Research on the Chloramphenicol Removal Performance of Co-Doped Porous Carbon Materials Derived from Co-Zn Bimetallic ZIFs

Abstract

Chloramphenicol antibiotics (CAPs) are broad-spectrum antibiotics, and excessive consumption has led to increasingly dangerous residues in the environment. The accumulation of these highly toxic and difficult-to-biodegrade CAPs and their long-term exposure in ecological environments can pose insidious and long-term hazards to human health and aquatic organisms. In this study, co-carbon composite nanocatalysts (Co_xZn_10−x-NC) with many carbon nanotubes on the surface were prepared via the one-step pyrolysis of bimetallic Co_xZn_10−x-ZIF with different Co/Zn ratios and used for the degradation of trace amounts of CAPs in a water column. The microstructure and chemical composition of the prepared catalysts were fully characterized using SEM, TEM, and XPS. The CAP degradation experiments demonstrated that Co₆Zn₄-NC in Co_xZn_10−x-NC possessed the highest catalytic activity level, removing 100% of the CAPs in 60 min. The CAPs had a corresponding reaction rate constant of 0.22 min⁻¹, and Co₆Zn₄-NC was able to completely mineralize 44.57% of them. Doping moderate amounts of Zn can effectively improve the carbon nanotube structure on the catalyst surface and promote the generation of monoatomic Co, thus improving catalytic activity. The results of the free-radical burst experiments and electron paramagnetic resonance (EPR) showed that the free-radical pathway mainly dominated within the Co₆Zn₄-NC+PMS system, in which SO₄^•− was the main ROS for CAP degradation.

1. Introduction

The increasing number of refractory pollutants, such as antibiotics, chlorine-containing compounds, and personal care products, in the environmental matrix has become a serious problem around the world [1]. Among them are CAPs, a kind of broad-spectrum antibiotic. Excessive consumption leads to chloramphenicol residues being left in the environment. The enrichment and accumulation of this highly toxic and difficult-to-biodegrade CAP and its long-term exposure in the ecological environment will cause hidden and long-term harm to humans and aquatic organisms. Therefore, the effective removal of CAPs is of great significance not only for the protection of the ecological environment but also for the health and safety of ordinary people.

At present, the main water treatment technologies to remove CAPs include the activated sludge process, activated carbon adsorption, membrane filtration, chemical treatments, and other traditional technologies. In recent years, to achieve high degradation efficiency, strong adaptability, and environmentally friendly properties, international researchers have extensively studied the advanced persulfate oxidation process (SR-AOP) [2,3,4]. The SR-AOP represents an efficient reaction methodology capable of catalytically transforming macromolecular organic contaminants into environmentally benign, mineralized end-products via the orchestrated action of diverse reactive oxygen species (ROS). Some examples of these potent ROS include SO₄^•– and ^•OH, which ensure a comprehensive and effective degradation process. During activation, the peroxy bond cleavage of peroxydisulfate (PDS) or peroxymonosulfate (PMS) usually produces these ROS [5,6]. Therefore, enhancing the activation of PMS to generate more ROS is crucial for improving the performance of the SR-AOP. Various catalysts based on transition metals have garnered significant attention owing to their good catalytic performance and low cost [7,8]. However, issues, such as the aggregation of metal nanoparticles and the leaching of metal ions, remain, resulting in a low catalytic activity and limited reusability.

Metal–organic frameworks (MOFs), which possess exceptional physical and chemical attributes, are tremendously promising for developing advanced heterogeneous catalysts [9,10]. These frameworks are typically formed through the self-assembly of transition metal ions and organic ligands. By undergoing direct carbonization, they are used to construct different transition metal–carbon or carbon-based composites with unique topologies [11]. Compared to the common carbon materials loaded with transition metals, those obtained from MOFs boast impressive morphologies, unique structures, large specific surface areas, and good dispersion [12,13]. Furthermore, the utilization of bimetallic MOFs as a template for the construction of advanced catalysts also facilitates the incorporation of other transition metals into the derived material. Benefiting from the synergistic effect between the different metals, the stability and catalytic activity of catalysts with multiple transition metal active sites in activating PMS are significantly improved [14,15]. For example, Fu et al. successfully synthesized porous CoFe₂O₄ nanocrystals (NCs) using a bimetallic organic framework as a precursor [16]. Using 0.1 g/L CoFe₂O₄ NCs, 2 mM PMS, and 10 mg/L CAP, the CAPs within the reaction system were completely removed after 120 min, and the mineralization rate reached 68%. The CoFe₂O₄ NCs derived from MIL-101 (Fe/Co) showed significantly larger specific surface areas, more porosity, and a higher level of catalytic activity than those of similar catalysts prepared by conventional processes. Li et al. synthesized a porous carbon skeleton-coated FeCo bimetallic catalyst Fe₄/Co@PC-700 using MIL-101(Fe/Co) to apply pressure [17]. The prepared Fe₄/Co@PC-700 was used as a heterogeneous electro-Fenton catalyst for the degradation of antibiotic contaminants, reflecting high catalytic activity levels over a wide pH range of 3–11 and different contaminant degradation pathways from those of the conventional electro-Fenton system.

