Sol–Gel Synthesis of Membrane Mo2C/Al2O3 Catalysts with Different Architectures and Their Catalytic Activity in the Reaction of Carbon Dioxide Conversion of Methane

Article abstract of DOI:10.1134/S002315841805004X

The methods of synthesis of composite membrane catalysts based on Mо₂C and Al₂O₃ support by a sol–gel method were developed. The samples of membrane catalysts with different architectures were developed. The difference in

Full text of DOI:10.1134/S002315841805004X

ISSN 0023-1584, Kinetics and Catalysis, 2018, Vol. 59, No. 5, pp. 635–643. © Pleiades Publishing, Ltd., 2018.
Original Russian Text © N.N. Gavrilova, M.A. Myachina, D.V. Ardashev, V.V. Nazarov, V.V. Skudin, 2018, published in Kinetika i Kataliz, 2018, Vol. 59, No. 5, pp. 612–621.
Sol–Gel Synthesis of Membrane Мo2С/Al2O3 Catalysts
with Different Architectures and Their Catalytic Activity
in the Reaction of Carbon Dioxide Conversion of Methane
N. N. Gavrilovaa, M. A. Myachinaa, *, D. V. Ardasheva, V. V. Nazarova, and V. V. Skudina
aMendeleev University of Chemical Technology of Russia, Moscow, 125047 Russia
*e-mail: myachinamary@gmail.com
Received March 27, 2018
Abstract—The methods of synthesis of composite membrane catalysts based on Мо2С and Al2O3 support by
a sol–gel method were developed. The samples of membrane catalysts with different architectures were
developed. The difference in the catalytic activity of membrane catalysts in carbon dioxide conversion of
methane depending on the porous structure and morphology of the catalytic layer was studied.
Keywords: membrane catalysis, membrane catalyst, molybdenum carbide, sol–gel method, carbon dioxide
conversion of methane
DOI: 10.1134/S002315841805004X
INTRODUCTION
tures should also be noted. Since many industrial pro-
cesses are carried out at high temperatures, it is obvi-
ous that this is not the most economically viable solu-
tion in developing membrane reactors.
Membrane catalytic reactors with membrane cata-
lysts are promising in performing high-temperature
chemical processes. It is known that during operation
of such reactors, the selectivity and yield of products
sharply increase compared with a conventional reactor
with a catalyst of the same composition [1–5].
Together with highly porous honeycomb and mono-
lith catalysts, membrane catalysts (catalytic membranes)
belong to a class of structured heterogeneous catalysts.
They require the use of support of a certain form. Mem-
brane catalysts allow simultaneous use of various driving
forces for forced transport of starting materials through
the porous catalyst structure.
Having said that, other membrane reactor modes
remain poorly studied: these are provided by the con-
tactor and distributor reactors. A distinctive feature of
the above modes is that the membrane catalyst in this
case does not have to possess separation characteris-
tics: it is sufficient to have a support on which a cata-
lytically active component to which the active compo-
nent is distributed, and this substantially simplifies its
synthesis and use.
Despite the absence of selective removal of a reac-
tion product, the use of porous membrane catalysts
with various compositions makes it possible to achieve
a several-fold increase in the productivity of a mem-
brane reactor compared to a conventional one. For
instance, such results were obtained by studying car-
bon dioxide conversion of methane [9, 10]. It has also
been shown that the different modes of operation of a
membrane reactor make it possible to carry out chem-
ical reactions at much lower temperatures than over
conventional heterogeneous catalysts [11, 12].
Analysis of publications of the recent decades
shows that the vast majority of studies in the field of
membrane catalysis has been devoted to one of the
possible modes of operation provided by a so-called
extractor membrane reactor [6–8]. Such a membrane
reactor is especially effective when carrying out revers-
ible chemical reactions under conditions close to equi-
librium when the selective removal of one of the com-
ponents (e.g., hydrogen) leads to an increase in the
conversion.
A disadvantage of extractor membrane reactor is
the fact that its productivity is limited by diffusive pro-
The increased efficiency of membrane catalysts is
cess of product removal through a membrane. When explained by the forced transport of substances
creating an industrial membrane catalytic reactor through its porous structure [6, 13, 14], which makes
operating in this mode, membranes with a significant it possible to overcome diffusion control. This is espe-
outer surface and porosity will be required to provide cially important for fast reactions, and almost all
acceptable performance. The complexity of creating industrial catalytic processes of conversion of gaseous
selective membranes that are stable at high tempera- hydrocarbons are fast.
635

636
GAVRILOVA et al.
