Synergy effect between tetracycline and Cr(VI) on combined pollution systems driving biochar-templated Fe3O4@SiO2/TiO2/g-C3N4 composites for enhanced removal of pollutants

Biochar-coupled Fe3O4@SiO2/TiO2/g-C3N4 composites were successfully constructed through simple sol–gel and calcination methods. The composites efficiently removed high-concentration toxic tetracycline (TC) by means of ·OH and ·O2-\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\rm O}_2^{-}$$\end{document}, whose removal rate exhibited 91.88% during 3 h, and the degradation rate constant reached up to 0.0068 min−1. The excellent performance can be attributed to the high specific surface area, enhanced visible light response, the introduction of magnetic nanoparticles and biochars expediting charge transfer, Z-scheme heterojunction enhancing the spatial separation of photo-generate carriers and, importantly extraordinary adsorption capacity of 147.96 mg g−1. Moreover, the composites showed the most excellent efficiency under the calcination temperature of 450 ℃, and exhibited good stability with tolerance to a wide range of pH and anions. Interestingly, a synergistic photocatalytic effect was discovered in the TC/Cr(VI) combined pollution systems, resulting in significantly improved removal of Cr(VI). Besides, the photocatalytic mechanism and degradation path of tetracycline were also elucidated. All these findings suggested the as-synthesized catalyst was an excellent photocatalyst for removal of TC/Cr(VI)-contaminated water. Synthesized biochar-based catalyst enhanced adsorption and photocatalytic activity. The catalys simultaneously removed 91% of TC and 79% of Cr(VI) in combined systems. Synergistic photocatalytic effect accelerated TC and Cr(VI) removal in combined systems. Synthesized biochar-based catalyst enhanced adsorption and photocatalytic activity. The catalys simultaneously removed 91% of TC and 79% of Cr(VI) in combined systems. Synergistic photocatalytic effect accelerated TC and Cr(VI) removal in combined systems.


Introduction
Nowadays, wastewater remediation is one of the critical issues due to the complexity of water environment, which has been detected to contain various types of chemicals, such as heavy metals, organic matter, etc., posing serious threats to the ecological environment and even human health (Li et al. 2021a;Sellaoui et al. 2021;Yu et al. 2022). Among these pollutants, Tetracycline (TC), an emerging antibiotic, and chromium Cr(VI), a common poisonous heavy metal, are difficult to remove from water by effective methods (Wang et al. 2020). It is urged to develop efficient methods for the removal of TC and Cr(VI).
Alternatively, photocatalysis has shown considerable potential for pollutant removal due to its critical advantages, such as the strong ability to completely degrade the recalcitrant pollutants, and low environmental risks without any oxidizer addition (Buzzetti et al. 2019;Long et al. 2020;Ravelli et al. 2009). Nevertheless, its practical large-scale application is limited due to particular challenges, such as limited light use efficiency and easy recombination of photo-generated charge carriers (Hu et al. 2021;Qiu et al. 2021a, b, c). Based on the energy band theory, with a heterojunction, the band gap energy can be significantly reduced, so light absorption and charge transfer will be increased (Wen et al. 2017). Therefore, the construction of heterojunction catalyst has drawn broad attention of researchers owing to its easy operation and high photocatalysis activity. The heterojunction photocatalysts (e.g., Bi 2 MoO 6 /ZnO, AgI/ CeO 2 , ZnO-TiO 2 and MoO 3 /Bi 2 O 3 /g-C 3 N 4 ) have been used to remove pollutants in water (Das et al. 2020;Li et al. 2021b;Zhang et al. 2019Zhang et al. , 2021. The introduction of CdS heterojunction greatly improves the absorption of ZnO in the visible region and the photoelectron-hole separation efficiency through the interface charge transfer (Wang et al. 2019). However, these photocatalysts are still unable to meet the requirements for practical applications, such as the development of a non-metal, green, high photocatalytic activity, and the facility in dispersion, separation and recovery of the photocatalysts from the solution (Kumar et al. 2020;Zhao et al. 2020). Among heterojunctions, as we all know, Z − scheme heterojunction exhibits higher photocatalytic activity compared with conventional type − II heterojunctions, due to high redox potential and separation of the charge carriers (Huo et al. 2021;Jourshabani et al. 2020). The graphitic carbon nitride (g-C 3 N 4 ), as a nonmetallic photocatalyst, exhibits wide visible light response but low redox ability (Wen et al. 2017), and titanium dioxide (TiO 2 ), as a most commonly used photocatalyst, shows low light use efficiency but excellent redox ability (Guo et al. 2019). Based on the characteristics of high efficiency, low toxicity, and particular band and electronic structures, g-C 3 N 4 and TiO 2 are considered to be ideal components for the construction of Z − scheme heterojunction.
Recently, a series of TiO 2 /g-C 3 N 4 -based heterojunctions were developed to enhance the photocatalytic performance of pollutant removal. For example, a synthesized magnetically g-C 3 N 4 /TiO 2 /Fe 3 O 4 @SiO 2 nanophotocatalyst exhibited the removal of 97% of ibuprofen within 15 min of visible light irradiance due to the low recombination rate of photogenerated carriers (Kumar et al. 2018a). Similarly, the construction of Fe 3 O 4 @ SiO 2 @g-C 3 N 4 /TiO 2 core-shell microsphere nanocomposite exhibited 27.7-fold higher for Methyl Orange degradation in comparison to TiO 2 due to the modified band structure (Narzary et al. 2020). In order to further improve the photo-response and redox ability, the TiO 2 /g-C 3 N 4 -based heterojunctions were modified to synthesize a magnetically terephthalic acid functionalized g-C 3 N 4 /TiO 2 photocatalyst, which showed enhanced degradation of pharmaceutical and personal care products (PPCPs) (Kumar et al. 2020). However, attention should still be paid to some challenges in the practical application of TiO 2 /g-C 3 N 4 -based heterojunctions: (a) low adsorption activity stemming from agglomeration, and (b) development of nonmetal heterojunction photocatalysts with lower costs (Qiu et al. 2021a). Recent years have witnessed the wide application of biochar-based materials in the field of photocatalysis for enhancing photocatalytic performance due to its excellent physicochemical properties, such as large specific surface area, abundant surface functional groups and mineral components (Kumar et al. 2018b;Lyu et al. 2020;Tam et al. 2020). Accordingly, we supposed that the introduction of biochars could also improve the photocatalytic performance of TiO 2 /g-C 3 N 4 -based heterojunctions associated with the advantage of extensive availability and conducive physio-chemical surface characteristics Yang et al. 2022). Furthermore, biochar could be expected to act as a promising alternative to semiconductor components of photocatalysts for reducing environmental risks. Nonetheless, the effects of the introduction of biochars on the optical electronic properties of the heterojunction, as well as the performance of pollutant removal have been poorly explored. Numerous studies tend to focus on the single pollutant system merely, while the photocatalytic removal of pollutants in the combined pollution systems has been barely characterized.
In this study, the biochar-coupled Fe 3 O 4 @SiO 2 /TiO 2 /g-C 3 N 4 composites were successfully synthesized through simple strategies for the photocatalytic oxidation of TC and removal of Cr(VI). To compared nonmetallic g-C 3 N 4 -based photocatalysts, metallic CdS-based photocatalysts were also prepared. The effects of controlled parameters, including calcination temperature, pH, and ion strength, on the removal performance of TC were studied. The removal performance, possible mechanisms, degradation pathways of TC, as well as reusability of the synthesized composites in the combined pollution systems were investigated.

