(362g) Experimental and Computational Characterization of Cathodes for LiS Batteries

Nowadays, lithium-ion batteries (LIBs) are playing a crucial role in the field of sustainable renewable energy sources due to their large versatility which allows them to be used in different applications such as energy storage systems, electric vehicles and electric portable devices. However, considering the further increase in global demand of batteries and electric performance, it is important to underline two key factors which are limiting the future employment of LIBs in the market: the supply of critical raw materials (CRMs) such as transition metals like nickel, manganese and cobalt, and the practical specific capacity of commercial LIBs. These limitations are leading to research on new energy storage systems, such as lithium-sulfur (LiS) and lithium-air (LiO₂) batteries, which are growing exponentially as future alternative of LIBs. Both systems have great advantages in respect to lithium-ion, such as high energy density, low cost, high availability of active material and they are more sustainable. In particular, LiS batteries seem to be more suitable to be commercialized in the coming years.

However, LiS systems suffer from two main limitations: the shuttling effect and the precipitation of active species, both leading to capacity fade and cell degradation. The first one consists in the migration of long-chain polysulfides from the cathode to the anode where they can react to form shorter polysulfides or lithium sulfides, which precipitate directly on the anode surface, involving both loss of active materials and anode corrosion, contributing to self-discharge. On the other side, the final products of charge and discharge processes (sulfur and lithium sulfide) are insoluble and precipitate, driving to loss of active species and deterioration of cathode structure due to the low electrochemical reversibility of these reactions. Both phenomena can be contained working on the cathode formulation by changing different features: the physico-chemical nature of the carbonaceous matrix used and/or the addition of different catalyst materials such as metal sulfides, metal oxides etc. However, the optimization of operating conditions generally requires several time-consuming and resource-demanding experiments. For this purpose, the employment of a computational model, able to describe the discharge/charge behaviour of the cell, and the effect of different operating conditions on the electrochemical performances, can be a suitable option to predict the behaviour of the system.

The present work has two main objectives: the first one is the experimental characterization of cathodes for LiS batteries, studying different compositions and formulations in order to optimize the electrochemical performances of the cells; the second one is the development of a physics-based model that can accurately describe all the elements present in a battery and all the phenomena that occur during charge and discharge processes at the pore-scale level. The model developed is a continuum physics-based model that describes the battery at the pore-scale, focusing on a realistic description of the porous cathode structure, which is represented by a separate domain, while the separator is modelled as a continuum domain using the Bruggeman correlation for the diffusion coefficients, and the anode consist of a lithium foil. The reaction chain is modelled with two chemical reactions, i.e., the dissolution of sulfur and the precipitation of the lithium sulfide, and five electrochemical reactions that characterize the reduction of the long-term chain polysulfide to sulfur anion. The two chemical reactions are modelled considering the difference between the concentration of the ions involved and the solubility product of the solid component as driving force; instead, the electrochemical reaction rate is evaluated from the current intensity generated from the reaction, which in turn is determined employing the Butler-Volmer equation. The model can predict the discharge behaviour of the cells and can describe the evolution during the discharge process of species concentrations, which can provide further insights on the shuttling phenomena and the sulfide precipitation, allowing to determine the effect of different conditions on these phenomena and on the electrochemical behaviour of whole system.

The preparation of the active material consists of different steps. First, sulfur and carbon are mixed together with a polyvinylidene difluoride (PVdF) binder using N-methyl-2-pyrrolidone (NMP) as solvent. In this step, three different features can be investigated: the mass ratio between sulfur, carbon and binder, the carbonaceous type employed and the technique used to mix all the materials. In our work, we started with a mass ratio of 60:30:10 between the three main components, progressively changing the amount of sulfur in order to increase the mass loading of the cathode to obtain a larger specific energy. The second feature investigated was the nature of the carbonaceous matrix, and two different types of conductive carbon, a mesoporous carbon and a nanofiber carbon, were considered. These two materials differ in morphology, porosity and specific surface which influence both the precipitation of lithium sulfides and the migration of polysulfides. Nanofibers show improvements for both effects, favoring the formation of small lithium sulfide particles which can be easily dissolved (avoiding the formation of lithium sulfide layer which are more insoluble) and, at the same time, enhancing the retention of sulfur inside the electrode. The third feature studied was the techniques employed to mix together the materials: a classical mechanical mixing performed by ball milling and a more peculiar method based on melt-infusion, which consists of melting the sulfur at high temperature enhancing its diffusion inside the pores of the conductive carbon. The melt-infusion method favours the retention of sulfur inside the electrode, decreasing the shuttling effect phenomena. After this step, the active material was cast on a carbon coated aluminium foil current collector using a Doctor Blade to control the thickness of the electrode, which was changed within 200 µm to 300 µm.

To assemble coin cells (2032), 15 mm diameter disks were cut from the cast material, to be used as cathode, while metallic lithium disks were used as anode, and 1 M LiTFSI in 1,3-dioxolane:dimethoxy ethane 1:1 (DOL:DME, 1:1) with 0.25 M LiNO₃ was used as electrolyte. The electrodes were dried in a vacuum oven before assembling inside an argon-filled glove box, and a commercial separator (Celgard 2500) was also used. The coin cells were used to perform electrochemical testing to characterize the cathodes produced and their electrochemical behaviour. Different types of tests were carried out: galvanostatic cycling and rate capability tests were employed to evaluate the cycle life of the cells and their behaviour at different current intensity, while cyclic voltammetry were used to evaluate thermodynamics and kinetics information about the redox processes. Other experiments can be employed: intermittent current interruption (ICI) was used to evaluate the evolution in time of the internal resistance and diffusion coefficient of the lithium ions in the electrode, potentiostatic intermittent titration technique is useful to obtain further information on the nucleation and growth of Li₂S crystals during discharge, field emission scanning electron microscope (FESEM) was used to characterize cathode morphology in order to represent more realistically the electrode structure. These tests can be employed to validate the model, comparing the results with the computational predictions, but also to inform the model and to describe more accurately the cell and the phenomena occurring during charge/discharge processes.

This work was funded by the Stellantis-CRF Polito framework agreement 2022-2026.