Independent bachelor thesis in chemistry, 15 Hp School of science and technology
Örebro University
The study of organofluorine analysis applied to total
oxidizable precursor (TOP) assay to understand per-
and polyfluoroalkyl substances (PFASs)
Crystal Ho
Supervisor: Leo Yeung Examiner: Tuulia Hyötiläinen Date: 2018-05-30
2
Abstract
Per- and polyfluoroalkyl substances (PFASs) are synthetic chemicals which are used in a wide range of applications such as aqueous film forming foam (AFFF) for firefighting, paint, coating and cosmetics. For almost a decade, PFASs have received worldwide attention because of the ubiquitous detection of PFAS in the environment and their bioaccumulative, persistent and give toxic effects on animals and humans. A number of study have shown many unknown organofluorine present in environmental samples. The total oxidizable precursor (TOP) method is an oxidation method to convert any precursor compounds of PFASs into measurable perfluorinated carboxylic acids (PFCAs) or sulfonic acids (PFSAs). The aims of this project are to evaluate if the use of total oxidizable precursor (TOP) assay may help improve the mass balance analysis between the quantifiable PFASs and extractable organofluorine in samples.
In this study, three water samples: landfill leachate water, contaminated groundwater and a diluted Sthamex AFFF with tap water were undergone TOP assay to reveal the present of any unidentified precursor compounds. To obtain the extractable organofluorine and quantifiable PFAS levels in the samples, combustion ion chromatography (CIC) and liquid
chromatography – tandem mass spectrometry (LC-MS/MS) were used, respectively. Results suggested the presence of unidentified precursor compounds in the samples with the aid of TOP assay reaction. Although after the oxidative conversion, a large portion of unidentified PFASs compounds were still present in the samples. The quantifiable PFASs in the samples were up to 7 % of the EOF; in other word, there were still around 93 % of unidentified compounds present in the samples. Further development of analytical method for characterization or identification of these unknown PFAS compounds is needed.
Keywords: Poly-/per-fluoroalkyl substances, total oxidizable precursor (TOP) assay, AFFFs,
Table of content
INTRODUCTION ... 4
Background ... 5
Poly-/per-fluoroalkyl substances (PFASs) ... 5
Principle behind liquid chromatography – tandem mass spectrometry (LC-MS/MS) ... 6
PFAA precursors ... 7
Total oxidizable precursor (TOP) assay ... 7
Combustion ion chromatography (CIC) ... 7
Aims and limitations ... 8
MATERIALS AND METHODS ... 8
Chemicals ... 8
Analytical standards ... 8
Samples ... 8
Total oxidizable precursor (TOP) assay ... 9
Solid phase extraction (SPE)... 9
Instrumental analysis ... 9
Liquid chromatography – Tandem mass spectrometry (LC-MS/MS) ... 9
Combustion ion chromatography (CIC) ... 10
Quality assurance and quality control ... 10
Mass balance analysis of organofluorine ... 11
RESULTS AND DISCUSSION ... 12
1. The conversion of 6:2 FTSA in the positive/negative reaction (PR/NR) ... 12
2. Landfill leachate ... 13
3. Contaminated groundwater ... 15
4. Sthamex AFFF diluted with tap water ... 17
5. Mass balance analysis of fluorine... 19
RECOMMENDATIONS ... 21
CONCLUSION ... 21
ACKNOWLEDGEMENTS ... 21
REFERENCES... 22
4
Introduction:
Per- and polyfluoroalkyl substances (PFASs) have received worldwide attention because of their bioaccumulative, persistent and toxic effects on animals (Buck et al. 2011) and humans (Grandjean and Clapp 2014), resulting in increased concern to environment and health agencies (Wang et al. 2017). PFAS are synthetic chemicals having a carbon backbone in which the hydrogen atoms (H) are fully (Figure 1a) or partially (Figure 1b) substituted with fluorine atoms (F) and have various carbon chain lengths; these carbon-fluorine backbone is usually associated with a function group (i.e., carboxylate, sulfonate, or phosphate). The carbon-fluorine moiety not only renders them chemical stability but also the chemical properties which are useful in water and oil repellent coating on shoes, carpets and
firefighting foam (Kissa 2001). For many years industries have been using PFASs in many different applications such as aqueous film forming foam (AFFF) for firefighting, paint, coating and cosmetics; and these compounds have been the rise of global emission (KEMI 2015).
a) b)
Figure 1. Picture showing the chemical structure of a) perfluoroalkyl substance (e.g. perfluorooctane sulfonate, PFOS) and b) polyfluoroalkyl substance (e.g.6:2 fluorotelomer sulfonate)) (Yeung 2018).
According to a survey conducted by Swedish Chemical Agency, over 3000 of PFAS
compounds existed and were available on the global market (KEMI 2015). In Sweden, only 11 PFASs are monitored in drinking water sources under the National Food Agency (NFA) (https://www.livsmedelsverket.se/en/food-and-content/oonskade-amnen/miljogifter/pfas-in-drinking-water-fish-risk-management). Some research laboratories have been able to recognize more than 50 PFASs, but were still leaving out a broad part of unidentified organofluorine in the environment. A new method may help recognize any precursor compounds that may degrade to persistent perfluorinated carboxylic acids (PFCAs); the method are known as total oxidizable precursor (TOP) assay and was developed based on the formation of hydroxyl radicals by adding potassium persulfate, sodium hydroxide and high temperature to transform known/unknown PFAS precursors to PFCAs (Houtz and Sedlak 2012). By comparing the PFCA concentrations before and after oxidation, the concentrations of PFCA precursors may be inferred.
The intention of current study was to measure extractable organofluorine (EOF) and PFAS levels in environmental samples before and after TOP assay using combustion ion
chromatography (CIC) and liquid chromatography – tandem mass spectrometer (LC-MS/MS). In addition, a mass balance analysis of fluorine in the environmental samples before and after TOP assay was conducted. Initially, the results were planned to be compared with those from another laboratory, the Eurofins. However, due to the tight and busy schedule of the company, this comparison study was not carried out.
