For the production of polycarbonate and plasticizers, which are used in items like medical equipment and mineral water bottles, bisphenol A (BPA) is a crucial raw element [1]. However, BPA also acts as an endocrine disruptor, making it simple to interfere with the endocrine system of animals and impact the growth and the reproductive ability [2]. Obesity, diabetes, cancer, and other conditions have been linked to BPA [3]. Therefore, it is evident that removing BPA pollution from water is crucial.

The peroxymonosulfate-based advanced oxidation process (PMS-AOPs) has garnered a lot of interest because of its benefits, which include high efficiency and a stable treatment effect [4]. Numerous studies have demonstrated that PMS offers a major benefit in eliminating BPA contamination from water bodies. Huang et al. prepared a Fe doped ZIF-8 derived Fe-N-C single atom catalyst to activate PMS, achieving a BPA removal rate of 94.3% within 10 min. Meanwhile, the system has been proven to be a non-radical oxidation pathway with singlet oxygen (¹O₂) as the main reactive oxygen species (ROS) [5]. Guo et al. used a single atom molybdenum catalyst dispersed on nitrogen doped carbon to catalyze the decomposition of PMS. The system can remove 100% BPA within 20 min [6]. Chen et al. also found that transition metal sulfides could activate PMS to remove BPA, which not only has high degradation efficiency, but also exhibits excellent stability and reusability. The sulfate radical (SO₄^•-) and hydroxyl group (^•OH) generated in the system mainly contribute to the degradation of BPA [7]. Thereby, PMS-AOPs is a promising organic wastewater purification technology.

Usually, PMS-AOPs has two oxidation pathways, namely the radical pathway with SO₄^•-, ^•OH, and other radicals as the primary ROS, and the non-radical pathway with catalyst-mediated electron transfer, high-valent metal or ¹O₂ as the main oxidants [8-10]. Among them, radicals have strong oxidation ability, which would simultaneously remove other organic substances in the water except for the target pollutants [11]. This could lead to the more consumption of PMS and an increase in sewage treatment costs. While, compared to the former, non-radical pathways have high selectivity for the specific pollutants removal and can effectively avoid the consumption of additional PMS [12]. Furthermore, the non-radical pathway offers additional benefits for the high-salt wastewater environment [13-15]. Therefore, the development of catalyst with high efficiency PMS activation led by non-radical routes is significant in the realm of water treatment.

In non-radical pathways, carbon-based catalyst function as efficient carriers and activators. Literature show that PMS can be activated by graphene, carbon nanotubes and other materials [16-18]. Nevertheless, compared to metal catalyst, carbon-based compounds have less catalytic activity. Hence, carbon-based catalyst can be added to transition metals for modification to increase catalytic activity [19, 20]. Catalyst performance on carbon-based materials can be greatly enhanced by transition metal loading. Cobalt-based catalyst in particular have garnered a lot of attention and use due to their high catalytic activity on PMS [21-24].

N-Doping is another way that carbon-based catalyst can function better. In general, the N-containing group can interact with the sp² hybridized carbon backbone via topological deviation and is anchored at defect sites through covalent bonds. Nitrogen atoms with lone pair electrons exhibit high electron mobility, inducing electron rearrangement, thereby forming electron-rich and electron-deficient centers, which promotes the activation of PMS. In addition, pyridine N defects can directly act as electron donors to capture highly electrophilic peroxide molecules, allowing them to adsorb onto adjacent positively charged carbon atoms, thereby facilitating the activation of peroxides [25-28]. Meanwhile, nitrogen doping could also enhance the interaction between PMS and carbon-based catalyst, promote electron transfer and accelerate the production of ROS, hastening the catalytic degradation process [29, 30]. Therefore, N doping is an effective way to improve the catalytic activity of carbon-based catalyst.

