Fabrication of ZnO@Ag@Ag3PO4 Ternary Heterojunction: Superhydrophilic Properties, Antireflection and Photocatalytic Properties

A ZnO seed layer was formed on the fluorine-doped tin oxide substrate by magnetron sputtering, and then a ZnO nanorod was grown on the ZnO seed layer by a hydrothermal method. Next, we prepared a single-crystal Ag seed layer by magnetron sputtering to form a ZnO@Ag composite heterostructure. Finally, Ag3PO4 crystals were grown on the Ag seed layer by a stepwise deposition method to obtain a ZnO@Ag@Ag3PO4 ternary heterojunction. The composite heterostructure of the material has super strong hydrophilicity and can be combined with water-soluble pollutants very well. Besides, it has excellent anti-reflection performance, which can absorb light from all angles. When Ag exists in the heterojunction, it can effectively improve the separation of photo-generated electrons and holes, and improve the photoelectric conversion performance. Based on the above characteristics, this nano-heterostructure can be used in the fields of solar cells, sensors, light-emitting devices, and photocatalysis.


Introduction
With the rapid development of the global economy, people's living standards have been greatly improved. At the same time, environmental and energy issues have become increasingly prominent [1,2]. To alleviate these problems, the development of clean and renewable energy has become a current research hotspot. Among these renewable energy sources, solar energy is a cheap, safe, clean, and easily accessible renewable energy source. Making full use of solar energy can solve environmental and energy problems [3][4][5][6][7][8][9][10].
varistors, and optoelectronic devices [13][14][15][16][17][18][19][20]. Its bandgap is larger than the photon energy (3.10 eV) of visible light at room temperature, resulting in high transmittance in the visible light band, making it an ideal UV light-absorbing material [21]. However, for pure ZnO, the photocatalytic performance is not very good due to the recombination of photo-generated holes (h+) and photo-generated electrons (e − ). In recent reports, it was pointed out that when heterostructured ZnO nanorods exist, they have stronger photocatalytic performance [22]. Recent studies have shown that when Ag is deposited on ZnO nanorods, Ag acts as a bridge and will inhibit the recombination of photogenerated electrons (e − )-h + in ZnO nanorods [23]. Ag 3 PO 4 has a lower valence band potential, generates photogenerated holes (h + ), and therefore has a strong oxidation effect [24,25]. It has excellent photocatalytic properties and can degrade organic pollutants in natural light [26]. The photocatalytic performance of Ag 3 PO 4 crystal under natural light is higher than that of conventional n-doped TiO 2 [27]. Its photocatalytic performance is limited by the separation of photogenerated charges, transmission efficiency, and ability to capture light. In recent reports, many researchers have selected appropriate metals or semiconductors to combine with target semiconductors to form specific multiple heterostructures [28][29][30][31][32][33][34][35], and this heterostructure can effectively regulate the energy band structure, change the distribution of surface charge, improve the photoelectric performance of the entire material. For example, there are reported a series of metal/oxide semiconductor heterostructures in recent reports, such as fluffy worm-like Ag core nanowires (NWs)-ZnO branched nanorods (ZnO BNRs) and dandelion-like Au core nanoparticles (NPs)-ZnO nanorods (ZnO NRs) heterostructures [36,37], and because of their heterostructure characteristics, they show stronger photocatalytic performance than ZnO nanorods. Similarly, by supporting Ag nanoparticles on the Ag 3 PO 4 semiconductor to form a heterogeneous photocatalyst, it can effectively improve the absorption efficiency of natural light, broaden the absorption range of the spectrum, effectively separate photo-generated carriers, and use the generated carriers to participate in redox reactions [38].
In this paper, a ZnO@Ag@Ag 3 PO 4 ternary heterojunction with high photocatalytic performance was prepared by the following steps. Firstly, control the morphology of ZnO and obtain more active sites to participate in the reaction to improve catalytic performance. Secondly, after recombination with silver, due to the different Fermi energy levels, the curved band structure can effectively separate photo-generated carriers and suppress their recombination, while expanding the absorption range of incident light. Finally, the semiconductor-metal interface (Schottky barrier) is formed by the deposition of Ag 3 O 4 , which prevents the recombination of photoinduced electrons(e − )-holes (h + ), and reduces the photo corrosion effect through this interface.

