Synthesis of BiVO₄–TiO₂–BiVO₄ three-layer composite photocatalyst: effect of layered heterojunction structure on the enhancement of photocatalytic activity

Abstract

BiVO₄–TiO₂–BiVO₄ three-layer heterostructure photocatalyst was successfully synthesized by a simple sol–gel method with a glass substrate. The structural and optical properties of the as-prepared samples were relatively characterized. UV-vis diffuse reflectance spectroscopy indicated that the BiVO₄–TiO₂–BiVO₄ three-layer composite possessed strong visible-light absorption. Compared to pure TiO₂, BiVO₄ and BiVO₄–TiO₂ bilayer composites, the BiVO₄–TiO₂–BiVO₄ three-layer composite photocatalyst exhibited much higher photocatalytic activity in decomposition of methylene blue and rhodamine B under visible light irradiation. The results of photoluminescence spectroscopy and photocurrent measurement indicated that three-layer structure could distinctly improve the separation and transmission of the photogenerated charges, which led to the enhanced activity. Moreover, the active species trapping experiments demonstrated holes (h⁺) and superoxide radicals (˙O²⁻) were major active species in the degradation process. Then, a possible reaction mechanism accounting for the excellent photocatalytic activity was proposed on the basis of the energy band structure of the composites.

---

Introduction

Semiconductor photocatalysis technology has received increasing attention for their potential applications in water splitting, carbon dioxide reduction and environmental remediation.^1–3 For all semiconductor photocatalysts, TiO₂ is the most commonly used due to its efficient photocatalytic performance, high physical and chemical stability, as well as its low cost and nontoxicity.^4,5 However, the wide band gap of TiO₂ (E_g = 3.2 eV) causes its inactivity under visible light, which occupies the major part of solar light. Therefore, in order to overcome the limitation, many studies have focused on the modification of the surface or bulk properties of TiO₂ to improve the photocatalytic activity, such as self-doping or doping with nonmetal atoms C⁶ N⁷ and S;⁸ coupling with noble metals particles Ag,⁹ Pt¹⁰ and Pd;¹¹ combination with narrow band gap semiconductors CdS,¹² WO₃,¹³ AgVO₄ (ref. 14) and CdSe;¹⁵ and dye sensitization.¹⁶ Bismuth vanadate (BiVO₄) has recently become an attractive material as a promising photocatalyst for the splitting reaction of water into hydrogen and oxygen,^17–19 and for the degradation of organic pollutants.^20–22 There are three crystalline phases of BiVO₄: tetragonal zircon structure, tetragonal scheelite structure and monoclinic scheelite structure. Among them, the monoclinic scheelite structure BiVO₄ (m-BiVO₄) with narrow band gap (2.4 eV) has been reported to own excellent activity of organic pollutants degradation under visible light irradiation.^23–25 Nevertheless, the photocatalytic activity of m-BiVO₄ is usually not satisfied because photogenerated electron–hole pairs tend to decay rapidly by recombination. The low charge separation efficiency may be the main drawback restricting the practical application of BiVO₄. To solve the problem, loading noble metals^26,27 and coupling other semiconductors with a heterojunction structure are commonly used as the effective modification strategies. Many studies have focused on the BiVO₄-based composite photocatalysts, such as CuO/BiVO₄,^28–30 Bi₂O₃/BiVO₄,³¹ CeO_x/BiVO₄,³² WO₃/BiVO₄ (ref. 33) and BiVO₄/TiO₂.^34–37 These composites have been demonstrated to achieve higher photocatalytic activity under visible light than pure BiVO₄.

Composite materials with a heterojunction structure have been investigated in photocatalysis and solar energy conversion, because the formation of heterojunction can significantly reduce the recombination and speed up the separation rate of photogenerated charge carriers as well as extend the wavelength range for visible light utilisation efficiently. As for the studies of BiVO₄/TiO₂ heterojunction photocatalysts, TiO₂ can be helpful to the rapid separation rate of electron–hole pairs in BiVO₄. Li et al.³⁴ first reported that BiVO₄/TiO₂ nanoheterostructure exhibited remarkable photocatalytic reactivity for the degradation of gaseous benzene under irradiation of various light sources with different wavelengths. They detected activated species (˙OH) by ESR, which was known to be attributed to the high activity. Tan et al.³⁵ fabricated TiO₂/BiVO₄ composite photocatalysts through the one-step microwave hydrothermal method. They showed that the anatase TiO₂ was dispersed on the surface of BiVO₄ by the EDS analysis and the TiO₂/BiVO₄ composites present much higher photocatalytic performance than pure BiVO₄ in the photodegradation reaction of rhodamine B under UV and simulated sun-light irradiation. In our previous studies, we had compared four different methods for synthesis of BiVO₄/TiO₂ photocatalytic nano-particles and evaluated the photocatalytic activities through the photodegradation of methylene blue under simulated sunlight irradiation.³⁶

