Abstract
Extractable organic fluorine (EOF) analysis is used to assess fluorinated organic content in, e.g., water and solid sample extracts. EOF content is often determined by combustion ion chromatography (CIC). However, contamination from inorganic fluoride (F(-)) in extracts risks overestimating actual EOF. Despite this known interference, a standardized and optimized methodology to mitigate such contamination remains lacking. In this study, we evaluated the presence of F(-) in EOF extracts and designed and validated a method for F(-) determination in small aliquots (∼250 μL) of complex aqueous and solvent matrices. The method is based on derivatization of F(-) to triphenylfluorosilane (TPSiF) and subsequent analysis with gas-chromatographic tandem mass spectrometry analysis (GC-MS/MS). Calibration was linear from 1 to 1000 μg L(-1) (R(2) > 0.99), and the limit of detection and the limit of quantification were 0.98 and 4.5 μg F(-) L(-1), respectively. The relative standard deviation was below 7% across the calibrated range. A certified reference material (CRM River water) and F(-)-spiked minced meat were used to assess F(-) recovery to 103 ± 3% (p = 0.10) and 95 ± 6% (p = 0.31) of certified/spiked concentrations. Analysis of sample extracts showed that (1) solid-phase extraction extracts of river water, (2) solvent extracts from aqueous-film-forming foam (AFFF)-impacted soil, and (3) extracts of a fish sample contained 29-49, 61 ± 4.7, and 17-90 μg F(-) L(-1), respectively. F(-) accounted for 3-59% of the measured EOF, indicating a substantial overestimation. Therefore, the analysis and discount of F(-) in EOF extracts are essential for accurate EOF determination by CIC.