Insight into the Novel Z-Scheme ZIF67/WO3 Heterostructure for Improved Photocatalytic Degradation of Methylene Blue Under Visible Light

In this work, a novel Z-scheme heterojunction of ZIF67/WO3 nanocomposite is successfully prepared via immobilizing WO3 on the MOFs ZIF67 and confirmed by measuring XRD, FT-IR spectroscopy, XPS, SEM, TEM, and EDS. UV–vis and Mott-Schottky measurements. The photocatalytic activity of the as-prepared ZIF67/WO3 is evaluated by degrading methylene blue (MB) solution under simulated visible light irradiation. Obviously, the optimal ZIF67/WO3-0.3 (designated as 0.3 g addition of WO3) delivers significant photocatalytic degradation exceeding 90% toward MB within 120 min, which is about 3.0 and 4.2 times higher than those of pure ZIF67 (30.0%) and WO3 (21.2%). With the addition of 0.02 g ZIF67/WO3-0.3 composite, a total volume of 50 mL of MB at 10 mg/L was thoroughly degraded in 120 min with an apparent rate constant of 0.0184 min−1. Reactive species scavenging experiments indicated that the holes (h+) and hydroxyl radical (·OH) are mainly responsible for MB degradation. Moreover, the probable reaction mechanism is elucidated by detecting the intermediates by performing LC–MS measurements. The excellent photocatalytic performance of the ZIF67/WO3 is predominantly determined by the structural properties of the composite and the Z-scheme charge transfer pathway, which promote the efficient separation of photogenerated carriers and facilitate the transfer of photogenerated electrons to active sites. Therefore, it is a promising strategy for immobilizing the WO3 on MOFs for constructing Z-scheme heterojunction to degrade the refractory organic dyes.


Introduction
At present, water pollution is a growing crisis worldwide that severely influences the survival of human beings and social development. Organic dyes with an aromatic structure are a kind of extensively used organic compounds in the textile, pharmaceutical, leather, and cosmetic industries [1]. Dyes used in industry are classified into three categories of cationic dyes, anionic and non-ionic. It is reported that 12% of the annual production of cationic dyes is wasted through industrial water streams polluting the environment [2,3]. If the organic dyes and their intermediates are directly discharged into the environment without any proper treatment, it results in decreasing dissolved oxygen concentration, photosynthetic activity, and water quality [4,5]. As a result, water-soluble dyes can harm aquatic plants and animals, cause them poisoning, and creates serious destructive environmental problems [6]. Meanwhile, they will contribute to severe environmental pollution and persistent ecosystem risk because of their carcinogenic and mutagenic nature [7]. To effectively eliminate the above-mentioned risks, various advanced treatment techniques have recently been developed including adsorption, flocculation, activated sludge, ion exchange, advanced oxide processes, etc. [8] Among them, semiconductor photocatalysis is considered one of the most promising methods for removing organic pollutants due to its cost-effectiveness, recyclability, sustainability, and high efficiency [9]. In the past decades, the dramatic development of photocatalysis technology has been witnessed due to its economic, recyclable, sustainable, and high efficiency [10].
In especially, semiconductor photocatalysts with the dramatic visible light response are extremely crucial for obtaining desirable photocatalytic efficiency for pollutants removal [11]. Thus, the exploration and development of the above-mentioned photocatalyst is a vital research hotspot. Among the various semiconductor photocatalysts, tungsten trioxide (WO 3 ) has a narrow band gap of 2.4-2.8 eV, which has been demonstrated as one of the most promising candidates because of its significant visible light absorption, low-cost, environmentally friendly, and stability, etc. [12] It can broaden the light absorbance to the visible light region compared with the widely used TiO 2 , ZnO and SnO 2 , etc. [13] However, the photocatalytic activity of WO 3 is unsatisfactory due to its small specific surface area, weak charge transmission and separation, and poor visible-light response, etc., which will impede the practical application in the degradation of dyes [14]. Therefore, it is quite necessary to solve the above-mentioned defects of WO 3 for improving its photocatalytic activity. In the last decades, numerous efforts have been devoted to the exploration and development of some effective methodologies to resolve such problems, including morphology engineering, the metal of nonmetal doping, defect construction, and heterojunction fabrication [15].
