5 January 2001
Chemical Physics Letters 333 (2001) 57±62
www.elsevier.nl/locate/cplett
Photocatalytic and photophysical properties of a novel series of solid photocatalysts, Bi2MNbO7 (M Al3; Ga3 and In3 ) Zhigang Zou a,*, Jinhua Ye b, Hironori Arakawa a a
National Institute of Materials and Chemical Research, 1-1 Higashi, Tsukuba, Ibaraki 305, Japan b National Research Institute for Metals, 1-2-1 Sengen, Tsukuba, Ibaraki 305, Japan Received 11 August 2000; in ®nal form 13 November 2000
Abstract Bi2 MNbO7
M Al3 ; Ga3 ; In3 solid photocatalysts were prepared by solid-state reaction and characterized by powder X-ray diraction and Rietveld structure re®nement. These photocatalysts crystallize in the same pyrochlore structure, but the lattice parameters decrease with decrease of M3
M3 Al3 ; Ga3 ; In3 ionic radii. The band gaps of Bi2 MNbO7
M Al3 ; Ga3 ; In3 were estimated to be about 2.9, 2.75 and 2.7 eV, respectively. The difference in the band gaps of the photocatalysts is due to their dierent conduction band levels, resulting in the dierence in photocatalytic activity of the photocatalysts. Ó 2001 Elsevier Science B.V. All rights reserved.
1. Introduction The photocatalytic splitting of water using oxide semiconductor materials attracts an increasing interest, since it can be a promising chemical route for energy renewal and storage [1±5]. However, the number of photocatalyst materials known up to now is yet limited, and their activity is still low. Therefore, there is an urgent need to develop new types of photocatalyst materials with higher activity [6]. Photocatalysis on semiconductors involves the direct absorption of a photon by the energy band gap of the materials, thereby generating electron±hole pairs. Such studies are often related to photophysical properties of photocatalysts.
*
Corresponding author. Fax: +81-298-61-4750. E-mail address:
[email protected] (Z. Zou).
Very recently, we have found that Bi2 InNbO7 is of paramagnetic and semiconducing behavior. Furthermore, we found that Bi2 InNbO7 acts as a photocatalyst under UV irradiation [7]. The Bi2 InNbO7 compound has pyrochlore crystal structure of consisting of the network of MO6
M In3 ; Nb5 as shown in Fig. 1. The compound has a band gap of about 2.7 eV and seems to have potential for improvement of its activity by modi®cation of the structure [7]. We suggest that substitution of In3 by Al3 and Ga3 in Bi2 InNbO7 might cause a slight modi®cation of crystal structure, resulting in an increase in hole(carrier) concentration and providing a change in photocatalytic and photophysical properties. It is known that a slight modi®cation of structure has a dramatic eect on the concentration and mobility of charge, which directly aect the photocatalytic and photophysical properties of semiconductors [8].
0009-2614/01/$ - see front matter Ó 2001 Elsevier Science B.V. All rights reserved. PII: S 0 0 0 9 - 2 6 1 4 ( 0 0 ) 0 1 3 4 8 - 8
58
Z. Zou et al. / Chemical Physics Letters 333 (2001) 57±62
Fig. 1. The schematic structural diagram of Bi2 MNbO7
M3 Al3 ; Ga3 ; In3 . Three-dimensional network of MO6 stacked along [0 0 1] and separated by a unit cell translation.
