Modelling the morphology of molecular crystals
Computer programs have been written which allow morphological calculations based on a knowledge of internal crystal structure to he carried out. Details of the programs are presented along with the guidelines developed for their use. The programs were used to compare and contrast the current methods employed for relating crystal shape to structure and to confront specific problems in that field. Calculations on a range of compounds show that the morphologies derived from the simple Donnay-Harker (DHI) model give almost as good a fit to the observed form as the more sophisticated attachment energy (AE) calculations except when strong bonding directions were present. In the first study of its type all the methods currently favoured in the literature including the Ising and PBC approaches as well as DH and AE models were applied to benzophenone. All the models gave the same theoretical morphology. One problem remaining in the field of relating crystal structure and morphology is that of polar morphology. None of the current methods can account for a polar morphology. Surface, bulk, isolated molecule charge distributions were used in a modification of the classical attachment energy model to account for urea exhibiting a polar morphology when growth from the vapour phase. For the disruptive type of tailor made additives an improvement in the current methodology is proposed with the calculation of an additional parameter. This additional parameter accounts for the morphology with an additive present and gives good agreement with the test case of benzamide crystals grown with benzoic acid as an additive. The additive approach also allowed the effects of toluene solvent on the crystal habit of benzophenone to be considered By treating toluene as a tailor made additive it was possible from calculations to identify the likely sites of toluene incorporation and the subsequent effect on crystal growth. The results from the calculation were consistent with experiment.