A magnetically reusable Ce-MOF/GO/Fe₃O₄ composite for effective photocatalytic degradation of chlortetracycline

Abstract

Herein, we report a novel 1/GO/Fe₃O₄ photocatalyst, comprising Ce(BTB)(H₂O) (MOF-1, H₃BTB = 1,3,5-benzenetrisbenzoic acid), graphene oxide (GO), and iron oxide (Fe₃O₄) for photocatalytic degradation of chlortetracycline (CTC). This design enables the effective transfer of electrons from the MOF to GO, thereby reducing the photoelectron–hole recombination rate. Therefore, the optimized 1/GO/Fe₃O₄ photocatalyst with H₂O₂ shows the highest photocatalytic activity toward CTC. The kinetic constant is 5.4 times that in the system of MOF-1 and hydrogen peroxide, which usually acted as efficient electron acceptors to improve the photocatalytic performance of MOFs. More importantly, light absorption is extended from the ultraviolet to the visible region. Furthermore, 1/GO/Fe₃O₄ can be quickly recycled under an applied magnetic field and displays outstanding stability and reusability. According to the radical trapping experiments and electron paramagnetic resonance results, hydroxyl radicals, superoxide radicals, and holes all contribute to excellent photocatalytic activity. The possible catalytic mechanism of 1/GO/Fe₃O₄ is tentatively proposed. This work aims to explore the synergistic effect between metal–organic frameworks (MOFs) and GO, and provide a theoretical basis for MOF-based composites to remove antibiotic contaminants in the environment.

---

Introduction

Over the past few decades, antibiotics have received increasing attention from the public and scientists as emerging contaminants in the natural water environment.¹ Long-term or excessive use of antibiotics may bring them into the environment through direct or indirect means, causing environmental pollution and endangering the health of organisms.^2,3 Chlortetracycline (CTC), as a kind of tetracycline antibiotic, is widely used for the disease treatment of animals and humans because of its easy preparation, low price, convenient use, and broad-spectrum antibacterial advantages.^4,5 However, only part of the CTC can be metabolized by animals and humans, and the unabsorbed residues will be released into the environment.⁶ It is difficult to eliminate antibiotics via traditional wastewater treatment technology because of their stable chemical structure and resistance to biodegradation. Consequently, new techniques are urgently needed to effectively remove residual CTC from water.

In the past few decades, photocatalysis has been approved as a promising green approach for the removal of antibiotics due to its simple operation, low energy consumption, and no secondary contamination. However, it takes a long time (>200 min) to achieve complete removal of CTC as per some studies.^7,8 Therefore, new better photocatalysts are needed to be developed to overcome this barrier. Metal–organic frameworks (MOFs) are chosen as target photocatalytic materials due to their semiconductor-like behavior under light irradiation. Compared with some traditional photocatalysts, MOFs have risen rapidly owing to their three-dimensional porous structure and adjustable electron transmission channels under illumination, which has set off a research frontier in the use of MOFs as photocatalysts.⁹ The porous structures of MOFs can provide abundant catalytic active sites and have more tight interfacial contact with pollutants, which is beneficial for realizing short-distance to trigger the a catalytic reaction. In addition, the two basic structural units (metal centers and organic ligands) of MOFs have diversified combinations and can be modified, indicating the well-regulated electronic structures. Furthermore, H₂O₂-involved photocatalysis has been involved in MOF-mediated pollutant degradation, which produces highly oxygen-reactive species through the reaction between MOFs and H₂O₂, exhibiting improved photocatalytic properties compared to pure MOFs.¹⁰

Despite the effectiveness of photocatalysis, it is still not satisfactory due to inefficient visible-light utilization and limited conductivity, so it requires further improvement. Graphene oxide (GO), as one of the ideal electron accepter candidates, has attracted scientific interest due to its superior conductivity and unique optical properties. The combination of MOFs with GO has been demonstrated to be one of the most effective ways to improve the catalytic efficiency of MOFs.¹¹ Benefiting from the synergistic effects between MOFs and GO, MOFs/GO composites can increase the separation efficiency of photogenerated carriers and significantly improve the utilization rate of visible light, which can lead to significant improvement in photocatalytic performance.^12,13 Recent studies have indicated that the MOFs/GO composites can degrade antibiotics and display outstanding photocatalytic performance. For example, UiO-67/CdS/rGO was prepared for the degradation of ofloxacin (OFL) under simulated sunlight.¹⁴ The GO/NH₂-MIL-125(Ti) hybrid showed improved photocatalytic performance towards the oxidation of gaseous pollutants (NO_x and acetaldehyde) compared with unmodified NH₂-MIL-125(Ti).¹⁵ A novel γ-Fe₂O₃/MIL-53(Fe)/GO composite was also fabricated, which exhibited outstanding photocatalytic performance for the degradation of norfloxacin. The kinetic constant was 4.5 times greater than that of pristine MIL-53 (Fe) and 1.4-fold higher than that in the system with H₂O₂ and MIL-53 (Fe).¹⁶ However, to date, there are a few reports about the elimination of CTC under visible light by MOF-based photocatalysts.

In addition, recycling powder catalysts is another problem for future practical applications. The loss of photocatalysts may happen and bring new secondary pollution. Therefore, developing magnetic catalysts is expected to solve this problem. In this field, some composite photocatalysts, such as ZnIn₂S₄/CoFe₂O₄,¹⁷ S-scheme heterojunction 3D/2D ZnIn₂S₄/BiFeO₃,¹⁸ AgBr/BiOBr/NiFe₂O₄,¹⁹ and Z-scheme heterojunction BiFeO₃/ZnFe₂O₄,²⁰ have been successfully synthesized and exhibited superior photocatalytic activity and excellent magnetic recovery properties.

