Application of TLSER method in predicting the aqueous solubility and n-octanol/water partition coefficient of PCBs, PCDDs and PCDFs.

The theoretical linear solvation energy relationship(TLSER) approach was adopted to predict the aqueous solubility and n-octanol/water partition coefficient of three groups of environmentally important chemicals-polychlorinated biphenyls(PCBs), polychlorinated dibenzodioxins and dibenzofurans (PCDDs and PCDFs). For each compound, five quantum parameters were calculated using AM1 semiempirical molecular orbital methods and used as structure descriptors: average molecular polarizability(alpha), energy of the lowest unoccupied molecular orbit( E(LUMO)), energy of the highest occupied molecular orbit( E(HOMO)), the most positive charge on a hydrogen atom (q+), and the most negative atomic partial charge(q-) in the solute molecule. Then standard independent variables in TLSER equation was extracted and two series of quantitative equations between these quantum parameters and aqueous solubility and n-octanol/water partition coefficient were obtained by stepwise multiple linear regression(MLR) method. The developed equations have both quite high accuracy and explicit meanings. And the cross-validation test illustrated the good predictive power and stability of the established models. The results showed that TLSER could be used as a promising approach in the estimation of partition and solubility properties of macromolecular chemicals, such as persistent organic pollutants.