Article pubs.acs.org/est
Transformation of Iopamidol during Chlorination Friedrich M. Wendel,† Christian Lütke Eversloh,† Edward J. Machek,‡ Stephen E. Duirk,‡ Michael J. Plewa,§ Susan D. Richardson,∥ and Thomas A. Ternes*,† †
Water Chemistry Department, Federal Institute of Hydrology (BfG), Am Mainzer Tor 1, D-56068 Koblenz, Germany Department of Civil Engineering, University of Akron, Akron, Ohio 44325, United States § Department of Crop Sciences and Safe Global Water Institute and NSF Science and Technology Center of Advanced Materials for the Purification of Water with Systems, University of Illinois at Urbana−Champaign, 1101 West Peabody Drive, Urbana, Illinois 61801, United States ∥ Department of Chemistry and Biochemistry, University of South Carolina, 631 Sumter St., Columbia, South Carolina 29208, United States ‡
S Supporting Information *
ABSTRACT: The transformation of the iodinated X-ray contrast media (ICM) iopamidol, iopromide, iohexol, iomeprol, and diatrizoate was examined in purified water over the pH range from 6.5 to 8.5 in the presence of sodium hypochlorite, monochloramine, and chlorine dioxide. In the presence of aqueous chlorine, only iopamidol was transformed. All other ICM did not show significant reactivity, regardless of the oxidant used. Chlorination of iopamidol followed a second order reaction, with an observed rate constant of up to 0.87 M−1 s−1 (±0.021 M−1 s−1) at pH 8.5. The hypochlorite anion was identified to be the reactive chlorine species. Iodine was released during the transformation of iopamidol, and was mainly oxidized to iodate. Only a small percentage (less than 2% after 24 h) was transformed to known organic iodinated disinfection byproducts (DBPs) of low molecular weight. Some of the iodine was still present in high-molecular weight DBPs. The chemical structures of these DBPs were elucidated via MSn fragmentation and NMR. Side chain cleavage was observed as well as the exchange of iodine by chlorine. An overall transformation pathway was proposed for the degradation of iopamidol. CHO cell chronic cytotoxicity tests indicate that chlorination of iopamidol generates a toxic mixture of high molecular weight DBPs (LC50 332 ng/μL).
■
INTRODUCTION Iodinated X-ray contrast media (ICM) are commonly used to enable the medical imaging of soft tissues (e.g., organs and blood vessels). Up to 120 g of the contrast medium is used in a single application, and about 75 million of these diagnosis examinations are performed with use of contrast media globally each year.1 ICM are designed to be inert in the human body and are excreted nonmetabolized within 24 h. Conventional wastewater treatment plants (WWTPs) have shown to be unsuited for a complete elimination of ICM, which have thus been found in WWTP effluents and surface water in concentrations up to several μg/L.2,3 Iopamidol for instance was found in the effluent of a WWTP in a concentration of 15 μg/L.2 Even in drinking water, ICM have been detected up to the higher ng/L range, e.g., up to 98 ng/L of iopamidol were found in the finished drinking water in a German waterworks.4 Furthermore, the removal of ICM is often accompanied by the occurrence of transformation products (TPs) or disinfection byproducts (DBPs). A wide range of biological TPs of ICMs were identified and detected after soil passage and in effluents of conventional WWTPs.3,5 Ozonation of iomeprol © 2014 American Chemical Society
containing waters was reported to primarily result in oxidation of the side chains.6 The release of inorganic iodine from ICM by ozonation or combined ozone and hydrogen peroxide treatment was shown by Ning and Graham.7 Dehalogenation reactions occurred in advanced treatments processes such as nonthermal plasma8 and electrochemical reductions.9,10 With regard to their chemical reactivity, nonionic ICMs behave very similar to each other as indicated by comparable rate constants for reactions with ozone7,11 hydroxyl radicals, or e−aq.12 Rate constants for the ionic ICM diatrizoate were slightly lower in comparison to the nonionic ICM. Recently, highly genotoxic and cytotoxic iodinated disinfection byproducts (iodo-DBPs) were identified in chloraminated and chlorinated drinking water from Canada and the United States.13,14 Iodo-DBPs were also shown to be more cytotoxic and genotoxic as compared to their brominated and Received: Revised: Accepted: Published: 12689
July 24, 2014 September 26, 2014 September 28, 2014 October 17, 2014 dx.doi.org/10.1021/es503609s | Environ. Sci. Technol. 2014, 48, 12689−12697
Environmental Science & Technology
Article
chlorinated analogues.15 Iopamidol was shown to be a potential source of these iodo-DBPs, and the toxicity of extracts obtained by XAD-extraction of chlorinated drinking water was shown to be significantly increased when iopamidol was added prior to chlorination.16 The DBPs discussed in this paper can be categorized into two groups. Smaller ones (e.g., trihalomethanes or iodoacetic acid) are generally referred to by their systematic names; and higher molecular weight compounds that still contain the hexasubstituted benzene ring are named by a number referring their molecular weight, e.g., DBP 705 stands for the disinfection byproduct with the molecular mass of 705 Da. The aim of this study was to elucidate the transformation of ionic and nonionic ICM, namely iopamidol, iopromide, iohexol, iomeprol, and diatrizoate (structures shown in Supporting Information (SI) Figure S1) during disinfection of drinking water with chloramine, chlorine dioxide, and hypochlorite. Duirk et al. already showed that chlorination of iopamidol leads to the formation of small iodo-DBPs with elevated toxicity.16 The main focus of the current study was to identify the chemical structures of higher molecular weight DBPs and to elucidate their transformation pathways.
