Methane coupling reaction in an oxy-steam stream through an OH radical pathway by using supported alkali metal catalysts

Article abstract of DOI:10.1002/cctc.201400018

A universal reaction mechanism involved in the oxidative coupling of methane (OCM) is demonstrated under oxy-steam conditions using alkali-metal-based catalysts. Rigorous kinetic measurements indicated a reaction mechanism that is consistent with OH radical formation from a H ₂O-O₂ reaction followed by C-H activation in CH ₄ with an OH radical. Thus, the presence of water enhances both the CH₄ conversion rate and the C₂ selectivity. This OH radical pathway that is selective for the OCM was observed for the catalyst without Mn, which suggests clearly that Mn is not the essential component in a selective OCM catalyst. The experiments with different catalyst compositions revealed that the OH^.-mediated pathway proceeded in the presence of catalysts with different alkali metals (Na, K) and different oxo anions (W, Mo). This difference in catalytic activity for OH radical generation accounts for the different OCM selectivities. As a result, a high C₂ yield is achievable by using Na₂WO₄/SiO₂, which catalyzes the OH^.-mediated pathway selectively. Make it methane: A universal reaction mechanism involved in the oxidative coupling of methane is demonstrated under oxy-stream conditions by using alkali-metal-based catalysts. Rigorous kinetic measurements indicated a reaction mechanism that is consistent with OH radical formation from an H₂O-O₂ reaction, followed by C-H activation in CH₄ with an OH radical.

Full text of DOI:10.1002/cctc.201400018

CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201400018
Methane Coupling Reaction in an Oxy-Steam Stream
through an OH Radical Pathway by using Supported Alkali
Metal Catalysts
Yin Liang,[a] Zhikao Li,[a] Mohamed Nourdine,[b] Salman Shahid,[c] and Kazuhiro Takanabe*[a]
A universal reaction mechanism involved in the oxidative cou-
pling of methane (OCM) is demonstrated under oxy-steam
conditions using alkali-metal-based catalysts. Rigorous kinetic
measurements indicated a reaction mechanism that is consis-
tent with OH radical formation from a H2O–O2 reaction fol-
lowed by CꢀH activation in CH4 with an OH radical. Thus, the
presence of water enhances both the CH4 conversion rate and
the C2 selectivity. This OH radical pathway that is selective for
the OCM was observed for the catalyst without Mn, which sug-
gests clearly that Mn is not the essential component in a selec-
tive OCM catalyst. The experiments with different catalyst com-
C
positions revealed that the OH -mediated pathway proceeded
in the presence of catalysts with different alkali metals (Na, K)
and different oxo anions (W, Mo). This difference in catalytic ac-
tivity for OH radical generation accounts for the different OCM
selectivities. As a result, a high C2 yield is achievable by using
C
Na2WO4/SiO2, which catalyzes the OH -mediated pathway
selectively.
Introduction
The oxidative coupling of methane (OCM), the direct synthesis
of C2 hydrocarbons from methane by using O2 and oxide cata-
lysts at high temperatures, has been investigated for many
years.[1] The simplified reaction pathway for the OCM that is
commonly accepted in the literature is shown in Scheme 1.
The sequential pathways to grow carbon chains account for
the attainable yield of OCM processes because the products
combust more rapidly than the CH4 reactant.
Until recently, the reaction was generally thought to be initi-
ated by the formation of methyl radicals on the catalyst surfa-
ces that then undergo gas-phase radical reactions (heteroge-
neous–homogeneous pathways).[2] The literature describes cat-
alysts based on MgO or rare earth oxides for the OCM.[3] In ad-
dition, Mn alkali catalysts have been investigated.[4] Rigorous
kinetic studies that use the state-of-the-art Mn/Na2WO4/SiO2
catalyst for the OCM revealed that the addition of water to the
reactant mixture can improve both the rate and selectivity.[5]
Rather than CH4 activation by the surface oxygen species (an
O*-mediated pathway), the OCM reaction was found to largely
proceed through CH4 activation by gas-phase OH radicals that
C
are formed catalytically from the H2O–O2 reaction (a OH -medi-
ated pathway).[5a]
This study addresses the site requirements for catalytic OH
radical generation in an oxy-steam stream at high tempera-
tures during an OCM reaction. Many OCM-active Mn-contain-
ing catalysts have been reported,[4] but whether these reduci-
ble Mn-originated sites are essential for water activation re-
mains uncertain. We demonstrate that Na2WO4/SiO2 (without
Mn species) yields a higher selectivity than its Mn-containing
counterparts because the catalyst is not active for the less se-
lective O*-mediated pathway, which allows the OH radical
pathway to predominate. These catalysts (without Mn) have
been investigated previously,[6] but the effects of H2O were not
considered. The melting points of Na2WO4 (971 K) and
Na2MoO4 (960 K) are lower than the typical OCM operating
temperatures (>1000 K), which suggests that under OCM con-
ditions, the catalyst surface is covered with these molten salts
that can, for example, facilitate a phase transfer of the SiO2
support from tridymite to cristobalite.[4d] Variation of the cata-
Scheme 1. The simplified reaction pathway for the OCM at high conversion
(dotted arrows: first order with respect to hydrocarbon concentration, solid
arrows: second order with respect to hydrocarbon concentration).
