Immobilized ionic liquids based on molybdenum- and tungsten-containing heteropoly acids: Structure and catalytic properties in thiophene oxidation

Article abstract of DOI:10.1134/S0965544117100164

Comparative analysis of heterogeneous catalysts for the peroxide oxidation of thiophene is conducted. The catalysts are imidazole ionic liquids (ILs) containing anions of phosphomolybdic and phosphotungstic acids immobilized on mineral supports. Immobilization is implemented by the covalent bonding of an IL fragment to a silica surface or by adsorption on alumina. The catalysts are active not only in the model oxidation of thiophene in isooctane, but also in the desulfurization of a straight-run diesel fraction and the “synthetic crude oil” derived from oil shale.

Full text of DOI:10.1134/S0965544117100164

ISSN 0965-5441, Petroleum Chemistry, 2017, Vol. 57, No. 10, pp. 859–867. © Pleiades Publishing, Ltd., 2017.
Original Russian Text © I.G. Tarkhanova, A.V. Anisimov, A.K. Buryak, A.A. Bryzhin, A.G. Ali-Zade, A.V. Akopyan, V.M. Zelikman, 2017, published in Neftekhimiya, 2017,
Vol. 57, No. 5, pp. 536–544.
Immobilized Ionic Liquids Based on Molybdenum-
and Tungsten-Containing Heteropoly Acids:
Structure and Catalytic Properties in Thiophene Oxidation
I. G. Tarkhanovaa, *, A. V. Anisimova, A. K. Buryakb, A. A. Bryzhina, c,
A. G. Ali-Zadea, A. V. Akopyana, and V. M. Zelikmana
aFaculty of Chemistry, Moscow State University, Moscow, Russia
bFrumkin Institute of Physical Chemistry and Electrochemistry, Russian Academy of Sciences, Moscow, Russia
cMendeleev University of Chemical Technology, Moscow, Russia
*e-mail: itar_msu@mail.ru
Received January 18, 2017
Abstract⎯Comparative analysis of heterogeneous catalysts for the peroxide oxidation of thiophene is con-
ducted. The catalysts are imidazole ionic liquids (ILs) containing anions of phosphomolybdic and phospho-
tungstic acids immobilized on mineral supports. Immobilization is implemented by the covalent bonding of
an IL fragment to a silica surface or by adsorption on alumina. The catalysts are active not only in the model
oxidation of thiophene in isooctane, but also in the desulfurization of a straight-run diesel fraction and the
“synthetic crude oil” derived from oil shale.
Keywords: thiophene oxidation, hydrogen peroxide, immobilized metal-containing ionic liquids, phospho-
molybdic and phosphotungstic heteropoly acids
DOI: 10.1134/S0965544117100164
In recent years, the use of catalyst composites syn- addition to a thiophene solution, in the desulfurization
thesized by the immobilization of ionic liquids (ILs) of a straight-run diesel fraction and “synthetic crude
on porous materials with a developed surface has oil” derived from oil shale.
become increasingly common [1]. Synthetic
approaches consist in the formation of ultrathin IL
EXPERIMENTAL
Catalyst Synthesis by the Covalent Immobilization
of an IL (I, Scheme 1)
layers containing a transition-metal derivative, typi-
cally, in the anionic form. Immobilization is imple-
mented by the covalent bonding of an IL fragment to a
silica surface or by adsorption. These methods make it
possible to significantly decrease the IL consumption
and simplify the procedure for separating the reaction
products from the catalyst.
IL-based heterogeneous catalysts are used in a
variety of processes; a promising direction is the use of
immobilized ILs—most often imidazole—for the oxi-
dative desulfurization of petroleum feedstocks [2–8].
Analysis of the literature suggests that the most active
composites include derivatives of Mo- and W-con-
taining Keggin-type heteropoly acids (HPAs), in par-
ticular, phosphomolybdates and phosphotungstates.
The preparation of silicas modified with an imidaz-
ole IL was described in detail in [8, 10]. The dehydrox-
ylation of the surface of a BASF Perlkat 97-0 silica gel
(specific surface area, Ssp = 500 m2/g; effective pore
diameter, deff = 10 nm) was implemented by azeo-
tropic distillation. To this end, the silica gel and tolu-
ene placed in a 100-mL round-bottomed flask
equipped with a Dean–Stark head and a reflux con-
denser were boiled until the cessation of water evolu-
tion. After that, 3-chloropropyltrimethoxysilane was
added to the mixture in the flask in a ratio of 1 : 4 with
respect to silica gel and boiled under stirring for 6 h.
