J. Mater. Sci. Technol., Vol.24 No.5, 2008

675

• Invited Review Doped-TiO2 Photocatalysts and Synthesis Methods to Prepare TiO2 Films Ying CUI1,2) , Hao DU1) and Lishi WEN1)† 1) Division of Surface Engineering of Materials, Institute of Metal Research, Chinese Academy of Sciences, Shenyang 110016, China 2) Graduate School of Chinese Academy of Sciences, Beijing 100039, China [Manuscript received March 3, 2008]

TiO2 is a promising photocatalyst. However, the low photocatalytic efficiency calls for the modification of TiO2 . Metal- and nonmetal-doping of TiO2 have been proved to be effective ways to enhance photocatalytic properties. This review provides a deep insight into the understanding of the metal- and nonmetal-doped TiO2 photocatalysts. This article begins with the introduction of the crystal structures of TiO2 and applications of TiO2 materials. We then reviewed the doped-TiO2 system in two categories: (1) metal-doped TiO2 photocatalysts system, and (2) nonmetal-doped TiO2 photocatalysts system. Both experimental results and theoretical analyses are elaborated in this section. In the following part, for the advantages of TiO2 thin films over particles, various preparation methods to obtain TiO2 thin films are briefly discussed. Finally, this review ends with a concise conclusion and outlook of new trends in the development of TiO2 -based photocatalysts. KEY WORDS: TiO2 films; Photocatalysis; Metal-doping; Nonmetal-doping; DFT

1. Introduction of TiO2 1.1 Crystal structures and physical properties of TiO2 TiO2 is a typical n-type semiconductor. It possesses three crystal polymorphs in nature: brookite, anatase, and rutile. Brookite is rare in nature. Rutile is a high temperature stable phase, while anatase is not as thermodynamically stable as rutile. The phase transformation temperature of TiO2 is illustrated in Table 1. The crystal structures of TiO2 are illustrated in Fig.1. The basic unit that composes TiO2 is TiO6 octahedrons. The crystal structure of TiO2 is determined by the distortion of each octahedron and the way how the TiO6 octahedrons are connected. For rutile, neighboring octahedrons share one corner and are stacked with their long axis alternating by 90◦ . In anatase, neighboring octahedrons share the edge. Brookite is considered to be built up from the octahedrons that both corner and edge are shared[9] . For the practical applications, it is anatase and rutile structures that are widely used. They are both composed of Ti ions surrounded by six oxygen atoms in an octahedral configuration. The primary structural differences between the two phases are the degree of the distorted octahedrons and the way in which octahedrons are connected. Rutile is made up of parallel chains of octahedrons, the octahedron is somewhat distorted, and every octahedron is connected with neighboring ten other octahedrons. The octahedrons in anatase show apparent prismatic distortions, every octahedron is connected with neighboring eight other octahedrons. Anatase has larger Ti-Ti distances and shorter Ti-O distances than rutile. Table 2 gives some X-ray data about the structure of titanium dioxide. Owing to the different crystal structures of TiO2 , each phase possesses dis†

Prof., to whom correspondence should be addressed, E-mail: [email protected], [email protected]

tinct physical properties. The comparisons of the typical physical and chemical properties of titanium dioxide are listed in Table 3. It is apparent that TiO2 is a wide band gap oxide semiconductor. TiO2 has high refractive index, excellent optical transmittance in the visible and near-infrared region, and high dielectric constant. Basically, both anatase and rutile have tetragonal structure. However, rutile is denser than anatase, and possesses higher refractive index. Thus, rutile is more able to scattering light and has screening effect against UV light. 1.2 Applications of TiO2 Traditionally, TiO2 has been widely used as a kind of pigment in industrial and architectural field[13] . In 1972, Fujishima and Honda first discovered the splitting of water on TiO2 electrode[14] . Since then, the photoelectrochemical solar energy conversion application has attracted tremendous interests of chemists, physicists, and material scientists[15] . However, the low quantum yield for photochemical conversion of solar energy hinders the development in this field. O0 Regan and M.Gr¨atzel invented a dye-sensitized nanocrystalline solar cell, which utilized the dye molecules to improve efficiency of solar cells[16] . So far, the highest conversion efficiency is exceeding 11.3%[17] . In 1977, Frank and Bard et al. first extended the application of TiO2 in the environmental purification field[18] . They photocatalyzed CN− and SO2− 3 , and opened a new area which is called heterogeneous photocatalysis. From then on, the interests to study TiO2 have been shifted to the photocatalytic field. When the energy of the illuminated light on TiO2 is higher than the band gap of TiO2 , photoinduced electronhole pairs can be generated. The photogenerated holes in the valence band possess a strong oxidizing power which accounts for the decomposition of most organic compounds. Meanwhile, photogenerated electrons in the conduction band reduce absorbates on the surface of TiO2 . By far, the most successful applica-

676

J. Mater. Sci. Technol., Vol.24 No.5, 2008

Table 1 Phase transformation temperature of TiO2 Phase transformation Brookite-rutile Anatase-rutile

Phase transformation temperature 500–600◦ C 800◦ C 850◦ C 250◦ C (4000 kPa) 680◦ C 620◦ C (dry envirionments) 570◦ C (humid argon)

Reference [1] [2] [3–5] [6] [7] [8] [8]

Table 2 Some X-ray data about the structure of titanium dioxide[10] Crystal structure Rutile Anatase Brookite

System Tetragonal Tetragonal Rhombohedral

Space group D414h − P 42 /mnm D419h − I41 /amd D215h − P bca

Lattice constants/nm b c/a a c 0.4584 – 0.2953 0.644 0.3733 – 0.937 2.51 0.5436 0.9166 0.5135 0.944

Fig.1 Crystal structure of TiO2 : (a) anatase, (b) brookite, (c) rutile

Table 3 Typical physical and chemical properties of titanium dioxide[9,11,12] Anatase Brookite Rutile Density/(g/cm−3 ) 3.83 4.17 4.24 Melting point/◦ C Turning into rutile Turning into rutile 1870 Boiling point/◦ C 2927 (at pressure pO2 101.325 kPa) Band gap/eV 3.2 – 3 Refractive index ng 2.5688 2.809 2.9476 np 2.6584 2.677 2.6506 Dielectric properties 55 78 110–117 Hardness on mineralogical scale (Mohs scale) 5–6.5

tion of TiO2 is in the degradation of organic molecules. It is reported that most organic pollutants could be totally destructed by TiO2 photocatalysts[19–22] . According to statistical data, Japan and U.S. play the dominant role in promoting products based on semiconductor photocatalysis[23] . Wang et al.[24] first reported a UV light induced super-hydrophilic TiO2 surface which displayed 0◦ contact angle. They attributed the highly amphiphilic character of TiO2 films to the formation of microstructural distribution of hydrophilic and oleophilic domains. These domains resulted from photogenerated Ti3+ defect sites which were favorable for dissociative water adsorption[25] . Because of the superhydrophilic properties, TiO2 films have been widely used in producing antifogging and self-cleaning materials, for example, automobile exterior view mirrors and ceramic building materials[26,27] . Owing to the large surface-to-volume ratio of

nano-crystalline TiO2 films, TiO2 has also been used as a gas sensor[28–30] . This application utilizes the change of electrical signal when gas is adsorbed on the surface of TiO2 films. In the biomaterial field, synthesizing oxide layers on bio-metal surfaces have attracted much attention. The oxide films are ideal biomaterials, because they provide corrosion resistance and good blood compatibility. Among the oxides, TiO2 is a promising candidate as a kind of biomaterial for its good blood compatibility and chemical stability[31–34] . Matsumoto et al. first discovered the roomtemperature ferromagnetism of Co-doped TiO2 [35] . This discovery triggered great interests among scientists in its potential as a dilute magnetic semiconductor (DMS) for device applications[36–38] . 1.3 Shortcomings of TiO2 photocatalysts It is well-known that TiO2 is a promising candi-

J. Mater. Sci. Technol., Vol.24 No.5, 2008

date for photocatalysis because it is chemically and biologically inert, photocatalytically active, easy to produce and to use, sunlight activated, non-toxic and cheap. TiO2 is very close to being an ideal photocatalyst, but unfortunately not. Like all the other things in the world, perfect photocatalyst seems not to exist. Despite those outstanding properties which TiO2 displays, there are many shortcomings of TiO2 as photocatalysts. The main problems are: (1) The major drawback of TiO2 is its wide band gap. This means that only a small portion of the solar light can be absorbed in the UV region, which occupies only 5% of the total solar spectrum. Thus, efficient absorption of the visible light which constitutes the major part of solar spectrum is prevented and the photocatalytic efficiency of TiO2 is hindered. (2) The recombination rate of photogenerated electron-hole carriers is rapid. The lifespan of photogenerated electron-hole carriers are approximately 10 ns. The rapid recombination rate of photogenerated electron-hole carriers decreases photo quantum efficiency of TiO2 . (3) At the initial stage of the study of photocatalysis, TiO2 was used in the form of powders. After photocatalytic reaction, a filtration step was requested to separate photocatalysts from slurry. This process adds extra cost to the development of commercial applications. 1.4 Modification methods of TiO2 photocatalysts Many methods have been proposed to overcome those shortcomings which constrain the wide usage of TiO2 photocatalysts. Some of these methods are briefly introduced here: (1) Surface sensitization. If the sensitizers are absorbed (either chemisorbed or physisorbed) on the surface of TiO2 , they can be much more easily excited than TiO2 . Hence, the excitation process can be efficiently improved[39–42] . The sensitizers can extend the range of excitation energies of the TiO2 into visible region. However, the sensitizers themselves can be degraded. Thereby, there is a need to add more sensitizers in reaction systems. Because of this reason, the publication of this kind of modification method has diminished these years. (2) Metal-doped and nonmetal-doped TiO2 . Doping has been proven to be a great method to enhance the photocatalytic activity. Metal-doping has long been used[43–46] . In recent years, nonmetal-doping has demonstrated great potential to be applied in the photocatalytic field[47–50] . However, controversial experimental results have been reported in literature, and different theoretical models have been proposed to analyze these phenomena. Yet, the mechanisms of the optical and photocatalytic properties of nonmetaldoped semiconductors are still open to discuss. (3) Coupling of different kinds of semiconductor photocatalysts with different energy levels[51–57] . The benefits of this method are: the absorption spectrum can be expanded and the carrier separation is effective. The wide band-gap semiconductor can be photoactivated by narrow band-gap semiconductors. In addition, the charge injection from the conduction band of the narrow band-gap semiconductor to that of TiO2 can lead to efficient charge separation by

