature (i.e., 450 ◦C).
The images are quite similar to the Co-promoted MoS2/␥-
Al2O3 HDS catalysts using thiophene as sulfurizing agent as
well [12].
nation of intermediate thiophene compounds (HDS) [59], the
C4/C4 mole ratio can be a convenient indicator of the rela-
=
tive rate of HYD to HDS reactions. It was observed a strong
increase of the SE at low nickel composition achieving a max-
imum in Ni/(Ni + Mo) mole ratio of 0.33 and then decrease
at higher content. An opposite trend but less pronounced was
noted for the C4/C4= mole ratio. Indeed, at approximately iso-
conversion (i.e., 50%), the relative rate of HYD to HDS for
Ni/(Ni + Mo) of 0.14 increases just by 20% compared with
the Ni mole fraction of 0.66, indicating that a slight antago-
3.3. HDS catalytic performance
The time-dependence of thiophene conversion for ␥-
Al2O3-supported Ni-Mo HDS catalysts with variable Ni
and fixed Mo compositions is shown in Fig. 10A. The Ni-
promoted Mo catalysts showed thiophene conversions four to
five times greater than those for the unpromoted Mo catalyst.
Furthermore, the Ni-Mo HDS catalysts illustrated practically
invariable HDS activity with the time on stream. Except for
1Ni-12Mo catalyst, which underwent a slight decrease of
the thiophene conversion similar to the monometallic (Ni or
Mo) catalysts, achieving the steady state after 10 h of reac-
tion. On the other hand, the C4 product distribution obtained
through the removal of sulphur from thiophene (C4H4S) was
not significantly affected by neither the time on stream nor
the presence of nickel since C4 is the major product for
Mo-containing catalysts (i.e. greater than 99%) [8]. Also,
Ni monometallic catalyst showed C4 selectivity greater than
90% and the remaining part was of lighter hydrocarbons
(C1–C3 compounds) (Fig. 10B). These findings indicate that
the promoter effect is closely linked to HDS activity rather
than the selectivity to C4 compounds. The C4/C4= mole ratio
again did not change significantly with the time on stream
however this was clearly affected by the nickel composition,
Fig. 10C.
=
nism effect on the C4/C4 mole ratio is present. Therefore,
synergistic effect increases markedly the C S bond cleavage
reaction but to a much lesser extent affects the relative rate
of hydrogen transfer reactions.
The optimal composition and the qualitative behaviour of
the promoter effect of nickel on Mo-based HDS catalyst are
rathersimilartotheearlierreportsofAl2O3-supportedNi-Mo
catalysts [60,61]. However, the optimal promoter effect (i.e.,
k
Ni-Mo/kMo) of the combustion-prepared catalyst was slightly
ˇ
lower (ca. 15%) than that recently given by Zdrazil over
impregnation-prepared Ni-Mo catalyst [62]. This is probably
due to the coverage of the most active HDS sites by deposition
of graphitic carbon upon sulfurizing process when thiophene
is employed as a sulfurizing agent and/or the lateral growing
of MoS2 by sintering.
An HDS apparent activation energy (Eaapp) derived from
Arrehenius plot of 59 8 kJ/mol for the catalyst with opti-
mal composition (i.e., 3Ni-12Mo) was obtained. This is in
rather good agreement with the figure reported by Borgna et
al. [63,64] and Ledoux et al. [65]. This suggests that carbon
deposited from the sulfurizing process does not significantly
inhibit the hydrodesulfurization reaction, probably due to its
participation in the formation of the HDS active sites, as
has been previously reported [12,66,67]. Therefore, the main
cause of the slightly lower promoter effect is due to the lat-
eral growing of MoS2 rather than the deposition of carbon. At
higher temperatures of 583 K, on the other hand, the apparent
activation energy decreases to 37 kJ/mol. This behaviour is
not ascribed to mass-transfer limitations since experiments
with different catalysts particles sizes revealed that these
limitations are absent. This result has been attributed to the
changes in the rate limiting step in the HDS reaction mech-
anism [68] and/or to a decrease of reactant surface coverage
as temperature increases [64,69].
