Non-hydrolytic SiO2-TiO2 mesoporous xerogels - Efficient catalysts for the mild oxidation of sulfur organic compounds with hydrogen peroxide

Article abstract of DOI:10.1016/j.cattod.2010.02.046

SiO₂-TiO₂ mixed oxides with molar ratios Ti/(Si + Ti) ranging from 0.03 to 0.09 were prepared by a non-hydrolytic sol-gel process using SiCl₄ and TiCl₄ precursors. The texture and the distribution of Ti atoms in the resulting xerogels strongly depended on the Ti concentration as precursor and the reaction temperature. Mesoporous xerogels with very high surface area (up to 1200 m² g^-1) and pore diameters (up to 7.9 nm) can be obtained at 150 °C with a Ti loading of 6 mol%. These mesoporous materials showed excellent catalytic performances in the sulfoxidation reaction of different bulky sulfides (methyl phenyl sulfide, diphenyl sulfide), and thiophenes (benzothiophene, dibenzothiophene, tetrahydrothiophene, 4,6-dimethyl dibenzothiophene) with H₂O ₂. In all the cases a very high selectivity toward H ₂O₂ consumption as oxidizing agent was observed (>98%).

Full text of DOI:10.1016/j.cattod.2010.02.046

Catalysis Today 157 (2010) 270–274
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Non-hydrolytic SiO2–TiO2 mesoporous xerogels—Efficient catalysts for the mild
oxidation of sulfur organic compounds with hydrogen peroxide
A.M. Cojocariua,c, P.H. Mutinb, E. Dumitriuc, A. Aboulaichb, A. Viouxb, F. Fajulaa, V. Huleaa,∗
a Institut Charles Gerhardt, UMR 5253, CNRS-UM2-ENSCM-UM1, Matériaux Avancés pour la Catalyse et la Santé, 8 rue de l’Ecole Normale, 34296 Montpellier Cedex 5, France
b Institut Charles Gerhardt, UMR 5253, CNRS-UM2-ENSCM-UM1, Chimie Moléculaire et Organisation du Solide, Université Montpellier 2, Place Eugène Bataillon,
34095 Montpellier Cedex 5, France
c Technical University of Iasi, 71 D. Mangeron Ave., 700050 Iasi, Romania
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 17 March 2010
SiO2–TiO2 mixed oxides with molar ratios Ti/(Si + Ti) ranging from 0.03 to 0.09 were prepared by a non-
hydrolytic sol–gel process using SiCl4 and TiCl4 precursors. The texture and the distribution of Ti atoms
in the resulting xerogels strongly depended on the Ti concentration as precursor and the reaction tem-
perature. Mesoporous xerogels with very high surface area (up to 1200 m2 g−1) and pore diameters (up
to 7.9 nm) can be obtained at 150 ◦C with a Ti loading of 6 mol%. These mesoporous materials showed
excellent catalytic performances in the sulfoxidation reaction of different bulky sulfides (methyl phenyl
sulfide, diphenylsulfide), andthiophenes(benzothiophene, dibenzothiophene, tetrahydrothiophene, 4,6-
dimethyl dibenzothiophene) with H2O2. In all the cases a very high selectivity toward H2O2 consumption
as oxidizing agent was observed (>98%).
