A search for good semi-metallic thermoelectric materials

Heavily doped semiconductors are the most studied class of thermoelectric materials. The presence of the bandgap breaks the symmetry between electrons and holes, allowing large Seebeck coefficient values which are 2-3 orders of magnitude higher than metals. However, the optimal chemical potential in these materials requires a high level of doping usually larger than 10¹⁹ cm^-3, resulting in low carrier mobility. Semimetals have properties in between semiconductors and metals with Seebeck coefficient values in between. Since there is no bandgap, the Seebeck coefficient values are small for structures where in electron-hole bands are similar. For example, for Dirac semimetals and at the Dirac point, the Seebeck coefficient is zero. However, if there is a large asymmetry between electron-hole bands (that is there is a large effective mass difference), then the Seebeck coefficient of intrinsic semimetals can be as large as those of heavily doped semiconductors. This combined with large intrinsic carrier mobility, makes semimetals potential thermoelectric candidates. In this talk, I will present our search results for semimetals with large Seebeck coefficient values using first principles calculations.