Next, the solid phase was decanted, washed with tolu-
ene, and air-dried. Quaternization was implemented
as follows: the support modified with 3-chloropropyl-
trimethoxysilane was placed in a glass tube and
impregnated with ethylimidazole (four- to fivefold
excess); the tube was evacuated to a residual pressure
The aim of this study is the synthesis and compar-
ative analysis of heterogeneous catalysts based on the
above compounds supported on silica and alumina by
different methods. The model substrate is thiophene,
which is the most difficult-to-oxidize compound in
the series of sulfur-containing derivatives contained in
petroleum feedstocks [9]. The catalysts are tested, in (10–2 torr), sealed, and subjected to thermal condi-
859

860
TARKHANOVA et al.
tioning at 180°C for 24 h. After that, the sample was was admixed with a weighed portion (1 g) of Perlkat
removed from the tube, washed with rectified ethanol
from excess ethylimidazole, and air-dried to constant
weight.
Phosphomolybdic (PMA) and phosphotungstic
(PTA) HPAs were deposited via the exchange reaction
(Scheme 1). To this end, 0.5 g of PMA (or 0.6 g of
PTA) was dissolved in ethanol (15 mL); the solution
modified by the above method; the resulting mixture
was stirred for 24 h. The liquid phase was decanted; the
solid residue was washed twice with alcohol and sub-
jected to a procedure described in [11]. Light-end
products were removed from the sample placed in a
tube in a vacuum at 85–90°С. This method was used
to synthesize PMo–SiO2 and PW–SiO2 catalysts.
Cl−
O
O
H3[P(Mo12O40)]
Si
N
N
Si
N
N
+
+
or
O
O
OMe
OMe
H3[P(W12O40)]
SiO2
SiO2
MomOn2−or P(W lOk)2−
O
O
N
N
Si
OMe
O
O
+
Si
OMe
Cl−
N
N
+
Scheme 1. Deposition of HPAs on a modified silica gel; m = 2–5; n = 7–13; l = 2–7; and k = 4–20.
Catalyst Synthesis by Adsorption (II, Schemes 2, 3)
nitrogen, unsealed, and subjected to thermal condi-
tioning at 100°C for 7 h. The resulting gray-brown salt
crystals were washed twice with ethyl acetate (10 mL)
Two milliliters of ethylimidazole and 3 mL of
dichloroethane were placed in a glass tube; the tube
was evacuated to 10–2 Torr under cooling with liquid and air-dried.
CH3
CH3
Cl−
N
N
Cl
+
Cl
+
N
N
Cl
Scheme 2. Synthesis of 1-chloroethyl-3-ethylimidazolium chloride.
A metal-containing IL (Scheme 3) was synthesized of PMA (3 g per 10 mL of ice water) cooled to 0°C.
by a modified procedure described in [12]. One gram After that, the aqueous phase was decanted, and the
of the molten IL—1-chloroethyl-3-ethylimidazolium solid residue was air-dried.
chloride—was added dropwise to an aqueous solution
Cl
H3C
H3[P(Mo12O40)]
CH3
N
N
or
Cl
N
Cl
+
N
H3[P(W12O40)]
+
Cl−
Cl−
3−
N
+
N
N
P[Mo(W)]12O40
H3C
CH3
+
N
Scheme 3. Synthesis of 1-chloroethyl-3-ethylimidazolium phosphomolybdate
and 1-chloroethyl-3-ethylimidazolium phosphotungstate.
In the case of PTA, 2.3 g of molten 1-chloroethyl- aqueous phase was decanted, and the solid residue was
air-dried.
3-ethylimidazolium chloride was added dropwise to
an aqueous solution of PTA (2.6 g in 10 mL of water)
The immobilization of 1-chloroethyl-3-ethylimid-
under stirring at room temperature. After that, the azolium phosphomolybdate and 1-chloroethyl-3-
PETROLEUM CHEMISTRY
Vol. 57
No. 10
2017

IMMOBILIZED IONIC LIQUIDS BASED
861
ethylimidazolium phosphotungstate on γ-Al2O3 (Ssp = (NH3 TPD) on a Unisit USGA-101 sorbent analyzer.