677

Fig.2 Energy diagram for the heterogeneous anatase/rutile TiO2 films at pH 7[58]

reducing the electron-hole recombination rate. The synergetic effect of composite semiconductor of anatase and rutile on the photocatalytic activity has also been reported[58] . Figure 2 shows the energy diagram for the heterogeneous anatase/rutile TiO2 films at pH 7. The transfer of photogenerated electrons from the anatase to rutile phase may reduce the electron-hole recombination rate. Hence, the composite semiconductor of anatase and rutile shows better photocatalytic activity than pure anatase phase. (4) Preparation of TiO2 films. In order to solve the shortcoming of separating TiO2 particles and avoid the filtration step, the techniques of immobilization of photocatalysts have been developed[59–61] . Various methods have been proposed to prepare TiO2 films[47,62–66] . Photocatalytic results demonstrate that the films are almost as effective as powder TiO2 catalysts which are used in slurry suspension systems. In the following part of this review, the authors would like to focus on two aspects among these modification methods: the introduction of preparation techniques of TiO2 thin films, and the discussion of the effects of doping on the electronic structures and photocatalytic properties of TiO2 . 2. Doped-TiO2 Photocatalysts Doped-TiO2 photocatalysts (either with metal ions or nonmetal ions) have been proven to enhance the photocatalytic activity by red-shifting the absorption edge to lower energies direction. Hence, these doped-TiO2 photocatalysts have been widely prepared and discussed for years. However, it remains difficult to summarize the details of the mechanisms of the optical and photocatalytic properties of the doped-TiO2 semiconductors due to the various experimental conditions, different kinds of preparation methods and the determination standards of photoreactivity. A theoretical analysis by computer simulation is expected to assist material scientists to clarify the impurity-doping effects at atomic level. It is well known that the band structure is one of the most important factors that evaluate the efficiency of TiO2 based photocatalysts. In this section, we discuss the theoretical mechanism of doped-TiO2 photocatalysts and review the experimental photocatalytic property

678

J. Mater. Sci. Technol., Vol.24 No.5, 2008

of the visible-light reactive TiO2 photocatalysts. 2.1 Metal-doped TiO2 photocatalysts 2.1.1 Experimental work Metal-doping was the dominant way at the initial stage of the study about doping effect on the photocatalytic properties of TiO2 . It was reported that the introduction of metal ions into the TiO2 matrix could significantly influence photoactivity, charge carrier recombination rates, and interfacial electron-transfer rates[43] . Choi and co-workers systematically studied the role of metal-doping in quantum-sized TiO2 photocatalysts[43] . They explained the role of metal ion dopants of TiO2 photocatalysts by photochemical and photophysical experiments. The ionic radii of the dopant metal ions they chose were similar to that of Ti4+ , thus it would be much easier for the substitution of the metal ions into the lattice of TiO2 matrix. All of the dopants showed that an optimum concentration existed in the degradation of CHCl3 . According to their work, doping with Fe3+ , Mo5+ , Ru3+ , Os3+ , Re5+ , V4+ , and Rh3+ at 0.1–0.5 at. pct could significantly increase the photoactivity. For Co3+ and Al3+ doping, the photoreactivity would decrease. Choi et al. concluded that the relative efficiency of a metal ion dopant depended on whether it served as a mediator of interfacial charge transfer or as a recombination center. They claimed that whether the dopant acted as an effective trap depended on the dopant concentration, the energy level of dopants within the TiO2 lattice, their d electronic configuration, the distribution of dopants within the particles, the electron donor concentration, and the incident light intensity. The most important factor that influenced the photoactivity of metal-doped TiO2 was the enhanced interfacial charge transfer in the presence of effective dopants. Fe3+ -doped TiO2 has been reported to be efficient in the degradation of many kinds of organic pollutants. The introduction of iron cations can exert strong influence on the charge carrier recombination time, and extend the absorption edge of TiO2 into the visible light region. Xin et al. used sol-gel method to prepare Fe3+ -doped TiO2 with different Fe3+ concentration[67] . According to their result, an optimal dopant concentration existed. The mechanism of photoinduced carriers separation and recombination processes was explained as follows: when the dopant content was less than 0.03 mol%, Fe3+ acted as the traps to capture the photoinduced electrons, which prohibited the recombination of photoinduced carriers and improved the photocatalytic activity of Fe3+ -doped TiO2 ; while when the dopant content exceeded 0.03 mol%, Fe2 O3 formed and became the recombination centers of photoinduced carriers. The level of conduction band of TiO2 was higher than that of Fe2 O3 , but the level of valence band of TiO2 is lower than that of Fe2 O3 . Hence, when the phase of Fe2 O3 formed, the electrons in conduction band and the holes in the valence band of TiO2 could transfer to Fe2 O3 . This led to the quick recombination of photoinduced electrons and holes. Ad´ an et al. prepared iron-doped anatase TiO2 by using microemulsion method and found that the introduction of Fe3+ cations into TiO2 structure induced an enhancement of the photocatalytic activity for the doping level up

to 1 wt pct[68] . They ascribed the deterioration of photoactivities of high iron doping TiO2 to the high levels of surface segregation of iron-containing amorphous oxidic phases. Klosek and Raftery synthesized V-doped TiO2 photocatalyst by ambient chemical method[69] . The prepared photocatalyst was active under visible light irradiation, because the vanadium centers could be excited by visible light and the electrons were transferred to the TiO2 conduction band. Anpo and co-workers prepared the Cr3+ doped TiO2 and found that the chromium ions effectively improved the photocatalytic activity under visible light[70] . Their study also demonstrated that an optimal doping concentration existed. They attributed the enhancement of visible-light induced photocatalytic activity to the excitation of 3d electrons of Cr3+ which were transferred to the conduction band of TiO2 . Lam et al.[71] implanted Cr ions into TiO2 thin films which were made by sol-gel method. Their results demonstrated that there was a need to control the amount of Cr ions in the TiO2 to obtain efficient photocatalytic degradation of formaldehyde vapor. Venkatachalam et al. made Zr4+ doped nano-TiO2 by sol-gel method[72] . Even though the introduction of Zr4+ in TiO2 matrix caused a wider band gap than pure TiO2 , it did hinder the growth of grain size, increase the surface area, decrease the anataserutile phase transformation and accelerate the surface hydroxylation. The enhancement of absorption of 4chlorophenol on the surface of catalysts and decrease of the particle size induced by the introduction of Zr4+ were responsible for the high activity of the catalysts. Hence, the Zr4+ doped nano-TiO2 possessed a better photocatalytic activity of the degradation of 4chlorophenol than both nano-TiO2 and commercial Degussa P25 product did. Di Paola et al. prepared polycrystalline TiO2 powders which were doped with transition metal ions (Co, Cr, Cu, Fe, Mo, V and W)[73] . Their study indicated that W-doped TiO2 was the most efficient one for the photodegradation of benzoic acid and 4-nitrophenol. Co-doped TiO2 was more efficient than pure TiO2 in the degradation of methanoic acid. The nature of the reacting molecules, the acid-base and electronic properties of the photocatalysts were taken into account to explain the differences of the photoreactivity. Brezov´a et al. investigated the photocatalytic decomposition of phenol using metal-doped TiO2 photocatalysts supported by the sol-gel method on glass fibers[46,74] . Their study demonstrated that the photoactivity of doped TiO2 was strongly dependent on the character and concentration of dopant ions. The best results in terms of phenol decomposition were obtained for dopant-free TiO2 , Li+ -, Zn2+ -, and Ptdoped TiO2 . The addition of Co3+ , Cr3+ , Ce3+ , Mn2+ , Al3+ , and Fe3+ ions in the TiO2 photocatalyst had a detrimental effect on its photocatalytic activity. Bouras et al. compared the photocatalytic properties of TiO2 with and without metal ions dopants[75] . They used sol-gel method to prepare Fe3+ , Cr3+ , and Co2+ doped TiO2 with dopant level ranged in a large domain. Their study showed that the introduction of Fe3+ ions induced energy states located within the band gap of TiO2 ; Co2+ doping caused a mixed behavior involving both doped titania and cobalt titan-

J. Mater. Sci. Technol., Vol.24 No.5, 2008

Fig.3 Schematic diagram of the metal ion-implantation method[78]

Fig.4 UV-vis absorption spectra of TiO2 (a) and Cr ion-implanted TiO2 photocatalysts (b–d). The amount of implanted Cr ions (µmol/g) is 0 (a), 0.22 (b), 0.66 (c), and 1.3 (d)[77]

Fig.5 Photoluminescence TiO2 [79]

spectra

of

metal

doped

ate; and Cr3+ doping led to the formation of an amorphous material by dissolving metal ions. Even though all doped materials they studied showed strong absorption in visible light region, their photocatalytic efficiencies were deteriorated as compared to pure TiO2 . The loss of crystallinity or transition from anatase to rutile seemed to account for the degradation of photocatalytic properties of doped TiO2 . Anpo et al. used metal ion-implantation method to modify the bulk electronic properties of TiO2 photocatalysts[76–78] . The schematic diagram of the metal ion-implantation method is shown in Fig.3. Transition metal ions, such as V, Cr, Mn, Fe and Ni, were implanted in the TiO2 matrix. The absorption edge of these doped-TiO2 photocatalysts showed

