First Report on the Occurrence and Bioaccumulation Of

Article

pubs.acs.org/est

First Report on the Occurrence and Bioaccumulation of Hexaﬂuoropropylene Oxide Trimer Acid: An Emerging Concern † ‡ † † † § ∥ ∥ Yitao Pan, , Hongxia Zhang, Qianqian Cui, Nan Sheng, Leo W. Y. Yeung, Yong Guo, Yan Sun, † and Jiayin Dai*, † Key Laboratory of Animal Ecology and Conservation Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing 100101, P. R. China ‡ University of Chinese Academy of Sciences, Beijing 100049, China § Man-Technology-Environment Research Centre (MTM), School of Science and Technology, Örebro University, SE-70182 Örebro, Sweden ∥ Key Laboratory of Organoﬂuorine Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Shanghai 200032, P. R. China

*S Supporting Information

ABSTRACT: Here, we report on the occurrence of a novel perfluoroalkyl ether carboxylic acid, ammonium perfluoro-2- [(propoxy)propoxy]-1-propanoate (HFPO-TA), in surface water and common carp (Cyprinus carpio) collected from the Xiaoqing River and in residents residing near a fluoropolymer production plant in Huantai County, China. Compared with the levels upstream of the Xiaoqing River, HFPO-TA concentrations (5200−68500 ng/L) were approximately 120−1600-times higher downstream after receiving fluoropolymer plant effluent from a tributary. The riverine discharge of HFPO-TA was estimated to be 4.6 t/yr, accounting for 22% of total PFAS discharge. In the wild common carp collected downstream from the point source, HFPO-TA was detected in the blood (median: 1510 ng/mL), liver (587 ng/g ww), and muscle (118 ng/g ww). The log BCFblood of HFPO-TA (2.18) was significantly higher than that of PFOA (1.93). Detectable levels of HFPO-TA were also found in the sera of residents (median: 2.93 ng/mL). This is the first report on the environmental occurrence and bioaccumulation of this novel chemical. Our results indicate an emerging usage of HFPO-TA in the fluoropolymer manufacturing industry and raise concerns about the toxicity and potential health risks of HFPO-TA to aquatic organisms and humans.

■ INTRODUCTION fonate (PFOS), its salts, and related substances.6 In 2015, the Per- and polyfluoroalkyl substances (PFASs) are synthetic Risk Assessment Committee from the European Union fluorinated chemicals that have been used since the 1950s.1 The adopted the German and Norwegian proposal to restrict the manufacture, use, and marketing of PFOA, its salts, and related unique amphiphilic properties of PFASs have made them useful 7 in a wide variety of industrial applications, such as the substances. fl Since these restrictions, manufacturers have started to production of uoropolymers, surface repellent coatings, 8 fi fi 2 produce shorter-chain perfluorinated and other fluorinated metal plating, and re- ghting foam. Legacy PFASs, typically 9 long-chain (seven perfluorinated carbons or longer) perfluor- compounds as alternatives, which include functionalized fl perfluoropolyethers (PFPEs) such as perfluoroether carboxylic oalkyl carboxylic acids (PFCAs) and per uoroalkanesulfonic 4 acids (PFSAs)1 are of great concern due to their environmental and sulfonic acids (PFECAs and PFESAs). By inserting one or fl persistence, bioaccumulation potential, and possible toxicity.3,4 more ether oxygens into the per uorinated carbon backbone, it 10 As a result, global regulations have been issued to reduce the is hoped that PFECAs and PFESAs are more degradable and 11−13 production and use of these compounds.3 In 2006, eight major have replaced PFCAs and PFSAs in many applications. In fluorochemical companies participated in the 2010/2015 chrome plating, chlorinated polyfluorinated ether sulfonic acids Perfluorooctanoate (PFOA) Stewardship Program proposed by the US Environmental Protection Agency, which aimed to Received: May 2, 2017 eliminate the production and emission of PFOA by 2015.5 In Revised: July 31, 2017 2009, the Stockholm Convention on Persistent Organic Accepted: August 7, 2017 Pollutants initiated regulation of the use of perfluorooctanesul- Published: August 7, 2017