[a] Y. Liang, Z. Li, Prof. K. Takanabe
Division of Physical Sciences and Engineering
KAUST Catalysis Center (KCC)
King Abdullah University of Science and Technology (KAUST)
4700 KAUST, Thuwal, 23955-6900 (Saudi Arabia)
E-mail: kazuhiro.takanabe@kaust.edu.sa
Homepage: http://catec.kaust.edu.sa
[b] M. Nourdine
Department of Chemical Engineering
Universiti Teknologi Petronas
Tronoh, Perak (Malaysia)
[c] S. Shahid
Department of Chemical Engineering
Indian Institute of Technology Kharagpur
West Bengal 721302, Kharagpur (India)
C
lyst components demonstrates that a selective OH -mediated
pathway within an OCM reaction depends on the catalyst for-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 8
&
1
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
mulation. We propose that alkali metals, not the reducible
oxide, may be the active component for high selectivity. To the
best of our knowledge, the experimentally obtained C2+ yields
(>27%) are among the highest reported to date.
Results and Discussion
Kinetic analysis using Na2WO4/SiO2 (low conversion)
We first focus on the initial steps of CH4 activation. The reac-
tion pathway that dominates at low conversions for a selective
catalyst, such as Na2WO4/SiO2, is shown in Scheme 2. Zero
Scheme 2. The reaction pathway for the CH4/O2/H2O reaction at low
conversion.
conversion rates, determined by extrapolating the rates mea-
sured at various conversions to zero, were used to reflect the
input conditions with the given reactant pressures.
The CH4 conversion rate on a Na2WO4/SiO2 catalyst (in the
absence of H2O) is proportional to PCH P1=2 (Figure 1A and 1B),
O2
4
which is consistent with the mechanism that involves the dis-
sociative adsorption of O2 that reacts with molecularly ad-
sorbed CH4.[4e,5a] The elementary reaction steps involved in the
surface O*-mediated pathway for Na2WO4/SiO2 are consistent
with those reported for a Mn-containing catalyst, described as
R1–R6 in Table 1.
To simplify the steps, the quasi-equilibrium steps for R1 and
R2 are combined into R7. The rate for methane activation
through this mechanism can be described as Equation (1)
(Table 1), which is first order with respect to CH4 and half order
with respect to O2 (Figure 1).
Table 1. Plausible elementary steps and rate equations for the surface O*
C
pathway and gas-phase OH pathway.
Reaction
K0
O2
R1
R2
R3
R4
R5
R6
R7
R8
O +*
O *+*
2
O *
H
2
O2
G
2
K00
K
G
2O*
H
G
CH4
CH +*
CH *
H
4
O
4
CH *+O* k CH +OH*+*
C
*
ꢀ!
4
3
K
OH
O
C
Figure 1. A) CH4 partial pressure dependence and B) O2 partial pressure de-
pendence of the CH4 conversion rates obtained by extrapolating the mea-
sured rates to zero (zero conversion rates) and C) C2 selectivity (C2H6+C2H4)
as a function of the CH4 conversion (under various conditions without a H2O
co-feed with a Na2WO4/SiO2 catalyst; 0.1 g cat, 1073 K, 6–48 kPa CH4, 1–
12 kPa O2, total pressure=101 kPa balanced by He).
*
*
2OH*
O*+H O*
*+H O
H
G
H
2
K
H2
H O*
G
2
2
KO
2
O +2*
2O*
H
G
2
K0
K0
OH
OH
C
C
C
C
*+OH*+OH
H
OH*
*+OH or H O*+O*
G
H
G
2
K
OH
C
C
R9
O +2H O
4OH
G
H
2
2
C k
C
OH
C
R10
Eq. (1)
Eq. (2)[a]
Eq. (3)[a]
Eq. (4)
Eq. (5)
CH +OH
CH +H O
ꢀꢀ!
4
3
2
r0 ¼ kO KCH KO1=2PCH PO1=2
4
4
2
2
*
rC ¼ kC P2CH
2
2
3
rCO ¼ kCO PCH PbO
x
x
3
2
After the kinetically relevant CꢀH bond activation step of
CH4, the methyl radicals can either react with another methyl
radical or with an oxygen species (Scheme 2, Eqs. (2) and (3) in
Table 1). The presence of O2 is essential to activate CH4 (either
r00 ¼ kOHKOHPCH PO1=4PH1=2
2O
4
2
rCH ¼ r0 þ r00 ¼ k0PCH PO1=2 þ k00PCH P1O=4PH1=2
2O
4
4
4
2
2
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
k
Pb 2 þ4k rCH ꢀkCO Pb
ð
Þ
COx
C2
4
x
O2
O2
[a] PCH
¼
2kC
3
2
C
through O* or OH ), which leads to a nonzero selectivity for
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 8
&
2
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
COx at zero CH4 conversion. A high O2 pressure results in a low
C2 selectivity at low CH4 conversions as indicated in Figure 1C.