In summary, catalysts with MOFs as precursors have shown significantly better physical and chemical properties compared with conventional catalysts, but the agglomeration of metal nanoparticles still leads to a reduction in active sites. Therefore, bimetallic MOFs with different Co/Zn ratios were selected as the self-sacrifice template in this study. Among all transition metals, Co has the best PMS activation performance, so it was selected as the preferred active site. As a result of its low boiling point, Zn will volatilize during high-temperature calcination, creating more pores for the derived transition metal–carbon materials, thus further enhancing the catalytic effect of the materials. Meanwhile, the introduction of Zn promotes the formation of single-atom cobalt, which will significantly reduce the agglomeration of active sites in the catalysts derived from MOFs, thus significantly enhancing the catalytic performance. In addition, the synthesized catalysts were employed to remove low concentrations of CAP in the solution. We delved into the impact of factors, including the Co/Zn ratio, the initial pH, the concentration of PMS, and the catalyst dosage, on both the removal efficiency and mineralization rate of CAPs. Given the practical deployment of the catalyst, we also assessed the influence of various coexisting ions in the solution and the catalyst’s reusability. Based on the characterization of the catalyst system, electron paramagnetic resonance analysis, and the quenching experiment, the possible reaction mechanism of CAP degradation was discussed.

2. Experimental

2.1. Reagents and Instruments

2-methylimidazole (98.0% purity) and zinc nitrate hexahydrate (99.0% purity) were purchased from Macklin Chemical Co., Ltd. (Shanghai, China), while cobalt nitrate hexahydrate (99.0% purity) and chloramphenicol (99.0% purity) were purchased from Shanghai Aladdin Industry Co., Ltd. Anhydrous ethanol (99.9% purity), methanol (99.9% purity), tert-butanol (99.0% purity), furfuryl alcohol (99.0% purity), and p-benzoquinone (99.0% purity) were purchased from Chengdu Cologne Chemicals Co., Ltd. (Chengdu, China).

We used a 3Flex 5.02 specific surface area analyzer (BET), Micromeritics, Norcross, GA, United States; a smartlab9 X-ray diffractometer (XRD), Rigku, Akishima City, Tokyo, Japan; an LabRAM HR Evolution Raman spectrometer (FTIR), HORIBA JobinYvon, Paris, France; an Apreo 2 scanning electron microscope (SEM), Thermo Fisher Scientific, Waltham, MA, United States; a JEM-2100F transmission electron microscope (TEM), JEOL Ltd., Akishima City, Tokyo, Japan; a Nexsa G2 X-ray photoelectron spectroscope (XPS), Thermo Fisher Scientific, Waltham, MA, United States; and an UltiMateTM3000 high-performance liquid chromatography equipment (HPLC), Thermo Fisher Scientific, Waltham, MA, United States.

2.2. Preparation of Materials

2.2.1. Preparation of Co_xZn_10−x-ZIF

To prepare the Co_xZn_10−x-ZIF, we modified an existing method for single-metal ZIF-67, and the typical synthesis process was shown in Figure 1 [18]. In the typical synthesis process, 1.81 g 2-methylimidazole (2-IM) (22.05 mmol) was dissolved in 25 mL deionized water under ultrasound, and the solution was recorded as solution A. Then, 314.32 mg cobalt nitrate hexahydrate (1.08 mmol) and 214.19 mg zinc nitrate hexahydrate (0.72 mmol) were diluted in 15 mL of water. Solution A was then added dropwise to solution B and constantly stirred. The two solutions were mixed and stirred continuously for 2 h and then left to stand for 24 h. Subsequently, the solid product was centrifuged (TGL-20M, Lu Xiangyi, Shanghai, China) at 10,000 rpm for 3 min, washed several times with deionized water, and dried under vacuum at 60 °C for 12 h to obtain a purple product named Co₆Zn₄-ZIF. The total amount of cobalt nitrate hexahydrate and zinc nitrate hexahydrate used in the synthesis of Co_xZn_10−x-ZIF was 1.8 mmol, where x refers to the molar ratio of Co in the two metals (the molar ratio is set to Co:Zn = 8:2:6:4:4:2:8, respectively). In addition, the synthesis of single-metal ZIF-67(Co) and ZIF-8(Zn) was similar to that for CoxZn10x-ZIF, where the ratio of Co to Zn in ZIF-67(Co) was 10:0, and the ratio of Co to Zn in ZIF-8(Zn) was 0:10.

