Reaction kinetics and mechanism of selective NO reduction on a CoZSM-5 catalyst as studied by SSITKA

Article abstract of DOI:10.1023/B:KICA.0000032181.81830.fd

The dynamics of ¹³C transfer from methane to carbon dioxide was studied under the steady-state reaction conditions of selective NO reduction with methane on a CoZSM-5 catalyst at various reactant (NO, CH₄, and O₂) concentrations and temperatures. It was found that the reaction occurs by a two-pathway mechanism with the participation of Co²⁺ sites (or CoO_x clusters) and paired Co²⁺-OH sites localized at the boundary between the clusters and the zeolite; in this case, the rate of the reaction at boundary sites was higher by more than one order of magnitude. Based on the numerical simulation of isotopic response curves, the concentrations of intermediate compounds and the rate constants of particular steps were evaluated; differences in the kinetics via the above reaction pathways were found and analyzed.

Full text of DOI:10.1023/B:KICA.0000032181.81830.fd

Kinetics and Catalysis, Vol. 45, No. 3, 2004, pp. 436–445. Translated from Kinetika i Kataliz, Vol. 45, No. 3, 2004, pp. 463–472.
Original Russian Text Copyright © 2004 by Sadovskaya, Suknev, Goncharov, Bal’zhinimaev, Mirodatos.
CATALYTIC REACTION MECHANISMS
Reaction Kinetics and Mechanism of Selective NO Reduction
on a CoZSM-5 Catalyst as Studied by SSITKA
E. M. Sadovskaya*, A. P. Suknev*, V. B. Goncharov*, B. S. Bal’zhinimaev*, and C. Mirodatos**
* Boreskov Institute of Catalysis, Siberian Division, Russian Academy of Sciences, Novosibirsk, 630090 Russia
** Institut de Recherches sur la Catalyse, F 69626 Villeurbanne Cedex, France
Received February 6, 2003
Abstract—The dynamics of 13C transfer from methane to carbon dioxide was studied under the steady-state
reaction conditions of selective NO reduction with methane on a CoZSM-5 catalyst at various reactant (NO,
CH4, and O2) concentrations and temperatures. It was found that the reaction occurs by a two-pathway mech-
anism with the participation of Co2+ sites (or CoOx clusters) and paired Co2+–OH sites localized at the boundary
between the clusters and the zeolite; in this case, the rate of the reaction at boundary sites was higher by more
than one order of magnitude. Based on the numerical simulation of isotopic response curves, the concentrations
of intermediate compounds and the rate constants of particular steps were evaluated; differences in the kinetics
via the above reaction pathways were found and analyzed.
and NOδ2+ species considerably decreased; this fact
INTRODUCTION
The selective reduction of NO with methane on a
CoZSM-5 catalyst is a potential process for the removal
of nitrogen oxides from waste gases. It is well known
that the rate of the reaction dramatically increases in the
presence of oxygen [1–4] because of the formation of
adsorbed [NOx] species, which are key intermediates in
the reduction of nitrogen oxides [2–6], on the catalyst
surface. A great number of publications have been
devoted to the nature and reactivity of [NOx] toward
hydrocarbons [4, 7–17]. According to IR-spectroscopic
data [10–12], NO is adsorbed on CoZSM-5 with the
formation of mono- and dinitrosyl complexes of cobalt,
whereas nitrite–nitrate complexes are additionally
formed in the presence of oxygen. Moreover, the for-
mation of adsorption complexes with an absorption
was indicative of the active participation of these spe-
cies in the reaction. Based on an analysis of isotope
14
transfer processes with the replacement of N16O by
15N18O in a mixture of NO + O2 [23], as well as NO +
O2 + CH4 [24], the rates of formation of adsorbed [NOx]
complexes as functions of oxygen concentration in
mixtures and the rates of their conversion into N2 were
determined. As a result, it was demonstrated that
molecular nitrogen is formed with the participation of
both NOδ2+ species and nitrite–nitrate complexes. In
the former case, the rate of the reaction was higher than
that in the latter case by a factor of ~2.5. Nevertheless,
some aspects that are crucial for understanding the
mechanism of this reaction remained questionable. In
particular, the conversion of carbon-containing inter-
mediates on the catalyst surface in CO2, as well as the
participation of these species in the transformations of
nitrogen-containing intermediates in N2, in accordance
with a mechanism proposed based on the isotope stud-
ies of 15N-label transfer, remained unclear.
band at about 2130 cm–1, which were assigned to NOδ2+
species additionally bound to the acidic OH groups of a
zeolite, has been reported in a number of publications
[10–12, 18–22]. Presently, the reactivity of [NOx] spe-
cies toward methane is still an open question. It is
believed [10–12, 14] that these are nitrite or nitrate
compounds. At the same time, data on the participation
of NOδ2+ species in the reduction of NO on CuZSM-5
[19] and HZSM-5 [20] catalysts were reported.
Previously, with the use of in situ diffuse-reflectance
IR spectroscopy [17], we detected three types of
adsorption complexes formed in the course of the
simultaneous adsorption of NO and O2 on a CoZSM-5
catalyst: cobalt mononitrosyls (a band at 1930 cm–1),
nitrite–nitrate complexes (1523 cm–1), and NO2δ+ spe-
Therefore, in this work, we studied the dynamics of
carbon-label transfer on the replacement of 12CH4 with
13CH4 under conditions of the steady-state reaction of
NO reduction with methane. The experiments were
performed under conditions similar to those in experi-
ments with labeled nitrogen oxide. Based on the
numerical simulation of isotopic response curves, the
reaction scheme of methane conversion into CO2 was
determined, and the concentrations and reaction rates
of intermediates were estimated as functions of reactant
concentrations and temperature. As a result, we
revealed differences in the kinetics of the reaction via
cies (2130 cm–1). On the addition of methane to a mix-
ture of NO and O2, the concentration of nitrite–nitrate
0023-1584/04/4503-0436 © 2004 MAIK “Nauka /Interperiodica”

