Photocatalytic activity of metal oxide supported gallium oxide for CO2 reduction with water

Article abstract of DOI:10.1246/BCSJ.20190366

The photocatalytic activities of Ga₂O₃ supported by various metal oxides for H₂O splitting and CO₂ reduction with water under UV light irradiation were investigated. The Ga₂O₃ photocatalyst supported by metal oxides was prepared via an impregnation method. Some oxides used as a supporting oxide of Ga₂O₃ photocatalyst enhanced activities of H₂O splitting and CO₂ reduction. Specifically, the CO production rates were well correlated to various properties of the supporting oxide. These oxides satisfied the three criteria greatly improving the CO production rate. First, their band gap was wider than that of Ga₂O₃ and the position of its conduction band was at a negative potential with respect to that of Ga₂O₃. Second, they were not miscible with Ga₂O₃ and third, mixed phases of α and γ or defective γ phase of Ga₂O₃ were well dispersed on them exhibiting columnar shape.

Full text of DOI:10.1246/BCSJ.20190366

Document type: Article
Photocatalytic Activity of Metal Oxide Supported Gallium Oxide
for CO2 Reduction with Water
Ryota Ito,1 Masato Akatsuka,1 Akiyo Ozawa,1,2 Muneaki Yamamoto,3
Tetsuo Tanabe,3 and Tomoko Yoshida*3
1Applied Chemistry and Bioengineering, Graduate School of Engineering, Osaka City University,
3-3-138 Sugimoto, Sumiyoshi-ku, Osaka 558-8585, Japan
2Corporate Research Laboratories, Research & Development Division, Sakai Chemical Industry, Co., Ltd.,
5-2 Ebisujima-cho, Sakai-ku, Sakai, Osaka 590-0815, Japan
3Advanced Research Institute for Natural Science and Technology, Osaka City University,
3-3-138 Sugimoto, Sumiyoshi-ku, Osaka 558-8585, Japan
E-mail: tyoshida@ocarina.osaka-cu.ac.jp
Received: December 24, 2019; Accepted: March 10, 2020; Web Released: May 15, 2020
Tomoko Yoshida
Tomoko Yoshida received her PhD from Kyoto University (1996) under the supervision of Professor Satohiro
Yoshida. In 1996, she joined Nagoya University as an Assistant Professor, and she was an Associate
Professor during 2003–2014. Moving to Osaka City University in 2015, she was promoted to a Professor.
Her research interests include CO2 reduction using semiconductor photocatalysts and establishment of a
new structural and electronic state analysis using synchrotron radiation spectroscopy.
Abstract
water using semiconductor photocatalysts has been widely
studied,4-15 and various photocatalysts have been developed.
Recently Ga2O3 has been used as a photocatalyst both for CO2
reduction and water splitting, and various efforts have been
made to improve its photocatalytic activity. Yamamoto et al.
have reported that Ag cocatalyst improves photocatalytic
activity of Ga2O3 for CO2 reduction under UV light irradiation
with water.9,10 Teramura et al. have succeeded in improving the
photocatalytic activity for CO2 reduction by cation (Zn, Pr,
or Yb) doping into Ga2O3 with using Ag as a cocatalyst.11-13
However, the role of the Ag cocatalyst on Ga2O3 for the
photocatalytic CO2 reduction has not been well understood. In
addition, it has been reported that Ga2O3 consisting of two
different phases such as β and γ showed high photocatalytic
activity without the Ag cocatalyst owing to possible promotion
of charge separation between the two phases.15-17
In a previous study, we successfully improved the photo-
catalytic activity of Ga2O3 for CO2 reduction and water split-
ting under UV light irradiation by using γ-Al2O3 as a catalyst
support without the Ag cocatalyst.18 Ga2O3 supported by Al2O3
consisted of two different phases (α and γ phases) and its
particle size changed with changing amounts of Ga2O3 loaded
on Al2O3. Thus, the control of the crystal phases and particle
sizes of Ga2O3 is likely a key factor for the improvement of its
photocatalytic activity.
