Partial oxidation of methane to H2 and CO over Rh/SiO2 and Ru/SiO2 catalysts

Article abstract of DOI:10.1016/j.jcat.2004.05.028

Partial oxidation of methane to synthesis gas over Rh/SiO₂ and Ru/SiO₂ catalysts was studied at 450°-800°C. Methane conversion selectivity to both H₂ and CO were higher over Rh/SiO₂ than over Ru/SiO₂.

Full text of DOI:10.1016/j.jcat.2004.05.028

Journal of Catalysis 226 (2004) 247–259
www.elsevier.com/locate/jcat
Partial oxidation of methane to H2 and CO over
Rh/SiO2 and Ru/SiO2 catalysts
Q.G. Yan a,∗, T.H. Wu b, W.Z. Weng b,c, H. Toghiani a, R.K. Toghiani a,
H.L. Wan b,c,∗, C.U. Pittman Jr. d,∗
a
Center for Advanced Vehicular Systems and Dave C. Swalm School of Chemical Engineering, Box 5405, Mississippi State University,
Starkville, MS 39762, USA
b
Institute of Physical Chemistry, Zhejiang Normal University, Zhejiang 321004, People’s Republic of China
c
Department of Chemistry and the State Key Laboratory for Physical Chemistry of Solid Surface, Xiamen University, Xiamen, Fujian 361005, PR China
d
Department of Chemistry, Box 9573, Mississippi State University, Starkville, MS 39762, USA
Received 25 March 2004; revised 17 May 2004; accepted 22 May 2004
Available online 2 July 2004
Abstract
Partial oxidation of methane to syngas (POM) over Rh/SiO and Ru/SiO catalysts was investigated at 450–800 C. Methane conversion
◦
2
2
and selectivity to both H and CO were higher over Rh/SiO than over Ru/SiO . The CO and H selectivities substantially decreased with
2
2
2
2
◦
increasing GHSV over Ru/SiO at 500 C, while these selectivities remained nearly constant over Rh/SiO . Both the CH conversions and
the CO and H selectivities increased slightly over Rh/SiO at 700 C as GHSV increases, while the CH conversions and the CO and H
2
2
4
◦
2
2
4
2
selectivities decreased slightly over Ru/SiO . Pulse reactions, transient reactions, and in situ microprobe Raman techniques were employed
2
to investigate the oxidation of methane over Rh/SiO and Ru/SiO catalysts. CO was the main product when Rh/SiO was exposed to
2
2
2
◦
methane pulsing at 700 C. No CO was detected during the first pulse over the Rh/SiO catalysts. However, CO was formed in every pulse
2
2
2
over Ru/SiO . Transient results showed that CO was formed prior to CO generation over Rh/SiO catalysts. CO was the primary product
2
2
2
2
◦
over Ru/SiO catalysts. In situ microprobe Raman spectroscopy at 450–600 C demonstrated that Ru/SiO surfaces contained significantly
2
2
larger amounts of metal oxide species than Rh/SiO during the POM reaction. The mechanisms of POM over the two catalysts are different.
2
A direct oxidation process mainly occurs over the Rh/SiO catalyst, while the dominant pathway over the Ru/SiO catalyst is the indirect
2
2
oxidation process.
2004 Elsevier Inc. All rights reserved.
Keywords: Partial oxidation; Methane; Syngas; Rh/SiO ; Ru/SiO ; Pulse reactions; Transient reaction; In situ microprobe Raman spectroscopy
2
2
1. Introduction
techniques are available for hydrogen production from hy-
drocarbons including steam reforming which is generally
accompanied by a water gas-shift conversion and a hydro-
gen purification process. Steam reforming [SR, Eq. (1)] has
constituted the dominant commercial process employed to
produce synthesis gas from methane over the past several
decades [5]:
The greenhouse effect and the limited reserves of fossil
fuels have encouraged studies of CO2-neutral and renewable
energy sources. One alternative energy source is hydrogen,
which can be used to generate both electricity and heat in
high efficiency fuel cells. Hydrogen is targeted as the next
generation fuel since it produces water instead of CO2. Fuel
cell technology and high-pressure hydrogen tank storage
are already in advanced development stages [1–4]. Several
CH4 + H2O → CO + 3H2,
ꢀH298 = 206 kJ/mol. (1)
However, this process requires a large energy input and
very complex processing. Therefore, to obtain hydrogen at a
lower energy input, the partial oxidation of methane [POM,
Eq. (2)] to hydrogen and carbon monoxide which was first
studied by Prettre et al. [6] has been studied intensively in
*
Corresponding authors.
E-mail addresses: yan@cavs.msstate.edu (Q.G. Yan),
hlwan@xmu.edu.cn (H.L. Wan).
0021-9517/$ – see front matter
doi:10.1016/j.jcat.2004.05.028
2004 Elsevier Inc. All rights reserved.

