The Order of Product Formation in the Partial Oxidation of Methane to Syngas

Article abstract of DOI:10.1023/A:1021698923193

The partial oxidation of methane to synthesis gas was studied in the presence of Pt- and Ni-containing catalysts. A substantial difference in the behavior of the isotopic label added with the CO₂ molecule in the case of the reaction on Pt/ZrO₂ and foam nickel. In the presence of Pt/ZrO₂, there was no noticeable transfer of the isotopic label to the CO molecules. On the Ni catalyst, ¹³CO was formed in substantial amounts, which could probably be explained by the redox reaction of ¹³CO₂ with metallic nickel under oxygen-free conditions behind the zone of the main reaction of methane oxidation.

Full text of DOI:10.1023/A:1021698923193

Kinetics and Catalysis, Vol. 43, No. 6, 2002, pp. 847–853. Translated from Kinetika i Kataliz, Vol. 43, No. 6, 2002, pp. 912–919.
Original Russian Text Copyright © 2002 by Tyulenin, Savkin, Sinev, Korchak.
The Order of Product Formation in the Partial
Oxidation of Methane to Syngas
Yu. P. Tyulenin, V. V. Savkin, M. Yu. Sinev, and V. N. Korchak
Semenov Institute of Chemical Physics, Russian Academy of Sciences, Moscow, 117977 Russia
Received February 5, 2002
Abstract—The partial oxidation of methane to syngas is studied in the presence of Pt- and Ni-containing cat-
alysts. The process kinetics does not provide unequivocal information on the order of formation of products
(including carbon oxides) when either methane–oxygen or methane–oxygen–CO2 mixtures are used. Experi-
ments with 13C-labeled carbon dioxide added show the difference in the behavior of the catalysts. In the pres-
ence of Pt/ZrO2, there is no noticeable transfer of the isotopic label to the CO molecules. On the nickel catalyst,
13CO is formed in substantial amounts, which can probably be explained by the redox reaction of 13CO2 with
metallic nickel under oxygen-free conditions behind the zone of the main reaction of methane oxidation.
INTRODUCTION
(2) The use of CO2 labeled with the 13C isotope and
the observation of the label transfer into CO.
Reliable information on the order of product forma-
tion in the conversion of methane to syngas (CO + H2)
is important for a search for new catalytic systems and
improving available systems and for optimizing the
process as a whole. From this standpoint, an important
question is whether carbon monoxide is the primary
gaseous product of oxidation (as considered for the
mechanism discussed in [1, 2]) or whether the main
pathway to its formation is the macrosteps of steam or
carbon dioxide reforming of methane, which follow the
primary oxidation of methane to carbon dioxide and
water (as proposed, for example, in [3–5]).
In the experiments of the first type, we used a setup
with a flow-type quartz reactor (i.d., 3.5 mm) and chro-
matographic analysis of products described in detail
earlier [6]. A specific feature of the reactor unit is the
presence of a reach-through hole normal to the axis in
the central hottest part of the furnace. The reactor was
mounted in this hole so that the catalyst layer is near the
longitudinal axis of the furnace. In the absence of the
reaction in this part of the reactor inside the furnace
(~10 mm), the constant temperature is maintained (with
an accuracy of ±2°C at working temperatures of up to
900°C). Outside of this hot zone, the temperature dras-
tically decreases (the gradient is several hundreds of
degrees for a furnace wall thickness of 2–3 mm). The
uniformity of the temperature field in the hot zone and
the drastic decrease in the temperature outside the fur-
nace allow for the highly reliable control of the reaction
space where methane conversion may occur, including
homogeneous reactions and reactions initiated by the
reactor wall surface or any other surfaces inside the
reactor. Moreover, the above configuration of the reac-
tor unit makes it possible to observe the temperature
field inside the catalyst layer and outside of it through
the open butt-ends and the transparent wall of the
quartz reactor without inserting any measuring ele-
ments that could change the conditions of heat and
mass transfer and affect the chemistry and kinetics of
the reactions. Moreover, the mobile fastening of the
furnace makes it possible to remove heating of the cat-
alyst zone of the reactor for the shortest time and to
replace the furnace with a socket which is not an active
heater but which decreases heat loss by the reaction
zone in the autothermal reaction due to its proper exo-
thermicity without external heating.