We have examined photocatalytic activities of Ga2O3 sup-
ported by various metal oxides for water splitting and CO2
reduction with water under UV light irradiation. The Ga2O3
photocatalyst supported by metal oxides was prepared by an
impregnation method. We have found that some oxides used as
a supporting oxide of Ga2O3 photocatalyst improve activities
of water splitting and CO2 reduction. In particular, the CO
production rates are well correlated to various properties of the
supporting oxide. These oxides satisfy the following three
criteria significantly enhancing the CO production rate. (1)
Their band gap is wider than that of Ga2O3 and the position of
its conduction band is at a negative potential with respect to
that of Ga2O3. (2) They are not miscible with Ga2O3 and (3)
mixed phases of α and γ or defective γ phase of Ga2O3 are well
dispersed on them exhibiting columnar shape. These criteria
would be useful for further improvement of the activity for
photocatalytic CO2 reduction with water by Ga2O3.
Keywords: Ga2O3 supported by metal oxide
j
Band gap and position of conduction band
Dispersion of Ga2O3 particles
j
1. Introduction
Recently, artificial photosynthesis has attracted much atten-
tion for resolving global warming and shortages of energy
fuel.1-3 As artificial photosynthesis, reduction of CO2 with
These observations motivated us to change the supporting
oxide from Al2O3 to other oxides for further improvement of
694 | Bull. Chem. Soc. Jpn. 2020, 93, 694–700 | doi:10.1246/bcsj.20190366
© 2020 The Chemical Society of Japan

Table 1. The time courses of CO, H2 and O2 production rates for all samples showing O2 evolution.
electron hole
balance
1 h
2 h
3 h
4 h
5 h
Samples
CO H2 O2 CO H2 O2 CO H2 O2 CO H2 O2 CO H2 O2 (after 5 reaction)
Ga⒬O⒭/La⒬O⒭
Ga⒬O⒭/Al⒬O⒭
Ga⒬O⒭/Gd⒬O⒭
Ga⒬O⒭/Yb⒬O⒭
Ga⒬O⒭/Nd⒬O⒭
Ga⒬O⒭
1.9 10.1 4.4 4.8 17.6 6.9 6.9 22.9 9.9 7.4 25.6 13.1 7.1 26.8 14.9
2.9 23.6 12.5 3.8 27.9 15.8 4.2 32.4 18.8 4.4 33.5 19.5 4.4 34.2 19.9
3.0 28.7 15.1 3.4 38.9 21.1 3.6 43.3 23.7 3.7 45.5 25.1 3.5 45.9 25.1
2.2 15.0 7.6 2.1 15.2 7.8 2.2 15.9 8.3 2.4 17.8 9.3 2.6 19.9 10.4
1.6 10.9 2.8 0.9 9.0 4.7 1.9 13.8 5.6 2.0 13.7 6.2 2.1 14.0 5.9
1.0 10.0 4.5 1.5 11.5 6.0 1.7 11.6 6.3 1.9 12.5 6.8 1.6 13.0 6.9
0.4 6.2 1.6 0.3 6.8 2.8 0.3 7.2 3.2 0.3 7.3 3.3 0.3 7.4 3.4
0.2 7.8 2.4 0.2 7.5 3.0 0.2 7.8 3.2 0.2 8.7 3.7 0.2 10.8 4.6
1.02
0.97
0.98
1.08
1.03
1.06
1.14
1.18
Ga⒬O⒭/ZSM-5
Ga⒬O⒭/ZrO⒬
the photocatalytic activity of Ga2O3. As the supporting oxide,
we have selected various oxides having varieties of features in
band gap widths, the position of conduction band, valency of
cations and so on. Selected oxides as the support were MgO,
Al2O3, TiO2, Cr2O3, MnO2, Fe2O3, ZnO, Y2O3, ZrO2, In2O3,
La2O3, CeO2, Nd2O3, Gd2O3, Yb2O3 and ZSM-5 type zeolite.
We have tried to find correlations between photocatalytic
activities of Ga2O3 supported by these supporting oxides and
their characteristics such as band structures and crystalline
structures.
(2) a Pt wire counter electrode, and (3) an Ag/AgCl reference
electrode, and an aqueous Na2SO4 solution (0.5 M, pH = 7.4).
The flat band potential (E0) of the oxide at pH 0 vs SHE
(standard hydrogen electrode) was extrapolated from the
observed potential (E) using the relation of E = E0-0.059pH
and the electrode potential of Ag/AgCl was set to be +0.199 V
(vs. SHE, 25 °C).
Crystalline structures of all Ga2O3/metal oxide samples
were examined by X-ray diffraction (XRD) analysis. XRD pat-
terns of the samples were recorded on a MiniFlex600 (Rigaku)
using Cu Kα as a radiation source with an operating voltage of
40 kV and current of 15 mA. The XRD patterns were collected
at 2θ angles of 20-70°. The 2θ step width was 0.02°, and the
scanning rate was 2°/min.
