A refined empirical model of the sigma-inductive effect

Abstract

The classic theory of the sigma-inductive effect fails in the case of more complex polycyclic molecules since it is not possible to define unambiguously the "number of paths" to be introduced into Eq. (2a). Therefore, three more sophisticated procedures were suggested and tested: 1) a model based on the charge distribution according to Smith-Eyring, or, 2) according to Del Re, and 3) a simple generalization of the classic description - Eq. (7). All the models meet some basic requirements, viz. they are compatible with the Taft equation and predict transmission of the inductive effect in either direction with the same intensity. If tested on experimental dissociation constants of carboxylic acids in mixed aqueous solvents, the third model appears to be the most satisfactory. The fit is actually better than expected for a purely topological approach, nevertheless it is impaired by solvent effects, in particular when polycyclic and open-chain acids are compared. In addition, the precision would be ultimately restricted if topologically equivalent but sterically different molecules were confronted. It follows that previous reasoning in favour of the through-space transmission of substituent effects is invalidated by our improved model. On the other hand, our results do not allow to claim the through-bonds transmission as the only process operating.