Effect of Support of S Catalysts on Hydrodeoxygenation of Guaiacol as a Model Compound of Biopetroleum

Article abstract of DOI:10.1134/S1070427218020155

S/Sup catalysts were prepared from 12-molybdophosphoric heteropoly acid and cobalt citrate, with Al₂O₃, SiO₂, TiO₂, and ZrO₂ used as supports. The synthesized catalysts were studied by low-temperature nitrogen adsorption, X-ray diffraction, temperature-programmed ammonia desorption, X-ray photoelectron spectroscopy, and high-resolution transmission electron microscopy. The catalytic properties of the catalysts were studied in a flow-through installation at 260 and 340°С, pressure of 3.0 MPa, feed space velocity of 80 h^–1, and Н2/feed ratio of 500 L_n.c. L^–1. The guaiacol hydrodeoxygenation rate increases with a decrease in the mean length of the active phase particles, irrespective of the kind of the oxide support. As for the support effect, the catalyst activity decreases in the order SiO₂, Al₂O₃, ZrO₂ ~ TiO₂. On the other hand, the catalysts supported on ZrO₂ and Al₂O₃ exhibit the highest stability. The causes of the observed trends and the possible relationships between the characteristics of the catalysts and active phase nanoparticles are discussed.

Full text of DOI:10.1134/S1070427218020155

ISSN 1070-4272, Russian Journal of Applied Chemistry, 2018, Vol. 91, No. 2, pp. 270−279. © Pleiades Publishing, Ltd., 2018.
Original Russian Text © V.A. Sal’nikov, P.P. Minaev, A.V. Mozhaev, A.A. Pimerzin, V.A. Mityagin, E.I. Alatortsev, O.L. Ovsienko, P.A. Nikul’shin, 2018, published
in Zhurnal Prikladnoi Khimii, 2018, Vol. 91, No. 2, pp. 222−231.
CATALYSIS
Effect of Support of СоМоS Catalysts on Hydrodeoxygenation
of Guaiacol as a Model Compound of Biopetroleum
V. A. Sal’nikova, P. P. Minaeva, A. V. Mozhaeva, A. A. Pimerzina,
V. A. Mityaginb, E. I. Alatortsevb, O. L. Ovsienkob, and P. A. Nikul’shina,b*
a Samara State Technical University, ul. Molodogvardeiskaya 224, Samara, 443001 Russia
b All-Russia Research Institute for Petroleum Processing, ul. Aviamotornaya 6, Moscow, 111116 Russia
*e-mail: p.a.nikulshin@gmail.com
Received October 31, 2017
Abstract—СоМоS/Sup catalysts were prepared from 12-molybdophosphoric heteropoly acid and cobalt citrate,
with Al2O3, SiO2, TiO2, and ZrO2 used as supports (Sup). The synthesized catalysts were studied by low-temper-
ature nitrogen adsorption, X-ray diffraction, temperature-programmed ammonia desorption, X-ray photoelectron
spectroscopy, and high-resolution transmission electron microscopy. The catalytic properties of the catalysts were
studied in a flow-through installation at 260 and 340°С, pressure of 3.0 MPa, feed space velocity of 80 h–1, and
Н2/feed ratio of 500 Ln.c. L–1. The guaiacol hydrodeoxygenation rate increases with a decrease in the mean length
of the active phase particles, irrespective of the kind of the oxide support. As for the support effect, the catalyst
activity decreases in the order SiO2 > Al2O3 > ZrO2 ~ TiO2. On the other hand, the catalysts supported on ZrO2
and Al2O3 exhibit the highest stability. The causes of the observed trends and the possible relationships between
the characteristics of the catalysts and active phase nanoparticles are discussed.
DOI: 10.1134/S1070427218020155
The use of biological organic raw materials for the
production of motor fuels would reduce the energy
dependence on natural fossil fuels [1–3]. Active studies
are being performed today with the aim of developing
catalystsandprocessesforproductionofmotorfuelsfrom
renewable hydrocarbon raw materials. The production
of biodiesel consisting of methyl esters of fatty acids
prepared by ester interchange from vegetable or animal
raw materials steadily increases [4]. However, this
process requires putting into operation new production
facilities, which increases the production cost of the
biofuel. An alternative solution is the development of
a process for joint hydrotreating of mixtures based
on renewable resources (vegetable oils, animal fats,
biopetroleum produced by pyrolysis of wood and waste
from woodworking industry) and petroleum fractions,
with its implementation at oil refineries [1–3, 5–7].
