Visible light sensitive photocatalyst, delafossite structured α-AgGaO2

Article abstract of DOI:10.1021/jp063406s

Delafossite structured α-AgGaO₂ powder was successfully synthesized through a cation exchange reaction. α-AgGaO₂ has a band gap of 2.4 eV, absorbs visible light up to 520 nm, and effectively decomposes 2-propanol to CO₂ via acetone by irradiating with either UV light (300-400 nm) or visible light (420-530 nm). The values of the quantum efficiency are similar (ca. 0.6%) under light irradiations with wavelengths of 365, 390, 430, 470, and 510 ± 10 nm, but steeply decrease with wavelengths longer than 530 ± 10 nm, which support a 2.4 eV band gap. In contrast, the other polymorph, β-AgGaO₂ powder, which has a band gap of 2.1 eV, shows a negligible activity when irradiating with either UV light or visible light. The higher oxidation activity of α-AgGaO ₂ is probably due to its larger band gap, which is formed at the top of its valence band in a lower energy region as compared to β-AgGaO ₂. Moreover, the first-principle calculations of α-AgGaO ₂ and β-AgGaO₂ clearly indicate that α-AgGaO₂ has a remarkably larger dispersed valence band as compared to β-AgGaO₂, which is advantageous to the photocatalytic activity due to the efficient hole conduction.

Full text of DOI:10.1021/jp063406s

23274
J. Phys. Chem. B 2006, 110, 23274-23278
Visible Light Sensitive Photocatalyst, Delafossite Structured r-AgGaO2
Yoshihiko Maruyama,† Hiroshi Irie,*,† and Kazuhito Hashimoto*,†,‡
Department of Applied Chemistry, The UniVersity of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan,
and Research Center for AdVanced Science and Technology (RCAST), The UniVersity of Tokyo, 4-6-1 Komaba,
Meguro-ku, Tokyo 153-8904, Japan
ReceiVed: June 1, 2006; In Final Form: August 17, 2006
Delafossite structured R-AgGaO2 powder was successfully synthesized through a cation exchange reaction.
R-AgGaO2 has a band gap of 2.4 eV, absorbs visible light up to 520 nm, and effectively decomposes 2-propanol
to CO2 via acetone by irradiating with either UV light (300-400 nm) or visible light (420-530 nm). The
values of the quantum efficiency are similar (ca. 0.6%) under light irradiations with wavelengths of 365, 390,
430, 470, and 510 ( 10 nm, but steeply decrease with wavelengths longer than 530 ( 10 nm, which support
a 2.4 eV band gap. In contrast, the other polymorph, â-AgGaO2 powder, which has a band gap of 2.1 eV,
shows a negligible activity when irradiating with either UV light or visible light. The higher oxidation activity
of R-AgGaO2 is probably due to its larger band gap, which is formed at the top of its valence band in a lower
energy region as compared to â-AgGaO2. Moreover, the first-principle calculations of R-AgGaO2 and
â-AgGaO2 clearly indicate that R-AgGaO2 has a remarkably larger dispersed valence band as compared to
â-AgGaO2, which is advantageous to the photocatalytic activity due to the efficient hole conduction.
Introduction
In addition to the recent increasing interest in the photocata-
lytic applications for environmental purification, visible light
sensitive photocatalysts are necessary to efficiently utilize solar
energy and indoor light. Thus, numerous studies have investi-
gated these photocatalysts. Typically, UV-active photocatalysts
are modified by anion doping,1-5 metal doping,6-8 and generat-
ing oxygen deficiency.9 However, such dopants and oxygen
vacancies act as charge recombination centers. Therefore, these
methods are disadvantageous to the photocatalytic activity.
Another common method is to produce photocatalysts with
smaller band gap energies without dopants by forming the
maximums of their valence bands in the higher energy region.
