Boosting Oxidative Desulfurization of Model and Real Gasoline over Phosphotungstic Acid Encapsulated in Metal–Organic Frameworks: The Window Size Matters

Article abstract of DOI:10.1002/cctc.201601450

It is desirable to develop new materials that can efficiently lower sulfur content in fossil fuels, such as gasoline and diesel oil. Polyoxometallic acids supported on metal–organic frameworks (MOFs) are an important class of heterogeneous catalysts for oxidative desulfurization. However, there has been no comprehensive study on the correlation between desulfurization activity and the window size of MOFs. A series of robust MOFs, which include MIL-100(Fe), UiO-66, and ZIF-8, with different window sizes were exploited as hosts to encapsulate phosphotungstic acid by the “bottle around the ship” method and utilized for the ultra-deep oxidative desulfurization (ODS) of model and real gasoline. Compared with UiO-66 and ZIF-8, which have very small window sizes, mesoporous MIL-100(Fe), which has a large window size, exhibited the best catalytic performance in the ODS of refractory sulfur compounds (benzothiophene, dibenzothiophene, and 4,6-dimethyl-dibenzothiophene) and in recycling experiments. The correlations between the desulfurization activity and the window size of the corresponding MOFs could provide insights for the design of new porous catalysts for ODS and other size-selective catalysis reactions in the future.

Full text of DOI:10.1002/cctc.201601450

Accepted Article
Title: Boosting Oxidative Desulfurization of Model and Real Gasoline
over Phosphotungstic Acid Encapsulated in Metal-Organic
Frameworks: the Window Size Matters
Authors: Xu-Sheng Wang, Lan Li, Jun Liang, Yuan-Biao Huang, and
Rong Cao
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201601450
Link to VoR: http://dx.doi.org/10.1002/cctc.201601450
A Journal of
www.chemcatchem.org

ChemCatChem
10.1002/cctc.201601450
FULL PAPER
Boosting Oxidative Desulfurization of Model and Real Gasoline
over Phosphotungstic Acid Encapsulated in Metal-Organic
Frameworks: the Window Size Matters
[
a,b]
[a]
[a]
[a]
[a]
Xu-Sheng Wang, Lan Li, Jun Liang, Yuan-Biao Huang,* and Rong Cao*
Dedication ((optional))
Abstract: The large-scale uses of fossil fuels containing
organosulfur compounds lead to air pollution and other related
environmental problems. Therefore, it is highly desirable to develop
new materials that can efficiently reduce the sulfur contents in fossil
fuels, such as gasoline, diesel oil. Particularly, metal-organic
frameworks (MOFs) supported polyoxometallic acids have been
proved to be an important class of heterogeneous catalysts for
oxidative desulfurization. However, there was no comprehensive
study on the correlation between the desulfurization activities and
window sizes of MOFs. In this work, a series of robust MOFs
including MIL-100(Fe), UiO-66 and ZIF-8 with different window sizes
were exploited as hosts to encapsulate phosphotungstic acid by
dibenzothiophene (4,6-DMDBT) due to their aromaticity and
[
1, 4-8]
steric hindrance.
In contrast, the existing d orbits of sulfur
atoms enable refractory sulfur-containing compounds to be
oxidated to sulfone or sulfoxide easily. Moreover, the higher
electron density of sulfur atoms is more favorable for oxidative
[
6, 7, 9]
desulfurization (ODS).
Thus, to meet the requirement of
strict environmental regulations, the ODS technique is
a
promising way to achieve an ultra-low level of sulfur content in
fossil fuels. In the biphasic oxidative desulfurization system, DBT
and other refractory organosulfur compounds were transferred
to polar phase and then oxidated to their corresponding sulfones
(Scheme 1). So far, polyoxometallic acids (POMs) based
catalysts have been intensively studied in the ODS process
because of their high catalytic activity in conversion of refractory
“bottle around ship” method and utilized for ultra-deep oxidative
desulfurization (ODS) of model and real gasoline. Compared with
UiO-66 and ZIF-8 with very small window sizes, the mesoporous
MIL-100(Fe) with larger window size exhibited the best catalytic
[
5-7, 9-18]
sulfur to sulfone.
However, these catalysts are
deactivated easily in the recycled runs because
proportion of POMs were leaching to the solutions.
a
large
Thus,
[
10-14]
performance
in
ODS
of
refractory
sulfur
and
compounds
(benzothiophene,
dibenzothiophene
4,6-dimethyl-
the use of porous material to confine POMs into their pores to
prevent them leaching out is a promising strategy to circumvent
this drawback.
dibenzothiophene) and in recycle experiments. The correlations
between the desulfurization activity and window size of
corresponding MOFs could provide deeper insights for designing
new porous catalysts for ODS and other size-selective catalysis
reactions in the future.
In recent years, metal-organic frameworks (MOFs) have been
widely applied in numerous fields including molecule separation
and heterogeneous catalysis owing to its high surface
area/porosity, tunable pore apertures/window size, and many
[
19-28]
other outstanding properties.
The use of highly porous
MOFs as hosts to confine effective POMs (POM@MOF) is a
promising strategy for ODS with high activity and good
recyclability. In particular, the diverse structures and tunable
sizes of both MOFs and POMs could obtain plentiful new
composite materials (POM@MOFs) as highly reactive catalysts
for ODS. For example, Ribeiro et al. immobilized a commercial
Introduction
Nowadays, new techniques for ultra-deep desulfurization of
fossil fuels have become more urgent because the large-scale
use of fossil fuels leads to serious air pollution and the stricter
environmental regulations have been promulgated by many
3
POM compound phosphotungstic acid (PTA, H PW12O40) into
[1-4]
countries.
