Li-/Na-/K-Dependent Nanopore Filling Pathways Determine Storage Mechanism, Capacity and Voltage in Hard Carbon

Hard carbon is increasingly explored as a universal anode across Li-, Na-, and K-based batteries, yet the mechanistic origin of nanopore filling and the pore features that govern ion-specific filling remain under debate. Using our previously reported Na results in zeolite-templated carbon (ZTC) nanopores as a reference benchmark, we test whether the pore-filling framework established for Na extends to Li and K and identify the limits of this mechanism across alkali chemistries. I will show through DFT, AIMD simulations, and convex-hull thermodynamics that Li undergoes a distinct two-regime filling mechanism, in which an ionic carbon-wall adsorption state is separated from a metallic clustering state by a composition-dependent stability threshold, whereas K shows a more continuous filling behavior with weaker separation between ionic and metallic clustered states. In comparison with the previously reported Na behavior, these results reveal how decreasing ionic charge density from Li to K changes the balance between ion–carbon wall stabilization and ion–ion clustering under nanopore confinement. Relative to graphite intercalation limits, ZTC nanopores support high stoichiometries for both Li and K, with Li approaching LiC 1.2 versus LiC 6 in graphite and K reaching KC 5.6 versus KC 8 . Compared with our previously reported Na benchmark of NaC 3 , these results show that similarly high alkali storage can emerge in nanoporous carbon, but through metal-dependent filling pathways. Convex-hull-derived voltage profiles further show that low-voltage pore filling is not a generic property of nanoporous carbon. Although all three metals transition from an initial high-voltage insertion regime into lower-voltage filling, only selected pore–metal combinations access the ~0.1 V window relevant to hard-carbon pore storage. Consistent with our prior Na results, Li also shows a systematic decrease in plateau voltage with increasing pore diameter, whereas K exhibits a non-monotonic response, indicating that pore-size features alone are insufficient to describe alkali-specific low-voltage storage. The highest predicted gravimetric capacities follow a consistent ordering of Li > Na > K, with Li reaching 1205–1616 mAh g⁻¹, Na up to 679 mAh g⁻¹, and K up to 387 mAh g⁻¹ in the ~1.5 nm pore window, showing that the same nanopore does not provide a universal storage response across alkali chemistries. For Li, the 1.10–1.50 nm pore-size window delivers capacities comparable to some of the highest values reported experimentally for nanoporous carbons, consistent with the present predictions. These findings identify pore-alkali metal matching, rather than pore size features by themselves, as the governing design principle for low-voltage and capacity alkali storage in nanoporous carbons.