Unravelling High-Molecular-Weight DBP Toxicity Drivers in Chlorinated and Chloraminated Drinking Water: Effect-Directed Analysis of Molecular Weight Fractions.

As disinfection byproducts (DBPs) are ubiquitous sources of chemical exposure in disinfected drinking water, identifying unknown DBPs, especially unknown drivers of toxicity, is one of the major challenges in the safe supply of drinking water. While >700 low-molecular-weight DBPs have been identified, the molecular composition of high-molecular-weight DBPs remains poorly understood. Moreover, due to the absence of chemical standards for most DBPs, it is difficult to assess toxicity contributions for new DBPs identified. Based on effect-directed analysis, this study combined predictive cytotoxicity and quantitative genotoxicity analyses and Fourier transform ion cyclotron resonance mass spectrometry (21 T FT-ICR-MS) identification to resolve molecular weight fractions that induce toxicity in chloraminated and chlorinated drinking waters, along with the molecular composition of these DBP drivers. Fractionation using ultrafiltration membranes allowed the investigation of <1 kD, 1-3 kD, 3-5 kD, and >5 kD molecular weight fractions. Thiol reactivity based predictive cytotoxicity and single-cell gel electrophoresis based genotoxicity assays revealed that the <1 kD fraction for both chloraminated and chlorinated waters exhibited the highest levels of predictive cytotoxicity and direct genotoxicity. The <1 kD target fraction was used for subsequent molecular composition identification. Ultrahigh-resolution MS identified singly charged species (as evidenced by the 1 Da spacing in 13C isotopologues), including 3599 chlorine-containing DBPs in the <1 kD fraction with the empirical formulas CHOCl, CHOCl2, and CHOCl3, with a relative abundance order of CHOCl > CHOCl2 ≫ CHOCl3. Interestingly, more high-molecular-weight CHOCl1-3 DBPs were identified in the chloraminated vs chlorinated waters. This may be due to slower reactions of NH2Cl. Most of the DBPs formed in chloraminated waters were composed of high-molecular-weight Cl-DBPs (up to 1 kD) rather than known low-molecular-weight DBPs. Moreover, with the increase of chlorine number in the high-molecular-weight DBPs detected, the O/C ratio exhibited an increasing trend, while the modified aromaticity index (AImod) showed an opposite trend. In drinking water treatment processes, the removal of natural organic matter fractions with high O/C ratio and high AImod value should be strengthened to minimize the formation of known and unknown DBPs.

Reactions of α,ß-Unsaturated Carbonyls with Free Chlorine, Free Bromine, and Combined Chlorine.

Chemical disinfectants employed in water and wastewater treatment can produce a variety of transformation products, including carbonyl compounds (e.g., saturated and unsaturated aldehydes and ketones). Experiments conducted under conditions relevant to chlorination at drinking water treatment plants and residual chlorine application in distribution systems indicate that α,ß-unsaturated carbonyl compounds readily react with free chlorine and free bromine over a wide pH range but react slowly with combined chlorine (i.e., NH2Cl). For nearly all of the 11 α,ß-unsaturated carbonyl compounds studied, the apparent second-order rate constants for the reaction with free chlorine increased in a linear manner with hypochlorite (OCl-) concentrations, yielding species-specific second-order rate constants for the reaction with OCl- ranging from 0.21 to 12 M-1 s-1. Predictions based on the second-order rate constants indicate that a substantial fraction (i.e., >60%) of several of the more prominent α,ß-unsaturated carbonyls (e.g., acrolein, crotonaldehyde) will be transformed to an appreciable extent in distribution systems by free chlorine. Products from the reaction of chlorine with acrolein, crotonaldehyde, and methyl vinyl ketone were tentatively identified using nuclear magnetic resonance (NMR) and gas chromatography coupled to high-resolution time-of-flight mass spectrometry (GC-HRT-MS). These products lacked unsaturated carbons and, in some cases, contained multiple halogens.

Trace Analysis of 61 Emerging Br-, Cl-, and I-DBPs: New Methods to Achieve Part-Per-Trillion Quantification in Drinking Water.

