Novel magnetic graphoxide/biochar composite derived from tea for multiple SAs and QNs antibiotics removal in water

Antibiotics pollution is an urgent public health issue. Biochar is a kind of promising composite for removal antibiotic in aqueous environment. In this study, a novel magnetic graphoxide/biochar composite (mGO/TBC) was synthesized by simple impregnation method and used as an efficient and recyclable persulfate (PS) activator for degradation and removal of sulfonamides (SAs) and quinolones (QNs) antibiotics. Based on the synergism pre-adsorption and degradation between graphoxide and biochar, the removal rates of mGO/TBC on sarafloxacin hydrochloride, sulfadimethoxine, sulfapyridine, sulfadoxine, sulfamonomethoxine, sulfachloropyridazine, enrofloxacin, and ciprofloxacin were increased above 95%. Moreover, the mGO/TBC could be reused at least seven times after degradation-recovery cycles. Quenching experiment and ESR analysis proved that 1O2, •OH, and SO4•− from mGO/TBC/PS system were the primary oxidation active species to degrade SAs and QNs. It is a promising substrate for antibiotic bioremediation with good application prospects.


Introduction
Antibiotics have been widely used in human and veterinary medicines (Soni et al. 2022). According to the report (Singhania et al. 2018), the number of therapeutic veterinary antibiotics products would reach above 100 thousand tons by 2030. Due to the increasing consumption, multiple antibiotics such as sulfonamides (SAs), quinolones (QNs), chloramphenicols, and tetracyclines have been frequently detected in different environmental media. The pollution of antibiotics was reported at concentrations ranging from ng/L to mg/L in surface water . Antibiotics could enter lakes, rivers, oceans, and even groundwater through surface runoff, wastewater discharge, or atmospheric particles deposition (Lapworth et al. 2012;Petrie et al. 2015). The presence of antibiotics in water environment is not only toxic to environmental organisms, but also could be enriched through food chain and cause potential threat to human health (Zhu et al. 2018). Moreover, the residues of antibiotics in environmental water could cause selection pressure on bacteria and the development of antibiotic resistance genes (Jia et al. 2020). Moreover, recent report had still also shown the low removal rates for sulfamethazine (44.6%) and ofloxacin (26.3%) in wastewater treatment plants with common treatment processes (Li et al. 2022). In the long run, the pollution caused by antibiotics will become more and more serious, and pollution control is also urgent.
Up to now, several technologies had been developed for removal of antibiotics, including adsorption, coagulation, membrane-filtration, chemical-oxidation, and biological treatments (Langbehn et al. 2021). Among them, biocharbased biological treatment method had exhibited more attractive antimicrobial and catalysis/adsorption property. Biochar is a stabilized carbon-rich material produced from the pyrolysis of oxygen-limited biomass (Aup-Ngoen and Noipitak 2020). Its high carbon content and porous structure provide abundant active sites for adsorbent antibiotics in water. In addition, the use of biochar as a catalyst to activate persulfate can produce active compounds that can totally Responsible Editor: Tito Roberto Cadaval Jr breakdown contaminants Jiang et al. 2018;Liu et al. 2020a). In comparison to the conventional "Fenton" method, advanced oxidation processes (AOPs) based on persulfate have many benefits including easy transportation and storage and broad pH adaptability (Ahmad et al. 2013;Furman et al. 2010). Persulfate-based AOPs (PS-AOPs) have a higher mineralization capacity and less energy demand than AOPs based on electrochemical and photochemical technologies, which making PS-AOPs a promising option for the treatment of multiple antibiotics pollutants (Yu et al. 2020). However, the adsorption and degradation performance of pristine biochar was limited because the relatively lower specific surface . In recent years, the remediation efficiencies based biochar techniques were highly improved through developing new composites to improve interactions and bonding abilities with antibiotic (Patel et al. 2022). Liang et al. developed a Fe 2 O 3 /biochars composite which had exhibited higher removal efficiency for norfoxacin than original biochar (Liang et al. 2022a). The study has shown a promising strategy for removing antibiotics from water.
The synthetic methods of modified biochar generally include co-pyrolysis (Liang et al. 2022b), chemical precipitation ), hydrothermal (Gholami et al. 2020) and microwave irradiation (Qu et al. 2020). Although modified nanomaterials have excellent adsorption capacity, they are not suitable for large-scale applications due to their huge cost and potential environment risks . Graphene oxide (GO) has demonstrated great affinity for various antibiotics with its high specific surface, large numbers of oxygen-containing groups and wide delocalized π-electron system . Magnetic particle-Fe 3 O 4 is popular in AOPs because of its simultaneous mixed valence states of Fe 2+ and Fe 3+ and the low activation energy required for Fe 2+ to activate PS (Xu et al. 2019). When GO was incorporated with magnetic particle, the magnetic composite not only can increase the solid-liquid separation efficiency but also improve its regenerative capacity by an external magnetic field (Khan et al. 2020). In our previous report, the adsorbent of magnetic GO to twelve QNs antibiotics in water were observed . Recently, Bu et al. found that surface functional groups of GO would reduce after pyrolysis (Bu et al. 2022). When GO and biochar composites was prepared with conventional pyrolyzation methods, the rigid covalent structure of GO cannot rearrange the disordered C atoms during pyrolysis to create sites for dangling bonds and free radicals. It would cause multiple interconnected hexagonal planes with internal dislocations and vacancies (Chia et al. 2015). Therefore, the inhomogeneity caused by these defects limit the adsorption capacity .
In this study, a novel tea-derived magnetic graphoxide/ biochar composite (mGO/TBC) was synthesized by simple impregnation method. The proposed strategy could avoid the loss of oxygen-containing functional groups in GO which caused by secondary pyrolysis process. The mGO/TBC composite had used as a recyclable PS activator degradation and shown excellent degradation performance for multiple SAs and QNs antibiotics. In addition, the modified mGO/TBC composite also has great application prospect for other antibiotics removal and degradation.
The standard stock solution of SAs and QNs was dissolved with acetonitrile and 0.03 mol/L NaOH respectively (Fisher Scientific, USA) and stored at -20 °C. Formic acid and methanol (for HPLC, ≥ 99.9%) were purchased from (Fair Lawn, NJ, USA). Ultrapure water (Millipore, Bedford, OH, USA) was used for the preparation of solutions. The tea was obtained from Xiangfeng Tea Group Co., Ltd.

