Chloramination of iopamidol- and bromide-spiked waters containing natural organic matter

Iopamidol (an iodinated X-ray contrast medium) and bromide are precursors in the formation of halogenated disinfection byproducts (DBPs). The interactions of these precursors are vital to elucidate the formation of halogenated DBPs during chloramination. This work investigated the formation of total organic halogen and select individual DBPs in two laboratory-chloraminated source waters containing iopamidol and bromide. Experiments were carried out in batch reactors containing Barberton source water (BSW) and Cleveland BSW (CSW), spiked with iopamidol (5 μM), bromide (15 μM), and 100 μM monochloramine. Total organic iodine concentrations were approximately equal regardless of source water since they are mostly unreacted iopamidol and iopamidol DBPs. Almost equal amounts of total organic chlorine (3–4 nM) were produced in the source waters, but higher quantities of total organic bromine were formed in BSW than CSW. Substantial quantities of regulated trihalomethanes (THMs) and haloacetic acids (HAAs) were formed in the source waters, along with appreciable concentrations of iodinated trihalomethanes (CHBrClI, CHCl2I, and CHBr2I). Low concentrations of iodo-HAAs were detected, especially at low pH. Overall, bromide concentrations appeared to suppress iodo-DBP formation during chloramination of iopamidol in the presence of natural organic matter. A good correlation (R1⁄4 0.801) between the yields of regulated DBPs and iodo-DBPs was observed.


INTRODUCTION
To alleviate the risk of water-borne diseases, chlorine has been employed to disinfect potable water since before the 1900s. Achieving adequate disinfection is paramount to public health; however, addition of chlorine results in the formation of appreciable concentrations of disinfection byproducts (DBPs), especially regulated trihalomethanes (THMs) and haloacetic acids (HAAs). Difficulty in complying with regulations has led some water utilities in the USA to switch from free chlorine to monochloramine (NH 2 Cl) as a secondary disinfectant (Luh & Marinas ). Generally, NH 2 Cl is used for maintaining residual disinfectant in distribution systems since it is less reactive with natural organic matter (NOM) and results in lower concentrations of regulated DBPs (Vikesland et al. ). However, studies have revealed that other DBPs are formed in chloraminated waters that elicit greater toxicity than regulated DBPs (Plewa et  The main objective of this study was to investigate the formation of total organic halogen (TOX) and the distribution of individual DBPs in chloraminated source waters containing iopamidol and Br À . The study specifically investigated how different NOM sources and pH affected halogen-specific TOX and DBP formation. Since both constituents are precursors in the formation and speciation of DBPs and bromide oxidation/incorporation is faster than iopamidol transformation, we wanted to investigate whether the presence of Br À might suppress the formation of choro-iodo/iodo-DBPs in chloraminated iopamidol-containing waters.

Reagents
All reagents used were of the highest available purity and are described in detail elsewhere (Ackerson et al. ). Purified water (18.2 MΩ-cm) used for preparing buffer solutions was produced from Barnstead ROPure Infinity/NANOPure system (Barnstead-Thermolyne Corp. Dubuque, IA, USA). An Orion 5-star pH meter equipped with a Ross ultra-combination electrode (Thermo Fisher Scientific, Waltham, MA) was used to measure pH. Phosphate buffer was used to maintain pH at 6.5 and 7.5, and borate buffer for pH 8.5 and 9.0. pH adjustments were carried out with 1 N H 2 SO 4 and 1 N NaOH. All glassware and polytetrafluoroethylene (PTFE) containers were soaked in a chlorine bath for 24 hours, rinsed with large amounts of purified water and dried before use.

Experimental methods
To investigate the impact of bromide and iopamidol on DBP and TOX formation and speciation, 1,000-and 250-mL Erlenmeyer flasks were filled with BSW or CSW dosed with buffer solutions (1 mM for TOX and 4 mM for DBPs), 5 μM iopamidol, and 15 μM Br À concentrations at pH 6.5-9.0. Phosphate buffer was used to maintain pH 6.5 and 7.5, while borate buffer was used at pH 8.5 and 9.0.
Monochloramine was added to achieve a final concentration of 100 μM under rapid mixing conditions on a magnetic stir plate with a PTFE stir bar. After uniform mixing, aliquots of the samples were transferred into 128-mL amber bottles and 40-mL amber vials with PTFE septa screw caps and stored headspace-free in the dark at 25 C for 0-48 hours prior to DBP and TOX analyses, respectively.

