No abstract

Introduction  Lithium-ion batteries using Li4Ti5O12 anode realize excellent performance of power, fast-charge, life, and safety [1]. However, the theoretical capacity of LTO is low (175 mAh/g) compared to that of graphite anode. We have been developing TiNb2O7 (TNO) as high-capacity anode material in order to replace the LTO anode [2]. The theoretical capacity of TNO is 387.6 mAh/g according to 5 Li insertion per formula unit, regarding electron transfer about Nb5+/Nb3+, Ti4+/Ti3+redox couples. Only a few researchers have reported on a large storage of more than 4 Li insertion (310 mAh/g). Hence, we report the lithium storage mechanism and electrode performance of hydrothermally synthesized TNO particles with the large storage of more than 4 Li insertion. Experimental  TNO particle samples were synthesized by hydrothermal method. Niobium (V) chloride was dissolved in ethanol, and become mixed with Titanium (IV) oxysulfate dissolved in dilute sulfuric acid. Ammonia solution was slowly dropped to mixed solution with stirring until pH values were changed to 8. The mixture was transferred to autoclave, and gave a thermal treatment for 5 hours at 160℃. After that, the mixture was washed with deionized water and was frozen and dried in vacuum. The precursor samples were heated from 700℃ to 1000℃ for 30min, and furnace cooled. To compare with electrode performance of TNO, we also prepared TNO particle sample by solid state reaction using conventional method. Electrochemical properties were carried out using a three-electrode cell with synthesized TNO anodes as working electrodes, and Li foils as reference and counter electrodes. The structural changes of TNO particles as lithium insertion proceeded were characterized using ex-situ XRD, and XAFS measurements at the BL16B2 beam line in the SPring-8 synchrotron radiation research facility. Results and discussion  Fig.1 shows charge (Li insertion) - discharge (Li extraction) curves of TNO electrodes at the first cycle. The TNO particles prepared by solid state reaction had a reversible capacity of 306 mAh/g, which was correspond to 3.95 of Li insertion amount per formula unit. On the other hand, hydrothermally synthesized TNO particles heated at 1000℃ had a large reversible capacity of 341 mAh/g (4.40 of Li/ formula unit) by the increase in capacity below 1.0 V (vs.Li+/Li). The reversible capacity per volume calculated using the true density of TNO was 1481 mAh/cm3, which is twice higher than that of graphite anode. Fig.2 shows the dQ/dV-V plots at the first discharge curves. The hydrothermally synthesized TNO anode had an additional peak around 1.0 V (vs. Li+/Li), which was significantly different from solid state reaction particles. Primary particle size (ca.100 nm) of hydrothermally synthesized TNO was much smaller than that (ca. 1 μm) of TNO synthesized by solid state reaction as shown in Fig.3. We consider that a short diffusion length of lithium ions, a large surface area, and highly crystallization of thermally synthesized TNO particle would make it possible to insert into a storage site of lithium ions above 4.0 per formula unit. We will give a description about the Li insertion and extraction mechanism of TNO at below 1.0 V (vs. Li+/Li) by using electrochemical analysis, ex-situ XRD, and XAFS measurements. References [1] N. Takami, H. Inagaki, Y. Harada, Y. Fujita, K. Hoshina, J. Electrochem. Soc. 156, A128, 156 (2009). [2] Y. Harada, N. Takami, H. Inagaki, Y. Yoshida, JP Patent No 5230713 filed on Oct. 29, 2010. Figure 1

INTRODUCTION As Li ion battery cathode materials, stratified formation materials such as LiCoO2 in which the Co are used mainly now. The Co-free spinel compound LiMn2O4 has been studied as Li-ion battery cathode materials. Simple composition and low cost of Mn is preferable for the battery material. However, there are the problems such as elution of the Mn with the cycle and the Jahn-Teller distortion by the Mn3+.1,2) Characteristic remarkable deterioration in this way occurs for a cycle. As solution to this problem, characteristic improvement by the substitution of the metallic element (Al, Co, Cr, Li, Mg, Ni, Ti, Zn) was reported.3) It is intended to examine a crystal structure change during charge and discharge process of the Al substituted spinel using the average and local structure analysis with ex-situ method. The average structures were calculated by Rietveld analysis and electron density by Maximum Entropy Method (MEM) using the neutron and synchrotron X-ray diffractions. In addition, we examined the local structure by XAFS and the PDF analysis using synchrotron X-ray total scattering. EXPERIMENTAL LiMn2-xAlxO4 (x=0, 0.2) were synthesized by the solid-phase method. The samples were determined for the valence state of Mn by potentiometric titration, and also characterized by XRD and ICP measurements. Charge-discharge test was performed using the HS cell (3.5 ~ 4.3V vs Li/Li+, the negative electrode: Li metal, separator: polypropylene, electrolyte: 1M LiPF6/EC:DMC (1:2)). Measured sample was obtained by the electrode after two cycles of charge and discharge process. These samples were measured at 4.04V, 4.15V, 4.3V in the charge process, and 4.12V, 4.02V, 3.5V in the discharge process. These samples were performed by neutron diffraction (BL20, BL09, J-PARC) and synchrotron X-ray diffraction (BL02B2, SPring-8). We attempted to Rietveld analysis, PDF and XAFS analysis using the measured data, and examined the electronic structure and the distortion during the charge discharge process. RESULTS AND DISCUSSION At first we performed the Rietveld analysis using the neutron diffractions about each sample of Al substituted material at charge process after 2 cycles. As a result, the occupancy of Li at 8a site changed with charge discharge process, and the occupancy of Li at 16d site was constant. Thus it was thought that only the Li at 8asite participates in electrochemical insertion and extraction. In addition, distortion parameter λ in the distance calculated from each analysis showed few changes. As for the Al　substituted spinel, it was suggested that the host structure was stable in 2nd cycles during charge discharge process. We performed Rietveld analysis by the synchrotron X-ray diffraction measurement and calculated electron density in MEM and demanded a line profile of M(Mn,Al)-O. The line profile of charge discharge process of LiMn1.8Al0.2O4showed Fig. 1. It was demonstrated that the M(Mn, Al)-O bond of the Al substituted sample was more stable than unsubstituted one in charge discharge process. We examined XANES and EXAFS. The EXAFS spectrum of LiMn2O4 and LiMn1.8Al0.2O4were shown Fig. 2. From EXAFS spectrum, the relaxation of the Jahn-Teller distortion of the Mn was suggested, because peak intensity at charge state increased. Furthermore, we examined the PDF by synchrotron X-ray total scattering. In unsubstituted sample at discharge, the decrease in peak was confirmed in comparison with before charge. It was thought that this decrease was attributed to the elution of Mn. This study reported in detail on that day. It was performed by help of NEDO (RISING) and shows thanks to the members concerned. References 1) H. Mao, J.N. Reimers, Q. Zhong and U. von Sacken, Electrochem, Soc., Prec., 94-28, 245 (1994). 2) M.M. Thackeray, A. De Kock, M.H. Rossouw, D. Liles, R. Bittihn and D. Hoge, J. Electrochem. Soc., 139, 2, 363 (1992). 3) Y. Idemoto, K. Horiko, Y. Ito, N. Koura, and K. Ui, Electrochemistry, 72, 755 (2004).