English

• Antibiotics are widely used to cure infectious humans and animals diseases [1]. However, with continuous increased use of antibiotics by human beings, these compounds are discharged into the aquatic environments, such as municipal wastewater treatment systems and river water, resulted in an increase in their concentration in the aquatic environment [2]. Ciprofloxacin (CIP), a typical quinolone antibiotic, is used to treat various bacterial infections, including respiratory tract, skin, and arthrodial infections [3]. It has been detected in surface water, groundwater and wastewater with concentration from ng/L to µg/L [4]. Furthermore, CIP is known to accumulate in aquatic matrices [5], posing potential risks to human health and aquatic organisms [6] or leading to the production of superbugs [7]. Given the stability of CIP, conventional treatment methods are inadequate for removing CIP from water. Therefore, it is urgent to find an effective way to remove CIP from water.

  Fenton treatment technology is one of the most commonly used advanced oxidation technologies, which has the advantages of high stability and has operated at atmospheric pressure and room temperature. However, the traditional Fenton technology needs to be performed in an acidic environment, and the iron ions may precipitate after treatment [8]. In the peroxymonosulfate based advanced oxidation processes (PMS-AOPs), PMS is easily activated to form highly reactive oxidative species (ROS), for instance sulfate radicals (SO₄^•−) and singlet oxygen (¹O₂), which are beneficial to the degradation of antibiotics [9]. Compared with traditional Fenton technology, PMS-AOPs has the advantages of high selectivity, wide pH application range and less iron ions residue, resulting in improved sustainability [10]. The single PMS-AOPs treatment technology still has a low removal efficiency of pollutants. Many studies have found that the combination of Fenton-like PMS-AOPs and photocatalysis (photo-Fenton-like) technology can produce a synergistic effect and greatly improve the removal efficiency of pollutants [11,12]. Under visible light excitation, PMS acts as an electron acceptor capable of inducing the separation of the yield photogenerated electron-hole pairs, which further promotes the activation of the PMS. However, practical applications are hindered by the high recombination rate of photogenerated electrons [13–15]. Therefore, it is a key issue to design and synthesize efficient visible-light photocatalysts.

  The commonly used metal oxide photocatalysts in PMS-AOPs process are TiO₂, ZnO, etc., but there are some problems, such as low utilization of visible light, limited adsorption capacity, poor selectivity and recyclability [16]. The α-Fe₂O₃ is kind of n-type semiconductor (E_g = 2.1 eV), which has attracted extensive attention due to its abundance, corrosion resistance, economic efficiency, and strong visible light adsorption [17]. Graphene aerogels (GA) possess unique three dimensional layer structures effectively dispersing the loaded iron component, so that the active sites are in full contact with the catalyst, facilitating the separation of photogenerated electron-hole pairs and allowing easy recovery catalyst [18,19]. Furthermore, metal-organic frameworks (MOFs) exhibit excellent visible light effects, and MIL-88A(Fe) is one of the widely used MOFs in photo-Fenton oxidation due to its rich content of Fe-O clusters [20–22]. In this work, to further improve catalytic activity, MIL-88A(Fe) was used as a sacrificial template to form α-Fe₂O₃, which was then dotted in graphene aerogel. This maintains the Fe-O cluster redox cycling and improves the photocatalytic performance of the material by utilizing the good electrical conductivity of GA.

  Herein, three-dimensional composites containing MIL-88A(Fe)-derived α-Fe₂O₃ and graphene aerogel (GA-Fe-X, X was the ratio of MIL-88A(Fe) and graphene oxide (GO)) were prepared via a one-step hydrothermal method and freeze-drying method from MIL-88A(Fe) and GO. The structural characteristics of the catalysts and the photodegradation efficiency of CIP over GA-Fe-X in the presence of PMS under various conditions were investigated. The stability and repeatability of the catalyst were investigated by cyclic experiments, and the possible reaction pathways and degradation mechanism were analyzed. The toxicity assessment by the toxicity estimation software tool (T. E. S. T) demonstrated that the intermediates of CIP photodegradation over GA-Fe-1 with the existence of PMS under visible light irradiation was friendly to the environment in this work.

  The reagents have been used as received without further processing. The information regarding chemical reagents is provided in Text S1 (Supporting information). The characterization methods, experimental procedure, analytical methods and the relevant parameters are provided in Text S2 (Supporting information).

  MIL-88A(Fe) was prepared through a hydrothermal synthesis [23]. The detailed preparation procedure of MIL-88A(Fe) and GA-Fe-X can be seen in Text S3 (Supporting information). The morphologies of GA and GA-Fe-1 were displayed in Fig. 1. The scanning electron microscope (SEM) image of the pure GA in Fig. 1a exhibited the porous network structure, which was consistent with the previous report [24]. GA-Fe-1 catalyst showed a porous 3D structure with nanoparticles (Fig. 1b). The transmission electron microscopy (TEM) image of GA-Fe-1 in Fig. 1c clearly showed the nanoparticles and the reduced GO sheets existed in the composite aerogel. Furthermore, the HRTEM image revealed a lattice spacing of 0.253 nm, which was consistent with the (110) crystalline surface of α-Fe₂O₃. In addition, the SEM image in Fig. S1 (Supporting information) showed that MIL-88A(Fe) sample was spindle-shaped crystalline particles with a size of about 300 nm and a length of 1.5–2.0 µm, as expected [25]. The results of SEM and TEM demonstrated that the MIL-88A(Fe) combined with the GA formed the α-Fe₂O₃ and reduced GO composites aerogels. EDS-mapping of GA-Fe-1 exhibited the uniform distribution of Fe, O and C in the composite aerogels (Fig. 1d), further demonstrated that the composite aerogel was successfully prepared.

