Unsteady-state direct partial oxidation of methane to synthesis gas in a fixed-bed reactor using AFeO3 (A = La, Nd, Eu) perovskite-type oxides as oxygen storage

Article abstract of DOI:10.1021/jp063490b

Direct partial oxidation of methane to synthesis gas on AFeO₃ (A = La, Nd, Eu) oxides by a novel sequential redox cyclic reaction in the absence of gaseous oxygen was investigated over a fixed-bed reactor. These oxides were prepared by the sol-gel method and characterized by X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) techniques. XRD analysis showed that all AFeO₃ (A = La, Nd, Eu) oxides, calcined at 1173 K, are single-phase perovskites. The CF₄-TPSR/MS and continuous reaction experiments indicated that the AFeO₃ (A = La, Nd, Eu) oxides provide mostly oxygen species, as the sole oxidant originated from lattice oxygen instead of gaseous oxygen, which can oxidize CH₄ to synthesis gas with high selectivity in the absence of gaseous oxygen. In terms of material economics and the amount of oxygen species for synthesis gas formation, the LaFeO₃ sample exhibits the best performance among these tested AFeC₃ oxides for synthesis gas production. The pulse experiments at different temperatures showed that the rate of oxygen migration during the CH₄ reaction with LaFeO₃ is strongly affected by the reaction temperature, and increases with rising temperature, which is favorable to much more CH ₄ selective oxidation at high temperature. The two types of oxygen species are identified by experiments of continuous reactions and pulses, and confirmed by XPS. Methane can be converted selectively to synthesis gas by consumption of lattice oxygen, and general carbonaceous deposits on the catalyst surface do not occur under the appropriate reaction conditions by sequential redox cycles. The performance of selective oxidation of CH₄ to synthesis gas can be recovered by reoxidation using gaseous molecular oxygen; the LaFeO₃ oxide maintains relatively high catalytic activity and structural stability in redox atmospheres.

Full text of DOI:10.1021/jp063490b

J. Phys. Chem. B 2006, 110, 22525-22531
22525
Unsteady-State Direct Partial Oxidation of Methane to Synthesis Gas in a Fixed-Bed
Reactor Using AFeO3 (A ) La, Nd, Eu) Perovskite-Type Oxides as Oxygen Storage
Xiao Ping Dai,†,‡ Ran Jia Li,‡ Chang Chun Yu,*,‡ and Zheng Ping Hao*,†
Research Center for Eco-EnVironmental Science, Chinese Academy of Sciences,
Beijing 100085, People’s Republic of China, and The Key Laboratory of Catalysis CNPC,
UniVersity of Petroleum, Beijing 102249, People’s Republic of China
ReceiVed: June 6, 2006; In Final Form: August 19, 2006
Direct partial oxidation of methane to synthesis gas on AFeO3 (A ) La, Nd, Eu) oxides by a novel sequential
redox cyclic reaction in the absence of gaseous oxygen was investigated over a fixed-bed reactor. These
oxides were prepared by the sol-gel method and characterized by X-ray diffraction (XRD) and X-ray
photoelectron spectroscopy (XPS) techniques. XRD analysis showed that all AFeO3 (A ) La, Nd, Eu) oxides,
calcined at 1173 K, are single-phase perovskites. The CH4-TPSR/MS and continuous reaction experiments
indicated that the AFeO3 (A ) La, Nd, Eu) oxides provide mostly oxygen species, as the sole oxidant originated
from lattice oxygen instead of gaseous oxygen, which can oxidize CH4 to synthesis gas with high selectivity
in the absence of gaseous oxygen. In terms of material economics and the amount of oxygen species for
synthesis gas formation, the LaFeO3 sample exhibits the best performance among these tested AFeO3 oxides
for synthesis gas production. The pulse experiments at different temperatures showed that the rate of oxygen
migration during the CH4 reaction with LaFeO3 is strongly affected by the reaction temperature, and increases
with rising temperature, which is favorable to much more CH4 selective oxidation at high temperature. The
two types of oxygen species are identified by experiments of continuous reactions and pulses, and confirmed
by XPS. Methane can be converted selectively to synthesis gas by consumption of lattice oxygen, and general
carbonaceous deposits on the catalyst surface do not occur under the appropriate reaction conditions by
sequential redox cycles. The performance of selective oxidation of CH4 to synthesis gas can be recovered by
reoxidation using gaseous molecular oxygen; the LaFeO3 oxide maintains relatively high catalytic activity
and structural stability in redox atmospheres.
