Removal of levofloxacin through adsorption and ...

Chemosphere 280 (2021) 130626

Available online 3 May 20210045-6535/© 2021 Elsevier Ltd. All rights reserved.

Removal of levofloxacin through adsorption and peroxymonosulfate activation using carbothermal reduction synthesized nZVI/carbon fiber

Weitong Tan a, Yang Ruan a, Zenghui Diao c, Gang Song a, Minhua Su a, Li’an Hou a, Diyun Chen a, Lingjun Kong a,b,*, Hongmei Deng a,**

a Guangdong Provincial Key Laboratory of Radionuclides Pollution Control and Resources, School of Environmental Science and Engineering, Guangzhou University, Guangzhou, 510006, China b Guangdong Key Laboratory of Environmental Catalysis and Health Risk Control, School of Environmental Science and Engineering, Institute of Environmental Health and Pollution Control, Guangdong University of Technology, Guangzhou, 510006, China c School of Environmental Science and Engineering, Zhongkai University of Agriculture and Engineering, Guangzhou, 510225, China

A R T I C L E I N F O

Handling Editor: Jun Huang

Keywords: Nano zero valent iron Carbothermal reduction Peroxymonosulfate Levofloxacin Cotton fiber

A B S T R A C T

Nano zero-valent iron (nZVI) is widely used for decontamination. The main issues associated with nZVI are agglomeration and oxidation in the long term. In this study, the carbothermal reduction of cotton fiber was conducted for the synthesis of nZVI supported on cotton carbon fiber (nZVI/CF) to address the agglomeration and oxidation of nZVI. Synergistic adsorption and peroxymonosulfate (PMS) activation using nZVI/CF for removing levofloxacin (LEV) are reported herein. The nZVI concentration and morphology were conveniently adjusted by soaking cotton fiber in ferric nitrate solutions of various Fe3+ concentrations. The carbothermal reduction of the cotton fiber at 900 ◦C contributed to the reduction of Fe3+ into nZVI. A nZVI/CF-900-0.3 system was obtained through the carbothermal reduction of cotton fiber soaked in 0.3 M ferric nitrate. Favorable adsorption of nZVI/CF-900-0.3 to LEV facilitated LEV degradation under PMS activation. Approximately 93.83% of LEV (C0 = 20 ppm) was removed within 60 min with 0.2 g/L of the catalyst and 1 mM PMS. It was preferable to use nZVI + CF-900 to activate PMS for degrading LEV, thus confirming the favorable effect of LEV adsorption on further degradation. The nZVI/CF-900-0.3 exhibited excellent long-term stability given that it was able to activate PMS after it was stored for 6 months. ⋅SO4

− played an important role in LEV degradation in the presence of PMS.

1. Introduction

Levofloxacin (LEV) is a third-generation water-soluble fluo-roquinolone antibiotic that is widely used to eradicate bacterial diseases (Mahmoud et al., 2020). LEV overuse increases the amount of LEV-containing wastewater released from pharmaceutical factories and hospitals, resulting in the formation of resistance genes as well as elevated endocrine disorder and cancer risks (Ma et al., 2019). LEV is refractory and nonbiodegradable owing to the high dissociation energy of the C–F bond in nature, which is difficult to break (Van Doorslaer et al., 2014; Zhong et al., 2021). Because the toxicity of halogen-containing organic molecules in aqueous environments has attracted considerable attention, investigations on LEV degradation

should focus on breaking of the C–F bond. Advanced oxidation processes (AOPs) represent a promising approach to degrading refractory organic pollutants.

AOPs based on sulfate radicals (⋅SO4− ) have increasingly attracted

attention for degrading recalcitrant organic contaminants in water due to their advantages of high redox potential (2.5–3.1 V), long lifetime (approximately 40 μs) (Li et al., 2018; Zhu et al., 2020), wide working pH range, and nonselectivity (Liu et al., 2018; Zhang and Wang, 2019). Generally, ⋅SO4

− can be generated through peroxymonosulfate (PMS) activation by heat (Nie et al., 2014), UV radiation (Takdastan et al., 2018), catalysts (Furman et al., 2010), and ultrasound (Cai et al., 2015). Among these approaches, PMS activation using heterogeneous envi-ronmentally friendly catalysts is a benign and economic process (Ma

* Corresponding author. Guangdong Provincial Key Laboratory of Radionuclides Pollution Control and Resources, School of Environmental Science and Engi-neering, Guangzhou University, Guangzhou, 510006, China. ** Corresponding author.

E-mail addresses: [email protected] (L. Kong), [email protected] (H. Deng).

Contents lists available at ScienceDirect

Chemosphere

journal homepage: www.elsevier.com/locate/chemosphere

https://doi.org/10.1016/j.chemosphere.2021.130626 Received 26 January 2021; Received in revised form 26 March 2021; Accepted 18 April 2021

Chemosphere 280 (2021) 130626

2

et al., 2020; Yao et al., 2021). Notably, iron, the fourth most abundant element in the Earth’s crust (Tanaka et al., 2019), is environmentally friendly and does not cause secondary pollution. Fe-based heteroge-neous catalysts that can activate PMS to eliminate pollutants in waste-water have been widely discussed due to their low cost, low toxicity, and environment-friendly nature, among other advantages (Pang et al., 2019). In particular, zero valent iron (ZVI), owing to its favorable reduction potential and electron transformation ability, has been widely studied in the context of PMS activation to eliminate organic pollutants from wastewater (Liu et al., 2021). Cao et al. (2019) investigated PMS activation with ZVI for tetracycline degradation. Yang et al. (2018) used ZVI-montmorillonite for PMS activation to remove bisphenol. Ma et al. (2018) reported the PMS activation performance of nZVI in lignin-derived hydrochar for phenol removal. In a previous study, we achieved efficient PMS activation by using ZVI to degrade dyes in wastewater (Pang et al., 2019). Surface corrosion of ZVI limits the transformation of zero-valent iron to Fe2+ or Fe3+, further limiting the PMS activation efficiency. Therefore, nZVI, characterized by a large surface area and high reactivity, was developed to increase the PMS activation efficiency.

However, the instability and aggregation of nZVI due to its large surface area and strong magnetic interaction were the main issues limiting its further application (Zhang et al., 2020; Zhou et al., 2014). The large surface area and high reactivity of nZVI improved its ability to activate PMS, but these characteristics led to the unfavorable outcomes of agglomeration and oxidation. To overcome these limitations, various types of porous materials, such as montmorillonite (Zhang et al., 2013), resin (Du et al., 2014), mesoporous silica (Petala et al., 2013), and biochar (Dong et al., 2017) have been employed as nZVI supporters to avoid its agglomeration and maintain its ability to activate PMS. Among them, biochar has attracted considerable attention owing to its large specific surface area (Hao et al., 2021; Lyu et al., 2020), porous structure (Zhang and Wang, 2020), and oxygen-containing groups (Wang et al., 2020). Hence, biochar is among the most ideal supporters of nZVI (Wang et al., 2019). Biochar-supported nZVI has been widely synthesized using liquid phase reduction through the dropwise addition of sodium boro-hydride (NaBH4) into a biochar and ferrite solution (Wang et al., 2017; Zhang, D.J. et al., 2019).

Biochar can be obtained through the carbonization of organic pre-cursors in an inert atmosphere. Recently, the use of organic waste to prepare biochar has attracted considerable interest from the viewpoint of recycling resources. It is practical to use organic waste as precursors of biochar. Up to 562 million tons of waste cotton fiber is assumed to be generated in China, which accounts for 28% of the total textile waste in China (Xu et al., 2018). Waste cotton fibers can be converted to porous carbon fiber through carbonization, a process in which cotton fibers are decomposed into small organic molecules in a reducing atmosphere (Wanassi et al., 2017; Xia et al., 2020). In a previous study, we illus-trated that nZVI can be obtained through the carbothermal reduction of iron-rich sludge, where the iron salt in the sludge can be reduced to ZVI in a reducing atmosphere (Kong et al., 2016). Thereafter, the prepara-tion of graphite-supported nZVI through carbothermal reduction was reported, wherein graphite was immersed in an iron solution (Li et al., 2019).