(6:2 and 8:2 Cl-PFESAs) have been used as mist suppressants to replace PFOS in China11 and have since been widely − detected in abiotic and biotic environments.11,14 18 In fluoropolymer manufacturing, certain PFECAs, such as perfluoro-2-propoxypropanoic acid (HFPO-DA), have been used as an alternative to PFOA. Since 2010, the ammonium salt of HFPO-DA (GenX produced by DuPont)12 has been produced at 10−100 tons per year in Europe,13 and has subsequently been observed in river waters downstream of fluorochemical industrial parks in Germany (107.6 ng/L),19 China (3825 ng/L),19 and the U.S. (631 ng/L).20 In addition, several other structurally similar chemicals have also been identified in the U.S., suggesting varied and widespread usage of PFECA homologues.10,20 Hexafluoropropylene oxide (HFPO) is a well-known key compound in organofluorine chemistry.21 Including HFPO- DA, which is the dimer acid of HFPO (structure shown in Figure S1), oligomeric HFPO can be applied as a monomer or intermediate in the synthesis of fluorinated chemicals.21 The trimer acid of HFPO, HFPO-TA (Figure S1), is used as a Figure 1. Sampling sites in Xiaoqing River. processing aid in the manufacture of fluorinated polymers, such as polytetrafluoroethylene and polyvinylidene fluoride,22 and is Common carp (Cyprinus carpio) were captured in the area an important building block in the synthesis of other between S12−S13 on December 1, 2015 (n = 15). Information fluorinated products, including surface-active agents,23 oil- on gender, body weight, and length can be found in Table S3. − repellent agents,24 ionic liquids,25 and industrial additives.26 28 Approximately 2−4 mL of whole blood was collected Available information on the physical and chemical properties immediately in EDTA-coated vacutainer tubes (BD Bioscien- of HFPO-TA are shown in Table S1. However, information is ces, USA). Liver and muscle samples were carefully dissected scarce with regard to its annual production, environmental from the fish, wrapped with aluminum foil, and maintained at occurrence, wildlife or human exposure, bioaccumulation −20 °C. potential, and toxic effects. Human subjects (22 male and 26 female) were recruited at In the present investigation, water and fish samples were Huantai County Hospital, located 8 km from the fluoropolymer collected from various sites in Xiaoqing River, China. Elevated plant. Participants were residents recruited at their first concentrations of PFCAs have been reported previously in presentation to the hospital in January 2016. All subjects had water19,29 and sediment samples of Xiaoqing River,30 which are lived in Huantai for at least two years and had never worked in likely due to discharge from one of the largest fluoropolymer the fluoropolymer plant. Blood samples were centrifuged production facilities in Asia,29 which has a reported annual immediately after collection with sera transferred and stored production of approximately 37000 t of polytetrafluoroethylene at −80 °C until analysis. The research protocol was approved (PTFE), 500 t of perfluorinated ethylene-propylene copoly- by the Ethics Committee of the Institute of Zoology, Chinese mers, 300 t of polyvinylidenefluoride (PVDF), and 40 t of Academy of Sciences, and the study hospital. ammonium perfluorooctanoate.31 Human blood samples from Sample Extraction. The water and biota samples were 32,33 local residents in Huantai County, where the fluoropolymer extracted based on previously published methods. Details production facility is located, were also collected. The of the extraction method on different matrices are provided in objectives of the present investigation were (1) to investigate the Supporting Information (SI). In brief, water samples were whether novel alternative HFPO-TA was present in freshwater extracted using a solid phase extraction (SPE) cartridge 32 and wild freshwater fish of Xiaoqing River, and (2) if so, to (Phenomenex strata X-AW, 200 mg/6 mL), whereas fish determine the tissue distribution and bioaccumulation potential blood, fish liver, and human serum were extracted using an ion- 33 in wild fish and (3) evaluate human exposure to HFPO-TA as pair extraction method. An alkaline digestion method was 32 well as other legacy PFASs in local residents. used for fish muscle samples. Additional cleanup using the SPE method was applied to fish liver and muscle samples. ■ MATERIALS AND METHODS Instrument Analysis. Target PFASs (structures shown in Figure S1), including PFCAs (C4−C14), PFSAs (C4, C6, C8), Sample Collection. Xiaoqing River is located in Shandong and Cl-PFESAs (4:2. 6:2, 8:2), were quantified using an Province, China, with a length of approximately 233 km and a Acquity UPLC coupled to a Xevo TQ-S triple quadrupole mass catchment area of 13000 km2. Parallel to the Yellow River, spectrometer (Waters, Milford, MA, USA). Because of the poor Xiaoqing River flows through four industrialized cities (Jinan, sensitivity (limit of quantification (LOQ): 5−20 ng/mL) of Binzhou, Zibo, and Dongying) before finally entering Laizhou HFPO-TA and HFPO-DA with the Xevo TQ-S, they were Bay of the Bohai Sea. From November 29 to December 1 of quantified using an API 5500 triple-quadrupole mass 2015, a total of 18 water samples were collected upstream (sites spectrometer (AB SCIEX, Framingham, MA, USA), which S1−S6), from the tributary receiving fluoropolymer plant showed much better quantification limits (0.05−0.1 ng/mL). effluent (sites S7−S10), and downstream (sites S11−18) of Multiple reaction monitoring (MRM) in ESI− mode was used Xiaoqing River (Figure 1 and Table S2). Approximately 1 L of in both mass spectrometers. Chromatographic separation was water from a depth of 1 m was collected in methanol-rinsed accomplished using an Acquity BEH C18 column (100 mm × polypropylene bottles and stored at −20 °C until analysis. 2.1 mm, 1.7 μm, Waters, MA, USA) with mobile phases of 2

