Analytical Science Programme in Chemistry with a focus on Forensics
Bachelor thesis, 15hp
Method development of total oxidizable
precursor assay for perfluoroalkyl acid
precursors in domestic sludge
Lydia Söderlund
Supervisor: Leo Yeung
Examinator: Anna Kärrman
Date: 2018-01-21
Abstract
Per- and polyfluoroalkyl substances (PFASs) are persistent organic pollutants used in industrial applications and are globally distributed in the environment. A group of PFASs that are difficult to measure with today’s method are perfluoroalkyl acid precursors (PFAA precursors) that, when degraded, serves as indirect sources of PFAAs. This study has optimized a previously developed method for quantification of PFAA precursors in soil; through total oxidizable precursor assay (TOP assay) under alkaline conditions, to be applicable on sewage sludge. To achieve and maintain an alkaline environment during the entire oxidative treatment, several parameters were tested: concentrations of NaOH, persulfate and sample; additional clean-up with graphitized non-porous carbon and reaction time. Solid phase extraction-weak anion exchange (SPE-WAX) was used for clean-up and separation of analytes, and LC-MS/MS was used for quantification. The optimal conditions with the highest levels of PFAAs detected was obtained with 1.33 M NaOH, 60 mM persulfate, 3.57 g/L sludge with a reaction time of 6 hours. The use of graphitized non-porous carbon reduced matrix effects on oxidative conversion resulting in a higher pH as well as a higher degree of oxidation, but with some analyte loss.
Keywords: Total oxidizable precursor assay (TOP assay), Perfluoroalkyl substances (PFASs),
Table of Content
Abstract ... 2
Introduction ... 4
Aims and Limitations ... 4
Background ... 5
Polyfluoroalkyl and/or perfluoroalkyl substances (PFASs) and oxidative conversion ... 5
Total oxidizable precursor assay (TOP assay) ... 6
PFAA precursors ... 6
Sewage Sludge ... 7
Method ... 8
Materials ... 8
Stock testing sample ... 8
Graphitized non-porous carbon treatment ... 8
Total oxidizable precursor (TOP) assay ... 8
Solid phase extraction ... 8
LC-MS/MS analysis ... 9
Results ... 10
pH after reaction ... 10
Oxidative Transformation... 11
Relative standard deviation... 12
Recovery ... 13
Discussion ... 14
pH ... 14
Graphitized non-porous carbon’s effect on target compounds... 14
Relative standard deviation... 15
Conclusion ... 16
Introduction
Poly- and perfluoroalkyl substances (PFASs) are anthropogenic compounds where one or more of the hydrogen atoms have been substituted with fluoride atoms, giving them properties useful for water repellence and tension lowering. Since the 1950’s, industries have used PFASs in a wide range of applications, resulting in a global emission of these compounds. While we can analyse more than 70 PFASs with current methods, many of them remain not studied because of unknown identity and the availability of authentic standards, leaving a large part of the organofluorine unidentified. To
overcome this issue, a new method, total oxidizable precursor assay (TOP assay), that transforms precursor compounds into detectable PFASs via oxidative treatment have been developed, for assessing the amount of PFAS precursor in a sample1.
Aims and Limitations
The aim of current study was to develop a method for the analysis of PFAA precursors in sewage sludge by oxidative conversion, and the method was initially based on a published method for soil. Parameters tested in current investigation included concentrations of NaOH and persulfate, graphitized non-porous carbon as an additional clean-up method before reaction, reaction time and
concentrations/amounts of sample (table 2). The goal was to obtain a pH of ≥ 12 during the entire reaction time and to qualitative evaluate the changes in levels of PFAA and PFAA precursors.
For every batch of reaction, a negative control sample without the oxidizing reagent was also prepared for contamination check. Positive and negative control samples containing 6:2 FTSA, a PFCA
precursor, was prepared and analyzed.
Solid Phase Extraction – Weak Anion Exchange (SPE-WAX) cartridge was used as a clean-up and separation step for the analysis. Recoveries of the SPE on selected PFASs were also evaluated in the samples. Levels of PFASs in the samples were analyzed with LC-MS/MS.
