Visible-light-driven photocatalytic degradation of doxycycline using TiO2/g-C3N4/biochar catalyst

TiO2/g-C3N4/biochar (TCNBC) catalysts were prepared by the hydrolysis method for the photocatalytic degradation of doxycycline antibiotic (DC), with biochar obtained from the pyrolysis of Phragmites australis. The catalysts were examined using scanning electron microscope (SEM), transmission microscopy (TEM), energy dispersive x-ray spectrometer (EDX), x-ray photoelectron spectroscopy (XPS), x-ray diffraction (XRD), Fourier-transform infrared spectroscopy (FTIR), photoluminescence spectroscopy (PL), and ultraviolet-visible diffuse reflectance spectroscopy (UV–Vis DRS) and nitrogen adsorption/desorption. The photocatalytic activity results showed that the TCNBC catalyst exhibited higher catalytic activity than pure TiO2 or g-C3N4. Its peak catalytic activity, achieving a decomposition efficiency of 91.93% and a mineralization efficiency of 81.50%, can be attributed to the synergistic effect of biochar, TiO2, and g-C3N4. Even after four cycles of use, the catalyst still maintained relatively high activity for the degradation of DC. The photocatalytic degradation efficiency of TCNBC decreased from 91.93% to 86.30% after four recycling events.


Introduction
Antibiotics are chemical compounds widely used in the treatment of microorganisms and infectious diseases, encompassing applications in both human and animal healthcare, as well as in aquaculture and livestock production [1,2].During the application of antibiotics, a large amount of wastewater containing antibiotics is generated and discharged into the environment, causing severe pollution.Antibiotics, characterized as organic compounds, exhibit stability and resistance to degradation by traditional biological and chemical treatment methods [1,2].Therefore, they are frequently detected in various natural environments, including river water [3], groundwater [4], surface water [5], soil [6], and water drink [7].The long-term presence of antibiotics in the natural environment can lead to the generation of antibiotic resistance genes and antibiotic-resistant bacteria, thereby contributing to the dissemination of antibiotic resistance, which poses a significant threat to human health and ecosystems [2].
Photocatalytic degradation of organic compounds is currently recognized as one of the most effective methods for wastewater treatment, especially for wastewater containing toxic and persistent organic pollutants [8,9].The photocatalytic process completely decomposes organic pollutants by forming active hydroxyl radicals and superoxide without generating sludge or waste [8].Among semiconductors photocatalysts, titanium dioxide (TiO 2 ) is undoubtedly the most viable photocatalyst for practical applications thanks to its favourable physicochemical properties, e.g.suitable band position, non-toxicity, low cost, chemical inertness, photochemistry and biocompatibility [8,10].However, due to their wide bandgap, pristine TiO 2 nanoparticles can only be excited by UV light [10].Therefore, various TiO 2 modification techniques have been applied to shift its optical absorption edge towards the visible light region.These techniques include doping with cations [11] or anions [12], ion implantation [13], narrow band gap semiconductors [14], sensitizers [15], and noble metal clusters with surface plasma resonance [10].Among the developed visible-light-driven photocatalysts, interest has grown in composite systems containing at least two semiconductors as they are expected to exhibit synergistic effects, e.g. the ability to absorb the light spectrum is broader as well as the separation efficiency between the photo-generated electron-hole pairs is high [16].
Recently, graphitic carbon nitride (g-C 3 N 4 ), a non-metallic polymeric photocatalyst introduced by Wang et al has emerged as a promising material for solar energy utilization due to its desirable band gap (2.7 eV), semiconductor properties, and high thermal and chemical stability [17].However, the quantum efficiency of g-C 3 N 4 is still low due to the high rate of charge recombination [18].Therefore, designing and preparing composites for large surface areas and high charge transfer will be helpful.Recent studies have demonstrated the improved photocatalytic properties of g-C 3 N 4 /TiO 2 composites compared to their individual component [19,20].To further enhance the photocatalytic activity of g-C 3 N 4 /TiO 2 and broaden its application scope, scientists have recently tended to study its immobilization on third matrix phases such as kaolinite [21], carbon [22], graphene [23], bentonite [19], diatomite [24], silica [25].
Biochar (BC), produced through the pyrolysis of biomass from carbon-rich sources (e.g., rice husks, peanuts, straws, reeds) under oxygen-limited conditions [26,27], has gained prominence as a promising material for environmental remediation.Its exceptional physicochemical properties include a large specific surface area, extensive porous structure, and many active functional groups on the surface.In addition to the superior adsorption capacity, BC can act as an electron acceptor and participate in electron transport.Moreover, its high stability makes BC an excellent support material for photocatalysts [19].Phragmites australis is a common reed predominantly found around lakes, rivers, and streams.It grows all year round throughout the countryside of Vietnam, especially in wetlands.This plant possesses a porous tissue structure primarily composed of cellulose, hemicellulose, and lignin, rendering it a promising material for developing adsorbents [28].Utilizing Phragmites australis as a raw source for fabricating BC hold both scientific and practical significance.Several BC-based materials have been synthesized recently, such as ZnO/BC [29], TiO 2 /BC [30], Co 3 O 4 /BC [31], CuO/BC [32] and BiOBr/BC [33].Recently, Wang et al successfully synthesized TiO 2 @g-C 3 N 4 @BC composite material derived from bamboo biomass with a photodegradation efficiency of the antibiotic ciprofloxacin reaching 89.2% under visible light radiation [34].Several studies show that the physicochemical properties of biochar are highly dependent on the type of raw material and operating parameters (heating rate, temperature, atmospheric conditions, residence time, etc) for biochar formation [35].Therefore, research on synthesising biochar-based composites created from specific raw biomass sources is necessary.However, to the best of our knowledge, research on the tertiary composition system of TiO 2 /g-C 3 N 4 /BC produced from raw reed biomass and its application in the field of photocatalysis remains limited.
In this study, the TiO 2 /g-C 3 N 4 loaded BC photocatalysts were synthesized via a simple hydrolysis method.The photocatalytic activity of the photocatalysts was evaluated based on the photodegradation of DC antibiotic under visible light (λ > 420 mm) irradiation.