In a previous study,[5b] the oxygen dependence b was associat-
ed with first-order CO formation (molecular O2 in the gas
phase) and half-order CO2 formation (dissociated O2 on the
surface (O*)). We emphasize that, in any case, the catalysts
should prevent the direct formation of COx.
sion rates are apparent (Figure 2A). The C2 selectivity
(C2H6+C2H4) during the same experiments is presented in Fig-
ure 2B. The selectivity also improved with increasing H2O pres-
sures (at a given conversion). At these low conversion levels,
the selectivity exceeded 90% at zero conversion, and mostly
for C2H6 formation with other products in smaller quantities
(Figure 2C). This suggests that the direct combustion of CH4 to
COx is minimized on the catalyst. C2H4 is likely formed by the
dehydrogenation of C2H6 as it is extrapolated to zero selectivity
at zero conversion (Figure 2C). This improvement in selectivity
Next, the CH4 conversion rates were measured by using
a Na2WO4/SiO2 catalyst at constant CH4 and O2 pressures with
varied H2O pressure. The rates as a function of the residence
time at various H2O pressures are presented in Figure 2A. The
CH4 conversion rates improved drastically as the H2O pressure
increased. Although the reason for the slight decrease in rates
with increasing residence time in the presence of H2O remains
unclear, the beneficial effects of H2O pressure on CH4 conver-
C
can be explained by the accelerated CH3 formation, which is
C
followed by a second-order recombination of CH3 to C2H6
C
[Eq. (2)] relative to the first-order transformation of CH3 to COx
with positive O2 dependence [Eq. (3)]. This high sensitivity to
O2 pressure increases the rate of COx formation over that of
C2H6, which results in the de-
creased C2 selectivity at higher
O2 pressure (Figure 1C). This
high initial selectivity is also con-
sistent with the observation that
the catalyst lacks steam-reform-
ing activity toward CO and H2.[5b]
The rates at zero conversion
were obtained similarly from the
data acquired if various H2O
pressures and CH4/O2 mixtures
were co-fed. The improved rates
in the presence of H2O were as-
sociated with a positive kinetic
order for H2O pressures. As indi-
cated in the kinetic analysis (Fig-
ure 2D), the incremental rates
between the H2O-present and
H2O-absent rates were propor-
1=2
H2O
tional to P1O=4
P
, which is consis-
2
tent with a reaction mechanism
in which CH4 is activated by
quasi-equilibrated OH radicals
from
a H2O/O2 mixture. The
quasi-equilibrated nature of OH
radical formation leads to R8 in
Table 1.[5,7] Thus, the quasi-
equilibrated steps R5–R8 lead
simply to R9. If an OH radical
reacts with CH4 in the gas phase,
R10 occurs to generate a methyl
radical. Notably, R10 is a pure
homogeneous (gas-phase) reac-
tion. The rate of this reaction
can be described as Equation (4).
The rate is proportional to
1=2
P1O=4
P
, which is consistent with
H2O
2
Figure 2D. The rate expression
for CH4 activation (at least at low
conversion) can be thus de-
scribed based on two separate
CH4 conversion terms, the O*-
Figure 2. Effect of water pressure on A) the CH4 conversion rates as a function of the residence time, B) C2 selectiv-
ity (C2H6 and C2H4) as a function of the CH4 conversion, C) and D) incremental CH4 conversion rate (obtained from
the measured differences between the rates with and without H2O) as a function of P1=4PH1=2 with a Na2WO4/SiO2
O2
2 O
catalyst (0.1 g, 1073 K, 10.0 kPa CH4, 1.7 kPa O2, 0–2.3 kPa H2O, total pressure=101 kPa balanced by He).
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 8
&
3
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
C
mediated and OH -mediated pathways as shown in Equa-
C
Table 2. Rate constants for the surface O* pathway and gas-phase OH
tion (5),[5a] in which k’=kO*KCH K1=2 and k’’=kOH KOH
C
C
.
O2
4
pathway and the rates under typical conditions (0.1 g, 10 kPa CH4, 1.7 kPa
O2, 1.7 kPa H2O).[a]
The measured apparent rate constants k’ and k’’, the rates
for r’ and r’’, and their rate ratio r’’/r’ under typical oxy-steam
conditions at 10 kPa CH4, 1.7 kPa O2, and 1.7 kPa H2O are sum-
marized in Table 2. The ratio r’’/r’ provides a clear explanation
Catalyst
k’
k“
r’
r”
r“/r’
Mn/Na2WO4/SiO2
Na2WO4/SiO2
Na2MoO4/SiO2
Na2WO4/Al2O3
K2WO4/SiO2
0.040
0.011
0.016
0.027
0.039
n.d.