Currently, there is not enough data to formulate the γ-AlOOH (boehmite) layer on the surface of initial
requirements for the porous structure of membrane support with further calcination at various tempera-
catalysts, which can be used as contactors or distribu- tures. For more details on the preparation of support
tors. Therefore, the creation of catalytic membranes for membrane catalysts, see [18].
with different architectures (different porous struc-
tures and surface morphologies) and their kinetic
studies are very topical. The solution to these problems
will make it possible to justify the requirements for the
structure of membrane catalysts and the choice of
their operation modes in a membrane reactor for car-
rying out catalytic reactions.
Carbon dioxide conversion of methane was chosen
as a model reaction because it occurs at high tempera-
tures and has a number of thermodynamic and kinetic
limitations that impede its industrial development.
The thermodynamic limitation is the necessity of a
high temperature, which corresponds to the maximum
yield of the product. However, an increase in tempera-
ture accompanied by noticeable acceleration of side
reactions, one of which leads to the strong carboniza-
tion of catalyst surface and its deactivation. Kinetic
constraints are caused by the low degree of use of the
internal surface of a catalyst. These circumstances
stimulate a search for new catalytic systems and the
development of appropriate process design. One of the
effective solutions to these problems can be the use of
a membrane catalytic reactor.
Molybdenum carbide has recently been considered
a promising catalytic system and thus we used it as an
active component in this work. It is known that Group
VI transition metal carbides are similar in their elec-
tronic structure to Group VIII metals and show cata-
lytic activity in the same reactions as the noble metals
[15]. Group VI metal carbides have a number of
advantages in comparison with platinum metals: lower
cost, sufficient mechanical strength, increased electrical
conductivity, and chemical stability. A distinctive feature
of these compounds is the resistance to coking and many
catalytic poisons [16, 17]. The absence of a noticeable
deposition of carbon on the surface of carbide catalysts
was noted in our previous studies [10–12].
Membrane catalysts were synthesized by the sol–
gel method using molybdenum blue as sol. Molybde-
num blue was prepared according to a technique
developed earlier [19] using ammonium heptamolyb-
date, hydrochloric acid, and glucose as a reducing
agent. Impregnation of support was carried out by
short-time (10 s) immersing of support into molybde-
num blue with further drying and calcination at 900°С
in N2 and a temperature ramp of 5°С/min [20].
Determination of the Main Characteristics
of Membrane Catalysts
The morphology of synthesized samples was stud-
ied using a scanning microscope MIRA with an X-ray
microanalyzer EDX INCA 350 X-Max (Tescan Orsay
Holding, Czech Republic). The distribution of the
active component over support was determined based
on energy dispersive spectroscopy (EDS) analysis of
sample cross sections.
The phase composition of modified support and
membrane catalysts was studied using an X-ray
Rigaku D/MAX 2500 diffractometer (Rigaku, Japan;
CuKα radiation). Phases in the composition of sam-
ples under study were identified according to ICDD
library.
The characteristics of the porous structure of the
samples were calculated from the isotherms of low-
temperature nitrogen adsorption. The studies were
carried out using a Gemini VII analyzer (Micromerit-
ics, USA) in the Mendeleev Center for Shared Use.
The specific surface area (Ssp) was determined by the
BET method. The overall volume of pores (V) was
found from a maximum value of relative pressure of
0.995. The volume of mesopores (Vmeso) and the distri-
bution of mesopore volume over sizes were determined
by the BJH method. The most probable diameter (d) was
diameter corresponding to the maximum on the differ-
ential curve of pore distribution over sizes.
The goal of this work was to synthesize membrane
catalysts Мo2С/x-Al2O3/α-Al2O3 with different archi-
tectures and determine their catalytic activity in car-
bon dioxide conversion of methane.
The average pore diameter of samples ( ) was esti-
d
mated by the equation
4V
S sp
d =
,
EXPERIMENTAL
Synthesis of Membrane Catalysts
where V is the overall pore volume; Ssp is the BET sur-
Tubular ceramic asymmetric microfiltration
corundum (α-Al2O3) membranes were used as initial
support. Their length was 20 cm; the outer diameter
was 10 mm, the wall width was 1 mm. The relative
porosity of support was 40%.
The structured support with the compositions
γ-Al2O3/α-Al2O3, δ-Al2O3/α-Al2O3, and α-Al2O3/
α-Al2O3 were prepared by forming an additional
face area of a sample.
When calculating the specific surface area and pore
volume of the catalytic layer, we assumed that the total
specific surface area and the total pore volume of the
sample were additive with the specific surface area and
pore volume of the support and the catalyst layer.