In this paper, a nitrogen-doped graphite coated cobalt-based carbon nanosphere catalyst was prepared by solvothermal-calcination approach (the experimental part is in Texts S1-S5 in Supporting information). First, the calcination temperature of the catalyst has been optimized. Subsequently, the reaction temperature, catalyst quantity, PMS amount, and starting pH were all tuned. Additionally, the effects of some typically coexisting inorganic anions have also been discussed. Simultaneously, the characteristics and surface functional groups of the catalyst were demonstrated by some characterization methods. In addition, the removal of different pollutants, practical application and catalyst stability were also investigated. Electron paramagnetic resonance (EPR), chronoamperometry, and quenching experiments were used to study the primary ROS in the system. Next, the activation mechanism of PMS was proposed by density functional calculation (DFT). Ultimately, liquid chromatography-quadrupole time-of-flight mass spectrometry (LC-QTOF-MS/MS) provided insight into the BPA breakdown process.

Investigations were conducted on the BPA removal rate, specific surface area (S_BET) change, phase change, degree of graphitization, magnetic characteristics, and metal ion leaching of catalysts synthesized at various calcination temperatures. The aforementioned results analysis and discussion (Fig. S1 and Text S6 in Supporting information) indicate that raising the calcination temperature can improve the catalyst specific surface area, speed up its graphitization, and enhance its capacity to remove BPA from water. It is also capable of successfully stopping metal ion leaching. Nonetheless, we determine that the ideal calcination temperature for catalyst preparation is 900 ℃, taking into account both the cost of preparation and catalytic performance.

Scanning electron microscopy (SEM) was used to analyze the micromorphology of the prepared Co@C-N-900 catalyst (Figs. 1a–d). It is evident that the majority of the prepared Co@C-N-900 catalyst is made up of regular spherical particles ranging from 100 nm to 1 µm. Furthermore, high resolution transmission electron microscope (HR-TEM) was used to evaluate the manufactured catalyst to further confirm the structure of the Co@C-N-900 catalyst (Figs. 1e–i). It is evident that the Co@C-N-900 catalyst is carbon sphere structures formed by wrapping metal particles in multiple layers of graphite. At the same time, the lattice fringes with dimensions of d = 0.338 nm and d = 0.205 nm are found, which are associated with the (002) crystal face of the graphite layer and the (111) crystal face of the metallic Co⁰. It demonstrated that the Co@C-N-900 consists of graphite carbon layer and metal Co⁰ nanoparticles, which is aligned with the XRD results (Text S6).

Simultaneously, the element mapping illustrated that the catalyst surface is uniformly distributed with a large number of C and Co elements, as well as N and O elements (Fig. 1j). The N element come from N-containing precursors (2-MI), suggesting the existence of N doping. Meanwhile, the presence of O element is speculated to be due to the adsorption of oxygen-containing substances or oxygen-containing functional groups on the catalyst surface. In addition, the presence of trace amounts of Zn element was also detected. However, HR-TEM did not detect the presence of substances containing Zn. The reason might be that the sublimation of Zn caused by high-temperature, resulting in a decrease in the content of Zn. The above analysis proved that a N-doped graphite carbon sphere coated cobalt nanomaterial catalyst has been successfully prepared.