9554 DOI: 10.1021/acs.est.7b02259 Environ. Sci. Technol. 2017, 51, 9553−9560 Environmental Science & Technology Article

Figure 2. Accurate mass measurement, LC retention time, and MS2 fragmentation patterns of HFPO-TA in standard (A), water (B), and fish blood samples (C). mM ammonium acetate in water (A) and methanol (B) at a Framingham, MA, USA) in ESI− mode. The instrument was flow rate of 0.3 mL/min. operated in full scan MS (100−1000 m/z) and MS/MS mode Quality Assurance and Quality Control. Extraction (50−1000 m/z) simultaneously through information-depend- blanks, method detection limits (MDL), quality control ent acquisition (IDA). The detailed parameters are provided in samples, and matrix recovery tests were conducted to ensure the SI. The molecular ion and fragment ion in water (m/z = accurate quantification of PFASs. All labware, solvents, and 495.9507, Δm = −2.627 ppm, and m/z = 184.9824, Δm = sampling equipment were prescreened to reduce possible −4.325 ppm) and fish blood (m/z = 494.9509, Δm = −2.222 contamination. In daily operation, two extraction blanks were ppm, and m/z = 184.9827, Δm = −2.703 ppm) suggested the included in every batch. No detectable contamination was presence of HFPO-TA in corresponding matrices. These found for most PFASs, except for consistent low levels of PFBA observations were further confirmed by the identical retention and HFPO-DA (from the SPE cartridge). Therefore, the levels time with that in the HFPO-TA standard (Figure 2). for these compounds were reported on a blank corrected basis, Data Analysis. Descriptive statistics are provided for PFAS and the MDLs were defined as the average plus three times the concentrations in water and biota samples. When the standard deviation of extraction blanks (shown in Table S5). concentrations of the PFASs were below the MDL, a value of Two QC samples (SRM1957, nonfortified human blood serum, MDL/2 was employed. Riverine mass discharge (t/yr) of National Institute of Standards and Technology, USA) were PFASs from Xiaoqing River was calculated by multiplying the used in every ten human serum samples, and the measured measured concentration (ng/L) with the annual water flux − mean levels of PFHpA (0.270 ± 0.024), PFOA (4.963 ± (m3/yr) and multiplying by 10 12 to harmonize with the units. 0.369), PFNA (0.843 ± 0.040), and PFHxS (3.854 ± 0.279) The measured PFAS concentration was derived from the were within the reported range (Table S6). Matrix recoveries (n average levels in water samples close to the river mouth (S15− = 4) were validated by spiking 2 ng of standard into a blank S18), whereas annual water flux was acquired from the matrix and subjected to the extraction method discussed above hydrological station adjacent to site S16 with a value of 6.5 × with values within 93−109% in water, 77−109% in serum, 72− 108 m3/yr.34 The bioconcentration factor (BCF) was calculated 124% in liver, and 80−125% in muscle (Table S7). The 1/x as the measured PFAS concentrations in fish blood and tissue weighted calibration curve was verified daily and exhibited (on a wet weight basis) divided by those in corresponding excellent linearity (R2 > 0.99). water samples (mean levels of S12 and S13). Tissue/blood The confirmation of the occurrence of HFPO-TA in the ratios were calculated to describe the distribution pattern of samples (i.e., some of the water and biota samples) were HFPO-TA in common carp. One-way analysis of variance conducted using a X500R Q-TOF System (AB SCIEX, (ANOVA) followed by Duncan’s multiple range tests were

9555 DOI: 10.1021/acs.est.7b02259 Environ. Sci. Technol. 2017, 51, 9553−9560 Environmental Science & Technology Article

Figure 3. PFAS concentrations (ng/L) in water samples along Xiaoqing River. used to test for differences in the BCF of PFASs. All statistical and HFPO-DA at sampling sites S1−S6 implied other point analyses were performed using IBM PASW statistics 18.0 sources upstream. The levels of ΣPFASs downstream of (SPSS Inc., USA) with a statistical significance threshold of p < Xiaoqing River remained relatively stable at 31600−35200 0.05. ng/L, which was possibly attributed to other tributaries. On the basis of an annual river water flux of 6.5 × 108 m3/yr,34 the ■ RESULTS AND DISCUSSION riverine discharge of ΣPFASs was estimated to be 21.3 t/yr Concentrations in Xiaoqing River. The concentrations (15.5 t/yr of PFOA, 4.6 t/yr of HFPO-TA; Table S9). and spatial distributions of PFASs in Xiaoqing River are Although instantaneous concentrations might result in a biased presented in Figure 3 and Table S8. Alternatives of PFASs, estimate, they can provide an approximation of HFPO-TA mass fl including HFPO-TA, HFPO-DA, and 6:2 Cl-PFESA, and 12 ux in Xiaoqing River. Sea waters were not collected in the current study; however, it is plausible that HFPO-TA could be legacy PFASs were all detected in the water samples. Results 19,30 showed that PFOA was the predominant compound, detected in Laizhou Bay. On the basis of earlier studies, accounting for 60 ± 18% of all PFASs, followed by HFPO- PFAS concentrations in Laizhou Bay were 3−10-times more TA (24 ± 12%), PFBA, PFHxA, PFPeA, PFHpA, and HFPO- dilute than that in the river mouth. If that is the case, the level DA (1.3 ± 1.0%). Along the main stream of Xiaoqing River, the of HFPO-TA could be approximately 3000−10000 ng/L, still ΣPFAS concentration increased by 3 orders of magnitude, from 1−2 orders of magnitude higher than that before the point 48.4 (S2) to 81900 ng/L (S11), and then decreased 2.5-fold to source input. Such high levels might be harmful to aquatic life 32800 ng/L (S18) before entering Laizhou Bay. The sharp in Laizhou Bay; however, there are no aquatic toxicity data increases in PFAS levels were attributed to the Dongzhulong available on this novel HFPO-TA compound. tributary, where a peak level of 282000 ng/L was observed at Tissue Distribution. The levels of total and individual sampling site S8, which was approximately 800-fold higher than PFASs in fish tissue are shown in Figure 4A and Table S10. All that upstream (355 ng/L, S7). This contamination was likely PFASs were detected in most blood and liver samples (>94%), caused by the fluoropolymer production plant located between but lower detection rates (0−47%) were found for C4−C6 sites S7 and S8, which has also been identified as a point source 19,29,30 PFCAs, PFBS, PFHxS, and 4:2 Cl-PFESA in muscle. The blood of PFASs in previous studies; PFOA was found to be the samples contained the greatest concentration of ΣPFASs (mean major compound in previous and current investigations. The value: 4350 ng/mL), followed by the liver samples (1200 ng/g peak level of PFOA in the present investigation (197000 ng/L) ww) and then the muscle samples (225 ng/g ww). The was lower than the values collected in the same location (e.g., fi ff 30 19 composition pro les of PFASs in di erent tissues are shown in 396000 and 724000 ng/L in April 2014), possibly due to Figure 4B. Similar to water samples, the concentrations of fluctuations in emissions and hydrological conditions over time. PFOA and HFPO-TA were at least one to 2 orders of Our results also showed that novel alternative HFPO-TA magnitude higher than that of other PFASs. PFOA was the ranked second highest after PFOA with a maximum level of predominant component in blood (median: 2190 ng/mL, 68500 ng/L at site S8, whereas HFPO-DA was observed with a ± Σ peak level of 2100 ng/L, comparable to that reported by accounting for 56 15% of PFASs), whereas HFPO-TA was 19 dominant in the liver (587 ng/mL, 47 ± 17%) and muscle Heydebreck et al. at the same location (3800 ng/L). The ± spatial distributions of HFPO-TA and HFPO-DA were highly samples (118 ng/mL, 51 16%). The ratios of HFPO-TA associated with the industrial point source (e.g., fluoropolymer between tissue and blood were calculated to further clarify its manufacturer). In contrast, no observable spatial trends for distribution and were then compared with other PFCAs with PFSAs, PFESAs, or C9−C14 PFCAs were observed. similar molecular chain lengths (e.g., PFOA and PFNA; Figure Our results revealed that the studied fluoropolymer plant 4C). Tissue/blood ratios of HFPO-TA in the liver samples (45 impacted the Dongzhulong tributary and consequently the ± 31%) were 5-fold greater than those in the muscle samples (9 majority of the Xiaoqing Basin. Other fluoropolymer facilities ± 7%). Compared with PFOA and PFNA, HFPO-TA had with smaller production capacities might also exacerbate the higher tissue/blood ratios but only reached statistical PFAS pollution.29,30 For example, the occurrence of HFPO-TA significance in the muscles.

