Emissions, Transport, and Fate of Emerging Per ... - ACS Publications

Organic Contaminants Control, School of Environment, Tsinghua University, Beijing 100084, China. ∥ ... area-based sediment-water distribution coeffi...
2 downloads 0 Views 3MB Size
Article pubs.acs.org/est

Cite This: Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Emissions, Transport, and Fate of Emerging Per- and Polyfluoroalkyl Substances from One of the Major Fluoropolymer Manufacturing Facilities in China Xiaowei Song,†,∥ Robin Vestergren,‡ Yali Shi,*,† Jun Huang,§ and Yaqi Cai†,∥,⊥

Downloaded via KAOHSIUNG MEDICAL UNIV on August 18, 2018 at 15:37:04 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.



State Key Laboratory of Environmental Chemistry and Ecotoxicology, Research Center for Eco-Environmental Science, Chinese Academy of Sciences, Beijing 100085, China ‡ IVL Swedish Environmental Research Institute. § State Key Joint Laboratory of Environment Simulation and Pollution Control (SKLESPC), Beijing Key Laboratory for Emerging Organic Contaminants Control, School of Environment, Tsinghua University, Beijing 100084, China ∥ University of Chinese Academy of Sciences, Beijing 100049, China ⊥ Institute of Environment and Health, Jianghan University, Wuhan 430056, China S Supporting Information *

ABSTRACT: Fluoropolymer manufacturing is a major historical source of perfluorooctanoic acid (PFOA) on a global scale, but little is known about the emissions, transport, and fate of emerging per- and polyfluoroalkyl substances (PFASs). Here, we performed a comprehensive spatial trend and interyear comparison of surface water and sediment samples from the Xiaoqing River, which receives water discharge from one of the major fluoropolymer manufacturing facilities in China. A suspect screening identified 42 chemical formulas, including the tetramer acid of hexafluoropropylene oxide (HFPO-TeA) and numerous tentatively detected isomers of C9−C14 per- or polyfluoroalkyl ether carboxylic acids (PFECAs). As revealed by the spatial trends and peak area-based sediment-water distribution coefficients, emerging PFASs with 3−9 perfluorinated carbons were transported unimpededly with the bulk water flow having no measurable degradation. Emerging PFASs with >9 perfluorinated carbons displayed more rapidly decreasing spatial trends than shorterchain homologues in surface water due to increasing sedimentation rates. The presence of HFPO oligomers, monoether PFECAs, monohydrogen-substituted perfluoroalkyl carboxylic acids (PFCAs) and monochlorine-substituted PFCAs could partly be explained by the active use of polymerization aids or the impurities therein. However, further research is encouraged to better characterize the emissions of low-molecular-weight PFASs from fluoropolymers throughout their life-cycle.



INTRODUCTION

inventories coupled to transport and fate models have demonstrated that FP manufacturing is the dominant historical source of perfluorooctanoic acid (PFOA) on a global scale.4−6 On a local scale, the emissions from FP manufacturing have resulted in elevated human exposure to PFOA primarily via contaminated drinking water.7,8 Elevated exposure to PFOA has subsequently been linked to several adverse health outcomes including ulcerative colitis,9 thyroid disease,10 kidney cancer and testicular cancer.11,12 The ubiquitous presence of PFOA in human and environmental samples combined with its intrinsic persistence,

Per- and polyfluoroalkyl substances (PFASs) are a large group of synthetic compounds that have been produced for numerous uses in consumer products and industrial processes since the 1950s.1 One of the most important commercial subgroups of PFASs is the fluoropolymers (FPs), which includes polytetrafluoroethylene (PTFE), polyvinylidene fluoride (PVDF), and fluorinated ethylene propylene resin (FEP).1 Due to their thermal and chemical stability and dielectric and low friction properties, FPs are used as inert tubes and linings in the chemical and pharmaceutical industries, as lightweight durable plastics in airplanes and for numerous other applications.2 Although FPs perform many critical functions in modern society, their manufacture results in significant emissions of low-molecular-weight PFASs which are used as emulsifying agents in the polymerization process.3,4 Emission © XXXX American Chemical Society

Received: December 26, 2017 Revised: July 24, 2018 Accepted: August 3, 2018

A

DOI: 10.1021/acs.est.7b06657 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

Figure 1. Spatial trends and interyear comparison of legacy and emerging PFASs in water and sediment samples using peak areas (C18 column).

bioaccumulation potential, and toxicity has prompted several changes in FP manufacturing and the use of PFASs as processing aids.13 In 2001, a major producer of the ammonium salt of PFOA (APFO) announced the phase-out of all perfluorooctyl substances produced by the electrochemical fluorination process (ECF).14 After this phase-out, several fluorochemical manufacturers using the telomerization process increased their production of PFOA to meet the market demand for high-performance polymerization aids.1,15 However, the emissions of PFOA from FP manufacturing have decreased drastically in Europe, North America, and Japan

after the initiation of a stewardship program between the U.S. EPA (Environmental Protection Agency) and eight major fluorochemical producers.15,16 This decrease in emissions was initially achieved by improved process technology that included the recycling of PFOA from waste streams,15 and ultimately by substituting APFO with various alternative polymerization aids.17 The major alternatives to APFO described in the literature are primarily the salts of per- or polyfluoroalkyl ether carboxylic acids (PFECAs) with 4−9 peror polyfluorinated carbons and 1−3 ether bonds, including Gen-X (CAS no. 62037−80−3), ADONA (CAS no. 958445− B

