Modifications to Dissolved Organic Matter and Disinfection Byproduct Formation during Solar Chlorine Photolysis and Related Oxidative Treatment, as Applied to Drinking Water Treatment

Thesis (Ph.D.)--University of Washington, 2019Chlorine is a common disinfectant used for water treatment worldwide, however not all pathogens can be inactivated using chlorine alone. Solar irradiation of chlorinated waters enhances inactivation of chlorine-resistant pathogens, through in situ formation of ozone and hydroxyl radical during photolysis of free available chlorine (FAC) at UVA/UVB wavelengths of sunlight (290 – 400nm). Organic and inorganic water constituents (e.g., natural organic matter (NOM), bromide, and carbonate) are also reactive with these photooxidants and can be linked to the formation of regulated drinking water disinfection byproducts (DBP). To evaluate the feasibility of this process as a novel water treatment, a variety of relevant surface water matrixes need to be treated via solar chlorine photolysis while monitoring consequent DBP. A suite of relevant chlorinated and brominated disinfection/oxidation byproducts (some regulated by the U.S. EPA) were monitored including oxyhalides, four trihalomethanes (THMs), and nine haloacetic acids (HAA9). Changes to NOM were characterized by monitoring spectral and redox properties, and molecular weight distributions during solar chlorine photolysis and compared to changes associated with conventional drinking water treatment (e.g., chlorination, ozonation, and advanced oxidation processes (AOP)). The results from this study show formation of DBPs vary significantly with water composition. Chlorate concentrations increased during solar photolysis of FAC (8mg/L as Cl2) at pH 8, and bromate formation followed similar trends (1.8à increase at pH 8 vs. pH 6) with 200 g/L bromide. Addition of NOM (Suwannee River) adds a competitor for radical and ozone species, and completely disrupted chlorate and bromate formation mechanisms; less reactive, natural waters limited oxyhalide formation to a lesser degree. Natural organic matter contributes to THM and HAA formation during chlorination and organic DBP yields increased further following solar chlorine photolysis (solar fluences > 4 J/cm2). Control experiments showed that increased organic DBP levels were not due to direct SRNOM photolysis and subsequent dark reactions with HOCl, but to co-exposure of SRNOM to HOCl and reactive species (e.g., O3, HO•, Cl•, ClO•) generated by FAC photolysis. ABTS·– was utilized to quantify changes in the electron donating capacity (EDC) of natural organic matter (NOM) isolates and two model compounds: phenol a model for reactive aromatic moieties, and benzoic acid a model for less reactive aromatic moieties. Low exposures of FAC (CTFAC = ∫[FAC]dt) significantly decreased the measurable EDC of NOM and phenol, potentially due to rapid halogenation or oxidation of reactive aromatic groups. Once these fast-reacting sites were depleted, the EDC decreased slowly with increasing chlorine exposure. Chlorination resulted in moderate decreases in UV absorbance of the bulk NOM and negligible changes in fluorescence EEM intensity (CTFAC ≤ 400 (mg/L as Cl2)à min). Reactive oxygen species such as ozone and hydroxyl radical (independently or in combination with FAC during solar chlorine photolysis) yielded loss of UV absorbance in the bulk NOM. Pre-treatment of DOM with low exposures of hydroxyl radical (UV/H2O2, X-ray radiolysis, or O2–/O3) led to increased EDC for NOM and benzoic acid, and increased THM and HAA formation after dark chlorination for NOM, benzoic acid, and phenol, likely due to hydroxylation of aromatics during pretreatment. Radical scavengers (50mM tert-butanol) suppressed EDC formation and hindered THM and HAA formation. Exposure to O3 itself (with radical scavengers added) actually appears to lead primarily to oxidation and/or ring cleavage, rather than hydroxylation which appears to occur due to HO· generated via O3 decay. X-ray irradiation of hypochlorite at pH 10 (favoring ClO· from reaction between HO• and OCl–) decreased EDC and spectral properties while increasing DBP yields, similar to sunlight-driven chlorine photolysis. Based on these findings, exposure of NOM to ozone, reactive oxygen and halogen species (e.g., HO•, Cl•, ClO•), and FAC during sunlight-driven chlorine photolysis involves hydroxylation of slow reacting aromatic moieties as a key pathway contributing to (a) increased electron donating capacity in aromatic components of the bulk NOM, and (b) increased reactivities toward FAC (and potentially other halogenating agents) to form halogenated organic DBPs. These findings provide an improved understanding of the processes leading to increased formation of DBP precursors during low to moderate exposures of ozone and reactive oxygen or halogen species, and could help identify strategies to minimize organic DBP formation during solar chlorine photolysis or other combinations of chlorine and ROS (such as UVC chlorine photolysis or sequential AOP/FAC treatment) likely to be applied during water and wastewater treatment