A matrix effect-free method for reliable quantification of perfluoroalkyl carboxylic acids and perfluoroalkane sulfonic acids at low parts per trillion levels in dietary samples

Abstract

In recent exposure modeling studies diet has been identified as the dominant pathway of human exposure to perfluorooctanoic acid (PFOA) and perfluorooctane sulfonic acid (PFOS). However, the paucity of highly sensitive and accurate analytical data to support these studies means that their conclusions are open to question. Here a novel matrix effect-free method is described for ultra-trace analysis of perfluoroalkyl carboxylic acids and perfluoroalkane sulfonic acids in dietary samples of varied composition. The method employs ion pair extraction of the analytes into methyl tert-butyl ether and subsequent solid phase extraction clean-up on Florisil and graphitized carbon. The target compounds are separated and detected using ultra performance liquid chromatography coupled to tandem mass spectrometry. Special care was taken to avoid procedural blank contamination and potential contamination sources were elucidated. The performance of the method was validated for five different food test matrices including a duplicate diet sample. Method detection limits in the low to sub pgg−1 range were obtained for all target analytes, which is 5–100 times more sensitive than previously reported for duplicate diet samples. Total method recoveries were consistently between 50 and 80% for all analytes in all tested food matrices and effects of co-extracted matrix constituents on ionization of the target compounds were found to be negligible. The precision of the method (defined as percentage relative standard deviation) at concentrations close to the respective method limits of quantification was <15% for all analytes. Accurate quantification at ultra-trace levels was demonstrated by laboratory control spike experiments. For the first time the presence of long-chain PFCAs in duplicate diet samples is reported. The method presented here can thus support an improved assessment of human exposure from dietary intake for a range of PFCA and PFSA homologues. Re-analysis of duplicate diet samples, which had been analyzed earlier using another analytical methodology, indicated that dietary intake of PFOA and PFOS may previously have been overestimated.