Environ. Sci. Technol. 2004, 38, 1600-1604
Photocatalytic Degradation of Organic Compounds over Combustion-Synthesized Nano-TiO2 K. NAGAVENI,† G. SIVALINGAM,‡ M. S. HEGDE,† AND G I R I D H A R M A D R A S * ,‡ Solid State and Structural Chemistry Unit and Department of Chemical Engineering, Indian Institute of Science, Bangalore-560012, India
The photocatalytic degradation of various organics such as phenol, p-nitrophenol, and salicylic acid was carried out with combustion-synthesized nano-TiO2 under UV and solar exposure. Under identical conditions of UV exposure, the initial degradation rate of phenol with combustionsynthesized TiO2 is 2 times higher than the initial degradation rate of phenol with commercial Degussa P-25 TiO2. The intermediates such as catechol (CC) and hydroquinone (HQ) were not detected during the degradation of phenol with combustion-synthesized TiO2, while both the intermediates were detected when phenol was degraded over Degussa P-25. This indicates that the rates of secondary photolysis of CC and HQ occur extremely faster than the rates at which they are formed from phenol and further implies that the primary hydroxylation step is rate limiting for the combustion-synthesized TiO2 aided photodegradation of phenol. The degradation rates of salicylic acid and p-nitrophenol were also investigated, and the rates were higher for combustion-synthesized titania compared to Degussa P-25 TiO2. Superior activity of combustionsynthesized TiO2 toward photodegradation of organic compounds can be attributed to crystallinity, higher surface area, more surface hydroxyl groups, and optical absorption at higher wavelength.
techniques is challenging due to its stability and high solubility in water. Salicylic acid is a good probe molecule to test a catalyst since it is structurally similar to many compounds that are potential environmental pollutants. Phenols and salicylic acid have been extensively studied (615) in a variety of photocatalytic reactors with different titania powders. The results indicate the complete mineralization of these compounds to CO2 and H2O through a mechanism involving hydroxylation of the aromatic ring followed by ringopening reactions. In most of the studies, commercially available titania powder, Degussa P-25, was employed either as received or modified to improve its photocatalytic activity. This form of titania was produced through hydrolysis of TiCl4 in a hydrogen flame. In a few studies, however, titania particles were prepared by reacting TiCl3 with ammonia for photocatalysis (6, 10). The solution combustion method is a single-step process and has been found to give fine particles/large surface area oxide materials such as alumina, ceria, titania, and zirconia. This method involves rapid heating (∼350 °C) of an aqueous redox mixture containing stoichiometric amounts of corresponding metal nitrates and urea/hydrazide fuels (16, 17). Titania prepared by this method is a highly crystalline, single anatase phase with high surface area and exhibits photocatalytic properties superior to those of the commercial TiO2 catalyst (Degussa P-25) for the photodegradation of various dyes under UV and solar radiation (18, 19). The photocatalytic activity of a semiconductor provides a simple and effective technique for wastewater treatment, and in the present work, the photocatalytic degradation of phenol, salicylic acid, and p-nitrophenol with UV and solar conditions was investigated over nanosized titania prepared by the solution combustion method. The photoactivity of combustion-synthesized titania was compared with that of commercial TiO2 (Degussa P-25). The intermediate products were investigated to understand the reaction pathways. It is shown that though the reaction mechanisms are similar for both the catalysts, combustion-synthesized titania degrades the organics at a faster rate compared to the commercial catalysts.
