Keywords

1 Introduction: Uses, Production and Global Distribution

Perfluorinated compounds (PFCs) have been manufactured since 1950s. Because of their properties, these compounds are employed for many industrial applications including stain repellents, textile, paints, waxes, polishes, electronics, adhesives and food packaging. Two of the most important PFCs are perfluorooctane sulphonate (PFOS) salts, components of fire-fighting foam concentrates, and perfluorooctane acid (PFOA), primarily used as emulsifier in industrial applications, for example in the production of fluoropolymers such as polytetrafluoroethylene (PTFE). Furthermore, PFOS and PFOA as well as other perfluoro carboxylic acids (PFCAs) are stable degradation products and/or metabolites of neutral PFCs like fluorotelomers alcohols (PFTOHs), perfluorinated sulphonamides (PFASAs) and perfluorinated sulphonamide ethanols (PFASEs). Figure 1 shows the most commonly used PFCs structures.

Fig. 1
figure 1

The most common structures of PFCs

Full size image

Because of their use in different industrial applications, these compounds enter to the environment through removal facilities, which are not able to degrade PFCs, or, for example, by their application in agricultural soils or by irrigation with contaminated water. These analytes are widespread around the world in water and soils as well in organism due the high affinity to low weight proteins [1]. The bioaccumulation derives to biomagnification through the food chain and, finally, arrives to human through diet and drinking water [2, 3]. PFCs have been detected in environmental and biological samples. They are present in remote areas as the Arctic (atmosphere [4], Arctic Ocean [5], biological samples [6, 7] and few reviews have been published [8, 9]) or Antarctic (biological samples as penguins or seals [10, 11]). Regarding the presence in human matrices, these analytes have been reported in blood from donors from different countries [12, 13], liver [14], urine, human breast milk [14–17] and cord blood [18, 19] being the breastfeeding and pregnancy a possible transferability route. Fish is another biological matrix, which has been analysed due to its involvement into marine trophic chain. This is one of the main entrance routes of PFCs into human organisms with other daily products as milk, meat or vegetables [2, 3, 20–23]. In 2006, EPA established the tolerable daily intake (TDI) for PFOA and PFOS [24, 25], and in 2008 the EFSA established as well, TDI levels at 150 ng/kg bw per day in the case of PFOS and 1,500 ng/kg bw for PFOA.

PFCs are considered as emerging organic pollutants since they have not been regulated. However, in the last decade, there are some of them which are proposed to be under regulation [26].

In 2006, EPA and the eight major PFCs producer companies [Arkema, Asahi, BASF Corporation (successor to Ciba), Clariant, Daikin, 3M/Dyneon, DuPont and Solvay Solexis] launched the “PFOA Stewardship Program”. The companies committed to phase out global emissions by 2015 [27]. Recently, PFOS has been included as a persistent organic pollutant (POP) under the Stockholm Convention for global regulation of production and use [28]. PFCs are also prime candidates for chemicals that will need authorization within the REACH regulation [26]. PFOS was added to the Annex III Substances subject to review for possible identification as priority substances or priority hazardous substances of the Directive 2008/105/EC of the European Parliament and Council of 16 December 2008 concerning the environmental quality standards (EQS) in the field water policy [29].

Wastewater has been also identified as a major source of PFCs in the environment since currently their elimination during wastewater treatments in wastewater treatment plants (WWTPs) is not completely achieved, and therefore, important amounts of these compounds can reach the natural environments through the treated effluents. Moreover, also sewage sludge produced in WWTPs is important source of PFCs’ contamination, since there is a redistribution of PFOS and long carbon chain PFCs into the sludge [30–32]. Several works during last years have informed about concentrations in sludge in the range between low ng/g and μg/g. These high concentrations are of concern because sewage sludge can be partially used in agricultural lands generating an indirect source of PFCs via consumption of crops, air-borne transport, surface waters and ground waters draining from these sites [33–35]. Clarke et al. scored different groups of organic contaminants commonly found in sewage sludge with respect to their potential significance for agricultural utilization [36], and in this classification PFCs obtained 10 scores over 11, based on their persistence in soil (more than 6 months), their potential accumulation in human food chain, their potential bioaccumulation and their possible soil ecotoxicity [36]. On the other hand, it should be considered that PFCs in sludge amended soil can be mobilized by rainfall [37], reaching phreatic waters.

Drinking water has been identified as one of the major sources of human exposure [38, 39] to PFCs. For this reason, in 2009, the EPA’s Office of Water (OW) have set a Provisional Health Advisories (PHA) maximum concentration values for PFOA and PFOS in drinking water. PHA values are 0.4 μg/L for PFOA and 0.2 μg/L for PFOS [40].

To protect human health and the environment against PFCs’ contamination, there is a need to assess their presence in the environment and main sources of human exposure and assess possible damages involving their occurrence, bioaccumulation, as well as, their environmental fate and behaviour. In this context, this chapter provides a summary of the state-of-the-art in the analysis of PFCs in environmental samples, and the occurrence of PFCs will be presented and discussed in a typical Mediterranean river: the Llobregat River.

