The overall goal of this project was to attain improved insight into the fundamental fate and transport processes that control per- and polyfluoroalkyl substance (PFAS) fate and transport as well as comingled chlorinated solvents and/or fuel hydrocarbons in groundwater at sites impacted by aqueous film-forming foam (AFFF). The project team focused on the release and transformation of polyfluorinated substances to perfluoroalkyl acids (PFAAs) in source zones as well as the impact of commonly employed remediation technologies for co-occurring chemicals on PFAS fate. The ultimate goal of improved insight into the fundamental fate and transport processes in these mixed chemical systems should help facilitate the development and optimization of treatment strategies for management of PFAS sites comingled with other chemicals of concern. Specific objectives of this project included:
- Investigation of the fundamental mechanisms controlling the release of PFAS from complex source-zone phases. (Task 1)
- Examination of the coupled diffusion and potential abiotic reactions of PFAS and comingled contaminants in low permeability materials. (Task 2)
- Assessment of the biotic transformation of the wide range of PFAS and co-occurring chemicals (chlorinated solvents and benzene, toluene, ethylbenzene, and xylene [BTEX]) present in the dissolved plume and the impacts of PFAS on co-occurring chemical bioremediation. (Task 3)
- Quantification of the impacts of remedial activities targeting co-occurring chemicals of conern (i.e., BTEX, chlorinated solvents) on the PFAS plume. (Task 4)
The hypothesis of this project is that the complex nature of PFAS present in AFFF results in complex and long-term release from both primary sources (which may contain non-aqueous phase liquids [NAPLs]) and low permeability zones. Further, coupled biotic and abiotic transformation of polyfluorinated PFAS to the problematic PFAAs are impacted by site mineral and geochemical conditions as well as the application of conventional treatment technologies. The application of enhanced aerobic transformation of aromatic hydrocarbons (biosparging), biostimulated reductive dechlorination of chlorinated ethenes, and application of in situ chemical oxidation (ISCO) have unknown impacts on the release and transformation of polyfluorinated PFAS to the more problematic PFAAs.
This project examined these impacts and interactions through carefully controlled batch and column laboratory experiments. However, to represent the true complexity of PFAS chemistry and co-occurring chemical interactions, field-collected groundwater and soil/aquifer samples were employed to serve as the primary source of these constituent compounds. Comprehensive analytical methodologies were employed, targeting not only the ~200 PFAS relevant to these sites, but also the potential for yet unidentified PFAS and volatile intermediates. Of particular interest are interactions of polyfluorinated PFAS with NAPLs, the diffusive transport and transformation of PFAS in clays, the biological transformation of PFAS present in AFFF, and the advective release and transformation of PFAS from field-collected materials under simulated remediation conditions.
The results from Task 1 indicate that at least for the two field-collected soils examined, PFAS elution clearly varied in relation to both the head group as well as perfluorinated chain length when the soils were flushed with artificial groundwater. Indeed, several PFAS were only very slow to elute (>100 pore volumes before significant elution was observed). This has important implications for site monitoring and remediation, as some classes, including potential PFAA-precursors did not appear in the effluent until after the highly oxidized PFAA were largely eluted. Elution of some slowly eluting PFAS such as perfluorooctanesulfonamide was significantly impacted by porewater pH, suggest that soil buffering capacity is highly relevant to PFAS transport and that remedial actions that produce large groundwater pH shifts may cause rapid release of PFAS that are typically less mobile.
Task 1 also revealed that PFAA partitioning into bulk Jet Fuel A was not observed for PFAA below eight carbons, though interfacial sorption coefficients for PFAS ranged from 0.06 – 0.26 centimeters. Perfluorocarboxylates with 11 – 14 carbons showed greater accumulation at the interface than shorter-chained compounds. Further, experiments to determine the critical micelle concentrations (CMC) for specific AFFF formulations indicated that the CMC was below the 3% AFFF application rate, indicating that micelles are present when AFFF are applied to fires. Reduced surface and interfacial tensions of AFFF-impacted waters after application can potentially increase soil wetting and infiltration. The presence of micelles also increases potential for non-aqueous phase liquid and other co-occurring chemical transport.
Under Task 2, experimentally determined aqueous diffusivities were in good agreement with previously performed experiments that measured PFAA uptake into activated carbon. The WilkeChang model was not able to accurately predict the values nor describe the trend for the range of PFAA used in this study. The non-uniform trend in the PFAA diffusivities with respect to the molar volumes likely is reflective of unique molecular interactions associated with highly fluorinated compounds. In the clay soil tube diffusion experiments, comparison of the results to numeric models showed that neglecting surface diffusion resulted in a severe underprediction (>10x for perfluorooctane sulfonate) of predicted aqueous concentrations. For the PFAA and clay soil examined herein, surface diffusion contributions became important (>10% of the overall diffusion) at Kd (concentration of antibody) values greater than approximately 0.5 L kg-1. Abiotic reactions with reactive iron, however, appear unlikely to be important with respect to reductive transformation of polyfluorinated PFAA precursors in AFFF. Ferrous iron minerals, however, may play a small role in precursor transformation under oxic conditions due to generation of hydroxyl radicals.