X-ray photoelectron spectroscopy (XPS) was used to investigate the surface element compositions and valence states of the Co@C-N-900 catalyst following the fresh, single used and four-times used. The entire spectra diagram (Fig. S2a in Supporting information) shows that the three samples had five signal peaks on their surface, which belong to C 1s, N 1s, O 1s, Co 2p, and Zn 2p (Figs. 2a–c, Figs. S2b and c in Supporting information). The N 1s spectrogram of the fresh, single used and four-times used Co@C-N-900 (Fig. 2a). Three peaks were obtained through spectral fitting, corresponding to pyridine N (398.55 eV), graphite N (401.00 eV), and oxide N (404.21 eV), respectively [31, 32]. Specifically, the production of oxidized N under N₂ atmosphere might come from the decomposition of nitrate, which introduces oxygen elements. Besides, after four cycles of use, the content of pyridine N and graphite N on the catalyst surface significantly decreased (Fig. S3a in Supporting information), while distinctive pyrrole N (400.50 eV) was generated [33]. This also indicates the successful doping of N elements, and suggests that N-doped sites on graphite will be the main site for PMS activation or BPA adsorption. Similarly, the O 1s spectra are displayed in Fig. 2b. The O 1s spectra display two prominent peaks attributed to C=O bond (531.60 eV) and H₂O molecules that had been adsorbed on the catalyst surface (533.05 eV) [34, 35]. At the same time, through XPS full spectrum, it can be observed that the intensity of the O 1s is enhanced after the use. This would be caused by an increase in H₂O molecules adsorbed on the catalyst surface. The conclusion is further corroborated by the rising proportion of peaks at 533.05 eV in the O 1s spectrums. After multiple cycles of the catalyst, the proportion of water molecules adsorbed on the catalyst surface increases (39.26% → 76.07%), while the original C=O bond content decreases (60.74% → 23.93%) (Fig. S3b in Supporting information). Subsequently, the Co 2p spectrums of the fresh, single used and four-times used Co@C-N-900 are discussed (Fig. 2c). All samples contain two peaks, which are divided into Co⁰ located at 778.54 eV and Co²⁺ located at 780.70 eV [36]. The presence of Co⁰ is consistent with the results of XRD and HR-TEM, while Co²⁺ may be due to strong interactions between Co atoms and C atoms at the contact interface between graphite layer and metal Co, which would result in charge reshuffling [37]. Obviously, the Co element signal changes slightly compared to the fresh catalyst and the reused catalyst (Fig. S3c in Supporting information). It suggests that the coating of in-situ graphite layer can effectively inhibit the oxidation of cobalt nanoparticles. Besides, although C and Zn are also detected, their modifications have minimal impact on the catalyst, so they are not covered in detail in this research (Texts S7 and S8, Figs. S2b and c, and Fig. S3d in Supporting information). The above research evidences that the N-doped sites on the catalyst surface may be responsible for PMS adsorption and activation. Meanwhile, the strong interaction between Co⁰ particles and the carbon layer enhances the electron transfer from Co⁰ to the carbon layer, which will be beneficial for PMS activation.

The surface functional groups of the fresh, single used and four-times used Co@C-N-900 were characterized by Fourier transform infrared spectroscopy (FTIR) (Fig. 2d). The bending vibration of C=O/C-N is represented by the absorption peak of 1187 cm^-1, and the tensile vibration of -C-N is represented by the absorption peak of 1386 cm^-1 [38, 39]. The successful synthesis of nitrogen-doped carbon materials can be indicated by two absorption peaks, which is also compatible with the findings of the XPS investigation. The absorption peaks at 1500-1700 cm^-1 are the stretching vibration of -C=C [40]. For -CH₂, the absorption peaks at 2855 cm^-1 and 2923 cm^-1 correspond to distinct vibration modes [41]. The stretching vibration mode of -OH in H₂O is represented by the absorption peak at 3425 cm^-1 [42]. It is evident from the aforementioned research that there is almost no change in the functional groups found in the fresh, single used and four-times used of Co@C-N-900, which indicates some stability for the Co@C-N-900. Notably, the alkalinity/acidity gas molecules (NH₃/CO₂) will be adsorbed differently by the acidity/alkalinity sites on the catalyst surface. Thus, the acidity/alkalinity adsorption sites on the surface of Co@C-N-900 catalyst were tested using CO₂-TPD and NH₃-TPD (temperature-programmed desorption experiment) (Figs. 2e and f). The TPD of CO₂ and NH₃ indicates that desorption signals occur between 50-600 ℃. In other words, there are both acidic and alkaline sites on the surface of Co@C-N-900. Furthermore, research has demonstrated that the Co@C-N-900 acidity/alkalinity sites are favorable for PMS activation and adsorption [43, 44]. Additionally, the fresh, single used and four-times used Co@C-N-900 were then identified using Tafel scanning in order to investigate the electron transfer rate and catalyst free corrosion potential, respectively (Fig. S4 in Supporting information). The largest electron transfer rate is seen in the fresh Co@C-N-900 because it has the lowest corrosion potential [45]. But after use, the catalyst rate of electron transfer drops. This can be because there is less graphene N on the catalyst surface following the reaction, which prevents electron transfer from occurring there [46].

