Water pollution has become one of the serious problems that threaten human life. Photocatalytic degradation is a promising green technology to remove organic and inorganic pollutants from the water. A lot of metal oxide semiconductors have been explored for photocatalytic pollutant degradation   . Zinc oxide (ZnO) semiconductor is extensively used as a photocatalyst in the photodegradation of organic and inorganic pollutants  . However, there are several drawbacks linked with the higher rates of charge carrier and recombination, and this can only absorb a small percentage of the solar light in the UV region owing to the wide band gap  . Hence, it restricts photocatalytic applications under the visible-light illumination  and it has posed a challenge to researchers for the effective utilization in the photocatalytic applications. In order to overcome these limitations, researchers did many attempts to tailor the band gap of ZnO including intrinsic   and extrinsic   doping and coupling ZnO with other semiconductor systems. It is well-known that the introductions of extrinsic or intrinsic defects, which cause defect related states in the form of band tail, promote the extension of photonic absorption to the broader visible region or band gap narrowing. Coupling ZnO with other narrow band gap semiconductors like CuO, Fe2O3, WO3, CdS, ZnS, etc. can reduce the band gap and extend the light absorption range of ZnO to a broader spectrum of the visible light, and promote electron-hole pair separation under sunlight irradiation in order to achieve higher photocatalytic activity  . The present authors have worked on coupled ZnO with CeO2 and Bi2O3 to construct Bi2O3-CeO2-ZnO heterostructure  . Bi2O3-CeO2-ZnO exhibited improved photocatalytic activity due to the extended visible-light photoabsorption range benefited from the presence of Bi2O3 (small band gap of 2.77 eV) and efficient separation of photo-induced electrons and holes caused by the vector transfer of electrons and holes among ZnO, CeO2 and Bi2O3  . On the other hand, usually the photocatalytic activity of a photocatalyst is strongly influenced by its surface area. The higher surface area, the higher number of active photocatalytic sites, the higher adsorption capability for the pollutants on the photocatalyst surface, and the higher photocatalytic activity    . Although there are several methods available to process these metal oxides, but hydrothermal technique is becoming very popular in recent years, because of its environmentally benign and facilitates in situ surface modification, which provides fabrication of nanoparticles with desired tailor-made properties     . In the present work, an attempt has been made to understand a unique comparative study on the effects of surface area, CB and VB positions and band gap of ZnO based metal oxide (ZnO/MxOy) heterostructures for their photocatalytic activities, whereas ZnO/MxOy are ZnO/MnO2, ZnO/Co3O4, ZnO/NiO and ZnO/In2O3. The photocatalytic efficiency of these coupled semiconductor systems have been tested by monitoring phenol degradation under solar irradiation.
2. Experimental Details
The ingredients were procured through different commercial suppliers as follows: Zn(NO3)2∙6H2O (Alfa Aesar, 99%), Ni(NO3)2∙6H2O (Sigma Aldrich, 99.99%, Co(NO3)2∙6H2O (Loba Chemie, 99.8%), In(NO3)∙xH2O (Sigma-Aldrich, 99.99%), Mn(NO3)∙4H2O (Alfa Aesar, 98%) and NaOH (Sigma-Aldrich, 97%). Ultra-pure water (18. 2 MΩ∙cm) was used in the experiments.
2.2. Hydrothermal Synthesis of ZnO/MxOy Nanocomposite
The synthesis of ZnO/MxOy nanocomposites were carried out by using 15 ml 0.1M Zn(NO3)2∙6H2O mixed with 15 ml 0.1M along with a desired metal nitrate (Ni(NO3)2∙6H2O, Co(NO3)2∙6H2O, Mn(NO3)2∙4H2O or In(NO3)∙xH2O). The mixed solution was ultrasonicated thoroughly and mixed for 30 minutes to get the homogenous mixture. Later, about 24 ml of the suspension was transferred into a 30 ml capacity Teflon liner and closed tightly. The percent fill was kept at 80% in all experiments. The Teflon liner was inserted into SS316 Stainless Steel autoclave and heated up to a particular temperature for a required period of time (Table 1). The autoclave was cooled down to the laboratory temperature after hydrothermal reaction and the resultant material in the Teflon liner was filtered and washed with distilled water and ethanol. Later, the product obtained was dried at around 80˚C in a hot air oven for 12 hours. Table 1 gives the experimental hydrothermal conditions used in the synthesis of ZnO-MxOy heterostructure nanocomposites.
