Introduction

With the development of industry, water pollution caused by antibiotics, heavy metals, and industrial dyes has become more and more serious [1,2,3,4]. At present, there are various technologies and strategies to selectively remove or purify pollutants, including adsorption [5], advanced oxidation [6], microwave catalysis [7] and photocatalysis [8, 9].Photocatalysis is an environmentally friendly and efficient method due to its mild conditions, high catalytic efficiency and simple experimental operation [10,11,12,13]. However, the limited solar energy absorption range and low quantum efficiency of photocatalyst, hinder its wider application in actual wastewater treatment [14,15,16,17]. Therefore, it is necessary to further explore more effective and higher catalytic activity of visible light reaction semiconductor photocatalytic materials.

BiOBr is an indirect semiconductor with special layered structure and a high degree of anisotropy, which exhibits great potential applications in photocatalytic degradation, carbon dioxide reduction, nitrogen fixation and hydrogen evolution [18,19,20,21,22]. BiOBr exhibits excellent photoelectric properties due to its special layered structure that can form an internal electric field. The BiOBr has a band gap of about 2.6 eV and two unique valence bands that show different oxidation capacities: one responds to visible light and the other to ultraviolet light [23, 24]. However, the conversion efficiency of BiOBr to solar energy is still limited by several shortcomings, such as inappropriate band structure, large band gap and unsatisfactory particle size, etc. [25, 26]. Various approaches have been adopted to address these limitations, including crystal face exposure, loading of precious metal particles, construction of heterojunctions, and composite modification. Sun et al. [27] synthesized BiOBr catalysts with different microstructure by adjusting solvothermal conditions. The result showed that the exposed BiOBr (110) crystal plane can obtain more oxygen defects. The BiOBr microflowers exhibited good photocatalytic degradation of gaseous o-dichlorobenzene under visible light. Jia et al. [28] prepared BiOBr/BiOAc1−xBrx S-scheme heterojunction photocatalyst. When the molar ratio of Br/Bi was 0.8, it showed good visible light catalysis for the degradation of tetracycline (TC) and Rhodamine B (RhB), with degradation efficiency of 99.2% and 99.4%. Wang et al. [29] synthesized a size-matched Sb2WO6/BiOBr photocatalyst using a precipitation-deposition method for the removal of NO.

Metal–organic frameworks (MOFs) have been widely used in various technical and scientific fields due to their controllable pore capacity and large specific surface area. Some representative MOFs (such as ZIF-8 [30], MOF-5 [31], MIL-125 [32], MIL-53 [33], MIL-101 [34] and UiO-66 [35]) have potential application value. MOF-5 has a cubic topology, which is formed by organic ligands and metal zinc ions in a specific conjugate structure [36,37,38]. This structure can promote charge transport and help reduce photo-generated electron–hole recombination [31, 39]. However, MOF-5 is easily hydrolyzed when the water content reaches 4% [40], and a large number of studies focused on how to explore a waterproof framework while ignoring the potential application of hydrolyzed MOF-5. Among them, Jeong et al. found that the effect of hydrolyzed MOF-5 can be eliminated in the heat treatment process [41].

As a widely used nitrogenous dye pollutant, RhB belongs to three types of carcinogens. Conventional wastewater treatment methods such as biological oxidation cannot effectively remove aromatic amines in wastewater, and may even produce potentially carcinogenic aromatic amines [42]. Therefore, this article uses the hydrothermal method to load MOF-5 on the BiOBr surface to form a scheme II heterojunction MOF-5/BiOBr system. The photocatalytic performance of MOF-5/BiOBr was evaluated by dissociating RhB in simulated solar degradation. A possible scheme II photocatalytic degradation mechanism was proposed by XPS and UV–Vis diffuse reflectance spectroscopy, free radical capture experiment, and Mott–Schottky results to explore the photocatalytic degradation mechanism.

Experimental

Reagent

Cetyltrimethylammonium bromide (CTAB), terephthalic acid (C8H6O4), and absolute ethanol were purchased from Sinopharm Chemical Reagent Co., Ltd. Rhodamine B (RhB) and bismuth nitrate pentahydrate (Bi(NO3)3⋅5H2O) were purchased from Aladdin Industries. Zinc acetate dihydrate (Zn(CH3CO2)2⋅2H2O) was purchased from Tianjin Guangxia Technology Development Co., Ltd. All reagents are analytically pure and no further purification is required. Deionized water was used throughout the experiment.