180 m2/g; pore diameter, deff = 1 nm) was conducted
at room temperature. One gram of 1-chloroethyl-3-
ethylimidazolium phosphoromolybdate was dissolved
in 15 mL of ethyl alcohol; the solution was admixed
with a weighed portion of γ-Al2O3 (4 g); the resulting
mixture was held in a Branson-1200 ultrasonic bath
for 15 min. The resulting solution was decanted; the
solid phase—a PMo/Al2O3 sample—was air-dried.
1-Chloroethyl-3-ethylimidazolium phosphoro-
tungstate (1.8 g) was dissolved in 15 mL of ethyl alco-
hol; the solution was admixed with a weighed portion
of γ-Al2O3 (7.2 g); the resulting mixture was stirred on
a magnetic stirrer for 12 h. The aqueous phase was
decanted; the solid PW/Al2O3 sample was air-dried.
The sample was calcined in a dry helium stream at
150°C and then cooled to room temperature. Adsorp-
tion of ammonia (diluted with nitrogen in a ratio of
1 : 1) was conducted at 60°C for 30 min. The physi-
cally sorbed ammonia was purged in a helium stream
at 100°C for 1 h. NH3 TPD experiments were con-
ducted in a temperature range of 60–600°C in a dry
helium stream (feed rate of 30 mL/min). The heating
rate was 8°C/min.
Catalytic Testing
and Product Analysis Procedures
The catalytic activity of the immobilized ILs was
measured for the following three model processes.
Thiophene oxidation with hydrogen peroxide in
isooctane. Ten milliliters of a model mixture (1 wt %,
0.076 M thiophene solution in isooctane), 0.08 g of
the catalyst, and an oxidizer—35 or 50% hydrogen
peroxide in different amounts (0.4–0.8 mL)—were
placed in a thermostated reactor. The mixture in the
reactor was stirred under heating (50–70°C), while
taking samples for analysis from the liquid phase when
required.
Organic compounds in the reaction mixture were
identified by nuclear magnetic resonance (NMR).
Spectra were recorded on a Bruker Avance 600 instru-
ment at room temperature. The products contained in
the aqueous phase were analyzed by electrospray ion-
ization (ESI) mass spectrometry. High-resolution
spectra were recorded on a Bruker micrOTOF II
instrument. The mass range was 50–3000 m/z. Mea-
surements were conducted at the Department of
Structural Studies, Zelinsky Institute of Organic
Chemistry, Russian Academy of Sciences.
Quantitative analysis of the liquid phase in the
reaction mixture was conducted by gas–liquid chro-
matography (GLC) using a Kristall 4000 chromato-
graph equipped with a 2-m-long packed column with
the SE-30 stationary phase (nonpolar silicon) and a
flame-ionization detector. The content of thiophene
and thiophene oxidation products was determined by
the internal standard method with linear temperature
programming of 90–220°C.
Catalyst Composition
and Structure Determination Procedures
The content of molybdenum and tungsten on the
catalyst surface was determined photometrically using
catechol, which forms a stable complex compound
with molybdate and tungstate in the presence of
sodium sulfite and sodium hydroxide [13, 14]. The
analyzed catalyst sample was heated in concentrated
sulfuric acid (1 mL) at 100°C for 3 h. After cooling, the
resulting solution was transferred to a graduated cylin-
der and diluted to 10 mL with distilled water. After
that, it was neutralized with an aqueous NaOH solu-
tion to a pH of 6. The catechol solution was prepared
as follows: 1.5 g of sodium sulfite was added to 50 mL
of a 0.4% sodium hydroxide solution in a graduated
cylinder. After salt dissolution, 1 g of catechol was
added under stirring; the solution volume was adjusted
to 100 mL with distilled water.
To provide the photometric determination of
molybdenum, 2 mL of a molybdic acid solution and
10 mL of a catechol solution prepared as described
above were mixed; the optical density of the resulting
mixture at 420 nm was measured. The amount of
tungsten was determined by a similar procedure, while
measuring the optical density at 350 nm. Electronic
spectra were recorded on an Ocean Optics
HR4000CG-UV-NIK instrument.
