Unraveling the role of chemical composition in the lattice thermal conductivity of oxychalcogenides as thermoelectric materials

Abstract

Oxychalcogenides represent a large chemical space with potential application as thermoelectric materials due to their low thermal conductivity. However, the nature of this behaviour is still under debate. Understanding the...

Predicting lattice thermal conductivity via machine learning: a mini review

“…Last but not least, we note that several other MLPs were proposed recently, such as the spectral neighbor analysis potential, the bond order potential, the force constant potential, and the spatial density neural network force fields, which have been also demonstrated to accurately evaluate the κ L of many systems at low computational cost 19,[121][122][123][124][125][126][127] . With the deep understanding of interatomic interactions, it is reasonable to expect that more and more reliable and universal MLPs could be developed in the future.…”

A unique [Sb6O2S13]12−finite chain in oxychalcogenide Ba6Sb6O2S13leading to ultra-low thermal conductivity and giant birefringence

“…In addition, inspired by a recent study that metal chalcogenides containing SCALPs exhibit extremely low intrinsic thermal conductivities, 20 we investigated the thermal transmission performance of Ba 6 Sb 6 O 2 S 13 . As shown in Fig.…”

High-Throughput Prediction of the Thermal and Electronic Transport Properties of Large Physical and Chemical Spaces Accelerated by Machine Learning: Charting the ZT of Binary Skutterudites

“…The use of machine learning techniques in the calculation of interatomic force constants, IFCs, has drastically reduced by up to 2 orders of magnitude the computational cost of predicting the lattice thermal conductivity of solids using the Boltzmann transport equation, BTE, 20,21 opening the door to explore large chemical spaces. 22,23 On the other hand, new efficient methods have been reported for calculating carrier scattering rates of semiconductors and insulators from first principles without any external parameters. 24−26 Very recently, some authors have combined both methodologies using MLIP and AMSET packages to predict ZT.…”

“…In recent years, new approaches have been developed to reduce the computational cost or the need for experimental data without compromising the accuracy of the results. The use of machine learning techniques in the calculation of interatomic force constants, IFCs, has drastically reduced by up to 2 orders of magnitude the computational cost of predicting the lattice thermal conductivity of solids using the Boltzmann transport equation, BTE, , opening the door to explore large chemical spaces. , On the other hand, new efficient methods have been reported for calculating carrier scattering rates of semiconductors and insulators from first principles without any external parameters. − Very recently, some authors have combined both methodologies using MLIP and AMSET packages to predict ZT . , While MLIP develops moment tensor potentials to reduce the computational cost of high-order force constants, AMSET predicts the electron transport properties beyond RTA . Following this strategy, Yuan et al explored the TE performance of CsK 2 X (X = Sb, Bi), and Bai et al examined LaAgOX (X = S, Se).…”