A novel 3D Fe₂O₃@ZnBi₂O₄ n–p heterojunction with high photocatalytic activity under visible light

Abstract

A novel n–p Fe₂O₃@ZnBi₂O₄ (FZB) heterojunction with a unique 3D structure was fabricated in two simple steps to break down MB under visible light. First, the polymer gel combustion technique was employed to fabricate a 3D Fe₂O₃ framework. Next, a microwave-assisted precipitation approach was used to incorporate 3D ZnBi₂O₄ flakes onto the framework surface. FZB can effectively collect a broad range of UV-Vis light and sunlight. Surprisingly, the core–shell p–n heterojunction structure makes it easier for photogenerated charges to move and separate. This is because the semiconductor parts are better connected and the electric field is inside the two junctions. UV-Vis-DRS, EIS, PL, and XPS analyses confirmed this phenomenon. As a result, n–p FZB, with a Fe₂O₃/ZnBi₂O₄ molar ratio of 2 : 1, showed the highest photocatalytic activity, increasing the reaction rate by 4.2 times compared to Fe₂O₃ and 2.8 times compared to ZnBi₂O₄. Exposure to light for 100 min at a concentration of 0.7 g L⁻¹ the FZB catalyst and pH = 9 led to the breakdown of more than 95% of the 50 ppm MB solution. In addition, the photodegradation of MB by n–p FZB increased the reaction rate by 2.15 times by adding hydrogen peroxide (H₂O₂), which is known as the photo-Fenton reaction. The efficacy showed remarkable photocatalytic activity, which increased the reaction rate by 5.95 times when persulfate was used. Finally, after examination of the energy band structure of the materials and the findings regarding the function of the oxidizing sites in the photocatalytic process, a reaction mechanism was proposed.

---

1. Introduction

The rapid growth of global industrialization has led to an increase in the demand for water resources. Moreover, industrial, agricultural, and urban regions release a substantial amount of garbage (approximately 2 million tons) into the environment daily, endangering human health and increasing the risk of water resource depletion. Industrial wastewater, which frequently contains numerous hazardous organic compounds, is challenging to break down and exhibits significant biological toxicity. Currently, more than 100 000 varieties of dyes are used in diverse textile manufacturing procedures. The majority of dyes, including rhodamine B (RhB), methyl orange (MO), and methylene blue (MB), are soluble in water, have low biodegradability, and pose potential risks to ecosystems.¹ Approximately 13% of the global population is already experiencing water scarcity, and this figure is projected to increase to 40% by the 2050s.² Thus, the global issue lies in discovering practical solutions for water treatment and recycling technologies that are environmentally benign, cost-effective, and highly efficient.³

Advanced oxidation uses highly reactive oxidizing agents to break down and remove organic pollutants from water or air. Studies have demonstrated that Advanced Oxidation Processes (AOPs) are efficient techniques for converting harmful organic pollutants into minerals and for eliminating contaminants from wastewater. Advanced oxidation processes (AOPs) that use ˙OH radicals (which have a standard electrode potential of 2.80 V/SHE) are very strong oxidizing agents that do not care what they are attacking. They can efficiently convert organic contaminants present in water into harmless smaller molecules such as CO₂ and H₂O.^3–6 Photocatalysis, typically using UV-Vis light, generates ˙OH radicals. Advanced oxidation techniques have been used in the textile, pharmaceutical, printing, and pulp industries to treat wastewater.^3,7,8 However, this technology is still in use. In practice, photocatalytic technology still has limitations in terms of quantum efficiency, durability, and catalyst cost. In the majority of individual semiconductor photocatalysts, the pace at which the photoelectron–hole pairs recombine is rapid, resulting in low quantum efficiency and performance.^9,10 Several widely recognized semiconductors exhibit excellent photocatalytic activities owing to their large bandgap energies. UV radiation specifically activates TiO₂, ZnO, CeO₂, and SnO₂. Consequently, their spectral response range is limited, preventing the effective absorption of sunlight, particularly visible light.^11–13 Sulfide and nitride semiconductors have band gap energies in the visible range; however, they are not very stable and do not last long in wet environments.¹⁴ Hence, it is imperative to address the aforementioned constraints by creating novel semiconductor materials derived from conventional semiconductors.

Hematite, also known as Fe₂O₃, is a semiconductor that exhibits n-type conductivity and has a band gap energy ranging from approximately 2.0 to 2.2 electron volts (eV). Hematite can capture light ranging from yellow to ultraviolet, specifically at a wavelength of λ ≤ 600 nm. It can absorb up to 40% of the energy in the solar spectrum. In addition, hematite exhibits favorable environmental characteristics, demonstrates excellent stability in aqueous solutions with pH greater than 3, and is among the most cost-effective options for semiconductor materials.^14–16 Thus, hematite is a highly intriguing choice for photocatalytic applications that specifically aim to harness visible light, particularly sunlight.¹⁷ Nevertheless, certain characteristics, such as the rapid recombination rate of electrons and holes, the limited diffusion length of holes (2–4 nm), and inadequate electrical conductivity, constrain the photocatalytic efficiency of Fe₂O₃, leading to diminished quantum optical efficiency.¹⁴ By combining with a second semiconductor, we can overcome these constraints and create a novel heterogeneous junction catalyst that offers exceptional benefits.^18–23 ZnBi₂O₄, a bismuthate compound, is considered a promising candidate owing to its exceptional characteristics. ZnBi₂O₄ is a well-established p-type semiconductor with a direct bandgap ranging from 2.2 to 3.0 eV,^24,25 which makes it highly appropriate for applications that utilize visible light or sun sources. ZnBi₂O₄ has a distinctive crystal structure, characterized by zigzag chains of metal–oxygen bonds formed by Zn and Bi atoms at the corners.²⁶ Moreover, ZnBi₂O₄ demonstrates chemical stability, a high optical current density, and the ability to undergo oxidation by hydroxyl (˙OH) ions.^27,28

Fe₂O₃ and ZnBi₂O₄ combinations can form n–p heterojunction semiconductors that are photocatalyst-driven under visible light or sun sources. Fe₂O₃ is an n-type semiconductor with the Fermi level located near the E_CB, whereas ZnBi₂O₄ is a p-type semiconductor with the Fermi level located near the E_VB. At the junction between the two semiconductors, electrons are transferred from Fe₂O₃ to ZnBi₂O₄ until their Fermi levels align, creating an electric field that can boost the transport and separation of photogenerated charges over ZnBi₂O₄.