Herein, a novel 1/GO/Fe₃O₄ composite was successfully prepared by combining MOF, GO, and Fe₃O₄ through reflux and ultrasonic procedures. The MOF, Ce(BTB)(H₂O) (MOF-1) (H₃BTB = 1,3,5-benzenetrisbenzoic acid), is a three-dimensional framework extended by BTB³⁻ ligands with rigid phenyl rings, showing two types of channels with cross-sections of ca. 4.6 Å and 5.6 Å, respectively.²¹MOF-1 has drawn much attention owing to its high surface area and unique ultramicroporous channels. The obtained novel 1/GO/Fe₃O₄ composite in this work was applied for the photocatalytic degradation of CTC. Additionally, the effects of several parameters, including the content of GO, catalyst dosage, and H₂O₂ concentration, on the photocatalytic performance were intensively evaluated. The recyclability of 1/GO/Fe₃O₄ was also examined. Moreover, the photocatalytic mechanism for CTC degradation was also proposed. To the best of our knowledge, this is the first time that an ultramicroporous MOF-based composite was prepared to degrade CTC. This work opens a new route to prepare novel high-performance MOF-based photocatalysts for water purification and environmental remediation.

Experimental section

The detailed preparation of MOF-1, GO, Fe₃O₄, and 1/GO/Fe₃O₄, materials characterization, and evaluation of photocatalytic activity are provided in the ESI† (see the ESI†).

Results and discussion

Components and structure

The PXRD patterns of the as-prepared GO, Fe₃O₄, MOF-1, and 1/GO/Fe₃O₄ are presented in Fig. 1a and Fig. S1 (ESI†). In the PXRD pattern of GO, there is a robust characteristic peak at 11.1° with a corresponding layer spacing of 0.8 nm.²² It is worth noting that no characteristic diffraction peaks of GO can be detected in the 1/GO/Fe₃O₄ composite owing to the small amount of GO (9%). Additionally, the PXRD pattern of the 1/GO/Fe₃O₄ composite shows almost the same characteristic peaks as those of MOF-1, which demonstrates that the incorporation of GO will not change the crystalline phase of MOF-1 and its framework structure is maintained. The diffraction peaks of Fe₃O₄ can also be found in the 1/GO/Fe₃O₄ composite, which preliminarily proves the successful preparation of the 1/GO/Fe₃O₄ composite.

Fig. 1 PXRD patterns (a) and FTIR (b) spectra of GO, Fe₃O₄, MOF-1, and 1/GO/Fe₃O₄.

To further confirm the existence of GO, Fe₃O₄, and MOF-1 in the 1/GO/Fe₃O₄ composite, FTIR spectra were recorded (Fig. 1b). For pure GO, the broadband at 3447 cm⁻¹ corresponds to the O–H stretching vibration, and the bands at 1132, 1382, 1626, and 1734 cm⁻¹ are due to the C–O–C, C–O, C=C, and C=O stretching vibrations, respectively.²³ A significant peak for Fe₃O₄ can be observed at 597 cm⁻¹, which is associated with Fe–O stretching vibrations. Furthermore, the FTIR spectrum of the 1/GO/Fe₃O₄ composite displays no change in the characteristic peaks of pure MOF-1, which manifests the incorporation of GO and Fe₃O₄ has no influence on the original structure of MOF-1 and agrees well with the PXRD patterns. A new peak appeared at about 590 cm⁻¹ in the 1/GO/Fe₃O₄ owing to the Fe–O stretching vibrations in the Fe₃O₄. This verifies that MOF-1, GO, and Fe₃O₄ have been successfully combined.

Morphological studies of MOF-1, GO, Fe₃O₄, 1/GO, and 1/GO/Fe₃O₄ were also conducted through SEM analysis. In Fig. S2a (ESI†), MOF-1 shows a prominent block structure. The image presented in Fig. S2b (ESI†) displays the GO layers with wrinkles.²⁴ Fig. S2c (ESI†) is the morphology of pristine Fe₃O₄, which consists of sphere-like particles with a diameter of 400 nm. After combining MOF-1 with GO, the irregular MOF-1 deposition on the surface of GO can be observed in Fig. S2d (ESI†), demonstrating the formation of a 1/GO composite. As presented in Fig. S2e and f (ESI†), the Fe₃O₄ particles are uniformly distributed over the surface of GO, which enables the catalyst to be quickly and conveniently recovered by applying a magnetic field. Also, the energy dispersive X-ray (EDX) elemental mapping images (Fig. S3a–e, ESI†) show Ce, C, O, and Fe elements in the selected area of the composite, and the distribution is uniform.