2 Analysis of PFCs

2.1 Sampling Process and Preservation

Storage and conservation of samples for PFCs’ analysis present some critical steps because losses or contamination of the samples can easily occur.

Martin et al. [41] summarized the key challenges in trace analysis of PFCs. They include blank contamination issues, purity of reference standards and matrix effects in the ionization process of the mass spectrometer. Blank contamination is one of the most relevant problems associated with PFCs’ analysis, and it is associated with fluoropolymer during sampling, storage and materials used in the laboratory. In order to avoid sources of contamination different measures have been suggested, as for example, pre-cleaning of the containers and materials prior sampling by rinsing with semi-polar solvents [38]. It is also important the materials involved during sampling, storage and the analytical process, being polypropylene (PP) containers [34, 42, 43], high density polyethylene (PE) bottles [44–46], or foil containers for solid samples [35] the recommended ones, because these materials cannot contaminate the samples. However, for solid samples losses by sorption to PP and PE containers can be considered negligible [46], whereas for aqueous samples in case of long-chain PFCs is higher [47, 48]. On the other hand, stored aqueous samples in glass have been widely discussed by different authors, whereas some works have reported sorption in glass when prepared samples were tested [49], this phenomena is not expected in real samples with complex matrices [50]. It is also important to consider possible losses due to volatilization of the PFTOHs and short C chain PFCs as PFBA. Therefore, it is recommended to avoid headspace in sampling bottles [51], and kept samples at low temperature after collection [44, 45, 52]. Szostek et al. studied the stability of PFTOHs in water under different storage conditions [53]. In this work it was concluded that aqueous samples can safely be stored in the freezer using glass vials and sealed with a septum lined with alumina foil. In addition, no biodegradation or biotransformation was observed under these conditions [53]. On the other hand, also it was studied the possible benefices associated with the use of biological inhibitors (such as formalin) to preserve the samples, but it was found that some suppress the MS responses during the analysis [54], and therefore should be avoided.

Drying procedures are usually applied for solid matrices using room temperature until a constant weight [42, 55], with soft temperature (40°C) along 3–4 days in porcelain bowls [46] or PP containers [34], and in an oven at 103°C overnight [52]. Other specific procedures consisted of direct freeze (−20°C) prior to any treatment in order to perform lyophilization [35, 43] or previous centrifugation to remove supernatant [56] and lyophilization [44]. Dried sludge or sediment is finely ground (<0.5 mm) [46] and homogenized with a mortar and pestle [35, 42–44, 55]. This homogenized sample is subsequently passed through a mesh sieve to remove pebbles or debris [43, 55]. Homogenized samples are kept frozen until analysis in PP containers [35, 42] or high-density PE bottles [44, 46]. Other authors interested in the analysis of PFCs in raw sewage sludge from Wets prepare the sample by centrifugation followed by filtration by GF/B glass filter and stored these filters at 4°C until extraction usually by solid–liquid extraction [45].

2.2 Sample Pre-treatments

Table 1 summarizes main sample pre-treatment that can be applied to the analysis of PFCs in different types of environmental matrices.

Table 1 Sample pre-treatments and instrumental analysis of different published works
Full size table

2.2.1 Water Samples

Extraction procedures for water analysis have been carried out using protocols based on solid phase extraction (SPE). Due to the different polarities of PFCs, different extraction SPE cartridges have been explored. Broadly, good recoveries were reported using Oasis WAX-SPE cartridges including short-chain (C4–C6) compound. These cartridges have been applied in many monitoring studies [39, 57, 58]. For longer-chain PFCs, less polar phases (C18 and Oasis HLB) may be applied [59–61]. Non-ionic PFCs may be extracted from the matrix by non-polar media (C18 SPE). Moderate polar media (Oasis HLB and Oasis WAX-SPE) have also been applied for extraction of non-ionic PFCs. However, one of the critical points in PFCs’ analysis is background contamination in the analytical blanks [62–64]. One known source of procedural contamination is contact with laboratory materials made of, or containing, fluoropolymers [54, 62]. Water samples may be filtered [54, 65] to separate solids from the liquid phase. However, filtration can result in losses by adsorption of PFCs on the filters, or on the contrary levels can increase by contamination from the filters, as was found by Schultz et al. for fibre, nylon, cellulose acetate and polyethersulphone filters [54]. They applied centrifugation as an alternative for separating theliquid from the solids.

Controversial studies reported the cross-contamination of samples during PFCs’ analysis using different SPE cartridges. Yamashita et al. [66] examined the source of blank contamination at various different steps, including sample collection, extraction and treatment of samples. PFOS and PFOA contamination in the SPE cartridges, OASIS HLB and Sep-Pak (C18), was evaluated. Both SPE cartridges were a cause of contamination by PFOS and PFOA. However, higher concentrations of PFOS and PFOA were reported for Sep-Pak cartridges. In the case of the Oasis HLB, PFOS, PFOA, PFHxS and PFBS were detected, but at lower concentrations than those found in the Sep-Pak cartridges. On the other hand, Taniyashu et al. [67] evaluated Oasis HLB and Oasis WAX columns for the extraction of PFCs. In this study, few target PFCs were detected in procedural blanks at a few pg/L in the final extract. However, PFOA, PFDA and PFUnA were still found at relatively high concentrations. In general, the performance of these columns was comparable. Recoveries were good (70–100%) for most compounds, but for short-chain PFCAs recoveries using Oasis WAX-SPE cartridges were higher. Losses due to evaporation during analysis and adsorption to the polypropylene sample container surface as discussed earlier were suggested causes for the lower recoveries.

2.2.2 Solid Samples

Sample pre-treatment in complex matrices as sludge or sediments is required to minimize matrix effects. It is important to be sure that transformation processes do not occur during this process. For example, the hydrolysis of fluorotelomer compounds to fluorotelomer alcohol during solvent extraction of soils was reported by Dasu et al. [68]. General approaches can be summarized in four different pre-treatments based on solid–liquid extraction or supernatant liquid extraction:

  1. 1.

    Ionic-pair extraction

  2. 2.

    Solid–acid liquid extraction

  3. 3.

    Solid–alkaline liquid extraction

  4. 4.

    Solid–liquid extraction

Table 1 summarizes the pre-treatments that can be found in some published works. The first four extraction procedures are based on the extraction of lyophilized or dried solid. The last one corresponds to non-dried solid extraction procedures.

2.2.2.1 Ionic-Pair Extraction

Ylinen et al. developed an ion-pair extraction procedure employing tetrabutyl ammonium (TBA) counter ions for the determination of PFOA in plasma and urine in combination with gas chromatography (GC) flame ionization detection (FID). Later on Hansen et al. [69] improved the sensitivity of the ion-pair extraction approach using methyl tertiary butyl ether (MTBE) and by inclusion of a filtration step to remove solids from the extract making it amenable for liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS) determination. Ion-pair extraction procedure has been the basis of several procedures for biota [70, 71] and food samples [72]. However, this method has shown some limitations, such as (1) co-extraction of lipids and other matrix constituents, (2) the absence of a clean-up step to overcome the effects of matrix compounds and (3) the wide variety of recoveries observed, typically ranging from <50% to >200%. Sludge samples usually contain high amounts of interferences and, once the sample is reconstituted, a filtration step previous to the analysis [73] or an additional clean-up step by Envicarb cartridges [74] is necessary. This pre-treatment improves the limits of quantification (MLOQ). Using this last approach, Eon et al. obtained recovery rates between 17% and 105% [73].

2.2.2.2 Solid–Acid Liquid Extraction

Current methods are based on the procedure described by Higgins et al. [75]. This method is based on the extraction of dried soils using acetic acid 1% at 60°C in an ultrasonic bath. Then, the mixture is centrifuged and the supernatant collected. The extraction of the solid residue is repeated twice; the extracts are combined and after centrifugation are cleaned up using SPE. Sometimes, a filtration step could be also required in order to eliminate the non-dissolved matter [44, 76]. This methodology has been used in the sludge characterization of different PFCs including PFCAs, PFSAs and fluorinated sulphonamides [43, 44, 52, 75, 76]. The recoveries were in most of the cases between 40% and 119%. In general, this treatment allows better quality parameters than ionic-pair extraction.

2.2.2.3 Solid–Alkaline Liquid Extraction

This procedure has been used for the analysis of acids, sulphates and sulphonamides fluorinated compounds by different authors. In general, alkaline methanol (with NaOH or NH4OH) is used. After the alkaline treatment, a neutralization step with HCl (if NaOH is used) [77] or acetic acid at 1% (if the alkaline digestion has been carried out by NH4OH) is needed [78]. Some more details can be found in Table 1. As a last clean-up step, SPE or dispersive-SPE (EnviCarb) is performed. The alkaline extraction allowed method limit of detection (MLOD) in the range of 0.14–1.43 ng/g dw in sludge [78] and MLOQ between 1.8 and 6.8 ng/g dw [77]. The recoveries ranged from 62% to 104% with RSDs between 2% and 7% (Table 1).

2.2.2.4 Solid–Liquid Extraction

This is the most commonly used methodology in the extraction of non-volatile PFCs from solid matrices due the facility and simplicity of the extraction and the required solvents. Some published works performed the extraction by methanol or acetonitrile with a shaker [79], in an ultrasonic bath with temperature between 40°C and 60°C along 20–30 min [34, 55, 56, 80, 81] or at 50°C along 2–7 days [82], depending on the analysed compounds.

In some other works, the extraction is performed using pressurized liquid extraction (PLE) by accelerated solvent extractor (ASE) [35, 83, 84]. The use of PLE instead of Soxhlet or hot vapour extraction allows decrease in the time of extraction due to the use of high pressure. The most frequent solvent extractor is methanol using a different number of cycles, temperatures (between 70°C and 100°C) and pressures (between 1,500 and 2,000 psi). After extraction, a clean-up process is in general required to decrease the matrix effects [35, 74], using SPE (C18, anionic exchange (WAX)) or by using Envicarb graphitized carbon adsorbent. The methodology allows, in general, a MLOQ between 0.05 and 2.77 ng/g of PFCs, recoveries ranging from 57% to 120% in most of the cases and 2–30% of RSD (see Table 1).

