Facile Fabrication of Durable Biochar/H2-TiO2 for Highly Efficient Solar-Driven Degradation of Enrofloxacin: Properties, Degradation Pathways, and Mechanism.

Widespread application of TiO2 for degradation of antibiotics is restricted by mainly the low photodegradation efficiency under solar irradiation. To expand the application of TiO2, the key factors that should be improved are visible-light response, yield of electrons and holes, and durability. Herein, we report a visible-light responsive and durable sugarcane-bagasse-derived biochar supported hydrogenated TiO2 (HSCB/H2-TiO2) photocatalyst with higher electron production fabricated by a facile one-pot hydrogenation. Mild hydrogenation temperature preserved the lotus-stem-like structure of sugarcane bagasse and gave the photocatalyst great separability. The superior durability of HSCB/H2-TiO2 was demonstrated by 12 rounds of repeated degradation of methylene blue (MB). In addition, the electron paramagnetic resonance (EPR) results demonstrated that the biochar skeleton contains abundant persistent free radicals (PFRs), which can provide excess electrons to form more •O2 -. Meanwhile, radical quenching experiment and EPR radical trapping results also revealed that •O2 - was the most dominant species for enrofloxacin (ENR) degradation. Thus, the as-fabricated photocatalyst shows excellent solar-driven degradation of ENR, and 95.6% of ENR was degraded in 180 min under simulated solar irradiation. Possible ENR degradation pathways and mechanism are also proposed based on the identified intermediates.

Introduction

Water, one of the most fundamental renewable natural resources, is essential for human health.[1] Over 70% of human weight consists of water. Despite abundant water resources, only 2.5% of water is drinkable. Antibiotics in water worsen the shortage of drinking water. Fluoroquinolones (FQs), one of the most consumed antibiotics in the world, are of special concern because they have been shown to be genotoxic in samples from hospital wastewaters. Enrofloxacin (ENR), a commonly used veterinary FQ, is of greater interest in environmental systems because of its environmental accumulation, poor biodegradability, high adsorption affinity, and the long-term chronic effects at low concentrations in water.[2,3] In recent years, adsorption,[4] electrocatalysis,[5] photocatalysis,[6] and photoelectrocatalysis[5] have been developed to successfully alleviate the pollution from ENR. However, most of them cannot achieve cost efficient[2,7] and ecofriendly[2] operation simultaneously. Solar-driven photocatalysis using semiconductor materials seems to be a promising method to remove ENR in water owing to its high efficiency, relatively favorable cost–performance ratio, and it is free of unstable oxidants. Moreover, most solar-driven catalysts designed for water purification are solids, allowing them to be easily separated and recycled.[8,9]

TiO2 is a promising semiconductor photocatalyst.[10,11] Great accessibility and relatively low environmental toxicity allow its wide application in antibiotics degradation.[12] However, the poor solar response of TiO2 significantly limits its photoefficiency.[13] Only 4% of total solar energy, mostly UV light, can be utilized by TiO2,[14] thus severely restricting its widespread application. Bandgap narrowing is one of the most promising methods to enhance the response to visible irradiation.[15] We reported that hydrogenation and fluorine doping could increase the solar-driven degradation performance of TiO2 particles to varying degrees.[9] Han’s team dispersed Pt/black TiO2 onto a light-diffuse-reflection surface to demonstrate highly efficient photocatalytic steam reforming of methane. The yield of H2 under visible-light irradiation increased by 3-fold compared to previous reports.[16]

Another major drawback of TiO2 particles is their low yield of electrons. Although electrons can be excited from the valence band of TiO2, they quickly recombine with holes and are not able to form reactive oxygen species (ROS). The high electron–hole recombination rate severely restricts the yield of light-induced electrons. Yang et al. reported a carbothermal reduction induced Ti3+ self-doped TiO2/graphene quantum dot (GQD) nanohybrid. Citric acid was used as the organic source in their study. The nanohybrid showed better photoinduced charge generation and enhanced charge transport on the surface layer.[17] Besides use as an electron–hole separator, carbon-based materials were also used as a crystal growth controller. Yang et al. found that TiO2 nanoparticles deposited on a cellulose surface were spindle-like and had very low agglomeration.[18]

Modifying TiO2 with biochar, a carbon-rich, low-cost product derived from the pyrolysis of biomass with limited oxygen, is an optimal solution to the drawbacks discussed above. Different sources of biochar modified TiO2 are now a rising stars in water and air purification,[19] soil remediation,[20] and the semiconductor industry[21] because of their excellent conductivity and high radiation adsorption.[22,23] Moreover, persistent free radicals (PFRs) are generated during the formation of biochar, which can increase the formation of •O2– in heterogeneous degradation.[24] Sugarcane bagasse (SCB), one of the most critical and abundant organic byproducts of the sugar industry,[25] is an ideal organic source of carbon. However, biochar modified TiO2 suffers from a major drawback, which is poor durability. It is difficult to control the mass loss during repeated tests because of either small particle size or the fragility of the biochar skeleton. Shen et al.[8] decorated biochar with magnetic materials to increase the separation ability of composites. Although the magnetic absorbent can be easily separated from the solution using a magnet, the addition of magnetic materials increases the cost of manufacture.

In this report, we synthesized a visible-light responsive and durable sugarcane-bagasse-derived biochar supported hydrogenated TiO2 (HSCB/H2-TiO2) photocatalyst via one-pot hydrogenation. The morphology and chemical composition of the photocatalyst were characterized thoroughly. Its solar-driven degradation performance toward ENR and durability were systematically studied. The yield of •OH and •O2– were evaluated and compared, and the contribution of different ROS to ENR degradation was determined. Finally, the degradation intermediates of ENR were identified, and a possible mechanism of ENR degradation by HSCB/H2-TiO2 was proposed based on the results.

