New Insights on Li4Ti5O12 Electrode/Electrolyte Interfaces: A X-Ray Photoelectron Spectroscopy and Scanning Auger Microscopy Study

Abstract

Spinel Li4Ti5O12 (LTO) is considered as a good alternative negative electrode material for Li-ion batteries [1], due to its negligible change of lattice parameter during insertion/extraction of lithium ions [2], resulting in a very low capacity decrease upon cycling. However, the reactivity of LTO toward commons carbonates based electrolytes has been evidenced by surface analysis [3] and an important gassing occuring at the electrode/electrolyte interface was reported [4]. Therefore it is essential to better understand the interfacial phenomenons. This work aims to achieve a precise understanding of the LTO electrode/electrolyte interfaces in relation with batteries electrochemical performances. The influence of various parameters (cycling temperature, electrode and electrolyte composition, cycling potential window) upon the solid electrolyte interphase (SEI) formation and dissolution through the first cycle was systematically investigated. The evolution of those interfaces after long cycling is currently being studied. Finally, Li4Ti5O12/LiMn2O4cells having potential assets in term of cost and safety [5], will be investigated, in order to point out the changes in the SEI formation due to interactions between the two electrodes. The samples are analysed by X-ray Photoelectron Spectroscopy (XPS) and Scanning Auger Microscopy (SAM), two complementary extreme surface characterization techniques (analysis depth 5-10 nm), operating at different spatial resolutions. The galvanostatic curves obtained for the Li4Ti5O12/Li half-cells first cycle at room temperature and at 85°C are presented on figure 1. The number of electrons exchanged per unit formula (Li4Ti5O12) at the end of the first discharge increases with the temperature indicating that more electrons are exchanged at higher temperatures to reach the same state of charge. The irreversible capacity also increases with the temperature at the first cycle. Those electrochemical results are directly correlated with the increase of the SEI thickness with the temperature, as observed by XPS and SAM experiments at the end of the first discharge. Indeed at room temperature, the two components doublets in the titanium spectrum (figure 2a) indicate a partial reduction of titanium ions consecutive to lithium insertion. The component at low binding energy in oxygen spectrum (figure 2b) is characteristic of oxygen ions in the LTO material, while the ones at higher binding energy reveal the presence of deposited species at the electrode surface. In addition to the components characteristic of PVdF binder and carbon black (CB) constituents of the electrode, carbon spectrum (figure 2c) presents components characteristic of C-O, O-C=O and CO3 chemical environments, confirming the formation of a SEI resulting from the degradation of electrolyte species. By contrast, at 85°C titanium and the component attributed to oxygen anions from the LTO lattice are not detected (figure 2c and 2d), meaning that the SEI is thicker than the XPS depth of analysis. The intensity decrease of the carbon black component (figure 2f) together with the intensity increase of C-O component, compared to room temperature, is consistent with an increase of the SEI thickness. The morphology of the electrode surface have been characterized by SEM (Figure 3a and 3c) and allow to identify LTO particles in light grey and carbon black aggregates in dark grey, for both temperatures. In contrast, the carbon elemental SAM mappings at room temperature (figure 3b) displays a higher carbon concentration on top of carbon black aggregates, whereas the mapping at 85°C (figure 3d) evidences a more homogeneous carbon distribution on top of the electrode, indicating a complete covering of the electrode surface by the SEI species. [1] J. B. Goodenough, Journal of the American Chemical Society 135 (4) (2013) 1167–1176. [2] K. Zaghib, Journal of Power Sources 196 (8) (2011) 3949–3954. [3] L. El Ouatani, Journal of The Electrochemical Society 156 (6) (2009) A468. [4] Y.-B. He, Scientific Reports 2. [5] I. Belharouak, Journal of the Electrochemical Society 159, A1165–A1170 (2012). Figure 1