Environ. Sci. Technol. 2003, 37, 2734-2742
Iron-Catalyzed Oxidation of Arsenic(III) by Oxygen and by Hydrogen Peroxide: pH-Dependent Formation of Oxidants in the Fenton Reaction STEPHAN J. HUG* AND OLIVIER LEUPIN Swiss Federal Institute for Environmental Science and Technology (EAWAG), U ¨ berlandstrasse 133, CH-8600 Du ¨ bendorf, Switzerland
The oxidation kinetics of As(III) with natural and technical oxidants is still not well understood, despite its importance in understanding the behavior of arsenic in the environment and in arsenic removal procedures. We have studied the oxidation of 6.6 µM As(III) by dissolved oxygen and hydrogen peroxide in the presence of Fe(II,III) at pH 3.5-7.5, on a time scale of hours. As(III) was not measurably oxidized by O2, 20-100 µM H2O2, dissolved Fe(III), or iron(III) (hydr)oxides as single oxidants, respectively. In contrast, As(III) was partially or completely oxidized in parallel to the oxidation of 20-90 µM Fe(II) by oxygen and by 20 µM H2O2 in aerated solutions. Addition of 2-propanol as an •OHradical scavenger quenched the As(III) oxidation at low pH but had little effect at neutral pH. High bicarbonate concentrations (100 mM) lead to increased oxidation of As(III). On the basis of these results, a reaction scheme is proposed in which H2O2 and Fe(II) form •OH radicals at low pH but a different oxidant, possibly an Fe(IV) species, at higher pH. With bicarbonate present, carbonate radicals might also be produced. The oxidant formed at neutral pH oxidizes As(III) and Fe(II) but does not react competitively with 2-propanol. Kinetic modeling of all data simultaneously explains the results quantitatively and provides estimates for reaction rate constants. The observation that As(III) is oxidized in parallel to the oxidation of Fe(II) by O2 and by H2O2 and that the As(III) oxidation is not inhibited by •OHradical scavengers at neutral pH is significant for the understanding of arsenic redox reactions in the environment and in arsenic removal processes as well as for the understanding of Fenton reactions in general.
Introduction With arsenic-contaminated groundwater affecting 30-50 million people in Bangladesh (1) and possibly up to10 million in Vietnam (2) and with the debate around lower limits in the United States (3), arsenic has been recognized as one of the most widespread and problematic water contaminants. The behavior of arsenic in the environment and in water treatment processes is mainly determined by its redox chemistry. As a water contaminant, As(III) is more problematic than As(V). Unlike the arsenate anions H2AsO4- and HAsO42-, the dominant As(III) species up to pH 8 is the * Corresponding author phone: +41-1-823-5454; fax: +41-1-8235210; e-mail:
[email protected]. 2734
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 12, 2003
nonionic H3AsO3, which does not adsorb as strongly to mineral surfaces (1) as As(V) (4). As(III) is thus more mobile in groundwaters and is also more difficult to remove in arsenic removal treatments (5, 6). Typically, preoxidation to As(V) is necessary in arsenic removal procedures (7). Although As(V) is thermodynamically favored under oxic conditions, As(III) is only slowly oxidized by dissolved O2 (e.g., with a half-life of around 9 d in air-saturated water with low iron contents and pH 7.6-8.5) (8). The As(III) oxidation by H2O2 is also slow at neutral and acidic pH, as only H2AsO3- and HAsO32but not H3AsO3 react with H2O2 (9). The half-life of As(III) in 1 mM H2O2 at pH 7.5 is 2.1 d. The influence of transition metals on the oxidation of As(III) with H2O2 has recently been studied (10), and it was found that Fe(II) and Fe(III) both increased the oxidation rates to a similar extent, but the effects of Fenton chemistry were not addressed. The iron-catalyzed oxidation of As(III) with O2 and H2O2 is important in both natural and technical systems. The connection between iron and arsenic redox reactions is one of the key factors in understanding the fate of