Fig. 11 illustrates the influence of the relative Ni molar
=
composition on the synergistic effect (SE) and the C4/C4
mole ratio. Since butane is a molecule generated by hydro-
genation (HYD) of C4-unsaturated compounds produced
through the hydrogenolysis of thiophene and/or ␤-H elimi-
It must be pointed out that the ␥-alumina-supported 3Ni-
12Mo HDS catalysts prepared using different urea/(Ni + Mo)
mole ratio and also combusted at different ignition temper-
atures (i.e. 300–600 ◦C) for 10 min showed similar catalytic
behaviour (i.e., HDS activity and product distribution) than
that presented in Fig. 11. This indicates that both parameters
do not affect significantly the catalytic properties of 3Ni-
12Mo HDS catalyst, probably due to a stabilizing effect of
the alumina upon the combustion process and/or to the carbon
material deposited during sulfurization process. Indeed, Chi-
anelli and co-workers have found that the interaction between
the carbon and the edge planes of MoS2 not only favours the
Fig. 11. Dependence of the synergistic effect (SE) and the C4/C4= mole ratio
on the Ni-Mo catalyst composition.

224
S.L. Gonza´lez-Corte´s et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 214–225
formation of surface “carbonsulfide” phase, but also stabi-
lizes texturally the sulfide particles [66,67], which would be
reflected in a remarkable catalytic stability.
[9] L. Qu, W. Zhang, P.J. Kooyman, R. Prins, J. Catal. 215 (2003) 7.
[10] A. Calafat, J. Laine, A. Lo´pez-Agudo, J.M. Palacios, J. Catal. 162
(1996) 20.
[11] J.L. Brito, J. Laine, J. Catal. 139 (1993) 540.
[12] S.L. Gonza´lez-Corte´s, T-C. Xiao, P.M.F.J. Costa, B. Fontal, M.L.H.
Green, Appl. Catal. A: Gen. 270 (2004) 209.
[13] K.C. Patil, S.T. Aruna, T. Mimani, Curr. Opin. Solid State Mater.
Sci. 6 (2002) 507.
4. Concluding remarks
[14] G. Xanthopoulou, G. Vekinis, Appl. Catal. B: Environ. 19 (1998)
37.
A new approach named urea–matrix combustion method
based on combustion and decomposition processes has
been successfully applied for the first time to prepare
alumina-supported Ni-Mo oxidic precursor. Several consec-
utive stages such as melting, dissolution and chemical reac-
tions take place before and during combustion process. The
deposition of the precursor salts and urea as fuel on alu-
mina at room temperature leads to the formation of poorly-
crystallized Mo-oxo species as the major phases. After the
combustion process well-dispersed Ni (at compositions not
greater than 6 wt.%) and polymolybdate species on alumina
surface are formed. There is not evidence of residual carbon
after combustion process. The sulfurization process using
thiophene in H2 as sulfurizing agent facilitates the forma-
tion of graphitic carbon and a strongly distorted-MoS2 type
structure.
[15] J.J. Moore, H.J. Feng, Prog. Mater. Sci. 39 (1995) 275.
[16] J.J. Moore, H.J. Feng, Prog. Mater. Sci. 39 (1995) 243.
[17] G. Xanthopoulou, G. Vekinis, Appl. Catal. A: Gen. 199 (2000) 227.
[18] G. Xanthopoulou, Appl. Catal. A: Gen. 185 (1999) L185.
[19] P. Bera, K.C. Patil, V. Jayaram, G.N. Subbanna, M.S. Hegde, J.
Catal. 196 (2000) 293.
[20] P. Bera, K.C. Patil, M.S. Hegde, Phys. Chem. Chem. Phys. 2 (2000)
3715.
[21] M. Adachi, C. Contescu, J.A. Schwarz, J. Catal. 162 (1996)
66.
[22] H. Kno¨zinger, P. Ratnasamy, Catal. Rev.-Sci. Eng. 17 (1978) 31.
[23] R. Murugan, H. Chang, J. Chem. Soc. Dalton Trans. 20 (2001)
3125.
[24] I.B. Sharma, S. Batra, J. Therm. Anal. 41 (1994) 1075.
[25] D.R. Lide (Ed.), CRC Handbook of Chemistry and Physics, 84th
ed., CRC press, Boca Raton, 2003.
