Voltage Fade of Layered Oxides: Its Measurement and Impact on Energy Density

Abstract
Voltage fade of layered, Li-intercalating transition metal oxides is caused by irreversible, structural changes. A method that uses a resistance-corrected average voltage is proposed to track and quantify voltage fade in a reproducible and time-efficient manner. it is used here to compare several layered oxides in terms of their degrees of fade. The materials studied include some that are of current technological importance, such as LiNi0.8Co0.15Al0.05O2 (NCA), Li1.05(Ni1/3Mn1/3Co1/3)0.95O2 (NMC), and Li1.2Ni0.15Mn0.55Co0.1O2, a Li- and Mn-rich NMC, also denoted as 0.5Li2MnO3•0.5LiMn0.375Ni0.375Co0.25O2, as well as some other Li-rich oxides with nano-composite structures. Electrochemical testing of these materials shows that voltage fade is common to many, if not all, layered oxides. For most materials, the decay rate of the resistance-corrected average voltage is on the order of a few millivolts per cycle, often with a slightly faster decay in the beginning. Particular attention is paid here to the Li- and Mn-rich NMC for which the rate of voltage fade is shown to increase when the oxide is cycled at elevated temperatures (55°C) and to high voltages (4.7 V vs. Li/Li+). After 20 cycles, voltage fade significantly reduces the material's energy output and can outweigh energy losses due to capacity loss and resistance rise under standard cycling conditions (i.e. 2.0 V–4.7 V vs. Li/Li+, 30°C). Despite its initial voltage fade, the Li- and Mn-rich NMC exhibits the highest oxide-specific energy density among the materials of interest. Voltage fade also appears to be related to the voltage hysteresis, which is particularly large in some of these materials. The interplay between reversible and irreversible processes, such as transition metal migration between the metal- and the Li-layer during the structural transformation, may explain some of the observed cycling characteristics.