Modeling and Simulation of Lithium-Ion Batteries with Silicon Anode and Ionic Liquid Electrolyte

Abstract

We owe the ubiquity of lithium-ion batteries (LIBs), and therefore much of our modern lifestyle, to the continuous development and improvement of this electrochemical energy storage technology. The latest approaches are pursuing the integration of novel materials with higher storage capacity, above all silicon. Here, especially nanostructured silicon anodes are promising due to their higher intrinsic stability. With their large electrochemical window, ionic liquids are an excellent counterpart to such low-potential anodes. Furthermore, the safety of LIBs is a prominent issue which can be addressed by applying non-flammable ionic liquids. Modeling and simulation support the development of such next-generation batteries. Modeling strategies based on physical and chemical concepts provide intrinsic understanding of the components. This thesis focuses on thermodynamically consistent transport theories for the description of different materials relevant for LIBs. This includes the application of a transport theory for highly concentrated electrolytes like ionic liquids and a chemo-mechanically coupled model for deforming anode materials like silicon. The theories include important aspects of the respective materials to accurately capture their transport behavior under operating conditions. Special attention is given to the relevance of reference frames. A variation of the electrolyte transport theory in the center-of-volume frame of reference is presented - in contrast to the widespread center-of-mass reference frame. This volume-based theory is applied to experimental measurements of ion mobilities obtained via electrophoretic nuclear magnetic resonance (eNMR). This method is powerful since it directly detects ion mobilities in highly concentrated electrolytes. The volume-based reference frame describes the experimental findings from ionic liquids and ionic liquid mixtures best. It was found that the relevant boundary condition in the eNMR setup is a vanishing volume flux in contrast to a vanishing momentum flux. On that basis, the focus is brought to transference numbers, which are frame-dependent transport parameters. They are not only an important performance indicator for electrolytes but also a necessary input for physics-based simulations. Transference numbers in different reference frames are presented together with the respective transformation rules. Finally, ionic liquids and silicon anodes are combined into a full cell modeling framework using the respective transport theories. Here, eNMR measurements provide some of the important input parameters for the ionic liquid electrolyte mixture. A volume-averaged 1d+1d modeling framework is used to study in particular different aspects of a silicon nanowire anode. Physics-based simulations enable the visualization of usually inaccessible quantities like for example the stress distribution in the anode. Nanostructured materials are beneficial due to the slow lithium-ion diffusion in silicon. Nevertheless, it is important to consider the volumetric expansion of silicon when simulating such anodes. Parameter studies on certain anode geometry parameters highlight the importance of supplying sufficient pore space for the volumetric expansion of silicon even with nanostructured anodes.