Modeling Phase Separation in Polymer-Derived Silicon Carbonitride Ceramics through Extended Machine Learning Molecular Dynamics
5 affiliations
- 1Institut de Recherche sur les Céramiques, UMR 7315 CNRS-Université de Limoges, Centre Européen de la Céramique, 12 rue Atlantis, 87068Limoges, France.
- 2University of Strasbourg, CNRS, ICube Laboratory UMR 7357, F-67412Illkirch, France.
- 3ADYNMAT CNRS consortium, F-67034Strasbourg, France.
- 4Université de Strasbourg, CNRS, Institut de Physique et Chimie de des Matériaux de Strasbourg, UMR 7504, F-67034Strasbourg, France.
- 5School of Environment and Science, Griffith University, Gold Coast4222, Australia.
Abstract
Polymer-derived ceramics combine the thermal stability of ceramics with the versatile properties of carbon domains, but modeling their atomic-scale evolution during processing remains elusive because of the limitations of traditional computational methods. To address this issue, here, we develop and apply a machine learning interatomic potential for hydrogen-containing silicon carbonitride-based (SiCNH) ceramics, trained on a diversified database of over 9000 configurations - including amorphous models, hightemperature states, surfaces, and crystal structure predic-tions�to capture the full complexity of these materials. This potential enables large-scale molecular dynamics simulations of 8000-atom systems, revealing the atomic-scale evolution of the polymer-derived ceramic during thermal treatment. A key result of this work is the occurrence of a phase separation, where carbon domains progressively nucleate from the amorphous SiCN matrix during thermal processing, forming distinct graphene-like sheets while preserving the integrity of the ceramic network. The resulting models reproduce the experimental atomic pair distribution functions, supporting our approach and providing microscopic explanations for the material combination of the ceramic and graphitic properties. In this process, defective 5-and/or 7-member carbon rings mediate the transformation to stable 6-member aromatic structures. These findings provide atomic-scale information about the thermal stability and structural transformation pathways of polymer-derived ceramics and demonstrate the applicability of the methodology to complex amorphous systems at experimentally relevant scales.
Related
- Modeling Reactions on the Solid-Liquid Interface with Next Generation Extended Lagrangian Quantum-Based Molecular Dynamics.
- Atomic-scale assessment of the crystallization onset in silicon carbonitride
- Molecular origin of pressure-induced re-entrant phase separation of CO₂–H₂O mixtures
- Is it possible to achieve double charge separation with a single photon? DFT insights on a donor–acceptor molecular dyad with double potential inversion
- Single-Source Precursor Synthesis of a Compositionally Complex Early Transitional Metal Carbonitride (Ti,Zr,Hf,Nb,Ta)NxC1−x
BibTeX
@article{mortier2026modeling,
title = {{Modeling Phase Separation in Polymer-Derived Silicon Carbonitride Ceramics through Extended Machine Learning Molecular Dynamics}},
author = {Fabien Mortier and Sylvian Cadars and Olivier Masson and Mauro Boero and Guido Ori and Yun Wang and Samuel Bernard and Assil Bouzid},
journal = {ACS Nano},
year = {2026},
month = oct,
doi = {10.1021/acsnano.6c11958},
url = {https://hal.science/hal-05783629v1},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Modeling Phase Separation in Polymer-Derived Silicon Carbonitride Ceramics through Extended Machine Learning Molecular Dynamics
AU - Fabien Mortier
AU - Sylvian Cadars
AU - Olivier Masson
AU - Mauro Boero
AU - Guido Ori
AU - Yun Wang
AU - Samuel Bernard
AU - Assil Bouzid
PY - 2026
DA - 2026/10/01
JO - ACS Nano
DO - 10.1021/acsnano.6c11958
UR - https://hal.science/hal-05783629v1
AB - Polymer-derived ceramics combine the thermal stability of ceramics with the versatile properties of carbon domains, but modeling their atomic-scale evolution during processing remains elusive because of the limitations of traditional computational methods. To address this issue, here, we develop and apply a machine learning interatomic potential for hydrogen-containing silicon carbonitride-based (SiCNH) ceramics, trained on a diversified database of over 9000 configurations - including amorphous models, hightemperature states, surfaces, and crystal structure predic-tions�to capture the full complexity of these materials. This potential enables large-scale molecular dynamics simulations of 8000-atom systems, revealing the atomic-scale evolution of the polymer-derived ceramic during thermal treatment. A key result of this work is the occurrence of a phase separation, where carbon domains progressively nucleate from the amorphous SiCN matrix during thermal processing, forming distinct graphene-like sheets while preserving the integrity of the ceramic network. The resulting models reproduce the experimental atomic pair distribution functions, supporting our approach and providing microscopic explanations for the material combination of the ceramic and graphitic properties. In this process, defective 5-and/or 7-member carbon rings mediate the transformation to stable 6-member aromatic structures. These findings provide atomic-scale information about the thermal stability and structural transformation pathways of polymer-derived ceramics and demonstrate the applicability of the methodology to complex amorphous systems at experimentally relevant scales.
SN - 1936-0851
SN - 1936-086X
ER -