Partial oxidation of methane to synthesis gas over alkaline earth metal oxide supported cobalt catalysts

Article abstract of DOI:10.1006/jcat.2001.3190

A study on the partial oxidation of methane to synthesis gas over the alkaline earth metal oxide supported Co catalysts showed that only MgO was a suitable support. The reaction behavior of the MgO-supported Co catalysts was considerably influenced by the calcination temperature (T_c) and Co loading. The 24 wt % Co/MgO catalyst precalcined at 800°C provided a high and stable activity. Those precalcined at higher temperatures provided a lower stable activity and those precalcined at lower temperatures a lower stability. Due to the oxidative states of Co, the structural properties of the Co catalysts were strongly affected by the nature of the support, Co loading, and calcination temperature. In the 24 wt % MgO-supported catalyst, only a solid of CoO and MgO was identified for T_c ≥ 800°C, while one or two of Co₃O₄ and MgCo₂O₄, were additionally detected for T_c < 800°C. Correlations among the species present in the catalyst, its reducibility, and its catalytic performance were established.

Full text of DOI:10.1006/jcat.2001.3190

Journal of Catalysis 199, 309–317 (2001)
doi:10.1006/jcat.2001.3190, available online at http://www.idealibrary.com on
Partial Oxidation of Methane to Synthesis Gas over Alkaline Earth Metal
Oxide Supported Cobalt Catalysts
1
H. Y. Wang and E. Ruckenstein
Department of Chemical Engineering, State University of New York at Buffalo, Amherst, New York 14260
Received September 15, 2000; accepted January 29, 2001; published online March 27, 2001
(
23), partial oxidation (12, 15), and their combination (24).
The partial oxidation of methane to synthesis gas was investi- Like NiO, CoO can also form an “ideal” solid solution over
gated over the alkaline earth metal oxide supported Co catalysts. the entire mole fraction range with MgO (25, 26). However,
unlike Ni, Co has two oxidative states (Co2 and Co ).
Consequently, additional species can be generated between
the cobalt and magnesium oxides during the preparation of
the catalyst, which are expected to depend on the prepara-
tion conditions, such as calcination temperature and metal
loading. The species present in a catalyst affect its reducibil-
ity and the size and morphology of the metal particles thus
generated and hence the catalytic pattern of the catalyst. Of
+
3+
Only MgO has proved to be a suitable support. The reaction behav-
ior of the MgO-supported Co catalysts was significantly influenced
by the calcination temperature (Tc) and Co loading. The 24 wt%
�
Co/MgO catalyst precalcined at 800 C provided a high and stable
activity; those precalcined at higher temperatures provided a lower
stable activity and those precalcined at lower temperatures a lower
stability. Because of the two oxidative states of Co, the structural
characteristics of the Co catalysts are strongly affected by the na-
ture of the support, calcination temperature, and Co loading. In the
2
4 wt% MgO-supported catalysts, only a solid solution of CoO and course, during reaction other species may form as well in
�
MgO was identified for Tc � 800 C, while one or two of the follow- the catalyst. It is therefore of interest to identify the species
ing compounds, Co3O4 and MgCo2O4, were additionally detected formed in the catalyst and their role in the catalytic process.
�
for Tc < 800 C. Correlations among the species present in the cata-
For this purpose, the reaction behavior of the Co/MgO cata-
lysts in the partial oxidation of methane was investigated
as a function of calcination temperature and Co loading.
For comparison purposes, the Co catalysts supported on
the other alkaline earth metal oxides were also included in
the investigation.
lyst, its reducibility, and its catalytic performance were established.
c 2001 Academic Press
�
Key Words: partial oxidation of methane; synthesis gas; Co cata-
lyst; effect of support; effect of calcination temperature; effect of
Co loading.
1
. INTRODUCTION
2. EXPERIMENTAL
The partial oxidation of methane, which was first inves- 2.1. Catalyst Preparation
tigated by Prettre et al. as early as 1946 (1), has gained re-
The catalysts were prepared by impregnating the sup-
newed interest during the past years (2–22). The catalysts
that have been reported to be active and selective toward
the synthesis gas formation were mainly supported noble
metals, such as Rh, Ru, Pd, Pt (2–8), supported Ni (9–17),
and Co (16–18) catalysts, as well as some pyrochlore type
oxides, such as Ln2Ru2O7 (19, 20), and perovskite type ox-
ides, such as LaMO3 (M = Ni, Rh, Co, Cr) (21, 22).
ports (MgO, CaO, SrO, and BaO) with aqueous solutions
�
of Co(NO3)2 � 6H2O, followed by overnight drying at 110 C.
The samples were then calcined in the open air of a fur-
�
nace for 8 h at 800 C, when CaO, SrO, and BaO were used
as supports, and at various temperatures between 500 and
�
1
000 C, when MgO was employed. The calcined catalysts
are denoted as Co(O)/MO (M = Mg, Ca, Sr, or Ba) and the
catalysts reduced in H2 as Co/MO. Co loading means wt%
Co in the completely reduced catalyst.
