Lanthanum-indium oxysulfide as a visible light driven photocatalyst for water splitting

Article abstract of DOI:10.1246/cl.2007.854

La-In-based oxysulfide is demonstrated to act as a photocatalyst for the reduction of H⁺ to H₂ and the oxidation of H₂O to O₂ in the presence of sacrificial reagents under visible light (420 ≤ λ ≤ 480 nm). Loading with IrO₂ is effective for promoting O₂ evolution, while Pt is effective as a cocatalyst for H₂ evolution. Copyright

Full text of DOI:10.1246/cl.2007.854

854
Chemistry Letters Vol.36, No.7 (2007)
Lanthanum–Indium Oxysulfide as a Visible Light Driven Photocatalyst for Water Splitting
Kiyonori Ogisu,1 Akio Ishikawa,1 Kentaro Teramura,1 Kenji Toda,2 Michikazu Hara,3 and Kazunari Domenꢀ1
1Depertment of Chemical System Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656
2Graduate School of Science and Technology, Niigata University, 8050 Ikarashi Ninocho, Niigata 950-2181
3Materials and Structures Laboratory, Tokyo Institute of Technology, 4259 Nagatuta, Midori-ku, Yokohama 226-8503
(Received April 9, 2007; CL-070380; E-mail: domen@chemsys.t.u-tokyo.ac.jp)
La–In-based oxysulfide is demonstrated to act as a photoca-
talyst for the reduction of Hþ to H2 and the oxidation of H2O
to O2 in the presence of sacrificial reagents under visible light
(420 ꢁ ꢀ ꢁ 480 nm). Loading with IrO2 is effective for promot-
ing O2 evolution, while Pt is effective as a cocatalyst for H2 evo-
lution.
3
2.5
2
1.5
Energy/eV
(a)
(b)
(c)
Certain sulfides, such as CdS1,2 and (AgIn)xZn2ð1ꢂxÞS2,3
exhibit good absorption in the visible-light region4–6 and display
activity for the photoreduction of Hþ to H2 in the presence of an
electron donor such as S2ꢂ and SO32ꢂ. However, sulfides are
generally unstable in water oxidation to form O2 because the
S2ꢂ anions are sensitive to oxidation by photogenerated holes.7,8
Recently, Ln2Ti2S2O5 (Ln = Pr–Er) oxysulfides have been
demonstrated to act as stable photocatalysts for both Hþ reduc-
tion and the oxidation of H2O to O2.9,10 In the present study,
La–In oxysulfide with d10 electric configuration is investigated
as another potential photocatalytic material for water splitting
under visible light.
Following the example of Kabbour et al.,11 the preparation
of LaInS2O was attempted in this study by heating a mixture
of La2S3, La2O3, and In2S3 at a stoichiometric molar ratio
(La2S3:La2O3:In2S3 = 1:2:3) in a sealed quartz tube under
vacuum at temperatures of 873–1273 K for 6–24 h.11 In this
study, we henceforth report the sample obtained at 1073 K for
12 h, which showed the highest photocatalytic activities among
all prepared samples. The sintered samples were then ground
and heated at 573 K for 1 h in air to remove absorbed sulfur9
to yield a yellow powder. The crystal structure of the resulting
material was examined by powder X-ray diffraction (XRD)
using a Rigaku Geigerflex RAD-B instrument with Cu Kꢁ
radiation. Ultraviolet–visible diffuse reflectance (UV–vis DR)
spectra were obtained using a Jasco V-560 spectrometer. Photo-
reduction of Hþ to H2 and photooxidation of H2O to O2 in the
presence of sacrificial reagents were carried out in a Pyrex reac-
