Visible-light-driven Cu(II)-(Sr1- yNay)(Ti 1- xMox)O3 photocatalysts based on conduction band control and surface ion modification

Article abstract of DOI:10.1021/ja105846n

Band-gap narrowing is generally considered to be a primary method in the design of visible-light-active photocatalysts because it can decrease the photo threshold to lower energies. However, controlling the valence band by up-shifting the top of the band or inducing localized levels above the band results in quantum efficiencies under visible light much lower than those under UV irradiation (such as those reported for N-doped TiO₂: Science 2001, 293, 269. J. Phys. Chem. B 2003, 107, 5483). Herein, we report a systematic study on a novel, visible-light-driven photocatalyst based on conduction band control and surface ion modification. Cu(II)-(Sr _1-yNa_y)(Ti_1-xMo_x)O₃ photocatalysts were prepared by a soft chemical method in combination with an impregnation technique. It is found that Mo⁶⁺ as well as Na ⁺ doping in the SrTiO₃ can lower the bottom of the conduction band and effectively extend the absorption edge to the visible light region. The Cu(II) clusters grafted on the surface act as a co-catalyst to efficiently reduce the oxygen molecules, thus consuming the excited electrons. Consequently, photocatalytic decomposition of gaseous 2-propanol into CO ₂ is achieved, that is, CH₃CHOHCH₃ + ⁹/₂O₂ → 3CO₂ + H₂O. For Cu(II)-(Sr_1-yNa_y)(Ti_1-xMo _x)O₃ at x = 2.0% under visible light irradiation, the maximum CO₂ generation rate can reach 0.148 μmol/h; the quantum efficiency under visible light is calculated to be 14.5%, while it is 10% under UV light irradiation. Our results suggest that high visible light photocatalytic efficiency can be achieved by combining conduction band control and surface ion modification, which provides a new approach for rational design and development of high-performance photocatalysts.

Full text of DOI:10.1021/ja105846n

Published on Web 10/08/2010
Visible-Light-Driven Cu(II)-(Sr1-yNa )(Ti Mo )O3
y
1-x
x
Photocatalysts Based on Conduction Band Control and
Surface Ion Modification
†
,‡
†
⊥
Xiaoqing Qiu, Masahiro Miyauchi,* Huogen Yu, Hiroshi Irie, and
,
†,§
Kazuhito Hashimoto*
Research Center for AdVanced Science and Technology, The UniVersity of Tokyo, 4-6-1
Komaba, Meguro-ku, Tokyo 153-8904, Japan, National Institute of AdVanced Industrial Science
and Technology, Tsukuba Central 5, 1-1-1 Higashi, Tsukuba, Ibaraki 305-8565, Japan, Clean
Energy Research Center, UniVersity of Yamanashi, 4-3-11 Takeda, Kofu, Yamanashi 400-8511,
Japan, and Graduate School of Engineering, The UniVersity of Tokyo, 7-3-1 Hongo, Bunkyo-ku,
Tokyo 113-8656, Japan
Received July 2, 2010; E-mail: m-miyauchi@aist.go.jp; hashimoto@light.t.u-tokyo.ac.jp
Abstract: Band-gap narrowing is generally considered to be a primary method in the design of visible-
light-active photocatalysts because it can decrease the photo threshold to lower energies. However,
controlling the valence band by up-shifting the top of the band or inducing localized levels above the band
results in quantum efficiencies under visible light much lower than those under UV irradiation (such as
those reported for N-doped TiO
report a systematic study on a novel, visible-light-driven photocatalyst based on conduction band control
and surface ion modification. Cu(II)-(Sr1-yNa )(Ti1-xMo )O photocatalysts were prepared by a soft chemical
method in combination with an impregnation technique. It is found that Mo as well as Na doping in the
SrTiO can lower the bottom of the conduction band and effectively extend the absorption edge to the
2
: Science 2001, 293, 269. J. Phys. Chem. B 2003, 107, 5483). Herein, we
y
x
3
6+
+
3
visible light region. The Cu(II) clusters grafted on the surface act as a co-catalyst to efficiently reduce
the oxygen molecules, thus consuming the excited electrons. Consequently, photocatalytic decomposition
of gaseous 2-propanol into CO
2
is achieved, that is, CH
3
CHOHCH
3
+ 9
/
2
O
2
f 3CO
2
+ H
2
O. For
Cu(II)-(Sr1-yNa )(Ti1-xMo )O at x ) 2.0% under visible light irradiation, the maximum CO
y
x
3
2
generation
rate can reach 0.148 µmol/h; the quantum efficiency under visible light is calculated to be 14.5%, while it
is 10% under UV light irradiation. Our results suggest that high visible light photocatalytic efficiency can be
achieved by combining conduction band control and surface ion modification, which provides a new approach
for rational design and development of high-performance photocatalysts.
1
. Introduction
Considering the current energy and environmental issues at
light, many efforts have been made to develop visible-light-
driven photocatalysts. Typically, to tailor the band-gap energies
3
of photocatalysts and extend their absorption to the visible light
region, cationic doping with transition metal ions, such as Cr,
Mn, Fe, Pb, Cu, and so on, or anionic doping with N, C, and S
the global level, heterogeneous photocatalysis using semicon-
ductors has a great potential for industrial uses. In general,
high performance of photocatalysts requires a high redox
potential of photogenerated charge carriers, that is, high
oxidation power of the holes in the valence band (VB) and
enough reduction power of the electrons in the conduction band
1
4
2
is commonly used. Even though most photocatalysts doped in
this way can exhibit visible light sensitivity, their quantum
efficiencies (QEs) under visible light are much lower than those
5
under UV irradiation. In the case of cationic doping, the
(
CB). Therefore, the most efficient photocatalysts are usually
sensitivity to visible light is mainly due to high levels of
impurities in the forbidden band of the photocatalysts, which
unfortunately serve as recombination centers for photoinduced
wide-gap semiconductors, such as TiO , ZnO, and SrTiO . These
2
3
wide-band-gap semiconductors only show activity under ultra-
violet (UV) light irradiation, which limits their practical
applications. For effective utilization of solar energy and indoor
(
3) (a) Zou, Z. G.; Ye, J. H.; Sayama, K.; Arakawa, H. Nature 2001,
414, 625. (b) Rodrigues, S.; Ranjit, K. T.; Uma, S.; Martyanov, I. N.;
Klabunde, K. J. AdV. Mater. 2005, 17, 2467. (c) Maeda, K.; Teramura,
K.; Lu, D. L.; Takata, T.; Saito, N.; Inoue, Y.; Domen, K. Nature
2006, 440, 295. (d) Qiu, X. F.; Zhao, Y. X.; Burda, C. AdV. Mater.
2007, 19, 3995. (e) Wang, D. F.; Kako, T.; Ye, J. H. J. Am. Chem.