Compared to the noble-metal based catalysts, the Ni- and
Co-based catalysts are particularly promising due to their
low prices. In recent papers, Ni/MgO catalysts, obtained
through the reduction of NiO–MgO solid solutions, were
reported to provide high and stable activities for the
methane conversion to synthesis gas via CO2 reforming
2
.2. Catalytic Reaction
All the catalysts were tested under atmospheric pressure
�
and at a Tfurnace of 850 C in a fixed-bed vertical quartz re-
actor (i.d. 3 mm), which was operated in a downflow mode
with the catalyst held on a quartz wool bed. The reduction
1
To whom correspondence should be addressed. Fax: +1-716-645-3822.
E-mail: feaeliru@acsu.buffalo.edu.
309
0021-9517/01 $35.00
Copyright �c 2001 by Academic Press
All rights of reproduction in any form reserved.

3
10
WANG AND RUCKENSTEIN
�
of the catalyst was carried out by increasing the tempera- gas (He, 35 ml/min) at 850 C for 0.25 h and then pulses of
�
�
ture from room temperature to 600 C at a rate of 20 C/min CH4 (250 �l) were injected in the carrier gas. The products
and from 600 to 850 C at a rate of 10 C/min, in a H2 flow were analyzed with an on-line gas chromatograph equipped
20 ml/min), without holding at 850 C. After reduction, the with a TCD and a Porapak Q column.
�
�
�
(
feed gases (CH4/O2 = 2/1) were introduced into the catalyst
bed at a flow rate of 60 ml/min. The reactants and products
were analyzed with an on-line gas chromatograph equipped
with a Porapak Q column. An ice-cold trap was located be-
tween the reactor exit and the GC sampling valve to remove
the water formed during reaction.
2
.3.3. Temperature-programmed
During each TPR run, a high-purity 2.5% H2/Ar mixture
35 ml/min) was passed over a calcined sample in a vertical
quartz tube reactor (i.d. 4 mm). The sample (10.0 or
reduction
(TPR).
(
2
0.0 mg) was held on a quartz wool bed and the temper-
ature of the sample was increased from 40 to 1000 C at
a rate of 20 C/min. The water, formed during reduction,
was trapped in a Hydro-Purge II column. The hydrogen
�
�
2
.3. Catalyst Characterization
2
.3.1. BET surface area and the exposed Co surface area. consumed in TPR was monitored continuously with a
The BET surface areas of the calcined catalysts, determined TCD, which was calibrated using known amounts of CoO
via nitrogen adsorption using a Micromeritics ASAP2000 and Co3O4.
instrument, are listed in Table 1.
2
.3.4. X-ray powder diffraction (XRD). XRD, per-
The exposed Co metal surface area of the reduced cata-
lyst was determined by CO pulse adsorption in a quartz
tube (i.d. 4 mm) at room temperature, assuming a 1/1 stoi-
chiometry. Fifty milligrams of the calcined catalyst powder
was charged and its reduction was carried out as described
in the previous section. After reduction, the catalyst was
purged with an ultra high purity carrier gas (He, 35 ml/min)
formed with a Siemens D500 X-ray diffractometer using fil-
tered CuK� radiation, was used to identify the major phases
of the calcined catalysts.
3. RESULTS
�
at 850 C for 0.25 h, and then the sample was cooled to room
temperature. CO pulses (10 �l per pulse) were then injected
3
.1. Partial Oxidation of Methane to Synthesis Gas
in the carrier gas and passed over the reduced catalyst. The
3.1.1. Comparison of the catalysts supported on different
CO uptake in each pulse was monitored using a thermal alkaline earth metal oxides. The time-dependent conver-
conductivity detector (TCD).
sions of CH and O and selectivities to CO and H of the 24
4
2
2
wt% Co catalysts supported on MgO, CaO, SrO, and BaO
precalcined at 800 C are plotted in Fig. 1. Over the Co/MgO,
2
.3.2. CH4 decomposition. The decomposition of pure
�
methane over the reduced catalyst was carried out in a
quartz tube (i.d. 4 mm) reactor containing 25.0 mg of cal-
cined catalyst. After reduction, which was carried out as
described previously, the catalyst was purged with a carrier
after an activation period of about 2 h, the conversion of
CH4 and selectivities to CO and H2 remained high and un-
changed during the period of study of 120 h and a H2/CO
ratio of about 2 was obtained. Over CaO-, SrO-, and BaO-
supported catalysts the initial (after 8 min) conversions of
CH4 and selectivities to CO and H2 were much lower than
those over MgO and, in addition, notable deactivation was
observed.
