Mechanism of the low-temperature interaction of hydrogen with α-oxygen on FeZSM-5 zeolite

Article abstract of DOI:10.1023/B:KICA.0000023792.95153.d8

The mechanism of a low-temperature reaction of hydrogen with the radical anion surface oxygen species (α-oxygen, O_α) formed by decomposing N₂O over FeZSM-5 zeolite was studied using kinetic and isotope techniques. It was found that the reaction is of first order with respect to hydrogen and the rate of the reaction is proportional to the concentration of O_α. The activation energy of the reaction, which was measured for H₂ or D₂ over a temperature range from +20 to -100°C, is equal to 3.2 or 5.3 kcal/mol, respectively. The reaction occurs with a considerable kinetic isotope effect (k_H/k _D), which varies over the range of 3.4-41 depending on the temperature. This fact indicates that the rate-limiting step of the reaction includes the dissociation of the hydrogen molecule. The temperature dependence of the isotope effect gave a value of 2.1 kcal/mol, which is close to the difference between the zero bond energies in the molecules of H₂ and D₂; this fact suggests that a tunnel effect does not significantly contribute to the reaction. The dissociative mechanism is consistent with data obtained by in situ IR spectroscopy. The interaction of hydrogen with α-oxygen is accompanied by the formation of new hydroxyl groups O _αH (absorption bands at 3635 and 3674 cm^-1) at the surface of the zeolite. The identification of these groups was supported by an isotope shift either on the replacement of H₂ by D₂ or on the replacement of ¹⁶O_α by O_α,. The stoichiometric ratio H₂:O_α, is consistent with the previously drawn conclusion on the paired arrangement of α-sites.

Full text of DOI:10.1023/B:KICA.0000023792.95153.d8

Kinetics and Catalysis, Vol. 45, No. 2, 2004, pp. 202–208. Translated from Kinetika i Kataliz, Vol. 45, No. 2, 2004, pp. 218–224.
Original Russian Text Copyright © 2004 by Dubkov, Starokon’, Paukshtis, Volodin, Panov.
CATALYTIC REACTION
MECHANISMS
Mechanism of the Low-Temperature Interaction
of Hydrogen with a-Oxygen on FeZSM-5 Zeolite
K. A. Dubkov, E. V. Starokon’, E. A. Paukshtis, A. M. Volodin, and G. I. Panov
Boreskov Institute of Catalysis, Siberian Division, Russian Academy of Sciences, Novosibirsk, 630090 Russia
Received October 9, 2002
Abstract—The mechanism of a low-temperature reaction of hydrogen with the radical anion surface oxygen
species (α-oxygen, éα) formed by decomposing N2O over FeZSM-5 zeolite was studied using kinetic and iso-
tope techniques. It was found that the reaction is of first order with respect to hydrogen and the rate of the reac-
tion is proportional to the concentration of éα. The activation energy of the reaction, which was measured for
ç2 or D2 over a temperature range from +20 to –100°ë, is equal to 3.2 or 5.3 kcal/mol, respectively. The reac-
tion occurs with a considerable kinetic isotope effect (kç/kD), which varies over the range of 3.4–41 depending
on the temperature. This fact indicates that the rate-limiting step of the reaction includes the dissociation of the
hydrogen molecule. The temperature dependence of the isotope effect gave a value of 2.1 kcal/mol, which is
close to the difference between the zero bond energies in the molecules of ç2 and D2; this fact suggests that a
tunnel effect does not significantly contribute to the reaction. The dissociative mechanism is consistent with
data obtained by in situ IR spectroscopy. The interaction of hydrogen with α-oxygen is accompanied by the for-
mation of new hydroxyl groups éαç (absorption bands at 3635 and 3674 cm–1) at the surface of the zeolite.
The identification of these groups was supported by an isotope shift either on the replacement of ç2 by D2 or
on the replacement of 16éα by 18éα. The stoichiometric ratio ç2 : éα is consistent with the previously drawn
conclusion on the paired arrangement of α-sites.
