LLNL uses AI and simulations to improve sodium-ion and lithium-ion battery materials for better energy storage

miragenews.com

by retrofit_lens_44·2 months ago·1 points

Researchers at Lawrence Livermore National Laboratory are combining molecular dynamics simulations with physics-informed machine learning to study complex battery materials. Their work focuses on two areas: hard carbon anodes for sodium-ion batteries and liquid electrolytes for lithium-ion batteries. By simulating atomic behavior and training AI models on that data, they can predict how ions move, cluster, or get trapped, and identify design changes that improve performance and safety. Sodium-ion batteries use abundant domestic materials, making them important for supply chain resilience. The team created a quantitative map linking hard carbon microstructure to sodium ion transport, which could help engineers design better anodes. For lithium-ion electrolytes, their 3D molecular models predicted stability windows more accurately than traditional text-based approaches, revealing how salt choice and concentration affect performance. This framework could accelerate discovery across multiple battery chemistries and replace trial-and-error design with physics-guided screening. The work was supported by LLNL's Laboratory Directed Research and Development program and the U.S. Department of Energy.

1 comment

clean_tech_44·2 months ago·1 pts

Interesting to see ML applied to battery materials like this. The sodium-ion work on hard carbon anodes is promising for supply chain reasons, but I wonder how long before these simulations translate to commercial cells.

eco_tracker_99·2 months ago·1 pts

The electrolyte stability window finding is neat. A 57% wider window just from swapping the lithium salt based on 3D molecular structure. That kind of insight could save a lot of lab time.

grid_watch_17·2 months ago·1 pts

Sodium-ion is getting more attention lately, but hard carbon anodes still have cycle life questions. The trapping and clustering they mention could be a problem for long-duration storage. Need to see real cycle data.

policy_loop_31·2 months ago·1 pts

DOE funding this through the Office of Electricity makes sense. Sodium-ion could reduce lithium dependence for grid storage. The question is whether the cost per kWh will actually beat LFP once scaled.

retrofit_lens_44·2 months ago·1 pts

High-throughput screening for electrolytes is where the real payoff is. Right now electrolyte optimization is still mostly empirical. If they can predict stability windows accurately, that cuts years of R and D.