The prepared heterostructures were characterized using Rigaku Smart Lab-II X-ray diffractometer (XRD) with CuKα radiation (λ = 1.540598 Å), Carl Zeiss MERLIN Compact field emission scanning electron microscopy (FE-SEM), JASCO 460 plus FTIR spectrometer, a Shimadzu UV-2450 spectrophotometer (UV-Vis DRS), NETZSCH, Germany, Model STA 2500 Regulus, thermo-gravimetric analysis (TGA), and BELSORP MINI 2, BEL, Japan, Brunauer-Emmett-Teller (BET) surface area measurements.
2.4. Photocatalytic Test
The photocatalytic properties of as synthesized ZnO-MxOy nanocomposites were carried out for the study of phenol degradation. The degradation was performed under solar light irradiation to the phenol sample with the catalyst ZnO-MxOy nanocomposites. The phenol samples with different concentrations, like 8, 16, 24 and 32 ppm, were prepared and the optimum phenol degradation concentration was investigated for the different catalysts. In typical procedure, the degradation
Table 1. Experimental conditions used in the fabrication of MxOy nanocomposites.
was carried out by preparing 100 ml of phenol solution with above mentioned different concentration in a 250 cm3 conical flasks. Later, 0.5 g of the catalysts such as ZnO-MxOy nanocomposites were added to individual flasks and closed tightly with rubber cork. The degradation reaction was carried out for the period of 3 h (from 11 am to 2 am in a sunny day in April 2018 in Mysuru, India) under solar light irradiation. After the completion of degradation reaction, the samples were filtered and the percentage of phenol degradation was investigated by performing chemical oxygen demand (COD) analysis.
3. Results and Discussion
3.1. The Powder XRD Analysis
Figure 1 shows the powder XRD patterns of pure ZnO and heterogeneous semiconductor ZnO-MxOy nanocomposites. It is clear indicates that, all the diffraction peaks of ZnO can be indexed to the hexagonal wurtzite structure according to JCPDS card N (JCPDS: 36-1451)  and peaks corresponds to heterostructured composites consist of different phases. The XRD pattern of heterostructured ZnO-Co3O4 composite indicates that it consists of two kinds of phases, hexagonal ZnO and spinel structured cubic Co3O4 (JCPDS card # 43-1003) confirming the coexistence of Zn and Co oxide structure together to form a composite however, the peaks of Co3O4 are weak compared with ZnO and it is matched with the data reported by Zhang et al.  . The XRD pattern of heterostructured ZnO-NiO composite indicates that it consists of two phases, hexagonal ZnO and cubic structured rocksalt type NiO phase (JCPDS-78-0643) showing the coexistence of Zn and Ni oxide structures together to form heterogeneous composite. The XRD pattern of heterostructured ZnO-MnO2 composite specifies that it consists of two kind of phases, hexagonal ZnO and birnessite δ-MnO2 crystal phase (JCPDS 80-1098)  which indicates that the coexistence of Zn and Mn oxide structures together to form heterogeneous heterostructure
Figure 1. Powder X-ray diffraction patterns characteristic samples of ZnO-MxOy nanocomposites.
composite  . The powder XRD pattern of heterostructured ZnO-In2O3 composite postulates that it consists of two kind of phases, hexagonal ZnO and bixbyite type In2O3 of cubic system crystal phase indicates that the coexistence of Zn and In oxide structures together to form heterogeneous composite. The XRD pattern of ZnO-In2O3 is in agreement with the earlier reported pattern  . The intensity of standard ZnO decreased in all heterogeneous composites, due to the major reflections of other metal oxides which were entrapped in intense ZnO reflections. This mechanism leads to the growth of heterogeneous composites with different crystal phases. The powder patterns show well crystallized metal oxides with a higher degree of crystallinity forming MxOy nanocomposites.
