Ruiheng Liu1 HongYi Chen1 Xun Shi1 Chen Lidong1 Wenqing Zhang2 Kunpeng Zhao1 Ctirad Uher3

1, Shanghai Institute of Ceramics, Chinese Academic of Science, Shanghai, , China
2, Southern University of Science and Technology, Shenzhen, , China
3, University of Michigan, Troy, Michigan, United States

High-throughput explorations of novel thermoelectric materials based on the Materials Genome Initiative paradigm only focus on digging into the structure-property space using nonglobal indicators to design materials with tunable electrical and thermal transport properties. As the genomic units, following the biogene tradition, such indicators include localized crystal structural blocks in real space or band degeneracy at certain points in reciprocal space. However, this nonglobal approach does not consider how real materials differentiate from others. Here, we successfully develops a strategy of using entropy as the global gene-like performance indicator that shows how multicomponent thermoelectric materials with high entropy can be designed via a high-throughput screening method. Optimizing entropy works as an effective guide to greatly improve the thermoelectric performance through either a significantly depressed lattice thermal conductivity down to its theoretical minimum value and/or via enhancing the crystal structure symmetry to yield large Seebeck coefficients. The entropy engineering using multicomponent crystal structures or other possible techniques provides a new avenue for an improvement of the thermoelectric performance beyond the current methods and approaches.