Introduction
engineering design still occur and should be solved before an
industrial-scale implementation.4
The production of synthesis gas by steam reforming is a
highly energy-consuming process, and about 60% of the cost
of the overall chemical production process is due to synthesis
gas generation by steam reforming.1 Compared with conven-
tional steam reforming, which requires high energy consumption
and intensive capital, the partial oxidation of methane is more
energy-efficient, requires a lower investment, and produces
synthesis gas with a H2/CO ratio of ca. 2, which is suitable for
methanol and Fischer-Tropsch synthesis. Therefore, it has
attained a lot of attention. However, partial oxidation of methane
occurs between gas-phase oxygen and methane, which limits
CO selectivity and hydrogen selectivity because of the complete
oxidation of methane. There are also serious safety concerns
due to the direct mixing of the highly explosive feed gas CH4
with oxygen at high temperatures.2 To avoid the above problems,
intensive efforts have been made to use membrane reactors for
the partial oxidation of methane to synthesis gas. In a membrane
reactor, methane and air flow on opposite sides of the
membrane, and methane is oxidized to CO/H2 by lattice oxygen
supplied from the air side of the membrane.3 However, many
practical problems such as reactor scale-up and sealing and
A novel two-step process for the production of synthesis gas
from methane and air has been recently proposed by Van Looij.
Methane is oxidized to CO2 and H2O, which then reform
methane for the production of synthesis gas in the absence of
gas-phase oxygen by using a metal oxide acting as an oxygen
storage.5 Manganese, iron, and cerium oxides and cerium-
containing mixed oxides have been studied, but poor thermal
stability, carbon deposition, or sintering at high temperatures
make it impossible to use them for practical application.6-11
Perovskite materials have been used quite effective for
catalytic oxidation reactions including hydrogenation and hy-
drogenolysis of hydrocarbons, CO oxidation, ammonia oxida-
tion, and catalytic combustion.12 Due to their variable structure
and chemical composition, much attention has been paid to
perovskite-type oxides with a general formula ABO3 (where A
and B are usually rare earth and transition metal cations,
respectively).13 Perovskite oxides, which usually exhibit excel-
lent redox properties, high oxygen mobility, and thermal
stability, are good oxygen suppliers. However, the use of these
materials for the direct conversion of methane to synthesis gas
without gas-phase oxygen is rarely reported. In this work, a
series of perovskite AFeO3 oxides (A ) La, Nd, Eu) are
prepared and characterized, and their catalytic activities in the
direct conversion of methane to synthesis gas in a fixed-bed
reactor are investigated in the absence of gas-phase oxygen by
* Corresponding authors. Telephone: 86-10-89707447 (C.C.Y.); 86-10-
62923564 (Z.P.H.). E-mail: lngcc@cup.edu.cn (C.C.Y.); zpinghao@rcees.ac.cn
(Z.P.H.).
† Chinese Academy of Sciences.
‡ University of Petroleum.
10.1021/jp063490b CCC: $33.50 © 2006 American Chemical Society
Published on Web 10/25/2006

22526 J. Phys. Chem. B, Vol. 110, No. 45, 2006
Dai et al.
TABLE 1: Tolerance Factor (t), Lattice Parameters (a, b, c), Interplanar Distance (d121), and Average A-O and Fe-O Bond
Distances of AFeO3 (A ) La, Nd, Eu) Oxides (r(Fe3+) ) 0.645 Å, r(O2-) ) 1.4 Å)
catalyst
A
3+ ionic radius (Å)
tolerance factor
a (Å)
b (Å)
c (Å)
d121 (Å)
A-O (Å)
Fe-O (Å)
LaFeO3
NdFeO3
EuFeO3
1.36
1.27
1.21
0.954
0.923
0.902
5.542
5.579
5.335
7.857
7.757
5.560
5.564
5.450
7.630
2.777
2.749
2.710
2.78
2.74
1.97
1.96
1.94
1.39
temperature programmed surface reaction, pulse reactions,
continuous reactions, and sequential redox cycles. Catalysts with
good performance and the reaction mechanism of oxygen
species are desirable to be obtained.
to calculate the amount of CH4, CO, and CO2 on the catalysts
taking into account a previous calibration of the mass spec-
trometer. The calibration of the mass spectrometer was carried
out with standard mixtures of known composition. The cracking
coefficients of methane, CO, and CO2 have been determined
and used to calculate their concentrations. The CO selectivity
is defined as CO/(CO + CO2) × 100%.
Experimental Section
Material Preparation. The perovskite AFeO3 oxide (A )
La, Nd, Eu) catalysts were prepared by the sol-gel method.
Glycine was added to an aqueous solution of metal nitrates to
have a ratio NH3/NO3 ) 1.05. Water was evaporated slowly
from the mixed solution at 353-363 K until a viscous gel was
obtained. The gel was then heated at 523 K for 30 min with a
fast exothermic reaction, and the yielded brown powder precur-
sor was slowly calcined at 1173 K for 10 h. The samples were
crushed, ground, and sieved for experiment.