Background:
Poly-/per-fluoroalkyl substances (PFASs):
The fluorine substitution of PFAS gives unique properties; PFAS has a structure of a surfactant: a hydrophobic tail and a hydrophilic head, which has the ability to lower the surface tension between two phases. The hydrophobic part consists of carbons bonding to fluorines which has a water repellent function; on the other side the hydrophilic part consists of a water-soluble group (Figure 2). Sometimes, the spacer group may serve as the ‘bridge’ between the tail and the head which linked them together. Although, these parts have different properties it may correspond to many beneficial functions together. The water-soluble group can combine a broad range of groups such as carboxylates, phosphates or sulfonates (KEMI 2015). Moreover, the hydrophilic part are usually polar and may dissolve in water and can transport into blood which are bound to serum proteins that may associate with health issues.
Figure 2. Structure of a fluorosurfactant – the 6:2 fluorotelomer sulfonate (6:2 FTSA).
The structures of PFAS are very stable since the carbon-fluorine bond is known to be the strongest covalent bond that makes these compounds resistant to the ‘normal environmental’ degradation process and are generally biologically and chemically stable (Kissa 2001). Thereby, the thermal and chemical stability of the compound makes them suitable for different ranges of applications (KEMI 2015). Within the PFAS group it contains two major classes; perfluoroalkyl sulfonic acids (PFSAs) and perfluoroalkyl carboxylic acids (PFCAs). These chemicals may enter the environment from both direct emission and through
transformation of precursor compounds (indirectly) (Buck et al. 2011). The direct emission (source) are referred to the PFAS emission during manufacturing process and indirect emission are transformed of their precursors (Armitage, MacLeod, and Cousins 2009). Some of the PFASs have a long fluorinated carbon chain (C8), which show some bioaccumulation potential (Cousins 2013). Perfluorooctane sulfonate (PFOS) and
perfluorooctanoate (PFOA) are some common detectable PFASs in the environment which contain eight carbons. These compounds have been frequently used before the environmental and health concerned were brought up; compounds with shorter chain have then been a better choice. Shorter chains (compounds having less than seven fluorinated carbon) have been a substitute for PFOS and PFOA since they may have a lower bioaccumulation in the environment (Yeung, 2018).
In Europe, PFOS has been classified as a substance with very high concern (SVHC) under registration, evaluation, authorization and restriction of chemicals (REACH) (Regulation (EC) no. 1907/2006) and have been prohibited in EU since 2008 (European Directive
6 2006/122/CE). In 2009, PFOS and its salts were added to the Stockholm Convention list of persistent organic pollutants
(http://chm.pops.int/TheConvention/ThePOPs/TheNewPOPs/tabid/2511/Default.aspx). According to a survey conducted by Swedish Chemical Agency, over 3000 of PFAS
compounds exists and available on the global market after the replacement of PFOS (KEMI 2015).
Although the direct emissions of some PFASs have been stopped, the indirect emission of other unidentified samples are still present in the environment and in our daily life, meaning PFAS will still releasing some harmful components (Yeung 2018). Polyfluorinated substances that are present in the environment are degraded to perfluoroalkyl acids (PFAAs) e.g. PFSAs or PFCAs by natural processes (Houtz and Sedlak 2012); this is referred to as PFAA
precursors.
Alternatives of PFOS have been introduced into global market; however, industry do not need to show the chemical structures of these new PFOS alternatives because of the proprietary reason. Therefore, it is not known what chemicals have been released into the environment. However, it have been observed that number of industries have been making new chemicals with polyfluorinated moiety associated with different functional groups such as a betaine or thiosulfonate group (KEMI 2015). These polyfluorinated moiety has shown to degrade and give rise to different PFCAs (Houtz and Sedlak 2012). Therefore, with the use of the TOP assay, by comparing the PFCA concentrations before and after, the total concentration of PFAA precursor may be inferred.
Principles behind liquid chromatography – tandem mass spectrometry (LC-MS/MS)
Liquid chromatography and electrospray ionization (ESI) mass spectrometry is typically used for PFAS analysis since the compounds are already charged. The samples are first introduced to the LC column in which the samples are separated. The LC is coupled with a triple
quadrupole mass spectrometer (QqQ-MS) which is used for quantitation or structural information (Figure 3). The first (Q1) and third (Q3) quadrupoles serve as two mass filters and the second quadrupole (Q2) acts as a collision cell causing fragmentations of the analyte through the interaction with a collision gas argon (Kang 2012). The ions are formed after LC separation using electrospray ionization (ESI); all ions are selected based on their m/z ratio in Q1 and further to the collision cell. In Q2, a specific ion is selected and fragmented to
produce a daughter ions; the fragments which are created in Q2 will be detected in Q3 (Herbert and Johnstone 2003).
Figure 3. A schematic shows different parts of a triple quadrupole mass spectrometer (Picture adapted from Jackson Milcah through Wikipedia.org).
The advantages of LC-MS/MS are speed, robustness, high specificity and selectivity. Some disadvantages are the relatively expensive instrumentation and that the detector response is affected by matrix effects (i.e., ion suppression)(Pitt 2009).
PFAA precursors
Precursors of PFAA are compounds that have the capability of transforming to PFAAs such as PFCAs or PFSAs. One of the active ingredient in the formulation of aqueous film forming foam (AFFF) for firefighting is PFAAs (KEMI 2015). At some AFFF impacted sites, high levels of 6:2 fluorotelomer sulfonate (6:2 FTSA), up to 176 ng/L (Ahrens et al. 2015) has been reported. Although 6:2 FTSA may not be a dominant AFFF component, it is believed that the 6:2 FTSA is a degradation product from many other fluorotelomer precursors. In addition, 6:2 FTSA is also PFCA precursor and has shown to degrade to shorter chain PFCAs (Wang et al. 2017).