fractions) using catalysts based on transition metal
sulfides in hydrodeoxygenation (HDO) of vegetable raw
materials with high concentration of oxygen-containing
compounds (15–40 wt %) is their rapid deactivation [8]
due to catalyst coking [3, 5, 7, 9] and to poisoning of
active sites with nitrogen- and phosphorus-containing
components present in the biomass. Deposition of
coke particles is considered as the major cause of
deactivation. Coke is formed on acid sites and plugs
support pores. Coking occurs via polymerization and
polycondensation. Their rate directly depends on the
kind of biomass components, process conditions, and
catalyst composition. For example, strong adsorption
of phenol compounds on the support prevented the
access of the other components to active sites under the
hydrotreating conditions [10–12]. The main problem in
the development of catalysts for processing vegetable
raw materials is that their activity in HDO can be
enhanced by increasing the catalyst acidity, but this
The major problem in joint hydrotreating of the
mixed raw material (biopetroleum and petroleum
232

EFFECT OF SUPPORT OF СоМоS CATALYSTS
271
Table 1. Composition and physicochemical characteristics of
the supports used
leads, as already noted, to the acceleration of coking
[11, 13, 14]. To study the catalyst deactivation in the
HDO process, Popov et al. [12] examined the adsorption
of guaiacol, phenol, and ethylphenols on the sulfide
CoMo/Al2O3 catalyst by IR spectroscopy. The surface
properties of the support and the kind and basicity of
oxygen-containing compounds proved to be decisive
factors in the interaction and poisoning of the sulfide
catalyst.
Support Concentration
of acid sites,
Textural characteristicsa
SBET
,
Vp,
μmol NH3/g
d, Å
m2 g–1 cm3 g–1
Al2O3
SiO2
200
53
208
360
0.62
0.99
96
90
The main role of the support consists in stabilization
of catalytically active particles in the highly dispersed
form, and Al2O3 is one of the most widely used supports
for HDO catalysts [15]. Its main disadvantage is strong
interaction with supported oxide precursors, prevent-
ing deep sulfidization of metals, and presence of acid
sites, leading to the coke formation on the surface in
the course of HDO. Therefore, in the development of
catalysts for hydrotreating of vegetable raw materials,
numerous studies are aimed at searching for new sup-
ports and catalytic systems capable of reducing the H2
consumption, increasing the selectivity to direct remov-
al of oxygen, and reducing the coke formation so as to
enhance the stability in the course of joint hydrotreating
under the conditions of competing hydrodeoxygenation,
hydrodesulfurization, hydrodenitrogenation, and hydro-
genation reactions.
TiO2
285
131
0.41
60
and 108
ZrO2
340
113
0.34
78
a (SBET) Specific surface area, (Vp) pore volume, and (d) effective
pore diameter.
tent of the metals in the ready catalysts was monitored
with an EDX800HS X-ray fluorescence analyzer.
The textural properties of the supports and catalysts
were determined with a Quantachrome Autosorb-1
adsorption porosimeter from the low-temperature
nitrogen adsorption. The specific surface area was
calculated using the BET model at the relative partial
pressures p/p0 in the range 0.05–0.3. The total pore
volume and the pore size distribution were calculated
from the desorption curve using the BJH model.
In this study we examined the effect exerted
by supports of sulfide catalysts prepared from
12-molybdophosphoricheteropolyacidandcobaltcitrate
on their activity and stability in hydrodeoxygenation of
guaiacol, which is widely used as a model compound of
biopetroleum [13, 14, 16–24].
The phase composition of the supports and
catalysts was determined with an ARLX’TRA X-ray
diffractometer (Cu anode, Kα radiation, λ = 1.54 Å,
38 mA, 43 kV, scanning rate 2 deg min–1).
Acid properties of supports were determined using
temperature-programmed ammonia desorption (NH3-
TPD) with a TPDRO 1100 analyzer. The samples were
pretreated as follows: calcination in a helium stream for
1 h at 550°С, cooling, saturation with ammonia at 60°С
in an NH3/N2 mixture for 15 min, removal of weakly
bound ammonia in a helium stream at 100°С for 1 h,
and cooling to 60°С. Analysis conditions: linear heating
at a rate of 8 deg min–1 to 750°С in a helium stream
(30 mL min–1).
EXPERIMENTAL
As supports we used Al2O3 prepared from
pseudoboehmite AlOOH (SASOL, Germany), SiO2
(Sigma–Aldrich), TiO2, and ZrO2 (ChemPur, Germany).
Their textural and acid characteristics are given in
Table 1.
