Adjusting the Energy Bands of WO3@ZnO Nanocomposite Heterojunction Through the Combination of WO3 Thin Film to Improve its Photoelectric Performance

At present, nanomaterials with high-quality photoelectric properties are urgently needed to be used in the manufacture of solar cells. In this study, the hydrothermal synthesis method was first used to grow ZnO nanorod arrays, and then a layer of WO3 thin film with controllable thickness was prepared on ZnO nanorod arrays by magnetron sputtering, forming a series of WO3@ZnO nanocomposite heterojunction. We found that the value of the photocurrent of the prepared nanocomposite samples is nearly 30 times higher than WO3 films under illumination, and it is more stable. The results show that this controllable microstructure can further modify the surface properties of ZnO nanorods, and possess the high visible absorption and photoelectric conversion efficiency. By controlling the thickness of the WO3 film, the band can be regulated and ultimately optimized the photoelectrochemical properties of the composite structure.


I. INTRODUCTION
In recent years, tungsten oxide has drawn widespread attention because of its characteristic properties, just like electrochromism, gas sensitivity, photoluminescence and superconductivity [1]- [9]. In addition, tungsten oxide semiconductor materials are non-toxic and have a narrow forbidden bandwidth (2.4 ∼ 2.8 eV) [10]. They can exploit a part of visible light and good stability to refrain photocorrosion [11]- [15], so they can be used as catalysts for visible light catalytic reactions to degrade organic compounds in sewage [16]. Tungsten trioxide (WO 3

) is the most stable
The associate editor coordinating the review of this manuscript and approving it for publication was Wei E. I. Sha . of tungsten oxides with a band gap of 2.8 eV and it is an important n-type semiconductor material [17]. It is widely used as an active photocatalyst of visible light [18]- [20]. However, in practical applications, the photoelectric performance of WO 3 needs to be improved. Recently, there have been extensive studies on the photoelectric performance of WO 3 , which have a bearing on solar energy utilization and photocatalytic. In recent years, the appearance of nanostructured tungsten oxide materials has a great impact on the research in the above fields. At present, many forms of WO 3 materials have been prepared, such as nanoparticles, nanotubes, nanosheets, nanorods and nanowires, etc., [21]- [26]. There are many methods for preparing WO 3 thin films, including electrochemical deposition, magnetron sputtering, sol-gel, hydrothermal synthesis, and the like [27], [28]. However, the photoelectric performance of the various shapes of pure WO 3 is still limited. According to the analysis reports, compositing other materials is a valid method for electron-hole pair separation. If the semiconductor materials are combined with other semiconductors to constitute a special composite heterojunction, which can availably control the structure of energy band and adjust the photo-generated charges spread of the composite materials, increase the life of carriers and enhance the performance of materials [29]- [31].
Among low-dimensional nano-materials, titanium dioxide (TiO 2 ) is the most studied material owing to its characteristics of high light conversion efficiency, good heat stability, high thermal conductivity, low cost and non-toxicity [32], [33]. TiO 2 nanofibers, mesoporous TiO 2 , etc. have important applications in the field of solar cells [34]- [36]. In 2016, Aadesh Pratap Singh, Bodh Raj Mehta et al. applied hydrogenation treatment to the top TiO 2 layer in the BiVO 4 /TiO 2 heterostructure to change the band edge to improve the photoelectrochemical performance [37]. An appropriate substitute material for TiO 2 is ZnO, which has a similar band structure and electron affinity to TiO 2 , and has a larger absorption ratio in the solar spectrum than TiO 2 [38]. It has a lot of research in the photoelectrochemical water splitting experiment [39]- [41]. ZnO is a semiconductor material with a wide band gap (Eg=3.37 eV). Therefore, the ultraviolet band has high absorption [42], [43]. In addition, the combination of electron-hole pairs occurs rapidly in a very short time (10 −9 s to 10 −12 s) [44]. There are many preparation methods of ZnO nanomaterial, including hydrothermal method, magnetron sputtering, chemical vapor deposition, sol-gel method and so on [45]- [52]. Therefore, the morphology of the ZnO structure can be highly adjusted by a simple method [53]. In those methods, the hydrothermal synthesis method is proverbially used in laboratories for it has mild reaction conditions, simple operation, good product crystallinity, no pollution, and the ability to prepare materials with advanced morphological characteristics [54]. Coupling ZnO with WO 3 can enlarge the spectral response range, increase the utilization of visible light, change the band structure and repress the combination of electron-hole pairs, thus improving the performance of the material.
In this experiment, loading an appropriate amount of WO 3 can maximize the photo-current. We have found the best amount of WO 3 that displays the maximum photoelectrochemical performance of ZnO nanorods. For this purpose, WO 3 @ZnO core-shell nanocomposites were synthesized by magnetron sputtering and hydrothermal method. Compared with a single WO 3 nanomaterial, the WO 3 @ZnO nanocomposites structure simply prepared by magnetron sputtering and hydrothermal method can availably suppress the recombination of electron-hole pairs. The specifics of the crystal structure, apparent morphology and optical properties of WO 3 and WO 3 @ZnO composite oxide materials were studied systematically. Moreover, the energy band gap has been calculated to comprehend the optical absorption mechanism of ZnO nanorods with optimum WO 3 loading.

