Introduction

The development of advanced secondary battery with high specific energy density is an urgent demand in the application fields of portable electronic products, electric vehicles, and large-scale energy storage. Lithium-ion batteries (LIBs) own the most probable prospect as the alternate candidate in the above-mentioned areas because of their high energy, the high power density, and the superior cycling stability [1,2,3,4,5,6]. Currently, the actual specific capacity of graphite anode material used in commercial LIBs is close to its theoretical value of 372 mAh g−1 [7,8,9], and the specific capacity of the next generation of alloy materials may even exceed 1000 mAh g−1, such as Si-based anode materials [10, 11] and various materials using metal-organic frameworks (MOF) as precursors [12,13,14,15,16]. However, the corresponding commercial cathode materials have low specific capacity, such as LiCoO2 (140–150 mAh g−1) [17,18,19], LiFePO4 (150–160 mAh g−1) [20,21,22], and LiMn2O4 (110–130 mAh g−1) [23,24,25]. So, we can see that LIBs suffer from the low specific capacities of cathode materials. Therefore, the development of high capacity cathode materials has become the key technology for upgrading the LIBs.

Monoclinic Li3V2(PO4)3 (LVP) has attracted much attention due to its high theoretical capacity (a theoretical capacity of 132 mAh g−1 can be obtained in the range of 3.0–4.3 V, and the theoretical capacity can be as high as 197 mAh g−1 in the range of 3.0–4.8 V), low synthesis cost, stable structure, and simple synthesis method [26, 27]. However, the main problem of this material is its low electronic conductivity, which is only about 2.4 × 10−7 S cm−1 at room temperature. Previous studies have shown that carbon coating or compositing can overcome the above disadvantages of LVP [28,29,30,31]. However, this is still unable to meet the demand in the field of high voltage, high power, and long working time power battery, which urges researchers to seek more effective methods to further improve the high rate performance and cycle life of LVP.

In recent years, the use of N-doped carbon materials to modify lithium-ion electrode materials has attracted extensive attention. This kind of carbon materials are mainly focused on graphite, graphene, and carbon nanotubes [32,33,34,35,36]. The principle of nitrogen-doped carbon coating modification is mainly reflected in the following aspects: (1) N-doping can modify the structure, chemical activity, and electronic properties of the materials; (2) N-doping can produce many extrinsic defects and active sites; and (3) nitrogen/carbon co modification can improve the stability of SEI film [37,38,39]. As far as we know, there are also reports that N-doped carbon materials, such as N-doped graphite and graphene, are used to modify LVP to improve its electrochemical performance. However, these N-doped carbon materials need complex nitrogen doping modification process, and the efficiency is also low. Based on this, in this work, we designed a simple rheological phase method, using melamine as nitrogen source and citric acid as carbon source to synthesize nitrogen carbon co coated LVP composites, and studied the effect of N-doping on the phase structure, morphology, and electrochemical properties of Li3V2(PO4)3/C (LVPC).

Experimental

LiOH·H2O, V2O5, NH4H2PO4, citric acid, and melamine with a stoichiometric ratio of 3.1:1:3:2-x:x (x = 0, 0.25, 0.5, 0.75, and 1) (by molar) were accurately weighed as lithium, vanadium, phosphorus, carbon, and nitrogen sources. Put the above raw materials in a dry agate mortar, fully grind the raw materials to fine powder under an infrared lamp, and mix them evenly. Then the raw materials were transferred to the PVC lining, and deionized water was added dropwise in the lining to adjust the powder raw materials to the rheological phase state. Then the lining was put into the reactor, tightly sealed, and put into the oven at 80 °C for 6 h. After that, the reactor was taken out for natural cooling and obtained bright blue-green viscous slurry and then dried in a drying oven at 100 °C for 12 h, and the blue-green honeycomb N-doped LVPC precursor was obtained. The dried precursor was taken out and placed in an agate mortar for full grinding, then transferred to a strip crucible, and placed in a high-temperature tubular furnace. It was pretreated at 350 °C for 3 h in Ar atmosphere, cooled and grinded again, and then sintered at 850 °C for 8 h in the tubular furnace to obtain the N-doped LVPC. According to the increasing order of doping nitrogen content, the synthesized products were recorded as LVPC/N-1, LVPC/N-2, LVPC/N-3, and LVPC/N-4, respectively. The specific synthetic routes are shown in Scheme 1. The synthesis process of LVPC is the same as that of N-doped LVPC, except that the reactant melamine was replaced with citric acid.

