Kinetics and Catalysis, Vol. 44, No. 4, 2003, pp. 536–546. Translated from Kinetika i Kataliz, Vol.

44, No. 4, 2003, pp. 587–598.

Original Russian Text Copyright © 2003 by Malakhova, Ermolaev, Danilova, Paukshtis, Zolotarskii.

Kinetics of Free Radical Generation

in the Catalytic Oxidation of Methanol
I. V. Malakhova, V. K. Ermolaev†, I. G. Danilova, E. A. Paukshtis, and I. A. Zolotarskii
Boreskov Institute of Catalysis, Siberian Division, Russian Academy of Sciences, Novosibirsk, 630090 Russia
Received May 28, 2002

Abstract—The formation of free radicals over the surface of platinum-containing catalysts in the methanol oxi-
dation reaction depending on the temperature, the composition of the reaction mixture, and the procedure used
for introducing platinum was studied by the matrix isolation method technique. The nature and transformations
of surface intermediates depending on the temperature and the presence of oxygen in the gas phase were studied
by Fourier transform IR spectroscopy. The main surface intermediate was the methoxy group. The following
three types of these groups were stabilized in alumina-based catalysts: (I) CH3O–Aloct (νs(ë–ç) = 2806 cm–1),
(II) CH3O–Altetr (νs(ë–ç) = 2825 cm–1), and (III) CH3O < (Al)2 (νS(ë–ç) = 2845 cm–1, δ‡s(ë–ç) = 1460 cm–1,
δs(ë–ç) = 1440 cm–1, r|| (ëç3) = 1185 cm–1, and ν(ë–é) = 1095 cm–1). At the same time, isolated methoxy
groups (νas(ë–ç) = 2997 cm–1, νas(ë–ç) = 2959 cm–1, νs(ë–ç) = 2857 cm–1, and δ(ëç3) = 1450 cm–1) and
hydrogen-bonded groups (ν(é–ç) = 3400–3550 cm–1), which resulted from chemisorption at siloxane bridges,
. .
were stabilized in silica gel–based catalysts. It was found that CH3 O and CH3O O radicals were formed only
over the surfaces of pure supports (SiO2 and Al2O3) and their mechanical mixtures with platinum. The total con-
centration of radicals was described by an extremal function of the composition of reactants, whereas the rela-
tive concentration depends on the nature of the support. This is conceivably due to the effect of coordinatively
unsaturated cations of the support, which are formed by dehydroxylation in the course of catalyst pretreatment.
An increase in the rate of formation of gas-phase radicals on mixed catalysts was explained by special proper-
ties of the platinum/support interface region, at which surface intermediates were formed in superequilibrium
concentrations under reaction conditions.

† INTRODUCTION .
(R O 2 ) were frozen in these reactions. Because of their
The high-temperature transformations of hydrocar- low chemical activity, these radicals were transferred
bons often include homogeneous reactions initiated by without considerable concentration changes from the
free radicals, which are desorbed from the surface of a reaction zone to a condensate-freezing unit. In this
catalyst. In the mechanistic studies of heterogeneous case, it is assumed a priori that the frozen radical can
and homogeneous reactions, it is of importance to cor- be identified by the EPR spectrum.
relate the formation of surface intermediates with the However, when a reaction results in the formation of
role of these intermediates in the formation of free rad- a few radicals with close parameters of the EPR spectra
icals and to determine the contribution of these free rad- . . .
icals to the overall reaction mechanism. (for example, ç O and ç O 2 [5, 6] or CH3 O and
The freezing of radicals in the matrix of products of
.
CH3 O 2 [7]), the spectrum of the condensate is a super-
a test reaction at 77 K followed by the recording of the position of the spectra of individual radicals. Because
EPR spectra of the condensate (the matrix isolation of this, their relative concentrations are difficult to eval-
method technique) is an instrumental technique for uate. To solve this problem, a special method was pro-
studying the composition of radicals in a gas phase and posed for determining the nature of frozen radicals
the kinetics of radical generation and decay over wide . .
temperature, pressure, and concentration ranges. This (R O 2 or R O ). This method is based on the assump-
technique was proposed by V.N. Panfilov and devel- tion that methoxy and methylperoxy radicals exhibit
oped under the supervision of A.B. Nalbandyan at the different stability to photolysis with the full light of
Institute of Chemical Physics in the Armenian Soviet high-pressure mercury lamps [7].
Socialist Republic [1, 2]. Matrix isolation method has With the use of photolysis for the identification of
received wide acceptance in the analysis of mecha- . .
nisms of the gas-phase oxidation of hydrocarbons from radicals, we found that CH3 O and CH3 O 2 radicals
different classes [2–5]. Presumably, peroxide radicals were formed in the methanol oxidation reaction on a
catalyst that was a mechanical mixture of platinum
† Deceased.
black and silica gel at a 1 : 1 ratio between the reactants

0023-1584/03/4404-0536$25.00 © 2003 MAIK “Nauka /Interperiodica”

