Photocatalytic conversion of methane and carbon dioxide over gallium oxide

Article abstract of DOI:10.1016/j.cplett.2007.12.051

Methane is photocatalytically converted to mainly ethane and hydrogen over gallium oxide at room temperature and also at higher temperatures. On the other hand, in the presence of carbon dioxide, although only hydrocarbons and hydrogen were formed on gall

Full text of DOI:10.1016/j.cplett.2007.12.051

Available online at www.sciencedirect.com
Chemical Physics Letters 452 (2008) 178–182
www.elsevier.com/locate/cplett
Photocatalytic conversion of methane and carbon dioxide
over gallium oxide
a
b
a,
*
Leny Yuliati , Hideaki Itoh , Hisao Yoshida
a
Department of Applied Chemistry, Graduate School of Engineering, Nagoya University, Nagoya 464-8603, Japan
b
Division of Environmental Research, EcoTopia Science Institute, Nagoya University, Nagoya 464-8603, Japan
Received 1 November 2007; in final form 19 December 2007
Available online 28 December 2007
Abstract
Methane is photocatalytically converted to mainly ethane and hydrogen over gallium oxide at room temperature and also at higher
temperatures. On the other hand, in the presence of carbon dioxide, although only hydrocarbons and hydrogen were formed on gallium
oxide photocatalyst at room temperature, CO was additionally produced photocatalytically at mild temperatures such as 473 K. The
additional thermal energy was suggested to enhance the reaction between methane and carbon dioxide by promoting the thermally acti-
vation steps in reaction mechanism, such as desorption of the products or electron migration from the bulk to the surface of Ga O .
2
3
Ó 2007 Elsevier B.V. All rights reserved.
1
. Introduction
Related to methane activation, our group has found that
non-oxidative coupling of methane (NOCM) to ethane
and hydrogen can proceed around room temperature over
some photocatalysts, such as silica-based photocatalysts
[3–7], zeolites [8], mesoporous silica [9], and ceria-based
photocatalysts [10]. However, there are only a few papers
Effective utilization of methane and/or CO is one of the
2
important topics to be developed nowadays. Unfortu-
nately, it is not easy to convert these stable molecules to
other useful chemicals at mild reaction condition. For
example, CO (dry) reforming of methane (DRM) [1,2],
about the reaction between methane and CO under photo-
2
2
the reaction between methane and CO , is one of the up-
irradiation [11–13]. For example, photoreduction of CO2
2
hill type reactions (DG > 0) shown as follows:
by methane to CO was reported to occur on ZrO and
2
MgO around room temperature [11,12], where unfortu-
0
¼ 170 kJ molꢀ1
CH
4
þ CO
2
! 2CO þ 2H
2
; DG2
ð1Þ
98 K
nately H was scarcely obtained from methane while other
2
In order to overcome the large DG of this endothermic
reaction, the reaction will normally require suitable catalyst
and high temperature, such as around 1023 K, which will
usually lead to quick catalyst deactivation due to the car-
bon deposition and structural changing of catalytic active
sites. Thus, alternative methods that can lower the reaction
temperature are desired.
products such as H O, surface formate species, etc., were
2
observed instead. Formation of acetone was also reported
as the main product in the photocatalytic reaction between
methane and CO over Cu/CdS–TiO /SiO at low temper-
2
2
2
ature such as 393 K [13].
In the present study, we report the first example for the
conversion of methane and CO to produce both CO and
2
It can be proposed that the use of photoenergy with
photocatalyst would break the thermodynamic barrier of
endothermic reaction to occur at lower temperature.
H at mild temperatures with a photocatalytic reaction sys-
2
tem using Ga O as the photocatalyst. Ga O is a wide
2
3
2
3
band-gap semiconductor and can be photoexcited under
UV irradiation; it has been reported to act as a photocata-
lyst for some reactions, such as water splitting [14,15], and
photodecomposition of benzene [16].
*
Corresponding author. Fax: +81 52 789 3178.
E-mail address: yoshidah@apchem.nagoya-u.ac.jp (H. Yoshida).
