Electronic structure and photocatalytic characterization of a novel photocatalyst AgAlO2

Article abstract of DOI:10.1021/jp055924t

A novel photocatalyst, AgAlO₂, was prepared by cation exchange reaction and characterized by powder X-ray diffraction. The result showed AgAlO₂ crystallizes in the layered orthorhombic structure with space group was Pna2₁. The energy band and electronic structures of AgAlO₂ were calculated based on the crystal structure. It was found that AgAlO₂ is an indirect band gap semiconductor. The valence band top mainly consists of O-2p orbitals and Ag-4d orbitals and the conduction band bottom is mainly constructed of Ag-5s5p orbitals. The band gap of AgAlO ₂ was estimated to be about 2.8(1) eV with UV-vis diffuse reflectance spectrometry. The photocatalytic activity of AgAlO₂ was characterized by photocatalytically decomposing the dye alizarin red (AR) under visible light irradiation, and AR could be decomposed about 70% under 2 h of visible light irradiation.

Full text of DOI:10.1021/jp055924t

J. Phys. Chem. B 2006, 110, 11677-11682
11677
Electronic Structure and Photocatalytic Characterization of a Novel Photocatalyst AgAlO2
Shuxin Ouyang,‡,⊥ Haitao Zhang,‡,⊥ Dunfang Li,‡,⊥ Tao Yu,§,⊥ Jinhua Ye, and Zhigang Zou*
†
,‡,⊥
Department of Materials Science and Engineering, Nanjing UniVersity, Nanjing 210093, P.R. China,
Ecomaterials and Renewable Energy Research Center (ERERC), Department of Physics, Nanjing UniVersity,
Nanjing 210093, P.R. China, National Laboratory of Solid State Microstructures, Nanjing UniVersity,
Nanjing 210093, P.R. China, and Ecomaterials Center, National Institute for Materials Science,
1
-2-1 Sengen, Tsukuba, Ibaraki 305-0047, Japan
ReceiVed: October 16, 2005; In Final Form: January 16, 2006
A novel photocatalyst, AgAlO
diffraction. The result showed AgAlO
was Pna2 . The energy band and electronic structures of AgAlO
It was found that AgAlO is an indirect band gap semiconductor. The valence band top mainly consists of
O-2p orbitals and Ag-4d orbitals and the conduction band bottom is mainly constructed of Ag-5s5p orbitals.
The band gap of AgAlO was estimated to be about 2.8(1) eV with UV-vis diffuse reflectance spectrometry.
The photocatalytic activity of AgAlO was characterized by photocatalytically decomposing the dye alizarin
2
, was prepared by cation exchange reaction and characterized by powder X-ray
crystallizes in the layered orthorhombic structure with space group
were calculated based on the crystal structure.
2
1
2
2
2
2
red (AR) under visible light irradiation, and AR could be decomposed about 70% under 2 h of visible light
irradiation.
I. Introduction
The electronic structures of some multi-metal oxides which
consist of Ag, such as AgInW2O8, AgTaO3 , and AgNbO3,
have been studied. Their valence bands (VB) are all constructed
by the hybridized Ag-4d orbitals and O-2p orbitals, and their
conduction bands (CB) are constructed by the hybridized Ag-
1
1
2
12
Since Fujishima et al.’s discovery of the photocatalytic
property of TiO2 three decades ago, many studies have been
carried out to develop novel semiconductor photocatalysts that
are able to split water into H2 and O2 with high efficiency.
Because electrons in these photocatalysts are excited from the
O-2p orbital to transition metal-d orbitals, their band gaps are
1
5
s orbitals and other orbitals, such as In-5s orbitals and W-5d
orbitals (in AgInW2O8), Ta-5d orbitals (in AgTaO3), and Nb-
4d orbitals (in AgNbO3). But in these studies, the individual
contribution of Ag to the valence band and the conduction band
has not been investigated clearly. To study the individual
contribution of Ag to the valence band and the conduction band,
Ag and the element Al, which only has 3s and 3p orbitals but
does not have d orbitals, were selected in our study. AgAlO2
has two typical structures, the hexagonal structure and the
orthorhombic structure. But the orthorhombic structure consists
of dense polyhedron layers. This means that such a structure
would be more favorable for transferring electron-hole pairs.
So the AgAlO2 with the orthorhombic crystal structure of space
group Pna21 was chosen to execute the calculation analysis.
