DESIGN AND DISCOVERY OF HIGHLY EFFICIENT THERMOELECTRIC MATERIALS

Thermoelectric energy conversion efficiency is directly related to the temperature difference over which the device operates, its average temperature of operation and the transport properties of the thermocouple material represented by Z T, the dimensionless figure of merit. To maximize ZT and thus the conversion efficiency, materials with a large Seebeck coefficient, high electrical conductivity and low lattice thermal conductivity are required. State-of-the-art thermoelectric materials, such as Bi,,Sb,Te, material, the process of maximizing ZT in a given temperature range has been classically conducted by two approaches: optimization of the electrical properties and thermal conductivity reductions. Although substantial improvements in the performance of state of the art materials have been achieved over the whole 1 OOK to 1500K temperature range of application, no one practical material seems to significantly break through the limit ZT,==l. Much larger values of ZT are theoretically possible and the investigation of radically new materials and the development of novel thin film heterostructures appear to be promising new approaches. Materials of interest include highly covalent semiconductors and semimetals with complex open crystal structures allowing for very low lattice thermal conductivity through ”void filling”, atomic “rattling”, valence fluctuations and long range disorder. Materials presenting a peculiar electronic density of states susceptible to lead to unusually large Seebeck coefficients are also attractive. Recent studies on new materials, such as skutterudites, Chevrel phases and rare earth compounds, illustrate how novel compositions with a good potential for achieving ZT much larger than 1 can be identified and their properties engineered. Y Se, or Si,-,Ge,, are conventional semiconducting solid solutions with high doping levels. For a give