Photo Or Solar Ferrioxalate Disinfection Technology Without External Hydrogen Peroxide Supply

Environ. Eng. Res. Vol. 12, No. 5, pp. 238～243, 2007 Korean Society of Environmental Engineers

Photo or Solar Ferrioxalate Disinfection Technology without External Hydrogen Peroxide Supply

Min Cho1,2, Joonseon Jeong1, Jaeeun Kim1, and Jeyong Yoon1,†

1School of Chemical and Biological Engineering, College of Engineering, Seoul National University, San 56-1, Sillim-dong, Gwanak-gu, Seoul, 151-742, Korea 2Korea Interfacial Science and Engineering Institute, Yangjae-dong, Seocho-gu, Seoul, 137-899, Korea

Received October 2007, accepted December 2007

Abstract

The Fenton reaction, which refers to the reaction between ferrous ions and hydrogen peroxide to produce the OH radical, has not been widely applied to the disinfection of microorganisms despite being economic and environmentally friendly. Cho et al. have previously proposed the neutral photo ferrioxalate system as a solution to the problems posed by the Fenton reaction in acidic conditions,1) but this system still requires an external hydrogen peroxide supply. In the present study, we developed a simple disinfection technology using the photo or solar ferrioxalate reaction without the need for an external hydrogen peroxide supply. E. coli was employed as the indicating microorganism. The study results demonstrated the effectiveness of the photo ferrioxalate system in inactivating E. coli without any external hydrogen peroxide supply, as long as dissolved oxygen is supplied. Furthermore, the solar ferrioxalate system achieved faster inactivation of E. coli than an artificial light source at similar irradiance.

Keywords: E. coli, Inactivation, Ferrioxalate, Photo or solar ferrioxalate, OH radical

1. Introduction rendering the application of this technology difficult and com- 1 plicated in terms of its operation and maintenance.4,8) 2+ The Fenton reaction (Fe /H2O2) has been widely applied to One method for producing in situ hydrogen peroxide is the the purification of non-biodegradable wastewater in the field of complete utilization of the photochemical reaction of ferrioxa- 2-4) III 3- 9) industrial wastewater treatment. Despite its dual advantages late ([Fe (C2O4)3] ). This approach requires the use of dis- of being economic and environmentally friendly, the Fenton solved oxygen (O2), supplied by bubbling with oxygen, which reaction, which refers to the reaction between ferrous ions and becomes the precursor for hydrogen peroxide through a series 5-7) hydrogen peroxide to produce the OH radical, has not been of reactions.4) widely used in microbial disinfection, due to the requirement of - - an acidic condition (pH of 3) and the formation of large amounts CO2 ･ + O2 → CO2 + O2 ･ equation (II) of iron sludge. - + Fe(II) + O2 ･ + 2H → Fe(III) + H2O2 equation (III) Fe(II) + H2O2 → ･OH equation (I) Accordingly, we examined the applicability of the photo fer- Cho et al. recently reported that a modified version of the rioxalate system as disinfection technology by generating in Fenton reaction, called the photo ferrioxalate system, can be situ hydrogen peroxide, and of the solar ferrioxalate system us- used to inactivate microorganisms at near neutral pH, even ing solar irradiation instead of artificial light (300-420 nm).10,11) under non biocidal light irradiation, as long as an appropriate Furthermore, we elucidated the role of reactive species formed 1) amount of oxalate is present. However, one limitation of this during the photochemical ferrioxalate reaction for microbial technology is the requirement for an external hydrogen pero- inactivation. xide supply, which is the key reagent of the Fenton reaction, Escherichia coli (E. coli) was selected as the indicator microorganism in this study, due to its similar characteristics to water- †Corresponding author borne bacterial pathogens such as Salmonella and Shigella and 12,13) E-mail: [email protected] its use in many previous disinfection studies. Tel: +82-2-880-8927, Fax: +82-2-876-8911 Photo or Solar Ferrioxalate Disinfection Technology without External Hydrogen Peroxide Supply 239