  Figure 1

  

  Figure 1.  SEM images of GA (a) and GA-Fe-1 (b). HRTEM image (c) and EDS-mapping (d) of GA-Fe-1. (e) XRD patterns of α-Fe₂O₃, GA, and GA-Fe-X. (f) Raman patterns of GA and GA-Fe-X.

  DownLoad: Full-Size Img PowerPoint

  The X-ray diffraction (XRD) patterns of GA and GA-Fe-X were displayed in Fig. 1e. For GA, a broad diffraction pattern near 24° appeared, indicated the successful synthesis of pure GA [18]. As shown in Fig. S2a (Supporting information), three characteristic diffraction patterns at 8°, 10.4° and 12.9° were observed, which have consistent with previous report, suggesting the successful preparation of MIL-88A(Fe) [26]. For GA-Fe-X, the characteristic pattern of MIL-88A(Fe) was not detected, although two new peaks at 33.1° and 35.6° corresponding to the (104) and (110) planes of α-Fe₂O₃ (PDF #33–0664) were observed. Moreover, the pattern intensity corresponding to α-Fe₂O₃ increased with the increased content of MIL-88A(Fe), while the diffraction pattern at 24° gradually decreased. These results implied that combination of MIL-88A(Fe) and GA into GA-Fe-X formed the crystalline structure of α-Fe₂O₃, which further demonstrated the successful synthesis of MIL-88A(Fe)-derived α-Fe₂O₃ and graphene composite aerogel. The Fourier transform infrared (FT-IR) spectra of Fig. S2b (Supporting information) showed GA-Fe-X had the similar FT-IR spectra, the peaks at 580 and 480 cm⁻¹ were assigned to the stretching vibration of the Fe-O bond in α-Fe₂O₃, respectively [27,28], further confirming the formation of α-Fe₂O₃. The Raman spectra of the catalysts were shown in Fig. 1f. The peaks of the catalysts at 1379 and 1594 cm⁻¹ corresponded to D band and G band. Compared with pure GA, the increased intensity ratio of D band and G band for the GA-Fe-X composite aerogels suggested the reduction of GO [29].

  N₂ adsorption-desorption isotherms and pore size distributions of the materials were next studied. The isotherms of GA-Fe-0.75, GA-Fe-1 and GA-Fe-1.5 (Fig. S3a in Supporting information) were attributed to type Ⅳ adsorption curves, indicated that these catalysts had a mesoporous structure [30]. Type Ⅰ and Ⅳ adsorption curves were observed for MIL-88A(Fe) and GA-Fe-0.5, suggested that both microporous and mesoporous structures were presented in these materials [31]. In addition, the pore size distributions (Fig. S3b in Supporting information) reconfirmed the microporous and mesoporous structure of the materials. With the increase of MIL-88A(Fe) content, an increase in the specific surface area of GA-Fe-X was also observed (Table S1 in Supporting information), which was attributed to the construction of extra porosity at the interface of reduced GO and MIL-88A(Fe)-derived α-Fe₂O₃ [32].

  CIP was picked as the target pollutant and the catalytic performance of GA-Fe-X composite aerogels under the illumination of visible light with the presence of PMS (PMS/Vis) was tested. As displayed in Fig. 2a, the removal efficiency of CIP was 55%, 67%, 92% and 82% over GA-Fe-0.5, GA-Fe-0.75, GA-Fe-1 and GA-Fe-1.5 with PMS/Vis within 120 min, respectively. Evidently, the highest degradation efficiency of CIP was achieved over GA-Fe-1, which was higher than both pristine GA and MIL-88A(Fe). Furthermore, the higher decomposition efficiency of CIP was obtained in the GA-Fe-1/PMS/Vis system compared with PMS, PMS/Vis, GA-Fe-1/PMS and GA-Fe-1/Vis system, demonstrating catalyst, visible light and PMS all accelerated the degradation of CIP (Fig. S4a in Supporting information). The kinetic process of CIP decomposition over GA-Fe-X, MIL-88A(Fe) and GA was fitted to a pseudo-first-order kinetic (Fig. 2b). The degradation rate constant (k) for the GA-Fe-1/PMS/Vis system was found to be 0.017 min⁻¹, which was 5.5 and 4.3 times higher than that of GA and MIL-88A(Fe), respectively. The results indicated that the MIL-88A(Fe) combined with the GO formed the α-Fe₂O₃ and reduced GO composites aerogels enhanced the degradation of CIP in the PMS/Vis system. Additionally, the concentration of SO₄²⁻ was detected by the ion chromatography and the results was recorded in Fig. S4b (Supporting information). It was observed that the SO₄^2- was increased with the increasing of reaction time. Ultimately, the solution exhibited a SO₄²⁻ concentration of approximately 420 mg/L after the reaction of 120 min.