9556 DOI: 10.1021/acs.est.7b02259 Environ. Sci. Technol. 2017, 51, 9553−9560 Environmental Science & Technology Article

5). Log BCF increased signiﬁcantly with increasing molecular chain length in each category of PFAS, which was in good

Figure 5. Log BCFblood of PFASs with increasing molecular chain length. Different letters indicate statistically significant differences in BCFs by Duncan’s multiple range test at p < 0.05.

agreement with previous studies focusing on PFCAs and PFSAs.37,38,43 For the first time, increasing trends were also Figure 4. (A) Concentrations of ΣPFASs, (B) composition profiles, observed in PFECAs and PFESAs. The log BCFs for PFESAs and (C) tissue:blood ratios in common carp. Variables with different were higher than those for PFSAs with the same number of letters indicate statistically significant differences by Duncan’s multiple carbons in the backbones (i.e., 4:2 Cl-PFESA > PFHxS; 6:2 Cl- range test at p < 0.05. PFESA > PFOS), suggesting that the inserted ester oxygen and/or the chlorine atom increased the bioaccumulation As expected, the levels of PFASs in common carp captured potential of PFASs. No clear pattern was found between downstream of Xiaoqing River near the emission source were PFECAs and PFCAs having the same number of carbons; tens to hundreds of times higher than those of other fish HFPO-DA (C6) had higher BCF than PFHxA (C6), whereas species from different regions.35,36 However, the PFAS HFPO-TA (C9) had a lower BCF than that of PFNA (C9). distribution and tissue/blood ratios were generally consistent The reason might be the branched carbons in HFPO-TA − with other studies.35 38 The observed concentrations in tissues (Figure S1) that lead to less hydrophobicity and complicate the in descending order (blood > liver > muscle) were in good comparison. The log BCFs for HFPO-DA, PFBA, PFPeA, and agreement with previous research,15,35,37,38 suggesting that all PFHxA were all relatively low (<1), suggesting lower PFASs, including HFPO-TA, share similar mechanisms of bioaccumulation potential for these compounds. In general, distribution. However, HFPO-TA tended to accumulate more log Kow and log BCF are used to predict bioaccumulation 37 in liver and muscle compared with those of PFOA and PFNA potential. The higher estimated log Kow (5.555) of HFPO-TA (Figure 4B and C). This discrepancy might be due to the by EPI Suite V4.11 suggested it was more bioaccumulative than ff ffi di erences in protein binding a nity and/or hydrophobic PFOA (log Kow = 4.814) and PFNA (log Kow = 5.483, Table ± properties. Because liver and muscle are rich in proteins and S11). However, the log BCFblood value for HFPO-TA (2.18 phospholipids, greater binding affinity or hydrophobicity may 0.44) fell between those for PFOA (1.93 ± 0.34) and PFNA lead to additional sorption,39,40 consequently leading to a (3.01 ± 0.37), suggesting that HFPO-TA was more higher distribution in liver and muscle. This hypothesis was bioaccumulative than PFOA but less bioaccumulative than supported by our recent findings that HFPO-TA was more PFNA. This deviation might be due to the high concentrations strongly bound than PFOA to human liver fatty acid binding of HFPO-TA in Xiaoqing River because the absorption from protein (hL-FABP), one of the most abundant proteins in the water into biota might be partly saturated at this high 41 ± 44−46 liver. The dissociation constant of HFPO-TA (Kd = 4.36 concentration. 1.17) was found to be much lower than that of PFOA (Kd = Human Exposure. Human exposure to PFASs was 8.03 ± 2.10), indicating a much stronger binding affinity of evaluated in 48 Huantai residents with an average age of 53 HFPO-TA to hL-FABP than to PFOA.41 Additionally, although years (Table 1). Detectable levels of HFPO-TA, C7−C13 the lone pair electrons of the O atom at the insertion of ester PFCAs, PFHxS, PFOS, and 4:2, 6:2, and 8:2 Cl-PFESAs were bonds in HFPO-TA might have decreased the hydrophobicity, measured in most serum samples (>97.9%). PFBA and the larger molecular size consequently increased its hydro- PFTeDA were detected in 87.5 and 62.5% of serum samples, phobicity42 compared with similar molecular structures of whereas the detection rates for HFPO-DA, PFPeA, PFHxA, PFOA and PFNA (Table S11, EPI Suite V4.11). and PFBS ranged from 16.7 to 39.6%. Results also showed that Bioaccumulation. The tissue-specific bioconcentration PFOA was dominant and accounted for 86 ± 9% of total factors (BCFs) for common carp are listed in Table S12. The PFASs. We previously reported a median PFOA level of 284.34 log BCFs for all PFASs ranged from 0.49 to 5.93 in the fish ng/mL in residents from Changshu, another important blood samples, approximately 0.4 and 1.0 log units higher than fluorochemical industrial zone in China.47 The median level those in the liver and muscle samples, respectively. Because all of PFOA (126 ng/mL) here was approximately 50% lower than target PFASs were frequently detected in blood but not in liver that detected in our previous study47 but was 5-times higher fl and muscle, log BCFblood was used to better re ect the than the reported levels by C8 Health Project (median: 24 ng/ differences in bioaccumulation potential among PFASs (Figure mL), which focused on the residents living near the DuPont