Synthesis of the mGO/TBC
The biochar was prepared with one-step carbonization activation method. The tea was washed with water and dried at 60 °C. The dried tea was smashed into biomass powder less than 100 mesh. Then, the biomass powder was carbonized at 700 °C for 3 h at a muffle furnace with a heating rate of 5 °C/ min. The tea biochar (TBC) was obtained. Magnetic graphene oxide (mGO) was synthetized based on our pervious method , Supporting information, Text S1). mGO/TBC was synthesized by impregnation method. TBC and mGO were mixed at ratio 1:1 and stirred for 6-8 h at room temperature. The mGO/TBC was synthesized and separated with an external magnetic field. After washing with deionized water, the mGO/TBC was obtained and dried at 60 °C. The characterizations of mGO/TBC were provided in the supplementary data (Text S2).

Adsorption and degradation experiments
Adsorption experiments were performed with 100 mL of mixed standard solution, 20 mg of mGO/TBC under 150 r/ min on a shaker. The concentration of antibiotics was determined with high performance liquid chromatography.
The adsorption isotherms experiment was performed with 1 mg/L to 250 mg/L of mixed standard solutions. The pH of solution was adjusted with 0.1 M hydrochloric acid and sodium hydroxide.
For degradation experiments, 20 mg of mGO/TBC were mixed 100 mL of 10 mg/L of QNs and SAs standard solution which was dissolved in distilled water . First, the mixture was vortexed 30 min for adsorption process. Then, 10 mM PS was added to the reaction system for the degradation process. The concentration of sarafloxacin hydrochloride, sulfadimethoxine, sulfapyridine, sulfado xine, sulfamonomethoxine, sulfachloropyridazine, enrofloxacin, and ciprofloxacin were determined by HPLC. The removal rate (%) was calculated by Eq. (1). C 0 and C t (mg L −1 ) are the initial and equilibrium antibiotic concentrations.

