Mn-doped FeF₃·0.33H₂O with enhanced electrochemical performance as cathode materials for lithium-ion batteries

Abstract

Lithium-ion batteries with the FeF₃·0.33H₂O cathode material enable a high energy density and safety. However, a major challenge of FeF₃·0.33H₂O is its low conductivity. In this work, Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05, and 0.07) are prepared via the solvent thermal method. Systematic investigations have studied the effect of Mn-doping on the physical and electrochemical properties. The results indicate that Mn-doping not only does not destroy the lattice structure of FeF₃·0.33H₂O, but also reduces the resistance and improves the diffusion coefficient of lithium ion, which provide it with better electrochemical properties. Fe_0.98Mn_0.03F₃·0.33H₂O delivers much excellent cycling performance and rate capacity than other materials. It has a 284.2 mAh g⁻¹ initial discharge capacity that remains at 258.9 mAh g⁻¹ after 50 cycles at 0.1 C, giving the high capacity retention rate of 91.1%. Additionally, the initial discharge capacity of Fe_0.98Mn_0.03F₃·0.33H₂O is 245, 231, 217, and 203 mAh g⁻¹ at 1, 2, 5, and 10 C in the voltage range of 1.5–4.5 V vs. Li^+/Li, respectively.

Introduction

Lithium-ion batteries (LIBs) that can maintain a high energy density, high working voltage, and long lifespan are necessary for applications in portable electronics, electric vehicles, and large-scale energy storage [1,2,3,4,5]. However, the greatest challenge for battery technology is meeting the demand for high-capacity and safe batteries. Therefore, significant interest is directed toward developing high performance electrode materials, which is the key for maximizing the performance of lithium-ion batteries.

Considerable focus has been placed on investigating iron-based fluorides due to their high theoretical energy density (712 mAh g⁻¹, three electrons transfer), low costs, and safety, which provide it with great potential for use as cathode materials in high-capacity lithium-ion batteries [6, 7]. However, there are several issues for these materials such as strong ionic bonds, a wide band gap and poor conductivity, which result in a low material rate performance that limits practical applications [8, 9]. FeF₃ is a multiphase crystal that mainly includes the following: anhydrous FeF₃ [10], FeF₃·0.33H₂O [11], FeF₃·0.5H₂O [12], and FeF₃·3H₂O [13]. Meanwhile, researchers have found that FeF₃·0.33H₂O has the most stable crystal structure and the best electrochemical performance [14,15,16]. The structure of FeF₃·0.33H₂O belongs to the orthogonal crystal system, the hexagonal tungsten bronze phase, in which iron atoms and fluorine atoms form a hexagonal tunnel, and water molecules are in the middle of the hexagon tunnel. Furthermore, the pore channels are benefit to the entry and exit of lithium ions, while the water molecules in the middle are conducive to the stability of the structure and the improvement of the conductivity. Researchers believe that it is this unique structure that makes this material’s electrochemical performance better than other iron fluoride-containing crystal water [17,18,19,20,21].

The methods, such as the ball mill method [22, 23], the liquid phase method [24, 25], the template method [26, 27], and the solvent thermal method [28, 29], can use to prepare the FeF₃ cathode material. The materials prepared by these different methods contain different crystal water. Moreover, the pure phase FeF₃·0.33H₂O can be directly synthesized by the simple solvent thermal method. This method is more prone to industrial production due to the controllable preparation conditions and low production costs.

The materials prepared by the solvent thermal method exhibit their excellent morphology and crystal structure. Nevertheless, the pure phase has poor conductivity and the actual capacity is relatively low [30]. Solving the issues, the approaches are often applied to improve the electrochemical properties of the cathode material, primarily cationic doping [31] and surface coating [32]. However, surface coating has the defect of being uneven, which easily falls off with the collecting fluid in the process of circulation and reduces the circulation performance. Doping is an effective way to fabricate high-performance material. For example, Fe_1-xCo_xF₃·0.33H₂O nanomaterials using the liquid phase method are synthesized by Sun et al. [33]. The result explains that an excellent electrochemical performance could be achieved in cationic-doped FeF₃·0.33H₂O cathode material for lithium-ion batteries. Yang et al. [34] calculate that Ti-doping can expand the hexagonal cavity of FeF₃·0.33H₂O through the first-principles theory, thereby improving the electrical conductivity of FeF₃·0.33H₂O. Therefore, one can expect that the performance of FeF₃·0.33H₂O can be improved by Mn-doping.

