Neural networks unlock the potential of high-entropy carbonitrides in extreme environments
The crystal structures of high-entropy carbide and carbonitride (TiZrTaHfNb)C0.75N0.25 were simulated at 3000, 3500, 4000, and 4500 K (liquid), respectively. The inset shows the nonmetallic sublattice. Carbon atoms are shown in gray and nitrogen in red. Credit: Alexander Kvashnin et al.
Melting point is one of the most important measurements of material properties and informs the potential applications of materials in various fields. Experimental measurements of melting point are complex and expensive, but computational methods can provide equally accurate results more quickly and easily.
A research group at Skoltec conducted a study to calculate the maximum melting points of high-entropy carbonitrides (compounds of titanium, zirconium, tantalum, hafnium, and niobium with carbon and nitrogen).
The results, published in the journal Scientific Reports, show that high-entropy carbonitrides can be used as promising materials in protective coatings for equipment operating under extreme conditions such as high temperatures, thermal shock, and chemical corrosion.
“In a new study, we modeled the structure of the high-entropy carbonitride (TiZrTaHfNb)CxN1−x, both solid and liquid, using atomic interaction potentials based on deep neural networks. We were able to predict the temperature and cooling temperature. We determined the melting point based on the nitrogen content and analyzed the relationship between structure and properties from the perspective of atomic interactions.
An increase in nitrogen content leads to an increase in the melting point, and the addition of nitrogen is associated with a change in the relative stability of the liquid phase compared to the solid phase,” says Alexander Kvashnin of the Skoltech Energy Transition Center. the professor commented. and research supervisor.
The research team created a new approach to train DeepMD’s potential to mimic the melting and crystallization process of TiZrTaHfNbCxN1-x, allowing it to calculate the melting point.
The potentials were trained based on atomic trajectories obtained by ab initio molecular dynamics, ensuring high prediction accuracy for nuclear power and energy.
This approach aims to extend the capabilities of classical molecular dynamics modeling and provides accurate modeling of melting processes through the prediction of melting temperatures not only for high-entropy carbonitrides, but also for other complex multicomponent materials. and analysis.
The authors identified a maximum melting point of composition (TiZrTaHfNb)C0.75N0.25 – 3,580±30 K. The addition of nitrogen improves the melting properties of high-entropy compounds and improves the thermophysical properties of functional and structural materials. change.
Further information: Viktor S. Baidyshev et al, Simulation of high-entropy carbonitride melting with deep learning potentials, Scientific Reports (2024). DOI: 10.1038/s41598-024-78377-4
Provided by Skolkovo University of Science and Technology
Citation: Neural networks unlock high-entropy carbonitride potential in extreme environments (December 20, 2024) https://phys.org/news/2024-12-neural-networks-potential-high-entropy. Retrieved December 22, 2024 from html
This document is subject to copyright. No part may be reproduced without written permission, except in fair dealing for personal study or research purposes. Content is provided for informational purposes only.