The energy band, crystal, and electronic structures of AgAlO2
were discussed in detail. Then, the sample was prepared and
the photocatalytic AR decomposition over the photocatalyst
under visible light irradiation was investigated to test the validity
of the material design method.
2
usually wider, exceeding 3.0 eV. This means that most of these
photocatalysts can only exhibit photocatalytic activity under UV
irradiation. The ultraviolet light accounts for merely about 4%
of the whole solar energy, but the visible light accounts for 43%
of the whole solar energy and also the main portion of the indoor
artificial illumination. To develop photocatalysts with high
activity under visible light, many investigations have been
carried out on new materials with a narrow band gap. Anion
2
doping of the transition metal oxides, such as TiO2-xNx ,
3
-6
TaON,
etc., could decrease the band gap. Though the light
absorption was increased to visible, the defects introduced by
doping would accelerate recombination of the electrons and the
holes. Some nitrides such as Ta3N5 3-5 were also reported as
photocatalysts under visible light, but not all of them were stable
under irradiation. On the other hand, the orbitals of transition
elements in multimetal oxide could hybridize the primal orbitals
to narrow the band gap. In this case, Zou et al. found that
InNbO4 and InTaO4 could evolve H2 from pure water under
visible light irradiation.7-10
II. Computational and Experimental Methods
To explore an approach to develop the novel visible light
corresponding to photocatalysts more efficiently, one of the
photocatalyst material design methods, which combines the first-
principles energy band calculation with the crystal structure
analysis, has been successfully developed in this paper.
Material Design Details. As mentioned above, in the studies
1
1
12
12
of AgInW O , AgTaO3, and AgNbO3, the individual
2
8
contribution of Ag to the valence band and the conduction band
has not been investigated clearly. To achieve the intention, a
suitable material should be dissectionally chosen. Ag O seems
2
to be the most adequate candidate, but the band gap of Ag2O is
*
Author to whom correspondence should be addressed. E-mail: zgzou@
too narrow to be a photocatalyst. The band gap breadths of silver
oxides are influenced nearly by the bond length of Ag-O. In
Table 1, the bond lengths of Ag-O in the familiar Ag2O are
smaller than 2.17 Å, and the bond length of Ag-O in AgTaO3,
which has high photocatalytic activity under ultraviolet light,
nju.edu.cn.
‡
Department of Materials Science and Engineering, Nanjing University.
§
†
⊥
National Laboratory of Solid State Microstructures, Nanjing University.
Ecomaterials Center, National Institute for Materials Science.
Ecomaterials and Renewable Energy Research Center, Nanjing
University.
1
0.1021/jp055924t CCC: $33.50 © 2006 American Chemical Society
Published on Web 05/27/2006

11678 J. Phys. Chem. B, Vol. 110, No. 24, 2006
Ouyang et al.
TABLE 1: The Ag-O Bond Length of Ag
2
O, AgTaO
3
, and
the thermal catalytic effect) that was used to locate the reaction
cell. The photocatalytic decomposition of AR was carried out
with 0.3 g of AgAlO2 powder suspended in 100 mL of AR
solution in a Pyrex glass cell under 2 h of irradiation. Before
irradiation the suspensions were magnetically stirred in the dark
for several hours to ensure establishment of an adsorption/
desorption equilibrium of dye on the sample surface. The AR
solution concentration was about 16 mg/L and the pH value of
the solution was nearly 7.5 after the dye reached the adsorption/
desorption equilibrium on the sample surface. The whole
photocatalytic reaction was conducted in air at about 30 °C (the
temperature of the reaction solution). The suspension samples
including the photocatalyst and AR solution were separated by
the centrifuge, and the solution samples were analyzed by the
UV-visible spectrometer (CARY 50 Probe, Varian, America)
and the total organic carbon (TOC) measurer (TOC-5000,
Shimadzu, Japan). The AR decomposition rate was calculated
as below: at the irradiation time of t min (including the
irradiation times 0, 20, 40, 60, 80, 100 and 120 min), the solution
samples (a certain volume, about 2 mL) after centrifugal
treatment were pH adjusted with an ammonia solution drip (pH
AgAlO
2
av Ag-O bond
space group
Ag-O bond length/Å
length /Å
P 3h m11
3
Ag
2
O
2.161, 2.161, 2.161
2.061, 2.061
2.048, 2.048
2.462, 2.462, 2.462
2.348, 2.360, 2.412, 2.477
2.161
2.061
2.048
2.462
2.399
1
4
Pn 3h mS
Pn 3h mZ
1
5
1
6
AgTaO
AgAlO
3
R 3h cH
1
7
2
Pna2
1
is 2.46 Å. AgAlO2, which is researched here, matches the
condition that the bond length of Ag-O is between 2.17 and
2
.46 Å; at the same time, the choice of the element Al could
meet the intention of this study.