2.2.2. Preparation of Co_xZn_10−x-NC

In order to volatilize the zinc inside the structure during the calcination process, which results in the formation of monatomic Co, the prepared Co_xZn_10−x-ZIF was heated up to 900 °C in a corundum crucible boat with a heating rate of 2 °C/min and calcined at 900 °C for 2 h. Subsequently, corresponding Co_xZn_10−x-NC solutions made using the different Co_xZn_10−x-ZIF preparations were obtained. At the same time, Co-NC and Zn-NC were derived by treating ZIF-8(Zn) and ZIF-67(Co) using a similar strategy, and the catalytic performance was compared with that of Co_xZn_10−x-NC.

2.3. Experimental Process

First, 100 mL of 20 mg/L CAP solution was added to a 250 mL beaker, followed by 10 mg Co_xZn_10−x-NC, and then it was stirred continuously for 30 min at room temperature (25 ± 1 °C). The initial pH was adjusted by adding predetermined amounts of NaOH (0.1 mM) and H₂SO₄ (0.1 mM) before the reaction. After continuously stirring until adsorption equilibrium had been achieved, a certain amount of PMS was added to initiate CAP degradation. At regular intervals, 2 mL CAP samples were extracted from the reaction system, and a membrane filter was used to separate the catalyst. To terminate the reaction, methanol (1 mL) was promptly mixed with the extracted CAP samples. Subsequently, the remaining CAP concentration in the reaction mixture was determined using HPLC. We used a C18 column (4.6 mm × 250 mm × 5 μm) for HPLC and a UV detector at a wavelength of 277 nm to monitor the concentration of CAP in the solution at different times. The mobile phase was a mixture of acetonitrile and water (60:40, v/v) at a flow rate of 0.1 mL/min. In addition, all the experiments were repeated three times under the same conditions.

To gain insights into the specific types of ROS generated in the Co_xZn_10−x-NC+PMS system, various concentrations of methanol, tert-butyl alcohol, p-benzoquinone, and furfuryl alcohol were introduced into the solution. After the reaction, the utilized Co₆Zn₄-NC was harvested by centrifugation, and several degradation experiments were conducted to assess its reusability.

3. Discussion

3.1. The Characterization Results of Co_xZn_10−x-NC

3.1.1. SEM Characterization Results

SEM images of Co₆Zn₄-ZIF and Co₆Zn₄-NC are shown in Figure 2. Firstly, the typical diamond dodecahedron morphology of Co₆Zn₄-ZIF can be observed in Figure 2a, which is similar to the morphology of ZIF-67 and ZIF-8 reported in the literature [18]. This indicates that the coexistence of the two metal elements does not have a significant effect on the micro-morphology of Co_xZn_10−x-ZIF. Figure 2b,c show the micro-appearance of calcined Co₆Zn₄-NC. The average particle diameter undergoes a notable reduction, decreasing from 2.2 to 0.8 μm. Additionally, the surface transforms from smooth to rough, with the emergence of numerous fibrous structures post-calcination. This change in the apparent morphology may be due to the formation of metal nanoparticles during calcination, as well as the structural collapse and shrinkage of Co₆Zn₄-ZIF. The large number of fibrous structures on the surface of Co₆Zn₄-NC may be caused by the carbon nanotubes produced by the volatilization of Zn at a high temperature during calcination. On this basis, we characterized the element distribution of Co₆Zn₄-NC using an element-mapping analysis, and the observed results are shown in Figure 2d,e. In Co₆Zn₄-NC, Zn and Zn are uniformly distributed in the material. There is still some Zn (0.98%), which may be attributed to incomplete calcination, but the contents of Co and N are 27.02% and 2.43%, respectively.

