Effect of a structure-size factor on the catalytic properties of complex oxide compositions in the reaction of deep methane oxidation

Article abstract of DOI:10.1134/S0023158407030111

The following catalysts for deep methane oxidation were studied using X-ray diffraction analysis, scanning electron microscopy, transmission electron microscopy, temperature-programmed reduction with hydrogen, and temperature-programmed ammonia desorption: highly dispersed supported aluminum-manganese and cobalt-zirconium catalysts; bulk and supported nanosized spinel ferrites. It was found that structure-size factors in complex oxide nanocomposites are responsible for differences in the catalytic properties (activity and thermal stability). The efficiency of the supported ferrite and cobalt-zirconium catalysts depends on both the oxygen-catalyst bond strength and the acid properties of catalyst surfaces.

Full text of DOI:10.1134/S0023158407030111

ISSN 0023-1584, Kinetics and Catalysis, 2007, Vol. 48, No. 3, pp. 414–429. © MAIK “Nauka /Interperiodica” (Russia), 2007.
Original Russian Text © M.R. Kantserova, S.N. Orlik, 2007, published in Kinetika i Kataliz, 2007, Vol. 48, No. 3, pp. 438–453.
Effect of a Structure–Size Factor on the Catalytic Properties
of Complex Oxide Compositions in the Reaction
of Deep Methane Oxidation
M. R. Kantserova and S. N. Orlik
Pisarzhevskii Institute of Physical Chemistry, National Academy of Sciences of Ukraine, Kiev, Ukraine
E-mail: mkantserova@ukr.net; orlyk@inphyschem-nas.kiev.ua
Received October 11, 2005
Abstract—The following catalysts for deep methane oxidation were studied using X-ray diffraction analysis,
scanning electron microscopy, transmission electron microscopy, temperature-programmed reduction with
hydrogen, and temperature-programmed ammonia desorption: highly dispersed supported aluminum–manga-
nese and cobalt–zirconium catalysts; bulk and supported nanosized spinel ferrites. It was found that structure–
size factors in complex oxide nanocomposites are responsible for differences in the catalytic properties (activity
and thermal stability). The efficiency of the supported ferrite and cobalt–zirconium catalysts depends on both
the oxygen–catalyst bond strength and the acid properties of catalyst surfaces.
DOI: 10.1134/S0023158407030111
INTRODUCTION
bulk and supported ferrites with the spinel structure
åIIFeI2II O4 (M is Mn, Co, or Ni) modified with surfac-
The development of energy generation by the flame-
less catalytic combustion of methane is one of the most tant additives.
promising areas in current fundamental and applied
catalysis. In this case, unlike torch combustion, 95–
100% methane oxidation to CO2 occurs at temperatures
lower than 800°C; this significantly increases the effi-
EXPERIMENTAL
Supported aluminum oxide catalysts. The sup-
ported aluminum oxide catalysts were prepared by the
treatment of supports with the supersaturated solutions
of active metal salts (barium nitrate, strontium nitrate,
lanthanum nitrate, copper nitrate, cobalt nitrate, chro-
mium nitrate, and manganese acetate) under nonequi-
librium conditions at an elevated temperature (method I)
[3, 4] and by the traditional impregnation of supports
with a mixture of soluble metal salts followed by drying
and calcination (method II). The total metal content of
a catalyst was 10–12 wt %: 5 wt % Mn (Cu, Co, or Cr),
the balance being lanthanum, barium, and strontium
modifying additives. The cobalt content of an unmodi-
fied cobalt catalyst was 10 wt %. Rare earth and alka-
line earth oxides were chosen as modifying additives
because of their ability to disperse fusible oxides of
transition metals, in particular, manganese (Tm = 535–
1080°C), and to stabilize low-temperature Al2O3 modi-
fications at overheating temperatures. Commercial A-1
and ShN-2 alumina samples and Al2O3 prepared by a
sol–gel method, which provides an opportunity to obtain
fine-powder materials with a particle size to 2–4 nm [5, 6],
were used as supports. Aluminum hydroxide was pre-
pared by the neutralization of a solution of Al(NO3)3
with a solution of NH4OH to pH 9–10 followed by
ciency of the process and provides an NOx emission
level of about 1 ppm. The absence of efficient low-tem-
perature catalytic systems with high performance char-
acteristics hinders the widespread use of the catalytic
combustion of gaseous hydrocarbon fuel. Supported
platinum and palladium catalysts are the most active
catalysts for the deep oxidation of methane. However,
because they are expensive, there is a need for the
development of new low-temperature catalysts without
noble metals.
The deep oxidation reactions of hydrocarbons, in
particular, methane, on catalysts containing noble met-
als are structure-sensitive; that is, the rate of these reac-
tions depends on the particle size of the active compo-
nent [1, 2]. The determination of the role of the struc-
ture–size factor in changing the catalytic properties of
oxide systems opens up new additional opportunities
for the development of less expensive catalysts for deep
methane oxidation with controllable characteristics. In
this work, we synthesized various nanosized complex
oxide systems and tested them in the deep methane oxi-
dation reaction. The following test systems were stud-
ied: aluminum oxide catalysts modified with rare earth
(REE) and alkaline earth elements (AEE); the cobalt–
zirconium oxide systems ëÓxOy/ZrO2 and ëÓxOy/(ZrO2–
zeolite), including Rh- and Pd-promoted systems; and
washing the resulting precipitate to remove NO–3 (reac-
414

EFFECT OF A STRUCTURE–SIZE FACTOR ON THE CATALYTIC PROPERTIES
415
tion with diphenylamine). The pressed precipitate as a small cylinders and dried at room temperature (xero-
hydrogel was treated in an autoclave at 150°C for 6 h. gel), was subjected to hydrothermal modification,
The resulting transparent sol was dried at 60–80°C until which provides an opportunity to regulate the pore
the precipitation of Al(OH)3, which was then calcined structure of the coprecipitated systems over a wide
range and, in a number of cases, facilitates the forma-
tion of fine crystalline phases with high specific surface
areas [14]. The modification was performed in an auto-
clave with saturated water vapor at 150°C for 6 h. After
the hydrothermal treatment, the xerogel was dried at
150°C and calcined at 400°C for 5 h.
in air at 600°C.
Bulk spinels. The bulk spinels were prepared by
coprecipitation from corresponding metal nitrate solu-
tions with an aqueous solution of sodium hydroxide at
a specified value of pH followed by keeping the suspen-
sion at room temperature. The resulting suspensions
were filtered; the precipitates were washed to a negative
reaction for NO–3 and dried at 110°C in air; thereafter,
The supported samples were prepared by the
impregnation of supports with a cobalt nitrate solution
under nonequilibrium conditions at an elevated temper-
ature followed by drying at 100°C and calcination at
350°C for 6 h. Reagent grade ZrO2, zirconia modified
with yttrium oxide (7 wt % Y2O3) [12], zirconium diox-
ide prepared by a sol–gel method, the H form of syn-
thetic pentasil zeolite TsVN (HTsVN), and the binary
system [65% ZrO2 (sol–gel)–35% HTsVN] were used
as supports. The last-named binary system was synthe-
sized in accordance with a published procedure [15] in
the following manner: A zeolite suspension was added
to an aqueous suspension of a zirconium hydrogel, and
the mixture was intensely stirred for 45 min to reach
higher homogeneity. Next, the sediment was pressed,
dried at 100°C, and calcined at 500°C for 3 h. The cat-
alysts based on the binary zeolite-containing support
were promoted with Rh and Pd additives (0.5 wt %)
using the solutions of palladium nitrate and rhodium
chloride (RhCl3 · 4H2O) followed by drying and calci-
nation. The concentration of cobalt oxide in all of the
samples was 10 wt % (on a metal basis).
they were calcined at 700°C for 7 h.
Note that homogeneous complex oxides are difficult
to prepare by methods commonly used for the prepara-
tion of spinels, for example, the coprecipitation of
hydroxides from corresponding metal salt solutions.
The spinels thus prepared are usually characterized by
low specific surface areas; they are inhomogeneous and
contain impurities that impair their physicochemical
properties [7, 8]. Surfactant additives can be used in
order to control the particle size of complex oxides [9].
According to patent data [10], bulk spinels, including
those modified with surfactant additives (to 6 wt % car-
boxymethyl cellulose), were prepared by the thermal
decomposition of the trinuclear heterometallic iron ace-
tate complexes [FeI2II MIIO(CH3COO)6(H2O)3] · 2H2O
(where MII is Mn, Co, or Ni). The advantage of this
method is that it provides an opportunity to obtain
nanosized complex oxides with controllable composi-
tions and particle sizes [11].
Supported spinel catalysts. The supported spinel
The phase composition of catalysts and supports
catalysts were prepared by the decomposition of pre- was determined by X-ray diffraction (XRD) using a
synthesized trinuclear transition metal carboxylate DRON-3M diffractometer (ëuKα radiation; λ =
complexes in porous support matrices [10]. Commer- 1.54184 Å). The average size of crystallites was deter-
cial ShN-2 alumina and zirconium dioxide modified mined from the Scherrer equation [16]. The specific
with yttrium oxide (7 wt % Y2O3) were used as supports surface area (Ssp) of samples was measured using the
[12]. The Y3+ cation was introduced into a support for thermal desorption of argon. The pore structure of par-
the low-temperature stabilization of a cubic ZrO2 mod- ticular samples was characterized based on the results
ification, which is characterized by a smaller particle of methanol adsorption (desorption) in a gravimetric
size, as compared to a monoclinic modification. The system equipped with a McBain–Bakr quartz helix bal-
addition of the Y3+ cation resulted in the formation of ance. The samples were preevacuated (P ~ 10–2 Torr) at
vacancies in the oxygen sublattice of ZrO2; this facili- 330°C. The total pore volume was evaluated from an
tates its crystallization in a cubic form [13]. Catalyst adsorption isotherm at P/Ps = 0.98 assuming that pores
samples were also prepared by mechanically mixing are filled with a condensed liquid adsorbate.
ferrites presynthesized from carboxylate complexes
The redox properties of catalysts and supports were
with a support (χ-Al2O3). The ferrite content was
studied by temperature-programmed reduction (TPR)
4−6 wt %.
with hydrogen over the range 20–900°C at a heating
Cobalt–zirconium oxide catalysts. The cobalt–zir-
conium oxide catalysts were prepared by coprecipita-
tion and impregnation. The (ëÓxOy–ZrO2) sample was
prepared by the coprecipitation of cobalt and zirconium
hydroxides with an aqueous ammonia solution from
cobalt and zirconium nitrate solutions at pH 9. The
resulting suspensions were filtered; the precipitates
rate of 10 K/min in a flow system equipped with a trap
with molecular sieves (–50°C) to remove water. The
amount of hydrogen consumed for the reduction of a
catalyst was measured by chromatography (thermal-
conductivity detector). The flow rate of a 10% ç2 mix-
ture in Ar was 50 cm3/min. The catalysts were pre-
heated at 250°C in a flow of argon (30 cm3/min) for 1 h.
were washed until the washings were free of NO–3 . The
The surface acid properties of samples were studied
washed and pressed precipitate, which was shaped into by temperature-programmed ammonia desorption
KINETICS AND CATALYSIS Vol. 48 No. 3 2007