However, the heterojunction between two semiconductors is constructed uncontrollably in these synthesis processes, which hinders the enhancement of photocatalytic performance. Because the formation of the effective heterojunction could accelerate the separation of electron–hole pairs, it is essential to rationally design layered heterojunction structures with improved photogenerated electron–hole separation to enhance the photocatalytic activity. As it is known that the photocatalysts with multilayer structure reveal high photocatalytic performance, for constructing multilayer structure could generate more effective heterojunction. The TiO₂/WO₃ (ref. 38) multilayer thin films heterostructure have been investigated and showed high photocatalytic activities under visible light irradiation. R. Dholam et al.³⁹ also demonstrated that Cr-doped-TiO₂ multilayer thin films exhibited higher hydrogen production rate through water-splitting with the enhanced photocurrent, which was attributed to higher absorption of visible light by Cr-doped-TiO₂ and the number of space charge ITO/TiO₂ interfaces in multilayer films. Herein, for the BiVO₄/TiO₂ composites, almost all the studies have been focused on synthesizing the granular structure composites for enhancing their photocatalytic activities, while the layered heterostructure structure of the composite has not yet been investigated extensively.

In this study, BiVO₄–TiO₂–BiVO₄ three-layer structure composites were synthesized by a simple sol–gel method using a glass substrate. BiVO₄ layers on the double surfaces act as a sensitizer to absorb photons and excite electron–hole pairs. Meanwhile, the middle TiO₂ layer can facilitate the separation and transmission of the photogenerated charges in BiVO₄. Therefore, constructing BiVO₄–TiO₂–BiVO₄ three-layer composites may possibly be a positive strategy to obtain a highly efficient visible light photocatalyst. This work creatively fabricated BiVO₄–TiO₂–BiVO₄ three-layer heterojunction photocatalysts possessing high efficiency visible light-driven photocatalytic activity. Their photocatalytic activities were evaluated by the degradation of MB and RhB under visible light irradiation. Furthermore, the possible mechanism of the enhanced photocatalytic activity with the layered structure was discussed based on tapping experiments and calculated band-edge potential positions.

Experimental

Preparation of catalysts

All the reagents were of analytical grade and were used without any further purification. The photocatalysts were synthesized through sol–gel method. Subsequently, 0.4 mmol Bi(NO₃)₃·5H₂O and 1 mmol citric acid were dissolved in a small amount of concentrated nitric acid and 20 mL of Milli-Q water to form a transparent solution. Then 0.4 mmol NH₄VO₃ was dissolved in 20 mL Milli-Q water at 70 °C to form a light yellow solution. These two solutions were mixed together afterwards, and the mixture was stirred for 30 min under ambient air to obtain a stable yellow slurry. And the pH value was adjusted to 4 with ammonia solution, resulting in the BiVO₄ sol. Tetrabutyl titanate (10 mL) was added dropwise to a solution of ethanol (16 mL) and glacial ethanoic acid (4 mL) under vigorous stirring, then an ethanol solution (80%, 20 mL) was added to the tetrabutyl titanate solution under vigorous magnetical stirring at room temperature, resulting in TiO₂ sol.

The BiVO₄ sol was stirred at 80 °C for 6 h to complete the hydrolysis reaction, then the precipitate was filtered and washed with Milli-Q water several times. Finally the precipitate was dried and calcined at 450 °C for 2 h to obtain pure BiVO₄ powder. The TiO₂ sol was dried at 80 °C for several hours and the resultant samples were calcined at 450 °C for 2 h to obtain pure TiO₂ powder.

The BiVO₄/TiO₂ layered structure composites were fabricated by transferring the above coating sol to flat on a glass substrate in an ambient atmosphere. Firstly the BiVO₄ sol was added dropwise to a glass substrate to form the first layer. Then the substrate coated with gel layer was dried and calcined at 450 °C for 2 h to obtain thin film sample. Subsequently the TiO₂ sol was added dropwise to the glass substrate with the thin film and then the substrate was dried and calcined at 450 °C for 2 h to obtain BiVO₄–TiO₂ bilayer composite, denoted as B–T. Finally the BiVO₄ sol was added dropwise to the glass substrate with bilayer composite again. The substrate was also dried and calcined at 450 °C for 2 h to obtain BiVO₄–TiO₂–BiVO₄ three-layer composite, labelled by B–T–B.

All the samples were scratched off from the glass substrate to form powder samples for some catalyst characterization.

