Environ. Sci. Technol. 2007, 41, 5485-5490
Dissolved Organic Nitrogen as a Precursor for Chloroform, Dichloroacetonitrile, N-Nitrosodimethylamine, and Trichloronitromethane W O N T A E L E E , * ,† PAUL WESTERHOFF,‡ AND J E A N - P H I L I P P E C R O U EÄ § HDR Engineering, Inc., 3200 East Camelback Road, Suite 350, Phoenix, Arizona 85018, Department of Civil and Environmental Engineering, Arizona State University, Tempe, Arizona 85287-5306, and Laboratoire de Chimie de l’Eau et de l’Environnement, Universite´ de Poitiers, Poitiers, France
Nitrogen-containing disinfection byproducts (N-DBPs) are potentially toxic. This study assessed the formation of three N-DBPs (dichloroacetonitrile (DCAN), trichloronitromethane (TCNM), and N-nitrosodimethylamine (NDMA)) and one regulated DBP (chloroform) upon adding free chlorine and monochloramine into solutions containing different fractions (hydrophobic, transphilic, hydrophilic, and colloidal) of dissolved organic matter (DOM) isolates (n ) 17). We hypothesized that N-DBP formation would increase for organic matter enriched in organic nitrogen. Formation potential tests were conducted with free chlorine or preformed monochloramine. Chloramination formed, on average, 10 times lower chloroform concentrations, but 5 times higher DCAN concentrations, as compared with free chlorine addition. The formation of the two halogenated N-DBPs (DCAN and TCNM) increased as the dissolved organic carbon (DOC) to dissolved organic nitrogen (DON) ratio decreased upon adding free chlorine, but the N-DBP formation was relatively constant upon adding monochloramine. NDMA, a nonhalogenated N-DBP, formed on average 0.26 nmol per mg of DOC (4.5 nmol per mg of DON) upon adding monochloramine; no NDMA formation occurred upon adding free chlorine. NDMA formation increased as the DOC/DON ratio decreased (i.e., increasing nitrogen content of DOM). NDMA formation also increased as the amino sugar to aromatic ratio of DOM increased. The results support the hypothesis that DON promotes the formation of N-DBPs.
Introduction Chlorine and chloramines are widely used low-cost disinfectants in water treatment plants, distribution systems, and wastewater effluents. These disinfectants, however, form halogenated and nonhalogenated disinfection byproducts (DBPs) of potential human health concerns (1, 2). Chlo* Corresponding author phone: (602)522-7700; fax: (602)522-7707; e-mail:
[email protected]. † HDR Incorporated. ‡ Arizona State University. § Universite ´ de Poitiers. 10.1021/es070411g CCC: $37.00 Published on Web 06/27/2007
2007 American Chemical Society
ramination forms lower concentrations of regulated DBPs (i.e., trihalomethanes (THMs) and haloacetic acids (HAAs)) than chlorination but forms a larger percentage of unidentified total organic halides (TOX) (3-7). In addition to the currently regulated DBPs, occurrence of nitrogen-containing DBPs (N-DBPs) (e.g., haloacetonitriles, halonitromethanes, and nitrosamines) has been reported (810). Although N-DBPs generally form at lower concentrations than THMs or HAAs, N-DBPs have been shown to be more carcinogenic and mutagenic than the currently regulated DBPs (THMs and HAAs) (11-13). Thus, N-DBPs are an emerging concern during the disinfection of drinking water, wastewater, and reclaimed water. Dissolved organic matter (DOM) reacts with chlorine or chloramines to form N-DBPs. Numerous formation mechanisms, based largely on model compound studies, have been proposed for individual N-DBP species. Dichloroacetonitrile (DCAN) and other haloacetonitriles can be produced from the chlorination of free amino acids (14), heterocyclic nitrogen in nucleic acids (15), proteinaceous materials, and combined amino acids bound to humic structures (16, 17). Many organic nitrogen compounds and humic acids have been shown to be trichloronitromethane (TCNM) precursors in drinking water (18). Generally, the accepted formation mechanisms of N-nitrosodimethylamine (NDMA) and related nitrosamines are nitrosation of organic nitrogen precursors (e.g., dimethylamine) by nitrous acid and/or nitrite (19, 20) or a reaction between monochloramine and organic amines (e.g., dimethylamine) (21, 22). NDMA formation is theoretically governed by a rate-limiting step involving the oxidation of DOM (23) and is generally enhanced when monochloramine is used for disinfection as compared to free chlorine (21, 22). Although some studies have suggested organic nitrogen moieties of DOM (i.e., measured in bulk as dissolved organic nitrogen (DON)) as the likely precursors for N-DBPs in drinking water (16-18, 24), controlled studies are lacking for N-DBP formation from chlorination versus chloramination of DOM. DON analysis is more analytically challenging to quantify than dissolved organic carbon (DOC) (25). Therefore, most studies of DBP formation rely upon DOC and/or UV absorbance measurements to quantify DBP organic precursors (26, 27). Little is known about the role of DON in DBP formation, and precursors leading to N-DBP formation in the water treatment system remain unclear. This study assessed the role of DON in the formation of N-DBPs (DCAN, TCNM, and NDMA) and a regulated DBP (chloroform) upon adding free chlorine or monochloramine. Under controlled laboratory conditions for chlorination or chloramination, formation potential experiments were conducted with solutions containing different fractions (hydrophobic, transphilic, hydrophilic, and colloidal) of DOM isolates, dissolved in nanopure water with only a pH buffer added (i.e., no bromide or inorganic nitrogen). To determine if DOM characteristics affect N-DBP formation, we correlated DBP yields with DOC/DON ratios, amino acid contents, and pyrolysis gas chromatography-mass spectrometry (pyrolysis GC-MS) analysis data.
