New Materials Library Helps AI Design Better Sodium-Ion Batteries With 90% Capacity Retention

AI-Guided Material Discovery Tackles Lithium Scarcity

Researchers at the University of Surrey have launched a materials library that could enable AI design of sodium-ion batteries with superior performance characteristics. Published in the Journal of Materials Chemistry A, the study addresses a pressing challenge: while sodium-ion batteries offer a sustainable alternative to lithium-ion batteries, they currently lag in energy density and cycle stability.1

1

Lithium scarcity and uneven global distribution make sodium's abundance particularly attractive for large-scale energy storage applications.2

2

Building the Foundation: 28 Sodium Manganese Oxide Compositions

The research team created a library of 28 different compositions by systematically substituting manganese in sodium manganese oxide with varying amounts of iron, titanium and vanadium. This cathode material directly influences how much energy a battery can store and how well it performs over time.1

1

Each composition underwent rigorous testing in sodium-ion batteries to map the relationship between chemical makeup and performance metrics. Dr. Daniel Commandeur, lecturer at the School of Chemistry and Chemical Engineering and corresponding author, emphasized that finding materials delivering the needed performance remains the core challenge in developing sodium-ion batteries.2

2

Iron, Titanium and Vanadium: Each Element's Distinct Role

The results revealed distinct performance patterns linked to specific elements. Iron substitution correlated with higher energy-storage capacity, making it valuable for applications prioritizing maximum energy density. Vanadium improved stability over repeated charging and discharging cycles, addressing one of the major barriers to wider sodium-ion battery adoption. Small amounts of titanium enhanced high-rate performance, enabling faster charging capabilities.1

1

Notably, some titanium-containing compositions retained more than 90% of their capacity after 100 cycles, demonstrating impressive longevity.2

2

Training Machine-Learning Models for Battery Optimization

The team developed an automated approach to analyze the substantial data generated by their experiments, using the resulting dataset to train a machine-learning model. While the current 28-composition library remains too small for reliable predictions, expanding it could eventually allow AI to identify promising battery compositions before physical synthesis and laboratory testing.1

1

This represents a first step toward building the larger datasets needed for machine learning to guide material discovery and accelerate development timelines for better sodium-ion batteries, according to Dr. Commandeur.2

2

Small Compositional Changes Yield Significant Performance Shifts

Eden Marks, Ph.D. student in space engineering and co-lead author, noted that relatively small changes in composition could dramatically shift the balance between energy storage, high-rate performance and cycle life. Mapping these relationships provides clearer direction for future research efforts.1

1

This granular understanding enables scientists to tailor sodium-ion battery cathodes to specific application requirements—whether the priority is storing more energy, charging faster or lasting through more charge cycles.2

2

Scaling Up: The Path to AI-Driven Battery Innovation

Expanding the approach to encompass a much wider range of cathode materials could create the larger, high-quality datasets necessary for machine learning to meaningfully accelerate battery material discovery. The researchers envision a future where AI models screen thousands of potential compositions computationally, identifying the most promising candidates for experimental validation and dramatically reducing the time from concept to commercial viability. This methodology could help improve sodium-ion battery performance across multiple dimensions simultaneously, making them competitive with established lithium-ion technology while leveraging sodium's superior abundance and accessibility for global manufacturing.