Halomethanesulfonic AcidsâA New Class of Polar Disinfection

Subscriber access provided by BUFFALO STATE

Environmental Measurements Methods

Halomethanesulfonic acids - a new class of polar disinfection by-products: standard synthesis, occurrence, and indirect assessment of mitigation options Daniel Zahn, Reinhard Meusinger, Tobias Frömel, and Thomas P. Knepper Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.9b03016 • Publication Date (Web): 12 Jul 2019 Downloaded from pubs.acs.org on July 17, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 16

Environmental Science & Technology

2

Halomethanesulfonic acids - a new class of polar disinfection by-products: standard synthesis, occurrence, and indirect assessment of mitigation options

3

Daniel Zahn1, Reinhard Meusinger2, Tobias Frömel1, Thomas P. Knepper*1

4

1Hochschule

5

2TU

6 7

*Corresponding author: Thomas P. Knepper, Hochschule Fresenius, University of Applied Sciences, Limburger Straße 2, 65510 Idstein, Germany, [email protected]

8

Abstract

1

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Fresenius, University of Applied Sciences, Limburger Straße 2, 65510 Idstein, Germany

Darmstadt, FB Chemie, Alarich-Weiss-Str. 4, 64287 Darmstadt, Germany

Halomethanesulfonic acids (HMSAs) are recently discovered polar disinfection by-products without commercially available reference materials. To allow for their accurate quantification, we successfully synthesized standards for the four presumably most prevalent HMSA congeners: chloromethanesulfonic acid, bromomethanesulfonic acid, dichloromethanesulfonic acid, and bromochloromethanesulfonic acid. After structure confirmation and quantification with highresolution mass spectrometry and nuclear magnetic resonance spectroscopy, we integrated them into a multi-layer solid phase extraction and hydrophilic interaction liquid chromatography – tandem mass spectrometry method dedicated to the analysis of polar water contaminants. With this method we monitored HMSAs in drinking water production plants from four European countries and tap water samples taken in six countries. HMSAs were detected in the low µg/L after the chlorination step during drinking water production, all tap waters samples, and two surface waters used for drinking water production. Concentrations in tap water samples ranged from 0.07 µg/L to 11.5 µg/L while the HMSA concentrations in surface waters were in the range of 100 ng/L. We utilized the HMSA formation potential to indirectly asses the behaviour of hitherto unknown HMSA precursors, consequently identifying ozonation, filtration through activated carbon and reverse osmosis as efficient removal tools for HMSA precursors, thus limiting their formation during subsequent water disinfection.

25 26

Key Words:

27

- Halogenated disinfection by-products (DBPs)

28

- Hydrophilic interaction liquid chromatography (HILIC)

29

- Persistent and mobile organic contaminants (PMOC)

30

- Very persistent, very mobile (vPvM) contaminants

31

- Quantification in tap and drinking water

32

- Formation potential

33

- Standard synthesis

34 35

1. Introduction

36 37 38

In industrialized countries drinking water is regularly treated with chlorine-containing disinfectants to drastically reduce the presence of pathogens, and thus circumvent the spread of waterborne diseases. The long term exposure to disinfection by-products (DBPs, for an overview of abbreviations see Table

ACS Paragon Plus Environment


39 40 41 42 43 44 45 46

S1) involuntarily formed during this water disinfection, however, has itself been associated with adverse health effects ranging from reproductive complications like a low birth weight, birth defects and stillbirth[1-3] to direct effects on the consumer health like an increased risk of bladder cancer[4]. Elevated bromide and iodide concentrations in many raw waters, which are expected to rise as a consequence of hydraulic fracturing, coal-fired power plants, industrial discharges or draughts caused by climate change[5] may lead to an increased formation of brominated and iodinated DBPs, which are often more toxic than their chlorinated congeners, thus possibly magnifying adverse health effects caused by DBPs within the next decades.

47 48 49 50 51 52 53 54 55 56 57 58 59

Of the potentially thousands of DBPs formed from natural organic matter or anthropogenic chemicals, approximately 700 have been identified and only a fraction is regulated[6]. The adverse health effects observed, however, cannot be explained by the regulated DBPs alone and more than 50% of the total organic halogen content (TOX) is still not accounted for[7], which in turn implies that important DBPs, either toxicologically or by weight, still remain unknown. The advent of novel analytical methodologies facilitated by advancements in computer-assisted data treatment, the incorporation of alternative chromatographic techniques (e.g. hydrophilic interaction liquid chromatography - HILIC) and an increasing availability of high-resolution mass spectrometers led to the identification of various hitherto unknown DBPs, including mutagenic nitrogenous DBPs[8], peptide DBPs[9], linear alkylbenzene sulfonate DBPs[10], trihalomethanols[11], N-chloro DBPs[12], and various DBPs of drugs and their metabolites[13, 14]. In fact, hundreds of previously not described DBPs were detected by Gonsior et al.[15], Zhang et al.[16], and Lavonen et al.[17] utilizing high-resolution mass spectrometry (HRMS), thereby clearly demonstrating that a vast number of DBPs that still remains unknown.

