Artificial Intelligence: Discovering a Material for Using “White Gold” Lithium in Batteries

Microsoft researchers used artificial intelligence and supercomputers; To reduce 32 million potential minerals to 18 within a week.

The screening process from inception to development of a working battery prototype took less than nine months, a process that would have taken more than two decades to implement using traditional laboratory research methods.

Both companies achieved this through advanced artificial intelligence and high-performance computing, which brings together large numbers of computers to solve complex scientific and mathematical tasks.

Jason Zander, Microsoft's executive vice president, told the BBC that one of the tech giants' missions is to “summarize 250 years of scientific innovation in the next 25 years.”

“We hope that technology like this will help us do that. I believe this is how this type of science will be done in the future,” he added.

Lithium is often referred to as “white gold” because of its market value and silver color. It's a key component in rechargeable batteries (lithium-ion batteries) that power everything from electric cars to smartphones.

The world could face shortages of the material by 2025 as demand for the metal rises and demand for electric vehicles increases, the International Energy Agency said.

According to the US Department of Energy, demand for lithium-ion batteries is expected to increase tenfold by 2030, so manufacturers continue to build battery factories.

Lithium mining can be controversial; It takes years to develop and has a significant impact on the environment. Extracting the mineral requires large amounts of water and energy, and the process can leave impacts and toxic waste on the landscape.

Leading a group of battery researchers at Imperial College London's Chemistry Department, Dr. Nuria Tapia Ruiz says any material with a small amount of lithium and good energy storage capabilities is the “holy grail” of lithium production. -ion ​​batteries.

He added: “Artificial intelligence and supercomputing will become important tools to help battery researchers predict new high-performance materials.”

But Dr Edward Friedman, a lecturer in chemical engineering at the University of Strathclyde, says the technology needs to be “cautioned”.

“This can lead to spurious results, or something that looks good initially but then turns into a known substance or something that can't be done in the lab,” he explains.

The AI-derived material, now known as N2116, is a solid-state electrolyte (electrolyte is a chemical medium that conducts electricity) tested by scientists who turned it from a raw material into a working prototype.

This material can be a sustainable solution for energy conservation. Because solid-state batteries are safer than traditional liquid or gel-like lithium.

In the future, fast-charging solid-state lithium batteries will be more energy efficient with thousands of charging cycles.

The way this technology works is by using a new type of artificial intelligence developed by Microsoft, trained on molecular data that can actually detect chemicals.

“This artificial intelligence relies entirely on scientific materials, databases and properties,” explained Microsoft executive vice president Zander, “and the data is very reliable for use in scientific discoveries.”

After the software narrowed down the candidate materials to eighteen, battery experts at the Pacific Northwest National Laboratory (PNNL) examined them and selected the final material to work on in the lab.

PNNL's Carl Muller said Microsoft's AI led them “to a much faster area of ​​productivity” than traditional business conditions.

He added: “We can modify, test and adjust the chemical composition of this new material and rapidly evaluate its technical suitability to become a working battery, showing the potential of advanced artificial intelligence in accelerating the innovation cycle.”