3D QSPR models for the removal of trace organic contaminants by ozone and free chlorine.

Endocrine-disrupting compounds (EDCs) and pharmaceuticals and personal care products (PPCPs) have been detected at low levels in water resources around the world and one impact of their detection is the continuous concern on their fate and removal by various water treatment processes. In this research, a 3D quantitative structure-property relationship (QSPR) model characterized by the utilization of 3D molecular structures is explored as a potential tool to prescreen these compounds and help focus research on more persistent compounds during typical water treatment processes. Monte Carlo (MC) statistical mechanics simulations were utilized to generate 3D molecular descriptors and physicochemical properties for the development of multiple linear regression analysis. The relevance of each parameter to removals of target compounds by ozone (O3) and free chlorine was determined based on data matrices generated in bench- and pilot-scale experiments. Calculated removals were correlated with experimental data with linear regression coefficients of 0.84 for ozonation and 0.71 for chlorination. The increased predictability of ozone removal reflects the fundamental simplicity of ozone reaction mechanisms, which is dominated by oxidation reactions. Interestingly, the weakly polar surface area, in addition to the pi surface area of these molecules, seems critical to ozone removal. The removal of these compounds by free chlorine is related to their ozone removal, ionization potential and three other parameters. The developed QSPR models help disclose the removal mechanism during ozonation and chlorination.