Chemical-disordered materials have a wide range of applications whereas the determination of their structures or configurations isone of the most important and challenging problems. Traditional methods are extremely i...Chemical-disordered materials have a wide range of applications whereas the determination of their structures or configurations isone of the most important and challenging problems. Traditional methods are extremely inefficient or intractable for large systemsdue to the notorious exponential-wall issue that the number of possible structures increase exponentially for N-body systems.Herein, we introduce an efficient approach to predict the thermodynamically stable structures of chemical-disordered materials viaactive-learning accompanied by first-principles calculations. Our method, named LAsou, can efficiently compress the samplingspace and dramatically reduce the computational cost. Three distinct and typical finite-size systems are investigated, including theanion-disordered BaSc(O_(x)F_(1−x))3 (x = 0.667), the cation-disordered Ca_(1−x)Mn_(x)CO_(3) (x = 0.25) with larger size and the defect-disordered ε-FeC_(x) (x = 0.5) with larger space. The commonly used enumeration method requires to explicitly calculate 2664, 1033,and 10496 configurations, respectively, while the LAsou method just needs to explicitly calculate about 15, 20, and 10configurations, respectively. Besides the finite-size system, our LAsou method is ready for quasi-infinite size systems empoweringmaterials design.展开更多
基金The authors are grateful for the financial support from the National Key R&D Program of China(No.2022YFA1604103)National Science Fund for Distinguished Young Scholars of China(Grant No.22225206)+5 种基金the National Natural Science Foundation of China(Nos.21972157,21972160 and 21703272)CAS Project for Young Scientists in Basic Research(YSBR-005),Key Research Program of Frontier Sciences CAS(ZDBS-LY-7007)Major Research plan of the National Natural Science Foundation of China(92045303)CAS Project for Internet Security and Information Technology(CAS-WX2021SF0110)Science and Technology Plan Project of Inner Mongolia Autono-mous Region of China(2021GG0309)funding support from Beijing Advanced Innovation Center for Materials Genome Engineering,Synfuels China,Co.Ltd,and Institute of Coal Chemistry(CAS).Q.P.would like to acknowledge the support provided by LiYing Program of the Institute of Mechanics,Chinese Academy of Sciences(Grant No.E1Z1011001).
文摘Chemical-disordered materials have a wide range of applications whereas the determination of their structures or configurations isone of the most important and challenging problems. Traditional methods are extremely inefficient or intractable for large systemsdue to the notorious exponential-wall issue that the number of possible structures increase exponentially for N-body systems.Herein, we introduce an efficient approach to predict the thermodynamically stable structures of chemical-disordered materials viaactive-learning accompanied by first-principles calculations. Our method, named LAsou, can efficiently compress the samplingspace and dramatically reduce the computational cost. Three distinct and typical finite-size systems are investigated, including theanion-disordered BaSc(O_(x)F_(1−x))3 (x = 0.667), the cation-disordered Ca_(1−x)Mn_(x)CO_(3) (x = 0.25) with larger size and the defect-disordered ε-FeC_(x) (x = 0.5) with larger space. The commonly used enumeration method requires to explicitly calculate 2664, 1033,and 10496 configurations, respectively, while the LAsou method just needs to explicitly calculate about 15, 20, and 10configurations, respectively. Besides the finite-size system, our LAsou method is ready for quasi-infinite size systems empoweringmaterials design.