TABLE 1
Physical Data for the Supported Co Catalysts
Surface areaa
(m /g cat.)
b
3.1.2. Effect of calcination temperature on the MgO-
Co loading Calcination
Co surface area
�
2
2
Support
(wt% )
temp. ( C)
(m /g cat.) � 100 supported Co catalysts. The time-dependent activities of
the MgO-supported catalysts precalcined at temperatures
MgO
MgO
MgO
MgO
MgO
MgO
MgO
MgO
MgO
CaO
SrO
6
12
24
24
24
24
24
36
48
24
24
24
800
800
500
700
800
900
1000
800
800
800
800
800
26
22
58
28
20
11
5
14
12
5
7
7
201
52
9
�
between 500 and 1000 C are plotted in Fig. 2. Only the con-
version of CH4 is presented because that of O2 was always
100% , and only the selectivity to CO is plotted because that
to H followed the same pattern. The results can be sum-
2
5
marized as follows: (i) the catalysts precalcined at temper-
—c
8
�
atures � 800 C exhibited stable activities and selectivities,
which decreased with increasing calcination temperature;
(ii) some deactivation with time on stream occurred over
25
79
8
�
4
1
the catalysts precalcined at 500 and 700 C; (iii) the initial
BaO
—c
conversion of methane and CO selectivity decreased with
�
a
increasing calcination temperature from 500 to 1000 C; (iv)
For the calcined catalysts.
For the reduced catalysts.
Too small to be detected.
b
c
activation periods were observed, particularly for the cata-
�
lysts calcined at temperatures � 800 C.

PARTIAL OXIDATION OF METHANE TO SYNTHESIS GAS
311
FIG. 1. Time-dependent conversions of CH4 (a) and O2 (b) and selectivities to CO (c) and H2 (d) over the alkaline metal oxide-supported Co
�
�
� 1 � 1
g .
catalysts (precalcined at 800 C). P = 1 atm, Tfurnace = 850 C, CH4/O2 = 2.0, space velocity = 720,000 ml h
3
.1.3. Effect of Co loading on the MgO-supported Co 3.2. Physicochemical Characterization
catalysts. The time-dependent activities of the MgO-
3
.2.1. The exposed metallic Co surface area of the reduced
supported catalysts with loadings between 6 and 48 wt%
�
catalyst. As shown in Table 1, for the 800 C calcined Co
catalysts supported on different alkaline earth metal oxides,
the exposed metal surface area after reduction decreased in
the sequence CaO � MgO � SrO � BaO. For the last sup-
port, the Co surface area was too small to be detected. For
the Co/MgO catalysts, Table 1 shows that for the same load-
ing (24 wt% ), the exposed Co surface area decreased with
increasing calcination temperature from 500 to 1000 C; for
the same calcination temperature (800 C), it was almost
the same for loadings between 6 and 36 wt% , but increased
notably with increasing Co loading from 36 to 48 wt% .
�
(
(
precalcined at 800 C) are plotted in Fig. 3, which shows:
i) over the 6 wt% catalyst, the conversion of CH4 was be-
low 20% , and no CO and H2 were formed; (ii) over those
with loadings � 12 wt% , the conversion of CH4 was above
8
0% , and CO and H2 were formed with high selectivities
(
2
>90% ); (iii) with the increase in Co loading from 12 to
4 wt% , the conversion of CH4 and the selectivity to CO
�
increased somewhat, but remained almost unchanged with
the further increase from 24 to 48 wt% .
�
3
.1.4. Effect of the amount of catalyst. As shown in
Fig. 4, over the 24 wt% Co/MgO catalyst (precalcined at
�
8
00 C), the conversion of CH4 and the selectivity to CO
3.2.2. Surface carbon (C) formed in the CH4 decomposi-
¯
increased with increasing amount of catalyst from 1.5 to tion. The decomposition of CH4 pulses over the reduced
�
5
.0 mg, but remained almost unchanged when the amount catalysts was carried out at 850 C and the amount of sur-
was increased from 5.0 to 15.0 mg.
face carbon (C) generated was calculated from the carbon
¯

3
12
WANG AND RUCKENSTEIN
�
A single peak at about 540 C was observed for Co(O)/CaO
�
(
Fig. 5b), and two peaks, one (the main) at about 505 C and
the other one as a shoulder at about 710 C for Co(O)/SrO
Fig. 5c). Two peaks, one (the main) at about 455 C and
the other one as a shoulder at about 790 C, were noted
for Co(O)/BaO (Fig. 5d). The reduction extent over the
MgO supported catalyst after TPR was estimated to be be-
low 5% , while nearly 100% reduction was reached over the
other alkaline earth metal oxide supported ones.
