Computationally assisted structure determination for molecular materials from X-ray powder diffraction data

For organic molecular materials, the key step in solving crystal structures from high-resolution powder diffraction data is the generation of reliable trial structures for final refinement. In this paper we demonstrate an efficient new methodology for generating trial structures that predicts the correct molecular packing in a crystal lattice given the unit cell dimensions and space group. The method uses a systematic search of intermolecular, atom-atom interactions to examine and rank all possible packing arrangements. The approach is demonstrated for a number of systems with increasing degrees of search complexity including indigo, 6,13-dichlorotriphendioxazine, phenanthrene, paracetamol, and benzophenone. The final example is of particular importance as it illustrates the first application of this type of methodology to a conformationally flexible system.