Surfactant-free synthesis of Ag@TiO2 and CuO@TiO2 photocatalyst: a comparative study of their photocatalytic activity for reduction of nitroaromatics in water

The reduction of harmful nitroaromatics to useful amino-aromatics have significant opportunities in synthetic chemistry. Here a visible-light-driven eco-friendly method for the selective reduction of aromatic nitro compounds to their corresponding amines in aqueous solution by using Ag@TiO2 and CuO@TiO2 is described. It was observed that both Ag@TiO2 and CuO@TiO2 are photo-catalytically more efficient compared to bare TiO2 and CuO@TiO2 has higher activity over Ag@TiO2 for the said conversion. The structural and morphological characterization of the as-synthesized catalysts has been done with SEM-EDX, TEM, powder XRD, ICP-AES, XPS, Photoluminescence, and UV-vis spectroscopic techniques. The nanocomposites Ag@TiO2 and CuO@TiO2 exhibit pure anatase phase with average crystallite size of 5.89 nm and 5.87 nm respectively as calculated from the Debye-Scherrer equation depending on the (101) plane. UV-visible results inferred enhanced optical properties of both the synthesized catalysts and revealed a reduced band gap (3.07 eV for Ag@ TiO2 and 2.5 eV for CuO@ TiO2) as compared to neat TiO2 (3.36 eV). Various nitro compounds were tolerated under 150 W LED as a light source (13.9 lumens for an area of 0.2 ft2) in an aqueous medium at room temperature (30 °C) using NaBH4 as a reducing agent to access corresponding amines in satisfying yields (78%–99%). The catalyst can be separated from the reaction mixture by simple centrifugal precipitation and reused for up to six consecutive cycles without apparent loss of its catalytic activity. The products were characterized by 1H-NMR spectroscopic techniques and compared with authentic samples.