mass spectrometry (MS) (AB Sciex API 4000, Applied Biosystems, Langen, Germany) with ESI operating in positive multiple reaction monitoring (MRM) mode to determine the residual ICM concentrations. Separation was achieved using a Chromolith column (2 μm, 100 × 4.6 mm2, Merck, Darmstadt, Germany). The remaining concentration of ICM in each sample was measured three times and regression analysis was performed on the data; degradation was deemed significant if it on average exceeded a 95% significance level (p < 0.05). Further details are provided in the SI. TOX-DBP Experiments (Only for Iopamidol). Experiments were carried in deionized water at pH 6.5 and 8.5 using 200 mM phosphate buffer, 1.29 mM iopamidol and 20 times excess (25.7 mM) of active chlorine. Experiments were started by adding chlorine stock solution to a stirred buffered ICM solution. Stirring was maintained for 3 min using a polytetrafluoroethylene (PTFE) coated magnetic stir bar. Subsequently, samples were transferred into eight 10 mL amber vials with PTFE septa and stored at 25 °C in an incubator for 0, 1, 2, 6, 12, 24, 48, and 72 h. At each discrete interval, 3.89 mL of the respective vial was diluted to 1 L with purified water, rapidly mixed for 3 min and divided as follows: (i) two 10 mL aliquots (in 16 mL amber vials) were quenched with aqueous sulfite solution (120% of [Cl2]0) for iodide analysis and with resorcinol (120% of [Cl2]0) for iodate analysis, respectively; (ii) a 30 mL aliquot (in a 40 mL amber vial) was quenched with sulfite solution for TOX analysis; and (iii) a 100 mL aliquot was quenched with sulfite and analyzed for DBPs. Characterization of DBPs. For identification of DBPs, chlorination experiments were performed at pH 6.5 and pH 8.5 using 200 mM phosphate buffer, 1.29 mM iopamidol and 20 times excess (25.7 mM) of aqueous chlorine. Aliquots (1 mL) of the solution were sampled after different reaction times, quenched with sodium thiosulfate (51.4 mM), and analyzed by HRMS and MSn fragmentation experiments (LTQ Orbitrap Velos, Thermo Scientific, Bremen, Germany). Separation was achieved using a PolarRP column (150 × 3.00 mm2, 4 μm; Phenomenex, Aschaffenburg, Germany). Further details are provided in the SI. Peak areas for different DBPs in their respective extracted-ion chromatograms (XIC) were used to obtain information about the temporal progression of these DBPs. Additional NMR experiments were performed on a Bruker Avance 700 MHz spectrometer (Bruker Corporation, U.S.A.) in case MSn did not yield sufficient information for identification of DBPs. Isolation and Degradation of DBPs. Iopamidol (1.0 g/L, 1.29 mM) was dissolved in phosphate buffer (200 mM) at pH 8.5. Then, 25.8 mM of sodium hypochlorite solution was added, the pH was adjusted to 8.5 again using hydrochloric acid (5%), and the mixture was stirred for the appropriate time to yield a large share of the desired product (1 h for early products, 24 h for follow-up products). After the addition of sodium thiosulfate (51.5 mM), the mixture was freeze-dried, and the residue was extracted with methanol using a Soxhlet extractor. The resulting solution was concentrated to about 10% using a gentle stream of nitrogen, and the DBPs were isolated by LC using a PolarRP column (250 × 10 mm2, 4 μm, Phenomenex, see SI for more details). The isolated DBPs were used in degradation experiments analogous to the procedure for parent ICM described above as well as for NMR measurements. Toxicological Experiments. For toxicological experiments, a large amount of the high molecular weight DBPs
■
MATERIALS AND METHODS Standards and Reagents. Purified water (18 MΩ·cm, DOC < 0.1 mg/L) was obtained from a Milli-Q system (Merck Millipore) or a Barnstead RO & NANOPure system (Barnstead-Thermolyne Corporation, Dubuque, IA). Other solvents (acetonitrile, methanol) were purchased from Merck (Darmstadt, Germany) at HPLC grade purity. Iodinated contrast media (iopamidol, iopromide, iomeprol, iohexol, diatrizoate) were supplied by Bayer Schering Pharma (Berlin, Germany) or purchased from United States Pharmacopeial Convention (Rockville, MD) and had a purity of >95%. Sodium hypochlorite solution was purchased from SigmaAldrich. The exact concentration of active chlorine was photometrically determined using standardized cuvette tests (Hach Lange, LCK 310). ClO2 solutions were diluted from a 0.3% stock solution prepared by a two component powder (based on sodium bisulfate and sodium chlorite) provided by TwinOxide (DeTongelreep, Netherlands). For monochloramine experiments, preformed monochloramine was used to avoid potential artifacts caused by reactions of excess free chlorine. The stock solutions were prepared by mixing 200 μL 12.5% NaOCl solution (0.4 mmol) with 36 μL aqueous NH3 (25%, 0.48 mmol) in 20 mM phosphate buffer. The solution was aged for 30 min prior to use in the experiments. An overview of all DBP standards and total organic halogen (TOX) standards used is provided in the SI. Kinetic Experiments. The pH of the solutions, ranging from 6.5 to 8.5 was controlled with sodium phosphate buffer (10 mM) that was prepared and brought to the desired pH by mixing 10 mM solutions of Na2HPO4 and KH2PO4 in the appropriate amounts. Reactions were carried out under pseudo first-order conditions, i.e., chlorine concentration (approximately 4 mg/L, 67.2 μM) was about 500 times higher than the initial ICM concentrations (100 μg/L, 0.13 μM). Experiments were started for each ICM separately by adding chlorine stock solution to a stirred buffered ICM solution. Samples (1 mL) were taken at discrete intervals and added to vials containing an excess of Na2S2O3 to quench the oxidative reactions. The samples were analyzed via liquid chromatography (LC)-tandem 12690
dx.doi.org/10.1021/es503609s | Environ. Sci. Technol. 2014, 48, 12689−12697
Environmental Science & Technology
Article
To determine the reaction order relative to active chlorine, the respective rate constant k1 was plotted as a function of the chlorine concentration [Cl2]T at several pH values (Figure 1a).