The target compounds included: PFCAs: C4-C14, C16, C18 (PFBA, PFPeA, PFHxA, PFHpA, PFOA, PFNA, PFDA, PFUnDA, PFDoDA, PFTrDA, PFTDA, PFHxDA and PFOcDA); PFSAs C4-C10, C12 (PFBS, PFPeS, PFHxS, PFHpS, PFOS, PFNS, PFDS and PFDoDS); and FTSAs C4-C8 (4:2 FTSA, 6:2 FTSA, 8:2 FTSA). Table 1 shows the abbreviations mentioned above.
Background
Polyfluoroalkyl and/or perfluoroalkyl substances (PFASs) and oxidative
conversion
Polyfluoroalkyl- and perfluoroalkyl substances (PFASs) are compounds that consist of a hydrophilic functional group attached to a carbon chain where the hydrogen atoms are partially or totally
substituted with fluorine atoms2. This structure gives properties of a surfactant, hydrophobic and
tension lowering, which makes them suitable components in a wide range of applications, such as paint, grease proof materials, coatings and aqueous film forming foams (AFFF) for firefighting3.
PFASs are divided into numerous families of which the most prominent compounds belong to the perfluoroalkyl acids (PFAAs)2,3 (table 1). PFAAs, a family that have earned great attention in research
and media, consist just like the other PFASs, of a fluorinated carbon chain but with an acidic
functional group attached to its terminal endsuch as carboxylic acid (for PFCAs) or sulfonic acid (for PFSAs)
Table 1. List of the targeted PFAS’s abbreviations, names and their number of carbons.
Abbreviation Name Carbon chain length
PFSA Perfluorosulfonic acid
PFBS Perfluorobutane sulfonic acid 4
PFPeS Perfluoropentane sulfonic acid 5
PFHxS Perflurohexane sulfonic acid 6
PFHpS Perfluoroheptane sulfonic acid 7
PFOS Perfluorooctane sulfonic acid 8
PFNS Perfluorononane sulfonic acid 9
PFDS Perfluorodecane sulfonic acid 10
PFDoS Perfluorododecane sulfonic acid 12
PFCA Perfluorocarboxylic acid
PFBA Perfluorobutanoic acid 4
PFPeA Perfluoropentanoic acid 5
PFHxA Perfluorohexanoic acid 6
PFHpA Perfluoroheptanoic acid 7
PFOA Perfluorooctanoic acid 8
PFNA Perfluorononanoic acid 9
PFDA Perfluorodecanoic acid 10
PFUnDA Perfluoroundecanoic acid 11
PFDoDA Perfluorododecanoic acid 12
PFTrDA Perfluorotridecanoic acid 13
PFTDA Perfluorotetradecanoic acid 14
FTSA Fluorotelomer sulfonic acid
4:2 FTSA 4:2 Fluorotelomer sulfonic acid 4
6:2 FTSA 6:2 Fluorotelomer sulfonic acid 6
In the 1950’s, industries began taking advantage of the unique characteristics of these substances in numerous applications, but the effects of PFAS emission to the environment were not studied until the early 2000’s4. That study showed a global distribution of PFAS in animal tissues and potential
accumulation to higher trophic levels. Further research found that high doses of PFAS could be lethal to newborn rodents; that PFASs could act as tumour inducing and immunotoxic agents in the body5;
that they can be transferred to the fetus during pregnancy6 and to infants through lactation7.
These findings, suggesting that PFASs are potentially hazardous to human health, persistent to degradation and bioaccumulative, have raised concern among scientists and general population. After the voluntary phase out of PFOS and PFOA, the European Union started to take action and restricted the use of certain PFAS compounds in 20068. Subsequently, the interests on this topic have grown and
more than 400 scientific articles about it are published every year2.
Total oxidizable precursor assay (TOP assay)
The most studied source of PFAS is direct emission, which involves release of the substances from products containing PFAS, industries and waste water treatment plants. However, PFAS can also enter the environment indirectly by oxidative transformation of other polyfluorinated compounds forming perfluoroalkyl acids (PFAAs)2. These polyfluorinated compounds, such as 6:2 FTSA and FOSA, are
referred to as PFAA precursors2,9 (see details below). One way to analyze them is to mimic the
oxidation process in a procedure called total oxidizable precursor assay (TOP assay), which have worked successfully in various matrices (e.g., water and soil)10. The method uses hydroxyl radicals
formed by thermolysis of persulfate under alkaline conditions to transform PFAA precursors into PFAA of corresponding chain length. In previous experiments with urban runoff water, the PFCA levels increased with a median of 69% after undergoing oxidative treatment1.
Demonstrated in earlier experiments, the reaction rate of the persulfate radicals to hydroxyl radicals transformation is 1 × 107 times faster under alkaline (eq. 2) than acidic (eq. 1) conditions11. Addition
of NaOH in the absence of persulfate during heating, however, have previously not shown to affect the precursor levels1. Therefore, pH is an important factor when conducting TOP assay and maintaining a
pH of 12 or above is required to assure a consistent degree of transformation between batches.
All pH: 𝑆𝑆𝑆𝑆4−· + 𝐻𝐻2𝑆𝑆 → 𝑆𝑆𝑆𝑆42−+ 𝑆𝑆𝐻𝐻· + 𝐻𝐻+ k < 60 M-1 s-1 (1)
Alkaline pH: 𝑆𝑆𝑆𝑆4−· + 𝑆𝑆𝐻𝐻− → 𝑆𝑆𝑆𝑆42− + 𝑆𝑆𝐻𝐻· k = 7 × 107 M-1 s-1 (2) Although the technique is a promising tool when examining PFAS precursors, there are various limitations that currently makes it a qualitative method such as insufficient knowledge about the oxidation process’s efficiency with the presence of organic contaminants in the matrix. Mixed results as a consequence of partial oxidation of some PFASs and a need for evaluation of the method for several PFAS precursor compounds, are some of the issues that the method struggles with. Therefore, the results of this experiment were evaluated mainly qualitatively12.
PFAA precursors
PFAA precursors refer to any compound that could degrade and give rise to PFAAs (PFSAs or PFCAs). For example, compounds that degrade to PFSAs are typically constructed of a PFSA unit connected to another functional group, such as an amide. Commonly used industrial precursors are N-ethylperfluoro-octanesulfonamide (EtFOSA) N-ethylperfluorooctane-sulfonamidoethanol (EtFOSE),
N-methylperfluorooctansulfonamide (MeFOSA) and N-methylperfluorooctanesulfonamidoethanol
(MeFOSE)9.
6:2 fluorotelomer sulfonic acid (6:2 FTSA) is one of the main components of aqueous film forming foam (AFFF) for firefighting13 and can, if environmentally degraded, contribute to emission of
PFCAs. Studies have shown that some microorganisms have the ability to through desulfonation and oxidation transform 6:2 FTSA into PFCAs13,14.
Sewage Sludge
Sewage sludge is a waste product from waste water treatment plants (WWTP). It is composed of organic matter, chemicals and all sorts of things that are flushed down the toilet. Because of the high nutrition value, the agriculture industry is reusing it as a fertilizer for fields and forests15.
However, WWTPs have been considered sources of PFASs release to the environment16. Recent
studies have detected large quantities of PFASs in sewage sludge from Swedish WWTPs; where the majority was of unidentified kind17. As there are some evidence for uptake of PFAS in plants, the use
of sewage sludge as a fertilizer may result in an accumulation of PFASs in foodstuff18. Therefore, it is
of great interest to analyse sewage sludge in the attempt to fully understand the toxicity and
bioaccumulation effects of PFASs. Further, evaluating the levels of PFAA precursors is a vital step in future investigations and risk assessments of PFASs in order to avoid underestimation of the total PFASs that humans and the environment are potentially exposed to.
As TOP assay has not been tested on sewage sludge before, this report initially adopted a method developed for soil and groundwater10, and optimized it from there.
Method
Materials
Potassium peroxodisulfate (K2S2O, 99.0%) and ammonium acetate (C2H7NO2, 99.0%) were purchased
from Sigma Aldrich. Sodium hydroxide pellets (NaOH, laboratory reagent grade), methanol (CH3OH,
HPLC grade), ammonia solution (NH4OH 25%, analytical reagent grade) were purchased from Fisher
Scientific.
All plastic containers and SPE equipment were sonicated with MilliQ-water and detergent, and then with ethanol, followed by a rinse with methanol before use.