Characterization of catalysts

The crystalline phases of pure BiVO₄, pure TiO₂, B–T and B–T–B composites were determined using X-ray diffraction (XRD) (D/MAX-RB, Rigaku, Japan). The diffraction patterns were recorded in the 2θ = 10–90° range with a Cu Kα source (λ = 0.15405) running at 40 kV and 30 mA. The specific surface areas of samples were determined by the Brunauer–Emmett–Teller (BET) method (NOVA 4200e, Quantachrome, USA). The morphologies of samples were examined using an emission scanning microscopy (SEM, S-4800, Hitachi, Japan). The transmission electron microscopy (TEM) and the high-resolution transmission electron microscopy (HRTEM) images were studied with a transmission electron microscope (F-20, FEI, USA) at an accelerating voltage of 200 kv. The UV-vis diffuse reflectance spectra (DRS) of the samples were recorded using a UV-vis spectrophotometer (T9s, Persee, China) equipped with an integrating sphere at room temperature. BaSO₄ was used as the reference. The photoluminescence spectra (PL) were recorded using a fluorescence spectrophotometer (F-4500, Hitachi, Japan) with a Xe lamp as the excitation light source.

Photocatalytic properties

The photocatalytic activities of samples were evaluated by degrading methylene blue (MB) and rhodamine B (RhB) under visible-light irradiation at room temperature using a 400 W Xe lamp with a 420 nm cut off filter as the light source. Experiments were performed at ambient temperature as follows: 40 mg powder photocatalyst was dispersed in 40 mL of MB solution (10 mg L⁻¹) and 40 mL of RhB solution (10 mg L⁻¹), respectively. Before illumination, the suspensions were stirred for 1 h to achieve an adsorption–desorption equilibrium between the photocatalysts and MB or RhB. During the irradiation, the reaction samples were collected at 30 min (60 min for RhB degradation) intervals and centrifuged to remove the photocatalyst particles. Finally, the centrifuged solution were recorded using a UV-vis spectrophotometer (T9s, Persee, China) at the maximum absorption wavelength of 664 nm for MB and 553 nm for RhB.

Photoelectrochemical measurements

The measurement of photocurrent was carried out on an electrochemical workstation (5060F, RST, China) in a standard three-electrode system with the samples, an Ag/AgCl electrode (saturated KCl), a Pt wire used as the working electrode, the reference electrode, and the counter electrode, respectively. 0.5 M Na₂SO₄ aqueous solution was introduced as electrolyte. A 100 W incandescent lamp with a 420 nm cut off filter was used as the light source. The preparation of the working electrode as follows: 5 mg samples were dispersed in a certain amount of ethanol and Nafion solution, followed by spreading on the bottom middle of an ITO glass in a circle with a diameter of 6 mm. Then the photocurrents of the photocatalysts with the light on and off were measured at 0.8 V.

Results and discussion

XRD analysis

The crystallographic structure of as-prepared BiVO₄, TiO₂, B–T bilayer and B–T–B three-layer composites were investigated by XRD analysis. Fig. 1 shows the XRD patterns of B–T and B–T–B composites. For comparison, the XRD patterns for pure BiVO₄ and TiO₂ are also given. It can be seen that the prepared BiVO₄ powder was in good agreement with the monoclinic phase of BiVO₄ (JCPDS card no. 14-0688), and the distinctive peaks at 2 theta of 18.67°, 28.82°, 34.49° and 59.26° were well-matched with the lattice planes of (1 1 0), (−1 2 1), (2 0 0) and (1 2 3), respectively. And diffraction peaks of TiO₂ can be well-indexed to anatase phase of TiO₂ (JCPDS card no. 21-1272). From Fig. 1, the main diffraction peaks of B–T and B–T–B composites were similar to those of pure BiVO₄ and they can be well-indexed to monoclinic BiVO₄. Moreover, the two small peaks at 2 theta of 25.28° and 37.80° in XRD patterns of B–T and B–T–B composites could be indexed to the characteristic peaks (1 0 1) and (0 0 4) of anatase TiO₂, indicating the existence of TiO₂.

Fig. 1 XRD patterns for the BiVO₄, TiO₂, B–T and B–T–B composites.

Morphology measurement

The morphology of the pure BiVO₄, B–T bilayer and B–T–B three-layer composite were observed by SEM, as shown in Fig. 2. It can be seen from Fig. 2a that the morphology of pure BiVO₄ presented rod-like nanoparticles which are no more than 100 nm wide. And B–T composite was irregular block and small particles due to the addition of TiO₂ in Fig. 2b, which could result in larger surface area (Table 1). Fig. 2c shows the morphology of B–T–B composite on the substrate. It had an obvious layered structure with a smooth surface because the outermost layer was BiVO₄. Moreover, the BET surface area of samples were determined by the nitrogen sorption tests, and the corresponding values were calculated to be 10.04, 51.10 and 16.55 m² g⁻¹ for pure BiVO₄, B–T composite and B–T–B composite, respectively as listed in Table 1.

