Investigation of nano-titanium oxide photocatalysts for the decomposition of nitrogen oxide in the flow system

Zhang Jinlong, He Bin, Masakazu Anpo^#
(Institute of Fine Chemicals, East China University of Science and Technology, Meilong Road 130, Shanghai 200237, China; ^# Department of Applied Chemistry, Osaka Prefecture University, Gakuen-cho, Sakai, Osaka 599-8531, Japan)

Received Dec. 10, 2002; Supported financially by the National Natural Science Foundation of China (No.29476231), Foundation of Ministry of Education P.R.C., Shuguang Plan of Commission of Education of Shanghai, and Nano-materials Special Item of Commission of Shanghai Science and Technology.


Abstract In order to investigate the mechanism of the decline phenomena of nano TiO₂ photocatalysts in the photocatalytic decomposition of NO in a large scale flow reaction system, the TPD (temperature-programmed desorption) spectra adsorbed species at the photocatalysts were investigated in detail before and after the photoreactions. It was found that the TPD peak assigned to the adsorbed NO species observed at around 524 K is directly associated with the decline in the photocatalytic reactivity of TiO₂. The TPD spectrum of the adsorbed NO species consisted of three different types of species. The g-adsorbed species of NO formed under UV irradiation on the active surface sites cause a remarkable decline in the photocatalytic reactivity of TiO₂. It was also found that such a decline caused by the strong adsorption of NO at the surface sites can be recovered to its original reactivity by heat treatment in atmosphere Ar and/or O₂ at around 573 K.
Keywords photocatalysis; titanium oxide; decomposition of NO; TPD spectra


1. INTRODUCTION
A great deal of attention has been aimed at the investigation of nano-photocatalysts which could be applied for improvement of human life environment. Although this field has expanded rapidly with regard to light energy, especially the conversion of solar energy into useful chemical energy using solid photocatalysts such as TiO₂,^[1-4] more recently, the application of photocatalysis to reduce toxic agents in the atmosphere and water is being actively investigated. Nitrogen oxide(NOx) is an especially harmful atmospheric pollutant which causes acid rain and photochemical smog. The direct decomposition of NOx into N₂ and O₂, has been a great challenge for many researchers.^[4-11] TiO₂ is especially attractive due to its various properties as a non-toxic, highly stable, low cost, and high activity, significantly, for chemical waste remediation. ^[12-16]
Up to date, there have been a few reports concerning the origin of the often observed decline in the photocatalytic reactivity of TiO₂ photocatalysts in both aqueous and gas phase reaction systems.^[16-24] The design and development of titanium oxide photocatalysts for use in a large scale will only be possible by investigating and understanding the various aspects of the reactivity including the origin and recovery of this decline. In the present study, we have carried out observations of the decline in the photocatalytic reactivity of titanium oxide photocatalysts for the decomposition of NO under a large scale flow system and investigated the origin by means of a temperature-programmed desorption (TPD) method. Moreover, we have investigated that how the lost photocatalytic reactivity can be recovered to its original reactivity level.

