Machine Learning Assisted Design of Catalysts for Metal-Air Batteries
SandBox is developing technologies which use DFT and proprietary machine learning and artificial intelligence algorithms to help its customers narrow the catalysts for evaluation by experiment and accelerate their discovery of new catalyst materials.
Metal-air batteries are attractive for long-ranged electrical vehicles and in-place energy storage due to their high operating voltages and energy densities. Because they exchange oxygen from the atmosphere, their weight and size are significantly reduced. Electrochemical catalysts for the oxygen reduction reaction (ORR) and oxygen evolution reaction (OER) are critical for fast charge and discharge rates, low overpotentials, and good charge/discharge efficiencies because of the naturally slow kinetics of these reactions. Currently, Pt is the highest performing catalyst, but it is very expensive, is not sustainable for broad applications, and has some long-term stability deficiencies. SandBox identifies low cost, high activity catalysts for ORR and OER using Density Functional Theory and machine learning to bring this energy storage and delivery technology into broad use. SandBox is developing technologies that use proprietary machine learning-enhanced approaches to construct databases containing asystematic ranking of potential catalysts for ORR and OER.
The development of catalysts to enable metal-air batteries will usher in a new era energy storage for electric vehicles (EVs) and in-place storage for intermittent wind and solar sources. These batteries will enable EVs to travel up to 10 times further than with current batteries and make in-home storage for solar efficient. There are also many military applications for these higher energy density batteries.