We were inspired to explore a one-step carbothermal reduction process for the synthesis of carbon fiber (CF)-supported nZVI in the carbonization of waste cotton, wherein the porous CF may enhance the ability of nZVI to enrich pollutants through adsorption and the nZVI leads to efficient PMS activation. The enrichment of LEV on the CF surface through adsorption is hypothesized to facilitate LEV degradation by nZVI-activated PMS. The objective of this work was to synthesize a novel nZVI/CF by carbonizing for synergistic adsorption and degrada-tion of LEV. In this work, cotton fiber was soaked into Fe(NO3)3⋅9H2O solutions of various concentrations, and the solutions served as iron precursors. The synthesized nZVI/CF was characterized through X-ray diffraction (XRD), scanning electron microscopy (SEM), and X-ray

photoelectron spectroscopy (XPS) analyses. The iron concentration and morphology were adjusted by controlling the Fe3+ solution concentra-tion. The contribution of LEV enrichment through adsorption and PMS activation to LEV removal was investigated. The long-term stability of nZVI/CF was reported. This approach is expected to be a promising strategy for exploring the use of nZVI/CF to enrich and degrade LEV in the context of environmental decontamination.

2. Material and methods

2.1. Materials

Ferric nitrate (Fe(NO3)3⋅9H2O) was purchased from Sinopharm Chemical Reagent Co, Ltd. Levofloxacin (LEV), potassium monop-ersulfate triple salt (PMS), nitric acid (HNO3), sodium hydroxide (NaOH) and zero-valent iron (ZVI) powder were purchased from Aladdin Bio-Chem Technology Co., Ltd (Shanghai, China). Cotton fiber was collected from solid waste transfer station in Guangzhou University, China. All solutions were prepared with ultra-pure water.

2.2. Synthesis of Fe based catalysts

The Fe based catalysts were prepared via the following procedure. Initially, 4 g cotton fiber were immersed in ferric nitrate solutions at concentrations of 0.03 M, 0.1 M, 0.3 M, and 0.6 M by stirring for 24 h. In this case, iron ions could be adsorbed onto the cotton fiber. The iron- soaked cottons were put in a natural environment to remove bulk water, and then further being dried at an oven under 60 ◦C for 12 h. The obtained iron cotton with different iron contents were filled into a ceramic ark and placed in a quartz tube of the programmed tube elec-trical furnace (LTKC-5-12, China). Firstly, the tube was purged with nitrogen to sweep residual air to obtain an inert atmosphere. Subse-quently, the temperature was raised to 600, 700, 800, 900 ◦C with a heat rate of 5 ◦C/min, the iron cotton precursor was carbonized at the determined temperature for 2 h. Finally, the resulted materials were taken out after cooled to room temperature (Scheme 1). The resulted samples were named as nZVI/CF-T-s, in which the letters of T and s presented the carbonization temperature and iron concentration, respectively.

Scheme 1. Schematic illustration of the preparation of nZVI/CF.

W. Tan et al.

Chemosphere 280 (2021) 130626

3

2.3. Characterization

The chemical phase and surface morphology of nZVI/CF-T-s were observed by X-ray diffraction (XRD, PANalytical, PW3040/60, Holland) using Cu Kα at 40 kV and 40 mA in the range of 5–80◦. The micro-morphology of nZVI/CF-T-s was characterized by a scanning electron microscope (SEM, Zeiss, Gemini 500, Germany). The concentration of LEV was determined using Shimadzu-2500 UV–visible spectrophotom-eter at 288 nm. The concentration of fluoride ion was determined using fluoride electrode (Thermo Orion, 9609BNWP, USA). The surface element composition and valence were determined by X-ray photo-electron spectroscopy (XPS) technology (Thermo Fisher Scientific, ESCALAB 250Xi, USA). The specific surface areas and pore structure are determined by fully automatic specific surface area and pore size test system (Micromeritics, ASAP, 2020, USA) at 77 K through a nitrogen adsorption-desorption method. To detect reactive radicals, electron paramagnetic resonance (EPR) was operated during the LEV degrada-tion process using a spectrometer (Bruker, A300, Germany) with dimethyl-pyrroline-N-oxide (DMPO) as free radical tracer. Total organic carbon (TOC) was analyzed by a TOC analyzer (Shimadzu, TOC-L, JAPAN). The intermediate products were analyzed by liquid chromatography-mass spectrometry system (LC-MS, Thermo Fisher, Q- Exactive, USA).

2.4. Adsorption and degradation tests

All the adsorption and catalytic degradation experiments were per-formed in a series of 250 ml of conical flasks. Firstly, 100 ml of LEV of 10, 20, and 30 mg/L were poured into the conical flasks. The solution pH was adjusted by NaOH (0.1 M) and HNO3 (0.1 M) solution to determined pH value. 20 mg catalysts were poured into the LEV solution to initiate the reaction. Then the mixtures were put into a constant temperature shaker at 25 ◦C. For the adsorption test, 5 ml of mixtures were taken out at each time interval, further being filtrated by 0.45 μm millipore membrane to remove the residual catalysts. The residual concentration of LEV was measured using a UV–visible spectrophotometer (Shimadzu, UV-2500, Japan) at 288 nm wavelength. For the catalytic degradation experiments, 1 mM PMS was added to the mixed solution to trigger the degradation reaction. At each determined time interval, 5 ml of mixtures were taken out and filtrated immediately for further analyzing the re-sidual concentration of LEV. In this process, the concentration of fluo-ride ion in solution was measured by using fluoride electrode (Thermo Orion, 9609BNW, USA). The iron contents in nZVI/CF-900-s were measured by a flam atomic absorption spectrometer (Thermo Scientific, FAAS, iCE 3500 AA System, USA). To minimize the impact of errors, all the experimental tests were repeated three times.

3. Results and discussion

3.1. Characterization of nZVI/CF-T-s

Fig. 1 depicts the XRD patterns of nZVI/CF-T-s prepared through the carbonization of cotton fiber after immersed in iron solutions of various concentrations. nZVI/CF-600-0.3, nZVI/CF-700-0.3, and nZVI/CF-800- 0.3 had similar diffraction peaks. Diffraction peaks corresponding to FeO (JCPADS NO.89–0687) and Fe3O4 (JCPAD NO.75–1609) were observed at the 2θ values of 42◦, 53◦, and 60◦ for nZVI/CF-600-0.3, nZVI/CF-700-0.3, and nZVI/CF-800-0.3 (Dai et al., 2016). The cotton fiber adsorbed iron was converted into iron oxides through the trans-formation of Fe species into Fe3O4 and reduction to FeO in a reducing atmosphere (Jiang et al., 2019). Moreover, the diffraction peaks of Fe3C (JCPDS NO.76–1877) at 2θ values of 42.9◦, 43.7◦, and 45.9◦ were observed for nZVI/CF-900-0.03 and nZVI/CF-900-0.1. As the carbon-ization temperature and iron concentration increased, the intensities of the diffraction peaks corresponding to the iron oxides and Fe3C greatly decreased, and the peaks even disappeared. Notably, distinct diffraction

peaks attributable to metallic α-Fe (JCPDS NO.06–0696) were observed at 2θ values of 44.8◦ and 65.2◦, and they were indexed to (1 1 0), (2 0 0), and (2 1 1) (Li et al., 2019). These results indicated that the carbon-ization temperature governed nZVI production. This finding is consis-tent with those of our previous works, in which nZVI was produced through carbothermal reduction at 900 ◦C (Zhang et al., 2019a, 2019b). In the cases of nZVI/CF-900-0.03, nZVI/CF-900-0.1, nZVI/CF-900-0.3, and nZVI/CF-900-0.6, α-Fe (JCPDS NO.06–0696) was the main phase, confirming that nZVI can be produced through the carbothermal reduction of cotton fiber at 900 ◦C. However, the intensity of the diffraction peak of α-Fe for nZVI/CF-900-0.03 was low, whereas the intensities of the diffraction peaks of α-Fe for nZVI/CF-900-0.1, nZVI/CF-900-0.3, and nZVI/CF-900-0.6 were similar. The low in-tensity of nZVI/CF-900-0.03 may be ascribed to the fact that it was immersed in a solution with a low iron concentration of 0.03 M.

