One way to obtain higher energy densities in lithium-ion batteries is to increase the nickel content in lithium nickel-cobalt-manganese oxide (NCM) cathode active materials (CAMs), ultimately approaching LiNiO2 (LNO). At the same time, a higher nickel content and specific capacity comes with more severe degradation of the CAM during battery cycling. Doping of the CAM with additional elements is a common approach to increase structural stability. For any CAM, comparisons of different dopants with each other and with undoped references are therefore commonly found in the literature.[1,2] In some cases, the effect of different dopants on the CAM particle morphology is also reported.[3,4] However, the influence of the process route of dopant introduction is rarely discussed. The most common ways are either the co-calcination of dopant source, lithium source and CAM precursor (pCAM), or the co-precipitation of dopant ions during pCAM precipitation. Impregnating the pCAM surface with a suspension of smaller dopant particles or a dopant solution, followed by solvent removal, can provide better dopant distribution in case of low dopant ion mobility or phase separation tendencies during calcination.Building on the previous finding that the primary particle morphology is the determining factor of LNO performance,[5,6] in this work, the primary particle morphology, but also dopant distribution in a series of 0.25 mol% Zr-doped LNO obtained from the three different process routes co-calcination, impregnation and co-precipitation, is investigated. This low dopant concentration is common in commercial CAM, as thus the content of redox-active transition metal content remains high to provide maximal energy density. Zr also tends to phase separate in the form of Li2ZrO3 at high concentrations, which is this way avoided as much as possible, since the chosen concentration is below the solubility limit of Zr in the LNO bulk. In this comparison, it is found that not the dopant itself, but the process route of its introduction affects the primary particle morphology of LNO, likely explained by the initial dopant localization early in the calcination.Differences in electrochemical performance, both in coin cells and long-term cycled pouch cells, as well as in specific capacitance and gas evolution are then also found to be due not to the dopant, as one would initially assume in comparisons of doped and undoped materials, but to the effect of the process route on LNO primary particle morphology.
Read full abstract