Disinfection byproducts (DBPs) are a ubiquitous source of chemical exposure in drinking water and have been associated with serious health impacts in human epidemiologic studies. While toxicology studies have pinpointed DBPs with the greatest toxic potency, analytical methods have been lacking for quantifying complete classes of most toxic DBPs at sufficiently low quantification limits (ng/L). This new method reports the parts-per-trillion quantification for 61 toxicologically significant DBPs from 7 different chemical classes, including unregulated iodinated haloacetic acids (HAAs) and trihalomethanes (THMs), haloacetaldehydes, haloketones, haloacetonitriles, halonitromethanes, and haloacetamides, in addition to regulated HAAs and THMs. The final optimized method uses salt-assisted liquid-liquid extraction in a single extraction method for a wide range of DBPs, producing the lowest method detection limits to-date for many compounds, including highly toxic iodinated, brominated, and nitrogen-containing DBPs. Extracts were divided for the analysis of the HAAs (including iodinated HAAs) by diazomethane derivatization and analysis using a GC-triple quadrupole mass spectrometer with multiple reaction monitoring, resulting in higher signal-to-noise ratios, greater selectivity, and improved detection of these compounds. The remaining DBPs were analyzed using a GC-single quadrupole mass spectrometer with selected ion monitoring, utilizing a multimode inlet allowed for lower injection temperatures to allow the analysis of thermally labile DBPs. Finally, the use of a specialty-phase GC column (Restek Rtx-200) significantly improved peak shapes, which improved separations and lowered detection limits. Method detection limits for most DBPs were between 15 and 100 ng/L, and relative standard deviations in tap water samples were mostly between 0.2 and 30%. DBP concentrations in real samples ranged from 40 to 17 760 ng/L for this study.

Does Granular Activated Carbon with Chlorination Produce Safer Drinking Water? From Disinfection Byproducts and Total Organic Halogen to Calculated Toxicity.

Granular activated carbon (GAC) adsorption is well-established for controlling regulated disinfection byproducts (DBPs), but its effectiveness for unregulated DBPs and DBP-associated toxicity is unclear. In this study, GAC treatment was evaluated at three full-scale chlorination drinking water treatment plants over different GAC service lives for controlling 61 unregulated DBPs, 9 regulated DBPs, and speciated total organic halogen (total organic chlorine, bromine, and iodine). The plants represented a range of impacts, including algal, agricultural, and industrial wastewater. This study represents the most extensive full-scale study of its kind and seeks to address the question of whether GAC can make drinking water safer from a DBP perspective. Overall, GAC was effective for removing DBP precursors and reducing DBP formation and total organic halogen, even after >22 000 bed volumes of treated water. GAC also effectively removed preformed DBPs at plants using prechlorination, including highly toxic iodoacetic acids and haloacetonitriles. However, 7 DBPs (mostly brominated and nitrogenous) increased in formation after GAC treatment. In one plant, an increase in tribromonitromethane had significant impacts on calculated cytotoxicity, which only had 7-17% reduction following GAC. While these DBPs are highly toxic, the total calculated cytotoxicity and genotoxicity for the GAC treated waters for the other two plants was reduced 32-83% (across young-middle-old GAC). Overall, calculated toxicity was reduced post-GAC, with preoxidation allowing further reductions.

Chlorination of Source Water Containing Iodinated X-ray Contrast Media: Mutagenicity and Identification of New Iodinated Disinfection Byproducts.

Iodinated contrast media (ICM) are nonmutagenic agents administered for X-ray imaging of soft tissues. ICM can reach µg/L levels in surface waters because they are administered in high doses, excreted largely unmetabolized, and poorly removed by wastewater treatment. Iodinated disinfection byproducts (I-DBPs) are highly genotoxic and have been reported in disinfected waters containing ICM. We assessed the mutagenicity in Salmonella of extracts of chlorinated source water containing one of four ICM (iopamidol, iopromide, iohexol, and diatrizoate). We quantified 21 regulated and nonregulated DBPs and 11 target I-DBPs and conducted a nontarget, comprehensive broad-screen identification of I-DBPs. We detected one new iodomethane (trichloroiodomethane), three new iodoacids (dichloroiodoacetic acid, chlorodiiodoacetic acid, bromochloroiodoacetic acid), and two new nitrogenous I-DBPs (iodoacetonitrile and chloroiodoacetonitrile). Their formation depended on the presence of iopamidol as the iodine source; identities were confirmed with authentic standards when available. This is the first identification in simulated drinking water of chloroiodoacetonitrile and iodoacetonitrile, the latter of which is highly cytotoxic and genotoxic in mammalian cells. Iopamidol (5 µM) altered the concentrations and relative distribution of several DBP classes, increasing total haloacetonitriles by >10-fold. Chlorination of ICM-containing source water increased I-DBP concentrations but not mutagenicity, indicating that such I-DBPs were either not mutagenic or at concentrations too low to affect mutagenicity.

