A theoretical study of enhancing thermoelectric efficiency in pnictogen-chalcogen alloys via doping, strain, and nanostructuring

Abstract

Thermoelectric devices have the potential to revolutionize modern energy technologies with environmentally-friendly, renewable energy materials. They have the advantages of being scalable, maintenance-free, compact, durable, and multi-purpose, capable of providing heating and cooling as well as waste heat recovery. However, to be favored among other technologies necessitates multi-fold enhancements in energy conversion efficiency, defined in the thermoelectric figure-of-merit ZT=α2σκT, requiring high Seebeck coefficient α, high electrical conductivity σ, and low thermal conductivity κ.

Finally, using similar DFT-based transport calculations with supercells, this thesis explores the effects of divalent, p-type dopants Pb, Ca, and Sn, to replicate enhanced band quasi-degeneracy behavior under ambient conditions. When compared to intrinsic Bi2Te3 with p-type Bi anti-site defects, dilute (sub-atomic-percent) Pb doping enhances α2σ up to 75%, in agreement with experiment. This is due to the preservation of highly-degenerate, low-effective-mass valence bands, which enhance σ without degrading α. Thermoelectric transport in Ca-doped Bi2Te3 has yet to be experimentally measured. This thesis therefore predicts that Ca doping provides up to 60% α2σ enhancement, owed to increase in valence band quasi-degeneracy. Increasing Ca concentration leads to the formation of high-effective-mass bands, which enhance α instead of σ. Lastly, dilute Sn doping causes up to 30% degradation in α2σ, due to decreased valence band quasi-degeneracy and increased hole effective mass. Contrary to popular hypotheses, the Sn-induced resonant state does not enhance α2σ. Rather, it produces a lower secondary α2σ peak at higher hole concentration. It is therefore not resonant states but increased band-edge quasi-degeneracy which transcends single crystal α2σ limits in Bi2Te3. When combined with nanostructuring, Pb and Ca doping can achieve ZT ~ 1.7.

This thesis starts by exploring the limits of increasing ZT in Bi2Te3 by suppressing κ while preserving a bulk-like power factor α2σ. Non-parabolic, two-band Boltzmann transport and Debye-Callaway models, modified to account for nano-grains and nano-pores, predict and demonstrate how optimized nanostructuring can provide ZT ~ 1.3, 50% higher than single crystal bulk.

To achieve further ZT enhancements, this thesis investigates three methods of increasing α2σ beyond bulk limits in V2VI3 and III-V semiconductors. First principles density functional theory (DFT) calculations are used to probe the effects of hydrostatic-pressure-induced lattice strain on the electronic structure of V2VI3 materials. The results, when coupled with Boltzmann transport analysis, show that strain can affect increases in maximum room-temperature α2σ by 20% in p-BiSbTe3 and by over 200% in n-Bi2Te3 and n-Bi2Te2Se, in agreement with experiment. This is due to favorable electronic structure alterations which increase band-edge quasi-degeneracy and thus significantly increase the electronic density-of-states (DOS) near the Fermi level (EF).