2016 Volume 63 Issue 7 Pages 679-683
Spinel-type LiNi0.5Mn1.5−xTixO4 powder library was prepared using the electrostatic spray deposition (ESD) method. Deposited powder was heat-treated at 700 °C for 3 hours in air. Phase identification of the X-ray diffraction patterns of the LiNi0.5Mn1.5−xTixO4 (0 ≤ x ≤ 0.5) prepared showed multi-phases of spinel, a slight amount of rock-salt-type LiyNi1−yO. In first charge-discharge capacities of LiNi0.5Mn1.5−xTixO4 (0 ≤ x ≤ 0.5) at current rates of 0.1 C and 1 C, LiNi0.5Mn1.5O4 (x = 0) and LiNi0.5Mn1.4Ti0.1O4 (x = 0.1) each showed a discharge capacity of about 100 mAh/g. LiNi0.5Mn1.4Ti0.1O4 was also found to retain its discharge capacity at a higher rate than LiNi0.5Mn1.5O4. The charge-discharge capacities were lower than those observed in the previous report, however, because the powders obtained included LiyNi1−yO type oxides as a second phase. Also, it was found that the decrease of rock-salt phase (LiyNi1−yO) was able to control by changing the atmospheric and powder shape-forming conditions.
Lithium-ion secondary batteries are used in many portable electronic devices and serve as environment-friendly power sources for electric vehicles (EVs) and hybrid electric vehicles (HEVs)1). Among the various kinds of cathode materials so far reported for these batteries, spinel-type LiMn2O4 compounds are composed of raw materials found in abundance in nature. LiMn2O4 is well known to exhibit three-dimensional diffusion paths of lithium ions and a high energy density. Its cyclic performance, however, is seriously degraded by Mn dissolution2,3) and Jahn-Teller distortion4,5). As a workaround, oxides partly substituted with other transition metal ions have been prepared to improve the problem. The compounds thus obtained showed decreased amounts of dissolved Mn in electrolyte and improved cyclic performance6). The capacity or voltage plateau of LiMxMn2−xO4 (M; Cr, Co, Fe, Cu, Al, Mg, and Ni) depends the type of transition metal7–12). LiNi0.5Mn1.5O4 showed a higher discharge capacity and high voltage plateau at around 4.7 V8). Further, the cyclic performance of Ti-substituted LiNi0.5Mn1.5−xTixO4 was markedly improved13–15).
In many cases, LiMxMn2−xO4 materials were prepared by the solid-state reaction method or sol-gel method. Powders prepared by the solid-state reaction method have a smaller specific surface area than powders prepared by other processes. In the sol-gel process, it can be difficult to control the homogeneity of the lithium component of the prepared powder. In this study we focused on the electrostatic spray deposition (ESD) method, one of the solution-based processes. The ESD method is effective for obtaining a homogeneous powder because solutions charged with high voltages are sprayed and dried immediately on grounded and heated substrates. Theorefore, in the present study we prepared a LiNi0.5Mn1.5−xTixO4 powder library using the ESD method.
Spinel-type LiNi0.5Mn1.5−xTixO4 powder libraries were prepared by the ESD method as mentioned above. Starting materials used were LiNO3, Ni(NO3)2·6H2O, Mn(NO3)2·6H2O, and TiO2 nano-slurry dispersed in ethanol. These were dissolved in a mixture of ethanol and butyl carbitol (C2H5OH : C4H9(OCH2CH2)2OH = 1:4), and adjusted to a concentration of 0.2 mol/L. A mixture of the starting material solution was electrostatically sprayed onto a grounded and heated (450 °C) YSZ substrate. Also, lithium component added to the mixture was increased to 10 % in order to compensate for the potential volatilization of lithium during the heat treatment. The deposited powder was heat-treated at 700 °C for 3 hours in air. The crystal structures of the sintered powder libraries were identified by powder X-ray diffraction (XRD; Rigaku Miniflex) with FeKα radiation. The particle sizes of the powders obtained were observed with a field emission scanning electron microscope (FE-SEM; JEOL JSM-7600). The electrochemical properties were characterized using a three-electrode beaker cell composed of a working electrode (W.E.), counter electrode (C.E.), and reference electrode (R.E.). A 1 × 1 cm2 W.E. was prepared from slurry by mixing LiNi0.5Mn1.5−xTixO4 powder, acetylene black (AB), and polyvinylidene difluoride (PVdF) at ratio of 80:10:10 (wt. %) with NMP and then coating the mixture onto Al foil. The C.E. and R.E. were prepared using Li foil. The W.E. mixture coated on Al foil was dried at 80 °C for 12 hours under atmospheric pressure and further dried at 120 °C for 4 hours under decompression using a rotary pump. The electrolyte used was 1 mol/L LiClO4 in EC/DEC. The cell was assembled in a grove box filled with Ar gas. A charge-discharge evaluation was performed at a current rate of 1 C (x = 0, 0.1, 0.3, 0.5) and 0.1 C (y = 0, 0.1) with cut-off potentials of 3.7–4.9 V versus Li / Li+.
Fig. 1 shows powder X-ray diffraction patterns of LiNi0.5Mn1.5−xTixO4 (0 ≤ x ≤ 1.5). In the range of 0 ≤ x ≤ 0.5, we observed not only spinel-type LiNi0.5Mn1.5−xTixO4, but also a slight amount of rock-salt type LiyNi1−yO as a secondary phase. In the range of 0.6 ≤ x ≤ 1.5, we observed multiple phases consisting of a spinel-type LiNi0.5Mn1.5−xTixO4, a rock-salt type LiyNi1−yO structure, and an unknown phase.