3.1.2. TEM Characterization Results

Figure 3 shows the microstructural characteristics of Co₆Zn₄-NC obtained by TEM. As shown in Figure 3a,b, dense black nanoparticles are present in the internal structure of Co₆Zn₄-NC, which may be Co-based metal nanoparticles formed by the Co metal clusters during calcination. Figure 3c also shows the fibrous microstructure on the surface of Co₆Zn₄-NC, which is mainly made of hollow tubes possibly formed by the volatilization of Zn during calcination. By magnifying the internal structure of Co₆Zn₄-NC, it should be noted that the metal nanoparticles inside the structure are surrounded by a dense structure (Figure 3d). By analyzing Figure 3e, the lattice stripe d₁ of the metal nanoparticles is 0.204 nm, corresponding to the crystalline surfaces of metal Co nanoparticles (PDF#15-0806), which proves that Co exists in the form of metal Co to some extent. The Co nanoparticles are surrounded by a dense structure that exhibits a lattice spacing d₂ of 0.34 nm, which corresponds to the (002) crystal plane of graphite carbon (PDF#41-1487). This indicates that Co nanoparticles in the Co₆Zn₄-NC structure are coated by dense graphite carbon, which is the result of Co-catalyzed carbon graphitization. At the same time, we also analyzed the nanotube-like structures on the surface. As shown in Figure 3f, the lattice spacing d₃ of the tubular structure was also 0.34 nm, which indicates that these structures are carbon nanotubes. During pyrolysis, these randomly oriented graphitized carbon structures are wrapped around the surface of larger zinc nanoparticles, and as the temperature increases, the gasification and movement of zinc nanoparticles result in the establishment of these graphitized carbon tubes.

3.1.3. XRD and Raman Characterization Results

Figure 4a shows the XRD characterization results of Co_xZn_10−x-NC, Co-NC, and Zn-NC. As shown in Figure 4a, a wide, weak peak at 26.13° and two sharp and high-intensity characteristic peaks at 44.35°and 51.66° can be observed in the XRD spectra of Co_xZn_10−x-NC and Co-NC. Among them, the peak at 26.13° is a characteristic peak of the (002) crystal facet belonging to graphitic carbon (PDF#41-1487), which is consistent with the distance between the crystal planes observed by high-resolution TEM characterization. The two peaks at 44.35°and 51.66° belong to the crystal planes of metal Co crystals (PDF#150806), which can also be observed in the TEM characterization. In addition, it was found that the intensity of the two characteristic peaks at 44.35° and 51.66° decreases with more zinc in Co_xZn_10−x-ZIF. This can be attributed to a reduction in the Co content within the structure or a decrease in the size of metal Co particles. When there is no Co in the calcined MOFs structure, the characteristic peak corresponding to Co in the obtained Zn-NC structure disappears, replaced by two high-strength and wide characteristic peaks at 26.13°and 43.95°. These indicate the existence of a disordered and defective carbon structure in the material (PDF#41-1487). The Raman spectra of Co_xZn_10−x-NC showed that various catalysts display two distinct peaks located approximately at 1342 cm⁻¹ and 1589 cm⁻¹ (Figure 4b). These peaks can be attributed to the stretching vibrations of disordered carbon (D band) and graphitized carbon (G band) within the sp²-hybridized carbon network. The graphitization degree of the catalyst can be assessed by calculating the ratio of disordered carbon to graphitized carbon (I_D/I_G). Based on the existing literature, the I_D/I_G values for the six different catalysts are 2.59, 2.53, 4.98, 4.034, 4.055, and 4.47, respectively [19]. The Co₆Zn₄-NC catalyst exhibits a higher relative intensity of I_D/I_G compared to those of the other five catalysts. This suggests that Co₆Zn₄-NC possesses more defects, offering more active sites for the reaction and, consequently, enhancing its catalytic performance to a certain degree.

3.1.4. BET Characterization Results

As shown in Figure 5, the pore structure of Co_xZn_10−x-NC was further evaluated by N₂ adsorption–desorption isotherms. All six catalysts show typical IV isotherms with an H4 hysteresis ring, which indicates that both the micropores and mesopores exist in all the catalysts [20].

Table 1. Specific surface area, pore volume, and average pore size of Co_xZn_10−x-NC.

Catalyst	Specific Surface Area (m2 g−1)	Pore Volume (cm3 g−1)	Average Aperture (nm)
Co-NC	198.49	2.46 × 10−1	7.62
Co8Zn2-NC	209.56	3.47 × 10−1	5.05
Co6Zn4-NC	274.94	3.84 × 10−1	4.97
Co4Zn6-NC	421.58	3.86 × 10−1	3.65
Co2Zn8-NC	554.22	3.99 × 10−1	2.79
Zn-NC	797.24	4.97 × 10−1	2.49

provides a summary of the specific surface area, the pore dimensions, and the pore volume for the six types of catalyst. With the increase in Zn content of the Co_xZn_10−x-NC precursor, the specific surface area and the pore volume of Co_xZn_10−x-NC gradually increase, while the average pore size gradually decreases. This may be due to the increase in Zn content, the increase in volatile components in the calcination process, and the formation of more cavities and carbon nanotube structures in this process. The structural collapse caused by the volatilization of Zn during calcination is also more serious. Therefore, Zn-NC shows the largest specific surface area and pore volume, while its average pore size is the smallest among the prepared materials.