Stock testing sample
A stock testing sample was prepared by extracting 1.5 g of domestic sludge (standard reference material (SRM) 2781 from National Institute of Standards and Technology) in 12 ml of 0,2 M NaOH in methanol in a 50 ml PP tube. The solution was homogenized by vortex followed by sonication for 30 min. After sonication, 18 ml of methanol was added, the tube was vortexed again and centrifuged for 10 min at 6000 rpm; the mixture was divided into a supernatant and a pellet. The supernatant was transferred to a new 50 ml tube and 12 ml of methanol was added to the remaining pellet. The centrifugation and transfer of supernatant procedure was repeated once, and the final volume of the supernatant was adjusted to 42 ml using methanol. Thereafter, the stock testing sample was divided into subsamples displayed in table 2; the subsamples in respective PP tubes were evaporated to dryness with nitrogen gas. When the samples were completely dry, 5 ml of MilliQ-water was used for resuspension. Following, NaOH, MilliQ-water and potassium persulfate were added to each sample to obtain the proportions shown in table 2; and the tubes were filled up to 15 ml with MilliQ-water.
Graphitized non-porous carbon treatment
Before the evaporation, some subsamples were subjected to graphitized non-porous carbon (GNPC) clean-up using a 1 ml Supelclean™ ENVI-Carb™ 1ml/100mg (Supelco) cartridge. Initially, the cartridge was rinsed with 1 ml of methanol 3 times, before loading the sample. The entire sample was loaded and collected, another 1 ml of methanol was added to eluate any analyte residues.
Total oxidizable precursor (TOP) assay
The samples were incubated in a water bath at 85°C for either 6h or 10h. When finished, the samples were placed in an ice bath to stop the reaction. The pH of the reaction solution was measured and then adjusted to pH 4 using concentrated hydrochloric acid before clean-up using solid phase extraction-weak anion exchange.
Table 2. The parameters that were tested in the reaction.
[NaOH] Sample [Persulfate] GNPC Reaction time
0.125 M 17.85 mg 60 mM With 6 hours
1.33 M 35.7 mg 600 mM Without 10 hours
53.55 mg 107.1 mg
Solid phase extraction
After reaction all samples were treated with solid phase extraction-weak anion exchange (Oasis SPE-WAX, Waters Corporation). First, the SPE cartridge was conditioned with 4 ml of 0.1% NH4OH in
methanol, 4 ml of methanol and 4 ml of MilliQ-water. Second, the sample was loaded and the cartridge washed with 20 ml of 0.01% NH4OH followed by 30 ml of MilliQ-water to remove any
3000 rpm for 2 min and then dried under vacuum for approximately 30 min. After that, the analytes were eluted in two fractions with 4 ml methanol for fraction 1 (fraction containing neutral and cationic compounds) and 4 ml of 0.1% NH4OH in methanol for fraction 2 (fraction containing anionic
compounds). The sample volume was concentrated to 0.5 ml under a gentle flow of nitrogen gas.
LC-MS/MS analysis
After the concentration step, the samples were transferred to LC vials for instrumental analysis with the following combinations of mobile phase and standards: mobile phase (0,2 M ammonium acetate in MilliQ-water) and recovery standard, in the proportions sample:mobile phase 80:20 for fraction 1 and 40:60 for fraction 2.
The instrument used for analysis was an Acquity UPLC system coupled to a XEVO TQ-S (Waters Co. Milford, USA) triple quadrupole mass spectrometer, operated in negative ionization mode using electrospray ionization. The analytes were separated in a 100 mm C18 BEH column (1.7 µm, 2.1 mm) using a gradient elution of mobile phase A (2mM ammonium acetate in a water and methanol mixture (70:30)), and B (2mM ammonium acetate in methanol).
The levels of PFCA, PFSA and FTSAs were quantified using internal standards that where added prior to analysis. LOD was calculated according to eq. 3 and LOQ to eq. 4. All reported values are above LOQ.
𝐿𝐿𝐿𝐿𝐿𝐿 = 𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑏𝑏𝑎𝑎𝑎𝑎𝑏𝑏+ 3 × 𝑠𝑠𝑠𝑠𝑚𝑚𝑚𝑚𝑠𝑠𝑚𝑚𝑠𝑠𝑠𝑠 𝑠𝑠𝑚𝑚𝑑𝑑𝑑𝑑𝑚𝑚𝑠𝑠𝑑𝑑𝐿𝐿𝑚𝑚𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑏𝑏𝑎𝑎𝑎𝑎𝑏𝑏 (3)
𝐿𝐿𝐿𝐿𝐿𝐿 =𝐿𝐿𝐿𝐿𝐿𝐿3 × 10 (4)
The LOD and LOQ values were calculated from the values of 7 blank replicates for samples treated with graphitized porous carbon, and 2 blank replicates for blanks not treated with graphitized non-porous carbon.
