The effect of Mg and K addition to a Mo2C/HY catalyst for the hydrogenation and ring opening of naphthalene

Article abstract of DOI:10.1016/j.catcom.2010.10.025

The effect of Mg (0.5-2 wt.%) or K (1 wt.%) added to a 20 wt.% Mo ₂C/HY catalyst for the hydrogenation and ring opening of naphthalene, is reported. Mg addition increased Mo₂C dispersion without affecting catalyst acidity, whereas K increased Mo₂C dispersion but decreased acidity by 30%. Both Mg-Mo₂C/HY and K-Mo₂C/HY increased naphthalene and coke-precursor hydrogenation whereas ring-opening selectivity was reduced compared to the Mo₂C/HY. Metal dispersion and catalyst acidity are key parameters that determine bifunctional catalyst performance. Although Mg and K addition improved Mo₂C dispersion they did not provide the optimum acidity needed for increased ring-opening selectivity.

Full text of DOI:10.1016/j.catcom.2010.10.025

Catalysis Communications 12 (2011) 454–458
Contents lists available at ScienceDirect
Catalysis Communications
journal homepage: www.elsevier.com/locate/catcom
Short Communication
The effect of Mg and K addition to a Mo2C/HY catalyst for the hydrogenation and ring
opening of naphthalene
⁎
Xuebin Liu, Shahrzad Jooya Ardakani, Kevin J. Smith
Department of Chemical and Biological Engineering, The University of British Columbia, 2360 East Mall, Vancouver, BC, Canada V6T 1Z3
a r t i c l e i n f o
a b s t r a c t
Article history:
The effect of Mg (0.5–2 wt.%) or K (1 wt.%) added to a 20 wt.% Mo2C/HY catalyst for the hydrogenation and
ring opening of naphthalene, is reported. Mg addition increased Mo2C dispersion without affecting catalyst
acidity, whereas K increased Mo2C dispersion but decreased acidity by 30%. Both Mg-Mo2C/HY and K-Mo2C/
HY increased naphthalene and coke-precursor hydrogenation whereas ring-opening selectivity was reduced
compared to the Mo2C/HY. Metal dispersion and catalyst acidity are key parameters that determine
bifunctional catalyst performance. Although Mg and K addition improved Mo2C dispersion they did not
provide the optimum acidity needed for increased ring-opening selectivity.
Received 2 September 2010
Received in revised form 24 October 2010
Accepted 27 October 2010
Available online 3 November 2010
Keywords:
Ring-opening
Hydrogenation
Catalyst
© 2010 Elsevier B.V. All rights reserved.
Molybdenum carbide
HY zeolite
Bifunctional catalyst
Carbon deposition
Catalyst deactivation
1. Introduction
20 wt.% Mo2C loading, whereas the selectivity to hydrogenation and RO
products remained relatively constant. The results from this study also
Production of synthetic crude oil (SCO) from the Canadian oilsands
is expected to double in the next 10 years. SCO is a light sweet crude
with some significant quality challenges, largely related to high
aromaticity. The heavy gas oil (HGO) bottom cut from SCO (about
38 vol.%) contains as much as 90% cycloparaffins plus aromatics,
which cause processing problems, particularly in fluid catalytic
cracking (FCC) units. Hydrogenation and ring opening (RO) of the
cycloparaffins and aromatics is one approach to address the quality
issues and refining challenges associated with SCO [1].
demonstrated that the catalysts underwent significant deactivation
during reaction with naphthalene. Carbon deposition (coke formation)
over the acidic HY support was identified as the main cause of the fast
deactivation with naphthalene as reactant [9]. Reduced acidity of the HY
support, as a consequence of increased Mo2C loading decreased the
catalyst deactivation. These results were described in terms of a
bimolecular reaction mechanism on acid sites that led to coke formation
competing with a monomolecular reaction mechanism on metal and
acid sites that yield hydrogenation and RO products [9–11].
