| Résumé | A mathematical approach to the accurate correction of the blank when applying isotope dilution (ID) and reverse ID is presented. The manner in which blank correction is undertaken is critical to the quality of the final results. Direct subtraction of a procedural blank from the gross analyte concentration is only valid when the blank contributes to the primary ID process and not to the reverse ID process. When the blank contributes to both processes, typically only a fraction of this blank concentration should be subtracted. The approach developed here was illustrated and validated by the determination of MeHg in tuna fish using ID and reverse ID SPME GC-ICP-MS and an enriched Me198Hg spike. Despite a 150-fold higher blank (equivalent to 7% of the analyte concentration in the sample) arising from use of a 1 M NaOAc/HOAc buffer solution compared to that obtained with use of a 0.5 M NH4OAc/HOAc buffer, final concentrations of 19.90 ± 0.34 and 19.88 ± 0.10 μmol kg−1 (one standard deviation, n = 3) respectively, were derived. These data are in good agreement with the assigned value of 19.91 ± 0.82 μmol kg−1 (as 95% confidence interval, derived from an international intercomparison exercise). The methodology was also applied to the determination of a 50-fold lower concentration of MeHg in a salmon fish. A method detection limit (3SD) of 0.9 nmol kg−1 based on processing of a 0.40 g subsample was obtained. |
|---|