2. Materials and Methods was cut on an Ultra-microtome (MT-X, RMC, Tucson Co., USA) and sectioned samples were stained in 2% uranyl acetate for 7 2.1. Preparation and Analysis of Chemicals min, followed by Reynold’s lead citrate for 7 min. Finally, grids were assayed using TEM (JEM-1010, JEOL Co., Japan). All solutions were prepared using de-ionized water (Millipore Co., USA), and analytical reagent grade chemicals (Aldrich Co. 2.4. Experimental Procedures and Sigma Co., USA). All glassware used was washed with dis- tilled water and then sterilized by autoclaving at 121°C for 15 The procedures for the disinfection experiment with the photo 3+ 2- min. Stock solutions of Fe (5 mM), oxalate (C2O4 , 300 mM) ferrioxalate system were similar to those described in our pre- 1) and hydrogen peroxide (H2O2, 10 mM) were prepared using the vious study. Experiments with artificial light were performed 9,14) same procedure as that described in detail in the studies. using 50 mL of the suspension in a 60 mL Pyrex reactor. The 3+ The concentrations of Fe and hydrogen peroxide were mea- reactor was closed to the atmosphere with a cap. The suspension 15) 16) sured using the phenanthroline and titanium sulfate methods, was irradiated from a distance of 2 cm and stirred vigorously at respectively. The oxalic acid and para-chlorobenzoic acid (pCBA; a temperature of 21 ± 1°C (maintained by air cooling). The in- 1.92 µM) concentrations were analyzed by high performance cident photon flow of the four Black Light Blue (BLB) lamps liquid chromatography (HPLC, Waters Co., USA) with a UV (18 W, Philips Co., Netherlands), which emit in the wavelength detector, employing carbohydrate (SUPELCOGEL C-610H, range of 300-420 nm, was estimated by ferrioxalate actinometry18) -6 300 mm × 7.8 mm, 9 μm) and C18 reverse-phase (XTerra Rp-18 to be 7.9 × 10 Einstein/L.sec while the irradiance was measured reverse-phase column, 150 mm × 2.1 mm, 5 μm) columns, by a radiometer (UVX radiometer, UVP Co., USA) at 360 nm respectively. was 0.8 mW/cm2. The initial pH was adjusted to 5.8 with phos- phate buffer solution (PBS; KH2PO4 + NaOH; 5 mM). Aerobic 2.2. Preparation and Analysis of E. coli conditions were obtained by bubbling high purity O2 or air gas through the solution for 20 min before starting the experiments E. coli (ATCC strain 8739) stock was prepared following the and anaerobic conditions were obtained using N2 bubbling. 12) procedure described in our previous work. In brief, E. coli The solution used for the photolysis experiment was prepared was cultured by growing in tryptic soy broth for 18 hr at 37°C. by adding appropriate amounts of Fe3+ (0.04-0.20 mM), oxalate The number of viable E. coli was measured by counting the E. (5 mM) and PBS (5 mM), followed by spiking the diluted E. coli coli colonies, which was formed using the spread plate method stock solution. The ferrioxalate solutions were protected from on nutrient agar (Difco Co., USA) after incubation at 37°C for light by means of aluminium foil, prior to photolysis, in order 24 hr. This non-selective nutrient agar medium was chosen for to prevent any photochemical reactions from occurring. The the E. coli count in order to give a more conservative population photo ferrioxalate experiments were started by removing the estimate. During the disinfection experiments, the initial popu- aluminium foil. During all experiments, because the oxalate was 5 lations of E. coli were adjusted within the range of 1.5 × 10 to continuously decomposed, the appropriate amount of stock oxa- 5 3.2 × 10 cfu/mL and 1 mL of solution was withdrawn at each late solution was injected into the reaction solution at each time sampling time and diluted 1/10 and 1/100. Three replicate 0.1 interval by an automatic injection system, so that it was main- mL samples of the diluted and undiluted solutions were spread tained at an average of 75% of its initial concentration. Tert- for the cell count and good reproducibility was achieved with butanol (t-BuOH; 20 mM) and pCBA were used as the OH rad- 10% standard deviation. ical scavenger and OH radical probe compounds, respectively.12,19) In general, 5-8 samples were taken during the disinfection 2.3. Transmission