The phase structures of NH4V3O8 samples with different morphologies were analyzed by XRD. Figure 1 represents the powder XRD pattern of prepared NH4V3O8 samples with different morphologies. From the curves, we can deduce that the diffraction peaks of the four samples can be indexed to a monoclinic crystalline NH4V3O8 phase (JCPDS Card No. 88-1473, space group P21/m). As seen, no other phases have been detected, showing that the products with high purity. For LiV3O8, the higher intensity of diffraction intensity (001), the better the degree of crystallinity[31]. Notably, with the decrease of the ethanol and glycerol, the diffraction intensity (001) peak increases, suggesting a bigger crystallite size of NH4V3O8[35]. Unfortunately, the good ordering of crystal is disadvantageous for the diffusion of Li+ intercalation and de-intercalation.
The morphology and micro-nano structure of the NH4V3O8 were characterized by SEM. Figure 2 shows the SEM images of the as-prepared NH4V3O8. When glycerol, ethanol, and ethylene glycol were used as the solvent, the NH4V3O8 showed nanoparticle morphology with an average size of ~ 150 nm (Fig. 2a, b). When the ratio of the solvent to ethanol and ethylene glycol was 2: 1, the NH4V3O8 showed ultra-small nanoparticles morphology with an average particle size of ~ 40 nm (Fig. 2c, d). Interestingly, when the ratio of the solvent to ethanol and ethylene glycol was 1: 1, the rectangular nanotube morphology formed (Fig. 2e, f). The average size of the tube is about ~ 140 nm. After removed ethylene glycol from the solvent, there are great morphology changes of the NH4V3O8, it changed from nanoscale order of magnitude to hierarchical microsheet (Fig. 2g, h). From the perspective of the crystal, morphology change is because of the oriented crystal plane.
To understand the superiority of the NH4V3O8 rectangular nanotube, the performance of the NH4V3O8 NP, SNP, RNT, and HMS as cathodes in lithium ion battery was investigated. Figure 3 exhibits the cycling performance and rate performance of the as-prepared NH4V3O8. The voltage range is 1.5-4.0 V (vs. Li/Li+). The cycle performance of the electrodes is shown in Fig. 3a. After 50 cycles RNT exhibits a discharge capacity of 189.5 mAh g− 1 at 15 mA g− 1, while the discharge capacity for NP, SNP, and HMS are only 157.8, 154, and 46.8 mAh g− 1, respectively. It shows that the RNT sample has a significantly improved lithium storage capability. We can get the Coulomb Efficiency (CE) of the sample RNT. The CE of RNT has been maintained at high levels, coincides with better cycling performance. Figure 3b shows the rate performance of the different NH4V3O8 samples at a current density from 15 mA g− 1 to 300 mA g− 1. The rate performance of RNT shows higher capacity retention than the other three samples. After 300 cycles the discharge capacity of RNT can revert to 131.1 mAh g− 1, while those for NP, SNP, and HMS are only 85.4, 72.1, and 40 mAh g− 1, respectively. The reason for the good electrochemical performance of NH4V3O8 is the structure of rectangular nanotube which promotes the contact area between NH4V3O8 and electrolyte, enhances the diffusion efficiency. Table 1 is the comparison of the electrochemical performance of as-prepared NH4V3O8 and other different morphology NH4V3O8 reported in the literature. Despite the initial discharge capacity and capacity retention of other NH4V3O8 material is a litter better than sample RNT, the number of cycles is small. In this paper, after 50 cycles, the discharge capacity retention rate maintains 75 %.
Table 1
Comparison of the capacity and capacity retention reported in recent papers for NH4V3O8
Morphology | Synthesis method | Initial discharge capacity | Cycle number | Capacity retention | Measured current density (mA g− 1) | References |
Regular hexagonal micro plates | Microwave assisted hydrothermal | 270 | 3 | - | 10 | [38] |
Flower | Microwave hydrothermal | 312 | 20 | 78 % | 15 | [39] |
Nanobelt | Microwave hydrothermal | 328 | 20 | 94 % | 15 | [39] |
Nanorods | Water-bath | 208 | 30 | 104 % | 15 | [40] |
rectangular nanotube | Solvothermal | 253 | 50 | 75 % | 15 | This paper |
The Nyquist plots of the as-prepared samples are compared in Fig. 4a. Firstly charged the battery to 4.0 V, and then measure electrochemical impedance spectroscope in the frequency range from 0.01 Hz to 100 kHz. NH4V3O8 RNT shows a much lower charge transfer resistance is 98.4 Ω, suggesting that it is easy to transfer for Li+ and electrons during the charge and discharge process. To better understand the plots, an equivalent circuit model was used to fit the impendence in Fig. 4a. The parameters Re, CPE, Rct, and W correspond to the Ohmic resistance of the electrolyte and electrode in the battery, constant phase element, the charge transfer resistance of corresponding electrochemical reactions, and the Warburg impedance related to Li+ diffusion, respectively. The value of Rct of different samples could be simulated from the EIS date by using the equivalent circuit. Figure 4b is the histogram of comparing each resistance value. It could be seen that the Rct value for RNT, HMS, NS, and NP was 83.2, 106.1, 86.8, and 97.2, respectively. The value of the total resistance of RNT is smaller than the other three samples. The results of electrochemical impedance spectroscope measurements are in good agreement with the electrochemical performance.
The charge/discharge curves of the sample RNT at different cycles shows in Fig. 5a.The discharge capacity of NH4V3O8 RNT at the 1st, 2nd, 10th, 25th, and 50th cycle is 269, 250, 233, 215, and 208 mAh g− 1, respectively. There are three discharge voltage plateaus at 2.75, 2.49, and 2.1 V, and the voltage plateaus indicate the multi-step Li+ intercalation/de-intercalation process. Figure 5b depicts the galvanostatic voltage profile of NH4V3O8 RNT at different specific currents. The charge capacity increase and the discharge specific capacity decrease with the variation of current densities from 15 to 300 mA g− 1, because of the sluggish Li+ diffusion kinetics at high current densities[41].
To clarify the better electrochemical performance of NH4V3O8 RNT, the CV measurement at different sweep rates was conducted. The CV analysis is performed to estimate the kinetic and mass transport during the redox process of cathode materials. Figure 6a shows the CV curves of the NH4V3O8 RNT measured in the voltage range of 1.5-4.0 V at a scan rate of 0.1 mV s− 1. The NH4V3O8 RNT shows a peak at 2.48 V and vanishes in the following cycles, corresponding to irreversible lithium insertion into NH4V3O8 in the first two cycles. Figure 6b shows the CV curves of the NH4V3O8 RNT measured in the voltage range of 1.5-4.0 V at different scan rates. As the scan rate increase from 0.1 to 0.8 mV s− 1, the shape of the redox peaks is well preserved, implying that small polarization, fast Li+ insertion/desertion, and good reversibility of NH4V3O8 RNT.