廢生物質衍生碳修飾多孔石墨氮化碳異質結用于海水中氧四環(huán)素的高效光降解

王聞達 麻金庫 魏玉珠 馬帥帥＊,

(1江蘇理工學院化學化工學院，常州 213001)

(2江蘇理工學院資源與環(huán)境工程學院，常州 213001)

0 Introduction

The release of antibiotics at sublethal concentrations from clinical facilities and maricultural operations contributes to an increased likelihood of antibiotic resistance gene (ARG) formation in pathogenic bacteria. This selective pressure encourages the assimilation and dissemination of ARGs within marine ecosystems, posing a significant threat to public health. Oxytetracycline (OTC), among various antibiotics, is commonly employed as a prophylactic agent in mariculture ponds due to its broad-spectrum efficacy and affordability[1]. However, its discharge into water bodies can result in severe ecological damage to offshore environments and potential harm to human well-being[2-3]. Photocatalysis has garnered significant interest as an innovative and environmentally friendly purification technology when compared to conventional membrane filtration[4], adsorption[5], electrochemical oxidation[6], Fenton reaction[7], and ozonation[8]. This is primarily because the degradation of pollutants occurs on a semiconductor photocatalyst,which can be regenerated concurrently under light irradiation, such as solar light.Among the various semiconductor photocatalysts,graphite carbon nitride (g-C3N4)is regarded as a particularly promising candidate, due to its combination of unique properties including suitable band gap, nontoxicity, and photochemical stability[9-11]. Nevertheless,the primitive g-C3N4remains plagued by problems such as rapid charge recombination, high band gap, low surface area, and unsatisfactory visible light harvesting,which limits its large-scale applications. To overcome these shortcomings approaches such as controlled nanomorphology, non-metallic doping, metal loading,carrier coupling, and heterostructure construction have been applied to further optimize the structure and composition of g-C3N4to achieve improved photocatalytic efficiencies[12-15]. Of these, the increase in specific surface area was considered to be the simpler and more practical method of modifying g-C3N4. Liu et al. fabricated mesoporous g-C3N4by a molten salt-assisted silica aerogel template method, which achieved 90.9%efficiency in photocatalytic degradation of RhB[16]. Jin et al. used a simple two-step condensation method to prepare g-C3N4nanotubes with high specific surface area and defects, which enhanced the photocatalytic activity of RhB under visible light irradiation[17].

In addition to these solutions, there has been a recent increase in attention to modulating g-C3N4with carbon materials to create g-C3N4/carbon hybrids or heterojunctions due to their significantly enhanced features on g-C3N4in transporting reagents in bulk,quantity of active sites, optical and electric properties[18-19].The carbon-based materials currently developed for composite semiconductor photocatalysts include carbon fibers, carbon nanotubes, reduced graphene oxide,mesoporous carbon,etc. Liu and Li et al. proposed a graphene oxide (GO) and reduced graphene oxide(rGO)/g-C3N4structure that significantly improved the efficiency of photogenerated electron - hole separation[20-21]. The metal-free two-dimensional (2D) g-C3N4/graphdiyne heterojunction designed by Lu et al. can shorten the transfer distance of photogenerated carriers and increase the hole mobility of g-C3N4due to the highπ-conjugated structure of graphdiyne[22]. Even though some desirable improvements in photocatalytic activities have been achieved, most of them require preparation with difficult synthetic steps,high cost of raw materials, extensive time and energy consumption, and cannot be produced at scale. In contrast, biomass-derived carbons from renewable wastes have been recognized as attractive candidates for the fabrication of functional carbon materials with superior properties due to their greenness, reproducibility, low cost, versatility, and electrical conductivity. For this reason, biomassderived carbons have been applied in wastewater treatment as efficient and appropriate catalyst supports. For instance, Li et al. synthesized Fe3O4/BiOBr/biomassderived carbon heterojunctions stacked on reed strawderived biochar via a one-step hydrolysis method, and the introduction of biochar significantly increased the open pathway to more available active sites and promoted the electronegativity of the separated photogenerated electron-hole pairs[23]. He et al. prepared chitinbased carbon/g-C3N4heterojunctions byin situcalcination of a mixture of chitin and urea and utilized the electron density imbalance of urea and the terminal group of chitin to modulate the microstructure of g-C3N4during calcination, resulting in an approximately 10 -fold increase in the specific surface area of g-C3N4with a stronger ability to inhibit charge recombination[24]. Despite these advances, the exploration of diverse biomass-derived carbons to assist in enhancing the photocatalytic properties of g-C3N4remains extremely attractive.

