Hierarchical Structuring of Mg2Si1-xSnx Alloys for Lower Thermal Conductivity and Thermal Conductivity as Probe for Structural Characterization

Licentiatavhandling, 2018

The thermoelectric effect was discovered nearly 200 years ago, but it started to be utilized for waste heat recovery fairly recently. From the economical point of view, such application requires a cheap and efficient material, which is able to convert heat energy into electricity.  Magnesium silicide-stannide alloys, also being environmentally friendly and stable at medium temperatures, have a further potential in increasing thermoelectric efficiency utilizing microstructuring, which has not been much applied to this system. The approach is aimed on reducing the lattice thermal conductivity, while keeping electron transport almost unaffected. In the current work it is suggested to manipulate the microstructure of Mg2Si1-xSnx alloys via formation of endotaxial phases utilizing phase separation. Compositions as well as arrangements of the endotaxial phases are controlled via the heat treatment protocols, which vary depending on the position of the binodal curve, the ratio of Si and Sn in Mg2Si1-xSnx and cooling rate.

Due to the discrepancy in the available quasi-binary phase diagrams Mg2Si-Mg2Sn below the solidus, one focus of the current study was on the estimation of the binodal curve position. For this, Si-rich Mg2Si1-xSnx alloys of different compositions were treated at different temperatures and their structure was investigated utilizing X-ray diffraction. The study showed an agreement of the results with one of the theoretically calculated binodal curve in the Si-rich region. However, the compositions of the Sn-rich phases did not fit to this or any other known models. Moreover, Sn-rich phases treated at higher temperatures contained less Si, whereas the solubility limit of Si in the Sn-rich phase is expected to grow with temperature, which can be a result of pinning effect provided by particle/grain boundaries.

The acquired approximate position of the binodal curve in the Si-rich region allows to control the phase separation process, and hence the microstructure. Thus, another focus of the thesis was put on creating the finest and most promising microstructure for thermoelectric materials, i.e. alternating lamellae-type endotaxial phases, which can, in principle, be achieved during spinodal decomposition. Such microstructure was found during the experiments of the current work. It is shown that when a compound enters the miscibility gap at temperatures that are too low for migration of the atoms over long distances, it rapidly decomposes forming lamellae with similar compositions. Alternatively, if a compound enters the miscibility gap at higher temperatures, higher cooling rates affect the phase separation similarly.

In addition, it was suggested to utilize the Transient plane source technique in quality control and advanced thermal conductivity characterization of manufactured thermoelectric legs. Hence, the thesis also includes the recent development of the so-called Structural probe technique, which makes it possible to convert the temperature vs. time function to the unique thermal conductivity vs. probing depth. Since the thermal conductivity is sensitive to the structural constitution of a material, such function allows to assess the microstructure variations with depth. The technique was successfully tested on homogeneous and inhomogeneous materials as well as the materials with macroscopic defects.