The goal of this study is to elucidate the order of
product formation in methane oxidation. Specifically,
we address the question of whether additional CO2
introduced to the reaction mixture is involved in the
formation of CO, and if it is, then at what stage and
under which conditions. To answer these questions, we
used the method of isotopic labeling (13C) of carbon
dioxide.
EXPERIMENTAL
The processes of partial oxidation and carbon diox-
ide reforming of methane to form syngas was studied
on the following catalysts: (1) the highly active sup-
ported Pt/ZrO2 catalyst analogous to that described in
[5] and (2) metallic foam nickel.
To solve the problem, we used the following methods:
(1) Comparison of the kinetic features of partial oxi-
dation and carbon dioxide reforming of methane to
syngas when they occur in separate (i.e., in the meth-
ane–oxygen and methane–CO2 mixtures) and coupled
(in the preliminary prepared methane–oxygen–CO2
mixture) regimes and
0023-1584/02/4306-0847$27.00 © 2002 MAIK “Nauka /Interperiodica”

848
TYULENIN et al.
Table 1. Composition of the reaction mixture in carbon di-
tion of the process, the sample changed its color from
faded gray to white and acquired a typical shine, point-
ing to the reduced surface and the absence of coke.
With an increase in the flow rate, a decrease in the yield
of CO and especially hydrogen was observed. An
increase in the pressure led to a decrease in the yields
of CO and hydrogen (in complete agreement with Le
Chatelier’s thermodynamic principle), while the above
nature of the product composition was preserved.
In oxygen-containing mixtures, a noticeable depen-
dence of the product composition on the feed rate was
observed: an increase in the flow rate led to an increase
in the concentration of CO and hydrogen and to a
decrease in the concentration of CO2 in the reaction
mixture.
An increase in the heat load by an increase in the
feed flow rate to 120–200 ml/min or 6.5–11 ml s–1 g–1
at a constant electric power of the furnace results in a
change in the process regime: a hot zone is formed in
the depth of the catalyst bed, which starts to move
toward the gas flow. Simultaneously, the temperature of
this zone increases. This transition regime lasts several
seconds. As a result, the hot zone moves to the front
layer of the catalyst bed and then the process becomes
stationary and autothermal. That is, at a constant flow
rate, it can occur when the voltage applied to the fur-
nace is shut off or drastically decreased depending on
the rate of the gas mixture supply.
oxide reforming of methane
W*,
ml/min
P(excess), [H2], [CH4], [CO2], [CO],
atm
vol % vol % vol % vol %
100
200
95
0.30
0.70
5.75
5.75
8.30
5.90
6.00
3.30
36.24 33.94
40.07 38.43
39.64 39.19
44.48 43.33
21.14
14.49
13.70
8.280
200
Note: Foam Ni, 300 mg; 49.5% CH –49.5% CO –~1% Ni ;
4
2
2
800°C. * W is the feed flow rate at atmospheric pressure.
Table 2. Changes in the composition of the reaction gas
when the initial mixture (1) 65% CH4–35% O2 is replaced by
(2) 60% CH4–35% O2–5% CO2
No. mix-
ture
[CO],
vol %
[CH4],
vol %
[CO2],
vol %
[H2],
vol %
1
2
26.40
30.30
23.90
29.55
9.55
40.10
38.80
11.40
Note: Foam Ni, 300 mg; 200 ml/min flow; the frontal temperature
is 520–535°C.
In this work we paid special attention to the appear-
ance of autothermal regimes, the determination of the
boundaries of their stability, and the regularities of
product formation under these conditions. Therefore,
we did not use any feedback to maintain a constant tem-
perature in the reactor. In the case of external heating,
its regime is set by applying a constant voltage to the
coils of the furnaces; that is, the supplied electric power
was controlled. The appearance of the very nonuniform
temperature profile due to the highly exothermic oxida-
tion of methane was observed by a change in the posi-
tion of the thermocouple in the reactor and (especially
in the runs when the thermocouple was not inserted into
the reactor) by sight.
The reactor unit was identical to that used in the iso-
topic labeling experiments. The only difference in this
case was the online mass spectrometric analysis of the
mixture described in detail in [7].