60 61 62 63 64 65 66 67 68 69 70

Among these novel DBPs are the very polar halomethanesulfonic acids (HMSAs), which were identified in water samples from five European countries[18] utilizing HILIC-HRMS and a multi-layer solid phase extraction (mlSPE) enrichment method[19]. While these DBPs were estimated to be present in the high ng/L or even µg/L range, any accurate quantification was impeded by the lack of commercially available reference materials. To enable their detailed investigation, we successfully synthesized four of these HMSAs, determined the concentrations of the purified standards with quantitative nuclear magnetic resonance spectroscopy (qNMR), and utilized them to assess the concentrations of the synthesized HMSAs in four drinking water production plants (DWPPs) and in 14 tap water samples taken in six countries. In addition, the HMSA formation potential was utilized to assess the removal of yet unknown HMSA precursor substances throughout these four DWPPs and identify treatment steps that may contribute to their mitigation.

71 72

2. Materials and Methods

73

2.1 Chemicals

74 75 76 77 78 79 80 81 82 83 84

Methanol (MeOH, ≥ 99.98%), ammonium hydroxide (30% in water), acetonitrile (ACN, ≥ 99.98%), sodium bromide (≥ 99%), formic acid (≥ 99.9%), dichloromethane (≥ 99%, not stabilized), sodium hypochlorite (5-10% available chlorine), and ascorbic acid (≥ 99%) were purchased from Carl Roth GmbH (Karlsruhe, Germany). Chloromethanesulfonyl chloride (≥ 90%), dichloromethanesulfonyl chloride (n/a), methanesulfonic acid (≥ 98%), 18O-water (97% 18O), ammonium formate (NH4Fo, ≥ 99%) were supplied by Sigma-Aldrich (Schnelldorf, Germany). Weak anion exchanger (WAX) and weak cation exchanger (WCX) cartridges and bulk materials were purchased from Waters (Darmstadt, Germany). Supelclean ENVI-Carb (particle size 120-400 mesh) was purchased from Supelco (Bellfonte, USA) CHROMABOND polypropylene cartridges (3 mL) and CHROMABOND polypropylene filters were supplied by Macherey-Nagel (Düren, Germany). Deionized water (18 MΩcm) was supplied by a Simplicity UV water purification system (Merck, Darmstadt, Germany). Regenerated cellulose syringe ACS Paragon Plus Environment

Page 2 of 16

Page 3 of 16


85 86

filters (0.2 µm) and glass fiber filters (0.45 µm, 55 mm) were purchased from GE Healthcare (Little Chalfont, UK)

87

2.2 Synthesis of HMSAs

88 89 90 91 92 93 94 95 96 97 98 99

Chlorinated MSAs were synthesized by hydrolysis of the respective chloromethanesulfonyl chlorides. 20 mL ultrapure water was added to chloromethanesulfonyl chloride (1.00 g, 6.71 mmol) and dichloromethanesulfonyl chloride (0.84 g, 4.56 mmol) and placed into an oven at 60°C. Hydrolysis was finished after 24 h (chloromethanesulfonic acid) and 72 h (dichloromethanesulfonic acid), respectively, and resulted in complete conversion of the respective sulfonyl chlorides. Afterwards, the samples were diluted with ultrapure water to 1 L, split into 25 mL aliquots and passed through Oasis WAX cartridges (6 mL, 500 mg, Waters, Darmstadt, Germany). The cartridges were eluted with 5 mL methanol containing 5% NH4OH. The eluates for chloro- (ClMSA) and dichloromethanesulfonic acid (Cl2MSA) were pooled, evaporated to dryness under a gentle stream of nitrogen, and reconstituted in 6 mL ultrapure water/acetonitrile 1:1 (v:v). Trifluoromethanesulfonyl chloride was hydrolyzed in 250 µL 18OH2O to 18O-trifluoromethanesulfonic acid (18O-TFMSA), diluted in H2O, and passed through an Oasis WAX cartridge (3 mL, 60 mg, Waters, Darmstadt, Germany) for purification.

100 101 102 103 104 105 106 107

The conversion of chlorinated to brominated HMSAs was achieved by a halogen exchange reaction. 3 mL aliquots of the ClMSA and Cl2MSA solutions were transferred into microwave vials, evaporated under a gentle stream of nitrogen, reconstituted in 4 mL saturated aqueous sodium bromide solution, and heated to 165°C in a pressurized container. For ClMSA 77.5% molar conversion to bromomethanesulfonic (BrMSA) acid was achieved after a reaction time of 16 h, while the reaction of Cl2MSA was stopped after 40.5 h which resulted in a molar conversion of 19.1% to bromochloromethanesulfonic acid (BrClMSA). Brominated MSAs were purified with Oasis WAX cartridges analogously to their chlorinated congeners.

108

2.3 Ion chromatography

109 110 111 112 113 114 115

A Metrohm (Herisau, Switzerland) ion chromatograph consisting of an IC Sample Processor, 771 IC Compact Interface, 762 IC Interface, 732 IC Detector (conductivity detector with suppressor), 733 IC Separation Center, 752 Pump unit, and an MF 709 Pump was utilized to quantify chlorine that was released during the hydrolysis of chlorinated methanesulfonyl chlorides, and thus served to indirectly monitor the hydrolysis of chlorinated methanesulfonyl chlorides. A Metrosep A Supp 5 column (4.0*150 mm; 5 µm, Methrohm, Herisau, Switzerland) and two eluents consisting of 3.2 mM aqueous Na2CO3 and 1 mM aqueous NaHCO3 were utilized for separation.