Fig. 2 SEM images of the as-prepared samples: (a) pure BiVO₄, (b) B–T bilayer composite and (c) B–T–B three-layer composite on the substrate.

Table 1 Surface area of pure BiVO₄, B–T bilayer composite and B–T–B three-layer composite

Photocatalysts	BiVO4	B–T	B–T–B
Surface area (m^2 g^−1)	10.04	51.10	16.55

---

To reveal the microstructure and the morphology of B–T and B–T–B composites, the TEM text was carried out, which is an efficient and widely used characterization in terms of heterojunction.^32,40 Fig. 3 shows the TEM and HRTEM images of the B–T and B–T–B composite. The TEM images (Fig. 3a and b) also presented a layered structure, which could form more effective heterojunction structure between BiVO₄ and TiO₂. As shown in Fig. 3c, the interplanar spacing of distinct lattice fringes is about 0.310 nm, which corresponded to the (1 2 1) lattice plane of the monoclinic BiVO₄. Meanwhile, the uniform fringes with an interval of 0.352 nm corresponded to the (1 0 1) lattice plane of anatase TiO₂. Thus, these results confirmed that the B–T composite was formed by TiO₂ covering in BiVO₄. From Fig. 3d, three different kinds lattice fringes were clearly exhibited, one of d = 0.352 nm matched the (1 0 1) lattice plane of anatase TiO₂, the other two d = 0.254 and 0.310 nm matched the (0 0 2) and (1 2 1) lattice planes of the monoclinic BiVO₄, respectively. Moreover explicit interfaces between BiVO₄ and TiO₂ were observed, which suggested that the B–T–B composite indeed possessed three-layer heterojunction structure. Therefore, the heterojunction nanostructure was formed for layered structure and the each layer was BiVO₄, TiO₂ and BiVO₄, respectively.

Fig. 3 TEM images of as-prepared samples (a) B–T and (b) B–T–B, HRTEM images of (c) B–T and (d) B–T–B.

Optical properties

The optical absorption properties of a semiconductor, which are related to its electronic structure, are recognized as the key factor in determining its photocatalytic activity.³⁶ The optical properties of as-prepared BiVO₄, TiO₂, B–T and B–T–B composites were investigated using UV-vis diffuse reflectance spectroscopy. Fig. 4 shows the UV-vis diffuse reflectance absorption spectra of as-prepared samples. As illustrated, pure TiO₂ only presented an absorbance in the UV-light region, with an absorption edge of 400 nm. While other samples exhibited a strong absorption in the UV and visible-light regions. The layered structure composites caused a red shift in the BiVO₄ absorption edge and the absorption intensities of layered structure composites were remarkably higher than that of pure TiO₂ and BiVO₄. The efficient visible light absorption abilities of the B–T–B composites ensured that they could generate sufficient electron–hole pairs under visible light irradiation. The optical band gap for the semiconductor photocatalysts was estimated using the following equation:

(Ahv) = α(hv − E_g)^n/2

where A is the absorption coefficient near the absorption edge; h is the Plank constant with the unit of eV; α is a constant; E_g is the absorption band gap energy; and n is 1 and 4 for a direct and indirect band gap semiconductor, respectively.^41,42 BiVO₄ have a direct band gap, and n is 1 herein. The inset in Fig. 4 shows the (Ahv)² versus the photon energy (hv) plots of the BiVO₄, TiO₂, B–T and B–T–B composites. The band gaps were estimated to be 3.12, 2.51, 2.42 and 2.35 eV, corresponding to the pure TiO₂, pure BiVO₄, B–T composite and B–T–B composite, respectively. Compared with pure TiO₂ and BiVO₄, the band gap of the B–T–B sample was obviously narrowed. The results demonstrated that B–T–B three-layer heterostructure possessed a much greater optical absorption region than that of pure TiO₂ and BiVO₄, which can be excited by visible light illumination.

Fig. 4 UV-vis DRS spectra of the as-prepared BiVO₄, TiO₂, B–T and B–T–B composites; the inset shows the band gap energies of the corresponding photocatalysts.

