Synthesis and electrochemical properties of monoclinic fluorine-doped lithium manganese oxide (Li_xMnO_2−yF_y) for lithium secondary batteries

Abstract

A series of monoclinic fluorine-doped lithium manganese oxide (Li_xMnO_2−yF_y) were prepared by the ion exchange of sodium for lithium in Na_xMnO_2−yF_y precursors that were obtained using a high-temperature solid-state reaction. The microstructure and composition of the samples were characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM), inductively coupled plasma atomic emission spectroscopy (ICP-AES), ion chromatography (IC), Fourier transform infrared spectroscopy (FTIR), and X-ray photoelectron spectroscopy (XPS). The different valence states of manganese in material Li_xMnO_2−yF_y were determined by redox titration method. The electrochemical performance of these samples as cathode materials were studied by the galvanostatic and cyclic voltammetry method. These materials had a high crystallite size, which was composed of 5–8 μm grains. The Li_0.86MnO_1.98F_0.02 materials delivered an initial discharge capacity of 129.2 mA h g⁻¹ and gradually increased to a maximum discharge capacity of 210 mA h g⁻¹ at a current density of 50 mA g⁻¹ after 50 cycles. Moreover, the material showed an excellent cycling behavior, even though its original structure transformed into the spinel phase during cycling. The results show that the partial substitution of monoclinic LiMnO₂ with fluorine can improve the cycle stability and high-rate discharge capability of cathode materials.

---

1. Introduction

Lithium manganese oxide (LiMnO₂) compounds have received attention as intercalation cathodes for rechargeable lithium ion batteries due to their high theoretical capacity, low cost, and nontoxicity, and there has been significant interest in monoclinic LiMnO₂ (space group C2/m, hereafter denoted as m-LiMnO₂) as a potential cathode material for replacing LiCoO₂, which is the currently used cathode material in commercial Li-ion rechargeable batteries.^1–3 However, m-LiMnO₂ is generally considered to be unstable compared to orthorhombic LiMnO₂ (space group Pmnm, hereafter denoted as o-LiMnO₂).⁴ One significant drawback is the difficult synthesis of layered m-LiMnO₂ using the conventional high-temperature solid-state reaction method because these methods always produce o-LiMnO₂.^5,6 Consequently, m-LiMnO₂ was first obtained by metastable synthesis routes such as ion exchange and hydrothermal reactions.^7,8

In this compound, the O²⁻ ions are arranged in a cubic close-packed structure, and the octahedral interstices are occupied by Li⁺ and Mn³⁺ with layered cation ordering. The neighboring MnO₆ octahedra share a common edge. Due to the presence of high-spin Mn³⁺ on the octahedral sites, the local site symmetry around Mn³⁺ is distorted from a regular octahedron, and the overall crystal structure is monoclinic due to the cooperative Jahn–Teller distortion. Therefore, it is difficult to inhibit the irreversible structural transformation from layered LiMnO₂ to a spinel LiMnO₂ during electrochemical cycling. According to the first-principles calculations, the Jahn–Teller effect should be absent in the low-spin state, and the layered structure could be stabilized in m-LiMnO₂.⁹

Cation doping protects the layered monoclinic structure against Jahn–Teller distortion¹⁰ and has superior Li insertion/extraction cycling properties for m-LiMnO₂. Many studies have focused on stabilization of the layered material structure, particularly for m-LiMnO₂, by partial substitution of Mn by Cr, Al, Co, Ni and Mg.^11–18 However, for m-LiMnO₂ to be suitable for practical applications such as high-energy and power density applications, further stabilization of the monoclinic structure is needed. Small amounts of additional dopants such as fluorine¹⁹ into the Li(Ni_1/3Co_1/3Mn_1/3)O₂ lattice have been reported to improve the structural stability and enhance the cycling performance. Zheng et al. prepared the Li[Li_0.2Mn_0.54Ni_0.13Co_0.13]O_2−xF_x (x = 0, 0.05 and 0.10) composite materials with different content of fluorine by sol–gel method using NH₄F as fluorine source. XRD and XPS analysis shows that the fluorine ions successfully substitute for oxygen sites. The F-doped Li[Li_0.2Mn_0.54Ni_0.13Co_0.13]O_2−xF_x materials show superior cycling performance.²⁰ In fact, Jouanneau et al. discussed the influence of LiF additions on Li(Ni_xCo_1−2xMn_x)O₂ and found that materials could be obtained using LiF as a sintering agent with almost no disturbance of the cell performance.^21–23

In this work, we attempt to prepare a partial substitution of O²⁻ by F⁻ layered monoclinic LiMnO₂ using a high-temperature solid-state reaction method with a subsequent ion-exchange reaction. The morphology, microstructure, composition, and electrochemical properties were investigated in detail to assess the suitability of these samples as cathode materials.