9557 DOI: 10.1021/acs.est.7b02259 Environ. Sci. Technol. 2017, 51, 9553−9560 Environmental Science & Technology Article

Table 1. Serum PFAS Levels in Local Residents from PFCAs in home-produced chicken eggs declined with a Huantai (n = 48) increasing distance from the same fluoropolymer plant in this study.55 Location information for the residents was not detection geometric 5th 95th rate (%) mean median percentile percentile acquired here, which hampers further exploration of the relationship between serum PFAS levels and residence distance. HFPO-DA 37.5 0.13 n.d. n.d. 1.72 Future studies are needed to evaluate this relationship. HFPO-TA 97.9 2.41 2.93 0.18 53.4 Environmental Implications. Earlier research has shown PFBA 87.5 0.29 0.35 n.d. 3.05 that large proportions of extractable organic fluorine in biota PFPeA 16.7 0.03 n.d. n.d. 0.16 and humans cannot be explained by known PFASs.56,57 Thus, PFHxA 16.7 0.03 n.d. n.d. 0.17 the identification of unknown fractions is of great importance PFHpA 100 0.25 0.25 0.05 1.75 and will improve our understanding of the current situation PFOA 100 134 126 22.0 638 regarding the manufacture, usage, and release of PFASs. In the PFNA 100 1.24 1.31 0.48 3.46 current study, relatively high levels of HFPO-TA were PFDA 100 0.96 1.01 0.26 3.87 measured in the surface water and fish samples downstream PFUnDA 100 0.53 0.60 0.19 1.30 from a fluoropolymer production plant, accounting for 24−51% PFDoDA 100 0.06 0.06 0.02 0.20 of total PFASs. The estimated annual riverine discharge of PFTriDA 100 0.07 0.08 0.03 0.17 HFPO-TA (4.6 t/yr) was approximately 30% of that for PFOA, PFTeDA 62.5 0.01 0.01 n.d. 0.03 indicated an emerging, significant amount of HFPO-TA being PFBS 39.6 0.01 0.01 n.d. 0.04 fl PFHxS 100 0.46 0.51 0.10 1.29 used in uoropolymer manufacturing. With rapidly increasing PFOS 100 5.79 6.54 1.95 13.7 demands in China and more stringent regulations for PFOA 4:2 Cl- 97.9 0.04 0.04 0.01 0.10 use, it is reasonable to believe that the production and usage of PFESA HFPO-TA as an alternative will continue to increase. We 6:2 Cl- 100 4.04 4.19 1.49 9.86 evaluated the bioaccumulation potential for HFPO-TA in PFESA common carp. Although HFPO-TA (BCFblood = 204 L/kg) was 8:2 Cl- 100 0.06 0.06 0.02 0.19 not bioaccumulative according to the range of promulgated PFESA “ ” − 43 Σ bioaccumulation B 1000 5000 L/kg, it could be regarded PFASs 158 147 29.3 725 as having a “tendency to accumulate in organisms” based on the a n.d., not detected. regulatory criteria of 1−1000 L/kg.58 In addition, the BCF of HFPO-TA was significantly higher than that of legacy PFOA, Washington Works plant in West Virginia, U.S.48,49 Addition- suggesting greater bioaccumulation potential in aquatic ally, the level in this study was still 40−100-times higher than organisms. Thus, more attention should be paid to its aquatic that recorded in other populations from China (1.39 ng/mL),50 toxicity and ecological risk, especially in regions suspected of Canada (2.17 ng/mL),51 and the US (3.07 ng/mL).52 Such being polluted, such as Laizhou Bay. The presence of HFPO- elevated PFOA levels suggest strong PFAS exposure from the TA in the sera of local residents also raises concerns about the nearby fluoropolymer industrial plant. potential health risks related to exposure. The replacement of fl Results also showed that HFPO-TA was the fourth highest in PFOA with HFPO-TA or other poly uorinated chemicals need median level (2.93 ng/mL) next to PFOA, PFOS, and 8:2 Cl- to be treated cautiously until further investigations regarding its PFESA (Table 1). The skewness of the HFPO-TA distribution metabolism, toxicity, and health risk are fully explored. was the greatest among all PFASs; 80% of subjects had HFPO- TA levels between nondetectable and 9.23 (mean: 2.63 ng/ ■ ASSOCIATED CONTENT − mL), whereas 20% ranged within 12.0 55.0 (mean: 36.8 ng/ *S Supporting Information mL). The high variability in the HFPO-TA levels suggests that ff The Supporting Information is available free of charge on the certain factors are a ecting the extent of exposure in the study ACS Publications website at DOI: 10.1021/acs.est.7b02259. subjects. No subject reported an employment history related to fluoropolymer production, and no age or gender differences in Additional information included standards and reagents, residents were observed in HFPO-TA levels (data not shown). synthesis of HFPO-TA, PFASs analysis, and qualitative Fish consumption frequency might be an important predictor analysis of HFPO-TA and other materials (PDF) because relatively high levels of HFPO-TA (median: 118 ng/g ww; Table S10) were observed in the muscle of common carp, ■ AUTHOR INFORMATION and more frequent consumption of contaminated fish from Corresponding Author Xiaoqing River might result in higher HFPO-TA exposure. * According to Shandong Statistic Year Book, the average fish Phone: +86-10-64807185; e-mail: [email protected]. consumption in the studied area was 17.0 g/day.53 Applying an ORCID average body weight of 60 kg for adults, the daily intake of Jiayin Dai: 0000-0003-4908-5597 −1 −1 HFPO-TA was estimated to be 33.4 ng kg day based on the Notes equation: daily intake (ng kg−1 day−1) = HFPO-TA level in The authors declare no competing financial interest. muscle (ng/g) × fish consumption (g/day)/body weight (kg). Another important factor might be residence distance from the fluoropolymer facility. A previous study has shown that PFASs ■ ACKNOWLEDGMENTS generated from a point source can lead to PFAS exposure in This work was supported by the National Natural Science humans via dust ingestion and dermal absorption with the Foundation of China (21737004 and 31320103915) and the estimated daily intake for residents inversely associated with Strategic Priority Research Program of the Chinese Academy of distance.54 Another piece of evidence is that the levels of Sciences (XDB14040202).