REACTION KINETICS AND MECHANISM OF SELECTIVE NO REDUCTION
437
tion of the known contribution of 13CH4 from this inten-
sity.
pathways with the participation of NOδ2+ species and
nitrite–nitrate complexes and, after a comparison with
isotope data on 15N18O, gained a clearer insight into the
reaction mechanism.
A catalyst fraction of 0.3–0.7 mm was used in all of
the above experiments; the catalyst weight was 0.1 g,
and the gas flow rate was 2 ml/s. Experiments on the
isotopic substitution of 13CH4 for 12CH4 were per-
formed under varying concentrations of CH4, NO, and
O2 in the starting mixture and temperature. Nine exper-
iments were performed in all at three different values of
each of the parameters. Table 1 summarizes the initial
concentrations and the conversions of CH4 and NO.
Based on the concentrations (C) of all isotopomers for
EXPERIMENTAL
The CoZSM-5 catalyst used in our experiments was
synthesized in accordance with the procedure described
elsewhere [14]. Note that it was prepared by the ion
exchange of HZSM-5 zeolite (with a SiO2/Al2O3 ratio
equal to 37) with a Co(NO3)2 solution followed by dry-
ing and calcination at 500°C. The cobalt content of the
catalyst was 1.8 wt %, which corresponds to 2 ×
1020 (atom Co)/(g Cat). X-ray diffraction data showed
that the catalyst was free of a cobalt oxide phase.
According to EXAFS data, cobalt occurred as small
oxide-like clusters with a tetrahedral oxygen environ-
ment [14].
13
each of the substances, the isotope fractions of C in
CH4 and CO2 were calculated as follows:
C(13CH4)
C(13CH4) + C(12CH4)
---------------------------------------------------
α
α
=
=
,
.
13CH4
C(13CO2)
---------------------------------------------------
13CO2
C(13CO2) + C(12CO2)
The SSITKA experiments were performed as
described below. Before an experiment, the catalyst
was held in a mixture of 6%O2 + He at 500°C for 1 h
and then cooled in the same flow to a temperature at
which the experiment was performed. Next, a reaction
mixture of NO + CH4 + O2 + He was introduced into the
reactor; after the attainment of steady-state reaction
conditions (usually, in 8–10 min), the reaction mixture
was stepwise replaced with a mixture of the same
chemical composition containing, however, carbon-
labeled methane 13CH4 (98% enrichment). As an isoto-
pic equilibrium was approached, the dynamics of 13C-
label transfer from methane to CO2 (CO was not found
in the reaction products) were measured. The experi-
ments were performed in a U-shaped quartz reactor
placed in a gradientless furnace. The diffusion of a gas-
flow front in this reactor was not greater than 1 s, as
determined with the use of argon added to the reaction
mixture in an amount of 0.1%. This allowed us to
ignore diffusion effects in the numerical simulation of
SSITKA data.
EXPERIMENTAL RESULTS
Figures 1a–4a demonstrate changes in the isotope
fraction of 13C in CH4 and CO2 molecules with time at
the reactor outlet after the replacement of a starting
mixture of 12CH4 + NO + O2 + He by a mixture of
13CH4 + NO + O2 +Ar + He. It can be seen that the frac-
tion of 13C in CH4 increased almost synchronously with
the argon concentration. In this case, the shapes of the
curves depended only slightly on the reaction condi-
tions. A small shift (on the order of 1 s) of the function
α
(t) with respect to the response curve of argon
13CH4
was commensurable with the time resolution of the
mass spectrometer, and it was most likely due to the
weak absorption of methane in zeolite channels.
As distinct from methane, the concentration of the
13C label in CO2 increased more slowly in all of the
The chemical and isotopic composition of a gas
mixture at the reactor outlet was detected with the use
of a VG SENSORLAB 200D quadrupole mass spec-
trometer, which simultaneously recorded up to
16 masses. The quantitative analysis of mass-spectro-
metric data was performed with the use of spectra
recorded for calibration gas mixtures with known con-
centrations of reaction mixture components. The con-
centrations of H2O, NO, O2, 12CO2, and 13CO2 were
determined from the intensities of parent peaks with
m/z = 18, 30, 32, 44, and 45, respectively. The concen-
trations of N2 and 13CH4 were determined from the
experiments, and the function α
(t) was noticeably
13CO2
shifted with respect to the response curves of argon and
methane. As was found previously [25, 26], the qualita-
tive features of the reaction mechanism can be deter-
mined from the shape of isotopic response curves plot-
ted as the logarithmic dependence of the isotope frac-
tion on time. This dependence is strictly linear if the
reaction occurs by a consecutive mechanism. As can be
seen in Figs. 1b–4b, in our case, the logarithmic func-
tion α
(t) is represented by a concave curve (convex
13CO2
intensities of peaks with m/z = 28 and 17, respectively, downwards), which is characteristic of reactions that
from which the known contributions of CO2 and H2O occur by a mechanism with parallel steps or by a con-
(CO+ and OH+ fragments) were subtracted beforehand. secutive mechanism with a buffer step. Based on the
12
The concentration of CH4 was determined from the currently available notions of the mechanism of this
intensity of a fragment with m/z = 15 after the subtrac- reaction [17, 24], a parallel (two-pathway) mechanism
KINETICS AND CATALYSIS Vol. 45 No. 3 2004

