Answers are grounded in this paper and cite their sources.
Machine learning classifies channel structures in lithium intermetallics, indicating potential for energy storage materials.
Lithium intermetallics with channel structures are of interest for energy storage applications. As a major subset of these intermetallics, ternary tetrelides Li-M-Tt (M = metal; Tt = Si, Ge, Sn) were selected to apply machine learning approaches to predict whether they adopt channel vs nonchannel structures. Through the use of a conventional machine learning method (support vector classifier, SVC) and a more interpretable one (sure independence screening and sparsifying operator, SISSO), models were developed to perform this structural classification. By combining predictions of candidates based on these models with the feasibility of their synthesis based on estimated formation energies, two new series of lithium-containing rare-earth silicides were confirmed to adopt channel structures: LiRESi (RE = Pr, Nd, Tm, Lu) with the hexagonal ZrNiAl-type structure and LiRESi₂ (RE = Pr, Nd) with the orthorhombic LiCaSi₂-type structure.
No takes yet. Share an insight, caveat, or question.
Gvozdetskyi et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: