SKM 2023 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

TUT: Tutorien

TUT 4: Strategic elements and sustainability (joint session MA/TUT)

TUT 4.2: Hauptvortrag

Sonntag, 26. März 2023, 16:30–17:00, HSZ 04

Sustainable Thermoelectric Materials Predicted by Data Mining and Machine Learning — •Kornelius Nielsch — Leibniz Institute of Solid States and Materials Research, Dresden, Germany — Institute of Materials Research at TU Dresden, Germany — Institute of Applied Physcis at TU Dresden, Germany

Generating electricity from temperature differences has proven itself in space. Thanks to this technology, the Voyager probes launched in 1977 are still sending signals today. In the meantime, the car industry and ship producers have become interested in thermoelectrics. The combustion of fossil fuels produces exhaust gas that is up to 1300 °C hot. Modern thermoelectric materials are continuously expanding the fields of thermoelectric applications. The experimental search for new thermoelectric materials remains largely restricted to a limited number of successful chemical and structural families, such as chalcogenides, skutterudites and zintl phases. In principle, computational tools such as density functional theory (DFT) offer the possibility of directing experimental synthesis efforts towards very different chemical structures. In practice, however, predicting thermoelectric properties based on first principles remains a difficult endeavour, and experimental researchers do not usually use computations directly to drive their own synthesis efforts. Strategies to bridge this practical gap between experimental requirements and computational tools will be discuss und presented in this tutorial talk. Ref: Energy Environ. Sci. 14, 3559 (2021) and Advanced Theory and Simulations 5, 2200351 (2022)