248
Q.G. Yan et al. / Journal of Catalysis 226 (2004) 247–259
the past decade [7–22]:
Gas-phase CH4 interacts with the catalyst surface, producing
CHx surface species. Elmasides and Verykios determined
that CO is the primary product, resulting from a surface
reaction between carbon and adsorbed atomic oxygen on
metallic Ru sites, while CO2 derives from CO oxidation on
oxidized Ru sites [31]. Mallens et al. [32,33] found differ-
ences using Rh versus Pt catalysts, in the selectivity toward
CO and H2 during POM. These differences were attributed
to the lower activation energy for methane decomposition
on Rh versus that on Pt. Mallens et al. suggested that the
catalyst’s ability to activate methane determines (1) the prod-
uct distribution and (2) the concentration of active surface
species of oxygen, carbon, and hydrogen. Fathi et al. [34]
studied the partial oxidation of methane to syngas over plat-
inum catalysts and proposed that the product distribution is
determined by both the concentrations and the types of sur-
face oxygen species present at the catalyst surface. Qin et
al. [35] suggested that the support might also influence the
concentration of adsorbed oxygen and, as a consequence,
the activation of methane and the product distribution. Li
et al. [36] studied the effect of gas-phase O2, reversibly
adsorbed oxygen, and oxidation state of the nickel in the
Ni/Al2O3 catalyst on CH4 decomposition and partial oxi-
dation using transient response techniques at 700 ◦C. They
concluded that the surface state of the catalyst affects the
reaction mechanism and plays an important role in POM
conversions and selectivities. Li et al. [36] also argued that
direct oxidation is the major POM route, and that the indi-
rect oxidation mechanism cannot become dominant under
their experimental conditions.
The reaction steps in CH4 oxidation to CO and H2 over
a Rh(1 wt%)/γ -Al2O3 catalyst were studied using in situ
DRIFTS at 973 K by Buyevskaya et al. [37,38]. The prod-
uct distribution and the resulting absorption band intensities
of the adsorbates were strongly influenced by oxygen cover-
age and carbon deposits on the surface. CH4 was dehydro-
genated to deposited carbon and H2, while simultaneously
being oxidized to CO2 and H2O. Surface OH groups in the
support were involved in the CHx conversion to CO via
reforming reactions. These authors assumed that surface car-
bon reacted with CO2 to contribute to CO formation. Weng
et al. [39,40] employed in situ TR-FTIR to investigate par-
tial methane oxidation to syngas over supported Rh and Ru
catalysts at 500 ◦C. Their TR-FTIR spectra indicated that a
significant difference in POM mechanisms occurred when
employing supported Ru versus supported Ru catalysts. CO
was the primary POM product over hydrogen-reduced and
working-state Rh catalysts according to these TR-FTIR ob-
servations. In contrast, CO2 was the primary POM product
over supported Ru catalysts. Therefore, the direct oxidation
of CH4 was proposed as the main pathway with Rh/SiO2,
while the reforming of unreacted CH4 to syngas dominated
using Ru/SiO2 and was accompanied by oxidation of a por-
tion of the CH4 to CO2 and H2O.
CH4 + 1/2O2 → CO + 2H2,
ꢀH298 = −35.7 kJ/mol.
(2)
Partial oxidation of methane (POM) to syngas is slightly
exothermic and more energy efficient than steam reforming
of CH4. A smaller reactor can be used to achieve high CH4
conversion and selectivity to H2 with short contact times
(∼ 10−3 s) [23]. Furthermore, POM is mechanically sim-
pler than the steam reforming, since it is completed in a
single reactor, without external heating. However, several
problems must be solved before POM can be developed on
the industrial scale. These include the cofeeding of CH4/O2
under explosive conditions [11], and the formation of local
hot spots within catalyst beds that can irreversibly damage
both the catalyst and the reactor [11,12]. Carbon deposition
may deactivate the catalysts [13,14]. Therefore, understand-
ing the reaction mechanism and the nature of the active cat-
alytic sites is important. Despite many past studies of the
partial oxidation of methane to syngas, engineering prob-
lems (heat and mass-transfer control) must still be solved
and catalysts must be improved [14]. Thus, a better under-
standing of mechanisms and microkinetics is required [15].
Two major POM reaction mechanisms have been pro-
posed. One involves total oxidation of a portion of the CH4
followed by reforming the unconverted CH4 with CO2 and
H2O to produce CO and H2. This was proposed by Pret-
tre et al. [6]. The other scheme, proposed by Schmidt et
al. [17,18], postulates direct partial oxidation of CH4 to CO
and H2 without the production of CO2 and H2O. Schmidt
et al. also found that the POM reaction process is mass-
transfer limited [19–22] and examined the role of boundary
layer mass transfer on the selectivity during partial oxida-
tion of CH4 oxidation to CO and H2 over Pt–Rh gauzes
and Pt-coated ceramic monoliths [19,20]. This selectivity is
strongly affected by the gas-flow rate and the catalyst geom-
etry. Schmidt et al. concluded that mass transfer across the
boundary layer over a catalyst surface strongly influences
reactor selectivity for fast reactions. Several variables dur-
ing H2 and CO production were examined. Selectivity was
improved by operating at higher gas and catalyst temper-
atures by maintaining high rates of mass transfer through
the surface boundary layer and by using catalysts with high
metal loadings. High flow rates minimized mass-transfer
limitations, causing gauze, foam monoliths, and extruded
monoliths to give similar selectivities and conversions. How-
ever, important differences resulted from different catalyst
geometries and thermal conductivities.
Supported noble metal catalysts are mainly used for the
POM reaction, including Rh, Ru, Pd, and Pt [24–27], and
supported Ni catalysts [28–30]. The type of catalyst em-
ployed may strongly influence the reaction steps of catalytic
POM. Elmasides and Verykios [31] investigated the partial
oxidation of methane to synthesis gas over Ru/TiO2 em-
ploying non-steady-state and steady-state isotopic transient
experiments, combined with in situ DRIFT spectroscopy.
The present investigation concerns the partial oxidation
of methane to syngas over Rh/SiO2 and Ru/SiO2 catalysts.

Q.G. Yan et al. / Journal of Catalysis 226 (2004) 247–259
249
Interaction of CH4/Ar or CH4/O2/Ar with both Rh/SiO2 and
Ru/SiO2 catalysts was examined by pulse and transient reac-
tion techniques with on-line mass spectrometry monitoring.
In situ microprobe Raman spectroscopy was used to study
the oxidation state of the catalyst’s surface under POM re-
action conditions. The catalysts were characterized by TPR,
XRD, and H2 chemisorption. The different results observed
over Rh/SiO2 versus those using Ru/SiO2 are explained.
He (m/z = 4), CH4 (m/z = 16), H2O (m/z = 18), CO or N2
(m/z = 28), O2 (m/z = 32), and CO2 (m/z = 44).
2.3.1. TPR experiments
Fresh catalyst was first pretreated in a He flow at 500 ◦C
for 1 h. The sample was then cooled to room temperature
and the He flow was switched to a 3% by volume H2 in N2
flow. After the effluent concentrations had reached constant
values for 30 min, the temperature was ramped at a rate of
15 ◦C/min to 1000◦C.
2. Experimental
2.3.2. O2-TPD experiments
2.1. Catalyst preparation
The catalyst was first pretreated at 700 ◦C under flowing
pure oxygen for 10 min and then cooled to room tempera-
ture. Then a He flow was introduced for 30 min. The temper-
ature was next increased at a rate of 20 ◦C/min to 1000 ◦C
while helium was next passed through the catalyst bed.
The pulse reaction experiments were performed at 700 ◦C
and 1 atm. Helium (30 ml/min) was used as the carrier gas,
and 1 ml CH4/Ar (1/10) per pulse was employed. The sam-
ple was continuously exposed to a flow of 30 ml/min He.
The desired gas was pulsed into this He stream. The reaction
products were quantified with a mass spectrometer.
The transient experiments were performed at 500 and
700 ◦C under 1 atm. The reactants were CH4/O2/Ar (2/1/20).
The supported Rh and Ru catalysts were prepared by so-
lution impregnation. The support SiO2 (Aldrich Chemical
Company, 60–80 mesh, 406 m2/g) was first impregnated
with either a RhCl3·3H2O or a RuCl3·3H2O aqueous solu-
tion. The water was evaporated and the solids were dried
overnight in a vacuum oven at 110 ◦C and calcined in air at
500 ◦C for 6 h. A series of 1.0, 2.0, and 4.0 wt% catalysts
were prepared. No chlorine content was detected in either of
the catalysts by XPS.
2.2. Catalytic performance tests
Catalytic reaction tests were carried out in a fixed-bed
quartz microreactor (200 mm, 6 mm i.d.). A 250-mg cat-
alyst sample was charged. The reaction system was first
purged with nitrogen, and the catalyst was reduced under
pure hydrogen flow (10 ml/min) at 500 ◦C for 0.5 h. Then,
a mixed CH4/O2 gas stream was introduced. The ratio of
CH4/O2 in the feed gas was held constant at 2.0 (molar ra-
tio). Activity testing was carried out at a GHSV of 1 × 105,
1.5×105, and 2×105 h−1, 450–800◦C and 1 atm. Analyses
of the reactant/product mixtures were conducted on a Model
103G gas chromatograph with a TC detector. A carbon sieve
TDX-01 column was used to separate H2, CO, O2, CH4, and
CO2 using argon as the carrier gas. The column was main-
tained at 110 ◦C under an argon flow of 30 ml/min. The
amount of carbon deposited on the catalysts after POM was
obtained via combustion chromatography as previously de-
scribed [13].
2.4. In situ microprobe Raman characterization
In situ microprobe Raman catalyst characterization ex-
periments were performed using a home-built high-tempera-
ture in situ microprobe Raman cell, allowing Raman spectra
of the catalysts to be obtained from room temperature to
700 ◦C under different gas atmospheres. Raman spectra were
recorded on a Dilor LabRam I cofocal microprobe Raman
system. The exciting wavelength of 514.5 nm was generated
with an Ar+ laser with a power of 15 mW and a spot size of
ca. 3 µm2. The laser beam was focused on the top of the cata-
lyst bed. In each experiment, the catalyst (ca. 2.5×10−3 ml)
was first subjected to a flow of O2 (99.995%) at 500 ◦C for
30 min. After O2 pretreatment, the Raman spectra of the cat-
alysts were obtained. Then the treatment was switched to a
H2/Ar (5/95, molar ratio) flow at 500 or 600 ◦C for 5 min and
the Raman spectra of H2-reduced sample were obtained. Fi-
nally, the H2-reduced catalyst bed was subjected to a flow of
CH4/O2/Ar (2/1/45, molar ratio) at 500 or 600 ◦C for 5 min,
and then the Raman spectra of the catalysts were obtained
under POM conditions at a flow rate of 5 ml/min. All the
Raman spectra were recorded under steady-state conditions.
2.3. Temperature-programmed reduction (TPR),
temperature-programmed desorption (TPD), pulsed
reactions, and transient responses to step changes of the
feed gases
TPR, TPD, pulsed reactions, and the transient responses
to step changes in the feed gases were each carried out in
a fixed-bed tubular quartz microreactor with an inner diam-
eter of 3 mm and 18 cm length. A 50-mg catalyst sample
was used in every run. An on-line Balzers quadruple mass
spectrometer (QMS200) continuously monitored the effluent
from the reactor. The effluents could contain H2 (m/z = 2),
2.5. X-ray diffraction (XRD) and metal dispersions
XRD patterns were obtained with a Philips PW 1840
powder diffractometer. Co-Kα radiation was employed, cov-
ering 2θ between 20 and 80◦. Metal dispersions on the cata-
lysts were measured by H2 chemisorption at room tempera-
ture. The percentage dispersion of Rh or Ru metal was calcu-

