An Environmentally Friendly and Cost-Effective Route to LiNi_0.5Mn_1.5O₄ Crystals: Structure, Morphology Evolution, and Electrochemical Properties

Chemicals of NiSO₄⋅6H₂O, MnSO₄⋅H₂O and NaOH in a mole ratio of 0.5:1.5:2.05 were mixed by ball-milling at a rotation speed of ∼280 rpm for 30 min, and nickel manganese hydroxides and sodium sulfate were produced by the room-temperature solid-state metathesis reactions between the Ni-Mn sulfates and NaOH during the mixing process. The obtained mixture was fully washed with distilled water to remove the water-soluble Na₂SO₄ and the excess NaOH and then dried at 80 °C in a blast air oven for 3 h. The dried powders were composed of Ni-Mn (oxy)hydroxides and taken as the precursor. The precursor was mixed with Li₂CO₃ in various mole ratios (1:1, 1:1.03, 1:1.05) by ball-milling for 30 min. The resulting mixture was sintered at different temperatures (750 °C, 800 °C, 850 °C and 900 °C) for 10 h and then naturally cooled to room temperature and ground. Finally, the sintered materials were annealed at 600 °C for 20 h and then naturally cooled to room temperature. The heating rate was 10 °C min⁻¹ during the sintering and annealing processes. The entire sintering and annealing process were carried out in tubular furnace with openings at both ends in air. Six samples synthesized under different conditions were named 3%–750 °C, 3%–800 °C, 3%–850 °C, 3%–900 °C, 0%–800 °C and 5%–800 °C, respectively. The schematic diagram of the whole preparation process for LiNi_0.5Mn_1.5O₄ is shown in Figure 1, and it is quite clear that the route is free of producing any toxic substance.

Zoom In Zoom Out Reset image size

The crystalline structure of the prepared samples was measured by X-ray diffraction (XRD, Ultima IV, Rigaku, Japan) with Cu K_α radiation, and the samples were mixed with standard silicon homogeneously before XRD detection for peak positions calibration. The XRD patterns were refined using GSAS Rietveld software. Particle size and morphology were observed by a scanning electron microscopy (SEM, GeminiSEM 300, ZEISS, Germany). Raman spectroscopy was measured by a Raman spectrometer (Raman, LabRAM HR, Horiba, Japan). X-ray photoelectron spectroscopy (XPS) was conducted by a scanning XPS microprobe system (Versaprobe-II, PHI, America) using monochromatic Al anode as X-ray source, and the data were calibrated by C1s peak at 284.8 eV. The XPS spectra was fitted using MultiPak software.

The electrochemical performance of the prepared samples was evaluated using CR2025 coin-type cells with lithium metal as the anode. 1 M LiPF₆ in EC/DMC/EMC (1:1:1 in volume) solution was used as electrolyte and Teklon UH2140 membrane was used as separator. The cathode was made by mixing of 80 wt.% LiNi_0.5Mn_1.5O₄, 10 wt.% super P and 10 wt.% polyvinylidene fluoride (PVDF) in N-methyl pyrrolidone (NMP) to form homogenous slurry. Then the slurry was coated on the aluminum foil using a doctor blade coater and then dried at 80 °C for 10 h in a vacuum oven. Cells were assembled in an Ar-filled glove box. The cells were charged and discharged galvanostatically using a battery test system (Land CT2001A). For the cycle performance, the cells were cycled at 0.2 C (1 C is equal to 147 mA g⁻¹) in the range of 3.5–5.0 V, and for the rate performance, the cells were charged at 0.2 C to 5.0 V and discharged at various rates to 3.0 V. Electrochemical impedance spectroscopy (EIS) was performed in the frequency range of 1–100000 Hz at a potentiostatic signal amplitude of 5 mV using an electrochemical workstation (CHI604E, Shanghai, China). All electrochemical measurements were carried out at a constant temperature of 30 °C.

