A Multi-Scale Approach for the Control of Thermoelectric Properties in Bismuth Telluride Alloys through Laser Powder Bed Fusion 295 views

Thermoelectric (TE) materials have garnered interest over the past several decades as a potential solution to rampant industrial heat waste due to their ability to convert between heat and electrical energy directly and reversibly. However, their widespread adoption has been hindered by their low conversion efficiency (≈5 %) and high production cost. Laser powder bed fusion (LPBF) is an additive manufacturing (AM) technique with the ability to create free form geometries through layer-by-layer construction and reduce cost through the minimization of material waste and manufacturing steps. This work summarizes a series of hierarchical approaches to leverage LPBF and purposefully affect the TE properties of bismuth telluride alloys across multiple length-scales. First, through an iterative process of augmented machine learning and experimental validation, optimized LPBF processing parameters were rapidly identified for the construction of highly dense (>99 %), crack-free Bi2Te2.7Se0.3 parts in nonstandard geometries. Second, an annealing-induced enhancement of TE efficiency was achieved, and potential mechanisms were explored. Specifically, manipulation of the point defect and charge carrier concentrations was proposed as a driver of improved electrical conductivity and Seebeck coefficient values, while simultaneous grain boundary manipulation provided the necessary pathway for a reduced thermal conductivity. Third, the ability to create intentional thermal and electrical transport behavior anisotropy was demonstrated through LPBF-induced crystallographic orientation control. Fourth, an ability to control the sign and magnitude of the Seebeck coefficient through LPBF processing parameter alterations was discovered. Possible methods for this experimental control were explored, including bismuth-rich oxide formation, matrix composition changes, and bulk porosity development. Ultimately, these studies collectively advance the control of TE properties using LPBF while providing a framework for broader AM material strategies.

PHD (Doctor of Philosophy)

Additive manufacturing; Laser powder bed fusion; Bismuth telluride; Thermoelectric materials; Machine learning; Functional properties

English

2024-04-22