Experimental Section

Chemical Reagents

Predried SCB (average length 2 cm) was obtained from Nanhua Sugar Mill, Guangxi, China. Hydrogen peroxide (H2O2, AR, 30%) was purchased from Chuandong Chemical, Chongqing, China. Glacial acetic acid (CH3COOH, AR, 99.5%) was purchased from Damao Chemical, Tianjin, China. Titanium butoxide (Ti(BuO)4, AR, 99.0%), tertiary butyl alcohol (TBA, C4H10O, AR), tryptophan (Trp, AR), trichloromethane (CHCl3, AR), AgNO3 (AR), dimethylpyridine nitrogen oxide (DMPO, AR), 2,2,6,6-tetramethylpiperidine (TEMP, AR), 2,2,6,6-tetramethylpiperidine nitrogen oxide (TEMPO, AR), and ENR (C19H22FN3O3, AR) were purchased from Aladdin, Shanghai, China. Ethanol (CH3CH2OH, AR, 99.5%) was purchased from Fuyu Fine Chemical, Tianjin, China. All chemical reagents were used as received without further purification. Deionized water was used in this study.

Preparation of SCB/TiO2 Composites

First, SCB was soaked in a premixed H2O2/CH3COOH solution (H2O2/CH3COOH = 1:1, v/v), and the solution was magnetically stirred for 30 min at 25 °C. Then, the solution was transferred to a 60 °C water bath for 12 h. After thorough washing with deionized water, the pretreated SCB was oven-dried at 80 °C for 24 h. Next, SCB/TiO2 composites were prepared via a simple hydrothermal process. Typically, 2 g of pretreated SCB was soaked in 150 mL of ionized water followed by the addition of 60 mL of ethanol. Then, 12.5 mL of Ti(BuO)4 was slowly injected (2 drops/s) to the mixture under vigorous magnetic stirring. After 30 min reaction, the obtained creamy solution and SCB were transferred to a Teflon-lined autoclave and subjected to a 24 h hydrothermal process to synthesize SCB/TiO2. The hydrothermal temperature was set at 160 °C. The SCB/TiO2 was washed three times with deionized water and then dried in an oven at 80 °C for 24 h.

Preparation of HSCB/H2-TiO2 Photocatalyst

The as-obtained SCB/TiO2 composites were subjected to a mild one-pot hydrogenation process (H2/Ar volume ratio was 1:9, and hydrogenation temperature was 400 °C) for 3 h. Hydrogenation was performed in a silica tube furnace at ambient pressure. The silica tube diameter was 90 mm and length was 800 mm. To eliminate the effect of oxygen, the furnace was ventilated with a H2–Ar mixed gas for 15 min prior to the heating. The gas flow rate was set to 600 μL/s, and the temperature ramp rate was set to 5 °C/min. Finally, the as-prepared synchronous hydrogenated biochar/H2-TiO2 was naturally cooled to room temperature. Plain hydrogenated biochar was prepared following the same hydrogenation process. Pure TiO2 and H2-TiO2 particles were prepared following our previous method.[9]

Characterization

The specific surface area was determined by N2 adsorption–desorption isotherm using a Micromeritics ASAP 2460 instrument and calculated by BET method. The surface morphology and lotus-stem-like multichannel structure were observed using a scanning electron microscope (SEM, Zeiss Sigma 300). The elemental composition of a selected area was characterized using an energy-dispersive X-ray spectrometer (EDS, Zeiss SmartEDX). The morphology of HSCB/H2-TiO2 composites was observed by high-resolution transmission electron microscopy (HRTEM, FEI Tecnai G2 F20 S-TWIN) with an acceleration voltage of 200 kV. The X-ray diffraction (XRD) patterns of samples were obtained using a D/MAX 2500 V X-ray diffractometer (Rigaku, Japan) in the 2θ range of 10–90° and at a scanning rate of 5°/min. The surface functional groups of biochar and HSCB/H2-TiO2 were characterized by Fourier transform infrared spectroscopy (FT-IR, SHIMADZU IRTrace-100). The core level Ti 2p, O 1s, and C 1s spectra were determined by X-ray photoelectron spectroscopy (XPS, Thermo Fisher Scientific ESCALAB 250XI+). The amount of PFRs in the HSCB were determined from electron paramagnetic resonance (EPR) spectra recorded using a Bruker E580 spectrometer (Germany).

Measurement of Photodegradation-Related Properties

EPR radical trapping tests were conducted on a Bruker E580 spectrometer (Germany) with DMPO, TEMP, and TEMPO as spin-trapping agents. Radical quenching experiments were carried out to determine the dominant ROS in the ENR degradation. HSCB/H2-TiO2 was used as photocatalyst, and CHCl3, AgNO3, Trp and TBA were used as scavengers for •O2– and •OH–, respectively. To visualize the electron–hole recombination rate, steady and transient photoluminescence (PL) emission spectra were obtained with a Hitachi F7100 with 330 nm light excitation. Contact angle tests were carried out using a JY-82B Kruss DSA goniometer under solar irradiation to determine the hydrophilicity/hydrophobicity of HSCB/H2-TiO2 composites and H2-TiO2 particles. The sessile drop method was used, and an adjustable plane parallel backlight was used as the source of solar irradiation. Transient photocurrent and Mott–Schottky plot of HSCB/H2-TiO2 composites and H2-TiO2 particles were recorded using an electrochemical workstation (Shanghai Chenhua, CHI 760E). Specifically, HSCB/H2-TiO2 and H2-TiO2 were ground and dispersed in ethanol. After 30 min sonication, the dispersed liquid was coated on the ITO substrate and calcined at 400 °C to remove impurities. The curves of transient photocurrent vs time (I–t curves) were recorded using a standard three-electrode system (Pt was used as the counter electrode, Ag/AgCl was used as the reference electrode, and the photocatalyst-coated ITO was used as the working electrode) with periodic on/off illumination. The on/off cycle was set for 10 s. The electrolyte was 1 mol/L Na2SO4 aqueous solution, and a broadband 250 W xenon lamp (consisting of 9% UV, 43% visible light, and 48% infrared light) was used as the light source.