Fig. 2 demonstrated that nZVI/CF-900-0.03, nZVI/CF-900-0.1, nZVI/CF-900-0.3, and nZVI/CF-900-0.6 had different morphologies. Nanoparticles were dispersed on the CF surface of nZVI/CF-900-0.03. According to our XRD analysis, these nanoparticles were ascribed to nZVI. In the case of nZVI/CF-900-0.1, the ZVI nanoparticles started to agglomerate as the iron concentration of the immersion solution increased. In the case of nZVI/CF-900-0.3, nanoparticle agglomeration intensified when the iron concentration of the immersion solution was increased to 0.3 M, but internal gaps were observed. Nanoparticle agglomeration was observed even in the case of nZVI/CF-900-0.6. These results indicated that the increased iron concentration led to the agglomeration of ZVI nanoparticles. The narrow-scan XPS spectra of the Fe 2P peaks of nZVI/CF-900-0.3 are illustrated in Fig. 3. The peaks located at 706 and 718.6 eV represent the binding energies of nZVI, thus confirming the presence of nZVI. The calculated iron concentrations of nZVI/CF-900-0.03, nZVI/CF-900-0.1, nZVI/CF-900-0.3, and nZVI/CF- 900-0.6 were 229.8, 679.4, 831.2, and 849.4 mg/g, respectively, as illustrated in Fig. 4a reported in Text S1. These results suggest that the amount of nZVI produced could be adjusted conveniently by controlling the iron concentration of the immersion solution. According to Fig. 4a, the iron concentrations of nZVI/CF-900-0.3 and nZVI/CF-900-0.6 were equal when the immersed iron concentrations were increased to 0.3 and 0.6 M. That is to say, the soaked iron on cotton fiber is saturated when

Fig. 1. XRD pattern of nZVI/CF-T-s.

W. Tan et al.

Chemosphere 280 (2021) 130626

4

the iron concentration is 0.3 M. The nZVI content of nZVI/CF-900-0.6 is equal to that of nZVI/CF-900-0.3.

The N2 adsorption and desorption isotherms of nZVI/CF-T-s are depicted in Fig. S1, and the pore size distributions were calculated using

the Barrett–Joyner–Halenda (BJH) method. According to the Interna-tional Union of Pure Applied Chemistry classification, the N2 adsorption- desorption isotherms of nZVI/CF-900-s are of type IV and have a typical H2 hysteresis loop. At P/P0 ˃0.4, all of the samples exhibited rapid

Fig. 2. SEM images of (a) nZVI/CF-900-0.03, (b) nZVI/CF-900-0.1, (c) nZVI/CF-900-0.3, (d) nZVI/CF-900-0.6.

Fig. 3. XPS spectra of Fe 2p region of (a) nZVI/CF-600-0.3, (b)nZVI/CF-700-0.3, (c) nZVI/CF-800-0.3, (d) nZVI/CF-900-0.3.

W. Tan et al.

Chemosphere 280 (2021) 130626

5

adsorption owing to the filling of the mesopores (Table S1). The Bru-nauer–Emmett–Teller (BET) surface area was small due to the low micropore volume. Notably, the surface area of nZVI/CF-900-0.3 was larger than that of nZVI/CF-900-0.6, although they had similar iron concentrations. This was ascribed to the agglomeration of nZVI for nZVI/CF-900-0.6, as confirmed by SEM (Fig. 2).

3.2. Adsorption performance of nZVI/CF-T-s in relation to LEV

The adsorption performance levels of nZVI, CF-900, nZVI + CF-900, and nZVI/CF-900-0.3 in relation to LEV are presented in Fig. 4. More-over, nonlinear fitted kinetic models (pseudo-first-order and pseudo- second-order) were used to evaluate the adsorption behaviors of LEV on nZVI, CF-900, nZVI + CF-900, and nZVI/CF-900-0.3. As illustrated in Fig. 4b and c and summarized in Table 1, the results obtained using the pseudo-second-order kinetic model had a good fit with the experimental data pertaining to the adsorption of LEV on nZVI/CF-900-0.3. CF-900 exhibited a good LEV adsorption capacity, and LEV adsorption on

commercial Fe powder was negligible. These results indicated that the CF mainly contributed to LEV adsorption. However, nZVI + CF-900 exhibited a higher LEV adsorption efficiency than that of nZVI/CF-900- 0.3, although they had similar carbon and Fe concentration. The main difference between nZVI + CF-900 and nZVI/CF-900-0.3 was in the microstructures of nZVI and carbon. In nZVI/CF-900-0.3, the nZVI was loaded onto the carbon surface, whereas in nZVI + CF-900, the nZVI and carbon were separate. The loaded nZVI occupied the carbon surface, which may have limited the ability of nZVI/CF-900-0.3 to adsorb LEV. This result further confirmed the adsorption of LEV on carbon. Then, LEV could be enriched on the carbon surface, and by extension, on the nZVI surface as well. Carbonization temperature had a positive effect on the LEV adsorption capacity of nZVI/CF-T-s. nZVI/CF-900-0.3 exhibited the highest adsorption capacity among the nZVI/CF-T-0.3 samples (Fig. 4d) because high-temperature carbonization facilitated LEV enrichment on CF through adsorption.

3.3. Reactivity of nZVI/CF-T-s for PMS activation

To explore the reactivity of nZVI/CF-T-s for PMS activation, the degradation efficiency of LEV was investigated, as shown in Fig. 5. The LEV degradation efficiency of the nZVI/CF-T-s/PMS system was considerably higher than that of the PMS-only system (8.66%), indi-cating that nZVI/CF-T-s played an important role in PMS activation for LEV degradation. In particular, the carbonization temperature greatly affected LEV degradation. nZVI/CF-900-0.3 exhibited the highest LEV degradation efficiency of 93.83% within 60 min, whereas the LEV removal efficiencies of nZVI/CF-600-0.3, nZVI/CF-700-0.3, and nZVI/ CF-800-0.3 were 9.7%, 17.7%, and 25.6%, respectively. Our evalua-tions of the LEV degradation kinetics of the PMS/nZVI/CF-T-0.3 systems are presented in the inset of Fig. 5a. As described in the figure, the LEV degradation kinetics fitted well with the pseudo-first-order kinetic

Fig. 4. (a) The concentration of Fe in nZVI/CF-900-s; (b) pseudo-first-order and (c) pseudo-second-order adsorption kinetic models of nZVI, CF-900, nZVI + CF-900, nZVI/CF-900-0.3 to LEV; (d) adsorption capacity of LEV on nZVI/CF-T-s. Reaction conditions: [nZVI] = 0.16 g/L, [CF-900] = 0.04 g/L, [nZVI + CF-900] = 0.2 g/L, [nZVI/CF-900-0.3] = 0.2 g/L, [LEV] = 20 mg/L, T = 25 ◦C, initial pH = 7.0.

Table 1 Nonlinear fitted pseudo-first-order and pseudo-second-order models of LEV adsorption on nZVI, CF-900, nZVI + CF-900 and nZVI/CF-900-0.3.

Sample Pseudo-first-order Pseudo-second-order

k1

(min− 1) qe (mg/ g)

R2 k2

(min− 1) qe (mg/ g)

R2

nZVI 0.282 1.048 0.999 0.426 1.144 0.999 CF-900 0.194 58.660 0.995 3.70 ×

10− 3 67.870 0.999

nZVI + CF- 900

0.237 47.649 0.995 6.63 ×10− 3

53.366 0.999

nZVI/CF- 900-0.3

0.352 18.977 0.992 0.035 20.324 0.999

W. Tan et al.

Chemosphere 280 (2021) 130626

6

model (-ln (C/C0) = kt). The calculated kinetic rate constant of LEV degradation with nZVI/CF-600-0.3, nZVI/CF-700-0.3, nZVI/CF-800- 0.3, and nZVI/CF-900-0.3 activation was 0.0039, 0.0151, 0.0177, 0.0716 min− 1, respectively. nZVI/CF-900-0.3 achieved the strongest PMS activation for LEV degradation. Because iron oxides constituted the main phase in nZVI/CF-800-0.3, nZVI/CF-700-0.3, and nZVI/CF-600- 0.3, although nZVI was the main active components in nZVI/CF-900- 0.3, the strong reactivity of nZVI/CF-900-0.3 was ascribed to the pres-ence of nZVI, which has been widely used for PMS activation in the degradation of organic pollutants (Pang et al., 2019). The leaching of Fe2+ through nZVI corrosion, as expressed in reaction Eq. (1), was observed, and this process could provide an electron and Fe2+ for PMS activation (Fig. S2). The electron and Fe2+ could combine with HSO5

− to produce ⋅SO4

− and ⋅OH, as shown in the reaction equation, whereas Fe3+

could react with HSO5− to produce ⋅SO5

− and Fe2+ (Eqs. (2)–(5)), resulting in an increased LEV degradation efficiency.

