Preparation of Ce Doped BiVO₄ Magnetic Composite and Its Photocatalytic Degradation Performance for Rhodamine B

Abstract

A Ce-doped photocatalytic composite with easy solid–liquid separation capability was prepared and a heterojunction was constructed between BiVO₄ and Fe₃O₄ via a co-precipitation method. A variety of characterization techniques were employed, such as X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR), X-ray photoelectron spectroscopy (XPS), ultraviolet–visible spectroscopy (UV-vis), transmission electron microscopy (TEM) and energy-dispersive X-ray spectroscopy (EDS), as well as other related methods. Its photocatalytic performance for the degradation of Rhodamine B (RhB) was also studied. The results indicate that the photocatalytic efficiency of BiVO₄/Fe₃O₄ is 1.4 times that of the pure BiVO₄ matrix. In particular, the photocatalytic efficiency of Ce_1.5%-BiVO₄/Fe₃O₄ was 2.2 times higher than that of the pure BiVO₄ matrix, and a 100% degradation rate of RhB was achieved within 30 min. The introduction of Fe₃O₄ not only forms a heterojunction with BiVO₄, increasing the active sites and surface oxygen vacancies of the material and effectively suppressing the recombination of photogenerated electron (e^-)-hole (h⁺) pairs, but it also enables the rapid separation of the material from the wastewater solution by the magnetic properties of Fe₃O₄. Additionally, the partial substitution of Ce for Bi in the BiVO₄ lattice reduces the bandgap energy, which enhances the utilization efficiency of visible light and improves the photocatalytic performance of the composite material. The mechanism of RhB degradation by Ce_1.5%-BiVO₄/Fe₃O₄ composite materials is also analyzed in this study. Quenching experiments and EPR tests revealed that h⁺ and ${·\mathrm{O}}_2^{-}$ were the primary reactive species in the degradation process.

1. Introduction

In recent years, with the accelerating pace of urbanization and industrialization, environmental issues such as energy shortages, greenhouse effects and water pollution have become increasingly severe [1,2]. Wastewater from the textile, dyeing, leather and papermaking industries contains substantial amounts of organic dyes, which pose severe threats to the environment if discharged without proper treatment. Therefore, effective treatment of dye wastewater is of paramount importance.

Bismuth vanadate (BiVO₄) is a visible-light-responsive semiconductor material that has attracted extensive research interest in photocatalysis due to its relatively narrow bandgap (~2.4 eV), excellent chemical stability, non-toxicity, and low cost [3,4]. However, recent studies have revealed that pure BiVO₄ suffers from high recombination rates of photogenerated electron (e^-)–hole (h⁺) pairs and low charge migration efficiency, resulting in suboptimal photocatalytic performance [5].

To enhance the photocatalytic activity of BiVO₄, various modification strategies have been employed, including doping with noble metal ions, non-metal ions, rare earth elements and p-type semiconductor materials. For instance, Yang et al. [6] achieved efficient charge carrier separation and transfer by uniformly coating decahedral BiVO₄ with Cu₂O nanospheres and Ag nanoparticles. Jiang et al. [7] prepared F-doped BiVO₄ photocatalysts via a facile one-step alcohol–hydrothermal method using NH₄F as the doping precursor, which exhibited enhanced crystallinity, surface oxygen vacancy density, and photogenerated charge carrier separation efficiency compared with pure-phase BiVO₄. Lu et al. [8] synthesized CeO₂-doped BiVO₄ nanophotocatalysts through a hydrothermal method, leading to the formation of a p-n heterojunction that reduced e^--h⁺ recombination and improved photocatalytic performance.

Additionally, facile recovery of photocatalysts from wastewater represents a critical practical challenge. Previous studies have demonstrated that Fe₃O₄ serves as a versatile support material, functioning not only as a functionalized matrix for metals, organic catalysts, N-heterocyclic carbenes and chiral catalysts, but also as a carrier for homogeneous catalytically active metals such as Pd, Pt, Cu, Ni, Co and Ir, thereby constructing stable heterogeneous catalysts with magnetic recyclability [9,10]. This property is crucial for recovering photocatalysts from wastewater and preventing secondary pollution.

In this work, BiVO₄-based magnetic composite photocatalysts were prepared using a co-precipitation method. First, BiVO₄ was coupled with Fe₃O₄ to form a composite system that not only enables magnetic recovery and reuse after organic wastewater treatment, but also promotes efficient separation of photogenerated electron–hole pairs through heterojunction formation. Subsequently, the material properties were optimized via doping with the rare earth element Ce, which significantly broadened the visible-light response range of BiVO₄ and reduced the recombination rate of photogenerated e^--h⁺ pairs, thereby enhancing photocatalytic degradation performance. The obtained materials were characterized by XRD, FTIR, XPS, UV-Vis, TEM, EDS, PL, Mott–Schottky analysis, EIS, BET, and EPR and their photocatalytic performance for RhB degradation was evaluated. Furthermore, the degradation mechanism was investigated through radical scavenging experiments and EPR tests, and the reusability of the photocatalysts was examined. The results indicate that the as-prepared Ce_1.5%-BiVO₄/Fe₃O₄ nanocomposite features low cost, excellent stability and magnetic separability, and exhibits significantly enhanced photocatalytic performance compared with pure BiVO₄. The catalytic system constructed in this study provides a scalable and feasible solution for water treatment technologies. These findings enrich the research system of Ce_1.5%-BiVO₄/Fe₃O₄ photocatalysts and offer technical support for promoting environmental sustainability.