2. Experimental

2.1 Preparation and characterization

First, Mn(CH₃COO)₂·4H₂O was heated with the ramping rate of 2 °C min⁻¹ and decomposed at 500 °C for 15 h in air to produce Mn₂O₃. Samples of monoclinic α-NaMnO₂ were prepared using a solid-state reaction of the stoichiometric mixture of Mn₂O₃ (0.01 mol, 1.58 g) and Na₂CO₃ (Na/Mn = 1.1, 0.011 mol, 1.16 g) in a closed crucible at 750 °C for 24 h in air. The mixture was quenched to room temperature and ground, and the process was repeated. The resulting samples were stored in a desiccator. In the subsequent ion-exchange reaction, 1 g of α-NaMnO₂ was mixed with 10 g LiCl in 40 ml n-hexanol and autoclaved at 150 °C for 12 h in a 50 ml Teflon-lined stainless steel vessel. The obtained m-LiMnO₂ products were washed with de-ionized water and then dried in an oven at 105 °C for 12 h. Fluorine-substituted materials were prepared with the molar ratio of F/Mn of 0.01/1, 0.03/1, and 0.05/1 with the homogeneous mixture of NH₄F of 0.001 mol (0.037 g); 0.003 mol (0.0111 g); and 0.005 mol (0.0185 g), respectively. The final fluorine-substituted materials were obtained by the above-mentioned method.

The microstructure of the as-prepared samples was characterized by X-ray diffraction (XRD, Rigaku D/max-2500) with CuKα radiation and scanning electron microscopy (SEM, HITACHI S-3500N). Elemental analysis for Na, Li, and Mn were performed using inductively coupled plasma atomic emission spectroscopy (ICP-AES, IRIS Advantage) and the content of fluorine by ion chromatography (ICS-2100) to determine the chemical composition of these materials. The Fourier transform infrared spectra (FTIR) of these samples were recorded from 400 to 4000 cm⁻¹ at room temperature using the KBr wafer technique in a Bio-Rad FTS 6000 FTIR instrument with a resolution of 2 cm⁻¹ in transmittance mode. The surface chemical state was recorded using X-ray photoelectron spectroscopy (XPS, PHI-5300 ESCA).

2.2 Determination of material composition

The sample was treated by HCl, HNO₃ and HF at room temperature, H₃BO₃ was added to complex with excessive fluoride ions. Working conditions of the instrument was optimized, appropriate analysis line of various elements was selected and the internal standard method was used to eliminate disturbance with yttrium selected as internal standard element, the content of lithium and manganese were determined by inductively coupled plasma emission spectroscopy (ICP-AES). The sample was decomposed by H₂SO₄. After water steam distillation at 160–180 °C, the fluoride in testing solution would be released and absorbed by NaOH solution realizing the separation from matrix and other elements. Then, the content of fluoride was determined by ion chromatography (IC).

2.3 Analysis of the valence states of manganese

The total amount of manganese was determined by chemical conversion to permanganic radical using acidified ammonium persulfate, the determination of Mn⁴⁺ by the acetylacetone-iodometric method being carried out in the presence of Mn³⁺ without its interference, and the amount of the tetravalent and trivalent manganese being determined by dissolving the lithium manganese oxides in a known excess of ammonium iron sulfate in 1 mol L⁻¹ sulfuric acid and back-titrating the unreacted Fe²⁺ with standardized potassium dichromate. Based on the results obtained, the amount of bivalent manganese and average manganic valence could be calculated.²⁴

2.4 Electrochemical performance

The working electrode was prepared by compressing a mixture of the active materials with acetylene black and a binder (polytetrafluoroethylene, PTFE) in a weight ratio of 75 : 20 : 5. Lithium metal was used for the counter and reference electrodes. The electrolyte was LiPF₆ (1 M) in a mixture of ethylene carbonate (EC), dimethyl carbonate (DMC), and ethyl methyl carbonate (EMC) in a weight ratio of 1 : 1 : 1. All procedures for handing and fabricating the electrochemical cells were performed in an argon-filled glove box. The galvanostatic method was used to measure the electrochemical capacity and cycle life of the electrodes under the charge–discharge current density of 50 mA g⁻¹ at room temperature using a LAND CT2001A instrument. The cut-off potentials for the charge and discharge were set between 2.0–4.3 V (vs. Li⁺/Li). Cyclic voltammetry (CV) was performed with a LK2500 electrochemical workstation, scanning from 2 to 4.3 V with a scan rate of 0.1 mV s⁻¹. The electrochemical impedance spectra (EIS) were measured using a Zahner IM6ex electrochemical workstation over the frequency ranges from 10 kHz to 10 MHz with a 5 mV AC input signal applied between the working and reference electrodes.

