Journal of Chromatography A, 1448 (2016) 98–106

Contents lists available at ScienceDirect

Journal of Chromatography A
journal homepage: www.elsevier.com/locate/chroma

Simultaneous determination of perfluoroalkyl iodides,

perfluoroalkane sulfonamides, fluorotelomer alcohols, fluorotelomer
iodides and fluorotelomer acrylates and methacrylates in water and
sediments using solid-phase microextraction-gas
chromatography/mass spectrometry
Cristina Bach ∗ , Virginie Boiteux, Jessica Hemard, Adeline Colin, Christophe Rosin,
Jean-François Munoz, Xavier Dauchy
ANSES, Nancy Laboratory for Hydrology, Water Chemistry Department, 40 Rue Lionnois, 54000 Nancy, France

a r t i c l e i n f o a b s t r a c t

Article history: Here, we developed and validated a headspace-solid-phase microextraction-gas chromatography/mass

Received 19 January 2016 spectrometry (HS-SPME-GC/MS) method for the determination of 14 volatile perfluorinated alkylated
Received in revised form 30 March 2016 substances (PFASs) in water and sediment samples according to SANTE 11945/2015 guidelines. Three
Accepted 10 April 2016
fluorotelomer alcohols (FTOHs), two perfluoroalkyl iodides (PFIs), three fluorotelomer iodides (FTIs), four
Available online 12 April 2016
fluorotelomer acrylates and methacrylates (FTACs and FTMACs) and two perfluoroalkyl sulfonamides
(FASAs) were analysed simultaneously to assess the occurrence of these compounds from their emission
Keywords:
sources to the outlets in water treatment plants. Several SPME parameters were optimised for both water
Perfluoroalkyl and polyfluoroalkyl
substances
and sediment to maximise responses and keep analysis time to a minimum. In tap water, the limits of
Fluorotelomers quantification (LOQs) were found to be between 20 ng/L and 100 ng/L depending on the analyte, with
Solid-phase microextraction mean recoveries ranging from 76 to 126%. For sediments, LOQs ranged from 1 to 3 ng/g dry weight
Gas chromatography–mass spectrometry depending on the target compound, with mean recoveries ranging from 74 to 125%. SPME considerably
Water reduced sample preparation time and its use provided a sensitive, fast and simple technique. We then used
Sediments this HS-SPME-GC/MS method to investigate the presence of volatile PFASs in the vicinity of an industrial
facility. Only 8:2 FTOH and 10:2 FTOH were detected in a few water and sediment samples at sub-ppb
concentration levels. Moreover, several non-target fluorotelomers (12:2 FTOH, 14:2 FTOH and 10:2 FTI)
were identified in raw effluent samples. These long-chain fluorotelomers have high bioaccumulative
potential in the aquatic environment compared with short-chain fluorotelomers such as 6:2 FTOH and
6:2 FTI.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction stain-protectants, food-contact paper coatings, paints, adhesives,

performance chemicals (e.g. in aqueous film-forming fire-fighting
Perfluoroalkyl and polyfluoroalkyl substances (PFASs) are a foams (AFFFs)), etc. [6,7]. Due to their ubiquity and persistence, a
group of chemicals that have been produced in high volumes since broad range of PFASs have been detected in the environment, in
the 1950s for a wide range of industrial and domestic applica- wildlife and in humans [8,9]. The most frequently detected PFASs
tions [1–4]. The OECD has registered more than 1000 substances are perfluoroalkyl sulfonates (PFSAs) and perfluoroalkyl carboxy-
in this class of compounds [5]. PFASs are particularly used as lates (PFCAs), despite their non-volatility and low water solubility.
surfactants, because they are lipophobic and hydrophobic. They Their long-range transport pathways are still controversial [10]. It
are also employed for the production of fluorinated-based poly- is well known that fluorotelomer alcohols (FTOHs), perfluoroalkyl
mers for paper and carpet treatments and are used in upholstery iodides (PFIs), fluorotelomer iodides (FTIs), fluorotelomer acrylates
and methacrylates (FTACs and FTMACs) and perfluoroalkyl sulfon-
amides (FASAs) are precursors and/or intermediates used in the
∗ Corresponding author.
telomerisation process to manufacture PFASs and fluorotelomer
E-mail address: [email protected] (C. Bach).
(FT)-based polymers [7,11–13]. Residual levels of these precursors

http://dx.doi.org/10.1016/j.chroma.2016.04.025
0021-9673/© 2016 Elsevier B.V. All rights reserved.
 C. Bach et al. / J. Chromatogr. A 1448 (2016) 98–106 99

Table 1
Per- and polyfluoroalkyl substances (PFASs) analysed in this study with their acronym, CAS registry number, formula, and molecular weight (MW).

Chemical Class Compound Acronym CAS Number MW (g/mol) Formula

(Acronym)

Perfluoroalkyl iodides Perfluorohexyl iodide PFHxI 355−43-1 446 C6 F13 I

(PFIs)
Perfluorooctyl iodide PFOI 507−63-1 546 C8 F17 I
(n:2) Fluorotelomer 4:2 Fluorotelomer iodide 4:2 FTI 2043−55-2 374 C4 F9 CH2 CH2 I
iodides (FTIs)
6:2 Fluorotelomer iodide 6:2 FTI 2043−57-4 474 C6 F13 CH2 CH2 I
8:2 Fluorotelomer iodide 8:2 FTI 2043−53-0 574 C8 F17 CH2 CH2 I
(n:2) Fluorotelomer 6:2 Fluorotelomer acrylate 6:2 FTAC 17527−29-6 418 C6 F13 CH2 CH2 OC(O)CH CH2
acrylates (FTACs)
8:2 Fluorotelomer acrylate 8:2 FTAC 27905−45-9 518 C8 F17 CH2 CH2 OC(O)CH CH2
(n:2) Fluorotelomer 6:2 Fluorotelomer methacrylate 6:2 FTMAC 2144−53-8 432 C6 F13 CH2 CH2 OC(O)C(CH3 ) CH2
methacrylates
(FTMACs)
8:2 Fluorotelomer methacrylate 8:2 FTMAC 1996−88-9 532 C8 F17 CH2 CH2 OC(O)C(CH3 ) CH2
(n:2) Fluorotelomer 6:2 Fluorotelomer alcohol 6:2 FTOH 647−42-7 364 C6 F13 CH2 CH2 OH
alcohols (FTOHs)
8:2 Fluorotelomer alcohol 8:2 FTOH 678−39-7 464 C8 F17 CH2 CH2 OH
10:2 Fluorotelomer alcohol 10:2 FTOH 865−86-1 564 C10 F21 CH2 CH2 OH
Perfluoroalkane N-Methyl perfluorooctane sulfonamide MeFOSA 31506−32-8 513 C8 F17 SO2 NH(CH3 )
sulfonamides (FASAs)
N-Ethyl perfluorooctane sulfonamide EtFOSA 4151−50-2 527 C8 F17 SO2 NH(C2 H5 )

