Lattice Parameter Hysteresis in Li- and Mn-Rich Layered Oxides and Its Dependence on State of Charge and Open Circuit Voltage

Abstract

Li- and Mn-rich layered oxides, also known as high energy NCM (HE-NCM, Li1+x[NiCoMn]1-xO2 with typically 0.1 < x < 0.2), are promising next-generation cathode active materials (CAMs) for Li-ion batteries, since they contain intrinsically more lithium ions than regular NCM materials (x < 0.05). The gain in capacity is accompanied by the drawbacks of voltage hysteresis (between charge and discharge) and voltage fade (during long-term cycling).1 The voltage hysteresis is a unique property of over-lithiated layered oxides, which does not only manifest itself in the voltage profile, but also in the cathode resistance,2 the redox activity of transition metals (TM) and oxygen,2 and the bond distances.3 Figure 1 shows our results of an in-situ X-ray diffraction (XRD) experiment during the initial cycles of a Li-rich NCM material. As seen in Figure 1a and b, the state of charge (SOC, equivalent to total xLi) dependence of the bulk lattice parameters a and c after the first activation cycle is characterized by a pronounced hysteresis between charge and discharge. Here, the hysteresis loop of the lattice parameter a is clearly different from that of c. Furthermore, the hysteresis of the lattice parameter a vanishes when it is correlated to the open circuit voltage (OCV) of the respective SOC (see Figure 1c), but c maintains the shifted maxima during charge and discharge (see Figure 1d). Based on the literature, these observations could be either explained by a path dependency of the reversible TM migration between TM and Li layer,4 the oxygen redox,2 or a combination of both.5 In addition to in-situ XRD data, we will show co-refined X-ray and neutron diffraction data of selected ex-situ samples, which have either the same SOC or OCV. By means of diffraction techniques, it is our goal to elucidate the lattice parameter variations in Li-rich CAMs and to clarify the often described contribution of TM migration to the hysteresis effects. Figure 1: In-situ XRD measurement of HE-NCM during the first 3 cycles. The pouch cell was cycled at a rate of C/10 between 2.0-4.8 V against metallic lithium with intermediate OCV steps to collect XRD patterns (at a laboratory diffractometer with Mo-Kα1 radiation). Using the trigonal model (space group R-3m), the lattice parameters a and c are shown as a function of SOC (a,b) and OCV (c,d), respectively. Acknowledgements: We want to acknowledge BASF SE for the support within the frame of its scientific network on electrochemistry and batteries and the BMBF (Federal Ministry of Education and Research, Germany) for its financial support within the ExZellTUM II project (grant number 03XP0081). We also thank the research neutron source FRM II for access to the instrument SPODI.