An efficient heterogeneous catalyst based on highly dispersed Na 7H2LaW10O36·32H2O nanoparticles on mesoporous silica for deep desulfurization

Article abstract of DOI:10.1016/j.apcata.2013.06.030

A lanthanide-containing polyoxometalate of Na₇H ₂LaW₁₀O₃₆·32H₂O (LaW ₁₀) have been chemically anchored to the amino-modified mesoporous silica through an organic linker. The catalysts have been characterized by ICP-ES, FT-IR, N₂ sorption measurements, UV-vis, MAS-NMR, and HR-TEM. It is found that the LaW₁₀ nanoparticles are highly dispersed on mesoporous silica, and the resulting catalysts show highly efficient deep desulfurization in only 35 min with almost 100% conversion of dibenzothiophene by using (SiO₂/NH₃⁺/LaW₁₀)/[bmim] BF₄ (bmim = 1-butyl-3-methylimidazolium) in the presence of H ₂O₂ as oxidant. Furthermore, deep desulfurization proceeds smoothly in model oil with S content as low as 100 ppm. A scaled-up experiment, in which the volume of model oil increases from 5 to 1000 mL with S content of 1000 ppm, indicates that almost 100% sulfur removal can be achieved in 45 min. Additionally, the catalyst could be recycled and reused at least ten times without any decrease of the catalytic efficiency. To the best of our knowledge, the (SiO₂/NH₃⁺/LaW₁₀)/[bmim]BF ₄ catalyst system with H₂O₂ as oxidant exhibits great potential for practical application as desulfurization systems reported so far.

Full text of DOI:10.1016/j.apcata.2013.06.030

Accepted Manuscript
Title: An Efficient Heterogeneous Catalyst Based on Highly
Dispersed Na7H2LaW10O36·32H2O Nanoparticles on
Mesoporous Silica for Deep Desulfurization
Author: Yang Chen Shen Zhao Yu-Fei Song<ce:footnote
id="fn1"><ce:note-para id="npar0010">they have the same
contribution.</ce:note-para></ce:footnote>
PII:
S0926-860X(13)00360-8
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.apcata.2013.06.030
APCATA 14290
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
3-4-2013
18-6-2013
22-6-2013
∗
Please cite this article as: Y. Chen, S. Zhao, Y.-F.S. <ce:inter-ref id="intr0005"
xlink:href="mailto:songyufei@hotmail.com">songyufei@hotmail.com</ce:inter-ref>
<i><ce:inter-refid="intr0010"xlink:href="mailto:songyf@mail.buct.edu.cn">songyf@mail.buct.edu.cn</ce
ref> An Efficient Heterogeneous Catalyst Based on Highly Dispersed
Na7H2LaW10O36·32H2O
Nanoparticles
on
Mesoporous
Silica
for
Deep Desulfurization, <i>Applied Catalysis A, General</i> (2013),
http://dx.doi.org/10.1016/j.apcata.2013.06.030
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

An Efficient Heterogeneous Catalyst Based on Highly Dispersed
Na7H2LaW10O36·32H2O Nanoparticles on Mesoporous Silica for
Deep Desulfurization
Yang Chen,‡ Shen Zhao,‡ Yu-Fei Song*
‡they have the same contribution.
*Corresponding author
State Key Laboratory of Chemical Resource Engineering,
Beijing University of Chemical Technology,
Beijing, 100029, P. R. China.
Tel/Fax: +86-10-64431832;
Email: songyufei@hotmail.com or songyf@mail.buct.edu.cn.
1
Page 1 of 19

Highlights
The LaW10 is highly-dispersed on modified mesoporous silica.
+
Deep desulfurization of DBT, BT, and 4,6-DMDBTcan be achieved by SiO2/NH3 /LaW10.
The desulfurization catalyst is highly efficient with the model oil of 1000 mL.
+
The SiO2/NH3 /LaW10 can be easily recovered and reused at least ten times.
Abstract
A lanthanide-containing polyoxometalate of Na7H2LaW10O36·32H2O (LaW10) have
been chemically anchored to the amino-modified mesoporous silica through an
organic linker. The catalysts have been characterized by ICP-ES, FT-IR, N2 sorption
measurements, UV–vis, MAS-NMR, and HR-TEM. It is found that the LaW10
nanoparticles are highly dispersed on mesoporous silica, and the resulting catalysts
show highly efficient deep desulfurization in only 35 min with almost 100%
+
conversion of dibenzothiophene by using (SiO2/NH3 /LaW10)/[bmim]BF4
(bmim=1-butyl-3-methylimidazolium) in the presence of H2O2 as oxidant.
Furthermore, deep desulfurization proceeds smoothly in model oil with S content as
low as 100 ppm. A scaled-up experiment, in which the volume of model oil increases
from 5 to 1000 mL with S content of 1000 ppm, indicates that almost 100% sulfur
removal can be achieved in 45 mins. Additionally, the catalyst could be recycled and
reused at least ten times without any decrease of the catalytic efficiency. To the best of
+
our knowledge, the (SiO2/NH3 /LaW10)/[bmim]BF4 catalyst system with H2O2 as
oxidant exhibits great potential for practical application as desulfurization systems
reported so far.
Keywords: Polyoxometalates, Desulfurization, Lanthanides, Mesoporous silica
2
Page 2 of 19

