QSAR models to predict effect of ionic strength on sorption of chlorinated benzenes and phenols at sediment-water interface.

It is hypothesised that the experimental sorption coefficient normalised to the organic carbon fraction of sediment (K(oc)exp) for non-ionic, hydrophobic, organic pollutant depends upon the molecular properties as well as background ionic strength of the aquatic system. The utility of this concept has been demonstrated by incorporating ionic strength as a parameter in the three quantitative structure activity relationships (QSARs) namely octanol-water partitioning coefficient model (Kow model), the linear solvation energy model (LSE model), and molecular connectivity indices theory (MCI model). Four chlorinated benzenes and two chlorinated phenols were employed in the present study. Sorption experiments using sediment from the Patalganga River were conducted in laboratory (bottle point method) at different ionic strengths (viz. 0.01, 0.05, and 0.10 M). The K(oc)cat values predicted using Kow model incorporating ionic strength compare reasonably well with the K(oc)exp values (r2 = 0.60 and standard error of estimator i.e. SEE = 0.35). The LSE model incorporating ionic strength too, was found to be equally good (r2 = 0.67, SEE = 0.33). An attempt has also been made to validate the QSARs developed in the present study utilising the sorption parameters experimentally measured by Dewulf et al. (1996) (Water Res. 30, 3130-3138) for sorption of toluene, ethylbenzenes, and xylenes onto the sediments from Belgian Continental Shelf and North Sea, as well as Mader et al. (1997) (Environ. Sci. Technol. 27, 1524-1531) for sorption of di-, tri-, tetra chlorobenzenes on pure mineral oxides namely Al2O3 and Fe2O3.