Metal-organic frameworks (MOFs) are a category of distinctive materials and primarily consist of metal ions and organic ligands. Owing to enormous prominent characteristics such as large surface area, excellent porous structure, stability, and various types, they have attracted great interest in various applications such as energy storage, catalysis, separation, drug delivery gas storage, and so on [16,17]. In recent years, in light of their significant optical properties, MOFs have been extensively used in solar energy-driven technologies such as photocatalytic removal of contaminants, hydrogen energy production, CO 2 reduction, etc. [18,19] In this process, the used MOFs not only ameliorate the photocatalytic performance of the main catalyst but also participate in the photocatalysis process [20]. ZIF67 is one of the possess high chemical and thermal stabilities MOF that act as active centers and has gradually attracted attention in photocatalytic degradation [21]. However, pure MOFs often have a high recombination rate of photo-generated electron-hole pairs, large electrical resistance, low photoresponsivity and so on, which prevent them from becoming excellent photocatalysts [22]. Accordingly, the development of MOFs materials and exploration of corresponding application methodologies are still of extreme urgency in the future.
In this work, a novel Z-scheme heterojunction composite photocatalyst has been developed by immobilizing WO 3 onto the ZIF67 MOFs via a facile two-step hydrothermal method and utilized as the photocatalyst for MB elimination. Microstructure and morphology analysis revealed that the WO 3 and ZIF67 in the ZIF67/WO 3 composite show no obvious variation except for the formation of the junction. More importantly, the synthesized ZIF67/WO 3 shows good photocatalytic activity toward MB removal compared with those of the pristine ZIF67 and WO 3 under visible light illumination and solution pH, initial concentration of pollutants and dosage of WO 3 exert significant influence on the photocatalytic degradation for MB. Furthermore, this photocatalytic reaction follows pseudo-first-order kinetic. After optimization, the best degradation efficiency of approximately 90% can be obtained by the optimal ZIF67/WO 3 -0.3 (denoted as the of addition 0.3 g WO 3 during the synthesis of the ZIF67/ WO 3 composite) at pH 9, dye concentration of 10 mg/l, and catalyst dosage of 20 mg in 120 min together with an apparent rate constant of 0.0184 min −1 . The enhancement of the photocatalytic degradation efficiency toward MB removal can easily be attributed to the Z-scheme charge transfer mechanism between ZIF67 and WO 3 , which significantly promote the separation and transmission of photogenerated carriers and finally alleviate the recombination. Additionally, the possible degradation pathway has been presumably referred to by performing the LC-MS investigation.

Synthesis of WO 3
The WO 3 was synthesized based on a literature-reported method with slight modification. Typically, 0.245 g sodium tungstate dihydrate was dissolved in 35 ml deionized water in the baker and then 35 ml concentrated HCl (37%) was slowly dropwise added into the above solution. After reacting for 30 min under energetically stirring, this solution was transferred into a Teflon-line stainless-steel autoclave and then heated at 180 °C for 12 h. After cooling to room temperature naturally, the product was collected by filtration and rinsing with deionized water and ethanol several times. Finally, the WO 3 was obtained by consecutively drying in an oven at 70 °C, grinding with mortar and calcined at 500 °C for 2 h in a muffle furnace [23].

Synthesis of ZIF67/WO 3 Catalyst
For the Synthesis of ZIF67/WO 3 , the ZIF67 sample was Initially synthesized according to the previously reported method with a slight modification [24]. Specifically, 0.58 g (1.99 mmol) of Co(NO 3 ) 2 ·6H 2 O was dissolved in 30 mL of methanol under vigorously stirring which was labeled as solution A, and 0.99 g (12.06 mmol) of 2-methylimidazole was dissolved in 10 mL of methanol which was marked as solution B. Sequentially, the Solution B was slowly dropwise added into solution A under magnetically stirring to make the homogeneous mixture solution [25]. Afterward, the above-prepared WO 3 was gradually dispersed into the mixture solution and then vigorously stirred for 24 h at a speed of 400 rpm. The ZIF67/WO 3 product was achieved by centrifuging the above reactants at 8000 rpm for 5 min then rinsing 3 times with ethanol and finally drying in a vacuum at 65 ℃.

Characterizations
The Fourier transform infrared (FT-IR) was conducted on Fourier transform infrared spectrophotometer (PE Company, America) from 400 to 4000 cm −1 with KBr discs. The XRD patterns were collected on an X-ray diffractometer equipped with Cu K X-ray radiation at 60 kV and 80 mA (German Bruker-AXS(D8). The solution pH was checked and regulated with a digital pH/ion meter Mettler Seven Compact (S220-K). The chemical compositions and valence states were investigated on X-ray photoelectron spectroscopy (XPS, ESCALAB 250Xi) with Al Ka radiation at 300 W. The specific surface area and pore structure of the samples were determined by N 2 adsorption-desorption at 77 K using micrometric 3flex analyzer. The specific surface area was obtained by the Brunauer-Emmett-Teller (BET) method, and the aperture distribution was calculated by the Barrett-Joyner-Heleda (BJH) method. The morphology was checked with Scan electron microscopy (SEM, Hitachi S-4300) with an operating voltage of 5 kV and a Transmission electron microscope (Hitachi H-7650, Japan). The photoluminescence spectra of the samples were determined by F-7000 beside the fluorescence spectrometer of Hitachi, Japan. The UV-Visible diffuse reflectance absorption spectra of the samples were tested on a TU-1901 UV-Visible double-beam spectrophotometer (integrating sphere, Beijing Pfeiffer General). The absorbance of the sample solution was determined by the TU-1901 UV-Visible double-beam spectrophotometer produced by the Beijing Pu-Analysis General Company.

