Despite the relative water solubility of per- and polyfluoroalkyl substances (PFAS), source zones retain PFAS even when decades have passed since the last aqueous film-forming foam (AFFF) applications. Retention of PFAS when present in complex AFFF mixtures, including successive AFFF applications, is unknown since most transport studies investigate single PFAS or a simple mixture at low concentrations. Interactions with AFFF and nonaqueous phase liquids (NAPL, e.g., jet fuel) are unknown, yet may form thermodynamically stable phases with distinct characteristics. Discharge of chemically complex AFFF to unsaturated soils involves PFAS interactions with the air-water interface that are also not fully understood.
This effort is being conducted in two phases. The overall objectives of Phase I of this project were to 1) identify the key PFAS and soil properties that control PFAS interaction with saturated soils when AFFF is applied at near application strength concentrations, 2) characterize the number and type of thermodynamically-stable phases that form when AFFF is mixed with jet fuel and the impact on PFAS transport, and 3) assess PFAS mobility under unsaturated zone conditions in order to identify the key hydraulic parameters and PFAS properties that control PFAS mobility and retention.
The main objectives of Phase II of this project are to 1) identify potential field sites impacted by LNAPL and PFAS in the form of thermodynamically stable microemulsions or mixtures; 2) document sites with PFAS-impacted LNAPL and identify factors for predicting the occurrence of such phases; 3) develop a predictive tool/model to simulate past, present and future PFAS-impacted LNAPL behavior; and 4) identify potential technologies to recover and separate PFAS from LNAPL.
The technical approach for Phase I consisted of highly controlled laboratory column experiments with soil and sediment under unsaturated (aerobic) and saturated (anoxic) conditions. Batch tests were performed to characterize the number of thermodynamically-stable phases that form when AFFF and NAPL are mixed and to quantify PFAS sorption at the air-water interface. For each experimental system, the research team quantified the transport characteristics of 42 PFAS present in 3M AFFF, including perfluorooctane sulfonate (PFOS) and perfluroooctanoate (PFOA). A total of 11 classes of PFAS were detected in 3M AFFF including four anionic, four cationic, and three zwitterionic classes.
The Phase I tasks were designed to test the following hypotheses:
Through SERDP projects ER-2126 (Higgins, lead investigator) and ER-2128 (Field, lead investigator), the team now possesses the capacity to quantify the individual PFAS, which is a technical skill necessary to unravel the complex nature of AFFF mixtures; their interactions with other surfactants, fuels, and solvents; and their behavior in soils/sediments.
The technical approach for Phase II consists of identifying sites impacted by both LNAPL and PFAS. Field sampling will focus on quantifying the number of phases obtained during groundwater well sampling at impacted sites followed by quantifying target and suspect PFAS concentrations associated with LNAPL and/or groundwater phases. Direct push cores will be collected from representative sites and the presence of LNAPL will be assessed using laser-induced fluorescence with the ultra-violet optical screening tool to guide the location (depth interval) of soil to be sampled. Data analysis tools will be used to identify the features from field data that contribute most to PFAS retention by LNAPL. Laboratory batch studies and flow experiments will be performed to determine the best approaches for mobilizing and recovering PFAS-impacted LNAPL. Additives known to reduce viscosity will be tested for effectiveness in mobilization, such as controlling pH and salinity. A series of flow experiments using site soils, realistic environmental conditions, and additives will be performed to recover the PFAS-impacted LNAPL. Using results from the flow experiment, methods to recover PFAS-impacted LNAPL will be simulated to test the practicality of PFAS-impacted LNAPL recovery in actual field settings. A review of potential methods to separate PFAS from LNAPL ex situ will also be performed.
