PFCA and PFSA in consumer products
PFASs are widely present in the investigated consumer products. The broad findings are summarised in Table 5. The only products with no or negligible PFAS concentrations tested were cleaners and wood glue. None of these samples exceeded an existing limit. In the majority of samples throughout all other sample groups, however, we found medium PFAS levels of up to 100 μg/kg or μg/m2 per substance. High concentrations over 100 μg/kg were reached in outdoor textiles, gloves, ski waxes and archived food contact paper samples. The degree of contamination does not seem to depend on quality level or price category of the tested products but was rather randomly distributed. Interestingly, a broad range of PFASs was found, rather than few distinctive substances, and does hence, not indicate a shift from the banned PFOS towards other PFASs conferring similar properties. Since there is no sample with direct use of PFOS, it is rather likely, that a contamination from unintentional production or cross-contamination occurred. The versatile pattern of PFCA with different chain lengths could rather indicate an unspecific formation route via harsh physical conditions like heat or UV exposure, i.e. hydrolysis and subsequent oxidation of FTOH from polymeric structures.
In regard of customer safety, food contact materials play a crucial role as intake pathway, since hydrophobic contaminations, such as PFASs, are discussed to migrate into the fat- and protein-containing food matrix, especially in the presence of compounds like phospholipids or other surface active compounds (Prieto et al. 2004; Still et al. 2013; Trier et al. 2011). Other critical consumer products are any kind of impregnation or nanosprays, since during intended usage (the generation of aerosols), it can hardly be avoided to inhale at least some of the product. Clothing may, at least in part, play a role considering the skin as an intake route (compare Trudel et al.); moreover, for children’s clothing and in particular children’s gloves, the oral route becomes prominent in terms of a textile-to-mouth contact (Trudel et al. 2008). Similar considerations render carpets critical instances because of a hand-to-mouth route for children, especially toddlers and young children, as well as the formation and inhalation of dust from carpets for all customer ages. Nevertheless, many samples still do not match with existing PFOS regulation and exceed levels of 1 μg/m2 in coated materials or 10 μg/kg for substances and formulations, especially of the product groups outdoor textiles, food contact papers and carpets. Such contaminations are avoidable, if care is taken that not only the used raw materials (e.g. fabrics in the case of textiles) but also any associated materials and contact surfaces are free of it (e.g. in sewing plants), and therefore, the routes of the chemicals need to be traced to minimise the contact of consumers with PFOS and other PFASs. Moreover, this may indicate an exposure risk for staff within the value added chain where the contamination may occur.
The regulation of PFOS use may have caused shifts of the found PFAS spectra to alternative molecules. Albeit exhibiting different characteristics, in this study, the use was mainly shifted to, or remained at PFCA with shorter chain lengths in some carpets (loads >25 μg/m2 for PFBA); in the same samples, the PFOS load was inevitably lower (<2 μg/m2) compared to carpets with high PFOS loads. Next to this, we found odd-numbered substances like PFPA or PFNA to be major contributors in some samples, e.g. paper-based FCM and textiles (see Table 2).
The data do not allow for a detailed estimation of the sources for the versatile PFASs. In case of the textiles, it can be expected that the suppliers of raw materials avoid any contact at least to the banned PFOS. That we, nevertheless, found it in samples of the final textile products may be due to a contamination that occurred during the multi-step fabrication of the final products. An optimisation of the coating process, a monitoring of the raw materials and a critical check of all technical and accessory agents could reduce the contamination of consumer products with PFASs. In the past, PFOA was used as a process agent for fluoropolymers like PTFE and PVDF, and therefore, PFOA levels may be a result of using these polymers as a fibre material in our textiles. However, in the analysed textile samples, PFHxA and PFDA were found besides PFOA in concentrations up to 18.8 μg/kg. Both are chemicals, which exhibit similar characteristics compared to PFOA with chain lengths of −2 and +2, respectively. Their presence and comparable ratios of 10:2 and 6:2 FTOH, however, indicate that FTOH moieties in PAPS or polymeric surfactants may be the initial source of PFHxA, PFDA and probably PFOA.
By covalently attaching them to a fluorine-free polymer skeletal, FTOH are the monomeric basis for the final polymeric structures, e.g. on textiles and other consumer products. The found FTOH levels may thus reflect the success of binding reactions of monomers to the sole structures and thus represent technical process remnants (Dinglasan-Panlilio and Mabury 2006).