Materials
Phragmites australis samples were collected from wet soil in Dong Thap province, Vietnam.The stems were dried under sunlight for four days before being finely ground to approximately 1-2 mm in size.The obtained biomass was rinsed with distilled water and dried in a vacuum oven at 70 °C until a constant weight was achieved.The resulting product was stored in a desiccator and used as raw P. australis biomass (cellulose 43.31%, hemicellulose 30.82% and lignin 20.37%) [36].

Methods
Heat biomass of reed stems (3.0 grams) under a nitrogen (N 2 ) atmosphere at 400 °C for 2 h (heating rate of 3 °C min −1 ).The resulting product was first washed with 1.0 M HCl and then washed several times with distilled water until the filtrate reached neutrality.The washed product was subsequently dried at 105 °C for 24 h and then stored in a desiccator for use as biochar (denoted as BC).
The urea (5.0 grams) was placed in a porcelain boat, sealed by aluminium foil and heated in an N 2 atmosphere at 520 °C for 2 h (heating rate 3 °C min −1 ).After cooling down naturally, the obtained yellow powder obtained was placed in a beaker containing 100 ml of distilled water and sonicated the mixture using a sonicator (4200 S3, SOLTEC) at room temperature for 30 min, then dried at 105 °C for 12 h to obtain the g-C 3 N 4 (denoted as CN).
Dissolve 4.0 grams of titanium(IV) oxysulfate in a 250 ml beaker containing 80 ml of distilled water to form a TiO 2 suspension.Add 0.5 grams of BC and 1.5 grams of CN to this suspension and vigorously stir the mixture at room temperature for 12 h (50.0%TiO 2 ; 37.5% g-C 3 N 4 and 12.5% BC).Next, slowly add 50 ml of 1.0 M NaOH to this mixture, hydrolyze it at laboratory temperature for 4 h, and let it stand for 1 h.The resulting mixture was collected and washed several times with distilled water until the filtrate became neutral.After washing, the product was dried at 105 °C for 12 h, then finely ground and calcined in N 2 atmosphere at 500 °C for 2 h (heating rate of 3 °C min −1 ) to obtain TiO 2 /g-C 3 N 4 /biochar composite (referred to as TCNBC).TiO 2 /g-C 3 N 4 (TCN) and pristine TiO 2 (T) samples were synthesized under the same conditions as above, in the absence of BC and simultaneously in the absence of both BC and CN.