To further ascertain the elemental composition and chemical state of the 1/GO/Fe₃O₄ composite, X-ray photoelectron spectroscopy (XPS) was performed. The XPS survey spectra verify the existence of Ce, Fe, O, and C elements in the 1/GO/Fe₃O₄ composite (Fig. 2a), which matches well with the results of EDX. The high-resolution XPS spectra of Ce in MOF-1 and 1/GO/Fe₃O₄ are given in Fig. 2b. The Ce 3d XPS spectrum of MOF-1 is composed of four components, which can be attributed to the existence of 904.3/900.5 eV of Ce(III) 3d_3/2 and 885.7/881.9 eV of Ce(III) 3d_5/2, whereas the XPS spectrum of the 1/GO/Fe₃O₄ sample contains peaks of 904.5/900.6 eV of Ce(III) 3d_3/2 and 886.1/882.2 eV of Ce(III) 3d_5/2.²⁵ Peaks corresponding to Ce(III) 3d_3/2 and Ce(III) 3d_5/2 shift to higher binding energies compared to that of MOF-1 due to the incorporation of GO. The shifts of these binding energies, thanks to the change of electron density when MOF-1, GO, and Fe₃O₄ are assembled into the 1/GO/Fe₃O₄ composite, demonstrate that the interface between them is formed by internal interaction. As is well known, a positive shift of the binding energy represents a decrease in the electron density.²⁶ So, it can be inferred that the photogenerated electrons transfer from MOF-1 to GO in the 1/GO/Fe₃O₄ composite, which is beneficial for constructing the 1/GO/Fe₃O₄ composite, facilitating the separation of photogenerated carriers and the improvement of photocatalytic performance. Fig. 2c presents the high-resolution XPS spectra corresponding to the Fe 2p of both Fe₃O₄ and 1/GO/Fe₃O₄ samples. The Fe 2p XPS spectrum of Fe₃O₄ can be divided into six peaks at 731.0 eV of satellite Fe(II) 2p_1/2, 726.5 eV of Fe(III) 2p_1/2, 723.0 eV of Fe(II) 2p_1/2, 718.2 eV of satellite Fe(II) 2p_3/2, 713.7 eV of Fe(III) 2p_3/2, and 709.9 eV of Fe(II) 2p_3/2, respectively. Another six peaks of Fe 2p are found in 1/GO/Fe₃O₄ at 732.9 eV of satellite Fe(II) 2p_1/2, 727.5 eV of Fe(III) 2p_1/2, 724.0 eV of Fe(II) 2p_1/2, 719.1 eV of satellite Fe(II) 2p_3/2, 714.5 eV of Fe(III) 2p_3/2, and 710.9 eV of Fe(II) 2p_3/2, respectively.²⁷ There is a positive shift in the binding energy of the 1/GO/Fe₃O₄ composite compared to pure Fe₃O₄, confirming the successful incorporation of Fe₃O₄ into the composite. The O 1s XPS spectrum of GO is presented in Fig. S4 (ESI†). The O 1s XPS spectrum of MOF-1 can be divided into three bonds with the binding energies of 532.4, 531.8, and 531.1 eV, originating from C–O, C=O, and Ce–O, respectively.²⁸ The O 1s spectrum of GO is composed of four components corresponding to C–O–C (533.4 eV), C–OH (532.8 eV), C=O (532.2 eV), and O–C=O (531.3 eV) groups, which further confirms that oxygen-containing functional groups are present in GO. The oxygen-containing functional groups can serve as reaction sites and interaction sites with MOF-1 and Fe₃O₄ nanostructures. The peaks due to O–H and Fe–O bonds are observed at 530.2 and 529.0 eV, respectively, in the O 1s XPS spectrum of Fe₃O₄. In the XPS spectrum of 1/GO/Fe₃O₄, the peaks due to C–O–C, C–OH, C=O, O–C=O, Ce–O, and Fe–O appeared at binding energies of 533.3, 532.7, 532.1, 531.5, 530.9, and 530.1 eV, which further confirms the presence of MOF-1, GO, and Fe₃O₄ in the 1/GO/Fe₃O₄ composite. So, the XPS analysis of all samples proved the successful formation of the 1/GO/Fe₃O₄ composite.

Fig. 2 XPS survey spectra of MOF-1, GO, Fe₃O₄, and 1/GO/Fe₃O₄ (a). The high-resolution XPS spectra of Ce 3d (b), and Fe 2p (c).

The UV-vis diffuse reflectance spectra (UV-Vis-DRS) of synthesized MOF-1 and 1/GO/Fe₃O₄ composite were investigated to evaluate the optical properties (Fig. 3). The optical absorption intensity of the 1/GO/Fe₃O₄ composite in the region of 230–850 nm is much stronger than the pristine MOF-1, indicating that the incorporation of GO can improve the optical absorption of MOF-1.^29,30 Also, 1/GO/Fe₃O₄ can be excited by visible light owing to the formation of Ce–O–C bonds between MOF-1 and GO. The reason is that the oxygenated functional groups are distributed on the surface of GO, which acts as anchoring sites for MOF-1.³¹ After incorporating black GO, the 1/GO/Fe₃O₄ composites become darker with more visible light absorption than pure MOF-1, which correspond to a previous report.³² This result combined with XPS characterization results further manifests the successful preparation of the 1/GO/Fe₃O₄ composite.

Fig. 3 UV-Vis DRS spectra of MOF-1 and 1/GO/Fe₃O₄.

Photocatalytic performance

The photocatalytic performance of different catalysts towards the degradation of CTC under visible light irradiation was assessed and is presented in Fig. 4a. Without the addition of the catalyst, the photocatalytic reaction did not proceed in 180 min. However, a small quantity of CTC was degraded when H₂O₂ was present, which may be attributed to the photolysis of H₂O₂ promoting the formation of reactive ˙OH. Interestingly, the degradation efficiency of CTC in the presence of MOF-1 and H₂O₂ was significantly higher than that of pristine MOF-1. Since the Ce cation with variable chemical valences (III and IV) can boost the decomposition of H₂O₂ to produce ˙OH via electron transfer, the photocatalytic performance was improved.³³ We also investigated the photocatalytic performance over GO before and after the addition of H₂O₂, which showed weak photocatalytic activity for CTC. However, the photocatalytic performance could be boosted obviously by the incorporation of GO. And 88.0% degradation of CTC was achieved within 180 min in the 1/GO/Fe₃O₄/H₂O₂/Vis system, which was much better than that in MOF-1. This may be attributed to that incorporating GO results in a better response to visible light irradiation and a lower recombination rate of the photogenerated carriers. The kinetic curve of CTC degradation with different photocatalysts followed the pseudo-first-order process by linear transformation ln(C₀/C) = kt (Fig. S7a, ESI†). Under identical conditions, the pseudo-first-order reaction kinetics (k) of 1/GO/Fe₃O₄ was 5.4 times higher than that of pristine MOF-1 in the H₂O₂/Vis system. By contrast, much boosted CTC degradation efficiency was achieved in photocatalytic reactions with H₂O₂, indicating accelerated catalytic reactions in these systems. Finally, to confirm the role of Fe₃O₄ in 1/GO/Fe₃O₄ in the photocatalytic reaction, the control experiment using only Fe₃O₄ was also assessed. As shown in Fig. S5 (ESI†), Fe₃O₄ can degrade CTC with the help of H₂O₂, but in the 1/GO/Fe₃O₄ complex, Fe₃O₄ was only responsible for endowing superparamagnetism to the composite. Extra experiments have also been conducted to figure out the role of the Fe₃O₄ in the 1/GO/Fe₃O₄ on the photocatalytic reaction. The photocatalytic properties of 1/GO and 1/GO/Fe₃O₄ were very close, indicating that the addition of Fe₃O₄ endowed the catalyst with magnetic properties without affecting the catalytic performance (Fig. S6, ESI†). Using GO as a substrate can increase aggregation and reduce dispersive forces within Fe₃O₄, forming an electrostatic interaction or chemical bond between Fe₃O₄ and GO, which was conducive to the protection of Fe₃O₄. Besides, the tight interfacial contact between Fe₃O₄ and GO further strengthened the protection of Fe₃O₄. Thus, the incorporation of Fe₃O₄ was only responsible for endowing superparamagnetism to the composite, enabling the composite to be recovered by an external magnetic field.