3. Results and discussion

3.1 Composition and structure

The composition of the as-prepared samples, which were calculated from the results of ICP-AES, and IC determination and valence state of Mn analysis, are summarized in Table 1. For all samples, most of the Na⁺ cation can be exchanged and lithium concentration can be controlled at about 0.90 in unit. All of the XRD patterns were refined by the Rietveld analysis, an example of the Rietveld refinement is shown in Fig. 1 for the composition of Li_xMnO_2−yF_y. All diffraction peaks, except for the market peaks, could be attributed to the small amount of monoclinic Na_0.7MnO_2.05 because NaMnO₂ was not stable and was very hygroscopic in air; NaMnO₂ was easily transformed to the more stable phase of Na_0.7MnO_2.05.²⁵

Table 1 Chemical composition of the m-LiMnO₂ and F-substituted m-LiMnO₂

Sample	Design formula	Precursor powders	Valence of manganese	Chemical composition
a	NaMnO2	Na0.96MnO2	3.04	Na0.96MnO2
b	LiMnO2	Na0.91MnO2	3.09	Na0.06Li0.85MnO2
c	LiMnO1.99F0.01	Na0.90MnO1.98F0.02	3.08	Na0.05Li0.86MnO1.98F0.02
d	LiMnO1.97F0.03	Na0.91MnO1.96F0.04	3.05	Na0.06Li0.88MnO1.96F0.04
e	LiMnO1.95F0.05	Na0.92MnO1.93F0.07	3.01	Na0.06Li0.87MnO1.93F0.07

---

Fig. 1 The XRD patterns of the as-prepared samples: (a) Na_0.96MnO₂, (b) Na_0.06Li_0.85MnO₂, (c) Na_0.05Li_0.86MnO_1.98F_0.02, (d) Na_0.06Li_0.88MnO_1.96F_0.04, and (e) Na_0.06Li_0.87MnO_1.93F_0.07.

The Table 2 list the comparison between the observed values and the JCPDS-87155 standard card data. The characteristics of the synthetic product crystal plane diffraction spacing and diffraction angle all can correspond well, the as-prepared samples could be assigned to the monoclinic structure with space group of C₂/m, which differs from o-LiMnO₂.^7–9 At the same time the characteristic peak of α-NaMnO₂ are not appear noticeably in XRD, this shows that ion exchange basic completely. Moreover, we also find that the observed d_hkl values is greater than the standard and the 2θ values is opposite. This may be due to the material contains no exchange of sodium (sodium ions with greater radius than lithium ion).

Table 2 Comparison between the observed and JCPDS-871255 values

h k l	dhkl (LiMnO2)/nm		2θ/(°)
	Observed	JCPDS	Observed	JCPDS
0 0 1	0.4849	0.4842	18.28	18.31
−2 0 1	0.2712	0.2711	33.02	33.01
1 1 0	0.2440	0.2435	36.80	36.89
−1 1 1	0.2397	0.2395	37.48	37.50
−2 0 2	0.2298	0.2295	39.12	39.21
1 1 1	0.2012	0.2006	45.02	45.15
−1 1 2	0.1942	0.1942	46.46	46.74
2 0 1	0.1878	0.1875	48.45	48.49
0 0 3	0.1626	0.1614	56.52	57.01
1 1 2	0.1560	0.1555	59.16	59.37
−1 1 3	0.1511	0.1505	61.26	61.56
−3 1 2	0.1493	0.1487	62.12	62.36
0 2 0	0.1407	0.1404	66.34	66.53
0 2 1	0.1350	0.1348	69.54	69.66

---

Lattice parameters of all samples obtained from the Rietveld refinement are listed in Table 3. With increasing fluorine content, cell volume increased from 0.0824 nm³ to 0.0839 nm³, the increase of cell volume is attributed to the replacement of O²⁻ (r = 0.132 nm) with greater F⁻ (r = 0.136 nm), which are in agreement with literature values.²⁶ From the table can be found that the increase of the content of fluorine can decrease the valence state of Mn in material. The results attributable to the low valence state of anion (F⁻) supersedes the high valence state O²⁻, and it also demonstrated that the fluorine ions successfully doped into the lattice structure.