Preparation of ZnO Nanorods
First, the substrate was pre-deposited for 5 minutes at a vacuum of less than 4 × 10 −4 Pa, an argon flow rate of 40 sscm, and a radio frequency (RF) sputtering power of 60 W. Magnetron sputtering instrument (JGP-560B type high vacuum magnetron sputtering coating system, Shenyang Scientific Instrument Co., Ltd., Chinese Academy of Sciences); Then, the substrate (with an fluorine-doped tin oxide (FTO) film on the surface) was aligned with the center of the ZnO target and sputtered for 5 minutes. ZnO deposited on FTO substrate at a thickness of 10 nm per minute, and the seed layer thickness was about 50 nm. After the sample is cooled, the precursor solution (zinc nitrate hexahydrate and hexamethylenetetramine) is placed at 95 • C using a hydrothermal method and reacted for 4 hours to grow ZnO nanorods. The mass ratio of Zn(NO 3 ) 2 ·6H 2 O to C 6 H 12 N 4 was 1: 1, and the experimental concentrations were 10, 20, 40 and 120 mmol/L, respectively. Finally, the samples were rinsed with deionized water more than three times and dried at room temperature. We label the sample with an experimental concentration of 10 mmol/L as 1#. The sample with an experimental concentration of 20 mmol/L is labeled 2#, the sample with an experimental concentration of 40 mmol/L is labeled 3#, and the sample with an experimental concentration of 120 mmol/L is labeled 4#, and the best-shaped samples were selected by the SEM test for the next step experiment.

Preparation of Ag Layer by Magnetron Sputtering
We use the results obtained in the previous experiment as the substrate for subsequent experiments (3#), and then a metal Ag layer was sputtered by magnetron. The vacuum was 4 × 10 −4 Pa, and the argon flow rate was 40 sscm, the DC sputtering power was 45 W. We first pre-sputtered for 2 minutes and then officially sputtered each sample for 90 s. After the sputtering experiment was completed, the samples were taken out after cooling for 60 minutes and labeled 5#.

Preparation of Ag 3 PO 4 by Step Deposition
In this section, AgNO 3 and Na 2 HPO 4 solutions were used as the deposition solution [39]. The concentrations of both solutions were 0.02 mol/L, and the volumes were 400 mL and 220 mL, respectively. The number of depositions is a set of variables, and the deposition process was as follows: 5# was placed in the AgNO 3 solution for 30 min, moved to Na 2 HPO 4 solution for 5 min, and finally washed with deionized water. All the above steps are regarded as one deposition, and different samples will be obtained for different deposition times. The deposition times of the samples were 1 time, 3 times, 5 times, and 7 times, respectively, and they were marked as 6#, 7#, 8#, and 9# after drying at room temperature.

Photocatalysis Experiment
The Rhodamine B (RhB) was selected as a photocatalytic degradation material and the concentration was 10 ppm. First, 50 mL RhB solution was transferred into a 100 mL beaker at each experiment, and the sample (7.0 × 1.5 cm 2 ) was tilted face down on the side of the cup wall. To maintain the adsorption equilibrium, the samples were allowed to stand in RhB solution for 30 min before the experiment. Then FX-300 fiber optic light source (25 mW/cm 2 ) was used for subsequent exposure, the light absorption test was performed by liquid UV-visible spectrophotometer (UV-2600A), the content of the degraded organic matter was detected by light absorption. To analyze the degradation process of the reaction solution, the RhB concentration spectrum was recorded once every 30 min by recording the absorption spectrum of RhB.

Experimental Instruments and Sample Characterization
Magnetron sputtering instrument (JGP-560B type high vacuum magnetron sputtering coating system, Shenyang Scientific Instrument Co., Ltd., Shenyang, China) The samples were analyzed by X-ray diffractometry (D/max-1400, Japanese Science, Tokyo Metropolis, Japan). scanning electron microscope (ULTRA55, Zeiss, Oberkochen, Germany) was used to study the surface structure and morphological characteristics of composite samples to analyze the morphology of samples; X-ray photoelectron spectroscopy (XPS) was used to record the chemical states of the elements on a PHI-5702 multi-functional X-ray photoelectron spectrometer (Physical Electronics, Chanhassen, MN, USA); ultraviolet-visible spectrophotometer (UV-2600A, Unico (Shanghai) Instrument Co., Ltd., Shanghai, China) was used to perform light absorption test on the degraded RhB solution; the contact angle tester (DSA30, Kruss, Germany) was used to analyze the hydrophilic and hydrophobic properties of the sample. The test contact angle range is 0 • -180 • , and the resolution is ±0.01 • .