These photocatalysts have been intensely investigated, especially
in the field of photocatalytic water splitting with specific d10/
d10s2 metal ions (Ag+,10,11 In3+ 12-14 Bi3+ 15
, etc.), but there are
,
limited reports in the field of photooxidation of organic
compounds.16,17
Delafossite structured oxides, which are denoted by ABO2
and include d10 metal ions (Cu+ or Ag+), are successful materials
for p-type transparent conductors (TCO) due to their structural
advantages.18-20 Figure 1a and b shows the crystal structures
of delafossite structured oxides. Delafossite has a layered
structure, which can be further classified into two polytypes,
3R (Figure 1a) and 2H (Figure 1b). Although the stacking
patterns of the layers differ slightly, both materials can be
regarded as the delafossite structure. The O-A-O dumbbell
bonding layers are interleaved with edge-sharing BO6 octahedra
blocks along the c-axis. These O-A-O layers and BO6 layers
act as separate conduction paths for holes and electrons,
respectively.18 In addition, each oxygen atom builds an OB3A
Figure 1. (a) Delafossite structure (3R-type), (b) delafossite structure
(2H-type), and (c) â-NaFeO2 structure.
* Corresponding author. Tel.: +81-3-5452-5080. Fax: +81-3-5452-
6593. E-mail:
irie-hrs@light.t.u-tokyo.ac.jp (H.I.); hashimoto@
tetrahedron, forming an O-2sp3 orbital, which does not form
localized O2p nonbonding states.19,20 These tetrahedra are
favorable for hole conduction. Furthermore, a small repulsion
light.t.u-tokyo.ac.jp (K.H.).
† Department of Applied Chemistry.
‡ RCAST.
10.1021/jp063406s CCC: $33.50 © 2006 American Chemical Society
Published on Web 10/24/2006

Visible Light Sensitive R-AgGaO2
J. Phys. Chem. B, Vol. 110, No. 46, 2006 23275
between O2- ions in two-coordinated O-A-O dumbbell bonds
causes a short A-O bonding. The interaction between the A+
and O2- ions then becomes stronger,21 and a highly dispersed
valence band structure is expected. This highly dispersed valence
band leads to a small effective mass of the holes with a higher
hole mobility. In contrast to the strong interactions between
them, interactions between the A+ ions are weak because they
work mainly through A+-O2--B3+-O2--A+ links, which are
two-dimensional in delafossite oxides. These weak and low-
dimensional interactions between the A+ ions induce a band
gap, which is bigger than that of A2O hemioxides.19,20 These
structural features, which are related to the effective hole
conduction and relatively large band gaps, are advantageous to
the p-type conductivity, transparency, and photocatalytic activity.
Copper delafossite oxides, CuMO2 (M ) Al, Cr, Fe, Ga, Y,
In, etc.), have been reported as p-type semiconductors without
doping,19,23,24 but most of them were black22,23 with small
indirect band gaps (1.26-1.65 eV).24 Hence, CuMO2 should
have an insufficient power to photooxidize organic compounds.
Among silver delafossite oxides (AgMO2), AgInO2 powder is
orange,22,23 indicating that the band gap is too narrow to generate
photooxidative activity to decompose organic compounds.
However, AgAlO2 powder is light gray.25 The gray color implies
that there is a relatively larger band gap, which can absorb UV
light, because the gray color should originate from the oxygen
vacancies. Therefore, we selected delafossite type AgGaO2, that
is, R-AgGaO2,26 because it should have an adequate band gap
to absorb visible light and to produce the photocatalytic activity.
In this report, we used R-AgGaO2 to investigate the photooxi-
dation of organic compounds.27 Because AgGaO2 has another
polymolph, that is, â-NaFeO2-type â-phase (Figure 1c), we also
examined the â-AgGaO2 phase.
