To meet the growing demand for batteries, alternative battery chemistries beyond lithium-ion (Li-ion) are required. A promising alternative is lithium-sulfur (Li-S). As most of the materials and morphology development in Li-S batteries is conducted at coin-cell level, it is essential to understand the mechanisms that limit coin-cell performance and how they translate to larger cell form factors. The ability to test assumptions and predictions through a model saves experimental time and cost, and enables simulation of behaviour under different conditions. Physics-based models are being used successfully for Li-ion cell design, and has created an opportunity to achieve this level of maturity with Li-S models. One key difference is that, for Li-S batteries, the mechanisms governing cell performance depend on factors such as electrolyte-to-sulfur, and electrolyte-to-cathode ratios, which change significantly when scaling-up from coin-cell to pouch-cell.

Li-S battery models are predominantly designed for pouch-cell batteries. Whilst these models provide the fundamental tool to analyse the behaviour of a pouch-cell, they have not been tested for coin-cells. The physical differences between coin and pouch-cells result in differing mechanisms, leading to the question: can a pouch-cell model be re-parameterised to accurately predict coin-cell behaviour? To answer this question, several published zero-dimensional (0D) pouch-cell models [1], [2], [3] were analysed and upgraded. The experimental data published by Boenke et al. [4] was used for parametrisation and validation.

The new model comprises of the key mechanisms from existing models, including those required to capture charge and discharge behaviour, precipitation dynamics and polysulfide shuttle. With the addition of several features, and updated derivation of terms including the cathode porosity and surface area, the model can accurately account for behaviour relating to sulfur utilisation and accessibility of active material. The model captures all behaviour relating to the scaling of cell format, including the geometric size and E/S ratio. Unlike existing Li-S models, this model can retrieve C-rate dependence in terms of both capacity and voltage. Whilst maintaining low dimensionality, this is achieved via transport of species between the cathode, separator and a reservoir of excess electrolyte, along with introducing a dependency between the change in area and the precipitation dynamics during cycling.

Despite its simplicity, these changes have significantly improved the causality and accuracy of model predictions when compared to experimental data. From here, further mechanisms or dimensionality can be added if needed, to capture additional features such as cell degradation, with the confidence that the correct causality for coin-cell behaviour is captured.

References:
[1] M. Marinescu, T. Zhang, and G. J. Offer, “A zero dimensional model of lithium-sulfur batteries during charge and discharge,” Physical Chemistry Chemical Physics, vol. 18, no. 1, pp. 584–593, 2016, doi: 10.1039/c5cp05755h.
[2] M. Cornish and M. Marinescu, “Towards Rigorous Validation of Li-S Battery Models,” Dec. 2021, [Online]. Available: http://arxiv.org/abs/2112.08722
[3] K. Kumaresan, Y. Mikhaylik, and R. E. White, “A Mathematical Model for a Lithium–Sulfur Cell,” J Electrochem Soc, vol. 155, no. 8, p. A576, 2008, doi: 10.1149/1.2937304.
[4] T. Boenke et al., “Sulfur Transfer Melt Infiltration for High-Power Carbon Nanotube Sheets in Lithium-Sulfur Pouch Cells,” Batter Supercaps, vol. 4, no. 6, pp. 989–1002, Jun. 2021, doi: 10.1002/batt.202100033.