2. Experimental
2.1 Preparation of Ga2O3/Metal Oxide Photocatalyst
Samples.
Ga2O3 supported by metal oxides (referred as
Ga2O3/metal oxide) were prepared by an impregnation meth-
od.18 In brief, Ga(NO3)3¢8H2O (Kishida Chemical Co. Ltd.
purity 99.0%) and one of the supporting oxides listed in
Table 1 were put in 200 mL of distilled water and stirred with
a magnetic stirrer in air and dried, followed by calcination at
823 K for 4 h. The amount of Ga2O3 supported on a metal oxide
was set to be 40 wt%. Unsupported Ga2O3 samples having
single phase either α-Ga2O3 or γ-Ga2O3 were also prepared for
comparison. α-Ga2O3 was obtained by calcining Ga2O3¢nH2O
(Mitsuwa Chemicals Co., Lid Ga2O3 78.8%) at 823 K for 5 h.19
For the preparation of γ-Ga2O3, Ga(NO3)3¢8H2O was dissolved
in ethanol (approximately 3 g of the reagent in 50 mL of the
solvent), and then an ethanol solution of 28 vol % aqueous
ammonia (volume ratio of ethanol/aqueous ammonia = 1) was
added slowly under continuous stirring at room temperature
until no further precipitates were formed. The resultant precip-
itates were filtered, rinsed in ethanol, and vacuum-dried in a
desiccator. Finally, the obtained solid was calcined at 773 K for
5 h to obtain γ-Ga2O3.20
Ga K-edge EXAFS was measured with the beam line BL5S1
at Aichi Synchrotron Radiation Center. Ga K-edge EXAFS
was obtained by a transmission mode. Morphology of the sam-
ples was observed by a field emission scanning electron micro-
scope (FE-SEM, JSM-6500F, JEOL Ltd.) with a backscattered
electron mode under an acceleration voltage of 15 kV.
2.3 Photocatalytic CO2 Reduction and Water Splitting
Tests. Photocatalytic CO2 reduction with H2O under UV light
irradiation was tested for 0.1 g of each of the Ga2O3/metal
oxide samples set in a fixed-bed flow reactor cell under CO2
gas flow. The UV light intensity was 40 mW/cm2 in the range
of 254 « 10 nm. Before the test, the sample was irradiated with
a 300 W Xe lamp for 1 h under CO2 gas flow with the flow rate
of 20.0 mL/min. The reduction test was conducted as follows.
First, a 10.0 mL NaHCO3 aqueous solution24 (1.0 M) with a
sample was prepared. The pH of the solution was 8.7. Then,
CO2 gas was introduced with a flow rate of 3.0 mL/min under
the UV light irradiation. This made the pH a little lower being
7.8, which was maintained during the reaction tests. After
stopping the UV light and CO2 gas flow, pH increased a little
to 8.4. The reaction products (CO, H2, and O2) were analyzed
by a gas chromatograph equipped with a thermal conductivity
detector.
2.2 Characterization. UV-vis diffuse reflectance spec-
tra were recorded at room temperature using a spectrometer
(JASCO V-670). The spectrum of Ba2SO4 was used as
reference.
The potential position of the conduction band of the support-
ing oxides was determined by the Mott-Schottky-plot from the
flat band potential measured with a three-electrode chemical
cell (VersaSTAT4 (AMETEK)).21-23 The following three elec-
trodes were used. (1) a metal oxide dispersed on a fluorine-
doped tin oxide substrate followed by calcination at 823 K,
3. Results and Discussion
3.1 Photocatalytic Activity of Ga2O3 Supported by Metal
Oxides. Figure 1 and 2 respectively show the production rates
of CO and H2 observed in the photocatalytic CO2 reduction and
water splitting tests for all samples after 5 h of the test. The
Bull. Chem. Soc. Jpn. 2020, 93, 694–700 | doi:10.1246/bcsj.20190366
© 2020 The Chemical Society of Japan | 695

Figure 1. Production rates of CO in the photocatalytic reduction of CO2 with H2O for Ga2O3 supported by various metal oxides
indicated in the figure after 5 h of reaction. The inset is magnified with about 40 times the production rates.