1. Introduction
Sulfur-containing compounds in transportation fuels such as gasoline and diesel oil
can be converted to SOx and NOx during combustion, which not only are responsible
for the formation of smog, sour gases, acid rain and haze, but also poison noble metal
catalysts for reducing SOx and NOx emissions. The recent presence of haze
continuously in China has drawn attention to the quality of fuels. It is known that
thiophene sulfides account for more than 80% of the total sulfur content in diesel fuel,
and benzothiophene (BT) and dibenzothiophene (DBT) account for 70% of thiophene
sulfides. As a result, the removal of thiophene sulfides is a top priority in industry.[1,
2] To date, hydrodesulfurization (HDS) has been applied in industry, and this
method is efficient in removing thiols, sulfides, and disulfides from fuels.[3] However,
conventional HDS is less effective for removing the refractory sulfur-containing
compounds such as DBT and its derivatives.[4] In addition, HDS requires harsh
conditions such as high temperature, high pressure, the catalysts with high activity,
and high operating cost. Under such circumstances, oxidative desulfurization (ODS)
has been explored intensively as a promising alternative to HDS in recent years since
the operating conditions are mild and no hydrogen is necessary.[5-7] To the best of
our knowledge, the development of non-HDS technologies for the production of clean
diesel fuel with little-to-no sulfur content under mild conditions remains to be highly
demanding and challenging.
Ionic liquids (ILs) exhibit many desirable properties such as nonvolatility, sparing
solubility in nonpolar organic solvents, good solvent power, and high thermal stability,
and thus show great potential in various research areas. Recently, the use of
extractive catalytic oxidative desulfurization (ECODS) in the presence of
polyoxometalates (POMs) as catalysts has attracted wide interest and attention.[8-12]
Our previous studies have demonstrated that highly efficient deep desulfurization can
be realized by using lanthanide-containing POMs of Na7H2LaW10O36·32H2O as
catalysts under mild conditions.[6, 7]
Polyoxometalates (POMs) constitute a group of discrete anionic metal-oxides of V,
3
Page 3 of 19

Mo, W and Nb etc. and they are thermally and oxidatively stable compared with
commonly used organic ligands. The chemical properties of POMs such as redox
potential, electron-transfer properties, acidities, and solubility etc. can be finely tuned
by adjusting the metal ions, the heteroatoms and the counter ions. These properties
endow POMs to be superior catalytic materials for oxidative catalysis.[13-20] It is
known that most oxidative desulfurization (ODS) processes reported so far are
homogeneous catalysis, which causes the catalyst separation an added step. As a
result, many catalysts are not applicable in the practical industrial application. To
overcome the above-mentioned limitation, LaW10 has been immobilized onto
mesoporous silica to prepare a heterogeneous catalyst that incorporates the properties
of both the active LaW10 and a high surface area of silica support. The
desulfurization experiments have indicated that an efficient and recyclable
+
heterogeneous catalyst of SiO2/NH3 /LaW10 presents great potential for industrial
application.
2. Experimental Section
2.1. Chemical materials
Na2WO4·2H2O (99%), acetic acid (99%), LaCl3·7H2O (99%), dibenzothiophene
(DBT, 98%), 4,6-dimethyldibenzothiophene (4,6-DMDBT, 95%), benzothiophene
(BT, 98+%), hydrogen peroxide (30 wt% H2O2), n-octane (99%)
(3-aminopropyl)triethoxysilane (APTES) and trifluoromethanesulfonic acid were
obtained from Alfa and were used without further treatment. Mesoporous silica (Silica
gel, 227196 SIGMA-ALDRICH, particle size: 40-63μm, pore size: 60 Å, amorphous
state)
and
the
ionic
liquids
[bmim]BF4,
[bmim]PF6,
(bmim=1-butyl-3-methylimidazolium) [omim]BF4, and [omim]PF6 (omim=1-octyl-3-
methyl-imidazolium) were purchased from Sigma-Aldrich and used directly without
further purification. All solvents were of analytical grade, purchased from Alfa Aesar,
and used without further purification.
4
Page 4 of 19

