Optimization of the procedure for the determination of polycyclic aromatic hydrocarbons and their derivatives in fish tissue: Estimation of measurements uncertainty.

Three alternative procedures were employed for the isolation of polycyclic aromatic hydrocarbons (PAHs; 15 of 16 US EPA priority pollutants and benzo[e]pyrene), their methyl-derivatives and sulphur analogues from fish tissue: (1) Soxhlet extraction, (2) batch extraction enhanced by sonication, and (3) saponification of the sample followed by re-extraction of analytes into hexane. Soxhlet extraction using hexane-acetone (1:1, v/v) was the most efficient extraction technique, with analyte recoveries in the range 70-108%. Within optimization of the clean-up step, several types of gel permeation chromatography (GPC) systems were tested: two types of polystyrene divinylbenzene copolymer gels (PSDVB), both 'soft' gel type (Bio-Beads S-X3) and 'rigid' gels type (PL gel and Envirogel) in combination with various mobile phases were compared. Bio-Beads S-X3 and mobile phase chloroform were the most appropriate for purifying of crude extracts before the final determinative step. High-performance liquid chromatography with fluorimetric detection (HPLC/FLD) was used for identification and quantification of PAHs in purified fish extracts. The uncertainties of PAHs measurements were estimated by employing two alternative approaches. Both provided similar results: the expanded uncertainties obtained for individual PAHs by the 'top-down' approach were in the range 9-53%, their values resulting from application of the 'bottom-up' approach were in the range 16-52%.