Science of the Total Environment 472 (2014) 912–922

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Distribution and fate of perfluoroalkyl substances in Mediterranean Spanish sewage treatment plants Julian Campo a,⁎, Ana Masiá a, Yolanda Picó a, Marinella Farré b, Damià Barceló b,c a

Food and Environmental Safety Research Group (SAMA-UV), Faculty of Pharmacy, University of Valencia, Av. Vicent Andrés Estellés s/n. 46100, Burjassot, València, Spain Department of Environmental Chemistry (IDAEA-CSIC), Jordi Girona 18-26, 08034 Barcelona, Spain Catalan Institute of Water Research, ICRA Catalan Institute for Water Research — ICRA, C/Emili Grahit, 101, Edifici H2O, Parc Científic i Tecnològic de la Universitat de Girona, E-17003 Girona, Spain

b c

H I G H L I G H T S • • • • •

Twenty-one PFASs were quantified in wastewater and dehydrated sludge samples. PFAS loads discharged to the Rivers were from 16 g day− 1 (Llobregat) to 67 g day− 1 (Ebro). STPs could be a focal point of PFAS contamination to the Rivers. Removal efficiencies in STPs confirm that PFASs are only partially eliminated. Distribution coefficients (Kd) for sludge cannot be compared to those for sediment.

a r t i c l e

i n f o

Article history: Received 2 September 2013 Received in revised form 11 November 2013 Accepted 11 November 2013 Available online 15 December 2013 Keywords: Sewage treatment plant Wastewater Sludge LC–MS/MS Removal efficiency Distribution coefficients

a b s t r a c t The concentrations of 21 perfluoroalkyl substances (PFASs: C4–C14, C16, C18 carboxylates, C4, C6–C8 and C10 sulfonates and C8 sulfonamide) were determined in influent, effluent and sludge from 16 different sewage treatment plants (STPs) located in the Ebro (6), Guadalquivir (5), Jucar (2) and Llobregat (3) Rivers, in two consecutive years (2010 and 2011). The analytes were extracted by solid phase extraction (SPE) and determined by Liquid Chromatography triple Quadrupole Mass Spectrometer (LC-QqQ-MS). All samples, except two sludges from Guadalquivir River STPs, were contaminated with at least one PFAS. Perfluorobutanoate (PFBA), perfluoropentanoate (PFPeA) and perfluorooctane sulfonate (L-PFOS) were the most frequently detected. The highest concentration in water was determined in 2010 in a Guadalquivir River STP (perfluorohexanoate, PFHxA: 5.60 μg L−1) and, in 2011, in an Ebro River STP (perfluorobutane sulfonate, L-PFBS: 0.31 μg L−1). In sludge samples, the maximum concentration in 2010 was 1.79 μg g−1 dry weight (dw) (L-PFOS, in a Llobregat River STP), and in 2011, 1.88 μg g−1 dw (PFBA, in one Guadalquivir River STP). High PFAS values in sludge could be related to positive removal efficiencies, and can be attributed to their adsorption. Distribution coefficients (Kd) were determined ranging between 0.32 L kg−1 (perfluorohexane sulfonate, L-PFHxS) and 36.6 103 L kg−1 (PFBA). The total PFAS loads discharged into the basins showed high values for the Ebro River STPs (66.9 g day−1) while in the others, the loads were between 3.97 g day−1, in the Jucar STPs, and 32.2 g day−1, in the Guadalquivir STPs. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Chemical properties of perfluoroalkyl substances (PFASs) are governed by their structure that consist of a hydrophobic perfluoroalkyl tail and a hydrophilic polar head (commonly sulfonate or carboxylate). This combination imparts strong water/oil repellency to these compounds and makes them effective at reducing surface tension. PFASs

⁎ Corresponding author. Tel.: +34 963543092; fax: +34 963544954. E-mail address: [email protected] (J. Campo). 0048-9697/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.scitotenv.2013.11.056

and their polyfluorinated precursors have been widely used in several industrial and commercial applications, such as cosmetics, lubricants, fire-fighting foams, stain resistant and moisture repelling coatings as well as in the synthesis of some polymeric materials (Kissa, 2001; Lewandowski et al., 2006; Qiu, 2007). Due to their extremely strong carbon–fluorine bonds, PFASs present significant thermal and chemical stability being persistent in the environment, where they can bio-accumulate and potentially have adverse effects on humans and wildlife (Austin et al., 2003; Joensen et al., 2009; Tomy et al., 2004). Laboratory studies with animals proved chronic and sub-chronic effects as hepatotoxicity, carcinogenicity and hormonal

J. Campo et al. / Science of the Total Environment 472 (2014) 912–922

effects (Lau et al., 2007). Human exposure to PFASs is of concern since these compounds tend to be associated with fatty acid binding proteins in the liver or albumin proteins in blood (Han et al., 2003), and have been detected in human serum (Haug et al., 2009), urine, saliva (Tao, 2009), seminal plasma (Guruge et al., 2005) and breast milk (Sundstrom et al., 2011; Tao, 2009). Perfluorooctane sulfonate (L-PFOS) and its synthetic starting material, perfluorooctyl sulfonyl fluoride (L-POSF), were the first PFASs to be listed as persistent organic pollutants at the Stockholm Convention (United Nations Environment Programme, 2010). The European Union also published a Directive prohibiting the general use of LPFOS after June 2008 (European Parliament and Council, 2006). Despite these attempts to decrease PFAS levels in the environment, a clear declining trend in their pollution has not been observed since there is still a growing demand for substances with their unique properties, and thus other short chain PFASs are replacing L-PFOS (Ahrens et al., 2009; Llorca et al., 2012). Over the last decade, advances in analytical techniques have resulted in studies describing the occurrence of PFASs in water (Pico et al., 2012; Taniyasu et al., 2003; van Leeuwen et al., 2009; Yamashita et al., 2005) including drinking water (Holzer et al., 2008; Llorca et al., 2012; Skutlarek et al., 2006), sediments (Higgins and Luthy, 2006), air (Jahnke et al., 2007), organisms (Giesy and Kannan, 2001; Martin et al., 2003), and even in ice caps (Young et al., 2007). Sewage Treatment Plants (STPs) seem to be ineffective in removing PFASs, and accordingly monitoring data are available worldwide, indicating that municipal sewage is a significant source of these compounds to the aquatic environment, potentially reaching treated water for human consumption. Specifically, various contamination levels have been reported in the influents and effluents of municipal STPs in China (Ma and Shih, 2010), Denmark (Bossi et al., 2008), Germany (Ahrens et al., 2009; Becker et al., 2008), Greece (Arvaniti et al., 2012), Korea (Guo et al., 2010), Singapore (Yu et al., 2009), Switzerland (Huset et al., 2008), and USA (Loganathan et al., 2007; Schultz et al., 2006; Sinclair and Kannan, 2006). However, few of these studies have reported the occurrence of PFASs in sludge samples (Arvaniti et al., 2012; Bossi et al., 2008; Guo et al., 2010; Loganathan et al., 2007; Yu et al., 2009; Sun et al., 2011; Yan et al., 2012; Stasinakis et al., 2013) and, additionally, there are limited data for the removal and the environmental load of these compounds from STPs (Loganathan et al., 2007; Sinclair and Kannan, 2006; Yu et al., 2009; Kunacheva et al., 2011; Arvaniti et al., 2012; Stasinakis et al., 2013). In Spain, there are very few peer-reviewed articles analysing concentrations of PFASs in wastewaters (Llorca et al., 2011, 2012), and even less reporting their removal efficiency. This work presents the results of, to our knowledge, the first extensive monitoring survey that was carried out in 2010 and 2011 in the main STPs from Ebro, Guadalquivir, Jucar and Llobregat Rivers, in Spain. Twenty-one currently used PFASs, belonging to different chemical classes, have been monitored. PFAS concentrations have been analysed in the influent, effluent and dehydrated sludge. With these data, removal efficiencies of such PFASs have been calculated and reported. The final objective of this study is to improve the knowledge about the causes of aquatic environment pollution considering the STPs as point sources of PFASs that can affect human health. 2. Materials and methods 2.1. Description of the study area This study covers 16 STPs that discharge treated wastewater to Ebro (6), Guadalquivir (5), Jucar (2) and Llobregat (3) Rivers. These areas were selected because of their economic and environmental importance (Navarro-Ortega et al., 2012). Ebro River (910 km) is the most important river in Spain (drainage basin of 85,534 km2). Guadalquivir River (657 km) is the main water source of the Andalusian region (more