INTRODUCTION
The kinetics and mechanism of the surface reactions
of α-oxygen are difficult to study because of its rate of
reaction is too high, it is immeasurably high even in the
case of methane [11]. Moreover, the situation is com-
plicated by the strong adsorption of organic molecules
on the zeolite surface, which hinders reliable kinetic
measurements, particularly at low temperatures.
The identification and characterization of surface
oxygen species that participate in the oxidation of
organic molecules are of considerable interest for
understanding the mechanism of oxidation catalysis. In
recent years, studies of the reactivity and chemical
properties of a so-called α-form of oxygen, which par-
ticipates in selective oxidation reactions on FeZSM-5
zeolites, have attracted the particular attention of
researchers [1–7]. This species is formed by the decom-
position of N2O at α-sites, which are the reduced mono-
In preliminary experiments, we found that hydrogen
reacts with α-oxygen at room temperature ~20 times
slower than methane. It is likely that this unusual ratio
between the reactivity of hydrogen and methane can be
explained by the special chemical properties of Oα. It is
of interest that an analogous ratio was observed by
Lipatkina et al. [14] in the interaction of hydrogen and
methane with O– radical anions on the surface of
reduced metal oxides. In addition to the lower reactivity
of hydrogen toward α-oxygen, hydrogen is weakly
adsorbed on the surface of zeolites. This fact makes it
possible to perform a reliable study of the kinetics and
mechanism of the low-temperature interaction of éα
with hydrogen, and this work is devoted to this study.
nuclear iron complexes Fe2α+ with a paired arrangement
in the zeolite matrix [8]. As found by Mössbauer spec-
troscopy and thermal-desorption measurements,
α-oxygen exhibits the nature of radical anions, and its
formation is accompanied by the oxidation of Fe2+ to
Fe3+ in the active site [8, 9]
.
Fe2α+ + N2O
Fe3α+Oα + N2 .
(I)
α-Oxygen can be formed in a high concentration,
and it is stable on the surface of FeZSM-5 zeolites at
temperatures lower than 300°ë; this provides a good
opportunity to study its properties. α-Oxygen acts
EXPERIMENTAL
Samples. The FeZSM-5 zeolite (Si/Al = 23; CFe =
much like the active oxygen of methane monooxygen- 0.3 wt % Fe) prepared by the impregnation of commer-
ase enzymes, oxidizes alkanes, cycloalkanes, and aro- cial ZSM-5-A zeolite in the H form (AngarskNefte-
matic hydrocarbons at room temperature with the OrgSintez) with a solution of FeCl3 was used for kinetic
selective formation of hydroxyl-containing products in studies. After the impregnation, the sample was cal-
all cases [10–13].
cined in a vacuum with a gradually increasing temper-
0023-1584/04/4502-0202 © 2004 MAIK “Nauka /Interperiodica”

MECHANISM OF THE LOW-TEMPERATURE INTERACTION OF HYDROGEN
203
ature up to 500°C and then in oxygen at 500°C. To
a-Oxygen concentration measurements. The
increase the concentration of α-sites, the resulting cat- concentration of α-oxygen [éα] on the surface of the
alyst was additionally activated by calcination in a vac- FeZSM-5 samples was calculated from the amount of
uum at 900°ë (1 h) and finally exposed to é2 (5 Torr) N2 liberated (N2O consumed) in reaction (I). To refine
at 550°ë. The concentration of α-sites in the activated the value of [éα], an isotope-exchange reaction
sample was Cα = 1.36 × 1019 site/g.
between 16Oα and 18O2 was performed at room temper-
ature. Both of the techniques gave similar results. The
average value obtained with the use of these techniques
was taken as the true value. The measurement proce-
dure for [éα] was described in detail elsewhere [11].
A sample of FeZSM-5-B zeolite (Si/Al = 72; CFe =
0.53 wt % Fe) in the H form, which was prepared by
hydrothermal synthesis by the introduction of iron as
FeCl3 into a parent gel, was used in IR-spectroscopic
studies [15]. The concentration of α-sites in the acti-
vated sample was Cα = 1.9 × 1019 site/g.