Based on the above results, the N-doped graphite carbon nanosphere coated cobalt nanoparticles catalyst has been successfully synthesized. The strong interaction between Co⁰ nanoparticles and graphite layer enhances the electron transfer. Meanwhile, the changes in N species on the catalyst surface suggest that these N doping sites are the main sites for PMS adsorption and activation. And the change of N species will affect the surface charge transfer ability of the catalyst, thereby affecting the catalytic performance. Additionally, the prepared Co@C-N-900 contains acid-base sites that facilitates PMS activation and adsorption. This suggests that N doping may be the main reason for the formation of acidic and alkaline sites on the catalyst surface. Otherwise, there was no significant change in the catalyst before and after the reaction, indicating that the prepared Co@C-N-900 has exceptional stability and dependability.

The effects of various operating parameters on the degradation of BPA in the Co@C-N-900/PMS were investigated. These variables include reaction temperature, pH, and the quantity of catalyst and PMS (Figs. 3a–d). When the catalyst concentration was adjusted from 0.25 to 0.45 g/L, the removal effectiveness of the Co@C-N-900 for BPA was enhanced. The presence of PMS causes the PMS to quickly become reactive oxygen species (ROS), which causes the BPA to be removed quickly. However, it is challenging to eliminate significant amounts of BPA from the system with additional catalyst dosage increases. The process efficiency may be decreased because the leftover metal sites may eliminate the free radicals generated at this point. The k_obs value increased from 0.177 min^-1 to 0.367 min^-1 at the same time that the catalyst amount increased from 0.25 g/L to 0.45 g/L (Fig. S5 in Supporting information). Ultimately, the following analysis employed Co@C-N-900 at a dosage of 0.35 g/L in light of the resource and energy consumption issues. PMS plays a crucial role in the degradation process as the primary source of active chemicals. The elimination efficiency of the Co@C-N-900 for BPA was initially markedly raised and then essentially stable with the adjustment in PMS dosage (0-0.125 g/L). Nevertheless, excess free radicals produced in a brief amount of time may be absorbed by self-quenching after PMS dosage is increased further. Consequently, an overabundance of PMS does not promote the breakdown of BPA [32]. The value of k_obs increased dramatically from 0.081 to 0.297 min^-1 with an increase in PMS dosage (Fig. S6 in Supporting information). In conclusion, PMS at a dosage of 0.075 g/L was employed in the subsequent analysis. Simultaneously, the impact of a gradient beginning pH value on the reaction process was also examined. The system BPA removal rate increased initially before declining as the gradient of the initial pH value increased. The reaction rate constant k_obs rises from 0.223 min^-1 to 0.297 min^-1 as the starting pH value rises from 2 to 6.5. The reaction rate constant k_obs drops from 0.297 min^-1 to 0.202 min^-1 when the starting pH value rises from 6.5 to 10 (Fig. S7 in Supporting information). Variations in the starting. pH did not substantially impact the BPA breakdown for the duration of the experiment. BPA was successfully eliminated after 30 min of treatment. The Co@C-N-900 can therefore be used in a variety of pH applications (pH 2-10). Additionally, the zero point charge (pH_zpc) on the Co@C-N-900 surface was investigated using the zeta potential method in order to investigate the impact of catalyst surface charge on the adsorption of BPA (Fig. S8 in Supporting information). Text S9 (Supporting information) goes into great detail about it. By adjusting the reaction temperature (10-40 ℃), the impact of temperature on the reaction process was investigated. The degradation rate of BPA in the Co@C-N-900/PMS system rose with temperature, rising from 95.39% to 97.86%. The rate constant of the degradation process (Fig. S9 in Supporting information) shows that when the reaction temperature rises from 10 ℃ to 40 ℃, k_obs increases from 0.190 min^-1 to 0.262 min^-1. In other words, the Co@C-N-900 capacity to activate PMS grew steadily as reaction temperature rose. Using the Arrhenius equation, the activation energy and degradation of BPA in the Co@C-N-900/PMS system (E_a = 5.31 kJ/mol) was determined (Fig. S10 in Supporting information). The benefits of the degradation process increase with decreasing activation energy. Stated differently, BPA degrades more quickly at higher temperatures. In addition, compared with other currently reported Fenton-like catalysts, Co@C-N-900 in this work also has a higher specific k-value (63.11 mol L^-1 min^-1) (k-value, multiply k_obs by pollutant concentration, then divide by catalyst dosage and PMS dosage) (Fig. 3e and Table S2 in Supporting information).

