Data Efficient and Stability Indicated Sampling for Developing Reactive Machine Learning Potential to Achieve Ultralong Simulation in Lithium- Metal Batteries

LK Xu and W Shao and HS Jin and Q Wang, JOURNAL OF PHYSICAL CHEMISTRY C, 127, 24106-24117 (2023).

DOI: 10.1021/acs.jpcc.3c05522

Modeling the formation of the solid-liquid interphase (SEI) is challenging due to its strict requirements of both simulation accuracy and length. Machine learning potential (MLP)-based molecular dynamics (MD) simulation is expected to play a role in this field, while currently its use is hindered by sampling efficiency and simulation stability. In this work, we tackle these two challenges together. We propose the stability-indicated sampling (SIS) algorithm for efficiently sampling training data using physical information (temperature). Unlike previous strategies, our method does not need prior knowledge of reaction networks or training multiple MLPs for uncertainty estimation. Our approach shows superior sampling efficiency without requirements of the diversity of initial training data, to realize >10 ns MLPMD simulation using an ab initio MD (AIMD) trajectory of just a few ps. We introduce the concept underlying instability consistency by showing the accuracy of reaction mechanisms, and the radial distribution function (RDF) can be improved by SIS-MLPMD, although their information is not explicitly used in our sampling decision. Furthermore, we show that long-time MLPMD simulation of lithium-metal batteries (LMB) can reproduce not only some well-known SEI components, including LiF, Li2O, LiOH, LiS, and incomplete N-S bond breaking in high-concentration systems, but also ionic aggregation structures of LiF, which is not shown in our AIMD training data but matches previous experimental results. Our work is expected to help accelerate future investigations, especially for efficient sampling fine-tuning data for pretrained MLPs and for studying long-time (>= ns scale) reaction dynamics in interfacial problems.

Return to Publications page