Download file (195KB)
Download as zip (190KB)

コレクション

引用

説明:

The design of thermal insulating coatings for advanced energy and electronic applications requires a deep understanding of both intrinsic and interfacial thermal transport. In this study, we present a data-driven workflow that integrates interfacial thermal resistance (ITR) modeling, low thermal conductivity screening, and inverse materials design for thin-film applications.
We first introduce our previously developed ITR prediction framework, built upon a curated experimental ITR database [1] and machine learning models [2], which enables the estimation of ITR based on interfacial chemical and structural information. Using this framework, we designed thermal insulating thin films with ultra-low thermal conductivities [3,4].
Building on this foundation, we developed predictive models for bulk thermal conductivity using periodic descriptors [5], trained on inorganic compounds from the AtomWork-Adv. (AWA) database [6]. These compact and chemically meaningful descriptors enable efficient learning, achieving a prediction accuracy of R² ~0.8 for bulk thermal conductivity. Applying these models, we screened over 150,000 known materials in AWA and explored potential multi-element (>5) compositions for low thermal conductivity. While none of the screened materials fully satisfied practical constraints such as non-toxicity and low cost, the approach identified unexplored compositional systems worthy of further investigation.
Experimental validation of these candidates is currently underway in collaboration with thin-film fabrication teams. This integrated pipeline—combining ITR modeling, bulk thermal property prediction, and multi-element composition design—offers a promising pathway toward the development of next-generation thermal management materials.

権利情報:

• In Copyright
  

キーワード: interfacial thermal resistance, Thermal Insulating Thin Film, Data-Driven Design, Multi-Element, periodic descriptor

会議: TACT 2025 International Thin Films Conference (2025-10-26 - 2025-10-29)

研究助成金:

• Japan Science and Technology Agency (JST) CREST JPMJCR21O2
• Grant-in-Aid for Scientific Research (C) 24K07349
• Japan Society for the Promotion of Science (JSPS) 25K08337

原稿種別: 論文以外のデータ

MDR DOI: https://doi.org/10.48505/nims.6165

公開URL:

関連資料:

その他の識別子:

連絡先:

更新時刻: 2026-01-26 16:30:07 +0900

MDRでの公開時刻: 2026-01-26 13:08:59 +0900