Emerging Chlorinated Polyfluorinated Polyether Compounds Impacting the Waters of Southwestern New Jersey Identified by Use of Nontargeted Analysis

Abstract

Per- and polyfluoroalkyl substances (PFAS) are a widespread, environmentally persistent class of anthropogenic chemicals that are widely used in industrial and consumer products and frequently detected in environmental media. Potential human health impacts from long-term exposure to legacy PFAS resulted in the industrial development and use of numerous replacement species in recent decades. Environmental investigative activities have been crucial in identifying the existence and environmental transport of emerging PFAS in environmental media. Previous investigations in an industrially impacted region of southwestern New Jersey has shown consistently elevated levels of legacy PFAS, motivating additional examination by non-targeted mass spectrometry to identify emerging PFAS contamination. This study applied non-targeted analysis to water samples collected in Gloucester and Salem Counties in southwestern New Jersey, revealing the existence of a series of novel chloro-perfluoro-polyether carboxylates and related PFAS species originating from an industrial PFAS user in the region. There is sparse publicly available toxicity information for the emerging chemical species, but estimated concentrations exceeded the state drinking water standards for perfluorooctanoic acid (PFOA) and perfluorononanoic acid (PFNA). Non-targeted analysis was used to estimate the effectiveness of point-of-entry water treatment systems for removal of the emerging species and reduced the abundance of PFAS by >90%.

Graphical Abstract

Introduction:

Per- and polyfluoroalkyl substances (PFAS) are a widely distributed, persistent class of anthropogenic chemicals defined by the inclusion of an aliphatic chain of two or more carbon atoms with fluorine-carbon bonds.¹ PFAS are chemically diverse, with hundreds of products in active commercial usage and thousands more known to be extant.^{2, 3} The widespread applications of PFAS include use as coatings, surfactants, and fire-suppressants, and as processing aids in the production of fluorpolymers^{4, 5}. These uses have resulted in the detection of legacy PFAS in a variety of globally distributed environmental and biological matrices, indicating worldwide exposure.^6–9 Legacy long-chain perfluoroalkyl acid (carboxylates and sulfonates) PFAS were identified as being highly persistent and potentially toxic,^9–12 and two of the most widely used compounds, perfluorooctane sulfonic acid (PFOS) and perfluorooctanoic acid (PFOA) were phased out by eight major manufacturers in the United States beginning in the early 2000s.^12–14 In response to the 2015 total phase-out target for PFOA and higher chain-length homologs, PFAS manufacturers began the development of replacement chemistries, which have likewise proven to be widely detectable in the environment.^{1, 9, 15} Non-targeted approaches for investigation and compound elucidation based on high-resolution mass spectrometry have been developed since traditional target-list based approaches for compound detection and quantitation are insufficient for detection of the ever-increasing list of emerging compounds. The application of non-targeted analysis to environmental samples has resulted in the detection of many novel PFAS structures in recent years.^{9, 16–21}

The New Jersey Department of Environmental Protection (NJDEP) leads an ongoing effort to investigate the presence of PFAS in an area in the southwest portion of the state that is located near industrial facilities with current or former PFAS use and/or manufacturing. NJDEP classifies PFAS compounds as contaminants of emerging concern²² and has promulgated Groundwater Quality Standards and Maximum Contaminant Levels (MCLs) in drinking water for PFOS (13 ppt) and PFOA (14 ppt) that are among the most stringent in the nation.^23–25 NJDEP also promulgated Groundwater Quality Standards and MCLs for perfluorononanoic acid (PFNA), at 13 ppt, motivated by concentrations of PFNA in public supply wells that were the among the highest reported in the world.^{26, 27} To determine the presence and distribution of these legacy PFAS chemicals in the environment, NJDEP collected soil samples and sediment²⁸ in addition to surface and groundwater in the northwestern portions of Gloucester and Salem counties. As part of a collaborative study with NJDEP, the US EPA Office of Research and Development conducted non-targeted analysis of the samples, to determine the presence of unanticipated emerging replacement PFAS compounds in this region. Further, NJ has recommended the use of granular activated carbon (GAC) or ion exchange resin filtration systems for removal of PFAS from private wells²⁹ and sought to measure the effectiveness of this treatment technology on the removal of emerging PFAS.

