Sodium-ion batteries could soon overcome their performance bottlenecks, thanks to a new materials library built by University of Surrey researchers to train future artificial intelligence models. Sodium offers a crucial advantage over conventional lithium-ion cells due to its abundance and wide accessibility across global supply chains. Yet, improving energy density and capacity retention across repeated charge cycles has stalled widespread commercial deployment until now.
Targeted Chemical Substitutions in Sodium Manganese Oxide
Writing in the Journal of Materials Chemistry A, the research team investigated how altering the chemical composition of sodium manganese oxide impacts its performance as a battery cathode. The cathode dictates both the total energy a cell stores and its operational efficiency. To map these traits, scientists assembled a library of 28 distinct compositions by substituting portions of manganese with varying quantities of iron, titanium, and vanadium.
Laboratory testing yielded distinct performance markers for each element. Iron increases energy-storage capacity, vanadium enhances stability over continuous charge and discharge cycles, and small amounts of titanium boost performance at elevated rates. Certain titanium-containing formulas retained more than 90 percent of their original capacity after 100 testing cycles.
Automating Data Collection for Machine Learning Integration
To handle the vast volume of experimental data, the team deployed an automated processing approach and used the resulting dataset to train an initial machine-learning model. While the current library remains too small for dependable predictive modeling, expanding the dataset could eventually allow artificial intelligence to forecast novel battery compositions prior to physical laboratory testing.
“We rely heavily on lithium-ion batteries, but lithium resources are relatively scarce and unevenly distributed around the world,” said Dr. Daniel Commandeur, a lecturer at the School of Chemistry and Chemical Engineering and corresponding author of the study. “Sodium is far more abundant, making sodium-ion batteries an important technology to develop.”
Mapping Fine Chemical Balances for Future Performance
Dr. Commandeur noted that the primary obstacle involves identifying materials that deliver required performance metrics. “This work is a first step towards building the larger datasets needed for machine learning to guide that search—helping us identify promising materials faster and accelerate the development of better sodium-ion batteries,” he added.
Expanding the framework to encompass a wider spectrum of cathode materials will generate the high-quality data necessary for machine learning acceleration. Eden Marks, a PhD student in space engineering and co-lead author of the research, pointed out the sensitivity of the chemical balances.
“What stood out most was how relatively small changes in composition could shift the balance between how much energy a material can store, how it performs at higher rates and how well it lasts,” Marks said. “Mapping those relationships gives us a much clearer picture of where to look next.”
Worth a look