Keywords:
Mesoporous
Xerogel
Non-hydrolytic sol–gel
Thiophenes
Sulfides
Oxidation
H2O2
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
On the other hand, mesoporous SiO2–TiO2 xerogels can be
obtained using non-hydrolytic sol–gel routes [10,11]. Thus, the
Ti-containing microporous and mesoporous materials are
attractive heterogeneous catalysts for the oxidation of organic
compounds with H2O2 [1–3]. The best catalysts are the crys-
talline Ti-silicates with zeolite framework (TS-1, TS-2, Ti-beta, etc.),
but their applications are limited to reactants with molecule size
smaller than 0.7 nm. The synthesis of amorphous Ti-containing
mesoporous templated silica (Ti-MCM-41, Ti-MCM-48) having uni-
form mesopores ranging from 2 to 10 nm in diameter opened new
opportunities for the oxidation of bulky molecules [4–6]. Homoge-
neous, and amorphous SiO2–TiO2 mixed oxides have been prepared
by conventional sol–gel processes based on the hydrolysis and
condensation of silicon and titanium alkoxides. However, the xero-
gels obtained by vacuum-drying are usually microporous [7,8] and
elaborated drying procedures are required to obtain mesoporous
materials [9], such as extraction with supercritical CO2 or washing
with hexane and silylation before drying. Generally, the amor-
phous SiO2–TiO2 mixed oxides exhibit much lower performance
than the crystalline titanosilicates in the oxidation of organic sub-
strates with aqueous solutions of H2O2 and thus their use is limited
to oxidation reactions with dry organic hydroperoxides.
reaction of SiCl4 and TiCl4 with diisopropyl ether offers a very sim-
ple and efficient route to obtain homogeneous SiO2–TiO2 xerogels
starting from cheap precursors, without any organic templating
agent. We recently reported that these materials are effective cat-
alysts for the mild oxidation of sulfur-containing compounds with
H2O2 [12]. This behavior has been related to the high degree
of condensation typical for SiO2–TiO2 xerogels obtained by non-
hydrolytic methods, which results in low amount of surface silanols
and low hydrophilicity.
In this work we study the influence of Ti content and synthe-
sis temperature on the texture and structure of the non-hydrolytic
SiO2–TiO2 xerogels, as well as their catalytic performance in the
sulfoxidation of bulky sulfides and thiophenes with H2O2. As
previously reported [2,13,14], these reactions can be considered
as model processes for two important industrial applications:
sulfoxides/sulfones synthesis and fuels catalytic oxidative desul-
furization. It is important to note that there are only few papers
dealing with oxidation of bulky sulfide using Ti catalysts [15].
2. Experimental
2.1. Catalysts preparation
∗
The SiO2–TiO2 mixed oxides were prepared by reaction of SiCl4
and TiCl4 with iPr2O. In a glove box under argon atmosphere, SiCl4
Corresponding author. Tel.: +33 0467163464; fax: +33 0467163470.
E-mail address: vasile.hulea@enscm.fr (V. Hulea).
0920-5861/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.cattod.2010.02.046

A.M. Cojocariu et al. / Catalysis Today 157 (2010) 270–274
271
Table 1
Composition and textural data of the different catalysts.
c
Sample
mol% Tia nominal
mol% Tia experimentalb
SBET (m2 g−1
)
Vp (cm3 g−1
)
Dp (nm)
Ti3-110
Ti5-110
Ti6-110
Ti7-110
Ti9-110
Ti3-150
Ti5-150
Ti6-150
Ti9-150
3.0
5.0
5.9
7.3
8.9
3.0
5.0
5.9
9.0
1.8
4.2
6.4
6.5
8.0
3.1
4.6
6.3
9.1
600
880
950
900
965
800
970
1215
920
0.3
0.7
0.9
0.9
1.0
0.5
1.2
2.4
2.6
2.0
3.2
3.8
4.0
4.1
2.5
4.9
7.9
11.3
a
molar ratio 100 × Ti/(Si + Ti).
b
c
Estimated error 1%.
Average pore diameter (4000Vp/SBET).
and TiCl4 (total amount 40 mmol), iPr2O (8.17 g, 80 mmol) and
CH2Cl2 (10 mL) were added successively with syringes into a 150-
mL, Teflon-lined stainless steel autoclave. The autoclave was heated
under autogenous pressure for 6 days at 110 or 150 ◦C. After cool-
ing, the autoclave was opened in a glove box. The gel was filtered
off, washed three times with dry CH2Cl2, dried under vacuum first
at room temperature for 2 h, then at 110 ◦C for 6 h, and calcined in
dry air at 500 ◦C for 5 h. The samples were labeled Tix–y where x is
the nominal molar percentage of Ti (100 × Ti/(Si + Ti)) and y is the
reaction temperature.
odically and analyzed by GC (DB-1, 60 m, 0.20 mm id, 0.25 ␮m film
thickness). The hydrogen peroxide concentration was measured by
standard iodometric titration.