contained in raw materials and/or in several FT-based products may fluoropolymer industries. The availability of commercial standards
be released in the environment, undergo atmospheric transport also guided the choice of compounds to analyse. The final list of
and be oxidised in the atmosphere and/or degraded under aero- selected volatile PFASs with their chemical formula and properties
bic conditions into persistent PFASs [7]. Moreover, direct emissions is given in Table 1.
from runoff effluents and wastewater from industrial plants, and Neutral and volatile PFASs in environmental samples are
accidental or untreated discharges can also be potential sources of generally analysed by gas chromatography coupled with mass
PFASs. Ultimately, they end up in soil, sediments and water bodies spectrometry (GC/MS) [2,17,23,24,29]. As an alternative to GC/MS,
[14]. the simultaneous determination of neutral FTs and ionic PFASs
Many of these fluorinated substances combine bioaccumulative by liquid chromatography-mass spectrometry (LC–MS) has also
potential, toxic effects and extreme persistence. The occurrence been reported and comprehensively reviewed in Jahnke and Berger
and fate of PFASs in the aquatic environment are recognised as an [30]. However, in LC–MS methods, the simultaneous analysis of
emerging concern. For example, FTIs are oxidised in the telomeri- non-ionic and ionic PFASs is impeded by the ionisation suppres-
sation process and produce FTOHs. Likewise, FTOHs and FASAs are sion of FTOHs caused by the buffered mobile phases needed for
considered emerging pollutants because they are easily degraded the chromatographic separation of ionic PFASs. Furthermore, PFIs
in the environment and/or are easily metabolised in organisms, and FTIs cannot generate protonated or deprotonated molecules by
leading to the formation of perfluorooctanoate (PFOA), perfluorooc- electrospray-ionisation [3,31]. With regard to sample treatments,
tane sulfonic acid (PFOS) and other derivatives, which are much conventional techniques such as liquid-liquid extraction (LLE) and
more toxic than their PFAS precursors [15–17]. Furthermore, some solid-phase extraction (SPE) have been widely used for the deter-
studies have reported the potential estrogenicity of FTOHs and mination of neutral PFASs in water samples [30]. Alternatively,
PFIs using in vitro bioassays such as the E-screen assay, reverse- according to Ruan, et al. [2], solid-phase microextraction (SPME)
transcription PCR and the MVLN-cell based transactivation assay may be a promising free-solvent technique for the simultaneous
using the human MCF-7 cell line. For FTOHs, their estrogenic effects determination of these compounds involving minimum sample
depend on the chain length of the alkyl group. Similarly, bioassays treatment to generate more data more efficiently. In addition, our
on PFIs have shown that the six-carbon chain length and iodine laboratory has previously demonstrated the advantages of SPME
substitution is a key attribute in estrogenicity. Moreover, in vivo extraction for the determination of polycyclic aromatic hydrocar-
metabolism studies in rats show that the genotoxic effects of 6:2 bons (PAHs) at sub-ppb quantification levels in drinking water and
FTMAC are due to FTOH-related metabolites [10,18,19]. However, it pipes [32].
is currently unclear whether precursors cause endocrine disruption In this study, a headspace (HS)-SPME-GC/MS method was devel-
under realistic environmental exposure conditions. oped and validated for the simultaneous determination of 14
In 2009, our research group conducted two sampling campaigns volatile PFAS of different chemical classes (PFIs, FTIs, FTOHs, FTACs,
on the occurrence of 10 PFASs (PFCAs and PFSAs) in water intended FTMACs and FASAs) in water and sediments to investigate the
for human consumption [20]. Although results did not reveal any behaviour and fate of PFASs from their emission sources to the
heavily contaminated sites, PFAS levels in treated water were outlets of water treatment plants.
sometimes higher in drinking water than in raw water. Based on
these data, a relationship between the atypical composition profiles 2. Experimental
of water resources and the vicinity of a fluoropolymer manufactur-
ing plant was established [21,22]. These early studies also revealed 2.1. Reagents, standards and materials
the need to investigate initial and intermediate precursors, espe-
cially FTs used by fluoropolymer manufacturing processes, as well Individual PFIs, FTIs, FTACs and FTMACs were obtained
as other PFASs. A list of target compounds was established to cross- as pure substances (≥95%) from Sigma-Aldrich (St-Quentin-
check data from previous publications in air [23,24], water [25–27] Fallavier, France) and dissolved at approximately 1000 mg/L in
and soil [2,28] as well as industrial-use information provided by methanol (MeOH) of absolute ULC/MS grade purchased from Bio-
100 C. Bach et al. / J. Chromatogr. A 1448 (2016) 98–106