Material Characterization. Phase analysis and lattice pa-
rameters were performed by X-ray powder diffraction (XRD)
using a ShiDU XRD-6000 with Ni-filtered Cu KR radiation.
The X-ray photoelectron spectra were obtained with a VG
ESCA LAB 210 system using Mg KR radiation under ultrahigh
vacuum (5 × 10-9 Pa). The spectra have been calibrated
relatively to the C(1s) bond energy of contamination carbon,
which was fixed at 284.6 eV.
Temperature Programmed Surface Reaction. Temperature
programmed surface reaction with CH4 (CH4-TPSR/MS) was
performed on a fixed-bed quartz microreactor (4.5 mm i.d.)
packed with 0.2 g catalyst samples. In the TPSR experiments,
the mixture of 11 vol % CH4/Ar was used at a gas flow rate of
23 mL/min; linear temperature was increased at a rate of 10
K/min. The analysis of the reactants and products was carried
out using an on-line mass spectrometer (AMTEK dycor System
1000), equipped with a fast response inlet capillary system. The
quadrupole mass spectrometer can detect eight mass channels
simultaneously with a minimum dwell time of 3 ms.
Pulse Reactions. The pulse reactions were performed by
injecting in a carrier gas (Ar) flowing continuously through the
catalyst bed. A pulse of CH4 gas was introduced with a six-
port gas-sampling valve equipped with measuring tubes (0.205
mL). The composition of the effluent gas from the reactor was
monitored by a quadrupole mass spectrometer. The pulse
measurements were carried out on catalyst (0.2 g) at 1173 K,
with a carrier flow rate of 23 mL/min.
Results and Discussion
Material Characterizations. A perovskite-type oxide has an
ABO3-type crystal structure wherein cations with a large ionic
radius have 12 coordination to oxygen atoms and occupy
A-sites, and cations with a smaller ionic radius have 6
coordination and occupy B-sites. The ionic radii of the two
cations A and B in a perovskite structure (ABO3) compound
must be in a limited range, i.e., rA > 0.090 nm and rB > 0.051
nm. The tolerance factor of the perovskite compound t (t ) (rA
14
x
+ rO)/ 2(rB + rO)) should be in the range of 0.8-1.0.
The tolerance factor (t) values calculated by the above
equation are listed for AFeO3 (A ) La, Nd, Eu) oxides in Table
1. It is seen that all tolerance factor (t) values of the AFeO3 (A
) La, Nd, Eu) oxides are in the range of 0.8-1.0. Thus, it is
possible for these oxides with stoichiometric composition to
form a single perovskite phase.
The XRD spectra of AFeO3 (A ) La, Nd, Eu) samples
calcined at 1173 K are shown in Figure 1. The results show
that AFeO3 (A ) La, Nd, Eu) samples are single-phase
perovskites, compared with the XRD data of the perovskite-
structure AFeO3 oxide (JCPDS Card Nos. 37-1498, 25-1149,
and 47-0066, respectively). It is notable that the crystallinity of
the samples is much higher for LaFeO3 and lower for NdFeO3
and EuFeO3. Also, Figure 1 clearly shows that the X-ray lines
are continuously shifted toward higher diffraction angles with
the decrease of the A ionic radii.
Table1 lists the observed values of lattice parameters and the
average A-O and Fe-O bond distances for AFeO3 (A ) La,
Continuous Reactions. Continuous reactions were carried
out in a quartz microreactor. The AFeO3 oxides were pretreated
with 11% O2/Ar at 23 mL/min at 1173 K for 30 min, and then
flow of the reactant gas (11% methane in argon) at 23 mL/min
was initiated and maintained for 4 min.
Sequential Redox Cycles. Sequential redox cycles over
LaFeO3 oxide were performed in a fixed-bed reactor for the
production of synthesis gas at 1173 K between reductive (CH4/
He) and oxidative atmospheres (O2/Ar) (denoted as “redox”).
The total gas flow rate was kept at 23 mL/min. Ar/He was used
as dilution gas and tracer gas.
The different reaction products were continuously measured
with the mass spectrometer at m/e ) 2 (H2), m/e ) 15 (CH4),
m/e ) 18 (H2O), m/e ) 28 (CO), m/e ) 32 (O2), and m/e ) 44
(CO2). The CH4, CO, H2, and CO2 pulses were injected in order
Figure 1. XRD spectra for AFeO3 (A ) La, Nd, Eu) samples calcined
at 1170 K.

Direct Partial Oxidation of CH4 to Synthesis Gas
J. Phys. Chem. B, Vol. 110, No. 45, 2006 22527
TABLE 2: O Species Amounts for CH4 Oxidation
less reactive
O speciesa
µmol/g catalyst µmol/g catalyst
very reactive
O speciesa
O species amt
for CO formationb
µmol/g catalyst
catalyst
LaFeO3
NdFeO3
EuFeO3
669.0
422.5
368.0
242.2
504.8
184.0
346.0
286.4
266.7
a Continuous reactions at 1173 K. b CH4-TPSR/MS.
the tolerance factor.17 This proves the easier oxygen mobility
for a high tolerance factor of perovskite oxide, such as LaFeO3
oxide.