In this work, biomass-derived carbon (BC)-modified porous g-C3N4(pg-C3N4) heterojunctions were fabricated from banana peel (BP) and urea by a simple one-pot thermal polycondensation process. The photocatalytic properties of pg-C3N4/BC in different ratios were determined by photodegradation of OTC in the artificial seawater under visible light irradiation, which exhibited significantly enhanced photocatalytic activity in comparison with pure pg-C3N4. A preliminary proposal for a possible visible-light-driven photodegradation mechanism in the pg-C3N4/BC system was made considering the directional behavior of charge migration. In particular, the objective of this study is to use biochar derived from massive biomass waste for the remediation of pollutants in industrial wastewater.

1 Experimental

1.1 Chemicals and reagents

Banana was purchased from local supermarket in Changzhou, China. Urea, OTC, isopropyl alcohol (IPA),ethylene diamine tetraacetic acid (EDTA), andpbenzoquinone (BQ) were purchased from Sinopharm Chemical Reagent Co., Ltd. All chemicals were of analytical grade and used without further purification.

1.2 Synthesis of pg-C3N4/BC composites

The specific synthetic procedures are schematically illustrated in Fig.1 and the details are described as follows. Firstly, the BP was chopped into bits and allowed to dry in an oven at 60 ℃for 24 h, and then powdered in a grinder. Subsequently, 13.16 g of urea was solved in 20 mL of distilled water.Then,a dedicated quantity of BP powder was admitted to the aqueous urea solution with vigorous stirring for 30 min followed by sonication for 30 min. After evaporating the water by baking the mixture at 60 ℃overnight, it was cooled at room temperature to allow urea to recrystallize. In the following step, the sample was calcined under N2flow for 2 h under 550 ℃with a 5 ℃·min-1heating rate. Similar procedures were replicated by altering the BP addition amounts to 0.040, 0.116, 0.193, and 0.386 g, respectively. The pg-C3N4/BC composite was labeled as pg-C3N4/BC-X(X=1, 3, 5, 10), whereX% refers to the theoretical load percentage of BC. For comparison,pure pg-C3N4and BC were produced by calcination of urea and BP at 550 ℃for 2 h,respectively.

