Accelerated prediction of atomically precise cluster structures using on-the-fly machine learning

作     者：Yunzhe Wang Shanping Liu Peter Lile Sam Norwood Alberto Hernandez Sukriti Manna Tim Mueller 

作者机构：Department of Materials Science and EngineeringJohns Hopkins UniversityBaltimoreMD21218USA 

出 版 物：《npj Computational Materials》 (计算材料学（英文）)

年 卷 期：2022年第8卷第1期

页      面：1644-1653页

核心收录：

学科分类：081704[工学-应用化学] 07[理学] 070304[理学-物理化学(含∶化学物理)] 08[工学] 0817[工学-化学工程与技术] 0703[理学-化学] 

基　　金：The work was supported by the Office of Naval Research under the grant No.ONR MURI N00014-15-1-2681,Calculations were performed using computational resources from the Maryland Advanced Research Computing Cluster(MARCC),the Stampede2 supercomputer at the Texas Advanced Computer Center(TACC)and the Gordon supercomputer in Department of Defense High Performance Computing Modernization Program TACC resources were provided through the XSEDE program with NSF award DMR-140068,Images of the atomic structures of clusters were generated using VESTA85 

主　　题：precise atomic cluster 

摘      要：The chemical and structural properties of atomically precise nanoclusters are of great interest in numerous applications,but predicting the stable structures of clusters can be computationally *** this work,we present a procedure for rapidly predicting low-energy structures of nanoclusters by combining a genetic algorithm with interatomic potentials actively learned *** this approach to aluminum clusters with 21 to 55 atoms,we have identified structures with lower energy than any reported in the literature for 25 out of the 35 *** benchmarks indicate that the active learning procedure accelerated the average search speed by about an order of magnitude relative to genetic algorithm searches using only density functional *** work demonstrates a feasible way to systematically discover stable structures for large nanoclusters and provides insights into the transferability of machine-learned interatomic potentials for nanoclusters.