Showering in Flint, MI: Is there a DBP problem?

Lead contamination in the City of Flint, MI has been well documented over the past two years, with lead levels above the EPA Action Level until summer 2016. This resulted from an ill-fated decision to switch from Detroit water (Lake Huron) with corrosion control, to Flint River water without corrosion control. Although lead levels are now closer to normal, reports of skin rashes have sparked questions surrounding tap water in some Flint homes. This study investigated the presence of contaminants, including disinfection by-products (DBPs), in the hot tap water used for showering in the homes of residents in Flint. Extensive quantitative analysis of 61 regulated and priority unregulated DBPs was conducted in Flint hot and cold tap water, along with the analysis of 50 volatile organic compounds and a nontarget comprehensive, broadscreen analysis, to identify a possible source for the reported skin rashes. For comparison, chlorinated hot and cold waters from three other cities were also sampled, including Detroit, which also uses Lake Huron as its source water. Results showed that hot water samples generally contained elevated levels of regulated and priority unregulated DBPs compared to cold water samples, but trihalomethanes were still within regulatory limits. Overall, hot shower water from Flint was similar to waters sampled from the three other cities and did not have unusually high levels of DBPs or other organic chemicals that could be responsible for the skin rashes observed by residents. It is possible that an inorganic chemical or microbial contaminant may be responsible.

Progressive Increase in Disinfection Byproducts and Mutagenicity from Source to Tap to Swimming Pool and Spa Water: Impact of Human Inputs.

Pools and spas are enjoyed throughout the world for exercise and relaxation. However, there are no previous studies on mutagenicity of disinfected spa (hot tub) waters or comprehensive identification of disinfection byproducts (DBPs) formed in spas. Using 28 water samples from seven sites, we report the first integrated mutagenicity and comprehensive analytical chemistry of spas treated with chlorine, bromine, or ozone, along with pools treated with these same disinfectants. Gas chromatography (GC) with high-resolution mass spectrometry, membrane-introduction mass spectrometry, and GC-electron capture detection were used to comprehensively identify and quantify DBPs and other contaminants. Mutagenicity was assessed by the Salmonella mutagenicity assay. More than 100 DBPs were identified, including a new class of DBPs, bromoimidazoles. Organic extracts of brominated pool/spa waters were 1.8× more mutagenic than chlorinated ones; spa waters were 1.7× more mutagenic than pools. Pool and spa samples were 2.4 and 4.1× more mutagenic, respectively, than corresponding tap waters. The concentration of the sum of 21 DBPs measured quantitatively increased from finished to tap to pool to spa; and mutagenic potency increased from finished/tap to pools to spas. Mutagenic potencies of samples from a chlorinated site correlated best with brominated haloacetic acid concentrations (Br-HAAs) (r = 0.98) and nitrogen-containing DBPs (N-DBPs) (r = 0.97) and the least with Br-trihalomethanes (r = 0.29) and Br-N-DBPs (r = 0.04). The mutagenic potencies of samples from a brominated site correlated best (r = 0.82) with the concentrations of the nine HAAs, Br-HAAs, and Br-DBPs. Human use increased significantly the DBP concentrations and mutagenic potencies for most pools and spas. These data provide evidence that human precursors can increase mutagenic potencies of pools and spas and that this increase is associated with increased DBP concentrations.

Ubiquitous Production of Organosulfates During Treatment of Organic Contaminants with Sulfate Radicals.