Fig.1 Schematic diagram of the synthesis of pg-C3N4/BC

1.3 Characterizations

The morphology of the samples was investigated by scanning electron microscopy (Sigma 500) at the acceleration voltage of 5 kV. The crystalline phases were identified using the SmartLab diffractometer(Rigaku) at 40 kV and 30 mA with CuKαX-ray radiation source, a nickel filter (λ=0.154 nm), and a 2θrange of 10°-60°. Transmission electron microscopy(TEM, Hitachi-9000) images was obtained at an accelerating voltage of 200 keV. Elements contents and their chemical states were obtained by X-ray photoelectron spectrometers (XPS, ESCALAB 250xi, Thermo Fisher Scientific, USA) equipped with an AlKαmonochromatic X-ray source (hν=1 486.7 eV) with a line width of 0.20 eV in an analysis chamber. The FTIR spectra were measured on a Nicolet iS50 spectrophotometer from 4 000 to 500 cm-1at room temperature using KBr as a diluting agent. The BET (Brunauer-Emmett-Teller) surface area was calculated based on the adsorption isotherm (ASAP2020HD88). Optical properties were analyzed by a UV-Vis spectrophotometer (UV-2600, Shimadzu). The electron spin resonance(ESR) spectra were acquired by the spectrometer(FA200, JEOL, Japan) with 5,5-dimethyl-1-pyrroline-N- oxide (DMPO). Photoluminescence (PL) spectra were recorded on an F-7000 spectrometer (Hitachi,Japan) with an excitation wavelength of 372 nm. Photoelectric properties of the sample were performed in a three-electrode chemical cell using a CHI 660E electrochemical analyzer with a Pt foil counter electrode and an Ag/AgCl reference electrode, and the solution of Na2SO4(0.5 mol·L-1) was used as the electrolyte.The working electrode was prepared as follows：2 mg of the photocatalyst powder was placed in 2 mL of ethanol containing 20 μL of a 5% Nafion solution to prepare slurry. Then, the powder was uniformly dispersed by grinding. 500 μL of the solution was deposited on an indium tin oxide (ITO) glass substrate (1 cm×1 cm).The obtained electrode was dried at 353 K for 30 min.The light source was a Xe lamp(300 W).

1.4 Static photocatalytic degradation

To evaluate the photocatalytic efficiencies of the fabricated photocatalysts, OTC was degraded in a quartz-jacketed photoreactor at room temperature with water circulation.In detail,20 mg of the as-synthesized photocatalyst was dispersed in 100 mL of artificial seawater (25 g·L-1of NaCl, 11 g·L-1of MgCl2, 4 g·L-1of Na2SO4,and 1.6 g·L-1of CaCl2in deionized water)containing 10 g·L-1of OTC. After being shaken in the dark for 60 min to achieve adsorption-desorption equilibrium, the resulting suspension was illuminated under visible light by a Xenon lamp (300 W) equipped with an ultraviolet filter (λ≥400 nm) at a distance of approximately 10 cm. Samples of the suspension were taken at 10-minute intervals and centrifuged to separate the photocatalyst and the absorbance of the reaction solution was measured with UV-Vis at 353 nm.

1.5 Photocatalysis in a continuous flow system

In Fig.2,a 300 W Xenon lamp in conjunction with a 400 nm cut-off filter was positioned above the reactor at a distance of 15 cm to serve as the light source. A quantity of 0.125 g of catalyst was applied and sintered onto a Ni mesh measuring 3.8 cm×1.8 cm, which was subsequently placed within a custom-made reactor(length： 4 cm, width： 2 cm, height： 1.5 cm). The antibiotics-contaminated artificial seawater and the reactor were connected using a peristaltic pump operating at a flow rate of 2 mL·min-1.The concentrations of antibiotics within the reactor were measured over time using a UV-Vis spectrophotometer.

Fig.2 Scheme illustration of the continuous flow system

2 Results and discussion

The XRD patterns of the as-synthesized pure pg-C3N4and pg-C3N4/BC-Xsamples are illustrated in Fig.3a. It was observed that the pure g-C3N4had two distinct diffraction peaks. The strong peak around 27.4° can be indexed as the (002) plane, which may be attributed to the characteristic interplanar stacking peak of the conjugated aromatic system. The weak diffraction peak around 13.1° is assigned to the (100)plane which corresponds to the in-plane structural packing motifs of tri-s-triazine[25]. In comparison with pristine pg-C3N4, the characteristic peaks of the pg-C3N4/BC-Xcomposites shifted from 27.4° to 26.8°as the BC content increased,demonstrating an increase in the distance between the pg-C3N4layers[26].Such phenomena may be attributed to the confined carbon construction in the pg-C3N4layer via the strongπ-πstacking interactions among the carbon and the pg-C3N4matrix coming from the aromatic character[27]. The FTIR analysis (Fig.3b) provided further confirmation of the structural characteristics of both pure pg-C3N4and pg-C3N4/BC-X. The pristine pg-C3N4exhibited a peak at 1 638 cm-1,which can be attributed to C—N stretching, and four peaks at 1 568, 1 421, 1 325, and 1 247 cm-1, indicative of aromatic C—N stretching vibrations[28-29]. Additionally, the peak at 809 cm-1is associated with the triazine ring modes, while the broad peaks in the range of 3 000-3 500 cm-1were attributed to terminal amino groups and surface-adsorbed hydroxyl species[30-31]. As anticipated, virtually all characteristic pg-C3N4-related peaks were reflected in pg-C3N4/BC-X, providing straightforward proof of heterotopic binding of pg-C3N4and BC.

