Effects of Al substitution for Ni and Mn on the electrochemical properties of LiNi0.5Mn1.5O4
G.B. Zhong,Y.Y. Wang,Z.C. Zhang,C.H. Chen
DOI: https://doi.org/10.1016/j.electacta.2011.03.093
IF: 6.6
2011-01-01
Electrochimica Acta
Abstract:Highlights ► Aluminum doped LiNi 0.5 Mn 1.5 O 4 powders are synthesized by a thermopolymerization process. ► The comparison of Al doping between Mn sites and Ni sites is made for the first time. ► Properly Al-doped LiNi 0.5 Mn 1.5 O 4 shows excellent cycling stability at 55 °C with the fading rate as low as 0.015% per cycle. ► At room temperature, the capacity fading rate may be even slower (less than 0.01%) and the rate capability is very good (114 mAh g −1 at 10 C). Abstract The effects of Al substitution for Ni or (and) Mn in LiNi 0.5 Mn 1.5 O 4 spinel on the structures and electrochemical properties are investigated. Powders of LiNi 0.5 Mn 1.5 O 4 , Li 0.95 Ni 0.45 Mn 1.5 Al 0.05 O 4 , LiNi 0.475 Mn 1.475 Al 0.05 O 4 and Li 1.05 Ni 0.5 Mn 1.45 Al 0.05 O 4 are synthesized by a thermopolymerization method. Their structures and electrochemical properties are studied by X-ray powder diffraction, scanning electron microscopy, infrared spectroscopy, cyclic voltammetry and galvanostatic charge–discharge testing. The introduction of Al in these LiNi 0.5 Mn 1.5 O 4 samples has resulted in structure variation, and greatly improved their cyclic performance and rate capability. The effects of Al substitutions for Ni and Mn in the LiNi 0.5 Mn 1.5 O 4 are different. Compared with LiNi 0.5 Mn 1.5 O 4 , Li 0.95 Ni 0.45 Mn 1.5 Al 0.05 O 4 demonstrates higher specific capacity at room temperature but faster capacity fading at elevated temperatures. Li 1.05 Ni 0.5 Mn 1.45 Al 0.05 O 4 displays a lower discharge capacity but better capacity retention at 55 °C. Moreover, the cyclic performance and rate capability of the Ni-substituted Li 0.95 Ni 0.45 Mn 1.5 Al 0.05 O 4 , Ni/Mn co-substituted LiNi 0.475 Mn 1.475 Al 0.05 O 4 and Mn-substituted Li 1.05 Ni 0.5 Mn 1.45 Al 0.05 O 4 at room temperature are similar, and have improved substantially compared with the Al-free LiNi 0.5 Mn 1.5 O 4 sample. Keywords Lithium-ion batteries Spinel Al doping Lithium nickel manganese oxide Thermopolymerization 1 Introduction With the ongoing global warming and depletion of crude oil, the environment and energy crisis exert great pressure on the existing energy infrastructures. Consequently, we have been largely increasing the utilization of renewable energies such as solar and wind, and, very vigorously developing electric vehicles (EV) and hybrid electric vehicles (HEV) to reduce CO 2 emissions. Also, large-scale energy storage systems are essential to store energy and provide stable energy output. At the forefront of these efforts is to seek an appropriate electricity storage device. Owing to their high energy density and excellent cycling performance, lithium-ion batteries become the most attractive candidates. They have proved to be the best power sources for portable devices [1–3] . Nevertheless, we need to improve further their performance substantially to meet the more demanding requirements of these large energy storage systems, especially on energy density and power density. If considering only the contribution from the cathode material, we prefer to use a material with a high capacity (in Ah kg −1 ) and a high working potential (in V). The product of both can be defined as the energy density of that material (in Wh kg −1 ). Fig. 1 compares the energy density of several cathode materials for lithium-ion batteries. In this figure, the integration area of the discharge curve represents the energy density of corresponding material. Thanks to its 4.7 V high operating voltage, which is higher than 3.4 V for LiFePO 4 , 3.9 V for LiCoO 2 and 4.1 V for LiMn 2 O 4 , the spinel-structured LiNi 0.5 Mn 1.5 O 4 delivers an energy density of 650 Wh kg −1 , about 20–30% higher than the others. In addition, the three-dimensional lithium diffusion pathways in the spinel lattices are beneficial to provide a high power density. Therefore, LiNi 0.5 Mn 1.5 O 4 has attracted great attentions during the past decade [4–11] . LiNi 0.5 Mn 1.5 O 4 displays a flat voltage profile at 4.7 V corresponding to the redox reactions of Ni 2+ /Ni 3+ and Ni 3+ /Ni 4+ redox couples, with the theoretical capacity of 146.7 mAh g −1 [12,13] . It is actually regarded as a special case of doped lithium magnetite spinels LiM x Mn 2− x O 4 (M = Cr, Fe, Co, Ni, Mg, Cu, etc.) [14–19] . This composition with the Ni:Mn molar ratio of 1:3 has attracted more attention than the others because it shows a dominant potential plateau around 4.7 V, while other compounds (LiM x Mn 2− x O 4 , M: Cr, Fe, Co, Cu) exhibit two plateaus at 4.0 and 5.0 V [19] . Usually, the capacity at 5.0 V or above is not utilizable because no suitable electrolyte system is available at present. Thus, LiNi 0.5 Mn 1.5 O 4 has become one of the most important cathode materials for lithium-ion batteries. Unfortunately, even this compound (LiNi 0.5 Mn 1.5 O 4 ) still has a non-negligible capacity fading during cycling, especially at elevated temperatures, due to the structural and chemical instabilities resulted from the presence of high spin Mn 3+ ions [20] . A number of groups have worked on improving the electrochemical properties of this interesting cathode material, especially on its cycling stability, by means of surface coating [20–23] and doping [4,11,24–27] . These methods, especially the doping, are proved to be quite effective. Aklalouch et al. have synthesized LiCr 2 γ Ni 0.5− γ Mn 1.5− γ O 4 (0 < γ < 0.2) by a sucrose-aided combustion method [24] . The samples with γ ≤ 0.1 have capacity retention of 96% after 40 cycles at 55 °C. Ito et al. have increased the lithium diffusion coefficient by Co doping [25] . The LiNi 0.5− x Co 2 x Mn 1.5− x O 4 (0 ≤ 2 x ≤ 0.2) exhibits improved cyclic performance at a high rate and at elevated temperatures. Park et al. have investigated LiNi 0.5− x Mn 1.5 Cr x O4 ( x = 0.00, 0.01, 0.03 and 0.05) and found that the initial capacity and capacity retention are improved as the Cr content increases [4] . Oh et al. have investigated a series of doped system LiNi 0.5− x Mn 1.5− y M x + y O 4 (M = Cr, Al and Zr) and found that the electrical conductivity increase with Cr- and Al-doping [26] . We can see that the foreign ions can substitute either Mn ions or Ni ions, or both of them. But to the best of our knowledge, the difference between the substitutions of Ni or (and) Mn has not been studied yet. To perform this study, we design three different spinels of Al substitutions, viz. Li 0.95 Ni 0.45 Mn 1.5 Al 0.05 O 4 , LiNi 0.475 Mn 1.475 Al 0.05 O 4 and Li 1.05 Ni 0.5 Mn 1.45 Al 0.05 O 4 . The content of Al is controlled to be 0.05 and the content of Li is altered to keep charge neutrality. LiNi 0.5 Mn 1.5 O 4 is also studied as a contrast. In the end, we have found LiNi 0.475 Mn 1.475 Al 0.05 O 4 gives rise to the best performance. 