Continuous Reactions over AFeO3 Oxides. The methane
oxidation experiments shown here are inherently transient since
the oxidation level of the solid declines with the consumption
of lattice oxygen. The duration of flow experiments was
therefore, limited to 4 min in all runs.
Figure 2. CH4-TPSR/MS profiles in 11% CH4/Ar over AFeO3 oxides.
Nd, Eu) samples. As seen, lattice symmetry changes from
orthorhombic to nearly cubic in the series EuFeO3 f NdFeO3
f LaFeO3. The lattice parameter c in LaFeO3 approaches the
value of a, and b is equal to the double length of the cubic
pseudocell dimension. The average A-O and Fe-O bond
distances were evaluated by taking into account half the value
Transient product concentrations for the reaction of 11% CH4/
Ar at 1173 K on AFeO3 (A ) La, Nd, Eu) oxides are shown in
Figure 3. Figure 3a indicates that the reactions were initially
very fast and produced significant quantities of carbon dioxide
and water but not the desired products (carbon monoxide and
hydrogen) at the preliminary stage. When the surface state was
changed, carbon dioxide and water formation declined rapidly
to a very low level, while methane intensity quickly declined,
which means that methane conversion was sharply enhanced.
With the surface oxygen species consumed, CO and H2
intensities increased steadily with time, while the CH4 intensity
decreased gradually, and the synthesis gas was the dominant
product. After 60 s of reaction in CH4/Ar, the intensities of CH4,
H2, and CO reached steady-state levels. The delay was observed
in the appearance of CO2 and the other products (such as CO
and H2), indicating that, in the absence of gas-phase oxygen,
methane first reacted with the very reactive oxygen species to
form CO2 and H2O. After these oxygen species declined, the
less reactive oxygen species reacted selectively with methane
to form CO and H2. From the trend of the methane intensity
change, it was noted that the consumption of lattice oxygen on
surface enhances oxygen mobility from the bulk toward the
surface. Exposure of LaFeO3 oxides to a reduced atmosphere
can generate oxygen vacancies, owing to the loss of lattice
oxygen and reduction of oxidation state of the Fe ion simulta-
neously.18 Similarly, Tejuca19 reported the reduction of Fe3+ to
Fe2+ and further to metallic Fe in LaFeO3 oxide.
of b as the cell dimension of the idealized cubic unit cell (A-O
15
x
) b 2/4, Fe-O ) b/4).
The values of unit cell parameters of LaFeO3, NdFeO3 and
EuFeO3 are similar to those reported in the literature.15,16 The
Fe-O bond is slightly stronger in EuFeO3 than in LaFeO3 and
NdFeO3; the Fe-O bond strength should be considered to be
related to catalytic performance and the chemical structure of
AFeO3 (A ) La, Nd, Eu) oxides. The interplanar distance (d121
)
increases with the increase of A3+ ionic radii, resulting in the
increase of the Fe-O bond distance, which is related to the
electronic structure of A-site cation.
Temperature Programmed Surface Reaction (CH4-TPSR/
MS). Figure 2 shows a variation of product formation and
methane conversion as a function of reaction temperature during
CH4-TPSR/MS over the AFeO3 (A ) La, Nd, Eu) oxides.
It was shown that the main products are CO and H2. Only
small amounts of CO2 and H2O were detected over the AFeO3
(A ) La, Nd, Eu) oxides. This demonstrates that the AFeO3
(A ) La, Nd, Eu) oxides provide mainly oxygen species, which
can oxidize CH4 to CO/H2 with high selectivity, and only a
small quantity of methane was oxidized to CO2. No reaction
was monitored over AFeO3 (A ) La, Nd, Eu) oxides below
860 K; greater CH4 conversion took place in CH4-TPSR on
AFeO3 (A ) La, Nd, Eu) oxides above 1000 K. Obviously the
amount of oxygen cannot be ascribed to only the surface oxygen,
because of the amount of oxygen at the first surface layer on
the basis of low BET surface area (ca. 10 m2/g oxide calcined
at 1173 K). Therefore, it is clear that the bulk lattice oxygen of
LaFeO3 oxide should participate in the partial oxidation of CH4
to synthesis gas. The oxygen species are transported by the
surface reaction and bulk ion conduction mechanism. It should
be noted that in Figure 2 the diffusion rates of oxygen have
obvious differences due to the various gradients of the CO and
H2 formation curves. The diffusion rate of oxygen has the
highest value in LaFeO3 oxide according to the gradients of
CO formation curves. The amounts of CO and H2 are higher
on LaFeO3 oxide than on NdFeO3 and EuFeO3 oxides, and this
indicates that the amount of oxygen species for synthesis gas
formation is the highest in LaFeO3 oxide, as shown in Table 2.
