A novel CdS photocatalyst based on magnetic fly ash cenospheres as the carrier: performance and mechanism

Abstract

A magnetic photocatalyst was synthesized by direct growth of CdS nanoparticles on the surface of magnetic fly ash cenospheres (MFACs) through chemical deposition. The photodegradation of danofloxacin mesylate (76.2%) demonstrated that CdS/MFACs exhibited higher photocatalytic activity than that of pure CdS (71.5%) under visible light irradiation. Moreover, the as-prepared photocatalyst was relatively stable and displayed good reusability after 5 cycles. A reasonable photocatalytic mechanism for CdS/MFACs was verified by reactive species trapping, and further confirmed by ESR and a radical scavenging species, DMPO. Furthermore, qualitative hydroxyl radical analysis showed that ˙OH was present during the photocatalytic degradation process and was only generated from the photogenerated electrons in the conduction band of CdS/MFACs. In addition, the experimental results showed that the photodegradation of danofloxacin mesylate is associated with ˙O₂⁻ and photogenerated holes. Finally, the photocatalysis degradation reaction kinetics and the mechanism of photodegradation of danofloxacin mesylate were also discussed.

---

1. Introduction

Antibiotics are increasingly found in the environment and this has attracted much attention. Although the concentration of antibiotic residue in the aquatic environment is low, it is widely recognized that antibiotic pollution contributes to the spread of antibiotic resistance.¹ Moreover, because of their inhibition of biomass growth, antibiotics are often not degraded via normal biological processes. Thus, specific treatments have been used to remove antibiotics from wastewater, such as adsorption,² ultrasonic degradation,³ and electrolysis.⁴ However, these approaches are incomplete and energy-inefficient. Photocatalytic technology is an advanced oxidation process for environmental protection and energy conservation that is also extensively used to remove medical organic pollutants.^5,6 P25 is one of the most extensively studied semiconductors. However, the wide band gap of P25 limits its practical application with solar light. Therefore, it is important to develop visible light-driven photocatalysts for using solar energy. CdS is an important II–VI semiconductor with a direct band gap of 2.4 eV. Owing to its unique properties, CdS is widely used in applications including solar cells,⁷ photoelectric conversion,⁸ photocatalysis,⁹ and photo-splitting of water.¹⁰ Hao et al.¹¹ reported the preparation of CdS nanorings by chemical bath deposition, and these as-prepared CdS nanorings exhibited a high photocatalytic efficiency for the degradation of rhodamine. Singh et al.¹² synthesized CdS nanorods via a solvothermal technique and investigated the photocatalytic degradation of salicylic acid. Nevertheless, all these pure CdS catalysts are difficult to recycle and the utilization rates are low. It has been reported that immobilizing the catalysts on a support can overcome these problems,^13,14 and is an efficient method for improving the utilization rate of the photocatalyst. The primary reason is that the photocatalytic reaction often occurs on the surface of the catalyst, which is directly exposed to light and the reactants. Consequently, loading CdS on other materials to increase the photocatalytic efficiency and immobility of the catalysts has been examined for common materials, including CNTs,¹⁵ SiO₂,¹⁶ polymers¹⁷ and halloysite nanotubes (HNTs).¹⁸ However, the supports are costly and difficult to obtain, which restricts their wide use.

Fly ash cenospheres (FACs) are a by-product generated in coal-fired power plants. They are composed of microscopic flakes and mullite cenospheres, which are considered to have great potential as a catalyst support material because of their excellent corrosion resistance and high thermal stability.¹⁹ For this reason, some composite photocatalysts prepared by coating the surface of FACs with photocatalysts have been reported.^20,21 In order to recycle the catalyst, magnetic composites prepared by coating the surface of FACs with a magnetic Fe₃O₄ layer have been reported.²² However, this method is complex and inefficient. Magnetic fly ash cenospheres (MFACs) can be recycled easily and reused efficiently several times because of their magnetic properties resulting from the presence of iron oxides.

For the photocatalytic mechanism of CdS, a number of researchers detected and identified various radicals formed on CdS during the photocatalytic process. Active oxygen species, such as hydroxyl radicals (˙OH),²³ superoxide radicals (˙O₂⁻)²⁴ and photogenerated holes (h⁺),²⁵ are regarded as key species. However, the details of the consumption of radicals are still controversial. For instance, Gao et al. reported that the photogenerated holes of CdS participate in redox reactions to form ˙OH during the photocatalytic process. Nevertheless, the VB edge potential of CdS was more negative than the standard redox potentials of ˙OH/OH⁻ (+2.38 eV). Theoretically, it was proposed that the photogenerated holes cannot form hydroxyl radicals. Currently, to our knowledge, there are few studies on this subject that use electron spin resonance (ESR) spectroscopy.²⁶