679

a shift toward visible light region, which was caused by the high energy implantation process as well as the interaction of the transition metal ions with the TiO2 catalysts. Figure 4 shows the UV-vis absoprtion spectra of Cr-doped and pure TiO2 . Compared with metal-doped TiO2 prepared by chemical methods, Anpo et al. claimed that the metal ionimplantation method was an efficient way to retain the photocatalytic properties of TiO2 under UV irradiation. Because the metal ions implanted in the TiO2 matrix did not work as electrons and holes recombination centers, but modified the electronic properties of metal-doped TiO2 . The modification of electronic structure of TiO2 enabled these photocatalysts to absorb and operate effectively not only under UV but also under visible light irradiation. By using solution combustion method, Nagaveni prepared W, V, Ce, Zr, Fe, and Cu metal-doped anatase TiO2 [79] . Figure 5 is the photoluminescence (PL) spectra of metal-doped TiO2 . As seen from the spectra, the PL intensities were quite sensitive to metal doping and the peak positions in the PL of the metal-doped TiO2 were consistent with those in pure TiO2 . The decrease in the emission intensities might originate from two reasons: one was the differences in electronic structure of the metal-doped TiO2 and the other was the introduction of new defect sites such as oxide ion vacancy. Photocatalytic activity experiments demonstrated that the photoactivity of metaldoped TiO2 was inferior to that of undoped-TiO2 . This was due to that the substituted metal ions acted as recombination centers instead of the trap sites of electrons and holes and the increase of the recombination rate of electron-hole pairs as confirmed by PL studies. 2.1.2 Theoretical analysis Computer simulation is a powerful tool to clarify the electronic structures of impurity states at atomic scale[80–84] . Casarin et al. used density functional cluster method to study the electronic properties of Nb-doped bulk and (110) surface of rutile TiO2 [80] . Their results showed that the electronic structure was significantly perturbed by Nb doping. The influence of Nb dopants on the electronic structure was quite similar to that in the bulk and at the surface of TiO2 . By using DV-Xα method, Nishikawa et al. calculated the electronic states of pure and metal-doped (V3+ , V4+ , V5+ , Cr3+ , Mn3+ , Fe3+ , Co3+ , Ni2+ , Ni3+ , and Rh3+ ) anatase and rutile TiO2 crystals[81] . Their study showed that the band gap decreased by replacing Ti4+ with metal ions. These results were in agreement with the experimental light absorption spectra, of which the metal-doped TiO2 shifted to the visible light region. The decrease of band gap was attributed to the overlap of 3d metal orbitals with 2p orbitals of oxygen ions in TiO2 and the increase in the degree of the covalent bond character between the replacing metal ion and an oxygen ion. Karvinen et al. used ab initio Hartree-Fock method to calculate the metal-doped TiO2 [82] . Their results demonstrated that the doping of Ti3+ , V3+ , Cr3+ , Mn3+ and Fe3+ could narrow the band gap of anatase TiO2 . For rutile TiO2 , doping of V3+ , Mn3+ and Fe3+ did not apparently alter the band gap. However, Cr3+ doping could broaden the band gap of rutile.

680

J. Mater. Sci. Technol., Vol.24 No.5, 2008

diation.

Fig.6 DOS of the metal-doped TiO2 (Ti1−x Ax O2 : A=V, Cr, Mn, Fe, Co or Ni). Gray solid lines: total DOS; black solid lines: dopant0 s DOS[84]

Umebayashi et al. investigated the electronic structures of 3d transition metals doped TiO2 by ab initio band calculation based on the density functional theory (DFT)[84] . Figure 6 shows the density of states (DOS) of the metal-doped TiO2 . It was seen that the introduction of 3d metal could create an occupied level in the band gap or valence band of TiO2 . The doping of V, Cr, Mn, or Fe would induce a localized level in the band gap. While for Co, the energy of localized level lay at the top of the valence band. For Ni doped TiO2 , the electrons from the Ni dopant were quite delocalized, and significantly contributed to the formation of the valence band with the O 2p and Ti 3d electrons. Together with the published absorption and photoconductivity spectra, they concluded that the t2g state of the dopant played an important role in the photoresponse of TiO2 under visible light irra-

2.2 Nonmetal-doped TiO2 photocatalysts The transition metal doping has been extensively studied to extend the optical absorption of TiO2 to visible light. However, the metal-doping of TiO2 photocatalysts also has some drawbacks, such as thermal instability and increase in the carrier-recombination centers. Whereas, the nonmetal-doped TiO2 seems to be more promising photocatalysts candidates. The vast research of nonmetal-doped TiO2 photocatalysts was widely carried out since the early 1990s. Among all the nonmetal-doped TiO2 photocatalysts, N-doped TiO2 is the most studied system[47,48,85–91] . Subsequently, several other visible-light active materials, such as B-[92,93] , C-[49,94–96] , F-[97–101] , P-[102,103] , S-[50,104–108] , I-[109–111] doped TiO2 were reported. 2.2.1 Nitrogen-doped TiO2 Among all these nonmetal-doped TiO2 systems, nitrogen-doping is the earliest studied system. Furthermore, it has attracted great attentions and becomes the most extensive studied system because of its relatively easier preparation method. N-doped TiO2 not only improves absorption in the visible region but also demonstrates remarkable photocatalytic properties. The first N-doped TiO2 was proposed by Sato in 1986[112] . He attributed the photoresponse of visible light to the NOx impurity introduced during preparation process. However, it was not until years later that researchers began to focus on the benign results which N-doped TiO2 brought. Up to date, there have been several models which have been proposed to describe the origin of the visiblelight photocatalytic activity of N-doped TiO2 . These models include: (1) Narrowing of the band gap of TiO2 . Morikawa et al. emphasized the importance of nitrogen-doped TiO2 in 2001[47,113] . Based on the spin-restricted local density approximation calculation, they proposed that the substitutional doping of N for O in the anatase TiO2 would lead the narrowing of the TiO2 band gap by mixing N 2p with O 2p states. They emphasized the importance of the substitutional doping of N for O in the TiO2 matrix in their experiment. Figure 7(a)–(b) shows the total density of states (TDOS) and projected density of states (PDOS) of doped TiO2 systems. Figure 7(c) shows the comparison of the photoactivity of TiO2−x Nx and

Fig.7 (a) TDOS of doped TiO2 , (b) PDOS of the doped anion sites. The dopants F, N, C, S, and P were located at a substitutional site for an O atom in anatase TiO2 . The results for N doping at an interstitial site (Ni -doped) and that at both substitutional and interstitial sites (Ni+s -doped) are also shown, (c) CO2 evolution as a function of irradiation time during the photodegradation of acetaldehyde gas under UV irradiation and visible irradiation (TiO2−x Nx samples: solid circle; TiO2 samples: open squares)[47]

J. Mater. Sci. Technol., Vol.24 No.5, 2008

TiO2 samples. It could be seen that the photoactivity of TiO2−x Nx was better than that of TiO2 samples in the visible range of irradiation, and the N-doped and undoped TiO2 demonstrated similar photoactivity in the UV region. Many other groups have also experimentally proved the band gap narrowing of nitrogendoped TiO2 [85,114] . (2) Mid-gap localized states of dopants. However, by means of spin-polarized GGA calculations, Valentin et al. found that the N 2p orbitals were localized above the O 2p valence bands, which explained the red shift of the optical absorption edge of N-doped TiO2 [48,115] . Excitation from the occupied high-energy states to the conduction band (CB) could account for the optical absorption edge shift to the lower energy of visible light region. They also found that N-doping in TiO2 induced the decrease of the formation energy of an oxygen vacancy, and oxygen vacancy formation usually accompanied the nitrogen doping. They claimed that whether the substitution or interstitial doping occurred depended on the specific preparation conditions, such as the oxygen concentration in the atmosphere and the annealing temperature. Lindgren et al. first reported on the preparation of nitrogen doped TiO2 thin films[116] . They experimentally confirmed that the states introduced by nitrogen lay close to the valence band edge. (3) Formation of color centers. Kuznetsov and Serpone carried out a series of experiments to study nonmetal-doped TiO2 (C-, N-, S-doped) and proposed that the red-shift of the absorption edge of TiO2 was due to the formation of oxygen vacancies and the formation of the color centers[117,118] . 2.2.2 Sulfur-doped TiO2 Asahi et al. predicted that the doping of N or S would be most effective among several nonmetals for the mixing of their 2p states with O 2p states. However, they stated that the S ion was hard to be introduced in TiO2 matrix for the large ionic radius of S element[47] . Umebayashi and co-workers successfully prepared S-doped TiO2 by various methods[50,104–106] . However, according to their experiment results, the S ions could either substitute lattice titanium atoms to form S-cation doped TiO2 or substitute lattice oxygen atoms to form S-anion doped TiO2 . It depended on the preparation conditions. If the S-doped TiO2 photocatalyst was prepared by chemically modified method, the S4+ could be detected from X-ray photoelectron spectra (XPS)[105] . Photocatalytic experimental results demonstrated that this S-cation doped TiO2 showed strong visible light absorption and high degradation rate of methylene blue, 2-propanol and partial oxidation of adamantane. Their total density of states (TDOS) calculation analysis displayed that there was a localized S 3s states existed in the mid-gap of the TiO2 , and electron transfer between the valence band and this level would be induced by the irradiation of visible light. If the S-doped TiO2 photocatalyst was prepared by oxidation annealing of TiS2 or ion implantation methods, the S ions would be implanted into the crystal structure of TiO2 and occupied the original O-atom sites[104,106] . XPS data also confirmed the substitution of S to O. Ab initio calculation showed that the mixing of S 3p states with valence band widened the valence band and band gap narrowing was the direct result of S-doping[108] .