�
�
(
�
The TPR profiles of the 24 wt% Co(O)/MgO catalysts
�
calcined at temperatures between 500 to 1000 C are pre-
�
sented in Fig. 6. The 500 C calcined 24 wt% Co(O)/MgO
�
exhibited two reduction peaks, one at about 320 C and the
�
other at about 695 C. The reduction extent of the sample
FIG. 2. Time-dependent conversion of CH4 (a) and selectivity to CO
(
b) over the 24 wt% Co/MgO catalysts precalcined at various temper-
�
atures. P = 1 atm, Tfurnace = 850 C, CH4/O2 = 2.0, space velocity =
�
1
� 1
g .
7
20,000 ml h
�
balance. Table 2 shows that, for the 800 C calcined Co cata-
lysts supported on different alkaline earth metal oxides,
the amount of C accumulated during the same period over
¯
MgO was comparable to that over BaO, but much lower
than those over CaO and SrO. The low carbon deposition
over BaO was due to the low exposed metal surface area.
Table 2 also shows that, for the 24 wt% Co/MgO catalysts,
the amount of C formed decreased monotonically with in-
¯
�
creasing calcination temperature from 500 to 1000 C.
3
.2.3. Reduction characteristics of the calcined supported
�
Co catalysts. The TPR spectra of the 800 C calcined
FIG. 3. Time-dependent conversion of CH4 (a) and selectivity to CO
2
4 wt% Co catalysts supported on different alkaline earth
�
(
b) over the Co/MgO catalysts (precalcined at 800 C) with different Co
metal oxides are plotted in Fig. 5. For Co(O)/MgO, a peak
�
loadings. P = 1 atm, Tfurnace = 850 C, CH4/O2 = 2.0, space velocity =
�
� 1 � 1
g .
started to form at a temperature higher than 800 C (Fig. 5a). 720,000 ml h

PARTIAL OXIDATION OF METHANE TO SYNTHESIS GAS
313
FIG. 4. Effect of the amount of catalyst on the CH4 conversion and
�
CO selectivity of the 24 wt% Co/MgO (precalcined at 800 C). Data ob-
�
tained after 22 h of reaction; reaction conditions: P = 1 atm, Tfurnace =
FIG. 5. TPR profiles of the 800 C calcined 24 wt% Co catalysts sup-
�
8
50 C, CH4/O2 = 2.0 (60 ml/min).
ported on MgO (a), CaO (b), SrO (c), and BaO (d).
�
was below 30% . At the calcination temperature of 700 C, TPR results. The data and the assignments of the XRD
�
in addition to a wide single peak at 430 C, a peak started to patterns for the calcined Co catalysts are listed in Table 3.
form at a temperature higher than 800 C. For calcination For the 800 C calcined 24 wt% Co(O)/MgO, the XRD
temperatures � 800 C only a single peak started to form at patterns could be attributed to (Co, Mg)O (solid solution
�
�
�
�
temperatures higher than 800 C.
of CoO and MgO) and/or MgO (Table 3). The reduction
�
For all the 800 C calcined Co(O)/MgO catalysts with Co peak that started to form at a temperature higher than
loadings between 6 and 24 wt% , a peak started to form 800 C (Fig. 5a) allowed us to conclude that a solid solu-
at a temperature higher than 800 C; an additional low- tion wasformed. Ca3Co4O9 wasidentified over Co(O)/CaO,
temperature reduction peak between 380 and 430 C was in agreement with the TPR spectrum (Fig. 5b) which pro-
�
�
�
observed over the catalysts with Co loadings of 36 and vided a single reduction peak. Only SrCoO2.52 was detected
4
8 wt% .
over Co(O)/SrO, while two reduction peaks were present in
the TPR spectrum (Fig. 5c). Most likely, the amount of the
species corresponding to the TPR shoulder was too small
to be detectable by XRD. Thus, the major phase present in
the Co(O)/SrO was SrCoO2.52. Similarly, the major phase
present in the Co(O)/BaO was found to be BaCoO2.8.
The XRD results for the 500 C calcined 24 wt%
Co(O)/MgO catalyst can be assigned to Co3O4 and/or
MgCo2O4, as well as to (Co, Mg)O and/or MgO (Table 3).
3
.2.4. Compounds present in the calcined supported Co
catalysts. The Co-containing species present in the cal-
cined catalysts were identified by combining the XRD and
�
TABLE 2
Amount of Carbon Deposited during CH4 Decomposition over the
�
Reduced Supported Co (24 wt% Loading) Catalysts at 850 C
C formed during different periods (�mol)
¯
Calcination First
Support temp. ( C) pulse 3 pulses 5 pulses 10 pulses 20 pulses
First
First
First
First
�
MgO
MgO
MgO
MgO
MgO
CaO
SrO
500
700
800
900
1000
800
800
800
8.2
6.5
2.2
1.5
1.2
9.0
6.6
2.2
16.8
9.4
5.1
3.3
2.8
22.7
10.1
4.3
19.1
11.0
6.6
4.0
3.6
28.2
12.7
6.0
23.5
14.1
9.7
5.8
4.9
36.1
17.1
9.6
29.4
18.4
14.9
9.0
6.9
45.7
22.8
16.3
FIG. 6. TPR profiles of the 24 wt% Co(O)/MgO catalysts calcined at
a) 500, (b) 700, (c) 800, (d) 900, and (e) 1000 C.