Fig. 4 Effect of various parameters: catalyst species (a), the content of GO (b), catalyst dosage (c), and H₂O₂ concentration (d) on the degradation of CTC in the Fenton-like system. Except for the investigated parameter, other parameters were set as follows: CTC concentration = 20 mg L⁻¹, catalyst dosage = 0.3 g L⁻¹, and H₂O₂ concentration = 30 mM.

A comparison of the degradation efficiency of CTC in this work with those of other reported catalysts is listed in Table S1 (ESI†). As presented, our designed 1/GO/Fe₃O₄ catalyst displays a better comprehensive evaluation concerning the degradation efficiency and recyclability than the reported catalysts. This result demonstrates that 1/GO/Fe₃O₄ appears to be an effective photocatalyst for the degradation of antibiotics in the aquatic environment.

Since the reaction parameters (GO content, catalyst dosage, and H₂O₂ concentration) have a significant influence on photocatalytic performance, further studies were conducted to study the influence of these reaction parameters on the degradation of CTC by 1/GO/Fe₃O₄. The degradation of CTC follows two-stage pseudo-first-order kinetics with different reaction parameters (i.e., the initial rapid decline stage and the subsequent slow decay stage). In the first stage, Ce(III) cations can quickly react with H₂O₂ to generate many ˙OH radicals to attack the target pollutants, thereby helping to degrade CTC rapidly. Owing to the consumption of a large amount of Ce cations in the first stage, the reaction rate in the second stage will be decided by the regeneration rate from Ce(IV) to Ce(III), which caused slower degradation of CTC. Similar results were obtained by another study.^34,35

The effects of the GO content in 1/GO/Fe₃O₄ on the photocatalytic performance toward CTC degradation are presented in Fig. 4b. The photocatalytic performance of the prepared composites based on pseudo-first-order reaction kinetic constants (k) followed the order (Fig. S7b, ESI†): 1/GO/Fe₃O₄-9 (0.0124 min⁻¹) > 1/GO/Fe₃O₄-15 (0.0075 min⁻¹) > 1/GO/Fe₃O₄-5 (0.0069 min⁻¹) > 1/GO/Fe₃O₄-22 (0.0058 min⁻¹). The 9 wt% 1/GO/Fe₃O₄ composite was the optimal photocatalyst since a nearly 88.0% degradation of CTC was achieved in 180 min of treatment. This improved performance is chiefly ascribed to the incorporation of GO inhibits the photogenerated charge recombination in MOF-1. The prolonged carrier lifetime produces higher active free radicals than MOF-1, which can help the photocatalytic degradation of CTC. However, a further increase in the content of GO will cause excess GO to compete for active sites, block available optical energy, and lead to a decrease in the photocatalytic degradation efficiency of CTC. Similar phenomena were also reported by previous studies.^22,36,37

Fig. 4c shows the effect of catalyst dosages on the CTC degradation. As the photocatalyst dosage increased from 0.05 to 0.30 g L⁻¹, the CTC degradation efficiency increased from 65.9 to 88.0%, and the corresponding k value increased from 0.0062 to 0.0124 min⁻¹ (Fig. S7c, ESI†). One possible reason is that the higher the dosage of the catalyst, the more the number of active sites available for the generation of ˙OH, thereby increasing the degradation efficiency. However, a further increase of catalyst dosages finally led to the decline of degradation efficiency down to 84.4%, since excess photocatalyst will cause an increase in turbidity and light-scattering because of catalyst aggregation. This will inhibit photon penetration and reduce the number of photogenerated oxidants, and thereby the degradation rate of pollutants considerably declined.³⁸ These results prove that the optimal amount of catalyst is about 0.30 g L⁻¹.