This work evaluates the results of nontargeted screening and novel structural elucidation of the water samples collected by NJDEP and the Delaware River Basin Commission, as well as the follow-up investigation conducted by NJDEP to determine the effectiveness of point-of-entry treatment (POET) systems installed at select private wells in the area.

Materials and Methods:

Sample collection

As part of the collaborative PFAS study in southwestern New Jersey, a water sampling campaign was undertaken in October and November, 2017. Surface water samples (n=32) were collected from non-potable sources (e.g. tidal creeks, ponds, etc. see Table S1, Figure S1) in Gloucester and Salem Counties. Groundwater samples (n=20) (Figure S2) were selected from well locations that had been previously analyzed by a commercial laboratory as part of Solvay or NJDEP studies and had reported concentrations of PFNA and/or PFOA above NJ water standards. Follow-up groundwater sampling was conducted in September of 2019 at a subset of locations with POET systems to determine if those systems were effective for emerging PFAS that were initially identified in the 2017 sampling (Figure S2). In each sampling event, water (^~1 L) was collected in high-density polypropylene (HDPE) vessels using previously described methods,³⁰ and stabilized with the addition of nitric acid for shipment to the US EPA facility in Research Triangle Park, North Carolina for non-targeted analysis.

Sample Processing

PFAS in the collected water was concentrated using weak-anion exchange (WAX) solid-phase extraction (SPE) according to our previously developed protocol for non-targeted PFAS analysis.³⁰ Briefly, the water was removed and the sampling containers were washed with methanol to reduce adsorption of PFAS. Following filtration with a Whatman GF/A glass microfiber filter, the combined water and rinsate was concentrated on a SPC-10 Sep-Pak concentrator with an Oasis WAX Plus SPE cartridge. Samples were eluted in basic methanol (0.1% NH₄OH in MeOH) and further concentrated by evaporation under nitrogen stream. The total concentration factor was ~ 1000x (i.e. 500 mL of water to 0.5 mL of extract). For samples collected from potable wells with POET systems, ¹³C₅ -perfluorononanoic acid (PFNA) was spiked in prior to filtering to act as a reference for semi-quantitative concentration estimates.

Sample Analysis

Water concentrates were diluted 1:3 with 2.5 mM ammonium acetate solution and analyzed by LC-MS. Initial screening of water samples was carried out in negative electrospray-ionization mode on a high resolution time-of-flight mass spectrometer as previously described.¹⁸ In-depth analysis was conducted on Thermo Orbitrap Fusion mass spectrometer (Thermo Fisher Scientific, Waltham, MA) with a heated electrospray ionization source operated in negative mode. Samples were measured using data dependent MS/MS acquisition with a scan range of 150–1500 m/z and Orbitrap resolution of 60,000 and 30,000 for MS1 and MS2 acquisition respectively. A complete detailing of instrumental settings can be found in the supplemental information.

Data Processing

For non-targeted data analysis, raw instrument files were processed using Thermo Compound Discoverer 2.1 to extract chemical features and tentatively match features against USEPA’s Distributed Structure-Searchable Toxicity (DSSTox) database. A more exhaustive description of the data processing settings can be found in the supplemental information. Chemical features were manually examined to confirm the feasibility of database matches, and chemicals with no match were manually assigned tentative structure/class annotations when the MS/MS contained fragments indicative of fluorination.