Powder X-ray diffraction patterns of LiNi0.5Mn1.5−xTixO4 (0 ≤ x ≤ 1.5).
Fig. 2 shows the unit cell parameters of the prepared powders calculated by the Appleman-Evans program16). The unit cell parameters for LiNi0.5Mn1.5−xTixO4 were linearly alternated according to the amount of Ti-substitution. This tendency was presumed to be attributable to the larger ionic radius of Ti4+ (0.0605 nm) versus that of Mn4+ (0.053 nm)17).
Lattice parameter of LiNi0.5Mn1.5−xTixO4 (0 ≤ x ≤ 1.5).
Fig. 3 shows FE-SEM images of the prepared LiNi0.5Mn1.5−xTixO4 (x = 0, 0.1, 0.3, 0.5). The particle sizes of all samples were in a range of 70~100 nm. These results shows about one-tenth of the particle size of the samples prepared by the solid-state reaction method15).
SEM images of LiNi0.5Mn1.5−xTixO4 (a) x = 0, (b) x = 0.1, (c) x = 0.3, (d) x = 0.5.
Fig. 4 shows charge-discharge curves of the 1st and 50th cycle of LiNi0.5Mn1.5−xTixO4 (x = 0, 0.1, 0.3, 0.5) under a current rate of 1 C. The slow charge curve between 4.6 V and 4.8 V was presumed to be associated with Ni2+/3+ and Ni3+/4+ 8,18). The slight change in slope observed around 4 V was attributable to the Mn3+/4+ redox, which suggested that the samples were deficient in oxygen.
Charge-discharge curves of LiNi0.5Mn1.5−xTixO4 (x = 0, 0.1, 0.3, 0.5).
The charge and discharge capacities decreased as the substitution amounts of Ti increased because Ti4+ in LiNi0.5Mn1.5−xTixO4 did not contribute to the redox reaction at the above potential region. Table 1 shows the discharge capacity of the 1st and 50th cycle and the retention rate after 50 cycles. The improvement in the retention rate after 50 cycles appeared to be attributable to the adjustment of the amount of substitution with the Ti element from x = 0 to 0.1.
| x | Theoritical capacity (mAh/g) | Discharge capacity (mAh/g) | Retention rate (%) | |
|---|---|---|---|---|
| 1st | 50th | |||
| 0 | 147 | 98 | 46 | 47 |
| 0.1 | 147 | 95 | 50 | 53 |
| 0.3 | 148 | 87 | 12 | 14 |
| 0.5 | 150 | 77 | 23 | 30 |
Fig. 5 and Fig. 6 plot comparisons of the first discharge capacities at current rates of 0.1 C and 1 C, respectively. The first charge-discharge capacities of LiNi0.5Mn1.5O4 (x = 0) and LiNi0.5Mn1.4Ti0.1O4 (x = 0.1) were the same at a current rate of not only 0.1 C, but also 1 C. At the current rate of 0.1 C, two plateaus were observed between 4.6 V and 4.8 V. As mentioned earlier, an oxidation reaction of Ni2+/3+ and Ni3+/4+ was presumed to have taken place.
Charge-discharge curves of LiNi0.5Mn1.5O4.
Charge-discharge curves of LiNi0.5Mn1.4Ti0.1O4.
All of the powders obtained by ESD had lower capacities than the powder obtained by solid state reaction15). LiyNi1−yO is thought to adversely affect the electrochemical properties19). Because of the oxygen deficiency and byproduct LiyNi1−yO resulting under the air atmospheric condition, we had to reconsider the heat-treatment condition in order to obtain a spinel single phase. We changed the heat-treatment condition from air to oxygen gas and from powder to pellet-shaped by uniaxial pressing. Fig. 7 shows the powder X-ray diffraction patterns of LiNi0.5Mn1.5−xTixO4 (x = 0, 0.1) heated under the new condition. The rock-salt phase (LiyNi1−yO) was decreased in the powder library of LiNi0.5Mn1.5−xTixO4 and a single phase of spinel-type compound was duly obtained.
Powder X-ray diffraction patterns of LiNi0.5Mn1.5−xTixO4 (x = 0, 0.1).
A spinel-type LiNi0.5Mn1.5−xTixO4 powder library was prepared using the electrostatic spray deposition (ESD) method. The deposited powder was heat-treated at 700 °C for 3 hours in air. The resulting powder library consisted of multiple phases of spinel and a rock-salt structure. The particle sizes of LiNi0.5Mn1.5−xTixO4 (0 ≤ x ≤ 0.5) observed by SEM were around 70–100 nm. The charge-discharge capacities of LiNi0.5Mn1.5−xTixO4 (0 ≤ x ≤ 0.5) powder composed of a slight amount of rock-salt type LiyNi1−yO was measured at two current rates, 0.1 C and 1 C. The first discharge capacities of LiNi0.5Mn1.5O4 (x = 0) and LiNi0.5Mn1.4Ti0.1O4 (x = 0.1) were both about 100 mAh/g. The capacity ratio after 50 cycles was improved by adjusting the amount of substitution with Ti element from y = 0 to 0.1. The charge-discharge capacities obtained were lower than those reported earlier5), however, because the powder product included a second phase (LiyNi1−yO). We then attempted to obtain a single-phase spinel-type structure by adjusting the heat-treatment condition from air to oxygen gas and from powder to pellet-shaped by uniaxial pressing. The rock-salt phase (LiyNi1−yO) was decreased in the resulting powder library of LiNi0.5Mn1.5−xTixO4 and a spinel single phase was almost obtained.