Photocatalytic activities

The photocatalytic performance of pure BiVO₄, TiO₂, B–T and B–T–B composites were evaluated by decomposing methylene blue (MB) under the visible light irradiation (λ ≧ 420 nm), as shown in Fig. 5. Before illumination, all the samples were dispersed in MB solution followed by stirring for 60 min in the dark in order to reach an adsorption–desorption equilibrium. The pure TiO₂ and BiVO₄ samples exhibited a nethermore photodegradation ratio of 21% and 35% after 180 min under visible light irradiation, whereas the B–T and B–T–B composites obviously presented a higher photocatalytic performance. The degradation ratio of MB using B–T–B three-layer composite was over 95% after 180 min light irradiation, which was the best photocatalytic activity. From the DRS, pure TiO₂ had no absorbance in the visible light region resulting in the negligible photocatalytic activity, while the layered structure of BiVO₄ and TiO₂ showed the enhanced photocatalytic performance with the light absorption to longer wavelengths. Fig. 5 also shows the adsorption ratio of MB in the dark for all samples. It can be observed that the adsorption in dark had an obvious effect on the photocatalytic activities of these photocatalysts. Therefore, the larger adsorption capacity of photocatalysts the better its photocatalytic performance was.

Fig. 5 Comparison of the adsorption in the dark and the degradation of MB under visible light irradiation using BiVO₄, TiO₂, B–T and B–T–B composites.

The kinetics of degradation of MB under visible light irradiation were also investigated. A pseudo-first-order kinetic mode was used to fit the degradation data using ln(C₀/C) = kt + b, where k is the photocatalytic reaction rate constant, C₀ is the concentration of MB after the adsorption–desorption equilibrium before illumination, and C is the concentration of MB at different illumination time. Fig. 6 illustrates the apparent kinetic rate constants (k) of the MB decomposition over the as-prepared samples, and the inset shows the plot of ln(C₀/C) versus reaction time (t). The rate constants were calculated to be 0.101 × 10⁻², 0.196 × 10⁻², 0.405 × 10⁻² and 1.396 × 10⁻² min⁻¹ for pure TiO₂, BiVO₄, B–T and B–T–B, respectively. The results revealed that the B–T–B composite displayed the largest value of k among the samples, which was 7 times compared to that of pure BiVO₄. These findings have provided further evidences that the formation of three-layer heterostructure for TiO₂ and BiVO₄ was much more energetically favorable to enhance the photocatalytic performance of the samples. Furthermore, the three-layer structure nanocrystalline presented the best photodegradation performance.

Fig. 6 Kinetic constants of the MB decomposition over different samples.

Moreover, stability is significant for photocatalysts in practical applications. To check the stability of the best photocatalyst (B–T–B), cycling experiments in the photocatalytic degradation of MB under visible light irradiation were evaluated. There was no obvious catalyst deactivation after four successive cycles as shown in Fig. 7, confirming that B–T–B was not photocorroded and held excellent stability during the degradation.

Fig. 7 Cycling runs for the photodegradation of MB over the B–T–B three-layer composite under visible light irradiation.

In addition to MB, a typical organic contaminant, RhB, was also chosen to further estimate the photocatalytic performance of samples. Fig. 8 presents the photocatalytic activity of the as-prepared samples which were evaluated by decomposing RhB under visible light irradiation. It can be observed that pure TiO₂ and pure BiVO₄ only showed 20% and 40% degradation ratios of RhB after 5 hours. However, the degradation ratios of B–T and B–T–B were increased to 81% and 91%, respectively. The B–T–B showed a much higher photocatalytic activity for RhB degradation which was about 2.3 times higher compared to that of pure BiVO₄. The finding further confirmed the enhanced photocatalytic activity of the three-layer heterojunction photocatalyst.

Fig. 8 Comparison of the adsorption in the dark and the degradation of RhB under visible light irradiation using BiVO₄, TiO₂, B–T and B–T–B composites.

Photocatalytic mechanism

The photocatalytic activity of photocatalyst depends on various factors, including the crystal structure, crystallinity, specific surface area, morphology, light energy utilization ratio, and quantum efficiency, etc.⁴³ As discussed above, the enhancement of photocatalytic activity of the B–T–B three-layer samples may partially come from the enhanced light absorption, increased BET surface area and improved charge separation. The B–T–B samples with lower BET surface areas than B–T samples exhibits the best performance. It is clear that the surface area is not a decisive factor for the photocatalytic activities in this study. In addition, the effective separation of the photogenerated electron–hole pairs was confirmed by PL analysis. Photoluminescence spectra of the photocatalysts are used to illustrate the efficiency of photogenerated carriers trapping, migration and recombination, because PL intensity mainly results from the recombination of excited electrons and holes.^44,45 Generally, the lower PL intensity often implies the lower recombination of electron–hole pairs under irradiation.^45,46 The PL emission spectra of pure BiVO₄, TiO₂, B–T and B–T–B composites with an excitation wavelength of 300 nm are shown in Fig. 9. The peaks with the largest intensity attributed to the pure BiVO₄ and TiO₂ appeared at around 540 nm and 470 nm, respectively. The B–T and B–T–B composites both had the two peaks. It can be observed that the PL intensity of the B–T and B–T–B composites were apparently weaker than that of pure BiVO₄ and especially of pure TiO₂. This finding implied that the recombination of the photogenerated charge carriers was greatly suppressed after the formation of the layered heterojunction structure between BiVO₄ and TiO₂, which resulted in the high photocatalytic activity of BiVO₄/TiO₂ layered structure composites.

