Three-Dimensional Carbon-Coated LiFePO₄ Cathode with Improved Li-Ion Battery Performance

Abstract

LiFePO₄ (LFPO)has great potential as the cathode material for lithium-ion batteries; it has a high theoretical capacity (170 m·A·h·g⁻¹), high safety, low toxicity and good economic benefits. However, low conductivity and a low diffusion rate inhibit its future development. To overcome these weaknesses, three-dimensional carbon-coated LiFePO₄ that incorporates a high capacity, superior conductivity and low volume expansion enables faster electron transport channels. The use of Cetyltrimethyl Ammonium Bromid (CTAB) modification only requires a simple water bath and sintering, without the need to add a carbon source in the LFPO synthesis process. In this way, the electrode shows excellent reversible capacity, as high as 159.8 m·A·h·g⁻¹ at 2 C, superior rate capability with 97.3 m·A·h·g⁻¹ at 5 C and good cycling ability, preserving ~84.2% capacity after 500 cycles. By increasing the ion transport rate and enhancing the structural stability of LFPO nanoparticles, the LFPO-positive electrode achieves excellent initial capacity and cycle life through cost-effective and easy-to-implement carbon coating. This simple three-dimensional carbon-coated LiFePO₄ provides a new and simple idea for obtaining comprehensive and high-performance electrode materials in the field of lithium cathode materials.

1. Introduction

Nowadays, with the rapid and sustainable development of energy, people have gradually shifted their attention from traditional fossil energy to new clean energy, and lithium-ion batteries (LIB) being a form of this [1,2,3,4,5,6,7]. Over the past few decades, LIBs have dominated the portable electronics market due to their much higher energy density compared with other energy storage systems. They have been widely used in traffic applications in hybrid electric vehicles (HEV), plug-in hybrid electric vehicles (PHEV) and electric vehicles (EV) to reduce environmental pollution, and further consideration is being given to store and utilize intermittent renewable energy such as solar and wind energy [8,9,10,11]. Since Padhi et al.’s groundbreaking report in 1997, LiFePO₄ has attracted significant attention and has been widely used as a LIB cathode material because of its high theoretical capacity (170 m·A·h·g⁻¹), high safety, low toxicity and the possibility of good economic benefits [12,13,14,15]. However, further development has been limited due to its low conductivity and slow lithium ion diffusion rate [16,17]. Researchers have made substantial efforts to address this issue, such as cation doping [18], surface coating [19,20,21,22,23,24], morphological control [25] and electrolyte modification [26].

Recently, with the deepening of the understanding of electrode materials, it has been found that the surface structure of electrode materials has an important effect on the electrochemical performance of Li-ion batteries. Carbon coating can efficiently enhance the conductivity of the electrode, improve the surface chemistry of the active substance and protect the electrode from direct contact with the electrolyte, thus increasing the cycle life of the battery. Carbon coating, together with nanotechnology, offers good electrical conductivity and rapid diffusion of lithium ions, which also results in good rate capability. Therefore, carbon coating is an efficient approach to promote the performance of electrode materials for lithium-ion batteries. As a cathode material for lithium-ion batteries, LFPO has a 500-billion market capacity, the third largest market capacity, but it has the key problem of poor electrical conductivity. The addition of a conductive carbon layer can significantly solve this problem; as previously reported [27,28,29,30], more graphene oxide can be added as a conductive carbon layer. For example, Li [31] et al. used graphene-modified LFPO to achieve a capacity of 208 m·A·h·g⁻¹ beyond the theoretical capacity at 0.1 C, and Park [32] et al. used graphene-modified LFPO to achieve a capacity of 171.9 m·A·h·g⁻¹ above theoretical capacity at 0.1 C. However, the production of GO is complex and dangerous, so it is imperative to find a method for the synthesis of conductive carbon layers with good safety that is simple and allows for mass production. Therefore, some researchers have begun exploring direct carbon coating. For example, Cui et al. [33] use CTAB to achieve carbon coating of porous silicon micron-sized particles for lithium battery anodes, Cheng et al. [34] use Bis-GMA to achieve carbon-coated NbO₂ for lithium battery anodes and Guo et al. [35] use PAN to achieve carbon-coated SnO₂ for lithium battery anodes.