681

2.2.3 Carbon-doped TiO2 By theoretically studying the effects of C-, N-, and S-doping on TiO2 , Wang and Lewis stated that the carbon-doped TiO2 was the most promising photocatalyst[119] . There were two reasons to support their conclusion. One was the significant overlap between the O 2p state and the carbon states near the valence band edge. The other one was the large valence band red shift of carbon-doped TiO2 . Wu et al. synthesized carbon-doped TiO2 micro/nanospheres and nanotubes by using chemical vapor deposition method[120] . The percentage of carbon in TiO2 microspheres and nanotubes was approximately 3% and 5%, respectively. Compared to the pure TiO2 , band gap narrowing occurred (2.78 eV for the carbondoped TiO2 microspheres; 2.72 eV for the carbondoped TiO2 nanotubes). Quantum chemical calculations demonstrated that the possible structure of doping was the insertion of carbon into TiO2 crystal lattice, either in cation or anion form. Many other researchers do not agree on the band gap narrowing of carbon-doped TiO2 [121–123] . Valentin et al. conducted density functional theory calculation and concluded that the introduction of carbon into TiO2 matrix could result in modest variations of the band gap and induce several localized occupied states in the gap[121] . Park et al. also stated that the visible light response of C-doped TiO2 was attributed to the presence of isolated band-gap states rather than band-gap narrowing[122] . 2.2.4 Fluorine-doped TiO2 Yu et al. prepared Fdoped TiO2 by hydrolysis of titanium tetraisopropoxide in a mixed NH4 F-H2 O solution[101,124] . The F doping not only improved the crystallinity of anatase but also suppressed the formation of brookite phase and prevented the phase transition of anatase to rutile. Absorption spectra showed that the band gap of F-doped TiO2 was narrower than that of pure TiO2 . Compared with Degussa P25, F-doped TiO2 showed better photocatalytic activity when the atomic ratios of F to Ti were in the range of 0.5–3 at 500◦ C. By using ion implantation technique, Yamaki et al. prepared F-doped TiO2 and investigated the details about damage recovery and impurity diffusion[97] . Their results demonstrated that TiO2−x Fx along with the recovery of the radiation damage occurred when the samples were annealed at 300, 600, 1000, and 1200◦ C for 5 h. Implanted F atoms could diffuse to the outer surface together with the recovery of disorder layer. By using fullpotential linearlized augmented plane wave (FLAPW) method, they calculated the electronic structures of the F-doped TiO2 . The substitution of F to O led to a modification of the electronic structure around the edge of conduction band of TiO2 . This modification could result in a reduction in the effective band gap, thereby inducing visible-light photoresponse. 2.2.5 Boron-doped TiO2 Zhao et al. reported a kind of TiO2 doped with both boron and Ni2 O3 [93] . They demonstrated that B was incorporated into TiO2 and Ni was separated from the anatase TiO2 . They attributed the red-shift of the absorption of TiO2 to the incorporation of boron, and the enhancement of photoactivity to the load of Ni2 O3 . By DFT calculation, B doped into substitutional site of TiO2 was proven to be the most stable state, and the nar-

682

J. Mater. Sci. Technol., Vol.24 No.5, 2008

rowing band gap of TiO2 was due to the p orbital of B mixing with O 2p orbitals. Ni2 O3 acted as electron traps and facilitated the photo-induced electron-hole separation. By using a sol-gel method, Chen and coworkers synthesized B-doped TiO2 nanoparticles[92] . They showed that doped boron was present in the form of B3+ in the B-doped TiO2 , forming a possible chemical environment like Ti-O-B. B3+ was likely to be incorporated into the interstitial position of TiO2 matrix. The doping of boron could inhibit grain growth and facilitate the transition from anatase to rutile before the B2 O3 phase formed. Compared with pure TiO2 , B-doped TiO2 showed higher photocatalytic property in the UV range. Theoretical analysis demonstrated that the substitution of B to O atoms would result in new midgap states, which could account for the red-shift of absorption edge and enhancement of photocatalytic efficiency[125] . 2.2.6 Phosphorus-doped TiO2 Mesoporous TiO2 is of great value as a photocatalyst because it possesses the high surface-to-volume ratio and offers much more active reactive sites during photocatalytic reaction. Based on the knowledge that the thermal stability and acidity of mesoporous material could be greatly improved by post-treatment with phosphoric acid, Yu et al. prepared P-doped mesoporous TiO2 by a surfactant-templated approach[126] . The P-doped mesoporous TiO2 demonstrated better photocatalytic ability than pure mesoporous TiO2 and commercial P25 product. This high photocatalytic ability was attributed to the extended band gap, larger surface area, and the existence of Ti ions in a tetrahedral coordination of P-doped mesoporous TiO2 . K¨or¨osi and D´ek´ any prepared high surface area (356–478 m2 /g) and high thermal stable P-doped mesoporous TiO2 by sol-gel method without the addition of surfactants[127] . The specific surface area, thermal stability, transition temperature from anatase to rutile, and photocatalytic activity largely depended on the phosphate content. Shi et al. prepared P-doped TiO2 nanoparticles by the sol-gel method[103] . They showed that P-doped TiO2 possessed narrower band gap and higher ability of nicotinamide adenine dinucleotide (NADH) photogeneration under visible light irradiation than pure TiO2 . 2.2.7 Iodine-doped TiO2 In 2005, Hong et al. prepared visible-light-activated iodine-doped TiO2 nanoparticles by using direct hydrolysis method[111] . According to their analysis, I5+ may substitute Ti4+ ions for their equivalent ionic radius. Optical absorption spectra indicated that I-doped TiO2 showed strong absorption in the range from 400 to 550 nm compared with commercial P25 and pure TiO2 . Photocatalytic experiments demonstrated that I-doped anatase TiO2 nanoparticles with 5 nm mean diameter possessed higher degree of mineralization than pure TiO2 under visible and UV, and visible light irradiation. Later, Long et al. carried out the first principle calculation to compare the electronic structure of Idoped TiO2 with that of N-doped and pure TiO2 [110] . According to their results, the band potential of Idoped anatase TiO2 shifted downwards, which implied the valence band of I-doped TiO2 should have a

stronger oxidative power than that of pure TiO2 . Additionally, the TiO6 octahedral distortion in I-doped anatase TiO2 was heavier than that in N-doped TiO2 . The dipole moment, which was generated from the distorted octahedral, of I-doped anatase TiO2 was larger than N-doped and pure TiO2 . The larger dipole moment was considered to be beneficial for the electron-hole separation, which could enhance photocatalytic properties upon visible light irradiation. Liu et al. synthesized I-doped mesoporous TiO2 with a bicrystalline framework by a two-step template hydrothermal route[109] . The prepared I-doped mesoporous TiO2 photocatalyst possessed excellent catalytic properties under both visible light and UVvisible light. This could be attributed to the bycrystalline framework (anatase and rutile), high crystallinity, larger surface area (157 m2 g−1 ), mesoporous structure, high absorbance in the visible light range (420 to 550 nm) of the I-doped mesoporous TiO2 . 2.2.8 Codoped TiO2 systems Many kinds of codoped TiO2 systems have been developed. The aim is to make good use of the synergistic effect which is introduced by doping TiO2 with more than one kind of elements. The studied co-doped TiO2 system includes: N, F codoped[128–131] , N, S codoped[132] , C, N codoped[133,134] , Br, Cl codoped[135] , and N, Fe (III) codoped TiO2 [136] . Li et al. synthesized N, F codoped TiO2 powders by spray pyrolysis technique[128–130] . They found that the prepared N, F codoped TiO2 powders possessed a porous and acidic surface. The photocatalytic experiment proved that the codoped TiO2 demonstrated higher photocatalytic activity under both UV and visible light irradiation than undoped TiO2 (which was prepared by the same technique) and commercial P25. Figure 8 presented the CO2 evolution from acetaldehyde decomposition over nonmetal-doped TiO2 powders with blue LED irradiation. It could be observed that the highest photocatalytic activity was obtained by N, F codoped TiO2 . Synergistic effect of the N, F elements codoping was utilized to elucidate the photocatalytic activity improvement because the doping of N ions could improved visible absorption and the doped F ions could enhance the surface acidity and the adsorption of reactant. Yu and coworkers prepared N, S codoped TiO2 by wet chemical method[132] . The prepared N, S codoped TiO2 showed higher photocatalytic activity for the oxidation of acetone or formaldehyde than commercial P25. It was the narrowing of the band gap, good crystallization, large surface area and two phase structures that explained the excellent photocatalytic properties of the N, S codoped TiO2 . Cong et al. also claimed that the narrowing of the band gap could account for the enhancement of photocatalytic properties of C, N codoped TiO2 [133] . Chen and coworkers prepared C, N codoped TiO2 by sol-gel method[134] . They found that more methylene blue could be absorbed on the surface of C, N codoped TiO2 than pure TiO2 . The doping of C could act as a role of photosensitizer like organic dyes, which could be excited and injected electrons into the conduction band of TiO2 . N doping could also promote electron transfer, induce intra-band-gap states close to the valence band edges, and shift flat-band potential position to a higher level than pure TiO2 . Thus,

J. Mater. Sci. Technol., Vol.24 No.5, 2008

683

accompanies the use of particles. Hence, it is a very economical way to develop TiO2 thin films. Many methods have been used to prepare TiO2 films, including wet chemical method, chemical vapor deposition, and physical vapor deposition method. In the following part, we would like to concentrate on some of the methods which are widely used to prepare TiO2 films.