Here we report the photocatalytic and photophysical characterization of the Bi2 MNbO7
M Al3 ; Ga3 ; In3 solid photocatalysts. A comparison of the photocatalytic properties of Bi2 MNbO7
M Al3 ; Ga3 ; In3 with the TiO2 photocatalyst is presented. 2. Experimental The polycrystalline samples of the Bi2 MNbO7
M Al3 ; Ga3 ; In3 photocatalysts were prepared by solid state reaction method using high purity grade chemicals of Bi2
CO3 3 , In2 O3 , Nb2 O5 , a-Al2 O3 and Ga2 O3 . The stoichiometric amounts of precursors were mixed and pressed into small columns. The columns were reacted in an aluminum crucible in air for two days at 1100°C. The same procedure was performed three times for the samples. The chemical composition of the samples before and after photocatalytic reactions was determined by scanning electron microscope-X-ray energy dispersion spectrum (SEM-EDS) with an acceler-
ating voltage of 25 kV. The crystal structure of the Bi2 MNbO7
M Al3 ; Ga3 ; In3 photocatalysts was determined by the powder X-ray diffraction method using CuKa radiation
k UV-vis diuse re¯ectance spectrum of 1:54178 A. Bi2 MNbO7
M Al3 ; Ga3 ; In3 was measured by using an UV-vis spectrometer (MPS2000). The surface area was determined by BET measurement (Micromeritics, Shimadzu, FlowPrep 060). The photocatalytic reaction was examined using a gas closed circulation system and an inner-irradiation type quartz cell with 400 W high-pressure Hg lamp. The gases evolved were determined with TCD gas chromatograph, which was connected with a circulating line. It is known that addition of noble metals or metal oxides to the surface of a semiconductor changes its surface properties and hence, its photocatalytic behavior [9]. Pt has been shown to be eective for the TiO2 photocatalyst [9], hence, we loaded Pt onto the catalyst surface in our experiments to obtain higher activity. The photocatalytic reaction was performed in an aqueous CH3 OH=H2 O solution (1.0 g powder
Z. Zou et al. / Chemical Physics Letters 333 (2001) 57±62
59
catalyst, 50 ml CH3 OH, 350 ml H2 O and 0.1 wt% Pt (Pt-loading instead of a H2 PtCl4 )). O2 evolution reaction was performed in an aqueous cerium sulfate solution (1.0 g powder catalyst, 1.0 m mol Ce
SO4 2 , 400 ml H2 O) since the solution is more stable than an aqueous silver nitrate under UV irradiation [10].
3. Results and discussion The chemical composition of the Bi2 MNbO7
M Al3 ; Ga3 ; In3 photocatalysts before and after photocatalytic reaction was determined using characteristic X-rays of MLa
M
Al3 ; Ga3 ; In3 , BiMa , and NbLa . The composition content was determined using the ZAF quanti®cation method. Oxygen content was calculated from the EDS results [8]. The SEM-EDS analysis showed that the photocatalysts have a homogenous atomic distribution with no other additional elements. The chemical composition of these samples was con®rmed to be the same before and after reactions. The crystal structure of the photocatalysts before and after reaction was investigated using X-ray powder diraction. The result is shown in Fig. 2. The data were collected at 295 K with a step scan procedure in the range of 2h 5±100°. The step interval was 0:024° and scan speed, 1° minÿ1 . The powder X-ray diraction analysis showed that all samples are single phases. Full-pro®le structure re®nement of the collected powder diraction data was performed using the Rietveld program REITAN [11]. The outcome of the ®nal re®nement indicated that the Bi2 MNbO7
M Al3 ; 3 3 Ga ; In photocatalysts have the pyrochlore type crystal structure, cubic system with space group Fd3m [7]. The Bi2 MNbO7
M Al3 ; Ga3 ; In3 photocatalysts crystallize in the same structure, but 2h angles of each re¯ection changed with In3 being substituted by Al3 and Ga3 , indicating a decrease in lattice parameter of the photocatalysts with decrease of the M ionic radii, < Ga3
0:62 A < In3
0:92 A.The Al3
0:57 A result of the ®nal re®nement is shown in Table 1. All the diraction peaks for the Bi2 MNbO7
Fig. 2. X-ray powder diraction patterns of samples before and after reactions.
M Al3 ; Ga3 ; In3 photocatalysts could be indexed based on the lattice parameters and the space group mentioned above. From these experimental results, we con®rmed that these photocatalysts have not changed in both the crystal structure and the chemical composition after reaction. Fig. 3 shows the H2 and CO evolutions from Pt=CH3 OH=H2 O solution under UV irradiation with the Bi2 MNbO7
M3 Al3 ; Ga3 ; In3 photocatalysts. The formation rate of H2 increased rapidly with decrease of the M3 ionic radii, < Ga3
0:62 A < In3
0:92 A.This Al3
0:57 A means that the activity of these photocatalysts increases with decrease of M3 ionic radii. The formation rate of H2 evolutions was estimated to be 0.71, 0.3, and 0.18 m mol gÿ1 hÿ1 in the ®rst 10 h for Al3 ; Ga3 and In3 , respectively. It is notable that the formation rate of H2 evolution with Bi2 AlNbO7 is much larger than that of TiO2 photocatalyst (TiO2 ±P25). This means that the activity of Bi2 AlNbO7 to decompose Pt=CH3 OH=H2 O solution is higher than that of TiO2 photocatalyst.