Results
pH after reaction
The first trials tested the lower sodium hydroxide concentration (0.125 M) along with different amount of persulfate (60 mM and 600 mM) and the effect of graphitized non-porous carbon (GNPC) clean-up. All samples resulted in an undesired pH and therefore, the concentration of sodium hydroxide was increased in the next trials.
In combination with the higher NaOH concentration (1.33 M), four different sample amounts were tested with further subparameters. Using a sample amount of 107.1 mg resulted in an undesired pH, in all test, both with high and low concentration of persulfate. However, an improvement from pH 1 to pH 7 could be observed when using GNPC. By reason of crystals observed in the sample when using 600 mM persulfate, the experiment proceeded with 60 mM persulfate.
Three lower sample amounts, with 53.55 mg as the highest, were tested and all three successfully obtained a high pH during a 6 hour reaction time, both with and without GNPC. An additional test using the parameters 1.33 M NaOH, 53.55 mg sample, 60 mM persulfate. Additionally one 53.55 mg GNPC treated sample was tested with a reaction time of 10 hours instead of 6. The prolonged reaction resulted in a decrease in pH from 12 to 7.
Regarding the control samples containing 6:2 FTSA, the results were similar to the SRM-tests: 0.125 M NaOH resulted in a low pH level as well as a 1.33 M NaOH with a reaction time of 10 hours. However, a combination of 1.33 M NaOH, 60 mM persulfate and 6 hours reaction time resulted in pH 12.
Oxidative Transformation
The presented data are obtained from the samples with the parameters 1.33 M NaOH, 60 mM persulfate, 53.55 mg sample and a reaction time of 6 hours.
Figures 2-4. Mean concentration (pg/g SRM) of PFCAs (figure 2), PFSAs (figure 3) and FTSAs (figure 4) in three replicates of SRM samples; P: positive (with persulfate), N: negative (without persulfate), E: with GNPC treatment.
0 100000 200000 300000 400000 500000 600000 700000 800000 900000 1000000 1100000 PE NE P N Conc (pg/g)
PFCAs in SRM
PFOcDA PFHxDA PFTDA PFTrDA PFDoDA PFUnDA PFDA PFNA PFOA PFHpA PFHxA PFPeA PFBA 0 150000 300000 450000 600000 750000 900000 1050000 1200000 1350000 1500000 PE NE P N Conc (pg/g)PFSAs in SRM
PFDoDS PFDS PFNS PFOS PFHpS PFHxS PFPeS PFBS 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 7500 PE NE P N Conc (pg/g)FTSAs in SRM
8:2 FTSA 6:2 FTSA 4:2 FTSA
Figure 5-7. The quantity of substance (nmol) of PFCAs (figure 5), PFSAs (figure 6) and FTSAs (figure 7) in the control samples. PC: positive control (with persulfate), NC: negative control (without persulfate), E: with GNPC treatment.
