Posted on 08/05/2025 6:26:13 AM PDT by Red Badger
AI just helped scientists find five new materials that might outperform lithium in future batteries.
These discoveries could enable cheaper, safer, and more powerful energy storage by using elements like magnesium and zinc.
Tackling the Lithium-Ion Problem
Researchers at the New Jersey Institute of Technology (NJIT) are using artificial intelligence to address a major challenge in the future of energy storage: finding low-cost, environmentally friendly alternatives to lithium-ion batteries.
In a study published in Cell Reports Physical Science, a team led by Professor Dibakar Datta used generative AI to rapidly identify new porous materials that could transform the development of multivalent-ion batteries. These next-generation batteries rely on more widely available elements such as magnesium, calcium, aluminum, and zinc. Compared to lithium-ion batteries, which are increasingly strained by supply and sustainability concerns, multivalent-ion batteries represent a promising and more affordable path forward.
Why Multivalent-Ion Batteries Are the Future
Multivalent-ion batteries differ from conventional lithium-ion versions by using ions that carry two or three positive charges instead of just one. This allows them to store much more energy, making them a compelling option for future energy storage technologies.
The challenge, however, lies in the larger size and stronger charge of these multivalent ions, which makes it difficult for them to move efficiently within standard battery materials. The NJIT team’s AI-powered approach was designed specifically to overcome this barrier by discovering materials better suited for handling these high-charge ions.
Multivalent-Ion Movement - The open, sponge‑like network inside a porous transition‑metal oxide lets the larger, doubly- or triply-charged ions travel during a battery’s charge and discharge cycles. Credit: New Jersey Institute of Technology Turning to Generative AI for Solutions
“One of the biggest hurdles wasn’t a lack of promising battery chemistries — it was the sheer impossibility of testing millions of material combinations,” Datta said. “We turned to generative AI as a fast, systematic way to sift through that vast landscape and spot the few structures that could truly make multivalent batteries practical.
“This approach allows us to quickly explore thousands of potential candidates, dramatically speeding up the search for more efficient and sustainable alternatives to lithium-ion technology.”
The Power of Dual-AI: CDVAE and LLM
To overcome these hurdles, the NJIT team developed a novel dual-AI approach: a Crystal Diffusion Variational Autoencoder (CDVAE) and a finely tuned Large Language Model (LLM). Together, these AI tools rapidly explored thousands of new crystal structures, something previously impossible using traditional laboratory experiments.
The CDVAE model was trained on vast datasets of known crystal structures, enabling it to propose completely novel materials with diverse structural possibilities. Meanwhile, the LLM was tuned to zero in on materials closest to thermodynamic stability, crucial for practical synthesis.
Discovery of 5 Breakthrough Structures
“Our AI tools dramatically accelerated the discovery process, which uncovered five entirely new porous transition metal oxide structures that show remarkable promise,” said Datta. “These materials have large, open channels ideal for moving these bulky multivalent ions quickly and safely, a critical breakthrough for next-generation batteries.”
The team validated their AI-generated structures using quantum mechanical simulations and stability tests, confirming that the materials could indeed be synthesized experimentally and hold great potential for real-world applications.
Beyond Batteries: A Scalable Materials Revolution
Datta emphasized the broader implications of their AI-driven approach: “This is more than just discovering new battery materials — it’s about establishing a rapid, scalable method to explore any advanced materials, from electronics to clean energy solutions, without extensive trial and error.”
With these encouraging results, Datta and his colleagues plan to collaborate with experimental labs to synthesize and test their AI-designed materials, pushing the boundaries further towards commercially viable multivalent-ion batteries.
Reference:
“Generative AI for discovering porous oxide materials for next-generation energy storage”
by Joy Datta, Amruth Nadimpally, Nikhil Koratkar and Dibakar Datta, 26 June 2025, Cell Reports Physical Science.
DOI: 10.1016/j.xcrp.2025.102665
Exactly—AI will conclude that a few casualties are definitely worth it to advance science.
We are their guinea pigs and lab mice.
Lol.
Will the owner of the AI engine used for this research claim rights to any future patents?
Solar cell shirts next on agenda ?.
AFAIK, AI designed stuff is not patentable...........
Graphene has been the subject of much research for improving battery life. So far, there have been no break throughs with it.
Battery technology is what is holding back electric. If there is ever a viable solution, invest everything you have in it.
Yep, low cost storage high capacity would change alot. It would make battery powered homes viable, assuming you have a solar field in your backyard.
“AI Just Found the Future of Batteries, And It’s Not Lithium”
might ... future ... could ...
el clicko-baito headline ...
Dilithium crystals are still the answer, you know. What I am looking for is the transformation that makes aluminum transparent, so it may be used for optical windows on interstellar spacecraft.
The Tesla Model 2, coming later this year, uses a modular aluminum-ion battery, which charges six times faster than lithium batteries and has higher energy density.
Lithium isn’t a “rare earth”; that term belongs to elements 58 through 71. Lithium is an alkali metal, like sodium or potassium.
That $20 million Biden gave her might just kill her by eating herself to death.
Call me a luddite, but I think AI is the embodiment of the old TZ episode, “To Serve Man”.
Dilithium crystals are still the answer, you know.
Beam me up Scottie
For the purpose of this discussion I ignore the the elephant in the room call Global Warming which is the bases for this search for better batteries.
But the real problem with these alternative metals that they are suggesting for lithium (Zink, Calcium, Magnesium and Aluminum) is that they are heavier than Lithium.
Better storage capacity is great. Less combustibility is great.
But the real hurdle to overcome is Energy Density.
Power storage to weight is the major problem.
15 gallons of gas weighs 90 pound and will take my car 420 miles.
Another advantage to gasoline is that as the gasoline burns the car gets lighter, the better the fuel milage.
With a battery, as the charge on the battery lessens, the battery weight stays the same.
As far as the issues with battery fires, I don't see many advantages with these other metals. They are still vary flammable and difficult to extinguish.
I will probably be proven wrong, but the technical issues with batteries may make it impossible to overcome the advantages of gasoline.
What to do with the battery when it is at the end of its life is still a problem yet to solve. Burned gasoline is recycled by nature.
Speaking of which, whatever happened to the "sodium battery" miracle?
AI works on weekends and doesn’t have to waste time on Zoom meetings
Or fill out TPS reports?
It’s my understanding that while an AI itself can’t be listed as the inventor (of course), the person that uses AI to design or create a product can absolutely patent it—so long as the applicant can describe the invention in enough detail that someone skilled in the field could make and use it (this is known as the “enablement” requirement).
To be patentable, an invention must involve a human inventor, be novel and non-obvious (not merely a discovery), and be disclosed in a patent application with a detailed description that meets the “enablement” standard.
The application must also meet the utility requirement (the invention must have a specific, real-world function) and the subject matter requirement (it must fall within a patentable category like a machine, process, or composition of matter—not an abstract idea, law of nature, or purely theoretical concept).
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.