250
Q.G. Yan et al. / Journal of Catalysis 226 (2004) 247–259
lated assuming a H/M atomic ratio of 1 [41]. The metal dis-
persion was estimated by hydrogen chemisorption at room
temperature. The catalyst sample was first reduced under hy-
drogen at 500 ◦C for 1 h, evacuated at 500 ◦C under high
vacuum for 30 min, and then cooled to room temperature
under vacuum for chemisorption.
and H2 selectivities decreased with increasing space veloc-
ity over 1 wt% Ru/SiO2, while on 1.0 wt% Rh/SiO2 the
CH4 conversion and the CO and H2 selectivities maintained
nearly constant. Both the CH4 conversions and the CO and
H2 selectivities increased slightly over Rh/SiO2 at 700 ◦C as
GHSV increases, while the CH4 conversions and the CO and
H2 selectivities decreased slightly over Ru/SiO2.
3. Results
3.2. X-ray diffraction, temperature-programmed reduction
(H2-TPR), and metal dispersions on catalysts
3.1. Catalytic performance
3.2.1. Catalyst XRD patterns
A CH4/O2 gas mixture with a molar ratio of 2/1 was used
as the feed gas, in POM reactions conducted between 450
and 800 ◦C at 1 atm, and at a GHSV of 150,000 h−1 over
1 wt% Rh/SiO2 and 1 wt% Ru/SiO2. The effect of tempera-
ture on catalyst reactivity was investigated. Product distribu-
tions, conversions, and selectivities are presented in Table 1.
The conversion of methane and selectivity to syngas both
increased with increasing temperature. The conversions and
selectivities were high at high temperatures (ꢀ 700 ◦C) for
both catalysts. But, conversion of CH4 and O2 and selec-
tivity of CO and H2 are obviously different for these two
catalysts at low temperatures (450–600◦C). Specifically, the
conversion of CH4 and selectivity to both H2 and CO over
Rh/SiO2 catalyst were higher than those over Ru/SiO2. The
O2 conversion was a little lower over Rh/SiO2 than that
over Ru/SiO2. After completing the POM reactions, com-
bustion chromatography was employed to detect any carbon
deposition on the catalysts. However, no evidence for carbon
deposition was detected for either catalyst after the POM re-
action.
The XRD patterns of 2.0 wt% Rh(O)/SiO2 and 2.0 wt%
Ru(O)/SiO2 are shown in Fig. 1. The use of (O) indicates
these catalysts had been calcined in the presence of oxygen.
X-ray diffraction analysis showed that a mixture of Rh2O3
and SiO2 phases was obtained for the 2.0% Rh(O)/SiO2
catalyst after calcinations. RuO2 and SiO2 phases were
found on the 2.0 wt% Ru(O)/SiO2 catalyst. The 1.0 wt%
Rh(O)/SiO2 catalyst was analyzed by XRD (not shown
here), but no diffractions corresponding to Rh or Rh2O3
were present. Only SiO2 was detected in Rh(O)/SiO2.
Thus, rhodium particles must be highly dispersed with sizes
smaller than the detection limit of the instrument. One of the
reasons for this high dispersion is that the rhodium content
is very low. XRD of 1.0 wt% Ru(O)/SiO2 exhibited the SiO2
phase with very weak RuO2 lines. The XRD described here
was performed on 2.0 wt% metal catalysts.
3.2.2. TPR of catalysts
TPR experiments were conducted from 25 to 500 ◦C to
investigate the reducibility of rhodium oxide and ruthenium
oxide on the 1.0% Rh(O)/SiO2 and 1% Ru(O)/SiO2 cat-
alysts. Fig. 2 illustrates hydrogen consumption by these
Table 2 shows the effect of space velocity on the CH4
and oxygen conversions and on syngas selectivities at both
500 and 700 ◦C. At 500 ◦C, the CH4 conversion and the CO
Table 1
Methane and oxygen conversion and the selectivity to H and CO during the partial oxidation of methane over 1.0 wt% Rh/SiO and 1.0 wt% Ru/SiO
catalysts
2
2
2
a
Catalyst
Temperature
( C)
Outlet gas composition (%)
Conversion (%)
CH
Selectivity (%)
CO
◦
H
O
CH
CO
CO
O
H
2
2
2
4
2
4
2
1.0 wt%
450
500
550
600
650
700
750
800
28.0
46.1
52.1
57.2
61.0
64.0
64.9
65.7
8.3
3.5
1.2
0.5
0
0
0
0
36.5
16.5
10.5
5.9
3.7
1.9
12.0
19.8
25.9
29.6
30.8
32.6
32.8
33.0
15.2
14.1
10.2
6.8
4.5
1.5
42.7
67.2
77.5
86.0
90.5
94.7
96.6
97.6
80.5
90.8
96.5
98.0
51.5
68.2
72.2
78.8
86.4
93.8
95.7
98.0
44.1
58.4
77.5
81.3
87.3
95.6
96.7
98.5
Rh/SiO
2
100
100
100
100
1.2
0.8
1.1
0.5
1.0 wt%
Ru/SiO
450
500
550
600
650
700
750
800
25.9
43.0
50.5
54.7
60.5
63.6
64.7
65.5
6.0
1.6
0.7
0
0
0
39.8
19.8
13.0
8.9
4.2
2.0
9.0
18.4
23.1
25.9
30.1
32.6
32.9
32.7
19.3
17.2
12.7
9.4
5.2
1.8
39.8
64.2
73.1
83.7
89.4
94.3
96.9
97.4
88.3
93.5
97.8
45.7
60.4
70.8
76.4
85.7
92.4
94.6
97.8
31.8
51.7
65.4
78.5
85.3
94.8
95.6
97.0
2
100
100
100
100
100
0
0
1.1
0.9
1.3
0.9
a
−1
.
Reaction conditions: pressure = 1 atm, CH /O = 2/1, GHSV = 150,000 h
4
2

