Accelerated Discovery of Battery Materials

More powerful and affordable electrochemical energy storage systems are key to accelerating the adoption of renewable energy and zero-emission electric vehicles (EV). While much of the electrochemical storage demand is currently satisfied by lithium-ion batteries (LIBs), the gradual saturation in achievable battery capacity, high cost, environmental impacts of critical raw materials, and increasing safety concerns have spurred growing interest in new battery materials and chemistries.

At IBM Research Almaden, we are developing a more powerful and sustainable battery, working in collaboration between materials discovery and advanced computing technology research. We have combined conventional and AI-assisted scientific methods to develop a brand-new battery chemistry without using costly heavy metals like cobalt and nickel, which have serious environmental and health impacts. Our heavy metal free battery based on more sustainable cathode materials (like iodine extracted from brine) has shown an ability to surpass conventional LIBs in a number of individual categories including high power and faster charging time.

To speed the development process, we are now using an AI-assisted workflow to discover safer and higher-performing electrolyte materials. The electrolyte discovery workflow includes high throughput screening of component materials based on their physico-chemical properties obtained by automated quantum chemical simulations built on ST4SD. Novel deep learning models then map the structure-composition-performance relationship of new electrolyte formulations, proposing the most promising candidates. The proposed electrolyte formulations can then be further optimized based on subsequent characterization and prediction of the solid-electrolyte interphase (SEI) layer. Implementation of this AI workflow and advanced computing capabilities helps to reduce the required volume of experiments needed to find safer (less flammable) and higher performing electrolyte systems, greatly speeding up the pace of materials discovery. Our team is also exploring how quantum computing could play a key role in this process to accurately simulate complex molecular interactions and ionic conduction behaviors. To spur continued innovation and ensure viable application, we have partnered with Mercedes-Benz Research & Development North America, battery electrolyte supplier Central Glass, and battery manufacturer Sidus Energy. Collaborating on the research and development process will allow us to validate the AI-workflow, the use of quantum computers and other tools for the discovery of sustainable battery materials.