Modeling and analysis of thermoelectric energy conversion efficiency in nanostructures.

116 p.

Recently, nanostructured materials have received tremendous attention because of their exceptional properties and great potential in applications. In this thesis, we investigated the thermoelectric transport properties of Si and Ge nanomembranes and nanowires as well as silica nanoclusters by using density functional theory calculations combined with Landauer-Buttiker formalism in the linear response regime. We also investigated the thermoelectric coefficients for quantum devices in the Coulomb blockade regime. The thesis is organized as follows.We first investigated the thermoelectric properties in one- and two-dimensional Si and Ge nanomembranes, i.e., silicene and germanene. We found that the figure of merit ZT in the one-dimensional zigzag-edged silicene and germanene nanoribbons decreases monotonously when the width of the nanoribbons increases. The decreasing figure of merit can be attributed to the rapid rising of the thermal conductance. Broader nanoribbons have more phonon modes contributing to the transport than those of the narrow nanoribbons, while electron conduction does not change so much. For armchair-edged silicene and germanene nanoribbons, ZT as a function of ribbon width shows an oscillating behavior. To improve the thermoelectric performance, we mixed the Si and Ge in the nanoribbons to enhance the lattice scattering. We found that the figure of merit at room temperature for hybrid narrow silicene and germanene nanoribbons is remarkably high, up to 2.5.Secondly, we investigated the thermoelectric energy conversion efficiency of Si and Ge nanowires, and in particular, that of SiGe core-shell nanowires. We showed how the presence of a thin Ge shell on a Si core nanowire increases the overall figure of merit. We found the optimal thickness of the Ge shell to provide the largest figure of merit for the devices. We also considered the Ge-core/Si-shell nanowires. Here, we found that an optimal thickness of the Si shell does not exist, since the figure of merit is a monotonically decreasing function of the radius of the nanowire. Moreover, we considered the nanowire in which the shell is an alloy of Si and Ge. We verified the empirical law relating the electron energy gap to the optimal working temperature, at which the efficiency of the device is maximized.Thirdly, we investigated the heat transfer between two silica nanoclusters. We found that when the gap between two neighboring clusters is in the range 4 A to 3 times the cluster size, the thermal conductance decreases due to the surface charge-charge interaction. When the gap is further increased to 5 times the cluster size, the volume dipole-dipole interaction dominates. On the other hand, when the gap is smaller than 4 A, quantum effects dominate, where electrons of both clusters are shared. This quantum interaction leads to the dramatic increase of thermal coupling between neighboring clusters when the gap distance decreases. This study provides a description of the transition between radiation and heat conduction in gaps smaller than a few nanometers.Finally, we investigated the thermoelectric coefficients, in particular the Seebeck efficient, for strongly interacting electrons in the Coulomb blockade regime. The Seebeck coefficient plays a fundamental role in identifying the efficiency of a thermoelectric device. Its theoretical evaluation for atomistic models is routinely based on density functional theory calculations combined with the Landauer-Buttiker approach to quantum transport. This combination, however, suffers from serious drawbacks for devices in the Coulomb blockade regime. We showed how to cure the theory through a simple correction in terms of the temperature derivative of the exchange-correlation potential. We also compared our results with those obtained from both rate equations and experimental measurements on carbon nanotubes, and we found good qualitative agreement in all cases.Our results are expected to be beneficial for the understanding of material performance and for designing future nanodevices.