All reagents were of analytical grade and were used without further purification. Orange II and rhodamine B used in this study were purchased from Nacalai Tesque. Ascorbic acid (AA), ammonium oxalate (AO) and t-butyl alcohol (TBA) were obtained from Wako Pure Chemicals. P-25 TiO₂ was purchased from Degussa Co. Ltd. Ultrapure water (18 MΩ cm) was prepared by an ultrapure water system (Advantec MFS Inc.).

The photocatalytic activities of TiO₂ were evaluated by the degradation of OII and RhB under visible light irradiation at ambient temperature. Typically, 30 mL of dye solution and 20 mg of photocatalyst were added to a 35 mL Pyrex glass cell. The initial concentration of dye in all experiments was 5 mg/L and the appropriate quantity of the photocatalyst powder was magnetically stirred before and during irradiation. Before irradiation, the photocatalyst suspension containing dye was allowed to equilibrate for 30 min in the dark. The sample solution was irradiated with a LED lamp (TOSHIBA LDA14L-G/100W) in conjunction with a UV cut filter (Y-44, HOYA), which was positioned on the side of the reaction cell. The luminous intensity was measured by a UV radio meter (UD- 400, TOPCON TECHNOHOUSE Co., Japan). The light intensity of the LED lamp after the filter was 5.3 mW/cm². After the desired irradiation time, TiO₂ was separated through the 0.45 μm Advantec membrane filter. The TiO₂ powder could be almost removed by the filtration. The absorbance of the remnant dye was measured using a UV-visible spectrometry (UV-1650PC, SHIMADZU Co., Tokyo, Japan). The relative concentration (C/C₀) of the OII solution was calculated by the relative absorbance (A/A₀) at 485 nm according to Beer-Lambert law. A₀ and A are the absorbance of the OII solution at the beginning time (t₀) of visible light irradiation and at time t, respectively. C₀ and C are the concentrations of OII at the beginning of visible light irradiation and at time t, respectively. The photodegradation of RhB (5 mg/L) was similar to that of OII except that the detection wavelength was 554 nm.

Radical scavenger studies were carried out to investigate the active species involved in the photodegradation of dye. The scavenging experiments of reactive oxygen species were similar to the photodegradation experiments. Three scavengers were selected, namely, t-butyl alcohol (^•OH radical scavenger), di-ammonium oxalate monohydrate (hole scavenger) and ascorbic acid (radical scavenger). Different quantity of t-butyl alcohol, di-ammonium oxalate monohydrate [9] and ascorbic acid [10] were added into the dye solution prior to addition of catalysts.

To determine the possible degradation mechanism of Orange II by TiO₂, different scavengers were introduced to quench the relevant active species. In this study, t-butyl alcohol (TBA), di-ammonium oxalate monohydrate (AO) and ascorbic acid (AA) were adopted to be the scavengers of hydroxyl radicals (^•OH), superoxide radical () and holes (h⁺), respectively. As shown in Figure 1, the photocatalytic degradation efficiency of OII (5 mg/L) with TiO₂ was about 100% after 6 h under visible light irradiation. The photodegradation of OII over the TiO₂ was affected slightly by the addition of TBA, demonstrating that ^•OH active species played a small role in the photocatalytic degradation of OII. However, the photocatalytic degradation efficiency of OII decreases significantly in the presence of AA, which indicates that is an important active species in the process of OII degradation. In addition, the photocatalytic activity of the TiO₂ was completely suppressed by AO, suggesting that h⁺ can be also involved in the process of OII degradation.

In order to investigate the active species involved in photodegrading RhB, scavenger studies were also carried out on TiO₂. As shown in Figure 2, photocatalytic degradation of RhB over TiO₂ was retarded with the presence of di-ammo- nium oxalate monohydrate (AO) and ascorbic acid (AA). The results strongly indicated that h⁺, and were the active species involved in the photodegradation of RhB. However, was more dominant in photodegrading RhB, with h⁺ as the most important active species. As aforementioned, the enhanced photocatalytic degradation of MB was contributed by the photosensitizing of RhB toward TiO₂. The electron ejected from HOMO to conduction band of TiO₂ could have been utilized for the reduction of surface adsorbed oxygen to produced [11] . The photocatalytic degradation of RhB was retarded most significantly with the presence of AO, conveying that oxidation reaction occurred

mainly via photogenerated holes, not via hydroxyl radical [12] . Therefore, the presence of holes scavenger has decreased the most the photocatalytic degradation of RhB. The presence of TBA had little effect on the decolorization rate, indicating that RhB was almost not degraded by ^•OH.

Figure 3(a) demonstrates the valence band (VB) and conduction band (CB) levels and the band gap energy of orange II and rhodamine B and TiO₂ vs NHE reference electrodes. The energy bands of rhodamine B, orange II and TiO₂

are 2.37 [13] , 2.03 [14] and 3.2 eV [13] , respectively. The band gap energies of orange II and rhodamine B are narrow enough to absorb visible light. Otherwise, as the more negative potential of OII and RhB lowest unoccupied molecular orbital (LUMO) level than the conduction band (CB) of TiO₂ [13] [14] , the electron transfer from the LUMO of dyes to the CB of TiO₂ is feasible. It is reported the redox potential of is ‒0.33 V vs NHE [15] , which is less negative than conduction band potential of TiO₂ (‒0.5 V vs NHE) [16] .

Under visible light irradiation, a dye sensitized mechanism has been depicted in Figure 3(b). Upon irradiation of visible light, a dye absorbs the light to create an electron and hole in the conduction and valence bands (LUMO and HOMO) of the dye [17] . The electron in the LUMO then transfers to the CB of TiO₂. The adsorbed molecular oxygen on the catalyst captures electron from the CB of TiO₂ to form. The oxidant radical reacts with adsorbed dye to degrade it. The holes in the HOMO react with adsorbed OH⁻ species to form ^•OH radical. However, the formation channel to ^•OH is minor under visible right [17] , which is similar with the result shown in Figure 1 and Figure 2. According to the results of the radical scavengers, OII and RhB were attacked by the super oxides, and holes, h⁺ for the degradation of dye. A possible degradation reaction mechanism is described below.