Fabrication of Bi₂S₃/ZnO heterostructures: an excellent photocatalyst for visible-light-driven hydrogen generation and photoelectrochemical properties

Abstract

Fabrication of heterostructures is considered as one of the effective strategies to improve photocatalytic performance for organic pollutant degradation and hydrogen production under solar light irradiation. Here, Bi₂S₃/ZnO heterostructures with a flower-like architecture have been successfully synthesized by a facile in situ generation method. The as-synthesized Bi₂S₃/ZnO heterostructures showed enhanced visible-light absorption and charge separation efficiency of photoinduced electron–hole pairs. This heterojunction exhibits 3 fold enhancement in organic pollutant degradation and 2.7 fold enhancement in photocatalytic hydrogen generation under visible irradiation compared to rod-shaped Bi₂S₃. The high current gain (ca. 8.79) and low photocorrosion in photoelectrochemical water splitting reveal the superior photocatalytic activity of the heterojunction under visible light. The superior photocatalytic activities are attributed to the synergetic effects of ZnO nanoparticles and rod-shaped Bi₂S₃ in the Bi₂S₃/ZnO heterostructures, which result in fast separation and slow recombination of photoinduced electron–hole pairs. The reusability and stability of the photocatalysts has been checked by recycling experiments. X-ray diffraction and scanning electron microscopy reveal that the structure and morphology of the heterostructures remain unchanged after photocatalytic cycling tests. The visible light active catalysts have potential for efficient solar light harvesting and overall water splitting. This work demonstrates the potential use of heterostructures as a highly efficient photocatalyst for dye degradation and hydrogen production under visible light irradiation.

---

Introduction

The challenges of environmental pollution and renewable energy production are both highly active areas of scientific research.^1,2 Visible light-induced photocatalysis is an effective clean approach for water splitting to generate clean solar fuels and converting solar energy to chemical energy.^3,4 On the other hand, water purification and environmental protection pose an equally daunting challenge.^5,6 Oxide-based semiconductors have been utilized as active photocatalysts in various applications such as photochemical degradation of organic contaminants, photochemical water splitting to produce hydrogen and photoelectrochemical cells.^7–9 Titanium dioxide (TiO₂) is one of the most popular photocatalysts and possesses high catalytic activity, low cost and non-toxicity.^10,11 However, TiO₂ can only absorb UV radiation (4% of the total solar irradiation) due to its wide band gap and low charge separation efficiency, which makes it unsuitable for visible light active photocatalysis.^12,13 Moreover, TiO₂ doping with C, N, B, or S or modification with metal nanoparticles (Au, Pt, Ag) results in plasmon-induced enhanced visible light absorption; however, the high cost and low environmental stability of noble-metal-doped TiO₂ restricts its economic potential.^14–16 Another semiconductor nanocrystal, zinc oxide (ZnO), having a direct band gap of 3.3 eV and a large excitation binding energy has been widely tested for phtotocatalytic applications.^17–19 Moreover, ZnO has some advantageous properties over TiO₂ such as high electron mobility due to its electronic structure, room temperature luminescence etc.^18,19 In this regard, the loading of multiple catalysts or secondary semiconductors can improve the catalytic efficiency of ZnO via the formation of heterojunctions.^20,21 This stimulated our interest in designing a coupled heterojunction with ZnO to improve solar light absorption in the visible region.

Significant efforts have been made on the fabrication of semiconductor-based heterostructures such as ZnO/TiO₂,²² Bi₂S₃/TiO₂^23,24 and CdS/TiO₂²⁵etc., which show potential applications in water splitting and organic pollutant degradation. In fact, high charge separation has been achieved via coupling of a large band gap semiconductor with a smaller one forming a heterojunction, which in turn enhances the photocatalytic efficiency by decreasing the recombination rate of the photogenerated electron–hole pairs.^26,27 In the case of a heterojunction, charge carriers are generated in one semiconductor and then vectorially transfer to the other material allowing for long lived electron hole pairs at the interface and are able to produce a potential gradient within the catalyst.⁵ The presence of multiple active sites within the heterostructures can provide a high surface area for the decomposition of organic pollutant molecules at the surface of the catalysts.²⁸ For example, TiO₂-based hybrid heterojunctions demonstrated high catalytic activity due to surface tunnelling of electrons between the surfaces of the semiconducting components of the heterostructure.²⁹ Until now, most of the studies have been focused on the hybridization of TiO₂ with semiconductors to improve the photocatalytic efficiency under visible light.^29–33 In this regard, few reports have been published on the modification of ZnO nanostructures with other semiconductors such as ZnO/TiO₂,³⁴ CdS/ZnO³⁵etc. for improved visible light-driven photocatalysis.

On the other hand, low band gap semiconductors such as sulphides, nitrides, graphitic carbon nitrides, oxynitrides, chalcogenides etc., have been widely investigated to construct visible light active photocatalysts.^36–39 Among these, bismuth sulphide (Bi₂S₃), a low band gap metal sulphide (1.3 eV), has shown absorption in the visible region and has been considered a potential candidate for photocatalytic applications.⁴⁰ However, the rapid recombination of photo-induced electrons and holes limits the catalytic application of Bi₂S₃ under visible light. Recently, a series of Bi₂S₃-based heterostructures such as MoS₂/Bi₂S₃,⁴¹ Bi₂S₃/CdS,⁴² Bi₂S₃/Bi₂WO₆,⁴³ Bi₂S₃/In₂S₃,⁴⁴ Bi₂S₃/BiVO₄,⁴⁵ Bi₂S₃/(BiO)₂CO₃⁴⁶etc., have been developed to improve the light absorption and charge separation efficiency. Recently, Bi₂S₃/g-C₃N₄ heterostructures exhibited high catalytic performance for organic dye (Rhodamine B) degradation, but the apparent kinetic rate was very low.⁴⁷ MoS₂/Bi₂S₃ and Bi₂S₃/In₂S₃ heterojunctions showed high catalytic activity for organic pollutant degradation but no report on hydrogen generation is available to date. Notably, semiconductor-based heterostructures are effective catalysts for water oxidation as the photogenerated electrons and holes on the surface of heterostructures have the potential to react at the surface active site. Thus, the excited electrons reduce the water to form hydrogen and the holes oxidize the water to generate oxygen.⁴⁸ The Bi₂S₃/CdS heterostructures displayed photocatalytic hydrogen generation under visible light irradiation but the hydrogen evolution rate is low.⁴² Recently, Wang et al.⁴⁹ have used Z-scheme CdTe–Bi₂S₃ heterojunctions for photoelectrochemical performance. Hence, Bi₂S₃-based heterojunctions have been extensively utilized for organic dye degradation but photocatalytic hydrogen generation has not been tested yet. Herein, Bi₂S₃ with a narrow band gap has been integrated with ZnO nanoparticles to fabricate visible light active heterostructures. Various characterization techniques such as XRD, SEM, TEM and FTIR have been employed for characterizing the structure, morphology and optical properties of the heterostructures. To the best of our knowledge, for the first time, we studied the photocatalytic performance and photoelectrochemical properties of Bi₂S₃/ZnO for hydrogen generation under visible light.

Experimental section

Materials

Bismuth nitrate pentahydrate (Bi(NO₃)₃·5H₂O, 99.99%), L-cysteine hydro-chloride (99.99%), mercaptosuccinic acid (MSA, 99%), ethylenediaminetetraacetic acid disodium salt (EDTA-Na, 99%), methanol were procured from Sigma Aldrich, USA. For the in situ generation of ZnO, zinc acetate dehydrate (Zn(CH₃COO)₂·2H₂O, 98%) and sodium hydroxide (NaOH, 98%) were procured from Merck, Germany.