was produced. Thus, hypochlorite (33.54 mmol) was added to 1.3 g (1.29 mM) of iopamidol in buffered solution (1.3 L, 200 mM phosphate buffer, pH 8.5). After 24 h, the reaction was quenched using an excess of sodium thiosulfate (16.65 g Na2S2O3·5H2O). The reaction mixture was freeze-dried, and the remaining substance was extracted with methanol in a Soxhlet extractor. The methanol was evaporated with a gentle stream of nitrogen and the toxicity of the residue was tested using Chinese hamster ovary (CHO) cell chronic cytotoxicity assays. For comparison, a control sample was prepared the same way without the addition of hypochlorite. CHO Cell Chronic Cytotoxicity Assay. Chinese hamster ovary (CHO) cell line AS52, clone 11−4−8 was used for the toxicity studies.17 The CHO cells were maintained in Ham’s F12 medium containing 5% fetal bovine serum, 1% antibiotics (100 U/mL sodium penicillin G, 100 μg/mL streptomycin sulfate, 0.25 μg/mL amphotericin B in 0.85% saline), and 1% glutamine at 37 °C in a humidified atmosphere of 5% CO2. This 96-well microplate assay measures the reduction in cell density as a function of the concentration over a period of 72 h (∼3 cell cycles).18 For each sample concentration, 4−8 replicates were analyzed, and the experiments were repeated twice. A concentration−response curve was generated for both the chlorinated sample and the unchlorinated control, and a regression analysis was conducted for each curve. The LC50 values were calculated, where the LC50 represents the sample concentration that induced a 50% reduction in cell density as compared to the concurrent negative control.
■
RESULTS AND DISCUSSION Degradation Experiments. The degradation of contrast media (100 μg/L) by monochloramine and chlorine dioxide was conducted with concentrations of the chlorinating agent at 3 mg/L in purified water. This concentration is close to the range of concentrations applied for drinking water disinfection.19,20 No significant degradation (p < 0.05) was observed for iopromide, iohexol, iomeprol, or diatrizoate, regardless of the pH adjusted. For iopamidol, a slight degradation of 7.9 ± 1.2% over the course of 24 h was observed with chlorine dioxide at pH 8.5, while no degradation was detected at pH 6.5. With 3 mg/L monochloramine, no significant degradation was found, regardless of the pH (SI Figures S2−S5). Iopromide, iohexol, iomeprol, and diatrizoate were also recalcitrant to oxidation by active chlorine, regardless of pH (SI Figures S6−S7); however, iopamidol (100 μg/L) was readily transformed in the presence of 5 mg/L active chlorine. At pH 6.5, 42% of the iopamidol were consumed within 24 h; at pH 8.5 iopamidol was even totally transformed within 24 h (SI Figure S8). Hence, the transformation of iopamidol with hypochlorite was studied in more detail. Degradation Kinetics and pH Dependence. Degradation experiments using a large excess of active chlorine resulted in an exponential decrease of iopamidol between pH 6.5−8.5 according to (pseudo) first order kinetics (SI Figure S9). Furthermore, the reaction rates increased with increasing pH, indicating that the hypochlorite anion might be the active chlorine species, although hypochlorous acid is reported to be the stronger oxidant species.21 The hypochlorite anion was previously proposed as the active chlorine species in 2011.3 That conclusion however relied mostly on the increased formation of iodo-THMs and iodo-acids at higher pH, while only very limited data on the degradation of iopamidol itself was available.
Figure 1. (a) Variation of the observed reaction constant k1 for the degradation of iopamidol as a function of [Cl2]T at different pH values. Error bars: standard error of the linear regression. (b) Experimental data and model calculations for the second-order rate constant k at different pH values. Error bars: standard error of the linear regression in part a.