[26] M.A.A. Elmasry, A. Gaber, E.M.H. Khater, J. Thermal. Anal. 52
(1998) 489.
The urea–matrix combustion method is a simple method
and the catalysts can be prepared in a short period of ther-
mal treatment. Urea employed as fuel not only increases
the combustion rate, but also undergoes a decomposition
process (endothermic reaction) that could contribute to the
reduction of the combustion temperature. This method gives
highlyactive␥-Al2O3-supportedNi-MoHDScatalyst, which
shows a comparable promoter effect to that of impregnation-
prepared catalyst. In addition, an opposite relation between
the activity and the hydrogenation properties was observed
indicating that highly active HDS catalyst requires low con-
sumption of hydrogen. Finally, both the ignition temperature
and the urea-oxidizer ratio produce no significant changes in
the HDS catalytic properties of Ni-Mo-based catalysts.
[27] J.B. Peri, J. Phys. Chem. 69 (1965) 211.
[28] J.H. de Boer, J.M.H. Fortuin, B.C. Lippens, W.H. Meijs, J. Catal. 2
(1963) 1.
[29] S.R. Jain, K.C. Adiga, V.R. Pai Verneker, Combust. Flame 40 (1981)
71.
[30] K. Suresh, K.C. Patil, in: K.J. Rao (Ed.), Perspective in Solid State
Chemistry, Narosa Publishing House, New Delhi, 1995, p. 376.
[31] D.A. Fumo, M.R. Morelli, A.M. Segada˜es, Mater. Res. Bull. 31
(1996) 1243.
[32] J.B. Holt, Z.A. Munir, J. Mater. Sci. 21 (1986) 251.
[33] J.A. Dean (Ed.), Lange’s Handbook of Chemistry, 12th ed., McGraw-
Hill, New York, 1979.
[34] M.T. Colomer, D.A. Fumo, J.R. Jurado, A.M. Segadaˆes, J. Mater.
Chem. 9 (1999) 2505.
[35] M. Morishita, A. Navrotsky, J. Am. Ceram. Soc. 86 (2003) 1937.
[36] A.C.F.M. Costa, M.R. Morelli, R.H.G.A. Kiminami, J. Mater. Synth.
Process. 9 (2001) 347.
[37] R. Keuleers, H.O. Desseyn, B. Rousseau, C. Van Alsenoy, J. Phys.
Chem. A. 103 (1999) 4621.
[38] B. Rousseau, C. Van Alsenoy, R. Keuleers, H.O. Desseyn, J. Phys.
Chem. A. 102 (1998) 6540.
[39] X. Carrier, J.F. Lambert, M. Che, J. Am. Chem. Soc. 119 (1997)
10137.
Acknowledgement
´
´
S.L. Gonzalez-Cortes is indebted to the ULA and
FONACIT (Venezuela) for financial support.
[40] L. Le Bihan, P. Blanchard, M. Fournier, J. Grimblot, E. Payen, J.
Chem. Soc., Faraday Trans. 94 (1998) 937.
[41] X. Carrier, J.F. Lambert, S. Kuba, H. Kno¨zinger, M. Che, J. Mol.
Struct. 656 (2003) 231.
References
[1] EP directive 2003/17/EC, Official J. Eur. Union L76, 46 (2003) 10.
[2] I.V. Babich, J.A. Moulijn, Fuel 82 (2003) 607.
[3] R. Prins, H.J. de Beer, G.A. Somorjai, Catal. Rev.-Sci. Eng. 31
(1989) 1.
[42] V. Vendange, D.J. Jones, Ph. Colomban, J. Phys. Chem. Solids 57
(1996) 1907.
[43] M.R. Waterland, D. Stockwell, A.M. Kelley, J. Chem. Phys. 114
(2001) 6249.
[4] R.R. Chianelli, Catal. Rev.-Sci. Eng. 25 (1984) 361.
[5] H. Topsøe, B. Clausen, Catal. Rev.-Sci. Eng. 26 (1984) 420.
[6] P.T. Vasudevan, J.L.G. Fierro, Catal. Rev.-Sci. Eng. 38 (1996)
161.