0.21
0.12
0.058
0.016
0.066
0.023
1.16
0.32
0.46
0.78
1.13
–
3.64
2.15
1.00
0.80
1.14
0.04
2.7
5.8
1.9
0.9
0.9
–
C
of how the OH -mediated pathway dominates the O*-mediated
C
/
pathway. The r’’/r’ ratio can be strongly correlated with KOH
K’ , which is the ratio of the H2O and CH4 adsorption coeffi-
CH4
Na2CO3/SiO2
cients. The presence of Mn in the catalyst clearly enhanced the
rate constants for the CH4 conversion with O* (without H2O) to
yield an apparent rate constant (k’) for Mn/Na2WO4/SiO2
(0.04 mmolgꢀ1 sꢀ1 kPaꢀ3/2) that exceeds that for Na2WO4/SiO2
(0.01 mmolgꢀ1 sꢀ1 kPaꢀ3/2). Thus, Mn has been considered effec-
tive previously if a CH4/O2 mixture (in the absence of H2O) was
introduced as a reactant.[4] The
[a] rCH ¼ r0 þ r00 , (surface O* pathway) r0 ¼ k0PCH P1O=2
, (gas-phase OH
C
4
4
2
pathway) r00 ¼ k00PCH PO1=4P1H=2
.
r, r’ and r“ in mmolgꢀ1 sꢀ1
,
k’ in
2 O
4
2
mmolgꢀ1 sꢀ1 kPaꢀ3/2, k” in mmolgꢀ1 sꢀ1 kPaꢀ7/4
.
apparent rate constant for the
C
OH -mediated pathway (k’’) was
found
to
be
0.12 mmolgꢀ1 sꢀ1 kPaꢀ7/4 for Mn/
Na2WO4/SiO2 compared with
0.21 mmolgꢀ1 sꢀ1 kPaꢀ7/4
for
Na2WO4/SiO2. Thus, r’’/r’ is 5.8 for
the catalyst without Mn, which
exceeds that for the catalyst
with Mn (2.7) to result in pre-
dominant CH4 activation through
C
OH -mediated pathway for
a
Na2WO4/SiO2. The C2+ selectivity
on using Na2WO4/SiO2 was nota-
bly higher than that with a Mn-
containing catalyst for all CH4
conversions, which is consistent
with that predicted by ChemKin
C
modeling in which a pure OH -
mediated pathway exhibits im-
proved C2 selectivity and yield.[5b]
The unique properties of
Na2WO4/SiO2 for the generation
of OH radicals from a H2O/O2
mixture may arise from redox
properties of the oxo anion (W).
To investigate this, the CH4 con-
version rates were measured by
using a Na2MoO4/SiO2 catalyst at
constant CH4 and O2 pressures
with varied H2O pressure, and
the results are shown in
Figure 3. As with a Na2WO4/SiO2
catalyst, the CH4 conversion
rates on using Na2MoO4/SiO2 in-
creased with increasing H2O par-
tial pressure (Figure 3A). The C2
selectivity also improved drasti-
cally with increasing H2O pres-
sures (at a given conversion; Fig-
Figure 3. Effect of water pressure on A) the CH4 conversion rates as a function of residence time, B) the C2 selectiv-
ity (C2H6 and C2H4) as a function of the CH4 conversion, and C) incremental CH4 conversion rate (obtained from
the measured differences between the rates with and without H2O) as a function of P1=4PH1=2 with a Na2MoO4/SiO2
O2
2 O
ure 3B and 3C), but this selectiv- catalyst (0.1 g, 1073 K, 10.0 kPa CH4, 1.7 kPa O2, 0–2.3 kPa H2O, total pressure=101 kPa balanced by He).
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 8
&
4
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
ity was slightly lower than that for the Na2WO4/SiO2 catalyst.
Without H2O addition (CH4/O2 mixture only), CO and CO2 are
formed significantly, and negligible C2H4 selectivity at low con-
version indicates that C2H4 is the secondary product from C2H6
dehydrogenation (Figure 3C). These incremental CH4 conver-
Table 3. Salt melting points, BET surface areas, and C2 selectivities
(C2H6+C2H4) at 2% CH4 conversion for the various catalysts (0.1 g, 1073 K,
10 kPa CH4, 1.7 kPa O2, total pressure=101 kPa balanced by He).
Catalyst
Salt m.p.