CoMoS/Sup catalysts were prepared by single im-
pregnation of supports with a solution of 12-molybdo-
phosphoric heteropoly acid and cobalt citrate. The solu-
tion volume corresponded to the moisture capacity. The
calculated molybdenum content of the catalysts was
about 12.6 wt %, and the cobalt content, 3.7 wt %. After
the impregnation, the synthesized catalysts were dried at
80, 100, and 120°С for 2 h at each temperature. The con-
To determine the physicochemical properties, the
dried catalysts were sulfidized with dimethyl disulfide
(DMDS, Arkema, France), dissolved in the kerosene
fraction, at 240°С for 10 h and at 340°С for 6 h (feed
space velocity FSV 2 h–1, pressure 3.0 MPa). The
catalysts were unloaded from the reactor without contact
with air.
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 91 No. 2 2018

272
SAL’NIKOV et al.
The sulfidized catalysts were examined by X-ray
The TEM images of all the samples were taken in the
light field under the conditions of underfocusing without
objective aperture (phase contrast) at a resolution of
approximately 200 000. The mean length of MoS2
crystallites, L, and the number of layers in the MoS2
stack, N, of the catalyst were determined from the data
for 400–600 particles located on 10–15 different regions
of the surface.
photoelectron spectroscopy (XPS) with an Axis Ultra
DLD photoelectron spectrometer (Kratos) using AlKα
radiation (hν = 1486.6 eV). The scale of binding ener-
gies Еb was preliminarily calibrated against the posi-
tion of core level peaks: Au4f7/2 (84.0 eV) and Cu2p3/2
(932.67 eV). The samples were applied onto a double-
sided scotch tape. The charging effect arising from the
electron photoemission was reduced to a minimum by ir-
radiation of the sample surface with slow electrons using
a flood gun. The C1s line (284.8 eV) from carbon present
on the catalyst surface was used for calibration. The en-
ergy step was 1 eV for the total spectrum and 0.1 eV for
separate lines: C1s, Al2p, Co2p, S2p, and Mo3d.
The dispersity D of active phase particles was
calculated on the basis of Kasztelan hexagonal model
[25] using the equation
(4)
The relative content of particles was determined from
the areas of peaks corresponding to specific particles.
The MoS2 content was calculated using the formula
where Me is the number of Mo atoms on edges of the
mean CoMoS crystallite; Mc, number of Mo atoms on
corners of the mean CoMoS crystallite; and MT, total
number of Mo atoms in the mean particle of the active
phase.
(1)
where AMoS , AMoS O , and AMo6+ are the peak areas cor-
responding to MoSx2,ymolybdenum oxysulfides MoSxOy,
2
We also calculated the degree of promotion of active
phase crystallite edges using the equation
and molybdenum in the oxygen surrounding, Mо6+.
The relative content of the СoMoS phase was
calculated using the formula
(5)
(2)
where D is the dispersity of active phase particles,
calculated from Eq. (4) using HR TEM data.
where ACoMoS, ACo S , and AСo are the peak areas
corresponding to the CoMoS phase, Co9S8, and cobalt
in the oxygen surrounding, Со2+.
2+
The catalytic properties in guaiacol HDO were
determined with a flow-through installation under
hydrogen pressure. A tubular reactor 8 mm i.d. was
charged with 0.3 g of the catalyst (fraction 0.25–0.5 mm)
diluted with 1 cm3 of silicon carbide SiC (particle size
0.2–0.4 mm). A mixture of guaiacol (3 wt %) and
DMDS (1 wt % S) in toluene, containing hexadecane
(1 wt %) as internal reference, was used as the feed. The
catalysts were sulfidized with DMDS in toluene under
the above-described conditions. Catalytic tests were
performed under the following conditions: temperature
260–340°С, hydrogen pressure 3.0 MPa, hydrogen/feed
ratio 500 Ln.c. L–1, FSV 80 h–1. The installation consisted
of units for setting, maintaining, and monitoring the
temperature, pressure, flow rate of the hydrogen-
containing gas, and feed flow rate. The temperature,
pressure, feed flow rate, and hydrogen flow rate were
9 8
In addition, we determined the effective content of
particles on the catalyst surface. For example, for the
CoMoS phase the effective content was calculated using
the formula
cCoMoS = [CoMoS]c(CoT)
(3)
where c(CoT) is the effective content of cobalt atoms on
the catalyst surface, determined by XPS (wt %).
The catalysts were examined by high-resolution
transmission electron microscopy (HR TEM) with
a Tecnai G2 20 device with a LaB6 cathode at an
accelerating voltage of 200 kV. Catalyst samples were
applied onto a copper grid coated with a carbon film.