9558 DOI: 10.1021/acs.est.7b02259 Environ. Sci. Technol. 2017, 51, 9553−9560 Environmental Science & Technology Article ■ REFERENCES (19) Heydebreck, F.; Tang, J. H.; Xie, Z. Y.; Ebinghaus, R. Alternative and legacy perfluoroalkyl substances: Differences between European (1) Buck, R. C.; Franklin, J.; Berger, U.; Conder, J. M.; Cousins, I. T.; and Chinese river/estuary systems. Environ. Sci. Technol. 2015, 49 de Voogt, P.; Jensen, A. A.; Kannan, K.; Mabury, S. A.; van Leeuwen, (14), 8386−8395. S. P. Perfluoroalkyl and polyfluoroalkyl substances in the environment: (20) Sun, M.; Arevalo, E.; Strynar, M.; Lindstrom, A.; Richardson, Terminology, classification, and origins. Integr. Environ. Assess. Manage. M.; Kearns, B.; Pickett, A.; Smith, C.; Knappe, D. R. U. Legacy and 2011, 7 (4), 513−41. emerging perfluoroalkyl substances are important drinking water (2) Kissa, E. Fluorinated surfactants and repellents, 2nd ed.; Marcel contaminants in the Cape Fear River watershed of North Carolina. Dekker Inc.: New York, NY, 2001. Environ. Sci. Technol. Lett. 2016, 3 (12), 415−419. (3) Lindstrom, A. B.; Strynar, M. J.; Libelo, E. L. Polyfluorinated (21) Millauer, H.; Schwertfeger, W.; Siegemund, G. Hexafluoropro- compounds: Past, present, and future. Environ. Sci. Technol. 2011, 45 pene oxide - a key compound in organofluorine chemistry. Angew. (19), 7954−7961. Chem., Int. Ed. Engl. 1985, 24 (3), 161−179. (4) Wang, Z. Y.; Cousins, I. T.; Scheringer, M.; Hungerbuehler, K. (22) Han, G. Synthesis and application of hexafluoropropene oxide Hazard assessment of fluorinated alternatives to long-chain perfluor- polymer. Zhejiang Chemical Industry (in Chinese) 2010, 41 (11), 1−4. oalkyl acids (PFAAs) and their precursors: Status quo, ongoing (23) Zhang, W.; Zhang, Y. M. Synthesis and surfactivity of non-ionic challenges and possible solutions. Environ. Int. 2015, 75, 172−179. fluorocarbon surfactant. Modern Chemical Industry (in Chinese) 2011, (5) U.S. Environmental Protection Agency. 2010/2015 PFOA 31 (4), 49−52. stewardship program; 2006. (24) Zhou, X. Synthesis and application of the HFPO trimer-based (6) Stockholm Convention. Governments unite to step-up reduction polymer oil repellent. Chemical Production & Technology (in Chinese) on global DDT reliance and add nine new chemicals under − international treaty; 2009. 2007, 14 (4), 5 8. (7) Committee for Risk Assessment. Background document to the (25) Kysilka, O.; Rybackova, M.; Skalicky, M.; Kvicalova, M.; Cvacka, J.; Kvicala, J. Fluorous imidazolium room-temperature ionic liquids opinion on the Annex XV dossier proposing restrictions on − perfluorooctanoic acid (PFOA), PFOA salts and PFOA-related based on HFPO trimer. J. Fluorine Chem. 2009, 130 (7), 629 639. substances; 2015. (26) Kysilka, O.; Rybackova, M.; Skalicky, M.; Kvicalova, M.; Cvacka, (8) Ritter, S. K. Fluorochemicals go short. Chem. Eng. News 2010, 88 J.; Kvicala, J. HFPO trimer-based alkyl triflate, a novel building block (5), 12−17. for fluorous chemistry. Preparation, reactions and F-19 gcosy analysis. Collect. Czech. Chem. Commun. 2008, 73 (12), 1799−1813. (9) United Nations Environment Programme. Technical paper on ́ the identification and assessment of alternatives to the use of (27) Meissner, E.; Wroblewska, A. Oligomerization of hexafluor- fl opropylene oxide in the presence of alkali metal halides. Pol. J. Chem. per uorooctane sulfonic acid in open applications; 2012. − (10) Strynar, M.; Dagnino, S.; McMahen, R.; Liang, S.; Lindstrom, Technol. 2007, 9 (3), 95 97. A.; Andersen, E.; McMillan, L.; Thurman, M.; Ferrer, I.; Ball, C. (28) Bazhin, D. N.; Gorbunova, T. I.; Zapevalov, A. Y.; Beketov, I. V.; Identification of novel perfluoroalkyl ether carboxylic acids (PFECAS) Saloutin, V. I. A study of the physico-chemical features of the and sulfonic acids (PFESAs) in natural waters using accurate mass [(perfluoroalkyl)methyl]oxirane amino derivatives based on the time-of-flight mass spectrometry (TOFMS). Environ. Sci. Technol. hexafluoropropylene oxide trimer. Russ. J. Gen. Chem. 2011, 81 (9), − 2015, 49 (19), 11622−11630. 