B. PREPARATION OF ZnO NANORODS
First, we clean and dry the required amount of Conductive Glass FTO, then attach the FTO to the substrate tray with high temperature tape and place it in the vacuum chamber. When the vacuum is less than 4 × 10 −4 Pa, set the argon flow to 40 sccm, the air pressure to 1 Pa, and the radiofrequency (RF) sputtering power to 60 W, then pre-deposition for 10 minutes to remove the surface impurities and improve sample purity. FTO was aimed to the center of the target to ensure the uniformity of the film and each sputtered for 2 min. In this study, the sputtering velocity is 10 nm per minute, so the seed layer is about 20 nm thick. After the sputtering is completed, the samples and target were cooled for 60 min and then taken out. Before hydrothermal growth of ZnO nanorods, hexamethylenetetramine (HMTA) and zinc nitrate hexahydrate (Zn(NO 3 ) 2 ·6H 2 O) were mixed with a ratio of 1:1 and then deionized water was added to prepare precursor solution having a concentration of 40 mol/L. The hydrothermal synthesis process will be performed at 95 • C for 4 hours. Then taking out the sample, rinse the surface attachment with ultra-pure water and dried at room temperature naturally.

C. PREPARATION OF WO 3 @ZnO NANOCOMPOSITE HETEROJUNCTION
The WO 3 @ZnO thin film was further synthesized on the above samples through magnetron sputtering. The ambient air pressure is 1 Pa and the RF sputtering power is 65 W. In this study, the sputtering velocity of this particle is 2.82 nm per minute. The composite samples with different WO 3 thickness (about: 20 nm, 40 nm, 60 nm, 80 nm, 100 nm) were prepared separately by controlling the length of the sputtering time. After all samples were sputtered, they were cooled in vacuum for 60 min and then removed. Finally, the samples were put into the sintering furnace at room temperature (25 • C) and heated to 450 • C at the rate of 7 • C per minute and then maintained for 3 hours to complete the annealing experiment. For better comparison of composite samples, the samples containing only pure ZnO nanorods were also annealed.

D. CHARACTERIZATION
The scanning electron microscopy and X-ray diffractometry were used to investigate the structure, elemental composition and structure of the prepared thin film composite samples. The ultraviolet reflection spectra were established by a solid-ultraviolet-visible spectrophotometer. The contact angle tester was used to test the hydrophilic and hydrophobic properties (DSA30, Krüss, Hambourg, Germany). The samples were passed through a three-electrode chemical workstation to measure the photoelectric properties of the samples, such as photocurrent and impedance.