Figure 2. Production rates of H2 in the photocatalytic reduction of CO2 with H2O for Ga2O3 supported by various metal oxides
indicated in the figure after 5 h of reaction. The inset is magnified with about 4 times the production rates.
results for α-Ga2O3 and γ-Ga2O3 are not shown because the
single phase of either α-Ga2O3 or γ-Ga2O3 hardly shows
photocatalytic activity.18 In the figure, supporting oxides are
ordered from left to right corresponding to the CO production
rates from higher to lower. Production rates of all products
(CO, H2 and O2) and selectivity toward CO evolution are
shown in Figures S1 and S2, corresponding to the results for
higher active and lower active oxides shown in Figures 1 and 2
respectively. And the time courses of the production rates of
CO, H2 and O2 and stoichiometric ratio for those oxide produc-
ing O2 are shown in Table 1. Figure 3 is an example of the time
courses of the production rates for Ga2O3/Al2O3.
In addition, we have conducted reduction tests under He
flow instead of CO2 for samples giving higher activity (Ga2O3/
La2O3 and Ga2O3/Al2O3) to confirm that CO originates from
CO2. The results are shown in Figure S3. Under the He flow,
CO production was hardly seen. Thus, we can conclude that
Figure 3. Time course of CO (black circle), H2 (gray circle)
CO and H2 were produced with the photocatalytic CO2 reduc-
and O2 (white circle) production rates for the sample
tion with water.
(Ga2O3/Al2O3).
As depicted in Figures 1 and 2, the production rates of both
CO and H2 separate the samples roughly into two groups, one
showing higher production rates than that of non-supported
Ga2O3 and the other lower production rates which were magni-
fied in the inset of the figures. Hereafter, Ga2O3/metal oxides
showing higher CO production rates are referred as higher
active samples and the others as lower active samples. Never-
696 | Bull. Chem. Soc. Jpn. 2020, 93, 694–700 | doi:10.1246/bcsj.20190366
© 2020 The Chemical Society of Japan

Table 2. Characteristics of all oxides used as the support; valency of metal ions, band gap determined by UV-
Vis absorption, conduction band position determined by the Mott-Schottky plot. Crystalline structures of
Ga2O3 supported by these oxides and production rates of CO2 and H2 after 5 hours of the reduction test.
Figure 5. Diffuse reflectance UV-vis spectra of oxides used
as a support for Ga2O3 catalysts showing lower activity,
(a) MnO2, (b) Fe2O3, (c) Cr2O3, (d) In2O3, (e) CeO2
(f) ZnO, (g) TiO2, (h) ZrO2, (i) MgO and (j) ZSM-5.
Figure 4. Diffuse reflectance UV-vis spectra of oxides used
as a support for Ga2O3 catalysts showing higher activity,
(a) Nd2O3, (b) Yb2O3, (c) Gd2O3, (d) La2O3, (e) Y2O3, and
(f) Al2O3.
theless, the order of the oxides showing higher H2 production
rates in Figure 2 are totally different from that of Figure 1. In
Table 2, observed production rates of CO and H2 for all sam-
ples are also given together with characteristics of the support-
ing oxides. The separation is also seen Table 2.
In contrast, those oxides showing optical absorption in longer
wavelength region reduced the CO production rate. The band
gap value calculated from these spectra is shown in Table 2. As
additional information, diffuse reflectance UV-vis spectra of
metal oxide supported Ga2O3 are shown in Figure S4 and S5.
Samples that showed water splitting and/or CO2 reduction
show similar spectra to that of Ga2O3.
3.2 Diffuse Reflectance UV-Vis Spectra.
Diffuse
reflectance UV-vis spectra of all supporting oxides are given
in Figures 4 and 5 with separation of higher and lower active
samples when used as the supporting oxide. One can clearly see
that those oxides showing absorption edge at shorter wave-
length than that of Ga2O3 result in higher CO production rates.
3.3 Mott-Schottky-Plot.
From the flat band potential
obtained by Mott-Schottky-plot, the potential position of the
conduction band was determined as given in Table 2. All
supporting oxides having their conduction band at potential
Bull. Chem. Soc. Jpn. 2020, 93, 694–700 | doi:10.1246/bcsj.20190366
© 2020 The Chemical Society of Japan | 697

Figure 7. FE-SEM images of (a) Ga2O3/La2O3 (b) Ga2O3/
Al2O3 (c) Ga2O3/MgO (d) Ga2O3/ZrO2.
of the oxides showing higher CO production rates together with
those of α-Ga2O3 and γ-Ga2O3, (Figure 6(a)) and lower ones
(Figure 6(b)).