Calculations. The lattice parameters of the AgAlO2 model
were optimized, and then the optimized AgAlO2 model was
calculated and the ground-state energy band by using the
standard CASTEP package.18 The lattice parameters and
coordinates of the atoms in AgAlO2 are listed in the Table 2.
The CASTEP code is a plane-wave pseudopotential total energy
calculation method that is based on the density functional theory
(DFT). In our calculational study, the electronic wave functions
12) and the UV-visible absorbance spectra were measured, and
were expanded in a plane-wave basis set up to a 340 eV cutoff.
The FFT grid of the basis was 30 × 40 × 30. And the self-
their UV-visible absorbance spectra all presented a character-
istic peak at 486 nm, because the absorbance value of the 0
min solution sample was A0, and the absorbance value of the t
min solution sample was At, so at the time of t min, the AR
decomposition rate DR(t) ) (A0 - At)/A0 × 100%.
-
6
consistent field (SCF) tolerance was 1 × 10 eV/atom. The
ultrasoft pseudopotential and 5 × 4 × 5 k-point for sample were
chosen in the calculation. The electronic exchange-correlation
energy was treated within the framework of the generalized
gradient approximation (GGA), which was superior to the local
density approximation (LDA). The calculations of the energy
band and density of states (DOS) were both executed.
Experiments. Cation exchange reaction was used to prepare
the polycrystalline AgAlO2.17 The precursor LiAlO2 was first
prepared by reaction of Al2O3 (99.0%) with a stoichiometric
amount of Li2CO3 (99.99%) at 750 °C for 12 h, and then
reground. The mixed powder was sintered at 1000 °C for 12 h
and the white LiAlO2 powder was obtained. The powder was
mixed with AgNO3 (99.8%) and KNO3 (99.0%) at a molar ratio
of 1:2:2 in an agate mortar. Then the mixture was heated in air
at 275 °C for 12 h. The resulting sheet sample that was washed
by distilled water became powder. At last, the sample was dried
in air at room temperature.
The photocatalytic decomposition of acetaldehyde was carried
out with 0.3 g of the powdered photocatalyst. The photocatalyst
was placed at the bottom of a Pyrex glass cell. The reaction
was in a gas-closed system with a gas-circulated pump. The
optical system for the catalytic reaction included a 300 W Xe
arc lamp (focused through a shutter window), a cutoff filter
(providing visible light of different wavelengths), and a water
filter (to prevent IR irradiation). The concentration of the CH3-
CHO gas mixture was 1270 and 680 ppm with nonfilter and λ
g 420 nm filter, respectively. The concentrations of CO2 and
acetaldehyde were detected by the gas chromatograph (GC-14B
with FID detector, Shimadzu). The yield of CO2 was calculated
for the mineralization rate of the CH3CHO.
III. Results and Discussion
The crystal structure of the powder sample was determined
by the X-ray diffraction method (Rigaku Denki, D/MAX-RA,
Japan) with Cu KR1 radiation. Intensity data were collected from
Crystal Structure and Energy Band. The calculating
AgAlO2 model was an orthorhombic crystal structure with space
group Pna21. Figure 1a represents a unit cell of AgAlO2 and
Figure 1b shows the polyhedron model of AgAlO2. Correspond-
ing to Figure 1a, Table 3 lists the bond lengths and bond angles
in the AgAlO2 crystal. Ag, Al, and O atoms in AgAlO2 all
occupy Wyckoff position 4a. The AgAlO2 crystal consists of
two kinds of tetrahedra, AgO4 and AlO4. In the XY-plane, the
AgO4 or AlO4 tetrahedra connects to each other to form zigzag
chains by sharing vertexes. Then the chains of AgO4 and the
1
5° to 75° in 2θ at room temperature; the scan speed was 10
-1
deg min , and the scan step was 0.02°. The diffuse reflectance
spectra of the sample were analyzed with a UV-visible
spectrometer (UV-2401, Shimadzu, Japan). The surface area was
determined by BET measurement (TriStar-3000, Micromeritics,
America).