Synthesis of Bi₂S₃ nanostructures

In a typical synthesis, 0.5 mmol of ethylenediaminetetraacetic acid disodium salt (EDTA-Na) was dissolved in 100 mL of distilled water and stirred for 10 minutes. Then 0.4 mmol of Bi(NO₃)₃ was mixed and the solution mixture was ultrasonicated until the solution became transparent. After that, 0.6 mmol of L-cysteine hydrochloride was added to the solution. Finally, the solution was transferred into a 100 mL Teflon-lined autoclave with a stainless steel shell and heated at 180 °C for 16 h. The Bi₂S₃ was collected and centrifuged, washed several times with distilled water and finally air-dried overnight at 50 °C for further characterization. Similarly, Bi₂S₃ nanostructures were also prepared using mercaptosuccinic acid as a sulphur source.⁵⁰ The effect of sulphur source concentration (0.6, 1.2, 2.4 mmol) and different reaction times (10, 16, 24 h) to follow the growth mechanism of Bi₂S₃ nanostructure formation has been studied in detail.

Synthesis of Bi₂S₃/ZnO heterostructures

In order to develop a coupled heterostructure of Bi₂S₃/ZnO, a facile in situ generation via co-precipitation technique has been employed using ethanol as a solvent.¹⁷ Briefly, 20 mL of 4 mmol zinc acetate dehydrate solution was heated at 70 °C for 30 min. Then a fixed amount of prepared Bi₂S₃ powder (1 mg mL⁻¹) was added and mixed thoroughly. In the next step, 20 mL of 4 mmol sodium hydroxide solution prepared at 70 °C in ethanol was added slowly and the mixture was hydrolyzed for 2 h at 60 °C to obtain ZnO NPs with average diameters of ∼5 nm. The pure semiconductors and heterostructures are presented as L-Bi₂S₃, L-Bi₂S₃/ZnO (using L-cysteine hydro-chloride as a source of sulphur) and M-Bi₂S₃, M-Bi₂S₃/ZnO (using MSA as a source of sulphur). To understand the growth mechanism of Bi₂S₃, extensive experimental analyses were performed for varying reaction times, metal salt to surfactant ratios and sulphur sources.

Instrumentation

The crystalline phase of Bi₂S₃ and the heterostructures was investigated by XRD (Philips X’Pert, The Netherlands) within the 2θ range of 10° to 80° at a slow scan rate of 1° min⁻¹ with Cu Kα radiation (at 40 kV and 40 mA). The structural morphology and EDS was recorded by Field Emission Scanning Electron Microscopy (LEO. 430i, Carl-Zeiss, Sigma). Transmission Electron Microscopy (TEM) images and the corresponding selected area electron diffraction (SAED) patterns were obtained on Tecnai G² 30ST (FEI) operating at 300 kV. The XPS study was performed using a PHI 5000 Versa Probe II spectrophotometer (Physical Electronics Inc., USA) with a monochromatized Al Kα (∼1486.6 eV) X-ray beam of size ∼100 μm. The samples were prepared in pallet form and the surfaces were sputtered with a 2 kV rastered Ar⁺ ion beam for one minute to clean the surface. The thermal stability of the bare Bi₂S₃ and heterojunction was investigated using thermogravimetric analysis (TGA) apparatus (NETZSCH, STA 449 F3, Jupiter). The test was carried out in air with a heating rate of 10 °C min⁻¹ from 100 °C to 900 °C. The porosity and specific surface area of the samples were determined through nitrogen adsorption at 200 °C on the basis of the BET equation using a Quantachrome, FL-33426. The UV-visible absorption spectra of the ethanolic solutions containing Bi₂S₃, ZnO and Bi₂S₃/ZnO were recorded using a Shimadzu, UV-3600 spectrophotometer. The hydrogen evolution was measured by online gas chromatography using a YL Instrument, 6500GC system with a thermoconductive detector. The photoelectrochemical measurements have been tested using a galvanostat–potentiostat (PGSTAT302N, Autolab, The Netherlands) with a standard three-electrode cell and Pt wire as a counter electrode and saturated Ag/AgCl as a reference electrode. The working electrode was a thin film of as-prepared material on glassy carbon. The photocatalytic I–V characteristics have been measured using simulated illumination (60 mW cm⁻²) by a white light source.

Photocatalytic test

L-Bi₂S₃, L-Bi₂S₃/ZnO, M-Bi₂S₃ and M-Bi₂S₃/ZnO have been tested for dye degradation using methyl orange (MO) as a model pollutant. It is a representative of hazardous azo dyes. The photodegradation of MO (initial concentration C₀ = 0.3 × 10⁻⁴ M) was carried out within a quartz cell reactor containing 50 mL model solution with a concentration of 1 mg mL⁻¹ under UV and visible lamps. A Xe-arc lamp (250 W) with an incident beam intensity of 100 mW cm⁻² (Oriel, Irvine, CA) was used as a light source for visible irradiation and a 395 nm cutoff filter was used. For a control experiment, bare L-Bi₂S₃ and M-Bi₂S₃ have been studied separately under similar reaction conditions. The appropriate amount of aliquots was collected from the reactor at successive time intervals. The percentage degradation (% DE) of MO was determined using the following equation:where C₀ is the initial absorption intensity of MO at λ_max = 463 nm and C is the intensity after light illumination.

Photocatalytic hydrogen generation

To study the hydrogen generation, the online gas chromatography method was used where the area under the Gaussian peak gives the amount of H₂ evolved. The H₂ production through water splitting was performed in a closed reactor in the presence of methanol solution with saturated argon media and vigorous stirring. For this experiment, a 25 volume% methanol solution was used at room temperature.⁵¹ Here, methanol acts as a sacrificial agent and use of methanol is very much advantageous compared to other alcohols because it reduces the formation of more carbon-based sub-products being the simplest organic molecule and also accelerates the main intermediate (free radical) formation process.⁵¹ Here we report H₂ generation up to 3 h.

Results and discussion

Structural characterization

XRD was carried out to study the crystalline phase of the as-prepared materials. Fig. 1a and b show the X-ray diffraction (XRD) patterns of L-Bi₂S₃, L-Bi₂S₃/ZnO, M-Bi₂S₃ and M-Bi₂S₃/ZnO. The X-ray patterns displayed narrow and sharp diffraction peaks, indicating the high crystallinity of the prepared samples. The strong peaks at 2θ values of 15.54, 17.41, 22.21, 23.57, 24.92, 28.43, and 31.71 correspond to characteristic diffraction from the (020), (120), (220), (101), (130), (211) and (221) planes of pure Bi₂S₃ respectively (JCPDS card no. 17-0320).⁴⁷ The diffraction peaks of ZnO at 2θ values of 31.76, 34.00 and 36.56 correspond to the (100), (002) and (101) planes indicating the wurtzite hexagonal phase of ZnO (JCPDS card no. 36-1451).⁵² The XRD pattern of the pure ZnO nanoparticles is shown in the inset of Fig. 1a which indicates that the nanoparticles are highly crystalline in nature. Therefore, the XRD patterns of the heterostructures reveal the presence of wurtzite ZnO within orthorhombic Bi₂S₃. Moreover, the presence of all the characteristic peaks of Bi₂S₃ and no additional peak in the heterostructures confirms the purity of Bi₂S₃/ZnO. Moreover, no remarkable changes were observed in the diffraction peak position and intensity of Bi₂S₃/ZnO at longer hydrothermal reaction times indicating the formation of stable orthorhombic Bi₂S₃ (Fig. S1, ESI†).

Fig. 1 XRD patterns of as-prepared (a) pure L-Bi₂S₃ and L-Bi₂S₃ decorated with ZnO NPs. Inset: XRD pattern of ZnO nanoparticles. (b) Pure M-Bi₂S₃ rod-shaped structures and M-Bi₂S₃ decorated with ZnO NPs. The samples have been synthesized at 16 h and metal salt to surfactant ratio 1 : 3.

