Effects of Organic Matter, Iron (Hydr)oxides, and Iron Reductive Dissolution on Per- and Polyfluoroalkyl Substances Sorption to Aqueous Film-Forming Foam-Impacted Solids

Abstract

Aqueous film-forming foams (AFFFs) are a major source of per- and polyfluoroalkyl substances (PFAS) contamination in groundwater and soil. Although PFAS sorption is known to depend on the properties of solid matrices, the roles of organic matter and iron (hydr)oxides remain poorly understood, especially for polyfluorinated compounds. To investigate the interplay between particulate organic matter and iron (hydr)oxides in PFAS-contaminated subsurface environments, we studied the partitioning of neutral, cationic, and anionic PFAS in AFFF to well-characterized solids using the mixed-mode solid-phase extraction (SPE) technique combined with the total oxidizable precursor (TOP) assay. Our results indicated that organic matter consistently enhanced PFAS sorption. In contrast, the effect of goethite, a representative iron-containing mineral, on PFAS sorption varied with PFAS charge. Using three different AFFFs, we found that anionic PFAS exhibited strong sorption regardless of organic matter content, while cationic and zwitterionic PFAS sorbed poorly at circumneutral pH. These trends were more pronounced for iron-coated sand, which showed higher affinity for anionic PFAS. Treatment with ascorbate, a mild reductant, released PFAS associated with iron (hydr)oxides. These findings highlight the importance of iron (hydr)oxides in the retention of anionic PFAS and could be leveraged to manage PFAS mobility through subsurface iron amendments or in situ iron (hydr)oxide formation/dissolution.

Graphical Abstract

Introduction

Per- and polyfluoroalkyl substances (PFAS) persist in the environment and can pose risks to human health and wildlife.^1–4 Aqueous film-forming foams (AFFFs),^{5, 6} are a major source of PFAS exposure. When AFFF is released on land, much of the mass of PFAS accumulates in soil, in the vadose zone or in shallow groundwater near the source.^{7, 8} Perfluorocarboxylic acids (PFCAs) and perfluorosulfonic acids (PFSAs) exhibit low pK_a values (<0.3–3.8),^{9, 10} and are therefore negatively charged under typical environmental conditions. Other PFAS, which are often termed precursors because they can be transformed into perfluoroalkyl acids (PFAAs) in the environment, exhibit different charge states due to the higher pK_a values of perfluoro-sulfonamides and amines (e.g., pK_a values for AmPr-FHxSA and TAmPr-FHxSA range from 3.3 to 6.5).^11–13

A key unresolved question is how iron (hydr)oxides—common, reactive mineral phases that exhibit pH-dependent charge behavior—modulate PFAS retention and release across these charge classes under field-relevant conditions. With points of zero charge (PZC) between 7 and 9.5,^14–18 iron (hydr)oxides are often positively charged at circumneutral pH, thereby enhancing electrostatic attraction of anionic PFAS. Consistent with this, prior researchers have reported delayed PFOS breakthrough in iron (hydr)oxide–coated media¹⁹ and higher distribution coefficients for anionic PFAS in Fe-rich solids.^{20, 21}

In contrast, most subsurface solids exhibit net negative charge at circumneutral pH values, which favors retention of cationic PFAS over anionic species.^{12, 21–26} For example, organic matter coatings on solids tend to be negatively charged. Particulate organic matter also promotes PFAS sorption via hydrophobic interactions between perfluoroalkyl chains and organic matter domains;^{20, 21, 24, 27–29} an attractive force that increases with PFAS chain length. Thus, the effects of iron on PFAS sorption may be confounded by the presence of high concentrations of natural organic matter.³⁰ Prior studies have illustrated the way that iron (hydr)oxides and organic matter affect PFAS coating, but these focused on PFAAs and a small number of other PFAS for which analytical standards are available. Consequently, the overall importance of iron (hydr)oxides to retention of the complex suite of PFAS in AFFF in the subsurface remains unclear.

To address this knowledge gap, we combined a mixed-mode solid-phase extraction (SPE) technique with the TOP assay (mixed-mode TOP assay) to quantify sorption of a suite of AFFF constituents.^{31, 32} We equilibrated diluted AFFFs from three sources with well-defined mixtures of sand and iron (hydr)oxides with and without particulate organic matter under field-relevant conditions to isolate the effects of iron (hydr)oxides on PFAS sorption. We also studied the importance of reductive dissolution of iron-(hydr)oxide coatings to the release of sorbed PFAS. Together, this integrated approach extends our knowledge of PFAS mobility and provides new mechanistic insight into PFAS–mineral–organic matter interactions.

