Highly efficient deep desulfurization of fuels by chemical oxidation

Article abstract of DOI:10.1016/j.cattod.2010.01.047

This paper describes the oxidation of several model S-containing molecules with hydrogen peroxide in L-L phase system using a heterogeneous catalyst under atmospheric pressure in the 333-353 K temperature range. Molybdenum and tungsten compounds are prepared by anion exchange with alkylammonium derivatives covalently anchored to silica gel. These solids are robust heterogeneous catalysts able to activate selectively hydrogen peroxide to remove sulfur compounds via oxidation (ODS). The influence of several reaction variables (the reaction temperature, the nature of the substrate, the solvent, the molar ratio of the oxidant (H₂O₂), the S-containing molecule, the catalysts nature and the reuse of the catalysts) on the performance was examined. The potential of this methodology is illustrated by the complete S removal from a 0.2 wt% dibenzothiophene mixture at 353 K in less than 1 h of reaction. Molybdenum catalysts exposed to hydrogen peroxide form peroxomolybdates moieties which are more active than acid precursors. The activated Mo catalysts are very active in ODS reaction and can be reused four times without lose of activity.

Full text of DOI:10.1016/j.cattod.2010.01.047

Catalysis Today 157 (2010) 390–396
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Highly efficient deep desulfurization of fuels by chemical oxidation
*
M.C. Capel-Sanchez, P. Perez-Presas, J.M. Campos-Martin , J.L.G. Fierro
´
´
Grupo de Energı´a y Quı´mica Sostenibles (EQS) Instituto de Catalisis y Petroleoquımica, CSIC. Marie Curie, 2 Cantoblanco, 28049 Madrid, Spain
A R T I C L E I N F O
A B S T R A C T
Article history:
This paper describes the oxidation of several model S-containing molecules with hydrogen peroxide in
L–L phase system using a heterogeneous catalyst under atmospheric pressure in the 333–353 K
temperature range. Molybdenum and tungsten compounds are prepared by anion exchange with
alkylammonium derivatives covalently anchored to silica gel. These solids are robust heterogeneous
catalysts able to activate selectively hydrogen peroxide to remove sulfur compounds via oxidation
(ODS). The influence of several reaction variables (the reaction temperature, the nature of the substrate,
the solvent, the molar ratio of the oxidant (H2O2), the S-containing molecule, the catalysts nature and the
reuse of the catalysts) on the performance was examined. The potential of this methodology is illustrated
by the complete S removal from a 0.2 wt% dibenzothiophene mixture at 353 K in less than 1 h of reaction.
Molybdenum catalysts exposed to hydrogen peroxide form peroxomolybdates moieties which are more
active than acid precursors. The activated Mo catalysts are very active in ODS reaction and can be reused
four times without lose of activity.
Available online 4 March 2010
Keywords:
Liquid-phase reactions
ODS
Oxidation
Hydrogen peroxide
ß 2010 Elsevier B.V. All rights reserved.
1. Introduction
can be oxidized by the electrophilic addition reaction of oxygen
atoms to the hexavalent sulfur of sulphones. The chemical and
The oil industry is facing increasing pressure to remove organic
sulfur compounds from transportation fuels. Each refinery must
meet unique circumstances, such as diesel blend components and
hydrogen availability.
physical properties of sulphones are significantly different from
those of hydrocarbons in fuel oil. Therefore, they can easily be
removed by separation operations such as distillation, solvent
extraction, adsorption, and decomposition.
Current hydrodesulfurization technology (HDS) allows desul-
furize aliphatic and acyclic sulfur-containing compounds quite
efficiently when adopted at industrial scale. This process, however,
is limited when treating dibenzothiophene (DBT), especially in the
case of DBTs having alkyl substituents at position 4 and/or 6 [1].
Thus, the production of ultra-clean light oil inevitably requires
severe operation conditions and especially highly active catalysts.
