THEORETICAL CRYSTAL STRUCTURE PREDICTION

Cornell University Ithaca, Ithaca NY

Investigators

Linked publications & trials

Abstract

Theoretical crystal structure prediction is still an unsolved and a very challenging problem in theoretical chemistry. A solution to the problem is very important for the development of intermolecular potentials, and therefore, for efficient computer-aided drug design. Two methods of global minimization, the diffusion equation method (DEM) and the distance scaling method (DSM), were applied to predict the crystal structures of the hexasulfur and benzene molecules. No knowledge about the systems other than the geometry of the molecules and the pairwise potentials was assumed; i.e., no assumptions were made about the space groups, cell dimensions, or the number of molecules in the unit cell. The crystal structures, known from experiment, were predicted correctly. To verify the power of the method, the problem of global minimization of the potential energy of crystals of both molecules was intentionally increased considerably in complexity: viz., the numbers of molecules in the unit cell were doubled (from 3 to 6 in the case of hexasulfur, and from 4 to 8 in the case of benzene), and the search for the global minimum was repeated; the method again located the global minimum for each molecule. These results demonstrate that the method is extremely powerful. Future work will be focused on crystal structure prediction of flexible molecules, using a new, much more powerful, Self-Consistent Basin-to-Deformed-Basin Mapping Method (SCBDBM). This new DEM/DSM-based global optimization method has been designed to significantly enhance the efficiency of deformational methods.