Q.G. Yan et al. / Journal of Catalysis 226 (2004) 247–259
251
Table 2
Effect of gas hourly space velocity (GHSV) on methane conversion and the selectivity to H and CO during the partial oxidation of methane over 1.0 wt%
2
a
Rh/SiO and 1.0 wt% Ru/SiO catalysts
2
2
◦
−1
)
Catalyst
Temperature ( C)
GHSV (h
CH conversion (%)
Selectivity (%)
4
CH
O
H
2
CO
4
2
1.0 wt%
Rh/SiO
500
500
500
700
700
700
100,000
150,000
200,000
100,000
150,000
200,000
68.8
67.2
67.0
93.5
94.7
94.6
91.7
90.8
90.0
100
100
100
69.5
68.2
66.2
93.4
93.8
94.7
59.1
58.4
56.5
95.3
95.6
96.7
2
1.0 wt%
Ru/SiO
500
500
500
700
700
700
100,000
150,000
200,000
100,000
150,000
200,000
66.3
64.2
61.1
94.5
94.3
93.9
95.1
93.5
91.0
100
100
100
66.5
60.4
53.6
91.7
92.4
91.6
57.9
51.7
45.4
95.3
94.8
94.6
2
a
Reaction conditions: pressure = 1 atm, CH /O = 2/1.
4
2
Fig. 2. Temperature-programmed reduction (TPR) profiles of the 1 wt%
Rh/SiO catalyst and the 1 wt% Ru/SiO catalyst.
Fig. 1. XRD patterns of Rh/SiO and Ru/SiO catalysts.
2
2
2
2
catalysts as a function of the temperature during the re-
duction. Reduction of the 1.0% Rh(O)/SiO2 catalyst took
place between 80 and 180 ◦C and was characterized by one
sharp peak with a maximum temperature around 120 ◦C. The
hydrogen consumption corresponds to a full reduction of
rhodium oxide to Rh0, i.e., (Rh2O3 + 3H2 → 2Rh + 3H2O).
The reduction profile for 1.0 wt% Ru(O)/SiO2 exhibited
peaks different from those for 1.0% Rh(O)/SiO2. Two re-
duction peaks were found in reductions of 1.0% Ru(O)/SiO2
atoms are present in 50-mg samples of these catalysts.
Thus, the 1.0 wt% Rh(O)/SiO2 catalyst must consume
7.29 × 10−6 mol H2 in TPR reactions if all the rhodium
were present as Rh2O3 and all Rh2O3 is completely re-
duced to Rh0 and H2O. Similarly, 1.0 wt% Ru(O)/SiO2
should consume 9.9 × 10−6 mol H2 if all RuO2 were re-
duced.
between 80 and 250 ◦C. These two peaks had Tmax
=
3.2.3. Metal dispersions
150 ◦C and Tmax = 200 ◦C, with the latter accounting for
significantly more hydrogen uptake. The lower tempera-
ture (150 ◦C) TPR peak has been assigned to the reduc-
tion of well-dispersed RuOx species [42], and the high-
temperature peak is attributed to the reduction of RuO2
particles (RuO2 + 2H2 → Ru + 2H2O) [43]. The hydro-
gen consumption for 1.0% Ru(O)/SiO2, calculated from
the area of the reduction peak, was about three times that
of 1.0% Rh(O)/SiO2. About 4.9 × 10−6 mol Rh or Ru
The metal dispersion of the reduced catalysts was ob-
tained by H2 chemisorption at 298 K. The percentage disper-
sion of Rh or Ru metal, calculated assuming a H/M atomic
ratio of 1, is shown in Table 3. The dispersion of rhodium
was 95% for 1.0 wt% Rh/SiO2 and ruthenium dispersion
was 33% for 1.0 wt% Ru/SiO2. The dispersion of Ru/SiO2 is
much less than that of Rh/SiO2. The dispersions for both the
2% Rh and 2% Ru catalysts were about half of those found
in the 1% Rh and Ru catalysts, respectively.

252
Q.G. Yan et al. / Journal of Catalysis 226 (2004) 247–259
Table 3
first pulse to form CO2. Readily observed tailing of the CO
The metal dispersion of the catalysts
peak occurred after each CH4 pulse. The CO tailed for 2–
3 min before reaching background level, while CO2 peaks
lasted only 5–6 s. We suggest that adsorbed CHx (x = 0–3)
reacts with oxygen species to form CO, while CO2 might
come from total oxidation methane (3CH4 + 4Rh2O3 →
8Rh + 3CO2 + 6H2O).
Catalyst
H/M (atomic ratio)
Dispersion (%)
1.0 wt% Rh/SiO
2.0 wt% Rh/SiO
1.0 wt% Ru/SiO
2.0 wt% Ru/SiO
0.95
0.50
0.33
0.15
95
50
33
15
2
2
2
2
The production of CO2 and CO in CH4 pulses over
reduced Rh/SiO2 at 700 ◦C is shown in Fig. 4b. Similar
to the response of oxidized Rh(O)/SiO2, the intensities of
CO and CO2 are high during the first pulse reduced over
Rh/SiO2. CO2 appears also only at the first pulse over the
reduced Rh/SiO2. No CO2 is detected after the first pulse.
The amount of CO produced was much lower than that
formed with Rh(O)/SiO2. CO peaks tailed when produced
by methane pulsing over both the oxidized Rh(O)/SiO2 and
the reduced Rh/SiO2 catalysts.
CH4 pulsing over the unreduced Ru(O)/SiO2 also pro-
duces CO and CO2 (Fig. 4c). Large amounts of CO and CO2
were detected in the first pulse, and the intensity was much
stronger than that found in the evolution of CO and CO2
from the Rh/SiO2 catalyst. However, unlike the Rh/SiO2 cat-
alyst, CO2 is formed at every pulse over the Ru(O)/SiO2
catalyst. Pulsing CH4 over reduced Ru/SiO2 also produced
both CO and CO2 with every pulse (Fig. 4d). The reaction
(2RuO2 + CH4 → 2Ru + CO2 + 2H2O) will take place over
Ru/SiO2. These products are very similar to those formed
over the unreduced Ru/SiO2, except that the intensities of
the CO and CO2 peaks are higher than those from reduced
Ru/SiO2. Tailing of the CO peak is also observed during
pulsing over Ru/SiO2 and the tailing time varies from 2 to
3 min. No CO2 tailing occurs.
Pulse methane reactions were conducted over Rh/SiO2
and Ru/SiO2 catalysts without added oxygen (Fig. 4a–4d).
Any methane oxidation in these experiments must be derived
from oxygen originating from the solid catalyst systems. In
each pulse experiment, 1 ml of reactants (CH4/Ar = 1/10)
was introduced. Each 1 ml contains about 4.46 × 10−6 mol
methane. If all the Rh is present as Rh2O3, the unreduced
Rh/SiO2 catalyst will consume ca. 1.82 × 10−6 mol CH4.
Similarly, the unreduced Ru/SiO2 should consume 2.50 ×
10−6 mol CH4 if all the Ru is present as RuO2. So a sin-
gle CH4/Ar pulse of 1 ml would be sufficient to reduce all
the rhodium and ruthenium oxides (∼ 1.0 wt%, ca. 4.9 ×
10−6 mol) in each of these 50-mg catalyst samples to the
metallic state.
Fig. 3. O -TPD profile of (a) the 1 wt% Rh/SiO catalyst and (b) the 1 wt%
2
2
Ru/SiO catalyst.
2
3.3. Temperature-programmed desorption of O2 (O2-TPD)
O2-TPD experiments were carried out to characterize the
metal oxides in Rh(O)/SiO2 and Ru(O)/SiO2. Both cata-
lysts were oxidized under a pure oxygen flow (10 ml/min)
at 700 ◦C for 30 min and then cooled to room temper-
ature before the TPD experiments were conducted. O2-
TPD of the two catalysts is depicted in Fig. 3. O2 started
to desorb from the 1.0 wt% Rh(O)/SiO2 catalyst around
550 ◦C and the desorption peak maximum was reached at ca.
800 ◦C, where the reaction 2Rh2O3 → 4Rh + 3O2 may oc-
cur. No oxygen desorption peak occurred during O2-TPD of
1 wt% Ru(O)/SiO2, perhaps because the reaction 3RuO2 →
Ru + 2RuO3 (or RuO4) takes place instead at high tempera-
tures [44].
3.4. Interaction of methane with the catalysts by pulsed
reactions
Oxygen-free CH4 pulsing reactions were carried out over
both 1.0 wt% Rh/SiO2 and 1.0 wt% Ru/SiO2 at 700 ◦C
and 1 atm pressure. The CO and CO2 produced are shown
in Fig. 4. Fig. 4a depicts the CO and CO2 formation re-
sponses during CH4 pulsing over unreduced Rh(O)/SiO2.
Large amounts of CO and CO2 were observed at the first
pulse. The formation of CO and CO2 may result from CH4
reacting with the active oxygen species on the Rh and/or
SiO2 support surface. However, no CO2 formation was ob-
served after the first pulse. The concentration of oxygen
species on the catalyst surface might be too low after the
3.5. Transient reactions of CH4/O2/Ar with the catalysts
Transient experiments were performed at 500 and 700 ◦C
under 1 atm using step changes in the feed gas from He
to CH4/O2/Ar. The responses of the product distribution to
these step changes, where oxygen has now been added, are
shown in Fig. 5.
Fig. 5a plots the responses to a step change in the feed
gas from helium to CH4/O2/Ar over unreduced Rh(O)/SiO2