Photocatalytic Experiments
The photocatalytic performance of WO 3 , ZIF67, and ZIF67/WO 3 was evaluated through degradation of MB in a photo-reactor under visible light (300 W, xenon lamps with a 420 nm cutoff filter). The process of the photocatalytic experiment was as follows: in a 50 mL quartz tube, 20 mg photocatalysts were dispersed in 50 mL MB blue solution of 20 mg/L. The mixture was stirred in dark for 30 min in to reach the adsorption saturation of MB on nano photocatalysts. Then the above MB solution with the existence of photocatalyst suspension was irradiated for photocatalytic degradation of MB. At each of the predetermined 20 min intervals, a 3.0 mL suspension was withdrawn and centrifuged at a speed of 8000 rpm for removing the residue catalysts. The supernatant was analyzed by UV-vis spectrophotometer at the maximum wavelength of 664 nm of MB for determining the concentration of residual MB [26].
Both the adsorption and photocatalytic degradation performance of MB by different catalysts were described according to the following equation: where R stands for the removal efficiency of MB (%), C t and C 0 represent the concentrations of MB at times of t and 0 (mg/L), respectively.
The degradation kinetics of MB was fitted by the quasifirst-order model, which can be manifested as follows: where k is the apparent rate constant and C 0 and C t are the concentrations at the 0 and t time, respectively.
The radical trapping assays were performed by adding benzoquinone (BQ), EDTA disodium salt (EDTA) and isopropanol (IPA) as the scavengers in the degraded MB solution for determining the possible generated reactive species involved in the photocatalytic process such as superoxide anion radical (·O 2 − ), hole (h + ) and hydroxyl radical (·OH), receptively.
Under the optimal experimental conditions, the photocatalytic assays were repeatedly recycled for 4 times to assess the reusability and stability of the photocatalysts. At the end of each cycle, the solid sample was collected then washed three times with deionized water and finally dried in an oven at 35 °C for 5 h for the subsequent utilization.

Photoelectrochemical Measurements
Photoelectrochemical measurements were conducted in an electrochemical workstation (CHI660E) by applying a standard three-electrode configure quartz cell with the photocatalyst coated FTO glass with an active area of ~ 1 cm 2 as the working electrode, a Pt foil as counter electrode and Ag/AgCl electrode as reference electrode, respectively, in which 0.5 M Na 2 SO 4 was utilized as the electrolyte. Note that the working electrodes were prepared with a drop-coasting approach. In detail, 40 mg photocatalyst samples were evenly dispersed into a 2 mL mixture solution containing naphthalene solution (0.2 mL) and ethanol (1.8 mL) with persistent sonication for 30 min to prepare the photocatalyst ink. Then, 80 μL above photocatalyst ink was uniformly dropped onto an FTO glass substrate with a controlled active area of ca. 1 cm 2 and dried in air for the following experiments.
The transient photocurrent measurements were conducted at intermittent light exposure with on-off for 10 s. The electrochemical impedance spectroscopy (EIS) was performed in the frequency range of 0.01-0 kHz with at 5 mV AC amplitude. Mott-Schottky plots were measured in the potential region from -1 to 0 V vs Ag/AgCl at a frequency of 1.0 × 10 3 Hz [27].