[K]
BET surface
area [m2 gꢀ1
C2 selectivity at 2%
CH4 conversion [%]
]
sion rates on using Na2MoO4/SiO2 were correlated as a function
Na2WO4/SiO2
Na2MoO4/SiO2
K2WO4/SiO2
971
960
1194
971
1.9
1.7
0.6
5.2
88.0
45.0
66.9
71.4
1=2
of P1O=4
P
(Figure 3D). The linear relationship of this plot is
H2O
2
consistent with a mechanism that involves CH4 activation
C
through quasi-equilibrated OH -mediated pathways. The r’’/r’
Na2WO4/Al2O3
C
ratio is 1.9, which indicates that a OH -mediated pathway is
predominant under oxy-steam conditions, yet below that of
Na2WO4/SiO2.
Further investigation was aimed to identify the active com-
ponents of the OH radical pathways. The K2WO4/SiO2, Na2WO4/
Al2O3, and Na2CO3/SiO2 catalysts were investigated in a similar
manner with particular attention paid to the effects of H2O.
The beneficial effects of H2O on both the CH4 conversion and
C2 selectivity were observed for all catalysts. The rate constants
and the r’’/r’ ratios are compiled in Table 2. Large k’ values and
the resultant small r’’/r’ ratios for Na2WO4/Al2O3 and K2WO4/
SiO2 indicate that the surface O*-mediated activation of CH4 is
relatively prominent on these catalysts. For improved C2 selec-
tivity, the catalyst must be inert under CH4 activation on the
surface (Al2O3 may not be because of its Lewis acidity), but suf-
ficiently sensitive to generate OH radicals (K may be less sensi-
tive than Na). Moreover, the absence of a W or Mo species (the
experiments performed with a Na2CO3 catalyst) also demon-
Figure 4. XRD patterns of the K2WO4/SiO2, Na2WO4/SiO2, and Na2MoO4/SiO2
catalysts before and after the reaction (AR).
C
strated exclusive activity for the OH -mediated pathway with
no measurable rates confirmed for the surface O*-mediated
pathway. These data indicate that W or Mo is not essential for
H2O activation with O2, but that alkali metal species (Na+, K+)
are more likely (W or Mo stabilizes those alkali species). In the
same experiment that uses quartz, no catalysis for CH4 conver-
sion was observed even in the presence of H2O (CH4 is inert in
both the presence and absence of water). The site requirement
is also related to the active oxygen species involved in the acti-
vation of CH4 and H2O, such as O* or OsꢀO* etc., which are
currently difficult to discern except that the kinetic analyses in-
dicate the involvement of dissociated O2. The alkali metal spe-
cies, which are often in the molten salt state at the reaction
temperature, generally have a high affinity for H2O (to form hy-
droxides easily). These kinetic analyses demonstrate the univer-
sality and significance of the reaction mechanism involved in
the OCM reaction that uses H2O as a co-reactant (i.e., the CꢀH
bond activation by OH radicals formed catalytically).
at high temperatures. We do not consider this phase transfer
to be critical to create a single active site for CH4 coupling (or
OH radical generation); however, it aids in the inhibition of the
combustion activity on the surface. The reaction contains H2O
as a sintering-facilitation reagent, and the typical reaction time
per sample is longer than 2 days. The catalytic performance re-
mains unchanged during the measurements.
Dual-reactor experiments with Na2WO4/SiO2 catalyst (high
conversions)
A sequential dual-reactor system (Scheme 3) was used to ach-
ieve high conversions and yields by using Na2WO4/SiO2, the
most selective catalyst investigated, with the introduction of
additional O2 between the first and second reactors. The C2+
yields (all hydrocarbons except for the CH4 reactant) measured
at various reaction temperatures and pressures are shown in
Figure 5. Some results of the CH4 conversion, carbon selectivity,
and C2 and C2+ yields under various conditions are summarized
in Table 4. In a single reactor with a CH4/O2 ratio of 3, the CH4
conversion reached ~37% with a C2 selectivity of ~64%. The
C2+ yield thus exceeds 25% already with a single reactor. Fur-
ther conversion of the CH4 was attempted with the second re-
actor at various temperatures. The C2+ yields improved only
slightly, as confirmed by the experiments under various O2
pressures and reaction temperatures in the second reactor. The
insensitivity of the O2 pressure at high CH4 conversion to the
C2 selectivity implies that all hydrocarbons have the same
oxygen dependence to form COx. Thus, the staged O2 ap-
Catalyst characterization
As described previously, Na2WO4 and Na2MoO4 are in a molten
state under the reaction conditions because their melting
points are below the reaction temperature (ꢁ1073 K; Table 3),
which leads to a facile transfer of the crystal structure to
a more rigid structure. The cristobalite phases dominate the
samples, especially with the molten salts (Na2WO4 and
Na2MoO4) and particularly after the OCM reactions (Figure 4).
With the knowledge that alkali metals enhance the phase
change, flux-assisted phase transformation is believed to occur
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 8
&
5
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
est hydrocarbon yields reported
for simple oxygen-addition ex-
periments.
