Modelling the depuration rates of polychlorinated biphenyls in Oncorhynchus mykiss with quantum chemical descriptors

Abstract

Using quantum chemical descriptors and partial least squares regression, a quantitative structure–activity relationship (QSAR) model is developed for the depuration rate constants (log k d) of 62 polychlorinated biphenyls (PCBs) in juvenile rainbow trout (Oncorhynchus mykiss). The values of the cross-validated regression coefficient () and standard deviation (SD) are 0.655 and 0.05, respectively. The high cross-validated coefficient and low standard deviation indicate that the QSAR model is well predictive. In the QSAR model, the following six descriptors are highly significant: (the most positive charge of a hydrogen atom), HOF (standard heat of formation), CCR (core–core repulsion), EE (electronic energy), α2 (squared average molecular polarisability), and S (molecular surface area). The significant descriptors show that the depuration of PCBs in rainbow trout may be mainly attributed to the biota–water partitioning process, and the reactive activity of PCB molecules may play a subordinate role.