Anaerobic Transformation of the Iodinated X-ray Contrast Medium

3 days ago - Federal Institute of Hydrology , Am Mainzer Tor 1, D-56068 Koblenz , Germany ... (DDPI), and its primary aniline desmethoxyacetyl iopromi...
0 downloads 0 Views 2MB Size
Article Cite This: Environ. Sci. Technol. XXXX, XXX, XXX−XXX

pubs.acs.org/est

Anaerobic Transformation of the Iodinated X‑ray Contrast Medium Iopromide, Its Aerobic Transformation Products, and Transfer to Further Iodinated X‑ray Contrast Media Maria Redeker, Arne Wick, Björn Meermann, and Thomas A. Ternes* Federal Institute of Hydrology, Am Mainzer Tor 1, D-56068 Koblenz, Germany

Downloaded via UNIV OF CALIFORNIA SANTA BARBARA on July 13, 2018 at 08:33:18 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: The iodinated X-ray contrast medium (ICM) iopromide and its aerobic transformation products (TPs) are frequently detected in the effluents of wastewater treatment plants and in different compartments of the aquatic environment. In this study, the anaerobic transformation of iopromide and its aerobic TPs was investigated in water−sediment systems. Iopromide, its final aerobic TP didespropanediol iopromide (DDPI), and its primary aniline desmethoxyacetyl iopromide (DAMI) were used as model substances. Five biologically formed anaerobic TPs of iopromide and DAMI and six of DDPI, and the respective transformation pathways, were identified. The TPs were formed by successive deiodination and hydrolysis of amide moieties. Quantification of the iodinated TPs was achieved by further development of a complementary liquid chromatography (LC)−quadrupole time-of-flight mass spectrometry (Q-ToF-MS) and LC−inductively coupled plasma − mass spectrometry (ICP-MS) strategy without needing authentic standards, despite several TPs coeluting with others. A database with predicted anaerobic TPs of ICMs was derived by applying the transformation rules found for the anaerobic transformation pathways of iopromide and diatrizoate to further ICMs (iomeprol and iopamidol) and their aerobic TPs already reported in the literature. The environmental relevance of the identified transformation pathways was confirmed by identifying an experimental TP and two predicted TPs using suspect screening of water taken from anaerobic bank filtration zones.



suspensions25 and solutions of peroxydisulfate,26 mineralization was not observed. Iopromide dissipated also in aerobic biological systems containing activated sludge,9,31−33 water and sediment, or water and soil8,32,34 up to >90%, but again mineralization was not observed. Iopromide was deiodinated in reductive processes such as electrolysis, 35,36 reductive catalysis with Pd on Al 2 O 3 catalysts37 or with reduced graphene oxides,38 and treatment with Fe0 at pH 3.39 Also, by use of special techniques such as e−aq produced by electron beam irradiation of aqueous solutions with sulfite or bicarbonate as an •OH scavenger40 or by γ-irradiation of N2-saturated solutions41 complete deiodination was achieved. In an anaerobic sludge digestion pilot plant,42 in lab-scale upflow anaerobic sludge blanket reactors,43,44 and in an anaerobic-anoxic/aerobic process,45 iopromide was removed by up to 40%, and the removal rate was increased up to 80% when metal−humic acid complexes, microbially reduced Pd(II), or an isolated strain of iopromide-

INTRODUCTION Iodinated X-ray contrast media (ICMs) are triiodinated aromatic substances used for the imaging of soft tissues and blood vessels during X-ray examinations. They are administered in high doses of up to 200 g,1 excreted nearly completely, and mainly nonmetabolized.2 About 1000 t of iopromide was sold worldwide to hospitals and pharmacies in 2016,3 and concentrations of several micrograms/liter have been detected in municipal wastewater.4−6 Removal of iopromide in wastewater treatment plants (WWTPs) was reported to be between negligible and >80%.4−8 In contact with activated sludge, iopromide is transformed into as many as 12 transformation products (TPs).8,9 Consequently, iopromide and its TPs have been detected in the effluents of WWTPs, in surface water, in bank filtrates, and even in groundwater and finished drinking water.4−6,8,10−19 Several techniques and processes have been investigated to reduce the concentrations of iopromide in the different compartments of the water cycle. Iopromide was transformed by up to >90% by oxidative processes such as ozonation20−23 and by UV irradiation in TiO2 suspensions,24,25 in solutions of peroxydisulfate and/or H2O 2,26−29 and in nonthermal plasma.30 However, except for UV irradiation in TiO2 © XXXX American Chemical Society