116

2.4 High resolution mass spectrometry

117 118 119 120 121 122 123

HRMS measurements were performed with an Orbitrap Velos Pro equipped with a ‘Heated Electrospray Ionization’ (H-ESI II) ion source (Thermo Scientific, Bremen, Germany). Structure confirmation of synthesized HMSAs was performed after direct infusion with a flow rate of 10 µL/min and negative ionization (2.8 kV) utilizing a full scan with a nominal resolution of 100,000 (at m/z 400, Figure S1, Table S2 to S5) and a product ion scan (Figure S2 –S5) after fragmentation in the HCD collision cell. In addition, direct infusion of diluted reaction mixtures with these conditions was utilized to monitor the progress of the halogen exchange from chlorinated to brominated HMSAs.

124

2.5 NMR

125 126 127

NMR spectra were recorded from 300 µL D2O solutions at 296 K on a Bruker DRX-500 spectrometer (Bruker BioSpin, Karlsruhe, Germany) in Shigemi tubes. Spectra were referenced to the HDO signal at 4.8 ppm. For each sample, 1H-NMR spectra were acquired (Figure S6), and supplemented by 13C-NMR



128 129

spectra for ClMSA and BrMSA (Figure S7). Quantification of synthesized HMSAs was performed based on the 1H-NMR signal utilizing methanesulfonic acid as internal standard.

130

2.6 Water samples

131 132 133 134 135 136 137

Four European drinking water production plants with various treatment technologies (including a pilot plant utilizing reversed osmosis) were sampled during April and August 2017. Grab samples from DWPPs were taken monthly in three consecutive months while tap water samples were taken once in November or December 2017. These samples were complemented by 14 tap water samples taken in December 2017. All samples taken after a chlorination step and all tap water samples in general were quenched by addition of 0.1 mM ascorbic acid. All samples were stored for up to 6 months at 4°C under exclusion of light. For additional sample information see Table S6 and S7.

138

2.7 Sample pretreatment

139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155

Water samples were filtered through glass fiber filters (0.45 µm, 55 mm, GE Healthcare, Little Chalfont, UK) and enriched with a mlSPE method and an evaporation method[19]. For mlSPE, 3 mL blank CHROMABOND cartridges were filled with 60 mg (±5 mg) of graphitized carbon black, WCX, and WAX from bottom to top and separated by polyethylene frits. After conditioning with 1 mL MeOH + 5% NH4OH, 1 mL MeOH + 2% HCOOH, 1 mL MeOH, and 3 mL deionized water, 100 mL aqueous sample (filtered and adjusted to pH 5.5 ± 0.1 with NH4OH or HCOOH) were passed through the cartridge. Once the cartridge was dried for 15 min with nitrogen, the SPE materials were eluted sequentially with 3 mL MeOH + 5% NH4OH, 3 mL MeOH + 2% HCOOH, and 1.5 mL MeOH/CH2Cl2 80:20 (v:v). After the first and third elution step the eluent was evaporated to dryness under a gentle stream of nitrogen at 50°C. Evaporation after the first elution step was performed to evaporate ammonia and thus limit the formation of ammonium formate which occurs when ammonia and formic acid containing eluents are mixed. The residue of the pooled eluates was reconstituted in 500 µL ACN/H2O 95:5 (v:v) and filtered through a syringe filter (regenerated cellulose, 0.2 µm) for subsequent HILIC - scheduled Multiple Reaction Monitoring (sMRM) analysis. Sample enrichment by evaporation was achieved by evaporating 2.5 mL of aqueous matrix at 50°C and 9 mbar. After evaporation to dryness, the residue was reconstituted in 500 µL ACN/H2O 95:5 (v:v) and filtered through a syringe filter (0.2 µm, GE Healthcare, Little Chalfont, UK) for subsequent HILIC-sMRM analysis.

156

2.8 Quality control

157 158 159 160 161 162 163 164 165 166 167 168

All quality control samples utilized for method performance assessment were prepared analogously to the other samples but spiked at different points of the procedure. Samples were either not spiked at all (βnative), spiked before enrichment (βpre) or spiked after enrichment (βpost). The standard concentration βStd. refers to the concentration the sample was spiked with. Apparent recoveries were calculated according to Equation 1. The method LOQ (Equation 2) was derived from the instrumental LOQ determined by a 12-point calibration, the enrichment factor, sample specific apparent recoveries, and a safety factor of two which was deemed to be sufficient after manual comparison of the baseline of pure solvent blanks, standards and 20 randomly selected samples. The contribution of the recovery (Equation 3) and matrix effects (Equation 4) to the apparent recovery was investigated in detail with five sample matrices containing artificial freshwater (preparation see Table S8), tap waters, and ground waters. Fortified samples (βfort) were spiked with one tenth of the concentration used for the determination of the apparent recovery (βfort, Std.), recovery, and matrix effect calculation.

169 170 171 172

Herein, trueness refers to the recovery of fortified samples when the reciprocal sample-specific apparent recovery is utilized as correction factor (Equation 5) and can thereby be utilized to assess if a correction with the sample-specific apparent recovery determined at one concentration is able to produce correct results over an extended concentration range. Throughout this work double ACS Paragon Plus Environment

Page 4 of 16

Page 5 of 16


173 174 175

determination refers to two independent replicates of a sample. Blank concentrations were calculated from ten independent replicates of deionized water prepared analogously to other samples with the evaporation and mlSPE method on five separate days.

176 177

Throughout all quality control experiments, concentrations were not corrected with order to directly assess the behavior of the analytes.

