The cycling performance and in operando gas analysis of LiNi0.5Mn1.5O4 (LNMO)/graphite cells with reasonably high loading, containing a "standard" carbonate-based electrolyte is reported. The gas evolution over the first couple of cycles was thoroughly investigated via differential electrochemical mass spectrometry (DEMS), neutron imaging and pressure measurements. The main oxidation and reduction products were identified as CO2, H2 and C2H4. In different sets of experiments graphite was substituted with delithiated LiFePO4 (LFP) and LNMO with LFP to distinguish between processes occurring at either anode or cathode and gain mechanistic insights. Both C2H4 and H2 were found to be mainly formed at the anode side, while CO2 is generated at the cathode. The results from DEMS analysis further suggest that the Ni redox couples play a profound role in the evolution of CO2 at the LNMO/electrolyte interface. Lastly, it is shown that the cycling stability and capacity retention of LNMO/graphite cells can be considerably improved by a simple cell formation procedure.

Gas generation as a result of electrolyte decomposition is one of the major issues of high-performance rechargeable batteries. Here, we report the direct observation of gassing in operating lithium-ion batteries using neutron imaging. This technique can be used to obtain qualitative as well as quantitative information by applying a new analysis approach. Special emphasis is placed on high voltage LiNi0.5Mn1.5O4/graphite pouch cells. Continuous gassing due to oxidation and reduction of electrolyte solvents is observed. To separate gas evolution reactions occurring on the anode from those associated with the cathode interface and to gain more insight into the gassing behavior of LiNi0.5Mn1.5O4/graphite cells, neutron experiments were also conducted systematically on other cathode/anode combinations, including LiFePO4/graphite, LiNi0.5Mn1.5O4/Li4Ti5O12 and LiFePO4/Li4Ti5O12. In addition, the data were supported by gas pressure measurements. The results suggest that metal dissolution in the electrolyte and decomposition products resulting from the high potentials adversely affect the gas generation, particularly in the first charge cycle (i.e., during graphite solid-electrolyte interface layer formation).

Continuous destruction of the solid electrolyte interphase (SEI) on the graphite-based negative electrode during cycling operation is a significant degradation mechanism that raises safety concerns and limits the cycle life of LiNi0.5Mn1.5O4 (LNMO)/graphite full-cells. Herein, we report on gassing phenomena which are typically concomitant with SEI destruction processes. Abrupt H2 evolution is observed by differential electrochemical mass spectrometry and pressure measurements at the end of discharge. Using a lithium reference electrode reveals that the gassing, which intensifies with cycling, is caused by an increase in the anode potential. Lithium is irreversibly consumed upon SEI formation, but this loss is not compensated for by the intrinsic degradation of LNMO in the first cycle. When the potential of the anode on discharge increases above approximately 0.9 V, the SEI is instantly damaged, causing gas generation, and eventually capacity fade. We show that this (“mediator-free”) cross-talk phenomenon can be suppressed to various degrees by either using a precycled graphite electrode or an LNMO material having a higher initial irreversible capacity loss.

Aging in Li-ion batteries is accompanied by structural and compositional changes of the active and passive material of the electrodes. Visualization of spatial correlation between morphological properties and chemical phase can help to find origins of aging and failure mechanisms and to understand the ongoing processes. Micro X-ray fluorescence (XRF) analysis allows for chemical imaging of elemental distributions of multiple 3d elements simultaneously, or distribution of the state of charge even at low concentrations or under operando conditions. Utilizing focused synchrotron radiation, representative volumes of the electrodes can be studied with sub-micrometer resolution on acceptable time scales. Cutting edge instrumentation in this field is the Maia fluorescence detector. It allows for ultra-fast mapping of areas of multiple mm² with high spatial resolution within minutes. This allows e.g. examination of large electrode areas. The backscattering geometry of the detector enables easy mounting of flat samples such as electrodes or pouch-cells. By variation of the energy in the vicinity of an absorption edge, conclusions on local variations of the state of charge can be drawn, probing the oxidation state of the redox-active element.  Furthermore, confocal XRF enables spatially resolved measurements of the deposition of transition metals on the graphitic anode without destruction. This will help to understand the impact on capacity fade and degradation. For the present study, samples of LiNi0.5Mn1.5O4 electrodes cycled vs. Li or graphite were investigated at beamline P06 at DESY, Hamburg Germany.1 The experiment revealed significant inhomogeneities in the local Ni/Mn ratio upon cycling which is correlated to thinning of the electrode. Higher cycling rates showed an increased impact and higher material losses. Inhomogeneities in the phase transformation were also found during the charge and discharge process. The graphitic anodes showed significant amounts of deposited transition metals, few wt%, both Ni and Mn, deeply penetrating into the pores of the electrode. On the other hand, a study on commercial LiFePO4 cathodes with outstanding rate capabilities showed an extremely high degree of homogeneity throughout the investigated areas. These methods are by no means limited to the investigated materials, but are easily transferred to other transition metal containing electrode materials. 1. U. Boesenberg et al., Chem.Mater., 27, 2525–2531 (2015)