Fig.8 CO2 evolution during the acetaldehyde decomposition of the nonmetal-doped TiO2 powders. The dotted line represents the blank test without photocatalyst[130]

the synergistic effect of carbon and nitrogen co-doping caused higher photocatalytic activity. Luo et al. synthesized Br and Cl codoped TiO2 by hydrothermal method[135] . The Br and Cl codoped TiO2 demonstrated higher photocatalytic activity than commercial P25 did for H2 and O2 production in Na2 CO3 . This enhancement of photocatalytic activity was attributed to the narrowing of band gap by the co-doping of Br− and Cl− . In order to obtain a photocatalyst with high photocatalytic ability under both UV and visible light irradiation, Cong et al. prepared nano-TiO2 codoped with N, and Fe (III)[136] . The best photocatalytic activity of the codoped system was achieved in the photocatalyst codoped with nitrogen and 0.5% Fe3+ . The mechanism of the good photocatalytic performance in this codoped system was that the nitrogen and Fe3+ ion led to the much narrowing of the band gap, which contributed to the improvement of the photocatalytic activity in the visible light region. In addition, suitable amount of Fe ion can trap the photogenerated electrons, while the nitrogen can trap part of photoinduced holes. Thus, the photoinduced electrons and holes could be separated and utilized efficiently. From the review of the studies of doped-TiO2 , it is obvious that there are intensive debates about the mechanism of the visible light response of the doped-TiO2 photocatalysts. In addition, the theoretical studies and experimental results are often controversial. The origin of these debates or controversy is due to the diversity of preparation methods and different calculation methods. In essence, photocatalytic activity itself is a complicated process which is affected by many factors. Usually, it is hard to separately investigate each factor on the photocatalytic activity exactly. Until now, any attempt to set up a universal law of the exact effects of dopants on TiO2 semiconductors and furthermore on the photocatalytic activity seems to be in vain. 3. Preparation Methods of TiO2 Films TiO2 thin films not only possess the virtues of photocatalysts, but also avoid the filtration process which

3.1 Sol-gel method[137–143] Sol-gel method is a kind of wet chemical method. When sol-gel method is applied to synthesize oxide films, the precursors can be alkoxide or inorganic salts. Usually, the precursors dissolve in water or organic solution. Then, the solute undergoes hydrolysis reaction and sol is obtained. After the polymerization and loss of solvent process, the liquid sol turns into solid gel. Thin films can be prepared on the substrates by spin-coating, dip-coating or spray-coating. The advantages of sol-gel method for preparing TiO2 films are: (1) no special apparatuses are required, easy to operate; (2) uniform films can be easily prepared, the purity of the films are high; (3) phase structure of films can be controlled; (4) applied to the industrial production. Negishi et al. prepared transparent TiO2 thin films via sol-gel method from tetraisopropyl titanium ethanol solution containing polyethylene glycol and diethylene glycol[137] . By decomposing gaseous acetaldehyde, they found that the quantum yield with the prepared TiO2 films was comparable to the commercial P25. Yu et al. used tetrabutylorthotitanate mixing with various amounts of polyethylene glycol (PEG) as precursor solutions to prepare TiO2 films[142] . The more PEG was added to the precursor solution, the larger the size and number of pores produced in the resultant films which resulted in the decrease of the transmittance of the films. The photocatalytic activity of TiO2 films depends on the pore size and number, the amount of -OH, the surface area and transmittance of TiO2 films. Yu et al. synthesized TiO2 films by sol-gel method and studied the effect of sulfuric acid treatment on the photocatalytic properties of TiO2 [143] . They found that the acid treatment enhanced photocatalytic activity of TiO2 due to the reduction of sodium ions and the increase in the adsorbed hydroxyl content on the surface of TiO2 films. A TEM image of TiO2 films prepared by sol-gel method is shown in Fig.9[143] . 3.2 Liquid phase deposition (LPD) method In 1988, Nagayama reported a novel method, which was called liquid phase deposition (LPD), to produce thin films[144] . During the deposition process, metal oxide thin films can be deposited onto the immersed substrates via a ligand exchange equilibrium reaction of metal-fluorocomplex ions and Fconsuming reaction by the addition of F-scavengers, such as boric acid or metal aluminum. The merits of LPD method are that it is very simple to operate and requires inexpensive equipment. The films prepared by LPD method possess dense structure and good chemical stability. Furthermore, LPD method can be applied to the preparation of films with large surface area, complex shapes, or on various kinds of substrates.

684

J. Mater. Sci. Technol., Vol.24 No.5, 2008

A typical TiO2 thin films preparation flow chat is given in Fig.10[152] . The disadvantages of LPD method are: (1) the films can only be prepared on the substrates with OH- radicle, (2) and the growth rate of the films is slow.

Fig.9 TEM of TiO2 film deposited on glass with 0.17 µm[143]

Fig.10 Flow chart to prepare TiO2 films by LPD method[152]

Fig.11 Schematic apparatus of CVD chamber[160]

TiO2 films have been synthesized by LPD method[145–151] . Normally, ammonium hexafluorotitanate and boric acid are used as treatment solution, with [TiF6 ]2− /H3 BO3 =1:2–4 as the most suitable molar ratio for the formation of thin films. Then, the clean substrates are immersed into the treatment solution for 48 h. The reaction is summarized as follows: [TiF6 ]2− + nH2 O ⇔ [TiF6−n (OH)n ]2− + nHF (1) H3 BO3 + 4HF ⇔ HBF4 + 3H2 O

(2)

3.3 Chemical vapor deposition (CVD) method[153–161] Chemical vapor deposition (CVD) method is widely used to prepare TiO2 films with high quality, high uniformity and controlled properties. CVD technique has many benefits, such as high deposition rate and easy to deposit films on complex substrates. However, the drawbacks of CVD are also prominent. The biggest one is the call for high temperature (600◦ C or above) during the CVD deposition process. Hence, many kinds of substrates are not fit for the preparation of TiO2 at this temperature. In order to lower the deposition temperature, plasma enhanced CVD (PECVD) and metal organic CVD (MOCVD) techniques have been developed. Figure 11 depicts the schematic apparatus of CVD chamber[160] . Battiston and coworkers deposited TiO2 thin films on different substrates, such as stainless steel, titanium, barium borosilicate glass and alumina by MOCVD[153] . The films are deposited at 420◦ C. Later, Battiston et al. combined the PECVD together with a metal organic precursor to develop a PEMOCVD method to deposit TiO2 films[161] . The TiO2 films were deposited at temperature between 393 and 523 K using titanium tetraisopropoxide as a precursor. Thus, TiO2 films could be deposited on some of the substrates which are very sensitive to temperature. 3.4 Magnetron sputtering method[116,162–176] Magnetron sputtering technique has been considered as a promising way to prepare TiO2 thin films. The merits of magnetron sputtering are large scale deposition and high deposition rate so that uniform and high quality films can be obtained. Typically, the preparation of TiO2 involves sputtering of Ti target in a vacuum system. Firstly, the furnace chamber is pumped down to ∼10−3 Pa, then argon and oxygen are introduced in the furnace and the deposition of TiO2 starts. The deposition rate of sputtering Ti target is higher than the sputtering of oxide target because metal targets generally have much higher sputtering yields than compound targets. Figure 12 shows the schematics of reactive magnetron sputtering system[172] . Sputtering pressure and substrate temperature are two important parameters in the magnetron sputtering process. Meng et al. systematically investigated the influence of sputtering pressure, substrate temperature on the properties of direct current (dc) reactive magnetron sputtered TiO2 films[163–166] . They found out that when the films were deposited at lower sputtering pressure (2×10−3 mbar), a dense structure with a mirror-like surface was obtained, while a porous structure with a rough surface and more voids was obtained under higher sputtering pressure (2×10−2 mbar). When the substrate temperature varied from ambient to 500◦ C, the orientation of the as-deposited films changed from (101) to random orientation, and finally towards (004) at temperature higher than 450◦ C.

J. Mater. Sci. Technol., Vol.24 No.5, 2008

685

range from 2.0 to 16.5%. Photocatalytic experiment demonstrated that the optimum concentration of substituted nitrogen was 6.0%. Cui et al. studied the influences of the sputtering power density and post annealing processes on the morphology and phase composition of TiO2 films by using mid-frequency dual targets magnetron sputtering technique[173,174] . Figures 13 and 14 shows the surface morphology of the TiO2 deposited at different power density and annealed at different temperatures[174] . According to their results, the crystallite size, phase composition, and stress state played a key role on the hydrophilic properties of TiO2 films. The contact angle versus annealing temperature is depicted in Fig.15[174] .

Fig.12 Schematic of reactive magnetron sputtering system[172]

However, the dc magnetron sputtering usually suffers from serious “cathode poisoning” problems which occur in the process of depositing insulator. The use of rf and mid-frequency dual targets magnetron sputtering technique can help to overcome this obstacle[171,174] . Kitano and coworkers prepared N-doped TiO2 films by rf magnetron sputtering method[171] . They changed the concentration of N in TiO2 by varying the concentration of N2 in the sputtering gas from 2% to 40%. The resulting concentration of N substituted within TiO2 was in the

3.5 Arc ion plating (AIP) technique[177–185] The arc ion plating technique involves a low voltage, high current plasma discharge which takes place between two electrodes. AIP is a deposition technique which possesses high ionization efficiency, fast deposition rate, high film density, high adhesion force and grain refinement. After DC bias AIP, pulse biased arc ion plating (PBAIP) is the second generation of AIP, which is a low temperature plasma process involving a pulsed bias of 400–1500 V applied on the substrate parallel to the DC bias circuit. Different from the DC bias, a pulsed bias suppresses micro-arc breakdown and enables deposition last long[177,178] . Figure 16 shows a schematic diagram of the PBAIP system. Zhang et al. prepared uniform and transparent TiO2 films on glass substrates by PBAIP technique[178] . According to their study, the films deposited at −300 V exhibited the

Fig.13 SEM images of the TiO2 thin films (prepared at the power density of 3.06 W/cm2 ) annealed at different temperatures: 800◦ C (a), 900◦ C (b), and 1000◦ C (c). These films are deposited on silicon wafer substrate with thickness approximately 143 nm. The shapes of the rutile phase are simulated and the crystallographic planes are indicated at the insets. The exposed facets of rutile crystal are mainly in the {110} planes[174]

Fig.14 SEM images of the TiO2 thin films (prepared at the power density of 2.22 W/cm2 ) annealed at different temperatures: 800◦ C (a), 900◦ C (b), and 1000◦ C (c). These films are deposited on silicon wafer substrate with thickness approximately 143 nm. The shapes of the rutile phase are simulated and the crystallographic planes are indicated at the insets. The exposed facets of rutile crystal are mainly in the {110} planes[174]

686

J. Mater. Sci. Technol., Vol.24 No.5, 2008

Fig.15 Contact angle as a function of annealing temperature. The substrate is silicon, the thickness of the films is approximately 143 nm. The phase composition is also illustrated. A: TiO2 films deposited at the power density of 3.06 W·cm−2 , B: TiO2 films deposited at the power density of 2.22 W·cm−2[174]

Fig.16 Schematic diagram of the PBAIP experimental setup[178]