416
Conversion of CH4, %
KANTSEROVA, ORLIK
itation of sparingly soluble metal salts on the catalyst
surface and to shorten the impregnation time [3].
100
80
60
40
20
0
In accordance with the Gibbs theory of homoge-
neous nucleation [17], the crystallites formed in the
course of crystallization from supersaturated solutions
are smaller than those obtained using dilute solutions.
However, these crystallites undergo rapid agglomera-
tion on heating. It was found [18] that, in the presence
of porous matrices, in which the resulting clusters are
isolated from each other, an increase in the temperature
did not cause the growth of particles. A limited concen-
tration of a solute within a pore resulted in the forma-
tion of a cluster with a limiting volume nmax (where n is
the number of atoms in the cluster); thereafter, the
growth of clusters became energetically unfavorable.
An analysis of changes in the Gibbs free energy func-
tion ∆G = f(n), which characterizes the growth of clus-
ters with increasing n, for dilute and supersaturated
solutions in an infinite volume and a closed pore dem-
onstrated that a minimum value corresponds to the
smallest cluster obtained from a supersaturated solution
in a closed pore. Consequently, published data [17, 18]
form the theoretical basis of the procedure proposed for
preparing supported nanosized catalysts from supersat-
urated solutions.
(a)
1
2
3
4
350 400 450 500 550 600 650
(b)
100
80
60
40
20
0
350 400 450 500 550 600 650
Temperature, °C
Fig. 1. The temperature dependence of methane conversion
on modified manganese catalysts prepared using method I
(a) before and (b) after treatment at 900°C. Catalyst sup-
ports: (1) A-1 (method II), (2) A-1, (3) ShN-2, (4) Al O
Figure 1 shows the temperature dependence of
methane conversion into CO2 on modified manganese
catalysts supported onto A-1, ShN-2, and alumina pre-
pared by the sol–gel method before and after treatment
at 900°C (method I). For comparison, data on the con-
version of methane on an analogous catalyst prepared
by ordinary impregnation of the A-1 support with metal
salts (method II) are given. As can be seen in Fig. 1, the
initial catalysts can be arranged in the following order
of increasing activity (depending on the nature of the
support and the preparation procedure):
2
3
(sol–gel).
(TPAD) in accordance with a procedure described else-
where [15].
The electron-microscopic study of particular sam-
ples was performed with the use of scanning and trans-
mission electron microscopes (REM-100U and JEOL
JEM 100 CX-II, respectively).
ShN-2 (method I) > A-1 (method I)
> Al2O3 (sol–gel; method I) > A-1 (method II).
The catalytic activity of samples (1 cm3; particle
size of 1–2 mm) was characterized by the conversion of
ëç4 into CO2 (which was determined in a quartz flow
reactor at atmospheric pressure and a space velocity of
6000 h–1 using a gas mixture containing 1% ëç4 in air)
and the temperature at which a 10 or 80% methane con-
version was reached (T10 or T80, respectively). The start-
ing substances and reaction products (ëç4, ëé2, and
CO) were analyzed by chromatography (thermal-con-
ductivity detector).
For the most active sample supported on ShN-2,
T10 = 400°C and T80 = 500°C, whereas the correspond-
ing values for a sample based on Al2O3 (sol–gel) are
400 and 550°C, respectively. Consequently, the temper-
ature at which 80% methane conversion was reached on
the catalyst prepared by method I is lower by 50°C than
that on the catalyst synthesized using the ordinary
impregnation method.
After the treatment of catalysts at 900°C for 5 h, the
order of activity remained unchanged; however, the
conversion of methane decreased by 10–15%, and the
activity of a sample on the support synthesized by the
sol–gel method decreased by 30%. Note that, even after
calcination at 900°C, the activity of the aluminum–
RESULTS AND DISCUSSION
Nanosize Effect in the Genesis
of Aluminum–Manganese Catalysts
The preparation of supported catalysts from the manganese catalyst prepared by method I was higher
supersaturated solutions of active metal salts under than the activity of an analogous sample prepared by
nonequilibrium conditions (method I) resulted in the ordinary impregnation and not subjected to high-tem-
dispersion of active components in porous support perature treatment. Thus, the activity and thermal sta-
matrices; it provided an opportunity to avoid the precip- bility of catalysts of the same chemical composition
KINETICS AND CATALYSIS Vol. 48 No. 3 2007

EFFECT OF A STRUCTURE–SIZE FACTOR ON THE CATALYTIC PROPERTIES
417
Table 1. Structure properties of aluminum oxides and aluminum–manganese catalysts (method I)
600°C, 2 h
900°C, 5 h
Sample
Ssp, m2/g
phase
L(Al2O3), nm Ssp, m2/g
phase*
L(Al2O3), nm
A-1
193
116
93
γ
χ
–
γ
χ
–
4
6
–
4
7
–
78
33
5
θ, (γ)
æ, (χ)
α
θ, γ, Mn2O3, Mn3O4
χ, Mn2O3, Mn3O4
α, La2O3 · 11Al2O3
10
11
ShN-2
Al2O3 (sol–gel)
>20
6
Mn–REE, AEE/A-1
184
68
98
53
7
Mn–REE, AEE/ShN-2
Mn–REE, AEE/Al2O3 (sol–gel)
9
65
>20
* Trace amounts of other phases are given in parentheses.
Table 2. Pore-structure characteristics of aluminum oxides and catalysts based on them
Total pore volume,
cm3/g
Micropore volume,
cm3/g
Predominant
pore diameter, nm
Sample
A-1
0.58
0.25
0.02
0.46
0.28
0.21
0.17
0.098
0.047
–
8.1
7.0–8.0
–
ShN-2
Al2O3 (sol–gel)
Mn–REE, AEE/A-1 (600°C)
Mn–REE, AEE/A-1 (900°C)
Mn–REE, AEE/ShN-2 (600°C)
Mn–REE, AEE/ShN-2 (900°C)
0.081
0.052
0.048
0.030
7.8
15.1
7.0
9.8
can be enhanced by the optimization of the preparation
procedure.
The X-ray diffraction pattern of a catalyst prepared
based on A-1 was identical to the X-ray diffraction pat-
tern of γ-Al2O3 with the same particle size (we failed to
detect a manganese-containing oxide phase). The X-ray
diffraction pattern of a catalyst based on ShN-2 exhib-
ited diffuse low-intensity diffraction peaks due to a
χ-Al2O3 phase with L = 7 nm; as in the case of the cat-
To find the reason for the higher activity of samples
synthesized by method I, we studied the phase compo-
sition of catalysts and corresponding supports after
treatment at 600°C (2 h) and 900°C (5 h) (Table 1).
According to XRD data, γ-Al2O3 and χ-Al2O3 are the
main phases of A-1 and ShN-2 supports, respectively. alyst based on A-1, phases corresponding to supported
The crystallite sizes (L), which were determined using
the lines 2θ = 45.74 and 67° for A-1 and 2θ = 67.3° for
ShN-2, were equal to 4 and 6 nm, respectively. Diffrac-
tion maximums were absent from the X-ray diffraction
pattern of Al2O3 (sol–gel); this fact suggests that the
sample was amorphous [19].
oxides were not observed. Rare earth and alkaline earth
oxides were not detected because their concentrations
in the catalysts are lower than the detection limit of this
technique. The X-ray amorphism of manganese oxides
can be explained by their low crystallinity, which
resulted from the synthesis temperature insufficiently
A study of the pore structure of samples (Table 2) high for the formation of a long-range order structure,
demonstrated that A-1 and ShN-2 supports are nano-
sized porous matrices with different total pore volumes.
In the A-1 support, the total volume of pores 4–18 nm
in diameter was 0.58 cm3/g with the predominance of
pores 8.1 nm in diameter. In the ShN-2 support, the
total volume of pores 4–18 nm in diameter was
0.25 cm3/g, and pores 7.5–8 nm in diameter were pre-
dominant. The sample of Al2O3 (sol–gel) was a wide-
pore material with a pore volume of 0.02 cm3/g and a
globule size of 17–21.5 nm, which was calculated from
the equation L = 6000/(Sspρtrue) [17].
or the small size of manganese oxide particles
(L ≤ 3 nm). An unmodified manganese catalyst with a
higher manganese oxide content (10% on a metal basis)
synthesized under analogous conditions was studied to
more accurately determine the phase composition and
crystallite size of catalysts. It was found that a tetrago-
nal modification of Mn3O4 with Mn2O3 impurities was
the main phase of manganese oxides. The crystallite
size of Mn3O4 was 5–6 nm, as determined from the line
2θ = 45.667°. These data suggest the formation of fine
KINETICS AND CATALYSIS Vol. 48 No. 3 2007