 KINETICS OF FREE RADICAL GENERATION IN THE CATALYTIC OXIDATION 537

and at T = 770–920 K with an activation energy of a standard procedure [12, 13] at the Omsk Branch of the
7 kcal/mol [7]. Boreskov Institute of Catalysis.
Nonadditive effects of an increase in the catalytic Commercial A-1 alumina (Ssp = 186 m2/g; Vpore =
activity of Pt(Pd) + oxide mechanical mixtures were 0.51 cm3/g) and KSKG silica gel from the Ryazan
reported in publications on the catalytic oxidation of ç2 refinery (Ssp = 280 m2/g; Vpore = 0.84 cm3/g) were used
and light hydrocarbons [9–11]. These effects were as catalyst supports. The silica gel was prewashed with
explained assuming the development of a radical chain concentrated nitric acid, ground by attrition, and pel-
mechanism. Thus, it was of interest to us to perform a leted at a pressure of ~50 atm. The above SiO2 and
comparative study of the generation of free radicals in Al2O3 were used for preparing mixed catalysts.
catalytic systems with active components introduced Platinum black was prepared according to Zelinskii
by different procedures in order to understand the [14]. A solution of H2PtCl6 (~100 ml) was treated with
mechanism of radical generation and the overall reac- a formaldehyde solution (10 ml) at 288 K; after cooling
tion mechanism. the mixture in an ice bath to 273 K, a KOH solution
To determine the effects of the active component, (~20 g in 50 ml of ç2é) was added in small portions.
the procedure of introducing it, and the nature of the Platinum black began to precipitate at pH 8–9. The pre-
support on the generation of radicals in methanol oxi- cipitate was heated to 323–333 K and settled at room
dation, we performed experiments in the order pure temperature for several hours. The platinum black was
support active component for the most commonly filtered off using a glass filter and washed with water to
used supports (SiO2 and γ-Al2O3), their mechanical remove the alkali. The resulting black was dried at 303–
mixtures with platinum, and supported platinum cata- 393 K for 6 h. After calcination (at 773 K for 6 h),
lysts. The methanol oxidation reaction was studied for Ssp was ~6 m2/g.
several reasons. On the one hand, methanol is one of the Mixed catalysts were mechanical mixtures of a sup-
main intermediates in the oxidation of methane and port and platinum black, which were ground by attri-
light hydrocarbons. On the other hand, methanol tion and pelleted. The weight concentrations of Pt were
belongs to the main products of organic synthesis with equal to that in the impregnated catalysts (3.1 and
continually expanding areas of application. The main 2.6 wt % Pt in InCatPt-1 and InCatPt-2, respectively).
areas of methanol application are the production of
formaldehyde and the use as an effective component of To retain the geometry of samples, all the catalysts
alcohol fuels. However, although many publications were ground by attrition in an agate mortar and pelleted
were devoted to the adsorption, decomposition, and at a pressure of ~50 atm. The average dimensions of
oxidation of methanol, the reaction mechanism is still working samples were 5 × 5 × 2 mm.
understood incompletely. Moreover, Krishtopa et al.
[3] found that the rates of radical generation in the oxi- Matrix Isolation Method with the Recording
dation of hydrocarbons and their oxygenates on oxide of EPR Spectra
catalysts in the order hydrocarbon alcohol
aldehyde carboxylic acid are in a ratio of 1 : 10 : The setup was described elsewhere [7]. Absolute
5 : 5. This suggests that the contribution of a homoge- methanol triply degassed by recrystallization and evac-
neous component to the oxidation of alcohols is consid- uation before each experiment was supplied from an
erable. evaporator to a reactor equipped with an electric fur-
In this work, we studied the reactions of surface nace. The catalyst was put over the grid of the reactor
intermediates formed under conditions of methanol (dint = 16 mm) as a one-grain layer. The temperature in
oxidation with the use of Fourier transform IR spectros- the catalytic bed zone was maintained to within 0.5 K.
copy. The flow rates of methanol vapor and oxygen, which
was supplied from a gas cylinder after the removal of
water vapor impurities, were controlled using fine con-
EXPERIMENTAL trol valves. The resulting radicals together with reaction
Radical products in the methanol oxidation reaction products and unreacted methanol were frozen at 77 K
at 720–920 K were studied by matrix isolation method in the side arm of a Dewar vessel, which was placed in
in the following series of catalysts: pure support (SiO2 the resonator of an EPR spectrometer and connected to
a flow vacuum system. The photolysis of radicals in the
or γ-Al2O3) mechanical mixture (Pt + support)
matrix of reaction products was performed with full
impregnated catalyst (Pt/SiO2 or Pt/Al2O3) pure Pt. light from a DRSh-500 lamp, which was focused by a
quartz lens to the end of the side arm of the Dewar ves-
Catalysts sel. Previously [4], it was found that, under the experi-
mental conditions specified, the height of a condensate
Pure Pt0 was a metal film deposited from H2PtCl6 on layer was 2–3 mm from the bottom of the arm, and the
the walls of a quartz reactor. condensate distribution did not increase errors in the
The impregnated catalysts Pt/SiO2 (InCatPt-1) and determination of the concentration of paramagnetic
Pt/Al2O3 (InCatPt-2) were prepared in accordance with centers.

KINETICS AND CATALYSIS Vol. 44 No. 4 2003

538 MALAKHOVA et al.

w ×10–11, radical/s case, oxygen (1–2 Torr) was added to the cell before
5 measuring the spectra at 523 K. The treatment of exper-
imental data included the decomposition of the result-
ing absorption spectra into individual Gaussian compo-
4 nents.

3 1 RESULTS AND DISCUSSION

Matrix-Isolation Study of Radical Generation
2 in the Gas Phase
With the use of photolysis for the identification of
1
. .
radicals, we found that CH3 O and CH3 O 2 radicals
2
were formed in the methanol oxidation reaction on sil-
0 ica gel at T = 720–920 K (the parameters of the EPR
0.1 1 10 spectra of these radicals g|| = 2.03, g⊥ = 2.008, and A⊥ =
CH3OH/O2 8.7 G were consistent with those reported previously
[7]). The maximum rate of radical generation on silica
Fig. 1. Dependence of the rates of radical generation in the gel at the reactant ratio methanol/oxygen = 1 : 1 was
methanol oxidation reaction (Texpt = 773 K) on the compo-
lower than that on a mechanical mixture of Pt + SiO2 by
sition of the reaction mixture for pure supports: (1) γ-Al2O3
a factor of 25 (Figs. 1, 2), and the apparent activation
(Ttr = 773 K) and (2) SiO2 (Ttr = 940 K). . .
energy of the overall formation of CH3 O and CH3 O 2
radicals on SiO2 was 38.4 kcal/mol. The value of the
The experiments were performed at methanol/oxy- activation energy of radical generation on the support is
gen ratios of 0.1 : 10 under the following pretreatment almost equal to the activation energy of methanol oxi-
conditions: Ttr = 943 K; oxygen pressure, ~2000 Pa; dation on silica gel (40 kcal/mol) determined in [17].
and treatment time ttr = 1 h. In some experiments, the This agreement in activation energies suggests that the
treatment temperature was changed for catalysts based formation of the adsorption centers of methanol mole-
on Al2O3 as specified below. cules is a rate-limiting step in the formation of radicals
All the experiments were performed in the kinetic over the surface of silica gel. According to IR-spectro-
region when radical reactions did not occur in a gas scopic data [17, 18], the adsorption of methanol
phase (working pressure of 3.7 Pa). An upper pressure resulted in the formation of SiOCH3 regardless of the
limit of the technique was ~13 Pa, which corresponds structure of the adsorption center; only the mechanism
to the breakdown of a diffusion pump. The sensitivity of adsorption was changed. Assuming that surface
of the EPR spectrometer was ~5 × 1010 paramagnetic intermediates are responsible for the yield of radicals to
centers under standard measurement conditions. a gas phase, the heterogeneous process of radical gen-
eration can be written as
IR Spectroscopic Experiments Si O Si + CH3OH SiOCH3 + SiOH,
The IR spectra of samples in the region 400–5000 cm–1 .
2 SiOCH3 + O* 2CH3O(gas) + Si O Si ,
were measured on a Shimadzu FTIR-8300 spectrome-
ter with a resolution of 4 cm–1 in a cell suitable for high- where O* is the active surface oxygen, which resulted
temperature spectroscopic measurements (up to 773 K). from the pretreatment with oxygen.
The catalyst samples were pressed in pellets 10–15 mg/cm2
in thickness before measuring the spectra. Next, they The situation changed when Pt0 (metal film depos-
were placed in the cell (V ≈ 500 cm3) and pretreated in ited from H2PtCl6) or Pt/SiO2 and Pt/Al2O3 (supported
oxygen and then in a vacuum at 773 K with the freezing by impregnation) were used as catalysts. In the case of
of volatile oxidation products for 1 h. Surface methoxy pure platinum, radicals were not detected (the rate of
groups were obtained by the treatment of a catalyst formation was Ⰶ108 radical/s) at all the tested reactant
sample with the vapor of absolute methanol (~6 Torr) at compositions and temperatures. In this case, the pres-
523 K for 1 h followed by evacuation at the specified sure in the system periodically changed at the ratio
temperature for 1 h (to remove alcohol molecules åÂéç/é2 = 1 : 1 and 873 and 893 K. It is evident that
bound to the surface by hydrogen or coordination these changes were due to the oxidation of the carbon
bonds). The resulting methoxy groups were decom- oxide formed on platinum, as described previously
posed by holding the samples at 573, 673, and 773 K [19]. The Pt/SiO2 catalyst exhibited an analogous
for 30 min. The spectra were measured at 7-min inter- behavior. However, in this case, along with oscillations,
vals. The experiments were performed both in the the pressure gradually increased to values that did not
absence and in the presence of oxygen. In the latter allowed us to perform experiments using this proce-