Table 3 Lattice parameters of the samples after Rietveld refinement

Sample	a (nm)	b (nm)	c (nm)	Cell volume (nm^3)
Na0.96MnO2	0.582	0.287	0.576	0.0962
Na0.06Li0.85MnO2	0.544	0.281	0.539	0.0824
Na0.05Li0.86MnO1.98F0.02	0.543	0.282	0.541	0.0828
Na0.06Li0.88MnO1.96F0.04	0.543	0.282	0.543	0.0831
Na0.06Li0.87MnO1.93F0.07	0.544	0.283	0.549	0.0839

---

The FTIR spectra of these as-prepared samples are indicated in Fig. 2. The bands in the region of 3000–3500 cm⁻¹ can be assigned to OH stretching vibrations. Moreover, the bending mode vibration of water is observed at 1600–1800 cm⁻¹. The weak absorptions at 2832 and 2886 cm⁻¹ are due to C–H stretching models, and the bands at around 1475 and 2465 cm⁻¹ result from the absorption of the C=O group.²⁷ Two obvious absorption bands are observed in the infrared (IR) spectra of these materials at around 875 and 993 cm⁻¹ (except for Na_0.96MnO₂ samples), which could be attributed to the metal hydroxide vibration band.²⁸ Furthermore, the IR bands in the region of 450–750 cm⁻¹ could be assigned to Mn–O stretching vibrations in lithium-manganese-oxide compounds.²⁹

Fig. 2 The IR spectra of (a) Na_0.97MnO₂, (b) Na_0.07Li_0.85MnO₂, and (c) Na_0.06Li_0.86MnO_1.98F_0.02.

For the Na_0.96MnO₂ and Na_0.06Li_0.85MnO₂ samples, one featured band at 530 cm⁻¹ is observed. However, for the F-doped sample of Na_0.05Li_0.86MnO_1.98F_0.02, an obvious blue shift is observed in the Mn–O stretching band from 532 to 550 cm⁻¹. The electronegativity of F (4.0), which is higher than that of O (3.5), can strengthen the Mn–O stretching bands, whereas the thermochemical radius of F⁻ (0.136 nm) is greater than the O²⁻ ionic radius (0.132 nm), leading to the blue shift.³⁰ Certainly, with the fluoride concentration increase, the blue shift will become more obviously which causes the oxygen defects concentration increase in the material. Due to the deficiency of oxygen ion in the electrode material caused by oxygen defects, there are positive charged oxygen vacancies in the structure; and then during the immersion of electrolytic solution to electrode surface, more oxygen defects produce thicker organic electrolytic liquid film, which arise the organic electrolytic liquid film resistance. This result implies that oxygen defects will reduce the electrochemical capacity and cycling performance of the electrode material; and this is in agreement with the electrochemical performance test results shown below.³¹

Fig. 3 shows the SEM images of the as-prepared samples. The layered Na_0.97MnO₂ and monoclinic Li_xMnO_2−yF_y samples have a similar bar-shaped morphology with a grain diameter of 5–8 μm and with small particles of about 2–3 μm in size. However, a topotactic reaction can be observed in a morphology changes from the SEM. Small particles can be found in Fig. 3. The things is especially obvious with the increase of fluorine content. The results should be attributed to the primary particles become well-developed when the amount of fluorine increases. Changes of morphology reveal that incorporation of fluorine alters the surface energy and catalyzes the growth of the primary particles, consistent with that reported for other cathode materials with fluorine doping.^32,33

Fig. 3 The SEM images of the as-prepared samples: (a) Na_0.96MnO₂, (b) Na_0.06Li_0.85MnO₂, (c) Na_0.05Li_0.86MnO_1.98F_0.02, (d) Na_0.06Li_0.88MnO_1.96F_0.04, and (e) Na_0.06Li_0.87MnO_1.93F_0.07.

3.2 Electrochemical properties

Fig. 4 exhibits the initial charge/discharge curves of Li_xMnO_2−yF_y (y = 0, 0.02, 0.04, 0.07) at a current density of 50 mA g⁻¹ (0.2 C) between 2.0 and 4.3 V. An initial discharge capacity of 133.2 mA h g⁻¹ is obtained for the Na_0.06Li_0.85MnO₂ sample, which is obviously higher than that of o-LiMnO₂ obtained using the conventional high-temperature solid reaction. With the increase of fluorine content, the discharge capacities of the initial discharge capacity is decreased to 129.2 mA h g⁻¹ at 0.2C for Na_0.05Li_0.86MnO_1.98F_0.02, the initial discharge capacities of Na_0.06Li_0.88MnO_1.96F_0.04 and Na_0.06Li_0.87MnO_1.93F_0.07 are 120.6 and 112.3 mA h g⁻¹, respectively. This is due to the existence of the inactive substances fluorine and the atomic weight of fluorine (19 g mol⁻¹) slightly larger than oxygen atomic weight (16 g mol⁻¹). Moreover, fluorine substituted materials exhibit higher charging voltage over the potential range and the discharge plateau is lower than the pristine one, which is an indication of the higher polarization effect. This similar phenomena, which is ascribed to the stronger bonding strength of Li–F (577 kJ mol⁻¹) bond than the Li–O bond (341 kJ mol⁻¹), are also observed and explained by previous researchers.^34,35

Fig. 4 The initial charge–discharge voltage profile (a), differential capacity plots (b) of Li_xMnO_2−yF_y (y = 0, 0.02, 0.04, 0.07) when cycled between 2.0 and 4.3 V at a current density of 50 mA g⁻¹.