  Figure 2

  

  Figure 2.  Photocatalytic degradation curve of CIP: over various catalysts (a), kinetic rate constants of CIP degradation over GA, MIL-88A(Fe) and GA-Fe-X (b), cycling experiments of GA-Fe-1 (c), the effects of different scavengers on CIP degradation (d), kinetic rate constants of different scavengers and relative contribution of active species (e). (f) EPR spectra of DMPO-^•OH, DMPO-SO₄^•−, and TEMP-¹O₂ in GA-Fe-1/PMS/Vis system.

  DownLoad: Full-Size Img PowerPoint

  The influencing factors on the CIP photo-Fenton like degradation was further investigated. Firstly, the effect of PMS concentration on CIP degradation was studied. As shown in Fig. S4c (Supporting information), the degradation rate of CIP was increased with the increasing concentration of PMS until 400 mg/L. After that, the CIP removal efficiency gradually decreased. This was rationalized as followed. As the concentration of PMS increased to 400 mg/L, the photogenerated electrons generated by photoexcitation reacted fully with the photogenerated holes, increasing the production of SO₄^•- [33]. As the concentration of PMS rose further, the SO₄^•− was self-consumed into the less reactive HSO₅⁻ [34], leading to a decrease in the degradation rate of the system. Thus, the optimal dosage of PMS used in this study was 400 mg/L. Determining the optimal amount of catalyst was equally important for system optimization. The results of catalyst dosage on CIP degradation were shown in Fig. S4d (Supporting information). With the increase amount of GA-Fe-1, the photocatalytic degradation efficiency of CIP gradually increased. However, when the amount of catalyst increased from 20 mg to 30 mg, no further improvement in the removal rate of CIP was observed. Considering at the application level, 20 mg was taken as the optimal amount in this study.

  In order to further analyze the effect of external factors on CIP degradation in PMS/Vis system, the impact of pH and inorganic salts was investigated. The experimental results varying pH level were displayed in Fig. S4e (Supporting information) and shown that as the pH value increased from 2.8 to 5.9, the decomposition rate of CIP increased. This might be attributed to the interaction between H⁺ and ^•OH/ SO₄^•− in the acidic condition, resulting in a decline in the active species and the CIP degradation efficiency. However, as the pH value further increased to 9.5, CIP degradation rate decreased. This was attributed to ^•OH formation in weakly alkaline and alkaline conditions, which had a lower oxidation capability than SO₄^•- [35,36]. The effect of anions on CIP degradation in GA-Fe-1/PMS/Vis system was shown in Fig. S4f (Supporting information). The results revealed that the addition of HCO₃⁻, H₂PO₄⁻, SO₄^2- and Cl⁻ all reduced CIP decomposition efficiency. For H₂PO₄⁻ and HCO₃⁻, they not only changed the pH level of CIP solution, but also reacted with ^•OH, SO₄^•− and h⁺ which generated weak oxidative species, such as ^•HPO₄⁻, ^•HCO₃ and ^•CO₃⁻ [20,37]. SO₄²⁻ could consume SO₄^•− generated from PMS activation, and thus hindered the CIP degradation efficiency [38]. The slight decline in CIP degradation rate in the presence of Cl⁻ was ascribed to the direct reaction of PMS (HSO₅⁻) to form less activate HOCl and Cl₂ [39]. Furthermore, the mineralization rate of CIP in the GA-Fe-1/PMS/Vis system was evaluated, the mineralization rate reached approximately 50% within 120 min (Fig. S5 in Supporting information). The as-prepared GA-Fe-1 was evaluated through five consecutive cycles for CIP degradation in PMS/Vis system. GA-Fe-1 maintained excellent reusability and the degradation exceeded 90.0% even in the fifth run (Fig. 2c). Besides, compared to fresh GA-Fe-1, the XRD pattern (Fig. S6 in Supporting information) and SEM image (Fig. S7 in Supporting information) of used GA-Fe-1 was unchanged, which was further proved the stability of the catalyst.

  The main active species in GA-Fe-1/PMS/Vis system was studied by quenching experiments in conjunction with the electron paramagnetic resonance (EPR) technique (results were presented in Figs. 2d-f). Ethylenediaminetetraacetic acid disodium salt (EDTA-2Na), tert–butyl alcohol (TBA), methanol (MeOH), 2, 2, 6, 6-tetramethylpiperidine (TEMP) and AgNO₃ were used as scavengers of h⁺, ^•OH, both SO₄^•− and ^•OH, ¹O₂ and e⁻, respectively. As shown in Fig. 2d, the degradation rate of CIP decreased to 81.4% when TEMP was added, indicating that ¹O₂ was the active species in the photo-Fenton like system [40,41]. EDTA-2Na also had a significant impact, indicated the participation of h⁺. When TBA and MeOH were introduced, the degradation efficiency decreased to 71.9% and 76.5%, respectively, suggesting both SO₄^•− and ^•OH participated in the CIP degradation process. However, the addition of AgNO₃ had no significant effect on the degradation rate of CIP, implying e⁻ was rarely involved in the CIP decomposition process. In order to clarify the contribution of various active species, their contribution was calculated by the following equation [42].