The objectives and subtasks of Phase II are designed to test the following hypotheses:
Under saturated conditions in Phase I, PFAS introduced as a complex AFFF were retained by clean sand and a natural diatomic carbon soil horizon under dynamic flow conditions, indicating significant adsorption even when organic carbon and surface charge were negligible. PFAS retention was related to PFAS structure and soil properties through a machine learning-based poly-parameter quantitative structure-parameter linear regression model. PFAS retention by clean sand increased over the first three successive AFFF additions, reaching a saturation limit after four applications. Thus, sorption of PFAS was enhanced by earlier AFFF applications. Viscous Winsor Type II microemulsions (gels) formed when AFFF mixed at the pore scale with NAPL during infiltration. Under unsaturated conditions, sorption of PFAS at environmentally relevant concentrations at the air-water interface was best modeled by Freundlich models. Correlations between air-water partition coefficients (Kia) and chromatographic retention times indicated Kia values can be predicted from chromatographic retention times for a wide range of PFAS. A mass transfer index was required to model the correlation between immobile water and tortuosity as saturation decreased in unsaturated soils and was critical for accounting for PFAS interfacial activity in unsaturated soils. Phase I laboratory data also indicated that LNAPL forms a unique phase with AFFF, forming microemulsions (LPME) that contain mg/L levels of PFAS. Field data from Navy sites indicated that LNAPL also forms a mixture with lower concentrations (µg/L) of PFAS (LPmix). This will be further explored during Phase II.
Greater retention of PFAS in porous media when compared to studies on single components or simple mixtures underscores the importance of working with actual AFFF at application strength. PFAS transport is dynamic and is nearly independent of soil properties, making PFAS transport difficult to model using conventional models. The experiments performed in Phase I provided the necessary data to identify key uncertainties in processes that control the nature and permanence of PFAS interactions with key components of source zone soils under unsaturated and saturated conditions, as well as in the presence of NAPL. The information gained from Phase I identified retention by LNAPL as a previously unidentified source of PFAS at field sites. Phase II will identify field strategies for identifying, mobilizing, recovering, and separating PFAS-impacted LNAPL at field sites. (Anticipated Phase II Completion - 2027)
Christie, E.C., T. Schwichtenberg, C. Schmokel, M.L. Kim-Fu, A.R. Moll, I.A. Titaley, K. Kostarelos, W.H. DiGuiseppi, and J.A. Field. 2023. Per- and Polyfluoroalkyl Substances in Field-Collected Light Non-Aqueous Phase Liquids. ACS Environmental Science & Technology Water, 3(3):885-891. doi.org/10.1021/acsestwater.2c00652.
Kostarelos, K., P. Sharma, E. Christie, T. Wanzek, and J. Field. 2021. Viscous Microemulsions of Aqueous Film Forming Foam (AFFF) and Jet Fuel a Inhibit Infiltration and Subsurface Transport. Environmental Science and Technology Letters, 8(2):142-147. doi.org/10.1021/acs.estlett.0c00868.
Schaefer, C.E., V. Culina, D. Nguyen, and J. Field. 2019. Uptake of Poly- and Perfluoroalkyl Substances at the Air-Water Interface. Environmental Science and Technology, 53(21):12442-12448. doi.org/10.1021/acs.est.9b04008.
Schaefer, C.E., D. Nguyen, and J.A. Field. 2020. Response to Comment on Uptake of Poly- And Perfluoroalkyl Substances at the Air-Water Interface. Environmental Science and Technology, 54(11):7021-7022. doi.org/10.1021/acs.est.0c02488.
Stults, J., T. Illangasakare, and C.P. Higgins. 2021. The Mass Transfer Index (MTI): A Semi-Empirical Approach for Quantifying Transport of Solutes in Variably Saturated Porous Media. Journal of Contaminant Hydrology, 242:103842. doi.org/10.1016/j.jconhyd.2021.103842.
Stults, J.F., Y.J. Choi, C. Rockwell, C.E. Schaefer, D. Nguyen, D. Knappe, T. Illangasekare, and C.P. Higgins. 2023. Predicting Concentration and Electrolyte Dependent Air-Water Interfacial Partitioning Parameters of PFASs Using Quantitative Structural Property Relationships (QSPRs). Environmental Science and Technology, 57(13):5203-5215. doi.org/10.1021/acs.est.2c07316.
Stults, J.F., Y.J. Choi, C.E. Schaefer, T. Illangasekare, and C.P. Higgins. 2022. Estimation of Transport Parameters of Per- and Polyfluorinated Alkyl Substances (PFASs) in Unsaturated Porous Media: Critical Experimental and Modelling Improvements, Environmental Science and Technology, 56(12):7963–7975. doi.org/10.1021/acs.est.2c00819.
Wanzek, T., J.F. Stults, M.G. Johnson, J.A. Field, and M. Kleber. 2023. Role of Mineral-Organic Interactions in PFAS Retention by AFFF Impacted Soil. Environmental Science and Technology, 57(13) 5231-5242. doi.org/10.1021/acs.est.2c08806.