Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Environmental Engineering and Earth Sciences

Committee Member

Tanju Karanfil, Ph.D., P.E., BCEE, Committee Chair

Committee Member

Cindy Lee, Ph.D.

Committee Member

David A. Ladner, Ph.D.

Committee Member

Brian Powell, Ph.D.


Disinfection is a process where a significant percentage of pathogenic microorganisms are killed or controlled. However, an unintended consequence is the formation of disinfection by-products (DBPs) as a result of reaction between different precursors and oxidants. Due to their various health effects, the US Environmental Protection Agency (USEPA) has been imposing increasingly stringent regulations for DBPs under the Disinfectants/DBP Rule (D/DBPR). Although water utilities have been changing their use of disinfectants (e.g., chloramine, chlorine dioxide, and ozone) to reduce the formation of regulated THMs (trihalomethanes) and HAAs (haloacetic acids), these practices may also cause the formation of unregulated DBPs (more cytotoxic and genotoxic than regulated DBPs) such as NDMA (N-nitrosodimethylamine) and HNMs (halonitromethanes) which are observed during chloramination and ozonation-chlorination, respectively. While NDMA is a probable human carcinogen at concentrations as low as 0.2 ng/L, HNMs are one of the most cyto- and geno-toxic classes among the emerging DBPs, with orders of magnitude higher cyto- and geno-toxicity than any of the regulated organic DBPs. Therefore, the best approach to control the formation of DBPs during water treatment is to maximize the removal of their precursors prior to oxidant addition. Although there are studies in the literature regarding the formation and control of DBPs during water treatment, these studies are limited to their own suite of experimental conditions. Therefore, this study was designed with two main foci: i) the control of NDMA and HNM precursors during nanofiltration and the effect of water chemistry on precursor control and ii) the formation and control of halogenated DBPs, especially brominated and iodinated species of HAAs, I-THMs and HANs, examined under different water chemistry composition to gain insights to their formation and control. The initial experimental results showed that NDMA precursors leach from virgin membranes. Therefore, the first main objective of this research was divided into two sub-objectives. The main goal of the first sub-objective was to investigate i) the potential leaching of NDMA precursors from different types of nanofiltration membranes and ii) membrane cleaning techniques using different types of background solutions to prevent the leaching. The results showed that, for three nanofiltration membranes, leaching of high levels of NDMA precursors (~180-450 ng/L) was observed. Leaching continued even after washing membranes with ~3900 L/m2 distilled deionized water. Among various cleaning techniques, washing the membrane with 1 mg/L Cl2 solution was found to be the most effective for reducing NDMA precursor leaching. In the second sub-objective of this research, the control of NDMA, HNM, and THM by selected nanofiltration membranes (NF) was examined. The aims of this objective were to investigate (i) the removal efficiencies of N-nitrosodimethylamine (NDMA), halonitromethane (HNM), and trihalomethane (THM) precursors by NF membranes from different source water types (i.e., surface water, wastewater impacted surface water, and municipal and industrial wastewater treatment effluents), (ii) the impact of membrane type, and (iii) the effects of background water components (i.e., pH, ionic strength, and Ca2+) on the removal of selected DBP precursors from different source waters. The results showed the overall precursor removal efficiencies were 57-83%, 48-87%, and 72-97% for NDMA, HNM, THM precursors, respectively. The removal of NDMA precursors decreased with increases in average molecular weight cut-off and negative surface charge of NF membranes tested, while the removal of THM precursors was only slightly affected. Rising pH increased the removal of NDMA precursors, but did not affect the removal of THM and HNM precursors in municipal WWTP effluent. On the other hand, pH changes had little or no effect on DBP removal from industrial WWTP effluent. In addition, regardless of the membrane type or background water type/characteristics ionic strength did not show any impact on DBP precursor removals. Lastly, an increase in Ca2+ concentration enhanced the removal of NDMA precursors while a slight decrease and no effect was observed for THM and HNM precursors in municipal WWTP effluent, respectively. In the second main objective, first of all, the formation and speciation of THMs, HAAs, and HANs in the presence of different types of dissolved organic matter (DOM) under various drinking water treatment conditions (i.e., initial Br- and Cl2, pH, and contact time) were studied. Results showed that only Cl/Br-DBPs were formed, while I-DBPs were not detected due to the oxidation of I- to IO3-. The formation of total THMs and HAAs increased as the aromaticity of the waters increased whereas the formation of HANs was not affected by aromaticity of DOM. On the other hand, calculated bromine substitution factors (BSF) for all studied waters indicated that BSF was higher in treated water where mostly hydrophobic fraction of DOM was removed in conventional water treatment train. The formation of THMs and HANs increased with increasing initial Br- concentration. Increasing the initial Br- concentration did not affect the total formation of nine HAAs although the five regulated HAAs decreased. In addition, while increasing initial Cl2 concentration enhanced the formation of THMs and HAAs, the formation of HANs initially increased at initial chlorine concentrations ranging from 4 to 8 mg/L, and then decreased at high Cl2 concentration (i.e.,16 mg/L). Increasing initial chlorine concentration also affected the bromine substitution in DOM such that while BSF for HAAs and HANs decreased, BSF for THMs remained constant. When the pH increased from 6.0 to 9.0, the formation of THMs was enhanced while an opposite trend was observed for HAAs and HANs. The formation of DBPs increased with the increasing chlorination time (2 to 48 h) while the bromine substitution decreased due to debromination at longer contact times. Lastly, the formation and toxicity of DBPs during chloramination of bromide and iodide containing waters were studied. The results revealed that increasing DOC concentration increased the formation of I-THMs at a certain point after which the concentrations decreased. Among all species, iodoform was the most influenced by the change in DOC concentration. In addition, the formation of I-THMs was lower in high aromatic content water (SUVA254=5.1 L/mg/m) while it was higher in low aromatic content water (SUVA254=2.1 L/mg/m). While increasing the initial I- concentration gradually increased the formation of iodoform , no considerable change was observed for other I-THM species; increasing bromide concentration from zero to 400 g/L did not make any considerable impact on the formation of I-THMs. From the pratical point of view, the key findings of this research were i) virgin nanofiltration membranes have potential to leach NDMA precursors but their leaching potential can be minimized with low dose of chlorine exposure, ii) the performance of nanofiltration membranes was not significantly impacted by background water characteristics while removing DBPs, which shows the succesful applicability of NF membranes for different water sources, iii) during oxidative water treatment, the presence of Br- and I- in source waters was shown to shift DBP speciation from chlorinated to brominated and iodinated species (under chloramination conditions), which are more cyto- and geno-toxic than their chlorinated analogues.