arsenic in the environment. For example, arsenic in the iron crusts of recent sediments from Lake Baikal was found to be enriched by a factor of 58 above background levels (11). In water treatment, oxidation followed by coprecipitation with iron (hydr)oxides is one of the most economic and efficient arsenic removal options (12). The oxidation of As(III) by Fe(II) and H2O2 (Fenton reaction) at pH 1-3 has been studied as early as 1963 (13) and has been tested for the treatment of groundwater with high arsenic concentrations at pH 2.5-3.0 (14). Treatment of As(III)-spiked groundwater by a combination of Fenton chemistry and filtration through scrap iron at neutral pH has recently been reported (15), but kinetic and mechanistic issues were not discussed. In a recent study of iron-catalyzed photochemical oxidation and removal of arsenic at neutral pH (16), we found that arsenic was partly oxidized in parallel to the dark oxidation of Fe(II) by dissolved O2. We have attempted to explain these observations with the formation of HO2•/•O2-, H2O2, and •OH radicals in the generally stated sequence of Fe(II,III) oxidation/reduction reactions and Fenton chemistry (17-21). However, the involvement of free •OH radicals was not consistent with quenching experiments in which •OH scavengers such as 2-propanol and 3-butanol had almost no effect on the oxidation of As(III) at neutral pH (16). An important role of •O2- as an oxidant for As(III) was also excluded on the basis of the results of the dark and photochemical reactions. A number of newer studies have suggested that oxidizing species different from free •OH radicals (e.g., Fe(IV) species) are formed in the Fenton reaction (22-25) and in the oxidation of Fe(II) with O2 (26). Fe(IV) intermediates in iron porhyrins have been isolated and studied by low-temperature spectroscopy (27). Spectroscopic evidence for a ferryl species and an Fe(III) dimer formed from ferryl and Fe(II) was also found in the reaction of uncomplexed Fe(II) with O3 at pH below 2 (28) and in a pulse radiolysis study (29). Rate constants for the reaction of ferryl with various compounds at low pH were reported (30). Furthermore, Fe(IV) species were suggested in the reaction of Fe(II)-NTA and Fe(II)EDDA (31) and of Fe(II)-DTPA (32) with H2O2 at neutral pH. The relative contribution of reaction pathways leading to Fe(IV) and •OH, respectively, seem to depend on various parameters, such as pH and the concentrations of organic and inorganic ligands (33-35). In the degradation of pollutants, it is generally observed that the efficiency of dark and photo-Fenton systems decreases with increasing pH (36). For example, atrazine 10.1021/es026208x CCC: $25.00
2003 American Chemical Society Published on Web 05/17/2003
TABLE 1. Experimental Parameters (Initial Concentrations after Mixing)a species (concn)
oxidation with O2
oxidation with oxidation with H2O2 at >pH 6 H2O2 at FeIIOH- > FeII(CO3)22- at pH 5.5-7.3 and FeIICO3 > FeII(CO3)22- > FeIIOH- at pH >7.3. In our model, we have included only the FeIICO3, as each additional species in the model also requires the addition of the corresponding reactions with Fe(IV) and with •O2-. The fitted log(k) for the reaction of H2O2 with FeIICO3 was between the values reported for FeIICO3 and for FeII(CO3)22- (see Table 3). For the reaction of FeIICO3 with H2O2, we propose the formation of an intermediate INT-CO3, which can form either carbonate radicals, Fe(IV), or both. (As the pKa of the bicarbonate/carbonate radical is reported as 1 × 107 M-1 s-1 (58). The description of Fe(III) removal by precipitation is important, as it affects the reduction of dissolved Fe(III) species by •O2- (R1-R3) and with it the fits in Figures 3 and 4. Following the suggestion by Sedlak and Hoigne (44), we have used the same rate constants for the reaction of FeIII(OH)2+ and FeIII(OH)2+ with •O2-. Assuming that only FeIII(OH)2+ reacts with •O2- still leads to good fits in Figures 1 and 2 but to unsatisfactory fits of the data