Received: March 1, 2018 Revised: June 16, 2018 Accepted: June 23, 2018

A

DOI: 10.1021/acs.est.8b01140 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

identified anaerobic TPs of iopromide, DDPI, DAMI, and diatrizoate as well as (ii) predicted anaerobic TPs of other aerobic iopromide TPs and (iii) predicted anaerobic TPs of other ICMs and their aerobic TPs.

degrading bacteria was added. The transformation reactions included transformations of the side chains as well as deiodinations.43,44 Accordingly, during bank filtration, iopromide concentrations decreased under both anaerobic and aerobic conditions. In addition, under anaerobic conditions a substantial reduction of the total adsorbable iodine (AOI) was observed. This indicates that, under anaerobic conditions, deiodination reactions may be occurring.15,16,18 Before reaching anaerobic zones, the water passes aerobic zones in which iopromide is already transformed into aerobic TPs. Therefore, it can be assumed that not only iopromide itself but also its aerobic TPs are further transformed if they reach the anaerobic zones. For the ICM diatrizoate, a transformation by reductive deiodinations and hydrolysis of amide groups has been observed in anaerobic environmental compartments.46 However, no TPs of iopromide that are formed in the anaerobic zones have been identified yet, and it is not clear to what extent the transformation reactions observed for diatrizoate can be transferred to iopromide as well as to other ICMs and their aerobic TPs. Moreover, since iopromide can be transformed already under aerobic conditions, studying its fate in anaerobic zones is more complex than that of aerobically stable compounds, and investigating the anaerobic transformation of iopromide itself would elucidate only part of its environmental fate in aerobic−anaerobic systems. Thus, the aim of this study was to investigate the microbiologically mediated anaerobic transformation of iopromide and selected aerobic TPs. Didespropanediol iopromide (DDPI), which is stable under aerobic conditions, was selected because it is the final TP of the aerobic transformation pathway of iopromide and has been shown to be ubiquitously present in wastewater-influenced bank filtrate and groundwater.8 Desmethoxyacetyl iopromide (DAMI), the primary aniline and an aerobic TP of iopromide, was identified in laboratory wastewater treatment plants.33 Both were chosen because their triiodinated aromatic core structure is similar to that of iopromide and they differ only by the length of the three aliphatic side chains. This enables examination of whether the distinct structural differences of the side chains influence the anaerobic transformation of ICMs. Thereby general anaerobic transformation rules, not only for iopromide and DDPI but also for other ICMs and their aerobic TPs, may be proposed. Identification and quantification of TPs were accomplished by complementary mass spectrometric techniques consisting of liquid chromatography−quadrupole time-of-flight mass spectrometry (LC-Q-ToF-MS) and liquid chromatography− inductively coupled plasma − mass spectrometry (LC-ICPMS). A previously developed methodology applied to quantify iodinated TPs without the need for authentic standards was further refined to overcome also the challenge of quantifying coeluting analytes individually. Finally, the transferability of the identified transformation processes to the real environment and to further ICMs was investigated. Also here, a possible preceding aerobic transformation had to be considered. For this, a suspect screening approach was developed. It was based on a database that contains the experimental TPs identified in this study and predicted TPs derived by combining the experimentally derived transformation rules from this study with that from previous studies. Native samples from anaerobic bank filtration sites were screened for the occurrence of (i) experimentally



