Antibiotics have been widely used in medical and health fields, especially recently in the context of the new coronavirus (COVID-19) pandemic [1-3]. While their widespread and excessive discharge to water bodies have caused serious threat to aquatic environments and human beings [4-6]. For instance, the persistence of antibiotics in aquatic environments are toxic to organisms, and it could also produce resistant pathogenic bacteria which is harmful to ecosystems and people [7,8]. Additionally, antibiotics are often detected in natural water bodies, and it ranged from ng/L to mg/L [9,10]. Therefore, understanding the transformation of antibiotics in water bodies is of great significance to effectively prevent their environmental risks.

Meanwhile, sulfides including H₂S, HS⁻ and S²⁻ also widely exist in aquatic environments, which are formed through anaerobic reduction of sulfate and other organic sulfur species by sulfate-reducing bacteria [11,12]. The concentration of sulfide in natural water bodies, such as sediments and porewater, is as high as hundreds of µmol/L to several mmol/L [11,13]. Sulfide is often considered as reductant for the reduction of organic pollutants in anaerobic aquatic environments [14,15]. It was recently discovered that active oxygen such as hydroxyl radicals (^•OH) was produced during oxidation of sulfide under aerobic conditions [16], and the cumulated ^•OH had the potential to further degrade the organic pollutants in aquatic environments. Thus, it is meaningful to use widespread presented sulfide to treat organic pollutants such as antibiotics in water environment. Additionally, latest research demonstrated that residual pyrogenic carbon in water bodies could facilitate the degradation of pollutants by sulfide oxidation under dark oxic conditions, which mediating ^•OH generation [17]. However, the reaction rate for pollutant aerobic transformation by sulfide is still limited, and pyrogenic carbon often suffer from the defects of lacking reaction active sites and reduction of catalytic activity in long-time usage [18]. Consequently, the development of improved catalysts with adequate activity and stability is still a goal to realize.

Active centers loading is a good method to improve the catalytic activity of carbon materials [19,20]. Transition metal-based materials (such as Co, Fe, Cu and Ni) often showed fine catalytic activity [21-23], especially high-activity Co-based materials widely used in various environmental redox reactions [24]. And active transition metal centers supported on carbon substrates could improve the catalytic performance and structural stability of material, which have emerged in advanced oxidation process during past few years [25,26], and it may also exhibit good catalytic performance to promote the organic pollutants aerobic transformation by sulfide. However, previous work had ignored to investigate the catalytic performance of metal supported carbon material on the transformation of organics by sulfide under aerobic conditions, and related mechanism especially the role of metal species is still unknown. Therefore, it is anticipated to develop a transition metal-supported carbon materials with good catalytic activity and stability for the aerobic transformation of organic pollutants such as antibiotics driven by sulfide, and figure out the reaction mechanisms for pollutants transformation.

The present study fabricated Co-loaded carbon nanotubes (CNTs) by a facile one-pot pyrolytic strategy, prepared Co-loaded CNTs could realize the efficient aerobic transformation of antibiotics by sulfide. Ciprofloxacin (CIP), one of the commonly used fluoroquinolone antibiotic and often detected in natural water [10,27,28], was chosen as a model antibiotic. First, related characterizations were used to clarify the structure of the prepared materials. Afterward, the prepared materials were used to elaborate their catalytic performance on CIP aerobic transformation by sulfide, total organic carbon (TOC) removal efficiency and transformation pathway of the target pollutant were also evaluated. Moreover, the formed sulfur species were identified and their effects on the CIP transformation were also estimated. Furthermore, electron paramagnetic resonance (EPR) and quenching experiments were conducted to figure out the active species that probably play critical roles in CIP aerobic transformation process and illustrate the reaction mechanism. Finally, the performance for decontamination applications, including organic pollutants acute toxicity control and stability of materials, were also investigated. This study may provide a reference for the application of Co-loaded CNTs on organic pollutants aerobic transformation driven by sulfide.

CIP, methanol, 5,5-dimethyl-1-pyrroline-N-oxide (DMPO, 99.99%) and humic acid (HA) were purchased from Sigma-Aldrich (USA); methyl phenyl sulfoxide (PMSO), Na₂S·9H₂O, tertiary butanol (TBA), fulvic acid (FA), 2,2,6,6-tetramethylpiperidinooxy (TEMP, 99.99%), isopropanol, furfuryl alcohol (FFA), and N,N-diethyl-p-phenylendiamine (DPD) were obtained from Aladdin Industrial Corporation (China); Na₂S₄ (polysulfide), CoCl₂·6H₂O, elemental sulfur, n-hexane, Na₂SO₃, Na₂SO₄ and Na₂S₂O₃ were obtained from Chinese National Medicines Corporation Ltd. (China).