The morphology and growth of Bi₂S₃ and Bi₂S₃/ZnO were investigated by field emission scanning electron microscopy (FESEM). Fig. 2a–c illustrates the assembled growth of Bi₂S₃ rods with increasing reaction time. It is clear from Fig. 2a that when the hydrothermal reaction proceeds for 10 h, the formation of rod-like structures is obtained which are completely overlapped. At 16 h, distinct rods with an average length of ∼3.81 μm (Fig. 2b) are formed which are connected with each other and radiated from a centre.

Fig. 2 FESEM images of Bi₂S₃ using L-cysteine as a structure controlling agent at three different reaction times: (a) 10 h, (b) 16 h, and (c) 24 h (metal : L-cysteine 1 : 1.5). Effect of surfactant concentration during the formation of Bi₂S₃ at three different metal/surfactant ratios: (d) 1 : 1.5, (e) 1 : 3 and (f) 1 : 6 at 16 h under similar conditions.

With further increase in the reaction time up to 24 h, the rods are more assembled and interestingly form a lotus flower-like morphology (Fig. 2c). This may happen due to over growth of Bi₂S₃ nuclei in the same direction.⁴² In addition, the metal salt to surfactant ratio has been varied to investigate the role of the surfactant as a morphology controlling factor. Fig. 2d–f display Bi₂S₃ synthesized at three different concentrations of surfactant at 16 h keeping all other conditions similar.

When the metal salt : L-cysteine ratio is 1 : 1.5, it can be observed that irregularly distributed rod-shaped structures of average length ∼4.7 μm are formed (Fig. 2d). In contrast, at 1 : 3, the assembled morphology (length ∼5.5 μm) consists of rods that have been formed with a bunched-like structure.

Moreover, at a higher concentration of L-cysteine (1 : 6 ratio), more assembled structures have been formed. Fig. 2e displays Bi₂S₃ superstructures of diameter ∼8 μm which are built of two-dimensional nanosheets. Hence, this highly assembled morphology is useful to provide more active sites in photocatalytic applications. In general, an assembled 1D morphology is more suitable for photocatalytic applications having a large surface to volume ratio, synergistic interactions as the long lengths of the nanorods will contact each other, and multiple functionalities.⁵³

Another kind of Bi₂S₃ nanostructure has been developed using mercaptosuccinic acid as a structure controlling agent. Fig. 3a–c illustrates the FESEM images of pure Bi₂S₃ at different magnifications, and clearly indicates solid rods radiated from a common centre and stacked uniformly to form a nanoflower-like morphology.

Fig. 3 FESEM images of (a)–(c) flower-like assembled structures of Bi₂S₃ using mercaptosuccinic acid as a structure controlling agent at three different magnifications and (d) EDS spectrum of Bi₂S₃ synthesized at 16 h and a metal salt to surfactant ratio of 1 : 3.

At high magnification, it has been clearly observed that the nanoflowers consist of solid rods of length ∼500 nm with the square edge of the sides at 60 to 70 nm (Fig. 3c). To investigate the elemental composition, X-ray energy-dispersive spectrometry (EDS) was carried out on the Bi₂S₃ nanoflowers. Fig. 3d indicates the coexistence of Bi and S elements on the rod-shaped structures. However, the presence of Al and Si signals originated from the glass substrate.

Additionally, the detailed structural and crystalline nature of the as-prepared Bi₂S₃ and Bi₂S₃/ZnO has been investigated through TEM and HRTEM images (Fig. 4). The TEM images (Fig. 4a and c) clearly reveal the uniform 1D structures of L-Bi₂S₃ and M-Bi₂S₃ and are well consistent with the FESEM images. The interplanar spacing of 0.375 nm reveals the characteristic (101) plane of Bi₂S₃ (Fig. 4b and d).⁵⁴ Therefore, the preferential growth of Bi₂S₃ rods occurred along the (101) direction. The calculated size of the bare ZnO NPs is in the range of 5 nm as determined from the TEM image (Fig. S2, ESI†).

Fig. 4 TEM images of (a) pure L-Bi₂S₃ nanostructures, (b) HRTEM of L-Bi₂S₃, inset: SEAD pattern, (c) pure M-Bi₂S₃ nanostructures and (d) HRTEM image of M-Bi₂S₃ at metal : surfactant = 1 : 3 in reaction time 16 h.

The microscopic images of the L-Bi₂S₃/ZnO heterostructures are displayed in Fig. 5. It can clearly be seen from the SEM images (Fig. 5a and b) that the spherical ZnO NPs have been formed uniformly deposited on the L-Bi₂S₃ surface. A further detailed study on the morphology of the L-Bi₂S₃/ZnO heterostructures has been performed through the TEM images. Fig. 5c clearly shows the rough surface of the rod-shaped L-Bi₂S₃ due to deposition of spherical hexagonal ZnO nanoparticles over the surface. The HRTEM image of L-Bi₂S₃/ZnO reveals its highly crystalline nature having a lattice spacing of about 0.33 nm related to the (130) plane of Bi₂S₃ (Fig. 5d).⁵⁵ The SAED pattern depicts the highly crystalline hexagonal structure of Bi₂S₃ (Fig. 5e). For further analysis of chemical compositions and the elemental distribution, EDS was carried out (Fig. 5f) where strong signals of Bi, S, O and Zn also confirm the formation of Bi₂S₃/ZnO heterostructures.

Fig. 5 (a and b) FESEM images of L-Bi₂S₃/ZnO heterostructures at two different magnifications. TEM image of (c) ZnO nanoparticles decorated on rod-shaped Bi₂S₃ structures, and (d) HRTEM, (e) SAED and (f) EDX spectrum of L-Bi₂S₃/ZnO heterostructures.

The morphology and crystalline nature of the heterostructures using M-Bi₂S₃ was further investigated using FESEM and TEM images. Fig. 6a shows the in situ generation of ZnO NPs on the rod-shaped M-Bi₂S₃ structures. It is evident that the assembled nanoflower-like structures are totally transformed into a nanofibre-based network after the formation of ZnO NPs. This is probably due to the strong interaction between ZnO and the surface ligand (MSA) of Bi₂S₃ which is responsible for the formation of the self-assembled nanoflower-like morphology. The FESEM image also illustrates that the single rod-shaped structure of Bi₂S₃ is completely covered by small ZnO NPs (Fig. 6b). Fig. 6c shows that the M-Bi₂S₃/ZnO rod-like nanostructures are covered with smaller nanoparticles which is consistent with the FESEM image (Fig. 6a and b) and the cross fringes in the HRTEM image indicate the crystallinity of the heterostructures with a lattice spacing value of 0.30 nm related to the (211) plane of Bi₂S₃ (Fig. 6d).⁵⁶ Owing to the highly crystalline nature of Bi₂S₃, it is difficult to observe crystal fringes of ZnO within the heterostructures.

Fig. 6 FESEM images of M-Bi₂S₃/ZnO heterostructures at (a) low magnification, and (b) high magnification. TEM image of (c) ZnO nanoparticles decorated on rod-shaped Bi₂S₃ structures, and (d) HRTEM of M-Bi₂S₃/ZnO heterostructures.

