There is a need to evaluate the presence of per- and polyfluoroalkyl substances (PFAS) without having to use expensive and time-consuming methods such as liquid chromatography-mass spectrometry (LC-MS) on every sample. To tackle this issue, a fast, simple, and low-cost method for determining total PFAS-based fluorine content was optimized using high-resolution continuum source graphite furnace molecular absorption spectrometry (HR-CS-GFMAS). HR-CS-GFMAS exploits the high fluorine sensitivity and selectivity of gallium fluoride (GaF) by complexing combustion-evolved fluorine with Ga in reaction modifier and measuring the complex at a wavelength of 211.248 nm. The prevailing total fluorine methods using HR-CS-GFMAS gave unsatisfying fluorine recovery and had a high limit of quantification (LOQ = 5.13 μg/L), required a PFOA reinjection step after each sample, and recovery for individual PFAS was limited. Herein, we made multiple method improvements including 1) introduction of modifiers at various concentrations and volume, 2) background correction curves, 3) optimization of pyrolysis and atomization temperatures, and 4) methodological improvement in GaF peak integration. The optimized method exhibited a LOQ of 1.25 μg/L and 100.7 ± 4.3% reproducibility. The method was applied to a comprehensive list of 45 PFAS in 14 groups and gave 79.1–103.8% fluorine recovery calculated based on PFAS concentrations quantified by LC-MS without the need for a PFOA reinjection step after every sample, cutting operational cost and analytical time in half. Effects of matrix ions including H⁺, Cl^-, Ca²⁺, Mg²⁺, and Al³⁺ were also evaluated to validate the robustness of the method. Moreover, analysis results of the AFFF-contaminated soils and wastewater treatment plant biosolid extracts using the optimized method were compared with fluorine measurement by combustion ion chromatography. Therefore, the optimized method appears promising for rapid and reliable PFAS screening in complex environmental samples, which can serve to guide which samples should be prioritized for resource-intensive LC-MS.

Youn Jeong Choi

Appearances