Photocatalytic Degradation of 1, 10-Dichlorodecane in Aqueous

Portage Avenue, Winnipeg, Manitoba, Canada, R3B 2E9. 1,10-Dichlorodecane (D2C10) is shown to be effectively photodegraded in aqueous suspensions of ...
0 downloads 0 Views 179KB Size
Environ. Sci. Technol. 2000, 34, 1018-1022

Photocatalytic Degradation of 1,10-Dichlorodecane in Aqueous Suspensions of TiO2: A Reaction of Adsorbed Chlorinated Alkane with Surface Hydroxyl Radicals TAHA M. EL-MORSI, WES R. BUDAKOWSKI, ALAA S. ABD-EL-AZIZ, AND KEN J. FRIESEN* Department of Chemistry, University of Winnipeg, 515 Portage Avenue, Winnipeg, Manitoba, Canada, R3B 2E9

h+vb/OH•ads + RX f Products

1,10-Dichlorodecane (D2C10) is shown to be effectively photodegraded in aqueous suspensions of TiO2 using a photoreactor equipped with 300 nm lamps. Solutions exposed to UV light intensities of 3.6 × 10-5 Ein L-1 min-1, established by ferrioxalate actinometry, showed negligible direct photolysis in the absence of TiO2. The degradation rate was optimal with 150 mg/L of TiO2 and a D2C10 concentration (240 µg/L) approaching its solubility limit. Kinetics of photodegradation followed a LangmuirHinshelwood model suggesting that the reaction occurred on the surface of the photocatalyst. The presence of h+vb and OH• radical scavengers, including methanol and iodide, inhibited the degradation supporting a photooxidation reaction. Electron scavengers (Ag+, Cu2+, and Fe3+) had small effects on the degradation rate. The lack of transformation of D2C10 in acetonitrile as solvent indicated that the major oxidants were OH• radicals. The presence of tetranitromethane, effectively eliminating the formation of free OH• radicals, did not affect the degradation rates significantly. This result, combined with observed increases in photolysis rates with the degree of adsorption of D2C10 onto the surface of the photocatalyst, confirmed that the reaction involved adsorbed 1,10-dichlorodecane and surface bound OH• radicals.

Introduction Aqueous suspensions of the photocatalyst TiO2 have been used to effectively degrade many organic pollutants (1-3). In this process the semiconductor is excited with UV light of wavelengths less than 380 nm to produce conduction band electrons (e-cb) and valence band holes (h+vb) which are capable of initiating photoreduction and photooxidation reactions, respectively.

TiO2 + hν f h+vb + e-cb

(1)

Although many organic chemicals undergo photooxidations, photoreduction reactions may occur depending on the redox properties of the chemical (2, 4). There has been much debate as to whether photoreactions occur on the * Corresponding author phone: (204)786-9043; fax: (204)775-2114; e-mail: [email protected]. 1018

9

surface of TiO2 or in the bulk solution. Although adherence of the reaction to the Langmuir-Hinshelwood kinetic model supports a surface reaction, additional evidence is required for its confirmation (1, 2). For photooxidations occurring in oxygenated, aqueous media, the mechanism may, furthermore, involve direct reaction of the organic chemical (RX) with h+vb, indirect reaction with trapped holes also described as adsorbed or surface-bound hydroxyl radicals (5, 6), or a dual mechanism involving both surface holes and radicals (7):

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 6, 2000

(2)

Several studies have demonstrated the significance of sorption of the organic compound onto the TiO2 surface for effective photodegradation (8-10), lending further support for a surface reaction. The major objective of this study was to assess the potential of TiO2 suspensions to photodegrade the chlorinated n-alkane, 1,10-dichlorodecane (D2C10). This involved determination of the optimum conditions for its photocatalytic degradation, elucidation of the type of photoreaction through the use of different e-cb, h+vb, and OH• scavengers, and investigation of the significance of sorption of D2C10 to the photocatalyst surface for its photodegradation. Chlorinated paraffins, complex commercial mixtures of polychlorinated n-alkanes (PCAs), are widely used as flame retardant additives in high-pressure lubricants, metal-cutting fluids, and plastics (11). Short-chain (C10 to C13) chlorinated paraffins have recently been detected in wastewater treatment effluents and in fish and marine animals indicating their widespread occurrence in the environment and their tendency for bioaccumulation (12). Understanding the photocatalytic behavior of 1,10-dichlorodecane, a PCA which is commercially available in pure form, should be valuable in assessing the potential of this technique for the photocatalytic degradation of chlorinated paraffins in contaminated water.

