Per- and Polyfluoroalkyl Substances (PFAS) in Surface Sediments: Occurrence, Patterns, Non-Physical Drivers and Contribution of Unidentified Precursors in French Aquatic Environments

International audienceSediment acts as a sink for per- and polyfluoroalkyl substances (PFAS), especially for perfluoroalkyl sulfonic acids (PFSAs) and perfluoroalkyl carboxylic acids (PFCAs). Recently, non-targeted methods have shown the presence of numerous unknown PFAS in environmental samples. The contribution of these PFAS can be estimated by the Total Oxidisable Precursor assay (TOP), an oxidative method used to convert polyfluoroalkyl precursors (pre-PFAAs) into perfluoroalkyl acids (PFAAs). While sediment properties (grain size and organic carbon fraction, foc) were reported as controlling factors for some PFAS, the influence of non-physico-chemical drivers (e.g. population density) still needs to be investigated. Thus, this work focused on sediments from France and used optimized analytical methods and geographical information system (GIS) to fill these knowledge gaps. Extraction solvent was modified from methanol (MeOH) + ammonium hydroxide to MeOH + ammonium acetate for the analysis of 16 PFAAs, 6 alternative PFAAs and 14 pre-PFAAs covering anionic, neutral and zwitterionic species (n:2 fluorotelomer sulfonamidoalkyl betaines, n:2 FTABs). The optimized method exhibited suitable performances for its application on sediments (n=50) collected in 2018. The TOP assay was adapted to this specific method extraction. Perfluorododecanic acid,perfluorooctane sulfonic acid (PFOS) as well as perfluorooctanoic acid (PFOA) were often detected (76, 74 and 68 % respectively). 6:2 FTAB, 8:2 FTAB and perfluoropentyl propanoic acid (FPePA) were also detected (38, 26 and 36 %, respectively). Median ∑PFAS was 1.3 ng.g-1 dry weight (dw) (range < LOD–23 ng.g-1 dw). 6:2 FTAB and 8:2 FTAB levels ranged from < 0.36 to 6.4 ng.g-1 dw and < 0.50 to 7.4 ng.g-1 dw, which is in the same order of magnitude than L-PFOS (< 0.08–5.1 ng.g-1 dw) advocating for the necessity to include these PFAS in targeted analysis. At nationwide scale, PFOS and ∑PFCAs accounted for 37 ± 25% and 31 ± 21% of ∑PFAS, respectively, while targeted pre-PFAAs accounted for 29 ± 27%, thereby supporting the importance of the latter in sediment and the potential contribution of untargeted pre-PFAAs monitored by the TOP assay (data under acquisition). Correlations between PFAS levels normalized by foc, unidentified pre-PFAAs and non-physico-chemical drivers with GIS are being investigated. Preliminary results showed that PFAS levels were not correlated with sample site position in the hydrographic system (proxy: Strahler number)