Linking structure and process in dendritic growth using persistent homology with energy analysis

We present a material analysis method that links structure and process in dendritic growth using explainable machine learning approaches. We employed persistent homology (PH) to quantitatively characterize the morphology of dendritic microstructures. By using interpretable machine learning with ener...

Full description

Saved in:
Bibliographic Details
Main Authors: Misato Tone, Shunsuke Sato, Sotaro Kunii, Ippei Obayashi, Yasuaki Hiraoka, Yui Ogawa, Hirokazu Fukidome, Alexandre Lira Foggiatto, Chiharu Mitsumata, Ryunosuke Nagaoka, Arpita Varadwaj, Iwao Matsuda, Masato Kotsugi
Format: Article
Language:English
Published: Taylor & Francis Group 2025-12-01
Series:Science and Technology of Advanced Materials: Methods
Subjects:
Online Access:https://www.tandfonline.com/doi/10.1080/27660400.2025.2475735
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We present a material analysis method that links structure and process in dendritic growth using explainable machine learning approaches. We employed persistent homology (PH) to quantitatively characterize the morphology of dendritic microstructures. By using interpretable machine learning with energy analysis, we established a robust relationship between structural features and Gibbs free energy. Through a detailed analysis of how Gibbs free energy evolves with morphological changes in dendrites, we uncovered specific conditions that influence the branching of dendritic structures. Moreover, energy gradient analysis based on morphological feature provides a deeper understanding of the branching mechanisms and offers a pathway to optimize thin-film growth processes. Integrating topology and free energy enables the optimization of a range of materials from fundamental research to practical applications.
ISSN:2766-0400