Herein, we demonstrate that by using CTAB to obtain three-dimensional carbon-coated LiFePO₄, the performance of the electrode materials can be significantly enhanced. The carbon layer at the surface of LiFePO₄ nanoparticles was obtained from a simple water bath and sintering. When used as the electrode material of LIBs, during the charge/discharge process, the carbon layer can effectively inhibit the volume expansion of LiFePO₄ nanoparticles in the process of insertion/de-insertion [36]. A thin layer of carbon covering the surface of LiFePO₄ can remarkably improve the conductivity of the electrode and speed up ion and electron transfer rates. In the meantime, the surface active sites of LiFePO₄ nanoparticles were increased by the covering of the carbon layer, thus increasing the capacity even closer to 170 m·A·h·g⁻¹. The electrode shows excellent reversible capacity as high as 159.8 m A h g⁻¹ at 2 C, good rate capability with 97.3 m·A·h·g⁻¹ at 5 C and prominent cycle ability, preserving ~ 84.2% capacity after 500 cycles. As shown in Table S1 [27,30,37,38,39,40,41], we compared the capacity of our 3D carbon-modified LFPO with that of other published materials, and it is clear that our material performs an excellent capacity at a relatively high current density, while the synthesis method is relatively simple. The three-dimensional carbon-coated LiFePO₄ is regarded as a promising electrode material for LIBs.

2. Materials and Methods

2.1. Materials Synthesis

LiFePO₄ and CTAB were purchased from Sinopharm Chemical Reagent Co. Ltd., (Shanghai, China) at analytical grade. All chemicals were used as received without further treatment. For a representative synthesis, 5 g LiFePO₄ and 0.75 g CTAB were added into 100 mL of H₂O at 80 °C under vigorous stirring for 8 h. After reaction, the precipitates were collected and cleaned thoroughly, followed by drying in a ventilated drying oven at 80 °C for 24 h. Finally, the obtained product was further annealed at 500 or 800 °C for 3 h in N₂ with a ramping rate of 2 °C·min⁻¹ and furnace-cooled to room temperature to obtain the target samples. We defined LiFePO₄ without CTAB as LFPO, added 0.75 g (mass fraction 15%) of CTAB at sintering temperature of 500 °C as LFPO_15%_500 °C and added 0.75g (mass fraction 15%) of CTAB at sintering temperature of 800 °C, defined as LFPO_15%_800 °C. We also prepared LFPO_5%_500 °C and LFPO_5%_800 °C, LFPO_10%_500 °C and LFPO_10%_800 °C and LFPO_20%_500 °C and LFPO_20%_800 °C in the same way. (In order to avoid ambiguity, we have explained the definitions of various abbreviations in Tables S3 and S4.)

2.2. Characterization

X-ray diffraction (XRD) data were collected using Cu Kα radiation (λ = 1.5418 Å) in the 2θ (5°–90°) at room temperature by a Bruker D8 Discover X-ray diffractometer (Bruker, Karlsruhe, Germany). Field-emission scanning electron microscopy (FESEM) and energy dispersive X-ray spectra (EDS) were recorded by JEOL-7100F (JEOL, Tokyo, Japan). Brurauer–Emmerr–Teller (BET) surface areas were tested with a micromeritics TriStar II Surface Area (Micromeritics, Norcross, GA, USA) and Porosity by adsorption of nitrogen at 80 °C. Transmission electron microscopy (TEM) and high-resolution transmission electron microscopy (HRTEM) images associated with selected area electron diffraction (SAED) were collected by a JEM-2100F microscope (JEM, Tokyo, Japan).