438
SADOVSKAYA et al.
Table 1. Conditions of SSITKA experiments
Initial concentration, vol %
O2
Conversion, %
Experiment no.
T, °C
NO
CH4
NO
CH4
1
2
3
4
5
6
7
8
9
450
450
450
450
450
450
450
400
480
0.3
0.3
0.3
0.3
0.3
0.2
0.35
0.3
0.3
3.0
3.0
3.0
0.3
0.6
3.0
3.0
3.0
3.0
0.3
0.4
0.6
0.4
0.4
0.4
0.4
0.4
0.4
31.4
39.9
48.5
32.3
27.1
53.1
39.0
44.9
66.8
15.6
15.7
12.8
12.5
11.2
19.1
17.7
6.2
44.9
of CO2 formation due to the reactions of methane with substitution is determined by the ratio of the rate of
reaction (r) to the concentration of intermediate com-
pounds (θ). Consequently, the ratio r/θ remained
unchanged on varying the concentrations of methane
and NO, whereas it increased with oxygen concentra-
tion and temperature. This means that the rate of con-
version of surface intermediates, which are formed by
the interaction of methane with [NOx], is practically
various adsorbed NO species seems most probable.
The shape of the isotopic response curves of
α
(t) remained unchanged on varying the tempera-
13CO2
ture and composition of a starting mixture; however,
the rate of isotopic substitution in CO2 changed notice-
ably (Figs. 1–4). Changes in the concentrations of
methane and NO had almost no effect on the isotopic independent of the concentration of NO; however, it
response dynamics (Figs. 1, 2), whereas the rate of 13C
transfer to CO2 noticeably increased as the concentra-
tion of oxygen and the temperature were increased
essentially depends on oxygen concentration and tem-
perature. Next, based on a numerical simulation of the
13
dynamics of C-label transfer, we attempted to deter-
(Figs. 3, 4). It is well known that the rate of isotopic mine the rates of reaction via either of the pathways and
α
13C
lnα
13Cé2
(a)
(b)
1.0
0.8
0.6
0.4
0.2
0
0
5
6
–1
1
2
3
4
–2
–3
–4
0
10
20
30
10
20
30
Time, s
Time, s
13
13
Fig. 1. Changes in (a) the isotope fraction of C in (1, 2) CH and (3, 4) CO and (b) the logarithm of the isotope fraction of
C
4
2
in (5, 6) CO with time at (1, 3, 5) C
= 0.3% and (2, 4, 6) CCH = 0.6% (experiment nos. 1 and 3 in Table 1, respectively).
2
CH4
4
Points, solid lines, and the dashed line indicate experimental data, calculated data, and argon, respectively. The arrow indicates the
point in time at with an isotope mixture was introduced.
KINETICS AND CATALYSIS Vol. 45 No. 3 2004

REACTION KINETICS AND MECHANISM OF SELECTIVE NO REDUCTION
439
α
13C
lnα
13Cé2
(a)
(b)
1.0
0.8
0.6
0.4
0.2
0
0
–1
–2
–3
–4
5
6
1
2
3
4
0
10
20
30
10
20
30
Time, s
Time, s
13
13
Fig. 2. Changes in (a) the isotope fraction of C in (1, 2) CH and (3, 4) CO and (b) the logarithm of the isotope fraction of
C
4
2
in (5, 6) CO with time at (2, 4, 6) C = 0.2 and (1, 3, 5) 0.35% (experiment nos. 6 and 7 in Table 1, respectively). Points, solid
lines, and the dashed line indicate experimental data, calculated data, and argon, respectively. The arrow indicates the point in time
at with an isotope mixture was introduced.
2
NO
α
13C
lnα
13Cé2
(a)
(b)
1.0
5
6
0
–1
–2
–3
–4
0.8
0.6
0.4
0.2
0
1
2
3
4
0
10
20
30
10
20
30
Time, s
Time, s
13
13
Fig. 3. Changes in (a) the isotope fraction of C in (1, 2) CH and (3, 4) CO and (b) the logarithm of the isotope fraction of
C
4
2
in (5, 6) CO with time at (1, 3, 5) C = 3.0 and (2, 4, 6) 0.3% (experiment nos. 2 and 4 in Table 1, respectively). Points, solid
2
O2
lines, and the dashed line indicate experimental data, calculated data, and argon, respectively. The arrow indicates the point in time
at with an isotope mixture was introduced.
to evaluate the concentrations of key intermediates and reactor length [26, 27]. The form of the right-hand
the rate constants of the steps of their conversion.
sides of the equations depends on the mechanism of
the reaction. Initially, it was assumed that the reaction
can occur via several parallel pathways and the steps
of methane interaction with [NOx] and all the subse-
quent reactions of C,N-containing surface com-
pounds to form CO2 are irreversible. Then, the reac-
tion scheme of methane conversion into CO2 can be
NUMERICAL SIMULATION RESULTS
A dynamic model for isotope transfer in a steady-
state reaction in a plug-flow reactor is represented as
a set of hyperbolic differential equations, which
reflect changes in the concentration of a label isotope
in the reaction components with time and along the represented as follows:
KINETICS AND CATALYSIS Vol. 45 No. 3 2004

