3.1. Microstructure characterization and composition analysis
XRD analysis was performed to investigate the crystal structures of the samples. Figure 1 shows the XRD diffraction patterns of the samples, all diffractograms show similar peaks. The signals at 2θ = 36.0°, 41.2°, 54.3° and 62.7° characterize the of TiO2, and two additional peaks at 2θ = 30.5° and 32.1° appear with related V2O5 and AgVO3, respectively. The AgVO3/V2O5-TiO2 composite showed a compound of rutile TiO2 phase (JCPDS 21-1276), V2O5 phase (JCPDS No.41-1426) [14]and AgVO3 phase (JCPDS No.29-1154) [19], revealing that the mixture was the main composition of TiO2, V2O5, AgVO3 crystal structure. The absence of particularly strong signals from V2O5 indicates the V2O5 highly dispersed. All characteristic peaks of AgVO3, V2O5 and TiO2 are observed in the XRD pattern of AgVO3/V2O5-TiO2.
The photographs of TiO2 and AgVO3/V2O5-TiO2 were exhibited in Fig. 2. During hydrothermal process, TiO2 forms nanorod arrays at FTO, as shown in Fig. 2(a) and (b). For comparison, Fig. 2(c), d display SEM images of the AgVO3/V2O5-TiO2 in preparation.There is no essential change in morphology, but significantly increased surface roughness in comparison with bare TiO2 precursor, which can be seen AgVO3, V2O5 its completely covered on the TiO2 nanorods surface. The diameter of these nanorods is 20-30nm under high magnification SEM observation.
In order to further investigate the microstructure and chemical composition distribution, TEM images and EDS elemental maps were tested in Fig. 3. Figure 3e represents the TEM images of AgVO3/V2O5-TiO2 and the corresponding EDS elemental profiles of Ti, O, V, and Ag, which exhibit the growth of V2O5 around the TiO2 nanorods and the attachment of AgVO3 to the outermost layer of the nanorods in the form of quantum dots (AgVO3 QDs structure presents discrete distribution in V2O5-TiO2 nanorod), confirming the successful formation of AgVO3/V2O5-TiO2 heterostructures. Figure 3a shows the TEM image of AgVO3/V2O5-TiO2 nanorod, an ultrathin coating layer with a thickness of about 10nm can be observed on the surface of TiO2 nanorod and and its surface is roughly covered with a large number of small particles (can be identified as AgVO3 QD), indicating the successful clad on TiO2 rods to form a shell structure as V2O5-TiO2, and AgVO3 QD dispersed separately on the surface of V2O5-TiO2, in good match with the elemental mapping images (Fig. 3e). Three HRTEM images gave further direct evidence for the formation of AgVO3/V2O5-TiO2, Fig. (3b, c and d) show an enlarged view of a partial area the AgVO3/V2O5-TiO2 in Fig. 2a, which can roughly distinguish TiO2 nanorod, V2O5 thin film layers, and AgVO3 QDs. The High-resolution TEM (HRTEM) images show different lattice fringes, lattice spacings of ~ 0.25nm correspond to the (101) crystal plane of rutile TiO2, and the lattice spacing of ~ 0.37nm in the middle V2O5 layer corresponds to their (001) crystal plane, and the (501) lattice spacing of the outermost AgVO3 QDs is 0.306 nm. It can be concluded that AgVO3 QDs has been successfully attached to the surface of V2O5-TiO2 nanoarray. All these findings well indicate that the AgVO3/V2O5-TiO2 nanoarrays were successfully synthesized on FTO substrates.