Fig.3 XRD patterns(a)and FTIR spectra(b)of pg-C3N4 and pg-C3N4/BC-X

The morphology and microstructures of the pg-C3N4, BC, and pg-C3N4/BC-5 samples were further investigated through the utilization of SEM and TEM techniques. The SEM and TEM images of the pg-C3N4synthesized using urea as a precursor (Fig. 4a, 4d)revealed a 2D structure composed of small, flat sheets with wrinkles, as well as a three-dimensional porous structure with minimal aggregation. Fig. 4b and 4e exhibit the SEM and TEM images of BC, revealing its composition as a disordered arrangement of carbon sheets. These amorphous sheets exhibited localized regions of crystallinity, characterized by stacked planes of carbon structures at the nanometer scale. In contrast, Fig.4c and 4f present the SEM and TEM images of the pg-C3N4/BC-5 sample, illustrating the successful integration of carbon sheets with pg-C3N4nanosheets,resulting in the formation of a 2D-2D structure. Furthermore, it is noteworthy that the pg-C3N4/BC-5 sample exhibited a wrinkled and porous lamellar structure.This porous morphology significantly augments the surface area of the pg-C3N4/BC-5 sample,thereby facilitating enhanced diffusion and transport of reactant substrates. Consequently, this structural characteristic potentially contributes to the improved efficiency of photocatalysis.

XPS was applied to investigate the surface chemical state of the as-prepared sample. The XPS survey spectra (Fig.5a) show the majority of elements of carbon,nitrogen,and oxygen in pg-C3N4and pg-C3N4/BC-5.The high-resolution XPS spectra of C1sand N1sfor g-C3N4and pg-C3N4/BC-5 samples are shown in Fig.5b and 5c. The C1sspectrum for pg-C3N4was deconvoluted into three peaks centered around 284.80, 286.38,and 288.18 eV, corresponding to C—C bands, C—O species, and N=C—N coordination bond[32-34], respectively.The peak ofsp2hybridized carbon in the pg-C3N4/BC-5 sample shift to higher binding energy, which should be attributed to spontaneous self-assembly electron redistribution between pg-C3N4and BC, demonstrating the formation of strong interaction between pg-C3N4and BC. The N1sspectrum of pg-C3N4can be well fitted with three deconvolution peaks centered at 398.29,400.06,and 403.66 eV,which are attributed to the triazine rings ofsp2-hybridized aromatic nitrogen(C=N—C), tertiary nitrogen coupled to three carbon(N—(C)3), and amino function groups (C—N—H)[35-37].Interestingly, the peak shifting phenomenon also occurred in the XPS spectra of N1swhich is similar to the high-resolution XPS spectra of C1sand has the same shifting direction for pg-C3N4/BC-5 compared with pg-C3N4.