III. RESULTS AND DISCUSSIONS
A. MORPHOLOGY AND STRUCTURE OF WO 3 @ZnO NANOCOMPOSITE HETEROJUNCTION Figure 1(A) exhibits the SEM photograph of ZnO nanorods (FTO+ZnO nanorods). The morphology of ZnO nanorods is decided by the growth conditions, for instance the density of precursor solution, reaction temperature and time. According to a series of previous reports, we selected the conditions of 40 mol/L, 95 • C reaction temperature and 3 h reaction time to generate ZnO nanorods. We can see that ZnO nanorods with uniform size grow over a large area on the surface of FTO and there is no connection between the nanorods. A cross-sectional SEM picture of ZnO nanorods array can be observed in Figure 1(B), which shows the nanorod arrays arranged vertically on the FTO. We can plainly see from Figure 1(C) that after sputtering WO 3 on the surface of the nanorods, many WO 3 nanoparticles are well adhered to the ZnO nanorods, resulting in increased the surface roughness. The SEM images in Figure 2 are FTO+ZnO+WO 3 (60 nm, 80 nm, 100 nm) corresponding to (A-C), respectively. As the thickness of WO 3 increased, the surface of the nanorods was covered gradually and many nanoclusters formed by the aggregation of WO 3 particles. As the WO 3 particles gradually filled the nanorods, the diameter of the nanorods first increased, and then decreased with the further increase in the content of WO 3 . Moreover, the inset in Figure 2 indicates that the cross-linking between the nanorods becomes serious. Overall, the ZnO nanorods loaded with 60 nm WO 3 have the best growth morphology, and it can be seen from Figure 2(A) that the surface of the nanorods is tightly coated.
XRD patterns of synthesized WO 3 @ZnO nanocomposite are represented in Figure 3. Sample 1# shows that the crystal structure of ZnO nanorods corresponds to that of wurtzite ZnO and there is no impurity peak (These peaks can be well indexed to the standard diffraction pattern of wurtzite ZnO structure (PDF # 01-089-1397)). The results reflect that the nanorods prepared by the hydrothermal method have good crystallinity and purity. The diffraction peaks of ZnO correspond to (100), (002), (101), (102), (110), (103) and (004) crystal orientation, respectively [55]- [57]. Among them, the  [58], [59]. By comparison, it was found that the value of the (100) crystal orientation peak was the highest, meaning that the growth of WO 3 along the (100) crystal plane was dominant. As shown in the figure, compared with the peaks of Sample 1#, the peaks belonging to ZnO in all composites haven't moved, which means that the transition metal oxide WO 3 will not replace the lattice of ZnO, and the thickness of WO 3 has a trivial impact on the crystallization phases of ZnO nanorods formed previously. At the same time, it was found that the characteristic peaks of WO 3 weren't obvious when WO 3 was initially sputtered, which may be due to the high dispersion of WO 3 or the low concentration of WO 3 , which was lower than the XRD detection limit. With the increase of sputtering thickness, the characteristic diffraction peak of WO 3 increases gradually. However, when the sputtering thickness is increased to a certain amount, the diffraction peak in (120) crystal direction reaches the maximum and then begins to decrease. Figure 4 shows the hydrophilicity of the samples. We know from Young's formula that when the static contact angle between liquid and solid material surface is less than 90 • (θ < 90 • ), the sample's surface is hydrophilic, which means that the liquid is easier to wet the material; if the angle is greater than 90 • (θ > 90 • ), the surface of the material is hydrophobic, that is, the liquid is not easy to wet the material and is easy to move on the surface [60]- [62]. Figure 4 shows that modifying the surface of ZnO nanorods with WO 3 can appropriately change the hydrophobicity of the material. We found that when there are only pure ZnO nanorods on the FTO, there are a large number of uniformly distributed voids between the ZnO nanorods, which increases the surface liquid/gas contact area fraction. Therefore, the hydrophobicity of the ZnO nanorod films which is modified by WO 3 increased (the contact angle increased gradually), as shown in Figure 4A, B, C.  However, we found that when the sputtering amount of WO 3 is relatively large, the modified nanorods will become larger accordingly and the flatness of the material will increase, so the contact angle is smaller and its hydrophobicity will be reduced to varying degrees.  Figure 5C is the UV-visible absorption spectrum. From Figure 5A, it can be seen that the light reflectivity increases rapidly in the near ultraviolet region, and there is a trailing edge of the reflection spectrum, which is a characteristic of direct band gap semiconductors. Since ZnO nanorods have a high absorption of light in the UV band, it can be seen from Figure 5B that the optical reflectivity is almost 0 in the band of 200 nm-380 nm. At the initial of sputtering WO 3 on the surface of ZnO nanorods, the optical reflectivity of the samples increased and the absorptivity decreased, but later, with the increase of sputtering thickness of WO 3 , the optical absorptivity increased gradually. When the thickness of WO 3 reached 60 nm, the light absorption of the sample was maximum, and then decreased gradually, even lower than the pure ZnO nanorods. We analyzed that this was because when we sputtered WO 3 on the ZnO nanorods surface at the beginning, the roughness of the nanorods was increased as mentioned before, so some reflective surfaces VOLUME 8, 2020 with different crystal directions were added on the surface of nanorods to promote the increase of light reflection. The surface of nanorods began to become dense and compact with the increase of WO 3 particles, the absorption increased, so the light reflectance began to decrease. However, when the amount of WO 3 was too much, the nanorods were covered in a large area and connected to each other, resulting in the reduction of the light absorption, so the light reflectance increased finally. Comparing Figure 5, it can be obtained that the light reflection of pure WO 3 film fluctuates in different degrees in the band of 500-800 nm, while the composite samples modified with WO 3 do not. Furthermore, the spectra of the composite samples loaded with WO 3 have a red shift. This is the result of the band-gap narrowing, and the composite of ZnO with WO 3 will produce defect levels in the band gap [63]. When the Zn 2+ ions in the nanocomposites are replaced by W 6+ ions, the defect energy level corresponding to the oxygen vacancy has come into being, which serves as the color center of visible light absorption. As a result, the reflectance of the composites in the visible band is much lower than that of pure WO 3 .