solve BV (Valkenswaards, the Netherlands). The FTOHs, FASAs 2.4. Instrumental analysis
and their mass-labelled standards (2-perfluorooctyl-[1,1-2 H2 -
1,2-13 C2 ]-ethanol (8:2 FTOH-13 C2 ), 2-perfluorodecyl-[1,1-2 H2 - The separation and detection of analytes were carried out on
1,2-13 C2 ]-ethanol (10:2 FTOH-13 C2 and n-ethyl-d5 -perfluoro-1- a 7890A GC coupled to an ion trap 240 MS equipped with an
octanesulfonamide (EtFOSA-d5 )) were purchased from Wellington autosampler 120, all from Agilent Technologies (Palo Alto, CA, USA).
Laboratories via BCP Instruments (Oullins, France), all at 50 mg/L in Chromatographic separation was carried out on a Rxi-624SilMS
MeOH. The 1H,1H,2H,2H-perfluorodecylacrylate-d3 (8:2 FTAC-d3 ) column (30 m × 0.25 mm I.D.; 1.4 ␮m film thickness) (Restek, Belle-
and 1H,1H,2H,2H-perfluorodecylmethacrylate-d5 (8:2 FTMAC-d5 ) fonte, PA, USA). The inlet was operated in splitless mode during
were obtained from Chiron AS via BCP Instruments (Oullins, France) fibre thermal desorption. The carrier gas was helium at a con-
both at 50 mg/L in MeOH. Ultrapure water was obtained from a stant flow of 1 mL/min. The oven temperature was programmed
Milli-Q system Elix Integral 10 (Millipore, Bedford, MA, USA). as follows: 50 ◦ C (7 min); 5 ◦ C/min to 188 ◦ C; 40 ◦ C/min to 300 ◦ C
(5 min). MS was operated in electron ionisation (EI) mode. For SPME
optimisation, full-scan mode was used in a mass range from 50
to 850 m/z. Calibration and quantification of samples were carried
2.2. Preparation of standard solutions
out using micro-selected ion storage mode (␮SIS) and tandem MS
(MS2 ) depending on target compounds (see Table 2). The source, ion
Stock solutions of individual standards (PFHxI, PFOI, 4:2 FTI, 6:2
trap, transfer line and manifold temperatures were held at 250 ◦ C,
FTI, 8:2 FTI, 6:2 FTAC, 8:2 FTAC, 6:2 FTMAC and 8:2 FTMAC) were
200 ◦ C, 310 ◦ C and 50 ◦ C, respectively.
each prepared at a concentration of 1000 mg/L in MeOH and stored
at 4 ◦ C. An intermediate stock solution combining these nine com-
2.5. Quality control
pounds at a concentration of 5 mg/L each was prepared in MeOH
from individual stock solutions and also stored at 4 ◦ C.
The limit of quantification (LOQ) was defined as the smallest
For SPME optimisation in full-scan acquisition mode (see
concentration of standard that results in a reproducible measure-
Section 2.4), water samples spiked with the 14 analytes at a con-
ment of peak areas consistent with the calibration curve. This
centration of 100 ␮g/L each in 10 mL of water were prepared from
concentration was used as the second point of the calibration curve.
appropriate dilution of the intermediate stock solution and com-
For water samples, the LOQs for a 10 mL water matrix ranged from
mercial individual standards (6:2 FTOH, 8:2 FTOH, 10:2 FTOH,
20 to 100 ng/L depending on the analyte. Different concentrations
MeFOSA and EtFOSA) (see Section 2.1).
were used for the matrix-matched calibration, starting from 10,
For calibration and quantification of water samples, a working
20, or 50 ng/L according to the LOQ of the target compound (see
stock solution of the 14 analytes at 50 ␮g/L each in MeOH was pre-
Table 3). For all compounds, the highest calibration point was
pared by appropriate dilution of the intermediate stock and the
1000 ng/L. Quantification was performed using the internal stan-
commercial standards. This multicompound working solution was
dard method, except for PFHxI and PFOI for which no commercial
used to prepare calibration standards and also for within-run and
labelled standards were available. 6:2 FTOH and 8:2 FTOH were
intra-sample controls (see Section 2.4). For internal calibration and
quantified using 8:2 FTOH-13 C2 ; 10:2 FTOH was quantified using its
quantification, a working stock solution of the five mass-labelled 13 C-labelled homologue (10:2 FTOH-13 C ); FTACs were quantified
standards was prepared at between 50 ␮g/L and 100 ␮g/L in MeOH 2
using 8:2 FTAC-d3 ; FTMACs were quantified using 8:2 FTMAC-d5
by dilution of commercial individual.
and FASAs were quantified using EtFOSA-d5 . Despite the lack of
For sediment samples, calibration standards and within-run
labelled standards for FTIs, 10:2 FTOH-13 C2 has proved to be an
and intra-sample controls were prepared using the working stock
adequate internal standard for these FT iodides in our experimental
solution for the 14 analytes (as described above) by adding an
conditions. The mass-labelled standards were added at the follow-
appropriate volume of this solution to 0.5 g dw of sediment.
ing concentrations: 250 ng/L for EtFOSA-d5 and 500 ng/L for 8:2
For internal standard calibration and quantification, the working
FTOH-13 C2 , 10:2 FTOH-13 C2 and 8:2 FTAC-d3 (see Table 2).
stock solution of five mass-labelled standard solutions at between
For sediment samples, the LOQs for 0.5 g dry weight (dw) of
50 ␮g/L and 100 ␮g/L (as described above) was used. An appropri-
sediment ranged from 1 to 3 ng/g dw depending on each target
ate volume of this solution was added to 0.5 g dw of sediment (see
compound. Matrix-matched calibration was carried out starting at
Section 2.4 below for more details).
0.4 ng/g dw, 1 ng/g dw, or 2 ng/g dw according to the LOQ of the
target compound (see Table 3). For all compounds, the highest cal-
ibration concentration was 20 ng/g dw. Labelled standards, used
2.3. SPME parameters as internal standards, were added at the following concentrations:
5 ng/g dw for EtFOSA-d5 and 8:2 FTMAC-d5 and 10 ng/g dw for 8:2
The starting point for the SPME extraction was based on the FTOH-13 C2 , 10:2 FTOH-13 C2 and 8:2 FTAC-d3 (see Table 2).
protocol described by Ruan et al. [2]. SPME was performed using The quadratic fit of the calibration curves obtained was checked
HS to extend SPME fibre life. SPME with direct immersion in water, for all water and sediment matrices, (R-squared values from regres-
a simple matrix, is common, but alternative SPME methods based sion analysis ≥0.98).
on HS have wide applications because they can be easily adapted Procedural blanks (10 mL of tap water spiked with labelled
to more complex matrices such as sediments [31]. standards) were also performed at regular intervals to check for
The following SPME operational parameters were investigated: potential contamination or carry over from sample to sample. Pro-
choice of fibre, temperature and time of fibre desorption, extrac- cedural blanks performed in this study were never 10 times above
tion temperature, extraction time, salt addition and water pH value. the LOQs of the analytes. The reliability of the results was ensured
When the HS mode is used, compounds need to be transported by within-run and intra-sample controls performed with each sam-
through the HS interface before reaching the fibre coating; ensuring ple batch. The within-run controls consisted of analytes at 250 ng/L
very efficient agitation can improve this transfer. We therefore did in tap water and at 10 ng/g dw of sediment inserted throughout the
not optimise this parameter [4]. The preliminary SPME conditions sample batch.
were 30 min at 30 ◦ C; agitation at 300 rpm and 7 min of desorp- Batches were validated only when the bias between the exper-
tion time. Fibre conditioning and injection temperatures were set imental and the theoretical concentration was ≤20%. Intra-sample
according to the fibre manufacturer’s instructions. controls consisted in spiking the analytes in the samples to
 C. Bach et al. / J. Chromatogr. A 1448 (2016) 98–106 101

Table 2
Gas chromatography-ion trap mass spectrometry conditions for the 14 PFAS precursors.