In this study, the CdS/MFACs composites were synthesized by chemical bath deposition, and CdS was directly grown on the surface of MFACs. The composites were characterized by X-ray diffraction (XRD), scanning electron microscope (SEM) and UV-vis diffuse reflectance spectra (UV-vis DRS). The photocatalytic performance for the degradation and mineralization of danofloxacin mesylate under visible light irradiation was investigated. Moreover, we used radical inhibitors during the photocatalytic process to examine the effects of radicals on the photocatalytic activity. In particular, the hydroxyl radicals generated from the photocatalytic process were studied using fluorescence and ESR spectroscopy. The photodegradation mechanism of danofloxacin mesylate was also discussed by using mass spectroscopy (MS), and the results indicated that danofloxacin mesylate was degraded stepwise and finally decomposed to CO₂, H₂O, and other gaseous compounds.

2. Materials and method

2.1 Materials and reagents

Hydrogen chloride (HCl, 37%), ethanol (CH₃CH₂OH), methanol (CH₃OH), methylbenzene (C₇H₈), N,N-dimethylformamide (DMF), thiourea (NH₂CSNH₂), ammonia solution (NH₃·H₂O), tert-butyl alcohol (tBuOH), and triethanolamine (TEOA) were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Succinic anhydride, coumarin, (3-aminopropyl)triethoxysilane (APTES), cadmium sulfate, 8/3-hydrate (CdSO₄·8/3H₂O), dimethyl sulfoxide (DMSO), 5,5-dimethyl-1-pyrroline-N-oxide (DMPO), were purchased from Aldrich. All the reagents were analytical grade and were used as received, without further purification.

2.2 Synthesis of CdS/MFACs

Screening and modifying magnetic fly ash cenospheres: the fly ash used in this study was collected from the cyclone of a mass-burning incinerator located in the city of Pingdingshan (China). Because iron ore was present in some of the raw materials, some of the fly ash cenospheres were magnetic, which aided the recycling of the photocatalyst. The details of the procedure for the preparation of CdS/MFACs are shown in Scheme 1. First, 30–50 μm magnetic fly ash cenospheres (MFACs) were selected using different screen mesh sizes. MFACs (5 g) were mixed with 1 mol L⁻¹ HCl (120 mL), and this mixture was stirred for 3 h at a constant temperature of 80 °C. The product was washed with deionized water until the pH was neutral and was dried under vacuum at room temperature, to obtain acid-activated MFACs. The acid-activated MFACs (3 g) and APTES (10 mL) were added to methylbenzene (120 mL), and this mixture was stirred for 12 h at 70 °C under a nitrogen atmosphere. The product was washed with methylbenzene and methanol three times, and amino modified MFACs (NH₂-MFACs) were obtained. To obtain carboxyl modified MFACs (COOH-MFACs), NH₂-MFACs (2 g) and succinic anhydride (3 g) were added to DMF (50 mL) with mechanical agitation for 24 h at room temperature. Subsequently, the solid product was washed with DMF and ethanol three times and dried under vacuum at room temperature for 12 h.

Scheme 1 Schematic of the preparation of CdS/MFACs.

Synthesis of the CdS/MFACs photocatalyst: in a typical synthetic procedure, COOH-MFACs (0.5 g) were dispersed well in deionized water (80 mL) with magnetic stirring, and CdSO₄ (2 mmol), thiourea (4 mmol) and ammonia (5 g) were added to the suspension. The mixture was kept at 60 °C with vigorous magnetic stirring for 3 h. After the reaction, the solid sample was washed with deionized water and ethanol three times and dried under vacuum at 30 °C for 3 h. The reaction process can be written as

Cd(NH₃)_n²⁺ + SC(NH₂)₂ + 2OH⁻ → CdS + 2H₂O + CN₂H₂ + 2nNH₃

2.3 Adsorption experiment

Danofloxacin mesylate solution (20 mg L⁻¹) was used as a target pollutant. The photocatalyst (20 mg) was added to danofloxacin mesylate solution (100 mL). The samples were placed in a thermostatic bath with magnetic stirring at 30 °C for 2 h. The concentration of free danofloxacin mesylate in the solution was measured at 275 nm by a UV-vis spectrophotometer. The adsorption capacity (Q) was calculated using
where C₀ (mg L⁻¹) is the initial concentration of danofloxacin mesylate, C (mg L⁻¹) is the residual concentration of danofloxacin mesylate in solution, V (L) is the volume of danofloxacin mesylate solution, W (g) is the mass of the photocatalyst, and Q (mg g⁻¹) is the adsorbed amount.

