Chain-forming zintl antimonides as novel thermoelectric materials

<p>Zintl phases, a subset of intermetallic compounds characterized by covalently-bonded "sub-structures," surrounded by highly electropositive cations, exhibit precisely the characteristics desired for thermoelectric applications. The requirement that Zintl compounds satisfy the valence of anions through the formation of covalent substructures leads to many unique, complex crystal structures. Such complexity often leads to exceptionally low lattice thermal conductivity due to the containment of heat in low velocity optical modes in the phonon dispersion. To date, excellent thermoelectric properties have been demonstrated in several Zintl compounds. However, compared with the large number of known Zintl phases, very few have been investigated as thermoelectric materials.</p> <p>From this pool of uninvestigated compounds, we selected a class of Zintl antimonides that share a common structural motif: anionic moieties resembling infinite chains of linked <i>M</i>Sb₄ tetrahedra, where $M$ is a triel element. The compounds discussed in this thesis (<italic>A</italic>₅<italic>M</italic>₂Sb₆ and <i>A</i>₃<i>M</i>Sb₃, where <i>A</i> = Ca or Sr and <i>M</i> = Al, Ga and In) crystallize as four distinct, but closely related "chain-forming" structure types. This thesis describes the thermoelectric characterization and optimization of these phases, and explores the influence of their chemistry and structure on the thermal and electronic transport properties. Due to their large unit cells, each compound exhibits exceptionally low lattice thermal conductivity (0.4 - 0.6 W/mK at 1000 K), approaching the predicted glassy minimum at high temperatures. A combination of Density Functional calculations and classical transport models were used to explain the experimentally observed electronic transport properties of each compound. Consistent with the Zintl electron counting formalism, <italic>A</italic>₅<italic>M</italic>₂Sb₆ and <i>A</i>₃<i>M</i>Sb₃ phases were found to have filled valence bands and exhibit intrinsic electronic properties. Doping with divalent transition metals (Zn²⁺ and Mn²⁺) on the <i>M</i>³⁺ site, or Na¹⁺ on the <i>A</i>³⁺ site allowed for rational control of the carrier concentration and a transition towards degenerate semiconducting behavior. In optimally-doped samples, promising peak <i>zT</i> values between 0.4 and 0.9 were obtained, highlighting the value of continued investigations of complex Zintl phases.</p>