Multiwalled CNTs (20–30 nm diameter, 10–20 µm length and higher than 95% purity) were purchased from XFNANO Materials Technology Co. (China), which was washed with nitric acid (1 g CNTs in 200 mL nitric acid) at 80 ℃ for 24 h before use to remove the metal species [14,29]. Co-loaded CNTs was prepared by a convenient one-pot pyrolytic strategy. Firstly, 4 or 20 mg CoCl₂·6H₂O (contain 1 or 5 mg Co respectively) were dissolved in 5 mL deionized water, and then CoCl₂·6H₂O water solution was added dropwise into the 100 mg CNTs until the formation of a slurry mixture, and the mixture was naturally dried at 30 ℃. Afterward, the dried precursor was heated at a rate of 10 ℃/min to 600 ℃, and kept at this temperature for 2 h under N₂ flow (100 cc/min), and then cooled naturally to 30 ℃ to obtain the Co-loaded CNTs. The obtained Co-loaded CNTs were named as Co₁-CNTs and Co₂-CNTs with 1 and 5 mg Co, respectively.

The crystalline phase of the materials was evaluated by an X-ray diffractometer (XRD, Panalytical X’Pert Powder, Panalytical B.V., Netherlands) with a Cu Kα radiation source, and the 2θ range from 10° to 80° with a scan rate of 2°/min. The morphological characterization and elemental mappings (including C, Co and O) for the prepared catalysts were conducted on a transmission electron microscope (TEM, JEM-ARM 2100F, JEOL, Japan). The surface atomic percent of C, Co and O elemental composition and functional groups of the materials were analyzed by X-ray photoelectron spectroscopy (XPS, ESCALAB 250, Thermo-VG Scientific Inc., USA), and it measured in a 1.4 × 10⁻⁷ Pa vacuum at 25 ℃. The curve fitting was performed using a Gaussian-Lorentzian peak shape after subtracting a Shirley background. Fourier transform infrared spectroscopy (FTIR, Nicolet iS50, Thermo Fisher, USA) was also performed to analyze the surface functional groups of prepared materials, in which the samples were scanned in the mid-infrared region from 4000 cm⁻¹ to 500 cm⁻¹ with a resolution of 2 cm⁻¹.

Electrochemical impedance spectroscopy (EIS) measurement was conducted by an electrochemical workstation (CHI660D, Chenhua, China) in 100 mL glass chamber with a three-electrode system. The platinum wire and Ag/AgCl electrode were applied as the counter and reference electrodes, respectively. The working electrode was prepared through depositing a mixture of 10 mg catalyst, 0.2 mL distilled water, 30 µL Nafion, and 0.2 mL ethanol onto a glassy carbon electrode. EIS tests were performed at 25 ℃, with init E = 1.1 V, amplitude = 0.005 V, quiet times = 2 s, and frequencies ranging from 10⁵ Hz to 10⁻² Hz.

CIP aerobic transformation experiments were conducted in 200 mL serum bottles containing 100 mL phosphate buffer solution (25 mmol/L) by sealing using air-permeable cellophane, which can prevent water from evaporating, and initial CIP concentration was set as 10 mg/L. A certain dosage of catalyst and natural organic matters (NOMs, if added) was added to the water solution and then ultrasonically dispersed (DY-10–240DT ultrasonic apparatus, Chongqing Dongyue Instrument Co., China) for 30 min. Then, specific amount of Na₂S·9H₂O was added to the reaction systems. After being mixed with Na₂S·9H₂O, the solution pH was adjusted to specific value using less than 0.5 mL of 4 mmol/L NaOH or HCl solution. The CIP transformation experiments were conducted by using an orbital shaker with 200 rpm at 25 ℃ under dark aerobic conditions. Dissolved oxygen concentration was measured by dissolved oxygen meter (FiveGo-F4, Mettler-Toledo, Switzerland). Each experiment was performed triplicate.