Although the traditional hydrodesulfurization (HDS)
the mesocages of MIL-101 for ODS with excellent catalytic
technique can removes thiols, sulfides, and disulfides in the
fossil fuels, it is difficult to remove the refractory sulfur-containing
aromatic hydrocarbon compounds, such as benzothiophene
[5]
performance. Nevertheless, the PTA molecules would leach
out during the ODS process due to its size (10.0 Å) is smaller
2
than the window size (16.0×16.0 Å ) of MIL-101, which led to
(
BT),
dibenzothiophene
(DBT),
and
4,6-dimethyl-
[12]
low activity of catalysts in the recycled runs. If the size of POM
is smaller than that of the nanocages but larger than the window
size of MOFs, POMs could be confined in the pores of MOFs to
prevent leaching. However, it should be noted that the catalytic
[
[
a]
X.-S. Wang, L. Li, J. Liang, Prof. Dr. Y.-B. Huang, Prof. Dr. R. Cao
State Key Laboratory of Structural Chemistry
Fujian Institute of Research on the Structure of Matter, Chinese
Academy of Sciences,
Fuzhou, Fujian, China 350002.
E-mail: ybhuang@fjirsm.ac.cn (Y.B. H); rcao@fjirsm.ac.cn (R. C)
X.-S. Wang
performance might decrease with narrowing the windows of
[29]
MOFs because of the slower mass transfer.
Therefore, the
judicious choice of MOFs with suitable window size which not
only can confine POMs into the nanocages but also maintain the
mass transfer speed is a key factor to enhance the catalytic
performance. However, to the best of our knowledge, there was
no report about the relationship between the oxidative
desulfurization activity and the window size of MOFs.
b]
University of the Chinese Academy of Sciences
Beijing, 100049, China
Supporting information for this article is given via a link at the end of
the document.((Please delete this text if not appropriate))
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601450
FULL PAPER
To fully understand the catalytic performance influenced by the
window size of porous catalysts, PTA encapsulated in the cages
of MOFs including MIL-100(Fe), UiO-66, and ZIF-8 with different
window sizes were successfully synthesized using the “bottle
around ship” strategy. The MOFs can assemble around the PTA
molecules, which could be encapsulated into their corresponding
cages. The resulting materials PTA@MOF (MOF = MIL-100(Fe),
UiO-66, ZIF-8) were fully characterized by powder X-ray
diffraction (PXRD), Fourier transform infrared spectroscopy (FT-
IR), inductively coupled plasma OES spectrometer (ICP-OES),
field emission scanning electron microscope (SEM),
transmission electron microscope (TEM), thermo gravimetric
measurements (TG), N
characterization methods of those catalysts are similar, MIL-
00(Fe) and PTA@MIL-100(Fe) as the representative materials
2
sorption experiment etc. Due to the
1
were selected to be discussed in detail, whereas the
characterization of other materials were presented in the
Supplementary Information.
The phase purity of MIL-100(Fe) was confirmed by PXRD
(
Figure 2a), and its pattern was in good agreement with the
simulated one. The PTA@MIL-100(Fe) composite was prepared
by a facile one-pot method via the reaction of iron powder,
Scheme 1. View of ODS using PTA@MOFs as heterogeneous catalysts, and
the catalytic reactions of organosulfur compounds in this study.
H
3
BTC, HF, HNO
3
and required amount of PTA in deionized
water. The obtained orange solids were denoted as
7%PTA@MIL-100(Fe),
16%PTA@MIL-100(Fe),
and
3
5%PTA@MIL-100(Fe) respectively, based on the PTA weight
With those considerations in mind, herein, three types of
robust MOFs MIL-100(Fe), UiO-66, and ZIF-8 with different
window sizes were selected as hosts to load PTA for the
content determined by ICP-OES. The PXRD patterns of the
PTA@MIL-100(Fe) material were similar to those of the parent
material MIL-100(Fe), which indicated that the crystal structure
of MIL-100(Fe) was retained after incorporation of PTA (Figure
[
30-35]
oxidative desulfurization of model and real gasoline.
MIL-
100(Fe) is an iron(III) carboxylate porous MOF with two types of
2
a). The characterization peaks of PTA were not observed in the
mesopores cages of free apertures of ca. 25.0 Å and 29.0 Å,
2
PXRD patterns of PTA@MIL-100(Fe) indicating that the PTA
were uniformly dispersed in the mesocages of MIL-100(Fe),
which can be employed as single active site for enhanced
catalysis. In addition, the TG analysis revealed that MIL-100(Fe)
and PTA@MIL-100(Fe) are thermal stable up to 400 C (Figure
S3), which are suitable for catalytic performance.
accessible through microporous windows of ca. 5.8×5.8 Å and
2
[30, 33, 34, 36]
8
.6×8.6 Å , respectively (Figure 1a).
UiO-66 was
constructed by Zr (OH) (CO 12 secondary building units and
6
O
4
4
2
)
dicarboxylate bridging ligands, which presents an interesting
three dimensional material with two types of microporous cages
of free aperture of ca. 10.0 Å and 14.0 Å, accessible through
o
[31, 35]
The successful incorporation of PTA was further confirmed by
FT-IR spectroscopy (Figure 2b). The characterization peak at
microporous windows of ca. 6.0 Å (Figure 1b).
ZIF-8 is
composed of 2-methylimidazolate and zinc(II) with a sod cage
ca. 12.0 Å, accessible through a narrow six-ring pore ca. 3.4 Å
-1
1080 cm can be assigned to the asymmetric vibrations of the
-1
[
32, 36]
P-O bond of PTA, and the peak at 980, 896, 825 cm was
attributed to the terminal W=O , edge share of W-O -W, and
corner share of W-O -W bonds, respectively. Compared with
the W-O -W characterization peak of pure PTA at 802 cm , a
red shift of the PTA@MIL-100(Fe) peak to 825 cm indicating
that strong interaction existed between PTA and MIL-
(
Figure 1c).