BaO
�
(

3
14
WANG AND RUCKENSTEIN
TABLE 3
Data and Assignments of XRD Patterns of the Calcined Supported Co Catalysts
Co loading
(wt% )
Calcination temp.
Co-containing species
�
a
Support
MgO
( C)
d ( A˚ )
Assignments
present
24
24
500
700
2.438, 1.440, 2.869, 1.560, 2.038;
Co3O4 and/or MgCo2O4;
(Co, Mg)O and or MgO
Co3O4 and/or MgCo2O4;
(Co, Mg)O and/or MgO
(Co, Mg)O and/or MgO
(Co, Mg)O and/or MgO;
Co3O4 and/or MgCo2O4
Ca3Co4O9
Co3O4 and MgCo2O4;
(Co, Mg)O
Co3O4;
(Co, Mg)O
(Co, Mg)O
(Co, Mg)O;
Co3O4
Ca3Co4O9
SrCoO2.52
BaCoO2.8
2
.106, 1.490, 1.216,1.270, 2.438
2.436, 1.430, 2.859, 1.557, 2.020;
.108, 1.491, 1.218, 1.271, 2.436
2.110, 1.493, 1.272, 1.218, 2.438
2.111, 1.493, 1.274, 1.219, 2.438;
MgO
2
MgO
MgO
24
48
800
800
2
.438, 1.429, 2.860, 1.556, 2.022, 1.652
CaO
SrO
BaO
24
24
24
800
800
800
2.686, 2.273, 3.587, 2.406, 2.083, 1.878
2.740, 3.105, 2.051, 1.642, 1.582, 1.372
3.401, 2.806, 2.164, 1.711, 2.379, 1.454
SrCoO2.52
BaCoO2.8
a
Concluded by combining TPR and XRD results.
�
�
Since Co3O4 can be reduced in H2 below 500 C (27, 28) (Co, Mg)O was present in the 900 and 1000 C calcined
and a complete reduction of (Co, Mg)O requires a temper- catalysts.
�
�
ature higher than 1000 C (Fig. 5a), the peak at about 320 C
can be attributed to Co3O4 and the one at about 695 C trum (Fig. 7e) indicated that Co3O4 and (Co, Mg)O were
to MgCo2O4. Though only two reduction peaks were ob- present in the 48wt% Co(O)/MgO(800 C) catalyst. Among
served for this catalyst (Fig. 6a), its reduction extent below the 800 C calcined catalysts with different Co loadings, the
The XRD patterns (Table 3) coupled with the TPR spec-
�
�
�
3
0% suggested that (Co, Mg)O was also present. Its reduc- TPR spectra for loadings of 4 and 12 wt% were similar to
tion peak was not observed most likely because the upper- that for 24 wt% , and the TPR spectrum for the 36 wt% was
limit temperature during the TPR experiment was not high similar to that for the 48 wt% (Fig. 7). Consequently, one
enough. Similarly, the XRD patterns (Table 3) coupled with can conclude that Co3O4 and (Co, Mg)O were present in
�
�
the TPR spectra (Figs. 6b and 6c) indicated that Co3O4 and the 36 and 48 wt% Co(O)/MgO(800 C) catalysts, but only
�
(
Co, Mg)O were present in the 700 C calcined catalyst and (Co, Mg)O in the Co(O)/MgO(800 C) catalysts with Co
�
solely (Co, Mg)O in the 800 C calcined catalyst. Based on loadings � 24 wt% .
the similarity of the TPR spectra among the catalysts cal-
�
cined at temperatures � 800 C, one can conclude that only
4
. DISCUSSION
4
.1. Influence of Support, Calcination Temperature, and
Co Loading on the Species Present in the Co Catalysts
First, it should be noted that for the same calcination con-
ditions and the same Co loading the species formed in Co
catalysts are dependent on the nature of the support. MgO,
CaO, SrO, BaO, and CoO all have a NaCl-type structure.
However, only MgO has a lattice parameter and bond dis-
tance close to CoO (Table 4). As a result, a solid solution of
CoO and MgO was formed in the MgO-supported catalyst
TABLE 4
Crystal Data for NaCl Type Metal
Oxides (29)
Compound a ( A˚ ) A–B distance ( A˚ )
MgO
CaO
SrO
BaO
CoO
4.2112
4.8105
5.1396
5.5391
4.2603
2.11
2.40
2.57
2.77
2.13
�
FIG. 7. TPR profiles of the 800 C calcined Co(O)/MgO catalysts with
Co loadings of 6 wt% (a), 12 wt% (b), 24 wt% (c), 36 wt% (d), and
4
8 wt% (e).