Figure 2a shows the XRD patterns of the precursor derived from the room-temperature solid-state reactions between Ni-Mn sulfates and NaOH. All peaks in the XRD patterns can be fully attributed to Ni-Mn hydroxides and oxyhydroxides, and there are no other peaks observed because the product of Na₂SO₄ and the excess NaOH have been washed out by distilled water. The presence of Ni-Mn oxyhydroxides in the precursor is due to the partial oxidation of Ni-Mn hydroxides during the preparation process in the air. The broaden diffraction peaks indicate the small size of the grains in the precursor, which was confirmed by the SEM observation. As shown in Fig. 2b, the size of the primary particles in the precursor is as small as about 20 nm. Considering that nanoparticles have large specific surface area and high reactivity, the nanosize precursor is expected to be conducive to the subsequent reaction with Li₂CO₃ to form LiNi_0.5Mn_1.5O₄.

Zoom In Zoom Out Reset image size

Upon the Ni/Mn distribution, LiNi_0.5Mn_1.5O₄ can exist in two crystal structures: the ordered structure with a space group of P4₃32 and the disordered structure with a space group of Fd3m. Early reports have demonstrated that an annealing treatment could lead to Ni/Mn ordering in the octahedral sites of LiNi_0.5Mn_1.5O₄, ^7,8,32 and the samples synthesized in the present work all underwent the same annealing treatment. Figures 3a and 3b show the XRD patterns of the LiNi_0.5Mn_1.5O₄ samples. As expected, the main diffraction peaks of each pattern all correspond to the ordered structure with a space group of P4₃32. As compared with the disordered structure, the ordered structure has additional diffraction peaks such as the peaks of (221) and (320) as shown in Figs. 3c and 3d. Moreover, the appearance of (220) peak indicates the occupation of transition metal ions in the tetrahedral Li site, ³¹ and the intensity of (220) peak was hardly affected by the sintering temperature, but gradually decreases with the increase of lithium excess from 0 to 5% as shown in Table I, which suggests that excess lithium can suppress the occupation of transition metal ions in the tetrahedral Li sites. All the XRD patterns were refined by GSAS Rietveld refinement program, and Fig. 3e shows the representative refinement of the XRD pattern for the 3%–800 °C sample, and the calculated lattice parameters of all samples are listed in Table I. When the lithium excess is fixed at 3% and the sintering temperature is raised from 750 to 900 °C, the lattice parameter increases from 8.1706 to 8.1745 Å. When the sintering temperature is fixed at 800 °C and the lithium excess is raised from 0 to 5%, the lattice parameter decreases from 8.1732 to 8.1714 Å. The change of the lattice parameter reflects the varied amount of Mn³⁺ in the sample. For the ideal LiNi_0.5Mn_1.5O₄, only Mn⁴⁺ should be present, but Mn³⁺ is often observed in the real sample because of the defects and impurity formation during the preparation process. ⁵ Since the ionic radius of Mn³⁺ (r = 0.645 Å) is larger than that of Mn⁴⁺ (r = 0.53 Å), the lattice parameter increases when Mn³⁺ appears in the sample. Similar variation in the amount of Mn³⁺ was observed by the XPS measurement. Figure 4 shows the XPS spectra of Mn 2p of the LiNi_0.5Mn_1.5O₄ samples. All spectra have a similar Mn 2p profile with two spin–orbit splitting peaks of Mn 2p_1/2 and Mn 2p_3/2, and each peak can be resolved into two peaks corresponding to the oxidation state of Mn³⁺ and Mn⁴⁺. ^26,28 The relative content of Mn³⁺ and Mn⁴⁺ on the surface of all samples can be obtained by fitting, and the content of Mn³⁺ is estimated to be 15.98, 26.36, 32.67, 36.11, 30.36, 18.89 for the 3%–750 °C, 3%–800 °C, 3%–850 °C, 3%–900 °C, 0%–800 °C and 5%–800 °C samples, respectively. Clearly, the variation in the amount of Mn³⁺ revealed by XPS is well consistent with that indicated by XRD.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The ordered structure of the synthesized LiNi_0.5Mn_1.5O₄ samples was further confirmed by the Raman spectroscopy which is a more sensitive tool to distinguish the two space groups of Fd3m and P4₃ 32. ⁸ Figure 5 shows the Raman spectra of the LiNi_0.5Mn_1.5O₄ samples. All spectra have a similar profile, indicating that all samples have a similar structure. The peaks at about 402 and 492 cm⁻¹ are associated with the stretching vibration of Ni–O bond, and the peak at about 635 cm⁻¹ is attributed to the stretching vibration of Mn–O bond in the structure. The sharp peaks at about 160, 218 and 239 cm⁻¹ coupled with a well splitting of the two peaks at around 587 and 607 cm⁻¹ are obviously observed, which are the characteristic features of the Ni/Mn ordering in the crystal structure of LiNi_0.5Mn_1.5O₄. ^33,34