Relative standard deviation
Table 3. The relative standard deviation of the three replicates of every sample. Compound PE NE P N PFBA 40% <LOQ 27% 73% PFPeA 52% <LOQ 15% 18% PFHxA 28% <LOQ 16% 19% PFHpA 20% 5% 17% 16% PFOA 25% 7% 16% 21% PFNA 24% 4% 20% 14% PFDA 25% 8% 19% 70% PFUnDA 28% <LOQ 18% 19% PFDoDA 23% 8% 33% <LOQ PFTrDA 21% 10% 23% <LOQ
PFTDA 22% <LOQ 39% <LOQ
PFHxDA 37% <LOQ <LOQ <LOQ
PFOcDA 73% <LOQ 29% 56% 0 500 1000 1500 2000 2500 3000 PCE NCE PC NC
n (nmol)
PFCA in control
PFOcDA PFHxDA PFTDA PFTrDA PFDoDA PFUnDA PFDA PFNA PFOA PFHpA PFHxA PFPeA PFBA 0 0,01 0,02 0,03 0,04 0,05 0,06 0,07 0,08 0,09 0,1 0,11 PCE NCE PC NC
n (nmol)
PFSA in control
PFDoS PFDS PFNS PFOS PFHpS PFHxS PFPeS PFBS 0 5 10 15 20 25 30 35 40 45 50 55 60 65 PCE NCE PC NC
n (nmol)
FTSA in control
8:2 FTSA 6:2 FTSA 4:2 FTSA
PFBS 23% 20% 13% 15%
PFPeS <LOQ <LOQ 11% 16%
PFHxS 19% 21% 9% 21%
PFHpS 17% 7% 14% 24%
PFOS 18% 10% 9% 21%
PFNS 18% 9% 12% 18%
PFDS 25% 9% 12% 17%
PFDoDS <LOQ <LOQ <LOQ <LOQ
4:2 FTSA <LOQ <LOQ <LOQ <LOQ
6:2 FTSA <LOQ 13% <LOQ 23%
8:2 FTSA <LOQ 12% 21% 11%
Recovery
The recovery of the SPE was measured in some samples and ranged between 0.4-50.6% in the real samples, 1.4-44.8% in the blanks and 1.4-99.6% in the FTSA controls. Recovery of the 6:2 FTSA in the control samples, however, was about 900%.
Discussion
pH
As discussed in the background section, pH is a vital parameter when conducting TOP assay, and maintaining a pH of 12 during the entire reaction time is desired for even and reproducible results between batches. Initially, the parameters, 0.125 M NaOH, 60 mM persulfate, with GNPC, reaction time of 6 hours. were chosen according to a study on water and soil sample10, and applied on positive
(with persulfate) and negative (without persulfate) triplicates of SRM and a control sample with a standard containing 6:2 FTSA. However, after reaction both the SRM and the control samples resulted in a pH 1. Thereafter, additional tests with various concentrations of NaOH, persulfate and sample were evaluated.
Sample amount: All tests with the initial sample amount (107.1 mg) resulted in a low pH. In the three
trials testing the lower amounts of sample with 1.33 M NaOH and 60 mM persulfate, however, all three resulted in a pH of 12, suggesting that the reaction worked accordingly with ≤ 53.55 mg SRM in 15 ml MilliQ-water.
Reaction time: Parameters that were confirmed in to yield a pH of 12 were applied, but with a reaction
time of 10h instead of 6h. However, the pH level dropped to pH 7 in the SRM sample and 9 in the control sample when the reaction time was increased. These results indicate that the matrix might not necessarily be the cause of a decrease in pH, since the control sample did not contain any matrix. A reason for the pH drop could be that all hydroxyl radicals were consumed during a prolonged reaction time, and that higher concentration of NaOH would be required for such conditions, although more comprehensive studies would be necessary to confirm that. Further, matrix effect, meaning that the oxidant reacted with other chemicals, could not be ruled out based on the current experiment. In conclusion, 1.33 M NaOH resulted in pH 12 when using a sample concentration of ≤ 53.55 mg in 15 ml MilliQ-water (3.57 g/L) with a reaction time of 6 hours. However, when increasing the reaction time to 10h, more NaOH is needed to maintain pH 12 during the entire reaction time.
Potassium persulfate: No difference between a higher or lower concentration of NaOH could be
observed, both resulted in undesired pHs. Besides, crystals of precipitated persulfate were observed in the samples, indicating that the reacting solution was saturated. Therefore, 60 mM persulfate was used instead of 600 mM in next trials. In future experiment, more thorough evaluations of the effect of persulfate concentration on the reaction is desirable to further optimize the reaction.
Graphitized non-porous carbon (GNPC): Regarding the pH, a test was conducted to see whether the
use of GNPC could reduce any matrix effects of the pH. Successfully, the test confirmed that a raise of the pH from 1 to 7 could be seen when using the initial sample amount (107.3 mg) along with 1.33 M NaOH and GNPC. The use of GNPC can reduce the amount of other substances consuming the hydroxyl radicals, yet a NaOH concentration of 1.33 M is still required to maintain the desired reaction.