Fe0 →Fe2+ + 2e− (1)

Fe2+ + HSO5− → Fe3+ + ⋅SO4

− + OH− (2)

Fe3+ + HSO5− → Fe2+ + ⋅SO5

− + H+ (3)

HSO5− + e− → ⋅SO4

− + OH− (4)

HSO5− + e− → SO4

2− + ⋅OH (5)

The nZVI/CF-900-s samples with various iron concentrations were obtained by adjusting the concentration of the Fe(NO3)3⋅9H2O solution. The LEV removal efficiency achieved through nZVI/CF-900-0.03, nZVI/ CF-900-0.1, nZVI/CF-900-0.3, and nZVI/CF-900-0.6 activation was 48.18%, 61.54%, 93.83%, and 87.95%, respectively (Fig. 5b). The rate constant of LEV degradation achieved through nZVI/CF-900-0.03,

nZVI/CF-900-0.1, nZVI/CF-900-0.3, and nZVI/CF-900-0.6 activation was 0.0208, 0.0263, 0.0716, and 0.0411 min− 1, respectively. The LEV degradation efficiency and rate constant of nZVI/CF-900-0.3 activation was the highest, indicating that nZVI/CF-900-0.3 exhibited a higher reactivity for PMS activation than did nZVI/CF-900-0.1 and nZVI/CF- 900-0.03. PMS activation for LEV degradation was enhanced as the Fe concentration increased. This result confirmed that nZVI played an important role in PMS activation for LEV degradation. Although the nZVI concentration of nZVI/CF-900-0.6 was similar to that of nZVI/CF- 900-0.3, its reactivity for PMS activation was lower than that of nZVI/ CF-900-0.3. nZVI/CF-900-0.6 was prepared through the immersion of cotton fiber in a 0.6 M iron solution. As shown in Fig. 4a, nZVI was agglomerated in nZVI/CF-900-0.6. The agglomerated nZVI could not efficiently activate PMS, which lowered the reactivity of PMS activation for LEV degradation.

To further confirm the reactivity of nZVI and carbon fiber in nZVI/ CF-900-0.3 for PMS activation, a comparison of the LEV removal effi-ciencies of the nZVI/CF-900-0.3/PMS, CF-900/PMS, nZVI/PMS, nZVI +CF-900/PMS, and PMS-only systems is depicted in Fig. 5c. CF-900 had a slightly affected on PMS activation. During adsorption (Fig. S3), the LEV removal rate of CF-900 (39.83%) was 14.34% lower than that of CF- 900/PMS (54.17%). This phenomenon can be ascribed to the fact that biochar produces persistent free radicals and electrons that react with PMS to produce ⋅SO4

− and ⋅OH (Fang et al., 2014; Jiang et al., 2019; Li et al., 2019). In addition, the LEV removal rates of nZVI, CF-900, nZVI +CF-900, and nZVI/CF-900-0.3 during the adsorption process were 1.05%, 39.83%, 40.27%, and 19.25%, respectively. The LEV degrada-tion efficiencies of the nZVI, nZVI + CF-900, and nZVI/CF-900-0.3 systems were up to 51.60%, 60.0%, and 93.83%, respectively, in the presence of PMS. The presence of nZVI strongly promoted LEV degra-dation by the nZVI/CF-900-0.3/PMS and nZVI + CF-900/PMS systems

Fig. 5. (a) LEV degradation efficiency of nZVI/CF-T-0.3 under different carbonization temperatures (inset: pseudo-first-order rate constant of LEV under different carbonization temperatures); (b) LEV degradation efficiency of nZVI/CF-900-s under different iron content (inset: pseudo-first-order rate constant of LEV under different irons content); (c) comparison for the degradation efficiency of LEV in various systems; (d) LEV degradation efficiency under different inorganic anions (inset: pseudo-first-order rate constant of LEV under different inorganic anions). Reaction conditions: [LEV] = 20 mg/L, [PMS] = 1 mM, [catalyst] = 0.2 g/L, T =25 ◦C, initial pH = 7.0.

W. Tan et al.

Chemosphere 280 (2021) 130626

7

compared with that by the CF-900/PMS and PMS systems, thus con-firming the efficient activity of nZVI in PMS activation. Although the nZVI/CF-900-0.3/PMS and nZVI + CF-900/PMS systems had similar nZVI concentrations, the nZVI/CF-900-0.3/PMS system exhibited significantly higher LEV degradation activity than the nZVI +

CF-900/PMS system did, indicating that the synthesized nZVI/CF-900-0.3/PMS system is a preferable PMS activator to CF-900/nZVI. Because nZVI + CF-900 is a mixture of CF-900 and nZVI, its components became separated in the solution. CF-900 exhibited a favorable adsorption ability to remove LEV, and nZVI provided the PMS activation ability to degrade LEV. The nZVI + CF-900 system exhibited a higher LEV adsorption efficiency than did the nZVI/CF-900-0.3 system, as shown in Fig. S3. This result may be ascribed to the fact that CF-900 played an important role in LEV adsorption, whereas the micropores on the CF of nZVI/CF-900-0.3 were blocked by the nano-sized ZVI. LEV could have been concentrated around the CF-900 surface through adsorption in the nZVI + CF-900/PMS and nZVI/CF-900-0.3/PMS sys-tems. Because free radicals were generated around the nZVI particles through PMS activation, these radicals were easily annihilated within several nanoseconds and could not attack the concentrated LEV around the CF-900 surface because the CF-900 and nZVI were separated in the nZVI + CF-900/PMS system (Xu et al., 2020). Once the nZVI was loaded on the CF in the nZVI/CF-900-0.3 system, the nZVI/CF-900-0.3 system not only exhibited advantages in enriching LEV through adsorption but also strongly activated PMS for LEV degradation, with an efficiency of 93.83%. This was ascribed to the fact that once the LEV molecules were enriched on the CF surface through adsorption, they were immediately degraded by the ⋅SO4

− and ⋅OH radicals. Thus, nZVI/CF-900-0.3 exhibited a preferable ability to remove LEV. Additionally, a high defluorination efficiency of 64.25% was observed for the nZVI/CF-900-0.3 system (Fig. S4), which indicated that the fluorine could be released from small organic molecules, which could reduce the toxicity of the intermediate product. The TOC removal efficiency of the nZVI/CF-900-0.3 system was up to 48.61%, which was higher than the values reported in previous studies (Pi et al., 2018, 2020) (Fig. S5). The TOC removal efficiency is consistent with the defluorination efficiency, indicating that LEV mineralization leads to defluorination.

An evaluation of the effects of anions such as Cl− , NO3− , HCO3

− , H2PO4

− on the LEV degradation efficiencies is depicted in Fig. 5d. The concentration of each anion was 10 mM. The results indicate that the influence of anions on the degradation efficiency of LEV has the following order: H2PO4

− ˃ HCO3− ˃ NO3

− ˃ Cl− . The LEV degradation effi-ciency decreased to 35.24% and the pH increased to 8.72 after the addition of HCO3

− . The alkaline solution may have inhibited ⋅SO4− pro-

duction. More importantly, HCO3− can strongly quench ⋅OH and ⋅SO4

− , as shown in Eqs. (6) and (7) (Ahmadi and Ghanbari, 2019). Thus, HCO3

−

limited PMS activation, leading to a decrease in LEV degradation effi-ciency (Eq. (13)). Similarly, H2PO4

− can react with ⋅SO4− and ⋅OH radicals

(Eqs. (8) and (9)) to limit LEV degradation. These results indicate that ⋅SO4

− and ⋅OH radicals contributed to LEV degradation. The rate con-stants of the reactions of HCO3

− and H2PO4− with PMS were close, and the

inhibition of these reactions was similar. NO3− marginally affected the

degradation of LEV because only ⋅SO4− could react with NO3

− (Eq. (12)). Although Cl− could consume ⋅SO4

− and ⋅OH radicals (Eqs. (10) and (11)), Cl⋅ and Cl2⋅- were produced through a reaction Eq. (14), and they could attack the unsaturated C = C bond in the LEV molecule (Yang et al., 2014), leading to a similar LEV degradation efficiency as that in the presence of NO3

− . More importantly, LEV degradation was slightly inhibited at 10 mM Cl− and NO3

− , suggesting the potential for applica-tion of nZVI/CF/PMS to low-salinity wastewater.