Zoom In Zoom Out Reset image size

Figure 6 shows the SEM images of the LiNi_0.5Mn_1.5O₄ samples. Typical octahedral-like morphology can be observed on these samples, and the sintering temperature shows a more significant effect on the morphology than the lithium excess. As shown in Figs. 6a–6d, the octahedral-like crystals with sharp edges and corners have been formed at 750 °C and a few of incompletely-developed tiny grains are scattered on the surface of some big octahedral-like crystals. When the sintering temperature was raised to 800 °C, the octahedral-like crystals remain intact with a slight growth in size and meanwhile the undeveloped tiny grains diminish. However, melting phenomenon is observed on the edges and corners of the octahedral-like crystals and simultaneously some tiny fragments are observed on the surface of the octahedral-like crystals at 850 °C. When the temperature is further raised up to 900 °C, the edges and corners of most octahedral-like crystals are broken up and the resulting tiny fragments distribute all around the big crystals. As shown in Figs. 6e, 6b and 6f, the crystals morphology is slightly affected by the lithium excess, and the three samples all show sharp edges and corners of the octahedral-like crystals, and the minor difference is the numbers of tiny grains attached on the surface of the octahedral-like crystals. As compared with the samples with 0% and 5% lithium excess, there are less tiny grains observed on the surface of the octahedral-like crystals for the 3% sample. In short, the 3%–800 °C sample possesses well-crystallized octahedral-like LiNi_0.5Mn_1.5O₄ crystals with the best integrity and the cleanest surfaces. Figure 7 shows the crystal size distributions for the six samples. With increasing the sintering temperature from 750 to 900 °C, in general the mean crystal size increases and the size distribution becomes broader at 850 and 900 °C, but the amount of lithium excess shows less effect. The mean sizes are 0.83, 1.16, 1.53, 2.44, 1.17, and 1.16 μm for the 3%–750 °C, 3%–800 °C, 3%–850 °C, 3%–900 °C, 0%–800 °C and 5%–800 °C samples, respectively. From the perspective of the crystal morphology and size, the 3%–800 °C sample appears to be a desirable outcome for the LiNi_0.5Mn_1.5O₄ prepared by the present route. As compared with the previous in-situ molten salt route, ^23,31 the present route can readily yield well-crystallized micron size octahedral-like LiNi_0.5Mn_1.5O₄ crystals at a moderate sintering temperature with a small lithium excess.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Well-crystallized octahedral-like LiNi_0.5Mn_1.5O₄ crystals are often found to have good electrochemical performance. ^13,30 Figure 8 shows the electrochemical performance of the LiNi_0.5Mn_1.5O₄ samples. As shown in the Figs. 8a and 8b, the charge/discharge curves at 0.2 C (1 C = 147 mA g⁻¹) of all samples show two voltage plateaus: a major 4.7 V plateau and a minor 4.1 V plateau which is corresponding to the redox couple of Mn³⁺/Mn⁴⁺. The 4.1 V plateau becomes more obvious as the temperature increases from 750 to 900 °C (Fig. 8a), which means the increased content of Mn³⁺ in the material, and becomes less obvious with the increase of the lithium excess from 0 to 5% (Fig. 8b), which means the decreased content of Mn³⁺. The observed variation