The Impact of Electrolyte Additives and Cycling Voltage on the Formation of a Rocksalt Surface Layer in LiNi0.8Mn0.1Co0.1 Electrodes

Abstract

High energy density lithium ion batteries that are cheaper, safer, and with longer lifetimes need to be developed in order to meet the increasing demand for applications such as electric vehicles. LiNi0.8Mn0.1Co0.1O2 (NMC811) can deliver a high capacity of ∼200 mAh/g with an average discharge potential of ∼3.8 V (vs. Li+/Li), making it a promising positive electrode material for high energy density lithium ion batteries. However, electrochemical tests of NMC811 from half cells and full cells show poor cycling performance when charged to potentials above 4.2 V. In our previous report1, it was shown the there are no significant structural changes in the bulk of the material during charge-discharge cycling. Instead, the parasitic reactions between the electrolyte and the highly reactive delithiated cathode surface at high potentials were suggested as the main reason for the failure of cells cycled above 4.2 V. Recently, Lin et al 2 showed that the surface of LiNi0.42Mn0.42Co0.16O2 went through a structural reconstruction from layered (Rm) to rocksalt (Fmm), which caused a significant increase in cell impedance under high voltage cycling conditions. Takamatsu et al 3 also reported that Co3+ at the surface of LiCoO2 was reduced to Co2+ after soaking in the electrolyte, however, the reduction of Co was suppressed with the presence of vinylene carbonate (VC) additives. This suggests that appropriate electrolyte additives might be able to suppress surface reconstructions of NMC materials. In this work, the impact of electrolyte additives and cell upper cut-off potential on the formation of a rocksalt surface layer in NMC811 cells was studied. NMC811/graphite pouch cells (220 mAh) were cycled for 100 cycles between 2.8 to 4.1 or 2.8 to 4.3 V. The rate used was C/5 for 5 cycles and followed by one C/20 cycle. The control electrolyte was 1M LiPF6in 3:7 v:v EC:EMC. The electrolyte additives studied in this work were 2% VC and “PES211”. PES211 is a blend of 2% prop-1-ene-1,3-sultone (PES) + 1% methylene methane disulfonate + 1% tris(trimethylsilyl) phosphate in control electrolyte. The cycled cells were discharged to 3.0 V and held for 24 h before disassembly in an argon-filled glovebox. The recovered positive electrodes were then washed with diethyl carbonate (DEC). Thick layers of carbon (~3 µm) and tungsten (~10 µm) were first deposited on the surface of the electrode to avoid beam damage during the FIB process. The samples were then analyzed with HAADF, EELS and NBD in aberration-corrected STEM mode. Figure 1A shows the HAADF-STEM images of the pristine NMC811 electrode near the surface before contacting any electrolyte. Every other column of the transition metal atoms, as indicated by the blue arrows, observed at the surface disappeared when moving into the bulk region, indicating a reconstructed rocksalt surface layer. This is likely due to the reaction between the electrode surface, that has high nickel content, with moisture in the air. The thickness of the rocksalt layer is ~2 nm. Figures 1B, 1C and 1D show the HAADF-STEM images of the electrodes after 100 cycles between 2.8 – 4.3 V with control, control plus 2% VC and control plus PES211 electrolyte in the cells respectively. The thickness of the surface layer on the control electrode was ~4 nm, while the thickness of surface layer in the electrodes in cells with 2% VC and PES 211 was about ~2 nm, almost the same as the pristine electrode. This suggests that both the VC and PES211 additives can suppress the formation of the rocksalt surface in NMC811 electrodes. Our previous report1showed that NMC811/graphite pouch cells with PES211 additives had worse cycling performance than the cells with only control electrolyte. Hence, at least for NMC811 cells, failure cannot only be ascribed to a growing rocksalt surface layer. Instead, other processes, for example associated with electrolyte oxidation, are believed to be responsible for failure. Figures 1E and 1F show the nano beam diffraction (NBD) from the surface and bulk of the control electrode shown in Figure 1B, respectively. The red dashed lines showed the remaining diffraction spots while the blue dashed lines showed the diffraction spots which disappeared when the beam moved from the bulk to the surface. This result confirms the that the surface was reconstructed. More results about EELS results will be discussed.