Optimisation of supercritical fluid extraction of polynuclear aromatic hydrocarbons from spiked soil and marine sediment standard reference material

 For the determination of 16 PAHs in soils and sediment samples by GC/FID and GC/MS, the dynamic off-line supercritical fluid extraction with both pure and modified carbon dioxide has been evaluated. The optimisation of extraction parameters was performed for four individual groups of PAHs according to their number of aromatic rings (2–3 rings, 4 rings, 5 rings and 6 rings) by varying pressure (200–510 bar), temperature (50–150 °C), extraction fluid volume (10–50 ml), and the methanol modifier concentration (0–10%). Using a five level spherical factorial experimental design the number of experiments required for optimisation was 45. In spiked soil samples extraction efficiencies of 80–100% were achieved for the individual groups of PAHs. At the optimal set of conditions 10–30% lower recoveries of PAHs were obtained for the standard reference material NIST SRM 1941a (marine sediment). The largest differences between extraction recoveries of native and spiked PAHs occurred at high molecular weight PAHs. Using SFE efficiency data for the standard reference material, cluster analysis proved that dividing the 16 PAHs into four groups according to their number of aromatic rings was appropriate and correct. Received: 2 February 1996/Revised: 26 November 1996/Accepted: 30 November 1996