2.3. Electrochemical Measurements

The 2016 coin batteries were assembled in a glove box filled with argon. In lithium half batteries, lithium metal was used as the anode, 1 M solution of LiPF₆ in ethylene carbonate (EC)/dimethyl carbonate (DMC) (EC:DMC = 1:1, w%) was used as the electrolyte and glass fiber (GF/A) from Whatman was used as the separator. The working electrode was prepared by mixing the as-synthesized samples with carbon black and polyvinylidene fluoride (PVDF) in a weight ratio of 70:20:10. After coating it with aluminum foil, the electrode film was uniformly cut into circular slices over an area of ~0.4 cm² and a mass loading of 1.0–2.5 mg·cm⁻². Galvanostatic charge/discharge measurements were performed at a potential window ranging from 2.5–4.2 V (vs. Li⁺/Li) using a multi-channel battery testing system by LAND CT3001A (LAND, Wuhan, China). The test current is represented by C, 1 C = 170 mA·g⁻¹. Cycle voltammetry (CV) and Electrochemical Impedance Spectra (EIS) were performed using an electrochemical workstation by CHI 760E (CHI, Shanghai, China). The EIS test frequency range was 0.01–100 kHZ, amplitude was 5 mV and the initial voltage was the open-circuit voltage of the battery. All tests were performed at room temperature. We show more detailed test standard parameters in Table S5 and in order to demonstrate the electrochemical measurements more clearly, we show the physical picture of the test in Figure S10.

3. Results and Discussion

Figure 1a depicts the X-ray diffraction patterns (XRD) of LFPO, LFPO_15%_500 °C and LFPO_15%_800 °C. Sharp diffraction peaks could be consistent in patterns corresponding to highly crystalline structures. All the main peaks could be well-matched to the monoclinic layered structure LiFePO₄ (JCPDS No. 01-081-1173); the lattice parameters are refined to be a = 10.3320 Å, b = 6.0100 Å and c = 4.6920 Å with α = β = γ = 90°, corresponding with a space group of P62. When the pure-LFPO nanoparticles were coated with carbon which was sintered by CTAB, the structure of the nanoparticles was preserved with a peak belonging to LFPO. These results demonstrate that carbon sintered by CTAB not only does not change the crystalline phase of LFPO, but also provides a protective layer for LFPO nanoparticles to inhibit volume expansion and improve electrical conductivity.

The LFPO crystal has a layered structure consisting of FeO₆ octahedra and PO₄ tetrahedra and an oxygen atom point-shared with neighboring PO₄ tetrahedra and FeO₆ octahedral, as shown in Figure 1b. The lithium atoms are sandwiched between these FeO₆ octahedra and PO₄ tetrahedra. The O atoms in LFPO are in a single configuration which is surrounded by 1 Fe and 1 P. In some lithium-based oxide cathode materials, one of the reasons for structural degradation is the movement of metal ions from the octahedral to a tetrahedral site. Since the thin carbon layer can protect the LFPO, the volume of carbon-coated LFPO is probably free from the migration of the Li ion.