Oxidation of organic contaminants by sulfate radical (SO4 •-) is becoming more popular for the treatment of hazardous waste sites by in situ chemical oxidation (ISCO) and industrial wastewater by advanced oxidation processes (AOPs). It is well documented that SO4 •- can produce similar oxygen-containing transformation products as hydroxyl radical-based treatment processes, but SO4 •- also has the potential to produce organosulfates by radical addition. Experiments conducted with a suite of 23 aromatic and 5 aliphatic compounds, including several contaminants typically detected at hazardous waste sites, demonstrated the formation of at least one stable sulfate-containing product for 25 of the compounds. These compounds likely exhibit higher mobility in the subsurface due to a lower affinity for surfaces (e.g., aquifer solids, activated carbon) than most other transformation products. Although the health risks associated with organosulfates are still uncertain, some aromatic organosulfates produced in this study (i.e. phenyl sulfate and p-cresyl sulfate) are known to be harmful uremic toxins. Further study of organosulfate formation, fate, and toxicity is needed before SO4 •--based treatment processes are more widely employed.

A novel automated method for the quantification of ten halobenzoquinones in drinking water using online solid-phase extraction coupled with liquid chromatography tandem mass spectrometry.

When assessing occurrence and exposure to toxic chemicals in drinking water, developing methods that are sensitive and efficient is paramount. A new method was developed for the quantification of ten halobenzoquinones (HBQs), a class of disinfection by-products (DBPs) in drinking water, which have been shown to be more toxic than most regulated DBPs. This method uses a small sample volume with online solid phase extraction (SPE) followed by liquid chromatography-tandem mass spectrometry (LC-MS/MS). Previous methods used offline SPE coupled with standard addition, and a pre-concentration step. This requires larger volumes of sample, solvent, and standards. The current method has comparable limits of quantification (0.2-166 ng/L), requires minimal sample preparation, and analysis is almost entirely automated. It also includes 2,6-dibromo-3­chloro-5-methyl-1,4-benzoquinone which was not included in previous methods. A stability test was conducted over a one-week period with different preservatives, including ascorbic acid, sodium thiosulfate, and formic acid. Method optimization included source temperature, SPE size, sample volume, and SPE loading time. This method was validated using drinking water collected from four different drinking water plants; spike recoveries of HBQs were between 70-111%, relative standard deviations of <20%, and linearity of >0.98. Further, using this method, we report the highest concentration of 2,6-dibromo-1,4-benzoquinone found in drinking water (254 ng/L).

Controlling disinfection byproducts from treated wastewater using adsorption with granular activated carbon: Impact of pre-ozonation and pre-chlorination.

This study measured chlorine- and chloramine-reactive precursors using formation potential (FP) tests of nine U.S. Environmental Protection Agency (EPA) regulated and 57 unregulated disinfection byproducts (DBPs) in tertiary-filtered wastewater before and after pilot-scale granular activated carbon (GAC) adsorption. Using breakthrough of precursor concentration and of concentration associated calculated cytotoxicity and genotoxicity (by correlating known lethal concentrations reported elsewhere), the performance of three parallel GAC treatment trains were compared against tertiary-filtered wastewater: ozone/GAC, chlorine/GAC, and GAC alone. Results show GAC alone was the primary process, versus ozone or chlorine alone, to remove the largest fraction of total chlorine- and chloramine-reactive DBP precursors and calculated cytotoxicity and genotoxicity potencies. GAC with pre-ozonation removed the most chlorine- and chloramine-reactive DBP precursors followed by GAC with pre-chlorination and lastly GAC without pre-treatment. GAC with pre-ozonation produced an effluent with cytotoxicity and genotoxicity of DBPs from FP that generally matched that of GAC without pre-oxidation; meanwhile removal of toxicity was greater by GAC with pre-chlorination. The cytotoxicity and genotoxicity of DBPs from FP tests did not scale with DBP concentration; for example, more than 90% of the calculated cytotoxicity resulted from 20% of the DBPs, principally from haloacetaldehydes, haloacetamides, and haloacetonitriles. The calculated cytotoxicity and genotoxicity from DBPs associated with FP-chloramination were at times higher than with FP-chlorination though the concentration of DBPs was five times higher with FP-chlorination. The removal of DBP precursors using GAC based treatment was at least as effective as removal of DOC (except for halonitromethanes for GAC without pre-oxidation and with pre-chlorination), indicating DOC can be used as an indicator for DBP precursor adsorption efficacy. However, the DOC was not a good surrogate for total cytotoxicity and genotoxicity breakthrough behavior, therefore, unregulated DBPs could have negative health implications that are disconnected from general water quality parameters, such as DOC, and regulated classes of DBPs. Instead, cytotoxicity and genotoxicity correlate with the concentration of specific classes of unregulated DBPs.