C. UV-VIS OPTICAL PROPERTIES
From the UV-visible reflectance spectra, the energy band gap for all the composite samples can be calculated as the following equation: where h is Planck constant (6.62607015 × 10 −34 J·s or 4.1356676969 × 10 −15 eV·s), c is the velocity of light (3 × 10 8 m·s −1 ), λ is the wavelength (nm) of absorption onset and Eg is the energy band gap (eV) [64], [65]. The results are shown in Table 1. We can see that the band gap of ZnO nanorod is 3.317 ev, which is close to the reported value. The band gap of the composite samples decreased with the raise of WO 3 load, and the absorption edge moves to longer wavelengths until the WO 3 load is 60 nm and it is reduced to 2.55 eV. Then the absorption edge moves towards the short wavelength direction. The above results indicate that the composition of WO 3 particle layers with different thicknesses changes the band gap of WO 3 @ZnO nanostructures, thus adjusting their ability to absorb visible light. Therefore, an optimal content is needed to decrease the composite effect of ZnO nanorods.  Figure 6 shows the photocurrent response with five light/dark cycles of intermittent illumination at a fixed bias voltage of 0.3 V (vs. Ag/AgCl). It can be found that the value of photocurrent rapidly decreases to about 0 mA/cm2 when the illumination is stopped, and the photocurrent whipped up when the illumination is turned on again. In Figure 6A, the photocurrent of each sample decays severely under illumination; in Figure 6B, only a slight photocurrent attenuation is found for each sample. The reason for the attenuation is that the energy brought by the sudden illumination makes the electron-hole pairs generate rapidly and produces a large photocurrent excitement, and then gradually begin to stabilize over time. It signifies that the composite film reveals better stability and repeatability compared with the pure WO 3 , and the photocurrent is increased nearly 30 times. Furthermore, we also observed that the photocurrent value first increased and then decreased with the increase of WO 3 which sputtered on ZnO nanorods surface. When the thickness of the sputtered WO 3 is 60 nm, the photocurrent value reaches the maximum of 0.412 mA/cm 2 . And after a continuous light/dark alternation, the optical current is stable at about 0.397 mA/cm 2 . Later, with the increase of the thickness of WO 3 sputtering on the surface of ZnO nanorods, the value of light current began to decrease. When WO 3 is sputtered to a thickness of about 100 nm, the photocurrent value was very close to that of the sample loaded with 20nm WO 3 , and dropped to about 0.353 mA/cm 2 . In Figure 6B, when the sample of only ZnO nanorods on FTO receives light, the electrons absorb energy to jump to the conduction band. Since this sample is a single material, the electrons that arrive at the conduction band will immediately release energy return to the valence band combine with the holes to stabilize, so the combination velocity is very high [66], [67]. At the beginning of sputtering WO 3 , the WO 3 material itself can also absorb energy to enhance the light conversion rate, and the heterojunction formed between ZnO and WO 3 can effectively inhibit the recombination of electron-hole pair, enhance its mobility to increase the photocurrent [68]. With the increase of the thickness of sputtered WO 3 , the increase in the amount of surface WO 3 results in an increase of surface resistance, which lessens the photocurrent ultimately.

