Significance of MTD and other descriptors in lipophilicity models for chlorinated aromatic compounds

The lipophilicity of chlorinated benzenes and 20 biphenyls has been modeled using the minimum topological difference method and MTD indices. The MTD descriptor has been optimized on its own or together with nCl (number of chlorine atoms per molecule), 2D- and 3D-descriptors such as van der Waals surface area and volume, solvent accessible surface area and solvent-accessible surface-bounded molecular volume, and the inter-planar angle of the substituted biphenyls. In lipophilicity models of chlorinated aromatic compounds the majority of descriptors, including MTD, contain essentially the information of nCl. MTD is a good lipophilicity descriptor of chlorinated aromatic compounds either alone or in association with other descriptors. For chlorobenzenes it gives r = 0.998, s = 0.069, F = 2281, and for the set of chlorinated biphenyls r = 0.974, s = 0.255, F = 338, a better result than gives the nCl descriptor (r = 0.967). The significance of MTD descriptor in different models is discussed. Beside nCl, the MTD index encodes ortho and para effects.