2. Results and Discussion

2.1. Catalyst Characterization

2.1.1. XRD Analysis

The XRD patterns of the prepared samples are shown in Figure 1. The characteristic diffraction peaks of BiVO₄ appear at 19.1, 29.1, 30.6, 34.6, 35.3, 40.0, 42.6, 46.2, 46.8, 47.4, 50.4 and 53.4°, corresponding to the (011), (121), (040), (200), (002), (211), (051), (132), (240), (042), (202) and (161) crystal planes of BiVO₄, respectively. All diffraction peaks are in excellent agreement with the standard BiVO₄ pattern (JCPDS Card No. 014-0688), and no impurity peaks are detected, confirming the successful synthesis of monoclinic BiVO₄ [11]. The main characteristic peaks of Ce_1.5%–BiVO₄, BiVO₄/Fe₃O₄ and Ce_1.5%–BiVO₄/Fe₃O₄ clearly appear, demonstrating that the original crystal structure of the material is maintained after Ce doping and Fe₃O₄ compositing [12]. The most prominent peaks of BiVO₄/Fe₃O₄ exhibit a slight leftward shift compared with the standard BiVO₄ pattern, likely attributable to lattice distortion induced by heterojunction formation [13]. Upon Ce doping, the XRD peaks shift slightly rightward relative to BiVO₄/Fe₃O₄. This phenomenon can be attributed to the fact that the ionic radius of Ce³⁺(1.14 Å) is slightly smaller than that of Bi³⁺(1.17 Å) and the partial substitution of Ce³⁺ for Bi³⁺ in the BiVO₄ lattice leads to a slight contraction of the BiVO₄ lattice, which proves the successful doping of Ce³⁺ [14]. Due to the relatively weak diffraction intensity of cubic Fe₃O₄, its characteristic peaks are barely observable in the BiVO₄/Fe₃O₄ pattern. The diffraction peak intensity of the samples significantly decreased after Ce doping, as observed in Ce_1.5BiVO₄ and Ce_1.5%-BiVO₄/Fe₃O₄, so this phenomenon is more likely attributed to reduced crystallinity or local disorder induced by Ce doping, leading to an overall decrease in diffraction intensity. Related research needs to be further deepened. In general, moderate reduction in crystallinity is beneficial for increasing surface defects and exposing more active sites to enhance photocatalytic efficiency [15,16,17].

2.1.2. FTIR

The FTIR spectra of BiVO₄, Fe₃O₄, BiVO₄/Fe₃O₄ and Ce_1.5%-BiVO₄/Fe₃O₄ are presented in Figure 2. The characteristic peaks at 3435 cm⁻¹ and 1628 cm⁻¹ correspond to the -OH stretching and bending vibrations of water molecules. The bands at 732 cm⁻¹and 832 cm⁻¹ are attributed to symmetric and asymmetric V-O stretching vibrations in BiVO₄ [18], which are observed in the spectra of BiVO₄, BiVO₄/Fe₃O₄ and Ce_1.5%-BiVO₄/Fe₃O₄. The characteristic Fe-O stretching vibration of Fe₃O₄ at 580 cm⁻¹ is red-shifted to 516 cm⁻¹ in BiVO₄/Fe₃O₄ and Ce_1.5%-BiVO₄/Fe₃O₄, indicating the formation of Fe-O-X chemical bonds in the composite [19,20], which suggests that Fe₃O₄ is homogeneously coupled with BiVO₄. In the spectrum of Ce_1.5%-BiVO₄/Fe₃O₄, the peak at 888 cm⁻¹ corresponds to Ce-O bonds [21,22].

2.1.3. XPS Analysis

Figure 3a displays the Bi 4f high-resolution XPS spectra. The pure BiVO₄ exhibits two distinct peaks at binding energies of 159.1 and 164.4 eV, corresponding to the Bi³⁺ 4f_7/2 and 4f_5/2 orbitals, respectively, which aligns with the reported electronic structure of monoclinic BiVO₄ [4]. Notably, the Bi 4f peak positions in the Ce_1.5%-BiVO₄ sample remain largely unchanged, indicating that Ce doping does not significantly alter the local chemical environment or valence state of Bi. In contrast, the Bi 4f peaks for the BiVO₄/Fe₃O₄ composite shift entirely towards higher binding energies. This positive shift is typically attributed to a decrease in electron cloud density around the Bi atoms, suggesting the presence of strong interfacial interactions between Fe₃O₄ and BiVO₄ that may facilitate electron transfer from BiVO₄ to Fe₃O₄ [23]. The V 2p spectrum (Figure 3b) of pure BiVO₄ shows characteristic peaks at 516.7 and 524.2 eV, assigned to the V⁵⁺ 2p_3/2 and 2p_1/2 states, respectively [24]. Consistent with the trend observed in the Bi 4f spectra, the V 2p peaks in BiVO₄/Fe₃O₄ also exhibit a distinct positive shift, further confirming a reduction in the surface electron density of V atoms [25,26].

Deconvolution of the O 1s spectra (Figure 3c) reveals that the spectrum for pure BiVO₄ can be fitted with two peaks at 529.8 and 531.5 eV. The O 1s spectrum of Ce_1.5%-BiVO₄ is fitted with peaks at 529.8 and 530.6 eV, where the peak at 530.6 eV is commonly associated with oxygen vacancies (O_Vs) or low-coordinated oxygen species. The O 1s spectrum of BiVO₄/Fe₃O₄ exhibits a more complex structure, which can be deconvoluted into three sub-peaks at 530.1 eV (lattice oxygen, Bi-O), 531.0 eV (oxygen vacancies, O_Vs), and 532.4 eV (chemisorbed oxygen species, such as -OH) [27]. Comparing the relative area of the peaks corresponding to oxygen vacancies reveals that the O_Vs signals for both Ce_1.5%-BiVO₄ and BiVO₄/Fe₃O₄ are significantly stronger than that of pure BiVO₄. This indicates that both Ce doping and the formation of the BiVO₄/Fe₃O₄ composite induce a higher density of oxygen vacancies [28].

The Ce 3d spectrum (Figure 3d) reveals peaks at 881.7 and 885.5 eV for Ce³⁺, and at 900.1 and 904.2 eV for Ce⁴⁺ [29,30,31], confirming the coexistence of Ce³⁺ and Ce⁴⁺ species. The Fe 2p spectrum (Figure 3e) displays broadened peaks at 711.4 and 724.2 eV, consistent with the coexistence of Fe²⁺ and Fe³⁺ in Fe₃O₄ [32]. The survey spectrum (Figure 3f) clearly identifies signals from Bi, V, O, Ce, and Fe, corroborating successful composite formation.