Soc. 2008, 130, 2724. (f) Nakamura, R.; Okamoto, A.; Osawa, H.;
Irie, H.; Hashimoto, K. J. Am. Chem. Soc. 2007, 129, 9596. (g) Weare,
W. W.; Pushkar, Y.; Yachandra, V. K.; Frei, H. J. Am. Chem. Soc.
2008, 130, 11355. (h) Li, G. H.; Dimitrijevic, N. M.; Chen, L.; Rajh,
T.; Gray, K. A. J. Phys. Chem. C 2008, 112, 19040.
†
Research Center for Advanced Science and Technology, The University
of Tokyo.
‡
National Institute of Advanced Industrial Science and Technology.
Clean Energy Research Center, University of Yamanashi.
Graduate School of Engineering, The University of Tokyo.
⊥
§
(
(
1) (a) Thompson, T. L.; Yates, J. T. Chem. ReV. 2006, 106, 4428. (b)
Kudo, A.; Miseki, Y. Chem. Soc. ReV. 2009, 38, 253. (c) Hoffmann,
M. R.; Martin, S. T.; Choi, W. Y.; Bahnemann, D. W. Chem. ReV.
1
995, 95, 69. (d) Chen, X.; Mao, S. S. Chem. ReV. 2007, 107, 2891.
2) Ohtani, B. Chem. Lett. 2008, 37, 216.
10.1021/ja105846n  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 15259–15267 9 15259

A R T I C L E S
Qiu et al.
1
a,6
charger carries. On the other hand, anionic doping generally
contributes to VB control by inducing localized levels above
the VB. The holes generated in the localized levels have less
oxidation power and mobility than those in the VB, which results
that CB control might be a good way to enhance visible light
absorption, leaving the oxidation power of the holes unchanged.
Moreover, combining CB control and efficient oxygen reduction
via Cu ions grafted on the photocatalysts’ surfaces is a novel
approach that is likely to provide an opportunity for designing
highly efficient visible-light-driven photocatalysts. We recently
5
in low photocatalytic performance. In fact, oxidative holes in
the VB can drive the complete mineralization of organic
pollutants, once the photogenerated electrons are efficiently
6
+
demonstrated that CB-controlled TiO co-doped with W and
2
2
3+
consumed in the presence of oxygen molecules in the air.
Ga showed visible light sensitivity after Cu(II) ions were
Therefore, it has significant importance to design photocatalysts
that utilize the high oxidation power of holes in the VB under
visible light irradiation.
grafted on the surfaces, even though the visible light absorption
1
0
was rather low.
Bearing these issues in mind, we selected strontium titanate
SrTiO ) as a wide-band-gap semiconductor for our study
system, based on the following considerations: (1) SrTiO , one
Recently, members of our laboratory designed visible-light-
(
3
2
+
7
driven photocatalysts by grafting Cu ions onto TiO
Under visible light irradiation, electrons in the VB of TiO
be transferred to Cu(II) clusters, i.e., interfacial charge transfer
2
surfaces.
3
2
can
of the most promising semiconductors, has versatile applications
in the electronics industry and photocatalysis due to its merits
8
(
IFCT), resulting in the formation of Cu(I) clusters. Cu(I)
of thermal and structural stability and resistance to photocor-
7
clusters can efficiently reduce oxygen molecules, thus consum-
ing the excited electrons. The holes generated in the VB can
decompose the organic compounds. In other words, this system
can take advantage of both the efficient oxygen reduction via
Cu ions and the high oxidation power of the holes in the VB of
11
rosion. (2) Most importantly, SrTiO
3
has a perovskite cubic
, which provides the
structure with the general formula ABO
3
flexibility to vary the composition of A and B sites to form
substituted nonstoichiometric perovskite and tune the electronic
12
structure. Further, the electronic structure of SrTiO
3
resembles
2
TiO induced by IFCT under visible light irradiation. However,
that of TiO
photogenerated holes can oxidize most organic compounds.
On the basis of our calculations of electronic density of states
DOS), we design conduction-band-controlled SrTiO by co-
2
, whose valence band is deep, and therefore
IFCT is an interface phenomenon, and the IFCT absorption is
11b
8
rather weak. It should be noted that tungsten trioxide (WO
has the proper band gap for good visible light absorption.
Despite the high oxidation power of the holes in the VB, WO
3
)
(
3
3
6+
doping. In the present study, we found that Mo as well as
Na is an optimum dopant for lowering the conduction band
level. We synthesized SrTiO
a simple hydrothermal reaction. After Cu(II) modification on
the surfaces of doped SrTiO , the photocatalysts exhibit high
was once considered to be inactive because of the low potential
of electrons in the CB. Encouragingly, upon modification of
WO with co-catalysts such as Pt, Pd, WC, CuO, or Cu(II)
3
+
6+
+
3
co-doped with Mo and Na by
clusters, the visible light activity was drastically enhanced via
9
3
the efficient oxygen reduction process. Previous studies sug-
activities for oxidation of gaseous 2-propanol (IPA) to CO
under visible light irradiation.
2
3
gested that the photogenerated electrons in WO are injected
into these co-catalysts, and these injected electrons cause multi-
electron reductions to produce active species such as hydrogen
9
2. Experimental Section
peroxides. However, the chemical stability of WO
3
is low, and
tungsten is an expensive, rare metal; thus, it is not suitable for
industrial applications. These elaborate works give us a hint
y 1-x x 3
2.1. Synthesis of (Sr1-yNa )(Ti Mo )O Samples. The
(Sr1-yNa
method. Sr(OH)
y
)(Ti1-xMo
x
)O
2
3
samples were prepared by a hydrothermal
, and Na MoO were used
2
·8H O, amorphous TiO
2
2
4
(
4) (a) Asahi, R.; Morikawa, T.; Ohwaki, T.; Aoki, K.; Taga, Y. Science
001, 293, 269. (b) Wang, D. F.; Ye, J. H.; Kako, T.; Kimura, T. J.
as the starting materials. All chemicals were analytical grade, used
without further purification. The synthesis procedure is described
2
Phys. Chem. B 2006, 110, 15824. (c) Kudo, A.; Niishiro, R.; Iwase,
A.; Kato, H. Chem. Phys. 2007, 339, 104. (d) Kato, H.; Kudo, A. J.
Phys. Chem. B 2002, 106, 5029. (e) Irie, H.; Maruyama, Y.;
Hashimoto, K. J. Phys. Chem.C 2007, 111, 1847. (f) Xie, T. H.; Sun,
X. Y.; Lin, J. J. Phys. Chem. C 2008, 112, 9753. (g) Miyauchi, M.;
Takashio, M.; Tobimatsu, H. Langmuir 2004, 20, 232. (h) Sun, X. Y.;
Lin, J. J. Phys. Chem. C 2009, 113, 4970. (i) Wang, D. F.; Ye, J. H.;
Kako, T.; Kimura, T. J. Phys. Chem. B 2006, 110, 15824. (j) Mrowetz,
M.; Balcerski, W.; Colussi, A. J.; Hoffmann, M. R. J. Phys. Chem. B
as follows: 5.315 g (0.02 mol) of Sr(OH)
2
·8H
2
O, 1.597 g (0.02
MoO were
mol) of amorphous TiO , and given amounts of Na
2
2
4
added into 80 mL of distilled water to form a white suspension.