Electron Microscopic (TEM) Assay experiments (20-240 min) to measure the microorganism popu- lations. The samples were quickly quenched with the appropriate The structural change of E. coli induced by the disinfection amount of sodium sulfite (Na2SO3, 1 mM), which consumed 17) 8 step was examined with TEM. E. coli suspensions (10 cfu/mL), any ferrioxalate-initiated oxidants and residual hydrogen pero- treated by the photo ferrioxalate system to achieve 1 log inacti- xide. The quenching reagent itself did not affect the viability of vation, were fixed in 2% para-formaldehyde for 2 - 4 hr at 4°C. E. coli. Samples were centrifuged (5000 rpm, 10 min, 4°C) and rinsed The ferrioxalate experiments with solar irradiation, termed three times in 0.05 M sodium cacodylate buffer (pH 7.2). The the solar ferrioxalate system, were performed with a 50 mL quartz pellet was suspended in 0.05 M of sodium cacodylate buffer reactor to allow the UV-C (< 300 nm) component of sunlight to containing 1% of osmium tetraoxide and kept at 4°C for 2 hr. penetrate into the sample solution. The oxygen condition was After the post fixation step, E. coli was rinsed twice using dis- maintained by bubbling air. The final concentration of Fe3+ and tilled water and stained in 0.5% uranyl acetate at 4°C overnight. oxalate was controlled to 0.1 mM and 5.0 mM, respectively, and Samples, gradually dehydrated using ethanol, were placed in hydrogen peroxide (0.2 mM) was added in one initial pulse propylene oxide (100%) for 15 min, infiltrated using propylene mode to achieve the rapid inactivation of the solar ferrioxalate oxide/Spurr’s resin (Low viscosity embedding kit, Electron system. Microscopy sciences Co., USA) for 30 min, and embedded in The disinfection experiments in the solar ferrioxalate system 100% Spurr’s resin and polymerized for 24 hr at 70°C. E. coli were conducted on a sunny summer day on the roof of our labo- 240 Min Cho, Joonseon Jeong, Jaeeun Kim, and Jeyong Yoon ratory building in Seoul (126°56´W, 37°27´N), Korea, in order to have a similar irradiance to that used for the photo ferrioxalate system. During the experiments, the irradiance and temperature varied within the range of 0.92 ± 0.22 mW/cm2 and 21-32°C, respectively. The intensity of solar irradiation was measured with a radiometer at 360 nm. Because of the increased temperature with the solar irradiation itself, additional control experiments were conducted in PBS (30 mM) at 30°C.20-22) During the control experiments of the photo ferrioxalate system, no E. coli inactivation was observed without BLB radiation Fig. 1. Reaction pathway in the photo or solar ferrioxalate system. or oxalate addition. Experiments were repeated three times and their averaged values with statistical deviations were used for To verify the formation of hydrogen peroxide and OH radical, the data analysis. The statistical analysis in this study was con- demonstrated in Fig. 1, several experiments were conducted to ducted using the Microsoft Excel program (version 2003, Micro- analyze the concentration of hydrogen peroxide and the degrada- soft Co., USA). tion of an OH radical probe compound (pCBA), and the results are described in Fig. 2. Fig. 2(a) shows the generation of hydro- 3. Results and Discussion gen peroxide in the photo ferrioxalate system with oxygen bubbling. In the presence of saturated oxygen, hydrogen peroxide 3.1. Reaction Pathway of the Photo and Solar Ferrioxalate was gradually generated, reaching at a concentration of 0.18 Systems mM at 140 min (Fig 2(a)), supporting the reactions (c) & (d) in Fig. 1. Fig. 2(b) shows the degradation profiles of pCBA in the The reaction pathway for the photochemical reaction of ferri- photo ferrioxalate system. pCBA was added three times sepa- oxalate is schematically presented in Fig. 1. When ferrioxalate III 3- rately during the hydrogen peroxide generation. Since pCBA is ([Fe (C2O4)3] ) absorbs light, it produces Fe(II) and the oxalyl 12) - an OH radical probe compound, the degradation of pCBA in radical anion (C2O4 ･) through the ligand to metal charge Fig. 2(b) apparently confirms the presence of OH radical indu- transfer. ced by reaction (a) in Fig. 1. In addition, the different rate of