  $ R(\text { reactive species })=\frac{k(\text { reactive species })}{k}=\frac{k-k(\text { quencher })}{k} $

  where R (reactive species) and k (reactive species) represented the comparative contribution and the reaction rate constants of the reactive species, respectively. The k(quencher) and k refer to the pseudo-first order rate constants with the addition of the bursting agent and without the bursting agent, respectively. The estimated contributions of SO₄^•−, ^•OH, e⁻, ¹O₂ and h⁺ to the degradation of CIP were found to be 1.5%, 52.6%, 0.5%, 18.5% and 83.2%, respectively (Fig. 2e). Clearly, the complex of non-free radical and free radical chemical reactions involved in the photocatalytic coupled PMS reaction system. These results indicated there were two reaction pathways, free radical and non-free radical, occurring in this GA-Fe-1/PMS/Vis system. Furthermore, EPR spectroscopy revealed no characteristic peaks at the beginning of the reaction, whilst clear signals were attributed to DMPO-^•OH, DMPO-SO₄^•− and TMPO-¹O₂, which were observed after 30 min (Fig. 2f), indicating the formation of ^•OH, SO₄^•- and ¹O₂ in the photo-Fenton-like process [43,44]. To further demonstrate the source of ¹O₂, the effect of nitrogen on CIP degradation was studied (Fig. S8 in Supporting information). Obviously, nitrogen hardly effect the degradation of CIP over GA-Fe-1 in PMS/Vis system, indicating ¹O₂ generated from the photo-Fenton-like process [45].

  The photoelectric properties, along with the band structure alignment of the materials, were next investigated by ultraviolet–visible diffuse reflectance spectroscopy (UV–vis DRS) and electrochemical tests. As presented in Fig. 3a, the combination of MIL-88A and GA increased the absorption range of the GA-Fe-X, particularly in the visible range. Moreover, GA-Fe-1 had the strongest light absorption capacity among the GA-Fe-X composites. The band gap energies (E_g) of the materials were calculated by the Tacu-plot method [30]. The calculated E_g of MIL-88A(Fe) and GA-Fe-1 were 2.6 and 2.3 eV, respectively (Fig. 3b). A narrower band gap leaded to improved production of photo-generated charge carriers under visible light illumination, resulting in the enhancement of photocatalytic activities [46]. From the Mott-Schottky plot (Fig. 3c), MIL-88A(Fe) and GA-Fe-1 had flat-band potentials (E_FB) of −0.20 and −0.65 eV versus NHE, respectively. The positive slopes indicated that MIL-88A(Fe) and GA-Fe-1 were both n-type semiconductor, and the conduction-band potential (E_CB) was 0.1 eV lower than E_FB for n-type semiconductor [21]. Consequently, the E_CB of MIL-88A(Fe) and GA-Fe-1 were calculated to be −0.30 and −0.75 eV versus NHE, respectively.

  Figure 3

  

  Figure 3.  (a) UV–vis DRS spectra of MIL-88A(Fe) and GA-Fe-X composites. The band gaps (b), the Mott-Schottky plots (c), PL spectra (d), the I-t curve (e) and EIS Nyquist impedance plots (f) of MIL-88A(Fe) and GA-Fe-1.

  DownLoad: Full-Size Img PowerPoint

  Photoluminescence (PL) intensity is closely related to the restructuring of photoexcited electron-hole pairs. Lower PL intensity often means the recombination ratio is lower resulting in greater separation of the photogenerated charge carriers [21]. Fig. 3d depicted the PL intensity of MIL-88A(Fe) and GA-Fe-1 at an excitation wavelength of 275 nm [43]. GA-Fe-1 clearly had a more effective charge separation efficiency than that of MIL-88A(Fe). The lifetime of excited electrons in the photocatalytic process of the catalyst was determined by the time-resolved photoluminescence decay spectra (TRPL). As could be seen from Fig. S9, after fitted the spectra, compared with MIL-88A(Fe) (τ = 2.05 ns), GA-Fe-1 had a shorter lifetime (τ = 1.97 ns), indicated that electrons were effectively transferred [47]. Transient photocurrent and electro-chemical impedance spectroscopy (EIS) measurements were also performed to evaluate the charge separation and transfer properties of the samples [48,49]. As shown in Fig. 3e, the transient photocurrent density of GA-Fe-1 was higher than that of MIL-88A(Fe), suggesting the GA-Fe-1 composite effectively suppressed the recombination of charge carriers. As illustrated in Fig. 3f, GA-Fe-1 had a smaller Nyquist arc radius than MIL-88A(Fe), which indicated a smaller charge-transfer resistance, resulting in a superior capability to accelerate electronic transmission [50]. These results demonstrated that GA-Fe-1 could promote the separation of photoexcited electron-hole pairs, which was conducive to the photo-Fenton-like reaction.