418
KANTSEROVA, ORLIK
manganese oxide particles in modified catalysts with a decrease in the phase transformation temperatures of
aluminum oxide with decreasing particle size.
lower manganese concentration (5 wt %).
The stabilizing effect of La, Ba, and Sr additives was
not found in the catalyst on amorphous Al2O3 (sol–gel).
As in the case of the pure support, the specific surface
area of the catalyst decreased by almost one order of
magnitude (to 7 m2/g), and the X-ray diffraction pattern
exhibited diffraction lines characteristic of α-Al2O3 and
the lanthanum aluminate La2O3 · 11Al2O3 [23], the
crystallite size of which is greater than 20 nm. Thus, the
Al2O3 (sol–gel) support is more reactive than the A-1
and ShN-2 supports toward both solid-phase interac-
tions with active catalyst components (La2O3) and
intrinsic phase transformations. This fact can explain
the lower activity and stability of catalysts based on this
support.
The XRD analysis of alumina supports and manga-
nese catalysts based on these supports with the modify-
ing additives of rare earth and alkaline earth elements
(before and after treatment at 900°C) suggests that the
catalysts can be arranged in the following order with
respect to thermal stability to phase transformations:
It is likely that the blocking of pores with active
components was responsible for the observed decrease
in the specific surface areas of modified catalysts based
on A-1 and ShN-2, as compared with those of pure sup-
ports (in both cases, the total pore volume decreased by
20%). A more dramatic decrease in the specific surface
area of the catalyst supported on ShN-2 can be due to
the fact that the total pore volume of this catalyst is
lower than that of A-1 by a factor of 2. The specific sur-
face area of the catalyst based on Al2O3 (sol–gel) can
decrease because of the solid-phase interaction of
active components with the support [5].
After treatment at 900°C for 5 h, the X-ray diffrac-
tion patterns of the supports exhibited diffraction lines
due to the following high-temperature modifications of
aluminum oxide: θ-Al2O3 with a small impurity of
γ-Al2O3 for A-1, æ-Al2O3 with an impurity of χ-Al2O3 for
ShN-2, and α-Al2O3 for the sample of Al2O3 (sol–gel).
The appearance of the above phases corresponds to the
sequence of phase transformations in γ-Al2O3, χ-Al2O3,
and Al2O3 (sol–gel) in the course of high-temperature
treatment [20], and it was accompanied by a noticeable
decrease in the degree of dispersion. The crystallite size
was estimated at ~10–11 nm (A-1 and ShN-2). In this
case, the specific surface area decreased to 78 (A-1) or
33 m2/g (ShN-2). The sample of Al2O3 (sol–gel) was
characterized by a well-crystallized α phase with a spe-
cific surface area of 5 m2/g.
Mn–REE, AEE/χ-Al2O3 ≥ Mn–REE,
AEE/γ-Al2O3 > Mn–REE, AEE/Al2O3 (sol–gel),
which is consistent with the activity order of catalysts
prepared using method I.
A number of samples were studied by TPR in order
to evaluate the oxygen–catalyst bond strength; the ini-
tial rate of reduction of oxides with hydrogen (reduc-
ibility) can serve as a relative characteristic of this bond
strength. Hydrogen consumption maximums are absent
from the TPR curves of the A-1 and ShN-2 supports;
this fact suggests a high strength of the M–O bond in
aluminum oxides and, consequently, the absence of cat-
alytic activity. The TPR curves of manganese catalysts
based on ShN-2 and A-1 are characterized by a wide
region of hydrogen consumption with a diffuse maxi-
mum at 370°C (Fig. 2). This result is consistent with
data published by Kapteijn et al. [24], who attributed
this maximum to the reduction of fine Mn2O3 in the alu-
minum–manganese catalyst. According to Stohmeir
and Hercules [21], Mn2O3 with a particle size of 5–6 nm
is the main manganese phase in the 4% Mn/Al2O3 cat-
alyst. The absence of diffraction maximums corre-
sponding to manganese oxides can be explained by
their smaller particle size, which resulted from the cat-
alyst preparation procedure. The catalyst based on
ShN-2 was characterized by a broader (170–650°C)
and less intense region of hydrogen consumption, as
compared with the sample based on A-1 (250–550°C).
This result can be explained by the difference between
the pore structures of the supports: the broader range of
predominant pore diameters in ShN-2 (7–8 nm) can
lead to an analogous size distribution of supported
oxides. The diffuse maximum in the TPR curve of the
catalyst based on ShN-2 suggests the appearance of an
internal size effect in the aluminum–manganese nano-
A comparison between the X-ray diffraction pat-
terns of the catalyst on ShN-2 before and after high-
temperature treatment indicates that the composition of
the support remained unchanged (χ-Al2O3), but the
crystallite size increased to 9 nm. In addition, diffrac-
tion peaks due to the manganese oxides Mn2O3 and
Mn3O4 were detected. The weak and diffuse reflections
of the above oxides are indicative of a small particle
size (L ~ 3 nm), which is smaller than the particle size
of the support. The absence of the æ-Al2O3 phase from
the diffraction pattern of the calcined catalyst suggests
a stabilizing effect of La, Ba, and Sr additives, which
inhibit the phase transformations of the support. In the
catalyst based on A-1, the phase composition of the
support somewhat changed (a θ phase appeared); the
size of γ-Al2O3 crystallites increased to 6 nm; and, as in
the study by Stohmeir and Hercules [21], diffraction
peaks due to the manganese oxides Mn2O3 and Mn3O4
were detected. The stabilizing effect of additives con-
sists in the inhibition of Al2O3 phase transformations. It
manifested itself in the fact that the specific surface area
of catalysts after heating at 900°C (5 h) decreased much
less than the Ssp of a pure support. The formation of
θ-Al2O3 even in the presence of stabilizing additives
can be explained by the phase size effect. According to
Uvarov and Boldyrev [22], this effect consists in a system [25]: a decrease in the reduction temperature of
KINETICS AND CATALYSIS Vol. 48 No. 3 2007

