PER AND POLYFLUOROALKYL SUBSTANCES IN WATER TREATMENT RESIDUALS: OCCURRENCE AND DESORPTION

Abstract

Per- and polyfluoroalkyl substances (PFAS) in surface and ground waters supplying municipal drinking water are a growing concern. However, PFAS concentrations in water treatment residuals (WTRs) - a solid byproduct of water treatment - have yet to be explored. In a first of its kind assessment, we examine PFAS occurrence in seven Ca-, Fe-, and Al-based drinking water treatment residuals (DWTRs) and one wastewater effluent treatment residual (WWETR) produced using aluminum chlorohydrate (ACH). Only perfluoroalkyl acids (PFAAs) were detected, with total PFAA concentrations in the seven DWTRs produced from naturally recharged water sources ranging from 0 to ~3.3 μg kg⁻¹; no PFAS were detected in either of the Ca-DWTRs. The ACH-WWETR contained the highest number and concentration of PFAAs (34 μg kg⁻¹). Desorption of resident PFAAs from the WTRs was negligible for the carboxylates (PFCAs). Some desorption of the sulfonates (PFSAs) was detected, particularly for PFOS which had the highest concentration among all resident PFAAs. The ACH-WWETR was further evaluated for its potential to attenuate additional PFAAs (3500 μg mL⁻¹ total PFAAs) in a biosolid-derived porewater matrix. Sorption was highest for long chain PFAAs and subsequent desorption of the adsorbed PFAAs ranged from 0% to no more than 26%, with the WWETR mass added strongly affecting both PFSA and PFCA sorption/desorption. These findings suggest that WTRs, if introduced into the environment, are unlikely to be a major source of PFAS. Also, the use of particular WTRs as amendments may provide a beneficial reduction in PFAS mobility.

INTRODUCTION

Per and polyfluoroalkyl substances (PFAS) are a family of exclusively anthropogenic compounds characterized by their polar and hydrophilic “heads” and hydrophobic and lipohobic fluoroalkyl ‘tails’. PFAS’ highly customizable and unique chemical characteristics has led to multiple uses including stain and water-resistant textiles, food packaging, fire-retardant and fire-extinguishing products, pesticides, paints, personal care products, and surfactants (Lau, Anitole et al. 2007) with more than 9000 PFAS compounds in existence as of 2021 (USEPA 2021). High usage, combined with PFAS’ resistance to degradation has resulted in PFAS being ubiquitous in the environment. PFAS are bioaccumulative and have been linked to adverse effects on wildlife (Kelly, Ikonomou et al. 2009, Jeon, Kannan et al. 2011) and humans (Zhang, Sun et al. 2011). PFAS exposure to humans, particularly via drinking water, is a concern as reflected by USEPA proposed legally binding thresholds known as Maximum Contaminant Levels (MCLs) for drinking water as 4 ng L⁻¹ for perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS) and hazard index (unitless) of 1.0 for combined perfluorononanoic acid (PFNA), perfluorohexanesulfonic acid (PFHxS), perfluorobutanesulfonic acid (PFBS) and hexafluoropropylene oxide-dimer acid (HFPO-DA or GenX) (USEPA 2016, USEPA 2023) and USEPA’s recent release of short-term PFAS regulation goals (USEPA 2021). States within the USA have established their own lower advisories for PFOA and PFOS (USEPA 2016, USEPA 2016). As of 2009, PFAS production is prohibited in the EU (UNEP 2009) while in the USA, only long-chain PFAS are largely phased out of industrial use, with provisions made to limit introduction of new PFAS without USEPA oversight (USEPA 2016, USEPA 2021).