3.1.5. XPS Characterization Results

As shown in Figure 6, the valence changes in the surface elements of Co₆Zn₄-NC were characterized using an XPS test. Figure 6a shows the presence of C, N, O, and Co elements on the surface of Co₆Zn₄-NC, while Zn can hardly be observed, which is similar to the results obtained by element mapping. There is almost no change in the intensity of the four main elements before and after the reaction, which, to a certain extent, indicates less ion leaching and more stable catalysts. Figure 6b displays the high-resolution spectrum of C 1s. Fitting allows for C 1s to be decomposed into three distinct peaks positioned at 284.8 eV, 285.8 eV, and 289.3 eV, corresponding to C-C, C-O, and C=O bonds in the catalyst, respectively. This is similar to the composition of carbon in the transition metal–carbon material derived from the MOFs previously reported. Figure 6c shows the XPS spectrum of Co 2p, which was decomposed into three peaks belonging to Co⁰ (778.5 eV), Co²⁺ (782.0 eV), and Co³⁺ (780.2 eV). Different from the characterization results of XRD and TEM, Co in the material exists not only in the form of zero valence, but also in the form of Co³⁺ and Co²⁺, and a corresponding cobalt oxide was not found in the previous test. This may be because during the calcination process, due to the existence of Zn, a part of Co was isolated and coordinated with N, resulting in the formation of monoatomic cobalt, while another part of the agglomerated Co atoms coordinated with the Co atoms themselves to form metal Co nanoparticles. To further validate this hypothesis, we analyzed the XPS high-resolution map of N in detail. As a key active component, second only to the metal center, N plays a crucial role in coordination with metal atoms. Different types of N possess distinct electronic structures and catalytic properties. Figure 6c displays the spectrum of N 1s, which was resolved in several components, including graphitic-N (401.3 eV), pyrrolic-N (399.7 eV), pyridinic-N (398.6 eV), and Co-N (399.0 eV). As we speculated, pyrrolidine N dominates the coordination with Co atoms during calcination to form monoatomic Co [21]. The elemental valence states of used Co₆Zn₄-NC were also analyzed via XPS characterization. As shown in Figure 6c, N is an important active site second only to Co and pyridine-N (398.6 eV). The content decreased from 8.1% to 5.6%, while that of Co-N (399.0 eV) decreased from 24.2% to 14.7%, and graphitized-N (401.3 eV) and pyrrolidine-N (399.7 eV) showed an upward trend. This indicates that pyridine-N and Co-N are possible active sites. In addition, Figure 6d also shows that after the reaction, the content of Co²⁺ (782.0 eV) decreased significantly from 34.2% to 22.7%, and the corresponding Co³⁺ content (780.2 eV) increased from 28.1% to 33.3%, indicating that Co is the main active site in Co₆Zn₄-NC.

3.2. Catalytic Performance of Co_xZn_10−x-NC

The catalytic activities of Co_xZn_10−x-NC were assessed based on the degradation efficiency of the CAPs in various reaction systems, with all the other reaction conditions held constant, except for the type of catalyst used. As illustrated in Figure 7a, all the six catalysts prepared in this study demonstrated the ability to degrade the CAPs to some extent within one hour. The catalytic activity of the catalyst was significantly improved when Zn was incorporated into the catalyst, and it was highest when the ratio of Co:Zn was 6:4. The reaction system with Co₆Zn₄-NC as the catalyst could completely remove the CAPs in the solution within 60 min. A clearer comparison of the catalyst’s activity can be made by calculating the reaction rate constant (k_app). As depicted in Figure 7b, Co₆Zn₄-NC exhibits the highest reaction rate constant of 0.22 min⁻¹ within 15 min, which significantly surpasses that of the Co-NC derived from single-metal ZIF-67 (Co). This suggests that the catalytic activity of Co_xZn_10−x-NC could be effectively improved by the appropriate addition of Zn, which may be attributed to the formation of a large number of carbon nanotubes on Co_xZn_10−x-NC. In addition, the addition of Zn results in some of the Co atoms being present as single atoms. However, the k_app of the Co_xZn_10−x-NC + PMS system decreased significantly from 0.22 min⁻¹ to 0.06 min⁻¹ with a further increase in Zn content, which may be due to the decrease in Co content. In addition, to some extent, the total organic carbon concentration (TOC) can explain the ability of the reaction system to completely mineralize the CAPs into carbon dioxide and water. It is observed that the reaction system with Co₆Zn₄-NC as a catalyst has the strongest mineralization ability, mineralizing 44.57% of the CAPs within 60 min, which is much higher than 36.11% of Co-NC and 25.32% of Zn-NC. To sum up, among the catalysts prepared in this study, Co₆Zn₄-NC showed the highest catalytic activity, so subsequent experiments will be carried out based on Co₆Zn₄-NC to study the interference factors and the reaction mechanism of CAP degradation.

