Closing the gap – inclusion of ultrashort-chain perfluoroalkyl carboxylic acids in the total oxidizable precursor (TOP) assay protocol

Joachim Janda a, Karsten Nödler a, Marco Scheurer a, Oliver Happel a, Gudrun Nürenberg a, Christian Zwiener b and Frank Thomas Lange *a
aTZW: DVGW-Technologiezentrum Wasser, Karlsruhe Str. 84, 76139 Karlsruhe, Germany. E-mail:
bEberhard Karls University Tübingen, Environmental Analytical Chemistry, Center for Applied Geosciences (ZAG), Hölderlinstr. 12, 72074 Tübingen, Germany

Received 10th April 2019 , Accepted 29th May 2019

First published on 30th May 2019

An improved protocol of the total oxidizable precursor (TOP) assay was developed for precursors to C2–C14 perfluoroalkyl carboxylic acids (PFCAs) and C4–C8 and C10 perfluoroalkyl sulfonic acids (PFSAs). The proposed protocol was tested and validated for contaminated soil samples. The perfluoroalkyl acids (PFAAs) present in the soil extract solutions after oxidation with persulfate were separated from the inorganic salts by vacuum-assisted drying of the digestion solution followed by solid–liquid extraction of the PFAAs with acetonitrile from the dry residue. Ion chromatography (for C2–C4 PFCAs) and reversed phase liquid chromatography (for all other PFASs), both coupled to tandem mass spectrometry, were used for quantification. High procedural recoveries of PFAAs between 68% and 123% with RSDs between 0.2% and 25% (n = 3) were achieved. The method was validated using selected polyfluoroalkyl phosphoric acid esters (PAPs) and bis-[2-(N-ethyl perfluorooctane-1-sulfonamido)ethyl] phosphoric acid ester (diSAmPAP) as model precursors in pure solutions and in the presence of soil matrix. The oxidation led to characteristic and reproducible PFCA patterns (in the case of PAPs) or PFOA (in the case of diSAmPAP) with total reaction yields between 92 ± 4% and 123 ± 13% (n = 3). The impact of the soil matrix on transformation yields was negligible. In a soil core from a PFAS-polluted agricultural site, precursors were concentrated in the upper 40 cm with long-chain precursors being prevalent. After oxidative digestion, the total molar PFAA-concentrations increased by factors of 1.6 to 5.0. More than 40 cm below ground precursors of TFAA, PFPrA and PFBA accounted for ∼50% of the reaction products, underlining the importance of their inclusion in mass balances based on the TOP assay.

Environmental significance

The oxidative conversion of precursors to perfluoroalkyl acids (PFAAs) by means of the total oxidizable precursor (TOP) assay is a versatile digestion method increasingly applied in recent years to obtain a measure for total PFASs in water or soil samples. However, the high salt concentration in the digested samples previously hampered the quantitative analysis of ultra-short chained perfluoroalkyl carboxylic acids (PFCAs). The introduction of a novel selective solid–liquid extraction of PFASs with acetonitrile from the dry residue of a soil extract after oxidative digestion and evaporation enables the quantitative analysis of the ultrashort-chained PFCAs including trifluoroacetate. The inclusion of these ultra-short chained acids closes the hitherto existing gap in the PFAS mass balance when applying the TOP assay.


Perfluoroalkyl and polyfluoroalkyl substances (PFASs) have been produced and applied in numerous industrial and consumer products for more than 60 years.1,2 Especially the use of PFASs in aqueous film forming foams (AFFFs) for fire-fighting3–7 and, less frequently, the application of compost, containing paper-fiber biosolids from the production of greaseproof food contact papers, to agricultural fields8,9 had led to local and widespread contaminations of soil, groundwater, and drinking water, respectively.10 Poly- and perfluorinated precursors to perfluoroalkyl acids (PFAAs) can be degraded in the environment to perfluoroalkyl carboxylic acids (PFCAs) and perfluoroalkyl sulfonic acids (PFSAs) as terminal products.11 Due to long range atmospheric or oceanic transport and degradation of volatile PFAA precursors, background levels of PFAAs occur ubiquitously in the environment.11–13 They have been detected in many environmental compartments: in soil,11 water,10 air,14 in plants15 as well as in wildlife and humans.16

Routine environmental analysis of PFASs by liquid chromatography-tandem mass spectrometry (LC-MS/MS) mainly comprises PFCAs and PFSAs.17–19 This approach might be appropriate for exposure assessment, e.g. for analysis of drinking water or crops, where terminal PFAS degradation products are prevailing. However, emissions through products (e.g. AFFFs or paper chemicals) often contain many PFASs, which can potentially act as PFCA and PFSA precursors. Only few of them can be analyzed by target analysis due to the lack of knowledge on compound identity and/or missing authentic and mass-labelled standards.

Sum parameters, such as adsorbable organic fluorine (AOF) for aqueous samples,20–22 extractable organic fluorine (EOF) for soil and other solids,23–26 and the total oxidizable precursor (TOP) assay for aqueous samples and solids,6,27 can be applied to assess the amount of unknown precursors by organofluorine mass balance (AOF, EOF) or by molar balance of perfluoroalkyl moieties (PFAAs prior to and after TOP assay). In contrast to AOF and EOF, which can also detect organofluorine from pollutants other than PFASs,21e.g. fluorinated pharmaceuticals or pesticides, the TOP assay is specific for perfluoroalkyl moieties. A comparison of existing sum parameters for organofluorine analysis is given in a recent review.28

Application of the TOP assay produces large amounts of sodium sulfate and potassium sulfate by reduction of the oxidizing agent potassium peroxodisulfate in the alkaline reaction batches. The theoretical final salt content, neglecting oxidation of matrix components, is 10.6 g K2SO4 and 8.5 g Na2SO4, in addition to 1.2 g excess NaOH per liter.27 This hampers the subsequent quantitative analysis of PFCAs with chain lengths shorter than C4 either due to signal suppression during direct injection reversed-phase (RP)-LC-MS/MS analysis or by interfering with the typically applied weak anion exchanger (WAX) solid phase extraction (SPE) for PFAAs. Therefore, most previous studies did not include the short-chain PFCA perfluoropropanoic acid (PFPrA, C3) while trifluoroacetic acid (TFAA, C2) was not included at all in the molar balances.6,27,29–31 Moreover, even without correspondingly high salt concentrations, TFAA and PFPrA in general are difficult to include in analytical methods targeting PFCAs and require special chromatographic approaches.32,33 However, in several products short-chain PFASs, e.g. with C4 and C6 perfluoroalkyl moieties, have become important substitutes of long chain legacy PFASs.34 Oxidation of these short-chain precursors during environmental degradation or in the TOP assay may produce significant amounts of PFAAs with a chain length shorter than C4, which were neither considered within the mentioned studies nor can be included to future studies unless the method will substantially be improved. This was shown in a recently published study in which the important contribution of PFPrA to the molar mass balance of 4:2 and 6:2 fluorotelomer precursors after TOP assay was highlighted. However, molar conversion yield was still inadequate with only 62% for 4:2 FTSA, underlining the need for the inclusion of TFAA, too.

