| Abstract Scope |
Ultrahigh-temperature processing and manufacturing require materials that remain stable under extreme thermal conditions, yet reliable melting-temperature data are sparse for many refractory ceramics, alloys, and multicomponent compounds. I will present a computational strategy for identifying and ranking candidate highest-melting materials using SLUSCHI-UP, a public atomistic melting-temperature platform that combines small-size solid–liquid coexistence simulations with universal machine-learning interatomic potentials. Candidate materials are screened first using rapid graph-neural-network melting-temperature predictions and then evaluated with SLUSCHI-UP coexistence simulations for physics-based validation. The resulting highest-melting-temperature leaderboard includes refractory carbides, nitrides, borides, oxides, metals, and multicomponent compositions relevant to ultrahigh-temperature synthesis and manufacturing. By coupling automated melting simulations with the MeltBench validation framework, this approach provides repeatable melting-temperature estimates, identifies promising unexplored chemical spaces, and supports data-driven selection of materials for extreme-temperature processing, aerospace, defense, and energy applications. |