Under the third objective, a 3M* AFFF formulation and a National Foam AFFF formulation had different responses to aerobic transformation of non-fluorinated and fluorinated components of AFFF. While there were initial decreases in a specific sulfonamide precursor observed in 3M AFFF, N-dimethyl ammonio propyl perfluorohexane sulfonamide (AmPr-FHxSA), in the live treatments, National Foam AFFF did not show a decoupling of the fluorotelomer compound 6:2 fluorotelomer sulfonamido betaine between the live treatment and autoclaved control. However, the non-fluorinated carbon (measured as total organic carbon; total organic carbon [TOC]) in the National Foam AFFF microcosms was apparently degraded, whereas the TOC in 3M stayed constant in the live treatment. More importantly, 3M AFFF completely inhibited BTEX biodegradation, though no inhibition was seen with AmPr-FHxSA alone. However, differences in extracellular and total metabolite abundances between AmPr-FHxSA, perfluorohexane sulfonamide (FHxSA), and perfluorohexane sulfonate (PFHxS) treatments compared to the PFAS-free control suggested AmPr-FHxSA (and perhaps other zwitterionic PFAS) may cause cell membrane leakage. Nevertheless, experiments with AmPr-FHxSA indicated that over the course of 70 days, it was transformed into FHxSA, and to a lesser extent PFHxS. Anaerobic microorganisms (trichloroethene [TCE] coculture, anaerobic BTEX biodegradation) appear to be more tolerant of AFFF than aerobic systems. Treatments with spiked AmPr-FHxSA had slowed TCE dehalogenation rates with no lag phase, suggesting that microorganisms did not recover from the inhibition. Addition of diethyl glycol monobutyl ether individually or via AFFF resulted in stimulation of TCE dehalogenation, presumably because of glycol fermentation.
Results from the final objective revealed that when aerobic microorganisms were specifically stimulated to mimic biosparging, transformation of multiple electrochemical fluorination (ECF) sulfonamide precursors was suspected based on mass balance, despite the surface soil being exposed to air and water for years prior to collection: source zone soils likely still have potential for PFAS transformation years after AFFF release. Both O2-sparged and N2-sparged columns released more fractional mass of ECF sulfonamides than unaltered, non-biologically active columns (i.e., in comparison to Task 1). After 200+ pore volumes of flushing, detectable concentrations of many PFAS, particularly long chain perfluoroalkane sulfonate and zwitterionic compounds, remained on the column soils. In contrast, the in situ chemical oxidation (ISCO) simulation experiments, perfluoro-n-hexanoic acid, perfluoro-n-octanoic acid, and PFHxS increased in both the permanganate- and persulfate-treated columns. All three example PFAS increased more rapidly and to higher concentrations in the permanganate-treated columns than the persulfate-treated columns, but it remains unclear whether this difference is attributed to the chemical oxidants themselves or the different concentrations of each oxidant.
In summary, this project confirmed that there are significant reservoirs of polyfluorinated substances still remaining on AFFF-impacted soils that can be very slowly released to groundwater. The release of these PFAS is dependent on both the perfluorinated tail length and the head group. Some of these PFAS (particularly those present in 3M AFFF) also inhibit microbial activity, with anaerobic communities appearing to be more tolerant to AFFF than aerobic microbial communities. Some ECF-derived polyfluorinated substances do appear to transform, albeit slowly, to PFAA such as PFHxS, though perfluoroalkyl sulfonamide such as FHxSA may be semi-stable intermediates. Natural abiotic subsurface reactions with polyfluorinated substances are likely not significant with respect to PFAS transformations. Finally, data collected under this project indicate that alterations in subsurface biogeochemistry, whether through alterations in soil porewater pH or changing redox conditions due to biosparging or ISCO, can significantly impact the time and magnitude of the release of PFAS mass from AFFF-impacted soils.
Garcia, R., A. Chiaia-Hernández, P. A. Lara-Martín, M. Loos, J. Hollender, K. Oetjen, C. P. Higgins and J.A. Field. 2019. Suspect Screening of Hydrocarbon Surfactants in AFFFs and AFFF-Contaminated Groundwater by High Resolution Mass Spectrometry. Environmental Science and Technology, 53:8068−8077. DOI: doi/10.1021/acs.est.9b01895
Maizel, A.C., S. Shea, A. Nickerson, C.E. Schaefer, and C. P. Higgins. 2021. Release of Per- and Polyfluoroalkyl Substances from Aqueous Film-Forming Foam Impacted Soils. Environmental Science and Technology, 55(21):14617-14627. DOI:10.1021.acs.est.1c02871.
Nickerson, A., A. Maizel, C. Olivares, C. E. Schaefer, and C. P. Higgins. 2021. Simulating Impacts of Biosparging on Release and Transformation of PFASs from AFFF-Impacted Soil. Environmental Science and Technology, 55(23):15744-15753. DOI:10.1021/acs.est.1c03448.
Nickerson, A., A. Rodowa, D. T. Adamson, J. A. Field, P. Kulkarni, J. Kornuc, and C. P. Higgins. 2021. Spatial Trends of Anionic, Zwitterionic, and Cationic PFASs at an AFFF-Impacted Site. Environmental Science and Technology, 55(1):313-323. DOI:10.1021.acs.est.0c04473.
Nickerson, A., A. Maizel, P. Kulkarni, D. T. Adamson, J. Kornuc, and C. P. Higgins. 2020. Enhanced Extraction of AFFF-Associated PFASs from Source Zone Soils. Environmental Science and Technology, 54(8): 4952–4962. DOI: doi/10.1021/acs.est.0c00792
Schaefer, C. E., D. M. Drennan, D. N. Tran, R. Garcia, E. Christie, C. P. Higgins, and J. A. Field. 2019. Measurement of Aqueous Diffusivities for Perfluoroalkyl Acids Journal of Environmental Engineering, 145(11):06019006. DOI: 10.1061/(ASCE)EE.1943-7870.0001585.
Schaefer, C. E., D. Drennan, A. Nickerson, A. Maizel and C. P. Higgins. 2021. Diffusion of Perfluoroalkyl Acids through Clay-Rich Soil. Journal of Contaminant Hydrology. 241: 103814. DOI: doi.org/10.1016/j.jconhyd.2021.103814