In the development of any alternative energy-efficient desulfuri-
zation process a radical approach, not limited to conventional HDS
technology, is required. To remove undesirable sulfur-containing
compounds or to convert them into more innocuous forms,
processes other than HDS have to be employed.
One of the most interesting options is oxidative desulfurization
(ODS) [2]. The greatest advantage of ODS as compared with the
conventional HDS processes is that it can be carried out in liquid
phase under very mild conditions at near room temperature and
under atmospheric pressure. In ODS reactions, the divalent sulfur
Among these oxidants, H2O2 is mostly chosen as an oxidant,
only producing water as a by-product [3–15]. Peracids produced
in situ from organic acids catalysts and H2O2 are reported to be
very effective for rapid oxidation of sulfur compounds in fuel oils
under mild conditions. Heteropolyacid catalysts in H2O2 oxida-
tion system have also exhibited high catalytic activity for the
oxidation of BTs and DBTs [7,8]. We have shown that a phase
transfer catalytic system can be very effective in the ODS
reaction [4]. However, the main obstacles for these catalysts to
the industrial application of the process at present are the
difficulties in separation and recovery. Therefore, the use of solid
catalysts in ODS processes has been developed in recent years.
Many types of solid catalysts have been attempted, such as Ti
molecular sieves [9–12], WOx/ZrO2 [13,14], vanadium oxide
[15,16] and molybdenum supported catalysts [17,18]. However,
the selectivity of these catalysts for sulfides in fuels is not high,
and some components of the fuel are also oxidized. Therefore,
the solid catalysts with high selectivity is highly desirable in
heterogeneous ODS system. Some attempts have been done for
the heterogeneization of tungsten acid compounds by immobi-
lization on a silica support with the use of silica-bound alkyl
ammonium salts [19,20].
* Corresponding author. Fax: +34 915854760.
E-mail address: j.m.campos@icp.csic.es (J.M. Campos-Martin).
URL: http://www.icp.csic.es/eac/
0920-5861/$ – see front matter ß 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.cattod.2010.01.047

M.C. Capel-Sanchez et al. / Catalysis Today 157 (2010) 390–396
391
Table 1
surface contents were estimated from the areas of the peaks,
corrected using the corresponding sensitivity factors.
Samples prepared and metal content determined by chemical analysis.
The liquid-phase catalytic sulfoxidation of sulfur-containing
organic compounds with hydrogen peroxide was carried out
batchwise in a mechanically stirred 250-mL thermostated glass
reactor equipped with thermometer, reflux condenser, and a
septum for withdrawing samples. In a typical experiment, 90 g of
the sulfur compound solution in n-hexadecane (Aldrich) were
heated to the reaction temperature. Then, a solution of hydrogen
Sample
Metal precursor
Metal content (wt%)
Mo-TMA
MoP-TMA
W-TMA
H2MoO4
11.3
12.2
12.4
12.4
H3PMo12O40
H2WO4
WP-TMA
H3PW12O40
Within this context, the aim of the present study was twofold.
We have studied the oxidation with H2O2 of several model
molecules in a liquid–liquid (L–L) phase system in the presence of
heterogeneous catalysts. The results reported here provide basic
information about the applicability of the oxidation method with
hydrogen peroxide in the selective transformation of organic sulfur
compounds contained in kerosene and light oils.
´
peroxide (70% (w/w), kindly supplied by Solvay Quımica, S.L.)
mixed with the solvent was added to the apolar phase, and finally
0.25 g of catalyst was added to the reactor. Aliquots were taken
from the reactor at different reaction times; the total amount
withdrawn from the reactor was less than 10% in order to avoid
interferences in the reaction results due to changes in the total
mass inside the reactor. The apolar phase was recovered by
decantation and analyzed by GC-FID equipped with a capillary
2. Experimental
column (HP-WAX, 25 m, Ø = 0.53 mm, 1.0
mm film thickness). The
hydrogen peroxide concentration was determined by standard
iodometric titration.
