Reliable QSAR for estimating Koc for persistent organic pollutants

Abstract

Several recent studies have shown that n-octanol/water partition coefficients may not be a good predictor for estimating soil sorption coefficients of persistent organic pollutants (POPs), defined here as chemicals with log K_ow  greater than 5. Thus, an alternative QSAR model was developed that seems to provide reliable estimates for the soil sorption coefficients of persistent organic pollutants (POPs). This model is based on a set of calculated topological molecular descriptors viz. W, J, ¹c, ²c, Jhetz, Jhetm, Jhete,  Jhetv,  and  Jhetp  and evaluated soil sorption data for 32 POPs. The regression analysis gave a penta-parametric correlation involving W, J, Jhetm, Jhetv and ¹c as the most significant QSAR model for modeling of POPs. The developed QSAR model was rationalized in terms of potential hydrophobic interactions between persistent organic pollutants and soil organic matrix.