2.4 Evaluation of photocatalytic activity

The activity of photocatalysts was evaluated by degrading danofloxacin mesylate under visible light irradiation. The photocatalyst (0.02 g) and danofloxacin mesylate solution (100 mL) were added to the photocatalytic reaction flask. The reaction was conducted in a thermostatic bath with magnetic stirring at 30 °C using a 250 W xenon lamp (Philips, MSD 250) as the light source. The xenon lamp was placed parallel to the reaction flask 10 cm away, and light with wavelengths below 420 nm was filtered out by a glass optical filter. Before the photocatalytic degradation, the suspensions were magnetically stirred in the dark to reach the adsorption equilibrium. To monitor the photocatalytic process, aliquots of danofloxacin mesylate solution (7 mL) were withdrawn from the reaction flask at intervals of 10 min and analyzed by a UV-vis spectrophotometer at 275 nm.

In the recycling experiments, the catalysts were initially separated three times with deionized water and several times with ethanol to remove the residual danofloxacin mesylate and degradation by-products. The catalysts were re-dispersed in the danofloxacin mesylate solution (100 mL, 20 mg L⁻¹) to undergo another cycle.

2.5 Detection of active oxygen species

In order to detect hydroxyl radicals (˙OH), holes (h⁺) and superoxide radicals (˙O₂⁻), which were expected to be the active photocatalytic species, 1.0 mM tBuOH (a quencher of ˙OH), 1 mM TEOA (a quencher of h⁺), or nitrogen (2 L min⁻¹) (to replace air) were added to the danofloxacin mesylate solution, and the photocatalytic activity was tested. In addition, the formation of ˙OH radicals was also evaluated by using coumarin as a fluorescence probe; coumarin reacts easily with ˙OH to form highly fluorescent 7-hydroxycoumarin. Instead of danofloxacin mesylate, coumarin (1.0 mmol L⁻¹) was added to the CdS/MFACs systems. The fluorescent intensity of the generated 7-hydroxycoumarin was monitored at 456 nm with an excitation wavelength of 346 nm on a Varian Cary Eclipse spectrofluorometer.

2.6 Characterization

The phases and atomic concentrations of different elements were investigated using a Bruker D8 Advance powder XRD system with a Cu Kα (λ = 1.5418 Å) radiation source. Diffraction patterns were collected from 20° to 70° at a speed of 5° min⁻¹. The surface morphology and the detailed crystal structure of the synthesized CdS/MFACs were analyzed by using a field emission scanning electron microscope (SEM) (JSM-7001F) equipped with energy-dispersive spectroscopy analysis (EDS). The UV-vis diffuse reflectance spectra (UV-vis DRS) of the samples were recorded with a UV 2450 spectrophotometer in the range of 200–800 nm. Nitrogen physisorption was detected by a Quantachrome micromeritics NOVA 2000e instrument. The adsorption and desorption isotherms were analyzed by using the Brunauer–Emmett–Teller (BET) method and the specific surface area and pore size distribution were calculated by the Barrett–Joyner–Halenda (BJH) method. The Raman spectra were collected at room temperature using a Thermo Scientific DXR Raman spectrometer. Thermogravimetric (TG) analysis was performed in a thermogravimetric analyzer (STA449C, NETZSCH). Photoluminescence (PL) spectra were obtained at room temperature with a Varian Cary Eclipse spectrofluorometer. Total organic carbon (TOC) analysis was performed on a TOC/TN analyzer (multi N/C 2100/2100 S). The degradation course of danofloxacin mesylate solution was monitored by Thermo LXQ mass spectrometry (MS). Active free radicals were identified by electron spin resonance (ESR) on a JEOL FA-200 spectrometer. For the measurements, samples (100 μL) were collected from the reaction solution after 4 min, and were immediately mixed with DMPO (20 mmol L⁻¹, 20 μL) to form a DMPO-radical adduct. Because of its instability in aqueous solutions, the ESR signal of ˙O₂⁻ was detected by dimethyl sulfoxide.

3. Results and discussion

3.1 Surface morphology

The morphologies of the catalysts were observed by SEM. Fig. 1a–f shows typical SEM images of the MFACs and CdS/MFACs. Fig. 1a, c and e shows a typical SEM image of the MFACs, which were regular spheres with diameters of approximately 30 μm. Moreover, there were few impurities on the surface and the spheres were hollow (Fig. S2†). Compared with the smooth surface in Fig. 1e, Fig. 2b, d and f clearly show that the surfaces of MFACs became rough because they were covered with a large amount of particles and scales. These results indicate that CdS nanoparticles were successfully deposited on the surfaces of the MFACs. Fig. S1† shows the EDS patterns of CdS/MFACs, which revealed the main elements in the chemical composition of MFACs (SiO₂, Al₂O₃) and the CdS loading on the MFACs surface. In particular, the Fe peak was mainly caused by the iron oxides in the MFACs.

Fig. 1 SEM images of different photocatalysts. MFACs (a and c), magnification of MFACs (e), CdS/MFACs (b and d), magnification of CdS/MFACs (f).

Fig. 2 XRD patterns of the CdS/MFACs photocatalysts (a) and micro Raman spectra of MFACs and CdS/MFACs (b).