Metal catalysts dispersed on acidic supports are preferred for
selective ring-opening and a number of Group VIII metals have been
reported to have good activity and selectivity for RO, including Rh, Ru, Ir,
Pt, Pd/Pt and Pt/Ir [2–7]. However, these metals are costly and have
relatively low resistance to sulphur poisoning. In previous work,
molybdenum carbide supported on zeolites of medium acidity were
identified as potential alternatives to conventional supported Group VIII
metal catalysts for RO [8]. Subsequent studies showed that increased
Mo2C loading decreased the catalyst acidity, as measured by temper-
ature-programmed desorption of n-propylamine, and increased the
Mo2C dispersion, as measured by XRD and XPS [9]. With naphthalene
as reactant, an increased Mo2C loading increased conversion up to a
Metal dispersion and catalyst acidity are key parameters that
determine bifunctional catalyst performance. In the present work, the
effect of Mg and K added to a bifunctional Mo2C/HY catalyst is
reported. The impact on Mo2C dispersion and acidity is discussed in
terms of the RO selectivity and stability of the catalysts measured
during naphthalene hydrogenation and ring-opening.
2. Experimental
2.1. Catalyst preparation
Catalysts with 20 wt.% Mo2C supported on HY zeolite and doped
with 0.5–2 wt.% Mg or 1 wt.% K, were prepared by co-impregnation of
the zeolite (ZeolystTM CBV720, SiO2/Al2O3 =30) using a quantified
aqueous solution of ammonium heptamolybdate tetrahydrate (MoO3
81.0–83.0%, Sigma) mixed with magnesium nitrate hexahydrate (99%,
⁎
Corresponding author.
E-mail address: kjs@interchange.ubc.ca (K.J. Smith).
1566-7367/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.catcom.2010.10.025

X. Liu et al. / Catalysis Communications 12 (2011) 454–458
455
Sigma) or potassium nitrate (99.97%, BDH Chemicals). After aging at
70 °C for 2 h, the impregnated support was dried at 120 °C for 12 h
and calcined at 500 °C for 4 h. Carburization of the calcined catalyst
precursor was done in a quartz U-tube reactor (i.d.=7 mm) by
temperature programmed reaction in a 100 ml (STP)/min of 20%CH4/
80%H2 (UHP) and a ramp rate of 5 °C/min up to 700 °C. The catalyst
was held at 700 °C for 4 h and then the gas was switched to H2 (UHP)
to remove excess carbon by holding at 700 °C for a further 2 h before
cooling to room temperature. Finally, the catalysts were passivated at
room temperature in a 1% O2 in He stream (150 ml (STP)/min) before
exposure to air. The resulting MgO-Mo2C/HY and K2O-Mo2C/HY
catalysts were subsequently characterized and tested as described
below.
grouped as hydrogenation (Hydro), ring opening (RO) and polyaro-
matics (POLY). RO products were mainly alkylcyclohexanes, alkyl-
benzenes and alkylindenes and POLY products were aromatics and
naphthenes with more than 10 carbon atoms, mainly alkyltetralins.
Further details of the activity tests and product analysis procedures
have been described previously [8,9].
3. Results and discussion
Following calcination, carburization and passivation, the Mg-Mo2C/HY
and K-Mo2C/HY catalysts had significantly reduced surface area, total pore
volume, micropore area and micropore volume (Table 1) compared to the
Mo2C/HY catalyst. However, the surface area was not strongly dependent
upon the amount of Mg added, with all the Mg-Mo2C/HY catalysts having
surface areas in the range 389–462 m2/g, compared to 522 m2/g for the
Mo2C/HY. The relative loss in BET and micropore area with Mg or K
addition to the Mo2C/HY, was almost the same. Fig. 1 shows the micropore
size distribution of the 2%Mg-Mo2C/HY catalyst compared to the HY
zeolite. Clearly, there was no significant change in the zeolite micropore
size distribution upon addition of Mg, although there was a small
reduction in micropore volume (Table 1). The pore size distribution data
of Fig. 2 shows the reduced volume associated with the mesopores in the
range 30–50 Å for the 2%Mg–Mo2C/HY compared to the Mo2C/HY and the
HY alone. Together these data suggest that the MgO/K2O(themetaloxides
formed during calcination) are located throughout the zeolite mesopores
outside of the zeolite micropores. The size of the Mo2C crystallites
(Table 1) confirm that the Mo2C was located outside of the zeolite
micropores [9] and so the reduction in the micropore volume must be due
to pore-mouth blocking by Mo2C or MgO/K2O crystals.