Graphitized non-porous carbon’s effect on target compounds
SRM samples: The levels of all detected compounds in the negative samples follow a similar trend
with lower levels after GNPC treatment, even for the precursors, suggesting a recovery loss during GNPC clean-up. Despite that, the amount of PFCAs (figure 2) and PFSAs (figure 3) in the SRM samples were more prevalent in the positive GNPC treatment (PE) than in the positive without GNPC treatment (P). Additionally, more precursor compounds were found to be transformed in PE than in P (figure 4). A possible explanation for this could be that the GNPC removes organic matter from the matrix that might also consume the hydroxyl radicals instead of the PFAA precursors.
The PFCAs with the highest levels in both PE and P were the C5-C8, indicating that the precursors dominating the SRM sludge are of 8 or lower carbon chain length.
The PFSAs were found in higher amounts than the PFCAs in the SRM, in both the positive and negative samples. In the samples without GNPC treatment, the levels did not differ significantly between negative and positive experiments.
Regarding the FTSAs, slightly lower levels were found in the GNPC treated samples. This strengthens the suggestion that GNPC treatment causes a loss of analytes. Besides, 8:2 FTSA were also found in the positive sample, suggesting that the total amount of precursors was not fully oxidized.
Control samples: The PFCAs found in the control samples were mainly C4-C6, as expected since the
precursor that were added have a six carbon backbone, and some C7 contaminations (figure 5). In contrast to the SRM samples, the levels of PFCAs were almost twice as high in the sample not treated with GNPC (PC) than in the one treated with GNPC (PCE). If the previous suggested explanation for this phenomenon is accurate, an outcome like this would be logical since what reduced the levels in the positive SRM samples without GNPC (PE) was the matrix effects and not the absence of GNPC. In this case, where there is no matrix, the yield is lower when treated with GNPC as seen in the negative SRM samples (figure 2, 3, 5, 6)
Quantifiable amounts of PFSAs were only found in the negative control sample without GNPC treatment. Accordingly, PFSAs are not the main transformation products of FTSA 13, meaning that
these compounds are likely to be contaminants or transformation products of contaminants.
When comparing the levels of precursors and the transformation products, they should relate to each other, since the molar ration between FTSA and PFCA is 1:1. The levels of PFCAs found the in the samples were about 20 in PCE and 40 in PC times greater than the levels of FTSA in their
corresponding negative samples.
The FTSA found in the negative samples also exceed the amount that was initially added (0.094 nmol). This can be explained by the very high recovery of the 6:2 FTSA found in the negative samples, around 900%. Also, small amounts of impurities of 8:2 FTSA, were found in the 6:2 FTSA standard used in the experiment.
In conclusion, while the pH was improved and the yields in the positive SRM samples better with GNPC treatment, loss of analytes was observed in the negative SRM samples and the control samples. Despite the favourable increased transformation to PFCAs, an analyte loss is not desired in
quantitative investigations, and would have to be accounted for in some way. As it is not possible to add internal standard before reaction, more thorough rinse of the cartridge could be one way to improve the recovery.
Relative standard deviation
The relative standard deviation (RSD) (table 2) corresponds to the amount of the compound found in the sample in some extent. Low levels resulted in a high RSD, and higher in lower RSD, with some exceptions (PFBA and PFPeA in the positive GNPC-treated samples). The negative samples have generally lower RSD than the positive, indicating that there could be an inconsistency in the reaction. For the PFSAs, the opposite trend is observed in P and N, where the RSD is lower in the positive samples than in the negative. Still, the differences between the PFSA levels in P and N and their RSD values are not significantly great.
Conclusion
To maintain a pH of ≥ 12 during a reaction time of 6h, the most suitable combination of parameters are ≤ 53.55 mg in 15 ml (3.57 g/L), 1.33 M NaOH and 60 mM persulfate. When increasing the reaction time, which might be required for a complete oxidation of the total amount of precursors, a higher concentration of NaOH is needed to avoid the pH drop observed when increasing the reaction time. Whether an increased reaction time yields a higher amount of PFCAs could not be validated in the current investigation and further studies are necessary.
Moreover, GNPCgives advantages in terms of reducing matrix components that, by consuming the hydroxyl radicals, could cause a decrease in pH. In other words, the use of GNPC could prevent a pH drop and thereby increase the efficiency of the oxidation, although a loss of analyte was observed. Further, the consistency of the transformation to PFCAs resulted in RSD values of 20-73% with GNPC treatment, and 15-39% without (table 3).