Preparation of MOF-5

Terephthalic acid was dissolved in 60 mL of DMF, followed by Zn(CH3CO2)2⋅2H2O added to the solution and stirred well [43]. The mixture was transferred into a 100 mL Teflon-lined autoclave and kept at 100 ℃ for 12 h. After the Teflon-lined autoclave was cooled to room temperature, the collected products were washed three times with DMF, dried at 60 °C for 6 h, and samples were collected.

Preparation of MOF-5/BiOBr

Weighed 0.7282 g CTAB was dissolved in 20 mL ethanol as solution a. 0.9701 g Bi(NO3)3⋅5H2O and a predetermined amount of MOF-5 were dissolved in 20 mL ethanol as solution b. Under vigorous stirring, add solution a to solution b for 1 h to form a homogeneous solution. The mixture was transferred to a 100 mL Teflon-lined autoclave and kept at 160 °C for 16 h. The product was washed three times with ethanol and water. The samples were collected by drying at 60 °C for several hours. MOF-5/BiOBr composite material represents X%MB (X% represents the mass ratio of MOF-5 to BiOBr, X is 10, 15, 20, 25, 30).

The method of synthesis of pure BiOBr is the same as that of MOF-5/BiOBr except that MOF-5 is not added.

Characterization

Scanning electron microscope (SEM, Zeiss Sigma300) and field emission transmission electron microscope (FE-TEM, Tecnai G2 F20) were used to test the microstructure and morphology of the samples. Using the Brucker D8 phase X-ray diffractometer (XRD) to measure the crystal structure of the sample. A Thermo Scientific K-Alpha X-ray Photoelectron Spectrometer (XPS) was used to obtain the element composition and valence state of the sample. UV–Vis diffusion spectroscopy (UV-2600 Lambda750 PE) was used to test the light absorption characteristics of the prepared samples at a wavelength of 200–800 nm. The specific surface area and porosity analyzer of Mike's model 2020 HD88 was used to test the samples.

Photocatalytic performance test

In order to evaluate the photocatalytic activity of synthetic materials, the degradation efficiency of RhB was tested. A sample of 10 mg was weighed and added into 100 mL of RhB solution with a concentration of 10 ppm, sonicate for 5 min, and magnetically stir the suspension in a dark environment for 1 h to establish an adsorption–desorption equilibrium. Then, under simulated sunlight, 4 mL of the suspension was taken every 3 min. The 4 mL suspension was centrifuged at a high speed to precipitate and separate the photocatalyst, and the supernatant was obtained. The absorbance of the supernatant was measured at 554 nm using a 721 visible spectrophotometer. The photocatalytic degradation of each sample follows the non-linear least squares fitting equation:

$$ {\text{A}}\; = \;{\text{X}}*{\text{ecp}}\left( { - {\text{k}}_{{{\text{app}}}} *{\text{t}}} \right)\; + \;{\text{E}}, $$
(1)

A is the amplitude of the process, X is the concentration of reactant after degradation of RhB, E is the end point, and kapp is the degradation rate constant.

Electrochemical test

The electrochemical impedance spectra and Mott–Schottky curves of the materials were measured using a three-electrode cell system at the Electrochemical Workstation (Model CHI660C) in Chenhua, Shanghai. BiOBr and MOF-5 (effective area 1 cm2) prepared on FTO conductive glass are used as working electrodes. The electrolyte is 0.5 mol/L sodium sulfate solution.

Results and discussion

The crystal structure of the sample was characterized by X-ray diffraction technique. As shown in Fig. 1, the main diffraction peaks of pure BiOBr are located at 2θ = 10.9°, 25.2°, 32.2°, 46.2°, and 56.7°, which are consistent with the standard spectrum of tetragonal BiOBr (JCPDS No. 09-0393) and correspond to the crystal planes (001), (101), (110), (200) and (212). The characteristic peaks of sample MOF-5 at 2θ = 6.77°, 9.70°, 13.71°, and 15.56° correspond to (200), (220), (400) and (420) crystal planes, which are consistent with previous literature reports [43]. Compared with that of pure BiOBr, the diffraction peak of MOF-5/BiOBr composite material, with the increase of MOF-5, the peak intensity of 2θ about 12.5° becomes more and more obvious. The results showed that the MOF-5 and BiOBr composite materials were successfully prepared.