The photocatalytic reaction system consisted of a 300 W Xe
arc lamp, cutoff filter, and a water cooler trough (preventing
TABLE 2: The Lattice Parameters and Coordinates of the Atoms in AgAlO
2
, Which Is an Orthorhombic Crystal Structure
with Space Group Pna2
1
Arris
exptl data
calcd data
lattice parameters
a (Å)
b (Å)
c (Å)
5.4306
6.9802
5.3751
5.2728
6.7628
5.3626
atom
Wyckoff site
x
y
z
x
y
z
atomic coordinates
Ag
Al
O1
O2
4a
4a
4a
4a
0.0532
0.0610
0.0311
0.1283
0.6268
0.1251
0.0723
0.6765
0.9969
0.0000
0.3209
0.4345
0.0598
0.0706
0.0342
0.1117
0.6427
0.1239
0.0679
0.6681
1.0069
-0.0013
0.3119
0.4346

Characterization of a Novel Photocatalyst AgAlO2
J. Phys. Chem. B, Vol. 110, No. 24, 2006 11679
Figure 1. The crystal structure of AgAlO
with two O1 atoms and two O2 atoms, and each Ag atom also link to two O1 atoms and O2 atoms, so the O1 atoms or the O2 atoms are the
juncture of the metal atoms Al and Ag; (b) the polyhedron modelsthe AgAlO crystal consists of two kinds of tetrahedron, AgO and AlO . The
AgO and AlO tetrahedron connects to each other to form a 3D net structure by sharing vertexes.
2
with orthorhombic structure: (a) the ball-and-stick modelsit is obvious that every Al atom connect
2
4
4
4
4
TABLE 3: The Bond Lengths and Bond Angles of AgAlO
2
except for some small peaks of Ag and LiAlO2 impurity, but
the metal Ag cannot exhibit photocatalysis, and the LiAlO2 that
was the precursor of AgAlO2 is not able to absorb light with a
wavelength larger than 290 nm, so the impurity of Ag and
LiAlO2 did not influence the characterization of photocatalytic
experiments of AgAlO2 in our experimental conditions. Ac-
cording to the diffuse reflectance spectra of the sample (in Figure
type value
bond lengths, Å
bond angles, deg
Ag-O
2.348, 2.360, 2.412, 2.477
1.754, 1.760, 1.761, 1.771
100.9, 103.9, 107.3, 131.6
99.3
Al-O
O1-Ag-O2
O1-Ag-O1
O2-Ag-O2
O1-Al-O2
O1-Al-O1
O2-Al-O2
Ag-O1-Ag
Ag-O2-Ag
Al-O1-Al
Al-O2-Al
Ag-O1-Al
Ag-O2-Al
109.8
107.7, 108.9, 109.0, 110.0
110.2
111.1
83.5
86.5
136.4
132.9
5), the band gap is estimated at about 2.8(1) eV, which is larger
than the calculated value. The reasons are as follows: on one
hand, because the discontinuity in the exchange-correlation
potential is not taken into account within the framework of
density functional theory (DFT), theoretical values of the energy
gap between unoccupied and occupied orbitals in semiconduc-
tors and insulators are underestimated compared with experi-
99.9, 103.0, 104.9, 115.2
100.9, 102.1, 109.1, 115.7
19
mental ones; on the other hand, it is considered that the optical
band gap does not originate from bulk, but from surface and
chains of AlO4 alternately link by sharing vertexes to form a
layer. Layers join each other also by vertexes to construct the
three-dimensional network.
2
0,21
defects.
Zhao et al. have studied the photooxidation mechanism of
2
2
dye AR with TiO2 as photocatalyst. In our experiment the
photocatalytic activity of AgAlO2 was characterized by the
experiment of photocatalytically decomposing the dye AR. The
results are shown in Figure 4. When the suspensions were
magnetically stirred in the dark for several hours to ensure
establishment of an adsorption/desorption equilibrium of dye
on the sample surface, the AR solution concentration decreased
only a little because of the adsorption, but after the Xe lamp
was turned on, the AR solution concentration decreased rapidly,
especially in the first 20 min, where the concentration decreased
about 42%, which means that this reaction is a light-responding
reaction. The AR solution decomposed about 70% after 2 h of
irradiation with visible light. It was found that the self-photolysis
decomposition rate of AR solution was about 25% in 2 h in the
same conditions. The measurement of TOC reveals the average
mineralization rate is 24.4% after 2 h of visible light irradiation.