2 Experimental The LiNi 0.5 Mn 1.5 O 4 , Li 0.95 Ni 0.45 Mn 1.5 Al 0.05 O 4 , LiNi 0.475 Mn 1.475 Al 0.05 O 4 and Li 1.05 Ni 0.5 Mn 1.45 Al 0.05 O 4 powders were synthesized by a thermopolymerization method. Stoichiometric amounts of lithium nitrate (LiNO 3 , 5% excess), aluminum nitrate (Al(NO 3 ) 3 ·9H 2 O), nickel nitrate (Ni(NO 3 ) 2 ·6H 2 O) and manganese acetate (Mn(CH 3 COO) 2 ·4H 2 O) were dissolved in deionized water to obtain a 0.2 M solution. Then acrylic acid (AA) was added to form an AA-H 2 O (1:2, v/v) solution. The solution was then kept in an oven at 150 °C for 10 h to complete the thermopolymerization reaction. The products were first calcined at 500 °C for 5 h, and cooled down to room temperature. After being grinded, the intermediate products were sintered at 900 °C for 15 h and subsequently annealed at 700 °C for 48 h. All the heat treatment processes were carried out in air atmosphere. The synthesized spinel powders were analyzed by X-ray diffraction (XRD) using a diffractometer (Philips X’pert Pro Super, Cu Kα radiation) in the 2 θ range from 10° to 80° with a scanning rate of 10° min −1 . The particle size and morphology were also observed by a scanning electron microscope (JSM-6390LA, JEOL). A Nicolet 8700 (Thermo Scientific Instrument Co. USA) was used for Fourier-transformed infrared (FTIR) study. To prepare the electrode laminates, we blended the active materials, acetylene black and poly(vinylidene fluoride) (PVDF) (80:10:10, w/w/w) with N-methyl-2-pyrrolidone in an agate mortar to obtain homogenous slurries, which were then cast on aluminum foils with a doctor blade and dried at 70 °C for 10 h to obtain the electrode laminates. Discs ( ϕ = 14 mm) of the laminates were punched, dried at 70 °C for 2 h in a vacuum oven and then were transferred into an argon-filled glove box (MBraun Labmaster 130). Afterwards, CR2032 half cells with Li as the counter electrode were assembled in the glove box with the electrolyte of 1 M LiPF 6 solution in ethylene carbonate (EC) – dimethyl carbonate (DMC) (1:1 w/w, Guotai Huarong New Chemical Materials Co.). The cyclic voltammograms (CV) of the cells were measured on a CHI 604A electrochemical workstation from 3.5 to 5.1 V at a scan rate of 0.1 mV s −1 . The cells were also galvanostatically cycled on a multi-channel battery cycler (Neware BTS2300, Shenzhen) in the voltage window from 3.5 to 5.0 V. The cells were also cycled at 55 °C by putting them in an oven. 3 Results and discussion The X-ray diffraction patterns of the spinel powders are shown in Fig. 2 . The results reveal that the products are well-defined cubic spinels and no trace of impurity phase (such as Li x Ni 1− x O) is detected. The structures are assigned to either Fd 3 ¯ m (F-type) or P4 3 32 (P-type) space group, which depends on the ordering of transition metal cations. It is well known that spinels showing cations ordering on the octahedral sites have P4 3 32 type symmetry [28] . In this structure the Ni 2+ and Mn 4+ transition metals occupy