Void growth in metal anodes in solid-state batteries: Recent progress and gaps in understanding

Abstract

Stripping of metal cations from the anode of a Li- or Na-ion cell into a ceramic electrolyte results in the formation of voids on the electrolyte/electrode interface. Such voids have been observed to grow to sizes in excess of 100 μm. Dendrites can nucleate and grow in the electrolyte from the vicinity of the voids during the plating phase of cycling of the cell, and lead to short-circuiting of the cell. Current theoretical understanding of the formation of these voids is in its infancy: the prevailing qualitative notion is that voids form within the metal anode when the stripping current density removes metal from the interface faster than it can be replenished. We review models that employ the Onsager formalism to develop a variational approach to model void growth by coupling power-law creep of the metal electrode and the flux of metal cations through a single-ion conductor solid electrolyte. These models, based on standard Butler-Volmer kinetics for the interfacial flux, predict that voids will shrink for realistic combinations of interfacial ionic resistance and electrolyte conductivity. Additional physics in the form of modified kinetics, such that the interfacial resistance is decreased by the presence of dislocations within the creeping metal electrode, are shown to give rise to initial growth of voids around impurity particles on the electrolyte/electrode interface. However, these voids ultimately collapse under the imposed stripping fluxes and no conditions have been identified for which isolated voids grow to more than 10μm in size. This is in contrast to the experimentally observed sizes of ∼ 100μm. The physical processes by which large voids form remain unclear but the current state-of-the-art understanding does provide clues of possible mechanisms that have not as yet been considered.