2.3 Instrumental Analysis

During this step is important minimize possible sources of contamination due Teflon or PVDF tubing other materials as PEEK. Other precaution that can be considered is the addition of extra chromatographic column prior to the injector in order to delay PFCs peaks due to system contamination [79].

Liquid chromatography-mass spectrometry (LC-MS) or LC-MS/MS has been in general the techniques of choice for the analysis of PFCs. Therein detailed information about the main experimental conditions used for analysis such as LC-MS/MS precursor-product ion transitions.

LC separation of PFCs has been mainly carried out with C18 and C8 columns. In spite of the wide use of RP-C18 columns for PFCs’ analysis, the interference producing the enhancement of spectral signal has been reported. RP columns with shorter alkyl chain bonded phases (e.g., C8, C6, phenyl and phenylhexyl) also separated the branch isomers, but to a lesser extent. To minimize the separation of branched isomers, the authors increased the LC column temperature to 35°C or 40°C [85, 86]. Taniyasu et al. [87] explored the chromatographic properties and separation of short-chain PFAs on RP-C18 and ion-exchange columns. The results showed that using RP-C18, the peaks of PFPrA and PFEtS were broad and not adequately resolved, whereas that of TFA was not retained in the analytical column eluting with the solvent front. This suggested that RP columns are not suitable for the analysis of short-chain PFAs, especially TFA. As a proper alternative, ion-exchange columns have superior retention properties for more hydrophilic substances enabling the analysis of short-chain PFCAs, TFA, PFPrA, PFBA, PFEtS, PFPrS and PFBS together with several long-chain PFCAs, in water samples.

Due to the complexity of environmental samples, it is possible that the co-occurrence of certain compounds can interfere the analyte determination. This problem has been partially solved using LC–MS/MS. However, certain interferences can affect the analyte ionization producing ion suppression or enhancement [35, 74]. The use of labelled PFCs during analytical process (surrogates or internal standards) helps to assess and normalize these instrumental effects.

LC-MS/MS performed using triple quadrupole mass spectrometer (QqQ) combined with multiple selected reaction monitoring (SRM) is one of the more widely applied analyzer [34, 43, 44, 73–78, 80, 81, 84, 88], as well as, to be one of the better suited for quantification of PFCs. Nowadays the performance of hybrid quadrupole linear ion trap (QqLit) [35] or hybrid quadrupole time of flight (QTOF) [52] has been also considered for trace quantification of PFCs. PFCs contain carboxylic, sulphonic, hydroxy or sulphonamide group. They have acidic properties and can therefore dissociate. Consequently, electrospray ionization in the negative mode (ESI(−)) has been the interface most widely used for the analysis of anionic perfluorinated surfactants, allowing limits of detection in the pg to ng/g range, although atmospheric pressure chemical ionization (APCI) and positive ESI have been employed for specific PFCs’ analysis [83]. In addition, ESI has been optimized for the determination of neutral compounds such as the sulphonamides FOSA, Et-FOSA and t-Bu-PFOS. The use of atmospheric pressure photoionization (APPI) has been explored in few works [68–70]. Takino et al. [68] found as the main advantage of this technology the absence of matrix effects, but the limits of detection were considerably higher than those obtained by LC–ESI-MS/MS.

Pseudomolecular ions are formed such as [M−K]− for PFOS (m/z 499), [M−H]− for PFOA (m/z 413) and FOSA (m/z 498), which are generally selected as precursor ions for MS2 experiments using ion trap and a triple quadrupole instruments. Berger et al. [89] have presented a comparison between IT, QqQ and TOF instruments. Tandem mass spectrometry showed excellent specificity, but the background is eliminated by the instrument, and thus it cannot be visualized. Applying TOF-MS gives an estimation of the amount of matrix left in the extract, which could impair the ionization performance and the high mass resolution of the TOF-MS instrument offers excellent specificity for PFCs’ identification after a crude sample injection. Recently, the analytical suitability of three different LC–MS/MS systems: QqQ, conventional 3D-IT and QqLIT, to determine trace levels of PFCs in fish and shellfish was compared [90]. In this study, the accuracy was similar in the three systems, with recoveries always over 70%. Precision was better for the QqLIT and QqQ systems (7–15%) than for the IT system (10–17%). The QqLIT (working in SRM mode) and QqQ systems offered a linear dynamic range of at least three orders of magnitude, whereas that of the IT system was two orders of magnitude. The main advantage of QqLIT system is the high sensitivity, at least 20-fold higher than the QqQ system. Another advantage of QqLIT systems is the possibility to use enhanced product ion (EPI) mode and MS3 modes in combination of SRM mode for confirmatory purposes of target analytes in complex matrices.

Other instrumental tools have been employed by Liu et al. [82] in the study of aerobic biodegradation of [14C] 6:2 PFTOH in a flow-through soil incubation system. The instrumental analysis was carried out by radioactivity and liquid chromatography/accurate radioisotope counting (LC/ARC).

3 Environmental Fate of PFCs

Partitioning and reactivity properties are important to understand and model the environmental behaviour of PFCs. Just during the recent years, it has been initiated the study of the reactivity properties of these compounds, and in addition should be pointed out that some data continue being contradictory. A starting point to study the mechanistic properties of PFCs, as well as, to evaluate and assess properties of new emerging PFCs is the study of selected physicochemical properties.