Interestingly, the size of PTA is ca. 10.0 Å,
which could be restricted to the nanocages of the above
mentioned MOFs including MIL-100(Fe), UiO-66 and ZIF-8,
avoiding the leaching out of the PTA from MOFs. More
significantly, the comprehensively study of the relationship
between window size and the activity of ODS would help pave
the way not only for the design of new porous catalysts towards
ultra-deep oxidative desulfurization and other catalytic reactions,
especially size-selective reactions, but also for the exploration of
their new applications such as gas storage/separation, drug
delivery and sensing.
a
b
[
7]
c
-1
c
-1
a
[
9]
1
1
00(Fe), which promoted the high loading of PTA in MIL-
00(Fe) as confirmed with the ICP-OES data.
2
The N sorption isotherms of MIL-100(Fe) and PTA@MIL-
1
3
00(Fe) with different PTA loading amount are shown in Figure
a. The parent material MIL-100(Fe) showed high BET (2055
2
3
m /g) and pore volume (0.993 cm /g). Interestingly, although the
very high amount of PTA (35 wt%) incorporation in MIL-100(Fe),
the PTA@MIL-100(Fe) exhibited high BET surface area of 1046
m /g, and large pore volume of 0.487cm /g (Table 1). The
existence of two types of mesopores (25.0 Å and 29.0 Å) and
the corresponding microporous windows (5.8×5.8 Å for
Results and Discussion
2
3
Catalysts preparation and characterization
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601450
FULL PAPER
Figure 1. Selected robust metal-organic frameworks with different sizes of windows in this study.
Figure 2. (a) PXRD and (b) FT-IR of MIL-100(Fe), 7%PTA@MIL-100(Fe), 16%PTA@MIL-100(Fe), and 35%PTA@MIL-100(Fe), (c) PXRD and (d) FT-IR of
6%PTA@MIL-100(Fe) and the recycled samples.
1
pentagonal windows and 8.6×8.6 Å for hexagonal windows) of
MIL-100(Fe) was illustrated by pore size distribution (PSD) of
MIL-100(Fe), which was calculated by nonlocal density
functional theory (NLDFT) model in Figure 3b. Notably, the
mesopores and microporous windows were also retained after
loading large amount of PTA, especially for 7%PTA@MIL-
substrates enter into the pores and contact with the active site
PTA. More interestingly, due to the relatively small windows and
large mesocages, the PTA (10.0 Å) were confined inside the
mesopores of MIL-100(Fe), which could prevent the leaching of
PTA during catalytic reactions.
SEM and TEM revealed that the octahedral shape of MIL-
100(Fe) was preserved after incorporation of PTA (Figure 4 and
Figure S1). The Fe, P, and W elements were clearly observed in
100(Fe), 16%PTA@MIL-100(Fe). The preserved mesopores
and windows of the PTA@MIL-100(Fe) could ensure that
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601450
FULL PAPER
2
Figure 3. The N adsorption-desorption isotherm at 77K (a), and pore size distribution (b) of the MIL-100(Fe), 7%PTA@MIL-100(Fe), 16%PTA@MIL-100(Fe),
and 35%PTA@MIL-100(Fe).
Table1. The BET surface area and pore volume of MIL-100(Fe) and PTA@MIL-100(Fe).
2
3
Catalyst
BET surface area (m /g)
Pore volume (cm /g)
MIL-100(Fe)
2155
2056
1627
1046
0.993
0.927
0.861
0.487
7%PTA@MIL-100(Fe)
16%PTA@MIL-100(Fe)
35%PTA@MIL-100(Fe)
Figure 4. (a) SEM, (b) HAADF-STEM image and (c-e) the corresponding EDX mapping of 16%PTA@MIL-100(Fe).
the energy-dispersive X-ray (EDX) of 16%PTA@MIL-100(Fe)
indicating that PTA have been incorporated into MIL-100(Fe)
(
Figure S2). The EDX elemental mapping further revealed that
Oxidative desulfurization performance
the elements of Fe, P, W were uniformly distributed throughout
the entire structure of MIL-100(Fe) (Figure 4b).
Optimization of oxidative desulfurization parameters
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601450
FULL PAPER
Figure 5. Graphs of optimizing the catalytic conditions. (a) Loading amount, (b) dosages, (c) temperature, and (d) O/S ratio over PTA@MIL-100(Fe).
Figure 6. (a) Oxidative desulfurization of DBT under different reaction time; (b) Kinetic profile for the oxidative of DBT.
Before fully understand the catalytic performance influenced by
the window size of porous catalysts, the effect of the oxidative
desulfurization parameters including PTA loading amount,
organosulfur compounds), reaction temperature and time were
investigated to find the optimal catalytic conditions (initial
reaction conditions: catalyst dosage 15 mg/0.3mol%, O/S molar
o
catalyst dosage, O/S molar ratio (the molar ratio of H
2
O
2
to
ratio 2, temperature 80 C, reaction time 1 h). As DBT is one of
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601450
FULL PAPER
dosage can improve the catalytic performance. Interestingly, the
conversion of DBT reached to 78% using only 0.6 mol% (30 mg)
PTA, which can be attributed to the well dispersed PTA
molecules in the nanocages and the formation of isolated active
site. When the dosage increased to 1.0 mol% (50 mg), the
activity was similar to that of using 0.6 mol%. The conversion
rate of the oxidative desulfurization was also affected by reaction
temperature. The catalytic results of 16%PTA@MIL-100(Fe) for
ODS at different temperatures ranging from 323 K to 373 K are
shown in Figure 5c. The conversion of DBT increased with
raising temperature firstly, and reached to maximum with 84% at
o
7
0 C due to the formation of active metallic peroxides at high
[
7, 9]
temperature.
However, the activity decreased when the
temperature continue to increase, which was ascribed to the
Figure 7. Molecular structures and the sizes of the representative
organosulfur compounds in this study.
increased decomposition rate of H O2 at higher temperature.
