Micro/Nano-Scale Structural and Thermoelectric Characterization of Magnesium Silicide-Based Composite Materials: Investigating Self-Diffusion of Constituent Elements, Band Alignment and Transport Properties

The ability of thermoelectric (TE) materials to produce electrical power from otherwise wasted thermal energy has propelled this field into the spotlight, with immense potential for addressing critical issues such as energy efficiency, sustainable power generation, and environmental conservation. The effectiveness of a TE generator is intricately linked to the performance of its constituent TE materials, which, in turn, is constrained within uniform bulk materials by the interplay of three fundamental TE transport properties: the Seebeck coefficient, electrical conductivity, and thermal conductivity. The precise manipulation of the electronic band alignment at the micro/nanoscale can serve to disentangle these TE properties by selectively filtering out detrimental spectral fractions of the charge carriers. Among the diverse array of TE materials, magnesium silicide-based materials are the preferred choice for this purpose due to their spontaneous formation of nanostructured composites comprising distinct phases with varying band gaps dependent on the Si:Sn ratio, driven by a miscibility gap in the solid solution series. Furthermore, these materials are emerging as compelling candidates due to their potential to serve as cost-effective and environmentally friendly components, showcasing exceptional TE properties within the mid-temperature range of 500-800 K. However, the conventional bulk property measurements often lack the sensitivity required to detect the subtle effects of charge carrier filtering, necessitating micro/nano-scale characterizations of TE properties (Seebeck coefficient and thermal conductivity) of multi-phase materials. In our research, we have integrated several locally resolved experimental techniques with electronic transport modeling. The transient Seebeck microprobe, with its micron-scale resolution, allows us to map the distribution of the Seebeck coefficient. Combining the Seebeck coefficient map and the compositional map, obtained by establishing a simple correlation between the grey value in backscattered electron (BSE) images and the chemical composition from EDX point analysis, enables a local correlation between the Seebeck coefficient and the phase composition in multiphase materials. Moreover, employing transport modeling offers insights into the local distribution of the reduced Fermi level and the alignment of valence and conduction bands, enabling the generation of a micro-scale carrier concentration map. Furthermore, utilizing Kelvin probe force microscopy with nano-scale spatial resolution can provide the value of the Fermi level position with respect to the vacuum level, enabling generation of the same carrier concentration map at the nano-scale. Our work also demonstrates the potential of selective doping, using Bi as a dopant, to precisely modify the band positions of individual phases, a prerequisite for material enhancement by energy filtering. This localized characterization technique also aids our understanding of the inter-diffusion of Mg, Si, and Sn from one phase to another, resulting in distinct changes in the thermoelectric properties of each phase. This knowledge is vital for optimizing thermoelectric properties, as it can either amplify or attenuate the effects of selective doping. Hence, this integrated approach represents a crucial step towards optimizing composite materials through effective energy filtering.