This suspension was transferred to a 100 mL Teflon-lined, stainless
steel autoclave and allowed to react at 140 °C for 20 h. The
autoclave was then cooled naturally to room temperature. The
products were harvested by pressure filtration with a membrane
filter (0.025 µm, Millipore), washed with diluted HCl and distilled
water, and dried in air at ambient temperature. To improve the
crystallinity, the products were calcined at 600 °C for 10 h. The
2
004, 108, 17269.
5) Irie, H.; Watanabe, Y.; Hashimoto, K. J. Phys. Chem. B 2003, 107,
483.
6) (a) Konta, R.; Ishii, T.; Kato, H.; Kudo, A. J. Phys. Chem. B 2004,
08, 8992. (b) Choi, W. Y.; Termin, A.; Hoffmann, M. R. J. Phys.
(
(
5
samples thus obtained were denoted as (Sr1-yNa
y
)(Ti1-xMo
x
)O
3
.
1
+
Chem. 1994, 98, 13669.
To get the samples with different Na contents and roughly fixed
Mo6 content, parallel control experiments were performed under
identical conditions.
+
(
(
7) (a) Irie, H.; Miura, S.; Kamiya, K.; Hashimoto, K. Chem. Phys. Lett.
2
008, 457, 202. (b) Irie, H.; Kamiya, K.; Shibanuma, T.; Miura, S.;
Tryk, D. A.; Yokoyama, T.; Hashimoto, K. J. Phys. Chem. C 2009,
13, 10761.
8) (a) Creutz, C.; Brunschwig, B. S.; Suntin, N. J. Phys. Chem. B 2005,
09, 10251. (b) Creutz, C.; Brunschwig, B. S.; Suntin, N. J. Phys.
Chem. B 2006, 110, 25181. (c) Hush, N. S. Electrochim. Acta 1968,
Control Experiment I: 5.315 g (0.02 mol) of Sr(OH)
.597 g (0.02 mol) of amorphous TiO , 0.106 g (0.513 mmol) of
NaMoO , and 10 g of NaCl were used as the starting materials.
2
2
·8H O,
1
1
2
1
4
1
3, 1005. (d) Hush, N. S. J. Electroanal. Chem. 1999, 460, 5. (e)
Hush, N. S. J. Electroanal. Chem. 1999, 470, 170.
(10) Yu, H. G.; Irie, H.; Hashimoto, K. J. Am. Chem. Soc. 2010, 132, 6898.
(11) (a) Bolts, J. M.; Wrighton, M. S. J. Phys. Chem. 1976, 80, 2640. (b)
Miyauchi, M.; Nakajima, A.; Fujishima, A.; Hashimoto, K.; Watanabe,
T. Chem. Mater. 2000, 12, 3. (c) Konta, R.; Ishii, T.; Kato, H.; Kudo,
A. J. Phys. Chem. B 2004, 108, 8992.
(
9) (a) Abe, R.; Takami, H.; Murakami, N.; Ohtani, B. J. Am. Chem. Soc.
2
008, 130, 7780. (b) Zhao, Z. G.; Miyauchi, M. Angew. Chem., Int.
Ed. 2008, 47, 7051. (c) Arai, T.; Horiguchi, M.; Yanagida, M.; Gunji,
T.; Sugihara, H.; Sayama, K. Chem. Commun. 2008, 5565. (d) Kim,
Y. H.; Irie, H.; Hashimoto, K. Appl. Phys. Lett. 2008, 92, 182107. (e)
Arai, T.; Yanagida, M.; Konishi, Y.; Iwasaki, Y.; Sugihara, H.;
Sayama, K. Catal. Commun. 2008, 9, 1254.
(12) (a) Pena, M. A.; Fierro, J. L. G. Chem. ReV. 2001, 101, 1981. (b)
Chan, N. H.; Sharma, R. K.; Smyth, D. M. J. Electrochem. Soc. 1981,
128, 1762.
1
5260 J. AM. CHEM. SOC. 9 VOL. 132, NO. 43, 2010

Cu(II)-(Sr1-yNa )(Ti1-xMo )O
y x 3
Photocatalysts
A R T I C L E S
3 y x 3
Figure 1. Density of states of (a) SrTiO and (b) (Sr1-yNa )(Ti1-xMo )O (at x ) 0.333). The vertical dashed line represents the Fermi level.
The remaining procedures were the same as those described above.
The sample was named sample V.
and a quartz cover, evacuated, and filled with fresh synthetic air.
To eliminate the sample surface organic contaminants, the vessel
was pre-illuminated with a Xe lamp (Luminar Ace 210, Hayashi
Control Experiment II: 5.315 g (0.02 mol) of Sr(OH)
.597 g (0.02 mol) of amorphous TiO , and 0.074 g (0.513 mmol)
of MoO instead of NaMoO were used as the starting materials.
2
·8H
2
O,
1
2
2
Tokei Works) until the rate of CO generation was less than 0.02
3
4
µmol/day. The vessel was re-evacuated and refilled with fresh
synthetic air. The pressure inside the vessel was kept at about 1
atm. Next, 300 ppm of gaseous IPA was injected into the vessel.
Prior to light irradiation, the vessel was kept in the dark for a
sufficient time to ensure the establishment of absorption/desorption
equilibrium of IPA on the surfaces of the photocatalysts. Subse-
quently, the vessel was irradiated under a Xe lamp. A combination
of glass filters (B-47, L-42, and C-40C, AGC Techno Glass) was
used to obtain visible light with wavelength from 400 to 530 nm.
The light intensity was determined by a spectroradiometer (USR-
The remaining procedures were the same as those described above.
The sample was named sample VI.
2
.2. Modification of (Sr1-yNa
y
)(Ti1-xMo
x
)O
3
Samples with
Cu(II) Ions. The Cu(II)-(Sr1-yNa
y
)(Ti1-xMo
x
)O
3
photocatalysts
were prepared by an impregnation technique that was described in
7
detail in our previous work. In a typical preparation, 1 g of
(
Sr1-yNa
y
)(Ti1-xMo
x
)O
3 2
sample was dispersed into 10 mL of CuCl
2+
solution in a vial reactor. The weight fraction of Cu relative to
(
sealed vial reactor was heated at 90 °C for 1 h using a water bath.
The suspension was then filtered and washed with a sufficient
amount of distilled water. The final products were dried at 110 °C
for 24 h and subsequently ground into a powder using an agate
mortar.