Adsorption experiments were conducted as previously described except without addition of any sulfur specie. The experiments for the influence of various sulfur species were also conducted as previously described except without addition of sulfide, while specific amounts of other sulfur species (1 mmol/L sulfur atom) were added to explore their influence on CIP aerobic transformation. Cycling experiments were performed to estimate the stability of the catalyst, and the used catalyst was centrifuged for 15 min and washed with deionized water for the next run. Quenching experiments were conducted to identify the active species in the process of CIP transformation, in which 300 mmol/L of isopropanol, TBA (both scavenger for ^•OH) or methanol (scavenger for SO₄^•−) was used as scavengers for different kinds of radicals, 300 mmol/L FFA was adopted as scavenger for ¹O₂, and 300 mmol/L PMSO was used as scavenger for high-valent Co species [25,30,31].

During CIP transformation process, 1 mL of liquid samples were taken at appropriate time intervals and immediately mixed with 1 mL methanol to stop the reaction (CIP transformation hardly occur in solution of 1:1 water and methanol), and then filtered through 0.22 µm membrane (Jinten Co., China). The CIP concentration was measured by high-performance liquid chromatography (HPLC, Agilent Infinity 1260, Agilent Technologies, USA), it equipped with a diode array detector and an Eclipse Plus C18 column (4.6 × 250 mm). The column temperature was set as 30 ℃, the composition of the mobile phase was 0.3% formic acid water solution/methanol (75:25, v/v) mixture at flow rate of 1 mL/min, the detection wavelength was 278 nm [32], and the sample injection volume was 50 µL.

The CIP transformation rate constant was calculated according to the pseudofirst-order kinetic model (Eq. 1)

(1)

where C₀ is the initial CIP concentration, C_t is the CIP concentration at a certain reaction time (t), and k is the reaction rate constant.

Transformation products of CIP were analyzed by liquid chromatography-mass spectrometry (LC-MS, Q Exactive Plus, Thermo Fisher, USA). The HPLC was operated prior to the MS analysis, and the electrospray ionization needle voltage was 3.5 kV. TOC variation was determined by a Multi N/C 2100 TOC analyzer (Analytik Jena, Germany) to estimate the mineralization levels of organics in the reaction process. And the biological toxicity of samples in the reaction process were monitored by luminescent bacteria toxicity detector (Delta Tox II, American Modern Water Company, USA), and all toxicity assessment experiments were performed at least triplicate.

The sulfide concentration was measured using the methylene blue method [33]. Elemental sulfur of the samples was first extracted by n-hexane (2 mL of sample mixed with 1 mL hexane, and volatile extraction for 10 min) and then monitored by HPLC (Agilent Infinity 1260, Agilent Technologies, USA). It also equipped a diode array detector and an Eclipse Plus C18 column, the mobile phase was 85% methanol and 15% water at flow rate of 1 mL/min, the detection wavelength was 240 nm and the sample injection volume was 50 µL [14]. The concentrations of sulfite, thiosulfate and sulfate were analyzed with an IC-5000 ion chromatograph (Thermo Fisher, USA) with an AS14A column and an electrochemical conductivity detector. H₂O₂ was analyzed by a modified DPD method at 551 nm by Cary 60 UV–vis spectrophotometer (Agilent Technologies, USA) according to previous study [17].

EPR was used to identify the generation of active species including ¹O₂ and radicals. Specifically, active species capture experiments were conducted in a mixed solution of 1 mL reaction sample and 100 mmol/L spin trapping agent TEMP (singlet oxygen trapping agent) or DMPO (radicals trapping agent). The EPR spectra were obtained by using EPR200-Plus spectrometer (Chinainstru & Quantumtech, China), and the analysis conditions were: modulation frequency of 100 kHz, resonance frequency of 9.85 GHz, modulation amplitude of 1.0 G, sweep width of 200 G, time constant of 40.96 ms, and receiver gain of 1.0 × 10³.

High resolution TEM images were recorded to investigate the morphology of CNTs (Fig. 1a) and Co-loaded CNTs (Figs. 1b and c). Images of Co₁-CNTs and Co₂-CNTs exhibited that obvious Co nanoparticles attached to the surface of carbon matrix, compared with nearly no metal nanoparticle on the surface of controlled CNTs sample, which indicated that the Co species were successfully loaded into Co₁-CNTs and Co₂-CNTs samples. Moreover, Figs. 1d and e showed the TEM elemental mapping images of the composite, C, Co and O for Co₁-CNTs and Co₂-CNTs, respectively. These results further demonstrated that Co species was successfully loaded into the surface of carbon matrix. Furthermore, the Co nanoparticles in Co₂-CNTs were commonly larger than the ones in Co₁-CNTs, which may be owing to higher dosage of Co species in Co₂-CNTs. And the ratio of Co atom content for Co₁-CNTs and Co₂-CNTs were about 0.46% and 2.71%, respectively.