Materials
The biochar raw material rice straw used in this research was obtained from Changzhou,

.1 Preparation of Fe 3 O 4 nanoparticles
Fe 3 O 4 nanoparticles were synthesized by a simple coprecipitation method. Briefly, A 50 mL mixture of Fe2+/ Fe3+ in the ratio of 1:1 was prepared with FeCl3·6H2O and FeSO4 ·7H2O, and vigorously stirred at 80 ℃ under the atmosphere of N 2 for 10 min. Then, 0.048 g·mL −1 of trisodium citrate dihydrate solution was added into the above mixture after pH was adjusted up to 10-11 by adding 5 M NaOH solution, which was continued whisking for 1 h to obtain the precipitated black Fe 3 O 4 nanoparticles.

Preparation of Fe 3 O 4 @SiO 2 composites
In order to prevent the corrosive effect from the photoelectrons, Fe 3 O 4 nanoparticles were covered by SiO 2 via the sol-gel method. 0.50 g of Fe 3 O 4 nanoparticles were dispersed into 80 mL of ethanol, and then 10 mL of 28% ammonium hydroxide and 18 mL of deionized water were added to the above mixture. The mixture was added with 5 mL of tetraethyl orthosilicate and kept stirring for 4 h. Finally, the composites were washed with ethanol and water several times and dried at 60 ℃.

Synthesis of biochar-coupled Fe 3 O 4 @SiO 2 /TiO 2 composites
The biochar-coupled Fe 3 O 4 @SiO 2 /TiO 2 (FSBT) photocatalyst was prepared by the similar method. 0.20 g of Fe 3 O 4 @SiO 2 nanoparticles and 0.50 g of modified rice straw-derived biochar prepared in our previous work (Dai et al. 2020) were dispersed in 50 mL of ethanol and sonicated for 20 min, respectively. After that, they were mixed, and 28% ammonium hydroxide was added dropwise into the suspension to adjust the pH to 8-9. 20 mL of titanium butoxide was added to the solution and vigorously stirred at 80 °C for 2 h. Subsequently, the mixture was heated hydrothermally by a Teflon-lined autoclave at 160 °C for 3 h.

Synthesis of biochar-coupled Fe 3 O 4 @SiO 2 /TiO 2 /g-C 3 N 4 composites
Biochar-coupled Fe 3 O 4 @SiO 2 /TiO 2 /g-C 3 N 4 composites (FSBTG) were prepared via a modified in accordance with the previous study (Kumar et al. 2018a). The g-C 3 N 4 was prepared by thermal condensation of melamine molecules. Typically, melamine was calcined at 550 ℃ for 4 h in a muffle furnace at a heating rate of 20 ℃ min −1 . Finally, the powder from FSBT was milled with prepared g-C 3 N 4 and calcined (300, 450, and 600 ℃) for 1 h in a tube furnace at a heating rate of 20 ℃ min −1 to synthesize FSBGT. The biochar-coupled Fe 3 O 4 @SiO 2 /TiO 2 /CdS photocatalyst (FSBTS) was also prepared by a similar method. The major steps are shown in Additional file 1: Fig. S1.

Characterization
The characterization details of as-synthesized catalysts are listed in the Additional file 1.