-3
Sr1-yNa
y
)(Ti1-xMo
x
)O was set to be 1 × 10 . Under stirring, the
2
40D, Ushio) and set to be 1 mW/cm . To fully understand the
photocatalytic performance of the samples, a parallel experiment
was conducted under UV light (310-400 nm) irradiation, which
is obtained by using a D-36 glass filter. The UV light intensity
was adjusted by controlling the absorption photon number of the
catalysts to be the same as that under visible light irradiation. During
the light irradiation, 1 mL of gaseous sample was periodically
extracted from the reaction vessel to monitor the concentrations of
2
.3. Sample Characterization. Elemental analyses of the
samples were performed using an inductively coupled plasma
atomic emission spectrometer (ICP-AES, P-4010, Hitachi) for Ti,
Sr, Mo, and Cu and a polarized Zeeman atomic absorption
spectrophotometer (Z-2000, Hitachi) for Na. The structural char-
acteristics of the samples were measured by powder X-ray
diffraction (XRD) at room temperature on a Rigaku D/MAX25000
diffractometer with a copper target (λ ) 1.54178 Å). Ni powder
IPA, acetone, and CO
Shimadzu Co., Ltd.).
2
using a gas chromatograph (model GC-8A,
2.5. Calculations of the Electronic Band Structure. The
electronic structure of the samples was calculated using plane-wave-
based density functional theory with the CASTEP program package.
The ionic cores of the elements were represented by scalar
relativistic, fully separable, ultrasoft pseudopotentials. The general-
ized gradient approximation (GGA-PBE) was applied, and the
kinetic energy cutoff was selected at 370 eV. The density of the
Monkhorst-Pack k-point mesh was 5 × 5 × 3. The Sr ions and Ti
ions were partly substituted by Na+ ions and Mo ions in the
served as an internal standard for peak position determination. The
data were collected from 2θ ) 20 to 70° in a step-scan mode (step,
0
.02°; counting time, 5 s). The morphologies of the samples were
investigated by field-emission scanning electron microscopy (SEM)
using a JEOL JSM-6700 apparatus and transition electron micros-
copy (TEM) on a JEOL JEM 2010 instrument under an acceleration
voltage of 200 kV. The specific surface areas of the samples were
determined from the nitrogen absorption data at liquid nitrogen
temperature using the Barrett-Emmett-Teller (BET) technique.
The samples were degassed at 200 °C and a pressure below 100
mTorr for a minimum of 2 h prior to analysis using a Micromeritics
VacPrep 061 instrument. The absorption spectra of the samples
were recorded using a UV-2550 spectrophotometer (Shimadzu).
The ionic characteristics and surface composition were studied
by X-ray photoelectron spectroscopy (XPS; Perkin-Elmer model
6+
SrTiO
compositions used for the calculation were SrTiO
)O at x ) 0.333. The system structures were
optimized by first-principles calculations.
3
cell to simulate the (Sr1-yNa
y
)(Ti1-xMo
x
)O
3
system. The
3
and
(Sr1-yNa
y
)(Ti1-xMo
x
3
3. Results and Discussion
3 3
The electronic structures of SrTiO and SrTiO co-doped with
Mo6 as well as Na were investigated on the basis of the plane-
+
+
5
600). The binding energy data are calibrated with the C 1s
signal at 284.6 eV.
.4. Evaluation of Photocatalytic Properties. The decomposi-
wave density function theory program package CASTEP. The
variations of DOS of SrTiO
1
3
with energy are shown in Figure
a. Zero energy corresponds to the Fermi level (E ). The VB
consists of O 2p, Sr 5s, and Ti 3d hybrid orbitals;
and contributes more electrons, while
the bottom of the CB above E is predominantly composed
of Ti 3d-like states. To understand the qualitative impact of
2
F
tion of gaseous IPA was chosen as a probe to evaluate the
photocatalytic activities of the samples in air. A 500 mL cylindrical
glass vessel was used as the photocatalysis reactor. The experiments
were carried out according to the following procedure: 300 mg of
photocatalyst powder was evenly spread on the bottom of a circular
of SrTiO
3
the O 2p is closer to E
F
F
2
+
6+
glass dish with an area of 5.5 cm , which was mounted in the middle
Na and Mo co-doping on the DOS, the unit cells of
)O were constructed by partially substitut-
of the vessel reactor. The vessel was sealed with a rubber O-ring
(Sr1-2yNa
y
)(Ti1-xMo
x
3
J. AM. CHEM. SOC. 9 VOL. 132, NO. 43, 2010 15261

A R T I C L E S
Qiu et al.
+
(
0.605 Å) in six-fold coordination. Therefore, for our study, Na
6+
ions are more likely to be introduced into the A site and Mo
ions into the B site of SrTiO
)O . The preliminary observations imply that
the perovskite structure is not disturbed by the addition of small
3
to form the compound
(
Sr1-yNa
y
)(Ti1-xMo
x
3
amounts of Mo6 and Na during the sample preparation procedure,
+
+
and the incorporation of Mo6 and Na induces no significant
change in the lattice size, due to the similarities in the ions radii.
When the Mo content was further increased to x ) 5%, several
+
+
additional peaks assigned to the secondary phase of SrMoO
4
appeared (Figure S2, Supporting Information). Because of the
detection limitation of XRD measurements, the structure of the
samples was further investigated using SEM, TEM, XPS, and
UV-visible spectroscopy.
y x 3
Morphologies of the (Sr1-yNa )(Ti1-xMo )O samples were
observed by SEM. As can be seen in Figure 3, all the samples
were an assembly of uniformly distributed nanoparticles. The
average grain size of the nanoparticles was about 50 nm.
y x 3
Figure 2. XRD patterns of (Sr1-yNa )(Ti1-xMo )O at x ) 0, 0.25, 0.9,
and 2.0%. Vertical bars below the patterns represent the standard diffraction
data from the JCPDS file for bulk SrTiO (No. 86-0176). Peaks marked
with an asterisk are due to the internal standard nickel.
3
6+
+
Introduction of a small amount of Mo as well as Na into
the SrTiO lattice did not change the morphology or the particle
size. Therefore, the effects of morphology and size on the
3
ing Na ions at the Sr lattice site and Mo6+ ions at the Ti4+
site in cubic perovskite SrTiO . As illustrated in Figure 1b, Mo
d-like states are located below the Ti 3d orbital and contribute
+
2+
1
6
photocatalytic activity can be excluded in the present study.
A TEM image of non-doped SrTiO (at x ) 0) further showed
3
3
4
that the sample consists of particle-like structures (Figure S3,
Supporting Information). The corresponding high-resolution
TEM image shows clear lattice fringes, demonstrating the single-
crystalline nature of these nanoparticles and the high quality of
the as-obtained samples. The shape and size of the samples were
not disturbed by surface modification with Cu(II) either, as
shown in Figure S3. Cu(II) clusters were hardly detected on
the sample surfaces by TEM observation, probably because the
amount of Cu(II) was extremely low and the Cu(II) clusters
are most likely to exist as an amorphous form due to the
to the bottom of the CB, resulting in the band narrowing.