Kinetic measurements. The rate of the reaction of
hydrogen with α-oxygen was determined by the graph-
Vacuum unit. Experiments on the decomposition of ical differentiation of the initial portion of the time
N2O and on the interaction of ç2 (D2) with α-oxygen dependence of hydrogen pressure in the course of the
were performed in a stainless steel static vacuum unit reaction. The pressure change at this segment did not
(pressures up to 10–7 Torr). The vacuum unit was exceed 5% of the starting value. The concentration of
equipped with a cryostat, which allowed us to control α-oxygen on the surface also decreased insignificantly
the reactor temperature over a range from 0 to –150°ë during this time. The rate of reaction per gram of sam-
with a high accuracy. Partial gas pressures in the reac- ple (wm) was calculated using the following expression:
tion volume were measured using a PPT Residual Gas
wm = (dP/dt)A(1/m), molecule s–1 g–1,
(1)
Analyzer mass spectrometer (MKS Instruments) and a
Baratron absolute pressure gage (MKS Instruments)
with a linearity range from 2 × 10–3 to 2 Torr. Before the
beginning of experiments, a sample (0.14 g) was placed
in a quartz microreactor, whose volume (5 cm3) is neg-
ligible compared to the reaction volume of the unit
(730 cm3). An important property of the reactor is that
it can be easily isolated from the reaction volume using
a valve. This allows the gas phase to be replaced within
the closed reactor without disturbing the adsorption
state of the sample surface. This significantly improved
the accuracy of the kinetic and adsorption measure-
ments.
where dP/dt is the slope of the line tangent to a curve of
P(t) at a given pressure; A = 2.5 × 1019 molecule/Torr is
a conversion factor; and m is the sample weight, g.
IR spectroscopy. A specially made quartz cell with
a high-vacuum stopcock was used for in situ studies by
diffuse-reflectance IR spectroscopy. A sample (0.5 g)
was placed in the cell and subjected to successive treat-
ments in the vacuum unit according to the above proce-
dure. After each of the treatments, the stopcock was
closed, the sample was cooled to 20°ë, and the cell was
detached from the unit. The spectra were measured at
room temperature with the use of a Shimadzu 8300
Fourier spectrometer in the region 2100–6000 cm–1
with a resolution of 2 cm–1. To accumulate the spectra,
100 scans were used. The spectra were converted from
a reflection scale to the Kubelka–Munk scale F(R) =
(1 – R)2/2R.
Sequence of experiments. The experiments were
performed in accordance with the following procedure:
(1) The standard treatment of a sample in a vacuum
at 530°ë for 1 h and then in é2 (1 Torr) at 530°ë
for 1 h.
(2) The formation of α-oxygen by the decomposi-
tion of N2O at an initial pressure of 0.2 Torr at 250°ë
for 25 min.
RESULTS AND DISCUSSION
(3) The reaction of α-oxygen with ç2 (D2) at tem-
peratures from +20 to –100°ë.
1. Reaction of Hydrogen with α-Oxygen
To determine a temperature region appropriate for
kinetic measurements, we initially studied the adsorp-
tion of ç2 on a sample of FeZSM-5-A over a tempera-
ture range from +20 to –150°ë. These experiments
demonstrated that the adsorption of hydrogen was neg-
ligibly small at temperatures above –100°ë.
The treatment allowed us to standardize the state of
the catalyst and to purify it from possible organic impu-
rities. To form Oα on the surface of FeZSM-5, after the
standard treatment, the sample was cooled in é2
(1 Torr) to 250°ë; the reactor was closed, and é2 in the
reaction volume was replaced by nitrous oxide at a
pressure of 0.2 Torr. After opening the reactor, the
decomposition of N2O was performed for 25 min until
the completion of reaction (I).
Figure 1 illustrates the results of two comparative
experiments on the interaction of ç2 with the surface of
the FeZSM-5-A sample at –100°ë. In the first experi-
To perform the reaction of α-oxygen with hydrogen, ment, hydrogen reacted with the zeolite after a standard
the reactor was closed after the decomposition of N2O treatment (Fig. 1a), whereas it in the second experiment
and cooled to the reaction temperature; the gas phase in reacted with the zeolite surface after the formation of
the reaction volume was replaced by ç2 (D2). The reac- α-oxygen from N2O on the surface by reaction (I)
tion of hydrogen with α-oxygen occurred after opening (Fig. 1b). It can be seen that only weak physical adsorp-
the reactor.
tion of ç2 was observed in the absence of α-oxygen.