Experimental Materials and Reagents. 1,10-Dichlorodecane (99%), anatase TiO2 (99.9%), and tetranitromethane were purchased from Aldrich Chemical Co. Transition metals were introduced as silver nitrate, ferric nitrate-nonahydrate, and copper nitrate. Potassium iodide (BDH) and methanol (HPLC grade, 99.97%, EM Science) were used as electron scavengers. Hydrogen peroxide (30%) was purchased from Mallinckrodt. Oxygen, nitrogen, and helium were supplied by BOC Gases. A stock solution of dichlorodecane with a concentration of 800 ng/ µL was prepared in 1:1 (v/v) water (Milli-Q purity) and acetonitrile. Aqueous solutions used for photolysis, prepared by diluting microliter volumes of the stock solution into 1 L of water, contained 0.00015 (v/v) acetonitrile. Photolyses. All solutions were irradiated in 50 mL Pyrex centrifuge tubes with a Rayonet Photochemical Reactor equipped with a carousel and 16 RPR 3000 Å lamps (13). Emission intensities of 3.6 × 10-5 Ein L-1 min-1 were established by ferrioxalate actinometry. All solutions were preequilibrated for 15 min with stirring prior to photolysis. After photolyzing for a specified time, solutions were extracted with hexane (extraction efficiencies 85 ( 3%), rotoevaporated, and concentrated to 300 µL for analysis. To determine whether D2C10 degraded by direct photolysis, corresponding experiments were conducted without the photocatalyst. In experiments requiring scavengers or modifiers, these were added to the water prior to introduction of D2C10 or TiO2. 10.1021/es9907360 CCC: $19.00

 2000 American Chemical Society Published on Web 02/04/2000

FIGURE 1. Degradation of 1,10-dichlorodecane in a dark control (9), in a solution without TiO2 (2) and in the presence of 150 mg/L of TiO2 (b). An initial concentration of 1,10-dichlorodecane of 240 µg/L was used for all solutions. When gases were used, the solutions were purged with the selected gas for 30-50 min prior to addition of any other substances. In experiments with acetonitrile as solvent, a 20 min sonication period was required to form the necessary TiO2 suspension. All experiments were run in duplicate unless otherwise reported. Adsorption. Adsorption studies were performed by allowing D2C10 to equilibrate in aqueous suspensions of TiO2 with stirring for specified periods of time under dark conditions. Slurries were centrifuged (2000 rpm for 40 min), and the aqueous solutions were extracted as described above. The amount of D2C10 sorbed to TiO2 was generally determined by mass balance. However, for several solutions the solid was analyzed directly to confirm sorption of D2C10 to TiO2. In these cases TiO2 was resuspended in clean doubly distilled water after withdrawing the original supernatant. The suspensions were filtered through 0.22 µm Millipore filters, and the solids were Soxhlet extracted for 4 h with hexane. Extracts were concentrated to volume as described above for analysis. Analyses. A Hewlett-Packard HP 5890 gas chromatograph equipped with a HP 5790 Mass Selective Detector (MSD) and a 30-m × 0.25-mm × 0.25-µm PTE-5 (Supelco) capillary column was used for analysis. A column head pressure of 14 psi and an overall helium flow rate of 55 mL/min were used. The column temperature program was as follows: 90 °C (2 min), 10 °C/min to 180 °C (2 min), 7 °C/min to 250 °C (5 min).

