Introduction

Lithium-ion batteries (LIBs) employed in hybrid electric vehicle energy storage device have been attracting great attention for their light weight, high safety, high energy, and long cycle durability without a memory effect [1, 2]. To develop desired electrode materials with high performance including excellent rate capability, high reversible discharge capacity and cycle stability are crucial for LIBs, especially for cathode materials [2, 3]. In recent decades, in order to achieve the increasing demand performance, various kinds of cathode materials have been researched and developed, such as lithium transition-metal phosphates. Due to the good structural stability, high theoretical specific capacity, and operating voltage [4, 5], lithium transition-metal phosphates have attracted significant attention in LIBs with high performance. Among the transition-metal phosphate compounds, Li3V2(PO4)3 (monoclinic) is the most prospective material ascribed to its acceptable ionic mobility, high theoretical specific capacity, operating voltage, and thermodynamically stable structure [6, 7]. However, the separated [VO6] octahedral arrangements result in the poor electronic conductivity of Li3V2(PO4)3 (2.4–10−7 S cm−1), which indeed limits its extensive application [8, 9]. Great efforts have been devoted to resolve those problems, such as coating Li3V2(PO4)3 with a carbon layer [10, 11], reducing its particle size [12, 13] and cation doping [14, 15]. Among them, cation doping is a facility and effective method to enhance the electrochemical properties of Li3V2(PO4)3 [16]. Up to now, a great number of cations, such as Zr3+, Mn2+, Al3+, Cu2+, Mg2+, Fe3+, Ge4+, and Cr3+ have been employed as dopants for Li3V2(PO4)3 and achieved some good results [14,15,16,17,18,19,20].

Previously, several methods have been adopted to synthesize cation doping Li3V2(PO4)3, which include hydrothermal technique, spray pyrolysis method, solid-state strategy, and sol-gel approach. For example, Park et al. [19] have prepared Mn-doping Li3V2(PO4)3 by the spray pyrolysis method. The electric conductivity of Li3V2(PO4)3 has been increased ascribed to the Mn doping, and the cathode electrode performances have been improved. Liu et al. [20] used two methods (the first step is the sol-gel method) to fabricate Fe-doping Li3V2(PO4)3 with improvement of electrochemical performance. However, these methods usually consist of redundant steps, such as a long time for reaction, washing several times, and calcination at various stages to prepare the final product. Thus, it is valuable to develop a facile synthesis route with low cost and efficiency for practical application. Up to now, we find no literature report on the preparation of Fe-doped LVP by combustion synthesis. In this work, we offer a facile route for the synthesis of Fe-doping Li3V2(PO4)3 using solution combustion synthesis. Compared with conventional synthesis methods, this method (solution combustion synthesis) is effective and economic due to time-saving processing and economic and simple experimental set-up [21,22,23]. Electrochemical measurement results imply the as-prepared Fe-doping Li3V2(PO4)3 is a permission candidated cathode material for application in lithium-ion batteries with high performance.

Experimental

Synthesis

In detail, 0.03 mol lithium nitrate (LiNO3), 0.02 mol ammonium metavanadate (NH4VO3), 0.02 mol ammonium dihydrogen phosphate (NH4H2PO4,), ferric nitrate (Fe (NO3)3·9H2O), 2 g glucose (C6H12O6·H2O,), and 0.07 mol glycine (NH2CH2COOH) were dissolved in 200 ml deionized water. All the chemicals were analytical reagent grade (Shanghai Zhanyun Chemical Co., Ltd.). The ratio between Fe and Li3V2(PO4)3 (0.02 mol ammonium metavanadate could prepare 0.01 mol Li3V2(PO4)3) was 0~2 (wt%). The whole process for preparation precursors takes 15–20 min. Then, the precursors prepared by combustion synthesis were heated at 800 °C for 5 h with N2 atmosphere, and then Fe-doping Li3V2(PO4)3 could be gotten.

Characterizations

The prepared samples were characterized with X-ray diffraction (XRD, MXP21VAHF) at room temperature, X-ray photoelectron spectra (XPS, PerkinElmer), scanning electron microscopy (SEM), and transmission electron microscopy (TEM).