In this work, we report on a facile strategy for the synthesis of Mn-doped FeF₃·0.33H₂O to prepare Fe_1-2x/3Mn_xF₃·0.33H₂O. The effects of different doping amounts on the physical and electrochemical properties are comprehensively investigated.

Experimental

Fabrication of Fe_1-2x/3Mn_xF₃·0.33H₂O

Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05, and 0.07) was prepared by the hydrothermal method. Fe(NO₃)₃·9H₂O and (CH₃COO)₂Mn·4H₂O in a molar ratio of 1:0, 0.994:0.01, 0.98:0.03, 0.967:0.05, and 0.954:0.07 were dissolved in a polytetrafluoroethylene container containing ethanol, cetyl trimethyl ammonium bromide, and Tween 80. The mixture was continuously stirred until it was dissolved to obtain reddish brown solution, and 15 ml 40 wt.% HF solution was slowly added with stirring. After 30 min, the polytetrafluoroethylene container was placed in a hydrothermal reaction vessel and sealed and then heated at 120 °C for 6 h in a blast drying oven. After cooling to room temperature, the light pink precipitate was obtained. Next, the precipitate was washed 4 times with ethanol to remove the unreacted HF and water, and then dried at 120 °C for 3 h to obtain a dried powder. The dried and mixed reactant mixture was reground and then heated in a tube furnace at 230 °C for 3 h under high-purity Ar protection to obtain Fe_1-2x/3Mn_xF₃·0.33H₂O.

Material characterizations

The X-ray diffusion (XRD) phase test of the synthesized samples was done with an XD-2 using a Cu K_α radiation source (λ = 0.15405 nm). All of the diffraction patterns were recorded at room temperature in the 2θ range from 10° to 80° with a scan rate of 4°/min and were operated at 36 KV and 20 mA. The morphological and crystalline structures of the samples were characterized by a scanning electron microscope (SEM, JSM-7001F, JEOL).

Electrochemical measurements

The electrochemical performance of the obtained particles was tested using a CR2032 coin-type cell. First, the active material Fe_1-2x/3Mn_xF₃·0.33H₂O (90 wt%), acetylene black (5 wt%), and polyvinylidene fluoride binder (5 wt%) were mixed and ground with N-methylpyrrolidone as a solvent to form a uniform slurry, which was coated on aluminum foil and dried under a vacuum at 120 °C for 24 h. Then, the foil was pressed and cut into disks, and a Celgard 2400 polypropylene microporous membrane were used as the separator and lithium foil were used as the negative electrode, respectively. The electrolyte consists of 1 M LiPF₆ in a mixture of ethylene carbonate (EC), dimethyl carbonate (DMC), and ethylene methyl carbonate (EMC) (volume ratio 1:1:1). The cells were assembled in an argon glove box.

The galvanostatic discharge-charge characteristics of the cells were performed over the potential range between 1.5 and 4.5 V (vs. Li/ Li⁺) using a Blue battery tester system (CT2001A Wuhan, China) at different rates of 0.1–10 C at room temperature. The cyclic voltammetry (CV) tests were measured using a CHI600A electrochemical workstation under a scan rate of 0.5 mV s⁻¹ between 1.5 and 4.5 V (vs. Li/ Li⁺). Electrochemical impedance spectroscopy (EIS) tests were performed by using an electrochemical workstation in the frequency range of 100 to 10 mHz with the perturbation of 5 mV. All of the characterizations and electrochemical measurements were conducted at room temperature (25 °C).

Results and discussion

Figure 1 shows the XRD patterns of Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05, and 0.07). Among all investigated samples, no obvious diffraction peaks for any impurities are detected, the major X-ray diffraction peaks at the 2θ angles of 13.8°, 23.6°, and 27.8°, matching the standard diffraction pattern of FeF₃·0.33H₂O, are indexed as the (110), (002), and (220) crystal planes, respectively. The results indicate that Mn-doping still maintains the material’s structure, which shows the orthorhombic structure of the Cmcm space group. Furthermore, with the increase of the Mn content, the diffraction peaks have not only enhanced intensity but also sharp shape owing to a high crystallinity of the material. Additionally, no diffraction peaks of Mn are detected in the Fe_1-2x/3Mn_xF₃·0.33H₂O samples, indicating that Mn is effectively doped, but it does not change the lattice structure of the FeF₃·0.33H₂O.

