Towards the use of experimental electron densities to estimate reliable lattice energies

We examine the reliability of lattice energies (and by inference, intermolecular interaction energies) estimated from experimental electron densities. Based on the modest number of results published to date, lattice energies obtained in this fashion typically overestimate more reliable values, sometimes by hundreds of kJ mol(-1). The causes of this behaviour are explored in detail, and include a misunderstanding of the contribution of the polarization energy, the failure to recognise the important link to a sublimation process, and the use of inappropriate atom-atom potentials to estimate dispersion and exchange-repulsion energies. It is hoped that this contribution will encourage further research of this nature, and a more critical evaluation of energies derived from high-quality diffraction experiments on molecular crystals.