Dr. David Singh Oak Ridge National Labs

There is increasing interest in thermoelectric materials motivated in part by recent progress and in part by the potential of these materials in various energy technologies. Thermoelectric performance is a multiply contra-indicated property of matter. For example, it requires (1) high thermopower and high electrical conductivity, (2) high electrical conductivity and low thermal conductivity and (3) low thermal conductivity and high melting point. The keys to progress are finding an optimal balance and finding ways of using complex electronic and phononic structures to avoid the counter-indications mentioned above. In this talk, I discuss some of the issues involved in the context of recent results. One key aspect is optimization of the doping level in a given thermoelectric material. While this has long been understood in terms of standard semiconductor parabolic band models, we find surprisingly different results for many thermoelectric materials when the actual first principles band structures are used. This has led to prediction of a number of useful thermoelectrics, some that are new, and surprisingly some that are old. This work was done in collaboration with David Parker, Xin Chen, Olivier Delaire and Mao-Hua Du and was supported by the Department of Energy through the S3TEC Energy Frontier Research Center.