Thimothée Fabre PH'D Defense

Batteries have a major role to play in the ecological transition, but even today many safety issues are regularly observed for the most commercialized system, lithium-ion batteries. One way to improve safety is to replace the flammable organic electrolyte with a water-based alternative.
However, due to its narrow theoretical electrochemical stability window of 1.23 V, water alone is not suitable for Li-ion battery applications. This limitation can be overcome by using a "water-in-salt" electrolyte (WISE) in which a lithium salt is introduced in very large quantity, exceeding thesolvent in mass and volume. Indeed, a high salt concentration significantly decreases the molecular activity of water by reducing the amount of so-called free water molecules, thus preventing the water decomposition reactions, the hydrogen evolution reaction (HER) and the oxygen evolution reaction (OER). Thus, the electrochemical stability window of the electrolyte can be extended up to 3 V with a high concentration of 21 mol/kg lithium
bis(trifluoromethanesulfonyl)imide (LiTFSI). Such concentration drastically modifies the interactions between the molecules of the solution, leading to a new nano-structural
arrangement responsible for a high ionic conductivity (10 mS/cm) despite a non-negligible viscosity. Although demonstrating promising properties, batteries made with these electrolytes
suffer from instabilities and numerous degradations leading to drying, corrosion and finally cell failure. The scientific literature on the subject is mainly directed towards the addition of additives, the combination of salt, or even hybridib creation with an organic solvent in order to improve cycling performances. The thesis work proposed here, is rather interested in a fundamental study based on a complete LiFePO4/TiS2 cell in the reference electrolyte 21 mol/kg of LiTFSI. A multi-scale model of the dynamics of water molecules in the electrolyte was proposed and validated by combining the results from pulsed field nuclear magnetic resonance with those of quasi-elastic neutron scattering, providing an in-depth understanding of their mobilities. Concerning the degradation mechanisms of the complete cell, the state of health of the active materials could be studied using multiple characterization techniques at each stage of the cycling test, from cell assembly, with contact with the electrolyte, to post-mortem analysis. Particular interest was paid to surface reactions with the use of X-ray photoelectron spectroscopy, and in particular on the formation of passive layers before and after the first charge, whose robustness and composition are crucial to gain stability. By successively modifying the cycling parameters, the causes of capacity increse or decrease can be associated with certain degradation mechanisms. The in-depth study of cycling curves and potential profiles of complete cells led to the detection of key parameters often neglected such as the balancing between the electrodes, depending on their
operating potential. The addition of a LiFePO4-based reference electrode also allowed the decorrelation between the electrochemical signature of the positive electrode and that of the
negative electrode. Eventually, the analysis of gas production allowed to establish that the main reaction responsible for the high irreversibility and instability of the battery cycling seems to be HER at the negative electrode, despite operating within the supposed electrochemical stability window of the electrolyte. Other parasitic reactions were also detected, providing a better understanding of the poor performance of lithium-ion aqueous batteries based on water-in-salt.