In a significant advancement for energy storage technology, a research team from the New Jersey Institute of Technology (NJIT) has unveiled five novel materials that could replace conventional lithium batteries. By employing cutting-edge generative artificial intelligence techniques, these new materials, which are based on abundant elements such as magnesium and calcium, promise to revolutionize multivalent batteries. This development comes at a time when there is a global race to find safer and more cost-effective alternatives to lithium.
Researchers from NJIT utilized artificial intelligence to tackle a critical challenge in the future of energy storage: the need for sustainable and affordable substitutes for lithium-ion batteries, as noted by sciencedaily.
In a study published in Cell Reports Physical Science, the NJIT team, led by Professor Deepakar Dutta, successfully used generative AI techniques to rapidly discover new porous materials that could transform multivalent ion batteries. These batteries, which include elements such as magnesium, calcium, aluminum, and zinc, present a promising and cost-effective alternative to lithium-ion batteries, which are currently grappling with supply issues and sustainability concerns.
Unlike traditional lithium-ion batteries that rely on lithium ions with only one positive charge, multivalent ion batteries utilize ions that can carry two or even three positive charges. This capability allows multivalent batteries to store significantly more energy, making them an ideal candidate for future energy storage solutions.
However, the larger size and higher charge of multivalent ions pose challenges for efficient incorporation into battery materials, a hurdle that the NJIT research aims to overcome directly.
Dutta explained, “One of the biggest obstacles was the lack of promising chemical compositions for batteries and the impracticality of testing millions of material combinations.” He continued, “We turned to generative AI as a rapid and systematic approach to analyze this vast field and identify the few structures that could make multivalent batteries truly viable.”
Furthermore, he added, “This approach allows us to explore thousands of potential candidates quickly, significantly accelerating the quest for more efficient and sustainable alternatives to lithium-ion technology.”
To address these challenges, the NJIT team developed a unique dual AI methodology: a Crystal-Dependent Variational Autoencoder (CDVAE) and a finely tuned Large Language Model (LLM). Together, these AI tools were able to explore numerous new crystalline structures rapidly, something that was previously unattainable with traditional laboratory experiments.
The CDVAE model was trained on extensive datasets of known crystalline structures, allowing it to propose completely new materials with diverse structural potential. At the same time, the LLM was fine-tuned to focus on materials that are closer to thermodynamic stability, which is crucial for practical applications.
Dutta mentioned, “Our AI tools have dramatically accelerated the discovery process, revealing five entirely new structural forms of porous transition metal oxides, which show very promising results.” He added, “These materials feature open and wide channels, ideal for the rapid and safe transport of these large multivalent ions, marking a significant milestone for next-generation batteries.”
The team validated the structures generated by AI through quantum mechanics simulations and stability tests, confirming that these materials can indeed be experimentally fabricated and hold substantial potential for real-world applications.
Dutta emphasized the broader implications of their AI-driven approach: “This is not just about discovering new battery materials; it’s about creating a rapid, scalable method for exploring any advanced materials, from electronics to clean energy solutions, without extensive trial and error.”
Encouraged by these promising results, Dutta and his colleagues plan to collaborate with experimental laboratories to assemble and test the AI-designed materials, pushing forward towards the commercial viability of multivalent ionic batteries.