Quantitative determination
High-performance liquid chromatography (HPLC, Water Alliance 2695, USA) was equipped with UV detector (Waters 2998) and fluorescence detector. The detection wavelengths of UV for SPD, SMM, SCP, SDM, and SDX were set at 270 nm. The detection wavelengths of fluorescence for CZP, ENR, and SAR were set at 280 nm. A Waters SunFire HPLC C18 column (250 mm × 4.6 mm, 5 μm) was performed for chromatographic separation. The column temperature was set at 35 °C with 20 μL sample injections volume. The mobile phase was conducted with acetonitrile as the organic phase A and ultrapure water containing 0.2% formic acid as the water phase B with initial flow rate 0.8 mL/ min. The percentage of organic phase B increased to 83% in 3.0 min, decreased to 80% in 10.5 min, 75% in 14.5 min, 40% in 22.5 min, and 83% in 25.0 min.

Characterization of material
The mGO/TBC was synthesized with simple impregnation method with mGO and TBC. Compared with the high temperature pyrolysis method, the proposed strategy was easier to operate and avoid the loss of oxygen-containing functional groups on GO. The micromorphology and element percentage of TBC, Fe 3 O 4 magnetic particle, mGO, and mGO/TBC were characterized by SEM. The porous, lumpy irregular structures was observed on SEM images of TBC, which was caused by the aromatization and carbonization during the biochar formation process (Sun and Li 2004) (Fig. 1a). As shown on Fig. 1b, c, Fe 3 O 4 was distributed on the surface of oxidized graphene sheets. Meanwhile, the modification of TBC had obviously changed the surface morphology of mGO (Fig. 1d). The wrinkling of the surface topographies and collapse of pore structure of TBC could be observed on mGO/TBC. At same time, the pores formed by hightemperature carbonization of biochar could be observed. The element mapping of mGO/TBC clearly illustrated that C, K, O, and Si were distributed on the surface of mGO/TBC ( Fig. 1e-l). K and Mg element were contributed from biochar ( Figure S1a), which indicated the TBC had successfully modified on the surface of oxidized graphene. FTIR spectra was used to further verify the surface functional groups (Fig. 2a). The peaks at 800 cm −1 , 1634 cm −1 , and 3440 cm −1 were belong to the tensile vibrations of C-H, C = C, and O-H, which was observed in spectra of mGO and mGO/TBC (Zhou et al. 2018). A new band at 1419 cm −1 was observed in the spectrum of TBC, which belong to O = C-O group. The O = C-O group was generated by the biochar during pyrolysis due to temperature changes, and the same peak was appeared in spectra of mGO/TBC (Yu et al. 2019). In addition, the extra peaks at 456 cm −1 belong to the stretching vibration of Fe-O ) was observed in the FTIR spectra of mGO and mGO/TBC. Furthermore, due to the inherent aromatic backbone of lignin precursors (Ge and Li 2018), TBC was assembled between magnetic graphene oxide sheets via π-π interaction, as indicated by the band of 1620 cm −1 (C = C stretching) in mGO/ TBC. These aromatic ring groups and functional groups that contain oxygen moved, demonstrating that biochar had been successfully linked to the surface of the mGO. Meanwhile, the intensity of characteristic peaks at 1096 cm −1 (stretching vibration of Si-O-Si) (Tang et al. 2015) decreased significantly, indicating that Si-O-Si modified magnetic particle on the mGO/TBC also participated in the adsorption and degradation process of SAs and QNs (Liu et al. 2020b). Besides, the band intensities of -OH was weakened after the reaction, indicating possible existence of hydrogen bonding due to chemisorption between the hydroxyl group and antibiotics (Tucureanu et al. 2016). The results further proved the successfully modification of TBC with mGO.
The specific surface area, pore volume and pore size distribution of TBC, mTBC, mGO, and mGO/TBC were analyzed by the Brunner-Emmet-Teller model and NLDFT theory by N 2 adsorption-desorption experiment. As shown in Fig. 2b, all proposed materials belong to type IV isothermal line (Liang et al. 2009). In the range of p/p 0 > 0.5, it was accompanied by type H 3 hysteresis loop, which might be caused by the condensation of N 2 in the hole. The pore sizes of four materials were between 2 and 50 nm (Table S1), indicating all materials were mesoporous (Feng et al. 2021). Compared with the initial biochar, the total pore volume (0.159 cm 3 /g) and BET surface area (27.142 m 2 /g) of mGO/TBC had increased threefold after mGO modification. The large surface was beneficial for the distribution of more active sites, which would further increase the removal rate. With their high separabilityand recyclability (Fig. 2c), the mGO/TBC composites dispersed in aqueous solution were a good contender for applications on the pollutant bioremediation.