Performance test
The adsorption and photocatalytic activities of as-synthesized several photocatalysts were evaluated using high concentration TC solutions as the targeted contaminant.
All adsorption experiments were conducted in a 250 mL conical flasks under the conditions of 100 mL of TC solution, 0.10 g of catalyst dose, agitator speed of 150 r min −1 and reaction temperature of 298 K. In the adsorption kinetic studies, the initial TC concentrations were set to 200 mg L −1 , and the solutions were sampled every ten minutes for concentration measurement. Meanwhile, the initial TC concentrations were adjusted to the range from 10 to 200 mg L −1 with an equilibrium time of 60 min. All catalysts were stirred for 120 min in the dark to ensure adsorption equilibrium before photocatalytic degradation experiments. The photocatalytic activity was assessed by the photodegradation of TC under simulated solar light using a 200 W Xenon lamp as the light source. ) on the photocatalytic reaction of 100 mg L −1 TC solution were also studied. To distinguish the role of the reactive species in the photocatalytic reaction, scavengers, such as tert-butyl alcohol (tBuOH), potassium dichromate (PD), EDTA-2Na and benzoquinone (BQ), were added to the 100 mg L −1 TC solution with catalysts for scavenger experiments. Besides, the photocatalytic activity of catalysts in the combined pollution systems with TC of 100 mg L −1 and Cr(VI) of 20 mg L −1 was further studied. The catalyst reusability in the combined pollution systems was also investigated by adding contaminant-loaded photocatalysts to 0.5 M NaOH solution after each use, stirring at 353 K for 1 h, and then replacing it with deionized water and stirring for another 1 h for complete desorption. The regenerative samples were washed, dried at 65 ℃ and applied for the next reusability test until removal efficiency was stabilized. Cr(VI) concentrations were detected via a diphenyl carbazide method using a UVvis spectrometer (UV-1901PC, Shanghai) at λ of 540 nm, and TC concentrations also were detected at λ of 357 nm. The photocatalytic intermediates in TC solution were identified using high-performance liquid chromatography coupled with mass (Thermo scientific Q Exactive Ultimate 3000 UPLC-MS).

Morphology and textural characteristics
The characters of the surface morphology of prepared composites were observed through SEM, EDS, TEM, and HRTEM ( Fig. 1; Additional file 1: Fig. S2). The g-C 3 N 4 exhibited a bulk irregular layered structure, and agglomerated to each other, resulting from the polycondensation of melamine during the process of calcination . The micrograph of FSBT (Fig. 1b) showed that titanium dioxide covered Fe 3 O 4 @SiO 2 hybrid nanoparticles dispersed over the porous biochar matrix. It could be found that relatively small-sized g-C 3 N 4 nanosheets and cubic CdS particles existed, which were coated to biochar-coupled FST nanoparticles to form biochar-coupled FSTG and FSTS composites, respectively. In addition, the different elements were distributed on the surface of two catalysts ( Fig. 1f-l; Additional file 1: Fig. S2f-l), obtained via the EDS elemental mappings of the SEM images, indicating the successful construction of catalysts. The TEM and high-resolution TEM (HRTEM) images clearly presented the different morphologies and phases of the composite catalyst. It can be seen that the TiO 2 nanoparticles wrap the CdS and g-C 3 N 4 to form two heterojunctions. The different phases identified by HRTEM images and selected area electron diffraction (SAED) (Fig. 1n, o) have lattice fringe spacings of 0.32, 0.31 and 0.35 nm, respectively, which are assigned to the (002) and (101) crystal plane of g-C 3 N 4 and TiO 2 , respectively (Wang et al. 2018b). In terms of FSBTS catalyst, the lattice fringes of interplanar spacing of about 0.319, 0.332 and 0.35 nm could be ascribed to the (002), (110) and (111) crystallites of hexagonal CdS, respectively (Additional file 1: Fig. S2n, o). The presence of high resolution lattice fringes and diffraction cycles of SAED patterns showed the highly crystalline nature of the conductor nanoparticles of assynthesized composite catalysts (Kumar et al. 2018a).
The textural properties of catalysts were studied through N 2 adsorption-desorption isotherms and pore size distribution curves (Additional file 1: Fig. S3). Compared with FST, FSBT and biochars, a higher capacity of N 2 adsorption-desorption was observed due to the introduction of the heterojunction. All catalysts show increasing hysteresis pattern without plateau at high relative pressure, indicating the shape might transform from disorder lamellar, slit and wedge-shaped pore structure to regular shape after secondary calcining. However, the H3 loop characterized the structure of synthesized catalysts based on the inconspicuous steady trend . Conversely, the hysteresis pattern of biochars is more likely to be a narrow silt-like structure . Specifically, when P/P 0 < 0.4, the isotherm showed an upward trend, indicating the existence of micropores, and when P/P 0 > 0.4, the hysteresis loop increased, indicating the existence of abundant mesoporous structures. The adsorption capacity of composites containing biochar or heterojunctions was higher than that of other catalysts, which could be ascribed to the development of more surface area and pore volume, resulting from further calcination during the process of preparation of heterojunctions. The continually rising trend isotherm of all catalysts suggested the presence of both microporous and mesoporous structures (Cai et al. 2013), and it was also further confirmed by pore size distribution curves. The specific surface area (S) of all catalysts accorded with the following rule: S FSTBG (280.30 m 2 g −1 ) > S FSTG (266.41 m 2 g −1 ) > S FSTBS (255.07 m 2 g −1 ) > S FSTS (247.07 m 2 g −1 ) > S FSBT (135.87 m 2 g −1 ) > S FST (111.05 m 2 g −1 ) > S BC (28.13 m 2 g −1 ), and a similar rule was also found in the total pore volume, which could be attributed to the aggregation of flaky g-C 3 N 4 and granular CdS. The detachment of unstable impurities from the surface of composites could also improve the surface characteristics of synthesized catalysts (Table 1).