KINETICS AND CATALYSIS Vol. 45 No. 2 2004

204
DUBKOV et al.
lnwm [molecule s–1 g–1]
PH , Torr
2
(a)
A
40
38
36
34
32
0.06
0.04
0.02
1
2
0
3.5
4.0
4.5
5.0
5.5
6.0
A
103/T, K–1
(b)
0.06
0.04
0.02
Fig. 2. Arrhenius plots for the interactions of α-oxygen with
(1) H and (2) D .
2
2
Based on these data, the temperature dependence of the
rate of ç2 oxidation was plotted on the Arrhenius coor-
dinates in Fig. 2 (curve 1). It can be seen that the experi-
mental results are adequately described by a linear func-
tion with an activation energy of 3.2 ± 0.3 kcal/mol.
In the case of D2, the activation energy calculated
from the Arrhenius function (Fig. 2, curve 2) is higher
and equals 5.3 ± 0.5 kcal/mol. Note that lower values of
activation energies were found for the reactions of ç2
and D2 with the é– radical anion oxygen species on the
surface of reduced V/SiO2, Mo/SiO2, and W/SiO2 cata-
lysts; these activation energies were approximately equal
and comprised 1.0 and 1.3 kcal/mol, respectively [14].
0
0
20 40 60 80 100 120 140 160
t, min
Fig. 1. Interaction of ç with FeZSM-5-A zeolite at
2
−100°ë: (a) without α-oxygen; (b) in the presence of 1.36 ×
19
–1
10 (atom O) (g α-oxygen) . A indicates the moment in
time at which the reactor was opened.
Another behavior was observed in the presence of
1.36 × 1019 atom O (g α-oxygen)–1 on the surface
(Fig. 1b). In this case, after opening the reactor, consid-
erable irreversible absorption of ç2 due to its consump-
tion in the reaction with α-oxygen was observed.
2. Reaction Order with Respect
to Hydrogen and α-Oxygen
To find the form of the rate equation that describes
the rate of the interaction of hydrogen with α-oxygen,
we initially determined the order of reaction with
respect to hydrogen. In this case, it was methodologi-
cally easier to use the reaction with deuterium because
the rate of its oxidation is lower than that of ç2 oxida-
tion (Table 1). For this purpose, we performed experi-
ments on the stoichiometric oxidation of D2 at 0°ë with
α-oxygen ([éα] = 1.36 × 1019 (atom O)/g) by varying
the initial pressure of hydrogen over the range of
0.018–0.143 Torr. Figure 3 demonstrates the effect of
the D2 pressure on the reaction rate. It can be seen that
the experimental points are adequately described by a
linear function, which suggests that the reaction is of
first order with respect to deuterium (hydrogen).
Table 1 summarizes the rates of ç2 and D2 reactions
with α-oxygen at temperatures from +20 to –100°ë.
Table 1. The temperature dependence of the rates of H2 and D2
oxidation with α-oxygen on FeZSM-5-A
wm, H
,
wm, D ,
2
2
T, °C
kH/kD
molecule s–1 g–1 molecule s–1 g–1
+20
0
3.25 × 1017
1.93 × 1017
8.50 × 1016
3.05 × 1016
1.76 × 1016
7.4 × 1015
9.6 × 1016
4.1 × 1016
1.08 × 1016
2.23 × 1015
7.5 × 1014
1.8 × 1014
3.4
4.7
–20
–50
–70
–100
7.9
For the complete kinetic description, it was also of
interest to find the dependence of the reaction rate on
the concentration of α-oxygen on the zeolite surface.
For this purpose, we oxidized deuterium at 0°ë varying
the initial amount of éα on the surface within the limits
of one order of magnitude. To form éα, an appropriate
13.7
23.4
41.1
19
Note: P0, H = P0, D = 0.056 Torr; [O ] = 1.36 × 10 (atom O)/g.