2
The molar ratio of O/S is a very important factor in industry. In
previous reports, most of catalytic processes need high O/S
molar ratio due to the side reactions involving the thermal
[
10, 13]
decomposition of hydrogen peroxide.
a high economic value, the use of H
Therefore, to achieve
2
O as small amount as
2
[
37]
possible is highly desirable. To investigate the influence of the
amount of H on the catalytic performance, O/S ratios from 1
2
O
2
to 10 were studied at optimal reaction temperature (Figure 5d).
With the O/S molar ratio raising, the activity increased at first,
then decreased, and the highest activity when the O/S was 4.
Notably, the conversion rate can reach to 70% at stoichiometric
ratio of O/S = 2, and 96% at O/S = 3. Further increasing the O/S
ratio to 4, the conversion rate can reach to 100%. Although the
2 2
high contents of H O can increase the catalytic activity, the
produced excess water (from ODS reaction and thermal
decomposition) will hinder the ODS reaction and lead to low
activity. These results indicated that the balance of the two
opposing factors is very important, which can be achieved by
[
9, 10, 13]
controlling of the oxidant reagent amount.
More
significantly, compared with other catalysts such as PTA-CeO
2
[10, 13]
and V
2
O
5
-TiO
2
,
the less required H
2
O
2
indicating that
Figure 8. Oxidative desulfurization of refractory organosulfur compounds
using catalysts with different window size.
16%PTA@MIL-100(Fe) is an effective catalyst for ODS. The
conversion of DBT versus reaction time at optimal conditions
over 16%PTA@MIL-100(Fe) was given in Figure 6a. The GC
yield can achieve 90% in 40 min, and fully conversion was
the most refractory organosulfur compounds, it was selected as
the model substrate to optimize the reaction conditions of
oxidative desulfurization over PTA@MIL-100(Fe).
-1
obtained in 1 hour. The rate constant k (0.065 min ) for the
t
oxidation of DBT was afforded from the linear plot of ln(C /C )
0
versus reaction time t as shown in Figure 6b and the calculation
procedures can be seen in supplementary information (S2.11
Calculation method of rate constant k). Moreover, control
experiments summarized in Table S1 shows that all catalysts
including PTA, MIL-100(Fe) and PTA@MIL-100(Fe) have no
The PTA loading amount had a significant effect for this
reaction (Figure 5a). As shown in Figure 5a, it should be noted
that the parent MIL-100(Fe) exhibited no activity in ODS process,
whereas PTA encapsulated in MIL-100(Fe) exhibited effective
catalytic performance. The catalytic activity of 16%PTA@MIL-
activity without the existence of H
PTA@MIL-100(Fe), the PTA or MIL-100(Fe) also shown
negligible activity with H existing, which confirmed the
2 2
O (Entry 2-4). Compared with
100(Fe) was more than twice of that of 7%PTA@MIL-100(Fe),
mainly because the high loading of PTA could provide more
active sites. However, further increasing the loading of PTA to
2
O
2
synergistic effect of two components (Entry 6-8).
Exploitation of window size–activity relationship
35%, the conversion of DBT decreased to 28% due to the
excess PTA partially blocked the pores and the reactants were
difficult to diffuse to the PTA active sites in the cages (Figure 3a
and Table 1). Consequently, 16%PTA@MIL-100(Fe) was
chosen as the optimized catalyst in the subsequent study for
ODS process. Notably, only DBT sulfone product was detected
by GC-MS. As shown in Figure 5b, increasing the catalyst
To explore the effect of the window size, three robust catalysts
PTA@MIL-100(Fe), PTA@UiO-66 and PTA@ZIF-8 were utilized
for oxidative desulfurization. The structures and the sizes of the
selected MOFs and three refractory organosulfur compounds
DBT, BT and 4,6-DMDBT are shown in Figure 1 and Figure 7,
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601450
FULL PAPER
PTA@ZIF-8 also exhibited the same ODS order for refractory
organosulfur compounds. This order can be attributed to
comprehensive effects of the electron density on the sulfur and
the steric hindrance around the sulfur atom. However, the
activities for PTA@UiO-66 followed the order of DBT > BT > 4,6-
DMDBT, which is different from PTA@MIL-100(Fe). This
difference can be attributed to that the big 4,6-DMDBT can only
reacted at the external surface of the catalyst and led to the low
activity. Moreover, the high activity of BT by PTA@UiO-66 was
result from the synergistic effect of UiO-66 and PTA. The defect
vacancies exist in the catalyst PTA@UiO-66 and the produced
open Lewis acid active site could also increase the
[
31]
desulfurization activity. The control experiments demonstrated
that UiO-66 showed high catalytic activity with respect to DBT,
whereas MIL-100(Fe) and ZIF-8 exhibited almost no activity
(
Entry 7, 10, 14, Table S1). This is mainly because the formation
IV
of Zr -peroxo groups on the surface of the UiO-66 by the
interaction with H , which is not easy to be formed on the
materials of MIL-100(Fe) and ZIF-8.
O
2
Figure 9. The recyclability and reusability of 16%PTA@MIL-100(Fe) in five
runs for DBT oxidative desulfurization.
2
[
31]
To evaluate the possible leaching of PTA from the three
robust catalysts, the W content in gasoline were analyzed using
ICP-OES after workup, respectively. Amazingly, the leaching W
into gasoline were negligible from the three robust catalysts
mainly due to the fact that the PTA molecules are confined in the
mesoporous cages and difficult to escape through the
microporous windows. Notably, a serious loss of PTA was
observed in the PTA@MIL-101(Cr) catalytic system due to the
window size of 1.6 nm which is larger than the size of PTA (1.0
respectively. As shown in Figure 8, PTA@ZIF-8 exhibited low
activity for the ODS and the conversion of 4,6-DMDBT, DBT and
BT was 25.7%, 28.6%, and 9.1%, respectively. The low activity
was ascribed to the very narrow windows (ca. 3.4×3.4 Å ) of
ZIF-8, which blocked the three organosulfur compounds to enter
in the cages of the catalyst.