TABLE 1: Absorbed Photon Numbers of Light Irradiation
a
on r-AgGaO2
TiO2
R-AgGaO2
9.6 × 1014
â-AgGaO2
2.2 × 1015
(anatase)
b
visible light
-
(λ ) 420-530 nm,
1000 µW cm-2
)
UV light
9.6 × 1014
1.3 × 1015
9.2 × 1014
(λ ) 300-400 nm,
750 µW cm-2
a Absorbed photon number/quanta cm-2 s-1
.
b Absorbed photon
number could not be determined for TiO2 (anatase) because its
absorption in the visible light region was virtually zero.
in a lab dish (irradiation area was 5.5 cm2) and placed in a 500
mL glass vessel with a quartz-made lid. The atmosphere in the
vessel then was substituted with dry synthetic air. A 300 ppm
(300 ppmv, that is, 6.7 µmol) concentration of the reactant gas
(IPA) was injected into the vessel and stored in the dark. After
the IPA concentration was confirmed constant, light irradiation
was initiated. IPA was photocatalytically oxidized via acetone
or directly into CO2 and H2O.30 To evaluate the photocatalytic
activity, the amounts of IPA, acetone, and CO2 were monitored
by a gas chromatograph (Shimadsu, GC-8A) equipped with
CO2-methanizer (Shimadsu, MTN-1). UV light with wave-
lengths between 300 and 400 nm and visible light between 420
and 530 nm were achieved by a Xe lamp (Hayashi Tokei,
Luminar Ace 210) using a glass filter (Asahi-Technoglass, UV-
D36B) and a combination of three glass filters (Asahi-Tech-
noglass, B-47, Y-44, C-40C), respectively. The experiments
were performed on R-AgGaO2, â-AgGaO2, and commercial
anatase TiO2 nanoparticles (Ishihara Sangyo, ST-01). The light
intensities for the visible and UV lights were adjusted to 1000
and 750 µW cm-2, respectively, to equalize the absorbed photon
numbers (9.6 × 1014 quanta cm-2 s-1) in the presence of
R-AgGaO2. Table 1 summarizes the absorbed photon numbers.
Additionally, IPA decomposition in the presence of R-AgGaO2
was measured under monochromatic lights to observe the
wavelength-dependence of the quantum efficiency (QE is the
same as the photonic efficiency) using a Czerny-Turner type
monochromator (Ritsu Ouyou Kogaku, MC-10N) and either a
Xe lamp or a Hg-Xe lamp (Hayashi Tokei, Luminar Ace 100).
The widths of the monochromatic lights were approximately
20 nm. Higher order diffracted lights were cutoff with the
appropriate glass filter. Table 2 summarizes the intensities,
wavelengths, and absorbed photon numbers.
The QE calculations were conducted using the same proce-
dure as previously reported.2,17 The initial acetone and CO2
generation rates by IPA decomposition were calculated using
the conventional least-squares method. Only one photon par-
ticipated in the decomposition process of IPA to acetone.30
Therefore, the QE values for acetone generation were calculated
using the following equation: QEacetone ) initial acetone
generation rate/absorption rate of incident photon. As compared
to the mechanism for acetone generation, the mechanism for
CO2 generation is complex. Therefore, it is difficult to determine
the number of photons participating in the reaction because there
are two reaction paths. Assuming the following formula, C3H8O
+ 5H2O + 18H+ f 3CO2 + 18H+, that is, six photons are
required to produce one CO2 molecule, the QE values for CO2
Experimental Section
R-AgGaO2 powder was prepared by a cation exchange
reaction.21,28,29 NaGaO2 powder was initially synthesized by a
solid-state reaction. Commercial Na2CO3 (Wako Chemical,
99.5%) and Ga2O3 (Kojundo Chemical, 99.9%) powders were
used as the starting materials. Stoichiometric amounts of the
starting materials for NaGaO2 were wet ball milled using ZrO2
balls as the milling medium in polyethylene bottles. The
resulting mixture was calcined at 700 °C for 5 h. The obtained
powder was reground and sintered at 900 °C for 5 h to produce
white â-NaGaO2 powder. â-NaGaO2 then was mixed with
AgNO3 (Wako Chemical, 99.8%) and KNO3 (Wako Chemical,
99%) with a molar ratio of 1:1.03:1 in an agate mortar, followed
by heating at 210 °C in air for 24 h. KNO3 was used to lower
the mixing enthalpy of the molten nitrates to accelerate the
substitution of Na+ by Ag+.21 The resulting powder was
pulverized and washed with enough distilled water to remove
the remaining nitrates, and dried in a refrigerator. The obtained
powder was â-AgGaO2, which was further heated in water at
60-70 °C for 12 h to complete the delafossite phase, R-Ag-
GaO2.