As revealed in Fig. 4d, the effect of H₂O₂ concentration on CTC degradation was studied. The increase of CTC degradation efficiency and the value of k (76.8 to 88.0%, 0.0088 to 0.0124 min⁻¹, respectively) is achieved by increasing H₂O₂ concentration from 10 to 30 mM (Fig. S7d, ESI†). This may be because the increase in the H₂O₂ concentration led to the generation of more ˙OH for CTC degradation. However, the decrease in the degradation efficiency and k value (81.3% and 0.0097 min⁻¹) was observed with the H₂O₂ concentration further increasing to 40 mM. This may result from the self-scavenging effect of excess H₂O₂ on ˙OH, which corresponds to the previous reports.^39,40

Reusability and stability of the composite

The circulating experiments were also performed to confirm the stability of the prepared photocatalyst. As presented in Fig. 5a, 88.0% degradation of CTC under visible light using the 1/GO/Fe₃O₄ composite was observed for the first run. After ten consecutive runs, the degradation rate of CTC decreased slightly. The minor decrease in the CTC degradation with the increase in the number of cycles could be ascribed to the adsorption of unwashed CTC and partial degradation intermediates on 1/GO/Fe₃O₄, which impeded further degradation of CTC through the photocatalytic process. Similar phenomena were also observed in previous research.⁴¹ Moreover, the PXRD patterns and FTIR spectra of 1/GO/Fe₃O₄ before and after the degradation (Fig. 5b and c) remained almost identical characteristic peaks, further confirming the excellent stability of 1/GO/Fe₃O₄. As displayed in Fig. 5d, the CTC degradation efficiency decreased as the pH value increased from 3 to 9, which may be due to the lower oxidation potential of ˙OH radicals (E⁰ = +2.8 V at pH 0; E⁰ = +2.0 V at pH 14) and the self-decomposition of H₂O₂.³⁴ Also, 1/GO/Fe₃O₄ exhibited relatively outstanding catalytic activity at near-neutral pH, proving that 1/GO/Fe₃O₄ can be used in a wide pH range. These results indicate that the 1/GO/Fe₃O₄ composite was the stable, reusable, and efficient photocatalyst with superior photocatalytic performance.

Fig. 5 Recycling test of 1/GO/Fe₃O₄ (a). The PXRD patterns (b) and FTIR spectra (c) of 1/GO/Fe₃O₄ before and after photocatalytic reactions. Effects of initial pH on degradation of CTC (d).

Possible photocatalytic mechanism

To verify the possible generation of radicals, two possible methods can be considered: (i) electron spin resonance (ESR) and (ii) radical trapping experiments. The ESR was firstly used to verify the existence of ˙OH and ˙O₂⁻ during the photocatalytic process in the presence of 5,5-dimethyl-1-pyrroline N-oxide (DMPO) (Fig. 6a and b).^42,43 No characteristic signals of ˙OH and ˙O₂⁻ can be detected in the dark, while they appeared after 1 min irradiation, and the peak intensity increased with the increase of the irradiation time. Therefore, the production of ˙OH and ˙O₂⁻ during CTC degradation can be confirmed. For radical trapping experiments, silver nitrate (Ag(NO)₃), p-benzoquinone (BQ), disodium ethylenediaminetetraacetic acid (EDTA-2Na), and isopropanol (IPA) were used to trap e⁻, ˙O₂⁻, h⁺, and ˙OH.^44–46 As shown in Fig. 6c, the scavenging effects of Ag(NO)₃, BQ, EDTA-2Na, and IPA prevented the degradation efficiency by 49%, 40%, 56%, and 48%, respectively, after 180 min of irradiation. EDTA-2Na displayed the most significant decline in the degradation of CTC, demonstrating that the h⁺ had the greatest influence on CTC degradation, followed by the existence of e⁻ and ˙OH. The presence of ˙O₂⁻ displayed the lowest impact on degradation, which may be because the production of ˙O₂⁻ is lower than that of e⁻, h⁺ and ˙OH.

Fig. 6 ˙OH (a) and ˙O₂⁻ (b) test in methanol with DMPO as radical trapper under dark and visible light conditions. Effects of different radical scavengers on the degradation of CTC in the 1/GO/Fe₃O₄ system (c).

To explore the charge-separation efficiency, photocurrent measurements (I–t) were performed, and the results displayed that the photocurrent for 1/GO/Fe₃O₄ enhanced in comparison with that for pure MOF-1 (Fig. 7a), demonstrating that the formation of the 1/GO/Fe₃O₄ composite contributed to separating the photogenerated electron–hole pairs.^47–49 This argument was also supported by the electrochemical impedance spectroscopy (EIS) results (Fig. 7b), where 1/GO/Fe₃O₄ showed a smaller diameter, indicative of smaller electron-transfer resistance.^50–53 This was further proven by cyclic voltammetry curves (CV), which provided valuable hints for the photoexcited electron transfer and recombination. In comparison, the anodic peak current and cathodic one at 1/GO/Fe₃O₄ significantly increased, which can be attributed to the outstanding conductivity of GO, accelerating electron transfer (Fig. 7c). Such different photoelectrochemical properties in 1/GO/Fe₃O₄ unequivocally prove that the doping of GO is of great importance.

Fig. 7 Transient photocurrent responses (I–t) (a), EIS Nyquist impedance (b), and CV plots (c) of MOF-1 and 1/GO/Fe₃O₄.

Mott–Schottky experiments were implemented to evaluate the conduction band (CB) potential of MOF-1. It was found that MOF-1 was an n-type semiconductor because of the positive slope of the Mott–Schottky plots. The flat band (FB) potential of the semiconductor is calculated by the intercept of the X-axis extended by the Mott curve, which was 0.1 V positive than the CB potential.^54,55 As shown in Fig. S8a (ESI†), the CB potential of MOF-1 was extrapolated to be about −1.04 eV vs. Ag/AgCl (−0.84 eV vs. NHE), which was more negative than the reduction potential for O₂ to ˙O₂⁻ (−0.33 eV vs. NHE), thus indicating that MOF-1 can combine with dissolved O₂ to generate ˙O₂⁻ (Fig. 8). Combined with the band gap value obtained by UV-Vis DRS spectra (Fig. S8b, ESI†), the corresponding valence band potential (VB) of the composite was calculated to be 2.56 eV vs. Ag/AgCl (2.76 eV vs. NHE).⁵⁶ Since the VB potential of MOF-1 was more positive than ˙OH/OH⁻ (1.99 eV vs. NHE),^57–59 so the photogenerated holes might oxidize OH⁻ to yield ˙OH. In our work, a heterojunction between MOF-1 and GO was formed by reflux, and thus created an inner electric field between them. When it was activated under visible light irradiation, the generated electrons by MOF-1 can be easily delocalized from its π* orbital and captured by GO, which further led to efficient carrier separation and thus an enhanced photocatalytic activity.