Characterization of the Sample
The morphology and composition of the WO 3 , ZIF67, and ZIF67/WO 3 were systematically analyzed by performing the SEM, TEM coupled with elemental EDS mapping measurements and the results were displayed in Fig. 1. As shown in Fig. 1a, d, the SEM and TEM images of WO 3 show irregular spherical nano granular morphology with apparent agglomeration and particle sizes of approximately 500 nm. As can be observed in the SEM and TEM images of ZIF67 in Fig. 1b, e, the ZIF/67 owns octahedron morphology with a particle size distribution range from 200 to 300 nm [28]. Additionally, as for the constructed ZIF67/ WO 3 , we can observe from Fig. 1c, f that both of the WO 3 with a smaller particle size are randomly distributed on the surface of the ZIF67 with a larger particle size, indicating the formation of heterojunction between ZIF67 and WO 3 . As reported in previous publications, this is highly conducive to the enhancement of the photocatalytic activity of the contaminant removal and the accelerated transmission of photogenerated carriers is the possible main reason. To further prove the surface element distribution of ZIF67/ WO3, SEM-EDS analysis was carried out, and the results are shown in Fig. S1. Furthermore, Fig. 1g-k illustrate the elemental EDS mapping images of ZIF67/WO 3 samples. We can see that the ZIF67/WO 3 composite is primarily composed of five kinds of elements including W, O, Co, C and N, indicating the successful preparation of ZIF67/WO 3 and in good accord with the SEM and TEM measurements. Figure 2a shows the FT-IR spectra of WO 3 , ZIF67, and ZIF67/WO 3 . For WO 3 , the characteristic bands in the FT-IR spectrum at around 735 cm −1 and 850 cm −1 can be ascribed to W=O and W-O-W stretching vibration bands of WO 3 -based species [29]. The other vibrational modes in this spectrum are associated with the O-H stretching of residual water (3410 cm −1 ), and a peak belonging to CO 2 (2340 cm −1 ). As for ZIF67, it is clearly shown that the peaks of ZIF-67 are mainly ascribed to the ligand 2-methylimidazole. The peaks located at 1300 cm −1 , 1386 cm −1 , and 1411 cm −1 are ascribed to the characteristic stretching and bending modes of the imidazole ring. Co-N at 420 cm −1 , C-N at 989 cm −1 and C=N at 1585 cm −1 are corresponding to ZIF-67 [24,30]. Besides the above characteristic peaks of ZIF67 and WO 3 in the FT-IR spectrum of ZIF67/ WO 3 , the new characteristic peaks centered at 1388 cm −1 and 1639 cm −1 may be caused by the interaction between the WO 3 and ZIF67 [31]. Furthermore, Fig. 2b displays the XRD patterns of WO 3 , ZIF67, and ZIF67/WO 3 . For pure WO 3 , the relatively strong peaks at 2θ = 23.1°, 23.6°, 24.4°, and 26.6° are attributed to the (002), (020), (200), and (120) lattice planes of the monoclinic structure of WO 3 , respectively, which is in good agreement with the standard card of WO 3 (PDF-#72-0677) [32]. The conspicuous diffraction peaks of ZIF67 at 2θ = 7.   . 2 a The FT-IR spectra and b XRD patterns and c partial magnification of the XRD patterns of WO 3 , ZIF67 and ZIF67/WO 3; d XPS survey spectra of WO 3 , ZIF67 and ZIF67/WO 3; high-resolution XPS of spectra of e Co, f W interaction between ZIF67 and WO 3 and their constituent atoms are inserted into each other's lattice [33]. According to the diffraction peak of the bulk semiconductor ZIF67,WO 3 and ZIF67/WO 3 , the grain size of each synthesized product was calculated according to the Scherrer formula, and the results are shown in Table 1 [34]. Sequentially, the XPS spectra were measured to confirm the elemental composition of the composite WO 3 , ZIF67, and ZIF67/WO 3 . Fig. 2d displays the full survey spectra of WO 3 , ZIF67, and ZIF67/WO 3 . It is obvious that C 1s, N 1s, O 1s, W4f, and Co 2p peaks were involved in ZIF67/WO 3 compared to the XPS spectra of ZIF67 and WO 3 . The spectrum belonging to WO 3 shows binding energy peaks at 36.8, 286.5, and 531.4 eV, which refers to W 4f, C 1s, and O 1s, respectively, verifying that WO 3 is the primary ingredient in the composite. On the other hand, binding energy peaks were detected in the XPS spectrum of the sample at binding energies of 783.7 eV, which can be ascribed to Co 2p. By magnification of the Co 2p peaks of ZIF67 and ZIF67/ WO 3 shown in Fig. 2e, we can see that the Co 2p peak of ZIF67/WO 3 shows a slight shift to the higher binding energy than that of the ZIF67. On a contrary, the peaks at the highresolution XPS spectra of W 4f of ZIF67/WO 3 in Fig. 2f show a shift to a lower binding energy ca. 0.21 eV. Undoubtedly, these are attributed to the chemical interaction with the chemical bond between ZIF67 and WO 3 [35]. Additionally, the identical phenomenon can also be demonstrated in other high-resolution spectra of C, N, and O elements in Fig. S2, powerfully confirming the formation of ZIF67/WO 3 . Furthermore, by integrating the intensity of the XPS peaks, the surface element atomic percentage for the WO 3 , ZIF67, and ZIF67/WO 3 specimens were calculated and summarized in Table S1.