The rate capabilities of the materials are operated at the same current densities between 2.0 and 4.3 V. Fig. 5 compares the discharge capacities of the electrodes at various C rates. The results of the electrochemical cycles indicated that the discharge capacity of monoclinic LiMnO₂ was small at the initial cycles, gradually increased with increasing cycle number, and the capacity finally stabilized. This result was consistent with the previously reported results of orthorhombic LiMnO₂ electrochemical cycle behavior. This phenomenon reveals that monoclinic LiMnO₂ cathode materials undergo an activation process during the electrochemical cycle, during which there is a phase transition from m-LiMnO₂ to cubic spinel Li_xMn₂O₄.³⁶ In comparison, the cycle stability of the Na_0.06Li_0.85MnO₂ sample is not so good with a low capacity retention of 80.8 mA h g⁻¹ after 50 cycles. However, the fluorine substituted materials show better rate capability at high charge and discharge rates. The capacity retention of these materials is significantly improved by fluorine substitution, the good rate capability of fluorine substituted materials could be attributed to the existence of the strong bonding by fluorine which stabilizes the host structure and increases the stability of cathode materials.³⁷ The results also show that the electrochemical performances of the materials were severely damaged with the fluorine content increased, this is related to the amount of oxygen defects in materials. Therefore, the fluorine content should be controlled in order to prevent more oxygen defects in synthetic materials. In other words, the partial fluorine substitution in the layered monoclinic structure can significantly depress this structural transformation and slow the rate of capacity decay.

Fig. 5 Cycle performance of the as-prepared samples at different discharge current densities (charging at 50 mA g⁻¹).

We also notice that the Na⁺ cation can be reserved at about 0.06 unit in all samples. The Na⁺ ionic radius (0.097 nm) is larger than the Li⁺ ionic radius (0.060 nm). Thus, in monoclinic Na-substituted Li_xMnO₂, the small change of the interlayer spacing to accommodate Na allows high rechargeability and good cycle life. This result could be explained by the extraction of a limited amount of sodium ions from the interlayer spacing while maintaining the layered structure upon cycling.³⁸

Initial investigation of Li_xMnO_2−yF_y were carried out using cyclic voltammetry, Fig. 6. A rate of 0.01 mV s⁻¹ between potential limits of 2 and 4.3 V was employed. Cycling commenced with oxidation from the open circuit potential of 2 V. Fig. 6(a) and (b) shows typical cyclic voltammograms of the Na_0.06Li_0.85MnO₂ and Na_0.05Li_0.86MnO_1.98F_0.02 samples subjected to 1, 2, 5, 10, 20 and 30 cycles and 0.1 mV s⁻¹ over the potential range between 2.0 and 4.3 V (vs. Li⁺/Li). The striking feature is the difference between the first and subsequent cycles. In fact the difference is to be found in the original CV for Na_0.06Li_0.85MnO₂ shows two anodic peaks at 3.25 V and 4.18 V as well as the corresponding cathodic peaks at 2.83 V and 3.85 V. On repeated cycling, the peaks at 4.18 V and 3.85 V split into two peaks, which becomes more obvious after 30 cycles. This peak splitting suggests a gradual transformation of layered LiMnO₂ into the spinel phase.³⁴ From Fig. 6(a), we can observe the anodic phase shows four oxidation peaks centered at 3.1 V, 4.0 V, 4.18 V, and 3.8 V after 2 cycles, but the cathodic phase shows only three reduction peaks with a major peak centered at 2.8 V and two minor peaks at 3.9 V and 4.1 V. The peak at 3.8 V does not have reduced peak, which corresponds to an irreversible phase change. Similar irreversible phase changes were observed by Paulsen et al. in O₂-type Li_2/3[Li_1/6Mn_5/6]O₂ and Li_2/3[Co_1/18Mn_17/18]O₂.³⁹ The reason for this irreversible peak/plateau is not presently known; however, because it is only observed in stacking faulted compounds, the irreversible peak could be correlated to crystallographic modification, oxygen removal from the lattice and/or removal of the stacking faults. For the Na_0.05Li_0.86MnO_1.98F_0.02 sample, the similar main cathodic and anodic peaks with Na_0.06Li_0.85MnO₂ sample in the first cycle are found in Fig. 6(b). From the CV curves of Fig. 6(b), Li⁺ can mostly deintercalate from the structure of sample Na_0.05Li_0.86MnO_1.98F_0.02 in the first anode process, due to the broader oxidizing peak about 3.8 V. But the reducing peak about 2.9 V is weaker in the first cathode process. This result indicates there is obvious capacity loss between the initial charge and discharge capacities, and the efficiency of reversible intercalate/deintercalate of Li⁺ is relatively lower, which is in accordance with Fig. 4. As the cycle goes by, the intensity of oxidizing peak about 3 V gradually becomes stronger, and the intensities of reducing peak about 2.8 V and 4 V gradually become stronger. However, the anodic or cathodic current densities change slightly in the following cycles, evidencing the higher electrochemical stability of the Na_0.05Li_0.86MnO_1.98F_0.02 sample as compared to the Na_0.06Li_0.85MnO₂ sample.