2
Interestingly, the catalyst PTA@UiO-66 with a large window
2
[5]
size of 6.0×6.0 Å gave very high activity with conversion of
nm), resulting low activity of catalysts in the next several runs.
9
4.8% and 100% for BT and DBT, respectively, which can be
Therefore, the judicious choice of MOFs with suitable window
size which not only can confine POMs into the nanocages but
also the accessible of the reactants is a key factor for oxidative
desulfurization and other catalytic reactions.
mainly attributed to that the two reactants can easily diffuse into
the inner pores of UiO-66. In contrast, the compound 4,6-
DMDBT with big size can only reacted at the external surface of
the catalyst, giving a 39.1% conversion to sulfone. However, the
mesoporous PTA@MIL-100(Fe) catalyst with larger window
[
5, 14, 16-18, 38-
Based on our results and other reported literatures,
a possible mechanism for this system was illustrated. The
4
3]
2
sizes (ca. 8.6×8.6 Å ) exhibited high activity for the three
substrates 4,6-DMDBT, DBT and BT and gave a conversion of
oxidative desulfurization started with the fast diffusion of DBT
from oil phase to acetonitrile phase. At the same time, the highly
porous MOF materials could enhance the adsorption capacity
92.8%, 100%, and 61.8%, respectively. The control experiments
[
3, 17, 44,
demonstrated that the large window size could facilitate the
diffusion of the reactants and thus accelerated the reaction rate
and gave high yield.
2 2
for organosulfur compounds and co-catalyst such as H O .
4
5]
VI
Thus the PTA supported on MOF can be oxidated to W -
peroxo by the enriched following oxidized the
2 2
H O ,
[
4]
Besides the effect of window size of catalysts, the ODS
reactions were also influenced by the physical and chemical
properties of substrates including the electron density on the
sulfur and the steric hindrance around the sulfur atom. The
electron density on the sulfur atom of 4,6-DMDBT, DBT, and BT
is 5.760, 5.758, and 5.739, respectively. It is known that lower
electron density on sulfur atoms was unfavorable for ODS.
Consequently, the substrate BT gave a low activity with 61.8%
conversion by PTA@MIL-100(Fe). As 4,6-DMDBT and DBT
have the similar electron density on sulfurs, the activity mainly
affect by the steric hindrance around the sulfur atom. Thus,
PTA@MIL-100(Fe) gave a higher conversion for DBT due to the
less steric hindrance around the sulfur atom. The oxidative
activity of refractory organosulfur compounds by PTA@MIL-
concentrated DBT to sulfide and sulfone. Moreover, the highly
dispersed POM confined in MOF cage can serve as single
active site which is beneficial for accelerating the ODS process.
In particular, the suitable window size of MOF could not only
confine the PTA into the pores of MOF to prevent leaching, but
also ensure the accessible of the DBT reactants.
Recycling of catalyst
The stability of catalysts is important for desulfurization industry.
Interestingly, compared with homogeneous catalyst, the
PTA@MIL-100(Fe) catalyst can be easily separated and reused
for five consecutive ODS cycles without loss of activity (Figure 9).
2
The PXRD patterns, FT-IR spectra, N sorption isotherms, pore
size distribution, BET and pore volume of the recycled catalyst
were similar to those corresponding fresh PTA@MIL-100(Fe)
catalyst, indicating that the structure of the catalyst was retained
1
00(Fe), as discussed above, decreased in the order of DBT >
[
5, 7, 38]
4,6-DMDBT > BT, which is consistent with previous works.
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601450
FULL PAPER
and their use for applications in oxidative desulfurization, but
also for the exploration of their new applications such as other
catalysis, gas storage, drug delivery.
Table 2. ODS performance for real gasoline by PTA@MIL-100(Fe) and
PTA@UiO-66.
Catalyst
Run
Times
S Remaining
(mass ppm)
Desulfurization/%
Experimental Section
PTA@MIL-100(Fe)
PTA@MIL-100(Fe)
PTA@UiO-66
1
2
1
2
73
85
92
75
79
36
Materials
The reagents phosphotungstic acid (Sinopharm, AR), iron
120
98
powder (Adamas, 99%), trimesic acid (H
terephthalic acid (H BDC, Adamas, 99%), 2-methylimidazole
(Adamas, 99%), ZrCl (Adamas, 99%), HNO (Sinopharm,
5%wt, 15 M), HF (Sinopharm, 40%wt, 22.5 M), DBT (Adamas,
9%), BT (Adamas, 99%), 4,6-DMDBT (Adamas, 98%),
3
BTC, Aladdin, 99%),
2
PTA@UiO-66
4
3
6
9
Footnotes: the initial S content in the real gasoline was ca. 473 mass ppm.
acetonitrile (MeCN, Sinopharm, 99%), n-heptane (Ourchem,
7%), 30% H (Sinopharm, AR), n-dodecane (Aladdin, 99%),
and real gasoline (TCI, reference materials of sulfur in gasoline
certified by the Japan Petroleum Institute, sulfur content 500
mass ppm level) were used as received without further
purification.
9
2 2
O
(
Figure 2, Figure S4 and Table S2). Furthermore, hot filtration
experiment was also investigated for the stability of PTA@MIL-
00(Fe). As shown in Figure 6a, the catalytic reaction stopped
1
after the removal of catalyst. In addition, the ICP-OES analysis
exhibited that the leaching of PTA into the model gasoline was
negligible which can be attributed to the fact that the PTA
molecules are confined in the mesoporous cages and difficult to
escape through the microporous windows.
Catalysts preparation
Oxidative desulfurization of real gasoline
Preparation of porous MIL-100(Fe)
The outstanding performance of the PTA@MIL-100(Fe) and
PTA@UiO-66 in the desulfurization of the model gasoline
motivated us to apply them to a real gasoline desulfurization
The porous material MIL-100(Fe) was prepared according to the
[
30]
method described in the literature
mixture of iron powder (0.45g, 8 mmol), H
mmol), HF (0.71 mL, 22.5 M, 16 mmol), HNO
.8 mmol), and deionized water (47 mL, 2.6 mol) was
. In a typical procedure, a
BTC (1.13 g, 5.4
(0.32 mL, 15 M,
3
(
sulfur content ca. 473 mass ppm, Table 2). 75% organosulfur
3
compounds in gasoline were removed by PTA@UiO-66 after
one cycle, and total 79% can be achieved after two cycles.
