Facile hydrothermal synthesis and electrochemical properties of orthorhombic LiMnO₂ cathode materials for rechargeable lithium batteries

Abstract

Pure-phase, LiMn₂O₄-mixed and aluminum-doped orthorhombic LiMnO₂ (o-LiMnO₂) cathode materials with high discharge capacity and excellent cyclic stability were prepared by one-step hydrothermal reaction of MnCl₂, EDTA, LiOH, AlCl₃ and NaClO solutions. Chemical composition and aluminum content were affected by temperature and the concentration of LiOH, NaClO and AlCl₃. A mixed phase of Mn₃O₄ and o-LiMnO₂, pure-phase o-LiMnO₂, and a mixed phase of o-LiMnO₂ and LiMn₂O₄ were formed with increasing the concentration of NaClO from 0.08 to 0.25 mol L⁻¹ at 180 °C for 24 h. Adding EDTA and NaClO facilitated the formation of o-LiMnO₂. Al/Mn molar percent ratios in doped o-LiMnO₂ were 0.34, 0.58, 0.91, and 1.22 when Al/Mn molar ratios in hydrothermal system were controlled at 0.05, 0.10, 0.15, and 0.20, respectively. Mixing LiMn₂O₄ and doping Al improved the discharge capacity and cyclic stability of o-LiMnO₂. o-LiMnO₂, the mixture with an o-LiMnO₂/LiMn₂O₄ mass ratio of 2.45, and doped o-LiMnO₂ with an Al/Mn molar percent ratio of 0.58 exhibited initial discharge capacities of 76, 139, and 82 mA h g⁻¹, and cycling capacities of 124, 144, and 156 mA h g⁻¹ after 100 cycles, respectively. This work facilitates the preparation and electrochemical performance improvement of o-LiMnO₂.

---

1. Introduction

Lithium ion batteries have been widely investigated for high-power applications including electric vehicles, plug-in hybrid electric vehicles, and hybrid electric vehicles.^1,2 For the first commercially successful cathode materials of lithium batteries, LiCoO₂ needs to be substituted due to its high cost, toxicity, and low discharge capacity of 140–150 mA h g⁻¹.^3–5 Alternatively, o-LiMnO₂ and spinel LiMn₂O₄ composites have been considered as promising cathode materials for rechargeable lithium ion batteries owing to their abundant resources, low cost, environmental friendship, safety, high voltage, and good rate capability, which has been a hot research topic in the field of chemical power source.^6,7 However, LiMn₂O₄ exhibits lower discharge capacity of 110–120 mA h g⁻¹ and cyclic stability due to accelerated Mn dissolution at an elevated temperature cycling and structural instability for Jahn–Teller distortion.^4,5,7–9 o-LiMnO₂ shows a higher theoretical discharge capacity of 285 mA h g⁻¹, which is almost twice that of spinel LiMn₂O₄.¹⁰ Additionally, o-LiMnO₂ shows a lower capacity fade rate, which is commonly affected by chemical structures and composition. Therefore, much effort has been devoted to the fabrication of pure-phased and doped o-LiMnO₂ particles for cathode materials of lithium batteries.^4,10–12

At present, o-LiMnO₂ samples are usually prepared by high temperature solid-state reactions,^4,11 sol–gel process,¹³ ion exchange,⁷ and hydrothermal method.^3,10,14 As for the above methods, micro-structured precursors including Mn₂O₃, MnCO₃, NaMnO₂, and γ-MnOOH should be first fabricated for the consequent transformation of o-LiMnO₂. Multi-step and high temperature reactions will increase the energy consumption and cost. Recently, o-LiMnO₂ nano-particles were prepared under mild hydrothermal conditions from KMnO₄, MnCl₂, and LiOH sources, but too much lithium source, Li/Mn molar ratio more than 30, was used.¹⁵ One-step hydrothermal method was used to prepare uniform nano-sized LiMnO₂ with low energy consumption, and the morphologies could be easily controlled, and the as-obtained LiMnO₂ showed excellent electrochemical performance.^3,16 Generally, for the hydrothermal synthesis of o-LiMnO₂, a high concentration of LiOH was required, and Li/Mn molar ratio in raw materials was usually larger than 13.^15–17 In order to reduce cost for industrial application, the preparation of o-LiMnO₂ using low Li/Mn molar ratio is highly desirable.

The weak cyclic stability of o-LiMnO₂ is associated with Jahn–Teller distortion for the change of manganese oxidation states during charge–discharge process, and the partial substitution of Mn by metal ions is commonly used to stabilize the layered structure.^11,12,18 Cr³⁺ and Al³⁺ ions prevent manganese ion migration during cycling, reducing the tendency of the layered structure converting to the more stable spinel.¹¹ Cr substitution decreases the capacity fade rate of o-LiMnO₂, and the reversible capacity maintains 160 mA h g⁻¹ after 50 cycles with C/5 rate.¹⁸ Co doped LiMnO₂ was formed using ion exchange of NaMnO₂ synthesized by a solid state reaction, and kept 200 mA h g⁻¹ even after 20 cycles at 0.1 mA cm⁻² over the potential range of 2.6–4.85 V.⁷ Ni doped LiMnO₂ also showed 198 mA h g⁻¹ with retention ratio of 96% after 25 cycles at C/3, and the layered structures exhibited no phase change during charge–discharge process.¹⁹ The indium- and sulfur-doped LiMnO₂ samples with orthorhombic structure as cathode materials of lithium batteries showed excellent electrochemical capacity of 267.9 mA h g⁻¹.²⁰ In other words, doping would remarkably improve LiMnO₂ electrochemical performance including specific capacity and cyclic stability. Hydrothermal synthesis of doped LiMnO₂ was usually conducted using the as-prepared precursors, such as MnCO₃ (ref. 11 and 18) and Mn₂O₃.^18–20 To the authors' knowledge, there was no facile one-step hydrothermal process to prepare metal doped o-LiMnO₂.

As for hydrothermal synthesis, pure-phased LiMnO₂ was difficult to synthesize likely due to the further oxidation of LiMnO₂ to form Li₂MnO₃.²¹ An appropriate redox rate should be controlled to obtain single-phased LiMnO₂ with low-valent manganese. Under refluxing conditions, sodium hypochlorite (NaClO) could oxidize Mn(II) source to form uniform ramsdellite-type manganese oxide with various morphologies.²² MnPO₄·H₂O and hureaulite containing low-valent Mn(III/II) can also be fabricated by adjusting reactant ratios and reaction temperatures.^22,23 Our previous studies have suggested that manganese minerals with different oxidation states of manganese could be formed when NaClO worked as oxidant.

In this work, pure-phased, LiMn₂O₄-mixed and aluminum-doped LiMnO₂ cathode materials were hydrothermally fabricated using 0.1 mol L⁻¹ MnCl₂ and NaClO solutions. Complexant EDTA was used to chelate Mn²⁺ to adjust the reaction rate to obtain single-phased o-LiMnO₂. The effect of reactant concentration and temperature on the chemical composition of products was studied. The influence of composition and Al doping on the unit cell parameters and the corresponding electrochemical performance of o-LiMnO₂ was investigated by X-ray diffraction spectrometry, Rietveld structure refinement, Fourier transform infrared spectroscopy, cyclic voltammetry, and electrochemical impedance spectroscopy.