In PMS-based SR-AOP systems, the concentrations of Co₆Zn₄-NC and PMS consistently have a significant effect on pollutant degradation efficiency. As illustrated in Figure 8a, with the increase in catalyst dosage from 0.05 g/L to 0.1 g/L, the CAP degradation efficiency obviously improved. This is because with a higher dose of Co₆Zn₄-NC, there are more active sites for PMS decomposition, which enhances the generation of ROS during CAP degradation in a Co₆Zn₄-NC+PMS system. However, the degradation efficiency of the CAPs in this Co₆Zn₄-NC+PMS system did not increase significantly by increasing the Co₆Zn₄-NC dosage from 0.1 g/L to 0.25 g/L, or it remained unchanged. This may be due to the fact that excess Co²⁺ in the Co₆Zn₄-NC+PMS system consumes a portion of SO₄^•− and ^•OH [22,23]. Figure 8b demonstrates the impact of the PMS concentration on the CAP removal efficiency in the Co₆Zn₄NC+PMS system. As shown in Figure 8b, when the PMS concentration increases from 0.5 mmol/L to 1.0 mmol/L, the CAP removal value increased significantly from 92.4% to 100%, and the corresponding k_app increased from 0.15 min⁻¹ to 0.22 min⁻¹. This indicates that an increase in PMS concentration can significantly enhance the catalytic performance of Co₆Zn₄-NC+PMS systems. Nevertheless, by improving PMS metrology, the catalytic performance of the Co₆Zn₄-NC+PMS system did not show a significant improvement, which may be attributed to there being insufficient catalysts in the reaction system to provide enough active sites for the excess PMS at the same time [24]. In addition, this may also lead to the self-quenching of some PMS [25]. Therefore, 0.1 g Co₆Zn₄-NC and 1.0 mmol/L PMS were chosen as the best parameters for this experiment.

In a Co₆Zn₄-NC+PMS system, the initial pH value can indirectly affect the surface charge of Co₆Zn₄-NC and the reactivity of the active site, so the initial pH is important for the removal of CAPs [26]. As shown in Figure 9, the Co₆Zn₄-NC+PMS system showed the highest catalytic activity level when the pH was seven. However, the catalytic activity of the Co₆Zn₄-NC+PMS system was similar when the pH was nine and was not significantly affected. When the pH was three and five, CAP removal in the Co₆Zn₄-NC+PMS system decreased significantly to 71.9% and 94.1%, and the corresponding k_app significantly decreased to 0.05 min⁻¹ and 0.13 min⁻¹. This might be because under acidic conditions, excess H⁺ in the reaction system will not only quench a part of the ROS produced by the self-decomposition of PMS but will also compete with Co₆Zn₄-NC for the O-O binding of PMS, thus hindering the self-decomposition of PMS activated by Co₆Zn₄-NC. As the pH of the Co6Zn4-NC+PMS system rose to 11, CAP removal was minimized, with only 50.4% of the CAPs removed in 60 min. Correspondingly, the reaction rate constant decreased significantly to 0.02 min⁻¹, which may be due to the formation of low-catalytic-activity Co(OH)₂ precipitation under alkaline conditions. In summary, the optimal reaction pH for the Co₆Zn₄-NC+PMS system is seven.