in Figures 3 and 4. The role of Fe(III) speciation and precipitation in Fenton systems at neutral pH is still difficult to describe and will require further attention. The reaction scheme in models 2a-2c is identical, but the rate constants for Fe(II) oxidation are different. In model 2b, all rate constants for the Fe(II) oxidation were left adjustable. This increased flexibility slightly improved the fits, but the output of model 2a (not shown) is almost indistinguishable from the output of model 2b (shown in the figures). Model 2c, with the rate constants for Fe(II) oxidation fixed to the values according to Millero et al., results in a somewhat faster oxidation of As(III) than observed in Figure 2B-D, but the fits for all other figures were the same as those of model 2b. Various studies have found slightly different rate constants for the oxidation of Fe2+ and FeIIOH+ with H2O2 (see last column in Table 3). Given the experimental uncertainties (e.g., 0.1-0.2 pH unit increase during the reactions and ∼10% relative errors in the As(III) determinations) and the simplification in the model (e.g., only one carbonate species), the log(k) values for the oxidation of Fe(II) species are in good agreement with the values reported by others. The most important part of the model, the pHdependent partitioning of the Fenton reaction into •OH and another oxidizing species (suggested to be Fe(IV)) was rather invariant to differences in the other rate constants or to the exact description of Fe(III) precipitation. Finally, it should be noted that the proposed model does not mean that no •OH radicals are formed at neutral pH. Rather, the model states that the yield of •OH formation drops sharply between pH 5 and pH 6 and 10-fold with each pH unit increase above pH 6. Compounds that only react with •OH can still be oxidized in the Fenton reaction at pH 7, although with low yields, and the addition of 2-propanol or other •OH scavengers inhibits their oxidation. Environmental and Technical Significance. The finding that As(III) is partly oxidized in parallel to the oxidation of Fe(II) is significant in all situations where oxygen is introduced into circumneutral waters containing As(III) and Fe(II). The co-oxidation of As(III) with Fe(II) by O2 might be one of the most important abiotic pathways for As(III) oxidation and might help to explain why both As(III) and As(V) are typically found in groundwater samples after they have been pumped to the surface. The oxidation of As(III) with H2O2 and Fe(II) is important in natural and technical systems at pH values up to 8, below which As(III) is not oxidized at appreciable rates by H2O2 alone. H2O2 can occur naturally at micromolar concentrations in rainwater and in soil waters, and it is used at millimolar concentrations in water treatment. The observation that the As(III) oxidation at circumneutral pH is not quenched by typical •OH scavengers such as 2-propanol or formate (in contrast to the almost complete quenching at low pH) is VOL. 37, NO. 12, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2741
important in natural and technical systems as it can be expected that naturally present dissolved organic matter would also not significantly affect As(III) oxidation by scavenging •OH. Since the rate and extent of the As(III) oxidation strongly depends on pH and As(III), Fe(II), O2, H2O2, and HCO3- in concentrations ranges that are common in natural and technical systems, a good quantitative understanding of these reactions is important. The proposed kinetic model provides a base for the understanding of the important factors in these systems and for the optimization of arsenic removal methods.
Acknowledgments We thank Thomas Ruettimann (EAWAG) for technical assistance; Bernhard Wehrli, Silvio Canonica, Urs von Gunten, and Barbara Sulzberger (EAWAG) for valuable discussions; and the Alliance for Global Sustainability for financial support in the framework of the arsenic mitigation project in Southeast Asia.