This behavior reflects the increase in Fe-O bond strength
deduced from bond distance considerations and confirms the
major role of lattice oxygen species in partial oxidation of
methane. The B-O bond energy lowers with the decrease of
After the reductive reaction with methane, the catalyst was
reoxidized by 11 mol % O2/Ar for 4 min at 1173 K isothermal
Figure 3. Responses of the transient from Ar to 11 mol % CH4/Ar
over AFeO3 oxides at 1173 K.

22528 J. Phys. Chem. B, Vol. 110, No. 45, 2006
Dai et al.
Figure 4. 4: CH4 conversion and CO selectivity vs temperature of pulse reaction over LaFeO3 oxide.
Figure 5. Total amount of products in 15 pulse reactions over LaFeO3
oxide at different temperatures.
Figure 6. Total CO selectivity and CO2 selectivity in 15 pulse reactions
over LaFeO3 oxide at different temperatures.
conditions. In the first 90 s, no oxygen was detected in the
effluent gas. The oxygen storage on oxides was recovered very
efficiently. During the reoxidation step, the signal of CO2 and
CO was not detected over LaFeO3 oxide, which implies that
almost no carbon was deposited on the oxides during the
methane reaction step under the reaction conditions.
The same trend occurred over NdFeO3 and EuFeO3 oxides
as shown in Figure 3b,c. The comparison of the catalytic
behaviors for the three oxides indicates that the NdFeO3 and
EuFeO3 oxides exhibit lower methane conversion, are less
selective for synthesis gas, and produce substantial quantities
of carbon dioxide and water. From the gradients of CO and H2
profiles, it can be seen that the diffusion rate of lattice oxygen
from the bulk toward the surface is lower over NdFeO3 and
EuFeO3 oxides than over LaFeO3 oxide. The O species amount
for CH4 oxidation is shown in Table 2. The maximum value of
less reactive oxygen species released by LaFeO3 oxide confirms
its best catalytic performance for synthesis gas formation. This
agrees with the results in CH4-TPSR experiments.
for synthesis gas production in CH4-TPSR and continuous
reactions. Therefore, it is deduced that LaFeO3 is the best
catalyst among the tested AFeO3 oxides, so it was chosen for
in-depth study.
Pulse Reactions over LaFeO3 Oxide at Different Tem-
peratures. The reactor temperature was raised under oxygen
flow, and a sweeping step with inert gases was carried out before
the pulse reaction. The results of catalytic tests carried out by
pulse reaction with CH4/Ar in the absence of gas-phase oxygen
are presented in Figures 4-6 for LaFeO3 oxide at different
temperatures.
For different temperatures, the reaction started with relatively
high conversion of methane in the first pulse, i.e., 20% for 1073
K and 42% for 1173 K (Figure 4a). Afterward, there was a
sharp decline of methane conversion from pulse 1 to pulse 2,
and a quick increase thereafter. On the other hand, the carbon
monoxide selectivity rose sharply from a very low level to above
90% within the first few pulses. There are pronounced differ-
ences between different temperatures. Higher temperature gives
a higher initial methane conversion and higher carbon monoxide
selectivity (Figure 4b). CH4 conversion and CO selectivity
increase with reaction temperature; especially the tendency is
sharper at high reaction temperatures. From Figures 5 and 6,
In terms of economics of material, considering the amount
of selective oxygen species and the diffusion rate of lattice
oxygen from the bulk toward the surface, the LaFeO3 sample
exhibits the best performance among these tested AFeO3 oxides

Direct Partial Oxidation of CH4 to Synthesis Gas
J. Phys. Chem. B, Vol. 110, No. 45, 2006 22529
Figure 7. Transient responses of the cyclic reaction between O2/Ar flow (10 s) and CH4/He flow (30 s) over LaFeO3 oxide at 1173 K: 1, H2; 2,
CO; 3, CH4; 4, H2O; 5, Ar; 6, CO2; 7, O2.
the total amounts of CO and H2 increased sharply when the
reaction temperature was increased from 1073 to 1173 K, due
to the higher diffusion rate at high temperature. The total
selectivity of CO increased with reaction temperature, and much
more of CH4 was selectively oxidized to CO and H2 at high
temperatures.