The degradation of BPA may be impacted by a variety of inorganic anions when administered in the real water environment. Consequently, the effects of four common inorganic anions (NO₃^-, HCO₃^-, Cl^-, NO₃^-, and H₂PO₄^-) on the breakdown of BPA were investigated (Fig. S11 in Supporting information). It is evident that the system is not significantly affected by the addition of the four common inorganic anions listed above. This demonstrates that the Co@C-N-900/PMS system may be used successfully in environments with high salinity. In addition, the adsorption and degradation of three distinct pollutants with 0.02 g/L (BPA, 2, 4-DCP, and CBZ) in the Co@C-N-900/PMS system were studied (Fig. S12 in Supporting information). It is evident that CBZ in the Co@C-N-900/PMS system is not easily degraded, while BPA and 2, 4-DCP have excellent degradation efficiency. This infers that the Co@C-N-900/PMS system exhibits selectivity in the oxidation of pollutants. Therefore, it could reduce the consumption of PMS on non-target organic matter, which can improve the selective utilization rate of PMS in practical applications [47]. Furthermore, studies were conducted on BPA removal in several real water bodies (rivers water, lakes water, and tap water). The sampling locations of lake water and river water are shown in Fig. S13 (Supporting information). As illustrated in Fig. S14 (Supporting information), the removal rate of BPA in lake water and tap water is basically the same. In river water, although the competition of organic matter for ROS in the system reduces the BPA degradation, the removal rate of BPA still reaches 92.68% after 30 min of reaction. The above analysis indicates that the Co@C-N-900/PMS system has excellent performance in different water bodies. To verify the good effect of the prepared material, control experiments were conducted (Fig. 3f). The dosage of Zn²⁺ and Co²⁺ was determined by the leaching concentration of water after reaction in Co@C-N-900/PMS system (Fig. S1f). The experimental results of Co²⁺/PMS system and Zn²⁺/PMS system showed that the removal amount of BPA in the leaching homogeneous system was very low, which demonstrate that the leaching is basically not affected by the system, and the catalytic effect comes from the heterogeneous catalysis of the prepared catalyst. The reusability of heterogeneous catalyst in applications must be taken into account. Therefore, cyclic tests are used to study the stability of the Co@C-N-900 (Fig. 4a). It is evident that the catalyst that was utilized for the first time performed incredibly well. Following the 30 min reaction, almost 97% of BPA was eliminated. After use, the catalyst is passed through a sand core filter before being immediately added to the following cycle after being cleaned in turns with organic solvent and deionized water. The findings demonstrated that the subsequent three cycles had BPA elimination rates of 91.87%, 88.48%, and 81.14%. To sum up, the Co@C-N-900 catalyst exhibits good stability and robust catalytic activity. In addition, continuous flow reaction experiments have found that, Co@C-N-900/PMS after 50 h of continuous operation, the removal rate of PMS remains above 60% (Figs. 4b and c). This further confirms the ability of the prepared catalyst to continuously mineralize pollutants.

