By regulating the building units, a series of defined phases (orth-, tri-, and tetra-MoVO) could be fabricated and applied in ZIBs. Those Mo-V oxides consisted of the pentagonal polyoxomolybdate [Mo6O21] with a central MO7 (M=Mo, V) pentagonal bipyramidal unit and edge-sharing MO6 octahedral.5,56,57 The existed various continuous channels in the  direction could give a positive effect on the Zn2+ diffusion (insets of Figure 1a, 1d, and 1g). The variable tunnel structures endowed the open channel frameworks for the three oxides, and resulted in varied diffusion kinetics of Zn2+. Compared with the tri-MoVO phase, orth-MoVO structure possesses more six- and seven-member rings in the  direction, while the tetra-MoVO only comprised four- and five-member rings. Power X-ray diffraction (XRD) patterns were employed to characterize the crystalline structure (Figure 1a, 1d, and 1g), in agreement with previously reported data.58 The well-defined XRD patterns indicate high crystallinity. The high-resolution transmission electron microscopy (HR-TEM) images in Figure 1b, 1e, and 1h show the lattice space of the three samples, and the morphological and elemental features are shown in Figure 1c, 1f, and 1i. The TEM images of the three samples show a 1D rod-like shape, with a diameter of approximately 200 nm. Moreover, the crystallographic growth occurred along the  direction. The elemental dispersion spectroscopy (EDS) mappings reveal that the distributions of the Mo, V, and O elements are completely consistent with the TEM images of the three samples. As shown in Figure S1, the ultra-visible-near infrared (UV-vis-NIR) diffuse reflectance spectra of the orth-, tri-, and tetra-MoVO exhibit an apparent absorption in the visible light region, owing to the electron transfer between Mo and V. The unique band structure indicates improved conductivity, which is favorable to the electrochemical process.
To estimate the Zn2+ storage ability of the as-prepared Mo-V oxides, the CR2016-type cell was assembled with Mo-V oxides as cathodes and Zn foil discs as anodes. The electrolyte was 2 M of aqueous Zn(CF3SO3)2 solution. Figure 2 shows the cyclic voltammetry (CV) and galvanostatic charge-discharge (GCD) curves of the orth-, tri-, and tetra-MoVO cathodes. Obviously, the orth-MoVO exhibited a Zn2+ insertion potential of approximately 0.55 V and a multistep extraction potential of approximately 0.55, 0.7, and 0.8 V (Figure 2a and 2d). In the subsequent 2nd and 3rd cycles, the Zn2+ insertion processes became facile and the CV curves exhibited good reversible properties. In Figure 2b, the tri-MoVO exhibited similar CV as the orth-MoVO and an additional Zn2+ insertion potential at 0.3 V. However, the tetra-MoVO exhibited one pair of redox peaks (Figure 2c), which suffered from severe irreversibility in the subsequent 2nd and 3rd cycles. As shown in Figure 2d-f, the orth-MoVO had a specific capacity of approximately 400 mAh g−1 at 0.1 A g−1 owing to the large amount of six- and seven-member ring tunnels, while the tri- and tetra-MoVO delivered less than 200 mAh g−1. Additionally, the specific capacity of the orth-MoVO was approximately without attenuation during the first three cycles. Hence, the four- and five-member ring tunnels of the tetra-MoVO contribute toward slow diffusion kinetics, which lead to a lower specific capacity. The poor reversibility of the tetra-MoVO was caused by the large charge repulsion force of Zn2+ and the narrow tunnels. Thus, the orth-MoVO is a more suitable cathode for ZIBs. The galvanostatic intermittent titration technique (GITT) measurements shows that the Zn2+ diffusion coefficient for the orth-MoVO range from 10−8 to 10−9 cm2 s−1 order of magnitude, revealing the fast ion migration.