A linear dependency for this plot was found with correlation coefficients R2 > 0.9 for every pH (p < 0.05), which shows that the reaction is also of first-order relative to the concentration of active chlorine. Figure 1a illustrates the proportional increase of the pseudo-first-order rate constant. Iopamidol, with a pKa of 10.7, remains predominantly uncharged in the experimental pH range.22 Plotting the pseudofirst order rate constant k1 against the pH produced an Sshaped curve with a maximum slope around pH 7.6 (SI Figure S10). Taking into account the pKa of hypochlorous acid (7.54),23 this underlines the assumption that OCl− rather than HOCl is the chlorine species reacting with iopamidol. The reaction was found to be first-order relative to the iopamidol concentration and first-order to the concentration of active chlorine. Under the assumption that other degradation pathways are negligible, a second-order reaction can be derived and the following rate law results: − 12691
d[iopamidol] = k[iopamidol][Cl 2]T dt
(1)
dx.doi.org/10.1021/es503609s | Environ. Sci. Technol. 2014, 48, 12689−12697
Environmental Science & Technology
Article
where [Cl2]T is the cumulative concentration of hypochlorous acid and the hypochlorite anion, and k, the second-order rate constant, is as follows:
k=
k1 [Cl 2]T
(2)
The experimentally determined second-order rate constants increased with increasing pH (Figure 1) in the pH range from 6.5 to 8.5. Kinetic Modeling. Under the assumption that the only chlorine species present are hypochlorous acid and the hypochlorite anion, the rate expression for the chlorination of iopamidol is as follows: −
d[iopamidol] = kHOCl[iopamidol][HOCl] dt + k OCl[iopamidol][OCl−]
(3)
To obtain values for the intrinsic rate coefficients of these two chlorine species, a simple kinetic model of the reaction was constructed. The pH dependency of the second-order rate constant k was calculated according to the following: k=
kHOCl[H+] + k OClK aHOCl [H+] + K aHOCl
(4)
The complete derivation is shown in the SI. The intrinsic rate coefficients kHOCl and kOCl were iteratively determined using Origin, by minimizing the quadratic deviation between the experimental second-order rate constants (the slopes of the linear regressions in Figure 1a) and the theoretical values (eq 4). Only for OCl− a reasonable value of the rate coefficient (kOCl = 0.94 ± 0.03 M−1 s−1) was obtained, while the uncertainty for the intrinsic rate constant for HOCl (kHOCl = (1.5 ± 2.5)·10−2 M−1 s−1) is higher than the value itself. It can be concluded that OCl− is the active oxidant reacting with iopamidol, while the reaction with hypochlorous acid is negligible. Figure 1b illustrates the good agreement of the calculated model and the experimental results. Fate of Iodine. In light of the high toxicity of iodo-DBPs, the question arises whether the iodine released from iopamidol is a potential source for their formation. In the degradation experiments with high iopamidol concentrations (1 g/L), an exponential decrease of total organic iodine (TOI) was observed (Figure 2). The TOI loss was faster at pH 8.5 than at pH 6.5, but overall slower than the iopamidol transformation. While iopamidol at pH 8.5 is consumed completely, one-third of the initial TOI remained under these reaction conditions. Regardless of pH, the sum of TOI and the iodate concentrations (i.e., [I]T) remained relatively constant throughout the experiment. The oxidation of iodide to iodate by HOCl and OCl− (via HOI) is well-known.24 It appears that a majority of the iodine released was oxidized to iodate. The concentrations of low-molecular weight iodo-DBPs were measured and calculated as a percentage of the TOI (taking into account their respective degree of iodination, SI Figures S35 and S36), as described by Echigo et al.25 It was found that only a small fraction of the TOI existed as known iodo-DBPs, while the majority appears to be associated with iopamidol or unknown DBPs. After 12 h, almost no low-molecular weight iodo-DBPs (less than 0.1% of the remaining TOI) were detected in the reaction mixture at both pH values. Since
Figure 2. Mass balance on total organic and inorganic iodide containing species at pH 6.5 (left) and pH 8.5 (right). [Cl2]T = 25.7 mM, [Iopamidol] = 1.29 mM, [Buffer]T = 200 mM, and temperature = 25 °C.
iopamidol itself is completely transformed at pH 8.5, more than 99.9% of the remaining TOI must consist of other DBPs, likely of high molecular weight. Although the concentrations of dichloroiodomethane and chlorodi-iodomethane increased slowly over time, these DBPs only accounted for a small portion of organically bound iodine throughout the whole process, reaching about 10% of TOI after 72 h at pH 8.5. Iodoacetic acids were not detected at all, although they were found by Duirk et al.16 in a previous study after chlorination of water containing NOM and iopamidol. Chloroform and trichloroacetic acid (TCAA) were detected as chlorinated DBPs with low molecular weights (SI Figure S34). Their relative concentrations were pH dependent as reported elsewhere.26 After 12 h at pH 6.5, a majority of the TOCl is associated with chlorinated THMs (i.e., chloroform, CHCl2I, and CHClI2) and HAAs (dichloroacetic acid (DCAA) and TCAA) (SI Figure S36). This is significantly different than after 12 h at pH 8.5, where those chloro-DBPs only account for a minor share (6.3% THMs, 14.7% HAAs), while most of the TOCl (79%) is unknown DBPs. At pH 6.5, the TOCl concentration after 12 h is approximately only one-third of the TOCl concentration detected at pH 8.5. Due to further formation of high molecular weight DBPs, after the longer reaction time of 24 h, the majority (78.8% at 12692
dx.doi.org/10.1021/es503609s | Environ. Sci. Technol. 2014, 48, 12689−12697
Environmental Science & Technology
Article
Figure 3. Proposed reaction pathway for the disinfection byproducts of iopamidol with aqueous chlorine.