Fig. 1

XRD patterns of Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05, and 0.07)

Figure 2 is SEM images of Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05, and 0.07). The FeF₃·0.33H₂O exhibits a rod-like structure that is approximately 3–9-μm long and 0.5–2-μm wide and can be seen in Fig. 2a. The surfaces of the pure phase FeF₃·0.33H₂O particles are relatively smooth. By contrast, the surfaces of the Mn-doped sample particles are rough, and this can greatly increase the specific surface area of the particles, and could help promoting the penetration of the electrolyte, these factors result in enhancing the electrochemical performance [35, 36]. The comparison shows that the particle size of Fe_0.98Mn_0.03F₃·0.33H₂O is the smallest in Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05 and 0.07). The electrode material with small particle size has larger specific surface area, and the electrolyte is more easily penetrated with the fastest ion diffusion rate. The SEM image reveals that it is effective to reduce the particle size and shorten the process of crystal growth via the Mn-doping. A smaller particle size can strongly reduce the resistance, possessing a superior electrochemical performance.

Fig. 2

SEM images of Fe_1-2x/3Mn_xF₃·0.33H₂O: ax = 0, bx = 0.01, cx = 0.03, dx = 0.05, ex = 0.07

Figure 3 shows charge/discharge performance curves of Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05, and 0.07) cathode material at 0.1 C. As shown in Fig. 3a, the pure phase FeF₃·0.33H₂O has a relatively low initial discharge specific capacity (234.8 mAh g⁻¹), which may be attributed to its poor conductivity. However, the doped samples have a high discharge specific capacity. The electrode can deliver the initial discharge capacities of 246.1, 284.2, 275.6, and 261.8 mAh g⁻¹ when doping amounts are x = 0.01, 0.03, 0.05, and 0.07, respectively. Noticeably, the change of the discharge capacity is not linearly related to the increase of Mn content. The discharge capacity decreases when the doping amount reaches the maximum, which suggests that the Mn doping amount should not be too high. It is worth noting that the Fe_0.98Mn_0.03F₃·0.33H₂O cathode material possesses the highest discharge capacity. Figure 3b displays the discharge capacity of the pure phase FeF₃·0.33H₂O decreases to 194.9 mAh g⁻¹, retaining 83% of initial capacity after 50 cycles. The discharge capacities of Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0.01, 0.03, 0.05, and 0.07) remain at 216.8, 258.9, 246.9, and 232.7 mAh g⁻¹, and the capacity retention rates are 88.1%, 91.1%, 89.6%, and 88.9%, respectively. It shows a great cycling stability of Mn-doped material. The Fe_0.98Mn_0.03F₃·0.33H₂O cathode material exhibits an excellent cycling performance after 50 cycles. All these above features indicate that an improved electrochemical performance is achieved with moderate Mn-doping. On the one hand, Mn-doping can effectively reduce the size of FeF₃·0.33H₂O particles as well as the specific surface area so as to shorten the diffusion path length for Li⁺ and electron transport. On the other hand, Mn²⁺ partially substituted Fe³⁺ reduces the band gap, which improves the conductivity of the material.

Fig. 3

Initial (a) and after 50 cycles (b) charge/discharge curves of Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05, and 0.07) at 0.1 C

Rate performance of Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05 and 0.07) from 0.1 to 10 C is presented in Fig. 4. The average discharge capacity at different rates is displayed in Fig. 5. Clearly, the higher discharge capacities of Mn-doped samples than that of FeF₃·0.33H₂O are achieved at different rates, suggesting good conductivity of Mn-doped samples. Besides, the lower discharge capacities of the five samples, can be obtained with the increasing of different rates (from 0.1 to 10 C), exhibiting different rate of reduction. Among them, Fe_0.98Mn_0.03F₃·0.33H₂O, which shows the higher discharge capacity and the slowest rate of reduction, reaches the average discharge capacities of 282.3, 271.3, 259.6, 243.4, 229.6, 215.8, and 202.0 mAh g⁻¹ at 0.1, 0.2, 0.5, 1, 2, 5, and 10 C rates, respectively. When the rate returned back to 0.1 C, the specific capacity of Fe_0.98Mn_0.03F₃·0.33H₂O can be quickly returned to 276.1 mAh g⁻¹ with a capacity retention of 97.8%. At the first current density of 0.1 C, the discharge capacity of 5 discharge cycles deteriorates dramatically because the initial irreversible reaction results in irreversible capacity.