Most water and wastewater treatment schemes have low to no effectiveness in PFAS removal with the exception of high-pressure membrane systems (e.g., nanofiltration and reverse osmosis) sorbent technologies (e.g., activated carbon) (Crone, Speth et al. 2019) and foam fractionation (Leung, Shukla et al. 2022). Recent work has shown that drinking water treatment residuals (DWTRs) produced from alum, a common coagulant aid utilized in water treatment, can irreversibly sorb PFOA and PFOS (Zhang, Sarkar et al. 2021). DWTRs and wastewater effluent treatment residuals (WWETRs) result from addition of chemicals to promote to remove impurities via co-precipitation, absorption, flocculation and settling (Turner, Wheeler et al. 2019). Most water treatment residuals (WTRs) are dominated by one of three major elements: aluminum (Al), iron (Fe) or calcium (Ca) as determined by the primary coagulant used. The most commonly utilized coagulants are alum and ferric sulfate compounds (Al₂(SO₄)₃*14(H₂O), FeCl₃ or Fe₂(SO₄)₃ * 9H₂O) and quick lime (CaO) or slaked lime (Ca(OH)₂) (Dassanayake, Jayasinghe et al. 2015, Turner, Wheeler et al. 2019). In addition to standard alum, preformed and pre-charged aluminum hydroxide polymers (e.g., polyaluminum chloride) and aluminum polymer salts (e.g., aluminum chlorohydrate, (ACH)) are available for commercial purchase. These aluminum polymers are designed to be more stable over a range of water pH and temperature, as well as outperforming alum in the removal efficiency of dissolved organics on a mass per mass basis (Verma, Bhunia et al. 2012, Zaman, Rohani et al. 2021). Due to the variability in drinking water treatment plant (DWTP) processes, water quality, polymer/coagulant utilization, and major coagulant selection, removal efficiency can vary greatly (Hendricks 2011). As a result, the relative mass yield of WTR can vary from scenario to scenario, making estimations of contaminant loading difficult without water sampling data.

Within the US, more than 2 million tons of DWTRs are produced daily as of 2003 (Prakash and Sengupta 2003) with quantities only expected to rise with increasing population densities. Disposal methodologies vary widely. A 2011 USEPA survey of 2,151 DWTPs estimated that ~37% of DWTPs landfilled their DWTRs, ~37% recycled (the method is not specified in the survey but likely through coagulant recovery), and ~19% land applied (USEPA 2011) with the rest including various unspecified methods. What little research exists for the beneficial reuse of land-applied WTR focuses primarily on use as a substitute for agricultural lime (Dassanayake, Jayasinghe et al. 2015) and phosphorus stabilization (often in conjunction with organic fertilizer applications, e.g., biosolids) (Judy, Silveira et al. 2019).

Currently, no characterization data exist for inherent PFAS concentrations in DWTRs. In this study, we determine the concentration of a suite of native PFAS in eight WTRs including those derived from Al, Fe and Ca-based materials. Key characteristics of the WTRs expected to influence PFAS partitioning were also measured including organic matter (OM) content, total and amorphous fractions of Al and Fe (amorphous Al and Fe, as determined by oxalate extractions, is considered to be a relatively reactive pool as a result of their typically small size and corresponding large surface area) (Sparks 1996), pH, and electrical conductivity (EC). Characterization of WTR Al, Fe and OM content was conducted as these characteristics have been shown to be correlated with PFAS partitioning in waste residuals (Gravesen, Lee et al. 2023), presumably as a result of the presence of sorption sites associated with PFAS retention, whereas pH and EC were selected to provide valuable context for sorption behavior. Furthermore, the degree to which resident PFAS desorb from WTRs was assessed, as was the degree to which amending biosolids-porewater with WTRs affects PFAS desorption, with the latter assessment simulating a beneficial reuse scenario where WTRs have been co-applied with biosolids with the intent to reduce PFAS mobility.

MATERIALS AND METHODS

Chemicals and Reagents.