2.2. Measurements
FT-IR spectra were recorded on a Bruker Vector 22 infrared spectrometer by using
29
KBr pellets. The solid state NMR experiments were carried out at 59.6 MHz for Si
on a Bruker Avance 300M solid-state spectrometer equipped with a commercial 5 mm
MAS NMR probe. The N2 adsorption-desorption isotherms were measured using
Quantachrome Autosorb-1 system at liquid nitrogen temperature. HRTEM images
were conducted on a JEOL JEM-2010 electron microscope operating at 200 kV. The
specimen was dispersed in ethanol using an ultrasonic washer, and dropped on a holey
ultrathin carbon film. Diffuse reflectance UV–vis (DRUV–vis) spectra of catalyst
samples were obtained using a Shimadzu UV-2101 PC spectrometer equipped with a
diffuse–reflectance attachment, with BaSO4 as the reference. Inductively coupled
plasma emission spectroscopy (ICP-ES, Shimadzu ICPS-7500) was used to measure
the concentration of La or Eu in the catalysts. The content of DBT, BT, and
4,6-DMDBT was analyzed on an Agilent 7820A GC system by using a 30 m 5%
phenymethyl silicone capillary column with an ID of 0.32 mm and 0.25 mm coating
(HP-5).
2.3. Preparation of lanthanide-containing POMs
Na7H2LaW10O36·32H2O was synthesized and characterized according to the
literature method. The ICP and FT-IR data were summarized in Table S1.[21]
2.4. Catalyst preparation
The surface modification of mesoporous silica by (3-aminopropyl) triethoxysilane
(APTES) was carried out as follows: In a typical preparation, 0.5 g of mesoporous
silica (particle size: 40-63μm, pore size: 60 Å) was pretreated at 393 K for 3 h under
reduced pressure (<10-2 torr). The activated mesoporous silica was refluxed in 50 cm3
of freshly-distilled toluene for 4 h. Then, 0.25 g of APTES was added and the reaction
mixture was stirred under reflux conditions for 24 h. After the solvent was evaporated
under vacuum, the solid was filtered and washed with distilled toluene, and then dried
for 8 h at 80 oC under vacuum.
5
Page 5 of 19

The amine-functionalized mesoporous silica was treated with trifluoromethane-
sulfonic acid to enhance POMs’ loading. Protonation was carried out by adding 6
mole equivalent of trifluoromethanesulfonic acid per mole of LaW10 to
amine-functionalized mesoporous silica. The more positively charged NH+ group
3
on the support, the stronger the electrostatic binding of LaW10 to the support.
Freshly protonated amine-functionalized mesoporous silica (0.5g) was added to 50
mL of methanol solution containing 0.3 g of LaW10 and refluxed for 24 h. The sample
o
was then filtered, washed with deionized water and dried at 100 C under vacuum.
+
The final materials are designated as SiO2/NH3 /LaW10.
2.5. Desulfurization procedure and analysis of S content
In a typical experiment, a solution of DBT, BT, and 4,6-DMDBT in n-octane was
used as model oil with an S content of 1000 ppm. The catalytic oxidative
desulfurization experiments were performed in a 50 mL two-necked flask, to which
0.08 mL of 30 wt% H2O2, 5 mL of model oil, 1 mL of [bmim]BF4, and 60 mg of
+
freshly prepared SiO2/NH3 /LaW10 were added. The resulting H2O2/DBT/LaW10
o
molar ratio was 150:30:1. The reaction mixture was stirred at 70 C. During the
reaction, the upper layer of the model-oil phase was periodically withdrawn and
analyzed by gas chromatography with a flame ionization detector (GC-FID). DBT, BT,
and 4,6-DMDBT were identified by using reference standards. The conditions were as
o
o
follows: injection port temperature 340 C; detector temperature 250 C; oven
temperature 70 oC; carrier gas: ultrapure nitrogen; sample injection volume 1 mL.
3. Results and discussion
3.1. Synthetic approach and characterization of the highly dispersed LaW10
nanoparticles on modified mesoporous silica
Scheme 1 shows the synthetic approach for preparation of supporting LaW10 on
silica supports. Firstly, 3-aminopropyl groups (APTES groups) were grafted on
mesoporous silica to obtain amine-functionalized material. Secondly, after the
6
Page 6 of 19