However, the VB remains almost the same as that of non-doped
+
SrTiO
3
. Na ions do not directly contribute to the energy
1
b
structure but construct the crystal structures. The theoretical
calculations demonstrate that Mo and Na+ are optimum
6+
dopants for lowering the CB level.
Elemental analyses of the samples were performed using ICP-
AES for Ti, Sr, Mo, and Cu and a polarized Zeeman atomic
absorption spectrophotometer for Na. The results are listed in
Table S1 (Supporting Information). The Mo contents in the
samples were close to the initial ones in hydrothermal solution.
Although relatively small amounts of Na were also detected in
7a
relatively low grafting temperature, in contrast to noble metal
9
a
deposition on the surface of photocatalysts.
The surface compositions and element chemical states of the
samples were investigated by XPS. Figure S4 (Supporting
Information) gives a series of full-scale XPS spectra of non-
3
the doped samples, we could synthesize SrTiO co-doped with
Mo and Na by a simple hydrothermal method. The relatively
low amount of Na is partly because of the high solubility of
Na in the aqueous solution under hydrothermal conditions. For
all the as-obtained samples, the Ti content was slightly higher
doped SrTiO
Cu(II)-(Sr1-yNa
SrTiO , only Sr, Ti, and O were detected. Moreover, the
spectrum of (Sr1-yNa )O at x ) 2.0% clearly showed
3
, (Sr1-yNa
y
)(Ti1-xMo
)O at x ) 2.0%. For the non-doped
x 3
x
)O
3
at x ) 2.0%, and
y
)(Ti1-xMo
3
than the Sr content. Hereafter, Mo- and Na-doped SrTiO
described as (Sr1-yNa )O
Figure 2 shows the XRD patterns of (Sr1-yNa
3
is
y
)(Ti1-xMo
x
3
y
)(Ti1-xMo
x
3
.
the additional peaks associated with Mo, while the diagnostic
peak of Na overlapped with that of Ti auger. For the
Cu(II)-(Sr1-yNa )(Ti1-xMo )O at x ) 2.0%, the signal of Cu
y x 3
y
)(Ti1-xMo
x
)O
3
samples that were obtained by the hydrothermal method with
heat treatment at 600 °C for 10 h. With varying Mo content x
e 2.0%, the peak position matched well the standard data for
bulk SrTiO (JCPDS card No. 86-0176). Therefore, all the
3
samples crystallized in a cubic perovskite structure with high
crystallinity, as indicated by the sharp XRD peaks. No peaks
associated with rutile or anatase TiO , MoO , or SrMoO were
2 3 4
detected. From the enlarged (110) peaks (Figure S1, Supporting
Information), there was no obvious shift of peak position with
was clearly observed, evidencing that Cu(II) was successfully
grafted on the surfaces. Figure 4a shows the Mo 3d core-level
XPS spectra. The Mo 3d signals increased with the content of
Mo in the samples. It is clear that the Mo 3d spectra consist of
well-defined photoelectron signals located at 235.6 and 232.5
eV, which can be assigned to Mo 3d3/2 and Mo 3d5/2 spin-orbital
1
7
components, respectively. One can safely conclude that
molybdenum in all the samples is present in one oxidation state
6
+
+
varying contents of Mo and Na . It is well known that SrTiO
is a typical ABO perovskite material in which the A site may
accommodate alkaline-earth, alkali, and rare-earth cations. Fujimoto
3
6+
17
(
Mo ), consistent with the literature reports. The Ti 2p core-
3
level spectra showed two well-resolved Ti 2p1/2 and Ti 2p3/2
spectral lines at 463.9 and 458.2 eV (Figure 4b), in agreement
13
+
2+
et al. demonstrated that Na ions substituted for Sr rather than
4+
18
4
+
+
with literature values for Ti . No shoulders associated with
Ti in SrTiO
3
because the 12-coordinated effective radius of Na
2+
14
ions (1.39 Å) is very close to that of Sr (1.44 Å). Devi and
15
6+
4+
Murthy found that Mo ions were prone to replace Ti due to
(14) Shannon, R. D. Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor.
Gen. Crystallogr. 1976, 32, 751.
6+
4+
the good ionic radius matching between Mo (0.62 Å) and Ti
(
(
15) Devi, L. G.; Murthy, B. N. Catal. Lett. 2008, 125, 320.
16) (a) Wang, H. H.; Xie, C. S.; Zhang, W.; Cai, S. Z.; Yang, Z. H.; Gui,
Y. H. J. Hazard. Mater. 2007, 14, 645. (b) Li, D.; Haneda, H.
Chemosphere 2003, 51, 129.
(
13) Fujimoto, M.; Chiang, Y. M.; Roshko, A.; kingery, W. D. J. Am.
Ceram. Soc. 1985, 68, C300.
1
5262 J. AM. CHEM. SOC. 9 VOL. 132, NO. 43, 2010

Cu(II)-(Sr1-yNa )(Ti1-xMo )O
y x 3
Photocatalysts
A R T I C L E S
y x 3
Figure 3. SEM images of (Sr1-yNa )(Ti1-xMo )O at (a) x ) 0, (b) x ) 0.25%, (c) x ) 0.9%, and (d) x ) 2.0%.
Figure 4. (a) Mo 3d core-level spectra of (Sr1-yNa
samples at x ) 0 and 2.0%. (c) Cu 2p core-level spectra of bare and Cu(II)-modified (Sr1-yNa )(Ti1-xMo )O samples at x ) 2.0%.
y x 3
y
)(Ti1-xMo
x
)O
3
y x 3
samples at x ) 0, 0.25, 0.9, and 2.0%. (b) Ti 2p core-level spectra of (Sr1-yNa )(Ti1-xMo )O
Ti3 or Ti2+ was observed at the lower energy side. When the
Mo content was roughly fixed and the Na content in the samples
varied, no obvious differences were seen in the chemical states
+
of Mo and Ti in the samples (Figure S5, Supporting Informa-
tion). Hitosugi et al. studied the microstructure of Nb -doped
5
+
anatase TiO
2
using XPS and found that the incorporation of
J. AM. CHEM. SOC. 9 VOL. 132, NO. 43, 2010 15263

A R T I C L E S
Qiu et al.
Figure 5. UV-visible absorption spectra of samples. (a) (Sr1-yNa
y
)(Ti1-xMo
x
)O
3
y x 3
at x ) 0, 0.25, 0.9, and 2.0%. (b) Cu(II)-(Sr1-yNa )(Ti1-xMo )O at x )
0
and 2.0%. The dashed curves correspond to those of the bare samples.
