The rechargeable Li-O2 battery has attracted interest due to its high theoretical energy density (about 10 times better than today’s Li-ion batteries). In this PhD thesis the cycling instability of the Li-O2 battery has been studied. Degradation of the battery has been followed by studying the interface between the electrodes and electrolyte and determining the chemical composition and quantity of degradation products formed after varied cycling conditions. For this in-house and synchrotron based Photoelectron Spectroscopy (PES) were used as a powerful surface sensitive technique. Using these methods quantitative and qualitative information was obtained of both amorphous and crystalline compounds. To make the most realistic studies the carbon cathode pore structure was optimised by varying the binder to carbon ratio. This was shown to have an effect on improving the discharge capacity. For Li-O2 batteries electrolyte decomposition is a major challenge. The stability of different electrolyte solvents and salts were investigated. Aprotic carbonate and ether based solvents such as PC, EC/DEC, TEGDME, and PEGDME were found to decompose during electrochemical cycling of the cells. The carbonate based electrolytes decompose to form a 5-10 nm thick surface layer on the carbon cathode during discharge which was then removed during battery charging. The degradation products of the ether based electrolytes consisted mainly of ether and carbonate based surface species. It is also shown that Li2O2 as the final discharge product of the cell is chemically reactive and decomposes carbonate and ether based solvents. The stability of lithium electrolyte salts (such as LiPF6, LiBF4, LiB(CN)4, LiBOB, and LiClO4) was also studied. The PES results revealed that all salts are unstable during the cell cycling and in contact with Li2O2. Decomposition layers thinner than 5 nm were observed on Li2O2. Furthermore, it is shown that the stability of the interface on the lithium anode is a chief issue. When compared to Li batteries (where oxygen levels are below 10 ppm) working in the presence of excess oxygen leads to the decomposition of carbonate based electrolytes to a larger degree.

Les systèmes électrochimiques lithium-air sont des concepts naissants mais exhibent des performances théoriques intéressantes qui laissent espérer une rupture technologique dans le domaine des batteries pour véhicule électrique. La possibilité d'atteindre une densité d'énergie supérieure à 500 Wh kg-1 est effectivement en ligne de mire. A contrario de la technologie lithium-air anhydre, les systèmes lithium-air aqueux n'ont, jusqu'à présent, fait l'objet d'aucune étude approfondie. Ce travail concerne donc le développement d'un système lithium-air aqueux, à trois électrodes, et vise également à améliorer nos connaissances fondamentales dans le domaine. La présente étude se focalise sur le compartiment positif de la cellule, dans lequel les réactions de l'oxygène sont mises en jeu. Dans un premier temps, une électrode spécifiquement dédiée à la réaction de dégagement d'oxygène a été élaborée à partir d'un acier 316L. L'étude de son comportement a révélé une bonne propension à catalyser la réaction de dégagement d'oxygène ainsi qu'une bonne stabilité sur 3 000 heures de fonctionnement. Néanmoins, d'importants problèmes de catalyse ont pu être observés et attribués à la présence des ions lithium dans l'électrolyte de la batterie. Les ions Li+ bloquent les transitions électrochimiques des sites actifs à l'origine des propriétés d'électrocatalyse. Le comportement d'une électrode à air, composée de carbone et d'oxydes de manganèse, a par la suite été caractérisé dans ce milieu. L'étude révèle deux phénomènes importants réduisant la performance de l'électrode et dont l'origine a également été attribuée aux ions lithium : un blocage des transitions (MnIII/MnIV), et une stabilisation des groupements oxygénés à la surface du carbone. Pour finir, il a été proposé d'optimiser le système électrolytique en limitant l'activité des ions Li+ en solution et ainsi d'améliorer le rendement en potentiel de charge/décharge de la batterie. / The electrochemical lithium-air devices are emerging concepts and their very high theoretical performances have attracted a lot of attention, especially for an application in the electrical vehicle. A target of at least 500 Wh kg-1 is aimed for. The aqueous lithium-air devices have not yet been studied in detail which is not the case for the anhydrous lithium-air technology. This thesis firstly deals with the development of an aqueous lithium-air cell based on a three electrodes setup, and secondly attempts to improve our theoretical knowledge of these systems. This study particularly focuses on the positive compartment of the cell in which oxygen reactions occur. The first section is dedicated to the development of an oxygen evolution electrode made of 316L stainless steel. The study shows its ability to catalyze the oxygen evolution reaction as well its good stability over 3 000 hours of operation. Nevertheless, major problems of catalysis were observed and assigned to the presence of lithium ions in the electrolyte. Li+ ions inhibit the electrochemical oxidation of the active sites, which are needed for the electrocatalytic properties. The behavior of an air cathode was characterized in the same medium. The results show two important phenomena which reduces the electrode performance and which are also attributed to lithium ions : a deactivation of the (MnIII/MnIV) transition and a stabilization of the oxygenated groups at the carbon surface. Finally, an optimization of the electrolytic system was proposed by limiting the Li+ ions activity in solution, which improves the charge/discharge potential efficiency of the battery.