MATERIALS AND METHODS Chemicals, Standards, and Solvents. Information on chemicals, standards, and solvents used for this study is provided in Supporting Information. Anaerobic Batch Experiments. Anaerobic sediment− water batch experiments were set up in 120 mL glass serum bottles in a glovebox under argon atmosphere. Oxygen-free Rhine water (25 mL) was added to 10 g of anaerobic sediment taken from a sulfate-reducing zone of a polishing pond. Iopromide, DDPI, and DAMI were spiked into separate bottles, in triplicate each, at a concentration of 2.6 μmol/L, which corresponds to an organic-bound iodine concentration of 1 mg/L. In addition, three nonspiked blank replicates were prepared to facilitate TP identification since all compounds that were formed also in the blank samples could be omitted from further consideration as potential TPs. To elucidate whether the transformation depends on microbial activity, three sterile batch systems were prepared with all three target substances. These batch systems were purged with argon and autoclaved (VE-75, Systec, Linden, Germany) twice before spiking. The preparation, incubation, and sampling of batch experiments using anaerobic techniques is described in more detail in Supporting Information. Bank Filtrate Samples. Grab samples were taken from five wells in anaerobic zones (iron- and/or sulfate-reducing conditions) of different bank filtration sites in Germany. The wells are located at 80−300 m distance from the respective river. Information on the redox conditions [Fe(II) concentrations and presence/absence of H2S smell] is provided in Table 1. Analytical Methods. Identification of Transformation Products. Identification of the TPs was performed with an HPLC 1290 Infinity system coupled with a 6550 iFunnel QToF mass spectrometer from Agilent Technologies (Waldbronn, Germany). The dual AJS (Agilent Jet Stream) electrospray ionization (ESI) source was operated in positive ionization mode. A Hydro-RP column (250 mm × 3 mm, 4 μm; Phenomenex, Aschaffenburg, Germany) was used for chromatographic separation. Mobile phases were ultrapure water and methanol, both amended with 0.1% formic acid. The chromatographic conditions were the same as those used for quantification by LC-ICP-MS (see next section). Further information on MS source parameters and data-dependent acquisition parameters is provided in Supporting Information. Samples from the batch experiments were screened for TPs formed in the spiked incubation experiments, but not in the nonspiked blank experiments, over the course of the incubation time. Molecular formulas of the TPs were determined by full-scan measurements. Chemical structures of the TPs were assigned by tandem mass spectrometric (MS2) experiments determining molecular formulas of the fragment ions. In addition, the samples were screened for hypothetical TPs formed by all combinations of reactions found in previous studies investigating reductive transformation processes of iopromide.35,36 For this, extracted ion chromatograms (EICs) with the exact masses (±10 ppm) of the hypothetical TPs were extracted from the total ion chromatograms (TICs). B

DOI: 10.1021/acs.est.8b01140 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology Quantitative Analysis of Samples from Batch Experiments. Quantification of iopromide, DDPI, and DAMI was carried out by LC-Q-ToF-MS using exact mass (±10 ppm) EICs from full-scan measurements. Iopromide-d3 (spiked at 250 μg/L) was used as internal standard. An external ninepoint calibration containing the three target compounds (1− 500 μg/L) and iopromide-d3 (250 μg/L) was prepared in ultrapure water. No authentic reference standards were available for the anaerobic TPs. However, to elucidate whether the quantitatively relevant TPs had been identified, the mass balances needed to be investigated. Iodine-containing compounds are amenable to quantification via ICP-MS. Determination of the iodinated TP concentrations was therefore based on modification (Figure 1, solid arrows) of a methodology