XRD patterns were employed to evaluate the dispersion status of Co species in Co-loaded CNTs. As shown in Fig. S1 (Supporting information), compared with nearly absence of Co species in CNTs, the Co phase in the Co₁-CNTs consisted of (220) Co₃O₄ and (111) CoO peaks. And the increase of (220) Co₃O₄ and (111) CoO diffraction line in Co₂-CNTs may indicate the presence of more Co₃O₄ and CoO oxides than in Co₁-CNTs sample, and the emerge of (200) Co⁰ peak showed the presence of Co⁰ in Co₂-CNTs [34,35].

Additionally, XPS was used to elucidate the near-surface chemical environment of prepared Co-loaded CNTs. As shown in Fig. S2 (Supporting information), the emergence of XPS peaks for Co2p revealed the successful introducing of Co species to both Co₁-CNTs and Co₂-CNTs. For Co₁-CNTs, the peaks at 781.8 and 797.7 eV were ascribed to Co 2p_3/2 and Co 2p_1/2 of the Co⁺², and the binding energies of 779.8 and 795.9 eV, corresponding to Co 2p_3/2 and Co 2p_1/2 of Co⁺³, which further proved the valence state of Co species in Co₁-CNTs were Co⁺² and Co⁺³. While for the Co₂-CNTs, excepted for the presence of Co⁺² and Co⁺³, the peaks located at 778.2 and 793.5 eV were ascribed to Co 2p_3/2 and Co 2p_1/2 of the Co⁰, which further confirmed the presence of Co⁰, Co⁺² and Co⁺³ in Co₂-CNTs [25,36].

Moreover, FTIR was used to study the functional groups of materials. As shown in Fig. S3 (Supporting information), the functional groups on the surface of CNTs and Co-loaded CNTs including: hydroxy groups (3400 cm⁻¹), alkyl groups (2900 cm⁻¹), carbonyl or quinone groups (1600 cm⁻¹) and carboxyl groups (1350 cm⁻¹) [37,38]. And it also indicated that Co species loading did not significantly change the functional groups on the surface of the CNTs.

Adsorption experiments in Fig. S4 (Supporting information) revealed that less than 10% of CIP removal after a long-time mixing with 100 mg/L catalysts (no sulfide addition), demonstrating the ignorable adsorption of CIP to catalysts. During the process of aerobic CIP transformation, the dissolved oxygen concentration ranged from 7.4 mg/L to 8.3 mg/L, and did not change significantly during process of the reaction. Fig. 2a displayed the effect of various materials on the aerobic CIP transformation by sulfide. The results showed that 100 mg/L CNTs addition could facilitate the aerobic transformation of CIP by sulfide compared with control one without materials addition with 4 h removal efficiency of CIP increased from 6.6% ± 0.7% to 52.7% ± 4.3%, and the 4 h removal efficiency of CIP further improved to 80.5% ± 6.3% and 94.1% ± 5.5% with presence of Co₁-CNTs and Co₂-CNTs, respectively. Moreover, pseudo-first order dynamic model could well fit the CIP transformation process, the reaction rate constant k increased from 0.013 h⁻¹ to 0.23 h⁻¹ with 100 mg/L CNTs addition, and k further improved to 0.44 and 0.71 h⁻¹ with 100 mg/L Co₁-CNTs and Co₂-CNTs addition (Fig. 2b), respectively. Above results showed that Co species loading could improve the catalytic activity of CNTs for the aerobic CIP transformation by sulfide, and the improvement of Co dosage further enhanced the catalytic performance of Co-loaded CNTs. This might be due to that the improvement of Co loading could provide more reactive sites.

Additionally, as shown in Fig. 2c, the TOC removal efficiency of CIP after 20 h of reaction improved from 0.35% to 23.59% with 100 mg/L CNTs addition, while further increased to 40.72% and 52.12% with 100 mg/L Co₁-CNTs and Co₂-CNTs, respectively. The above results demonstrated that the Co-loaded CNTs/sulfide system not only boosted the transformation of organic pollutants but also further facilitated the mineralization process.