protonation of the amine-functionalized mesoporous silica, the freshly activated
material was added into the methanol solution of LaW10 and refluxed in order to load
LaW10 onto the mesoporous silica.[22] The electrostatic interactions allow LaW10 to
be chemically anchored strongly to the surfaces of the functionalized supports,
+
resulting in the formation of POM-immobilized mesoporous silica SiO2/NH3 /LaW10.
+
It should be noted that the lanthanum content in SiO2/NH3 /LaW10 is 87 μmol/g
analyzed by inductively coupled plasma emission spectroscopy (ICP-ES).
Scheme 1. Schematic representation of the LaW10 immobilization and dispersion result on
mesoporous silica.
Figure 1. The 29Si CP/MAS NMR profile of (a) SiO2 and (b) SiO2/NH2.
7
Page 7 of 19

29
As shown in Figure 1, the Si CP/MAS NMR spectrum of SiO2/NH2 exhibit two
visible resonances at -59 and -67 ppm, which correspond to RSi(OH)(OSi)2 (T2) and
RSi(OSi)3 (T3), respectively (R = CH2CH2CH2NH2).[23] Additionally, the resonances
at -101 and -110 ppm that are associated with (SiO)3SiOH (Q3) and Si(OSi)4 (Q4) can
be observed, indicating that (3-aminopropyl) triethoxysilane (APTES) groups have
been bound to the surface with two or three Si-O-Si linkages. Since no peak appears
at −45 ppm, which is normally the chemical shift of silicon in liquid (3-aminopropyl)
trialkoxysilane, this result indicates that no free silane molecules have been physically
adsorbed on the amine-functionalized mesoporous silica surface.[24] Contrast
29
experiment shows that no T2 and T3 resonance can be observed in the Si CP/MAS
NMR spectrum of SiO2.
+
+
Figure 2. FT-IR spectra of (a) SiO2, (b) SiO2/NH2, (c) SiO2/NH3 (d) SiO2/NH3 /LaW10 and (e)
LaW10.
+
+
FT-IR spectra of SiO2/NH3 /LaW10, SiO2/NH3 , SiO2/NH2, SiO2 and LaW10 have
been shown in Figure 2. The insertion of amino propyl groups on the mesoporous
silica is confirmed by the presence of N–H bending mode vibrations at 1552 cm−1 in
+
the FT-IR spectrum of SiO2/NH2.[25] In the FT-IR spectrum of SiO2/NH3 /LaW10,
the original N-H bending at 1552 cm−1 is shifted to 1519 cm−1 (Figure 2d), indicating
8
Page 8 of 19

+
the transformation of NH2 to NH3 . FT-IR spectrum of LaW10 shows four
characteristic peaks of W-O stretching bands at 944, 844, 785, and 710 cm−1, which
have shifted to 946, 883, 792, and 716 cm−1, respectively, in the corresponding FT-IR
+
spectrum of SiO2/NH3 /LaW10.
Table 1. Physico-chemical properties of the materials.
Materials
Surface area
(m2/g)
490
Pore volume
(cm3/g)
0.82
Average pore
diameter (nm)
SiO2
6.6
5.6
4.9
SiO2/NH2
322
0.73
+
SiO2/NH3 /LaW10
135
0.34
+
Figure 3. Adsorption–desorption isotherms of (a) SiO2, (b) SiO2/NH2 and (c) SiO2/NH3 /LaW10.
+
Figure 4. Pore size distribution of (a) SiO2, (b) SiO2/NH2 and (c) SiO2/NH3 /LaW10.
9
Page 9 of 19

The BET measurements were carried out on the organo-functionalized mesoporous
silica with and without LaW10. The specific surface area, pore volume, and pore
diameters (estimated from N2 adsorption–desorption isotherms) are presented in Table
1. BET surface areas and BJH pore distributions were calculated using N2 adsorption
at 77 K. It can be seen clearly that the aminosilylation and introduction of LaW10
significantly affected the surface area and pore distribution of the modified samples.
The samples display a type IV isotherm with H1 hysteresis (Figure 3). The textural
properties of mesoporous silica are substantially maintained over amine
functionalization and on subsequent anchoring of LaW10.[26] The parent mesoporous
silica sample exhibits a maximum pore diameter (6.6 nm) and surface area (490 m2/g)
(Table 1). The aminosilylation of the mesoporous silica results in a shift of the pore
maximum to smaller diameters (Figure 4) and a decrease in surface area (322 m2/g).
Moreover, the introduction of LaW10 leads to a further decrease in surface area and
pore volume.
+
Figure 5. HRTEM images of (a) SiO2, (b) SiO2/NH2, (c) SiO2/NH3 /LaW10, (d) (e) LaW10 highly
dispersed on mesoporous silica and (f) magnification of selected region in (e) of mesoporous
silica.
10
Page 10 of 19