2.1. Catalyst preparation
Catalysts were prepared by ionic exchange of quaternary
ammonium functionalized silica supports (Silicycle: SiliaBond1
TMA Chloride). An excess (3:1) of metallic salts (Na2MoO4,
Na2WO4, phosphomolybdic acid and phosphotungstic acid) were
solubilized in water (25 mL), and then 4 g of functionalized silica
was added. The suspension was maintained under vigorous
stirring at room temperature for 24 h. The solid was filtered off
and washed ten times with 25 mL of water and dried at room
temperature. The codes of the samples are compiled in Table 1.
3. Results and discussion
The amount of metal incorporated into the samples (Table 1) is
very similar in all cases. This finding can be ascribed to the fact that
the exchange procedure, which depends on the number of
ammonium groups on the functionalized silica surface, is similar
in all samples.
The DRS-UV–vis spectra of samples show absorption associated
with the ligand metal charge transfer (LMCT) from the oxygen to
the transition metal ion (Fig. 1). Mo-TMA sample spectrum shows
an absorption band at 256 nm, which is characteristic of molybdate
species (Mo(IV)) [21]. The electronic spectrum of MoP-TMA sample
is somewhat more complex showing two maxima at 214 and
300 nm which is typical of ammonium phosphomolybdate salt
[22]. Tungsten-containing samples are similar but the peaks shift
to lower wavelengths, W-TMA spectrum shows a peak around
200 nm, usually present in tungstic acid [23], but WP-TMA
spectrum exhibits two peaks at 200 and 252 nm, which are
assigned to phosphotungstic acid [24].
2.2. Catalyst characterization
The metal loading of the catalysts was determined using
inductively coupled plasma absorption spectrometry, with a
PerkinElmer Optima 3300 DV instrument. Ultraviolet–visible
spectra were measured on a Varian Cary 5000 UV–vis spectropho-
tometer equipped with an integrating sphere. A BaSO4 disc was
used as reference. All spectra were acquired under ambient
conditions.
X-ray photoelectron spectra were recorded with a VG Escalab
200R spectrometer, equipped with a hemispherical electron
The binding energies of the most intense Mo3d5/2 and W4f7/2
peaks indicate that the ions remain in the highest oxidation state
(Mo6+ BE = 232.0 eV and W6+ BE = 36.0 eV, respectively). No change
analyzer and an Mg K
a (hn = 1253.6 eV) X-ray source. The binding
energies (BE) were referenced to the C 1s peak at 284.9 eV. Atomic
Fig. 1. UV–vis spectra of prepared samples.

392
M.C. Capel-Sanchez et al. / Catalysis Today 157 (2010) 390–396
has been observed in the value of the binding energy irrespectively
of the transition element and presence of phosphorus.
Summing up, the data provided by both UV–vis and photoelec-
tron spectroscopic techniques confirm unambiguously that the
oxidation state and anion structure of tungsten and molybdenum
compounds remain unchanged after the ion exchange.
3.1. Influence of the effect of the catalysts nature
All the samples synthesized were then tested in the removal of
S-containing compounds in
a fuel by means of oxidative
desulfurization (ODS) reaction. For the sake of ease understanding,
a model synthetic fuel consisting of a solution of DBT in n-
hexadecane was used. The oxidant employed consisted in a
hydrogen peroxide solution in acetonitrile. The concentration of
DBT was set at 0.2 wt% (240 ppm S). Although all catalysts were
active in the ODS reaction (Fig. 2) some differences were observed
among the samples used. Molybdenum catalysts are clearly more
active than tungsten counterparts (Fig. 2). At short reaction time
molybdenum catalysts reach high DBT conversion levels, while the
reaction is slower for tungsten samples. In addition, the P-
containing metal catalysts are more active than their P-free
counterparts. This effect is evident for both molybdenum- and
tungsten-containing catalysts (Fig. 2). Based on these preliminary
results, as the MoP-TMA catalyst prepared with phosphomolybdic
acid displayed better performance in ODS reaction, it was selected
for the next steps of the study, similar results have been found with
homogeneous systems [25].
