2.1 Preparation of In(OH)₃

In(OH)₃ was prepared by hydrolysis of In(NO₃)₃ (0.05 M) in an aqueous solution of ammonium (12%) at room temperature. When the solution pH reached about 8, the white precipitate was collected, and washed thoroughly with distilled water till the total concentration of ionic species was lower than 10 ppm. The resulting slurry was then dispersed into water under ultrasound radiation until a transparent sol was formed. Finally, the sol was dried at 60 °C, followed by thermal treatment at 120 °C for 16 h.

2.2 Activity test

The photocatalytic oxidation experiment of benzene was conducted in a tubular quartz microreactor operating in a continuous-flow mode. The reactor was surrounded by four 8-W fluorescent UV bulbs (TUV 8W/G4 T5, Philips, wavelength 254 nm). The catalyst loading was 0.5 g (50–70 mesh). Unless otherwise mentioned, the reaction temperature was ca. 120 °C, which results from the light source without controlling. Certain amount of saturated benzene vapor (at 273.15 K) was introduced into the reactor along with a continuous O₂ gas stream. Prior to irradiation, adsorption of benzene on the photocatalyst reached equilibrium. The concentrations of benzene and product CO₂ were analyzed by an on-line gas chromatograph (HP 6890), equipped with a thermal conductivity detector, a flame ionization detector, and a Porapak R column.

2.3 Characterization

X-ray diffraction patterns (XRD) were collected using a Bruker D8 Advance X-ray diffractometer (Cu Kα, 1.5418 Å). UV–vis diffuse reflectance spectra (DRS) were recorded on a UV–vis–NIR spectrometer (Varian Cary 500 Scan, BaSO₄ as a background reference). Infrared spectra (IR) were recorded on a Nicolet 670 FTIR spectrophotometer. TEM image of the sample was recorded on a JEOL model JEM 2010 EX instrument at an acceleration voltage of 200 kV. Nitrogen adsorption and desorption isotherms of the samples were measured on a Coulter Omnisorp 100 CX system. X-ray photoelectron spectrum (XPS) analysis was performed with a Quantum 2000 spectrometer (American physical electronic Co.) using Al Kα X-ray beam (1486.6 eV). All the samples were degassed in vacuum until the pressure is lower than 10⁻⁸ Torr. Binding energies were calibrated for charging effect in reference to the C_1s peak, which was assumed to have a binding energy of 284.8 eV.

3.1 Photocatalytic properties of In(OH)₃

Fig. 1 shows photodegradation of benzene over In(OH)₃ and P25 TiO₂ under 254 nm UV light irradiation. For P25 TiO₂, the initial conversion of benzene is 9.5%, and the amount of product CO₂ is 180 ppm, corresponding to 26% of mineralization of benzene. However, during benzene photooxidation, it suffers from rapid deactivation so that the conversion of benzene decreases from 9.5% to 0% after 300 min, and the color of the catalyst changes from white to brown. The quick deactivation is related to the large amount of the carbonaceous deposits, as discussed later. The deactivated TiO₂ cannot be regenerated even it is calcined under oxygen atmosphere at 200 °C for 5 h. In contrast to P25 TiO₂, In(OH)₃ presents a much higher steady-state photodegradation activity (from 200 to 800 min) for benzene under the same conditions. After 300 min of the photoreaction, the conversion of benzene and the mineralization rate over In(OH)₃ achieve the maximum, 33.4% and 56%, respectively, while the P25 TiO₂ is deactivated completely. For In(OH)₃, the gradual increasing conversion rate of benzene in the range 0–300 min may be resulting from the increasing temperature of the reaction system under UV light irradiation. After 800 min of reaction, In(OH)₃ also suffers from a gradual deactivation, but after 30 h of reaction it remains 25% of conversion and 40% of mineralization rate of benzene. These results show that In(OH)₃ has not only higher photocatalytic activity and stronger photocatalytic mineralization power for benzene, but also better durability of photoactivity.