913

than 7 million inhabitants) with a catchment of 57,527 km2. Jucar River (498 km) serves 103,0979 people with urbanized and industrial uses located mainly in its lowest part (basin 21,632 km2). Llobregat River (156 km) is one of Barcelona's major drinking water resources within a catchment of 4,957 km2. Fig. 1 shows the location of the STPs monitored in the four River Basins. The STPs analysed in the Ebro River were Logroño, Pamplona, Tudela, Zaragoza, Lleida and Tortosa; in the Jucar River, Cuenca and Alzira; in the Llobregat River, Igualada and Manresa (in 2011 Abrera STP was added); and in the Guadalquivir River, Cordoba, Loja, Moron de la Frontera, Ranilla and Copero (these last two treat the wastewaters from Sevilla city, 4th largest city in Spain). Their characteristics are summarized in Table S1. 2.2. Sampling Sampling campaigns were carried out in October of 2010 and of 2011 in each STP (15 in 2010 and 16 in 2011). Integrated samples of influent and effluent were taken using automatic 24-hour volumeproportional composite sampling (the equipment takes a constant sample at variable time intervals after a certain sewage volume has passed the sampling point). In 2010, punctual samples were analysed in Alzira, Loja and Cuenca STPs meanwhile in 2011, only the samples of Cuenca STP were punctual. Sludge samples were provided by the plant operators and they were transferred to aluminium foil. Before the analyses, influents and effluents were vacuum filtered through 1 μm glass fibre filters followed by 0.45 μm nylon membrane filters (VWR, Barcelona, Spain). During the studied periods the influent and effluent flows did not change. 2.3. Chemicals and reagents A total of 21 PFASs were selected, with a variety of uses as well as physicochemical characteristics and toxicity. Most of them are perfluorocarboxylates (PFCAs), but there are also perfluorosulfonates (PFSs) and perfluorosulfonamides (PFSAs). Exhaustive description of selected target PFASs is provided in Supplementary content Table S2. 2.4. Sample preparation and instrumental analysis 2.4.1. Water extraction An off-line solid-phase extraction (SPE) procedure was used for the pre-concentration of water samples (Pico et al., 2012). Very briefly, a mix of isotopically labelled standards at 100 ng mL− 1 of each (62.5 μL) were added to water samples (250 mL) that were vacuum passed through STRATA-X Polymeric Reversed Phase cartridges (Phenomenex SPE cartridge 200 mg sorbent/6 mL cartridge, Torrance, CA, USA), previously preconditioned with 4 mL 0.1% of NH4OH in methanol, 4 mL of methanol and 4 mL of H2O. The cartridges were air-dried and, then, analytes were eluted with 4 mL of 0.10% of NH4OH in methanol drop by drop. Extracts were evaporated to dryness, re-constituted with 250 μL of methanol and analysed. 2.4.2. Sludge extraction Sludge sample (5 g) was weighed in a 50 mL falcon tube spiked with a mix of internal standards (62.5 μL), and homogenized with 10 mL of 1% acetic acid in water (Pico et al., 2012). The mixture was agitated intensively in a vortex for 1 min, ultra-sounded 15 min at 40 °C and centrifuged at 3000 rpm for 2 min. The supernatant was passed to a second falcon tube. Then, 2.50 mL of methanol-acetic acid 1% (90:10 v/v) was added to the first falcon tube and the mixture was again agitated, ultra-sounded and centrifuged as previously. The supernatant was poured into the second falcon tube. This procedure was repeated with 10 mL of 1% acetic acid in water. After, adjust the final extract volume to 250 mL with water, the extraction was completed following the process described in the Section 2.4.1. Moisture content of sludge was

914

J. Campo et al. / Science of the Total Environment 472 (2014) 912–922

Ebro River Llobregat River

Jucar River

Guadalquivir River Fig. 1. Location of the STPs.

calculated as the percentage of the difference in weight of the wet and dry sample (dried over night at 100 °C) divided by the weight of the wet sample. This moisture was used to report concentrations in dry weight. 2.4.3. LC–MS/MS determination The chromatographic instrument was an HP1200 series LC – with an automatic injector, a degasser, a quaternary pump and a column oven – combined with an Agilent 6410 triple quadrupole (QqQ) mass spectrometer, equipped with an electrospray ionization (ESI) interface (Agilent Technologies, Waldbronn, Germany). Data were processed using MassHunter Workstation Software for qualitative and quantitative (internal standard methodology based on peak areas) analysis (A GL Sciences, Tokyo, Japan). Detailed information relating to instrumental determination is listed in Tables S3 and S4. 2.4.4. Quality assurance/quality control To monitor the results a strong quality control was performed. The analytical methods were carefully validated. In wastewater samples, mean relative recoveries ranging from 55 to 94%, with relative standard deviations between 8 and 18%, were achieved for selected PFASs (detailed information about mean relative recoveries obtained is available in Supplementary information, Table S5). For sludge samples, recoveries were between 79 and 111% and, relative standard deviation was in all cases below 25% at the limits of quantification (LOQs). These limits were determined by the injection of spiked water extracts (n = 3) and estimated as the minimum detectable amount of analyte with a signal-to-noise ratio of 10:1. LOQs were 0.04–8.00 ng g−1 for sludge samples and 0.01–2.00 ng L−1 for influent and effluent waters. Calibration curves were prepared daily obtaining R2 N 0.98. Prior and after each sampling batch (between 25 and 30 samples), calibration curves were constructed by injecting standards at different concentrations (1, 2.5, 5, 10, 25, 50, 65, 75 ng L−1). Each 15 samples, one instrumental and one procedural blanks as well as one positive control were analysed to serve as quality control. Traces of PFOA, always below the LOQs, observed in both blank types indicated some background contamination from the injection system and tubing of the LC– MS/MS, but there was no sample contamination during sampling and analysis. The background contamination can only minimally affect the quantification accuracy and then, it was not taken into account. Each wastewater and sludge sample was analysed in triplicate and the average concentration was reported.