Fig. 3. Influence of the reaction temperature in the conversion of dibenzothiophene.
Conditions: substrate:H2O2 = 1:10 (molar ratio), solvent: acetonitrile, 0.2 wt% DBT
in hexadecane, 0.25 g catalyst MoP-TMA.
3.2. Influence of the reaction temperature
The effect of temperature on the kinetics of the elimination of
DBT was investigated (333–353 K). The conversion profiles of
dibenzothiophene in the ODS reaction versus the reaction time at
three different temperatures (333, 343 and 353 K) are shown in
Fig. 3. A rise in the reaction temperature from 333 to 343 K led to a
remarkable increase in the reaction rate at every time of reaction.
However, this increase in the temperature from 343 to 353 K was
less marked. Moreover, the conversion of DBT was complete in
only 30 min of reaction time at 353 K, however at 333 K only 55% of
the DBT conversion was reached. In view of these results, the
reaction temperature employed for next studies was set at 353 K.
Though this temperature is the highest investigated, a very high
excess of hydrogen peroxide was used in this study, which is quite
far from an industrial application. Therefore it is inferred that
reaction rate will decrease when hydrogen peroxide concentration
is decreased [4].
3.3. Influence of the H2O2:substrate ratio
The above experiments were carried out with an excess of
hydrogen peroxide (H2O2/DBT = 10 molar ratio). However, accord-
ing to the stoichiometry only 2 mol of H2O2 are consumed per mole
of sulphone (R-SO2) produced (seeScheme 1). Thus, itwas ofinterest
to explore the influence of the hydrogen peroxide concentration on
the kinetics of DBT conversion. Several H2O2-to-substrate molar
ratios were selected, including 10/1, 5/1 and 2.5/1. The lowest H2O2
concentration was set slightly above (H2O2/DBT = 2.5/1 mol) the
stoichiometric (H2O2/DBT = 2/1 mol) ratio in order to avoid limita-
tions of the oxidant reagent and to achieve complete conversion of
DBT. The hydrogen peroxide-to-DBT ratio had a strong influence on
the reaction rate (Fig. 4). The reaction rate decreases with the H2O2/
DBT ratio. The reaction rate is the highest for H2O2/DBT = 10/1 molar
ratio (Fig. 4) for which complete DBT conversion was attained at
30 min. However, although the reaction rate is lower with H2O2/
DBT = 2.5/1 molar ratio, 100% of DBT conversion can be reached at
theendofthe reaction. Here, weshowthatcompleteDBTconversion
is obtained with oxidant-to-sulfur ratio close to the stoichiometry
(2.5/1), and this ratio was used in the following studies, because this
hydrogen peroxide:sulfur ratio is more interesting from the
industrial point of view.
Fig. 2. Influence of the catalyst nature in the conversion of dibenzothiophene.
Conditions: substrate:H2O2 = 1:10 (molar ratio), T: 353 K, solvent: acetonitrile,
0.2 wt% DBT in hexadecane.

M.C. Capel-Sanchez et al. / Catalysis Today 157 (2010) 390–396
393
Scheme 1. Simplified reactions of sulfur compounds following the ODS process.
3.4. Influence of the nature of the substrate
of these three model S-containing molecules allow not only to rank
them in terms of reactivity but also to evaluate the potential of the
ODS process in fuel desulfurization. Comparison of the reactivity of
different sulfur-containing compounds revealed that reactivity
depends on the type of S-containing molecule to be oxidized. The
relative ranking of the reaction rate was DBT > DMDBT ꢀ BT
(Fig. 5). In principle, this order of reactivity is unexpected because
the BT molecule, with a relatively accessible S-atom, would give
rise to a higher oxidation rate than in the more sterically hindered
DBT is representative of the S-compounds present in middle
distillate fuels. However, other S-containing molecules are present
in liquid fuels, and their reactivity against ODS reaction should be
evaluated. For this reason, we have compared the performance of
our catalytic system in the ODS of other sulfur-containing
compounds: benzothiophene (BT), dibenzothiophene (DBT) and
4,6-dimethyl-dibenzothiophene (DMDBT). The relative ODS rates
Fig. 5. Conversion–time curves for different S-compounds. Conditions: solvent:
acetonitrile, 0.2 wt% of S-compound in hexadecane, 0.25 g catalyst MoP-TMA, T:
353 K.
Fig. 4. Influence of the H2O2:substrate ratio in the conversion of dibenzothiophene.