EXAFS spectra of α-Ga2O3 and γ-Ga2O3 prepared as the
reference are quite similar to those previously observed for
well crystalized α-Ga2O3 phase and defective γ-Ga2O3 phase,
respectively.18 This is consistent with the results of XRD
analysis.
As indicated in the figures, RSFs of all samples show two
dominant peaks characteristic to Ga2O3. The first peak appear-
ing at around 1.5 ¡ is assigned to the backscattering from an
adjacent oxygen atom to a Ga atom (Ga-O bond), and the
second peak around 2.7 ¡ indicates the presence of the second-
neighboring gallium atoms (Ga-(O)-Ga bond).28-30 It should
be noted in Figure 6(a) that α-Ga2O3 and γ-Ga2O3 show two
characteristic shoulders around 3.5 ¡ and 4.5 ¡, and 0.9 ¡ and
2.1 ¡, respectively, which are indicated as white arrows and
black arrows, respectively. Referring to them, higher active
samples given in Figure 6(a) show either or both of these small
peaks, i.e. Ga2O3/Nd2O3 and Ga2O3/Gd2O3 (b, c) have peaks
like α-Ga2O3, and Ga2O3/Al2O3 and Ga2O3/La2O3 (d, e) have
peaks like defective γ-Ga2O3. In Figure 6(b), the peaks derived
from Ga2O3 in Ga2O3/MgO (f) and Ga2O3/ZnO (k) are signif-
icantly shifted, which well corresponds to the formation of
composite oxides observed by the XRD analysis. In the other
samples of Figure 6(b), characteristic peaks of α and γ were not
appreciable.
Figure 6. Fourier transforms of EXAFS spectra for metal
oxide supported Ga2O3, α-Ga2O3, γ-Ga2O3, (a) Ga2O3/
Y2O3 (b) Ga2O3/Nd2O3 (c) Ga2O3/Gd2O3 (d) Ga2O3/
Al2O3 (e) Ga2O3/La2O3 (f) Ga2O3/MgO (g) Ga2O3/CeO2
(h) Ga2O3/ZrO2 (i) Ga2O3/TiO2 (j) Ga2O3/ZSM-5 (k)
Ga2O3/ZnO. White arrow indicated the characteristic peak
of α-Ga2O3 and black arrows indicated those of γ-Ga2O3.
negative to that of Ga2O3 except MgO gave higher production
rates, while those supporting oxides having the conduction
band at a more positive position reduced the production rate.
3.4 XRD Measurements. Results of XRD analysis of all
samples are also given in Table 2. In terms of crystalline struc-
ture, the supporting oxides are separated into three groups; (1)
supporting oxides miscible with Ga2O3 or coexisting with
Ga2O3, (2) those dissolving some Ga2O3 and vice versa, and
(3) those forming composite oxides with Ga2O3 or forming
GaOOH, which hardly showed photocatalytic activity. In most
of the samples showing higher CO and H2 production rates,
Ga2O3 on the supporting oxides exhibited single or mixed
phases of α and γ phases of Ga2O3.20,25 The XRD analysis of
α-Ga2O3 and γ-Ga2O3 prepared as the references showed that
the former was well crystalized while the latter was defective,
which is quite consistent with previous work.26,27
3.6 FE-SEM Observation.
Figure 7 shows images of
selected samples taken with a field emission scanning electron
microscope (FE-SEM) in a backscattered electron mode, of
which contrast well corresponds to the mass of constituent
atoms of the samples.
On Ga2O3/La2O3 (a) and Ga2O3/Al2O3 (b), Ga2O3 particles
showing columnar shape with their length of around 200-
500 nm and diameter of a few nanometer are clearly distin-
guished from supporting metal oxides of La2O3 and Al2O3 with
micrometer sizes. In addition, Ga2O3 particles were well
dispersed. The particle sizes of Ga2O3 on La2O3 were a little
larger than those on Al2O3, which would be the cause of lower
3.5 XAFS Measurements. Figure 6 shows radial structure
functions (RSFs) obtained from Fourier transformation of
¹1
EXAFS spectra in the range from 3 to 12 ¡ with separation
698 | Bull. Chem. Soc. Jpn. 2020, 93, 694–700 | doi:10.1246/bcsj.20190366
© 2020 The Chemical Society of Japan

specific surface area of Ga2O3/La2O3. All other higher active
samples also showed the columnar shaped Ga2O3 particles. (not
shown here).