3.2 Crystal structure

Fig. 2a shows the XRD patterns of MFACs and CdS/MFACs photocatalysts. The characteristic MFACs diffraction peaks are clearly visible in the MFACs XRD pattern, which are consistent with the previously reported results,²⁷ although a sharp high peak for nanocrystallite CdS was visible in the XRD pattern of CdS/MFACs. The peaks around 2θ of 26.46°, 31.67°, 44.09° and 51.75° corresponded to the characteristic peaks (111), (200), (220), and (311) of CdS (JCPDS 01-0647), which indicates that the MFACs surface was successfully coated with CdS and was in a good agreement with the SEM results.

Vibrational spectroscopy is a useful tool for examining the crystal phase. Raman spectroscopy of semiconductors is a fast, nondestructive tool for evaluating crystalline material properties, including surface conditions and homogeneity. Fig. 2b shows the micro Raman spectra of MFACs and CdS/MFACs at room temperature. There are no obvious peaks in the Raman spectra of MFACs (line B₁). There were three peaks at 293.1, 587.2, and 892.8 cm⁻¹ in the Raman spectra of CdS/MFACs (line B₂), which corresponded to the first-order, second-order and third-order longitudinal optical phonon (LO) modes of CdS, respectively. This was in good agreement with previous studies of CdS nanostructures.²⁸

3.3 UV-vis spectra of photocatalysts

Fig. 3a shows the diffuse reflectance absorption spectra of MFACs, P25, CdS and CdS/MFACs. According to this figure, MFACs showed a weak absorption between 200 and 380 nm in the ultraviolet region. In contrast, CdS/MFACs showed a broad absorption throughout the ultraviolet light region. Additionally, the absorption edge was shifted toward the longer wavelength side compared with P25. Moreover, CdS/MFACs had intense absorptions in the visible range of 400–550 nm, which promoted organic pollution degradation under visible light. The absorption almost disappeared for wavelengths over 600 nm. The absorption started sharply at a wavelength of 520 nm, indicating no absorption of the impurity energy levels and that the absorption mainly resulted from the band-gap transition of electrons from the valance band to the conduction band of CdS. In addition, the Tauc relation for optical band-gap calculations of photocatalysts is written as

(αhv)ⁿ = k(hv − E_g)

where α is the absorption coefficient, k is the parameter that is related to the effective masses associated with the valence and conduction bands, n is 2 for a direct transition, hv is the absorption energy, and E_g is the band gap energy.²⁹ Fig. 3b shows a Tauc plot for the CdS systems where (αhv)² was plotted against the photon energy (hv). The band gaps estimated from the extrapolated Tauc plot were 2.26 eV for free CdS/MFACs, and 3.26 eV for P25.

Fig. 3 UV-vis reflectance absorption spectra of photocatalysts (a) and band gap energies of photocatalysts (b).

3.4 BET and VSM of composite photocatalysts

BET gas sorptometry measurements were conducted to examine the adsorption–desorption isotherms and the pore-size distribution (inset) of the CdS/MFACs (Fig. 4a). The small adsorption at low partial pressure and the hysteresis loop at higher partial pressure can be observed from the shape of the adsorption–desorption isotherms. This type of behavior is generally classified as a type IV isotherm.³⁰ According to the BJH analysis, the specific surface area was 20.73 m² g⁻¹ and the pore size distribution was in the range of 2–5 nm, which corresponded to a mesoporous microstructure.

Fig. 4 Nitrogen adsorption–desorption isotherm and Barrett–Joyner–Halenda (BJH) pore size distribution plot (inset) of the CdS/MFACs (a) and magnetization curves of CdS/MFACs (inset: photo of magnetic separation) (b).

Fig. 4b shows the magnetic hysteresis curve measured at room temperature for the CdS/MFACs. The saturation magnetization of the CdS/MFACs was 11.5 emu g⁻¹, which suggested that these composites might be easily separated from the solution phase through an external magnetic field.

3.5 TG curves of MFACs and CdS-MFACs

Fig. 5 shows the thermogravimetric curves (TG) of MFACS and CdS-MFACs in the temperature range of 25 to 1200 °C. The TG curve of MFACs in Fig. 5A shows a small weight loss of 3%, which was possibly caused by adsorbed water, structural water, and impurity particles in MFACs, because MFACs are mainly composed of inorganic materials. For CdS-MFACs, the first stage occurred from 25 to 735 °C with a weight loss of 4.8%. This may result from the loss of water molecules and some impurity particles in CdS-MFACs. The second stage started from 735 °C with loss of 30.5% corresponding to the decomposition of CdS nanoparticles.

Fig. 5 TG curve of MFACs (A), TG curve of CdS-MFACs (B).