2.2. Catalyst characterization and testing
X-ray diffraction (XRD) was performed on the passivated catalysts
using a Siemens D500 diffractometer with a Cu Kα X-ray source and
crystallite size (dc) estimates were made using the Scherrer equation, as
described previously [8,9]. The BET surface area and pore volume of the
passivated catalysts were determined from N2 adsorption–desorption
isotherms measured at −196 °C using a Micromeritics ASAP 2020
analyzer. Catalysts were degassed at 250 °C for 24 h under vacuum
(5 μm Hg) before being analyzed. Thermogravimetric analysis (TGA) of
the catalysts before and after reaction with naphthalene was done in a
flow of dry air (16 ml (STP)/min) and a ramp rate of 5 °C/min to the final
temperature of 850 °C using a TGA-50 thermogravimetric analyzer
(Shimadzu, Japan). The weight gain and loss were quantified according
to the methods described previously [9] and differential TGA (DTGA)
was accomplished using standard TGA software.
Catalyst acidities were determined by n-propylamine (nPA) pulsed
adsorption and TPD using a Micromeritics Autochem II 2920 unit
equipped with a thermal conductivity detector (TCD). Approximately
40 mg of catalyst was pretreated in He (30 ml (STP)/min) at 120 °C for
1 h to remove water, followed by a 1 h treatment in H2 (30 ml (STP)/
min) at 400 °C to activate the passivated Mo2C, and finally a 2 h flush
in He (30 ml (STP)/min) at 500 °C to remove adsorbed H2. While
under He flow, the catalyst was cooled to 120 °C and a flow of He
(30 ml (STP)/min) saturated at room temperature with n-PA (99+%,
Aldrich) was injected into the He flow in 0.5 ml pulses. Pulses of n-PA
were injected repeatedly until the TCD signal showed no further
adsorption. Following the pulsed adsorption experiments, the system
was purged for 1 h in a He flow (30 ml (STP)/min) to remove residual
amine. Subsequently, TPD of n-PA from 120 °C to 700 °C was
performed at a heating rate of 5 °C/min. The detailed interpretation
of the n-PA adsorption/desorption data have been described previ-
ously [9].
The acidity data, shown in Table 1 for the Mo2C/HY catalysts doped
with different quantities of Mg, give an average acidity of 0.93
0.09 mmol/g, in good agreement with the acidity of the un-doped
Mo2C/HY catalyst (0.97 mmol/g). Hence we conclude that the addition
of Mg had no significant effect on the acidity compared to the Mo2C/HY
catalyst, whereas the K decreased the acidity by approximately 30%
(0.68 mmol/g). The impact of the Mg- and K-oxides is expected to
depend on their dispersion, basicity, and their ability to ion exchange
with the HY zeolite. The amount of Mg and K added to the Mo2C/HY was
well below the ion exchange capacity of the HY. Clearly, some K was
exchanged with the acid sites of the HY support during preparation, but
this did not occur in the case of Mg.
The XRD analysis (Fig. 3) of the catalysts showed that the added Mg
or K decreased the Mo2C crystal size (Table 1 and Fig. 3). The improved
Mo2C dispersion is also reflected in the activity data of Table 2. The
naphthalene conversion on the Mg- and K-doped Mo2C/HY catalysts
was marginally higher than on the Mo2C/HY, whereas the selectivity to
hydrogenated products increased significantly and the polyaromatic
and RO product selectivity decreased. These results suggest an increased
hydrogenation capability of the doped catalysts, that not only affected
the product selectivity, but that would also be expected to impact
catalyst deactivation by reducing coke formation.
The RO performance of each of the catalysts was determined in a
stainless steel fixed-bed reactor (i.d.=9 mm) at 300 °C and a total
pressure of 3.0 MPa with 5% naphthalene in n-heptane as the model
reactant. The reaction products were analysed by gas chromatography
(GC) and gas chromatography/mass spectroscopy (GC/MS) and
Table 1
Properties of Mg-Mo2C/HY, K-Mo2C/HY and Mo2C/HY catalysts used for naphthalene hydrogenation and ring-opening.
