AM1 calculations of rotation around essential single bonds and preferred conformations in conjugated molecules

Equilibrium twist angles, rotational barriers around essential single bonds, and preferred conformations for over 60 conjugated organic molecules were calculated using the semiempirical AM1 (Austin model 1) method. Comparison with ab initio and experimental data shows that AM1 can be applied quite successfully to conformational problems of this type. Relatively large errors are, however, found for compounds in which lone pair–hydrogen and especially lone pair–lone pair interactions are decisive for their conformational behavior. AM1-calculated rotational barriers in conjugated molecules, however, are found to be much too low. Moreover, AM1 does not seem to even correctly reproduce the trend of rotational barriers within a series of structurally related compounds. E/Z energy differences obtained by AM1 are also frequently considerably too low. In contrast to rotational barriers, their trends and thus conformational preferences, however, are quite satisfactorily calculated by this method. Ionization potentials obtained by AM1 are too high by about 0.5 eV. However, trends are quite well predicted.