PFAS standards were purchased from Wellington Laboratories (Guelph, Canada) including isotopically mass-labeled compounds (MFTA-MXA and MPFAC-24ES) and native standards (NS) (PFAC30PAR and FTA-MXA; Table S1). Methanol (≥ 99.9%, HPLC grade), glacial acetic acid (≥ 99.7%, HPLC grade), ammonium acetate (CH₃COONH₄, 100%), acetone (≥ 99.5%, HPLC grade), 2-propanol (≥ 99.9%, Optima grade), ammonium hydroxide (≥ 99.9%, Optima grade), Supelclean^™ ENVI-Carb SPE bulk packaging were purchased from Sigma Aldrich Inc (Burlington, MA). Nanopure water was produced using a Synergy^®-R EMD-Millipure water purification system using Synergy Pak^® 3 DI cartridges.

WTRs.

A total of eight WTRs were examined in this study, including four Al DWTRs, two Ca DWTRs, one Fe DWTR, and one Al WWETR, which was derived from addition of ACH to clean wastewater prior to reinjecting into the aquifer. DWTRs resulted from cleanup of either surface water or groundwater as summarized in Table 1 along with the coagulant used. The ACH was freeze-dried upon receipt whereas all other residuals were spread and air-dried on high-density polyethylene (HDPE) plastic sheets upon receipt.

Table 1.

General information on water treatment residuals (WTRs) evaluated.

WTR	Type	Water Source	Coagulant utilized	Year of production	Approximate population served
Ca-1	Ca	Ground	Calcium Oxide	2018	200,000
Ca-2	Ca	Ground	Hydrated Lime	2015	40,000
Fe	Fe	Surface	Ferric Sulfate	2020	700,000
Al-1	Al	Surface	Likely Aluminum Sulfate1	2009	40,000
Al-2	Al	Surface	Likely Aluminum Sulfate1	2005	20,000
Al-3	Al	Surface	Aluminum Sulfate	2007	320,000
Al-4	Al	Surface	Aluminum Sulfate	2021	550,000
ACH WWETR	Al	WWTP2 effluent	Aluminum Chlorohydrate	2021	NA

¹Where coagulant utilized by the DWTP could not be confirmed, the most likely coagulant has been listed based on regional production practices and WTR chemical characterization results (Table 2);

²Wastewater treatment plant

WTR Chemical Characterization.

All WTRs were analyzed for PFAS, total Al, Fe, and Ca contents, non and poorly-crystalline Fe and Al mineral contents, OM content, and pH and EC. Total Fe and Al were determined by microwave assisted acid digestion in reverse aqua regia and analyzed on ICP-OES (Spectro Arcos, Kleve, Germany) (USEPA 2007). Oxalate extractable Fe (Fe_ox) and Al (Al_ox) were determined by acid ammonium oxalate solution and analyzed on ICP-OES (Perkin Elmer, NexION 300X) (Loeppert and Inskeep 1996). Percent OM was determined by loss on ignition (LOI) at 450°C using a modification of a previously published method(Nelson and Sommers 1996). pH and EC were determined by a 1:2 solid to solution ratio in nano-pure water, a background matrix of salt was omitted considering the high native salt content of the WTRs.

WTR Extraction for PFAS.

PFAS extractions followed a modified previously published methodology used for solids (Choi, Lazcano et al. 2019). WTRs were ground and sieved to ≤ 2 mm prior to weighing 0.5 g samples in triplicate for each WTRs into 15-mL polypropylene (PP) centrifuge tubes. Isotopically labeled surrogates (2-5 ng) were added to each sample prior to being extracted three times sequentially using 7-mL 99:1 v/v methanol/30% ammonium hydroxide solution and1-h sonication in a heated bath at 30 °C. Samples were then centrifuged at 2700 x g for 20 min after each extraction and supernatants pooled into a single 50-mL PP tube. The pooled extract was concentrated under a gentle nitrogen (N) gas stream at 35°C using an Organomation N-EVAP 12 position N evaporator (Catalog no. 11155-NT) to 0.5-1 mL volume visually and final volume determined gravimetrically. Concentrate was then transferred to microcentrifuge tubes containing 20-40 mg ENVI-Carb and 20 μL glacial acetic acid for cleanup and then vortexed for 30 sec followed by centrifugation at 13,000 x g for 30 min. An aliquot of the cleaned supernatant was transferred to a 2-mL glass injection vial with a final solvent composition of 400/400/268 (v/v/v/) methanol/isopropanol/0.003% ammonium hydroxide in nano pure water. Final samples were sonicated for 5 min then stored at 4° C until analysis, at which time they were sonicated again for 5 min prior to injection.