Among the most unique prop-
erties of the Na2WO4/SiO2 cata-
lyst is the high affinity of its sur-
face for H2O compared with hy-
drocarbons under steady-state
conditions. As discussed in detail
previously,[5b] the absence of
strong hydrocarbon adsorption
Scheme 3. Schematic image of the sequential dual-reactor system with O2 addition.
prevents the preferential com-
bustion of a C2H4 product (with
p electrons) on the surface. Moreover, the relative rates for the
kinetically relevant CꢀH bond cleavage in each hydrocarbon
through an OH radical pathway reflects only a Brønsted–
Evans–Polanyi relationship, the CꢀH bond is stronger in C2H4
(452 kJmolꢀ1) than in CH4 (439 kJmolꢀ1), to yield a comparative-
ly lower rate of CꢀH abstraction by the OH radical for C2H4
than for CH4 (1.2ꢁ1012 or 1.6ꢁ1012 cm3 molꢀ1 sꢀ1, respectively,
at 1073 K), which results in an improved C2 yield.[5b]
This study demonstrates the improved rates and yields of
a methane coupling reaction under oxy-steam conditions. The
kinetic analysis of the experiments that use a Na2WO4/SiO2 cat-
alyst is consistent with the quasi-equilibrated OH radical forma-
tion, which preferentially abstracts hydrogen from CH4 over
C2H4. The OCM reactions at high pressures remain a challenge
because the carbon growth reactions are generally second
order with respect to the hydrocarbon or relevant radical con-
centrations (Scheme 1).[3g] The formation of higher hydrocar-
bons, the concentrations of which increase with high CH4 con-
version, leads to reduced C2 yields because of their high rates
of oxidation to COx.[3g,5b]
Figure 5. C2+ yield as a function of the CH4 conversion using a sequential
dual reactor for the CH4/O2/H2O reaction with a Na2WO4/SiO2 catalyst (first
reactor: 0.8 g, 1153 K, 10.0 kPa CH4, 0.8–3.3 kPa O2, 0–2.3 kPa H2O; second re-
actor: 0.4 g, 1073–1153 K, 0.8–2.4 kPa O2 added to the stream from the first
reactor, total pressure=101 kPa balanced by He).
Conclusions
proach (the O2 membrane reactor) is only useful at low CH4
conversions as discussed previously (Scheme 2). Under such
conditions, a sufficient C2 concentration triggers the C2 activa-
tion to form C3 (and C4), which combust more rapidly
(Scheme 1). The selectivity of C3 or higher hydrocarbons (pri-
marily C3H6) was ~3%. At 1123 K in the second reactor, the
CH4 conversion reached ~52% with a C2H4 selectivity of ~40%.
Overall, the C2+ yield was 27.6%, which is higher than that re-
ported over Mn/Na2WO4/SiO2 (~26%),[5b] and among the high-
The presence of water in a CH4/O2 mixture enhanced both the
CH4 conversion rate and the C2 selectivity for the oxidative
coupling of methane (OCM) that used alkali-metal-based cata-
lysts. The reaction mechanism is consistent with OH radical for-
mation from a H2O–O2 reaction, followed by CꢀH activation in
hydrocarbons with an OH radical. The contribution of this OH
radical pathway that is selective for OCM over the surface O*
pathway predominantly accounts for the different OCM selec-
Table 4. Catalytic results from experiments that used a Na2WO4/SiO2 catalyst by using sequential dual reactors with O2 added between the first and
second reactors (first reactor: 0.8 g, 1153 K, 10.0 kPa CH4, 3.3 kPa O2, 2.3 kPa H2O; second reactor: 0.4 g, 1073–1153 K, 2.4 kPa O2 added to the stream from
the first reactor, total pressure=101 kPa balanced by He).[a]
Conversion [%]
CH4
Selectivity [%]
C3
Yield [%]
C2H4
C2H6
CO2
CO
C2
C2+
1st reactor (only) 1153 K
36.9
53.1
51.7
44.7
48.0
37.2
39.5
43.2
15.8
9.2
10.6
13.5
4.3
2.9
2.9
3.4
13.9
23.8
21.8
18.2
17.8
26.4
24.7
21.2
23.5
24.6
25.9
25.3
25.2
26.4
27.6
27.0
2
2
nd reactor 1153 K
nd reactor 1123 K
2nd reactor 1073 K
[a] H2 was detected with a selectivity below 6% for hydrogen balance.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 8
&
6
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
C
The XRD patterns of the products before and after the OCM reac-
tion were obtained by using a Bruker DMAX 2500 X-ray diffractom-
eter using CuKa radiation (l=0.154 nm). The N2 sorption studies
were conducted by using a Micromeritics ASAP 2420 to determine
the BET surface area. The results are compiled in the Supporting In-
formation.
tivities. The OH -mediated pathway also proceeded with cata-
lysts without Mn, W, or Mo, which suggests clearly that these
components are not essential as an OH radical generator from
a H2O/O2 mixture. The universal mechanism for the selective
OCM reaction is proposed to use the H2O–O2 reaction on sup-
ported alkali metal catalysts.