EFFECT OF A STRUCTURE–SIZE FACTOR ON THE CATALYTIC PROPERTIES
Hydrogen consumption
419
manganese oxide to 170°C with decreasing particle
size. After the treatment of catalysts at 900°C, the max-
imum intensity of hydrogen consumption dramatically
decreased. As found by XRD, this was due to the con-
version of Mn2O3 into Mn3O4, which undergoes reduc-
tion at a higher temperature. A comparison of the TPR
curves of manganese catalysts calcined at 900°C dem-
onstrated that the TPR peak intensity of the sample
based on ShN-2 was higher than that on A-1; this sug-
gests the retention of a considerable amount of Mn2O3
in the former sample. This can explain the higher activ-
ity of the sample based on ShN-2 after treatment at
900°C.
4
3
2
1
To study the effect of the nature of the transition
metal on the activity of aluminum oxide catalysts, we
synthesized (using method I) and tested samples based
on A-1 modified with rare earth and alkaline earth ele-
ments. In these samples, the oxides of copper, chro-
mium, and cobalt and their mixtures, in particular, with
manganese oxide, were used as active components
(Fig. 3). The experimental data suggest that the cata-
lysts are arranged in the following order of activity in
deep methane oxidation:
200
300
400
500
600
700
Temperature, °C
Fig. 2. TPR curves of aluminum–manganese catalyst sam-
ples based on A-1 and ShN-2 after thermal treatments under
various conditions: (1) Mn–REE, AEE/A-1 (900°C), (2)
Mn–REE, AEE/A-1 (600°C), (3) Mn–REE, AEE/ShN-2
(900°C), and (4) Mn–REE, AEE/ShN-2 (600°C).
MnLaBaSr > MnCuLaBaSr > CuLaBaSr
> CuCrLaBaSr > CoREEBaSr > CoLaBaSr > Co.
Conversion of CH4, %
100
90
80
70
The MnLaBaSr catalyst was more active than the
samples containing the oxides of other transition metals
(Cu, Cr, and Co) or their binary compositions. There-
fore, aluminum–manganese catalysts are promising for
use in the test reaction and it is reasonable to optimize
their composition further.
60
1
2
3
4
5
6
7
50
40
30
20
10
0
Thus, the study of the structure and texture charac-
teristics of aluminum–manganese catalysts for deep
methane oxidation demonstrated that the supporting of
an active component and stabilizing additives (La, Ba,
and Sr) under nonequilibrium conditions resulted in the
formation of nanosized active phases in the porous
matrix of an Al2O3 support, the retention of nanosized
alumina particles, and the stabilization of low-tempera-
ture alumina modifications (γ-Al2O3 and χ-Al2O3). This
is responsible for the higher activity (100% methane
conversion was reached at 550–600°C) and thermal sta-
bility of aluminum–manganese systems synthesized by
the proposed procedure, as compared with the well-
400
450
500
550
600
650
Temperature, °C
Fig. 3. The temperature dependence of methane conversion
on modified oxide catalysts based on A-1: (1) Co, (2)
CoREEBaSr, (3) CoLaBaSr, (4) CuLaBaSr, (5) CuCrLa-
BaSr, (6) MnCuLaBaSr, and (7) MnLaBaSr.
known catalysts of analogous composition prepared lysts in the reaction of deep methane oxidation seems
using traditional impregnation [26, 27]. The appear- promising.
ance of an internal size effect in the aluminum–manga-
The most effective spinels are the ferrites MFe2O4
nese nanosystem was found, which consisted in a
(M = Mn, Co, Ni, and Zn), which exhibited higher
decrease in the phase transformation temperatures of
activity in the test reaction than that of aluminates.
aluminum oxide and the reduction temperature of man-
They are less toxic and less expensive than chromites;
ganese oxide with decreasing particle size.
their stability under oxidizing conditions is higher than
that of cobaltites [28, 29].
Effect of the Size Factor on the Catalytic Properties
Table 3 summarizes the structure characteristics
of Spinel Ferrites
(phase composition, crystallite size, and Ssp) and cata-
Along with highly dispersed aluminum oxide sys- lytic activity (temperatures at which 10, 50, and 100%
tems, the use of nanosized complex spinel oxide cata- methane conversion was reached) of bulk ferrites.
KINETICS AND CATALYSIS Vol. 48 No. 3 2007

420
KANTSEROVA, ORLIK
Table 3. Structure characteristics and catalytic activity of bulk ferrites
Temperature* at which specified methane
Indexed phases
(modification)
conversion was reached, °C
2
*
Ssp , m /g
Sample
L (CSR), nm
10%
50%
100%
CoFe2O4
NiFe2O4
MnFe2O4
CoFe2O4
CoFe2O4
9
8
47.0
0.6
<350
550
375
525
450
600
425
–
410
**
500
**
NiFe2O4
MnFe2O4
CoFe2O4
NiFe2O4
CoFe2O4
NiFe2O4
35.0
0.8
460
650
515
**
550
**
15
27.0
0.4
650
**
>50
>50
7.5
8
15.0
–
515
–
600
–
(from inorganic salts)
NiFe2O4
16.0
–
410
–
480
–
600
–
(from inorganic salts)
CoFe2O4 + surfactant
54.0
5.0
<350
500
375
500
370
650
450
600
500
**
NiFe2O4 + surfactant
33.0
7.0
550
**
* Top, initial catalyst; bottom, catalyst after heating at 900°C for 5 h.
** The specified conversion was not reached on the sample over a temperature range to 650°C.
According to XRD data, all of the tested ferrite sam-
ples exhibited a cubic spinel structure and the particle
size of these samples depended on preparation proce-
dure. Spinels with a particle size from 7.5 to 15 nm
were formed in ferrite synthesis by the thermal decom-
position of polynuclear complexes, whereas analogous
samples prepared by the decomposition of inorganic
salts contained particles of size >50 nm. Modification
with surfactant additives had no effect on the particle
size of nickel ferrite and somewhat increased the degree
of dispersion of cobalt ferrite.
> CoFe2O4(from inorganic salts) (0).
The best results were obtained for cobalt and nickel
ferrites with a particle size of 7.5–9 nm. The retention
of the particle size of catalysts in the nanometer range,
which is optimal for the occurrence of the given reac-
tion, is responsible for high catalytic activity in deep
methane oxidation [2, 30].
All of the ferrites have the same structure of the
inverse spinel [31]; therefore, it is most likely that dif-
ference in their catalytic activity depends on the prop-
erties of the M2+ cation, which is a constituent of spinel.
The lower catalytic activity of manganese ferrite is
related to its instability on heating under oxidizing con-
ditions. According to XRD, thermal analysis, and IR-
spectroscopic data [31], the growth of particles and the
decomposition of manganese-containing spinel into
oxides (MnO, which was oxidized to Mn2O3; Fe2O3)
were observed on heating above 400°C; this was
accompanied by the agglomeration of particles and a
decrease in Ssp (Table 3). The heating of cobalt- and
nickel-containing catalysts over the temperature range
400–700°C was not accompanied by the decomposition
of spinels.
Nanosized cobalt and nickel ferrites (L = 8–9 nm),
which were synthesized from heterometallic trinuclear
complexes, were much more active than analogous
spinels prepared by coprecipitation from metal salt
solutions (Table 3): the temperatures of 10 and 100%
methane conversion on nickel and cobalt ferrites were
lower by 50 and 100°C, respectively.
A comparison of the specific catalytic activity of
spinels, which were prepared using various techniques,
at relatively low temperatures (to 450°C) indicated the
effect of the size factor on the reaction rate of deep
methane oxidation. This effect consists in an increase in
the specific catalytic activity of cobalt and nickel fer-
rites with decreasing particle size (the specific catalytic
activity of spinels w × 103, ml ëç4 m–2 min–1, is given
in parentheses):
The higher activity of cobalt ferrite can be due to the
fact that an oxygen environment in the lattice of a com-
plex oxide facilitates the transition of a transition metal
ion to the highest oxidation state (Co3+). Moreover,
according to Tret’yakov [32], because of a relatively
low energy of disordering, the surface layer of cobalt
NiFe2O4 (9.1) > CoFe2O4 + surfactant (8.0)
> CoFe2O4 (6.3) > NiFe2O4 (from inorganic salts) (3.4)
KINETICS AND CATALYSIS Vol. 48 No. 3 2007

EFFECT OF A STRUCTURE–SIZE FACTOR ON THE CATALYTIC PROPERTIES
421
Table 4. Structure characteristics and catalytic activity of supports and catalysts
Temperature* at which
specified methane conversion was
2
*
Ssp , m /g
Sample
Indexed phases (modification)
reached, °C
10%
50%
100%
CoFe2O4/Al2O3
NiFe2O4/Al2O3
χ-Al2O3 (cubic)
96.0
63.0
91.0
61.0
78.0
14.0
91.0
18.0
116.0
33.0
9.5
425
475
375
450
460
550
410
540
540
560
485
550
560
625
500
590
650
650
″
″
″
″
600
650
CoFe2O4–Al2O3
(mix)
650
>650
600
NiFe2O4–Al2O3
(mix)
>650
Al2O3 (ShN-2)
Inactive
CoFe2O4/ZrO2–Y
NiFe2O4/ZrO2–Y
ZrO2 (cubic, monoclinic [traces])
ZrO2 (cubic, monoclinic [traces])
ZrO2 (cubic, monoclinic [traces]), NiFe2O4
ZrO2 (cubic, monoclinic [traces])
400
390
–
463
470
–
600
550
–
13.0
–
ZrO2–Y
33.0
570
650
>650
* Top, initial catalyst; bottom, catalyst after heating at 900°C for 5 h.
ferrite can be rapidly rearranged with a retention of the
bulk in a stable state. This provides a rapid supply of
oxygen, which is required for the occurrence of the
reaction in accordance with the mechanism of deep
hydrocarbon oxidation, to the catalyst surface.
According to XRD data (Fig. 4), the main phase of
zirconium dioxide modified with Y2O3 is cubic ZrO2
with small impurities of a monoclinic modification
(curve 9). The X-ray diffraction patterns of supported
samples exhibited diffraction maximums correspond-
ing to the cubic modifications of supports: zirconium
dioxide with small impurities of a monoclinic modifi-
cation and aluminum oxide (χ-Al2O3); the reflections of
NiFe2O4 and CoFe2O4 phases were absent (curves 3, 4, 7).
The X-ray diffraction pattern of a catalyst prepared by
mechanically mixing a ferrite with ZrO2 (curve 8) is
identical to the X-ray diffraction patterns of a supported
catalyst (curve 7) and a support (curve 9). The X-ray
diffraction pattern of a catalyst prepared by mechani-
cally mixing Al2O3 with nickel ferrite (curve 1) is anal-
ogous to the diffraction pattern of the pure support cal-
cined at 400°C for 5 h (curve 2). The X-ray diffraction
patterns of supported samples exhibited no diffraction
peaks corresponding to nickel and cobalt ferrites, the
X-ray amorphism of which can be explained by high
dispersion (L ≤ 3 nm). Moreover, reflections corre-
sponding to a cubic spinel phase, which is isomorphous
to cubic ZrO2 and Al2O3 phases, can be masked by
more intense diffraction peaks of supports. An addi-
tional argument for the formation of spinels in a porous
support matrix is that the diffraction patterns of sup-
ported catalysts exhibited no reflections characteristic
of a rhombohedral Fe2O3 phase, which is a spinel
decomposition product.
After high-temperature treatment, the activity of
catalysts decreased as a result of the agglomeration of
spinel particles and the decrease of Ssp (Table 3). The
modification of catalysts by surfactant additives had no
effect on the activity of nickel ferrite, whereas the tem-
perature at which high methane conversions were
reached on cobalt ferrite decreased by 40°C. In this
case, the catalysts exhibited higher thermal stability
than that of analogous unmodified samples. It is likely
that, on the addition of surfactant additives at the stage
of catalyst preparation, a nanoparticle is surrounded by
a layer of bulky structures, whose geometric size is
responsible for the appearance of a steric barrier, which
prevents the contact and agglomeration of nanoparti-
cles.
Table 4 summarizes the results of a study of the
structure characteristics and activity of supports and
supported ferrite catalysts. The decrease of Ssp in cata-
lysts based on ShN-2, as compared with that of the sup-
port, can be due to the blocking of pores with spinel
particles because the total pore volume also decreased
by 20%. As demonstrated above, the ShN-2 support—
a nanosized porous matrix with a predominant pore
diameter of 7–8 nm—is a low-temperature modifica-
tion of χ-Al2O3 (Tables 1, 2).
KINETICS AND CATALYSIS Vol. 48 No. 3 2007