0
009-2614/$ - see front matter Ó 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.cplett.2007.12.051

L. Yuliati et al. / Chemical Physics Letters 452 (2008) 178–182
3. Results and discussion
179
2
. Experimental
2
.1. Material characterizations
3.1. Characterizations
Ga O was obtained commercially from Kishida Chem-
The present Ga O sample has a BET specific surface
area around 2 m g . The crystal phase of the present
2
3
2
3
ꢀ1
2
icals. BET specific surface area was measured by a Yuasa
Ionic Monosorb. Powder X-ray diffraction (XRD) pattern
was recorded on a Rigaku diffractometer RINT 2500 using
Cu Ka radiation (50 kV, 100 mA). Diffuse reflectance (DR)
UV–visible spectrum was recorded at room temperature on
a JASCO V-550 equipped with an integrating sphere cov-
sample is b-Ga O as identified by XRD. As confirmed
2
3
by DR UV–vis spectroscopy, the b-Ga O exhibited a
2
3
strong absorption band in the UV region (<290 nm),
where the band-gap energy was estimated to be 4.7 eV.
The Ga K-edge XANES spectrum of Ga O showed
2
3
ered with BaSO . Ga K-edge XANES spectrum was
two distinguishable peaks at 10375.2 and 10379.1 eV
which have been assigned to gallium in tetrahedral
coordination, Ga(Td), and gallium in octahedral coordi-
nation, Ga(Oh), respectively. The analysis of the spec-
trum by deconvolution using curve-fitting method
revealed that the ratio of Ga(Td) to Ga(Oh) was close
to unity, confirming the sample was clearly in b-Ga O
4
recorded at AR-NW10A station of Photon Factory, Insti-
tute of Materials Structure Science, High Energy Accelera-
tor Research Organization (KEK-PF) in Tsukuba, Japan.
2
.2. Photocatalytic reaction test
2
3
The photocatalytic reactions were carried out in the sim-
phase, in good agreement with previously reported liter-
ature [17].
ilar way to the previous studies [3–10] in a closed quartz
3
reactor (30 cm ). The catalyst sample (0.2 g) was spread
2
over a flat bottom of reactor (14 cm ) and treated in
3.2. Photocatalytic reaction results
1
3.3 kPa of oxygen atmosphere at 1073 K for 1 h, followed
by evacuation at 1073 K for 1 h to clean up the catalyst sur-
The reaction results are shown in Table 1. Without
face. Methane (200 lmol) or a mixture of methane
CO , methane was converted to mainly ethane and
2
(
200 lmol) and carbon dioxide (200 lmol) were introduced
hydrogen over Ga O around room temperature (at
2
3
at various reaction temperatures. The reaction temperature
was measured with a thermocouple placed in the holder on
the reactor. The thermocouple holder was filled with the cat-
alyst, so that the measured temperature would be very close
to the reaction temperature inside the reactor. The reactor
was heated by using an electric furnace, which specially
designed with a hole on its bottom to let the UV light reach
the quartz reactor. The catalyst sample then was irradiated
by a 300 W Xe lamp for 3 h from the bottom of the reactor.
Light intensity measured in the range of 220–300 nm wave-
314 K) upon UV irradiation without external heating
(Entry 1). The mass balance among the products could
be confirmed, i.e., the ratio of produced H2 to the
expected one calculated from the amount of produced
hydrocarbons was close to unity. Only a trace amount
of desorbed hydrocarbon products was observed after
additional heating the sample at 573 K. No other oxy-
genated products were observed in this condition. More-
over, no product was obtained in the dark or without
photocatalyst, suggesting that the NOCM (Eq. (2))
occurred photocatalytically on Ga O , as follows:
2
length was about 9 mW/cm at the reactor. For comparison,
2
3
thermal reactions without UV irradiation were also carried
out in the similar way. The products in gaseous phase were
collected and analyzed by gas chromatography.
hm
0
2
¼ 68:6 kJ molꢀ1
2CH
! C
H
6
þ H
; DG
ð2Þ
4
2
2
98 K
Ga2O3
Table 1
a
2 3
Results of reaction tests over b-Ga O
Entry
Light
Reaction temperature
K)
Reactants
(lmol)
Products
(lmol)
2 2
H exp./H
calculated
(
CH
4
CO
2
C
2
H
6
C
2
H
4
C
3
H
8
C
3
H
6
C
4
H
10
H
2
CO
1
2
3
4
5
6
7
8
a
On
On
On
On
Off
On
On
Off
314
314
473
473
473
573
673
673
200
200
200
200
200
200
200
200
0
200
0
200
200
200
200
200
0.51
0.44
1.02
1.04
0
1.04
0.70
tr.