3. Results and discussion 3.1. Occurrence of selected PFASs in wastewater samples Of the 21 analytes screened in this study, 15 were detected in 2010 and 20 in 2011 in influent and effluent waters (Table 1). In both years,

Table 1 Mean concentration and frequency of occurrence of each PFAS in the wastewater samples (IN: influent, OUT: effluent) analysed in 2010 and 2011. Family/compound

2010

Perfluorocarboxylatesa PFBA PFPeA PFHxA PFHpA PFOA i,p-PFNA PFNA PFDA PFUdA PFDoA PFTrDA PFTeDA PFHxDA PFODA Perfluorosulfonatesa L-PFBS L-PFHxS L-PFHpS L-PFOS i,p-PFNS L-PFDS Perfluorosulfonamidesa PFOSA ∑PFASs

2011

IN

Mean Frequency Frequency (ng L−1) (N° (N° occurrences) occurrences) OUT IN OUT

103 20.5 7.76 1.87 103 14.3 19.0 n.d. 8.73 36.7 4.58 1.62 6.80 n.d. n.d. n.d. 63.6 12.0 15.5 14.6 118 3.28 n.d. n.d. n.d. 2157

11.0 13.4 8.09 4.87 9.58 16.4 n.d. 5.52 28.1 2.57 0.07 5.10 n.d. n.d. n.d. 37.7 8.57 14.1 2.19 76.7 0.04 n.d. n.d. n.d. 195

Mean (ng L−1)

26 26 9–8b 6 18 9–10b 0 16 8 5–3b 4 2 0 0 0 12 8 8 8 12–10b 2 0 0 0 26

24.6 53.7 9.35 1.07 13.0 22.4 0.40 21.2 0.58 12.9 13.8 13.2 0.02 0.04 300 35.1 19.1 41.9 8.83 78.1 5.62 n.d. 0.20 0.20 615

25.8 57.9 14.5 17.5 7.48 14.9 0.40 33.7 21.6 5.62 13.3 0.02 0.02 0.04 190 50.7 57.9 37.7 2.91 91.0 0.04 n.d. 0.20 0.20 567

30 30 20 6 20 23 2 18 10 7–5b 6 2 2 2 2 46 12 4–2b 10 6–8b 4 0 2 2 30

n.d.: not detected. a Concentration given in influent and effluent for the different families (bold) is mean values of the compounds, while frequency is the total number of samples with at least one PFAS. b First frequency is in the influents and the second in the effluents, when a unique value is given, the frequency is the same in the influents and effluents.

J. Campo et al. / Science of the Total Environment 472 (2014) 912–922

water samples were contaminated with at least one PFAS. Most frequently detected PFCAs were PFBA, PFNA and PFPeA, and most frequent PFSs, L-PFOS and L-PFBS, being present in more than 33% of the STPs (detection frequency details for the four Rivers in the two sampling campaigns are provided in Supplementary information, Fig. S1). In 2010, PFOSA was not detected, but in 2011, it was identified in 1 out of 6 of the Ebro River STPs. In both years, the Jucar River STPs showed the highest frequency of PFCAs (PFBA, PFDA, PFOA, PFPeA and PFUdA were present in all of the STPs) while those of PFSs were lower (only L-PFHpS was detected in both STPs whereas L-PFOS, L-PFBS, and L-PFHxS were detected in 1 STP). The presence of these PFASs was also high in the Llobregrat River STPs (PFBA and PFOA were found in all the STPs, and L-PFOS, L-PFHxS and PFPeA were found in 2 STPs). PFBA (detected in 4 STPs in 2010 and in 5 in 2011) and L-PFOS (in the 6 STPs in 2010) were the most frequent PFASs in STPs that discharge to the Ebro River. These two compounds together with PFHpA, PFNA, PFOA and PFPeA were also the most frequent (≥ 3 out of 5 STPs) in the Guadalquivir River STPs. Summarising, 5.3% of PFASs considered in this research were detected in 75–100% of the studied STPs, 24% in 50–75%, 19% in 25–50% and 52% in 0–25% (Fig. S2). The mean contribution of each type of PFAS according to their functional group and chain length, in percentage, to the total amount of PFASs determined in the STPs discharging wastewater to the Ebro, Guadalquivir, Jucar and Llobregat is shown in Fig. 2. Descriptive statistics for L-PFOS, PFBA and PFOA, which appear at the highest concentrations, in influent, effluent and sludge of STPs discharging wastewaters in the four Rivers are shown in Fig. 3. Detailed information of maximum concentrations detected in each River STPs is presented in Fig. S3. In 2010, the maximum concentration in the influent samples of the Ebro STPs was for L-PFOS with 689 ng L−1. Among the PFCAs, the highest value was for PFBA (86.7 ng L−1). The concentration of these analytes in the effluent samples was reduced between 30 and 70%, with values of 501 ng L− 1 and 26.3 ng L−1 for L-PFOS and PFBA, respectively (Fig. S3a). Mean contribution of the representative type of compounds, in percentage, to the total amount of PFASs in the Ebro STP wastewaters confirms their dominance (Fig. 2a). PFPeA, PFHpA, PFNA and L-PFBS concentrations ranged from 18.6 to 42.5 ng L−1, being always lower in the effluents than in the influents. These decreases could be attributed to their sorption onto the activated sludge, which could explain the high concentrations detected in those samples, particularly for L-PFOS and PFBA. Guadalquivir River STPs showed the highest concentrations (Fig. S3b), with 5.60 μg L−1 of PFHxA in Cordoba's influent (more than 90% of total PFAS contribution, Fig. 2b). Concentrations of L-PFHps (45.2 ng L−1), PFOA (27.1 ng L−1), L-PFOS (24.9 ng L−1) and PFBA (23.4 ng L−1) were lower. In the effluent samples, the concentrations of PFOA (25.0 ng L−1) and PFBA (27.0 ng L−1) were similar, meanwhile those of PFHxA, L-PFHps and L-PFOS decreased (concentrations around 1.00 ng L−1). The Jucar River STPs presented the lowest PFAS concentrations not surpassing the 150 ng L− 1 in any case. The highest value detected was for the PFDA, 128 ng L−1 in influent and 103 ng L−1 in effluent, followed by L-PFOS with 52.5 and 35.1 ng L− 1, respectively (Figs. 2c and S3c). In the Llobregat River STPs, the highest concentration in the influent samples was for L-PFOS (213 ng L− 1), followed by PFOA (51.5 ng L−1) and L-PFHxS (40.1 ng L−1). Their maximum concentrations in the effluent samples were 31.4, 55.4 and 35.1 ng L−1, respectively (Figs. 2d and S3d). In 2011 sampling campaigns, the PFAS higher concentrations were in the Ebro STP influents for PFBA (302 ng L−1), PFODA (300 ng L−1) and L-PFOS (109 ng L−1), while in the effluents the maximum values were 305 ng L−1 for L-PFBS, 195 ng L−1 for L-PFOS, 190 ng L− 1 for PFODA and 153 ng L−1 for PFBA (Figs. 2a and S3a). In the Guadalquivir STPs, the higher concentrations in the influent samples were observed for L-PFOS (106 ng L−1), PFBA (85.4 ng L−1) and PFNA (52.8 ng L−1).

