Mechanism and kinetics of hydrogen oxidation on silver

Article abstract of DOI:10.1007/s11172-012-0316-y

The kinetic study of the stationary oxidation of hydrogen on silver was carried out. The dependences of the reaction rate on the ratios of hydrogen and oxygen in the reaction mixture under a constant total pressure of 1000 Pa were obtained at 423, 448, and 473 K. The reaction seems to proceed via the intermediate formation of surface hydroxyls. A kinetic equation of the reaction that satisfactorily described the experimental data was derived.

Full text of DOI:10.1007/s11172-012-0316-y

Russian Chemical Bulletin, International Edition, Vol. 61, No. 12, pp. 2225—2229, December, 2012
2225
Mechanism and kinetics of hydrogen oxidation on silver
E. V. Dokuchits,a,b A. V. Khasin, and A. A. Khassin
a
a
aG. K. Boreskov Institute of Catalysis, Siberian Branch of the Russian Academy of Sciences,
5
prosp. Akad. Lavrent´eva, 630090 Novosibirsk, Russian Federation.
Eꢀmail: oschtan@catalysis.ru
b
Novosibirsk State University,
2
ul. Pirogova, 630090 Novosibirsk, Russian Federation
The kinetic study of the stationary oxidation of hydrogen on silver was carried out. The
dependences of the reaction rate on the ratios of hydrogen and oxygen in the reaction mixture
under a constant total pressure of 1000 Pa were obtained at 423, 448, and 473 K. The reaction
seems to proceed via the intermediate formation of surface hydroxyls. A kinetic equation of the
reaction that satisfactorily described the experimental data was derived.
Key words: silver, oxygen, kinetics of hydrogen oxidation.
Catalytic oxidation of hydrogen with oxygen on silver
is interesting as a model redox catalytic reaction. At the
same time, the reaction is practically significant for the
solution of problems of power engineering that arise, in
particular, in the development of fuel cells.
determination of the silver surface using this isotherm. Oxygen
dissolved in the nearꢀsurface layer of silver practically cannot be
removed from the metal volume under the studied conditions
8
and exerts no effect on the adsorption equilibrium of oxygen.
9
The reaction was studied by the flow "differential" method
under the total atmospheric pressure of gaseous mixtures conꢀ
taining hydrogen, oxygen, and helium. Experiments were carꢀ
ried out at 423, 448, and 473 K under the constant pressure of the
reaction mixture of hydrogen and oxygen: РН2 + РО2 = 1000 Pa
(РН2 and РО2 are the partial pressures of hydrogen and oxygen,
respectively, in the gaseous mixture). Reaction mixtures of reꢀ
quired compositions were prepared using a system consisting of
three reductors of gas dosage that provided a constant rate of the
gaseous mixture passing through the reactor. In all experiments,
Based on the early studies1 it was concluded that the
reaction follows the mechanism including two consequent
steps: chemisorption of oxygen and interaction of gaseous
hydrogen with adsorbed oxygen. However, the subsequent
works showed that the mechanism is more complicated
—3
4
,5
and includes the intermediate formation of hydroxyls.
The purpose of this work is to study the mechanism
and kinetics of the catalytic oxidation of hydrogen. To
describe the kinetic regularities, we used the results of
previous works on separate studies of particular steps of
–
1
the feeding rate of the gaseous mixture was 3 L h
.
The reactor was made of quartz as a cylinder with the bottom
permeable for the gas (the bottom was made of superthin silica
fiber). The catalyst was deposited on the bottom. One layer with
the catalyst granules uniformly covering all the sectional area of
the reactor was placed in the reactor. An optimum geometry of
the reactor provided the uniform distribution of the reaction
mixture flow over its cross section.
4
—6
the process of hydrogen oxidation.
Experimental
The silver catalyst was prepared by the reduction of silver
oxide with pure hydrogen. Silver oxide was precipitated from an
aqueous solution of silver nitrate (99.9 wt.%, reagent grade,
GOST 1277ꢀ75) with potassium hydroxide (special purity grade).