Compound Retention Ion acquisition Parent ion (m/z) CID voltage Quantifiers (m/z) Qualifiers (m/z)
time (min.) mode [ion assignment] (V) [ion assignment] [ion assignment]

PFHxIa 6.26 ␮SIS – – 231 [C5 F9 ]+ + 181 281[C6 F11 ]+

[C4 F7 ]+
PFOIa 11.55 ␮SIS – – 331 [C7 F13 ]+ + 231 381 [C8 F15 ]+
[C5 F9 ]+
4:2 FTI 14.21 MS 2 374 [C6 H4 F9 I]+ 0.46 227 [C6 H3 F8 ]+ 207[C2 F3 I]+ , 177
[CHF2 I]+
6:2 FTOH 17.82 ␮SIS – – 275 [C8 H3 F12 ]+ + 295 363 [C8 H4 F13 O]+
[C7 H2 F11 ]+
6:2 FTI 18.81 MS 2 474 [C8 H4 F13 I]+ 0.61 327 [C8 H3 F12 ]+ 281[C6 F11 ]+ , 277
[C3 HF6 I]+
8:2 FTOH 21.47 ␮SIS – – 405 [C10 F14 H3 O]+ 463 [C10 H4 F17 O]+ , 395
[C9 H2 F15 ]+
6:2 FTAC 22.22 MS 2 418 [C11 H7 F13 O2 ]+ 0 418 [C11 H7 F13 O2 ]+ –
8:2 FTI 22.67 MS 2 574 [C10 H4 F17 I]+ 0.63 427 [C10 H3 F16 ]+ 377 [C5 HF10 I]+
10:2 FTOH 24.69 MS 2 505 [C12 H3 F18 O]+ 0 505 [C12 H3 F18 O]+ 417 [C11 F15 ]+
6:2 FTMAC 24.76 MS 2 432 [C12 H9 F13 O2 ]+ 95 417 [C11 H6 F13 O2 ]+ 377 [C9 H6 F13 O]+ , 327
[C8 H3 F12 ]+
8:2 FTAC 25.50 MS 2 518 [C13 H7 F17 O2 ]+ 0 518 [C13 H7 F17 O2 ]+ –
8:2 FTMAC 27.79 MS 2 532 [C14 H9 F17 O2 ]+ 115 517 [C13 H6 F17 O2 ]+ 477 [C11 H6 F17 O]+ , 427
[C10 H3 F16 ]+
MeFOSA 32.57 MS 2 448 [C9 H3 F17 N]+ 1.40 428 [C9 H2 F16 N]+ 378 [C6 H2 F16 ]+
EtFOSA 33.16 MS 2 448 [C9 H3 F17 N]+ 1.10 428 [C9 H2 F16 N]+ 378 [C6 H2 F16 ]+
8:2 FTOH-2 C13 21.42 ␮SIS – – 408 + 409 –
10:2 FTOH-2 C13 24.64 MS 2 508 0 508 + 509 –
8:2 FTAC-d3 25.46 MS 2 521 0 521 –
8:2 FTMAC-d5 27.69 MS 2 538 115 519 –
EtFOSA−d5 33.09 MS 2 450 1.10 430 –

CID, collision-induced dissociation; ␮SIS, ion storage mode; MS2 tandem mass spectrometry; see Table 1 for definition of compound acronyms.
a
External standard calibration.

Table 3
Recoveries in tap and surface water and sediments spiked at the limit of quantification (LOQ) concentration; number of replicates = 5.

Tap water Surface water Sediments

Compound LOQ(ng/L) % Recovery(SD) LOQ(ng/L) % Recovery(SD) LOQ(ng/g dw) % Recovery(SD)

PFHxI 100 76 (7) 100 116 (8) 2 125 (10)

PFOI 100 108 (15) 100 83 (6) 2 82 (9)
4:2 FTI 20 126 (8) 20 103 (17) 1 90 (18)
6:2 FTOH 100 124 (6) 100 165 (5) 1 111 (6)
6:2 FTI 20 96 (9) 20 105 (8) 1 74 (5)
8:2 FTOH 100 117 (7) 100 213 (8) 1 105 (15)
6:2 FTAC 100 84 (5) 100 165 (11) 1 93 (5)
8:2 FTI 20 105 (11) 20 64 (12) 1 93 (20)
10:2 FTOH 50 99 (4) 50 112 (3) 1 102 (12)
6:2 FTMAC 50 102 (16) 50 145 (23) 3 92 (5)
8:2 FTAC 100 113 (5) 100 76 (11) 1 99 (8)
8:2 FTMAC 100 84 (8) 100 127 (11) 3 99 (12)
MeFOSA 20 85 (7) 20 103 (4) 1 94 (8)
EtFOSA 20 88 (3) 20 96 (5) 1 103 (9)

See Table 1 for definition of compound acronyms.

detect possible matrix effects. Intra-sample controls were rou- optimising analytical methods, we set up SPME parameters in tap
tinely performed on all samples and were considered valid when water (free of target compounds) because it is more similar to real
recovery was between 60% and 140%, as recommended in the matrices.
SANTE/11945/2015 guidelines [33].

3.1.1. The choice of fibre

3. Results and discussion The 65 ␮m PDMS/DVB fibre has been reported to be suitable
for the simultaneous determination of PFIs and FTIs in environ-
3.1. HS-SPME optimisation for water samples mental matrices [2]. However, moderately polar compounds such
as FTOHs, FASAs and FTACs may show more efficient extraction
For water samples, extraction parameters were optimised with with SPME fibres based on carboxen (CAR) adsorbents [4]. The
a 20 mL SPME vial filled with 10 mL of tap water spiked using the following SPME fibres were assessed: a 65 ␮m PDMS/DVB fibre,
working solution at 50 ␮g/L in MeOH. According to Pawliszyn [4], a 75 ␮m CAR/PDMS fibre and 50/30 ␮m DBV/CAR/PDMS Stable
once the SPME extraction has been optimised, new matrices may Flex fibre (Supelco via Agilent Technologies, Palo Alto, CA, USA).
not be compatible with the selected fibre coating or show decreased As shown in Fig. 1A, the lowest responses were observed using
extraction efficiency. Although ultrapure water is usually used for 85 ␮m CAR/PDMS. Performance varied with the compound for
102 C. Bach et al. / J. Chromatogr. A 1448 (2016) 98–106

Fig. 1. Optimisation of SPME parameters for 10 mL of tap water spiked with the 14 PFASs at 100 ␮g/L. All experiments were performed in quintuplicate. A shows the
comparison of responses with different SPME fibres. B gives the effect of temperature on responses. C shows the salting-out effect on the responses at different NaCl
concentrations. D shows the influence of pH on responses.

