Dendrite-free alkali-metal electrodeposition from contact-ion-pair state induced by mixing alkaline earth cation


Alkali metals, such as lithium and sodium, have been expected to be used for rechargeable metal-anode batteries owing to their low electrode potentials and large capacities. However, the well-known fatal problem, “dendritic growth” causing a dangerous short circuit, is faced while charging the batteries. Here, through a comprehensive study with electrochemical experiments, Raman and soft X-ray emission spectroscopies, density-functional-theory calculation, and molecular dynamic simulations, we provide an advanced guideline for electrolyte design in which a mixture of alkaline earth (Mg, Ca, Ba) salts is used to inhibit dendrite growth of alkali metals (Li, Na) during electrodeposition. Especially, focusing on CaTFSA2, as a salient exemplary alkaline-earth-cation additive, we demonstrate that dendrite-free morphology upon alkali-metal electrodeposition can successfully be attained by modifying their solvation structures in the dual-cation electrolyte systems. Adding divalent Ca2+ promotes alkali cation (Li+ or Na+) to form the contact ion pairs (CIPs) with the counter anions, which replaces the solvent-separated ion pairs (SSIPs) commonly existing in single-cation electrolytes. Such CIPs related to alkali cations would separate Ca2+ ions distantly to shield the strong coulomb interaction among the divalent cations. The stronger binding of the CIPs would retard the desolvation kinetics of alkali cations and, consequently, realizes a severely constrained alkali-metal electrodeposition in a reaction-limited process that is required for the dendrite-free morphology. This work provides prospects to construct dual-cation electrolytes for dendrite-free alkali-metal-anode batteries utilizing the concerted interactions between monovalent and multivalent cations.


Supplementary material

Supporting Information
Additional electrochemical experiments such as Li-Li sym-metric cells, additional Raman results, radial distribution functions obtained by MD simulations.