DOI: 10.1021/acs.est.7b06657 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

dispersant that is synthesized through an oligomerization process (see Supporting Information (SI) Figure S1) with the ammonium salt of HFPO-TrA as the main active ingredient (CAS no. 13252−14−7).30,31 The majority of APFO and T-5 produced at the facility is subsequently used as inert processing aids in PTFE and PVDF production, respectively. The subsequent emissions are believed to enter the environment mostly by wastewater and to a lesser extent by air.4,15,30,31 In addition to these major production lines of FPs, the Dongyue industrial park has a large complex of research and development facilities where alternative fluorinated surfactants are synthesized, performance tested, and possibly emitted to the environment.32 Previous studies performed by our research group in 2014 analyzed surface water and sediment samples along the entire Xiaoqing River and its major tributaries to the estuary of the Bohai Sea.24 In this study, we reanalyzed samples from the most highly contaminated river basin around the FP facility; the sample collection was composed of 25 surface water and 24 sediment samples that were collected on April 24−26, 2014. A follow-up sampling campaign was conducted on June 1−6, 2016, at approximately the same sampling sites as those in 2014 (See Figure 1). In total, 25 surface water (W1−W20, WT1-WT3, WT5, and WT6) and 24 sediment samples (S1− S19, ST1−ST3, ST5, and ST6) were collected every 5−10 km from the main river stream and the four main tributaries in 2016. Surface water samples (0−0.5 m depth) were collected using 500 mL polypropylene (PP) bottles that were precleaned with methanol and Milli-Q water. Water samples were kept at 4 °C in a fridge prior to the analysis of PFASs. Sediment samples were collected with a stainless steel shovel and were subsequently homogenized, freeze-dried under vacuum and stored at −20 °C until extraction. In addition to the water and sediment samples, technical APFO mixtures from three major producers in China that were collected in 2014 (see details in Shi et al.24) were analyzed for emerging PFASs using suspect screening. Standards and Reagents. The standards for native and mass-labeled PFASs (PFAC-MXB, MPFAC-MXA, including perfluorobutanoic acid (PFBA), perfluoropentanoic (PFPeA), perfluorohexanoic acid (PFHxA), perfluoroheptanoic acid (PFHpA), PFOA, perfluorononanoic acid (PFNA), perfluorodecanoic acid (PFDA), perfluorobutanesulfonic acid (PFBS), perfluorohexanesulfonic acid (PFHxS), perfluorooctanesulfonate (PFOS), 13C4−PFBA, 13C4−PFOA, 13C2−PFDA, 13C4− PFOS, and 18O2−PFHxS), HFPO−DA, M3HFPO−DA, ADONA and C8 chlorinated polyfluoroalkyl ether sulfonic acid (Cl-PFESA) were purchased from Wellington laboratories (Guelph, ON, Canada). The standards of HFPO-TrA (95%) and the tetramer acid of HFPO (HFPO-TeA, 95%) were purchased from J&K Chemical Co. Ltd. (Beijing, China). Nuclear magnetic resonance (Bruker-500 MHz NMR) was used to confirm the structure and purity of HFPO-TrA and HFPO-TeA standards (see SI Figure S2−S5). A complete list of chemicals and materials used for the sample treatment is provided in the SI. The chemical names and structures of positively identified and tentatively detected PFASs are shown in SI Figure S6. Sample Treatment. Both surface water and sediment samples were treated according to a previously validated methodology for anionic PFASs.24 In brief, sediment samples were extracted using methanol after lyophilization and then loaded into solid-phase extraction (SPE) cartridges. Surface

44−8) and the unnamed products from Asahi (CAS no. 908020−52−0) and Solvay (CAS no. 329238−24−6).18 However, recent studies employing nontarget screening approaches have discovered several additional classes of PFASs in wastewater or river water downstream of FP manufacturers, including monochlorine-substituted perfluoroalkyl carboxylic acids (Cl-PFCAs), monohydrogen-substituted PFCAs (H-PFCAs), poly hydrogen-substituted PFCAs (x H-PFCAs), monoether PFECAs, and polyether PFECAs.19−23 These studies clearly demonstrate that there is an array of substances being released into the environment, but the origin of these emerging PFASs in FP manufacturing and the magnitude of emissions, environmental transport and fate remain poorly understood. A parallel development in FP manufacturing, with major consequences for the global environmental inventories of PFASs, is the rapidly increasing production in Asia, particularly in China.15 Free from global restrictions and with no domestic emissions limit for PFOA, several FP manufacturers in China have continued to use ECF-based APFO as processing aids.15,24 Recent studies further conclude that some Chinese FP manufacturing facilities dominate the global emissions of PFOA with approximately 80 tonnes of emissions in the year 2014.25 Interestingly, recent monitoring studies of the Xiaoqing River, downstream from a major FP manufacturing plant in China, have also detected oligomers of hexafluoropropylene oxide (HFPO, C3F6O), including dimer acid (HFPO−DA, C6HF11O3) and trimer acid (HFPO-TrA, C9HF17O4) in concentrations ranging from 3.80 to 6.85 μg/ L.26,27 These studies provide indisputable evidence that Chinese FP manufacturers employ both alternative PFASs and APFO as processing aids. However, the interyear fluctuations in emissions and the presence of other emerging PFASs from this important FP production facility have not been studied to date. In this paper, we present a comprehensive spatial trend and interyear comparison of surface water and sediment samples from the Xiaoqing River using a suspect screening approach. The specific objectives were to investigate the occurrence and evaluate the environmental transport, fate, and riverine discharges of emerging PFASs from FP manufacturing. Throughout the paper we use the terminology “emerging PFASs” to denote the classes of recently discovered substances which have not yet been subject to any regulatory actions.



MATERIALS AND METHODS Study Area and Sampling. The Xiaoqing River, located in Shandong province, flows parallel to the Yellow river (Huang He) before emptying into the Laizhou Bay of the Bohai Sea. Highly elevated concentrations of perfluoroalkyl carboxylic acids (PFCAs) and HFPO oligomers in water samples from the Xiaoqing River have previously been reported. These concentrations have been linked to an industrial park that is dedicated to the production of fluorine and silicon materials operated by the Dongyue group.27,28 The annual production capacity of FPs at this facility in 2015 was 4.43 × 104 tonnes of PTFE resin, 1.30 × 104 tonnes of vinylidene fluoride (VDF), 1.00 × 104 tonnes of hexafluoropropylene (HFP), 1.00 × 104 tonnes of fluorine rubber (FKM), 8.40 × 103 tonnes of PVDF, and 5.50 × 103 tonnes of FEP.29 In addition to the polymeric products, low-molecularweight PFASs are produced at the facility including ECF-based APFO24 and T-5. T-5 is the trade name for an emulsifying C

DOI: 10.1021/acs.est.7b06657 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology water samples were prefiltered through a glass microfiber filter (0.70 μm, 47 mm), which was subsequently washed with 5 mL of methanol to prevent the adsorption of PFASs before extraction by SPE. Oasis WAX single-use cartridges (6 cc/150 mg) preconditioned with 4 mL of 0.1% ammonium hydroxide (in methanol), 4 mL of methanol, and 4 mL of ultrapure water were used to extract PFASs. After loading the samples, the cartridges were washed with 4 mL of buffer solution (25 mmol/L acetic acid/ammonium acetate, pH = 4) and 8 mL of ultrapure water and centrifuged for 10 min at 3000 rpm to remove the residual water. Finally, the analytes were eluted using 4 mL of methanol and 4 mL of 0.1% ammonium hydroxide (in methanol), which were combined and evaporated to 1 mL under nitrogen gas prior to injection. Targeted Analysis of PFASs. Quantitative analysis of PFASs was achieved by HPLC (Ultimate 3000, Thermo Fisher Scientific Co.) coupled with electrospray ionization tandem mass spectrometry (ESI-MS/MS, API 4500, Applied Biosystems/MDS SCIEX, USA). An Acclaim 120 C18 column (5 μm, 4.6 mm × 150 mm, Thermo Fisher Scientific Co.) was used to separate the target PFASs. The HPLC instrument was equipped with a dual pump (HPG-3200RS), autosampler (WPS-3000RS), column compartment (TCC-3000RS), and DCMS Link software. The target compound abbreviations, mass transitions and experimental conditions used for ESIMS/MS are shown in SI Table S1. Suspect Screening Analysis. The identification of suspect PFASs was performed using the HPLC-Orbitrap MS (Thermo Fisher Scientific Inc., Waltham, MA), which was operated using negative electrospray ionization (ESI−). Data were acquired in the full scan mode (60−1000 m/z) with a resolution of 120 000. The chromatographic separation of PFASs was achieved using both a high-resolution ACQUITY HSS PFP Column (1.8 μm, 100 Å, 50 mm × 2.1 mm, Waters Co.) and an Acclaim 120 C18 column (5 μm, 4.6 mm × 150 mm, Thermo Fisher Scientific Co.) in suspect screening. Additional details regarding the instrumentation for ESI-MS/ MS and Orbitrap MS are provided in the SI. The list of suspect compounds included 107 PFASs that have previously been reported in the vicinity of FP manufacturers and in lake trout with a nonspecified source is given in SI Table S2.19−23,33 Since these emerging PFASs were typically homologous series that differed by CF2, CF2CH2, or CF2O, we also included longer- and shorter-chain homologues relative to those previously detected in our suspect list. Evidence for identification included occurrence of a series of homologous masses, expected retention time (i.e., increasing with increasing length of the alkyl chain), similar chromatographic peak shape for members of a homologous series, occurrence of isotopes and/or adduct peaks, and interpretation of the MS/MS spectra.34 In the full scan mass spectrum, several suspicious peaks were found in the extracts. The Xcalibur software (Thermo Scientific) was used to predict molecular formulas via accurate molecular mass. Proposed molecular formulas were refined by mass error (Δm) within a threshold of 5 parts-per-million (ppm) accuracy. The compound structures were subsequently elucidated by the corresponding MS/MS fragmentation patterns. If available, MS/MS experiments were also conducted on the standards of emerging PFASs to confirm the molecular formulas and structures by comparison of the characteristic structure fragment ion and retention time. For the categorization and reporting of emerging PFASs identified by Orbitrap MS we

used the confidence level system proposed by Schymanski et al.35 Quality Assurance and Quality Control. All procedural, field, and solvent injection blanks were consistently below instrumental detection limits. Quantification was performed against a 10-point internal standard calibration curve (0.05, 0.1, 0.2, 0.5, 1, 2, 5, 10, 20, 50 μg/L), that was spiked with 2 ng of M3HFPO−DA, 13C4−PFBA, 13C4−PFOA, 13C2−PFDA, 13 C4−PFOS, and 18O2−PFHxS, using a 1/x2 weighted regression (R2 > 0.98). The accuracy and precision for target PFASs were evaluated by spike-recovery experiments where 5 ng of native PFAS homologues were added to the water and sediment samples. The mean matrix spike recoveries (n = 5) of HFPO−DA, HFPO-TrA, HFPO-TeA and ADONA in water and sediment samples ranged from 89 ± 15% to 102 ± 9.3% and 83 ± 9.4% to 103 ± 12%, respectively (shown in SI Table S3). The recoveries of other PFASs ranged from 94 ± 8.4% to 112 ± 3.1% and 95 ± 4.8% to 107 ± 12% for water and sediment, respectively (all results are shown in SI Table S3). Highly contaminated surface water samples (W12−W20, WT1−WT3, WT5, and WT6) were diluted and requantified to ensure that the peak areas were within the calibration range. The limits of quantification (LOQ) were defined as the lowest concentration of each compound resulting in a signal-to-noise ratio (S/N) ≥ 10. Data Treatment. Sediment-water distribution coefficients (Kd‑con) were calculated for paired samples that were collected along the spatial gradient of target PFASs (C4−C10 PFCAs and HFPO-TrA) according to eq 1. Kd‐con =

Csed × 103 Cwater

(1)

where Csed is the concentration (ng/g dry weight (dw)) of PFASs in sediment and Cwater is the concentration of PFASs in water (ng/L). For tentatively detected PFASs that could not be quantified against authentic reference standards, the peak areabased sediment-water distribution coefficients (Kd‑area) were calculated according to eq 2. Kd‐area =

A sed × 200 A water

(2)

where Ased is the peak area of PFASs in sediment, Awater is the peak area of PFASs in water. A conversion factor of 200 was included to correct for the difference in sample weight (1 g sediment) and volume (200 mL) used for extraction. The validity of the estimated sediment−water distribution was subsequently evaluated in a training set of C4−C10 PFCAs and HFPO-TrA for which both Kd‑area and Kd‑con were calculated. Statistical analysis was performed in IBM PASW statistics 18.0 (SPSS Inc., 1993−2007). Correlations of peak area data were performed by Spearman rank analysis with a statistical significance threshold of p < 0.05. Pearson’s correlation analysis (p < 0.05) was used to investigate the correlations between log Kd‑con and log Kd‑area of C4−C10 PFCAs and HFPO-TrA. The Mann−Whitney U-test was applied to examine statistically significant differences of log Kd‑area for different series of PFASs. The riverine discharges of ∑PFCAs and ∑HFPO oligomers were calculated using concentration data in 2014 and 2016 and the annual water flow data according to the approach reported in our previous paper.24 D

DOI: 10.1021/acs.est.7b06657 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

Figure 2. Chromatograms (C18 column, transition: 184.9/118.9) of HFPO−DA, HFPO-TrA, and HFPO-TeA in standard, surface water, and sediment samples with HPLC-MS/MS.

(confirmed structures, SI Table S4) and to quantify their concentrations using the internal standards of PFCAs. In the following subsections we present the peak area-based relative abundance, homologue distribution and isomeric profiles of emerging PFASs in relation to previous studies.19−23 A more detailed interpretation of mass-spectral evidence of their detection is presented in the SI. HFPO Oligomers (DA, TrA, TeA). Figure 2 shows the chromatograms of three detected HFPO oligomers in standard solution, water and sediment samples, respectively. The absence of multiple peaks using both the C18 and PFP columns combined with identical mass-spectra compared to the standard demonstrated that no additional isomers were present in the environmental samples. It is worth noting that the response of the [M]− peak of HFPO-TrA and HFPO-TeA was relatively low compared to its fragments, which might be due to in-source fragmentation. Thus, we investigated the possibility of using different MS/MS transitions (184.9/118.9) from the one reported by Pan et al.27 for HFPO-TrA (494.9/ 184.9). As shown in SI Figures S7−S10, the characteristic fragments including [C2F5]− (m/z = 118.99256), [C3F7O]− (m/z = 184.98429), and [C6F13O2]− (m/z = 350.96962) had significantly higher sensitivity than the parent ions for HFPOTrA and HFPO-TeA. Therefore, we used the 184.9/118.9 transition for the quantification of both HFPO-TrA and HFPO-TeA, which could be distinguished by their different retention times. Cl-PFCAs (ClCnF2n−2HO2, n = 5−12). Cl-PFCAs were tentatively identified with C5−C10 homologues detected in both surface water and sediment, while C11 and C12 homologues were only detected in sediment samples (chromatograms in SI Figure S11 and mass-spectra in SI Figure S12). In contrast to the study of Liu et al.,20 where ClPFCA homologues were reported as single peaks, at least two

For emerging PFASs, riverine emissions were estimated based on their relative peak area ratio to ∑PFCAs.



RESULTS AND DISCUSSION Identification of Emerging PFASs. In total, 42 emerging PFASs (39 tentatively identified and 3 with structures confirmed by reference standards) were detected in surface water and/or sediment samples. SI Table S4 shows their observed mass, protonated formula, corresponding mass error (Δm < 2 ppm to predict molecular formula) and the confidence level for identification. Among the identified compounds, HFPO-TeA was positively confirmed and quantified for the first time in the environment, also C9− C14 monoether PFECAs were tentatively detected for the first time. The series of emerging PFASs were detected based on expected retention time, accurate mass, MS/MS fragmentation pattern, similar chromatographic peak shape within each class, isotope and/or adduct peaks which were consistent with the literature.19−23 Given these detection criteria, there was little ambiguity in the molecular formulas for emerging PFASs, although the exact structures of detected compounds could not be determined. In fact, the use of a PFP column for identification demonstrated that all classes of emerging PFASs, except HFPO oligomers and polyether PFECAs, had multiple isomeric peaks that further complicated the elucidation of the structure. Thus, without mass-spectral library information or clear diagnostic evidence, it was not possible to determine the exact structures of the individual peaks. Therefore, according to the confidence level system reported by Schymanski et al.,35 the 39 emerging PFASs detected in our study would fall into category 3 (tentative candidate structures, SI Table S4). For the three HFPO oligomers, the available authentic standards allowed us to identify these substances at the highest confidence level E

DOI: 10.1021/acs.est.7b06657 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