the other two fibres. Although DVB/CAR/PDMS is recommended time of 10 min. We therefore set desorption temperature and time
for a large range (C2 –C20 ) of volatile compounds [34], extraction at 240 ◦ C and 10 min, respectively.
capacities decreased with increasing molecular weight. This rela-
tionship may be due to the adsorption mechanism by which the
species is retained on the surface and not in the bulk of the fibre
coating in addition to particularities due to the specific chemical 3.1.3. Extraction temperature
structure of the analyte tested. However, in terms of overall perfor- Different extraction temperatures were investigated with the
mance, DVB/CAR/PDMS produced slightly better results compared HS-SPME: 30, 40, 50 and 60 ◦ C. It has been reported that PFIs and
to DVB/PDMS. Therefore, the 50/30 ␮m DVB/CAR/PDMS fibre was FTIs are more easily vaporised and effectively adsorbed onto the
selected for the development of the analytical method. PDMS/DVB fibre at 25 ◦ C compared with higher temperatures for
soil samples [2]. Contrary to expectations, upon varying extrac-
tion temperatures from 30 to 60 ◦ C, two opposing effects were
observed depending on molecular weight and the chemical class
3.1.2. Temperature and time of fibre desorption of the analyte (see Fig. 1B). For lighter iodide compounds (PFHxI,
Once the most suitable fibre was chosen, two desorption tem- PFOI, 4:2 FTI, 6:2 FTI), the best responses were observed at 40 ◦ C. In
peratures in the inlet of the GC were tested (see Fig. A1 in fact, for these iodide analytes, the efficiency of the fibre gradually
Supplementary content): 200 ◦ C and 240 ◦ C. For most tested com- decreased with increasing temperature, with partitioning from the
pounds, the best responses were obtained at 240 ◦ C, especially aqueous phase into vial HS being favoured [32]. For several heavier
for high molecular weight compounds, because their desorption compounds such as 8:2 FTI, 6:2 FTMACs, MeFOSA and EtFOSA, an
was better at higher temperatures. The following desorption times increase in temperature improved diffusion into the HS and then
were investigated: 5, 7 and 10 min. No substantial differences adsorption onto the fibre coating, producing better responses. For
were observed in the responses of FTOHs and FASAs in three time FTOHs, FTACs and 8:2 FTMAC, no substantial differences in fibre
setups (see Fig. A2 in Supplementary content). In contrast, the efficiency between 40 ◦ C and 60 ◦ C were observed. Finally, a tem-
other target compounds, especially FTIs and FTMACs, seemed to perature of 50 ◦ C was selected as the best compromise to detect
be affected, showing high responses using the longest desorption intermediate and heavier compounds, which are likely found in
 C. Bach et al. / J. Chromatogr. A 1448 (2016) 98–106 103

Fig. 2. Influence of extraction time on responses for 10 mL of tap water spiked at 100 ␮g/L with a multicompound solution of 14 PFASs in MeOH. All experiments were
performed in quintuplicate.

water (high boiling points) and accumulate in sediments (longer 3.2. HS-SPME optimisation for sediments
fluorinated chains) [35].
Using HS-SPME, the extraction of molecules from the solid
3.1.4. Extraction time matrix cannot be directly performed without considerably increas-
The extraction time is an important parameter dominated by the ing the temperature of extraction, which may lead to the loss of
analyte diffusion process to reach equilibrium between the three relatively volatile or thermolabile analytes. One of the most com-
phases: water/air/fibre coating [4]. Ideally, in SPME methods, all the monly used approaches is to analyse solids using pure water as an
analytes reach an equilibrium between these three phases, unless interface, which removes analytes from the solid matrix first, fol-
the desired limits of detection are attained [4]. The following expo- lowed by SPME [4]. Ruanet al. [2] reported good recoveries of PFIs
sure times were tested: 5, 10, 20, 30, 40, 50 and 60 min. As shown in and FTIs in soil samples using this approach. We therefore used the
Fig. 2, in agreement with Ruanet al. [2], it is clear that 30 min was HS-SPME method previously developed for water samples and only
sufficient to reach adsorption equilibrium for most analytes. For assessed two parameters for sediment extraction: sample weight
EtFOSA only, the equilibrium between the aqueous matrix and the and the volume of pure water.
DBV/CAR/PDMS fibre was not reached after 40 min; nevertheless
the resulting sensibility was considered sufficient. We thus chose 3.2.1. Sample weight
an extraction time of 30 min. The following weights of sediment samples were investigated:
0.2, 0.5, 0.7, 1, 1.5 and 2 g spiked with around 7–9 ng of the 14
3.1.5. Salt addition PFAS using the multicompound solution of 100 ng/L in MeOH. Sed-
In the SPME method, salt addition increases the ionic strength iments were previously lyophilised and then ground to obtain a
of the water sample. A salting-out effect occurs and molecules pass homogeneous sample and a higher surface for partitioning with
more readily from water to the vial headspace [4]. For ion strength water. 10 mL of pure water were added to the sediment sample in
optimisation, 0, 1, 2 and 5% (w/v) of NaCl were separately added into a 20 mL SPME vial. The fibre was exposed by applying the condi-
the vial with the spiked tap water solution. As shown in Fig. 1C, the tions previously optimised for water samples. Thus the mixture
highest salt content (5%) decreased the efficiency (salting-in effect) was extracted for 30 min, 300 rpm at 50 ◦ C in the autosampler.
of the SPME fibre. This salting-in effect with higher levels of Na2 SO4 Although there were higher standard deviations for some com-
has previously been observed with PFIs in soil samples [2]. For most pounds, increasing the sampling weight produced different effects
analytes, the salting-out effect showed the best response at 2% of in the responses, depending on the chemical class of the analyte
NaCl (w/v)—and thus was the selected salt content. (see Fig. 3A). Although there were no substantial differences in
the responses of PFHxI among the sample weights tested, the best
3.1.6. pH values response was obtained at 0.2 g for PFOI. In contrast, the highest
In several pre-treatment procedures, pH was adjusted in envi- responses were observed at 0.5 g and 0.7 g dw from 4:2 FTI to 8:2
ronmental samples to generate the stable speciation of molecules FTOH. A progressive decrease in responses was observed from 6:2
and extract them better. The conversion of the analytes into neutral FTAC to EtFOSA likely due to their low volatility compared with
or acidic forms using pH adjustment can improve method sensi- the lighter analytes. We thus set the sample weight at 0.5 g dw as
tivity [4].Three pH values were tested: 2, 6 and 7 in tap water to the best compromise for all analytes, keeping in mind that, if sam-
assess the efficiency of the HS-SMPE procedure. Unlike Ruan et al. pled sediments are highly contaminated, the sample weight can be
[2], varying pH values did not appear to be a limiting factor for the reduced to 0.2 g dw of sediment.
extraction of PFIs in water samples as shown in Fig. 1D. FTIs and
FTOHs in water at pH values close to 7 appeared to decrease the 3.2.2. Volume of ultrapure water
extraction capabilities of the fibre. In contrast, FTACs, FTMACs and In this case, water is the interphase between the solid matrix
FASAs produced better responses in acidic pH. Taking into account and air; its volume is thus an important parameter. The following
that the average pH values of real water samples range from 6.5 to volumes were investigated: 5, 7, 10, 12 mL of pure water added to
7.5, we did not adjust the sample pH. Moreover, detection perfor- 0.5 g dw of sediment in a 20 mL SPME vial. Sediment was spiked
mances for FTACs, FTMACs and FASAs were still reached in these with around 7–9 ng of the 14 PFAS using the multicomponent solu-
pH conditions. tion of 100 ng/L in MeOH. As shown in Fig. 3B, responses for FTACs
104 C. Bach et al. / J. Chromatogr. A 1448 (2016) 98–106

