Catalytic processes on membrane palladium alloys: I. Carbon monoxide disproportionation

Article abstract of DOI:10.1134/S002315841002014X

The results of a kinetic study of the disproportionation reaction of carbon monoxide on the surface of B1-B3 palladium alloy membranes over wide ranges of residence times and temperatures are reported. The specific rates and activation energies of the reaction with the formation of CO₂ and atomic carbon adsorbed on the surface were determined. It was found that the rate of reaction on all of the alloys from the B series other than B3-2 was independent of the consumption of CO and the residence time. The catalytic activity of multicomponent palladium alloy membranes in the CO disproportionation reaction was much lower and the activation energy was higher than the corresponding parameters of Pd-(6-10)Ru and Pd-5.5Ni binary alloy membranes. Atomic carbon formed in the reaction on all of the alloys from the B series other than the B3-2 alloy did not form carbon deposits as polymer films, pyrolytic carbon, or soot strongly bound to the surface of each other to result in a decrease in the rate of reaction. On the surface of the B3-2 alloy, atomic carbon initiated the formation of a polymer film weakly bound to the surface, which decreased the rate of reaction. A reaction mechanism was proposed to adequately describe the experimental results.

Full text of DOI:10.1134/S002315841002014X

ISSN 0023ꢀ1584, Kinetics and Catalysis, 2010, Vol. 51, No. 2, pp. 255–265. © Pleiades Publishing, Ltd., 2010.
Original Russian Text © D.I. Slovetskii, E.M. Chistov, 2010, published in Kinetika i Kataliz, 2010, Vol. 51, No. 2, pp. 270–280.
Catalytic Processes on Membrane Palladium Alloys:
I. Carbon Monoxide Disproportionation
D. I. Slovetskii and E. M. Chistov
Topchiev Institute of Petrochemical Synthesis, Russian Academy of Sciences, Moscow, 119991 Russia
eꢀmail: slovetsk@ips.ac.ru
Received March 4, 2008; in final form, August 6, 2009
Abstract—The results of a kinetic study of the disproportionation reaction of carbon monoxide on the surꢀ
face of B1–B3 palladium alloy membranes over wide ranges of residence times and temperatures are
reported. The specific rates and activation energies of the reaction with the formation of CO2 and atomic carꢀ
bon adsorbed on the surface were determined. It was found that the rate of reaction on all of the alloys from
the B series other than B3ꢀ2 was independent of the consumption of CO and the residence time. The catalytic
activity of multicomponent palladium alloy membranes in the CO disproportionation reaction was much
lower and the activation energy was higher than the corresponding parameters of Pd–(6–10)Ru and Pd–
5.5Ni binary alloy membranes. Atomic carbon formed in the reaction on all of the alloys from the B series
other than the B3ꢀ2 alloy did not form carbon deposits as polymer films, pyrolytic carbon, or soot strongly
bound to the surface of each other to result in a decrease in the rate of reaction. On the surface of the B3ꢀ2
alloy, atomic carbon initiated the formation of a polymer film weakly bound to the surface, which decreased
the rate of reaction. A reaction mechanism was proposed to adequately describe the experimental results.
DOI: 10.1134/S002315841002014X
INTRODUCTION
The disproportionation reaction was studied on
binary alloy membranes in a flow of CO or a mixture
of CO with hydrogen [7–9]. In the mixture, the
hydrogenation of CO molecules to form С1–С2
hydrocarbon and water molecules occurred along with
a decrease in the rate of reaction (I). Water molecules
can compete with hydrogen molecules in adsorption at
the same active surface sites [11] to decrease the rate of
formation of СО2 in reaction (I) and enter into the
water gas reaction with CO molecules
The recovery of hydrogen from various gas mixꢀ
tures, including synthesis gas prepared by hydrocarꢀ
bon conversion, using palladium alloy membranes is
an efficient currently available technique for the proꢀ
duction of highꢀpurity hydrogen [1–4]. At the same
time, palladium alloy membranes can serve as cataꢀ
lysts for the hydrogenation of carbon oxides and the
dehydrogenation of hydrocarbons, alcohols, and
ethers [5–9]. Thus, in the extraction of pure hydrogen
from the synthesis gas of hydrocarbon conversion with
membranes of promising multicomponent palladium
alloys, soot formation was observed on the surfaces of
membranes, membrane module units, hydrogen flow
lines, and lines for the removal of depleted synthesis
gas [10]. Moreover, it was experimentally found that,
upon passing pure carbon monoxide through a quartz
reactor with membrane samples of B1 and B2 alloys
and a number of construction materials (St3 and
Kh18N10T steels, PSRꢀ72 and PMFꢀ8 solder alloys,
nickel, copper, and Pyrex), the disproportionation of
CO (Boudoir reaction) occurred at 873 K [10]:
Н2О + СО
СО2 + Н2
(II)
to increase the yield of СО2
.
Thus, to interpret the results of a study of the
simultaneous occurrence of all of the above reactions
in mixtures of CO with hydrogen, including mixtures
prepared by hydrocarbon conversion [1, 4], the disꢀ
proportionation reaction should be initially studied on
palladium alloy membranes in pure CO; then, the
water gas reaction and the hydrogenation of CO
should be studied.
This work is the first stage of a study of catalytic
reactions on multicomponent B series (B1–B3) pallaꢀ
dium alloy membranes, which are widely used in the
extraction of pure hydrogen from hydrocarbon conꢀ
version gas mixtures [1–4]. The aim of this work was to
experimentally study the rate of the disproportionꢀ
S
i
2CO
CO2 + C,
(I)
which was accompanied by soot formation, especially
intensive on St3 steel, nickel, and PSrꢀ72 solder alloy.
The strong dependence of the yield of СО2 on the
material of samples at the same parameters (gas presꢀ ation reaction in a flow of pure CO over the working
sure and temperature) suggests the heterogeneous temperature range (573–925 K) in order to predict the
character of reaction (I).
maximum possible rate of reaction and to minimize
255

256
SLOVETSKII, CHISTOV
cessing of chromatograms on a personal computer.
13
Highꢀpurity argon or helium was used as the carrier
gas in the chromatograph at a flow rate of 1 l/h. Zeolite
NaX or Porapak served as sorbents (column length,
2 m; column temperature, 60°С). The chromatograph
was calibrated against gases and calibration mixtures at
room temperature and atmospheric pressure. To pump
CO and products through the reactor, chromatograph,
and the measuring system, a small excess pressure was
10
7
9
8
12
11
sufficient (p = 0.105 MPa). The gas from the flow
meter was supplied to a burner mounted in an exhaust
hood and combusted.
1
2
5
Errors in concentration measurements were no
higher than 3% (СО) or 5% (СО2) and in gas flow
6
4
3
rate measurements, 5%
.
