Ab-Initio Accuracy at Unprecedented Length and Time Scales for Lithium Metal Simulations using Machine Learning

The mechanical properties of lithium metal are key parameters in the design of next generation lithium metal batteries. They are difficult to probe experimentally due to the high reactivity and low melting point of lithium as well as the microscopic scales at which lithium exists in batteries where it is found to have enhanced strength, with implications for dendrite suppression strategies. Computationally, there is a lack of good empirical potentials and Ab-Initio calculations are too costly. In this work, we generate data and train two Machine Learning Interaction Potentials (MLIPs) with Density Functional Theory (DFT) data to state-of-the-art accuracy in reproducing experimental results. Using Molecular Dynamics (MD), we predict a number of thermal and mechanical properties of lithium and their temperature dependence.