Article pubs.acs.org/est
Comparison of Halide Impacts on the Efficiency of Contaminant Degradation by Sulfate and Hydroxyl Radical-Based Advanced Oxidation Processes (AOPs) Yi Yang,† Joseph J. Pignatello,‡ Jun Ma,*,† and William A. Mitch*,§ †
State Key Laboratory of Urban Water Resource and Environment, Harbin Institute of Technology, Harbin, 150090, China Department of Environmental Sciences, Connecticut Agricultural Experiment Station, 123 Huntington Street, P.O. Box 1106, New Haven, Connecticut 06504-1106 § Department of Civil and Environmental Engineering, Stanford University, Stanford, California 94305 ‡
S Supporting Information *
ABSTRACT: The effect of halides on organic contaminant destruction efficiency was compared for UV/H2O2 and UV/S2O82− AOP treatments of saline waters; benzoic acid, 3-cyclohexene-1-carboxylic acid, and cyclohexanecarboxylic acid were used as models for aromatic, alkene, and alkane constituents of naphthenic acids in oil-field waters. In model freshwater, contaminant degradation was higher by UV/ S2O82− because of the higher quantum efficiency for S2O82− than H2O2 photolysis. The conversion of •OH and SO4•− radicals to less reactive halogen radicals in the presence of seawater halides reduced the degradation efficiency of benzoic acid and cyclohexanecarboxylic acid. The UV/S2O82− AOP was more affected by Cl− than the UV/H2O2 AOP because oxidation of Cl− is more favorable by SO4•− than •OH at pH 7. Degradation of 3-cyclohexene-1-carboxylic acid, was not affected by halides, likely because of the high reactivity of halogen radicals with alkenes. Despite its relatively low concentration in saline waters compared to Cl−, Br− was particularly important. Br− promoted halogen radical formation for both AOPs resulting in ClBr•−, Br2•−, and CO3•− concentrations orders of magnitude higher than •OH and SO4•− concentrations and reducing differences in halide impacts between the two AOPs. Kinetic modeling of the UV/H2O2 AOP indicated a synergism between Br− and Cl−, with Br− scavenging of •OH leading to BrOH•−, and further reactions of Cl− with this and other brominated radicals promoting halogen radical concentrations. In contaminant mixtures, the conversion of •OH and SO4•− radicals to more selective CO3•− and halogen radicals favored attack on highly reactive reaction centers represented by the alkene group of 3-cyclohexene-1-carboxylic acid and the aromatic group of the model compound, 2,4-dihydroxybenzoic acid, at the expense of less reactive reaction centers such as aromatic rings and alkane groups represented in benzoic acid and cyclohexanecarboxylic acid. This effect was more pronounced for the UV/S2O82− AOP.
■
INTRODUCTION There is an increasing need for techniques to remove organic contaminants from brackish industrial and municipal wastewaters. Brackish industrial wastewaters include produced waters from oilfield operations (i.e., the saline wastewaters removed from oil reservoirs along with petroleum1−4) and energy extraction-related wastewaters associated with hydraulic-fracturing operations. Such wastewaters can feature salinities 8-fold higher than seawater.5,6 In addition, reverse osmosis membrane treatment often employed for indirect potable reuse of municipal wastewaters generates a significant volume of brine containing the salts and organic contaminants rejected by the membranes. In many situations, desalination of the wastewaters may not be required before discharge to surface receiving waters. However, the organic contaminants in these waters may need to be detoxified to protect organisms in receiving waters and downstream water supplies.2,3,7 Additionally, brackish waters can serve as a valuable water supply in the arid regions in which Americans are increasingly settling. The U.S. Bureau © 2014 American Chemical Society
of Reclamation’s Desalination and Water Purif ication Technology Roadmap identified several nonpotable beneficial reuse options for brackish wastewaters (e.g., brackish water aquaculture) but indicated that the development of technologies for organic contaminant destruction in these waters is needed to meet future water demands.8 Broad-spectrum destruction technologies are desirable to handle the array of organic contaminants that may occur in these waters. Hydroxyl radical-based advanced oxidation processes (AOPs), including the ultraviolet photolysis of hydrogen peroxide (UV/H2O2), have been widely used for broad-spectrum removal of organic contaminants from freshwaters, because hydroxyl radicals (•OH) react with many organic chemicals at near diffusion-controlled rates.9 Sulfate Received: Revised: Accepted: Published: 2344