The samples of membrane catalysts were studied by
the above methods both before and after catalytic
KINETICS AND CATALYSIS
Vol. 59
No. 5
2018

SOL–GEL SYNTHESIS OF MEMBRANE Мo2С/Al2O3 CATALYSTS
637
tests. Comparison of the obtained data on morphol-
RESULTS AND DISCUSSION
Synthesis of Support for Membrane Catalysts
ogy, elemental and phase composition made it possi-
ble to conclude that the catalytic systems under study
are stable and resistant to carbonization.
Supports for membrane catalysts were prepared
according to the techniques described in [18]. The idea
behind the procedure was to accumulate boehmite
suspension on the outer surface of support by filtration
with further drying and heat treatment of samples. The
use of highly dispersed crystalline particles of boeh-
mite makes it possible to obtain on the initial mem-
brane (support) modifying layers with a high specific
surface area and developed porous structure. Changes
in the porous characteristics (to be more specific, in
the predominant size of pores and width of pore distri-
bution over sizes) was maintained by varying the final
temperature of support calcination. The calcination
temperature determined both the characteristics of the
porous structure of samples and the phase composi-
tion of alumina that constituted the additional modi-
fying layer.
The content of alumina in synthesized support was
~3 wt %, which corresponded to the layer thickness of
70 μm. This follows from Fig. 1 on which microscopic
images of a typical support sample prepared as
described above are shown (a γ-Al2O3/α-Al2O3 sam-
ple).
The synthesized support for a membrane catalyst is
a tubular element consisting of several layers: initial
support consisting of coarse-grained α-Al2O3 parti-
cles, a microfiltration layer consisting of small-grained
particles of α-Al2O3, and supported layer, which con-
sisted of aggregates of primary particles with a nearly
spherical shape. The phase composition of these latter
particles depended on the calcination temperature
(Table 1). The morphology of the layer does not
depend significantly on the calcination temperature,
and microscopic images of other samples are thus not
shown.
The catalytic activity of the membrane catalysts
was estimated in carbon dioxide conversion of meth-
ane, which was carried out in a membrane catalytic
reactor operating in the contactor mode. The initial
mixture of CH4 and CO2 (in the equimolar ratio) was
fed from the outer side of a membrane. The flow rate
of the gaseous mixture was varied from 50 to 320
cm3/min (based on normal conditions).
The quantitative analysis of the reaction mixture
was carried out using a Khromatek-Kristall 5000 gas
chromatograph (Khromatek, Russia) equipped with
two thermal conductivity detectors. For more details
on analyses, see [12].
To compare the catalytic activity of synthesized
samples of membrane catalysts, we calculated rate
constants of methane dissociation (k), which was
accepted as a rate-limiting step of carbon dioxide con-
version of methane [21]. The specific rate constant
(kw) per unit weight was calculated by dividing the rate
constant k by the weight of the active component:
k
m
kw =
,
where m is the weight of the active component Мо2С.
The specific rate constant per unit surface area was
determined by dividing the rate constant k by the sur-
face area of the catalyst layer s:
k
mS
ks =
,
(
)
sp
where Ssp is the specific surface area of the catalyst
layer Мо2С/х-Al2O3, where x = γ, δ, or α.
The contact time τ was calculated by the formula:
273Vcat
A distinctive feature of the samples considered here
is the presence of pores of two types in the support
(supported) layer. Pores of the first type are transport
pores and they are necessary for the catalytic reaction
over the membrane catalyst. The size of such pores
should allow the free transport of reagents. Pores of
the second type (smaller) are mainly in the aggregates
of the primary Al2O3 particles. They are likely to pro-
vide the high specific surface area and are necessary
for the subsequent support of the catalytically active
phase.
τ =
,
νT
where Vcat is the volume of membrane catalyst, includ-
ing support and the catalytic layer; ν is the volume
flow rate of the mixture under normal conditions; and
T is temperature in the reaction zone.
The Knudsen number Kn was calculated for meth-
ane molecules at 850°С and an atmospheric pressure
according to the equation [14]
As can be seen from the presented data, the initial
support is a typical macroporous material with a low
specific surface area (0.4 m2/g) (Table 1). An addi-
tional layer leads to a significant increase in the spe-
cific surface area; that is, a developed porous structure
with prevailing mesopores is formed.
Figure 2 shows the differential curves of the meso-
pore volume distribution over sizes for synthesized
supports of membrane catalysts. It follows from data
kBT
λ
d
Kn =
=
,
4 2πr2Pd
where λ is the free path of the molecule; kB is the
Boltzmann constant; T is the gas temperature; is the
average pore diameter in the sample; Р is the pressure
in a reactor; r is the effective radius of methane mole-
cule (at 850°С, the effective radius of methane mole-
cule is 1.5 × 10–10 m [22]).
d
KINETICS AND CATALYSIS Vol. 59 No. 5 2018

638
GAVRILOVA et al.
