Lithium-Sulfur (Li-S) batteries promise high energy density and low cost, but is hindered by the rapid capacity and energy fading. In this thesis, several integrated computational models were developed to connect the electrode-electrolyte interface mechanisms with the discharge curves and capacity loss. The new insights were used to design the electrodes and electrolyte.Between the Li anode and the electrolyte, the formation of an electrical insulating solid electrolyte interphase (SEI) is responsible for the initial irreversible capacity loss. Assuming the electron tunneling from the electrode to the electrolyte was blocked by the SEI inorganic components at a critical thickness, based on the results from density functional theory (DFT), an analytical model was developed to connect the initial irreversible capacity loss with the anode surface area. Good agreement with experimental measurements confirmed that the initial irreversible capacity loss was due to the self-limiting electron tunneling property of the SEI.In typical cathodes, elemental sulfur is embedded in a carbon matrix. During discharging, soluble long-chain Li polysulfides (Li-PSs) were generated before the insoluble insulating Li2S. In this thesis, it was found that the solvation status of Li-PS (fully, partially, or not dissolved) had a profound impact on discharging curves. The open circuit voltage (OCV) was first predicted using DFT calculated free energies at finite temperatures including solvation energy. This model revealed that the solvation stabilized the Li-PS. Thus, the formation of fully solvated Li-PS led to two-plateaued OCV; while non-solvated Li-PS is not favorable, so the direct transition from S to Li2S led to one-plateaued OCV. Practically, the solvation status of the Li-PS is related to electrolyte volume. A mechanism based analytical model was developed to illustrate the micrometer level porosity determined discharging curve by connecting electrolyte amount, pore volume, Li-PS solubility and carbon matrix surface area. This model was used to optimize the porosity to maximize the volumetric energy density of Li-S batteries.Dissolved Li-PSs can diffuse in the electrolyte, and precipitate as insulating Li2S on electrode surfaces, leading to quick capacity and energy drop. Nanopore and sub-nanopore size were important in controlling the Li-PS solvation status to prevent Li-PS shuttling. The Li-PS formation could be suppressed when partially solvated, suggested by DFT. Since decreasing pore size to nanometer level and increasing electrolyte concentration could both create partially solvated Li-PS, a new strategy to mitigate "Li-PS shuttle problem" based on this synergetic effect was proposed by modeling and verified by experiments. A more idealized structure would be sulfur filled in carbon nanotubes (CNT), with open rings only permeable to Li-ions. Reactive molecular dynamics (MD) simulations showed that the DFT determined optimum open ring size could be achieved by controlling the CNT oxidation.Highly concentrated electrolytes can achieve the partially solvated Li-PS, and expand the electrochemical stability window by forming SEI. To solve the high viscosity issue, co-solvent was designed by adding low viscosity and electronically stable dichloromethane (DCM) to highly-concentrated LiTFSI in ethyl acetate (EA). The concentration of the DCM was designed to obtain a unique solvation structure, where clusters of partially solvated Li+, TSFI-, and EA network were surrounded by the DCM so that the former inherited the expanded electrochemical window and the latter accelerated the Li transport.Overall, this thesis demonstrated that atomic electrode-electrolyte interface structure, interaction, and properties can be directly connected to the cell-level discharge performance, thus modeling these connections provided an integrated approach for battery materials design.
Read