Results and Discussion:

Identification of a series of novel PFAS compounds

Examining the most abundant chemical features identified by the non-targeted screening revealed the anticipated legacy compounds (PFOA, PFNA, and other perfluoroalkyl acids Table S2), but also an abundant mass with a chlorinated isotope signature and a mass defect consistent with heavy halogenation.³¹ The MS1 scan revealed a series of related, coeluting ions showing similar spectral patterns (Figure S3). Because the ions coeluted, they were assumed to be in-source artifacts related to a single source compound. MS/MS scans for each precursor ion showed similar fragmentation patterns, supporting this assignment (Figure S4). The three were ultimately assigned as the [M-H]-, [2M-H]-, and [M-H-CO₂-CF₂]- ions of C₈HClF₁₄O₄. The in-source [M-H-CO₂-CF₂]- fragment is the major observed ion for these species and was used as the representative ion when determining species abundance. The fragility of the precursor [M-H]- ion prevented it from being observed under all MS source conditions.

The assigned molecular formula was consistent with a chloro-perfluoro-polyether carboxylate (ClPFPECA) compound registered to Solvay Specialty Polymers,^{15, 28, 32, 33} which is an active fluoropolymer user and manufacturer in this region. MS/MS fragmentation was used to elucidate a structural ID based on major fragments (Figure 1), which was consistent with a known product associated with fluoropolymer production.¹⁵ Structurally, the isomerization within the fluoroether propyl unit(s) and the position of the terminal chlorine on the ultimate or penultimate carbon are difficult to determine by MS/MS alone, and the fluorochemical literature suggests that a mixture of isomers is likely.

Figure 1.

Annotated MS/MS fragmentation spectrum of the ion at m/z 460.9268 and corresponding tentative structures. The position of the Cl on the sixth or seventh carbon is ambiguous based on MS/MS alone.

Chromatographically, a homologous series, each of member of which with multiple isomer peaks, was identified (Figure S5), but for simplicity we use the general structure from Wang¹⁵ and reference the compounds as PFPECA(e,p) based on molecular formula of the oligomer units (Figure 2). Ultimately, we were also able to identify a hydro- perfluoroether carboxylic acid (HPFPECAs) series with a similar structure (Table S3, Figure 2), which is similar to the breakdown products of perfluoroalkyl ether precursors observed in other industrially impacted waters.^{17, 18, 34} Notably, the manufacture of fluoropolymers does not occur at the NJ location, and therefore the polyfluoroalkyl HPFPECAs are likely an impurity retained as an industrial intermediate of manufacturing, where dehalogenation is a common step or a byproduct in fluoroether synthesis.³⁵

Figure 2.

General structure of the perfluoroether carboxylic acid oligomeric series. Both a chloro- (ClPFPECA) and hydro- (HPFECA) form were identified with a similar backbone. The R group position is likely variable.^{35, 36}

Five major oligomer forms of the ClPFPECA(e,p) were identified, with e = 0–2, and p = 0–2 (Table S3). Longer chain oligomeric forms (MW > ~628) were identified in soil collected contemporaneously²⁸ but their abundance was not significantly different from blanks; lack of detection was likely due to limited water solubility.

Distribution of novel PFAS compounds

The sum of all ClPFPECA oligomers was substantially higher in relative abundance in tidally influenced surface water (mostly the Delaware River and tributary mouths) than in non-tidal surface water (Figure 3). This is likely driven by wastewater discharge to the Delaware River in the vicinity of Solvay as the most likely contributor of the emerging PFAS to water concentrations (Figure S6). Detection in groundwater well samples indicates that groundwater contamination has occurred, but it is unclear whether the origin is surface water-to-groundwater exchange or some other mechanism. The consistent detection in groundwater of ClPFPECA(0,1) suggests a surface-to-groundwater migration process, given the lack of true groundwater plume from direct groundwater release (Figure S7) and should be the subject of future work.

Figure 3.

Total peak abundance for the sum of detected ClPFPECA oligomer forms in ground (GW), non-tidally influenced surface (NSW), and tidally influenced surface (TSW) water, normalized to the highest abundance detected across all samples.

When examining the ratio between oligomer forms, the ClPFPECA(0,1) form dominates in every water sample collected (mean = 94.4% of total ClPFPECA abundance), and is the only constituent detected in groundwater. The ratio of oligomer forms is markedly different between types of surface water samples as well, with heavy bias towards the (0,1) oligomer (Figure S8). In non-tidal water ClPFPECA(0,1) makes up 88% of the total ClPFPECA abundance, while in tidal waters the fraction is 98%. The fraction observed in soil averaged only ~40%²⁸ and suggests that the non-tidal waters may be influenced by aerial deposition, as soil is, rather than direct discharge.