KINETICS AND CATALYSIS Vol. 44 No. 4 2003

 KINETICS OF FREE RADICAL GENERATION IN THE CATALYTIC OXIDATION 539

dure. According to EXAFS data [20], deposited plati- w × 10–12, radical/s

num was stabilized as Pt0 in the lattice of silica gel. At 2.5
the same time, according to IR-spectroscopic data [18],
the SiHx structures (x = 1, 2) and the partial oxidation
products of methanol (CO and hydrogen) were formed 2.0
1 2
upon the degradation of methoxy groups on the surface
of silica gel at elevated temperatures (873 K). Thus, we 1.5
believe that in the case of the Pt/SiO2 catalyst the fol-
lowing reactions of methanol oxidation take place:
1.0
on platinum
CH3OH + 1/2O2 CO + H2O + H2, 0.5
CO + 1/2O2 CO2,
on silica gel 0
0.1 1 10
Si–OCH3 CO + H2 + Si–Hx . CH3OH/O2
In this case, a temperature shift of the oscillation
Fig. 2. Dependence of the rates of radical generation in the
regime, as compared with published data [19], was due methanol oxidation reaction (Texpt = 773 K) on the compo-
to the effect of the support, which, in turn, can result in sition of the reaction mixture for mixed catalysts: (1) Pt +
the formation of CO and hydrogen; this explains the Al2O3 (Ttr = 773 K) and (2) Pt + SiO2 (Ttr = 940 K).
increase in the pressure in the system.
Radicals were also not detected over the Pt/Al2O3
catalyst (the rate of formation was Ⰶ108 radical/s). The sponding main values were equal to g|| = 2.03, g⊥ =
pressure in the system increased from 6 × 10–3 to 10−2 Torr 2.008, and A⊥ = 8.7 G; we attributed them to the sum of
. .
in the first 20 min of reaction; however, it subsequently CH3 O and CH3 O 2 radicals.
stabilized and remained constant (~10–2 Torr) up to the
end of the experiment. According to EXAFS data [20] In repeated experiments with the supported
for supported Pt/Al2O3 catalysts, three types of plati- Pt/Al2O3 sample, whose treatment temperature was
num compounds can be recognized: (1) platinum metal changed to 773 K, radicals were not detected at all the
clusters, in which Pt–Pt interatomic distances are test reactant compositions. Radicals were not formed in
noticeably shorter than that in the bulk metal; (2) PtOx the case of the impregnated catalyst; this fact allowed
surface oxides; and (3) isolated platinum ions stabilized us to make an assumption about the nature of active
at tetrahedral cationic vacancies in γ-Al2O3. Because of centers responsible for radical generation on the sur-
the complicated overall behavior of active component face. In a study of the acceptor properties of an alumina
centers, their contributions to the reaction cannot be surface by the paramagnetic probe method [22], it was
distinguished under the conditions of our procedure. found that the formation of Al3+ ions in a tetrahedral
Therefore, we performed experiments for studying the environment in the structure of the oxide and coordina-
reaction on a pure support. tively unsaturated ions on the surface are characteristic
of pure alumina. The introduction of platinum dramati-
Early attempts to detect radicals on pure γ-Al2O3 cally decreased the concentration of coordinatively
were unsuccessful. It is well known [21] that acid sites unsaturated aluminum ions; in this case, the number of
of different strengths are formed on the surface dehy- “disappeared” centers was commensurable with the
droxylation of γ-Al2O3. The dependence of the concen- number of supported platinum atoms. According to
tration of acid sites on the temperature of calcination Lunina [22], a decrease in the acceptor properties of
exhibited two maximums at 773 and 1073 K alumina is associated with the injection of electrons
[21 (Fig. 7.5, p. 197)]. Because the concentrations of from the metal to the support. Because of electron
acid sites were minimum at 940 K (standard sample transfer, the alumina surface adjacent to a metal crystal-
treatment conditions), we chose another temperature of lite acquires a negative charge. In this case, the depth of
773 K as the pretreatment temperature. The other con- acceptor traps (the strength of acceptor sites) near crys-
ditions remained unchanged. tallites decreases so that they cease to be acceptor sites.
As a result of a change in treatment conditions, we This approach can explain an apparent contradiction
detected radicals on the pure support (γ-Al2O3) and its between our results and data obtained by Pak [23], who
mechanical mixture at reaction temperatures of 723, reported the formation of methylperoxy radicals in the
753, and 773 K. The EPR spectra of the radicals were oxidation of methanol on a Pt/Al2O3 (0.64 wt % Pt)
analogous to the spectra of radicals formed in the meth- supported catalyst. According to Lunina [22], the con-
anol oxidation reaction on silica gel; they appeared as centration of acceptor sites at this platinum content
an asymmetric line typical of paramagnetic species decreased by a factor of ~2.3 as compared with the con-
with an axial anisotropy of the g-tensor. The corre- centration of acceptor sites on pure alumina. The ratio

KINETICS AND CATALYSIS Vol. 44 No. 4 2003

540 MALAKHOVA et al.

(b)

2845
2825
1095

(‡)

