Prediction of loss of lithium caused by anode overhang effects – Model validation and parameter study

Summary:

Most lithium-ion batteries contain areas of active material with no opposing counter-electrode as a result of the production process. This overhang, which is formed by the graphite anode in most lithium-ion battery designs, contributes to the overall cell performance as a sink or a source of cyclable lithium inventory over time periods ranging from days to months. Therefore, overhang effects can strongly affect the capacity evolution of lithium-ion batteries both during accelerated aging tests and in field, even causing phenomena like a rise of capacity in the first phase of aging.
To allow for the prediction of changes in the cyclable lithium inventory caused by overhang effects, the authors performed an extensive experimental study analyzing the overhang of more than 100 coin cells with double-side coated anodes by color recognition. The results were then applied to validate the feasibility of a 3d-capable Newman type battery model to predict the lithium transport within the anode overhang, showing great agreement between experimental data and simulation, thus proving the capability of the model used to predict long-range lithium transport phenomena.
In a subsequent parameter study using the validated Newman-type model, the authors identified overhang size, active material porosity, active material tortuosity as well as ionic conductivity and conducting salt diffusion coefficient of the electrolyte as major influencing factors on the lithium transport velocity into the overhang. Further investigation of the internal states of the model reveals that directly after a change in SOC of the cell, lithium transport into the overhang is limited by the conducting salt diffusion in the electrolyte and leads to a total depletion of conducting salt within the overhang area. Subsequently, however, lithium transport transforms into a process restricted by the ionic conductivity of the electrolyte. In an empirical fitting approach, the diffusion limited period of lithium transport can be approximated by a square root of time function, whereas in the long term, lithium transport is best fitted by an exponential term.