Effect of charging/discharging on plasticity and fracture in Si anodes for batteries

Lithium-ion batteries provide the power for portable electronics and found many other interesting applications. It is desirable to design batteries with high energy density and long cycle life. Silicon is among the highest Li-storing anode materials in batteries, but the large capacity is accompanied by significant volume expansion that causes mechanical failure and capacity fading after few charging/discharging cycles. There is evidence that the mechanical behavior of lithiated amorphous silicon depends on the history of charging and discharging. The goal of this thesis is to better understand the mechanical properties and behavior of lithiated silicon, specifically the dependence of its properties on the state of charging/discharging, through atomistic simulation, constitutive modelling and finite elements method calculations. Amorphous high-storage-capacity Li-Si flows at lower stresses than crystalline materials but the plastic flow stress decreases with charging and discharging, indicating important non-equilibrium aspects to the flow behavior. In this thesis, a mechanistically-based constitutive model for rate-dependent plastic flow in amorphous materials, during charging and discharging is developed based on two physical concepts: (i) excess energy is stored in the material during electrochemical charging and discharging due to the inability of the amorphous material to fully relax during the charging/discharging process and (ii) this excess energy reduces the barriers for plastic flow processes and thus reduces the flow stress. Plastic flow stress in our model is a result of a competition between the time scale of charging/discharging and the time scale of glassy relaxation. The two concepts, as well as other aspects of the model, are validated using molecular simulations on a model Li-Si system. Furthermore, I formulate and implement a finite element method based on the developed constitutive model to capture the full complexity of coupled chemical-mechanical evolution including plastic flow that arises in these amorphous battery materials. Fracture is the main cause of degradation and capacity fading in lithiated silicon during cycling, thus it is essential to develop mechanistic models for the fracture of Li-Si to interpret the experiments and facilitate the design. Here, I perform systematic atomistic simulations of crack propagation for different Li compositions discharged samples. I observe void nucleation and coalescence as the primary mechanism of crack growth in all samples. Discharging increases the structural disorder which results in decrease in the flow stresses but simultaneously facilitates void nucleation and growth. The fracture toughness and energy is increased by discharging, indicating that the flow and fracture of lithiated silicon depends on the history of charging/discharging. Because of the similarities between the fracture mechanism of Li-Si and ductile fracture, Gurson's model is used to help interpret the simulation results. Qualitative agreement between the trends predicted by Gurson's model based fracture simulations and MD simulations of cracks in Li-Si is demonstrated. Gurson-type models predict that the fracture energy scales with the yield stress and void spacing. In all tested cases the nucleated voids spaced within few nanometers of the crack tip, which explains the low fracture energy of Li-Si.