Data on the molecular composition of the com-
pounds and their distribution over the catalyst surface
were derived by surface-assisted laser desorption/ion-
ization (SALDI) mass spectrometry. Mass spectra for
the test samples were recorded in the RN Pep Mix
mode on a Bruker Ultraflex instrument equipped with
a nitrogen laser (wavelength, 337 nm; energy, 110 μJ)
using a time-of-flight mass analyzer. The spectra were
recorded in the negative ion detection mode using a
reflectron. Cluster ions were identified with respect to
isotopic distribution using the IsoPro simulation soft-
ware [15].
Oxidative desulfurization of a 180–360°C straight-
run diesel fraction produced at the Ryazan refinery.
Ten milliliters of the fuel (total sulfur of 8800 ppm),
53–111 mg of the catalyst, and 1.13 mL of an oxidizer
(35% hydrogen peroxide) were added to a jacketed
reactor equipped with a magnetic stirrer. The mixture
was stirred at a temperature of 60°C for 3 h. After that,
to remove the oil oxidation products and the catalyst
residues, the mixture was sequentially washed with
water, twice with a 95% dimethylformamide solution
The acidic properties of the catalysts were studied in water, and again with water (in all cases, a volume
by temperature-programmed desorption of ammonia ratio of 1 : 1).
PETROLEUM CHEMISTRY Vol. 57 No. 10 2017

862
TARKHANOVA et al.
Table 1. Catalyst sample composition, synthesis method, and catalytic properties in the model reaction (10 mL of a thio-
phene solution in isooctane, 0.08 g of the catalyst, 0.4 mL of 35% H2O2)
Metal content
in catalyst, wt %
Maximum thiophene
conversion, %
No.
Catalyst
Synthesis method
Covalent bonding
PMo–SiO2
1
2
3
4
4.5
12
2
45
51
53
66
PW–SiO2
PMo/Al2O3
PW/Al2O3
Adsorption
6
Oxidative desulfurization of synthetic crude oil
Comparison of the spectra in Figs. 1a and 1b shows
derived by oil shale extraction (Slantsy-2 settlement, that, in the case of molybdenum derivatives, the for-
Leningrad oblast) in straight-run gasoline (40–160°C mation of catalysts on the SiO2 surface leads to the
fraction). The extraction pyrolysis of oil shale is
described in detail in [16]. Ten milliliters of the feed-
stock (total sulfur content of 1375 ppm), 5–17 mg of
the catalyst, and 0.2 mL of an oxidizer (35% hydrogen
peroxide) were placed in a thermostated reactor. The
mixture was stirred at 60°C for 3 h. The product was
purified as described for the second model process.
Analysis of the fuel was conducted by X-ray fluores-
cence method on a Spectroscan Max (M-049-S/98)
X-ray fluorescence spectrometer.
HPA degradation and the formation of polymolyb-
dates, which are evident in the mass spectra as the fol-
lowing negative ions: Mo2O–7 , Mo4O1–2 , and Mo5O13
[8]. At the same time, the structure of the W-contain-
ing catalyst comprises the metal in the form of HPA
anions, as evidenced by comparison of the spectra in
Figs. 2a and 2b. The cluster compositions of feed PTA
and the imidazole derivative of this HPA on the silica
gel are close: the mass spectrum of the latter exhibits
phosphotungstate ion peaks containing up to 12 metal
ions.
HPA anions are more stable on the modified Al2O3
surface. The presence of HPA in the PMo/Al2O3 and
PW/Al2O3 catalysts is confirmed by the occurrence of
peaks of ions corresponding to large clusters in the
mass spectra: Mo6O2–1, PMo5O1–8, PMo4O1–5, and
PMo3O1–2 (Fig. 1c) and PW8O2–8 , PW10O3–4 , and
RESULTS AND DISCUSSION
Catalyst Composition and Physicochemical
Characteristics
Spectrophotometric analysis data on the determi-
nation of the content of Mo and W on the surface of
the synthesized catalysts are shown in Table 1.
According to the data, the metal concentration on
the catalyst surface depends on the synthesis method
and the nature of the support; the use of method I pro-
vides an increase in the metal content. This finding is
apparently attributed to the fact that the specific sur-
face area of the silica gel is larger than that of Al2O3.