Experimental Section Introduction Semiconductor photocatalysis has been the focus of numerous investigations because of its application for the quantitative destruction of undesirable chemical contaminants in water and air (1-5). Semiconductors under light illumination of energy greater than the band gap undergo charge-transfer processes that ultimately result in oxidation of a wide variety of organic molecules. Although several oxide semiconductors (ZnO, Fe2O3, and WO3) have photocatalytic properties, the polycrystalline powders of supported or unsupported titanium dioxide, in the anatase phase, have been utilized in the major part of the investigations performed so far. The reasons for this choice are its high photostability, nontoxicity, low cost, and availability (2). Among various aromatic compounds, phenol and phenolic compounds are common contaminants in industrial wastewater. The degradation of phenol by conventional * Corresponding author fax: +91-80-360-0683; e-mail: giridhar@ chemeng.iisc.ernet.in. † Solid State and Structural Chemistry Unit. ‡ Department of Chemical Engineering. 1600
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 5, 2004
Catalyst Preparation and Characterization. The pure anatase phase nano-titania was prepared by the solution combustion method. The solution combustion method is a single-step, simple, and fast route in which an aqueous redox mixture containing stoichiometric amounts of metal salts and water-soluble fuel is heated rapidly. In a typical combustion synthesis, a Pyrex dish (300 cm3) containing an aqueous redox mixture of titanyl nitrate (2 g) and glycine (0.8878 g) in 15 mL of water was introduced into the muffle furnace preheated to 350 °C. The solution undergoes dehydration, and a spark appears at one corner, which spreads throughout the mass, yielding a voluminous solid product. This process is reproducible, and catalysts prepared in different batches showed similar characteristics and the same photocatalytic activity. The combustion-synthesized nano-TiO2 disperses very well in water. However, unlike Degussa P-25, it does not form a turbid solution in water. Though the combustion-synthesized catalyst is smaller in size compared to Degussa, it is easy to separate from water by centrifugation, for the above reason. Further details of the preparation have been described elsewhere (18, 19). The catalyst has been characterized by XRD, Raman, TEM, BET, TG-DTA, FTIR, and UV-vis spectroscopy techniques. 10.1021/es034696i CCC: $27.50
2004 American Chemical Society Published on Web 01/28/2004
Photochemical Reactor. The UV experiments were conducted in a photochemical reactor made of a jacketed quartz tube of 3.4 cm i.d., 4 cm o.d., and 21 cm length. A medium-pressure mercury vapor lamp (MPML) of 125 W (Mysore Lamps, India) was used as an illuminating source (18). The lamp radiated predominantly at 365 nm (3.4 eV). The average energy of the light emitted was 3.5 eV with a corresponding photon flux of 5.86 × 10-6 mol of photons/s (18). Cold water was circulated through the annulus of the reactor to avoid heating of the solution during the reaction. The source assembly was placed concentrically inside a Pyrex glass container of 5.7 cm i.d. and 16 cm height filled with organic solution. All the solar experiments were carried out in a cylindrical borosilicate glass reactor with an i.d. of 8 cm and volume of 400 cm3. Direct sunlight was used in this study. The experiments were conducted between 10 a.m. and 2.30 p.m. when the average solar intensity was 0.753 kW/m2 (19) and the solar intensity fluctuations were minimal. Samples were collected at regular intervals for subsequent analysis. Sample Analysis. The samples were centrifuged and filtered through 0.45 µm Millipore membrane filters to remove the catalyst particles before analysis. The UV-vis spectrophotometer (Shimadzu, UV-2100) was used for the determination of UV absorbance in the range of 190-700 nm. Calibration based on the Beer-Lambert law was used to quantify the organics concentration. Chemical analysis of the filtrate was carried out in an HPLC system (Waters Inc.). The HPLC system consisted of an isocratic pump (Waters 501), an injector (Rheodyne) with a sample loop of 500 µL and reversed-phase C-18 column, and a UV detector (GBC, Australia). The eluent stream consisted of 82.6 wt % water, 2.4 wt % acetic acid, and 15 wt % acetonitrile (15%) and was pumped at a flow rate of 2 mL/min. The UV detector was operated at 270 nm, and the output was stored digitally using a data acquisition system. Because the polarities of phenol, catechol, and hydroquinone are different, three distinct peaks were observed in the chromatograph for the mixture at the operating conditions employed. Various known concentrations of catechol in water were injected into the HPLC system, and the peak areas were obtained. A linear calibration was obtained for the variation of the catechol concentrations with peak areas. The calibration for hydroquinone was also carried out in a similar method.