Supporting Information Available The input file for ACUCHEM with the complete reaction scheme of model 2b. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Kinniburgh, D. G.; Smedley, P. L. Arsenic contamination of groundwater in Bangladesh; Final Report Summary; Bangladesh Department for Public Health Engineering, British Geological Survey: 2000. (2) Berg, M.; Tran, H. C.; Nguyen, T. C.; Pham, H. V.; Schertenleib, R.; Giger, W. Environ. Sci. Technol. 2001, 35, 2621-2626. (3) National Research Council. Arsenic in drinking water; National Academy Press: Washington, DC, 1999. (4) Manning, B. A.; Goldberg, S. Soil Sci. 1997, 162 (12), 886-895. (5) Hering, J. G.; Chen, P. Y.; Wilkie, J. A.; Elimelech, M.; Liang, S. J. Am. Water Work Assoc. 1996, 88, 155-167. (6) Meng, X. G.; Bang, S.; Korfiatis, G. P. Water Res. 2000, 34 (4), 1255-1261. (7) Borho, M.; Wilderer, P. Water Sci. Technol. 1996, 34, 25-31. (8) Kim, M. J.; Nriagu, J. Sci. Total Environ. 2000, 247, 71-79. (9) Pettine, M.; Campanella, L.; Millero, F. J. Geochim. Cosmochim. Acta 1999, 63, 2727-2735. (10) Pettine, M.; Millero, F. J. Mar. Chem. 2000, 70, 223-234. (11) Muller, B.; Granina, L.; Schaller, T.; Ulrich, A.; Wehrli, B. Environ. Sci. Technol. 2002, 36, 411-420. (12) United States Environmental Protection Agency, Office of Water (4606). EPA 815-R-00-028, 2000; Vol. 2000. http://www.epa.gov/ safewater/ars/treatments_and_costs.pdf. (13) Woods, R.; Kolthoff, I. M.; Meehan, E. J. J. Am. Chem. Soc. 1964, 86, 1698. (14) Banerjee, K.; Helwick, R. P.; Gupta, S. Environ. Prog. 1999, 18, 280-284. (15) Krishna, B. M. V.; Chandrasekaran, K.; Karunasagar, D.; Arunachalam, J. J. Hazard. Mater. 2001, B84, 229-240. (16) Hug, S. J.; Canonica, L.; Wegelin, M.; Gechter, D.; von Gunten, U. Environ. Sci. Technol. 2001, 35, 2114-2121. (17) Haber, F.; Weiss, J. Proc. R. Soc. London, Ser. A 1934, 332-351. (18) Barb, W. G.; Boxendale, J. H.; George, P.; Hargrove, K. R. Trans. Faraday Soc. 1951, 47, 591-616. (19) Rush, J. D.; Bielski, B. H. J. J. Phys. Chem. 1985, 89, 5062-5066. (20) Koppenol, W. H. Redox Rep. 2001, 6, 229-234. (21) Liochev, S. I.; Fridovich, I. Redox Rep. 2002, 7, 55-57. (22) Wink, D. A.; Nims, R. W.; Desrosiers, M. F.; Ford, P. C.; Keefer, L. K. Chem. Res. Toxicol. 1991, 4, 510-512. (23) Bossmann, S. H.; Oliveros, E.; Go¨b, S.; Siegwart, S.; Dahlen, E. P.; Payawan, L., Jr.; Straub, M.; Wo¨rner, M.; Braun, A. M. J. Phys. Chem. A 1998, 102, 5542-5550. (24) Kremer, M. L. Phys. Chem. Chem. Phys. 1999, 1, 3595-3605. (25) Pignatello, J. J.; Liu, D.; Huston, P. Environ. Sci. Technol. 1999, 33, 1832-1839. (26) Reinke, L. A.; Rau, J. M.; McCay, P. B. Free Radical Biol. Med. 1994, 16, 485-492.