440
SADOVSKAYA et al.
[Z1–C...N]1
[Zi–C...N]1
[Zn–C...N]1
…
…
…
[Z1–C...N]j
[Zi–C...N]j
[Zn–C...N]j
…
…
…
CO2
CO2
CO2
r1
ri
CH4
rn
,
fraction of the 13C label as a constituent of θij; ri is the
where [Zi–C...N]j refers to intermediate C,N-contain-
ing compounds.
rate of the reaction via the ith pathway, molecule g–1 s–1;
n is the number of reaction pathways; m is the number
of reaction steps in every pathway; V is the reactor vol-
ume, cm3; U is the gas flow rate, cm3/s; G is the catalyst
sample weight, g; NA is Avogadro number; t is time, s;
and ξ is the dimensionless length of the catalyst bed.
A model of 13C-label transfer from CH4 to CO2 cor-
responding to the above scheme has the form
n
∂CCH αCH
∂C
α
∂t
CH4 CH4
U
V
G
4
4
----------------------------
-------
VN
------------------------- +
= –
riαCH ,
4
∑
∂ξ
i = 1
The number of reaction pathways (n), the rates of
reaction via these pathways (ri), and the concentrations
of C,N-containing compounds θij were determined by
numerical analysis of the set of Eqs. (1) so that the
experimental curves of α(t) were identical with the cal-
culated curves. As a result, we found that the conver-
sion of methane into CO2 occurs via two main path-
ways and evaluated the rates of reaction via these path-
ways. Note that the results of the simulation do not
exclude the occurrence of other pathways; however, the
contribution of these pathways to the overall rate of
CO2 formation is negligibly small. Unfortunately,
SSITKA did not allow us to detail the reaction
sequences of C,N-containing intermediates because of
the inadequate time resolution of the procedure. There-
fore, we evaluated only the total concentrations of
C,N-containing compounds formed in each of the main
pathways. Table 2 summarizes the rates of reaction via
the two designated pathways and the concentrations of
C,N-containing intermediates. The calculated functions
∂αij
∂t
i1---------
θ
= riαCH – riαi1 ,
4
…………………………
(1)
∂αij
ij---------
θ
= riαij – 1 – riαij ,
∂t
n
∂CCO αCO
∂C
α
∂t
CO2 CO2
U
V
G
2
2
----------------------------
-----------
------------------------- +
=
riαim .
∑
∂ξ
VNAi = 1
The initial and boundary conditions are as follows:
t = 0: αij = 0, αCO = 0, αCH = 0,
2
4
ξ = 0: αCH = 0.98,
4
where CCH and CCO are the concentrations of CH4
4
2
and CO2 in a gas phase, vol %; θij is the surface concen-
tration of a C,N-containing compound formed via the
ith pathway at the jth step, molecule/g; αij is the isotope
α
13C
lnα
13Cé2
(a)
(b)
1.0
5
0
6
0.8
–1
1
2
3
4
0.6
0.4
0.2
0
–2
–3
–4
0
10
20
30
Time, s
10
20
30
Time, s
13
13
Fig. 4. Changes in (a) the isotope fraction of C in (1, 2) CH and (3, 4) CO and (b) the logarithm of the isotope fraction of
C
4
2
in (5, 6) CO with time at (1, 3, 5) T = 400 and (2, 4, 6) 480°C (experiment nos. 8 and 9 in Table 1, respectively). Points, solid lines,
2
and the dashed line indicate experimental data, calculated data, and argon, respectively. The arrow indicates the point in time at with
an isotope mixture was introduced.
KINETICS AND CATALYSIS Vol. 45 No. 3 2004