Results and Analysis
ZnO nanorods are one-dimensional nanomaterials whose structure and morphology have a crucial effect on performance [40]. Figure 1 shows the SEM images of ZnO nanorods under different concentrations. Figure 1a is the SEM image of 1# growth concentration 10 mmol/L. It can be seen that the nanorods in the formed nanorod array are extremely short and unevenly distributed. It was because the concentration of the precursor was too low to crystallization it. Figure 1b shows the SEM image of 2# growth concentration 20 mmol/L. It can be seen that with the increase of the solution concentration, the length and diameter of the nanorods gradually increase, and some hexagonal rod-like structures are gradually formed, but the overall is still uneven. The SEM images of 3# growth concentration 40 mmol/L have been shown in Figure 1c, it was shown that the growth rate and aspect ratio of ZnO nanorods were increased. It meant that the nucleation site increase during hydrothermal was beneficial for the production of ZnO nanorods with high quality and density. Figure 1d is an SEM image of 4# growth concentration 120 mmol/L. It is seen that there is adhesion between the nanorods due to the too high concentration of the precursor solution. Previous research reports have found that the morphology of ZnO has a large effect on performance [40][41][42]. When the precursor concentration was 40 mmol/L, there was no adhesion between the nanorods, and they were evenly distributed. ZnO nanorods are one-dimensional nanomaterials whose structure and morphology have a crucial effect on performance [40]. Figure 1 shows the SEM images of ZnO nanorods under different concentrations. Figure 1a is the SEM image of 1# growth concentration 10 mmol/L. It can be seen that the nanorods in the formed nanorod array are extremely short and unevenly distributed. It was because the concentration of the precursor was too low to crystallization it. Figure 1b shows the SEM image of 2# growth concentration 20 mmol/L. It can be seen that with the increase of the solution concentration, the length and diameter of the nanorods gradually increase, and some hexagonal rod-like structures are gradually formed, but the overall is still uneven. The SEM images of 3# growth concentration 40 mmol/L have been shown in Figure 1c, it was shown that the growth rate and aspect ratio of ZnO nanorods were increased. It meant that the nucleation site increase during hydrothermal was beneficial for the production of ZnO nanorods with high quality and density. Figure 1d is an SEM image of 4# growth concentration 120 mmol/L. It is seen that there is adhesion between the nanorods due to the too high concentration of the precursor solution. Previous research reports have found that the morphology of ZnO has a large effect on performance [40][41][42]. When the precursor concentration was 40 mmol/L, there was no adhesion between the nanorods, and they were evenly distributed. The performance of ZnO nanorods is not ideal, and the photo-generated carrier recombination speed is fast. To suppress the recombination of photogenerated carriers, 3# having the best morphology was selected, and an Ag layer with a thickness of 100 nm was sputtered on the 3# surface. Figure 2 shows the SEM images of 5#. It can be clearly seen that, after the magnetron sputtering of the Ag layer, there was a thin film outline on the ZnO nanorods' surface. This meant that the Ag layer attached to the ZnO nanorods' surface successfully. Figure 3 was the SEM images of 6#, 7#, 8#, and 9#. It can be seen that when the number of depositions is one, there are only a small amount of Ag3PO4 nanoparticles on the surface. When the number of deposition times is three, there are a relatively uniformly distribution of Ag3PO4 nanoparticles on the surface. As the number of depositions increases to five, many Ag3PO4 nanoparticles of the same shape and size appear on the surface. When the time of depositions increased to seven, we can see from the picture that the grown Ag3PO4 nanoparticles are too large and irregular in shape. The performance of ZnO nanorods is not ideal, and the photo-generated carrier recombination speed is fast. To suppress the recombination of photogenerated carriers, 3# having the best morphology was selected, and an Ag layer with a thickness of 100 nm was sputtered on the 3# surface. Figure 2 shows the SEM images of 5#. It can be clearly seen that, after the magnetron sputtering of the Ag layer, there was a thin film outline on the ZnO nanorods' surface. This meant that the Ag layer attached to the ZnO nanorods' surface successfully. Figure 3 was the SEM images of 6#, 7#, 8#, and 9#. It can be seen that when the number of depositions is one, there are only a small amount of Ag 3 PO 4 nanoparticles on the surface. When the number of deposition times is three, there are a relatively uniformly distribution of Ag 3 PO 4 nanoparticles on the surface. As the number of depositions increases to five, many Ag 3 PO 4 nanoparticles of the same shape and size appear on the surface. When the time of depositions increased to seven, we can see from the picture that the grown Ag 3 PO 4 nanoparticles are too large and irregular in shape.   