Scheme 1
scheme 1

Schematic illustration for formation process of LVPC/N

Full size image

X-ray diffraction (XRD) measurements were carried on Rigaku MiniFlex 600 using Cu Kα radiation at λ = 1.54 Å. Scanning electron microscopy (SEM) and transmission electron microscope (TEM) were performed on FEI-Quanta 250 FEG and JEM-2100F, respectively. The obtained LVPC/N composite was mixed with super carbon black and poly tetra fluoroethylene (PTFE) microemulsion in a mass ratio of 8:1:1, and the mixture was dispersed in isopropanol, pressed into film, and dried at 80 °C for 10 h in a vacuum oven. Then the dried film was cut into a disk with a diameter of 10 mm and pasted on Al net. Electrochemical tests were carried using CR2016 coin-typed cells and the above electrode as cathode and Li metal as a counter electrode. Commercial polyethylene film and 1 mol L−1 LiPF6 in ethylene carbonate and dimethyl carbonate (1:1 v/v) were employed as separator and electrolyte, respectively. Galvanostatic charge/discharge tests were conducted on the CT 2001A battery tester (LAND) in the voltage range of 3.0–4.8 V (vs. Li+/Li). Cyclic voltammetry (CV, 3.0–4.8 V (vs. Li+/Li)) measurements and electrochemical impedance spectroscopy (EIS) (the frequency range is between 100 kHz and 10 mHz, potential amplitude: ± 5 mV) were tested on CHI 660E electrochemical workstation.

Results and discussion

Figure 1 shows the XRD spectra of pure carbon-coated LVP and carbon nitrogen co coated LVP materials. It can be seen from the figure that all LVPC/N samples have sharp diffraction peaks and no impurity peaks compared with LVPC sample. All of them belong to monoclinic LVP indexed to a space group of P21/n (JCPDS No. 72-7074). Moreover, all samples have no other characteristic peaks, and the peaks of N-doped carbon are not found, indicating that the N-doped carbon is amorphous or mainly ascribed to its low content and the strong intensity of LVP peaks. It also states that N-doped carbon will not affect the crystal structure of the LVP. According to Raman spectra (Fig. 1b), the intensity ratio of D to G band (ID/IG) could estimate the defects in the graphitic carbon structure [40]. The ID/IG of LVPC/N-3 (0.96) is higher than that of LVPC (0.90), due to more defects structure caused by the incorporation of N atoms into carbon lattice.

Fig. 1
figure 1

(a) XRD patterns of LVPC and LVPC/N composites, (b) Raman spectra of LVPC and LVPC/N-3, (c) XPS surveys and (d) high resolution C 1s spectra of LVPC and LVPC/N-3, (e) XPS spectrum of N 1s for LVPC/N-3, and (f) the schematic illustration of N-doped carbon for LVPC/N composite

Full size image

Moreover, XPS was performed to reveal the surface bonding information of the samples. In Fig. 1c, the signals of O, V, N, C, P, and Li elements are detected from the survey spectrum of LVPC/N, whereas there is no signal of N element in LVPC, which demonstrates that the N atoms have been successfully doped into LVPC/N composite and detected as 2.03 at.%. As displayed in Fig. 2d, the C 1s spectra of two samples can be resolved into three peaks located at 284.6, 285.6, and 288.8 eV, corresponding to the C–C, C–N/C–O, and C=O bond, respectively [41]. By comparison of the area percentage of C–C in total C, it is clearly seen that LVPC/N-3 shows a lower value of 55.6% than that of LVPC (60.6%), indicating more disorder carbon structure by N-doping. And in Fig. 1e, the characteristic N 1s spectrum of LVPC/N-3 can be divided into three N structures including pyridinic N (398.1 eV), pyrrolic N (399.6 eV), and graphitic N (401.2 eV) [42]. The schematic illustration of N-doped carbon structure was depicted as in Fig. 1f. With the well-designed structure, it could bring the mass vacant sites for Li-ion storage and boost the improvement of electron conductivity.