Desorption of Resident PFAS.

Desorption of the PFAS residing on the WTRs was conducted on the three WTRs that had the highest PFAS concentrations which included the ACH WWETR, Fe DWTR and one of the Al DWTRs. 0.5 g of each WTR was equilibrated with 250-mL of deionized water in nominal 250-mL PP bottles for 96 hours (Zhang, Sarkar et al. 2021) at 40 rpm on a Glas-Col laboratory rotator (Terre Haute, IN USA) followed by centrifugation (Jouan centrifuge, 1711 g for 20 minutes). The supernatant was removed for PFAS analysis. Triplicate process blanks were included consisting of media-less deionized water. The pH of the resulting supernatant ranged between 5.5-6.

PFAS Adsorption/Desorption.

An initial evaluation of the attenuation potential of the ACH -WETR was done using biosolid-derived porewater amended with additional PFAAs. This porewater was selected instead of deionized water to represent what may be present in a wastewater or what may potentially leach from a land-applied biosolid. The biosolids utilized to produce the porewater were a 2021 Class A biosolids produced from a Cambi^® Thermal Hydrolysis process combined with anaerobic digestion. Generated porewater was produced by a 24-h static equilibration with deionized water at a mass to volume ratio of 1:5 in a 50-ml PP conical tube. The supernatant after centrifugation (1711 g for 30 minutes) was diluted by 30% and amended with seven perfluoroalkyl acids (PFAAs) including C6, C7, C8 and C9 perfluoroalkyl carboxylic acids (PFCAs) and C4, C6 and C8 perfluoroalkyl sulfonates (PFSAs) (Table S2). The final total PFAA concentration was approximately 3500 ng L⁻¹ with individual PFAA concentrations ranging from 226 to 777 ng L⁻¹ (Table S2). The dominant cation in the final biosolid-derived porewater was Ca²⁺ at 150 μg L⁻¹ with all other cations not less than one order of magnitude in concentration. The dissolved organic carbon (DOC) concentration in the diluted biosolids-derived porewater and supernatant from the sorption and desorption steps were measured using a Shimadzu TOC-L CPH/CPN analyzer.

PFAA sorption was assessed by equilibrating ACH-WWETR at varying masses (0.005, 0.02, 0.05. 0.25 and 0.5 g) in triplicate with 9-mL of PFAA-amended biosolids-derived porewater for 48 h in 15-mL PP tubes rotated end-over-end at 40 rpm. Experimental controls included media-less porewater. After equilibration, samples were centrifuged (1711 g for 20 minutes), supernatant was collected for PFAS analysis. Nine mL of deionized water was added to the post sorption residuals, and equilibrated for 96 h, after desorption, samples were centrifuged, and supernatant collected for PFAS analysis (Fig. S1). Supernatant pH values for both sorption and desorption were in the 5 - 5.5 range. The percent PFAAs adsorbed was calculated by the difference in the PFAA mass initially present (resident PFAAs in the ACH-WWETR and in the amended porewater) and the mass found in the supernatant after equilibration with the ACH-WWETR.

Solid-phase Extraction (SPE).

Supernatants for all sorption and desorption steps were processed with SPE using HLB (30 μg, 6 mL, 200 mg, Waters, Oasis HLB a hydrophilic-lipophilic-balanced) cartridges. A thermos 24-port manifold was used for small sample volumes (9 mL) while an EZPFC semi-automated SPE (FMC. INC) was used for the large sample volumes (250 mL). The SPE cartridges were pre-conditioned using 5 mL methanol and 5 mL deionized water at flow rate of about 2 mL min⁻¹. The samples were then passed through the cartridge at about 2 mL min⁻¹ using a vacuum manifold connected to a vacuum pump. After all the samples have been passed through the cartridges, the SPE cartridges were washed with 5-mL 0.4% formic acid 5/95 v/v methanol/water solution and dried under vacuum at ~20 mmHg for 2 h. Cartridges were eluted with 5 mL methanol under gravity. The final extracts were blown down under gentle N until almost dry and then reconstituted with 250-μL methanol and 250-μL water prior to PFAS analysis.

