Abstract
Understanding and controlling the atomistic-level reactions governing theformation of the solid-electrolyte interphase (SEI) is crucial for theviability of next-generation solid state batteries. However, challenges persistdue to difficulties in experimentally characterizing buried interfaces andlimits in simulation speed and accuracy. We conduct large-scale explicitreactive simulations with quantum accuracy for a symmetric battery cell,{\symcell}, enabled by active learning and deep equivariant neural networkinteratomic potentials. To automatically characterize the coupled reactions andinterdiffusion at the interface, we formulate and use unsupervisedclassification techniques based on clustering in the space of local atomicenvironments. Our analysis reveals the formation of a previously unreportedcrystalline disordered phase, Li$_2$S$_{0.72}$P$_{0.14}$Cl$_{0.14}$, in theSEI, that evaded previous predictions based purely on thermodynamics,underscoring the importance of explicit modeling of full reaction and transportkinetics. Our simulations agree with and explain experimental observations ofthe SEI formations and elucidate the Li creep mechanisms, critical to dendriteinitiation, characterized by significant Li motion along the interface. Ourapproach is to crease a digital twin from first principles, without adjustableparameters fitted to experiment. As such, it offers capabilities to gaininsights into atomistic dynamics governing complex heterogeneous processes insolid-state synthesis and electrochemistry.