Further more, as the window size of PTA@MIL-100(Fe) facilitate
all substrates accessible to active sites, the removal of
organosulfur compounds in gasoline can reach to 85% and 92%
respectively after one cycle and two cycles, respectively. The
high efficiency of PTA@MIL-100(Fe) indicated that it is a
promising candidate catalyst for future industrial desulfurization
application.
4
o
ultrasonicated for 20 min and heated at 160 C for 24 h. The
obtained orange solid was further washed with water and
ethanol three times, respectively. The orange solid was dried at
o
7
0 C over night finally. The resulting solid was referred as MIL-
00(Fe).
1
Preparation of porous PTA@MIL-100(Fe)
The encapsulation of PTA in MIL-101(Fe) was prepared
following the same procedure as for the pure MIL-100(Fe).
Typically, a mixture of iron powder (0.45 g, 8 mmol), H
1.13 g, 5.4 mmol), HF (0.71 mL, 22.5 M, 16 mmol), HNO
3
BTC
(0.32
(
3
Conclusions
mL, 15 M, 4.8 mmol), and deionized water (47 mL, 2.6 mol)
were ultrasonicated for 20 min, then 1.44 g, 2.88 g, or 5.76 g of
PTA was added into the above solution and stirred for further 10
min. Then, the autoclave was transferred to a stainless vial and
In conclusion, PTA molecules encapsulated in three robust
MOFs with different window sizes have been successfully
synthesized by a facial one-pot “bottle around ship” method. The
study of the relationship of window size–activity for ODS
performance revealed that the activity and the leaching amount
of PTA increased with the increasing of the window size. The
control experiments demonstrated that PTA@MIL-100(Fe)
exhibited high activity and good recyclability in the oxidative
desulfurization of model gasoline and real gasoline because of
its suitable window size and confined effect, which facilitates
organosulfur compounds accessible to active sites (PTA). The
study of the relationship between the structure and activity
paves the way not only for the synthesis of new porous catalysts
o
heated at 160 C for 24 h. The obtained orange solids were
further washed with water and ethanol throughly and dried at 70
o
C over night. The solids were denoted as 7%PTA@MIL-
100(Fe), 16%PTA@MIL-100(Fe), and 35%PTA@MIL-100(Fe)
respectively, based on the weight content of PTA determined by
ICP-OES.
Preparation of porous UiO-66
The procedure of preparation of UiO-66 was similar to the
[31]
method described in the literature . In a typical procedure, 699
mg ZrCl
4
(3 mmol) and 498 mg terephthalic acid (3 mmol) were
o
dissolved in 60 mL DMF, and then heated at 120 C for 24 hours.
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601450
FULL PAPER
After cooling to room temperature, the obtained white colloid
was filtered, and washed three times with DMF and ethanol
each and dried at 70 C overnight.
organosulfur compounds using catalysts with different window
size, the reaction conditions were set as: catalyst dosage
0.6mol%, O/S molar ratio 4, temperature 70 C, reaction time 24
o
o
Preparation of porous PTA@UiO-66
h.
The encapsulation of PTA in UiO-66 was prepared by one pot
method following the same procedure as for the pure UiO-66,
except 1.44g PTA was added in the solution during the
synthesis procedure. The resulting solid was denoted as
The desulfurization procedure of real gasoline is similar to the
above model gasoline. In a typical reaction, 16%PTA@MIL-
100(Fe) (30 mg), MeCN (2 mL), and real gasoline (2 mL) were
2 2
placed in a reaction tube (15 mL) under stirring. Then, H O (20
1
0%PTA@UiO-66, based on the weight content of PTA
μL) was added to the mixture and the reaction started after the
tube was put into an oil bath at 70 °C. After 24 hours, the
reaction stopped by putting the tube into an ice bath. The
catalyst were washed throughly with ethyl acetate and dried in
an 50 °C oven. Then it was used as catalyst for the second
desulfurization cycle. The qualitative analysis of sulfur content in
the real gasoline was determined through ultraviolet
fluorescence method by Institute of Quality Inspection for
Products (Fuzhou, Fujian province, China).
determined by ICP-OES.
Preparation of porous ZIF-8
The porous material ZIF-8 was prepared according to the
[
32]
method described in the literature
10 mg (9.874 mmol) 2-methylimidazole dissolved in 25 mL of
MeOH was poured into a 25 mL MeOH solution containing 734
mg (2.469 mmol) Zn(NO ·6H O under stirring. After 24 h, the
. In a typical procedure,
8
3
)
2
2
solid was obtained by centrifugation, and washed with fresh
MeOH throughly. At last, the product is dried at 70 C overnight.
o
Hot filtration and recycling experiment
Preparation of porous PTA@ZIF-8
In a typical hot filtration experiment, 16%PTA@MIL-100(Fe) (30
mg), MeCN (2 mL), and model gasoline (2 mL) were placed in a
reaction tube (15 mL) under stirring. Then, H O (20 μL) was
2 2
added into the mixture. The reaction started after the reaction
tube was put into an oil bath at 70 °C. After 20 min, the solid
catalyst was removed by syringe filter, and the conversion rate
The encapsulation of PTA in ZIF-8 was prepared following the
same procedure as for the pure ZIF-8, except 5 mL (10 mM) of
predesigned PTA solution was added during the synthesis
procedure. The resulting solid was denoted as 10%PTA@ZIF-8,
based on the weight content of PTA determined by ICP-OES.
of DBT was detected by GC. The filtrate was further treated at
o
7
0 C, and analysed in a period of 20 min.
Catalytic test
In the recycling experiment, the catalyst was centrifugated
after two hours reaction and washed throughly with ethyl acetate.