Fig. 9 Photoluminescence (PL) spectra of pure BiVO₄, TiO₂, B–T and B–T–B composites.

The photocurrent responses of the samples in an electrolyte under visible light irradiation directly correlate with the generation and transfer of the photogenerated charge carriers in the photocatalytic process, which can further confirm the separation of electron–hole pairs.⁴² Fig. 10 compares the photocurrent response of pure BiVO₄, TiO₂, B–T and B–T–B composites with the light on and off. Distinctly, the current abruptly increased and decreased when the light source was switched on and off. As shown in Fig. 10, the photocurrent response of B–T–B and pure BiVO₄ were much higher than that of B–T and pure TiO₂. The B–T–B sample exhibited the most enhanced photocurrent response among the samples. This result indicated that more efficient separation of photogenerated electron–hole pairs and faster transfer of photoinduced charge carriers occurred in the B–T–B composite with the layered heterojunction structure. Meanwhile, the photocurrent responses of all the samples were broadly in line with their photocatalytic performance.

Fig. 10 Photocurrent responses of pure BiVO₄, TiO₂, B–T and B–T–B composites under visible light irradiation.

To clarify the reasons of the high photocatalytic activity, it is necessary to determine the main active species^47–49 in the process of the degradation of MB. Therefore, we performed the trapping experiments with different scavengers. In this way, 0.2 mM benzoquinone (BQ, a quencher of ˙O₂⁻),^47–49 10 mM sodium oxalate (Na₂C₂O₄, a quencher of h⁺)^49,50 and 10 mM isopropanol alcohol (IPA, a quencher of ˙OH radical)^49,51 were adopted. Fig. 11 shows the variation of MB degradation with different scavenger added over the B–T–B. The degradation efficiency changed slightly when the scavenger IPA was added, which suggested that ˙OH radical was not the dominant active species in the photocatalytic reaction system of B–T–B three-layer composites. Meanwhile, there was an obvious decline in the photodegradation rate of MB after addition of Na₂C₂O₄, indicating that h⁺ was an important active species in the photocatalytic degradation. Moreover, the degradation efficiency of MB was almost completely depressed with the addition of the scavenger BQ. Considering the above results, h⁺ and ˙O₂⁻ were major active species in the degradation process, while ˙OH had little impact on the MB photodegradation process.

Fig. 11 Effects of different scavengers on the photocatalytic degradation of MB over the B–T–B three-layer composite under visible light irradiation.

Based on the above data, a possible mechanism for the MB photodegradation over B–T–B three-layer heterojunction photocatalyst under visible light irradiation can be proposed. Here BiVO₄ served as a sensitizer for visible-light-induced redox process while TiO₂ was a substrate. The conduction band (CB) and valence band (VB) potentials of BiVO₄ and TiO₂ at the point of zero charge can be calculated by the following empirical equation:^52,53

E_VB = χ − E_e + 0.5E_g (1)

E_CB = E_VB − E_g (2)

where χ is the electronegativity of the semiconductor, which is the geometric mean of the electronegativity of the constituent atoms, E_e is the energy of free electrons on the hydrogen scale (∼4.5 eV), and E_g is the band gap energy of the semiconductor. The χ values for BiVO₄ and TiO₂ are ca. 6.04 eV (ref. 54) and 5.90 eV,⁵³ respectively. The band gap energies of BiVO₄ and TiO₂ were adopted as 2.51 eV and 3.12 eV (inset in Fig. 4), respectively. The calculation result showed that the bottom of the conduction band of BiVO₄ was 0.28 eV, while the top of the valence band was around 2.79 eV. The energy of the conduction band and valence band of TiO₂ were −0.16 and 2.96 eV, respectively.

