MACHINE LEARNING FORCE FIELD WITH CLASSICAL/QUANTUM MOLECULAR DYNAMICS

Machine Learning Force Field and Molecular Dynamics

Machine learning force fields offer near-quantum accuracy at a fraction of the computational cost, enabling simulations of large, realistic systems. They significantly accelerate molecular dynamics studies of complex battery materials, capturing essential atomistic behaviors over longer timescales.

Neural Network Potentials

Neural Network Potentials

Neural network-based machine learning models can be trained on atomic structures using DFT-derived energies, forces, and spins. This accelerates large-scale molecular dynamics simulations with quantum-chemical accuracy.

Our Publications:

  1. Hanzeng Guo, Volodymyr Koverga, Selva Chandrasekaran Selvaraj, Anh T. Ngo. Unveiling the Lithium-Ion Transport Mechanism in Li2ZrCl6 Solid-State Electrolyte via Deep Learning-Accelerated Molecular Dynamics Simulations. https://arxiv.org/abs/2508.05598
  2. Selva Chandrasekaran Selvaraj, Daiwei Wang, Donghai Wang, Anh T Ngo. Mechanism and Stability of Li-Dynamics in Amorphous Li-Ti-PS Based Mixed Ionic-Electronic Conductor. https://arxiv.org/abs/2506.11199
  3. Selva et. al., 2024, J. Electrochem. Soc. 171 050544  https://iopscience.iop.org/article/10.1149/1945-7111/ad4ac9/meta