Nb5 into the lattice of TiO
+
resulted in the formation of minor
room temperature, are displayed in Figure 5a. Non-doped SrTiO
(at x ) 0) shows the intense interband absorption in the UV
2
3
3+
19
components of Ti to keep the charge balance. However, for
our study, Mo6 as well as Na+ doping induced negligible
change of the Ti 2p spectrum. Due to the good ionic radius
matching between Mo and Ti in six-fold coordination, Mo
ions are prone to replace Ti of SrTiO
+
6+
+
region. It should be noted that doping of Mo as well as Na
narrows the band gap of SrTiO and that the absorption edges
shift toward a longer wavelength region with increasing content
3
6+
4+
6+
4
+
15
6+
6+
+
3
.
Besides the ionic
of Mo . While the Mo content is roughly fixed, the Na
radius matching, electroneutrality is another factor to be fulfilled
for an ABO perovskite. The important characteristic of ABO
perovskites is their susceptibility to partial substitution of cations
or deficiencies of cations at both A and B sites. It is well
documented that a strontium vacancy is a favorable point defect
content has no obvious effect on the optical absorbance
properties (Figure S6a, Supporting Information). From the DOS
calculations, it can be concluded that the band gap narrowing
is derived from the lowering of the bottom of the CB while the
potential of the top of the VB is maintained. Compared with
3
3
1
2
1
2,20
12b
in SrTiO
3
,
especially when Ti/Sr > 1. According to the
that of SrTiO
3
, the decrease of the CB bottom levels of
)O samples at x ) 2.0% was approximately
above elemental analysis, Sr content was slight lower than that
of Ti, which suggests that, in addition to the small amount of
Na substituted in A sites, Sr vacancies could play an
important role in keeping the charge balance. After modification
with Cu(II), the Cu 2p3/2 core-level XPS signal was evidently
observed at 932.3 eV (Figure 4c), which is consistent with the
y
(Sr1-yNa )(Ti1-xMo
x
3
0.7 eV, estimated from the change in the optical band-gap
energies (Figure S6b, Supporting Information). After surface
modification with Cu(II), the Cu(II)-SrTiO exhibited two
3
+
13
additional absorption bands in the range of 800-600 and
460-400 nm, as shown in Figure 5b. The absorption in the
7
b
21
results from our previous analysis. In our previous work on
the Cu(II)-TiO system, details of the Cu(II) state were
range of 800-600 nm is attributed to Cu(II) d-d transition.
2
The weak absorption band at 460-400 nm can be ascribed
7
a
investigated by the combination of X-ray absorption fine
structure (XAFS) measurements and XPS. It was found that
Cu(II) was grafted on the surfaces in a distorted, amorphous
CuO-like structure with a five-coordinated square pyramidal
form. By analogy with the previous study, the chemical state
y
to IFCT from the VB to Cu(II). For Cu(II)-(Sr1-yNa )-
x 3
(Ti1-xMo )O at x ) 2.0%, only the additional Cu(II) d-d
transition was observed. The IFCT band is masked by the
interband transition, due to the narrower band gap of
(Sr1-yNa )(Ti1-xMo )O samples at x ) 2.0%. This optical
y x 3
property is expected to find important application in photoca-
talysis, since more visible light energy might be effectively used
by combination of IFCT and band-gap narrowing.
and environment of Cu(II) in the Cu(II)-(Sr1-yNa )(Ti1-xMo )O
y x 3
samples might be described in the same way. Furthermore, the
amount of Cu was determined to be 0.099 wt % by the ICP
measurement, nearly equal to the starting ratios used in the
preparation.
Photocatalytic properties of the samples were evaluated by
the decomposition of IPA in air under visible light irradiation.
The wavelength of the visible light source was 400-530 nm,
y x 3
The optical absorbance properties of (Sr1-yNa )(Ti1-xMo )O
2
samples, as measured by diffuse reflectance spectroscopy at
and the light intensity was 1 mW/cm (Figure S7, Supporting
Information). A typical change in the gas concentration during
(
17) (a) Choi, J. G.; Thompson, L. T. Appl. Surf. Sci. 1996, 93, 143. (b)
Decanio, S. J.; Cataldo, M. C.; Decanio, E. C.; Storm, D. A. J. Catal.
the decomposition of IPA over Cu(II)-(Sr1-yNa
at x ) 2.0% under visible light irradiation is given in Figure
a. Prior to light irradiation, almost all of the IPA was adsorbed
onto the surfaces of the photocatalysts. With the onset of visible
light irradiation, the amount of acetone increased rapidly. After
the sharp increase, the acetone concentration started to decrease.
y x 3
)(Ti1-xMo )O
1
989, 119, 256. (c) Gr u¨ nert, W.; Stakheev, A. Y.; M o¨ rke, W.;
Feldhaus, R.; Anders, K.; Shpiro, E. S.; Minachev, K. M. J. Catal.
992, 135, 269.
18) Oku, M.; Wagatsuma, K.; Kohiki, S. Phys. Chem. Chem. Phys. 1999,
6
1
(
(
1
, 5327.
19) Hitosugi, T.; Kamisaka, H.; Yamashita, K.; Nogawa, H.; Furubayashi,
Y.; Nakao, S.; Yamada, N.; Chikamatsu, A.; Kumigashira, H.; Oshima,
M.; Hirose, Y.; Shimada, T.; Hasegawa, T. Appl. Phys. Express 2008,
2
Meanwhile, the amount of CO initially increased slowly.
Accompanying the decease of acetone, the concentration of CO
began to increase quickly. In the later period, the rate of CO
2
1
, 111203.
2
(
20) (a) Tanaka, T.; Matsunaga, K.; Ikuhara, Y.; Yamamoto, T. Phys. ReV.
B 2003, 68, 205213. (b) Mizoguchi, T.; Sato, Y.; Buban, J. P.;
Matsunaga, K.; Yamamoto, T.; Ikuhara, Y. Appl. Phys. Lett. 2005,
generation became slower. The experimental observation is
8
7, 241920.
(21) Filho, N. L. D. Mikrochim. Acta 1999, 130, 233.
1
5264 J. AM. CHEM. SOC. 9 VOL. 132, NO. 43, 2010

Cu(II)-(Sr1-yNa )(Ti1-xMo )O
y x 3
Photocatalysts
A R T I C L E S
Figure 6. (a) Representative time-dependent gas concentrations during IPA decomposition over Cu(II)-(Sr1-yNa
light irradiation. (b) Comparative studies of CO generation over Cu(II)-(Sr1-yNa )(Ti1-xMo )O
visible light irradiation with the same initial photon flux. (c) Photocatalytic activities of (Sr1-yNa )(Ti1-xMo
)(Ti1-xMo )O at x ) 2.0% and bare SrTiO
y
)(Ti1-xMo
x
)O
3
at x ) 2.0% under visible
2
y
x
3 3
samples with various Mo contents and bare SrTiO under
y
x
)O
3
at x ) 2.0% without Cu(II) co-catalyst under
visible light and UV irradiation. (d) Photocatalytic activities of Cu(II)-(Sr1-yNa
y
x
3
3
under UV light irradiation. As
for the Figure 6c,d, the UV light intensity was calibrated to keep the absorbed photon numbers the same as that under visible light irradiation.
