Prediction of molecular dipole moments from bond moments: testing of the method by DFT calculations on isolated molecules

Abstract

The well-known principle of vector addition of bond moments, used as standard approximation when interpreting and predicting the experimental dipole moments, has been now checked on the calculated moments of isolated molecules with the accuracy not attainable by measurements in solution. Chloro derivatives of adamantane 1a–3a and 1,4-derivatives of bicyclo[2.2.2]octane 4–16 were used as model compounds: their dipole moments were calculated at the levels B3LYP/6-311+G(d,p) and B3LYP/AUG-cc-pVTZ//B3LYP/6-311+G(d,p). In the case of two collinear dipoles, the resulting dipole moments were always slightly greater than the additive values but the additive approximation would be sufficient for practical purposes. The small difference can be explained by mutual induction of the two dipoles and can be roughly estimated by simple electrostatic calculation in spite of several uncertain parameters. The same applies to the adamantane derivatives 2a and 3a, in agreement with previous experiments in solution. In the case of the unsymmetrical CH₂Cl groups in bicyclooctane derivatives 13–16, the additive values are always too great since the group moment does not lie exactly in the C–Cl bond. Nevertheless, the principle of vector addition can be used even in this case when an empirically determined group moment is used instead of a bond moment. Summarizing, the vector addition of bond moments is not exact theoretically but can be used rather broadly with a precision comparable to the accuracy of the experimental dipole moments estimated from solution measurements.