PFAS are a large and varied class of more than 3000 compounds and currently it is difficult or even impossible to quantify more than a few dozen of the many known PFAS due to unavailability of authentic standards.
Currently, there are defined, or “targeted” analytical methods for approximately two to three dozen PFAS compounds in specific environmental matrices such as water, soil, sludges and sediment. Targeted analysis of PFAS, where the identities of the compounds to be determined are known, is typically done using HPLC with triple quadrupole molecular mass spectrometry (HPLC-MS/MS), which is the basis of USEPA Methods 537.1 and 533. These methods are validated for approximately 30 PFAS and only apply, as written, to finished drinking water. The method typically underperforms when applied to more complex matrices due to the presence of high levels of interfering species that can suppress the response of the analytes. It is also necessary to have authentic analytical standards of each compound to be quantified, in both native and (ideally) stable-labeled isotope forms.
In addition, there are thousands of unknown PFAS chemicals, including degradants and metabolites, most for which we know little or nothing about the risks they might pose to health or the environment. A non-targeted approach for screening for known and unknown PFAS would seem to be the best approach for source identification.
The current tools for non-targeted detection of unknown PFAS in environmental samples include high performance liquid chromatography with high resolution mass spectrometry (HPLC-HRMS). HPLC-HRMS can obtain high resolution mass spectra of separated sample components and provides accurate mass empirical formulae and useful fragmentation patterns that can be interpreted to postulate a structure. This technique works best when the PFAS are cleanly separated from other sample components by the HPLC. Interfering species at high concentrations that are commonly found in complex environmental samples such as soil and sediment extracts and surface water can complicate spectral interpretation or even completely suppress signals from PFAS at low concentrations.
A “hybrid” non-targeted/targeted approach is the Total Oxidizable Precursor (TOP) method. In this approach, PFAS are treated with hydroxyl radicals generated by heating persulfate. PFAS sulfonamides and telomers are converted to terminal Perflurocarboxylic acids (PFCA) which can then be quantitated using HPLC-MS/MS with authentic standards. This approach was used to study the fate of AFFF derived PFAS in a wastewater treatment plant. The TOP approach is useful for capturing some PFAS precursors that may be missed in a strictly targeted approach. However, the oxidation of the precursors results in the loss of valuable information since the identity of the original PFAS is lost. This makes identifying the source of the contamination very difficult, as the precursors can be unique to a given source.
We are researching a more facile Total Organo Fluorine method using a Inductively-coupled Plasma ionization coupled to a tandem mass spectrometer with an HPLC inlet (HPLC-ICP-MS/MS) which has been shown by Jamari et. al for screening PFAS in surface waters. Quantification limits were on the order of 1-3 µg/L (ppb). Quantification is only organofluorine content, as opposed to molecular structure, since all the very strong C-F bonds are broken to form F- and subsequently BaF+ ions. Sample components do not appear to affect signal response. Therefore, authentic standards of each compound are not necessary; a surrogate standard with approximately the same fluorine content and eluting reasonably close to the analytes of interest is sufficient. Stable isotope standards are also not necessary, as matrix effects are eliminated by the action of the plasma torch.