Fig. 2
figure 2

SEM images of LVPC (a) and LVPC/N-3 composite (b, c), TEM image of LVPC/N-3 composite (d), energy-dispersive spectrometer (EDS) of LVPC (e), and LVPC/N-3 composite (f)

Full size image

The morphology of LVPC and LVPC/N-3 composite was investigated by SEM in Fig. 2a and b. As displayed in Fig. 2a, the LVPC consists of plate-like particles with about 2–5-μm diameters. Through Fig. 2b, it is easy to see that LVPC/N-3 composite shows a stacked bulk-like morphology with reduced sizes. Interestingly, there are uniform strip grains owing to the influences of N-doped carbon (Fig. 2c). It can be clearly seen from Fig. 2d that there is uniform N-doped carbon coating on the surface of LVP material, and the thickness is about 20 nm, which is in favor of the Li+ ions insertion/extraction, bringing about the better electrochemical properties. EDS results in Fig. 2e described the existence of N compared to Fig. 2f, which more demonstrated that N-doped carbon is successfully composited with LVP. Table 1 illustrates the At % ratio of LVPC and LVPC/N-3 composite, further confirming the successful preparation of N-doped carbon-coated LVP.

Table 1 The element content ratio of LVPC and LVPC/N-3 composite
Full size table

Figure 3a shows the charge-discharge curves of LVPC and four LVPC/N composite electrodes in the second cycle at a voltage range of 3.0–4.8 V and a current density of 30 mA g−1. As shown in the figure, the charge-discharge curves of all LVPC/N composites and LVPC are similar. The charging curves display four flat platforms, and the discharge curves display three platforms, which correspond to the phase change in the charge and discharge process, respectively, which are consistent with the literature reports [34, 43]. It shows that N-doped C does not affect the charge-discharge process behavior of the LVPC. The discharge capacities of LVPC, LVPC/N-1, LVPC/N-2, LVPC/N-3, and LVPC/N-4 in the second cycle are 163.5, 157.6, 179.5, 193.1, and 146.3 mAh g−1, respectively. Among them, LVPC/N-3 electrode has the highest discharge specific capacity, which is obviously improved compared with LVPC material, and the best ratio of citric acid to melamine is 1.25:0.75 (5:3). In addition, compared with LVPC, the charge-discharge voltage plateau of LVPC/N-2 and LVPC/N-3 is significantly longer, and the gap between charge and discharge platform is smaller, which indicates that the two electrode materials have higher capacity and better reversibility.

Fig. 3
figure 3

Initial discharge curves (a) and cycle performances (b) of LVPC and four LVPC/N composites and rate capabilities (c) and CV curves (d) of LVPC and LVPC/N-3 composite

Full size image

Figure 3b illustrates the cycle performance of LVPC and four LVPC/N composite electrodes in the same conditions. The initial discharge specific capacities of LVPC, LVPC/N-1, LVPC/N-2, LVPC/N-3, and LVPC/N-4 are 165.7, 160, 182.2, 195.0, and 149.7 mAh g−1, respectively. After 100 cycles, the specific capacities and retention rates are 129.7 (78.3%), 125.5 (78.4%), 143.3 (78.6%), 165.7 (85.0%), and 104.7 mAh g−1 (70.0%), respectively. It is found that the discharge capacities and cycle performance of LVPC/N-1 and LVPC are almost the same; LVPC/N-2 and LVPC/N-3 materials have obvious improvement; especially LVPC/N-3 has the highest discharge capacity and the best cycle performance. However, the electrochemical performance of LVPC/N-4 material is poor, which may be due to the high proportion of melamine in the raw material, resulting in the thickness of the coating layer too low and the conductivity of the material poor. The results show that the amount of N-doped carbon should be suitable for the electrochemical performance of LVP.

In order to test the electrochemical performance of LVPC/N-3 material under different current density, the rate performance under continuous step current density compared with LVPC was carried out. The test results are shown in Fig. 3c. It can be seen from the figure that the discharge capacities of the materials are decreasing with the increase of current density, because the higher the current density is, the faster the charge-discharge is, and the greater the electrode polarization is. However, the discharge capacity of LVPC/N-3 is always higher than that of LVPC. And after 40 charging and discharging cycles at different current densities, the discharge capacity of LVPC/N-3 can still reach 165 mAh g−1 and the retention rate is 86%, while the capacity of LVPC material is only 132 mAh g−1 (80%), which indicates that N-doped carbon is more beneficial to the rate performance of LVP.