The HRTEM images of the parent mesoporous silica (Fig. 5a) and grafted samples
(Fig. 5b and 5c) provide strong evidence that the mesoporous structure of supports has
been retained. As shown in Figure 5d and 5e, LaW10 nanoparticles are highly
dispersed on mesoporous silica and these nanoparticles are isolated and almost evenly
distributed throughout the entire mesoporous silica without severe agglomeration. In
addition, according to the HRTEM image, the LaW10 particle size is in the range of
1–2 nm and the lattice fringe of LaW10 nanoparticles can also be clearly observed,
which indicates the LaW10 molecules are mono-dispersed on the mesoporous silica.
+
Figure 6. The UV/Vis spectra of (a) LaW10 and (b) SiO2/NH3 /LaW10.
As shown in Fig 6, the diffuse-reflectance UV/Vis spectroscopy of LaW10 exhibits
+
characteristic absorption bands centered at 265 nm, whereas that of SiO2/NH3 /LaW10
exhibits O→W ligand–metal charge-transfer (LMCT) transition at around 262 nm,
which is very close to that of neat LaW10. The result indicates the presence of LaW10
in the corresponding mesoporous silica. The band shift suggests the interactions exist
between LaW10 and mesoporous silica. It should be noted that LaW10 is under the
constraint environment of mesoporous silica after immobilization.
3.2. Influence of different extractants on sulfur removal
Previous studies have shown that the selection of ionic liquids could have a
significant influence on the catalytic desulfurization.[6,7] In this study, different ionic
liquids have been added to the catalytic system, respectively, to investigate the effect
11
Page 11 of 19

+
on the catalytic desulfurization of DBT by SiO2/NH3 /LaW10 (Table 2). In a typical
+
catalytic oxidation reaction, DBT and freshly prepared SiO2/NH3 /LaW10 have been
added to the model oil in the presence of H2O2 and different ionic liquids. Catalytic
oxidative desulfurization of DBT with [omim]BF4 or [omim]PF6 as extractant shows
sulfur removal of 51% or 95% in 80 min (Table 2). In contrast, when [bmim]BF4 or
[bmim]PF6 is applied as extractant, >99% sulfur removal could be reached for
[bmim]PF6 in 40 min and [bmim]BF4 in 35 min. As such, [bmim]BF4 is used for the
following studies.
Table 2. Effect of different ionic liquids on sulfur removal of DBT.
Entry
Catalyst
IL
t [min]
80
S removal [%]
+
1
2
3
4
SiO2/NH3 /LaW10
[omim]BF4
[omim]PF6
[bmim]BF4
[bmim]PF6
51
95
99
99
+
SiO2/NH3 /LaW10
80
+
SiO2/NH3 /LaW10
35
+
SiO2/NH3 /LaW10
40
Reaction conditions: T=70oC H2O2 0.080 mL, model oil 5 mL (S content 1000 ppm), [omim]BF4 or [omim]PF6 or
[bmim]BF4 or [bmim]PF6 1 mL, H2O2/DBT/LaW10 150:30:1.
3.3. Effect of temperature on sulfur removal of DBT
To investigate the effect of temperature on sulfur removal, oxidation of DBT has
+
been performed at different temperatures (Table 3). The SiO2/NH3 /LaW10 shows
desulfurization of DBT for 36%, 69% and 92%, 99%, at 30, 40, 50 and 60oC in 80
min, respectively. In contrast, this system reaches more than 99% sulfur removal in 35
o
min at 70 C. Therefore, DBT can be completely removed in a relatively short
reaction time at 70 oC.
Table 3. Effect of different temperatures on sulfur removal of DBT
Entry
temperature
30oC
t [min]
80
S removal [%]
1
2
3
4
5
36
69
92
99
99
40oC
50oC
60oC
70oC
80
80
80
35
Reaction conditions: H2O2 0.080 mL, model oil 5 mL (S content 1000 ppm), [bmim]BF4, 1 mL H2O2/DBT/LaW10
150:30:1.
3.4. Effect of H2O2/DBT molar ratio on the desulfurization of model oil
12
Page 12 of 19