Fig. 6 Cyclic voltammograms of the (a) Na_0.067Li_0.85MnO₂ and (b) Na_0.05Li_0.86MnO_1.98F_0.02 subjected to 1, 2, 5, 10, 20 and 30 cycle.

The Fig. 5 also shows that the variation characteristics of capacity with cycle number for the Na_0.06Li_0.85MnO₂ and Na_0.05Li_0.86MnO_1.98F_0.02 materials. The first discharge of the Na_0.06Li_0.85MnO₂ sample is associated with a specific capacity of 133.2 mA h g⁻¹ but this falls significantly to 117.7 mA h g⁻¹ at the third cycle, then slowly increase the 190 mA h g⁻¹ after 20 cycles and the subsequent decay slowly. For the Na_0.05Li_0.86MnO_1.98F_0.02 sample, the varying pattern is similar with the Na_0.06Li_0.85MnO₂ sample. The first specific capacity on discharge is 129.2 mA h g⁻¹ but this falls significantly to 95 mA h g⁻¹ at the 5th cycle, and subsequent slowly increase to 219 mA h g⁻¹ after 40 cycles. However, the greatest difference can be found in the thereafter cycle process, the capacity on discharge is very similar at around 218 mA h g⁻¹. This is in accord with the CV in Fig. 6 and reinforces the view that the electrochemical evidence points to a change occurring during the first charge.

In order to further verify the results of electrochemical test and the influence of fluorine doping for the materials. The XRD was used to confirmed by crystal structure measurements of the cycled cathode powder. Fig. 7 shows the XRD pattern of the Na_0.067Li_0.85MnO₂ and Na_0.05Li_0.86MnO_1.98F_0.02 subjected to 30 cycles. The diffraction peaks in the XRD pattern broadened after electrochemical cycling, which is attributed to the creation of local lattice strain and tiny crystallites during cycling.⁴⁰ The results also show that the samples have been converted to the cubic spinel phase LiMn₂O₄ upon cycling, the results are consistent with the electrochemical experiments. However, the layered structure LiMnO₂ phase can obvious exist for the Na_0.05Li_0.86MnO_1.98F_0.02 materials in the electrochemical process. This fully shows that fluorine doping can effectively inhibit the structure of the shift.

Fig. 7 XRD patterns of (a) Na_0.06Li_0.85MnO₂ and (b) Na_0.05Li_0.86MnO_1.98F_0.02 after 30 cycles.

The Nyquist plots of the electrochemical impedance spectra (EIS) analyses are shown in Fig. 8. All plots consist of both a semicircle in the high-frequency region and a slope in the low-frequency region. The electrode reaction process is primarily controlled by a mixed process of lithium ion diffusion (Warburg impedance, which is inversely proportional to the diffusion coefficient D_Li⁺) and the surface electrochemical reaction at a temporary steady state of EIS. Compared to the Na_0.07Li_0.85MnO₂ sample from Table 4, the Warburg impedance of fluorine-substituted materials is lower, which contributes to their improved high rate capability. On the contrary, the charge-transfer resistance increases after fluorine substitution due to the poor electrochemical activity of the anions, while the Warburg impedance obviously decreases as compared to that of the Na_0.07Li_0.85MnO₂ sample. Clearly, the Warburg impedance of the fluorine-substituted material Na_0.06Li_0.86MnO_1.98F_0.02 is much smaller than that of the Na_0.07Li_0.85MnO₂ sample, which is consistent with the improved high-rate capability demonstrated above.

Fig. 8 Electrochemical impedance spectra (EIS) of the as-prepared samples at the discharged state (about 3.0 V (vs. Li⁺/Li)) after the 30th cycle.