422
I(2θ)
KANTSEROVA, ORLIK
Consumption of H2
(‡)
1
2
(‡)
3
1
2
4
5
0
20
40
60
80
200 300 400 500 600 700 800 900
(b)
(b)
6
7
8
3
4
9
5
0
20
40
60
2θ, deg
80
100
6
200 300 400 500 600 700 800
Temperature, °C
Fig. 4. X-ray diffraction patterns of support samples and
ferrite catalysts: (1) NiFe O –Al O , (2) Al O (400°C;
2
4
2
3
2 3
3
5 h), (3) NiFe O /Al O , (4) CoFe O /Al O , (5) Al O ,
Fig. 5. TPR curves of ferrite catalyst samples: (1) NiFe O ,
2 4
(2) CoFe O , (3) Fe O /ZrO , (4) CoFe O /ZrO , (5)
2
4
2
3
2
4
2
2
3
2
(6) NiFe O /ZrO (900°C; 5 h), (7) NiFe O /ZrO ,
2
4
2
2
4
2 4 2 4 2 2 4 2
(8) NiFe O –ZrO , and (9) ZrO .
NiFe O /Al O , and (6) CoFe O /Al O .
2 4 2 3 2 4 2 3
2
4
2
2
The X-ray diffraction pattern of a nickel-containing curve 6), which can be attributed to a cubic NiFe2O4
phase. It is likely that the appearance of this phase was
related to the growth of spinel grains due to the agglom-
eration of nanoparticles. The absence of reflections cor-
responding to a Fe2O3 phase suggests that the NiFe2O4
spinel did not decompose in the course of the high-tem-
perature treatment of the supported catalyst.
catalyst supported on ZrO2, which was treated at 900°C
for 5 h, exhibited diffraction peaks due to a cubic zirco-
nia modification with small impurities of a monoclinic
modification and a shoulder at about 2θ = 35.7° (Fig. 4,
Table 5. TPR study of ferrite catalyst samples
A comparison between supported catalysts based on
ferrites showed that the samples prepared by the
decomposition of polynuclear complexes in the pores
of a support (χ-Al2O3) were more active and more sta-
ble than the samples prepared by mechanically mixing
ferrites with the support. A porous support matrix is a
structural restrictor in the course of spinel agglomera-
tion at high temperatures, which facilitates the retention
of the catalytic activity of samples after high-tempera-
ture treatment. The difference in the catalytic activity of
supported ferrites can be explained by the fact that
cobalt ferrite is less stable than nickel ferrite [31]: after
treatment at 700°C, spinel began to decompose into
iron and cobalt oxides; the latter can react with the sup-
port to form inactive ëÓAl2é4 [33]. The use of zirco-
nium dioxide as a support for ferrites allowed us to
Peak
Reduction onset
temperatures, °C
Sample
temperature
(To), °C
Tmax1
Tmax2
CoFe2O4
295
365
285
350
285
210
570
555
525–580
450
–
–
NiFe2O4
CoFe2O4/Al2O3
NiFe2O4/Al2O3
CoFe2O4/ZrO2
NiFe2O4/ZrO2
ZrO2
–
435
–
465
610
600
–
385
630
KINETICS AND CATALYSIS Vol. 48 No. 3 2007

EFFECT OF A STRUCTURE–SIZE FACTOR ON THE CATALYTIC PROPERTIES
423
decrease the temperature of the onset of reaction (10% vation of a C–H bond on the surface of an oxide catalyst
methane conversion) and the temperature at which high can involve both homolytic and heterolytic decomposi-
methane conversions were reached, as compared with tion. Because the C–H bond in methane is weakly
samples based on Al2O3.
acidic (pKα = 46), the active surface sites capable of
deprotonating ëç4 should be strongly basic. However,
these sites rapidly undergo self-poisoning because of
strong CO2 adsorption. The acid–base pair of a metal
cation (acid site) and an oxygen anion (basic site) is
considered an active center for the deprotonation of the
methane C–H bond on the surface of metal oxides.
Deprotonation can result in the formation of adsorbed
OH groups and the CH–3 ion at the metal cation [36].
Table 5 and Fig. 5 summarize the results of a TPR
study of the redox properties of ferrite catalysts. The
TPR profile of a CoFe2O4 sample is characterized by
lower temperatures of the onset of reduction (To) and
maximum ç2 consumption (THm2ax ), as compared with
that of NiFe2O4. This suggests a lower bond strength of
oxygen in the cobalt-containing ferrite. The TPR curve
of modified zirconium dioxide exhibited a maximum in
the region 570–740°C with THm2ax at 630°C. A shift of
the THm2ax peak to the lower temperature region suggests
According to Wu et al. [37], the ignition temperature of
deep propane oxidation decreased because of an
increase in the acid site strength of catalysts based on
γ−Al2O3 and ZrO2. Yan et al. [38] believed that the pro-
tons of a support in cobalt-containing catalysts served
as methane activation sites in the SCR process
a higher lability and/or reactivity of lattice oxygen in
modified zirconium dioxide than that in Al2O3 (ShN-2),
(CH4 + NO + O2
N2 + CO2 + H2O). A study of the
which was not reduced over the test temperature range
acid properties of supported ferrites using TPAD dem-
onstrated that the surface concentration of acid sites in
a sample based on Al2O3 (3.1 × 10–3 (mmol NH3)/m2)
was higher than that in a sample based on ZrO2 (1.8 ×
(to 900°C). The supported catalysts exhibited a
H2
decrease in To and T , as compared with bulk spinel
max
samples; this can be due to an increase in the degree of
dispersion of the ferrites. The TPR profile of the
CoFe2O4/Al2O3 catalyst is characterized by a broader
and less intense region of hydrogen consumption than
that of NiFe2O4/Al2O3. This can be due to the formation
of CoAl2O4 in supported cobalt ferrite; because of this,
the amount of weakly bound oxygen, which is required
for the occurrence of the reaction, decreased.
10–3 (mmol NH3)/m2). In this case, nickel ferrite sup-
ported on zirconium dioxide was characterized by the
presence of stronger acid sites: the temperature of a
NH3
desorption maximum of NH3 (T ) was equal to
max
NH3
max
300°C (against T
= 160°ë for NiFe2O4/Al2O3). The
higher activity of NiFe2O4/ZrO2 can also be related to
the occurrence of stronger surface acid sites, which are
of importance for deep hydrocarbon oxidation [37].
Consequently, a relationship between the activity and
the redox and acid properties of supported ferrite cata-
lysts was established.
The supported zirconium oxide catalysts exhibited
an increase in the intensity of hydrogen consumption,
as compared with aluminum oxide catalysts. This can
be due to an increase in the mobility of lattice oxygen
in modified ZrO2 upon doping with ferrites. Moreover,
unlike ZrO2, Al2O3 and MeFe2O4 crystallize in a spinel-
type structure; therefore, the isomorphous solid solu-
tions of MeFe2O4 with Al2O3 can be formed along with
ferrites. This increases the degree of interaction
between ferrite and the support and, consequently,
decreases the amount of weakly bound oxygen in cata-
lysts supported on Al2O3, as compared with those on
ZrO2. The TPR data demonstrated that the amount and
lability (reactivity) of oxygen increase in the order
Thus, as a result of the study of the spinel catalysts,
we found that the ferrite preparation procedure affected
the catalyst activity: the most active catalysts were
nanosized cobalt and nickel ferrites prepared by the
thermolysis of polynuclear iron complexes. In the pres-
ence of the ëÓFe2O4 catalyst, the temperature of 100%
ëç4 conversion was 500°C, which is lower than that on
well-known spinel-type catalysts by 100–150°C [39–
41]. It was found that the use of a support (Al2O3) and
the addition of surfactants increased the thermal stabil-
ity of catalysts. The effect of the size factor on the reac-
tion rate of deep methane oxidation at relatively low
temperatures (to 450°C) was found. This effect consists
in an increase in the specific catalytic activity of cobalt
and nickel ferrites with decreasing ferrite particle size.
CoFe2O4 > NiFe2O4 > NiFe2O4/ZrO2
> CoFe2O4/ZrO2 > NiFe2O4/Al2O3
> CoFe2O4/Al2O3 > ZrO2 > Al2O3.
This order is consistent with the order of activity of
the test supports and catalysts that were not subjected to
high-temperature treatment.
The activation of a C–H bond is required for the het-
erogeneous catalytic oxidation of hydrocarbons; the
mechanism of this activation is under discussion [34–
38]. The activation of alkanes is related to the occur-
Effect of the Structure–Size Factor on the Catalytic
Properties of Cobalt–Zirconium Nanosystems
The use of compositions based on ZrO2, which
rence of atomic oxygen and acid–base sites on the cat- exhibit high thermomechanical properties, and zeolites,
alyst surface [34]. Burch et al. [35] noted that the acti- in which supported elements occur in cavities, seems
KINETICS AND CATALYSIS Vol. 48 No. 3 2007

