Abstract: Many challenges in material science are accompanied by relatively small data sets of thousands or fewer data points that pose unique challenges for statistical approaches. Collecting significantly more data is frequently not an option because of the time required or high expense per data point.

I will discuss our efforts to augment such small data sets with approximate physics-based models to perform holistic computational structure screening of more than 12,000 solid lithium-ion conductor material candidates for batteries. We discovered many new crystalline solid materials with fast single-crystal lithium-ion conductivity at room temperature through density functional theory simulations guided by machine learning (ML) methods. The discovery of new solid lithium superionic conductors is of critical importance to the development of safe all-solid-state lithium-ion batteries.

With a predictive universal structure−property relationship for fast-ion conduction not well understood, the search for new solid lithium-ion conductors has relied largely on trial-and-error computational and experimental searches over the last several decades. When compared with a random search of materials space, we find that the ML-guided search is 2.7 times more likely to identify fast lithium-ion conductors, with at least a 44 times improvement in the log-average of room-temperature lithium-ion conductivity. The F1 score of the ML-based model is 0.50, 3.5 times better than the F1 score expected from completely random guesswork.

In a head-to-head competition against six Ph.D. students working in the field, we find that the ML-based model doubles the F1 score of human experts in its ability to identify fast lithium-ion conductors from atomistic structure with a 1000-fold increase in speed, clearly demonstrating the utility of this model for the research community. In addition to having high predicted lithium-ion conductivity, all materials reported here lack transition metals to enhance stability against reduction by the lithium-metal anode and are predicted to exhibit low electronic conduction, high stability against oxidation, and high thermodynamic stability, making them promising candidates for solid-state electrolyte applications on these several essential metrics.

Bio: Evan Reed is a faculty member in Materials Science and Engineering at Stanford University.  He received a B.S. in applied physics from Caltech and Ph.D. in physics from the Massachusetts Institute of Technology.  His recent work focuses on theory and modeling of two-dimensional materials, statistical learning for chemical and energy storage applications, structural phase changes, and high-pressure shockwave compression.