We characterized the individual SEM images of the pure LFPO, LFPO_15%_500 °C and LFPO_15%_800 °C (Figure 2a–c). For the pure LFPO, the morphology is of regular shape with a particle size of 100–300 nm (Figure 2a), and LFPO nanoparticles have a fairly smooth surface. As depicted in Figure 2b,c, when the annealing temperature is 500 °C, the surface of particles began to become rough and was covered with a small amount of floccule (Figure 2b). When increasing the annealing temperature to 800 °C, it is clear that the surface of particles became rougher and was covered in a great deal of flocculent substance (Figure 2c). It can be seen that when the CTAB doping amount is 15%, the floccule on the surface of LFPO nanoparticles increases with the increase of the sintering temperature. In order to verify whether the flocculent substance is carbon, the EDS test was performed on LFPO_15%_800 °C, as shown in Figure 2d–f. Observation from the EDS image shows that the surface of LFPO nanoparticles is covered with a layer of carbon, which can be seen from the brightness of EDS as a carbon layer and their even distribution on the surface of the LFPO. At the same time, EDS also shows the uniform distribution of P, O and Fe (Li cannot be detected due to its relatively small molecular mass), indicating that the addition of a carbon layer does not impact the structure of LFPO nanoparticles. As can be seen from Figure 2g–i, N₂ adsorption–desorption measurements indicated that LFPO_15%_800 °C has a BET surface area of ∼21.0 m²·g⁻¹ (Figure 2i) bigger than that of LFPO (~16.0 m²·g⁻¹, Figure 2g) and LFPO_15%_500 °C (~17.2 m²·g⁻¹, Figure 2h). BET surface area increases with the rise of temperature. The test results of BET correspond well with SEM and EDS. With the increase of the carbon layer, the BET surface area does increase, indicating that the carbon layer coating the LFPO has a significant effect. As shown in Figure S1 from the BET surface areas of LFPO_5%_500 °C (~16.0 m²·g⁻¹, Figure S1a) and LFPO_5%_800 °C (~15.2 m²·g⁻¹, Figure S1d), LFPO_10%_500 °C (~16.4 m²·g⁻¹, Figure S1b) and LFPO_10%_800 °C (~16.1 m²·g⁻¹, Figure S1e), LFPO_20%_500 °C (~15.4 m²·g⁻¹, Figure S1c) and LFPO_20%_800 °C (~18.8 m²·g⁻¹, Figure S1f), it is not hard to see that the doped CTAB and the sintering temperature must be suitable to increase the BET surface area and thus the capacity. At the same time, TEM, HRTEM and corresponding SAED tests were also performed to further observe the carbon layer on the surface of LFPO nanoparticles. As shown in Figure 3a–c and Figure S2a, the surface of LFPO nanoparticles is surrounded by a thin layer of carbon flocculation, and the thickness gradually increases with the increase of temperature. Through HRTEM of LFPO, LFPO_15%_500 °C and LFPO_15%_800 °C in Figure 3d–f and Figure S2b, it can be further observed that through three-dimensional carbon coating, the carbon layer on the surface of LFPO nanoparticles increases from 0 to 5 nm and then to 8 nm, indicating that the carbon layer covering is indeed realized. The corresponding selected area electron diffraction (SAED) patterns in the illustrations in Figure 3d,f represent the single crystallinity of LFPO and that carbon coating does not change the lattice orientation of LFPO nanoparticles. As shown in Figure S3, it can be clearly seen that with the increase of temperature, the mass loss rate of LFPO increases, indicating that the amount of carbon on the surface of LFPO increases. According to previous reports, [42,43,44,45] at about 400 °C, LiFePO₄ will be oxidized into Li₃Fe₂(PO4)₃ and Fe₂O₃, resulting in a mass increase. Therefore, the TG curve will rise first and then decline, and the declining part represents the content of carbon. According to the results of TEM and TGA, we conclude that the thickness of the carbon layer increases with the increase of the sintering temperature. All these results show that CTAB-modified LFPO can indeed achieve three-dimensional carbon-coated LFPO and improve the BET surface area.

Coin-type cells are assembled to investigate the lithium storage performances. CV curves of LFPO, LFPO_15%_500 °C and LFPO_15%_800 °C are measured at a scan rate of 0.1 mV·s⁻¹ from 2.5 to 4.2 V (vs. Li⁺/Li) at room temperature (Figure 4a). Generally speaking, three samples exhibit semblable CV curves, implying their identical electrochemical behaviors. The CV curves show that the reduction peaks appear at ~3.6 V, corresponding to the reduction of the formation of a solid electrolyte interphase (SEI) layer. One oxidation peak at ~3.3 V can be observed which may be imputed to the oxidation of Li to Li⁺. The areas of CV curves for LFPO_15%_800 °C are larger than LFPO and LFPO_15%_500 °C, indicating the higher capacity of LFPO_15%_800 °C. Figure S4 shows the typical CV curves of LFPO_15%_800 °C at various scan rates from 0.1 to 1.0 mV·s⁻¹ in a voltage range of 2.5 to 4.2 V. The CV curves exhibit a similar shape, demonstrating that the lithiation/de-lithiation processes are highly reversible with few side reactions.