Fig. 3. Optimisation of SPME parameters for sediments spiked at 7–9 ng depending on analytes. A represents the effect of sample weight on responses. B represents the
influence of ultrapure water on responses. All experiments were performed in quintuplicate.

dropped with increasing water volume. In contrast, for 4:2 FTI, 8:2 CH3 -SO2 ]+ for EtFOSA. Quantification ion transition (448 > 428) of
FTOH, 10:2 FTOH and FASAs, the increase in water volume appeared both FASAs (MeFOSA and EtFOSA) correspond to the loss of one
to produce better results. For the other compounds, no significant HF (m/z = 20). Non-resonant mode was used for FTMACs generat-
differences were observed. The 10 mL volume was selected because ing the most intensive transition with the loss of a CH3 . For FTACs,
the resulting sensitivity was considered sufficient for all analytes. In the M+• > M+• transition was monitored due to their great stabil-
contrast with our results, Ruan et al. [2] used 50 mL of pure water ity in resonant and non-resonant mode and also because, upon
in a 150 mL vial to extract analytes from soil samples. However, applying the highest CID voltage, both FTACs showed much higher
their SPME procedure was not automated; they were therefore not fragmentation.
restricted to the commercial SPME vial volume.

3.4. Validation of the analytical method

3.3. GC–MSn analysis
The analytical method was validated according to the require-
Electron ionisation (EI) was investigated for analyte fragmen- ments described in the SANTE/11945/2015 guidelines[33]. The
tation. As previously reported [31], molecular ions for several limits of detection (LODs) and the LOQs of target compounds were
analytes (PFIs, FTOHs and FASAs) were not observed due the high determined on the basis of a signal-to-noise ratio of 3 and 10,
fragmentation pattern promoted by EI. As recommended by Jahnke respectively. The validation was conducted with matrices represen-
and Berger [30], positive ionisation (PCI) was checked and this soft tative of the planned sampling campaign, namely tap and surface
ionisation generated the protonated form of molecule. However, water and sediments. These matrices were previously analysed
the sensitivity of the analytes decreased dramatically compared to ensure the absence of analytes. In water matrices, recoveries
to EI (data not shown), likely due to external ionisation in our were checked at LOQ (see Table 3) and at 500 ng/L for all analytes.
ion trap. Therefore, the detection and identification of target com- As shown in Table 3, mean recoveries ranged from 76 to 126%
pounds was carried out in EI. Reference spectra were generated in tap water for LOQ-spiked levels. Matrix effects were observed
using commercial standards and retention times were used to in surface water, although labelled standards were used, recov-
distinguish among possible analogue structures. A representative eries of 6:2 FTOH, 8:2 FTOH, 6:2 FTAC and 6:2 FTMAC exceeded
chromatogram and MS parameters for the 14 PFASs are given in the upper limit of 140% set by the SANTE guidelines. However,
Fig. 4 and Table 2, respectively. ␮SIS acquisition mode was used the standard deviations (SD) for five replicates did not exceed
for PFIs, 6:2 FTOH and 8:2 FTOH because molecular ion and/or spe- 20%, thus these recoveries were consistent. For 6:2 FTOH and 8:2
cific fragments were not obtained or their relative abundances were FTOH in surface water spiked at 500 ng/L, the same phenomenon
too low. Although the [M-H]+ for 6:2 FTOH (363 m/z) and 8:2 FTOH was observed (see Table A1 in Supplementary content). To over-
(463 m/z) were generated, their intensities were around 10%, thus come these matrix effects, we corrected the concentration values
the MS2 could not be performed. For 10:2 FTOH, although MS2 was obtained for water samples in the sampling campaign using the
performed, CID voltage was not applied. We therefore only used recovery value obtained after spiking the corresponding samples.
MS2 to decrease noise and therefore increase the signal-to-noise The LOQs obtained for this analytical method (see Table 3) were
ratio for this molecule. For PFIs, higher fragmentation was observed 100-fold higher compared to method performances reported in
and no specific fragments of these molecules were found, only frag- surface water by Mahmoudet al. [27]. This study used method
ments typical of fluorinated compounds were observed (181, 231, detection limits (MDLs), which are defined as the lowest con-
281, 331, 381 m/z). centrations at which analytes in a sample do not cause matrix
For quantification, several ions were simultaneously selected interferences. These MDLs are typically determined using spiked
for PFHxI (231 + 181), PFOI (331 + 231) and 6:2 FTOH (275 + 295) ultrapure water. In our analytical method, the LOQs were deter-
to increase sensitivity. For the other analytes, MS2 was selected mined and validated in real water and sediment samples. These
because it has several advantages, especially selectivity. The best matrices are more representative of field samples compared with
responses in resonant mode were obtained for FTIs and FASAs. laboratory reagent water. Thus, our matrices were more complex
For FTIs, the most intensive transition corresponded to [M-HF-I]+ . and resulted in higher LOQs.
For FASAs, instead of the molecular ion M+• , the m/z = 448 was With regard to sediments, recoveries were checked at LOQs and
generated corresponding to [M-H-SO2 ]+ for MeFOSA and to [M- at 15 ng/g dw for all analytes. Mean recoveries ranged from 74% to
 C. Bach et al. / J. Chromatogr. A 1448 (2016) 98–106 105

Fig. 4. Extracted ion chromatograms of a calibration standard of the analytes at 500 ng/L in tap water. For definition of compound acronyms and GC/MS conditions, refer to
Tables 1 and 2 respectively.

