Environ. Sci. Technol. 2004, 38, 285-289
Improvement of Catalyst Durability by Deposition of Rh on TiO2 in Photooxidation of Aromatic Compounds HISAHIRO EINAGA,* TAKASHI IBUSUKI, AND SHIGERU FUTAMURA National Institute of Advanced Industrial Science and Technology, AIST Tsukuba West, 16-1, Onogawa, Tsukuba, Ibaraki 305-8569, Japan
Gas-solid heterogeneous photocatalytic oxidation of aromatic compounds in air was carried out at room temperature with a fixed bed flow reactor. The deposition of Rh on TiO2 catalyst improved the catalyst durability in benzene photooxidation. The Rh deposition reduced the amounts of carbonaceous materials on the catalyst surface that were the cause of catalyst deactivation. The highest reaction rate was obtained at the Rh loading of 0.51.0 wt %. The Rh/TiO2 catalyst was gradually deactivated in prolonged benzene photooxidation, due to the increasing amount of carbonaceous materials on the catalyst surface. XPS studies showed that the gradual deactivation was related to the changes in the oxidation state of the surface Rh metals. The catalysts were regenerated by the treatment of hydrogen reduction after the photoirradiation in humidified air, indicating that Rh0 was essential for the improvement of the catalyst durability. The catalyst was also regenerated by the heat treatment in N2 flow instead of the hydrogen reduction. Rh deposition was also effective for the improvement of catalyst durability in toluene photooxidation.
Introduction It is known that volatile organic compounds (VOCs) are triggering serious environmental problems such as stratospheric ozone depletion and tropospheric ozone increase, depending on their chemical structures. They should be removed from exhaust gases emitted from various moving and stationary sources. Hitherto, photocatalytic oxidation systems using TiO2 catalyst have been extensively studied for VOC removal (1). These systems are promising as pollution control technologies since they can efficiently decompose low concentrations of VOCs under ambient conditions. In addition, additive fuels are unnecessary for the systems, in marked contrast with incineration and catalytic oxidation. It has been reported that many kinds of organic compounds are decomposed to CO2 on photoirradiated TiO2 surfaces (2-8). Catalyst deactivation is observed in some cases of VOC photooxidation, depending on the structure and concentration of the substrate (9-13). In many cases, deactivation has been ascribed to the accumulation of lessreactive intermediates and byproducts on TiO2 surfaces (14). * Corresponding author phone: +81-29-861-8679; fax: +81-29861-8266; e-mail:
[email protected]. 10.1021/es034336v CCC: $27.50 Published on Web 11/14/2003
2004 American Chemical Society
We have investigated the photocatalytic oxidation of several types of hydrocarbons over TiO2 catalysts and have reported that TiO2 catalyst is significantly deactivated in the decomposition of aromatic compounds (15). The formation and decomposition behavior of the intermediates and byproducts on TiO2 surfaces affects the rates for the photooxidation of hydrocarbons. Regeneration of used catalysts is also indispensable for the practical application of photocatalytic oxidation technologies. In the previous paper, we described that the deactivated TiO2 catalysts used for the benzene photooxidation were regenerated by a photochemical process (15). The carbonaceous materials formed on the catalyst surface were decomposed to COx by photoirradiation in a humidified air. To practically utilize the photocatalytic oxidation technologies for purification of air polluted with relatively high concentrations of aromatic compounds, it is necessary to improve the durability of TiO2. In this paper, we report that the durability of TiO2 catalyst in the photooxidation of aromatic compounds is greatly improved by deposition of Rh on TiO2. The formation and decomposition behavior of carbonaceous materials on the photoirradiated catalysts is compared with each other, and our findings show that the Rh deposit on TiO2 retards their accumulation on the catalyst surface. The Rh/TiO2 catalyst is gradually deactivated in prolonged photooxidation. We also describe the cause of deactivation and the regeneration process for the Rh/TiO2 catalyst.
Experimental Section Commercially available TiO2 P25 (BET surface area: 43 m2 g-1) was used as the catalyst and the precursor. Rh was photochemically deposited on TiO2 surfaces. For the preparation of 0.5 wt % Rh/TiO2, RhCl3‚3H2O (0.0243 mmol; 99.5%, minimum assay; Wako Pure Chemical Ind., Ltd.) was dissolved into an ethanol (150 mL)-water (100 mL) solution in a Pyrex vessel. TiO2 powder (0.5 g) was then dispersed into the solution with vigorously stirring, and N2 was bubbled into the suspension for 20 min to remove the dissolved O2. The suspension was photoirradiated with a 500-W highpressure Hg lamp for 3 h. After the irradiation, the Rh/TiO2 powder was filtered off, washed with water, and dried at 383 K. It was confirmed by the analysis of filtrate on ICP-MS (VG Elemental Plasma Quad 2 Plus) that more than 99% of the Rh had been removed from the ethanol-water solution. The TEM images showed that the sizes of Rh particles on the catalyst were in the range of 3-10 nm in diameter. Separately, Rh was deposited on TiO2 on addition of NaBH4 to the aqueous suspension containing TiO2 and RhCl3‚3H2O instead of the photoirradiation. Rh/TiO2 was also prepared by the impregnation of TiO2 with RhCl3‚3H2O followed by the heat treatment at 673 K at first in an O2 stream for 3 h and then in H2 for 3 h. Pd/TiO2 and Pt/TiO2 catalysts were prepared by the photodeposition method using Pd(CH3COO)2 and H2PtCl6‚3H2O as the precursors, respectively. The photocatalytic reactor used in this study was a single pass type, and the details were described elsewhere (16). The reactor was composed of an inner rod (outer diameter: 8 mmΦ; length: 500 mm) and an outer tube (inner diameter: 13 mmΦ), which were fabricated from Pyrex glass. The volume of the vessel was 42 mL. The catalyst was coated onto the inner rod from the aqueous slurry, and the rod was then dried at 383 K. Coating was repeated several times until the weight of the catalyst became 0.24 g ((0.005 g). The reactor was photoirradiated with four 20 W black lights (TOSHIBA FL20S‚BLB-A). Reaction gases were prepared from benzene in N2, N2 (>99.99%, total hydrocarbons < 1 ppm), VOL. 38, NO. 1, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
285
FIGURE 2. Time course plots for the COx formation when the used catalysts were photoirradiated in humidified air. (b), TiO2; (O), Rh/TiO2; water vapor, 1.7% in air; flow rate 100 mL min-1; catalyst, 0.24 g.