Results and Discussion Optimum Conditions for Photocatalytic Degradation. Photocatalytic degradation was relatively rapid with 62% disappearance of D2C10 in 15 min of photolysis as shown in Figure 1. A dark control showed no significant change in concentration over the photoperiod indicating that thermal reactions were negligible. Reaction of D2C10 in the absence of TiO2 was indistinguishable from the dark control, indicating that direct photolysis was not contributing significantly to the measured rates of photolysis with TiO2. Photodegradation rates increased as the initial concentrations of D2C10 increased from 120 µg/L to its solubility limit (14) of 240 µg/L. Therefore, in all subsequent experiments the initial concentration of 240 µg/L was used to maximize sensitivity. Photolysis rates reached a maximum at a TiO2 concentration of 0.15% (w/v), decreasing at lower TiO2 levels as the total number of active sites provided by the photocatalyst decreased and at higher TiO2 concentrations

FIGURE 2. Effect of the hole scavenger, methanol, on the photodegradation of 1,10-dichlorodecane in 150 mg/L of TiO2 at a pH of 2.8. Methanol concentrations: 0 mM (1), 1 mM (2), 10 mM (b), and 100 mM (9).

FIGURE 3. Effect of iodide ion (I-), on the photocatalytic degradation of 1,10-dichlorodecane in 150 mg/L of TiO2 at KI concentrations of 0 M (1), 10-5 M (2), and 10-3 M (9) and the effect of 10-5 M KI in the presence of 10-5 M Cu2+ (b). as light attenuation effects increased (15, 16). Therefore, in all further experiments the TiO2 concentration was maintained at 150 mg/L with a solution pH of 5.4. Identification of the Reactive Species. Information regarding the type of photocatalytic reaction may be provided by determining the influence of different electron and hole scavengers on the rate of degradation of D2C10. The Effect of Hole Scavengers. The rate of degradation of D2C10 was strongly inhibited with increasing methanol concentration as shown in Figure 2. The effects were slightly greater at pH 12 relative to pH 2.8. Since methanol is a known h+vb scavenger as well as an efficient scavenger of free or adsorbed OH• radicals (7), this result suggested that D2C10 was being photooxidized by direct interaction with the valence-band hole or by reaction with OH• radicals. Increases in the methanol concentration increased the competition for the oxidizing species, decreasing the degradation rate of D2C10. The photocatalytic degradation of D2C10 was also inhibited by the presence of iodide ion (I-) (Figure 3), lending further support for an oxidative degradation of D2C10. Iodide ion is a scavenger which reacts with h+vb and adsorbed OH• radicals (15, 17), reducing the number of oxidizing species available on the TiO2 surface for reaction with D2C10. Addition of 10-5 M Cu2+ to the reaction medium containing iodide minimized the effects of I-. By effectively scavenging conduction band electrons (e-cb), Cu2+ is reduced to unstable Cu+ which may VOL. 34, NO. 6, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1019

FIGURE 4. The photodegradation of 1,10-dichlorodecane in 150 mg/L of TiO2 without added scavengers (1), with 1.0 mM Ag+ and an oxygen concentration of 0.5 mM (9), with 1.0 mM Ag+ and 0.03 to 0.06 mM oxygen (b), and with 1.5 mM Cu2+ and 1.0 mM Ag+ (2).

FIGURE 5. The effect of electron scavengers on the photocatalytic degradation of 1,10-dichlorodecane in 150 mg/L of TiO2: 10-3 M Cu2+ (9),10-3 M Fe3+ (b), 10-3 M Ag+ (2), and a solution without added scavengers (1).

remove I- from the solution as CuI:

Cu2+ + e-cb f Cu+

(3)

Cu+ + I- f CuI

(4)