Photocatalytic degradation of DC
In this study, a beaker containing 200 ml of an aqueous suspension of 25 mg.L −1 DC and 0.1 g of catalyst.The light source is a 50 W − 220 V Compact lamp (Dien Quang, Vietnam) equipped with a wavelength cut-off filter (λ 420 nm, d = 77 mm).Before illumination, the suspension was stirred magnetically in the dark for 60 min to ensure adsorption/desorption equilibrium.Four millilitres of the suspension was withdrawn at a certain time interval and centrifuged to remove the solid.The concentration of DC in the supernatant was determined by the UV-vis spectrophotoscopy at λ max = 346 nm.The adsorption efficiency of DC (A%) was calculated from: ) and C 0e (mg.l −1 ) are the concentration of DC at initial and equilibrium times, respectively.The photodegradation efficiency (D%) of the photocatalyst was calculated according to expression (2) [37]: e t e 0 0 = ẃhere C 0e (mg.l −1 ) and C t (mg.l −1 ) are the DC concentration at sorption equilibrium time and at irradiated time of t (min), respectively.According to the Langmuir-Hinshelwood kinetics model, the photocatalytic process of DC dye can be expressed as an apparent pseudo-first-order kinetic equation [38]: where k is the apparent pseudo-first-order rate constant.220) and (215) planes of anatase TiO 2 (JCPDS, file No. 00-021-1272), respectively [39,40].Characteristic diffraction peaks at 13.1°and 27.4°correspond to the (100) and (002) planes of CN [41].For the two composites TCN and TCNBC, in addition to the appearance of the anatase phase, there was also a very low-intensity CN peak, demonstrating the presence of T and CN in the composite samples.This could be due to the low crystallinity of CN in the catalyst [41] and the uniform coating of CN on the TiO 2 nanoparticles, preventing the order stacking pattern of CN in the long-range [41].According to the XRD results, the average crystallite size calculated by the Scherrer equation at (101) is shown in table 1.The size of the nanoparticles decreases in the order of T (13.8 nm) < TCN (10.5 nm) < TCNBC (5.5 nm), demonstrating that CN and biochar inhibited the crystal growth of TiO 2 particles from aggregation during calcination due to their layered structure [41].