(a)
(b)
γ-Al2O3
Boundary layer
α-Al2O3
Support
α-Al2O3
γ-Al2O3
100 μm
2 μm
(d)
(c)
20 μm
2 μm
Fig. 1. Microscopic image of cross section (a, b) and the surface of support of the membrane catalyst γ-Al O /α-Al O (c, d).
2
3
2 3
presented in Table 1 and Fig. 2 that the calcination proven to be efficient in the synthesis of catalysts and
temperature strongly affects the porous structure of
the samples. Sample γ-Al2O3/α-Al2O3 has a narrow
distribution of pores over sizes with a most probable
diameter of ~10 nm, sample δ-Al2O3/α-Al2O3 has a
wider distribution of pores with a most probable diam-
eter of ~30 nm. The maximal prevalent size of pores
(44 nm) is characteristic of the α-Al2O3/α-Al2O3 sup-
port sample.
covers, including selective layers in membranes.
In the case of supported materials, the sol–gel pro-
cess is usually based on the use of sols (stable dispersed
nanoparticles). The preparation of a membrane cata-
lyst includes the following typical steps: deposition of
sol on support (a stage that determines the nature of
distribution of the component and the layer thick-
ness), drying (a stage when sol transforms into gel),
and heat treatment (a stage when the active molybde-
num carbide phase is formed).
Synthesis of Membrane Catalysts
Sometimes when synthesizing materials by a sol–
gel method, it is necessary to repeat the deposition–
The samples of structured supports obtained were
used to prepare membrane catalysts. To create a cata-
lyst layer, a sol–gel method was chosen, which has drying–heat treatment cycle, which leads to a signifi-
KINETICS AND CATALYSIS
Vol. 59
No. 5
2018

SOL–GEL SYNTHESIS OF MEMBRANE Мo2С/Al2O3 CATALYSTS
639
Table 1. Characteristics of the porous structure of supports for membrane catalysts
Ssp, m2/g
Vmeso, cm3/g
V, cm3/g
Tcalcination, °С
Samples studied
d, nm
d,nm
α-Al2O3 (initial support)
γ-Al2O3/α-Al2O3
δ-Al2O3/α-Al2O3
1250
900
0.4
0.0010
0.0007
–
–
1.6/35.0
1.0/17.6
0.6/5.9
0.0130/0.35
0.0120/0.32
0.0095/0.25
0.0120/0.33
0.0110/0.30
0.0074/0.20
10
30
44
33
48
63
1000
1100
α-Al2O3/α-Al2O3
Data on the specific surface area and pore volume refer to the whole sample based on the support x-Al O layer (where x = γ, δ, or α).
2
3
A dash means the absence of pores in this region (the region of mesopores) for the sample of the initial support.
cant increase in the time and cost of materials thus sisted of two samples prepared on the same support
prepared.
(γ-Al2O3/α-Al2O3) with different distributions of the
active component (uniform and crusted) in the addi-
tional layer of support, which was achieved by chang-
ing the properties of molybdenum blue dispersion,
specifically the sign of a particle charge. Figure 3
shows the difference in the morphologies of mem-
brane catalysts samples with uniform and crusted
types of distribution.
The nature of catalyst distribution over support was
determined from EDA of the samples based on the
analysis of linear distributions of the main elements
(molybdenum, carbon, and aluminum) over the cross
section of membrane catalysts. Examples of one of
these dependencies (molybdenum distribution) are
given in Figs. 3a, 3d.
In the case of crusted-type distribution, the main
part of active component is on the external surface of
support (see Fig. 3a); that is, the external surface of
support becomes covered with a thin catalytic layer of
Mo2C.
The other type of distribution of the active compo-
nent is uniform distribution over the bulk of support
layer (see Fig. 3d). On the outer surface and along the
We used molybdenum blue for the preparation of
membrane catalysts. This is a dispersed mixture of
oxуgen-containing compounds of variable-valence
molybdenum. The above dispersed systems can be
obtained by reduction of molybdates in the acidic
medium. The reducing agents may vary. Because the
main task was to obtain molybdenum carbide, only
organic substances were considered as reducing
agents, which in further heat treatment served as a
source of carbon [23].
In this work, molybdenum blue was synthesized
using glucose as a reducing agent. Heat treatment of a
xerogel in an inert medium that was obtained by dry-
ing the supported molybdenum blue initially leads to
the formation of molybdenum oxycarbide MoOC,
which then transforms into carbide Мо2С. The tem-
perature of carbide formation was 870°С [24].
