Some issues relating to the use of perfluorooctanesulfonate (PFOS) samples as reference standards

Samples of potassium perfluorooctanesulfonate (PFOSK) from three suppliers were analyzed by LC-ESI-MS/MS for purity and by LC-ESI-MS for the percentage of linear isomer present. Our data indicated that the purity ranged from 80% to 98% and the percentages of linear isomer from 67% to 79%. The proportion of branched isomers present in the samples was also estimated using (19)F NMR. These results agreed quite closely with those found by LC-ESI-MS indicating that there is essentially no difference in overall SIM response factor for the branched isomers vs. that of the linear isomer. Several further observations relevant to the use of standards when analyzing for PFOS were encountered during this study. It appears unlikely that matrix effects attributable to the cation (sodium or potassium) present in PFOSNa or PFOSK internal standards is an issue. In seeking potential matrix effects, it was found that the chromatography was improved substantially when the standard was injected as a solution in 80:20 methanol/water rather than 100% methanol. Notably, in concert with the improvement in chromatography, an increase of about 10% in response was observed. In some closely related studies, when (18)O(2) mass-labeled perfluorohexanesulfonate was used as an internal standard, the actual and theoretical concentration ratios matched closely those for related native sulfonates as long as they did not co-elute. However, when they did co-elute, the peak intensities of the native species were enhanced by about 5%, while those of the labeled compound were suppressed by a similar amount. If this effect were not taken into account, the concentration of the native would be inflated by 10%.