Catalyst
BET area
m2/g
Total pore volume
cm3/g
Micropore area
m2/g
Micropore volume
cm3/g
Average pore diam.
nm
dc of Mo2C by XRD
nm
Total acidity
mmol/g
HY
783
522
451
389
430
462
444
0.49
0.36
0.34
0.30
0.30
0.30
0.32
543
385
322
272
324
358
337
0.25
0.18
0.15
0.13
0.15
0.17
0.16
2.52
2.76
2.99
3.09
2.82
2.60
2.84
–
1.80
0.97
0.85
0.91
–
20%Mo2C/HY
22
13
11
13
10
14
0.5%Mg-20%Mo2C/HY
1.0%Mg-20%Mo2C/HY
1.5%Mg-20%Mo2C/HY
2.0%Mg-20%Mo2C/HY
1.0%K-20%Mo2C/HY
1.02
0.68

456
X. Liu et al. / Catalysis Communications 12 (2011) 454–458
Fig. 3. XRD of Mg-Mo2C/HY and K-Mo2C/HY catalysts.
Fig. 1. Micropore size distribution of HY and 2%Mg-20%Mo2C/HY catalyst.
dispersion that results in stronger interaction between theMo2C and the
HY support, yielding Mo2CxOy species and making the Mo2C less
oxidizable.
The results of the DTGA analysis of the un-used and used catalysts
from the present study are shown in Figs. 4 and 5, respectively. There
were two peaks that showed a weight increase during the DTGA of the
un-used catalysts (Fig. 4). The peaks appearing at low temperature
(b350 °C) are attributed to the oxidization of well dispersed Mo2C,
while those emerging at higher temperature are attributed to the
oxidation of a Mo2C “bulk-phase” species [9]. As shown in Fig. 4, the
second peak decreased with increased Mg loading, and this peak nearly
disappeared at a Mg loading of 2%, indicating that at this level of Mg
loading the Mo2C was well dispersed onto the HY support, consistent
with the XRD results (Fig. 3). A similar analysis of the second weight
gain peak of the DTGA spectra shows that the Mo2C dispersion of the
K-doped catalyst was higher than that of the 0.5% Mg-doped catalyst,
but not as high as that obtained for the 1% Mg-doped catalyst. These
observations are also consistent with the XRD results of Table 1 and
Fig. 3. The summary of the TGA analysis, shown in Table 3, demonstrates
that the total weight gain of the un-used catalysts decreased from 6.1%
to 4.0% as the Mg loading increased from 0.5% to 2%. As noted previously,
the weight gain was mainly caused by the oxidation of Mo2C. This
suggests that the amount of oxidizable Mo2C on the Mo2C/HY catalysts
decreased in the presence of Mg, probably because of a higher Mo2C
Results from TGA/DTGA of the Mg-Mo2C/HY and K-Mo2C/HY
catalysts after 5 h of reaction with 5% naphthalene in heptane are
shown in Fig. 5 and Table 3. Comparison of the spectra in Figs. 4 and 5
shows a larger weight loss below 150 °C for the used catalysts than the
un-used catalysts. This means that besides water, some volatile
hydrocarbons remained on the catalysts after reaction. The weight
loss of the used catalysts at low temperature influenced the analysis of
the weight gain at low temperature (because of some peak overlap).
Therefore, the weight gain of the used catalysts (due to Mo2C
oxidation) was smaller than that of the corresponding un-used
catalysts, as shown in Table 3.
Weight loss peaks between 350 and 550 °C (Fig. 5) are attributed
to carbon removal from the used catalysts [9]. However, there is some
overlap between these weight loss peaks and the weight gain peak
between 350 and 550 °C shown in the DTGA of the un-used
catalysts (Fig. 4), and this weight loss was accounted for in the
used catalyst weight gain data reported in Table 3. After reaction,
the Mg- and K-doped Mo2C/HY catalysts had a weight loss of 3.9
0.6 wt.%, significantly smaller than the 7 wt.% of the un-doped Mo2C/HY
catalyst and indicative of less coke deposition and catalyst deactivation
for the Mg- and K-doped Mo2C/HY catalysts compared to the Mo2C/HY.
With increasing Mg content from 0.5% to 2%, the amount of coke formed
on the catalyst did not change significantly. We speculate that coke
accumulation on the HY support was not effectively inhibited by Mg,
since addition of the Mg did not impact the catalyst acidity.
Furthermore, the lower coke content of the Mg-doped Mo2C/HY
catalysts compared to the Mo2C/HY must be due to improved Mo2C
20% Mo C/HY
2% Mg-20% Mo C/HY
2
Zeolite HY
2
0.012
0.010
0.008
0.006
0.004
0.002
0.000
Table 2
Summary of napthalene conversion and product selectivities measured over Mg-Mo2C/
HY and K-Mo2C/HY catalystsa.
