Abstract
The classical rules for Cope rearrangements predict a transition state with chair form to be favored over the boat form. On the other hand, bridged homotropylidenes, which allow only a boatβform transition state by steric reasons, have extremely low barriers. A controversy about the correct pathway and the different possible intermediates and transition states of the reaction has gone on for years. In this work, the hypersurfaces of barbaralane, in comparison with the boatβ and chairβform of hexaβ1,5βdiene, are computed by the ab inito CASSCF (6,6)/6β31G** method starting with UMP2/6β31G** natural orbitals (NO's). All three hypersurfaces show characteristic features, and, moreover differ from each other. A hitherto undiscussed intermediate, bicyclo[2.2.0]hexane, was localized on the boatβhexaβ1,5βdiene pathway. So it is noteworthy that our transition state for the boatβhexaβ1,5βdiene does not correspond to the transition states found by other authors for this conformation. The computed enthalpies of activation of boatβ and chairβhexaβ1,5βdiene, and barbaralane are in good agreement with the experimental data.