of the 4.1 V plateau is well correlated with the variation of the calculated lattice parameters listed in Table I. Figures 8c and 8d show the cycle performance of the LiNi_0.5Mn_1.5O₄ samples at 0.2 C. As shown in Fig. 8c, the discharge capacities at 0.2 C can reach up to 130.5, 131.9, 131.5 and 135.1 mAh g⁻¹ with corresponding capacity retention rates of 90.16, 90.66, 88.33 and 84.87% after 100 cycles for the samples obtained at 750, 800, 850 and 900 °C, respectively. The accelerated capacity decay begins at 850 °C and becomes more obvious at 900 °C. It is noted that the edges and corners of the octahedral-like LiNi_0.5Mn_1.5O₄ crystals begins to melt at 850 °C and was seriously broken up at 900 °C as shown in Figs. 6c and 6d. Therefore, it is thought that the accelerated capacity decay for the samples obtained at 850 and 900 °C is related to the damaged integrity of the octahedral-like crystals. When the octahedral-like crystals were damaged at higher sintering temperatures, the new formed surfaces and nanosize fragments would increase the reactivity of the sample, which, coupled with the increased amount of Mn³⁺ generated at higher sintering temperatures, would increase the Mn dissolution from the LiNi_0.5Mn_1.5O₄ induced by the disproportionation reaction of Mn³⁺ in the electrolyte and thus accelerate the loss of the active material and the decay of the capacity during cycling. ³⁵ Figure 8d shows the cycle performance of the LiNi_0.5Mn_1.5O₄ samples obtained at different lithium excess. The discharge capacities at 0.2 C can reach up to 123.7, 131.9 and 131.1 mAh g⁻¹ for the samples obtained at the lithium excess of 0, 3 and 5%. The lower discharge capacity of the 0%–800 °C sample is due to lithium deficiency caused by the lithium volatilization at a high sintering temperature, and a 3% lithium excess seems to be appropriate for the compensation of lithium volatilization at the sintering temperature of 800 °C. As the lithium excess increases, the capacity retention rates show a slight increase from 87.63, 90.66 to 92.59% at 0.2 C after 100 cycles, which should be associated with the decreased content of Mn³⁺ that decreases the possible Mn dissolution. Figures 8e and 8f show the rate performance of the LiNi_0.5Mn_1.5O₄ samples. When the discharge rate increases from 0.2 to 10 C and subsequently decreases back to 0.2 C, the discharge capacity of all samples can be basically recovered. When the lithium excess is fixed at 3%, the discharge capacities at 10 C can reach up to 77.7, 98.8, 73.9 and 80.8 mAh g⁻¹ for the samples obtained at 750, 800, 850 and 900 °C, respectively. When the sintering temperature is fixed at 800 °C, the discharge capacities at 10 C can reach up to 96.8, 98.8 and 90.3 mAh g⁻¹ for the samples obtained at the lithium excess of 0, 3 and 5%, respectively. It is seen that the discharge capacities at 10 C for the three samples obtained at 800 °C are all higher than 90 mAh g⁻¹, which suggests that it is very important to control the sintering temperature for achieving high-rate LiNi_0.5Mn_1.5O₄. Among these samples, the 3%–800 °C sample gives the best rate performance. Overall, the charge/discharge tests show that the well-crystallized octahedral-like LiNi_0.5Mn_1.5O₄ crystals with better integrity of the 3%–800 °C sample gives excellent electrochemical performance.