Keywords: alkali metals
· kinetics · radicals · oxidative
Experimental Section
coupling · reaction mechanisms · supported catalysts
For the catalyst preparation, SiO2 (Sigma–Aldrich, Silica Gel, Davisil
Grade 646, 35–60 mesh) or Al2O3 (Evonik, Fumed Aluminum oxide
Aeroxide Alu130) was used as a support to immobilize 5 wt%
Na2WO4·2H2O (Sigma–Aldrich, 99%), Na2MoO4·2H2O (Sigma–Al-
drich, ꢁ99.5%), K2WO4 (Sigma–Aldrich, 94%), or Na2CO3 (Fluka,
ꢁ99.9999%) by wet impregnation. This sample was heated under
[1] a) J. S. Lee, S. T. Oyama, Catal. Rev. Sci. Eng. 1988, 30, 249–280; b) Z.
Zhang, X. E. Verykios, M. Baerns, Catal. Rev. Sci. Eng. 1994, 36, 507–556;
c) J. H. Lunsford, Angew. Chem. Int. Ed. Engl. 1995, 34, 970–980; Angew.
Chem. 1995, 107, 1059–1070; d) M. Yu. Sinev, Z. T. Fattakhova, V. I. Lomo-
nosov, Yu. A. Gordienko, J. Nat. Gas Chem. 2009, 18, 273–287; e) U. Za-
vyalova, M. Holena, R. Schlçgl, M. Baerns, ChemCatChem 2011, 3, 1935–
1947.
a flow of dry air at 1173 K for 8 h at a rate of 2 Kminꢀ1
.
[2] K. Takanabe, J. Jpn. Pet. Inst. 2012, 55, 1–12.
The rates and selectivities of the CH4/O2/H2O reactions were mea-
sured by using flow reactors using a U-shaped quartz cell (4 mm
I.D.). The samples (0.02 g) were held on quartz wool without dilu-
tion. The temperature was maintained by using a Honeywell con-
troller coupled to a resistively heated furnace and measured with
a K-type thermocouple set outside the catalyst bed. The CH4
(99.9995%), 20% O2 in He, and He (99.999%) were purchased from
Abdullah Hashim Industrial Gases & Equipment Co., Ltd (AHG) and
used after further purification by filtration. The flow was regulated
by mass flow controllers. A saturator with a well-controlled temper-
ature (278–293 K) was used to introduce the H2O gas.
[3] a) G. E. Keller, M. M. Bhasin, J. Catal. 1982, 73, 9–19; b) T. Ito, J.-X. Wang,
C.-H. Lin, J. H. Lunsford, J. Am. Chem. Soc. 1985, 107, 5062–5068; c) K.
Otsuka, A. A. Said, K. Jinno, T. Komatsu, Chem. Lett. 1987, 77–80; d) K.
Machida, M. Enyo, J. Chem. Soc. Chem. Commun. 1987, 1639–1640;
e) J. A. Labinger, K. C. Ott, J. Phys. Chem. 1987, 91, 2682–2684; f) K. D.
Campbell, J. H. Lunsford, J. Phys. Chem. 1988, 92, 5792–5796; g) J. A. La-
binger, Catal. Lett. 1988, 1, 371–376; h) J. M. DeBoy, R. F. Hicks, J. Chem.
Soc. Chem. Commun. 1988, 982–984; i) E. Morales, J. H. Lunsford, J. Catal.
1989, 118, 255–265; j) G. J. Hutchings, J. R. Woodhouse, M. S. Scurrell, J.
Chem. Soc. Faraday Trans. 1 1989, 85, 2507–2523; k) J. A. Roos, S. J. Korf,
R. H. J. Veehof, J. G. van Ommen, J. R. H. Ross, Appl. Catal. 1989, 52, 131–
145; l) S. C. Reyes, E. Iglesia, C. P. Kelkar, Chem. Eng. Sci. 1993, 48, 2643–
2661; m) S. C. Reyes, C. P. Kelkar, E. Iglesia, Catal. Lett. 1993, 19, 167–180;
n) C. A. Mims, R. Mauti, A. M. Dean, K. D. Rose, J. Phys. Chem. 1994, 98,
13357–13372; o) S. Lacombe, H. Zanthoff, C. Mirodatos, J. Catal. 1995,
155, 106–116; p) P. M. Couwenberg, Q. Chen, G. B. Marin, Ind. Eng. Chem.
Res. 1996, 35, 3999–4011; q) A. J. Colussi, V. T. Amorebieta, J. Catal. 1997,
169, 301–306; r) J. Coronas, J. Santamarꢂa, Catal. Today 1999, 51, 377–
389; s) Y. S. Su, J. Y. Ying, W. H. Green, J. Catal. 2003, 218, 321–333.