Experimental Procedures DOM Isolates and Solutions. DOM was isolated at the Universite´ de Poitiers on a variety of source waters from the U.S. and France. Methods detailing the DOM isolation procedures and isolate characterization are provided elsewhere (28). Solutions containing DOM were prepared by dissolving fractionated DOM isolates in deionized water (NANOpure Infinity) and sonicated prior to filtration through VOL. 41, NO. 15, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
5485
TABLE 1. DOC/DON Ratios, SUVA Values, THAA Content, and Pyrolysis GC-MS Analysis of DOM Isolate Solutionsa proportion (%) in the pyrolysis GC-MS analysis sample ID
source
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Britanny River Colorado River Naintre WWTP Effluent Ribou Reservoir Seine River South Platte River Nordic Lake Britanny River Colorado River Ribou Reservoir Nordic Lake Suwannee River South Platte River Suwannee River Suwannee River Naintre WWTP Effluent Ribou Reservoir
polyamino type of DOC/DON SUVA THAA content aromatics saccharides sugars proteins unknown fraction (mg of C/mg of N) (m-1 (mg/L)-1) (nmol/mg of DOC) (%) (%) (%) (%) (%) HPO HPO HPO HPO HPO HPO-A FA TPI TPI TPI TPI-A TPI-A TPI-N TPI-N HPIA+N colloids colloids
19 36 14 21 15 26 58 11 15 9 45 40 6 28 28 9 10
3.4 2.2 2.4 2.3 3.0 3.0 5.0 1.5 1.0 1.3 4.0 3.5 0.9 2.8 2.2 4.6 3.7
159 n/a 912 204 848 157 n/a 557 n/a 264 n/a 143 n/a 471 n/a 4529 n/a
43 34 25 25 29 45 n/a 14 19 n/a n/a n/a 13 40 45 12 6
12 15 5 21 8 13 n/a 13 16 n/a n/a n/a 4 35 15 11 34
10 5 14 6 28 6 n/a 25 23 n/a n/a n/a 17 7 5 38 30
15 25 31 33 23 19 n/a 36 27 n/a n/a n/a 57 11 20 32 24
20 21 25 15 12 19 n/a 12 15 n/a n/a n/a 9 7 15 7 6
a HPO: hydrophobic; HPO-A: hydrophobic acid; FA: fulvic acid; TPI: transphilic; TPI-A: transphilic acid; TPI-N: transphilic neutral; and HPIA+N: hydrophilic acid plus neutral.
ashed (at 500 °C) glass fiber filters (GF/F; pore size ) 0.7 µm; Millipore Corp.). The colloidal DOM isolates were not filtered. Table 1 summarizes the sources and characteristics of DOM isolates. DOC/DON ratios of DOM isolates ranged from 6 to 58 mg of DOC per mg of DON. Specific ultraviolet absorbance (SUVA) values ranged from 1 to 5 m-1 (mg/L)-1. SUVA, the measurement of UV absorbance at 254 nm divided by the DOC concentration, indicates the degree of aromaticity of DOM (29). Solutions were bromide-free; nitrate, nitrite, and ammonia concentrations of solutions were measured to be less than the detection limits, 5 µg/L as N. Chlorination and Chloramination of DOM Isolates. Hypochlorite solutions (Fisher Scientific) were diluted and used to prepare free chlorine and to preform monochloramine stock solutions. Monochloramine stock solutions were prepared daily by mixing hypochlorite and ammonia chloride solutions (both solutions preadjusted to pH 8.3) at a ratio of 0.8 mol/mol. The monochloramine stock solution was refrigerated for at least 3 h before use. Solutions containing DOM isolates at a DOC concentration of 3 mg/L were chlorinated and chloraminated in sealed 1 L amber glass bottles at 20 °C in darkness. Free chlorine DBP formation potential experiments were conducted following the approach of Reckhow et al. (17), using a dose of 5 mg of Cl2 per mg of DOC with a reaction time of 7 days (pH 7.0 with 5 mM phosphate buffer). Monochloramine DBP formation potential experiments were conducted following the approach of Mitch et al. (30), using a high dose of 45 mg of Cl2 per mg of DOC with a reaction time of 10 days (pH 7.0 with 10 mM phosphate buffer), which was based upon logic similar to that behind the THM formation potential test. When the preformed monochloramine was added into DOM solutions at pH 7, part of the monochloramine converted to dichloramine (