was further refined to overcome the challenge of quantifying coeluting TPs individually. Instead of using LC-ICP-MS chromatograms directly for quantification, the exact mass EICs obtained from LC-Q-ToFMS measurements were screened to find, for each iodinated TP or TP isomer, at least one sample in which no coelution with (isomers of) other iodinated TPs occurred. In those samples, the molar concentrations of the iodinated TP isomers were determined by LC-ICP-MS (see Figure 1). From the iodinated TP concentrations obtained by quantification with LC-ICP-MS and the corresponding peak areas from the LC-QToF-MS EICs (normalized by peak area of the internal standard iopromide-d3, which had been spiked at 250 μg/L), response factors (RF) were calculated for each TP isomer by dividing its respective concentration (cLC‑ICP‑MS) by the corresponding normalized LC-Q-ToF-MS peak area (ALC‑Q‑ToF‑MS) (eq 1): c RF = LC‐ICP‐MS ALC‐Q‐ToF‐MS (1) The normalized peak areas of the LC-Q-ToF-MS EICs of all samples were multiplied by these isomer-specific factors to obtain the calculated TP concentrations (ccalc) in all samples (eq 2): ccalc = (RF)(ALC‐Q‐ToF‐MS)

(2)

If a response factor could be calculated from more than one sample for a given TP or TP isomer, the arithmetic mean of the response factors calculated in those samples was used. However, for a number of TP isomers, no sample existed in which the coeluting isomers of other TPs were not present. In these cases, the response factor of the targeted isomer was calculated by subtracting the concentration of the coeluting isomers (ccoe), for which the response factors could be determined from samples taken at earlier incubation times, from the total concentration of targeted and coeluting isomers (cLC‑ICP‑MS,tot) (eq 3): c LC‐ICP‐MS,tot − ccoe RF = ALC‐Q‐ToF‐MS (3) Figure 1. Quantification strategy for iodine-containing TPs of iopromide, DDPI, and DAMI.

Examples for calculations of response factors of one TP isomer without coelution and one TP isomer with coelution are provided in Tables S1 and S2. The limits of quantification (LOQs) for single TP isomers were the concentrations at which the signal-to-noise ratios in the EICs were ≥10. They were determined by extrapolation or interpolation of the calculated concentrations in the samples from batch experiments. TPs that are completely deiodinated could only be assessed semiquantitatively by their peak areas in the LC-Q-ToF-MS measurements. Iodide was quantified via ion chromatography (IC)-MS. Information on the respective ion chromatographic and MS parameters is provided in Supporting Information. Suspect Screening of Transformation Products in Bank Filtrate Samples. To investigate whether the identified anaerobic transformation pathways of iopromide and DDPI in the batch experiments can also occur under environmental conditions and whether they can also be transferred to other ICMs and their aerobic TPs, a suspect screening method was applied. For this, a database (Table S3) with elemental compositions of the identified anaerobic TPs of iopromide, DDPI, and

reported previously (Figure 1, dashed arrows) by the authors.46 In that study, based on measurements with LCESI-MS/MS for peak assignment, complementary LC-ICP-MS measurements were applied for quantification of anaerobic TPs of diatrizoate. Molar concentrations of the iodinated TPs were calculated from concentrations of the isotope 127I determined by LC-ICP-MS. This was possible due to the elementalselective as well as species-unspecific response of the ICPMS,47−49 which could be verified for model substances and further ICMs (see Supporting Information). Thus, speciesunspecific standards could be used for quantification of the iodinated TPs. Details on LC-ICP-MS setup and parameters are provided in Supporting Information. In the present study, the EICs of the target compounds and their iodine containing TPs showed up to 11 peaks per sample, some of which coeluted with others (Figure S7). By LC-ICPMS alone, it is possible to determine the total concentrations of coeluting analytes but not to distinguish what proportion of the concentration belongs to which TP. Therefore, the method C

DOI: 10.1021/acs.est.8b01140 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

exponential; the times in which the concentration dissipated to 50% of the original concentration c0 (DT50) were 5.3 ± 1.5 days (iopromide), 6.5 ± 0.5 days (DDPI), and 9.7 ± 1.5 days (DAMI) (see Table S4). Throughout the whole incubation period, the redox potentials and oxygen concentrations were