PW12O4–0 (Fig. 2c). However, comparative analysis of
the spectra of pure PMA and the PMo/Al2O3 sample
shows that the formation of the latter leads to the par-
tial degradation of large particles: the cluster ions con-
tain no more than 6 Mo atoms, while the mass spectra
of pure PMA exhibit ion clusters with 12 Mo atoms. In
the case of formation of an IL based on PTA
(PW/Al2O3 sample), the anions do not undergo degra-
The use of the different supports and synthesis
methods also affects the structure of ions on the sur- dation: the spectrum exhibits peaks of large ion clus-
face, as evidenced by SALDI mass spectrometry data ters containing 10–12 W atoms. These results are con-
shown below. Figures 1a and 2a show the mass spectra
of feed PMA and PTA.
sistent with the literature data on a higher stability
of PTA compared with that of PMA and the degrada-
tion of HPA anions during the deposition of PTA on a
γ-Al2O3 surface [17, 18].
Thus, analysis of the mass spectra of the catalysts
synthesized by the different methods suggests that
imidazole derivatives of HPAs can be immobilized on
a γ-Al2O3 surface without any cardinal degradation of
anions. This finding is apparently attributed to the
These data suggest that the peaks in the mass spec-
tra of PMA (Fig. 1a) correspond to the PMo11O3–5,
PMo9O3–3, PMo8O3–0, PMo5O1–8 , and PMo4O16 ions.
The mass spectra of PTA (Fig. 2a) exhibit peaks of the
PW12O4–0, PW7O2–0, PW6O–17, PW5O1–5, PW4O14,
–
PW3O1–2 , PW2O–5 , and PWO–6 ions.
The mass spectra of the catalysts synthesized by specific features of synthesis by the adsorption
method I are shown in Figs. 1b and 2b. method: in this case, the stage of formation of a metal-
PETROLEUM CHEMISTRY
Vol. 57
No. 10
2017

IMMOBILIZED IONIC LIQUIDS BASED
863
(а)
(b)
PMo8O30
PMo4O16
Mo2O7
1000
800
Mo4O12
150
PMo9O33
600
100
50
PMo5O18
Mo5O13
400
200
PMo11O35
0
0
2250
2000 m/z
300
500
700
900
1100
250
750
1250
1750
500
1000
1500
400
600
800
1000
1200 m/z
(c)
PMo5O18
PMo4O15
300
200
100
0
PMo3O12
PMo6O21
1500
500
1000
2000
2500
3000
m/z
Fig. 1. SALDI mass spectra of (a) PMA, (b) the PMo–SiO , and (c) PMo/Al O catalysts recorded in the negative ion detection
2
2 3
mode.
containing IL precedes the deposition of the IL onto fact suggests that the latter catalyst contains stronger
the surface. In the procedure involving a silica gel, the
catalyst is formed owing to the reaction between the
HPA and the IL already immobilized on the support;
in this case, the anions of PMA and PTA undergo
complete and partial degradation, respectively, in full
accordance with the literature data on the order of sta-
bility of HPAs [17].
acid sites. The total acidity for the PMo/Al2O3 and
PMo–SiO2 samples is 65 and 59 μmol NH3/gcat,
respectively. For comparison, similar measurements
were conducted for pure supports. It was found that
the feed SiO2 sorbs ammonia at a level of about
20 μmol/g; that is, the immobilization of an IL leads
to an increase in the number of acid sites (most prob-
ably, Brønsted sites) on the surface. At the same time,
the total acidity of pure Al2O3 is 285 μmol/g; the TPD
curve exhibits a peak at 220°C. Thus, the deposition of
an IL onto the surface of this support leads to a signif-
The acidic properties of the catalysts supported on
γ-Al2O3 and SiO2 were characterized by the TPD
method. The TPD curve for the PMo–SiO2 sample
exhibits a peak at 156°C; in the case of PMo/Al2O3,
the peak is shifted to higher temperatures (200°C); this icant (fourfold) decrease in acidity.
PETROLEUM CHEMISTRY Vol. 57 No. 10 2017

864
TARKHANOVA et al.