1829 1833. (11) Wang, S. W.; Huang, J.; Yang, Y.; Hui, Y. M.; Ge, Y. X.; Larssen, (29) Wang, P.; Lu, Y. L.; Wang, T. Y.; Fu, Y. N.; Zhu, Z. Y.; Liu, S. J.; T.; Yu, G.; Deng, S. B.; Wang, B.; Harman, C. First report of a chinese Xie, S. W.; Xiao, Y.; Giesy, J. P. Occurrence and transport of 17 PFOS alternative overlooked for 30 years: Its toxicity, persistence, and perfluoroalkyl acids in 12 coastal rivers in south Bohai coastal region of presence in the environment. Environ. Sci. Technol. 2013, 47 (18), China with concentrated fluoropolymer facilities. Environ. Pollut. 2014, − 10163−10170. 190, 115 122. (12) Dupont. Dupont GenX processing aid for making fluoropol- (30) Shi, Y. L.; Vestergren, R.; Xu, L.; Song, X. W.; Niu, X. M.; ymer resins; 2010. Zhang, C. H.; Cai, Y. Q. Characterizing direct emissions of (13) European Chemicals Agency. Registered substances; 2013. perfluoroalkyl substances from ongoing fluoropolymer production (14) Ruan, T.; Lin, Y. F.; Wang, T.; Liu, R. Z.; Jiang, G. B. sources: A spatial trend study of Xiaoqing river, China. Environ. Pollut. Identification of novel polyfluorinated ether sulfonates as PFOS 2015, 206, 104−112. alternatives in municipal sewage sludge in China. Environ. Sci. Technol. (31) Dongyue Group Limited. 2012 Annual Report; 2012. 2015, 49 (11), 6519−6527. (32) Taniyasu, S.; Kannan, K.; So, M. K.; Gulkowska, A.; Sinclair, E.; (15) Shi, Y. L.; Vestergren, R.; Zhou, Z.; Song, X. W.; Xu, L.; Liang, Okazawa, T.; Yamashita, N. Analysis of fluorotelomer alcohols, Y.; Cai, Y. Q. Tissue distribution and whole body burden of the fluorotelorner acids, and short- and long-chain perfluorinated acids chlorinated polyfluoroalkyl ether sulfonic acid F-53B in crucian carp in water and biota. J. Chromatogr. A 2005, 1093 (1−2), 89−97. (Carassius carassius): Evidence for a highly bioaccumulative con- (33) Hansen, K. J.; Clemen, L. A.; Ellefson, M. E.; Johnson, H. O. taminant of emerging concern. Environ. Sci. Technol. 2015, 49 (24), Compound-specific, quantitative characterization of organic: Fluo- 14156−14165. rochemicals in biological matrices. Environ. Sci. Technol. 2001, 35 (4), (16) Shi, Y. L.; Vestergren, R.; Xu, L.; Zhou, Z.; Li, C. X.; Liang, Y.; 766−770. Cai, Y. Q. Human exposure and elimination kinetics of chlorinated (34) Ministry of Water Resources of the People’s Republic of China. polyfluoroalkyl ether sulfonic acids (Cl-PFESAs). Environ. Sci. Technol. Hydrological data of Huaihe River Basin, Annual Hydrological Report; 2016, 50 (5), 2396−2404. China Water & Power Press, 2014. (17) Pan, Y. T.; Zhu, Y. S.; Zheng, T. Z.; Cui, Q. Q.; Buka, S. L.; (35) Murakami, M.; Adachi, N.; Saha, M.; Morita, C.; Takada, H. Zhang, B.; Guo, Y.; Xia, W.; Yeung, L. W. Y.; Li, Y. R.; Zhou, A. F.; Levels, temporal trends, and tissue distribution of perfluorinated Qiu, L.; Liu, H. X.; Jiang, M. M.; Wu, C. S.; Xu, S. Q.; Dai, J. Y. Novel surfactants in freshwater fish from Asian countries. Arch. Environ. chlorinated polyfluorinated ether sulfonates and legacy per-/ Contam. Toxicol. 2011, 61 (4), 631−641. polyfluoroalkyl substances: Placental transfer and relationship with (36) Lam, N. H.; Cho, C. R.; Kannan, K.; Cho, H. S. A nationwide serum albumin and glomerular filtration rate. Environ. Sci. Technol. survey of perfluorinated alkyl substances in waters, sediment and biota 2017, 51 (1), 634−644. collected from aquatic environment in Vietnam: Distributions and (18) Liu, Y. W.; Ruan, T.; Lin, Y. F.; Liu, A. F.; Yu, M.; Liu, R. Z.; bioconcentration profiles. J. Hazard. Mater. 2017, 323, 116−127. Meng, M.; Wang, Y. W.; Liu, J. Y.; Jiang, G. B. Chlorinated (37) Martin, J. W.; Mabury, S. A.; Solomon, K. R.; Muir, D. C. G. polyfluoroalkyl ether sulfonic acids in marine organisms from Bohai Bioconcentration and tissue distribution of perfluorinated acids in Sea, China: Occurrence, temporal variations, and trophic transfer rainbow trout (oncorhynchus mykiss). Environ. Toxicol. Chem. 2003, behavior. Environ. Sci. Technol. 2017, 51 (8), 4407−4414. 22 (1), 196−204.