Conditions: solvent: acetonitrile, 0.2 wt% DBT in hexadecane, 0.25 g catalyst MoP-
TMA, T: 353 K.

394
M.C. Capel-Sanchez et al. / Catalysis Today 157 (2010) 390–396
Fig. 6. Influence of solvent in the conversion of dibenzothiophene. Conditions: 0.2 wt% DBT in hexadecane (240 ppm S), substrate:H2O2 = 2.5:1 (molar ratio), T: 353 K.
S-containing DBT and DMDBT molecule. The explanation for the
above trend can be visualized in terms of S–C bond stabilization in
the BT molecule. In the case of DBT, and to a lesser extent that of
DMDBT, the inducing effect of the aromatic rings elicits an increase
in the electron charge density of the S-atom, meaning that the
molecule can be oxidized more easily by hydrogen peroxide.
operation cycles (Fig. 7). The used catalysts after 5 operation cycles
were analyzed by XPS. It was shown that the binding energy of the
most intense Mo3d5/2 peak (BE = 233.0 eV) shifts by about 1 eV
toward higher binding energy with respect to the fresh sample
(BE = 232.0 eV) (Fig. 8). This change is attributed to the formation
of peroxomolybdates moieties (Scheme 2). This molybdenum
species are formed during the activation of polymolybdopho-
sphates with hydrogen peroxide in organic solution [28–31], and
they are very active in oxidation reactions [32]. Based on this
3.5. Solvent effects
In all of previously depicted experiments, acetonitrile (ACN)
was used as solvent. ACN is appropriate solvent because it is able to
extract and solve the reaction products and exhibits a low surface
tension, which facilitates the transfer of products and reagents at
the polar–apolar interface, increasing notably the mass transfer
along the interphase. However, ACN is partially dissolved in the
apolar (fuel) phase [26], then ACN molecules are present in the fuel
phase, in consequence the final nitrogen concentration in the fuel
is very high. To avoid this, a later separation unit is mandatory to
remove the fraction of ACN transferred to the fuel phase.
Accordingly, other alternative solvents have also been explored
in this work. Several solvents have been proposed in the
bibliography but the best results in ODS has been reported using
g
-butyrolactone (GBL) [27].
The nature of the solvent also plays a very important role in
liquid-phase catalytic reactions. The solvent has an important
effect on the outcome of the reaction, yields, by-product formation,
and reaction kinetics, although this effect is strongly dependent on
the type of catalyst used and on the nature of the substrate. The
DBT conversion profile versus time (Fig. 6) shows a lower activity
in the experiment with GBL solvent, when standard reaction
conditions are used (stirring: 400 rpm). GBL is quite more viscous
than ACN and in consequence there are some mass transfer
limitations at low stirring rate. But, the same experiment at high
stirring speed (low mass transfer limitations) GBL yields to slightly
higher conversion level than ACN (Fig. 6).
3.6. Catalyst reuse
Finally, we carried out catalyst reuse experiments. The filtered
catalyst was dried after washing with acetonitrile and reused
without further treatment. Surprisingly, the catalytic activity
increases with operation cycles, reaching a maximum after 5
Fig. 7. Influence of the reuse of catalyst in the conversion of dibenzothiophene.
Conditions: 0.2 wt% DBT in hexadecane (240 ppm S), substrate:H2O2: = 2.5:1
(molar ratio), T: 353 K.