Earlier we found [20] that molybdenum blue is
capable of structuring and, at a certain concentration
of the dispersed phase, transforms into a gel. This
make is it possible to deposition without adding addi-
tional viscosity control agent. Thus, it is possible to
significantly reduce the number of stages and catalyst
preparation time.
Because the main goal of this work was to prepare
membrane catalysts with different architectures, it was
important to find conditions that would allow to con-
trol the depth of penetration of molybdenum blue par-
ticles into the support and the thickness of supported
layer. This is achieved by varying the colloidal chemi-
cal properties of the molybdenum blue dispersions
(pH of the dispersion medium and the concentration
of the dispersed phase) and conditions of their deposi-
tion (the contact time).
0.00020
10 nm
0.00015
2
0.00010
30 nm 44 nm
0.00005
4
3
Based on the experiments carried out in this work
on supporting of dispersions under various conditions,
we found that for the formation of the catalytic layer
only on the additional layer of support, the contact
time of support with dispersion in the course of depo-
sition should not be greater than 10 s at a concentra-
tion of the dispersed phase of 2 wt %.
1
0
50
100 150 200 250 300
Pore diameter, nm
Fig. 2. Differential curves of mesopore size distribution for
supports of membrane catalysts: the initial support
α-Al O (1), γ-Al O /α-Al O (2), δ-Al O /α-Al O (3),
2
3
2
3
2
3
2
3
2 3
The above method was used to prepare several
samples of membrane catalysts. The first group con-
and α-Al O /α-Al O (4).
2
3
2 3
KINETICS AND CATALYSIS Vol. 59 No. 5 2018

640
(a)
GAVRILOVA et al.
(b)
(c)
Mo
30 μm
30 μm
10 μm
(f)
(e)
(d)
Mo
30 μm
10 μm
10 μm
Fig. 3. Microscopic image of cross section (a, b, d, e) and the external surface (c, f) of membrane catalysts with crusted (a, b, c)
and uniform (d, e, f) types of distribution.
cross section of the sample, spherical aggregates are
clearly seen, while the volume between particles is not
filled, which should not have a bad effect on the per-
meability of these samples.
The sample of the second group differed in the
porous structure of support (Table 1), and the amount
of supported active component (1% Mo2C) and the
nature of its distribution (uniform in the bulk of sup-
port) were the same.
Table 2 shows the characteristics of the porous
structure of synthesized catalyst samples. As can be
seen from the data presented in Tables 1 and 2, the
deposition of Мо2С leads to an insignificant decrease
in both the total specific surface area of membrane
catalyst samples and the area recalculated per catalyst
layer, which is probably due to the partial filling of the
pores of the support layer with a catalyst. The total vol-
ume of pores and their size also change, which is
mainly due to a decrease in the contribution of meso-
pores.
Catalytic Properties of Membrane Catalysts
The synthesized samples of membrane catalysts
with different architectures were tested in carbon diox-
ide conversion of methane. It was found that mem-
brane catalysts Mo2C/γ-Al2O3/α-Al2O3 show catalytic
activity at ≥800°С. With increasing temperature, the
conversion and yield of products increased. The reac-
tion products are hydrogen (Н2) and carbon monoxide
(CO). Other reaction products, including C2+ hydro-
carbons, were not detected by chromatographic anal-
yses.
Figure 4 shows an example of dependence of meth-
ane conversion on the contact time at 850°С.
As can be seen from Fig. 4a, for the samples of the
second series with different structures of porous sup-
port, the conversion increased with an increase in the
specific surface area and a decrease in the pore size in
the samples. Sample Mo2C/γ-Al2O3/α-Al2O3, which
is characterized by the narrowest distribution over
Table 2. The characteristics of the porous structure of membrane catalysts Mo2C/х-Al2O3/α-Al2O3
Catalyst sample and the type
Ssp, m2/g
Vmeso, cm3/g
V, cm3/g
d, nm
d,nm
of distribution
Mo2C/γ-Al2O3/α-Al2O3, crusted
Mo2C/γ-Al2O3/α-Al2O3, uniform
Mo2C/δ-Al2O3/α-Al2O3, uniform
Mo2C/α-Al2O3/α-Al2O3, uniform
1.0/14.0
1.3/22.5
1.1/17.0
0.6/4.8
0.005/0.090
0.005/0.100
0.0043/0.083
0.0028/0.045
0.004/0.080
0.0031/0.060
0.0039/0.080
0.0010/0.008
3.7
3.0
4.0
5.5
20
15
16
19
Data on the specific surface area and pore volume refer to the whole sample based on the catalytic layer Mo C/х-Al O (where x = γ, δ, or α).