Fig. 12 presents the transfer behavior of photogenerated electron–hole in the B–T–B heterojunction composites. BiVO₄ was considered as an intrinsic semiconductor, so the Fermi level in BiVO₄ laid in the middle of the conduction band and the valence band which located at around 1.54 eV.³² Generally speaking, the Fermi level of n-type semiconductor laid below the bottom of the conduction band by ca. 0.1–0.2 eV,⁵⁵ so the Fermi level in TiO₂ was −0.06 eV. The heterojunction was formed when these two types of semiconductor materials were joined together. The two semiconductors had a uniform Fermi level after the system was in equilibrium. In our work, the Fermi level turned to be around 1.54 eV which was the same to the Fermi level of BiVO₄. Consequently the E_CB and E_VB of TiO₂ decreased from −0.16 to 1.44 eV, 2.96 to 4.56 eV respectively. The BiVO₄ layer in the heterojunction photocatalysts acted as a sensitizer to absorb photons as well as excited electron–hole pairs under visible light irradiation. For B–T–B three-layer structure composites, BiVO₄ layers on the double surfaces were more advantageous to absorb the photons, so as to improve the photocatalytic efficiency. The photoinduced electrons on the BiVO₄ layers surface can easily transfer to TiO₂ conduction band via heterojunction interfaces, leaving the holes on the BiVO₄ valence band. The conduction electrons were good reductants which could capture the absorbed O₂ on the surface of the catalyst and reduce it to ˙O₂⁻. Moreover, the generated ˙O₂⁻ was assigned to the degradation of MB. At the same time, the photogenerated holes left on the valence band of BiVO₄ could oxidize organic pollutant directly. The trapping and transferring process of electrons enhanced the efficiency of separation of electron–hole pairs, which contributed to organics degradation by the active holes (h⁺) and superoxide radical anion (˙O₂⁻) species. As a summary, the results of PL spectra, photocurrent measurement and trapping experiments were well in agreement with the supposed photocatalytic mechanism.

Fig. 12 Schematic illustration of mechanism of the B–T–B three-layer photocatalyst under visible light irradiation.

Conclusions

In summary, BiVO₄–TiO₂–BiVO₄ three-layer heterostructure photocatalysts has been successfully synthesized by a sol–gel method using a glass substrate. The structural studies revealed that B–T–B composite was composed of monoclinic BiVO₄ and anatase TiO₂, and the composite showed obvious layered structure. The B–T and B–T–B composites exhibited higher photocatalytic activity for the degradation of MB and RhB than that of pure BiVO₄ and TiO₂ under visible light illumination, and the three-layer composite presented the best photocatalytic activity. What's more, the results of PL and photocurrent indicated that the formation of layered structure could generate more efficient heterojunction, and further induced the effective separation and transmission of the photogenerated electron–hole pairs. We demonstrated that the rationally designed multilayer structure composites could effectually enhance the photocatalytic performance.

Acknowledgements

We gratefully acknowledge the financial support provided by the Project of the National Natural Science Foundation of China (Grant No. 21271022).