The crystal structures of the prepared powders were identified
by X-ray diffraction (XRD; Spectris, Panarytical PW-1700). The
surface areas were determined by BET (Micrometrics,
TriStar3000). The diffuse reflectance spectra were obtained with
a UV-vis spectrometer (Rigaku, UV2550). The grain sizes of
the powders were observed by scanning electron microscopy
(SEM; S-4200, Hitachi Ltd.).
The photocatalytic oxidation activities were evaluated by the
decomposition of gaseous 2-propanol (IPA) while irradiating
with light. Each 300 mg powder sample was uniformly spread
generation were calculated using the following equation, QECO
2
) 6 × initial CO2 generation rate/absorption rate of incident
photon.
Plane-wave-based density functional method calculations
were carried out for R- and â-AgGaO2 by employing an ab initio

23276 J. Phys. Chem. B, Vol. 110, No. 46, 2006
Maruyama et al.
TABLE 2: Intensities and Absorbed Photon Numbers of Monochromatic Light Irradiation on r-AgGaO2
wavelength/nm
365
80a
390
100
430
100
470
100
510
100
530
100
550
100
570
100
intensity/µW cm-2
absorbed photon number/
quanta cm-2 s-1
1.0 × 1014
1.2 × 1014
1.1 × 1014
9.6 × 1013
8.7 × 1013
8.9 × 1013
9.3 × 1013
9.7 × 1013
a Hg-Xe lamp was used as a light source.
Figure 3. UV-vis absorption spectra of R-AgGaO2 and â-AgGaO2.
Figure 2. X-ray powder diffraction patterns of (a) R-AgGaO2 and (b)
â-AgGaO2. The inset of (a) indicates no changes in XRD profiles for
R-AgGaO2 between before and after 72 h UV light (the same UV light
condition in Table 1) irradiation.
Figure 4. (hνR)1/2 plot for (a) R-AgGaO2 and (b) â-AgGaO2. The
absorption coefficient R is estimated by the Kubelka-Munk conversion
of reflectance.
total energy and molecular dynamics program, VASP (Vienna
Ab initio Simulation Program).31,32 The wave functions and
potentials of the core orbitals were replaced by a projector
augmented wave (PAW) by Blo¨chl,33 and the exchange-
correlation energy was calculated by the generalized gradient
approximation (GGA). At first, the lattice structures of R- and
â-AgGaO2 were optimized for their atomic positions using
lattice parameters22,34 reported for related structures. The kinetic
energy cutoff was 500 eV for structure optimization. The density
of states (DOSs) and band structures were calculated from the
optimized structures with an energy cutoff of 400 eV.
Result and Discussion
Figure 5. Changes in acetone and CO2 concentrations by IPA
decomposition as a function of time in the presence of R-AgGaO2 or
â-AgGaO2 or TiO2 (anatase) under (a) visible light (420 < λ < 530
nm) and (b) UV light (300 < λ < 400 nm).
R-AgGaO2 and â-AgGaO2 were obtained as yellow and
orange powders, respectively. Figure 2a and b shows the XRD
patterns of the obtained R- and â-AgGaO2 samples, respectively.
