Difference of Relaxation Behavior Between LiNiO2 and Li(Ni, Co, Al)O2 after Lithium Extraction

Abstract

Introduction LiNiO2 (LNO) is one of the famous cathode materials of layered rock salt structure having larger discharge capacity in addition to lower toxicity and lower cost in comparison with LiCoO2. For the utilization of higher voltage region as the stable charge-discharge capacity, it is important to investigate the static and kinetic properties of this cathode material. We have investigated the structural variation of electrode materials after the termination of lithium insertion and/or extraction to reveal the structural transient from kinetic to equilibrium states clearly. We named this technique as “Relaxation analysis” and conducted this method to various types of cathode or anode materials, e.g. γ-Fe2O3 [1,2], LiMn2O4[3], LiFePO4[4], LiCoO2[5], or graphite [6], to clarify the lithium insertion and extraction process. We have recently carried out the relaxation analysis on LNO after charging to relatively higher voltage region (x ≤ 0.12 in the form of LixNiO2) [7], in which two phases with similar structures (Li-rich and Li-lean phases with the space group ) coexist [8,9]. It has been revealed that excess amount of Li-lean phase, formed during the lithium extraction process, separates into Li-rich phase and Li-lean phase with decreased lithium concentration. Partial substitution of Co and Al instead of Ni (Li(Ni, Co, Al)O2; NCA) is known to provide more stable charge-discharge cycle performance [10] and it is also expected that relaxation analysis would help to understand this improved property. In the present study, relaxation analysis is applied to NCA to compare the structural relaxation with the previously studied LNO. Experimental Working electrode was prepared by mixing Li(Ni0.933Co0.031Al0.036)O2 powder (Sumitomo Metal Mining Co., Ltd.) with Acetylene Black (AB) as a supplemental conductor and PVdF powder as an adhesive agent with the weight ratio of 80:10:10. Lithium foil and 1 mol∙dm-3 LiPF6 in EC/DMC (2:1 v/v%, Kishida chemical Co., Ltd) were employed as the counter electrode and the electrolyte, respectively to construct a two electrode cell (Hohsen Co.). Lithium ion was electrochemically extracted from NCA at a constant current of 0.01C rate to achieve x = 0.12, 0.09 and 0.06 of Lix(Ni0.933Co0.031Al0.036)O2. After the termination of Li extraction, we immediately removed the working electrode from the cell in a glove box to avoid the local cell reaction between the electrode material and the current collector. XRD data were collected from 15 ° to 75° in 2θ for CuKα radiaton (RINT-TTR, Rigaku Co., Japan) at various relaxation time and were served for Rietveld structure analysis using RIEVEC code [11]. Results and discussion Fig.1 represents the measured X-ray diffraction pattern for x = 0.09 of Lix(Ni0.933Co0.031Al0.036)O2, which is obtained 48 hours after the termination of lithium extraction. The Rietveld refined pattern assuming the two phases (Li-rich and lean phases) is also drawn. The measured and the calculated patterns agree well with sufficiently small R wp value. The calculated mole fractions of Li-lean phase are plotted in Fig. 2. For all samples, the molar ratio of Li-lean phase decreases with relaxation time, indicating that Li-lean phase partly transforms into the Li-rich phase for all the compositions. Such a behavior is essentially the similar to the previously reported LNO. The relaxation time variation of the lattice parameter c is represented in Fig. 3. While LNO possesses a large difference in c -length between Li-rich and lean phases, NCA has small difference, especially in x = 0.12. In addition, the relaxation time variations of the c -parameter are very small in NCA for all compositions. It is supposed that small difference in c-length between the Li-rich and lean phases as well as small variation of cduring the relaxation provide the better charge/discharge cycle performance in NCA even at high voltage region.