To further clarify the successful deposition of AgVO3 and V2O5, X-ray photoelec-tron spectroscopy (XPS) was performed to characterize the surface chemical states of AgVO3/V2O5-TiO2. Figure 4a shows the survey XPS spectra results in line with the constituents. The distinct peaks of Ti, O, V, C, and Ag elements for AgVO3/V2O5-TiO2 and the peaks of Ti, C and O elements for TiO2 can be found. Altogether, these observations indicate that TiO2 nanorods are modified by V2O5 and AgVO3 sequentially. To further clarify the interfacial interaction between AgVO3/V2O5 and TiO2, high-resolution XPS spectra of C 1s, Ti 2p, O 1s, V 2p, and Ag 3d in TiO2 and AgVO3/V2O5-TiO2 hybrid were compared. For the C 1s spectrum of in TiO2 and AgVO3/V2O5-TiO2, the peak at 284.8eV is assigned C = C, indicating that adding AgVO3 and V2O5 does not introduce and change the C structure. The XPS data for TiO2 and AgVO3/V2O5-TiO2 showed characteristic Ti 2p3/2 (458.48eV) and Ti2p1/2 (464.57eV) peaks for Ti4+, the binding energy at AgVO3/V2O5-TiO2 is reduced by about 0.22eV compared to the same peak of TiO2 (Fig. 4(c)). This small shift may be due to the presence of additional V5+ and Ti3+ in AgVO3/V2O5-TiO2. The O 1s XPS spectra of both TiO2 and AgVO3/V2O5-TiO2 in Fig. 4(d) show two typical peaks at ~ 530.2 and ~ 532.0 eV, which are attributed to lattice oxygen (Ti-O species) and a surface-adsorbed hydroxyl group (OH−), respectively. As compared to TiO2, it is clear that the O1s XPS spectrum of AgVO3/V2O5-TiO2 adds strong peaks (532.7eV) corresponding to the lattice oxygen (V-O or Ag-O). To gain insight into the oxidation state of V, we performed XPS characterization of all samples, as shown in Fig. 4e. The V 2p peak at 516.8(V4+) ~ 517.6(V5+) eV and 524.4eV are ascribed to the V 2p3/2 and V 2p1/2 levels, and the bare TiO2 does not show any peaks of V, respectively [23–25]. As can be seen from Fig. 4(f), bare TiO2 and V2O5-TiO2 samples have no Ag peak position and Ag 3d peaks appear in AgVO3-TiO2 and AgVO3/V2O5-TiO2 samples, which is completely consistent with the experiment. From Fig. 4(f), it can be seen that the binding energies of 368.2eV and 374.2eV correspond to two distinct peaks of 3d5/2 and 3d3/2 for Ag 3d, respectively[21, 26]. Combining with the above XRD, SEM, TEM, and XPS maps, the above results prove.
3.2. UV-vis diffuse reflectance and bandgag mechanism
Efficient solar light absorption is vital factor for photoelectrocatalytic activity. Figure 5a displays bare TiO2 absorbed light with an absorption onset at about 400nm, complying well with the theoretical bandgap (be calculated to 3.07eV) of rutile TiO2[4, 27]. Figure 5(a) compares the UV–vis diffuse absorption spectra of the bare TiO2, AgVO3-TiO2 V2O5-TiO2 and AgVO3/V2O5-TiO2. Figure 5(a) and (b) show the composite photocatalysts can adjust the absorption cut-off wavelength. The visible light absorption intensity (VA) was estimated with VA (I×nm) of 7, 61, 144, 189 corresponds to the bare TiO2, AgVO3-TiO2, V2O5-TiO2 and AgVO3/V2O5-TiO2[4], respectively, as shown in Fig. 5(a). The prepared AgVO3/V2O5-TiO2 heterostructure has high light absorption properties in visible light (400nm ~ 700nm). Figure 5(b) shows the Tauc plot of UV-vis spectra, which reveals that the Eg of the bare TiO2, AgVO3-TiO2, V2O5-TiO2 and AgVO3/V2O5-TiO2 are about 3.07, 2.92, 2.12 and 1.41eV, respectively.
The reason for the high visible light absorption performance of the 1D AgVO3/V2O5-TiO2 heterostructures is described as follows. The conduction band, valence band and Fermi level of TiO2, AgVO3 and V2O5 are shown in Fig. 5(c), it is shows the wide band gap of TiO2 (~ 3.07eV) with coupling two band gaps small semiconductor materials V2O5 (~ 2.40eV) and AgVO3 (~ 2.55eV) [4, 28, 29], but these three semiconductor materials are not premium heterojunctions if they cannot be synthesized into one material [6]. Fermi-level alignment refers to the fact that when semiconductor materials of different energy levels come into contact generally involves a redistribution of charge, which causes the shift in band edge positions [29, 30]. This indicates that when the heterostructure is formed, the Fermi energy of the AgVO3/V2O5-TiO2 semiconductor has to be the same. This leads to the CB and VB of both AgVO3 and V2O5 to lie above TiO2 as shown in Fig. 5(c). Specifically, when the sunlight shines on the surface of the AgVO3/V2O5-TiO2, since the CB of AgVO3 and V2O5 are higher than bare TiO2, a double type-II band alignment exists [6], and the photo-generated carriers are transferred from the CB of AgVO3 and V2O5 to TiO2. This double electron transfer paths can greatly enhance the separation of photogenerated electrons and holes. Similarly, the holes in the VB of TiO2 transfer to the VB of AgVO3 and V2O5 and facilitates the degradation of pollutants. Therefore, although the current configuration is not ideal for double type-II band gap heterojunctions, Efficient charge separation can still be achieved by Fermi-level alignment. This process can efficiently accelerate the separation of photo-induced e−/h+ pairs and prolong the e−/h+ pairs lifetime. The the double type-II band gap information is confirmed by the photocatalytic and the photoelectrochemical test results. This in turn can be used to elucidate the photocatalytic mechanism.