The exact atomic ratio and electrochemical performances of the orth-MoVO was investigated in detail to clarify the Zn2+ storage mechanism. Inductively coupled plasma-atomic emission spectroscopy (ICP-AES), thermogrametric (TG) and in-situ XRD analysis revealed the molecular formula of the orth-MoVO can be defined as Mo2.55VO9.43 (Figure S2 and S3). The specific capacity gradually decreased from 400 mAh g−1 at 0.1 A g−1 to 170 mAh g−1 at 2 A g−1 (Figure 2g and 2h). And the orth-MoVO delivered a specific capacity of 347 mAh g−1 at 0.1 A g−1 after 100 cycles (Figure 2i). Additionally, the specific capacity reached 145 mAh g−1 at 2 A g−1 after 1000 cycles. However, the rate capabilities were quite poor for the tri- and tetra-MoVO, and both of them underwent severe capacity degradation at 0.1 A g−1 (Figure S4). The discharge plot was divided into two parts for the GCD curves. One part was a platform at 0.65 V, while the other part was an inclined line from 0.55 V to 0.2 V. Figure S6 shows the GCD plots at 0.1 A g−1 with Zn(CF3SO3)2 dissolved in water and acetonitrile, respectively. The 1st discharge plot in water has a higher platform than that in acetonitrile, which indicates a more facile Zn2+ diffusion process in water, compared with that in acetonitrile. The 2nd discharge plots in water and acetonitrile are similar, delivering the same specific capacity. This phenomenon indicates that water exerts a lubricant effect on the 1st Zn2+ insertion process.59 However, the 2nd discharge plot illustrates that the extraction of Zn2+ is incomplete. The residual Zn2+ may have acted as pillars for expanding and stabilizing the crystalline orth-MoVO structure, which facilitates the subsequent insertion process. In addition, the discharge plots for orth-MoVO rule out the intercalation of H+ in aqueous electrolyte.
In-situ XRD was carried out to clarify the crystal structure evolution of the orth-MoVO during the Zn2+ insertion/extraction process. Figure 3a shows the detailed pseudo-colouring XRD patterns evolution when the orth-MoVO undergoes a discharge/charge process (Figure S7). The XRD pattern of the prepared cathode is similar to that of the pristine orth-MoVO. During the Zn2+ insertion process, the diffraction peaks at 22.3° and 26.7° slightly moved toward a higher degree, while the peaks at 7.0°, 8.2°, 9.4°, 11.2°, 24.2°, 27.1°, 27.4°, 27.6°, 28.6°, and 29.6° shifted toward a lower degree. Moreover, the peaks at 27.4° and 27.6° merged together. The peaks at 7.0°, 8.2°, 9.4°, and 22.3° are attributed to the (020), (120), (210), and (001) crystal planes. These results indicate that the lattice space of (020) increases, while that of (001) decreases during the Zn2+ insertion process. During the Zn2+ extraction process, all peaks approximately revert to the pristine state, which indicates excellent reversibility of the phase structure. In-situ Fourier transfer infrared spectroscopy (FT-IR) was conducted (Figure S8). The peak at 901 cm−1 exhibited a red-shift while the peak at 937 cm−1 exhibited a blue shift during the Zn2+ insertion process. After the extraction of Zn2+, the two peaks merged at 925 cm−1. These two peaks are considered to have been induced by the V-O and Mo-O vibrations, which indicate that Zn2+ penetrates into the tunnels and coordinates with the O atoms. Additionally, a portion of Zn2+ remains in tunnels after the extraction process, and acts as pillars for stabilizing the orth-MoVO. The extended X-ray absorption fine structure (EXAFS) ensues, which indicates bond variations at different discharge/charge stages. Figure S9 shows the K-edge FFT function spectra of Mo, V, and Zn. After a full discharge/charge process, the first dominant peaks of Mo and V slightly increased, and there were approximately no changes for Zn.