Efficacy of POET system removal of novel PFAS compounds

POET systems had been previously installed on private potable wells with levels of PFNA exceeding the state MCL, either prior to or following the 2017 sampling event. Follow-up nontargeted analysis conducted in 2019 on the influent water, midpoint of the two-stage treatment system, and the effluent from the POET system was used to assess the effectiveness of the treatment system at reducing levels of ClPFECA. The absolute concentration of these chemicals cannot be determined without reference standards, which are unavailable due to the lack of commercial vendor and the proprietary nature of the chemicals; although, some efforts to estimate concentrations by using structurally similar surrogate compounds have been made.^{28, 37} An internal standard quantity of PFNA was used as a surrogate for estimation in these samples. Assuming the MS response of PFNA and the ClPFPECA is similar the concentration of ClPFPECA can be estimated as ClPFPECA_concentration = ClPFPECA_abundance / ¹³C₅-PFNA_abundance * ¹³C₅-PFNA_concentration. The highest estimated concentration for the PFECAs in the POET influent water ~300ng/L, far in excess of NJ’s limits for similarly sized PFOA, PFOS, and PFNA. Comparisons of concentrations demonstrated a 99% reduction for the major ClPFPECA oligomer for all the POET systems analyzed (Figure 4).

Figure 4.

Estimated ClPFPECA(0,1) concentrations from samples collected at the influent (INF), between the stages (MID) and effluent (EFF) of two-stage POET systems installed at several private wells in New Jersey. Concentration estimate is based on a simple abundance ratio with ¹³C₅ - perfluorononanoic acid (PFNA) internal standard.

The relative abundance ratios for the influent and effluent samples were used to estimate removal efficiencies for the ClPFECA and legacy PFAS compounds. The average reduction in peak-area abundance across POET systems for identified PFAS ions was 95±8%. The sorption efficiency for PFAS is variable dependent on the underlying PFAS chain length (and polarity)^{38, 39} but since the emerging PFAS compounds are structurally similar to the PFAS they were intended to replace, it is unsurprising that they respond similarly to treatment (i.e. effective removal >90%). Perhaps less desirably, the affinity for removal systems also corresponds to potential for bioaccumulation – the log estimated bioconcentration factor for ClPFPECA(0,1) is similar to that for PFOA and PFNA (ClPFPECA(0,1): 3.85, PFOA: 3.57, PFNA: 3.96) based on EPA’s Episuite prediction methodology. At this stage there is no information about the long-term performance of the POET systems for ClPFPECA species, so considerations such as breakthrough will need to be monitored.

Implications

This study demonstrates the utility of NTA applied to environmental samples collected near industrially impacted sites and provides a paradigm for research-driven support of local governmental goals. The application of nontargeted analysis in conjunction with legacy chemical monitoring enabled the identification of the emerging chlorinated perfluorinated ether acids, as well as their polyfluorinated versions, at abundance levels similar to those of regulated legacy PFAS (PFOA, PFNA, and PFOS). The emerging ClPFPECA compounds do not have available standards for quantitation, and the only publicly available safety evaluation is based on negative results in genotoxicity assays.³² PFAS are generally not genotoxic, so much additional work is necessary to understand the potential ecological and human health effects. For example, fluorinated polyether compounds similar in size to those detected in this study were detected at parts per billion levels in the blood serum of individuals with drinking water exposure,⁴² which, when combined with the physiochemical properties, suggests ClPFPECA compounds detected in this study have the potential for bioaccumulation.

The usage of GAC and ion exchange resin-based POET systems is effective at removing novel PFAS as well as the legacy PFAS for which the POET was installed, but the long-term breakthrough potential of this treatment method and environmental accumulation potential of the ClPFPECA compounds require further investigation.