This study presents an improved TOP assay protocol, which separates the target PFASs to be analyzed nearly quantitatively from the interfering salt burden in the oxidation batches by a novel solid–liquid extraction approach, prior to instrumental analysis employing ion chromatography (IC) and reversed phase (RP) LC, both coupled to tandem mass spectrometry (MS/MS). This improved protocol allows for the first time an inclusion of C2 and C3 PFCAs in the TOP assay molar balance for precursors.

Methods and materials

Detailed information on chemicals and materials used in this study is given in the ESI.

Samples and sample pretreatment

About 1000 ha of agricultural land (as at January 2019) in the Upper Rhine Valley (Baden-Württemberg, Germany) have been polluted with PFAS-contaminated compost presumably in the period 2000–2008. Among other PFASs polyfluoroalkyl phosphoric acid esters (PAPs) and bis-[2-(N-ethyl perfluorooctane-1-sulfonamido)ethyl] phosphoric acid ester (diSAmPAP) were proven to occur as precursors to PFAA in this region.9 A soil core was taken from such a contaminated field and was separated in segments of 10 cm thickness down to 150 cm depth. Four segments thereof (SEG1: 0–10 cm, SEG2: 20–30 cm, SEG3: 40–50 cm, and SEG4: 60–70 cm) were analyzed in this study. A non-polluted reference sample (hereinafter referred to as reference soil) was collected from a non-contaminated agricultural field in Rheinstetten-Forchheim, also located in the Upper Rhine Valley.

Soil samples were sieved (≤5 mm) and used for further preparations. A subsample of approx. 250 g was freeze-dried and ground in a ball mill. The authors are aware that this process is detrimental to the integration of volatile compounds to the TOP assay. However, when working with soil extracts (see the following sections) the methanol (MeOH) is fully evaporated before the TOP assay,6 which finally also results in a loss of volatiles. Therefore, drying/grinding was preferred, as the procedure allows for optimal homogenization of the sample. The dry soil samples were stored in polypropylene (PP) bottles and kept in the dark at room temperature. Analyses were carried out within three weeks after sample pretreatment.

Preparation of soil extracts

A subsample of 0.5 g of pretreated soil was weighed into a 50 mL centrifuge tube and 5.0 mL of MeOH were added. Extraction with neutral MeOH was chosen for the reason of comparability with the German standard method DIN 38414-14 for the PFAA analysis from soil.19 After vortex mixing (∼10 s), the mixture was sonicated for 15 min at 25 °C, followed by shaking for 30 min on a vortex mixer at 1800 rpm. After 5 min centrifugation at 3000 rpm (2968×g), the liquid phase was completely transferred into a 15 mL PP tube and the extraction process was repeated with fresh MeOH. The combined extracts were concentrated to 5.0 mL using a gentle stream of N2 at 33–34 °C. For the determination of the native concentrations of the analytes, i.e. the concentrations before the TOP assay, a 1.0 mL aliquot of this extract was evaporated to dryness under a gentle stream of N2, reconstituted in 10 μL of internal standard working solution (IS-mix) and 990 μL of MeOH/H2O (9:1, v/v) and analyzed as described in the subchapter “Instrumental analysis”.

TOP assay

An aliquot of the methanolic soil extract (1.0 mL, corresponding to 0.1 g of the original soil) was transferred to a 10 mL centrifuge tube (Nalgene™, Oak Ridge, polypropylene copolymer (PPCO) with screw cap) and carefully dried under a stream of N2. After evaporation of the solvent, 8.0 mL K2S2O8 solution (20 g L−1) and 0.15 mL 10 N NaOH solution were added to the dried extract and the mixture was vortexed (10 s). The tube was completely filled with ultra-pure water to avoid headspace and firmly closed. For oxidative digestion, samples were heated to 85 °C in an oven for 20 h before cooling in an ice bath.

A 10 μL aliquot of the IS-mix was placed into a 12 mL round bottom glass test tube followed by a complete transfer of the cooled digestion solution to the test tube. The PPCO tube was rinsed with 0.5 mL MeOH, which was also added to the glass tube. The liquid phase was removed at 10 mbar using a rotational vacuum concentrator with the following temperature program: from 25 °C to 50 °C within 1 min, holding this temperature for 4.5 h before cooling down the samples to 30 °C for 30 min.

Sample clean-up

Final method. To extract the analytes, 1.4 mL acetonitrile (ACN) was added to the dried residue, which was thoroughly crushed and homogenized afterwards with a metal spatula. After adding 6–7 glass beads, the test tube was sealed with a PP stopper and mixed for 30 min at 2000 rpm on a vortex shaker. The solid sample components were allowed to settle and the supernatant was transferred to a 2 mL PP centrifuge tube. This fraction was centrifuged (10 min at 12.000×g) before transferring the supernatant to another 2 mL tube. The extraction process was repeated by using 0.7 mL ACN and 30 s mixing time and the supernatants were combined. After evaporating the ACN by a gentle stream of N2 and redissolving the residues in 1.0 mL of a mixture of MeOH and ultra-pure water (9[thin space (1/6-em)]:[thin space (1/6-em)]1, v/v), the resulting extract was transferred to a 1.5 mL PP vial for instrumental analysis with RP-LC-MS/MS and IC-MS/MS (see below).
Optimization process. The optimization of the clean-up step included an evaluation of the number of extraction steps necessary for high extraction yields. For this purpose, blank digestion solutions containing only ultra-pure water and the reagents (K2S2O8, NaOH) were prepared. After reacting and cooling down, they were spiked with the analytes (10 ng each, corresponding to 1 μg L−1) in triplicate without prior addition of IS-mix. After the drying step (rotary vacuum concentrator), three consecutive extractions with ACN (1st step: 1.4 mL, 2nd and 3rd steps: 0.7 mL each) were performed as described above. The resulting extracts were collected separately and individually spiked with 10 μL aliquots of IS-mix before drying in order to distinguish between extraction losses and signal suppression due to co-extracted matrix.