Introduction
In recent years, the release of pollutants from organic compounds such as nitroaromatic compounds and organic dyes has markedly increased as a result of growth in industrialization [1]. In an aqueous medium, derivatives of nitroaromatic compounds are stable and not effortlessly degradable, so they are highly toxic and harmful not only to human health but also to the environment. Therefore, the conversion of these nitroaromatic compounds to useful amino derivatives becomes highly desirable. The aromatic amino compounds are key intermediates for agrochemicals, pharmaceuticals, polymers, dyes, drugs, etc [2]. So far, the said conversion has been studied in the presence of various transition metals such as Au, Pd, Pt, Rh, Ag, Cu, Ir, Ni, Co, Fe, etc, and using several reducing agents such as hydro silanes, NaBH 4 , CO/H 2 O, hydrazine, HCOONH 4 , ammonium borane, etc [3][4][5][6][7][8][9]. However, many of them associated with some drawbacks such as high cost, use of toxic chemicals, use of the high amount of catalyst, excess reducing agent, high temperature and pressure. In this context, photocatalysis which uses sustainable and abundant solar or visible light, ambient temperature, and pressure, could be an alternate green approach [10]. Many protocols were reported for photocatalytic reduction of nitroaromatics using Fe 2 O 3 /BiVO 4 [11], N-doped TiO 2 nanoparticles [12], Ni (II) and Zn(II)-porphyrin metal-organic framework [13,14], Ag &Pd@ MIL-125-NH 2 @cellulose acetate film [15], In 2 S 3 -CNT nanocomposite [16], graphene @CoS 2 [17] etc. The semiconductor-based photocatalysts have immerged as fascinating candidates due to their heterogeneous nature, adaptability to holding multiple oxidation states, lower band gap, and eco-friendliness. For the reduction of nitroaromatic compounds various reports are available utilizing noble and non-noble metal-based transition metal oxides semiconductors including Cu 2 O/TiO 2 , CuCo 2 O 4 /RGO, Au/TiO 2 , Pd/ZnO, Cu/Ag-ZnO, Fe 3 O 4 @Cu, Au/Fe 3 O 4 and so forth [18][19][20][21]. TiO 2 is one of the most widely used high-performance photocatalysts with a band gap of 3.2-3.4 eV [22]. However, TiO 2 nanoparticles have some limitations due to the natural agglomeration into larger counterparts, for which its surface area decreases and reduces its applicability. Hence, the use of different stabilizers like polymers, inorganic solids, ligands, additives, surfactants etc are required in their fabrication to control size, shape and to enhance the stability of the nanoparticles. However, the presence of these substances might block the active site of the catalyst surface and could negatively affect the catalytic and electrochemical properties of the synthesized nanoparticles. So, these additives, surfactants etc need to be removed from the catalyst surface which involves energy and time intensive steps adding the risk of structural change of the synthesized catalyst [23,24]. To overcome this problem, various dopants, including both metals and non-metals, could be incorporated over the surface of TiO 2 [25,26]. The transition metal ions are worthier due to the presence of partially filled d-orbitals. The inclusion of transition metals into the lattice of TiO 2 persuades the development of new energy levels near the conduction band and retards the rapid recombination of electron-hole pairs, thereby increasing the photocatalytic activity of TiO 2 [27]. Some reports are available for the synthesis of non-metallic co-doped TiO 2 such as I-S/TiO 2 , N-C/TiO 2, N-B/TiO 2, etc, monometallic co-doped TiO 2 such as Ag/TiO 2, Cu/TiO 2, Au/TiO 2, Cu-S/TiO 2 , Cu-N/TiO 2, etc and bimetallic co-doped TiO 2 such as Cu-La/TiO 2 , Co-Ni/TiO 2, Hg-Ag/TiO 2 , Ni-Ag/TiO 2 and so forth [28][29][30][31][32][33]. When transition metals such as Ag and Cu are loaded onto the photocatalyst they may act as a carrier of photogenerated electrons from the conduction band of TiO 2 and the transfer of electrons between the transition metal and TiO 2 facilitates the separation of charge carriers improving the photocatalytic activity of TiO 2 . The Cu dopant is more favorable due to its high electronic conductivity, low cost, and high availability in the earth's crust compared to Ag. Copper as a dopant in the TiO 2 matrix introduces structural defects by initiating d states at the intermediate of the TiO 2 band gap [34].
For the reduction of nitro aromatics to their corresponding amino aromatics, two strategies can be adopted depending on the source of hydrogen, which includes direct hydrogenation and transfer hydrogenation. However, the first method has some demerits, such as the use of high-pressured hydrogen under high temperatures, high cost associated with the storage and transportation of hydrogen, complex experimental setup, high energy, and so forth. Contrarily, transfer hydrogenation has several merits, notably easy experimental set-up, paucity of dangerous hydrogen, use of available and economical hydrogen sources, low energy, and so forth [35]. In this context, here we have reported the synthesis of Ag@TiO 2 and CuO@TiO 2 photocatalysts, investigated and compared their photocatalytic activity towards the reduction of nitro aromatics using NaBH 4 as reductant and water as a green solvent. The present catalytic system works under ligand-free, and mild reaction condition is advantageous to those that make use of toxic organic solvents, ligands, the high-pressure reactor of hydrogen gas, and high temperature. In comparison to other traditional metal-mediated reductions (Fe, Sb, and Zn) [36], when CuO@TiO 2 is used as a catalyst in an aqueous medium, only 1 mmol of NaBH 4 is required for the reduction of nitro compounds, and the catalyst could be reused for up to six consecutive cycles without apparent loss of its catalytic activity. The comparison of the photocatalytic activity of Ag@TiO 2, and CuO@TiO 2 , short time requirement, absolute heterogeneity, magnificent reusability, use of green reaction conditions, and supremely low metal loading (0.0827mol% of Ag and 0.1856 mol% of Cu per 10 mg) as well as less amount of the reducing agent (1 mmol) added extra light to the present protocol.