0.01
0.01
0.13
0.12
0
0.12
0.17
0
0.05
0.05
0.10
0.12
0
0.13
0.06
0
tr.
0
0.02
0.03
0
0.04
0.04
0
0.01
0.01
0.01
0.02
0
0.01
tr.
0
0.74
0.53
2.41
2.48
0
4.32
6.32
0
0
tr.
0
1.08
0
1.49
2.29
0
1.1
0.9
1.5
0.9
–
1.3
b
b
b
1.7
–
ꢀ2
Sample was 0.2 g, irradiation time was 3 h, UV light intensity was around 9 mW cm (in the range of 220–300 nm). Prior to the reaction, the sample
was treated at 1073 K. The experimental error on the reaction tests was estimated within 4%.
b
2
The origin of the high H ratio would be from the possibility that consecutive coupling reactions proceeded producing hydrogen with longer
hydrocarbons products that could not be detected in the present experimental method, see text. tr. = trace.

180
L. Yuliati et al. / Chemical Physics Letters 452 (2008) 178–182
In addition, since some other hydrocarbons were
methane during the reaction at 473 K could react with
CO , the amount would be insignificant since CH would
observed such as ethene, propene, and butane, dehydroge-
nation of the produced hydrocarbons and coupling
between produced hydrocarbons and methane and/or
among the produced hydrocarbons themselves would also
proceed as minor reactions. Among the photocatalysts that
2
4
prefer to undergo the coupling reaction than the decompo-
sition reaction under this mild condition over Ga O . No
2
3
change was observed in color of photocatalyst after the
reaction. CO might also be produced via some reactions,
such as shown in the following equations
have been studied in the same reaction condition, Ga O3
2
showed a similar activity to silica–titania photocatalyst,
which exhibited the highest activity to date [3,4].
CO
CO
2
2
þ CH
þ CH
4
4
! CO þ CH
! CO þ CH
2
3
O þ H
2
;
ð5Þ
ð6Þ
When CO was introduced as the reactant at 314 K, no
OH
2
significant effect was observed in the photocatalytic
NOCM (Entry 2); the amounts of the products were very
However, in the present experimental condition, no other
oxygenated product was detected from GC–MS, so that
we could exclude the possibility that CO was produced
via these reactions.
similar to those in the absence of CO . However, interest-
2
ingly, a trace amount of CO was detected in the presence
of CO , suggesting that a trace amount of CO might be
2
2
It was confirmed that CO was produced when hydrocar-
bons such as ethane and/or propane were reacted with CO2
under UV irradiation at 473 K. However, in the present
system CO would be mainly produced from methane since
methane was the major gas in the present system. More-
over, as can be seen from the reaction results, no remark-
ably change was observed in the hydrocarbons products
at each temperature (Entries 1–4). While the detailed mech-
anism would need to be investigated, it is obvious here that,
in a total perspective as the result, we could obtain CO and
H from methane and CO .
reduced and/or a trace amount of CH might be oxidized
during the reaction.
4
When the conversion of methane was carried out in the
absence of CO under UV irradiation at higher tempera-
2
ture, i.e., 473 K (Entry 3), the coupling products increased
as the reaction temperature increased. It can be considered
that the higher reaction temperature (473 K) would help
thermally activation steps in the reaction mechanism, so
that the higher amount of the products would be obtained.
On the other hand, when the photoreaction of methane
2
2
in the presence of CO was carried out at 473 K, the cou-
2
It was confirmed that no product could be detected
pling products increased and CO product was also obvi-
ously detected (Entry 4). The amounts of coupling
products obtained here were similar to those obtained from
when the reactions between methane and CO were carried
2
out at 473 K without UV irradiation (Entry 5), suggesting
that these reactions required photoenergy. Thus, it can be
proposed here that these reactions shown in Eqs. (7) and
the NOCM in the absence of CO (Entry 3). Since the
2
amount of produced H was much higher than the one cal-
2
(
8) would proceed photocatalytically over Ga O ,
2 3
culated from the obtained hydrocarbons products accord-
ing to the Eq. (2), the H would be produced not only
2
hm; D
0
¼ 70:7 kJ molꢀ1;
2
CH
4
ƒƒ! C
2
H
6
þ H
2
; DG
ð7Þ
via NOCM but also via other reactions. Even though we
could not clarify the detailed mechanism of these reactions
at the present, we consider that CO would be produced via
CO reforming as shown in Eq. (1). When the H was
473 K
Ga2O3
hm; D
0
¼ 126 kJ molꢀ1 ð8Þ
CH
4
þ CO
2
ƒƒ! 2CO þ 2H
2
; DG
473 K
Ga2O3
2
2
assumed to be produced via NOCM and CO reforming
As shown in Eqs. (7) and (8), these reactions are still up-hill
reactions and unfavorable even when employing mild reac-
tion temperatures such as 473 K. The catalyst used in these
reactions could be refreshed by the pretreatment procedure
at 1073 K to exhibit almost the same activity as the fresh
one, indicating that the photocatalyst is stable and
reusable.