Materials and Methods

Materials.

All materials and chemicals used in this study were of the highest available purity, unless otherwise specified, and were used without further purification. Two fluorotelomer-based AFFFs (Ansulite and Forafac) and one electrochemical fluorination-based AFFF (Peterson Bucket) were used, with additional details provided in Text S1. All PFAS analytical standards for analysis were purchased from Wellington laboratories (Guelph, ON, Canada).

Solids Loading and Elution.

Four different solids were used to assess the effects of solid composition on the affinity of PFAS (Figure S1): a) 100% Ottawa sand (particle size <0.6–0.8 mm; Fisher Scientific) (Sand); b) a mixture of 95% (w/w) Ottawa sand and 5% (w/w) goethite (particle size <0.4 μm;³³ Sigma Aldrich) (Sand+Fe); c) a mixture of 70% (w/w) Ottawa sand and 30% (w/w) Pahokee peat soil (Catalog number 2BS103P; International Humic Substances Society; see Table S1 for details) (Sand+Peat); d) a mixture of 65% (w/w) Ottawa sand, 30% (w/w) Pahokee peat soil, and 5% (w/w) goethite (Sand+Peat+Fe).

Approximately 25 g of each set of solids was equilibrated with 25 mL of AFFFs diluted with a synthetic freshwater solution (pH 8; details are available in Table S2) in 50 mL methanol-rinsed polypropylene (PP) tubes. After equilibration, 10 g solid subsamples were air-dried and stored at −20°C for up to two weeks prior to PFAS extraction and quantification. The remaining 15 g subsample of PFAS-containing solids was eluted with 150 mL of the synthetic freshwater in 10-cm glass columns. The details of solid loading with AFFFs and the elution procedure are provided in Text S1, which also describes the measurement of the PZC of goethite. After elution, solids were air-dried and stored at −20°C for no more than two weeks prior to extraction and quantification of PFAS. All experiments were performed in triplicate.

PFAS Quantification with TOP Assay and Mixed-Mode TOP Assay.

PFAS were extracted from a 0.5 g subsample of the air-dried solids after the loading and elution steps. PFAS were quantified before and after oxidation as part of the TOP assay as described previously.^{25, 34} PFAAs measured before the TOP assay were quantified using the analytical standards and differences in PFCAs concentrations before and after the TOP assay (ΔPFCAs) were used as a measure of precursors. Details of the PFAS extraction and the TOP assay procedures are provided in Text S2. AFFF samples used in the study were characterized with the TOP assay (Figure S2–3), and mixed-mode TOP Assay was used to provide electrostatic charge information.³¹ Detailed procedures for mixed-mode TOP are provided in Text S3.

Reductive Dissolution of Iron-Coated Sand.

Ottawa sand pre-rinsed with LC-MS grade water was iron-coated using an FeSO₄•7H₂O (Fisher Scientific) solution, which exhibited a distinct red-orange color (Figure S4), following a modified version of a previously described iron-coating procedure.³⁵ The amount of Fe(III) on the sand was determined with the ferrozine method following Fe extraction with acidic ammonium oxalate (AAO).³⁶ Detailed procedures for Fe coating, Fe extraction, and ferrozine-based quantification³⁷ are provided in Text S4.

A 20 g subsample of the iron-coated sand or 20 g of uncoated sand were mixed with 20 mL of Ansulite, anionic rich AFFF, diluted 300-fold in 0.1 M HEPES buffer (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; Fisher Scientific) adjusted to pH 8 with 0.1 N NaOH and HCl. Samples containers were covered with aluminum foil prior to placement on an orbital shaker at 20 rpm for 7 days.^{30, 38} The 0.1 M HEPES, which exhibits a low affinity for iron species,³⁹ was used to buffer pH during the iron reduction and ascorbate addition. After equilibration samples were centrifuged at 10,000 g for 10 min. The resulting pellets were collected and air-dried.

To investigate the potential release of PFAS during iron reduction, PFAS-containing iron-coated sand was subjected to reductive dissolution using ascorbate as the reductant. The details of iron reductive dissolution experiments are available in Text S5.

Analytical Methods.