2.1.4. UV-Vis Diffuse Reflectance Spectroscopy and Mott–Schottky Plots

The UV-Vis diffuse reflectance spectra are shown in Figure 4a. All samples exhibit absorption in the ultraviolet region (200–500 nm). Both BiVO₄/Fe₃O₄ and Ce_1.5%-BiVO₄/Fe₃O₄ show further red-shifted absorption edges extending to 750 nm, substantially enhancing visible light absorption and providing a reliable basis for improved photocatalytic performance. The band positions of BiVO₄, Ce_1.5%-BiVO₄, Fe₃O₄, BiVO₄/Fe₃O₄, and Ce_1.5%-BiVO₄/Fe₃O₄ were determined via UV-Vis diffuse reflectance spectra and Mott–Schottky curves (Figure 4d). The bandgap (E_g) (an effective optical response of the composite) of the synthesized photocatalysts is typically estimated using the Kubelka–Munk method, which is expressed by the equation (αhν)ⁿ = A (hν-E_g). Herein, α, h, ν, and A denote the absorption coefficient, Planck constant, light frequency, and a constant, respectively. The value of n depends on the transition characteristics of the semiconductor, and E_g can be derived by plotting (αhν)ⁿ against it. Since BiVO₄ is a direct allowed transition semiconductor [33], n = 1/2 is adopted. Plotting (αhν)^1/2 versus hν (Figure 4b) yields a bandgap of approximately 2.24 eV for BiVO₄. Fe₃O₄ is a metal oxide, so n = 2 is used [34]. By plotting (αhν)^1/2 versus hν, the bandgap of Fe₃O₄ is estimated to be about 2.55 eV (Figure 4c). The calculated bandgap values (Figure 4b) are 2.00 eV for Ce_1.5%-BiVO₄, 1.80 eV for BiVO₄/Fe₃O₄, and 1.65 eV for Ce_1.5%-BiVO₄/Fe₃O₄. The Ce_1.5%-BiVO₄/Fe₃O₄ composite possesses the smallest bandgap (1.65 eV), which requires the lowest excitation energy for electron transition from the valence band (VB) to the conduction band (CB), thereby maximizing the visible light response and utilization efficiency. The bandgap value of 1.65 eV for Ce_1.5%-BiVO₄/Fe₃O₄ is unusually low and it should be considered as an apparent/effective optical response of the composite, rather than a conventional intrinsic bandgap. The strong narrowing of the bandgap can be attributed to the chemical bonds formed between BiVO₄ and Ce species [35]. Enhancing visible light utilization is highly beneficial for improving the photocatalytic efficiency of the catalysts [36].

The Mott–Schottky plots in Figure 4d show that the flat-band potentials (V_FB) of the BiVO₄, BiVO₄/Fe₃O₄, and Ce_1.5%-BiVO₄/Fe₃O₄ samples are 0.27, 0.44, and 0.62 V (vs. Ag/AgCl), respectively. Compared with BiVO₄, the flat-band potentials of BiVO₄/Fe₃O₄ and Ce_1.5%-BiVO₄/Fe₃O₄ exhibit a significant positive shift, indicating reduced band bending and a lower Schottky barrier. This can be attributed to the decreased Fermi level, which enhances the photogenerated electron transfer efficiency of the composite photocatalysts [37]. The even more positive flat-band potential observed for the Ce-doped composite indicates that Ce incorporation may significantly modify the band structure of BiVO₄. Ce ions, with their unique 4f electronic configuration, act as electron acceptors in the BiVO₄ lattice, introducing localized defect states near the conduction band. These defect states may trap photogenerated electrons, further reducing the electron density in the surface region of BiVO₄ and leading to a more pronounced positive shift in the flat-band potential. In addition, the Mott–Schottky curves measured at frequencies of 1500 Hz and 2000 Hz display positive slopes for both BiVO₄ and Fe₃O₄, confirming that both are n-type semiconductors [38]. The flat-band potentials (vs. Ag/AgCl) of BiVO₄ and Fe₃O₄ are determined to be 0.27 and −0.88 V, respectively, from the Mott–Schottky plots. Typically, the CB potential (vs. NHE, pH = 7) is approximately 0.2 eV more negative than the flat-band potential. Based on Equation (1), the CB potentials of BiVO₄ and Fe₃O₄ are calculated to be 0.47 and −0.68 V (vs. NHE, pH = 7), respectively. Furthermore, according to Equation (2), the corresponding VB potentials are determined to be +2.71 and +1.87 V (vs. NHE, pH = 7).

$$E_{\mathrm{NHE}, \mathrm{pH}=7}=E_{\text{Ag/AgCl}}+0.197$$

(1)

$$E_{\mathrm{g}}=E_{\mathrm{VB}}-E_{\mathrm{CB}}$$

(2)

2.1.5. PL and EIS Nyquist Analysis

The separation efficiency of photogenerated electron–hole pairs was characterized by photoluminescence (PL) spectroscopy. The PL emission peaks originate from the recombination of photogenerated carriers, and their intensity directly reflects the separation efficiency of electron–hole pairs. Generally, a lower fluorescence intensity indicates greater suppression of carrier recombination and a higher separation efficiency [39]. As shown in Figure 5a, under excitation at 375 nm, BiVO₄, Ce_1.5%-BiVO₄, BiVO₄/Fe₃O₄ and Ce_1.5%-BiVO₄/Fe₃O₄ exhibit a distinct emission peak at approximately 508 nm, which is attributed to the recombination of photoexcited electrons and holes. Notably, the PL intensity of Ce_1.5%-BiVO₄/Fe₃O₄ is significantly lower than that of pure BiVO₄, indicating that the introduction of Ce effectively suppresses the recombination of photogenerated carriers and enhances the separation of electron–hole pairs [40]. Meanwhile, Fe₃O₄ possesses favorable electronic conductivity, which facilitates the migration of photogenerated carriers [41]. Therefore, the improved charge separation and migration efficiency synergistically enhance the photocatalytic performance of the composite.

