Preparation of Co-Doped Graphitic Carbon Nitride and Degradation of MB by Heterogeneous Photo-Fenton-like Reaction System

: In this study, Co-porous-g-C 3 N 4 (porous Co-Doped graphitic carbon nitride) was synthesized by a one-step calcination method successfully. Co-porous-g-C 3 N 4 was characterized by transmission electron microscopy, X-ray photoelectron spectroscopy, UV-vis diffuse reflectance, and specific surface area determination. The Photo- Fenton-like system was successfully constructed under the condition of coexistence of Co-porous-g-C 3 N 4 and H 2 O 2 . The synergistic effect of Fenton-like system and photocatalysis was realized. Through the optimization of experimental conditions, the best photocatalytic performance of this heterogeneous reaction system was achieved. The degradation rate of MB (methylene blue, 40 mg/L) could reach 98.3% at 90 min. Co-porous-g-C 3 N 4 had photocatalytic degradation activity for different organic pollutants and had wonderful recyclability and stability. The main degradation effect was attributed to ·O 2 - which could be generated through two channels in the Photo- Fenton-like system. Moreover, the rate of photogenerated electron-hole complexation could be decreased due to the reaction of low-valent cobalt and photoelectrons in this heterogeneous reaction system.


1.Introduction
Photocatalytic technology, an excellent oxidation technology, was widely used in disposing organic pollutants [1], and typical photocatalysts are normally divided into metal oxide [2][3][4], selenide and sulfide [5][6][7], and non-metallic semiconductor [8][9]. However, there are still some limits in practical application, due to the inherent properties of semiconductor. Notably, photocatalytic technology has good cooperativity with other advanced oxidation technologies, for example Fenton oxidation technology [10]. In a nutshell, Photo-Fenton reaction system would be constructed by visible-light, hydrogen peroxide and photocatalysts that was modified by transition metal, such as Fe, Co, Cu.
Graphitic carbon nitride (g-C3N4), a green and excellent photocatalyst, possesses many advantages such as visible light response, environmental friendliness, suitable band gap (2.7 eV) and stability [11]. It is widely used in many fields as organic pollutants treatment [12], H2 evolution [13], harmful gas treatment [14], supercapacitors [15], N2 fixation [16] and so on. However, g-C3N4 has some defects that can limit the application. But these defects can be improved by various modification methods, such as metal and non-metal doping [17][18], Z-scheme and S-scheme heterojunction [19][20], polymerizing g-C3N4 with other semiconductors [21][22] and precursor pretreatment [23]. It is worth noting that the ionic radius of transition metals and g-C3N4 are similar, so the transition metal doping is beneficial to g-C3N4 to form more active sites and improve the photocatalytic ability. Therefore, the transition metal doping is one of the best modification methods.
Transition metal (cobalt) doping was used in various semiconductor composites. The physical and chemical properties of cobalt element are similar to iron and nickel elements. Wook Ahn et al. [24] designed a ternary transition (Ni, Co, Fe) metal-doped porous organic framework electrocatalyst (NCF-MOF). The novel NCF-MOF electrocatalyst exhibited a well-coordinated electronic structure, which has a larger active specific surface area and more mass transfer pathways.
Jianping Ai et al. [25] synthesized Co-ZnO powder by hydrothermal method and Co doping which possessed high crystallinity and magnetism at room temperature. The photocatalytic efficiency of Co-ZnO was studied, the methyl orange degradation reached 89.8% at 3 h.
At the meantime, graphitic carbon nitride also can be modified by cobalt element. Co-doping can not only retain the advantages of g-C3N4, but also can make up for the defects of g-C3N4. Yalin Yu et al. [26] prepared cobalt-oxygen co-doped graphite carbon nitride (Co-OCNVN) by thermal polymerization method. The conclusion showed that the cobalt element existed in the form of cobalt-nitrogen bond (Co 2+ -Nx), instead of cobalt oxygen compound (CoxOy). The photocatalytic had a higher activity for the degradation of Bisphenol A (BPA). Also, D. Das et al. [27] synthesized a cobalt-doped graphite carbon nitride (CoGCN) photocatalyst through a thermal polymerization process. Although, the structure of g-C3N4 was not changed due to the doping of cobalt, the results proved that the specific surface area of CoGCN and the degradation of Eosinbluish increased. However, the layered structure of g-C3N4 has not been changed by simple cobalt-doping, which led to little promotion of photocatalytic ability. In addition, the precursor modification is also a advantageous way to optimize the photocatalytic ability of g-C3N4.
Herein, we intend to prepare Co-g-C3N4 by hydrochloric acid (37%) and cobalt chloride, which possess poriferous microstructure and excellent photocatalytic ability. Under the combined action of Cog-C3N4, visible light and hydrogen peroxide (30%), the Fenton-like system was established. And the degradation effect of methylene blue was investigated in different conditions. As a consequence, cobaldoping with the auxiliary of hydrochloric acid (%) is an effective pathway to dispose organic pollutants.