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 91 No. 2 2018

EFFECT OF SUPPORT OF СоМоS CATALYSTS
Scheme. Guaiacol HDO pathways.
273
2-Methylcatechol
o-Cresol
p-Cresol
Guaiacol
Catechol
Phenol
Cyclohexanol
Cyclohexene
Cyclohexane
Benzene
maintained with accuracies of ±1°C, ±0.05 MPa,
±0.2 mL h–1, and ±0.2 L h–1, respectively. Liquid
catalyzates were taken off at 0.5–1.0-h intervals and
analyzed by GLC with a Kristall-5000 chromatograph
(quartz capillary column with grafted OV-101 phase).
The products were identified from the retention times
using commercially available samples and by gas
chromatography–mass spectrometry with a Shimadzu
GCMS-QP2010 Ultra device. To evaluate the degree
of the catalyst deactivation, each experiment was
performed for 12–15 h. All the samples preserved the
activity and selectivity after 10-h continuous tests.
The guaiacol conversion xGua was calculated by the
formula
(6)
where c0Gua is the guaiacol concentration in the model
mixture (wt %), and сGua is that in the catalyzate (wt %).
The degree of oxygen removal, xHDO, was evaluated as
conversion of guaiacol and oxygen-containing HDO
products (see scheme) [14, 24, 26]:
(7)
where n0Gua and nGua are the guaiacol amounts in the feed
and catalyzate (mol); nPhOH, nCre, nCyc, nCat, and nMeCat
are the amounts of phenol, cresols, cyclohexanol +
cyclohexanone, catechol, and methylcatechol in the
catalyzate, respectively (mol).
The rate constant of guaiacol HDO was calculated
from the pseudo-first-order equation
(8)
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 91 No. 2 2018

274
SAL’NIKOV et al.
Table 2. Composition and physicochemical characteristics of the synthesized CoMoS/Sup catalysts in the activated (sulfide)
state
Distribution of
molybdenum particles,
%
Geometric characteristics
of active phase particles
Textural characteristics
Catalyst
support
(Sup)
сCoMoS
wt %
,
(Сo/Mo)edge
mean number
mean
of MoS2
length,
layers in
nm
SBET
,
Vp,
d, Å
MoS2
64
MoSxOy Mo6+
m2 g–1 cm3 g–1
stack
Al2O3
175
0.25
38
2.4
1.4
20
16
1.4
0.51
and 96
SiO2
TiO2
ZrO2
262
84
0.48
0.21
0.14
80
38
38
3.0
4.1
3.3
1.7
1.6
1.7
69
80
80
16
10
8
15
10
12
0.7
3.8
4.8
0.61
1.08
1.04
64
where F is the guaiacol feeding rate (g s–1); W, catalyst
weight (g); and xGya(HDO), guaiacol conversion and
degree of hydrodeoxygenation (%).
prepared from heteropoly acids have the predominant
size of 2–4 nm [14, 27–30].
The supports used differed in the textural
characteristics and concentration of acid sites. Silicon
oxide was characterized by the maximal values of the
specific surface area (360 m2 g–1) and pore volume
(0.99 cm3 g–1) but low acidity (53 μmol NH3/g), whereas
ZrO2 had the lowest specific surface area (113 m2 g–1)
and high acidity (340 μmol NH3/g). The mean pore
diameter for all the supports was close to 80–100 Å, and
only for titanium oxide the pore size distribution was
bimodal with peaks at 60 and 108 Å (Table 1).
The degree of catalyst deactivation in guaiacol
hydrodeoxygenation, DdGua, and oxygen removal,
DdHDO, was calculated using the formula [14, 24]
(9)
where DdGua(HDO) is the degree of catalyst deactivation in
HDO of guaiacol and oxygen removal after continuous
operation for 20 h in the accelerated deactivation mode at
The physicochemical characteristics of the
synthesized CoMoS/Sup catalysts in the sulfide state
and the geometric characteristics of the formed active
phase particles are given in Table 2. The synthesized
catalysts differed in the textural properties and in the
content and characteristics of active phase particles. The
specific surface area and pore volume of the catalysts
on different supports decreased in the order SiO2 >
Al2O3 > TiO2 ≈ ZrO2, which is due to properties of the
initial supports (Table 1). The pore size distribution
significantly changes after application of the active
phase. The catalyst sulfidization is accompanied by the
formation of coke from citric acid and feed molecules.
