Estimation of water solubility of polycyclic aromatic hydrocarbons using quantum chemical descriptors and partial least squares

Quantitative Structure-Property Relationship (QSPR) modeling is a powerful approach for predicting the properties of environmental organic pollutants from their structure descriptors. In this study, QSPR models were established for estimating the water solubility of Polycyclic Aromatic Hydrocarbons (PAHs). Quantum chemical descriptors computed with density functional theory at the B3LYP/6-31G(d) level and Partial Least Squares (PLS) analysis with an optimizing procedure were used to generate QSPR models for the logarithm of the water solubility of PAHs. Two optimized models with high correlation coefficients (R-2=0.966 and 0.970) were obtained for estimating logarithmic mass and molar concentration of water solubility, respectively. The internal statistics results of a cross-validation test (Q(cum)(2) =0.928 and 0.937, respectively) showed both the models had high precision and good prediction capability. The logarithmic water solubility values predicted by the models are close to those observed. The PLS analysis indicated that PAHs with larger electronic spatial extent and lower total energy values tend to be less soluble.