α
2
2
KINETICS AND CATALYSIS Vol. 45 No. 2 2004

MECHANISM OF THE LOW-TEMPERATURE INTERACTION OF HYDROGEN
205
wm, D × 10–16, molecule s–1 g–1
wm, D × 10–16, molecule s–1 g–1
2
2
12
4
3
2
1
9
6
3
0
0.04
0.08
0.12
0.16
P, Torr
0
0.5
1.0
1.5
[Oα] × 10–19, (atom é)/g
Fig. 3. Effect of the D pressure on the rate of its interaction
Fig. 4. Effect of the α-oxygen concentration on the rate of
its interaction with D at 0°ë.
2
with α-oxygen at 0°ë.
2
amount of N2O was decomposed at 250°ë before the reac-
tion with D2.As can be seen in Table 2, the reaction rate wm
changed by approximately one order of magnitude and
increased from 5 × 1015 to 4.1 × 1016 molecule s–1 g–1 as
[éα] increased from 1.4 × 1018 to 1.36 × 1019 (atom O)/g.
Data shown in Fig. 4 indicate that the rate is propor-
tional to the concentration of éα on the sample surface.
Thus, the apparent rate of the reaction of hydrogen
with α-oxygen over the test temperature region can be
described by the following rate equation:
3. Kinetic Isotope Effect
Data given in Table 1 allowed us to calculate the
kinetic ç2–D2 isotope effect in the reaction with α-oxy-
gen, which provides important information on the reac-
tion mechanism. It can be seen in Table 1 that the rates
of H2 and D2 oxidation under identical conditions are
very different; that is, the reaction occurs with a high
isotope effect (kH/kD), which changed from 3.4 to 41 as
the temperature was decreased from +20 to –100°ë.
This demonstrates that bond cleavage in the hydrogen
molecule occurs at the rate-limiting step of the reaction.
wm, H
2
(2)
= Ak0, H e–E /RT PH [Oα ] molecule s–1 g–1,
a
A considerably lower isotope effect equal to 4.3 at
−100°C was found in the reaction of hydrogen with é–
on the surface of V/SiO2 [14]. Taking into account the
high values of the kinetic isotope effect in our case, we
believe that the reaction of H2 with α-oxygen can occur
with a considerable tunnel effect contribution, espe-
cially at low temperatures. To test this hypothesis, we
plotted the temperature dependence of the kinetic iso-
tope effect in the Arrhenius coordinates based on the
2
2
where A = 2.5 × 1019 molecule/Torr is a proportion-
ality factor, k0, H = 3.2 × 10−18 (atom O)–1 s–1 is a pre-
2
exponential factor; Ea, H = 3.2 ± 0.3 kcal/mol is the
2
activation energy; and PH is the pressure of hydrogen,
2
Torr; [éα] is the concentration of α-oxygen, (atom O)/g. In
the case of deuterium, k0, D = 3.2 × 10−17 (atom O)–1 s–1
2
and Ea, D = 5.3 ± 0.5 kcal/mol. Differences between
2
Table 2. Effect of the concentration of α-oxygen on the rate
of D2 oxidation on a FeZSM-5-A sample at 0°C
the preexponential factors for H2 and D2 are within the
accuracy of the determination from the corresponding
Arrhenius functions.
w
× 10–16
,
w
× 10–3,
[Oα] × 10–19
(atom O)/g
,
m, D2
α, D2
molecule s–1 g–1 molecule (atom O)–1 s–1
The ratio of the reaction rate (wm) to the value of
[éα] gives the activity of a single éα atom in the oxida-
tion of deuterium (wα). As can be seen in Table 2, this
value is approximately constant and does not depend on
the surface concentration of éα. This indicates that
α-oxygen is equivalent in its reactivity. At the same
time, we can note a small deviation of experimental
points from a straight line for the rates of H2 and D2 oxi-
dation in Fig. 2. This may be a consequence of a non-
uniformity of éα, which becomes noticeable at low
temperatures.