The reactor was filled in turn with the following
palladium alloy membranes (wt %):
Fig. 1. Experimental setup: (
) temperature measurement and control unit; (
preparation, purification, and drying unit; ( ) regulating
valve; ( ) chromatograph; ( ) flow meter; ( 12) shutoff
valves; and (13) pressureꢀandꢀvacuum gage for measuring
gas pressure in the reactor.
1
) reactor; (
2
) electric heater;
(3
4) CO
5
B1 (Pd–15.6Ag–3.06Au–0.6Pt–0.6Ru–0.24Al);
B2 (Pd–15Ag–1.5In–0.5Y)[10];
B3ꢀ1 (Pd–15Ag–1.5In–0.2Y–0.5 W);
B3ꢀ2 (Pd–15Ag–1.5In–0.2Y–0.5 Ta);
6
7
8–
the rate of product (primarily, soot) formation in the
membrane extraction of pure hydrogen.
B3ꢀ3 (Pd–15Ag–1.5In–0.2Y–0.5Nb);
B3ꢀ4 (Pd–15Ag–1.5In–0.2Y– 0.5Mo) [13].
The B1 alloy was used as a bundle of capillary tubes
with an outer diameter of 0.245
×
10–2 m, a wall thickꢀ
EXPERIMENTAL
ness of 120 m, and a length of 0.275 0.005 m, and
μ
Experiments were performed in a system (Fig. 1)
analogous to that described previously [10], but over
wider ranges of temperatures (573–925 K) and memꢀ
brane alloy compositions. The system included a
the other alloys were used as the bundles of wire 0.05
×
10–2 m in diameter and from 0.15 to 0.18 m in length
(Table 1, columns 1–7), which were degreased,
washed with hydrocarbon solvents, and dried. The
length of alloy elements was responsible for the length
of the active reaction zone (Lr). The bundle of capilꢀ
lary tubes of the B1 alloy more or less uniformly filled
fusedꢀsilica cylindrical flow reactor
diameter of 1.03
10–2 m and 70 10–2 m long, which
was heated with an electric heater , a temperature
control and measurement unit , a unit for the generꢀ
1 with an inner
×
×
2
3
the reactor cross section (Sr = 0.833
×
10–4 m2), and
ation of CO by the decomposition of formic acid in
accordance with a published procedure [12] in the
presence of phosphoric acid (Н3РО4) with chemical
gas purification and drying with phosphorus pentoxide
the bundles of wire of the other alloys were chaotically
arranged at the bottom to fill less than 5% of the reacꢀ
tor cross section. The surface roughness of alloy samꢀ
ples, which were made by extrusion and broaching,
was less than 1%.
at 193 K
ply line before the reactor inlet for gas flow regulation,
an LKhM 8MD chromatograph , and a gas flow
meter . The heated lines were made of copper tubes.
4, a regulation valve 5 arranged in the gas supꢀ
Before being charged, the reactor was heated to
873 K and purged with pure dry carbon monoxide at a
rate of 0.003 mol/h. In this case, the concentration of
CO2 measured at the reactor outlet at the silicaꢀwall
6
7
The reactor temperature was measured using a
Chromel–Alumel thermocouple arranged at the outer
surface of the reactor wall to exclude a contribution
from reaction on the thermocouple surface. The gas
flow rate was measured by leakage into calibrated 1ꢀ or
5ꢀl volumes, which were preevacuated with a fore vacꢀ
uum pump, with the automatic detection of the time
of pressure increase from initial (~150 Torr) to final
values (170–200 Torr) using a Sapfir gage. Gas flow
surface area of 0.0227 m2 was no higher than 0.01%;
this demonstrated the low catalytic activity of fused
silica (0.67
reactor with a test material, the required flow rate of
CO was adjusted and measured at the pressure
0.105 MPa and = 300 K. Then, the reactor was
×
10–5 mol m–2 h–1). After loading the
p
=
T
heated to 873 K, and the presence of impurities in a
flow of CO after the outlet of the reactor filled with
alloy samples was checked at T = 873 K and minimum
flow rates (0.003–0.006 mol/h). The products were
analyzed by chromatography with the use of highꢀ
purity argon as the carrier gas and a Porapak column
for the determination of hydrogen, CO, and CO2.
Only the peaks of CO and CO2 were detected on Poraꢀ
meter
immediately at the outlet of valve
the reactor outlet; for this purpose, shutoff valves
7
allowed us to measure the flow rates of CO
or a gas mixture at
12
5
8–
were mounted in gas inlet and outlet lines. The gas
mixture composition at the reactor outlet was deterꢀ
mined using chromatograph
6 equipped with the
ECOCHROM software for the measurement and proꢀ
KINETICS AND CATALYSIS Vol. 51
No. 2
2010

CATALYTIC PROCESSES ON MEMBRANE PALLADIUM ALLOYS: I
257
Table 1. Reactor parameters for studying the kinetics of reactions on various membrane alloys
Salloy
ꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀꢀ × 10–2, m–1
Vr – Valloy
Alloy
Lr
×
102, m
Vr
×
106, m3 Valloy
×
106, m3 Salloy
×
104, m2 Vr
–
Valloy
106, m3
×
В1
В2
27.5
22.9
12.2
13.2
15.0
12.7
13.5
3
1.20
0.50
0.80
0.70
0.80
0.73
5
403
39.7
21.7
11.7
12.4
14.3
11.9
12.8
3.5
18.6
3.4
5.0
3.9
5.2
4.9
15
14.6
15.9
18.0
15.3
16.2
3
В3ꢀ1
В3ꢀ2
В3ꢀ3
В3ꢀ4
61.5
56.0
62.3
58.3
10
(Δх/х), %
pak. Because the peak of nitrogen overlapped with the ples (Valloy) (Table 1, columns 3, 4, and 6) using the
peak of CO because of the same retention times, the equation
measurements were repeated on a column with zeoꢀ
τ
= (Vr
–
V
alloy)/{ QCO
(
p0
/
p
)(
Т
/
Т
0)},
(2)
lite, on which the peaks of hydrogen, nitrogen, oxyꢀ
gen, and methane were not detected. The upper estiꢀ
mates of component concentrations with the use of
experimental detection sensitivity values were
where p0 = 0.1 MPa, Т0 = 273 K, and
p
and
T
are the
gas pressure and temperature in the reactor, respecꢀ
tively. Error in the determination of was 8.5%. At a
constant length (Lr) of the active reactor zone (Vr
Valloy = const), the residence time was inversely proꢀ
portional to gas flow rate.
Upon measuring the dependence of the concentraꢀ
tion of CO2 on the flow rate of CO at
reactor temperature was varied at a constant flow rate
of CO and the temperature dependence of the conꢀ
centration of CO2 in the flow at the reactor outlet and
the rate of reaction was obtained.