The growth of bare Bi₂S₃ and its heterostructures can be explained on the basis of coordination interaction between Bi³⁺ and the surfactant. In the hydrothermal synthesis of the Bi₂S₃ nanostructures and in situ generation method for heterojunction formation, the following reactions may be involved:

Bi(NO₃)₃·5H₂O → Bi₃⁺ + 3NO₃⁻ + 5H₂O (2)

L-Cys + Bi(NO₃)₃ → Bi(III)–L-Cys (3)

MSA + Bi(NO₃)₃ → Bi(III)–MSA complex (4)

Initially, L-cysteine contains various functional groups such as –NH₂, –COOH, and –SH^57–59 and MSA contains –SH, and –COOH groups which have a tendency to coordinate with inorganic cations and metals. Thereafter, the amino group reacts with the neighbouring carboxylic group of the surfactant to form a dipeptide or polypeptide because these complex forms are more stable relative to others.^42,54 Moreover, this intermediated state serves as a template in the successive nucleation of Bi₂S₃ nanocrystals. Finally, the rod-shaped Bi₂S₃ originates from continuous growth of Bi₂S₃ nuclei along one direction which may be caused by a typical Ostwald ripening process.⁴² From the effect of different reaction times, it can be concluded that the transformation of the assembled flowers from an undefined structure may be caused due to recrystallization along the preferential growth axis to form the rod-composed flower as shown in the FESEM images (Fig. 2). To get the ZnO-loaded Bi₂S₃ rods, direct adsorption and in situ generation of 5 nm ZnO NPs on the Bi₂S₃ have been followed. Possible mechanisms of heterostructure formation are displayed in Scheme 1a and b.

Scheme 1 (a) Schematic illustration of ZnO nanoparticle decorated Bi₂S₃ assembled structures synthesized by a simple hydrothermal process using L-cysteine hydrochloride as a structure controlling agent. (b) Schematic illustration of ZnO nanoparticle decorated Bi₂S₃ rods synthesized by a simple hydrothermal method using mercaptosuccinic acid as a structure controlling agent.

X-ray photon spectroscopy (XPS) was carried out to further elucidate the chemical compositions as well as the oxidation states of the pure semiconductor and heterostructures. The overall XPS spectra (Fig. 7a) indicate the presence of strong peaks of Bi, S, Zn, O and C. Here, the 1S peak of C acts as a reference point coming from the background.

Fig. 7 XPS spectra of the as-prepared heterostructures. (a) The overall spectra of M-Bi₂S₃ and M-Bi₂S₃/ZnO. (b) The Bi 4f and S 2p spectra, (c) Zn 2p spectra and (d) O 1s spectra of M-Bi₂S₃/ZnO. Here, the scatter and solid lines indicate the experimental and fitted data.

The strong peaks at binding energies of 157.84 eV and 163.16 eV with typical spin orbit doublet splitting of 5.32 eV can be assigned to the binding energy of Bi 4f_7/2 and Bi 4f_5/2 respectively, closely matched with the Bi³⁺ in Bi₂S₃ (Fig. 7b).⁶⁰ The peaks found between Bi 4f_7/2 and Bi 4f_5/2 at 161.83 eV and 160.56 eV correspond to S 2p_1/2 and S 2p_3/2 (1.27 eV), which indicates the existence of S²⁻ within Bi₂S₃.⁴² Moreover, after the heterojunction formation, the blue shift of the binding energy for Bi 4f_7/2 (∼0.3 eV), Bi 4f_5/2 (∼0.27 eV) and S (∼0.19 eV) indicates the strong interaction between Bi₂S₃ and ZnO followed by the formation of a heterojunction. In addition, the strong oxidation peaks located at 1021.9 eV and 1045.1 eV of Zn 2p_3/2 and 2p_1/2 respectively (Fig. 7c) suggest the Zn²⁺ state of ZnO.⁴⁹ It is also observed that the binding energy of pure ZnO shifts towards a lower value of 0.23 eV after forming a heterojunction. Fig. 7d exhibits the asymmetric profile of O 1s which can be fitted to two symmetrical peaks at 530.2 and 531.72 eV, indicating the presence of two different kinds of O species in the sample. The peaks at 531.72 eV and 530.2 eV should be associated with the lattice oxygen (O_L) of ZnO and chemisorbed oxygen (OH) by the surface.⁶¹ Thus, the XPS spectra suggest the co-existence of Bi₂S₃ and ZnO within the heterostructures.

Thermo gravimetric analysis (TGA) was carried out to determine the loading of ZnO NPs after in situ deposition as well as the thermal stability of the heterostructure. The loading of ZnO calculated from the TGA curves is found to be 10% and 15% for L-Bi₂S₃ and M-Bi₂S₃ respectively (Fig. 8a and b). The initial mass losses in the range of 220 °C to 230 °C originate from the removal of adsorbed O₂ and water molecules. The further mass loss above 380 °C is mainly attributed to the decomposition of Bi₂S₃ into Bi metal.

Fig. 8 TG curves of (a) pure L-Bi₂S₃ and L-Bi₂S₃/ZnO, and (b) pure M-Bi₂S₃ and M-Bi₂S₃/ZnO. (c) Diffuse reflectance spectra of L-Bi₂S₃ (pink line), L-Bi₂S₃/ZnO (blue line), M-Bi₂S₃ (black line) and M-Bi₂S₃/ZnO (red line) respectively. Inset: Absorption spectra of ZnO nanoparticles (5 nm). (d) Photoluminescence spectra of ZnO (5 nm), L-Bi₂S₃/ZnO and M-Bi₂S₃/ZnO.

Notably, after 440 °C, the pure Bi₂S₃ curve is fairly stable whereas the Bi₂S₃/ZnO heterostructures are unstable and indicate a strong weight loss above 478 °C which implies the formation of ZnO.⁴² The porous structure of the pure semiconductor as well as the heterostructures was studied by the nitrogen adsorption desorption isotherm method. The specific surface areas calculated by the BJH method are 3.13 m² g⁻¹ and 9.53 m² g⁻¹ for the pure M-Bi₂S₃ and M-Bi₂S₃/ZnO heterostructures, respectively. This indicates that the specific BET surface area is increased when the heterojunction is formed which may be useful for photocatalytic applications.⁵⁶ The hysteresis loop present in the BET curve (Fig. S3, ESI†) suggests the type II pattern of the semiconductor heterostructures. Hence, incorporation of ZnO nanoparticles effectively generates porosity and active surface sites within the heterostructure.

The optical properties of the pure semiconductor and semiconductor heterostructures have been evaluated by diffuse reflectance and photoluminescence (PL) spectroscopy. Fig. 8c shows the diffuse reflectance spectroscopy of the bare semiconductors and heterostructures where both the L-Bi₂S₃/ZnO and M-Bi₂S₃/ZnO heterostructures demonstrated broad absorption in the visible range that extends up to the near IR range. For the bare ZnO, the onset of light absorption is in the UV region with a band gap of ∼3.37 eV. The light absorption of the Bi₂S₃/ZnO heterostructures increases in the visible region compared to bare ZnO or Bi₂S₃ indicating that heterostructures are useful for efficient solar light harvesting applications. The photoluminescence spectra of ZnO display an emission peak at ∼514 nm upon excitation to 320 nm, as shown in Fig. 8d. The intensity of the emission peak lowered significantly to 519 nm and 517 nm when L-Bi₂S₃ and M-Bi₂S₃ attached to the ZnO NPs. Thus, the quenching in emission intensity (65.5% for M-Bi₂S₃/ZnO and 83.25% for L-Bi₂S₃/ZnO) for the heterostructures (Fig. 8d) indicates the strong electronic interaction and improved charge carrier separation efficiency.¹⁷

Photocatalytic MO degradation

The photocatalytic activity has been studied under visible light using methyl orange (MO) as a model pollutant. It is well known that MO is a very stable dye which is resistant to self-photo degradation. Fig. 9a clearly shows the degradation of MO in the presence of Bi₂S₃ and Bi₂S₃/ZnO under visible light irradiation. The M-Bi₂S₃/ZnO heterostructures exhibit significantly enhanced photocatalytic activity as compared with bare M-Bi₂S₃ and L-Bi₂S₃. Bare L-Bi₂S₃ has negligible photocatalytic activity (26%) but bare M-Bi₂S₃ shows high activity (63%). In contrast, the photodegradation efficiency of MO reached 82% for M-Bi₂S₃/ZnO and 55% for L-Bi₂S₃/ZnO after 150 min of visible light irradiation (Fig. 9a), indicating that heterostructures possess superior photocatalytic activity. The photocatalytic activity of the M-Bi₂S₃/ZnO heterostructures is ∼8.6 times higher compared to bare M-Bi₂S₃.

