For understanding of existing battery systems such as lithium-ion batteries and for the development of future battery systems, advanced numerical simulation tools are important. One beneficial aspect of computational methods is that they can provide insight into physical and chemical aspects, which sometimes cannot be probed by experimental methods. Accuracy and reliability are key issues for such numerical simulations and depend on appropriate physical models, boundary conditions, and accurately determined physico-chemical parameters1. For numerical simulations of battery systems, the ion-transport model for concentrated electrolyte solutions introduced by Newman and Thomas-Alyea is frequently used and depends on three ion transport parameters, namely the conductivity, the transference number and the binary diffusion coefficient. In addition to the transport parameters, the thermodynamic factor which is derived from the mean molar activity coefficient is required for the correct description of the thermodynamic behavior of a binary electrolyte solution. Since the microscopic geometry of actually used porous electrodes and separators are largely unknown, a homogenization approach is applied for the macroscopic description of porous media. In this case, the influence of the microstructure on the macroscopic behavior is modelled by additional geometric parameters such as the porosity ε and the tortuosity τ. For an accurate description of the interface reactions the intercalation and deintercalation kinetics have to be known, often described by a Butler-Volmer behavior. Lastly solid state diffusion of lithium ions within the active materials may limit the battery performance, especially at high charge and discharge currents. A vast spectrum of physico-chemical parameters can be found in the literature (e.g., Valoen and Reimers2), but since the investigated electrolytes and the experimental techniques differ from study to study, it is difficult to find a consistent parameter set for a given electrolyte. In addition, some of the parameters are often fitted to match experimental performance data and may thus be fitting parameters rather than intrinsic physico-chemical parameters with predictive capability. We will present an overview of the necessary physico-chemical parameters for simulation of lithium ion battery performance and will show methods by which they can be measured rigorously. For example, novel methods will be shown with which one can determine transport, thermodynamic and geometrical properties like for instance the transference number, the thermodynamic factor and the separator and electrode tortuosity3-5. Figure 1 exemplarily shows one of those results, namely the determination of electrode tortuosities based on impedance spectroscopy measurements in symmetrical cells. Our concentration and temperature dependent studies of physico-chemical parameters as well as the methodology for their unambiguous deter-mination will allow for an improved correlation between lithium ion battery experiments and simulations. More importantly, the use of intrinsic physico-chemical paramters will enable predictive simulations rather than data fitting models, which helps in understanding as well as predicting present and future lithium ion battery technologies. Figure 1: Comparison of determined tortuosities for different electrodes at porosities between 30% and 35% and comparison with the tortuosity obtained by the commonly used Bruggeman estimate (τ=ε -0.5). Acknowledgements We gratefully acknowledge the funding by the Bavarian Ministry of Economic Affairs and Media, Energy, and Technology for its financial support under the auspices of the EEBatt project.
Read full abstract