  The mechanism of CIP degradation was further analyzed by X-ray photoelectron spectroscopy (XPS) before and after the reaction of GA-Fe-1 (Fig. 4). The high-resolution C 1s spectrum exhibited binding energy at 284.4 eV (C—C/C=C), 285.3 eV (C—O) and 288.8 eV (C=O), respectively (Fig. 4b). Fig. 4c revealed peaks at 533.1, 531.7, and 530.4 eV, ascribed to C—O/C=O, OH, and Fe-O bonds, respectively [20]. In Fe 2p region (Fig. 4d), two peaks at 711.5 and 725.5 eV belonged to Fe 2p_3/2 and Fe 2p_1/2 of the Fe³⁺ in α-Fe₂O₃, respectively. The used sample showed significant changes, the disappearance of the satellite peak (Sat Fe³⁺) might indicate a change in the electronic structure of iron, which might be due to electron excitation from the valence band to the conduction band or from the 3d orbit of the iron atom to a higher energy level. This suggested that Fe³⁺/Fe²⁺ involved in the activated oxidation of PMS [51]. In accordance with the above studies, the mechanism of CIP degradation over GA-Fe-1 in the system of PMS/Vis was proposed in Fig. S10 (Supporting information). GA-Fe-1 was able to produce photogenerated electrons and holes via a photo-Fenton-like process, in the presence of PMS and the illumination of visible light. The photoproduced electrons migrated between GA and the MIL-88A(Fe) generated α-Fe₂O₃, thus accelerating the transfer efficiency of photo-excited electrons. Furthermore, the photogenerated electrons were able to activate PMS to generate SO₄^•− (Eq. 1). Subsequently, SO₄^•- could react with OH⁻ to produce SO₄²⁻ and ^•OH (Eq. 2). SO₄²⁻ could then react with the photoexcited holes to form SO₄^•− (Eq. 3). Meanwhile, Fe²⁺ and Fe³⁺ from α-Fe₂O₃, could be cyclically converted whilst Fe²⁺ simultaneously activated PMS (HSO₅⁻) to produce SO₄^•− (Eq. 4). The produced ^•OH and SO₄^•− were attributed to the free radical path during the CIP degradation. Additionally, the photogenerated h⁺ could react with PMS to form SO₅^•−, which could subsequently react with H₂O to produce ¹O₂ (Eqs. 5-7) [52]. The generated ¹O₂ had attributed to the non-radical pathway during the CIP photo-Fenton-like degradation. Finally, the efficient degradation of CIP was primarily carried out by ^•OH and SO₄^•− and ¹O₂ (Eq. 8).

  $ \mathrm{e}^{-}+\mathrm{HSO}_5^{-} \rightarrow \mathrm{SO}_4{ }^{--}+\mathrm{OH}^{-} $

  (1)

  $ \mathrm{SO}_4{ }^{\cdot-}+\mathrm{OH}^{-} \rightarrow \mathrm{SO}_4^{2-}+\cdot \mathrm{OH}^{2-} $

  (2)

  $ \mathrm{SO}_4^{2-}+\mathrm{h}^{+} \rightarrow \mathrm{SO}_4{ }^{\cdot-} $

  (3)

  $ \mathrm{Fe}^{2+}+\mathrm{HSO}_5^{-} \rightarrow \mathrm{SO}_4{ }^{\cdot-}+\mathrm{OH}^{-}+\mathrm{Fe}^{3+} $

  (4)

  $ \mathrm{h}^{+}+\mathrm{HSO}_5^{-} \rightarrow \mathrm{H}^{+}+\mathrm{SO}_5{ }^{\cdot-} $

  (5)

  $ \mathrm{SO}_5{ }^{\cdot-}+\mathrm{H}_2 \mathrm{O} \rightarrow 1.5^1 \mathrm{O}_2+2 \mathrm{HSO}_4^{-} $

  (6)

  $ \mathrm{SO}_5^{\bullet-}+\mathrm{SO}_5^{\bullet-} \rightarrow{ }^1 \mathrm{O}_2+2 \mathrm{SO}_4^{2-} $

  (7)

  $ \mathrm{SO}_4^{\bullet-} / \cdot \mathrm{OH} /{ }^1 \mathrm{O}_2+\mathrm{CIP} \rightarrow \text { Products } $

  (8)

  Figure 4

  

  Figure 4.  XPS spectra of GA-Fe-1before and after the cycling tests, the XPS spectra survey (a), C 1s (b), O 1s (c), and Fe 2p (d).

  DownLoad: Full-Size Img PowerPoint

  The CIP degradation intermediates were analyzed by LC-MS [53]. Based on the results obtained from LC-MS, a possible degradation pathway for CIP in the GA-Fe-1/PMS/Vis system was presented in Fig. 5. Firstly, piperazine ring was attacked by ¹O₂ and ^•OH to form a dialdehyde derivative P1 (m/z 362), progressive oxidation of piperazine ring side chain by ^•OH, P1 (m/z 362) produced P2 (m/z 334) with monoaldehyde moiety, and further broke the C₃H₆NO side chain to form an intermediate P3 (m/z 263), which was converted to P4 (m/z 245) after defluorination [54]. The oxidation of aromatic amine resulted in generation of P5 (m/z 274) [55]. Simultaneously, free oxidative radicals attacked the piperazine ring of CIP, resulting in the formation of the intermediate P6 (m/z 306), which could further convert into compound P7 (m/z 263) [56]. Through cyclopropane ring cleavage in compound P7 (m/z 263), intermediate P8 (m/z 195) was produced [57]. These steps yielded various aromatic intermediates without opening the quinoline ring. In later steps, these intermediates decomposed into smaller compounds, such as 1, 2, 4, 5-tetrahydroxybenzene (P9 (m/z 143)) [58]. In pathway Ⅲ, the carboxylic bonds on the pyridine ring were broken by ^•OH, firstly forming intermediate P10 (m/z 288) [59]. The cyclopropane ring was then cleaved, yielding P11 (m/z 248). Further oxidation leaded to the rupture of the pyridine ring, forming P12 (m/z 240), which underwent decarboxylation to produce P13 (m/z 196) [43]. Ultimately, these products were finally mineralized and translated into CO₂ and H₂O by the influence of free radicals and non-free radicals.