Figure 4a shows the X-Ray Diffraction pattern (XRD) of ZnO nanorods grown on FTO conductive glass (3#) and the XRD test of 5#. It was shown that there were four characteristics of the 3# at 31.75°, 34.39°, 36.23°, and 62.80° which correspond to its (100), (002), (101), and (103) crystal planes of ZnO. Among these diffraction peaks, the (002) diffraction peaks crystal plane was the strongest, indicated that ZnO growth along the (002) crystal plane preferentially, this was because the surface free energy of the (002) crystal plane was lowest [43,44]. For 5#, there was a characteristic diffraction peak of Ag (111) at 38.36° in figure 4a, this was due to the Ag layer on 5# [45,46]. A shift of diffraction pattern peaks is attributed to the fact that when some silver was sputtered onto the ZnO nano-surface by magnetron sputtering, it entered into the ZnO lattice, which caused the plane reflection X-rays of ZnO in samples 3 and 5 to be different. Figure 4b shows the XRD diffraction patterns of 6#, 7#, 8#, and 9#. It can be seen there were other nine diffraction peaks corresponding to the (110), (200), (210), (102), (310), (322), (320), and (321) crystal planes of Ag3PO4, respectively [47,48]. Moreover, the intensity of these peaks increases as the number of deposits increases. This is because when the number of deposits is small, there are few Ag3PO4 nanoparticles, so the characteristic diffraction peaks are not visible. As the number of deposits   Figure 4a shows the X-Ray Diffraction pattern (XRD) of ZnO nanorods grown on FTO conductive glass (3#) and the XRD test of 5#. It was shown that there were four characteristics of the 3# at 31.75°, 34.39°, 36.23°, and 62.80° which correspond to its (100), (002), (101), and (103) crystal planes of ZnO. Among these diffraction peaks, the (002) diffraction peaks crystal plane was the strongest, indicated that ZnO growth along the (002) crystal plane preferentially, this was because the surface free energy of the (002) crystal plane was lowest [43,44]. For 5#, there was a characteristic diffraction peak of Ag (111) at 38.36° in figure 4a, this was due to the Ag layer on 5# [45,46]. A shift of diffraction pattern peaks is attributed to the fact that when some silver was sputtered onto the ZnO nano-surface by magnetron sputtering, it entered into the ZnO lattice, which caused the plane reflection X-rays of ZnO in samples 3 and 5 to be different. Figure 4b shows the XRD diffraction patterns of 6#, 7#, 8#, and 9#. It can be seen there were other nine diffraction peaks corresponding to the (110), (200), (210), (102), (310), (322), (320), and (321) crystal planes of Ag3PO4, respectively [47,48]. Moreover, the intensity of these peaks increases as the number of deposits increases. This is because when the number of deposits is small, there are few Ag3PO4 nanoparticles, so the characteristic diffraction peaks are not visible. As the number of deposits  Among these diffraction peaks, the (002) diffraction peaks crystal plane was the strongest, indicated that ZnO growth along the (002) crystal plane preferentially, this was because the surface free energy of the (002) crystal plane was lowest [43,44]. For 5#, there was a characteristic diffraction peak of Ag (111) at 38.36 • in Figure 4a, this was due to the Ag layer on 5# [45,46]. A shift of diffraction pattern peaks is attributed to the fact that when some silver was sputtered onto the ZnO nano-surface by magnetron sputtering, it entered into the ZnO lattice, which caused the plane reflection X-rays of ZnO in samples 3 and 5 to be different. Figure 4b shows the XRD diffraction patterns of 6#, 7#, 8#, and 9#. It can be seen there were other nine diffraction peaks corresponding to the (110), (200), (210), (102), (310), (322), (320), and (321) crystal planes of Ag 3 PO 4 , respectively [47,48]. Moreover, the intensity of these peaks increases as the number of deposits increases. This is because when the number of deposits is small, there are few Ag 3 PO 4 nanoparticles, so the characteristic diffraction peaks are not visible. As the number of deposits increases, the amount of Ag 3 PO 4 nanoparticles increases, and the intensity of diffraction peaks related to Ag 3 PO 4 gradually increases. These results showed that ZnO@Ag@Ag 3 PO 4 has been prepared successfully.
