The occurrence of natural organic matter (NOM) in source waters presents a concern in water treatment plants due to formation of toxic disinfection byproducts (DBPs). In this doctoral thesis, the fate of NOM during ozonation, biofiltration, and chlorination was investigated to identify important aspects in these processes that can be manipulated for better DBP control. Specifically, this thesis studied (1) reaction mechanisms of ozone with dissolved organic nitrogen (DON), an important fraction of NOM that forms nitrogenous DBPs, (2) role of ozone and •OH-mediated attack on NOM and DBP formation during chlorination, and (3) impact of ozonation on biodegradability of DBP precursors.
The reaction of ozone with DON (Chapter 4) was observed to produce various transformation products including nitrate (NO3-) and ammonium (NH4+). This observation was shown in batch ozonation experiments involving NOM standards, surface water and wastewater effluent samples. A strong correlation was found between NO3- formation and O3 exposure (R2 > 0.82) during ozonation of both model DON solutions and real water samples. High NO3- yields were obtained for solutions containing primary amines such as glycine. Experiments with glycine showed that NO3- was formed via an intermediate with a second-order rate constant of 7.7 ± 0.1 M-1s-1 while NH4+ was formed by an electron-transfer mechanism with O3 as confirmed from a •OH yield of 24.7 ± 1.9%. The NH4+ concentrations, however, were lower than the •OH yield (0.03 mol NH4+/mol •OH) suggesting other •OH-producing reactions that compete with NH4+ formation. This study showed evidence that NO3- formation during ozonation of DON is induced by an oxygen-transfer to nitrogen forming hydroxylamine and oxime, while NH4+ formation is induced by electron-transfer reactions involving C-centered radicals and imine intermediates.
These reactions of ozone and other generated reactive oxygen species with NOM also affect NOM’s reactivity during post-chlorination, consequently affecting formation potentials of nitrogen-containing and carbon-based DBPs. Chapter 5 presents the effects of varying exposures of O3 and •OH on the resulting DBP concentrations and their associated toxicity generated after subsequent chlorination. DBP formation potential tests (target Cl2 residual after 24 h = 1 – 2 mg/L) and in vitro bioassays were conducted after ozonation of coagulated surface water at O3- and •OH-dominated conditions. Although ozonation led to a 24 – 37% decrease in formation of total trihalomethanes (THM4), haloacetic acids (HAA8), haloacetonitriles (HAN4), and trihaloacetamides (THAM), an increase in formation of total trihalonitromethanes (THNM2), chloral hydrate (CH), and haloketones (HK2) was observed. This effect however was less pronounced for samples ozonated at conditions favoring ozone (e.g., pH 6 and in the presence of t-BuOH) over •OH reactions (e.g., pH 8 and in the presence of H2O2). Compared to ozonation only, addition of H2O2 consistently enhanced formation of all DBP groups (20 – 61%) except trihalonitromethanes. This proves that •OH-transformed NOM is more susceptible to halogen incorporation. Analogously, adsorbable organic halogen (AOX) concentrations increased under conditions that favor •OH reactions. The ratio of unknown to known AOX, however, was greater at conditions that promote O3 reactions. Although significant correlation was found between AOX and genotoxicity with p53 bioassay, toxicity tests using 4 in vitro bioassays showed relatively small differences between various ozonation conditions.
Following ozonation, the biodegradability of DBP precursors was investigated in Chapter 6, with emphasis on two operational factors: ozone exposure and empty bed contact time (EBCT). Ozone exposure was varied through addition of H2O2 during ozonation at 1 mgO3/mgDOC followed by biological filtration using either activated carbon (BAC) or anthracite. Ozonation led to a 10% decrease in dissolved organic carbon (DOC), without further improvement from H2O2 addition. Compared to ozonation without H2O2, raising H2O2 concentrations to 2 mmol/mmolO3 resulted in increased DBP formation potentials during post-chlorination of the ozonated water (target Cl2 residual after 24 h = 1 – 2 mg/L) as follows: THM4 (37%), HAA8 (44%), CH (107%), HK2 (97%), HAN4 (33%), trichloroacetamide (TCAM, 43%), and AOX (27%), but a decrease in concentrations of THNM2 (43%). Coupling ozonation with biofiltration prior to chlorination effectively lowered the formation potentials of all DBPs including CH, HK2, and THNM2, all of which increased after ozonation. The dynamics of DBP formation potentials during BAC filtration at different EBCTs followed first-order reaction kinetics. Minimum steady-state concentrations were attained at an EBCT of 10 – 20 min, depending on the DBP species. The rate of reduction in DBP formation potentials varied among individual species before reaching their minimum concentrations. CH, HK2, and THNM2 had the highest rate constants of between 0.5 and 0.6 min-1 followed by HAN4 (0.4 min-1), THM4 (0.3 min-1), HAA8 (0.2 min-1), and AOX (0.1 min-1). Relative to concentrations after ozonation, the reduction in formation potential for most DBPs (e.g., at 15 min EBCT) was less than 50% but was higher than 70% for CH, HK2, and THNM2. The formation of bromine-containing DBPs increased with increasing EBCT likely due to an increase in Br-/DOC ratio.