In this study, a novel 3D Fe₂O₃@ZnBi₂O₄ n–p heterojunction was fabricated in two steps. Initially, the gel–polymer combustion technique produced Fe₂O₃ using tartaric acid and polyvinyl alcohol as the combustion agents. Next, ZnBi₂O₄ was introduced onto the Fe₂O₃ surface via microwave-assisted precipitation. In addition, the physicochemical properties and photocatalytic activities of the heterojunctions were studied and discussed in detail. Furthermore, the physicochemical characteristics and photocatalytic activity of the heterojunctions were comprehensively examined and analyzed. Additionally, experimental investigations into the function of oxidizing agents in the breakdown of MB have been conducted. A reaction mechanism for dye degradation was proposed.

2. Experimental

2.1 Materials

Bi(NO₃)₃·5H₂O, Zn(NO₃)₂·6H₂O, Fe(NO₃)₃·9H₂O, NH₄OH, H₂O₂, Na₂S₂O₈, and methylene blue (MB) were purchased from Xilong Scientific Co. Tartaric acid, polyvinyl alcohol (PVA, M = 145 000 g mol⁻¹, 99%), AgNO₃ (99.9%), disodium ethylenediamine tetraacetate (Na₂EDTA, 99%), isopropanol (Isp, 99.5%), phenol (Phe(OH), 99%), and 1,4-benzoquinone (p-BQ, 98%) were purchased from Sigma-Aldrich. All the reagents were used without further purification.

2.2 Synthesis of Fe₂O₃ materials

Synthesis of the Fe₂O₃ material was done using the combustion technique. 1.32 grams of PVA and tartaric acid (4.5 g) were combined in 80 mL of deionized water at 50 °C using a magnetic stirrer. The 0.5 M Fe(NO₃)₃ solution was added to the mixture slowly, 40 mL at a time, and the mixture was stirred well. Subsequently, the pH of the solution was modified to 4 using a 2 M NH₄OH solution. The reaction mixture was continuously agitated at 80 °C until water evaporated and transformed into a gelatinous substance. The gel was allowed to remain undisturbed overnight for aging and subsequently dried at 100 °C for 2 h. Finally, the gel was calcined at 500 °C for 2 h, leading to the formation of a brick-red powder containing Fe₂O₃.

2.3 Synthesis of Fe₂O₃@ZnBi₂O₄ materials

n–p Fe₂O₃@ZnBi₂O₄ was produced using a microwave-assisted precipitation method. The sample had a molar ratio of 2 : 1 between Fe₂O₃ and ZnBi₂O₄. The solution was prepared by dissolving 1.455 g of Bi(NO₃)₃·5H₂O in a 5% HNO₃ solution. Subsequently, 50 mL of 0.03 M Zn(NO₃)₂ solution was added, and the mixture was stirred using a magnetic stirrer for 30 minutes. In parallel, 0.48 g of previously synthesized Fe₂O₃ was dispersed in 100 mL of sonicated deionized water. This suspension was then introduced into the Bi–Zn solution and stirred for an additional 45 minutes. The pH of the resulting mixture was adjusted to 10 by the dropwise addition of a 2 M NH₄OH solution under continuous stirring. After stirring for 30 minutes, the reaction mixture was transferred to a 500 mL flask and subjected to microwave heating at 400 W for 45 min. The temperature of the reaction mixture was lowered to match the surrounding temperature; it was separated by filtering, and the resultant solid was cleaned by rinsing it with deionized water until it achieved a pH of 7. Finally, the mixture was subjected to a temperature of 500 °C for 3 h, followed by drying at 90 °C for 3 h, which resulted in the production of Fe₂O₃@ZnBi₂O₄ (FZB21). For comparative analysis, materials having Fe₂O₃/ZnBi₂O₄ molar ratios of 1 : 2 and 1 : 1 were produced and labeled as FZB12 and FZB11, respectively. The ZnBi₂O₄ sample synthesized under similar conditions, in the absence of Fe₂O₃, is denoted as ZBO.

2.4 Characterization

The composites were analyzed using X-ray diffraction (XRD) on a Bruker D8 Advance instrument (Germany) to investigate their phase compositions and crystal structures. The interfacial bonding characteristics of the material were analyzed using Fourier transform infrared spectroscopy (FTIR) with a PerkinElmer Spectrum Two instrument (United States). The optical absorption and bandgap properties of the material were analyzed by ultraviolet-visible diffuse reflectance spectroscopy (UV-Vis DRS) using a Jasco V670 (Japan), and photoluminescence (PL) measurements were conducted using a machine 205 Horiba iHR550 (Japan). The surface morphology and structure of materials were analyzed using a Hitachi S-4800 scanning electron microscope (SEM) from Japan and high-resolution transmission electron microscopy (HR-TEM) with an S-4800 NIHE model. The oxidation status of the elements in each catalyst was evaluated using X-ray photoelectron spectroscopy (XPS) with an ESCA-3400 apparatus (Shimadzu). The analysis was conducted using Mg Kα 175 radiation (1253.6 eV, 10 kV, 20 mA). Mott–Schottky plots, photocurrent curves, and Nyquist plots were obtained using a CHI 650E electrochemical workstation with a three-electrode cell in a 0.1 M Na₂SO₄ electrolyte solution.

2.5 Evaluating photocatalytic performance

The catalytic activity of the FZB powder was assessed by detecting the degradation of the MB dye under exposure to halogen light (150 W, Osram, Germany). The reaction consists of two steps.

To establish equilibrium between the adsorption and desorption of MB on the surface of the catalyst, the reaction mixture was stirred in the absence of light for 1 h as a step in the adsorption process.

Once the catalyst surface reaches the point of MB adsorption–desorption equilibrium, the breakdown process can be initiated by exposing it to light irradiation. To carry out the photocatalytic reaction, the light source was activated until most of the color was eliminated from the MB solution. A magnetic stirrer and circulating water were used to agitate the reaction mixture until it reached a steady state at ambient temperature. The catalyst was separated from the suspension by subjecting 3 mL of the mixture to centrifugation every 20 min. The MB concentration was determined using an ultraviolet-visible spectrophotometer.

3. Results

3.1 Charaterization of the 3D Fe₂O₃@ZnBi₂O₄ n–p (FZB) heterojunction

Fig. 1 displays the XRD patterns, which show the structure and crystalline phase composition of n–p FZB. The two crystalline phases constituting the FZB materials were α-Fe₂O₃ and ZnBi₂O₄. The lattice planes (012), (104), (110), (113), (024), (116), (214), and (300) exhibit the α-Fe₂O₃ phase, according to JCPDS card No. 79-0007.^29,30 The ZnBi₂O₄ (ZBO) tetragonal phase is clearly indicated by the diffraction peaks corresponding to the lattice planes (100), (201), (002), (211), (210), (301), (311), (212), (321), (322), (203), (421), (402), and (331). This matches the information in JCPDS card No. 00-043-0449.³¹ These results indicated the successful synthesis of FZB heterostructures with good crystallinity.

Fig. 1 XRD patterns of α-Fe₂O₃, ZBO, and FZB.

