Li y (Ni 0.425 Mn 0.425 Co 0.15 ) 0.88 O 2 materials were synthesized by a slow rate electrochemical deintercalation from Li 1.12 (Ni 0.425 Mn 0.425 Co 0.15 ) 0.88 O 2 during the first charge and the first discharge in order to study the structural modifications occurring during the first cycle and especially during the irreversible "plateau" observed in charge at 4.5 V vs Li + /Li. Chemical Li titrations showed that the lithium ions are actually deintercalated from the material during the entire first charge process, excluding the possibility that electrolyte decomposition causes the "plateau". Redox titrations revealed that the average transition metal oxidation state is almost constant during the "plateau", despite further lithium ion deintercalation. 1 H MAS NMR data showed that no Li + /H + exchange was associated to the "plateau" itself. Rietveld refinement of the XRD pattern for a material reintercalated after being deintercalated at the end of the "plateau", as well as redox titrations, revealed an M/O ratio larger than that of the pristine material, which is consistent with the oxygen loss proposed by Dahn and coauthors for the LiNi x Li (1/3-2x/3) Mn (2/3-x/3) O 2 materials to explain the irreversible overcapacity phenomenon observed upon overcharge. X-ray and electron diffraction showed that the transition metal ordering initially present within the slabs is lost during the "plateau" due to a cation redistribution. To explain this behavior a cation migration to the vacancies formed by the lithium deintercalation from the transition metal sites (3a) is assumed, leading to a material densification.

The search for new materials that could improve the energy density of Li-ion batteries is one of today's most challenging issues. Many families of transition metal oxides as well as transition metal polyanionic frameworks have been proposed during the past twenty years. Among them, manganese oxides, such as the LiMn2O4 spinel or the overlithiated oxide Li[Li1/3Mn2/3]O2, have been intensively studied owing to the low toxicity of manganese-based materials and the high redox potential of the Mn(3+)/Mn(4+) couple. In this work, we report on a new electrochemically active compound with the 'Li4Mn2O5' composition, prepared by direct mechanochemical synthesis at room temperature. This rock-salt-type nanostructured material shows a discharge capacity of 355 mAh g(-1), which is the highest yet reported among the known lithium manganese oxide electrode materials. According to the magnetic measurements, this exceptional capacity results from the electrochemical activity of the Mn(3+)/Mn(4+) and O(2-)/O(-) redox couples, and, importantly, of the Mn(4+)/Mn(5+) couple also.