The Effect of Electron Scavengers. Electron scavengers were useful in confirming that the photocatalytic degradation of D2C10 was an oxidative process. If the substrate was being reduced by reaction with conduction band electrons, then e-cb scavengers would inhibit its rate of degradation. If the substrate was being oxidized, then the presence of electron scavengers could enhance its rate of photocatalytic degradation by minimizing the rate of recombination of e-cb and h+vb (18). Molecular oxygen is a well-known e-cb scavenger, being converted to the superoxide anion radical (19), O2.-, which eventually regenerates adsorbed oxygen, O2ads. The photocatalytic degradation of D2C10 was rapid with dissolved oxygen levels of 0.5 mM and was not affected significantly by N2 purging which reduced O2 levels to 0.03-0.06 mM. It appears that large concentrations of oxygen are not required for this photooxidation, perhaps because of the cyclic nature of the process. Silver ion, another effective electron scavenger, is reduced by e-cb to metallic Ag

Ag+ + e-cb f Ag(s)

(5)

(20) increasing the availability of h+vb or OH• for reaction with D2C10. A slight increase in the rate of reaction of D2C10 in the presence of Ag+, particularly at low oxygen concentrations (see Figure 4), is attributed to a decrease in the degree of electron-hole recombination (21). Addition of 1.5 mM Cu2+ to a solution containing 1.0 mM Ag+ inhibited the degradation rate of D2C10 (Figure 4), since the reactive Cu+ ions formed (21) compete with D2C10 for the oxidants, h+vb or OH•:

Cu+ + h+vb/OH• f Cu2+

(6)

The photocatalytic degradation rates of D2C10 in the presence of Ag+, Cu2+, and Fe3+, at reduced oxygen levels are illustrated in Figure 5. The presence of Fe3+ had little effect on the degradation rate. A large initial rate of reaction suggests a relatively rapid reduction of ferric to ferrous, a fact which was confirmed by the disappearance of the yellow color of the solution. This initial reduction prevents electron-hole recombination (20, 21) and enhances the degradation of D2C10 but will not continue to do so after the majority of Fe3+ has 1020

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 6, 2000

FIGURE 6. The effect of 10-2 M hydrogen peroxide on the photodegradation of 1,10-dichlorodecane in 150 mg/L of TiO2 with oxygen concentrations of 0.03-0.06 mM (2) and 0.5 mM (b). A solution with no added H2O2 and 0.5 mM oxygen is shown (9) for reference. been transformed to Fe2+. The slow degradation after consumption of Fe3+ is attributed to the removal of active centers from the reaction medium (22). As illustrated in Figure 5, Ag+ is the most efficient e-cb scavenger, since the product Ag(s) precipitates out of the solution. On the other hand, the tendency of Cu2+ to adsorb to the surface of TiO2 (23) and the creation of the unstable Cu+ ion (20) account for the slow degradation rates observed in the presence of Cu2+. Thus the degradation rate of D2C10 is slow in solutions containing copper with or without another electron scavenger present. Hydrogen Peroxide. Addition of 10 mM H2O2 to solutions containing either natural or reduced levels of oxygen caused a slight reduction in the rate of photocatalytic degradation of D2C10 (Figure 6). H2O2 is capable of reacting with h+vb and OH• (19) in addition to scavenging conduction band electrons. The electron scavenging ability of H2O2 is therefore offset by the formation of a less reactive oxidant species, the perhydroxyl radical (HO2•), which blocks the oxidation sites (19) leading to decreased reaction rates. The Site of the Reaction. To differentiate between direct photooxidation of D2C10 by reaction with h+vb or indirect photooxidation via reaction with OH• radicals, experiments were carried out in acetonitrile to minimize the formation of OH• radicals. The complete lack of transformation of D2C10 in acetonitrile indicated that oxidation was proceeding by reaction of D2C10 with OH• radicals (24, 25).

FIGURE 7. Fitting the kinetics of photocatalytic degradation of 1,10-dichlorodecane with TiO2 to the Langmuir-Hinshelwood model (r 2 ) 0.99), where C is the initial concentration of 1,10-dichlorodecane (mM) and r is the initial reaction rate (mM/min).

FIGURE 8. The rate of adsorption of 1,10-dichlorodecane to TiO2, showing the mass of solute sorbed to the photocatalyst as a function of the equilibration time.