After precipitation, silver oxide was thoroughly washed with disꢀ
tilled water and dried. The weight of the studied sample was 1.28 g,
the silver granule size was 0.5—1.0 mm, and the specific surface
area was 0.75 m g . The value of specific surface area was
determined from oxygen adsorption at 473 K using the earlier7
obtained isotherm of oxygen adsorption on a similar silver samꢀ
ple. The oxygen adsorption isotherm on silver was determined
by both direct measurement of volumes of adsorbed oxygen and
volumes of hydrogen consumed in the reaction with adsorbed
oxygen. Therefore, the used isotherm of oxygen adsorption on
silver describes the equilibrium between gaseous oxygen and oxꢀ
ygen adsorbed on the silver surface, which allowed a fairly exact
The reaction proceeded at very low changes in partial presꢀ
sures of the starting substances: the conversion of the reaction
mixture in experiments did not exceed 1%. A similar method of
9
the reaction performance is named "differential." At the same
time, a necessary accuracy of determination of reaction rate was
achieved by the freezing out and accumulation of water formed
in the course of the reaction. Water was frozen out from
the reaction mixture flow leaving the reactor in a spiral sampler
cooled with liquid nitrogen. The sampler was made of a stainless
steel capillary with an internal diameter of 2.5 mm. The length
of the sampler directly cooled with liquid nitrogen was 50 cm.
The rest part of the sampler was heated to a constant temperaꢀ
ture (473 K) that was maintained inside the experimental setup
and chromatograph. At first the sampler was cooled with liquid
nitrogen, rapidly in 1 min the reaction mixture was introduced
into the sampler; the duration of delivery of the reaction mixture
2
–1
Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 12, pp. 2206—2210, December, 2012.
066ꢀ5285/12/6112ꢀ2225 © 2012 Springer Science+Business Media, Inc.
1

2226
Russ.Chem.Bull., Int.Ed., Vol. 61, No. 12, December, 2012
Dokuchits et al.
ranged from 15 to 45 min, depending on the composition of the
reaction mixture and, therefore, on the expected rate of accuꢀ
mulation of water formed. When the sampling was finished, the
sampler was rapidly heated with silicone oil to 473 K, and one
minute after the start of heating the sample was transferred to
the chromatograph. The carrier gas (helium) and a mixture of
gases were fed through the sampler and chromatograph in such a
way that a continuous flow of helium through the chromatoꢀ
graph was ensured, and a flow of helium or the analyzed mixture
could be introduced through the sampler at particular moments.
Due to this, when the current sampling terminates, the sampler
was filled with helium and the next sampling start to the moment
of beginning of the next analysis.
Gaseous mixtures were analyzed on a Kristallꢀ2000 chroꢀ
matograph using a heat conductivity detector. Helium served as
a carrier gas because it ensures a high sensitivity of the detector
to the analyzed gases. Gases and mixtures of gases were subjectꢀ
ed to the mandatory standard purification: oil was removed using
Petryanov´s filter, carbon dioxide and water were removed using
columns packed with zeolite and active carbon. Hydrogen and
a mixture of hydrogen with helium were also purified from oxyꢀ
gen in a column with the nickel—chromium catalyst. The conꢀ
tent of water vapor was determined using a column packed with
the PorapakꢀT sorbent, and the amounts of oxygen and hydroꢀ
gen were determined on columns packed with zeolites. The comꢀ
positions of the initial mixtures were determined in control exꢀ
periments. The water vapor pressure in the starting and final
reaction mixtures was measured in the sampler with freezing
out. The level of the water vapor content in the starting mixtures
was lower than the detection limit.
The dependences are described by curves with maxima
near the stoichiometric composition. In the region of oxyꢀ
gen excess, at low molar fractions of hydrogen in the reacꢀ
tion mixture, the reaction follows the kinetic equation of
the first order with respect to hydrogen
W = kН2РН2,
(1)
where W is the reaction rate, mol (of H or H O) m s–1
–2
;
2
2
–2 –1
–1
kН2 is the rate constant, mol m
s
Pa ; РН2 is the parꢀ
tial pressure of hydrogen, Pa.
In the region of hydrogen excess, at molar fractions of
hydrogen in the reaction mixture close to unity, the reacꢀ
tion follows the kinetic equation of the first order with
respect to oxygen
W = kO2РO2
,
(2)
where kO2 is the rate constant, mol m–2 s–1 Pa–1; РO2 is
the partial oxygen pressure, Pa.