the 4a and 12b lattice positions, respectively, rather than distribute on the 16d sites in the case of Fd 3 ¯ m. In fact, with a very slow scanning rate, LiNi 0.5 Mn 1.5 O 4 with the structure of P4 3 32 should have also shown some small diffraction peaks at 15.3°, 39.7°, 45.7°, 57.5° and 65.6° [28] . Nevertheless, the scanning rate in this work was too fast for us to detect them. On the other hand, FTIR spectroscopy has proved to be quite effective to distinguish Fd 3 ¯ m from P4 3 32. According to Amatucci et al. [29] , P-type and F-type can be distinguished by two methods: (1) the P-type has eight IR absorption bands and F-type has only five due to peak broadenings and (2) the intensity ratios among the absorption signals are different in the P-type and F-type samples. For example, with increasing degree of lattice ordering, the intensity of Ni–O band at 588 cm −1 increases while that of Mn–O band at 619 cm −1 decreases. Thus, for this study, the infrared spectra of the samples are shown in Fig. 3 . It can be seen that the space groups of LiNi 0.5 Mn 1.5 O 4 and Al-doped spinels are different, although they were heated and annealed at the same condition. The space group of LiNi 0.5 Mn 1.5 O 4 is P4 3 32 while the Al-doped spinels demonstrate the more disordered space group Fd 3 ¯ m. Fig. 4 shows the scanning electron microscopy (SEM) images of the LiNi 0.5 Mn 1.5 O 4 and the Al-doped spinels powders. It can be found that the particle morphologies of these spinels are almost the same. The particles exhibit surface facets, which suggest that the samples are all well crystallized. And the particle sizes of these powders are similar, about 1–3 μm. Fig. 5 shows the cyclic voltammograms of half-cells with the LiNi 0.5 Mn 1.5 O 4 and the Al-doped spinels samples as the working electrodes. For the Al-doped samples (curves b, c, and d), there are two strong split peaks at around 4.7 V which correspond to Ni 2+ /Ni 3+ and Ni 3+ /Ni 4+ redox couples, and a weak peak at around 4.0 V which corresponds to the redox couples of Mn 3+ /Mn 4+ . For the pure LiNi 0.5 Mn 1.5 O 4 sample (curve a), it seems to show only one peak at around 4.7 V, suggesting a very slight difference of the voltage steps corresponding to Ni 2+ /Ni 3+ and Ni 3+ /Ni 4+ redox couples in this case. The inset graph in Fig. 5 is the magnified 4.0 V region framed in the rectangle. We can see that the peak of LiNi 0.5 Mn 1.5 O 4 at 4.0 V region is negligible. The intensity of the peak increases after the Al-doping. The peaks at around 4.0 V are particularly pronounced for the sample Li 0.95 Ni 0.45 Mn 1.5 Al 0.05 O 4 , in which Al 3+ ions are used to substitute Ni 2+ ions. Because the valence of foreign Al ions is higher than that of Ni ions, the valence of Mn must partly decrease from 4+ to 3+ in order to keep the charge neutrality of the whole compound. Electrochemical charge–discharge studies were performed in a galvanostatic mode between 3.5 and 5.0 V. The first-cycle voltage profiles of the Al-free sample LiNi 0.5 Mn 1.5 O 4 and three Al-doped spinels are shown in Fig. 6 . Being consistent with above CV results, the charge–discharge curves also show two main voltage plateaus in the 4.7 V (main) and 4.0 V (minor) regions. Obviously, the Al-free sample LiNi 0.5 Mn 1.5 O 4 has a negligible 4.0 V plateau while the Ni-substituted