Table 4 Calculated electrochemical parameters from the AC impedance spectra of the ion-substituted Li_xMnO₂ samples using ZView 2.8 software^a

Sample (ICP-AES)	Rct (Ω)	CPE (×10^−5)	Ws (Ω)
a Rct represents the charge-transfer resistance, CPE represents the constant phase element, and Ws represents the Warburg impedance.
Na0.06Li0.85MnO2	122.31	1.04	270.71
Na0.05Li0.86MnO1.98F0.02	132.81	1.08	36.25
Na0.06Li0.88MnO1.96F0.04	152.91	1.05	112.32
Na0.06Li0.87MnO1.93F0.07	221.62	2.91	117.31

---

3.3 XPS spectra analysis of materials

The XPS spectra can provide chemical information such as the oxidation state as well as the semi-quantitative composition of the surface, thus enabling observation of the surface properties. Fig. 9(a) shows the Mn 2p XPS spectra of the Na_0.05Li_0.86MnO_1.98F_0.02 sample. The binding energies of Mn 2P_1/2 and Mn 2p_3/2 were determined to be 653.86 eV and 641.80 eV, respectively. The peaks at the binding energy of about 641.60 eV are similar to those observed in Mn₂O₃ and MnOOH.⁴¹ The binding energy peak situated at 653.86 is close to the binding energy of Mn 2P_1/2 (653.90 eV) observed in the Mn₂O₃ sample. The binding energy of Na 1s was determined to be 1071.26, which is similar to that observed in NaF.⁴² Moreover, the binding energy of F 1s is situated at 686.1 eV, which is similar to that observed at 686.1 eV in MnF₂.⁴³ To further determine the existence of fluorine within the crystal structure, XPS measurement was carried out on the four samples. Fig. 9(b) shows the XPS spectra of F 1s for the pristine and F-doped materials. All curve fittings were based on C 1s (C–C bonding energy) calibration to 284.5 eV. As shown in the figure, with the increase of the amount of F-doping, the area of the F 1s peak proportionately increases, reflecting an increase in the concentration of the fluorine ions in the doped material as the composition design. This observation is constant with the consecutive changes of lattice constants shown in Section 3.1, which further proves that the fluorine ions have been successfully doped into the lattice structure.

Fig. 9 The XPS spectra of the Na_0.06Li_0.86MnO_1.98F_0.02 sample and F 1s for Li_xMnO_2−yF_y.

4. Conclusion

The fluorine doped monoclinic LiMnO₂ could be successfully synthesized via a high-temperature solid-state reaction and subsequent ion exchange by hydrothermal method. The partial fluorine substitution with high electronegativity anions strengthens the Mn–O bands in the layered monoclinic structure, which stabilizes the structure. Though the F-doped samples had lower initial capacities, they showed better cycle performance compared with F-free samples. The Na_0.05Li_0.86MnO_1.98F_0.02 electrode delivers an initial discharge capacity of 129.2 mA h g⁻¹ between 2 and 4.3 V at a current density of 50 mA g⁻¹ at room temperature, which gradually increased to a maximum discharge capacity of 219 mA h g⁻¹ after 40 cycles. The research indicated that the different performances can be attributed to the different morphologies of the respective materials as well as the effect of sodium ion concentration within the structure. Therefore, controlling the morphology of the material and the amount of sodium ions in its structure may be advantageous for developing high quality cathode materials which can be diversified for specific applications in Li-ion batteries.

Acknowledgements

This work was supported by The National Natural Science Foundation of China (21061015) and The Nature Science Foundation from Xinjiang autonomous region (2015211A037).