Absorption
Absorption

2805
1

3002
1575
1595
1460
1185

1445 2
1375
1390

1 3
1315

3015
4
4
~
~

1200 1400 1600 1700 2800 2900 3000

ν, cm–1 ν, cm–1

Fig. 3. Changes in the IR spectra of a sample of 2.6 wt % Pt + γ-Al2O3 (treated with methanol vapor at 523 K with the subsequent
evacuation at T = 523 K) (a) at 1050–1700 cm–1 and (b) in the region of CH vibrations depending on the temperature of heating
(K): (1) 523, (2) 573, (3) 673, and (4) 773.

between the rates of radical generation found in this energy of radical generation in our experiments should
work for the pure support and reported by Pak [23] for be equal to Eapp for platinum metal. However, the acti-
Pt/Al2O3 (0.64 wt % Pt) under identical conditions was .
vation energy of CH3 O formation found for a Pt +
the same (a factor of 2.32). It is likely that, as the weight
concentration of Pt was increased up to 2.6 wt % in our SiO2 catalyst [7] is lower than the value reported by
experiments, platinum blocked centers at which the for- Zum Mallen and Schmidt [19] by a factor of 3. In our
mation of surface intermediates responsible for the gen- opinion, the platinum–support interface plays a priority
eration of gas-phase radicals took place. role in the case of mixed catalysts. This difference in
activation energies can be explained on the assumption
Thus, in the order support active component, that the desorption of radicals from the catalyst surface
we detected radical products only in the case of pure is a rate-limiting step in radical generation. Taking into
supports and their mechanical mixtures with platinum. account that the catalyst was a mechanical mixture of a
Figure 1 demonstrates the dependence of the rates of support and platinum in our experiments, we believe
radical generation on the methanol/oxygen ratio for dif- that a methoxy radical was formed on the platinum.
ferent supports. Next, it migrated to the platinum–support interface,
The rates of radical generation on pure alumina .
where the desorption of CH3 O is energetically
were higher than those on pure SiO2 by almost one
order of magnitude, and they insignificantly depended favored, because of a spillover. The effect of pretreat-
on temperature. ment on the rate of radical generation is indirect evi-
dence for the above consideration. Thus, the pretreat-
Figure 2 demonstrates the effect of the composition ment in the presence of oxygen increased the rate of
of a reaction mixture on the rate of radical generation for radical generation by a factor of 2–2.5 for both sup-
mixed catalysts. The maximal rates of radical generation ports.
upon the addition of platinum to SiO2 or Al2O3 increased
in both supports (by a factor of 6 or 25, respectively). We obtained additional corroboration of the hypoth-
According to published data [19], in the oxidation of esis that the support plays a crucial role in the genera-
methanol on platinum foil, hydroxyl and methoxy radi- tion of radicals by analysis of the composition of result-
cals were formed with E‡ = 20 kcal/mol. One might ing radicals. We found that the nature of radicals des-
assume that the increase in the rate of radical generation orbed from the surface of a pure support or its
resulted from their additional desorption from the sur- mechanical mixture with platinum remained practically
face of platinum. In this case, the apparent activation unchanged. At the same time, the composition of radi-

KINETICS AND CATALYSIS Vol. 44 No. 4 2003

 KINETICS OF FREE RADICAL GENERATION IN THE CATALYTIC OXIDATION 541

cals strongly depended on the initial reactant concen- 250oC 300oC 400oC 500oC 250oC 300oC 400oC 500oC
. .
trations. Thus, ~80% CH3 O and 20% CH3 O 2 radicals
(‡) (b)
were formed on SiO2 and Pt + SiO2 catalysts at the com-
1
position methanol/oxygen = 1 : 1. As the concentration 2
of oxygen was increased, the concentration of methylp- 6 3

∆I, arb. units

eroxy radicals somewhat increased, whereas the meth- 4
ylperoxy radical was not detected in reducing mixtures 5
(where the alcohol concentration was higher than the
.
oxygen concentration); however, the ëç O radical 3
with proton splitting of ~127 G appeared. The methylp-
eroxy radicals were practically not detected in catalysts
based on γ-Al2O3, whereas the formyl radical was
formed in an excess of methanol. The table summarizes
changes in the composition of frozen radicals depend- 0 30 60 90 30 60 90
ing on the alcohol/oxygen ratio in the initial reaction Time, min
mixture.1
Fig. 4. Changes in the absorption band intensities of surface
groups and gas-phase methane on heating a γ-Al2O3 sam-
In Situ IR Spectroscopic Study of Surface Groups ple: (a) without oxygen and (b) on the addition of oxygen
(1 Torr) at the beginning of heating. (1) νs(C–H) (CH3O <
The treatment of samples based on Al2O3 with (Al)2), 2845 cm–1; (2) νs(C–H) (CH3O – Aloct), 2805 cm–1;
methanol at 523 K followed by evacuation at the above
(3) νs(C–H) (CH3O – Altetr), 2825 cm–1; (4) νs(C–é)
temperature resulted in the formation of three types of
surface methoxy groups (two terminal and one bridging (çCéé–), 1390 cm–1; and (5) νs(C–H) (ëç4), 3015 cm–1.
group). These groups can be identified by the following The temperatures of heating over the corresponding time
intervals are specified at the top of the figure.
absorption bands observed in the IR spectra (Fig. 3,
spectrum 1): (I) CH3O–Aloct (νs(ë–ç) = 2806 cm–1),
(II) CH3O–Altetr (νs(ë–ç) = 2825 cm–1), and (III) 250oC 300oC 400oC 500oC 250oC 300oC 400oC 500oC
CH3O < (Al)2 (νS(ë–ç) = 2845 cm–1, δ‡s(ë–ç) =
1460 cm–1, δs(ë–ç) = 1440 cm–1, r|| (ëç3) = 1185 cm–1, 6 (‡) (b)
and ν(ë–é) = 1095 cm–1) [24], which form hydrogen
bonds with the neighboring OH groups of alumina 1
2
(ν(é···ç) = 3450–3565 cm–1). The composition of the 3
∆I, arb. units