References
1. Houtz, E. F. & Sedlak, D. L. Oxidative Conversion as a Means of Detecting Precursors to Perfluoroalkyl Acids in Urban Runoff. Environ. Sci. Technol. 46, 9342–9349 (2012). 2. Buck, R. C. et al. Perfluoroalkyl and polyfluoroalkyl substances in the environment:
Terminology, classification, and origins. Integr. Environ. Assess. Manag. 7, 513–541 (2011). 3. Hutzinger, O. The handbook of environmental chemistry. Fuel Oxygenates: Handbook of
Environmental Chemsitry2 (1980). doi:10.1007/b_10457
4. Giesy, J. P. & Kannan, K. Global Distribution of Perfluorooctane Sulfonate in Wildlife.
Environ. Sci. Technol. 35, 1339–1342 (2001).
5. Butenhoff, J. L. & Rodricks, J. V. Human Health Risk Assessment of Perfluoroalkyl Acids.
Toxicological Effects of Perfluoroalkyl and Polyfluoroalkyl Substances by DeWitt, Jamie C
(2015). doi:10.1007/978-3-319-15518-0
6. Lau, C. et al. Exposure to perfluorooctane sulfonate during pregnancy in rat and mouse. II: Postnatal evaluation. Toxicol. Sci. 74, 382–392 (2003).
7. Kärrman, A. et al. Exposure of perfluorinated chemicals through lactation: Levels of matched human milk and serum and a temporal trend, 1996-2004, in Sweden. Environ. Health Perspect.
115, 226–230 (2007).
8. Council of the European Union & Parlament. Directive 2006/122/ECOF perfluorooctane sulfonates. Regulation 166–168 (2006).
9. Martin, J. W., Asher, B. J., Beesoon, S., Benskin, J. P. & Ross, M. S. PFOS or PreFOS? Are perfluorooctane sulfonate precursors (PreFOS) important determinants of human and environmental perfluorooctane sulfonate (PFOS) exposure? J. Environ. Monit. 12, 1979 (2010).
10. Houtz, E. F., Higgins, C. P., Field, J. A. & Sedlak, D. L. Persistence of perfluoroalkyl acid precursors in AFFF-impacted groundwater and soil. Environ. Sci. Technol. 47, 8187–8195 (2013).
11. Lee, Y. C., Lo, S. L., Kuo, J. & Lin, Y. L. Persulfate oxidation of perfluorooctanoic acid under the temperatures of 20-40°C. Chem. Eng. J. 198–199, 27–32 (2012).
12. Banzhaf, S., Filipovic, M., Lewis, J., Sparrenbom, C. J. & Barthel, R. A review of contamination of surface-, ground-, and drinking water in Sweden by perfluoroalkyl and polyfluoroalkyl substances (PFASs). Ambio 46, 335–346 (2017).
13. Zhang, S., Lu, X., Wang, N. & Buck, R. C. Biotransformation potential of 6:2 fluorotelomer sulfonate (6:2 FTSA) in aerobic and anaerobic sediment. Chemosphere 154, 224–230 (2016). 14. Wang, N. et al. 6:2 Fluorotelomer sulfonate aerobic biotransformation in activated sludge of
waste water treatment plants. Chemosphere 82, 853–858 (2011).
15. Sundin, A. M. (Naturvårdsverket). Användningsmöjligheter för avloppsslam. http://www.naturvardsverket.se/Stod-i-miljoarbetet Available at:
http://www.naturvardsverket.se/Stod-i-miljoarbetet/Vagledningar/Avlopp/Avloppsslam/Anvandningsmojligheter-for-avloppsslam/. (Accessed: 6th October 2015)
16. Weinberg, I., Dreyer, A. & Ebinghaus, R. Waste water treatment plants as sources of
air. Environ. Pollut. 159, 935–941 (2011).
17. Kärrman, A., Yeung, L., Eriksson, U. & Swedish National Environmental Protection Agency.
A pilot study on unidentified poly- and perfluoroalkyl substances (PFASs) in sewage in Sweden.
18. Stahl, T. et al. Carryover of perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS) from soil to plants. Arch. Environ. Contam. Toxicol. 57, 289–298 (2009).