3.2. FTIR Analysis
The FTIR analysis of heterostructured ZnO-MxOy nanocomposites was performed. Figure 2 shows the FTIR spectra and the bands are observed near 3445 cm−1 and 1630 cm−1 correspond to the O-H stretching and bending vibration of water molecules adsorbed on the surface of nanocomposites due to the porosity as noticed in the images of electron microscopic study.
The carbonate peaks are prominent in all the composites and it is again attributed to the absorption of CO2 from the atmosphere. The finger print region of metal oxygen bond shows the characteristic absorption bands in the wavenumber region 400 to 600 cm−1  . This finger print range shows the metal oxygen bond presence of different metal oxygen of ZnO-MxOy nanocomposites. Usually ZnO shows flower like, or interlocked rod like structure, and other metal oxides like CO3O4 , In2O3, NiO, and MnO2, show a different morphology (either tabular, or massive rounded to sub-rounded). The bonding of composites have metal and oxygen bond between Zn-O, Ni-O, Mn-O, In-O and Co-O along with the absorption peak at wavenumber range 1402 - 1506 cm−1 indicates the presence of CO2, which was due to the atmospheric CO2 interference during analysis   . Apparently, the FTIR results further support the formation
Figure 2. Characteristic FTIR spectra of ZnO-MxOy nanocomposites.
of the heterostructured nanocomposites and agree well with XRD results.
3.3. FESEM Analysis
Figure 3 shows FE-SEM images of ZnO-MxOy nanocomposites and it is observed that ZnO-Co3O4 and ZnO-In2O3 show fine interlocked nanorods showing porous structure, whereas ZnO-NiO and ZnO-MnO2 show still finer structures with smaller rods and also showing the interlocked porous structure. This is a very typical characteristic of ZnO-MxOy nanocomposites. It is further observed that the particles have narrow-size distribution. The average particle size distribution of ZnO-MxOy nanocomposites is shown in Figure 4. Among the nanocomposites synthesized, ZnO-In2O3 have highest particle size (140 nm) and ZnO-NiO have the least particle size (40 nm). This kind of interlocked structure shows higher porosity and in turn higher BET surface area. It is also evident from FTIR spectra that, ZnO-MxOy nanocomposites have the surface adsorption of CO2 and O-H.
Figure 3. FESEM images of characteristic ZnO/MxOy nanocomposites.
Figure 4. Particle size distribution of ZnO-MxOy nanocomposites.
3.4. DLS Analysis
Particle size distributions of as prepared nanocomposites are shown in Figure 4. The analysis was carried out by DLS instrument, in which the size measurement was based on the Brownian motion of particle in the solvent media. The results as evident from Figure 4, indicate that the average particle size (nm) of the nanocomposites prepared by hydrothermal reaction conditions. The average particle size of pure ZnO, composites ZnO-Co3O4, ZnO-In2O3, ZnO-MnO2 and ZnO-NiO was 38, 47, 143, 46 and 44 respectively. The pure ZnO shows the lowest and composite ZnO-In2O3 shows highest average particle size. The results obtained from DLS corresponding with the results of FESEM images. It was described that the highest particle size from the composite ZnO-In2O3 due to the high atomic/ionic radius (167 pm) of Indium atom, when compared to the other metals from the composite. Further the composites, prepared from the hydrothermal reaction show the size within 100 nm range except ZnO-In2O3 and indicates that the confirmation of nanocomposite formation.
3.5. UV-Vis Spectroscopic Analysis
The DRS UV-Vis spectra of as-synthesized ZnO-MxOy nanocomposites were recorded to determine their light absorption characteristics and Figure 5 shows the DRS UV-Vis spectra of heterostructured nanocomposites. ZnO-MnO2 shows the highest visible light absorption while ZnO-NiO shows the lowest visible light absorption. The band gaps of the various heterostructures were obtained from Tauc plot displayed in Figure 5(b)   . Obviously, the band gap of the four heterostructures followed the trend of ZnO-NiO (3.1 eV) > ZnO-In2O3 (2.71 eV) > ZnO-Co3O4 (2.64 eV) > ZnO-MnO2 (2.19 eV).