The primary ROS in the system must be identified to investigate the PMS activation mechanism. Firstly, quenching studies were carried out to identify the primary ROS that BPA destroyed in the Co@C-N-900/PMS system [48-51]. After the addition of MeOH or TBA into the Co@C-N-900/PMS system, the BPA removal rates still reached 95.72% and 96.74% (Fig. 4d). It is demonstrated that the system is not significantly quenched by MeOH or TBA. That is to say, ^•OH and SO₄^•- are not the primary ROS in the Co@C-N-900/PMS system. On the other hand, when p-BQ was introduced to the Co@C-N-900/PMS system, the clearance rate of BPA was 80.76%. It suggests that the system might contain some O₂^•-. Meanwhile, the addition of FFA resulted in a 56.54% clearance rate of BPA. Significant quenching effect was observed, indicating the presence of a large amount of ¹O₂ in the Co@C-N-900/PMS system. To further confirm the presence of ¹O₂ in the system, the quenching experiment of β-carotene on ¹O₂ in pure methanol solvent was investigated. The findings disputed that the presence of β-carotene in methanol results in a removal rate of only 4.11% for BPA in the system. Specifically, degradation experiments of BPA without β-carotene in pure methanol and physical adsorption of the Co@C-N-900 were also examined. The physical adsorption results showed that the prepared Co@C-N-900 removed about 4.05% of BPA from the system by adsorption in pure methanol. However, the BPA removal rate in pure methanol using the Co@C-N-900/PMS system was around 85%. Overall, it can be concluded that β-carotene in pure methanol completely quenched the ROS in the Co@C-N-900/PMS system, resulting in almost no degradation of BPA. Therefore, it can be inferred that the main ROS in the Co@C-N-900/PMS system is ¹O₂. The quenching effect of p-BQ may be due to the reaction between reducing p-BQ and PMS in the system, which consumes some PMS and reduces the removal efficiency of BPA.

Utilizing electron paramagnetic resonance spectroscopy (EPR), the primary ROS in the Co@C-N-900/PMS system was examined additional. The agents utilized to trap ^•OH, SO₄^•- and O₂^•- were DMPO and the agent used to capture ¹O₂ was TEMP [52]. As seen in Fig. 4e, no signals were detected for the addition products of ^•OH, SO₄^•- or O₂^•- to DMPO were detected, indicating that ^•OH, SO₄^•- and O₂^•- do not exist in the Co@C-N-900/PMS system. Meanwhile, a three-wire EPR signal of equal strength is recognized, which is TEMP-¹O₂ composed of the sum of TEMP and ¹O₂. Thus, it is evident from the analysis of the aforementioned ROS verification experiment that the primary ROS in the Co@C-N-900/PMS system is simply ¹O₂.

Based on the above analysis, combined with the influence of inorganic anions and the high selectivity of the system for pollutant mineralization, it can be inferred that the system belongs to a non-radical pathway. Meanwhile, XPS results have ruled out the presence of high valence metals. Therefore, it is necessary to further exclude the non-radical pathway with catalyst-mediated electron transfer. The chronoamperometry is used to examine whether the Co@C-N-900/PMS system is an electron transfer system (electrochemical measurements in Text S10 in Supporting information). As illustrated in Fig. 4f, the current is steady at 300 s, and it clearly trends downward as PMS is introduced into the system. Afterwards, when the current reaches a steady state again, the system is supplemented with BPA. At this point, due to the instantaneous change in system concentration, there is a slight fluctuation in the current, but it soon returns. Overall, the addition of BPA had no effect on the current of the system. It indicates that there is no formation of metastable complexes between the Co@C-N-900 and PMS in the system. In other words, non-radical pathway with catalyst-mediated electron transfer is not the main cause of BPA degradation in the Co@C-N-900/PMS system [39]. Overall, the Co@C-N-900/PMS system is a non-radical pathway dominated by ¹O₂.