Micromachines 2020, 11, x 6 of 12 increases, the amount of Ag3PO4 nanoparticles increases, and the intensity of diffraction peaks related to Ag3PO4 gradually increases. These results showed that ZnO@Ag@Ag3PO4 has been prepared successfully. X-ray photo-e − spectroscopy (XPS) can provide more info about the chemical state of ZnO@Ag@Ag3PO4. Figure 5 shows the XPS of 7#. It can be seen there are Zn, Ag, P, and O in 7#. The position of the Zn2p band for ZnO@Ag@Ag3PO4 (Figure 5a) was at about 1022.1 eV, which meant that the Zn element was the form of Zn 2+ here [49]. Figure 5b shows the detailed spectra for Ag. It can be seen there were two bands at 368.05 and 374.15 eV which ascribed the Ag 3d5/2 (metallic Ag0 in Ag layer) and Ag 3d3/2 (Ag+ in Ag3PO4), respectively; these results were similar with related reports [50,51]. The Ag 3d5/2 and Ag 3d3/2 bands could be further divided into two different binds at 368.0, 367.5 eV and 374.0, 373.4 eV, respectively [52]. These results indicated there was metallic Ag0 in ZnO@Ag@Ag3PO4. The peak at 531.55 eV was related to lattice oxygen (O 1s) in ZnO and Ag3PO4 and the peak at 133.6 eV corresponds to the binding energy of phosphorus 2p (P 2p) in Ag3PO4. From previous reports, there was a minor difference in these data, which was attributed to the ternary heterojunction.  Figures 6a,b show the static contact angles of 3# and 5#, respectively. It was shown that the static contact angle of ZnO was 13.33°. When a 100 nm Ag layer is present on the surface, the static contact angle increases from 13.33° to 25.43°, and the surface hydrophilicity decreases. The reason is that ZnO is a metal oxide, which itself has a certain hydrophilicity. However, Ag is not easily X-ray photo-e − spectroscopy (XPS) can provide more info about the chemical state of ZnO@Ag@Ag 3 PO 4 . Figure 5 shows the XPS of 7#. It can be seen there are Zn, Ag, P, and O in 7#. The position of the Zn2p band for ZnO@Ag@Ag 3 PO 4 (Figure 5a) was at about 1022.1 eV, which meant that the Zn element was the form of Zn 2+ here [49]. Figure 5b shows the detailed spectra for Ag. It can be seen there were two bands at 368.05 and 374.15 eV which ascribed the Ag 3d5/2 (metallic Ag0 in Ag layer) and Ag 3d 3/2 (Ag+ in Ag 3 PO 4 ), respectively; these results were similar with related reports [50,51]. The Ag 3d5/2 and Ag 3d3/2 bands could be further divided into two different binds at 368.0, 367.5 eV and 374.0, 373.4 eV, respectively [52]. These results indicated there was metallic Ag 0 in ZnO@Ag@Ag 3 PO 4 . The peak at 531.55 eV was related to lattice oxygen (O 1s) in ZnO and Ag 3 PO 4 and the peak at 133.6 eV corresponds to the binding energy of phosphorus 2p (P 2p) in Ag 3 PO 4. From previous reports, there was a minor difference in these data, which was attributed to the ternary heterojunction.