178

Equation 1:

𝐴𝑝𝑝𝑎𝑟𝑒𝑛𝑡 𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑦 (%) =

179

Equation 2:

𝑀𝑒𝑡ℎ𝑜𝑑 𝐿𝑂𝑄 = 𝑒𝑛𝑟𝑖𝑐ℎ𝑚𝑒𝑛𝑡 𝑓𝑎𝑐𝑡𝑜𝑟 ∗

180

Equation 3:

𝑅𝑒𝑐𝑜𝑣𝑒𝑟𝑦 (%) =

181

Equation 4:

𝑀𝑎𝑡𝑟𝑖𝑥 𝑒𝑓𝑓𝑒𝑐𝑡𝑠 (%) =

182

Equation 5:

𝑇𝑟𝑢𝑒𝑛𝑒𝑠𝑠 (%) =

183

2.9 HILIC-sMRM

184 185 186 187 188 189

Quantification of HMSAs was achieved with a Nexera X2 UHPLC system (Shimadzu, Kyoto, Japan) interfaced with an AB Sciex QTrap 5500 (Darmstadt, Germany) and controlled with the Analyst software (Analyst 1.6.2). An Acquity BEH Amide column (100*2.1 mm; 1.7 µm, Waters, Milford, USA) was operated with an ACN/water gradient containing 5 mM NH4Fo at pH 3 at a flow rate of 500 µL/min. Analysis was performed in sMRM mode using two transitions per analyte. For additional method information please consult the supplementary information Table S9 to Table S11.

190

2.10 Chlorination experiment – formation potential

191 192 193 194

To assess the formation potential of HMSAs, sodium hypochlorite (100 mg/L, calculated as available chlorine) was added to 10 mL aliquots of DWPP samples. After 3 h reaction time at 20°C under exclusion of light, residual chlorine was quenched with a 1.2-fold molar access of ascorbic acid. Afterwards the samples were enriched with evaporation and analyzed by HILIC-sMRM.

195

3. Results and Discussion

196

3.1 Synthesis of chlorinated and brominated HMSAs

197 198 199 200 201 202 203 204 205 206 207 208 209 210 211

Halomethanesulfonic acids are not commercially available, and thus had to be synthesized to allow for accurate quantification. Of the six initially identified chlorinated and brominated HMSAs, four were successfully synthesized in significant quantities (ClMSA, BrMSA, Cl2MSA, and ClBrMSA). Unless otherwise stated, HMSAs from here on refers to these four congeners in the context of this paper. In addition, 18O-TFMSA was synthesized as internal standard for HMSA analysis. Standard synthesis was performed in two steps. At first, commercially available methanesulfonyl chlorides (chloromethanesulfonyl chloride, dichloromethanesulfonyl chloride and trifluoromethanesulfonyl chloride) were hydrolyzed at 60°C. The equimolar formation of chloride during the hydrolysis reaction allowed an indirect monitoring of the reaction with ion chromatography. The reactions were stopped after 24 h (ClMSA) and 72 h (Cl2MSA), respectively, when the chloride concentration corresponded to a complete hydrolysis. After purification of the products with SPE utilizing WAX cartridges, the identity of the reaction products was unequivocally confirmed with HMRS (Figure S1, Table S2 and S3), MS/HRMS (Figure S2 and S3) and NMR (Figure S6 and S7) Trifluoromethanesulfonyl chloride was hydrolyzed in H218O and not controlled with IC, but instead stopped when the water-insoluble educt had completely disappeared. Its structure was confirmed with HRMS and MS/HRMS.

𝛽𝑝𝑟𝑒 ― 𝛽𝑛𝑎𝑡𝑖𝑣𝑒 𝛽𝑆𝑡𝑑.

𝑖𝑛𝑠𝑡𝑟𝑢𝑚𝑒𝑛𝑡𝑎𝑙 𝐿𝑂𝑄

𝛽𝑝𝑟𝑒 ― 𝛽𝑛𝑎𝑡𝑖𝑣𝑒 𝛽𝑝𝑜𝑠𝑡 ― 𝛽𝑛𝑎𝑡𝑖𝑣𝑒

(

𝛽𝑓𝑜𝑟𝑡,𝑆𝑡𝑑.

in

∗ 100%

100 % 𝑎𝑝𝑝𝑎𝑟𝑒𝑛𝑡 𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑦 (%)

∗2

∗ 100%

𝛽𝑝𝑜𝑠𝑡 ― 𝛽𝑛𝑎𝑡𝑖𝑣𝑒

𝛽𝑓𝑜𝑟𝑡 ― 𝛽𝑛𝑎𝑡𝑖𝑣𝑒

18O-TFMSA

𝛽𝑆𝑡𝑑.

