Vacancy-enhanced Li solid-solution anodes for solid-state batteries: computational screening enabling high critical current density

Abstract

Sluggish vacancy-mediated self-diffusion in lithium metal limits the critical current density (CCD) of Li anodes in solid-state batteries by promoting interfacial void formation during stripping. To enhance the effective vacancy concentration while preserving the bcc Li host, we screen 50 substitutional solutes using density functional theory (DFT) within a funnel based on automatic phase diagram construction, solution energies, solute-vacancy binding energies, and experimental solubility limits. Using an adaptation of the Lomer model for the dilute limit, we show that only Ag, Ge, and Cd solid solutions are predicted to reach $1~\mathrm{mA~cm^{-2}}$ within their solubility limits. The thermodynamically estimated CCD of Ag at $\sim 9$ at.% reaches $\sim 4~\mathrm{mA~cm^{-2}}$ at a theoretical capacity of $1522~\mathrm{mAh~g^{-1}}$. Migration barrier calculations indicate that the enhancement may be partially offset by vacancy trapping, necessitating further validation via molecular dynamics simulations. Notably, Hume-Rothery rules alone prove insufficient as prefilters, whereas machine-learning models trained on the DFT data can eliminate most unsuitable solutes early in the funnel and extend the workflow to related solid-solution searches.

Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv.