Electrochemical Behavior of NH₄F-Pretreated Li_1.25Ni_0.20Fe_0.13Co_0.33Mn_0.33O₂ Cathodes for Lithium-ion Batteries

Abstract

We report a novel method to fabricate lithium-ion batteries cathodes with the NH₄F pretreatment. In this study, NH₄F-pretreated Li_1.25Ni_0.20Fe_0.13Co_0.33Mn_0.33O₂ hollow nano-micro hierarchical microspheres were synthesized for use as cathode materials. The X-ray diffraction patterns of NH₄F-pretreated Li_1.25Ni_0.20Co_0.33Fe_0.13Mn_0.33O₂ were analyzed with the RIETAN-FP software program, and the results showed that the samples possess a layered α-NaFeO₂ structure. The effects of pretreatment with NH₄F on the electrochemical performance of the pristine material were evaluated through charge/discharge cycling, the rate performance, and electrochemical impedance spectroscopy (EIS). Pretreatment with NH₄F significantly improved the discharge capacities and coulombic efficiencies of Li_1.25Ni_0.20Co_0.33Fe_0.13Mn_0.33O₂ in the first cycle and during subsequent electrochemical cycling. The sample pretreated with an appropriate amount of NH₄F (NFCM 90) showed the highest discharge capacity (209.1 mA h g⁻¹) and capacity retention (85.2% for 50 cycles at 0.1 C). The EIS results showed that the resistance of the NFCM 90 sample (76.32 Ω) is lower than that of the pristine one (206.2 Ω).

1. Introduction

Economic development has led to increased research on electrochemical energy storage techniques for new energy sources, such as lithium-ion [1,2,3], lithium–sulfur [4], and sodium-ion batteries [5], of which lithium-ion batteries are the most promising and widely investigated [6,7,8,9,10]. The development of lithium-ion batteries has been restricted because the capacities of anode materials are much better than those of cathode materials. LiNi_1/3Co_1/3Mn_1/3O₂ (NCM111) which has been widely investigated is a better cathode material choice than LiCoO₂ in terms of reversible capacity, thermal and structural stability, and cost. However, the rapid capacity fading and inferior rate capability of NCM limit its sustainability and prevent its wide use in high-power applications.

Many different methods have been used to solve the above problems, e.g., formation of special morphologies, bulk doping, surface coating, and pretreatment. There are already relevant experiments on the above that constructing electrode materials with different microstructures, coating the surfaces of lithium-ion battery cathodes with materials, such as carbon and CoAl₂O₄, and doping positive electrode materials with metal ions, such as Fe and Zn and so on. Spinel-phase LiM₂O₄ with oxygen vacancies can be formed on the surfaces of cathode materials through pretreatment. The formed spinel phase improves the diffusion of lithium ions and the escape of O₂ gas from the cathode surface during the cycle, which helps improve electrochemical performance [11]. The most common electrolyte component, namely LiPF₆, produces a small amount of HF via reaction with water and corrodes the oxide electrode material during charging and discharging [12]. Electrode corrosion by a highly active organic electrolyte can be resolved by the addition of fluorides, such as CeF [13], Mg₂F [14], ZrF_x [15,16], AlF₃ [17], FeF₃ [18], CaF₂ [19], and LiF [20,21]. The LiFe_xNi_1/3-xCo_1/3Mn_1/3O₂ hollow micro-spheres as cathodes for LIBs have been shown to have the best electrochemical performance at x = 0.133 [22], although hollow microspherical structure enhances performance, the iron substitution increases microspheres aggregation, and also reduces electrochemical properties as decreasing lithium ion diffusion.

In this study, we adopt a simple and reliable method for electrochemical performance improvement. The method involves NH₄F pretreatment of Li_1.25Ni_0.2Fe_0.13Co_0.33Mn_0.33O₂ cathodes, which were obtained by using MnO₂ microspheres as self-templates. X-ray diffraction (XRD), scanning electron microscopy (SEM), energy-dispersive X-ray spectroscopy (EDS), and electrochemical tests showed that the structure, morphology, and electrochemical properties of Li_1.25Ni_0.2Fe_0.13Co_0.33Mn_0.33O₂ after NH₄F pretreatment need to be considered.