HCO3− + ⋅SO4

− → SO42− +⋅CO3

− + H+, k = 1.6–9.1 × 106 M− 1 s− 1 (6)

HCO3− + ⋅OH → H2O + ⋅CO3

− , k = 8.5–10 × 106 M− 1 s− 1 (7)

H2PO4− + ⋅SO4

− → SO42− + ⋅HPO4

− + H+, k = 1.2–16 × 106 M− 1 s− 1 (8)

H2PO4− +⋅OH → H2O + ⋅HPO4

− , k = 2.0–15 × 104 M− 1 s− 1 (9)

Cl− +⋅SO4− → SO4

2− +⋅Cl, k = 1.3–6.6 × 108 M− 1 s− 1 (10)

Cl− + ⋅OH → ClOH− , k = 3–4.3 × 109 M− 1 s− 1 (11)

NO3− + ⋅SO4

− → SO42− + ⋅NO3

− , k = 5–210 × 104 M− 1 s− 1 (12)

HSO5− + HCO3

− → HCO4− + HSO4

− (13)

Cl⋅ + Cl− → Cl ⋅−2 (14)

An evaluation of the effects of PMS concentration, initial pH, nZVI/ CF-900-0.3 dosage and LEV concentration on LEV degradation effi-ciency is depicted in Fig. 6. Because PMS is a source of sulfate radicals, the addition of PMS promoted LEV degradation. Fig. 6a demonstrates the influence of PMS concentration on LEV degradation efficiency. Approximately 89.04%, 93.83%, and 92.36% of LEV was degraded at the PMS concentrations of 0.5 mM, 1.0 mM, and 1.5 mM, respectively. The LEV degradation efficiency increased as the PMS concentration increased up to 1.0 mM, whereas it decreased when the PMS concen-tration reached 1.5 mM. Excessive PMS had a detrimental effect on LEV degradation due to the competitive reactions of free radicals, as shown in Eqs. (15)–(17) (Guan et al., 2011; Jiang et al., 2019; Takdastan et al., 2018). The generated ⋅SO4

− was consumed, and ⋅HO2 with low reactivity was produced (Eq. (18)) in the presence of excess PMS (PMS concen-tration of 1.5 mM in this work) (Ghanbari et al., 2019, 2020).

⋅SO4− + ⋅SO4

− →2SO 2−4 or S2O2−

8 (15)

⋅OH + ⋅OH → H2O2 (16)

⋅SO4− + ⋅OH → HSO −

5 (17)

HSO−5 + ⋅SO−

4 + H2O→⋅HO2 +2SO2−4 + 2H+ (18)

To investigate the effect of initial pH on LEV degradation through nZVI/CF-900-0.3 activation, the LEV degradation efficiency with the initial solution pH of 3.0–11.0 is depicted in Fig. 6b. Almost 93.83% of the LEV was removed within 60 min over a wide pH range of 5.0–9.0. However, only 82.17% of the LEV was removed when the solution pH was 3. Moreover, the LEV degradation efficiency decreased when the pH was 11.0. The highest removal efficiency was obtained at a pH of 7.0. The degradation kinetic constant of LEV was the highest when the so-lution pH was 7.0. Sulfate and hydroxyl radicals mainly contribute to the degradation of organic pollutants. The main active sulfate radical is more abundant under neutral conditions (Noroozi et al., 2020). Thus, the LEV degradation activity of nZVI/CF-900-0.3 for PMS activation was the strongest under a neutral condition. In alkaline conditions, ⋅OH and ⋅SO4

− could react with hydroxide ions (OH− ), thus causing the con-sumption of free radicals [Eqs. (19)–(22)] (Ma et al., 2020) Hydroxide is adsorbed on the surface of the catalyst under alkaline conditions, which induces a negative charge on the catalyst surface; this is inconducive to the binding of a neutral or electron-rich form of LEV (Kaur et al., 2019). Moreover, the released Fe2+ could react with hydroxyl groups (OH− ) in the alkaline solution to produce iron hydroxide, thus reducing PMS activation by Fe2+. Generally, ⋅OH is easily produced in acidic solutions through the Fenton reaction, but the degradation efficiency and kinetic constant rate are greatly reduced at a pH solution of >3.0. The excessive H+ could react with ⋅OH in the presence of electrons to form H2O ac-cording to reaction Eq. (23) (Cai et al., 2020), leading to the consump-tion of electrons and ⋅OH. Moreover, it has been widely reported that ⋅SO4

− could be consumed by H+ in an acidic solution [Eq. (24)] (Cai et al., 2020), leading to a decrease in the LEV degradation efficiency. Thus, ⋅SO4

− mainly contributed to LEV degradation. LEV degradation using PMS activated by nZVI/CF-900-0.3 could be achieved over a wide pH range, indicating that nZVI/CF-900-0.3 is promising for use in PMS activation in environmental remediation.

⋅SO4− + OH− → ⋅OH + SO 2−

4 (19)

W. Tan et al.

Chemosphere 280 (2021) 130626

8

⋅SO4− + H2O → HSO4

− + ⋅OH (20)

⋅OH + ⋅OH →H2O2 (21)

S2O 2−8 + H2O2 →2H+ + 2SO 2−

4 + O2 (22)

H+ + ⋅OH + e− →H2O (23)

H+ + ⋅SO−4 + e− →HSO−

4 (24)

The evolution of solution pH values after the reaction is displayed in Fig. S6. When the initial solution was acidic with a pH of 3.0, the final environment remained acidic (pH = 3.71). When the initial solution pH was 5.0, 7.0, and 9.0, the final pH values were 5.12, 5.14, and 5.13, respectively. A similar result was obtained by Jiang et al. (2019). One explanation was the buffering ability of biochar (Hu et al., 2018).

The LEV degradation efficiency increased to 89.34% after only 10 min as nZVI/CF-900-0.3 dosage increased from 0.1 to 0.3 g/L (Fig. 6c). As shown in the inset image, the k value increased with the nZVI content in the nZVI/CF-900-0.3/PMS system. The high nZVI content favored high yields of free radicals through PMS activation, resulting in high LEV degradation efficiency. A comparison of the LEV degradation efficiency between the nZVI/CF/PMS systems and the reported system is presented in Table S2. And LEV degradation efficiency of 93.83% was achieved within 60 min in the case of the nZVI/CF-900-0.3/PMS system, which was the highest among the studied systems, especially at an initial LEV concentration and catalyst concentration of 20 ppm and 0.2 g/L, respectively. Although a removal efficiency of 89.4% was achieved with the CoFeO2@CN system, the initial concentration was only 10 ppm (Pi et al., 2020). The nZVI/CF-900-0.3/PMS system benefited from its high LEV removal efficiency and low catalyst dosage compared with the tested system. The favorable results may be ascribed to the

advantageous activity of nZVI/CF. In addition, an investigation of the influence of the initial LEV concentration on the LEV degradation effi-ciency is depicted in Fig. 6d, which indicates that 10 ppm LEV was almost removed within 40 min. As the LEV concentration was increased from 20 ppm to 30 ppm, the degradation efficiency decreased from 93.83% to 88.88%. The rate constants for LEV degradation at the con-centrations of 10, 20, and 30 ppm was 0.0380, 0.0716, and 0.0211 min− 1, respectively.