contrast to previous studies which reported C3−C6 homologues as the dominant compounds,22,23 surface water samples from the Xiaoqing River featured C8 and C9 as the dominant homologues based on peak area. Polyether PFECAs (CnF2n−1HOn, n = 4−7). C4−C7 polyether PFECAs differing by CF2O (m/z = 65.9917) were tentatively detected in surface water and sediment extracts (SI Figure S22 and Figure S23), while no polyether PFECAs with longer chains were detected. Consistent with previous literature,22,23 no isomers were found in this series of emerging PFASs where C4 and C5 homologues were the most abundant compounds based on peak area. Spatial Trends and Semiquantitative Emission Estimates. In accordance with previous studies,27 the quantitative target analysis from the Xiaoqing River showed exceptionally high concentrations of ∑PFCAs (54.7−4.96 × 105 ng/L) with PFOA as the dominating homologue in water samples from 2014 and 2016. The water concentrations of HFPO-TrA (0.73 in 2016 (See SI Figure S24). The strong intercorrelations of emerging PFASs and PFCAs indicate that they undergo similar transport and fate processes which further imply that (i) horizontal transport, where the bulk flow of water is the dominant transport process and (ii) no significant degradation of emerging PFASs occurs over the hydraulic residence time of the watershed.24,37 The weaker intercorrelations for PFDA and the rapid drop in detection frequency in surface water with increasing distance to the source for emerging PFASs with C9−C10 chain-lengths further indicate that sorption to particulate matter and sedimentation is a relatively more important transport mechanism for these compounds. To further investigate the fate of emerging PFASs, Kd‑area was calculated for PFASs that were tentatively detected in both water and sediment. As shown in SI Figure S25, log Kd‑con was strongly correlated with log Kd‑area in a training set of C4− C10 PFCAs and HFPO-TrA (Pearson’s correlation, p < 0.05, r2 = 0.98). The strong correlation indicated that the peak areabased Kd values could provide a reasonable estimate of the relative hydrophobicity for emerging PFASs despite that they could not be accurately quantified. SI Figure S26 displays log Kd‑area for emerging PFASs and C4−C10 PFCAs with an increasing number of carbons. Overall, log Kd‑area increased with perfluoroalkyl chain-length, although some variability was observed between the classes of emerging PFASs. The median

facility situated in the tributary of the Xiaoqing River was the dominant source of all emerging PFASs that were identified in this study (Figure 1). Downstream samples showed decreasing peak areas of PFASs with increasing distance to the source, which reflects dilution, sedimentation and/or transformation of PFASs which is discussed in more detail below. The presence of PFCAs and emerging PFASs in upstream and downstream samples at similar peak area ratios may be explained by atmospheric emissions and deposition to the river catchment and/or surface water runoff from the manufacturing facility. The interyear comparison for both sediment and water showed a fairly consistent absolute and relative abundance of PFCAs and emerging PFASs, which indicates that the emissions have remained approximately constant over this period of time. Based on the peak areas relative to ∑PFCAs, the discharges of ∑monoether PFECAs and ∑polyether PFECAs were estimated to be 2.84 tonnes/year and 3.59 tonnes/year, respectively. The emissions of H-PFCAs, x HPFCAs and Cl-PFCAs were approximately an order of magnitude lower (SI Table S11). However, it should be noted that differences in response factors between individual compounds and classes render these discharge estimates highly uncertain, and targeted analysis using authentic standards is strongly encouraged to better constrain the emissions. Furthermore, fluctuations in emissions or water flow conditions could introduce uncertainty to riverine discharge estimates. For example, water samples downstream of the FP manufacturer that were collected in 2016 displayed a minimum in peak area for sampling site 12, which was not observed in 2014. Continuous monitoring of the Xiaoqing River surface water with higher temporal resolution would therefore be needed to reduce uncertainty in riverine discharges and to evaluate temporal trends. G

DOI: 10.1021/acs.est.7b06657 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology values of log Kd‑area of H-PFCAs (C8, C10), and monoether PFECAs (C8, C9) were slightly lower than corresponding PFCAs with the same number of carbons. The log Kd‑area of C10 x H-PFCA was significantly lower than that of C10 PFCA (Mann−Whitney U-test, P < 0.05). Slightly higher median values of log Kd‑area of polyether PFECAs (C4, C5), Cl-PFCAs (C5, C6, C8), and HFPO-TrA compared to log Kd‑area of the corresponding PFCAs with the same number of carbons that was observed although the difference was not statistically significant (Mann−Whitney U-test, P > 0.05). The estimated trends in relative hydrophobicity for emerging PFASs gained from Kd‑area were generally in line with the trends in retention times on a C18 column (SI Figure S27). Retention times for different classes of PFASs with the same number of carbon atoms were, in order from longest to shortest: polyether PFECAs ≥ monoether PFECAs > HFPO oligomers ≥ ClPFCAs ≥ PFCAs > H-PFCAs > x H-PFCAs. The trends in relative hydrophobicity from both Kd‑area and retention times could generally be explained by their predicted molecular volume changes that result from the insertion of ether bonds or the substitution of fluorine with chlorine or hydrogen atoms.17,38 The contradictory result of monoether PFECAs, which are less hydrophobic than PFCAs based on the log Kd‑area value, but more hydrophobic based on retention time, can possibly be explained by different interactions with the solid phase in sediments (which contains both mineral surfaces and organic carbon) compared to the C18 stationary phase.36 Nevertheless, improved sorption studies using controlled batch experiments and isomer-specific authentic reference standards would be needed to gain an understanding of the sorption mechanisms for this diverse class of PFASs.39 Origin of Emerging PFASs in FP Manufacturing. The co-occurrence of numerous emerging PFASs in the Xiaoqing River and other sites impacted by FP manufacturing demonstrates that the detection of these substances is not an isolated occurrence; this suggests that specific processes and/ or use patterns that may be common for several sites around the world are causing the emissions. In Figure 3, an overview of the emerging PFASs detected in this study is presented and compared with previous studies. For HFPO oligomers, the observed homologue patterns were in good agreement with our knowledge about the polymerization aids being used at the different locations. The dominance of HFPO-TrA in the Xiaoqing River is consistent with the reported use of T-5 as a polymerization aid in PVDF production by the Dongyue group.27,30 Contrastingly, the dominance of HFPO−DA at the European and North American sites can be explained by the use of Gen-X as a replacement for PFOA.40 The common findings of HFPO oligomers, mono- and poly ether PFECAs in Cape Fear Basin22,23 and in the Xiaoqing River may further indicate that these compounds have a common source that is related to the oligomerization process. Such an explanation is partly supported by the fact that monoether PFECAs were detected as impurities in the standard solution of HFPO-TrA (SI Figure S28), which is produced by the same synthesis route as T-5. However, since the HFPO-TrA standard was purified to >95% it may not be entirely representative of the impurities present in the T-5 technical mixtures. Given that HFPO and HFPO oligomeric acyl fluoride homologues (C2F5[CF2OCF(CF3)]nCOF, n = 1, 2, 3···) are produced for several different products including functionalized perfluoropolyethers, olefin and acrylate monomers (SI Figure S29),31,41−43 additional

byproducts may also be formed and released to waste streams during this process. For x H-PFCAs, two hypotheses have previously been proposed by Newton et al.21 to explain their occurrence in environmental samples. The first hypothesis suggests that x HPFCAs are intentionally synthesized by telomerization and used as alternatives to PFOA in FP manufacturing.21 Given the consistent observations of multiple isomeric structures of x HPFCAs from two different locations, this hypothesis seems rather unlikely since the telomerization process would result in isomeric pure products.1 The second hypothesis stipulates that x H-PFCAs can be formed as byproducts from PVDF polymerization using VDF as the starting material. 21 Considering the large volumes of PVDF produced by the Dongyue FP facility,29 the second hypothesis seems plausible for explaining the presence of x H-PFCAs in the Xiaoqing River. However, further studies are needed to elucidate the origin of these compounds in FP production. The occurrence of H-PFCAs in the Xiaoqing River samples may be linked to two different emission processes related to PVDF production. First, H-PFCAs have been reported in patents as alternative polymerization aids for PVDF.20 Second, H-PFCAs were detected as impurities in the HFPO-TrA standard (SI Figure S28) which, analogously to monoether PFECAs, suggests that they can be present in T-5 technical products. Given the relatively low abundance of H-PFCAs in surface water, we find it more probable that these substances are impurities rather than intentionally used as polymerization aids. However, these two proposed release mechanisms are not mutually exclusive and may occur in parallel with each other. For Cl-PFCAs there are patents from both Europe and the U.S. describing the use of these substances as alternative polymerization aids to replace APFO.30,42 Interestingly, ClPFCAs were also detected in technical mixtures of APFO from the Chinese market (see SI Figure S30), which demonstrates that these substances can be formed as impurities from the ECF process. Given that the peak areas of ∑Cl-PFCAs in surface water and sediment were typically 2 orders of magnitude lower than those of ∑PFCAs (Figure 1), it seems plausible that the emissions of these substances are related to small-scale testing during process optimization and/ or impurities in APFO technical mixtures. Implications and Future Directions. The present study revealed that several classes of emerging PFASs, without welldocumented uses in FP manufacturing, are emitted in quantities of several tonnes per year into the Xiaoqing River. Based on the apparent persistence and the high mobility in water, the emerging PFASs identified here have a similar intrinsic mobility and potential to contaminate drinking water reservoirs as PFOA.8 Although the Xiaoqing River is not used for municipal drinking water, human exposure may occur from the consumption of private well water and the use of river water for the irrigation of crops.8,44,45 Given the efficient transport of C3−C9 emerging PFASs in surface water, it is also likely that these substances will become globally distributed contaminants if the emissions continue. As the origins for many of these substances in FP manufacturing remain speculative and probably originate from multiple processes, it is difficult to estimate both the historical and future emissions. A systematic analysis of commercially used polymerization aids and finished polymeric products would help to resolve whether the emerging PFASs detected here and elsewhere are intentionally used as alternatives to PFOA or are impurities H

DOI: 10.1021/acs.est.7b06657 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology from the synthesis of surfactants or polymer reactions.15 Such data could subsequently be combined with FP production and consumption data of polymerization aids to construct global and regional emission inventories of emerging PFASs to aid informed decision-making about emission reduction and substitution strategies.15



(3) Kissa, E. Fluorinated Surfactants and Repellents, 2nd ed.; Marcel Dekker, Inc.: New York, 2001; Vol. 97. (4) Prevedouros, K.; Cousins, I. T.; Buck, R. C.; Korzeniowski, S. H. Sources, fate and transport of perfluorocarboxylates. Environ. Sci. Technol. 2006, 40 (1), 32−44. (5) Armitage, J. M.; Schenker, U.; Scheringer, M.; Martin, J. W.; MacLeod, M.; Cousins, I. T. Modeling the global fate and transport of perfluorooctane sulfonate (PFOS) and precursor compounds in relation to temporal trends in wildlife exposure. Environ. Sci. Technol. 2009, 43 (24), 9274−9280. (6) Cousins, I. T.; Kong, D.; Vestergren, R. Reconciling measurement and modelling studies of the sources and fate of perfluorinated carboxylates. Environ. Chem. 2011, 8 (4), 339−354. (7) Bartell, S. M.; Calafat, A. M.; Lyu, C.; Kato, K.; Ryan, P. B.; Steenland, K. Rate of decline in serum PFOA concentrations after granular activated carbon filtration at two public water systems in Ohio and West Virginia. Environ. Health Perspect. 2010, 118 (2), 222−228. (8) Hoffman, K.; Webster, T. F.; Bartell, S. M.; Weisskopf, M. G.; Fletcher, T.; Vieira, V. M. Private drinking water wells as a source of exposure to perfluorooctanoic acid (PFOA) in communities surrounding a fluoropolymer production facility. Environ. Health Perspect. 2011, 119 (1), 92−97. (9) Steenland, K.; Zhao, L.; Winquist, A.; Parks, C. Ulcerative colitis and perfluorooctanoic acid (PFOA) in a highly exposed population of community residents and workers in the Mid-Ohio valley. Environ. Health Perspect. 2013, 121 (8), 900−905. (10) Winquist, A.; Steenland, K. Perfluorooctanoic acid exposure and thyroid disease in community and worker cohorts. Epidemiology 2014, 25 (2), 255. (11) Han, X.; Nabb, D. L.; Russell, M. H.; Kennedy, G. L.; Rickard, R. W. Renal elimination of perfluorocarboxylates (PFCAs). Chem. Res. Toxicol. 2012, 25 (1), 35−46. (12) Han, X.; Snow, T. A.; Kemper, R. A.; Jepson, G. W. Binding of perfluorooctanoic acid to rat and human plasma proteins. Chem. Res. Toxicol. 2003, 16 (6), 775. (13) Lindstrom, A. B.; Strynar, M. J.; Libelo, E. L. Polyfluorinated compounds: Past, present, and future. Environ. Sci. Technol. 2011, 45 (19), 7954−7961. (14) 3M. Letter to US EPA, Re: Phase-out plan for POSF-based products(226−0600); US EPA Administrative Record 226, 2000. (15) Wang, Z.; Cousins, I. T.; Scheringer, M.; Buck, R. C.; Hungerbühler, K. Global emission inventories for C4−C14 perfluoroalkyl carboxylic acid (PFCA) homologues from 1951 to 2030, Part I: Production and emissions from quantifiable sources. Environ. Int. 2014, 70, 62−75. (16) USEPA. 2010/2015 PFOA Stewardship Program. http://www. epa.gov/oppt/pfoa/pubs/stewardship/ (last updated on January 16 2013). (17) Wang, Z.; Cousins, I. T.; Scheringer, M.; Hungerbuehler, K. Hazard assessment of fluorinated alternatives to long-chain perfluoroalkyl acids (PFAAs) and their precursors: Status quo, ongoing challenges and possible solutions. Environ. Int. 2015, 75, 172−179. (18) Wang, Z.; Cousins, I. T.; Scheringer, M.; Hungerbühler, K. Fluorinated alternatives to long-chain perfluoroalkyl carboxylic acids (PFCAs), perfluoroalkane sulfonic acids (PFSAs) and their potential precursors. Environ. Int. 2013, 60, 242−248. (19) Gebbink, W. A.; van Asseldonk, L.; van Leeuwen, S. Presence of emerging per- and polyfluoroalkyl substances (PFASs) in river and drinking water near a fluorochemical production plant in the Netherlands. Environ. Sci. Technol. 2017, 51 (19), 11057−11065. (20) Liu, Y.; Pereira, A. D. S.; Martin, J. W. Discovery of C5-C17 poly- and perfluoroalkyl substances in water by in-line SPE-HPLCOrbitrap with in-source fragmentation flagging. Anal. Chem. 2015, 87 (8), 4260−4268. (21) Newton, S.; McMahen, R.; Stoeckel, J. A.; Chislock, M.; Lindstrom, A.; Strynar, M. Novel polyfluorinated compounds identified using high resolution mass spectrometry downstream of

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.est.7b06657. Standards and reagents, instrument conditions, detailed identification text, target analysis data (Table S1, S3, and S5−S8), suspect analysis list (Table S2, S4), peak area ratios (Table S9 and S10), estimated riverine emissions (Table S11), oligomerization process of T-5 product (Figure S1), NMR results of HFPO-TrA and TeA standard solution (Figure S2−S5), chemical structures or molecular formulas of PFASs (Figure S6), mass spectra and/or chromatograms of emerging PFASs (Figure S7−S23), correlations of PFASs peak area in surface water (Figure S24), correlations between log Kd‑con and log Kd‑area value (Figure S25), box and whisker plots of log Kd‑area (Figure S26), retention time of PFASs on C18 column (Figure S27), impurities in HFPO-TrA standard solution (Figure S28), schematic picture of industrial applications for HFPO and HFPO oligomeric acyl fluoride (Figure S29), impurities in APFO products (Figure S30), and the interyear concentration spatial trends of predominant target PFASs (Figure S31) are available (PDF)



AUTHOR INFORMATION

Corresponding Author

*Phone: +86 (10) 62849676; fax: +86 (10) 62849182; e-mail: [email protected]. ORCID

Yali Shi: 0000-0001-9946-1525 Jun Huang: 0000-0001-9207-8953 Yaqi Cai: 0000-0002-2805-5535 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was jointly supported by the National Natural Science Foundation of China (No. 21722705, 21537004, 21677154, 21777182), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB14010201), and the Swedish Research Council FORMAS (No. 2014-514). The authors gratefully appreciate Dr. Jana Johansson from Stockholm University for her useful suggestions to improve this manuscript.



REFERENCES

(1) Buck, R. C.; Franklin, J.; Berger, U.; Conder, J. M.; Cousins, I. T.; de Voogt, P.; Jensen, A. A.; Kannan, K.; Mabury, S. A.; van Leeuwen, S. P. J. Perfluoroalkyl and polyfluoroalkyl substances in the environment: Terminology, classification, and origins. Integr. Environ. Assess. Manage. 2011, 7 (4), 513−541. (2) Bowman, J. S. Fluorotechnology is critical to modern life: The fluorocouncil counterpoint to the madrid statement. Environ. Health Perspect. 2015, 123 (5), A112−A113. I