PFAS Analysis.

Samples were stored and shipped at 4° C to Purdue University for analysis. PFAS analysis was performed on a Shimadzu ultrahigh pressure liquid chromatograph (uPLC) coupled to a Sciex quadrupole time of flight (QToF; Sciex TripleTOF^® 5600 System) operated in negative ion electrospray ionization (ESI) for the analysis of resident PFAS on the WTRs. QToF data acquisition was done in SWATH mode and data processed with Sciex OS software (ver 2.1.1). For the sorption/desorption study with added PFAAs, a uPLC coupled to a triple Shimadzu 8040 quadrupole mass spectrometer was used for which data was acquired using multiple reaction monitoring (MRM) mode and processed using Labsolution software for quantitative analysis. Additional chromatographic information is in the supplemental information (SI) along with information on MSMS transition, internal standard and surrogates used for quantitation, and limit of quantitation (LOQ) are provided in Table S1. The percent recoveries for PFAAs from matrix spikes in the SPE process for the supernatants and for PFAS in the extraction of the WTRs are provided in Table S3 and Table S4, respectively.

RESULTS AND DISCUSSION

Native PFAS Concentrations in WTRs.

Of the 52 PFAS that were targeted, only PFAAs were detected, with the most commonly detected PFAAs being L- (linear) and Br- (branched) PFOS, which were detected in 6 of the 8 WTRs. PFHxA, PFOA, and L-PFHxS, which were each detected in two WTRs, were the next most commonly detected PFAAs (Fig. 1 and Table 3). Several PFAAs were below the limits of detection (<LOD). Total PFAA concentrations ranged from < LOD (DWTRs Ca-1 and Ca-5) to ~34 μg kg⁻¹ for the ACH WWETR (Fig. 1).

Figure 1.

Resident PFAA composition of the analyzed WTRs. Individual PFAA concentrations are averages of three replicates with bars representing standard errors. ‘L-’ and ‘Br-‘ in the legend refer to linear and branched PFSAs, respectively.

Table 3.

Desorption of resident PFAAs from two DWTRs and one WWETR.

	ACH WWETR		Fe DWTR		Al-1 DWTR
PFAS	Resident concentration (μg kg−1)	Desorption (%)	Resident concentration (μg kg−1)	Desorption (%)	Resident concentration (μg kg−1)	Desorption (%)
PFBS	0.47 ± 0.12	27%	< LODc	NAC	< LODc	NAd
PFHxS	3.52± 0.45	0%	< LODc	NAC	0.18 ± 0.03	15%
PFOS	16.05 ± 0.26	2%	1.35 ± 0.34	49%	2.45 ± 0.39	38%
PFPeA	3.72 ± 0.96	0%	< LODc	NAd	< LODc	NAd
PFHxA	2.13 ± 0.07	0%	0.50 ± 0.28	0%	< LODc	NAd
PFHpA	3.01 ± 0.32	0%	< LODc	NAd	< LODc	NAd
PFOA	3.78 ± 0.12	<LOQb	< LODc	NAd	0.62 ± 0.06	<LOQ
PFNA	1.86 ± 0.16	<LOQ	< LODc	NAd	NAC	NAd

^aPFAAs desorption from 0.5 g WTR using 250 mL DI water, equilibrated for 96 h. n=3;

^bbelow limit of quantification (< LOQ);

^cbelow limit of detection (< LOD);

^dNot applicable

Variations in WTRs’ Resident PFAS Concentrations.