0.14
0.48
0.67
0.85
1.02
1.36
0.50
1.74
2.36
2.78
3.55
4.10
3.6
3.6
3.5
3.3
3.6
3.0
Note: P0, D = 0.056 Torr.
2
KINETICS AND CATALYSIS Vol. 45 No. 2 2004

206
ln(kH/kD)
DUBKOV et al.
F(R)
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0
4.0
3.5
3.0
2.5
2.0
1.5
1.0
1
2
3
4
3.0
3.5
4.0
4.5
5.0
5.5
6.0
103/T, K–1
3000
3200
3400
3600
3800
ν, cm–1
Fig. 5. Arrhenius plot for the isotope effect in the reaction
of hydrogen with α-oxygen.
Fig. 6. IR spectra of FeZSM-5-B zeolite in the region of OH
vibrations: (1) after standard treatment, (2) after the reac-
16
tion of ç with α-oxygen ( é ) at 20°ë, (3) difference
data given in Table 1 (Fig. 5). The difference between
2
α
spectrum, and (4) difference spectrum after the reaction of
ç with α-oxygen ( é ) at 20°ë.
the activation energies of D2 and H2 oxidation ∆Ea =
18
2
α
Ea, D – Ea, H , which was found from the slope of this
2
2
function, was 2.1 ± 0.3 kcal/mol, which is close to the
difference between the zero bond energies in D2 and H2
molecules, which is equal to 1.7–1.8 kcal/mol [17].
This indicates that the reaction of α-oxygen with
hydrogen, as well as with methane [16], occurs without
a considerable contribution from the tunnel effect. Note
that high values of the kinetic isotope effect are also
typical of oxidation reactions with the participation of
the active oxygen of methane monooxygenases [18].
vibrations. This demonstrates that the reaction of ç2
with α-oxygen results in the formation of two types of
new OH groups on the zeolite surface. Note that these
results are qualitatively consistent with the published
data [23].
We performed an additional experiment to clarify
whether α-oxygen atoms are the constituents of the
resulting OH groups. In this experiment, before the
reaction with hydrogen, α-oxygen was enriched in the
18O isotope (90% degree of enrichment) by the
4. Mechanism of the Reaction of H2 with α-Oxygen
18
exchange with é2 heavy oxygen at room tempera-
ture [11]
We used IR spectroscopy in order to reveal a more
detailed mechanism for the reaction and to identify the
compounds formed on a zeolite surface as a result of
this reaction. Figure 6 demonstrates the IR spectra of
FeZSM-5-B zeolite in the region of OH-group stretch-
ing vibrations. The spectrum of an initial sample after
standard treatment (Fig. 6, spectrum 1) exhibited three
main absorption bands. They belong to the vibrations of
intrinsic OH groups of the zeolite: terminal Si–OH
groups (3470 cm–1), bridging Al–OH–Si groups
(3610 cm–1) and OH groups bound to extralattice Al
species (3670 cm–1) [19–22]. After the formation of
α-oxygen ([éα] = 1.9 × 1019 (atom O)/g) by the decom-
position of N2O at 250°ë, the spectrum remained
unchanged.
(16é)α + 18é2 = (18é)α + 16é18é.
(II)
After the isotopic exchange of 16éα with 18éα followed
by the reaction with ç2, the absorption bands of the OH
groups shifted to the low-frequency region to give a
new intense absorption band at 3663 cm–1 with a shoul-
der at 3621 cm–1 (Fig. 6, spectrum 4). The occurrence
of this isotope shift demonstrates that éα atoms are the
constituents of the resulting OH groups.
Figure 7 demonstrates analogous spectra obtained
after the reaction of D2 with α-oxygen. It can be seen
that on the replacement of ç2 by D2 the absorption
The spectrum qualitatively changed after the inter- bands of the resulting hydroxyl groups shifted to the
action of ç2 with the zeolite surface on which α-oxy- low-frequency region to give new absorption bands at
gen was present (Fig. 6, spectrum 2). As can clearly be 2712 and 2681 cm–1 (Fig. 7, spectra 2, 3). Evidently,
seen in the difference spectrum (Fig. 6, spectrum 3), these absorption bands belong to surface OαD groups,
obtained by subtracting the spectrum of an initial sam- which were formed in the reaction of D2 with α-oxy-
ple, a new intense absorption band at 3674 cm–1 with a gen. This indicates that the reacting hydrogen is the
shoulder at 3635 cm–1 appeared in the region of OH source of hydroxyl groups on the zeolite surface.