9559 DOI: 10.1021/acs.est.7b02259 Environ. Sci. Technol. 2017, 51, 9553−9560 Environmental Science & Technology Article

(38) Labadie, P.; Chevreuil, M. Partitioning behaviour of perfluorooctanoic acid (PFOA) around a fluorochemical industrial perfluorinated alkyl contaminants between water, sediment and fish park in China. Environ. Int. 2017, 101,1−6. in the Urge River (nearby Paris, France). Environ. Pollut. 2011, 159 (56) Yeung, L. W. Y.; Miyake, Y.; Taniyasu, S.; Wang, Y.; Yu, H. X.; (2), 391−397. So, M. K.; Jiang, G. B.; Wu, Y. N.; Li, J. G.; Giesy, J. P.; Yamashita, N.; (39) Zhang, L. Y.; Ren, X. M.; Guo, L. H. Structure-based Lam, P. K. S. Perfluorinated compounds and total and extractable investigation on the interaction of perfluorinated compounds with organic fluorine in human blood samples from China. Environ. Sci. human liver fatty acid binding protein. Environ. Sci. Technol. 2013, 47 Technol. 2008, 42 (21), 8140−8145. (19), 11293−11301. (57) Yeung, L. W. Y.; Miyake, Y.; Wang, Y.; Taniyasu, S.; Yamashita, (40) Xie, W.; Ludewig, G.; Wang, K.; Lehmler, H. J. Model and cell N.; Lam, P. K. S. Total fluorine, extractable organic fluorine, membrane partitioning of perfluorooctanesulfonate is independent of perfluorooctane sulfonate and other related fluorochemicals in liver the lipid chain length. Colloids Surf., B 2010, 76 (1), 128−136. of Indo-Pacific humpback dolphins (Sousa chinensis) and finless (41) Sheng, N.; Cui, R. N.; Wang, J. H.; Guo, Y.; Wang, J. S.; Dai, J. porpoises (Neophocaena phocaenoides) from South China. Environ. − Y. Cytotoxicity of novel ﬂuorinated alternatives to long-chain Pollut. 2009, 157 (1), 17 23. perﬂuoroalkyl substances to human liver cell line and their binding (58) Gao, Y. P.; Ji, Y. M.; Li, G. Y.; Mai, B. X.; An, T. C. capacity to human liver fatty acid binding protein. Arch. Toxicol. 2017 Bioaccumulation and ecotoxicity increase during indirect photo- chemical transformation of polycyclic musk tonalide: A modeling (under minor revision).90217 − (42) Gomis, M. I.; Wang, Z. Y.; Scheringer, M.; Cousins, I. T. A study. Water Res. 2016, 105,47 55. modeling assessment of the physicochemical properties and environmental fate of emerging and novel per- and polyfluoroalkyl substances. Sci. Total Environ. 2015, 505, 981−991. (43) Conder, J. M.; Hoke, R. A.; De Wolf, W.; Russell, M. H.; Buck, R. C. Are PFCAs bioaccumulative? A critical review and comparison with regulatory lipophilic compounds. Environ. Sci. Technol. 2008, 42 (4), 995−1003. (44) Kudo, N.; Katakura, M.; Sato, Y.; Kawashima, Y. Sex hormone- regulated renal transport of perfluorooctanoic acid. Chem.-Biol. Interact. 2002, 139 (3), 301−316. (45) Kudo, N.; Sakai, A.; Mitsumoto, A.; Hibino, Y.; Tsuda, T.; Kawashima, Y. Tissue distribution and hepatic subcellular distribution of perfluorooctanoic acid at low dose are different from those at high dose in rats. Biol. Pharm. Bull. 2007, 30 (8), 1535−1540. (46) Liu, C. H.; Gin, K. Y. H.; Chang, V. W. C.; Goh, B. P. L.; Reinhard, M. Novel perspectives on the bioaccumulation of PFCs - the concentration dependency. Environ. Sci. Technol. 2011, 45 (22), 9758− 9764. (47) Wang, J. S.; Zhang, Y. T.; Zhang, W.; Jin, Y. H.; Dai, J. Y. Association of perfluorooctanoic acid with hdl cholesterol and circulating miR-26b and miR-199−3p in workers of a fluorochemical plant and nearby residents. Environ. Sci. Technol. 2012, 46 (17), 9274− 9281. (48) Barry, V.; Winquist, A.; Steenland, K. Perfluorooctanoic acid (PFOA) exposures and incident cancers among adults living near a chemical plant. Environ. Health Perspect. 2013, 121 (11−12), 1313− 1318. (49) Steenland, K.; Zhao, L. P.; 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. (50) Pan, Y. Y.; Shi, Y. L.; Wang, J. M.; Cai, Y. Q.; Wu, Y. N. Concentrations of perfluorinated compounds in human blood from twelve cities in China. Environ. Toxicol. Chem. 2010, 29 (12), 2695− 2701. (51) Haines, D. A.; Murray, J. Human biomonitoring of environmental chemicals-early results of the 2007−2009 Canadian Health Measures Survey for males and females. Int. J. Hyg. Environ. Health 2012, 215 (2), 133−137. (52) Centers for Disease Control and Prevention. Fourth national report on human exposure to environmental chemicals, February 2015. (53) Shandong Statistical Bureau. Shandong statistical yearbook; 2015. (54) Su, H. Q.; Lu, Y. L.; Wang, P.; Shi, Y. J.; Li, Q. F.; Zhou, Y. Q.; 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. (55) Su, H. Q.; Shi, Y. J.; Lu, Y. L.; Wang, P.; Zhang, M.; Sweetman, A.; Jones, K.; Johnson, A. Home produced eggs: An important pathway of human exposure to perfluorobutanoic acid (PFBA) and

9560 DOI: 10.1021/acs.est.7b02259 Environ. Sci. Technol. 2017, 51, 9553−9560