Structure-property relationships of lithium-ion battery cathodes using stochastic 3D modeling and resistor network simulations based on synchrotron tomography

Macroscopic effective transport properties of battery materials are predominantly influenced by the morphology of their microstructure. In order to bridge the gap between these different length-scales, a stochastic 3D microstructure model is combined with physical simulations on the electrode scale. More precisely, the stochastic model is used to generate virtual, but realistic microstructures of cathodes in lithium-ion batteries. The model is calibrated to tomographic image data and validated with respect to various geometric descriptors and effective transport properties. By systematically varying the parameters of the model, a data base of artificial microstructures is created with varying volume fractions and size distributions of active material particles. Furthermore, a resistor network method is used to quickly compute the effective conductivity and effective diffusivity of these microstructures on the electrode scale, where the microstructure is represented by a simplified graph structure. Subsequently, the available data base is used to investigate quantitative structure-property relationships, which link geometric descriptors of the microstructure to effective transport properties of the electrode. Finally, the results are used to validate a previously established empirical formula that uses geometric descriptors of the microstructure to predict the effective conductivity.