The band gap of AgAlO2 is 2.8(1) eV, which means that
AgAlO2 could show photoabsorption in the visible light region
(λ g 420 nm), but the photoabsorption is weak. In the catalytic
reaction, the surface area is an important factor for the catalytic
2
The calculated energy band is shown in Figure 2a. The Fermi
energy, defined as the highest occupied energy level, has been
taken as the valence band maximum (VBM). The VBM is
located at the boundary of the Brillouin zone, U point. And the
conduction band minimum (CBM) is situated at the middle of
the Brillouin zone, Γ point. This means that AgAlO2 is an
indirect-gap semiconductor material. A minimum band gap
between VBM and CBM is about 1.1(6) eV, which it means
that AgAlO2 has the ability to respond to the visible light.
Figure 2b shows the total density of states (TDOS) and partial
density of states (PDOS), corresponding to the energy region
in Figure 2a. Table 4 presents the projected weights of wave
functions at the valence band top and the conduction band
bottom. It is clear that the valence band top mainly consists of
O-2p orbitals and Ag-4d orbitals, where Ag-4d5s5p, and Al-
3s3p orbitals hybridize with O-2p. The DOS peaks at the top
of the valence band are constructed of both O-2p and Ag-4d
orbitals, and the contribution of Ag-5s5p and Al-3s3p orbitals
is negligibly small. The conduction band bottom is mainly
constructed of the Ag-5s5p orbitals, and the contribution of the
Al and O electronic orbitals is small. These means that AgAlO2
is a suitable material to be selected to investigate the individual
contribution of Ag to the valence band and the conduction band.
activity. The surface area of the AgAlO sample is only 3.9
2
m /g. Since an efficient photocatalytic reaction process occurs
on the photocatalyst surface, the smaller surface area of AgAlO2
might lead to the decrease of the effective photocatalytic
reaction. The study of the effects of surface area on the effective
photocatalytic reaction is in progress, which is believed to further
help in understanding the new photocatalysts.
Photophysical and Photocatalytic Properties. The XRD
spectra show that our sample is AgAlO2 of space group Pna21
(in Figure 3), because the indexed result is in good agreement
with that reported in the JCPDS database card No. 21-1070,

11680 J. Phys. Chem. B, Vol. 110, No. 24, 2006
Ouyang et al.
Figure 2. The calculated energy band and density of states (DOS) of AgAlO
semiconductor and its E
orbitals, while the conduction band bottom is mainly constructed of the Ag-5s5p orbitals.
2
: (a) the energy bandsthe AgAlO
) 1.1(6) eV; (b) the density of statessthe valence band top mainly consists of the hybridized Ag-4d orbitals and O-2p
2
crystal is an indirect band gap
g
2
Figure 4. AR decomposition over the AgAlO sample was performed
under visible light (λ > 420 nm) at room temperature (ca. 30 °C) in
air [the powdered photocatalyst, 0.3 g; the AR solution, 16 mg/L, 100
mL (pH ∼7.5)]. The AR solution decomposed about 70% after 2 h of
irradiation with visible light, and the photocatalytic efficiency of
AgAlO was about three times the self-photolysis efficiency (the self-
2
photolysis decomposition rate of AR solution was about 25% in 2 h in
the same conditions).
2 2
Figure 3. The XRD patterns of AgAlO , and (1) Ag and (V) LiAlO
peaks.
TABLE 4: Projected Weights of Wave Functions at the
Valence Band and the Conduction Band
elements
Ag
orbitals
VB, %
CB, %
5s
2
2
45
1
22
42
2
5
4
p
d
the decrease of concentration of the AR might result from self-
photolysis decomposition. The light was irradiated through the
suspension, but the reaction did not begin, which means the
reaction is not a photochemical reaction. After the photocatalytic
reaction, the AgAlO2 sample was analyzed by XRD again. The
XRD spectra showed that the sample remained AgAlO2 of space
group Pna21 and also included the Ag and LiAlO2 impurity (in
Figure 3). It is known that the photooxidation/photodissolution
of catalyst might occur on the photocatalyst surface in the
photocatalytic reaction. The samples before and after reaction
were analyzed by XPS. There was no shift for the peaks of
Ag-3d and Al-2p in the surface of samples before and after
reaction. This suggests that the surface of AgAlO2 is stable in
the photocatytic reaction, which means that the sample was
stable in the photocatalytic reaction. The above proves the
reaction of AR decomposition was photocatalysis.