2. EXPERIMENTAL
2.1 Photocatalytic flow reaction system
The photocatalytic reactions were carried out in the flow system described in the preceding paper.^[25] The flow rates of the reacting gases (NO, O₂) and pretreatment gases (Ar, O₂) were adjusted by a mass flow controller. An electronic furnace was used to pretreat the catalysts under an O₂ and Ar flow at suitable temperatures. During the photoreactions, to keep the Hg-lamp cool and to absorb the infrared beams emitted, water was cycled in cooling pipes made of quartz glass.
    A NOx meter (New Cosmos Electric Motors, Ltd.) and gas chromatography were used to measure the NO and product (N₂, O₂ and N₂O) concentrations. The reaction cell was made of quartz (length: 190 mm; inner diameter: 10 mm). 150 mg of the TiO₂ photocatalyst was introduced into the cell. To prevent any movement of the catalyst in the cell, glass wool was placed at both the bottom and top sides of the reaction cell. The NOx meter and gas chromatographer were connected to a recorder which continuously measured changes in the concentrations of NO and the products of N₂, O₂ and N₂O.
2.2 Catalysts
Five different types of nano powdered TiO₂ catalysts (grain size, 0.02-0.1 mm) supplied by the Catalysis Society of Japan as standard reference TiO₂ (JRC-TIO-1, 2, 3, 4, 5) were used as the photocatalysts.^[25] Detailed information on these standard reference TiO₂ catalysts is available from the Catalysis Society of Japan.
2.3 Photocatalytic Decomposition of NO
Photocatalytic reactions were carried out under a flow system.^[25] The flow reactant gases and the pretreatment gases were introduced from the bottom side of the catalyst layers. 150 mg of the TiO₂ photocatalyst was loaded into the reaction cell which was then connected to the flow system with silicone grease. The TiO₂ photocatalyst was heated in O₂ using an electronic furnace, and then treated with Ar flow gas at 20 cc/min. When the temperature of the photocatalyst reached the suitable temperature, O₂ and Ar flow gases were introduced into the cell for 2 h at 20 cc/min, respectively. The system was then cooled down to the reaction temperature, the O₂ flow was discontinued although the Ar flow was continued further for 1 h at a flow rate of 20 cc/min. After these pretreatment proceduces, the reaction cell was cooled down to room temperature under an Ar gas flow.
    In the photocatalytic reaction, the concentration of the NO reactant gas (mixture of NO and Helium) was 10 ppm, and the flow rate of the reactant gas was 100 cc/min. After reaching a steady state, UV irradiation was started and continued for 2 h. A Toshiba SHL-100UV high-pressure Hg lamp was used. The distance of the lamp from the reaction cell was 10 cm and a color filter of UV-27 (l> 270 nm) was used. The photon density of the Hg lamp passing through the UV-27 filter was 10790 Lux, measured by a Digital Lux meter (model LX-1332).
2.4 TPD measurements
After the photocatalytic decomposition reaction of NO in the hermetic system, the TPD of the TiO₂ photocatalyst was measured. 10 mg of TiO₂ was loaded into the TPD cell. Before each run, the TiO₂ photocatalyst was heated to 573 K at 5 K/min in 3 kPa (23 Torr) of O₂ in order to create an original oxide surface, and then evacuated for 2 h. After degassing the sample at 573 K for 1 h, the TiO₂ sample was cooled to room temperature. Then after 1 kPa (7 Torr) of NO was introduced into the TPD cell, the TiO₂ sample was irradiated for 2 h by UV light. The TiO₂ was degassed in the system for 15 min prior to TPD measurements. Temperature-programmed desorption was measured by a M-QA100TS mass analysis meter equipped with a PC-9821 computer. During TPD measurements, the TiO₂ catalyst temperature was increased at a constant rate of 5K/min controlled by a Digital Programm Temperature Adjuster FP21.

3. RESULTS AND DISCUSSION
3. 1 The origin of the decline of the photocatalytic activity
Standard reference TiO₂ photocatalysts JRC-TIO-1-5 were used in the flow system to study the origin of the decline in efficiency during long time utilization. Figure 1 shows that JRC-TIO-4 gradually loses its photocatalytic reactivity during photoreactions carried out on a large scale and for a long period. After 2 h, the conversion of NO for each catalyst levels off and becomes 1/4-1/5 of the reactivity at the beginning of the reaction. Thus, the decline in the photocatalytic reactivity of TiO₂ for the decomposition of NO in the absence of O₂ and/or H₂O could be confirmed. The reaction products of the decomposition of NO in the flow system were also confirmed to be N₂, O₂ and N₂O, as with the closed reaction system.


Fig.1 Reaction time profiles of the photocatalytic decomposition of NO at room temperature on the TiO₂ photocatalyst pretreated with O₂ at 573 K.
Pretreatment: after introduction of O₂ (20 cc/min) and Ar (20 cc/min) at 573 K, heated in Ar flow (20 cc/min) at 295 K (●), 373 K (■), 473K (▲) and 573 K (▽), respectively. Gas component: 10 ppm NO, 100 cc/min. Catalyst: JRC-TIO-4, 150 mg.

    The adsorbed compounds cause the photocatalytic deactivation in the NO decomposition reaction.^[26] It was found that such a decline in reactivity after the reaction of NO under a large scale flow system could be recovered to its original level by applying heat treatment to the used catalysts under an O₂ or Ar flow at around 573 K for one h, as shown in Fig. 2.


