Âé¶¹ÒùÔº


Neural networks unlock potential of high-entropy carbonitrides in extreme environments

New research used neural networks to study melting of high-entropy carbonitrides
Simulated crystal structures of high-entropy carbides and carbonitrides (TiZrTaHfNb)C0.75N0.25 at 3000, 3500, 4000, and 4500 K (liquid) respectively. In the insets the non-metal sublattices are shown. Carbon atoms are shown by gray, while nitrogen is red. Credit: Alexander Kvashnin et al.

The melting point is one of the most important measurements of material properties, which informs potential applications of materials in various fields. Experimental measurement of the melting point is complex and expensive, but computational methods could help achieve an equally accurate result more quickly and easily.

A research group from Skoltech conducted a study to calculate the maximum of a high-entropy carbonitrides—a compound of titanium, zirconium, tantalum, hafnium, and niobium with carbon and nitrogen.

The published in the Scientific Reports journal indicate that high-entropy carbonitrides can be used as promising materials for protective coatings of equipment operating under —high temperature, thermal shock, and chemical corrosion.

"In the new study, we used deep neural network-based potentials of interatomic interaction to model the structure of high-entropy carbonitride (TiZrTaHfNb)CxN1−x in both solid and liquid states. This allowed us to predict the heating and cooling temperatures depending on the nitrogen content, determine the melting point, and analyze the structure-property relationship in terms of interatomic interactions.

An increase in the leads to an increase in the melting point, which is associated with a change in the relative stability of the liquid phase compared to the solid phase when nitrogen is added," commented Professor Alexander Kvashnin from the Skoltech Energy Transition Center and the study supervisor.

The team has created a new approach for training the DeepMD potential to mimic the melting and crystallization processes of the TiZrTaHfNbCxN1-x, which enabled them to calculate its melting point.

The potential was trained on atomic trajectories obtained by molecular dynamics, which ensured high accuracy of predictions in atomic forces and energies.

The approach aimed at expanding the capabilities of classical modeling, which allows for accurate modeling and analysis of the melting process with prediction of the melting temperature not only of high-entropy carbonitrides, but also of other complex multicomponent materials.

The authors identified the maximum melting point for the composition (TiZrTaHfNb)C0.75N0.25—3,580±30 K. By adding nitrogen, the melting characteristics of high-entropy compounds can be improved, while the thermophysical properties of functional and structural materials will change.

More information: Viktor S. Baidyshev et al, Melting simulations of high-entropy carbonitrides by deep learning potentials, Scientific Reports (2024).

Journal information: Scientific Reports

Citation: Neural networks unlock potential of high-entropy carbonitrides in extreme environments (2024, December 20) retrieved 16 June 2025 from /news/2024-12-neural-networks-potential-high-entropy.html
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.

Explore further


26 shares

Feedback to editors