Adsorption performance of mGO/TBC
First, the adsorption equilibrium and diffusion mechanisms were studied with three classical kinetic models included the pseudo-first-order (Eq. (S1)), pseudo-second-order (Eq. (S2)), and intra-particle diffusion models (Eq. (S3)). The adsorption efficiencies of mGO/TBC on five SAs and three QNs was studied within 0-240 min.
The higher regression coefficients (R 2 , Table S2), the smaller difference between the experimental results and the model values, the pseudo-second-order model gives a better R 2 (between 0.9628 and 0.9849) than the first (between 0.9476 and 0.9804), indicating that the adsorption process of SAs and QNs on mGO/TBC was primarily controlled by chemisorption (Ho et al. 2000). The rate constants (k 2 ) were approximately 2-3 times higher than the previously published values for frequently used biochar (Reguyal and Sarmah 2018;Zhang et al. 2021), which was benefit for the adsorption process.
Represented by the diffusion model of SPD, the adsorption kinetic curves with Weber-Morris model were shown in Fig. 2d. The intraparticle diffusion model showed three linear stages. At first, antibiotics diffused on the surface of mGO/TBC with a fast adsorption rate. Then, antibiotic molecules were diffused into mGO/TBC channels. At the adsorption equilibrium stage, the adsorption rate was prolonged, the antibiotic in the solution was at a low grade and the active sites on the surface of mesoporous carbon were saturated. In addition, the linear relationship between Q t and t 0.5 in the fitting curves did not pass through the origin, indicating that the adsorption rate of mGO/TBC for antibiotics was controlled by intraparticle diffusion and external diffusion.
Adsorption isotherm is an effective method for studying adsorption properties and the microcosmic mechanism between adsorbents and adsorbents. The adsorption process of SAs and QNs on mGO/TBC was described. The results of the Langmuir and Freundlich model fits were shown in Table S3. Compared with the Freundlich model (R 2 between 0.8902 and 0.9774), the Langmuir model (R 2 between 0.9721 and 0.9999) provide relatively higher correlation coefficients, which showed that the monolayer adsorption dominated adsorption on mGO/TBC with uniform distribution of functional groups, and multilayer adsorption also plays a role. Similarly, the Langmuir model was equally applicable in the fitting of adsorption isotherm models for antibiotics adsorbed in graphene oxide (Chen et al. 2015;Rostamian and Behnejad 2016), graphene nanosheet (Rostamian and Behnejad 2016), and porous carbon (Zbair et al. 2018). The maximum adsorption capacity of SPD obtained from the Langmuir model on the mGO/TBC was 351.048 mg/g (Table S3).

