Most known per- and polyfluoroalkyl substances (PFAS) bioaccumulate by binding to proteins or partitioning to phospholipids, leading to their prevalence in liver and blood. However, the recent discovery of high concentrations of unidentified extractable organofluorine (EOF) in the blubber of a killer whale (Orcinus orca) from Greenland suggests that some fluorinated substances preferentially bioaccumulate in storage lipids. To further investigate this, the present work examined blubber from 4 killer whales (3 from Greenland, 1 from Sweden) via gas chromatography-atmospheric pressure chemical ionization-ion mobility mass spectrometry. Using collision cross sections, we prioritized features suspected to be highly fluorinated and then selected 5 for manual annotation. Custom synthesized standards confirmed 10:2 and 12:2 fluorotelomer methylsulfone, 10:2 and 12:2 fluorotelomer chloromethylsulfone, and 6:2 bisfluorotelomer sulfone in all blubber samples from Greenland at concentrations ranging from <0.4 to 72.5 ng/g, explaining 34–75% of blubber EOF, but none in the Swedish sample. None of these substances were observable in liver, suggesting preferential accumulation in storage lipids. To the best of our knowledge, this is the first report of neutral fluorotelomer sulfones in wildlife and the first identification of lipophilic, highly fluorinated PFAS.

Suspect screening strives to rapidly monitor a large number of substances in a sample using mass spectral libraries. For hydrophobic organic contaminants (HOCs), these libraries are traditionally based on electron ionization mass spectra. However, with the growing use of state-of-the-art mass spectrometers, which often use alternative ionization approaches and separation techniques, new suspect screening workflows and libraries are urgently needed. This study established a new suspect screening library for 1,590 HOCs, including exact mass and a combination of measured and model-predicted values for retention time (RT) and collision cross section (CCS). The accuracy of in silico predictions was assessed using standards for 102 HOCs. Thereafter, using gas chromatography-atmospheric pressure chemical ionization-ion mobility-mass spectrometry, a suspect screening workflow constrained by the full scan mass spectrum of (quasi-)molecular ions (including isotope patterns), RT, CCS, and fragmentation mass spectra, together with a continuous scoring system, was established to reduce false positives and improve identification confidence. Application of the method to fortified and standard reference sediment samples demonstrated true positive rates of 79% and 64%, respectively, with all false positives attributed to suspect isomers. This study offers a new workflow for improved suspect screening of HOCs using multidimensional information and highlights the need to enrich mass spectral databases and extend the applicable chemical space of current in silico tools to hydrophobic substances.