CO production rate without losing electron energy as clearly
seen in Table 2.
4.2 Modification of the Chemical Nature of Ga2O3. As
depicted in Table 2, in all supporting oxides resulting in higher
CO production rate, the valency of their cation is either +2 or
+4, different from that of Ga (+3). The different coordination
number would cause instability in Ga2O3 and also make the
boundary between Ga2O3 and the support oxide unstable result-
ing in insufficient electric junction. MgO and ZnO reacted with
Ga2O3 to form composite oxides. On ZSM-5, TiO2 and CeO2
Ga2O3 turned to GaOOH. Such changes in surface chemical
nature caused the lower CO production rate.
It should be mentioned that without Ga2O3, both La2O3 and
Al2O3, which showed high activity when they were used as a
support, hardly showed catalytic activity, as shown in
Figure S6 in supporting information. This clearly indicates
that the active site of CO2 reduction is on Ga2O3 particles but
not supporting oxides or other newly formed compounds.
4.3 Microstructure of Ga2O3. As depicted in Table 2, the
crystalline structure of Ga2O3 in all samples showing higher
CO production rate consisted of either α and γ phase or their
mixture making columnar shaped small particles. This is con-
sistent with our previous finding15 that on photocatalytic reduc-
tion using Ga2O3 without support oxides, the mixture of two
different phases such as β and γ phases or defective γ phase of
Ga2O3 shows high activity. Considering that without a support-
ing oxide defective γ-Ga2O3 alone did not show high activity,18
defective γ-Ga2O3 on supporting oxides would be further
modified. Possible modifications are distortion of crystalline
structure caused by mismatch to that of the supporting oxide,
dissolution of cations of the supporting oxide, appearance of
boundaries between the Ga2O3 particles and the supporting
oxides including the formation of a heterojunction. Another
important factor is dispersion of Ga2O3 without segregation as
seen in Figure 6.
On MgO (Figure 7(c)), no columnar shaped particles were
observed because MgO and Ga2O3 reacted to form a composite
oxide. Although on ZrO2 (Figure 7(d)) are seen the columnar
shaped Ga2O3 particles, they seem aggregated and less dis-
persed. As discussed later, the columnar shaped Ga2O3 particles
are very likely playing a key role in the photocatalysis. In this
respect, crystal structures of the supporting oxides seem to play
an important role. And, a valency of cations in the supporting
oxides of three is a critical factor to increase the CO production
rates as shown in Table 2. The corresponding oxides show
layered structure like perovskite or its modified type as Ga2O3,
while most of the other oxides showing lower CO2 production
rates are cubic or tetragonal. This suggests that on the latter
oxides, Ga2O3 is rather free to aggregate as larger particles,
while on the former oxides, their lattice similarity to Ga2O3
would inhibit free growth of Ga2O3 or dissolution of their
cations into Ga2O3. The cation dissolution could introduce
lattice distortion that appeared as defective γ-phase and inhibits
the growth of Ga2O3 particles.
4. Discussion
As seen in Figures 1 and 2 and Table 2, the supporting
oxides can be roughly divided into two groups, one giving
higher activity and the other giving lower activity than the
activity of non-supported mixed phased Ga2O3.
Although the band gap energies of those supporting oxides
resulting in higher production rates both for CO and H2 are
larger than 4 eV, no correlation appears between the production
rates of CO and H2, suggesting both reactions proceeded sepa-
rately as already observed in our previous work on Ga2O3/
Al2O3.18 Furthermore, only the CO production rates can be
correlated to various characteristics of the supporting oxides as
discussed in the following. Therefore, the following discussion
is focused to correlate the CO production rate to (1) material
properties of the supporting oxide, (2) modifications of chemi-
cal nature and (3) microstructure or crystalline structure of
Ga2O3.
In summary, the present research clearly shows following,
(1) The active sites for the photocatalytic CO2 reduction with
water on Ga2O3 supported by metal oxides are on columnar
shaped Ga2O3 particles which are well dispersed on the sup-
porting oxide and exhibited mixed phase of α and γ or defective
γ phase.
4.1 Material Properties. There are two properties of the
supporting oxides differentiating the changes of the CO pro-
duction rate:
(2) Any changes of chemical nature in Ga2O3 given by the
supporting oxides reduced the CO production rate.
(1) Those supporting oxides having wider band gap than that of
Ga2O3 showed higher CO production rates than that of non-
supported Ga2O3. This indicates that there is some threshold in
energy of photo-excited electrons for the CO2 reduction which
is very near the bottom of the conduction band of Ga2O3.
