Elusive contribution of the experimental surface molecular electrostatic potential and promolecule approximation in the empirical estimate of the crystal density.

The aim of this study is to probe the crystal density (D(c)) description in terms of pertinent molecular characteristics and properties. In this purpose, the electrostatic potential was derived from available experimental electron density multipole parameters of molecular compounds with different D(c) magnitudes. The surface electrostatic potential has been analyzed through the positive and negative statistical variances. The surface of the molecule is here corresponding to particular isodensity values according to Bader's topological theory. Following the successful Politzer's method based on quantum mechanics calculations to empirically describe macroscopic properties, the crystal density was regressed on the molecular density and the surface electrostatic potential variance. This latter appears to be a poor statistical descriptor of the crystal density when the experimentally derived electrostatic potential is used and it does not significantly improve the fit of D(c) to molecular density alone. Compared to Politzer's approach based on gas phase isolated molecules, the experimental electrostatic potential is biased by the interactions in the crystal lattice. As an alternative to other sophisticated methods, the promolecule isodensity surface offers a quite useful and straightforward way to define the molecular volumes. The reported description of the crystal density for a set of 50 molecules using the promolecule approach yields satisfactory results.