Micromachines 2020, 11, x 6 of 12 increases, the amount of Ag3PO4 nanoparticles increases, and the intensity of diffraction peaks related to Ag3PO4 gradually increases. These results showed that ZnO@Ag@Ag3PO4 has been prepared successfully. X-ray photo-e − spectroscopy (XPS) can provide more info about the chemical state of ZnO@Ag@Ag3PO4. Figure 5 shows the XPS of 7#. It can be seen there are Zn, Ag, P, and O in 7#. The position of the Zn2p band for ZnO@Ag@Ag3PO4 (Figure 5a) was at about 1022.1 eV, which meant that the Zn element was the form of Zn 2+ here [49]. Figure 5b shows the detailed spectra for Ag. It can be seen there were two bands at 368.05 and 374.15 eV which ascribed the Ag 3d5/2 (metallic Ag0 in Ag layer) and Ag 3d3/2 (Ag+ in Ag3PO4), respectively; these results were similar with related reports [50,51]. The Ag 3d5/2 and Ag 3d3/2 bands could be further divided into two different binds at 368.0, 367.5 eV and 374.0, 373.4 eV, respectively [52]. These results indicated there was metallic Ag0 in ZnO@Ag@Ag3PO4. The peak at 531.55 eV was related to lattice oxygen (O 1s) in ZnO and Ag3PO4 and the peak at 133.6 eV corresponds to the binding energy of phosphorus 2p (P 2p) in Ag3PO4. From previous reports, there was a minor difference in these data, which was attributed to the ternary heterojunction.  Figures 6a,b show the static contact angles of 3# and 5#, respectively. It was shown that the static contact angle of ZnO was 13.33°. When a 100 nm Ag layer is present on the surface, the static contact angle increases from 13.33° to 25.43°, and the surface hydrophilicity decreases. The reason is that ZnO is a metal oxide, which itself has a certain hydrophilicity. However, Ag is not easily  Figure 6a,b show the static contact angles of 3# and 5#, respectively. It was shown that the static contact angle of ZnO was 13.33 • . When a 100 nm Ag layer is present on the surface, the static contact angle increases from 13.33 • to 25.43 • , and the surface hydrophilicity decreases. The reason is that ZnO is a metal oxide, which itself has a certain hydrophilicity. However, Ag is not easily oxidized into a polar metal oxide in the air, and itself has certain hydrophobicity. For 7#, it could be seen from Figure 6c that there was complete wetness and the static contact angle was 0 • on the surface-this is because the Ag 3 PO 4 was an inorganic metal salt and was hydrophilic.
Micromachines 2020, 11, x 7 of 12 oxidized into a polar metal oxide in the air, and itself has certain hydrophobicity. For 7#, it could be seen from Figure 6c that there was complete wetness and the static contact angle was 0° on the surface-this is because the Ag3PO4 was an inorganic metal salt and was hydrophilic.  Figure 7 shows the light reflection of 3#, 5#, and 7#. It can be seen that there was reflection from 380 nm to 800 nm of 3#, 5#, and 7#. The reflectance of 3# was approximately 50.94%. The reflectance of 5# was 16.52% and was 34.42% lower than that of ZnO, which meant that the Ag layer in this material was beneficial to light absorption [53][54][55]. The reflectivity of 7# was 8.24%, and was 8.28% lower than 5#, meaning that core-shell composites had a high light absorption.  Figure 8a shows the photocatalytic test results of 3#, 5#, and 10# (5# without Ag layer). The results show that the photocatalytic performance of 5# is higher than that of 3# and 10#, which means that when the surface has an Ag layer, the photocatalytic performance can be improved. The reason is that when there is an Ag layer, the photogenerated e --h + will be separated and utilized due to the potential, thereby reducing its recombination rate. The separation efficiency of the photo-generated carriers was added [56][57][58]. Simultaneously, the Ag layer played an important role in collecting eby acting as an eacceptor; it could separate photogenerated carriers effectively and inhibited the recombination of photo-generated carriers produced by ZnO, photocatalytic reduction, and the oxidation reactions of eand h + . Figure 8b is the photocatalytic test results of 6#, 7#, 8#, and 9#. It can be seen that with the increase of the number of depositions, the photocatalytic performance increases. When the number of depositions is three times (7#), the photocatalytic performance reaches the best. This is because the Ag3PO4 nanoparticles on the 7# surface are relatively uniformly distributed and are hydrophilic, which can better contact and react with the substances in the solution. Besides, lower reflectivity  Figure 7 shows the light reflection of 3#, 5#, and 7#. It can be seen that there was reflection from 380 nm to 800 nm of 3#, 5#, and 7#. The reflectance of 3# was approximately 50.94%. The reflectance of 5# was 16.52% and was 34.42% lower than that of ZnO, which meant that the Ag layer in this material was beneficial to light absorption [53][54][55]. The reflectivity of 7# was 8.24%, and was 8.28% lower than 5#, meaning that core-shell composites had a high light absorption.