KINETICS AND CATALYSIS Vol. 45 No. 2 2004

MECHANISM OF THE LOW-TEMPERATURE INTERACTION OF HYDROGEN
207
of methane reaction with Oα [16]. The reaction equa-
F(R)
1.2
tion can be represented as follows:
.
H2 + 2(O )α
2(OH)α.
(III)
1.0
Based on the kinetic measurements, we believe that
reaction (III) actually includes the following two ele-
mentary steps:
0.8
0.6
0.4
0.2
0
1
.
.
k1
H2 + (O )α
(OH)α + H ,
(IV)
(V)
.
.
2
3
k2
H + (O )α
(OH)α .
At the former step, the ç2 molecule reacts with the éα
atom; this reaction is accompanied by H-atom abstrac-
tion and the formation of the (OH)α group. At the latter
step, the H atom rapidly reacts with the nearest Oα
atom. The high isotope effect indicates that the former
step is slower, and it limits the rate of reaction. As a
result, the reaction is of first order with respect to both
reactants.
2600 2800 3000 3200 3400 3600 3800
ν, cm–1
Fig. 7. IR spectra of FeZSM-5-B zeolite in the regions of
OH and OD vibrations: (1) after standard treatment and
α-oxygen formation, (2) after the reaction of D with
α-oxygen ( é ) at 20°ë, and (3) difference spectrum.
2
16
α
According to the above mechanism, α-oxygen
should react with hydrogen in a stoichiometric ratio of
2 : 1. Table 3 summarizes data on the stoichiometry of
ç2 and D2 reactions with éα-oxygen at various temper-
atures. It can be seen that the experimental stoichiomet-
ric ratio is lower than 2; on the average, it is 1.6 : 1. An
analogous result was obtained previously [13]. It is
likely that the underestimated stoichiometric ratio
results from the fact that a portion of the hydrogen
atoms formed at step (IV) were bound to surface zeolite
sites other than éα It is believed that α-sites should be
located in sufficiently close proximity to each other in
order for reaction (III) to proceed. This is consistent
with the previously drawn conclusion on the paired
arrangement of α-sites [8].
It is believed that the formation of OαH groups of
two types is associated with the presence of two types
of α-sites on the surface of FeZSM-5 zeolite. These
sites differ in the coordination of constituent Fe2+ ions,
and they can be detected by Mössbauer spectroscopy [8].
Thus, according to IR-spectroscopic data, the inter-
action of hydrogen with α-oxygen occurs by a dissocia-
tive mechanism, which is analogous to the mechanism
Table 3. Stoichiometry of the reactions of H2 and D2 with α-
oxygen on FeZSM-5-A
Despite its radical–anion nature, the α-form of oxy-
gen differs essentially from other reported radical–
anion é– species on reduced metal oxides [9, 24]. The
α-form is generated only from N2O on active Fe2+ com-
plexes in zeolites (α-sites), and it can selectively oxidize
various hydrocarbons. The observed differences in the
kinetics of hydrogen oxidation with the participation of
these oxygen species (higher values of the activation ener-
gies and kinetic isotope effects in the case of éα) also sug-
gest considerable differences in their properties.
Amountofreacting Amount of Oα on Stoichiometric
T, °C H2 or D2 × 10–18, thesurface×10–19
,
ratio
molecule/g
(atom O)/g
Oα : H2 (Oα : D2)
H2 + (Oα) reaction
–100
–70
0
8.56
7.74
7.85
8.60
1.36
1.6
1.6
1.6
1.6
1.25
1.23
REFERENCES
20
1.36
1. Sobolev, V.I., Panov, G.I., Kharitonov, A.S., Romanni-
kov, V.N., Volodin, A.M., and Ione, K.G., J. Catal., 1993,
vol. 139, p. 435.