September 18, 2013 January 29, 2014 January 30, 2014 January 30, 2014 dx.doi.org/10.1021/es404118q | Environ. Sci. Technol. 2014, 48, 2344−2351
Environmental Science & Technology
Article
radical-based AOPs, including the ultraviolet photolysis of peroxodisulfate (UV/S2O82−), have recently received attention for freshwaters. The sulfate radical (SO4•−) also reacts with a wide range of organic contaminants with near diffusion-limited rate constants, because of its high reduction potential (2.5−3.1 V10), even compared with that of •OH (2.7 V in acid solution and 1.8 V in neutral solution9). Sulfate radical-based AOPs have been shown to effectively remove aromatic components in produced waters.11 In general, SO4•− is more prone to oneelectron oxidation reactions than •OH, expanding the range of contaminant transformation pathways; for example, SO4•− promotes the decarboxylation of carboxylic acids.12,13 However, little research has addressed the application of these AOPs to brackish waters. Although both the UV/H2O2 and UV/S2O82− AOPs have been applied to produced waters from oil sands petroleum extraction operations, the impacts of salts were not systematically evaluated.2,4 Halide ions, especially Br−, are known to be the most important •OH scavengers in seawater.14 Although one might anticipate that significant scavenging of •OH by Br− would render AOP treatment inefficient, this scavenging forms radical (especially Br•, Br2•−, ClBr•−) and nonradical (HOX, X2, and X3−) Reactive Halogen Species (RHS) that are also capable of oxidizing organics. Previous studies on the impact of halides on AOP treatment have focused on the effect of Cl−, but found conflicting results. Chloride inhibited destruction of organics in various •OHbased AOPs, predominantly at low pH.15−20 While Cl− inhibited the destruction of polychlorinated biphenyls by an SO4•−-based AOP,21,22 it promoted the degradation of 2,4dichlorophenol.23 The formation of trichlorinated products indicated the activity of halogenated oxidants.23 Grebel et al. evaluated the role of Cl− and Br− for contaminant destruction in a UV/H2 O 2 -based AOP.24 Modeling of •OH and RHS concentrations indicated that radical RHS concentrations (primarily ClBr•−, and Br2•−) may exceed that of •OH by several orders of magnitude during AOP treatment of brackish waters. Bromide was found to be particularly important for promoting RHS formation, despite its occurrence at far lower concentrations than Cl− in most saline waters (e.g., 675-fold lower in seawater). In the absence of Br−, formation of Cl2•− from •OH scavenging by Cl− (eqs 1−3) is significant only at low pH (eq 2). However, in the presence of Br−, formation of ClBr•−, and Br2•− is facilitated by reaction of the intermediate XOH•− with X− (eq 4). Because ClBr•− and Br2•− are more selective oxidants than •OH,9,10 the presence of halides reduced the efficiency with which electron-poor model organic compounds were transformed, but only slight reductions in treatment efficiency were observed for electronrich target contaminants (e.g., 17β-estradiol, which contains an electron-rich aromatic ring). Interestingly, in the presence of natural organic matter (NOM), the presence of seawater levels of halides increased the removal rate of the electron-rich model target compound, resorcinol. The increase in removal rate was attributed to the conversion of nonselective •OH to selective RHS, that would more efficiently target the electron-rich target compound, rather than the electron-poor functional groups within NOM. While the formation of brominated phenols from AOP treatment of phenol in the presence of seawater levels of halides indicated the active participation of RHS in contaminant destruction, yields of halogenated products were very low. •
OH + X− ↔ XOH•−
XOH•− + H+ ↔ X• + H 2O
(2)
X• + X− ↔ X 2•−
(3)
XOH•− + X− ↔ X 2•− + OH−
(4)
SO4•− + X− ↔ X• + SO4 2 −
(5)
The purpose of this study is to compare the effects of Cl− and Br− on SO4•−-based compared to •OH-based AOP treatment of contaminants. It was hypothesized that SO4•−based systems would be more sensitive to Cl− because oxidation of Cl− may lead directly to X• (eqs 5 and 3 versus eqs 1−4 for HO•−-based AOPs). Given the recent interest in reclamation of produced waters and hydraulic fracturing wastewaters, benzoic acid (BA), 3-cyclohexene-1-carboxylic acid (3CCA) and cyclohexanecarboxylic acid (CCA), were selected as model analogues for aromatic, alkene, and alkane naphthenic acid contaminants of concern in produced waters. Hydraulic fracturing wastewaters represent a particularly interesting case, as the water released upon depressurization of wells (flowback water) can exhibit salinities far higher than seawater, and Br−/Cl− ratios nearly five times higher than in seawater (Table SI-1 in the Supporting Information).5,6