Furthermore, SEM and TEM analysis were used to study the morphology and structure of the samples. The SEM image in Fig. 2A shows that the Fe₂O₃ material produced using the polymer gel combustion technique possessed a three-dimensional structure characterized by flat branches. These branches are formed from the amalgamation of several tiny planar particles during gel combustion. Besides, the FZB material (Fig. 2B) exhibited a 3D structure composed of two components. This structure consisted of flower-like clusters composed of ZnBi₂O₄, which were accompanied by long slender blades that covered the surface of the Fe₂O₃ branches. The Fe₂O₃ flat branches were seen to undergo rounder and larger size transformation after being subjected to microwave heating and calcination during the synthesis of FZB. In addition, the high-resolution TEM (HR-TEM) results confirmed the formation of an FZB heterojunction. The lattice fringes were approximately ∼0.370 nm which is ascribed to the (012) plane of α-Fe₂O₃,^32,33 in which ∼0.292 nm belongs to the (102) plane of ZnBi₂O₄. Together with the XRD findings mentioned earlier, this provides evidence that the FZB material was effectively synthesized using a two-step process.

Fig. 2 SEM images of (A) Fe₂O₃ and (B) FZB; (C) TEM, (D) HR-TEM images of FZB, (E–I) EDS mapping of FZB, and (J) EDS spectra of FZB.

Moreover, EDS spectra and elemental mapping of FZB further confirmed the effective synthesis of FZB. Fig. 2E–J illustrate that the sample contained Zn, Fe, Bi, and O and exhibited a very regular distribution. No extraneous foreign components were identified in any of the samples.

The chemical bonds of the catalysts were studied by FT-IR spectroscopy (Fig. 3). Fig. 3A shows that the absorption peaks at approximately 462 and 555 cm⁻¹ of the Fe₂O₃ material can be attributed to the stretching vibrations of Fe–O.^20,34 For ZnBi₂O₄, the absorption peaks at 845 and 1390 cm⁻¹ correspond to the Bi–O and Bi–O–Bi vibrations, respectively. Additionally, the peaks at 480, 528, and 575 cm⁻¹ indicate the vibration of Zn–O bonds.^25,26,31,35 It is worth mentioning that the absorption peaks in the FT-IR spectrum of the FZB material are very strong and show bond metal oxygens, such as Zn–O, Fe–O, Bi–O, and Bi–O–Bi. Additionally, vibration peaks were observed in the range of 3300–3500 cm⁻¹ attributed to O–H from absorbed H₂O. The FT-IR results further support the successful synthesis of n–p FZB heterojunctions.

Fig. 3 (A) FTIR, (B) UV-Vis-DRS, (C) Tauc plots, and (D) PL spectra of Fe₂O₃, ZBO, and FZB.

The UV-Vis-DRS spectrum is associated with the optical absorption range and bandgap energy of the material. Fig. 3B demonstrates that in the ultraviolet light spectrum, the synthetic materials exhibited excellent absorption capabilities, with the order of absorption being ZBO < FZB < Fe₂O₃. In the visible light spectrum, the FZB and Fe₂O₃ materials exhibited significant absorption, whereas the ZBO material exhibited significantly weaker absorption. The optical absorption characteristics can be verified by estimating the position of the optical absorption edge and bandgap of the material using the Tauc diagram (Fig. 3B and C).^36,37 The bandgaps of ZBO, Fe₂O₃, and FZB were measured at 2.84, 1.95, and 1.98 eV, respectively. The absorption wavelength of each material was determined using the formula λ = hc/E_g, resulting in values of 436, 635 nm for Fe₂O₃, and 626 nm for FZB. Thus, it can be deduced that ZBO exhibits a high capacity for absorbing ultraviolet rays and only partially absorbs visible light with a wavelength λ ≤ 436 nm. However, both the FZB and Fe₂O₃ materials possess a strong ability to absorb light across the ultraviolet and visible regions of the light spectrum. The advantageous attribute of FZB is highly advantageous for the practical implementation of the FZB material as a solar photocatalyst, as it enables the utilization of sunlight as a light source to initiate the photocatalytic process. The new catalyst (FZB) has notable advantages in this regard.

In addition, the optical characteristics of the FZB material were studied using photoluminescence (PL) spectroscopy, which enables the investigation of both electron–hole recombination and charge separation in materials. The photoluminescence (PL) spectra of ZnBi₂O₄, Fe₂O₃, and FZB were recorded at an excitation wavelength of 400 nm. Fig. 3D shows that all emission peaks are situated exclusively within the visible light range, specifically in the deep-level emission (DLE) region. The strongest emission peak at 436 nm is related to the rapid recombination probability of electron–hole pairs in ZnBi₂O₄, and the peak position agrees well with the bandgap (2.84 eV) of ZnBi₂O₄. Meanwhile, the broad peak with the highest intensity at 440 nm was associated with the atomic transitions of iron vacancies in Fe₂O₃.³⁸ The weak shoulder peaks at 458 and 488 nm and a broad peak at 583 nm in the ZnBi₂O₄ PL spectra correspond to the ³P_0,1 → ¹S₀ transitions of the Bi³⁺ ions.³⁹

The formation of complexes on the heterogeneous surface reduced the PL intensity and peak position shift. The peak observed at 438 nm for the FZB composite exhibited a sharp decrease in intensity, and there was a slight red shift compared to that of pristine ZnBi₂O₄ (436 nm) and a slight blue shift compared to that of pure Fe₂O₃ (440 nm). The significant decrease in intensity indicates that after the formation of the heterojunction between Fe₂O₃ and ZnBi₂O₄, the rapid recombination of photogenerated free radicals was significantly inhibited. The broad monotonically shifted emission peak of the FZB nanocomposite was attributed to the presence of defect sites, leading to the generation of localized energy levels in the forbidden energy gap.

The surface chemistry of the synthesized ZnBi₂O₄ and FZB was analyzed using XPS. Fig. 4 shows the binding energies of the Zn 2p, O 1s, Fe 2p, and Bi 4f orbitals.

Fig. 4 XPS spectra of (A) FZB heterojunction and high-resolution (B) O 1s, (C) Fe 2p, and (D) Bi 4f XPS spectra.

The Zn 2p orbitals (Fig. 4A) of the FZB sample exhibit a complex electronic structure, as evidenced by the presence of two distinct peaks in the X-ray photoelectron spectroscopy (XPS) analysis. The peaks observed at 1021.54 eV and 1044.64 eV correspond to the energy levels of Zn 2p_3/2 and Zn 2p_1/2,⁴⁰ respectively. These energy levels are characteristic of spin–orbit splitting in the Zn2p orbitals, providing valuable information about the electronic configuration of zinc atoms within the composite material. The slight redshift observed in the binding energy peaks of the FZB sample compared to those of ZnBi₂O₄ (ZBO) is a significant indicator of the successful formation of the composite material. This shift in the binding energies suggests a change in the chemical environment surrounding the zinc atoms, likely due to the interaction between the Fe₂O₃ and ZnBi₂O₄ components.⁴¹ The redshift may be attributed to factors such as charge transfer, changes in coordination, or alterations in the local electronic structure resulting from the integration of iron oxide into the zinc–bismuth oxide matrix.