sample influenced the formation and conversion of sur- 4

face groups in the course of thermal treatment. Only 5
methoxy groups were formed on the surface of alumina 3
at 523 K; other surface groups, which will be consid-
ered below, were also formed upon the treatment of
Pt-containing catalysts.
An increase in the cell temperature to 573 K in the
absence of oxygen resulted in a decrease in the intensi-
ties of absorption bands at 2845 and 2825 cm–1 and an
increase in the intensity of an absorption band at 0 30 60 90 30 60 90
Time, min
2805 cm–1 in the IR spectra of the pure support (Fig. 4a).
This corresponds to the decrease in the concentrations
of bridging and terminal methoxyls (types III and II) Fig. 5. Changes in the absorption band intensities of surface
groups and gas-phase methane on heating a Pt + γ-Al2O3
and the increase in the concentration of terminal meth- sample (see Fig. 4 for notation): (a) without oxygen and
oxy groups (type I). According to Moravek et al. [25], (b) on the addition of oxygen (1 Torr) at the beginning of
the conversion of methoxy complexes CH3O< heating.
CH3O– can occur by an associative mechanism of cat-
alytic dehydration via bimolecular substitution with the
participation of two neighboring surface sites. An decrease in the concentrations of all types of methoxyls
increase in the temperature to 673 K resulted in a (Fig. 4a) and the formation of new surface compounds,
which are characterized by the absorption bands
1 We compared the compositions of radicals only for mechanical ν(ë−ç) = 3002, νas(ë−é) = 1595, νS(ë–é) = 1390,
mixtures. Because the rates of radical generation in the mixed
systems were considerably higher, errors in the relative radical
and δ(ë–ç) = 1375 cm–1 in the IR spectra. According
concentrations decreased with the use of photolysis for identifica- to Matyshak et al. [26], they are due to the formation of
tion. HCOO– formate structures. The concentration of

KINETICS AND CATALYSIS Vol. 44 No. 4 2003

 542 MALAKHOVA et al.

Composition of radical products in the methanol oxidation reaction on mixed catalysts at a reaction temperature of 773 K
Radical composition
Methanol/oxygen ratio
Pt + SiO2 Pt + Al2O3

10 : 1
. . Not detected
2.2% H C O, 97.8% CH3 O
5:1
. . . . .
94% CH3 O , 4% CH3 O 2 , 2% H C O ~100% CH3 O , trace H C O
2:1
. . .
93% CH3 O , 7% CH3 O 2 100% CH3O
1:1
. . . .
80% CH3 O , 20% CH3 O 2 91% CH3O , 9% CH3 O 2
1:2
. . . .
76% CH3 O , 24% CH3 O 2 88% CH3O , 12% CH3 O 2
1 : 10 Not detected Not detected

HCOO– passed through a maximum in the course of adsorbed formaldehyde. Thus, we can assume the
the decomposition of methoxy groups at 673 K. occurrence of the following surface reactions:
Yamashita et al. [27] studied the behavior of surface CH3O<(Al)2 CH3O–Al,
formates in the decomposition of methanol on Cr2O3; 400°C 500°C
they explained an analogous effect that was manifested CH3O–Al HCOO–Al CO2(gas) + H2O(gas),
as the extremal dependence of formate concentrations 400°C
by the formation of methoxy–formate complexes. It is CH3O–Al CH2Oads
likely that such complexes are also formed on the sur- + HO–Al
500°C
CO2(gas) + H2O(gas).
face of alumina to increase the rate of formate decom-
position at temperatures higher than 673 K. At 773 K, In this case, formates are formed with the participa-
the complete decomposition of formate and methoxy tion of neighboring lattice oxygen atoms of the support.
structures occurred with the release of ëé2 and water, In accordance with Fig. 4, the addition of gaseous
oxygen to the system did not change the rate and tem-
which were readsorbed with the formation of surface
perature of decomposition of bridging and terminal
ammonium bicarbonate on cooling the sample to 298 K methoxyls; however, differences in the products of their
(absorption bands at 1230, 1445, and 1650 cm–1) [28]. reactions were observed. According to IR-spectro-
Moreover, the IR spectrum of the sample heated at 673 K scopic data, the formation of carboxylate–carbonate
exhibited the appearance of a weak absorption band at surface structures, which are characterized by absorp-
1690 cm–1, which can be attributed to the absorption of tion bands at 1445, 1530–1550, and 1575 cm–1, was
observed on heating to 673 K. Methane (ν(ë–ç) =
3015 and δ(ë–ç) = 1305 cm–1) and CO (broad absorp-
250oC 300oC 400oC 500oC 250oC 300oC 400oC 500oC tion band at 2110–2165 cm–1 with a minimum at
2140 cm–1) appeared in the gas phase. Carbon monox-
(‡) (b) ide may be a decomposition product of formaldehyde.
6 The concentration of surface formate structures formed
1 at 673 K was lower than that in the absence of gaseous
2 oxygen by a factor of ~2. This is likely due to the rapid
∆I, arb. units

3
4 final oxidation of these structures to the gaseous prod-
5 ucts of deep oxidation. Thus, in the presence of gas-
3 phase oxygen, the conversion of methoxy groups
occurred via an additional reaction path to form surface
carbonate–carboxylate complexes and ëç4, and CO in
the gas phase. The mechanism of methane formation is
unclear; probably, it is formed by the decomposition of
formate–methoxy complexes
2–
0 30 60 90 30 60 90 [HCOO–···CH3O–] CH4(gas) + C O 3 .
Time, min An analysis of the IR spectra of γ-Al2O3 and its
mechanical mixture with platinum suggests that differ-
Fig. 6. Changes in the absorption band intensities of surface
groups and gas-phase methane on heating a Pt/γ-Al2O3 ences in the behavior of reaction intermediates were
sample (see Fig. 4 for notation): (a) without oxygen and observed even when catalyst surfaces were treated with
(b) on the addition of oxygen (1 Torr) at the beginning of methanol vapor at 523 K followed by evacuation at the
heating. specified temperature. In the sample of a mechanical