PARTIAL OXIDATION OF METHANE TO SYNTHESIS GAS
315
�
after calcination at 800 C, while a compound, in which Co methane. Sintering decreases not only the number of active
n+
was present as Co (n � 3), was present as major phase in sites, but also accelerates the carbon deposition, since large
the other alkaline earth metal oxide supports (Table 3).
Second, one should emphasize that the species present
metal ensembles stimulate coke formation (30).
The size and morphology of metal clusters in the reduced
in the MgO supported catalysts are strongly affected by catalysts is strongly affected by its reducibility. The higher
the calcination temperature and Co loading. Because of the reducibility, the larger the size of metal particles is ex-
the two oxidative states of Co, other species can be gen- pected to be. When onlya solid solution ispresent, hence for
�
erated between cobalt and magnesium oxides besides the an MgO supported catalyst precalcined at Tc � 800 C, the
solid solution. Indeed, depending upon the calcination tem- size of the metal clusters is expected to be small. Parmaliana
perature and Co loading, XRD (Table 3) and TPR (Figs. 6 et al. brought transmission electron microscopy (TEM) evi-
and 7) brought evidence for the existence of one, two, or dence for a quasi-uniform distribution of small Ni particles
2+
three of the following compounds: Co3O4, MgCo2O4, and (d = 25 A˚ ) originated through the reduction of Ni ions
(
Co, Mg)O. For the 24 wt% Co catalysts, Co3O4, Co2MgO4, dispersed in the MgO matrix of a NiO–MgO solid solution
�
and (Co, Mg)O were present in the 500 C calcined cata- (31). Being small, the clusters do not favor coke formation.
lyst, Co3O4 and (Co, Mg)O in the 700 C calcined one, Indeed, Table 2 indicates that the carbon deposition during
�
and solely (Co, Mg)O in the catalysts calcined at tempera- the decomposition of pure CH4 was much lower over the
�
�
tures � 800 C. For the 800 C calcined Co(O)/MgO catalysts, Co/MgO than over the Co/CaO and Co/SrO (while the car-
Co3O4 and (Co, Mg)O were present in the 36 and 48 wt% bon deposition was almost equally low over the Co/BaO, no
catalysts, but only (Co, Mg)O in the catalysts with Co load- metallic sites were present in this catalyst). Being extracted
ings � 24 wt% .
from the substrate, the generated metallic clusters remain
partially embedded into the MgO lattice. Parmaliana et al.
brought TEM evidence that such embedding is present in
the reduced NiO–MgO solid solution systems (31). Being
partially embedded into the MgO lattice, they are more re-
sistant to sintering than the crystallites supported on the
conventional supports (32). In conclusion, the high stabil-
ity of the Co/MgO catalyst is related to the formation of a
solid solution of CoO and MgO during calcination.
4
.2. Influence of the Species Present on the Reducibility
of Co Catalysts
The reducibilityofa catalyst isdirectlyrelated to the com-
pounds formed during calcination. Due to the irreducibility
of MgO and because the oxygen belonging to Co interacts
also with Mg, it is difficult to reduce the CoO dispersed
�
in the lattice of MgO. Indeed, the 800 C calcined 24 wt%
Co(O)/MgO required the highest reduction temperature
4
.3.2. Effect of calcination temperature on the reaction
(
Fig. 5) and its extent of reduction after the TPR was be-
behavior of Co/MgO catalysts. The calcination temper-
ature has a significant influence on the stability of the
Co/MgO catalysts. The 24 wt% Co catalysts precalcined at
low 5% . For this reason, the exposed Co surface area over
the Co/MgO was much lower than that over the Co/CaO
(
Table 1). Most likely, the low exposed Co surface area over
�
temperatures � 800 C exhibited stable activities during the
the Co/SrO and Co/BaO (Table 1) was caused by the sin-
tering of the metallic crystallites and their burying induced
by the sintering of SrO and BaO.
period of study, while some deactivation was observed over
�
the catalystsprecalcined at 500and 700 C (Fig. 2). The XRD
and TPR indicated that only the solid solution was present
The reducibility of the three Co-containing species iden-
tified in the MgO-supported catalysts decreased in the se-
quence Co3O4 > Co2MgO4 > (Co, Mg)O. Consequently,
under the conditions that favor the formation of a solid so-
lution the percentage of cobalt that can be reduced to metal
is expected to decrease. Indeed, the exposed metal surface
area of the 24 wt% Co/MgO catalysts decreased monotoni-
cally as the calcination temperature was increased from 500
�
in the catalysts calcined at � 800 C and that Co3O4 and
�
MgCo2O4 were additionallypresent in the 500 C and Co3O4
�
in the 700 C calcined ones (Table 3). Being more reducible
than the solid solution, they generated large metal clusters
that stimulated carbon deposition. Indeed, as shown in Ta-
ble 2, the carbon deposition during the CH4 decomposition
�
was much greater over the 500 and 700 C calcined catalysts
�
than over those calcined at temperatures � 800 C. In addi-
�
to 1000 C (Table 1).
tion, due to the relatively weak interactions between these
clusters and support, enhanced sintering occurred. Carbon
deposition and sintering were responsible for deactivation.