Fig. 4

Cyclic performances of Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05, and 0.07) at different rates

Fig. 5

The average values of discharge capacities at different C rates

To evaluate the cycling performance of FeF₃·0.33H₂O and Fe_0.98Mn_0.03F₃·0.33H₂O, the samples are cycled 50 times at 1 C to obtain cycling performance curves, as shown in Fig. 6. The capacity loss in the first 10 times is obvious, which is caused by the occurrence of side reactions and the formation of the SEI film, leading to a partial irreversible capacity. The first discharge capacities of Fe_0.98Mn_0.03F₃·0.33H₂O and FeF₃·0.33H₂O are 245.2 and 175.3 mAh g⁻¹. After 50 cycles the samples maintain a capacity of 219.6 and 134.6 mAh g⁻¹ with the capacity retention of 89.6% and 76.8%, respectively, and especially the coulombic efficiencies reach nearly 99%. The excellent cycling stability of Fe_0.98Mn_0.03F₃·0.33H₂O is attributed to the following: (i) Mn-doping effectively increases the unit cell volume by reducing the activation energy barrier required for Li⁺ diffusion and increasing the Li⁺ diffusion coefficient; and (ii) the size of FeF₃·0.33H₂O particles is reduced by Mn²⁺ partially substituted Fe³⁺ providing a transport channel for the insertion and conversion reaction of Li⁺, which facilitates the migration of lithium ions and improves the conductivity of the material.

Fig. 6

Cyclic performances of Fe_0.98Mn_0.03F₃·0.33H₂O and FeF₃·0.33H₂O at 1 C and 1.5–4.5 V

Figure 7 exhibits the CV curves of Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05, and 0.07) obtained at the scanning voltage of 1.5 to 4.5 V and the scanning rate of 0.5 mV s⁻¹. In the CV curves, the peak in the positive direction of the ordinate axis is the oxidation peak, and that in the negative direction is the reduction peak. The potential difference between the two peaks reflects the polarization degree of the electrode. The electrode reaction process of FeF₃ in a lithium-ion battery is [37, 38].

$$ \mathrm{Fe}{\mathrm{F}}_3+\mathrm{Li}\to \mathrm{LiFe}{\mathrm{F}}_3\left(4.5-1.7\mathrm{V}\right) $$

(1)

$$ \mathrm{LiFe}{\mathrm{F}}_3+2\mathrm{Li}\to \mathrm{Fe}+3\mathrm{LiF}\left(1.7-1.5\mathrm{V}\right) $$

(2)

Fig. 7

CV curves of Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05, and 0.07) at a scan rate of 0.5 mV s⁻¹

It can be seen from the figure that two pairs of redox peaks are observed, corresponding to the conversion between the Fe³⁺ and Fe²⁺ phases in the Li⁺ insertion/extraction behaviors and the chemical transformation reaction between the Fe²⁺ and Fe⁰ phases, which are consistent with Fig. 3. The potential interval of Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05, and 0.07) are 1.07, 0.98, 0.61, 0.66, and 0.83 V, respectively, indicating that Fe_0.98Mn_0.03F₃·0.33H₂O (△E = 0.61) has the smallest polarization.