Preliminarily mixed gaseous mixtures were used
that contained methane, oxygen, and CO2. In the exper-
iments with labeled CO2, the fraction of 13C was 84%.
At all rates of flow in the studied range, the apparent
conversion of oxygen is at least 92–95% in the auto-
thermal regime.
The experimentally observed nature of the depen-
dence of the product yield on the flow rate points to the
primary formation of CO and hydrogen, because the
selectivity to these products increases with a decrease
in the contact time (that is, with an increase in the flow
rate). However, because CO2 reforming of methane can
occur on the same catalyst, such a dependence does not
shed light on the ratio between two possible pathways
of, for instance, CO formation: direct (from methane
and oxygen) and consecutive (via the primary complete
oxidation of methane to CO2 and water and further car-
bon dioxide and steam reforming).
To obtain further information on the reaction path-
ways, we carried out experiments with the replacement
of the methane–oxygen mixture with the triple meth-
ane–oxygen–CO2 mixture. In these experiments the
concentration of oxygen in the mixture was maintained
constant (35 vol %), and some portion of methane
(5 vol %) was replaced by CO2 after the beginning of
the autothermal regime at a flow rate of 200 ml/min.
The results obtained are summarized in Table 2.
RESULTS AND DISCUSSION
1. General Features of CO2 Reforming of Methane and
Methane Partial Oxidation in Separate and Combined
Regimes
As the above data show, the addition of CO2 to the
Nickel catalyst. On foam nickel, the efficient for- initial mixture results in an increase in the CO concen-
mation of syngas components was observed in meth- tration in the dry product gas and some decrease in the
ane–CO2 mixtures at temperatures >800°C and flow hydrogen concentration. In other words, the addition of
rates of the feed gas <5 ml s–1 g–1 (see Table 1). The cat- CO2 to the reaction mixture undoubtedly affects the
alyst activity increased with time, and upon the comple- process, but it is impossible to draw an unequivocal
KINETICS AND CATALYSIS Vol. 43 No. 6 2002

THE ORDER OF PRODUCT FORMATION IN THE PARTIAL OXIDATION
849
conclusion about the formation of CO and hydrogen via in the measured temperature to ~780°C. When the ther-
carbon dioxide reforming of methane. An alternative mocouple is used without a jacket, its readings increase
explanation for the effect of CO2 addition may be a shift to ~1060°C in the absence of additional insulation and
in the equilibrium in the reaction of carbon dioxide to 1070°C in the presence of additional insulation.
reforming of methane and an indirect (via this equilib- However, the pyrometric temperature of the thermo-
rium) effect on the position of the equilibrium in the junction is several hundred degrees lower. The reasons
water-gas shift reaction and in the steam reforming of for such a difference are not quite clear, but the exother-
methane. More definite conclusions can be drawn mic process of free radical decay on the metallic sur-
based on the experiments with 13C-labeled CO2 (see face of the thermocouple probably contributes most
below).
greatly to the observed phenomena. There is further
evidence for the fact that heat transfer from the hot zone
of methane oxidation does not determine thermocouple
heating behind the catalyst bed:
(i) The independence of thermocouple readings at a
distance of 5–10 mm downstream the catalyst bed on
the additional thermal insulation of the reactor;
A very important question is the localization of the
hot zone. If we place a small amount of quartz wool
before the front layer of the catalyst, then it has a higher
pyrometric temperature than the catalyst (even in its
front layer) under the conditions of the settled autother-
mal process. The readings of the thermocouple whose
thermojunction is deepened to ~1 mm in the front layer
of the catalyst are noticeably lower than the pyrometric
temperature of the hot zone (520–535 and ~750–
800°C, respectively). When the thermocouple is pulled
out of the catalyst bed, its readings start to depend on
the flow rate in a complex manner: an increase in the
load leads to an increase in the temperature measured by
the thermocouple to a certain limit (due to the intensifica-
tion of heat evolution); then, the cooling of the thermo-
junction with unreacted gases starts to dominate and leads
to readings that are too low. The pyrometric temperature
continues to increase with an increase in the load.