Q.G. Yan et al. / Journal of Catalysis 226 (2004) 247–259
253
(a)
(b)
(c)
(d)
◦
Fig. 4. CO and CO produced by pulsing CH at 700 C over (a) the preoxidized Rh(O)/SiO , (b) the prereduced Rh/SiO , (c) the preoxidized Ru(O)/SiO ,
2
4
2
2
2
and (d) the prereduced Ru/SiO catalyst.
2
at 500 ◦C. It is obvious that CO2 (4.0 s) appears before CO
(4.5 s).
sharply to reach their maximum and then decrease slowly to
a constant level. Only a very tiny quantity of CO2 was de-
tected (5.0 s) after the step change.
The product responses to a feed gas change from He to
CH4/O2/Ar over unreduced Ru(O)/SiO2 at 500 ◦C are shown
in Fig. 5e. CO2 appeared at 2.6 s after this step change, while
CO appeared at 3.0 s. Fig. 5f depicts the transient prod-
uct response to the feed gas step change over the reduced
Ru/SiO2 catalyst at 500 ◦C. The CO2 (3.3 s) appears prior to
CO (4.5 s).
The product responses to a feed gas change from He to
CH4/O2/Ar over unreduced Ru(O)/SiO2 at 700 ◦C are shown
in Fig. 5g. CO2 (2.2 s) appears before CO (4.2 s) after the
step change. This contrasts with the behavior induced by the
oxidized Rh(O)/SiO2 catalyst at 700 ◦C. Furthermore, CO2
reaches a maximum first and then decreases rapidly to the
background level. CO is rapidly formed, rising quickly fol-
lowed by a long gradual further increase.
Fig. 5b plots the products versus time as the step change
from helium to CH4/O2/Ar was made at 500 ◦C over the re-
duced Rh/SiO2. This Rh/SiO2 catalyst had been previously
subjected to a 30 min reduction by H2 at 500 ◦C. CO and
CO2 appear at the same time (3.7 s) and then increase to a
steady level after the gas feed switch.
Fig. 5c plots the product responses to changing the feed
gas from helium to CH4/O2/Ar over oxidized Rh(O)/SiO2
at 700 ◦C. CO (2.7 s) appears at once and sharply increases
to a steady level. CO appears slightly before CO2 (4.0 s).
The amount of CO2 increases to a steady state level but the
quantity of CO2 formed is much lower than the amount of
H2 and CO generated.
The responses of the step change in the feed gases from
helium to CH4/O2/Ar were also obtained over the reduced
Rh/SiO2 catalyst (Fig. 5d) which had been previously sub-
jected to a flow of H2 for 30 min at 700 ◦C. CO (3.9 s) and H2
appear immediately after the change from He to CH4/O2/Ar
was made at 700 ◦C. The amounts of H2 and CO increase
Fig. 5h depicts the transient product response to the He to
CH4/O2/Ar step change over the reduced Ru/SiO2 catalyst
at 700 ◦C. CO2 (3.0 s) still appears prior to CO (3.6 s); how-

254
Q.G. Yan et al. / Journal of Catalysis 226 (2004) 247–259
(a)
(b)
(c)
(d)
◦
◦
Fig. 5. Transient responses upon a step change in feed gas from He to CH /O /Ar at 500 and 700 C over (a) preoxidized Rh/SiO at 500 C, (b) prereduced
4
2
2
◦
◦
◦
◦
Rh(O)/SiO at 500 C, (c) preoxidized Rh/SiO at 700 C, (d) prereduced Rh(O)/SiO at 700 C, (e) preoxidized Ru(O)/SiO at 500 C, (f) prereduced
Ru/SiO at 500 C, (g) preoxidized Ru(O)/SiO at 700 C, and (h) prereduced Ru/SiO at 700 C.
2
2
2
2
◦
◦
◦
2
2
2
ever, only very small quantities of CO2 were found to form
over reduced Ru/SiO2 catalyst.
(Section 3.3). Thus, metallic rhodium surfaces are avail-
able from 500 to 800 ◦C. The Raman spectrum of 4 wt%
Ru/SiO2 catalyst was recorded at 600 ◦C under O2 (Fig. 6b).
Two bands attributable to ruthenium oxide appear at 489
and 609 cm−1 [46,47]. These bands vanished when the O2-
pretreated 4 wt% Ru/SiO2 was subsequently placed under a
flow of H2/Ar at 600 ◦C, and then these bands reappeared
when a flow of CH4/O2/Ar gases was introduced at the same
temperature. Thus, oxides of ruthenium form on the catalyst
under POM conditions at 600 ◦C.
3.6. In situ microprobe Raman spectroscopy
The Raman spectrum of 4 wt% Rh/SiO2 was recorded at
500 ◦C under an O2 flow (Fig. 6a). A broad band centered
at 491 cm−1. It was attributed to Rh2O3 [42]. Many stud-
ies of Rh catalysts have proved that complete oxidation of
Rh to Rh2O3 takes place under an O2 atmosphere at tem-
peratures above 500 ◦C [45]. The Rh2O3 band at 491 cm−1
disappeared when the O2-pretreated Rh/SiO2 sample was re-
duced in a flow of H2/Ar at 500 ◦C. No Rh2O3 491 cm−1
band appeared when the H2/Ar pretreated Rh/SiO2 was
switched to a flow of CH4/O2/Ar at the same tempera-
ture. Thus, surface oxides of rhodium do not form in ob-
servable quantities under POM conditions at 500 ◦C. At
higher temperatures the thermal desorption of oxygen from
rhodium surfaces was shown to become more pronounced
4. Discussion
4.1. Characterization of Rh/SiO2 and Ru/SiO2 catalysts
and effects of oxidation states on POM mechanism over
Rh/SiO2 and Ru/SiO2 catalysts
Rh2O3 is the most stable oxide of rhodium [48]. Com-
plete oxidation of Rh to Rh2O3 occurs above 500 ◦C un-