To elucidate the influence of H2O2 on the desulfurization, different H2O2/DBT
molar ratios have been applied under the optimized experimental conditions of T = 70
oC, 5 mL of model oil (S content 1000 ppm), 1 mL of [bmim]BF4, DBT:LaW10 = 30:1.
With the increase of the H2O2/DBT molar ratio from 1:1 to 8:1 (Table 4, entries 1–5),
the catalytic desulfurization of DBT increases gradually. When excess amount of
H2O2 is applied, DBT is completely oxidized to DBTO2, whereas only part of DBT
transforms into DBTO2 when one eqiv of H2O2 is applied. (Figure S1). The observed
competition between H2O2 decomposition and DBT oxidation is in line with the
results of Li and co-workers.[27]
Table 4. Effect of H2O2/DBT molar ratio on sulfur removal of DBT
Entry
H2O2/DBT
H2O2 [μL]
t [min]
80
S removal [%]
1
2
3
4
5
1
2
3
5
8
16
32
44
74
97
99
99
80
48
80
80
35
128
30
Reaction conditions: T = 70oC, model oil = 5 mL (S content 1000 ppm), [bmim]BF4 = 1 mL, DBT/LaW10 = 30:1.
3.5. Effect of DBT/LaW10 molar ratio on the desulfurization of model oil
As shown in Table 5, the time for deep desulfurization could be shortened from 45
to 35 min when the DBT/LaW10 molar ratio increases from 60:1 to 60:4 in the
presence of 0.080 mL of H2O2 (Entries 1–3). It can be found that when DBT/LaW10
molar ratio reaches to 60:8, sulfur removal reaches 86% in 35 min and 91% in 80 min,
which might be caused by the catalyst poison. Furthermore, slight increase of the
amount of H2O2 to 0.128 mL leads to almost 100% sulfur removal in only 20 min at T
= 70 oC with the H2O2/DBT/LaW10 molar ratio of 480:60:8 (Entry 9).
Table 5. Effect of DBT/LaW10 molar ratio on sulfur removal of DBT
Entry
H2O2 [μL]
t [min]
45
S removal [%]
H2O2/DBT/LaW10
300:60:1
1
2
3
4
5
6
7
80
80
99
99
99
86
91
99
99
300:60:2 = 150:30:1
300:60:4 = 75:15:1
300:60:8 = 75:15:2
300:60:8 = 75:15:2
480:60:1
35
80
35
80
35
80
80
128
128
35
480:60:2 = 240:30:1
30
13
Page 13 of 19

8
9
480:60:4 = 120:15:1
480:60:8 = 240:15:2
128
128
30
20
99
99
Reaction conditions: T = 70oC, model oil = 5 mL (S content 1000 ppm), [bmim]BF4 = 1 mL.
3.6. Kinetic study on catalytic oxidation of sulfur compounds
To obtain the kinetic parameters for oxidation of DBT, experiments have been
performed with the H2O2/DBT/LaW10 = 150:30:1 at 70 oC. Percentage sulfur removal
and ln(Ct/C0) are plotted against reaction time in Figure 9, in which C0 and Ct are
initial DBT concentration and DBT concentration at time t, respectively. The linear fit
of the data reveals that the catalytic reaction exhibits pseudo-first-order kinetics for
the desulfurization of sulfides (R2=0.9946). The rate constant k of the oxidation
reaction can be determined to be 0.10 min-1. From the data of k(70oC), k(60oC),
k(50oC) and k(40oC) (Figure 7, Figure S2), the calculated activation energy (Ea) for
DBT oxidation is 58 kJ/mol. The oxidation of DBT to DBTO2 could be complete in
about 35 min. Thus, the catalyst exhibits high catalytic efficiency for oxidation of
sulfide to sulfone, and the catalytic reaction strictly obeys pseudo-first-order kinetics
with 100% selectivity for DBTO2.
-dCt/dt=kCt
(1)
(2)
(3)
ln(C0/Ct)=kt
ln k  ln A Ea / RT
Figure 7. Sulfur removal of DBT and ln(Ct/C0) as functions of reaction time at 70oC. H2O2 =
0.080 mL, model oil = 5 mL (S content 1000 ppm), [bmim]BF4 = 1 mL, H2O2/DBT/LaW10
150:30:1.
=
14
Page 14 of 19