Fig. 8 Schematic diagram of the enhanced photocatalytic mechanism over the 1/GO/Fe₃O₄ composite.

To further investigate the possible photocatalytic degradation mechanism, CTC aqueous solutions before and after degradation were studied by liquid chromatography–mass spectrometry (LC–MS),^60–62 and the possible intermediate products (m/z: [M + H]⁺/z) were determined. Fig. S9a and b (ESI†) exhibit the extraction ion chromatography (EIC) of [CTC + H]⁺ ions at m/z = 479 ([M + H]⁺) before irradiation, which was the mass spectrum of CTC. After photocatalytic degradation, the initial concentration of CTC decreased at a retention time of 8.26 minutes, indicating the degradation of CTC (Fig. S9a and c, ESI†). Combined with the previous reports, Fig. S9d–g and Table S2 (ESI†) show the corresponding MS spectra and structure information of the intermediate products. It can be seen that different m/z peaks, including 501, 419, 385, 279, 261, 249, 227, and 185 were detected. The possible degradation pathways are illustrated in Scheme S2 (ESI†) through the comprehensive analysis of the above intermediates and other related references. In the black pathway, ions containing m/z = 501 ([M + Na]⁺) had a molecular weight difference of 23 from CTC, indicating that it was possible to gain weight from Na⁺.⁶³ For the red pathway, the intermediate product of m/z = 385 ([M + H]⁺) was obtained by dechlorination and double bond breaking.⁶³ In the green pathway, CTC was attacked by radicals and converted into products m/z = 419 ([M + H]⁺)⁶⁴ and m/z = 249 ([M + H]⁺)⁶⁵ by removing an acetyl group and conducting a ring-opening reaction, respectively.⁶⁶ For the orange pathway, the intermediate products of m/z = 227 ([M + H]⁺) and m/z = 185 ([M + H]⁺) were obtained by the degradation.⁶⁷ In the blue pathway, the intermediate product of m/z = 279 ([M + H]⁺) was obtained by dechlorination, which was further decomposed after the removal of the aromatic ring, and the intermediate product m/z = 261 ([M + H]⁺) was obtained due to the breaking of the double bond.⁶⁸ Ultimately, CO₂, H₂O, and other inorganic substances could be produced through the mineralization of organic intermediates.

Through the above analysis and the corresponding results, the possible mechanism was proposed, in which the e⁻ in the CB of MOF-1 rapidly transfers to GO due to both the shorter electron-migration distance and the tight solid–solid contact interface, while h⁺ will remain in the VB of MOF-1. Besides, the transfer pathway was further verified by XPS. The binding energies of Ce 3d in 1/GO/Fe₃O₄ shifted to higher binding energies compared with those of pure MOF-1 (Fig. 2b), proving that the photoexcited e⁻ in the CB of MOF-1 can flow to the GO. Meanwhile, the accumulated e⁻ in the GO can react with O₂ to generate ˙O₂⁻ to further decompose CTC, while h⁺ on the VB of MOF-1 might directly oxidize OH⁻ to yield ˙OH to degrade CTC. Besides, the photogenerated electrons in photocatalytic reactions can also facilitate the Ce(III)/Ce(IV) conversion, and consequently, Ce(III) will sustainably react with H₂O₂ to produce ˙OH. Accordingly, synergistic effects can be achieved in the H₂O₂-containing catalytic system for the degradation of CTC. Ultimately, the produced reactive species, including ˙O₂⁻, ˙OH and h⁺, together oxidize CTC to the smaller molecules or ions (CO₂, H₂O, Cl⁻, etc.).

Conclusions

This work reports a facile strategy to synthesize a highly efficient and reusable 1/GO/Fe₃O₄ catalyst, which displays significantly enhanced photocatalytic activity compared with pure MOF-1. This increase is achieved by utilizing GO, which inhibits the recombination of charge carriers, improves the utilization rate of visible light, and helps to improve the photocatalytic activity of MOF-1. Under the optimized conditions, 88.0% degradation of 20 mg L⁻¹ CTC has been achieved in 180 min under visible light irradiation. In addition, the degradation efficiency of CTC does not change significantly after ten consecutive catalytic degradations, demonstrating the satisfactory stability and reusability of the catalyst. More importantly, the catalyst can be quickly and conveniently separated from the aqueous solution after water treatment benefiting from magnetism endowed by Fe₃O₄. This research not only provides more insights for expanding the attractive application fields of MOF-based materials but also provides new photocatalysts for the removal of CTC or other antibiotics from wastewater in the future.

Author contributions

Yuting Bai: methodology, formal analysis, and writing – original draft; Derek Hao: writing – review & editing; Sisi Feng: conceptualization, writing – review & editing, and supervision; Liping Lu: resources and project administration; and Qi Wang: writing – review & editing.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the Natural Science Foundation of China (Grant No. 21671124) and the Fund for Shanxi “1331 Project” Key Innovative Research Team (1331KIRT). A portion of this work was performed at the Scientific Instrument Center of the Shanxi University of China.