Fig. 9 (a) Photocatalytic degradation of methyl orange in the presence of catalyst, bare semiconductors L-Bi₂S₃ and M-Bi₂S₃, and heterostructures L-Bi₂S₃/ZnO and M-Bi₂S₃/ZnO under visible light irradiation. (b) Effect of argon, isopropanol, and Cu²⁺ on the photocatalytic activity of M-Bi₂S₃/ZnO for methyl orange degradation.

The enhanced photocatalytic activity of heterostructures can be achieved due to strong absorption in the visible range with high surface area of the assembled structure.⁶² As evident from the TGA data, loading of ZnO NPs on the M-Bi₂S₃ nanofibres may create more catalytic centres within the heterostructure and thereby show high catalytic activity. The enhanced catalytic activity of M-Bi₂S₃/ZnO may be associated with the multiple reflections and scattering of light within the interconnected rods which results in enhanced light absorption as reflected in the absorption spectra (Fig. 8c).²⁸ Similarly, L-Bi₂S₃/ZnO demonstrated higher catalytic activity in comparison with bare L-Bi₂S₃. The photocatalytic activity under UV light irradiation was also studied where M-Bi₂S₃/ZnO degrades only 48% of MO concentration (Fig. S4, ESI†).

Possible reactions involved in photocatalytic dye degradation are given below.

Bi₂S₃/ZnO → h_VB⁺ + e_CB⁻ (7)

O₂ + e⁻ → O₂˙⁻ (8)

O₂˙⁻ + h⁺ → HO₂˙ (9)

2H₂O + O₂˙⁻ → 4OH˙ (10)

In order to investigate the photocatalytic mechanism of the heterostructure-based catalysts, a further study has been conducted in the presence of sacrificial agents and the contribution of excess electron–holes (Cu²⁺, isopropanol and saturated argon media) (Fig. 9b).⁶² Semiconductors react with photoinduced electrons and holes from water to form various reactive species including O₂˙⁻ and OH˙.⁶³ Herein, Cu²⁺ was used to understand the role of excess electrons as it reacts with electrons very quickly and converts into Cu⁺. The remarkable change of MO degradation in the presence of 2 × 10⁻⁶ M Cu²⁺ (it changes 68% to 43%) is presented in Fig. 9b after 6 h visible light irradiation. This suggests a role of electrons during the photocatalytic reaction. In order to confirm the role of O₂˙⁻, a photodegradation experiment has been performed under argon saturated atmosphere. In general, the argon-saturated inert medium may suppress the O₂˙⁻ radical production which may be the possible reason for inhibiting the degradation of MO by M-Bi₂S₃/ZnO.⁶² It only degrades 20% of MO after 4 h of visible light irradiation. Thus, it can be concluded that oxygen has a crucial role in the photocatalytic reaction. Moreover, photocatalysis has been performed in the presence of 0.1 M isopropanol (acts as a hole scavenger) and oxygen to investigate the role of photoexcited holes in the photocatalytic reactions. The decomposition kinetics of MO increased up to 94% in the presence of isopropanol which clearly indicates the role of holes in the photodegradation of MO.

Photocatalytic H₂ generation

Fig. 10a shows photocatalytic H₂ generation by Bi₂S₃ and Bi₂S₃/ZnO heterostructures via water splitting using 25 volume% of methanol solution as a sacrificial agent. In general, the semiconductors with more negative CB can reduce water efficiently, and thus increase the hydrogen production rate. Herein, a relatively low band gap semiconductor, Bi₂S₃, acts as a recombination centre for ZnO and the reduction process occurs at the ZnO surface. The H₂ generation enhanced with the increase in irradiation time. The amount of H₂ production is higher for heterostructures, L-Bi₂S₃/ZnO (2791 μmol). For comparison, the H₂ evolution activity has been investigated for bare semiconductors L-Bi₂S₃ and M-Bi₂S₃ under visible light irradiation. The heterojunction shows 2.74-fold enhancement in hydrogen generation compared to L-Bi₂S₃ (1020 μmol). The H₂ generation for M-Bi₂S₃ and M-Bi₂S₃/ZnO is 263 μmol and 2450 μmol respectively. In contrast to dye degradation, L-Bi₂S₃/ZnO shows superior performance for photocatalytic hydrogen generation in comparison to M-Bi₂S₃/ZnO.

Fig. 10 (a) Photocatalytic hydrogen generation in the presence of catalyst L-Bi₂S₃, M-Bi₂S₃, L-Bi₂S₃/ZnO, and M-Bi₂S₃/ZnO for 3 h under visible light from an aqueous solution containing 25 volume% methanol at pH 7. (b) Effect of pH on hydrogen evolution from aqueous solution after 1 h and 3 h of irradiation.

The exact reason is not clear to us but the enhanced catalytic activity may be associated with the assembled morphology of L-Bi₂S₃/ZnO which leads to intimate contact between metal sulphide and the zinc oxide NPs. The three-dimensional interconnected assembled structure may also facilitate the transport of photogenerated electrons and holes to the binding sites; thereby water oxidation occurred in the presence of a sacrificial agent which is consistent with the literature report.⁶⁴ Moreover, both the heterostructures may provide facile electron transfer compared with bare semiconductors which can effectively enhance the hydrogen generation.⁶⁵

Furthermore, the effect of pH on the photocatalytic hydrogen evolution process has been studied to optimize the reaction conditions. Fig. 10b illustrates that higher activity for hydrogen evolution is achieved at pH 9, ∼8089 μmol, whereas under low pH conditions, i.e. acidic conditions, it generates only 1159 μmol due to protonation of the photocatalyst in acidic solution.⁶⁵ Generally under acidic conditions the driving force for hydrogen generation decreases as the redox potential of H⁺/H₂ becomes more negative. On the other hand, a highly alkaline medium is also not preferable for hydrogen evolution because of insufficient protons.⁶⁶

In view of practical applications, reusability and stability of photocatalysts is an important parameter. Here, we checked the recycling of M-Bi₂S₃/ZnO up to 5 successive cycles and the result is displayed in Fig. 11a, which reveals nearly 10% loss in MO degradation. In hydrogen generation, a recycling experiment of L-Bi₂S₃/ZnO has been performed up to the 5th cycle displayed in Fig. 11b. After the 5th cycle, no remarkable decline was found in the hydrogen evolution rate. Therefore, it can be concluded that the M-Bi₂S₃/ZnO and L-Bi₂S₃/ZnO heterostructures are stable and reusable visible light active photocatalysts for organic pollutant degradation as well as for hydrogen generation. To ensure the structural stability and morphology of the catalyst, XRD and FESEM were re-examined before and after catalytic reactions. As shown in Fig. 11c, the XRD patterns of the ZnO decorated Bi₂S₃ heterostructures almost remain the same and all the characteristic peaks are present. The FESEM image of the heterostructures reflects a similar kind of morphology after the photocatalytic reaction. Thus the semiconductor-based coupled heterostructures are stable under long visible light irradiation and represent a fruitful approach for solar light energy harvesting.

Fig. 11 Recycling test of (a) M-Bi₂S₃/ZnO during MO degradation and (b) L-Bi₂S₃/ZnO during hydrogen generation. (c) The XRD pattern and (d) FESEM image of L-Bi₂S₃/ZnO after and before the catalytic reaction of MO degradation and H₂ generation at pH 7.