Electrochemical impedance spectroscopy (EIS) is commonly employed to investigate the migration resistance and separation efficiency of photogenerated charge carriers in semiconductors. The interfacial resistance at the electrode surface is reflected by the arc radius in the Nyquist plots. Generally, a smaller interfacial resistance leads to higher charge transfer efficiency, which corresponds to superior material performance [42]. As shown in Figure 5b, the Nyquist plots of all composite samples exhibit a semicircular arc in the high-frequency region. Compared with pure BiVO₄ and pure Fe₃O₄, the BiVO₄/Fe₃O₄ and Ce_1.5%-BiVO₄/Fe₃O₄ samples display smaller arc radii. Among them, the Ce_1.5%-BiVO₄/Fe₃O₄ composite exhibits the smallest arc radius, indicating that this system possesses the lowest migration resistance for photogenerated charge carriers, thereby effectively suppressing their recombination. The above photoelectrochemical analysis reveals that the introduction of an appropriate amount of Ce and Fe₃O₄ significantly increases the number of photogenerated charge carriers.

2.1.6. TEM Analysis

Figure 6 presents the TEM images of BiVO₄, Fe₃O₄, BiVO₄/Fe₃O₄ (Figure 6a–c) and high-angle annular dark-field (HAADF) imaging with elemental mapping of Ce_1.5%-BiVO₄/Fe₃O₄ (Figure 6d–j), which reveal uniform distribution of O, V, Fe, Ce, and Bi elements, confirming successful doping of Ce into the BiVO₄ matrix. High-resolution TEM images (Figure 6k–l) show lattice fringes with interplanar spacings of 0.253 nm, corresponding to the (311) plane of cubic Fe₃O₄ [43], and 0.300 nm, corresponding to the (121) plane of monoclinic BiVO₄. This slight decrease in interplanar spacing may result from lattice contraction induced by Ce doping [44,45]. The clear interfacial region between BiVO₄ and Fe₃O₄ demonstrates heterojunction formation, which is beneficial for photocatalytic enhancement [28].

The EDS spectra of BiVO₄, BiVO₄/Fe₃O₄, and Ce_1.5%-BiVO₄/Fe₃O₄ (Figure 7) confirm the presence of Bi, V, and O in BiVO₄; Bi, V, O, and Fe in BiVO₄/Fe₃O₄; and Bi, V, O, Fe, and Ce in Ce_1.5%-BiVO₄/Fe₃O₄, further validating successful material synthesis.

Table 1. The mass fraction of the composite elements in BiVO₄, BiVO₄/Fe₃O₄ and Ce_1.5%-BiVO₄/Fe₃O₄ samples obtained from EDS mapping.

Sample Elements	O (%)	V (%)	Fe (%)	Ce (%)	Bi (%)
BiVO4	18.41	17.76	-	-	63.83
BiVO4/Fe3O4	12.80	12.21	3.38	-	71.62
Ce1.5%-BiVO4/Fe3O4	13.58	28.17	24.08	4.13	30.04

shows the semi-qualitative analysis data of each element in the elemental mapping images of BiVO₄, BiVO₄/Fe₃O₄, and Ce_1.5%-BiVO₄/Fe₃O₄ samples. It summarizes the mass fractions of key elements (O, V, Fe, Ce, Bi) in BiVO₄, BiVO₄/Fe₃O₄, and Ce_1.5%-BiVO₄/Fe₃O₄ samples prepared by the co-precipitation method, with the data obtained from EDS mapping analysis. The data presented in Table 1 were acquired from random regions of different samples during characterization, and are affected by sampling depth and local inhomogeneity. Although they can reflect the inhomogeneous elemental distribution in the corresponding regions, the elemental contents among different samples are not directly comparable. Therefore, these semi-quantitative results can only confirm the successful incorporation of the elements, but cannot fully represent the true proportions of each element in the samples.

2.1.7. BET Analysis

Figure 8 shows the N₂ adsorption–desorption isotherms and pore size distribution curves of pure BiVO₄ and Ce_1.5%-BiVO₄/Fe₃O₄. N₂ adsorption–desorption measurements were performed to investigate the effects of Ce doping and Fe₃O₄ composite modification on the specific surface area and pore structure of BiVO₄. The Barrett Joyner Halenda (BJH) model was applied to analyze the pore size distribution of the samples. As displayed in Figure 8a,b, the Ce_1.5%-BiVO₄/Fe₃O₄ composite exhibits a typical Type IV isotherm, confirming the existence of a mesoporous structure. The specific surface areas of BiVO₄ and Ce_1.5%-BiVO₄/Fe₃O₄ are 4.78 and 14.63 m²/g, while their average pore diameters are 6.5290 and 6.0273 nm, respectively. Compared with the pure material, the larger specific surface area of the composite effectively enlarges the contact interface between the catalyst and reaction substrates and provides more active sites for the photocatalytic reaction [46].

