Towards reliable semi-quantification of PFAS in aquatic biota: A similarity-based approach and practical...

Numerous per- and polyfluoroalkyl substances (PFAS) are persistent, bioaccumulative, toxic and highly mobile in aquatic environments, making aquatic biota a relevant matrix to study PFAS contamination. Currently, hundreds of PFAS have been revealed by non-target screening. Quantifying exposure levels to these compounds is crucial for assessing risks and advising regulatory measures. For this, a comprehensive extraction protocol and accurate concentration estimation is required. In this study, the exhaustiveness of a sample preparation protocol was evaluated on a reference mussel matrix spiked with standards and aqueous film-forming foams (AFFF), covering 143 compounds. In addition, a similarity-based semi-quantification method was developed for PFAS in marine biota. The approach relies on the use of the most similar available standard, selected based on chemical fingerprints (ECFP6, MACCS) and fluorine content. Sample preparation proved effective for most compounds, and normalization against internal standards corrected losses and matrix effects. This method was applied to a spiked matrix, a field-collected mussel sample, and a fish CRM, with mean and median fold errors below 1.73. Despite matrix complexity, low concentrations could still be determined (e.g., 23 pg/g dw of FHpSA in mussel, fold error 1.13). The behavior of PFAS in ESI was evaluated and the in-source fragmentation of 6:2 FTSAm was described for the first time. Because it relies on systematic and reproducible calculations applicable to any PFAS, the proposed method can be readily implemented and, combined with the recommendations provided, aims to improve the reliability of PFAS semi-quantification in non-targeted analysis of biological samples.