tion vessel connected to a gas-circulation system. H2 evolution
was examined using an aqueous solution (200 mL) containing
0.1 g of the oxysulfide loaded with Pt metals by in situ photode-
position, and 0.01 M Na2S and 0.01 M Na2SO3 as sacrificial
electron donors. O2 evolution was examined using an aqueous
0.01 M AgNO3 solution containing 0.1 g of the oxysulfide load-
ed with IrO2 by the impregnation method with Na2IrCl6 solution
and then the treatment in air at 573 K for 1 h. La2O3 was used as
a buffer material to maintain the pH of the solution at 7–8.
The reaction solution was evacuated several times to remove
air and then irradiated under visible light using a 300-W Xe lamp
with a cutoff filter (ꢀ > 420 nm) to eliminate UV light and water
filter to remove infrared light.
300
500
700
Wavelength/nm
Figure 1. UV–vis diffuse reflectance spectra for (a) LaInO3,
(b) La–In oxysulfide, and (c) Sm2Ti2S2O5.9
ples matched those for LaInS2O reported by Kabbour et al.11
Yet, the structural detail of the LaInS2O phase is unknown.
Given the presence of In2O3 in the XRD pattern, the oxysulfide
obtained by the present preparation procedure is considered
to be a mixture of In-poor La5In3S9O3 and minor impurity phas-
es, most likely La1:33In1:33S412 and LaIn2S4.13 As a single-phase
La5In3S9O3 powder sample could not be obtained from
starting materials with a nominal composition of La:In = 5:3,
La5In3S9O3 may exist in the present samples as a metastable
phase in an overall In-rich (La:In = 1:1) system. Therefore,
the prepared samples containing those phases henceforth are
denoted as La–In oxysulfide.
Figure 1 shows the UV–vis DR spectra for the present La–In
oxysulfide and LaInO3. For comparison, UV–vis DR spectrum
of Sm2Ti2S2O5 is also shown. Plane–wave–based density
function theory (DFT) calculations have suggested that the
valence band (EVB) of Sm2Ti2S2O5 is made up of the O2p and
S3p hybridized orbitals and the conduction band (ECB) consists
of Ti3d; as a result, Sm2Ti2S2O5 has a smaller band-gap energy
(ꢃ2:1 eV) compared with that of Sm2Ti2O7 (ꢃ3:5 eV).9
Similarly, the EVB of the La–In oxysulfide appears to consist
of the O2p and S3p orbitals. On the other hand, the ECB of
La–In oxysulfide and LaInO3 in both cases would be composed
of hybridized In5s5p orbitals.14 Accordingly, the La–In oxysul-
fide has a smaller band-gap energy (ꢃ2:6 eV) than LaInO3
(ꢃ4:1 eV).
Figure 2 shows the time course of repeated H2 evolution
over La–In oxysulfide loaded with 1.0 wt % Pt under visible-
light irradiation (ꢀ > 420 nm) in the presence of Na2S–Na2SO3.
The reaction system was evacuated every 5 h. In the early stage
of the reaction (2 h), H2PtCl6 was reduced to Pt as an H2 evolu-
tion promoter on the catalyst surface. The rate of H2 evolution,
however, remained essentially stable after this induction period.
The XRD pattern of the catalyst after the H2 evolution reaction
was the same as before the reaction. The Pt-loaded La–In
oxysulfide therefore functions as a stable photocatalyst for the
reduction of Hþ to H2 under visible-light irradiation.
Most of the crystalline peaks produced by the present sam-
Copyright ꢀ 2007 The Chemical Society of Japan