As a result, small mesopores of 38 Å size appear. The
CoMoS/Al2O3 catalyst had a biporous structure with the
mean pore diameter of 38 and 76 Å. The SiO2-based
catalyst had no secondary coke mesopores; after the
impregnation, the pore volume decreased, and the mean
pore diameter decreased to 80 Å. Finally, the catalysts on
340°С and increased guaiacol content (%); xinit
,
Gua(HDO)
initial conversion of guaiacol and degree of oxygen
removal at 260°С (%); and xsGua(HDO), steady-state
conversion of guaiacol and degree of oxygen removal
at 260°С after operation in the accelerated deactivation
mode (%).
RESULTS AND DISCUSSION
According to the diffraction patterns of the supports
used, Al2O3 occurred as γ-phase; TiO2, in the form of
anatase; and ZrO2, in the form of monoclinic (major)
and tetragonal phases. XRD patterns of the all the
catalyst samples contained no reflections corresponding
to bulk transition metal sulfides or to any oxysulfide
particles, which is due to the ultradispersed state of the
active component, as the sulfide active phase crystallites
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 91 No. 2 2018

EFFECT OF SUPPORT OF СоМоS CATALYSTS
(b) (c)
275
(a)
(d)
10 nm
10 nm
10 nm
10 nm
Fig. 1. TEM images of samples of CoMoS/Sup catalysts. Sup: (a) Al2O3, (b) SiO2, (c) TiO2, and (d) ZrO2.
X, %
X, %
(a)
(b)
d, nm
Fig. 2. Distribution Х of sulfide particles with respect to (a) size d and (b) number of MoS2 layers in the stack N in CoMоS/Sup catalysts
(Sup: Al2O3, SiO2, TiO2, ZrO2).
surface areas. In addition, the degree of the Mo sulfidi-
zation on the catalyst surface and the promoter content
in the CoMoS active phase, determined by deconvolu-
tion of the XPS spectra similarly to the methods previ-
ously published in [13, 14], appreciably differ.
TiO2 and ZrO2 supports mainly contained pores 38 Å in
diameter, suggesting high extent of their carbonization,
probably due to relatively low specific surface area of
the initial supports and high surface concentration of the
metals. Nevertheless, the pore volume of the CoMoS/
TiO2 catalyst was close to that of the sample supported
on Al2O3.
Higher content of sulfidized МоS2 particles in
CoMoS/ZrO2 and CoMoS/TiO2 samples (Table 2) is
caused by weaker oxide–oxide interaction with the
support and by possible partial chemical transformation
of the support itself. For example, it was noted previously
that TiO2 itself can undergo sulfidization in the course
of activation of hydrotreating catalysts and can exert a
specificorientingeffectoncrystallitesoftheactivephase,
fixed on the support via edges [15, 31, 32]. On the other
hand, the CoMoS/Al2O3 catalyst was characterized by
the lowest degree of the metal sulfidization (64% of Мо
in the form of MoS2), and CoMoS/SiO2, by the presence
of the promoter in the active CoMoS phase. The surface
concentration of cobalt in the active CoMoS phase and
the specific surface area of all the synthesized catalysts
vary in opposite directions.
Black thread-like bands in the TEM images (Fig. 1)
correspond to layers of MoS2 crystallites. The interplanar
spacing in them is about 0.65 nm, which is characteristic
of the (002) basal plane of crystalline molybdenum
disulfide MoS2. The size distribution of sulfide particles
is shown in Figs. 2a and 2b. The mean length of particles
of the CoMoS active phase in the catalysts varied from
2.4 to 4.1 nm, and the mean number of layers in the
(Co)MoS2 crystallite was within 1.4–1.7 (Table 2). The
mean length of the active phase particles depending on
the kind of support decreased in the order TiO2 > SiO2 ≈
ZrO2 > Al2O3, and the mean number of MoS2 layers in
the stack, in the order ZrO2 ≈ SiO2 > TiO2 > Al2O3.
Such changes are associated with different strength
of the interaction of the active phase with the chemically
different supports, and also with their different specific
Catalytic tests were performed under the conditions
typical of hydrotreating of petroleum hydrocarbon raw
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 91 No. 2 2018

276
SAL’NIKOV et al.