+
Further investigation of the activity and potential application of SiO2/NH3 /LaW10
has been carried out. First of all, it has been demonstrated that 100% desulfurization
could be achieved in 5 min in model oil with S content as low as 100 ppm, indicating
that deep desulfurization can be realized in oil with low S contents. Secondly,
scaled-up experiment by increasing the model oil from 5 mL to 1000 mL shows that
+
deep desulfurization can be reached in 45 mins, which suggest the SiO2/NH3 /LaW10
has a great potential for practical application.
Table 6. Effect of S content decrease and scaled-up experiments
Entry
volume of model oil
1000 mL
S content
1000 ppm
100 ppm
t [min]
S removal [%]
1
4
45
5
99
99
5 mL
Reaction conditions: T=70oC, H2O2 0.080 mL, model oil 5 mL (S content 1000 ppm), [bmim]BF4 1 mL,
H2O2/LaW10 150:1.
3.7. Sulfur removal of BT and 4,6-DMDBT
+
The performance of SiO2/NH3 /LaW10 with H2O2 as oxidant for different substrates
including BT, DBT, and 4,6-DMDBT was evaluated (Table 7). It is known that, due
to steric hindrance, it is very difficult to remove 4,6-DMDBT in the HDS process.
+
Under our experimental conditions, the catalyst SiO2/NH3 /LaW10 proves to be highly
efficient for removing sulfur of different substrates with 99% conversion of DBT in
only 35 min, 92% conversion of BT and 86% conversion of 4,6-DMDBT in 80 min,
respectively.
Table 7. Sulfur removal of DBT, BT, and 4,6-DMDBT.
Entry
Substrates
DBT
t [min]
35
S removal [%]
1
3
4
99
92
86
BT
80
4,6-DMDBT
80
Reaction conditions: T = 70oC, H2O2 = 0.080 mL, model oil = 5 mL (S content 1000 ppm), [bmim]BF4 = 1 mL,
H2O2/substrate/LaW10 = 150:30:1.
3.8. recycling catalyst
The main advantage of using solid catalysts in a liquid-phase reaction is the ease of
separation and reuse of the catalyst in catalytic cycles. The recycled
+
SiO2/NH3 /LaW10 catalytic system has been applied for further oxidation of DBT.
15
Page 15 of 19

After the desulfurization procedure, the upper layer was separated by decantation or
using separatory funnel. Then, 50 mL of deionized water was added to the water
phase to decrease the viscosity of IL. After that, the catalyst was separated by
centrifugation and filtration. The diluted ionic liquid in deionized water was
concentrated under vacuum to recycle the ionic liquid. As such, the IL and catalyst
are both reusable. The catalyst could be recycled and reused at least ten times without
any decrease of the catalytic efficiency (Figure S3).
3.9. Proposed desulfurization process
+
Figure 8. The proposed catalytic oxidative desulfurization process using the (SiO2/NH3 /LaW10)
catalyst system.
As shown in Figure 8, the immiscibility of [bmim]BF4 with model oil and catalyst
results in the formation of triphasic catalyst system, in which DBT is in the oil phase,
H2O2 is in the IL phase, and LaW10 immobilized onto the amine-modified
mesoporous silica is in the solid phase. It should be mentioned that the presence of
+
strong electronic interactions between LaW10 and the protonated -NH3 group makes
LaW10 anchored tightly onto SiO2. At the interface of water phase/solid phase,
LaW10 is continuously oxidized to active peroxo species in the presence of H2O2.
With the extraction of DBT from oil phase to IL phase, it can be oxidized by the
16
Page 16 of 19