The CV curves of LVPC/N-3 and LVPC electrode materials at a scanning rate of 0.1 mV s−1 in the voltage range of 3.0–4.8 V are shown in Fig. 3d. From the figure, we can see that there are four oxidation peaks in the CV curves of the two materials, which are 3.64 V, 3.70 V, 4.13 V, and 4.56 V, respectively, which corresponds to the phase transition of Li3V2(PO4)3←→Li2.5V2(PO4)3, Li2.5V2(PO4)3←→Li2V2(PO4)3, Li2V2(PO4)3←→Li1V2(PO4)3, and Li1V2(PO4)3←→V2(PO4)3. The three reduction peaks are 3.54 V, 3.64 V, and 4.01 V, respectively, and the reduction peak at 4.01 V represents the behavior of solid solution. The obvious difference is that the redox peak area of LVPC/N-3 is larger, which indicates that LVPC/N-3 has higher capacity. CV test results are consistent with constant current charge-discharge test results.

Figure 4 shows the EIS curves and equivalent circuit fitting diagram of LVPC and LVPC/N-3 electrodes after 1, 20, 50, and 100 charge and discharge cycles at 30 mA g−1 in the voltage range of 3.0–4.8 V. It can be seen from the figures that the Nyquist curve of LVPC is composed of a semicircle and a straight line, while the Nyquist curve of LVPC/N-3 consists of two flattened semicircles and a straight line, forming an obvious solid electrolyte interface (SEI). The fitting impedance value and calculated Li+ ions diffusion coefficient (DLi+) results are shown in Table 2. The DLi+ can be calculated by the following equations: Z′ = Re + Rf + Rct + σω-1/2 and DLi+ = 0.5 R2T2/ A2n4F4C2σ2 [44,45,46]. The comparison shows that the Rct of the two materials is similar, which means that the charge transfer rate of lithium ions on the surface of the material and the internal diffusion rate of the material are the same. However, the Rf of LVPC/N-3 is significantly larger, and the DLi+ should also increase, which means that the Li+ transfer process of LVPC/N-3 is changed during the charging and discharging process. The reason for this phenomenon is probably due to the N-doped carbon, which easily reacts with electrolyte to form SEI. SEI film is insoluble in organic solvents and can be stably stored in organic electrolyte. Lithium ions can pass through the passivation film freely, but solvent molecules cannot pass through it, which can effectively prevent the destruction of electrode materials caused by co-embedding of solvent molecules, thus greatly improving the cycle performance and service life of the electrode. Therefore, we speculate that N-doping does not change the intrinsic conductivity of LVPC but improves the lithium-ion migration rate and material stability by changing the coating layer; thus, improving the electrochemical performance of LVPC and LVPC/N-3 shows better rate performance and electrochemical cycle stability.

Fig. 4
figure 4

The Nyquist plots of the LVPC (a) and LVPC/N-3 (c) and the relationship between Z′ and ω−1/2 of LVPC (b) and LVPC/N-3 (d) in the low-frequency region

Full size image
Table 2 EIS equivalent circuit fitting parameters and the values of DLi+ for the LVPC and LVPC/N-3 composite at different cycle
Full size table

Conclusion

N-doped carbon-coated LVP composites were synthesized by one-step rheological phase method using melamine as nitrogen source and citric acid as carbon source. The effects of different ratios of melamine and citric acid on the structure and electrochemical properties of LVP were studied. The XRD results showed that the N-doped carbon did not affect the crystal structure of LVPC, and the electrochemical performance test results showed that when the ratio of citric acid to melamine was 5:3, the properties of LVPC/N-3 were the best. The initial discharge capacity of LVPC/N-3 is 195 mAh g−1 (LVPC, 165 mAh g−1) at 30 mA g−1 in the voltage range of 3.0–4.8 V, and the capacity can still be maintained at 165 mAh g−1(85%) (LVPC: 130 mAh g−1, 78%) after 100 cycles. The effect of nitrogen doping on the LVPC was further explained from EIS. We found that the thickness and properties of the coating can be changed by adding melamine. On the one hand, N-doping can improve the conductivity of the coating. Even if the thickness of the coating is smaller, the conductivity of the material does not decrease. At the same time, the thinner N-doped carbon layer increases the tap density and the specific capacity of the material. On the other hand, N-doped carbon layer is conducive to the formation of SEI film, which can stabilize the electrode material and improve its cycle performance. Therefore, we conclude that this method can effectively improve the structure and electrochemical properties of LVP and can also be used for the modification of other electrode materials.