具有声子玻璃电子晶体特征的三元层状热电材料的高通量计算

Layered ternary compounds, composed of electron transport layers and strong phonon scattering layers, have ultra-low thermal conductivities but limited power factors. Due to the wide chemical and structural space, their thermoelectric properties could be largely improved. This project will combine high throughput computations and data mining to quickly screen copper and silver based ternary compounds in layered structures, especially those with a stoichiometric ratio of 1:1:2. Within this project, we will establish a high throughput thermoelectric materials design platform and database to obtain nontoxic and stable advanced layered thermoelectric materials. We will try to obtain a better understanding of the main issues currently associated with layered compounds, such as complicated structures, low power factor and nonideal carrier concentration. The correlation between properties of interest, such as stability, power factor and defect formation energy, and basic chemical and structural properties will also be established through a careful data mining study. The eventual goal is to establish design principles to guide experimental synthesis and advance layered thermoelectric materials development.

由电子传输层和强声子散射层构成的三元层状热电材料具有极低的热导率，其丰富的元素组合和多元的晶体结构提供了较大的热电性能优化空间。本项目拟采用高通量计算模拟结合数据挖掘的方法，针对铜基和银基具有层状结构的三元热电材料尤其是化学计量比为1:1:2的材料进行系统的理论研究，期望就这类材料现在面临的主要问题（结构复杂、功率因子低、载流子浓度不理想）提出理论解释及优化方案。旨在建立针对热电材料设计的高通量计算平台和相关数据库，获得无毒的、稳定的、性能优异的新型层状热电材料系统，研究材料稳定性、电输运、缺陷等与组分和晶体结构的关联作用，建立层状热电材料的优化准则，为实验开发提供理论指导并加速新型层状热电材料的性能突破。

由电子传输层和强声子散射层构成的三元热电材料具有极低的热导率，其丰富的元素 组合和多元的晶体结构提供了较大的热电性能优化空间。本项目运用高通量第一性原理计算，对P3m1和I-42d结构的XYZ2（X，Y，Z分别代表三价，一价，负二价元素）材料的稳定性、电子结构、电输运性能等进行理论预测并将结果存入数据库。对数据库所有化合物进行稳定性分析，并对稳定化合物进行基于zT值的筛选，结合机器学习算法寻找并解释兼具高稳定性和高zT值的XYZ2型热电材料的关键因素。通过本研究，我们发现了碱金属塞贝克有效质量主导型以及d区元素-第五主族复杂因子主导型的两类高性能XYZ2热电材料，得到600K，1020cm-3载流子浓度下zT值高于1.5的高性能热电材料29种，其中p型空间群为P3m1的AgBiS2在上述条件下的理论计算zT值高达2.676。同时化合物的成分变化引起的轨道能量不同，最终形成带边简并度形貌不同的电子结构，是产生上述两种不同高性能热电材料的物理本质。