Zoom In Zoom Out Reset image size

The electrochemical behavior of these samples was further investigated by the EIS measurement, and Figures 9a and 9b shows the EIS of the LiNi_0.5Mn_1.5O₄ samples recorded at the fully discharged state after 5 cycles at 0.2 C. The spectra of the six samples have a similar profile which is composed of a semicircle in a high-to-medium frequency region and a slope line in the low-frequency region. The semicircle approximately represents the charge transfer impedance (R_ct ) between the electrode and the electrolyte, and the slope line at low frequency represents the Warburg impedance (Z_W ) related to the diffusion of lithium ions. ³⁶ A similar trend of the variation in the semicircle size is seen upon the sintering temperature and the lithium excess. When the sintering temperature increases from 750, 800, 850 to 900 °C or the lithium excess increases from 0, 3 to 5%, the size of the semicircle decreases significantly at first and then increases moderately, and the 3%–800 °C sample has the smallest semicircle, indicative of the smallest charge transfer impedance. The Figs. 9c and 9d show the curves of Z' vs ω^−1/2 in the low-frequency region. The Warburg factor σ can be obtained by the linear fitting of the curve of Z' vs ω^−1/2 according to the formula of Z_real = R_e + R_ct + σω^−1/2 , and then the diffusion coefficient of lithium ions can be calculated based on the equation of D = R² T² /2A² n² F⁴ C² σ² . ³⁷ The diffusion coefficient of lithium ions for these samples are listed in Table II. It is seen that the diffusion coefficient also varies upon the sintering temperature and the lithium excess, which is roughly inverse to the variation in the charge transfer impedance, and exactly the 3%–800 °C sample has the largest diffusion coefficient of lithium ions. A lower charge transfer impedance and a larger lithium ions diffusion coefficient reflect better ionic and electronic conduction in electrode, which explains why the 3%–800 °C sample has the best rate capability. In conclusion, the electrochemical measurements demonstrate that the present route can yield high-performance LiNi_0.5Mn_1.5O₄ crystals.

Zoom In Zoom Out Reset image size

Through the above measurements and analysis, we can find that the trend of the variations in the charge transfer impedance and the diffusion coefficient of lithium ions appears to be well correlated with the evolution of the crystal morphology which is influenced by the sintering temperature and lithium excess, and the octahedral-like LiNi_0.5Mn_1.5O₄ crystals with better integrity and cleaner surface has better electrochemical performance. It is noted that the best rate performance is not observed on the samples with the highest content of Mn³⁺ or with the lowest occupation of transition metal ions in the tetrahedral Li site or with the smallest crystal size. This indicates that the intrinsic bulk property of the LiNi_0.5Mn_1.5O₄ is not the rate-determining factor for the tested composite cathode and cell. Several available reports have noticed similar behavior and pointed out that the lithium ions transport in the composite electrode or the electrolyte might be the rate-limiting step, ^13,38 and the present results reach a similar conclusion. The well-developed octahedral-like LiNi_0.5Mn_1.5O₄ crystals offer favorable surfaces to form better interface which benefits the mitigation of kinetic limitations for the composite cathode and cell to realize superior rate capability. Therefore, from the perspective of material synthesis and application, particular attention should be paid to the surface and morphology of the LiNi_0.5Mn_1.5O₄, and it is of vital importance to produce well-developed micron-sized LiNi_0.5Mn_1.5O₄ via an environmentally benign and economically viable route.

Well-crystallized micron size octahedral-like LiNi_0.5Mn_1.5O₄ crystals can be readily synthesized from the nano precursor derived from the room-temperature solid-state reactions. The structure, Mn³⁺ content, crystal size and morphology and electrochemical behaviors of the LiNi_0.5Mn_1.5O₄ are all affected by the sintering temperature and lithium excess. It is found that the crystal morphology and integrity have a significant effect on the electrochemical properties, and the octahedral-like LiNi_0.5Mn_1.5O₄ crystals with better integrity has better electrochemical performance. The present work provides a green and economical route to effectively synthesize high-quality LiNi_0.5Mn_1.5O₄ crystals.

This work is supported by National Natural Science Foundation of China (grant numbers 51874155 and 51664031).