Indices "H2" and "O2" in the designations of the reacꢀ
tion rate constants in Eqs (1) and (2) imply that the reacꢀ
tion rate constants in these equations are not true parameꢀ
ters of the specific steps but they only characterize the
occurrence of the processes in the region of oxygen or
hydrogen excess, where the first order of the reaction with
respect to hydrogen or oxygen, respectively, is observed.
These are experimental values, and their physical sense
can be revealed only by studying the reaction mechanism.
The regions in which Eqs (1) and (2) are obeyed are difꢀ
ferent. The linear dependence of the reaction rate on the
partial hydrogen pressure covers a considerable region of
composition of the reaction mixture: from Х  0.02—0.03
to Х  0.40. At the same time, the linear dependence of the
reaction rate on the partial oxygen pressure is valid only in
a narrow range of molar fractions of oxygen (Y = 1 – X):
depending on the temperature between Y  0.02—0.07 and
Y  0.01—0.02. The experimental values of the kinetic
parameters of the reaction are given in Table 1.
Results and Discussion
The dependences of the reaction rate (rate of water
formation) on the composition of the reaction mixture
(
expressed through the molar fraction of hydrogen in the
reaction mixture with oxygen, Х) are shown in Fig. 1.
W•10 /mol m s–1
9
–2
1
2
3
The results of separate studying of the reaction steps by
1
0
isotope oxygen exchange, adsorption kinetic methods
4
,6
6
4
2
0
using mass spectrometry, and surface electron spectrosꢀ
1
1,12
copy
suggest the scheme for the reaction mechanism
(
Scheme 1).
Scheme 1
0
.2
0.4
0.6
0.8
X
Fig. 1. Comparison of the experimental data at 423 (1), 448 (2),
and 473 К (3) with the results of calculation (solid lines) using
kinetic Eq. (9).
In Scheme 1, Oads, OHads, and Hads are intermediate
species adsorbed on the silver surface; are the equivalent

Kinetics of hydrogen oxidation on silver
Russ.Chem.Bull., Int.Ed., Vol. 61, No. 12, December, 2012
2227
Table 1. Experimental and kinetic parameters of the reaction in an excess of oxygen (А) and hydrogen (В) and at the stoichiometric
composition (С)
T/K
A
B
C
k
E
k
E
W•109
/mol m–2 s–1
E
H2
O2
–2 –1
mol m–2 s–1 Pa–1
/kJ mol
–1
/mol m
s
Pa–1
/kJ mol
–1
/kJ mol
–1
/
–12
–12
–11
–12
4
4
4
23
48
73
2.93•10
7.22•10
1.23•10
47.9
47.9
47.9
8.95•10
84.5
84.5
84.5
1.62
3.32
6.56
46.5
46.5
46.5
–11
–10
3.05•10
1.14•10
sites of the surface capable of binding the intermediate
species, and their number is equal to that of silver atoms
on the Ag(111) plane of the silver single crystal. All steps
are considered as elementary processes. Steps (I) and (III)
are slower than steps (II) and (IV), which are fast and
occur immediately after steps (I) and (III). Assumptions
about the rates of the steps are based on the results of
separate analysis the kinetics of these steps. The dissociaꢀ
tive adsorption of oxygen on silver in the region of medium
coverages of the surface occurs with an activation energy
formation in reaction (IV). The conditions of stationarity
of the reaction can be expressed as follows:
m(dOH/dt) = 2w – 2w = 0,
(6)
I
III
where m is a value of the monolayer capacity, mol m–2
.
Inserting Eqs (3) and (4) into Eq. (6), we obtain for
the rate of steps w and w
I
III
kIРO2(1 – OH)2 = kIIIРH2OH
.
(7)
–1 13
The following designations are used in Eqs (3)—(7):
of ~60 kJ mol , whereas the activation energy of step (II)
wI and wIII are the rates of steps (I) and (III), mol m–
s ;
2
–1
–
1
is close to 20 kJ mol . At ambient temperature, under the
conditions of static experiment, step (II) is completed
within ~1 min, whereas ~180 min are required for the
k and k are the rate constants of steps (I) and (III),
I
III
mol m s–1 Pa ; OH is the coverage of the surface with
–
2
–1
4
hydroxyls.
overall step (V) including steps (III) and (IV) to occur.