Z. Zou et al. / Chemical Physics Letters 333 (2001) 57±62
2.9 2.75 2.7 3.2 O2
25 10 7 17
CO
32 7 5 15
H2
710 300 180 550 0.51 0.52 0.51 53.80 Pyrochlore Pyrochlore Pyrochlore Anatase rutile Bi2 AlNbO7 Bi2 GaNbO7 Bi2 InNbO7 TiO2
P25
10.7171(2) 10.7342(2) 10.7793(2)
Ce
SO4 2 =H2 O Ch3 OH=H2 O
Lattice parameter a (A) Type of structure Photocatalyst
Table 1 Rates of gas evolution and physical properties of the photocatalysts
Surface area (m2 gÿ1 )
Rate of gas evolutions
l mol hÿ1
Band gap (eV)
60
Fig. 3. Photocatalytic H2 and CO evolutions on Bi2 MNbO7
M3 Al3 ; Ga3 ; In3 and TiO2 from Pt=CH3 OH=H2 O solution.
The CO evolution was observed as the oxidation product in this reaction from Pt= CH3 OH=H2 O solution (see Fig. 3). The CO evolution increases with illumination time as does H2 evolution. However, the rate of CO evolution is much lower than that of H2 evolution. The formation rate of CO increased with decrease of M3
M3 Al3 ; Ga3 ; In3 , showing the same tendency as observed in H2 evolutions. The rate of CO evolution in Bi2 AlNbO7 is also much larger than that in TiO2 photocatalyst. O2 evolution reaction was performed in an aqueous cerium sulfate solution, wherein the following stoichiometric reaction takes place; 4Ce4 2H2 O ! 4Ce3 O2 4H . The result is shown in Table 1. This means that the photocatalysts have potential for O2 evolution from aqueous solution and the potential activity for O2 evolution increases with decrease of M3 ionic radii. The rate of O2 evolution in Bi2 AlNbO7 is again larger than that in the TiO2 photocatalyst, showing the same result as H2 evolution from CH3 OH=H2 O solution. Fig. 4 shows the result of diuse re¯ection spectra of the Bi2 MNbO7
M Al3 ; Ga3 ; In3 photocatalysts. The onset of diuse re¯ection
Z. Zou et al. / Chemical Physics Letters 333 (2001) 57±62
Fig. 4. Diuse re¯ectance
M3 Al3 ; Ga3 ; In3 .
spectrums
of
Bi2 MNbO7
spectra of these photocatalysts showed an obvious shift to lower wavelength with decrease of M3 ionic radii. The band gaps of the Bi2 MNbO7
M Al3 ; Ga3 ; In3 photocatalysts were estimated to be about 2.9, 2.75 and 2.7 eV from onset of diuse re¯ection spectra, respectively. This means that the band gaps of these photocatalysts
61
increase with decrease of M3 ionic radii. Scaife [12] showed that the band structure of oxides is generally de®ned by d-level and O 2p-level. However, when the compound contains octahedral MO6 the valence band energy should be assumed by the O 2p-levels of oxygen in MO6 and the conduction band assumed by d-levels of metal in MO6 . The band structures of Bi2 MNbO7
M Al3 ; Ga3 ; In3 are shown in Fig. 5. The valence band potentials of the photocatalysts should be the same because Bi2 MNbO7
M Al3 ; Ga3 ; In3 have same crystal structure. The structure of the Bi2 MNbO7
M Al3 ; Ga3 ; In3 photocatalysts can be described as consisting of the three-dimensional network of MO6
M Nb5 ; Al3 ; Ga3 ; In3 , stacked along [0 0 1] as shown in Fig. 1 [7]. The octahedral MO6
M Al3 ; Ga3 ; In3 and Nb5 are connected to form chains and the Bi ions are located in the three-dimensional network of MO6 [7]. The conduction band potentials of the photocatalysts might be dierent because of the dierence in octahedrons in Bi2 MNbO7
M Al3 ; Ga3 ; In3 . The conduction bands of Bi2 MNbO7
M Al3 ; Ga3 ; In3 should be slightly lowered as we go from Al3 to Ga3 to In3 , since the band gaps
Fig. 5. Suggested energy levels of Bi2 MNbO7
M3 Al3 ; Ga3 ; In3 and TiO2 .