Total oxidizable precursor (TOP) assay
Total oxidizable precursor (TOP) assay is a method which aids with the conversion of PFSA or PFCA precursors into measurable PFCAs. As discussed above, PFAS can enter the environment either direct or indirect emissions. The indirect emission is through oxidative transformation of some polyfluorinated compounds such as 6:2 FTSA. To be capable to analyze these polyfluorinated compounds, total oxidizable precursor (TOP) assay can be used. This method is based on the formation of hydroxyl radicals through thermolysis of potassium persulfate in basic conditions (pH 8-14) with the supply of heat. The hydroxyl radicals will oxidize the precursor compounds. Reactions 1 and 2 shows how the hydroxyl radicals is converted from persulfate with heat into sulfate radicals and further into hydroxyl radicals (Houtz and Sedlak 2012).
Reaction 1 Reaction 2
Combustion ion chromatography (CIC)
There are several analytical methods that can detect more than 50 PFASs in sample, but it is still difficult to detect all the PFAS, especially the unidentified PFAS in the sample. However, with the aid of the combustion ion chromatography the total organofluorine (TOF) in the samples can be measured and this allows assessment of unknown PFASs. A mass balance analysis of fluorine is used to determine the amount of quantifiable PFAS which is accounted for the total organofluorine (TOF) in a sample whereas the amount of unidentified PFAS can be measured. The total organofluroine consists of both organic and inorganic fluorine. For the extractable organoflourine some unidentified organofluorine and PFASs can be detected. The amount of quantifiable PFASs is measured by LC-MS/MS.
8
Aims and limitations:
The aims of this project were I) to estimate if the use of TOP assay may help to improve the mass balance analysis of PFAS and the total/extractable organofluorine in samples, since many unknown organofluorines are present in the environmental samples; and II) to evaluate the reproducibility of TOP assay by analyzing the same samples in two different labs. Target compounds include:
PFCAs: PFBA, PFPeA, PFHxA, PFHpA, PFOA, PFNA, PFDA, PFUnDA, PFDoDA, PFTrDA, PFTDA, PFHxDA, PFOcDA;
PFCA precursors: 4:2 FTSA, 6:2 FTSA, 8:2 FTSA, 5:3 FTCA, 6:2 FTUCA, 7:3 FTCA, 8:2 FTUCA, 10:2 FTUCA;
PFSAs: PFBS, PFPeS, PFHxS, PFHpS, PFOS, PFNS, PFDS, PFDoDS; PFSA precursor: FOSA.
More details are given in Appendix – Table 1.
Materials and methods
Chemicals
Ammonia solution (NH4OH 25%), methanol (HPLC grade), sodium hydroxide pellets (NaOH) were purchased from Fisher Scientific. Potassium peroxodisulfate (K2O8S2, 99%), acetic acid (Glacial), N-methylpiperidine 99% and ammonium acetate (C2H7NO2) were from Sigma Aldrich. Hydrochloric acid (1 N) was purchased from Scharlab. Methanol (HPLC grade) was used to clean all the equipment and followed by Milli-Q water before use.
Analytical Standards
Potassium salts of PFOS, 13C
4-PFOS and perfluorobutanesulfonate (PFBS); sodium salts of perfluorodecanesulfonate (PFDS), perfluorohexanesulfonate (PFHxS) and 18O2-PFHxS, d3 -MeFOSAA, d5-EtFOSAA, perfluorooctanesulfonamide (FOSA), 13C8-FOSA,
perfluorohexanoate (PFHxA), 13C
2-PFHxA, perfluoroheptanoate (PFHpA), PFOA, 13C4 -PFOA, perfluorononanoate (PFNA), 13C5-PFNA, perfluorodecanoate (PFDA), 13C2-PFDA, perfluoroundecanoate (PFUnDA), 13C2-PFUnDA, perfluorooctanesulfonamido acetate (FOSAA; N-methyl and N-ethyl substituted: MeFOSAA, EtFOSAA), were obtained from Wellington lab (Guelph, ON).
Samples
The samples analyzed were three types of water samples; I) Leachate landfill water II) Contaminated groundwater III) 950 ml tap water and 50 ml Sthamex AFFF. The first two water samples were collected from different contaminated sites and given by Eurofins. The third samples was prepared by Dr. Anna Kärrman from Örebro University, more information can be found in (KEMI 2015). The landfill leachate water had a concentration of 65 000 ng/L which was measured by Eurofins which 11 PFAS were detected. For contaminated
groundwater, 16 PFAS were detected and had a concentration of 30 000 ng/L.
Table 1. Information of the samples.
Sample Landfill leachate water Contaminated groundwater
950 ml tap water and 50 ml Sthamex AFFF
Total oxidizable precursor (TOP) assay
Total oxidizable precursor assay was performed on the three water samples. The method was based on published method (Houtz and Sedlak 2012) with some modifications. In brief, 15 mL pp-tubes was used containing 8 mL of water sample, 1 mL of 150mM NaOH in Milli-Q and 6 mL of Milli-Q. To perform TOP assay both positive and negative reactions were required as a controller. For positive reaction (PR), 0,25 gram of potassium peroxodisulfate was also added into the tubes; whereas no potassium peroxodisulfate was added into the negative reaction (NR) tubes. Blanks using Milli-Q water for both positive and negative reactions were also added into the procedure and was treated as ‘normal’ samples to check any contamination. Before and after the water bath the pH value was measured with a pH-meter. Thereafter, the samples were placed into a water bath at 85°C for 6 hours to initiate the reactions. The reaction halted when the reacting samples were put on ice. Right before the solid phase extraction (SPE) the pH was adjusted to 2-6 with hydrochloride (HCl).
Solid phase extraction (SPE)
After the reaction all samples were extracted with solid phase extraction using a weak anion exchange sorbent (Oasis SPE – WAX 6 cc 30 µm (Water corporation, Milford, USA)) following published method (Yeung, Eriksson, and Kärrman 2016). The SPE cartridge was first conditioned with 4 mL of 0,1 % NH4OH in methanol, 4 mL of methanol and followed by 4 mL of Milli-Q. The next step was to load 15 mL of the sample into the cartridge. After all the samples was loaded onto the cartridge, 4 mL of 20 % MeOH in Milli-Q water was added into the original tube, and then the 4 mL solution loaded back onto the cartridge. The
cartridge was washed with 20 mL of 0,01 % of NH4OH to remove any inorganic fluoride that might have been enriched during the extraction, followed by 10 mL of Milli-Q and lastly 4 mL of the 25 mM ammonium acetate buffer. For elution, two fractions were obtained by adding 4 mL of MeOH into the cartridge in which the neutral fraction eluted first and the second anionic fraction was eluted when 4 mL of 0,1 % NH4OH in MeOH was added; they were collected separately.