Table 3. Conversion of guaiacol and oxygen-containing compounds (HDO) on oxide supports
260°С
340°С
Support
xGua
xHDO
SPhe/Cat
xGua
xHDO
SPhe/Cat
SiCa
0
0
–
0.1
2.8
3.5
0
0
25
34
16
13
0
–
Al2O3
7.3
5.6
1.7
0
0
0
0.7
0.8
1.2
1.9
SiO2
0
8.3
1.5
6.9
TiO2
ZrO2
0.5
0
a Silicon carbide (inert diluent).
materials [15]: 260–340°С, 3.0 MPa, Н2/feed ratio
500 Ln.c. L–1, FSV 80 h–1. Prior to testing the catalysts,
we studied the guaiacol transformations under the same
conditions on the inert diluent (SiC) and on the oxide
supports (Table 3). Silicon carbide was fully inert, and
guaiacol underwent no thermal transformations under
these conditions. The catalyst supports exhibited certain
catalytic properties, especially at 340°С, in shifts of
the methyl and methoxy groups (see scheme), which is
associated with the acid–base properties of the supports
and with the presence of functional ОН groups. At
260°С, only TiO2 appeared to be active in removal of
oxygen-containing compounds (HDO) (0.5%).
less active. The selectivity SPhe/Cat, characterizing
the ability of the catalyst to effect specifically the
hydrogenation, is maximal for the CoMoS/Al2O3 and
CoMoS/ZrO2 catalysts. The major products of guaiacol
HDO were phenol and catechol formed by guaiacol
demethoxylation and demethylation (see scheme);
o-cresol, cyclohexanol, and methylcatechol were present
in small amounts (up to 5 wt % each), and p-cresol was
present in trace amounts. The yield of the target reaction
products (benzene, cyclohexene, cyclohexane) was up
to 5 wt % each.
In the course of the tests, we performed accelerated
catalyst deactivation and evaluated the stability of
the catalyst operation under the HDO conditions.
In the course of accelerated deactivation in HDO
of the feed containing 6 wt % guaiacol at 340°С, the
catalysts on Al2O3 and SiO2 supports showed the
highest activity both in overall guaiacol conversion
and in hydrodeoxygenation. However, after returning
to the initial test conditions (260°С, HDO, 3 wt %
guaiacol), the catalyst on SiO2 support showed the
lowest activity. High conversion of guaiacol on the
CoMoS/SiO2 catalyst may be due to the activity of the
support itself (Table 3). Thus, the catalysts supported on
ZrO2 and AL2O3 exhibited the highest stability under
hydrodeoxygenation conditions (degree of deactivation
about 37 and 46%, respectively, compared to 72% for
CoMoS/SiO2).
The results of testing the CoMoS/Sup catalysts are
given in Table 4. The steady-state guaiacol conversion
after 10-h tests at 260°С was 29–36%, and the degree
of oxygen removal was 13–19%. The catalysts on the
SiO2 and ZrO2 supports showed the highest activity,
and the samples supported on Al2O3 and TiO2 were
The revealed trends in hydrocatalytic transformations
are associated with the properties of the synthesized
catalysts(Figs. 3, 4).Therateconstantofoxygenremoval
in guaiacol HDO increases with a decrease in the mean
length of the active phase particles, irrespective of the
kind of the oxide support (Fig. 3). However, the catalyst
L, nm
Fig. 3. Rate constant of oxygen removal kHDO in guaiacol
HDO at 300°С as a function of the length L of the active
CoMoS phase particles of CoMoS/Sup catalysts. The kind of
the support is indicated at points; the same for Fig. 5.
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 91 No. 2 2018

EFFECT OF SUPPORT OF СоМоS CATALYSTS
277
Table 4. Convenrsion of guaiacol and oxygen-containing compounds on CoMoS/Sup catalists at 260 and 340°С and degree of
their deactivation
Conversion, %, and selectivity at temperature, °С
Degree of
Catalyst support
(Sup)
deactivation, %
260
340
260 afer ADa
xGua
xHDO
SPhe/Cat
xGua
xHDO
SPhe/Cat
xGua
xHDO
SPhe/Cat
DdGua
DdHDO
Al2O3
SiO2
32
36
29
36
13
19
13
17
8.0
1.2
1.6
8.0
86
91
73
75
52
53
30
49
35
20
9
17
10
15
23
4.0
2.3
3.2
5.0
3.6
0.8
3.2
1.0
46
72
50
37
69
88
75
71
TiO2
ZrO2
14
a Accelerated deactivation in guaiacol HDO (6 wt %) at 340°С for 20 h and return to a temperature of 260°С.
TOFHDO × 104, s–1
activity in HDO did not correlate with the Co content of
the CoMoS phase particles. For example, for the TiO2-
supported sample at approximately 4 wt % Co content
of the CoMoS particles (Table 2), the HDO rate constant
was two times lower compared to the other supported
catalysts studied (Fig. 3).
We evaluated the catalytic activity as the turnover
frequency normalized on the number of active edge sites
and plotted the 3D dependence of the turnover frequen-
cy (TOF) in the oxygen removal on the degree of pro-
motion of the crystallite edges (Co/Mo)edge and on the
mean length of the active phase of the catalysts (Fig. 4).