Figure 8 shows the EIS curves of Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05, and 0.07) cathode materials after 1 cycle, 20 cycles, and 50 cycles at 1 C. The EIS profiles exhibit a semicircle in the high frequency region and an oblique line in the low frequency region. The equivalent circuit model is shown in the figure, which shows that the fitted graph and the experimentally measured pattern match well. As showed in Table 1, the fitted values of the impedance spectra can be obtained by the relevant equivalent circuit model. In the equivalent circuit, R_s denotes the impedance of the electrode system, R_ct represents the charge transfer impedance, CPE is used instead of the interface impedance, and W is the Warburg impedance of the solid phase diffusion, which is related to the diffusion dynamics of Li⁺ in the solid lattice. It can be seen from the figure that there are no significant differences between the R_s values of the samples. Moreover, the R_s and R_ct values of the Mn-doped composites are lower than those of FeF₃·0.33H₂O with the increasing cycle number. Especially, the R_s and R_ct values of Fe_0.98Mn_0.03F₃·0.33H₂O show a minimal change. The R_ct values of FeF₃·0.33H₂O are 367.52, 399.10, and 468.95 Ω after 1 cycle, 20 cycles, and 50 cycles, respectively, while the R_ct values of Fe_0.98Mn_0.03F₃·0.33H₂O are 46.96, 59.09 and 70.14 Ω after 1 cycle, 20 cycles, and 50 cycles, respectively, much less than those of the FeF₃·0.33H₂O. This phenomenon confirms that Fe_0.98Mn_0.03F₃·0.33H₂O presents minimum polarization when the electrode reaction occurs, which is consistent with the CV result. The smaller polarization is of benefit to the insertion and extraction of Li⁺, effectively enhancing the stability of the positive electrode material. Therefore, benefiting from an appropriate amount of Mn-doping to reduce the charge transfer impedance, the electrochemical reversibility of the material can be improved.

Fig. 8

EIS curves of Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05, and 0.07) cathode materials after 1 cycle, 20 cycles, and 50 cycles at 1 C

Table 1 Simulated impedance parameters of Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05, and 0.07)

Figure 9 illustrates the linear relationship between Z_re and the reciprocal square root of the lower angular frequencies ω^-1/2. According to Eqs. 3 and 4 [39, 40], we can calculate the D_Li⁺ of Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05, and 0.07).

$$ {Z}_{re}={R}_S+{R}_{ct}+{\sigma}_w\cdot {\omega}^{-0.5} $$

(3)

$$ {D}_{Li+}=\frac{R^2{T}^2}{2{A}^2{n}^4{F}^4{C}_{Li}^2{\sigma_w}^2} $$

(4)

Fig. 9

Graph of Z_re plotted against ω^-1/2 after 50 cycles for Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0.01, 0.03, 0.05, and 0.07) cathode material

Table 2 shows that the electrochemical kinetics parameters of Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05, and 0.07) after 50 cycles at 1 C. The R_s values of Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0.07) is lower than those of the other materials, whereas it is clear that Fe_0.98Mn_0.03F₃·0.33H₂O shows the highest D_Li+, 1.95 × 10⁻¹³ cm² s⁻¹, among all materials, which corresponds to its highest initial discharge specific capacity. This result indicates Mn-doping helps to improve the conductivity of the material and accelerate the charge transfer reaction at the electrode/electrolyte interface, possessing good multiplier performance. More importantly, the D_Li+ increases from 10⁻¹⁴ cm² s⁻¹ to 10⁻¹³ cm² s⁻¹, suggesting that Mn-doping is beneficial to Li⁺ stripping/embedding, which not only increases the conductivity of lithium ions, but also decreases the polarization, thus improving the reaction kinetics of the electrode.

Table 2 Electrochemical Kinetic Parameters of Fe_1-2x/3Mn_xF₃·0.33H₂O (x = 0, 0.01, 0.03, 0.05, and 0.07) after 50 cycles at 1 C

Conclusion

In conclusion, the Fe_1-2x/3Mn_xF₃·0.33H₂O cathode material was successfully fabricated via the solvent thermal method. The capability and cycling stability of these doped materials have been improved. Notably, Fe_0.98Mn_0.03F₃·0.33H₂O exhibits a high capacity of 245.2 mAh g⁻¹ and it still retains 219.6 mAh g⁻¹ at 1 C even after 50 cycles, which are higher than those of FeF₃·0.33H₂O, indicating that Mn-doping can improve the multiplier performance of FeF₃·0.33H₂O. Meanwhile, the EIS results reveal the charge transfer resistance of FeF₃·0.33H₂O decreases greatly after Mn-doping. The Fe_0.98Mn_0.03F₃·0.33H₂O has highest Li⁺ diffusion coefficient (1.95 × 10⁻¹³ cm² s⁻¹). This outstanding battery performance of Fe_0.98Mn_0.03F₃·0.33H₂O can primarily be attributed to Mn-doping, which reduces the size of particles, thereby accelerating the diffusion rate and reducing the polarization of the electrochemical reaction. Therefore, Fe_0.98Mn_0.03F₃·0.33H₂O promises to be a cathode material for power lithium-ion batteries.