Expt. Mg (K) Conv. Selectivities (%)
Yield (%)
wt.%
(%)
Decalin Tetralin Hydro RO
Poly Hydro RO
E1
E2
E3
E4
E5
E6
0
91
90
94
93
93
2.6
2.9
3.4
3.0
3.0
6.9
61.7
76.4
75.1
76.9
75.2
84.0
64.3
79.3
78.5
79.9
78.2
90.9
22.4 13.3 58.5
20.4
0.5
1.0
1.5
2.0
13.8
12.3
11.0
14.6
5.3
6.9 71.4
9.2 73.8
9.2 74.3
7.3 72.7
3.8 85.4
12.4
11.6
10.2
13.6
5.0
1.0 (K) 94
0
50
100 0
50
100 0
50
100
a
Pore size, Å
Pore size, Å
Pore size, Å
Reaction conditions: time-on-stream: 1 h; catalyst weight: 500 mg; T: 300 °C; P:
3 MPa; WHSV (referred to the sum of H2 and naphthalene): 1 h−1; Feed: H2 and a
solution of 5 wt.% Naphthalene in heptane; H2/naphthalene=20 mol.
Fig. 2. Mesopore size distribution of HY, 20%Mo2C/HY and 2%Mg-20%Mo2C/HY catalysts.

X. Liu et al. / Catalysis Communications 12 (2011) 454–458
457
Table 3
TGA analysis of various catalysts.
Sample
Before reaction
Weight gain Weight loss Weight gainb Weight lossc
After reactiona
wt.%
5.8
wt.%
wt.%
wt.%
20%Mo2C/HY
0
0
0
0
0
0
2.8
4.4
3.8
3.7
3.7
4.3
7.0
4.4
2.9
4.1
4.3
4.0
0.5%Mg-20%Mo2C/HY 6.1
1.0%Mg-20%Mo2C/HY 5.1
1.5%Mg-20%Mo2C/HY 4.8
2.0%Mg-20%Mo2C/HY 4.0
1.0%K-20%Mo2C/HY
5.9
a
Reaction conditions: time-on-stream: 5 h; catalyst weight: 500 mg; T: 300 °C; P:
3 MPa; WHSV (referred to the sum of H2 and naphthalene): 1 h−1; Feed: H2 and a
solution of 5 wt.% naphthalene in heptane; H2/naphthalene=20 (mole).
b
In some cases, the weight gain corresponded to two peaks in the DTGA spectra.
Weight loss refers to the weight loss after consideration of the overlap with the
c
weight gain (between 350 and 550 °C) in TGA of un-used catalysts.
Fig. 4. DTGA of un-used Mg-Mo2C/HY and K-Mo2C/HY catalysts.
catalyst acidity compared to the Mo2C/HY. Accordingly, the mononu-
clear pathway would be enhanced, resulting in increased selectivity to
tetralin and reduced selectivity to polymer products, as observed
(Table 2). There was also a reduction in coke formation and RO
selectivity, even though the acid strength of the Mg-doped Mo2C/HY
was the same as that of the Mo2C/HY catalyst (Table 1). The increased
hydrogenation of naphthalene and tetralin limited the formation of
coke precursors that may undergo bimolecular reactions on acid sites
that form coke. In the case of K addition to the Mo2C/HY that resulted
in reduced acidity, ring-opening selectivity was reduced compared to
the Mo2C/HY catalyst, even though the Mo2C dispersion also
increased relative to the un-doped catalyst. In the present work,
addition of Mg and K resulted in lower coke formation (i.e. lower
weight loss after reaction – see Table 3), in the former case because of
increased hydrogenation that was a consequence of increased Mo2C
dispersion, in agreement with previous work [9], whereas in the latter
case, because of both increased Mo2C dispersion and reduced acidity.
In the case of both Mg and K, however, the reduced coke formation
was accompanied by a significant reduction in RO selectivity,
especially in the case of the K-doped Mo2C/HY catalyst that had
relatively low acidity. Both Mg and K increased the hydrogenation
capability of the Mo2C/HY catalyst but did not provide the optimum
acidity for increased RO selectivity. Together these results point to the
difficulty in obtaining high RO selectivity while limiting coke
formation.
dispersion (since the acidities were the same), and hence the improved
hydrogenationof theMg-doped catalysts reduced cokeformation. In the
case of the K-Mo2C/HY catalyst, however, both decreased acidity and
increased Mo2C dispersion reduced coke formation compared to the
Mo2C/HY catalyst.