Instrumental analysis

Instrumental analysis was performed by applying two separate methods based on IC-MS/MS (TFAA, PFPrA, and perfluorobutanoic acid (PFBA)) and RP-LC-MS/MS (all other analytes), respectively. A suppressor unit is not required for the IC-MS/MS application as the buffer contains only volatile components. Therefore, the method does not require any additional laboratory equipment except for the IC column and a different buffer. Further details of the analytical system used and the instrumental parameters are given in the ESI.

In order to avoid unfavorable chromatographic effects, such as peak broadening, fronting or asymmetry due to the high organic solvent content in the samples, an injector program was implemented for samples measured with RP-LC-MS/MS to adjust the ratio of water and organic solvent: after drawing 10 μL from the sample, the injector needle was rinsed with isopropanol at the flushing position of the autosampler for 5 s. Then 40 μL of 2 mM ammonium formate solution were drawn into the sample loop and this mixture was injected. In IC-MS/MS the final extract was injected without any additional measures.

TOP assay with selected precursors

Four different precursors were subjected to the TOP assay: 1H,1H,2H,2H-perfluorooctyl-1H,1H,2H,2H-perfluorodecyl phosphate (6:2/8:2 diPAP), bis(1H,1H,2H,2H-perfluorodecyl) phosphate (8:2 diPAP), 1H,1H,2H,2H-perfluorododecyl phosphate (10[thin space (1/6-em)]:[thin space (1/6-em)]2 monoPAP), and diSAmPAP. Experiments with these precursors were performed in the presence of the reference soil matrix and also free of matrix. In the first case, a 1 mL aliquot of a previously prepared extract from the reference soil was given into the reaction vessel and spiked with 0.15 mL (10:2 monoPAP, 8:2 diPAP or diSAmPAP) or 0.25 mL (6:2/8:2 diPAP) of solutions in MeOH (β = 0.1 μg mL−1). In matrix-free experiments the same amount of each precursor was used without adding the soil extract. After careful removal of the methanol in a N2-stream the TOP assay and clean-up were performed as described above. All experiments were performed in triplicates.

Individual spiking experiments were performed with perfluoroalkane sulfonamides (FASAs). Perfluorooctane sulfonamide (FOSA) was already shown to convert virtually equimolar to perfluorooctanoic acid (PFOA) in the TOP assay27 and, therefore, a comparison with the literature would be possible. It could be assumed that shorter chained homologues will also be converted to the corresponding PFCAs. Subsamples of the reference soil were spiked using a working solution containing four FASAs (C2, C4, C6 and C8, FASA-mix) at levels of 25 μg per kg dw or 50 μg per kg dw. These samples were prepared in duplicate each.

Quality control

Each set of samples contained a procedural blank in order to monitor for contamination. Analytes were quantified by isotope dilution quantification. The calibration standards (seven concentration levels in the range 0.1–25 μg L−1 prepared in MeOH/H2O, 9:1, v/v), were prepared separately for the two instrumental methods: calibration standards for IC-MS/MS contained TFAA, PFPrA, PFBA as well as the mass-labelled compounds TFAA-M2 and PFBA-M4. The standards for RP-LC-MS/MS contained all PFCAs, PFSAs, FASAs, and isotopically labelled PFASs addressed in this study.

Procedural recoveries of the PFCAs and PFSAs were assessed by spiking subsamples of the uncontaminated reference soil with PFCAs and PFSAs in triplicate (100 μg per kg dw each) and conducting the whole method. Prior to the sample preparation, the spiked soil was thoroughly vortex mixed and the ACN was allowed to evaporate overnight at room temperature. In order to prevent interferences between spiked PFAAs and PFCAs formed from FASAs during the TOP assay, no FASAs were spiked in these experiments.

Results and discussion

Optimization of sample clean-up

After application of the TOP assay, excess sulfate and hydroxide in the mmol L−1 range hamper the quantitative determination of PFASs, especially of short-chain PFCAs. Therefore, solid–liquid extraction (SLE) and liquid–liquid extraction (LLE) were tested for matrix separation. LLE with ACN or ethyl acetate was not successful (for details see ESI, subchapter 1.5). SLE of the analytes from dried residues of the TOP assay with acetone and ACN were highly promising and less than 0.01% of sulfate was transferred to the organic extract (ESI, Table S4). Application of acetone, however, had two remarkable differences compared to ACN: at longer contact times (∼1–1.5 h) with the dry residues of the oxidized solution a yellow color of the extraction solvent successively appeared and small amounts of a low-volatile liquid were generated hampering the solvent exchange. This coloring is most likely due to base-catalyzed self-condensation of acetone leading to isophorone, a substance with a yellowish color.35 Since condensation reactions generate water, the low-volatile liquid probably consisted of isophorone and water. SLE with ACN showed no such effects. Therefore, after determining the number of extraction steps required to virtually completely recover the PFAAs generated by oxidation, the final method was established on the basis of ACN. Experimental details are given in subchapter “Sample clean-up”. The resulting recoveries are shown in Fig. 1.
image file: c9em00169g-f1.tif
Fig. 1 Extraction recoveries of PFASs after three consecutive SLE steps with ACN. Error bars indicate standard deviations (1σ) of triplicates.

Except for perfluorohexane sulfonamide (FHxSA; 70%), excellent extraction yields between 87% (perfluorodecane sulfonic acid, PFDS) and 121% (perfluoropentane sulfonic acid, PFPeS) were achieved after two extraction steps. A third extraction step did not lead to any substantial improvement and, therefore, was not applied in further experiments.