D2 + (Oα) reaction
2. Panov, G.I., Sobolev, V.I., and Kharitonov, A.S., Appl.
Catal., 1993, vol. 98, p. 1.
–70
0
7.77
7.52
7.65
7.70
1.25
1.24
1.27
1.33
1.6
1.7
1.7
1.7
3. Filatov, M.J., Pelmenschikov, A.G., and Zhidomi-
rov, G.M., J. Mol. Catal., 1993, vol. 8, p. 243.
4. Vereshchagin, S.N., Kirik, N.P., Shishkina, N.N., and
0
Anshits, A.G., Catal. Lett., 1998, vol. 56, p. 145.
5. Voskoboinikov, T.V., Chen, H.-Y., and Sachtler, W.M.H.,
20
Appl. Catal., 1998, vol. 19, p. 279.
KINETICS AND CATALYSIS Vol. 45 No. 2 2004

208
DUBKOV et al.
6. Yoshizawa, K., Shiota, Y., Yumura, T., and Yamabe, T., 15. US Patent 5077026, 1991.
J. Phys. Chem. B, 2000, vol. 104, p. 734.
16. Dubkov, K.A., Sobolev, V.I., Talsi, E.P., Rodkin, M.A.,
Watkins, N.H., Shteinman, A.A., and Panov, G.I., J. Mol.
Catal., 1997, vol. 123, p. 155.
7. Ribera, A., Arends, I.W.C.E., de Vries, S., Perez-
Ramirez, J., and Sheldon, R.A., J. Catal., 2000, vol. 195,
p. 287.
8. Dubkov, K.A., Ovanesyan, N.S., Shteinman, A.A.,
Starokon, E.V., and Panov, G.I., J. Catal., 2002, vol. 207,
p. 341.
17. Ozaki, A., Isotopic Studies of Heterogeneous Catalysis,
Moscow: London: Academic, 1977.
18. Khenkin, A.M. and Shilov, A.E., New J. Chem., 1989,
vol. 13, p. 659.
9. Starokon, E.V., Dubkov, K.A., and Panov, G.I., Top.
Catal., 2003, vol. 23, nos. 1–4, p. 137.
19. Jacobs, P.A. and von Ballmoos, R., J. Phys. Chem.,
1982, vol. 86, p. 3050.
10. Kharitonov,A.S.,Aleksandrova, T.N., Panov, G.I., Sobo-
lev, V.I., Sheveleva, G.A., and Paukshtis, E.A., Kinet.
Katal., 1994, vol. 35, no. 2, p. 296.
11. Dubkov, K.A., Sobolev, V.I., and Panov, G.I., Kinet.
Katal., 1998, vol. 39, no. 1, p. 79.
12. Rodkin, M.A., Sobolev, V.I., Dubkov, K.A., Watkins, N.H.,
and Panov, G.I., Stud. Surf. Sci. Catal., 2000, vol. 130,
p. 875.
13. Dubkov, K.A., Paukshtis, E.A., and Panov, G.I., Kinet.
Katal., 2001, vol. 42, no. 2, p. 230.
20. Quin, G., Zheng, L., Xie,Y., and Wu, C., J. Catal., 1985,
vol. 95, p. 609.
21. Kustov, L.M., Kazansky, V.B., Kubelkova, L., and Jiru, P.,
J. Phys. Chem., 1987, vol. 91, p. 5247.
22. Bordiga, S., Platero, E.E.,Arean, C.O., Lamberti, C., and
Zecchina, A., J. Catal., 1992, vol. 137, p. 179.
23. Zholobenko, V., Mendeleev Commun., 1993, vol. 2,
p. 67.
24. Panov, G.I., Dubkov, K.A., and Paukshtis, E.A., Cataly-
sis by Unique Metal Ion Structures in Solid Matrices,
Centi, G., Wichterlowa, B., and Bell, A., Eds., Dor-
drecht: Kluwer Academic, 2001, p. 149.
14. Lipatkina, N.I., Shvets, V.A., and Kazanskii, V.B., Kinet.
Katal., 1978, vol. 19, no. 4, 979.
KINETICS AND CATALYSIS Vol. 45 No. 2 2004