Moreover, the influence of catalyst dosages on the aerobic transformation of CIP was also studied, and the high catalytic performance Co₂-CNTs was chosen as a representative catalyst. As shown in Fig. 2d, the reaction rate constant k improved from 0.0098 h⁻¹ to 0.80 h⁻¹ with the dosages of Co₂-CNTs increased from 0 to 200 mg/L. It might be due to that more active sites or substance at higher dosages of Co₂-CNTs concentration were involved in the transformation of organics by sulfide.

In order to understand the pathway of CIP aerobic transformation with presence of sulfide and various catalysts, LC-MS was used to clarify the intermediate products of CIP. 14 kinds of main intermediate products were detected as shown in Fig. S5 (Supporting information), and different kinds of catalysts addition not change types of CIP intermediates. Based on above results, plausible transformation pathways of CIP were proposed in Fig. 3, including defluorination, oxidation, hydroxylation, decarboxylation, quinoline ring opening, cleavage of C—N and C—C bonds (piperazine ring). Initially, the fluorine and quinolone substituent of CIP could be directly hydroxylated with formation of P1 and P2, respectively. Additionally, the piperazine ring can be oxidized with formation of carbonyl groups (P3). Alternatively, C—C bond of piperazine ring could also be attacked with formation of P4. Afterward, the different sequences for cleavage of C—N bonds or fluorine substituent in P4 leading to 3 reaction paths. The preferentially broken of different C—N bonds and then cleavage of fluorine substituents leading to the 2 reaction paths (P5 → P8 → P12; P7 → P10 → P11 → P12); while the preferentially cleavage of fluorine substituent and then broken of several C—N bonds resulted in the pathway of: P6 → P9 → P11 → P12. Subsequently, the cleavage of C—N bonds and completed removal of piperazine ring for P12 with formation of P13. Eventually, the opening of quinoline ring in P13 with P14 formed, and P14 might further oxidize and even mineralize to other small molecule substances, CO₂ and H₂O.

Previous studies implied that intermediate sulfur species showed significant influence on the transformation of organics [14,39,40], therefore the formed sulfur species and transformation pathway of sulfide during CIP aerobic transformation process were explored. The variation trend of various sulfur species including sulfide, thiosulfate, elemental sulfur, sulfite and sulfate are shown in Fig. S6 (Supporting information). Fig. S6a exhibited that 100 mg/L different kinds of catalysts enhanced sulfide transformation compared with the control one, and Co₂-CNTs further accelerate the transformation of sulfide compared with CNTs or Co₁-CNTs. Furthermore, for the control system without catalyst addition, sulfide initially oxidized to elemental sulfur and further transformed to thiosulfate (Figs. S6b and c), but was hard to go on to sulfite and sulfate, with concentration of sulfite and sulfate lower than 0.1 mmol/L (Figs. S6d and e). While with presence of different kinds of catalysts, small amounts of sulfite and elemental sulfur formed in the early stage of reaction and then gradually transformed to other sulfur species; but high levels of thiosulfate and sulfate formed during the process of aerobic CIP transformation. Above results indicated that catalysts, especially Co-loaded CNTs, showed higher electron accepting ability to promote the oxidation of sulfide and other intermediate sulfur species, and this might due to the good electron transfer ability for these materials.

To further understand the influence of different kinds of sulfur species on CIP aerobic transformation, the reactions between various sulfur species (including polysulfide, thiosulfate, sulfite and sulfate) and CIP with presence of Co₂-CNTs were also explored. As shown in Fig. S7 (Supporting information), aerobic CIP transformation could hardly occur in presence of various sulfur species (1 mmol/L sulfur atom) and catalyst. Consequently, sulfide oxidation process would play key roles for the aerobic CIP transformation, while transformation of other intermediate sulfur species did not directly promote CIP transformation. Moreover, recent studies also displayed that the process of sulfide oxidation could promote the formation active species [16,17], and thus the formed active species might further promote the transformation of organics. Above results implied that Co-loaded CNTs might facilitate the transformation of sulfide and thus enhance the aerobic CIP transformation.