Reagents used and characterization techniques
All reagents like titanium isopropoxide, isopropyl alcohol, acetic acid, methanol, silver nitrate, sodium nitrate, sodium chloride, copper chloride etc were purchased from Merck and nitro compounds were obtained from Sigma-Aldrich. The reagents were chemically pure and used as received without any further purification or drying.
Powder x-ray diffraction patterns (PXRD) of TiO 2 , Ag@TiO 2 and CuO@TiO 2 were recorded on a Bruker AXS D8 Advance at STIC, Kochi University, Kochi. The amount of Ag and Cu loading in the synthesized catalysts were determined by Inductively Coupled Plasma Atomic Emission Spectrometric (ICP-AES) analysis at SAIF, IIT Bombay using ARCOS, Simultaneous ICP Spectrometer. The UV-Visible and Photo Luminescence spectra of the catalysts were recorded using Hitachi U-3900H UV-Vis Spectrophotometer and Fluoromax 4P Spectrofluorometer, respectively. The HR-TEM with SAED pattern of Ag@TiO 2 and CuO@TiO 2 were done using JEOL/JEM 2100 operating at 200 kV at STIC, Kochi University, Kochi. The SEM images and EDX spectra of Ag@TiO 2 and CuO@TiO 2 were obtained using JEOL 6390 LV operating at 30 kV at STIC, Kochi University, Kochi. The x-ray Photoelectron spectra of Ag@TiO 2 and CuO@TiO 2 were recorded using PHI 5000 Versa Probe II, FEI, Inc. at ACMS, IIT Kanpur. The 1 H NMR spectra of the isolated products were recorded using Bruker Ascend TM, 500 MHz Spectrometer at CSIC, Dibrugarh University, Dibrugarh.

Synthesis of TiO 2 nanoparticles
TiO 2 nanoparticles were prepared via the sol-gel route by following a previously reported protocol with slight changes in reaction conditions [37,38]. 12 ml isopropyl alcohol and 4 ml titanium isopropoxide were introduced into a beaker, and the solution was stirred for about 3 h. To this solution, 12 ml acetic acid and 24 ml methanol was added followed by stirring for 5 h whereupon a yellow precipitate was formed. The precipitate was filtered, washed with deionized water, and dried at 80°C for 2 h. It was then calcined at 400°C for 3 h.

Synthesis of Ag@TiO 2 nanoparticles
0.75 g of as-synthesized TiO 2 powder was added to an aqueous solution of 0.1M AgNO 3 (15 ml) and stirred for 6 h. To this slurry, a 1.0 percent (W/V) aqueous solution of NaNO 3 (15 ml) was added and stirred for another 7 h. The suspension was dried at room temperature and calcined at 400°C for 3 h.

Synthesis of CuO@TiO 2 nanoparticles
0.75 g of as-synthesized TiO 2 powder was added to an aqueous solution of 0.1M CuCl 2 (15 ml) and stirred for 6 h. To this slurry, a 1.0 percent (W/V) aqueous solution of NaCl (15 ml) was added and stirred for another 7 h. The resulting suspension was dried at room temperature and calcined at 400°C for 3 h.

Photocatalytic experiment
2.3.1. General procedure for photocatalytic reduction of nitro compounds A mixture of nitro compound (1 mmol), catalyst (10 mg), NaBH 4 (1 mmol), and water(5 ml) was taken in a 50 ml round bottom flask surrounded by a circulating water jacket, and the reaction mixture was stirred for the required time in presence of visible light using a 150W LED light which was kept 25 cm apart from the reaction flask. The intensity of the light as calculated by using a lux meter was found to be 13.9 lumens for an area of 0.2 ft 2 . The progress of the reaction was monitored by TLC (1:9 ethyl acetate-hexane as eluent). Upon completion of the reaction, the catalyst was recovered from the mixture by simple centrifugal precipitation. The organic product was extracted with ethyl acetate (3 × 10 ml) and dried over anhydrous Na 2 SO 4 . The extracted product was purified by silica gel column chromatography using ethyl acetate/hexane (1:9) as eluent. The obtained products were characterized by 1 H -NMR spectroscopy.  [39]. The XRD patterns of each of the Ag@TiO 2 and CuO@TiO 2 demonstrated that all the diffraction peaks could be ascribed to the anatase phase (figure 1(a)) [40,41]. Moreover, no obvious shift in the peak positions (inset top right in figure 1(b)) would suggest the non-incorporation of Ag and Cu particles into the lattice of TiO 2 ; contrarily, they most likely adsorb on the surface of TiO 2 nanoparticles. The two diffraction peaks at 2theta value of 38.3°and 64.4°for Ag@TiO 2 and three peaks at 2theta values of 38.1°, 48.1°and 75.4°for CuO@TiO 2 could be indexed to the planes of Ag and Cu particles respectively [42,43]. Moreover, figure 1(b) clearly showed that the XRD peak intensity of CuO@TiO 2 is higher compared to pure TiO 2 and Ag@TiO 2 , which indicated that the incorporation of Cu enhances the crystallinity and improves the structural quality of TiO 2 . The average crystallite sizes of the samples were calculated from XRD data using the Debye-Scherrer equation based on the plane (101) at the 2θ values of 25.4°. The average crystallite sizes were found to be 6.2 nm, 5.9 nm, and 5.8 nm for TiO 2 , Ag@TiO 2, and CuO@TiO 2 , respectively.