Results and discussion
Characterization of the chemical functional groups of the samples for T, CN, BC, TCN and TCNBC was investigated by FTIR spectroscopy.Figure 1(b) shows that all samples have an absorption peak around 3400 cm −1 , which is thought to be related to the stretching vibrations of −OH in water molecules [42].A peak at 480 cm −1 is observed in the spectra of T, TCN and TCNBC, corresponding to Ti-O-Ti characteristics, indicating TiO 2 formation in the samples [42].In addition to the absorption peaks related to the −OH and C−O vibrations mentioned above, biochar exhibits absorption peaks at 2924 cm −1 and 1619 cm −1 , attributed to the asymmetry vibration of C−H and sp 2 character of C=C, respectively [42].The peak at 1462 cm −1 can be attributed to the stretching vibration of COO− [42].The region between 900 and 1300 cm −1 and 1385 cm −1 is considered involved in C−O vibrations and the vibrational mode of O−H, C−C and C−H bending [43,44].Sample CN exhibits a broad absorption region between 3200 cm −1 -3600 cm −1 , which is attributed to the stretching   [43,44].Moreover, sample CN displays a series of bands typical of the stretching vibrations of the C−N heterojunctions (1636 cm −1 , 1459 cm −1 , and 1410 cm −1 ) and high-intensity strain vibrations of tri-units s-triazine at 813 cm −1 [43,44].Peaks at 1317 cm −1 and 1240 cm −1 are associated with aromatic C-N stretching [43,44].Notably, the absence of absorption peaks at 3000 cm −1 and 2200 cm −1 rules out C≡N triple bond formation [43,44].The TCN and TCNBC composite samples also show characteristic peaks relative to the component materials, especially another peak observed around 1640 cm −1 , representing the C=N of CN, thus confirming that CN was successfully added to TCN and TCNBC [43,44].
The textural properties of the T, CN, BC, TCN and TCNBC samples were studied by the nitrogen adsorption/desorption isotherms at 77 K (figure 1(c) and table 1).All the transmission isotherms in figure 1(c) exhibit a type IV isotherm transmission, and since a loop exists at the relative pressures between (P/P 0 ) from 0.4 to 0.6, it indicates the presence of mesopores [41].The TCN and TCNBC composites exhibit hysteresis loops at both medium and high P/P 0 , implying the existence of both small and large-diameter mesopores.Specific surface area (S BET ) values of T, CN, BC, TCN and TCNBC samples were 2.10 m 2 /g, 36.80 m 2 /g, 22.80 m 2 /g, 16.06 m 2 /g and 32.69 m 2 /g.The TCNBC composite has a higher specific surface area than pure T and TCN, indicating more porous structures.It can be seen that the CN and BC matrix phases have significantly improved the TiO 2 surface area in the TCNBC composite.A specific surface area can provide more potential active sites.Therefore, the large specific BET surface area of TCNBC composite compared to pure TiO 2 and TCN may be one of the reasons for its higher catalytic activity [45].Furthermore, forming a heterojunction structure g-C 3 N 4 /TiO 2 in the TCNBC composite would enhance the separation efficiency between the photo-generated electron-hole pairs, thereby further improving the efficiency of optical decomposition [46].
Morphology and microstructure of the obtained materials were observed by SEM and TEM.The T sample appears in the form of agglomerated particles with an average diameter of about 14.1 nm, which was further confirmed by TEM images (SEM-figure 2(a), TEM-figures 2(b) and (c)).Sample of CN displayed a typical layered structure indicating the formation of a graphite phase after calcination (figure 2(d)) [47].Figure 2(e) shows SEM images of BC as flat filaments with smooth surfaces.The morphology of TCN consists of fine particles with around 7-9 nm in diameter, which dispersed highly in the CN matrix (SEM-figure 2(f), TEMfigures 2(g) and (h)).The SEM image of the TCNBC composite sample shows spherical TiO 2 particles with small particle sizes (∼5 nm) and they seem to be successfully embedded onto the CN and BC sheets (SEM-figure 2(i), TEM-figures 2(j) and (k)).The high-resolution TEM (HRTEM) image (figure 2(l)) shows lattice fringes with a distance of d(101) (0.35 nm) of anatase TiO 2 [48].
The EDX elemental mapping of a selected area on the TCNBC composite sample is shown in figure 3.All four elements C, N, O and Ti, are uniformly distributed throughout the TCNBC composite network (figures 3(d)-(g)).The existence of C, N, O and Ti on the surface of the TCNBC sample, along with the merged image of C, N, O and Ti (figure 3(c)) confirms the spatial distribution of elements in the structure.EDX analysis further reveals that the atomic composition of the surface is as follows: C 17.42%, N 15.48%, O 48.45% and Ti 18.65% (the inset of figure 3(b)).These results demonstrate that TiO 2 nanoparticles are successfully dispersed on g-C 3 N 4 and biochar, and that the synthesized TCNBC is free of impurities.
Figure 4 presents the core-level XPSs, which further confirms the presence of Ti, O, C, and N elements.The core-level XPS of C1s is divided into three peaks: the peak at 277.6 eV contributed to the adventitious carbon.The 286.0 eV, 287.8, and 289.4 eV peaks could be attributed to C-N bands, (N) 2 -C=N bands, and O=C=O bands, respectively [49,50].The XPS of O 1s in figure 4(b) presents two peaks; one peak of high intensity contributing to the Ti-O bond at ac 531.6 eV, and the one at 522.7 eV with lesser intensity, which is due to the O-H bond of water molecules [49].The spectra of N1s in figure 2(c) are also deconvoluted into three peaks: a small peak at 391.5 eV belongs to N in C=N-C bonding, while the other at 400.0 eV corresponds to N-(C) 3 and the highest peak at 401.3 eV which attributes to N-H structures [37].The sharp peaks at 459.9 eV, 465.6 eV, and 473.3 eV correspond to Ti2p (figure 2(d)), indicating the presence of the Ti 4+ state [40,49,51].
Optical properties of samples T, CN, BC, TCN and TCNBC were evaluated using UV-Vis DRS spectra (figure 5).The BC sample displayed a slight absorption band covering the entire region of the ultraviolet-visible light (figure 5(a)) [52].Figure 5(a) shows that the pure T sample has an upward-moving optical absorption edge at a wavelength of about 410 nm.The CN sample has an upward-moving optical absorption edge in the wavelength region of about 470 nm.After T was coated on the surface of CN, the optical absorption edge of the TCN composite had a red shift from 410 nm to 873 nm due to the strong optical absorption of CN, showing that CN dominates and affects the optical absorbance of TCN.If TiO 2 is coated on the surface of CN and BC simultaneously, there is a redshift from 410 nm to 1000 nm due to the strong optical absorption of both CN and BC, proving that TCNBC composite absorptive light in the visible light range.This shift represents the smaller crystal size of the TCNBC composite compared with other samples prepared in this work.
To calculate the band gap values, the Tauc plot relation is used [34,53]: where α is the absorption coefficient, h is the Planck constant, v is the wavenumber, A is a constant and E g is the energy band gap.
The bandgap energy value from the absorption data was calculated by plotting (αhν) 1/2 against the photon energy E g = hν (figure 5(b)) [53].Calculation results obtained E g values of samples T, CN, TCN, and TCNBC are 3.02 eV, 2.64 eV, 1.42 eV and 1.22 eV, respectively (figure 5(b) and table 1).Thus, a transition to lower energy has occurred for TCNBC composites, and the composite's ability to absorb visible light is also increased [53].This result implies that the formation of tight chemically bonded interfaces between the T, CN, and BC phases in the TCNBC composite can cause the TCNBC photocatalysts to shift to the lower energy region.It is also consistent that doping a semiconductor into a graphite-like carbon matrix can induce vacancies, leading to band gap narrowing [53].The small bandgap energy of TCNBC shows that the surface charge of the oxide increases in the composite, which can cause a modification of the basic process of electron-hole pair formation during irradiation [53].Photoluminescence (PL) spectra were further applied to study the optical properties.It is known that a decrease in emission intensity means a lower recombination rate of photo-generated carriers [54,55].The CN material exhibits a strong emission peak around 450 nm, indicating a high recombination efficiency of the photo-generated electron-hole pairs (inset of figure 5(c)).After inserting the T layer between the CN layers, the TCN composite exhibits a weaker emission peak than that of CN, suggesting that the photo-induced electrons in the conduction band of CN can be rapidly introduced T (figure 5(c)), thus directly preventing the recombination of electrons and holes [54,55].The TCNBC composite has a lower strength than all three T, CN and TCN samples because the photoelectrons can efficiently migrate from the TiO 2 surface to BC [54,55].The absorption of BC for luminescence leads to quenching effects [54,55], which leads to enhance separation efficiency between photo-generated electron-hole pairs [54,55].