τ
–
Ӷ0.001% hydrogen and
Ӷ0.01% nitrogen and oxygen
at the concentrations
= 0.5–2.5 vol %. Conseꢀ
C
СО2
quently, the rate of formation of CO2 by the oxidation
of CO with water vapor and residual oxygen
T = 873 K, the
СО + О2
СО2 + О
(III)
was much lower than the rate of formation of CO2 in
reaction (I) so that reactions (II) and (III) can be
ignored.
After completion of measurements in the first samꢀ
ple, the CO drying agent was regenerated at the outlet
of reactor 4, the next sample was loaded, and a cycle of
the above measurements and calculations was
repeated in turn with the other membrane alloy samꢀ
ples.
The disproportionation kinetics was studied over a
wide ranges of parameters (T = 573–925 K and gas
flow rates of 0.0022–0.0670 mol/h). Low CO2 conꢀ
centrations were measured chromatographically using
Porapak columns and helium as the carrier gas. In the
course of these experiments, the concentrations
C
СО2
and ССО at the reactor outlet were calculated from
chromatograms. The measured concentrations of CO2
were corrected for CO2 formation at the wall surface of
the silica reactor. From the refined values of the conꢀ
RESULTS
Table 2 summarizes the CO2 concentrations meaꢀ
sured at the reactor outlet depending on the rate QCO
at the inlet and calculated using Eq. (1), as well as the
specific rates of reaction on various membrane alloys
at 873 K. The specific rate of reaction reached a minꢀ
imum on the B1 alloy and increased in the following
centration
and measured CO flow rates, the speꢀ
C
СО2
cific rate of reaction (I) on the membrane palladium
alloy was calculated as a function of temperature and
residence time (
τ
):
order: B1
В3ꢀ4
В3ꢀ3
В3ꢀ1
В2
,
T,τ Q T,τ }/S
alloy
(1)
w T,τ = {C
(
)
(
)
(
)
r
СО2
CO
B3ꢀ2. As the temperature was decreased, it monotonꢀ
ically decreased on all of the alloys (Table 3).
where
is the concentration of CO2 molecules,
C
СО2
Form the measured temperature dependence of the
specific rate in the Arrhenius coordinates, the activaꢀ
tion energies of the disproportionation reaction were
calculated. The experimental points lied in straight
lines (Figs. 2a, 2b) within the limits of relative error for
all of the multicomponent membrane alloys, which
mole fractions (henceforth, Salloy is the geometric surꢀ
face area of an alloy sample (Table 1, column 5)).
Errors in the calculated absolute values of
20 and 10% in the relative values (at Salloy = const).
The tentative gas residence time ( ) with the surꢀ
Δ
wr/wr were
τ
face of alloy samples at constant loading, temperature,
and pressure was calculated from the measured rate
was
Δ
logwr
) shows the same dependences for Pd–10Ru
the volumes of the reaction zone (Vr) and alloy samꢀ and Pd–5.5Ni binary membrane alloys calculated
/ logwr < 7%. For comparison, Fig. 2b
(
QCO) at the reactor outlet and the difference between (lines 7, 8
KINETICS AND CATALYSIS Vol. 51
No. 2
2010

258
SLOVETSKII, CHISTOV
Table 2. Dependence of the concentration of CO2 and the rate of CO disproportionation on membrane alloys on the gas flow
rate at the reactor inlet at T = 873 K and p = 0.105 MPa
B1
B2
B3ꢀ1
CCO
wr
×
103,
QCO
mol/h
,
wr
×
103,
QCO
mol/h
,
wr
103,
×
CCO
,
CCO
,
,
2
2
2
QCO, mol/h
mol m–2 h–1
mol m–2 h–1
mol m–2 h–1
vol %
vol %
vol %
0.0500
0.0300
0.0280
0.0200
0.0140
0.0086
0.0062
0.08
0.18
0.19
0.33
0.53
0.78
0.96
1.00
1.34
1.32
1.64
1.81
1.69
1.49
0.0360
0.0170
0.0120
0.0080
0.0057
0.0047
0.0044
0.0039
0.0034
0.17
0.37
0.51
0.77
1.13
1.35
1.49
1.65
1.77
15.4
15.6
0.066
0.044
0.026
0.015
0.014
0.010
0.005
0.003
0.16
0.18
0.27
0.45
0.56
0.86
1.72
2.50
17.1
12.8
11.4
11.1
12.7
14.0
14.0
12.2
15.4
15.6
16.1
15.9
16.6
16.1
15.1
wr × 103 mol m–2 h–1
molecule cm–2 s–1
1.55 0.14
1013
15.80 0.39
1014
12.60 0.90
1014
2.6
×
2.6
×
2.1
×
σ
2 = 0.04
σ
2 = 0.23
σ
2 = 1.3
B3ꢀ2
CCO
B3ꢀ3
CCO
B3ꢀ4
CCO
103,
QCO
mol/h
,
wr
×
103,
QCO
mol/h
,
wr
103,
×
,
,
,
wr
×
2
2
2
QCO, mol/h
mol m–2 h–1
mol m–2 h–1
mol m–2 h–1
vol %
vol %
vol %
0.0340
0.0190
0.0140
0.0085
0.0048
0.0038
0.50
0.94
1.13
1.44
2.14
2.50
30.4
32.0
28.2
21.8
18.4
17.0
0.0630
0.0130
0.0110
0.0093
0.0079
0.0066
0.0041
0.0036
0.0032
0.04
0.20
0.23
0.28
0.31
0.41
0.56
0.66
0.73
4.01
4.17
0.0140
0.0110
0.0099
0.0074
0.0056
0.0039
0.0037
0.0031
0.14
0.16
0.20
0.23
0.31
0.39
0.45
0.54
3.43
3.09
3.43
2.92
2.92
2.57
2.92
2.92
4.01
4.65
3.85
4.33
3.69
3.85
3.85
wr × 103 mol m–2 h–1
molecule cm–2 s–1
30.2 2.4
4.05 0.22
1013
3.03 0.22
1013
5.1
×
1014
2 = 3.6
= 3)
6.8
×
5.1
×
σ
σ
2 = 0.05
σ
2 = 0.8
(n
14.8 1.0
2 = 157
= 6)
σ
(n
based on published data [7–9, 14]. In these cases, relꢀ
The calculated Reynolds numbers (characterizing
ative errors were greater: 12% for Pd–5.5Ni and turbulence) for all of the alloy samples over the tested
20% for Pd–(6–10)Ru. The activation energies of range of parameters varied from 0.2 to 4.2, and addiꢀ
the reaction of CO disproportionation varied from 15 tional turbulization did not occur in the longitudinal
to 95 kJ/mol for multicomponent alloys from the flow around long smooth samples. Thus, the gas flow
series of B1–B3 and from 9 (Pd–5.5Ni) to 18 kJ/mol in the reactor was laminar. The experimental data
(Pd–(6–10)Ru) for twoꢀcomponent alloys (Table 4).