III 3- 2- - pCBA degradation in Fig. 2(b) could be explained by the diffe- [Fe (C2O4)3] + hv → Fe(II) + 2C2O4 + C2O4 ･ equation (IV) rent concentrations of hydrogen peroxide (Fig. 2(a)) and OH radical (reaction (a)) during the experimental time. This reaction occurs even in the case of irradiation using a solar or artificial lamp within the wavelength range of 300-420 3.2. E. coli Inactivation in the Photo Ferrioxalate System nm. The rate of this initiation reaction is linearly proportional to its primary quantum yield (less than unity), because of its back The effect of oxygen on the E. coli inactivation by the photo 5) - reaction and the rate of light absorption. The C2O4 ･ produced ferrioxalate system was investigated according to the presence as a result of photo-reduction is not significantly involved in of a hydrogen peroxide supply (Fig. 3). Three important obser- any other reactions, due to its rapid decarboxylation dissocia- vations can be made from the results of Fig. 3. First, E. coli was - tion to the carbon dioxide radical anion (CO2 ･) and carbon apparently inactivated in oxygenated conditions (-○-), even in 2) dioxide. the absence of an external hydrogen peroxide supply, as 96 min was required for 2 log inactivation. As shown in reaction (a) of - - 6 -1 C2O4 ･ + CO2 ･ + CO2 (reaction rate : 2×10 s ) equation (V) Fig. 1, this observation is important since the external supply of hydrogen peroxide is the key reagent to generate the OH radical - When dissolved oxygen is present, the photo-generated CO2 ･ responsible for microbial inactivation in the photo ferrioxalate faces two major comparative pathways, namely reactions (b) system. Therefore, a photo ferrioxalate system that does not and (c) in Fig. 1, depending on the concentrations of ferrioxa- require an external hydrogen peroxide supply will possess a late and oxygen (O2). major advantage in field applications. Second, no significant difference in the E. coli inactivation was apparent according to - III 3- CO2 ･ + [Fe (C2O4)3] → Fe(II) + CO2 + the type of oxygen supply, pure oxygen (O2 (-○-)) or air bubb- 2- 3C2O4 ･(Reaction (b)) equation (VI) ling (-●-). This observation provides another advantage in the practical application of the photo ferrioxalate system. Third, a - CO2 ･ + O2 →→ CO2 + H2O2 (Reaction (c)) equation (VII) large lag phase of approximately 60 min (-○- in Fig. 3) was observed in the photo ferrioxalate disinfection practice, possi- - In an oxygen saturated condition, CO2 ･ is mainly involved bly due to the time required for (1) destroying significant sur- in a reaction with oxygen, i.e. that is reactions (c) and (d), to face components of the microorganism12,23) and (2) attaining a produce hydrogen peroxide, after which the OH radical is pro- certain quantity of in situ hydrogen peroxide generation. The duced from the reaction between ferrous ion and hydrogen shortened lag phase of E. coli inactivation (-■- in Fig. 3.) resu- peroxide (reaction (a) in Fig. 1). lting from the pulse input of hydrogen peroxide (0.2 mM) at the Photo or Solar Ferrioxalate Disinfection Technology without External Hydrogen Peroxide Supply 241