Fig.5 High-resolution XPS spectra of(a)survey,(b)C1s,and(c)N1s spectra of pg-C3N4 and pg-C3N4/BC-5

The N2adsorption-desorption isotherms were used to determine the BET specific surface areas (SBET) of the pg-C3N4and pg-C3N4/BC-Xsamples. Fig.6 illustrates that all the samples exhibited a type-Ⅳisotherm model with a hysteresis loop within thep/p0range of 0.8-1.0, indicating the presence of mesopores in both the pg-C3N4and pg-C3N4/BC-Xsamples[38-39].By analyzing the desorption curve, the BET surface area of the pg-C3N4/BC-5 sample was calculated to be 117 m2·g-1,surpassing that of the pg-C3N4sample. This phenomenon suggests that the introduction of BC probably prevents pg-C3N4from restacking[40]. This phenomenon implies that the incorporation of BC likely inhibits the restacking of pg-C3N4. It is widely acknowledged that a high concentration of surface-active sites is advantageous for the adsorption and transportation of pollutant molecules within the interconnected porous structure,thereby promoting enhanced photocatalytic performance. Furthermore, the specific surface area exhibited an initial increase with the rising of BC content,followed by a subsequent decrease as the BC content continued to rise, owing to the potential aggregation induced by excessive BC, consequently leading to a reduction in the specific surface area (as depicted in the inset of Fig.6).

Fig.6 N2 adsorption-desorption isotherms and BET surface areas(Inset)of pg-C3N4 and pg-C3N4/BC-X

The optical absorption properties of materials were analyzed using UV-Vis diffuse reflectance absorption spectroscopy (DRS). Fig.7a illustrates that pristine pg-C3N4demonstrates a typical light-responsive capability within the visible light range, with an absorption edge extending up to 460 nm. In contrast, the pg-C3N4/BC-Xsamples exhibited enhanced light absorption across the entire UV-visible light spectrum, accompanied by an increase in adsorption intensity, when compared to pristine pg-C3N4.The narrow gap of thesp2carbon cluster embedded in BC is widely recognized for its exceptional light absorption across a broad range of wavelengths[41]. Consequently, the incorporation of BC material in pg-C3N4/BC-Xcomposites can significantly enhance their light harvesting efficiency. The findings indicate that these composites can capture a greater amount of light, leading to the generation of more electron-hole pairs and ultimately augmenting their photocatalytic activity under visible light irradiation. Moreover, the conversion of DRS spectra into Tauc plots were employed to examine the energy-band structure of pg-C3N4and pg-C3N4/BC-5 (Fig.7b). The calculated band gaps for pg-C3N4and pg-C3N4/BC-5 were determined to be 2.59 and 1.27 eV,respectively.Additionally, the observed reduction in band gap width suggests that electron transitions can occur with lower energy,thereby enhancing the photocatalytic capacity[42]. It is widely recognized that the investigation of energy band positions is crucial for comprehending the redox mechanism of the photocatalyst. Fig.7c shows the XPSvalence band (VB) spectra of pg-C3N4and pg-C3N4/BC-5. The VB positions of pg-C3N4and pg-C3N4/BC-5 were determined to be located at approximately 1.65 and 0.39 eV by linear extrapolation, respectively. So theECB(conduction band energy) of pg - C3N4and pg-C3N4/BC-5 were equal to -0.94 and -0.88 eV,respectively. The above measurements and calculation helped to estimate the band structures in Fig.7d. Compared with pg-C3N4,pg-C3N4/BC-5 had a lower conduction band position and VB position, which may be caused by the introduction of BC with rich electronabsorbing groups[43]. The more negative the conduction band, the stronger its reducing ability, and the more conducive to the generation of free radicals[44], which should be evidence that pg-C3N4/BC-5 had a stronger catalytic ability.