Photodegradation of DC over TCNBC catalyst
The photocatalytic activity of the obtained catalysts was studied through DC photodegradation experiments under visible light irradiation (λ > 420 nm), as shown in figure 6(a).No photolysis of DC was observed after 2 h of visible light irradiation, demonstrating that DC is stable under visible light irradiation [38,56].Pure T exhibited low DC decomposition under visible light irradiation, as it is hardly excited by visible light radiation, consistent with previously reported results [38,56].In contrast, in the present work, pristine CN performed well in photodegradation under visible light.The TCN composite sample exhibited higher photocatalytic activity in DC degradation than pristine g-C 3 N 4 , and the TCN photocatalyst could photodegrade 65.35% DC in 120 min.The TCNBC photocatalyst has the highest photodegradation efficiency compared to the three samples, TCN and TCN, and it can photodegradable DC, reaching 91.93% in 120 min.
To better understand the reaction kinetics of the DC degradation catalyzed by different photocatalysts, experimental data were fitted by a pseudo-first-order model (Langmuir-Hinshelwood mechanism ) (equation ( 3)).As seen in the inset of figure 6(b), the photocatalytic degradation curves in all cases are consistent with the apparent first-order kinetic model (R 2 > 0.93).Furthermore, a significant difference in the photocatalytic activity was observed for T, CN, TCN, and TCNBC (figure 6(b)), implying that the incorporation of T was simultaneous with CN as well, and BC significantly improved the photocatalytic activity of T. Figure 6(b) shows that TCNBC composite has the highest degradation rate constant for DC photodegradation.The degradation rate constant values of T, CN, TCN, and TCNBC samples are about 9.6877 × 10 −4 min −1 , 0.0074 min −1 , 0.0150 min −1 and 0.0383 min −1 , respectively.TCNBC has a decomposition rate about 5.17 times and 2.55 times higher than pristine CN and composite TCN, respectively, and many times higher than pure T. The adsorption involves substantially to the catalytic degradation because the photoctalytic process is considered following the Langmuir-Hinshelwood mechanism in which a reaction of a molecules proceeding on a catalyst surface, followed in the degradation of this molecules by photo induced radicals.Comparing the firstorder apparent degradation rate constant of the present catalyst with previously published literature (as listed in table 2).The results show that the catalytic activity of the TCNBC composite is relatively high compared to the values reported in the literature.