Of all the WTRs (7 DWTRs and 1 WWETR), the number and concentration of individual PFAAs detected were greatest for the ACH-WWETR (Fig. 1). This is consistent with the use of ACH WWETR to treat WWTP effluent, which is expected to have higher concentrations and a more diverse suite of PFAS than typically found in sources for potable water (Rayne and Forest 2009, Houtz, Sutton et al. 2016). To the best of our knowledge, all the DWTRs examined were produced by DWTPs utilizing naturally recharged surface and ground water sources (Table 1). The high basicity (i.e., ratio of hydroxide ions to Al groups) of the ACH WWETR and operational pH range in comparison to other alum coagulants (e.g., other polyaluminum chlorides and aluminum sulfate) has proven to increase its effectiveness to coagulate suspended particles and dissolved organic carbon (DOC) (Yonge, Duranceau et al. 2016, Zaman, Rohani et al. 2021), thus also likely to have a higher propensity to bind PFAS. ACH has the highest soluble aluminum concentration among all commercially available aluminum-based coagulants and has the highest basicity. ACH contains 23 percent to 24 percent Al₂O₃ and a basicity of about 80 - 85% in contrast to alum having about 8.3% Al₂O₃ and no basicity (Gebbie, 2001). High basicity products have a higher cationic charge and are more efficient in coagulating negatively charged contaminants; thus, ACH coagulants are highly effective in removing suspended and dissolved particles required for the final polishing of WWTP effluent (Verma, Bhunia et al. 2012, Zaman, Rohani et al. 2021).

No PFAS were detected in the two Ca-DWTRs. The absence of detectable PFAS may be due to minimal PFAS concentrations in the source water, which was groundwater for both (which typically contains less contaminants that surface water, which is why the less efficient Ca coagulants can be used), as well as the low removal efficiency of organic contaminants and dissolved organic carbon from solution by Ca coagulants (Hendricks 2011, Suez 2021).

The single Fe-DWTR examined contained total PFAS concentrations within the range of the Al-DWTRs sourced from natural waters (Al-1, Al-2, Al-3 and Al-4); Fig. 1). While removal efficiency comparisons cannot be done in absence of source water PFAS characterization, the characterization of the Fe-DWTR indicates that even while the total Fe content was high (>270,000 mg kg⁻¹), the oxalate-extractable (Fe_ox) content was low (~50 mg kg⁻¹). This is in contrast with the high total fraction of Al_ox content (~71% to 93%) found for the Al-WTRs. This suggests that the treatment regime responsible for the production of this media utilized a low pH floc methodology (Hendricks 2011).

Desorption of Resident PFAAs.

The percent PFAA desorbed was estimated by the mass extracted from the WTRs after the desorption step relative to the resident PFAAs determined. For ACH-WWETR, with substantially higher individual and total PFAA concentrations compared to the WTRs, desorption was observed for only the shortest chain PFAA detected (PFBS) at 27% and PFOS at ≤ 2%; all other PFAAs were not observed or < LOQ (Table 3; calculations shown in the SI). PFOS was the only PFAA present in all three WTRs evaluated for desorption with % desorbed as follows: ACH WWETR (2%) < Al DWTR (38%) < Ca DWTR (49%).

PFAA Sorption/Desorption with ACH WWETR.

PFAA sorption from the PFAA-amended biosolid-derived porewater (~3500 ng L⁻¹ total PFAAs) increased with increasing ACH-WWETR concentration (i.e., 0.6 to 56.9 g L⁻¹) and increasing PFAA chain length (Fig. 2a). At the two highest ACH-WWETR concentrations (28.7 and 56.9 g/L), PFHxS, PFOS, PFOA and PFNA were ≥ 90% sorbed. Subsequent desorption of the adsorbed PFAAs ranged from negligible to no more than 26% regardless of the ACH-WWETR concentration (Fig. 2b) although trends with increasing ACH-WWETR concentration varies with PFAS. Longer chain PFAAs showed little to no desorption at the highest amount of ACH-WWETR used. Both PFOS and PFOA desorption was limited to < 2% desorption at the highest WTR concentration (56.9 g/L). In sorption/desorption studies of PFOS and PFOA in acidified deionized water (pH 3.0, total PFAAs 1000 μg L⁻¹) with Al-WTR (10 g/L), Zhang et al. (Zhang, Sarkar et al. 2021) also observed high sorption of PFOS (99.5%) and PFOA (97.4%). Subsequent desorption studies by Zhang et al. (Zhang, Sarkar et al. 2021) also showed low percent desorption (PFOA < 8.1 %; PFOS < 1.9 %).