Fig.17 Macroscopic image of TiO2 films deposited at different bias[178]

highest refractive index at 550 nm and the lowest root mean square roughness (Rrms) about 2.51 and 0.113 nm. Macroscopic image of TiO2 films was shown in Fig.17. Yumoto et al. carried out experiment to examine the photocatalytic decomposition of NO2 on TiO2 films deposited by AIP technique[181] . The decomposition of NO2 gas by TiO2 photocatalyst was high when the arc current was 60 A because of the low oxygen deficiency. They correlated the photocatalytic activity to the structure, composition and crystallinity of the TiO2 films. Chang et al. studied the photocatalytic properties of TiO2 films made by AIP[179] . Their experiment results illustrated that the photocatalytic activity of deposited TiO2 films was strongly dependent on the anatase phase quantity, which could be increased at higher oxygen partial pressures, higher deposition temperature, and lower substrate bias. Chang et al. also studied the photocatalytic performance of Cr or N doped TiO2 made by AIP method[180] . When the nitrogen partial pressure was 5%, the photocatalytic efficiency of the N-doped TiO2 was better than pure TiO2 . Further increase of nitrogen partial pressure could reduce the photocatalytic activity of the N-doped TiO2 . The doping of Cr reduced photoactivity because Cr (III) was considered to promote the recombination of electron and hole pairs. The drawback of this technique is the macropar-

ticles in the plasma. Hence, the wide applications have been prevented. However, great efforts have been made to overcome this obstacle. The most effective way to eliminate the macroparticles is the use of curved magnetic filters to effectively separate the plasma from the macroparticles. This technique refers to the combination of cathodic arc with the magnetic filters[185] . Zhao and Tay studied the influence of working pressure on properties of TiO2 thin films at room temperature[186] . The deposited films were amorphous with residual stress less than 0.5 GPa and smooth surface. The refractive index and extinction coefficient both increased with the decrease of working pressure. While the film transmittance increased with the increase of working pressure. Bendavid et al. found that the structure of TiO2 film strongly depended on the substrate bias[182] . Thus, by adjusting proper substrate bias, TiO2 films with amorphous, anatase, or rutile phase could be prepared. 4. Conclusions and Outlook TiO2 is a really wonderful material in photocatalysis field. It can decompose all the organic pollutants. However, the limits of this material are also obvious. Just as stated above, how to effectively utilize solar energy and effectively separate electrons and holes are still challenges. Scientists have already de-

J. Mater. Sci. Technol., Vol.24 No.5, 2008

veloped many ways to solve these problems. (1) Combining different techniques to achieve higher activity. Photoelectrochemically assisted catalysis has been developed since the middle of 1990s. With the application of an electric field, the photogenerated electrons could be driven away via the external circuit to the counter electrode compartment, the efficiency of oxidation induced by holes could be greatly improved at the semiconductor-electrolyte interface. (2) Design and development of novel photocatalysts. Up to now, many kinds of new photocatalysts have already been invented. The expansion of photocatalyst family is a proof of the prosperity of photocatalyst science. Based on the benign properties of TiO2 , we have confidence that TiO2 is and will continue to be a promising photocatalyst material in the future. REFERENCES [1 ] J.G.Li and T.Ishigaki: Acta Mater., 2004, 52, 5143. [2 ] X.Bokhimi, A.Morales, E.Ortiz, T.Lopez, R.Gomez and J.Navarrete: J. Sol-Gel Sci. Technol., 2004, 29, 31. [3 ] K.N.P.Kumar, K.Keizer and A.J.Burggraaf: J. Mater. Chem., 1993, 3, 1141. [4 ] B.Hunsche, M.Vergˇ ohl and A.Ritz: Thin Solid Films, 2006, 502, 188. [5 ] D.Wicaksana, A.Kobayashi and A.Kinbara: J. Vac. Sci. Technol. A, 1992, 10, 1479. [6 ] R.R.Bacsa and M.Gr¨ atzel: J. Am. Ceram. Soc., 1996, 79, 2185. [7 ] K.Okada, N.Yamamoto, Y.Kameshima, A.Yasumori and K.J.D.MacKenzie: J. Am. Ceram. Soc., 2001, 84, 591. [8 ] O.K.Varghese, D.W.Gong, M.Paulose, C.A.Grimes and E.C.Dickey: J. Mater. Res., 2003, 18, 156. [9 ] U.Diebold: Surf. Sci. Rep., 2003, 48, 53. [10] G.V.Samsonov: The Oxide Handbook, 2nd edn, IFI/Plenum Press, New York, 1982, 9. [11] L.Gao, S.Zheng and Q.H.Zhang: Nano-titanium Dioxide Photocatalysis Materials and its Application, 1st edn, Chemical Industry Press, Beijing, 2002, 35. [12] H.Tang, K.Prasad, R.Sanjinbs, P.E.Schmid and F.L´evy: J. Appl. Phys., 1994, 75, 2042. [13] J.H.Braun, A.Baidins and R.E.Marganski: Prog. Org. Coat., 1992, 20, 105. [14] K.Honda and A.Fujishima: Nature, 1972, 238, 37. [15] A.Fujishima and D.A.Tryk: in Photoelectrochemical Conversion, ed. K.Honda, Springer, Tokyo, 1999, 196. [16] B.O0 Regan and M.Gr¨ atzel: Nature, 1991, 353, 737. [17] M.Gr´ atzel: Inorg. Chem., 2005, 44, 6841. [18] S.N.Frank and A.J.Bard: J. Phys. Chem., 1977, 81, 1484. [19] A.L.Linsebigler, G.Q.Lu and J.T.Yates: Chem. Rev., 1995, 95, 735. [20] M.R.Hoffmann, S.T.Martin, W.Choi and D.W.Bahnemann: Chem. Rev., 1995, 95, 69. [21] Y.Hu and C.W.Yuan: J. Mater. Sci. Technol., 2006, 22, 239. [22] H.F.Liu, X.N.Cheng, J.Yang, X.H.Yan and H.B.Shi: J. Mater. Sci. Technol., 2007, 23, 123. [23] A.Mills and S.K.Lee: J. Photochem. Photobiol. A, 2002, 152, 233. [24] R.Wang, K.Hashimoto, A.Fujishima, M.Chikuni, E.Kojima, A.Katamura, M.Shimohigoshi and T.Watanabe: Nature, 1997, 388, 431.

687

[25] R.Wang, K.Hashimoto, A.Fujishima, M.Chikuni, E.Kojima, A.Kitamura, M.Shimohigoshi and T.Watanabe: Adv. Mater., 1998, 10, 135. [26] M.Nakamura, M.Kobayashi, N.Kuzuya, T.Komatsu and T.Mochizuka: Thin Solid Films, 2006, 502, 121. [27] A.Fujishima, T.N.Rao and D.A.Tryk: J. Photochem. Photobiol. C, 2000, 1, 1. [28] K.Zakrzewska, M.Radecka and M.Rekas: Thin Solid Films, 1997, 310, 161. [29] C.Garzella, E.Comini, E.Tempesti, C.Frigeri and G.Sberveglieri: Sens. Actuators. B, 2000, 68, 189. [30] Y.Shimizu, T.Hyodo and M.Egashira: Sens. Actuators. B, 2007, 121, 219. [31] F.Zhang, Z.Zheng, Y.Chen, X.Liu, A.Che and Z.Jiang: J. Biomed. Mater. Res., 1998, 42, 128. [32] J.X.Liu, D.Z.Yang, F.Shi and Y.J.Cai: Thin Solid Films, 2003, 429, 225. [33] S.Oh and S.Jin: Mater. Sci. Eng. C, 2006, 26, 1301. [34] H.C.Cheng, S.Y.Lee, C.C.Chen, Y.C.Shyng and K.L.Ou: Appl. Phys. Lett., 2006, 89, 173902. [35] Y.Matsumoto, M.Murakami, T.Shono, T.Hasegawa, T.Fukumura, M.Kawasaki, P.Ahmet, T.Chikyow, S.Koshihara and H.Koinuma: Science, 2001, 291, 854. [36] D.H.Kim, J.S.Yang, K.W.Lee, S.D.Bu, T.W.Noh, S.J.Oh, Y.W.Kim, J.S.Chung, H.Tanaka, H.Y.Lee and T.Kawai: Appl. Phys. Lett., 2002, 81, 2421. [37] J.Y.Kim, J.H.Park, B.G.Park, H.J.Noh, S.J.Oh, J.S.Yang, D.H.Kim, S.D.Bu, T.W.Noh, H.J.Lin, H.H.Hsieh and C.T.Chen: Phys. Rev. Lett., 2003, 90, 017401. [38] R.Janisch and N.A.Spaldin: Phys. Rev. B, 2006, 73, 035201. [39] A.Hagfeldt and M.Gr¨ atzel: Chem. Rev., 1995, 95, 49. [40] J.M.Rehm, G.L.McLendon, Y.Nagasawa, K.Yoshihara, J.Moser and M.Gr¨ atzel: J. Phys. Chem., 1996, 100, 9577. [41] D.Chatterjee and A.Mahata: Appl. Catal. B, 2001, 33, 119. [42] D.Chatterjee and S.Dasgupta: J. Photochem. Photobiol. C, 2005, 6, 186. [43] W.Choi, A.Termin and M.R.Hoffmann: J. Phys. Chem., 1994, 98, 13669. [44] K.Wilke and H.D.Breuer: J. Photochem. Photobiol. A, 1999, 121, 49. [45] M.I.Litter: Appl. Catal. B, 1999, 23, 89. [46] D.Dvoranov´ a, V.Brezov´ a, M.Maz´ ur and M.A.Malati: Appl. Catal. B, 2002, 37, 91. [47] R.Asahi, T.Morikawa, T.Ohwaki, K.Aoki and Y.Taga: Science, 2001, 293, 269. [48] C.D.Valentin, G.Pacchioni and A.Selloni: Phys. Rev. B, 2004, 70, 085116. [49] S.U.M.Khan, M.Al-Shahry and W.B.Ingler: Science, 2002, 297, 2243. [50] T.Umebayashi, T.Yamaki, H.Itoh and K.Asai: Appl. Phys. Lett., 2002, 81, 454. [51] J.E.Evans, K.W.Springer and J.Z.Zhang: J. Chem. Phys., 1994, 101, 6222. [52] K.Vinodgopal and P.V.Kamat: Environ. Sci. Technol., 1995, 29, 841. [53] M.D.Amiridis and J.P.Solart: Ind. Eng. Chem. Res., 1996, 35, 978. [54] L.Y.Shi, C.Z.Li, H.C.Gu and D.Y.Fang: Mater. Chem. Phys., 2000, 62, 62. [55] W.K.Ho, J.C.Yu, J.Lin, J.G.Yu and P.S.Li: Langmuir, 2004, 20, 5865. [56] S.J.Wang, X.T.Zhang, G.Cheng, X.H.Jiang, Y.C.Li, Y.B.Huang and Z.L.Du: Chem. Phys. Lett., 2005, 405, 63. [57] L.Wu, J.C.Yu and X.Z.Fu: J. Mol. Catal. A, 2006, 244, 25. [58] J.G.Yu, J.C.Yu, W.K.Ho and Z.T.Jiang: New J. Chem., 2002, 26, 607.