The UV-vis diffuse reflection spectrum (UV-vis DRS) was applied to investigate the absorbance properties and band gap of the as-samples. As displayed in Fig. 3a, the bare WO 3 presents relatively weak absorbance in UV-vis light region with an absorption edge extending to 450 nm, while the ZIF67 possesses a stronger absorption in UV-vis light region special for another distinct characteristic absorption  [36]. The index n relies on the type of electronic transition of semiconductors and the n is 1 for the direct and 4 for the indirect semiconductor in general. Therefore, the calculated band gap of WO 3 and ZIF67 were 2.98 eV and 1.98 eV which are shown in Fig. 3b. Photoluminescence (PL) spectrum test was further employed to investigate the separation of photogeneration electron-hole pairs which is important for photocatalytic reactions. As can be seen in Fig. 3c, the pure WO 3 exhibits a PL emission peak at 470 nm with the highest intensity, indicating the WO 3 can be excited by photons with high energy. Conversely, the bare ZIF67 has no obvious PL emission peak in this region. Interestingly, the PL peak intensity of ZIF67/WO 3 shows an apparent decrease than that of the WO 3 , verifying that the photogenerated electrons of WO 3 can effectively transfer to the ZIF67 and incorporation of WO 3 and ZIF67 can accelerate the separation of photogenerated electron-hole pairs, which is highly conducive to the photocatalytic reactions [37]. To further verify the electron transfer between ZIF67 and WO 3 , the time-resolved fluorescence decay spectra of WO 3 and ZIF67/WO 3 were tested and illustrated in Fig. 3d. Compared with that of the WO 3 , the ZIF67/WO 3 shows a shorter PL lifetime, confirming the photogenerated electrons of WO 3 can be effectively extracted by ZIF67 and matched optical energy band structures between ZIF67 and WO 3 are the dominant reason. Convincedly, this offers powerful evidence on the formation of Z-scheme heterojunction between ZIF67 and WO 3 and we can further presumably speculate on the transmission process of photogenerated electrons from CB of WO 3 to VB of E VB of ZIF67 [38]. Figure 4 displays the N 2 adsorption-desorption isotherm curves of WO 3 , ZIF67, and ZIF67/WO 3 . As can be seen in Fig. 4a, the WO 3 exhibits a type IV isotherms curve with H 3 hysteresis loops, indicating that the above samples possess mesoporous [39,40]. Figure 4b shows the N 2 adsorption-desorption isotherms for the ZIF-67 sample, which is categorized as type I isotherm according to the IUPAC classification, indicating the microporous characteristics of the ZIF-67 sample. After a systematical calculation ofthe surface area pore size, and pore volume (at p/p 0 = 0.95) of WO 3 , ZIF67, and ZIF67/WO 3 , distribution is summarized in Table 1. As can be seen, the BET surface area (S BET ) of WO 3 is ~ 4.8 m 2 g −1 , the prepared ZIF67 holds a high BET surface area of 1476.55 m 2 ·g −1 . The Barrett-Joyner-Halenda (BJH) pore size distribution curve further reveals the microporous nature of the ZIF-67 sample. Such favorable microstructure with high surface area and microporous property can not only provide the efficient transport pathways for interfacial charge carrier transferring and trapping but also confine reactants in nano space, which are both beneficial for photocatalytic reaction [41,42]. By comparison, the ZIF67/WO 3 delivers a moderate S BET value of 968.34 m 2 g −1 . Quite obviously, the decreased specific surface area of ZIF67/ WO 3 compared with ZIF67 can mainly be attributed to the insertion of WO 3 inside and immobilization of the surface of the ZIF67 which is conducive to the abundant supplication of catalytical active sites [43]. Some of the pores appeared as mesopores and hysteresis in ZIF67-WO 3 (0.3). This is because in the ZIF67-WO 3 (0.3) are formed cavities between the particles.

Photocatalytic Performance of the Catalysts
In general, the compositional dosage of photocatalyst has a vital impact on the photocatalytic activity for MB elimination. As shown in Fig. 5a, it can be found that pure WO 3 and ZIF67 show inferior photocatalytic removal performance for MB (mere 21.2% and 30%) within 120 min, whereas the ZIF67/WO 3 composites exhibit significantly higher photocatalytic activity toward MB removal with a following orders: ZIF67/WO 3 -0.1 (74.8%) < ZIF67/WO 3 -0.2 (79.5%) < ZIF67/WO 3 -0.3 (90.0%) > ZIF67/WO 3 -0.4 (71.4%) > ZIF67/WO 3 -0.5 (74.3%). Quite obviously, the ZIF67/WO 3 -0.3 endows an optimum degradation rate of MB. Figure 5b shows the linear relationship between -ln (C t /C 0 ) and t (where C 0 and C t are the concentration at time 0 and t, respectively, t is the reaction time and k is the rate constant), we can see that the degradation process of MB by photocatalysts follows the pseudo-first order kinetics, in good line with the previous publications [44]. As shown in Fig. 5c, the ZIF67/WO 3 -0.3 shows the highest apparent rate constants (k app ) for MB degradation as high as 0.026 min −1 compared with those of ZIF67 (0.008 min −1 ) and WO 3 (0.015 min −1 ). This is mainly responsible for the more photoactive sites, accelerated photogenerated charge carriers transportation and reduced recombination of h + /e − pairs on the ZIF67/WO 3 than that of WO 3 [45]. The above analyses further confirm that the compositional content of WO 3 and ZIF67 can significantly influence the photocatalytic performance of ZIF67/WO 3 [46]. dAdditionally, Fig. 5d depicts the UV-vis spectra of the MB solution with various reaction Fig. 5 a Photodegradation of MB with the presence of WO 3 , ZIF67 and ZIF67/WO 3 with different WO 3 contents; b The corresponding linear relationship curves between -ln (C t /C 0 ) and t derived from a, c the histogram of the extracted apparent rate constants (k) from b; d UV-visible spectra of adsorbed MB solution with an extension of the reaction time, and the inset is the images of MB solution before and after degradation times and the inset is images of the MB solutions before and after degradation. We can see that the characteristic absorbance peak of MB shows a gradual decline with the extension of reaction time as well as the color of MB is significantly faded shown in the inset of Fig. 5d. Because of this, all the following investigations on the photocatalytic assays are uniformly performed based on the ZIF67/WO 3 -0.3 photocatalyst.