Figure 2.

A) Average % PFAA sorbed by varying concentrations of ACH WWETRs (0.6-56.9 g ACH-WWETR/L aqueous phase) in PFAA-amended biosolid-derived porewater; and B) subsequent % desorption by deionized water. Bars represent standard deviations.

The equilibration of ACH-WWETR with the biosolids porewater matrix resulted in decreasing supernatant color with increasing mass of ACH-WWETR (Fig. S2). The visible brownish yellow color was presumed to be DOC, which was quantified to be 238 ppm DOC in the diluted biosolids-derived porewater used in the sorption/desorption experiment. In the sorption step, we observed decreasing color (DOC) in the porewater with increasing ACH-WWETR mass. Measured DOC in the supernatants after the sorption step supported this observed change in color. DOC in the post sorption supernatant reduced with increasing ACH-WWETR (Table S5). However, of significance, the decrease in color (DOC) and increase in PFAS sorption occurred with increasing ACH-WWETR suggesting that the DOC did not hinder PFAA sorption by ACH-WWETR. Kothawala et al. 2017 (Kothawala, Kohler et al. 2017) had similar behavior with two types of DOCs and two types of ion exchange resins; the DOC did not significantly impact PFAS sorption to the resins. Hydrophobic interaction can happen between the hydrophobic region of DOC (e.g., aliphatic or aromatic groups) and the hydrophobic tail (–CF₂ chain) of the PFAS. Also, divalent cation bridging can occur with DOC, which enhances sorption between the negatively charged sites on DOC and anionic PFAS. In the desorption step, the reappearance of some color in the higher two ACH-WWETR concentrations (Fig. S2 and Table S5) suggests that some of the DOC subsequently desorbed, which may be linked to some of the apparently inconsistent trends between PFAS in the desorption step. Additional work is needed to better understand mechanism and solution matrix effects, which was outside the scope of this study.

Implications.

Little is known regarding potential PFAS flux into the environmental resulting from land-application of DWTRs and considering the increasingly stringent regulatory environmental related to PFAS in the environment, concerns exist that land application of any waste residual may result in liability and/or environmental harm. Here, we have provided a first of its kind assessment of PFAS concentrations in WTRs as well as desorption from WTRs. For the materials examined here, we found PFAS concentrations of ~34 μg kg⁻¹ in an ACH-WWETR, < 5 μg kg⁻¹ for Al and Fe DWTRs and < LOD in Ca-based DWTRs. Even in the ACH-WWETR, where the highest total PFAS concentrations were found, concentrations are extremely low compared to a recent survey of US biosolids, which reported concentrations ranging from 323 to 1100 μg kg⁻¹.(Schaefer, Hooper et al. 2022) Desorption of resident PFAS from DWTRs was < LOD for all PFCAs analyzed. For PFSAs, desorption ranged from 2-49% for PFSAs, with PFOS desorbing at the highest rate (39% in the Al-DWTR and 49% in the Fe-DWTR). However, in experiments examining sorption/desorption by ACH-WWETRs of spiked PFAAs (~ 3500 ng/L total PFAAs) from biosolids-derived porewater, sorption of PFAAs (including the PFSAs PFHxS and PFOS) increased with increased ACH-WWETR concentrations. Subsequent desorption of all PFAAs were generally < 18% with one exception, demonstrating a more comprehensive capacity for both PFCAs and PFSAs to be sorbed on WTRs. Taken together, these findings suggest that DWTRs are unlikely to be a major source of PFAS via desorption and also that amending environmental matrices with WTRs may provide a beneficial reduction in PFAS mobility.