The influence of different N doping on the surface charge distribution of catalyst was investigated through charge difference distribution (Text S11 in Supporting information) [53]. The catalyst doped with graphite N, and pyridine N and nitrogen oxide are defined as C-BDN, C-SMN, and C-YHN. As shown in the Fig. S15 (Supporting information), the N atom in C-BDN or C-SMN is surrounded by a large number of electrons, while the C atoms adjacent to the N atom exhibit an electron deficient state. That is to say, N doping causes charge rearrangement on the catalyst surface. The electrons in the surrounding C atoms transfer to the N atom, causing the N atom to become a basic site for charge accumulation, while the C atom that loses electrons forms an acidic site for positive charge accumulation. Similarly, nitrogen oxide doping (C-YHN) also produces such an effect. However, those electrons lost by the C atom in C-YHN are not only concentrated around N atoms but also distributed around O atoms. The above results are consistent with the previous TPD results, indicating that the main alkaline sites on the catalyst surface are N doping sites, while the acidic sites are C atoms around N defects. Furthermore, the specific number of electron transfers was obtained through the Bader method. The electrons obtained by the N atom or N oxide group in C-BDN, C-SMN, and C-YHN is 1.145 e, 1.105 e, and 1.712 e, respectively. Usually, the dissociated PMS is a negatively charged ion, which is prone to electrostatic adsorption with acidic sites on the catalyst surface, thereby facilitating the adsorption and activation of PMS. Thus, the above results suggest that N-doping effectively modulates the surface charge distribution of the catalyst, creating favorable sites for PMS adsorption and activation.

The adsorption model of PMS on the catalyst surface following geometric optimization is depicted in Fig. 5a. On the surfaces of C-BDN, C-SMN, and C-YHN, PMS had an adsorption energy (E_ads) of -1.05 eV, -0.10 eV, and -4.56 eV. Among them, the N oxide doped catalyst has the highest E_ads, while graphite N doped catalyst has the lowest E_ads. In general, the bigger the E_ads, the better the adsorption performance [54]. Therefore, the N oxide doped catalyst are most likely to adsorb PMS. This might be due to the accumulation of more negative charges in N oxide group, resulting in stronger electrostatic interactions with PMS. Meanwhile, the charge transfer of PMS adsorbed on the catalyst surface was analyzed by charge difference distribution (Fig. 5b). Clearly, PMS is surrounded by negative charges, while the catalyst surface in contact with PMS accumulates positive charges. That is, electrons move from the N-doped graphite layer on the surface towards the adsorbed PMS direction. Moreover, the electrons gain and loss of PMS adsorbed on the surface of three distinct doped catalyst were determined using Bader method. The findings demonstrate that after adsorption on C-BDN, C-SMN, and C-YHN, PMS can receive 0.48 e, 0.66 e, and 0.51 e. The electrons obtained by PMS decrease with the increase of E_ads. This infers that excessive adsorption is not conducive to PMS activation, and the catalyst doped with graphite N has the best electron transfer capacity. Simultaneously, the peroxide O-O bond length (l_O-O) in PMS adsorbed on the catalyst surface was obtained. The findings demonstrate that the l_O-O of PMS adsorbed on the surfaces of C-BDN, C-SMN, and C-YHN is 1.416 Å, 1.439 Å, and 1.410 Å, in that order. But, the original l_O-O of PMS is 1.345 Å. It indicates that the peroxide O-O bond in PMS becomes longer after receiving electrons. At this point, the PMS is highly active, and could undergo contact with nearby PMS to produce ¹O₂ [55]. Obviously, among the three types of N-doped catalyst, graphite nitrogen doped catalyst provide more electrons, making the peroxide O-O bond longer and easier to convert into ROS. Therefore, graphite N doping sites are the optimal sites for PMS activation.

Based on the above analysis, it can be concluded that the successful doping of N on the graphite layer effectively regulates the acid-base sites on the catalyst surface, which is beneficial for PMS adsorption and activation. Although there are differences in the adsorption performance and charge transfer quantity of different types of N doping, they are all beneficial for the adsorption and activation of PMS. The graphite N-doped sites may be the optimal PMS activation sites on the catalyst surface. Finally, based on the DFT calculation results, the PMS activation mechanism is proposed: After PMS is adsorbed on the N-doped graphite layer on the catalyst surface, electrons are transferred from the catalyst to PMS, resulting in the elongation of the peroxide bond of PMS. As of right now, the elongated peroxide bonds in PMS are very active and can be converted into ¹O₂ when it contacts with another PMS.