With respect to food contact paper baking forms, a temporal trend in terms of a ‘before and after 2010’ scenario was observed. The PFOS limit of 1.0 μg/m2 is exceeded only rarely and moderately for recently collected samples. In archived samples of the same product group, however, massive loads with perfluoroalkyl carboxylic acids were found. Such archived baking forms (purchased before 2010), e.g. popularly used for muffins, frequently show very high contaminations with PFASs (e.g. PFNA and PFDA up to 478.2 and 658.1 μg/kg, respectively). However, recently bought samples showed much lower values of max 13.5 and 18.0 μg/kg for PFOA and PFPA. Notably, in all recent food contact paper samples (purchased in 2010), PFPA was the main contributor to total PFAS load. In contrast to our findings, a recent study of food packaging papers from the Greek market found only a minor contamination of these products with selected PFASs (Zafeiraki et al. 2014); however, PFPA was not in their spectrum of monitored substances. Also, in other samples, PFPA was one of the most abundant and most frequently detected PFASs.
A dominant role of PFPA was not clear prior to this study, and PFPS was not part of the initially selected PFAS spectrum. Since PFBA and PFBS often occur with similar patterns, PFPS should be included in the substance spectra for future studies.
FTOH in consumer products
The FTOH load of the investigated consumer products differed between product groups and inside the groups. In addition, considerable differences between the levels of PFAAs and FTOHs were observed. While the PFAA contents in the examined cleaning agents were negligible (<0.5 μg/kg, except for PFOS in one case (1.1 μg/kg)), the FTOH levels of cleaning agents were comparably high (up to 73,000 μg/kg 8:2 FTOH). In outdoor textiles, the FTOH levels topped 180 μg/m2 with PFOS levels of 10 μg/m2. This fact makes it hard to draw a general conclusion but requires exposure estimation from case to case.
In general, 6:2, 8:2 and 10:2 FTOH were identified in FTOH positive samples with 8:2 being the dominating congener. Only in mixed paper samples different patterns were observed, possibly due to the application of longer chain FTOH side groups in the specific surfactants used.
FTOH loads within the product groups were not distributed homogeneously, i.e. the concentrations differed from one sample to another. In the case of outdoor textiles, FTOH levels varied by a factor of about 25.
Highest FTOH levels were found in impregnating sprays (up to 719,000 μg/kg 8:2 FTOH), which is consistent with the assumption that polymeric surfactants used in impregnating applications are a major source of FTOH (Dinglasan-Panlilio and Mabury 2006). Thus, not using such sprays in closed rooms is a major and serious task to prevent direct customer exposure.
Outdoor textiles are another product group exhibiting high levels of FTOH (up to 380 μg/m2 8:2 FTOH). These have been identified as major sources of FTOH in indoor environments as FTOH are released from these products (Schlummer et al. 2013; Langer et al. 2010).
Initial studies (Dinglasan-Panlilio and Mabury 2006) identified the presence of residual unbound FTOHs of varying chain lengths (C6–C14) in several commercially available and industrially applied polymeric and surfactant materials, normally used for impregnating of leather surfaces, papers or other materials. The authors concluded ‘that residual alcohols, left unreacted and unbound from the manufacturing process of fluorinated polymers and surfactants, could be a significant source of the polyfluorinated telomer alcohols and sulfonamides released into the environment’. The examined fluorinated materials contained 0.04–3.8 % residual-free fluorinated alcohols on an applied fluorinated alcohol basis (dry mass basis). This study suggests that elimination or reduction of these residual alcohols from all marketed fluorinated polymers and fluorosurfactants is the key in reducing the prevalence of perfluoroalkyl acids formed in the environment.
Ski waxes showed nearly the complete PFAS spectrum, and other samples were dominated by individual compounds. In case of FTOH, normally 6:2, 8:2 and 10:2 FTOH were identified with 8:2 being the dominating congener. Accordingly, Plassmann and Berger reported the finding of several PFAS in snow and soil samples from ski areas (Plassmann and Berger 2013). Adding to this context, a massive PFAS exposure of people professionally working with ski waxes has been reported (Nilsson et al. 2010). Moreover, degradation of FTOH to PFCA in humans was also shown using ski waxers by Nilsson et al., indicating the importance of such studies (Nilsson et al. 2013).
Only in mixed paper samples different patterns were observed, possibly due to the application of longer chain FTOH side groups in the specific surfactants used.
Regarding the co-occurrence of different PFAS subgroups, a significant linkage for FTOH and PFCA, but not for FTOH and PFSA was shown. The load with PFCAs could yet be depending on the post-coating processes of the textiles such as poor polymerisation conditions, insufficient linkage to the basic textile polymeric material (e.g. polyester), or the use of raw materials of suboptimal quality. As technically expected, due to different industrial synthesis routes and no degradation to sulfonic acids, the load with PFOS shows no correlation to the respective loads with FTOH.