3.6 Photocatalytic process

The photocatalytic activity of the CdS/MFACs was evaluated through the degradation of danofloxacin mesylate under visible light irradiation. Because the adsorption capacity of photocatalysts has a significant effect on the degradation behavior of organic pollutants, dynamic adsorption experiments were conducted initially. According to the adsorption capacity of CdS/MFACs for danofloxacin mesylate solution in the dark (Fig. S3†), we chose 20 min as the dynamic adsorption time before visible light irradiation.

Fig. 6a shows the photocatalytic performance of different samples under visible light irradiation. According to the figure, a blank test (danofloxacin mesylate without any catalyst) under irradiation showed little decrease in the concentration of danofloxacin mesylate, which meant that danofloxacin mesylate was stable under illumination. The concentration of danofloxacin mesylate in the system within MFACs showed some small changes in its adsorption, which was mainly caused by the porous structure of MFACs. Moreover, both pure CdS and CdS/MFACs showed high photocatalytic activity because of the low energy band gap of CdS. The photocatalytic degradation efficiency of CdS/MFACs reached 76.3% under visible light irradiation within 60 min, which was higher than that of pure CdS and P25. Similar results have been reported in the literature.³¹ Generally, the photocatalytic reaction often occurs on the surface of the catalyst, which is directly exposed to light and the reactants.³² Hence, we assumed that the morphology was the main reason for the photocatalytic differences between pure CdS and CdS/MFACs. There were several plausible explanations for the enhanced photocatalytic activity: (1) compared with pure CdS, the fly ash cenospheres improve dispersion of the composite in danofloxacin mesylate solution, which also leads to good absorption and use of light;^33,34 (2) the FACs acted as a dispersion support to inhibit grain growth, which contributed to making full use of the light for photocatalysis.³⁵

Fig. 6 Photocatalytic performance of different samples under visible light irradiation: no catalyst (a1), MFACs (a2), P25 (a3), CdS (a4), CdS/MFACs (a5) (a). TOC removal of danofloxacin mesylate: MFACs (b1), CdS (b2), CdS/MFACs (b3) (b). Kinetic plots for the photodegradation of danofloxacin mesylate: catalyzed with no catalyst (c1), MFACs (c2), P25 (c3), pure CdS (c4), CdS/MFACs (c5) (c). Repeated experiments (before (#) is the adsorption stage) (d).

The total organic carbon (TOC) values were related to the total concentration of organic compounds in the solution, and the decrease of TOC values during the irradiation time can reflect the degree of mineralization. The changes in the TOC values during the degradation of danofloxacin mesylate with CdS/MFACs are shown in Fig. 6b. The lower removal of TOC was attributed to the formation of simple organic compounds during the degradation of danofloxacin mesylate.

The degradation kinetics of danofloxacin mesylate were investigated by fitting the experimental data to the Langmuir–Hinshelwood model

Because of the low concentration of the reactant (K_C → 0), the Langmuir–Hinshelwood model was reduced to a pseudo first-order kinetics equation

where C_e represents the concentration of danofloxacin mesylate at the adsorption equilibrium, C denotes the concentration at a given reaction time, t, and k is the reaction rate constant determined by plotting −ln(C/C_e) vs. t. As shown in Fig. 6c, the rate constants of the photocatalysts were calculated to be 0.0003, 0.0023, 0.008, 0.016, and 0.0208 min⁻¹ for no catalyst, MFACs, P25, pure CdS, and CdS/MFACs, respectively. The order of the rate constants can be summarized as CdS/MFACs > pure CdS > P25 > MFACs > no catalyst, which agreed well with the results for the photodegradation of danofloxacin mesylate shown in Fig. 6a. The result also demonstrated that CdS/MFACs showed the highest photocatalytic activity.

Generally, reusability and stability are the most important factors for the practical utility of the photocatalyst. As shown in Fig. 6d, the catalysts retained a high photocatalytic activity with no decrease during the photodegradation of danofloxacin mesylate, even after 5 cycles, indicating that CdS/MFACs were relatively stable.

4. Photocatalytic mechanism

To investigate the process of photodegradation and the destruction of danofloxacin mesylate, the mass spectra of the irradiated solutions of danofloxacin mesylate were observed at different time intervals. The spectra are shown in the ESI (Fig. S4†). When danofloxacin mesylate dissolved in water to form a solution, the sulfonic acid groups dissociated and danofloxacin was the main component. Consequently, a characteristic peak at 358 (the danofloxacin peak) can be seen. The probable degradation route is shown in Fig. 7. One possible scenario is that e_aq⁻ reduced the ketone and the oxoquinoline double bond formed alcohol product B. The conversion of compound B to F occurred because B lost its CH₃ group, and the further reaction of B and F with e_aq⁻ resulted in the formation of products C and G, respectively, which were assigned as products of the decarboxylation. Products D and H arose from product C losing the –C₅N₂H₁₁ and cyclopropyl groups, respectively. Product I can be formed when product H lost –OH.