4
.3. The Partial Oxidation of Methane to Synthesis Gas
.3.1. Effect of support on the reaction behavior of Co
4
On the other hand, with increasing calcination tempera-
�
catalysts. Among the four alkaline earth metal oxides in- ture from 500 to 1000 C, the initial conversion of methane
vestigated, only MgO provided a high and stable activity in partial oxidation of methane decreased (Fig. 2). This re-
during the period of study of 120 h; various extents of de- flects the trend of the exposed Co surface area (Table 1)
activation were observed over CaO, SrO, and BaO (Fig. 1). and implies that the metallic sites are the active sites for the
The carbon deposition and the metal sintering are the main synthesis gas formation. The concentration of CoO near the
causes for catalyst deactivation in the partial oxidation of surface decreased with increasing calcination temperature,

3
16
WANG AND RUCKENSTEIN
because the higher the calcination temperature the deeper rior activity. Due to its low reducibility, the metal crystallites
is the penetration of CoO into the MgO lattice. For this formed during the reduction of the solid solution were small
reason, fewer metallic sites were present on the surface of and hence little carbon deposition occurred. Originating
the catalysts precalcined at higher temperatures and the ini- from a solid solution, the crystallites were partially embed-
tial activity decreased with increasing calcination temper- ded into the substrate and thus were resistant to sintering.
ature. In addition, an activation period can be noted over The reaction behavior of Co/MgO catalysts was strongly
�
the catalysts precalcined at temperatures � 800 C (Fig. 2a), affected by the calcination temperature and Co loading.
indicating that additional metallic sites were generated dur- Deactivation was noted for the low calcination tempera-
�
ing reaction via the reduction of the solid solution by CH4 tures of 500 and 700 C. Besides the solid solution, Co3O4
�
and/or H2. Consequently, the solid solution catalyst acts like and MgCo2O4 were identified in the 500 C calcined cata-
a “reservoir” which can provide additional active metallic lyst and Co3O4 in the 700 C calcined one, compounds which
�
sites during reaction.
can be more easily reduced than the solid solution. The rel-
atively large metal particles, formed during the reduction
of the more reducible Co-containing species (Co3O4 and
MgCo2O4), stimulated carbon deposition. For calcination
4
.3.3. Effect of Co loading on the reaction behavior of
Co/MgO catalysts. The reaction behavior of Co/MgO cata-
lysts is affected by Co loading. As shown in Fig. 3, over the
�
temperatures � 800 C, the activity decreased with increas-
6
wt% catalyst, the conversion of CH4 was below 20% and
ing calcination temperature. Since only the solid solution
onlytotaloxidation ofCH4 occurred. However, the exposed
Co surface area of this catalyst after reduction was not neg-
ligible (Table 1). This suggests that the initially available
metallic sites were oxidized. When the Co loading was in-
creased to 12 wt% , partial oxidation took place. With the
increase in Co loading from 12 to 24 wt% , the conversion of
CH4 and the selectivity to CO increased somewhat, but re-
mained almost the same with a further increase from 24 to
�
was present in the catalysts calcined at � 800 C, the higher
the calcination temperature the deeper was the penetration
of CoO into the MgO lattice. For this reason, fewer metallic
sites were present on the surface of the catalysts precalcined
at higher temperatures.
REFERENCES
4
8 wt% . The almost equal methane conversions might have
1. Prettre, M., Eichner, C., and Perrin, M., Trans. Faraday Soc. 43, 335
occurred either because the reaction was diffusion limited,
or because the number of metallic sites (including those
generated during reaction) became sufficiently large for
thermodynamic equilibrium to be achieved. The results ob-
tained regarding the effect of the amount of catalyst provide
useful information in this direction. Indeed, Fig. 4 shows
that, for the 1.5 mg of catalyst, the methane conversion was
about 86% , with a CO selectivity of about 96% . As the
amount of catalyst was increased, the CH4 conversion and
CO selectivity increased, but remained almost unchanged
when the amount of catalyst became larger than 5.0 mg.
Consequently, the reactions occurred essentially within a
thin layer of about 5.0 mg (or even less) at the beginning of
the catalyst bed and equilibrium was reached.
(
1946).
2
3
. Hickman, D. A., and Schmidt, L. D., J. Catal. 138, 267 (1992).
. Bhattacharaya, A. K., Breach, J. A., Chand, S., et al., Appl. Catal. A:
Gen. 80, L1 (1992).