2742
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 12, 2003
(27) Nakamoto, K. J. Mol. Struct. 1997, 408/409, 11-16. (28) Logager, T.; Holcman, J.; Sehested, K.; Pedersen, T. Inorg. Chem. 1992, 31, 3523-3529. (29) Rush, J. D.; Bielski, B. H. J. J. Am. Chem. Soc. 1986, 108, 523525. (30) Jacobsen, F.; Holcman, J.; Sehested, K. Int. J. Chem. Kinet. 1998, 30, 215-221. (31) Rush, J. D.; Koppenol, W. H. J. Am. Chem. Soc. 1988, 110, 49574963. (32) Rahhal, S.; Richter, H. W. J. Am. Chem. Soc. 1988, 110, 31263133. (33) Luzzatto, E.; Cohen, H.; Stockheim, C.; Wieghardt, K.; Meyerstein, D. Free Radical Res. 1995, 23, 453-463. (34) Goldstein, S.; Meyerstein, D.; Czapski, G. Free Radical Biol. Med. 1993, 15, 435-445. (35) Goldstein, S.; Meyerstein, D. Acc. Chem. Res. 1999, 32, 547-550. (36) Aplin, R.; Waite, T. D. Water Sci. Technol. 2000, 42, 345-354. (37) Gallard, H.; de Laat, J.; Legube, B. New J. Chem. 1998, 22, 263268. (38) Balmer, M.; Sulzberger, B. Environ. Sci. Technol. 1999, 33, 24182424. (39) Haywood, M. G.; Riley, J. P. Anal. Chim. Acta 1976, 85, 219-230. (40) Johnson, D. L.; Pilson, M. E. Q. Anal. Chim. Acta 1972, 58, 289299. (41) Yamamoto, M.; Urata, K.; Murashige, K.; Yamamoto, Y. Spectrochim. Acta Part B 1981, 36, 671-677. (42) Zwank, L.; Schmidt, T. C.; Haderlein, S. B.; Berg, M. Environ. Sci. Technol. 2002, 36, 2054-2059. (43) Braun, W.; Herron, J. T.; Kahaner, D. ACUCHEM, Computer Program for Modeling Complex Reaction Systems; National Bureau of Standards: Gaithersburg, MD. (44) Sedlak, D. L.; Hoigne´, J. Environ. Sci. Technol. 1994, 28, 18981906. (45) Voelker, B. M.; Morel, F. M. M.; Sulzberger, B. Environ. Sci. Technol. 1997, 31, 1004-1011. (46) Hug, S. J.; Laubscher, H.-U.; James, B. R. Environ. Sci. Technol. 1997, 31, 160-170. (47) Millero, F. J.; Sotolongo, S.; Izaguirre, M. Geochim. Cosmochim. Acta 1987, 51, 793-801. (48) Asmus, K. D.; Mockel, H.; Henglein, A. J. Phys. Chem. 1973, 77, 1218-1221. (49) Klaning, U. K.; Bielski, B. H. J.; Sehested, K. Inorg. Chem. 1989, 28, 2717-2724. (50) Cornell, R. M.; Schwertmann, U. The Iron Oxides; VCH Verlagsgesellschaft: Weinheim, Germany, 1996. (51) Voegelin, A.; Hug, S. J. Environ. Sci. Technol. 2003, 37, 972-978. (52) The Radiation Chemistry Data Center of the Notre Dame Radiation Laboratory, 2002; http://allen.rad.nd.edu/. (53) Buda, F.; Ensing, B.; Gribnau, M. C. M.; Baerends, E. J. Chem.Eur. J. 2001, 7, 2775-2783. (54) Stumm, W.; Lee, G. F. Ind. Eng. Chem. 1961, 53, 143-146. (55) Ensing, B.; Buda, F.; Blochl, P.; Baerends, E. J. Angew. Chem. Int. Ed. 2001, 40, 2893-2895. (56) King, D. W.; Farlow, R. Mar. Chem. 2000, 70, 201-209. (57) Czapski, G.; Lymar, S. V.; Schwarz, H. A. J. Phys. Chem. A 1999, 103, 3447-3450. (58) Schneider, W. Chimia 1988, 42, 9-20. (59) Stuglik, Z.; Zagorski, Z. P. Radiat. Phys. Chem. 1981, 17, 229233. (60) Huie, R. E.; Shoute, L. C. T.; Neta, P. Int. J. Chem. Kinet. 1991, 23 (6), 541-552. (61) Millero, F. J.; Sotolongo, S. Geochim. Cosmochim. Acta 1989, 53, 1867-1873. (62) Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. J. Phys. Chem. Ref. Data 1988, 17, 513-886. (63) Draganic, Z. D.; Negron-Mendoza, A. S., K.; Vujosevic, S. I.; Navarro-Gonzales, R.; Albarran-Sanchez, M. G.; Draganic, I. G. Radiat. Phys. Chem. 1991, 38 (3), 317-321. (64) Lilie, J.; Hanrahan, R. J.; Henglein, A. Radiat. Phys. Chem. 1978, 11, 225-227. (65) Smith, R. M., Martell, A. E., Eds. Critical Stability Constants; Plenum Press: New York, 1976.
Received for review October 3, 2002. Revised manuscript received April 2, 2003. Accepted April 7, 2003. ES026208X