(а)
(b)
PW3O12
PW2O5
PW3O12
PW2O5
1.25
1.00
0.75
0.50
0.25
PW4O14
6000
4000
2000
0
PWO6
PW4O14
PW5O15
PW5O15
PW6O17
PW7O19–20
PW6O17
PW7O20
PW12O40
PW12O40
0
500 1000 1500 2000 2500 3000 m/z
500 1000 1500 2000 2500 3000 m/z
(c)
PW6O17
W7O20
1250
1000
750
PW12O40
W3O11–12
PW8O28
500
250
0
PW10O34
500
1500
2500
3500
4500
1000
2000
3000
4000 m/z
Fig. 2. SALDI mass spectra of (a) PTA, (b) the PW–SiO , and (c) PW/Al O catalysts recorded in the negative ion detection
2
2 3
mode.
Comparison of the Activity and Stability
of the Catalysts in Thiophene Oxidation
To determine the activity and stability of the synthe-
sized catalysts, a set of tests on the oxidation of the model mediate formation of sulfoxide and sulfone.
substrate—thiophene—was conducted. According to the
literature [5, 19–23], thiophene oxidation occurs in
accordance with Scheme 4 (see below) through the inter-
O
HO
O
S
O O
S
HO
S
[O]
[O]
[O]
[O]
CH
+
−H2SO4
O
−H2SO4
[O]
CO2
Scheme 4. Thiophene oxidation under the action of hydrogen peroxide.
PETROLEUM CHEMISTRY
Vol. 57
No. 10
2017

IMMOBILIZED IONIC LIQUIDS BASED
865
The reaction can occur to form a sulfate anion and
60
50
40
30
20
10
0
benzoic acid or, in the case of deeper oxidation, car-
bon dioxide. Data on the activity of the synthesized
catalysts in model mixture 1 are shown in Table 1 and
Fig. 4.
To avoid a deep oxidation of thiophene (to CO2 and
H2SO4), the [H2O2] : [S] molar ratio of 5 was used at
the initial stage. However, Fig. 4 shows that, under
these conditions, the maximum thiophene conversion
did not exceed 65%. To reveal the cause of cessation of
the reaction, the organic and aqueous phases of the
reaction solution were analyzed by various methods.
According to proton magnetic resonance (PMR) and
C13 NMR, the organic phase contained only the feed
thiophene and the solvent. ESI mass spectrometry
analysis did not reveal the presence of carboxylic
acids—possible products of the reaction occurring via
the styrene formation pathway—in the aqueous phase.
The addition of barium chloride to the aqueous phase
led to the formation of a white precipitate of sulfate
and, possibly, barium carbonate. The results suggest
that, under these conditions, thiophene undergoes
deep oxidation, which, according to the reaction stoi-
chiometry, requires at least a tenfold excess hydrogen
peroxide with respect to thiophene. It should be noted
that, in this set of experiments, identical weighed por-
tions of the catalysts were used to determine the possi-
ble contribution of the acidic properties of the support
to the catalytic activity of the synthesized samples.
This assumption follows from analysis of the literature
data [5, 23]. In fact, it is known that Brønsted acid sites
have a positive effect on thiophene derivatives in the
case of heterogeneous catalytic oxidation [23]. In
addition, the introduction of acids (most commonly,
formic or acetic) to the reaction solution contributes to
the occurrence of a deeper process in homogeneous or
biphasic systems [24, 25]. However, Table 1 and the
TPD data show that a change in the acidic properties
did not have a significant effect on the activity of the
samples; this finding suggests that the nature of the
anion plays the dominant role in catalysis over immo-
bilized ILs.
PMo–SiO2 PW–SiO2 PMo/Al2O3 PW/Al2O3
Fig. 3. Stability of the catalysts immobilized on Perlkat and
γ-Al O in three consecutive cycles. Initial process condi-
2
3
tions: m = 0.08 g, [H O ] : [S] = 5, 3 h, 60°C.
cat
2 2
SiO2 catalyst (Table 1). On the other hand, according
to Table 1, the PMo/Al2O3 catalyst contained a mini-
mal amount of Mo (2 wt %); however, it was more
active than the PMo–SiO2 catalyst with a significantly
higher metal content. This effect—the inverse effect of
molybdenum content in heterogeneous composites on
their activity in the oxidation of thiophene deriva-
tives—is known from the literature [26]. Apparently, a
high metal content accelerates the hydrogen peroxide
decomposition, which is a process that competes with
the target reaction.