Photoelectrochemical performance

In order to investigate the photoelectrochemical activity of the Bi₂S₃/ZnO heterostructures, it is important to find out the band edge potential of both Bi₂S₃ and ZnO as band edge potentials play a crucial role in determining the migration routes of photo-generated electrons and holes. Additionally, photocatalytic activity directly depends on optical absorption, phase structure, morphology and separation efficiency of photo-generated charge carriers. The valence band (VB) and conduction band (CB) of both the semiconductors were calculated using the following empirical equations,⁴²

E_VB = x − E_fe + 1/2E_g (11)

E_CB = x − E_fe − 1/2E_g (12)

where E_VB and E_CB are the valence and conduction band edge potential respectively, and x is the geometric mean of the electronegativity of the constituent atoms. E_fe is the energy of an electron on the hydrogen scale (4.5 eV) and E_g is the band gap of the semiconductor. The VB and CB edge potentials of Bi₂S₃ calculated using the above equations are 1.45 eV and 0.08 eV while for ZnO they are 3 eV and −0.2 eV respectively.^67,68 This difference in band edge potential is useful for better charge separation and migration which is favourable for photoelectrochemical performance.^68,69 The photoelectrochemical activity of the catalyst has been examined by making a layer on an FTO-coated glass substrate through linear sweep voltammetry (LSV) under dark and light conditions with light irradiation. A 300 W xenon lamp with a water filter of 1 M NaNO₂ solution was used as the visible light (≥395 nm) source. Fig. 12a–c shows the difference in photo current density measured via LSV of the M-Bi₂S₃, ZnO and M-Bi₂S₃/ZnO catalysts in 0.1 M KOH solution using Pt wire as a counter and Ag/AgCl as a reference electrode at a scan rate of 20 mV s⁻¹. The photoelectrochemical performance in terms of current density is presented in Table 1. Bi₂S₃ shows a photocurrent response and the photocurrent density reached up to 0.11 mA cm⁻² (Fig. 12a). A weak photocurrent was obtained for bare ZnO upon illumination in the applied potential range as shown in Fig. 12b. In contrast, the Bi₂S₃/ZnO heterostructures demonstrated an enhanced photocurrent density of 0.25 mA cm⁻² which is a 56% enhancement in comparison with bare ZnO NPs and Bi₂S₃.

Fig. 12 Photoelectrochemical current density vs. potential plot of (a) M-Bi₂S₃, (b) ZnO (5 nm) and (c) M-Bi₂S₃/ZnO via the LSV method without light (dark) and under light illumination conditions at a scan rate 20 mV s⁻¹ using 0.1 M KOH solution as an electrolyte, Pt wire as a counter electrode and Ag/AgCl as a reference electrode and (d) time dependence of photocurrent density at an external bias of 0.26 V vs. Ag/AgCl with illumination, at a scan rate of 20 mV s⁻¹ for bare M-Bi₂S₃ and heterojunction M-Bi₂S₃/ZnO.

Table 1 Comparative study of current density measured from the photoelectrochemical studies performed in 0.1 M KOH solution as an electrolyte, Pt wire as a counter electrode and saturated Ag/AgCl as a reference electrode within the voltage range −0.4 V to 0.8 V under dark and light conditions

Catalyst	Dark current (Idark) (mA cm^−2)	Photo current (Ilight) (mA cm^−2)	Photo current gain Ilight/Idark
M-Bi2S3	0.039	0.112	2.88
ZnO	0.004	0.007	1.79
M-Bi2S3/ZnO	0.029	0.255	8.79

---

This result implies that the M-Bi₂S₃/ZnO heterostructures show higher photoelectrochemical activity and a photo current gain which is ca. 3.05 times increased compared to the single component M-Bi₂S₃. Notably, bare Bi₂S₃ also shows high photocurrent but due to fast recombination problems its photocurrent gain reduces. Thereby, the introduction of ZnO with Bi₂S₃ enhances light absorption and charge separation efficiency. The stability of pure semiconductor as well as heterostructures was investigated by the chronoamperometry method under light illumination for 200 seconds (Fig. 12d).⁶⁹ In the presence of light, a certain change in current density is displayed for M-Bi₂S₃ whereas a gradually increasing tendency is exhibited for the coupled system M-Bi₂S₃/ZnO because of a charge transfer mechanism.¹⁷ As the photogenerated electrons migrate through the CB, more and more electrons are gathered at the active surface site and gradually increase the current density. Remarkably, Fig. 12d shows the long time i–t response (up to 600 s) and the current density remains almost stable which indicates that the catalysts are durable against photocorrosion. Hence, M-Bi₂S₃/ZnO shows enhanced photoelectrochemical activity compared to bare Bi₂S₃ and ZnO.

The photocurrent for Bi₂S₃ and the heterostructures shows 7.39 and 8.06 fold enhancement in current density compared with dark conditions at a voltage of 1.0 V respectively (Fig. S5, ESI†). This enhancement in current density may happen due to enhanced charge carrier separation in the presence of light irradiation. The linear nature of the I–V plot indicates the good ohomic contact between the semiconductor and the ITO-coated glass substrate which is a good sign for device applications.⁷⁰ When light falls on the catalyst surface, excess photocarriers are generated which leads to charge separation and electrons are efficiently transferred to the electrode when voltage is applied.⁷¹ The fluorescence decay curves at an excitation wavelength of 400 nm for ZnO and Bi₂S₃/ZnO have been examined to investigate the decay dynamics of the heterostructures.

It is observed that the photocurrent is higher for heterostructures than single semiconductors. These charge carrier separation phenomena also enhance the mobility of electrons. The on/off photocurrent has been measured at 20 second intervals at a bias voltage 500 mV (Fig. 13a), which exhibits two distinct states – one is a high current state when the light is on and the other is a low current state while the light is off. This time-dependent response of the Bi₂S₃/ZnO-coupled system is preferable for optoelectronic device applications. In the photocatalytic process, super oxide radicals and holes have a crucial role in organic dye degradation which has been previously confirmed by lots of tests. Hence, the photo-excited electron and holes are responsible for generating these free radicals. Moreover, the enhanced charge separation have been studied via picosecond resolved time correlated single photon counting (TCSPC). Fig. 13b shows a decay curve of ZnO with an average life time of 4.34 ns. Interestingly, it decreases to 0.338 ns when Bi₂S₃ is attached with ZnO. From this sharp decrease in average life time, it can be concluded that fast electron transfer happens within the heterostructures through the CB of ZnO to Bi₂S₃.¹⁷ This charge transfer reduces the recombination rate, which can also be understood from the PL study, confirming that more electrons may participate for photocatalytic water splitting and H₂ generation (Table 2).

Fig. 13 (a) Comparative time-dependent photocurrent density response of the (i) M-Bi₂S₃/ZnO heterostructures and (ii) pure M-Bi₂S₃ with illumination switched on and off in air at a bias of 500 mV and (b) the picosecond-resolved PL spectra of ZnO (5 nm), M-Bi₂S₃ and M-Bi₂S₃/ZnO heterostructures.