Composition and structural characteristics
The crystal structures of the synthesized catalysts were analyzed via XRD (  (311) (222), respectively. Same peaks could be observed in the catalysts containing TiO 2 , g-C 3 N 4 , or CdS, but the peak intensity decreased gradually due to the very low quantity detected on the surface of the catalysts. Interestingly, the peak intensity of TiO 2 -based heterojunctions was enhanced with the increase of calcination temperature from 300 to 600 ℃, which could be ascribed to the decrease in thickness of g-C 3 N 4 . When the calcination temperature reached up to 600 ℃, new peaks at 36° and 41.1° were observed, indicating the formation of rutile (101) and (111) plane under high calcination temperatures . As the synthesis temperature of CdS-based catalyst increased to 600 °C, the composite catalyst generated new hexagonal (101) and (103) crystal planes (Wang et al. 2018b).
The functional groups of the as-prepared catalysts were studied by FT-IR spectra ( Fig. 2b; Additional file 1: Fig.  S4b). The peaks at 460 and 540 cm −1 correspond with the stretching vibrations of Fe-O (Nematollahzadeh et al. 2015). The peaks at 808 cm -1 indicate the bending vibration of s-triazine units , several peaks in the range of 880-1335 cm -1 can be attributed to the stretching vibration of N-H/N-H 2 , C-N and C=N, and these peaks are attributed to the existence of g-C 3 N 4 (Kumar et al. 2018b;Li et al. 2019). The relatively wide peaks near 3140 cm −1 stem from the bending vibration of terminal NH 2 or NH groups at the defect sites of the aromatic ring (Yan and Yang 2011). Besides the above peaks, several new peaks indicated stretching vibrations of Ti-O, Ti-O-Ti and Si-O were observed at 400-600 cm −1 in as-synthesized catalysts (Nematollahzadeh et al. 2015). The peaks at 1500-1700 cm −1 and 3340-3550 cm −1 are attributed to the stretching vibration of the H-O-H group and -OH group (Cheng et al. 2016;Dai et al. 2020), which are derived from the biochar, TiO 2 , and CdS, suggesting the successful construction of composite catalysts .
The elemental states of the synthesized photocatalysts were analyzed via XPS in the C 1s, N 1s, Ti 2p and Fe 2p binding energy regions. As shown in Fig. 2c-f, the C 1s peaks at 284.4 and 286.2 can be attributed to the C atoms in the C=C/C-C and C-OH bonds, respectively, which originate from the graphitic and amorphous C of NaOHactivated biochar, whereas the peak centered at 288.5 eV indicates the C atoms in the structure of N = C-(N) 2 Wang et al. 2020). The obvious N 1 s peak at 399.0 eV can be considered as the sp 2 -hybridized aromatic N atoms derived from C=N-C, the weak peak at 400.3 eV is assigned to (C) 3 -N linking structural motifs of C 6 N 7 , (C) 2 -NH/C-NH 2 , or Ti-ON, and the peak at 401.1 eV is indicated to N atoms of aromatic cycles of N-(C) 3 Wang et al. 2020;Zhang et al. 2018). As for Ti 2p, two obvious peaks located at 459.1 and 464.9 eV are attributed to the Ti 2p 3/2 and Ti 2p 1/2 state of Ti 4+ , respectively (Ma et al. 2021). A satellite peak at 472.1 eV can be clearly seen for the Ti(IV) oxidation state (Herath et al. 2022). For Fe 2p, the peaks at around 711.5 and 724.5 eV are assignable to Fe 2p 3/2 and Fe 2p 1/2 spin-orbit, respectively (Djellabi et al. 2019). Moreover, it can be curve-fitted with five peaks. Among them, the ones at the binding energies of 710.9 and 724.4 eV are in accordance with Fe 2+ Raha and Ahmaruzzaman 2020), the one at 712.5 is consistent with Fe 3+ , and the ones at 719.3 and 732.8 eV are the satellites peaks of Fe 3 O 4 .

Optical and electronic characteristics
The optical properties of a catalyst directly influence its photocatalytic activity. The light response characters of the as-synthesized catalysts were analyzed by ultraviolet-visible/diffuse reflection spectroscopy (UV-vis/DRS). In Fig. 3a, g-C 3 N 4 exhibits an obvious light response at 200-450 nm, indicating a certain visible response. For g-C 3 N 4 -based photocatalysts, the light absorption over the range of 200-800 nm can be attributed to the decreased reflectivity by introducing black biochars, which improved the electronic transition efficiency by capturing photons (Li et al. 2015). The band gap plots are also plotted according to Tauc relation (Kumar et al. 2017) in Fig. 3a. It can be found that the band gap of composite catalysts is lower than that of g-C 3 N 4 , which could be attributed to the formation of heterojunctions between the employed photocatalysts (Jourshabani et al. 2020). In addition, the band gap energy of semiconductors can be decreased by the introduction of biochar. It can be interpreted as the reason that biochar can sensitize the semiconductor with another composite, create a mid-gap energy state via doping non-metals such as N, S, C, O etc., and form a local trapping state below the conduction band, meanwhile, the N can also reduce the band gap of the composite by forming a mid-gap energy state . A red shift was found for CdS based photocatalysts from Fig. 3b, indicating a stronger ability to capture photons compared with CdS. This could result from the formation of heterojunction among biochars, Fe 3 O 4 @SiO 2 , TiO 2 and CdS.
The photocurrent responses and electrochemical impedance spectroscopy (EIS) were studied to evaluate the separation efficiency of photo-generated carriers. It can be seen in Fig. 3c that a steady photocurrent was generated, and the switching on of current was done after every 50 s. FSBTG-450 and FSTG-450 increase significantly in photocurrent compared to g-C 3 N 4 , suggesting the reduction of the photogenerated charge carrier recombination (Guo et al. 2021). It could be deemed that introducing biochar can further improve the separation of photo-generated charge carriers of photocatalysts, in accordance with the phenomenon that the FSBTG-450 shows a higher photocurrent intensity than FSTG-450. The results were also found in previous studies (Lyu et al. 2020;Tam et al. 2020). Significantly, a prominent decrease occurred in photocurrent of CdS-based photocatalysts in comparison to CdS (Fig. 3d), although biochar, Fe 3 O 4 @SiO 2 , TiO 2 could promote the absorption of visible light. The reduced charge carrier recombination of the composite catalysts can be reconfirmed by EIS analysis in Fig. 3e, in which the composite catalysts exhibit lower charge transfer resistance and small Nyquist semicircle compared with single g-C 3 N 4 . It can be explained via the following main reasons: (1) the aromatic ring structure of biochar can accelerate the electron shuttling between different reactants, which could be mediated via surface redoxactive moieties; (2) biochar containing large amounts of quinone can act as an electron reservoir, while the electron storage capacity of biochar depends on the types of biomass and the pyrolysis conditions; (3) the formation of heterojunction between the semiconductor and biochar is responsible for potentiation in charge separation, and the similar increase in charge separation can occur at metal Fe or iron oxides and semiconductor heterojunctions, which can be promoted by light irradiation . Interestingly, inconsistent results are presented between Fig. 3d and Fig. 3f. FSBT-300 and FSBTS-300 revealed lower charge transfer resistance than CdS on account of Nyquist semicircle, a possible reason could be that the heterojunction is beneficial to the migration of carriers but ineffective to the separation of photogenerated electron/hole.