EPR spectroscopy was employed to analyze the active species formed in the process of aerobic CIP transformation by sulfide. As shown in Fig. 4a, no obvious EPR signals appeared with trapping agent of ¹O₂ (TEMP) in the control system. While significant EPR signals of TEMPO (1:1:1 triplet peak, TEMP oxidized by ¹O₂) appeared in the system with presence of different kinds of catalysts. Furthermore, the TEMPO intensity in the Co₂-CNTs system was about triple and double times higher than that in the CNTs and Co₁-CNTs systems, indicating the highest ¹O₂ production in the Co₂-NG/sulfide system. However, no obvious EPR signals appeared with trapping agent of radicals (DMPO) no matter with or without presence of catalysts (Fig. 4b). Above results demonstrated that CNTs and Co-loaded CNTs addition could facilitate the formation of ¹O₂ but not radicals during the process of CIP aerobic transformation by sulfide, and Co-loaded CNTs, especially Co₂-CNTs, further enhance the formation of ¹O₂.

To further figure out which active species playing a key role in the process of CIP aerobic transformation by sulfide, the active species quenching experiments were performed. As shown in Fig. 4c, CIP removal efficiency Co₂-NG/sulfide system was 94.34% without addition of scavenger, and slightly reduced to 87.59%, 87.39%, 83.59% and 86.58% in the presence of 300 mmol/L TBA, isopropanol, methanol and PMSO, respectively. However, the addition of 300 mmol/L FFA, a unique scavenger for ¹O₂, significantly reduced the CIP removal ratio to 18.37%. And the reaction rate constant k decreased from 0.71 h⁻¹ without addition of scavenger to 0.51, 0.52, 0.44, 0.50 and 0.06 h⁻¹ in the presence of 300 mmol/L TBA, isopropanol, methanol, PMSO and FFA, respectively (Fig. S8 in Supporting information). Above results indicated that ¹O₂ would be the dominating active specie that play a critical role in the process of CIP aerobic transformation by sulfide with presence of Co₂-CNTs, while free radicals and high-valent Co species may not play dominant role in CIP transformation process.

Previous studies disclosed that oxidation of sulfide could provide electron for the transformation of O₂ to H₂O₂, and the formed H₂O₂ further transformed to active oxygen [16,17]. Thus, the intermediate H₂O₂ formed in the process of aerobic CIP transformation by sulfide with and without presence of Co-loaded CNTs were determined. As shown in Fig. S9 (Supporting information), different kinds of catalysts addition promoted the formation of H₂O₂, and formed H₂O₂ ranged from 50 µmol/L to 100 µmol/L in the various catalysts addition systems compared with about 20 µmol/L for the control one. Furthermore, higher concentration of H₂O₂ formed in the Co-loaded CNTs systems compared to CNTs one, which may be also due to that Co loading could provide more reactive sites for the H₂O₂ formation.

XPS spectrograms for Co (Fig. S10 in Supporting information) used to study the near-surface chemical environment of Co-loaded CNTs after 4 cycles of reaction, results showed that the chemical valence state of Co species not significantly changed after the reaction, indicating that Co species in Co-loaded CNTs may mainly acted as the electron transfer mediator to boost the active oxygen formation and CIP degradation. Moreover, EIS data demonstrated that Co-loaded CNTs had a stronger current response and lower charge transfer impedance than CNTs (Fig. 4d), and Co₂-CNTs showed the lowest charge transfer impedance, suggesting faster electron transfer ability of Co-loaded CNTs (especially Co₂-CNTs) [30,32]. This might be due to the better electron transfer capability for the Co species than CNTs substrate, and higher Co loading and presence of Co⁰ in Co₂-CNTs giving it better electron conductivity [41]. Additionally, related studies demonstrated that higher electron transfer ability of catalysts could activate the peroxides with the formation of ¹O₂ [25,42]. Thus, above results implied that higher electron conductivity Co active sites in Co-loaded CNTs might act as electron mediator to be responsible for the transformation of intermediate H₂O₂ to ¹O₂, consequently speeding up the aerobic transformation and even mineralization of CIP.

To determine the acute toxicity variation of CIP and sulfide-containing water systems in the reaction process, the inhibition rate of luminescent bacteria was estimated, in which a higher inhibition rate means higher toxicity of samples [30]. As shown in Fig. 5a, the inhibition rate for the 30 µmol/L CIP and 5 mmol/L sulfide water sample was 98.67%, nearly not changed (95.35%) after 20 h reaction without materials addition, and reduced to 72.17% in the CNTs system; however, it further decreased to 45.15% and 30.42% for the Co₁-CNTs and Co₂-CNTs systems, respectively. This may be due to that Co-loaded CNTs could promote the formation of ¹O₂ and further facilitate the transformation of CIP and their intermediate products, which leading to the reduction of acute toxicity from CIP and their reaction intermediate products.