Fig. 2 NO conversion of the photocatalytic decomposition of NO on the TiO₂ photocatalyst recovered by treatment under a flow of Ar or O₂ for 1h at different temperatures. Pretreatment: after introduction of O₂ (20cc/min) and Ar (20cc/min) at 573 K, heated in Ar flow (20cc/min) at 573 K. Gas component: 10 ppm NO, 100 cc/min. Catalyst: JRC-TIO-4, 150 mg.

    In this treatment, the declined reactivity of the TiO₂ photocatalyst could be greatly recovered by heating in Ar at 473 K.^[27] After Ar heat treatment at up to 573 K, the reactivity of the TiO₂ photocatalyst recovered almost to its original level. Thus, in order to address the observed decline, we found that a complete recovery of the reactivity to its original level could be achieved by heat treatment with Ar or even with O₂ at 573 K. These results clearly indicate that the decline phenomena of TiO₂ for the decomposition of NO are caused by the adsorption species on the surface inhibiting the subsequent reactions.


Fig.3 TPD profiles of the adsorbed NO, N₂O, N₂ and O₂ recorded after the photocatalytic decomposition of NO on JRC-TIO-4 at room temperature. Catalyst: 10 mg. Vacuum pressure: 10^-5-10^-6 Pa (8x10^-8 - 8x10^-9 Torr). Heating rate: 5 K/min.

3.2 TPD measurements of the TiO2 photocatalysts after photocatalytic reactions
In order to investigate the kinds of adsorbed species which may be associated with the decline in the photocatalytic reactivity of TiO₂, temperature-programmed desorption (TPD) measurements of the catalysts were carried out after the reactions. From Fig. 3, it is clear that the decline is strongly associated with the adsorbed species on the photocatalyst, and not due to changes in the surface structure of the photocatalyst, since the photocatalytic reactivity can be recovered by Ar heat treatment at 573 K. As it can be seen in Fig. 3, with the TPD spectrum, the desorption peaks assigned to N₂, N₂O and NO can be observed. The chemical strectures of N₂, N₂O and NO were confirmed by a mass analyzer, however, the desorption peaks assigned to O₂ and NO₂ cannot be detected. The appearance of peaks due to N₂O and N₂ in addition to NO clearly indicate the decomposition of the adsorbed NO. Therefore, it can be confirmed that the decline phenomena is not caused by the adsorption of O₂. Furthermore, a single peak is observed at around 348 K due to the desorption of N₂O, and it was found that the adsorbed species of N₂O is completely dissociated up to 473 K. Figure 2 shows that the decline in photocatalytic reactivity could not be recovered to its original level without heat treatment of the catalyst with Ar or O₂ at 573 K. These results clearly indicate that the adsorbed species of N₂O which were completely decomposed into N₂ and O₂ at around 473 K are not associated with the decline. The TPD spectrum assigned to N₂ consists of two different components, i. e., a peak at around 373 K and another peak higher than 573 K. The peak at around 373 K is the same as that observed for the N₂O species which completely decomposed into N₂ and O₂ at up to 473 K, while the other N₂ peak is higher than 573 K showing the strong adsorption property of N₂ which cannot be dissociated even with treatment at 573 K with Ar or O₂. These results clearly indicate that the N₂ adsorbed species does not induce the decline in the photocatalytic reactivity. In addition to these peaks, three peaks assigned to the adsorbed NO species were observed at around 343 K, 381 K and 524 K. The dissociative peaks at around 343 K and 381 K could be eliminated by treatment at 473 K, however, the declined photocatalytic reactivity could not be recovered to its original level. It is, therefore, possible that the dissociative peaks of NO at 343 K and 381 K have little associated with the decline in photocatalytic reactivity. For the dissociation of the adsorbed NO species at around 524 K, the TPD peak can be observed between 473 K and 573 K. With the original TiO₂ photocatalyst, such desorption peaks could not be observed. In fact, they could only be observed with TiO₂ catalysts which showed a decline in photocatalytic reactivity, as shown in Fig. 2. From these results, we can conclude that the decline in the photocatalytic reactivity of TiO₂ in the decomposition of NO at 295 K is strongly associated with the formation of strong adsorption species of NO on the TiO₂ surface, which are perhaps able to adsorb on the strong acidic sites working as active surface sites on the photocatalyst. These NO species can be eliminated by treatment of the catalysts with Ar or O₂ at temperatures as high as 573 K.