Removal of TC from water by adsorption
Adsorption plays an important role in the process of photocatalytic removal of pollutants by catalysts. The kinetics and isotherm models were used to describe the adsorption capacity and mechanism of TC with the synthesized catalysts (Fig. 4). In order to research the adsorption kinetic characteristics, the experimental data were fitted by pseudo-first-order and pseudo-secondorder models, respectively. The higher correlation coefficients of the two models suggest that both physical and chemical mechanisms were involved in this adsorption process. TC sorption by catalysts rose rapidly during the first 20 min, then gradually slowed down until sorption equilibrium after around 60 min. Compared with FST and FSBT, the equilibrium adsorption capacity of the catalysts introducing g-C 3 N 4 increased by 116.47 mg g−1 and 102.34 mg g−1, respectively, and that of the catalysts introducing CdS increased by 119.61 mg g−1and 120.95 mg g−1, respectively. Moreover, the adsorption capacity of FSBTS gradually decreased as calcination temperature increased, and the catalysts calcined at 300 °C showed the best performance. Maximal adsorption capacity was presented under 450 °C for FSBTG. The results could be explained via the reasons that the structure and physicochemical properties connected with the adsorption of rice straw-derived biochar can be easily changed as calcination temperature increases. It has been suggested in previous studies that the biochars at high temperatures shows structural defects, blocked micropores on the surface (Rosales et al. 2017), and a decrease of carboxylic functional groups (Zhao et al. 2017), resulting in the decrease of adsorption active sites. Similarly, both Langmuir and Freundlich isotherm models can fit the isotherm experimental data well. q e gradually increased along with the increase of C e due to the more powerful driving force and contact area between high concentration TC and adsorbent (Dai et al. 2020). FSBTS-300 and FSTS-300 showed maximum excellent adsorption capacities reaching up to 173.32 mg g −1 and 171.82 mg g −1 , while lower maximum adsorption capacities of FSBTG-450 and FSTG-450 came up to 147.96 mg g −1 and 104.70 mg g −1 , respectively. Metal components decreased and nonmetallic g-C 3 N 4 , barely adsorbing TC, replaced Fe 3 O 4 @SiO 2 and TiO 2 to occupy the active sites, which caused the above phenomenon. In general, the Langmuir model had a higher nonlinear relationship coefficient, indicating that TC adsorption may be single-layer molecular adsorption, and chemical adsorption is involved in this process (Zeng et al. 2019). The key parameter n of the Freundlich model is related to the degree of heterogeneity of adsorption sites, and all n values in this study were < 1, indicating the high heterogeneity of the solution during the TC adsorption process (Srivastava et al. 2006). The presence of oxygen-containing groups, such as hydroxyl and carboxyl groups, can contribute to the formation of H-bonding between photocatalysts and TC (Rosales et al. 2017). It is reasonable to assume that the coating of nanoparticles on carbon structure could also increase the adsorption sites, which can be confirmed via the result that the introduction of metal nanoparticles, such as TiO 2 , Fe 3 O 4 , and SiO 2 , further improved the adsorption of biochar or biochar-based catalysts. Consequently, besides H-bonding, the interaction of surface complexes could be considered as a dominant mechanism in TC adsorption (Tan et al. 2016).