Free radical quenching experiments were conducted, as depicted in Fig. 7, to identify the radicals generated during PMS activation by nZVI/ CF-T-s. Ethanol (EtOH) and tert-butanol (TBA) were used as the free radical quenchers. EtOH has the capacity to quench ⋅SO4

− at a rate constant of 1.6–1.7 × 107 M− 1 s− 1, whereas TBA, which was used as an ⋅OH scavenger, displayed more advanced reactivity with ⋅OH ((3.8–7.6) × 108 M− 1 s− 1) than with ⋅SO4

− ((4–9.1) × 105 M− 1 s− 1) (Pi et al., 2020). LEV degradation in the presence of EtOH and TBA is illustrated in Fig. 7a. When TBA was added into the mixed system of nZVI/CF-900-0.3/PMS, the LEV removal efficiency slightly decreased. However, the LEV removal efficiency was significantly inhibited to 28.67% within 60 min when 5 M EtOH was added to the system. ⋅SO4

−

was the main radical that contributed to LEV degradation in the nZVI/CF-900-0.3/PMS system. The rate constants of k(No scavenger), k(EtOH), and k(TBA) were calculated to be 0.0716, 0.0153, and 0.0303 min− 1, respectively (Fig. S7), confirming that both ⋅OH and ⋅SO4

−

contributed to LEV degradation. The results of an EPR test are addi-tionally displayed in Fig. 7b to confirm the produced radicals. When DMPO was added into the reactive solution in a typical ratio of 1:2:2:1 (αH = αN = 14.8), it confirmed the existence of ⋅OH. DMPO-⋅SO4

− adducts (αH = 0.78, αH = 1.48, αH = 9.6, and αN = 13.2) and ⋅SO4

− (Ding et al., 2019; Li et al., 2019). The results further confirmed that although both ⋅SO4

− and ⋅OH were generated in the nZVI/CF-900-0.3/PMS system, the ⋅SO4

− mainly contributed to LEV degradation.

Fig. 6. Effects of reaction parameters on the degradation of LEV: (a) PMS doses (inset: pseudo-first-order rate constant of LEV under different PMS doses.); (b) pH conditions (inset: pseudo-first-order rate constant of LEV under different pH conditions); (c) catalyst dosages (inset: pseudo-first-order rate constant of LEV under different catalyst dosages); (d) initial concentration (inset: pseudo-first-order rate constant of LEV under different initial concentration). Reaction conditions: [LEV] = 20 mg/L (if needed), [PMS] = 1 mM (if needed), [catalyst] = 0.2 g/L (if needed), T = 25 ◦C, initial pH = 7.0 (if needed).

W. Tan et al.

Chemosphere 280 (2021) 130626

9

Given that the stability of nZVI is important for its application, the stability of nZVI/CF-900-0.3 for PMS activation was measured, as shown in Fig. 8, after storage for 6 months. The stored nZVI/CF-900-0.3 system exhibited favorable PMS activation for LEV degradation, thus confirm-ing its excellent long-term stability. This indicates that the nZVI embedded in the biochar micropores did not easily come into contact with air, which delayed the formation of an oxidation layer on the nZVI surface and preserved its high reactivity (Uzum et al., 2009; Wang et al., 2013). Moreover, the nZVI/CF-900-0.3 system exhibited strong magnetism, meaning that it could be separated easily using an external magnet (as shown in the inset of Fig. 8). The desirable long-term sta-bility of the catalyst activity of nZVI/CF-900-0.3 can be ascribed to the electron-rich carbon that prevents the oxidation of nZVI. Thus, nZVI/CF-900-0.3 is promising system for use in environmental reme-diation applications.

3.4. Possible degradation pathways of LEV for nZVI/CF-900-0.3

To explore possible LEV degradation pathways, the intermediate products were detected using LC-MS. Several degradation intermediates of LEV in the nZVI/CF-900-0.3/PMS system were found, as summarized in Table S3. On the basis of these results, one of the degradation path-ways of LEV in the nZVI/CF-900.-0.3/PMS system was proposed

(Scheme 2). First, LEV molecules (m/z = 362) are dehydrogenated to m/ z 344 (Xu et al., 2017). Then, nonselective ⋅SO4

− and ⋅OH may attack piperazine nitrogen atoms, leading to the formation of m/z 336 through demethylation (Ma et al., 2019). The peak at m/z 279 was obtained by de-piperazinly of the product at a peak of m/z 336 (Xu et al., 2017). The peak at m/z 261 was produced through defluorination of the product corresponding to m/z 279 (Xu et al., 2017). Finally, all of the interme-diate products were converted into other micromolecules of CO2 and H2O as the reaction time of the nZVI/CF-900-0.3/PMS system increased.

Thus, the strong catalytic performance of nZVI/CF-900-0.3 in acti-vating PMS can be ascribed to the following mechanism: firstly, LEV molecules are absorbed and enriched on the nZVI/CF surface. Then, nZVI provides an electron and Fe2+ for activating PMS to generate ⋅SO4

−

and ⋅OH. These radicals attack the LEV molecules, which are eventually transformed into smaller molecules.

4. Conclusion

In this study, nZVI/CF-900-0.3 was prepared with a one-step car-bothermal reduction process in an inert gas atmosphere. nZVI was successfully supported on cotton carbon fiber. The resulting nZVI/CF- 900-0.3 exhibited a high LEV degradation efficiency in the presence of PMS. Under a neutral condition, nearly 93.83% of the LEV with an initial concentration of 20 ppm was removed in the presence of nZVI/CF-900- 0.3 with a dosage of 0.2 g/L and PMS concentration of 1 mM within 60 min. Both ⋅SO4

− and ⋅OH contributed to LEV degradation, but ⋅SO4− was

the main component causing LEV degradation. Moreover, the proposed catalyst exhibited excellent long-term stability after storage for 6 months. Thus, we have provided a promising and stable nZVI/CF-900- 0.3 synthesized through carbothermal reduction for eliminating LEV from the environment.

Credit author statement

Weitong Tan: Conceptualization, Methodology, Data curation, Investigation, Visualization, Writing - original draft preparation. Yang Ruan: Resources, Software, Writing - review & editing. Zenghui Diao: Writing - review & editing. Gang Song: Writing - review & editing. Minhua Su: Writing - review & editing. Li’an Hou: Writing - review & editing. Diyun Chen: Writing - review & editing. Hongmei Deng: En-glish writing & editing. Lingjun Kong: Conceptualization, Data cura-tion, Funding acquisition, Supervision.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence

Fig. 7. (a) Effect of radical scavengers on LEV degradation (b) EPR test results in nZVI/CF-900-0.3/PMS systems. Reaction conditions: [scavenger] = 5 M, [LEV] =20 mg/L, [PMS] = 1 mM, [catalyst] = 0.2 g/L, T = 25 ◦C, initial pH = 7.0.

Fig. 8. LEV removal efficiency of nZVI/CF-900-0.3 after 6 months. (inset: the magnetic separation of the nZVI/CF-900-0.3 dispersed in LEV). Reaction con-ditions: [LEV] = 20 mg/L, [PMS] = 1 mM, [catalyst] = 0.2 g/L, T = 25 ◦C, initial pH = 7.0.

W. Tan et al.

Chemosphere 280 (2021) 130626

10

the work reported in this paper.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (21976042), The Project of Guangdong Provincial Key Labo-ratory of radioactive contamination control and resources (2017B030314182), Science and Technology Planning Project of Guangdong Province, China (2021A0505030076), Guangdong Key Laboratory of Environmental Catalysis and Health Risk Control (2018B030322014), Guangdong Basic and Applied Basic Research Foundation (2019A1515011543), Guangdong Province Universities and Colleges Pearl River Scholar Funded Scheme (2018, 2019), University scientific research project of Guangzhou Education Bureau (201831803), The research project of Guangzhou University (YK2020012).

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi. org/10.1016/j.chemosphere.2021.130626.

References

Ahmadi, M., Ghanbari, F., 2019. Organic dye degradation through peroxymonosulfate catalyzed by reusable graphite felt/ferriferrous oxide: mechanism and identification of intermediates. Mater. Res. Bull. 111, 43–52.

Cai, C., Zhang, H., Zhong, X., Hou, L., 2015. Ultrasound enhanced heterogeneous activation of peroxymonosulfate by a bimetallic Fe-Co/SBA-15 catalyst for the degradation of Orange II in water. J. Hazard Mater. 283, 70–79.

Cai, H.Y., Li, J., Yin, H.X., Yao, G., Lai, B., 2020. Degradation of atrazine in aqueous solution through peroxymonosulfate activated by Co-modified nano-titanium dioxide. Water Environ. Res. 92 (9), 1363–1375.

Cao, J.Y., Lai, L.D., Lai, B., Yao, G., Chen, X., Song, L.P., 2019. Degradation of tetracycline by peroxymonosulfate activated with zero-valent iron: performance, intermediates, toxicity and mechanism. Chem. Eng. J. 364, 45–56.