SEM-EDX
The surface morphology of Ag@TiO 2 and CuO@TiO 2 nanocomposite were studied by SEM, and their elemental analyses were obtained from the EDX analysis. The SEM images of Ag@TiO 2 and CuO@TiO 2 (figures 2(a) and (c)) clearly showed the non-homogeneous surface morphology for both the hetero nanostructures, while their EDX spectra showed the presence of desired elements in proportions (figures 2(b) and (d)). Moreover, the EDX analysis clearly enumerates the low loading of Ag and Cu in the respective nanocomposites.

TEM
To get more information about surface morphology, particle size, and crystal structure of Ag@TiO 2 and CuO@TiO 2 nanocomposites TEM analysis was carried out. The TEM image of Ag@TiO 2 nanocomposite affirms the presence of non-spherical particles while in the case of CuO@TiO 2, spherical particles were formed with an almost uniform size (figures 3(a) and 4(a)). Moreover, TEM micrographs of both distinctly showed the deposition of black-colored particles which might be indicative of the presence of metal (Ag/CuO) nanoparticles on the surface of gray-colored TiO 2 nanoparticles. The fringe spacing of 0.35 and 0.351 nm as observed in the HR-TEM images of Ag@TiO 2 and CuO@TiO 2 (figures 3(b) and 4(b)) match that of the anatase (101) plane whereas the lattice fringes of 0.204 nm and 0.213 nm could be attributed to the Ag (200) and Cu (200) plane respectively [44,45]. The concentric rings were observed due to the reflection from (101), (004), (200), (211), and (204) planes in the SAED patterns of Ag@TiO 2 and CuO@TiO 2 and agree well with the XRD patterns of both (figure 3(c) and 4(c)). The average particle sizes of Ag@TiO 2 and CuO@TiO 2 photocatalysts were found 44 nm and 42 nm, respectively from the particle size distribution curves (figure 3(d) and 4(d)).

XPS
The chemical states of Ag, Cu, and Ti in all three compounds were determined with XPS spectroscopy. Figures 5(a) and (d) demonstrate the XPS survey spectra of Ag@TiO 2 and CuO@TiO 2 , respectively. The sharp peaks located at binding energies of 458.4 and 464.1 eV for both Ag@TiO 2 and CuO@TiO 2 could be ascribed to Ti 2P 3/2 and Ti 2P 1/2 , respectively, indicative of Ti 4+ in TiO 2 lattice (figures 5(b) and (e)) [46]. The highresolution XPS of Ag 3d core level in Ag@TiO 2 exhibited two peaks at binding energies of 368.5 eV and 374.5 eV corresponding to Ag 3d 5/2 and Ag 3d 3/2 that unambiguously told the presence of Ag element ( figure 5(c)) [46,47]. The Ag 3d 5/2 peak appeared at a binding energy of 368.5 eV indicating the presence of metallic Ag. Moreover, the binding energy difference (ΔB.E. = 6.0 eV) between Ag 3d 5/2 and Ag 3d 3/2 is indicative of metallic Ag 3d states. The two distinct peaks at 934.6 eV and 954.5 eV ( figure 5(f)), corresponding to Cu 2p 3/2 and Cu 2p 1/2, respectively enumerating the presence of Cu 2+ [48]. The strong shake-up peaks observed at (S1) 932.5 and (S2) 942.6 eV further confirm the presence of Cu (II) [49]. This evidence clearly revealed that in Ag@ TiO 2 the loaded Ag exists as metallic Ag and in CuO@TiO 2 , the loaded copper exists as Cu 2+ .

ICP-AES
The ICP-AES analysis showed that 10 mg of Ag@TiO 2 photocatalyst contains 0.0827 mol% of Ag whereas the same amount of CuO@TiO 2 contains 0.1856 mol% of Cu.