The XPS spectrum of O 1s, as depicted in Fig. 4B, reveals two distinct peaks that provide valuable information regarding the oxygen species present in ZBO. The lattice oxygen (O²⁻) at 530.33 eV represents the oxygen ions incorporated into the crystal lattice structure of the material. The surface-adsorbed oxygen (O⁻) at 532.02 eV corresponds to the oxygen species absorbed on the surface of the material.⁴² The XPS analysis of the FZB sample revealed a notable shift in the two peaks towards higher energy levels, indicating significant changes in the electronic structure of the material. The most striking difference is observed in the relative intensity of the 532.65 eV peak, which is markedly higher in FZB than in ZBO. This increase in the peak intensity is a strong indicator of a higher concentration of oxygen vacancies within the FZB sample. Oxygen vacancies play a crucial role in the electronic and optical properties of metal oxides and often serve as active sites for various chemical reactions.

Fig. 4C shows two prominent peaks of the Fe 2p core level at 710.7 eV, and 724.4 eV attributed to Fe 2p_3/2 and Fe 2p_1/2 states.^43,44 The reported binding energy values are consistent with the typical values for iron oxides, suggesting that the iron in the sample is likely to be in an oxidized state.

Fig. 4D shows that the two distinct peaks of 158.4 and 163.6 eV are related to the Bi 4f_7/2 and Bi 4f_5/2 spin-orbital pairs of Bi³⁺, for ZnBi₂O₄.⁴⁵ The binding energies of Bi 4f_7/2 (158.6 eV) and Bi 4f_5/2 (163.8 eV) shifted to higher binding energies for FZB. This shift in binding energies suggests a change in the chemical environment of bismuth atoms when ZnBi₂O₄ is combined with Fe₂O₃.⁴⁶ The lower binding energies observed in FZB may indicate a reduction in the oxidation state of Bi or alterations in its electronic structure.

Electrochemical impedance spectroscopy (EIS) and transient photocurrent measurements were performed to investigate the charge-transfer behavior. As illustrated in Fig. 5A, the Nyquist plot reveals that the arc radius of the two-component nanocomposite was smaller than that of the individual components. This indicates that the FZB heterostructure encounters fewer obstacles in the charge transfer across the interface, resulting in a higher charge separation efficiency. The reduced charge transfer resistance in the FZB heterostructure can be attributed to the synergistic effect between the two components, which facilitates electron–hole pair separation and enhances overall photocatalytic performance. This improvement in charge transfer dynamics is further supported by transient photocurrent measurements, which show a higher and more stable photocurrent response for the nanocomposite compared to the individual components. To validate these findings, the photocurrent responses of the samples were recorded over four on/off cycles (Fig. 5B). FZB exhibited significantly higher photocurrent responses than the Fe₂O₃ and ZnBi₂O₄ photocatalysts. These results confirm that the internal electric field generated at the n–p heterojunction interface between ZnO and ZnBi₂O₄ enhances charge separation and transport efficiency, consistent with the EIS and PL measurement results.

Fig. 5 (A) EIS curves and (B) transient photocurrent responses of Fe₂O₃, ZBO, and FZB.

3.2 Photocatalytic activity of FZB

The photocatalytic activity was assessed by measuring the degradation efficiency of MB under visible-light irradiation. The experimental variables influencing photocatalytic activity, including the molar ratio of Fe₂O₃/ZnBi₂O₄, solution pH, and catalyst concentration, were investigated. The results are shown in Fig. 6. Experimental investigations were conducted to examine the impact of the molar ratio of Fe₂O₃/ZnBi₂O₄ on the pH of natural solutions. The data presented in Fig. 6A demonstrate that the heterojunctions were better than the pure components in the decolorization of MB. The MB degradation efficiency after 160 min of irradiation was 69.2% for Fe₂O₃ and 83.3% for ZBO. The degradation efficiencies of FZB12, FZB11, and FZB21 were 96.5, 98.1, and 99.5%, respectively. Furthermore, the Langmuir–Hinshelwood pseudo-first-order kinetic model was employed to compute the rate constant of the photocatalytic reaction. As shown in Fig. 6B, the reaction rate constant (k_app) for FZB21 is 0.0326 min⁻¹, which is higher than those of FZB12 (0.0213 min⁻¹) and FZB11 (0.0262 min⁻¹) and specifically surpasses those of Fe₂O₃ (0.0078 min⁻¹) by approximately 4.2 times and ZnBi₂O₄ (ZBO) (0.0116 min⁻¹) by 2.8 times. The improved effectiveness of decolorization is ascribed to the synergistic effect between ZBO and Fe₂O₃ in the heterojunction structure, which successfully limits the recombination of the photoexcited charge carriers.

Fig. 6 Effects of the molar ratio of Fe₂O₃/ZnBi₂O₄ (A)–(B), solution pH (C) and (D), initial MB concentration (E) and (F), and photocatalyst concentration (G) and (H) on MB degradation using the FZB heterojunction under visible light.

The efficacy of photodegradation is frequently affected by the solution pH, partly because of its association with the surface charge of the catalyst, which subsequently affects the adsorption of organic compounds onto the catalyst surface. In this experiment, the pH of the solution was calibrated within an interval of 4–10. Fig. 6C and D illustrate that the decolorization of MB by FZB21 was slow at solution pH values ranging from 4 to 7; however, its rate greatly increased at solution pH values ranging from 8 to 10. After 45 min of exposure to light, the reaction rate constants varied between 0.017 min⁻¹ and 0.020 min⁻¹ within the pH range of 4–7. Within the pH range of 8–10, the values varied from 0.0572 to 0.0853 min⁻¹. The breakdown of MB was observed to be especially slow at pH values below 7, mostly because of the positive charge present on both the catalyst surface (pH_z of FZB21 = 7.8) and MB⁺ ions. Thus, only a restricted amount of MB was absorbed onto the catalyst surface and underwent breakdown. At pH values greater than 7, the FZB21 surface had a negative charge, which facilitated the effective adsorption of MB onto its surface. Hence, the rate constant of the reaction increases and attains its maximum value at pH 10. Nevertheless, because of the highly alkaline pH of 10, this study employed a pH of 9.