Both samples show homogeneous crystalline R- and â-AgGaO2
phases. R-AgGaO2 is mainly the 3R delafossite phase. Peak
broadenings are observed only in Figure 2a when the peaks were
diffracted from crystallographic planes neither perpendicular nor
horizontal to the c-axis, which is attributed to the long-range
stacking disorder between the 2H and 3R polytypes along the
c-axis.35 The surface areas of R- and â-AgGaO2 are 2.80 and
1.49 m2 g-1, respectively. According to the SEM observations,
both are agglomerated and have similar powder sizes of 200-
1000 nm (not shown here). â-AgGaO2 is shaped into spherical
particle, but R-AgGaO2 is shaped into hexagonal platelets, which
is consistent with the previous report.22 Figure 3 shows the UV-
visible absorption spectra of R- and â-AgGaO2. Both types of
AgGaO2 absorb in the visible light range, but the wavelength
for the absorption-edge of R-AgGaO2 is shorter than that of
â-AgGaO2. The constant absorption is larger than a wavelength
of approximately 520 nm for R-AgGaO2 and approximately 600
nm for â-AgGaO2. The reason for the constant absorption is
unclear, but may be due to the oxygen vacancies; it is well-
known that the oxygen vacancy level is just below the
conduction minimum in the forbidden band, which results in
the absorption in a longer wavelength region.9 Delafossite
structured oxides are regarded as indirect gap semiconductors,36-38
and the band gap of the R-AgGaO2 is estimated from the tangent
lines in the plots of the square root of the Kubelka-Munk
functions against the photon energy as shown in Figure 4a. The
tangent line extrapolated to R1/2 ) 0 indicates that the band
gap of R-AgGaO2 is 2.4 eV. Similarly, that of â-AgGaO2 is
2.1 eV, as shown in Figure 4b.39
Figure 5a and b shows the photocatalytic activity in the
presence of R-AgGaO2, â-AgGaO2, and anatase powders while
irradiating with UV (300-400 nm) and visible light (420-530
nm), respectively. The anatase powder shows a photocatalytic
activity under UV light irradiation to produce acetone and CO2,
but did not display a photocatalytic activity under visible light

Visible Light Sensitive R-AgGaO2
J. Phys. Chem. B, Vol. 110, No. 46, 2006 23277
irradiation. The increases in the acetone and CO2 concentrations
under both visible or UV light irradiation were negligible in
the presence of â-AgGaO2, which confirms â-AgGaO2 does not
have photocatalytic activity. On the contrary, in the presence
of R-AgGaO2 under either light irradiation, the acetone and CO2
concentrations increased by decomposing IPA. Judging from
the fwhm (full width at half-maximum) of XRD patterns, the
crystallinity of â-AgGaO2 is higher than that of R-AgGaO2 in
the present samples. It means that even the lower crystallized
R-AgGaO2 possesses much higher photocatalytic activity than
the higher crystallized â-AgGaO2. Therefore, we can conclude
that R-AgGaO2 generally possesses much higher photocatalytic
activity than the higher crystallized â-AgGaO2. These results
are due to the differences in the band gaps. â-AgGaO2 has a
smaller band gap, which forms in the top of its valence band in
a higher energy region and results in an insufficient power to
oxidate IPA. In contrast, R-AgGaO2 has a larger band gap so
that it exhibits the appropriate oxidation power. In addition to
the band gap, the larger dispersed valence band of R-AgGaO2
is attributed to the photocatalytic activity. The QE for IPA
decomposition in the presence of R-AgGaO2 was 0.35% under
UV light, which was much lower than that in the presence of
TiO2 (6.0%). However, the QE of R-AgGaO2 under visible light
is 0.39%, which is much higher than values reported for other
photocatalysts sensitive to visible light (AgNbO3, GaN-ZnO
solid solution, <0.1% under visible light).40
Figure 6. Dependence of the quantum efficiency (QE) on the
wavelength of monochromatic light (λ ) 365, 390, 430, 470, 510, 530,
550, 570 ( 10 nm). For λ ) 365 and 430 nm, experiments were
performed twice.
The stability of R-AgGaO2 under UV light irradiation was
confirmed using the XRD, XPS (X-ray photoemission spec-
troscopy, Perkin-Elmer Japan, 5600), and UV-visible spec-
trometer. The UV light condition was controlled to be the same
as the IPA decomposition test in Figure 5b, as shown in Table
1. Even after 72 h irradiation, the XRD profile did not change,
as compared to that before light irradiation, as shown in the
inset of Figure 2a. XRD peaks derived from metallic Ag were
not observed at all. The data are not shown here, but neither
Ag3d, Ga3d, nor O1s XPS spectra were changed after UV light
irradiation. The UV-visible absorption spectrum after UV light
irradiation slightly changed as shown in Figure 3. After UV
light irradiation, the value of 1-reflectance increased very slightly
in the wavelength region larger than 500 nm, which corresponds
to Ag plasmon absorption. Because the change in 1-reflectance
was very small and no changes in XRD and XPS were observed,
the present R-AgGaO2 was quite stable under UV light condition
in Table 1.