2
of methane (DRM), a good material balance among the
products can be obtained, suggesting that these two reac-
tions, NOCM and DRM, would mainly proceed over
Ga O . To the best of our knowledge, this result would
2
3
be the first successful photocatalytic CO2 reforming of
methane to produce both CO and H2.
One might consider that the CO product was not pro-
duced via DRM but produced via different reactions, such
as a reverse water gas shift (RWGS) reaction,
In order to clarify the contribution of thermal energy,
the photoreactions were carried out at higher temperatures.
As the reaction temperatures increased, the amount of cou-
pling products (hydrocarbons) tended to decrease (Entries
CO þ H ! CO þ H O;
ð3Þ
2
2
2
4, 6, and 7), while the amount of reforming product (CO)
increased. Some consecutive coupling reactions, CO2
or Boudouard reaction,
reforming of higher hydrocarbons, and/or dehydrogena-
tion of hydrocarbons to form carbon species might pro-
C þ CO ! 2CO:
2
ð4Þ
In the separated experiment, we confirmed that RWGS
reaction did not occur under the reaction condition men-
tioned above (at 473 K under UV irradiation), thus, CO
product would not be produced via this route. Even though
deposited carbon species formed through decomposition of
ceed, leading to the formation of H and CO, but with a
2
slightly lower amount of hydrocarbons products.
Assuming that the amounts of CO product (x) during a
certain reaction time would be proportional to the rate
constant (k), the value of lnx can be plotted against

L. Yuliati et al. / Chemical Physics Letters 452 (2008) 178–182
181
be active for NOCM, such as alumina, silica–titania, and
silica-supported cerium oxides [3–5,10], also did not show
good photocatalytic activities for this reaction in the pres-
1
.5
0
673 K
ence of CO . Therefore, it can be suggested here that the
2
suitable photocatalyst for DRM should be able to activate
not only methane but also CO molecules at one time. In
2
the present study, Ga O was found to be one of the suit-
2
3
0
5
73 K
able photocatalysts for both NOCM and DRM. Further
investigation on the unique ability of Ga O to activate
2
3
both methane and CO would be necessary.
2
4
73 K
R = 0.98
4. Conclusions
Methane and carbon dioxide can be photocatalytically
0.0015
0.0018
0.0021
converted to higher hydrocarbons, CO and hydrogen via
mainly NOCM and DRM over Ga O photocatalyst.
-
1
1
/T (K )
2
3
While the photocatalytic NOCM could proceed at room
temperature, photocatalytic DRM required both photoen-
ergy and thermal energy such as a mild temperature of
Fig. 1. Pseudo-Arrhenius plot for CO
values were determined from the amounts of CO product (lmol/3 h).
2
reforming of methane (DRM). x
473 K. The mild external heating upon photoirradiation
would help the thermal steps with low activation energy
in the photocatalytic DRM, e.g., desorption of products
or migration of photoexcited electron. However, the main
activation step on this DRM would be promoted by
photoenergy.
reciprocal temperatures (1/T) as a pseudo-Arrhenius plot,
as shown in Fig. 1. Since the reactions were carried out
in a closed system, the rate constant (k) should be evalu-
ated from the amount of product at minimum reaction
time, which could provide the initial rate. However, due
to the detection limit, it is difficult to obtain a certain
amount of CO product when the reaction was carried out
at shorter time. Thus, the data after 3 h photoreaction were
tentatively employed in the present study. In separated
experiments, it was confirmed that CO was continuously
produced even after 24 h reaction under UV irradiation
at 473 K, although the reaction rate was not constant but
slightly decreased with time. This suggests that the data
after 3 h photoreaction are applicable for the analysis to
represent the initial rate of the reaction. From this primi-
Acknowledgements
The Ga K-edge XANES measurement was performed
under the approval of the Photon Factory Program Advi-
sory Committee (Proposal No. 2003G248). This work was
supported by Grant-in-Aid for Scientific Research on Pri-
ority Areas (No. 19028023, ‘Chemistry of Concerto Catal-
ysis’) from Ministry of Education, Culture, Sports, Science
and Technology, Japan.