2. Materials and methods

2.1. Materials and chemicals

MnCl₂·4H₂O (AR, ≥99.0%), ethylene diamine tetraacetic acid tetrasodium salt (C₁₀H₁₂N₂O₈Na₄·4H₂O, EDTA, AR, ≥40%), AlCl₃·6H₂O (AR, ≥97.0%), LiOH·H₂O (AR, >99.0%), anhydrous ethyl alcohol (AR, ≥99.7%) and NaClO solution (AR, available chlorine content ≥5.2%) were all purchased from China National Medicine Group Shanghai Chemical Reagent Company. High purity metallic lithium pellets (99.9%) were supplied by Wuhan Newthree Technology Co. Ltd., China. All reagents used were of analytical grade, unless otherwise noted.

2.2. Preparation and characterization of o-LiMnO₂

MnCl₂·4H₂O (0.989 g) and EDTA (2.262 g) were first dissolved in distilled deionized water (DDW) in a beaker, and NaClO solution (6.717 g) and LiOH·H₂O (1.253 g) were then added to the above solution system. The suspended solution was adjusted to 50 mL by adding DDW and subsequently transferred into a Teflon-lined stainless steel autoclave with a volume of about 100 mL, sealed and maintained at 120, 150, and 180 °C, respectively, for 24 h. The concentration of MnCl₂, EDTA, LiOH, and NaClO was controlled as 0.1, 0.1, 0.6, and 0.1 mol L⁻¹, respectively. The suspension was then cooled to room temperature and washed by repeated filtration until the conductivity of filtrate was below 20.0 μS cm⁻¹, and then dried in an oven at 60 °C for 24 h.

In order to synthesize Al doped o-LiMnO₂, AlCl₃·6H₂O was added to the above reaction system, and the concentration of LiOH and NaClO was slightly changed accordingly. When the concentration of AlCl₃ was 0.005, 0.010, 0.015, and 0.020 mol L⁻¹, LiOH concentration was controlled as 0.6, 0.6, 0.6, and 0.8 mol L⁻¹, and NaClO concentration was adjusted as 0.10, 0.11, 0.12, and 0.12 mol L⁻¹, respectively. The as-prepared Al doped LiMnO₂ samples were washed and dried as above.

The as-obtained products were identified by X-ray diffraction spectrometry (XRD, Bruker D8 Advance diffractometer with Cu Kα) at a scan rate of 1° per min with a step of 0.02°. Structure refinement of XRD data was carried out by Rietveld method using TOPAS software (Bruker). The micro-morphologies of samples were characterized by scanning electron microscopy (FE-SEM, JEOL, JSM-6700F) and transmission electron microscopy (TEM, Hitachi-7650). The particle size distribution of the as-obtained samples was analyzed using software ImageJ 1.46 (National Institutes of Health, USA) by SEM images. Fourier transform infrared spectroscopy (FTIR, Nicolet 8700) was carried out using a DTGS detector by making pellets with KBr powder and LiMnO₂/KBr mass ratio about 1 : 50, and the resolution was set at 4 cm⁻¹ with a scan number of 64. BET surface area of LiMnO₂ was tested with a Quantachrome Autosorb-1-C automated N₂ gas adsorption system. The concentration of Al and Mn was measured using an inductively coupled plasma atomic emission spectrometry (ICP-AES, Varian Vista-MPX), and further examined by Varian AAS 240FS atomic absorption spectrometer (Varian, Australia).

2.3. Electrochemical measurements of o-LiMnO₂

The electrochemical characteristics of the synthesized materials were evaluated by galvanostatic charge–discharge using 2032 coin-type cells assembled in an argon-filled glove box, and the simulated battery test controlled by a battery testing system (Shenzhen Neware Electric Co. Ltd., China). The cathode mixtures consisted of 75 wt% active materials (including o-LiMnO₂, a mixture of o-LiMnO₂ and LiMn₂O₄, and Al doped o-LiMnO₂) dried at 110 °C for 12 h, 15 wt% acetylene black and 10 wt% polytetrafluoroethylene binder on an aluminum mesh. The anode and reference electrodes were the same lithium foil. The electrolyte was 1 mol L⁻¹ LiClO₄ in a mixture of EC, DMC, and EMC with equal volume in this work. Celgard 2400 membrane worked as a separator. The charge–discharge cycles were carried out between 4.3 V and 2.4 V (vs. Li⁺/Li) at a current density of 100 mA g⁻¹ at 25 ± 1 °C. The cathode foils were characterized by XRD after different cycles of charge–discharge test.

The electrochemical behaviors of active materials including o-LiMnO₂ and the mixture of o-LiMnO₂ and LiMn₂O₄ were studied by cyclic voltammetry in the above electrolyte at a scan rate of 0.1 mV s⁻¹. Electrochemical impedance spectroscopy (EIS) was used to characterize the interface properties of the modified electrodes. EIS was employed using a 5 mV rms AC perturbation voltage of frequencies ranging from 0.01 Hz to 100 kHz. EIS data were analyzed by fitting to equivalent circuit models using ZSimpWin 3.10 software package (EChem Software). Electrochemical control was achieved by a computer assisted CHI660B electrochemical system (Shanghai Chenhua Instrument Co. Ltd., China).

3. Results

3.1. The preparation of o-LiMnO₂

Hydrothermal reactions were conducted in different solution systems containing MnCl₂, LiOH, EDTA, and NaClO at 120, 150, and 180 °C, respectively, for 24 h. As shown in Fig. 1a, a mixture of Mn₃O₄ (JCPDS: 01-1127), o-LiMnO₂ (JCPDS: 86-0352), and Li₂MnO₃ (JCPDS: 73-0152) was formed when 0.08–0.10 mol L⁻¹ NaClO was used in the presence of 0.1 mol L⁻¹ MnCl₂, 0.6 mol L⁻¹ LiOH, and 0.1 mol L⁻¹ EDTA at 120 °C. A mixed phase of birnessite (JCPDS: 87-1749) and Li₂MnO₃ was obtained when the concentration of NaClO was increased to 0.15 mol L⁻¹. That is, high concentration of NaClO facilitates the formation of high valent manganese oxides. The influence of hydrothermal temperature on the product composition was also investigated. When reaction temperature was increased to 150 °C, a mixed phase of Mn₃O₄, o-LiMnO₂, LiMn₂O₄ (JCPDS: 88-1026) and Li₂MnO₃ was formed using 0.08 mol L⁻¹ NaClO solution as exhibited in Fig. 1b. When NaClO concentration was increased to 0.1 mol L⁻¹, a mixture of o-LiMnO₂ and LiMn₂O₄ was formed, which are both active cathode materials for rechargeable lithium batteries.^6,7 Fortunately, when hydrothermal reaction was conducted at 180 °C, pure-phased o-LiMnO₂ was achieved if NaClO concentration was kept at 0.10 mol L⁻¹. However, a mixture of o-LiMnO₂ and Mn₃O₄ was formed in the presence of 0.08 mol L⁻¹ NaClO. Moreover, a mixture of o-LiMnO₂ and LiMn₂O₄ was formed when NaClO concentration was further increased to 0.15 and 0.25 mol L⁻¹ (Fig. 1c). In another word, hydrothermal temperature and enough NaClO supply played critical roles in the formation of pure o-LiMnO₂.

Fig. 1 XRD patterns of products synthesized in hydrothermal systems containing 0.1 mol L⁻¹ MnCl₂, 0.1 mol L⁻¹ EDTA, LiOH and NaClO for 24 h: 0.6 mol L⁻¹ LiOH with different concentration of NaClO at 120 °C (a), 150 °C (b), and 180 °C (c); 0.1 mol L⁻¹ NaClO with different concentration of LiOH at 180 °C (d).

Decreasing hydrothermal temperature and increasing the concentration of NaClO facilitate the formation of lithium-manganese-oxide compounds with higher-valent manganese. Only when the hydrothermal reaction was conducted by using 0.1 mol L⁻¹ NaClO at 180 °C for 24 h, could pure-phased o-LiMnO₂ be formed, and it was represented as M₀. When hydrothermal reaction was performed at 150 °C with the same reagent compositions as above, a mixture of o-LiMnO₂ and LiMn₂O₄ was obtained with o-LiMnO₂/LiMn₂O₄ mass ratio of 2.45 : 1 analyzed by XRD (Fig. 1b), which was designated as M_m.

Above results suggested that Mn₃O₄ and o-LiMnO₂ were first formed, and then LiMn₂O₄ and Li₂MnO₃ were formed with increasing the concentration of NaClO. Diffraction peak at 18.5°, 2θ, can be assigned to Li₂MnO₃, the common impurity phase in o-LiMnO₂ materials.^5,13–15,24 Although the diffraction peaks of Li₂MnO₃ and LiMn₂O₄ overlap at about 18.5°, the chemical compositions can be identified by other peaks. LiMn₂O₄ exhibits a relative strong diffraction peak at 63.8°, which is observed in the XRD patterns of products as shown in Fig. 1b and c. Li₂MnO₃ shows a relative weak diffraction peak at 20.8°, however, there is no diffraction peaks at this diffraction angle in this work. Therefore, it could be safely concluded that a mixture of o-LiMnO₂ and LiMn₂O₄ was formed when 0.1 mol L⁻¹ NaClO was used at 150 °C (Fig. 1b). When hydrothermal reaction was conducted at 180 °C by using 0.15 and 0.25 mol L⁻¹ NaClO, the diffraction peaks for the impurity phase in o-LiMnO₂ materials could also be ascribed to LiMn₂O₄ (Fig. 1c).

The effect of LiOH concentration on the formation process of o-LiMnO₂ was investigated in hydrothermal system containing 0.1 mol L⁻¹ MnCl₂, 0.1 mol L⁻¹ NaClO, 0.1 mol L⁻¹ EDTA at 180 °C for 24 h. As shown in Fig. 1d, pure-phased Mn₃O₄ was formed when LiOH concentration was controlled as in the range of 0.1–0.2 mol L⁻¹, and a mixture of Mn₃O₄ and o-LiMnO₂ was formed when LiOH concentration was increased to 0.4 mol L⁻¹. Single-phased o-LiMnO₂ was formed when LiOH concentration was increased to 0.6 and 0.8 mol L⁻¹.

EDTA affected the reaction process possibly due to the complexation reaction with Mn²⁺. As for the hydrothermal reaction of 0.1 mol L⁻¹ MnCl₂, 0.1 mol L⁻¹ NaClO, 0.6 mol L⁻¹ LiOH at 180 °C for 24 h, a mixture o-LiMnO₂, Mn₃O₄, and Li₂MnO₃ was formed rather than pure-phased o-LiMnO₂ (Fig. S1†), indicating a definite necessity of EDTA for the formation of pure-phased o-LiMnO₂. Fig. 2a and b show the SEM images of M₀ and M_m, respectively. As a mixture of o-LiMnO₂ and LiMn₂O₄, M_m exhibited a non-uniform size distribution, and the bigger particles were likely due to o-LiMnO₂ since similar morphologies were observed for the pure-phased o-LiMnO₂ (Fig. 2a).

Fig. 2 SEM images of M₀ (a), M_m (b), M₃ (c), M₆ (d), M₉ (e) and M₁₂ (f).

3.2. The preparation of aluminum doped o-LiMnO₂

In order to prepare aluminum doped o-LiMnO₂, AlCl₃ was added to the hydrothermal system, in which pure-phased o-LiMnO₂ (M₀) was formed at 180 °C for 24 h. When Al(III)/Mn(II) molar ratio in reaction system was increased, the concentration NaClO and LiOH should be slightly increased to fabricate single-phased aluminum doped o-LiMnO₂ samples. Single-phased aluminum doped o-LiMnO₂ samples could be formed when Al/Mn molar ratios in reaction system were controlled as 0, 0.05, 0.10, 0.15, and 0.20 under the following conditions: LiOH concentration was 0.6, 0.6, 0.6, 0.6, and 0.8 mol L⁻¹, and NaClO concentration was 0.1, 0.1, 0.11, 0.12 and 0.12 mol L⁻¹, respectively. Al/Mn molar percent ratios in o-LiMnO₂ samples were determined to be 0.34, 0.58, 0.91, and 1.22 when Al/Mn molar ratios in hydrothermal system were controlled as 0.05 : 1, 0.10 : 1, 0.15 : 1, and 0.20 : 1, respectively. Fig. S2† shows the XRD patterns of the as-obtained samples, and they were designated as M₃, M₆, M₉ and M₁₂ according to Al/Mn molar ratios in doped o-LiMnO₂ being 0.34, 0.58, 0.91, and 1.22, respectively. It was noted that the Al/Mn molar ratios in o-LiMnO₂ phase were significantly positively correlated with those in hydrothermal system, and linearly dependent coefficient R² was 0.9971. Fig. 2c–f shows the SEM images of M₃, M₆, M₉ and M₁₂, and flake-like structures were observed. The content of doped aluminum had no obvious effect on the micro-morphologies of o-LiMnO₂.

The particle size distribution of the as-obtained samples was further analyzed using software ImageJ as shown in Fig. S3.† M_m shows a particle distribution range from 50 to 150 nm. M₀ was formed with bigger particles in the range of 100–200 nm owing to the higher hydrothermal temperature. Aluminum doped o-LiMnO₂ samples exhibit a broad range of particle sizes, which increased with an increase of aluminum content in doped o-LiMnO₂. The micromorphologies and particle size were further characterized by TEM as shown in Fig. S4.† After aluminum doping, the particle size slightly increased with an increase in the content of aluminum in o-LiMnO₂ likely due to that adding AlCl₃ changed the complex state of Mn²⁺ for the formation of Mn²⁺ complexes of chlorides, such as [MnCl₃]⁻, [MnCl₄]²⁻, and [MnCl₆]⁴⁻. As for the aluminum doped o-LiMnO₂ powder, the particles had remarkable difference in size, but similar micromorphologies were observed (Fig. 2 and S4†). The influence mechanism of AlCl₃ on the micromorphology and particle size of o-LiMnO₂ needs to be further studied in future.

Fig. 3 illustrates the FTIR spectra of M₀, M_m, M₃, M₆, M₉, and M₁₂. Most of the bands in the low wave-number region of 400–800 cm⁻¹ are attributed to the vibrations of MnO₆ octahedra, and they are main fingerprints of o-LiMnO₂ and LiMn₂O₄.^25,26 The dominant absorption peaks for pure-phased o-LiMnO₂ (M₀) at 487 and 633 cm⁻¹ were assigned to the asymmetry (ν_as) and symmetry stretching modes (ν_s) of Mn–O,⁴ respectively. A smaller feature at 439 cm⁻¹ was likely owing to LiO₆ octahedra.^27,28 Absorption peak at 1636 cm⁻¹ was assigned to the adsorbed water molecules. As for LiMn₂O₄, the high frequency bands of the FTIR spectra located at 516 and 632 cm⁻¹ are attributed to the asymmetric stretching modes of the MnO₆ group.^26,28,29 Doping Al affected the absorption bands for the symmetry stretching of o-LiMnO₂, and red shifts occurred from 633 cm⁻¹ for M₀ to 623, 603, 613, and 617 cm⁻¹ for M₃, M₆, M₉, and M₁₂, respectively. The shift for the absorption peak in FTIR spectra is usually associated the change of interatomic distance in crystal structures, and the similar change was observed in high-frequency components in the band of MO₆ vibration around 600 cm⁻¹ when Ni was substituted for cobalt in LiCoO₂.³⁰

Fig. 3 FTIR spectra of M₀, M_m, M₃, M₆, M₉, and M₁₂.

Table 1 shows cell parameters obtained by Rietveld structure refinement, and specific surface area (SSA) for M₀, M_m, M₃, M₆, M₉, and M₁₂. As the content of doped Al in o-LiMnO₂ increasing, parameter a decreased, while b increased. Moreover, parameter c was first slightly decreased and then increased, and cell volume slightly shrinked with increasing the content of doped aluminum due to that Al³⁺ ionic radius (0.0535 nm) is smaller than that of Mn³⁺ (0.0645 nm).^5,31 The most significant decrease along a axis occurred after substitution, which would possibly improve the electrochemical stability due to that the Jahn–Teller distortion of the [Mn³⁺O₆] octahedra induces an elongation along the a direction.⁵

Table 1 Cell parameters and specific surface area (SSA) of the as-obtained samples

Samples	a (nm)	b (nm)	c (nm)	Cell volume (nm^3)	Rwp (%)	SSA (m^2 g^−1)	Al/Mn molar ratio (%)
M0	0.4577	0.5751	0.2811	0.07398	12.6	9.1	0
Mm	—	—	—	—	—	21.3	—
M3	0.4569	0.5756	0.2811	0.07392	13.5	9.7	0.34
M6	0.4561	0.5759	0.2810	0.07383	13.6	10.2	0.58
M9	0.4548	0.5764	0.2813	0.07373	14.7	10.6	0.91
M12	0.4542	0.5764	0.2812	0.07362	13.5	13.5	1.22

---

It should be pointed out that these experiments have been repeated more than three times for that sufficient active materials should be supplied for the characterization of electrochemical performance. The as-prepared powder samples were characterized by XRD and FTIR each time, and same results were obtained. The SEM images also showed the uniform morphologies, which further suggested that the pure products were synthesized. The composition of Al doped o-LiMnO₂ was also analyzed by Varian AAS 240FS atomic absorption spectrometer (Varian, Australia), besides the ICP-AES as mentioned above. That was to say, the composition of Al doped o-LiMnO₂ was confirmed by two analysis techniques, further suggesting that the experiments were repeatable.

3.3. The electrochemical performance of o-LiMnO₂

Fig. 4 shows discharge specific capacity and cyclic stability of M₀, M₃, M₆, M₉, M₁₂, and M_m at a current density of 100 mA g⁻¹. M₀ had a relatively low initial discharge specific capacity of 76 mA h g⁻¹, and increased to about 139 mA h g⁻¹ after 25 cycles. The cyclic capacity was kept stable, and discharge specific capacity was about 124 mA h g⁻¹ after 100 cycles. Among the as-obtained samples, M_m showed the largest discharge specific capacity. The initial discharge capacity of M_m was 139 mA h g⁻¹, and increased to 171 mA h g⁻¹ after 25 cycles, and retained about 144 mA h g⁻¹ after 100 cycles. As reported, a mixture of LiMnO₂ and Li₂MnO₃ with mass ratio of 1.40 : 1 exhibited the best electrochemical performance.³² In this work, the mixture of o-LiMnO₂ and LiMn₂O₄ with o-LiMnO₂/LiMn₂O₄ mass ratio about 2.45 : 1 exhibited excellent electrochemical performance.

Fig. 4 Discharge specific capacity vs. cycle number at the current density of 100 mA g⁻¹ for M₀, M₃, M₆, M₉, M₁₂, and M_m after heat-treatment at 110 °C for 12 h.

Doping aluminum improved the electrochemical performance of o-LiMnO₂ as displayed in Fig. 4. The discharge capacity increased at the initial few cycles and then kept steady after 25 cycles. The increase of discharge capacity in the first few cycles was likely due to the electrochemical activation of cathode materials and the improvement of lithium ion diffusion channels in active materials, resulting in higher lithium utilization during initial cycling stage. The cyclic capacity of doped o-LiMnO₂ was first increased and then decreased with an increase in the content of aluminum. The initial discharge capacity of M₆ was 82 mA h g⁻¹, and kept steady at about 160 mA h g⁻¹ after 25 cycles, and slightly decreased to 156 mA h g⁻¹ after 100 cycles. The as-obtained o-LiMnO₂ samples exhibited better electrochemical performance than those synthesized by other methods.^5,7,33

Intermediate products in charge–discharge processes were studied using XRD after different cycles. As illustrated in Fig. 5a, a mixed phase of o-LiMnO₂ and LiMn₂O₄ was formed after charge–discharge test of the first cycle. o-LiMnO₂ was almost completely transformed into LiMn₂O₄ after five cycles, and a mixture of LiMn₂O₄, monoclinic LiMnO₂ (m-LiMnO₂, JCPDS: 87-1255) and γ-MnO₂ was formed after 100 cycles. The similar results were obtained for the aluminum doped o-LiMnO₂ sample of M₆ after 100 cycles as shown in Fig. 5b. Li₂Mn₂O₄ (JCPDS: 84-1523) was also formed in this process. As reported, o-LiMnO₂ was converted to other manganese oxides, such as Li_xMn₂O₄,^7,33,34 which was then gradually transformed into m-LiMnO₂ and MnO₂.³⁵ In this work, similar intermediate products were also observed.

Fig. 5 XRD patterns of intermediate products after different cycles of charge–discharge test for M₀ (a), and 100 cycles of charge–discharge test for M₀ and M₆ (b).

Electrochemical impedance spectroscopy was collected to analyze the change trend of electrical conductivity and ionic migration rate. The semicircle at the high frequency region was related to the solid electrolyte interface film of the newly assembled cell. The curve at low frequency region can be explained as the diffusion-controlled process in the electrolytes. Fig. 6 shows the typical families of Nyquist plots obtained from electrodes of M₀, M_m, M₃, M₆, M₉, and M₁₂ and measured using 3.15 V of open circuit voltage at 25 °C before charge–discharge test. The particle size, morphologies and specific surface area of the above samples were similar as shown in Fig. 2, S3,† and Table 1, and the same electrochemical reaction mode R(QR)WC was used to compare their reaction impedance (R_ct) and lithium ion diffusion rates in this work.³⁶ As shown in Table 2, R_ct of M_m was higher than that of M₀. It was found that doping aluminum improved conductivity and the corresponding R_ct was decreased. A Warburg type element in the low frequency region corresponded to solid-state diffusion of lithium ions and a steep line at the lower frequencies. Diffusion rate of lithium ions in solid phase could be reflected by the linear slope in the low frequency region, and the Z_w values for all the samples were compared as shown in Table 2. As for M₀, M₃, M₆, M₉, and M₁₂, the diffusion rate of lithium ions increased with a decrease in the cell volume of Al doped o-LiMnO₂.

Fig. 6 Electrochemical impedance spectra of M₀, M₃, M₆, M₉, M₁₂, and M_m as Nyquist plots measured at 25 °C before charge–discharge test.

Table 2 EIS model (R(QR)WC) parameters for M₀, M_m, M₃, M₆, M₉ and M₁₂ before discharge–charge tests

Samples	M0	Mm	M3	M6	M9	M12
Rct (Ω)	229.4	297.9	176.1	145.0	252.5	203.0
Error (%)	1.246	0.935	0.779	0.913	1.357	0.982
Zw (S s^−1/2)	0.011	0.009	0.015	0.015	0.007	0.009
Error (%)	5.387	3.685	3.663	3.559	5.111	3.173

---

4. Discussion

4.1. The influence on o-LiMnO₂ formation process

Hydrothermal reactions have been widely conducted to synthesize o-LiMnO₂ from different low-valent manganese oxide precursors including Mn₃O₄, Mn₂O₃ and MnOOH, whose micro-morphologies could possibly be preserved during the transformation of target products.^{3,10,14,17,20,24} In the present work, Mn₃O₄ was found to be one of the intermediate products in the formation process of o-LiMnO₂. When the concentrations of MnCl₂, LiOH, and EDTA were controlled as 0.1, 0.6, and 0.1 mol L⁻¹, respectively, Mn₃O₄ was first formed, and then oxidized to Li₂MnO₃ and birnessite-type manganese oxides with increasing the concentration of NaClO at 120 °C (Fig. 1a). The formation of layered birnessite was attributed to the high concentration of Na⁺, which participated in the reaction and intercalated into the interlayer to support the layered structure. When hydrothermal temperature was increased to 150 °C, the content of Li₂MnO₃ in solid products increased with increasing the concentration of NaClO (Fig. 1b). Li₂MnO₃ was formed and transformed into LiMnO₂ at a higher hydrothermal temperature.^15,21 Birnessite was not formed likely due to the formation of Li₂MnO₃ and LiMn₂O₄ at 150 °C.²¹ These results suggested that birnessite could be used as precursor to prepare nano-sized Li₂MnO₃, LiMn₂O₄ and LiMnO₂ with various morphologies because the morphologies and particle size of birnessite could be controlled easily as reported in our previous work.³⁷ The transformation of LiMnO₂ was a complicated multi-step process, and influencing factors included hydrothermal temperature, LiOH concentration, the additives of organics and cation, etc.

Hydrothermal temperature plays an important role in the formation of LiMnO₂. Different temperatures were needed when Mn₃O₄, Mn₂O₃ and MnOOH worked as precursors. For example, LiMnO₂ samples were hydrothermally synthesized using Mn₂O₃ precursor at 150 °C,²⁰ 160 °C (ref. 10, 14 and 24) and 170 °C.¹⁷ When MnOOH was used instead, the hydrothermal reaction was conducted at 200 °C.³ In the case of Mn₃O₄ used as a precursor, hydrothermal treatment was performed with high concentration of LiOH at 170 °C for 4 days.³⁸ Hydrothermal temperature was somewhat higher when Mn₃O₄ was used likely due to its lower reaction activity than that of Mn₂O₃, and oxidant was needed to prepare o-LiMnO₂. In this work, pure-phased o-LiMnO₂ was formed by adding oxidant NaClO at 180 °C for 24 h. Hydrothermal temperature should be elevated to synthesize the target product of o-LiMnO₂ possibly due to the further redox reactions between organics and manganese oxides, such as birnessite, Li₂MnO₃ and LiMn₂O₄.³⁹

In order to save raw materials and manufacturing cost, the dosage of LiOH was concerned in the preparation process of LiMnO₂. Only Mn₃O₄ was formed when LiOH of low concentration (≤0.2 mol L⁻¹) was used, and a mixture of Mn₃O₄ and LiMnO₂ was formed when LiOH was 0.4 mol L⁻¹. Pure-phased target product was formed when 0.6 mol L⁻¹ LiOH was used, and crystallinity was increased when 0.8 mol L⁻¹ LiOH was used instead (Fig. 1d). LiOH was reacted with Mn₃O₄ and Mn₂O₃ to form o-LiMnO₂, and high concentration reactant facilitated this transformation process.^17,24,38 High concentration of LiOH was usually supplied to fabricate o-LiMnO₂ when low-valent manganese oxides were used as precursor for hydrothermal reaction. It was possible due to that water molecules were produced during the formation of o-LiMnO₂, which reduced the formation rate of the o-LiMnO₂ phase and increased the reaction time,¹⁷ particularly for the mass transport of Li⁺ in the pores of manganese oxides during the hydrothermal process. As for the one-step synthesis of LiMnO₂, a mixture of LiMnO₂ and Li₂MnO₃ was formed when LiOH/MnSO₄ molar ratio was less than 4 : 1.²¹ Although Li/Mn molar ratio was increased to 30, the final product was a o-LiMnO₂ crystalline phase with trace amounts of Li₂MnO₃,¹⁵ and the mixture was also observed as intermediate product in this work (Fig. 1a and b). The rapid generation of manganese oxides was unfavorable to the formation of target product due to that big crystal particle would inhibit the further reaction. As reported, LiMnO₂ crystal grain would not be formed when hydrothermal temperature was too high.²⁴ High concentration of Li⁺ facilitated the formation of o-LiMnO₂ mostly due to an increase in Li⁺ mobility, which accelerated its reaction with Mn(III/II) source to form lithium manganese oxides, and was also beneficial to the formation of a layered structure. In the present work, Li/Mn molar ratio could be controlled as 6 : 1, which was lower than those as reported.^3,15,24 Low concentration of Li⁺ was performed to reduce cost, and complexation reaction resulted in a slow crystallization process and decreased the negative influence of low mass transfer rate of Li⁺ on the formation of o-LiMnO₂. When EDTA was not added to the reaction system, pure phased LiMnO₂ was not formed, and a mixed phase of o-LiMnO₂, Mn₃O₄, and Li₂MnO₃ was generated instead (Fig. S1†). The formation rate of manganese oxides would be remarkably decreased due to the complexation reaction of EDTA and Mn²⁺ before hydrothermal reaction. As reported in our previous work, rapid transformation from Mn(OH)₂ precipitate to Mn₃O₄ occurred in air.⁴⁰ However, Mn₃O₄ was not formed in alkali solution after adding EDTA into the hydrothermal system in the present work. The formation rate of manganese oxides would be remarkably decreased due to the formation of the Mn²⁺ complexes of EDTA before hydrothermal reaction. The formation of complexes by adding organics tetrabutylammonium bromide and tetraoctylammonium bromide into MnSO₄ solution could prevent its rapid reaction with KMnO₄ to form cryptomelane-type manganese oxides with different degrees of crystallinity.³⁹ LiMnO₂ was formed with Li/Mn molar ratio as low as 1.05 when Mn₂O₃ precursor was used in miscible solvents with a low reaction rate.¹⁷ In order to further reduce cost, some cheaper complexant would be used to substitute EDTA in future work.

The addition of AlCl₃ consumed OH⁻ to form AlO₂⁻, and also affected the oxidation capacity of NaClO in strong alkaline solution systems.^22,23 The preparation of doped o-LiMnO₂ by one-step hydrothermal synthesis was seldom reported, which was likely due to the precipitation reaction of metal cations and hydroxide ions. In this work, soluble AlO₂⁻ was formed using AlCl₃ in LiOH solution of high concentration. In order to adjust the alkalinity and oxidation capacity of NaClO, the concentrations of LiOH and NaClO were slightly increased when AlCl₃ concentration was increased. LiOH concentrations were controlled as 0.6, 0.6, 0.6, 0.6, and 0.8 mol L⁻¹, and NaClO concentrations were controlled as 0.1, 0.1, 0.11, 0.12, and 0.12 mol L⁻¹, when Al/Mn molar ratios in reaction system were controlled as 0, 0.05, 0.10, 0.15, and 0.20, respectively. When the concentrations of LiOH and NaClO were not adjusted accordingly, a mixture of Mn₃O₄ and o-LiMnO₂ with a small amount of LiMn₂O₄ was formed (figure not given). It should be pointed out that Al/Mn molar percent ratios in o-LiMnO₂ phase were only 0.34, 0.58, 0.91, and 1.22 when Al/Mn molar ratios in reaction system were controlled as 0.05 : 1, 0.10 : 1, 0.15 : 1, and 0.20 : 1, respectively. The formation of anion AlO₂⁻ did not well facilitate the formation of doped LiMnO₂ with high content of Al³⁺ ions. The detailed reaction processes and influence mechanisms are currently underway.

4.2. The effect on o-LiMnO₂ electrochemical performance

Chemical composition, crystal structure, crystallinity and cell parameters possibly affected the electrochemical performance of active electrode materials. The cyclic voltammetric behavior of M₀ and M_m was studied after two cycles of charge–discharge activation. As shown in Fig. 7, oxidation/reduction current peaks A/A′ (3.9 V) and B/B′ (4.1 V) were associated with the reversible deinsertion/insertion processes of lithium ion between LiMn₂O₄ and MnO₂, and a pair of peaks C/C′ (3.2 V/2.7 V) corresponded to the insertion/extraction of lithium between cubic LiMn₂O₄ and tetragonal Li₂Mn₂O₄,^5,41 and these intermediate products could be determined (Fig. 5). More than 50% Jahn–Teller ions (Mn³⁺) presented in Li₂Mn₂O₄ participated in the transformation process of cubic LiMn₂O₄ to tetragonal Li₂Mn₂O₄, and the phase transformation results in the volume shrinkage of unit cell that imparts a high degree a stress in grains. Grain fracture may occur with the repeating insertion/extraction of Li⁺ in this process, which finally results in the capacity loss.^9,41 The reduction current peak intensity of M_m was stronger than that of M₀ likely due to the participation of LiMn₂O₄ in M_m, which worked as an intermediate product after a few charge–discharge cycles (Fig. 5). The discharge capacity of M_m was higher than that of M₀ because the former needed not to be activated in the initial stage (Fig. 4). Small particle size and large specific surface area facilitate the high electrochemical performance of LiMnO₂ and LiMn₂O₄ cathode materials.^13,42 It was noted that the specific surface area of M_m was about 21.3 m² g⁻¹, two times larger than that of M₀ (9.1 m² g⁻¹), and it may lead to a better electrochemical performance. As a fact, the as-obtained o-LiMnO₂ would be oxidized and transformed into LiMn₂O₄ through heat-treatment at 300–600 °C in air for 2 h (Fig. S5†). LiMn₂O₄, M₆₀₀, from the transformation of o-LiMnO₂ using heat treatment at 600 °C with specific surface area about 12.6 m² g⁻¹, exhibited an initial discharge capacity of 134 mA h g⁻¹, and kept steady at about 120 mA h g⁻¹ after 20 cycles, and decreased to 117 mAh g⁻¹ after 100 cycles (Fig. S6†). The decrease of electrochemical performance further indicated the influence of specific surface area. Specific surface area, particle size, and theoretical specific capacity all affect the electrochemical lithium storage performance. Mixing o-LiMnO₂ also improved the electrochemical performance of LiMn₂O₄ in the present work. Therefore, forming a mixed phase of LiMn₂O₄ and o-LiMnO₂ would improve the electrochemical performance of single o-LiMnO₂ and LiMn₂O₄, which was another highlight of this work.

Fig. 7 Cyclic voltammograms of M₀ and M_m in 1.0 mol L⁻¹ LiClO₄ +EC + DMC + EMC with equal volume within potential range of 2.0–4.3 V (vs. Li/Li⁺) with a scan rate of 0.1 mV s⁻¹.

The influence of cell parameters on the electrochemical performance of LiMnO₂ was further analyzed by discharge–charge profiles at a current density of 100 mA g⁻¹. A remarkable discharge voltage plateau at 3.3–2.4 V was observed for pure-phased M₀ (Fig. 8). Plateau at 3.3–2.4 V was owing to the discharge of orthorhombic LiMnO₂ or tetragonal Li₂Mn₂O₄.^3,4 A new voltage plateau at 4.3–3.6 V was formed and kept stable after 20 cycles, likely due to the formation of spinel LiMn₂O₄, which was further transformed into tetragonal Li₂Mn₂O₄ (Fig. 5), and Li_xMn₂O₄ (0 < x < 2) was also could be formed.^4,7,20,35,41 The discharge capacity increased with an increase in the cyclic number during the charge–discharge test, and kept stable after 20 cycles likely due to the activation process.³ After 100 cycles, the discharge voltage at 4.3–3.6 V was remarkably lowered for M₀ and M_m possibly due to the disproportionation of Li₂Mn₂O₄ to Mn²⁺ and MnO₂, resulting in capacity fade.^35,41,43 However, there is no obvious change for the discharge plateau of M₆ (Fig. 8c), which was attributed to that aluminum doping affected the crystal lattice and electrochemical performance of LiMnO₂.⁵ The Jahn–Teller distortion of the [Mn³⁺O₆] octahedra induces an elongation along the a direction.⁵ However, parameter a decreased with an increase in the content of doped Al in LiMnO₂ (Table 1). The distorted axial Mn³⁺–O bond length of 0.229 nm is greater than the typical Al³⁺–O bond length (0.19 nm) in octahedral coordination,⁵ and the most significant changes due to substitution were expected to occur along a axis. These changes inhibited the shrinkage distortion during charge–discharge process, and excellent cyclic stability was achieved.^4,5,44 In this work, appropriate content of doped aluminum with Al/Mn molar percent ratio about 0.58 exhibited the best electrochemical performance. It should be pointed out that the appropriate decrease in cell volume with compression along a axis and stretch along b axis facilitated the cyclic stability of o-LiMnO₂ by Al doping due to the shrinkage of cell volume for the formation of LiMn₂O₄ during the charge–discharge process.³¹ However, an increase in cell volume facilitated the electrochemical stability of LiMn₂O₄ owing to the expansion of cell volume for the formation of Li₂Mn₂O₄ during the charge–discharge process.^44,45 Aluminum doping decreases Jahn–Teller distortion and improves the cyclic stability due to that the phase transformation from o-LiMnO₂ to LiMn₂O₄ occurs in the charge–discharge test. In this work, although LiMn₂O₄ was formed as an intermediate product during the charge–discharge test, the discharge capacity of o-LiMnO₂ was about 160 mA h g⁻¹, which was much higher than that of LiMn₂O₄.^4,5,7–9

Fig. 8 Discharge–recharge profiles of M₀ (a), M_m (b), and M₆ (c) for different cycles at a current density of 100 mA g⁻¹.

During the charge–discharge processes, lithium ions migrate out of the lattice along c axis, and oxygen position does not change with Li⁺ extraction, resulting in the formation of vacant octahedral sites.^43,46 Subsequently, half of the manganese ions migrate statistically into the vacant octahedral sites along b axis, and the migration of metal ions is accompanied by a slight shift for oxygen atoms to the ideal cubic closest packing positions with cubic symmetry.^9,46,47 From the results of FTIR (Fig. 3) and Rietveld refinement (Table 1), it could be concluded that the bond energy of Mn–O towards the equatorial plane of MnO₆ octahedron was lowest in M₆, which facilitated the migration of manganese ions into the vacant octahedral sites.^46,47 The transformation of o-LiMnO₂ to spinel LiMn₂O₄ easily occurred in the initial stage of charge–discharge test (Fig. 5a). To a certain degree, electrochemical reaction impedance in the initial stage positively correlated with the difficulty of phase change. In this work, M₆ had the lowest reaction impedance before charge–discharge test (Fig. 6, Table 2). Phase transformation easily occurred in the electrochemical process, and exhibited stable electrochemical lithium storage performance after activation of a few cycles.

As for the charge processes of M₀, pure-phased LiMnO₂, a short charging voltage plateau at about 3.15 V was observed at the first cycle. The similar results were also reported in the literature.^5,7,14 The presence of the short charge plateau was possibly due to the activation of o-LiMnO₂ and the formation of LiMn₂O₄ in this process. As seen from the multi-cycle charge curves of M₀ and M₆, voltage plateau at about 3.15 V was extended due to the increase of the content of LiMn₂O₄ formed in charge–discharge test.⁹ As for M_m, this charge plateau was more obvious after charge–discharge activation of five cycles for the high content of LiMn₂O₄ in cathode materials (Fig. 8b). These results were consistent with the cyclic voltammograms as shown in Fig. 7. The formation of LiMn₂O₄ could be further confirmed by the XRD analysis of electrode materials as illustrated in Fig. 5 after charge–discharge test of the first cycle. To further improve the electrochemical cyclic stability of o-LiMnO₂, the phase transition from layered to spinel structure might be remarkably retarded by doping other metal ions, such as Ni.¹⁹ This hydrothermal method also provides the possibility for the preparation of other metal doped o-LiMnO₂.

Decreasing reaction impedance (R_ct) may effectively improve the electrochemical performance of o-LiMnO₂ and LiMn₂O₄ cathode materials. As seen in Fig. 6, the R_ct values of the as-obtained samples were greater than 200 Ω. In order to improve the electrochemical performance, carbon materials, such as carbon nanotube and graphene, modified o-LiMnO₂ might be directly synthesized using this hydrothermal system, and this study is currently underway.

5. Conclusions

Pure-phased, LiMn₂O₄-mixed, and aluminum-doped o-LiMnO₂ cathode materials of rechargeable lithium batteries with high discharge capacity and excellent cyclic stability have been successfully fabricated using Mn(II) salt by a facile one-step hydrothermal treatment at 180 °C for 24 h. The molar ratio of Li⁺/Mn²⁺ can be controlled as low as 6, and the complexation reaction of EDTA and Mn²⁺ facilitates the formation of o-LiMnO₂. Al/Mn molar percent ratio in aluminum-doped LiMnO₂ is in the range of 0 to 1.22 when AlCl₃/MnCl₂ molar ratio is controlled from 0 to 20% in reaction system. LiOH and NaClO concentration should be slightly increased to obtain single-phased aluminum-doped o-LiMnO₂ when Al/Mn molar ratio increased in hydrothermal system. Mixing LiMn₂O₄ and doping aluminum improve the electrochemical lithium storage capacity and cyclic stability of o-LiMnO₂. Aluminum doped LiMnO₂ with Al/Mn molar percent ratio of 0.58 shows the best electrochemical performance. Pure-phased o-LiMnO₂, a mixture of o-LiMnO₂ and LiMn₂O₄ (LiMnO₂/LiMn₂O₄ mass ratio = 2.45 : 1), and aluminum doped LiMnO₂ (Al/Mn molar ratio = 0.58%) exhibits the initial discharge capacity of 76.0, 139.1 and 81.8 mA h g⁻¹, and cycling capacity of 124.1, 143.7 and 155.8 mA h g⁻¹ after 100 cycles, respectively. This preparation method and the study on the electrochemical properties widen the potential application of pure-phased and doped o-LiMnO₂ in the field of rechargeable lithium-ion battery.

Acknowledgements

The authors thank the National Natural Science Foundation of China (Grant nos: 41171375 and 20807019), the Program for New Century Excellent Talents in University of China (no. NCET-12-0862), the Young Outstanding Talent Foundation of Hubei Province of China (no. 2012FFA031), and the Fundamental Research Funds for the Central Universities (Program no.: 2013PY029, 2013PY030, 2011PY015, 2009SC007, 2012SC31 and 2013QC016) for financial support. The authors also gratefully acknowledge Dr Wenfeng Tan at Huazhong Agricultural University and Dr Xiaoqing Qiu at Fuzhou University for useful suggestions.

References

1. M. Armand and J.-M. Tarascon, Nature, 2008, 451, 652–657.
2. H. Nishide and K. Oyaizu, Science, 2008, 319, 737–738.
3. X. L. Xiao, L. Wang, D. S. Wang, X. M. He, Q. Peng and Y. D. Li, Nano Res., 2009, 2, 923–930.
4. C. Liu, J. M. Nan, X. X. Zuo, X. Xiao and D. Shu, Int. J. Electrochem. Sci., 2012, 7, 7152–7164.
5. J. B. Cook, C. Kim, L. P. Xu and J. Cabana, J. Electrochem. Soc., 2013, 160, A46–A52.
6. J. B. Goodenough and Y. Kim, Chem. Mater., 2010, 2, 587–603.
7. P. G. Bruce, A. R. Armstrong and R. L. Gitzendanner, J. Mater. Chem., 1999, 9, 193–198.
8. J. Cho, Y. J. Kim, T.-J. Kim and B. Park, Chem. Mater., 2001, 13, 18–20.
9. Y.-I. Jang, B. Y. Huang, H. F. Wang, D. R. Sadoway and Y.-M. Chiang, J. Electrochem. Soc., 1999, 146, 3217–3223.
10. H. Xu, J. Sun and L. Gao, Ionics, 2013, 19, 63–69.
11. Z. P. Guo, S. Zhong, G. X. Wang, H. K. Liu and S. X. Dou, J. Alloys Compd., 2003, 348, 231–235.
12. G. Ceder and S. K. Mishra, Electrochem. Solid-State Lett., 1999, 2, 550–552.
13. Z. P. Guo, K. Konstantinov, G. X. Wang, H. K. Liu and S. X. Dou, J. Power Sources, 2003, 119–121, 221–225.
14. H. M. Ji, G. Yang, X. W. Miao and A. Q. Hong, Electrochim. Acta, 2010, 55, 3392–3397.
15. J. L. Xie, X. Huang, Z. B. Zhu and J. H. Dai, Ceram. Int., 2011, 37, 419–421.
16. R. I. Walton, Chem. Soc. Rev., 2002, 31, 230–238.
17. Y. He, Q. Feng, S. Q. Zhang, Q. L. Zou, X. L. Wu and X. J. Yang, ACS Sustainable Chem. Eng., 2013, 1, 570–573.
18. Z. P. Guo, G. X. Wang, H. K. Liu and S. X. Dou, Solid State Ionics, 2002, 148, 359–366.
19. K. S. Park, M. H. Cho, S. H. Park, K. S. Nahm, Y. K. Sun, Y. S. Lee and M. Yoshio, Electrochim. Acta, 2002, 47, 2937–2942.
20. Z. Su, Z. W. Lu, X. P. Gao, P. W. Shen, X. J. Liu and J. Q. Wang, J. Power Sources, 2009, 189, 411–415.
21. X. K. Huang, Q. S. Zhang, H. T. Chang, J. L. Gan, H. J. Yue and Y. Yang, J. Electrochem. Soc., 2009, 156, A162–A168.
22. G. H. Qiu, J. Wang, W. Zhao, X. H. Feng, W. F. Tan, Q. Zhang and F. Liu, Chin. J. Chem., 2010, 28, 2301–2307.
23. H. Yin, F. Liu, X. H. Chen, X. H. Feng, W. F. Tan and G. H. Qiu, Microporous Mesoporous Mater., 2012, 153, 115–123.
24. Q. Liu, Y. X. Li, Z. L. Hu, D. L. Mao, C. K. Chang and F. Q. Huang, Electrochim. Acta, 2008, 53, 7298–7302.
25. R. Chitrakar, K. Sakane, A. Umeno, S. Kasaishi, N. Takagi and K. Ooi, J. Solid State Chem., 2002, 169, 66–74.
26. Y. M. Hon, S. P. Lin, K. Z. Fung and M. H. Hon, J. Eur. Ceram. Soc., 2002, 22, 653–660.
27. L. A. Riley, S. V. Atta, A. S. Cavanagh, Y. F. Yan, S. M. George, P. Liu, A. C. Dillon and S.-H. Lee, J. Power Sources, 2011, 196, 3317–3324.
28. S. Chitra, P. Kalyani, T. Mohan, R. Gangadharan, B. Yebka, S. Castro-Garcia, M. Massot, C. Julien and M. Eddrief, J. Electroceram., 1999, 3, 433–441.
29. C. Wu, Z. X. Wang, F. Wu, L. Q. Chen and X. J. Huang, Solid State Ionics, 2001, 144, 277–285.
30. D. Ostrovskii, F. Ronci, B. Scrosati and P. Jacobsson, J. Power Sources, 2001, 94, 183–188.
31. W.-H. Ryu, J.-Y. Eom, R.-Z. Yin, D.-W. Han, W.-K. Kim and H.-S. Kwon, J. Mater. Chem., 2011, 21, 15337–15342.
32. G. Yang, H. M. Ji, P. Gao, A. Q. Hong, H. R. Ding, S. Roy, J. Pinto and X. F. Jiang, J. Electrochem. Soc., 2011, 158, A1071–A1078.
33. Y.-J. Gu, Y.-B. Chen, H.-K. Wu, Y.-M. Wang, L. Chen, M. Wang, X.-W. Huang, H. Zheng, X.-B. Liu, Ch. Lu and Y. Hu, Energy Materials, 2009, 4, 40–43.
34. C.-H. Lu and H.-C. Wang, J. Eur. Ceram. Soc., 2004, 24, 717–726.
35. J. Cho, T.-J. Kim and B. Park, J. Electrochem. Soc., 2002, 149, A288–A292.
36. L. L. Xiong, Y. L. Xu, C. Zheng, Z. W. Zhang and J. B. Li, J. Solid State Electrochem., 2011, 15, 1263–1269.
37. H. J. Cui, G. H. Qiu, X. H. Feng, W. F. Tan and F. Liu, Clays Clay Miner., 2009, 57, 715–724.
38. S.-T. Myung, S. Komaba and N. Kumagai, Electrochim. Acta, 2002, 47, 3287–3295.
39. G. H. Qiu, H. Huang, S. Dharmarathna, E. Benbow, L. Stafford and S. L. Suib, Chem. Mater., 2011, 23, 3892–3901.
40. Q. H. Qiu, Q. Li, Y. Yu, X. H. Feng, W. F. Tan and F. Liu, J. Soils Sediments, 2011, 11, 1323–1333.
41. S. B. Tang, M. O. Lai and L. Lu, J. Power Sources, 2007, 164, 372–378.
42. Y. C. Chen, K. Xie, Y. Pan and C. M. Zheng, J. Power Sources, 2011, 196, 6493–6497.
43. J. Cho and M. M. Thackeray, J. Electrochem. Soc., 1999, 146, 3577–3581.
44. Y. Shin and A. Manthiram, Chem. Mater., 2003, 15, 2954–2961.
45. Y. Shin and A. Manthiram, J. Electrochem. Soc., 2004, 151, A204–A208.
46. W. P. Tang, H. Kanoh and K. Ooi, J. Solid State Chem., 1999, 142, 19–28.
47. J. Akimoto, Y. Takahashi, Y. Gotoh, K. Kawaguchi, K. Dokko and I. Uchida, Chem. Mater., 2003, 15, 2984–2990.

---

Footnote

† Electronic supplementary information (ESI) available. This includes the effect of EDTA on product composition, the preparation of Al doped o-LiMnO₂, the micromorphology, particle size distributions of the as-obtained pure and Al doped o-LiMnO₂, the heat treatment of as-obtained o-LiMnO₂ in air, and the electrochemical performance of the as-obtained o-LiMnO₂ and LiMn₂O₄. See DOI: 10.1039/c3ra47932c

---