New battery material that uses less lithium found in AI-powered search

Microsoft announced Tuesday that a team of scientists used artificial intelligence and high-performance computing to engineer 32.6 million possible battery materials — many of which do not occur in nature — in 80 hours, a task the team said previously would have took 20 years. The results set the stage for an ambitious effort to create a new generation of batteries less dependent on toxic and environmentally harmful lithium.

The company shared some of the best candidates with the government’s Pacific Northwest National Laboratory in Richland, Washington, which studied the most promising and built a prototype battery using the brand new material.

While the dime-sized prototype isn’t yet ready for a prime-time role powering today’s watches and car keys, it functions using less lithium than commercially available options and has the ability to recharge. The feat also demonstrates the potential of new technologies to revolutionize an underappreciated but fast-growing field of material science.

“It’s really not an exaggeration to say that almost every major problem we face as a society can be, at least in part, helped by the availability of better materials,” said Christopher M. Wolverton, professor in the Department of Materials Science and Engineering at Northwestern University. Wolverton, who was not involved in Microsoft’s battery project but has worked on similar projects of his own, called the company’s claim that it viewed 32.6 million pieces of material in 80 hours “astonishing.”

“Thirty-two million tell me right away that they weren’t at the screening [just] known materials,” he added. “They were checking for hypothetical new materials that they hoped to discover.”

Like other outside experts interviewed, he was able to review news releases from Microsoft and the Pacific Northwest National Laboratory, but was unable to see the scientific paper describing the work before its release. The paper, which has not yet been peer-reviewed, was published Tuesday on the scientific preprints site arXiv.

Powerful mix and match

Although estimates vary, experts generally believe there are about 200,000 known materials in the world, and materials scientists continue to search for new ones that may hold the key to solving some of the planet’s serious challenges.

To combat climate change, scientists need to find the best materials to capture and store atmospheric carbon dioxide. To reduce the huge accumulations of plastic waste on beaches, clogging landfills and endangering human health, scientists must find safe ways to break it down into fuel and raw materials. To respond quickly to new pathogens capable of causing future pandemics, scientists must design effective drugs that use proteins to attack viruses and bacteria or stimulate the body’s defenses.

All depend on discovering the right materials, a process that is slowed by hypothesis-driven trial-and-error approaches.

High-performance computing and artificial intelligence now allow scientists to quickly use the elements in the periodic table the way an artist can use a palette of colors, mixing them and arriving at new configurations.

“This is a new way of doing science,” said Nathan Baker, the paper’s author and senior director of chemistry and materials partnerships at Microsoft. The project uses the company’s Azure Quantum Elements platform, which was unveiled in June to use advanced computing power to speed up the discovery process.

Starting with all the known materials, Baker said, “we can go through the periodic table and drop new atoms into [various] locations through a substitution process.’

Azure Quantum Elements quickly eliminated materials that were poor conductors, too unstable, too reactive, or too expensive to be used for batteries. The high-performance computing used by Microsoft essentially uses the power of about 5,000 typical laptops. Baker emphasized that after they had early results, Microsoft consulted materials experts at the government lab with the belief that, “Okay, a computer gave us the answer. … Let’s prove it.”

Microsoft officials and the national lab said they will continue the battery project as part of a three-year collaboration aimed at accelerating the pace of innovation and discovery.

“This is not just a headline-grabbing announcement. That’s the nature of things to come,” said Chirag Decate, vice president and analyst at Stamford, Connecticut-based research and consulting firm Gartner, who was not on the research team.

Searching for a better battery

The need for better battery technology has emerged as a major challenge for scientists.

“Many of the current battery technologies are extremely inefficient,” Dekat said. “If you look at one of the reasons why the promise of a green future has so far eluded us, despite advances in solar and wind power, [power]if you look at the root cause, it’s actually the battery technology challenges.”

Lithium-ion batteries now power most electric cars, as well as most electric scooters and electric bicycles, which have become ubiquitous in modern cities around the world. In 2022, the global lithium-ion battery market size was estimated at $46.2 billion, and the industry is expected to reach $189.4 billion by 2032.

But the current crop of batteries cannot charge quickly or hold a charge for long periods. Also, some proved to be fickle.

In 2023, the New York City Fire Department reported that 18 people died in fires related to electric vehicle batteries.

Microsoft is looking for materials to build what it calls a solid-state battery; they have a higher energy density than liquid ion batteries and do not pose a risk of fire or leakage. The material used for the prototype contains some lithium, but up to 70 percent less than the amount found in existing batteries.

Microsoft’s work on the battery project “started in earnest about nine months ago,” Baker said.

The team trained their AI system by showing it examples of different materials, helping it learn their crystal structures and energy properties. “As we look at different compositions, it starts to learn how structure and composition relate to energetics,” Baker said.

The Microsoft team then uses AI to function as a funnel. A large amount of opportunities were entered at the top of the funnel and then passed through various filters that narrow down the list of candidates. The best came out of the bottom of the funnel.

Given that the battery material must be stable, the AI ​​began by filtering the number of candidates down to just under 590,000 that would remain in the shape needed to work in a battery. From there, the AI ​​essentially asked each candidate: would it react to oxygen? Will it do something weird when an electric current is applied through it? How do you get lithium atoms to move through the material, a necessary process for a functioning battery?

Ultimately, Microsoft contacted the Pacific Northwest National Laboratory, which has extensive experience in battery research.

“They asked, what are the properties that you typically need in a material to make a better battery?” said Vijay Murugesan, co-author of the paper and head of the lab’s materials science group in the physical sciences department. He called it a huge challenge to develop a solid material that would only allow lithium ions to move from one side of the battery to the other.

When he looked at a list of 120 to 130 top candidates, Murugesan said his reactions ranged from “Why? What? Really?” for some of the more surprising candidates to “I see this as an obvious next step in what we’re doing.”

The scientists said they were encouraged by the fact that some of the materials recommended by the artificial intelligence had already been noted by experts as showing promise.

The lab team analyzed the best candidates and made a handful of the hypothetical materials dreamed up by the computers. The new material chosen for the prototype contains some lithium along with sodium, chloride and yttrium. The lab is conducting additional work on the material used for the prototype and is preparing to study at least two more candidate materials.

In an emailed response to the preliminary information presented by Microsoft, Aaron Walsh, professor and chair of material design at Imperial College London, said: “AI provides a new generation of approximate but practical tools that allow us to tackle problems that seemed impossible before, such as the rapid exploration of large chemical spaces reported here.

Leave a Comment

Your email address will not be published. Required fields are marked *