Photocatalytic Degradation Under Different Conditions
As all we know, the dosage of photocatalysts, concentration of MB, and pH value of the reaction system can significantly influence the degradation efficiency of model contaminants by photocatalysts. As can be seen in Fig. 6a, with the amount of the ZIF67/WO 3 -0.3 photocatalyst increase range from 10 to 50 mg, the final degradation rate of MB provides a tendency of gradual increase in the beginning and then decrease at a bigger amount. 20 mg of the photocatalyst results in the best degradation rate toward MB as high as 86.5%. Quite obviously, an insufficient amount of photocatalysts such as 10 mg will lead to incomplete utilization of the light source, which will generate fewer amounts of photoactive species for MB degradation [47]. Therefore, properly increasing the amount of the ZIF67/WO 3 -0.3 can produce enough active centers and increase the contact area between MB and photocatalysts, which is conducive to the photocatalytic degradation rate of MB. On a contrary, when the amount of the ZIF67/WO 3 -0.3 is persistently increased from 30 to 50 mg, the degradation rate of MB shows a dramatic decrement. This is probably attributed that the excessive photocatalysts  From Fig. 6b, c, we can see the degradation of MB follows the pseudo-first-order kinetics based on the abovementioned calculation approach and the system of 20 mg ZIF67/WO 3 -0.3 for MB removal yields an optimal apparent rate constant of 0.0176 min −1 . Figure 6d displays the degradation rate of MB under different initial dye concentrations. As can be observed, the final photocatalytic degradation rate of MB by ZIF67/WO 3 -0.3 shows a significant decline with the increment of MB concentration ranging from 10 mg/l to 50 ml/L in 120 min and the optimal removal of MB up to 89.8% can be obtained at the concentration of 10 mg/l. This should be ascribed to the decrease of effective photons reaching the surface of the photocatalysts with an increase in the dye concentration and the recombination of charge carriers [48]. Figure 6d illustrates the fitted pseudo-first-order reaction kinetics curves with the above-mentioned equation and Fig. 6e shows the histogram of the extracted k app values, we can see that the k app for the 10 mg/L MB is 0.0201 min −1 , a significantly higher than those of others. Figure 6g shows the photocatalytic degradation curves of MB at different pH. As can be seen, the degradation rate for MB gradually increases from 38 to 90% with the increase in pH range from 3 to 9. Conversely, the dramatic reduction in the degradation rate toward MB can be witnessed when the pH finally increases to 11. Additionally, Fig. 6h, i show the fitted kinetics curves and the extracted k app by using the abovementioned pseudo-first-order reaction kinetic model. The k app for MB degradation at pH = 9 is 0.0201 min −1 . Therefore, it can be found that the optimal conditions for photocatalytic degradation of MB are 30 mg ZIF67/WO 3 -0.3 photocatalyst, 10 mg/L initial concentration of MB and pH 9. Note that the following photocatalytic reactions are carried out under these conditions. Moreover, this photocatalyst is comparable with other relative works recently reported in previous literature (Table. S2).