Instrumental analysis

Particularly long-chained PFASs exhibit high surface activity, which may result in substantial losses of analytes from the liquid phase of aqueous solutions.36 To avoid such losses, a solvent with a high proportion of MeOH (MeOH/H2O, 9:1, v/v) was selected for the reconstitution of the analytes after the SLE with ACN and its subsequent evaporation. Lower contents of MeOH led to decreasing recoveries of the long-chain PFCAs covered in this study (data not shown). In RP-LC-MS/MS, however, injection of samples with a high organic content compared to the starting composition of the mobile phase can lead to unsatisfactory or even inacceptable peak shapes. Thus, an injection program was implemented to adjust the MeOH/H2O ratio. Preliminary injections with TFAA in MeOH/H2O (9:1, v/v) under the conditions of the RP-LC-MS/MS method led to no signal for this analyte (data not shown). At a concentration of 2.5 μg L−1 the peaks of PFPrA and PFBA in RP-LC showed substantially larger widths at base (wb) and lower signal-to-noise ratios (SNRs) at a level of 25 μg L−1 compared to the signals of other analytes in this method. In addition, the PFPrA peak was asymmetric. Both can be attributed to mixing effects after injection of 50 μL sample via the injector program. Therefore, a second method was used for C2–C4 PFCAs based on IC. The IC method was originally developed for the analysis of TFAA37 and was extended for the determination of PFPrA and PFBA in this study. This method does not require a suppressor unit, which is typically used to suppress the conductivity of the eluent and to remove non-volatile constituents of the mobile phase. Furthermore, no injector program was necessary. The peaks obtained by IC-MS/MS showed excellent symmetry and achieved higher SNRs than analytes in RP-LC (SNRPFPrA = 3780, SNRPFBA = 6870). Thus, IC-MS/MS was used for quantitation of PFCAs with C2 to C4 and the signals of PFBA and PFPrA in RP-LC were only used for qualitative confirmation, since only one mass transition was available for the PFCAs with C < 4. TFAA, however, was analyzed without additional confirmation. Typical chromatograms resulting from both instrumental methods are shown in Fig. 2.
image file: c9em00169g-f2.tif
Fig. 2 Chromatograms obtained with IC-MS/MS and RP-LC-MS/MS for a calibration standard (5.0 μg L−1), a native soil extract, and an oxidized extract. The substances analyzed by RP-LC are displayed here separated by classes, but are all recorded in one measurement. The numbered peaks correspond to the analytes as follows: (1) TFAA, (2) PFPrA, (3) PFBA, (4) PFPeA, (5) PFHxA, (6) PFHpA, (7) PFOA, (8) PFNA, (9) PFDA, (10) PFUnDA, (11) PFDoDA, (12) PFTrDA, (13) PFTeDA, (14) PFBS, (15) PFPeS, (16) PFHxS, (17) PFHpS, (18) PFOS, (19) PFDS, (20) FEtSA, (21) FBSA, (22) FHxSA, (23) FOSA. Marked (*) signals were used for qualitative purposes only; TFAA, PFPrA and PFBA were quantified by IC-MS/MS.

Validation experiments

Details about the analytical characteristics of the instrumental methods applied in this study, such as figures of merit, are summarized in the ESI (subchapter 3.2).

The calibration curves in the working range (0.1–25 μg L−1) were linear (coefficients of determination R2 > 0.99) for all analytes. Limits of quantification (LOQ) based on the instrumental LOQ and concentration factors achieved during sample preparation were between 0.2 μg per kg dw (perfluorododecanoic acid, PFDoDA) and 1.1 μg per kg dw (PFOA). Details are summarized in the ESI (Table S5). An adjustment of the procedure to lower contamination levels might be achieved by – either individually or in combination – e.g. (i) exposing larger sample aliquots to the specified TOP assay, (ii) using both larger sample aliquots and more persulfate/NaOH, and (iii) lowering the volume of MeOH/H2O-mixture (9:1, v/v) for preparing the final extract. The latter seems to be directly applicable while using the other strategies may result in (i) decreased oxidation potential for the precursors due to the presence of elevated organic matrix components or (ii) decreased extraction yields of the oxidation products (i.e. PFCAs) due to the presence of increased amounts of salt residues and the effects need to be evaluated before application.

Contaminated soil samples can contain PFCAs and PFSAs, e.g. from microbial degradation of polyfluorinated precursors,11,38 but PFAAs may also be formed in oxidative treatment of the samples.27 In order to ensure that (i) they can be reliably extracted from soil, (ii) they do not degrade in the subsequent TOP assay, and (iii) they can be recovered during clean-up, procedural recoveries were determined. For this purpose, subsamples of the reference soil were spiked in triplicate at 100 μg per kg dw for each PFAA. The solvent of the spike solution was allowed to evaporate over night before the samples were processed. Mean recoveries and precisions obtained after conducting the complete procedure are summarized in the ESI (Table S6). Except for PFPrA (68%), very good procedural recoveries were achieved for the PFCAs and PFSAs between 82% (PFOA) and 123% (perfluorotetradecanoic acid, PFTeDA) with precisions between 0.2% (PFOA) and 25% RSD (perfluorotridecanoic acid, PFTrDA) (n = 3). The results clearly demonstrate the applicability of the improved method of the TOP assay with an enhanced parameter list of PFCAs from C2 to C14.

Application to selected precursors

To assess the reaction products formed from precursors three PAPs (10:2 monoPAP, 6:2/8:2 diPAP, 8:2 diPAP) and diSAmPAP typically used for grease proofing of paper and board were subjected to the TOP assay. These PFASs – among others – had been detected previously in soil samples from the respective region.9 To investigate the impact of matrix on the TOP assay in later analyses, the four precursors were treated in presence and absence of soil matrix. For this purpose, an extract of the non-contaminated reference soil was used and compared to the matrix-free sample (ultra-pure water). The molar reaction yields for the PFCAs obtained are presented in Fig. 3.
image file: c9em00169g-f3.tif
Fig. 3 Molar reaction yields for PFCAs formed from oxidation of PAPs and diSAmPAP by the top assay in presence and absence (matrix free) of a soil matrix. The recoveries are based on the maximum achievable concentrations from the applied PAPs and diSAmPAP; error bars: 1σ, n = 3.