Results and Discussion The XRD pattern of titania shows broad peaks, indicating smaller crystallite size. The pattern can be indexed to TiO2 in the anatase phase only. The rutile and brookite phases of TiO2 were not present. The crystallite size determined from the XRD pattern using the Sherrer formula is in the range of 6 nm. A typical XRD pattern of TiO2 is shown in Figure 1. The Raman spectrum also confirms the pure anatase phase of combustion-synthesized titania. TEM studies show that the crystallites of TiO2 are homogeneous with a mean size of 6-8 nm, which agrees well with the XRD measurements. A TEM picture of combustion-synthesized titania is given in Figure 2. The electron diffraction pattern confirms the polycrystalline nature of combustion-synthesized material, and the rings can be indexed to only anatase TiO2. The BET surface area of this titania was 240 m2/g. TGA studies of combustion-synthesized titania show an 11% weight loss, indicating more surface hydroxyl groups and water molecules. The optical absorption spectra of combustion-synthesized TiO2 show two optical absorption thresholds at 570 and 467 nm that correspond to band gap energies of 2.18 and 2.65 eV, respectively. Combustion-synthesized TiO2 absorbs appreciably at wavelengths less than 600 nm (19). The commercial catalyst (Degussa P-25 TiO2) used in the present study
FIGURE 1. XRD pattern of combustion-synthesized TiO2.
FIGURE 2. TEM of combustion-synthesized nano-TiO2. was pure anatase phase with an average particle size of 30 nm, a BET surface area of about 50 m2/g, and an optical absorption edge of 400 nm (3.2 eV). Photocatalytic Degradation of Phenol. No observable degradation of phenol over TiO2 was observed either without irradiation or without catalyst. The adsorption capacity of the combustion-synthesized and Degussa P-25 titania for phenol was evaluated in aqueous media. A 100 mL sample of 1 mM phenol was mixed with 0.1 g of the catalyst, the suspension was stirred, and the concentrations of the organics were measured in the dark over a 30 h time period. The result indicates that adsorption was only 2% over combustion-synthesized TiO2 while it was 10% over Degussa P-25 TiO2 after 30 h. The adsorption was not appreciable within 2-3 h. Hence, the initial concentration was taken to be Co in all cases. The initial rates of the reaction are determined by conducting a series of concentration versus time experiments at different initial concentrations, and each run is extrapolated back to the initial conditions (19). The slopes calculated on the basis of three to five points were nearly constant. The initial point, t ) 0, was taken after addition of the catalyst. The effect of catalyst loading on the photocatalytic degradation of phenol was studied by varying the amount of TiO2 from 0.25 to 2 kg/m3; the initial rates reached saturation for catalyst loading above 1 kg/m3. Thus, 1 kg/m3 VOL. 38, NO. 5, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1601
FIGURE 3. Effect of initial concentration on the degradation of phenol with combustion-synthesized TiO2 and a catalyst loading of 1 kg/m3. catalyst loading was chosen for the degradation of all the organics. The effect of the initial concentration of phenol on the photocatalytic degradation rate was investigated over the concentration range of 0.34-1.0 mmol/L (30-94 ppm) with combustion-synthesized titania, and the experimental results are presented in Figure 3. It can be seen from the figure that the concentration has a significant effect on the degradation rates and the rate of decrease in the phenol concentration is faster when the initial concentration is less. Kinetic studies showed that phenol degradation under UV irradiation is in accordance to the Langmuir-Hinshelwood (L-H) equation (20, 21). Recently, we have proposed a generalized model for the photocatalytic degradation of any species (18). It takes the form r ) -(dC/dt) ) [krKeC/(1 + KeC)], where kr and Ke are the apparent reaction rate constant and apparent equilibrium of adsorption constant. The rate form can be linearized for initial concentrations as 1/ro ) (1/krKe)(1/Co) + 1/kr, where Co is the initial phenol concentration and ro is the initial degradation rate (18). The intercept of the plot of 1/ro and 1/Co determines the kinetic coefficient that varies directly with the rate of degradation, and the parameter Ke represents the equivalent of the adsorption coefficient. The rate constants, kr, for combustion-synthesized and Degussa P-25 TiO2 are 0.287 and 0.016 µmol L-1 s-1, respectively. The adsorption equivalent L-H parameter Ke is 0.522 × 103 and 7.937 × 103 L/mol for combustion-synthesized and Degussa P-25 TiO2, respectively. The higher value of kr indicates the higher degradation rate of phenol over combustionsynthesized titania. The Ke value further indicates the amount of strong or irreversible adsorption of phenol over Degussa P-25 compared to combustion-synthesized TiO2. Figure 4a