424
KANTSEROVA, ORLIK
Table 6. Structure characteristics and catalytic activity of supports and catalysts
Temperature* at which
specified methane conver-
sion was reached, °C
Sample
Indexed phases* (modification)
L, nm Ssp, m2/g
10%
325
80%
397
(CoıéÛ–ZrO2)
ZrO2 (cubic)
12–13 141
Co3O4
14
5
–
400
–
589
–
ZrO2 (cubic [60%], monoclinic [30%])
28–30
CoıéÛ/ZrO2
Co3O4
12
30
19
8
367
370
363
409
490
500
467
520
(reagent grade)
ZrO2 (monoclinic)
CoıéÛ/ZrO2
(sol–gel)
ZrO2 (tetragonal [75%], monoclinic [15%])
12
62
9.6
CoıéÛ/(ZrO2–Y)
Co3O4
13
50
–
280
–
475
–
ZrO2 (cubic [80%], monoclinic [10%])
16–17
Co3O4
18–20
16–17
28
–
370
–
480
–
ZrO2 (cubic [80%], monoclinic [10%])
CoıéÛ/[ZrO2(sol–
gel)–HTsVN]
Co3O4
15
–
78
50
–
350
440
575
470
450
400
483
605
ZrO2 (tetragonal), pentasil
HTsVN
pentasil
–
>650
623
ZrO2 (reagent grade) ZrO2 (monoclinic)
26–28
10–12
15–16
11
80
58
ZrO2 (sol–gel)
(ZrO2–Y)
ZrO2 (tetragonal [75%], monoclinic [15%])
ZrO2 (cubic [90%], monoclinic [10%])
592
563
* Top, initial catalyst; bottom, catalyst after heating at 800°C for 1 h.
effective in order to decrease the temperature of cata- Y2O3 at the stage of support preparation facilitated the
lytic methane oxidation [42].
uniform distribution of the additive between highly dis-
persed ZrO2 particles and prevented their agglomera-
tion.
Table 6 and Fig. 6 show data on the catalytic activity
of the test supports and cobalt-containing catalysts
based on them. The conversion of methane on the
Catalysts based on zeolite HTsVN did not exhibit
Coxéy/γ-Al2O3 catalyst is given for comparison. The high activity: the complete oxidation of methane
samples based on zirconium dioxide exhibited the high- occurred at a temperature higher than 650°C. The use
est activity in the test reaction. This activity was inde- of the [ZrO2 (sol–gel)–HTsVN] binary system as a sup-
pendent of procedures used for the preparation of both port for cobalt oxides increased the activity of the cata-
the support and the catalyst. The lowest temperatures of lyst, as compared with that of a sample based on
80 and 100% methane conversion were observed in a HTsVN, and decreased the temperature of the onset of
coprecipitated sample (Coxéy–ZrO2). A comparison of reaction, as compared with that on the ëÓxéy/ZrO2
the activities of zirconium-containing catalysts showed (sol–gel) catalyst. The activity of a catalyst on a binary
that the introduction of a modifier additive (Y3+) and the support reached a maximum upon promotion with Rh
use of a sol–gel method for preparing the support and Pd: the temperatures of the onset of reaction and
resulted in a decrease in both To (10% conversion of 80% conversion of methane were lower by 50–90°C
ëç4) and temperatures at which high methane conver- than those on an unpromoted sample. According to the
sions were reached. The activity of samples decreased electronic theories of catalysts, the introduction of a
after high-temperature treatment because of the donor noble metal impurity into the composition of an
agglomeration of catalyst particles, which was accom- oxide catalyst facilitates the activation of hydrocarbons
panied by a decrease in the specific surface area. The because of a greater (than that of oxides) tendency to
catalyst based on ZrO2 modified with yttrium exhibited electron-acceptor interactions with oxygen. Moreover,
the highest thermal stability. Evidently, the addition of promoters prevent agglomeration and recrystallization
KINETICS AND CATALYSIS Vol. 48 No. 3 2007

EFFECT OF A STRUCTURE–SIZE FACTOR ON THE CATALYTIC PROPERTIES
Conversion of CH4, %
425
100
80
60
40
20
0
(‡)
1
2
3
4
5
6
7
200
300
400
500
600
700
57.80 nm
(‡)
100
80
60
40
20
0
(b)
1
2
3
4
5
350 400 450 500 550 600 650
Temperature, °C
57.80 nm
(b)
Fig. 6. The temperature dependence of methane conversion
on cobalt–zirconium oxide catalysts (a) before and (b) after
treatment at 800°C. Catalysts: (1) (Co é –ZrO ), (2)
x
y
2
Co é /ZrO (reagent grade), (3) Co é /ZrO (sol–gel), (4)
Fig. 7. Electron micrographs of the samples of (a) ZrO
2
modified with Y O and (b) a catalyst based on it; magnifi-
x
x
y
y
2
2
x
y
2
Co é /ZrO –Y, (5) Co é /[ZrO (sol–gel)–HTsVN], (6)
x
y
2
2 3
Co é /Al O , and (7) Co é /HTsVN.
cation, 82000×.
x
y
2
3
x y
processes in a catalyst based on the binary zeolite-con-
taining composition during high-temperature treatment
to enhance its thermal stability. A catalyst supported on
γ-Al2O3 exhibited the lowest activity: the temperatures
of the onset of reaction and 80% conversion of methane
were higher by 250–300°C than those on samples based
on ZrO2 and its modified forms.
To reveal reasons for different activities of cobalt-
containing catalysts, we studied the structure character-
istics, redox properties, and acid properties of a number
of samples.
Table 6 summarizes the structure characteristics of
supports and the most active catalysts. An analysis of
the X-ray diffraction pattern of ZrO2 (reagent grade)
demonstrated the presence of only a single phase: a
monoclinic modification of zirconium dioxide with a
particle size of 26–28 nm. According to XRD data, a
tetragonal modification with a particle size of 10–12 nm
was predominant in the sample of ZrO2 (sol–gel);
the main phase of ZrO2 modified withY2O3 was a cubic
modification of zirconium dioxide (L = 15–16 nm) with
small monoclinic impurities. The use of a sol–gel
method for the preparation of zirconium dioxide and
the introduction of a modifying additive of Y2O3 had a
significant effect on the support particle size: the degree
of dispersion increased by 45–55%. The X-ray diffrac-
tion pattern of a catalyst supported on modified zirco-
nium dioxide exhibited diffraction maximums corre-
Depending on the preparation procedure and the
nature of the support, the catalysts are arranged in the
following order of activity:
Coxéy–ZrO2 > Coxéy/(ZrO2–Y)
> Coxéy/ZrO2(sol–gel) > Coxéy/[ZrO2(sol–gel)–TsVN]
> Coxéy/ZrO2(reagent grade) > Coxéy/HTsVN
> Coxéy/Al2O3.
This order is consistent with the order of activity of
pure supports; that is, the supports have a determining
effect on the catalytic properties of supported cobalt
oxides rather than serve as inert materials.
KINETICS AND CATALYSIS Vol. 48 No. 3 2007