With few exceptions, concentrations of PFCAs formed during digestion of the PAPs and diSAmPAP were similar in H2O (matrix free) and in the presence of soil matrix, so the matrix influence did not lead to substantially different concentration ratios. In the case of 10:2 monoPAP, PFCAs with chain lengths from C4 (PFBA) to C11 (perfluoroundecanoic acid, PFUnDA) occurred after the TOP assay, with perfluorononanoic acid (PFNA) being predominant (35% yield) followed by PFDA (soil matrix: 16%, matrix free: 18%) and PFOA (soil matrix: 16%, matrix free: 15%). From 8:2 diPAP, PFCAs from C4 (PFBA) to C9 (PFNA) were formed. Considering the results of 10:2 monoPAP, even shorter-chained PFCAs were expected to result from an 8:2 fluorotelomer (FT) precursor, but TFAA and PFPrA were not detected. However, comparing the concentration patterns with 10:2 monoPAP, two PFCAs as main products with similar molar recoveries, perfluoroheptanoic acid (PFHpA; 30–38%) and PFOA (36–38%), were found instead of one predominant product in the case of 8:2 diPAP. The reaction yields of the individual PFCAs formed from 8:2 diPAP are in good agreement with the results published by Houtz and Sedlak.27 During oxidation of6:2/8:2 diPAP all homologue PFCAs from C2–C9 were formed. Compared to the PFCA distribution after oxidation of 8:2 diPAP, it can be concluded that the formation of TFAA and PFPrA is attributed to the 6:2 FT chain, while the longest-chained product, PFNA, can only be derived from the 8:2 FT chain. The predominant products after oxidation of 6:2/8:2 diPAP were perfluoropentanoic acid (PFPeA), PFHxA, and PFHpA with molar recoveries between 13% (PFHpA in ultra-pure water) and 23% (PFPeA in soil matrix). The low probability of the occurrence of short-chain PFCAs from FT based precursors may be due to a stepwise degradation by ˙OH (and to a lesser extent sulfate) radicals starting at the non-fluorinated C atoms of the FT chain.39 Consequently, TFA and PFPrA only occur after oxidation of a precursor with a comparatively short FT chain like the 6:2 FT chain in 6:2/8:2 diPAP. Oxidative treatment of diSAmPAP led to a single reaction product, PFOA (soil matrix: 123%, matrix free: 118%), which is in accordance with previous results of the TOP assay applied to perfluorooctane sulfonamide-based precursors.6,27

The molar recoveries of the PFCA obtained from the individual precursors were used to determine mass balances. This resulted in total recoveries between 92 ± 4% (6:2/8:2 diPAP in H2O) and 123 ± 13% (diSAmPAP in soil matrix), and the balances were thus virtually closed.

The results reveal that TFAA and PFPrA can be formed in the TOP assay of FT-based precursors, as supposed by earlier users of the TOP assay6,29–31 and recently was proofed for PFPrA.31 Similar product formation has already been shown in oxidation experiments applying UV and H2O2 to 6:2 FT sulfonate.39

To investigate mixtures of sulfonamide-based precursors during the oxidative treatment, four FASAs were spiked onto the reference soil at two concentration levels in duplicate. After conducting the method, the FASAs were analyzed as their corresponding PFCAs. Based on the added amounts, mean recoveries and RSDs were calculated, which are listed in Table 1.

Table 1 Mean recoveries ([x with combining macron]Rec) and relative standard deviations (RSDs) of four PFCAs produced by oxidation of four FASAs at two concentration levels (25 μg per kg dw and 50 μg per kg dw, in duplicates)
Precursor Top assay product [x with combining macron] Rec/% RSD/%
FEtSA TFAA 88 12
FHxSA PFHxA 84 6
FOSA PFOA 103 16

Houtz and Sedlak investigated the behavior of FOSA in the TOP assay and found that it is quantitatively transformed to PFOA.27 A similar approach was applied here to three additional FASAs with C2, C4 and C6 chains, which were correspondingly transformed to TFAA, PFBA and perfluorohexanoic acid (PFHxA). The transformation recoveries and RSDs were satisfactory (between 65 ± 7% for perfluorobutyl sulfonamide (FBSA) and 103 ± 16% for FOSA).

For the first time, TFAA was included in an approach to estimate PFAA precursors based on the TOP assay. FT-based substances of 8:2 and 10:2 chains in turn lead to longer chained PFCAs. Thus, the developed method is suitable for legacy PFASs as well as modern short-chained alternative PFAS products.

Application of the improved protocol to a soil core

A soil core from a contaminated area in Baden-Württemberg, Germany, was divided into segments of 10 cm thickness, of which four segments (0–10 cm, 20–30 cm, 40–50 cm, and 60–70 cm) were analyzed for PFASs. PFAS concentrations were determined in the original sample and in samples after oxidation by the TOP assay. The results revealed PFOS and PFDA as the dominating PFASs in all soil horizons in the original samples, but with 9–10 times higher concentrations in the upper horizons, SEG1 and SEG2 compared to SEG3 and SEG4 (Fig. 4; individual concentrations in Tables S7 and S8). They can be considered as degradation products of polyfluorinated precursors.
image file: c9em00169g-f4.tif
Fig. 4 Concentrations of analytes in four segments of a vertical soil profile in the original sample (full bars) and after oxidative treatment (shaded bars); error bars: 1σ, n = 3; dw: dry weight.

After oxidative treatment in the TOP assay, concentrations of C2–C14 PFCAs increased considerably in the upper horizons (SEG1 and SEG2) with PFOA as the predominating compound. The individual PFCA levels of analytes quantified in the native sample of SEG1 increased by an average factor of 26 after oxidation. Some of the PFCAs (TFA, PFPrA and PFPeA) were not even present before oxidation. In the deeper horizons, SEG3 and SEG4, there was only a small increase of C2–C10 PFCAs after oxidation (on average 1.3-fold for analytes quantified in the native samples) and the concentrations were about 20 times lower than in SEG1 and SEG2.

SEG1 and SEG2 together, i.e. the segments from the plough horizon, contain 92% of the total mass concentration of PFASs in all four segments after oxidative treatment. This is consistent with previous observations of PFASs in soil profiles, which showed that the highest concentrations are present in the upper soil horizons 0–40 cm.40,41

The PFCA formation patterns in the upper horizons, SEG1 and SEG2, indicate a predominance of C8–C12 FT compounds and/or perfluorooctane sulfonamidoethanol (FOSE)-based PFASs. Due to the high sorption coefficients of long-chained PFASs, transport to deeper soil layers is accordingly slow, and the contaminants remain in the upper soil layers after application onto the topsoil. The almost equal concentrations in SEG1 and SEG2 are most likely due to mechanical tillage in agriculture. Typical ploughing depths are about 20–30 cm.