Dai, Y., Hu, Y.C., Jiang, B.J., Zou, J.L., Tian, G.H., Fu, H.G., 2016. Carbothermal synthesis of ordered mesoporous carbon-supported nano zero-valent iron with enhanced stability and activity for hexavalent chromium reduction. J. Hazard Mater. 309, 249–258.

Ding, M.M., Chen, W., Xu, H., Shen, Z., Lin, T., Hu, K., Kong, Q., Yang, G., Xie, Z.L., 2019. Heterogeneous Fe2CoTi3O10-MXene composite catalysts: synergistic effect of the ternary transition metals in the degradation of 2,4-dichlorophenoxyacetic acid based on peroxymonosulfate activation. Chem. Eng. J. 378, 122177.

Dong, H.R., Deng, J.M., Xie, Y.K., Zhang, C., Jiang, Z., Cheng, Y.J., Hou, K.J., Zeng, G.M., 2017. Stabilization of nanoscale zero-valent iron (nZVI) with modified biochar for Cr (VI) removal from aqueous solution. J. Hazard Mater. 332, 79–86.

Du, Q., Zhang, S.J., Pan, B.C., Lv, L., Zhang, W.M., Zhang, Q.X., 2014. Effect of spatial distribution and aging of ZVI on the reactivity of resin-ZVI composites for arsenite removal. J. Mater. Sci. 49 (20), 7073–7079.

Fang, G.D., Gao, J., Liu, C., Dionysiou, D.D., Wang, Y., Zhou, D.M., 2014. Key role of persistent free radicals in hydrogen peroxide activation by biochar: implications to organic contaminant degradation. Environ. Sci. Technol. 48 (3), 1902–1910.

Furman, O.S., Teel, A.L., Watts, R.J., 2010. Mechanism of base activation of persulfate. Environ. Sci. Technol. 44 (16), 6423–6428.

Ghanbari, F., Ahmadi, M., Gohari, F., 2019. Heterogeneous activation of peroxymonosulfate via nanocomposite CeO2-Fe3O4 for organic pollutants removal: the effect of UV and US irradiation and application for real wastewater. Separ. Purif. Technol. 228, 115732.

Ghanbari, F., Khatebasreh, M., Mahdavianpour, M., Lin, K.Y.A., 2020. Oxidative removal of benzotriazole using peroxymonosulfate/ozone/ultrasound: synergy, optimization, degradation intermediates and utilizing for real wastewater. Chemosphere 244, 125326.

Guan, Y.H., Ma, J., Li, X.C., Fang, J.Y., Chen, L.W., 2011. Influence of pH on the formation of sulfate and hydroxyl radicals in the UV/peroxymonosulfate system. Environ. Sci. Technol. 45 (21), 9308–9314.

Hao, M.J., Qiu, M.Q., Yang, H., Hu, B.W., Wang, X.X., 2021. Recent advances on preparation and environmental applications of MOF-derived carbons in catalysis. Sci. Total Environ. 760, 143333.

Hu, L.M., Zhang, G.S., Liu, M., Wang, Q., Wang, P., 2018. Enhanced degradation of Bisphenol A (BPA) by peroxymonosulfate with Co3O4-Bi2O3 catalyst activation: effects of pH, inorganic anions, and water matrix. Chem. Eng. J. 338, 300–310.

Jiang, S.F., Ling, L.L., Chen, W.J., Liu, W.J., Li, D.C., Jiang, H., 2019. High efficient removal of bisphenol A in a peroxymonosulfate/iron functionalized biochar system: mechanistic elucidation and quantification of the contributors. Chem. Eng. J. 359, 572–583.

Kaur, M., Umar, A., Mehta, S.K., Kansal, S.K., 2019. Reduced graphene oxide-CdS heterostructure: an efficient fluorescent probe for the sensing of Ag(I) and sunset yellow and a visible-light responsive photocatalyst for the degradation of levofloxacin drug in aqueous phase. Appl. Catal. B Environ. 245, 143–158.

Kong, L.J., Zhu, Y.T., Liu, M.X., Chang, X.Y., Xiong, Y., Chen, D.Y., 2016. Conversion of Fe-rich waste sludge into nano-flake Fe-SC hybrid Fenton-like catalyst for degradation of AOII. Environ. Pollut. 216, 568–574.

Li, M.Q., Luo, R., Wang, C.H., Zhang, M., Zhang, W.X., Klu, P.K., Yan, Y.B., Qi, J.W., Sun, X.Y., Wang, L.J., Li, J.S., 2019. Iron-tannic modified cotton derived Fe-0/ graphitized carbon with enhanced catalytic activity for bisphenol A degradation. Chem. Eng. J. 372, 774–784.

Scheme 2. Possible degradation pathway of LEV in nZVI/CF-900-0.3/PMS system.

W. Tan et al.

Chemosphere 280 (2021) 130626

11

Li, Z.T., Wang, L., Meng, J., Liu, X.M., Xu, J.M., Wang, F., Brookes, P., 2018. Zeolite- supported nanoscale zero-valent iron: new findings on simultaneous adsorption of Cd(II), Pb(II), and As(III) in aqueous solution and soil. J. Hazard Mater. 344, 1–11.

Liu, G.S., Zhou, Y.N., Teng, J., Zhang, J.N., You, S.J., 2018. Visible-light-driven photocatalytic activation of peroxymonosulfate by Cu-2(OH)PO4 for effective decontamination. Chemosphere 201, 197–205.

Liu, X., Pang, H., Liu, X., Li, Q., Zhang, N., Mao, L., Qiu, M., Hu, B., Yang, H., Wang, X., 2021. Orderly porous covalent organic frameworks-based materials: superior adsorbents for pollutants removal from aqueous solutions. Innovation 2 (1), 100076.

Lyu, H.H., Tang, J.C., Cui, M.K., Gao, B., Shen, B.X., 2020. Biochar/iron (BC/Fe) composites for soil and groundwater remediation: synthesis, applications, and mechanisms. Chemosphere 246, 125609.

Ma, Q.L., Nengzi, L.C., Li, B., Wang, Z.J., Liu, L.C., Cheng, X.W., 2020. Heterogeneously catalyzed persulfate with activated carbon coated with CoFe layered double hydroxide (AC@CoFe-LDH) for the degradation of lomefloxacin. Separ. Purif. Technol. 235, 116240.

Ma, Q.L., Zhang, H.X., Zhang, X.Y., Li, B., Guo, R.N., Cheng, Q.F., Cheng, X.W., 2019. Synthesis of magnetic CuO/MnFe2O4 nanocompisite and its high activity for degradation of levofloxacin by activation of persulfate. Chem. Eng. J. 360, 848–860.

Ma, Q.Q., Cui, L., Zhou, S., Li, Y., Shi, W.J., Ai, S.Y., 2018. Iron nanoparticles in situ encapsulated in lignin-derived hydrochar as an effective catalyst for phenol removal. Environ. Sci. Pollut. Res. 25 (21), 20833–20840.

Mahmoud, M.E., El-Ghanam, A.M., Mohamed, R.H.A., Saad, S.R., 2020. Enhanced adsorption of Levofloxacin and Ceftriaxone antibiotics from water by assembled composite of nanotitanium oxide/chitosan/nano-bentonite. Mat. Sci. Eng. C. 108, 110199.

Nie, M.H., Yang, Y., Zhang, Z.J., Yan, C.X., Wang, X.N., Li, H.J., Dong, W.B., 2014. Degradation of chloramphenicol by thermally activated persulfate in aqueous solution. Chem. Eng. J. 246, 373–382.

Noroozi, R., Gholami, M., Farzadkia, M., Jafari, A.J., 2020. Degradation of ciprofloxacin by CuFe2O4/GO activated PMS process in aqueous solution: performance, mechanism and degradation pathway. Int. J.Environ. Anal. Chem. 1, 1–22.

Pang, Y.X., Ruan, Y., Feng, Y., Diao, Z.H., Shih, K.M., Hou, L.A., Chen, D.Y., Kong, L.J., 2019. Ultrasound assisted zero valent iron corrosion for peroxymonosulfate activation for Rhodamine-B degradation. Chemosphere 228, 412–417.

Petala, E., Dimos, K., Douvalis, A., Bakas, T., Tucek, J., Zboril, R., Karakassides, M.A., 2013. Nanoscale zero-valent iron supported on mesoporous silica: characterization and reactivity for Cr(VI) removal from aqueous solution. J. Hazard. Mater. 261, 295–306.