∗

∗ 100%

) ― 100%

100 % 𝑎𝑝𝑝𝑎𝑟𝑒𝑛𝑡 𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑦 (%)


∗ 100%


212 213 214 215 216 217 218 219 220 221 222 223 224 225 226

Aliquots of the purified chlorinated methanesulfonic acids were used as educts for the synthesis of their brominated congeners. Halogen exchange from chlorine to bromine was achieved via the Finkelstein reaction in the presence of a molar excess of bromide at elevated temperatures (165°C) in a microwave. The reaction was monitored qualitatively with HRMS. A slow reaction rate was observed for dichloromethanesulfonic acid, and thus its reaction was stopped after 40.5 h when a partial conversion (19.1%, determined by qNMR) to its brominated congeners was achieved. The reaction of ClMSA to BrMSA reached a molar conversion rate of 77.5% (determined by qNMR) After purification (WAX-SPE) and structure confirmation (HRMS see Figure S1, Table S4 and S5, MS/HRMS see Figure S4 and S5 and NMR, see Figure S6 and S7), all chlorinated and brominated HMSAs were quantified with qNMR utilizing MSA as standard (Figure S6). 1H-NMR signals were used for quantification and the structure was further confirmed with 13C-NMR data for ClMSA and BrMSA. 1H-NMR spectra showed a higher data quality and thus the time-intensive generation of 13C-NMR data was omitted for dihalogenated HMSAs. Probably owing to the large size of bromine atoms and a resulting steric hindrance, dibromomethanesulfonic acid was only synthesized in trace amounts and could thus not be quantified with NMR.

227

3.2 Method development and quality control

228 229 230 231 232 233 234 235 236 237 238

The synthesized standards were utilized to implement HMSAs and the internal standard 18O-TFMSA in a sample pre-treatment (mlSPE and evaporation) and HILIC-sMRM methodology that is dedicated to the analysis of persistent and polar water contaminants, including TFMSA[20]. For chloro- and dichloromethanesulfonic acid, the SO3●- radical anion and Cl- were the fragments of choice for MRM transitions. For brominated species, the cleavage of Br- was dominant and thus the only fragment of significant intensity. Therefore, the cleavage of 79Br- and 81Br- from their respective bromomethanesulfonic acid isotopolog was utilized, which in turn resulted in a reduced selectivity compared to the formation of two structurally independent fragments. For chlorobromomethanesulfonic acid, formation of 79Br- from the 79Br/35Cl isotopolog and formation of 79Br- from the 79Br/37Cl isotopolog were selected as MRM transitions which led to an increased selectivity despite the fact that Br- was available as only fragment of significant intensity.

239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258

To assess the method performance, apparent recoveries (sum of recovery and matrix effects) and LOD values (Figure 1 A and B) of HMSAs were determined for both deployed enrichment methods in 70 water samples, ranging from waste water treatment plant effluent to finished drinking water. Monohalogenated HMSAs showed apparent recoveries below 100%, especially after enrichment with mlSPE while their dihalogenated congeners showed apparent recoveries well above 100%. This behavior was investigated in detail with five matrices including artificial freshwater, drinking waters, and ground waters that were prepared in triplicate. One replicate was spiked before enrichment, one replicate was spiked after enrichment, and one replicate was not spiked to establish the base concentration in each matrix. This allowed a separate assessment of the recovery (Figure 1 C, comparison of replicate spiked before enrichment to replicate spiked after enrichment) and the matrix effects (Figure 1 D, comparison of replicate spiked after enrichment to standard), and demonstrated that SPE recoveries are partially poor for monohalogenated HMSAs while good recoveries close to 100% were achieved for all analytes with the evaporation method. The highly variable apparent recoveries well above 100% that were observed for dihalogenated HMSAs were clearly attributed to pronounced ion enhancement, which demonstrates that matrix effects may be highly relevant in the analysis of very polar chemicals. These pronounced, highly variable matrix effects resulted in LOQ values that differ significantly between samples, and thus sample specific LOQ values had to be used. Coelution with inorganic ions is a potential cause for this pronounced ion enhancement and was also observed in the analysis of polar pharmaceuticals and pharmaceutical metabolites by Boulard et al.[21], who counteracted this effect with isotope-labeled internal standards. Appropriate internal ACS Paragon Plus Environment

Page 6 of 16

Page 7 of 16

259 260 261


standards were not commercially available for the analyzed HMSAs and the synthesized 18O-TFMSA showed a different behavior in the test matrices and was thus not deemed suitable to compensate matrix effects. 18O-TFMSA was consequently only utilized to compensate volumetric errors.

262 263 264 265 266 267

Figure 1: Mean and standard deviation of apparent recoveries (A) and LOQs (B) throughout all 70 water samples. The influence of recoveries (C) and matrix effects (D) were investigated in detail with a set of five test matrices. The relative deviation between double determinations (E) was investigated with nine samples. Quantification of eight fortified samples was utilized to assess the trueness (F) of the method. Samples below the lower LOQ or above the upper LOQ were excluded. Dashed lines represent desired values (A, C, D, and F) or upper limits of acceptance (E).

268 269

In order to compensate for the absence of isotope-labeled internal standards, a spiked replicate of each sample was prepared to calculate sample-specific apparent recoveries, which were employed to ACS Paragon Plus Environment


270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288

compensate varying recoveries and matrix effects. To judge the effectiveness of these measures, independent double determinations of nine samples of the original sample set were prepared and demonstrated to deviate predominantly by less than 20% from the mean (Figure 1 E). To assess the trueness of the results (Figure 1 F), eight samples were fortified with 25 ng/L HMSAs each (7.5 ng/L for BrClMSA) for enrichment with mlSPE and 1 µg/L each (300 ng/L BrClMSA) for enrichment with evaporation, respectively. These concentrations were found to be insufficient for enrichment of BrMSA and ClMSA with mlSPE. Mean corrected concentrations (subtraction of concentration in unspiked sample, corrected with apparent recovery) of the fortified samples were found to deviate only slightly from the true values with standard deviations between 10 and 20%. The evaporation method was shown to generate the more reliable results and was thus used whenever the analyte concentration was sufficient, otherwise the mlSPE method was used (for a detailed overview of the enrichment method used for each analyte in each sample and the corresponding sample-specific LOQs see Table S12). Overall, these results demonstrate that the utilization of sample specific apparent recoveries determined at one concentration was sufficient to compensate the pronounced matrix effects observed for dihalogenated HMSAs, and thus the developed method is suitable to generate results which are expected to be at least within 20% of the true value. Given the complexity of the method, the lack of commercially available (isotope-labeled) standards and the wide concentration range covered during this method assessment, the method at hand was deemed suitable for the quantification of HMSAs within these specifications.

