Crystal structure prediction of CL-20 polymorphs using a tailor-made polarizable force field

The prediction of crystal structures is of great significance in the community of energetic materials since the crystal packing pattern is closely linked to the energetic and safety features. Here, we developed a tailor-made polarizable force field (TMPFF) calibrated by the benchmarks obtained from ab initio methods and applied it with the USPEX evolutionary algorithm to predict crystal structures from energetic molecular structures. The results showed that the prediction results of three polymorphs of hexaazaisowurtzitane (CL-20) all agree well with experimental measurements, for both lattice parameters and packing patterns. This work provides a new solution to predict polymorphs of CHON-based energetic molecules with high accuracy and good efficiency and shall accelerate the cognition process of new energetic crystals by theoretical approaches prior to experimental synthesis.