The solution of square Eq. (7) gives the expression for
the stationary coverage of the surface with hydroxyls and
the kinetic equation of the reaction
Overall step (V) follows the kinetic equation of the first
4
,6
order with respect to hydroxyls, i.e., it follows the kinetic
law of step (III) in Scheme 1. Therefore, step (III) is the
rateꢀdetermining step in overall step (V) and, hence, ocꢀ
curs more slowly than fast step (IV).
(8)
(9)
It can be accepted that the mechanism of the fast steps
is insignificant for the elucidation of kinetic regularities of
the reaction, and their parameters are not included in the
kinetic equation of the reaction.1 The surface concentraꢀ
tions of intermediate particles Oads and Hads, formed in
slow steps (I) and (III), are very low, since these particles
immediately undergo further transformations and can be
neglected. Thus, only slow steps (I) and (III) can be conꢀ
sidered in the further kinetic analysis.
The experimental kinetic data in Fig. 1 are compared
with the curves calculated by the kinetic reaction Eq. (9).
It is seen that the kinetic equation satisfactorily describes
the experimental data. The divergences between the exꢀ
perimental dependences and results of calculations by Eq.
4
(
9) do not exceed 5—10% on the average, which correꢀ
sponds to the usual inaccuracy of experimental studies.
Perhaps, these divergences are caused by an inaccuracy of
determination of the reaction rate constants. The comꢀ
parison performed suggests the validity of the proposed
mechanism and the kinetic model for the studied reaction.
The rate constants of slow steps (I) and (III) in reꢀ
action Scheme 1, which were used in the calculations of
the reaction rate by kinetic Eq. (9), are listed in Table 2.
The reaction rate constants were estimated using the ineꢀ
qualities
Taking into account the above statements, the rates of
these steps are determined by the following equations:
w = k Р (1 – OH)2,
(3)
(4)
I
I O2
wIII = kIIIРH2OH
.
The reaction rate (W) is as follows:
W = 2wIII
,
(5)
kI = kO2,
(10)
(11)
since each elementary act of transformation in reacꢀ
tion (III) is accompanied by the elementary act of transꢀ
kIII = 0.5kH2,

2228
Russ.Chem.Bull., Int.Ed., Vol. 61, No. 12, December, 2012
Dokuchits et al.
Table 2. Kinetic parameters of slow steps (I) and (III) that follow from an analysis of the reaction kinetics
T/K
Step (I)
Step (III)
k
EI
/kJ mol
kIII•101
–2
2
EIII
/kJ mol
I
–1
/mol m s–1 Pa–1
–1
/
mol m–2 s–1 Pa–1
–
–
11
11
4
4
4
23
48
73
1.26•10
3.05•10
1.0•10
68.7
68.7
68.7
2.2
4.5
7.0
38.6
38.6
38.6
–
10
where kO2 and kH2 are the experimental reaction rate conꢀ
stants in an excess of hydrogen and oxygen, respectively.
The physical meaning of Eq. (11) is that each elementary
act of reaction (III) is immediately accompanied by the
elementary act of reaction (IV).
From the results of the studies it can be inferred that
the adsorbed layer on the silver surface, under the condiꢀ
tions of the stationary reaction, consists predominantly of
hydroxyls. Figure 2 shows the dependences of the coverꢀ
age of the silver surface with hydroxyls on the molar fracꢀ
tion of hydrogen. These dependences were obtained using
Eq. (8). These data are similar, although their mutual
arrangement exhibits no regularity. Probably, this is a conꢀ
sequence of an inaccuracy of determination of the rate
constants when using kinetic Eq. (9).
The experimental data are well described by a simple
scheme of the mechanism including only two steps: the
dissociative adsorption of oxygen and the interaction of
adsorbed oxygen species with gaseous hydrogen. Howevꢀ
er, the mechanism disagrees with the results of separate
studying of the reaction steps, which undoubtedly prove
that the reaction involves surface hydroxyls as intermediꢀ
ates and the activated complex of the rateꢀdetermining
step includes a hydrogen molecule and a hydroxyl (see
step (III) of Scheme 1).