The sample extracts were further divided into two ‘sets’: spike and non-spike samples; the spike sample was spiked with internal standard (IS) before the analysis with LC-MS/MS; whereas the non-spike sample was analyzed by combustion ion chromatography (CIC)
without spiking any IS. Although the samples had been extracted and cleanup by SPE, sample matrices might be still present in the sample causing ion suppression. Therefore, IS was spiked in the spike sample to determine the PFAS concentrations in the sample extract to correct any matrix effect. In order to determine the presence of any unidentified
organofluorine in the sample, mass balance analysis of fluorine was conducted by comparing the levels of organofluorine from CIC with the levels of PFAS from LC-MS/MS. The
reported PFAS levels were not recovery corrected.
Instrumental analysis
Liquid chromatography – tandem mass spectrometry (LC-MS/MS)
The samples extracts after SPE were transferred to LC vials for instrument analysis. The vials contained 200 µl of the sample, 300 µl of aqueous mobile phase (2mM ammonium acetate with 5mM 1-methylpiperidine added) and spiked with 1 ng of internal standard (table 2 in Appendix). The samples were analyzed by an Acquity UPLC system coupled to a triple quadrupole mass spectrometer XEVO TQ-S tandem mass spectrometer (MS/MS -Water Corporation) with negative ionization mode using the electrospray ionization. For separation a BEH column with a size of a 2.1 x 100 mm C18 (1.7 µm) was used. The mobile phases were prepared according to table 2, the mobile B and water phase were mixed together to get a
10 proportion of H2O:MeOH 70:30 to get mobile A. For the water phase; 0,150 g of ammonium acetate (2 mM) in 1000 mL Milli-Q water and for PAPs analysis 600 µl of 1-methylpiperidine were used. The mobile B had 0,045 g of ammonium acetate (2 mM) in 300 mL MeOH and 180 µl of 1-methylpiperidine.
Table 2. Description of the preparation proportion of the mobile phases for LC-MS/MS
Mobile phases Mobile A Mobile B
PFAS analysis H2O:MeOH 70:30, 2 mM NH4Ac, 5 mM 1-MP
MeOH, 2 mM NH4Ac, 5 mM 1-MP
Combustion ion chromatography (CIC)
The samples were prepared for the analysis by transferring 200 µl of the samples after SPE into plastic vials and then placed in the instrument. The total organofluorine in the samples were analyzed by a combustion ion chromatography, consisting of a 930 Compact IC flex with an 920 absorbent module (Metrohm) and coupled to a combustion module (Analytik Jena). For eluent carbonate buffer containing 64 mM sodium carbonate and 20 mM sodium bicarbonate was used. In brief, the sample were placed into the instrument and analyzed at 1000-1050°C for combustion. The procedure will convert all the organofluorine into
hydrogen fluoride (HF) which will be absorbed into the Milli-Q water. With the aid of the ion chromatography the concentration of F- in the samples was analyzed (Yeung, Eriksson, and Kärrman 2016).
Quality assurance and quality control
Two pre-trials of the chosen experimental method were performed before the ‘real’
experiment to gain a better knowledge about each part. With aid of some statistics it could be proven that after each trial the percentages of contamination decreased (data not shown). In order to minimize the risk of unintended contamination, the samples were stored in HDPE bottles collected by Eurofins. All the laboratory equipment were pre-cleaned three times with methanol and Milli-Q water. Procedural blanks (n=2) were included into each experiment; positive (PR) and negative (NR) controllers containing 6:2 FTSA were treated the same way as other samples. These controllers were included to show the oxidation reaction had taken place. Each water sample was analyzed with triplicate of PR and NR conditions to get a sense of the precision of the experiment. The 6:2 FTSA in the NR controllers and the procedural blanks did not show any peak for detectable target compounds which suggested no
Table 3 shows the limit of quantification (LOQ) value for each corresponding compound and this might be an explanation for why some compounds could not be detected.
Table 3. Limits of quantification (LOQ) value for each target compound.
Compound LOQ (ng/mL) Compound LOQ (ng/mL)
PFBS 0,02 PFOA 0,02 PFPeS 0,02 PFNA 0,02 PFHxS 0,02 PFDA 0,02 PFOS 0,02 PFUnDA 0,02 PFNS 0,02 4:2 FTSA 0,02 PFDoDS 0,02 6:2 FTSA 0,02 PFHxA 0,02 8:2 FTSA 0,02 PFHpA 0,02
Compound LOQ (ng/mL) Compound LOQ (ng/mL)
PFBA 0,04 6:2 FTUCA 0,04 PFPeA 0,04 7:3 FTCA 0,04 PFDoDA 0,04 8:2 FTUCA 0,04 PFTrDA 0,04 10:2 FTUCA 0,04 PFTDA 0,04 PFHxDA 0,04 PFOcDA 0,04 5:3 FTCA 0,04
Mass balance analysis of organofluorine
The measured PFAS concentration (ng/g) in the samples given from the LC-MS/MS were converted into fluoride concentration (ng F/g) using the following equation.
12
Results and discussion
1. The conversion of 6:2 FTSA in the positive/negative reaction (PR/NR):
In the current study, 6:2 FTSA was used as an indicator to show if the oxidation takes place or not, because it has been shown that it can be oxidized based on previous studies (Söderlund 2018; Larsson 2018) under these experimental conditions. To assure that the oxidation conversion is working, the detection in the NR controller or the disappearance in PR
controller of 6:2 FTSA was monitored in the three different batches of experiments. In the PR controller, if the 6:2 FTSA is consumed, the level decreases which implies that the oxidation is ongoing; at the same time different transformation products such as PFBA, PFPeA, PFHxA and PFHpA will be formed accordingly. However, in the PR controller, if 6:2 FTSA is not decreasing or increasing, this indicates that the reaction is not working or contamination occurs. As for the NR controller, the 6:2 FTSA should remain. If the 6:2 FTSA decreases in the NR controller, then it suggest some other unknown reaction is taking place.