When the current density is 2 C, the first discharging capacity of LFPO_15%_800 °C is 159.8 m·A·h·g⁻¹ with the initial coulombic efficiency of ~75%. Although the initial coulomb efficiency is only 75%, it increases with the charging and discharging process. As shown in Figure S5, good coulomb efficiency is achieved at different C rates. With the increase of the number of cycles, the capacity gradually stabilized and reached 156.9 m·A·h·g⁻¹ after 100 cycles, showing a good cycling capacity (Figure 4b). However, the first discharging capacity of LFPO_15%_500 °C is 121.6 m·A·h·g⁻¹ and reached 111.2 m·A·h·g⁻¹ after 100 cycles; the first discharge capacity of LFPO is even lower, being only 103.9 m·A·h·g⁻¹, and reached 100.0 m·A·h·g⁻¹ after 100 cycles. The high electrode capacity of LFPO_15%_800 °C is probably because of (i) the carbon layer covering the surface of LFPO nanoparticles, which enhances the integral conductivity, and (ii) the carbon layer on the surface of LFPO nanoparticles inhibiting the volume expansion between Li insertion/de-insertion processes [21,27,28,43]. Therefore, the capacity is even closer to 170 m·A·h·g⁻¹. It should be noted that LFPO_15%_500 °C exhibits capacity behavior and coulombic efficiency similar to LFPO_15%_800 °C. This result indicates that carbon coating can also be achieved at 500 °C, which improves the battery capacity and stability. It may be because the carbon layer is closely covered, so the BET area does not increase significantly. This is also confirmed by the charging-specific capacity at 5 C current density (Figure S6); LFPO_15%_500 °C and LFPO_15%_800 °C have a similar specific capacity and are both higher than LFPO. As shown in Figure S7, we also tested the cycling performance of LFPO_5%_500 °C and LFPO_5%_800 °C, LFPO_10%_500 °C and LFPO_10%_800 °C and LFPO_20%_500 °C and LFPO_20%_800 °C at 2 C current densities. Their capacity performance corresponds to the size of the BET surface area. Their capacity corresponds to the BET surface area. The larger the BET surface area, the higher the specific capacity. It can be seen that the BET surface area does not increase with the increase of sintering temperature [46,47], nor is a higher CTAB doping amount [48] better. LFPO_10%_500 °C and LFPO_20%_800 °C are higher than LFPO, LFPO_5%_500 °C and LFPO_10%_800 °C are close to LFPO and LFPO_5%_800 °C and LFPO_20%_500 °C are lower than LFPO. This also indicates that only can the sintering temperature and the CTAB doping amount’s effective collocation have a good and specific capacity gain effect. Figure 4c shows the charging and discharging voltage platform of LFPO_15%_800 °C, with the discharging platform at ~3.3 V and the charging platform at ~3.5 V. Moreover, each cycle platform is similar and each circulation platform corresponds well to the redox peak in Figure 4a, indicating its good cycling stability. The same is true for LFPO and LFPO_15%_800 °C in the case of 2 and 5 C (Figures S8 and S9).