3.3 Photocatalytic property
Using the photocatalytic test evaluation described in Sect. 2.3, the photocatalytic degradation efficiency against Methylene blue under 150W simulated solar (AM 1.5G, 100mW/cm2) irradiation. The absorption spectra decolorization curves of the MB solution after different photo-degradation time treatments with bare TiO2 and AgVO3/V2O5-TiO2 are displayed in Fig. 6(a). After 120 min of simulated sunlight, the cannot be excited by visible light, so the MB degradation efficiency can only reach 49.9%, while the MB degradation rate of AgVO3/V2O5-TiO2 reached 95.8%. In Fig. 6(b) and (c), the photocatalytic performance of AgVO3/V2O5-TiO2 was outstanding, the photocatalytic degradation rates were 49.9%, 85.4%, 80.2%, 81.6%, and 95.8% for TiO2, AgVO3-TiO2, V2O5-TiO2, respectively.
According to kinetic principles, the recombination of photogenerated e−/h+ pairs can occur within 10− 9s to 10− 12s. When electron acceptors (pollutants) can be pre-sorbed on the catalyst surface, this is more favorable for photocatalytic performance. and the kinetics of the studied catalysts are shown in Fig. 6(d). The reaction rate constant of AgVO3/V2O5-TiO2 is 0.025min− 1, which is approximately five times that of TiO2 (k = 0.005min− 1). The photocatalytic mechanism is shown in Fig. 5(c).The reaction rate constant order of the samples is as follows: TiO2 < AgVO3-TiO2 < V2O5-TiO2 < AgVO3/V2O5-TiO2. As mentioned above, the AgVO3/ V2O5-TiO2 sample demonstrated superior photocatalytic MB reduction activity compared to their sampls, with the main source of this superior activity is related to the reasonable heterojunction structure and the resultant photo-induced charge transfer properties.
3.4 Photoelectrochemical performance and water splitting
The photoelectrochemical (PEC) performances of the TiO2, V2O5-TiO2, AgVO3-TiO2 and AgVO3/V2O5-TiO2 nanorod arrays grown on FTO substrates were characterized. All the samples had been tested for 3h to investigate the stability. The separation, capture and migration of the photo-generated carriers on the surface of catalysts were investigated by Photocurrent, Electrochemical impedance spectroscopy (EIS) and Photoluminescence spectra (PL). Figure 7(a) shows a comparison of the current density responses of samples. For the bare TiO2 (6μA), there was less photocurrent response due to its the intrinsic limitation of the quantum yield. The loads of AgVO3-TiO2 and V2O5-TiO2 could further enhance the photocurrent response of TiO2, while AgVO3/V2O5-TiO2 (115μA) presented higher response compared with AgVO3-TiO2 (15μA) and V2O5-TiO2 (88μA). Importantly, AgVO3/V2O5-TiO2 exhibits the highest photocurrent, indicating a more efficient separation and longer lifetime of the charge carriers, which is in good accordance with the order of their photocatalytic measurements. The charge transfer characteristics of the photoelectrodes were further obtained by EIS analysis of different samples, and the results are presented in the form of Nyquist plots as shown in Fig. 7(b). The EIS response demonstrated the AgVO3/V2O5-TiO2 possessed a smaller impedance radius compared with TiO2, AgVO3-TiO2 and V2O5-TiO2, and creates a more suitable environment for holes transfer across the interface to the electrolyte [31]. The electron transport recombination properties was further confirmed by the results of PL. In Fig. 7(c), compared to TiO2, the PL emission intensity of AgVO3-TiO2 and V2O5-TiO2 and AgVO3/V2O5-TiO2 were dramatically decreased, and the AgVO3/V2O5-TiO2 shows the lowest intensity. The Nyquist plots and quenched PL intensity imply the effective charge transfer by the double type-II scheme heterojunction. Based on the PC, EIS, PL results, it provides strong support for the effective separation of e−/h+ pairs of AgVO3/V2O5-TiO2. It could be concluded that the two-step hydrothermal treatment has been shown to optimize the interface TiO2 nanorod array, the reorganization rate of e−/h+ pairs was slowed down and the photocatalytic performance of PEC was significantly improved.