2
2
3
KINETICS AND CATALYSIS
Vol. 59
No. 5
2018

SOL–GEL SYNTHESIS OF MEMBRANE Мo2С/Al2O3 CATALYSTS
641
(а)
(b)
80
60
40
20
80
60
40
20
Mo2C/γ-Al2O3/α-Al2O3
Mo2C/δ-Al2O3/α-Al2O3
Mo2C/α-Al2O3/α-Al2O3
Mo2C/γ-Al2O3/α-Al2O3—crusted
Mo2C/γ-Al2O3/α-Al2O3—uniform
0
0.2
0.4
τ, s
0.6
0.8
0
0.2
0.4
τ, s
0.6
0.8
Fig. 4. Dependences of methane conversion on the contact time at 850°С on catalysts with different architectures: Mo C/δ-
2
Al O /α-Al O and Mo C/α-Al O /α-Al O with uniform distribution of the catalyst (a), Mo C/γ-Al O /α-Al O with differ-
2
3
2
3
2
2
3
2
3
2
2
3
2 3
ent types of catalyst distribution (crusted and uniform) (b).
pore sizes and the smallest prevailing diameter,
As can be seen from Table 3, in the series of samples
showed the maximum catalytic activity. At a reaction Mo2C/х-Al2O3/α-Al2O3, with a change in the compo-
temperature of 850°С and a contact time of 0.84 s, the
conversion increased from 8 to 60% for the samples of
this series.
sition of support layer from α-Al2O3 to γ-Al2O3, an
increase in the rate constant is observed. Such a differ-
ence could be attributed to the difference in the specific
surface area of support, but if the activity is recalculated
per unit surface area of the catalyst (ks), the above feature
is preserved. The Mo2C/γ-Al2O3/α-Al2O3 samples show
the maximum catalytic activity. Of these, the sample
with uniform distribution of the catalyst is the most
active.
Figure 4b shows dependence of methane conver-
sion on the contact time over samples with different
distributions of the active component on support
γ-Al2O3/α-Al2O3. The activity of a sample with
crusted-type distribution was lower compared with a
uniformly distributed catalyst. At a contact time of 0.84 s,
the difference in methane conversion reached 30%.
However, the results for a sample with crusted-type cat-
alyst distribution are better than those for samples based
on δ-Al2O3/α-Al2O3 and α-Al2O3/α-Al2O3 (Fig. 4a).
When considering the series of supports for mem-
brane catalysts, it should be noted that γ-Al2O3 exhib-
its its own catalytic activity in many reactions, and
It should be noted that contact time at which the
dependences were obtained corresponded to the time of
passage of the reaction mixture through the entire mem-
brane catalyst sample. If only the volume of the catalytic
layer (Mo2C/х-Al2O3) with a thickness of ~70 μm is
taken into account in calculation, the values of the real
contact time decrease by 4 orders of magnitude.
To compare the activity of membrane catalysts with
different architectures, the rate constants of methane
dissociation were calculated. According to our previ-
ous studies [20, 21], activity of membrane catalysts
with different architectures, the rate constants of
methane is rate-limiting in carbon dioxide conversion
of methane. Since we have a first-order reaction, the
experimentally obtained dependences of methane
concentration in the reaction products on the contact
time were processed in semilogarithmic coordinates
(Fig. 5), which allowed us to estimate the rate constant
of methane dissociation.
4.0
1
2
1. y = –0.27x + 3.90
R2 = 0.97
3.5
3.0
2.5
2. y = –0.31x + 3.90
R2 = 0.95
4
3. y = –1.65x + 3.90
R2 = 0.88
4. y = –0.98x + 3.90
R2 = 0.99
3
0
0.5
1.0
1.5
τ, s
Fig. 5. Dependence of methane concentration in the reac-
tion products on the contact time at 850°С on the samples
of membrane catalyst: (1) Mo C/α-Al O /α-Al O ; (2)
2
2
3
2 3
Table 3 shows the values of rate constants and the
values of the specific catalytic activity described by
specific constants.
Mo C/δ-Al O /α-Al O ; (3) Mo C/γ-Al O /α-Al O
2
2
3
2
3
2
2 3
2
3
2 3
(uniform); (4) Mo C/γ-Al O /α-Al O (crusted). R is the
2
2
3
approximation validity coefficient.
KINETICS AND CATALYSIS Vol. 59 No. 5 2018

642
GAVRILOVA et al.