Fig.7 (a)UV-Vis DRS spectra of BC,pg-C3N4,and pg-C3N4/BC-X;(b)Tauc plots converted from DRS spectra,(c)XPS-VB spectra,and(d)energy band structure of pg-C3N4 and pg-C3N4/BC-5

PL experiments are performed to examine the efficacy of charge carrier separation in the photocatalysts that have been prepared. Fig.8 illustrates the PL spectra of the original pg-C3N4and pg-C3N4/BC-Xsamples,with an excitation wavelength of 325 nm. It is evident that the emission spectrum of pure pg-C3N4displayed heightened intensity. In contrast, it was observed that the intensities in the pg-C3N4/BC-Xcomposites noticeably decreased with an increase in the BC content.This decrease in PL intensity indicates a more effective separation of photo-excited charge carriers at the interface between BC and pg-C3N4, implying a lower rate of recombination. As a result, the incorporation of BC in the composites can significantly prolong the lifetime of photogenerated electrons and holes during the transfer process, thereby enhancing the quantum efficiency and improving the photocatalytic efficiency.

Fig.8 PL emission spectra of the prepared samples

The transfer kinetics of photogenerated carriers were further analyzed by transient photocurrent response and electrochemical impedance spectroscopy.As shown in Fig.9a, pg-C3N4/BC-5 exhibited the highest photocurrent density far surpassing those of pg-C3N4and pristine BC, indicating that BC introduction effectively promoted the electron-hole pairs separation. To provide further support for the enhanced interfacial charge separation in the binary heterojunction system, electrochemical impedance spectroscopy (EIS)was performed (Fig.9b).It was observed that the arc radius of the Nyquist circle for pg-C3N4/BC-5 was smaller than that of pg-C3N4, indicating a reduced resistance for carrier transfer. This suggests that the efficient migration of photogenerated electrons and holes within the components effectively inhibits their recombination, allowing a greater number of free charges to participate in the photocatalytic degradation process.

Fig.9 (a)Photoelectrochemical response and(b)EIS plots of BC,pg-C3N4,and pg-C3N4/BC-5

To examine the photocatalytic efficacy of the samples,OTC was selected as the indicator for organic contaminants in artificial seawater. As evidenced in Fig.10a,the adsorption equilibrium was attained within a 60-minute duration for all catalysts under dark conditions. After reaching the adsorption equilibrium,Fig.10a illustrates the alteration in OTC concentration(ρ/ρ0) with varying illumination durations for all photocatalysts. An experiment was conducted in which no photocatalyst was utilized, and the findings indicate that there was no significant reduction in OTC under visible light exposure. This implied that OTC remained stable and the occurrence of self-photolysis can be dismissed. The photocatalytic degradation of OTC using BC was slightly increased from 24.1% (under dark) to 26.3% (under light irradiation). It suggests that BC only acts as an adsorbent and does not have any photocatalytic properties due to the stability of BC under light. In contrast, the utilization of pure pg-C3N4as the catalyst led to a gradual reduction in the concentration of OTC, resulting in a degradation efficiency of 45.3%after 70 min of irradiation. Nevertheless, the incorporation of BC significantly improved the photocatalytic degradation performance of the pg-C3N4/BC-Xsamples.The pg-C3N4/BC-5 sample achieved a degradation efficiency of 96.6% after undergoing 70 min of irradiation in the same reaction. However, the introduction of higher BC content in pg-C3N4/BC-Xcomposites(pg-C3N4/BC-10) resulted in a significant decline in photocatalytic activity. This observation highlights the significance of maintaining an appropriate BC content to attain optimal photocatalytic performance. This phenomenon can be elucidated in the following manner：the excessive presence of residual BC leads to the formation of agglomerates that cover the surface of pg-C3N4, as evidenced by the BET results presented in Fig.6. Consequently, this coverage partially obstructs the absorption of light, thereby causing a reduction in the efficiency of OTC photocatalytic degradation. In addition, it is well known that carbon can adsorb OTC molecules onto surfaces of the composites, and the adsorption capacity increased with the increase of carbon contents. Thus, when the carbon content was above that in pg-C3N4/BC-5,the excessive adsorbed OTC molecules on the surfaces of pg-C3N4/BC-10 partly decelerate the interfacial charge transfer, resulting in a decrease of OTC photodegradation efficiency[45]. The results of the kinetic study indicate that the photocatalytic degradation of OTC using these catalysts follows pseudo-first-order kinetics, as shown in Fig.10b and 10c.The experimental results reveal that the photocatalytic degradation efficiency was significantly improved by incorporating BC into pg-C3N4, as evidenced by the notably higher reaction rate constant (k) of 0.047 min-1for pg-C3N4/BC-5 compared to the pristine pg-C3N4withkof 0.005 min-1. This enhancement was estimated to be approximately 8.4 times greater, indicating the effective role of BC in enhancing photocatalytic performance. Furthermore, to facilitate the practical implementation of photocatalysts for water purification, they must exhibit not only high efficiency but also exceptional stability to enable multiple uses. Consequently,the stability and reusability of the pg-C3N4/BC-5 photocatalyst were assessed through five successive reaction cycles.As depicted in Fig.10d,the photocatalytic activity of the pg-C3N4/BC-Xsamples remained largely unaffected even after undergoing five cycles. Although a slight decline in activity was observed,it is plausible to attribute this to the inevitable loss of catalyst during the recycling process. This finding provides additional evidence of the exceptional reusability and stability exhibited by the pg-C3N4/BC-Xphotocatalysts.