While further research examining questions such as how influent composition (e.g., dissolved organic matter, salts, and native PFAS concentrations) affects final PFAS partitioning and loading into WTRs is warranted, our findings suggest that PFAS loads into terrestrial ecosystems resulting from land-application of DWTRs would be very low to negligible compared to background soil concentrations (Brusseau, Anderson et al. 2020, Zhu, Khan et al. 2022). Although more research investigating a broader suite of contaminants in a larger number of DWTRs would be beneficial, this initial work suggests that land application of DWTRs is unlikely to be a significant source of PFAS flux into the environment, a finding with important implications for concerns related to beneficial reuse of DWTRs.

Table 2.

Mean and standard deviation (SD) for several chemical WTR properties.

WTR	Total Al1		Total Fe1		Total Ca1		Oxalate Extractable Fe2		Oxalate Extractable Al2		% OM3		pH4		EC4
	mg kg−1		mg kg−1		mg kg−1		mg kg−1		mg kg−1		% by Mass				dS m−1
	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD
Ca-1	818	97	356	33	383886	4971	0.1	<0.1	475	104	1.7	0.0	9.4	<0.1	0.93	0.03
Ca-2	2365	165	424	51	360936	5573	0.2	<0.1	1895	36	2.6	0.3	9.7	<0.1	1.12	0.02
Fe	1527	24	271931	4919	3349	172	54	5.5	712	107	43	0.3	4.4	<0.1	0.33	0.01
Al-1	151010	5473	6016	231	6600	61	1.2	0.1	107570	9380	36	0.3	7.2	<0.1	0.87	0.03
Al-2	116703	890	6090	133	4255	92	0.8	<0.1	102775	1682	43	0.2	7.2	0.2	1.13	0.05
Al-3	90863	1583	3675	610	5098	755	1.1	0.1	72259	7493	24	1.4	6.9	0.2	1.46	0.05
Al-4	93319	417	16879	210	25905	351	1.5	<0.1	76240	4614	19	0.3	7.2	<0.1	0.38	0.05
ACH WWETR	130375	2106	24913	1134	1132	143	4.7	0.1	120789	4481	44	0.7	5.9	0.1	0.45	0.02

¹Total Al, Fe, P, and Ca determined by USEPA 3051a digestion and analysis on ICP-OES.

²Oxalate extractable Al and Fe determined by previously published methodology (Loeppert and Inskeep, 1996).

³Organic matter determination achieved by loss on ignition (LOI) per ESTL method (i.e., 450°C).

⁴Electrical conductivity (EC) and pH determined by ARL methods (ESTL, 2014).

ABBREVIATIONS

ACH

aluminum chlorohydrate

Al_ox

oxalate-extractable aluminum

DOC

dissolved organic carbon

DWTP

drinking water treatment plant

DWTR

drinking water treatment residuals

EC

electrical conductivity

Fe_ox

oxalate-extractable iron

HDPE

high-density polyethylene

HF PO-DA

hexafluoropropylene oxide-dimer acid or GenX

ICP-OES

inductively-coupled optical emission spectrometry

LOD

limit of detection

LOQ

limit of quantitation

MRM

multiple reaction monitoring

OM

organic matter

PFAS

Per- and polyfluorinated alkyl substances

PFAAs

Perfluoroalkyl acids

PFSAs

Perfluoroalkyl sulfonates

PFOA

Perfluorooctanoic acid

PFNA

Perfluorononanoic acid

PFHxS

Perfluorohexanesulphonic acid

PFOS

Perfluorooctanesulfonic acid

PP

polypropylene

USEPA

United States Environmental Protection Agency

SPE

solid-phase extraction

UPLC

ultrahigh pressure liquid chromatograph

WTR

water treatment residual

WWETR

wastewater effluent treatment residual