KINETICS AND CATALYSIS Vol. 44 No. 4 2003

 KINETICS OF FREE RADICAL GENERATION IN THE CATALYTIC OXIDATION 543

mixture of Pt + γ-Al2é3, three types of methoxy groups, complexes, we can conclude that the role of Pt consists
which were described above, as well as formate and in the activation of oxygen. It is likely that oxygen is of
carboxylate (νas(ë–é) = 1575 and νS(ë–é) = 1440 cm–1) first importance in the formation of methoxy radicals.
surface complexes, were identified using IR-spectro- The following reaction scheme can be proposed for
scopic data (Fig. 3b, spectrum 1). The lower tempera- radical generation on the mechanical mixture of Pt + γ-
ture of the formation of HCOO– and RCOO– structures Al2é3:
on the addition of platinum to alumina can result from êt + 1/2é2(gas) O*,
the formation (on the oxidative treatment of the sample)
of chemisorbed oxygen, which is active in the oxidation CH3 CH3
of methanol or methoxyls, on the platinum at the step O + 䊐 +O*
+ .OCH3.
of the initial treatment. Endo et al. [29] observed the O O
Al Al Al Al Al Al
formation of formates in the oxidation of methanol with
molecular oxygen on Pt(111) under conditions of a This reaction path can lead to the formation of addi-
high vacuum. They found that formate degradation at tional free radicals from the support, along with radi-
temperatures higher than 420 K resulted in the forma- cals formed by a spillover. It is likely that radical gen-
tion of deep oxidation products: carbon dioxide and eration on the γ-Al2é3 surface occurs analogously, but
water. Figures 3 and 5 illustrate the decomposition of in this case activated oxygen is formed in the process of
the resulting surface groups in the absence and pres- surface dehydration.
ence of gaseous oxygen, respectively (a change in the As demonstrated above, along with methoxy radi-
intensity of the absorption band νs(ë–ç) = 2825 cm–1, cals, peroxy and formyl radicals were detected in the
which characterizes the terminal methoxy groups oxidation of methanol. We believe that they are formed
CH3O–Altetr, is not shown in Fig. 5; the decrease in the in the secondary reactions of methoxyl with activated
concentration of these groups upon decomposition is oxygen:
proportional to the decrease in the concentration of
bridging methoxy groups). The decomposition of all CH 3 O . + O* CH 3 O .2 *
.
CH 3 O 2 ( gas ) ,
the types of methoxy groups on the sample of a
mechanical mixture under oxygen-free conditions .
CH 3 O 2 *
.
CH 3 O 2 ; CH 3 O .2 *
.
HCO + H 2 O .
occurred similarly to the decomposition of methoxyls
on pure alumina but at a lower rate: the IR spectra Here, CH3 O .2 * (ads) is the peroxy radical resulting
exhibited the absorption bands of the three types of from the interaction with oxygen in the coordination
methoxy groups even at 773 K (Fig. 3, spectrum 4 and sphere of Al+3, usually in the electronically excited
Fig. 5a). The resulting surface formate complexes did state. As evidenced by the results of photolysis, one of
not degrade with the formation of ëé2(gas) + ç2é(gas). the reaction paths of this species consists in the forma-
Methane traces were detected in the gas phase upon tion of the formyl radical.
heating above 673 K. In the presence of oxygen, the
In the study of the supported catalysts by kinetic
decomposition of all the types of surface methoxyl
experiments using matrix isolation method, the forma-
complexes on the sample of a mechanical mixture of
tion of radicals was not detected at all the tested com-
Pt + γ-Al2é3 was considerably accelerated (Fig. 5b).
positions of the reaction mixtures. This finding was
The heating of the sample to 773 K decreased the con-
supported by IR-spectroscopic data. The treatment of
centration of bridging methoxyls by a factor of ~15
the surface of supported platinum catalysts with meth-
(proportionally to a change in the intensity of the
anol vapor at 523 K followed by evacuation at the spec-
absorption band νs(ë–ç) = 2845 cm–1), as compared
ified temperature resulted in the formation of not only
with a decrease in the concentration of CH3O< by a fac-
the three types of surface methoxy groups and surface
tor of ~1.5 under oxygen-free conditions. A portion of
formate complexes, which were described for the
the resulting formate structures was oxidized at 773 K
mechanical mixture of Pt + γ-Al2é3, but also platinum
with the release of ëé2 and water into the gas phase. In
carbonyls. The IR spectra exhibited intense absorption
the gas phase, methane traces were detected at 573 K or
bands at 1790, 2025, 2060, 2080, and 2110 cm–1, which
higher.
characterize bridging and terminal CO complexes on
A comparison between the rates of decomposition êt0 platinum metal clusters of different sizes and êtδ+
of methoxyl complexes for γ-Al2O3 and its mechanical [30, 31]. It is likely that CO and hydrogen are the main
mixture with platinum suggests (Figs. 4b, 5b) that the decomposition products of methoxyl complexes
rate was much higher in the latter case in the presence formed on the surface of supported samples under oxy-
of gas-phase oxygen. In kinetic experiments with the gen-free conditions at 523–573 K. The IR spectra
use of matrix isolation method, we found that the exhibited a considerable increase in the absorption
amount of the resulting methoxy radicals in the band intensities of platinum carbonyls. At 673–773 K,
mechanical mixture of Pt + γ-Al2é3 was several times only the products of full oxidation were detected
greater than that in the case of pure alumina. Assuming among the decomposition products of methoxyls. The
that methoxy radicals appeared in the gas phase introduction of oxygen into the system at 523 K decom-
because of the decomposition of surface methoxyl posed carbonyls on platinum metal. On heating above