References

1. L. Joseph, B.-M. Jun, M. Jang, C. M. Park, J. C. Muñoz-Senmache, A. J. Hernández-Maldonado, A. Heyden, M. Yu and Y. Yoon, Chem. Eng. J., 2019, 369, 928–946.
2. K. Changanaqui, E. Brillas, H. Alarcón and I. Sirés, Electrochim. Acta, 2020, 331, 135194.
3. X. Deng, S. Zhang, N. Ye, L. Zhang and Y. Xiang, Dyes Pigm., 2022, 206, 110660.
4. L. Yi, B. Li, Y. Sun, S. Li, Q. Qi, J. Qin, H. Sun, D. Fang and J. Wang, Sep. Purif. Technol., 2020, 250, 117257.
5. J. Zhao, Q. Song, Q. He, D. D. Dionysiou, F. Wu, Y. Feng and X. Zhang, Chemosphere, 2021, 268, 129269.
6. H. Lv, Y. Duan, X. Zhou, G. Liu, X. Wang, Y. Wang, M. Yuan, Q. Meng and C. Wang, Catal. Sci. Technol., 2020, 10, 8230–8239.
7. J. Liu, X. Yu, L. Wang, M. Guo, W. Zhu and S. Tian, Water Sci. Technol., 2019, 80, 1249–1256.
8. J. Fu, X. Yu, Z. Li, Y. Zhang, W. Zhu and J. Liu, Water, Air, Soil Pollut., 2021, 232, 12.
9. Y. Wang, L. Feng, J. Pang, J. Li, N. Huang, G. S. Day, L. Cheng, H. F. Drake, Y. Wang, C. Lollar, J. Qin, Z. Gu, T. Lu, S. Yuan and H. Zhou, Adv. Sci., 2019, 6, 1802059.
10. Q. Xia, X. Yu, H. Zhao, S. Wang, H. Wang, Z. Guo and H. Xing, Cryst. Growth Des., 2017, 17, 4189–4195.
11. L. Nirumand, S. Farhadi, A. Zabardasti and A. Khataee, Ultrason. Sonochem., 2018, 42, 647–658.
12. B. Bouider, S. Haffad, B. S. Bouakaz, M. Berd, S. Ouhnia and A. Habi, J. Inorg. Organomet. Polym. Mater., 2023, 33, 4001–4011.
13. Z. Wu, Z. Chen, J. Chen, X. Ning, P. Chen, H. Jiang and H. Qiu, Environ. Sci.: Nano, 2022, 9, 4609–4618.
14. S. Zhang, Y. Wang, Z. Cao, J. Xu, J. Hu, Y. Huang, C. Cui, H. Liu and H. Wang, Chem. Eng. J., 2020, 381, 122771.
15. X. Li, Z. Le, X. Chen, Z. Li, W. Wang, X. Liu, A. Wu, P. Xu and D. Zhang, Appl. Catal., B, 2018, 236, 501–508.
16. Q. Wu, Y. Liu, H. Jing, H. Yu, Y. Lu, M. Huo and H. Huo, Chem. Eng. J., 2020, 390, 124615.
17. X. Jiang, D. Fan, X. Yao, Z. Dong, X. Li, S. Ma, J. Liu, D. Zhang, H. Li, X. Pu and P. Cai, J. Colloid Interface Sci., 2023, 641, 26–35.
18. D. Zhang, R. Zhang, J. Liu, X. Pu and P. Cai, J. Am. Ceram. Soc., 2023, 106, 4785–4793.
19. X. Jiang, D. Kong, B. Luo, M. Wang, D. Zhang and X. Pu, Colloids Surf., A, 2022, 633, 127880.
20. X. Jiang, Z. Wang, M. Zhang, M. Wang, R. Wu, X. Shi, B. Luo, D. Zhang, X. Pu and H. Li, J. Alloys Compd., 2022, 912, 165185.
21. Z. Lin, R. Zou, J. Liang, W. Xia, D. Xia, Y. Wang, J. Lin, T. Hu, Q. Chen, X. Wang, Y. Zhao and A. K. Burrell, J. Mater. Chem., 2012, 22, 7813–7818.
22. J. Lin, H. Hu, N. Gao, J. Ye, Y. Chen and H. Ou, J. Water Process Eng., 2020, 33, 101010.
23. Y. Chen, B. Zhai, Y. Liang and Y. Li, Mater. Sci. Semicond. Process., 2020, 107, 104838.
24. R. Al Gaashani, A. Najjar, Y. Zakaria, S. Mansour and M. A. Atieh, Ceram. Int., 2019, 45, 14439–14448.
25. N. Zhou, Y. Ma, B. Hu, L. He, S. Wang, Z. Zhang and S. Lu, Biosens. Bioelectron., 2019, 127, 92–100.
26. D. Zhang, Z. Yang, J. Hao, T. Zhang, Q. Sun and Y. Wang, Chemosphere, 2021, 276, 130226.
27. N. Esfandiari, M. Kashefi, M. Mirjalili and S. Afsharnezhad, Mater. Sci. Eng., B, 2020, 262, 114690.
28. W. Yi, C. Han, Z. Li, Y. Guo, M. Liu and C. Dong, Environ. Sci.: Nano, 2021, 8, 258–268.
29. C. H. Nguyen, M. L. Tran, T. T. V. Tran and R. S. Juang, Sep. Purif. Technol., 2020, 232, 115962.
30. Y. Wang, Z. Qiang, W. Zhu, W. Yao, S. Tang, Z. Yang, J. Wang, J. Duan, C. Ma and R. Tan, ACS Appl. Nano Mater., 2021, 4, 8680–8689.
31. Y. M. Hunge, A. A. Yadav, A. G. Dhodamani, N. Suzuki, C. Terashima, A. Fujishima and V. L. Mathe, Ultrason. Sonochem., 2020, 61, 104849.
32. T. Wang, D. Yue, X. Li and Y. Zhao, Appl. Catal., B, 2020, 268, 118399.
33. X. Chen, X. Liu, H. Wang, K. Cui, R. Weerasooriya, S. He, G. Li, J. Pan and K. Zhou, Environ. Eng. Res., 2022, 27, 200519.