Al
3s
2
3
2s
p
2
0
24
0
16
1
7
O1
O2
2p
2s
1
2p
24
7
The wavelength dependence of the photocatalytic activity of
a semiconductor is often used to distinguish if the reaction is
really driven by light. In our experiment, the light wavelength
dependence of the AR decomposition rate was observed from
λ g 350 nm (full arc of the Xe lamp, without filter) to λ g 500
nm with different cutoff filters (in Figure 5). If a reaction is
driven by light, the variation of the light wavelength will directly
affect the number of photons entering the reaction system and
then the photocatalytic properties. In the present work, it is
obvious that the variation of the photocatalytic properties over
AgAlO2 was closely relevant to the light wavelength, suggesting
that the AR catalytic degradation over AgAlO2 was truly driven
by light. And when the 500 nm cutoff filter was used, there
was the same decreased concentration of the AR solution (about
In our photocatalytic experiments, the contrast experiment
of the photocatalytic decomposing AR on TiO2 was not
executed, because AgAlO could not exhibit photocatalytic
2
activity at the irradiation light wavelength of λ g 500 nm (this
cutoff wavelength is only a little larger than the light absorbable
limitation wavelength of AgAlO2, about 50-60 nm), and it
could be concluded that the photocatalytic decomposing AR
10%) with and without the AgAlO2 photocatalyst, whcih means

Characterization of a Novel Photocatalyst AgAlO2
J. Phys. Chem. B, Vol. 110, No. 24, 2006 11681
consists of hybridized Bi-6s orbitals and O-2p orbitals, and the
conduction band is mainly constructed of Bi-6p orbitals; (3) in
2
3
CaIn2O4, the valence band consists of O-2p orbitals, and the
conduction band is constructed of In-5s orbitals. But the band
structure of AgAlO2 is different from any mentioned above.
Therefore, though AgAlO2 does not exhibit as high activity as
the materials mentioned above, special element composition and
band structure make AgAlO2 a novel photocatalyst. And such
band structure supplies some different ideas for developing new
photocatalysts.
For the photocatalyst AgAlO2, the valence band top mainly
consists of O-2p orbitals and Ag-4d orbitals. This seems that
the mobility of holes generated by band-gap excitation is
expected to be high, because the holes could be excited from
the hybridized Ag-4d orbitals and O-2p orbitals, so the electronic
structure of the valence band top is more preferable for
photocatalysis than the InTaO4, which consists of a valence band
2
Figure 5. The UV-visible diffuse reflectance spectra of AgAlO and
the wavelength dependence of the AR decomposition rate with different
2
cutoff filters after light irradiation for 120 min over AgAlO at room
temperature (ca. 30 °C) in air: when the full arc of the Xe lamp was
use as the light source, the photocatalytic decomposition rate reached
2
9
top of only O-2p orbitals. But the energy band in Figure 2a
shows the valence band is very flat. The relativity of E and V
is know clearly V ) 3kE(k)/p, where E is the energy of the
electron and V is the velocity of the electron. If 3kE(k) is small
9
5%, and when the 500 nm cutoff filter was used in the photocatalytic
experiment, the decomposition rate decreased to about 10% (the value
was nearly equal to the self-photolysis decomposition rate in the same
conditions).
(band is flat), V is also small. So, the valence band being flat
means that holes are easily localized round Ag or O or move
very slowly, though they could be transferred along O-Ag-O
chains. This implies that the contribution of holes to photoca-
talysis is very limited. However, in contrast to the valence band,
the conduction band is mainly constructed of Ag-5s5p orbitals,
and the conduction band is very abrupt, suggesting that the
mobility of the electrons was strong. Therefore, to a certain
extent, such a condition is a possible advantage for restraining
recombination of the charge carriers.
on AgAlO2 was an intrinsic photocatalytic reaction but not a
dye-assisted photocatalytic mechanism, but the photocatalytic
decomposing of AR over TiO2 was a dye-assisted photocatalytic
reaction.22 Because of the dye-assisted photocatalytic mecha-
nism, assisted by the different dyes, the photocatalysts could
exhibit photocatalytic activity at the cutoff wavelength much
more than at their light absorbable limitation wavelength, even
more than 200 nm, such as CaIn2O4.23
It was also observed that the AgAlO2 photocatalyst is effective
in the gas-medium photocatalytic reaction under 300 W Xe arc
lamp irradiation and visible light irradiation (λ g 420 nm). The
results of the acetaldehyde photocatalytic decomposition are
showed in Figure 6.