2. Experimental

2.1. Materials Synthesis

MnO₂ microspheres were prepared via a previously reported method [23]. The prepared MnO₂, Ni(NO₃)₂·6H₂O (AR, 98%, Aladdin, Shanghai, China), Co(NO₃)₂·6H₂O (AR, 99%, Aladdin, China), Fe(NO₃)₃·9H₂O (AR, Aladdin, China), and LiOH·H₂O (AR, 98%, Aladdin, China) were added to ethanol in specific proportions (metal-ion molar ratios Li:Fe:Ni:Co:Mn = 1.25:0.13:0.2:0.33:0.33) to form a suspension. The suspension was stirred at 60 °C until the ethanol evaporated. The mixture was calcined at 850 °C for 15 h in air to give rise to Li_1.25Ni_0.2Fe_0.13Co_0.33Mn_0.33O₂ [24,25].

The NH₄F-pretreated Li_1.25Ni_0.2Fe_0.13Co_0.33Mn_0.33O₂ materials were prepared as follows. NH₄F was dissolved in deionized water. The obtained Li_1.25Ni_0.2Fe_0.13Co_0.33Mn_0.33O₂ powder was homogeneously dispersed in the NH₄F solution under stirring. Samples with Li_1.25Ni_0.2Fe_0.13Co_0.33Mn_0.33O₂ and NH₄F molar percentages of 95% and 5% (NFCM 95), 90% and 10% (NFCM 90), 85% and 15% (NFCM 85), and 80% and 20% (NFCM 80) were prepared. The mixed solutions were dried until the solvent completely evaporated and the resulting materials were calcined at 400 °C for 5 h in air.

2.2. Characterization

XRD (Rigaku SmArtlab-3KW, Tokyo, Japan, Cu Kα radiation) was performed in the range 15° to 70° to determine the crystal structures of the pristine and NH₄F-pretreated materials. The XRD results were analyzed with the RIETAN-FP program. The elemental compositions of the samples were determined by EDS (Oxford INCA, Oxfordshire, UK). The particle morphology was examined by scanning electronic microscopy (SEM, STEM Sigma500, Carl Zeiss, Oberkochen, Germany) and high-resolution transmission electron microscopy (HRTEM, JEOL JEM-2100F, Tokyo, Japan).

2.3. Electrochemical Measurements

The electrochemical performances were investigated at 25 °C with the LAND CT2001A battery tester (LANHE, Guangzhou, China), which using CR2025-type coin cell including cathode and lithium metal anode separated by the Celgard 2400 film. The cathode consisted of stoichiometric amounts of the active materials, conductive carbon, and poly (vinylidene fluoride) (mass ratios 8:1:1). N-Methy-2-pyrrolidinone (>99.5%, Aladdin, China) was added to the mixture to obtain a slurry, which was spread on an aluminum foil current collector and then dried in a vacuum at 100 °C for 12 h. Then use the tableting mechanism to produce a round electrode with a surface area of about 1.1 square centimeters. Use of an argon-filled glove box (water < 1 ppm, O₂ < 1 ppm) is essential for assembly of the coin cells. The electrolyte was 1 M LiPF₆ in a 1:1 v/v mixture of ethylene carbonate and diethyl carbonate (Kejing Group, Hefei, China). The NEWARE test system was used to investigate the charge/discharge performances of the cells in the voltage range 2.5–4.5 V at 25 °C. Electrochemical impedance spectroscopy (EIS) was performed with an electrochemical workstation (PARSTAT 2273, AMETEK, Berwyn, PA, USA) in the range 0.1 Hz to 100 kHz. Cyclic voltammetry (CV) was performed at a scanning rate of 0.1 mV s⁻¹ in the range 2–5 V.

3. Results and Discussion

Figure 1a shows the XRD patterns of pristine and NH₄F-pretreated samples. The patterns show that all samples were layered α-NaFeO₂ structure (rhombohedral lattice with the R-3m space group). The clear separations of the (006) from the (012) peak and the (108) from the (110) peak confirm the formation of a highly ordered hexagonal layered structure.

