Solid Electrolyte Interphase Formation Model for Lithium-Ion Batteries Based on STEM Observations and Discharge Curve Analysis

摘要

We proposed a model for solid electrolyte interphase (SEI) formation in lithium-ion batteries (LIBs) on the basis of scanning transmission electron microscopy (STEM) observations and discharge curve analysis. A highly precise SEI formation model is needed because such formation is a key factor affecting capacity fading of LIBs from the viewpoint of lithium-ion loss. Conventionally, lithium-ion loss is thought to increase proportionally to the square root of the storage duration (tl/2), and capacity is thought to decrease proportionally to tl/2. However, capacity fading and lithium-ion loss are sometimes not proportional to tl/2, especially under high temperature (> 40 ℃) and state of charge (SOC) storage conditions. We used discharge curve analysis to evaluate the correlation between capacity fading of LIBs and lithium-ion loss caused by SEI formation. We also analyzed the local structure of an SEI using STEM and electron energy loss spectroscopy (EELS) to understand why we observed different types of Li-ion loss behavior depending on the storage conditions. We observed more prominent peaks of the inorganic elements (Li2C03) in the SEI layer in high temperature and SOC storage conditions. As previously noted, capacity fading is not proportional to tl/2 under these conditions. On the basis of these results, we supposed that the formation of the inorganic elements would suppress the formation of an organic SEI. We then formulated both organic and inorganic SEIs and made a combined SEI formation model in which the organic and inorganic SEIs were grown in parallel. In the combined SEI model, we assumed that the formation of the organic SEI would be restricted because the presence of the inorganic SEI would reduce the active surface area on which the organic SEI could grow. We also assumed that the organic SEI would grow proportionally to tl/2 on the active surface area and that the inorganic SEI would exponentially reduce the reaction surface area. Using this model, we could effectively explain lithium-ion loss behavior caused by SEI formation under various conditions.