Micromachines 2020, 11, x 7 of 12 oxidized into a polar metal oxide in the air, and itself has certain hydrophobicity. For 7#, it could be seen from Figure 6c that there was complete wetness and the static contact angle was 0° on the surface-this is because the Ag3PO4 was an inorganic metal salt and was hydrophilic.  Figure 7 shows the light reflection of 3#, 5#, and 7#. It can be seen that there was reflection from 380 nm to 800 nm of 3#, 5#, and 7#. The reflectance of 3# was approximately 50.94%. The reflectance of 5# was 16.52% and was 34.42% lower than that of ZnO, which meant that the Ag layer in this material was beneficial to light absorption [53][54][55]. The reflectivity of 7# was 8.24%, and was 8.28% lower than 5#, meaning that core-shell composites had a high light absorption.  Figure 8a shows the photocatalytic test results of 3#, 5#, and 10# (5# without Ag layer). The results show that the photocatalytic performance of 5# is higher than that of 3# and 10#, which means that when the surface has an Ag layer, the photocatalytic performance can be improved. The reason is that when there is an Ag layer, the photogenerated e --h + will be separated and utilized due to the potential, thereby reducing its recombination rate. The separation efficiency of the photo-generated carriers was added [56][57][58]. Simultaneously, the Ag layer played an important role in collecting eby acting as an eacceptor; it could separate photogenerated carriers effectively and inhibited the recombination of photo-generated carriers produced by ZnO, photocatalytic reduction, and the oxidation reactions of eand h + . Figure 8b is the photocatalytic test results of 6#, 7#, 8#, and 9#. It can be seen that with the increase of the number of depositions, the photocatalytic performance increases. When the number of depositions is three times (7#), the photocatalytic performance reaches the best. This is because the Ag3PO4 nanoparticles on the 7# surface are relatively uniformly distributed and are hydrophilic, which can better contact and react with the substances in the solution. Besides, lower reflectivity  Figure 8a shows the photocatalytic test results of 3#, 5#, and 10# (5# without Ag layer). The results show that the photocatalytic performance of 5# is higher than that of 3# and 10#, which means that when the surface has an Ag layer, the photocatalytic performance can be improved. The reason is that when there is an Ag layer, the photogenerated e − -h + will be separated and utilized due to the potential, thereby reducing its recombination rate. The separation efficiency of the photo-generated carriers was added [56][57][58]. Simultaneously, the Ag layer played an important role in collecting e − by acting as an e − acceptor; it could separate photogenerated carriers effectively and inhibited the recombination of photo-generated carriers produced by ZnO, photocatalytic reduction, and the oxidation reactions of e − and h + .