Fig. 1
figure 1

X-ray diffractometer pattern of MOF-5, 10%MB, 15%MB, 20%MB, 25%MB, 30%MB and BiOBr(XRD condition: Bruker D8 Advance, 5°–90°, 10°/min)

Full size image

The morphology of the sample was characterized by scanning electron microscope. As shown in Fig. S1a, the width of the MOF-5 cubic crystal is about 2 μm. As shown in Fig. S1b, BiOBr is self-assembled from nanosheets into microspheres of approximately 3 μm. Fig. S1c is the SEM image of the 20% MB composite. It can be seen from the figure that the structure of MOF-5 has undergone significant changes, and the cubic crystals have been transformed into needle-shaped nanosheets [36,37,38], which are loaded on the surface of BiOBr. In addition, energy dispersive X-ray (EDX) mapping was performed on the element distribution of the 20% MB sample. The distribution diagrams of Bi, O, Br, and Zn elements are shown in the Fig. S1. SEM shows that the composite has been successfully prepared, which is consistent with the XRD results.

TEM and high resolution TEM (HRTEM) are used to further prove that MOF-5 is supported on BiOBr. As shown in the figure, the TEM image in Fig. S2 is 20%MB, which shows that it is composed of sheet structure. Compared with the SEM image, the sheet is more dispersed, which is due to the selection of more broken areas. Fig. S2b depicts an HRTEM image of 20% MB, a lattice fringe spacing of 0.277 nm, corresponding to the crystal plane of the tetragonal BiOBr (110), confirming the XRD results. Since MOF-5 has no lattice stripes, the blank space is MOF-5. MOF-5 was grown on BiOBr nanosheets, indicating that there is an interface contact between the two semiconductors, which is conducive to the separation of photogenerated carriers.

The valence state, surface composition and chemical state of the sample were analyzed by XPS spectroscopy. As shown in Fig. 2, the O 1s of BiOBr is on two peaks around 532.17 eV and 530.08 eV, which are attributed to the BiOBr and the oxygen species adsorbed on the Bi-O bond surface of the sample, such as hydroxyl or H2O. With the increase of MOF-5 loading, the binding energy decreased, and the lattice oxygen peak of 20% MB shifted to 529.99 eV. The two peaks at 69.65 eV (Br 3d3/2) and 68.47 eV (Br 3d5/2) are assigned to the spin–orbit splitting of Br 3d, indicating the presence of Br− anions (Fig. 2b). In the XPS spectrum of Bi 4f (Fig. 2c), the observed binding energy peaks are around 164.59 eV (Bi 4f5/2) and 159.29 eV (Bi 4f7/2), confirming that the bismuth species in the product is Bi3+. On Fig. 2d, Zn 2p3/2 was observed at 1021.91 eV and 1044.93 eV, indicating the presence of BDC ligand complexes, which are consistent with the framework group of zinc carboxylate. Compared with BiOBr, the binding energies of O 1s, Bi 4f and Br 4d at 20% MB peak position decreased. The negative shift of binding energy shows an increase in BiOBr electron density, indicating that the surface binding interaction between MOF-5 and BiOBr and the charge transfer from the surface of MOF-5 to BiOBr. The changes in binding energies Bi 4f, Br 4d and O 1s indicate that the chemical environment of Bi 4f, Br 4d and O 1s has changed. The surface electron density caused by the change in charge redistribution can be attributed to the MOF-5 load. Also, Fig. 2e shows the survey spectra of XPS with 20% MB. By semi-quantitative analysis of the elements in the XPS spectrum, the elemental ratios of Bi/Zn, O/Zn, Br/Zn can be obtained: 19.71:3.09; 26.91:3.09; 13.19:3.09. It is consistent with the result of EDS in SEM, It further confirmed the successful preparation of MOF-5 and BiOBr.

Fig. 2
figure 2

X-ray photoelectron spectroscopy (XPS) spectra of pure BiOBr and 20% MB

Full size image

As shown in the Fig. 3, the specific surface area and BJH pore structure of the samples BiOBr, MOF-5 and 20%MB composite materials were studied. According to the IUPAC classification, the isotherms of BiOBr, MOF-5 and 20% MB correspond to the type IV and type H3 hysteresis loop, indicating the existence of mesopores and the material recombination has almost no effect on the adsorption capacity of the sample. In addition, the BET specific surface areas of MOF-5, BiOBr and 20%MB are 3.7763 m2/g, 16.9192 m2/g and 16.1288 m2/g. Compared with the BiOBr, the BET specific surface area of 20% MB is slightly reduced, but higher than that of MOF-5, indicating that the specific surface area of the composite is not the main factor affecting the photocatalytic activity [44, 45] (Table 1).