  Figure 5

  

  Figure 5.  Proposed pathways of CIP over GA-Fe-1 in PMS/Vis system.

  DownLoad: Full-Size Img PowerPoint

  In order to evaluate the environmental impact of the CIP degradation system, a toxicity study for the degradation intermediates were carried out. The T. E. S. T. based on the quantitative constitutive relationship (QSAR) was used to estimate the biotoxicity of CIP and its possible intermediates. Fig. S11 (Supporting information) showed the decreasing toxicity of CIP intermediates was signed from toxic to harmful, and then to harmless, which the parameter values between 1 and 10 were toxic, between 10 and 100 were harmful, and greater than 100 were harmless [60]. Obviously, the toxicity of most intermediates was less than that of CIP. The results demonstrated that this novel GA-Fe-1/PMS/Vis system could effectively degrade CIP and also diminished the toxicity of its intermediate products.

  In summary, a series of novel GA-Fe-X catalysts were synthesized via hydrothermal methods with GO and MIL-88A(Fe) in different ratios. Characterization of GA-Fe-X was carried out by XRD, FTIR, SEM, and BET, which confirmed the successful synthesis of GA-Fe-X. Under visible light conditions, the GA-Fe-1 and PMS had a synergistic effect on the photocatalytic degradation of CIP, 92% removal efficiency of CIP was achieved with a rapid CIP degradation k value of 0.017 min⁻¹. Moreover, the GA-Fe-1 had a three-dimensional structure, with better structural stability and cyclability. The results of quenching experiments and EPR showed that ^•OH, SO₄^•−, and ¹O₂ were the main active species in the system. The possible degradation pathways of CIP were hypothesized based on the results of LC-MS. The toxicity of the degradation intermediates decreased in comparison to CIP decreasing environmental impact. The materials in this work could be used for water pollution treatment.

  Declaration of competing interest

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

  CRediT authorship contribution statement

  Ning Liu: Writing – original draft. Man Tian: Writing – original draft. Ye Zhang: Conceptualization. Jinming Yang: Data curation. Zhihao Wang: Formal analysis. Wangxi Dai: Investigation. Guixiang Quan: Software. Jianqiu Lei: Software. Xiaodong Zhang: Writing – review & editing. Liang Tang: Validation.

  Acknowledgments

  The authors thank National Natural Science Foundation of China (Nos. 12075152, 42177405, 12075147) for the financial support. The authors also thank Ella Clark from KU Leuven for her assistance with proofreading.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.111063.

  
  
• 1. [1]

     L. Perez-Bou, A. Gonzalez-Martinez, J. Gonzalez-Lopez, et al., Environ. Pollut. 342 (2024) 123115. doi: 10.1016/j.envpol.2023.123115

  2. [2]

     T. Fayaz, N. Renuka. S.K. Ratha, Chemosphere 349 (2024) 140822. doi: 10.1016/j.chemosphere.2023.140822

  3. [3]

     G.F. Zhang, X. Liu, S. Zhang, et al., Eur. J. Med. Chem. 146 (2018) 599–612. doi: 10.1016/j.ejmech.2018.01.078

  4. [4]

     K. Wang, D. Gao, J. Xu, et al., Environ. Sci. Pollut. Res. Int. 25 (35) (2018) 35064–35073. doi: 10.1007/s11356-018-3413-0

  5. [5]

     S. Ghosh, S. Pourebrahimi, A. Malloum, et al., Mater. Today Commun. 37 (2023) 107500. doi: 10.1016/j.mtcomm.2023.107500

  6. [6]

     K. Sathishkumar, S. Naraginti, K. Lavanya, et al., Environ. Res. 235 (2023) 116558. doi: 10.1016/j.envres.2023.116558

  7. [7]

     N.A. Alygizakis, P. Gago-Ferrero, V.L. Borova, et al., Sci. Total. Environ. 541 (2016) 1097–1105. doi: 10.1016/j.scitotenv.2015.09.145

  8. [8]

     J. Qi, G. Jiang, Y. Wan, et al., Chem. Eng. J. 466 (2023) 142960. doi: 10.1016/j.cej.2023.142960

  9. [9]

     Y. Yang, J. Ye, Y. Zhai, et al., Surf. Interfaces. 51 (2024) 104713. doi: 10.1016/j.surfin.2024.104713

  10. [10]

      S. Wang J. Wang, Chem. Eng. J. 456 (2023) 141086. doi: 10.1016/j.cej.2022.141086

  11. [11]

      N. Liu, J.L. Wang, M. Tian, et al., J. Colloid. Interface Sci. 603 (2021) 270–281. doi: 10.1016/j.jcis.2021.06.082

  12. [12]