915

In the effluents, L-PFOS (140 ng L−1) and PFNA (74.8 ng L− 1) kept their high concentrations (Fig. S3b), while PFBA (149 ng L−1) increased one order of magnitude with regard to the influent, doubling its presence in the total distribution of PFASs (Fig. 2b). Jucar STP wastewaters presented again the lowest PFAS concentrations, except for the PFBA that reached values of 148 and 135 ng L−1, in influent and effluent, respectively (Fig. S3c). PFOA concentrations were also high, with 107 ng L−1 in the influent and 67.9 ng L− 1 in the effluent. Maximum concentrations of the other PFASs were b70.0 ng L− 1. Fig. 2c shows how the contribution of the studied PFASs was quite different in influent and effluent, without any dominating compound. In the Llobregat River STPs, the higher concentrations in influent samples were detected for L-PFHxS (130 ng L−1), L-PFOS (89.2 ng L−1) and PFNA (64.4 ng L−1) (Fig. S3d). L-PFOS (283 ng L−1), PFBA (81.1 ng L−1) and PFOA (60.4 ng L− 1) increased their levels at the Llobregat's STP outlets (Fig. 2d). The increased PFAS concentrations in effluent wastewaters suggest an additional contribution within the wastewater stream such as biodegradation of precursor compounds (Arvaniti et al., 2012; Schultz et al., 2006; Sinclair and Kannan, 2006; Stasinakis et al., 2013). Comparison of these results with data previously published for other STPs (see Tables S6 and S7) shows that the concentration ranges in both years were similar to those reported in previous studies, particularly for PFHpA, PFOA, PFNA, PFDA, PFUdA, PFDoA, PFTrDA, L-PFBS, L-PFHxS and L-PFHpS. Values for PFHxA were also in a comparable range, but only in 2011. Levels of PFPeA, in this study, were related to those of Germany (Ahrens et al., 2009). PFHxA showed significantly higher concentration, in 2010, than all those detected in other studies, meanwhile L-PFOS and PFTeDA values were lower (Greece; Arvaniti et al., 2012). The United States Environmental Protection Agency (US-EPA) established provisional health advisory (PHA) concentrations of 400 ng L− 1 for PFOA and 200 ng L−1 for L-PFOS (US EPA, 2009, 2011), which were considered protective in a short-term exposure scenario (acute toxicity). According to the results obtained in this study, the effluents of Pamplona (in 2010) and Igualada (in 2011) exceeded the PHA concentration for L-PFOS (501 and 284 ng L−1, respectively) (Fig. S3a and d). Safety values of 600 ng L−1 for PFHxS and PFBS, and of 1000 ng L−1 for PFHxA and PFPeA were also established by other US local Agencies (Minnesota Pollution Control Agency, 2007; US EPA, 2009). Regarding these values, most of the samples may not pose an immediate human health risk, with an exception of Cordoba's effluent in 2010 (PFHxA at 5.60 μg L−1). Mass loads of individual PFASs (mg day−1), and mass loads normalized to inhabitants (milligrammes per day and 1000 inhabitants), were calculated from PFAS concentrations in each sampling campaign using the corresponding daily effluent volume, and the served people in each STP (Tables S8 and S9). Detected PFAS mean daily loads, in 2010, in the Ebro River basin ranged from 0.04 mg day−1 of i,p-PFNS up to 9.02 103 mg day−1 of L-PFOS. In the following year, loads were between 0.29 and 3.23 103 mg day−1 of PFTeDA and PFBA, respectively. In the Guadalquivir River, the maximum mean loads were 1.06 103 (PFOA) and 5.19 103 (PFBA) mg day− 1. Similar values were observed in the Jucar basin, in the range of 0.70 (L-PFBS)–1.23 103 mg day − 1 (L-PFOS), in 2010, and 1.41 (PFUdA)–4.75 103 mg day− 1 (PFBA), in 2011. Finally, the mean loads reaching the Llobregat River were, in 2010, between 83.2 (PFNA) and 1.80 103 mg day− 1 (PFOA), and between 0.57 (PFPeA) and 3.10 103 mg day− 1 (L-PFOS), in 2011. According to these results, it has not been possible to establish a general pattern for the basins researched (10 different compounds were found in the estimates with differences of up to 6 orders of magnitude). However, making a rough calculation of the total PFAS loads that can be discharged into the rivers, and having in mind the number of STPs in each catchment, the highest loads were those of the Ebro River with around 67.0 g of PFASs reaching every day the watercourses during

916

b)

IN

IN

2010

2010

a)

OUT

OUT

PFCAs 12-18 C SLUDGE

PFCAs 4-7 C

SLUDGE

PFCAs 8-11 C IN

PFSAs 8 C

IN

2011

2011

OUT

SLUDGE

SLUDGE

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

d)

IN

IN

2010

2010

c)

PFSs 8-10 C

OUT

OUT

OUT

PFCAs 12-18 C PFCAs 4-7 C

SLUDGE

SLUDGE

PFCAs 8-11 C PFSAs 8 C

IN

IN

2011

2011

PFSs 4-7 C OUT

SLUDGE

SLUDGE

0%

PFSs 8-10 C

OUT

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Fig. 2. Percentage of PFCAs, PFSs and PFSAs according to the low (C4–7) medium (C8–10) and high (C12–18) length of the chain present in (a) Ebro, (b) Guadalquivir, (c) Jucar and (d) Llobregat STP influent (IN), effluent (OUT) and sludge samples, in both sampling campaigns.

J. Campo et al. / Science of the Total Environment 472 (2014) 912–922

PFSs 4-7 C

J. Campo et al. / Science of the Total Environment 472 (2014) 912–922

917

PFOS concentration

a)

PFBA concentration

b)

PFOA concentration

c)

Fig. 3. Non-parametric probability distribution of (a) L-PFOS, (b) PFBA and (c) PFOA levels in influent, effluent and sludge of the four river STPs. (Top and bottom of each box represent 75th and 25th percentile, respectively, the top and bottom of each whisker represent maximum and minimum values; line across inside of each box represents median; and circles and stars beyond whiskers mean outliers).

