235
In-House Valence Evaluation of Cathode Active Materials in Lithium-Ion Batteries

Wednesday, 6 March 2019
Areas Adjacent to the Forum (Scripps Seaside Forum)
T. Yoneda, K. Tantrakarn, T. Izumi, T. Omori, S. Tokuda, S. Adachi, K. Sato (Shimadzu Corporation), M. Kobayashi, T. Mukai, H. Tanaka, and M. Yanagida (Nat'l Inst. of Advanced Industrial Science and Technology)
In the development of lithium-ion batteries (LIBs), cathode materials are being further improved to achieve not only higher energy capacity but also longer cycle life. Analyzing the valence of the cathode active materials during charge and discharge cycles is one of the practical methods to understand the redox mechanism and degradation mechanism, which can help in achieving higher energy capacity and longer life. The valence has been evaluated using X-ray absorption fine structure (XAFS) measurements with a synchrotron X-ray source. To conduct in-house analyses of valence, we developed a polychromatic simultaneous wavelength dispersive X-ray fluorescence (PS-WDXRF) spectrometer, which is comprising an X-ray tube, a slit, a flat analyzing crystal, and a silicon strip detector (SSD) [1]. The X-ray dispersed by the crystal is simultaneously detected by the SSD, and the detected signals at each channel provide intensity information at corresponding energy. Because there are no moving parts in the optical setup, the spectrometer is possible to detect the chemical changes, including valence changes, of 3d transition metals with high precision.

Two PS-WDXRF spectrometers were developed. One spectrometer comprises a 1280-channel SSD and a Ge (220) crystal to detect X-rays in the range around 5.38 - 6.52 keV, which is designed for manganese. Another spectrometer comprises same components with another one and the range of detecting X-ray energy is 6.27 – 7.89 keV, which is designed for cobalt and nickel.

The first step was to obtain and analyze data of pressed powders of manganese oxides using the spectrometer for manganese. Thus, powders of MnO (II), Mn2O3 (III), MnO2 (IV), and KMnO4 (VII) were placed in the evacuated chamber of the experimental setup. An X-ray tube voltage of 20 kV and a current of 100 mA was applied, and the target was made from Rh and no filter was used. A 5-min measurement was repeated five times for each sample. The PS-WDXRF analysis showed that the specimens achieved clearly different peak energies which belong to more fine electron transition fluorescence lines. This means that the WDXRF has sufficient energy resolution, and the valence information is obtained by the peak energy shift derived from the chemical shift [2, 3].

The second step was to obtain and analyze data of pressed powders of cobalt oxides and nickel oxides using the spectrometer for cobalt and nickel. Thus, powders of CoO(II), LiCoO2(III), NiO(II) and LiNiO2(III) were placed in the evacuated chamber of the experimental setup. An X-ray tube voltage of 20kV and a current of 100 mA was applied, and the target was made from W and no filter was used. This experiment demonstrated the discriminability of fluorescence peak energy shift derived from the chemical shift, in spite of the worse channel resolution. These results indicate that there is a potential for valence evaluation of cathode materials in a short time with highly precise energy resolution.

These spectrometers were applied to the evaluation of actual cathode active materials in LIBs which is including manganese, cobalt or nickel. Some different states of cathode materials were prepared and the valence estimation was performed using the PS-WDXRF spectrometers. The relationship between the valence result and state of charge/discharge was confirmed, which shows that the valence number of LIBs cathode material is in accordance with the reaction formula.

The recent results of valence evaluation on actual cathode active materials in LIBs will be presented in detail.

[1] K. Sato, A. Nishimura, M. Kaino, S. Adachi, X-Ray Spectrom. 46, 330–335 (2017).
[2] J. Kawai, M. Takami, C. Satoko, Phys. Rev. Lett. 65, 2193–2196 (1990).
[3] K. Sakurai, H. Eba, Nucl. Instr. Meth. Phys. Res. B199, 391–395 (2003).