Fig. 7 The degradation of danofloxacin mesylate.

The other possibility is that danofloxacin can be converted to compounds E, J, and K by three successive hydroxylations. Similarly, product N was obtained by decarboxylation of hydroxylation product K. The degradation of K to L and N to O were both via losing –F. Somewhat differently, the conversion of K to L was through a substitution. N can form product O if it is oxidized further. Product P was formed from M by losing –CH₃. Likewise, Q was also formed from P by losing cyclopropyl and –OH. The intermediates were then fragmented into compounds R, S, T, U, V and other low molecular weight intermediates, and finally mineralized to CO₂, H₂O and other gaseous components.

During the processes of photocatalytic degradation of danofloxacin mesylate, an array of reactive oxygen species, such as h⁺, ˙OH, or ˙O₂⁻, were thought to be involved. In order to investigate why the CdS/MFACs degraded danofloxacin mesylate so efficiently, a qualitative analysis of the reactive species and hydroxyl radicals was conducted. Triethanolamine (TEOA), tert-butyl alcohol (tBuOH), and N₂ acted as the quenchers for h⁺, ˙OH, and ˙O₂⁻, respectively, in the photocatalytic process. Fig. 8 shows the photocatalytic activity of CdS/MFACs on the degradation of danofloxacin mesylate with different quenchers. The addition of tBuOH to the danofloxacin mesylate solution has little effect on the photocatalytic activity of CdS/MFACs. However, under N₂, there was little oxygen in the system to limit the generation of superoxide radicals, which has a significant suppressing effect on the degradation of the danofloxacin mesylate solution. This indicates that ˙O₂⁻ is one of the main oxidative species in the degradation processes. Moreover, the photocatalytic degradation of danofloxacin mesylate is clearly inhibited after the addition TEOA, which means that h⁺ was also the main active species in CdS/MFACs for degrading danofloxacin mesylate under visible light irradiation.^36–38 However, it was not clear whether the formation of ˙OH inhibits the degradation or does not enhance the catalytic activity. Thus, the following experiment was conducted.

Fig. 8 Photocatalytic activities of CdS/MFACs during the degradation of danofloxacin mesylate in the presence of different quenchers under visible light irradiation.

Fig. 9 shows that as the irradiation time increased, coumarin was oxidized by the ˙OH to the strongly fluorescent 7-hydroxycoumarin and the inset shows that the fluorescence intensity increased continuously. This confirmed that ˙OH was generated by CdS/MFACs in water when it was irradiated with visible light and ˙OH did not promote the photocatalytic activity for danofloxacin mesylate. Moreover, the formation of reactive radicals was confirmed with an ESR spin-trapping technique by using 5,5-dimethyl-1-pyrroline-N-oxide (DMPO), which was used to detect the active species generated by visible light irradiation.³⁹

Fig. 9 Fluorescence spectral changes and fluorescence intensity at 456 nm of coumarin solution (inset) during irradiation of CdS/MFACs in a 1 mM aqueous coumarin solution (excitation at 346 nm).

ESR spectroscopy is frequently used to investigate photocatalytic reactions at a molecular level. In particular, this technique has been used to follow the photoinduced generation of charge carriers on photocatalysts and their transfer to inorganic adsorbed molecules, such as O₂ and H₂O. Fig. 10B shows the four characteristic peaks of DMPO–˙OH with an intensity of 1 : 2 : 2 : 1, which is similar to results reported by other groups for the ˙OH adduct.⁴⁰ However, no such signals were detected in the dark. This means that irradiation is essential for generation of ˙OH on the surface of the CdS/MFACs, which also confirmed that ˙OH was present in the system. In order to detect ˙O₂⁻ in the system, DMSO was used to replace water. This is because DMSO can reduce the interference of ˙OH with the reaction and the disproportionation of superoxide species in water to allow the slow reactions between ˙O₂⁻ and DMPO to occur.⁴¹ Fig. 10C shows the characteristic peaks, corresponding to the spin-adduct DMPO–˙O₂⁻, which indicated that ˙O₂⁻ was generated in the CdS/MFACs/DMPO/DMSO system.

Fig. 10 DMPO spin-trapping ESR spectra of CdS/MFACs in the dark (A), and with DMPO–˙OH (B) and DMPO–˙O₂⁻ (C) under visible light irradiation.