Preparation of porous Co-g-C3N4 photocatalyst
The poriferous Co-g-C3N4 was synthesized as follows: 2.5g melamine was dispersed in hot deionized water under magnetic stirring, and then 37% hydrochloric acid was dropped into the solution which contained melamine slowly. The mixture solution naturally cooled at room temperature. An appropriate amount of cobalt chloride was added into the mixture under magnetic stirring and maintained 30 min to keep homogeneous, the precursor solution was prepared. Subsequently, the precursor solution was dried and then put into a crucible with cover. This crucible was annealed at 500℃ for 2 h and 520℃ for 2 h in air atmosphere with the heating rate of 5℃·min -1 . After cooling to room temperature, the photocatalysts were prepared. The photocatalyst treated with hydrochloric acid was recorded as Co-porous-g-C3N4, and the photocatalyst not treated with hydrochloric acid was recorded as Co-g-C3N4.

Photocatalytic exploration Experiment of Co-porous-g-C3N4
Photocatalytic optimization analyses were carried out on a photocatalytic reactor, equipped with a steel shell, a magnetic stirrer, a condenser equipment and a 500 W xenon lamp. A certain amount of asprepared Co-porous-g-C3N4 was added into the cylindrical quartz tube that included 50 mL methylene blue reaction solution under magnetic stirring. After stirring in dark for 40 min, the reaction system switched on the xenon lamp to simulate visible light, and then added 30 % H2O2. During the process of continuous irradiation, 5 mL of the reaction mixture that was filtered by 0.45 µm water-based membrane system was drew at a given illuminated time intervals and then the concentration of MB was monitored by a UV-vis spectrophotometer at 664 nm. The degradation effect of Co-porous-g-C3N4 would be reflected in the concentration ratio of MB in different time.

Characterization of Co-porous-g-C3N4
In order to analyse the internal composition of as-synthesized Co-porous-g-C3N4, the photocatalyst had been characterized by several ways, such as X-ray diffraction, UV-Vis DRS spectra, X-ray photoelectron spectroscopy, transmission electron microscope and so on. XRD (D/MAX-2500/PC) patterns of g-C3N4 and Co-porous-g-C3N4 were revealed in Figure 1, which 2θ ranged from 10° to 70°. There are two distinct peaks at 12.9° and 27.4°. These two peaks are related to the (100) inter-planar stacking and (002) inter-planar stacking of g-C3N4. Compared to g-C3N4 and Co-porous-g-C3N4, the positions of characteristic peak were basically the same, but the peak intensity of Co-porous-g-C3N4 was stronger due to the high crystalline degree of polymer. This may be due to the addition of hot water and hydrochloric acid during the melamine treatment, which will partially hydrolyze melamine into cyanuric acid. The presence of cyanuric acid is more conducive to the polymerization of g-C3N4.  (Figure 2c-d) images of Co-g-C3N4 reveal distinct porous structure that distribute regularly at the whole photocatalyst, indicating that the Co-porous-g-C3N4 photocatalyst possesses larger surface area and more active sites compared with g-C3N4 which was prepared in previous study [28]. Moreover, there are no distinct black particles of metals or metal compounds that was observed by the images of Co-porous-g-C3N4, the result Corresponding to the analysis of XRD.  The corresponding element composition of Co-porous-g-C3N4 was characterized by EDS (JSM-6700F) analysis. As showed in Table 1, the C, N, O and Co atoms were observed, the C and N originated from g-C3N4 and then the Co originated from cobalt chloride added during the photocatalyst preparation process. Besides, Co element content is 0.61wt%. The result indicated that the cobalt-doped porous g-C3N4 photocatalyst was successfully synthesized. For ascertaining the chemical states and the connection states of element, XPS (JSM-6700F) spectra of Co-porous-g-C3N4 was studied in Figure 3. As showed in Figure 3, the existence of C 1s, N 1s and Co 2p have been researched in detail. There are two distinct peaks of C 1s located at 285.3 and 288.5 eV in Figure 3a, which ascribe to sp 2 -hybridized C-C and triazine ring N-C=N, respectively. In Figure  3b, three peaks of N 1s spectrum were studied, that located at 398.2, 399.5 and 400.6 eV, their energies correspond to sp 2 -hybridized nitrogen (N-C=N), sp 3 -tertiary nitrogen (N-(C)3) and CH-N bond, respectively. It is also obvious that three constituent peaks of Co 2p were shown in Figure 3c, two peaks of which located at 781.8 and 796.7 eV, associating with Co 2p3/2 and Co 2p3/2, respectively. The other peak located at 785.4 eV which ascribed to the satellite peak of Co element. These peaks of Co 2p did not match metallic Co, Co oxide or Co (Ⅲ), indicating that the Co element existence state was Co-N bond in Co-porous-g-C3N4. These results illustrate the Co-porous-g-C3N4 photocatalyst possesses great structural stability and the Co element was doped into g-C3N4 successfully. The as-prepared Co-porous-g-C3N4 and g-C3N4 were further investigated by nitrogen adsorption and desorption isotherms (ASAP2460-2) in Figure 4. The isotherm of g-C3N4 shows a type IV [28], which https://doi.org/10.37358/RC. 22.2.8522 indicates the structure is mesopores (2 nm-50 nm) with a clear hysteretic regression line. However, the isotherm of Co-porous-g-C3N4 shows a type V due to the curve has not any inflection poins that is close to the x-axis in the low-medium pressure area and the force of between photocatalyst and nitrogen is weak. Besides, the type H3 hysteresis loop begins to appear in the medium pressure area, and the adsorption isotherm rises faster, which represents the adsorption isotherm feature of flake granular materials. These results indicate the Co-porous-g-C3N4 is mesoporous in a irregular structure and contain various forms of mesopores. Meanwhile, the surface area, pore size and volume of Co-porous-g-C3N4 photocatalyst are showed in Table 2, the surface area of Co-porous-g-C3N4 is 17.86 m 2 /g, which is 3.74 times larger than g-C3N4 [28]. The average pore size of Co-porous-g-C3N4 is the almost same as g-C3N4, but the pore volume of Co-porous-g-C3N4 has about 5.5 times larger than g-C3N4.