Analytical methods
Extracted and combusted TOX samples were analyzed on a Dionex ICS-3000 column (Dionex Corporation, Sunnyvale, CA) with conductivity detector and an ASRS ® 300 4 mm anion self-regenerating suppressor. Detection of TOX was accomplished with an AS20 analytical column (4 × 250 mm) and a guard column (Dionex Corporation, Sunnyvale, CA) with 10 mM KOH as the mobile phase (flow rate of 1 mL/ min). The specific TOX species analyzed were total organic bromine (TOBr), total organic chlorine (TOCl), and total organic iodine (TOI). TOBr, TOCl, and TOI were detected respectively on the ICS-3000 as Br À , Cl À and I À . The recoveries of all TOX species were 98-100% with a limit of quantitation of 0.50 μM and a detection limit of 0.25 μM.
Using liquid-liquid extraction, samples for DBP analysis were extracted in methyl tert butyl ether (MtBE) with 1,2dibromopropane as the internal standard. The extracted organic phase was divided into 1.5 mL (for THM and HAN analyses) and 0.5 mL aliquots for derivatization of HAAs.
Derivatization was conducted with diazomethane to form methyl esters. Table S2 shows the lists  Gas chromatography-tandem mass spectrometry (GC-MS/ MS) with a ThermoScientific Quantum GC-triple quadrupole mass spectrometer coupled to a TRACE GC Ultra gas chromatograph (ThermoScientific, Waltham, MA) was used to analyze iodo-HAAs (Table S2). The temperature program for the iodo-HAAs analysis is shown in Table S5 while the Table S6.

RESULTS AND DISCUSSION
TOX speciation TOI did not change substantially when the source waters were spiked with 5 μM iopamidol, 15 μM Br À , and 100 μM NH 2 Cl. In both Barberton and Cleveland source waters, TOI exhibited marginal losses at the end of 48 hours ( Figure 1 and Figure S2, Supplementary Information). This is due to unreacted iopamidol, iopamidol transformation products (IDOL-DBPs), or iodide released from the IDOL-DBPs that has been oxidized and incorporated into DBP precursors in the NOM structure. Monochloramine undergoes hydrolysis, especially at pH < 8.5 (Vikesland et al. ). The hydrolysis product, HOCl, is in equilibrium with OCl À , which can act as a nucleophile, reacting with one of the amide side chains of iopamidol and forming IDOL-DBPs (Wendel et al. ).
The DBPs continue to react with the chlorinated oxidants present and release the iodine from the IDOL-DBP ring structure ( Figure S1). It has been observed that chloramine cannot degrade iopamidol effectively (Tian et al. ). The iodide is then oxidized to HOI and subsequently incorporated into NOM structures, possibly forming different TOI structures.
Like TOI, TOBr did not show substantial differences as a function of pH. However, different quantities of TOBr were produced with respect to the different source waters ( Figure 1 and Figure S2). An appreciable level of bromine incorporation (67-90%) was found in BSW. By contrast, only 25-32% bromide was incorporated into chloraminated CSW. The amounts of TOBr generated during chloramination were 12-35% and two-to threefold less in BSW and CSW, respectively, than when these source waters were chlorinated in a previous study (Ackerson et al. ).
Chlorination and chloramination of organic matter isolates in finished water from Bloomington water treatment plant (IL), containing Br À and iodide also showed marginal a difference between TOBr levels (Yang et al. ). Other studies have likewise reported both marginal and substantial differences between TOBr concentrations in chlorinated