To detect the leaching amount of PTA from catalysts, both of the
phases of gasoline and acetonitrile were analyzed by detecting
the W leaching.
Preparation of model gasoline and the property of real
gasoline
20 mmol DBT and 2 mL dodecane (internal standard) were
mixed with n-heptane to form 1000 mL model gasoline with
sulfur content ca. 950 mass ppm. Other two kinds of model
gasoline (BT and 4,6-DMDBT) were prepared with the similar Acknowledgements ((optional))
procedure and the sulfur contents are 950 mass ppm. For the
real gasoline, the sulfur content is ca. 473 mass ppm and the
density is 0.8287 g/cm at 15 C.
Catalytic tests using model and real gasoline
The refractory organosulfur compounds can be extracted by
MeCN and readily oxidized to the corresponding sulfones in the
bi-phase ODS process (Scheme 1). The bi-phase ODS system
was carried out in a reaction tube (15 mL) equipped with a
magnetic stirrer at different temperature. The catalytic
performance of the catalysts were investigated by oxidation of
refractory organosulfur compounds (DBT, BT or 4,6-DMDBT) in
The authors gratefully acknowledge the financial support of the
3
o
973 Program (2014CB845605 and 2013CB933200), NSFC
(21671188, 21273238, 21521061, and 21331006), Strategic
Priority Research Program of the Chinese Academy of Sciences
(XDB20000000), Youth Innovation Promotion Association, CAS
(2014265), and Chunmiao Project of the Haixi Institute of the
Chinese Academy of Sciences (CMZX-2014-004).
Keywords: desulfurization • heterogeneous catalysis • metal-
organic frameworks • phosphotungstic acid • real gasoline.
the presence of H
identical to MeCN. In a typical reaction, 16%PTA@MIL-100(Fe)
30 mg, 0.6 mol%), MeCN (2 mL), and model gasoline (2 mL)
were placed in a reaction tube (15 mL) under stirring. Then,
(20 μL) was added to the mixture and the reaction started
2 2
O oxidant. The volume of model gasoline is
[
1]
B. Pawelec, R. M. Navarro, J. M. Campos-Martin, J. L. G.
Fierro, Catal. Sci. Technol. 2011, 1, 23.
(
[2]
A. Gómez-Paricio, A. Santiago-Portillo, S. Navalón, P.
Concepción, M. Alvaro, H. Garcia, Green Chem. 2016, 18,
2 2
H O
5
08.
after the tube was put into an oil bath at 70 °C. Finally, the
reaction stopped by putting the tube into an ice bath. The
quantitative analysis of various sulfur compounds in the model
gasoline and the polar solvent were performed on a gas
chromatograph (Angilent G7890A) equipped with a FID detector
and a gas chromatography–mass spectrometry (Varian 450-
GC/240-MS). To study the oxidative desulfurization of refractory
[
3]
F. Shi, M. Hammoud, L. T. Thompson, Appl. Catal. B:
Environ 2011, 103, 261.
V. Chandra Srivastava, RSC Adv. 2012, 2, 759.
S. Ribeiro, A. D. S. Barbosa, A. C. Gomes, M. Pillinger, I.
S. Gonçalves, L. Cunha-Silva, S. S. Balula, Fuel Process.
Technol. 2013, 116, 350.
[
[
4]
5]
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601450
FULL PAPER
[
6]
C. M. Granadeiro, L. S. Nogueira, D. Julião, F. Mirante, D. [28] C.-Y. Sun, S.-X. Liu, D.-D. Liang, K.-Z. Shao, Y.-H. Ren,
Ananias, S. S. Balula, L. Cunha-Silva, Catal. Sci. Technol. Z.-M. Su, J. Am. Chem. Soc. 2009, 131, 1883.
016, 6, 1515.
X. S. Wang, Y. B. Huang, Z. J. Lin, R. Cao, Dalton Trans.
2
[29] J. Juan-Alcañiz, J. Gascon, F. Kapteijn, J. Mater. Chem.
2012, 22, 10102.
[
7]
2
014, 43, 11950.
[30] B. Van de Voorde, M. Boulhout, F. Vermoortele, P.
Horcajada, D. Cunha, J. S. Lee, J. S. Chang, E. Gibson,
M. Daturi, J. C. Lavalley, A. Vimont, I. Beurroies, D. E. De
Vos, J. Am. Chem. Soc. 2013, 135, 9849.
[
[
8]
9]
C. S. Song, Catal. Today 2003, 86, 211.
G. Q. Luo, L. H. Kang, M. Y. Zhu, B. Dai, Fuel Process.
Technol. 2014, 118, 20.
[
[
[
[
[
[
[
10] M. Zhang, W. Zhu, S. Xun, H. Li, Q. Gu, Z. Zhao, Q.
Wang, Chem. Eng. J. 2013, 220, 328.
11] W. Zhu, B. Dai, P. Wu, Y. Chao, J. Xiong, S. Xun, H. Li, H.
Li, ACS Sustain. Chem. Eng. 2015, 3, 186.
12] J. Zhang, A. J. Wang, Y. J. Wang, H. Y. Wang, J. Z. Gui,
Chem. Eng. J. 2014, 245, 65.
13] L. C. Caero, E. Hernández, F. Pedraza, F. Murrieta, Catal.
Today 2005, 107-108, 564.
14] G. Rodriguez-Gattorno, A. Galano, E. Torres-García,
Appl. Catal. B: Environ 2009, 92, 1.
15] X.-M. Yan, P. Mei, L. Xiong, L. Gao, Q. Yang, L. Gong,
Catal. Sci. Technol. 2013, 3, 1985.
16] D. Julião, A. C. Gomes, M. Pillinger, L. Cunha-Silva, B.
de Castro, I. S. Gonçalves, S. S. Balula, Fuel Process.