3.6. TGA Analysis
The thermal analysis of as synthesized ZnO-MxOy nanocomposites was carried out to get information about change in composites properties as function of temperature. Figure 6 shows the TGA curves obtained under inert nitrogen environment for the nanocomposites. TGA curve indicate that the change in weight loss (%). The TGA curve of ZnO-Co3O4 shown that, the gradual decrease
Figure 5. UV-Vis absorbance spectrum and Tauc plot of ZnO-MxOy nanocomposites.
Figure 6. TGA curves of ZnO-MxOy nanocomposites.
in the weight at 105˚C, attributed to the evaporation of any residual water. Further decrease in the weight loss up to 290˚C was due to the evaporation of CO2. The presence of CO2 was evident from the FTIR analysis as well as the sudden loss in the weight% at temperature 290˚C to 350˚C was due to the decomposition of Co3O4 which in turn form Co3O4 and O2. The total weight loss of ZnO-Co3O4 was ~15% and it shows high thermal stability. TGA curve at the temperature range between 40˚C - 180˚C and 190˚C - 375˚C of ZnO-NiO composite shows the weight loss occurred due to evaporation of water and CO2 respectively. Again the sudden increase in the weight at temperature range between 380˚C - 390˚C occurred due to the re-adsorption of CO2 with the influence of fluid N2. Further the sudden decrease in the weight loss at temperature between 440˚C - 520˚C was due to the decomposition of composites. The TGA curve of ZnO-MnO2 composite shows the curve with gradual decrease in the weight loss at temperature 40˚C - 150˚C due to evaporation of water molecules. Further the weight loss at temperature 420˚C - 550˚C was due to the thermal decomposition of ZnO-MnO2 to MnO2, and also the total weight loss observed was only ~8%, which depicts the higher thermal stability. The TGA of ZnO-In2O3 composite shows that the gradual decrease of weight loss at 40˚C - 120˚C and 150˚C - 240˚C indicate the evaporation of water and CO2 respectively. Further the sudden decrease in the weight loss at 280˚C - 330˚C was due to the decomposition of composite, with a larger weight loss (35%).
3.7. BET Surface Area
The N2 sorption isotherms of all samples can be classified as type IV with H3 hysteresis loops (Figure 7), which means the existence of slit-like mesopores (2 - 50 nm) formed between aggregated nanoparticles  . The hysteresis loops of all samples spanned a large P/P0 range (0.3 - 1.0), corresponding to the presence of mesopores. The surface area of the various samples follows the order: ZnO-MnO2 (58.4 m2/g) > ZnO-NiO (41.6 m2/g) > ZnO-Co3O4 (39.6 m2/g) > ZnO-In2O3 (22.7 m2/g). In comparison with DLS results, the lowest surface area is shown by ZnO-In2O3 with the larger particle size. The pore size distribution of the samples (Figure 7) further confirms the formation of mesopores. As can be seen, the pore size distribution curve is quite broad (from 2 to 100 nm) with small and large mesopores (in the range of 2 - 50 nm). The small mesopores may reflect porosity within the nanoparticles and the large mesopores are associated to the pores formed between the nanoparticles  .
3.8. Photocatalytic Properties of ZnO-MxOy Nanocomposites
The photocatalytic activities of the as prepared nanocomposites were performed by using phenol as a model organic pollutant in an aqueous solution for the degradation process. Figure 8 shows Chemical Oxygen Demand level of phenol
Figure 7. BET N2 adsorption isotherm of ZnO-MxOy nanocomposites.
Figure 8. COD of ZnO-MxOy nanocomposites.
degradation in different concentrations of various composites under sunlight irradiation for the period of 3 hours. The order of phenol degradation follows the trends: ZnO-Co3O4 > ZnO-In2O3 > ZnO-MnO2 > ZnO-NiO. Apparently, the order of phenol degradation does not follow the order of surface area, but it may occur because of the chemical composition of composites. The photocatalytic degradation was maximum by using ZnO-Co3O4, where 97% of 8 ppm of phenol have been degraded followed with 95% degradation by ZnO-In2O3, 84% degradation by ZnO-MnO2 and 29% degradation by ZnO-NiO respectively. The degradation efficiency decreased when the concentration of phenol increased from 8 ppm to 32 ppm, which may be due to phenol molecules around the photocatalytic active sites resulted in inhibiting the penetration of light to the surface of the catalyst. Hence, the generation of relative amount of ?OH and on the surface of the catalyst decreased  . The heterostructure catalysts were further tested for repetitive processes of photocatalytic degradation. It was found that the photocatalytic degradation efficiency did not decrease even after 3 cycles of photodegradation.