The crystal structure influences the energy band and electronic
structures. And another essential of the crystal structure, which
was usually neglected, is that the crystal structure impacts the
move efficiency of the charge carriers. The AgAlO structure
2
has been introduced above, and the detailed crystal structure
Discussion. First, for the past several decades, research on
the photocatalysis surrounded several elements, such as Ti, Ta,
Nb, V, W, Mo, In, Ga, and Bi, but AgAlO2 does not include
these elements. Second, the multimetal oxide materials which
could photocatalytically decomposed organic contaminations
will be discussed below. According to calculation results, the
AgO4 tetrahedron planes perform the main transmission job of
the charge carriers, and the contribution of the AlO4 tetrahedron
planes is negligibly small. So the bond angles of Ag-O-Ag
and O-Ag-O are regarded here. In the AgAlO2 structure, the
bond angles of O-Ag-O and Ag-O-Ag range from 99.3° to
131.6° and from 83.5° to 86.5°, respectively (in Table 3). But
because of the high symmetry, in the perovskite structure, the
bond angles of O-metal-O and metal-O-metal are all about
180°. This means the charge carriers could move to the sample
surface through fewer atoms in the perovskite structure than in
24
25
with high activity are listed here: (1) Bi2WO6, BiVO4, PbBi2-
Nb2O9, 26 and ZnBi12O20 sand in these photocatalysts, the
valence bands mainly consist of hybridized s orbitals, p orbitals,
and d orbitals, and the conduction bands are mainly constructed
by s orbitals, p orbitals, or hybridized s orbitals and p orbitals;
27
(
2) in the photocatalyst CaBi2O4,28 the valence band mainly
Figure 6. CH
00 W Xe lamp, full arc; catalyst, 0.3 g; a gas-closed system with a gas-circulated pump; initial concentration of CH
The decomposition rate of CH CHO, (b) the mineralization rate of CH CHO; (b) Reaction conditions: 300 W Xe lamp, λ g 420 nm; catalyst, 0.3
g; a gas-closed system with a gas-circulated pump; initial concentration of CH CHO, about 680 ppm. (9) The decomposition rate of CH CHO, (b)
the mineralization rate of CH CHO.
3
CHO photocatalytic decomposition over AgAlO
2
under ultraviolet + visible light (a) and visible light (b). (a) Reaction conditions:
3
3
CHO about 1270 ppm. (9)
3
3
3
3
3

11682 J. Phys. Chem. B, Vol. 110, No. 24, 2006
Ouyang et al.
the AgAlO2 structure. From the crystal structure viewpoint, the
more ideal model for the most effective transfer of the charge
carriers is mentioned below: in a crystal that is constructed of
MO6 octahedral layers, these octahedra should be equilateral
polyhedra and connect each other by sharing vertexes, at the
same time the metal-O-metal and O-metal-O angles are
nearly 180° (this means that the crystal structure has high
Acknowledgment. The generous financial support of this
work from the National Natural Science Foundation of China
(Nos. 20373025 and 50472067) is gratefully acknowledged.
References and Notes
(1) Honda, K.; Fujishima, A. Nature 1972, 238, 37.
(2) Asahi, R.; Morikawa, T.; Ohwaki, T.; et al. Science 2001, 293,
symmetry, which usually belongs to a cubic or tetragonal
269.
(3) Domen, K.; Hara, M.; Kondo, J. N.; Tanaka, T. Bull. Chem. Soc.
Jpn 2000, 73, 1307.
4) Hara, M.; Chiba, E.; Ishikawa, A.; et al. J. Phys. Chem. B 2003,
07, 13441.
system), such as SrTiO3;3
0,31
in a crystal that is composed of
MO4 tetrahedral layers, these tetrahedra should be freakish
polyhedra to augment bond angles and connect each other by
sharing vertexes, at the same time the metal-O-metal and
O-metal-O angles are nearly 180° (this implies that the crystal
structure has lower symmetry, which usually belongs to a
triclinic or monoclinic system); but as to the crystals which have
general symmetry, the crystal structure that consist of both MO6
octahedra and MO4 tetrahedra should be more favorable, such
as InVO4.9
Now the projected weights of wave functions at the valence
band and the conduction band and the detailed crystal structure
in AgAlO2 have been analyzed intensively. To further develop
the Ag-system multimetal oxides, Al should be exchanged for
the element, which could improve the existing electronic
structure and crystal structure in AgAlO2. A series of new Ag-
system multimetal oxides have been investigated, and their
properties will be reported somewhere.