In current study, in the NR controllers, 6:2 FTSA remained, and there was no detectable of other transformation products (Table 4 and Table 3 in Appendix); the variability of the 6:2 FTSA in the NR controller was found to be 5,4 % across the three batches of experiment, which suggest the NR experiments performed were reproducible.
In the PR controllers, all 6:2 FTSA was consumed (not detectable) and the transformation products of 6:2 FTSA included PFBA, PFPeA, PFHxA and PFHpA were detected (Table 4 and Table 3 in Appendix). The transformation products of 6:2 FTSA were 15,2 % of PFPeA and 10,2 % of PFHxA, 7,4 % of PFBA and 2,7 %. This result indicates that the reaction was ongoing as expected as all the 6:2 FTSA were consumed and transformed into other products. The variability of the batches ranged between 0,9 to 2,5 %, which shows that on three
different days, results were reproducible in the PR controller.
Table 4. Results showing the reactant, the 6:2 FTSA and its transformation product in the NR and PR controllers on 3 different experimental days. Result given the mean in pmole, standard deviation (SD) and percent (%) (empty cells indicates sample below LOQ)
pmole % Mean SD Mean SD NR 6:2 FTSA (C8) 1,32 0,07 5,4 PFBA (C4) PFPeA (C5) PFHxA (C6) PFHpA (C7) PR 6:2 FTSA (C8) PFBA (C4) 0,099 0,017 7,4 0,9 PFPeA (C5) 0,202 0,043 15,2 2,5 PFHxA (C6) 0,136 0,036 10,2 2,2 PFHpA (C7) 0,037 0,016 2,7 1,1
2. Landfill leachate
The results of landfill leachate waters in the NR and PR conditions are given in Table 5 and Figure 4. A total number of nine PFASs showed detectable concentrations in the sample in the NR condition; whereas eight PFASs showed detectable levels in PR condition. By comparing the NR and PR conditions in Table 5, it can be seen that 6:2 FTSA and 4:2 FTSA have been decreased fully and the levels of some compounds such as PFBS, PFOS, PFBA, PFPeA, PFHxA, and PFHpA were increased (Table 5 and Figure 4). The increases in PFBA, PFPeA, PFHxA and PFHpA were believed to be due to the degradation of the FTSAs based on the results of Table 4. The increases of PFBS and PFOS might be due to the degradation of some unknown PFBS or PFOS precursors.
Figure 4. The composition profiles (%) of different PFASs in landfill leachate water. Negative reaction (NR) did not undergo any oxidative conversion by TOP assay; the positive reaction (PR) underwent oxidative conversion.
NR
14
Table 5. The mean, standard deviation (SD), and percent change (%) of target analytes in the PR and NR conditions in the sample. The empty cells indicates compound was found to below its limit of quantification (LOQ). Percent change is
calculated by the concentrations of (PR-NR)/NR*100. This symbol ( * ) in the percent change (%) indicate that the values are significant after Student t Test. The significance level was set at α= 0.05.
Landfill leachate water
PR (n=3) NR (n=3)
Mean SD Mean SD Percent change
(ng/mL) (%) PFBS 0,0846 0,009 0,058 0,001 45* PFPeS PFHxS 0,011 0,001 0,012 0,002 -3 PFHpS PFOS 0,013 100 PFNS PFDS PFDoDS PFBA 0,083 0,026 0,028 0,005 196* PFPeA 0,785 0,041 0,544 0,040 44* PFHxA 0,315 0,045 0,175 0,016 80* PFHpA 0,137 0,010 0,088 0,001 56* PFOA 0,140 0,003 0,112 0,011 25* PFNA PFDA PFUnDA PFDoDA PFTrDA PFTDA PFHxDA PFOcDA 4:2 FTSA 0,0162 0,0026 -100 6:2 FTSA 0,540 0,220 -100 8:2 FTSA 5:3 FTCA 6:2 FTUCA 7:3 FTCA 8:2 FTUCA 10:2 FTUCA
3. Contaminated groundwater
The results of contaminated groundwater in the NR and PR conditions are given in Table 6 and Figure 5. A total number of 11 PFASs showed detectable concentrations in the sample in the PR and NR conditions. By comparing the NR and PR conditions in Table 6, it can be seen that the contamination of 6:2 FTSA has been decreased and the levels of some compounds such as PFBS, PFPeS, PFOS, PFBA, PFPeA, PFHxA, PFHpA and PFOA increased (Table 6 and Figure 5). The increases in PFBA, PFPeA, PFHxA, and PFHpA were believed to be due to the degradation of the 6:2 FTSA. An extremely high increase of PFBA with 1577 % was observed, which suggest some other unknown PFBA precursor present in the sample that degraded to PFBA.
The increases of PFBS, PFPeS and PFOS might be due to the degradation of some unknown precursors.
Figure 5. The composition profiles (%) of different PFASs in contaminated groundwater. Negative reaction (NR) did not undergo any oxidative conversion by TOP assay; the positive reaction (PR) underwent oxidative conversion.
NR
16
Table 6. The mean, standard deviation (SD), and percent change (%) of target analytes in the PR and NR conditions in the sample. The empty cells indicates compound was found to below its limit of quantification (LOQ). Percent change is
calculated by the concentrations of (PR-NR)/NR*100. This symbol ( * ) in the percent change (%) indicate that the values are significant after Student t Test. The significance level was set at α= 0.05.
Contaminated groundwater
PR (n=3) NR (n=3)
Mean SD Mean SD Percent change
(ng/mL) (%) PFBS 0,30 0,016 0,247 0,017 21* PFPeS 0,462 0,03 0,306 0,0385 51* PFHxS 1,51 0,154 1,09 0,0903 39 PFHpS 0,135 0,0076 0,106 0,013 27 PFOS 1,496 0,102 1,07 0,067 40* PFNS PFDS PFDoDS PFBA 0,520 0,021 0,031 0,015 1577* PFPeA 1,02 0,094 0,608 0,020 68* PFHxA 2,24 0,308 0,867 0,043 158* PFHpA 0,165 0,001 0,133 0,011 24* PFOA 0,247 0,011 0,167 0,012 48* PFNA PFDA PFUnDA PFDoDA PFTrDA PFTDA PFHxDA PFOcDA 4:2 FTSA 6:2 FTSA 0,0683 0,0051 0,903 0,057 -92* 8:2 FTSA 5:3 FTCA 6:2 FTUCA 7:3 FTCA 8:2 FTUCA 10:2 FTUCA
4. Sthamex AFFF diluted with tap water
The results of Sthamex AFFF diluted with tap water in the NR and PR conditions are given in Table 7 and Figure 6. A total number of 16 PFASs showed detectable concentrations in the sample in the PR and NR conditions. By comparing the NR and PR conditions in Table 7, it can be seen that 6:2 FTSA and 8:2 FTSA have a slightly decrease and at the same time PFNA also decreased. The levels of some compounds such as PFBA, PFPeA, PFHxA, PFHpA, PFOA, PFDA, PFDoDA, PFTrDA and PFTDA increased (Table 7 and Figure 6). The increases in PFBA, PFPeA, PFHxA, PFHpA and PFOA were believed to be due to the
degradation of the 6:2 and 8:2 FTSAs. At the same time 6:2 and 8:2 FTSAs were still detected in the PR condition, which suggests that the reaction did not go as expected or there are unknown precursors that might give rise to 6:2 and 8:2 FTSAs. However, increases have been found for some PFCAs suggesting that the reaction was still undergoing. A report “Chemical analysis of selected fire-fighting foams on the Swedish market” identified 6:2 fluorotelomer sulfonamide alkylbetaine (6:2 FTAB) was one of the active ingredient in Sthamex AFFF (KEMI 2015). Therefore, when 6:2 FTAB oxidized it might give rise to 6:2 FTSA; at the same time the 6:2 FTSA in the sample might have been oxidized but 6:2 FTSA might be formed from 6:2 FTAB. The increases of PFDA, PFDoDA, PFTrDA and PFTDA are due to some fluorotelomer-based compounds such as 12:2 or 14:2. The decrease of PFNA in the reaction is unknown because the perfluorinated compounds are persistent.
The incomplete oxidation of 6:2 and 8:2 FTSAs in PR suggests two possibilities: I) more amount of potassium peroxodisulfate (the reactant) is needed to complete the oxidation or II) longer reaction time is needed.
Figure 6. The composition profiles (%) of different PFASs in contaminated groundwater. Negative reaction (NR) did not undergo any oxidative conversion by TOP assay; the positive reaction (PR) underwent oxidative conversion.
NR
18
Table 7. The mean, standard deviation (SD), and percent change (%) of target analytes in the PR and NR conditions in the sample. The empty cells indicates compound was found to below its limit of quantification (LOQ). Percent change is
calculated by the concentrations of (PR-NR)/NR*100. This symbol ( * ) in the percent change (%) indicate that the values are significant after Student t Test. The significance level was set at α= 0.05.
Sthamex AFFF diluted with tap water
PR (n=3) NR (n=3)
Mean SD Mean SD Percent change
ng/mL (%) PFBS PFPeS PFHxS PFHpS PFOS PFNS PFDS PFDoDS PFBA 4,21 0,251 0,64 0,109 557* PFPeA 3,07 0,274 2,04 0,221 50* PFHxA 4,10 1,06 0,18 0,02 2178* PFHpA 1,07 0,148 0,04 0,001 2575* PFOA 2,17 0,059 0,15 0,042 1347* PFNA 0,320 0,061 0,536 0,097 -40* PFDA 1,38 0,151 0,70 0,077 97* PFUnDA 0,203 0,011 0,177 0,006 15 PFDoDA 0,945 0,142 0,474 0,027 99* PFTrDA 0,109 0,004 0,075 0,017 45* PFTDA 0,652 0,124 0,309 0,067 111* PFHxDA 0,236 0,087 0,129 0,017 83 PFOcDA 0,988 0,234 0,552 0,102 79 4:2 FTSA 0,019 0,005 0,048 0,0003 -60* 6:2 FTSA 11,5 1,57 12,71 1,30 -10* 8:2 FTSA 15,1 0,698 15,07 1,35 0,2 5:3 FTCA 6:2 FTUCA 7:3 FTCA 8:2 FTUCA 10:2 FTUCA
5. Mass balance analysis of fluorine:
The results for the mass balance of organofluorine between known PFAS and EOF for landfill leachate water and contaminated groundwater can be seen in Table 8a and b. The results for Sthamex AFFF are not reported because of some issues in IC analysis (Figure 7b). Figure 7a shows the chromatogram for landfill leachate water, which shows a clear separation of fluoride. However, for Sthamex AFFF the chromatogram shows some strange peak form; a ‘shark-fin” like peak was observed on the chromatogram. The peak appears a few times broader than the other peaks showed in Figure 7a. This might be due to two or more
compounds co-eluting or the column was overloaded. Dilution of the sample may help with reducing over-loading the column or other column with different separating principle may help resolve co-eluting species. Therefore, the results of EOF Sthamex AFFF were not reported.
Normal chromatogram showing fluoride in landfill leachate water
Figure 7a. A normal chromatogram of landfill leachate water which was obtained from the CIC.
Abnormal chromatogram showing fluoride in Sthamex AFFF diluted with tap water
Figure 7b. The abnormal chromatogram of Sthamex AFFF diluted with tap water which was obtained from the CIC.