The perfluoro alkyl sulphonates (PFASs) and PFCAs are strong acids that exist in equilibrium between the neutral form and the anionic form. In general, both the anionic and neutral forms, as it happens with PFOA, are soluble in water. Although the Henry’s law constant values suggests partitioning to air for the neutral, protonated form, predicting the amount that partitions into air is complicated because there is uncertainty over the degree to which carboxylic and sulphonic acids partition from the water to atmosphere. The uncertainty arises with regard to the value of the acid dissociation constant (i.e. pK a), or the fraction of the acid form present at environmentally relevant pH. PFCA and PFAS have been detected in air, water and soil samples collected throughout the world. The oceans have been suggested as the final sink and route of transport for perfluorinated carboxylic and sulphonic acids, where they have been detected on the surface and at depths over 1,000 m [91].

Some PFAS/PFCA have the potential for long-range environmental transport (LRET) by a combination of dissolved-phase ocean and gas-phase atmospheric transport; however, determining which is the predominant transport pathway is complicated by the uncertainty over water to atmosphere partitioning. Furthermore, there is evidence that transport and subsequent oxidation of volatile alcohol PFAS/PFAC precursors may contribute to the levels of PFAS/PFCA in the environment.

The evaluation of PFCs in remote areas such as the Antarctica peninsula is one of the very few forms of evaluation of LRET. The global fate of POPs is associated with different biogeochemical cycles and geophysical drivers. The occurrence of PFCs into remote areas such as the Antarctica could be partially explained by the theory of cold condensation, concerning the chemical movements or chemical transformations from sources under the impact of environmental forces, such as temperature, and interaction with other environmental compartments (soil, oceans, etc.) [92]. In addition, the physicochemical characteristics of PFCs should be considered, since these properties dictate their environmental behaviour [5].

Just few previous studies have reported the presence of PFCs in different biota samples from the Antarctica continent [11, 70, 93], whereas this information could be of importance to establish Global PFCs’ distribution and also the basis of LRET of these compounds.

Different studies have shown that PFTOH can be degraded by microorganisms and by abiotic processes. 8:2 FTOH and FTOH of other chain lengths, and related chemicals in mixed microbial cultures, activated sludge and soil systems have been shown to be easily degraded to form PFOA and related perfluorinated acids. Some studies have also shown that –CF2- groups can be mineralized, forming shorter-chain perfluoro acids. If FTOH are absorbed from ingestion, inhalation, dermal or ocular exposure or formed in vivo by from other compounds they can be metabolized by mammals and other organisms to form perfluorinated acids and other fluorinated compounds. FTOH can be degraded by abiotic processes in water and air to produce PFCA and various intermediates. FTOH are fairly volatile. Based on atmospheric half-lives determined in chamber studies, FTOH can be transported globally. Deposition or degradation in areas far from the source can result in PFCA contamination in high latitudes and other remote locations and contribute to global background levels of PFCA and PFAS.

Therefore, two mechanisms should be considered to explain the LRET capabilities of PFCs. The first suggests atmospheric distribution of neutral, volatile compounds (flyers), such as fluorotelomer alcohols and perfluorinated sulphonamido alcohols. Flyer compounds are susceptible to suffer atmospheric long-range transport because of their partitioning properties (log Kaw values estimated between 0 and 1 and log Kow around 5), which indicate that these classes of chemicals can be classified as flyers according to the Globo-POP model [5, 92]. This is also in agreement with the findings of Dreyer et al. (2009). Then, after their transport and cold condensation, these flyer compounds can biodegrade as it has already been indicated in previous studies [80] or suffer in situ oxidation to form ionic PFCs [94].

The second mechanism is related to the properties of ionic PFCs (negligible vapour pressure, water solubility and moderate sorption to solids), which predicts their accumulation in surface waters (swimmers) [95, 96]. Some studies have evaluated the influence of these mechanisms and have been revealed that the dominant phenomenon is the hydrospheric transport for PFOS, PFOA and PFNA [97]. For example, fluorotelomer alcohols have short atmospheric lifetimes in the order of 10–20 days [98]. The geographical isolation of Antarctica combined with both, short atmospheric lifetimes of fluorotelomer alcohols and the low yield of the oxidation pathway, significantly reduces the potential for effective atmospheric delivery to the Antarctic continent. Therefore, atmospheric input of flyer PFCs to the Antarctica is principally a function of rapid and direct delivery of contaminated wind masses.

Concerning to degradation processes, some recent data show that perfluorooctane sulphonyl fluoride (POSF) and its derivatives can be degraded under environmental conditions to form perfluoroalkyl sulphonates and carboxylic acids. Reaction of POSF (CF3(CF2) n -SO2F) with methyl or ethyl amines is used to produce N-ethyl or N-methyl perfluorooctane sulphonamidoethanols. Similar reactions are used to make shorter- and longer-chain analogues to POSF and POSF derivatives. FOSE compounds, such as N-methyl and N-ethyl FOSEs can be degraded though a series of intermediates to form both PFCAs and perfluoroalkyl sulphonates. Other chemical intermediates produce other FOSA derivatives, including phosphate esters, fatty acids esters, silanes, carboxylates and polymers with acrylate, urethane and other linkages.