Chemistry Letters Vol.36, No.7 (2007)
855
evacuation
8
6
4
2
0
50
40
30
20
10
400
500
600
Wavelength/nm
0
0
5
10
15
20
25
Reaction Time/h
Figure 4. Dependence of rate of H2 and O2 evolution on cutoff
wavelength of incident light, and UV–vis DR spectrum of La–In
oxysulfide. Circles denote H2 evolution (1 wt % Pt-loaded catal.,
0.1 g; 0.01 M Na2S–0.01 M Na2SO3 solution), and triangles
denote O2 evolution (2 wt % IrO2-loaded catal., 0.1 g; 0.01 M
AgNO3 solution, La2O3, 0.2 g).
Figure 2. Time course of repeated H2 evolution over La–In
oxysulfide (Pt-loaded catalyst, 0.1 g; 0.01 M Na2S–0.01 M
Na2SO3 solution, 200 mL).
15
(c)
of La–In oxysulfide containing sulfide phases influenced the QE
for O2 evolution.
10
La–In oxysulfide was demonstrated to catalyze the reduction
of Hþ to H2 and the oxidation of H2O to O2 under visible-light
irradiation in the presence of a sacrificial electron donor (Na2S–
Na2SO3) or acceptor (Agþ), respectively. This oxysulfide, with
a band gap of 2.6 eV, was thus confirmed to be a photocatalyst
with reduction and oxidation abilities, having conduction and
valence bands at suitable potentials for the reduction of Hþ
and oxidation of H2O. O2 evolution was effectively enhanced
by loading with IrO2, while Pt proved to be suitable as a cocata-
lyst for H2 evolution.
5
0
(b)
(a)
0
1
2
3
4
5
Reaction Time/h
Figure 3. Time course of O2 evolution over (a) CdS, (b) La–In
oxysulfide, and (c) 2 wt % IrO2/La–In oxysulfide under visible-
light irradiation (ꢀ > 420 nm) (catal., 0.1 g; 0.01 M AgNO3
solution, 200 mL; La2O3, 0.2 g).
References
1
2
3
4
N. Buhler, K. Meier, J.-F. Reber, J. Phys. Chem. 1984, 88,
¨
3261.
M. Matsumura, S. Furukawa, Y. Saho, H. Tsubomura, J. Phys.
Chem. 1985, 89, 1327.
I. Tsuji, H. Kato, H. Kobayashi, A. Kudo, J. Am. Chem. Soc.
2004, 126, 13406.
Figure 3 shows the time courses of O2 evolution over bare
and IrO2-loaded La–In oxysulfide, and over CdS for comparison.
O2 evolution was not observed over CdS owing to the photode-
composition of CdS by photogenerated holes. In contrast, over
the bare La–In oxysulfide, O2 evolution was observed immedi-
ately with the onset of irradiation. Loading with 2 wt % IrO2
increased the evolution rate by approximately three-fold,
indicating that IrO2 is an effective O2 evolution promoter for
the La–In oxysulfide. After an initial period (1 h), the rate
of O2 evolution decreased over time owing to the deposition
of metallic silver on the surface of the catalyst.
K. Domen, J. N. Kondo, M. Hara, T. Takata, Bull. Chem. Soc.
Jpn. 2000, 73, 1307.
5
6
A. Kudo, H. Kato, I. Tsuji, Chem. Lett. 2004, 33, 1534.
J. Sato, N. Saito, H. Nishiyama, Y. Inoue, J. Phys. Chem. B
2003, 107, 7965.
7
8
R. Williams, J. Chem. Phys. 1960, 32, 1505.
H. Gerischer, J. Electroanal. Chem. Interfacial Electrochem.
1975, 58, 263.
Figure 4 shows the relationship between the H2 and O2
evolution rates and the cutoff wavelength of incident light.
The steady rate of H2 evolution and the initial rate of O2 evolu-
tion decreased with increasing cutoff wavelength, confirming
the position of the absorption edge of the La–In oxysulfide and
the procession of these photocatalytic reactions via band-gap
transitions. No O2 evolution was observed under visible-light
irradiation with longer wavelength than 500 nm because of our
9
A. Ishikawa, T. Takata, J. N. Kondo, M. Hara, H. Kobayashi,
K. Domen, J. Am. Chem. Soc. 2002, 124, 13547.
10 A. Ishikawa, T. Takata, T. Matsumura, J. N. Kondo, M. Hara,
H. Kobayashi, K. Domen, J. Phys. Chem. B 2004, 108, 2637.
11 H. Kabbour, L. Cario, Y. Moelo, A. Meerschaut, J. Solid State
¨
Chem. 2004, 177, 1053.
12 A. Likforman, M. Cuittard, Acta Crystallogr., Sect. C 1993, 49,
1270.
13 S. A. Amirov, N. A. Shnulin, G. G. Guseinov, S. Kh,
Mamedov, Kristallografiya 1984, 29, 787.
detection limit to be ꢃ0:1 mmol hꢂ1
.
The apparent quantum efficiencies (QE)15 of H2 and O2 for
La–In oxysulfide (Fig. 2 and Fig. 3b) were estimated to be ꢃ0:2
and ꢃ0:1%, respectively, indicating that La–In oxysulfide
exhibited superior QE for H2 evolution and lower QE for O2
evolution compared to those of Sm2Ti2S2O5 [QE: 0.1% (H2),
0.2% (O2)].9 Although many factors should affect photocatalytic
activity, it seems that ECB of In5s5p orbitals with large disper-
sion14 mainly led to the higher QE of H2 evolution and the purity
14 J. Sato, H. Kobayashi, Y. Inoue, J. Phys. Chem. B 2003, 107,
7970.
15 Quantum efficiency values were calculated with using the
coefficients (H2: 2, O2: 4), steady H2 or initial O2 evolution
rate, the rate of absorption of incident photons [Sm2Ti2S2O5:
8:6 ꢄ 1021 photons hꢂ1 at 440 ꢁ ꢀ ꢁ 650 nm, La–In oxysul-
fide: 7:4 ꢄ 1021 photons hꢂ1 at 420 ꢁ ꢀ ꢁ 600 nm].
Published on the web (Advance View) June 2, 2007; doi:10.1246/cl.2007.854