DBP formation and speciation
Addition of monochloramine to source waters containing iopamidol and bromide generated both regulated THMs and iodo-THMs. Generally, most of the regulated THMs exhibited a slow initial formation, especially at high pH, probably due to monochloramine being the dominant oxidant and being extremely stable at high pH ( Jafvert & Valentine ).
All four regulated THMs -CHCl 3 , CHBrCl 2 , CHBr 2 Cl, and CHBr 3were detected in both source waters. Appreciable quantities of CHCl 3 were formed in BSW compared to CSW (Figure 2). Nonetheless, formation rates were faster in CSW than BSW. Almost 74-91% and 50-64% of CHCl 3 formed at 48 hours was detected at 6 hours in CSW and BSW, respectively, except for CSW at pH 6. Since the direct reaction of monochloramine with iopamidol is minimal NHBrCl and other bromamines have been assumed to exhibit slow reaction with NOM (Duirk & Valentine ). This may account for the slow rate of formation seen for CHBr 3 especially as pH increased, reducing brominated oxidant formation.
The formation of two bromo-chloro-THMs was observed in both source waters. However, pH and source water impacted the concentrations of the mixed brominated/chlorinated THMs formed. Mixed halo-THM (i.e., CHBr 2 Cl and CHBrCl 2 ) formation was substantially greater in BSW than CSW ( Figure S3). This was likely the result of BSW being more reactive with brominated/chlorinated oxidants than CSW and IDOL-DBPs. Ackerson et al. () have previously shown that BSW has more DBP precursors than CSW and IDOL-DBPs, generally resulting in unknown TOX and unknown DBPs. Therefore, the formation of these species may be a simultaneous incorporation of chlorine and bromine found in NH 2 Cl, HOCl, NHBrCl, and HOBr.
Since the reactivity of each oxidant or haloamine species is pH dependent, their relative abundance at each pH will vary. This pH-dependent oxidant formation would likely have impacted the formation of these THMs. iopamidol was CHBrClI. When iodide was present as the primary iodo-DBP precursor, they found that iodide favored CHI 3 formation in fulvic and humic acids, as well as the same source waters from China. It is possible that HOI formed from IDOL-DBPs, bromamines, and the other oxidants listed above may have concurrently incorporated into NOM to produce CHBrClI. CHCl 2 I was detected in both BSW and CSW ( Figure 3). Also, up to 6.5 nM of CHClI 2 was found only in BSW at pH 6.5 and 7.5.
CHBr 2 I was produced in both source waters (Figure 3), whereas CHBrI 2 was identified only in BSW in small quantities at pH 6.5. All the iodo-THMs exhibited a decreasing formation trend with increasing pH. The same iodo-THM species were formed in two source waters from China spiked with iopamidol, Br À , and NH 2 Cl at pH 7 (Wang    suggested that the active bromine species with a valence of þ1 (designated as Br(I)), which is formed by the reactions between Br À and NH 2 Cl, may be very reactive with DHAA precursors. These DHAA precursors have been assumed to contain fast-reacting chromophores, which enhance bromine incorporation (Korshin et al. ).
This was consistent with the findings that the depletion of the fast-reacting chromophores in NOM was as a result of the rapid formation of DCAA (Korshin et al. ).
About 3-23% of the total DCAA observed in CSW was pro-  The percentage of each class of DBPs was >1% of TOCl at pH 6.5 and 7.5 in BSW, except chloro-iodo-THMs (C,I-THMs) and chloro-iodo-HAA (C,I-HAA), which produced <1% of TOCl ( Figure S8). As the pH increased, the percentage of each class of DBPs decreased, while unknown TOCl (UTOCl) increased substantially. At pH 6.5, UTOCl represented approximately 47% of TOCl, which increased with pH to 96% UTOCl at pH 9. Only chloro-THMs and chloro-HAAs produced >1% of TOCl at all pH levels in BSW. A lower percentage of TOCl formed in CSW was composed of the specific DBPs quantified, and UTOCl increased with pH ( Figure S8). In general, the classes of chlorinated DBPs produced in CSW were <1% except for B,C,I-THMs and chloro-HAAs at low pH. The low formation of the measured chloro-DBPs in the source waters could be ascribed to the weak reactivity of NH 2 Cl with NOM, resulting in predominantly unknown chloro-DBPs.
Relationship between the yields of regulated DBPs and

Iodo-DBPs
The two classes of DBPs identified in this study were regu- The study showed a strong positive correlation (R 2 ¼ 0.801) between the yields of regulated DBPs and

CONCLUSIONS
In the presence of bromide and iopamidol, chloramination of source waters produced significant quantities of halogen-specific TOX and DBPs. Of importance to this study were iodo-DBPs, which were formed in significant quantities at low pH in particular. Iodo-DBPs are central to the study because they are highly genotoxic and cytotoxic Although substantial concentrations of iodo-DBPs were formed due to high amounts of iopamidol and bromide used in these controlled laboratory reactions, lower concentrations would be expected at environmentally relevant concentrations. However, due to the high toxic potency of iodo-DBPs, their toxicity would still be of concern