(ii) The observed coking of the thermocouple junc-
tion at a distance of 5 mm from the catalyst bed; a
growth of the friable coke layer leads to a decrease in
the temperature measured by the thermojunction (from
~450 to 125°C); at some points, the coke layer peels off
and the measured temperature drastically increases to
320°C and then the whole cycle repeats.
The possibility of secondary reactions of radicals
(formed in the catalytic activation of methane) on the
material of the chromel–alumel thermocouple with a
change in the apparent conversion and product compo-
sition was shown earlier (see, for instance, [6, 7]). In
this work, we failed to detect a change in the composi-
tion of the gaseous products of the reaction when the
jacket-free thermocouple was placed downstream the
catalyst bed. Apparently, we can only assume addi-
tional anaerobic methane conversion because oxygen is
almost completely absent from the gas phase. The
extent of such conversion is too small to be detected
against the background of the high methane conversion
at earlier process stages.
The above experiments show that the process prob-
ably occurs in the following manner: the active catalyst
generates active species (the products of incomplete
oxidation or free radials, which is more likely); the lat-
ter diffuse to some distances to the gas phase toward the
gas flow to the zone upstream the catalyst and initiate a
homogeneous reaction there, where the mixture is still
rich in oxygen. Nevertheless, the concentration of oxy-
gen in the mixture is insufficient for purely homoge-
neous gas-phase combustion. Note that there are no
steps in the homogeneous oxidation of methane where
a CO2 molecule is formed bypassing the primary for-
mation of CO (see, for instance, [8, 9]). In other words,
CO2 cannot be the primary product because it is only
formed via the reactions of CO with oxygen-containing
species, mostly radicals:
Pt/ZrO2 catalyst. At relatively low flow rates of the
mixtures 65% CH4–35% O2 and 60% CH4–35% O2–
5% CO2, methane reforming on the Pt/ZrO2 catalyst
only occurs in the case of external heating; the main prod-
ucts of the reaction are CO2 and water, and CO and hydro-
gen are formed in small amounts (usually, H2 ӷ CO).
If the mixture supply is sufficient for compensating
the heat loss (that is, if the flow rates of the methane–
oxygen mixture are rather high (>100 ml/min)), the
“light-off” of the reaction with fast self-heating occurs at
nearly room temperature. The regime of such a self-sus-
taining process is determined by the ratios of the resi-
dence times of the reaction mixture in different zones.
The most interesting case is when the amount of the
catalyst is small (the formal contact time is <0.01 s). In
this case, the following picture is observed for the set-
tled regime: at a relatively low temperature of the cata-
lyst (≤600°C), the space upstream the catalyst bed is
strongly heated along the mixture flow due to the exo-
thermic process of methane oxidation (to 800–900°ë).
Almost all the oxygen of the initial mixture is con-
sumed by the reaction before reaching the catalyst bed
for a period of ~0.01 s.
If a chromel–alumel thermocouple is placed in a
quartz jacket downstream the Pt/ZrO2 catalyst at differ-
ent distances, then its readings correspond to the tem-
perature profile along the reactor axis. In the typical set-
tled process, the temperature right behind the active
catalyst bed is ~580°C. Additional thermal insulation of
the reactor with an asbestos socket leads to an increase
.
.
CO + RO
CO2 + R
(I)
or
.
.
CO + RO2
CO2 + RO ,
(II)
KINETICS AND CATALYSIS Vol. 43 No. 6 2002

850
I, rel. units
TYULENIN et al.
I, rel. units
1
2
3
1
2
3
0
10
20
30
40
m/e
0
10
20
30
40
m/e
Fig. 1. Mass spectra of (1) the initial mixture 60% CH –
4
4
13
35% O –5% CO and the reaction mixtures when (2)
2
2
13
12
CO and (3) O are added in the autothermal regime on
2
2
the Pt/ZrO catalyst.
foam nickel.
2
2. Methane Conversion
where R is an organic (e.g., hydrocarbon) radical or a
hydrogen atom.
in the Mixture Containing 13CO2
Experiments with the addition of 13C-labeled carbon
Interestingly, the process selectivity can be changed
if a small amount of less active catalyst capable of com-
plete methane oxidation to CO2 (MgO) is placed
upstream the Pt/ZrO2 bed. The relative fraction of CO2
in the reaction gases increases, but the selectivity does
not change if the SiO2 (silica gel) packing is used.