Longer- and shorter-chain perfluoro sulphonyl derivatives have also been produced intentionally and as unintended reaction products. Based on existing data from the open literature and CBI data, it is expected that most, if not all, of these POSF and other chain length sulphonyl fluorides and their derivatives will be degraded to carboxylic acids and/or sulphonate over time. Most of these compounds will have environmental and metabolism half-lives of weeks to months. Some will be degraded faster and some will degrade more slowly, but all will eventually be degraded.

Very little data are available on the behaviour of other perfluorochemicals in the environment and in vivo but the existing data suggest that they will also be degraded to form PFAC. For example, recent studies have shown that ingested mono- and di-polyfluoroalkyl phosphates (PAPs) can be degraded in rats to form PFOA and other PFAC in the body. They can also be degraded by microbial processes in soil and wastewater to form perfluorinated acids [99].

A limited number of studies on the degradation of fluorotelomer-based polymers have been submitted, but some studies have shown that fluorotelomer-based polymers are subject to hydrolysis, photolysis and biodegradation to some extent. Studies have shown half-lives of a few days to hundreds of years. In addition, preliminary research on degradation of fluorotelomers has shown that some urethanes and acrylates biodegrade; however, half-lives and kinetics of the fluorotelomers are not yet well defined.

These studies have shown that the perfluorinated portion of some polymers is released as the polymer is degraded by microbial or abiotic processes to form telomer alcohols or other intermediates and that they eventually form PFCA. Polymers based on POSF and other chain length chemistries show similar degradation rates and release intermediates that further degrade to form perfluorinated acids and sulphonates. Studies have shown that some polymers can undergo indirect photolysis in soil and in aquatic systems and be degraded with half-lives of days to several years.

4 Occurrence of PFCs: The Llobregat River as Case Study

During the last decade, an important work has been carried out in order to assess the occurrence of PFCs in the aquatic environment. In this section, the occurrence of PFCs in river basins will be revised taken as central example the Llobregat River, as example of a Mediterranean river suffering a high industrial pressure.

Different examples of works assessing the occurrence of PFCs in river water are summarized in Table 2.

Table 2 PFCs – published works
Full size table

Due to their persistency and wide use in the past PFOS and PFOA are the best-studied compounds. Currently, the use and production of these two compounds are almost stopped; however, due to their high resistance to degradation and because they are the end products of other PFCs in use, PFOS and PFOA continue being present in high concentrations in surface waters and sediments. Due to their physicochemical characteristics PFOA is mainly found in water, whereas PFOS is retained in higher proportion into the sediments.

4.1 Waste Water Treatment Plants

Wastewater treatment plants (WWTPs) are major sources of PFCs to the natural environment, through treated effluents and also when contaminated sludge is used in agricultural lands. The first work assessing the occurrence of PFCs in sewage sludge was performed by Higgins et al. [75] who studied the occurrence PFCs in sediments and sludge from WWTPs in San Francisco (1998–2004) [75]. Concentrations from 1.2 to 2,610 ng/g dw, were reported, being PFOS the compound at higher concentrations. Following this work, several works have been devoted to assess the content of PFCs in sewage sludge. PFOS is exceptionally stable chemical compound that is highly resistant to degradation and due to its higher partition coefficient in comparison with other PFCs, especially PFCAs, is present in high concentrations in sewage sludge worldwide. Furthermore, PFOS is the end-point of the degradation of fluorochemicals used in a variety of industrial and commercial applications. Compounds that may be transformed to PFOS are 2-(N-ethyl perfluorooctane sulphonamido) acetic acid (N-EtFOSAA) and 2-(N-methyl perfluorooctane sulphonamido) acetic acid (N-MeFOSAA), among others. These compounds have been also identified in general in WWTP sewage sludge at levels often exceeding PFOS. This could indicate that part of the PFOS is directly generated in the degradation process of related products.