3. Results and discussion
3.1. Synthesis and characterization of SiO2–TiO2 mixed oxides
The non-hydrolytic sol–gel process used here provides a simple,
one-step route to amorphous SiO2–TiO2 materials, starting from
cheap chloride precursors. This process is based on the following
etherolysis and condensation reactions (M = Si or Ti):
2.2. Catalysts characterization
i
i
M–Cl + Pr2O → M–OiPr + PrCl
(1)
(2)
The Ti content of xerogels was determined by ICP-MS at the
“Service Central d’Analyse” USR-59/CNRS. N2 physisorption exper-
iments were done at −196 ◦C on a Micromeritics ASAP 2010
sorptometer with samples previously degassed under vacuum at
200 ◦C for 15 h. The pore size distribution was obtained from the
desorption branch using the BJH method and the Harkins and Jura
standard isotherm. Diffuse reflectance UV–vis (DRUV–vis) spectra
were recorded under ambient conditions on a PerkinElmer Lambda
14 spectrometer equipped with a BaSO4 coated integration sphere;
the samples were diluted in BaSO4; the spectra were plotted using
the Kubelka–Munk function.
M–Cl + M–OiPr → M–O–M + PrCl
i
In the case of SiO2–TiO2 gels, the condensation reactions around
Si atoms are catalyzed by Ti species. The molar ratio Ti/(Si + Ti)
investigated ranged between 0.03 and 0.09. Two reaction tempera-
tures were used, 110 and 150 ◦C. Elemental analysis indicated that,
within experimental error, the Ti content in the materials was close
to the nominal Ti content, indicating that all the Ti atoms were
incorporated into the silica matrix (Table 1). Both Ti content and
synthesis temperature strongly influenced the textural properties
of solids, as shown in Fig. 1. The pore volume and pore diameter
increased with the Ti content and synthesis temperature (Table 1).
Thus, the xerogel Ti3-110 contained a significant amount of micro-
pores, whereas the xerogels synthesized at higher temperature or
with a higher Ti loading were mesoporous (Fig. 1), with average
pore diameters ranging from 3.2 to 11.3 nm. The calcined xero-
gels Ti6-150 and Ti9-150 exhibited truly outstanding textures, with
pore volumes around 2.5 cm3 g−1 usually associated with aerogels.
The increase of the pore volume and pore diameter with Ti content
2.3. Catalytic studies
The catalytic oxidation reactions were performed at atmo-
spheric pressure in a thermostated glass batch reactor. In a typical
experiment, 50 mg of catalyst were suspended under stirring
(650 rpm) in a mixture containing 1.5 mmol of organic substrate,
3 mmol H2O2 (aqueous solution, 50 wt%) and 15 mL of acetonitrile
as solvent. Samples of the reaction mixture were withdrawn peri-
Fig. 1. N2 adsorption–desorption isotherms at −196 ◦C of the catalysts prepared at 110 (left) and 150 ◦C (right).

272
A.M. Cojocariu et al. / Catalysis Today 157 (2010) 270–274
Fig. 2. Normalized DRUV–vis spectra of the catalysts prepared at 110 (left) and 150 ◦C (right).
and reaction temperature can be ascribed to the increase of the
condensation rate leading to gels with higher degree of crosslink-
ing, able to withstand the high capillary pressures imposed on the
gel network during the drying process that are responsible for the
collapse of unsufficiently crosslinked gels [10].
the DBT oxidation was highly selective to the sulfur atom, leading
directly to the corresponding sulfone (Eq. (3)).
The distribution of the Ti atoms in the silica framework was
investigated using DRUV–vis spectroscopy. DRUV–vis is commonly
used to establish the environment of Ti-sites in Ti-containing
silicates. Thus, bands at 210–230 nm have been attributed to
oxygen-to-metal charge transfer (CT) at isolated tetrahedral Ti(IV)
centers [16,17]. The CT band at 260–280 nm has been ascribed
to partially polymerized penta- and hexacoordinated Ti atoms
[18,19]. Conversely, the presence of a segregated anatase phase
leads to a characteristic absorption at 320–340 nm [20,21].