Removal of TC from water by photocatalytic degradation
The photolysis experiments under simulated sunlight with different as-synthesized photocatalysts were conducted to investigate their photolysis efficiency. As shown in Fig. 5a, b, in the absence of any kinds of catalysts, the TC concentration remained unchanged with increasing irradiation time, indicating negligible photolysis of TC without photocatalyst. The composite catalysts exhibited relatively high photocatalytic efficiency, and FSBTG-450 and FSBTS-300 showed the most excellent efficiency of 88.20% and 91.88%, much higher than that of pure g-C 3 N 4 (12.65%) and CdS (47.21%). Similarly, the catalysts synthesized by high-temperature calcination showed relatively low removal performance, and 450 ℃ and 300 ℃ were the optimal temperatures for FSBTG and FSBTS, respectively, which could be ascribed to the formation of anatase TiO 2 with higher weight and activity at the suitable conditions ). Due to the faster rate of adsorption than that of degradation, the TC removal efficiency rose rapidly during the first 20 min, subsequently, the activity sites were occupied by adsorbed TC and intermediates of incomplete degradation, resulting in gradual slowness in removal rate. This phenomenon also suggests that adsorption plays a critical role in all processes of TC removal. The removal of TC followed degradation kinetics, as shown in Fig. 5c, d; Additional file 1: Table S3, the removal rate of catalyst follows the following order: FSTG-450 (0.0071 min −1 ), FSBTG-450 (0.0068 min −1 ), BTG-450 (0.0051 min −1 ), FSBTG-300 (0.0025 min −1 ), FSBTG-600 (0.00027 min −1 ), and g-C 3 N 4 (0.00026 min −1 ). k value of FSBTG was 1.33 times that of BTG, suggesting that Fe 3 O 4 @SiO 2 composites enhanced the degradation rate on TC. Based on the characterizations of catalysts, it can be seen that the catalysts introducing g-C 3 N 4 and CdS had larger specific surface area, wider visible light response range, and high photogenerated carrier separation capacity. Different from CdS, introducing g-C 3 N 4 to FSBT showed lower adsorption capacity, which improved photocatalytic activity at the cost of decreasing the adsorption activity. During the process of TC removal by FSBTG, photocatalytic oxidation, which played a leading role, was relatively slow and difficult compared with adsorption. Therefore, FSBTG and FSBT showed similar adsorption capacities. It can be discovered that the incorporation of biochar had almost no effect on photocatalytic performance, although biochar was considered as an electron reservoir in previous studies . As for temperature, the same results could be found in adsorption and photocatalytic degradation. It was reported that biochar synthesized at high heat temperatures showed a higher electron storage capacity (Kluepfel et al. 2014). In this study, the catalysts with the high calcination temperature exhibited decreased performance, suggesting that the electron storage capacity of biochar may be not the primary factor affecting photocatalytic performance, and there is no obvious evidence about electron storage stemming from biochar. In addition, a previous study showed that the catalytic activity of carbon-based materials also probably depends on the internal electronic state of the hybrid covalent system (Duan et al. 2018). Oxygen-containing functional groups are the main active sites for the catalytic oxidation of carbonaceous catalysts (Wang et al. 2018b), and the introduction of metal oxides and biochars may also cause O doping to g-C 3 N 4 , resulting in the generation of more active sites and high photocatalytic activity. Compared with photocatalysts synthesized in recent studies, as illustrated in Table 2, various photocatalysts exhibited significantly different removal performances for TC due to the difference in characters of catalysts, reaction conditions, as well as photocatalytic mechanisms (Saadati et al. 2016). Above all, the as-synthesized FSBTG-450 catalyst could be considered as an efficient material with high removal performance for high concentration TC solutions.

pH
It was deemed that the ionization state of pollutants and the surface charge of the photocatalyst are affected via pH (Kumar et al. 2018a). Similar results were exhibited for g-C 3 N 4 -and CdS-based photocatalysts on the effect of pH. The solution pH was adjusted using 0.1 M HCl and NaOH, and the effect of pH on the photocatalytic removal of TC by FSBTG-450 is shown in Additional file 1: Fig. S5a. After 1 h of light irradiation, 88.78%, 87.68%, 83.60%, 77.18%, 76.40% and 77.33% of TC were removed when the pH of the solution was not adjusted, 3, 5, 7, 9 and 11, respectively. Generally, the state of TC is affected by the pH of the solution. When pH < 3.4, 3.4 < pH < 7.6, 7.6 < pH < 9.7 and pH > 9.7, TC exhibits the state of H 4 TC + , H 3 TC, H 2 TC − and HTC 2− , respectively (Jang and Kan 2019). It was found that the surface of adsorbents was positively charged below pH zpc , otherwise it was negatively charged (Dai et al. 2020). From Additional file 1: Fig. S8, the zero point of charge (pH zpc ) of FSBTG is 4.43. Obviously, the catalyst exhibited a relatively high removal efficiency when pH < pH zpc , which could be attributed to the electrostatic interaction between the catalyst and TC ions or molecules, but activity was inhibited when pH > pH zpc due to the electrostatic repulsion. However, the effect of pH on the removal of TC was not very significant, indicating that other mechanisms might play critical roles in this process. To preliminarily evaluate the applicability of FSBTG-450 in the actual water environment, various 0.1 M anions, including Cl − , NO − 3 , SO 2− 4 , and PO 3− 4 , were used to explore the effect on the photocatalytic removal of TC. As shown in Additional file 1: Fig. S5c, the photocatalyst exhibited 88.36% of removal efficiency during 1 h without any additional anion. The removal efficiency was 82.46%, 91.86%, 88.62%, and 9.47% in the presence of anions of SO 2− 4 , Cl − , NO − 3 , and PO 3− 4 , respectively. Apparently, the effect of Cl − on TC removal was positive, that of NO − 3 was negligible, while a slightly inhibitory effect of SO 2− 4 on the reaction was observed, which can be ascribed to the reactive species scavenging reaction from SO 2− 4 : ·OH + SO 2− 4 → OH − + SO ·− 4 , h + + SO 2− 4 → SO ·− 4 . Moreover, TiO 2 existing on the surface of the composite catalyst can adsorb SO 2− 4 through van der Waals forces and hydrogen bonds, occupying the active sites of the catalyst (Xekoukoulotakis et al. 2011). The strong inhibitory effect was observed in the presence of PO 3− 4 due to the strong adsorption affinity of PO 3− 4 for TiO 2 and biochar, resulting in a large number of surface sites that were competitively occupied (Kudlek et al. 2016;Wang et al. 2020). In addition, ·OH was caught and the transfer of h + from g-C 3 N 4 and CdS to TiO 2 through biochars and Fe 3 O 4 @SiO 2 was hindered (Kudlek et al. 2016). FSBTS-300 showed similar results.