GAC to BAC: Does it make chloraminated drinking water safer?

Biological activated carbon (BAC) is widely used as a polishing step at full-scale drinking water plants to remove taste and odor compounds and assimilable organic carbon. BAC, especially with pre-ozonation, has been previously studied to control regulated disinfection by-products (DBPs) and DBP precursors. However, most previous studies only include regulated or a limited number of unregulated DBPs. This study explored two full-scale drinking water plants that use pre-chloramination followed by BAC and chloramine as the final disinfectant. While chloramine generally produces lower concentrations of regulated DBPs, it may form increased levels of unregulated nitrogenous and iodinated DBPs. We evaluated 71 DBPs from ten DBP classes including haloacetonitriles, haloacetamides, halonitromethanes, haloacetaldehydes, haloketones, iodinated acetic acids, iodinated trihalomethanes, nitrosamines, trihalomethanes, and haloacetic acids, along with speciated total organic halogen (total organic chlorine, bromine and iodine) across six different BAC filters of increasing age. Most preformed DBPs were well removed by BAC with different ages (i.e., operation times). However, some preformed DBPs were poorly removed or increased following treatment with BAC, including chloroacetaldehyde, dichloronitromethane, bromodichloronitromethane, N-nitrosodimethylamine, dibromochloromethane, tribromomethane, dibromochloroacetic acid, and tribromoacetic acid. Some compounds, including dibromoacetaldehyde, bromochloroacetamide, and dibromoacetamide, were formed only after treatment with BAC. Total organic halogen removal was variable in both plants and increases in TOCl or TOI were observable on one occasion at each plant. While calculated genotoxicity decreased in all filters, decreases in overall DBP formation did not correlate with decreases in calculated cytotoxicity. In three of the six filters, calculated toxicity increased by 4-27%. These results highlight that DBP concentration alone may not always provide an adequate basis for risk assessment.

The DBP exposome: Development of a new method to simultaneously quantify priority disinfection by-products and comprehensively identify unknowns.

Disinfected drinking water contains hundreds of disinfection by-products (DBPs) that are formed by the reaction of disinfectants with natural and anthropogenic organic matter, bromide, and iodide. Understanding what these DBPs are is important because millions of people worldwide consume drinking water every day, and human epidemiologic studies have reported cancer, miscarriage, and birth defects from consuming such waters. While more than 600 DBPs are reported in the literature, very few studies quantify complete classes of chlorinated, brominated, and iodinated DBPs. Also, very few studies conduct comprehensive non-target analyses of unknown DBPs to characterize the complete DBP exposure (the exposome). We developed a new gas chromatography (GC)-mass spectrometry (MS) method that simultaneously quantifies 39 priority unregulated DBPs from six different chemical classes (haloacetaldehydes, haloketones, haloacetamides, haloacetonitriles, halonitromethanes, and iodinated-trihalomethanes) and analyzes unknown DBPs with mass accuracy <600 ppm under full-scan conditions. Using a new type of time-of-flight (TOF) mass spectrometer, which combines selected ion monitoring (SIM)-level sensitivity with mass accuracy of ±0.05 Da, method detection limits of 3-61 ng/L were achieved. These levels were found to be quite comparable to those of a widely used single quadrupole mass spectrometer (2-90 ng/L) operated in SIM mode. However, analysis using this TOF mass spectrometer offers two additional advantages over traditional quadrupole-MS: (1) full-scan data, which provides additional confidence for target analytes, as well as complete mass spectra for unknown analysis, and (2) two decimal place mass accuracy, which allows additional confidence for target analytes and importantly, molecular formula indication for unknowns. High resolution accurate mass TOF was also used to validate identification of selected compounds. This new method was demonstrated on finished drinking waters from three different drinking water plants, where target quantification and non-target unknown analyses were performed simultaneously during the same run. This enabled the quantification of 39 DBPs, along with the non-target identification of many other drinking water contaminants, including two additional non-target DBPs: N,N-dimethylacetamide and N-nitrosodibutylamine.