pH 6.5, 82.3% at pH 8.5) of the TOCl formed at both pH is associated with unknown DBPs. At least some of these are structurally similar to iopamidol, as is shown below. The formation of small molecular weight chlorinated DBPs happens
more rapidly at pH 6.5, likely due to the higher oxidation potential of HOCl.21 Iodate formation and TOI degradation initially exhibited a lag phase of several hours at pH 6.5, but not at pH 8.5. The formation of TOCl also starts only after several 12693
dx.doi.org/10.1021/es503609s | Environ. Sci. Technol. 2014, 48, 12689−12697
Environmental Science & Technology
Article
Figure 4. Temporal progression of the main high molecular iodo-DBPs of iopamidol. c0 (Iopamidol) = 1 g/L, pH 8.5, with an excess (20:1 n/n) of chlorine.
hours at pH 6.5, but within 1 h at pH 8.5. It appears that some initial TPs formed from iopamidol directly still contain all iodine and only release it in the following reactions that occur more slowly at lower pH. Also, since at least one-third (34.4%) of the original TOI remained, it appears that iopamidol forms stable products that still contain some of the original iodine (Figure 3). The fact that the known small chlorinated DBPs do not suffice to explain the entire TOCl suggests that stable chlorinated DBPs of much higher molecular weight are also present in the reaction mixture. Identification of High Molecular Weight DBPs. High molecular weight chlorination products were identified using their exact masses, isotope patterns, and MSn fragmentations. Two key DBPs, namely DBP 705 and DBP 778, could be isolated in sufficient quantities to confirm their structures by NMR. The data for the identification of these two compounds and for DBP 778’s direct precursor, DBP 777, are discussed here in detail. Data for all other DBPs can be found in the SI (Figures S12−S24). DBP 705. The exact mass of DBP 705 (705.8402 as M+H+) suggests the chemical formula C14H19I3O3N6. The MS/MS fragmentation (SI Figure S11) yields two major products at m/ z 615 (−91) and m/z 688 (−18). The latter corresponds to the loss of one molecule of water, the former to the cleavage of the
CN-bond in the amide moiety of one of the molecule’s carboxamide side chains. This shows that at least one of these side chains remained unchanged. The MS3 spectrum of m/z 688 led to three products: m/z 615 (−71), m/z 597 (−91), and m/z 670 (−18). The fragment at m/z 615 corresponds to the cleavage of the previously dehydrated side chain, m/z 670 to a second dehydration. Fragment m/z 597 is formed by the cleavage of the second carboxamide side chain. This indicates that both of these side chains are still present in DBP 705. This and the fact that all the iodine is still present suggests that transformation occurred at the acylamino side chain. It is therefore proposed that product 705 is formed from iopamidol by cleavage of side chain A at the amide CN-bond. The NMR of DBP 705 (SI Figures S25−S27) confirmed the presence of an aromatic amine (seen as two singlets at 5.36 and 5.39 ppm due to the existence of rotamers). The fact that both carboxamide side chains are still present could also be confirmed due to the fact that all their signals (OH at 4.52/ 4.68 ppm, CH2OH at 3.4−3.6 ppm, CHNH at 3.76 ppm, NHCO at 7.52/8.03 ppm) correlate to the iopamidol data.22 DBP 777 and DBP 778 differ from iopamidol only very slightly. DBP 777 has the same exact mass as iopamidol and a very similar fragmentation pattern (SI Figures S22 and S23), but a different retention time. DBP 778’s mass is larger by 0.98 Da. This difference could be explained by the exchange of an 12694
dx.doi.org/10.1021/es503609s | Environ. Sci. Technol. 2014, 48, 12689−12697
Environmental Science & Technology
Article
dissolved in purified water led only to the formation of DBP 704 by a cleavage of side chain B. All other DBPs were probably formed via DBP 705. This was shown by chlorinating the isolated DBP 705 in a buffer at the conditions also used for the iopamidol chlorination. In these experiments, all DBPs except DBP 777, DBP 778, and DBP 704 were formed, underlining the existence of two separate transformation pathways. Further transformation of DBP 705 includes either a successive exchange of iodine by chlorine yielding DBP 613 and DBP 521 or an oxidation of the amino moiety to a nitro group (DBP 735). Combination of both reactions furnishes DBP 643 and DBP 551. Furthermore, cleavage of side chain B (or B′) was observed at many stages of the transformation pathway. DBP 705, DBP 613, and DBP 521 are precursors in the formation of dimers DBP 1406, DBP 1314, DBP 1222, DBP 1130, and DBP 1038. This appears to be a result of oxidative-coupling reactions between the primary amines in DBP 705, DBP 613, and DBP 521 in highly oxidizing environments similar to the high concentration experiments.30 To obtain further insights into the sequence of these reactions, chlorination experiments were performed with isolated DBPs. The chlorination of DBP 735 showed that only DBP 661 was slowly formed through cleavage of side chain B, while no halogen exchange occurred. It is assumed that this is analogously true for the formation of DBP 569 from DBP 643 and DBP 477 from DBP 551. Since DBP 643 and DBP 551 are not formed from DBP 735, the exchange of iodine for chlorine occurred only at the aniline DBPs (DBP 705, DBP 613, and DBP 521). The nitro-DBPs are subsequently formed from those by oxidation of the amino moiety. There was no indication that the nitro-DBPs or the dimer compounds underwent any further halogen exchange. Mammalian Cell Toxicology. The CHO cell chronic cytotoxicity data are presented in Figure 5. The chlorinated and unchlorinated iopamidol samples were prepared and concentrated as presented in the Materials and Methods section. The concentration range of the chlorinated sample was from 0 to 120 μg of the residual DBP mixture per well (total volume 200 μL medium). The concentration range for negative control
