Modelling the sorption behaviour of perfluoroalkyl carboxylates and perfluoroalkane sulfonates in soils

Abstract

A simple parametric model was developed to predict the sorption of perfluoroalkyl substances (PFASs) in soils. Initially, sorption and desorption solid-liquid distribution coefficients (Kd and Kd,des respectively) of eight PFASs (five perfluoroalkyl carboxylates, PFCAs, and three perfluoroalkane sulfonates, PFSAs) in seven soils with organic carbon (OC) content ranging from 1.6 to 41% were quantified using batch experiments. The information obtained helped to fill the gaps in a literature-based database of Kd values of PFASs, which was lacking data on soils with high OC content. The overall dataset finally comprised 435 entries. Normalized sorption coefficients for the soil OC and mineral fraction contents (KOC and KMIN respectively) were deduced for each PFAS by correlating the corresponding Kd values obtained under a wide range of experimental conditions with the fraction of organic carbon (fOC) of the soils. Furthermore, the sorption mechanisms in each phase were shown to depend mainly on PFAS chain length. The dependence of KOC and KMIN values on PFAS chain length defined the basic equations to construct the model for predicting PFAS sorption, applicable to both PFCAs and PFSAs with chain lengths ranging from 3 to 11 fluorinated carbons. The validation of the proposed model confirmed its ability to predict the Kd of PFASs based only on the soil OC and silt+clay contents and PFAS chain length. Therefore, it can be used in the first stages of a risk assessment process aiming at estimating the potential mobility of PFASs in soils after a contamination event.