Experimental investigations of the impact of the initial concentration of the MB solution on the effectiveness of MB decolorization are shown in Fig. 6E and F. The results indicated that the decline in the MB concentration exhibited a progressive reduction as the initial MB concentration increased from 30 to 60 ppm. At a concentration of 30 ppm, MB exhibited a reaction rate constant of 0.0698 min⁻¹, 40 ppm at 0.0536 min⁻¹, 50 ppm at 0.0294 min⁻¹, and 60 ppm at 0.0206 min⁻¹. The reduction in efficiency with an increasing initial MB concentration is attributed to the light-blocking phenomenon. The deeper hue of the solution obstructs the ability of the catalyst to absorb light, resulting in a reduction in the concentration of photoexcitation charge carriers, which is the process responsible for organic matter biodegradation.

Furthermore, the effect of the catalyst concentration was also investigated. Fig. 6G and H indicate that the efficiency of MB degradation approached 0.0310 min⁻¹ when the catalyst concentration was 0.5 g L⁻¹ and significantly increased to 0.0322 min⁻¹ when the catalyst concentration was 0.7 g L⁻¹. Nevertheless, the efficiency of MB degradation declined to 0.0283 and 0.0259 when the catalyst concentration was raised to 1.0 and 1.2 g L⁻¹, respectively. The reduction in efficiency can be attributed to two factors: first, the surplus catalyst induces a light-blocking phenomenon, and second, it facilitates competition for the active sites on the catalyst surface.

To study the parameters influencing the efficiency of MB degradation, the optimal conditions for the photocatalytic reaction were determined as follows: FZB21 material, pH 9, MB concentration 60 ppm, catalyst content 0.7 g L⁻¹.

Additionally, we performed a series of studies to evaluate the influence of H₂O₂ and K₂S₂O₈ on the MB degradation efficiency under optimal conditions. Fig. 7A and B indicate that in the absence of light, the degradation of MB was significantly challenging in the presence of a catalyst and H₂O₂, yielding a rate constant of k = 0.00173 min⁻¹ and a degradation efficiency of only 13.9%. In contrast, exposure to light resulted in a substantial reduction in the MB concentration. The reaction rate constants were k = 0.0322 min⁻¹ for 0 mM H₂O₂, k = 0.0338 min⁻¹ for 2 mM H₂O₂, k = 0.0645 min⁻¹ for 4 mM H₂O₂, and k = 0.0695 min⁻¹ for 8 mM H₂O₂. The addition of H₂O₂ increased the reaction rate by a factor of 2.0 with 4 mM H₂O₂ and by a factor of 2.15 with 8 mM H₂O₂. Doubling the concentration of H₂O₂ from 4 to 8 mM resulted in a small increase in the reaction rate constant, indicating that the 4 mM concentration of H₂O₂ was nearly optimal. The introduction of H₂O₂ initiated the photo-Fenton reaction (H₂O₂/Vis, Fe₂O₃/H₂O₂), augmenting the concentration of ˙OH and ˙HO₂ in the solution (reactions (1)–(3)). This facilitated the expedited breakdown of MB^47–49

Fe³⁺ + e⁻ → Fe²⁺ (1)

Fe²⁺ + H₂O₂ → Fe³⁺ + ˙HO + OH⁻ (2)

Fe³⁺ + H₂O₂ → H⁺ + Fe²⁺ + ˙HO₂⁻ (3)

Fig. 7 Effect of H₂O₂ (A) and (B) and K₂S₂O₈ (C) and (D) on MB degradation using the FZB heterojunction under visible light.

Furthermore, the effect of S₂O₈²⁻ ions on the photodegradation efficiency of FZB was also studied (Fig. 7C and D). The reaction rate constant augmented by 1.5 times (k = 0.0486 min⁻¹) for 1 mM K₂S₂O₈, 5.7 times (k = 0.1836 min⁻¹) for 2.0 mM K₂S₂O₈, and 5.95 times (k = 0.1917 min⁻¹) for 3 mM K₂S₂O₈. Electrons generated using light interact with Na₂S₂O₈ on the surface of the catalyst to create ROS radicals ˙SO₄⁻ and ˙OH (reactions (1), (4)–(7))^50,51 which accelerates the reaction. The reaction slows the recombination of photogenerated electrons and holes, and K₂S₂O₈, acts as an electron trapping agent. At the same time, it creates reactive oxygen species (ROS), which make the strong oxidizing agent more effective at breaking down methylene blue (MB).

Fe³⁺ + S₂O₈²⁻ → ˙S₂O₈²⁻ + Fe²⁺ (4)

S₂O₈²⁻ + Fe²⁺ → ˙SO₄⁻ + SO₄²⁻ + Fe³⁺ (5)

S₂O₈²⁻ + e⁻ → ˙SO₄⁻ + SO₄²⁻ (6)

˙SO₄⁻ + H₂O → ˙HO + SO₄²⁻ + H⁺ (7)

The use of K₂S₂O₈ significantly improved the effectiveness of MB degradation relative to that of H₂O₂. After 80 minutes of irradiation, the MB decomposition efficiency of FZB/H₂O₂ exceeded 99.4%, but the MB decomposition efficiency of FZB/K₂S₂O₈ attained 99.9% after merely 40 minutes of irradiation. The reaction rate constant for FZB/K₂S₂O₈ was 0.1917 min⁻¹, which is 2.8 times greater than that of FZB/H₂O₂ at 0.695 min⁻¹. There are several reasons why the MB decomposition reaction with K₂S₂O₈ was much faster than that with H₂O₂. (i) The first is that the O–O bond energy in K₂S₂O₈ is lower than that in H₂O₂ (213.3 kJ mol⁻¹), which makes S₂O₈²⁻ more reactive than O₂²⁻.⁵² (ii) The second step involves the addition of K₂S₂O₈, which generates radicals ˙SO₄⁻, ˙S₂O₈²⁻, and ˙HO (reactions (1), (2) and (4)–(7)), while the addition of H₂O₂ produces ˙HO and ˙HO₂⁻ (reactions (1)–(3)). The standard reduction potential of ˙SO₄⁻ (2.5–3.1 V) exceeds that of ˙HO (1.7–2.7 V) and ˙HO₂⁻ (1.8 V). At neutral pH, the ˙SO₄⁻ radical has a higher standard oxidation potential than the hydroxyl radical. It can successfully interact with organic molecules across a wide pH range of 2–8. (iii) The third is the sulfate radical, which is more selective and effective than ˙HO radicals and can transfer electrons to organic molecules with aromatic or unsaturated bonds.⁵³ (iv) Finally, the half-life of ˙SO₄⁻ is 30–40 μs, exceeding that of ˙HO at 1 μs, and it can oxidize contaminants at an exceptionally high rate of 106–109 M s⁻¹.^54,55 The prolonged half-life enhances substantial interactions and mass transfer between ˙SO₄⁻ radical and reactants.