These results provide a better understanding of the pro-
cess in the presence of this Pt-containing catalyst. It is
likely that, in the presence of the hot metallic surface,
the homogeneous oxidation of methane in the gas
dioxide to the initial methane-oxygen mixture were
carried out to reveal the ratio between the routes of CO
formation. In these experiments, after the autothermal
regime is settled, the ionic current of the mass spec-
trometer was recorded for m/e ranging from 1 to 48
(Figs. 1, 2). In this range, masses corresponding to the
fragments of the hydrogen, methane, and water mole-
cules (m/e = 1), molecular hydrogen ion (m/e = 2), frag-
ments of the methane (m/e = 12–16) and water mole-
cules (m/e = 16–18), 12CO (m/e = 28), 13CO (m/e = 29),
phase is initiated before the catalyst. We should stress molecular oxygen ion (m/e = 32), 12CO2 (m/e = 44), and
13CO2 (m/e = 45) were registered. The peak at m/e = 16
that the presence of the active catalyst is necessary for
is a complex combination of molecules of methane,
water carbon oxides, oxygen, and doubly ionized
molecular oxygen. The criterion for the formation of
CO from CO2 in our experiments is the appearance of
ions with m/e = 29 in the mass spectrum. Note that the
low-intensity signal of these ions is also registered
when the initial mixture is analyzed; it is likely that the
spectrum contains the signal of 13CO, which is a frag-
ment of the 13CO2 molecule and which is formed under
the action of the electron impact in the ionization cham-
ber of the mass spectrometer. Furthermore, even in the
mixtures that do not contain 13C-labeled molecules,
low-intensity signals with m/e = 29 and 45 and signals
with m/e = 28 and 44 were observed for the room-tem-
perature mixture that does not contain added CO2.
Apparently, all of them are the components of residual
gases in the vacuum chamber. Tables 3 and 4 show the
values of the peak intensities corresponding to the
molecular ions of carbon oxides with different isotopic
compositions registered under different conditions.
that, because the limiting reactant (oxygen) cannot be
converted in full in the mixture with this composition
under conditions of a purely gas-phase process at the
same temperatures. The product composition is deter-
mined by the process in the precatalytic zone: in com-
plete agreement with the regularities of homogeneous
oxidation, the main products are CO and hydrogen and
the fraction of CO2 increases when a material active in
the complete oxidation (MgO in our case) is placed in
the zone before the Pt-containing catalyst. When the
arrangement of zones (material beds) is reversed, there
is no effect from adding the second component.
Visual examination of the process of autothermal
regime settling in the presence of foam nickel (see
above) and the regularities of product formation led us
to assume that the predominant gas-phase oxidation of
methane in the precatalytic space also takes place in the
case of this catalyst.
KINETICS AND CATALYSIS Vol. 43 No. 6 2002

THE ORDER OF PRODUCT FORMATION IN THE PARTIAL OXIDATION
851
Table 3. Intensities of mass spectrum signals corresponding to reactants and the main products in the oxidation of methane
on the Pt/ZrO2 catalyst
m/e
Conditions
1
2
15
18
28
29
32
44
45
(28/44) (29/45)
Room temperature, 0.00318 0.00196 0.84448 0.07009 0.11150 0.00979 0.75111 0.00275 0.06331 40.55 0.00782
13CO2
ATR*, 13CO2
ATR, 12CO2
0.00704 0.01755 0.51893 0.52911 0.39557 0.01421 0.02719 0.23563 0.06007 1.22 0.106
0.00813 0.03494 0.48036 0.52353 0.46072 0.00837 0.03114 0.29301 0.00509 1.20 1.645
* ATR stands for the autothermal regime.
Table 4. Intensities of mass spectrum signals corresponding to reactants and the main products in the oxidation of methane
on foam Ni
m/e
Conditions
1
2
15
18
28
29
32
44
45
(28/44) (29/45)
Room temperature, 0.00318 0.00196 0.84448 0.07009 0.11150 0.00979 0.75111 0.00275 0.06331 40.55
0.00782
13CO2
ATR*, 13CO2
ATR, 12CO2
0.00455 0.10709 0.26829 0.22807 0.36673 0.04180 0.04011 0.26099 0.04941 0.988 0.741
0.00399 0.10724 0.17977 0.23041 0.49042 0.00934 0.04265 0.39113 0.00561 0.976 1.664
* ATR stands for te autothermal regime.