(a) (b) Fig. 2. Hydrogen peroxide formation (a) and pCBA degradation (b) supporting the presence of OH radicals in the photo ferrioxalate system (oxy- 3+ -6 3+ gen bubbling, [Fe ] = 0.1 mM, [oxalate]0 = 5.0 mM, incident photon flow = 7.9 × 10 Einstein/L.sec, [oxalate]/[Fe ] ≥ 25, pH = 5.8-6.5, temperature = 21 ± 1°C, [pCBA]0 = 1.92 µM).

Fig. 3. Inactivation of E. coli in the photo ferrioxalate system without Fig. 4. Effect of ferric ion dose on E. coli inactivation in the photo 3+ any external H2O2 supply ([Fe ] = 0.1 mM, [oxalate]0 = 5.0 mM, inci- ferrioxalate system without any external hydrogen peroxide supply -6 2 3+ dent photon flow = 7.9 × 10 Einstein/L.sec, irradiance = 0.8 mW/cm , (H2O2) ([Fe ] = 0.04 - 0.2 mM, [oxalate]0 = 5.0 mM, incident photon -6 2 [oxalate]/[Fe3+] ≥ 25, pH = 5.8-6.5, temperature = 21 ± 1°C) flow = 7.9 × 10 Einstein/L.sec, irradiance = 0.8 mW/cm , [oxalate]/ [Fe3+] ≥ 25, pH = 5.8-6.5, temperature = 21 ± 1°C) beginning supports the last explanation about the required time for hydrogen peroxide generation. When hydrogen peroxide 180 to 78 min). The effect of the ferric ion concentration on E. was injected once, the efficacy of microbial inactivation for 2 coli inactivation was found to be similar regardless of the exter- 1) log E. coli inactivation was elevated by 53%, which is consis- nal hydrogen peroxide supply in the photo ferrioxalate system. tent with a relatively small lag phase in the hydrogen peroxide supplying system.1) 3.4. E. coli Inactivation in the Solar Ferrioxalate System

3.3. Effect of Ferric Ion The solar ferrioxalate system is defined as the ferrioxalate system in which the artificial light is replaced with solar irradia- The effect of the ferric ion concentration on E. coli inactivation tion. Fig. 5 shows the results of E. coli inactivation in the solar was examined in the photo ferrioxalate system without an exter- ferrioxalate system. Two log E. coli inactivation was achieved nal hydrogen peroxide supply, with a ferric ion concentration four times faster with the ferrioxalate system in the presence of ranging from 0.04 to 0.20 mM (Fig. 4). E. coli inactivation was solar irradiation (BLB irradiance of 0.9 mW/cm2, temperature increased with increasing ferric ion concentration. As the initial of 21-32°C, -■- in Fig. 5) than with the photo ferrioxalate sys- ferric ion concentration increased from 0.04 to 0.20 mM, the time tem at similar experimental conditions (BLB irradiance of 0.8 required for 2 log E. coli inactivation decreased by 57% (from mW/cm2, temperature of 21 ± 1°C, -■- in Fig. 3). The reaction 242 Min Cho, Joonseon Jeong, Jaeeun Kim, and Jeyong Yoon

In the photochemical ferrioxalate system, several reactive - - species (OH radical, H2O2, O2 ･ and CO2 ･) have the potential to cause E. coli inactivation (Fig. 1). Several additional experiments were therefore conducted to elaborate the role of reactive species for E. coli inactivation formed in the photo ferrioxalate system. In the absence of oxygen by nitrogen gas (N2) bubbling, it is - clear that only CO2 ･ could exist among the reactive species of the photo ferrioxalate system, due to the inhibition of reactions (a) and (c) in Fig. 1. However, no E. coli inactivation was observed in nitrogen condition (data not shown), indicating that - CO2 ･ does not cause any significant E. coli inactivation under the given experimental conditions. When photo ferrioxalate experiments were conducted in the presence of an excess amount of the OH radical scavenger (20 mM of t-BuOH), no E. coli inactivation was detected (data not Fig. 5. Inactivation of E. coli in the solar ferrioxalate system (one - - 3+ initial pulse input of 0.2 mM H2O2, air bubbling, ([Fe ] = 0.1 mM, shown). Because other reactive species (H2O2, O2 ･ and CO2 ･), 2 [oxalate]0 = 5.0 mM, irradiance = 0.92 ± 0.22 mW/cm at 360 nm, except the OH radical, exist in the OH radical scavenging condi- temperature = 21-32°C). tion, this result confirms that the OH radical, which was produced by the Fenton reaction (reaction (a) in Fig. 1),24,25) is the temperature in the solar ferrioxalate system was gradually incr- main species responsible for E. coli inactivation in the photo eased from 20°C to 32°C, possibly due to radiant heat. Since no ferrioxalate system. significant E. coli inactivation was observed (data not shown), Fig. 6(b), showing the TEM image of E. coli when 1 log inac- and as E. coli was left at 30°C for 60 min in PBS, the increased tivation was attained in the photo ferrioxalate system, clearly temperature alone cannot explain the significantly enhanced inac- illustrates the apparent disruption of various surface components, tivation. However, it is well known that microbial inactivation such as cell wall and membrane, and inner components of E. coli becomes faster at higher temperature, when chemical treatments by OH radicals formed by the photo ferrioxalate reaction. Consi- 12) such as OH radical, ozone, chlorine are applied. Whereas, as dering the high reactivity of the OH radical (oxidation potential: shown in Fig. 5 (-□-), E. coli was only slightly inactivated by 2.7 V) toward most E. coli components, it is plausible to assume solar irradiation alone (1 log inactivation after 52 min) without that it must undergo rapid reaction with the various cell compo- ferrioxalate. This level of inactivation is expected by solar irra- nents.26,27) diation itself.21) Therefore, the combined effect of the increased biocidal sensitivity at high temperature and the solar biocidal 4. Conclusion irradiation explains the superior performance of the solar ferrioxalate system over the photo ferrioxalate system shown in This study has confirmed the effective utilization of a photo Fig. 5. ferrioxalate system with artificial light (300-420 nm) for inactivating E. coli without an external hydrogen peroxide supply, 3.5. Role of Reactive Species in the Photo Ferrioxalate System as long as a level of dissolved oxygen is maintained. Even fas-