Figure 7. (a) Density of states and (b) band structures calculated for
R-AgGaO2 (left) and â-AgGaO2 (right) using VASP.
the absorption edge and the QE values are nearly constant, but
the QE values steeply decrease when irradiating with wave-
lengths longer than 530 ( 10 nm. These results clearly
demonstrate that R-AgGaO2 has a continuous band structure
with a band gap of 2.4 eV.
The QE values of IPA decomposition on R-AgGaO2 were
on the same order regardless of the irradiation light source (UV
light (0.35%) and visible (0.39%) light), indicating that R-Ag-
GaO2 has a continuous valence band. This means that the visible
light absorption of R-AgGaO2 should be originated from an
interband transfer, not from a transition from/to discrete energy
Ab initio calculations were performed to evaluate the
electronic structures of R- and â-AgGaO2. Figure 7a and b
shows the density of states and band structures, respectively,
for R- and â-AgGaO2 calculated by VASP. The valence band
maximums for both R- and â-AgGaO2 consist of Ag4d and O2p
orbitals. Additionally, both are indirect semiconductors where
an indirect transition is favorable for carrier migration. The
indirect band gap of R-AgGaO2 agrees well with previous
reports on the electronic structures of delafossite CuMO2 (M
) Al, Ga, In) with a conduction band minimum on the Γ point
and a valence band maximum near the F-point as calculated by
the Hartree-Fock-Roothan method36,37 or the full-potential
linear augmented plane wave method.38 As shown in Figure 7a
and b, the band structures of R- and â-AgGaO2 obviously differ.
As expected from the structural property of delafossite oxides,
the valence band of R-AgGaO2 is highly dispersed, which means
that the photogenerated holes have a small effective mass. This
probably explains the higher photocatalytic activity in the
presence of R-AgGaO2 as compared to that in the presence of
2,5
levels generated by dopants, that is, anion-doped TiO2 due
to the following reason. As soon as the free charge carriers are
photogenerated in a semiconductor and rapid communications
between the different subband states in the valence band and
conduction band are complete, then the photocarriers will occupy
the lowest states in their corresponding bands (i.e., they become
thermalized). The carrier temperature becomes equal to the
ambient temperature and the carriers are indistinguishable,
regardless of the initial state immediately occupied upon
photoexcitation. That is, the excess photoenergy is dissipated
as heat rather than as chemical potential. To confirm this
explanation, we performed IPA decomposition on R-AgGaO2
under monochromatic light irradiation. Figure 6 shows the
results. When irradiating with wavelength shorter than 520 nm,

23278 J. Phys. Chem. B, Vol. 110, No. 46, 2006
Maruyama et al.
(4) Ohno, T.; Mitsui, T.; Matsumura, M. Chem. Lett. 2003, 32, 364.
(5) Irie, H.; Watanabe, Y.; Hashimoto, K. Chem. Lett. 2003, 32, 772.
(6) Borgarello, E.; Kiwi, J.; Gratzel, M.; Pelizzetti, E.; Visca, M. J.
Am. Chem. Soc. 1982, 104, 2996.
(7) Sepone, N.; Hermann, J. M. Langmuir 1994, 10, 693.
(8) Choi, W.; Termin, A.; Hoffmann, M. J. Phys. Chem. 1994, 97,
13669.
(9) Nakamura, I.; Negishi, N.; Kutsuna, S.; Ihara, T.; Sugihara, S.;
Takeuchi, K. J. Mol. Catal. A 2000, 161, 205.
(10) Kato, H.; Kobayashi, H.; Kudo, A. J. Phys. Chem. B 2002, 106,
12441.
â-AgGaO2. In addition, the calculated band gap of R-AgGaO2
is 0.73 eV, which is larger than that of â-AgGaO2, 0.34 eV.