DOI: 10.1021/acs.est.7b06657 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology manufacturing facilities near Decatur, Alabama. Environ. Sci. Technol. 2017, 51 (3), 1544−1552. (22) Strynar, M.; Dagnino, S.; McMahen, R.; Liang, S.; Lindstrom, A.; Andersen, E.; McMillan, L.; Thurman, M.; Ferrer, I.; Ball, C. Identification of novel perfluoroalkyl ether carboxylic acids (PFECAs) and sulfonic acids (PFESAs) in natural waters using accurate mass time-of-flight mass spectrometry (TOFMS). Environ. Sci. Technol. 2015, 49 (19), 11622−11630. (23) Sun, M.; Arevalo, E.; Strynar, M.; Lindstrom, A.; Richardson, M.; Kearns, B.; Pickett, A.; Smith, C.; Knappe, D. R. U. Legacy and emerging perfluoroalkyl substances are important drinking water contaminants in the Cape Fear River watershed of North Carolina. Environ. Sci. Technol. Lett. 2016, 3 (12), 415−419. (24) Shi, Y.; Vestergren, R.; Xu, L.; Song, X.; Niu, X.; Zhang, C.; Cai, Y. Characterizing direct emissions of perfluoroalkyl substances from ongoing fluoropolymer production sources: A spatial trend study of Xiaoqing River, China. Environ. Pollut. 2015, 206 (1), 104−112. (25) Wang, T.; Vestergren, R.; Herzke, D.; Yu, J.; Cousins, I. T. Levels, isomer profiles, and estimated riverine mass discharges of perfluoroalkyl acids and fluorinated alternatives at the mouths of Chinese rivers. Environ. Sci. Technol. 2016, 50 (21), 11584−11592. (26) Heydebreck, F.; Tang, J.; Xie, Z.; Ebinghaus, R. Alternative and legacy perfluoroalkyl substances: Differences between European and Chinese river/estuary systems. Environ. Sci. Technol. 2015, 49 (14), 8386−8395. (27) Pan, Y.; Zhang, H.; Cui, Q.; Sheng, N.; Yeung, L. W. Y.; Guo, Y.; Sun, Y.; Dai, J. First report on the occurrence and bioaccumulation of hexafluoropropylene oxide trimer acid: An emerging concern. Environ. Sci. Technol. 2017, 51 (17), 9553−9560. (28) Su, H.; Lu, Y.; Wang, P.; Shi, Y.; Li, Q.; Zhou, Y.; Johnson, A. C. Perfluoroalkyl acids (PFAAs) in indoor and outdoor dusts around a mega fluorochemical industrial park in China: Implications for human exposure. Environ. Int. 2016, 94, 667−673. (29) Dongyue Group Limited, 2015 Annual report, http://www. dongyuechem.com/ (in Chinese). (30) Wang, H.; Fang, J.; Li, W.; Li, X.; Hao, H. A preparation method of free-PFOA fluoropolymer. CN101745338A, 2010-06−23. (patent in Chinese). (31) Wang, H.; Li, X.; Shi, H. Influence of PFOA-free fluorocontaining surfactant on synthesis of PVDF resin. Coat. Technol. & Abstr. (in Chinese) 2015, No. 11, 9−12. (32) Wang, P.; Fang, M.; Guan, Y.; Fu, G.; Xu, Y. Progress on application of the substitutes for PFOA in fluoro-polymer synthesis. Zhejiang Chem. Indust. (in Chinese) 2013, 44 (11), 1−4. (33) Fakouri Baygi, S.; Crimmins, B. S.; Hopke, P. K.; Holsen, T. M. Comprehensive emerging chemical discovery: Novel polyfluorinated compounds in Lake Michigan trout. Environ. Sci. Technol. 2016, 50 (17), 9460−9468. (34) Schymanski, E. L.; Singer, H. P.; Longrée, P.; Loos, M.; Ruff, M.; Stravs, M. A.; Ripollés Vidal, C.; Hollender, J. Strategies to characterize polar organic contamination in wastewater: Exploring the capability of high resolution mass spectrometry. Environ. Sci. Technol. 2014, 48 (3), 1811−1818. (35) Schymanski, E. L.; Jeon, J.; Gulde, R.; Fenner, K.; Ruff, M.; Singer, H. P.; Hollender, J. Identifying small molecules via high resolution mass spectrometry: Communicating confidence. Environ. Sci. Technol. 2014, 48 (4), 2097−2098. (36) Higgins, C. P.; Luthy, R. G. Sorption of perfluorinated surfactants on sediments. Environ. Sci. Technol. 2006, 40 (23), 7251− 7256. (37) McLachlan, M. S.; Holmström, K. E.; Reth, M.; Berger, U. Riverine discharge of perfluorinated carboxylates from the European continent. Environ. Sci. Technol. 2007, 41 (21), 7260−7265. (38) Gomis, M. I.; Wang, Z.; Scheringer, M.; Cousins, I. T. A modeling assessment of the physicochemical properties and environmental fate of emerging and novel per- and polyfluoroalkyl substances. Sci. Total Environ. 2015, 505, 981−991. (39) Wang, Z.; Macleod, M.; Cousins, I. T.; Scheringer, M.; Hungerbühler, K. Using COSMOtherm to predict physicochemical

properties of poly- and perfluorinated alkyl substances (PFASs). Environ. Chem. 2011, 8 (4), 389−398. (40) Chemours. DuPont GenX processing aid for making fluoropolymer resins. https://www.chemours.com/Industrial_ Bakery_Solutions/en_GB/assets/downloads/Chemours_GenX_ Brochure_Final_07July2010.pdf (accessed July 5th 2016). (41) Wang, H.; Wang, Z.; Zhang, Y. Preparation and application of hexafluoropropylene oxide. Organo-Fluorine Indust. (in Chinese) 2006, No. 03, 23−26. (42) Fang, M.; Xu, Y.; Fu, G.; Wang, P.; Wu, X.; Yang, J. A polymerization method of VDF homopolymer or copolymer emulsion. CN101906182A, 2009-6−8. (patent in Chinese). (43) Zhu, H.; Yu, K.; Wang, Z. Application of hexafluoropropylene oxide oligomer. Organo-Fluorine Indust. (in Chinese) 2004, No. 03, 46−48 + 40. (44) Zhang, H.; Vestergren, R.; Wang, T.; Yu, J.; Jiang, G.; Herzke, D. Geographical differences in dietary exposure to perfluoroalkyl acids between manufacturing and application regions in China. Environ. Sci. Technol. 2017, 51 (10), 5747−5755. (45) Felizeter, S.; McLachlan, M. S.; De Voogt, P. Uptake of perfluorinated alkyl acids by hydroponically grown lettuce (Lactuca sativa). Environ. Sci. Technol. 2012, 46 (21), 11735−11743.

J

DOI: 10.1021/acs.est.7b06657 Environ. Sci. Technol. XXXX, XXX, XXX−XXX