Llorca et al. [35] investigated the presence of PFCs in sewage sludge from five WWTPs along the Llobregat River. The results showed that PFCs longer than 10 C chains were at lower ng/g concentration levels, or below. In general, the concentrations of perfluorocarboxylic acids were ranging from 0.4 to 30.3 ng/g. PFOA, PFNA, PFDA and PFDoA were present in all the samples at concentrations higher than 1.0 μg/kg. These high concentrations were in agreement with other works. For example, Zhang et al. [100], Guo et al. [101], Li et al. [55] or Ma et al. [102]. On the other hand, in most of the samples the long-chain acidic compounds were not detected and just PFOA was found to be in high amounts, but this concentration can be associated with the biodegradation of other long-chain congeners currently in use [103, 104]. This predominance of shorter C chains is supported by Ma et al. [78]. The authors found a dominance of even-chain length PFCAs in all of the WWTP sludge samples investigated. It is suggested that a strong aerobic degradation of fluorotelomer alcohols in WWTPs ends in shorter fluorinated compounds. In addition, developing substitute materials to replace long-chain PFCs, or new processes to eliminate their presence as impurities in other products, has been a significant technical challenge. There has been considerable progress in the development and introduction of substitutes and alternatives. Many substitutes are shorter-chain compounds that still provide the needed functionality, but lack the bioaccumulation potential of the long-chain PFCs. In this sense, considerable amounts of these products can also reach sludge of WWTP, partially contributing to these concentrations. In spite of the lack of data reporting the profile of PFCs present in sewage sludge during the past, available data seem to show a strong decrease in the presence of long-chain PFCs, and at the same time an increase in short C-chain compounds. In the studies of Llorca et al. in the sewage sludge of WWTP discharging into the Llobregat River, FOSA was another of the more frequently found compounds, with concentrations ranging from 0.3 to 10.7 μg/kg. There was not found perfluorosulphonates at higher levels than MLOQ with an exception of PFOS, which was detected at concentrations ranging from 53 to 121 ng/g, being the compound that was present at higher levels, as it was expected. Picó et al. within the Framework of the SCARCE project also analysed sewage sludge from the WWTP of Igualada in the Anoia tributary of Llobregat and also found PFOS at concentration as high as 1,790 ng/g. Zhou et al. [30] reported the sorption of PFCs on the heterogeneous protein composition of activated sludge and the different sorption kinetics according to their carbon chain length and different functional groups [105], which could explain the high concentrations of PFOS found by Llorca et al. [35] and other authors. There is a general agreement among results in sewage sludge were the concentrations of PFOS is in general three to ten times higher than the concentrations of PFOA. This difference could be associated with the different sorption kinetics in function the different functional groups, in agreement with Zhou et al. [105].

The lack of total elimination of PFCs in wastewater treatments has been proved, and many works have reported high concentrations of PFCs in treated effluents, being therefore one of the main inputs to receiving waters.

PFCs have been studied in effluent water from Llobregat WWTP located in El Prat de Llobregat by Sanchez-Avila et al. [60]. The sampling point was located near to river surface water sampling point location which was also investigated by the same author. The WWTP effluent values were below 0.77 ng/L for PFBS, below 0.03 ng/L for PFHxS, 14.1 ng/L for PFOS, 61.9 ng/L for PFOA and below 0.06 ng/L for PFNA. Comparing the levels found in surface river water and in effluent water, the concentrations in this last one were higher. These results suggested that PFCs are discharged into the river through WWTPs effluents arriving to potable water treatment plants and, finally, to humans through tap water. However, the WWTPs processes redistribute some of the PFCs from influent water to sludge. The study realized by Zhou et al. [30] showed the favourable sorption of PFCs on the heterogeneous protein composition of activated sludge which could be explain the highest levels found by Llorca et al. [35] in WWTPs from Llobregat River. Zhou et al., in another published work, studied the different sorption kinetics in function of carbon chain length and different functional groups [105]. The calculated distribution coefficients indicate that PFOS had a higher sorption tendency to activated sludge than PFOA. Becker et al. [106] studies supported this last result. The authors showed that, in WWTP, the calculated mass flow of PFOA was fully discharged into the river while about half of PFOS was retained in the sewage sludge. The study carried out by Pico et al. in the WWTP of Anoia also showed that PFOS was accumulated in the sludge whereas PFCAs including PFHpA, PFOA, PFNA and PFDA were mostly in water.

4.2 Surface Waters

In general, concentration gradients can vary up to several orders of magnitude between different areas along the same river or lake, reflecting in general proximity to known industrial sources and WWTPs concentrated near populated regions. For example, different Japanese works found PFOS and PFOA in surface river samples with concentrations in the range from 0.30 to 157 ng/L for PFOS, or between 1.6 and 104 ng/L also for PFOS and 3.8–311 ng/L for PFOA [65]. The study carried out by Skutlared et al. [38] in the Ruhr River in Europe should be mentioned. In this study the occurrence of 12 PFCs was assessed including PFBA, PFPeA, PFHxA, PFHpA, PFOA, PFBS and PFOS. The results showed extremely high concentrations of some compounds. PFOA was present at concentrations till 33,900 ng/L in the Moehne River, and the authors found the main source of contamination in an agricultural area near Brilon-Scharfenberg. In addition, it was proved that this source of contamination leads to the consecutive pollution of Lake Moehn, the Ruhr River and corresponding drinking waters. In another example in China Wang et al. [107] studied the environment around a manufacturing facility. The authors also observed a decreasing trend of the PFOS, PFOA and PFHxS concentrations in soils, water and chicken eggs with the increased distance to the production factory, indicating the production site to be the primary source of PFCs in the region.

Regarding the occurrence of PFCs in the Llobregat River, very few works have assessed the content of PFCs. In 2009 Sanchez-Avila et al. [60] investigated the levels of PFBS, PFHxS, PFOS, PFOA, PFNA, in surface waters from Catalonia, including a sample from an industrialized area of the Llobregat River. In this sampling site, PFOS and PFOA were the compounds found at higher concentrations, but should be mentioned that all the compounds investigated were also found at quantifiable concentrations: 0.88 ng/L for PFBS, 0.64 ng/L for PFHxS, 9.13 ng/L for PFOS, 9.63 ng/L for PFOA and 1.62 ng/L for PFNA. The concentration levels reported in this study were comparable with the data reported for other river waters from industrialized areas. However, the profile of compounds was quite different than the one found in other European countries, such as Germany, where the occurrence of PFOA in surface water is generally found in higher concentrations than PFOS.