sample Li 0.95 Ni 0.45 Mn 1.5 Al 0.05 O 4 presents the longest 4.0 V plateau. The capacities of the 4.0 V plateaus for the Mn-substituted sample Li 1.05 Ni 0.5 Mn 1.45 Al 0.05 O 4 and the Ni/Mn co-substituted sample LiNi 0.475 Mn 1.475 Al 0.05 O 4 are similar. In addition, compared to LiNi 0.5 Mn 1.5 O 4 , the Mn-substituted sample Li 1.05 Ni 0.5 Mn 1.45 Al 0.05 O 4 delivers a lower discharge capacity of about 121 mAh g −1 while the Ni substituted and the Ni/Mn co-substituted spinels have a comparable capacity of about 130 mAh g −1 . The cyclic performances of these spinel samples at 1 C rate and room temperature are shown in Fig. 7 . It is clearly observed that Al-doping can clearly improve the cyclic performance of LiNi 0.5 Mn 1.5 O 4 spinel. The Al-free sample LiNi 0.5 Mn 1.5 O 4 gives a poor cyclic performance, with the discharge capacity notably decreasing from 130 to 116 mAh g −1 after 100 cycles. On the other hand, the reversibility of each Al-doped sample is very good. Both the Li 0.95 Ni 0.45 Mn 1.5 Al 0.05 O 4 and the LiNi 0.475 Mn 1.475 Al 0.05 O 4 samples exhibit the best and nearly the same cycling stability. The capacity only diminishes slightly from 132 to 130 mAh g −1 after 100 cycles, or the capacity fading rate is around 0.015% per cycle. For the Li 1.05 Ni 0.5 Mn 1.45 Al 0.05 O 4 sample, although its initial capacity (121 mAh g −1 ) is a little lower than the former two, a capacity of 120 mAh g −1 is still retained after 100 cycles or the capacity fading rate is less than 0.01% per cycle. The results indicate that cyclic performance of LiNi 0.5 Mn 1.5 O 4 at room temperature can be markedly improved by Al-doping, no matter on Ni or Mn sites. A thermodynamic consideration can explain the reason for the improvement by Al-doping. The Gibbs energy of formation at 298 K for NiO, MnO 2 and Al 2 O 3 is −211.7, −465.1 and −1582.3 kJ mol −1 , respectively [30] . Obviously, the Al–O bonding strength is much higher than those of Mn–O and Ni–O bondings. Thus, the excellent cycling property of Al-doped samples may be attributed to the enhanced stability of the crystal structure reinforced by introducing some Al–O bonds. In fact, aluminum doping is also found to greatly improve the cycling stability of LiMn 2 O 4 [31] . As mentioned before, the LiNi 0.5 Mn 1.5 O 4 spinel may experience severe capacity loss when cycled at an elevated temperature (50–60 °C) [24] . Here, the results of 55 °C-cycling at 1 C rate are shown in Fig. 8 . It can be seen that the Al-free sample LiNi 0.5 Mn 1.5 O 4 indeed suffers a rapid capacity fading from initially 130 mAh g −1 to 95 mAh g −1 , or only 73.1% retention, at the 100th cycle. The cyclic performance of the Ni-substituted sample Li 0.95 Ni 0.45 Mn 1.5 Al 0.05 O 4 also becomes worse at 55 °C. Its capacity retention after 100 cycles decreases to 87.9%, which is 98.5% at room temperature. In clear contrast, the Mn-substituted sample Li 1.05 Ni 0.5 Mn 1.45 Al 0.05 O 4 and the Ni/Mn co-substituted sample LiNi 0.475 Mn 1.475 Al 0.05 O 4 still present excellent cyclic performance with a capacity of 121 and 128 mAh g −1 (98.4% and 98.5% of their initial capacity, respectively) after 100 cycles. To study the evolution during the cycling tests at 55 °C, the charge/discharge curves of the first, 50th and 100th cycles of the samples are shown in Fig. 9 . It is