Data-driven Design of Eco-friendly Thermoelectric High Entropy Sulfides

High entropy compounds with compositional complexity can be designed which may provide new high performance thermoelectric materials. To overcome the combinatorial explosion problem in their design, a data driven model was... [ view full abstract ]

High entropy compounds with compositional complexity can be designed which may provide new high performance thermoelectric materials. To overcome the combinatorial explosion problem in their design, a data driven model was developed and demonstrated for copper containing sulfides, which are eco-friendly thermoelectric materials and are of great interest due to the world’s demands for green energy generation. Using the data driven model, two high entropy sulfides, metallic Cu₅SnMgGeZnS₉ and semiconducting Cu₃SnMgInZnS₇, were designed. They were then successfully fabricated as single phase dense ceramics using mechanical alloying and spark plasma sintering. XRD confirmed that they were single phase and EDX showed that the cations were homogenously distributed. Their phase stability and atomic local structures were also investigated using density functional theory calculations. Finally, a zT value of 0.58@773K was obtained for Cu₅Sn_1.2MgGeZnS₉, where additional Sn was used to tune the carrier concentration. The other compound Cu₃SnMgInZnS₇ is a semiconductor for which no suitable dopant has yet been identified. This work demonstrated a simple approach to find new functional high entropy compounds in the largely unexplored multi-element chemical space.