Thermoelectric properties of complex mesowires

Thermoelectric (TE) materials convert electric currents into temperature gradients and vice versa. As such they can be used both for solid state refrigeration and for conversion of waste heat into electricity, processes of great interest for a cleaner environment and a more efficient energy household. However, this prospect has been hampered by the relatively low efficiency of commercially available TE materials, restricting their technical applicability. The prediction made more than twenty years ago that this can be remedied by nanotechnological means has triggered extensive research efforts worldwide, most focusing on the introduction and stabilization of nanostructures in bulk TE materials. The main problem with bulk nanostructuring, however, is that it usually does not only alter the grain sizes but also the material`s density, composition, doping level, amount of foreign phases, and so on. As a consequence our understanding of the effects that nanostructuring has on the charge and heat carriers of the material is very limited. This situation calls for a new approach. ThermoWire shall overcome these limitations by studying TE properties on single, well-shaped mesoscopic (tens of nanometers to micrometers) structures of TE materials of great current interest. Measurements on single mesoscopic samples hold great promise for the complex TE materials we propose to study here: (1) rattling compounds, (2) materials with strongly correlated electrons, and (3) topological insulators. In these three materials classes mesostructuring is particularly interesting for very different reasons: in (1) it will reveal whether indeed very long-wavelength phonons dominate the heat transport, in (2) it will test whether giant thermoelectric powers are robust against mesostructuring, and in (3) it will reveal whether high surface conductivities may boost the TE efficiency. Unfortunately, the reproducible bottom-up fabrication of mesowires is an unpredictable and tedious undertaking, and for most of the new materials that shall be investigated no routes of production are known. Therefore, the method that ThermoWire shall employ is a top-down approach. Bulk single crystals of selected compounds of all three material classes shall be synthesized, shaped by a focused ion beam technique into mesowires of varying diameters, crystallographic direction, and surface constitution, and measured one-by-one on a dedicated mesowire measurement platform. This will allow us to disentangle the different effects, compare the results with theory, and thus distinctly advance the field. Having recently developed the techniques, we are well-poised to go about these challenging tasks.

Thermoelectric Properties of Complex Mesowires Thermoelectricity is a field that combines thermal and electrical processes, most typically in the solid state. The key property is the thermopower, a measure of the magnitude of an induced "thermoelectric" voltage in response to a temperature difference across that material. Materials with large thermoelectric effects are of interest for applications, for instance for waste heat recovery or, more recently, advanced functionalities in quantum devices. However, the microscopic mechanisms of thermoelectric properties are poorly understood. This is particularly true for complex materials with strongly interacting electrons and/or phonons, and in materials with nontrivial electronic topology, which hold great promise for such devices. In this project we set out to study several carefully selected materials as bulk (macroscopic) single crystals and, where appropriate, in structured form ("mesowires", with diameters from about hundred nanometers to a few micrometers). As an example, in the topological insulator TlBiSe2, we discovered an intriguing light sensitivity of the electrical resistivity below a certain temperature. The robustness of this effect in a mesostructure confirms its topological origin and opens the door to exciting possibilities in devices. In type-I clathrates, which have high potential for thermoelectric applications due to their low lattice thermal conductivity, we found that the two different types of lattice vibrations, localized and propagating phonons, interact with each other in a way reminiscent to the electronic Kondo effect, thereby localizing the short-wavelength part of the propagating phonons. Whether the remaining long-wavelength phonons can be cut off by mesostructuring was investigated in clathrates mesowires. An unexpected highlight of the project was the discovery that materials previously classified as Kondo semimetals or semiconductors host topologically nontrivial states in their bulk. Though this made mesowire studies on these materials less pertinent, we put considerable effort in pushing forward this frontier topic.