2.2. Photocatalytic Performance

2.2.1. Photocatalytic Activity

Figure 9a displays the adsorption and photodegradation curves of RhB with and without catalysts. It can be observed that all samples basically reach the adsorption equilibrium within 45 min. The degradation efficiency first increases and then decreases with increasing Ce content. Ce_1.5%-BiVO₄/Fe₃O₄ exhibits the optimal performance, achieving 100% RhB removal within 30 min. Within the same 30 min period, the degradation rates of RhB using BiVO₄, BiVO₄/Fe₃O₄, Ce_1.5%-BiVO₄, Ce_1.0–BiVO₄/Fe₃O₄, and Ce_2.0–BiVO₄/Fe₃O₄ as photocatalysts were only 37.6%, 75.2%, 85.3%, 81.3%, and 83.9%, respectively. Meanwhile, Figure 9a compares the degradation efficiencies of BiVO₄, BiVO₄/Fe₃O₄, Ce_1.5%-BiVO₄, and Ce_1.5%-BiVO₄/Fe₃O₄, which require 65, 45, 40, and 30 min for complete RhB degradation respectively. Fe₃O₄ exhibited negligible photocatalytic activity toward RhB degradation under visible light irradiation. The modified composites demonstrate significant performance enhancement, with BiVO₄/Fe₃O₄, Ce_1.5%-BiVO₄ and Ce_1.5%-BiVO₄/Fe₃O₄ showing 1.4-fold, 1.6-fold, and 2.2-fold improvements in degradation efficiency compared with pure BiVO₄. Heterojunction formation between BiVO₄ and Fe₃O₄ promotes photogenerated charge separation and suppresses recombination. Ce doping further enhances performance. This is mainly due to the fact that when an appropriate amount of Ce is doped into BiVO₄/Fe₃O₄, the Ce⁴⁺ ions capture some electrons and reduce to Ce³⁺ ions, thereby preventing the recombination of e^– and h⁺ in BiVO₄/Fe₃O₄ photocatalysts and improving the photocatalytic degradation efficiency. In addition, Ce ions possess a rich energy level structure, which narrows the bandgap of BiVO₄/Fe₃O₄ photocatalysts and enhances visible light absorption (as shown in Figure 4b). However, excessive Ce doping introduces defects that act as recombination centers, deteriorating photocatalytic activity. The reaction rate constants (Figure 9b) are 0.028, 0.056, 0.064, and 0.079 min⁻¹ for BiVO₄, BiVO₄/Fe₃O₄, Ce_1.5%-BiVO₄ and Ce_1.5%-BiVO₄/Fe₃O₄, respectively. As shown in

Table 2. Goodness-of-fit indicators (R² values).

Sample	BiVO4	BiVO4/Fe3O4	Ce1.5%-BiVO4	Ce1.5%-BiVO4/Fe3O4
R2	0.9472	0.9682	0.9503	0.9900
Adjusted R2	0.9367	0.9602	0.9421	0.9874
Kα/(min−1)	0.0278	0.056	0.064	0.079

, the goodness-of-fit indicators (R² values) of the degradation models for BiVO₄, BiVO₄/Fe₃O₄, Ce_1.5%-BiVO₄, and Ce_1.5%-BiVO₄/Fe₃O₄ are 0.95, 0.97, 0.95, and 0.99, respectively, all of which are greater than 0.9, indicating an excellent model fit [47]. The adjusted R² values are 0.94, 0.96, 0.94, and 0.99, respectively, with corresponding differences of 0.01, 0.01, 0.01, and 0.00, all less than 0.1, demonstrating that the models are reasonable and reliable [48]. Meanwhile, total organic carbon (TOC) analysis was performed on Ce_1.5%-BiVO₄/Fe₃O₄. After 30 min of visible light irradiation, the TOC removal efficiency of the target pollutant solution reached 90.20%, verifying its promising application potential in wastewater treatment. As summarized in

Table 3. Degradation performance of RhB over different photocatalysts.

Catalyst	RhB Concentration (mg/L)	ct/c0	Degradation Time (min)	Reference
BiVO4-Ce	5	0.8	240	[49]
BiVO4/Fe3O4	10	0	120	[50]
Ce1.5%-BiVO4/Fe3O4	10	0	30	This work

, the photocatalytic degradation performance of various reference catalysts toward RhB was investigated in this study, all of which exhibited relatively limited activity. This table is only for general comparison with existing literature, rather than a one-to-one quantitative benchmark comparison under identical experimental conditions. From Table 3, the Ce_1.5%-BiVO₄/Fe₃O₄ composite prepared in this work demonstrated the superior photocatalytic activity among all tested catalysts, with the detailed performance parameters presented in the table.

2.2.2. Magnetic Separation Performance

The magnetic separability of Ce_1.5%-BiVO₄/Fe₃O₄ was evaluated. Figure 9c shows the uniform dispersion of the catalyst in solution, while Figure 9d presents a separation effect diagram of sample and solution. The results indicate that under the influence of a magnet, the solution became clear within 1 min, achieving rapid separation of the Ce_1.5%-BiVO₄/Fe₃O₄ sample after degrading RhB from the aqueous phase. Therefore, Ce_1.5%-BiVO₄/Fe₃O₄ possesses excellent paramagnetic properties, which can be utilized to achieve rapid separation and recycling of Ce_1.5%-BiVO₄/Fe₃O₄ from the aqueous phase.

2.2.3. Reusability Performance

The stability of Ce_1.5%-BiVO₄/Fe₃O₄ was assessed through five consecutive cycles of RhB degradation (30 min per cycle). As shown in Figure 10a, the RhB removal rate gradually decreases with cycling but remains above 90% after five cycles, demonstrating excellent reusability and commercial potential [51,52].

As shown in Figure 10b,c, for the recycled Ce_1.5%-BiVO₄/Fe₃O₄ composite after cyclic degradation experiments, no obvious shift in the XRD characteristic diffraction peaks is observed, and no significant difference appears in the FTIR spectra. This confirms that the crystal structure of the material remains stable after cycling tests. However, the intensity of the diffraction peaks decreased, indicating that a small number of defects may form inside the material after recycling, which may lead to a slight reduction in photocatalytic efficiency. XPS analysis was conducted to further investigate the interfacial behavior and variations in the chemical states of elements within the Ce_1.5%-BiVO₄/Fe₃O₄ composite. In the Bi 4f and V 2p spectra, the characteristic peaks of the Ce_1.5%-BiVO₄/Fe₃O₄ composite shifted toward higher binding energies by approximately 0.2 and 0.4 eV, respectively. Similarly, the O_L peak in the O 1s spectrum exhibited a notable shift to a higher binding energy (Figure 10d–h). In contrast, the Fe 2p_3/2 spectrum was deconvoluted into two components corresponding to Fe²⁺ and Fe³⁺ species, with their binding energies shifting to lower values by about 0.6 and 0.1 eV, respectively, indicating electron enrichment around the iron atoms (Figure 10g). Nevertheless, such a shift may also arise from other factors as the test is based on ex situ measurements, so further investigation is required to verify the conclusions. Moreover, upon light irradiation, the proportion of Fe³⁺ species decreased from 28.13% to 24.87%, while that of Fe²⁺ increased from 71.87% to 75.13% [28]. A similar trend was observed for the Ce species, with the Ce⁴⁺ ratio decreasing from 38.11% to 36.78% and the Ce³⁺ ratio increasing from 61.89% to 63.21% (Figure 10h). This may be attributed to the fact that under visible light irradiation, Fe³⁺ and Ce⁴⁺ in the system can effectively capture photogenerated electrons and be reduced to Fe²⁺ and Ce³⁺, thereby significantly enhancing the overall charge separation efficiency at the BiVO₄/Fe₃O₄ interface.