shows the degradation profile of phenol for combustion-synthesized and Degussa P-25 TiO2 with an initial concentration of 0.5 mmol/L with a catalyst loading of 1 kg/m3 under UV. The phenol disappearance was essentially complete at 120 min with the combustionsynthesized TiO2, while Degussa P-25 shows only 18% conversion at that time. The initial rates of degradation of phenol with combustion-synthesized TiO2 and Degussa P-25 are 0.023 and 0.012 µmol L-1 s-1, respectively. The reason can be explained from the observed rate constants. It can be seen from the values that the kinetic rate constant for combustion-synthesized TiO2 is 1 order higher than that for Degussa P-25 and the adsorption equilibrium constant is 1 order less than that for Degussa P-25. It implies that the phenol adsorption is not very strong over the combustion1602
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 5, 2004
FIGURE 4. Degradation profiles of phenol with an initial concentration of 0.5 mmol/L and a catalyst loading of 1 kg/m3 with combustionsynthesized and Degussa P-25 TiO2 under (a) UV and (b) solar irradiation. synthesized TiO2 compared to Degussa P-25 and hence led to higher degradation and thus superior activity (18). Experiments were also carried out to examine the degradation of phenol under solar radiation with an initial concentration of 0.5 mmol/L and a catalyst loading of 1 kg/ m3, and the degradation profile is shown in Figure 4b. It is seen that complete degradation of phenol required only 200 min with combustion-synthesized TiO2, while Degussa shows a decrease in the phenol concentration only up to 25%. This indicates that the photocatalytic activity of the combustionsynthesized TiO2 is higher than that of the commercial Degussa P-25 catalyst for both UV and solar exposure. Phenol Intermediates and Reaction Pathway. Various groups (9, 14, 15) have reported the pathway of phenol degradation using TiO2 as a photocatalyst. The results indicate that the degradation proceeds through the stepwise formation of intermediates. Catechol (CC), hydroquinone (HQ), hydroxyhydroquinone (HHQ), and pyrogallol (PG) are the products of the initial stages of the degradation. These aromatic intermediates undergo further photocatalytic oxidation to ring cleavage to yield carboxylic acids and aldehydes, which give CO2 and H2O due to decorboxylation as shown in Figure 5. Thus, the photocatalytic decomposition of phenol mainly proceeds via its hydroxylated compounds. The HPLC analysis of the solution from phenol degradation under UV with combustion-synthesized TiO2 was carried out to investigate the formation of intermediates and compare it with the phenol degradation with Degussa P-25 TiO2. When the degradation of phenol was carried out with combustionsynthesized TiO2 in UV, no CC or HQ was detected during the reaction and the concentration of phenol continuously
FIGURE 5. Scheme of phenol degradation with reaction pathways. decreased with time. However, when the degradation of phenol was conducted with Degussa P-25, two major peaks for CC and HQ were detected and the concentration of HQ and CC increased with the decrease in the phenol concentration. At 180 min over Degussa P-25, phenol conversion to open ring fragments was 34.5%, with combined amounts of HQ, CC, and unreacted phenol of 23.4%, 22.3%, and 19.8%, respectively. Two possible explanations for these observations can be given. One of them could be that the degradation of phenol in the presence of combustion-synthesized TiO2 catalysts is by a different mechanism. To validate the hypothesis, experiments were carried out for the degradation of phenol with combustion-synthesized TiO2 with solar exposure. When the solution was analyzed by HPLC, small amounts of intermediates (CC, HQ) were detected. This indicates that the mechanism of degradation of phenol by both the catalysts is the same. The observed reaction behaviors can be explained from the various limiting steps of the reactions mentioned in Figure 5. From the experimental data, it appears that secondary hydroxylation (k24, k35, k25) and ring-opening reactions (k46, k56) are extremely faster compared to the primary hydroxylation reactions (k12, k13) for the degradation of phenol by combustion-synthesized catalyst. Most of the phenolic species including CC and HQ among others are impurities for water. Since no detectable amounts of these species are formed when phenol is degraded by combustion-synthesized TiO2, the degree of pollution in the water bodies is reduced compared with that during degradation by the commercial material Degussa P-25. The concentrations of the intermediates formed during the degradation of phenol by Degussa P-25 increase initially with time and then decrease with time because of the competition of the primary and secondary hydroxylation steps. A preliminary analysis of the data indicates that the photocatalytic degradation rate coefficients for the primary and secondary hydroxylation