UV-vis
The UV-vis spectra of TiO 2 , Ag@TiO 2, and CuO@TiO 2 along with their band gaps are depicted in figure 6. The band gaps were calculated by using the Tauc plot [ energy (eV) versus (αhυ) 1/2 ] and were found to be 3.36 eV, 3.07 eV, and 2.5 eV for pure TiO 2 , Ag@TiO 2, and CuO@TiO 2 , respectively. The indirect band gap value of bare TiO 2 is in accordance with the literature [50]. The UV-vis absorption spectrum of bare TiO 2 showed absorption peaks only in the UV region and totally zero absorption in the visible region. On the contrary, Ag@TiO 2 exhibited slight absorption in the visible region with the highest absorption at 358 nm ( figure 6(b)), while CuO@TiO 2 showed significant absorption in the visible region with a strong absorption peak at 364 nm (figure 6(c)) [41,51]. From these observations, it may be ascribed that visible light irradiation may produce electron-hole pairs in the case of Ag@TiO 2 and CuO@TiO 2 , although the possibility is smaller for Ag@TiO 2 , but cannot generate from bare TiO 2 . These visible light-generated electron-hole pairs might be the main reactive species for the reduction of nitroaromatics. So, it is confirmed that metal loading certainly produces an intermediate band between valence and conduction bands of bare TiO 2 and thereby decreases the band gap which in turn makes visible light responsive catalytic activity of Ag@TiO 2 and CuO@TiO 2 .

Photoluminescence (PL)
The PL spectroscopy can predict the electron-hole separation and recombination in the photocatalyst sample. The room temperature PL emission spectra of TiO 2 , Ag@TiO 2, and CuO@TiO 2 excited at a wavelength of 340 nm are depicted in figure 7. It was reported that bare TiO 2 nanoparticles showed a recombination emission peak at around 420 nm [52]. In our case anatase phase TiO 2 exhibited five visible emission peaks at 438 nm, 452 nm,467 nm, 480 nm, and 492 nm. The two visible peaks at 438 nm and 452 nm could be ascribed to band edgefree excitation and self-trapped excitation of TiO 2 octahedra, respectively. The blue emission peak at 467 nm is indicative of the presence of oxygen vacancies in TiO 2 nanoparticles. The emission peaks observed at 480 nm and 492 nm might be due to the electronic transition between ionized oxygen vacancies and interstitial oxygen or surface defects [44,[53][54][55]. As observed from figure 7, the intensity of all the peaks decreases for Ag@TiO 2 and CuO@TiO 2 compared to pure TiO 2, and CuO-loaded TiO 2 showed the lowest PL intensity. The PL spectra intensity reduction represents the recombination rate decreasing of the photogenerated electron-hole pairs, and consequently, many photoexcited charge carriers take part in the photochemical reaction which leads to higher photocatalytic activity [56]. So, the PL study affirmed the highest photocatalytic activity of CuO@TiO 2 compared to the bareTiO 2 and Ag@TiO 2.
From the PL spectra, the band gap of the photocatalysts was calculated and found 3.0, 2.6, and 2.2 eV respectively for TiO 2 , Ag@TiO 2, and CuO@TiO 2 . The band gap observed from PL spectra is lower than that obtained from UV-vis spectra since the PL emissions are not 100% radiative in nature. There is always an associated portion of non-radiative emissions due to which the band gap calculated by the PL study will always be less than the original band gap.