M.C. Capel-Sanchez et al. / Catalysis Today 157 (2010) 390–396
395
Fig. 8. Mo3d core-level spectra of fresh, reused and activated MoP-TMA sample.
finding, we decide to activate a fresh MoP-TMA catalyst with a
H2O2 solution in acetonitrile before their use in reaction. Sample
was activated adding 1 g catalyst to 60 ml (5 wt% H2O2 in ACN)
under stirring for 30 min. This activated catalyst became highly
active, and their catalytic activity is very close to that reused along
5 cycles. This activated catalyst was reused 4 times in reaction
without activity drop (Fig. 9). Taking in account these changes in
the catalytic activity of MoP-TMA after activation with hydrogen
peroxide, Mo-TMA catalysts was activated also and tested in
reaction, this activated catalyst was more active than the fresh one,
but it yields a lower activity than activated MoP-TMA, similar
observation obtained with fresh samples.
Scheme 2. Schematic formation of peroxomolybdates.
4. Conclusions
This work demonstrates that the desulfurization of DBT, and
also of BT and DMDBT, can be effectively accomplished by
oxidation with hydrogen peroxide under very mild reaction
conditions; namely atmospheric pressure and temperatures close
to ambient (333–353 K). S removal by the ODS reaction from BT,
DBT and DMDBT is easy and can be almost completed under the
very mild conditions imposed by this reaction. This very high
reactivity makes the ODS reaction promising for the deep
desulfurization of middle distillates.
The ODS process can be considered as no competitor of the
traditional HDS one. It appears to be complementary of it since
hydrotreatments not only remove S and other heteroatoms from
refinery streams but also improve the quality of the fuels. The ODS
is specifically designed to be integrated in the last processing steps
to decrease remaining sulfur until the levels fixed by environmen-
tal legislations. Thus, it seems appropriate to compare the ODS
with the revamping or modification of an HDS unit able to perform
a deep desulfurization and reach the same low S-levels which can
be achieved with the coupled standard HDS and ODS technologies.
Molybdenum catalysts are modified by contact with hydrogen
peroxide yielding peroxomolybdates moieties; these species are
more active than acid precursors. Fresh molybdenum catalysts can
be activated with hydrogen peroxide, and these activated systems
are very active in ODS reaction and can be reused four times
without lose in activity.
Fig. 9. DBT conversion in several reaction cycles of the activated catalyst at 15 min.
Conditions: 0.2 wt% DBT in hexadecane (240 ppm S), substrate:H2O2: = 2.5:1
(molar ratio), T: 353 K.

396
M.C. Capel-Sanchez et al. / Catalysis Today 157 (2010) 390–396
[18] V.V.D.N. Prasad, K.-E. Jeong, H.-J. Chae, C.-U. Kim, S.-Y. Jeong, Catal. Commun. 9
References
(2008) 1966.
[19] F. Bigi, A. Corradini, C. Quarantelli, G. Sartori, J. Catal. 250 (2007) 222.
[20] B. Sreedhar, P. Radhika, B. Neelima, N. Hebalkar, A.K. Mishra, Catal. Commun. 10
(2008) 39.
[21] A. Ko¨ckritza, M. Kanta, M. Walterb, A. Martin, Appl. Catal. A: Gen. 334 (1–2) (2008)
112.
[22] G. Ranga Rao, T. Rajkumara, B. Varghese, Solid State Sci. 11 (1) (2009) 36.
[23] L. Barrio, J.M. Campos-Martin, J.L.G. Fierro, J. Phys. Chem. A 111 (2007) 2166.
[24] A. Kukovecz, Z. Ko´nya, I. Kiricsi, J. Mol. Struct. 563/564 (2001) 409.
[25] A. De Angelis, P. Pollesel, D. Molinari, W.O. Parker Jr., A. Frattini, F. Cavani, S.
Martins, C. Perego, Pure Appl. Chem. 79 (2007) 1887.