obvious that LiNi 0.475 Mn 1.475 Al 0.05 O 4 ( Fig. 9 c) and Li 1.05 Ni 0.5 Mn 1.45 Al 0.05 O 4 ( Fig. 9 d) present high stability during the cycling. But for the Al-free sample LiNi 0.5 Mn 1.5 O 4 , the capacity decreases rapidly and the polarization aggravates with cycle number. Upon charging, the voltage increases to 5.0 V rapidly and most of the capacity measured is obtained during the constant voltage process for the 100th cycle. The variation is indicated to be caused by the difference in the space group. Obviously, LiNi 0.5 Mn 1.5 O 4 ( Fd 3 ¯ m ) is more stable than LiNi 0.5 Mn 1.5 O 4 (P4 3 32) so that the former shows superior cycling stability at the elevated temperatures compared to the latter, as also observed by Sun et al. [28] . To observe any structural change during the cycling tests at 55 °C, the cathode electrodes after 100 cycles were characterized by SEM ( Fig. 10 ). Before cycling, the typical spinels particle morphology is octahedral with smooth surfaces ( Fig. 4 ). After 100 cycles at 55 °C, the morphology of the Al-free sample LiNi 0.5 Mn 1.5 O 4 remains intact with smooth surfaces ( Fig. 10 a), but the Ni-substituted sample Li 0.95 Ni 0.45 Mn 1.5 Al 0.05 O 4 displays completely different morphology ( Fig. 10 b). The particles have an irregular shape with some unidentified species on the surface. The Ni/Mn co-substituted sample LiNi 0.475 Mn 1.475 Al 0.05 O 4 and the Mn-substituted sample Li 1.05 Ni 0.5 Mn 1.45 Al 0.05 O 4 also show a small degree of change on the surface ( Fig. 10 c and d). When discussing the Mn dissolution in the Al-doped samples, two opposite factors should be taken into account. (1) The presence of Mn 3+ would tend to cause Mn dissolution and hence capacity fade. (2) Al-doping stabilizes the crystal structure and tends to suppress the Mn dissolution. Therefore, if the amount of Mn 3+ is not too much, the latter factor may dominate, resulting in a stable cycling. This is the case for LiNi 0.475 Mn 1.475 Al 0.05 O 4 and Li 1.05 Ni 0.5 Mn 1.45 Al 0.05 O 4 . On the other hand, if the amount of Mn 3+ is too much, the former factor may dominate and the capacity fades fast, which is the case for Li 0.95 Ni 0.45 Mn 1.5 Al 0.05 O 4 . To compare the rate capabilities of LiNi 0.5 Mn 1.5 O 4 and Al-doped spinels, all coin cells were galvanostatically charged at room temperature under a current density of 0.5 C rate and discharged at different C-rates from 0.5 C to 10 C and then back to 0.5 C ( Fig. 11 ). With increasing the current, the Al-free sample LiNi 0.5 Mn 1.5 O 4 exhibits a gradual decrease in capacity with lower discharge voltages. At the rate of 10 C, the obtained capacity is only 53 mAh g −1 and the operating discharge voltage is lower than 3.8 V. In contrast, the Al-doped spinels demonstrate a higher discharge voltage and much higher capacity retention. For example, the average discharge voltage is around 4.2 V at 10 C rate with a capacity of 114 mAh g −1 ( Fig. 11 c). Fig. 12 demonstrates the discharge capacity of the samples at different C-rates. The Li 1.05 Ni 0.5 Mn 1.45 Al 0.05 O 4 and LiNi 0.475 Mn 1.475 Al 0.05 O 4 spinels present the best rate performance. They can deliver a capacity of 111 and 114 mAh g −1 at 10 C rate, respectively, about 88% of their capacity at 0.5 C rate. The capacity retention of the sample Li 0.95 Ni 0.45 Mn 1.5 