Table 2 Dynamics of picosecond-resolved luminescence transients and decay parameters of ZnO (5 nm) and Bi₂S₃/ZnO

Sample	Excitation wavelength (nm)	Detection wavelength (nm)	τ 1 (ns)	τ 2 (ns)	τ avg (ns)
ZnO NPs	400	500	0.82 (51.8%)	29.30 (7.5%)	4.42
Bi2S3/ZnO	400	500	0.25 (42.0%)	5.02 (7.9%)	0.338

---

Depending on the above results and discussion, a probable mechanism of photocatalytic activity can be described as follows. In the presence of solar light, both the semiconductors absorb light and the electrons in the VB get excited up to a higher potential of −1.53 eV for Bi₂S₃ and −0.2 eV for ZnO.⁵⁵ Therefore, the effective charge transfer process proceeds within the semiconductor due to high photon energy.⁴² Generally, the excited electrons and holes always wish to transfer to the nearest recombination centre. The CB electrons of the Bi₂S₃ rods can transfer to the CB of the ZnO NPs and simultaneously the excited holes of ZnO can migrate to the Bi₂S₃ rods and consequently reduce the first recombination process shown in Scheme 2. These photo-generated electrons and holes generate oxidative radicals (in the presence of O₂) such as h⁺, OH˙, and O₂˙⁻ by the oxidation of O₂ and the reduction of H₂O which are mainly responsible for dye degradation. In hydrogen evolution, the sacrificial agent methanol acts as a hole scavenger and generates more OH˙ free radicals and accelerates the H₂ production rate. As the band position of Bi₂S₃ is not suitable for H₂ production, it can be concluded that photochemical reactions occur at the ZnO surface and incorporation of a narrow band gap semiconductor helps in charge separation, leading to more efficient activity in H₂ generation.^69–72 On the other hand, the visible light activity can be explained on the basis of a hole-trapping mechanism which has been explained in the scheme diagram. Therefore, semiconductor heterojunctions are a suitable route for creating more recombination centres and thus increase the life time of electrons as well as reduce the recombination rate.

Scheme 2 Proposed diagram for the possible band edge energy for charge separation and migration in the Bi₂S₃/ZnO heterostructures.

Conclusions

In summary, an assembled nanoflower morphology of Bi₂S₃ decorated with ZnO nanoparticles has been successfully synthesized by a facile in situ deposition method. The strong absorption of Bi₂S₃/ZnO heterostructures in the visible region and enhanced charge separation in the heterostructures makes them promising candidates for solar light harvesting applications. The experimental results indicate that M-Bi₂S₃/ZnO demonstrates high catalytic activity for organic pollutant degradation, whereas L-Bi₂S₃/ZnO shows superior performance for photocatalytic H₂ generation through water splitting. The photocatalytic activity of the heterostructures is ∼8.6 times higher compared to bare Bi₂S₃. The L-Bi₂S₃/ZnO heterostructures (2791 μmol) exhibit 2.74 fold enhancements in photocatalytic hydrogen generation compared to L-Bi₂S₃ (1020 μmol) under visible light. Notably, Bi₂S₃/ZnO demonstrates high photoelectrochemical activity with a photo current density of 0.25 mA cm⁻², which is 56% and 97% higher than that of bare M-Bi₂S₃ and ZnO respectively under similar reaction conditions. The band edge potential of the heterostructures is suitable for sufficient charge separation which causes enhancement in photo current. Hence, the present synthetic methodology can be employed to prepare efficient, low cost heterostructure-based photocatalysts to substitute the common use of noble metal-based catalysts, and Bi₂S₃/ZnO has potential in environmental remediation and water-splitting applications.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

One of the authors (SB) is thankful to Department of Science & Technology (DST), India, for providing an INSPIRE fellowship. Another author (SG) is thankful to the Council of Scientific & Industrial Research (CSIR), India, for providing a CSIR-Senior Research Associateship (Scientists’ Pool Scheme). The authors also wish to thank Dr Sunirmal Jana and Dr Rajat Banerjee of this institute for their help by providing some of the instrumental facilities. The authors acknowledge Director, CSIR-CGCRI for his kind permission to publish the work.

