Understanding and Modelling the Thermodynamics and Electrochemistry of Lithiation of Tin Sulfide Compounds As Novel Anode Materials for Lithium Ion Batteries

Abstract

Due to its theoretical capacity of 960 mAh/g, which is approximately two and a half times that of graphite (372 mAh/g), as well as its natural abundance, Sn is as an interesting anode active material to replace the currently used graphite in lithium ion batteries. However, Sn-based anodes have not been commercialized because they suffer from prohibitively large volume expansions due to the formation of brittle LixSn alloys during lithiation (>185 % for full lithiation to Li17Sn4). These volume changes lead to extensive crack formation, loss of electrical contact between particles, pulverization of the electrode during cycling, lithium trapping, and the growth of an unstable SEI. One way to take advantage of the high capacity of Sn and simultaneously overcome its degradation mechanisms is to use Sn-based chalcogenide compounds such as SnS2 and SnS as anode active materials. During lithiation of these compounds, a composite electrode structure consisting of active Sn particles embedded in an inactive Li2S buffer matrix is formed, which is expected to accommodate the mechanical stress of the alloying and de-alloying reactions during cycling.Although several authors have studied the lithiation of SnS2 via in-situ electron microscopy techniques, there are still many open questions regarding 1) the amount of lithium which can be intercalated into the layered structure of SnS2, 2) the structural changes of SnS2 during lithiation, and 3) if the reaction occurs via a one-phase or a two-phase process. Additionally, there are no published phase diagrams of the Li–Sn–S system at room temperature, which could be used to explore the changes in the electrode constitution during lithiation and de-lithiation. In fact, only Hwang et al. [1] modelled the phase equilibria in the Li–Sn–S at 0 K using density functional theory (DFT) calculations, but their data are incomplete since they omitted the Sn2S3, Li7Sn3 and Li5Sn2 phases and did not take into account the electrochemically and chemically observed solubility of lithium in SnS2. Therefore, in this work, the CALPHAD (CALculation of PHAse Diagrams) and computational thermodynamics methods were combined with key experiments to 1) understand the electrochemical lithiation of SnS2, 2) model the phase development during the equilibrium and non-equilibrium lithiation of the SnxSy tin sulfide compounds, and 3) evaluate the energetics of the competing lithiation reactions for the first time.Firstly, a thermodynamic description for the ternary Li–Sn–S system was developed by combining the descriptions of the Li–Sn system from Reichmann et al. [2] with that for the Sn–S system from Lindwall et al. [3] and the Gibbs free energy function for Li2S from the Scientific Group Thermodata Europe (SGTE) [4]. The ternary database was used to calculate the Li–Sn–S phase diagram at room temperature and simulate the titration curves for the equilibrium lithiation of the tin sulfide compounds. At the same time, homogeneous and heterogeneous Sn–S powder samples were processed as slurries, coated onto copper-foil substrates and subjected to electrochemical testing vs Li+/Li. The galvanostatic intermittent titration technique was used to characterize the quasi-equilibrium open circuit potentials of the electrodes during full and partial de-/lithiation in the potential range of 0.01-2.5 V. Additionally, a full profile, Rietveld refinement of the ex-situ XRD patterns of selected electrodes was performed in order to assess the phase development and structural changes of the electrode active materials during titration. These data were used to suggest a one-phase mechanism for the equilibrium lithiation of SnS2 after voltage relaxation. Finally, the electrochemical and crystallographic data were used to develop a new thermodynamic model to describe the intercalation of lithium ions into the layered SnS2 structure and calculate the measured open circuit voltage values. Using this description, it could be shown that lithiated SnS2 is indeed a metastable phase, which forms due to the energetic feasibility of the intercalation mechanism compared to the equilibrium conversion reaction.