As can be seen, the turnover frequency for the CoMo
sites increases with a decrease in (Co/Mo)edge and in the
length of the active phase particles. The active sites of
the catalyst supported on SiO2 show the highest HDO
activity. Depending on the kind of the support, the
turnover frequency for the CoMo sites decreases in the
order SiO2 > Al2O3 > ZrO2 ≈ TiO2.
L, nm
Fig. 4. 3D plot of the turnover frequency TOFHDO in guaiacol
HDO on edge sites of active CoMoS phase particles vs. mean
particle length L and degree of promotion (Co/Mo)edge
.
It is more difficult to correlate the degree of deactiva-
tion of the catalytic systems with their physicochemical
characteristics. We revealed no direct relationships be-
tween the stability of the catalysts and the acidity of sup-
ports, which were observed previously in experiments
with zeolite-containing and carbonized supports [24].
For the SiO2-supported catalysts containing virtually no
acid sites (53 μmol NH3/g), the degree of deactivation
exceeds that of the catalysts on the other supports stud-
ied (Al2O3, TiO2, and ZrO2), exhibiting higher acidity
(>200 μmol NH3/g). On the other hand, the isoelectric
point for the supports used [33] varies in the order SiO2
(1.7–3.5) < TiO2 (3.5–7.0) < ZrO2 (5.5–7.0) < Al2O3
Fig. 5. Degree of deactivation DdHDO of CoMoS/Sup catalysts
in guaiacol HDO as a function of the cobalt content cCoMoS in
particles of the CoMoS phase.
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 91 No. 2 2018

278
SAL’NIKOV et al.
(6.2–9.5), which, on the whole, correlates with the vari-
ation of the degree of the catalyst deactivation in guaia-
col HDO (Table 4). Thus, deactivation of supported
sulfide catalysts in guaiacol HDO cannot be unambigu-
ously attributed to the support acidity. The differences
may also be due to acid properties of the active sites of
the sulfide active phase nanoparticles, on which coke is
known to be formed also [19, 20, 34]. On the other hand,
CoMoS/SiO2 contained the smallest amount of the pro-
moted CoMoS sites, capable of hydrogenation of the ad-
sorbed coke precursors, which may also be responsible
for the strongest degree of deactivation of this catalyst
compared to the other catalysts. Indeed, the linear de-
pendence of the degree of the catalyst deactivation in
guaiacol HDO on the content of the promoted CoMoS
sites (Fig. 5) shows that the role of the cobalt promoter
in such catalytic processes consists in stabilization of
the sulfide active phase, which is an important factor for
the subsequent commercial use.
(project no. 14.574.21.0139, unique project identifier
RFMEFI57417X0139).
REFERENCES
1. Helwani, Z., Othman, M.R., Aziz, N., et al., Fuel Process.
Technol., 2009, vol. 90, pp. 1502–1514.
2. Lappas, A.A., Bezergianni, S., and Vasalos, I.A., Catal.
Today, 2009, vol. 145, nos. 1–2, pp. 55–62.
3. Demirbas, A., Appl. Energy, 2011, vol. 88, pp. 17–28.
4. Flach, B., Lieberz, S., Rondon, M., et al., EU Biofuels
Annual 2016, NL6021, https://gain.fas.usda.gov.
5. Bui, V.N., Toussaint, G., Laurenti, D., et al., Catal. Today,
2009, vol. 143, pp. 172–178.
6. Fogassy, G., Thegarid, N., Toussaint, G., et al., Appl.
Catal. B: Environmental, 2010, vol. 96, pp. 476–485.
7. Choudhary, T.V. and Phillips, C.B., Appl. Catal. A:
General, 2011, vol. 397, pp. 1–12.
8. Furimsky, E. and Massoth, F.E., Catal. Today, 1999,
vol. 52, pp. 381–495.
CONCLUSIONS
9. Furimsky, E., Appl. Catal. A: General, 2000, vol. 199,
pp. 147–190.
The activity of CoMoS/Sup catalysts in hydrodeox-
ygenation of guaiacol strongly depends on the kind of
the support and decreases in the order SiO2 > Al2O3 >
ZrO2 ≈ TiO2. The hydrodeoxygenation rate constant in-
creases with a decrease in the mean length of the ac-
tive CoMoS phase particles, irrespective the kind of the
support. The number of turnovers for the CoMo sites
of the synthesized catalysts increases with a decrease in
the degree of promotion of (Co/Mo)edge crystallite edges
and in the length of the active phase particles.
10. Holmgren, J., Marinageli, R., Nair, P., et al., Hydrocarbon
Process., 2008, pp. 95–103.