Fig. 4 Effect of UV irradiation time on the TPD profiles recorded after the photocatalytic decomposition of NO. Irradiation time: (a) 120 min, (b) 30 min, (c) 10 min, (d) 0 min. Catalyst: 10 mg. Vacuum pressure: 10^-5~10^-6 Pa (8x10^-8 ~ 8x10^-9 Torr). Heating rate: 5 K/min.

3. 3 The effect of UV irradiation on the TPD spectrum of NO
Figure 4 shows the changes in the NO TPD spectra with the irradiation time in the photocatalytic decomposition of NO. The intensities of the TPD spectra become stronger with the extending UV irradiation time. The deconvolutions of the NO TPD patterns are shown in Fig. 5. The spectra consist of three different components, the a, b, and g species, each having a different dissociation temperature. The deconvolution spectra recorded after the reactions under various UV irradiation times are shown in Table 1, where it can be seen that the intensities of the a, b, and g TPD peaks remarkably increase with the UV irradiation time. In particular, the intensity of the a TPD peak due to NO is greatly increased compared with those of the b and g species. It is likely that the NO adsorbed species assigned to the a peak plays a more important role in the photocatalytic decomposition of NO than the b and g species. The TPD peak positions of the a, b, and g species were found to shift and change with the UV irradiation time. The peaks of the a, and b species shifted to lower temperature regions with prolonged UV irradiation. However, the peak of the g species which seems to play a vital role in the decline of the photocatalytic reactivity of TiO₂ for the decomposition of NO shifted to higher temperature regions, in contrast to the a and b species. These results clearly indicate that the g-species which are formed by UV irradiation adsorb strongly onto the surface of  TiO₂, i. e., on the active sites for the photocatalytic decomposition reaction of NO, causing a remarkable decline in the photocatalytic reactivity of TiO₂ under UV irradiation. Therefore, to develop highly active TiO₂ photocatalysts, the surface of TiO₂ must be modified to avoid and resist the strong NO adsorption properties of this species, investigations of which are presently underway in our laboratory.

Table 1. The results of the deconvolution of the TPD spectra of TiO₂ after the photocatalytic reaction for different UV irradiation time intervals.

Fig. 5 TPD profiles recorded after the photocatalytic decomposition of NO. (a): observed TPD profiles of a, b, and g shown in (c); (b): sum of the deconvoluted peaks shown in (c), (c): deconvoluted peaks, a, b and g.

    It has been reported that the properties of adsorption and dissociation of NO are different on metals and oxides when their crystal faces are different.^{[26, 27]} Because there are differing anatase and rutile constitutions in the standard reference catalysts JRC-TIO-1 ~ JRC-TIO-5, and their physicochemical properties are also different, these factors will naturally affect the TPD patterns. The peak positions of NO, N₂O, N₂ and O₂ on the standard reference catalysts used are shown in Table 2 along with the differences observed in the dissociation species, dissociation temperatures and numbers of TPD peaks when the physicochemical properties of TiO₂ are different. The TPD patterns of NO on the standard reference catalysts are shown in Fig. 6. From Fig. 6, NO peaks of lower temperature can be observed in all the standard reference catalysts, and those peaks can be removed when the treatment temperatures are raised up to 473 K. The intensities of the peaks of the dissociation NO species in JRC-TIO-3 and JRC-TIO-4 are strong, showing that their reactivities are higher in the photocatalytic decomposition of NO. It is possible to assign the peak at the lower temperature to the a species, i. e. the adsorbed NO species assigned to the a peak plays a more important role in the photocatalytic decomposition of NO than the b and g species. However, the direct relationship between the differences in the anatase and rutile structures, surface area, surface -OH groups, band gap and their dissociation patterns have not yet been concluded.