Stability and Capture Experiment
Additionally, the stability and recoverability of the ZIF67/ WO 3 -0.3 were also evaluated by performing four consecutive cycle degradation experiments for confirming the practical applicability. As shown in Fig. 7a, the photocatalytic degradation efficiency of MB by ZIF67/WO 3 -0.3 has negligible change after four cycles, powerfully indicating that ZIF67/WO 3 -0.3 has good stability. Furthermore, it can be also observed that there is no significant change in the XRD patterns of ZIF67/ WO 3 -0.3 before and after photocatalytic reaction (Fig. 7b), simultaneously demonstrating that the ZIF67/WO 3 -0.3 remains superior structural stability after the photocatalytic reaction [49]. Radical capture experiments were performed for unraveling the dominant reactive species during the photocatalytic degradation process of MB by ZIF67/WO 3 -0.3. As shown in Fig. 7c, when the Ethylenediaminetetraacetic acid disodium salt (EDTA) with a concentration of 10 mM was added to the  ) did not participate in the oxidation reaction. Also, it can be found that the above-calculated CB value (-0.362 eV) of ZIF67/WO 3 is more positive than that (− 0.046 eV vs NHE) of the E 0 (O 2 /·O 2 − ), This is because the CB of ZIF67/WO 3 did not reach the O 2 − potential (− 0.33 eV vs NHE), and did not participate in the oxidation reaction in the degradation process [50].

Photoelectrochemical Analysis
Photoelectrochemical measurements were performed to indepth explore the separation, transmission and recombination of the photogenerated electrons/holes (e − /h + ) pairs and unravel the photocatalytic mechanism during the photocatalytic process. In Fig. 7d, we can clearly see that the ZIF67/ WO 3 -0.3show higher photocurrent density without obvious decay compared with those of WO 3 and ZIF67, elucidating that the photogenerated e − /h + has a faster transmission rate in the ZIF67/WO 3 -0.3 than those of others, which is conducive to the photocatalytic degradation of MB. EIS was further performed to understand the dynamic parameters and investigate the charge transport behavior. Figure 7e shows the EIS spectra for disclosing the photogenerated carriers transfer and the separation performance. There are no distinct semicircles in the high frequency region for the Nyquist plots, and the intersections with the real axis represent the ohmic resistance of the electrode. It is clearly seen that ZIF67/WO 3 exhibited a smaller arc than that of bare WO 3 , indicating the ZIF67/WO 3 -0.3 has smaller charge transfer resistance than that of the WO 3 . Noting that the EIS Nyquist radius reflects the migration rate happening at the surface of samples, indicates that ZIF67/WO 3 -0.3 enhanced the separation of photo-induced electron-hole pairs and had a higher interfacial charge transfer rate than pure WO 3 and ZIF67. Figure 7f shows the Mott-Schottky plot for WO 3 and ZIF67.
In Fig. 7f, ZIF67 and WO 3 demonstrates positive slopes in the linear regions of the Mott-Schottky plots, indicating both the ZIF67 and WO 3 have n-type semiconductor behavior. For an n-type semiconductor, E CB locates very close to the flat band potential (E FB ), which can be obtained by the X-intercept when 1/C 2 is zero. Therefore, the WO 3 and ZIF67 was approximately-0.362 and -0.524 V (vs. AgCl, pH 7), and the E FB of WO 3 and ZIF67 was 0.249 eV and 0.087 eV, the E CB of WO 3 and ZIF67 was 0.349 eV and 0.187 eV (vs. NHE, pH 0) [38]. Because the band gap energy of the WO 3 and ZIF67 was 2.95 eV and 1.98 eV obtained from Fig. 3a, b, the E CB and E VB of WO 3 were 0.349 eV and 3.33 eV, meanwhile, the E CB and E VB of ZIF67were 0.187 eV and 2.167 eV, respectively (E (vs. NHE, pH 0) = E (vs. AgCl, pH 7) + 0.0591 pH + 0.197 eV) [51].
According to the band gap structures of ZIF67 and WO 3 , the separation processes of photoexcited electron holes could be exhibited in Fig. 8. There are two possible charge separation ways for ZIF67/WO 3 -0.3 composite. In detail, one is the traditional double-transfer mechanism and the other is the  [52]. While the holes in the VB of WO 3 will migrate to the VB of ZIF67. As a result, the electrons will accumulate to the CB of WO 3 and the holes will gather to the VB of ZIF67. If this is true, the accumulated electrons on the CB of WO 3 can not reduce O 2 to form O 2 − radicals due to the CB of WO 3 being more positive than the potential of O 2 / •O 2 − (− 0.046 eV vs. NHE). Moreover, the holes of ZIF67 cannot oxidize OH or H 2 O to give OH due to the VB potential of ZIF67 is lower than the standard redox potential of OH − /OH (2.40 eV vs. NHE) and H 2 O/ •OH (2.72 eV vs. NHE). However, the radical trapping experiment indicated that h + and •OH are the predominant active species for the ZIF67/WO 3 -0.3 photocatalytic system. Therefore, the separation and transfer process of the photogenerated electron hole charges should not follow the common heterojunction process [53].