[26] V.P. Sazonov, D.G. Shaw, J. Phys. Chem. Ref. Data 31 (4) (2002) 989.
[27] L.F. Ramirez-Verduzco, E. Torres-Garcia, R. Gomez-Quintana, V. Gonzalez-Pena, F.
Murrieta-Guevara, Catal. Today 98 (2004) 289.
[1] C. Song, X. Ma, Appl. Catal. B: Environ. 41 (2003) 207.
[2] A. Anisimov, A. Tarakanova, Russ. J. Gen. Chem. 79 (2009) 1264.
[3] S. Otsuki, T. Nonaka, N. Takashima, W.H. Qian, A. Ishihara, T. Imai, T. Kabe, Energy
Fuels 14 (2000) 1232.
[4] J.M. Campos-Martin, M.C. Capel-Sanchez, J.L.G. Fierro, Green Chem. 6 (2004) 557.
[5] Y. Shiriraishi, T. Naito, T. Hirai, Ind. Eng. Chem. Res. 42 (24) (2003) 6034.
[6] L.Y. Kong, G. Li, X.S. Wang, B. Wu, Energy Fuels 20 (2006) 896.
[7] F.M. Collins, A.R. Lucy, C. Sharp, J. Mol. Catal. A: Chem. 117 (1997) 397.
[8] H. Mei, B.W. Mei, T.F. Yen, Fuel 82 (2003) 405.
[9] V. Hulea, F. Fajula, J. Bousquets, J. Catal. 198 (2001) 179.
[10] Y. Shiraishi, T. Hirai, I. Komasawa, J. Chem. Eng. Jpn. 35 (2002) 1305.
[11] L.Y. Kong, G. Li, X.S. Wang, Catal. Today 93–95 (2004) 341.
[12] S.F. Cheng, Y.M. Liu, J.B. Gao, L.L. Wang, X.L. Liu, G.H. Gao, P. Wu, M.Y. He, Chin. J.
Catal. 27 (2006) 547.
[28] L. Salles, R. Thouvenot, J.-M. Bre´geault, Dalton Trans. (2004) 904.
[29] L. Salles, C. Aubry, R. Thouvenot, F. Robert, C. Dore´mieux-Morin, G. Chottard, H.
Ledon, Y. Jeannin, J.-M. Bre´geault, Inorg. Chem. 33 (1994) 871.
[13] E. Torres-Garcia, G. Canizal, S. Velumani, L.F. Ramirez-Verduzco, F. Murrieta-
Guevara, J.A. Ascencio, Appl. Phys. A 79 (2004) 2037.
[14] X.-M. Yan, P. Mei, J. Lei, Y. Mi, L. Xiong, L. Guo, J. Mol. Catal. A: Chem. 304 (2009) 52.
[15] Y. Shiraishi, T. Naito, T. Hirai, Ind. Eng. Chem. Res. 42 (2003) 6034.
´
[30] L. Salles, C. Aubry, F. Robert, G. Chottard, R. Thouvenot, H. Ledon, J.-M. Bregeault,
New J. Chem. 17 (1993) 367.
´
[31] C. Aubry, G. Chottard, N. Platzer, J.-M. Bregeault, R. Thouvenot, F. Chauveau, C.
´
˜
´
[16] H. Gomez-Bernal, L. Cedeno-Caero, A. Gutierrez-Alejandre, Catal. Today 142
Huet, H. Ledon, Inorg. Chem. 30 (1991) 4409.
[32] G. Wahl, D. Kleinhenz, A. Schorm, J. Sundermeyer, R. Stowasser, C. Rummey, G.
Bringmann, C. Fickert, W. Kiefer, Chem. Eur. J. 5 (11) (1999) 3237.
(2009) 227.
[17] J.L. Garcia-Gutierrez, G.A. Fuentes, M.E. Hernandez-Teran, F. Murrieta, J. Navar-
rete, F. Jimenez-Cruz, Appl. Catal. A-Gen. 305 (2006) 15.