NH group for an oxygen atom. Since there are no terminal nitrogen-containing groups present in iopamidol, it is assumed that one side chain is first rearranged to yield DBP 777, which then reacts further to DBP 778. Both carboxamide side chains can still be seen in the MSn fragmentation of DBP 777; so it has to be side chain A that was rearranged into an ether, leaving a terminal amide function, which hydrolyzes to a carboxylic acid in DBP 778. Since fragmentation of side chain A is not observed in iopamidol nor in either of these DBPs, NMR was required to validate these structural suggestions. DBP 777 is an intermediate product and could not be isolated in sufficient quantities, but the NMR of DBP 778 (SI Figures S28−S30) confirmed that both side chains B and B′ are still present (since their signals were only shifted very slightly from iopamidol), and that side chain A had been inverted: the carboxylic acid could be seen as a broad signal around 13 ppm, while the adjacent CH3CH group showed a multiplet at 3.44−3.78 ppm (for CH3CH) and a multiplet at 1.42−1.58 ppm (CH3CH), with a strong coupling between these two signals in the correlation spectroscopy (COSY) spectrum (SI Figure S29). The DBPs found in this study are different from ICM DBPs and biological TPs discussed in the literature. Seitz et al.6 found only minor transformation of the side chains in Iomeprol with ozone. Lütke Eversloh et al.9 found dehalogenated products of iopromide formed during electrochemical treatment of reverse osmosis concentrates, but no substitution of the liberated iodine with chlorine. The biological TPs found by Kormos et al.27 and Schulz et al.5 showed major transformation of the side chains, but the tri-iodinated benzene ring remained stable throughout their biological and chemical processes. In some cases, the transformations also included cleavage of the C−N bond in the amide moieties, as observed in this study. Proposed Transformation Pathway. The temporal progression of the formation of major DBPs identified is shown in Figure 4. The concentrations of DBP 705 and DBP 777 increased quickly from the beginning, while the formation of all other DBPs occurred later. This indicates that DBP 705 and DBP 777 are the precursor compounds for all or most of the other DBPs. DBP 705 was formed by cleavage of side chain A. This cleaved side chain A might be a source to form chlorinated DBPs. This would explain the TOCl formation at the beginning of experiments as well as the stable TOI since iodine is not released in the formation of DBP 705. The initial reaction site of chlorine at the iopamidol molecule to form DBP 705 is not clear because chlorine-containing intermediates were not observed. Therefore, the precise mechanism is not yet known. Mechanisms involving either Nchlorination or O-chlorination of the amide group are possible. A recent study suggests that the most favorable mechanism in the chlorination of amide-containing pharmaceuticals like acetaminophen and carbamazepine involves the formation of an iminol intermediate followed by reaction with HOCl to form N-chlorination products.28 For example, for acetaminophen, which has an amide linkage similar to iopamidol, one study reported a primary chlorination reaction product containing one atom of chlorine.29 DBP 777 can only be formed from iopamidol by an inversion of side chain A. The mechanism of this step is unknown, although the existence of a transitional state containing a fivemembered ring can be assumed. Analogous reactions could not be found in the literature. It is likely that DBP 778 was formed by hydrolysis of the amide moiety of DBP 777. Chlorination of isolated DBP 778
Figure 5. Concentration−response curves of the chlorinated iopamidol sample and the unchlorinated negative control. 12695
dx.doi.org/10.1021/es503609s | Environ. Sci. Technol. 2014, 48, 12689−12697
Environmental Science & Technology
Article
sample (without chlorine) was from 0 to 350 μg/well. The iopamidol DBP mixture was significantly more toxic than the iopamidol negative control. Using an analysis of variance test for significance, the lowest concentration of the chlorinated sample that induced a cytotoxic response was 30 μg/well. The LC50 value of 66.4 μg/well (or 332 ng/μL) was calculated from the regression analyses of the concentration response curve. An LC50 value for the negative control was not conducted because it was nontoxic throughout the concentration range in which there was no solvent interference. These data clearly indicate that chlorination of iopamidol generated a toxic mixture of DBPs. Since iodoacetic acids are not formed and most THMs and other volatile DBPs are lost by freeze-drying, the toxicity is very likely caused by other DBPs. Practical Relevance. Some of the DBPs (i.e., the dimer compounds) formed at these high iopamidol concentrations are unlikely to be found under conditions occurring in waterworks. To ensure that the results reported above are transferable to waterworks conditions, the experiments were repeated with (i) natural water containing NOM and (ii) a much lower iopamidol concentration. Using water from the river Rhine (for parameters, see SI) and 1 g/L iopamidol as well as a 20fold excess of active chlorine, the same high molecular weight DBPs were found as in the experiment with purified water, e.g., DBP 778, DBP 705, DBP 735, DBP 643, and DBP 551, as well as dimers (SI Figure S32). The experiments were also repeated with an iopamidol concentration of 100 μg/L and a 20-fold excess of active chlorine, both in purified water and in Rhine water. In this case, DBPs had to be measured using the LC-tandem MS system in the MRM mode, which requires the instrument to be tuned for each isolated DBP. However, it has to be noted that only DBP 705 and DBP 735 could be monitored, as at this time the other DBPs were not available at a sufficient quality and purity. Both DBPs were found in the experiment with purified water. In Rhine water, iopamidol was hardly degraded (98% remaining even after 24 h), but DBP 705 could be detected. This shows that the transformation of iopamidol under these conditions must at least in part follow the transformation pathway described in this paper. Measuring the other high molecular weight DBPs to close the mass balance requires isolated standards and the development of a sensitive and reliable analytical method. To assess the environmental relevance of iopamidol transformation, future research will also have to take into account the potential toxicity of the high molecular weight DBPs.