      H. He, Y. Wang, J. Li, et al., Chem. Eng. J. 427 (2022) 131962. doi: 10.1016/j.cej.2021.131962

  13. [13]

      N. Liu, W.X. Dai, F.H. Fei, et al., Sep. Purif. Technol. 297 (2022) 121545. doi: 10.1016/j.seppur.2022.121545

  14. [14]

      H. Deng, Y. Hui, C. Zhang, et al., Chin. Chem. Lett. 35 (6) (2024) 109078. doi: 10.1016/j.cclet.2023.109078

  15. [15]

      X. Ma, H. Du, M. Tan, et al., Sep. Purif. Technol. 339 (2024) 126644. doi: 10.1016/j.seppur.2024.126644

  16. [16]

      L. Jing, Y. Xu, M. Xie, et al., J. Colloid. Interface Sci. 659 (2024) 520–532. doi: 10.1016/j.jcis.2024.01.011

  17. [17]

      H. Zhang, M. Li, J. Cao, et al., Ceram. Int. 44 (2018) 19958–19962. doi: 10.1016/j.ceramint.2018.07.262

  18. [18]

      M. Farbod. M. Madadi Jaberi, Fuller Nanotub. Car. N. 29 (2021) 244–250. doi: 10.1080/1536383x.2020.1832995

  19. [19]

      N. Liu, D.H. Wang, J.W. Xu, et al., Surf. Interfaces. 46 (2024) 104192. doi: 10.1016/j.surfin.2024.104192

  20. [20]

      Q. Wang, H. Zhou, J. Qian, et al., J. Mater. Sci. Technol. 190 (2024) 67–75. doi: 10.1016/j.jmst.2023.11.045

  21. [21]

      N. Liu, W.Y. Huang, X.D. Zhang, et al., Appl. Catal. B 221 (2018) 119–128. doi: 10.1016/j.apcatb.2017.09.020

  22. [22]

      M. Tang, K. Tang, D. Wang, et al., Sep. Purif. Technol. 329 (2024) 125245. doi: 10.1016/j.seppur.2023.125245

  23. [23]

      H. Wang, C. Zhang, L. Zheng, et al., J. Solid. State Chem. 323 (2023) 124051. doi: 10.1016/j.jssc.2023.124051

  24. [24]

      Q. Xiang, H. Zhang, Z. Liu, et al., Chem. Eng. J. 480 (2024) 147825. doi: 10.1016/j.cej.2023.147825

  25. [25]

      S.G. Khasevani. M.R. Gholami, Mater. Res. Bull. 106 (2018) 93–102. doi: 10.1016/j.materresbull.2018.05.024

  26. [26]

      X.Y. Liao, F. Wang, F. Wang, et al., Appl. Catal. B 259 (2019) 118064. doi: 10.1016/j.apcatb.2019.118064

  27. [27]

      Q. Ding, X.J. Zou, J. Ke, et al., J. Colloid. Interface Sci. 649 (2023) 148–158. doi: 10.1016/j.jcis.2023.06.095

  28. [28]

      G. Sivakumari, M. Rajarajan, R. Naveenkumar, et al., Emergent. Mater. 6 (2023) 1243–1258. doi: 10.1007/s42247-023-00518-8

  29. [29]

      X.L. Xu, W.H. Shi, P. Li, et al., Chem. Mater. 29 (2017) 6058–6065. doi: 10.1021/acs.chemmater.7b01947

  30. [30]

      Q. Chen, J. Li, L. Cheng, et al., Chem. Eng. J. 379 (2020) 122389. doi: 10.1016/j.cej.2019.122389

  31. [31]

      R. Ghasemzadeh. K. Akhbari, Inorg. Chem. 62 (2023) 17818–17829. doi: 10.1021/acs.inorgchem.3c02616

  32. [32]

      M. Muschi. C. Serre, Coord. Chem. Rev. 387 (2019) 262–272. doi: 10.1016/j.ccr.2019.02.017

  33. [33]

      Y. Fang, H. Wang, X. Wang, et al., Sep. Purif. Technol. 307 (2023) 122771. doi: 10.1016/j.seppur.2022.122771

  34. [34]

      X. Zhang, J. Duan, Y. Tan, et al., Sep. Purif. Technol. 293 (2022) 121123. doi: 10.1016/j.seppur.2022.121123

  35. [35]

      D. Li, H.M. Li, Q. Wen, et al., Adv. Funct. Mater. 34 (2024) 2313631. doi: 10.1002/adfm.202313631

  36. [36]

      S.C. Ma, D. Yang, Y.A. Guan, et al., Appl. Catal. B 316 (2022) 121594. doi: 10.1016/j.apcatb.2022.121594

  37. [37]

      H. Cheng, H. Liu, C. Huang, et al., Sep. Purif. Technol. 330 (2024) 125311. doi: 10.1016/j.seppur.2023.125311

  38. [38]

      S.T. Li, L. Chang, K. Wang, et al., Chemosphere. 341 (2023) 140117-140117. doi: 10.1016/j.chemosphere.2023.140117

  39. [39]

      J. Jiang, D. Shi, S. Niu, et al., J. Hazard. Mater. 468 (2024) 133749. doi: 10.1016/j.jhazmat.2024.133749

  40. [40]