(a) (b) Fig. 6. Transmission electron microscopy images of E. coli inactivated with the photo ferrioxalate system: (a) before inactivation and (b) after one log inactivation. Photo or Solar Ferrioxalate Disinfection Technology without External Hydrogen Peroxide Supply 243 ter inactivation was observed in the solar ferrioxalate system. relation between inactivation of E. coli and OH radical These results support the potential of the photo or solar ferrio- concentration in TiO2 photocatalytic disinfection,” Water xalate system for disinfecting contaminated surface water or Research, 38, 1069-1077 (2004(b)). groundwater in remote rural places where sunlight is abundant. 13. Huang, Z., Maness, P. C., Blake, D. M., Wolfrum, E. J., Smolinski, S. L., and Jacoby, W. A., “Bactericidal mode of Acknowledgements titanium dioxide photocatalysis,” J. Photochem. Photobiol. A: Chem., 130, 163-170 (2000). This work was supported by the Brain Korea 21 Program in 14. Bader, H., Sturzenegger, V., and Hoigné, J., “Photometric the Ministry of Education in Korea and the Ministry of Environ- method for the determination of low concentrations of hyd- ment as “The Eco-technopia 21 project”. rogen peroxide by the peroxidase catalyzed oxidation of N,N-Diethyl-p-Phenylenediamine (DPD),” Water Research, References 22, 1109-1115 (1988). 15. Tamura, H., Goto, K., Yotosuyanagi, T., and Nagayama, 1. Cho, M., Lee, Y., Chung, H., and Yoon, J., “Inactivation of M., “Spectrophotometric determination of iron(III) with Escherichia coli by photochemical reaction of ferrioxalate 1,10-Phenanthroline in the presence of large amounts of at slightly acidic and near-neutral pHs,” Applied and Envi- iron(III),” Talanta, 21, 314-318 (1974). ronmental Microbiology, 70, 1129-1134 (2004(a)). 16. Eisenberg, G. M., “Colorimetric determination of hydrogen 2. Huston, P. L., and Pignatello, J. J., “Reduction of perchlo- peroxide,” Ind. Eng. Chem. Anal., 15, 327-328 (1943). roalkanes by ferrioxalate-generated carboxylate radical 17. Khadre, M. A., and Yousef, A. E., “Sporicidal action of preceding mineralization by the photo-Fenton reaction,” ozone and hydrogen peroxide: a comparative study,” Inter- Environmental Science and Technology, 30, 3457-3463 national Journal of Food Microbiology, 71, 131-138 (2001). (1996). 18. Hatchard, C. G., and Parker, C. A., “A new sensitive chemi- 3. Hislop, K. A., and Bolton, J. R., “The photochemical gene- cal actinometer. II. Potassium ferrioxalate as a standard chemi- ration of hydroxyl radicals in the UV-vis ferrioxalate. H2O2 cal actinometer,” Proc. Roy. Soc. Lon. A., 235, 518-536 (1956). system,” Environmental Science and Technology, 33, 19. Piechowski, M. V., Thelen, M. A., Hoigné, J., and Buler, R. 3119-3126 (1995). E., “tert-Butanol as an ·OH scavenger in the pulse radiolysis 4. Lee, Y., Jeong, J., Lee, C., and Yoon, J., “Influence of vari- of oxygenated aqueous systems,” Ber. Bunsenges. Phys. ous reaction parameters on 2,4-D removal in photo ferriox- Chem., 96, 1448-1454 (1992). alate/H2O2 process,” Chemosphere, 51, 901-912 (2003). 