REACTION KINETICS AND MECHANISM OF SELECTIVE NO REDUCTION
441
Table 2. Calculated values of the rates of CO2 formation via different pathways and the concentrations of surface [Zi–C...N]
complexes
Rate of CO2 formation × 10–18
,
Concentration × 10–18
,
Initial concentration, vol %
molecule g– 1 s–1
molecule/g
T, °C
NO
O2
CH4
r1
r2
r1/r2
[Z1–C...N] [Z2–C...N]
450
0.3
0.3
0.3
0.3
0.3
0.3
0.2
0.3
0.35
0.3
0.3
0.3
3.0
3.0
3.0
0.3
0.6
3.0
3.0
3.0
3.0
3.0
3.0
3.0
0.3
0.4
0.6
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.19
0.25
0.30
0.18
0.16
0.25
0.30
0.25
0.28
0.08
0.25
0.74
0.08
0.11
0.14
0.10
0.10
0.11
0.14
0.11
0.12
0.04
0.11
0.28
2.4
2.3
2.2
1.8
1.6
2.3
2.1
2.3
2.3
2.0
2.3
2.6
0.16
0.22
0.26
0.28
0.20
0.22
0.22
0.22
0.20
0.20
0.22
0.06
0.8
1.0
1.5
1.4
1.3
1.0
1.4
1.0
1.1
1.0
1.0
1.1
450
450
400
450
480
with surface nitrite–nitrate complexes in the second
case.
α
(t) obtained based on the two-pathway reaction
13CO2
scheme are practically identical to the experimental
functions (Figs. 1–4); differences between them lie
within the limits of experimental error.
It is well known that the formation of nitrite–nitrate
complexes is associated with the presence of transition
metal cations or transition metal oxide clusters in zeo-
lite channels [4, 9–17]. As for NOδ2+, the formation of
DISCUSSION
these species has been observed on a number of cata-
lysts (NaZSM-5, CoZSM-5, and FeZSM-5) including
HZSM-5 [11, 12, 17–22]. Therefore, it is believed that
they are associated with the strong proton sites of zeo-
lites, and they can be stabilized in the presence of tran-
sition metal cations [12, 17, 19, 20]. This also follows
from data obtained previously using in situ diffuse-
reflectance IR spectroscopy on the same catalyst sam-
ple that exhibited a clear correlation between the rate
An analysis of the dynamics of carbon-label transfer
demonstrated that the formation of CO2 in the reduction
of nitrogen oxide with methane occurs via two parallel
pathways. The rate of methane conversion into CO2 via
the first pathway is several times higher than that via the
second pathway. At 450°C in an excess of oxygen (CO
=
2
3.0%), the ratio between the rates of CO2 formation via
these pathways is approximately constant; it is practi-
cally independent of the NO and methane concentra-
tions in the mixture and equals ~2.3. This is consistent
with the ratio between the rates of reactions that occur
via NOδ2+ species and nitrite–nitrate complexes, which
were evaluated previously from isotope experiments
with 15N18O [24]. On this basis, we assume that the for-
mation of CO2 via the first (rapid) pathway results from
of formation of NOδ2+ species and a decrease in the
band intensities of zeolite hydroxyl groups [17]. More-
over, the results of experiments with labeled 15N18O
suggest that the oxygen exchange of CoOx clusters with
the mononitrosyl complexes of cobalt and nitrite–
nitrate complexes occurs much more rapidly than with
NOδ2+ species. Thus, it is reasonable to consider that
nitrite–nitrate complexes and NOδ2+ species are related
to different sites: the former are related to Co2+ ions
the interaction of methane with NOδ2+ species, whereas
CO2 is formed as a result of the interaction of methane (or CoOx clusters), whereas the latter are related to the
KINETICS AND CATALYSIS Vol. 45 No. 3 2004