(Tables 2, 3) in combination with reactor and sample
KINETICS AND CATALYSIS Vol. 51
No. 2
2010

CATALYTIC PROCESSES ON MEMBRANE PALLADIUM ALLOYS: I
259
Table 3. Dependence of the concentration and the rate of formation of CO2 on the surface temperature of membrane alloys at
p = 0.105 MPa and constant CO flow rates
В1;
QCO = 0.0280 mol/h
В2;
QCO = 0.0077 mol/h
В3ꢀ1; QCO = 0.0260 mol/h
wr
×
103,
wr
×
103,
wr
103,
×
CCO
,
vol %
CCO
,
vol %
CCO , vol %
2
T, К
T
, К
T, К
mol m–2 h–1
mol m–2 h–1
mol m–2 h–1
2
2
751
767
770
790
810
841
873
0.062
0.103
0.117
0.128
0.150
0.165
0.193
0.43
0.72
0.81
0.89
1.04
1.15
1.34
743
763
794
829
851
870
873
892
921
0.26
0.29
0.38
0.56
0.77
0.80
0.80
0.87
1.00
5.04
5.62
783
802
825
853
863
873
883
0.055
0.06
0.13
0.21
0.21
0.24
0.29
2.32
2.50
7.34
5.50
10.90
14.90
15.50
15.50
16.90
19.40
8.88
8.88
10.15
12.26
В3ꢀ2;
QCO = 0.020 mol/h
В3ꢀ3;
QCO = 0.048 mol/h
В3ꢀ4; QCO = 0.011 mol/h
wr
×
103,
wr
×
103,
wr
103,
×
CCO
,
vol %
CCO
,
vol %
CCO , vol %
2
T, К
T
, К
T, К
mol m–2 h–1
mol m–2 h–1
mol m–2 h–1
2
2
573
623
646
741
803
853
873
0.35
0.46
0.53
0.72
0.83
0.92
1.07
12.50
16.43
18.92
25.71
29.64
32.86
38.21
754
824
847
873
883
0.075
0.272
0.375
0.518
0.572
0.58
2.03
2.89
3.99
4.41
793
836
873
924
0.09
0.12
0.16
0.25
1.52
2.19
3.24
4.76
Pd–10 Ru [8, 9]
Pd–5.5 Ni [8, 9]
Pd–(6, 10) Ru [7, 16]
Pd–5.5 Ni [7, 16]
wr
×
103,
wr
×
103,
wr
×
103,
wr
103,
×
T, К
T, К
T, К
T, К
mol m–2 h–1
mol m–2 h–1
mol m–2 h–1
mol m–2 h–1
523
583
633
683
98.23
125.33
263.89
362.12
523
583
633
683
30.67
37.14
45.20
53.08
523
573
623
673
98
120
260
360
523
573
623
673
30
35
44
50
Table 4. Activation energy of CO disproportionation reaction (I) on the surface of multicomponent and binary membrane alloys
Alloy addition
Alloy
Ea
/
R
×
10–3, K
Ea, kJ/mol
transition metals, wt % Group III metals, wt %
В1
5.9 0.2
5.6 0.4
49.2 2.0
46.7 3.3
95.0 2.5
15.0 0.3
87.4 2.5
46.2 1.7
17.90
0.6 Pt–0.6 Ru
–
0.24Al
В2
1.5In–0.5Y
1.5In–0.2Y
1.5In–0.2Y
1.5In–0.2Y
1.5In–0.2Y
–
В3ꢀ1
11.4 0.3
1.8 0.04
10.5 0.3
5.5 0.2
2.14
0.5 W
В3ꢀ2 (τ ≤ 40 s)
В3ꢀ3
0.5 Ta
0.5 Nb
0.5 Mo
6–10 Ru
5.5 Ni
В3ꢀ4
Pd–(6–10) Ru
Pd–5.5 Ni
1.07
8.95
–
KINETICS AND CATALYSIS Vol. 51
No. 2
2010

260
SLOVETSKII, CHISTOV
1
103) [mol m−2 h− ]
CCO
log(wp
1.0
×
2
2.5
1
(a)
2.0
2
1.5
1.0
0.5
0
3
1.0
0.5
0
5
4
5
6
50
100
150
200
, s
4
τ
3
Fig. 3. Dependence of the concentration
idence time with the surfaces of the following alloys: (
B3ꢀ2, ( ) B3ꢀ1, ( ) B2, ( ) B1, ( ) B3ꢀ3, and ( ) B3ꢀ4.
on the resꢀ
C
СО2
1)
2
3
4
5
6
2
1
reaction rate coefficients to the coefficients of the
mass transfer of CO molecules to the surface of alloys
over the entire test temperature range:
{wrSalloy/(Vr −Valloy)CСО}/{DdifSe/(Vr −Valloy)δ}
−0.5
1.1
1.2
1.3
(3)
≤ 0.5 −11 ×10−6,
(
)
2.5
(b)
where Ddif
is the diffusion coefficient of CO molecules in a gas
flow [17]; is the thickness of a boundary layer on the
samples. In the case of B1 alloy, to obtain an upper estiꢀ
mate from relationship (3), was taken to be approxiꢀ
0.25
(
T
= 573–925 K) = (0.70–1.54)
×
10–4 m2/s
δ
8
δ
2.0
1.5
1.0
mately equal to the diameter of samples (
δ
≤
×
10⎯2 m) at the uniform reactor filling of about 50%.
For the other alloys (B2, B3) whose filling was no
higher than 5%, the overall mass transfer to the surface
of all of the samples was estimated. In this case, the
7
boundary layer thickness was δ ≤ (0.2–0.3)
×
10–2 m.
From estimates obtained using relationship (3), it folꢀ
lows that the diffusion massꢀtransfer coefficient was
higher than the reaction rate coefficient of disproporꢀ
tionation by five or six orders of magnitude for all of
the samples over the entire test range of parameters.
Consequently, the gasꢀphase diffusion of CO moleꢀ
cules to the surface of alloys did not limit the rate of
disproportionation, which occurred in the kinetic
region for all of the test membrane alloys over the
entire test range of parameters. Thus, the rate of disꢀ
proportionation is independent of the gas flow rate in
the reactor. Indeed, differences between the rates of
reaction and the average rates of reaction given in
Table 2 do not exceed the relative errors of measureꢀ
ments for all of the alloys other than B3ꢀ2. Therefore,
6
1.2
1.4
1.6
1.8
2.0
1
103/ , K−
T
Fig. 2. The temperature dependence of the specific reacꢀ
tion rate of CO disproportionation on alloys in the Arrheꢀ
nius coordinates (lines); points refer to experimental data
for the following alloy samples: (
) B3ꢀ1, ( ) B2, ( ) B3ꢀ2, (
Pd 10Ru.