For the landfill leachate water, the quantifiable PFAS between PR and NR did not differ much (Table 8a) as compared to contaminated groundwater that significant increase in the PR condition was found. For the landfill leachate water the quantifiable PFAS accounted for
20 4,3 % in NR and 4,6 % in PR conditions. For contaminated groundwater the quantifiable PFAS accounted for 4,9 % in NR and 7 % in PR conditions. The use of TOP assay was used to demonstrate the present of identified PFCA or PFSA precursors. However, after TOP assay precursor compounds may give rise to some unidentified intermediate, and therefore, will not detected. The results shows that the quantifiable PFAS increased in contaminated
groundwater which helps to increase the percentage of known PFCAs in the sample by 3 %. These results suggest significant amount of fluorine still remain unidentified after TOP assay.
Table 8a. The concentration (ng F/mL) of quantifiable and unidentified PFAS in PR and NR for samples 1. n=3 indicate it is based on triplicates.
Landfill leachate water PR (n=3) NR(n=3)
Mean SD Mean SD C onc en tr at ion ng F /m L Quantifiable PFAS concentrations by LC-MS/MS 1,02 0,04 0,93 0,16 EOF by CIC 22,1 3,5 21,5 0,8 Unidentified PFAS 21,1 20,5 C om pos iti on % %of quantifiable PFAS to EOF 4,6 4,3 %of unidentified PFAS 95,4 95,7
Table 8b. The concentration (ng F/mL) of quantifiable and unidentified PFAS in PR and NR for samples 2. n=3 indicate it is based on triplicates. Contaminated groundwater PR (n=3) NR(n=3) Mean SD Mean SD C onc ent ra ti on ng F /m L Quantifiable PFAS concentrations by LC-MS/MS 5,23 0,50 3,47 0,23 EOF by CIC 74,4 5,6 71,4 6,3 Unidentified PFAS 69,1 67,9 C om pos iti on % %of quantifiable PFAS to EOF 7,0 4,9 %of unidentified PFAS 93,0 95,1
Part of the largely unidentified PFASs in the samples could be some ultrashort-chain C2 and C3 PFCAs, because these compounds were not analyzed in this study. In the earlier study (Larsson 2018), the aim of the study was to measure how much the ultra-short chain (C2-C3)
PFASs could contribute to overall concentrations of PFAS in aqueous film forming foam for firefighting (AFFF). The C2 and C3 compounds contributed for up to 1 % to the overall concentrations. Since that study was mainly focused in firefighting foam, it should not be applied to current study which were on environmental water samples. Another study on environmental water samples of ultrashort-chain C2 and C3 have been reported to contribute to over 40 % of the total detectable compounds (Yeung, Stadey, and Mabury 2017). Some of the unidentified percentages of PFAS in current study might be the ultra-short chain C2 and C3. If this speculation is valid, the 40 % of the unidentified PFAS might be these ultra-short chain compound. This will narrow down the amount of the unidentified PFAS compounds in the samples.
Recommendations
One of the initial aims was to evaluate the reproducibility of TOP assay by analyzing the same samples in two different labs – the Eurofins. However, this did not occur and would have been splendid to be able to compare the results from different labs to measure the reproducibility.
Conclusion
From the analyses of the water samples, it can be concluded that PFSA and PFCA precursors may be detected with the aid of TOP assay reaction. It also showed that unidentified PFCA precursors were present in the samples which may have contributed to the increase
concentrations of detectable compounds in quantifiable PFAS.
After oxidative conversion, it still showed a large portion of unidentified PFAS compounds present in the samples. This study could only determine a small fraction of the total
organofluorine and PFAS levels in environmental samples. The known PFASs in the samples are up to 7 % and on the other hand there are still around 93 % of unidentified compounds in the samples. Since there are still lot of unidentified components out there, further
development of analytical method for the unidentified PFASs is needed.
Acknowledgements
I would like to thank my supervisor Leo Yeung for all his help and support throughout the project. I would also like to thank everyone working at MTM for their wonderful help and advice during my lab work there.
22
References
Ahrens, Lutz, Karin Norström, Tomas Viktor, Anna Palm Cousins, and Sarah Josefsson. 2015. “Stockholm Arlanda Airport as a Source of Per- and Polyfluoroalkyl Substances to Water, Sediment and Fish.” Chemosphere 129 (June): 33–38.
https://doi.org/10.1016/j.chemosphere.2014.03.136.
Armitage, James M., Matthew MacLeod, and Ian T. Cousins. 2009. “Modeling the Global Fate and Transport of Perfluorooctanoic Acid (PFOA) and Perfluorooctanoate (PFO) Emitted from Direct Sources Using a Multispecies Mass Balance Model.”
Environmental Science & Technology 43 (4): 1134–40.
https://doi.org/10.1021/es802900n.
Buck, Robert C., James Franklin, Urs Berger, Jason M. Conder, Ian T. Cousins, Pim de Voogt, Allan Astrup Jensen, Kurunthachalam Kannan, Scott A. Mabury, and Stefan P. J. van Leeuwen. 2011. “Perfluoroalkyl and Polyfluoroalkyl Substances in the
Environment: Terminology, Classification, and Origins.” Integrated Environmental
Assessment and Management 7 (4): 513–41. https://doi.org/10.1002/ieam.258.
Cousins, Ian T. 2013. “Nordic Research on Per- and Polyfluoroalkyl Substances (PFASs).”
Environmental Science and Pollution Research 20 (11): 7926–29.
https://doi.org/10.1007/s11356-013-2000-7.
Grandjean, Philippe, and Richard Clapp. 2014. “Changing Interpretation of Human Health Risks from Perfluorinated Compounds.” Public Health Reports (Washington, D.C.:
1974) 129 (6): 482–85. https://doi.org/10.1177/003335491412900605.
Herbert, Christopher G., and R. A. W. Johnstone. 2003. Mass Spectrometry Basics. Boca Raton: CRC Press.
Houtz, Erika F., and David L. Sedlak. 2012. “Oxidative Conversion as a Means of Detecting Precursors to Perfluoroalkyl Acids in Urban Runoff.” Environmental Science &
Technology 46 (17): 9342–49. https://doi.org/10.1021/es302274g.
Kang, Ju-Seop. 2012. “Principles and Applications of LC-MS/MS for the Quantitative Bioanalysis of Analytes in Various Biological Samples.” In Tandem Mass
Spectrometry - Applications and Principles, edited by Jeevan Prasain. InTech.
https://doi.org/10.5772/32085.
