AI Helps Scientists Create New Battery Type

By Quinn Peacock

In 2022, we saw our first major breakthrough of AI in mainstream society, with Chat GPT being released. Now, we are seeing how it is able to revolutionize the scientific community and our daily lives with its help in the new development of Lithium-ion batteries.
Every phone, laptop, electric car, and almost every other rechargeable electronic device uses a Lithium-ion battery. According to Apple, these batteries charge faster, last longer, and have a higher power density for more battery life in a lighter package than traditional batteries, such as the AA. However, they require significant resources and the production of Lithium can cause large amounts of land and water pollution.
The safest Lithium-ion batteries require something called a solid electrolyte, which is a material that transfers the electrically charged atoms (ions) between the battery electrodes.
While Lithium cannot be replaced, as it is the electrode, the material of the electrolyte can be changed to make the battery more efficient or safer. Unfortunately for researchers trying to find the best material to use, there are millions of combinations of materials that can work.
To solve this problem, The Pacific Northwest National Laboratory (PNNL) used Microsoft's Azure Quantum Elements tool to search for a better material to use for solid electrolytes. They used a “mix and match” system to create 32 million different combinations of electrolytes. Then, the software discarded materials that either couldn’t exist due to instability, or had the right conductive properties for the battery. Lastly, with some help from researchers, all toxic or rare materials were also eliminated.
This whole process took a total of 80 hours, which would have taken decades without AI. It narrowed 32 million possible combinations into the 23 best ones. Furthermore, 5 of the 23 combinations were already known to exist, so PNNL researchers picked one and tested it. They synthesized the material in the lab and tested it in a real battery, and it worked.
This new breakthrough is exciting for the whole scientific community, as it shows how much faster new advancements can be made with the help of AI. This leaves us with the standing question: when is the next major breakthrough coming?