The radicals reacted (Equations (1) and (2)) to form chlorine free radicals with a low redox potential. To some extent, this also shows that the radical pathway contributes significantly to the degradation of CAPs in Co₆Zn₄NC+PMS systems. Similar to Cl⁻, HPO₄²⁻ can also combine with the ROS produced in the reaction system to produce phosphate free radicals, whose activity level is lower than that of sulfate and hydroxyl radicals, thus inhibiting the reaction system dominated by free-radical pathways to a certain extent. As shown in Figure 10c, under different concentrations of HPO₄²⁻, the removal rate of CAPs is reduced to 92.2%, 86.6%, and 82.1%. Compared with the other three inorganic anions, HCO₃⁻ affects the SR-AOP system in different ways. HCO₃⁻ does not only quench a part of the ROS produced in the system but also affects the pH of the reaction solution of the Co₆Zn₄-NC+PMS system, thus further affecting the removal efficiency of the CAPs. Among the four inorganic anions, HCO₃⁻ had the most obvious inhibitory effect on the Co₆Zn₄-NC+PMS system. Under the action of 1 mM, 3 mM, and 5 mM HCO₃⁻, the CAP removal decreased significantly to 80.1%, 77.6%, and 76.1%.

$${\mathrm{C}\mathrm{l}}^{-}+{{\mathrm{SO}}_4}^{•-}\to {\mathrm{Cl}}^{•}+{{\mathrm{SO}}_4}^{2-}$$

(1)

$${\mathrm{C}\mathrm{l}}^{-}+\bullet \mathrm{OH}\to {\mathrm{ClOH}}^{•-}$$

(2)

3.3. Potential Reaction Mechanism

To investigate the underlying reaction mechanism of the Co₆Zn₄-NC+PMS system, it is crucial to identify the ROS generated by Co₆Zn₄-NC during the activation of PMS, so we carried out a free-radical-quenching experiment. Because there may be both radical and non-radical pathways during the degradation of the CAPs, MeOH was chosen as the scavenger of ^•OH and SO₄^•−, TBA was selected as the quenching agent of ^•OH, FFA was selected as the scavenger of ¹O₂, and p-BQ was selected as the scavenger of O₂^•−. As shown in Figure 11a, the removal of CAPs in the presence of a high concentration of MeOH (1000 mM) decreased significantly from 100% to 29.4%. This demonstrates that the radical pathway is the dominant mechanism for PMS activation within the Co₆Zn₄-NC+PMS system, while ^•OH and SO₄^•− are the main ROS of CAP degradation. The main types of radical were further identified by TBA, and it was observed that 74.4% of CAPs were removed in the presence of 1000 mM TBA, which indicated that SO₄^•− contributed more to the degradation of CAPs among the radicals. The ^•OH also played a certain role. CAP removal was reduced by 13.2% in the presence of 50 mM FFA, suggesting that ¹O₂ may also contribute to CAP degradation. The existing studies have shown that O^•− may be the intermediate product in the reaction process of the SR-AOP system and was converted into ¹O₂ in the subsequent reaction. Since ¹O₂ promotes the degradation of CAPs, the role of O₂^•− in the reaction process was also investigated by adding p-BQ [27]. However, CAP removal was slightly decreased by only 1.9% in the presence of 10 mM p-BQ, indicating that O₂^•− has no obvious contribution to the location of CAPs, and O₂^•− is not a minor source of ¹O₂.

On this basis, we tested the types of ROS produced by PMS self-separation using EPR characterization. As shown in Figure 11b, the Co₆Zn₄-NC+PMS system shows a high-intensity signal belonging to DMPO-X, and this increases significantly from 5 min to 10 min, suggesting that ^•OH and SO₄^•− are generated during the reaction. Figure 11c shows a weak characteristic signal belonging to TEMP-¹O₂ (1:1:1), but its intensity is weaker than that of DMPO-X, and the signal strength does not change significantly during the reaction. However, Figure 11d shows the results of the quenching experiment. Almost no O₂^•− was generated during the reaction.

Based on the existing literature, the electron transfer process constitutes a significant aspect of the non-radical pathway [28]. Furthermore, this electron transfer process is less influenced by time compared to that of short-lived, instantaneous free radicals like SO4^•− and ^•OH. To further explore the contribution of the electron transfer process to the degradation of CAPs in the Co₆Zn₄-NC+PMS system, we pre-mixed catalyst and PMS and added the CAPs at different time nodes. As shown in Figure 12, the CAPs were added at 6 and 9 min after pre-mixing to start the reaction, and the removal efficiency of the CAPs decreased to 85.7%, 73.6%, and 58.2%, respectively, with increased addition time. This is mainly because the short-lived transient radicals generated on the Co₆Zn₄-NC surface are rapidly consumed over time. However, the contribution of the electron transfer process in the Co₆Zn₄-NC+PMS system was very small, leading to a significant decrease in the removal of CAPs [29,30].

