Prediction of Mechanical Properties and Thermodynamic Stability of Ti-N system using MTP Interatomic Potential
Pradeep Kumar Rana, Atharva Vyawahare, Rohit Batra, Satyesh Kumar Yadav

TL;DR
This study develops a moment tensor potential for the Ti-N system, accurately predicting mechanical properties and stability across various compounds, validated against DFT calculations and enabling exploration of new stable phases.
Contribution
The paper introduces a reliable MTP interatomic potential for Ti-N, capable of predicting properties and stability of multiple compounds, including new phases, with high accuracy.
Findings
MTP potential predicts formation energies with RMSE of 2.1 meV/atom (training) and 6.8 meV/atom (testing).
Elastic constants trends are consistent with DFT benchmarks.
Structures with N/Ti ratios 0 to 1 can be thermodynamically stable.
Abstract
Ti-N material system have range of compounds with different stoichiometry like Ti2N, Ti3N2, Ti6N5, Ti4N3 alongwith Ti , TiN and solid solutions of N in Ti with a maximum of 23% solubility. In this work, we develop an interatomic potential based on moment tensor potential (MTP) that could reliably predict mechanical properties and thermodynamic stability of all Ti-N system. Taking into account the structural similarity and dissimilarity of various Ti-N system to choose training dataset was crucial for development of the potential. Root mean square error (RMSE) in prediction of formation energy using MTP potential compared to one calculated using density functional theory (DFT) for training dataset is 2.1 meV/atom and for testing dataset is 6.8 meV/atom. The frequency of absolute error in formation energy peaks at a maximum value of 3.8 meV/atom for system that was part of training…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsMachine Learning in Materials Science · X-ray Diffraction in Crystallography · Boron and Carbon Nanomaterials Research
