On-Line Electrochemical Mass Spectrometry Investigations on Gassing Behavior of Li4Ti5O12 Electrodes and Its Origins

Abstract

Advanced lithium ion batteries with the use of Li4Ti5O12 (LTO) as unhazardous, thermal stable and “zero-strain” anode material are promising candidates for applications in the field of long-term energy storage systems. However, it is known from post-mortem gas analysis [1, 2] that LTO and organic carbonate electrolytes containing pouch bag cells swell during charging and storage due to evolving significant amounts of hydrogen gas. However, a correlation between gas evolution and cell chemistry couldn’t be formulated yet. In our study, we analyzed LTO-LiFePO4 (LFP) full-cells and LTO-Li half-cells with three different water contents of the electrolyte in-operando with On-Line Electrochemical Mass Spectroscopy [3] to detect evolved gases followed by post-mortem ATR-FTIR and x-ray diffraction analysis of charged LTO electrodes. The main focus of this work is on the side reactions promoted by added moisture to the electrolyte solution to evaluate proposed mechanisms given by He et al. as well as Amine and coworkers [1, 2]. As an example, Figure 1 shows the charging profile of an in-operando LTO/LFP full-cell containing 4000 ppm H2O. The voltage profile (green, left axis) follows the typical trend with its charge/discharge plateau at ≈ 2 V.Here, it is observed that the charging time takes longer than the expected 5 hours at a rate of C/5, equating to an overall “charge capacity” of 193,9 mAh/gLTO. The red and blue line (right axis) gives in-operando concentrations for hydrogen and carbon dioxide, respectively. During the preliminary OCV step, hydrogen or carbon dioxide evolution is not detected. Both gases evolve upon charging, whereas the H2 concentration approaches a plateau and the CO2 concentration decreases after reaching an initial peak value. The hydrogen amount after the charging process was calculated to be 6x10-6 mol after charging. If one considers the reduction potential for water and LTO in the Pourbaix diagram, the redox potential of the latter is below that of the H2 evolution reaction, which implies that water can be reduced in contact with a charged LTO anode. Thus, the evolution of hydrogen is enabled by a parasitic reduction reaction of water while reducting the Ti4+ species of LTO during Li+-ion intercalation in the charging step, which leads to an increase of the apparent “charging capacity” of LTO beyond its theoretically expected value. We propose a reaction mechanism [4] which corresponds to i) the H2O reduction process proposed by Belharouk et al. [2] and ii) the solvent reactions published by D. Aurbach [5] and P. Bruce and coworkers [6], respectively. The parasitic one-electron reduction of water leads to hydrogen gas evolution and the additional formation of solid carbonate compounds. These compounds were analyzed by FTIR-ATR and assigned to oligomers of general formula LiO2C(OC2H4)nOCO2Li. As a result, reduction of water triggers a decomposition reaction of the solvent. Moreover, gassing can only occur in charged batteries, wherein the LTO anode's potential is lower than that for the reduction of water (potential of pristine LTO electrodes with ≈3 V vs. Li/Li+ is above the H2O reduction potential). This is in accord with the evidence that no hydrogen can be detected in stored pouch bag cells without preliminary charge ([1]).