Table 3 shows that, in the presence of the Pt/ZrO2 were considerable. It is close to that observed in the
catalyst, the replacement of the initial 12CO2 mixture by reaction on the Pt/ZrO catalyst (~0.98 and ~1.21,
13CO2 does not lead to a considerable change in the respectively), whereas the ratio of the concentrations of
ratio of the intensities of peaks at m/e = 28 and 44, carbon oxides containing 13C changes by a factor of 7
which should have taken place in the case of the sub- when Pt/ZrO2 is replaced by foam nickel. It can be eas-
stantial contribution of the consecutive route of CO for- ily shown that, at the same conversions of carbon diox-
mation via the intermediate complete oxidation of ide in the reaction with methane, the ratio of concentra-
methane to CO2 according to the reaction
12CH4 + 13CO2 12CO + 13CO + 2H2.
tions of 13CO2 and13CO in the case of reaction (III)
should be half that of 12CO2 and 12CO in the reaction
(III)
12CH4 + 12CO2
212CO + 2H2.
(IV)
This conclusion is supported by the analysis of the
ratio of peak intensities at m/e = 28, 29, 44, and 45 in
the mass spectra registered with an increase from room
temperature to ~950 K: at a certain decrease (~5%) in
the intensity of the peak at m/e = 45 and the correspond-
ing increase in the intensity of the peak at m/e = 29
compared to the background values, the ratio of inten-
sities calculated taking into account the background
values (shown in boldface in Table 3) differs greatly
from the peak intensity ratio for the molecular ions
12CO and 12CO2. This also shows the preferred forma-
tion of CO molecules via the direct pathway immedi-
ately from methane and without the complete oxidation
of methane to CO2.
Comparison of data shown in Tables 3 and 4 shows
a substantial difference in the behavior of the isotopic
label added with the CO2 molecule in the case of the
reaction on Pt/ZrO2 and foam nickel. The ratio of signal
intensities at m/e = 28 and 44 in the case of the reaction
on the foam nickel is almost independent of the isotopic
Indeed, if carbon monoxide were only formed via
the consecutive mechanism, the expected ratio of signal
intensities in the carbon dioxide reforming of methane
could be estimated on the basis of the following
assumptions. Let us denote the yields of 13CO from
13CO2, 12CO from 13CO2, and 12CO from 12CO2 by
Y13/13, Y12/13, Y12/12, respectively. Under the same condi-
tions, we can write for reactions (III) and (IV)
Y12/12 = Y12/13 + Y13/13
Y13/13 = Y12/13
Y13/13 = 0.5Y12/13
.
Because
,
we conclude that
or
J(28)/J(44) = 2J(29)/J(45),
composition of CO2 added to the reaction. This would where J(m/e) are the intensities of the peaks corre-
be hard to explain if the contribution of reaction (III) sponding to a given m/e value in the mass spectrum.
KINETICS AND CATALYSIS Vol. 43 No. 6 2002

852
TYULENIN et al.
However, radically different relationships are et al. [11] concluded, based on their study of the pro-
observed in the experiment, and they cannot be cess of methane and CO2 interaction with Ni-contain-
explained by the occurrence of carbon dioxide reform- ing catalysts, that the main pathway of CO formation in
ing of methane:
carbon dioxide reforming of methane on these systems
is the interaction of CO2 with the products of methane
dissociation on the reduced nickel surface. Hydrogen is
formed in the dissociative chemisorption of methane,
which also requires a reduced metallic surface. If one
could show that residual oxygen, which is almost
always present in the reaction mixture of oxy reforming
and combined reforming, does not prevent the occur-
rence of this process at a high rate, there would be
strong evidence for the involvement of CO2 in this reac-
tion with methane. Note, however, that the experiments
were carried out in this work in strictly anaerobic con-
ditions. This fact makes it difficult to use these results
in analyzing the processes occurring in triple (i.e., oxy-
gen-containing) initial mixtures. Nevertheless, the
results reported in [11] support the possibility of the
above redox process on Ni-containing catalysts.
