Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)

Legacy Department

Environmental Engineering and Earth Science


Karanfil, Tanju

Committee Member

Carraway , Elizabeth

Committee Member

Freedman , David L.

Committee Member

Schlautman , Mark A.


Nine halonitromethanes (HNMs) were among the more than fifty unregulated high-priority disinfection by-products (DBPs) that were monitored for and detected in drinking waters in a 2000-2002 nationwide occurrence study funded by the US Environmental Protection Agency (USEPA). The toxicology studies conducted in recent years have shown that HNMs are one of the most cytotoxic and genotoxic classes among the emerging DBPs, with orders of magnitude higher cytotoxicity and genotoxicity than any of the regulated organic DBPs. Furthermore, brominated HNMs were found to be more toxic than their chlorinated analogs, with dibromonitromethane being the most cyto- and genotoxic. Today, our limited understanding of the HNM formation in drinking water is mainly about trichloronitromethane (TCNM or chloropicrin). The formation of other eight HNM species has been largely neglected due to the lack of commercial HNM standards and analytical methods which just become available in recent years.
The main objective of this research was to investigate the formation and distribution of HNMs in drinking waters. Specifically, the following research objectives were: (1) examing the HNM formation potentials under different disinfection/oxidation conditions typically used in water treatment, and the relationships between HNM formation potential and some water quality parameters commonly measured at water treatment plants, (2) examing the roles of pH, bromide and nitrite on the formation and distribution of HNMs at typical drinking water treatment conditions, and (3) investigating the formation of HNMs from selected nitrogenous organic model compounds (i.e., eight amino acids and four amino sugars) at different pH, bromide and nitrite levels.
HNM formation potentials in drinking waters were examined under different oxidation conditions using samples obtained from various drinking water sources. The results showed that ozonation-chlorination produced the highest HNM yields followed by chlorination, ozonation-chloramination, and chloramination. Higher or about the same HNM yields were observed in treated waters compared to raw waters, indicating that hydrophilic natural organic matter (NOM) components that are not effectively removed by conventional treatment processes are likely the main precursors of HNMs. This was further confirmed by examining HNM formation potentials of NOM fractions obtained with resin fractionation. Hydrophilic NOM fractions (HPI) showed significantly higher HNM yields than hydrophobic (HPO) and transphilic (TPH) fractions. Regression analyses of HNM formation potentials with various water quality parameters showed the best regression was between HNM yields and the ratios of dissolved organic nitrogen to dissolved organic carbon concentrations along with specific ultraviolet absorbance at 254nm (SUVA254) during ozonation-chlorination in the water samples tested.
HNM formation and distribution were also examined under various disinfection conditions of typical water distribution systems in the United States, and the effects of pH, bromide and nitrite on HNM formation and distribution were evaluated. During these experiments, HNM formation was also compared with the regulated trihalomethanes (THMs). Preozonation prior to chlorination enhanced HNM formation and resulted in the formation of various HNMs species. Trihalogenated HNMs (THNMs) were the major HNM species formed. Brominated HNM species were observed in the presence of bromide. Dihalonitromethanes (DHNMs) were detected in some cases but it was usually below the minimum reporting levels (MRLs). These results demonstrated that if TCNM was the only species monitored, the overall formation of HNMs in the samples would be underestimated. Treated water collected after conventional treatment showed higher or similar reactivity toward HNM formation than the raw water, confirming that the conventional treatment did not effectively remove the HNM precursors and hydrophilic natural organic matter (NOM) components are the likely precursors of HNMs. Formation of THMs sharply contrasted to the HNM formation. THM formation was much higher in the raw water and decreased after pre-ozonation. These results indicated that precursors and pathways of formation are not the same for HNMs and THMs. HNM formation generally increased with pH. During ozonation-chlorination, bromide increased overall formation of HNMs and shifted HNM formation to brominated species. The presence of nitrite increased HNM formation under both chlorination and ozonation-chlorination conditions, but did not have any influence on THM formation. HNM formation during ozonation-chloramination was about 1 μg/L even at high bromide and nitrite concentrations. Monochloramination alone did not form any measurable HNMs. These results indicated that the use of chloramine can be an effective way to reduce HNM formation at typical drinking water treatment conditions.
Eight amino acids (glycine, alanine, serine, cysteine, aspartic acid, glutamic acid, lysine and histidine) and four amino sugars (glucosamine, galactosamine, N-acetylglucosamine and N-acetylneuraminic acid) were tested as model compounds to examine HNM formation with formation potential tests under two disinfection scenarios, chlorination and ozonation-chlorination. In addition, the effects of bromide, nitrite and pH on HNM formation were also evaluated. Formation and distribution of THMs were also analyzed during the experiments and compared with HNMs. HNM formation was about minimum reporting level with chlorination at pH 6 and 8 for most of the model compounds. Preozonation enhanced the formation of HNMs for all the model compounds at pH 8. The most important finding was that glycine and lysine were the two amino acids that formed significantly higher amount of HNMs with ozonation-chlorination compared to other model compounds. Glycine is one of the amino acids most commonly present in the surface waters in North America. In particular, at pH 8, almost fifteen and one hundred times more reactivity of glycine than lysine and other tested model compounds, respectively, indicated that it could be one of the major precursors of HNMs during ozonation-chlorination. The presence of bromide had a small impact on HNM formation during chlorination, whereas with ozonation-chlorination HNM formation was increased when bromide was presented. Bromide incorporation was more efficient at higher pH values with chlorine or ozone-chlorine. The formation of THMs from most of the amino acids and sugars reacted with chlorine or ozone-chlorine increased in the presence of bromide. In the presence of nitrite, higher HNM formation was observed for ozonation-chlorination, while no significant impact of nitrite was observed for chlorination. There was no pH effect on HNM formation with chlorine for all model compounds. However pH effect on HNM formation for the model compounds was different during ozonation-chlorination.