Fig. 3
figure 3

MOF-5, BiOBr, 20% MB: a N2 adsorption–desorption isotherm; b BJH pore size distribution diagram

Full size image

Tested the ultraviolet–visible diffuse reflectance spectrum (UV–Vis DRS), as shown in the Fig. 4. Obviously, the light absorption edge of BiOBr is 440 nm, which indicates that BiOBr microspheres have partial absorption in the visible light region. However, MOF-5 shows obvious light absorption in the ultraviolet region. It can be seen from Fig. 4a that compared with BiOBr, 20% MB has only a slight red shift. The reason for the red shift may be attributed to the interaction of the heterojunction formed by MOF-5 and BiOBr. According to the reflectance data, the Eg of the sample is further estimated by the formula:

$$ \left( {\alpha {\text{h}}\nu } \right)^{{\text{n}}} \; = \;A\left( {{\text{h}}\nu \; - \;{\text{E}}_{{\text{g}}} } \right), $$
(2)
Fig. 4
figure 4

a Ultraviolet–visible diffuse reflectance spectra of MOF-5, BiOBr and composite materials; b band gap distribution of BiOBr, MOF-5 and 20% MB (UV–Vis diffusion spectroscopy condition: UV-2600 Lambda 750 PE, Wavelength range: 200–800 nm, Absorption ratio reference: BaSO4; Eg calculation formula: [αhν)n = A(hν − Eg)]

Full size image
Fig. 5
figure 5

Photocatalytic degradation of RhB by MOF-5, BiOBr, 10%MB, 15%MB, 20%MB, 25%MB and 30%MB (experimental conditions: [RhB]0 = 10 mg/L, [Catalyst = 10 mg], Vsolution = 100 ml, λ > 420 nm, irradiation time: 18 min, nonlinear least square fits-lines)

Full size image

The Eg of BiOBr, MOF-5 and 20%MB are 2.69 eV, 2.92 eV, and 2.62 eV.

The photocatalytic activity of the prepared samples was evaluated by simulating the degradation of RhB under sunlight irradiation. Fig. 3a is a graph showing the degradation of RhB by pure BiOBr and MB composites under xenon lamp irradiation for 18 min. It can be seen from the figure that the MB materials loaded with different masses of MOF-5 show different photocatalytic degradation activities, and the order was 20%MB > 25%MB > 15%MB > 30%MB > 10%MB > BiOBr > MOF-5, the degradation rates were 99.7%, 98.4%, 95.5%, 93.8%, 89.4%, 76.6% and 2.1%.

The entire photocatalytic degradation process conforms to non-linear least squares ftting equation. The rate constant, standard deviation (σ) and correlation coefficient value (R2) of 20% MB, 25% MB, 15% MB, 30% MB, 10% MB, BiOBr and MOF-5 (Table 2) are shown in Table 2. Obviously, the 20% MB composite has the highest rate constant.

Table 1 BET specific surface area, pore size and pore volume of BiOBr, MOF-5 and 20% MB
Full size table
Table 2 The standard deviation (σ) for photocatalytic degradation RhB with correlation coefficient value (R2) a and the rate constants by the non-linear least squares fitting
Full size table

The function of free radicals in the photodegradation process was measured by the active substance capture experiment (Fig. S3), and p-benzoquinone (BQ, 5.0 mg), triethanolamine (TEOA, 0.1 mL) and isopropanol (IPA, 0.4 mL) were selected as a capture agent for capture experiments. When TEOA and BQ were added to the photocatalytic system, the degradation of RhB by 20% MB was negligible, and the photocatalytic activity was completely inhibited. The results show that h+ and ·O2− play important roles in photocatalytic reactions. When IPA was added to the photocatalytic system, the degradation of RhB did not change, indicating that ·OH had no effect in the degradation process. In order to evaluate the stability of the prepared photocatalyst, the RhB degradation experiment of the 20% MB composite was repeated under the same conditions. As shown in Fig. S4, after four cycles of experiments, the photocatalytic activity of the 20% MB composite material showed only a slight change (1.3%), indicating that the prepared 20% MB composite material has good photocatalytic stability.