Fig.10 (a)Photocatalytic degradation efficiency of OTC under visible-light irradiation over pg-C3N4,BC,and pg-C3N4/BC-X,and corresponding(b)degradation kinetic curves and(c)k values,and(d)cyclic stability test for degradation of OTC over pg-C3N4/BC-X

To assess the potential of pg-C3N4/BC-5 as a viable technology, a continuous flow reaction system(Fig.2) was employed to investigate the likelihood of photocatalytic degradation of OTC in an artificial seawater solution. The results depicted in Fig.11 demonstrated a rapid decline in the concentration of OTC within the cell, occurring within a span of 40 min. This can be attributed to the ongoing transfer of OTC molecules from the solution to the surface of pg-C3N4and pg-C3N4/BC-5.Despite demonstrating a 45.3%degradation efficiency of OTC in seawater (Fig.10a), the efficacy of pg-C3N4in the dynamic system was deemed unsatisfactory at 9.7%. Photocatalysts exhibiting superior adsorption capabilities performed more effectively in the dynamic system. Notably, the pg-C3N4/BC-5 photocatalyst achieved a removal rate of 39.1% over an extended duration, indicating the synergistic effect of adsorption and photocatalysis in facilitating OTC removal in a dynamic system.

Fig.11 Photocatalysis performance of OTC in the artificial seawater in continuous flow test