Solar-Driven Catalytic Degradation of ENR and Durability Tests

For the degradation of ENR, the typical process was that 20 mg of HSCB/H2-TiO2 (containing 0.25 mg of H2-TiO2) was first dispersed and soaked in 200 mL of 3.6 mg/L ENR solution with magnetic stirring in the dark for 2 h to reach the adsorption–desorption equilibrium. The sample was then transferred to a transparent cylindrical glass reactor. The wall of the reactor was filled with circulating cooling water to maintain the reactor at room temperature. A broadband 250 W xenon lamp (consisting of 9% UV, 43% visible light, and 48% infrared light) was used as the source of solar radiation. The source was placed 10 cm away from the reactor. The entire reaction system (source of irradiation and the reactor) was cooled using a ventilation device to avoid overheating. An injector equipped with a 0.45 nm filter was used for sampling at different time intervals. The filter was used to ensure that the collected solution was free from photocatalyst. The residual concentration of ENR in the solution was analyzed using liquid chromatograph/mass spectrometer (LC/MS, AB SCIEX 5500). A similar process was applied for ENR degradation by HSCB or H2-TiO2. The dosage for HSCB and H2-TiO2 was 20 mg and 25 mg, respectively. For the repeated MB degradation test, the process was similar to the degradation of ENR except the dark adsorption time, degradation time, and sampling intervals were 15 min, 3 h, and 5 min, respectively. The concentration of MB was analyzed using a UV–visible spectrophotometer. The photocatalysts were recycled by a simple back wash and dry process.

Results and Discussion

The specific surface area of HSCB, HSCB/H2-TiO2, and pure TiO2 nanoparticles and their pore characteristics were investigated by N2 adsorption–desorption isotherm (Figure S1), and the results can be found in Table S1. The BET surface area of HSCB was 43.7281 m2/g, which provided great adsorption capacity. Due to the adherence of H2-TiO2, the BET surface area of HSCB/H2-TiO2 was larger than that of HSCB. Meanwhile, the increased average pore diameter suggests that H2-TiO2 was densely loaded on the HSCB surface. Compared to pure TiO2 nanoparticles, the specific area and pore volume were decreased slightly. However, support from the biochar may provide easy separation to the composites. In addition, the biochar support has great capability for absorbing the pollutants, preventing them fromdissolving into the bulk solution, concentrating a target pollutant-rich environment at the HSCB/H2-TiO2 interface.[26] In comparison with HSCB/H2-TiO2, pure TiO2 nanoparticles would allow the dissolution of pollutants. Only when the pollutants collide with TiO2 again could the degradation be further conducted. These results may indicate that the HSCB/H2-TiO2 can adsorb pollutants efficiently and have excellent reusability.

The morphology of as-prepared hydrogenated SCB-derived biochar (HSCB) and HSCB/H2-TiO2 was recorded using SEM (Figure a–c,e). EDS mapping of a selected area is displayed in Figure d,f. As shown in Figure a,b, the lotus-stem-like multichannel hierarchical structure of HSCB was an ideal skeleton for the anchoring of TiO2. Interestingly, uniformly dispersed holes (UDHs) with an average diameter of ∼0.8 μm were observed on the side walls of HSCB. They act like channels connecting the internal structure of HSCB and the external environment, allowing more efficient mass transfer. In addition, they also neutralize the effect of the decrease in surface area and the change in pore size distribution caused by blocking the pore entrances with H2-TiO2. Figure c,e shows the microstructure of the channels and the surface of HSCB/H2-TiO2, respectively. As shown in Figure c, the typical hierarchical structure remained unchanged with full coverage of H2-TiO2. The multichannel organic framework provided a significantly large surface area and became a good site for the anchoring of TiO2. Uniform and dense deposition of H2-TiO2 can be observed on the surface of HSCB in Figure e. The EDS spectra (Figure d,f) further confirmed the universal distribution of H2-TiO2.

Figure 1

(a, b) SEM images of HSCB. (c) SEM image of the front end of HSCB/H2-TiO2; insets, enlarged images of the selected (yellow square) area. (d) EDS elemental mapping of corresponding insets. (e) SEM image of the side wall of HSCB/H2-TiO2. (f) EDS elemental mapping of corresponding insets. (g) TEM image of HSCB. (h–j) TEM images of HSCB/H2-TiO2 with different magnifications.

