Date of Award
Doctor of Philosophy (PhD)
Tritt, Terry M
Daw , Murray
He , Jian
Liebenberg , Donald
Thermoelectric materials provide a solid state route for the direct conversion between thermal and electrical energy. Hence, thermal gradients can be used to generate power of up to several KW, or driven DC power can be used to pump heat in or out of a system. Recent pushes in thermoelectric research are for power generation materials with higher efficiency. These materials can be used to recover energy from the waste heat of vehicles' exhaust systems, and as power generators for extraterrestrial exploration when coupled with a radioactive heat source.
The goal of this dissertation will be to demonstrate a new synthesis technique for the current state of the art thermoelectric material for high temperature power generation, silicon germanium (SiGe). This technique is referred to as the single element (SE) spark plasma sintering (SPS) technique because the single elements of silicon, germanium, and their n and p type dopants are alloyed together during the SPS consolidation process. This novel synthesis technique is two orders of magnitude faster than the original technique for alloying this material and one order of magnitude faster than the current technique used for alloying this material. The innovation and newness of this technique presented to the scientific community is closely tied to material science, while the understanding of the resulting thermoelectric material and its properties are perceived through a physics lens.
In order to fully demonstrate that the SE SPS technique alloys SiGe several scientific studies and investigations are performed. First, SiGe is alloyed using the current state of the art method, mechanical alloying (MA). Powders of MA SiGe are traditionally consolidated by a conventional hot press (HP). These materials are employed by NASA for deep space power generation on radio-isotope thermoelectric generators (RTGs). Hence, there is readily available published data for MA+HP SiGe used in RTGs. The SiGe powder that is MA by the author is consolidated using the SPS process, MA+SPS. Therefore, an initial study was conducted to ensure that the SPS consolidation process was not having any adverse effects SiGe as compared to the HP technique. Essentially it will be shown that SiGe produced by the MA+HP method and the MA+SPS method are equivalent. This guarantees that the synthesis and characterization techniques used at the complex and advanced materials laboratory (CAML) by the author agree with published standards.
Second, once the first study has demonstrated that no adverse effects occur by using the SPS to consolidate SiGe, a study was conducted to show that undoped single elements of silicon and germanium can be alloyed in the SPS. To confirm that undoped SiGe is truly alloyed using the SE SPS technique, the structural properties of the resulting materials were investigated. Based on the densities, x-ray diffraction patterns, derived lattice constants, and Vegard's law it will be shown that the SE SPS method does successfully alloy multiple compositions of undoped SiGe.
The third and most important study demonstrated that SiGe alloyed using the SE SPS synthesis technique can be successfully doped to a n and p type thermoelectric (TE) material. This required an investigation of all of the TE transport properties of these materials. A significant investigation and commentary will be provided for the lattice thermal conductivity of SiGe. The need for this investigation arises from the difference in synthesis processes between the traditional MA and the novel SE SPS techniques. The MA powder is already alloyed into micron sized powders that are consolidated by the HP for an extended time (>1 hour), which allows for grain growth. The SE SPS method relies on diffusion being promoted by the electric field assisted sintering technique and occurs over a very short period of time (<30 minutes). Therefore it can not be assumed that grain growth is not affected by the time dependent processes of sintering and diffusion with the SE SPS process. As will be discussed grain size plays a role in the lattice thermal conductivity of SiGe. It is surprising and physically interesting that the MA+HP standards and the SE SPS samples have lattice thermal conductivities that indicate the dominant scattering mechanism is the same.
The physical insight provided by the fourth study is made possible by the existence of the new SE SPS synthesis method for SiGe. The MA method is optimized by the addition of GaP to the n-type SiGe materials during processing. The explanation for this optimization is a subject of debate within the community. Although, a staunch conclusion can not be made due to the need for more samples and carrier concentration data, this initial study does indicate that one physical explanation within the debate for the improvement of n-type SiGe with GaP additions is more coherent with scientific experimentation.
The fifth study is aimed to provide suggestions for future studies for improving this material. This includes brief investigations on the effects of various nano-structure inclusions on lattice thermal conductivity of SiGe alloys. The study is meant to be used as a tool for future students who wish to investigate the interesting physical properties of this system.
In conclusion, this dissertation will answer the question, 'Can a new synthesis technique for the current state of the art thermoelectric material for high temperature power generation, silicon germanium (SiGe), be formulated?' In answering this question new physical insight will be contributed to the community. First, a new synthesis technique that is advantageously faster and simplistic in its doping methods will be provided that will allow for rapid empirical investigations. Second, the scattering mechanisms for lattice thermal conductivity have been investigated for this new technique and shown to be the same as the traditionally alloyed SiGe. Third, the SE SPS synthesis of SiGe allows for new investigations on the effects of GaP doping of n-type SiGe, which could bring to resolution the controversy/misunderstanding surrounding this doping process. And finally, suggestions for future studies are provided by brief investigations on the lattice thermal conductivity of nano-composite SiGe materials and from questions that arose while reviewing the literature of others working on SiGe.
Thompson, Daniel, "Thermoelectric Properties of Silicon Germanium: An In-depth Study to the Reduction of Lattice Thermal Conductivity" (2012). All Dissertations. 984.