Excited photoelectrons in the wider band gap oxides could be
transferred to the conduction band of Ga2O3 and accordingly
enhance the reduction.
(2) Those oxides having their conduction band more positive
than that of Ga2O3 reduced the CO production rate. This clearly
correlates to the band gap, because photoexcited electrons in
the conduction band of Ga2O3 would easily lose their energy
through transition to the conduction band of supporting oxides
more positive than Ga2O3 and are not used for CO2 reduction.
Therefore, those oxides of which the conduction band position
is lower than ¹1.03 eV (that of Ga2O3) resulted in the higher
(3) In order to promote the photocatalytic activity on CO2
reduction with water, there seems to be a threshold in energy
of photoexcited electrons, which is near the bottom of the
conduction band of Ga2O3 or a band position of supporting
oxides more negative than that of Ga2O3.
5. Conclusion
We have examined photocatalytic activities of Ga2O3 sup-
ported by various metal oxides (MgO, Al2O3, TiO2, Cr2O3,
MnO2, Fe2O3, ZnO, Y2O3, ZrO2, La2O3, CeO2, Nd2O3, Gd2O3,
Yb2O3, ZSM-5) for CO2 reduction with water under UV light
irradiation without a noble metal cocatalyst and compared their
catalytic activities in terms of band gap widths, crystalline
structures and other characteristics of the supporting oxides.
Both water splitting and CO production proceeded simultane-
ously but independently at different active sites with each other.
Bull. Chem. Soc. Jpn. 2020, 93, 694–700 | doi:10.1246/bcsj.20190366
© 2020 The Chemical Society of Japan | 699

Yagi, e-J. Surf. Sci. Nanotechnol. 2014, 12, 299.
10 M. Yamamoto, T. Yoshida, N. Yamamoto, T. Nomoto, S.
Yagi, Nucl. Instrum. Methods Phys. Res. B 2015, 359, 64.
11 H. Tatsumi, K. Teramura, Z. Huang, Z. Wang, H. Asakura,
S. Hosokawa, T. Tanaka, Langmuir 2017, 33, 13929.
12 Z. Huang, K. Teramura, H. Asakura, S. Hosokawa, T.
Tanaka, J. Mater. Chem. A 2013, 36, 19351.
13 Z. Wang, K. Teramura, Z. Huang, S. Hosokawa, Y. Sakatad,
T. Tanaka, Catal. Sci. Technol. 2016, 6, 1025.
14 A. Iwase, S. Yoshino, T. Takayama, Y. H. Ng, R. Amal, A.
Kudo, J. Am. Chem. Soc. 2016, 138, 10260.
15 Y. Kawaguchi, M. Yamamoto, A. Ozawa, Y. Kato, T.
Yoshida, Surf. Interface Anal. 2019, 51, 79.
16 X. Wang, Q. Xu, M. Li, S. Shen, X. Wang, Y. Wang, Z.
Feng, J. Shi, H. Han, C. Li, Angew. Chem., Int. Ed. 2012, 51,
13089.
CO production rates for all Ga2O3/metal oxides are well
correlated to various properties of supporting oxide as there are
four main factors increasing the CO production rate: (1) wider
band gap than Ga2O3 (2) more negative conduction band
position than Ga2O3 (3) columnar shaped Ga2O3 particles
consisting of mixed phases of α and γ or defective γ phase well
dispersed on supporting oxide. (4) the valency is three for the
cation of supporting oxides.
From all results, we can conclude that:
(i) The active sites for the CO production are on columnar
shaped Ga2O3 particles, which are well dispersed on the sup-
porting oxide and consist of mixed phases of α and γ or
defective γ phase. Although the boundary between the Ga2O3
particles and the supporting oxides could be the possible active
sites, their area would not be large enough.
(ii) To allow the formation of the nano to sub-micrometer sized
Ga2O3 particles with the columnar shape on the support-
ing oxides, crystalline structure and valency of the cation in the
supporting oxides are quite important.
(iii) There is a threshold in energy of photoexcited electrons to
promote CO2 reduction with water which is close to the band
gap energy of Ga2O3.
17 S. Jin, X. Wang, X. Wang, M. Ju, S. Shen, W. Liang, Y.
Zhao, Z. Feng, H. Y. Playford, R. I. Walton, C. Li, J. Phys. Chem.