References

1. J. Reed and G. Ceder, Chem. Rev., 2004, 104, 4513–4534.
2. P. Suresh, A. K. Shukla and N. Munichandraiah, Electrochem. Solid-State Lett., 2005, 8, A263–A266.
3. B. Ammundsen, J. Desilvestro, T. Groutso, D. Hassell, J. B. Metson, E. Regan, R. Steiner and P. J. Pickering, J. Electrochem. Soc., 2000, 147, 4078–4082.
4. J. E. Greedan, N. P. Raju and I. J. Davidson, J. Solid State Chem., 1997, 128, 209–214.
5. Y. J. Wei, H. Ehrenberg, N. N. Bramnik, K. Nikolowski, C. Baehtz and H. Fuess, Solid State Ionics, 2007, 178, 253–257.
6. T. J. Kim, D. Son, J. Cho and B. Park, J. Power Sources, 2006, 154, 268–272.
7. A. R. Armstrong and P. G. Bruce, Nature, 1996, 381, 499–500.
8. F. Capitaine, P. Gravereau and C. Delmas, Solid State Ionics, 1996, 89, 197–202.
9. Z. F. Huang, X. Meng, C. Z. Wang, Y. Sun and G. Chen, J. Power Sources, 2006, 158, 1394–1400.
10. R. Prasad, R. Benedek and M. M. Thackeray, Phys. Rev. B: Condens. Matter Mater. Phys., 2005, 71, 134111.
11. G. Ceder and S. K. Mishra, Electrochem. Solid-State Lett., 1999, 2, 550–552.
12. C. W. Park, I. R. Mangani, H. W. Ryu, C. J. Park, J. S. Lee, S. J. Song, J. H. Moon and J. Kim, J. Phys. Chem. Solids, 2007, 68, 1126–1130.
13. A. D. Robertson, A. R. Armstrong and P. G. Bruce, Chem. Mater., 2001, 13, 2380.
14. L. H. Yu, H. X. Yang, X. P. Ai and Y. L. Cao, J. Phys. Chem. B, 2005, 109, 1148–1154.
15. B. Patoux, M. Dolle and M. M. Doeff, Chem. Mater., 2005, 17, 1044–1054.
16. J. H. Kim, C. W. Park and Y. K. Sun, Solid State Ionics, 2003, 164, 43–49.
17. B. J. Hwang, Y. W. Tsai, C. H. Chen and R. Santhanam, J. Mater. Chem., 2003, 13, 1962–1968.
18. W. K. Pang, J. Y. Lee, Y. S. Wei and S. H. Wu, Mater. Chem. Phys., 2013, 139, 241–246.
19. G. H. Kim, S. T. Myung, H. S. Kim, C. S. Yoon and Y. K. Sun, J. Electrochem. Soc., 2005, 152, A1707–A1713.
20. J. M. Zheng, X. B. Wu and Y. Yang, Electrochim. Acta, 2013, 105, 200–208.
21. S. Jouanneau and J. R. Dahn, J. Electrochem. Soc., 2004, 151, A1749–A1754.
22. Q. G. Zhang, T. Y. Peng, D. Zhan and X. H. Hu, J. Power Sources, 2014, 250, 40–49.
23. J. L. Xie, X. Huang, Z. B. Zhu and J. H. Dai, Ceram. Int., 2011, 37, 419–421.
24. W. S. Fyfe, Anal. Chem., 1951, 23(1), 174–175.
25. R. Chitrakar, H. Kanoh, Y. S. Kim, Y. Miyai and K. Ooi, J. Solid State Chem., 2001, 160, 69–76.
26. S. H. Kang, I. Belharouak, Y. K. Sun and K. Amine, J. Power Sources, 2005, 146, 650–653.
27. P. Barboux, J. M. Tarascon and F. K. Shokoohi, J. Solid State Chem., 1991, 94, 185–196.
28. E. Wolska, P. Piszora, W. Nowicki and J. Darul, Int. J. Inorg. Mater., 2001, 6, 503–507.
29. L. K. Kang, M. M. Zhang, Z. H. Liu and K. Ooi, Spectrochim. Acta, Part A, 2007, 67, 864–869.
30. H. D. B. Jenkins, H. K. Roobottom, J. Passmore and L. Glasser, Inorg. Chem., 1999, 38, 3609–3620.
31. X. Q. Wang, H. Nakamura and M. Yoshio, J. Power Sources, 2002, 110, 19–26.
32. Y.-S. He, L. Pei, X.-Z. Liao and Z.-F. Ma, J. Fluorine Chem., 2007, 128, 139–143.
33. W. Oh, S. H. Park, J. H. Kim, Y. C. Bae and Y. K. Sun, J. Power Sources, 2006, 157, 464–470.
34. G. H. Kim, M. H. Kim, S. T. Myung and Y. K. Sun, J. Power Sources, 2005, 146, 602–605.
35. Y. S. He, L. Pei, X. Z. Liao and Z. F. Ma, J. Fluorine Chem., 2007, 128, 139–143.
36. Y.-II Jang, B. Y. Huang, H. F. Wang, R. S. Donald and Y. M. Chiang, J. Electrochem. Soc., 1999, 146(9), 3217–3223.
37. P. Yue, Z. X. Wang, X. H. Li, X. H. Xiong, J. X. Wang, X. W. Wu and H. J. Guo, Electrochim. Acta, 2013, 95, 112–118.
38. S. Bach, J. P. Pereira-Ramos and P. Willmann, Electrochim. Acta, 2006, 52, 504–510.
39. J. M. Paulsen, C. L. Thomas and J. R. Dahn, J. Electrochem. Soc., 1999, 146, 3560–3565.
40. S. W. Mai, M. Q. Xu, X. L. Liao, L. D. Xing and W. S. Li, J. Power Sources, 2015, 273, 816–822.
41. B. J. Tan, K. J. Klabunde and P. M. A. Sherwood, J. Am. Chem. Soc., 1991, 113, 855–861.
42. V. I. Nefedov, Y. V. Salyn, G. Leonhsardt and R. Scheibe, J. Electron Spectrosc. Relat. Phenom., 1997, 10, 121–124.
43. R. G. Hayes, N. Edelstein and D. A. Shirley, Electron Spectroscopy, North-Holland Publishing Company, Amsterdam, 771-781, 1972.

---