English

An Experimentally Driven Automated Machine Learned lnter-Atomic Potential for a Refractory Oxide

Materials Science 2021-04-20 v1 Machine Learning Computational Physics

Abstract

Understanding the structure and properties of refractory oxides are critical for high temperature applications. In this work, a combined experimental and simulation approach uses an automated closed loop via an active-learner, which is initialized by X-ray and neutron diffraction measurements, and sequentially improves a machine-learning model until the experimentally predetermined phase space is covered. A multi-phase potential is generated for a canonical example of the archetypal refractory oxide, HfO2, by drawing a minimum number of training configurations from room temperature to the liquid state at ~2900oC. The method significantly reduces model development time and human effort.

Keywords

Cite

@article{arxiv.2009.04045,
  title  = {An Experimentally Driven Automated Machine Learned lnter-Atomic Potential for a Refractory Oxide},
  author = {Ganesh Sivaraman and Leighanne Gallington and Anand Narayanan Krishnamoorthy and Marius Stan and Gabor Csanyi and Alvaro Vazquez-Mayagoitia and Chris J. Benmore},
  journal= {arXiv preprint arXiv:2009.04045},
  year   = {2021}
}
R2 v1 2026-06-23T18:24:19.205Z