The RIETAN-FP program was used to analyze XRD patterns of pristine and NH₄F-pretreated samples, the calculated structural paraments are listed in

Table 1. The lattice parameters of all samples refined by the RIETAN-FP.

x	a (Å)	c (Å)	V (Å)	Rwp (%)	Re(%)	S
NFCM 80	2.861(5)	14.25(1)	101.06(6)	4.309	4.054	1.0630
NFCM 85	2.861(4)	14.24(9)	101.03(5)	4.227	3.998	1.0574
NFCM 90	2.861(4)	14.25(7)	101.10(0)	4.183	4.161	1.0054
NFCM 95	2.861(7)	14.25(4)	101.09(5)	4.303	4.221	1.0194
Pristine	2.861(4)	14.26(7)	101.17(1)	4.353	4.048	1.0754

. The values of

$$\left(\equiv \sqrt{{{\displaystyle \sum }}_{i=1}^N{\omega }_i{\left[y_i-f_i\left(x\right)\right]}^2/{{\displaystyle \sum }}_{i=1}^N{\omega }_i{y_i}^2}\right)$$

(1)

$$R_e\left(\equiv \sqrt{\left(N-n\right)/{{\displaystyle \sum }}_{i=1}^N{\omega }_i{y_i}^2}\right)$$

(2)

and S (≡ R_wp/R_e) are sufficiently small, which confirms that the final refinements are reliable. The c/a values are 4.986 and 4.982 (both greater than 4.90, Table 1) and the intensity ratios [I(003)/I(104)] are 1.54 and 1.41 (both greater than 1.2, Figure 1a) in the XRD patterns of pristine and NFCM 90 samples. This shows that NFCM 90 sample has stable layered structure and low degree of cation mixing. Figure 1b shows the XRD powder pattern (crosses) of the NFCM 90 sample together with the Rietveld refinement (solid curve); the green line corresponds to the recorded curve, the blue sticks correspond to the peak positions, the purple line corresponds to the fitted curve, and the red line corresponds to the error curve.

The morphologies and microstructures of the obtained electrode materials are shown in Figure 2. The submicron MnO₂ spheres, which were used for self-templating, affected the morphologies and particle sizes of the obtained materials. The materials clearly consist of sub-micron spheres of particle size 1–2 µm. There are no obvious differences among the spherical shapes and particle sizes of the pristine and NH₄F-pretreated samples. This shows that treatment with NH₄F did not significantly change the morphology of pristine material.

Figure 3 shows TEM and HRTEM images of pristine and NFCM 90 samples. The TEM images show that treatment with NH₄F changed the electrode material surface. Figure 3b shows a lattice spacing of 0.47 nm corresponding to the (003) plane of the layered phase, and Figure 3d shows lattice spacings of 0.24 and 0.20 nm corresponding to the (222) and (004) planes, respectively, of the spinel phase. The above results prove the presence of the spinel phase on the pretreated material surfaces.

The chemical composition of the NFCM 90 sample was determined by EDS. The elemental mappings (Figure 4) show the distributions of Ni, Co, Mn, Fe, and F elements, in the ratios Fe:Ni:Mn:Co = 4.27:6.03:11.34:10.96, which are close to the theoretical molar ratios. The weight percentage of fluorine in the positive electrode material is close to the expected value, which indicates successful preparation of the positive electrode material.

The electrochemical performances of pristine and NFCM 90 samples were investigated. Figure 5a shows the discharge curves of samples at 0.1 C (1 C = 278 mA h g⁻¹). After 50 cycles, the discharge capacities of pristine and NFCM 90 samples are 129.4 and 178.2 mA h g⁻¹, respectively. The capacity retention rates for pristine and NFCM 90 samples are 76.3% (169.4 mA h g⁻¹) and 85.2% (209.1 mA h g⁻¹), respectively. Figure 5b shows the charge/discharge curves of the initial cycle for pristine and NFCM 90 samples at 0.1 C. The continuity of the voltage curve during the charge-discharge cycles confirms the stability of electrode material. The charge capacities of pristine and sample NFCM 90 are 195.1 and 250 mA h g⁻¹, and the discharge capacities are 169.4 and 209.1 mA h g⁻¹, respectively. The NFCM 90 sample has a higher capacity retention, higher than those of pristine samples after 15 cycles (92.3% > 90.4%) and 30 cycles (87.5% > 82.8%). Pretreatment with an appropriate amount of NH₄F can prevent the cathode surface from reacting with the electrolyte. Partial substitution of oxygen in the crystal by fluorine improves the cycling stability of the NH₄F-pretreated samples. The performances of lithium batteries can be degraded by treatment with an excessive amount of NH₄F. Electrolyte corrosion by NH₄F therefore needs to be considered.

Figure 5c shows the discharge curves of pristine and NFCM 90 samples at various rates from 0.1 to 2 C. The voltage charge must be kept constant for 30 min to reach the predetermined voltage when the constant current charge at rates higher than 0.5 C. The discharge capacities of the NFCM 90 sample at 0.1, 0.5, 1, and 2 C are 206.4, 161, 106, and 47.5 mA h g⁻¹, respectively. The capacity of the NFCM 90 sample returns to 182.8 mA h g⁻¹ (88.6%) when the current rate finally returns to 0.1 C, whereas the pristine sample capacity returns to 141.6 mA h g⁻¹ (85.7%). This suggests that the NFCM 90 sample has better reversibility. Formation of spinel phase after NH₄F pretreatment can improve the lithium ion diffusion rate, form oxygen vacancies to limit the release of O₂ from the cathode surface, and improves the structure of the solid electrolyte interphase (SEI) film on the cathode surface. These factors may be responsible for the better rate capacity of the NFCM 90 sample.

Figure 5d shows the cycle voltammetry results for the pristine and NFCM 90 samples. Lithium ion extraction ability is related to oxidation peak, and lithium ion insertion ability is related to reduction peak. The specific capacity of the electrode material is related to the redox peak area. The potential interval between the anode peak and the cathode peak represents the electrode polarization capacity, which indicates the reversibility of lithium ion insertion and extraction. The potential interval for the NFCM 90 sample is 0.39 V, which is smaller than the pristine one (0.44 V), indicating a smaller electrode polarization. The redox peak for the NFCM 90 sample is larger than that for the pristine sample, which indicates a higher specific capacity.

Figure 6a shows the EIS results for pristine and NFCM 90 samples. Figure 6b shows the fitted experimental results and equivalent circuits (inset). At the interface between the electrode material and electrolyte, the capacitance double-layer and Li-ions form a semicircle from high frequency to intermediate frequency in the AC impedance spectrum under the action of the interfacial exchange resistance. The exchange resistance of Li-ions at the interface between electrode material and electrolyte corresponds to the intercept of the semicircle on the x-axis, which represents the charge-transfer resistance (R_ct) in the analog circuit. The detailed fitted results are listed in

Table 2. Parameters obtained from simulation of elements in an equivalent circuit model.

Sample	Re (Ω)	Rct (Ω)
Pristine	4.491	206.2
NFCM 90	8.753	76.32

. The R_ct (76.3 Ω) of the NFCM 90 sample electrode is smaller than the pristine one (206.2 Ω), which indicates that the R_ct of the NFCM 90 sample in the lithium-ion battery is smaller, and this confirms that NH₄F pretreatment can effectively inhibit the reaction between Li_1.25Ni_0.2Fe_0.13Co_0.33Mn_0.33O₂ particles and the electrolyte, as well as retard thickening of the SEI film. The NFCM 90 sample is suitable as lithium-ion anode material because it has better conductivity, in addition to electrolyte oxidation and electrolyte/electrode interface degradation being effectively alleviated through NH₄F pretreatment.

4. Conclusions

Pretreatment with NH₄F and self-templating with MnO₂ microspheres enabled successful synthesis of Li_1.25Ni_0.2Fe_0.13Co_0.33Mn_0.33O₂. The XRD and SEM results show that pretreatment did not obviously affect the crystal structures, morphologies, nor particle sizes of the obtained Li_1.25Ni_0.2Fe_0.13Co_0.33Mn_0.33O₂ samples. NH₄F pretreatment improved initial coulombic efficiency and stability in subsequent electrochemical cycling, as well as rate capability. The NFCM 90 sample had a rate capacity of 209.1 mA h g⁻¹ at 0.1 C and a capacity retention of 85.2% (178.2 mA h g⁻¹) after 50 cycles at 0.1 C. These values are better than the pristine ones. We developed a feasible method for synthesizing NH₄F-pretreated lithium-rich layered oxides, which contributes to promote electrochemical performances of Li-ion batteries.