34. J. Tang and J. Wang, Environ. Sci. Technol., 2018, 52, 5367–5377.
35. W. Sharmoukh and H. N. Abdelhamid, J. Cluster Sci., 2023, 34, 2509–2519.
36. N. Liu, W. Huang, X. Zhang, L. Tang, L. Wang, Y. Wang and M. Wu, Appl. Catal., B, 2018, 221, 119–128.
37. X. Zhang, Y. Yang, W. Huang, Y. Yang, Y. Wang, C. He, N. Liu, M. Wu and L. Tang, Mater. Res. Bull., 2018, 99, 349–358.
38. A. R. B. Bayantong, Y. J. Shih, C. D. Dong, S. Garcia Segura and M. D. G. de Luna, Environ. Sci. Pollut. Res., 2021, 28, 5472–5481.
39. M. Liu, D. Zou, T. Ma, Z. Liu and Y. Li, Inorg. Chem. Front., 2019, 6, 1388–1397.
40. J. Tang and J. Wang, Chemosphere, 2020, 241, 125002.
41. Z. Zhu, C. Zhu, C. Hu and B. Liu, J. Colloid Interface Sci., 2022, 607, 595–606.
42. J. Liang, X. Xu, W. Qamar Zaman, X. Hu, L. Zhao, H. Qiu and X. Cao, Chem. Eng. J., 2019, 375, 121908.
43. J. Liang, X. Xu, Q. Zhong, Z. Xu, L. Zhao, H. Qiu and X. Cao, J. Hazard. Mater., 2020, 398, 122861.
44. X. Zhao, X. Yi, X. Wang, J. Zhang, B. Liu, X. Liu, S. Guo and W. Chu, Nanotechnology, 2020, 31, 235707.
45. J. Zhou, J. Ding, H. Wan and G. Guan, J. Colloid Interface Sci., 2021, 582, 961–968.
46. X. Bai, X. Wang, X. Lu, Y. Liang, J. Li, L. Wu, H. Li, Q. Hao, B.-J. Ni and C. Wang, J. Hazard. Mater., 2020, 398, 122897.
47. S. Bao, H. Liang, C. Li and J. Bai, J. Photochem. Photobiol., A, 2020, 397, 112590.
48. Y. Lv, Y. Liu, J. Wei, M. Li, D. Xu and B. Lai, Chem. Eng. J., 2021, 417, 129188.
49. Q. Li, J. Zhao, H. Shang, Z. Ma, H. Cao, Y. Zhou, G. Li, D. Zhang and H. Li, Environ. Sci. Technol., 2022, 56, 5830–5839.
50. F. Wang, H. Fu, F. Wang, X. Zhang, P. Wang, C. Zhao and C. Wang, J. Hazard. Mater., 2022, 423, 126998.
51. N. M. Latiff, X. Fu, D. K. Mohamed, A. Veksha, M. Handayani and G. Lisak, Carbon, 2020, 168, 245–253.
52. X. Chen, S. Xiao, H. Wang, W. Wang, Y. Cai, G. Li, M. Qiao, J. Zhu, H. Li, D. Zhang and Y. Lu, Angew. Chem., Int. Ed., 2020, 59, 17182–17186.
53. 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, e202305538.
54. H. Derikvand, N. Tahmasebi and S. Barzegar, J. Phys. Chem. Solids, 2023, 181, 111528.
55. C. Wan, L. Zhou, L. Sun, L. Xu, D. Cheng, F. Chen, X. Zhan and Y. Yang, Chem. Eng. J., 2020, 396, 125229.
56. Y. Fu, K. Zhang, Y. Zhang, Y. Cong and Q. Wang, Chem. Eng. J., 2021, 412, 128722.
57. A. Gordanshekan, S. Arabian, A. R. Solaimany Nazar, M. Farhadian and S. Tangestaninejad, Chem. Eng. J., 2023, 451, 139067.
58. S. Meena, M. Sethi, S. Meena, P. Saini, K. Kumar, S. Saini, S. Shekhawat, M. L. Meena, A. Dandia, S. D. Lin and V. Parewa, Environ. Res., 2023, 231, 116181.
59. X. Chen, Z. Li, J. Zhou, S. Chen, Y. Huang, W. Wang, W. Wang, Q. Xu and X. Xi, J. Alloys Compd., 2023, 960, 170892.
60. R. Tang, D. Gong, Y. Deng, S. Xiong, J. Zheng, L. Li, Z. Zhou, L. Su and J. Zhao, J. Hazard. Mater., 2022, 423, 126944.
61. Y.-X. Li, W.-L. Duan, B.-Y. Ren, J. Luan and F. Guo, Sep. Purif. Technol., 2023, 311, 123337.
62. K. Yan, R. Li, Z. Yang, X. Li, Y. Wang and G. Wu, iScience, 2021, 24, 102421.
63. Y. Liu, Y. Gao, B. Yao and D. Zou, Chemosphere, 2020, 238, 124543.
64. Y. Zhang, F. Ma, M. Ling, H. Zheng, Y. Wu and L. Li, Chem. Eng. J., 2023, 464, 142762.
65. R. Zhang and K. Zeng, Diamond Relat. Mater., 2021, 115, 108343.
66. F. Guo, X. Huang, Z. Chen, L. Cao, X. Cheng, L. Chen and W. Shi, Sep. Purif. Technol., 2021, 265, 118477.
67. R. Pulicharla, R. Drouinaud, S. K. Brar, P. Drogui, F. Proulx, M. Verma and R. Y. Surampalli, Chemosphere, 2018, 207, 543–551.
68. N. Liu, W. Huang, M. Tang, C. Yin, B. Gao, Z. Li, L. Tang, J. Lei, L. Cui and X. Zhang, Chem. Eng. J., 2019, 359, 254–264.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d3cp04499h

---