1-17. Composition-structure-based descriptors for screening electrochemical storage Materials
Jianjun Liu
Integrated Computational Materials Scientific Research Center, Shanghai institute of Ceramics, Chinese Academy of

Sciences, 1295 Dingxi Road, Shanghai 200050, China

Abstract: Establishing quantitative/qualitative description for material composition-structure effect on property is a fundamental base for designing new materials and optimizing material performance based on high-throughput computational methods and machine-learning models, as well as an important part of materials genomic engineering. Taking electrochemical storage material as an example, the goal in this talk is to establish quantitative relationship of composition-structure-property via revealing coupling mechanism between electron transfer and electrochemical reaction in an atomic/molecular level. As such, some high-performance electrochemical storage materials are screened and designed by combining high-throughput calculations, machine-learning, and experimental confirmation. For electrochemical storage materials, competitivity between electroactivity and structural stability is a common issue of electrochemical field because it is physical origin of competitivity between storage capacity and cycling performance. For solving poor cycling performance of Li-excessive Li-ion cathodes, the quantitative relationship between reversible capacity and valance-structure, atomic size, and atomic electronegativity is established for designing high-performance cathodes through high-throughput screening and machine-learning. For solving the shuttling effect of Li-S battery, through calculating binding strength and surface electronegativity of cathodes, we established quantitative screening factor of cathodes according to difference of atomic electronegativities between cation and anion. These predicts are partially verified by our experiment and available data. This composition-structure screening factors such as atomic electronegativity and radii have further been extended into electrocatalytic field such HER. Therefore, These studies indicate that combining calculation and experimental techniques cannot provide deep insight on understanding materials’ physical and chemical properties, but also establish effective strategy for new material design and performance optimization of classical materials.

      
基于组成与结构的电化学储能材料筛选因子

刘建军

 中科院上海硅酸盐研究所，集成计算材料研究中心，上海市定西路 1295 号，200050 

摘要:建立材料组成结构对性能影响的定量/定性描述是高通量计算与机器学习筛选与设计新材 料、优化材料性能的重要基础，也是材料基因工程重要组成部分。本报告以电化学储能材料为 例，通过在电子-原子分子层次揭示电荷转移与电化学反应耦合机制建立定量化组成结构-性能 关系，通过高通量计算、机器学习、实验验证设计系列高性能电化学储能材料。电化学储能材 料的电化学活性与结构稳定性竞争是储能容量与循环性能竞争关系是电化学传统问题。针对富锂相正极材料可逆循环性能差的问题，以电子空穴定量表征为基础，建立局域结构的价键结构、原子电负性、原子半径等对析氧反应与可逆容量的定量关系，高通量筛选与数据挖掘结合，设 计一系列高性能富锂相正极材料。针对锂硫电池穿梭效应问题，通过计算正极材料对多硫分子吸附能与表面电负性，利用组成原子电子亲和势差值建立材料定量筛选因子，部分计算结果得 到实验验证，这种原子电负性与原子半径等组成结构特性作为电化学活性筛选因子的策略进一 步扩展到析氢电催化材料筛选，计算结果与实验一致。 因此建立基于组成结构的材料筛选因子 不仅对理解材料基本特性提供帮助，也为筛选设计新材料、优化材料性能提供重要理论基础。