FIGURE 1. Time course plots for (a) the conversion and (b) the COx selectivity in benzene photooxidation with TiO2 and 0.5 wt % Rh/TiO2 catalysts. (0, 9), COx selectivity; (O), carbon balance; benzene, 250 ppm; water vapor, 1.7%; reaction temp, 30 °C; flow rate 100 mL min-1; catalyst, 0.24 g. and O2 (>99.99%, total hydrocarbons < 1 ppm). Reaction gases were humidified with a water-bubbling apparatus. The concentration of water vapor was 1.7%. In a typical run, the concentration of benzene was 250 ppm and the gas flow rate was set at 100 mL/min, where the residence time of the reaction gas in the reactor was 25 s. Unless otherwise stated, freshly prepared catalysts were used for the reactions. As the pretreatment, the catalyst was photoirradiated in a humidified air. The reaction was started after the adsorptiondesorption equilibrium was achieved between the gas phase and the catalyst surface in the photoreactor. The reaction temperature was kept at 303 K. The benzene concentration in the reaction gas was determined on a GL science GC390B gas chromatograph with an FID. CO2 and CO were separately analyzed by using a GL science GC390B gas chromatograph with a TCD, an FID, and a methane converter. XPS measurements were performed on an ULVAC-PHI Quantum 2000 spectrometer. Powdered samples were loaded on a carbon tape. Monochromatic AlKR radiation (hν ) 1486.6 eV) operating at 107.2 W was used for sample excitation. Spectra were recorded at a constant pass energy of 23.5 eV. The region between 300 and 320 eV contains information on the Rh 3d5/2 and 3d3/2 electron excitations. A flood gun was used to counteract the charging of the catalyst sample.
Results and Discussion Catalytic Properties of Rh/TiO2. Figure 1(a) shows typical time courses for benzene photooxidation over TiO2 and Rh/ TiO2 catalysts in a humidified air. We have reported that water vapor facilitates the decomposition of carbonaceous materials on TiO2 (15, 16) and inhibits the TiO2 deactivation in benzene photooxidation. With pure TiO2, benzene conversion immediately decreases with time and levels off at 13% after 3 h. When the 0.5 wt % Rh/TiO2 catalyst is used for the benzene photooxidation instead of TiO2, the decrement of benzene conversion becomes much smaller, although the conversion for the first 5 min is comparable with 286
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 1, 2004
that with the TiO2 catalyst. After 10 h, the conversion is almost constant and is about 3 times higher than that with the TiO2 catalyst. Here, the total amount of benzene reacted during the photooxidation is estimated to be 3.0 × 10-4 mol for the Rh/TiO2 catalyst. The turnover number for the benzene decomposition is estimated to be 2.5 molecules per a Ti site by taking the total amount of benzene reacted and normalized by the Ti density of the single-crystal surfaces of anatase (001) plane (7.00 atoms nm-2) (17), which is atomically the densest plane and is usually assumed to terminate the crystallites predominantly (18). Figure 1(b) shows time courses for the formation of CO2 and CO in the benzene photooxidation with Rh/TiO2. The carbon balance is 88-90% for the first 3 h due to the formation of carbonaceous materials on the catalyst surface, as previously reported for TiO2 (16). However, it increases up to 97-103% after about 4 h. The mole fractions of CO2 and CO are 93 and 7%, respectively. They are almost unchanged with time and are similar to those with TiO2 in the previous studies (16, 19). We have reported that benzene is completely decomposed to CO2 over Pt/TiO2 catalysts (19). The Pt/TiO2 catalysts oxidize the byproduct CO formed in the benzene decomposition. However, Rh/TiO2 is less reactive in CO photooxidation than Pt/TiO2. The rate for CO photooxidation with 1.0 wt % Rh/TiO2 [0.35 × 10-6 mol/g min (reaction conditions: CO 100 ppm in air, water vapor 2.0%)] is 20 times lower than for 1.0 wt % Pt/TiO2 [6.6 × 10-6 mol/g min (19)], although it is higher than for TiO2 (