1
) B3ꢀ3, (
2) B1, (3) B3ꢀ4,
) Pd–5.5Ni, and (8)
(4
5
6
7
⎯
parameters (Table 1) allowed us to evaluate the disproꢀ
portionation reaction conditions on multicomponent
alloys with consideration for the effect of diffusion
mass transfer. For this purpose, in accordance with a
the deviations of particular points from average values
2
greater than the doubled dispersion
σ
can be due to
published method [15, 16], we calculated the ratios of only random errors, and they were excluded in the calꢀ
KINETICS AND CATALYSIS Vol. 51
No. 2
2010

CATALYTIC PROCESSES ON MEMBRANE PALLADIUM ALLOYS: I
261
Table 5. Rate of formation (wC) and the specific (QCtexp) and total (QCtexp alloy) yields of atomic carbon in the disproportionꢀ
S
ation of CO on the surface of membrane alloys in a maximum exposure time (texp = 12 h) in experiments performed at T = 873 K
and p = 0.105 MPa
Alloy wC
В1
×
10–14
,
atom cm–2 s–1 QC exp
t
×
10–18
,
atom/cm2
QCtexp
,
mg/cm2
QCtexpSalloy, mg
0.26
2.60
2.10
0.68
0.51
5.10
1.10
11.20
9.10
0.023
0.230
0.182
0.059
0.044
0.450
9.10
7.38
В2
В3ꢀ1
В3ꢀ3
В3ꢀ4
В3ꢀ2
11.20
3.68
2.94
2.20
2.57
22.00
26.00
culation of the average values of reaction rates on B1
and B3ꢀ1 alloys. The rates of reaction on B1, B2, B3ꢀ
1, B3ꢀ3, and B3ꢀ4 alloys did not depend on the flow
rate of CO, and the concentrations of CO2 molecules
at the reactor outlet monotonically increased as the
residence time was increased (Fig. 3).
In disproportionation reaction (I), the yield of carꢀ
bon atoms is equal to the yield of CO2 molecules. This
allowed us to calculate the specific and total yields of
carbon, which resulted from reaction (I) on the surꢀ
faces of various alloys at
mum time of measurements ( exp
from data in Table 2.
T
= 873 K during the maxiꢀ
t
= 12 h) (Table 5),
A monotonic decrease in the rate of reaction
beyond the limits of relative errors ( 10%) was
detected in B3ꢀ2 alloy as the flow rate of CO was
decreased (Table 2). In this case, deviations from the
From these data, it follows that the amount of carꢀ
bon atoms formed on the alloy surfaces in the experiꢀ
ment times as a result of the reaction was sufficient for
the formation of deposits from 335 to 680 nm in thickꢀ
ness at distances between carbon layers characteristic
of graphite (0.335 nm) and soot particles. They can
completely block the surface and decrease the rate of
reaction to its termination. However, even a monoꢀ
tonic decrease in the reaction rate with exposure time
was not detected on the alloys other than B3ꢀ2 alloy.
At the same time, a continuous polymer film 350–400
nm in thickness formed on the surface of B3ꢀ2 alloy
incompletely blocked the surface. The molecules of
CO continued to arrive at the alloy surface from a gas
phase and enter into reaction (I); this suggests the
occurrence of the diffusion of CO molecules and the
release of the resulting gasꢀphase products through the
polymer film.
arithmetic mean value of the rate of reaction at all
2
points (
n
= 6) were greater than 70%, and
σ
was
higher than the average value by a factor of 5. At the
same time, on this alloy, which is characterized by a
maximum rate of reaction at residence times τ ≤ 40 s,
a break of a curve was detected in the dependence of
the concentration of CO2 on residence time (Fig. 3).
As shown above, this break cannot be explained by
reaction rate limitations due to diffusion mass transfer.
Moreover, after the experiments performed at the conꢀ
stant temperature
40 s, and the exposure
T
= 873 K, the residence times
τ
>
texp ~12 h, a continuous transꢀ
parent violet film was detected on the surface of B3ꢀ2
alloy; this film was easily removed upon sample bendꢀ
ing. The violet color was due to the interference of
incident and reflected light; it suggests that the film
thickness was about the wavelength of this light
(~350–400 nm). Undoubtedly, the formation of this
film was responsible for an almost twofold decrease in
the rate of reaction as the film thickness increased over
Therefore, the average reaction rate of CO disproꢀ
portionation on B3ꢀ2 alloy (Table 2) was calculated
from the points that corresponded to the rates QCO
≥
the residence time interval
blocking the alloy surface.
τ
= 60–200 s (Fig. 3) by
0.0140 mol/h and residence times τ≤40 s, at which the
formation of a polymer film was not detected. The
activation energy of reaction on this alloy was also calꢀ
The formation of polymer films on the surfaces of
other alloys was not observed. After the experiments at
culated at
face.
τ
= 40 s in the absence of a film on the surꢀ
T
= 873 K and exposure times to 12 h in a flow of CO,
only a barely perceptible dark carbon deposit was
detected on the surfaces of B2 and B3ꢀ1 alloys; this
deposit was easily removed from the surface with a
flow of air. Therefore, carbon did not exhibit strong
adhesion to the alloy surface and to individual carbon
particles. The absence of chemical bonds clearly indiꢀ
cated that the detected deposit was different from
pyrolytic carbon. It can be a deposit of soot or graphite
particles.
An analysis of these data, especially, small amounts
of carbon deposits and weak bonds to the surfaces of
B2 and B3ꢀ1 alloys, where the rates of reaction were
much higher than on the other alloys (B1, B3ꢀ3, and
B4), led us to a conclusion that the surface was continꢀ
uously cleaned from carbon removed with a gas flow in
the course of reaction on all of the membrane alloys
other than B3ꢀ2 alloy.
KINETICS AND CATALYSIS Vol. 51
No. 2
2010

262
SLOVETSKII, CHISTOV
DISCUSSION
great molecular dissociation energy (1066 kJ/mol)
[20], which is higher than the atomic bond energies in
MO (486–1063 kJ/mol) and MC crystals (611–
1237 kJ/mol) in the majority of the test membrane
alloys [21]. Therefore, it is most likely that the dissociꢀ
ation of CO molecules results from irreversible
adsorption at two or more closely spaced alloy metal
atoms, which comprise an active site s2 with the forꢀ
Let us consider a possible mechanism of heterogeꢀ
neous reaction (I) based on the above experimental
data (Tables 2–5, Fig. 3) and the results of a study of
the structure and bond energies of CO molecules
adsorbed on the surface of Group VIII metals [18] and
Pd–Ru membrane alloys [7–9, 14] using various techꢀ
niques.
mation of adsorbed
and
atoms on the surface.