This grossly underestimated band energy gap is characteristic
of the density functional calculation. In R-AgGaO2, the Ag-O
bonds are two-coordinated and the Ag+ ion network is two-
dimensional, which lead to a large dispersion of the valence
band and a sufficient band gap. However, in â-AgGaO2, the
Ag-O bonds are four-coordinated and the Ag+ ion network is
three-dimensional, which lead to a small dispersion of the
valence band and an insufficient band gap. These results are
consistent with the above studies on the structural properties
and photocatalytic activities. Ag2O has O-Ag-O dumbbell
bonds similar to delafossite oxides, and the electronic structure
of Ag2O calculated by VASP has a highly dispersed valence
band. However, the band gap is almost zero (not shown), which
may be due to the O-Ag-O three-dimensional cross-linking
as Cu2O.19,20 These findings are compatible with a fact that
Ag2O is a p-type semiconductor41 with a 1.8 eV band gap, which
is insufficient to photooxidize organic compounds. In contrast,
silver complex-oxides, including Ag+ ions such as AgNbO3,
have relatively large band gaps. However, the bonds and the
interactions between the Ag+ ion and O2- ions are weak, which
decrease the dispersion of the calculated valence band and lower
the hole conductivity. Thus, the weak bonds and interactions
may explain the low photooxidation activity of these silver
complex-oxides.42 However, R-AgGaO2 is an excellent photo-
catalyst, which is sensitive to visible light, because it has the
appropriate band gap for visible light absorption and a high hole
conductivity, which both originate from its unique layered
crystal structure.
(11) Konta, R.; Kato, H.; Kobayashi, H.; Kudo, A. Phys. Chem. Chem.
Phys. 2003, 5, 3061.
(12) Zou, Z.; Ye, J.; Arakawa, H. Chem. Phys. Lett. 2000, 332, 271.
(13) Ye, J.; Zou, Z.; Oshikiri, A.; Matsuyama, M.; Shimoda, M.; Imai,
M.; Shishido, T. Chem. Phys. Lett. 2000, 356, 221.
(14) Zou, Z.; Ye, J.; Sayama, K.; Arakawa, H. Nature 2001, 414, 625.
(15) Kudo, A.; Omori, K.; Kato, H. J. Am. Chem. Soc. 1999, 121, 11459.
(16) Tang, J.; Zou, Z.; Ye, J. Angew. Chem. 2004, 116, 4563.
(17) Murase, T.; Irie, H.; Hashimoto, K. J. Phys. Chem. B 2005, 109,
13420.
(18) Hosono, H.; Kamiya. Bull. Ceram. Soc. Jpn. 2003, 38, 825.
(19) Kawazoe, H.; Yasukawa, M.; Hyodo, H.; Kurita, M.; Yanagi, H.;
Yasukawa, M.; Hosono, H. Nature 1997, 389, 939.
(20) Yanagi, H.; Kawazoe, H.; Kudo, A.; Yasukawa, M.; Hosono, H.
J. Electroceram. 2000, 4, 407.
(21) Clayton, J. E.; Cann, D. P.; Ashmore, N. Thin Solid Films 2002,
411, 140.
(22) Shannon, R. D.; Rogers, D. B.; Prewitt, C. T. Inorg. Chem. 1974,
10, 713.
(23) Nagarajan, R.; Duan, N.; Jayaraj, M. K.; Li, J.; Vanagja, K. A.;
Yokochi, A.; Draeseke, A.; Tate, J.; Sleight, A. W. Int. J. Inorg. Mater.
2001, 3, 265.
(24) Benko, F. A.; Koffyberg, F. P. J. Phys. Chem. Solids 1987, 48,
431.
(25) Gessner, W. Z. Anorg. Allg. Chem. 1967, 352, 145.
(26) Sheets, W. C.; Mugnier, E.; Barnabe, A.; Marks, T. J.; Poep-
pelmeier, K. R. Chem. Mater. 2006, 18, 7.
(27) Delafossite structured oxide photocatalysts have also been reported
for CuAlO2 to generate hydrogen in extreme conditions (Koriche, N.;
Bouguelia, A.; Aider, A.; Trari, M. Int. J. Hydrogen Energy 2005, 30, 693)
and for R-AgAlO2 to degrade methylene blue (Zou, Z.; Yang, S. O.; Ye, J.