Technol. 2015, 131, 78.
17] Y. Liu, S. Liu, S. Liu, D. Liang, S. Li, Q. Tang, X. Wang, J.
Miao, Z. Shi, Z. Zheng, ChemCatChem 2013, 5, 3086.
18] X. L. Hao, Y. Y. Ma, H. Y. Zang, Y. H. Wang, Y. G. Li, E.
B. Wang, Chem. Eur. J. 2015, 21, 3778.
[31] C. M. Granadeiro, S. O. Ribeiro, M. Karmaoui, R. Valenca,
J. C. Ribeiro, B. de Castro, L. Cunha-Silva, S. S. Balula,
Chem. Commun. 2015, 51, 13818.
[32] J. Cravillon, R. Nayuk, S. Springer, A. Feldhoff, K. Huber,
M. Wiebcke, Chem. Mater. 2011, 23, 2130.
[33] P. Horcajada, S. Surble, C. Serre, D. Y. Hong, Y. K. Seo,
J. S. Chang, J. M. Greneche, I. Margiolaki, G. Ferey,
Chem. Commun. 2007, 27, 2820.
[34] G. Ferey, C. Serre, C. Mellot-Draznieks, F. Millange, S.
Surble, J. Dutour, I. Margiolaki, Angew. Chem. Int. Ed.
2004, 43, 6296.
[35] J. H. Cavka, S. Jakobsen, U. Olsbye, N. Guillou, C.
Lamberti, S. Bordiga, K. P. Lillerud, J. Am. Chem. Soc.
2008, 130, 13850.
[36] X. C. Huang, Y. Y. Lin, J. P. Zhang, X. M. Chen, Angew.
Chem. Int. Ed. 2006, 45, 1557.
[37] J. M. Campos-Martin, M. C. Capel-Sanchez, P. Perez-
Presas, J. L. G. Fierro, J. Chem. Technol. Biotechnol.
2010, 85, 879.
[
[
[
[
19] J. Jiang, O. M. Yaghi, Chem. Rev. 2015, 115, 6966.
20] Y. B. Huang, Q. Wang, J. Liang, X. Wang, R. Cao, J. Am.
Chem. Soc. 2016, 138, 10104.
21] A. H. Assen, Y. Belmabkhout, K. Adil, P. M. Bhatt, D. X.
Xue, H. Jiang, M. Eddaoudi, Angew. Chem. Int. Ed. 2015,
[38] X. Hu, Y. Lu, F. Dai, C. Liu, Y. Liu, Microporous
Mesoporous Mater. 2013, 170, 36.
[39] S. Ribeiro, C. M. Granadeiro, P. Silva, F. A. Almeida Paz,
F. F. de Biani, L. Cunha-Silva, S. S. Balula, Catalysis
Science & Technology 2013, 3, 2404.
[
5
4, 14353.
[40] C. M. Granadeiro, A. D. S. Barbosa, S. Ribeiro, I. C. M. S.
Santos, B. de Castro, L. Cunha-Silva, S. S. Balula,
Catalysis Science & Technology 2014, 4, 1416.
[41] X. Zhang, Y. Shi, G. Liu, Catal. Sci. Technol. 2016, 6,
1016.
[
[
[
22] X. S. Wang, J. Liang, L. Li, Z. J. Lin, P. P. Bag, S. Y. Gao,
Y. B. Huang, R. Cao, Inorg. Chem. 2016, 55, 2641.
23] L. Li, X. Wang, J. Liang, Y. Huang, H. Li, Z. Lin, R. Cao,
ACS Appl. Mater. Inter. 2016, 8, 9777.
24] J. S. Qin, D. Y. Du, M. Li, X. Z. Lian, L. Z. Dong, M.
Bosch, Z. M. Su, Q. Zhang, S. L. Li, Y. Q. Lan, S. Yuan, H.
C. Zhou, J. Am. Chem. Soc. 2016, 138, 5299.
25] H. Furukawa, K. E. Cordova, M. O'Keeffe, O. M. Yaghi,
Science 2013, 341, 1230444.
26] J. Liu, L. Chen, H. Cui, J. Zhang, L. Zhang, C. Y. Su,
Chem. Soc. Rev. 2014, 43, 6011.
27] W.-H. Zhang, Q. Liu, J.-P. Lang, Coord. Chem. Rev.
[42] Y. Ding, W. Zhu, H. Li, W. Jiang, M. Zhang, Y. Duan, Y.
Chang, Green Chem. 2011, 13, 1210.
[43] N. Tang, Y. Zhang, F. Lin, H. Lu, Z. Jiang, C. Li, Chem.
Commun. 2012, 48, 11647.
[44] W.-W. He, S.-L. Li, W.-L. Li, J.-S. Li, G.-S. Yang, S.-R.
Zhang, Y.-Q. Lan, P. Shen, Z.-M. Su, J. Mater. Chem. A
2013, 1, 11111.
[
[
[
[45] J.-X. Qin, P. Tan, Y. Jiang, X.-Q. Liu, Q.-X. He, L.-B. Sun,
2015, 293-294, 187.
Green Chem. 2016, 18, 3210.
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201601450
FULL PAPER
FULL PAPER
Xu-Sheng Wang, Lan Li, Jun Liang,
Yuan-Biao Huang,* and Rong Cao*
Page No. – Page No.
Boosting Oxidative Desulfurization of
Model and Real Gasoline over
Phosphotungstic Acid Encapsulated
in Metal-Organic Frameworks: the
Window Size Matters
The window size to success: Compared with UiO-66 and ZIF-8 with small window
size, phosphotungstic acid encapsulated in the mesoporous MIL-100(Fe) with
larger window sizes exhibited efficient desulfurization performance of model and
real gasoline due to the confine of the PTA into the pores of MOF and ensure the
accessible of all the organic sulfurs into the confined active sites in the pores.
This article is protected by copyright. All rights reserved.