Per- and polyfluoroalkyl substances (PFAS) are a diverse class of synthetic chemicals that have garnered significant attention due to their persistence, widespread occurrence, and adverse health effects. Thousands of PFAS are registered globally, occupying a wide chemical space and requiring diverse methods for their identification and quantification. Despite vast improvements in analytical coverage over the last two decades, there are increasing concerns that unknown or emerging compounds continue to be overlooked. To address these concerns, a number of new analytical strategies have emerged: one is the so-called fluorine mass balance (FMB) approach, which involves subtracting the fluorine attributed to target PFAS (∑PFAS) from the total- or extractable organic- fluorine (TF and EOF, respectively) to deduce the quantity of unknown PFAS in a sample. This approach can be used to prioritize samples with high levels of unidentified fluorine for further interrogation. A second approach involves high resolution mass spectrometry (HRMS)-based suspect and non-target screening, which aims to identify novel PFAS in environmental samples. This thesis develops and/or applies these emerging analytical methods in order to improve our understanding of PFAS sources and occurrence in the environment, with FMB experiments applied throughout Papers I to III, and suspect- and nontarget screening used in Papers II-IV.
In Paper I, an FMB of different components of artificial turf (backing, filling, and blades) revealed high levels of total fluorine in all samples (ranges of 16−313, 12−310, and 24−661 μg of F/g in backing, filling, and blades, respectively), while EOF and target PFAS occurred in <42% of all samples (<200 and <1 ng of F/g, respectively). Further experiments confirmed the absence of both fluoride and perfluoroalkyl acid precursors in these samples. Collectively, these results point toward the occurrence of a polymeric organofluorine, consistent with patent literature, and shines a light on the use of fluoropolymers in plastic and rubber production which might complicate disposal of these products.
In Paper II, both FMB and HRMS-based suspect screening were applied to liver samples from a variety of marine mammals. As part of this work, an ionization efficiency-based model for quantification of substances lacking analytical standards was trained and validated. Thereafter, the model was used to quantify PFAS detected by suspect screening, and ultimately reduced the quantity of unidentified organofluorine from 13-70% (median: 32%) down to 0-27% (median: 17%).
Paper III delved further into FMB and non-target analysis of marine mammals, this time focusing on blubber, where unexpectedly high levels of unknown EOF were previously uncovered in the blubber of a Greenlandic killer whale. Using a combination of ion exchange solid phase extraction, gas chromatography-ion mobility-high resolution mass spectrometry (GC-IM-HRMS), and collision cross section (CCS)-based prioritization (i.e. CCS [Å2] < 0.2 Å2 × m/z + 100 Å2) the number of plausible organofluorine peaks was reduced from several thousand down to several hundred. Structures were proposed for the most abundant on this list based on fragmentation. Five novel fluorotelomer sulfones were identified at confidence level 1 (CL 1: identified with standard) and quantified, accounting for up to 75% of the EOF in blubber.
Finally, in Paper IV methanol extracts of municipal wastewater treatment plant sludge, as well as sludge and dust standard reference materials (SRMs), were characterized by liquid chromatography-IM-HRMS, and the same CCS filter used in Paper III was applied, together with two additional PFAS prioritisation strategies (mass defect and mass/number of carbon atoms). A total of
Fluorine mass balance, suspect, and non-target screening are critical tools for expanding our understanding of PFAS contamination in diverse environmental and biological matrices. Integrating these advancements is essential for more comprehensive exposure assessments and informed policy decisions.