Figure 4. N2-adsorption-desorption isotherm
of Co-porous-g-C3N4 Table 2. BET specific surface area and pore size and volume of Co-porous-g-C3N4 UV-Vis DRS spectra of Co-porous-g-C3N4 and g-C3N4 were performed in Figure 5. Unlike the typical light yellow of g-C3N4, the color of Co-porous-g-C3N4 is light green, the reason is that color is influenced by band gap. As shown in Figure 5, the absorption wavelength of 461 nm (g-C3N4) shift to 534 nm (Co-porous-g-C3N4) and the band gap can be calculated with Eg = 1240/λg (nm), the band gap of g-C3N4 (2.69 eV) changed to Co-porous-g-C3N4 (2.32 eV), with a decrease by 0.37 eV. The visible light responsiveness of Co-porous-g-C3N4 is enhanced with the decrease of the band gap, the utilization rate of visible light is increased. So the photocatalytic performance of Co-porous-g-C3N4 is improved.

Photocatalytic degradation activity of Co-porous-g-C3N4 towards MB
For the sake of exploring the photocatalytic ability of Co-porous-g-C3N4, the degradation effect would be texted with MB model under a certain conditions and the consequences were showed in Figure  6. As depicted in Figure 6a, the doping amount of Co element was experimented, which the doping dosage was 0, 0.3, 0.6, 0.9, 1.2, 1.5 wt% under the treatment of 1 mL hydrochloric acid. To explore the catalytic ability of Co-porous-g-C3N4 as a photocatalyst, the photocatalytic degradation of MB by Co-porous-g-C3N4 was investigated under certain conditions. As depicted in Figure 6a, the doping amount of Co element was experimented. The doping dosage was 0, 0.3, 0.6, 0.9, 1.2, 1.5 wt% under the treatment of 1 mL hydrochloric acid. When the doping dosage was 0.9 wt%, the degradation effect of MB was the best, and the degradation rate reached 72% at 120 min. It's worth noting that the photocatalytic effect would be decreased when the increasing of Co doping amount, and only half of MB could be degraded by that photocatalyst contained 1.2-1.5 wt% Co element. This may be due to the fact that the excessive doping of Co element will lead to the combination of catalyst active centers, which in turn affects the degradation performance of the catalyst. However, the photocatalyst without the treatment of 1 mL hydrochloric acid behaved worse in degrading MB under the same condition. When the doping dosage was 0.9 wt%, the degradation rate only reached 43% at 120 min in Figure 6b. But the degradation performance of the catalyst could also be improved due to the doping of Co.
In addition, the dose of Co-porous-g-C3N4 was explored from 0.6 g/L, 0.8 g/L, 1.0 g/L, to 1.2 g/L, and the results were showed in Figure 6c. In this reaction system, due to the different adsorption abilities of the different dose of Co-porous-g-C3N4, the concentration of MB had been controlled in 11-14 mg/L after 40 min adsorption in dark. And since there was some error in MB concentration (11-14 mg/L) at the end of the dark reaction, the degradation effect was compared by pure photocatalytic rate. The best degradation was achieved when the dose of Co-porous-g-C3N4 was 1.0 g/L. The MB concentration decreased from 11.6 mg/L to 2.8 mg/L in 40-160 min, the pure photocatalytic degradation rate was 75.9%. When the doses of Co-porous-g-C3N4 were 0.6 g/L, 0.8 g/L and 1.2 g/L, the pure photocatalytic degradation rates were 49.1%, 52.3% and 31.3% respectively. Interestingly, when the dose was 1.2 g/L, the degradation effect decreased instead. The reason may be that excessive catalysts will accelerate the photogenerated electron-hole recombine rate and the simulated solar light source will be shaded. Therefore, the optimal dose of Co-porous-g-C3N4 was 1.0 g/L.