Additionally, Co₂-CNTs was also chosen as a representative catalyst to study the impact of various factors on CIP aerobic transformation by sulfide. The effect of pH on the CIP aerobic transformation was also investigated (Fig. 5b), and the pH variation were less than 0.1 during the process of aerobic CIP transformation owing to the pH regulation capacity by phosphate buffer. The reaction rate k improved from 0.098 h⁻¹ to 0.78 h⁻¹ as pH increased from 4.0 to 6.0, while reduced to 0.061 h⁻¹ with pH further increased to 10.0. Above results indicated the optimal pH value for the reaction was 6.0, and it was beneficial to aerobic CIP transformation by sulfide in the pH range of 6.0 to 7.0. Thus, the neutral conditions in real natural water bodies are favorable for CIP aerobic transformation by sulfide.

Furthermore, the influence of sulfide concentration on the aerobic transformation of CIP was also studied (Fig. S11 in Supporting information). The reaction rate constant k improved from 0.15 h⁻¹ to 0.69 h⁻¹ when the sulfide concentration increased from 1 mmol/L to 5 mmol/L, while it was nearly invariable with further increasing sulfide concentration to 20 mmol/L. Results implied that 5 mmol/L sulfide was enough for the aerobic transformation of 10 mg/L CIP, and sulfide variation trend data also demonstrated that less than 5 mmol/L of sulfide consumed in the CIP aerobic transformation process (Fig. S5a).

Moreover, the effect of widely presented NOMs (including HA and FA) on CIP aerobic transformation by sulfide with Co₂-CNTs was also investigated. As shown in Fig. 5c, the reaction rate k slightly decreased from 0.71 h⁻¹ without NOMs to 0.61 and 0.52 h⁻¹ with 10 and 50 mg/L HA, respectively. And k decreased from 0.71 h⁻¹ without NOMs to 0.46 and 0.23 h⁻¹ with 10 and 50 mg/L FA, respectively. Above results indicated the presence of different kinds of NOMs could hinder the aerobic CIP transformation by sulfide with Co₂-CNTs to a certain extent. On the one hand, this might be due to the presence of NOMs blocking the reaction active sites of Co₂-CNTs [25,43]. On the other hand, NOMs may compete for the formed reactive species in the process of organics transformation [44,45].

The cycling stability of Co₂-CNTs in aerobic CIP transformation by sulfide was also studied. The CIP removal efficiency remained at 94%, and moreover, the reaction rate k slightly decreased from 0.71 h⁻¹ to 0.60 h⁻¹ after 4 cycles of reaction (Fig. 5d), demonstrating the outstanding stability of the prepared Co₂-CNTs. TEM results implied 4 cycles of reaction did not significantly change the morphology of Co-loaded CNTs (Fig. S12 in Supporting information), and XPS results showed that the chemical valence state of Co species also not significantly changed after reaction (Fig. S10). Above results further proved the fine structure stability of Co-loaded CNTs after cycling reaction. Furthermore, only 0.027 and 0.069 mg/L of leached Co species for Co₁-CNTs and Co₂-CNTs, respectively. Both were far below the permissible limit concentration (1 mg/L) according to the Chinese National Standard (GB 25,467–2010) [46], and the leached solution of Co₁-CNTs and Co₂-CNTs did not significantly boost the CIP aerobic transformation by sulfide (Fig. S13 in Supporting information).

This work prepared Co-loaded CNTs by a convenient method for the efficient aerobic transformation of antibiotics CIP driven by sulfide. It was found that ¹O₂ was the main active specie for aerobic CIP transformation with presence of sulfide and Co-loaded CNTs. The excellent electron transfer ability of Co-loaded CNTs might play an important role to promote the electron transfer from sulfide to form intermediate H₂O₂ and ¹O₂. This work not only offer a better understanding for the Co-loaded CNTs mediated aerobic antibiotics transformation by sulfide, but also provide a reference for the application of Co-loaded carbon materials on the organics aerobic transformation by sulfide in aqueous environments.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The authors wish to thank the National Natural Science Foundation of China (Nos. 52200186, U19A20108, 52025101 and 52070025), China Postdoctoral Science Foundation (No. 2021M693720) and Chongqing Municipal Education Commission (No. KJCX2020001) for financially supporting this study.

Supplementary materials

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2023.108694.