Adsorption-degradation performance
The adsorption-degradation performance of mTBC, mGO, and mGO/TBC were studied represented by SDX. The results were shown in Fig. 3a. Once only PS was added in the system, the removal rate of SDX was 59.44% after 60 min. The removal rate was increased to 77.60% when both catalysts and PS were added in the system. After combination preadsorption with PS-AOP, the removal rate was increased to 99.74% in 60 min. The catalyst mGO/TBC significantly improved the removal rate of SDX. The results indicated that the combination of mGO and TBC had high enhanced catalytic performance for SDX. The adsorption of mGO/TBC on SDX had provided a higher driving force for the degradation. The synergistic effect of adsorption and catalysis of mGO/ TBC caused the high removal rate on SDX. The catalytic capacity of four catalysts was mGO/TBC > mTBC > mGO. In addition, the comparison curves of simple adsorption and adsorption-degradation of the eight antibiotics were shown in Fig. 3b, the removal rate of ENR, CZP, SAR, SDM, SPD, SMM, SCP, and SDX were all increased from 36.60 to 99.31% with the adsorption-degradation strategy. The results indicated that mGO/TBC was a promising substrate for multiple antibiotic bioremediations.

Effect of catalyst dosage
First, the effect of the mGO/TBC dosage was shown in Fig. 3c. when the dosage of mGO/TBC was increased from 0.05 to 0.40 g/L, the removal rate of SAs and QNs except SCP was increased from 14.3 to 30.6%. The removal rate of Fig. 3 The removal rate of SDX with different systems (a); the removal rate of eight antibiotic with mGO/TBC under adsorption and adsorption-degradation procedure (b); the effect of catalyst dosage (c), PS dosage (d), pH (e), and reaction temperature (f) on removal rate of SAs and QNs in mGO/TBC/PS system SCP was 98.97% with 0.05 g/L dosage in 60 min. More catalyst could provide more reactive groups for PS activation, which resulting in a higher removal rate. However, limited by the diffusion and aggregation of the catalyst, the removal rate would not increase when the dosage was higher than 0.30 g/L. Therefore, 0.30 g/L mGO/TBC was used for further experiment.

Effect of PS concentration
Persulfate was an essential reagent in PS-AOPs. Potassium persulfate was commonly used for oxidizing antibiotics in the aquatic environment. First, the oxidation property of potassium persulfate was appropriate and the pollutants were less affected by the pH. On the other hand, the decomposition of PS could provide potassium elements, which can improve the activation performance of mGO/TBC . Figure 3d showed the effect of PS dosage on removal rate of SAs and QNs in mGO/TBC/PS system. The removal rates of SMM, SDX and SDM were 56.43-58.56% with 4 mM PS after 60 min. When the dosage of PS was increasing to 10 mM, the removal rate of eight antibodies were increased to 74.36-98.97% within 60 min. The increased removal rate indicated that PS was effectively decomposed and activated, resulting in more active species (Vicente et al. 2011). However, a decrease removal rate was observed for the five sulfonamides when the concentration of PS was further increased, which might due to the free radical quenching caused by the addition of excessive PS. Therefore, 10 mM PS was used for next experiments. Figure 3e shows the degradation of SAs and QNs with mGO/ TBC/PS system at different pH. In general, SO 4

Effect of pH
•− production is faster and more efficient under acidic conditions than alkaline conditions (Akbari et al. 2016).With increase of pH, the removal rate of SDM, SDX, and SMM decreased slightly, while the other five substances could be degraded above 89.40% within 60 min when pH changed from 3 to 9. A wide pH range indicates that mGO/TBC/PS has a high degradation efficiency and could be used in environment water. Therefore, the pH 7 of the mGO/TBC/PS system was selected in a neutral environment that was suitable for most water samples without adjustment.