■
MATERIALS AND METHODS Materials. Sigma-Aldrich BioXtral NaCl contained 0.0038 mol percent Br− 22 and was reported to contain ≤0.001% I−. Sigma benzoic acid (BA; 99.5+%), 3-cyclohexene-1-carboxylic acid (3CCA; 97%), cyclohexanecarboxylic acid (CCA; 98%), 2,4-dihydroxybenzoic acid (2,4-DHBA; 97%), potassium peroxodisulfate (K2S2O8; 99+%), ammonium molybdate tetrahydrate, EMD sodium perchlorate (98+%), Baker sodium phosphate monobasic monohydrate, sodium phosphate dibasic, sodium bicarbonate, sodium hydroxide (98.5%), hydrogen peroxide (H2O2) solution (30%), methanol, and acetic acid were used as received without purification. H2O2 and K2S2O8 stock solutions were standardized spectrophotometrically based on their molar absorption coefficients: ε = 40 M−1 cm−1 at 240 nm for H2O225 and ε = 20 M−1 cm−1 at 254 nm for peroxodisulfate.26 Experimental Procedures. UV irradiation was conducted with a semicollimated beam system consisting of 4 × 15 W low pressure mercury lamps emitting predominantly at 254 nm, as described previously.24 The light proceeded through a 12.5 cm diameter aperture down onto 500 mL samples stirred within an open-top cylindrical crystallization dish (12.5 cm diameter × 6.5 cm depth) nearly 0.3 m below. A shutter was used to open or close the aperture. The surface irradiance (1.68 × 10−7 Einsteins L−1 s−1) was determined by iodide-iodate actinometry.27 Except where mentioned, experimental solutions contained 1 mM H2O2 or S2O82− and organic target compounds at 10−100 μM in deionized water buffered at pH 7 with 10 mM phosphate buffer. The objective was to compare the effects of salts on H2O2 or S2O82−-based AOPs rather than to evaluate optimized degradation rates for these model compounds. In all cases, the pH remained within ±0.1 pH unit of the target pH over the course of the experiments. In addition to this simulated freshwater matrix, simulated saline water matrices included a saline water containing seawater levels of halides and carbonates (540 mM NaCl, 0.8 mM NaBr, and 2.3 mM NaHCO3), an ionic strength control with 540 mM sodium perchlorate, and a saline water with salt concentrations
(1) 2345
dx.doi.org/10.1021/es404118q | Environ. Sci. Technol. 2014, 48, 2344−2351
Environmental Science & Technology
Article
concentrations. While model organic precursors and their products would influence the absolute concentrations via scavenging, the kinetic model served to clarify the effects of inorganic constituents on radical concentrations. Equation 8 describes the formation rate of either •OH or SO4•− because of the photolysis of either H2O2 or S2O82−, respectively.
near the high end of those observed in hydraulic fracturing wastewaters (1 M NaCl, 10 mM NaBr, and 2.3 mM NaHCO3). H2O2 or K2S2O8 was added to the stirred solution immediately before irradiation. All of the experiments were conducted in duplicate or triplicate at ambient temperature (20 ± 2 °C). Samples were periodically withdrawn and supplemented with 20 μL methanol per mL sample to quench any radicals formed by thermolysis of H2O2 or K2S2O8. No degradation of target compounds was observed in dark controls containing H2O2 or K2S2O8. Generally, BA, 3CCA, CCA, and 2,4-DHBA were analyzed by HPLC with UV detection. However, for analysis of the mixture of all three compounds, these compounds were analyzed by HPLC-MS. The concentrations of H2O2 and K2S2O8 were measured by the iodometric method. Analytical details are provided in the Supporting Information. The degradation of a target contaminant in UV/H2O2 and UV/S2O82− AOPs can be described using eqs 6 and 7, respectively, where kd is a first order rate constant for direct photolysis of the target contaminant (s−1), k•OH and kSO4•− are second-order rate constants for the reaction of the target contaminant with •OH and SO4•−, respectively, and kother represents second order rate constants for reactions of the target organic with other radicals (e.g., CO3•− or, in saline waters, Br2•−, Cl2•−, and ClBr•−). Control experiments in the absence of H2O2 and K2S2O8 indicated no significant loss of any target contaminant, indicating that direct photolysis was negligible. The concentrations of H2O2 or K2S2O8 declined