Q.G. Yan et al. / Journal of Catalysis 226 (2004) 247–259
255
(e)
(f)
(g)
(h)
Fig. 5. Continued.
der an O2 atmosphere. Rh2O3 thermally decomposes only
at 1100 ◦C [49], but it easily reduces with hydrogen at
200 ◦C [50]. Facile reduction of Rh(O)/SiO2 by H2 was ob-
served at 120 ◦C in TPR experiments (Fig. 2). Ruthenium
oxides including Ru2O3, RuO2, RuO3, and RuO4 are more
resistant to reduction than Rh2O3 [51]. The reduction of
1 wt% Rh(O)/SiO2 takes place between 80 and 180 ◦C with
the H2 uptake peak occurring at Tmax = 120 ◦C (Fig. 2). In
contrast, Ru(O)/SiO2 exhibits two hydrogen uptake peaks,
with the smaller peak maximizing at 150 ◦C and the larger
at 200 ◦C. A single reduction peak from Rh(O)/SiO2 is
consistent with excellent surface dispersion of the rhodium
oxides on SiO2. The two peaks (150 and 200 ◦C) in the
TPR profile of Ru(O)/SiO2 suggest the presence of two
different ruthenium oxides [42] where Ru–O bonding is
stronger than rhodium–oxygen bonding. The 150 ◦C TPR
peak has been assigned to the reduction of well-dispersed
RuOx species [42], while the high-temperature peak is at-
tributed to the reduction of RuO2 particles (RuO2 + 2H2 →
Ru + 2H2O) [43].
Different surface metal oxidation states and/or the rel-
ative bond strengths of transition metals with oxygen can
result in different mechanisms [32,33,35]. Stronger M–O
bond strengths give more stable transition metal oxides. This
could lead to differences in activity and mechanism during
POM.
CH4 pulse experiments provide some mechanistic in-
sight. A noticeable difference was observed in the reaction
of pure methane over the unreduced versus the reduced
1.0 wt% Rh and Ru catalysts at 700 ◦C. CO is both the
direct product and the major product over unreduced and re-
duced 1.0 wt% Rh/SiO2. This result is in accord with previ-
ous in situ time-resolved FTIR spectroscopy studies [39,40],
in which CO was the primary POM reaction product over
both reduced Rh metal and working state Rh/SiO2 catalysts.
In contrast, CO2 is a primary product during POM over
Ru/SiO2. CO2 was formed in every pulse when oxygen-free
CH4 was pulsed over Ru/SiO2 at 700 ◦C. Therefore, it is
clear that the oxygen, contained in the ruthenium oxides, was
responsible for both CO2 and CO formation occurring over
Ru/SiO2. In contrast, CO2 was only formed during the first

256
Q.G. Yan et al. / Journal of Catalysis 226 (2004) 247–259
CHx species on the rhodium catalyst’s surface to CO2 before
CO desorbs. Thus, the amount and state of the metal oxides
present on the Rh/SiO2 and Ru/SiO2 catalysts are of sem-
inal importance to the methane activation pathway and the
product distribution.
CO is the primary product over Rh/SiO2 catalysts af-
ter the change from He to CH4/O2/Ar was made at 700 ◦C
for the transient reaction (Fig. 5c and 5d) and only a very
tiny amount of CO2 was detected. CO2 is the primary prod-
uct for both preoxidized and prereduced Ru/SiO2 catalysts
at 500 and 700 ◦C, implying that the concentration of oxy-
gen species on the catalyst may have a significant influence
on the reaction schemes for POM. Based on TAP (tempo-
ral analysis of products) studies of POM over individual
metal or supported metal catalysts, such as Rh sponge [52],
Rh/Al2O3 [52,53], and Pt gauze [32], some researchers have
concluded that the formation of primary products depends
on the amount/concentration of oxygen species on the cat-
alyst surface. This suggests that the amount of oxygen on
Rh/SiO2 surfaces where direct oxidation of methane to syn-
gas occurs will be less than that on Ru/SiO2 surfaces where
the POM occurs via an indirect combustion/reformingmech-
anism.
(a)
In situ microprobe Raman spectroscopy characterization
of the 4 wt% preoxidized Rh/SiO2 surface exhibited the loss
within 2 min of the broad 491 cm−1 band of Rh2O3 (found
at 500 ◦C under O2) [42], when the gas flow was switched
to H2/Ar at 500 ◦C. The Raman spectrum of this cata-
lyst recorded under a simulated POM feed (CH4/O2/Ar =
2/1/45) is very similar to that recorded under the H2/Ar at-
mosphere. No Rh2O3 was detected. Similar spectra were
obtained when heating the catalyst in a CH4/O2/Ar between
500 and 600 ◦C. Clearly, most of the Rh species over the
Rh/SiO2 are in the metallic state during POM. Mallens et
al. [32] reached the same conclusion from their TAP exper-
iment, where by assuming the Rh2O3 oxide stoichiometry,
only 0.4 wt% of Rh2O3 is present during the simultane-
ous interaction of CH4 and O2 at a stoichiometric feed ratio
(CH4/O2 = 2/1) over the rhodium sponge.
Raman bands (489 and 609 cm−1) attributed to RuO2
[46,47] were observed for the 4 wt% Ru/SiO2 under O2
at 500 to 600 ◦C. These bands disappeared when the pre-
oxidized 4 wt% Ru/SiO2 was exposed to H2/Ar at 500 to
600 ◦C and reappeared when the H2/Ar pretreated Ru/SiO2
was switched to a flow of CH4/O2/Ar at the same tempera-
ture. Thus, the Ru species of the Ru/SiO2 catalyst are almost
fully oxidized under the POM conditions. In situ microprobe
Raman characterizations of Rh/SiO2 and Ru/SiO2 catalyst
(Fig. 6) under POM conditions correspond to pulse and tran-
sient reaction results, showing that a significant difference
in mechanism exists over these two catalysts. The concen-
tration of oxygen species on the catalyst surface causes the
mechanistic change. No Raman band of Rh2O3 species was
detected in a working-state Rh/SiO2. Even at the top of the
catalyst bed, most of the Rh is in the metallic state during
POM. CH4 is activated and dissociated to CHx (x = 0–3)
(b)
◦
Fig. 6. In situ Raman spectra of (a) 4 wt% Rh/SiO at 500 C and (b) 4 wt%
2
Ru/SiO at 600 C when each of these catalysts was under O , CH /O /Ar
◦
2
2
4
2
(2/1/45), or 5% H /95% Ar.
2
pulse and never thereafter when methane was pulsed over
Rh/SiO2. Thus, surface oxygen on rhodium sites favors the
formation of CO. Apparently, the very small rhodium oxide
surface concentration is too low to continue to fully oxidize