The spectra of all the calcined xerogels showed two large
absorption bands in the range of 210–230 nm (ascribed to the iso-
lated four-coordinated Ti atoms) and 260–280 nm (attributed to Ti
oligomers (Fig. 2)). The relative intensity of the band at 260–280 nm
increased with the Ti loading, as the probability of Ti–O–Ti bond for-
mation increases. The formation of anatase domains occurred only
in the case of Ti9-150, as shown by the weak band at 350 nm. It is
known that anatase is responsible for the direct decomposition of
H2O2 into oxygen and water.
(3)
In the absence of catalyst, using the same experimental conditions,
DBT conversion was negligible. For the sake of comparison, the oxi-
dation of DBT was carried out with a TS-1 catalyst (400 m2 g−1
,
1.2 wt% Ti) prepared and characterized as described previously
[22]. Under the above-mentioned conditions only 5% of DBT was
converted after 360 min, indicating that this molecule hardly pen-
etrates within the TS-1 micropores with 0.53–0.56-nm aperture.
The kinetics profiles of DBT oxidation over the SiO2–TiO2 mixed
oxides prepared by non-hydrolytic sol–gel are plotted in Fig. 3. For
all the catalytic materials the oxidation reaction occurred with sig-
nificant conversions. These data indicate, on the one hand, the high
intrinsic activity of the Ti-sites, and on the other hand, the highly
beneficial effect of the catalyst mesoporosity for the accessibility
and mobility of DBT molecule inside the pores. The reaction rate
strongly depends on the Ti loading and the preparation tempera-
ture. Thus, for a given preparation temperature, the DBT conversion
increased with the Ti loading in the catalysts; for a given Ti loading,
the DBT conversion increased with the preparation temperature. A
notable exception is Ti9-150, which showed only moderate activ-
ity. This could be related to the presence of anatase domains in
this sample (Fig. 2), favoring the direct decomposition of the H2O2
into oxygen and water (H2O2 → H2O + (1/2)O2). Indeed, the mea-
sured H2O2 efficiency (defined as the ratio between the number of
H2O2 moles required for the DBT oxidation and the number of H2O2
moles converted) was only 80% for Ti9-150, whereas it was higher
3.2. Catalytic studies
The catalytic potential of SiO2–TiO2 mixed oxides was evaluated
in the sulfoxidation reaction of dibenzothiophene (DBT), a bulky
and weakly reactive molecule, using H2O2 as oxidizing agent. The
reactions were carried out at 60 ◦C in acetonitrile, a solvent fully
miscible with both organic substrate and the aqueous hydrogen
peroxide solution. Typically, the experiments were carried out with
a H2O2/DBT ratio of 2. Under the present experimental conditions
Fig. 3. Kinetics profiles of DBT oxidation on catalysts prepared at 110 (left) and 150 ◦C (right). Conditions: T = 60 ◦C, H2O2/DBT = 2/1 (mol/mol).

A.M. Cojocariu et al. / Catalysis Today 157 (2010) 270–274
273
Table 2
DBT oxidation over Ti6-150 under different conditions.
−1
Parameter
Value
Initial rate (mol L−1 h−1 gcatal
)
DBT conversion after 120 min (%)
50
60
70
2.6
3.0
3.5
96
97
100
Temperature (◦C)a
2/1
3/1
4/1
3.0
3.8
4.4
97
99.5
100
H2O2/DBT (mol/mol)b
Oxidizing agentc
H2O2 50% in H2O
TBHP 70% in H2O
TBHP 5.5 M in decane
3.0
3.2
3.1
97
94
96
a
Conditions: H2O2/DBT = 2/1 (mol/mol).
b
T = 60 ◦C.
c
T = 60 ◦C, oxidizing agent/DBT = 2/1 (mol/mol); solvent = acetonitrile.