The electrochemical impedance (EIS) experiment is used to analyze the photoelectric properties of the material. As shown in the Fig. S5, compared with the monomer, the radius of the impedance semicircle on the 20%MB Nyquist diagram is the smallest, the electron transfer process is faster, and the electron transfer resistance on the surface of the composite material is lower. According to the above results, the photoelectric performance of 20% MB is more effective than pure BiOBr (Fig. 6).

Fig. 6
figure 6

Mott–Schottky diagram of a BiOBr and b MOF-5(M-S diagram condition: Model CHI660C, 0.5 mol/L sodium sulfate solution, − 1.0 to 1.0 V)

Full size image

The flat band potentials of BiOBr and MOF-5 can be obtained by Mott–Schottky diagram.

$$ \frac{1}{{{\text{C}}^{2} }}\; = \;\frac{2}{{{\text{A}}^{2} {\text{e}}\varepsilon \varepsilon_{0} {\text{N}}_{{\text{A}}} }}\left( {{\text{E}}\; - \;{\text{E}}_{{{\text{FB}}}} \; - \;\frac{{{\text{kT}}}}{{\text{V}}}} \right), $$
(3)

Through the Mott–Schottky curve, the EFB value can be obtained. The slope of the tangent line is positive, indicating that the pure BiOBr and MOF-5 have n-type semiconductor characteristics. It is generally believed that for n-type semiconductors, the flat band potential is positive 0.2 V than the conduction band potential. The calculated conduction band potentials of BiOBr and MOF-5 are − 0.28 V and − 1.04 V.

Photocatalytic mechanism

Through UV–Visible diffuse reflectance spectroscopy and Mott–Schottky calculations, the conduction band (CB) potential and valence band (VB) potential of BiOBr are − 0.28 V and 2.41 V, respectively, and the conduction band (CB) potential and valence band (VB) potentials of MOF-5 are − 1.04 V and 1.88 V, respectively. As shown in the figure, under simulated sunlight, both BiOBr and MOF-5 can be excited to obtain photo-generated electron–hole pairs. Since the CB value (− 0.28 V) of BiOBr is more corrected than the standard redox potential (− 0.33 V vs. NHE) of ·O2/·O2−, the accumulated electrons on the CB of BiOBr cannot have a high enough reduction potential to form ·O2/·O2−. At the same time, because the VB value of MOF-5 (1.88 V) is more negative than the potential of H2O/·OH (1.99 V vs NHE), the holes on the VB of MOF-5 can hardly generate H2O/·OH radicals. By XPS test, the charge was found to migrate from the MOF-5 surface to BiOBr. Therefore, the photocatalytic reaction mechanism belongs to type II heterojunction.

BiOBr and MOF-5 act as an oxidation photocatalyst with a lower Fermi level and a reducing semiconductor with a higher Fermi level. After BiOBr and MOF-5 contact with each other and form a heterojunction, the charge is transferred from MOF-5 to BiOBr until the Fermi level reaches equilibrium. Under the irradiation of light, the photogenerated electrons in the CB of BiOBr will migrate to the VB of MOF-5 through the intimate interface, and recombine with holes under the driving force of the product. In the field of electronics, both useful electrons and holes with higher redox potential are retained on the CB of MOF-5 and the VB of BiOBr. Moreover, RhB is excited by light to produce e− and RhB*. Since the LUMO of RhB is − 1.04 eV, which is lower than the CB of MOF-5 and BiOBr, the e− generated by RhB excited by light can migrate to the CB of MOF-5 and BiOBr. This is the dye sensitization effect of RhB [46, 47]. As a result, the electrons on the CB of MOF-5 can react with ·O2 to form ·O2−, and the holes on the VB of BiOBr can oxidize H2O to ·OH. Subsequently, these active substances will decompose pollutants.

Conclusion

In summary, a series of type II heterojunction MOF-5/BiOBr composites were prepared by hydrothermal method. Compared with the monomer BiOBr, the degradation rate of RhB by 20wt%MB composite material increased by 4 times within 18 min after xenon lamp irradiation. This catalyst is expected to play an important role in the treatment of environmental wastewater. In addition, the microstructure of the heterojunction with high dispersibility can provide more reaction sites. This work provides enlightenment for the design of high-efficiency heterojunction photocatalysts using a simple synthesis route.