Active substances during degradation
E − and h + are the main reactive species produced on the surface of the photocatalysts during the photocatalytic removal process. After that, they combine with dissolved O 2 and H 2 O to form other reactive species including ·OH and · O − 2 (Kumar et al. 2018a). In order to identify the dominant active substances controlling the photocatalytic removal of TC under light irradiation, different scavengers of 1 mM, tBuOH (·OH scavenger), PD (e − scavenger), EDTA-2Na (h + scavenger) and BQ ( ·O − 2 scavenger) were added into the reaction systems, respectively (Kumari et al. 2022). The experiment was also conducted without any scavenger during degradation for reference. From Additional file 1: Fig. S5d, it was discovered that the removal efficiency decreased a little after adding PD and EDTA-2Na, so e − and h + had a slight effect on the removal efficiency of TC. Nevertheless, the photocatalytic activity reduced clearly due to the introduction of BQ, indicating the importance of · O − 2 in TC removal. Moreover, the degradation efficiency of TC reduced greatly after the addition of tBuOH, elucidating that ·OH was the primary active species during the TC degradation process. To further validate the generation of active species during the reaction process, the ESR spin-trap with DMPO technique was carried out. As shown in Additional file 1: Fig. S6a, the TEMPO signal peak strength fell with longer illumination times, which indicated that FSBTG could generate electrons and holes due to the anaerobic condition of acetonitrile solvent, so the heterojunction enhanced the spatial separation of photo-generate carriers (Wang et al. 2020). Moreover, there was no EPR signal for the DMPO-·OH or DMPO-·O − 2 adducts under dark conditions, indicating that no active free radicals were generated. However, four characteristic peaks for DMPO-·OH and six characteristic peaks for DMPO-·O − 2 were found in the presence of light and enhanced with increasing light irradiation time (Li et al. 2022a), demonstrating that FSBTG could produce ·OH and · O − 2 during the photocatalytic reaction. The contribution of reactive oxygen species (ROS) in photodegradation was obtained in the following order: ·OH > ·O − 2 > h + > e − according to Additional file 1: Fig.  S6d. As a result, ·OH and · O − 2 were generated during the photocatalytic degradation of TC, which was consistent with the results of the quenching experiments. H 2 O could react with h + to produce ·OH, whereas · O − 2 , generated by the reaction between e − and O 2 , could react further with H 2 O to produce ·OH (Chong et al. 2010;Fang et al. 2015). Therefore, it is a revelation that the photocatalytic efficiency in TC solutions with high concentration can be promoted by adjusting the ratio of H 2 O and O 2 for the practical application of photocatalysts.

Simultaneous removal of TC and Cr(VI) in the combined pollution systems
The pollutant removal performance of FSBTG and FSBTS was further discussed in TC/Cr(VI) combined pollution systems. The UV-vis spectroscopic spectra at different times are shown in Fig. 6a, b. Obviously, during the whole reaction course, except for the typical absorption peaks, new peaks emerged and these two peaks synchronously decreased with offset, suggesting that the pollutants were altered, not merely adsorbed. The pollutant removal performance of photocatalysts in single and TC/ Cr(VI) combined pollution systems are shown in Fig. 6c, d. Interestingly, in terms of FSBTG, the removal of Cr(VI) in the single system without adjustment of pH can be disregarded, with only 2.7% of Cr(VI) being removed within 120 min under light irradiation. The significantly increased removal of 69.99% and 77.95% occurred in the combined pollution systems, as shown in Additional file 1: Table S3, and the degradation rates, k value, could reach up to 0.0093 min −1 for FSTG and 0.0108 min −1 for FSBTG, which were over 46 and 54 times more than that of FSBTG in the single system. Likewise, the removal of TC in the combined pollution system also slightly increased compared with the single system. k values were 1.99 and 2.24 times for FSTG (0.0141 min −1 ) and FSBTG (0.0152 min −1 ), respectively. The significantly enhanced removal of Cr(VI) can be explained by the presence of the synergistic photocatalysis effect in the combined pollution systems. In a Cr(VI) single pollution system without adjustment of pH, the Cr(VI) is barely removed because of the high electron cloud density of O atoms in Cr 2 O 7 2-, resulting in the difference in the generation of the reducing active intermediate (RAI) and oxidizing active intermediate (OAI). The redox between TC and Cr(VI) occurring in the combined pollution systems without adjustment of pH can be interpreted by the fact that the H + originating from TC can promote Cr 2 O 7 2to form RAI and OAI . It was observed that TC removal in the combined pollution systems was slightly higher than that in the TC single pollution system. , Cr (III) may adhere to the surface of catalysts through adsorption, accelerating the migration of photon-generated carriers . It can be supposed that a complexation interaction occurred between the adsorbed Cr(III) on the surface of catalysts and oxygen-containing functional groups of TC. In addition, the adsorption of Cr(III) on the surface of catalysts can improve the removal of Cr(VI) based on the disproportionation reaction between the adsorbed Cr(III) and Cr(VI) (Wang et al. 2020). As for FSBTS, Cr(VI) could be removed in the single system due to photo-generated electrons based on efficient separation of carriers. However, it was found that further enhanced removal of Cr(VI) and slightly inhibited removal of TC in the combined pollution systems, synergistic photocatalysis effect and the competition between Cr 2 O 7 2and e − /h + might be responsible for the above phenomenon.
The photocatalytic experiments were carried out for five cycles in the TC/Cr(VI) combined pollution systems to evaluate the stability of the recycled catalysts. After the first catalytic run, the recovered photocatalysts were regenerated and then used for the next photocatalytic experiment under the same conditions. As shown in Fig. 6e, f, the removal efficiency of TC and Cr(VI) by the recycled catalyst exhibited a trend of gradual decrease with the increase in cycle time, and especially, a significant adverse effect on the removal of Cr(VI) was observed. After five photocatalytic cycles, the TC removal rate of FSBTG was reduced from 91.03% to 57.32%, and the removal rate of Cr(VI) was reduced from 78.56% to 16.08%. This may be due to the reaction between the hot alkaline solution and chromium ions in the catalyst regeneration process to form a precipitate that is tightly combined with the catalyst, occupying the active sites of the catalyst, and causing the catalytic activity to decrease with the frequency of recycling. Further searching for suitable biochar-based catalyst regenerators is needed to achieve effective regeneration of the catalyst without significant adverse effects on its activity. Also, the removal of TC and Cr(VI) for FSBTS was greatly influenced by cycle time. Besides the above reason, local aggregation and deactivation of FSBTS during the regeneration process and the loss of photocatalysts during wash/dry process could weaken the removal of pollutants (Huang et al. 2018).