918

J. Campo et al. / Science of the Total Environment 472 (2014) 912–922

2010 and 2011. In the other basins, total discharges were lower with values between 3.97 g day−1 in the Jucar and 32.2 g day−1 in the Guadalquivir (Table S8). Normalization of STPs daily loads to the number of served inhabitants showed that, in the Ebro basin, means were between 0.01 (PFOSA, in 2011) and 31.1 mg day−1 per 1000 inhabitants (L-PFOS, in 2010) (Table S9). In the Guadalquivir and Jucar Rivers, means were also in the same range, with 0.04 (L-PFHxS)–21.5 (PFPeA) mg day−1 per 1000 inhabitants (both in 2011), and 0.01 (PFPeA and L-PFBS, 2010)– 41.6 (PFBA, in 2011) mg day−1 per 1000 inhabitants, in the latter. Normalized load means in the Llobregat, ranged from 0.01 mg day−1 per 1000 inhabitants of PFDA, in 2011, up to 106 mg day−1 per 1000 inhabitants of L-PFOS, in 2010. Regarding the maximum loads provided by each STP in each basin, the highest value was detected in Igualada with 448 mg day−1 per 1000 inhabitants of L-PFOS, followed by Pamplona with 124 mg day−1 per 1000 inhabitants of L-PFOS, and by Moron with 108 mg day−1 per 1000 inhabitants of PFPeA. A lower normalized mean was observed in Alzira with only 41.6 mg day−1 per 1000 inhabitants of PFBA. In the literature there are few data available regarding PFAS loading in STPs. Maximum values obtained in Spanish STPs were similar to those reported by Loganathan et al. (2007) with daily flows ranging up to 227 mg day−1 per 1000 inhabitants for PFOA in a Georgia STP. Likewise, mean daily loads showed in the present study, are comparable with the 42.0 and 44.6 mg day− 1 per 1000 inhabitants, detected by Yu et al. (2009) in a Singapore STP for PFOA and L-PFOS, respectively. Mass loads calculated in two Greece STPs were in the same range of those obtained in Spain, with maximum values between 820 (PFNA) and 19 103 (PFPeA) mg day− 1 (Arvaniti et al., 2012). However, normalized mass loads proposed in that research for one of the STPs were significantly lower than those calculated here with mean values of only 0.02 (PFHxA) — 14.2 (PFTrDA) mg day−1 per 1000 inhabitants.

3.2. Occurrence of selected PFASs in sludge samples In 2010, 13 PFASs and, in 2011, all (21) were detected in the sludge samples. Sludge was not analysed in Loja, Moron and Manresa STPs, in 2010, and in Logroño and Ranilla STPs, in 2011. Only the samples of Copero treatment plant, in 2010, and those of Cordoba, in both years, were not contaminated with any PFAS (Table 2). Fig. 3 shows the descriptive statistical for the STPs that discharges in each river. Most of the PFASs detected in influent and effluent samples were also present in sludge. Among PFCAs, the most frequently detected were PFBA and PFPeA. The PFNA, ubiquitous in the sewage of all STPs, was also found in their sludges with the exception of those of the Jucar River STPs. The PFSs most commonly found in the samples of this study were L-PFOS and L-PFBS, which were present in 2–6 STPs out of 6 depending on the sampling campaign. Similarly to that reported in the wastewaters, PFOSA was not detected in any sludge sample of the first year, but in the second, it was detected in 1–3 STPs out of 5. Differently to wastewaters, in both years, sludge samples of Llobregat STPs showed the highest frequencies of PFCAs (PFBA, PFDA, PFHpA, PFPeA and PFUdA), and of some PFSs (L-PFBS and L-PFOS). In 2010, L-PFDS, L-PFHpS and L-PFHxS were not observed in any sludge sample, but in the following year the first two were detected in 1 STP out of 3, and the last one in 2 STPs out of 3. PFBA, PFPeA, L-PFBS and L-PFOS were found in both years in all the Jucar River STPs. Sludge samples of Guadalquivir STPs did not present any PFAS contamination in 2010, but in 2011, PFBA and L-PFOS were detected in 4 treatment plants out of 5 (the other PFASs were identified in 3 STPs out of 5). PFBA and L-PFOS were present in the sludge of all Ebro River STPs, while PFHpA, PFNA, PFPeA and L-PFBS were detected only in 2–3 STPs out of 5, in 2010, and out of 6, in 2011. In summary, PFBA and L-PFOS were found in 80–100% of the sludge samples studied.

Table 2 Mean concentration and frequency of occurrence of each PFAS in the sludge samples analysed in 2010 and 2011. Family/compound

2010

2011

Mean Mean Frequency Frequency (ng g−1 dw) (N° (ng g−1 dw) (N° occurrences) occurrences) Perfluorocarboxylatesa PFBA PFPeA PFHxA PFHpA PFOA i,p-PFNA PFNA PFDA PFUdA PFDoA PFTrDA PFTeDA PFHxDA PFODA Perfluorosulfonatesa L-PFBS L-PFHxS L-PFHpS L-PFOS i,p-PFNS L-PFDS Perfluorosulfonamidesa PFOSA ∑PFASs

56.5 203 2.50 11.1 3.27 1.83 n.d. 1.52 12.7 0.18 0.10 n.d n.d. n.d. n.d. 143 38.2 n.d. n.d. 229 0.13 n.d. n.d. n.d. 503

9 9 6 1 5 2 0 4 3 2 2 0 0 0 0 9 6 0 0 9 1 0 0 0 9

95.9 399 193 1.33 6.41 21.7 17.6 27.2 160 8.68 0.10 4.17 18.8 0.13 14.1 13.3 0.78 0.01 1.98 41.4 2.15 0.01 0.67 0.67 920

14 14 11 6 9 6 5 8 7 8 12 6 5 5 5 14 10 6 5 14 6 5 5 5 14

dw: dry weight; n.d.: not detected. a Concentration given for the different families (bold) is mean values of the compounds, while frequency is the total number of samples with at least one PFAS.

In 2010, sludge samples of the Llobregat STPs not only presented high frequencies but also concentrations, showing 1790 ng g−1 in dry weight (dw) of L-PFOS, and 149 ng g−1 dw of PFBA (Fig. S4). The high concentration of PFASs in the influent and their possible adsorption onto the activated sludge accounted for its apparent decrease in the effluent, and its high concentration in the sludge. According to Schultz et al. (2006), there is no known biodegradation pathway for these compounds. The PFAS residue levels in the Jucar River samples were low, reaching the highest level at 175 ng g−1 dw (L-PFBS). In the Ebro STP sludge samples, the highest concentration detected was for PFBA (442 ng g−1 dw), followed by that of L-PFOS (57.0 ng g−1 dw). In 2011, sludge samples of the Llobregat STPs presented low PFAS levels reaching only 3.26 and 0.13 ng g−1 dw of L-PFOs and PFBA, respectively. The highest value obtained was 93.7 ng g−1 dw for PFTeDA (Fig. S4). In the Jucar STPs, the concentrations detected were also low with values ranging between 28.6 ng g−1 dw (PFBA) and 68.3 ng g−1 dw (L-PFOS). Conversely, PFASs showed high concentrations in the Guadalquivir STP sludge samples for PFBA (1.88 μg g−1 dw), PFPeA (1.08 μg g−1 dw), PFDA (0.67 μg g−1 dw) and L-PFOS (0.50 μg g−1 dw). Finally, PFAS concentration in the Ebro STP sludge samples was also high for the PFBA (0.31 μg g−1 dw) but PFPeA and PFDA showed also elevated values, 0.72 and 0.38 μg g−1 dw, respectively (Fig. S4). Comparing the PFAS concentrations detected in other studies (see Table S10) with those obtained in the present one (Table 2), most of them were in the same range. To mention some differences, maximum values of PFUdA and PFOA found in New York State, USA (Sinclair and Kannan, 2006) were higher than those detected in this research, contrary to L-PFOS concentrations in Greek sludge that were lower (Arvaniti et al., 2012). Only the mean values of PFDA were higher, particularly those detected in 2011 in Guadalquivir and Ebro samples, meanwhile PFDoA, PFDS and PFHxS concentrations were quite lower than those reported by Arvaniti et al. (2012),

J. Campo et al. / Science of the Total Environment 472 (2014) 912–922

Schultz et al. (2006) and Sinclair and Kannan (2006). Up to our knowledge, the only research reporting PFAS occurrence in Spanish STP sludge samples was carried out in Catalonia (Llorca et al.,

2011). Results obtained in that study were similar to those reported here. Only the mean concentrations found here for PFBA, in both years, and those of PFNA, PFDA and PFODA, in 2011, were higher,

a) 150

Remvoal efficiency (%)

100

50

0

-50

-100

-150

b) 150

Remvoal efficiency (%)

100

50

0

-50

-100 -399

-557

-150

c) 150

100

Remvoal efficiency (%)

919

50

0

-50

-100

-150

Fig. 4. PFAS RE of the (a) Ebro, (b) Guadalquivir, (c) Jucar and (d) Llobregat STPs in 2010.

