Novel Nanostructured Thermoelectrics
While the ultimate CEEM goal is the development of materials with improved thermoelectric power generation capabilities, the path to that goal involves discovering and using the important physics of the electrical transport, thermoelectric potentials and heat transport in the new structures that we prepare. This involves importantly understanding and engineering the inclusion, transport and scattering of mobile charge carriers and the propagation and scattering of heat-carrying lattice vibrations in the materials, all of which are crucially dependent on the physical and electronic structure of the materials. And since the materials that are being created at CEEM are new, both structurally and chemically, there is a great deal of significant chemistry and materials science discovery that is needed and results from the work.
Our Scientific objective is to develop new thermoelectric materials using semimetal nanoparticles in semiconductors to simultaneously increase the Seebeck coefficient and electrical conductivity, and reduce the electronic and lattice thermal conductivities. We explore a wide range of materials, including rare earth bismuthides, antimonides, phosphides, nitrides, and arsenides and clathrates. Our calculations show that ZT > 2 is possible, and the design can be optimized for a particular temperature range while a stack of such layers is optimized for kW generation from large temperature differences. We target two types of nanostructured TE materials: (1) bulk systems with nanostructure-modulated heterostructures made by incorporating metallic nanodomains into a semiconducting matrix, and (2) packaged, highly oriented nanowire arrays which can be assembled on a large scale with electrochemical control of composition and impurity levels, and with well-defined nanowire dimensions and sculptured nanoscale geometries. As a TE p-n semiconductor patterned nanowire array, the latter will make use of recent observations demonstrating the exceptionally low thermal conductivities, at or below the Slack limit, that can be realized by using appropriate nanoscale wire widths and surface roughnesses.
Figure above: STEM image of SrTiO3/GdTiO3 superlattice structuregrown by molecular beam epitaxy. Clear atomic level abruptinterfaces are made between the two
oxides with Perovskitestructure.
Figure above: SEM image of highly ordered Si nanowire arrays embedded in spin-on
glass for thermoelectric applications and thermoreflectance imaging of the integrated
heater on a device fabricated for cross-plane Seebeck measurement.