442
SADOVSKAYA et al.
paired sites Co2+···OH localized at the interface
believe that the former (faster) reaction pathway occurs
between CoOx clusters and the zeolite lattice, which, on Co2+···OH sites, whereas the latter (slower) occurs
probably, contains strong proton sites. Therefore, we on Co2+ sites:
+[NOδ2+
]
[Z1–C…N]
[Z2–C…N]
CO2 (+ N2 + H2O) fast
CO2 (+ N2 + H2O) slow,
CH4
+[ONOx]
where [ONOx] refers to nitrite–nitrate complexes, and
Previously [23], we evaluated the number of paired
[Z1–C…N] and [Z2–C…N] refer to intermediates local- Co2+···OH sites in special isotope experiments with
ized on Co2+···OH and Co2+ sites, respectively.
labeled 15N18O, in which the concentrations of NO and
O2 were varied over a wide range. We found that, start-
ing with concentrations of NO and O2 equal to 0.6 and
1.5%, respectively, the concentration of NOδ2+ species
ceased to increase. Assuming that in this case the limit-
ing coverage of sites is reached, we can estimate the
number of Co2+···OH sites, which was approximately
10% of the total number of cobalt atoms in the catalyst.
Thus, taking the number of Co2+ sites equal to the
total number of cobalt atoms (2 × 1020 molecule/g) and
the number of Co2+···OH sites equal to 2 × 1019 mole-
cule/g, we calculated the rates of reaction via either of
the pathways referenced to the number of sites and the
coverages with [Z1–C…N] and [Z2–C…N] intermedi-
ates (Table 3). It can be seen that the turnover number
As mentioned above, a change in the concentration
of methane and NO has practically no effect on the ratio
between the rates of reaction via different pathways. At
the same time, Table 2 indicates a clear trend toward
decreasing the contribution of the former pathway to
the overall reaction rate as the concentration of oxygen
in the mixture was decreased. In other words, the
dependence of the rate of the reaction on Co2+···OH
sites on the concentration of oxygen was stronger than
that on Co2+ sites. A change in the temperature also
more strongly affected the rate of reaction on Co2+···OH
sites. Differences in the kinetics of the reaction via the
identified pathways became more clearly pronounced
on going to relative rates normalized to the number of
active sites, that is, the turnover numbers of the reac- of the reaction on Co2+···OH sites is greater than that on
tion.
Co2+ sites by a factor of ~20–25, whereas the coverage
Table 3. Turnover numbers of the reaction via different pathways, surface coverages with [Zi–C...N] complexes, and appar-
ent rate constants of their conversion
Initial concentration, vol %
Turnover number × 102, s–1
Surface coverage, %
k
app, s–1
T, °C
NO
O2
CH4
pathway 1
pathway 2
[Z1–C...N] [Z2–C...N] [Z1–C...N] [Z2–C...N]
450
0.3
0.3
0.3
0.3
0.3
0.3
0.2
0.3
0.35
0.3
0.3
0.3
3.0
3.0
3.0
0.3
0.6
3.0
3.0
3.0
3.0
3.0
3.0
3.0
0.3
0.4
0.6
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.95
1.25
1.50
0.90
0.80
1.25
1.50
1.25
1.40
0.40
1.25
3.70
0.04
0.055
0.07
0.05
0.05
0.055
0.07
0.055
0.06
0.02
0.055
0.14
0.8
1.1
1.3
1.4
1.0
1.1
1.1
1.1
1.0
1.0
1.1
0.3
0.4
1.18
1.14
1.15
0.64
0.80
1.14
1.36
1.14
1.40
0.40
1.14
12.30
0.10
0.11
0.09
0.07
0.08
0.11
0.10
0.11
0.11
0.04
0.11
0.25
0.5
0.75
0.7
450
450
0.65
0.5
0.7
0.5
0.55
0.5
400
450
480
0.5
0.55
KINETICS AND CATALYSIS Vol. 45 No. 3 2004

REACTION KINETICS AND MECHANISM OF SELECTIVE NO REDUCTION
443
kapp[Z –C...N], s–1
ln(kapp[Z –C..N]) [s–1]
i
i
3
1
1.2
1.0
0.8
0.6
0.4
2
2(×10)
1
0
–1
–2
–3
1
2
–4
0
0.5
1.0
1.5
2.0
2.5
3.0
1.30
1.35
1.40
1.45
1.50
Concentration of é2, vol %
1/T × 103, K–1
Fig. 5. Dependence of the apparent rate constant of the con-
Fig. 6. Arrhenius plots for the rate constants of conversion
version of [Z –C···N] intermediates formed via (1) the first
i
of [Z –C…N] intermediates formed via (1) the first and (2)
and (2) the second pathways on the concentration of oxy-
gen. Points and curves refer to experimental results and data
calculated from Eq. (3).
i
the second pathways of the reaction of selective nitrogen
oxide reduction with methane.
Co2+···OH are higher than those on Co2+ by more than
one order of magnitude. Moreover, note that the depen-
dence of the rate constants of conversion of [Zi–C…N]
intermediates upon the concentration of oxygen on
Co2+···OH sites is stronger than that on Co2+ sites
(Fig. 5). Based on these relationships, we can evaluate
the rate constant of the interaction of [Zi–C…N] inter-
mediates with oxygen. Representing the conversion of
[Zi–C…N] into CO2 as a sequence of elementary steps
at one of which the reaction with oxygen occurs and
of active sites with [Z1–C…N] and [Z2–C…N] interme-
diates is about 1% or lower. These low coverages sug-
gest that the rate constants of the steps of [Z1–C…N]
and [Z2–C…N] transformations are comparatively
high, and the steps of formation of [NOx] complexes
and/or the steps of [NOx] interaction with methane are
the rate-limiting steps of the reaction via both of the
pathways. As can be seen in Tables 2 and 3, among all
of the initial reactants (NO, O2, and CH4), the concen-
tration of methane has the greatest effect on the rate of
the reaction. Indeed, a twofold increase in the concen-
tration of CH4 increased the rate of the reaction by a
factor of 1.5–1.8, whereas approximately the same
change in the concentration of NO had practically no
effect on the rate of the reaction. As the concentration
of oxygen was increased by one order of magnitude, the
rates of the reaction increased by only 10–20%. This
allowed us to assume that the rate of the reaction via
either the former or the latter pathway is limited by the
expressing the rate of this step as ri = koi xθ[Z –C…N] CO
i
2
j
and the rates of all the other steps as ri = kijθ[Z –C…N]
i
j
(where kij is the rate constant of the jth step via the ith
pathway), the rate of the reaction via either of the path-
ways can be expressed as follows:
θ
[Zi–C…N]j
∑
----------------------------------
ri =
,
(2)
1
1
steps of methane interaction with NOδ2+ species and
--------------- +
---
∑
koi xCO
kij
2
nitrite–nitrate complexes. It is most likely that the rela-
tively weak dependence of the rate of the reaction on
the concentration of NO and O2 is due to the reversibil-
ity of the steps of formation of NOx complexes.
where
θ[Z –C…N] s the sum of the concentrations of
∑
i
j
all the C,N-containing intermediates formed via a given
pathway, that is, the value of θ[Z –C…N] , which is deter-
mined from experimental data. Equation (2) can be put
in the form
i
Although the steps of conversion of [Zi–C…N]
intermediates into CO2 are fast, a study of the kinetics
of these steps is important because it will provide an
opportunity to gain deeper insight into the reaction
mechanism. The ratio of the turnover number to the
coverage with [Zi–C…N] intermediates is nothing else
than the apparent rate constant of conversion of these
intermediates. The data in Table 3 indicate that the
apparent rate constants of [Zi–C…N] conversion on
ri
1
1
app
[Zi–C…N]