As showed in Figure 6d, the photocatalytic process was significantly inhibited which corresponding to a MB degradation rate of 64% when the MB solution was weakly acidic (pH=5). It could be attributed to the binding of H + in solution with photoelectrons, which inhibited the photocatalytic degradation process. However, the MB was basically completely degraded after 2 h, when the MB solution was neutral or weakly alkaline (pH=7, pH=9). It was more economical to choose neutral conditions (pH=7) as the optimal solution pH. As showed in Figure 7a, the degradation rates of MB were 60% and 18% at 90 min respectively when only Co-porous-g-C3N4 or H2O2 was applied. In contrast, the photocatalytic degradation performance could be enhanced substantially due to the construction of Photo-Fenton-like reaction system when Co-porous-g-C3N4 coexisted with H2O2. The degradation rate of MB could reach 95% when the dose of H2O2 (30%) was increased to 6.2 mmol/L. However, when the dose of H2O2 was increased to 7.2 mmol/L, the degradation rate of MB did not continue to increase but decreased. The reason may be that excess H2O2 can lead to low degradation rate due to the deactivation of active substances by violent collisions.
As shown in Figure 7b, Co-porous-g-C3N4 showed excellent photocatalytic effect when MB were 30 mg/L and 40 mg/L. The degradation rate could still reach 92% at 90 min when MB was increased to 50 mg/L. It can be seen that the high concentration MB could be degraded by Co-porous-g-C3N4 well. https://doi.org/10.37358/RC. 22.2.8522 Photocatalytic experiments were carried out with various organic pollutants, such as phenol, tetracycline and RhB (20mg/L). As showed in Figure 7c, different types of organic pollutants all could be degraded. The degradation effect of tetracycline was excellent. Thus, a variety of pollutants could be degraded by this system.
After completion of the single experiment, the catalyst was cleaned by physical methods such as simple ultrasonic treatment and rinsing. As showed in Figure 7d, the degradation rates of the three cycles were 99%, 98% and 95% respectively, indicating the wonderful stability of Co-porous-g-C3N4. To explore the types of active species that play a major role in this system, radical quenching experiments were conducted. TBA, BQ and EDTA-2Na were used for quenching ·OH, ·O2-and h + , respectively. As shown in Figure 8a, the main degradation effect was attributed to ·O2-while the secondary degradation effect was attributed to h + and ·OH. The mechanism of the degradation of organic pollutants by Co-porous-g-C3N4 was shown in Figure 8b.
Firstly, the holes with oxidation properties (h + ) could be formed by the photoelectron transitions. In addition, the reaction of O2 with Co 2+ or H2O2 with Co + were both ways to generate ·O2 -(Eq. 3-1, Eq. 3-2 and Eq. 3-3). Finally, the circulation would be completed when High-valent cobalt turned to low-valent cobalt by absorbing photoelectrons (Eq. 3-4 and Eq. 3-5). The recombination rate of photogenerated electron-hole could be slowed down greatly in this system.

Conclusions
Highly active porous photocatalyst Co-porous-g-C3N4 was synthesized by calcination using cobalt chloride as source of Co and concentrated hydrochloric acid-treated melamine as precursor. Photo-Fenton-like reaction system was successfully constructed, and MB was almost completely degraded within 90 min with a degradation rate of 98.3%. Moreover, the catalytic degradation of various organic pollutants was also achieved. Through three recycling experiments, the good recyclability and stability of Co-porous-g-C3N4 were verified. And the main degradation effect was attributed to ·O2-. In this study, the synergistic effect of Fenton-like system and photocatalysis was realized, which had a good application in degrading organic wastewater.