The effect of temperature on the thiophene oxida-
tion rate is also complex. This dependence was studied
in a range of 50–70°C using the example of the
PMo/Al2O3 catalyst; it was found that the thiophene
conversion achieves a maximum at 60–65°C. This
dependence can be attributed to the fact that, at a low
temperature, the catalyst cannot exhibit an optimum
activity, while a high temperature leads to an increase
in the oxidizer decomposition rate. An optimum pro-
cess occurs in a narrow temperature range.
To provide a more complete conversion of thio-
phene, a 50% H2O2 solution was used at a molar ratio
with thiophene of about 10. To decrease the possible
effect of the side reaction— peroxide decomposition—
and increase the efficiency of use of H2O2, it was
charged in batches (0.4 mL + 0.4 mL after 3 h). This
technique is known in the literature on the peroxide-
mediated oxidation of various substrates [27, 28]. Our
data are shown in Table 2.
Figure 3 shows that the highest activity in the first
cycle was exhibited by the PMo/Al2O3 catalyst synthe-
sized by adsorption. However, after testing this sample
in several consecutive cycles, the thiophene conver-
sion significantly decreased, while the PMo–SiO2
catalyst was found to be much more stable. The
PW/Al2O3 catalyst exhibited fairly high stability;
moreover, in the second cycle, the activity of this sam-
ple considerably increased; this finding suggests that
the catalyst underwent conditioning. This effect also
took place for the catalysts synthesized by covalent
bonding; however, it was less pronounced.
It is evident from the table that the charging of
H2O2 in batches was efficient: it provided an increase
in the thiophene conversion almost to 90%.
Of two tungsten catalysts, the PW/Al2O3 catalyst
provided a lower reaction rate in the first cycle. This
The catalyst activity in petroleum feedstocks is
shown in Table 3. The catalysts supported on γ-Al2O3
result could be attributed to the fact that the tungsten proved to be the most active in the oxidation of the
content in this sample is lower than that in the PW– diesel fraction. The highest activity was exhibited by
PETROLEUM CHEMISTRY Vol. 57 No. 10 2017

866
TARKHANOVA et al.
Table 2. Total conversion of thiophene (%) for the different modes of peroxide charging (10 mL of a thiophene solution
in isooctane, 0.08 g of the catalyst, 0.8 mL of 50% H2O2)
Thiophene conversion, %
PMo–SiO2
0.4 mL + 0.4 mL of H2O2
PW–SiO2
0.4 mL + 0.4 mL of H2O2
Time, h
0.8 mL of H2O2
0.8 mL of H2O2
1
3
6
45
58
66
40
69
86
52
69
75
47
70
88
Table 3. Total conversion of sulfur-containing derivatives (%) in petroleum feedstocks (Mo(W) : H2O2 : S molar ratio of 1 : 6 : 100)
Catalyst
Conversion in diesel fraction
Conversion in synthetic crude oil
PMo–SiO2
PW–SiO2
PMo/Al2O3
PW/Al2O3
42
50
86
64
85
73
82
47
the PMA-based catalyst with a low metal content on
the surface.
ACKNOWLEDGMENTS
The authors thank A.O. Chizhov for assistance in
recording and interpreting mass spectrometry data.
This work was supported by the Russian Founda-
tion for Basic Research, project no. 15-03-01995;
OOO NIKSA; and Gercelia Limited.
In the case of an oil shale extract, the catalysts
based on molybdenum derivatives showed the best
results, regardless of the nature of the mineral support.
In general, comparison of the model solutions shows
that all of the catalysts, except for PW/Al2O3, exhibit
high activity in synthetic crude oil derived from oil
shale. This feature of this model mixture is apparently
associated with the structure of organic sulfur com-
pounds contained in it. In fact, the oil shale was pro-
cessed at high temperatures in the absence of oxidiz-
ers, such as atmospheric oxygen. Under these condi-
tions, thiophenic compounds can undergo partial
conversion to respective sulfides, which are much
more readily oxidized by hydrogen peroxide in the
presence of Mo- and W-containing catalysts [5, 16].
REFERENCES
1. Supported Ionic Liquids: Fundamentals and Applications,
Ed. by R. Fehrmann, A. Riisager, and M. Haumann
(Wiley–WCH, Weinheim, 2014).
2. M. Li, M. Zhang, A. Wei, et al., J. Mol. Catal. A:
Chem. 406, 23 (2015).
3. X. Y. Shi and J. F. Wei, J. Mol. Catal. A: Chem. 280,
142 (2008).