Table 3
Catalytic behavior of Ti6-150 for the oxidation of various substrates (S) with H2O2.
Substrate (solvent)
H2O2/S (mol/mol)
Ci (mol L−1
)
Temperature (◦C)
Reaction time (min)
S conversion (%) into:
H2O2 efficiency (%)
Sulfoxide
Sulfone
5
20
30
55
2
0
45
98
100
THT (MeCN)
MPS (MeCN)
DPS (MeCN)
2/1
2/1
2/1
0.1
0.1
0.1
40
40
40
100
100
100
20
30
45
15
5
0
85
95
100
30
120
180
46
15
0
54
85
100
60
360
0
0
45
93
BT (MeCN)
2/1
0.1
60
80
98
98
60
360
0
0
30
97
4,6-DMDBT (MeOH)
10/1
0.01
than 98% for all other catalysts. The influence of the preparation
temperature on the DBT conversion can be related to the signifi-
cantly higher surface area and pore size of the catalysts prepared
at 150 ◦C (Table 1).
alytic sites. As expected, the sulfides are oxidized much faster than
the thiophene derivatives. It is known that the reactivity of sulfur
compounds can be correlated to the nucleophilicity of the sulfur
atom, which is higher in sulfides than in thiophenes [2,14]. Finally,
it is important to point out that 4,6-DMDBT, a less reactive molecule
which is present in the fuels processed by the hydrodesulfurization
plants, can be effectively oxidized with H2O2 in diluted methanol
solution. Methanol was used as solvent in this test because 4,6-
DMDBT is not soluble in acetonitrile.
In order to obtain information on the stability of active Ti species
in the SiO2–TiO2 matrix, Ti6-150 was reused several times in the
DBT oxidation reaction performed at 60 ◦C for 60 min in acetonitrile
with a ratio H2O2/DBT of 2. The catalyst was separated by filtration
afterreaction, washedwith acetonitrile, andthen placed into afresh
reagent mixture. The conversion of DBT for the four runs was 90%,
91%, 89%, and 87%, indicating that our catalyst was stable under
operating conditions.
In order to investigate the effect of the oxidation temperature
and H2O2/substrate ratio in the sulfoxidation, several tests were
performed with DBT as a model substrate and Ti6-150 as cata-
lyst. The initial rate and the conversion of the organic substrate
are summarized in Table 2. As expected, both the initial rate and
the conversion of DBT increase when the reaction temperature
increases. As shown in Table 2, an excess of H2O2 has a favor-
able effect on the catalytic performances. However, it is important
to note that the DBT conversion reached 100% after 180 min even
without any excess of H2O2, indicating a high catalytic activity for
6Ti-150 and a very selective consumption of hydrogen peroxide for
the sulfoxidation reaction.
The catalytic oxidation of DBT was also carried out in the pres-
ence of anhydrous TBHP (5.5 M in decane) and 70% TBHP in aqueous
solution as oxidants. The results included in Table 2 indicate that
the catalytic behaviour of H2O2 and TBHP were very close.
Some sulfides and thiophenes with different molecule sizes
and reactivities, such as tetrahydrothiophene (THT), methyl phenyl
sulfide (MPS), diphenyl sulfide (DPS), benzothiophene (BT) and 4,6-
dimethyl dibenzothiophene (4,6-DMDBT) were also oxidized with
H2O2 in the presence of Ti6-150. The oxidation of sulfides led to the
corresponding sulfoxides and sulfones, while the oxidation of the
thiophene derivatives led directly to the corresponding sulfone. The
catalytic results and the oxidation conditions are summarized in
Table 3. Very high conversion and H2O2 efficiency were obtained for
all the substrates. These data indicate that even the larger molecules
easily penetrate the mesopores of Ti6-150 to reach the active Ti cat-
On the other hand, to demonstrate that the activity of our cata-
lyst was not due to the leaching of Ti species, an additional catalytic
test was performed with DBT: the catalyst was filtered off at 60 ◦C
after 15 min reaction (35% conversion) then the solution was fur-
ther reacted for 60 min at 60 ◦C. Only a small increase in conversion
of DBT (from 35% to 37%) was observed.