''
Os2
'
Сs2
The experimental data demonstrate the occurrence
of heterogeneous reaction (I) on all of the membrane
alloys. In this case, the change in the average concenꢀ
tration of CO molecules in a gas phase along the reacꢀ
tor was lower than 2–5% in the course of reaction on
all of the alloys at a constant temperature. Conseꢀ
quently, all of the active surface sites on all of the alloys
were permanently occupied by adsorbed particles over
the entire test ranges of residence times and temperaꢀ
tures. From the experimental data (Tables 2, 3), it folꢀ
lows that the rate of reaction and hence the composiꢀ
tion and structure of sites active in reaction (I) and the
concentrations of particles adsorbed on them depend
on temperature and alloy composition and do not
depend on residence time.
The multicenter dissociative and oneꢀcenter reversꢀ
ible forms of the adsorption of CO and hydrogen
molecules were detected on the surfaces of all of the
Group VIII metals [18] and PdRu and PdNi binary
alloys [7, 14].
Similar lowꢀtemperature and highꢀtemperature
adsorption sites will occur on the surfaces of other
binary and multicomponent membrane palladium
alloys, in particular, alloys from the series B1–B3,
because their constituents are addition elements of
Group I (Ag and Au), Group III (Al, In, and Y),
Group V (Ta and Nb), Group VI (Mo and W), and
Group VIII (Ru and Pt) metals, which possess high
catalytic properties, in addition to palladium
(Table 4).
It is well known that the adsorption of CO moleꢀ
cules on all of the Group VIII metals [18] and the test
palladium membrane alloys [7–9] occurrs at active
sites of two types to form weak and strong bonds,
respectively. The weak bond occurred in the reversible
lowꢀtemperature adsorption at one active site, and it
was characterized by the linear structure of an adsorpꢀ
tion complex, whereas the strong one occurred in the
irreversible dissociative adsorption with the formation
of a bridging structure due to the bond of CO moleꢀ
cules with two or more active sites of different types
[7–9, 14].
Based on the above analysis of the experimental
results and published data, we suggest that the probaꢀ
ble mechanism of CO2 formation in a flow of CO on
the test multicomponent membrane palladium alloys
from the series B1–B3, as well as on membrane PdRu
and PdNi binary alloys, involves the following steps:
СО + s1
(COs2)*
(CO )* + s ,
COs1,
(IV)
CO + s2
COs +
+
,
''
Сs2 Os2
(V)
'
(VI)
''
Os2
''
s2
1
2
1
CO +
(CO )* ,
''
s2
(VII)
''
Os2
2
Thus, two forms of adsorption with maximums at
690 10 and 960 30 K were detected on the surface
of a binary PdRu membrane alloy by thermal desorpꢀ
tion from the surface presaturated with CO molecules
(CO )*
2
CO2
↑
+ ,
''
s2
(VIII)
''
s2
where s1 are the lowꢀtemperature sites of reversible
adsorption and s2 are the highꢀtemperature sites of disꢀ
at
presaturated with CO molecules at lower temperatures
523 K, only one peak at 690 20 K was detected;
T = 773–973 K. On desorption from the surface
sociative (irreversible) adsorption of CO, and
,
and
''
s2
are the portions of multicomponent s2 sites. It is difꢀ
'
s2
Т
≤
ferent from the mechanism of thermal adsorption folꢀ
lowed by vacuum desorption [7] only in the presence
of possible step (VII) [14].
this peak shifted to 720 K and broadened to 1000 K as
the adsorption temperature was increased to 573 K [7].
Both of the peaks were detected upon cooling from
773 to 295 K in a continuous flow of CO [14, 19]. In
the presence of a single peak, only CO molecules were
detected in thermal desorption products, whereas CO
and CO2 molecules were detected upon the appearꢀ
ance of a second peak in the thermal desorption specꢀ
tra [7]. These data allowed us to relate first type sites s1
(peak at 690 K) to the reversible adsorption–desorpꢀ
tion of CO molecules and second type sites s2 to the
dissociative adsorption of CO (peak at 960 K) [7, 14]
because the presence of a peak at 960 K is a necessary
condition for the formation of CO2.
At low temperatures (Т < 573 K), only the reversꢀ
ible adsorption of CO occurred to result, in combinaꢀ
tion with the reverse reaction, desorption (IV), in an
equilibrium at this step. As the temperature was
increased (
Т
≥
573 K), the irreversible adsorption of
CO molecules at active sites of the second type was
detected; it resulted in dissociation (V) with the forꢀ
mation of equal amounts of oxygen and carbon atoms
adsorbed at different parts of these sites
The bond energies of adsorbed oxygen and carbon
atoms at Group VIII metals (Pd, Pt, Ru, and Ni) and,
and
.
''
Оs2
'
Сs2
The dissociative adsorption of CO cannot occur probably, other transition metals were 260–420 and
with the participation of only one site because of the 80–160 kJ/mol, respectively [18]. Subsequently, the
KINETICS AND CATALYSIS Vol. 51
No. 2
2010

CATALYTIC PROCESSES ON MEMBRANE PALLADIUM ALLOYS: I
263
adsorbed oxygen atoms were completely consumed in
Thus, all of the active surface sites of both types,
reactions (VI)–(VII) for the formation of CO2 moleꢀ which participated in the disproportionation process,
were liberated and occupied once again as a result of
the subsequent dissociative and reversible adsorption
of CO molecules on all of the test membrane pallaꢀ
dium alloys. The above mechanism of reactions (IV)–
(XI) explains the observed absence of changes in the
reaction rate depending on the residence time of a flow
of CO with the surfaces of all of the alloys other than
B3ꢀ2.
In the case of B3ꢀ2 alloy, whose composition difꢀ
fered from that of the other alloys only in the presence
of tantalum as an addition element, the Ta–C bond
was the strongest, as compared with the Pd–C, Y–C,
Mo–C, Nb–C, and W–C bonds, and the formation
of heavier Сn molecules was required for the rupture of
this bond. In this case, at least a portion of carbon
atoms and molecules could be more strongly bound to
the surface atoms of tantalum (М1). The chemisorpꢀ
tion of CO molecules and the subsequent reactions up
to the formation of a polymer film, which was weakly
bound to the surface, can occur at the adsorbed СM1
cules, which were desorbed from the surface into a gas
flow (VIII). As a result of reactions (VI)–(VIII), the s1
and surface sites were purified.
''
s2
Carbon atoms remained as
at the surface sites.
'
Сs2
Their subsequent behavior depended on the character
and strength of bonds with sites. From IR spectroꢀ
'
s2
scopic data, it follows that CO molecules can be
adsorbed in a strongly bound bridging form in the folꢀ
lowing probable configurations:
C=/=O=M1
C=/=O=M1
M1
M1
M1
M
M
C=/=O=M
C=/=O=M ;
(IX)
C=/=O=M1
,
А
B
where M and М1 are palladium and additional eleꢀ
ment atoms, including Group VIII metals, that conꢀ
stituted s2 active sites at various faces of the crystal latꢀ
tice, and the / sign indicates the site of bond rupture
atoms or
carbon molecules by the loosening of
С
nM1
the Ta–C bond in the course of polymerization.
with the formation of adsorbed
and
atoms.