KINETICS AND CATALYSIS Vol. 44 No. 4 2003

544 MALAKHOVA et al.

573 K, the IR spectra exhibited a continuous increase in unchanged. In this case, the pressure in the system
absorption bands due to CO in the gas phase. Moreover, decreased to 4 × 10–2 Torr in 3 min, whereas the signal
a considerable increase in the intensities of absorption intensity increased by a factor of ~6. That is, the rate of
bands due to formate structures was observed, which is radical desorption after the termination of supplying
indicative of the rapid oxidation of methoxyl com- methanol increased by almost two orders of magnitude.
plexes by oxygen activated on dispersed platinum metal After 10 min, the pressure in the system increased again
(Fig. 6). In the absence of oxygen, methoxyl complexes and stabilized at a value of ~2 × 10–2 Torr after 35 min
on the surface of supported catalysts decomposed with with a slow increase in the signal intensity. The
the release of CO; in an excess of oxygen, the mecha- repeated supply of methanol to the reactor resulted in
nism of rapid oxidation to formates came into play. an increase in the pressure to limiting values in 3 min,
Thus, the absence of radicals from the gas phase in the and the signal intensity continued to increase for 50 min
case of supported platinum catalysts is likely due to a after stopping the flows of methanol and oxygen.
low steady-sate concentration of surface intermediates An analogous so-called stop effect was described
because of their rapid conversion into nonparamagnetic [34–36] for the heterogeneous oxidation reactions of
stable decomposition products. According to published alcohols and amines on amphoteric metal oxides, when
data, the rate of methanol decomposition on the the termination of the flow of a reactant resulted in a
Pt/Al2O3 and Ni/Al2O3 supported metal catalysts is considerable change in the rate of formation of final
higher than the rate of methanol decomposition on the products. The appearance of the stop effect for radical
support by a factor of ~106 [32]. Correspondingly, both products is indicative of a significant contribution of
the steady-state concentration of intermediates and the radical reactions under conditions of a nonstationary
rate of radical desorption into the gas phase should state of the catalyst surface; this opens up the way to the
decrease to the same extent. determination of elementary steps in the reaction mech-
The treatment of samples based on silica gel with anism of methanol oxidation. Because the composition
absolute methanol at 523 K followed by evacuation at of the gas phase smoothly changed (from the ratio
the specified temperature resulted in the formation of methanol/oxygen = 10 : 1 to 1 : 10) when the flow of
surface methoxy groups with the simultaneous degra- methanol was terminated, we actually observed the for-
dation of methanol to CO and ç2é. The IR spectra mation of radicals over the entire range of concentra-
exhibited the absorption bands νas(ë–ç) = 2997, νas(ë– tions.
ç) = 2959, νs(ë–ç) = 2857, and δ(ëç3) = 1450 cm–1,
which characterize isolated methoxy groups formed in
the reaction of methanol with isolated hydroxyl groups CONCLUSIONS
of silica gel, and an absorption band with ν(é–ç) = A comparison between experimental conditions in
3400–3550 cm–1, which characterizes hydrogen- the studies of methanol oxidation using Fourier trans-
bonded methoxy groups formed by the chemisorption form IR spectroscopy and matrix isolation method indi-
of methanol at siloxane bridges [17]. The isolated cates that the initial conditions (catalysts, catalyst treat-
methoxy groups exhibit enhanced thermal stability: ment conditions, temperature regions, and reactant
they underwent no degradation or oxidation on heating compositions) were identical or overlapping, whereas
the samples to 800 K [33] (the temperature range of the modes of reactant supply to the catalyst surface
radical detection). were different. Flow conditions with a varied composi-
An increase in the concentration of methoxy groups tion of the reactants in a flow were used (methanol/oxy-
on the surface of a mixed catalyst increased the rate of gen = 0.1–10) in the matrix isolation method technique,
radical generation due to the adsorption substitution whereas the stationary conditions of methanol chemi-
sorption at 523 K were used in the IR-spectroscopic
Pt + 1/2O 2 O*, experiments, and the reactions of surface intermediates
. in the presence or in the absence of oxygen in the gas
O* + 2 SiOCH3 Si O Si + 2CH 3 O phase were studied over the temperature region of rad-
ical detection. In the case of the IR-spectroscopic
and/or the spillover of methoxy radicals from platinum. experiments, the ratio between the reactants (the
amount of chemisorbed methanol/the amount of oxy-
gen in the gas phase) was 4 : 1, which corresponded to
Radical Stop Effect the center of the range of compositions studied by
An interesting effect was found on the Pt + γ-Al2O3 matrix isolation method.
catalyst at the ratio methanol/oxygen = 10 : 1 with the As demonstrated above, methoxy and formate
use of matrix isolation method. Because the pressure in groups (main intermediates in the catalytic reactions of
the system increased from 10–4 to 10–1 Torr (beyond the methanol oxidation) were detected on the surfaces of
working range of a diffusion pump; the signal intensity catalysts based on γ-Al2O3 in all the test systems. It
of radicals remained at a noise level) after radical accu- would be expected that the detected surface species are
mulation for ~45 min, the supply of methanol was ter- the precursors of gas-phase radicals, which were
minated, whereas the flow of oxygen remained detected in the matrix isolation method experiments.

KINETICS AND CATALYSIS Vol. 44 No. 4 2003

 KINETICS OF FREE RADICAL GENERATION IN THE CATALYTIC OXIDATION 545

The chemisorption of the methanol molecule on alu- gates. It is likely that this boundary layer exhibits spe-
mina resulted in the methoxy anion stabilized on the cial properties. The special properties of this boundary
Al3+ cation and a proton, which reacted with the neigh- layer can be due to a superequilibrium amount of meth-
boring basic O2– to form the hydroxyl group. Methoxy oxy groups under reaction conditions. Methoxy parti-
groups on the surface of γ-Al2O3 are mobile and reac- cles that were formed on platinum, which tend to stabi-
tive. This manifests itself in the rearrangement of the lize on the support by the spillover mechanism, and
species of types II and III into the methoxy groups of methoxy groups that migrated by the reverse spillover
type I and in the formation of formate structures with mechanism from the support to the active component
the simultaneous consumption of methoxyls. In the simultaneously diffused to the platinum–support inter-
case of silica gel, chemisorbed methanol is the methoxy face. When the surface sites of the support near the
group stabilized at isolated silicon cations (at the sites interface are blocked by chemisorbed methanol, the
of isolated hydroxyl groups) or hydrogen-bonded methoxy particles that tend to the interface from plati-
methoxyls formed on the chemisorption of methanol at num can provide an additional reaction path for the for-
siloxane bridges. Because the methoxy group stabilized mation of gas-phase radicals. In terms of this reaction
at isolated silicon cations are thermally stable up to 800 K scheme, we can explain the lower activation energy of
[18], it is likely that in this temperature region methox- the formation of gas-phase radicals for the SiO2 + Pt
yls that are desorbed from siloxane bridges contribute to catalyst, as compared with the energies of radical gen-
the formation of gas-phase radicals. As estimated previ- eration on the support and the active component.
ously [37], the number of active surface sites on silica The absence of data on the molar absorption coeffi-
gel after treatment at 973 K was ~1.5 × 10–4 mol/g, and cients of methoxy and formate groups on catalyst sur-
the concentration ratio was [SiOH]/[SiOSi] = 2 : 1. At faces does not allow us to quantitatively compare the
the same time, according to published data [38], the rates of conversion of surface groups and the rates of
concentration of active surface sites for various alumi- formation of gas-phase radicals; this comparison will
num modifications was ~1018 site/m2, which gave a be the subject matter of further studies.
value of ~3 × 10–3 mol/g based on the surface area of the
support used in this work. Thus, the contribution of sur- REFERENCES
face methoxyls to the formation of gas-phase radicals
on alumina should be at least one order of magnitude 1. Nalbandyan, A.B. and Mashtayan, A.A., Elementarnye
higher than that on silica gel. These estimated values protsessy v medlennykh gazofaznykh reaktsiyakh (Ele-
are consistent with the results of matrix isolation mentary Processes in Model Gas-Phase Reactions), Yere-
van: ArmSSR Acad. Sci., 1975.
method experiments (Fig. 1), when the rates of forma-
. . 2. Nalbandyan, A.B. and Vardanyan, I.A., Sovremennoe
tion of total CH3 O and CH3O O radicals on the pure sostoyanie problemy gazofaznogo okisleniya organi-
supports were compared. cheskikh soedinenii (State of the Art and Problems of
The introduction of platinum changed the structure Gas-Phase Oxidation of Organic Compounds), Yerevan:
of active surface sites. According to published data ArmSSR Acad. Sci., 1986.
[20], supported catalysts based on alumina contained a 3. Krishtopa, L.G., Ermolaev, V.K., Ismagilov, Z.R., and
set of PtOn mixed oxides (where n ≥ 1), PtOxCly incor- Zamaraev, K.I., Proc. VIII Soviet–French Seminar,
porated into the oxygen sublattice of Al2O3, and Pt0 Novosibirsk, 1990, p. 96.
metal clusters of different size. In supported platinum 4. Ermolaev, V.K., Pak, S.R., Krishtopa, L.G., Ismagi-
catalysts based on silica gel, Pt0 clusters with a size of lov, Z.R., and Zamaraev, K.I., Khim. Fiz., 1988, vol. 7,
~10–15 Å were formed. The amount of oxidized plati- no. 8, p. 1141.
num in the catalyst increased after high-temperature 5. Garibyan, E.G., Muradyan, A.A., and Garibyan, T.A.,
pretreatment in oxygen; in this case, finely dispersed Arm. Khim. Zh., 1978, vol. 31, no. 7, p. 466.
Pt0 clusters (up to 15 Å in size) disappeared, but the 6. Boyd, S.L., Boyd, R.J., and Barclay, L.R.C., J. Am.
metal core of large clusters (> 60 Å) was retained [20]. Chem. Soc., 1990, vol. 112, no. 15, p. 5724.
The disappeared clusters were converted into PtO2; it is 7. Malakhova, I.V. and Ermolaev, V.K., Mendeleev Com-
likely that they were also incorporated into the lattice of mun., 1998, vol. 3, p. 83.
the support. Based on the experimental results (see 8. Applegate, B.E., Pushkarsky, M.B., and Miller, T.A.,
Fig. 6), we believe that the inclusion sites and/or J. Phys. Chem. A, 1999, vol. 103, p. 1538.
boundaries serve as the sites of formation (stabiliza- 9. Kosaki, K.Y., Miyamoto, A., and Murakami, Y., Bull.
tion) of formate structures and, consequently, regions Chem. Soc. Jpn., 1982, vol. 55, p. 2706.
where the deep oxidation of methanol occurs. 10. Sal’nikov, V.S. and Tsyrul’nikov, P.G., Kinet. Katal.,
It is likely that the size of Pt aggregates in mixed cat- 1988, vol. 29, no. 5, p. 1117.
alysts remained unchanged (according to optical micro- 11. Kuz’mina, G.M., Cand. Sci. (Chem.) Dissertation,
scopic data, the characteristic size of these aggregates Novosibirsk: Institute of Catalysis, 1972.
is 0.8–70 µm). Under reaction conditions, the surface 12. Belyi, A.S., Smolikov, M.D., Fenelonov, V.B.,
of aggregates was covered with chemisorbed oxygen Gavrilov, V.Yu., and Duplyakin, V.K., Kinet. Katal.,
atoms, which formed a boundary layer near Pt0 aggre- 1986, vol. 27, no. 6, p. 1414.