References

1. M. R. Hoffmann, S. T. Martin, W. Choi and D. W. Bahnemann, Chem. Rev., 1995, 95, 69–96.
2. T. Jafari, E. Moharreri, A. S. Amin, R. Miao, W. Song and S. L. Suib, Molecules, 2016, 21, 900–929.
3. Z. G. Zou, J. H. Ye, K. Sayama and H. Arakawa, Nature, 2001, 414, 625–627.
4. D. J. Martin, K. Qiu, S. A. Shevlin, A. D. Handoko, X. Chen, Z. Guo and J. Tang, Angew. Chem., Int. Ed., 2014, 53, 9240–9245.
5. A. Kudo and Y. Miseki, Chem. Soc. Rev., 2009, 38, 253–278.
6. S. Ghosh, N. A. Kouame, S. Remita, L. Ramos, F. Goubard, P.-H. Aubert, A. Dazzi, A. Deniset-Besseau and H. Remita, Sci. Rep., 2015, 5, 18002–18011.
7. T. Hisatomi, J. Kubota and K. Domen, Chem. Soc. Rev., 2014, 43, 7520–7535.
8. Y. H. Ng, A. Iwase, A. Kudo and R. Amal, J. Phys. Chem. Lett., 2010, 1, 2607–2612.
9. H. Wang, L. Zhang, Z. Chen, J. Hu, S. Li, Z. Wang and J. Liu, Chem. Soc. Rev., 2014, 43, 5234–5244.
10. A. Fujishima and K. Honda, Nature, 1972, 238, 37–38.
11. J. Tian, Z. Zhao, A. Kumar, R. I. Boughton and H. Liu, Chem. Soc. Rev., 2014, 43, 6920–6937.
12. L. Xu, G. A. Sewvandi, S. Uemura, T. Kusunose, S. Nakanishia and Q. Feng, New J. Chem., 2017, 41, 10998–11008.
13. G. Wang, H. Wang, Y. Ling, Y. Tang, X. Yang, R. C. Fitzmorris, C. Wang, J. Z. Zhang and Y. Li, Nano Lett., 2011, 11, 3026–3033.
14. J. Virkutyte and R. S. Varma, New J. Chem., 2010, 34, 1094–1096.
15. M. D. Hernandez-Alonso, F. Fresno, S. Suarez and J. M. Coronado, Energy Environ. Sci., 2009, 2, 1231–1257.
16. S. Piskunov, O. Lisovski, J. Begens, D. Bocharov, Y. F. Zhukovskii, M. Wessel and E. Spohr, J. Phys. Chem. C, 2015, 119, 8686–18696.
17. S. Sardar, P. Kar, H. Remita, B. Liu, P. Lemmens, S. K. Pal and S. Ghosh, Sci. Rep., 2015, 5, 17313–17327.
18. J. Dutta, M. A. Mahmood and S. Baruah, Mater. Chem. Phys., 2011, 130, 531–535.
19. Q. Shen, X. Zhao, S. Zhou, W. Hou and J.-J. Zhu, J. Phys. Chem. C, 2011, 115, 17958–17964.
20. K. M. Lee, C. W. Lai, K. S. Ngai and J. C. Juan, Water Res., 2016, 88, 428–448.
21. L. Zhang, L. Du, X. Yu, S. Tan, X. Cai, P. Yang, Y. Gu and W. Mai, ACS Appl. Mater. Interfaces, 2014, 6, 3623–3629.
22. G. Marci, V. Augugliaro, M. J. López-Muñoz, C. Martin, L. Palmisano, V. Rives, M. Schiavello, R. J. D. Tilley and A. M. Venezia, J. Phys. Chem. B, 2001, 105, 1026–1032.
23. R. Brahimi, Y. Bessekhouad, A. Bouguelia and M. Trari, Catal. Today, 2007, 122, 62–65.
24. S. Boumaza, B. Bellal, A. Boudjemaa and M. Trari, Sol. Energy, 2016, 139, 444–451.
25. N. Qin, Y. Liu, W. Wu, L. Shen, X. Chen, Z. Li and L. Wu, Langmuir, 2015, 31, 1203–1209.
26. S. J. Jum, G. K. Hyun, A. J. Upendra, W. J. Ji and S. L. Jae, Int. J. Hydrogen Energy, 2008, 33, 5975–5980.
27. S. K. Dutta, S. K. Mehetor and N. Pradhan, J. Phys. Chem. Lett., 2015, 6, 936–944.
28. X. Li, J. Yu and M. Jaroniec, Chem. Soc. Rev., 2016, 45, 2603–2636.
29. J. Zhang, G. Xiao, F.-X. Xiao and B. Liu, Mater. Chem. Front., 2017, 1, 231–250.
30. L. Jing, J. Zhou, J. R. Durrant, J. Tang and D. Liu, Energy Environ. Sci., 2012, 5, 6552–6558.
31. C. Sotelo-Vazquez, R. Quesada-Cabrera, M. Ling, D. O. Scanlon, A. Kafizas, P. K. Thakur, T.-L. Lee, A. Taylor, G. W. Watson, R. G. Palgrave, J. R. Durrant, C. S. Blackman and I. P. Parkin, Adv. Funct. Mater., 2017, 27, 1605413.
32. S. J. A. Moniz, S. A. Shevlin, X. An, Z.-X. Guo and J. Tang, Chem. – Eur. J., 2014, 20, 15571–15579.
33. M. Xie, X. Fu, L. Jing, P. Luan, Y. Feng and H. Fu, Adv. Energy Mater., 2013, 4, 1300995.
34. F.-X. Xiao, ACS Appl. Mater. Interfaces, 2012, 4, 7055–7063.
35. C. Eley, T. Li, F. Liao, S. M. Fairclough, J. M. Smith, G. Smith and S. C. E. Tsang, Angew. Chem., Int. Ed., 2014, 53, 7838–7842.
36. A. Abdi, A. Denoyelle, N. Commenges-Bernole and M. Trari, Int. J. Hydrogen Energy, 2013, 38, 2070–2078.
37. L. X. Hao, G. Chen, Y. G. Yu, Y. S. Zhou, Z. H. Han and Y. Liu, Int. J. Hydrogen Energy, 2014, 39, 14479–14486.
38. F. Dong, Z. Ni, P. Li and Z. Wu, New J. Chem., 2015, 39, 4737–4744.
39. Q.-Z. Huang, Y. Xiong, Q. Zhang, H.-C. Yao and Z.-J. Li, Appl. Catal., B, 2017, 209, 514–522.
40. H. He, S. P. Berglund, P. Xiao, W. D. Chemelewski, Y. Zhang and C. B. Mullins, J. Mater. Chem. A, 2013, 1, 12826–12834.
41. M. Li, J. Wang, P. Zhang, Q. Deng, J. Zhang, K. Jiang, Z. Hu and J. Chu, Sci. Rep., 2017, 7, 42484–42492.
42. R. P. Panmand, Y. A. Sethi, R. S. Deokar, D. J. Late, H. M. Gholap, J.-O. Baeg and B. B. Kale, RSC Adv., 2016, 6, 23508–23517.
43. C. Zhang, Y. Zhou, Y. Zhang, S. Zhao, J. Fang and X. Sheng, New J. Chem., 2017, 41, 11089–11096.
44. J. Zhou, G. Tian, Y. Chen, Y. Shi, C. Tian, K. Pan and H. Fu, Sci. Rep., 2014, 4, 4027–4035.
45. X. Gao, H. B. Wu, L. Zheng, Y. Zhong, Y. Hu and X. W. Lou, Angew. Chem., Int. Ed., 2014, 53, 5917–5921.
46. F. Dong, X. Feng, Y. Zhang, C. Gao and Z. Wu, RSC Adv., 2015, 5, 11714–11723.
47. D. Lu, H. Wang, X. Zhao, K. K. Kondamareddy, J. Ding, C. Li and P. Fang, ACS Sustainable Chem. Eng., 2017, 5, 1436–1445.
48. S. S. Thind, C. C. Mustapic, J. Wen, C. D. Goodwin and A. Chen, New J. Chem., 2017, 41, 10542–10549.
49. Q. Liu, J. Huan, N. Hao, J. Qian, H. Mao and K. Wang, ACS Appl. Mater. Interfaces, 2017, 9, 18369–18376.
50. H. Bao, X. Cui, C. M. Li, Y. Gan, J. Zhang and J. Guo, J. Phys. Chem. C, 2007, 111, 12279–12283.
51. M. G. Mendez-Medrano, E. Kowalska, A. Lehoux, A. Herissan, B. Ohtani, S. Rau, C. Colbeau-Justin, J. L. Rodriguez-Lopez and H. Remita, J. Phys. Chem. C, 2016, 120, 25010–25022.
52. B. Li, Y. Zhang, R. Du, L. Gan and X. Yu, Langmuir, 2016, 32, 11639–11645.
53. S. Hernández, V. Cauda, A. Chiodoni, S. Dallorto, A. Sacco, D. Hidalgo, E. Celasco and C. F. Pirri, ACS Appl. Mater. Interfaces, 2014, 6, 12153–12167.
54. Y. Wang, J. Chen, P. Wang, L. Chen, Y.-B. Chen and L.-M. Wu, J. Phys. Chem. C, 2009, 113, 16009–16014.
55. S. Khanchandani, S. Kundu, A. Patra and A. K. Ganguli, J. Phys. Chem. C, 2013, 117, 5558–5567.
56. J. Zhang, L. Zhang, N. Yu, K. Xu, S. Li, H. Wang and J. Liu, RSC Adv., 2015, 5, 75081–75088.
57. A. Priyam, S. Ghosh, S. C. Bhattacharya and A. Saha, J. Colloid Interface Sci., 2009, 331, 195–201.
58. S. Ghosh, A. Priyam, S. C. Bhattacharya and A. Saha, J. Fluoresc., 2009, 19, 723–731.
59. S. Ghosh, M. Ray, M. Das, A. Chakrabarti, A. H. Khan, D. D. Sarma and S. Acharya, Phys. Chem. Chem. Phys., 2014, 16, 5276–5283.
60. M. Han and J. Jia, J. Power Sources, 2016, 329, 23–30.
61. X. Zhang, J. Qin, Y. Xue, P. Yu, B. Zhang, L. Wang and R. Liu, Sci. Rep., 2014, 4, 4596–4604.
62. S. Ghosh, N. A. Kouame, L. Ramos, S. Remita, A. Dazzi, A. Deniset-Besseau, P. Beaunier, F. Goubard, P.-H. Aubert and H. Remita, Nat. Mater., 2015, 14, 505–511.
63. W. Cui, J. Li, F. Dong, Y. Sun, G. Jiang, W. Cen, S. C. Lee and Z. Wu, Environ. Sci. Technol., 2017, 51, 10682–10690.
64. J. Sun, J. Zhang, M. Zhang, M. Antonietti, X. Fu and X. Wang, Nat. Commun., 2012, 3, 1139.
65. S. Luo, F. Chai, L. Zhang, C. Wang, L. Li, X. Liu and Z. Su, J. Mater. Chem., 2012, 22, 4832–4836.
66. F. Ozel, E. Aslan, B. Istanbullu, O. Akay and I. H. Patir, Appl. Catal., B, 2016, 198, 67–73.
67. B. Tian and Y. Wu, Appl. Catal., B, 2017, 204, 33–42.
68. C. Liu, Y. Yang, W. Li, J. Li, Y. Li and Q. Chen, Sci. Rep., 2016, 6, 23451–23463.
69. C. Liu, Y. Yang, W. Li, J. Li, Y. Li and Q. Chen, Int. J. Hydrogen Energy, 2016, 41, 1–9.
70. J. Kundu, S. Khilari and D. Pradhan, ACS Appl. Mater. Interfaces, 2017, 9, 9669–9680.
71. X. Zhang, Y. Liu, G. Zhang, Y. Wang, H. Zhang and F. Huang, ACS Appl. Mater. Interfaces, 2015, 7, 4442–4448.
72. Y.-Q. Zhang, S. Li and B.-P. Zhang, RSC Adv., 2016, 6, 103215–103223.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c7nj03424e

---