125% at LOQ-spiked levels (see Table 3) and from 75% to 118% at these unexpected peaks was possible by comparing them with the
15 ng/g dw spiked level. Although no matrix effect was observed in fragmentation pattern obtained for 10:2 FTOH and 8:2 FTI spectra.
sediments, we systematically corrected positive results using the Mass spectra of unknown peaks and of reference compounds are
recovery values obtained after spiking the corresponding sediment shown in Figs. A3 and A4 . As shown in Fig. A3, according to M-H
sample. of 10:2 FTOH, 12:2 FTOH and 14:2 FTOH (563, 663 and 763 m/z),
the most intensive ions (505, 605 and 705 m/z) correspond to M-
58 [M-H2 F3 ]+ . The typical polyfluorinated telomer fragments 131,
3.5. Analysis of real samples
95 and 69 m/z are common to all spectra corresponding to [C3 F5 ]+ ,
[C3 H2 F3 ]+ and [CF3 ]+ , respectively. According to molecular ions for
The HS-SPME/GC–MS method of this study was used to
8:2 FTI (574 m/z) and 10:2 FTI (674 m/z) (see Fig. A4), the most
analyse water samples and sediments from four sampling cam-
intensive fragment corresponded to the [M-HF-I]+ ion for both
paigns at two different locations in France. Roughly 300 samples
compounds, reflecting typical fragmentation of FTIs. However, the
(water + sediments) were analysed. Only traces of 8:2 FTOH rang-
concentration of 10:2 FTI in the raw effluent samples was likely
ing from 102 and 246 ng/L were observed in 11 samples of surface
lower than 8:2 FTI; thus the 10:2 FTI mass spectrum was noisy
water from a river located 40 km south-west of a fluoropolymer
because it was close to the limits of overall sensitivity of MS. These
manufacturing plant. Two sediment samples from the same river
compounds were not quantified because the analytical standards
showed concentrations of 8:2 FTOH ranging from 1.1 to 5.7 ng/g
were not available in our laboratory and, furthermore, two of them
dw and concentrations of 10:2 FTOH ranging from 4.9 to 7.4 ng/g
(12:2 FTOH, 14:2 FTOH) are not even marketed.
dw. FTOHs are not regulated by the European Union or other
institutional entities; thus in terms of potential health risks, no con-
clusions can be drawn. All other PFASs in this study were either not
4. Conclusion
detected or were detected below the LOQs, and thus not quantified
in water and sediments.
We developed and validated an HS-SPME-GC–MS analytical
method for the simultaneous determination for 14 PFASs in water
3.6. Detection of non-target compounds and sediments. Recoveries, repeatability and LOQs were studied
showing good improvements according to the SANTE/11945/2015
Raw effluent samples obtained directly from an industrial facil- guidelines. Although LOQs obtained were higher in water matri-
ity were analysed in the four sampling campaigns. Our aim was ces in comparison with a previous paper, the SPME considerably
to assess the presence of target compounds and potentially detect reduced sample preparation time. Moreover, this extraction
non-target substances in the outlet of the industrial plant. The HS- method is a free-solvent technique and it is easier and quicker to
SPME-GC/MS method was used to analyse raw effluent samples, set up than traditional techniques such as solid-phase and liquid-
although the method had not been validated for this matrix. With liquid extraction. Thus, SPME is very useful for sampling campaigns
regard to GC–MS, full scan acquisition mode from 60 to 800 m/z in which a high number of samples are to be collected and analysed.
was used for detecting unknown compounds using the same sep- Here, we simultaneously identified and quantified PFIs, FTOHs, FTIs,
aration conditions and MS temperatures presented in Section 2.4. FTACs, FTMACs and FASAs in water and sediment samples. Four
Five target compounds were detected in all four raw effluent sam- sampling campaigns in the vicinity of an industrial facility were
ples from one industrial site: 6:2 FTOH, 6:2 FTI, 8:2 FTOH, 10:2 performed to investigate the occurrence of these volatile PFASs.
FTOH and 8:2 FTAC. Moreover, three unexpected chromatographic Only 8:2 FTOH and 10:2 FTOH were detected in a few water and
peaks were also observed in these samples corresponding to 10:2 sediment samples at sub-ppb concentration levels. Several non-
FTI, 12:2 FTOH and 14:2 FTOH. The elucidation of the structures of target FTs were identified in raw effluent samples, namely 12:2
106 C. Bach et al. / J. Chromatogr. A 1448 (2016) 98–106

