Battery Cathodes and Energy Storage

Cathode development has stalled around a handful of chemical families. NMC, LFP, and NCA dominate commercial cells, and most R&D consists of incremental doping or stoichiometric tweaks within these chemistries. The combinatorial landscape of multi-element oxides, sulfides, and phosphates dwarfs what any synthesis campaign can sample. Promising composition regions go untested simply because human-guided exploration cannot keep pace with the scale of the search space. Meanwhile, the energy transition demands step-change improvements that marginal refinements to decades-old formulations will not deliver.

Computational cathode screening typically starts from a pre-enumerated candidate list and evaluates it with density functional theory. This is expensive and inherently limited to compositions someone has already thought to propose. Combinatorial synthesis covers narrow windows dictated by available precursors. Machine-learning surrogates interpolate well within known chemical families but extrapolate poorly to genuinely novel chemistries. All three approaches share the same bottleneck: they evaluate candidates rather than construct them, so the most impactful regions of composition-structure space (those farthest from existing data) stay out of reach.