Recently, under the frame of the SCARCE project Picó et al. have investigated the presence of 21 PFCs in different sampling sites along the Llobregat River during 2 sampling campaigns. In this case, PFOA was one of the more frequently found compounds. However, PFOS was found at higher concentrations up to 2.7 μg/L. The presence of these compounds showed an important spatial distribution. In agreement with data reported in other European rivers. However, the Llobregat River is affected in a great manner by climate episodes, such as flows, which can re suspends contaminants in general contained in the sediments. The higher concentrations were found near the mouth as it was expected, because it corresponds to a heavily populated and industrialized area. The compound found in higher concentration was PFHpA with concentrations around 30 ng/L in more polluted samples. In general, more frequent compounds and also those in higher concentration were short-chain compounds, indicating a tendency to replace more persistent long-chain PFCs by new short-chain ones.

4.3 Sediments

There are few available data reporting the levels of PFCs in sediment samples from Llobregat River basin within the SCARCE project. Different PFCAs including PFBA, PFPeA, PFHxA, PFHpA, PFOA, PFNA, PFDA, PFUnA and PFDoA were detected in sediment samples but at concentrations up to 7.5 ng/g. Although the number of PFSAs is restricted to PFBS and PFOS, they reach concentrations up to 11 ng/g. However, other related works indicated that these compounds are distributed between the two compartments, water and sediment. An example was published by Picó et al. [108]. The authors investigated different points from l’Albufera de Valencia (Spain) assessing the presence of PFCs between MLOD and 10.9 ng/g where the highest values corresponded to PFOA and PFOS. In the same study, the levels of PFCs in surface waters were assessed showing the distribution of these compounds between water and sediments. Another study performed by Gómez et al. [59] focused the investigation on the analysis of different sediments from Cantabrian Sea samples (North of Spain). The results of PFCs were in most of the cases below MLOD in sediment river samples.

4.4 Drinking Water

In order to elucidate the possible source of PFCs in drinking water, some authors have compared the levels found in the catchment sites in surface river or lakes and in tap water, and it was showed that certain relations can be established. This reveals the ineffective removal of PFCs by the purification processes performed at water purification plants [38, 109]. For example, Skutlarek et al. [38] showed extremely high concentrations for PFOA in tap water in the zone of Ruhr area, which is in agreement with the concentrations found in environmental surface waters of the same areas. In another example PFOS and PFOA were measured at concentration levels around 9 and 3 ng/L, respectively, in The Lake Maggiore (Switzerland), and the results of the analysis of drinking water produced from the lake gave almost identical results revealing the poor performance of sand filtration and chlorination which is applied by the local waterworks [109].

Similar conclusion can be extracted from the work carried out by Takagi et al. [110], who studied the presence of PFCs in different waters including raw water and drinking water from Japan. In raw water the results for PFOS and PFOA were between 0.26–22 ng/L and 5.2–92 ng/L, respectively, and in tap water similar results were also obtained in the ranges between 0.16–22 ng/L and 2.3–84 ng/L for PFOS and PFOA, respectively. In addition, other PFCs that in general are not assessed and that are generated during the water treatment processes should be considered [48, 103, 111, 112].

Regarding the study of the Llobregat River, the occurrence of PFCs in drinking water produced from Llobregat River has been assessed by Llorca et al. under the frame of the SCARCE project. Among the 21 compounds considered in this study, the presence of PFBA, PFPeA, PFOA, PFNA, PFBS, PFHxS and PFOS has been found in final drinking water at concentrations between 0.07 and 35 ng/L. The more polluted drinking waters were those corresponding to catchment locations in more industrialized and polluted areas, such as Barcelona city, as it was expected.

These results were in agreement with a previous work carried out by Ericson et al. [39, 57] who studied the presence of these contaminants in drinking water from different areas of Catalonia. In the study performed by Ericson et al., the concentrations of PFCs were in the range between 0.02 and 69 ng/L in tap waters.

5 Future Trends

In recent years, the research has been focused on the study of degradation mechanisms of fluorochemicals as PFOA and PFOS. Cheng et al. found that sonolysis (sonochemical) is able to decompose PFOS and PFOA present in groundwater beneath a landfill following a pseudo first-order rate constant [113].

Because of the poor degradability of these ones in the treatment facilities, these compounds are discharged directly into the rivers. Once these recalcitrant compounds reach the environment, they can arrive to the drinking water through the drinking water prepared from surface water [114] or enter into food chain through the irrigation of agricultural lands with contaminated waters [115] or by the bioaccumulation, and consequent biomagnification, through the food chain [3]. The study of these compounds in river waters as well in flora and fauna is of high importance since they are not regulated and should be under control in order to elucidate possible focuses of PFCs into, for example, the Llobregat River [116].