Usually, ˙OH and ˙O₂⁻ are both important redox initiators for the photocatalytic decomposition of organic compounds.⁴² However, ˙OH had a larger oxidation capacity than ˙O₂⁻. The degradation and mineralization of danofloxacin mesylate is mainly caused by ˙O₂⁻ and h⁺ rather than ˙OH. In other words, there were two main sources of ˙OH radicals: the reaction between holes and surface hydroxyl groups (h⁺ + H₂O(OH⁻) → ˙OH + H⁺), and the adsorbed H₂O₂ (e⁻ + H₂O₂ + H⁺ → ˙OH + H₂O). However, in the CdS/MFACs/danofloxacin mesylate system, because the valence band edge potential of CdS (1.88 eV vs. NHE) is less positive than E° (˙OH/OH⁻) (2.38 eV vs. NHE), the h⁺ on the surface of CdS cannot oxidize OH⁻ to yield ˙OH.⁴³ As shown in Fig. 11, the conduction band of CdS (−0.52 eV) was more negative than the redox potential of O₂/˙O₂⁻ (−0.28 eV), which indicated that the electrons transferred to the surface of semiconductor particles could interact with O₂ to form ˙O₂⁻. However, the photogenerated holes on the valence band of CdS cannot form ˙OH radicals, but they directly oxidize the danofloxacin mesylate. Moreover, ˙OH is mainly generated by ˙O₂⁻,⁴⁴ and the possible formation process for ˙OH is

˙O₂⁻ + e⁻ + 2H⁺ → H₂O₂

H₂O₂ + e⁻ → ˙OH + OH⁻

Fig. 11 Photocatalytic mechanism of CdS/MFACs under visible light irradiation.

Therefore, the amount of ˙OH was restricted by the concentration of ˙O₂⁻ in the system. A series of reduction–oxidation reactions would happen when some ˙O₂⁻ combinated with danofloxacin mesylate, so that only a small amount of ˙OH was formed from ˙O₂⁻. This was consistent with the result of the degradation experiments with free radical scavengers. In summary, the photodegradation of danofloxacin mesylate over the CdS/MFACs photocatalyst is mainly affected by ˙O₂⁻ and the photogenerated holes in the CdS valance bands.

5. Conclusions

A CdS/MFACs photocatalyst was successfully prepared by the chemical bath deposition method. The ultraviolet visible absorption spectrum, photocatalytic activity and recycling of the CdS/MFACs were also investigated. CdS/MFACs had a stronger absorption, higher photocatalytic activity (76.3%) and better recycling under visible light irradiation. Furthermore, the results shown that the loading of the CdS nanoparticles on the surface of MFACs is good, and that the CdS/MFACs composites can be quickly separated by an external magnetic field after photocatalysis. The photocatalytic mechanism was proposed based on the energy band position. By using coumarin, the changes in ˙OH during CdS/MFACs photocatalysis in the reaction solution were also determined. Moreover, the free radicals were also detected by ESR and the photodegradation and mineralization of danofloxacin mesylate were mainly associated with ˙O₂⁻ and the photogenerated holes.

Acknowledgements

We gratefully acknowledge the Natural Science Foundation of China (no. 21306068), the Natural Science Foundation of Jiangsu Province (no. BK20130487), the financial support of the China Postdoctoral Science Foundation (no. 2012M521015) and the Innovation Programs Foundation of Jiangsu Province (no. CXZZ13_0693 and CXZZ13_0665).