Electrochemical measurements

The electrodes were fabricated by a mixture of Fe-doping Li3V2(PO4)3 materials (80 wt%), acetylene black (10 wt%), and poly (10 wt%). The compounds were scattered in N-methyl-2-pyrrolidone (NMP) to form sizing agent, depositing on an Al foil. All the fabricated electrodes were stoved at 80 °C for 10 h in a vacuum oven. The half cells (CR2032-type) were fabricated in an Ar-filled glove box. The electrolyte was a non-aqueous solution with 1 M LiPF6 dissolved in a mixture of 1:1:1 of ethylene carbonate (EC)/dimethyl carbonate (DMC)/ethylene methyl carbonate (EMC). The charge and discharge properties of the cells with different current densities were tested in 3.0–4.8 V. For the capacity rate testing, the current gradually changed from 0.5 to 1, 2, 5, and 10 C and decreased to 0.5 C. Cyclic voltammetry (CV) curves were studied with CHI710D (Chenhua, Shanghai) electrochemical workstation at the rate of 0.1 mV s−1.

Results and discussion

Figure 1a demonstrates the XRD results of the as-prepared products. All diffraction peaks can be labeled as the crystalline monoclinic Li3V2(PO4)3 (JCPDS 97-009-6962), and no other peaks could be found. It indicated the Li3V2(PO4)3 has been successfully prepared by reduction combustion synthesis precursors at a temperature of 800 °C for 5 h. Figure 1b shows the chemical information of samples characterized by XPS. According to Fig. 1b, the Li, P, C, V, O, N, and Fe elements could be found. The ~ 55.7 eV band is due the contribution of Li+ in Li3V2(PO4)3 [24]. The ~ 140 eV band is attributed to 2P3/2 in (PO4)3 [25]. The peak at 531.5 eV was the O1s spectrum generated by (PO4)3 [26]. The peak located at 285 eV is corresponding C1s in carbon [27]. The ~ 400 eV peak confirms the N1s in samples [28]. In the high-resolution XPS of V2p for 0 wt% and 1 wt% samples, as shown in Fig.1c, the V2p with a binding energy of around 517 eV ascribed to V2p, which correspond to V3+ in Li3V2(PO4)3 [29]. Additionally, the high-resolution XPS of Fe2p for 0 wt% and 1 wt% samples are shown in Fig.1d. It can be found that new peaks of Fe2p with a binding energy around 711 eV [30] appeared in the Fe-doped sample and no characteristic peak of Fe in 0% sample. The XPS results indicate the Fe-doped Li3V2(PO4)3 has been successfully prepared.

Fig.1
figure 1

a XRD patterns of the Li3V2(PO4)3/C samples with the different Fe contents. b Survey XPS spectra of samples, high-resolution XPS analysis of c V2p and d Fe2p for 0 wt% and 1 wt% samples

Figure 2 illustrates the SEM characterizations of all samples. As illustrated in Fig.2, the 0% sample and 2% sample has a bulk structure and composed of irregular particles. The 1% sample has a porous structure and composed of a sheet structure, which is favorable for lithium ion transmission [31]. By comparison, the Fe-doped Li3V2(PO4)3 have less agglomeration than pure Li3V2(PO4)3, particularly for the 1% sample. This is due to the Fe source (ferric nitrate) which was added in the sample. In this work, it used solution combustion method (SCS) to synthesize precursors. For the SCS, ferric nitrate is the oxidant in the system. When ferric nitrate was induced in the system, the precursors were prepared by the reaction, and the ratio of fuel and oxidant was increased. It could affect the chemical energy and gases liberated in the reaction process [32, 33]. Because of the dispersant effect of more gases on the products [34], the 1% sample has a porous structure. When the Fe content was added to the 2% sample, the surpassing oxidation in the reaction system and the surpassing energy generated in the system, and the particles began to grow and agglomerated. So, the morphology of the product has been changed. Figure 3 exhibits the original images of 1% Fe-doped Li3V2(PO4)3/C powders and the map distribution of V and Fe elements. It is visible in Fig. 3b that the powders have homogeneous distribution of V and Fe elements. Because all raw materials were dissolved in water, all elements in the reaction system uniformly mixed at a molecular or atomic level. And the elements in the samples could distribute homogeneously.