KEMI. 2015. “Occurrence and Use of Highly Fluorinated Substances and Alternatives.”
Swedish Chemical Agency, no. 7/15.
Kissa, Erik. 2001. Fluorinated Surfactants and Repellents. 2nd ed., rev. And expanded. Surfactant Science Series, v. 97. New York: Marcel Dekker.
Larsson, Pontus. 2018. “A Study to Understand the Information Gap between Total Organofluorine Analysis and Total Oxidizable Precursor Assay on
Polyfluoroalkyl/Perfluoroalkyl Substances (PFASs).” Örebro University. Pitt, James J. 2009. “Principles and Applications of Liquid Chromatography-Mass
Spectrometry in Clinical Biochemistry.” The Clinical Biochemist. Reviews 30 (1): 19– 34.
Söderlund, Lydia. 2018. “Method Development of Total Oxidizable Precursor Assay for Perfluoroalkyl Acid Precursors in Domestic Sludge.” Örebro University.
Wang, Zhanyun, Jamie C. DeWitt, Christopher P. Higgins, and Ian T. Cousins. 2017. “A Never-Ending Story of Per- and Polyfluoroalkyl Substances (PFASs)?” Environmental
Science & Technology 51 (5): 2508–18. https://doi.org/10.1021/acs.est.6b04806.
Yeung, L. W. Y. 2018. “Source, Fate and Transport of Highly Fluorinated Contaminants in the Environments.” Örebro University, 2018.
Yeung, L. W. Y., Ulrika Eriksson, and Anna Kärrman. 2016. “A Pilot Study on Unidentified Poly- and Perfluoroalkyl Substances (PFASs) in Sewage in Sweden.” Örebro
University, no. diva2:1112388.
Yeung, Leo W.Y., Christopher Stadey, and Scott A. Mabury. 2017. “Simultaneous Analysis of Perfluoroalkyl and Polyfluoroalkyl Substances Including Ultrashort-Chain C2 and C3 Compounds in Rain and River Water Samples by Ultra Performance Convergence Chromatography.” Journal of Chromatography A 1522 (November): 78–85.
24
Appendix A
Table 1. The target list of PFASs abbreviation, carbon chain lengths and their names.
Carbon chain length Abbreviation Name
PFSA Perfluorosulfonic acid
4 PFBS Perfluorobutane sulfonic acid
5 PFPeS Perfluoropentane sulfonic acid
6 PFHxS Perflurohexane sulfonic acid
7 PFHpS Perfluoroheptane sulfonic acid
8 PFOS Perfluorooctane sulfonic acid
9 PFNS Perfluorononane sulfonic acid
10 PFDS Perfluorodecane sulfonic acid
12 PFDoS Perfluorododecane sulfonic acid
PFCA Perfluorocarboxylic acid
4 PFBA Perfluorobutanoic acid
5 PFPeA Perfluoropentanoic acid
6 PFHxA Perfluorohexanoic acid
7 PFHpA Perfluoroheptanoic acid
8 PFOA Perfluorooctanoic acid
9 PFNA Perfluorononanoic acid
10 PFDA Perfluorodecanoic acid
11 PFUnDA Perfluoroundecanoic acid
12 PFDoDA Perfluorododecanoic acid
13 PFTrDA Perfluorotridecanoic acid
14 PFTDA Perfluorotetradecanoic acid
FTSA Fluorotelomer sulfonic acid
6 4:2 FTSA 4:2 Fluorotelomer sulfonic acid
8 6:2 FTSA 6:2 Fluorotelomer sulfonic acid
Table 2. A List of internal standard (IS) used before the instrumental analysis. Name Trace IS_6:2 FTSA 429 > 409 IS_8:2 FTSA 529 > 509 IS_PFBA 216.97 > 172 IS_PFPeA 265.97 > 222 IS_PFHxA 314.97 > 270 IS_PFHpA 366.97 > 322 IS_PFBS 301.9 > 98.9 IS_PFHxS 402.9 > 102.9 IS_PFOS 502.97 > 98.96 IS_PFOA 416.97 > 372 IS_PFNA 471.99 > 427 IS_PFOSA 505.9 > 77.8 IS_PFDA 514.97 > 470 IS_PFUnDA 564.07 > 520 IS_PFTDA 714.9 > 670 IS_PFDoDA 614.97 > 570 IS_PFHxDA 814.9 > 770 IS_6:2 FTUCA 358.9 > 293.91 IS_10:2 FTUCA 558.84 > 493.82 IS_8:2 FTUCA 458.9 > 393.94
26
Table 3. The result given in pmole and percent (%) whereas 6:2 FTSA converts into its transformation products.
S1 pmole 6:2 FTSA % 6:2 FTSA 1,34 PFBA 0,094 9,4 PFPeA 0,198 19,8 PFHxA 0,128 12,8 PFHpA 0,027 2,7 S2 pmole 6:2 FTSA % 6:2 FTSA 1,25 PFBA 0,084 8,4 PFPeA 0,160 16,0 PFHxA 0,104 10,4 PFHpA 0,027 2,7 S3 pmole 6:2 FTSA % 6:2 FTSA 1,39 PFBA 0,117 11,7 PFPeA 0,247 24,7 PFHxA 0,176 17,6 PFHpA 0,055 5,5
To obtain how much pmole and percentages of 6:2 FTSA contained in a specific compound Equation 1a and b were used.
Table 4. The concentration (ng/mL) of the transformation products of 6:2 FTSA
PFBA PFPeA PFHxA PFHpA
0,094 0,198 0,128 0,027
0,084 0,160 0,104 0,027
0,117 0,247 0,176 0,055
pg mole (6:2 FTSA)=.-/01, #' 2#*- #' ,1- %#23#4$5 #' /$,-6-7,"#$%.#'
)$)*+,-Equation 1a. Used to calculate how much pg mole of 6:2 FTSA was detected in specific compounds
Equation 1b. Used to convert the pg mole into percent (%)