D. THE PHOTOELECTRIC PERFORMANCE OF COMPOSITES
When two semiconductors contact to form an interface structure, due to the different band gaps width of the two materials, band steps will be formed at the bottom of the conduction band and the top of the valence band of the two materials. The alignment of the energy bands on both sides of the interface depends on the charge transfer of the materials on both sides. Figure 7 shows a schematic diagram of the photocurrent generated by the ZnO@WO 3 nano heterojunction under light. The heterojunction was used as a photoanode, and the photoelectric response was tested by an electrochemical workstation at a bias voltage of 0.3V. When light strikes both materials, the electrons in the valence band use photon energy to jump into the conduction band and leave holes. At this time, when a bias voltage is applied, the electrons flow into the WO 3 conduction band and flow out through the external Pt electrode. The holes migrate from WO 3 valence band to the ZnO valence band and then are transported together with the holes in the ZnO valence band through the FTO conductive glass to the external circuit. Finally, the external circuit collects and utilizes the carriers flowing out from the two electrodes. Combined with the alternating current (AC) impedance test results show in Figure 7, the photoelectric performance of the sample can be further analyzed. It can be seen from Figure 7 that there is only one semicircle in the spectrum of all samples, so it can be judged that they all have only one time constant. The equivalent circuit diagram is shown in Figure 8. It is observed that the impedance value is very high when there are only ZnO nanorods, but it is greatly reduced when WO 3 is sputtered on the surface of ZnO nanorods to form a heterojunction. As the thickness of WO 3 increases to about 60-80 nm, the impedance value decreases. However, the impedance of the samples began to increase after 80 nm. This is because when a small amount of WO 3 is sputtered on the surface of nanorods, the formation of heterojunction contributes to the improvement of photoelectric performance (as described above). But with the increase of WO 3 , it will cover the nanorods and form heterojunctions in a large area. Moreover, the nanorods will be corroded gradually, so the impedance increase. Therefore, we can sputter WO 3 with a thickness of about 60nm on the ZnO nanorods surface to obtain the best photoelectric performance.

IV. CONCLUSION
In summary, a series of WO 3 @ZnO core-shell nanorods array composites with different amounts of WO 3 were successfully synthesized on FTO conductive glass by combining magnetron sputtering and hydrothermal. The SEM and XRD have been tested the morphology and structure. The study found that WO 3 @ZnO composite has higher photosensitivity compared with pure WO 3 thin films, which can maintain good reversibility and stability under dark/visible light cycle switching. And when a small amount of WO 3 particles are modified on the surface of ZnO nanorods, the photocurrent can be effectively enhanced. WO 3 can change the overall band structures through abating the effective band-gap to extend the light absorption edge of ZnO from the ultraviolet to the visible light region. Therefore, this material can greatly increase the photoelectric conversion efficiency in the application of solar cells and improve the utilization rate of sunlight.