4. Conclusion
The non-hydrolytic sol–gel process based on the etherolysis and
condensation of chloride precursors offers a simple one-step route
to prepare mesoporous SiO2–TiO2 xerogels with a good dispersion
of Ti species and outstanding textural properties.

274
A.M. Cojocariu et al. / Catalysis Today 157 (2010) 270–274
We have investigated the sulfoxidation reaction of sulfur
[8] Z. Liu, R.J. Davis, J. Phys. Chem. 98 (1994) 1253.
[9] R. Hutter, T. Mallat, A. Baiker, J. Catal. 153 (1995) 177.
[10] V. Lafond, P.H. Mutin, A. Vioux, Chem. Mater. 16 (2004) 5380.
[11] P.H. Mutin, A. Vioux, Chem. Mater. 21 (2009) 582.
[12] A.M. Cojocariu, P.H. Mutin, E. Dumitriu, F. Fajula, A. Vioux, V. Hulea, Chem.
Commun. (2008) 5357.
[13] V. Hulea, P. Moreau, F. Di Renzo, J. Mol. Catal. 111 (1996) 325.
[14] V. Hulea, A.L. Maciuca, F. Fajula, E. Dumitriu, Appl. Catal. A: Gen. 313 (2006)
200.
[15] C. Jin, G. Li, X. Wang, Y. Wang, L. Zhao, D. Sun, Micropor. Mesopor. Mater. 111
(2008) 236.
[16] M.R. Boccuti, K.M. Rao, A. Zecchina, G. Leofanti, G. Petrini, Stud. Surf. Sci. Catal.
48 (1989) 133.
compounds with aqueous solution of hydrogen peroxide over
SiO2–TiO2 xerogels. It has been shown that these large-pore cat-
alytic materials are active for the selective oxidation of bulky
sulfides and thiophenes to the corresponding sulfoxides/sulfones.
An excellent selectivity based on hydrogen peroxide consumption
was observed for these reactions, which makes these materials very
promising for fine chemicals synthesis and environmental applica-
tions.
[17] L. Marchese, T. Maschmeyer, E. Gianotti, S. Coluccia, J.M. Thomas, J. Phys. Chem.
B 101 (1997) 8836.
References
[18] G.N. Vayssilov, Catal. Rev. Sci. Eng. 39 (1997) 209.
[19] C. Galacho, M.M.L. Ribeiro Carrott, P.J.M. Carrott, Micropor. Mesopor. Mater.
100 (2007) 312.
[20] L. Marchese, E. Gianotti, V. Dellarocca, T. Maschmeyer, F. Rey, S. Coluccia, J.M.
Thomas, Phys. Chem. Chem. Phys. 1 (1999) 585.
[21] M.C. Capel-Sanchez, J.M. Campos-Martin, J.L.G. Fierro, J. Catal. 234 (2005) 488.
[22] A.M. Cojocariu, P.H. Mutin, E. Dumitriu, A. Vioux, F. Fajula, V. Hulea, Chemo-
sphere 77 (2009) 1065.
[1] R.A. Sheldon, I.W.C.E. Arends, H.E.B. Lempers, Catal. Today 41 (1998).
[2] V. Hulea, F. Fajula, J. Bousquet, J. Catal. 198 (2001) 179.
[3] M. Ziolek, Catal. Today 90 (2004) 145.
[4] A. Corma, M.T. Navarro, J. Perez-Pariente, J. Chem. Soc., Chem. Commun. (1994)
147.
[5] T. Blasco, A. Corma, M.T. Navarro, J. Perez-Pariente, J. Catal. 156 (1995) 65.
[6] W. Zhang, T.J. Pinnavaia, Catal. Lett. 38 (1996) 261.
[7] S. Klein, S. Thorimbert, W.F. Maier, J. Catal. 163 (1996) 476.