tive study, the activation energy (E ) was estimated to be
a
ꢀ
1
References
ca. 10 kJ mol (at 473–673 K, data from Entries 4, 6,
and 7). In the case of thermal DRM, the reported activa-
ꢀ
1
[1] J.R.H. Ross, Catal. Today 100 (2005) 51.
tion energy was relatively high, such as ca. 178 kJ mol
[2] Y.H. Hu, E. Ruckenstein, Adv. Catal. 48 (2004) 297.
ꢀ
1
over CeO -doped Ni/Al O , 154 kJ mol over CeO , and
23 kJ mol over Pt/SrTiO catalysts [18–20]. Since the
3
2
2
3
2
[3] L. Yuliati, H. Itoh, H. Yoshida, Stud. Surf. Sci. Catal. 162 (2006)
961.
ꢀ
1
1
estimated Ea was much lower than that for a thermal
DRM, it can be suggested that the thermal energy was used
for the mild activation steps such as thermal desorption of
products or migration of photoexcited electron from the
bulk to the surface of Ga O . Since no CO product was
[4] L. Yuliati, H. Itoh, H. Yoshida, Stud. Surf. Sci. Catal. 172 (2007) 457.
[
[
5] Y. Kato, H. Yoshida, T. Hattori, Chem. Commun. (1998) 2389.
6] H. Yoshida, M.G. Chaskar, Y. Kato, T. Hattori, Chem. Commun.
(
2002) 2014.
[7] L. Yuliati, T. Hattori, H. Yoshida, Phys. Chem. Chem. Phys. 7 (2005)
2
3
195.
detected when the reaction was carried out in the dark even
at 673 K (Entry 8), the main activation step in the photore-
action at these reaction temperatures should be promoted
by photoenergy.
[8] Y. Kato, H. Yoshida, A. Satsuma, T. Hattori, Micro. Mesopor. 51
2002) 223.
9] L. Yuliati, M. Tsubota, A. Satsuma, H. Itoh, H. Yoshida, J. Catal.
38 (2006) 214.
(
[
2
[
10] L. Yuliati, T. Hamajima, T. Hattori, H. Yoshida, Chem. Commun.
In separated experiments, it was confirmed that the well-
(2005) 4824.
known TiO photocatalyst did not show activity for DRM,
[11] Y. Kohno, T. Tanaka, T. Funabiki, S. Yoshida, Phys. Chem. Chem.
Phys. 2 (2000) 5302.
2
even though some trace amounts of coupling products
could be observed. It suggests that not all semiconductors
could work as photocatalysts for the conversion of meth-
[12] K. Teramura, T. Tanaka, H. Ishikawa, Y. Kohno, T. Funabiki, J.
Phys. Chem. B 108 (2004) 346.
[13] D. Shi, Y. Feng, S. Zhong, Catal. Today 98 (2004) 505.
ane and CO . Some photocatalysts that were reported to
[14] A. Kudo, I. Mikami, J. Chem. Soc. Faraday Trans. 94 (1998) 2929.
2

182
L. Yuliati et al. / Chemical Physics Letters 452 (2008) 178–182
[
[
[
15] T. Yanagida, Y. Sakata, H. Imamura, Chem. Lett. 33 (2004)
26.
16] Y. Hou, X. Wang, L. Wu, Z. Ding, X. Fu, Environ. Sci. Technol. 40
2006) 5799.
17] K. Nishi et al., J. Phys. Chem. B 102 (1998) 10190.
[18] N. Laosiripojana, W. Sutthisripok, S. Assabumrungrat, Chem. Eng.
J. 112 (2005) 13.
[19] N. Laosiripojana, S. Assabumrungrat, Appl. Catal. B 60 (2005).
[20] A. Topalidis, D.E. Petrakis, A. Ladavos, L. Loukatzikou, P.J.
Pomonis, Catal. Today 127 (2007) 238.
7
(