Experiments were conducted to evaluate the functions of the active species in the photocatalysis process, employing radical capturing techniques and utilizing the pg-C3N4/BC-5 composites. IPA, EDTA, and BQ were employed as scavengers for ·OH, hole (h+)and·O2-,respectively.According to Fig.12a,the photocatalytic efficiency was moderately influenced, resulting in a degradation rate reduction of 82.3% when IPA(1 mmol·L-1) was introduced into the OTC solution.This suggests that·OH had minimal impact on the degradation reaction. However, the addition of EDTA (1 mmol·L-1) and BQ (1 mmol·L-1) resulted in a notable decrease in the degradation rate of OTC, with reductions of 26.4% and 8.2% respectively. This suggests that the presence of ·O2-and h+is crucial in facilitating the photodegradation of OTC over pg-C3N4/BC-5.The ESR spectra were conducted to further analyze the roles of ·OH and ·O2-. As shown in Fig.12b and 12c,no signal peaks of DMPO - ·OH and DMPO - ·O2-adducts appeared in dark conditions, indicating that the light irradiation was necessary to generate ·OH and·O2-.The signal peaks of DMPO-·OH and DMPO-·O2-were detected under visible-light irradiation, which demonstrated that ·OH and ·O2-could be produced by pg-C3N4and pg-C3N4/BC-5. Besides, a typical weak quadruple peak spectral characteristic signal of DMPO-·OH with an intensity ratio of 1∶2∶2∶1 was detected in the pg-C3N4/BC-5 photocatalytic system[44], demonstrating that low activity ·OH occurs in the degradation of OTC. At the same time, a typical strong six-peak characteristic signal of DMPO-·O2-can be observed[46]. The above conclusions indicate that ·O2-is the main dominant active free radical in the pg-C3N4/BC-5 photocatalytic system. Based on the aforementioned findings and subsequent analysis, a plausible mechanism for the degradation of organic pollutants through a visible-lightdriven photocatalytic reaction in the presence of pg-C3N4/BC-5 sample is proposed in Fig.12d. The incorporation of BC into the pg-C3N4/BC-5 sample has resulted in an augmented specific surface area, thereby facilitating the availability of more reactive sites.Furthermore, owing to the exceptional optical characteristics of the pg-C3N4/BC-5 composites, a substantial number of charge carriers are generated upon exposure to visible light irradiation. The BC exhibits conductive properties and functions as a transient electron acceptor. Consequently, the photogenerated electrons originating from the CB (conduction band) of pg-C3N4can be transferred to BC via the contact surface between pg-C3N4and BC. This facilitates the efficient separation of photogenerated electrons and holes, thereby diminishing the recombination rate of photogenerated electrons (e-) and h+in pg-C3N4/BC-5. Simultaneously,the transferred electrons engage in a reaction with free O2in the solution, resulting in the production of superoxide radicals (·). The valence band (VB) energy of pg-C3N4/BC-5 (0.39 eV) exhibited a more negative value compared to the standard redox potential of OH-/·OH (1.99 eV), resulting in the ineffectiveness of h+in converting OH-to ·OH[47]. However, ·can react further with H2O and electrons, leading to the formation of ·OH[48]. Consequently, the generated ·,·OH,and h+can all serve as active species in engaging in redox reactions with phenanthrene adsorbed on the catalyst surface, thereby facilitating the mineralization and decomposition of pollutants. The degradation mechanism of OTC can be explained by the following equations：

Fig.12 Trapping experiments of the active species during the photocatalytic degradation of OTC for pg-C3N4/BC-5(a);ESR spectra of(b)DMPO-·OH and(c)DMPO-·adducts;Mechanism for the enhanced photocatalytic activity(d)

3 Conclusions

In conclusion, the pg-C3N4/BC composites were effectively synthesized via a straightforward one-step calcination method utilizing a precursor mixture of BP and urea. The present study comprehensively evaluated the photocatalytic efficiency of the as-synthesized photocatalysts under visible-light conditions for the degradation of OTC, a representative organic pollutant in seawater. The results indicate that the pg-C3N4/BC composites exhibited enhanced photocatalytic performance in the removal of OTC compared to the pristine pg-C3N4. Furthermore, the pg-C3N4/BC composites with an optimal BC content demonstrate superior efficiency compared to all other prepared photocatalysts. The study determined that the reaction rate constant of pg-C3N4/BC exhibited a notable increase of approximately 8.4 times compared to that of pg-C3N4during the degradation of OTC. This significant augmentation in photocatalytic activity can be ascribed to the amplified specific surface area, enhanced light-harvesting capability, and efficient separation of photo-generated charge carriers. Furthermore, the pg-C3N4/BC-5 composites exhibited favorable stability, as evidenced by their negligible decline in activity following five consecutive reactions. This study has successfully showcased the potential of the synthesized pg-C3N4/BC composites as efficient photocatalysts under visible light irradiation for effective environmental remediation.

Acknowledgments:This work was financially supported by the Natural Science Foundation of Jiangsu Province (Grant No.BK20181048) and the Postgraduate Research & Practice Innovation Program of Jiangsu Province (Grant No. SJCX22_1457),and the authors express their gratitude to the Jiangsu University of Technology (JSUT), located in Changzhou, China, for its substantial resources and support.