k
= ------------------------------ = 1 ⁄ --------------- +
--- . (3)
∑
ox
j


θ
ki CO
ki
[Zi–C…N]j
∑
2
Thus, we expressed the apparent rate constants of the
conversion of [Zi–C…N] intermediates in terms of the
rate constants of elementary steps and the concentra-
KINETICS AND CATALYSIS Vol. 45 No. 3 2004

444
SADOVSKAYA et al.
tion of oxygen. Equation (3) adequately describes the (second pathway) is the rate-limiting step of the reac-
experimental dependence of the rate constants of the tion via either of the two pathways. The rate of the reac-
tion on Co2+···OH sites is higher than that on Co2+ by
more than one order of magnitude. The main factors
responsible for the rapid occurrence of the reaction on
Co2+···OH sites are, first, the presence of strong acid
conversion of [Zi–C…N] intermediate compounds on
the concentration of oxygen (Fig. 5).
The found rate constants of the steps of interaction
with oxygen via the first and second pathways are k1ox
380 s–1 and ko2x = 50 s–1, respectively.
=
sites (which stabilize NOδ2+ species, which are more
reactive toward methane) at the interface with oxide-
like cobalt clusters and, second, dissociative oxygen
adsorption, which occurs at interfacial sites. The latter
circumstance is favorable for the more rapid conversion
of C,N-containing intermediates, which are formed in
the reaction of NOδ2+ with methane, into reaction prod-
ucts.
The higher value of the rate constant ko1x , as com-
pared with ko2x , could be due to the features of molecu-
lar oxygen adsorption. Previously [23], we found that
the mechanism of formation of NOδ2+ species includes
a step of dissociative oxygen adsorption. It is most
likely that this adsorption occurs at the interface
between cobalt oxide clusters and zeolite. We con-
cluded that oxygen did not undergo dissociative
adsorption on the clusters themselves. This conclusion
was based on the fact that an oxygen label, which rap-
idly entered into the composition of CoOx clusters, did
not appear in the gas-phase oxygen in the course of the
isotopic substitution of 15N18O for 14N16O. It is reason-
able to assume that the dissociatively adsorbed oxygen
ACKNOWLEDGMENTS
This work was performed within the framework of a
joint research program between the Boreskov Institute
of Catalysis, Siberian Division, Russian Academy of
Sciences and the Institut de Recherches sur la Catalyse
(Lyon, France) and supported by the Russian Founda-
tion for Basic Research (grant no. 00-03-22004).
species that reacts with NO to form NOδ2+ can partici-
REFERENCES
pate in the oxidation of [Zi–C…N] intermediate com-
pounds. In this case, the probability of this interaction
with [Z1–C…N] intermediates, which are also local-
ized at the interface between oxide clusters and zeolite,
is higher than that with [Z2–C…N]. TheArrhenius plots
of the rate constants of conversions of [Z1–C…N] and
[Z2–C…N] intermediates (Fig. 6) can also be inter-
preted in terms of these concepts. It can be seen that the
apparent activation energy via the first reaction path-
way (44 kcal/mol) is twice as high as that via the second
pathway (22 kcal/mol). The former of the estimated
values can correspond to the activation energy of the
step of dissociative oxygen adsorption, and the latter
can correspond, for example, to the activation energy of
the diffusion of adsorbed oxygen to active sites, where
nitrite–nitrate complexes are formed. Note that the
above considerations are preliminary, and the possibil-
ity of dissociative oxygen adsorption on the oxide-like
cobalt clusters cannot be completely excluded,
although the rates are lower than that on interfacial
sites. Additional detailed studies with the use of labeled
molecular oxygen should be performed in order to clar-
ify this question.
1. Vassallo, J., Miro, E., and Petunchi, J., Appl. Catal., B,
1995, vol. 7, p. 65.
2. Burch, R. and Scire, S., Appl. Catal., B, 1994, vol. 3,
p. 295.
3. Li, Y. and Armor, J.N., Appl. Catal., B, 1993, vol. 2,
p. 239.
4. Witzel, F., Sill, G.A., and Hall, W.K., J. Catal., 1994,
vol. 149, p. 229.
5. Petunchi, J.O. and Hall, W.K., Appl. Catal., B, 1993,
vol. 2, p. 17.
6. Shelef, M., Montreuil, C.N., and Jen, H.W., Catal. Lett.,
1994, vol. 26, p. 277.
7. Konin, G.A., Il’ichev, A.N., Matyshak, V.A., Khomen-
ko, T.I., Korchak, V.N., Sadykov, V.A., Doronin, V.P., Bun-
ina, R.V., Alikina, G.M., Kuznetsova, T.G., Paukshtis, E.A.,
Fenelonov, V.B., Zaikovskii, V.I., Ivanova, A.S.,
Beloshapkin, S.A., Rozovskii, A.Y., Tretyakov, V.F.,
Ross, J.R.H., and Breen, J.P., Top. Catal., 2001, vol. 16,
p. 193.
8. Tretyakov, V.F., Burdeynaya, T.N., Davydova, M.N.,
Matyshak, V.A., and Glebov, L.S., Top. Catal., 2001,
vol. 16, p. 243.
9. Cowan, A.D., Dumpelmann, R., and Cant, N.W.,
Thus, the study of the dynamics of 13C-label transfer
demonstrated that the reaction of NO reduction with
methane on CoZSM-5 in the presence of molecular
oxygen occurs via two parallel pathways: on Co2+ sites
(or CoOx clusters) and on paired Co2+···OH sites local-
ized at the interface of oxide-like cobalt clusters and
zeolite. The interaction of methane with NOδ2+ species
(first pathway) or surface nitrite–nitrate complexes
J. Catal., 1995, vol. 151, p. 356.
10. Li, Y., Slager, T.L., and Armor, J.N., J. Catal., 1994,
vol. 150, p. 388.
11. Bell, A.T., Catal. Today, 1997, vol. 38, p. 151.
12. Aylor, A.W., Lobree, L.J., Reimer, J.A., and Bell, A.T.,
Stud. Surf. Sci. Catal., 1996, vol. 101, p. 661.
13. Adelman, B.J., Beutel, T., Lei, G.-D., and Sachtler, W.M.H.,
J. Catal., 1996, vol. 158, p. 327.
KINETICS AND CATALYSIS Vol. 45 No. 3 2004