920

J. Campo et al. / Science of the Total Environment 472 (2014) 912–922

d) 150

Remvoal efficiency (%)

100

50

0

50

-100

-150

Fig. 4 (continued).

while those of PFDoA, L-PFDS and PFOSA were lower than the ones reported by Llorca et al. (2011). 3.3. Removal of PFASs by STPs The removal efficiency (RE, %) of PFASs was calculated from the analyte concentration in influent (Cin) and effluent (Cef) as: RE ð% Þ ¼ ½ðCin −Cef Þ=Cin   100%: According to this equation, and evaluating the RE of the most ubiquitous compounds, in 2010, the elimination of PFCAs ranged from − 557% (PFNA) to 100% (PFPeA, PFUdA), meanwhile the removal of PFSs was in the range 0% (L-PFBS, L-PFHxS and L-PFHpS) to 100% (L-PFOS). PFSA efficiency could not be calculated since PFOSA was not detected in any water sample. PFAS removal in the STPs was globally variable and often poor, with concentrations in the effluent sometimes higher than in the corresponding influent. This behaviour including the “negative” RE has been widely reported for many organic contaminants and for PFASs in particular (Arvaniti et al., 2012; Schultz et al., 2006). Possible explanations for these poor or negative removal rates are, among many others considered (e.g. sampling, method biases, matrix effects, atmospheric deposition), the existence of an additional source of these compounds within the wastewater stream such as chemical and biological degradation of precursor compounds or desorption from particulate matter during wastewater treatment. Furthermore, occasionally, PFBA, PFDA, PFOA, PFPeA, L-PFBS, and L-PFHpS were detected in the influent at the LOQ level and at moderate concentrations at the effluent (up to 380 ng L−1). These cases were not included in the RE calculations because the abnormally high negative values obtained are unrealistic. These results reinforce the hypothesis that these compounds could be released from the precursors. Studying the PFAS removal in the STPs by year, in 2010, the less efficient in the Ebro River was Logroño, with 0% of PFBA, PFPeA and L-PFBS elimination, and the most efficient was Tortosa, with 99% of i,p-PFNS removal (Fig. 4). In the Guadalquivir basin, Ranilla showed very low (−557% of PFNA) and very high (100% of PFPeA) elimination efficiencies, even though removals of 100% were also calculated for L-PFOS in Loja and Copero. In the Jucar River STPs, the efficiencies were between − 121% of PFBA, in Alzira, and 82% of PFOA, in Cuenca. Finally, in the Llobregat basin, the removal ranged from − 11% of PFHpA to 100% of PFPeA and PFUdA, all in Manresa. Next year, 2011, in the Ebro basin, Tortosa showed low elimination efficiency of PFBA (− 342%), and Tudela very high (100% of L-PFOS).

In the Guadalquivir River, Ranilla STP presented the lowest efficiency, with −1217% of PFBA elimination, but also the highest, with 100% of PFPeA, L-PFOS and L-PFHxS removal. Additionally, Copero STP also removed 100% of L-PFOS. In the Jucar River STPs, the removal ranged from − 9200% of PFDA, in Alzira, to 100% of PFUdA, in Cuenca (Fig. S5). Finally, in the Llobregat River, the efficiencies were between − 580% of PFHpA, in Igualada, and 100% of PFUdA and L-PFHxS in Abrera, Igualada and Manresa. Summarizing, and regarding mean values over both years, 38% of the PFASs analysed in this study were not eliminated or even reduced in the treatment plants, 56% reached elimination efficiencies between 25 and 75% and none of the efficiencies exceeded the 75% (for detailed data see Fig. S6). Among the PFAS removal efficiency reported in other studies, the highest value obtained by Arvaniti et al. (2012) was 73% for L-PFHxS, which is in general higher than those calculated here in 2010 but lower than those of 2011, in Guadalquivir and Llobregat STPs. Arvaniti et al. (2012) also reported mean negative removal efficiencies for PFPeA and PFOA and no clear trends for L-PFOS, PFHxA, PFHpA and PFUdA with values ranging from −408% to 64% in the case of L-PFOS. Schultz et al. (2006) found positive efficiencies for L-PFBS and PFNA, and negative ones for PFDA, while unambiguous patterns could not be established for the recoveries of L-PFHxS, L-PFOS, L-PFDS, PFHxA, PFHpA and PFOA. According to the results of the present study, PFPeA, PFHxA and PFOA were efficiently removed in 2010 but not in 2011, in agreement with previous findings. PFHpA showed unclear RE trends. L-PFOS and PFUdA were quite efficiently removed (up to 100%), only in Ebro's STP negative REs were observed for L-PFOS in 2011. 3.4. Estimation of the solid–liquid distribution coefficients The solid–liquid distribution coefficient (Kd) was estimated as Kd ¼ Cs =Cw where Kd is the sorption coefficient (L kg−1), Cs is the concentration of target compound in sludge samples (ng kg−1), and Cw is the concentration of target compound in influent samples (ng L−1). Results as mean obtained for each year are outlined in Table 3. Most previous studies carried out in sediments or in suspended particulate matter clearly established that Kd increases with the perfluorocarbon chain length and depends on the functional group (Ahrens et al., 2010; Higgins and Luthy, 2006; Kwadijk et al., 2010; Pico et al., 2012). Kd calculated for the PFASs in this study showed values ranging from

J. Campo et al. / Science of the Total Environment 472 (2014) 912–922 Table 3 Mean distribution coefficient values, Kd (in L kg−1) for PFASs obtained from 2010 and 2011 samples. Family/Compound Perfluorocarboxylates PFBA PFPeA PFHxA PFHpA PFOA i,p-PFNA PFNA PFDA PFUdA PFDoA PFTrDA PFTeDA PFHxDA PFODA Perfluorosulfonates L-PFBS L-PFHxS L-PFHpS L-PFOS i,p-PFNS L-PFDS Perfluorosulfonamides PFOSA

2010

2011

Mean

5.26 103 9.27 103 n.c. 76.3 1.18 n.c. 41.7 138 230 11.9 n.c. n.c. n.c. n.c.