[44] JCPDS Powder Diffraction File, Int. Centre for Diffraction Data,
Swarthmore, PA, 1989.
[45] G. Mestl, T.K.K. Srinivasan, Catal. Rev.-Sci. Eng. 40 (1998) 451.
[46] F.D. Hardcastle, I.E. Wachs, J. Raman Spectrosc. 21 (1990) 683.
[47] D.S. Kim, K. Segawa, T. Soeya, I.E. Wachs, J. Catal. 136 (1992)
539.
[7] R. Prins, Adv. Catal. 46 (2001) 399.
[8] H. Topsøe, B.S. Clausen, F.E. Massoth, in: J.R. Anderson, M.
Boudart (Eds.), Hydrotreating Catalysis, Science and Technology,
Springer-Verlag, Berlin, 1996.
[48] H. Hu, I.E. Wachs, J. Phys. Chem. 99 (1995) 10911.
[49] G. Mestl, J. Mol. Catal. A: Chem. 158 (2000) 45.

S.L. Gonza´lez-Corte´s et al. / Journal of Molecular Catalysis A: Chemical 240 (2005) 214–225
225
[50] S. Kasztelan, J. Grimblot, J.P. Bonnelle, J. Phys. Chem. 91 (1987)
1503.
[61] V.H.J. de Beer, T.H.M. van Sin Fiet, J.F. Engelen, A.C. van Haandel,
M.W.J. Wolfs, C.H. Amberg, G.C.A. Schuit, J. Catal. 27 (1972)
357.
[51] H. Jeziorowski, H. Knoezinger, Appl. Surf. Sci. 5 (1980) 335.
[52] F. Tuinstra, J.L. Koenig, J. Chem. Phys. 53 (1970) 1126.
[53] D.S. Knight, W.B. White, J. Mater. Res. 4 (1989) 385.
[54] A. Ermolieff, A. Chabli, F. Pierre, G. Rolland, D. Rouchon, C. Van-
nuffel, C. Vergnaud, J. Baylet, M.N. Se´me´ria, Surf. Interface Anal.
31 (2001) 185.
[62] M. Zdrazˇil, Catal. Today 86 (2003) 151.
[63] A. Borgna, E.J.M. Hensen, L. Coulier, M.H.J.M. de Croon, J.C.
Schouten, J.A.R. van Veen, J.W. Niemantsverdriet, Catal. Lett. 90
(2003) 117.
[64] A. Borgna, E.J.M. Hensen, J.A.R. van Veen, J.W. Niemantsverdriet,
J. Catal. 221 (2004) 541.
[55] F. Delannay, Appl. Catal. 16 (1985) 135.
[56] K.C. Pratt, J.V. Sanders, V. Christov, J. Catal. 124 (1990) 416.
[57] T.F. Hayden, J.A. Dumesic, J. Catal. 103 (1987) 366.
[58] R.M. Stockmann, H.W. Zandbergen, A.D. van Langeveld, J.A.
Moulijn, J. Mol. Catal. A: Chem. 102 (1995) 147.
[59] R. Prins, in: G. Ertl, H. Knozinger, J. Weitkamp (Eds.), Handbook
of Heterogeneous Catalysis, vol. 4, Wiley-VCH, Weinheim, 1997, p.
1908.
[65] M.J. Ledoux, O. Michaux, G. Agostine, J. Catal. 102 (1986) 275.
[66] G. Berhault, L. Cota Araiza, A. Duarte Moller, A. Mehta, R.R.
Chianelli, Catal. Lett. 78 (2002) 81.
[67] G. Berhault, A. Mehta, A.C. Pavel, J. Yang, L. Rendon, M.J. Yaca-
man, L. Cota Araiza, A. Duarte Moller, R.R. Chianelli, J. Catal. 198
(2001) 9.
[68] A.N. Startsev, Catal. Rev.-Sci. Eng. 37 (1995) 353.
[69] E.J.M. Hensen, M.J. Vissenberg, V.H.J. de Beer, J.A.R. van Veen,
R.A. van Santen, J. Catal. 163 (1996) 429.
[60] S.P. Ahuja, M.L. Derrien, J.F. Le Page, Ind. Eng. Chem. Prod. Res.
Dev. 9 (1970) 272.