688

J. Mater. Sci. Technol., Vol.24 No.5, 2008

[59] J.Sabate, M.A.Anderson, H.Kikkawa, M.Edwards and C.G.Hill, Jr: J. Catal., 1991, 127, 167. [60] K.Vinodgopal, S.Hotchandani and P.V.Kamat: J. Phys. Chem., 1993, 97, 9040. [61] J.A.Byrne, B.R.Eggins, N.M.D.Brown, B.Mickinney and M.Rouse: Appl. Catal. B, 1998, 17, 25. [62] C.Guillard, B.Beaugiraud, C.Dutriez, J.M.Herrmann, H.Jaffrezic, N.Jaffrezic-Renault and M.Lacroix: Appl. Catal. B, 2002, 39, 331. [63] Y.J.Chen and D.D.Dionysiou: Appl. Catal. B, 2006, 69, 24. [64] H.Choi, E.Stathatos and D.D.Dionysiou: Appl. Catal. B, 2006, 63, 60. [65] H.M.Yates, M.G.Nolan, D.W.Sheel and M.E.Pemble: J. Photochem. Photobiol. A, 2006, 179, 213. [66] J.C.Tao, Y.Shen, F.Gu, J.Z.Zhu and J.C.Zhang: J. Mater. Sci. Technol., 2007, 23, 513. [67] B.F.Xin, Z.Y.Ren, P.Wang, J.Liu, L.Q.Jing and H.G.Fu: Appl. Surf. Sci., 2007, 253, 4390. [68] C.Ad´ an, A.Bahamonde, M.Fern´ andez-Garcia and A.Martinez-Arias: Appl. Catal. B, 2007, 72, 11. [69] S.Klosek and D.Raftery: J. Phys. Chem. B, 2001, 105, 2815. [70] J.F.Zhu, Z.G.Deng, F.Chen, J.L.Zhang, H.J.Chen, M.Anpo, J.Z.Huang and L.Z.Zhang: Appl. Catal. B, 2006, 62, 329. [71] R.C.W.Lam, M.K.H.Leung, D.Y.C.Leung, L.L.P.Vrijmoed, W.C.Yam and S.P.Ng: Sol. Energ. Mat. Sol. C, 2007, 91, 54. [72] N.Venkatachalam, M.Palanichamy, B.Arabindoo and V.Murugesan: J. Mol. Catal. A, 2007, 266, 158. [73] A.Di Paola, E.Garcia-L´ opez, S.Ikeda, G.Marci, B.Ohtani and L.Palmisano: Catal. Today, 2002, 75, 87. [74] V.Brezov´ a, A.Blaˇzkov´ a, L0 .Karpinsk´ y, J.Groˇskov´ a, ˇ B.Havlinov´ a, V.Jorik and M.Ceppan: J. Photochem. Photobiol. A, 1997, 109, 177. [75] P.Bouras, E.Stathatos and P.Lianos: Appl. Catal. B, 2007, 73, 51. [76] M.Anpo, M.Takeuchi, K.Ikeue and S.Dohshi: Curr. Opin. Solid State Mater. Sci., 2002, 6, 381. [77] M.Anpo and M.Takeuchi: J. Catal., 2003, 216, 505. [78] M.Anpo and M.Takeuchi: Int. J. Photoenergy, 2001, 3, 89. [79] K.Nagaveni, M.S.Hegde and G.Madras: J. Phys. Chem. B, 2004, 108, 20204. [80] M.Casarin, C.Maccato and A.Vittadini: Phys. Chem. Chem. Phys., 1999, 1, 3793. [81] T.Nishikawa, T.Nakajima and Y.Shinohara: J. Mol. Struct., 2001, 545, 67. [82] S.Karvinen, P.Hirva and T.A.Pakkanen: J. Mol. Struct., 2003, 626, 271. [83] M.S.Park, S.K.Kwon and B.I.Min: Phys. Rev. B, 2002, 65, 161201. [84] T.Umebayashi, T.Yamaki, H.Itoh and K.Asai: J. Phys. Chem. Solids, 2002, 63, 1909. [85] J.Premkumar: Chem. Mater., 2004, 16, 3980. [86] H.Irie, Y.Watanabe and K.Hashimoto: J. Phys. Chem. B, 2003, 107, 5483. [87] X.B.Chen and C.Burda: J. Phys. Chem. B, 2004, 108, 15446. [88] X.Chen, Y.B.Lou, A.C.S.Samia, C.Burda and J.L.Gole: Adv. Funct. Mater., 2005, 15, 41. [89] T.Tachikawa, M.Fujitsuka and T.Majima: J. Phys. Chem. C, 2007, 111, 5259. [90] S.K.Joung, T.Amemiya, M.Murabayashi and K.Itoh: Appl. Catal. A, 2006, 312, 20. [91] O.Diwald, T.L.Thompson, T.Zubkov, Ed.G.Goralski, S.D.Walck and J.T.Yates, Jr: J. Phys. Chem. B, 2004, 108, 6004. [92] D.M.Chen, D.Yang, Q.Wang and Z.Y.Jiang: Ind. Eng. Chem. Res., 2006, 45, 4110.

[93] W.Zhao, W.H.Ma, C.C.Chen, J.C.Zhao and Z.G.Shuai: J. Am. Chem. Soc., 2004, 126, 4782. [94] M.Shen, Z.Y.Wu, H.Huang, Y.K.Du, Z.G.Zou and P.Yang: Mater. Lett., 2006, 60, 693. [95] Y.Choi, T.Umebayashi and M.Yoshikawa: J. Mater. Sci., 2004, 39, 1837. [96] R.Hahn, A.Ghicov, J.Salonen, V.P.Lehto and P.Schmuki: Nanotechnology, 2007, 18, 105604. [97] T.Yamaki, T.Umebayashi, T.Sumita, S.Yamamoto, M.Maekawa, A.Kawasuso and H.Itoh: Nucl. Instrum. Methods Phys. Res. Sect. B, 2003, 206, 254. [98] S.N.Subbarao, Y.H.Yun, R.Kershaw, K.Dwight and A.Wold: Inorg. Chem., 1979, 18, 488. [99] W.Ho, J.C.Yu and S.Lee: Chem. Commun., 2006, 10, 1115. [100] D.Li, H.Haneda, N.K.Labhsetwar, S.Hishita and N.Ohashi: Chem. Phys. Lett., 2005, 401, 579. [101] J.C.Yu, J.G.Yu, W.K.Ho, Z.T.Jiang and L.Z.Zhang: Chem. Mater., 2002, 14, 3808. [102] L.Lin, W.Lin, Y.X.Zhu, B.Y.Zhao and Y.C.Xie: Chem. Lett., 2005, 34, 284. [103] Q.Shi, D.Yang, Z.Y.Jiang and J.Li: J. Mol. Catal. B, 2006, 43, 44. [104] T.Umebayashi, T.Yamaki, S.Tanaka and K.Asai: Chem. Lett., 2003, 32, 330. [105] T.Ohno, M.Akiyoshi, T.Umebayashi, K.Asai, T.Mitsui and M.Matsumura: Appl. Catal. A, 2004, 265, 115. [106] T.Umebayashi, T.Yamaki, S.Yamamoto, A.Miyashita, S.Tanaka T.Sumita and K.Asai: J. Appl. Phys., 2003, 93, 5156. [107] Q.W.Zhang, J.Wang, S.Yin, T.Sato and F.Saito: J. Am. Ceram. Soc., 2004, 87, 1161. [108] S.Matsushima, K.Takehara, H.Yamane, K.Yamada, H.Nakamura, M.Arai and K.Kobayashi: J. Phys. Chem. Solids, 2007, 68, 206. [109] G.Liu, Z.G.Chen, C.L.Dong, Y.N.Zhao, F.Li, G.Q.Lu and H.M.Cheng: J. Phys. Chem. B, 2006, 110, 20823. [110] M.C.Long, W.M.Cai, Z.P.Wang and G.Z.Liu: Chem. Phys. Lett., 2006, 420, 71. [111] X.T.Hong, Z.P.Wang, W.M.Cai, F.Lu, J.Zhang, Y.Z.Yang, N.Ma and Y.J.Liu: Chem. Mater., 2005, 17, 1548. [112] S.Sato: Chem. Phys. Lett., 1986, 123, 126. [113] T.Morikawa, R.Asahi, T.Ohwaki, K.Aoki and Y.Taga: Jpn. J. Appl. Phys., Part 2, 2001, 40, L561. [114] S.Sakthivel, M.Janczarek and H.Kisch: J. Phys. Chem. B, 2004, 108, 19384. [115] C.D.Valentin, G.Pacchioni, A.Selloni, S.Livraghi and E.Giamello: J. Phys. Chem. B, 2005, 109, 11414. [116] T.Lindgren, J.M.Mwabora, E.Avendano, J.Jonsson, A.Hoel, C.G.Granqvist and S.E.Lindquist: J. Phys. Chem. B, 2003, 107, 5709. [117] V.N.Kuznetsov and N.Serpone: J. Phys. Chem. B, 2006, 110, 25203. [118] N.Serpone: J. Phys. Chem. B, 2006, 110, 24287. [119] H.Wang and J.P.Lewis: J. Phys.: Condens. Matter., 2006, 18, 421. [120] G.S.Wu, T.Nishikawa, B.Ohtani and A.C.Chen: Chem. Mater., 2007, 19, 4530. [121] C.D.Valentin, G.Pacchioni and A.Selloni: Chem. Mater., 2005, 17, 6656. [122] J.Y.Lee, J.Park and J.H.Cho: Appl. Phys. Lett., 2005, 87, 011904. [123] H.Kamisaka, T.Adachi and K.Yamashita: J. Chem. Phys., 2005, 123, 084704. [124] J.G.Yu, J.C.Yu, B.Cheng, S.K.Hark and K.Iu: J. Solid State Chem., 2003, 174, 372. [125] H.Geng, S.W.Yin, X.Yang, Z.G.Shuai and B.G.Liu: J. Phys.: Condens. Matter., 2006, 18, 87. [126] J.C.Yu, L.Z.Zhang, Z.Zheng and J.C.Zhao: Chem. Mater. 2003, 15, 2280.