In total, the concentrations of PFAAs obtained after oxidative treatment of SEG1 add up to 5.0 μmol per kg dw, while the sum of the molar PFAA concentrations in the native SEG1 is 1.0 μmol per kg dw (not considering FOSA, because it is a precursor). Thus, 80% of the PFAAs analyzed after the TOP assay result from oxidation of precursors. In the other segments, precursors account for 72% (SEG2), 37% (SEG3), and 42% (SEG4) of the total PFCA concentrations. The analyzed precursors, FOSA, 8:2 diPAP and bis(1H,1H,2H,2H-perfluorooctyl) phosphate (6:2 diPAP), contribute small amounts (mass concentrations are included in the ESI, Table S8) to the molar concentrations of PFAA precursors. In the upper two segments examined, the contributions are 4.9% (FOSA), 7.6% (8:2 diPAP), and 3.4% (6:2 diPAP) in SEG1 and 6.2% (FOSA), 8.5% (8:2 diPAP), and 3.3% (6:2 diPAP) in SEG2, respectively. This also emphasizes the prevalence of long-chain precursors in the upper soil horizons. Regarding the total (molar) precursor concentrations of all segments, SEG1 and SEG2 together contain 98%. FT-based precursors with shorter or longer chain lengths (e.g. with 4:2, 6:2 or 14:2 alkyl chains) also contribute to the PFCA formation potential but they play only a minor role in the product pattern of the contamination. PFASs with short alkyl chains can move faster through the soil due to comparably low sorption coefficients, but the added fractions in SEG3 and SEG4 correspond to only 2% considering the total (molar) concentration of formed PFCAs after oxidation. The sum of PFAAs formed in SEG1 and SEG2 is 3200 μg per kg dw, whereas the sum in SEG3 and SEG4 is only 51 μg per kg dw.

Biotic and abiotic degradation of polyfluorinated precursors produce intermediates and PFAAs as terminal products. For example, microbial degradation of 8:2 FT or 6:2 FT precursors produces PFOA (C8) or PFHxA (C6) as the main product, respectively. FOSE-based precursors mainly yield PFOS as the final product and FOSA can be considered as an intermediate.11,38 Therefore, the occurrence of mainly PFOS and PFDA in the native samples of the upper horizons can be regarded as the result of microbial degradation of FOSE- and 10:2 FT-based PFAS or as product impurities. The sum of PFAAs in SEG1 is 1.0 μmol per kg dw and 1.6 μmol per kg dw in SEG2, which is about 20% and 28% of the total PFAAs after oxidative treatment (5.0 μmol per kg dw in SEG1 and 5.9 μmol per kg dw in SEG2). Interestingly, the upper two horizons do not show any TFAA and PFPrA and only rather small amounts of other short- and medium-chain PFCAs (C4–C8). The sums of detected C2–C8 PFCAs were 13.2 μg per kg dw in SEG1, 33.3 μg per kg dw in SEG2, 5.7 μg per kg dw in SEG3 and 4.7 μg per kg dw in SEG4. This can be interpreted by a faster displacement of short- and medium-length PFCAs (C2–C8) into deeper soil layers compared to C9–C14 PFCAs after their formation from precursors. The occurrence of short-chain PFAAs in groundwater samples from the same region as these samples further support this hypothesis.33


The analysis of PFCA patterns after oxidation with the TOP assay allows for the characterization of different kinds of PFAS precursors in soil and water samples, which may be unknown or not amenable to the analytical approach. Application of the TOP assay to a soil core demonstrated that long-chained precursors, such as legacy fluorinated paper chemicals or their breakdown products persist in the upper soil horizons (i.e. >40 cm) over long time periods. In contrast, short-chain breakdown products can be transported to deeper layers and, therefore, ultrashort dead-end products (C2- and C3-PFCAs) have to be considered to get a more complete picture of a PFAS contaminations. For this aim, an improved protocol of the TOP assay applicable to both, legacy and modern PFAS, as presented here, is required to obtain a highly efficient transformation of PFAS precursors for different sample matrices. Increased concentrations of oxidation chemicals furthermore require sample clean-up and fractionation to remove most of the high salt concentration for a reliable and accurate quantification of PFCAs by LC-MS-MS. The results of this work reveal the occurrence of PFAS precursors in the top horizons of a vertical soil profile. Since the contamination dates back at least 10 years the results also reveal the high sorption tendency and the persistence of those PFAS precursors. Deeper horizons down to 60 cm are characterized by very low concentrations of precursors with rather short perfluorinated carbon chain lengths. In conclusion, the application of the TOP assay in combination with powerful separation methods for a broad range of PFAA products (C2–C14) is recommended for any environmental research to get a more comprehensive picture of the input and fate of PFAS.

Conflicts of interest

There are no conflicts of interest to declare.


We highly acknowledge the donation of soil core samples from the Regierungspräsidium Karlsruhe in collaboration with the Landratsamt Rastatt. The non-polluted soil reference sample was received homogenized and dried from the Landwirtschaftliches Technologiezentrum Augustenberg (LTZ). The authors also thank Franziska Klein for sulfate analyses and the Baden-Württemberg Ministry of the Environment, Climate Protection and the Energy Sector for financial support within the program BWPLUS (project EOFplus, grant number L7517011).