steps for phenol degraded by Degussa P-25 are 4 × 10-3 and 10-2 min-1, respectively. However, the degradation rate coefficient for the secondary hydroxylation step for phenol degraded by combustion-synthesized TiO2 would be much higher (around 0.1 min-1), and thus no intermediate formation is observed. The fast consumption of the intermediates may be due to the increase of adsorption of these intermediates when the primary substrate does not adsorb extensively. Thus, the difference in reactivities of the catalysts may also be due to the differences in adsorption of the intermediates on the catalysts. Thus, on an overall basis, the total disappearance rate of phenol is at least 1 order higher for combustion-synthesized catalyst compared to Degussa P-25. Thus, the combustion-synthesized catalysts are superior in catalytic activities as well as in terms of the toxic impurities formed during degradation of phenol. p-Nitrophenol is used in the production of pesticides and synthetic dyes. It is also used in insecticides and herbicides. Since it has significant water solubility, it is often present in
FIGURE 6. Degradation profiles of salicylic acid with an initial concentration of 30 ppm (0.217 mmol/L) with combustion-synthesized TiO2 and a catalyst loading of 1 kg/m3.
FIGURE 7. Degradation profiles of p-nitrophenol with an initial concentration of 1 mmol/L (139 ppm) with combustion-synthesized TiO2 and a catalyst loading of 1 kg/m3. wastewater (12). To check the photoactivity of combustionsynthesized titania, degradation of salicylic acid and pnitrophenol was carried out under UV, and the activity was compared with that of the Degussa catalyst. Figure 6 shows the degradation profile of salicylic acid with an initial concentration of 0.217 mmol/L (30 ppm) and a catalyst loading of 1 kg/m3 under UV irradiation. It can be seen from the figure that combustion-synthesized TiO2 shows a 90% decrease in the concentration of salicylic acid in 200 min while the conversion was 35% and reached a saturation of C/Co of 0.65 with Degussa P-25. The initial rates of degradation of salicylic acid with combustion-synthesized TiO2 and Degussa P-25 are 0.075 and 0.028 µmol L-1 s-1, respectively, 2.6 times higher for combustion-synthesized TiO2 than Degussa P-25 TiO2. Figure 7 shows the degradation profile of p-nitrophenol under UV exposure with an initial concentration of 1 mmol/L (139 ppm) and a catalyst loading of 1 kg/m3. It is clear from the figure that combustion-synthesized TiO2 shows complete degradation of p-nitrophenol in 60 min while Degussa P-25 shows saturation around 0.5 mmol/L (69.5 ppm). The initial rates for the degradation of p-nitrophenol with combustionsynthesized and Degussa P-25 TiO2, respectively, are 1.64 and 0.37 µmol L-1 s-1, 4.4 times higher for combustionsynthesized TiO2 than Degussa P-25 TiO2. It is not entirely clear to us why combustion-synthesized TiO2 shows higher photocatalytic activities in both UV and VOL. 38, NO. 5, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1603
solar exposure compared to Degussa P-25 TiO2. The reason can be (a) a smaller particle size along with a higher surface area, (b) a higher crystallinity achieved due to the high temperature reached during combustion synthesis, (c) a higher density of adsorbed water or surface hydroxyl groups, and (d) a relatively smaller band gap. Heat treatment of combustion-synthesized TiO2 at 400 °C for 48 h reduces the surface area to 63 m2/g, and the particle size increases to 16 nm. But the solar photoactivity of the heated catalyst was comparable with that of the freshly prepared catalyst. The photoactivity of combustion-synthesized titania is, therefore, not a mere function of the surface area and crystallite size. The crystallinity of TiO2 is known to influence the degradation of phenol (11). TiO2 synthesized by this method is highly crystalline (19), and the crystallinity extended till the surface of the crystal seems to be the main reason for the high catalytic activity. A higher amount of surface hydroxyl groups and more efficient absorption of UV and solar light are also other reasons for a higher catalytic activity (19). Further study on the surface characterization of combustion-synthesized TiO2 is necessary to find the exact reason for the higher photocatalytic activity. This study has shown that combustion-synthesized nanoTiO2 catalyst is more effective than commercial Degussa P-25 TiO2 for the photodecomposition of phenol, salicylic acid, and p-nitrophenol. No detectable amounts of toxic impurities such as CC and HQ were present during the degradation of phenol with combustion-synthesized catalyst compared to the Degussa catalyst, which makes the combustion-synthesized catalyst ecofriendly. The high photocatalytic activity of combustion-synthesized titania compared to the Degussa catalyst was attributed to a higher surface area, crystal structure (anatase), and density of the surface hydroxyl groups.