Removal mechanism of photocatalyst for TC and Cr(VI)
The removal of TC and Cr(VI) was achieved via adsorption and photocatalytic redox reactions of photocatalyst, in which huge surface area gave it countless active sites, and the possible mechanism is shown in Fig. 7.
Adsorption played a critical role in the whole removal process, in which both physisorption and chemisorption were involved, with the latter as the dominant via isotherm and kinetic analysis. Physical adsorption was dominated by pore filling effect and electrostatic interaction, based on the appropriate pore size distribution and surface charge of photocatalysts, which contributed to the adsorption effect (Dai et al. 2020;. The hydrogen bonding force and surface complexation were deemed to be the critical mechanisms for TC adsorption due to the presence of oxygen-containing functional groups and metal oxides, which were discovered to enhance TC adsorption in this study (Dai et al. 2020;Li et al. 2022b). The surface complexation interaction between metal chromium ions and oxygen-containing functional groups of catalysts was also considered as the main force for Cr(VI) adsorption Xiao et al. 2019;Zhu et al. 2021). The adsorbed pollutants can be further decomposed by reactive oxygen species, generation of which is significantly dominated by utilization of light and transport efficiency of photogenerated charges (Buzzetti et al. 2019). The enhanced visible light response and separation of photogenerated charges obtained from UV-vis DRS, photocurrent response, and EIS analysis indicated improved optical and electronic activities, which contributed to the photocatalytic reaction of catalysts. The ·OH and ·O 2 − were regarded as the dominant reactive species based on the quenching experiment. The photogenerated e − and h + transfer process of FSBTG-450 can be considered as Z-scheme rather than type II heterojunction structure due to the product of ·O 2 − (Wang et al. 2020). In the as-synthesized magnetic biochar-based composite catalyst, the photogenerated h + in the valence band (VB) of g-C 3 N 4 and the e − in the conduction band (CB) of TiO 2 recombined, and the components of biochar and Fe 3 O 4 @SiO 2 accelerated the transmission of electrons, while lots of electrons and holes survived in the CB of g-C 3 N 4 and the VB of TiO 2 , respectively. The ·O 2 − generated in the CB of g-C 3 N 4 via reaction between e − and O 2 , and ·OH generated in the VB of TiO 2 via reaction between h + and H 2 O were deemed to be responsible for the oxidation of TC (Guo et al. 2021;Kumar et al. 2020). The Cr(VI) also can be removed by the e − generated in the CB of g-C 3 N 4 . Although the presence of competition for e − between the removal of Cr(VI) and the generation of ·O 2 − , the Cr(VI) can be removed effectively due to the introduction of H + derived from TC.
To analyze the degradation pathway of TC, the intermediate products during the photocatalytic process were detected by means of UPLC-MS. The results showed 16 major intermediates during the degradation process. Several photocatalytic degradation pathways of TC over FSBTG-450 are proposed in Fig. 8. There are three types of functional groups, including double bond, phenolic group and amine group, which are vulnerable to attack by radicals due to the high electron density (Wang et al. 2018a). In pathway I, TC was attacked by reactive oxygen species to generate the intermediate with m/z of 427 via dehydroxylation and N-demethylation due to the low energy of N-C bond (Guo et al. 2021;Qin et al. 2021). Subsequently, the intermediates with m/z of 385 and 302 were generated by dehydroxylation, demethylation and benzene ring opening reaction. The two intermediates can be further decomposed by similar reaction mechanisms to produce the intermediates with m/z of 306, 282, and 148. In pathway II, the formation of the intermediate of m/z = 496 was mainly caused by the dehydrogenation of -OH in Carbon 3 position, -N(CH 3 ) oxidation, and carboxylation of -CH in Carbon 8 position after the breakage of double bond in Carbon 7-8 position. The intermediate was further oxidized to produce the intermediate of m/z = 433, which subsequently was decomposed to low-molecular-weight ketone or carboxylic acid compounds with m/z of 256, 253, 196, 151, 119, and 118 by ring cleavage along with hydroxylation. Similarly, in pathway III, the low-molecular intermediates with m/z of 283 and 214 were produced by rings-opening, along with additional decarbonylation and demethylation reactions (Qin et al. 2021). Finally, these low molecular intermediates could be completely decomposed to CO 2 and H 2 O.

Conclusions
The biochar-coupled Fe 3 O 4 @SiO 2 @TiO 2 /g-C 3 N 4 composites have been successfully synthesized, and exhibited extraordinary removal performance for high-concentration TC. The introduction of biochar and magnetic nanoparticles can improve visible light response, charge transfer, and separation of photo-generate carriers. Except for the significant inhibition of PO 4 3− , other anions and pH had no obvious effect on removal performance. The heterojunction enhanced the spatial Fig. 7 The plausible removal mechanism of TC and Cr(VI) by FSBTG separation of h + and e − , and ·OH and ·O 2 − were identified as the main reactive species responsible for TC degradation and Cr(VI) removal. Due to the synergistic photocatalysis effect, the removal efficiency of Cr(VI) was notably enhanced in the combined pollution systems. In addition, removal mechanisms involved in adsorption and Z-scheme photocatalytic reaction and the pathway of TC were proposed. This work can provide a reference for the removal of high-concentration TC/Cr(VI) in combined pollutant systems and the potential application of biochar-based materials in waste treatment.