62
Z. Zou et al. / Chemical Physics Letters 333 (2001) 57±62
of Bi2 MNbO7
M Al3 ; Ga3 ; In3 increase with decrease of M3 ionic radii. This dierence in the conduction band levels and the corresponding dierence in the band gaps leads to dierent photocatalytic behavior, as can be seen from Table 1. Also, it is evident from Table 1 that the dierent materials have very similar surface areas; hence the dierences in photocatalytic activity cannot be attributed to variations in surface area. It is interesting to note that Bi2 MNbO7
M Al3 ; Ga3 ; In3 show photoabsorption in the visible light region
k > 420 nm, but the photoabsorption is weak. This means that the photocatalysts have ability to respond wavelength of visible light region. However, these photocatalysts do not work under visible light irradiation
k > 420 nm in our experiment. Alig et al. have shown that direct absorption of photons by the band gap of oxides can generate electron±hole pairs in the solid [13,14]. However, the energy requirement is generally higher than the band gap of the oxides. In order to increase the activity of these catalysts, two approaches are possible. One way would be to further modify the catalyst surface to increase the range of wavelengths at which the catalyst is active. Another would be to increase the irradiation energy. BET measurement showed that the surface areas of Bi2 MNbO7
M Al3 ; Ga3 ; In3 are 0.51, 0.52 and 0.51 m2 gÿ1 , respectively. Since this is only about 1% of the surface area of the TiO2 photocatalyst (see Table 1), it demonstrates the much higher eciency of the new photocatalysts. It is evident that further increase in activity might be expected from increasing the surface area. Studies are currently underway in our laboratory to investigate the eects of surface area and surface property modi®cation. In summary, the experimental results show that the Bi2 MNbO7
M Al3 ; Ga3 ; In3 photocatalysts are sensitive to UV irradiation and that it is possible to obtain H2 and O2 from both Pt=CH3 OH=H2 O and Ce
SO4 2 =H2 O solutions.
Although the Bi2 MNbO7
M Al3 ; Ga3 ; In3 photocatalysts have a suitable band, the photocatalysts do not work under visible light irradiation to directly decompose pure water, even in Pt=CH3 OH=H2 O solution. Increasing the surface area of Bi2 MNbO7
M Al3 ; Ga3 ; In3 might increase activity of photocatalysts and increase responding wavelength range. The study of photocatalytic and photophysical properties will provide useful information on the mechanism of photocatalysts and also on making the photocatalysts with high activity. Acknowledgements The authors would like to thank Dr. K. Sayama for valuable discussions and appreciate the contributions of Dr. K. Kawaguchi and K. Hara. References [1] K. Honda, A. Fujishima, Nature 238 (1972) 37. [2] T. Kawai, T. Sakata, Nature 286 (1980) 474. [3] B.S. Georey, E.M. Thomas, J. Phys. Chem. B 101 (1997) 2508. [4] K. Yeong, S. Samer, J.H. Munir, E.M. Thomas, J. Am. Chem. Soc. 113 (1991) 9561. [5] K. Sayama, K. Yase, H. Arakawa, K. Asakura, K. Tanaka, K. Domen, T. Onishi, J. Photochem. Photobiol. A 114 (1998) 125. [6] T. Takata, A. Tanaka, M. Hara, J. Kodo, K. Domen, Catal. Today 44 (1998) 17. [7] Z. Zou, J. Ye, H. Arakawa, Catalysis Lett. 68 (2000) 235. [8] Z. Zou, J. Ye, K. Oka, Y. Nishihara, Phys. Rev. Lett. 80 (1998) 1074. [9] H.G. Kim, D.W. Hwang, J. Kim, Y.G. Kim, J. Lee, Chem. Commun. (1999) 1077. [10] E.A. Meulenkamp, A.R. Wrr, Electrochimica Acta. 41 (1996) 109. [11] F. Izumi, J. Crystallogr. Assoc. Jpn. 27 (1985) 23. [12] D.E. Scaife, Solar Energy 25 (1980) 41. [13] R.C. Alig, S.W. Bloom, Phys. Rev. B 22 (1980) 5565. [14] A.L. Linsebigler, G. Lu, J.T. Yates Jr., Chem. Rev. 95 (1995) 735.