The use of Li-ion batteries in portable electronic products is today widespread and on-going research is extensively dedicated to improve their performance and energy density for use in electric vehicles. The largest contribution to the overall cell weight comes from the positive electrode material, and improvements regarding this component thereby render a high potential for the development of these types of batteries. A promising candidate is LiNi_0.5Mn_1.5O₄ (LMNO), which offers both high power capability and energy density. However, the instability of conventional electrolytes at the high operating potential (~4.7 V vs. Li⁺/Li) associated with this electrode material currently prevents its use in commercial applications.

This thesis work aims to investigate practical approaches which have the potential of overcoming issues related to fast degradation of LNMO-based batteries. This, in turn, necessitates a comprehensive understanding of degradation mechanisms. First, the effect of a well-known electrolyte additive, fluoroethylene carbonate is investigated in LNMO-Li₄Ti₅O₁₂ (LTO) cells with a focus on the positive electrode. Relatively poor cycling performance is found with 5 wt% additive while 1 wt% additive does not show a significant difference as compared to additive-free electrolytes. Second, a more fundamental study is performed to understand the effect of capacity fading mechanisms contributing to overall cell failure in high-voltage based full-cells. Electrochemical characterization of LNMO-LTO cells in different configurations show how important the electrode interactions (cross-talk) can be for the overall cell behaviour. Unexpectedly fast capacity fading at elevated temperatures is found to originate from a high sensitivity of LTO to cross-talk.

Third, in situ studies of LNMO are conducted with neutron diffraction and electron microscopy. These show that the oxygen release is not directly related to cation disordering. Moreover, microstructural changes upon heating are observed. These findings suggest new sample preparation strategies, which allow the control of cation disorder without oxygen loss. Following this guidance, ordered and disordered samples with the same oxygen content are prepared. The negative effect of ordering on electrochemical performance is investigated and changes in bulk electronic structure following cycling are found in ordered samples, accompanied by thick surface films on surface and rock-salt phase domains near surface.

This thesis presents experimental investigations of the electronic structure of emerging and novel cathode materials used in lithium- and sodium-ion batteries. The investigated materials include a range of oxide materials containing the elements nickel and manganese. Central goals are to find fundamental explanations for favorable, respectively, unfavorable electrochemical cycling behavior and to arrive at a better understanding of the roles that the different elemental constituents of the compounds play. The experiments are based on the application of X-ray Absorption Spectroscopy (XAS) and Resonant Inelastic X-ray Scattering (RIXS) in the soft X-ray region and have been performed at synchrotron radiation facilities such as The Advanced Light Source (USA), The Swiss Light Source (Switzerland) and SPring-8 (Japan).

 XAS and RIXS of spinel LiNi_0.44Mn_1.56O₄ at the O K-edge as well as the Ni and Mn L-edges were measured for two different crystal structures, namely, transition-metal-ordered and -disordered, respectively. The results show that both Ni and O contribute strongly as redox centers for the charge compensation during electrochemical cycling. The Ni L-RIXS spectra show evidence of a more stable Ni--O bond in the disordered material.

 In the layered manganese oxide materials Li[Li_0.2Ni_0.2Mn_0.6]O₂, Na_0.67[Mg_0.28Mn_0.72]O₂, and Na_0.78[Li_0.25Mn_0.75]O₂, as well as the disordered Li_1.9Mn_0.95O_2.05F_0.95 one observes that reversible O redox leads to two distinct features in O K-RIXS. Both features resonate in a narrow incident energy range suggesting that localized O hole states are formed, one close to the elastic peak and the other as a strong emission peak at an energy loss of about 8 eV. These features appear reversibly on the voltage plateau of the charge-discharge curve and can be used to identify a certain type of O redox reactions.

The work also includes investigations that compare two different compositions of the structurally related material Li₂MnO₃ grown epitaxially as thin films. Evidence is found for anionic activity during the initial cycle that is of a different kind than the above as no evidence for localized O holes is found. Instead, excess Li in the transition metal layer is shown to lead to a more rapid loss of covalency in the Mn--O bonds.

In short, this work presents some of the first explorations into the role of different types of anionic redox centers in cathodes, by means of XAS and RIXS thereby also demonstrating the utility and power of synchrotron based techniques for gaining atomic-level understanding of battery electrode materials.