References

1. Z. G. Xiong and X. S. Zhao, J. Am. Chem. Soc., 2012, 134, 5754–5757.
2. S. Sun, W. Wang, L. Zhou and H. Xu, Ind. Eng. Chem. Res., 2009, 48, 1735–1739.
3. A. D. Paola, E. García-López, G. Marcì and L. Palmisano, J. Hazard. Mater., 2012, 211, 3–29.
4. T. Hirakawa, K. Sato and A. Komano, J. Phys. Chem. C, 2010, 114, 2305–2314.
5. Y. Liu, J. Ma, Z. Liu and C. Dai, Ceram. Int., 2010, 36, 2073–2077.
6. C. Lettmann, K. Hildenbrand and H. Kisch, Appl. Catal., B, 2001, 32, 215–227.
7. J. Wang, D. N. Tafen and J. P. Lewis, J. Am. Chem. Soc., 2009, 131, 12290–12297.
8. T. Ohno, M. Akiyoshi and T. Umebayashi, Appl. Catal., A, 2004, 265, 115–121.
9. K. Naoi, Y. Ohko and T. Tatsuma, J. Am. Chem. Soc., 2004, 126, 3664–3668.
10. L. Yang, W. Yang and Q. Cai, J. Phys. Chem. C, 2007, 111, 16613–16617.
11. A. Honciuc, M. Laurin and S. Albu, J. Phys. Chem. C, 2010, 114, 20146–20154.
12. G. S. Li, D. Q. Zhang and J. C. Yu, Environ. Sci. Technol., 2009, 43, 7079–7085.
13. C. W. Lai and S. Sreekantan, Int. J. Hydrogen Energy, 2013, 38, 2156–2166.
14. J. Wang, H. Ruan and W. Li, J. Phys. Chem. C, 2012, 116, 13935–13943.
15. W. Ho and C. Y. Jimmy, J. Mol. Catal. A: Chem., 2006, 247, 268–274.
16. D. Achey, S. Ardo and H. L. Xia, J. Phys. Chem. Lett., 2011, 2, 305–308.
17. J. A. Seabold and K. S. Choi, J. Am. Chem. Soc., 2012, 134, 2186–2192.
18. A. Kudo, Int. J. Hydrogen Energy, 2006, 31, 197–202.
19. A. Kudo, R. Niishiro and A. Iwase, Chem. Phys., 2007, 339, 104–110.
20. D. K. Zhong, S. Choi and D. R. Gamelin, J. Am. Chem. Soc., 2011, 133, 18370–18377.
21. A. Zhang, J. Zhang and N. Cui, J. Mol. Catal. A: Chem., 2009, 304, 28–32.
22. W. Liu, Y. Yu and L. Cao, J. Hazard. Mater., 2010, 181, 1102–1108.
23. H. Fan, T. Jiang and H. Li, J. Phys. Chem. C, 2012, 116, 2425–2430.
24. G. Li, D. Zhang and J. C. Yu, Chem. Mater., 2008, 20, 3983–3992.
25. Y. Zhang, G. Li and X. Yang, J. Alloys Compd., 2013, 551, 544–550.
26. L. Ge, J. Mol. Catal. A: Chem., 2008, 282, 62–66.
27. A. Zhang and J. Zhang, Appl. Surf. Sci., 2010, 256, 3224–3227.
28. H. Xu, H. Li and C. Wu, J. Hazard. Mater., 2008, 153, 877–884.
29. H. Jiang, H. Endo and H. Natori, Mater. Res. Bull., 2009, 44, 700–706.
30. C. Yu, K. Yang and C. Y. Jimmy, J. Alloys Compd., 2011, 509, 4547–4552.
31. L. Li and B. Yan, J. Alloys Compd., 2009, 476, 624–628.
32. S. Gu, W. Li and F. Wang, Appl. Catal., B, 2015, 170, 186–194.
33. P. Chatchai, Y. Murakami and S. Kishioka, Electrochim. Acta, 2009, 54, 1147–1152.
34. Y. Hu, D. Li and Y. Zheng, Appl. Catal., B, 2011, 104, 30–36.
35. L. Zhang, G. Tan and S. Wei, Ceram. Int., 2013, 39, 8597–8604.
36. S. Wang, W. Li and F. Chen, Aust. J. Chem., 2015, 68, 1268–1275.
37. M. Xie, X. Fu and L. Jing, Adv. Energy Mater., 2014, 4, 1300995.
38. H. Shinguu, M. M. H. Bhuiyan and T. Ikegami, Thin Solid Films, 2006, 506, 111–114.
39. R. Dholam, N. Patel and A. Santini, Int. J. Hydrogen Energy, 2010, 35, 9581–9590.
40. J. Su, X. Zou and G. Li, J. Phys. Chem. C, 2011, 115, 8064–8071.
41. X. Zhang, L. Zhang and T. Xie, J. Phys. Chem. C, 2009, 113, 7371–7378.
42. F. Wang, W. Li and S. Gu, RSC Adv., 2015, 5, 89940–89950.
43. Y. Xu, C. Liu and M. Chen, Int. J. Nanopart., 2011, 4, 268–283.
44. M. Long, W. Cai and J. Cai, J. Phys. Chem. B, 2006, 110, 20211–20216.
45. F. Duan, Y. Zheng and M. Chen, Appl. Surf. Sci., 2011, 257, 1972–1978.
46. H. Li, W. Li and S. Gu, Catal. Sci. Technol., 2016, 6, 3510–3519.
47. H. Lin, H. Ye and S. Chen, RSC Adv., 2014, 4, 10968–10974.
48. J. Cao, B. Luo and H. Lin, J. Hazard. Mater., 2012, 217, 107–115.
49. J. Ke, C. Niu and J. Zhang, J. Mol. Catal. A: Chem., 2014, 395, 276–282.
50. C. Zhang, L. Ai and L. Li, J. Alloys Compd., 2014, 582, 576–582.
51. P. Ju, P. Wang and B. Li, Chem. Eng. J., 2014, 236, 430–437.
52. T. Fan, C. Chen and Z. Tang, RSC Adv., 2016, 6, 9994–10000.
53. G. Dai, S. Liu and Y. Liang, Appl. Surf. Sci., 2013, 264, 157–161.
54. M. Guan, D. Ma and S. Hu, Inorg. Chem., 2010, 50, 800–805.
55. S. R. Morrison, Electrochemistry at Semiconductor and Oxidized Metal Electrodes, Plenum Press, New York, 1980.

---