4. E. Kowsari, Ionic Liquids: New Aspects for the Future,
Ed. by J.-I. Kadokawa (InTech, Rijeka, 2013), ch. 11,
p. 277.
Thus, comparative analysis of the catalysts based
on immobilized metal-containing ILs has shown that
the order of catalyst activities is determined by the
nature of the substrate. At a high content of thiophene
or thiophene derivatives (model mixtures 1, 2), the
best results in the first cycle are shown by the alumina-
supported PMA-based PMo/Al2O3 catalyst synthe-
sized by the adsorption method. This catalyst contains
a minimum amount of the metal in the form of anions,
which are PMA fragments comprising up to six Mo
atoms. However, the stability of this catalyst is low.
The PTA-based catalysts containing large fragments
of heteropoly anions are more stable. In addition, the
PMA-based catalyst synthesized by covalent bonding,
which comprises polymolybdates formed as a result of
HPA degradation during synthesis, is also stable; how-
ever, it is less active.
5. A. V. Anisimov and A. V. Tarakanova, Ross. Khim. Zh.
(Zh. Ross. Khim. Ob-va Im. D.I. Mendeleeva) 52 (4),
32 (2008).
6. X. Shi, X. Han, W. Ma, et al., J. Mol. Catal. A: Chem.
341, 57 (2011).
7. J. Zhang, A. Wang, X. Li, and X. Ma, J. Catal. 279, 269
(2011).
8. I. G. Tarkhanova, A. V. Anisimov, S. V. Verzhi-
chinskaya, et al., Pet. Chem. 56, 158 (2016).
9. S. Otsuki, T. Nonaka, N. Takashima, et al., Energy
Fuels 14, 1232 (2000).
10. I. G. Tarkhanova, M. G. Gantman, and V. M. Zelik-
man, Appl. Catal., A 470, 81 (2014).
11. X. Y. Shi and J. F. Wei, J. Mol. Catal. A: Chem. 280,
142 (2008).
12. Y. Sasaki, K. Ushimaru, K. Iteya, et al., Tetrahedron
Lett. 45, 9513 (2004).
PETROLEUM CHEMISTRY
Vol. 57
No. 10
2017

IMMOBILIZED IONIC LIQUIDS BASED
867
13. A. I. Busev, Analytical Chemistry of Molybdenum (Izd. 22. L. Kong, G. Li, and X. Wang, Catal. Lett. 92, 163
AN SSSR, Moscow, 1962) [in Russian].
(2004).
14. A. I. Busev, Analytical Chemistry of Tungsten (Nauka,
23. L. Kong, G. Li, and X. Wang, Catal. Today 93, 341
Moscow, 1976) [in Russian].
(2004).
15. A. K. Buryak and T. M. Serdyuk, Usp. Khim. 82, 369
24. S. Otsuki, T. Nonaka, N. Takashima, et al., Energy
(2013).
Fuels 14, 1232 (2000).
16. S. V. Kardashev, S. V. Lysenko, S. V. Egazar’yants,
25. F. Al-Shahrani, T. Xiao, S. A. Liewellyn, et al., Appl.
et al., Khim. Tekhnol. Topl. Masel, No. 6, 3 (2015).
Catal., B 73, 311 (2007).
17. I. V. Kozhevnikov, Chem. Rev. 98, 171 (1998).
26. D. Wang, E. W. Qian, H. Amano, et al., Appl. Catal., A
253, 91 (2003).
18. J. L. Garsia-Gutierrez, G. A. Fuentes, M. E. Hernan-
dez-Teran, et al., Appl. Catal., A 334, 366 (2008).
27. Y. Bi, M. Zhou, H. Hu, et al., React. Kinet. Catal. Lett.
72, 73 (2001).
19. J. Т. Sampanthar, H. Xiao, J. Dou, et al., Appl. Catal.,
B 63, 85 (2006).
28. Z. P. Pai, D. I. Kochubey, P. V. Berdnikova, et al., J.
Mol. Catal. A: Chem. 332, 122 (2010).
20. L. Chen, S. Guo, and D. Zhao, Chin. J. Chem. Eng.
15, 520 (2007).
Translated by M. Timoshinina
21. B. Zhang, Z. Jiang, J. Li, et al., J. Catal. 287, 5 (2012).
PETROLEUM CHEMISTRY Vol. 57 No. 10 2017