AI detects new compound that could reduce lithium use in batteries by 70%

With further research and development, solid-state batteries have the potential to revolutionise the energy storage industry and accelerate the transition toward a greener future.
Used metal lithium alkaline batteries and lightbulbs on grass background

Microsoft and the Pacific Northwest National Laboratory (PNNL), part of the US Department of Energy, have made a significant discovery in battery technology using AI and supercomputing. The research has led to the identification of a new material that has the potential to reduce the reliance on lithium in batteries by up to 70%.

Lithium is vital in rechargeable batteries in electric vehicles, smartphones, and other devices. However, the mining and extraction processes for lithium have severe environmental implications. These processes require large amounts of water and energy, resulting in toxic waste and scars on the landscape. Consequently, battery researchers are actively seeking alternatives to address these concerns.

Microsoft researchers analyzed 32 million inorganic materials using AI and supercomputers in less than a week, narrowing the list to 18 promising candidates. This screening process would have taken over two decades using conventional lab research methods. The PNNL team then selected the most promising material, currently known as N2116, for further development in the lab.

N2116 is a solid-state electrolyte, offering several advantages over traditional liquid or gel-like lithium. Solid-state batteries are considered safer and have the potential for higher energy density. The material has already been successfully tested as a power source for a light bulb, demonstrating its viability as a battery component.

Why does it matter?

The significance of this discovery lies in its potential to address the increasing demand for lithium-ion batteries and alleviate concerns about lithium shortages. The International Energy Agency predicts a potential shortage by 2025, with demand expected to increase tenfold by 2030, as the US Department of Energy reported. These projections underscore the urgency of finding alternative materials for battery technology. The process, from initial discovery to developing a working battery prototype using N2116, took less than nine months. This rapid progress showcases the potential of AI and supercomputing to accelerate scientific discovery and technology development.