Generation of virtual lithium-ion battery electrode microstructures based on spatial stochastic modeling

Abstract

It is well known that the microstructure of the active material in lithium-ion battery electrodes has a strong influence on the battery’s performance. In order to improve functional properties of lithium-ion batteries, designing optimized electrode morphologies is an important task. As exploring a large set of possible design concepts via laboratory experiments is very expensive in cost and time, model-based simulations have become an important tool to explore a broad range of possible microstructures on the computer. They allow a preselection of promising design concepts. This procedure, which is called virtual materials design, involves two main tasks. First, a tool for creating virtual, but realistic electrode morphologies is needed. This tool must be able to generate a broad range of electrode microstructures on the computer. In a second step, the performance of these virtual electrodes must be evaluated using spatially resolved numerical transport simulations. In the present paper, the first part of this procedure is addressed. A general framework based on tools of stochastic geometry is presented, which can be used to create a broad range of different electrode microstructures on the computer. To demonstrate the wide spectrum of possible outcomes of the microstructure generator as well as its ability to describe real electrode microstructures, we show how the microstructure of three types of electrodes, which exhibit rather different morphologies, can be described using different adaptions of the framework. A comparison of structural characteristics of the model outputs and tomographic image data of real electrodes indicates a good fit of the model. Moreover, we show how design concepts can be implemented for generating virtual electrode microstructures that can be used as input for spatially resolved transport simulations.