+
reasonable, as it is well known that the photocatalytic decom-
position of IPA proceeds via formation of acetone as an
intermediate, followed by further decomposition of acetone to
additional Na source (sample V) or MoO
3
as the dopant instead
3
of NaMoO in our hydrothermal conditions (sample VI). The
+
optimum doping density of Na was less than that expected on
the basis of theoretical calculations, probably due to the creation
2
2
2 2
the final products CO and H O. After 260 h of irradiation,
of Sr2 vacancies. Our SrTiO
+
sample co-doped with Mo as
6+
the concentration of CO
µmol), which is nearly 3 times the amount of initially injected
2
was approximately 900 ppm (ca. ∼18
3
+
6+
well as Na exhibits better activity than SrTiO
3
doped with Mo
IPA (300 ppm), indicating the complete decomposition of IPA
alone (sample VI), but the activity decreased with further
increasing the Na content in the sample. One possible reason
9
+
(
CH
3
CHOHCH
3
+
/
2
O
2
f 3CO
2
+ 4H
2
O). Comparative
studies of photocatalytic activities of the photocatalysts with
various Mo contents under the same visible light source are
3
for this is that Na ions are favored for co-doping in SrTiO ,
2
4
which does not improve the photocatalytic activity. In the
present study, optimum doping densities for Mo6 and Na were
2.0% and 0.25%, respectively (see details in the Supporting
Information, Figure S8). On the basis of Figure 6b, the QE for
+
+
shown in Figure 6b. CO
negligible because it cannot absorb visible light. In contrast,
the Cu(II)-modified, non-doped SrTiO (at x ) 0) was active
under visible light, owing to IFCT from the VB of non-doped
SrTiO to the Cu(II) ions, but its visible light activity was
relatively low. More interestingly, the visible light activity of
Cu(II)-(Sr1-yNa )O samples increased with the
Mo doping up to x ) 2.0%, since the visible light absorption
2
3
generation over bare SrTiO was
3
CO
QE ) R
involved in CO
and Ra
is the absorption rate of incident photons. The details
for this calculation are described in the literature
2
generation was calculated using the following equation:
r
a
a
, where Rr
3
p
/R
p
) 6RCO2/R
p
p
is the reaction rate of photons
generation rate,
2
generation, RCO2 is the CO
2
y
)(Ti1-xMo
x
3
p
6+
7a22
and in the
6+
increased, as shown in our UV-visible spectra. When the Mo
Supporting Information (Figure S9). Some calculated results are
summarized in Table 1. Under the same irradiation conditions,
doping content was increased to more than 2.0%, the visible
light activity of the sample decreased, owing to the formation
the absorption rate of incident photons and the CO
2
generation
rate increased as Mo content increased. For the Cu(II)-
)O at x ) 2.0%, the CO generation rate
2
3
6+
of inactive SrMoO
4
as an impurity. Furthermore, we also
+
optimized the doping density of Na by using NaCl as an
(Sr1-yNa
y
)(Ti1-xMo
x
3
2
(
(
22) Ohko, Y.; Hashimoto, K.; Fujishima, A. J. Phys. Chem. A 1997, 101,
(24) (a) Fernandez, A.; Lassaletta, G.; Jimenez, V. M.; Justo, A.; Gonazalez-
Elipe, A. R.; Herrmann, J. M.; Tahiri, H.; Ait-Ichou, Y. Appl. Catal.,
B 1995, 7, 49. (b) Fujishima, A.; Rao, T. N.; Tryk, D. A. Electrochim.
Acta 2000, 45, 4683.
8
057.
23) Bi, J. H.; Wu, L.; Zhang, Y. F.; Li, Z. H.; Li, J. Q.; Fu, X. Z. Appl.
Catal., B 2009, 91, 135.
J. AM. CHEM. SOC. 9 VOL. 132, NO. 43, 2010 15265

A R T I C L E S
Qiu et al.
Table 1. Rate of Incident Photons (Ri
), Absorption Rate of Incident Photons (R ), CO Generation Rate (RCO2), and Quantum Efficiencies
p p 2
a
a
(
2
QE) for CO Generation Over Different Samples
y x 3
Cu(II)-(Sr1-yNa )(Ti1-xMo )O
b
2
b
3
x ) 0
x ) 0.25%
x ) 0.9%
x ) 2.0%
Cu(II)-TiO
Cu(II)-WO
i
16
14
16
16
15
16
15
16
14
16
15
R
R
R
p
(quanta/s)
(quanta/s)
1.30 × 10
2.30 × 10
0.035
1.30 × 10
1.30 × 10
1.00 × 10
0.093
1.30 × 10
1.02 × 10
0.148
1.30 × 10
3.30 × 10
0.029
1.30 × 10
3.91 × 10
0.680
a
14
p
6.90 × 10
CO2 (µmol/h)
QE (%)
0.043
6.1
15.2
9.2
14.5
8.8
17.5
a
2
b
Visible light source: wavelength, 400-530 nm; intensity, 1 mW/cm . Data from ref 7a.
7
a
Figure 7. Proposed photocatalysis processes. λ represents the reorganization energy.
reached a maximum at 0.148 µmol/h, and QE was calculated
2
or nitrogen-doped TiO , and its QE is as high as that reported
to be 14.5%. It is worth noting that the activity is much higher
for very efficient visible light photocatalysts like WO . Fur-
3
than that of Cu(II)-TiO
2
and comparable to that of
thermore, it is also noteworthy that our QE under visible light
is the same as that under UV light. This QE trend has never
7a
Cu(II)-WO
3
,
suggesting that Cu(II)-(Sr1-yNa )O
y
)(Ti1-xMo
x
3
samples are promising visible-light-driven photocatalysts for
practical uses.
In order to specify the role of Cu(II) ions, we evaluated the
been seen in doped semiconductor photocatalysts such as
5
nitrogen-doped TiO
2
. Nitrogen levels in N-doped TiO
2
are
isolated above its VB, and its QE under visible light is much
decomposition of IPA over (Sr1-yNa
Cu(II) ions under UV or visible light irradiation, and the results
are shown in Figure 6c. The bare (Sr1-yNa )O at x
2% without Cu(II) ions was inactive under both UV and
y
)(Ti1-xMo
x
)O
3
without
5
lower than that under UV light.
From what has been discussed above, we can draw several
important conclusions: (1) The doped sample without Cu(II)
modification is inactive for CO generation under UV and visible
2
light irradiation, although lowing the conduction band level can
enhance the visible light absorption. (2) After Cu(II) modifica-
tion, the photocatalytic activity of samples was greatly improved.
y
)(Ti1-xMo
x
3
)
visible light irradiation, even though its absorption edge
extended to the visible light region. These results reveal that
the modification of the doped SrTiO
indispensable to achieve its photocatalytic activity. The UV light
activities over bare SrTiO and Cu(II)-(Sr1-yNa )O
3
with Cu(II) ions is
3
y
)(Ti1-xMo
x
3
(
3) The QE of Cu(II)-(Sr1-yNa
light irradiation is similar with that under UV light irradiation.
4) The activity of Cu(II)-(Sr1-yNa )O under
either visible light or UV light irradiation is much better than
that of SrTiO under UV light irradiation with the same
absorbed photon number. (5) The activity of Cu(II)-(Sr1-yNa )-
)O can be controlled by tuning the dopants content.
y x 3
)(Ti1-xMo )O under visible
at x ) 2.0% are shown in Figure 6d in order to discuss its
electric band structure. The UV light intensity was calibrated
to absorb the same photon numbers as that under visible light
irradiation (Figure S7, Supporting Information). Even though
(
y
)(Ti1-xMo
x
3
3
the metal ions were doped into Cu(II)-(Sr1-yNa
its UV light activity was better than that of bare SrTiO
experiments were performed under the same absorbed pho-
ton numbers, indicating that the QE of Cu(II)-(Sr1-yNa )-
)O at x ) 2.0% under UV light was about 10%, which
was higher than that of bare SrTiO under UV light. It is also
noteworthy that the photocatalytic activity of Cu(II)-(Sr1-yNa )-
)O under UV light irradiation (shown in Figure 6d)
y x 3
)(Ti1-xMo )O ,
y
3
. These
(
Ti1-xMo
x
3
Based on the above results, the photocatalytic mechanisms are
speculated in Figure 7. Although the absorption edge was
y
(
Ti1-xMo
x
3
extended to the visible light region, the rate of CO
was negligible over (Sr1-yNa )O without Cu
modification. This can be understood by considering the low
2
generation
3
2
+
y
)(Ti1-xMo
x
3
y
(
Ti1-xMo
x
3
6+
was very close to that under visible light irradiation (shown in
Figure 6b). To the best of our knowledge, the QE of our
materials is the best among the reported perovskite materials
potential of photoinduced electrons in the CB. Mo as well as
+
Na doping shifted the bottom of the CB more positive than
the potential for the single-electron reduction of oxygen (E )
1
5266 J. AM. CHEM. SOC. 9 VOL. 132, NO. 43, 2010

Cu(II)-(Sr1-yNa )(Ti1-xMo )O
y x 3
Photocatalysts
A R T I C L E S
2
5
. This is because doping with Mo6+ as
3
-
0.064 V vs SHE, pH ) 0), resulting in inefficient consump-
that of Cu(II)-SrTiO
well as Na increases the light absorption by lowering the
bottom of the conduction band. Therefore, high photocatalytic
+
tion of photoinduced electrons by oxygen. Once the sample was
grafted with Cu(II), the potential of Cu /Cu was about 0.16
2
+
+
V (vs SHE, pH ) 0); thus, it can be expected to act as a co-
y x 3
decomposition of IPA over Cu(II)-(Sr1-yNa )(Ti1-xMo )O
7
catalyst to reduce oxygen molecules efficiently. Therefore, for
samples is achieved by utilization of the high oxidation power
of holes in the VB in combination with the efficient oxygen
reduction via Cu(II). On the basis of this study, possible
candidates for visible-light-active materials are largely extended
to other materials. That is, some photocatalytically inactive
materials, such as low-conduction-band materials or only UV-
light-active materials, could become good candidates for efficient
visible-light-active photocatalysts on the basis of our new
findings. In other words, our work indicates a strategic way to
develop new visible-light-active photocatalysts.
Cu(II)-SrTiO (at x ) 0), under visible light irradiation, the
3
holes in the VB resulting from IFCT decomposed IPA, while
photoinduced electrons were consumed via an efficient oxygen
reduction process with Cu ions, which accounts for the enhanced
activity compared to that of bare SrTiO
3
. For Cu(II)-SrTiO
3
doped with Mo6 as well as Na , by virtue of CB control, in
addition to the IFCT process, the electrons in the VB can also
be promoted to the CB under visible light irradiation. The
photoindued electrons in the CB were further transferred to
Cu(II) clusters and drove the oxygen reduction, which retarded
the photoindued electron-hole recombination effectively. Con-
sequently, high visible light photocatalytic activity is achieved
+
+
Acknowledgment. This work was performed under the man-
agement of the Project To Create Photocatalyst Industry for
Recycling-Oriented Society, supported by the New Energy and
Industrial Technology Development Organization (NEDO) in Japan.
over Cu(II)-(Sr1-yNa
y x 3
)(Ti1-xMo )O samples as a result of
lowering the bottom of the CB and surface Cu(II) modification.
Supporting Information Available: Table S1, composition of
the sample derived from ICP and atomic absorption spectro-
photometry; Figures S1 and S2, XRD patterns of samples;
4
. Conclusion
In summary, we have developed a novel visible-light-driven
Figure S3, TEM images of SrTiO
3
3
and Cu(II)-SrTiO ; Figures
photocatalyst based on CB control and surface modification with
Cu(II). Our studies show that Mo as well as Na doping in
the lattice of SrTiO can lower the bottom of the CB and
6
+
+
S4 and S5, full-scale XPS spectra of samples and comparative
studies of Mo 3d core-level and Ti 2p core-level spectra of
3
6
+
effectively extend the absorption edge to the visible light region.
When the bottom of the CB was lowered, Cu(II) modification
was indispensable to achieve photocatalytic activity because the
potential of the photoinduced electrons in the CB is too low to
initiate the single-electron reduction of oxygen. The role of
Cu(II) modification has two main positive effects: (i) IFCT for
visible light absorption and (ii) co-catalyst for efficient oxygen
reduction to consume the photoinduced electrons. Under the
same visible light irradiation condition, the activity of
y x 3
(Sr1-yNa )(Ti1-xMo )O samples with roughly fixed Mo
+
content and different Na contents; Figure S6, (a) comparative
study of the UV-visible spectra of samples with roughly fixed
Mo6 content and different Na contents and (b) relationships
+
+
between (ahν)1 and photon energy of (Sr1-yNa
/2
)(Ti1-xMo
)O
3
y
x
samples; Table S2, specific surface area values of samples;
Figure S7, information on the light source for photocatalytic
experiments; Figure S8, comparative studies of CO generation
2
over different samples under visible light irradiation; Figure S9,
details of the calculations of quantum efficiency. This material
is available free of charge via the Internet at http://pubs.acs.org.
y x 3
Cu(II)-(Sr1-yNa )(Ti1-xMo )O at x ) 2.0% is much higher than
(25) Bard, A. J., Parsons, R., Jordan, J., Eds. Standard Potentials in Aqueous
Solution; Marcel Dekker: New York, 1985.
JA105846N
J. AM. CHEM. SOC. 9 VOL. 132, NO. 43, 2010 15267