289 290 291 292 293

The quantification of ClMSA was hindered by a procedural blank, which was more pronounced after enrichment with mlSPE due to the higher enrichment factor realized. The mean and standard deviation of the ClMSA blank were calculated from ten independent blank replicates prepared on five days. The mean ClMSA blank was subtracted from all samples and the tenfold standard deviation of the blank was set as LOQ.

294 295

3.3 Quantification of HMSAs in drinking water production plants and tap water samples

296 297 298 299 300 301 302 303 304 305 306

The HILIC-sMRM method was deployed to quantify the synthesized HMSAs throughout four DWPPs with diverse treatment options (Figure 2 A, only raw water and finished drinking water shown). As a consequence of the pronounced procedural blank, ClMSA was rarely detected above the LOQ. A formation of HMSAs in the low µg/L range was observed after water chlorination, confirming the initial assumption that HMSAs are DBPs. The DWPPs were sampled in three consecutive months from April to June for DWPP 1, 2, and 3 and June to August for DWPP 4. In finished drinking waters a steady increase in HMSA concentrations was observed from April to June. While this corresponds roughly to increasing total organic carbon (TOC) values in the corresponding raw water (e.g. DWPP 3: April 1.64 mg/L, May 1.88 mg/L, and June 2.13 mg/L), the information is insufficient to clearly indicate a seasonal trend, for which long term studies will be required. The low HMSA concentrations in the raw and drinking water of DWPP 3 taken in April cannot be explained by the variations in the TOC value.

307 308 309 310 311

HMSAs were not only detected in finished drinking water but also in the raw waters of DWPP 1 and 3, which are located in regions where chlorination is prevalent. While the concentrations were significantly lower than for finished drinking water, quantities of up to 100 ng/L were observed for the most prominent congeners, which could originate from an influx of chlorinated water, either from chlorination in waste water treatment plants or during drinking water production.

312 313 314 315

The analysis of 14 tap water samples taken in six countries revealed a wide spread presence of HMSAs (Figure 2 B), and an apparent prevalence of dihalogenated congeners. While Cl2MSA was the most prevalent congener in most tap water samples, BrClMSA was present in higher concentrations in Barcelona, Jinan, and the USA samples, which indicates elevated bromine concentrations in these tap ACS Paragon Plus Environment

Page 8 of 16

Page 9 of 16


316 317 318 319 320 321 322 323

waters. In some samples a co-eluting interference impeded the accurate quantification of BrMSA, however estimated concentrations were in any case below 100 ng/L, which is in compliance with the majority of the analyzed tap water samples, where BrMSA had only a minor contribution to the total HMSA load. Therefore, monohalogenated HMSAs, and especially BrMSA, seem to be much less prevalent in tap water samples (4.3% BrMSA, 94% Cl2- and BrClMSA) compared to finished drinking water directly taken from DWPPs (31.2% BrMSA, 62.3% Cl2- and BrClMSA), which may be explained by a reaction of monohalogenated HMSAs with residual chlorine during drinking water distribution forming their dihalogenated congeners.

324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345

With the exception of Frankfurt (170 ng/L), Wiesbaden (70 ng/L), and London (330 ng/L) total HMSA concentrations were in the single to low double digit µg/L range. The low concentrations detected in Frankfurt, Wiesbaden, and London are at least partially reflected by the concentration of trihalomethanes, taken from readily available analytical data, either from water suppliers or annual water quality reports, which may be used as an indicator for the total DBP load (e.g. Dresden 5.2 µg/L[22], Wiesbaden 0.76 µg/L[23]). While the comparability of these literature data is limited due to variations in the reporting period, the number of samples, and the involved laboratories, they provide at least a rough estimate of the prevalence of DBPs. The highest concentrations for HMSAs were detected in Duarte with a total HMSA concentration of 11.5 µg/L, and thus significantly higher concentrations than in the other samples taken in California, USA (Glendora: 0.8 µg/L, North Hollywood: 1.2 µg/L). This, however, is not reflected in the prevalence of frequently monitored haloacetic acids (annual range Duarte: 10% of formation potential), are shown as a grey triangle. Error bars indicate the range of double determinations and a green asterisk marks treatment steps during which a significant reduction of the formation potential took place. TOC values in these samples are depicted as orange dots.