According to the above discussion and the experimental results, a possible Z-scheme mechanism for the degradation of the organic pollutant over the ZIF67/WO 3 -0.3 nanocomposites was proposed. ZIF67 has CB and VB values of 0.187 and 2.167 eV, respectively. Under visible light irradiation, electron-hole pairs are generated on the surfaces of ZIF67 and WO 3 . The photogenerated electrons of WO 3 first transfer to the CB of WO 3 and then migrate to the VB of ZIF67 to combine with holes. Therefore, the photo-induced electrons and holes of ZIF67 are separated effectively, and the photoelectrons are continuously transferred to the CB interface of ZIF67. In this way, a large number of electrons are accumulated on the CB interface of ZIF67, and a large number of holes are accumulated in the VB interface of WO 3 . Under the action of holes, H 2 O reacts with h + to produce ·OH, and ·OH reacts with contaminants to form CO 2 and H 2 O. Meantime, a great deal of H + and O 2 react with electrons to generate H 2 O 2 at the conduction band interface of ZIF67. After that, H 2 O 2 reacts with electrons to generate ·OH, and ·OH reacts with MB to form CO 2 and H 2 O. In this way, the photogenerated electrons and holes are separated continuously, which largely increases photocatalytic reaction rate speed, and finally the photocatalytic activity of the catalyst is greatly improved [54]. Based on the above analysis, the specific reactions are as follows: Figure 9 depicted the referred degradation pathway by identifying the probable intermediates with LC-MS measurement during the photocatalytic degradation process. As in, the N-CH 3 bond was broken through demethylation because of its low bond energy, and it was identified at m/z = (1)280, which corresponds to the loss of methylene groups from the MB molecule as a result of ·OH attack. Successive bond scissions occurred in the polycyclic MB, resulting in the opening of the benzene ring resulted in intermediates being formed at m/z values of (13)147, (14)117, and (15)73 [55]. Meanwhile, ·OH radical generated during the photocatalytic degradation process, might attack the sulfhydryl (C-S + = C) group present in MB, which is in direct coulombic interaction with the surface of ZIF67/WO 3 to convert the sulfhydryl group to sulfoxide (C-S=O-C) group. It was supposed that the formation of the reaction intermediates takes place by cleavage of one or more of the methyl groups substituent on the amine group. The formation of (2) azure A, (3) azure B and (4) Thionin through the demethylation cleavage during the photocatalytic degradation [56]. As the degradation time increases, both S and N in the molecular structure of MB underwent redox reactions to form at m/z values of (8)198, (9)168, (10)149, (11)129 and (12)101 [57]. Finally, it leads to the slow degradation of small molecules. All these small molecular fractions were identified as the end products of the photodegradation process, which could be further decomposed to H 2 O, CO 2 , NO 3 − , and SO 4 2− .

Conclusion
In summary, a novel ZIF67/WO 3 composite has been constructed by consecutive two-step hydrothermal strategies. Compared with the pristine WO 3 and ZIF67, this ZIF67/ WO 3 exhibits higher photocatalytic removal performance toward MB model contaminants under visible light. Furthermore, the initial concentration of MB, dosage of used photocatalysts, solution pH exerted significant influence on the final photocatalytic activity of the ZIF67/WO 3 -0.3 for MB elimination. its relative photocatalytic activity was assessed by degrading MB. After systematical investigation of the above conditions, the optimal ZIF67/WO 3 -0.3 (denoted as 0.3 g of WO 3 addition into the ZIF67 precursor solution) endowed the best elimination rate as much as 90% in 120 min and shows desirable consecutive reusability (condition: pH 9, 20 mg of photocatalyst and 10 mg/L of MB). Also, the MB degradation by ZIF67/WO 3 -0.3 faultlessly followed the pseudo-first-order kinetics with an apparent rate constant of 0.0184 min −1 . The radicals capture experiments demonstrated that photoinduced h + played a dominant role in MB degradation and whereas ·OH is the subordinate reactive species, which was mainly reflected by the decline of MB removal rate at different degrees with the addition IPA and EDTA. After a detailed analysis of the intermediates during the degradation process, the possible mechanisms and pathways of MB degradation are presumably inferred, which primarily undergo demethylation, deamination, conjugate structure and benzene ring structure destruction by the derived photoactive species such as h + and ·OH. By analyzing the calculated energy band structure of WO 3 and ZIF67, a novel Z-scheme photogenerated electron transfer path is the main determination for the enhancement of photocatalytic properties of ZIF67/WO 3 compared with the WO 3 and ZIF67, which can dramatically accelerate the photogenerated charges transfer and reduce the recombination of photogenerated e−/h + pairs. Moreover, the enlarged specific surface area and pore structure together with ameliorated light absorbance for ZIF67/WO 3 also play a vital role in its high efficiency of MB elimination. In conclusion, this work may provide new inspirations for improving the activities of photocatalysts based on the novel designation for ordinary semiconductors.