References

1. E. S. Elmolla and M. Chaudhuri, J. Hazard. Mater., 2010, 173, 445–449.
2. L. Ji, W. Chen, S. Zheng, Z. Xu and D. Zhu, Langmuir, 2009, 25, 11608–11613.
3. A. Ghauch, H. Baydoun and P. Dermesropian, Chem. Eng. J., 2011, 172, 18–27.
4. L. Sun, C. Wang, M. Ji and X. Kong, Chem. Eng. J., 2013, 215, 50–56.
5. Y. Yan, Y. Wu, Y. Yan, W. Guan and W. Shi, J. Phys. Chem. C, 2013, 117, 20017–20028.
6. C. Gadipelly, A. Pérez-González, G. D. Yadav, I. Ortiz, R. Ibáñez, V. K. Rathod and K. V. Marathe, Ind. Eng. Chem. Res., 2014, 53, 11571–11592.
7. L. A. Silva, S. Y. Ryu, J. Choi, W. Choi and M. R. Hoffmann, J. Phys. Chem. C, 2008, 112, 12069–12073.
8. G. Shen and C.-J. Lee, Cryst. Growth Des., 2005, 5, 1085–1089.
9. Q. Xiang, B. Cheng and J. Yu, Appl. Catal., B, 2013, 138, 299–303.
10. Q. Li, B. Guo, J. Yu, J. Ran, B. Zhang, H. Yan and J. R. Gong, J. Am. Chem. Soc., 2011, 133, 10878–10884.
11. Q. Hao, J. Xu, X. Zhuang, Q. Zhang, Q. Wan, H. Pan, X. Zhu and A. Pan, Mater. Lett., 2013, 100, 141–144.
12. R. Singh and B. Pal, J. Mol. Catal. A: Chem., 2013, 378, 246–254.
13. C. Sriwong, S. Wongnawa and O. Patarapaiboolchai, Chem. Eng. J., 2012, 191, 210–217.
14. B. Wang, Q. Li, W. Wang, Y. Li and J. Zhai, Appl. Surf. Sci., 2011, 257, 3473–3479.
15. H. Y. Li, S. L. Bai, Y. J. Guan and Z. B. Wang, Adv. Mater. Res., 2012, 396, 768–771.
16. C. Peng, X. Cheng, S. Chen, X. Li, T. Li, D. Zhang, Z. Huang and A. Zhang, Photochem. Photobiol., 2012, 88, 1433–1441.
17. B. Pan, Y. Xie, S. Zhang, L. Lv and W. Zhang, ACS Appl. Mater. Interfaces, 2012, 4, 3938–3943.
18. W. Xing, L. Ni, P. Huo, Z. Lu, X. Liu, Y. Luo and Y. Yan, Appl. Surf. Sci., 2012, 259, 698–704.
19. A. Arizmendi-Morquecho, A. Chávez-Valdez and J. Alvarez-Quintana, Appl. Therm. Eng., 2012, 48, 117–121.
20. Y.-t. Yu, Powder Technol., 2004, 146, 154–159.
21. P. Huo, Y. Yan, S. Li, H. Li and W. Huang, Desalination, 2010, 256, 196–200.
22. Z. Lu, W. Zhou, P. Huo, Y. Luo, M. He, J. Pan, C. Li and Y. Yan, Chem. Eng. J., 2013, 225, 34–42.
23. P. Gao, J. Liu, D. D. Sun and W. Ng, J. Hazard. Mater., 2013, 250–251, 412–420.
24. Z. Liu, B. Wu, Y. Zhu, F. Wang and L. Wang, J. Colloid Interface Sci., 2013, 392, 337–342.
25. Y. Guo, X. Shi, J. Zhang, Q. Fang, L. Yang, F. Dong and K. Wang, Mater. Lett., 2012, 86, 146–149.
26. Z. Wang, W. Ma, C. Chen, H. Ji and J. Zhao, Chem. Eng. J., 2011, 170, 353–362.
27. P. Huo, Z. Lu, X. Liu, X. Liu, X. Gao, J. Pan, D. Wu, J. Ying, H. Li and Y. Yan, Chem. Eng. J., 2012, 198, 73–80.
28. S. Rengaraj, S. Venkataraj, S. H. Jee, Y. Kim, C.-w. Tai, E. Repo, A. Koistinen, A. Ferancova and M. Sillanpää, Langmuir, 2010, 27, 352–358.
29. Y. Wang, H. Gao, J. Sun, J. Li, Y. Su, Y. Ji and C. Gong, Desalination, 2011, 270, 258–263.
30. Z. Liu, H. Bai and D. Sun, Appl. Catal., B, 2011, 104, 234–238.
31. G. Lin, J. Zheng and R. Xu, J. Phys. Chem. C, 2008, 112, 7363–7370.
32. J. Wang, B. Li, J. Chen, N. Li, J. Zheng, J. Zhao and Z. Zhu, J. Alloys Compd., 2012, 535, 15–20.
33. J. Lv, T. Sheng, L. Su, G. Xu, D. Wang, Z. Zheng and Y. Wu, Appl. Surf. Sci., 2013, 284, 229–234.
34. J. Zhang, H. Cui, B. Wang, C. Li, J. Zhai and Q. Li, Appl. Surf. Sci., 2014, 300, 51–57.
35. P. Sharma, V. Durga Kumari and M. Subrahmanyam, Chemosphere, 2008, 72, 644–651.
36. Z. Chen, W. Wang, Z. Zhang and X. Fang, J. Phys. Chem. C, 2013, 117, 19346–19352.
37. T. Xu, L. Zhang, H. Cheng and Y. Zhu, Appl. Catal., B, 2011, 101, 382–387.
38. L. Ye, J. Liu, Z. Jiang, T. Peng and L. Zan, Appl. Catal., B, 2013, 142–143, 1–7.
39. C. Chen, P. Lei, H. Ji, W. Ma, J. Zhao, H. Hidaka and N. Serpone, Environ. Sci. Technol., 2004, 38, 329–337.
40. N. Zhang, S. Liu, X. Fu and Y.-J. Xu, J. Phys. Chem. C, 2011, 115, 9136–9145.
41. D. T. Sawyer and J. S. Valentine, Acc. Chem. Res., 1981, 14, 393–400.
42. O. Carp, Prog. Solid State Chem., 2004, 32, 33–177.
43. H. Cheng, B. Huang, Y. Dai, X. Qin and X. Zhang, Langmuir, 2010, 26, 6618–6624.
44. M. R. Hoffmann, S. T. Martin, W. Choi and D. W. Bahnemann, Chem. Rev., 1995, 95, 69–96.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra06593j

---