The effect of different molecular descriptors on PFAS transport and sorption behaviour in saturated porous media

Per- and polyfluoroalkyl substances (PFAS) are widespread and persistent in groundwater. PFAS pose a significant threat to human health due to their toxicity and bioaccumulation. Despite past progress, the fate and transport of PFAS mixtures in the saturated zone—as well as physicochemical and structural features that control this transport—are not fully understood. This study showcases a new integrated method that links PFAS sorption characteristics in saturated porous media to a wide range of molecular descriptors. The approach consists of: 1) soil column experiments using a suite of PFAS, 2) a detailed porous media profile analysis following experiments combined with mass balance calculations, 3) characterisation of PFAS sorption behaviour from modelling PFAS breakthrough curves, 4) a clustering analysis of PFAS based on 137 molecular descriptors, and 5) regression modelling between selected descriptors and sorption coefficients. The experimental results identified four distinct PFAS transport behaviour classes ranging from conservative transport to nearly complete retention. Clustering analysis highlighted five molecular descriptors controlling this sorption behaviour. Unexpectedly, chain length was not the primary factor driving PFAS transport behaviour, but rather topological features and aliphatic chain length. Regression analysis showed that, for prediction of sorption magnitude, Fluorine atoms emerged as the best from all molecular descriptors. Extended flushing of the columns over several weeks, combined with a soil profiling, demonstrated significant desorption with almost all sorbed PFAS ultimately remobilised. Variations in PFAS concentrations and mixture composition altered transport behaviour, indicating competition for sorption sites. This integrated framework enables systematic investigation of PFAS sorption and can support improved groundwater contamination assessment and water management strategies.