406

3.6 Perspectives and implications

407 408 409 410 411 412 413 414 415 416 417 418 419 420

HRMS and NMR data demonstrated that four HMSAs were successfully synthesized, which is a strong indicator that the methodology presented herein may also be utilized to synthesize other chlorinated, brominated and iodinated congeners, and thus extend the existing method to monitor a broader spectrum of HMSAs. Despite the current limitation of the quantitative method to four HMSA congeners, concentrations in tap water were regularly in the µg/L range, indicating that they may be a significant contribution to the total DBP load of disinfected drinking water. The pronounced ion enhancement for dihalogenated HMSAs may point towards a high relevance of matrix effects in the analysis of very polar chemicals that warrants further investigation. Initial QSAR data generated with the EPA Toxicity Estimation Software Tool[28] conclude that HMSAs are not toxic. However, since very polar or ionic chemicals such as HMSAs may be out of the scope of commercially available QSAR tools[29], which results in less reliable predictions, and experimental data is not available, potential adverse health effects especially caused by long term exposure cannot be excluded at this point. Thus, experimental data is required to assess the potential risk posed by these high concentration drinking water contaminants and evaluate if mitigation efforts should be undertaken.

421 422 423 424 425 426 427 428

In addition to drinking water HMSAs were also present in two surface water samples and may thus also be environmental water contaminants. This may potentially act as one of few examples of the contamination of the aquatic environment through drinking water. The high polarity and assumed persistency of HMSAs may lead to their classification as “very persistent very mobile” (vPvM) chemicals, which recently generated a high scientific[30] and regulatory[31] interest as a consequence of their ability to spread unimpaired throughout the water cycle. The sorption of HMSAs to activated carbon is limited, and they are thus likely not amenable to commonly deployed TOX procedures. This is a strong indicator that the “analytical gap” postulated by Reemtsma et al.[30] for the most polar ACS Paragon Plus Environment

Page 12 of 16

Page 13 of 16

429 430 431 432 433


water contaminants is not only relevant for trace analytical methods, but also for sum parameters as long as they rely on an enrichment of the analytes from water. The mlSPE method[19] utilized herein for the enrichment of HMSAs alongside other polar contaminants from aqueous matrices may be useful to extend the polarity range of TOX measurements and thus narrow the “analytical gap” for this important sum parameter.

434 435

Acknowledgement

436 437 438 439 440 441 442 443

The authors thank the European Union Joint Programming Initiative “Water Challenges for a Changing World” (Water JPI) and the BMBF for funding the PROMOTE project (FKZ: 02WU1347B) as well as Annika Harloff (Hochschule Fresenius) who participated in the synthesis of the standards and Victoria Zilles (Hochschule Fresenius) who performed a large share of the lab work during the monitoring. In addition, we thank Oasen and Hessenwasser for each supplying the samples for one of the drinking water production plants, Vittorio Albergamo (Institute for Biodiversity and Ecosystem Dynamics – University of Amsterdam) for organizing the shipment of the Oasen samples, and Stefanie Schulze (Helmholzzentrum für Umweltforschung) for coordinating the DWPP sampling campaign.

444 445

Supporting Information

446 447

Summary of abbreviations, HRMS data, MS/HRMS data, and NMR data for HMSA structure confirmation, information about water samples, method parameters, sample specific LOQs

448 449

Literature

450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474

[1] M.J. Nieuwenhuijsen, M.B. Toledano, N.E. Eaton, J. Fawell, P. Elliott, Chlorination disinfection byproducts in water and their association with adverse reproductive outcomes: a review, Occup. Environ. Med., 57 (2000) 73-85. [2] F. Bove, Y. Shim, P. Zeitz, Drinking water contaminants and adverse pregnancy outcomes: a review, Environ. Health Perspect., 110 (2002) 61-74. [3] K. Waller, S.H. Swan, G. DeLorenze, B. Hopkins, Trihalomethanes in Drinking Water and Spontaneous Abortion, Epidemiology, 9 (1998) 134-140. [4] C.M. Villanueva, K.P. Cantor, S. Cordier, J.J. Jaakkola, W.D. King, C.F. Lynch, S. Porru, M. Kogevinas, Disinfection byproducts and bladder cancer: a pooled analysis, Epidemiology, 15 (2004) 357-367. [5] S. Regli, J. Chen, M. Messner, M.S. Elovitz, F.J. Letkiewicz, R.A. Pegram, T.J. Pepping, S.D. Richardson, J.M. Wright, Estimating Potential Increased Bladder Cancer Risk Due to Increased Bromide Concentrations in Sources of Disinfected Drinking Waters, Environmental Science & Technology, 49 (2015) 13094-13102. [6] S.D. Richardson, T.A. Ternes, Water Analysis: Emerging Contaminants and Current Issues, Analytical Chemistry, 90 (2018) 398-428. [7] S.W. Krasner, H.S. Weinberg, S.D. Richardson, S.J. Pastor, R. Chinn, M.J. Sclimenti, G.D. Onstad, A.D. Thruston, Occurrence of a new generation of disinfection byproducts, Environmental Science & Technology, 40 (2006) 7175-7185. [8] D. Vughs, K.A. Baken, A. Kolkman, A.J. Martijn, P. de Voogt, Application of effect-directed analysis to identify mutagenic nitrogenous disinfection by-products of advanced oxidation drinking water treatment, Environmental science and pollution research international, (2016). [9] Y. Tang, Y. Xu, F. Li, L. Jmaiff, S.E. Hrudey, X.-F. Li, Nontargeted identification of peptides and disinfection byproducts in water, Journal of Environmental Sciences, 42 (2016) 259-266. [10] T. Gong, X. Zhang, Y. Li, Q. Xian, Formation and toxicity of halogenated disinfection byproducts resulting from linear alkylbenzene sulfonates, Chemosphere, 149 (2016) 70-75. ACS Paragon Plus Environment