426
KANTSEROVA, ORLIK
Table 7. TPR study of the redox properties of supports and
The phase composition of the most active Coxéy–
ZrO2 and Coxéy/(ZrO2–Y) samples was also studied
after treatment at 800°C for 1 h. The X-ray diffraction
pattern of the coprecipitated catalyst exhibited diffrac-
tion maximums that corresponded to the cubic and
monoclinic modifications of zirconium dioxide. The
appearance of the latter modification corresponds to the
sequence of ZrO2 phase transformations in the course
of high-temperature treatment [43], and it was accom-
panied by a considerable decrease in dispersion. The
diffraction pattern also exhibited reflections due to a
Co3O4 phase; it is likely that the appearance of this
phase was associated with a decrease in the degree of
dispersion of cobalt oxide because of nanoparticle
agglomeration.
Note that high-temperature heating caused a more
dramatic decrease in Ssp and a decrease in the activity of
Coxéy–ZrO2 and Coxéy/ZrO2 (sol–gel) samples, as
compared with the catalyst based on ZrO2 (reagent
grade). This fact suggests the appearance of an internal
size effect in a zirconium oxide nanosystem [22]: a
decrease in the temperature of agglomeration with
decreasing catalyst particle size. Moreover, the temper-
atures of phase formation and aggregation considerably
decreased with the use of the sol–gel preparation
method [5].
An increase in the treatment temperature of a cata-
lyst based on modified ZrO2 resulted in the growth of
active component particles. This manifested itself in the
X-ray diffraction pattern as a decrease in diffraction
peak widths; the degree of dispersion of the support
remained unchanged. Thus, the modification of zirco-
nia with yttrium oxide resulted in a deceleration of
agglomeration processes and phase transformations;
consequently, the specific surface area decreased to a
lesser extent.
To determine the binding energy of surface oxygen,
we studied catalyst samples using TPR (Table 7).
The TPR profiles of ZrO2 (reagent grade) and ZrO2
(sol–gel) exhibited weakly intense high-temperature
hydrogen consumption regions. However, the ZrO2
catalysts
Hydrogen TPR data:
reduction onset
and peak temperatures, °C
Sample
To
Tmax1
Tmax2
Tmax3
(CoxOy–ZrO2)
123
297
282
376
698
429
–
CoxOy/ZrO2
(reagent grade)
501
CoxOy/ZrO2
(sol–gel)
252
312
387
588
CoxOy/(ZrO2–Y)
CoxOy/Al2O3
≈60
454
642
383
481
725
429 454; >800
555
–
–
–
ZrO2
(reagent grade)
ZrO2 (sol–gel)
(ZrO2–Y)
580
670
–
–
–
≈60 400–450 >750
sponding to Co3O4 phases (L = 13 nm) and a cubic
modification of the support with a particle size of 16–
17 nm. Electron micrographs (Fig. 7) support the nan-
odispersion of Y2O3-modified zirconium dioxide and a
catalyst based on it; an analysis of these micrographs
suggests sufficiently developed surfaces of the support
and the catalyst with a particle size of 6–25 nm. These
data are consistent with XRD data and the values of Ssp.
An analysis of the diffraction pattern of a sample
based on ZrO2 (reagent grade) exhibited the presence of
a Co3O4 phase (L = 12 nm) and a monoclinic modifica-
tion of the support. In a precipitated sample of
Coxéy−ZrO2 and a catalyst based on ZrO2 (sol–gel),
peaks corresponding to only a support phase were
detected, whereas we failed to detect a cobalt-contain-
ing oxide phase. It is likely that the X-ray amorphism
of cobalt oxides is explained by their high degree of dis-
persion because a Co3O4 phase was detected in samples
based on modified and laboratory (reagent grade) zirco-
nium dioxides, which were synthesized at the same (sol–gel) exhibited a decrease in To and THm2ax , as com-
temperature (350–400°C). Moreover, reflections due to
a cubic phase of highly dispersed Co3O4, which is iso-
morphous to a cubic ZrO2 phase, in the diffraction pat-
tern of the Coxéy–ZrO2 sample can be masked by zir-
conia diffraction maximums, which are more intense.
Evidently, the high dispersion of the active component
and zirconia in the Coxéy–ZrO2 and Coxéy/ZrO2 (sol–
gel) catalysts is responsible for their higher activity
than that of samples based on ZrO2 (reagent grade) and
modified zirconium dioxide (Fig. 6).
pared with the corresponding values for ZrO2 (reagent
H2
grade). The shifts of To and peak T
to a lower tem-
max
perature region indicate that the lability (reactivity) of
lattice oxygen in ZrO2 (sol–gel) is higher. The reduc-
tion of the ZrO2-Y support, which has a higher capacity
for oxygen [13], occurred in two regions: a low-temper-
ature region (60–640°C) with THm2ax1 at 400–450°C and
a high-temperature region (>672°C) with TmH2ax2
>
In addition to peaks that correspond to a tetragonal 750°C. This suggests the occurrence of both relatively
modification of ZrO2 and pentasil, the diffraction pat- weakly bound and relatively strongly bound oxygen
tern of the [ZrO2 (sol–gel)–HTsVN] catalyst exhibited species. The TPR data are consistent with the catalytic
a number of reflections that suggested the occurrence of properties of the supports in methane oxidation with the
a Co3O4 phase (L = 15 nm).
labile oxygen of zirconium dioxide.
KINETICS AND CATALYSIS Vol. 48 No. 3 2007

EFFECT OF A STRUCTURE–SIZE FACTOR ON THE CATALYTIC PROPERTIES
427
Table 8. TPAD study of the surface acid properties of supports and catalysts
NH3 desorption
peak temperatures, °C
Acid site concentrations, (mmol NH3)/m2
Sample
KN1 H × 103,
KN2 H × 103,
3
3
ΣKNH × 103
3
Tmax1
Tmax2
170–250°C
250–450°C
(CoxOy–ZrO2)
200
170
180
250
210
200
180
200
210
220
280
250
270
395
450
450
–
0.28
0.52
0.72
0.28
3.8
1.6
–
0.45
0.51
0.98
0.12
2.8
1.0
–
0.73
1.03
1.7
CoxOy/ZrO2 (reagent grade)
CoxOy/ZrO2 (sol–gel)
CoxOy/(ZrO2–Y)
0.4
CoxOy/HTsVN
6.6
CoxOy/[ZrO2(sol–gel)–HTsVN]
CoxOy/Al2O3
2.6
2.2
ZrO2 (reagent grade)
ZrO2 (sol–gel)
–
–
–
1.39
2.8
–
–
–
(ZrO2–Y)
–
–
–
0.36
In the presence of the active oxide component activity and the concentration and strength of acid sites
Co3O4, which can form nonstoichiometric phases with
oxygen, the reducibility of catalysts increased, as com-
pared with individual supports. This is reflected in a
(Table 8). Nevertheless, the most active catalysts based
on zirconium dioxide were characterized by the pres-
ence of stronger acid sites at an insignificant total con-
centration of acid sites. The introduction of zeolite into
the composition of a ZrO2 (sol–gel) support increased
the strength and concentration of acid sites in the
Coxéy/[ZrO2 (sol–gel)–HTsVN] catalyst (Table 8,
Fig. 8). It is likely that the presence of stronger acid
sites on the surface of the Coxéy/[ZrO2 (sol–gel)–
HTsVN] catalyst was responsible for a decrease in the
temperature of the onset of reaction on this sample, as
compared with that on Coxéy/ZrO2 (sol–gel).
decrease of To and THm2ax and in an increase in the inten-
sity of hydrogen consumption in catalysts. The Coxéy–
ZrO2 and Coxéy/ZrO2 (sol–gel) samples exhibited a
shift of To and THm2ax to the region of lower temperatures,
as compared with the Coxéy/ZrO2 (reagent grade) sam-
ple. This can be due to the higher dispersion of both
cobalt oxides and zirconium dioxide in these samples.
It is likely that the presence of nonreducible zeolite as
a constituent of the [ZrO2 (sol–gel)–HTsVN] support
was responsible for higher temperatures at which 80–
100% conversion of methane was reached on the binary
support, as compared with that for the sample based on
ZrO2 (sol–gel). The TPR profile of the Coxéy/Al2O3
Intensity
2
sample was characterized by high values of To and TmH2ax
and a less intense region of hydrogen consumption than
those for catalysts based on zirconium dioxide. This
difference in reducibility can be explained by the fact
that, unlike ZrO2, γ-Al2O3 and Co3O4 crystallize in a
spinel-type structure, which is responsible for the for-
mation of an isomorphous solid solution of Co3O4 with
Al2O3 or spinel with CoAl2O4. This resulted in a
decrease in the amount of weakly bound oxygen, which
is necessary for the occurrence of a reaction. Thus, the
results obtained using TPR demonstrated that the
amount of labile oxygen increased in an order that is
consistent with the order of activity of the test catalysts
and individual supports.
1
100
200
300
400
500
600
Temperature, °C
Fig. 8. Spectra of TPAD from the surface of cobalt catalysts
supported on (1) ZrO (sol–gel) and (2) [ZrO (sol–gel)–
An analysis of the surface acid properties of the test
catalysts did not exhibit a simple correlation between
2
2
HTsVN].
KINETICS AND CATALYSIS Vol. 48 No. 3 2007