Two control experiments, either without the In(OH)₃ photocatalyst or without UV light, were performed to prove the need for photocatalyst and UV light during the photodegradation of benzene. As shown in Fig. 1(c) and (d), negligible photoactivity for benzene is presented, showing that benzene cannot be photodecomposed by UV light, or by the photocatalyst alone. It was concluded that the concurrence of photocatalyst and UV light is necessary for the photodegradation of benzene.

Activity stability of the In(OH)₃ can be further ameliorated by adding water vapor to the reactant gas. As shown in Fig. 2(B), the benzene photooxidation reaction reaches a steady state after 6 h in the presence of water vapor, and no notable deactivation is observed during the photoreaction.

3.2 Spectral characterizations of the catalyst before and after benzene photooxidation

As shown in Fig. 3, as-synthesized In(OH)₃ is a white solid with cubic crystallites. Some colorful by-products are formed from benzene oxidation on the catalyst, as evidenced by UV–vis DRS (Fig. 4). The reflectance percentage is decreased for both the catalysts after photoreaction, which is due to the deposits of by-products. Based on the magnitude of such decreased reflectance, the relative amount of the by-products on TiO₂ is more than that on In(OH)₃ after the reaction. It is also evidenced by the FTIR spectra in Fig. 5. The absorption in the range 1200–1800 cm⁻¹ appears obviously on TiO₂. The band at 1483 cm⁻¹ and in the region of 1260–1410 cm⁻¹ could be assigned to be the signals of benzene derivatives and the corresponding aliphatic C–H bending vibration. Two weak bands shown in spectrum (A) at 1688 and 1711 cm⁻¹ indicate the formation of CO bonding, implying the occurrence of the attack by the activated oxygen species or hydroxyl radical (OH) upon the carbon structures [12,13]. However, such change is not observed for In(OH)₃. Spectrum (B) presents some bands related to the absorbed water and hydroxyl groups [14] in the range of 1300–1800 cm⁻¹.

The XPS spectrum (Fig. 6) and the XRD plot (Fig. 7) show that no obvious change in the oxidation state of In³⁺ (In3d_5/2, 444.7 eV; In3d_3/2, 452.2 eV [15]) and the crystal phases of the In(OH)₃ sample occurs after the photoreaction, indicating that the IR and DRS spectral changes observed above are exactly due to carbonaceous by-products formed from benzene oxidation.

A number of studies showed that indium hydroxide is a typical semiconductor with 5.15 eV of wide band gap [16,17], and hence it can be excited exclusively by 254-nm UV light irradiation. Several factors can contribute to the observed difference in photoactivity between the two catalysts. First of all, the catalysts have different spectral properties, which would affect photon harvesting and thus the photocatalytic efficiency for benzene oxidation. When the reaction is performed with UV-light cut-off at 365 nm, no photooxidation of benzene is observed with In(OH)₃, but the reaction is still efficient with TiO₂. Secondly, direct hole oxidation and hydroxyl radical (OH) attack have been proposed as the reaction pathways for gas-phase photodegradation of benzene over TiO₂ [4]. Einaga and co-workers have demonstrated that during benzene photooxidation the OH⁻ groups on TiO₂ surface are consumed, which can be regenerated by water vapor for sustainment of the catalyst activity [8]. In this respect, the excess OH⁻ groups preserved by In(OH)₃ would be beneficial for benzene photooxidation. Thirdly, the fate of both photogenerated holes and electrons are vital, once the semiconductor is excited by UV light. The reactivity of such photogenerated carriers would be different from one catalyst to another toward surface species, such as benzene, O₂, and water. Since In(OH)₃ has a larger band gap than that of TiO₂, the photogenerated holes may have more positive potentials than those of TiO₂ for direct hole oxidation of benzene or for generation of OH radicals to attack aromatic rings. Moreover, In(OH)₃ has a higher surface area than TiO₂ (the values of the surface areas are 123.8 m² g⁻¹ and 50 m² g⁻¹, respectively). This serves as another reason that In(OH)₃ is superior to TiO₂ as a photocatalyst for benzene degradation. The detailed mechanism is under further investigation.