These results demonstrate that the FZB catalyst effectively degrades MB under visible light, demonstrating its superior photocatalytic capabilities. The incorporation of minor quantities of oxidants (hydrogen peroxide and sodium persulfate), especially persulfate, significantly improved this characteristic.

3.3 Proposed mechanism

Experiments with selective scavengers at 1 mM doses were conducted to investigate the involvement of the reaction species in the degradation of MB. Na₂EDTA sequestered h⁺, p-BQ sequestered ˙O₂⁻, Isp sequestered ˙OH, AgNO₃ sequestered e⁻, and PheOH sequestered ˙SO₄⁻.^56–60Fig. 8A and B illustrate that the rate of MB breakdown diminished owing to the inhibitory effects of the scavengers. The use of p-BQ, Isp, AgNO₃, and Na₂EDTA resulted in MB degradation efficiencies of 87.8, 77.7, 56.9, and 48.9%, respectively. The efficiency of MB degradation decreased marginally with the introduction of p-BQ and Iso but declined significantly with the incorporation of AgNO₃ and EDTA. This indicates that ˙O₂⁻, ˙OH, e⁻, and h⁺ oxidants collaboratively decompose MB by photocatalysis, with e⁻ and h⁺ assuming the most significant roles. Moreover, the incorporation of persulfate and PheOH scavengers markedly lowered the degradation efficiency of MB by 33.3% and 68.4%, respectively, after 40 and 80 min of illumination, respectively, underscoring the essential coordinating function of the ˙SO₄ radical in MB degradation.

Fig. 8 Effect of radical scavengers (A and B) on MB degradation using FZB. (C) Mott–Schottky plots of Fe₂O₃, ZBO, and FZB.

Fig. 8C illustrates that the Mott–Schottky (M–S) plots exhibit a positive slope for n-type semiconductors and a negative slope for p-type semiconductors. The M–S plot of Fe₂O₃ exhibited a positive slope, signifying n-type semiconductor characteristics. The M–S curve of ZnBi₂O₄ has a negative slope, confirming its p-type characteristics. The flat band potentials of Fe₂O₃ and ZnBi₂O₄ were ascertained to be +0.28 V and +2.20 V vs. Ag/AgCl at pH 7 (corresponding to +0.485 V and +2.425 V vs. NHE), respectively. In n-type semiconductors, the conduction band (E_CB) is generally positioned 0.1 V above the flat potential, whereas in p-type semiconductors, the valence band (E_VB) is typically situated 0.1 V below the flat potential. The E_CB of Fe₂O₃ was established at +0.385 V, whereas the E_VB of ZnBi₂O₄ was computed to be +2.525 V.

The E_VB of Fe₂O₃ and the E_CB of ZnBi₂O₄ were computed using the formula E_VB = E_CB + E_g, where E_g represents the energy bandgap. The band gap values of Fe₂O₃ and ZnBi₂O₄ are 1.95 eV and 2.84 eV, respectively, with the E_VB of Fe₂O₃ estimated at +2.355 V and the E_CB of ZnBi₂O₄ estimated at −0.315 V. The photocatalytic mechanism for the Fe₂O₃@ZnBi₂O₄ nanocomposite is depicted in Fig. 9, based on these analytical and computational findings.

Fig. 9 Hypothesized mechanisms for MB degradation using the FZB catalyst under visible light.

The introduction of p-type ZnBi₂O₄ onto the surface of n-type Fe₂O₃ resulted in the formation of an n–p heterojunction. This configuration induces fluctuations in the energy bands of ZnBi₂O₄ and Fe₂O₃ until achieving Fermi equilibrium. An internal electrostatic field is formed from Fe₂O₃ to ZnBi₂O₄, which facilitates charge separation. Upon exposure to visible light, ZnBi₂O₄ and Fe₂O₃ are stimulated to produce electron–hole pairs (e⁻/h⁺) (reaction (8)). The conduction band level of ZnBi₂O₄ was inferior to that of Fe₂O₃ following the formation of the n–p heterojunction. This indicates that the excited electrons of ZnBi₂O₄ can readily transition into the conduction band (CB) of Fe₂O₃. Conversely, the valence band (VB) level of Fe₂O₃ is more positive than that of ZnBi₂O₄, allowing the excited holes of Fe₂O₃ to transfer to the VB level of ZnBi₂O₄. The internal electric field at the interface between Fe₂O₃ and ZnBi₂O₄ propels these reactions. The internal electric field of the p–n heterojunction efficiently segregates the photogenerated charges.⁶¹ This significantly reduced the recombination of e⁻/h⁺ pairs in the Fe₂O₃/ZnBi₂O₄ heterojunction. Photoluminescence studies, photocurrent response, electrochemical impedance spectroscopy, and radical scavenger investigations validated the capacity for charge separation.

Moreover, the electrons at the heightened conduction band level of ZnBi₂O₄ can reduce the adsorbed O₂ on the surface to form superoxide anion radicals (˙O₂⁻) (reaction (8)), while the holes at the lowered valence band level of Fe₂O₃ can oxidize the adsorbed H₂O on the surface to yield hydroxyl radicals (˙OH) (reaction (9)). Owing to the abundance of electrons at the conduction band level in Fe₂O₃, Fe³⁺ can undergo oxidation to Fe²⁺ (reaction (11)), resulting in the formation of ˙OH radicals (reaction (12)). The radicals exhibiting potent oxidizing characteristics, along with vacancies in the valence band of ZnBi₂O₄ and electrons in the conduction band of Fe₂O₃, contribute to the mineralization of the dye into CO₂, H₂O, and other inorganic byproducts (reaction (19)).

In the presence of H₂O₂, photo-Fenton processes occur, generating additional ˙OH and ˙HO₂⁻ radicals. This accelerated the reaction and enhanced its efficacy for the decomposition of MB. It was specifically the presence of K₂S₂O₈ that sped up the breakdown of MB by creating ˙SO₄⁻ and ˙S₂O₈⁻ radicals, which then increased the amount of ˙OH (reactions (15)–(18)). It is known that the radicals ˙SO₄⁻, ˙S₂O₈⁻, and ˙HO are all strong oxidizing agents. The ˙SO₄⁻ radical strongly reacts with aromatic chemicals. The addition of persulfate significantly accelerated the MB breakdown reaction in the presence of the FZB catalyst and visible light (reaction (20)).