Acknowledgments Financial support from the Department of Science and Technology, Government of India, is gratefully acknowledged.
1604
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 5, 2004
Literature Cited (1) Ollis, D. F.; Al-Ekabi, H. Photocatalytic Purification and treatment of Water and Air; Elsevier: Amsterdam, 1993. (2) Fox, M. A.; Dulay, M. T. Chem. Rev. 1993, 93, 341-357. (3) Linsebrigler, A. L.; Lu, G.; Yates, J. T. Chem. Rev. 1995, 95, 735758. (4) Hoffmann, M. R.; Martin, S. T.; Choi, W.; Bahnemann D. W. Chem. Rev. 1995, 95, 69-96. (5) Ollis, D. F.; Pelizzetti, E. Serpone, N. Environ. Sci. Technol. 1991, 25, 1523-1529. (6) Palmisano, L.; Augugliaro, V.; Sclafani, A.; Schiavello, M. J. Phys. Chem. 1988, 92, 6710-6713. (7) Sclafani, A.; Palmisano, L.; Schiavello, M. J. Phys. Chem. 1990, 94, 829-832. (8) Matthews, R. W.; McEvoy, S. R. J. Photochem. Photobiol., A 1992, 64, 231-246. (9) Chen, J.; Eberlein, L.; Langford, C. H. J. Photochem. Photobiol., A 2002, 148, 183-189. (10) Fotou, G. P.; Pratsinis, S. E. Chem. Eng. Commun. 1996, 151, 251-269. (11) Tanaka, K.; Capule, M. F. V.; Hisanaga, T. Chem. Phys. Lett. 1991, 187, 73-77. (12) San, N.; Hatipoglu, Kocturk, G.; Cinar, Z. J. Photochem. Photobiol., A 2002, 146, 189197. (13) Yawalkar, A. A.; Bhatkhande, D. S.; Pangarkar, V. G.; Breenackers, A. C. M. A. J. Chem. Technol. Biotechnol. 2001, 76, 363-370. (14) Turchi, C. S.; Ollis, D. F. J. Catal. 1989, 119, 483-496. (15) Okamoto, K.; Yamamoto, Y.; Tanaka, H.; Tanaka, M.; Itaya, A. Bull. Chem. Soc. Jpn. 1985, 58, 2015-2022. (16) Patil, K. C.; Amal Sekar, M. M. Int. J. Self-Propag. High-Temp. Synth. 1994, 3, 181-196. (17) Aruna, S. T.; Patil, K. C. J. Mater. Synth. Process. 1996, 4, 175179. (18) Sivalingam, G.; Nagaveni, K.; Hegde, M. S.; Madras, G. Appl. Catal., B 2003, 45, 23-38. (19) Nagaveni, K.; Sivalingam, G.; Hegde, M. S.; Madras, G. Appl. Catal., B, in press (http://dx.doi.org/10.1016/j.apcatb.2003. 09.013). (20) Matthews, R. W. J. Catal. 1988, 111, 264-272. (21) Al-Ekabi, H.; Serpone, N. J. Phys. Chem. 1988, 92, 5726-5731.
Received for review July 2, 2003. Revised manuscript received December 14, 2003. Accepted December 17, 2003. ES034696I