2.3. Degradation Mechanism

Radical scavenging experiments were conducted to identify the primary active species in the Ce_1.5%-BiVO₄/Fe₃O₄ system. Potassium iodide (KI), isopropanol (IPA), and benzoquinone (BQ) were employed as scavengers for holes (h⁺), hydroxyl radicals (·OH) and superoxide radicals (${·\mathrm{O}}_2^{-}$), respectively [53]. As shown in Figure 11a, the addition of IPA causes only a 3.06% decrease in degradation efficiency (96.94% retention), whereas KI and BQ addition reduces efficiency to 54.89% and 49.46%, respectively. These results confirm that h⁺ and ${·\mathrm{O}}_2^{-}$ are the predominant reactive species, while ·OH plays a weak role. EPR tests were performed using DMPO and TEMP as spin-trapping agents (Figure 11b–d) to further identify the reactive oxygen species generated in the Ce_1.5%-BiVO₄/Fe₃O₄ photocatalytic system. Compared with the smooth baseline measured in the dark, the characteristic sextet signal of ${·\mathrm{O}}_2^{-}$ was significantly enhanced under photoexcitation (Figure 11b), confirming the abundant production of ${·\mathrm{O}}_2^{-}$ radicals. No obvious characteristic peaks were observed under dark conditions, while a weak signal corresponding to ·OH was detected upon light irradiation (Figure 11c), indicating that ·OH participated in the degradation process. Meanwhile, the 1:1:1 triplet signal intensity of h⁺ under illumination was remarkably lower than that in the dark (Figure 11d), suggesting that h⁺ was continuously consumed during the photocatalytic reaction. These EPR results were consistent with those of radical quenching experiments, further verifying that the active species produced by Ce_1.5%-BiVO₄/Fe₃O₄ played a crucial role in the photocatalytic degradation of RhB [45,54].

There are two possible reaction mechanisms for the photocatalytic degradation of RhB by Ce_1.5%-BiVO₄/Fe₃O₄ under visible light irradiation. According to the traditional type-II photocatalytic mechanism, BiVO₄ and Fe₃O₄ are simultaneously excited under visible light irradiation. Subsequently, a large number of electrons jump from VB to CB within the same semiconductor, leaving h⁺ in the VB. CB potential of Fe₃O₄ (E_g = −0.68 eV vs. NHE) is more negative than that of BiVO₄ (E_g = +0.47 eV vs. NHE), so the electrons in the CB of Fe₃O₄ can jump to the CB of BiVO₄. In contrast, VB potential of BiVO₄ (E_g = +2.71 eV vs. NHE) is more positive than that of Fe₃O₄ (E_g = +1.87 eV vs. NHE), which allows the migration of h⁺ from BiVO₄ to Fe₃O₄. However, the type-II photocatalytic mechanism cannot explain the generation of ·OH and ${·\mathrm{O}}_2^{-}$ radicals, because the VB potential of Fe₃O₄ cannot reach the potential of H₂O/·OH (+1.99 eV), and the CB potential of BiVO₄ also cannot reach the potential of O₂/${·\mathrm{O}}_2^{-}$ (−0.33 eV). Based on these findings from Mott–Schottky analysis, UV-Vis, radical trapping, XPS and EPR, this study speculates that the most plausible charge-transfer pathway of Ce-BiVO₄/Fe₃O₄ photocatalyst follows a Z-scheme charge transfer mechanism [55]. The detailed mechanism of electron transfer remains to be further investigated due to current limitations in research techniques. Based on the available data, the photocatalytic mechanism is speculated as illustrated in Figure 12. Under visible light irradiation, electrons are excited from the VB to CB in both BiVO₄ and Fe₃O₄, generating holes in the VB (Equation 3). Owing to the conductivity of Fe₃O₄, electrons in the CB of BiVO₄ (E_g = +0.47 eV νs. NHE) migrate to the VB of Fe₃O₄ (E_g = +1.87 eV νs. NHE) at the heterojunction interface through a Z-scheme charge transfer pathway [55] (Equation (4)), where electron–hole recombination occurs, prolonging electron lifetime and promoting charge separation. Subsequently, electrons in the CB of Fe₃O₄ (E_g = −0.68 eV νs. NHE) reduce Fe³⁺ to Fe²⁺ (Equation 5), and Fe²⁺ reacts with adsorbed O₂ to generate ${·\mathrm{O}}_2^{-}$ (−0.33 eV) (Equation 6) [53]. Meanwhile, some electrons are captured by Ce⁴⁺ to form Ce³⁺ (Equation 7), preventing electron–hole recombination. The Ce³⁺ species can also react with adsorbed oxygen to produce ${·\mathrm{O}}_2^{-}$ (Equation 8). Simultaneously, holes in the VB of BiVO₄ (E_g = +2.71 eV νs. NHE) directly oxidize RhB molecules. Ultimately, RhB is decomposed by h⁺ (Equation 9) and ${·\mathrm{O}}_2^{-}$ (Equation 10). It should be noted that Figure 12 is merely a schematic based on the band structures of pure-phase BiVO₄ and pure-phase Fe₃O₄, and does not fully represent the intrinsic true energy level structure of the composite photocatalyst. Ce doping introduces localized energy levels and lattice distortions that alter the band structure of BiVO₄, while the heterojunction interface between BiVO₄ and Fe₃O₄ induces band bending. These effects, which are not captured by pure-phase data, mean the diagram cannot fully reflect the actual electronic structure of the Ce_1.5%-BiVO₄/Fe₃O₄ ternary composite. Compared with the degradation mechanism of RhB reported by Luo et al. [49] for the Ce-doped BiVO₄ system, the mechanism proposed herein exhibits distinct characteristics. In the Ce-BiVO₄ system, the photocatalytic activity primarily relies on Ce⁴⁺ acting as an electron trap to form Ce³⁺, which subsequently promotes the generation of ·OH as the dominant reactive species. Furthermore, the separation efficiency in that system is modulated by the doping concentration through its influence on the space-charge region potential [49]. In contrast, the BiVO₄/Fe₃O₄ composite developed in this work employs a Z-scheme heterojunction to facilitate interfacial charge separation. Coupled with the synergistic effect between the Fe³⁺/Fe²⁺ redox cycle and the Ce⁴⁺/Ce³⁺ transition, the present system utilizes ·O₂^- and h⁺ as the main active species for direct pollutant degradation. Moreover, the role of Ce in this composite extends beyond simple electron trapping to include band-structure modulation, forming multiple charge-transfer channels with Fe₃O₄ that collectively enhance carrier separation and the generation of reactive species. This integrated “heterojunction-doping” synergy thus establishes a more efficient and robust photocatalytic mechanism.