[4] a) C. A. Jones, J. J. Leonard, J. A. Sofranko, J. Catal. 1987, 103, 311–319;
b) X. Fang, S. Li, J. Lin, J. Gu, D. Yang, J. Mol. Catal. 1992, 6, 427; c) D.
Wang, M. P. Rosynek, J. H. Lunsford, J. Catal. 1995, 155, 390–402; d) A.
Palermo, J. P. H. Vazquez, A. F. Lee, M. S. Tikhov, R. M. Lambert, J. Catal.
1998, 177, 259–266; e) S. Pak, P. Qiu, J. H. Lunsford, J. Catal. 1998, 179,
222–230; f) S. Pak, J. H. Lunsford, Appl. Catal. A 1998, 168, 131–137; g) S.
Ji, T. Xiao, S. Li, L. Chou, B. Zhang, C. Xu, R. Hou, A. P. E. York, M. L. H.
Green, J. Catal. 2003, 220, 47–56; h) S. Hou, Y. Cao, W. Xiong, H. Liu, Y.
Kou, Ind. Eng. Chem. Res. 2006, 45, 7077–7083; i) M. R. Lee, M.-J. Park, W.
Jeon, J.-W. Choi, Y. W. Suh, D. J. Suh, Fuel Process. Technol. 2012, 96, 175–
182.
The reactant and product concentrations were measured by using
a VARIAN gas chromatograph 450GC with a programmed system
that involved a molecular sieve 5 A column and a HayeSep Q
column with a thermal conductivity detector, and a VARIAN CP-
Wax 52 CB capillary column with a flame ionization detector. This
configuration enables the differentiation of all C1–C4 hydrocarbons.
For H2 detection, a micro-gas chromatograph (Agilent Technologies
3000A) equipped with thermally conductive molecular and plot U
columns was used. Conversions, selectivities, and yields are report-
ed on a carbon basis as cumulative integral values as follows
[Eqs. (6)–(9)]:
ðtotal mols of carbon in productsÞ
Methane conv: ½%ꢂ ¼
ꢃ 100
ðtotal mols of CH4 inÞ
ð6Þ
ð7Þ
[5] a) K. Takanabe, E. Iglesia, Angew. Chem. Int. Ed. 2008, 47, 7689–7693;
Angew. Chem. 2008, 120, 7803–7807; b) K. Takanabe, E. Iglesia, J. Phys.
Chem. C 2009, 113, 10131–10145; c) V. Lomonosov, Yu. Gordienko, M.
Sinev, Top. Catal. 2013, 56, 1858–1866.
[6] A. Palermo, J. P. H. Vazquez, R. M. Lambert, Catal. Lett. 2000, 68, 191–
196.
[7] a) L. C. Anderson, M. Xu, C. E. Mooney, M. P. Rosynek, J. H. Lunsford, J.
Am. Chem. Soc. 1993, 115, 6322–6326; b) K. B. Hewett, L. C. Anderson,
M. P. Rosynek, J. H. Lunsford, J. Am. Chem. Soc. 1996, 118, 6992–6997;
c) K. B. Hewett, M. P. Rosynek, J. M. Lunsford, Catal. Lett. 1997, 45, 125–
128.
ðtotal mols of carbon in productsÞ
ðtotal mols out including CH4Þ
or
ꢃ 100
ðmols of carbon in the specific productÞ
ðtotal mols of carbon in productsÞ
ð8Þ
ð9Þ
Select: ½%ꢂ ¼
ꢃ 100
Yield ½%ꢂ ¼ Methane conversion ½%ꢂ ꢃ Selectivity ½%ꢂ=100
For rigorous kinetic analysis, linear regression was used to extrapo-
late the rates measured at various conversions to the rates at zero
conversion. The obtained rates at zero conversion reflect the input
conditions with the given reactant pressures strictly, which also
minimizes the contribution of the heat generated by the reaction
at low conversion levels.
Received: January 8, 2014
Revised: February 3, 2014
Published online on && &&, 0000
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 8
&
7
&
ÞÞ
These are not the final page numbers!

FULL PAPERS
Y. Liang, Z. Li, M. Nourdine, S. Shahid,
K. Takanabe*
Make it methane: A universal reaction
mechanism involved in the oxidative
coupling of methane is demonstrated
under oxy-stream conditions by using
alkali-metal-based catalysts. Rigorous ki-
netic measurements indicated a reaction
mechanism that is consistent with OH
radical formation from an H2O–O2 reac-
tion, followed by CꢀH activation in CH4
with an OH radical.
&& – &&
Methane Coupling Reaction in an Oxy-
Steam Stream through an OH Radical
Pathway by using Supported Alkali
Metal Catalysts
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 8
&
8
&
ÞÞ
These are not the final page numbers!