4#
ZnO Adhesion between nanorods 5# ZnO@Ag UV-visible light reflection is higher than pure ZnO, but hydrophobic 6# ZnO@Ag@Ag3PO4 UV-visible light reflection is higher than the ZnO@Ag, but hydrophilic 7# ZnO@Ag@Ag3PO4 The photocatalytic performance reaches the best. 8# ZnO@Ag@Ag3PO4 UV-visible light reflection is higher than the ZnO@Ag, but hydrophilic 9# ZnO@Ag@Ag3PO4 UV-visible light reflection is higher than the ZnO@Ag, but hydrophilic Based on the above results, the reason why the performance of ZnO@Ag@Ag3PO4 heterostructures was better than Ag3PO4 nanoparticle and pure ZnO nanorods is discussed in more detail. The Figure 9 illustrated the band structure of ZnO@Ag@Ag3PO4 heterostructures. When ZnO@Ag@Ag3PO4 was located at an organic solution and radiated by UV light while photon energy was equal to or higher than the bandgap of ZnO nanorods, the ein the valence band (VB) could be excited to the conduction band (CB) and the h + would be generated in VB with the same number. As illustrated in Figure 9, the new Fermi energy level of heterostructure was lower than the energy level at the bottom of CB, the photoexcited ecould be moved from ZnO nanorods to Ag through above potential. At the same time, there is an appropriate Fermi level in Ag, which could serve as an excellent eacceptor for facilitating the transfer of efrom Ag3PO4 nanoparticle quickly at nature light irradiation [59][60][61]. Thereby, the photoexcited ecould be migrated to Ag quickly, and the photoinduced h + still situated on Ag3PO4 nanoparticle and ZnO nanorods, which promoted the separation of photoexcited e − -h + effectively. As illustrated in Figure 6b, the enriched eon Ag nanocrystals could promote the multiple-ereduction reaction effectively, the corresponding reaction formula is as follows [62][63][64], O2 + 4H + + 4e − → 2H2O.  Figure 8b is the photocatalytic test results of 6#, 7#, 8#, and 9#. It can be seen that with the increase of the number of depositions, the photocatalytic performance increases. When the number of depositions is three times (7#), the photocatalytic performance reaches the best. This is because the Ag 3 PO 4 nanoparticles on the 7# surface are relatively uniformly distributed and are hydrophilic, which can better contact and react with the substances in the solution. Besides, lower reflectivity means more light energy is converted into chemical energy. The 7# had the best photocatalytic properties. Figure 8c is the continuous photocatalytic test of 7#. The results showed that with the increase in degradation time, the prepared samples continued to degrade RhB. The detailed sample number and related characteristics are shown in Table 1. Based on the above results, the reason why the performance of ZnO@Ag@Ag 3 PO 4 heterostructures was better than Ag 3 PO 4 nanoparticle and pure ZnO nanorods is discussed in more detail. The Figure 9 illustrated the band structure of ZnO@Ag@Ag 3 PO 4 heterostructures. When ZnO@Ag@Ag 3 PO 4 was located at an organic solution and radiated by UV light while photon energy was equal to or higher than the bandgap of ZnO nanorods, the e − in the valence band (VB) could be excited to the conduction band (CB) and the h + would be generated in VB with the same number. As illustrated in Figure 9, the new Fermi energy level of heterostructure was lower than the energy level at the bottom of CB, the photoexcited e − could be moved from ZnO nanorods to Ag through above potential. At the same time, there is an appropriate Fermi level in Ag, which could serve as an excellent e − acceptor for facilitating the transfer of e − from Ag 3 PO 4 nanoparticle quickly at nature light irradiation [59][60][61]. Thereby, the photoexcited e − could be migrated to Ag quickly, and the photoinduced h + still situated on Ag 3 PO 4 nanoparticle and ZnO nanorods, which promoted the separation of photoexcited e − -h + effectively. As illustrated in Figure 6b, the enriched e − on Ag nanocrystals could promote the multiple-e − reduction reaction effectively, the corresponding reaction formula is as follows [62][63][64], O 2 + 4H + + 4e − → 2H 2 O.

Conclusions
In summary, the ZnO@Ag@Ag3PO4 ternary nano-heterostructure was obtained by magneto-sputtering, hydrothermal method, and deposition successively. The SEM, XRD, and XPS have been tested the morphology, structure, and elements. The study found that the morphology of ZnO nanorods was the best when the concentration of the precursor solution was 40 mmol/L. The presence of Ag in the heterojunction can effectively improve the separation of photoinduced electrons and holes, and therefore has excellent photoelectric conversion properties. The hydrophilicity increased and light reflectance decreased after the deposition of Ag3PO4.

Conclusions
In summary, the ZnO@Ag@Ag 3 PO 4 ternary nano-heterostructure was obtained by magneto-sputtering, hydrothermal method, and deposition successively. The SEM, XRD, and XPS have been tested the morphology, structure, and elements. The study found that the morphology of ZnO nanorods was the best when the concentration of the precursor solution was 40 mmol/L. The presence of Ag in the heterojunction can effectively improve the separation of photoinduced electrons and holes, and therefore has excellent photoelectric conversion properties. The hydrophilicity increased and light reflectance decreased after the deposition of Ag 3 PO 4 .

Conflicts of Interest:
The authors declare no conflict of interest.