475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526

[11] J. Han, X. Zhang, J. Liu, X. Zhu, T. Gong, Characterization of halogenated DBPs and identification of new DBPs trihalomethanols in chlorine dioxide treated drinking water with multiple extractions, Journal of Environmental Sciences, 58 (2017) 83-92. [12] M. Nihemaiti, J. Le Roux, C. Hoppe-Jones, D.A. Reckhow, J.-P. Croué, Formation of Haloacetonitriles, Haloacetamides, and Nitrogenous Heterocyclic Byproducts by Chloramination of Phenolic Compounds, Environmental Science & Technology, 51 (2017) 655-663. [13] N. Negreira, J. Regueiro, M. López de Alda, D. Barceló, Degradation of the anticancer drug erlotinib during water chlorination: Non-targeted approach for the identiﬁcation of transformation products, Water research, 85 (2015) 103-113. [14] I. Carpinteiro, R. Rodil, J.B. Quintana, R. Cela, Reaction of diazepam and related benzodiazepines with chlorine. Kinetics, transformation products and in-silico toxicological assessment, Water Research, 120 (2017) 280-289. [15] M. Gonsior, C. Mitchelmore, A. Heyes, M. Harir, S.D. Richardson, W.T. Petty, D.A. Wright, P. Schmitt-Kopplin, Bromination of Marine Dissolved Organic Matter following Full Scale Electrochemical Ballast Water Disinfection, Environmental Science & Technology, 49 (2015) 90489055. [16] H. Zhang, Y. Zhang, Q. Shi, H. Zheng, M. Yang, Characterization of Unknown Brominated Disinfection Byproducts during Chlorination Using Ultrahigh Resolution Mass Spectrometry, Environmental Science & Technology, 48 (2014) 3112-3119. [17] E.E. Lavonen, M. Gonsior, L.J. Tranvik, P. Schmitt-Kopplin, S.J. Köhler, Selective Chlorination of Natural Organic Matter: Identification of Previously Unknown Disinfection Byproducts, Environmental Science & Technology, 47 (2013) 2264-2271. [18] D. Zahn, T. Frömel, T.P. Knepper, Halogenated methanesulfonic acids: A new class of organic micropollutants in the water cycle, Water Research, 101 (2016) 292-299. [19] N. Köke, D. Zahn, T.P. Knepper, T. Frömel, Multi-layer solid-phase extraction and evaporation— enrichment methods for polar organic chemicals from aqueous matrices, Analytical and Bioanalytical Chemistry, 410 (2018) 2403-2411. [20] U. Berger, S. Schulze, D. Zahn, R. Montes, J. B. Quintana, T. P. Knepper, T. Reemtsma, in preparation. [21] L. Boulard, G. Dierkes, T. Ternes, Utilization of large volume zwitterionic hydrophilic interaction liquid chromatography for the analysis of polar pharmaceuticals in aqueous environmental samples: Benefits and limitations, Journal of Chromatography A, 1535 (2018) 27-43. [22] https://www.drewag.de/wps/portal/drewag/cms/menu_main/privatkunden/produkte/wasser#suppl yzone4, accessed at 09.04.2018. [23] https://www.eswe-versorgung.de/privatkunden/wassernetz/trinkwasserqualitaet.html, accessed at 10.04.2018. [24] https://dnnh3qht4.blob.core.windows.net/portals/2/2017%20CCRs/CA%20%20Duarte%202016%20CCR%20Revised.pdf?sr=b&si=DNNFileManagerPolicy&sig=m6AQOUWOZ7VX Ccgc8vLOqjpX536P1WNJ0v/bdacZQl4=, accessed at 09.04.2018. [25] https://files.swwc.com/ca/ccr/ccr-glendora-2014.pdf, accessed at 09.04.2018. [26] https://www.csun.edu/sites/default/files/DWP%20Consumer%20Confidence%20Report%20DWQR% 202016%20Web%20High%20Resolution.pdf, accessed at 09.04.2018. [27] W. Zong, F. Sun, H. Pei,W. Hu, R. Pei, Microcystin-Associated Disinfection by-Products: The Real and Non-Negligible Risk to Dinking Water Subject to Chlorination. Chemical Engineering Journal, 279 (2015) 498-506. [28] E. Toxicity Estimation Software Tool, v 4.2, https://www.epa.gov/chemical-research/toxicityestimation-software-tool-test. [29] H.P.H. Arp, T.N. Brown, U. Berger, S.E. Hale, Ranking REACH registered neutral, ionizable and ionic organic chemicals based on their aquatic persistency and mobility, Environmental Science: Processes & Impacts, 19 (2017) 939-955.


Page 14 of 16

Page 15 of 16

527 528 529 530 531


[30] T. Reemtsma, U. Berger, H.P.H. Arp, H. Gallard, T.P. Knepper, M. Neumann, J.B. Quintana, P.d. Voogt, Mind the Gap: Persistent and Mobile Organic Compounds—Water Contaminants That Slip Through, Environmental Science & Technology, 50 (2016) 10308-10315. [31] Workshop: PMT and vPvM substances under REACH. Voluntary measures and regulatory options to protect the sources of drinking water, Berlin, Germany

532 533

TOC Graphic:

534



299x162mm (300 x 300 DPI)


Page 16 of 16

Halomethanesulfonic AcidsâA New Class of Polar Disinfection

Recommend Documents

Halomethanesulfonic AcidsâA New Class of Polar Disinfection