Pi, Y.Q., Gao, H.Q., Cao, Y.D., Cao, R.L., Wang, Y.B., Sun, J.H., 2020. Cobalt ferrite supported on carbon nitride matrix prepared using waste battery materials as a peroxymonosulfate activator for the degradation of levofloxacin hydrochloride. Chem. Eng. J. 379, 122377.

Pi, Y.Q., Ma, L.H., Zhao, P., Cao, Y.D., Gao, H.Q., Wang, C.F., Li, Q.L., Dong, S.Y., Sun, J. H., 2018. Facile green synthetic graphene-based Co-Fe Prussian blue analogues as an activator of peroxymonosulfate for the degradation of levofloxacin hydrochloride. J. Colloid Interf. Sci. 526, 18–27.

Takdastan, A., Kakavandi, B., Azizi, M., Golshan, M., 2018. Efficient activation of peroxymonosulfate by using ferroferric oxide supported on carbon/UV/US system: a new approach into catalytic degradation of bisphenol A. Chem. Eng. J. 331, 729–743.

Tanaka, S., Kaneti, Y.V., Septiani, N.L.W., Dou, S.X., Bando, Y., Hossain, M.S.A., Kim, J., Yamauchi, Y., 2019. A review on iron oxide-based nanoarchitectures for biomedical, energy storage, and environmental applications. Small Methods 3 (5).

Uzum, C., Shahwan, T., Eroglu, A.E., Hallam, K.R., Scott, T.B., Lieberwirth, I., 2009. Synthesis and characterization of kaolinite-supported zero-valent iron nanoparticles and their application for the removal of aqueous Cu2+ and Co2+ ions. Appl. Clay Sci. 43 (2), 172–181.

Van Doorslaer, X., Dewulf, J., Van Langenhove, H., Demeestere, K., 2014. Fluoroquinolone antibiotics: an emerging class of environmental micropollutants. Sci. Total Environ. 500, 250–269.

Wanassi, B., Ben Hariz, I., Ghimbeu, C.M., Vaulot, C., Ben Hassen, M., Jeguirim, M., 2017. Carbonaceous adsorbents derived from textile cotton waste for the removal of Alizarin S dye from aqueous effluent: kinetic and equilibrium studies. Environ. Sci. Pollut. Res. 24 (11), 10041–10055.

Wang, J., Shen, M., Gong, Q., Wang, X.H., Cai, J.Y., Wang, S.L., Chen, Z.Q., 2020. One- step preparation of ZVI-sludge derived biochar without external source of iron and its application on persulfate activation. Sci. Total Environ. 714, 136728.

Wang, R.Z., Huang, D.L., Liu, Y.G., Zhang, C., Lai, C., Wang, X., Zeng, G.M., Gong, X.M., Duan, A., Zhang, Q., Xu, P., 2019. Recent advances in biochar-based catalysts: properties, applications and mechanisms for pollution remediation. Chem. Eng. J. 371, 380–403.

Wang, S., Gao, B., Li, Y., Creamer, A.E., He, F., 2017. Adsorptive removal of arsenate from aqueous solutions by biochar supported zero-valent iron nanocomposite: batch and continuous flow tests. J. Hazard. Mater. 322, 172–181.

Wang, T., Su, J., Jin, X.Y., Chen, Z.L., Megharaj, M., Naidu, R., 2013. Functional clay supported bimetallic nZVI/Pd nanoparticles used for removal of methyl orange from aqueous solution. J. Hazard. Mater. 262, 819–825.

Xia, M.L., Shao, X.H., Sun, Z.H., Xu, Z.H., 2020. Conversion of cotton textile wastes into porous carbons by chemical activation with ZnCl2, H3PO4, and FeCl3. Environ. Sci. Pollut. Res. 27, 25186–25196.

Xu, K., Ben, W.W., Ling, W.C., Zhang, Y., Qu, J.H., Qiang, Z.M., 2017. Impact of humic acid on the degradation of levofloxacin by aqueous permanganate: kinetics and mechanism. Water Res. 123, 67–74.

Xu, L., Fu, B.R., Sun, Y., Jin, P.K., Bai, X., Jin, X., Shi, X., Wang, Y., Nie, S.T., 2020. Degradation of organic pollutants by Fe/N co-doped biochar via peroxymonosulfate activation: synthesis, performance, mechanism and its potential for practical application. Chem. Eng. J. 400, 125870.

Xu, Z.H., Zhang, T.Q., Yuan, Z.H., Zhang, D.F., Sun, Z.H., Huang, Y.X., Chen, W.F., Tian, D.Q., Deng, H.X., Zhou, Y.W., 2018. Fabrication of cotton textile waste-based magnetic activated carbon using FeCl3 activation by the Box-Behnken design: optimization and characteristics. RSC Adv. 8 (66), 38081–38090.

Yang, S.S., Wu, P.X., Liu, J.Q., Chen, M.Q., Ahmed, Z., Zhu, N.W., 2018. Efficient removal of bisphenol A by superoxide radical and singlet oxygen generated from peroxymonosulfate activated with Fe-0-montmorillonite. Chem. Eng. J. 350, 484–495.

Yang, Y., Pignatello, J.J., Ma, J., Mitch, W.A., 2014. Comparison of halide impacts on the efficiency of contaminant degradation by sulfate and hydroxyl radical-based advanced oxidation processes (AOPs). Environ. Sci. Technol. 48 (4), 2344–2351.

Zhang, B., Wang, D.P., 2019. Preparation of biomass activated carbon supported nanoscale zero-valent iron (nzvi) and its application in decolorization of methyl orange from aqueous solution. Water-Sui 11 (8).

Zhang, D.J., Li, Y., Sun, A.W., Tong, S.Q., Su, G.Y., Jiang, X.B., Li, J.S., Han, W.Q., Sun, X. Y., Wang, L.J., Shen, J.Y., 2019a. Enhanced nitrobenzene reduction by modified biochar supported sulfidated nano zerovalent iron: comparison of surface modification methods. Sci. Total Environ. 694, 133701.

Zhang, H.M., Ruan, Y., Liang, A.P., Shih, K.M., Diao, Z.H., Su, M.H., Hou, L.A., Chen, D. Y., Lu, H., Kong, L.J., 2019b. Carbothermal reduction for preparing nZVI/BC to extract uranium: insight into the iron species dependent uranium adsorption behavior. J. Clean. Prod. 239, 117873.

Zhang, M., Wang, Y.C., 2020. Effects of Fe-Mn-modified biochar addition on anaerobic digestion of sewage sludge: biomethane production, heavy metal speciation and performance stability. Bioresour. Technol. 313, 123695.

Zhang, X.R., Zhang, Z.X., Li, Y.F., Zhang, L., Wang, D.S., Huang, M., Liu, J.H., Guo, J.X., 2020. Preparation of graphene-Co/Ni/Fe3O4 nanocomposites and their electrocatalytic activity for reduction of p-nitrophenol. J. Nanosci. Nanotechnol. 20 (4), 2592–2597.

Yao, L., Yang, H., Chen, Z., Qiu, M., Hu, B., Wang, X., 2021. Bismuth oxychloride-based materials for the removal of organic pollutants in wastewater. Chemosphere 273, 128576.

Zhang, Y.Y., Jiang, H., Zhang, Y., Xie, J.F., 2013. The dispersity-dependent interaction between montmorillonite supported nZVI and Cr(VI) in aqueous solution. Chem. Eng. J. 229, 412–419.

Zhong, Y., Shih, K., Diao, Z., Song, G., Su, M., Hou, L.a., Chen, D., Kong, L., 2021. Peroxymonosulfate activation through LED-induced ZnFe2O4 for levofloxacin degradation. Chem. Eng. J. 417, 129225.

Zhou, Y.M., Gao, B., Zimmerman, A.R., Chen, H., Zhang, M., Cao, X.D., 2014. Biochar- supported zerovalent iron for removal of various contaminants from aqueous solutions. Bioresour. Technol. 152, 538–542.

Zhu, S.J., Xu, Y.P., Zhu, Z.G., Liu, Z.Q., Wang, W., 2020. Activation of peroxymonosulfate by magnetic Co-Fe/SiO2 layered catalyst derived from iron sludge for ciprofloxacin degradation. Chem. Eng. J. 384, 123298.

W. Tan et al.