36.6 103 10.5 103 222 9.3 103 44.1 103 222 35.2 103 n.c 335 167 0.34 223 222 0.59

20.9 103 9.94 103

77.7 n.c. n.c. 1.04 103 3.70 n.c.

1.48 103 0.32 2.10 8.88 222 n.c.

897

n.c.

222

5.4 103 25.4 103 20.3 103 290 101

444 130 n.c.

n.c.: not calculated.

0.34 for PFTrDA to 36.6 103 L kg−1 for PFBA. Values were chaotic and do not confirm any of the already established tendency in sediment. On the contrary, low chain PFASs as PFBA, PFPeA and PFHxA and PFBS show abnormally high Kd whereas PFTrDA and PFODA show values too low. Important differences were observed calculating these Kd from one year to another, which at the first sight could seem a non-sense since this value should be constant. One of the possible reasons to explain the high Kd values for short chain PFASs as well as the differences observed can be the formation of low chain PFASs by degradation of precursor molecules. The low values observed for some long chain PFASs are only for those with very low occurrence and then, could be little representative and attributable to error inherent to the sampling method. The results obtained in this study, even that could seem chaotic, agree with various studies conducted in sewage sludge (Arvaniti et al., 2012; Stasinakis et al., 2013; Zhou et al., 2010). These studies also report chaotic Kd values, which are not related to the alkyl chain length and vary depending on the type of sludge analysed: primary, secondary, mixed liquor, dehydrated sludge, etc. that could also support the release of these compounds from precursors. This hypothesis is reinforced by the fact that in those samples (in which PFASs were not detected in influent but were in the effluent systems) were accompanied of high concentrations also in the sludge samples. Other factors that could influence the Kd particularly in sludge, and that has been previously reported (Arvaniti et al., 2012), are the negative charge of the sludge that could create electrostatic repulsion and reduce the adsorption, at the same time that the existence of cations could act as ion bridges enhancing the absorption of PFASs, which are negatively charged.

4. Conclusions All samples analysed in this study were contaminated with at least one PFAS except some sludge from Guadalquivir's STPs. A high number of PFASs were detected in 2010 (15 and 13 in wastewater and sludge, respectively), and in 2011 (20 and 21). The compounds most frequently identified were perfluorobutanoate (PFBA), perfluoropentanoate (PFPeA) and perfluorooctane sulfonate (L-PFOS). Despite the fact that production of L-PFOS ceased in 2002, the results of this study indicate that this compound is still actively spilled to the environment.

921

Concentrations of most PFASs were higher in effluents than in influents, particularly in 2011, suggesting their release from precursors during wastewater treatment. On the other hand, positive efficiencies could be more related to PFAS sorption onto the sludge than to capable degradation processes. Both facts could indicate that sewage treatment plants constitute a focal point of contamination to the rivers, although this conclusion is not detrimental to the important role that they play in wastewater depuration. An improved understanding of the presence and behaviour of the different PFASs needs a better knowledge of the precursors present in the wastewaters. Longer temporal researches (including seasonal campaigns and hourly sampling over different periods of time), as well as sample analysis after each treatment stage are possible approaches that could help to assess whether the results observed here are representative of the PFAS fate throughout the sewage treatment process in Spanish plants. Such facilities are responsible of the high PFAS loads that can reach every day the watercourses (Ebro River: 66.9 g day−1; Guadalquivir: 32.2 g day−1, Jucar 17.0 g day−1 and Llobregat 15.9 g day−1), endangering the water quality of ecosystems where they can bio-accumulate and, potentially, produce adverse effects on humans. Acknowledgements This work has been supported by the Spanish Ministry of Economy and Competitiveness through the Projects “Assessing and Predicting Effects on Water Quantity and Quality in Iberian Rivers Caused by Global Change” (SCARCE, CSD-665-2009) and “Evaluation of Emerging Contaminants in the Turia River Basin: From Basic Research to the Application of Environmental Forensics” (EMERFOR CGL2011-29703-C02-02). Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.scitotenv.2013.11.056. References Ahrens L, Felizeter S, Sturm R, Xie Z, Ebinghaus R. Polyfluorinated compounds in waste water treatment plant effluents and surface waters along the River Elbe, Germany. Mar Pollut Bull 2009;58:1326–33. Ahrens L, Taniyasu S, Yeung LWY, Yamashita N, Lam PKS, Ebinghaus R. Distribution of polyfluoroalkyl compounds in water, suspended particulate matter and sediment from Tokyo Bay, Japan. Chemosphere 2010;79:266–72. Arvaniti OS, Ventouri EI, Stasinakis AS, Thomaidis NS. Occurrence of different classes of perfluorinated compounds in Greek wastewater treatment plants and determination of their solid–water distribution coefficients. J Hazard Mater 2012;239–240:24–31. Austin ME, Kasturi BS, Barber M, Kannan K, MohanKumar PS, MohanKumar SMJ. Neuroendocrine effects of perfluorooctane sulfonate in rats. Environ Health Perspect 2003;111:1485–9. Becker AM, Gerstmann S, Frank H. Perfluorooctane surfactants in waste waters, the major source of river pollution. Chemosphere 2008;72:115–21. Bossi R, Strand J, Sortkjaer O, Larsen MM. Perfluoroalkyl compounds in Danish wastewater treatment plants and aquatic environments. Environ Int 2008;34:443–50. Environmental Protection Agency US. Provisional health advisories for perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS). http://water.epa.gov/action/ advisories/drinking/upload/2009_01_15_criteria_drinking_pha-PFOA_PFOS.pdf, 2009. Environmental Protection Agency US. Drinking water health advisories and science support. http://water.epa.gov/action/advisories/drinking/drinking_index.cfm, 2011. European Parliament and Council. Directive 2006/122/ECOF; 2006. Giesy JP, Kannan K. Global distribution of perfluorooctane sulfonate in wildlife. Environ Sci Technol 2001;35:1339–42. Guo R, Sim WJ, Lee ES, Lee JH, Oh JE. Evaluation of the fate of perfluoroalkyl compounds in wastewater treatment plants. Water Res 2010;44:3476–86. Guruge KS, Taniyasu S, Yamashita N, Wijeratna S, Mohotti KM, Seneviratne HR, et al. Perfluorinated organic compounds in human blood serum and seminal plasma: a study of urban and rural tea worker populations in Sri Lanka. J Environ Monit 2005;7:371–7. Han X, Snow TA, Kemper RA, Jepson GW. Binding of perfluorooctanoic acid to rat and human plasma proteins. Chem Res Toxicol 2003;16:775–81. Haug LS, Thomsen C, Becher G. A sensitive method for determination of a broad range of perfluorinated compounds in serum suitable for large-scale human biomonitoring. J Chromatogr A 2009;1216:385–93. Higgins CP, Luthy RG. Sorption of perfluorinated surfactants on sediments. Environ Sci Technol 2006;40:7251–6.