KINETICS AND CATALYSIS Vol. 44 No. 4 2003

546 MALAKHOVA et al.

13. Smirnova, I.E., Belyi, A.S., Smolikov, M.D., and Dup- 26. Matyshak, V.A., Khomenko, T.I., Lin, G.I., Zavali-
lyakin, V.K., Kinet. Katal., 1990, vol. 31, no. 3, p. 687. shin, I.N., and Rozovskii, A.Ya., Kinet. Katal., 1999,
14. Zelinskii, N.D., Izbrannye trudy (Selected Works), Mos- vol. 40, no. 2, p. 295.
cow: Nauka, 1968. 27. Yamashita, K., Naito, S., and Tamaru, K., J. Catal.,
15. Ingold, K.U. and Morton, J.R., J. Am. Chem. Soc., 1964, 1985, vol. 94, p. 353.
vol. 86, no. 16, p. 3400. 28. Shen, J., Cortright, R.D., Chen, Y., and Dumesic, J.A.,
16. Bhattacharya, D. and Willard, J.E., J. Phys. Chem., J. Phys. Chem., 1994, vol. 98, p. 8067.
1982, vol. 86, p. 962.
29. Endo, M., Matsumoto, T., Kuboto, J., Domen, K., and
17. Wovchko, E.A., Camp, J.C., Glass, J.A., and Yates, J.T., Hirose, Ch., J. Phys. Chem., B, 2000, vol. 104, p. 4916.
Jr., Langmuir, 1995, vol. 11, p. 2592.
30. Sheppard, N. and Nguyen, T.T., Advances in Infrared
18. Pelmenschikov, A.G., Morosi, G., Gamba, A., and Raman Spectroscopy, Clark, R.J.H. and Hester R.E.,
Zecchina, A., Bordiga, S., and Paukshtis, E.A., J. Phys. Eds., London: Philadelphia-Rhein, 1978, vol. 5, p. 67.
Chem., 1993, vol. 97, p. 11979.
19. Zum Mallen, M.P. and Schmidt, L.D., J. Catal., 1996, 31. Matyshak, V.A., Khomenko, T.I., Bondareva, N.K., Pan-
vol. 161, p. 230. chishnyi, V.I., and Korchak, V.N., Kinet. Katal., 1998,
vol. 39, no. 1, p. 100.
20. Zamaraev, K.I. and Kochubei, D.I., Kinet. Katal., 1986,
vol. 27, no. 5, p. 1031. 32. Chen, B. and Falconer, J.L., J. Catal., 1993, vol. 114,
21. Satterfield, C.N., Heterogeneous Catalysis in Practice, p. 214.
New York: McGraw-Hill, 1980. 33. Paukshtis, E.A., J. Mol. Catal., A: Chem., 2000, vol. 158,
22. Lunina, E.V., Kataliz. Fundamental’nye i prikladnye p. 37.
issledovaniya (Catalysis: Fundamental and Applied 34. Kouber, J., Pasek, J., and Ruzicka, V., New Horizons in
Studies), Petrii, O.A. and Lunin, V.V., Eds., Moscow: Catalysis, Amsterdam: Elsevier, 1980, p. 853.
Mosk. Gos. Univ., 1987, p. 262.
35. Koubek, J., Paser, J., and Ruzicka, V., Catalyst Deactiva-
23. Pak, S.N., Cand. Sci. (Chem.) Dissertation, Novosibirsk: tion, Amsterdam: Elsevier, 1980, p. 251.
Institute of Catalysis, 1990.
36. Thullie, J. and Renken, A., Chem. Eng. Sci., 1991,
24. Busca, G., Rossi, P.F., Lorenzelli, V., Benaissa, M., vol. 46, p. 1083.
Travert, J., and Lavalley, J.-C., J. Phys. Chem., 1985,
vol. 89, p. 5433. 37. Paukshtis, E.A. and Danilova, I.G., Kinet. Katal., (in
25. Moravek, V., Lamotte, J., and Lavalley, J.C., Proc. VI Int. press).
Symp. on Heterogeneous Catalysis, Sofia, 1987, vol. 1, 38. Narayanan, C.R., Srinivasan, S., Datye, A.K., Gorte, R.,
p. 377. and Biaglow, A., J. Catal., 1992, vol. 138, p. 659.

KINETICS AND CATALYSIS Vol. 44 No. 4 2003