– in the presence of the Pt/ZrO2 catalyst:
J(28)/J(44) > 10J(29)/J(45);
– in the presence of foam nickel:
J(28)/J(44) ≈ 1.37J(29)/J(45).
The relationships observed can be explained by the
absence of any considerable involvement of carbon
dioxide in the reaction with methane in the presence of
oxygen. As concerns foam nickel, the experimentally
determined difference in the ratios of signal intensities
at m/e = 29 and 45 and m/e = 28 and 44 is much smaller
than possible in the case of reactions (III) and (IV)
(only ~30%). If we also take into account the possibil-
ity of steam reforming of 12CH4 molecules, this differ-
ence should be even greater. This led us to assume the
presence of the pathway of carbon dioxide conversion
into carbon monoxide via the reaction that is not con-
nected to methane conversion. It is most likely that the
interaction of CO2 with reduced nickel is such a pathway:
CONCLUSION
The results of this work show that the main pathway
for the formation of products of partial methane oxida-
tion (carbon monoxide and hydrogen) in the presence
of Pt-containing catalysts is the direct one, which is the
oxidation of methane with oxygen. Carbon dioxide
added to the reaction mixture is not involved in the
reaction at short contact times. On a foam nickel cata-
lyst, CO2 added to the reaction mixture partially con-
verts to oxidize reduced metal; however, in this case,
the direct pathway of CO formation from methane by
oxidation with oxygen also exists.
Special attention should be given to the process
occurring in the autothermal regime. Unfortunately, at
this stage it is hard to be certain regarding the mecha-
nism of its formation, its dynamics, and spatial local-
ization of steps. However, we can state that macroki-
netic factors play an important role here with a high
degree of probability. Two limiting cases are consid-
ered in the literature:
(i) Light-off of the active catalyst surface due to the
exothermic effect of the catalytic reaction [12]; in this
case, the noncatalytic reaction in the gas phase does not
occur;
(ii) light-off of the gaseous mixture heated by the
surface [13], which is considered as a hot but chemi-
cally inert wall.
The case considered in this work differs in that the
surface probably affects the process not only (or not
mainly) thermally, but it intensively generates active
species, including free radicals, and initiates the chem-
ical gas-phase reaction. Undoubtedly, secondary reac-
tions of chain carriers can occur on the same surface,
Ni + CO2
NiO + CO.
(V)
This reaction can be coupled with the reduction of
NiO by any reducing agent available in the reaction
mixture:
NiO + CH4
NiO + CO
NiO + H2
Ni + (H2, H2O, CO, CO2),
(VI)
(VII)
Ni + CO2,
Ni + H2O.
(VIII)
It can be easily shown that only the combination of
reactions (V) and (VI) can be considered as carbon
dioxide reforming of methane, whereas the combina-
tion of reactions (V) and (VIII) and the reverse gives us
the water-gas shift equilibrium (which is also accompa-
nied by the transfer of the label and an increase in the
ratio [13CO]/[13CO2]).
The key factor that determined the difference in the
behavior of the two catalysts under consideration may
be the different capabilities of the active metals to be
oxidized by carbon dioxide. In the case of nickel, this
process is quite possible (the calculation of the heat
effect according the data from [10] gives us the value
∆H = 38.5 kJ/mol), which allows the occurrence of
steps (V) and (VII) in the framework of the single redox
process resulting in the formation of isotopically
labeled carbon monoxide in considerable amounts. At
the same time, a reaction analogous to (V) for platinum
is much less probable (∆H ≈ 210 kJ/mol), and the car-
bon dioxide reforming on Pt-containing catalysts can
probably occur only under strictly anaerobic conditions
via a mechanism other than steps (V) and (VI).