Comparison with literature data
In a similar approach, Herzke et al. (2012) analysed 30 consumer products from the Norwegian market and found beneath PFOS and PFBA shorter chained PFAS, such as PFBS, to be a major contributor to the total PFAS load of consumer products (especially in non-stick ware and waterproofing agents, but not in food contact cardboards) (Herzke et al. 2012). On average, they found less individual PFAS compared to our results and comparably high loads only in leather. However, the sample size of two per matrix was relatively small and due to the inhomogeneity of PFAS contamination, highly loaded samples were rarely found in this study, e.g. one aqueous fire-fighting foam and one waterproofing agent had extraordinary PFAS loads of up to 900 mg/L (1286 mg/kg; with an estimated density of 0.7 kg/L) PFHpS and 1.2 mg/L (1.7 μg/kg) PFDoA, respectively. Fire-fighting foams were not subject of our analysis, but in all analysed waterproofing agents (in our study, nanosprays and impregnating sprays), the highest PFCA load was found for PFOA with max 28.9 μg/kg, and max 5.3 μg/kg PFDoA, which fits with the results of Herzke et al., for the majority of samples analysed. Also, for FTOH, the results were comparable, with the exceptions given above. FTOH were also the majorly investigated substance group of a study by Fiedler et al. (2010). Also, here, in 15 samples analysed, among them 9 impregnating agents, the highest values were found for FTOH 8:2 and 10:2 with 52,000 and 32,000 μg/L in impregnating agents, which is in the same range as in this study. PFOS was not detected in any sample, but also, PFOA was found in levels of 100–400 μg/L. High values for PFAS in ski wax were also found by Freberg et al. (2010). Most alarming, corresponding PFAS levels were also found in the blood of customers having professional contact with these products (Freberg et al. 2010; Nilsson et al. 2010).
The widespread use of PFASs in consumer-near products is reported by Huset et al. (2011) who analysed municipal landfill leachates. The authors found PFCA and PFSA levels of up to 2.8 and 2.3 μg/L, respectively, while short-chain PFASs (C4–C7) were more abundant than the corresponding longer chain homologues. Results of our study underline the dominance of short chain PFAS in two product categories of consumer products with significant levels of perfluoroalkyl compounds: leather and carpets and particularly PFPA in paper-based FCM.
A comprehensive overview on PFAS in outdoor textiles is provided in a non-peer-reviewed publication by the interest group Greenpeace e.V. organisation. In this large-scale approach, conditionally, they found a variety of PFAS in qualitatively and quantitatively comparable concentrations to the results of our study (Santen and Kallee 2012). However, they reported no textile exceeding the EU regulation for PFOS. This may be due to the selection of samples, and a role may also play the sampling years for our (2010) and the Greenpeace study (2012). Berger and Herzke (2006) reported high amounts of extractable FTOH (sum of 4:2 to 10:2) in different textiles samples including outdoor textiles. Median levels of their report compare well to FTOH levels reported here.
Implication for risk assessment
Exposure estimations can be optimised when detailed information is available for as many as possible samples. Thus, this and further data can contribute to reliable safety considerations (e.g. Trudel et al. 2008) and point out the importance of a large-scale monitoring approach for PFASs in a broad spectrum of consumer products and food contact materials. An important issue for the risk assessment, however, is the risk of cumulative exposure of all PFASs in sum (Borg et al. 2013), e.g. the 16 monitored PFCA and PFSA sum up to 113.7 μg/kg in children’s gloves, with PFPA being the main contributor at a concentration of 47.7 μg/kg. Borg et al. monitored the levels in blood of respective customer groups, and our data provide information on contact points. In either case (Borg et al. analysed 17 substances), considering only selected compounds for exposure assessment may lead to an underestimation of the risk. Future approaches should therefore target an enhancement of analyte portfolio towards further polyfluoroalkyl and branched substances in order to allow for comprehensive exposure estimation.
Generally, the PFAS exposure of consumers by products like outdoor materials (textiles) needs to be better understood. Markedly, also children’s textiles and carpets are highly loaded with diffuse mixes of PFAS and may contribute to the exposure of children and toddlers. According to the tolerable daily intake (TDI) defined for PFOA (1.5 μg/kg body weight/day), the maximum exposure for a toddler of 10 kg is 15 μg per day (EFSA 2008). With up to 41.0 μg/m2 in outdoor textiles, a significant portion of the TDI may be contributed by these products upon textile-mouth or hand-mouth contact. Up to now, it can only be estimated that these limits are not passed from a toddler wearing gloves via the hand to mouth route for a single compound or even in a cumulative exposure scenario (Trudel et al. 2008). However, these considerations need to be subject of future research.
This study confirmed the presence of PFASs in a wide variety of consumer products including sensitive samples such as children’s clothing. Moreover, many products were identified which do not comply with the present European PFOS regulation. Other samples, such as food contact papers, impregnation sprays and ski waxes showed massive loads with PFCA instead PFSA.