REACTION KINETICS AND MECHANISM OF SELECTIVE NO REDUCTION
445
14. Sadykov, V.A., Beloshapkin, S.A., Paukshtis, E.A., 20. Hadjiivanov, K., Saussey, J., Freysz, J.L., and Lavalley, J.C.,
Alikina, G.M., Kochubei, D.I., Degtyarev, S.P., Bulga-
kov, N.N., Veniaminov, S.A., Netyaga, E.V., Bunina, E.V.,
Kharlanov, A.N., Lunina, E.V., Lunin, V.V., Matyshak, V.A.,
and Rozovskii, A.Ya., Pol. J. Environ. Stud., 1997,
vol. 6, p. 21.
Catal. Lett., 1998, vol. 52, p. 103.
21. Iwamoto, M., Hidenori, Y., Mizuno, N., Zhang, W.-X.,
Mine,Y., Furukawa, H., and Kagawa, S., J. Phys. Chem.,
1992, vol. 96, p. 9360.
22. Beutel, T., Adelman, B.J., and Sachtler, W.M.H., Appl.
Catal., B, 1996, vol. 9, p. 1.
23. Sadovskaya, E.M., Suknev, A.P., Pinaeva, L.G., Gon-
charov, V.B., Bal’zhinimaev, B.S., Chupin, C., and Miro-
datos, C., J. Catal., 2001, vol. 201, p. 159.
24. Sadovskaya, E.M., Suknev, A.P., Goncharov, V.B.,
Sadykov, V.A., and Bal’zhinimaev, B.S., Proc. 9th Int.
Symp. on Heterogeneous Catalysis, 2000, p. 49.
15. Cowan, A.D., Cant, N.W., Haynes, B.S., and Nelson, P.F.,
J. Catal., 1998, vol. 176, p. 329.
16. Lombardo, E.A., Sill, G.A., D’Itri, J.L., and Hall, W.K.,
J. Catal., 1998, vol. 173, p. 440.
17. Pinaeva, L.G., Sadovskaya, E.M., Suknev, A.P., Gon-
charov, V.B., Sadykov, V.A., Balzhinimaev, B.S.,
Decamp, T., and Mirodatos, C., Chem. Eng. Sci., 1999,
vol. 54, p. 4327.
25. Shannon, S. and Goodvin, J., Chem. Rev., 1995, vol. 95,
no. 3, p. 679.
18. Hoost, T.E., Laframboise, K.A., and Otto, K., Catal.
Lett., 1995, vol. 33, p. 105.
26. Sadovskaya, E.M., Bulushev, D.A., and Bal’zhini-
maev, B.S., Kinet. Katal., 1999, vol. 40, no. 1, p. 61.
19. Valyon, J. and Hall, W.K., J. Phys. Chem., 1993, vol. 97, 27. Happel, J., Isotopic Assessment of Heterogeneous Catal-
p. 1204.
ysis, Orlando: Academic, 1986.
KINETICS AND CATALYSIS Vol. 45 No. 3 2004