  1. K. Prevedouros, I. T. Cousins, R. C. Buck and S. H. Korzeniowski, Sources, Fate and Transport of Perfluorocarboxylates, Environ. Sci. Technol., 2006, 40(1), 32–44 CrossRef CAS PubMed .
  2. E. Kissa, Fluorinated Surfactants and Repellents, 2nd Edition Revised and Expanded, Marcel Dekker, 2001 Search PubMed .
  3. C. A. Moody and J. A. Field, Perfluorinated Surfactants and the Environmental Implications of Their Use in Fire-Fighting Foams, Environ. Sci. Technol., 2000, 34(18), 3864–3870 CrossRef CAS .
  4. C. A. Moody, J. W. Martin, W. C. Kwan, D. C. Muir and S. A. Mabury, Monitoring Perfluorinated Surfactants in Biota and Surface Water Samples Following an Accidental Release of Fire-Fighting foam into Etobicoke Creek, Environ. Sci. Technol., 2002, 36(4), 545–551 CrossRef CAS PubMed .
  5. O. Weiss, G. A. Wiesmüller, A. Bunte, T. Göen, C. K. Schmidt, M. Wilhelm and J. Hölzer, Perfluorinated Compounds in the Vicinity of a Fire Training Area – Human Biomonitoring among 10 Persons Drinking Water from Contaminated Private Wells in Cologne, Germany, Int. J. Hyg. Environ. Health, 2012, 215(2), 212–215 CrossRef CAS PubMed .
  6. E. F. Houtz, C. P. Higgins, J. A. Field and D. L. Sedlak, Persistence of Perfluoroalkyl Acid Precursors in AFFF-Impacted Groundwater and Soil, Environ. Sci. Technol., 2013, 47(15), 8187–8195 CrossRef CAS PubMed .
  7. J. A. Field, D. L. Sedlak and L. Alvarez-Cohen, Characterization of the Fate and Biotransformation of Fluorochemicals in AFFF-Contaminated Groundwater at Fire/Crash Testing Military Sites. Final Report, 2017,, accessed Nov 22, 2017 Search PubMed .
  8. F. T. Lange, J. Müller and B. Körner, Entwicklung eines fluorspezifischen Gruppenparameters "EOF" für Boden und weitere Feststoffmatrices,, accessed Apr 9, 2019, 2017 Search PubMed .
  9. G. Nürenberg, K. Nödler, F. T. Lange, C. Schäfer, K. Huber and M. Scheurer, Nachweis von polyfluorierten Alkylphosphatestern (PAP) und Perfluoroktansulfonamidoethanol-basierten Phosphatestern (SAmPAP) in Böden, Mitt. Umweltchem. Ökotox., 2018, 24(2), 28–32 Search PubMed .
  10. C. Eschauzier, P. de Voogt, H.-J. Brauch and F. T. Lange, Polyfluorinated Chemicals in European Surface Waters, Ground- and Drinking waters, in Polyfluorinated Chemicals and Transformation Products, ed. T. P. Knepper and F. T. Lange, Springer, 2012, pp. 73–102 Search PubMed .
  11. H. Lee and S. A. Mabury, Global Distribution of Polyfluoroalkyl and Perfluoroalkyl Substances and Their Transformation Products in Environmental Solids, in Transformation Products of Emerging Contaminants in the Environment: Analysis, Processes, Occurrence, Effects and Risks, ed. D. A. Lambropoulou and L. M. L. Nollet, Wiley, 2014, pp. 797–826 Search PubMed .
  12. C. J. Young and S. A. Mabury, Atmospheric Perfluorinated Acid Precursors: Chemistry, Occurrence, and Impacts, in Reviews of Environmental Contamination and Toxicology, Vol. 208, Perfluorinated Alkylated Substances, D. A. Whitacre and P. de Voogt, Springer, 2010, pp. 1–109 Search PubMed .
  13. N. Yamashita, S. Taniyasu, G. Petrick, S. Wei, T. Gamo, P. K. S. Lam and K. Kannan, Perfluorinated Acids as Novel Chemical Tracers of Global Circulation of Ocean Waters, Chemosphere, 2008, 70, 1247–1255 CrossRef CAS PubMed .
  14. A. Gawor, C. Shunthirasingham, S. J. Hayward, Y. D. Lei, T. Gouin, B. T. Mmereki, W. Masamba, C. Ruepert, L. E. Castillo, M. Shoeib, S. C. Lee, T. Harner and F. Wania, Neutral Polyfluoroalkyl Substances in the Global Atmosphere, Environ. Sci.: Processes Impacts, 2014, 16(3), 404–413 RSC .
  15. R. Ghisi, T. Vamerali and S. Manzetti, Accumulation of Perfluorinated Alkyl Substances (PFAS) in Agricultural Plants: A review, Environ. Res., 2019, 169(2), 326–341 CrossRef CAS PubMed .
  16. C. Lau, K. Anitole, C. Hodes, D. Lai, A. Pfahles-Hutchens and J. Seed, Perfluoroalkyl Acids: A Review of Monitoring and Toxicological Findings, Toxicol. Sci., 2007, 99(2), 366–394 CrossRef CAS PubMed .
  17. U.S. Environmental Protection Agency, Determination of Selected Per- and Polyfluorinated Alkyl Substances in Drinking Water by Solid Phase Extraction and Liquid Chromatography/Tandem Mass Spectrometry (LC/MS/MS),, accessed Nov 22, 2018 Search PubMed.
  18. DIN 38407-42:2011-03, German Institute for Standardization (DIN e.V.), German Standard Methods for the Examination of Water, Waste Water and Sludge – Jointly Determinable Substances (Group F) – Part 42: Determination of Selected Polyfluorinated Compounds (PFC) in Water – Method Using High Performance Liquid Chromatography and Mass Spectrometric Detection (HPLC/MS-MS) after Solid–Liquid Extraction (F 42), Beuth, 2011.
  19. DIN 38414-14:2011-08, German Institute for Standardization (DIN e.V.), German Standard Methods for the Examination of Water, Waste Water and Sludge – Sludge and Sediments (Group S) – Part 14: Determination of Selected Polyfluorinated Compounds (PFC) in Sludge, Compost and Soil – Method Using High Performance Liquid Chromatography and Mass Spectrometric Detection (HPLC-MS/MS) (S 14), Beuth, 2011.
  20. A. Wagner, B. Raue, H.-J. Brauch, E. Worch and F. T. Lange, Determination of Adsorbable Organic Fluorine from Aqueous Environmental Samples by Adsorption to Polystyrene–Divinylbenzene Based Activated Carbon and Combustion Ion Chromatography, J. Chromatogr. A, 2013, 1295, 82–89 CrossRef CAS PubMed .
  21. S. Willach, H.-J. Brauch and F. T. Lange, Contribution of Selected Perfluoroalkyl and Polyfluoroalkyl Substances to the Adsorbable Organically Bound Fluorine in German Rivers and in a Highly Contaminated Groundwater, Chemosphere, 2016, 145, 342–350 CrossRef CAS PubMed .