922

J. Campo et al. / Science of the Total Environment 472 (2014) 912–922

Holzer J, Midasch O, Rauchfuss K, Kraft M, Reupert R, Angerer J, et al. Biomonitoring of perfluorinated compounds in children and adults exposed to perfluorooctanoatecontaminated drinking water. Environ Health Perspect 2008;116:651–7. Huset CA, Chiaia AC, Barofsky DF, Jonkers N, Kohler HP, Ort C, et al. Occurrence and mass flows of fluorochemicals in the Glatt Valley Watershed, Switzerland. Environ Sci Technol 2008;42:6369–77. Jahnke A, Berger U, Ebinghaus R, Temme C. Latitudinal gradient of airborne polyfluorinated alkyl substances in the marine atmosphere between Germany and South Africa (53°N–33°S). Environ Sci Technol 2007;41:3055–61. Joensen UN, Bossi R, Leffers H, Jensen AA, Skakkebaek NE, Jørgensen N. Do perfluoroalkyl compounds impair human semen quality? Environ Health Perspect 2009;117(6): 923–7. Kissa E. Fluorinated surfactants and repellents, Marcel Dekker 97; 2001. Kunacheva C, Tanaka S, Fujii S, Boontanon SK, Musirat C, Wongwattana T, et al. Mass flows of perfluorinated compounds (PFCs) in central wastewater treatment plants of industrial zones in Thailand. Chemosphere 2011;83:737–44. Kwadijk CJAF, Korytar P, Koelmans AA. Distribution of perfluorinated compounds in aquatic systems in The Netherlands. Environ Sci Technol 2010;44:3746–51. Lau C, Anitole K, Hodes C, Lai D, Pfahles-Hutchens A, Seed J. Perfluoroalkyl acids: a review of monitoring and toxicological findings. Toxicol Sci 2007;99:366–94. Lewandowski G, Meissner E, Milchert E. Special applications of fluorinated organic compounds. J Hazard Mater 2006;136:385–91. Llorca M, Farre M, Pico Y, Barcelo D. Analysis of perfluorinated compounds in sewage sludge by pressurized solvent extraction followed by liquid chromatography–mass spectrometry. J Chromatogr A 2011;1218:4840–6. Llorca M, Farre M, Pico Y, Muller J, Knepper TP, Barcelo D. Analysis of perfluoroalkyl substances in waters from Germany and Spain. Sci Total Environ 2012;431: 139–50. Loganathan BG, Sajwan KS, Sinclair E, Senthil Kumar K, Kannan K. Perfluoroalkyl sulfonates and perfluorocarboxylates in two wastewater treatment facilities in Kentucky and Georgia. Water Res 2007;41:4611–20. Ma R, Shih K. Perfluorochemicals in wastewater treatment plants and sediments in Hong Kong. Environ Pollut 2010;158:1354–62. Martin JW, Mabury SA, Solomon KR, Muir DCG. Bioconcentration and tissue distribution of perfluorinated acids in rainbow trout (Oncorhynchus mykiss). Environ Toxicol Chem 2003;22:196–204. Minnesota Pollution Control Agency. Perfluorochemicals and health: environmental health. http://www.pca.state.nm.us/index.php/waste/waste-and-cleanup/cleanupprograms-and-topics/topics/perfluorochemicals-pfcs.html?menuid=&redirect=1, 2007. Navarro-Ortega A, Acuña V, Batalla RJ, Blasco J, Conde C, Elorza FJ, et al. Assessing and forecasting the impacts of global change on Mediterranean rivers. The SCARCE Consolider project on Iberian basins. Environ Sci Pollut Res 2012;19:918–33. Pico Y, Blasco C, Farre M, Barcelo D. Occurrence of perfluorinated compounds in water and sediment of L'Albufera Natural Park (Valencia, Spain). Environ Sci Pollut Res 2012;19:946–57.

Qiu Y. Study on treatment technologies for perfluorochemicals in wastewater. Kyoto Publisher University; 2007. Schultz MM, Higgins CP, Huset CA, Luthy RG, Barofsky DF, Field JA. Fluorochemical mass flows in a municipal wastewater treatment facility. Environ Sci Technol 2006;40: 7350–7. Sinclair E, Kannan K. Mass loading and fate of perfluoroalkyl surfactants in wastewater treatment plants. Environ Sci Technol 2006;40:1408–14. Skutlarek D, Exner M, Farber H. Perfluorinated surfactants in surface and drinking waters. Environ Sci Pollut Res 2006;13:299–307. Stasinakis AS, Thomaidis NS, Arvatini OS, Asimakopoulos AG, Samaras VG, Ajibola A, et al. Contribution of primary and secondary treatment on the removal of benzothiazoles, benzotriazoles, endocrine disruptors, pharmaceuticals and perfluorinated compounds in a sewage treatment plant. Sci Total Environ 2013;463–364:1067–75. Sun H, Gerecke CA, Giger W, Alder AC. Long-chain perfluorinated chemicals in digested sewage sludges in Switzerland. Environ Pollut 2011;159:654–62. Sundstrom M, Ehresman DJ, Bignert A, Butenhoff JL, Olsen GW, Chang SC, et al. A temporal trend study (1972–2008) of perfluorooctanesulfonate, perfluorohexanesulfonate, and perfluorooctanoate in pooled human milk samples from Stockholm, Sweden. Environ Int 2011;37:178–83. Taniyasu S, Kannan K, Horii Y, Hanari N, Yamashita N. A survey of perfluorooctane sulfonate and related perfluorinated organic compounds in water, fish, birds, and humans from Japan. Environ Sci Technol 2003;37:2634–9. Tao L. Biomonitoring of perfluorochemicals in human blood, breast milk, saliva, and urine. Environmental health sciences. Albany, New York, United States: State University of New York; 2009247 (PhD). Tomy GT, Budakowski W, Halldorson T, Helm PA, Stern GA, Friesen K, et al. Fluorinated organic compounds in an Eastern arctic marine food web. Environ Sci Technol 2004;38:6475–81. United Nations Environment Programme. New POPs SC-4/17: listing of perfluorooctane sulfonic acid, its salts and perfluorooctane sulfonyl fluoride. United Nations Environment Programme: Stockholm Convention on Persistent Organic Pollutants (POPs), Geneva, Switzerland; 2010. van Leeuwen SPJ, Swart CP, van der Veen I, de Boer J. Significant improvements in the analysis of perfluorinated compounds in water and fish: results from an interlaboratory method evaluation study. J Chromatogr A 2009;1216:401–9. Yamashita N, Kannan K, Taniyasu S, Horii Y, Petrick G, Gamo T. A global survey of perfluorinated acids in oceans. Mar Pollut Bull 2005;51:658–68. Yan H, Zhang CJ, Zhou Q, Chen L, Meng XZ. Short- and long-chain perfluorinated acids in sewage sludge from Shanghai, China. Chemosphere 2012;88:1300–5. Young CJ, Furdui VI, Franklin J, Koerner RM, Muir DCG, Mabury SA. Perfluorinated acids in arctic snow: new evidence for atmospheric formation. Environ Sci Technol 2007;41: 3455–61. Yu J, Hu J, Tanaka S, Fujii S. Perfluorooctane sulfonate (PFOS) and perfluorooctanoic acid (PFOA) in sewage treatment plants. Water Res 2009;43:2399–408. Zhou Q, Deng S, Zhang Q, Fan Q, Huang J, Yu G. Sorption of perfluorooctane sulfonate and perfluorooctanoate on activated sludge. Chemosphere 2010;81:453–8.