The mechanism of DC degradation over TCNBC catalyst
The mineralization of DC over TCNBC is estimated by measuring COD values during reaction time (figure 7(a)).It is clear that the COD concentration gradually decreased with time, indicating a gradual decomposition of DC.A remarkable 88.50% COD reduction was achieved for DC antibiotics after 120 min of light illumination.Therefore, it can be concluded that the photocatalytic degradation of DC is mainly caused by CO 2 .The experiments of radical scavengers were also performed.For this experiment, 10 mM potassium iodide (KI), potassium bromate (KBrO 3 ), tert-Butanol (t-BN) and ascorbic acid (AA) were added to the DC solution, respectively.It was found that, in comparison with the degradation efficiency in the absence of a scavenger, the degradation efficiency was found to reduced to nearly 33.61%, 58.58%, 80.14%, and 71.88% compared with the   The photodegradation mechanism of DC can be explained by the theoretical bandgap in the solid physic.The valence and conduction band potentials (VB and CB) of a semiconductor can be determined according to the following equations [43,53].
where χ is the absolute electronegativity of the semiconductor, and the values of χ were 4.73 eV and 5.81 eV for g-C 3 N 4 and TiO 2 , respectively [41,43].The E e is the energy of free electrons versus hydrogen (4.5 eV), and E g is the bandgap value of the semiconductor [41].The valence band potential (E VB ) and conduction band potential (E CB ) were calculated as 2.775 eV and −0.155 eV for TiO 2 and 1.545 eV and −1.085 eV for g-C 3 N 4 .Due to the different positions of valence and conduction band potentials between TiO 2 and g-C 3 N 4 , type II heterojunction can occur in coupling effects [43].Pure TiO 2 semiconductors are hardly excited by visible light due to their large band gap (2.93 eV).Pristine g-C 3 N 4 , due to its narrow bandgap energy (2.63 eV) is excited by visible light to generate electron-hole pairs (equation (7)).Since the boundary potential of the conduction band (−1.085 eV) of g-C 3 N 4 is more negative than that of TiO 2 (−0.155 eV), photo-induced electrons in the CB of g-C 3 N 4 are transferred directly to the CB of TiO 2 (equation ( 8)), and then that is the surface of TCNBC photocatalyst [43].On the other hand, biochar on the TCNBC composite surface can also accept electrons on the conduction band of TiO 2 [43].The photosensitive electrons on the catalyst surface will react with O 2 dissolved in the solution and generate superoxide radicals •O 2 -(equation ( 9)).Superoxide radicals continue to react with adsorbed H 2 O molecules on the surface of the catalyst or H + ions present in the solution to form hydroperoxyl • OH radicals (equations ( 10)-( 13)).The E VB of TiO 2 (2.775 eV) is more positive than the E VB of g-C 3 N 4 (1.545 eV), the holes in the valence band of TiO 2 will move to the valence band of g-C 3 N 4 (equation ( 14)).While the E VB potential of g-C 3 N 4 (1.545 eV) is more negative than that of • OH/OH − (1.99 eV) [43], the holes in the valence band of g-C 3 N 4 will react with the OH − ion to forming the • OH radical (equation ( 15)), •OH hydroxyl radicals have strong oxidizing properties, so they will participate in the photodegradation of DC molecules (equation ( 16)) [43].Many documents have also shown that holes in the valence band of g-C 3 N 4 can directly oxidize DC molecules (equation ( 17)) [66,67].Based on the literature [43,66,[68][69][70] and the results above, a visible-lightdriven photocatalysis mechanism is possible at the interface of the proposed TCNBC heterojunction exported and shown in figure 8.