Al 0.05 O 4 is 79% at 10 C rate. It is lower than that of Mn-doped spinels but still much higher than 38% for the Al-free sample LiNi 0.5 Mn 1.5 O 4 . The calculated energy density (power density) of the LiNi 0.475 Mn 1.475 Al 0.05 O 4 sample is 474 Wh kg −1 (5.38 kW kg −1 at 10 C rate), based only on the mass of the cathode material, while it is only 197 Wh kg −1 (4.82 kW kg −1 at 10 C rate) for the Al-free sample LiNi 0.5 Mn 1.5 O 4 . Moreover, when the current returns to 0.5 C rate after cycles at various C-rates, the discharge capacity of all the samples can return to initial levels. The improved rate performance should be attributed to the increased electronic conductivity by Al-doping. To prove this speculation, the DC resistance of the cells was measured by employing an interruption for 1 min after every charge step for 5 min [32] . The values of the calculated resistance of the cells versus the charge capacity are shown in Fig. 13 . Obviously, all of the Al-doped spinels demonstrate similar resistances, which are substantially lower than that of pristine LiNi 0.5 Mn 1.5 O 4. This result is consistent with the work of Sun et al. [28] . LiNi 0.5 Mn 1.5 O 4 with the P4 3 32 structure has a higher resistance than that with the Fd 3 ¯ m structure during delithiation. Finally, we compare the data of 55 °C cycling and rate performance of our samples with those reported in the literature ( Table 1 ). The capacity fading rate of our samples is about 0.015% per cycle for LiNi 0.475 Mn 1.475 Al 0.05 O 4 and Li 1.05 Ni 0.5 Mn 1.45 Al 0.05 O 4 . This value is lower than the literature data (usually above 0.04%). For the rate performance, for our sample LiNi 0.475 Mn 1.475 Al 0.05 O 4 , the discharge capacity at 10 C rate is as high as 114 mAh g −1 , which is much higher than that of Co-doped LiNi 0.5 Co 0.05 Mn 1.45 O 4 [34] and LiMn 1.42 Ni 0.42 Co 0.16 O 4 [35] , Mg-doped LiMg 0.07 Ni 0.43 Mn 1.5 O 4 [36] and Cr-doped LiNi 0.45 Cr 0.1 Mn 1.45 O 4 [11] . Therefore, the Al doping is a useful way to improve the high temperature cycling performance and rate performance of the LiNi 0.5 Mn 1.5 O 4 -based cathode materials. 4 Conclusions LiNi 0.5 Mn 1.5 O 4 and Al-doped Li 0.95 Ni 0.45 Mn 1.5 Al 0.05 O 4 , LiNi 0.475 Mn 1.475 Al 0.05 O 4 and Li 1.05 Ni 0.5 Mn 1.45 Al 0.05 O 4 spinels have been synthesized by a thermopolymerization method. The Al-doping changes the space group of LiNi 0.5 Mn 1.5 O 4 from ordered P4 –3 32 to disordered Fd 3 ¯ m under the same heat treatment conditions. It can significantly improve the cycling stability of LiNi 0.5 Mn 1.5 O 4 . The capacity retentions of Al-doped spinels are over 99% after 100 cycles at room temperature. Even at 55 °C, the capacity retention of Li 1.05 Ni 0.5 Mn 1.45 Al 0.05 O 4 can reach 98% after 100 cycles. In addition, the rate performance of LiNi 0.5 Mn 1.5 O 4 is also greatly improved by the Al-doping. The effects of Al substitutions for Ni and (or) Mn ions in the LiNi 0.5 Mn 1.5 O 4 are somewhat different in that the Li 0.95 Ni 0.45 Mn 1.5 Al 0.05 O 4 shows faster capacity fading at elevated temperature than the Li 1.05 Ni 0.5 Mn 1.45 Al 0.05 O 4 sample. The latter one gives higher capacity retention at 55 °C and displays better rate capability. As a compromise, the Ni/Mn co-substituted sample LiNi 0.475 Mn 1.475 Al 0.05 O 4 shows the best