The dispersion and crystalline structure of HSCB (Figure g) and HSCB/H2-TiO2 (Figure h–j) were further studied by HRTEM analysis. Comparing HSCB with HSCB/H2-TiO2, full coverage of H2-TiO2 on the HSCB surface can be observed. H2-TiO2 particles were anchored tightly on the HSCB surface, and clear lattice fringes were observed (Figure i), representing the well-defined crystalline structure of H2-TiO2. A fringe spacing of 0.35 nm indicates the successful synthesis of anatase TiO2.[27] Among anatase, brookite, and rutile TiO2, anatase TiO2 has the highest visible-light adsorption and fastest organic degradation efficiency.[28]

The XRD patterns of H2-TiO2 and HSCB/H2-TiO2 were recorded and are shown in Figure a. The XRD pattern of anatase TiO2 (PDF no. 21-1272) was also recorded for reference. It can be seen that all characteristic peaks of HSCB/H2-TiO2 were consistent with those of anatase H2-TiO2. The (101) diffraction peak of HSCB/H2-TiO2 appeared at 2θ of 25.3°, representing an interplanar distance of 0.352 nm,[29] consistent with the TEM result. The peak intensity of HSCB/H2-TiO2 clearly decreased compared with H2-TiO2, indicating that the composition significantly affected the crystalline structure of H2-TiO2.

Figure 2

(a) XRD patterns of H2-TiO2 and HSCB/H2-TiO2. Pattern of anatase TiO2 is shown as reference. (b, c) FT-IR spectra and XPS full scan, respectively, of HSCB and HSCB/H2-TiO2. (d, e) High-resolution XPS O 1s spectra of HSCB/H2-TiO2 and HSCB, respectively. The deconvoluted results are shown as shaded regions. (f, g) High-resolution C 1s XPS spectra and the deconvoluted results for HSCB/H2-TiO2 and HSCB, respectively. (h) High-resolution Ti 2p XPS spectrum of HSCB/H2-TiO2. (i) Solid state EPR results of different samples.

To further characterize the composite, the FTIR spectra of HSCB and HSCB/H2-TiO2 were recorded, as shown in Figure b. The absorption peaks of HSCB at 3448 and 1612 cm–1 can be attributed to the hydroxyl group or absorbed water (O–H stretching mode)[30] and the skeletal vibration of carbon framework,[31] respectively. The spectra of HSCB/H2- TiO2 showed a similar pattern; a new characteristic peak appeared at ∼620 cm–1, can be attributed to the vibration of Ti–O–Ti and Ti–O–C bonds.[31] The strong signal of Ti–O–R (R = Ti or C) bonds in HSCB/H2-TiO2 indicated that the chemical bonds between H2-TiO2 particles and the HSCB surface were formed during the hydrothermal and hydrogenation process. In addition, compared to HSCB, the adsorption peak at 3448 cm–1 in HSCB/H2-TiO2 red-shifted slightly and its intensity was increased. This can be attributed to the increased hydrophilicity of surface-loaded H2-TiO2. According to previous studies, pure TiO2 is less hydrophilic (contact angle 72° ± 1°) without UV irradiation,[32] which would restrict the absorption of water. In the present study, HSCB/H2-TiO2 exhibited increased hydrophilicity even under solar irradiation. We measured the hydrophilicity under ambient environment via contact angle tests, and the results are shown in Figure S2. HSCB and HSCB/H2-TiO2 had good contact with water, and their contact angles were 33.69° and 40.86°, respectively. The tendency to contact with water allowed the HSCB/H2-TiO2 surface to absorb more water molecules, resulting in the amplification of absorption peaks at 3448 cm–1. Moreover, the decrease in contact angle could be considered as evidence that TiO2 was excited. Accordingly, the hydrophilicity of TiO2 would not be any different when irradiated by solar light, but it becomes more hydrophilic under UV irradiation.[32] This is because the electrons present in TiO2 could be easily excited by UV irradiation. After the electron excitation, the holes left in the conduction band would diffuse and accumulate at the TiO2 surface where they were captured by the reduced species (normally −OH) present in the electrolyte,[33] and the original less hydrophilic surface would be modified into a super hydrophilic surface.

XPS is a powerful technique to investigate the chemical state and environment on the surface of HSCB/H2-TiO2. The full-scale XPS spectra of HSCB and HSCB/H2-TiO2 are shown in Figure c. Typical Ti 2p, O 1s, and C 1s peaks could be observed in Figure c, confirming that TiO2 particles were successfully deposited on the biochar.

The O 1s high-resolution XPS spectra of HSCB/H2-TiO2 and HSCB were deconvoluted and are shown in Figure d,e, respectively. Compared with HSCB, a new peak appeared at ∼532.1 eV in the HSCB/H2-TiO2 sample. This peak could be assigned to Ti–O[17,34−39] from the TiO2 lattice, and it demonstrated the successful deposition of H2-TiO2 onto the HSCB surface, which is consistent with the TEM and FTIR results. Additionally, a slight shift could be observed in HSCB/H2-TiO2. The peak located at ∼530.9 eV represented the −COOH,[37] and it shifted from 530.9 to 531.4 eV. The shift was a significant signal for the change in the surface chemical environment. This change may be the result of the formation of oxygen vacancy.[38,39]

The C 1s deconvoluted XPS spectra of HSCB/H2-TiO2 and HSCB are shown in Figure f,g, respectively. In the HSCB, there are three identical peaks that can be assigned to C (∼284.7 eV),[17,34,36,38,40] C–OR/C–OH (∼285.6 eV), and C=O (∼287.7 eV).[38] According to the previous report, the peak representing C–OR/C–OH normally appeared at around 286.1 eV.[40] The obvious change in the peak location (286.1 to 285.6 eV) is mainly because of the formation of PFRs inside the biochar. Similar peaks can be located in HSCB/H2-TiO2, except the C=O peak shifted from ∼287.7 to 287.1 eV and there is a drastic drop in the concentration of C–O peak. With the load of TiO2, Ti–O–C bonds were formed on the HSCB surface and the surface chemical environment was changed. Consequently, the shift toward lower binding energy region happened in the characteristic peak of C–OR (from 288.9 to 287.6 eV) and the concentration of C–O decreased (from 25.9% to 14.3%). The characteristic peak of Ti–C bond located at ∼281 eV was not observed. This indicated that the substitution of TiO2 lattice oxygen atom by carbon atom did not happen.[35,36,38] The discussion above is highly in accordance with the XPS O 1s results and the FT-IR spectra.