ISNEPP 2005 Abstract 2005).
(28) Thilo, E.; Gessner, W. Z. Anorg. Allg. Chem. 1966, 347, 151.
(29) Li, J.; Sleight, A. W. J. Solid State Chem. 2004, 177, 889.
(30) Ohko, Y.; Hashimoto, K.; Fujishima, A. J. Phys. Chem. A 1997,
101, 8057.
Conclusion
We investigated a novel photocatalyst, which is sensitive to
visible light, from the viewpoints of the electronic structure of
the valence band and the crystal structure. R-AgGaO2 has a
relatively high photocatalytic activity and can decompose
2-propanol via acetone and ultimately produces CO2 under
visible light irradiation (λ < 520 nm). R-AgGaO2 has the proper
band gap energy to absorb visible light and to generate a
photocatalytic activity. The calculated electronic structure of
R-AgGaO2 also supports that it is advantageous for the mobility
of the photogenerated holes, which leads to the rather high
activity. This study demonstrates that searching for a material
with a delocalized and dispersed valence band by mixing a
metal-d orbital with an O2p orbital is a candidate for generating
a high-performance, visible light sensitive photocatalyst.
(31) Kresse, G.; Hafner, J. Phys. ReV. B 1993, 48, 17.
(32) Kresse, G.; Furthmu¨ller, J. Phys. ReV. B 1996, 54, 16.
(33) Blo¨chl, P. E. Phys. ReV. B 1994, 50, 17953.
(34) Data from JCPDS Card, #21-1076.
(35) Jacob, A.; Parent, C.; Boutinaud, P.; Flem, G. L.; Dormerc, J. P.;
Ammar, A.; Elazhari, M.; Elaatmani, M. Solid State Commun. 1997, 103,
529.
(36) Buljan, A.; Alemany, P.; Ruiz, E. J. Phys. Chem. B 1999, 103,
8060.
(37) Robertson, J.; Peacock, P. W.; Towler, M. D.; Needs, R. Thin Solid
Films 2002, 411, 96.
(38) Nie, X.; Wei, S. H.; Zhang, S. B. Phys. ReV. Lett. 2002, 88, 066405.
(39) According to ab initio calculations, â-AgGaO2 and R-AgGaO2 are
indirect semiconductors.
Acknowledgment. This work was supported by a Grant-
in-Aid for Scientific Research on Priority Areas (417) from the
Ministry of Education, Culture, Sports, Science, and Technology
(MEXT) of Japan.
(40) We examined the photocatalytic activity of IPA decomposition in
the presence of other visible light sensitive water-splitting photocatalysts,
which have been already reported, under visible light irradiation with
1014-1015 absorbed photon numbers. The quantum yield is 0.041% for
AgNbO3 (band gap ≈ 2.8 eV10) and 0.039% for a GaN-ZnO solid solution
(band gap ≈ 2.6 eV; Maeda, K.; Tanaka, T.; Hara, M.; Saito, N.; Inoue,
Y.; Kobayashi, H.; Domen, K. J. Am. Chem. Soc. 2005, 127, 8286).
(41) Fortiu, E. Phys. Status Solidi 1964, 5, 515.
References and Notes
(1) Asahi, R.; Morikawa, T.; Aoki, K.; Taga, Y. Science 2001, 293,
(42) AgNbO3 has a relatively large band gap of 2.8 eV, contains 12-
coordinated long A-O bonds (∼2.79 Å on average), and is considered to
have weak bonds and interaction. The band structure calculated by VASP
shows a larger band gap (1.68 eV) and small dispersion of the valence
band, which should lead to a large effective mass and small hole
conductivity.
269.
(2) Irie, H.; Watanabe, Y.; Hashimoto, K. J. Phys. Chem. B 2003, 107,
5483.
(3) Umebayashi, T.; Yamaki, T.; Tanaka, S.; Asai, K. Chem. Lett. 2003,
32, 330.