A materials approach to improving the efficiency of thermoelectric cooling devices (Chalcogenide)

Solid state thermoelectric devices are not yet efficient enough to replace conventional cooling technologies. Device efficiency will only improve with the development of better materials, as cooling is caused by simply flowing current through material junctions. Efficiency increases with each material's thermoelectric figure of merit (Z), which is proportional to the square of its thermopower and inversely proportional to its electrical resistivity and thermal conductivity. Since these properties are interdependent, developing materials with a favorable balance of transport properties is a formidable task. This dissertation focuses on two classes of materials which could yield this opportune balance: high-symmetry, large-volume semiconductors and intermediate-valent (IV), Ce-containing, intermetallic compounds.; Theory indicates that a semiconductor's band extremum multiplicity near the Fermi energy (N_V) directly influences thermoelectric efficiency: complex, high-symmetry semiconductors are more likely to have a higher N_V, leading to a higher Z. Solid state reactants containing tetrahedral anions were used to form such large-volume, high-symmetry compounds. Efforts culminated in structural studies of two families of semiconductors which contain discrete tetrahedral [Ge(S,Se) ₄]4− anions: the new cubic Pb_{2− x}Sn_xS_{4− y}S_y solid solution, and the previously-reported hexagonal La₃CuGe(S,Se)₇ phases. These compounds are all brightly colored, suggesting that their carrier mobilities are too low for them to be good thermoelectric materials.; Empirical evidence shows that Z peaks for an optimal carrier density, indicating that this is another important property for potential thermoelectric materials. Room temperature carrier concentrations for CeSbTe (a poor metal) and the NdxCe_3−xPt ₃Sb₄ Kondo insulators are reported from Hall coefficient measurements.; Metal-like IV compounds have been studied extensively because they display anomalously-large thermopowers. This work describes: searches for new Ce-Co-(Sb, In, Bi, C) phases, screening for IV behavior in magnetic susceptibility and lattice parameter data of known rare-earth transition-metal carbides, and thermopower studies of the new CeNi_1−xCo _xAl₄, CeNiIn_4−xGa _x, and CeNiAl_1−xIn _x solid solutions (all isotypic to the known IV CeNiAl ₄). None of these compounds display thermopowers of sufficient magnitude to warrant further study.