Search for new Zintl phases for thermoelectricity

Abstract

In order to find promising candidates for materials with a high thermoelectric (TE) figure of merit, various Zintl phases were designed or chosen from literature and synthesised by melting elements together. These included Yb2SnGe2, Yb2InGe2, Sr2PbGe2, Sm4Bi2Te, Yb4Sb2S and SmGa6Te10. Depending on the composition, they were synthesised either directly in an open cold boat or in a sealed tantalum crucible. For the latter, an induction furnace was specifically designed. The synthesised polycrystals were analysed at room temperature (RT), using Xray powder diffraction (XRD) and, in some cases, SEM/EDX analysis. From these data, phase relations were analysed. If the sample's quality was deemed good enough, the TE properties were measured. This involved measurements of the electrical resistivity (2 K - 870 K), as well as high temperature measurements of the Seebeck coeffcient (RT - 870 K) and the thermal conductivity (420 K - 820 K). Additionally, the Hall coeffcient was measured at low temperatures (2 K - RT), with a magnetic field of 9 T. Both Yb2SnGe2 and Yb2InGe2 display typical metallic behaviour, with a linearly rising electrical resistivity, a high charge carrier concentration and small Seebeck coeffcient values. Together with a high thermal conductivity, these TE properties suggest that neither of these two compounds is a promising thermoelectric material and the chances of improving them through doping are very small. The other materials could not be obtained in a single-phase form. Sr2PbGe2 samples, synthesised both in the cold boat and in a sealed Ta crucible at 1200 °C, always contained free Pb, which did not disappear after annealing. The synthesis of the charge compensated Sm4Bi2Te in a sealed Ta crucible at 1500 °C resulted in a sample that contained both Sm4Bi2Te and SmTe. Annealing could not eliminate the foreign SmTe phase. An attempt to synthesise the charge compensated Yb4Sb2S in a sealed Ta crucible at 1600 °C resulted in a Yb4Sb2:6S0:4 matrix with YbS inclusions. Annealing did not improve the quality of the sample. The results indicate a limited incorporation of the chalcogens Te and S into anti-Th3P4 structure-type pnictogen phases. Finally, the attempt of synthesising SmGa6Te10 in a sealed Ta crucible at 1500 °C produced mainly a mixture of GaTe and Ga2Te3. These materials do not seem to be obtainable easily by simple melt synthesis. For this reason, and because some of them contain toxic components, they were not investigated in the framework of the current master's thesis..