As shown in Figure 13, it can be considered that in the process of CAP degradation in the Co₆Zn₄-NC+PMS system, the ROS produced by Co₆Zn₄-NC promoting the self-decomposition of PMS include SO₄^•−, ^•OH, and ¹O₂, in which SO₄^•− plays a leading role, followed by ^•OH and ¹O₂, which also play an auxiliary role, and the contribution of electron transfer process is almost negligible.

Reusability is also an important indicator to evaluate the performance of Co₆Zn₄-NC. Therefore, after CAP degradation, the catalyst was reused five times after cleaning and drying under the same reaction conditions, and the CAP removal efficiency of the Co₆Zn₄-NC+PMS system was tested at the same time point. As shown in Figure 14, there was no significant change in the removal of the CAPs in the first two cycles, while the removal rate of the CAPs decreased slightly to 98.1% in the third cycle. CAP removal decreased to 92.6% in the fourth cycle and further decreased to 88.7% in the fifth continuous cycle. Moreover, the concentration of Co ions in the solution at the end of the reaction was 0.47 mg/L, which indicated that the leaching of Co ions during the reaction was minor and would not cause serious secondary environmental pollution. Based on the experiment results, the main reason for the decrease in the performance of CAP degradation is the reduction in active sites attributed to the leaching of Co ions during the reaction. Furthermore, the organic compounds adsorbed on the catalyst surface during multiple cycles may also reduce the exposed active sites, leading to a decrease in CAP removal. In conclusion, after five cycles, the Co₆Zn₄-NC+PMS system could still remove 88.7% of the CAPs within 60 min, showing good stability and reusability. We also compared the catalytic properties of Co₆Zn₄-NC with those of existing catalysts in the literature. As shown in

Table 2. Comparison of catalytic activity of Co₆Zn₄-NC with previously reported catalysts.

Catalyst	Pollutant (mg/L)	Catalyst Dose (g/L)	Oxone (mmol/L)	Removal Efficiency Kapp (min−1)	Ref.
CoFe2O4 NCs	CAP(10)	0.1	2	0.08	[16]
CuCo@C	CAP(20)	0.1	2	0.12	[31]
Co3O4-BC	CAP(30)	0.3	10	0.34	[32]
CuCoO−1	CAP(20)	0.1	0.5	98%, 15 min	[33]
Fe0	CAP	0.5	0.2	0.03	[34]
Co-NC	CAP(20)	0.1	1	0.07	This work
Co6Zn4-NC	CAP(20)	0.1	1	0.22	This work

, at lower catalyst doses and oxidant concentrations, Co₆Zn₄-NC showed an excellent catalytic performance capable of efficiently degrading CAP in a short period of time, and its catalytic performance was higher than most of the reported catalysts.

4. Conclusions

In this study, doping appropriate amounts of Zn into the Co_xZn_10−x-ZIF structure effectively increased the specific surface area and porosity of Co_xZn_10−x-NC, which promoted interaction between PMS, the CAPs, and the Co₆Zn₄-NC surface. Furthermore, Zn doping significantly promoted the formation of single-atom Co, inhibited the formation of Co nanoparticles, and reduced their size, which improved the activity of the catalysts to a certain extent. CAP degradation experiments showed that CoxZn10-x-NC could effectively remove the low concentration of CAPs in the solution, and the catalytic activity of Co_xZn_10−x-NC gradually increased with the increase in Zn content. The catalytic activity of Co_xZn_10−x-NC was highest when the Co/Zn ratio was 6:4. Using 0.1 g/L of Co₆Zn₄-NC, 1 mM of PMS, and 20 mg/L of CAP, the Co₆Zn₄-NC + PMS system was able to remove 100% of the CAPs within 60 min, and the corresponding k_app was 0.22 min⁻¹. With the further increase in Zn content, the catalytic activity of Co_xZn_10−x-NC decreased gradually, which may be due to the significant decrease in Co content at the main active site. In the Co₆Zn₄-NC+PMS system, the ROS generated by PMS decomposition included SO₄^•−, ^•OH, and ¹O₂, among which SO₄^•− dominated in degrading the CAPs, followed by ^•OH and ¹O₂, which also played an auxiliary role. The contribution of the electron transfer process was almost negligible. In addition, Co₆Zn₄-NC has excellent stability and reusability and can maintain an 88.7% CAP removal rate after five repetitions, showing great application potential. However, the leaching of Co during the reaction process and the high cost of catalyst preparation are still practical issues that limit the further application of Co₆Zn₄-NC. Moreover, since the degradation of CAP in the reaction system is dominated by free radicals, the coexisting ions in the environment have a certain influence on the reaction system.