W-peroxo species to DBTO2. ICP-ES measurement of the water phase shows no
presence of La and W, indicating no bleach of LaW10 from solid catalyst to IL phase.
The La (LaW10) content is 86.6 μmol/g in the recycled catalyst, which is almost the
same as new prepared catalyst. Additionally, the catalyst can be recycled and reused
at least ten times without any decrease of the catalytic efficiency.
Conclusions
Immobilization of Na7H2LaW10O36·32H2O onto amine-functionalized mesoporous
+
silica leads to the formation of new heterogeneous catalyst of SiO2/NH3 /LaW10.
Deep oxidative desulfurization of DBT, BT, and 4,6-DMDBT has been achieved by
+
applying the catalytic system of (SiO2/NH3 /LaW10)/[bmim]BF4 in the presence of
H2O2 as oxidant. Furthermore, such desulfurization proceeds smoothly and efficiently
in model oil with an S content as low as 100 ppm. The scaled-up experiment, in which
the volume of model oil increases from 5mL to 1000 mL, shows that deep
desulfurization can be realized in only 45 min. It has been demonstrated that 99%
desulfurization of DBT in only 35 min, 92% conversion of BT and 86% conversion of
o
4,6-DMDBT can be achieved at 70 C in 80 min, respectively. In addition, the
heterogeneous catalyst can be recycled at least ten times without any decrease of the
+
catalytic efficiency. Therefore, the SiO2/NH3 /LaW10 is one of the most efficient
heterogeneous desulfurization catalysts reported so far with great potential for further
application.
Acknowledgements
This research is supported by the National Science Foundation of China (21222104, 21076020),
the program for New Century Excellent Talents of the Ministry of Education of China
(NCET-09-0201), the Beijing Nova Program (2009B12), the Fundamental Research Funds for the
Central Universities (ZZ1227).
References
[1] P. S. Kulkarni, C. A. M. Afonso, Green Chem. 12 (2010) 1139-1149.
[2] Y. Chi, C. Li, Q. Jiao, Q. Liu, P. Yan, X. Liu, U. Welz-Biermann, Green Chem. 13 (2011)
1224-1229.
17
Page 17 of 19

[3] M. Francisco, A. Arce, A. Soto, Fluid Phase Equilib. 294 (2010) 39-48.
[4] A. Röthlisberger, R. Prins, J. Catal. 235 (2005) 229-240.
[5] A. R. Hansmeier, G. W. Meindersma, A. B. de Haan, Green Chem. 13 (2011) 1907-1913.
[6] J. Xu, S. Zhao, W. Chen, M. Wang, Y.-F. Song, Chem. Eur. J. 18 (2012) 4775-4781.
[7] J. Xu, S. Zhao, Y. Ji, Y.-F. Song, Chem. Eur. J. 19 (2013) 709-715.
[8] W. Zhu, H. Li, X. Jiang, Y. Yan, J. Lu, L. He, J. Xia, Green Chem. 10 (2008) 641-646.
[9] H. Li, W. Zhu, Y. Wang, J. Zhang, J. Lu, Y. Yan, Green Chem. 11 (2009) 810-815.
[10] J. Zhang, W. Zhu, H. Li, W. Jiang, Y. Jiang, W. Huang, Y. Yan, Green Chem. 11 (2009)
1801-1807.
[11] H. Lü, W. Ren, H. Wang, Y. Wang, W. Chen, Z. Suo, Appl. Cat. A: Gen. 453 (2013) 376-382.
[12] H. Lü, W. Ren, W. Liao, W. Chen, Y. Li, Z. Suo, Appl. Cat. B: Environ. 138–139 (2013) 79-83.
[13] Polyoxometalate Themed Issue, C. L. Hill, Chem. Rev. 98 (1998) 1-2.
[14] I. A. Weinstock, Chem. Rev. 98 (1998) 389-390.
[15] N. Mizuno, M. Misono, Chem. Rev. 98 (1998) 199-218.
[17] D.-L. Long, R. Tsunashima, L. Cronin, Angew. Chem. Int. Ed. 49 (2010) 1736-1758.
[18] Polyoxometalate Themed Issue, L. Cronin, A. Müller, Chem. Soc. Rev. 41 (2012) 7333-7334
[19] Y.-F. Song, R. Tsunashima, Chem. Soc. Rev. 41 (2012) 7384-7402.
[20] Polyoxometalate Themed Issue, Guest Editor: D.-L. Long, L. Cronin, Dalton. 41 (2012)
9815-9816.
[21] R. D . Peacock, T. J. R. Weakley, J. Chem. Soc. A (1971) 1836-1839.
[22] B. J. S. Johnson, A. Stein, Inorg. Chem. 40 (2001) 801-808.
[23] K. Yamaguchi, C. Yoshida, S. Uchida, N. Mizuno, J. Am. Chem. Soc. 127 (2004) 530-531.
[24] A. Bordoloi, F. Lefebvre, S. B. Halligudi, J. Catal. 247 (2007) 166-175.
[25] S. Shylesh, A. P. Singh, J. Catal. 244 (2006) 52-64.
[26] J. Horniakova, T. Raja, Y. Kubota, Y. Sugi, J. Mol. Catal. A: Chem. 217 (2004) 73-80.
[27] W. Huang, W. Zhu, H. Li, H. Shi, G. Zhu, H. Liu, G. Chen, Ind. Eng. Chem. Res. 49 (2010)
8998-9003.
18
Page 18 of 19

Graphic Abstract:
19
Page 19 of 19