Q.G. Yan et al. / Journal of Catalysis 226 (2004) 247–259
257
species on Rh0 sites and then converted to CO through sur-
face reactions, such as H dissociation from CHx and the
oxidation of surface C or CHx (x = 1–3) species by surface
oxygen species (e.g., O2−).
A more significant effect on Ru/SiO2 occurred when GHSV
increased from 100,000 to 150,000 h−1. Methane conver-
sion decreased from 66.3 to 61.1%; CO selectivity dropped
from 57.9 to 45.4%, and H2 selectivity reduced from 66.5
to 53.6%. These results also support direct oxidation with
Rh/SiO2 and indirect combustion/reforming over Ru/SiO2,
in agreement with POM mechanisms proposed based on the
in situ micro-Raman, pulse and transient reactions studies.
In contrast to Rh/SiO2, the Ru species on Ru/SiO2 near
the entrance of the catalyst bed are almost fully oxidized
under the reaction conditions. Transient and pulse reactions
show that oxidized Ru species catalyze the complete oxi-
dation of CH4 to CO2, causing complete oxygen consump-
tion from the feed (CH4/O2) within a narrow zone at the
entrance of the catalyst bed. Ru located in the rest of cat-
alyst bed is reduced by the CH4 remaining in the feed to the
metallic state. Reforming of the resulting CH4/CO2/H2O to
syngas occurs at such reduced Ru sites. The oxygen species
on the catalysts could be surface O2−. Surface O2− is re-
sponsible for both the selective and complete oxidation of
methane, depending on its concentration. The significant cat-
alytic differences of Rh/SiO2 and Ru/SiO2 may be related to
the large differences in the Rh–O (405.0 kJ/mol) and Ru–O
(528.4 kJ/mol) bond strengths [54]. This renders the reduc-
tion of Ru/SiO2 more difficult than that for Rh/SiO2. For ex-
ample, temperature-programmed reductions (Fig. 2) demon-
strate a lower reduction temperature for 1 wt% Rh/SiO2
relative to 1 wt% Ru/SiO2. The Ru(O)/SiO2 surface chem-
istry differs from that of Rh(O)/SiO2. O2 dissociates to form
Ru–O species on metallic Ru sites, eventually generating a
RuO2 surface phase that is not easily decomposed. RuO2
inhibits C–H dissociation, desorption of CHx, and reaction
of CHx fragments with dissociatively adsorbed oxygen. The
high surface RuO2 concentration on Ru/SiO2 during low-
temperature (450–600◦C) POM contrasts sharply with the
surface composition of Rh/SiO2. CH4 cannot dissociate over
metallic ruthenium sites and so cannot react via a direct sur-
face oxidation process. However, CH4 can be oxidized via
a nonselective indirect mechanism [55] (CH4 + 2RuO2 →
CO2 + 2H2O + 2Ru0) where CO2 is a primary product over
the ruthenium catalyst. Therefore, POM over Ru/SiO2 oc-
curs mainly via an indirect nonselective oxidation process.
The selectivity of Ru/SiO2 is lower than that of Rh/SiO2.
The low Rh–O bond strength leads to a much lower sur-
face oxygen coverage on Rh/SiO2 catalyst during POM.
Surface oxygen on oxidized Rh(O)/SiO2 begins to ther-
mally desorbed at 550 ◦C according to the O2-TPD exper-
iments (Fig. 3). Reduction of rhodium oxides started at
80 ◦C and peaked at 120 ◦C in TPR experiments, so the
Rh/SiO2 surface must be metallic Rh during POM since H2
is present. Consequently, partial oxidation of CH4 to syngas
over Rh/SiO2 proceeds mainly via a direct oxidation [56,57],
in which CH4 and O2 react at the metal surface via their dis-
sociated and adsorbed species.
4.2. Primary product determination by transient reactions
A main concern in the POM mechanism debate is
whether CO2 or CO is the primary product. For the direct
oxidation scheme, CO is the primary product whereas CO2
should be formed prior to CO in the indirect process. The
TAP reactor has been used to monitor CO and CO2 for-
mation during the POM reaction [25,32–34,37,38,52,53,58,
59]. Responses to pulses of O2, CH4, or CH4/O2 over cat-
alysts under different conditions have provided mechanistic
information. Most TAP studies focused on the POM over
platinum and rhodium. No TAP study of POM with a ruthe-
nium catalyst has been reported, although Ru supported on
TiO2 is an effective catalyst for POM via a direct route [22,
31,60,61]. Mechanistic results from TAP studies of a similar
catalyst still do not agree about the primary reaction prod-
ucts [32,38]. Weng et al. [39,40] used in situ TR-FTIR spec-
troscopy to follow the primary products over SiO2-supported
rhodium and ruthenium catalysts from 500 to 600 ◦C. CO
was the primary POM product over hydrogen-reduced and
working-state Rh catalysts. In contrast, CO2 was the pri-
mary POM product over supported Ru catalysts. Therefore,
direct oxidation of CH4 was proposed with Rh/SiO2, while
the reforming of unreacted CH4 to syngas dominated using
Ru/SiO2. This was accompanied by oxidation of a portion
of the CH4 to CO2 and H2O.
Transient reactions (Fig. 5) were used in the current
study to determine the primary POM products over SiO2-
supported rhodium and ruthenium catalysts at temper-
atures at 500 and 700 ◦C. When O2-pretreated 1 wt%
Rh/SiO2 was subjected to a step change from helium to
CH4/O2/Ar (2/1/45) at 700 ◦C, CO appeared before CO2
(Fig. 5c). Similar results were obtained for prereduced
Rh/SiO2 (Fig. 5d). These results agree with the TAP study
of POM at 600 ◦C over unsupported Pt and Rh catalysts
by Mallens et al. [32,33] where direct oxidation of CH4
occurred upon pulsing with O2, CH4, or CH4/O2. At compa-
rable temperatures, the Rh catalyst shows a higher methane
conversion and greater selectivity to both CO and H2 than Pt.
Fathi et al. [34] also showed that POM could be attributed
to a direct mechanism using a TAP-2 reactor over Pt gauze
at temperatures above 800 ◦C. In this current study, tran-
sient reactions prove that CO2 is the primary POM product
over both preoxidized and prereduced 1 wt% Ru/SiO2 at 500
and 700 ◦C. CO is only detected after CO2. CO results from
combustion of CH4 to CO2 and H2O, followed by reform-
ing. Buyevskaya et al. [25,38], Guerrero-Ruiz et al. [62], and
The gas hourly space velocity (GHSV) CH4/O2 = 2/1
(Table 2) has little effect on the performance of Rh/SiO2 at
500 ◦C when increased from 100,000 to 150,000 h−1. The
CH4 conversion and selectivities to H2 and CO decreased
slightly, CH4 conversion from 68.8 to 67.0%, CO selectivity
from 59.1 to 56.5%, and H2 selectivity from 69.5 to 66.2%.

