RESUMEN
The methods that can accurately measure the concentrations of nonextractable residues (NERs) of hydrophobic organic contaminants (HOCs) in soil are still lacked in current studies. In this study, three methods, namely methanolic saponification treatment (MST), silylation treatment (ST), and acid deashing treatment (ADT), were investigated and then combined to extract the NERs of six types of polycyclic aromatic hydrocarbons (PAHs) from nine soil samples. The NER concentrations of PAHs obtained by ST (2.43-521.73 ng g-1) were comparable to or significantly higher than those obtained by MST (1.94-291.54 ng g-1), owing to the properties of soil and target compounds. Additionally, ADT could further release a considerable amount of PAH NERs (0.39-276.99 ng g-1) from the soils that had been treated with ST. The mechanism was that acid solution dissolved mineral components, significantly increasing the pore size of the soil matrices from 9.37-15.57 nm to 17.11-27.51 nm. The average percentage of each PAH obtained by ADT (the ratio of the amount obtained by ADT to the total NER content) exhibited a negative correlation with their ring numbers (R2 = 0.62, p < 0.05), whereas the percentage of targets recovered through ST increased linearly with their log KOW values (R2 = 0.75, p < 0.05). Moreover, there is a positive correlation (R2 = 0.73, p < 0.05) between the NER percentages of phenanthrene (obtained by ST-ADT) and the specific surface areas of soils, and the NER percentages of benzo(g,h,i)perylene is positively correlated to the content of total organic carbon (R2 = 0.62, p < 0.05). These results suggested that the amounts and locations of NERs were influenced by both the physicochemical characteristics of PAHs and soils. These findings provide some basic understandings of the entrapped mechanisms of PAH NERs, helping to establish strategies for improving their detection accuracy.