Cornell University researchers have combined machine learning with powerful X-rays to unravel the mystery behind the unusual behavior of a certain material that could find use in thermoelectric energy conversion and other applications.
Researchers have long shown that cubic-phase germanium telluride (GeTe) exhibits an unexpected increase in lattice thermal conductivity as temperature is increased, but knowing that the property exists is one thing, and explaining why it exists is quite another.
“The new technology is a game changer,” said Zhiting Tian, ​​associate professor in the Sibley Department of Mechanical and Aerospace Engineering at Cornell University's College of Engineering. Nature Communications This provides a plausible explanation for GeTe's unexpected behavior, and more generally, the work improves researchers' understanding of heat transport in a class of materials known as phase-change materials.
What Tian and her team discovered was that when a sample of GeTe was heated until it changed phase from a rhombohedral to a cubic structure, the bonds between the next-nearest neighbors of the same atom (Ge-Ge and Te-Te) were significantly strengthened: When the sample's temperature was increased from 693 to 850 Kelvin, the strength of the Ge-Ge bond increased by 8.3%, and the strength of the Te-Te bond increased by an astonishing 103%.
Using machine learning-assisted first-principles calculations supported by X-ray scattering measurements, the research group computationally reproduced for the first time the trend in increasing thermal conductivity. They then borrowed commonly used chemistry techniques to perform a bonding analysis, confirming that these increasingly stronger second-nearest neighbor bonds play a key role in the previously unexplained increase in the lattice thermal conductivity of GeTe.
“Computationally, it was very challenging to explore the effect of temperature and, for example, to take into account higher-order scattering,” Tian says, “But we were able to harness the potential of machine learning, which allowed us to extract the interactions more efficiently and take into account multiple effects at once, including temperature dependence, four-phonon scattering, and coherence contributions.”
Phase change materials such as GeTe are valued for their usefulness in a variety of optical and electronic applications: the optical and electrical properties of these materials change dramatically depending on which of several stable phase states they are in, and these phase states can be easily reversed.
Additionally, GeTe has been explored as an alternative to the semiconductor lead telluride as a thermoelectric material because of lead's inherent toxicity, according to Tian, ​​a faculty fellow at Cornell University's Atkinson Center for Sustainability.
This work demonstrates an efficient and thorough path toward accurate modeling of materials near phase transitions or at high temperatures expected for phase change, thermoelectric, and other energy applications.
“We also identified other materials, such as tin telluride and tin selenide, that showed similar conductivity increases,” Tian said, “so we hope that our work will stimulate interest in exploring the thermal transport behavior of other phase-change materials in greater depth.”
More information:
Samuel Killer et al. “Anomalous lattice thermal conductivity increase with temperature in cubic GeTe correlates with enhanced second nearest neighbor bonding.” Nature Communications (2024). DOI: 10.1038/s41467-024-51377-8
Courtesy of Cornell University
Citation: Machine learning explains material's unexpected thermal conductivity (August 22, 2024) Retrieved August 25, 2024 from https://phys.org/news/2024-08-machine-material-unexpected-thermal.html
This document is subject to copyright. It may not be reproduced without written permission, except for fair dealing for the purposes of personal study or research. The content is provided for informational purposes only.