(p–n)FZB + hυ → e⁻ + h⁺ (8)

e⁻ + O₂ → ˙O₂⁻ (9)

h⁺ + H₂O → ˙OH + H⁺ (10)

e⁻ + Fe³⁺ → Fe²⁺ (11)

Fe²⁺ + H₂O + e⁻ → ˙OH + Fe³⁺ + H⁺ (12)

Fe²⁺ + H₂O₂ → Fe³⁺ + ˙HO + OH⁻ (13)

Fe³⁺ + H₂O₂ → H⁺ + Fe²⁺ + ˙HO₂⁻ (14)

Fe³⁺ + S₂O₈²⁻ → ˙S₂O₈⁻ + Fe²⁺ (15)

S₂O₈²⁻ + Fe²⁺ → ˙SO₄⁻ + SO₄²⁻ + Fe³⁺ (16)

S₂O₈²⁻ + e⁻ → ˙SO₄⁻ + SO₄²⁻ (17)

˙SO₄⁻ + H₂O → ˙HO⁻ + SO₄²⁻ + H⁺ (18)

e⁻, h⁺, ˙OH, ˙O₂⁻ + MB → CO₂ + H₂O + inorganic products (19)

e⁻, h⁺, ˙OH, ˙O₂⁻, ˙HO₂⁻ + MB → CO₂ + H₂O + inorganic products (20)

e⁻, h⁺, ˙OH, ˙O₂⁻, ˙SO₄⁻, ˙S₂O₈⁻ + MB → CO₂ + H₂O + inorganic products (21)

4. Conclusion

In conclusion, a novel 3D Fe₂O₃@ZnBi₂O₄ n–p heterojunction was successfully synthesized using two simple methods. The structural, morphological, optical, and electrochemical properties of the photocatalyst were examined. The photocatalytic activity of FZB under visible light was evaluated under various settings. With a molar ratio of 2/1 for Fe₂O₃ and ZnBi₂O₄, FZB had the most photocatalytic activity, 4.2 times that of Fe₂O₃ and 2.8 times that of ZnBi₂O₄. Over 95% of the 50 ppm MB solution broke down after 100 minutes of light exposure. This happened under the best conditions, with a Fe₂O₃/ZnBi₂O₄ molar ratio of 2 : 1, a pH of 9, and a catalyst concentration of 0.7 g L⁻¹. Furthermore, a photocatalytic reaction involving H₂O₂ and persulfate activation of FZB was conducted, resulting in rate constants that increased by 2.15 and 5.95, respectively. The small bandgap energy (E_g = 1.98 eV) of FZB makes it a better photocatalyst because it absorbs UV and visible light with λ ≤ 626 nm. The construction of the FZB n–p heterojunction efficiently segregates photogenerated charges, reducing their recombination rate, and the electric field in the heterojunction boosts the charge transfer. Moreover, the distinctive 3D architecture of FZB can produce a greater number of catalytically active sites. The investigation of the roles of reaction species and the energy band structures of the materials culminated in the proposal of a reaction mechanism.

Data availability

Data for this article, including origin data, are available at https://dx.doi.org/10.6084/m9.figshare.28458698.

Conflicts of interest

The authors declare that they have no competing financial interests or personal relationships that could influence the work reported in this study.

Acknowledgements

This work was supported by the Industrial University of Ho Chi Minh City (IUH), Ho Chi Minh, Vietnam, under grant number 50/HĐ-ĐHCN.

