Exploring the structural stability, thermal and mechanical properties of nanoporous carbon nitride nanosheets using a transferrable machine learning interatomic potential

authored by: Bohayra Mortazavi, Rabczuk Timon, Xiaoying Zhuang
Abstract: Nanoporous carbon nitride nanosheets (NPCNNs) currently stand as one of the most promising classes of two-dimensional materials, exhibiting exceptional properties, wide range of applications and continuous experimental realization of novel structures. However, from a theoretical standpoint, detecting their most stable configurations and accurately evaluating their mechanical and thermal properties are challenging. To effectively address these obstacles, here for the first time we have developed a transferable machine learning interatomic potential (MLIP), capable of substantially facilitating the detection of dynamically stable configurations. Through comprehensive analysis of the mechanical properties of diverse NPCNNs, it is shown that the developed model could with negligible computational cost and remarkable accuracy reproduce directional-dependent stress–strain curves and failure mechanisms predicted by density functional theory calculations. It is furthermore shown that the developed model could conveniently detect the complex corrugated stable configurations, and moreover highlight that the resulting mechanical and electronic properties might be considerably different from their dynamically unstable flat counterparts. The developed transferable MLIP can thus offer an accurate and fast solution to explore the structural stability, and mechanical responses at ground state, and moreover investigate the temperature dependent properties of the NPCNNs, thus substantially facilitating deeper insights into their physical properties and potential applications.
Organisation(s): PhoenixD: Photonics, Optics, and Engineering - Innovation Across Disciplines
Institute of Photonics
External Organisation(s): Bauhaus-Universität Weimar
Tongji University
Type: Article
Journal: Machine learning for computational science and engineering
Volume: 2025
No. of pages: 11
ISSN: 3005-1428
Publication date: 26.11.2024
Publication status: Published
Peer reviewed: Yes
Electronic version(s): https://doi.org/10.1007/s44379-024-00008-6 (Access: Open)
: Details in the research portal "Research@Leibniz University"

BibTeX

@article{d4aae261f4a74ab486a4b31b553f95b4,
title = "Exploring the structural stability, thermal and mechanical properties of nanoporous carbon nitride nanosheets using a transferrable machine learning interatomic potential",
abstract = "Nanoporous carbon nitride nanosheets (NPCNNs) currently stand as one of the most promising classes of two-dimensional materials, exhibiting exceptional properties, wide range of applications and continuous experimental realization of novel structures. However, from a theoretical standpoint, detecting their most stable configurations and accurately evaluating their mechanical and thermal properties are challenging. To effectively address these obstacles, here for the first time we have developed a transferable machine learning interatomic potential (MLIP), capable of substantially facilitating the detection of dynamically stable configurations. Through comprehensive analysis of the mechanical properties of diverse NPCNNs, it is shown that the developed model could with negligible computational cost and remarkable accuracy reproduce directional-dependent stress–strain curves and failure mechanisms predicted by density functional theory calculations. It is furthermore shown that the developed model could conveniently detect the complex corrugated stable configurations, and moreover highlight that the resulting mechanical and electronic properties might be considerably different from their dynamically unstable flat counterparts. The developed transferable MLIP can thus offer an accurate and fast solution to explore the structural stability, and mechanical responses at ground state, and moreover investigate the temperature dependent properties of the NPCNNs, thus substantially facilitating deeper insights into their physical properties and potential applications.",
author = "Bohayra Mortazavi and Rabczuk Timon and Xiaoying Zhuang",
year = "2024",
month = nov,
day = "26",
doi = "10.1007/s44379-024-00008-6",
language = "English",
volume = "2025",
journal = "Machine learning for computational science and engineering ",
issn = "3005-1428",
publisher = "Springer International Publishing",
number = "1",
}

Back