Photocatalyst + hν → e^- + h⁺

(3)

e_CB^-_(BiVO4) + h_VB⁺_(Fe3O4) → recombination

(4)

e^- + Fe³⁺ → Fe²⁺

(5)

$${\mathrm{Fe}}^{2+}+{\mathrm{O}}_2\to {·\mathrm{O}}_2^{-}{\mathrm{Fe}}^{3+}$$

(6)

Ce⁴⁺ + e^- → Ce³⁺

(7)

$${\mathrm{Ce}}^{3+}+{\mathrm{O}}_2\to {·\mathrm{O}}_2^{-}{\mathrm{Ce}}^{4+}$$

(8)

RhB + h⁺ → degradation products

(9)

$$\mathrm{RhB}+{·\mathrm{O}}_2^{-}\to \mathrm{degradation}\mathrm{products}$$

(10)

3. Materials and Methods

3.1. Reagents and Instruments

Sodium metavanadate (NaVO₃, 99.0%), bismuth chloride (BiCl₃), cerium nitrate hexahydrate (Ce(NO₃)₃·6H₂O, 99.95%), nano-sized iron(II,III) oxide (Fe₃O₄, 99.5%), ammonium hydroxide (NH₃·H₂O, 25.0–28.0%) and concentrated hydrochloric acid (HCl, 36.0–38.0%) were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). All reagents were of analytical grade and used as received without further purification.

Instruments used: D8 Advance X-ray diffractometer (XRD, Bruker, Billerfeld, Germany); Nicolet 6700 Fourier transform infrared spectrometer (FTIR, Thermo Fisher Scientific, Waltham, MA, USA); ESCALAB XI+ X-ray photoelectron spectrometer (XPS, Thermo Fisher Scientific, Waltham, MA, USA); Cary 5000 UV-Vis spectrophotometer (Agilent Technologies, Santa Clara, CA, USA); JEM-F200 transmission electron microscope (TEM, JEOL, Tokyo Metropolis, Tokyo, Japan). Photoluminescence emission (PL) spectra of the samples were detected using a fluorescence spectrophotometer (F97Pro, Shanghai Lengguang Technology Co., Ltd., Shanghai, China) equipped with a 450 W Xe lamp as the excitation source. Electrochemical impedance spectroscopy (EIS) measurements were performed using Na₂SO₄ (0.3 mol/L) with frequency ranging between 0.1 and 106 Hz and an amplitude of 0.01 V. The Mott–Schottky curves were obtained by an electrochemical workstation (CHI660E, Shanghai Chenhua Instruments Co., Ltd., Shanghai, China) with a potential range of −0.5 to 1 V and a test frequency of 3000 Hz. The surface area and pore size were measured by a surface area and porosity analyzer (ASAP2460-2, Mike Company, Norcross, GA, USA).

Electron Paramagnetic Resonance (EPR)

All EPR spectra were recorded at room temperature using a Bruker EMXnano spectrometer (USA), with specific conditions for each active species as follows:

${\mathrm{O}}_2^{-}$: A total of 10 mg of catalyst was dispersed in 1 mL of methanol containing 50 mM DMPO. The mixture was irradiated with visible light (300 W xenon lamp, λ > 420 nm) for 10 min, and the spectrum was recorded immediately. A characteristic six-line spectrum of the DMPO-${·\mathrm{O}}_2^{-}$ adduct was observed (a(N) ≈ 1.5 mT, a(H) ≈ 1.1 mT).

OH: A total of 10 mg of catalyst was dispersed in 1 mL of deionized water containing 50 mM DMPO. The mixture was irradiated following the same method described above. A characteristic four-line spectrum of the DMPO- ·OH adduct was observed (a(N) ≈ a(H) ≈ 1.5 mT).

h⁺: A total of 10 mg of catalyst was dispersed in 1 mL of deionized water containing 0.1 mM TEMPO. EPR spectra were recorded before visible light irradiation (dark state—the mixture was kept in the dark for 5 min, and then the EPR spectrum was recorded to establish the baseline signal intensity of TEMPO) and after 20 min of irradiation (irradiated with visible light from a 300 W xenon lamp with λ > 420 nm for 20 min, and the spectrum was recorded immediately afterward), respectively. Photogenerated holes oxidize TEMPO (1:1:1 triplet, g ≈ 2.006) into an EPR-silent oxoammonium cation, resulting in a decrease in signal intensity under light irradiation compared with the dark state.