The average charge capacities of 164, 150, 151, 138, 125 and 97 m·A·h·g⁻¹ are obtained for LFPO_15%_800 °C at rates of 0.1, 0.2, 0.5, 1, 2 and 5 C, respectively (Figure 4d). Remarkably, when the current density returns to 0.1 C, it also indicates a capacity of up to 169 m·A·h·g⁻¹. This performance demonstrates the excellent high-rate capability and outstanding cyclability of the LFPO_15%_800 °C. Besides superior specific capacity and good rate capability, LFPO_15%_800 °C and LFPO_15%_500 °C also show excellent cycling performance (Figure 4e). After 500 cycles, LFPO_15%_800 °C maintains a good capacity of 100.5 m·A·h·g⁻¹ at a current density of 5 C and maintains ~84.2% of the first cycle discharge capacity, corresponding to a capacity decay of 0.0316% at each cycle. All these results show that LFPO nanoparticles can bind tightly to the carbon layer and greatly improve the overall cycling stability. This is because LFPO nanoparticles are uniformly fixed and firmly connected to the CTAB sintered nanofibers, resulting in a process of charge and discharge and an adaptive three-dimensional grid structure with strain relaxation ability formed to stabilize the charge and discharge cycles.

In Figure 5a, we can clearly see that a modified LFPO battery can light up 51 LED lights with CUG in the row. It is safe to say that the modified LFPO battery maintained high power, high capacity and high stability well during the charging/discharge process. Nyquist plots (Figure 5b) present a semicircle and a quasi-straight line, which are associated with the charge transfer resistance (R_ct) and the impedance of Li⁺ diffusion in solid materials (Warburg impedance, Z_w), respectively. The specific fitting data of EIS is presented in Table S2; the R_ct of LFPO_15%_500 °C and LFPO_15%_800 °C are lower than those for LFPO and the Z_w of LFPO_15%_500 °C and LFPO_15%_800 °C are higher than those for LFPO, indicating that the electronic transmission capability of the LFPO_15%_500 °C and LFPO_15%_800 °C electrodes are superior to LFPO. Based on the data obtained from the EIS test, the diffusion coefficient values of the lithium ions (D) can be calculated using the formula D = 0.5 (RT/AF²σ_wC)², where R is the gas constant, T is the temperature, A is the area of the electrode surface, F is Faraday’s constant, σ_w is the Warburg factor and C is the molar concentration of Li ions [49]. The calculated lithium diffusion coefficient value for LFPO is 5.9 × 10⁻¹³ cm²·s⁻¹, LFPO_15%_500 °C is 8.9 × 10⁻¹³ cm²·s⁻¹ and LFPO_15%_800 °C is 3.36 × 10⁻¹² cm²·s⁻¹. It is obvious that the lithium diffusion coefficient values of LFPO_15%_500 °C and LFPO_15%_800 °C are much higher than that of LFPO. The results show that the three-dimensional carbon-coated LFPO can improve the lithium ion diffusion rate, which can improve the specific capacity of LFPO.

4. Conclusions

In this work, three-dimensional carbon-coated LFPO was successfully fabricated through a facile water bath and a calcination method. The LFPO and CTAB formed a thin carbon layer on the LFPO surface by N₂ sintering after the water bath, thereby enhancing the conductivity of the electrode. The electrochemical mechanisms of LFPO and modified LFPO were investigated. As a cathode material for LIBs, the modified LFPO exhibited excellent specific capacity (159.8 m·A·h·g⁻¹ at 2 C), extraordinary rate performance (97.3 m·A·h·g⁻¹ up to 5 C) and ultralong cycling stability (~84.2% maintained even after 500 cycles at 5 C). Such mesmerizing performances are ascribed to (i) the carbon layer covering the surface of LFPO nanoparticles, which enhances the integral conductivity, and (ii) the carbon layer on the surface of LFPO nanoparticles inhibiting the volume expansion between Li insertion/de-insertion processes. It can be seen that LFPO cathode can achieve good initial capacity and cycle life by carbon coating, which is cost-effective and easy to realize. This work suggests that using a carbon layer covering to modify electrode material is a promising strategy for obtaining high comprehensive performance electrode materials in the field of lithium cathode materials.