X-ray photoelectron spectroscopy (XPS) was performed to further demonstrate the changes of Mo, V, and C during the discharge/charge process (Figure S10). This XPS results indicated that CF3SO3- underwent an insertion/extraction process along with the Zn2+ ions, which is consistent with the EDS results presented in Figure S11. The XPS and EDS analysis further revealed that the Zn2+ extraction process was not completely reversible in the first cycle. The TEM image of the orth-MoVO morphology under 0.2 V (Figure S12b) shows a precipitation layer on the surface. This precipitation layer decomposed during the charging process (Figure S12c). As shown in Figure S13, the voltage of the orth-MoVO cathode remained stable below 0.25 V after the discharge process. However, the voltage sharply jumped to about 0.6 V after dipping in HCl, which may be induced by the dissolution of the precipitation layer.
Subsequently, HR-TEM was employed to track the structural transfer of the orth-MoVO during the GCD process. The HR-TEM image of the orth-MoVO in Figure 3b and the corresponding FFT pattern in the  zone axis reveal a crystalline structure, which is in agreement with the results of scanning transmission electron microscopy (STEM, Figure S14). After discharged to 0.2 V, the insertion of Zn2+ resulted in crystalline transformation (Figure 3c). The lattice space of (001) decreased from 0.408 to 0.371 nm, while the lattice space of (020) increased from 1.27 to 1.39 nm. This phenomenon is in agreement with the in-situ XRD analysis results. The angle between (010) and (001) became 77.23° (Figure 3c), then returned to 88.8° after charging to 1.6 V (Figure 3d). Additionally, the lattice spaces of (001) and (020) underwent reversible changes during the charging process. Figure S15 shows the Rietveld-refined XRD of the orth-MoVO at 0.2 V and the corresponding atomic structure in the  direction. It is proposed that Zn2+ inserted into the six- and seven-member rings of the orth-MoVO. The Zn2+ storage mechanism is illustrated in Figure 3e according to the abovementioned results. In the discharge process, Zn2+ and CF3SO3 entered into the orth-MoVO tunnels. Then, part of Zn2+ and CF3SO3- remained in the tunnels to stabilize the orth-MoVO after the 1st charge process. Thereafter, the Zn2+ ions can insert and extract reversibly from the orth-MoVO in the subsequent GCD processes.
With consideration to the most favorable path of Zn2+ diffusion, the possible adsorption sites on the oxygen atoms of MoO6 and/or VO6 in seven-, six- and five-member MO6 units were considered. The density functional theory (DFT) calculation was carried out to theoretically evaluate the adsorption properties of Zn2+. The formation energies of Zn2+ ion embedded into the three investigated systems are employed. The porous channel consisting of heptagonal and hexagonal MO6 clusters (channel) were selected for the orth- and tri-MoVO, while pentagonal channel for tetra-MoVO, with the consideration to the most favorable path of Zn2+ diffusion. Additionally, different possible adsorption sites on the oxygen atoms around the porous channel were considered, and the one with strongest adsorption was chosen as the most stable adsorption site. According to the calculation results in Figure 4a and S16, the Zn atoms is prone to entering the hepta-member channel with two Zn2+ simultaneously in, which is consistent with the experimental estimations. The 11 diffusion pathways were considered to give the optimal model of Zn2+ diffusion. In Figure 4b, Zn2+ tends to alternately enter the hepta- and hexa-member channels of orth-MoVO, both of which allow for 2 Zn2+ insertions. The system becomes unstable if the third Zn2+ entered the hepta- and hexa-member channel. Consistent with the Figure S13, about one third of discharge capacity is ascribed to the 4 Zn2+ insertion process, whereas the other two thirds discharge capacity is obtained from the precipitation reaction on the surface of the orth-MoVO. Figure 4c and S17 show the optimal atomic model after accommodating Zn2+ for orth-MoVO. However, the situation is different for tri-MoVO. In Figure S18 and S19, the system is unstable for Zn2+ insertion in hepta- or hexa-member channels. As shown in Figure S20, the only possible channel to accommodate Zn2+ is the penta-member channel in tetra-MoVO. Considering the channel density and the structural stability, the orth-MoVO exhibits the maximal capacity accommodation and highest kinetic properties for ZIBs.