3.9. Plausible Mechanism of Phenol Degradation
The band gaps and band positions were calculated and the band structure diagrams of the different nanocomposites were presented in Figure 9, which is
Figure 9. Photocatalysis mechanism of ZnO-MxOy nanocomposites.
helpful in understanding the mechanism of phenol degradation. The enhancement in the photocatalytic activity of ZnO-Co3O4 and ZnO-In2O3 can be attributed to the small band gap of ZnO-In2O3 (2.71 eV) and ZnO-Co3O4 (~2.64 eV) which results in improving the visible light absorption, and the enhancement in the charge separation occurs because of coupling ZnO with Co3O4 and In2O3 separately    . In case of Co3O4-ZnO nanocomposite, the electrons are accumulated in the conduction band (CB) of ZnO and the holes in the valance band (VB) of Co3O4 because the difference in the band edges positions. Similarly, in In2O3-ZnO nanocomposite, the electrons migrate to the CB of ZnO and the holes to the VB of In2O3 leading to inhibiting the electron hole recombination and enhancing the charge carrier separation efficiency and thus enhanced the photocatalytic activity. The electronic structure and band gap movement and also defect states have been studied in detail using XPS and photoluminescence spectroscopy. These results will be published elsewhere as a part of a detailed mechanism for heterostructure metal oxides nanocomposites.
Although the band gap of ZnO-MnO2 (2.19 eV) is smaller than ZnO-In2O3, ZnO-Co3O4 and ZnO-NiO shows lower photocatalytic activity and the reason is charge separation efficiency of ZnO-MnO2 is very low when compared with ZnO-Co3O4 and ZnO-In2O3. This process is indicated clearly in Figure 9, where the electrons and holes are accumulated in the CB and VB of MnO2 which leads to high electron-hole recombination rate and thereby low photocatalytic activity. Moreover, the CB edge of MnO2 is lower than the potential required to produce superoxide radicals ( ), hence only hydroxyl radicals (?OH) take part in the phenol degradation over ZnO-MnO2 leading to further decrease in the photocatalytic activity  . The low photocatalytic activity of ZnO-NiO under sunlight irradiation due to the large band gap of ZnO-NiO (3.1 eV) leads to low light absorption and hence low photocatalytic activity occurred.
ZnO-Co3O4, ZnO-In2O3, ZnO-MnO2, and ZnO-NiO heterostructures were fabricated via facile hydrothermal route. The successive formation of heterostructures and their purity were confirmed by XRD analysis. FT-IR analysis further verified that fabricated heterostructures are metal oxides and no other impurities. SEM and BET studies revealed that the heterostructures obtained exhibit porous nature. TGA analysis shows that the highest and lowest thermal stability were exhibited by ZnO-MnO2 and ZnO-In2O3 respectively. The band gap of the prepared heterostructures follows the trend: ZnO-NiO (3.1 eV) > ZnO-In2O3 (2.71 eV) > ZnO-Co3O4 (2.64 eV) > ZnO-MnO2 (2.19 eV). The photocatalytic activity was influenced by the electronic structure and the band gap of heterostructure rather than its surface area. The photocatalytic activity follows the order: Co3O4-ZnO > In2O3-ZnO > ZnO-MnO2 > ZnO-NiO. The low charge separation efficiency of ZnO-MnO2 and its inappropriate CB edge position are the reasons for lower photocatalytic activity than Co3O4-ZnO and ZnO-In2O3 with higher band gaps.
Authors wish to acknowledge the financial support and the laboratory facilities to carry out this work under UGC-UPE project, Govt. of India. FE-SEM work was carried out at Mangalore University, India.