11. Mortensen, P.M., Grunwaldt, J.D., Jensen, P.A., et al.,
Appl. Catal. A, 2011, vol. 407, pp. 1–19.
12. Popov, A., Kondratieva, E., Mariey, L., et al., J. Catal.,
2013, vol. 297, pp. 176–186.
13. Varakin, A.N., Salnikov, V.A., Nikulshina, M.S., et al.,
Catal. Today, 2017, vol. 292, pp. 110–120.
14. Nikulshin, P.A., Salnikov, V.A., Varakin, A.N., et al.,
Catal. Today, 2016, vol. 271, pp. 45–55.
The catalysts supported on ZrO2 and Al2O3 showed
the highest stability in the course of accelerated deacti-
vation in guaiacol hydrodeoxygenation. Deactivation of
supported sulfide catalysts in guaiacol hydrodeoxygen-
ation does not correlate with the acidity of the supports
used and depends also on the content of the active Co-
MoS phase particles. The possible ways to improve the
hydrodeoxygenation catalysts based on transition metal
sulfides can involve a decrease in the length of the active
phase particles, an increase in the content of promoted
CoMoS active sites, and optimization of the isoelectric
point of the support.
15. Kogan, V.M., Nikulshin, P.A., Dorokhov, V.S., et al.,
Russ. Chem. Bull., 2014, vol. 63, no. 2, pp. 332–345.
16. Tran, N.T.T., Uemura, Y., and Ramli, A., Procedia Eng.,
2016, vol. 148, pp. 1252–1258.
17. Badawi, M., Paul, J.F., Cristol, S., et al., Catal. Commun.,
2011, vol. 12, no. 10, pp. 901–905.
18. Pinheiro, A., Hudebine, D., Dupassieux, N., et al., Energy
Fuels, 2009, vol. 23, pp. 1007–1014.
19. Centeno, A., Laurent, E., and Delmon, B., J. Catal., 1995,
vol. 154, pp. 288–298.
20. Bui, V.N., Laurenti, D., Afanasiev, P., et al., Appl. Catal.
B: Environmental, 2011, vol. 101, pp. 239–245.
21. Bui, V.N., Laurenti, D., Delichere, P., et al., Appl. Catal.
B: Environmental, 2011, vol. 101, pp. 246–255.
ACKNOWLEDGMENTS
22. Yunquan, Y., Hean, L., Gangsheng, T., et al., Chin. J.
Chem. Eng., 2008, vol. 16, no. 5, pp. 733–739.
The study was financially supported by the Ministry
of Education and Science of the Russian Federation
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 91 No. 2 2018

EFFECT OF SUPPORT OF СоМоS CATALYSTS
279
23. Bykova, M.V., Bulavchenko, O.A., Ermakov, D.Yu.,
29. Nikulshin, P.A., Tomina, N.N., Pimerzin, A.A., et al.,
et al., Katal. Prom–sti., 2010, no. 5, pp. 45–52.
Appl. Catal. A: General, 2011, vol. 393, pp. 146–152.
24. Nikul’shin, P.A., Sal’nikov, V.A., Zhilkina, E.O., et al.,
30. Nikul’shin, P.A., Mozhaev, A.V., Ishutenko, D.I., et al.,
Katal. Prom–sti., 2014, no. 4, pp. 63–73.
Kinet. Catal., 2012, vol. 53, no. 5, pp. 620–631.
25. Kasztelan, S., Toulhoat, H., Grimblot, J., and Bon-
31. Breysse, M., Geantet, C., Afanasiev, P., et al., Catal.
Today, 2008, vol. 130, pp. 3–13.
nelle, J.P., Appl. Catal., 1984, vol. 13, pp. 127–159.
26. Olcese, R.N., Bettahar, M., Petitjean, D., et al., Appl.
Catal. B: Environmental, 2012, vols. 115–116, pp. 63–73.
32. Ramirez, J., Macias, G., Cedeno, L., et al., Catal. Today,
2004, vol. 98, pp. 19–30.
27. Ferrari, M., Delmon, B., and Grange, P., Micropor.
Mesopor. Mater., 2002, vol. 56, pp. 279–290.
33. Kosmulski, M., J. Colloid Interface Sci., 2009, vol. 337,
pp. 439–448.
34. Saidi, M., Samimi, F., Karimipourfard, D., et al., Energy
Environ. Sci., 2014, vol. 7, p. 103.
28. Nikulshin, P.A., Tomina, N.N., Pimerzin, A.A., et al.,
Catal. Today, 2010, vol. 149, pp. 82–90.
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 91 No. 2 2018