The explanation could be that first CO2 is likely produced
by the reaction between methane and the chemisorbed oxygen
species on oxide.20 After the concentrations of surface oxygen
species decline, lattice oxygen is diffused from the bulk to the
surface when the pulse reaction of CH4 is carried out continu-
ously. With the oxygen species on the surface of LaFeO3 oxide
consumed, the increased concentration of lattice oxygen vacan-
cies provides pathways of oxygen transport through the lattice,
Figure 8. CH4 conversion and CO selectivity as a function of redox
and leads to enhancement of oxygen diffusivity. It is well-known
that oxygen vacancies can be formed easily when a perovskite
is heated at higher temperature. The migration rate of oxygen
during the CH4 reaction with LaFeO3 oxide was strongly
affected by the reaction temperature, especially at high reaction
temperatures. The rate of oxygen migration from the bulk to
the surface increased with rising temperature, and more CO and
H2 could be formed due to selective oxidation of CH4. The
difference in CH4 conversion at different temperatures is due
to the high activation energy of the activated oxygen, which
implies the importance of the collective properties of perovs-
kites, e.g., the bond energy of the lattice oxygen.21
Production of Synthesis Gas by Sequential Redox Cycles
over LaFeO3. The oxidation of hydrocarbons is supposed to
occur on the perovskite surface by means of an intrafacial
catalytic mechanism, in which the adsorbed oxygen is partly
consumed and regenerated during a continuous cycle.4,22,23
DuPont commercialized a circulating fluidized bed concept in
which catalyst is shuttled between a net oxidizing and a net
reducing environment.24 Sequential redox cycles were carried
out in a fixed-bed reactor for the production of synthesis gas
over LaFeO3 oxide catalyst at 1173 K between O2/Ar (11 mol
% O2) and CH4/He (11 mol % He). The total gas flow rate was
kept at 25 mL/min. Ar/He was used as dilution gas and tracer
gas. Related results are shown in Figures 7 and 8.
cycle number.
Figure 7 shows the product responses during the cyclic
reactions between O2/Ar oxidation for 10 s and CH4/He
reduction for 30 s over the LaFeO3 catalyst at 1173 K. As
described previously (Figure 3a), CO2 was formed first and then
CO and H2 were detected for the first cycle. However, from
the second redox cycle to the sixth redox cycle, only a small
quantity of CO2 was produced, while CO and H2 were the
dominant products. Fe4+ was present in small amounts also in
the LaFeO3 perovskite, and partial reduction of Fe4+ to Fe2+
,
Fe3+ to Fe2+ could likely occur at higher temperature even if it
does not lead to the loss of the perovskite structure.15 It has
also been reported that, once the reaction temperature exceeded
1173 K, the sintering of iron oxide occurred.25 With the increase
of cycle number from the seventh redox cycle, the amount of
CO2 increased, which may be due to the change of the redox
properties of Fe. From the 50 redox cycles at 1173 K, the
LaFeO3 oxide maintained relatively high catalytic activity and
structural stability in XRD characterization (shown in Figure
9), but the products provide evidence of the appearance of
deactivation phenomena (shown in Figure 8). When the lattice
oxygen consumed at reductive atmospheres (CH4/He) is not
sufficiently restored at oxidative atmospheres for 10 s, the
oxidation level of the catalyst may not reach the initial state,

22530 J. Phys. Chem. B, Vol. 110, No. 45, 2006
Dai et al.
types of oxygen presented on the pre- and post-reaction LaFeO3
oxide surface. These two oxygen species can be defined as a
very reactive oxygen species for the CO2 formation and a less
reactive oxygen species for the synthesis gas formation. The
line at ca. 528.9 eV is assigned to the oxygen in the lattice, and
the one at ca. 531.1 eV is assigned to adsorbed oxygen
containing species.26-28 Omata et al.29 reported that two kinds
of oxygen species existed in substituted SrCoO3 perovskite, and
they are responsible for selective oxidation and complete
oxidation, respectively. This is in agreement with the results of
our experiments. It is clear that the lattice oxygen decreases
after reacting with 11 mol % CH4/Ar at 1173 K. The prereaction
oxide has a higher value of Olat/Oads (ca. 0.74) than the
postreaction oxide (ca. 0.49) on the surface, and the value of
Olat/Oads decreases from 0.58 to 0.49 with the reaction time
increased from 75 to 180 s with 10% CH4/Ar. The different
amounts of the lattice oxygen indicate that the oxygen vacancies
increase due to the reaction with methane. The amount of
adsorbed oxygen containing species is difficult to determine
since the oxides are inevitable owing to contact with air. This
is consistent with the results obtained from unsteady-state
reactions, which indicate CO2 formation first and then CO
production. As Voorhoeve et al. proposed,23 two different
mechanisms of catalytic oxidation should be distinguished: (i)
a suprafacial process in which the catalyst surface provides a
set of electronic orbitals of proper symmetry and energy for
the bonding of reactants and intermediates, and (ii) intrafacial
reactions in which the catalyst participates as a reagent that is
partly consumed and regenerated in a continuous cycle. In the
former mechanism, the oxygen vacancies at the catalyst surface
play a relevant role for the adsorption and activation of the
reactant molecules, whereas in the latter mechanism, the metal
cation of the catalyst oxide should have the capacity to change
the oxidation state and thus to participate in a redox cycle.