The kinetic equation of the reaction (9) based on the
proposed Scheme 1 satisfactorily describes the experimenꢀ
tal data. Therefore, it can be concluded that the surface
disproportionation of the hydroxyls
4
,6
does not substantially contribute to the occurrence of the
reaction. Under the studied conditions, the stationary reꢀ
action proceeds via the hydrogenation of the surface
hydroxyls.
Based on the study of the interaction of hydrogen with
6
oxygen preꢀadsorbed on silver and taking into account
15
the available literature data on disproportionation, it was
suggested that this process can proceed via two parallel
routes: by hydrogenation (III) and by the disproportionꢀ
ation of the surface hydroxyls (VI). However, the results
of this work do not indicate that, under the studied condiꢀ
tions, the stationary reaction of hydrogen oxidation proꢀ
ceeds via the route including the disproportionation of the
surface hydroxyls.
The studies performed gave a detailed account of the
effect of the hydrogen to oxygen ratio in the reaction mixꢀ
ture on the rate of hydrogen oxidation on silver.
The data of previous works on separate analysis of the
kinetics of the interaction steps suggest that the reaction
proceed through the intermediate formation of surface
hydroxyls. Taking into account these results, we analyzed
the kinetics of the stationary reaction, which made it posꢀ
sible to derive the kinetic reaction equation that satisfacꢀ
torily describes the experimental data.
OH
1
2
3
0
0
0
.75
.50
.25
To conclude, the adsorbed layer on the silver surface
consists predominantly of hydroxyls under the conditions
of the stationary reaction.
References
0
.2
Fig. 2. Stationary coverage of the surface with hydroxyl groups
OH) vs molar fraction of hydrogen (Х) in the reaction mixture
of hydrogen and oxygen calculated using Eq. (8) at 423 (1),
48 (2), and 473 K (3).
0.4
0.6
0.8
X
1
. D. L. Chapman, W. K. Hall, Roy. Soc. Proc. A, 1929,
124, 478.
(
2. A. F. Benton, J. C. Elgin, J. Am. Chem. Soc., 1929, 51, 7.
3. S. Ya. Pshezhetskii, M. L. Vlodavets, J. Phys. Chem. USSR
(Engl. Transl.), 1950, 24 [Zh. Fiz. Khim., 1950, 24, 353].
4

Kinetics of hydrogen oxidation on silver
Russ.Chem.Bull., Int.Ed., Vol. 61, No. 12, December, 2012
2229
4
5
6
7
8
9
. A. V. Khasin, G. K. Boreskov, Kinet. Katal., 1969, 10, 613
Kinet. Catal. (Engl. Transl.), 1969, 10].
. V. Sh. Gruver, A. V. Khasin, G. K. Boreskov, Kinet. Catal.
Engl. Transl.), 1971, 12 [Kinet. Katal., 1971, 12, 156].
. E. V. Dokuchits, A. V. Khasin, A. A. Khassin, Reac. Kinet.
Mech. Catal., 2011, 103, 261.
10. G. K. Boreskov, A. V. Khasin, T. S. Starostina, Dokl. Akad.
Nauk SSSR, 1965, 164, 606 [Dokl. Chem. USSR
(Engl. Transl.), 1965].
11. C. T. Campbell, M. T. Paffett, Surf. Sci., 1984, 143, 517.
12. C. T. Campbell, Surf. Sci., 1985, 157, 43.
13. A. V. Khasin, Kinet. Katal., 1993, 34, 42 [Kinet. Catal.
(Engl. Transl.), 1993, 34].
14. M. I. Temkin, Adv. Catal., 1979, 28, 173.
15. S. W. Jorgensen, A. G. Sault, R. G. Madix, Langmuir, 1985,
1, 526.
[
(
. G. K. Boreskov, I. V. Lokhova, A. V. Khasin, React. Kinet.
Catal. Lett., 1981, 16, 167.
. Y. L. Sandler, W. M. Hickem, Proc. 3rd Intern. Congr. Catalꢀ
ysis, Amsterdam, NorthꢀHolland, Amsterdam, 1965, 1, 227.
. G. K. Boreskov, Heterogeneous Catalysis, Transl. from Rusꢀ
sian under Editorship of K. I. Zamaraev, A. V. Khasin, Nova
Science Publishers, Inc., New York, 2003.
Received June 15, 2012;
in revised form November 19, 2012