Waste Heat Recovery with Oxide Thermoelectrics
 
Investigators: Matthew L. Scullin, Jayakanth Ravichandran, Subroto Mukerjee, Joel Moore, Arun Majumdar, R. Ramesh
 
Thermoelectric materials are those which show an appreciable Seebeck coefficient ("thermopower", Volts/Kelvin) and efficiency in converting a thermal gradient to an electric current in the solid-state. Over fifty percent of the world's 10+ Terrawatts of power produced is lost as waste heat. The ability to capture only a tenth of a percent of this energy globally at ten percent efficiency would produce enough power to eliminate the need for 1,500 coal-fired plants or 5.8 million barrels of oil, and would be a billion dollar industry at 10 cents / kWh. Much of this waste heat is high-quality -- temperature gradients in excess of 100-200 °C -- where thermoelectric materials stable at high temperatures can capture and convert this otherwise wasted energy into useful electricity.

Attractive for their high-temperature thermoelectric performance are complex oxide materials, which, because of their high carrier effective masses, show a large thermopower that increases with temperature combined with a wide tunability in electrical conductivity and excellent chemical stability. The "thermoelectric figure of merit" zT -- a dimensionless, intrinsic measure of the efficiency of a thermoelectric material -- is a product of the square of the thermopower, the electrical conductivity, and the inverse of the thermal conductivity. Finding a combination of these properties which yields a high-efficiency material is far from trivial, as they are all coupled and changing one usually adversely affects another. The current state-of-the-art, Bismuth Telluride, has a zT just above 1 at room temperature. The oxide with the highest zT, Niobium-doped strontium titanate, achieves zT = 0.4 at 1000K. This research explores the effect of double-doping in this strontium titanate system, where the addition of La in place of Sr offers a large electrical conductivity while the introduction of oxygen vacancies drives up the electron effective mass and in turn the thermopower. Through selecting the combination of dopant concentrations that yields both the maximum effective mass and carrier concentration, zT in this system has been pushed as high as 0.25 at 300K, and is expected to be above 1 at 1000K, making it a viable material for high-temperature energy conversion applications such as in power plants, factories, and automobiles.
 
Fig. 1: Thermoelectric efficiency zT as a function of temperature for increasing La-doping in highly oxygen deficient SrTiO3-δ thin films.