The samples were analyzed following established procedures.^{31, 34} PFAS containing extracts were analyzed by liquid chromatography-tandem mass spectrometry (LC-MS/MS) with an Agilent 1260 Infinity II HPLC system, coupled with an Agilent 6470B triple quadrupole mass spectrometer. PFCAs and PFSAs were separated using reverse-phase chromatography with a gradient solvent program. Details regarding QA/QC procedures and analytical conditions are provided in Text S6 and Tables S3–4.

Results and Discussion

Effect of Iron (Hydr)oxide on PFAA Sorption at Low- and High-f_oc.

The sorption of anionic PFAAs increased when goethite was present (Figure 1a–c). For example, in Peterson Bucket-impacted solids, PFOS sorption was about four times higher with goethite (Sand+Fe), relative to sand alone (Sand). Given that the solution pH (7.4–7.6; Figure S5) was below goethite’s PZC (7.9–8.2; Figure S6), this pattern is consistent with electrostatic attraction to positively charged goethite contributing to PFAS partitioning.^19–21

Figure 1.

(a-c) PFAAs (i.e., PFCAs and PFSAs) and (d-f) ΔPFCAs associated with solids after elution with synthetic freshwater. Different mixtures of sand, Pahokee peat soil, and goethite were pre-loaded with (a, d) Ansulite, (b, e) Peterson Bucket, or (c, f) Forafac for one week to reach equilibrium. ΔPFCAs were estimated based on the difference in PFCAs concentrations before and after the TOP assay. Error bars represent the standard deviation of data across triplicates. The subpanels in panels (a, c) show different y-axis ranges to improve the visibility of low sorption values. Different letters on bars denote a significant difference (p < 0.05) between total sorbed amounts (sum of the heights of stacked bars) determined using a one-way ANOVA with Tukey’s honestly significant difference (HSD) post hoc testing. The schematic diagram of loading and elution steps are described in Figure S1.

In high-f_oc (14%; the upper end of the environmental range)⁴⁰ containing solids characteristic of peaty soil horizons, wetlands and organic-amended soils, PFAAs partitioning to solids were up to about 30% higher when goethite was present (Sand + Peat + Fe) than in Sand + Peat, consistent with electrostatic contributions from goethite. In these experiments, the peat induced a pH decrease (the pH values dropped to pH 4.3–4.5; Figure S5), which also may have increased the surface charge of the goethite and, in turn, enhanced electrostatic attraction of anionic PFAAs to the solids, alongside OM–goethite interactions such as charge screening and competition at anion-exchange sites.

Effect of Iron (Hydr)oxide on Precursor Sorption.

The partitioning of PFAS from three different AFFF formulations depended upon the composition of the AFFF. These differences could be explained by the charge state of the precursors. Negatively charged compounds accounted for about half of the polyfluorinated compounds in the Ansulite AFFF, whereas negatively charged compounds accounted for only about 10% and 3% of the polyfluorinated compounds in Peterson bucket and Forafac, respectively (Figure 2b and c). Ansulite-impacted solids (high in anionic PFAS; Figure 2a) showed results that mirrored PFAA trends: goethite increased sorption in both matrices—an increase in partitioning of about 3.6 times was observed in the absence of peat and an increase of about 30% was observed in the presence of high concentrations of organic matter—consistent with the positively charged goethite increasing electrostatic attraction of the negatively charged precursors. It is worth noting that the axis scale for Figures 1d–f are about six times higher than that of Figures 1a–c.

Figure 2.

Charge fractions of PFAS in diluted AFFFs (a) Ansulite, (b) Peterson Bucket, and (c) Forafac, measured using mixed-mode SPE combined with the TOP assay (ΔPFCAs). The neutral fraction includes both uncharged compounds and zwitterionic species with a net charge of zero. The ability of the mixed-mode TOP assay to separate these compounds in diluted AFFFs was validated by Antell et al. (2025).³¹

For the Peterson Bucket–impacted solids (90% cationic/neutral; Figure 2b), goethite reduced sorption by about 30% in the presence of peat. In the absence of peat, a smaller mass of PFAS underwent sorption during the loading step (see second column of Figure S7). During elution, a small increase in PFAS retention was observed (see second column of Figure S7), consistent with uptake of the relatively small amount of anionic PFAS by goethite (Figure 2b). In Forafac-impacted solids (96% cationic/neutral; Figure 2c), goethite-containing treatments exhibited about 65% less sorption in the absence of peat. No differences in sorption were observed in the presence of peat because at these high-f_oc values the organic matter-related sorption masked any effects of goethite (Figure 1f).

Across all conditions, the sorption of precursors (i.e., ΔPFCAs) from Ansulite were relatively low after both loading and elution (Figure 1a and the first column of Figure S7). As a result, more ΔPFCAs were flushed from those solids compared to other AFFF-impacted solids. The effect was most pronounced under high-f_oc (14%) conditions, which (i) promoted hydrophobic partitioning/ion-pairing of neutral or cationic species and (ii) electrostatic conditions that were unfavorable to anionic precursors; accordingly, at high-f_oc, anionic precursors sorbed least, while neutral/cationic were preferentially retained. This finding aligns with previous studies showing that zwitterionic and cationic PFAS exhibit reduced mobility in the solids, whereas anionic PFAS are more mobile due to the repulsion from net-negatively charged organic matter in the solids.^{12, 21–26} Additionally, precursor sorption (Figure 1d–f) exceeded PFAA sorption (Figure 1a–c) by several orders of magnitude across all solids, underscoring the importance of targeting PFAS beyond PFAAs to better manage AFFF-impacted sites. Because we employed an environmentally relevant eluent (synthetic stormwater) and well-defined solids, a pH decrease at high organic-matter loadings was inherent to the system; accordingly, pH-matched comparisons warrant future investigation.

Sorption and Release of PFAS from Iron-Coated Sand.

After 7 days of equilibration, organic matter-free iron-coated sand (Figure 3a) exposed to anionic-PFAS-rich Ansulite AFFF in 0.1 M HEPES buffer (pH 8) adsorbed about an order of magnitude more ΔPFCAs (Figure 3b) and 76% more PFAAs than uncoated sand (Figure 3c). This finding demonstrates the pronounced influence of iron (hydr)oxides on PFAS solid-aqueous partitioning under environmentally relevant pH conditions with typical iron concentrations and no (or very low) particulate organic matter content. The observed enhancement exceeded that of the goethite amendment described above—likely due to the absence of organic matter—and aligns with previous reports of strong interactions between iron (hydr)oxides and PFOS and PFOA.^41–43

Figure 3.

Iron (a) and PFAS (ΔPFCAs (b) and PFAAs (c)) associated with sand and Fe-coated sand after 7 days of equilibration with Ansulite AFFF; and iron (d) and PFAS (ΔPFCAs (e) and PFAAs (f)) retained on Fe-coated sand after 24 h of equilibration with and without 10 mM ascorbate. Iron coating was extracted with the AAO method³⁶ and quantified using the ferrozine method.³⁷ ΔPFCAs were quantified using the total oxidizable precursors (TOP) assay. Error bars represent the standard deviation of triplicate samples. Different letters above bars in PFAS panels (b, c, e, f) denote significant differences (p < 0.05) in total PFAS sorbed on solids (sum of stacked bars), determined using a one-way ANOVA with Tukey’s honestly significant difference (HSD) post hoc testing. Iron mass balance between solid- and liquid-phase is available in Figure S9.

To assess the potential for reductive dissolution to release PFAS from the iron-coated sand, PFAS-contaminated solids were equilibrated for 24 h in fresh HEPES buffer (pH 8.0), with and without ascorbate. The ascorbate treatment resulted in the reductive dissolution of approximately half of the Fe(III), whereas only about 14% of the Fe-coating was released from the control sample subjected to that same pH 8.0 solution without ascorbate (Figure 3d). Both the control and ascorbate treated samples showed reduced sorption compared to the initial values; however, ascorbate exposure resulted in significantly less retention of ΔPFCAs and PFAAs than the control, by factors of 2.2 and 1.5, respectively (Figure 3e and f). An additional control experiment confirmed that ascorbate does not directly react with PFAS (Figure S8). These results indicate that reductive dissolution of Fe(III)-(hydr)oxides most likely promotes PFAS release due to the loss of sorption sites (Figure 3d and S9).⁴⁴ Alternatively, under reducing conditions, Fe(III)-(hydr)oxides can be converted into phases with a lower PZC (e.g., magnetite or siderite exhibit PZC below 7),^{45, 46} which would have less favorable electrostatic interactions with anionic PFAS. Further research is needed to determine the relative importance of these two processes.