Effect of reaction temperature
Reaction temperature was also an essential factor in the chemical oxidation process. The reaction temperature at 25 °C, 35 °C, and 45 °C were optimized for the mGO/ TBC/PS system. As shown in Fig. 3f, the removal rate of SAs and QNs was increased with the increase of reaction The degradation ratio of mGO/TBC on ENR (b) and SPD (c) after seven times adsorption-degradation-recovery cycles temperature. The results might be explained by the possibility that higher temperature could provide more energy to reactant molecules, which enhancing reaction rates and facilitating the breakdown of persulfate into more radicals (Olmez-Hanci et al. 2013).

Effect of inorganic anions
Multiple ions are present in aquatic ecosystems that would interfere with the degradation process. Figure 4a showed the influence of 10 mM Cl − , HCO 3 − , SO 4 2− , and H 2 PO 4 − on the removal rate of SAs and QNs with mGO/TBC/PS system. The presence of the four common anions had slight effects on most targets. The removal rates of SDM, SAR, and CZP were decreased mainly by HCO 3 − and H 2 PO 4 − , which decreased the removal rates from 2.42 to 12.03%. That was because HCO 3 − could react rapidly with SO 4 •− and • OH to inhibit the free radical pathway. •− .
For the H 2 PO 4 − , the effect of removal rates could combine with Fe-OH on the catalyst's surface to form innersphere complexes.
In summary, the mGO/TBC/PS system could avoid the interaction of common anions on the removal rate for most antibiotics. The proposed system could meet a wider range of application conditions.

Application on environmental water samples
Under the optimized conditions, the practical application performance of proposed mGO/TBC/PS system was validated by environmental water samples. Three environmental water samples were collected from the upper reaches, middle reaches and lower reaches of HaiHe River were collected. The samples were treated as the procedure 2.3. The results were shown in Table 1. The SAs and QNs were not detected in the three water samples. The removal rates were 82.16% and 100.00% when the concentration of SAs and QNs in spiked samples were 10 mg/L. As a complex multicomponent sample, river water had a negative effect on the degradation of SMM in the middle reaches (adsorption capacity decreases 16.20% at 10 mg/L level) and slightly effects on SPD in the lower reaches (adsorption capacity decreases 13.90% at 10 mg/L level). Overall, mGO/TBC maintained high removal performance for SAs and QNs in practical water samples, which had indicated that mGO/TBC was applicable to the complex matrix components with multiple antibiotics. The proposed strategy had broad potential for wide application in water treatment.

Reusability of mGO/TBC
To evaluate the reusability of mGO/TBC, the mangetic field of phase structures of mGO/TBC before and after application were firstly investigated. As shown in Figure S2, no significant peak shift was observed except the intensity changed, indicating mGO/TBC could maintain the structural integrity after adsorption of the eight antibiotics in water. Actually, mGO/TBC had excellent water stability (Akiyama et al. 2011). The reusability of mGO/TBC was studied represented by SPD and ENR. As shown in Fig. 4b and c, the degradation ratio of mGO/ TBC/PS system on SPD and ENR were still 98.00% and 79.64% after seven adsorption-degradation-recovery cycles, which indicating the excellent regeneration property of mGO/TBC after antibiotics removal. The proposed strategy was further compared with recent studies of similar type. As shown in Table S5, the proposed method has shown high removal efficiency and reusability for multiple antibiotics.

