Lithium sulfur (Li-S) batteries are considered a promising energy storage technology for that its theoretical specific capacity of 1.672 Ah g^-1 is over five folds of currently commercialized lithium-ion batteries. The charge/discharge processes of Li-S batteries, however, involve complex reactions of sulfur species and are not yet fully understood. Physical models are crucial to elucidate these mechanisms occurring in Li-S batteries in assisting a multitude of experimental investigations.

However, previously developed models are usually overly simplistic and lack generality. The first one-dimensional Li-S model employed dilute solution theory including diffusion and migration, where transport properties base only on interactions between the solute and the solvent. In the case of multicomponent transport, ion pairing and ion association are important. Therefore, we integrate concentrated solution theory to the macroscopic model to incorporate effects of multi-valent states of sulfides.

Comparing our concentrated solution theory results to those of dilute solution theory, the discharge curves resembled one another, suggesting concentration polarization took little part in the overall potential loss under current parameterization. Thus, we investigate interrelated components of potential loss, namely the surface overpotential, ohmic potential drop, and concentration overpotential. Fig 1 shows that the modeled discharge rate capability is limited by the anode surface overpotential. This result is inconsistent with previous studies, where the low discharge rate has been explained in terms of mass transport limitation and slow charge transfer at the cathode due to precipitation of Li₂S. We consider the lack of reliable parametrization is to blame for the inconsistency, and seek to resolve the parameter-prediction problems inherent in computational models.

To simulate multicomponent diffusion in concentrated solutions, we compute the full transport properties through the representation of Onsager coefficients matrix (Fig 2) and thermodynamic factors using equilibrium molecular dynamics (MD). In addition, we report a theoretical analysis of sampled-current voltammetry based on concentrated-solution to compute the exchange current density, which is not previously established. We emphasize the mathematical consistency among the formulation of modeling equations, experimental methods, and MD methods to achieve fidelity in simulations.