''
Оs2
'
Cs2
The violet film deposited on the surface of B2 alloy
was similar to a film deposited on a glass reactor surꢀ
face in the course of a study of the decomposition of
CO molecules in a nonequilibrium glowꢀdischarge
plasma at reduced pressures [22]. The formation of
C(3P) and О(3Р) atoms as a consequence of the elecꢀ
tronꢀimpact dissociation of CO molecules was
detected in the bulk of the plasma. In this case, the forꢀ
mation of a film was detected on the walls of a disꢀ
charge tube. According to IR spectroscopic and Xꢀray
diffraction data, the composition of the film, which
was analogous to the film on the surface of B3ꢀ2 alloy
in these experiments in appearance and color, correꢀ
After dissociation and the interaction of adsorbed oxyꢀ
gen atoms with CO molecules followed by the desorpꢀ
tion of CO2 (VI)–(VIII), carbon atoms adsorbed at
the neighboring M and М1 metal atoms ( ) in conꢀ
'
Сs2
figuration
A
(IX) remained on the surfaces of alloys. It
is likely that, in this configuration, neighboring carbon
*
atoms interacted with the formation of excited
'
Сns2
2). The internal excitation energies of
molecules (
n
≥
the resulting C2–C8 carbon molecules (600–
4100 kJ/mol [21]) was sufficient for the rupture of
their bonds with two (160–320 kJ/mol) or more
'
s2
sites and the desorption of free carbon as a mixture of
Сn molecules from the surface to a gas flow. As a result
of the desorption of carbon molecules from the surꢀ
sponded to the [Сх Оу]n polymer, in which the x/y ratio
varied from 2 : 1 to 3 : 2 and color varied from black to
violet depending on the composition of the gas phase.
A study of the material balance of chemical reactions
occurring in the discharge demonstrated that the film
formation was initiated by the deposition of carbon
atoms onto the surface of molybdenum glass, and the
growth propagation was due to the adsorption of CO
molecules and C atoms on them from the gas
phase [18].
An analogous mechanism—the subsequent
adsorption of CO molecules on C atoms adsorbed at s2
sites on the surface of B3ꢀ2 alloy with the formation of
[СхОу]n polymers—is also possible in the disproporꢀ
tionation of CO.
In our opinion, the absence of coke or soot formaꢀ
tion from the surfaces of other alloys studied in this
work provides support for the above reaction mechaꢀ
nism with consideration for the formation of carbon
molecules, the desorption of them into the gas phase,
and the removal with a gas flow. Otherwise, polymer
face, all of the sites were liberated:
'
s2
*
+ (n
– 1)
+
Cn + 2 . (X)
'
s2
↑
'
'
'
'
Сs2
Сs2
Сns2 s2
The main portion of desorbed carbon molecules was
removed with a gas flow, whereas a portion of heavier
molecules could be deposited on the surface as a film,
which was detected on the surface of the B2 and B3ꢀ1
test multicomponent membrane alloys. This deposit
of carbon molecules or particles was readily removed
by a flow of air after the removal of the samples into the
atmosphere. The occurrence of these reactions exꢀ
plains the experimental fact that strong chemical
bonds with the surface and between deposited carbon
particles were absent from B1, B2, B3ꢀ1, B3ꢀ3, and
B3ꢀ4 alloys and a polymer film was absent from the
surface of B3ꢀ2 alloy. At the same time, reactions (X)
resulted in the reduction of s2 sites
+
s2.
(XI)
'
''
s2 s2
KINETICS AND CATALYSIS Vol. 51
No. 2
2010

264
SLOVETSKII, CHISTOV
films, coke, or soot would be formed on the surfaces of one order of magnitude, as compared with PdRu alloy,
all of the alloys to affect the rate and activation energy or by a factor of more than 2–3, as compared with
of forward reaction (I).
PdNi alloy.
(2) The reaction rate of disproportionation monoꢀ
An analysis of the experimental results and pubꢀ
lished data showed that the reaction rates and activation
energies (Еа) reflect the desorption of CO2 molecules
over the temperature range of CO disproportionation
reaction (573–925 K) on multicomponent (B1–B3)
and twoꢀcomponent membrane Pd–10Ru and
Pd⎯5.5Ni palladium alloys [7]. According to the
above disproportionation mechanism of reactions
(IV)–(XI), these results reflect the temperature distriꢀ
tonically decreased with decreasing temperature on all
of the test multicomponent and twoꢀcomponent
membrane palladium alloys. The change in the activaꢀ
tion energy of the reaction on palladium alloys with
various compositions (Еа = 9–95 kJ/mol) reflects the
shape of the temperature distribution of the concenꢀ
tration of s2 active sites, which depends on the nature
and concentration of alloy elements.
bution of the concentrations of active surface sites (s2
of the strong bridging form of the adsorption of CO
molecules for corresponding alloys [7, 18].
)
(3) The formation of atomic carbon in the reaction
of CO disproportionation on all of the test membrane
palladium alloys other than B3ꢀ2 does not result in the
formation of coke or soot deposits, which can decrease
the rate of reaction at residence times to 200 s and total
exposure times to 12 h, on their surfaces.
(4) On the surface of B3ꢀ2 alloy, atomic carbon
adsorbed as a result of the reaction initiates the formaꢀ
tion of a polymer film, which is weakly bound to the
surface but partially blocks it to cause a monotonic
decrease in the rate of reaction over the residence time
range of 40–200 s by a factor of almost 2. The reaction
rate and activation energy for this alloy were deterꢀ
At gas pressure
p
≥
0.1 MPa, the formation of CO2
resulted from consecutive reactions: dissociative CO
adsorption (VI) with the formation of Оs2 adsorbed
oxygen atoms followed by the oxidation of CO moleꢀ
cules adsorbed at other sites s1 (V) or immediately at
Оs2 atoms (VII). One way or the other, all of these steps
occurred at the active sites of alloy surfaces. At a gas
pressure of 0.1 MPa in the reactor and
flow of CO molecules to the surface of alloy samples
was
1024 molecule cm–2 s–1. Therefore, the adsorpꢀ
Т = 873 K, the
3
×
mined at the CO residence time with the surface
40 s in the absence of a film.
τ
=
tion surface sites s1 and s2 should be completely occuꢀ
pied by CO molecules with the formation of adsorbed
(5) We proposed a reaction mechanism to adeꢀ
quately explain the experimental results.
and
atoms in a time shorter than
'
Сs2
3
10–6 s [23,
×
''
Оs2
24]. From the experimental data, it follows that
''
Оs2
atoms decayed in the reactions of CO2 formation with
a characteristic time of no shorter than 10 s (Fig. 3).