Furthermore, reductive dissolution of iron (hydr)oxides can liberate organic matter tied up in mineral-organic associations.^{47, 48} The co-release of this organic matter may mobilize additional PFAS, particularly cationic and zwitterionic PFAS that associate strongly with organic matter.^{12, 22–24} The released organic matter also may stimulate microbial activity. In the subsurface solids subject to frequent redox oscillations—such as those in tropical forests—reduction of iron minerals followed by reprecipitation can alter iron speciation,^{49, 50} leading to shifts in dominant anaerobic microbial respiration pathways.⁵¹ Collectively, these changes in microbial processes may influence PFAS biotransformation dynamics. However, it is important to note that iron released during reductive dissolution can also enhance PFAS sorption through cation bridging together with Ca²⁺ and Mg²⁺ (Figure S10). Specifically, reduced iron may form bridges between anionic PFAS and negatively charged organic matter, offsetting the mobilizing effects of organic matter mineralization and iron dissolution.⁵² At the same time, anion enrichment from solid leaching can counteract the electrostatic effects of iron minerals and their reduced forms. The complex interplay among iron (hydr)oxides, organic matter, and PFAS under reducing conditions warrants further investigation. Because Fe(III)-(hydr)oxides were expected to have the strongest interactions with anionic PFAS, reduction experiments focused on these compounds; the behavior of cationic and zwitterionic PFAS remains to be established. Ascorbate reduced about half of the Fe(III)-(hydr)oxides on the sand, as evidenced by the release of dissolved Fe(II) (Figure S9). Although microbes are usually involved in reductive dissolution of iron (hydr)oxides in the subsurface, ascorbate provided a fast and reproducible approach for assessing the role of Fe(III)-(hydr)oxide transformation. Additional research is needed to assess the importance of Fe(III) reduction to PFAS fate under conditions that more closely mimic biological processes that occur in the subsurface. Finally, AFFF formulations contain organic solvents, such as diethylene glycol butyl ether (DGBE), and hydrocarbon surfactants, such as sodium dodecyl sulfate (SDS), alkylbenzene sulfonates, and alcohol ethoxylates. These co-contaminants are known to influence PFAS sorption.^{27, 53} However, it is unlikely that they were transformed by the reductant (ascorbate) used in these experiments. Research is needed to assess the impacts of biotransformation of these co-contaminants on the sorption of PFAS under reducing redox conditions.

Environmental Implications.

Our study employs a mixed-mode TOP assay to broaden targets and resolve charge effects, demonstrating that iron minerals increase anionic PFAS sorption but exert limited influence on zwitterionic and cationic species. The charge-dependent effects on PFAS sorption demonstrated here using well-characterized solids expand upon prior studies that highlighted the role of iron minerals in enhancing anionic PFAAs sorption.^41–43 Conversely, other researchers have reported inhibitory effects of iron minerals on zwitterionic PFAS retention in environmental solids, though without isolating the specific contribution of the iron minerals.^{20, 21} Thus, PFAS sorption behavior is expected to vary depending on AFFF composition and site-specific soil characteristics (e.g., iron and organic matter content), underscoring the importance of characterizing both AFFF formulations and geochemical conditions at contaminated sites. While our study employed a reductionist approach within a controlled range of organic matter and iron mineral concentrations, it lays a foundation for future research that examines a broader range of environmental variables, including different iron mineral phases and phyllosilicates.

The demonstrated ability of iron (hydr)oxides to strongly retain anionic PFAS regardless of organic matter contents highlights a promising strategy for mitigating PFAS mobility through targeted geochemical amendments. In addition, the tendency of reductive dissolution of iron (hydr)oxides to release anionic PFAS indicates the importance of controlling redox and provides a potential mechanism for controlling PFAS mobility in the subsurface. Our results suggest that in situ formation of iron (hydr)oxides in the subsurface could be leveraged to create reactive barriers that slow the migration of anionic PFAS. This approach may be particularly useful in managing PFAS plumes in contaminated groundwater or enhancing retention in engineered systems. While dynamic redox conditions will influence PFAS release and transformation, future research should investigate Fe-based amendments or natural precipitation processes that can be optimized to trap PFAS in situ and limit their mobility.

Supplementary Material

Additional details of PFAS analysis methods including LC-MS/MS molecular and product ion scans; information of PFCAs and PFSAs; acronyms used in the study; and ion information for the AFFF formulations. A list of acronyms is provided in Table S5.

Synopsis.

Understanding the interplay between organic matter and iron (hydr)oxides and iron reductive dissolution offers insights into the fate and potential remediation strategies for PFAS on AFFF-impacted solids.