The ab initio prediction of yet unknown molecular crystal structures by solving the crystal packing problem

Summary

Recent progress in solving the crystal packing problem from molecular information only for organic molecules containing heteroatoms leads to good agreement between ‘predicted’ and experimentally determined crystal structures. Until now these complex computations have not been used for a ‘true’ prediction prior to experiment. In this contribution experimentally unknown crystal structures of heteroatom-containing molecules are predicted for the first time at the level of atomic coordinates. The examples are easily verifiable; the first prediction deals with a novel racemic modification of cyclo-di-alanyl, of which only the enantiomerically pure cyclo-l-alanyl-l-alanyl crystal structure is known, and the second prediction with a new crystal modification of cyclo-bis(dehydro-alanyl).