The catalytic activity of Ag@TiO 2 and CuO@TiO 2 photocatalysts towards the reduction of various nitroaromatics
To test and compare the catalytic activity of Ag@TiO 2 and CuO@TiO 2 for selective reduction of nitroaromatics to corresponding amines a set of reactions was carried out taking 4-nitrobenzene as the model substrate to optimize the reaction conditions. A thorough catalyst screening established that CuO@TiO 2 was the most active photocatalyst among the three nanocomposites. From our observations, it was noticed that no product formation takes place without the catalysts, solvents, and reducing agent even after a prolonged time (table 1, entries 1,2,3 & 6). The bare TiO 2 with hydrazine hydrate as a reducing agent gave no product in the water while NaBH 4 afforded only 36% product yield in water but could not be in acetonitrile solvent (table 1, entries 4,5&16). Contrarily, Ag@TiO 2 and CuO@TiO 2 photocatalysts afforded satisfactory product yield in water using NaBH 4 as the hydrogen source (table 1, entries 7-11). We examined the performance of various reductants such as NaBH 4, LiBH 4 , and NH 2 NH 2. H 2 O and found NaBH 4 was the best hydrogen source in our cases (table 1, Entries 7-15). NaBH 4 is a safe, comparatively milder reductant and is widely used. It undergoes hydrolysis in an aqueous solvent to give H 2 for the effective reduction of nitro compounds. Optimization of the model reaction with respect to various solvents, viz. acetonitrile, ethanol, isopropanol, and H 2 O established water as the best solvent (table 1, entry10). While we carried out the model reaction by taking 10 mg of Ag@TiO 2 and CuO@TiO 2 each with the same amount of NaBH 4 and water, it was observed that CuO@TiO 2 gave 99% product yield in ½ h while Ag@TiO 2 offered 80% of desired product yield in 4 h (table 1, entries 9&10). These results enumerate the highest catalytic efficiency of CuO@TiO 2 .
The possible reasons for the highest photocatalytic activity of CuO@TiO 2 over Ag@TiO 2 and TiO 2 might be due to the lowering of the band gap (as observed from Tauc plot) which allows easy transfer of photo-induced electrons from the conduction band of TiO 2 to Cu 2+ ion [57]. The transfer of electrons between the transition metal and TiO 2 facilitates the separation of charge carriers thereby preventing the electron-hole recombination rate and resulting in improved photocatalytic activity of TiO 2 . In addition, the lowest PL spectra intensity of CuO@TiO 2 compared to Ag@TiO 2 and TiO 2 (figure 7) represents the lowest electron-hole pairs recombination rate, and the highest number of photoexcited charge carriers taking part in the photochemical reaction which leads to higher photocatalytic activity of CuO@TiO 2 compared to Ag@TiO 2 and TiO 2 photocatalyst. The UVvisible analysis also revealed the highest absorbance of CuO@TiO 2 in the visible region (figure 6(c)) compared to Ag@TiO 2 (figure 6(b)) and bare TiO 2 ( figure 6(a)). These factors certainly made CuO@TiO 2 the best photocatalyst among the three. In the present case, 10 mg of CuO@TiO 2 as the catalyst (0.1856 mol% of Cu and 1.4878 mol% of Ti), 5 ml of H 2 O as the solvent, and 1 mmol of NaBH 4 gave the best reaction condition (table 1, entry10) to receive the highest product yield in presence of visible light (150W LED light kept 25 cm apart from the reaction flask with light intensity 13.9 lumens for an area of 0.2 ft 2 ).
Employing the optimized reaction condition and 150 W visible light a series of substrates were studied and in all the cases nitroaromatics underwent reduction to corresponding aromatic amines with a satisfactory yield of the desired products (table 2). It was noticed that by using excess NaBH 4 (2 mmol) as the hydrogen source, 1,4-dinitro benzene was also reduced by CuO@TiO 2 selectively to 4-nitroaniline (table 2, Entry 8) with 85% isolated yield whereas 70% isolated yield of the desired product was obtained with Ag@TiO 2 photocatalyst. Nitro aromatics bearing electron withdrawing and electron donating substituents gave their corresponding anilines with good to excellent isolated yield. However, the position of the substituents on the aromatic ring affected the yields of the product. Our investigations showed that ortho-and para-substituents at the ring afforded higher yields of the product (up to 99%) while the presence of substituents at the meta-position offered lower yields (table 2, Entries 3,5&7) [58]. The N, N-dimethyl amino-4-nitrobenzene gave 85% and 94% product yield with Ag@TiO 2 and CuO@TiO 2 , respectively (table 2, Entry 10). Without undergoing facile dehalogenation, the halo-substituted nitroaromatics were also reduced selectively to their corresponding halogenated aromatic amines (table 2, Entry [2][3][4][5][6][7]9). In all the cases, CuO@TiO 2 was found to be more efficient than Ag@TiO 2 photocatalyst. The comparative study of the reported catalyst for reduction of nitroaromatics with that of existing literature is illustrated in table 3.
The schematic representation of Ag@TiO 2 and CuO@TiO 2 photocatalytic system for the reduction of nitroaromatics is shown in scheme 1. A plausible mechanism has been proposed for the reduction of nitroaromatics on the basis of band structure of CuO@TiO 2 . When visible light is irradiated on the reaction mixture, the CuO@TiO 2 semiconductor facilitates the excitation of electrons from valence band to conduction band leaving holes at valence band which is responsible for the redox reaction. Here, water present in the reaction system act as hole scavenger and the lone pair present on O atom took hold the holes at valence band thereby inhibiting the process of electron-hole pair recombination. The photoinduced electrons in the conduction band and the protons were than employed to reduce nitroaromatics and the reaction proceeds through the sequential formation of nitroso [A] and phenyl hydroxylamine [B] intermediates and resulted in the desired corresponding anilines as final products [59].