258
Q.G. Yan et al. / Journal of Catalysis 226 (2004) 247–259
Boucouvalas et al. [63] came to a similar conclusion about
POM based on in situ diffuse reflectance infrared Fourier
transform spectroscopy (DRIFTS), kinetics, and isotopic
tracing and TAP reactor studies.
[5] J.P. Van Hook, Catal. Rev.-Sci. Eng. 21 (1981) 1.
[6] M. Prettre, Ch. Eichner, M. Perrin, Trans. Faraday Soc. 43 (1946) 335.
[7] A.T. Ashcroft, A.K. Cheetham, J.S. Foord, M.L.H. Green, C.P. Grey,
A.J. Murrel, P.D.F. Veron, Nature 344 (1990) 319.
[8] V.A. Tsipouriari, Z. Zhang, X.E. Verykios, J. Catal. 179 (1998) 283.
[9] C.T. Au, H.Y. Wang, J. Catal. 167 (1997) 337.
[10] D. Dissanayake, M.P. Rosynek, K.C.C. Kharas, J.H. Lunsford, J. Ca-
tal. 132 (1991) 117.
[11] E. Ruckenstein, Y.H. Hu, Ind. Eng. Chem. Res. 37 (1998) 1744.
[12] R. Drago, S. Jurczyk, K. Kob, N. Bhattacharyya, A. Masin, J. Catal.
Lett. 51 (1998) 177.
[13] Q. Yan, Z. Chao, T. Wu, W. Weng, H. Wan, Stud. Surf. Sci. Catal. 130D
(2000) 3549.
5. Conclusions
Methane conversion and selectivity to both H2 and CO
were higher over Rh/SiO2 than Ru/SiO2. The CO and H2
selectivities substantially decreased with increasing GHSV
over Ru/SiO2 at 500 ◦C but remained nearly constant over
Rh/SiO2. Both the CH4 conversions and the CO and H2 se-
lectivities increased slightly over both catalysts at 700 ◦C as
GHSV increases, while the CH4 conversions and the CO and
H2 selectivities decreased slightly. CO was the main product
at 700 ◦C when Rh/SiO2 was exposed to methane pulses,
whether the catalyst was preoxidized or prereduced. CO2
was only detected during the first pulse over Rh/SiO2 cat-
alyst. More CO2 was formed over Ru/SiO2 than Rh/SiO2
during the first pulse; CO2 was formed at every pulse over
Ru/SiO2.
[14] K. Nakagawa, N. Ikenaga, Y. Teng, T. Kobayashi, T. Suzuki, J. Ca-
tal. 186 (1999) 405.
[15] H.D. Gesser, N.R. Hunter, Catal. Today 42 (3) (1998) 183.
[16] C. Mirodatos, Stud. Surf. Sci. Catal. 119 (1998) 99.
[17] D.A. Hickman, L.D. Schmidt, AIChE J. 39 (1993) 1164.
[18] D.A. Hickman, L.D. Schmidt, Science 259 (1993) 343.
[19] D.A. Hickman, L.D. Schmidt, J. Catal. 136 (2) (1992) 300.
[20] D.A. Hickman, L.D. Schmidt, J. Catal. 138 (1) (1992) 267.
[21] L.D. Schmidt, M. Huff, S.S. Bharadwaj, Chem. Eng. Sci. 49 (24A)
(1994) 3981.
[22] Y. Boucouvalas, Z.L. Zhang, X.E. Verykios, E. Xenophon, Catal.
Lett. 40 (3,4) (1996) 189.
[23] L.D. Schmidt, M. Huff, S.S. Bharadwaj, Chem. Eng. Sci. 49 (24A)
(1994) 3981.
[24] M.C.J. Bradford, M.A. Vannice, Catal. Today 50 (1999) 87.
[25] O.V. Buyevskaya, D. Wolf, M. Baerns, Catal. Lett. 29 (1994) 249.
[26] P.M. Witt, L.D. Schmidt, J. Catal. 163 (1996) 465.
[27] P.M. Torniainen, X. Chu, L.D. Schmidt, J. Catal. 146 (1994) 1.
[28] W.J.M. Vermeiren, E. Blomsma, P.A. Jacobs, Catal. Today 13 (1992)
427.
CO formed before CO2 over preoxidized or prereduced
Rh/SiO2 at 700 ◦C in transient reactions, induced by chang-
ing the feed gas from helium to CH4/O2/Ar. CO2 is the
primary product over Ru/SiO2 catalysts during transient re-
actions.
In situ microprobe Raman spectroscopy demonstrated
that the oxide levels present on Ru/SiO2 were far higher dur-
ing POM than those on Rh/SiO2. The mechanisms of POM
over these two catalysts are different. On the Rh/SiO2 cat-
alyst, POM is mainly a direct oxidation process, while on
the Ru/SiO2 catalyst, the dominant pathway of POM is the
indirect oxidation process.
[29] D. Dissanayake, M.P. Rosynek, J.H. Lunsford, J. Phys. Chem. 97
(1993) 3644.
[30] V.R. Choudhary, A.M. Rajput, B. Prabhakar, J. Catal. 139 (1993) 326.
[31] C. Elmasides, X.E. Verykios, J. Catal. 203 (2) (2001) 477.
[32] J.E.P. Mallens, J.H.B.J. Hoebink, G.B. Marin, J. Catal. 167 (1997) 43.
[33] E.P.J. Mallens, J.H.B.J. Hoebink, G.B. Marin, Catal. Lett. 33 (1995)
291.
[34] M. Fathi, F. Monnet, Y. Schuurman, A. Holmen, C. Mirodatos, J. Ca-
tal. 190 (2000) 439.
[35] D. Qin, J. Lapszewicz, X. Jiang, J. Catal. 159 (1996) 140.
[36] C. Li, C. Yu, S. Shen, Catal. Lett. 67 (2–4) (2000) 139.
[37] K. Walter, O.V. Buyevskaya, D. Wolf, M. Baerns, Catal. Lett. 29 (1,2)
(1994) 261.
[38] O.V. Buyevskaya, K. Walter, D. Wolf, M. Baerns, Catal. Lett. 38 (1,2)
(1996) 81.
[39] W. Weng, M. Chen, Q. Yan, T. Wu, Z. Chao, Y. Liao, H. Wan, Catal.
Today 63 (2000) 317.
[40] W. Weng, Q. Yan, C. Lou, Y. Liao, et al., Stud. Surf. Sci. Catal. 136
(2001) 233.
Acknowledgments
This project is generously supported by the Ministry of
Science and Technology of China (No. G1999022408), Na-
tional Natural Science Foundation of China (Nos. 20023001,
2002100), the Center for Advanced Vehicular Systems
(CAVS), and the support from the National Science Foun-
dation (EPS0132618) and the US Environmental Protection
Agency (USMGR0129001) are gratefully acknowledged.
[41] C. Hoang-van, Y. Kachaya, S.J. Teichner, J.A. Dalmon, Appl. Catal. 46
(1989) 81.
[42] C.T. Williams, C.G. Takoudis, M.J. Weaver, J. Phys. Chem. B 102
(1998) 406.
[43] L. Ji, J. Lin, H.C. Zeng, Chem. Mater. 13 (7) (2001) 2403.
[44] N.N. Greenwood, A. Earnshaw, Chemistry of the Elements, Pergamon,
Oxford, 1984, p. 1297.
References
[1] J.M. King, M.J. O’Day, J. Power Sources 86 (1–2) (2000) 16.
[2] M.R. Gardiner, J. Cunningham, R.M. Moore, Fuel Cells and Alterna-
tive Fuels/Energy Systems, Society of Automotive Engineers, 2001,
[Special publication] SP-1635, p. 65.
[45] D.D. Beck, T.W. Capehart, C. Wong, D.N. Belton, J. Catal. 144 (1993)
311.
[46] S.Y. Mar, C.S. Chen, Y.S. Huang, K.K. Tiong, Appl. Surf. Sci. 90
(1995) 497.
[3] D.P. Wilkinson, Electrochem. Soc. Interface 10 (1) (2001) 22.
[4] S.C. Amendola, S.L. Sharp-Goldman, M.S. Janjua, N.C. Spencer,
M.T. Kelly, P.J. Petillo, M. Binder, Int. J. Hydrogen Energy 25 (10)
(2000) 969.
[47] Y.S. Huang, P.C. Liao, Sol. Energy Mater. Sol. Cells 55 (1998) 179.
[48] D.R. Lide (Ed.), CRC Handbook of Chemistry and Physics, seventy-
eighth ed., Chemical Rubber Company Press, Boca Raton, FL, 1997,
pp. 4–50.

Q.G. Yan et al. / Journal of Catalysis 226 (2004) 247–259
259
[49] A. Frydman, D.G. Castner, M. Schmal, C.T. Campbell, J. Catal. 152
[56] Y.H. Hu, E. Ruckenstein, J. Phys. Chem. A 102 (1998) 10,568.
[57] Y.H. Hu, E. Ruckenstein, J. Phys. Chem. B 102 (1998) 230.
(1995) 164.
[50] J. Emsley, The Elements, second ed., Clarendon, Oxford, 1991, p. 172.
[51] V.I. Parvulescu, S. Coman, P. Palade, D. Macovei, C.M. Teodorescu,
G. Filoti, R. Molina, G. Poncelet, F.E. Wagner, Appl. Surf. Sci. 141
(1999) 164.
[52] H.K. Hofstad, J.H.B.J. Hoebink, A. Holmen, G.B. Marin, Catal. To-
day 40 (1998) 157.
[53] D. Wang, O. Dewaele, A.M. De Groote, G.F. Froment, J. Catal. 159
(1996) 418.
[54] J.B. Pedley, E.M. Marshall, J. Phys. Chem. Ref. Data 12 (1984) 967.
[55] H.Y. Wang, E. Ruckenstein, J. Catal. 186 (1999) 181.
[58] M. Soick, O. Buyevskaya, M. Höhenberger, D. Wolf, Catal. Today 32
(1996) 163.
[59] M. Soick, D. Wolf, M. Bearn, Chem. Eng. Sci. 55 (2000) 2875.
[60] Y. Boucouvalas, Z.L. Zhang, A.M. Efstathiou, X.E. Verykios, Stud.
Surf. Sci. Catal. 101 (1996) 443.
[61] C. Elmasides, D.I. Kondarides, W. Grünert, X.E. Verykios, J. Phys.
Chem. 103 (1999) 5227.
[62] A. Guerrero-Ruiz, P. Ferreira-Aparicio, M.B. Bachiller-Baeza, I.
Rodríguez-Ramos, Catal. Today 46 (1998) 99.
[63] Y. Boucouvalas, Z.L. Zhang, X.E. Verykios, Catal. Lett. 27 (1994) 131.