Figure h shows the Ti 2p XPS spectra of HSCB/H2-TiO2. Two typical deconvoluted peaks located at ∼458.8 eV and ∼464.5 eV corresponding to Ti4+ 2p3/2 and Ti4+ 2p1/2 could be found in both samples. Two new peaks located at ∼456.2 eV and ∼461.9 eV corresponded to the Ti3+ 2p3/2 and Ti3+ 2p1/2 orbitals, respectively.[17] The Ti3+ XPS signal confirmed the existence of Ti3+ in the composite and was also a strong evidence for the formation of oxygen vacancy in the TiO2 lattice, because oxygen vacancies and Ti3+ must be present together to reach charge equilibrium.[17]

The concentration of defects has been considered as a key factor affecting the photodegradation efficiency of photocatalytic materials. During hydrogenation, defects such as Ti3+ and oxygen vacancies could be formed in the TiO2 lattice and PFRs would be formed in the HSCB.[41] The PFRs would produce reactive oxygen species (ROS) when the HSCB contacted water.[24,42,43] Thus, EPR was applied to determine the number of defect species (such as Ti3+, oxygen vacancies and PFRs in this case) in HSCB and HSCB/H2-TiO2. Compared to traditional X-ray-based techniques, EPR can be used to determine the number of unpaired electrons inside materials by recording the resonance of electrons. As shown in the EPR spectra of pure TiO2, H2-TiO2, HSCB, and HSCB/H2-TiO2 (Figure i), pure TiO2 barely displayed any signal, and H2-TiO2 presented clear signals owing to the hydrogenation-induced Ti3+ and oxygen vacancies. Both HSCB and HSCB/H2-TiO2 showed very strong EPR signal owing to the large amount of PFRs present in biochar[43] and the defect sites presented in H2-TiO2.[9] A previous study revealed that, in the absence of H2O2, when HSCB contacted aqueous solution, the PFRs would transfer an electron to dissolved oxygen forming •O2–.[42] HSCB/H2-TiO2 showed even stronger EPR signal compared to HSCB, which can be attributed to significantly increased defect sites in the H2-TiO2. The enriched defects might be introduced by the synergistic effect of hydrogenation[44] and carbothermal reduction.[17] With the help of abundant defects, more electrons can be easily excited (due to the lower energy level of defects), leading to higher production of ROS. Based on the results of XPS and EPR analysis, it can be predicted that the organic/inorganic composites will achieve a higher photodegradation efficiency than TiO2 or HSCB.

Measurement of Photodegradation-Related Properties

The amount of excited electrons from H2-TiO2 was found to be critical for successful photodegradation in water. However, the excited electrons were highly inclined to recombine with holes. To increase the photodegradation efficiency, a good method is to decrease the electron–hole recombination rate. Steady-state and transient PL spectra were used to evaluate the recombination rate. It can be seen in Figure a,b that HSCB/H2-TiO2 had weaker steady-state PL intensity and shorter carrier lifetime compared to H2-TiO2, indicating that it has a lower electron–hole recombination rate under solar irradiation.[45] The efficient separation between electrons and holes may be explained by the numerous hydrogenation-induced defect sites. They act like electron sinks to facilitate the separation of electron–hole pairs.[46]

Figure 3

(a–c) Steady-state PL spectra, transient PL spectra, and current versus time (I–t) curves, respectively, for H2-TiO2 and HSCB/H2-TiO2. (d–f) EPR-DMPO radical trapping tests against (d) H2-TiO2, (e) HSCB, and (f) HSCB/H2-TiO2. (g, h) EPR-e– and EPR-1O2 trapping results, respectively. (i) Result of ENR degradation by HSCB/H2-TiO2 with different scavengers for radical quenching.

Measurement of transient photocurrent response is an effective method to quantify the amount of excited electrons from H2-TiO2. The I–t curves of H2-TiO2 and HSCB/H2-TiO2 with on–off cycles recorded under xenon lamp irradiation are shown in Figure c. Both as-prepared samples responding to the xenon lamp showed similar I–t patterns, demonstrating their excellent visible-light absorption. Furthermore, the photocurrent intensity of HSCB/H2-TiO2 was almost quintupled compared with that of H2-TiO2. The increase can be attributed to electrons provided by the PFRs from biochar and the excited electrons from surface loaded H2-TiO2. Once the electrons were excited by the incident irradiation, they were rapidly transferred to the back-contacted ITO electrode via graphitized HSCB.[47] Therefore, higher photocurrent density and more efficient production of ROS could be achieved simultaneously, and the degradation ability of the composite would be increased.

