Computational investigation of materials for energy storage applications

The need for sustainable, economical, and high energy density rechargeable batteries are mandatory to develop next generation consumer electronics, transportation, and stationary energy devices to meet the growing energy demands of the world. In order to meet those challenges, the fundamental understanding of thermodynamics and diffusion processes in battery materials can help achieve the current goals. Atomic-scale simulations using density functional theory coupled with experimental characterization techniques have the capabilities to reveal local atomic environment of battery electrodes and electrolytes which is directly related to the ionic conductivity, stability of electrode-electrolyte interphase, electrode potential, energy density, and rate capabilities. Developing models using atomistic simulations are powerful tools because structural features can be directly compared to experimental characterization results and develop deeper insights into battery processes. In this thesis, a number of atomistic models were developed using computational characterization techniques, which were compared with experiments to develop an accurate understanding of next-generation battery materials for high-energy density applications. These atomistic models were used to compare the catalytic processes in zin-air batteries and the intercalation process in different ion intercalation materials for dual-ion batteries for high energy density and economical application. Furthermore, the computational cost of running electronic structure calculations limits the lengthscale and timescale of simulations to study kinetic processes at experimental timescales, which involve rare events on potential energy surfaces. This work developed a generalized interatomic potential for bulk lithium using a machine learning package PyAMFF to replace density functional theory calculations to study the metal deposition process in batteries.