Figure 9. Comparison of XRD profiles over pre- and postreaction
LaFeO3 catalyst.
Figure 10. XPS of O(1s) over pre- and postreaction LaFeO3 oxides:
1, prereaction oxide; 2, postreaction oxide (75 s in 11% CH4/Ar); 3,
postreaction oxide (180 s in 11% CH4/Ar).
Conclusions
but the most reactive oxygen species can be recovered. They
can react with CH4 to form synthesis gas. The results of catalyst
reoxidation in the redox cycle indicate that the oxide has a high
reactivity toward the oxygen and the oxygen can be consumed
completely. It should be noted that no CO2 and CO were
observed, which could result from the oxidation of carbonaceous
deposits on the catalyst surface by O2/Ar. This indicates that
methane pyrolysis does not occur at the appropriate reaction
condition. The ratio of H2/CO is about 1.93 and near the
stoichiometric ratio of the partial oxidation of methane.
Figure 8 shows the variations of CH4 conversion and selec-
tivity for CO as a function of redox cycles. The LaFeO3 oxide
exhibits better catalytic activity for the production of synthesis
gas. The CH4 conversions depend on the cycle number and
remain at high level of 60-70%, but decrease slightly with the
increase of cycle number, while the selectivities to CO with
the increase of the cycle number and remain at higher level
(>96%). This indicates that oxygen species for the CO forma-
tion can be recovered, while the reactive oxygen species for
the CO2 formation is hardly restored at the appropriate condition.
Surface Oxygen Species of LaFeO3 Oxides. It is well-
known that the surface oxygen species play an important role
in the catalytic oxidation reaction. XPS measurements were
performed on pre- and postreaction LaFeO3 oxides, as shown
in Figure 10.
A novel sequential redox cyclic reaction for methane partial
oxidation to synthesis gas is reported using perovskite-type
oxides as oxygen storage free of gaseous oxygen. Based on the
redox cyclic reaction, AFeO3 oxides (A ) La, Nd, Eu) were
prepared via the sol-gel method and identified by their XRD
patterns. All AFeO3 (A ) La, Nd, Eu) oxides, calcined at 1173
K, are single-phase perovskites. The Fe-O bond is affected by
A
3+ ionic radii, which could be related to the catalytic activity
of partial oxidation of methane. From the CH4-TPSR/MS
experiments, the AFeO3 (A ) La, Nd, Eu) oxides provide mostly
oxygen species, which can oxidize CH4 to synthesis gas with
high selectivity.
The investigation of the continuous reaction of CH4 oxidation
at 1173 K in the absence of gas-phase oxygen indicates that
methane was reacted first with the very reactive oxygen species
to form CO2, and then the less reactive oxygen species reacted
selectively with methane to form CO and H2. AFeO3 (A ) La,
Nd, Eu) oxides are the only oxygen source in the oxidation of
CH4 to synthesis gas. The LaFeO3 sample has the highest
selectivity to synthesis gas, and it is the best catalyst among
the tested AFeO3 oxides for synthesis gas production.
The pulsed experiments at different temperatures imply that
the rate of oxygen supply during the CH4 reaction with LaFeO3
oxide was strongly influenced by the reaction temperature. Two
types of oxygen species are present on the LaFeO3 oxides, and
they are responsible for CO2 formation and synthesis gas
formation. The rate of oxygen migrating from the bulk to the
surface increases with an increase in temperature, and more CH4
The oxides were pretreated by switching to CH4/Ar for 75
and 180 s at 1173 K and then changing to He for cooling to
room temperature. The XPS analysis of the O(1s) core level
spectra shows that it is composed of two peaks, indicating two

Direct Partial Oxidation of CH4 to Synthesis Gas
J. Phys. Chem. B, Vol. 110, No. 45, 2006 22531
(9) Otsuka, K.; Wang, Y.; Sunada, E.; Yananaka, I. J. Catal. 1998,
175, 152.
is selectively oxidized to CO and H2. The synthesis gas
production prefers to occur over the LaFeO3 oxide at high
temperature.
(10) Fornasiero, P.; Balducci, G.; Di Monte, R.; Ka<crn>spar, J.; Sergo,
V.; Gubitosa, G.; Ferrero, A.; Graziani, M. J. Catal. 1996, 164, 173.
(11) Otsuka, K.; Wang, Y.; Nakamura, M. Appl. Catal., A: Gen. 1999,
183, 317.
(12) O’Connell, M.; Norman, A. K.; Hu¨ttermann, C. F.; Morris, M. A.