20. Joyce, T. M., McGuigan, K. G., Elmore-Meegan, M., and 5. Balzani, V., and Carassiti, V., Photochemistry of coordi- Conroy, R. M., “Inactivation of fecal bacteria in drinking nation compounds, Academic Press, London (1970). water by solar heating,” Applied and Environmental Mic- 6. Zepp, R. G., Faust, B. C., and Hoigné, J., “Hydroxyl radical robiology, 62, 399-402 (1996). formation in aqueous reactions (pH 3-8) of iron (II) with 21. Sinton, L. W., Finlay, R. K., and Lynch, P. A., “Sunlight hydrogen peroxide: the photo-Fenton reaction,” Environ- inactivation of fecal bacteriophages and bacteria in sewage- mental Science and Technology, 26, 313-319 (1992). polluted seawater,” Applied and Environmental Microbio- 7. Faust, B. C., A review of the photochemical redox reactions logy, 65, 3605-3613 (1999). of Iron(III) species in atmospheric, oceanic, and surface 22. Oates, P. M., Shanahan, P., and Polz, M. F., “Solar disinfec- waters : Influences on geochemical cycles and oxidant for- tion (SODIS): simulation of solar radiation for global asse- mation, In: Helz, G. R., Zepp, R. G., Crosby, D. G., Aquatic and ssment and application for point-of-use water treatment in Surface Photochemistry, Lewis, Boca Raton, FL, pp. 3-37 Haiti,” Water Research, 37, 47-54 (2003). (1994). 23. Young, S. B., and Setlow, P., “Mechanisms of killing of 8. Nogueira, R. F. P., and Guimaraes, J. R., “Photodegradation Bacillus subtilis spores by hypochlorite and chlorine dio- of dichloroacetic acid and 2,4-dichlorophenol by ferrioxalate/ xide,” J. Applied Microbiology, 95, 54-67 (2003). H2O2 system,” Water Research, 34, 895-901 (2000). 24. Haag, W. R., Hoigné, J., “Photo-sensitized oxidation in na- 9. Zuo, Y., and Hoigné, J., “Formation of hydrogen peroxide tural water via OH radicals,” Chemosphere, 14, 1659-1671 and depletion of oxalic acid in atmospheric water by pho- (1985). tolysis of iron(III)-oxalato complexes,” Environmental Sci- 25. von Sonntag, C., “Disinfection by free radicals and UV-ra- ence and Technology, 26, 1014-1022 (1992). diation,” Wat. Supply, 4, 11-18 (1986). 10. Acher, A. J., and Juven, B. I., “Destruction of coliforms in 26. Maness, P. C., Smolinski, S., Blake, D. M., Huang, Z., Wol- water and sewage water by dye-sensitized photo-oxidation,” frum, E. J., and Jacoby, W. A., “Bactericidal activity of Applied and Environmental Microbiology, 33, 1019-1022 photocatalytic TiO2 Reaction: toward an understanding of (1977). its killing mechanism,” Applied and Environmental Micro- 11. Safarzadeh-Amiri, A., Bolton, J. R., and Cater, S. R., “Fer- biology, 65, 4094-4098 (1999). rioxalate-mediated solar degradation of organic contami- 27. Sunada, K., Watanabe, T., and Hashimoto, K., “Studies on nants in Water,” Solar Energy, 56, 439-443 (1996). photokilling of bacteria on TiO2 thin film,” J. Photochem. 12. Cho, M., Chung, H., Choi, W., and Yoon, J., “Linear cor- Photobiol. A: Chem., 156, 227-233 (2003).