In previous work, the reaction of naphthalene over Mo2C/HY was
discussed in terms of two competing reaction mechanisms [9–11].
Accordingly, naphthalene is hydrogenated to tetralin and small
amounts of decalin on Mo2C sites by a mononuclear mechanism.
The hydrogenated intermediates migrate to acid sites where further
hydrogenation, by spillover hydrogen migrating from Mo2C sites, and
isomerization and rupture of saturated naphthenic rings occurs,
leading to the desired ring-opened products. However, when an
effective supply of hydrogen to the zeolite acid sites is limited,
reactions can follow a bimolecular mechanism. In this scheme, a
carbenium ion of alkylbenzene is formed on a Brönsted acid site from
one tetralin molecule, which is a consequence of hydride transfer
reactions of naphthalene over acidic sites. This carbenium ion
molecule then attacks one naphthalene molecule leading to alkyl-
naphthalene and multinuclear aromatics that are further transformed
into coke by chain reactions.
Results from the present study show that addition of Mg to the
Mo2C/HY increased the Mo2C dispersion without decreasing the
4. Conclusions
The addition of Mg to Mo2C/HY catalysts resulted in improved
dispersion of the Mo2C and no significant change in total catalyst
acidity. Addition of K also improved Mo2C dispersion but decreased
the total acidity by 30%. Increasing the dispersion of the Mo2C
improved the hydrogenation of napthalene and reduced coke
deposition on the catalyst by limiting the formation of coke precursors
that undergo bimolecular coke-formation reactions on acid sites.
However, in both cases, ring-opening selectivity was also reduced,
especially in the case of the K-Mo2C/HY catalyst. Although Mg and K
provide increased naphthalene and coke-precursor hydrogenation,
they do not provide the optimum acidity needed for increased RO
selectivity.
Acknowledgement
Funding for the present study from the Natural Resources Canada
through the Climate Change Technology and Innovation Initiative is
gratefully acknowledged.
Fig. 5. DTGA of Mg-Mo2C/HY and K-Mo2C/HY catalysts after reaction in naphthalene.
Reaction conditions give in Table 2.

458
X. Liu et al. / Catalysis Communications 12 (2011) 454–458
[6] M. Jacquin, D.J. Jones, J. Rozière, A.J. López, E. Rodríguez-Castellón, J.M.T. Menayo,
References
M. Lenarda, L. Storaro, A. Vaccari, S. Albertazzi, J. Catal 228 (2004) 447.
[7] M. Santikunaporn, J.E. Herrera, S. Jongpatiwut, D.E. Resasco, W.E. Alvarez, E.L.
Sughrue, J. Catal. 228 (2004) 100.
[8] S.J. Ardakani, X. Liu, K.J. Smith, Appl. Catal. A:Gen 324 (2007) 9.
[9] X. Liu, K.J. Smith, Appl. Catal. A:Gen 335 (2008) 230–240.
[10] M.A. Arribas, A. Martinez, Appl. Catal. A 230 (2002) 203.
[11] K. Sato, Y. Iwata, Y. Miki, H. Shimada, J. Catal. 186 (1999) 45.
[1] H. Du, C. Fairbridge, H. Yang, Z. Ring, Appl. Catal. A 294 (2005) 1.
[2] G.B. McVicker, M. Daage, M.S. Touvelle, C.W. Hudson, D.P. Klein, W.C. Baird Jr., B.R. Cook,
J.G. Chen, S. Hantzer, D.E.W. Vaughan, E.S. Ellis, O.C. Feeley, J. Catal. 210 (2002) 137.
[3] F. Roessner, U. Roland, J. Mol. Catal. A 112 (1996) 401.
[4] K. Sato, Y. Iwata, T. Yoneda, A. Nishijima, Y. Miki, H. Shimada, Catal. Today 45
(1998) 367.
[5] D. Eliche-Quesada, J. Mérida-Robles, P. Maireles-Torres, E. Rodríguez-Castellón, A.
Jiménez-López, Appl. Catal. A 262 (2004) 111.