(
1
9
4
(
5) Hara, M.; Hitoki, G.; Takata, T.; et al. Catal. Today 2003, 78, 555.
(6) Hara, M.; Takata, T.; Kondo, J. N.; Domen, K. Catal. Today 2004,
0, 313.
7) Zou, Z.; Ye, J.; Arakawa, H. Chem. Phys. Lett. 2000, 332, 271.
8) Zou, Z.; Ye, J.; Sayama, K.; Arakawa, H. Nature (London) 2001,
14, 625.
(9) Ye, J.; Zou, Z.; Arakawa, H.; et al. J. Photochem. Photobiol. A:
Chem. 2002, 148, 79.
10) Zou, Z.; Arakawa, H. J. Photochem. Photobiol. A: Chem. 2003,
58, 145.
(
(
(
1
(
(
11) Tang, J.; Zou, Z.; Ye, J. J. Phys. Chem. B 2003, 107, 14265.
12) Kato, H.; Kobayashi, H.; Kudo, A. J. Phys. Chem. B 2002, 106,
12441.
(13) Kabalkina, S. S.; Popova, S. V.; Serebrjanaja, N. R.; et al. Dokl.
Akad. Nauk SSSR 1963, 152, 853.
(
14) Wyckoff, R. W. G. Am. J. Sci. 1921-1938, Ser. 5, 1; 1922, Ser. 3,
84.
(15) Norby, P.; Dinnebier, R. E.; Fitch, A. N. Inorg. Chem. 2002, 41
14), 3628.
1
(
(
(
16) Wolcyrz, M.; Lukaszewski, M. Z. Kristallogr. 1986, 177, 53.
17) Li, J.; Sleight, A. W. J. Solid State Chem. 2004, 177, 889.
IV. Conclusions
The energy band, crystal, and electronic structures of AgAlO2
were investigated intensively. The calculated results demon-
strated that AgAlO2 is a suitable material to be selected to
investigate the individual contribution of Ag to the valence band
and the conduction band. As an indirect band gap semiconduc-
tor, the calculated indirect band gap Eg ) 1.1(6) eV and the
experimental estimated band gap was 2.8(1) eV. The photo-
catalytic activity of AgAlO2 was characterized by photocatalytic
decomposing AR under visible light irradiation. The advantages
and disadvantages of the AgAlO2 crystal and electronic struc-
tures were also analyzed. It will be useful to carry out further
research in developing the new Ag-system multimetal oxides.
It was shown that the energy band analysis combining with the
crystal structure analysis is a feasible approach to develop new
photocatalysts which could exhibit photocatalytic activity under
visible light.
(18) Segall, M. D.; Lindan, P. L. D.; Probert, M. J.; et al. J. Phys.:
Condens. Matter 2002, 14, 2717.
(19) Hohenberg, P.; Kohn, W. Phys. ReV. B 1964, 136, 864-871.
(20) Hedin, L.; Lundqvist, S. Solid State Phys. 1969, 23, 1.
(21) Hybertsen, M. S.; Louie, S. G. Phys. ReV. B 1986, 34, 5390.
(22) Liu, G.; Li, X.; Zhao, J. J. Mol. Catal. A: Chem. 2000, 153, 221.
(23) Tang, J.; Zou, Z.; Ye, J. Chem. Mater. 2004, 16, 1644.
(24) Kudo, A.; Hijii, S. Chem. Lett. 1999, 10, 1103.
(
29.
25) Kudo, A.; Ueda, K.; Kato, H.; et al. Catal. Lett. 1998, 53 (3-4),
2
(
26) Kim, H. G.; Hwang, D. W.; Lee, J. S. J. Am. Chem. Soc. 2004,
1
26, 8912.
(27) Tang, J.; Ye, J. Chem. Phys. Lett. 2005, 410, 104.
(28) Tang, J.; Zou, Z.; Ye, J. Angew. Chem., Int. Ed. 2004, 43, 4463.
(29) Matsushima, S.; Obata, K.; Nakamura, H.; et al. J. Phys. Chem.
Solid 2003, 64, 2417.
(30) Hutton, J.; Nelmes, R. J. J. Phys. C 1981, 14, 1713.
(31) Domen, K.; Nalto, S.; Onlshl, T.; et al. J. Phys. Chem. 1982, 86,
3657.