To evaluate the PEC performance of the TiO2, AgVO3-TiO2, V2O5-TiO2 and AgVO3/V2O5-TiO2, we performed linear sweep voltammogram (LSV) and Mott–Schottky (M-S) in an alkaline electrolyte. Figure 7(d) shows the photocurrent–potential (J–V) curves for the sample photoanodes under one standard simulated sunlight irradiation (100mW/cm2). In Fig. 7(d), LSV curves of TiO2, AgVO3-TiO2 and V2O5-TiO2 and AgVO3/V2O5-TiO2 showed OER activity with initialpotentials of 1.93, 1.71, 1.64 and 1.47V to reach current density of 10mA/cm2, respectively. Using the equation η = ERHE −1.23, the overpotential (η) of is calculated, as shown in Fig. 7(e). Clearly, the AgVO3/V2O5-TiO2 can serve as efficient OER electrocatalyst for practical application at high current density and exhibits the lowest onset overpotential (246.2mV), which is less than 446.9mV for bare TiO2 (693.1mV). This enhancement can be attributed to the double type-II band gap of heterostructure. In addition, Fig. 7(f) shows indicates that the LSV curve of AgVO3/V2O5-TiO2 is lower (145.6mV) in the light source onset potential than in the dark, indicating a photocatalytic effect. Mott–Schottky (M-S) analysis has been proved to be an effective tool in studying the electronic properties, by means of which can be determined flat band potentials (VFD) and donor density (ND) of samples. Here, the flat band potential of electrodes is determined by the Mott–Schottky equation [31–33]:

with ε0 (8.86×10−12F/m) and ε (90F/m)-the passive oxide and vacuum permittivity, respectively; e-the electron charge (1.6×10-19C); ND-the donor density inside the passive oxide; VFB-the flatband potential; k-the Boltzmann constant; T-the temperature (the value of kT/e0 is 0.026V at 25°C). By extrapolation to
the flat band potential VFB can be determined. From Fig. 7(g), the slopes of Mott-Schottky plots show positive values for all samples, suggesting that the AgVO3 QDs and V2O5 effect will not change the n-type conductivity of TiO2. The VFB of TiO2, V2O5-TiO2, AgVO3-TiO2 and AgVO3/V2O5-TiO2 derived from the extrapolation linear M-S plot to potential bias axis are 0.46V, 0.82, 1.06 and 1.27 versus RHE, respectively. As can be seen in Fig. 7(g), the slope of the Mott–Schottky plot for bare TiO2 is much steeper than for the others amples. More specifically, According to Equ. (2), ND values of the TiO2, V2O5-TiO2, AgVO3-TiO2 and AgVO3/V2O5-TiO2 photoelectrodes were calculated to be 4.3ⅹ1017, 9.4×1017, 6.4×1017 and 2.15×1018cm−3, respectively. After AgVO3 and V2O5 modification, the ND could be remarkably increased almost 5 times as compared with pristine TiO2. This result reveals a drastic increase of free charge carriers in the AgVO3 and V2O5 modified TiO2 electrodes. To further confirm our deduction, the black and red lines show the fitting of the linear range of the Mott-Schottky plots for both in the dark and light AgVO3/V2O5-TiO2 based photoanodes. The fits shown in Fig. 7(h) yield VFB=1.27V and ND=2.15×1018cm−3 for in the light AgVO3/V2O5-TiO2 samples and VFB=1.18V and ND= 2.09×1018cm−3 for in the dark AgVO3/V2O5-TiO2 samples, which shows that the increased quantity ∆ND=6×1016cm−3 derives from the effect of light and is consistent with the LSVcurve.