C 2015, 119, 18221.
18 R. Ito, M. Akatsuka, A. Ozawa, Y. Kato, Y. Kawaguchi, M.
Yamamoto, T. Tanabe, T. Yoshida, ACS Omega 2019, 4, 5451.
19 K. Shimizu, M. Takamatsu, K. Nishi, H. Yoshida, A.
Satsuma, T. Hattori, Chem. Commun. 1996, 1827.
20 Y. Hou, L. Wu, X. Wang, Z. Ding, Z. Li, X. Fu, J. Catal.
2007, 250, 12.
Supporting Information
21 J. Boltersdorf, B. Zoellner, C. M. Fancher, J. L. Jones, P. A.
Maggard, J. Phys. Chem. C 2016, 120, 19175.
Diffuse reflectance UV-vis spectra of metal oxide supported
Ga2O3 samples and results of their photocatalytic reduction
of CO2 or He with H2O are given in supporting information.
This material is available on https://dx.doi.org/10.1246/bcsj.
20190366.
22 D. Kato, K. Hongo, R. Maezono, M. Higashi, H. Kunioku,
M. Yabuuchi, H. Suzuki, H. Okajima, C. Zhong, K. Nakano, R.
Abe, H. Kageyama, J. Am. Chem. Soc. 2017, 139, 18725.
23 X. Qi, G. She, X. Huang, T. Zhang, H. Wang, L. Mu, W.
Shi, Nanoscale 2014, 6, 3182.
24 K. Teramura, K. Hori, Y. Terao, Z. Huang, S. Iguchi, Z.
Wang, H. Asakura, S. Hosokawa, T. Tanaka, J. Phys. Chem. C
2017, 121, 8711.
References
1
A. Kubacka, M. Fernandez-García, G. Colon, Chem. Rev.
2012, 112, 1555.
2
D. Kim, K. K. Sakimoto, D. Hong, P. Yang, Angew. Chem.,
25 Q. Liu, D. Guo, K. Chen, Y. Su, S. Wang, P. Li, W. Tang,
J. Alloys Compd. 2018, 731, 1225.
Int. Ed. 2015, 54, 3259.
3
L. Yu, G. Li, X. Zhang, X. Ba, G. Shi, Y. Li, P. K. Wong,
26 M. Higashiwaki, K. Sasaki, H. Murakami, Y. Kumagai, A.
Koukitu, A. Kuramata, T. Masui, S. Yamakoshi, Semicond. Sci.
Technol. 2016, 31, 034001.
27 L. Li, B. Ma, H. Xie, M. Yue, R. Cong, W. Gao, T. Yang,
RSC Adv. 2016, 6, 59450.
28 Y. Kato, M. Yamamoto, A. Ozawa, Y. Kawaguchi, A.
Miyoshi, T. Oshima, K. Maeda, T. Yoshida, e-J. Surf. Sci.
Nanotechnol. 2018, 16, 262.
J. C. Yu, Y. Yu, ACS Catal. 2016, 6, 6444.
4
K. Iizuka, T. Wato, Y. Miseki, K. Saito, A. Kudo, J. Am.
Chem. Soc. 2011, 133, 20863.
T. Takayama, K. Tanabe, K. Saito, A. Iwase, A. Kudo,
Phys. Chem. Chem. Phys. 2014, 16, 24417.
K. Li, A. D. Handoko, M. Khraisheh, J. Tang, Nanoscale,
2014, 6, 9767.
5
6
7
Ş. Neatu, J. A. Macia’-Agullo’, P. Concepcio’n, H. Garcia,
29 M. Akatsuka, T. Yoshida, N. Yamamoto, M. Yamamoto, S.
Ogawa, S. Yagi, J. Phys.: Conf. Ser. 2016, 712, 012056.
30 K. Nishi, K. Shimizu, M. Takamatsu, H. Yoshida, A.
Satsuma, T. Tanaka, S. Yoshida, T. Hattori, J. Phys. Chem. B 1998,
102, 10190.
J. Am. Chem. Soc. 2014, 136, 15969.
8
H. Tsuneoka, K. Teramura, T. Shishido, T. Tanaka, J. Phys.
Chem. C 2010, 114, 8892.
9
M. Yamamoto, T. Yoshida, N. Yamamoto, H. Yoshida, S.
700 | Bull. Chem. Soc. Jpn. 2020, 93, 694–700 | doi:10.1246/bcsj.20190366
© 2020 The Chemical Society of Japan