Thus, the oxidation of CO by the interaction with
adsorbed oxygen atoms is the rateꢀlimiting step of the
reaction of CO2 formation. An increase in the reactor
pressure above 0.1 MPa and in the flux of CO moleꢀ
cules to the alloy surfaces cannot affect the rates of CO
oxidation steps and the formation of CO2 molecules
with their desorption into a gas phase because the
REFERENCES
1. Slovetskii, D.I., Drag. Kamni Drag. Met., 2003, no. 1,
p. 119.
2. Knapton, A.G., Platinum Met. Rev., 1977, vol. 21,
p. 44.
3. German Patent 2305595, 1974.
4. Mordkovich, V.Z., Baichtock, Yu.K., and Sosna, M.H.,
Platinum Met. Rev., 1992, vol. 36, p. 90.
purification of sites and the regeneration of s2 active
'
s2
5. Tarasova, V.E. and Pavlova, L.F., in Metally i splavy kak
membrannye katalizatory (Metals and Alloys As Memꢀ
brane Catalysts), Gryaznov, V.M. and Klabunovskii, E.I.,
Eds., Moscow: Nauka, 1981, p. 120.
sites occurred only after the formation and desorption
of CO2 and Сn molecules. Consequently, the rate of
CO disproportionation reaction (I) is independent of
gas pressure. The absence of the pressure dependence
of the rate of the reaction was also supported by pubꢀ
lished experimental data [7–9].
6. Roshan, N.R. and Polyakova, V.P., in Metallicheskie
monokristally (Metal Single Crystals), Moscow: Nauka,
1990, p. 235.
7. Gur’yanova, O.S., Serov, Yu.M., Lapidus, A.L., Dmitꢀ
riev, R.G., Gul’yanova, S.G., Gryaznov, V.M., and
Based on the above consideration, we can make the
following conclusions:
Minachev, Kh.M., Izv. Akad. Nauk SSSR, Ser. Khim.
,
(1) The specific rate of reaction (wr) of CO disproꢀ
portionation with the formation of CO2 molecules and
1987, no. 11, p. 2428.
8. Gur’yanova, O.S., Serov, Yu.M., Gul’yanova, S.G.,
and Gryaznov, V.M., Kinet. Katal., 1988, vol. 29, no. 4,
p. 850.
adsorbed atomic carbon is independent of the resꢀ
'
Сs2
idence time of CO with the surfaces of all of the test
multicomponent membrane palladium alloys other
9. Gryaznov, V.M., Gul’yanova, S.G., and Serov, Yu.M.,
than B3ꢀ2. At
order B3ꢀ
T = 833–925 K, wr decreases in the
В2 В3ꢀ1 В3ꢀ3 В3ꢀ4
Usp. Khim., 1989, vol. 58, p. 58.
2
10. Koshel’, V.I., Dobradin, A.A., and Chistov, E.M.,
Vestn. Akad. Nauk BSSR, Ser. Fiz. Energ. Nauk, 1983,
no. 2, p. 102.
B1. At the same process parameters, it was lower than
that on twoꢀcomponent membrane alloys: at least by
KINETICS AND CATALYSIS Vol. 51
No. 2
2010

CATALYTIC PROCESSES ON MEMBRANE PALLADIUM ALLOYS: I
265
11. Rozovskii, A.Ya. and Lin, G.I., Teoreticheskie osnovy
protsessa sinteza metanola (Theoretical Foundations of
Methanol Synthesis), Moscow: Khimiya, 1987, p. 74.
Simplest Gases with Group VIII Metals), AlmaꢀAta:
Nauka, 1979.
19. Serov, Yu.M., Gryaznov, V.M., and Gul’yanova, S.G.,
Izv. Vyssh. Uchebn. Zaved., Khim. Khim. Tekhnol.
1981, vol. 24, no. 3, p. 321.
,
12. Rapoport, F.M. and Il’inskaya, A.A., Laboratornye
metody polucheniya chistykh gazov (Laboratory Methꢀ
ods of Preparation of Pure Gases), Moscow: Gosꢀ
khimizdat, 1963, pp. 44–45, 238–242.
20. Vedeneev, V.I., Gurvich, L.V., Kondrat’ev, V.N.,
Medvedev, V.A., and Frankevich, E.L., Energii razryva
khimicheskikh svyazei: Spravochnik (Bond Dissociation
Energies: A Handbook), Moscow: Akad. Nauk SSSR,
1962.
21. Termicheskie konstanty veshchestv: Spravochnik (Therꢀ
mal Constants of Substances: A Handbook), Moscow:
Akad. Nauk SSSR, 1979, issues V–VIII.
13. Gol’tsov, V.A., Latyshev, V.V., Timofeev, N.I., and
Machikina, N.Yu., Izv. Vyssh. Uchebn. Zaved., Tsvetn.
Metall., 1977, no. 4, p. 117.
14. Serov, Yu.M., Doctoral (Chem.) Dissertation, Moscow:
Peoples’ Friendship University of Russia, 1999, p. 28.
22. Slovetskii, D.I., Khimiya plazmy (Plasma Chemistry),
15. Boreskov, G.K., Kataliz v proizvodstve sernoi kisloty
(Catalysis in Sulfuric Acid Production), Moscow: Gosꢀ
khimizdat, 1954, p. 71.
Moscow: Energoatomizdat, 1984.
23. Mikhalenko, I.I. and Yagodovskii, V.D., Kinet. Katal.
,
2004, vol. 45, no. 2, p. 256 [Kinet. Catal. (Engl. Transl.),
vol. 45, no. 2, p. 239].
16. Boreskov, G.K., Geterogennyi kataliz (Heterogeneous
Catalysis), Moscow: Nauka, 1988, pp. 233–248.
24. Mikhalenko, I.I., Chernova, A.E., Roshan, N.R.,
Chistov, E.M., and Slovetskii, D.I., Materialy mezhd.
nauch. konf. po nestatsionarnye energoꢀ i resursosbereꢀ
gayushchie protsessy i oborudovanie v khimicheskoi,
nanoꢀ i biotekhnologii (Proc. Int. Conf. on Unsteadyꢀ
State Energyꢀ and ResourceꢀSaving Processes and
Equipment for Chemical Technology, Nanotechnolꢀ
ogy, and Biotechnology), Moscow, 2008, p. 174.
17. Dubovkin, N.F., Spravochnik po uglevodorodnym topliꢀ
vam i ikh produktam sgoraniya (Handbook of Hydroꢀ
carbon Fuels and Their Combustion Products), Mosꢀ
cow: Gosenergoizdat, 1962.
18. Popova, N.M., Babenkova, L.V., and Savel’eva, G.A.,
Adsorbtsiya i vzaimodeistvie prosteishikh gazov s metalꢀ
lami VIII gruppy (Adsorption and Interaction of the
KINETICS AND CATALYSIS Vol. 51
No. 2
2010