Thermoelectric properties of mixed-valence chalcogenides

• Applicant: Professor Dr. Oliver Oeckler
• Subject Area: Solid State and Surface Chemistry, Material Synthesis ; Mineralogy, Petrology and Geochemistry
• Term: from 2015 to 2022
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 282777288

This project aims at new concepts to correlate the thermoelectric properties of mixed-valence transition-metal chalcogenides with their crystal structures, real structures and electronic structures. This includes the additional benefit of dynamically disordered copper atoms that reduce the thermal conductivity according to the "phonon-liquid electron-crystal" concept. Although thermoelectric materials can reversibly interconvert heat and electric energy, their efficiency must be significantly increased in order to use them e.g. for waste-heat harvesting. We focus on the synergism of empirical optimization with fundamental understanding and theoretical models for the prediction of thermoelectric properties. A broad range of diffraction, spectroscopic and microscopic (TEM) methods shall provide a reliable basis for advanced discussion. Compounds inspired by minerals such as Cu5FeS4 (bornite) with distorted antiflourite-type structure, CuFe2S3 (cubanite) with a wurtzite-related structure, chain structure like KFe2S3 (rasvumite) and structures with cation layers like Cu5.5FeS6.5 (nukundamite) offer unique possibilities of combining Cu-atom mobility at high temperatures with the effects of mixed valence states (in some cases after suitable substitutions). The structural variety is supplemented by copper bismuth sulfides with complex structures analogous to sulfosalts. The cheap and nontoxic sulfides are an attractive starting point; however, anion substitution by Se or Te is expected to enhance the electrical conductivity and to simultaneously reduce the thermal conductivity by introducing disorder. The variation of the synthesis conditions and further cation substitution leads to disorder on various length scales, including spin disorder due to the mixed valence of magnetic cations. In a new approach to combine structural and electronic disorder with tuning charge-carrier concentration and mobility, we aim at decoupling not only thermal from electrical conductivity, but also at an independent tuning of the Seebeck coefficient. The substitution of Cu by Ag and/or Li on cation sites allows separating the influence of mixed-valence effects of Cu from those of Fe. Oxidation states and chemical bonding in iron-containing compounds will be probed by Mößbauer spectroscopy, supplemented by EPR, XPS and susceptibility measurements which are also suitable for the other elements involved. In combination with DFT calculations, we expect a deep understanding of Cu-Cu and Fe-Cu interactions as well as spin disorder. The feedback of such results on the optimization of thermoelectric materials yields an intriguing interplay of fundamental research and materials science.

• DFG Programme: Research Grants