3.2. Photocatalyst Preparation

3.2.1. Synthesis of BiVO₄

A total of 2 mmol of BiCl₃ and 2 mmol of NaVO₃ were dissolved separately in 1.0 mL of 6 mol·L⁻¹ dilute HCl and 19.0 mL of deionized water, respectively. Under ultrasonication, the bismuth chloride solution was added dropwise to the sodium metavanadate solution. The pH was adjusted to 8–9 with ammonia water. Subsequently, the mixture was mechanically stirred continuously at 500–600 r·min⁻¹ at 80 °C for 0.5 h. After completion of the reaction, the resulting product was centrifuged at 4000 r·min⁻¹ for 10 min at room temperature. The supernatant was discarded and the solid product was collected. The solid was washed with deionized water and anhydrous ethanol three times each, followed by centrifugation at 4000 r·min⁻¹ for 10 min after each washing step. The final product was vacuum-dried at 80 °C for 12 h. The dried sample was ground and calcined in a muffle furnace at 400 °C for 2 h (heating rate: 5 °C·min⁻¹).

3.2.2. Synthesis of Ce-BiVO₄, BiVO₄/Fe₃O₄ and Ce-BiVO₄/Fe₃O₄ Composites

The preparation procedure is illustrated in Scheme 1. A total of 2 mmol of BiCl₃ and 2 mmol of NaVO₃ were dissolved separately in 1.0 mL of HCl (6mol·L⁻¹) and 19.0 mL of deionized water. To the sodium metavanadate solution, 0.0648 g of Fe₃O₄ was added (Fe₃O₄:BiVO₄ mass ratio = 10%). For cerium-doped samples, Ce(NO₃)₃·6H₂O was added at Ce:V molar ratios of 0.5%, 1.0%, 1.5% and 2.0%, respectively. Under ultrasonic irradiation, the bismuth chloride solution was added dropwise into the sodium metavanadate solution, and the pH was adjusted to 8–9 with ammonia water. The mixture was reacted at 80 °C for 0.5 h under mechanical stirring at 500–600 r·min⁻¹. After the reaction, the product was centrifuged at 4000 r·min⁻¹ for 10 min at room temperature. The solid product was washed three times with deionized water and anhydrous ethanol, respectively, followed by centrifugation at 4000 r·min⁻¹ for 10 min after each washing step. The final product was vacuum-dried at 80 °C for 12 h, ground, and then calcined at 400 °C for 2 h with a heating rate of 5 °C·min⁻¹.

3.3. Photocatalytic Degradation Experiments

A 10.00 mg·L⁻¹ rhodamine B (RhB) solution was prepared, and its initial absorbance was measured using a UV–visible spectrophotometer at λ= 554 nm. Then, 0.1000 g of photocatalyst was added to 100 mL of the RhB solution with a concentration of 10.00 mg·L⁻¹. The suspension was magnetically stirred at 400–500 r·min⁻¹ for 45 min at room temperature in the dark and without aeration or oxygen supply to achieve adsorption–desorption equilibrium. Subsequently, the photocatalytic reaction was carried out under visible light irradiation using a 300 W xenon lamp (the light intensity was maintained at 600 mW/cm²) with optical filters, at a distance of approximately 10 cm from the liquid surface. Cooling water was continuously circulated to maintain room temperature. Samples were periodically collected, separated and measured until complete RhB degradation was achieved. After each photocatalytic test, BiVO₄/Fe₃O₄ and Ce_1.5%-BiVO₄/Fe₃O₄ were magnetically recovered for cyclic reusability evaluation.

The degradation efficiency was calculated using Equation (11):

$$RhB removal rate={\displaystyle \frac{C_0-C_{\mathrm{t}}}{C_0}}\times 100\%$$

(11)

c₀ and c_t represent the initial and instantaneous concentrations of RhB.

The kinetic constants of the samples were calculated via the Langmuir–Hinshelwood model.

$$r=-{\displaystyle \frac{dc}{dt}}=kK{c}_{\mathrm{t}}+K{c}_0$$

(12)

In the formula, r is the reaction rate, c₀ is the initial concentration of the reactant, c_t is the reactant concentration at time t, t is the illumination time, K is the reaction rate constant under the given experimental conditions, and k is the adsorption constant of the reactant.

In the experiments in this paper, the initial concentration is relatively low, and Kc₀ can be neglected. Therefore, the model is simplified to a first-order equation:

$$r=-{\displaystyle \frac{dc}{dt}} = kK{c}_{\mathrm{t}}= K_{\alpha }{c}_{\mathrm{t}}$$

(13)

i.e.,

$$-\ln \left({\displaystyle \frac{C_{\mathrm{t}}}{C_0}}\right)=K_{\alpha }t$$

(14)

where K_α is the first-order reaction rate constant.

4. Conclusions

In this study, BiVO₄/Fe₃O₄ and Ce-BiVO₄/Fe₃O₄ composite photocatalysts were successfully synthesized via a facile co-precipitation method. The materials were characterized by XRD, FTIR, XPS, UV-Vis, TEM, EDS, PL, Mott–Schottky analysis, EIS, BET, and EPR. Photocatalytic investigations revealed that BiVO₄/Fe₃O₄, Ce_1.5%-BiVO₄, and Ce_1.5%-BiVO₄/Fe₃O₄ exhibited 1.4-fold, 1.6-fold, and 2.2-fold enhancements in RhB degradation efficiency compared with pure BiVO₄, respectively. Heterojunction formation between BiVO₄ and Fe₃O₄ promoted photogenerated charge separation, while Ce doping introduced electron-trapping centers via Ce⁴⁺/Ce³⁺ redox cycling and reduced the bandgap energy, thereby expanding visible light absorption and improving photocatalytic efficiency. Radical scavenging experiments and EPR tests confirmed that h⁺ and ${·\mathrm{O}}_2^{-}$ are the primary active species responsible for RhB degradation. The magnetic properties of Fe₃O₄ enable rapid catalyst recovery and reuse, offering promising prospects for practical applications in RhB wastewater treatment.