Experimental degradation study of a commercial lithium-ion battery

Abstract

In this study, we analyze data collected during the aging of 196 commercial lithium-ion cells with a silicon-doped graphite anode and nickel-rich NCA cathode. The cells are aged over a large range of calendar and cycle aging conditions. For different cells, these conditions are constant, alternating or randomly changing. The total test time and reached cycles are 697 days and 1500 equivalent full cycles for the calendar and cyclic aging tests respectively. A periodic check-up procedure was consistently performed at 20°C controlled temperature, which allows for good comparability between different aging scenarios. In addition to capacity fade and resistance increase, we provide data of the degradation modes LLI, LAMNE, and LAMPE obtained from half-cell fitting of the full cell’s and both electrode’s differential voltage. During calendar aging, we quantify the influence of the check-up procedure, which significantly increases the aging observed after 697 days. For certain conditions, the degradation induced by the check-up even exceeds the pure calendar aging, which reveals that the lifetime of lithium-ion batteries was underestimated in previous studies and models. Fitting of the Arrhenius equation on calendar aging at four temperatures at different SOCs shows a varying activation energy, which lies between 23.6kJmol−1 and 29.9kJmol−1 for the capacity. The activation energy of resistance increase is almost twice as high. The influence of SOC on the calendar aging is highest at about 85% SOC, presumable due to a cathode-driven shuttle mechanism. Next to current rates, temperature and mean SOC, the depth-of-discharge has the highest impact on cycle aging. Additionally it is shown that lower charging rates favor the onset of accelerated aging in certain conditions. For all tests, the raw data is made publicly available, which fosters the development of data-based aging models or state-estimation algorithms.