Table 3. The values of specific rate constants of methane dissociation at 850°С for the samples of membrane catalysts
Mo2C/х-Al2O3/α-Al2O3 with different architectures
Catalyst sample and the type
of distribution
Kn
of support
Kn of membrane
kw, s–1 g–1 ks, s–1 m–2
k, s–1
H2/СО
catalyst
Mo2C/γ-Al2O3/α-Al2O3, crusted
Mo2C/γ-Al2O3/α-Al2O3, uniform
Mo2C/δ-Al2O3/α-Al2O3, uniform
Mo2C/α-Al2O3/α-Al2O3, uniform
0.98
1.65
0.31
0.27
24.5
41.3
7.8
0.44
0.46
0.11
0.28
0.4
0.9
0.4
0.3
13
13
8
20
27
25
21
6.7
6
under conditions of carbon dioxide conversion, it sig-
When the catalyst is supported, the pore sizes
nificantly accelerates methane dissociation [25]. This decrease, and the average diameter also decreases (see
makes it possible to explain the difference in the activ- values in Tables 1 and 2). When the pore size
ity of catalysts of the same composition Mo2C/γ- decreases, the Knudsen number increases, which, in
turn, should lead to an increase in the fraction of mol-
ecules colliding with the pore walls and to more com-
plete utilization of the inner catalytically active sur-
face.
Al2O3/α-Al2O3, but with different types of active com-
ponent distribution. As can be seen from Table 3, the
ratios of the reaction products H2/СО for the two
samples are substantially different. A lower value of
H2/СО for the crusted distribution is probably due to
the presence of the γ-Al2O3 surface that is not covered
by the catalyst and which is responsible for the side
reaction.
CONCLUSIONS
The developed method of synthesis of membrane
catalysts based on Мо2С and modified supports of
Al2O3 makes it possible to obtain samples with differ-
ent architectures and morphology of a catalytic layer.
However, the own catalytic activity of support
γ-Al2O3 in carbon dioxide conversion of methane does
not explain the difference in the behavior of samples
Mo2C/х-Al2O3/α-Al2O3 (where x = γ, δ, or α).
The synthesized Mo2C/х-Al2O3/α-Al2O3 samples
(where x = γ, δ, or α) show catalytic activity in carbon
dioxide conversion of methane, which was carried out
in a contactor reactor with a membrane catalyst. It was
shown that the porous structure of membrane cata-
lysts substantially affects the results of carbon dioxide
conversion. Of the samples studied, the Mo2C/γ-
Al2O3/α-Al2O3 catalysts show the highest activity. The
difference in the catalytic activity of the samples can
be explained by the appearance of the Knudsen mech-
anism of mass transfer in the porous structure of the
catalysts.
It was suggested in [26] that an increase in the
activity of membrane catalysts is possibly associated
with the appearance of the Knudsen mechanism of
mass transfer in the porous structure of catalysts. It is
known that, as a result of gas dispersion in a porous
medium when pore sizes are less than the free path of
molecules, the free molecular mode of transport can
occur. In this case, there is an effect of decreasing the
probability of collisions of molecules with each other
in the gas phase to extremely low values. In this case,
the collision of molecules with the walls of the pores
starts to dominate. This leads to a more complete use
of the catalytic surface.
ACKNOWLEDGMENTS
The probability of the Knudsen regime of transport
in the pores can be estimated from the Knudsen num-
ber (Kn), which characterizes the ratio of the free path
of molecules to the characteristic pore size (for exam-
ple, to the average diameter). It is generally assumed
that for Kn > 10 a free molecular flow or Knudsen dif-
fusion is observed [14].
This work was supported by Mendeleev University
of Chemical Technology of Russia, project no. 028-
2018.
REFERENCES
1. Alekseeva, O.K., Alekseev, S.Yu., Amirkhanov, D.M.,
Kotenko, A.A., Chelyak, M.M., and Shapir, B.L.,
Membrany. Kriticheskie Tekhnologii, 2003, vol. 19,
no. 3, p. 20.
The values of Knudsen numbers determined for the
studied samples of supports and membrane catalysts
are given in Table 3. It can be seen that the transition
to free molecular flow should be observed in
γ-Al2O3/α-Al2O3 support samples. For these samples,
the calculated values of Kn exceed 10. For other sam-
ples of support, a transient transport mode should be
expected.
2. Tereshchenko, G.F., Malygin, A.A., Ermilova, M.M,
Orekhova, N.V., Volkov, V.V., Lebedeva, V.I., Petrova, I.V.,
Tsodikov, M.V., Teplyakov, V.V., Trusov, L.I., and
Moiseev, I.I., Kataliz v promyshlennosti. Spetsial’nyi
Vypusk, 2008, p. 36.
