Shotgun crystal structure prediction using machine-learned formation energies  

作　　者：Liu Chang Hiromasa Tamaki Tomoyasu Yokoyama Kensuke Wakasugi Satoshi Yotsuhashi Minoru Kusaba Artem R.Oganov Ryo Yoshida 

机构地区：[1]The Institute of Statistical Mathematics,Research Organization of Information and Systems,Tachikawa,Tokyo,190-8562,Japan [2]Technology Division,Panasonic Holdings Corporation,Kadoma,Osaka,571-8508,Japan [3]Skolkovo Institute of Science and Technology,Skolkovo Innovation Center,Moscow,121205,Russia [4]The Graduate Institute for Advanced Studies,The Graduate University for Advanced Studies(SOKENDAI),Tachikawa,190-8562,Japan

出　　处：《npj Computational Materials》2024年第1期13-26,共14页计算材料学(英文)

基　　金：supported in part by a Ministry of Education,Culture,Sports,Science and Technology(MEXT)KAKENHI Grant-in-Aid for Scientific Research on Innovative Areas(grant number 19H05820);Japan Society for the Promotion of Science(JSPS)Grants-in-Aid for Scientific Research(A)(grant number 19H01132);Early-Career Scientists(grant number 23K16955);JST CREST(grant numbers JPMJCR19I3,JPMJCR22O3,and JPMJCR2332);Computational resources were partly provided by the supercomputer at the Research Center for Computational Science,Okazaki,Japan(projects 23-IMSC113 and 24-IMS-C107).

摘　　要：Stable or metastable crystal structures of assembled atoms can be predicted by finding the global or local minima of the energy surface within a broad space of atomic configurations.Generally,this requires repeated first-principles energy calculations,which is often impractical for large crystalline systems.Here,we present significant progress toward solving the crystal structure prediction problem:we performed noniterative,single-shot screening using a large library of virtually created crystal structures with a machine-learning energy predictor.This shotgun method(ShotgunCSP)has two key technical components:transfer learning for accurate energy prediction of pre-relaxed crystalline states,and two generative models based on element substitution and symmetry-restricted structure generation to produce promising and diverse crystal structures.First-principles calculations were performed only to generate the training samples and to refine a few selected pre-relaxed crystal structures.The ShotunCSP method is less computationally intensive than conventional methods and exhibits exceptional prediction accuracy,reaching 93.3%in benchmark tests with 90 different crystal structures.

关 键 词：CRYSTAL STRUCTURE CRYSTALLINE 

分 类 号：O61[理学—无机化学]