In order to determine the quantity of ROS generated in aqueous solution, EPR DMPO spin-trapping experiments with H2-TiO2 (Figure d), HSCB (Figure e), and HSCB/H2-TiO2 (Figure f) were performed, and the spectra of •O2– and •OH under dark conditions and light irradiation were obtained. As shown in the DMPO trapping results, without irradiation no •OH signal was observed in all three samples and weak •O2– signal can be detected in HSCB and HSCB/H2-TiO2. This was because the PFRs contained in the biochar can be sensitized and generate electrons when in contact with water. The as-generated electrons are then transmitted to oxygen to form •O2–. After 5 min of solar irradiation, typical •OH and •O2– signals could be observed in all samples. For HSCB, compared with the signals under dark conditions, the •O2– signal remained almost unchanged and the •OH signal appeared. The •OH signal from HSCB is because the as-formed •O2– can easily dismutate with H+ to form H2O2, which can be photosensitized to form •OH[42] then captured by the DMPO. Moreover, the obvious •OH and •O2– signals from H2-TiO2 confirm that H2-TiO2 can be successfully excited by solar irradiation. After excitation, electrons in the valence band can jump into the conduction band and leave holes behind in the valence band. Those electrons and holes will then transport to the surface of H2-TiO2 to react with O2 and H2O, forming •O2– and •OH, respectively. Based on the even stronger •O2– and •OH signals from HSCB/H2-TiO2, we can predict that HSCB/H2-TiO2 generates more ROS in water compared with other samples and excellent degradation performance can be predicted.

Besides •OH and •O2–, we also investigated other active oxidation species by ESR, such as 1O2 and electrons. TEMP and TEMPO were used as spin trapping agents for 1O2 and e–, respectively. The ESR spin trapping spectra of e– and 1O2 are shown in Figure g,h, respectively. It can be seen from ESR-e– spectra that H2-TiO2 did not generate any e– without irradiation. Even under dark conditions, the ESR-e– signal of HSCB and HSCB/H2-TiO2 decreased, revealing that they can generated electrons without visible irradiation, which was inconsistent with the ESR-•O2– results. When visible irradiation was applied, significant decrease in the intensity of the ESR-•O2– signal could be observed in all three samples, indicating that a huge amount of electrons were generated. Proposed by other researchers, 1O2 can be generated via oxidation of •O2– (eq ).[48,49] Thus, the ESR-1O2 spectra could be used as proof of the simultaneous existence of •OH and •O2–. The results of ESR-1O2 demonstrated that HSCB cannot generate •OH even with irradiation applied.

Radical quenching tests were conducted to estimate the contribution from different ROS, and the results are shown in Figure i. HSCB/H2-TiO2 was used as photocatalyst. AgNO3, Trp, CHCl3, and TBA were used as scavengers for e–, 1O2, •O2–, and •OH–, respectively. It can be seen that after 2 h dark reaction, the concentration of ENR did not decrease in the samples without •O2–. The concentration slightly dropped in the samples in which •O2– was not quenched. This indicates that biochar can generate •O2– without any irradiation. The result is highly in accordance with the analysis of EPR spin trapping results. After solar irradiation was applied, ENR concentration drops quickly in the samples in which •O2– was not quenched. The degradation reached equilibrium in 180 min, and the final degradation rate could reach 81.2% (ENR + TBA) and 95.6% (ENR). In contrast, in the samples without •O2–, the degradation rates of ENR were greatly inhibited and its concentration hardly decreased after 120 min. The final degradation rates were only 26.1% (ENR + CHCl3) and 8% (ENR + TBA + CHCl3). From the quenching results, it can be concluded that •O2– contributes to the ENR degradation more than •OH–.

Measurement of Durability of the Photocatalyst

MB is a commonly used organic dye and chemical indicator. The solar-irradiation degradation of MB by HSCB/H2-TiO2 was recorded to simulate the degradation of organic dyes, and its degradation by HSCB and H2-TiO2 particles was used for comparison. The final degradation rate for each round is shown in Figure a. HSCB/H2-TiO2 can degrade MB effectively under solar irradiation. At the end of first round, about 95.4% of MB was removed by HSCB/H2-TiO2, while the HSCB and H2-TiO2 only removed 12% and 49% of MB. After 12 rounds of repeated degradation, HSCB/H2-TiO2 could still degrade ∼55.1% of MB, while less than 9% and 11% MB could be degraded by H2-TiO2 and HSCB, respectively. The recovery rate of all photocatalysts at the end of each round is also recorded in Figure a. The weight of recovered HSCB/H2-TiO2 decreased smoothly and gradually. The final weight of HSCB/H2-TiO2 after 12 rounds remained around 84% of the initial weight. The HSCB underwent 15% weight loss, retaining 84.8% of its initial weight. For comparison, the weight loss of H2-TiO2 particles was as high as 51%. The small particle size of H2-TiO2 might be responsible for the unexpected loss. It is highly possible that small particles were stuck on the filter paper during separation. This is strong evidence for the superior durability of composites and the great separation ability of HSCB/H2-TiO2. We attributed the high durability of our composite to the mild synchronous hydrogenation. After the deposition of the Ti precursor, the mild hydrogenation facilitated further growth of TiO2 on the surface of cellulose. Strong bonding between the biochar and H2-TiO2 reduced the leaching of surface-loaded catalysts. On the other hand, low hydrogenation temperature allowed the biochar to maintain its external shape almost undamaged, which was beneficial for easy separation. Moreover, during repeated tests, the mass loss of HSCB/H2-TiO2 was acceptable. It was indicated that the template had excellent strength. We compared the HSCB/H2-TiO2 with previous reports. The results are listed in Table . The as-synthesized photocatalyst showed superior photocatalytic performance in ENR degradation.

Figure 4

(a) Repeated MB degradation results: (top) MB degradation rate by different samples at the end of each round; (bottom) recovery rate of samples at the end of each round. (b) ENR degradation by HSCB/H2-TiO2 at different pH values. (c) Degradation efficiencies of various samples at pH = 7 for ENR. The first 2 h was reacted in dark reaction, followed by a 240 min of solar irradiation catalytic degradation. The error bar is based on the results of parallel tests. (d) Repeated ENR degradation results using HSCB/H2-TiO2.