Recyclability and stability
For practical applications, it is necessary to evaluate the long-term stability of the photocatalyst during the reaction.The TCNBC catalyst was reused four times.After each photocatalytic activity test, the catalyst was separated by centrifugation, then eluted several times with methanol and finally dried at 105 °C for 24 h.The photocatalytic degradation efficiency of TCNBC decreased from 91.93% to 86.30% after four test cycles (figure 9(a)).The XRD spectra and FTIR spectra of the TCNBC samples before and after being reused at cycle four were almost unchanged; therefore, it can be concluded that TCNBC is stable in photocatalytic degradation reactions (figures 9(b), (c)).

Conclusions
The TCNBC composite was successfully synthesized by hydrolyzing a mixture containing CN, BC, and TiOSO 4 solution in an alkaline environment.TiO 2 in TCNBC is the form of the anatase phase, and its crystal size is around 5 to 10 nm.The TCNBC composite demonstrated high photocatalytic activity for decomposing DC in the visible light region.The kinetics of the photodegradation of DC on the TCNBC composite followed the Langmuir-Hinshelwood first-order kinetics.Furthermore, the TCNBC could catalyze for mineralizing DC and is stable during operation.The present study proposes a promising avenue for developing cost-effective photocatalysts produced from biomass, such as reed stalks.

Figure 6 .
Figure 6.(a) Photocatalytic degradation of DC over T, CN, TCN and TCNBC catalysts and the absence of catalyst; (b) Degradation rate constant k (min −1 ) for the photocatalytic degradation of DC over T, CN, TCN and TCNBC samples and the plot of pseudo-firstorder kinetics for the photocatalytic degradation of DC (insert image).Experimental conditions: catalyst dosage: 0.5 g.l −1 , DC 0 : 25 mg.l −1 and initial pH: 3.0.

Figure 7 .
Figure 7. (a) COD value decreases with irradiation time over TCNBC catalyst and (b) Effects of addition of KI, KBrO 3 , t-BN and AA on the degradation of DC in the presence of TCNBC under visible light.(Experimental conditions: catalyst dosage: 0.5 g.l −1 , DC 0 : 25 mg.l −1 and initial pH: 3.0).

Table 1 .
The average crystallite size (D), band gap energy (E g ), and textural properties of the synthesized samples.

Table 2 .
Comparison of rate constant of the present catalyst in the literature for the degradation of doxycycline. of a scavenger as each KI, KBrO 3 , tBA, or AA is added, respectively (figure 7(b)).KI, KBrO 3 , t-BA and AA are scavengers that react with photo-induced hole (h + ), electron (e − ), hydroxyl radical (•OH) and superoxide anion radical •O , 2 -respectively [63-65].tBA reduced the degradation efficiency significantly to 82.65%, which means that t-BA reacts with the hydroxyl radical (•OH radicals).Next, AA reduced the decomposition efficiency to 74.19%, contributing to the reaction of AA with •O , 2 -followed, KBrO 3 captures electron (e -) the degradation efficiency reduces to 42.99%.The strong reducing property of AA is widely utilized to probe the role of photoinduced holes in CB on the degradation of DC.The AA causes the loss of decomposition efficiency to only an 11.81% reduction.Therefore, it can be concluded that the formation of • OH and •O 2 [44,45]-radicals in the solution play a significant role in the photocatalytic degradation of DC antibiotics.The photo-generated electron holes in the photocatalyst react with the water and oxygen molecules adsorbed on the catalyst surface, generating • OH and •O 2 -radicals that react to decompose DC[44,45].