Heterogeneous vs Homogeneous Oxidation. Although a heterogeneous process appeared more plausible, several approaches were used to determine if the oxidation of D2C10 by OH• radicals was a heterogeneous reaction occurring on the TiO2 surface or a homogeneous reaction in the bulk solution. The initial rates of photodegradation of D2C10 followed the modified Langmuir-Hinshelwood kinetic model (26) for solid-liquid interfaces (see Figure 7)

1 1 1 ) + r k kKC

(7)

where the initial rate (r) is calculated as the product of the first-order rate constant (k, min-1) and the initial D2C10 concentration (C, µg/L) and K is the adsorption coefficient of the reactant (27). Although the data fit a model of prior substrate adsorption followed by photodegradation up to D2C10 concentrations of 240 µg/L, further experimental evidence is required for its substantiation (24). Free OH• radicals in the bulk solution may arise from adsorbed OH• radicals which diffuse away from the surface of the photocatalyst (24) or by reduction of hydrogen peroxide formed from a reaction involving the superoxide anion radical (6). There is evidence that diffusion of surface-bound OH• radicals from the TiO2 surface into the bulk solution is minimal (6, 24). The powerful oxidizing agent tetranitromethane (TNM) was used to remove electrons and other reductants from solution (6). TNM will eliminate the species responsible for the formation of free OH• radicals in solution, yet it will not remove adsorbed OH• radicals since TNM is not strongly adsorbed to the surface of TiO2 (6, 28). The photooxidation of D2C10 was not affected by the presence of TNM indicating that the photooxidation of D2C10 involved surface bound rather than free OH• radicals. An adsorption study demonstrated that D2C10 reaches sorptive equilibrium between water and TiO2 in approximately 70 h (Figure 8). The process appears to be biphasic with a large initial rate of sorption within the first 20 min, followed by a slower sorption rate. Since all photolytic experiments were carried out following a 15-20 min equilibration period after solution preparation, D2C10 was at least 75% sorbed prior to photolysis. Therefore, it appears that adsorbed D2C10 was being oxidized by adsorbed OH• radicals. This is further supported by the fact that the photodegradation rate correlated fairly closely with the time of adsorption as shown in Figure 9. The degradation rate increased nearly 2-fold as the degree of adsorption of D2C10 onto the TiO2 surface increased with longer equilibration times. This result

FIGURE 9. Plots showing the effect of equilibration time on the degree of sorption of 1,10-dichlorodecane to TiO2 (b) and on the photodegradation rate (2). supports the common assumption that the substrate is always in contact with the semiconductor surface (8, 9, 29). Therefore, 1,10-dichlorodecane is effectively photooxidized in aqueous suspensions of TiO2. The mechanism of the reaction involves surface bound OH• radicals as evident by the lack of degradation in a nonaqueous solution. Removal of free OH• from the bulk medium with TNM did not decrease the degradation rate, indicating that adsorbed OH• radicals are the important oxidants. This was supported by the adherance of the kinetics to the Langmuir-Hinshelwood model and correlation of adsorption time with photodegradation rates of D2C10.

Acknowledgments The authors gratefully acknowledge The University of Winnipeg for financial support and thank Professors M. M. Emara and H. M. Abdelbary, Al-Azhar University, Cairo, Egypt for valuable discussions.

Literature Cited (1) Hoffmann, M. R.; Martin, S. T.; Choi, W.; Bahnemann, D. W. Chem. Rev. 1995, 95, 69-96. (2) Pelizzetti, E.; Minero, C. Electrochim. Acta 1993, 38, 47-55. (3) Pelizzetti, E.; Minero, C.; Maurino, V.; Hidaka, H.; Serpone, N.; Terzian, R. Annal. Chim. 1990, 80, 81-87. (4) Calza, P.; Minero, C.; Pelizzetti, E. Environ. Sci. Technol. 1997, 31, 2198-2203. VOL. 34, NO. 6, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1021