428
KANTSEROVA, ORLIK
5. Rodionov, Yu.M., Slyusarenko, E.M., and Lunin, V.V.,
Thus, we found that the nature of the support and the
Usp. Khim., 1996, vol. 65, no. 9, p. 865.
procedures used for the preparation of both the support
and the catalyst affected the activity of cobalt–zirco-
nium nanosystems. We prepared low-temperature cata-
lysts for deep methane oxidation under conditions
developed for the formation of the nanosized phases of
6. Stinson, S.C., Chem. Eng. News, 1982, vol. 60, no. 32,
p. 26.
7. Villa, R., Cristiani, C., and Groppi, G., J. Mol. Catal. A,
2003, vol. 204, p. 637.
8. Sitidze, Yu. and Sato, Kh., Ferrity (Ferrites), Moscow:
a support (LZrO = 12–13 nm) and an active component
2
Mir, 1964.
(LCo O ≤ 3 nm). In terms of activity, these catalysts are
9. Ivanova, A.S., Fedotov, M.A, and Litvak, G.S., Kinet.
Katal., 2002, vol. 43, no. 1, p. 150 [Kinet. Catal. (Engl.
Transl.), vol. 43, no. 1, p. 139].
10. Ukrainian Patent 71394 A, 2004.
11. Ukrainian Patent 2001075296, 2002.
12. Kharlanov, A.N., Lunina, E.V., and Lunin, V.V., Zh. Fiz.
Khim., 1999, vol. 73, no. 5, p. 898 [Russ. J. Phys. Chem.
(Engl. Transl.), vol. 73, no. 5, p. 790].
13. Labaki, M., Lamonier, J.-F, Siffert, S., Zhilinskaua, E.A.,
and Aboukais, A., Tezisy dokladov VI Rossiiskoi konfer-
entsii po mekhanizmam kataliticheskikh reaktsii (Proc.
VI Russian Conf. on Mechanisms of Catalytic Reac-
tions), Moscow, 2002, p. 51.
14. Choudhary, V.R., Banerjee, S., and Uphade, B.S., Appl.
Catal., A, 2000, vol. 197, no. 2, P. L183.
15. Mironyuk, T.V., Orlik, S.N., and Struzhko, V.L., Teor.
Eksp. Khim., 2001, vol. 37, no. 4, p. 256.
x
y
comparable with the well-known catalysts based on zir-
conium dioxide and with catalytic systems containing
noble metals [44–46]. We found the manifestation of an
internal size effect in the zirconium oxide nanosystem:
a decrease in the temperature of agglomeration with
decreasing catalyst particle size.
We found that the most active catalyst among sup-
ported ferrite and cobalt–zirconium nanosystems was
characterized by the greatest amount and the highest
reactivity of oxygen. The presence of strong acid sites
on the catalyst surface was responsible for the lower
temperature of the onset of the deep methane oxidation
reaction. This is consistent with the current concepts of
the mechanism of deep methane oxidation on oxide cat-
alysts.
The results of the study of complex oxide nanocom-
posites in deep methane oxidation suggest that the
problem of developing efficient catalysts that do not
contain noble metals can be successfully solved based
on the effect of structure–size (especially, nanosize)
factors on the activity and thermal stability of catalysts
and a relationship between the functional (redox and
acid) and catalytic properties of the above oxide com-
positions. The nanosized supported zirconia- and alu-
mina-based catalytic systems and spinel-structure
MFe2O4 nanocomposites (M = Mn, Co, or Ni) with con-
trollable structure–size characteristics, redox proper-
ties, and acid properties form the basis for the develop-
ment of commercial catalysts for the purification of
methane-containing vent gas emissions and the com-
bustion of hydrocarbon gas fuel in power plants.
16. Hammond, C., The Basics of Crystallography and Dif-
fraction, Oxford: Oxford Univ. Press, 1997, p. 145.
17. Dzis’ko, V.A., Tarasova, D.V., and Karnaukhov, A.P.,
Fiziko-khimicheskie osnovy sinteza okisnykh kataliza-
torov (Physicochemical Foundations of the Synthesis of
Oxide Catalysts), Novosibirsk: Nauka, 1978, p. 384.
18. Suzdalev, I.P. and Suzdalev, P.I., Usp. Khim., 2001,
vol. 70, no. 3, p. 203.
19. Murrell, L.L., Catal. Today, 1997, vol. 35, no. 3, p. 225.
20. Moroz, E.M., Zh. Prikl. Khim., 1996, vol. 69, no. 11,
p. 1764.
21. Stohmeir, B.R. and Hercules, D.M., J. Phys. Chem.,
1984, vol. 88, no. 21, p. 4922.
22. Uvarov, N.F. and Boldyrev, V.V., Usp. Khim., 2001,
vol. 79, no. 4, p. 307.
23. Choundary, T.V., Benerijce, S., and Choundary, V.R.,
Appl. Catal., A, 2002, vol. 234, nos. 1–2, p. 1.
24. Kapteijn, F., Langeveld, A.D., Moulijn, J.A., Andreini, A.,
Vuurman, M.A., Turek, A.M., Jehng, J.M., and Wachs,
I.E., J. Catal., 1994, vol. 150, no. 1, p. 94.
ACKNOWLEDGMENTS
This study was performed within the framework of
a project of the program “Nanostructured Systems,
Nanomaterials, and Nanotechnologies” of the National
Academy of Sciences of Ukraine.
25. Sergeev, B.G., Ross. Khim. Zh., 2002, vol. 46, no. 5,
p. 22.
26. Tsikoza, L.T., Ismagilov, Z.R., Shkrabina, R.A., Ko-
ryabkina, N.A., Ushakov, V.A., Kuznetsov, V.V., and Ov-
syannikova, I.A., Zakonomernosti glubokogo okisleniya
veshchestv na tverdykh katalizatorakh (Deep Oxidation
of Substances on Solid Catalysts), Novosibirsk: Inst.
Kataliza, 2000.
27. Artizzu, P., Garbowski, E., Primet, M., Brulle, Y., and
Saint-Just, J., Catal. Today, 1999, vol. 47, nos. 1–4,
p. 83.
REFERENCES
1. Hicks, R.F., Qi, H., Young, M.L., and Lee, R.G.,
J. Catal., 1990, vol. 122, no. 2, p. 280.
2. Briot, P., Denis, A.A., and Primet, J.M., Appl. Catal.,
1990, vol. 59, no. 1, p. 141.
3. Ukrainian Patent 2003065833, 2004.
28. Garbowski, E., Guenin, M., Marion, M.C., and Primet, M.,
Appl. Catal., 1990, vol. 64, p. 209.
4. Kosmambetova, G.R., Kantserova, M.R., Orlik, S.N.,
and Kazimirov, V.P., Teor. Eksp. Khim., 2003, vol. 39, 29. Kurina, L.N. and Vodyankin, A.Yu., Zh. Prikl. Khim.,
no. 4, p. 241.
1997, vol. 70, no. 8, p. 1389.
KINETICS AND CATALYSIS Vol. 48 No. 3 2007

EFFECT OF A STRUCTURE–SIZE FACTOR ON THE CATALYTIC PROPERTIES
429
30. Zarur, A.J. and Ying, J.Y., Nature, 2000, vol. 403, 38. Yan, J., Kung, M.C., Sachtler, W.M.H., and Kung, M.C.,
no. 6765, p. 65.
J. Catal., 1998, vol. 175, no. 2, p. 294.
31. Gavrilenko, K.S., Mironyuk, T.V., Il’in, V.G., Orlik, S.N.,
and Pavlishchuk, V.V., Teor. Eksp. Khim., 2002, vol. 38,
no. 2, p. 110.
39. RF Patent 5035655/04 RF, 1995.
40. USSR Inventor’s Certificate no. 4 323 850, 1993.
41. Comino, G., Gervasini, A., and Ragaini, V., Catal. Lett.,
32. Tret’yakov, Yu.D., Tverdofaznye reaktsii (Solid-Phase
1997, vol. 48, nos. 1–2, p. 39.
Reactions), Moscow: Khimiya, 1978.
42. Li, Y. and Armor, J.N., Appl. Catal., B, 1994, vol. 3,
33. Tsyrul’nikov, P.G., Shitova, N.B., and Kudrya, E.N., and
Savel’eva, E.G., Zakonomernosti glubokogo okisleniya
organicheskikh veshchestv (Total Oxidation of Organic
Compounds), Novosibirsk: Inst. Kataliza, 2002.
no. 4, p. 275.
43. Ivanova, A.S., Zh. Fiz. Khim., 1995, vol. 69, no. 11,
p. 1790.
34. Margolis, L.Ya. and Korchak, V.N., Usp. Khim., 1998, 44. Choudhary, V.R., Uphade, B.S., and Pataskar, S.G.,
vol. 67, no. 12, p. 1175.
Appl. Catal., A, 2002, vol. 227, nos. 1–2, p. 29.
35. Burch, R., Crittle, D.J., and Hayes, M.J., Catal. Today,
45. Xiao, T., Ji, S., Wang, H., Coleman, K.S., and Green, M.L.,
1999, vol. 47, nos. 1–4, p. 229.
J. Mol. Catal. A, 2001, vol. 175, nos. 1–2, p. 111.
36. Choudhary, V.R. and Rane, V.H., J. Catal., 1991,
46. Kantserova, M.R., Kosmambetova, G.R., and Orlik, S.N.,
Tezisy III Ukrainskoi nauchno-tekhnicheskoi konferen-
tsii po katalizu (Proc. III Scientific and Engineering
Conf. on Catalysis), Slavyanogorsk, 2002, p. 115.
vol. 130, no. 2, p. 411.
37. Wu, H.C., Liu, L.C., and Yang, S.M., Appl. Catal., A,
2001, vol. 211, no. 2, p. 159.
KINETICS AND CATALYSIS Vol. 48 No. 3 2007