4. Hickman, D. A., and Schmidt, L. D., Science 259, 343 (1993).
. Torniainen, P. M., Chu, X., and Schmidt, L. D., J. Catal. 146, 1 (1994).
. Nakagawa, K., Ikenaga, N., Suzuki, T., et al., Appl. Catal. A: Gen. 169,
5
6
2
81 (1998).
7
. Fathi, M., Hofstad, K. H., Sperle, T., et al., Catal. Today 42, 205 (1998).
8. Ruckenstein, E., and Wang, H. Y., J. Catal. 187, 151 (1999).
9. Dissanayake, D., Rosynek, M. P., Kharas, K. C. C., and Lunsford,
L. H., J. Catal. 132, 117 (1991).
0. Vermeiren, W. J. M., Blomsma, E., and Jacobs, P. A., Catal. Today 13,
4
1. Choudhary, V. R., Rajput, A. M., and Prabhakar, B., J. Catal. 139, 326
(1993).
2. Tang, S., Lin, J., and Tan, K. L., Catal. Lett. 51, 169 (1998).
3. Tsipouriari, V. A., Zhang, Z., and Verykios, X. E., J. Catal. 179, 283
1
1
27 (1992).
1
1
5
. SUMMARY
(
1998).
4. Provendier, H., Petit, C., Estourn e` s, C., et al., Appl. Catal. A: Gen. 180,
63 (1999).
1
The structural characteristics of the Co catalysts are
1
strongly affected by the nature of the support, calcination 15. Ruckenstein, E., and Hu, Y. H., Appl. Catal. A: Gen. 183, 85 (1999).
1
1
1
6. Slagtern, A˚ ., Swaan, H., Olsbye, U., et al., Catal. Today 46, 107 (1998).
7. Santos, A., Men e´ ndez, M., Monz o´ n, A., et al., J. Catal. 158, 83 (1996).
8. Sokolovskii, V. D., Coville, N. J., Parmaliana, A., et al., Catal. Today,
temperature, and Co loading, and so are their reducibility
and catalytic performance. Among the alkaline earth metal
oxides (MgO, CaO, SrO, and BaO)-supported Co catalysts
42, 191 (1998).
�
(
precalcined at 800 C), only Co/MgO has proved to be a
1
2
9. Ashcroft, A. T., Cheetham, A. K., Foord, J. S., et al., Nature 344, 319
(1990).
0. Jones, R. H., Ashcroft, A. T., Waller, D., et al., Catal. Lett. 8, 169 (1991).
highly efficient and stable catalyst. It provided a CH4 con-
version of 91% and a CO selectivity of 98% at the high
�
1
� 1
21. Slagtern, A., and Olsbye, U., Appl. Catal. A: Gen. 110, 99 (1994).
space velocity of 720,000 ml g
h , without any deactiva-
2
2. Lago, R., Bini, G., Pena, M. A., and Fierro, J. L. G., J. Catal. 167, 198
1997).
3. Ruckenstein, E., and Hu, Y. H., Appl. Catal. A: Gen. 133, 149 (1995).
tion for the duration of the experiment (120 h). A solid so-
(
�
lution of CoO and MgO was identified in the 800 C calcined
2
MgO-supported catalyst, which is responsible for its supe- 24. Ruckenstein, E., and Hu, Y. H., Ind. Eng. Chem. Res. 37, 1744 (1998).

PARTIAL OXIDATION OF METHANE TO SYNTHESIS GAS
317
2
5. Hagan, A. P., Lofthouse, M. G., Stone, F. S., and Trevethan, 29. N a´ ray-Szab o´ , I., “Inorganic Crystal Chemistry.” Akad e´ miai Kiad o´ ,
M. A., in “Preparation of Catalyst II” (B. Delmon, P. Grange, P. Budapest, 1969.
A. Jacobs, and G. Poncelet, Eds.), p. 417. Elsevier, Amsterdam, 30. Rostrup-Nielsen, J. R., in “Catalysis Science and Technology” (J. R.
1
979.
Anderson and M. Boudart, Eds.), Vol. 5, pp. 1–118. Springer, Berlin,
1984.
2
6. Gazzoli, D., Occhiuzzi, M., Cimino, A., Cordischi, D., Minelli,
G., and Pinzari, F., J. Chem. Soc. Faraday Trans. 92, 4567 31. Parmaliana, A., Arena, F., Frusteri, F., et al., J. Catal. 141, 34 (1993).
(
1996).
32. Highfield, J. G., Bossi, A., and Stone, F. S., in “Preparation of Catalyst
III” (G. Poncelet, P. Grange, and P. A. Jacobs, Eds.), p. 181. Elsevier,
Amsterdam, 1983.
2
2
7. Chung, K. S., and Massoth, F. E., J. Catal. 64, 320 (1980).
8. Arnoldy, P., and Moulijn, J. A., J. Catal. 93, 38 (1985).