Recyclability experiments
The scope of recyclability is the main attractiveness of heterogeneous catalysis. Therefore, we have checked the recyclability of Ag@TiO 2 and CuO@TiO 2 for nitroaromatic reduction taking 4-chloro nitrobenzene as the model substrate. After completion of a catalytic reaction, the photocatalysts were centrifuged at 5,000 rpm for 13 min, washed carefully with ethanol and water followed by drying in an oven at 70°C and subjected to a fresh run. To our delight, the catalysts have retained their efficiency even after the 6th cycle, however, a 4% decrease in product yield for both cases was observed (figures 8(a) and (b)). To know any physical changes of the catalysts that may occur during the reaction, we recorded the TEM and XRD pattern of the reused CuO@TiO 2 catalyst Table 2. Substrate study of various substituted nitroaromatics to corresponding aromatic amines according to the optimized condition obtained from table 1 using Ag@TiO 2 and CuO@TiO 2 photocatalyst. a a Reaction conditions: Nitroaromatics (1 mmol), H 2 O (5 ml), NaBH 4 (1 mmol), Ag@ TiO 2 /CuO@TiO 2 (10mg, 0.0827 mol% of Ag/ 0.1856 mol% of Cu), b NaBH 4 (2 mmol), room temperature (30°C).
after the 6th cycle. Interestingly the recovered catalyst after the 6th run showed an almost similar TEM image (figure 8(c)) and XRD pattern ( figure 8(d)) which indicates unchanged morphology as well as existence of all the crystal planes as that of the fresh catalyst.

Hot filtration test
To check the heterogeneity of the catalysts we have performed a hot filtration test (figure 9) under the optimized reaction conditions using 4-bromo nitrobenzene as the model substrate. The photocatalysts were separated   from the reaction mixture after 1h by simple centrifugation. Then the reaction was allowed to continue with the filtrate for another 4h and observed a little yield of the product. For confirmation of this result, ICP-AES analysis was done and the report ascertained negligible leaching of Ag and Cu metal (<0.08 mol ppm of Ag and < 0.05 mol ppm of Cu) into the reaction mixture. This demonstrated the non-leaching or negligible leaching of metals in the course of the reaction thereby signifying its stability and heterogeneous nature.

Conclusion
In conclusion, we have prepared Ag@TiO 2 and CuO@TiO 2 photocatalysts via the sol-gel route and evaluated them as highly efficient and selective catalysts for the reduction of a variety of harmful nitroaromatic compounds. The photocatalytic activity of both the catalysts was compared and found CuO@TiO 2 to be more efficient for the photocatalytic reduction of nitroaromatics containing both electron-donating and electronwithdrawing groups to their corresponding amines under visible light irradiation. The photocatalytic reduction of nitroaromatics using CuO@TiO 2 is less time-consuming compared to Ag@TiO 2 , afforded higher product yield, and also the reusability was found to be better than using Ag@TiO 2 photocatalyst. The enhanced photocatalytic activity of CuO@TiO 2 over Ag@TiO 2 might be due to the large surface area, smaller band gap and increased number of photoinduced charge carriers taking part in the photochemical reaction. The magnificent reusability, short time requirement, absolute heterogeneity, use of green reaction conditions, and supremely low metal loading (0.0827mol% of Ag and 0.1856 mol% of Cu per 10 mg) as well as less amount of the reducing agent (1 mmol) added extra light to the present protocol.