Removal Mechanism of mGO/TBC on antibiotics
At first, the hydrophobic interaction, hydrogen bonding and π-π conjugation would induce rapid adsorption between mGO/TBC and antibiotics included SAs and QNs. The process was verified by elution experiments. As shown on  of SAs and QNs could be eluted after initial adsorption process. After the total removal experiment, there were not SAs and QN could be detected, which indicated the SAs and QNs had been degraded.
Methanol (MeOH) had been proved in numerous experiments to efficiently quench both • OH and SO 4 •− . Tert-butanol (TBA) was a good quencher for • OH produced in solution or surface boundaries. Additionally, furfuryl alcohol (FFA) could detect whether 1 O 2 exists (Ho et al. 2019). Therefore, quenching experiments were carried out with MeOH, TBA, and FFA as scavengers which were used to further identify the role of reactive species to the degradation of SAs and QNs. As shown in Table S4, the removal rate of eight SAs and QNs were ranged from 95.40 to 99.48% without bursting agents. The removal rate of SAs and QNs decreased to varying degrees with the addition of MeOH and TBA (5 mL), which indicating the co-existed SO 4 •− and • OH in the catalytic degradation system. Compared with MeOH, the inhibitory effect of TBA on the removal rate of antibiotics was more significant. The removal rate of SMM decreased to 41.29% with the addition of TBA. Meanwhile, FFA was not also dramatically able to quench the reactions, and the removal rate decreased 22.10-10.43%. The result suggested that • OH played a vital role, SO 4 •− and 1 O 2 play a relatively minor role in the removal of SAs and QNs in mGO/TBC/PS system.
The types of reactive oxygen species formed in the mGO/ TBC/PS system was further confirmed by ESR spectra with DMPO and TEMP as trapping agents. No signs of free radicals or 1 O 2 were observed when PS was present (Fig. 5a, b). However, varying degrees of DMPO-• OH, DMPO-SO 4 •− signals (peak strength of 1:2:2:1) and TEMP-1 O 2 (peak strength of 1:1:1) were observed in all four activator-derived combined preadsorption and PS-AOPs   6 Material synthesis and mechanism scheme of QNs and SAs degradation in mGO/TBC/PS system mGO/PS systems, the mGO/TBC/PS system had stronger signal response. It indicating that mGO/TBC had a more effective effect on activating PS than TBC, mTBC, and mGO.
The XPS spectra of mGO/TBC and the used mGO/TBC were studied. The activation mechanism of PS activation was investigated. As shown in Fig. 5c and d, the C (1 s) spectra could be deconvoluted into four peaks assigned to C = C, C-C (284.  . It could be found that the percentage of -C = C/C-C and C-O increased greatly and COOH appeared as new species. This phenomenon showed that oxygen-containing functional groups (OC and OCO) in mGO/ TBC may be one of the reasons affecting the activation performance of PS (Dai et al. 2019). According to FT-IR results (Fig. 2a), the peak value of oxygen-containing functional groups also changed, consistent with the results. Furthermore, XPS could not be detected due to the low iron content, but Fe (II) in mGO/TBC was a key component that activates PS to degrade SAs and QNs .
Based on the above analysis, the possible generation mechanism of reactive free radicals in the mGO/TBC/PS system were proposed and illustrated in Fig. 6. As shown in Eq. (9), nFe 3 O 4 could directly interact with S 2 O 8 2− by donating electrons from Fe (II) to generate SO 4 •− , and SO 4 •− continued to interact with water to produce • OH (Eq. 10). S 2 O 8 2− can also act with water to produce HSO 5 − , which in turn produces 1 O 2 (Eqs. 11-13), which is related to the promotion of C = O (Sun et al. 2017). These results were consistent with the quenching test and ESR test results (Chen et al. 2020;Fu et al. 2019).

Conclusions
In summary, a highly active biochar-based magnetic graphene oxide composite named mGO/TBC was successfully synthesized by simple impregnation method. The mGO/ TBC had shown high degradation performance on multiple SAs and QNs antibiotics based on the synergistic adsorption and persulfate degradation. According to the fitting results, mGO/TBC fitted the pseudo-second-order kinetic model and the maximum adsorption capacity of SPD was 351.048 mg/g. The SAs and QNs could be swiftly eliminated throughout a broad pH range from 3.0-9.0 due to adsorption and oxidation of radicals (SO4 •− and • OH) and non-radicals including 1 O 2 . Meanwhile, mGO/TBC was renewable and could be recycled up to seven times. It had good resistance to various co-existing anion interferences and good adaptability for water samples.

Data availability
The datasets used or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations
Ethics approval and consent to participate Not applicable.

Consent for publication Yes.
Competing interests The authors declare that they have no competing interests.