Table 1

Photocatalytic Degradation Efficiencies of Various Photocatalysts for ENR

photocatalyst	light	cat. dosage (g L–1)	ENR concn (mg L–1)	degradation efficiency	ref
Ti-KBC	visible light (15 W)	2.5	100	58.97% in 60 min	(50)
GaOOH/ZnBiTaO5	mercury lamp (500 W)	1	10	58.27% in 60 min	(51)
P2–Mn3O4	xenon lamp (500 W)	0.1	a	72.3% in 10 min (with the addition of PMS)	(52)
BN/BiPO4	UV light (250 W)	3.33	10	91.5% in 2 h	(53)
HSCB/H2-TiO2	xenon lamp (250 W)	0.1	3.6	95.6% in 180 min	this work

Not reported.

ENR Degradation Test and Possible Mechanism

The effect of different solution pH on the ENR degradation by HSCB/H2-TiO2 was monitored, and the results are showed in Figure b. It can be seen that the most ENR could be degraded in neutral solution. The effect of different pH value could be explained because in the acidic environment, the surface-loaded TiO2 and ENR were both positively charged while they were both negatively charged in the alkaline environment. Therefore, when the solution pH deviated from neutral, electrostatic repulsion would occur between the catalyst and the targeted antibiotic and prevented them from contacting, which would drastically impede the ENR degradation efficiency.

Figure c displayed the results of the degradation of ENR using HSCB/H2-TiO2, HSCB, and H2-TiO2. As can be seen, the adsorption percentage of H2-TiO2 and HSCB/H2-TiO2 was 6% and 1.6%, respectively. The adsorption percentages were in accordance with their specific surface area. For HSCB, the ENR removal percentage was 20.5% during the 2 h dark reaction. HSCB exhibited the lowest specific surface area among HSCB, H2-TiO2, and HSCB/H2-TiO2. However, based on the results of active oxidation species trapping ESR spectra, HSCB generated e– even under dark conditions. Therefore, the ENR removal was not only contributed by the physical adsorption but also the oxidation by the active oxidation species in the water. During the following 240 min irradiation, the concentration barely reduced. It can be concluded that the degradation reached equilibrium after the 2-h dark reaction. For the H2-TiO2 and HSCB/H2-TiO2 samples, when the simulated irradiation was applied, the concentration of ENR dropped drastically in the first 15 min for both samples. This was because the ENR underwent self-sensitized photolysis and light-excited electron direct photolysis. According to previous research,[54] solar irradiation can be absorbed by ENR, and the exposure to irradiation can destroy ENR’s resistance to radical degradation. From 15 to 30 min, the ENR continued to be drastically degraded by HSCB/H2-TiO2, while the rate of degradation of ENR by H2-TiO2 decreased. Based on the discussion of the excess amount of electrons provided by HSCB and the improved hydrophilicity of HSCB/H2-TiO2, more electrons and holes were created and transferred to the surface of the composites, contacting with surface-absorbed ENR molecules and facilitating the degradation. After 30 min, the rate of degradation of ENR by HSCB/H2-TiO2 started to decrease because ENR contains carbonyl, C–F bonds, and piperazinyl, which are very hard to break. Self-sensitized photolysis and electron direct photolysis were not strong enough to break them. However, in the heterogeneous photocatalytic system, strongly oxidizing •O2– and •OH can be generated, and under their attacks, decarboxylation, defluorination, or piperazinyl N4-dealkylation occurred effectively.[55] Eventually, hydroxyl and superoxide radical oxidation will degrade ENR into CO2, H2O, F–, and NO3–. The whole degradation test lasted for 4 h, and it reached equilibrium in 180 min. The equilibrium degradation efficiency reached as high as 95.6%. The photostability of HSCB/H2-TiO2 was tested by five rounds of repeated ENR degradation. The results in Figure d demonstrated that our catalysts had great photostability. After 5 rounds of repeated tests, the degradation efficiency could still reach 77%. In addition, we measured the total organic carbon (TOC) removal efficiency using HSCB/H2-TiO2 to further explore its mineralization capability. The results are shown in Figure S3. It could be seen that around 77% of TOC could be removed in 240 min under visible irradiation, demonstrating the excellent mineralization capability of HSCB/H2-TiO2.

H2-TiO2 and HSCB/H2-TiO2 were characterized by UV–vis/diffuse reflection spectroscopy (DRS0 (Figure a) and with an electrochemical workstation. Data on TiO2 was also collected as control group. Their bandgap (Eg) and conductive band edge (ECB) were estimated via Tauc plots (Figure b) and Mott–Schottky plots (Figure c), respectively. The valence band edge (EVB) of samples was calculated by combining the value of ECB with Eg, and the results are shown in Figure d, together with the potentials of O2/•O2– (−0.33 V) and OH–/•OH (3.2 V). It can be seen from Figure a that the light absorption range of TiO2 was from 200 to 500 nm and those of H2-TiO2 and HSCB/H2-TiO2 were extended to 800 nm. Their absorption edges were estimated to be 420 nm (TiO2), 445 nm (H2-TiO2), and 560 nm (HSCB/H2-TiO2), and the corresponding Eg values, estimated using Tauc plots,[56] were 3.37 eV (TiO2), 3.16 eV (H2-TiO2), and 2.8 eV (HSCB/H2-TiO2). The addition of HSCB improved the light absorption edge and narrowed the bandgap width, enhancing the light utilization capacity of catalysts.