Such a mechanism can be, for instance, the interac- and this fact complicates the picture even more. It is dif-
tion of CO2 with the products of dissociative chemi- ficult to say to what degree the light-off of the gaseous
sorption of methane on reduced metal. Thus, Bychkov mixture in the zone upstream the catalyst is determined
KINETICS AND CATALYSIS Vol. 43 No. 6 2002

THE ORDER OF PRODUCT FORMATION IN THE PARTIAL OXIDATION
853
5. O’Connor, A.M. and Ross, J.R.H., Catal. Today, 1998,
by heating from the hot catalyst surface and to what
degree this is due to the additional initiation of active
species escaping from the surface and diffusing toward
the flow. To answer all the related questions, additional
experiments with pyrometric measurements of the tem-
perature profile along the reactor axis (in the catalyst
bed and outside in both directions) are needed. Model-
ing is also needed that would take into account both
heterogeneous and homogeneous reactions of all spe-
cies which are present in the gas phase in methane oxi-
dation and the processes of mass and heat transfer.
Although there are recent papers in which the modeling
of complex heterogeneous–homogeneous processes is
approached (see, for instance, [14–20]), this task
remains rather complex and has no general solution for
schemes that differ in chemical and macrokinetic sys-
tems “gas-phase reaction–solid catalyst.”
vol. 46, p. 203.
6. Tulenin, Yu.P., Sinev, M.Yu., Savkin, V.V., and Kor-
chak, V.N., Stud. Surf. Sci. Catal., 1997, vol. 110, p. 757.
7. Tyulenin, Yu.P., Sinev, M.Yu., Savkin, V.V., Korchak, V.N.,
andYan,Yu.B., Kinet. Katal., 1999, vol. 40, no. 3, p. 405.
8. Warnatz, J., Combustion Chemistry, Gardiner, W.C., Jr.,
Ed., New York: Springer, 1984.
9. Tsang, W. and Hampson, R.F., J. Phys. Chem. Ref. Data,
1986, vol. 15, p. 1087.
10. Fiziko-khimicheskie svoistva okislov (Physicochemical
Properties of Oxides), Samsonov, G.V., Ed., Moscow:
Metallurgiya, 1969.
11. Bychkov, V.Yu., Krylov, O.V., and Korchak, V.N., Kinet.
Katal., 2002, vol. 43, no. 1, p. 94.
12. Rozovskii, A.Ya., Kinetika topokhimicheskikh reaktsii
(The Kinetics of Topochemical Reactions), Moscow:
Nauka, 1980.
13. Kondrat’ev, V.N. and Nikitin, E.E., Kinetika i mekha-
nizm gazofaznykh reaktsii (Kinetics and Mechanism of
Gas-Phase Reactions), Moscow: Nauka, 1974.
14. McCarty, J.G., Methane Conversin by Oxidative Pro-
cesses, Wolf, E.E., Ed., New York: Van Nostrand Rein-
hold, 1992, p. 320.
ACKNOWLEDGMENTS
This work was supported by the Russian Foundation
for Basic Research (grant no. 01-03-32524). We thank
Professor O.V. Krylov for his interest in this work.
15. Sinev, M.Yu., Catal. Today, 1992, vol. 13, p. 561.
16. Sinev, M.Yu., Catal. Today, 1995, vol. 24, p. 389.
REFERENCES
17. Schmidt, L.D., Deutchmann, O., and Goralski, C.T., Jr.,
Stud. Surf. Sci. Catal., 1998, vol. 119, p. 685.
1. Hickman, D.A., Haupfear, E.A., and Schmidt, L.D.,
Catal. Lett., 1993, vol. 17, no. 3/4, p. 223.
18. Beretta, A., Groppi, G., Majocchi, L., and Forzatti, P.,
Appl. Catal. A, 1999, vol. 187, no. 1, p. 49.
2. Hickman, D.A. and Schmidt, L.D., Science, 1993,
19. Schmidt, L.D., Stud. Surf. Sci. Catal., 2001, vol. 136,
vol. 259, no. 5093, p. 343.
p. 1.
3. Dissanayake, D., Rosynek, M.P., Kharash, K.C.C., and
20. Deutchmann, O., Schwiedernoch, R., Maier, L.I., and
Chatterjee, D., Stud. Surf. Sci. Catal., 2001, vol. 136,
p. 251.
Lunsford, J.H., J. Catal., 1991, vol. 132, p. 117.
4. Krylov, O.V., Usp. Khim., 1992, vol. 61, p. 2040.
KINETICS AND CATALYSIS Vol. 43 No. 6 2002