■
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This study was financially supported by the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG, project number TE 533/4-1) and the National Science Foundation (NSF, project numbers NSF1124865 and NSF1124844). The authors would like to thank Manfred Wagner and Stefan Spang (MPI for Polymer Research, Mainz, Germany) for NMR measurements as well as Elizabeth Crafton and Nana Ackerson (University of Akron, U.S.A.) for laboratory assistance.
■
(1) Christiansen, C. X-ray contrast mediaAn overview. Toxicology 2005, 209 (2), 185−187. (2) Ternes, T. A.; Hirsch, R. Occurrence and behavior of X-ray contrast media in sewage facilities and the aquatic environment. Environ. Sci. Technol. 2000, 34 (13), 2741−2748. (3) Kormos, J. L.; Schulz, M.; Ternes, T. A. Occurrence of iodinated X-ray contrast media and their biotransformation products in the urban water cycle. Environ. Sci. Technol. 2011, 45 (20), 8723−8732. (4) Seitz, W.; Jiang, J. Q.; Weber, W. H.; Lloyd, B. J.; Maier, M.; Maier, D. Removal of iodinated X-ray contrast media during drinking water treatment. Environ. Chem. 2006, 3 (1), 35−39. (5) Schulz, M.; Löffler, D.; Wagner, M.; Ternes, T. A. Transformation of the X-ray contrast medium lopromide in soil and biological wastewater treatment. Environ. Sci. Technol. 2008, 42 (19), 7207−7217. (6) Seitz, W.; Jiang, J. Q.; Schulz, W.; Weber, W. H.; Maier, D.; Maier, M. Formation of oxidation by-products of the iodinated X-ray contrast medium iomeprol during ozonation. Chemosphere 2008, 70 (7), 1238−1246. (7) Ning, B.; Graham, N. J. D. Ozone Degradation of Iodinated Pharmaceutical Compounds. J. Environ. Eng.−ASCE 2008, 134 (12), 944−953. (8) Gur-Reznik, S.; Azerrad, S. P.; Levinson, Y.; Heller-Grossman, L.; Dosoretz, C. G. Iodinated contrast media oxidation by nonthermal plasma: The role of iodine as a tracer. Water Res. 2011, 45 (16), 5047−5057. (9) Eversloh, C. L.; Henning, N.; Schulz, M.; Ternes, T. A. Electrochemical treatment of iopromide under conditions of reverse osmosis concentratesElucidation of the degradation pathway. Water Res. 2014, 48, 237−246. (10) Zwiener, C.; Glauner, T.; Sturm, J.; Worner, M.; Frimmel, F. H. Electrochemical reduction of the iodinated contrast medium iomeprol: Iodine mass balance and identification of transformation products. Anal. Bioanal. Chem. 2009, 395 (6), 1885−1892. (11) Ternes, T. A.; Stüber, J.; Herrmann, N.; McDowell, D.; Ried, A.; Kampmann, M.; Teiser, B. Ozonation: A tool for removal of pharmaceuticals, contrast media and musk fragrances from wastewater? Water Res. 2003, 37 (8), 1976−1982. (12) Jeong, J.; Jung, J.; Cooper, W. J.; Song, W. H. Degradation mechanisms and kinetic studies for the treatment of X-ray contrast media compounds by advanced oxidation/reduction processes. Water Res. 2010, 44 (15), 4391−4398. (13) Plewa, M. J.; Wagner, E. D.; Richardson, S. D.; Thruston, A. D.; Woo, Y. T.; McKague, A. B. Chemical and biological characterization of newly discovered lodoacid drinking water disinfection byproducts. Environ. Sci. Technol. 2004, 38 (18), 4713−4722. (14) Richardson, S. D.; Fasano, F.; Ellington, J. J.; Crumley, F. G.; Buettner, K. M.; Evans, J. J.; Blount, B. C.; Silva, L. K.; Waite, T. J.; Luther, G. W.; McKague, A. B.; Miltner, R. J.; Wagner, E. D.; Plewa, M. J. Occurrence and mammalian cell toxicity of iodinated disinfection byproducts in drinking water. Environ. Sci. Technol. 2008, 42 (22), 8330−8338.
