Surface Passivation of Lithium-Ion Electrodes: A Path to High-Performance Energy Storage

The electrode/electrolyte interface is one of the most important determinants of the performance of a rechargeable battery. The manganese spinels offer a high-voltage cathode, but the surface reaction 2Mn3+ = Mn2+ + Mn4+ influences the surface phase formed at the sintering temperature; and at voltages greater than 4.3 V versus Li+/0, the cathode oxidizes the liquid organic electrolyte used in a Li-ion battery to create a passivation solid electrolyte interphase (SEI)1 that may impede or promote Li+ transfer across it. Coating the spinel surface with a designed SEI is not stable on cycling since the electrodes change volume during charge/discharge, so doping and thermal processing are used to create an SEI that can be stable over a long cycle life.2,3,4 However, this process is empirical without direct observation of how the doping and processing influence the structure and chemistry of the SEI that forms. We report some direct observations of the surfaces of the manganese spinels Li[Mn2]O4 (LMO) and Li[Ni0.5Mn1.5]O4 (LNM); they are charged, respectively at voltages versus Li+/0 ≥ 4.3 V and V ≥ 4.7 V. Figure 1 shows an aberration-corrected high-angle annular dark-field (HAADF) scanning transmission electron microscopy (STEM) image of an as-prepared LMO particle; the image reveals a thin, manganese-rich Mn2+[Mn23+]O4 surface phase (“ring” phase) resulting from the surface disproportionation reaction. EELS data in Figure 2 confirms the Mn2+[Mn23+]O4 surface phase and indicates a subsurface Li-rich Li[LixMn2-x]O4 spinel phase and that the bulk of the particle retains the LMO spinel phase. 5 This observation is consistent with a model of the surface disproportionation reaction and a displacement of the surface Li+ from the surface Mn2+[Mn2 3+]O4 to a subsurface phase. However, an Mn2+[Mn2 3+]O4 surface phase would block Li+ transport across it; the strong preference energy for tetrahedral-site Mn2+ therefore indicates a surface layer of Mn2+[Lix+Mn2-3x3+Mn2x4+]O4 with x ≤ 2/3. Figure 3 shows the HAADF STEM images of an LNM particle. This manganese spinel also forms a surface phase with tetrahedral-site Mn2+, the ring phase, but an additional Ni-rich rock-salt phase LiyNi1-yO phase can also be identified. The amount of the rock-salt phase increases with the loss of oxygen above 700°C; but the Ni2+ is reincorporated into the spinel phase on slow cooling or annealing in air below 700°C6 . We also examined the influence of increasing the Mn/Ni ration in the LNM phase. With increasing Mn concentration in Li[Ni0.5- xMn1.5+x]O4, a thicker surface ring phase and suppression of the Ni-rich rock-salt phase improved dramatically the high-temperature charge/discharge cycling. We also attempted to modify the surface of LMO and LNM samples with different chemical treatments including an acid treatment, a chemical delithiation, and a reactive oxygen plasma treatment in order to see if the new surface structures and chemistries would be beneficial or detrimental to cyclability. Electrochemical data had not been taken on the samples with surface treatments at the time of this report, so no conclusions about the surface structure’s effect on electrochemical performance can be made. The acid treatment changed the morphology of LMO to a highly porous material so HAADF STEM could not be used to identify the structure post treatment. Chemical delithiation removed Li from LMO as confirmed with inductively coupled plasma optical emission spectrometry (ICP-OES) and preserved the morphology of the original sample. Chemically delithiated LMO was then imaged with HAADF STEM (Figure 4). What is observed is the preservation of the bulk cubic spinel structure, the presence of the ring phase at the surface, and a rock-salt phase, which has not been previously observed in LMO samples except under damaging high-dose electron beam conditions. We believe the rock-salt phase in chemically delithiated LMO is due to electron-beam sensitivity, which doesn’t occur in LMO due to the presence of Li+ and their corresponding electrons in the Mn4+/3+ band. The oxygen plasma treatment applied to LMO preserves the morphology of the original sample, but creates a much thicker surface ring phase (Figure 5). We think the highly oxidizing nature of the oxygen plasma may lead to the oxidation of surface oxygen, leading to the loss of molecular oxygen from the sample. We have thoroughly characterized the surface of LMO and LNM related materials and have discovered the role of Mn, Ni, and O in the surface reconstruction of their respective materials. We have also identified effects in electrochemical performance that can be attributed to the surface structure of cathode materials. Chemical post-treatments were applied to LMO with varying results, proving the surface structure can be modified, which could lead to the formation of passivating surface phases. Understanding how the constituent elements and post chemical treatments affect the surface will lead to tailoring of the surface of the high voltage cathode materials and could ultimately lead to better cyclability