All-solid-state alkali-ion batteries (ASSABs) utilizing alkali superionic conductor (ASIC) solid electrolytes (SEs) are an efficient solution to address safety issues from flammable organic solvent-based liquid electrolytes and tend to be more energy-dense. In this dissertation, we illustrate the application of density functional theory (DFT) methods to gain profound insights into the key properties of ASIC solid electrolytes and accelerate their optimization and discovery. This dissertation is divided into three projects.
In the first project, we developed a high-throughput (HT) first-principles screening workflow and applied it to Li−P−S ternary and Li−M−P−S chemical spaces, where M is a non-redox-active element. Given the extraordinary superionic conductors in the silver composition space such as α−AgI and the structural similarity between Ag and Li compounds, we substituted Li for Ag in Ag−P−S and Ag−M−P−S chemical spaces and predicted two novel superionic conductors Li3Y(PS4)2 and Li5PS4Cl2. An efficient tiered screening strategy combining quick topological analysis and short ab initio molecular dynamics (AIMD) simulations was developed to rapidly exclude candidates which are unlikely to exhibit satisfying conductivity.
In the second project, we investigated the influence of Na defects and cation/anion dopants in Na3PS4 and evaluated the phase stability, dopant formation energy as well as Na+ conductivity. Aliovalent M4+ (M = Si, Ge, Sn) for P5+ substitution in c-Na3PS4 introduces Na+ excess, which is shown to significantly enhance ionic conductivity. AIMD simulations predict that 6.25% Si-doped c-Na3PS4 has a conductivity of 1.66 mS/cm, which is very close to the experimental value of 0.74 mS/cm at a 6% Si-doping ratio. Further analysis using topological methods and the van Hove correlation function explains how dopants affect the channel volume and how correlation motions of Na+ enhance Na diffusion in this material. Besides cation doping, we showed that anion doping, e.g., Cl- substitution for S2-, is an alternative strategy for introducing vacancy defects in t-Na3PS4 to improve both electrochemical stability and Na+ conductivity, with a conductivity of ~1.14 mS/cm demonstrated experimentally. In collaboration with NMR experiments, we show via AIMD simulations that maximizing the concentration of Cl dopants while maintaining relatively low Na deficiency can further enhance Na conductivity to 2 mS/cm.
In the third project, we systematically assessed cation mixing effects on Na3Pn'xPn''1-xS4 and Na4-xTt1-xPnxS4 (Pn = P, As, Sb; Tt = Si, Ge, Sn) in terms of thermodynamic/electrochemical/moisture stability and conductivity. Isovalent and aliovalent cation mixing in prototype structures including Na3PnS4 and Na11Sn2PS12 as a subset, gives the lever of tuning Na concentration and cation types/ratio simultaneously to achieve the twin goals of better stability and larger diffusion channel. We demonstrated that Na3Pn'xPn''1-xS4 exhibits negative or very small positive mixing enthalpies, including these cation-mixed compositions are likely to form solid solutions. The newly experimental Na11Sn2PS12 prototype with I4_1/acd space group is an excellent model to study and holds the promise to perform even better as a solid electrolyte after substituting with other Tt4+ and Pn5+ elements. We proposed a new compound of Na11Sn2AsS12 which can potentially achieve higher conductivity and chemical stability than already experimentally reported compounds.