Mechanism of Performance Improvement in LiNi0.5Mn1.5-XTixO4 High Voltage Spinel Full-cells paired with Graphite and Li4Ti5O12 Negative Electrodes

Abstract

Listen

Translate article

The LiNi0.5Mn1.5O4 (LNMO) high voltage spinel has been recognized as a tantalizing option for the cathode in next generation of Li-ion batteries. The combination of high operating voltage (4.75 V vs Li) and excellent rate capability, which arises from three-dimensional Li+ ion diffusion, makes this material particularly attractive for automotive applications. Recent studies, however, suggest that the most critical barrier for the commercialization of LNMO in Li-ion batteries is electrolyte decomposition and the concurrent degradative reactions at electrode/electrolyte interfaces, which consume active Li+ ions and reduce cycle life.[1,2] Elemental substitution for Ni and/or Mn in LNMO has been the most widely accepted strategy to control the crystallographic properties and stabilize the performance of high voltage spinel/Li half-cells. Despite demonstrating excellent half-cell performance, LiNi0.5-xMxMn1.5O4/graphite full-cells (x = 0.05 and 0.1, M = Fe, Co, Cu, Al, Ga, and Mg) deliver poor cycle live, similar to that of the pristine LNMO; i.e., capacity retention after 100 cycles falls to the range of 45 – 53 %, and shows similar rates of capacity fading ( Fig. 1).In contrast, Noguchi et al.[3] reported that Ti-substitution for Mn in LNMO improves its cycle life in full-cells paired with amorphous carbon anodes. They showed that the capacity retention of the LiNi0.5Mn1.5-xTixO4 full-cells improved with increasing Ti content in the range from x = 0 to 0.19. In our recent work[4], this improvement was also observed for higher Ti contents, up to x = 0.4, as shown in Fig. 2. The LiNi0.5Mn1.5-xTixO4/graphite full-cells also showed less oxidation of the electrolyte during cycling compared with that of Ti-free LNMO, as evidenced by higher Coulombic efficiency and lower self-discharge rates.[4] However, the improvement mechanism of the LiNi0.5Mn1.5-xTixO4full-cells has not been determined.In an attempt to identify the improvement mechanism operant in the LiNi0.5Mn1.5-xTixO4 full-cells, various surface and bulk analyses of cycle-aged and/or HF-etched cathodes were undertaken, including TEM, TOF-SIMS, FT-IR, and ICP in combination with AC-impedance analysis. Figure 3 shows that LiNi0.5Mn1.5-xTixO4 particle surfaces are Ti and O rich after etching in 1 wt% HF solution, and Mn and Ni deficient. In contrast, Ti-free LNMO does not show any elemental gradient after etching under the same conditions. This result suggests that the Ti-rich surface, which is formed as a result of the Mn dissolution, may passivate the LiNi0.5Mn1.5-xTixO4 during prolonged cycling. The detailed post-mortem analyses results for LiNi0.5Mn1.5-xTixO4will be discussed in comparison with those for Ti-free LNMO to support the above hypothesis.Since the LiNi0.5Mn1.5-xTixO4spinel is a relatively promising new material, its electrochemical performance needs to be optimized by tuning chemical compositions and synthesis parameters. Therefore, we will also discuss what influence these factors have on crystal structure, phase purity, and electrochemical performance.