      K. Zhu, W. Qin, Y. Chen, et al., Nano Today. 58 (2024) 102462. doi: 10.1016/j.nantod.2024.102462

  41. [41]

      Y. Huang, K. Zhu, Z. Hu, et al., J. Hazard. Mater. 466 (2024) 133611. doi: 10.1016/j.jhazmat.2024.133611

  42. [42]

      N. Li, R. Li, X. Duan, et al., Environ. Sci. Technol. 55 (2021) 16163–16174. doi: 10.1021/acs.est.1c06244

  43. [43]

      J. Sun, C. -H. Shen, J. Guo, et al., J. Colloid. Interface Sci. 588 (2021) 19–30. doi: 10.1016/j.jcis.2020.12.043

  44. [44]

      F. Ye, Y. Su, R. Li, et al., Appl. Catal. B 337 (2023) 122992. doi: 10.1016/j.apcatb.2023.122992

  45. [45]

      Q. Peng, X. Tang, K. Liu, et al., Minerals. 10 (2019) 10010002.

  46. [46]

      T. Wang, C. Zhao, L. Meng, et al., Appl. Catal. B 334 (2023) 122832. doi: 10.1016/j.apcatb.2023.122832

  47. [47]

      Y. Sun, X. Wang, Q. Fu, et al., ACS. Appl. Mater. Interfaces. 13 (33) (2021) 39331–39340. doi: 10.1021/acsami.1c09650

  48. [48]

      Q. Wang, C. Zhang, R. Huo, et al., J. Environ. Manage 367 (2024) 122046. doi: 10.1016/j.jenvman.2024.122046

  49. [49]

      Q. Wang, W. Ma, J. Qian, et al., Environ. Pollut. 347 (2024) 123707. doi: 10.1016/j.envpol.2024.123707

  50. [50]

      W. Huo, M. Wang, H. Wei, et al., Appl. Surf. Sci. 639 (2023) 158199. doi: 10.1016/j.apsusc.2023.158199

  51. [51]

      P.S. Bagus, C.J. Nelin, C.R. Brundle, et al., Phys. Chem. Chem. Phys. 24 (2022) 4562–4575. doi: 10.1039/d1cp04886d

  52. [52]

      C. Yao, Z. Wang, Z. Xu, et al., J. Alloys. Compd. 941 (2023) 168920. doi: 10.1016/j.jallcom.2023.168920

  53. [53]

      Z. Li, X. Yuan, H. Tang, et al., Environ. Sci. -Wat. Res. 8 (2022) 2744–2760. doi: 10.1039/d2ew00320a

  54. [54]

      C. Liu, S. Mao, M. Shi, et al., J. Hazard. Mater. 420 (2021) 126613. doi: 10.1016/j.jhazmat.2021.126613

  55. [55]

      F. Rong, Q. Lu, H. Mai, et al., ACS. Appl. Mater. Interfaces. 13 (2021) 21138–21148. doi: 10.1021/acsami.0c22825

  56. [56]

      X. Jia, J. Zhang, Q. Huang, et al., Environ. Res. 241 (2024) 117639. doi: 10.1016/j.envres.2023.117639

  57. [57]

      H. Mohebali, G. Moussavi, M. Karimi, et al., Sep. Purif. Technol. 315 (2023) 123670. doi: 10.1016/j.seppur.2023.123670

  58. [58]

      Y. Wang, J. Chen, J. Gao, et al., Sep. Purif. Technol. 272 (2021) 118884. doi: 10.1016/j.seppur.2021.118884

  59. [59]

      K.D. Jayeola, D.S. Sipuka, T.I. Sebokolodi, et al., Chem. Eng. J. 479 (2024) 147482. doi: 10.1016/j.cej.2023.147482

  60. [60]

      Q. Man, F. Sun, Z. Zhang, et al., Sep. Purif. Technol. 349 (2024) 127915. doi: 10.1016/j.seppur.2024.127915

• 

  Figure 1  SEM images of GA (a) and GA-Fe-1 (b). HRTEM image (c) and EDS-mapping (d) of GA-Fe-1. (e) XRD patterns of α-Fe₂O₃, GA, and GA-Fe-X. (f) Raman patterns of GA and GA-Fe-X.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Photocatalytic degradation curve of CIP: over various catalysts (a), kinetic rate constants of CIP degradation over GA, MIL-88A(Fe) and GA-Fe-X (b), cycling experiments of GA-Fe-1 (c), the effects of different scavengers on CIP degradation (d), kinetic rate constants of different scavengers and relative contribution of active species (e). (f) EPR spectra of DMPO-^•OH, DMPO-SO₄^•−, and TEMP-¹O₂ in GA-Fe-1/PMS/Vis system.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) UV–vis DRS spectra of MIL-88A(Fe) and GA-Fe-X composites. The band gaps (b), the Mott-Schottky plots (c), PL spectra (d), the I-t curve (e) and EIS Nyquist impedance plots (f) of MIL-88A(Fe) and GA-Fe-1.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  XPS spectra of GA-Fe-1before and after the cycling tests, the XPS spectra survey (a), C 1s (b), O 1s (c), and Fe 2p (d).

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Proposed pathways of CIP over GA-Fe-1 in PMS/Vis system.

  下载: 全尺寸图片 幻灯片
•