ASSOCIATED CONTENT
S Supporting Information *
Structure of the analytes; DBP analysis and quantification; TOX extraction and analysis; iodate analysis; derivation of equation (4); chromatographic methods; reaction with chloramine; reaction with chlorine dioxide; reaction with hypochlorite; pH dependence of iopamidol degradation; fragmentation patterns; NMR spectra: DBP 705, DBP778; DBPs detected at lower concentrations; chlorination of iopamidol in river water; TOI/TOCl formation trends; speciation of DBPs; and cited literature. This material is available free of charge via the Internet at http://pubs.acs.org.
■
REFERENCES
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. 12696
dx.doi.org/10.1021/es503609s | Environ. Sci. Technol. 2014, 48, 12689−12697
Environmental Science & Technology
Article
(15) Plewa, M. J.; Wagner, E. D.; Muellner, M. G.; Hsu, K. M.; Richardson, S. D. Comparative Mammalian Cell Toxicity of N-DBPs and C-DBPs. In Disinfection by-Products in Drinking Water: Occurrence, Formation, Health Effects, and Control; Karanfil, T., Krasner, S. W., Xie, Y., Eds.; American Chemical Society: Washington, DC, 2008; Vol. 995, pp 36−50. (16) Duirk, S. E.; Lindell, C.; Cornelison, C. C.; Kormos, J.; Ternes, T. A.; Attene-Ramos, M.; Osiol, J.; Wagner, E. D.; Plewa, M. J.; Richardson, S. D. Formation of toxic iodinated disinfection byproducts from compounds used in medical imaging. Environ. Sci. Technol. 2011, 45 (16), 6845−6854. (17) Wagner, E. D.; Rayburn, A. L.; Anderson, D.; Plewa, M. J. Analysis of mutagens with single cell gel electrophoresis, flow cytometry, and forward mutation assays in an isolated clone of Chinese hamster ovary cells. Environ. Mol. Mutagen. 1998, 32 (4), 360−368. (18) Plewa, M. J.; Kargalioglu, Y.; Vankerk, D.; Minear, R. A.; Wagner, E. D. Mammalian cell cytotoxicity and genotoxicity analysis of drinking water disinfection by-products. Environ. Mol. Mutagen. 2002, 40 (2), 134−142. (19) EPA. Effectiveness of Disinfectant Residuals in the Distribution System, 2004. (20) World Health Organization. Monochloramine in Drinking-water, 2004. (21) Deborde, M.; von Gunten, U. Reactions of chlorine with inorganic and organic compounds during water treatmentKinetics and mechanisms: A critical review. Water Res. 2008, 42 (1−2), 13−51. (22) Florey, K. H. Analytical Profiles of Drug Substances; Academic Press Inc.: 1988; Vol. 17. (23) Albert, A., Serjeant, E. P. The Determination of Ionization Constants. A Laboratory Manual, 3rd ed.; Chapman and Hall: New York, 1984. (24) Bichsel, Y.; von Gunten, U. Oxidation of iodide and hypoiodous acid in the disinfection of natural waters. Environ. Sci. Technol. 1999, 33 (22), 4040−4045. (25) Echigo, S.; Zhang, X.; Minear Roger, A.; Plewa Michael, J., Differentiation of Total Organic Brominated and Chlorinated Compounds in Total Organic Halide Measurement: A New Approach with an Ion-Chromatographic Technique. In Natural Organic Matter and Disinfection By-Products; American Chemical Society: Washington, DC, 2000; Vol. 761, pp 330−342. (26) Liang, L.; Singer, P. C. Factors influencing the formation and relative distribution of haloacetic acids and trihalomethanes in drinking water. Environ. Sci. Technol. 2003, 37 (13), 2920−2928. (27) Kormos, J. L.; Schulz, M.; Kohler, H. P. E.; Ternes, T. A. Biotransformation of selected iodinated X-ray contrast media and characterization of microbial transformation pathways. Environ. Sci. Technol. 2010, 44 (13), 4998−5007. (28) Sakic, D.; Sonjic, P.; Tandaric, T.; Vrcek, V. Chlorination of Nmethylacetamide and amide-containing pharmaceuticals. Quantumchemical study of the reaction mechanism. J. Phys. Chem. A 2014, 118 (12), 2367−2376. (29) Glassmeyer, S. T.; Shoemaker, J. A. Effects of chlorination on the persistence of pharmaceuticals in the environment. Bull. Environ. Contam. Toxicol. 2005, 74 (1), 24−31. (30) Laha, S.; Luthy, R. G. Oxidation of aniline and other primary aromatic amines by manganese-dioxide. Environ. Sci. Technol. 1990, 24 (3), 363−373.
12697
dx.doi.org/10.1021/es503609s | Environ. Sci. Technol. 2014, 48, 12689−12697