Catal. Today 1999, 47, 123.
(13) Porta, P.; Cimino, S.; Rossi, S. D.; Faticanti, M.; Minelli, G.; Pettiti,
I. Mater. Chem. Phys. 2001, 71, 165.
(14) Burggraaf, A. J.; Cot, L. Fundamentals of Inorganic Membrane
Science and Technology; Elsevier: Amsterdam, 1996; p 435.
(15) Ciambelli, P.; Cimino, S.; Rossi, S. D.; Lisi, L.; Minelli, G.; Porta,
P.; Russo, G. Appl. Catal., B 2001, 29, 239.
(16) Niu, X. S.; Du, W. M.; Du, W. P. Sens. Actuators, B 2004, 99,
399.
(17) Voorhoeve, R. J. H.; Burton, J. J.; Garton, R. L. AdVanced Materials
in Catalysis; Academic Press: New York, 1977; p 129.
(18) Ramos, T.; Atkinson, A. Solid State Ionics 2004, 170, 275.
(19) Tejuca, L. G.; Fierro, J. L. G.; Tasc´on, J. M. D. AdV. Catal. 1989,
36, 237.
(20) Machida, M.; Eguchi, K.; Arai, H. J. Catal. 1990, 123, 477.
(21) Ciambelli, P.; Cimino, S.; Lasorella, G.; Lisi, L.; Rossi, S. D.;
Faticanti, M.; Minelli, G.; Porta, P. Appl. Catal., B 2002, 37, 231.
(22) Seiyama, T. Catal. ReV.sSci. Eng. 1992, 34, 281.
(23) Voorhoeve, R. J. H.; Remeika, J. P.; Trimble, L. E. Ann. N. Y.
Acad. Sci. 1976, 272, 3.
(24) Contractor, R. M.; Garnett, D. I.; Horowitz, H. S.; Bergna, H. E.;
Patience, G. S.; Schwartz, J. T.; Sisler, G. M. In New DeVelopment in
SelectiVe Oxidation II; Corberan, C., Bellon, S. V., Eds.; Elsevier:
Amsterdam, 1994; p 233.
Methane can be converted selectively to synthesis gas, which
is suitable for Fischer-Tropsch synthesis, under the appropriate
reaction conditions by redox cycles. The property of selective
oxidation of CH4 to synthesis gas can be recovered by
reoxidation using gaseous molecular oxygen. The LaFeO3 oxide
maintained relatively high catalytic activity and structural
stability in redox atmospheres. The results clearly indicate that
the partial oxidation of methane over LaFeO3 proceeds via a
suprafacial process and intrafacial reaction mechanism as
Voorhoeve et al. proposed.
Acknowledgment. Financial support from the Chinese
Natural Science Foundation (Project Nos. 20306016 and
20322201) and the National Basic Research Program of China
(2004CB217808, 2004CB719500) is gratefully acknowledged.
References and Notes
(1) Haggin, J. Chem. Eng. News 1990, 23, 27.
(2) Vanderstraeten, B.; Tuerlinckx, D.; Berghmans, J.; Vligen, S.; Van’t
Oost, E.; Smit, B. J. Hazard. Mater. 1997, 56, 237.
(3) Pantu, P.; Kim, K.; Gavalas, G. R. Appl. Catal., A 2000, 193, 203.
(4) Zeng, Y.; Tamhankar, S.; Ramprasad, N.; Fitch, F.; Acharya, D.;
Wolf, R. Chem. Eng. Sci. 2003, 58, 577.
(5) Van Looij, F.; Van Giezen, J. C.; Stobbe, E. R. Catal. Today 1994,
21, 495.
(6) Stobbe, E. R.; de Boer, B. A.; Geus, J. W. Catal. Today 1999, 47,
161.
(25) Li, R. J.; Shen, S. K. J. Mol. Catal. (China) 2001, 15, 181.
(26) Tavares, A. C.; da Silva Pereira, M. I.; Mendonc¸a, M. H.; Nunes,
M. R.; Costa, F. M.; Sa´, C. M. J. Electroanal. Chem. 1998, 449, 91.
(27) Haenen, J.; Visscher, W.; Barendrecht, E. J. Electroanal. Chem.
1986, 208, 297.
(7) Li, R. J.; Yu, C. C.; Dai, X. P.; Shen, S. K. Chin. J. Catal. 2002,
23, 381.
(8) Trovarelli, A.; de Leitenburg, C.; Dolcetti, G. CHEMTECH 1997,
27, 32.
(28) Alifanti, M.; Kirchnerova, J.; Delmon, B. Appl. Catal., A 2003,
245, 231.
(29) Omata, K.; Yamazaki, O.; Tomita, K.; Fujimoto, K. J. Chem. Soc.,
Chem. Commun. 1994, 1647.