electrochemical performance with a high specific capacity during cycling at room and elevated temperatures, and excellent rate capability. Acknowledgments This study was supported by National Science Foundation of China (grant nos. 20971117 and 10979049 ) and Education Department of Anhui Province (grant no. KJ2009A142 ). We are also grateful to the Solar Energy Operation Plan of Academia Sinica. References [1] J.M. Tarascon E. Wang F.K. Shokoohi W.R. Mckinnon S. Colson J. Electrochem. Soc. 138 1991 2859 [2] F.K. Shokoohi J.M. Tarascon B.J. Wilkens Appl. Phys. Lett. 59 1991 1260 [3] T. Ohzuku K. Ariyoshi S. Yamamoto Y. Makimura Chem. Lett. 2001 1270 [4] S.B. Park W.S. Eom W.I. Cho H. Jang J. Power Sources 159 2006 679 [5] R. Alcantara M. Jaraba P. Lavela J.L. Tirado Electrochim. Acta 47 2002 1829 [6] Y.K. Sun C.S. Yoon I.H. Oh Electrochim. Acta 48 2003 503 [7] K. Ariyoshi Y. Iwakoshi N. Nakayama T. Ohzuku J. Electrochem. Soc. 151 2004 A296 [8] T.A. Arunkumar A. Manthiram Electrochim. Acta 50 2005 5568 [9] J.C. Arrebola A. Caballero L. Hernan J. Morales J. Power Sources 180 2008 852 [10] H.F. Xiang X. Zhang Q.Y. Jin C.P. Zhang C.H. Chen X.W. Ge J. Power Sources 183 2008 355 [11] D.Q. Liu Y.H. Lu J.B. Goodenough J. Electrochem. Soc. 157 2010 A1269 [12] Y.S. Lee Y.K. Sun S. Ota T. Miyashita M. Yoshi Electrochem. Commun. 4 2002 989 [13] M. Mohamedi A. Makino K. Dokko T. Itoh I. Uchida Electrochim. Acta 48 2002 79 [14] K. Amine H. Tukamoto H. Yasuda Y. Fujita J. Electrochem. Soc. 143 1996 1607 [15] Q.M. Zhong A. Bonakdarpour M.J. Zhang Y. Gao J.R. Dahn J. Electrochem. Soc. 144 1997 205 [16] C. Sigala A.L. La Salle Y. Piffard D. Guyomard J. Electrochem. Soc. 148 2001 A812 [17] H. Shigemura H. Sakaebe H. Kageyama H. Kobayashi A.R. West R. Kanno S. Morimoto S. Nasu M. Tabuchi J. Electrochem. Soc. 148 2001 A730 [18] J.S. Kim J.T. Vaughey C.S. Johnson M.M. Thackeray J. Electrochem. Soc. 150 2003 A1498 [19] T. Ohzuku S. Takeda M. Iwanaga J. Power Sources 82 1999 90 [20] Y.K. Sun K.J. Hong J. Prakash K. Amine Electrochem. Commun. 4 2002 344 [21] J.S. Kim C.S. Johnson J.T. Vaughey S.A. Hackney K.A. Walz W.A. Zeltner M.A. Anderson M.M. Thackeray J. Electrochem. Soc. 151 2004 A1755 [22] J. Liu A. Manthiram J. Electrochem. Soc. 156 2009 A66 [23] J. Liu A. Manthiram J. Electrochem. Soc. 156 2009 A833 [24] M. Aklalouch J.M. Amarilla R.M. Rojas I. Saadoune J.M. Rojo J. Power Sources 185 2008 501 [25] A. Ito D. Li Y. Lee K. Kobayakawa Y. Sato J. Power Sources 185 2008 1429 [26] S.H. Oh K.Y. Chung S.H. Jeon C.S. Kim W.I. Cho B.W. Cho J. Alloys Compd. 469 2009 244 [27] J. Liu A. Manthiram J. Phys. Chem. C 113 2009 15073 [28] J.H. Kim S.T. Myung C.S. Yoon S.G. Kang Y.K. Sun Chem. Mater. 16 2004 906 [29] M. Kunduraci J.F. Al-Sharab G.G. Amatucci Chem. Mater. 18 2006 3585 [30] R.L. David Handbook of Chemistry and Physics 2005 CRC Press Boca Raton, FL http://www.hbcpnetbase.com [31] S.T. Myung S. Komaba N. Kumagai J. Electrochem. Soc. 148 2001 A482 [32] H.Y. Xu S. Xie N. Ding B.L. Liu Y. Shang C.H. Chen Electrochim. Acta 51 2006 4352 [33] Y.K. Sun Y.S. Lee M. Yoshio K. Amine Electrochem. Solid State 5 2002 A99 [34] S.W. Oh S.T. Myung H.B. Kang Y.K. Sun J. Power Sources 189 2009 752 [35] J. Liu A. Manthiram Chem. Mater. 21 2009 1695 [36] C. Locati U. Lafont L. Simonin F. Ooms E.M. Kelder J. Power Sources 174 2007 847