The extraction of aged polycyclic aromatic hydrocarbon (PAH) residues from a clay soil using sonication and a Soxhlet procedure: a comparative study.

Abstract

A sonication method was compared with Soxhlet extraction for recovering polycyclic aromatic hydrocarbons (PAH) from a clay soil that had been contaminated with tar materials for several decades. Using sonication over an 8 h extraction period, maximum extraction of the 16 US EPA priority PAH was obtained with dichloromethane (DCM)-acetone (1 + 1). The same procedure using hexane-acetone (1 + 1) recovered 86% of that obtained using DCM-acetone (1 + 1). PAH recovery was dependent on time of extraction up to a period of 8 h. The sonication procedure showed that individual PAH are extracted at differing rates depending on the number of fused rings in the molecule. Soxhlet extraction [with DCM-acetone (1 + 1)] over an 8 h period recovered 95% of the PAH removed by the sonication procedure using DCM-acetone (1 + 1), indicating that rigorous sonication can achieve PAH recoveries similar to those obtained by Soxhlet extraction. The lower recovery with the Soxhlet extraction was explained by the observed losses of the volatile PAH components after 1-4 h of extraction. The type of solvent used, the length of time of extraction and extraction method influenced the quantification of PAH in the soil. Therefore, the study has implications for PAH analyses in soils and sediments, and particularly for contaminated site assessments where the data from commercial laboratories are being used. The study emphasizes the importance of establishing (and being consistent in the application of) a vigorous extraction, particularly for commercial laboratories that handle samples of soil in batches (at different times) from a single site investigation or remediation process. The strong binding of PAH to soil, forming aged residues, has significant implications for extraction efficiency. This paper illustrates the problem of the underestimation of PAH using the US EPA method 3550, specifically where a surrogate spike is routinely employed and the efficiency of the extraction procedure for aged residues is unknown. The implications of this study for environmental monitoring, particularly where numerous batches of samples from a single site assessment or remediation program are submitted to commercial laboratories, is that it would be advisable for these laboratories to check their existing method's extraction efficiencies by conducting a time course sonication extraction on their particular soil to determine the optimum extraction time.