The tortuous path to energy density: Microstructure-based cell models for graded electrode design

Microstructure design of battery electrodes is key to enable fast charge and higher capacity in next-generation lithium-ion batteries but exploring the possible design space traditionally relies on slow trial-and-error. Here we introduce microstructure-aware cell models based on a novel electrode microstructure descriptor, the spatial tortuosity τ(x), which is directly extracted from 3D image-based simulations. The study shows that state-of-the-art approaches for cell characterisation and simulation are only valid for homogeneous battery electrodes while this new approach naturally generalizes to graded and multi-layer electrode designs. By preserving the computational efficiency of porous electrode models while embedding microstructure information, our workflow runs in minutes per design and provides physically interpretable levers for architecture optimisation under fast (dis)charge. Combined with conditionalised generative AI methods, this pipeline enables high-throughput exploration of the electrode microstructure space accelerating the design of electrodes that simultaneously improve transport, reaction uniformity, and accessible capacity.