J. Mater. Sci. Technol., Vol.24 No.5, 2008

[127] L.K¨ or¨ osi and I.D´ek´ any: Colloids and Surf. A, 2006, 280, 146. [128] D.Li, H.Haneda, S.Hishita and N.Ohashi: Chem. Mater., 2005, 17, 2588. [129] D.Li, H.Haneda, S.Hishita and N.Ohashi: Chem. Mater., 2005, 17, 2596. [130] D.Li, N.Ohashi, S.Hishita, T.Kolodiazhnyi and H.Haneda: J. Solid State Chem., 2005, 178, 3293. [131] D.G.Huang, S.J.Liao, J.M.Liu, Z.Dang and L.Petrik: J. Photochem. Photobiol. A, 2006, 184, 282. [132] J.G.Yu, M.H.Zhou, B.Cheng and X.J.Zhao: J. Mol. Catal. A, 2006, 246, 176. [133] Y.Cong, F.Chen, J.L.Zhang and M.Anpo: Chem. Lett., 2006, 35, 800. [134] D.M.Chen, Z.Y.Jiang, J.Q.Geng, Q.Wang and D.Yang: Ind. Eng. Chem. Res., 2007, 46, 2741. [135] H.M.Luo, T.Takata, Y.G.Lee, J.F.Zhao, K.Domen and Y.S.Yan: Chem. Mater., 2004, 16, 846. [136] Y.Cong, J.L.Zhang, F.Chen, M.Anpo and D.N.He: J. Phys. Chem. C, 2007, 111, 10618. [137] N.Negishi, T.Iyoda, K.Hashimoto and A.Fujishima: Chem. Lett., 1995, 9, 841. [138] Y.Y.Xiong, T.Ma, L.H.Kong, J.F.Chen, X.Q.Wu, H.H.Yu and Z.X.Zhang: J. Mater. Sci. Technol., 2006, 22, 353. [139] M.Bockmeyer and P.L¨ obmann: Chem. Mater., 2006, 18, 4478. [140] M.Maeda and S.Yamasaki: Thin Solid Films, 2005, 483, 102. [141] N.P.Mellott, C.Durucan, C.G.Pantano and M.Guglielmi: Thin Solid Films, 2006, 502, 112. [142] J.G.Yu, X.J.Zhao and Q.N.Zhao: Thin Solid Films, 2000, 379, 7. [143] J.C.Yu, J.G.Yu and J.C.Zhao: Appl. Catal. B, 2002, 36, 31. [144] H.Nagayama, H.Honda and H.Kawahara: J. Electrochem. Soc., 1988, 135, 2013. [145] H.Kishimoto, K.Takahama, N.Hashimoto, Y.Aoi and S.Deki: J. Mater. Chem., 1998, 8, 2019. [146] X.P.Wang, Y.Yu, X.F.Hu and L.Gao: Thin Solid Films, 2000, 371, 148. [147] J.G.Yu, H.G.Yu, B.Cheng, X.J.Zhao, J.C.Yu and W.K.Ho: J. Phys. Chem. B, 2003, 107, 13871. [148] A.Dutschke, C.Diegelmann and P.Lobmann: J. Mater. Chem., 2003, 13, 1058. [149] S.Deki, S.Iizuka, M.Mizuhata and A.Kajinami: J. Electroanal. Chem., 2005, 584, 38. [150] J.G.Yu, H.G.Yu, C.H.Ao, S.C.Lee, J.C.Yu and W.K.Ho: Thin Solid Films, 2006, 496, 273. [151] D.Gutierrez-Tauste, X.Domenech, N.Casan-Pastor and J.A.Ayllon: J. Photochem. Photobiol. A, 2007, 187, 45. [152] M.K.Lee, J.J.Huang and C.F.Yen: J. Electrochem. Soc., 2007, 154, 117. [153] G.A.Battiston, R.Gerbasi, M.Porchia and A.Marigo: Thin Solid Films, 1994, 239, 186. [154] A.Borr´ as, C.L´ opez, V.Rico, F.Gracia, A.R.Gonz´ olezElipe, E.Richter, G.Battiston, R.Gerbasi, N.McSporran, G.Sauthier, E.Gy¨ ogy and A.Figueras: J. Phys. Chem. C, 2007, 111, 1801. [155] F.K.Lee, G.Andreatta.and J.J.Benattar: Appl. Phys. Lett., 2007, 90, 181928. [156] H.Y.Lee, Y.H.Park and K.H.Ko: Langmuir, 2000, 16, 7289. [157] B.C.Kang, S.B.Lee and J.H.Boo: Surf. Coat. Technol., 2000, 131, 88.

689

[158] D.J.Won, C.H.Wang, H.K.Jang and D.J.Choi: Appl. Phys. A, 2001, 73, 595. [159] C.K.Jung, B.C.Kang, H.Y.Chae, Y.S.Kim, M.K.Seo, S.K.Kim, S.B.Lee, J.H.Boo, Y.J.Moon and J.Y.Lee: J. Cryst. Growth, 2002, 235, 450. [160] M.Kang, J.H.Lee, S.H.Lee, C.H.Chung, K.J.Yoon, K.Ogino, S.Miyata and S.J.Choung: J. Mol. Catal. A, 2003, 193, 273. [161] G.A.Battiston, R.Gerbasi, A.Gregori, M.Porchia, S.Cattarin and G.A.Rizzi: Thin Solid Films, 2000, 371, 126. [162] J.M.Mwabora, T.Lindgren, E.Avenda˜ no, T.F.Jaramillo, J.Lu, S.E.Lindquist and C.G.Granqvist: J. Phys. Chem. B, 2004, 108, 20193. [163] L.J.Meng, V.Teixeira, H.N.Cui, F.Placido, Z.Xu and M.P.dos Santos: Appl. Surf. Sci., 2006, 250, 7970. [164] L.J.Meng, M.Andritschky and M.P. dos Santos: Appl. Surf. Sci., 1993, 65/66, 235. [165] L.J.Meng, M.Andritschky and M.P. dos Santos: Thin Solid Film, 1993, 223, 242. [166] L.J.Meng and M.P. dos Santos: Thin Solid Film, 1993, 226, 22. [167] Y.Nakano, T.Morikawa, T.Ohwaki and Y.Taga: Appl. Phys. Lett., 2005, 86, 132104. [168] D.Guerin and S.Ismat Shah: J. Vac. Sci. Technol. A, 1997, 15, 712. [169] S.Z.Chen, P.Y.Zhang, D.M.Zhuang and W.P.Zhu: Catal. Commun., 2004, 5, 677. [170] C.J.Tavares, J.Vieira, L.Rebouta, G.Hungerford, P.Coutinho, V.Teixeira, J.O.Carneiro and A.J.Fernandes: Mater. Sci. Eng., B, 2007, 138, 139. [171] M.Kitano, K.Funatsu, M.Matsuoka, M.Ueshima and M.Anpo: J. Phys. Chem. B, 2006, 110, 25266. [172] J.O.Carneiro, V.Teixeira, A.Portinha, A.Magalh˜ aes, P.Coutinho, C.J.Tavares and R.Newton: Mater. Sci. Eng. B, 2007, 138, 144. [173] Y.Cui, H.Du, S.W.Lee, C.Sun and L.S.Wen: Mater. Sci. Forum, 2007, 544-545, 27. [174] Y.Cui, H.Du, J.Q.Xiao and L.S.Wen: J. Mater. Sci. Technol., 2008, 24, 172. [175] L.Sirghi, T.Aoki and Y.Hatanaka: Thin Solid Films, 2002, 422, 55. [176] M.Yamagishi, S.Kuriki, P.K.Song and Y.Shigesata: Thin Solid Films, 2003, 442, 227. [177] L.S.Wen and R.F.Huang: Surf. Coat. Technol., 2005, 193, 1. [178] M.Zhang, G.Q.Lin, C.Dong and L.S.Wen: Surf. Coat. Technol., 2007, 201, 7252. [179] J.T.Chang, C.W.Su and J.L.He: Surf. Coat. Technol., 2006, 200, 3027. [180] J.T.Chang, Y.F.Lai and J.L.He: Surf. Coat. Technol., 2005, 200, 1640. [181] H.Yumoto, S.Matsudo and K.Akashi: Vacuum, 2002, 65, 509. [182] A.Bendavid, P.J.Martin and H.Takikawa: Thin Solid Films, 2000, 360, 241. [183] P.J.Martin and A.Bendavid: Thin Solid Films, 2001, 394, 1. [184] M.D.Huang, Y.P.Lee, C.Dong, G.Q.Lin, C.Sun and L.S.Wen: J. Vac. Sci. Technol. A, 2004, 22, 250. [185] B.K.Tay, Z.W.Zhao and D.H.C.Chua: Mater. Sci. Eng. R, 2006, 52, 1. [186] Z.W.Zhao and B.K.Tay: J. Appl. Phys., 2007, 101, 013505.