CALYPSO structure prediction method and its wide application

Atomistic structure prediction from “scratch” is one of the central issues in physical, chemical, materials and planetary science, and it will inevitably play a critical role in accelerating materials discovery. Along this thrust, CALYPSO structure prediction method by taking advantage of structure smart learning in a swarm was recently developed in Prof. Yanming Ma’s group, and it has been demonstrated through a wide range of applications to be highly efficient on searching ground state or metastable structures of materials with only the given knowledge of chemical composition. The purpose of this paper is to provide an overview of the basic theory and main features of the CALYPSO method, as well as its versatile applications (limited only to a few works done in Ma’s group) on design of a broad range of materials including those of isolated clusters/nanoparticles, two-dimensional reconstructed surfaces, and three-dimensional bulks (at ambient or high pressure conditions) with a variety of functional properties. It is to say that CALYPSO has become a major structure prediction technique in the field, with which the door for a functionality-driven design of materials is now opened up.

This paper was originally published in Computational Materials 112, Part B, 2016, 406–415