Fig.6 TPD profiles of the adsorbed NO species recorded after the photocatalytic decomposition of NO at room temperature for the different TiO₂ photocatalysts.

Table 2 The results of the analysis of the observed TPD patterns due to the adsorbed NO, N₂O, N₂, and O₂ on standard reference TiO₂ photocatalysts after the photocatalytic decomposition of NO at room temperature.

sh=shoulder, br=broad

4 CONCLUSIONS
In order to investigate the origin of the decline in the photocatalytic reactivity of TiO₂ in the decomposition of NO under a flow reaction system, the TPD spectra of the TiO₂ photocatalysts were investigated in detail after the photoreactions. It was found that the TPD peak assigned to the adsorbed NO species observed at around 524 K is directly associated with the decline in the photocatalytic reactivity under UV irradiation. The TPD spectrum of the adsorbed NO species consisted of three different NO adsorbed species of which the g-adsorbed species formed on the active surface sites under UV irradiation were found to cause the drastic decline in the photocatalytic reactivity of TiO₂. It was also found that the recovery of the reactivity to its original level can be achieved by heat treatment with Ar and/or O₂ at around 573 K.

REFERENCES
[1] Anpo M, Yamashita H. in "Surface Photochemistry", Anpo, M., Ed., John Wiley & Sons: Chichester, U. K., 1996, 117.
[2] Anpo M. Catal. Surv. Jpn., 1997, 1: 169.
[3] Hoffmann M R, Martin S T, Choi W et al. Chem. Rev., 1995, 95: 69.
[4] Fox M A, Dulay M T. Chem. Rev., 1993, 93: 341.
[5] Ollis D F, Al-Ekabi H  Eds. Photocatalytic Purification and Treatment of Water and Air. Elsevieer: Amsterdam, 1993.
[6] Ichihashi Y, Yamashita H, Anpo M. Study. Surf. Sci. Catal., 1997, 105: 1609.
[7] Anpo M, Ichihashi Y, Takeuchi M et al. Res. Chem. Intermed., 1998, 24 (2): 143.
[8] Serpone N, Pelizzetti E, Photocatalysis. Wiley, New York, 1989.
[9 Fan J, Yates J T. J. Am. Chem. Soc., 1996, 118: 4686.
[10] Lichtin N N, Avudaithai M. Environ. Sci. Technol., 1996, 30: 2014.
[11] Yamashita H, Zhang S G, Matsumura Y et al. Appl. Surf. Sci., 1997, 121&122: 305.
[12] Alekabi H, Serpone N. J. Phys. Chem., 1988, 92: 5726.
[13] Yamashita H, Honda M, Harada M et al. J. Phys. Chem. B, 1998, 102: 10707.
[14] Yamashita H, Kawasaki S, Ichihashi Y et al. J. Phys. Chem. B, 1998, 102: 5870.
[15] Nimlos M R, Jacoby W A, Blake D M et al. Environ. Sci. Technol., 1993, 27: 732.
[16] Park D R, Zhang J L, Ikeue K et al. J. Catal., 1999, 185: 114.
[17] Phillips L A, Raupp G B. J. Mol. Catal., 1992, 77: 297.
[18] Obee T N, Brown R T. Environ. Sci. Technol., 1995, 29: 1223.
[19] Larson S A, Falconer J L. Catal. Lett., 1997, 44: 57.
[20] Yamashita H, Ichihashi Y, Harada M et al. J. Catal., 1996, 158: 97.
[21] Peral J, Ollis D F. J. Catal., 1992, 136: 554.
[22] d'Hennezel O, Ollis D F. J. Catal., 1997, 167: 118.
[23] Cunningham J, Hodnett B K. J. Chem. Soc. Far. Trans., 1981, 77: 2777.
[24] Larson S A, Falconer J L. Catal. Lett., 1997, 44: 57.
[25] Zhang J L, Ayusawa T, Minagawa M et al. J. Catal., 2001, 198: 1.
[26] Pande N K, Bell A T. J. Catal. 1986, 97: 137.
[27] Okuhara T, Misono M. Hyoumen. 1990, 28: 87.

[ Back ] [ Home ] [ Up ] [ Next ]Mirror Site in USA  Europe  China  GBNet