To better understand the BPA degradation in the Co@C-N-900/PMS system, intermediate products were analyzed using LC-QTOF-MS/MS. As shown in Fig. 6, the Co@C-N-900/PMS system had four different types of intermediates (Figs. S16-S20 in Supporting information). Two BPA degradation pathways in the Co@C-N-900/PMS system were hypothesized based on the four identified intermediates. Pathway 1 is the β-scission in BPA: ¹O₂ route attacking the isopropyl group located between the two phenyl groups [56]. Thus, 4-(2-hydroxypropan-2-yl)phenol (Product-1, C₉H₁₂O₂), is produced. Pathway 2: ¹O₂ leads to the oxidation of hydroxyl groups on the aromatic ring of BPA, followed by further β-scission [57]. The oxidation of hydroxyl groups on the aromatic ring of BPA produces bene-1, 2-diol (Product-2, C₁₅H₁₆O₃). Further β-scission of bene-1, 2-diol produces intermediate products gentisyl alcohol (Product-3, C₇H₈O₃) and pyrogallol (Product-4, C₆H₆O₃). These intermediate products further react with ¹O₂ and eventually mineralize into H₂O and CO₂ for removal [58]. The aforementioned investigation indicates that BPA can eventually break down into H₂O and CO₂ regardless of the method used to degrade it, removing BPA from water.

BPA is a crucial raw element for synthetic plasticizers. However, BPA is a known endocrine disruptor that can readily alter an animal endocrine system, endangering both humans and the environment. BPA needs to be taken out of water immediately. In this work, solvothermal roasting is used to create the nano-carbon ball material catalyst Co@C-N-900, which is created by transition metal particles coated with N-doped graphite and activated by PMS to remove BPA from water. The findings indicated that the breakdown rate of BPA could reach around 98% in 30 min at 0.35 g/L Co@C-N-900, 0.075 g/L PMS, pH 6.50, and reaction temperature of 30 ℃. Additionally, the features of Co@C-N-900 were demonstrated by a number of characterizations and DFT calculations. It has been proved that the coating of N-doped graphite layer can improve the charge transfer ability of catalyst surface, promote the activation ability of PMS, and ensure the good catalytic activation performance. Concurrently, the Co@C-N-900/PMS system has been effectively applied to high salinity environment, and great performance in a variety of water bodies. It has also been successfully used to high salinity environments. Combined with the results of free radical quenching experiment, EPR and I-t curve, the mechanism of Co@C-N-900/PMS system was explored by using ¹O₂ as the main ROS non-free radical pathway. Ultimately, two possible BPA degradation pathways were identified using LC-QTOF-MS/MS data. The first involves ¹O₂ attacking the isopropyl group between two phenyl groups in BPA, leading to β-scission, and the second involves ¹O₂ causing additional β-scission following the oxidation of the hydroxyl group on the aromatic ring of BPA.

As can be observed, the Co@C-N-900 has a great deal of potential for use in the treatment of organic wastewater since it can efficiently activate PMS to breakdown BPA in water bodies. The Co@C-N-900/PMS system furthermore offers a crucial point of reference for non-radical degradation.

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

Huiyuan Deng: Writing – original draft, Validation, Methodology, Investigation, Data curation, Conceptualization. Na Zhao: Methodology, Investigation. Junjie You: Writing – review & editing, Writing – original draft, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Zhicheng Pan: Funding acquisition. Bo Xing: Conceptualization. Yuling Ye: Conceptualization. Bo Lai: Conceptualization. Yuxi Wang: Methodology, Investigation. Tongrui Lu: Methodology, Investigation. Xiaonan Liu: Writing – review & editing, Funding acquisition.

Acknowledgments

The authors would like to acknowledge the financial support from Sichuan Science and Technology Program (No. 2023NSFSC0847), Scientific Research and Innovation Team Program of Sichuan University of Science and Technology (No. SUSE652A003), Talent Introduction Project of Sichuan University of Science and Engineering (No. 2021RC03), Talent Introduction Project of Sichuan University of Science and Engineering (No. 2021RC05), the Undergraduate Training Program for Innovation and Entrepreneurship (No. CX2024042), The Innovation Fund of Postgraduate, Sichuan University of Science & Engineering (No. Y2024094).

Supplementary materials

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110650.