FTOH, 14:2 FTOH and 10:2 FTI. Long-chain PFASs have high bioac- [13] B.S. Larsen, P. Stchur, B. Szostek, S.F. Bachmura, R.C. Rowand, K.B. Prickett, S.H.
cumulative potential in the aquatic environment. Since 2002, FT Korzeniowski, R.C. Buck, Method development for the determination of
residual fluorotelomer raw materials and perflurooctanoate in
industries have pledged to partly substitute in their manufacturing fluorotelomer-based products by gas chromatography and liquid
process line the linear C8 substances with C4 analogues, which do chromatography mass spectrometry, J. Chromatogr. A 1110 (2006) 117–124.
not appear to induce toxic effects. However, the presence of long- [14] K. Prevedouros, I.T. Cousins, R.C. Buck, S.H. Korzeniowski, Sources, fate and
transport of perfluorocarboxylates, Environ. Sci. Technol. 40 (2006) 32–44.
chain FTs (>C8 ) in raw samples was observed. The simple and rapid [15] C. Lau, K. Anitole, C. Hodes, D. Lai, A. Pfahles-Hutchens, J. Seed, Perfluoroalkyl
analytical method presented in this paper will encourage the mon- acids a review of monitoring and toxicological findings, Toxicol. Sci. 99 (2007)
itoring of PFASs to better assess their transport pathways and their 366–394.
[16] K. Dasu, L.S. Lee, R.F. Turco, L.F. Nies, Aerobic biodegradation of 8:2
fate in the aquatic environment.
fluorotelomer stearate monoester and 8:2 fluorotelomer citrate triester in
forest soil, Chemosphere 91 (2013) 399–405.
Acknowledgements [17] K. Dasu, J. Liu, L.S. Lee, Aerobic soil biodegradation of 8:2 fluorotelomer
stearate monoester, Environ. Sci. Technol. 46 (2012) 3831–3836.
[18] M. Maras, C. Vanparys, F. Muylle, J. Robbens, U. Berger, J.L. Barber, R. Blust, W.
The authors gratefully acknowledge financial support from the De Coen, Estrogen-Like properties of fluorotelomer alcohols as revealed by
Seine-Normandy Water Agency (AESN), the Rhône-Mediterranean- MCF-7 breast cancer cell proliferation, Environ. Health Perspect. 114 (2006)
100–105.
Corsica Water Agency (AERM), the Syndicat des Eaux d’Ile-de-
[19] S.S. Anand, T.L. Serex, C. Carpenter, E.M. Donner, R. Hoke, R.C. Buck, S.E.
France (SEDIF) and the Ile-de-France Regional Health Agency (ARS). Loveless, Toxicological assessment of tridecafluorohexylethyl methacrylate
They are also grateful to the wastewater treatment plant oper- (6:2 FTMAC), Toxicology 292 (2012) 42–52.
ators for their participation in this study, their assistance during [20] V. Boiteux, X. Dauchy, C. Rosin, J.-F. Munoz, National screening study on 10
perfluorinated compounds in raw and treated tap water in France, Arch.
sampling and for providing plant information. Finally, the authors Environ. Contam. Toxicol. 63 (2012) 1–12.
would like to thank the fluoropolymer manufacturer for providing [21] X. Dauchy, V. Boiteux, C. Rosin, J.-F. Munoz, Relationship between industrial
several standards and access to its raw effluent. discharges and contamination of raw water resources by perfluorinated
compounds. part I: case study of a fluoropolymer manufacturing plant, Bull.
Environ. Contam. Toxicol. 89 (2012) 525–530.
Appendix A. Supplementary data [22] X. Dauchy, V. Boiteux, C. Rosin, J.-F. Munoz, Relationship between industrial
discharges and contamination of raw water resources by perfluorinated
compounds: part II: case study of a fluorotelomer polymer manufacturing
Supplementary data associated with this article can be found, in plant, Bull. Environ. Contam. Toxicol. 89 (2012) 531–536.
the online version, at http://dx.doi.org/10.1016/j.chroma.2016.04. [23] X. Liu, Z. Guo, E.E. Folk Iv, N.F. Roache, Determination of fluorotelomer
025. alcohols in selected consumer products and preliminary investigation of their
fate in the indoor environment, Chemosphere 129 (2015) 81–86.
[24] S. Oono, K.H. Harada, M.A.M. Mahmoud, K. Inoue, A. Koizumi, Current levels of
References airborne polyfluorinated telomers in Japan, Chemosphere 73 (2008) 932–937.
[25] B. Szostek, K.B. Prickett, R.C. Buck, Determination of fluorotelomer alcohols by
[1] A.B. Lindstrom, M.J. Strynar, E.L. Libelo, Polyfluorinated compounds: past, liquid chromatography/tandem mass spectrometry in water, Rapid Commun.
present, and future, Environ. Sci. Technol. 45 (2011) 7954–7961. Mass Spectrom. 20 (2006) 2837–2844.
[2] T. Ruan, Y. Wang, T. Wang, Q. Zhang, L. Ding, J. Liu, C. Wang, G. Qu, G. Jiang, [26] S. Taniyasu, K. Kannan, M.K. So, A. Gulkowska, E. Sinclair, T. Okazawa, N.
Presence and partitioning behavior of polyfluorinated iodine alkanes in Yamashita, Analysis of fluorotelomer alcohols, fluorotelomer acids, and short-
environmental matrices around a fluorochemical manufacturing plant: and long-chain perfluorinated acids in water and biota, J. Chromatogr. A 1093
another possible source for perfluorinated carboxylic acids? Environ. Sci. (2005) 89–97.
Technol. 44 (2010) 5755–5761. [27] M.A.M. Mahmoud, A. Kärrman, S. Oono, K.H. Harada, A. Koizumi,
[3] U. Berger, I. Langlois, M. Oehme, R. Kallenborn, Comparison of three types of Polyfluorinated telomers in precipitation and surface water in an urban area
mass spectrometers for HPLC/MS analysis of perfluoroalkylated substances of Japan, Chemosphere 74 (2009) 467–472‘.
and fluorotelomer alcohols, Eur. J. Mass Spectrom. 10 (2004) 579–588. [28] H. Peng, K. Hu, F. Zhao, J. Hu, Derivatization method for sensitive
[4] J. Pawliszyn, Handbook of Solid Phase Microextraction, Chemistry Industry determination of fluorotelomer alcohols in sediment by liquid
Press, Beijing, 2009. chromatography–electrospray tandem mass spectrometry, J. Chromatogr. A
[5] OCDE, Lists of PFOS, PFAS, PFOA, PFCA, related compounds and chemicals that 1288 (2013) 48–53.
may degrade to PFCA, ENV/JM/MONO, 2006, 15, (2007). [29] W. Liu, S. Takahashi, Y. Sakuramachi, K.H. Harada, A. Koizumi, Polyfluorinated
[6] A. Jahnke, Polyfluorinated Alkyl Substances (PFAS) in the Marine telomers in indoor air of Japanese houses, Chemosphere 90 (2013)
Atmosphere—Investigations on Their Occurrence and Distribution in Coastal 1672–1677.
Regions, in, Von der Fakultät III—Umwelt und Technik der Universität [30] A. Jahnke, U. Berger, Trace analysis of per- and polyfluorinated alkyl
Lüneburg als Dissertation angenommene Arbeit, 2007, pp. 115. substances in various matrices—how do current methods perform? J.
[7] M.J.A. Dinglasan, Y. Ye, E.A. Edwards, S.A. Mabury, Fluorotelomer alcohol Chromatogr. A 1216 (2009) 410–421.
biodegradation yields poly- and perfluorinated acids, Environ. Sci. Technol. 38 [31] T. Portolés, L.E. Rosales, J.V. Sancho, F.J. Santos, E. Moyano, Gas
(2004) 2857–2864. chromatography–tandem mass spectrometry with atmospheric pressure
[8] Z. Wang, I.T. Cousins, M. Scheringer, R.C. Buck, K. Hungerbühler, Global chemical ionization for fluorotelomer alcohols and perfluorinated
emission inventories for C4–C14 perfluoroalkyl carboxylic acid (PFCA) sulfonamides determination, J. Chromatogr. A 1413 (2015) 107–116.
homologues from 1951 to 2030, Part I: production and emissions from [32] J. Tillner, C. Hollard, C. Bach, C. Rosin, J.-F. Munoz, X. Dauchy, Simultaneous
quantifiable sources, Environ. Int. 70 (2014) 62–75. determination of polycyclic aromatic hydrocarbons and their chlorination
[9] Z. Wang, I.T. Cousins, M. Scheringer, R.C. Buck, K. Hungerbühler, Global by-products in drinking water and the coatings of water pipes by automated
emission inventories for C4–C14 perfluoroalkyl carboxylic acid (PFCA) solid-phase microextraction followed by gas chromatography–mass
homologues from 1951 to 2030, part II: the remaining pieces of the puzzle, spectrometry, J. Chromatogr. A 1315 (2013) 36–46.
Environ. Int. 69 (2014) 166–176. [33] SANTE/11945/2015, Guidance document on analytical quality control and
[10] C. Wang, T. Wang, W. Liu, T. Ruan, Q. Zhou, J. Liu, A. Zhang, B. Zhao, G. Jiang, method validation procedures for pesticides residues analysis in food and
The in vitro estrogenic activities of polyfluorinated iodine alkanes, Environ. feed., European Commission health and consumer protection directorate
Health Perspect. 120 (2012) 119–125. general. http://ec.europa.eu/food/plant/docs/plant pesticides mrl guidelines
[11] R.C. Buck, J. Franklin, U. Berger, J.M. Conder, I.T. Cousins, P. de Voogt, A.A. wrkdoc 11945 en.pdf.
Jensen, K. Kannan, S.A. Mabury, S.P.J. van Leeuwen, Perfluoroalkyl and [34] S. Risticevic, H. Lord, T. Gorecki, C.L. Arthur, J. Pawliszyn, Protocol for
polyfluoroalkyl substances in the environment: terminology, classification, solid-phase microextraction method development, Nat. Protocols 5 (2010)
and origins, Integr. Environ. Assess. Manage. 7 (2011) 513–541. 122–139.
[12] C.M. Butt, D.C.G. Muir, S.A. Mabury, Biotransformation of the 8:2 [35] E. Kissa, Fluorinated Surfactants and Repellents, 2nd edition, Marcel Decker,
fluorotelomer acrylate in rainbow trout. 1. In vivo dietary exposure, Environ. New York, 2001.
Toxicol. Chem. 29 (2010) 2726–2735.