QSPR modeling of n-octanol/air partition coefficients and liquid vapor pressures of polychlorinated dibenzo-p-dioxins

Abstract

The molecular geometries of 75 polychlorinated dibenzo-p-dioxins (PCDDs) were optimized using B3LYP/6-31G* method. The calculated structural parameters were taken as theoretical descriptors to establish two new novel QSPR models for n-octanol/air partition coefficients (log KOA) and subcooled liquid vapor pressure (log PL) of PCDDs. The R2 values of the two models are 0.983 and 0.985 respectively. Their standard deviations of prediction in modeling (SD) are 0.174 and 0.230 respectively. The results of leave-one-out (LOO) cross-validation for training set show that the two models exhibited optimum stability and good predictive power.