Fig. 2
figure 2

SEM images of the Li3V2(PO4)3/C samples with the different Fe contents. a, b 0%. c, d 1%. e, f 2%

Fig. 3
figure 3

a Original SEM images of 1% Fe-doped Li3V2(PO4)3/C powders and b V and c Fe map distribution

To further observe the structure and morphology of the products, TEM images of pure and Fe-doping Li3V2(PO4)3 products are displayed in Fig. 4. It is clear that pure Li3V2(PO4)3 (Fig. 4a) and 2% Fe-doping Li3V2(PO4)3 (Fig. 4c) have a bulk structure, and the 1% Fe-doped Li3V2(PO4)3 (Fig. 4b) shows a highly porous structure with interconnected macropores. High-resolution transmission electron microscope (HRTEM) investigations of 1% Fe-doped Li3V2(PO4)3 (Fig. 4d–f) demonstrate the modified Li3V2(PO4)3 are covered with amorphous carbon, which can improve the electron conductivity of the products. To study the crystalline nature of Li3V2(PO4)3, electron diffraction of Li3V2(PO4)3 was performed. Figure 4f exhibits the selected electron diffraction pattern, and the distinctive diffraction patterns confirm the highly crystalline nature of the 1% Fe-doped Li3V2(PO4)3. The clear diffraction spots further represent the monoclinic crystal system of Li3V2(PO4)3 phase which is consistent with the XRD results.

Fig. 4
figure 4

TEM images of the Li3V2(PO4)3/C samples with the different Fe contents. a 0%. b 1%. c 2%. df HR-TEM of Li3V2(PO4)3/C with Fe content of 1%

Figure 5 a exhibits the circulation performance of different Fe content samples at 0.1 C in 3.0–4.8 V. It is clear that the discharge capacity of neat Li3V2(PO4)3 after 20 cycles was only 85 mAh g−1. The Fe-doped samples have a higher capacity than pure Li3V2(PO4)3 due to the incorporation of Fe3+ which improved electrical conductivity and structural stability. The 1% sample presents the highest discharge capacity of 180 mAh g−1 and the retention is 91.4%. In comparison with the 1% sample, the 2% sample shows a lower capacity with poor capacity retention. Due to too much Fe3+ substitution in Li3M2(PO4)3 lattice, the closet bond lengths of Li1–O, Li2–O, and Li3–O contacts are significantly reduced. The binding interaction between O and Li gradually increases, reducing the insertion/extraction mobility of lithium ions during discharge/charging [35, 36]. Therefore, with the increase in Fe amount, an excess of Fe3+ in the Li3V2(PO4)3 lattice decreases the electrochemical activities and contributes to the lower electrochemical properties of the electrodes. Figure 5b shows the rate performance of the samples at different rates increasing from 0.5 to 10 C and then reducing to 0.5 C each 5 cycles. It obviously shows that the discharge capacity is decreased as the current rate increases which indicates the electrochemical activities of Li3V2(PO4)3 is determined by ion diffusion. And the 1% sample electrode possesses the capacity of 145, 138, 125, 88, and 143 mAh g−1 at 0.5 C, 1 C, 2 C, 5 C, 10 C, and 0.5 C, respectively.

Fig. 5
figure 5

a The resulting cycling data versus capacity for the samples (at a rate of 0.1 C between 3.0–4.8 V). b Comparison of the cycling performances versus C-rates

EIS measurements can reveal diffusion process information during charge/discharge process [37]. The kinetic properties of undoped and Fe-doped samples were studied by EIS. Figure 6 presents the Nyquist plots of all the three Li3V2(PO4)3 electrodes, and each plot composed of a small intercept at high frequency (corresponding to the ohm resistance of electrolyte (Re)), a depressed semicircle at medium frequency (including the charge transfer resistance between the electrode and electrolyte interface (Rct) and the double-layer capacitance between electrolyte and electrode (Cdl)), and a linear part at low frequency (associated with the Warburg resistance Zw) [38, 39]. It is obvious that the 1% Fe-doped sample shows the lowest Rct. The result is in good agreement with the electrochemical performance obtained in Fig. 5. The improvement is due to the appropriate Fe doped into the Li3V2(PO4)3 crystal lattice which can increase the degree of disorder in the lattice and enhance the electronic transfer conductivity.