(5) Carraway, E. R.; Hoffman, A. J.; Hoffmann, M. R. Environ. Sci. Technol. 1994, 28, 786-793. (6) Mao, Y.; Schoneich, C.; Asmus, K.-D. J. Phys. Chem. 1991, 95, 10080-10089. (7) Sun. Y.; Pignatello, J. J. Environ. Sci. Technol. 1995, 29, 20652072. (8) Kesselman, J. M.; Lewis, N. S.; Hoffmann, M. R. Environ. Sci. Technol. 1997, 31, 2298-2302. (9) Zhao, J.; Wu, T.; Wu., K.; Oikawa, K.; Hidaka, H.; Serpone, N. Environ. Sci. Technol. 1998, 32, 2394-2400. (10) Martin, S. T.; Kesselman, J. M.; Park, D. S.; Lewis, N. S.; Hoffmann, M. R. Environ. Sci. Technol. 1996, 30, 2535-2542. (11) Environment Canada, Health and Welfare Canada. Priority Substances List Assessment Report: Chlorinated Paraffins; Catalogue No. En 40-215/17E; Government of Canada: 1993. (12) Muir, D.; Bennie, D.; Fisk, A.; Tomy, G.; Stern, G. Am. Chem. Soc. Environ. Div. 1999, 39(1), 176-178. (13) Choudhary, G. G.; Webster, G. R. B. Chemosphere 1985, 14, 9-26. (14) Drouillard, K. G.; Hiebert, T.; Tran, P.; Tomy, G. T.; Muir, D. C. G.; Friesen, K. J. Environ. Toxicol. Chem. 1998, 17, 1261-1267. (15) Martin, S. T.; Lee, A. T.; Hoffmann, M. R. Environ. Sci. Technol. 1995, 29, 2567-2573. (16) Matthews, R. Water Res. 1990, 24, 653-660. (17) Rabani, J.; Yamashita, K.; Ushida, K.; Stark, J.; Kira, A. J. Phys. Chem. B. 1998, 102, 1689-1695. (18) Halmann, M. M. Photodegradation of Water Pollutants; CRC Press: New York, 1996; pp 7-17. (19) Turchi, C. S.; Ollis, D. F. J. Catal. 1990, 122, 178-192.

1022

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 6, 2000

(20) Sakata, T,; Kawai, T.; Hashimoto, K. J. Phys. Chem. 1984, 88, 2344-2350. (21) Butler, E.; Davis, A. J. Photochem. Photobiol. A: Chem. 1993, 70, 273-283. (22) Palmisano, L.; Schiavello, M.; Sclafani,. A.; Martin, C.; Martin, I.; Rives, V. Catal. Lett. 1994, 24, 303-315. (23) Zang, L.; Liu, C.-Y.; Ren, X.-M. J. Chem. Soc., Chem. Commun. 1994, 1865-1866. (24) Serpone, N.; Pelizzetti, E.; Hidaka, H. In Photocatalytic Purification and Treatment of Water and Air; Ollis, D. F., Al-Ekabi, H., Eds.; Elsevier Science, New York, 1993; pp 225-250. (25) O’Shea, K.; Conde, A. In Photocatalytic Purification and Treatment of Water and Air; Ollis, D. F., Al-Ekabi, H., Eds.; Elsevier Science, New York, 1993; pp 707-712. (26) D’Oliveira, J.-C.; Al-Sayyed, G.; Pichat, P. Environ. Sci. Technol. 1990, 24, 990-996. (27) Glaze, W. H.; Kenneke, J. F.; Ferry, J. L. Environ. Sci. Technol. 1993, 27, 177-184. (28) Ferry, J. L.; Glaze, W. H. J. Phys. Chem. B 1998, 102, 2239-2244. (29) Cunningham, J.; Al-Sayyed, G. J. Chem Soc., Faraday Trans. 1990, 86, 3935-3941.

Received for review July 1, 1999. Revised manuscript received December 3, 1999. Accepted December 8, 1999. ES9907360