KINETICS AND CATALYSIS
Vol. 59
No. 5
2018

SOL–GEL SYNTHESIS OF MEMBRANE Мo2С/Al2O3 CATALYSTS
643
3. Basov, N.L., Ermilova, M.M., Orekhova, N.V., and 16. Szymanska-Kolasa, A., Lewandowski, M., Sayag, C.,
Yaroslavtsev, A.B., Rus. Chem. Rev., 2013, vol. 82,
p. 352.
Brodzki, D., and Djega-Mariadassou, G., Catal.
Today, 2007, no. 19, p. 35.
4. Sanchez, J. and Tsotsis, T.T., Membr. Sci. Technol.,
17. Darujati, A.D.S. and Thompson, W.J., Chem. Eng.
1996, vol. 4, p. 529.
Sci., 2006, no. 61, p. 4309.
5. Tan, X. and Li, K, Inorganic Membrane Reactors: Fun-
18. Gavrilova, N.N., Kruglaya, T.I., Myachina, M.A.,
Nazarov, V.V., and Skudin, V.V., Steklo Keram., 2017,
no. 1, p. 29.
damentals and Applications, Chichester: Wiley, 2015.
6. Dittmeyer, R., Höllein, V., and Daub, K., J. Mol. Catal.
A: Chem., 2001, vol. 173, nos. 1–2, p. 135.
7. Basile, A., Paola, L.Di., Hai, F.I., and Piemonte, V.,
Membrane reactors for energy applications and basic
chemical production, Woodhead publishing, 2015.
8. Makertihartha, I.G.B.N., Dharmawijaya, P.T., Zunita, M.,
and Wenten, I.G., Adv. Sci. Lett., 2017, vol. 23, no. 6,
p. 5726.
9. Tsodikov, M.V., Teplyakov, V.V., Fedotov, A.S., Moi-
seev, I.I., Kozitsyna, N.Y., Bychkov, V.Y., and
Korchak, V.N., Russ. Chem. Bull., 2011, no. 1, p. 55.
10. Bukharkina, T.V., Bazhenova, M.D., Gavrilova, N.N.,
Kryzhanovskii, A.S., and Skudin, V.V., Khim. Prom.
Segodnya, 2013, no. 11, p. 4.
11. Bukharkina, T.V., Gavrilova, N.N., Kryzhanovskii, A.S.,
Skudin, V.V., and Shul’min, D.A., Membrany i Mem-
brannye Tekhol., 2013, vol. 3, no. 2, p. 139.
19. Bazhenova, M.D., Gavrilova, N.N., and Nazarov, V.V.,
Colloid J., 2015, vol. 77, no. 1, p. 3.
20. Gavrilova, N.N., Nazarov, V.V., and Skudin, V.V.,
Kinet. Catal., 2015, vol. 56, no. 5, p. 670.
21. Adamu, A., Kislov, V.R., and Skudin, V.V., Usp. Khim.
Khim. Tekhnol., 2014, vol. 28, p. 17.
22. Chapman, S. and Cowling, T.G., The mathematical
theory of non-uniform gases: an account of the kinetic the-
ory of viscosity, thermal conduction and diffusion in gases,
Cambridge: Cambridge University Press, 1970, p. 423.
23. Vityaz’, P.A., Skudin, V.V., Gavrilova, N.N., Sudnik, L.V.,
Markova, L.V., Bazhenova, M.D., and Kryzhanovskii, A.S.,
Poroshkovaya Metallurgiya. Respublikanskii mezhve-
domstvennyi sbornik nauchnykh trudov, no. 36, p. 154.
24. Bazhenova, M.D., Gavrilova, N.N., Kryzhanovskii, A.S.,
Nazarov, V.V., Skudin, V.V., Vityaz’, P.A., and Sudnik, L.V.,
Khim. Prom-st. Segodnya, 2014, no. 1, p. 4.
12. Bukharkina, T.V., Gavrilova, N.N., and Skudin, V.V.,
Katal. Prom-sti., 2015, vol. 15, no. 4, p. 14.
25. Arutyunov, V.S. and Krylov, O.V., Russ. Chem. Rev.,
13. Nagy, E., Int. J. Comp. Mater., 2012, vol. 2, p. 79.
2005, vol. 74, no. 12, p. 1111.
14. Karniadakis, G., Beskok, A., and Aluru, N., Microflows
and nanoflows—Fundamentals and simulation, Appl.
Mathematics, New York: Springer, 2005, p. 818.
26. Pina, M.P., Menendez, M., and Santamaria, J., Appl.
Catal., B, 1996, no. 11, p. 19.
15. Grigoryan, E.A. and Merzhanov, A.G., Nauka, Proiz-
Translated by Andrey Zeigarnik
vodstvu, 1998, no. 3, issue 5, p. 30.
KINETICS AND CATALYSIS Vol. 59 No. 5 2018