A new route to apply the charge distribution (CHARDI) method to structures based on heteroligand coordination polyhedra is presented. The previous algorithm used scale factors computed in an iterative way based on the assumption (which turned out to be not always correct) that a real over-under bonding effect affects mainly the anionic charges of each single anion, without grossly modifying the total charge of each type of anion. The new, more general approach is not based on any a priori assumption but treats separately the homoligand sub-polyhedra and attributes to each type of atom a fraction of the charge of the atom coordinated to it, computed in a self-consistent iterative way. The distinction between the bonding and non-bonding contact is also redefined in terms of the mean fictive ionic radii (MEFIR), without the need of an empirical parameter, used in the previous algorithm. CHARDI equations are generalized in terms of the new approach and a series of examples is presented.
Keywords: bond strength; charge distribution; effective coordination number; heteroligand polyhedra.