Fig. 6
figure 6

Nyquist plots of the two electrodes in the frequency range of 100 kHz to 10 mHz

To further study the electrochemical reaction during cycles, the CV curves of the 1% sample electrode in 3.0–4.8 V at a scanning rate of 0.1 mV s−1 are presented in Fig.7. Because of formation and structure rearrangement of the solid electrolyte interface (SEI) film, the CV curves were tested after five cycles[40-41]. The curves present four oxidation peaks and three reduction peaks. The oxidation peaks located at ~ 3.63 (A1) and 3.72 V (A2) are corresponding to the first Li+ extraction from Li3V2(PO4)3 in two steps ascribed to the ordered phase at mixed V3+/V4+, and the third anodic peak (A3) about 4.13 V is attribute to the second Li+ extraction from Li2V2(PO4)3 [42-44]. The first two Li+ extractions related to the V3+/V4+ couple. The fourth oxidation peak (A4) at about 4.63 V generates by the removal of the third Li+ (LiV2(PO4)3 → V2(PO4)3), which attributed to the V4+/V5+ couple [1]. The border shape with lower densities of peak A4 is ascribed to the lower electronic/ionic conductivity of V2(PO4)3, and, therefore, it has slower reaction kinetics and lower electrochemical activity. Moreover, the peak around B3 is due to the insertion of two Li+ into V2(PO4)3 by a solid solution process, which is ascribed to the V5+/V4+ couple. The reduction peaks located at B2 and B1 are ascribed to a two-step insertion of the first Li+ transform Li2V2(PO4)3 to Li2.5V2(PO4)3 and then to Li3V2(PO4)3, and it corresponds to the V4+/V3+ pairs. To study the high current rate and circulation performance of the 1% sample, the electrodes in 3.0–4.8 V at different current densities for 500 cycles are shown in Fig. 7b. At the current of 1 C, 10 C, and 20 C, the initial special capacity is 175.7, 142.0, and 67.0 mAh g−1, respectively, and the capacity was still 142.5, 132.5, and 65.9 mAh g−1 after 500 cycles, respectively. The results indicate that the material has good circulation performance. The capacity of the cathode decreases with current rate increasing. Due to the low current density, the polarization of the electrode is also small, and the charging and discharging time is long enough for Li+ to be inserted/extracted, so the specific capacity is higher. However, when the current density is too high, the charging and discharging time is too short, which leads to the incomplete insertion/extraction for Li+, so the specific capacity will decrease accordingly. In general, the circulation performance of Fe-doped Li2V2(PO4)3 has been greatly improved.

Fig. 7
figure 7

a CV curves of 1% Fe-doped LVP/C electrodes in the potential range of 3.0–4.8 V (vs. Li+/Li) at a scan rate of 0.1 mV/s. b The resulting cycling data for the samples at different rates

Conclusions

Fe-doping Li3V2(PO4)3/C materials were successfully synthesized from combustion synthesis precursors. Because appropriate Fe doped into the Li3V2(PO4)3 crystal lattice which could increase the degree of disorder in lattice and enhance the electronic transfer conductivity, the 1% Fe-doped composites as cathode electrode material for LIBs exhibit enhanced electrochemical performance. In 3.0~4.8 V, it has a specific discharge capacity of 180 mAh g−1 at the rate of 0.1 C after 20 cycles, 142.5 mAh g−1 at 1 C, and 132.5 mAh g−1 10 C after 500 cycles. Moreover, it shows stabilized specific discharge capacity of 65.9 mAh g−1 after 500 cycles at the rate of 20 C, and the capacity retention is 98%. Thus, it could infer the Fe-doping Li3V2(PO4)3/C material is a potential cathode material for application in LIBs with high performance. Furthermore, this work provides a new insight into the construction of cation doped Li3M2(PO4)3 for improving electrochemical performance.