Nitrosamine, dimethylnitramine, and chloropicrin formation during strong base anion-exchange treatment

Strong base anion-exchange resins represent an important option for water utilities and homeowners to address growing concerns with nitrate, arsenate, and perchlorate contamination of source waters. Most commercially available anion-exchange resins employ quaternary amine functional groups. Previous research has provided contradictory evidence regarding whether these resins serve as sources of nitrosamines, considered as highly carcinogenic nitrogenous disinfection byproducts (N-DBPs), even without disinfectants. For three common varieties of commercial anion-exchange resins, we evaluated the importance of releases of nitrosamines, and two other N-DBPs (dimethylnitramine and chloropicrin), when the resins were subjected to typical column flow conditions with and without free chlorine or chloramine application upstream or downstream of the columns. In the absence of disinfectants, fresh trimethylamine- and tributylamine-based type 1 and dimethyl-ethanolamine-based type 2 anion-exchange resins usually released 2-10 ng/L nitrosamines, likely due to shedding of manufacturing impurities, with excursions of up to 20 ng/L following regeneration. However, the lack of significant nitrosamine release in a full-scale anion-exchange treatment system after multiple regeneration cycles indicates that releases may eventually subside. Resins also shed organic precursors that might contribute to nitrosamine formation within distribution systems when chloramines are applied downstream. With free chlorine or chloramine application upstream, nitrosamine concentrations were more significant, at 20-100 ng/L for the type 1 resins and ∼400 ng/L for the type 2 resin. However, chloropicrin formation was lowest for the type 2 resin. Dimethylnitramine formation was significant with free chlorine application upstream but negligible with chloramines. Although no N-DBPs were detected in cation-exchange-based consumer point-of-use devices exposed to chlorinated or chloraminated waters, our results indicate that inclusion of anion-exchange resins in these devices, as in laboratory deionized water systems, would likely be problematic.