Figure 5

(a) UV–vis/DRS spectra of H2-TiO2 and HSCB/H2-TiO2. Data for TiO2 was used as control group. (b, c) Tauc plot and Mott–Schottky plot for the samples. (d) Scheme of the energy band structures of as-prepared samples.

From the Mott–Schottky plots, it can be seen that the samples had positive slopes, implying they all are n-type semiconductors. The flat band edges of samples were estimated to be −0.6 V (TiO2), −0.49 V (H2-TiO2), and −0.35 V (HSCB/H2-TiO2) vs Ag/AgCl. According to previous researchers, the flat band edge of an n-type semiconductor is about 0.2 V more positive than its ECB. Therefore, the ECB values were estimated to be −0.4 V (TiO2), −0.29 V (H2-TiO2), and −0.15 V (HSCB/H2-TiO2) vs Ag/AgCl or −0.6 V (TiO2), −0.49 V (H2-TiO2), and −0.35 V (HSCB/H2-TiO2) vs NHE. The results showed that HSCB/H2-TiO2 had a less negative VB level than that of H2-TiO2, which revealed that the reduction ability of HSCB/H2-TiO2 was slightly lower. Nevertheless, from Figure d, the VB level of HSCB/H2-TiO2 was more negative than the redox potential of O2/•O2– (−0.33 V), so the catalyst still had the ability to reduce O2 to yield •O2. Based on the Eg and ECB results, the EVB can be calculated to be 2.77 V (TiO2), 2.67 V (H2-TiO2), and 2.45 V (HSCB/H2-TiO2) vs NHE. The CB level of HSCB/H2-TiO2 was more positive than the redox potential of OH–/•OH (3.2 V), which revealed that •OH radicals could be successfully produced by oxidized OH– ion or H2O despite the oxidation ability of HSCB/H2-TiO2 being weaker than that of H2-TiO2.

Based on the above results and discussions, a possible photocatalytic mechanism can be proposed. The one-pot hydrogenation generated abundant PFRs in the biochar support and plentiful defect sites in the surface-loaded H2-TiO2 simultaneously. When the catalyst contacts water, the PFRs in the biochar would be sensitized and electrons would be transferred from the PFRs to the surface of the photocatalyst. In addition, hydrogenation can effectively narrow the bandgap width of TiO2 (from 3.37 to 3.16 eV), allowing irradiation with longer wavelength to be utilized by the photocatalysts. With the addition of sugarcane biochar, the bandgap was further narrowed to 2.8 eV. Furthermore, the addition of the biochar skeleton not only increased the separation efficiency of electron–hole pairs but also provided additional electrons to further increase the photodegradation efficiency. When simulated solar irradiation was applied, the defect-induced lower energy band allowed more electrons to be excited from the valence band of H2-TiO2. The defects also acted as electron sinks to prevent the electrons from recombining with the holes. With the synergetic help of biochar and H2-TiO2, excess electrons accumulated on the surface of the photocatalyst, reacting with O2 to form excess •O2–, leading to high degradation efficiency against ENR.

Photocatalytic Degradation Pathways of ENR

Eleven intermediates of ENR degradation were identified by LC-MS, and detailed information about them can be found in Table S2, and the mass spectra of intermediates are provided in Figure S4. According to the dominant ROS in the heterogeneous photodegradation system and the intermediates, three possible degradation pathways were proposed in Figure . The first pathway was caused by the direct photolysis by irradiation and the attack by •O2–. P1 was produced via de-ethylation with the attack of •O2– and then was transformed into P2 via photolysis. For the second pathway, ENR was direct attacked by electrons and P3 was produced via defluorination. The piperazinyl of P3 can be continuously attacked by •O2– to form P4 and P5. In the degradation of ENR, •O2– and •OH mainly attack piperazinyl[57] and quinolone,[57,58] respectively. Thus, the third pathway of ENR degradation was caused by oxidation via •OH. P6 was produced through the breaking of the quinolone ring from P4. Then P6 went through decarboxylation to produce P7. P7 was continuously degraded by the synergetic attack of •O2– and •OH to produce intermediates P8–P11.

Figure 6

Possible ENR degradation pathways catalyzed by HSCB/H2-TiO2 under solar irradiation.

Conclusion

In summary, highly durable HSCB/H2-TiO2 composites capable of excellent solar-driven degradation of ENR were successfully synthesized via a facile one-pot hydrogenation for the first time. The mild hydrogenation process preserved the lotus-stem-like multichannel skeletal structure of SCB, which allowed more sites for the composites to contact water and irradiation. Characterizations revealed that H2-TiO2 particles were well dispersed on the surface of biochar and strongly connected with the support. Twelve rounds of repeated degradation tests were conducted, and the highly durable HSCB/H2-TiO2 could still degrade ∼55.1% of simulated pollutant with well-controlled mass loss. EPR DMPO trapping results and radical trapping tests demonstrated the large amount of •O2– in the heterogeneous photodegradation system and the major contribution of •O2– in efficient ENR degradation. Solid state EPR demonstrated the existence of PFRs in the biochar, which were the main reason for the excess amount of •O2–. Decreased electron–hole recombination rate of composites was determined by PL spectra, and increased transient photocurrent indicated that more photoexcited electrons were generated. Combined with the synergetic contributions from HSCB and H2-TiO2, efficient ENR degradation was achieved: 95.6% of ENR could be removed by HSCB/H2-TiO2 in 180 min. Moreover, possible photodegradation pathways of ENR were proposed based on the identified intermediates. Therefore, the highly durable HSCB/H2-TiO2 composites not only expand the industrial application of TiO2 but also provide new choices for the environmental and high-value application of sugarcane bagasse.