References

1. T. Robinson, G. McMullan, R. Marchant and P. Nigam, Bioresour. Technol., 2001, 77, 247–255.
2. J. J. Rueda-Marquez, I. Levchuk, P. Fernández Ibañez and M. Sillanpää, J. Cleaner Prod., 2020, 258, 120694.
3. G. Ren, H. Han, Y. Wang, S. Liu, J. Zhao, X. Meng and Z. Li, Nanomaterials, 2021, 11, 1804.
4. N. Gaur, D. Dutta, A. Singh, R. Dubey and D. V. Kamboj, Front. Environ. Sci., 2022, 10 DOI:10.3389/fenvs.2022.872514.
5. C. Karthikeyan, P. Arunachalam, K. Ramachandran, A. M. Al-Mayouf and S. Karuppuchamy, J. Alloys Compd., 2020, 828, 154281.
6. N. Alahmadi, Separations, 2022, 9, 264.
7. M. A. Al-Nuaim, A. A. Alwasiti and Z. Y. Shnain, Chem. Pap., 2023, 77, 677–701.
8. Y. Chu, C. Zhao, Y. Zheng, X. Ren, S. Yuan, L. Zhao, Y. Wu and Y. He, J. Environ. Sci., 2025, 154, 444–456.
9. H. Baniamerian, M. Safavi, M. Alvarado-Morales, P. Tsapekos, I. Angelidaki and S. Shokrollahzadeh, Environ. Res., 2018, 166, 497–506.
10. D. C. Ashiegbu and H. J. Potgieter, Heliyon, 2023, 9, e20674.
11. W. Wu, S. Zhang, X. Xiao, J. Zhou, F. Ren, L. Sun and C. Jiang, ACS Appl. Mater. Interfaces, 2012, 4, 3602–3609.
12. M. Tahir and N. A. S. Amin, Appl. Catal., B, 2015, 162, 98–109.
13. H. Kumari, S. Sonia, R. Ranga, S. Chahal, S. Devi, S. Sharma, S. Kumar, P. Kumar, S. Kumar, A. Kumar and R. Parmar, Water, Air, Soil Pollut., 2023, 234, 349.
14. M. Mishra and D. M. Chun, Appl. Catal., A, 2015, 498, 126–141.
15. H. N. Luong, L. Ngoc, T. Nguyen, T. M. Dinh, N. Dan, N. Huynh, L. T. Duy, C. K. Tran and V. Q. Dang, Sci. Technol. Dev. J., 2024, 27, 3453–3462.
16. S. Lubis, Sheilatina and Murisna, J. Phys.:Conf. Ser., 2018, 1116, 042016.
17. C. N. C. Hitam and A. A. Jalil, J. Environ. Manage., 2020, 258, 110050.
18. F. Ge, X. Li, M. Wu, H. Ding and X. Li, RSC Adv., 2022, 12, 8300–8309.
19. T. Xie, Y. Liu, H. Wang and Z. Wu, Sci. Rep., 2019, 9, 1–12.
20. X. Zheng, M. Huang, Y. You, H. Peng and J. Wen, Mater. Res. Bull., 2018, 101, 20–28.
21. N. Mao, Sci. Rep., 2019, 9, 12383.
22. E. C. Pastrana, P. G. Ramos, L. A. Sánchez and J. M. Rodriguez, Thin Films - Deposition Methods and Applications, 2022,  DOI:10.5772/intechopen.105818.
23. H. Baniamerian, M. Safavi, M. Alvarado-Morales, P. Tsapekos, I. Angelidaki and S. Shokrollahzadeh, Environ. Res., 2018, 166, 497–506.
24. A. Bahadoran, S. Masudy-Panah, J. R. De Lile, J. Li, J. J. Gu, B. Sadeghi, S. Ramakrishna and Q. Liu, Int. J. Hydrogen Energy, 2021, 46, 24094–24106.
25. N. T. M. Tho, D. N. N. Khanh, N. Q. Thang, Y. I. Lee and N. T. K. Phuong, Environ. Sci. Pollut. Res., 2020, 27, 11127–11137.
26. R. Nithya and S. Ayyappan, J. Photochem. Photobiol., A, 2020, 398, 112591.
27. A. Bahadoran, S. Masudy-Panah, J. R. De Lile, J. Li, J. J. Gu, B. Sadeghi, S. Ramakrishna and Q. Liu, Int. J. Hydrogen Energy, 2021, 46, 24094–24106.
28. O. Baaloudj, A. A. Assadi, M. Azizi, H. Kenfoud, M. Trari, A. Amrane, A. A. Assadi and N. Nasrallah, Appl. Sci., 2021, 11(9), 3975.
29. M. Qayoom, K. A. Shah, A. H. Pandit, A. Firdous and G. N. Dar, J. Electroceram., 2020, 45, 7–14.
30. M. Hjiri, J. Mater. Sci.: Mater. Electron., 2020, 31, 5025–5031.
31. A. Habibi-Yangjeh, M. Pirhashemi and S. Ghosh, J. Alloys Compd., 2020, 826, 154229.
32. C. Wang, L. Wang and K. Tang, Int. J. Electrochem. Sci., 2013, 8, 4543–4550.
33. X. Liang, J. Qian, Y. Liu, Z. Zhang and D. Gao, RSC Adv., 2020, 10, 29077–29081.
34. S. Singh, G. Hitkari and G. Pandey, Inorg. Nano-Met. Chem., 2018, 48, 477–485.
35. V. A. Mane, D. V. Dake, N. D. Raskar, R. B. Sonpir, E. Stathatos and B. N. Dole, Results Chem., 2023, 6, 101083.
36. T. V. Ha Luu, N. N. Dao, H. A. Le Pham, Q. B. Nguyen, V. C. Nguyen and P. H. Dang, RSC Adv., 2023, 13, 5208–5218.
37. B. D. Viezbicke, S. Patel, B. E. Davis and D. P. Birnie, Phys. Status Solidi B, 2015, 252, 1700–1710.
38. P. Thomas, SB Academic Review, 2016, 18, 108–116.
39. A. M. Djaballah, R. Bagtache and M. Trari, J. Mater. Sci.: Mater. Electron., 2022, 33, 18410–18419.
40. A. Habibi-Yangjeh, M. Pirhashemi and S. Ghosh, J. Alloys Compd., 2020, 826, 154229.
41. A. Bahadoran, S. Masudy-Panah, J. R. De Lile, J. Li, J. Gu, B. Sadeghi, S. Ramakrishna and Q. Liu, Int. J. Hydrogen Energy, 2021, 46, 24094–24106.
42. J. H. Pan and W. I. Lee, Chem. Mater., 2006, 18, 847–853.
43. L. Jiang, J. Wang, X. Wu and G. Zhang, Water, Air, Soil Pollut., 2017, 228(12) DOI:10.1007/s11270-017-3646-4.
44. P. Cai, S. M. Zhou, D. K. Ma, S. N. Liu, W. Chen and S. M. Huang, Nano–Micro Lett., 2015, 7, 183–193.
45. B. Li, C. Lai, G. Zeng, L. Qin, H. Yi, D. Huang, C. Zhou, X. Liu, M. Cheng, P. Xu, C. Zhang, F. Huang and S. Liu, ACS Appl. Mater. Interfaces, 2018, 10, 18824–18836.
46. S. Dong, X. Ding, T. Guo, X. Yue, X. Han and J. Sun, Chem. Eng. J., 2017, 316, 778–789.
47. J. Liu, B. Wang, Z. Li, Z. Wu, K. Zhu, J. Zhuang, Q. Xi, Y. Hou, J. Chen, M. Cong, J. Li, G. Qian and Z. Lin, J. Alloys Compd., 2019, 771, 398–405.
48. R. Ameta, A. K. Chohadia, A. Jain and P. B. Punjabi, Fenton and Photo-Fenton Processes, 2018,  DOI:10.1016/B978-0-12-810499-6.00003-6.
49. K. O'Dowd and S. C. Pillai, J. Environ. Chem. Eng., 2020, 8, 104063.
50. M. Liu, L. Zhang, B. D. Xi, S. Yu, X. Hu and L. A. Hou, RSC Adv., 2017, 7, 51512–51520.
51. V. D. Cao, L. X. Nong, V. H. Nguyen, T. V. Tran, H. T. Vu, C. V. Nguyen and S. T. Do, IOP Conf. Ser.:Mater. Sci. Eng., 2020, 736, 042019.
52. X. Ao and W. Liu, Chem. Eng. J., 2017, 313, 629–637.
53. Q. Zhao, Q. Mao, Y. Zhou, J. Wei, X. Liu, J. Yang, L. Luo, J. Zhang, H. Chen, H. Chen and L. Tang, Chemosphere, 2017, 189, 224–238.
54. U. Ushani, X. Lu, J. Wang, Z. Zhang, J. Dai, Y. Tan, S. Wang, W. Li, C. Niu, T. Cai, N. Wang and G. Zhen, Chem. Eng. J., 2020, 402, 126232.
55. S. Guerra-Rodríguez, E. Rodríguez, D. N. Singh and J. Rodríguez-Chueca, Water, 2018, 10(12), 1828.
56. C. Zhao, J. Wang, X. Chen, Z. Wang, H. Ji, L. Chen, W. Liu and C. C. Wang, Sci. Total Environ., 2021, 752, 141901.
57. K. Tian, F. Shi, M. Cao, Q. Zheng and G. Zhang, Catalysts, 2022, 12(9), 1058.
58. C. Chen, Z. Wu, S. Zheng, L. Wang, X. Niu and J. Fang, Environ. Sci. Technol., 2020, 54(13), 8455–8463.
59. S. Li, H. Shang, Y. Tao, P. Li, H. Pan, Q. Wang, S. Zhang, H. Jia, H. Zhang, J. Cao, B. Zhang, R. Zhang, G. Li, Y. Zhang, D. Zhang and H. Li, Angew. Chem., Int. Ed., 2023, 62(28), e202305538.
60. X. Liu, S. Li, Z. Ren, H. Cao, Q. Yang, Z. Luo, L. He, J. Zhao, Q. Wang, G. Li, X. Liu, L. Chen, H. Li and D. Zhang, Environ. Sci. Technol., 2024, 58, 17797–17806.
61. K. Wang, Z. Guan, X. Liang, S. Song, P. Lu, C. Zhao, L. Yue, Z. Zeng, Y. Wu and Y. He, Ultrason. Sonochem., 2023, 100, 106616.

---