  22. C. Bach, X. Dauchy, V. Boiteux, A. Colin, J. Hemard, V. Sagres, C. Rosin and J.-F. Munoz, The Impact of Two Fluoropolymer Manufacturing Facilities on Downstream Contamination of a River and Drinking Water Resources with Per- and Polyfluoroalkyl Substances, Environ. Sci. Pollut. Res., 2017, 24(5), 4916–4925 CrossRef CAS PubMed .
  23. L. W. Y. Yeung, A. O. De Silva, E. I. H. Loi, C. H. Marvin, S. Taniyasu, N. Yamashita, S. A. Mabury, D. C. G. Muir and P. K. S. Lam, Perfluoroalkyl Substances and Extractable Organic Fluorine in Surface Sediments and Cores from Lake Ontario, Environ. Int., 2013, 59, 389–397 CrossRef CAS PubMed .
  24. P. Wang, T. Wang, J. P. Giesy and Y. Lu, Perfluorinated Compounds in Soils from Liaodong Bay with Concentrated Fluorine Industry Parks in China, Chemosphere, 2013, 91(6), 751–757 CrossRef CAS PubMed .
  25. G. Codling, A. Vogt, P. D. Jones, T. Wang, P. Wang, Y.-L. Lu, M. Corcoran, S. Bonina, A. Li, N. C. Sturchio, K. J. Rockne, K. Ji, J.-S. Khim, J. E. Naile and J. P. Giesy, Historical Trends of Inorganic and Organic Fluorine in Sediments of Lake Michigan, Chemosphere, 2014, 114, 203–209 CrossRef CAS PubMed .
  26. B. Tan, T. Wang, P. Wang, W. Luo, Y. Lu, K. Y. Romesh and J. P. Giesy, Perfluoroalkyl Substances in Soils around the Nepali Koshi River: Levels, Distribution, and Mass Balance, Environ. Sci. Pollut. Res., 2014, 21(15), 9201–9211 CrossRef CAS PubMed .
  27. E. F. Houtz and D. L. Sedlak, Oxidative Conversion as a Means of Detecting Precursors to Perfluoroalkyl Acids in Urban Runoff, Environ. Sci. Technol., 2012, 46(17), 9342–9349 CrossRef CAS PubMed .
  28. M. Ateia, A. Maroli, N. Tharayil and T. Karanfil, The Overlooked Short- and Ultrashort-chain Poly- and Perfluorinated Substances: A Review, Chemosphere, 2019, 220, 866–882 CrossRef CAS .
  29. K. C. Harding-Marjanovic, E. F. Houtz, S. Yi, J. A. Field, D. L. Sedlak and L. Alvarez-Cohen, Aerobic Biotransformation of Fluorotelomer Thioether Amido Sulfonate (Lodyne) in AFFF-Amended Microcosms, Environ. Sci. Technol., 2015, 49(13), 7666–7674 CrossRef CAS PubMed .
  30. E. F. Houtz, R. Sutton, J.-S. Park and M. Sedlak, Poly- and Perfluoroalkyl substances in Wastewater: Significance of Unknown Precursors, Manufacturing Shifts, and Likely AFFF Impacts, Water Res., 2016, 95, 142–149 CrossRef CAS PubMed .
  31. D. Martin, G. Munoz, S. Mejia-Avendaño, S. V. Duy, Y. Yao, K. Volchek, C. E. Brown, J. Liu and S. Sauvé, Zwitterionic, cationic, and anionic perfluoroalkyl and polyfluoroalkyl substances Integrated into Total Oxidizable Precursor Assay of contaminated groundwater, Talanta, 2019, 195, 533–542 CrossRef CAS PubMed .
  32. S. Taniyasu, K. Kannan, L. W. Y. Yeung, K. Y. Kwok, P. K. S. Lam and N. Yamashita, Analysis of Trifluoroacetic Acid and Other Short-Chain Perfluorinated Acids (C2–C4) in Precipitation by Liquid Chromatography-Tandem Mass Spectrometry: Comparison to Patterns of Long-Chain Perfluorinated Acids (C5–C18), Anal. Chim. Acta, 2008, 619, 221–230 CrossRef CAS PubMed .
  33. J. Janda, K. Nödler, H.-J. Brauch, C. Zwiener and F. T. Lange, Robust Trace Analysis of Polar (C2–C8) Perfluorinated Carboxylic Acids by Liquid Chromatography-Tandem Mass Spectrometry: Method Development and Application to Surface Water, Groundwater and Drinking Water, Environ. Sci. Pollut. Res., 2019, 26(8), 7326–7336 CrossRef CAS PubMed .
  34. Z. Wang, I. T. Cousins, M. Scheringer and K. Hungerbühler, Fluorinated Alternatives to Long-Chain Perfluoroalkyl Carboxylic Acids (PFCAs), Perfluoroalkane Sulfonic Acids (PFSAs) and their Potential precursors, Environ. Int., 2013, 60, 242–248 CrossRef CAS .
  35. H. Siegel and M. Eggersdorfer, Ketones, in Ullmann's Encyclopedia of Industrial Chemistry, Wiley, 2010, pp. 187–207 Search PubMed .
  36. U. Berger, M. A. Kaiser, A. Kärrman, J. L. Barber and S. P. J. van Leeuwen, Recent Developments in Trace Analysis of Poly- and Perfluoroalkyl Substances, Anal. Bioanal. Chem., 2011, 400(6), 1625–1635 CrossRef CAS PubMed .
  37. M. Scheurer, K. Nödler, F. Freeling, J. Janda, O. Happel, M. Riegel, U. Müller, F. R. Storck, M. Fleig, F. T. Lange, A. Brunsch and H.-J. Brauch, Small, Mobile, Persistent: Trifluoroacetate in the Water Cycle-Overlooked Sources, Pathways, and Consequences for Drinking Water Supply, Water Res., 2017, 126, 460–471 CrossRef CAS PubMed .
  38. J. Liu and S. Mejia Avendaño, Microbial Degradation of Polyfluoroalkyl Chemicals in the Environment: a Review, Environ. Int., 2013, 61, 98–114 CrossRef CAS PubMed .
  39. X. Yang, J. Huang, K. Zhang, G. Yu, S. Deng and B. Wang, Stability of 6:2 Fluorotelomer Sulfonate in Advanced Oxidation Processes: Degradation Kinetics and Pathway, Environ. Sci. Pollut. Res., 2014, 21(6), 4634–4642 CrossRef CAS PubMed .
  40. A. Dreyer, S. Thuens, T. Kirchgeorg and M. Radke, Ombrotrophic Peat Bogs are not Suited as Natural Archives to Investigate the Historical Atmospheric Deposition of Perfluoroalkyl Substances, Environ. Sci. Technol., 2012, 46(14), 7512–7519 CrossRef CAS PubMed .
  41. J. W. Washington, H. Yoo, J. J. Ellington, T. M. Jenkins and E. L. Libelo, Concentrations, Distribution, and Persistence of Perfluoroalkylates in Sludge-Applied Soils near Decatur, Alabama, USA, Environ. Sci. Technol., 2010, 44(22), 8390–8396 CrossRef CAS PubMed .


Electronic supplementary information (ESI) available. See DOI: 10.1039/c9em00169g
Center for Pediatric and Adolescent Medicine, University Hospital Heidelberg, Im Neuenheimer Feld 669, 69120 Heidelberg, Germany

This journal is © The Royal Society of Chemistry 2019