Degradation of Norfloxacin in an Aqueous Solution by the Nanoscale Zero-Valent Iron-Activated Persulfate Process

In this study, nanoscale zero-valent iron (nZVI) was synthesized and used to activate persulfate (PS) for the degradation of norfloxacin (NOR). The nZVI/PS system exhibited a high reactivity towards NOR, and the degradation efficiency of NOR (100mg/L) reached 93.8% with 0.1 g/L nZVI, 12mM PS, and an initial pH of 7.0 within 7min. The NOR degradation followed a pseudo-first-order kinetic model, and the effects of parameters such as nZVI dosage, PS concentration, initial pH, and temperature were investigated systematically. Overloading of nZVI lowered the degradation efficiency owing to the quenching effect of excessive Fe. The higher PS concentration and temperature favored the degradation of NOR. The influence of pH was not obvious, and the degradation was effective in a wide pH range. In addition, the radical quenching experiments and electron paramagnetic resonance (EPR) indicated that both sulfate radical (SO4 ) and hydroxyl radical (OH) were the dominant radicals in the degradation process, in which the latter played a more important role. Finally, three degradation pathways were proposed based on the result of intermediates identified by liquid chromatography-mass spectrometry. Overall, this study indicated that the nZVI/PS system could provide a promising alternative for NOR wastewater treatment.


Introduction
Nowadays, the existence of antibiotics in the ecosystem, released by pharmaceutical and domestic wastewater, has been recognized as a serious environment problem [1]. Norfloxacin (NOR), a broad-spectrum antibiotic, is one of the most important fluoroquinolone antibiotics widely used in both human and veterinary medicine [2]. In China, more than 500 ton of NOR is used every year [3]. The adsorption of NOR is quite low for both humans and animals, and the majority of NOR (60-70%) is discharged into the environment through feces and urine [4]. The extensive consumption of NOR and continuous release of effluents containing NOR lead to its bioaccumulation. This may induce antibiotic resistance in bacteria and poses threats to humans' health and the ecological environment [5]. Furthermore, owing to its low degradability and high solubility, it is challenging to remove it from water with conventional technologies [6]. The presence of NOR has been frequently detected in surface water and groundwater from ng/L to μg/L in many countries around the world [7][8][9]. Consequently, it is essential to seek effective techniques for the removal of NOR from water. Toward this goal, various treatment methods have been developed, such as adsorption [4], biodegradation [10], and advanced oxidation processes (AOPs) [11].
Among these methods, adsorption is the easiest to apply. However, it is just an immobilization method that does not remove contaminants. In addition, the adsorption ability of an adsorbent is limited for high concentrations of contaminant; furthermore, it takes a long time to achieve adsorption equilibrium [12]. For biodegradation, the adsorption process is complicated, limiting its practical application.
AOPs are regarded as effective treatment for the removal of recalcitrant contaminants [13]. In recent years, persulfatebased in situ chemical oxidation (ISCO) technology has become a promising and effective method for the remediation of organic pollutants [14]. Persulfate (PS, S 2 O 8 2-) could be an alternative oxidant based on high stability and wide pH operation range [15]. Moreover, PS could be activated via heat [16], UV [17], and transition metals [18], which generated a highly reactive sulfate radical (SO 4 ⋅-, E 0 = 2:5-3:1 V). Compared with OH ⋅ , SO 4 is more stable and exhibits a strong oxidative ability in a wide pH range [19]. Among the transition metal ions, Fe 2+ is commonly used to activate PS (Equation (1)) because it is environment friendly and relatively inexpensive [20]. However, a large quantity of Fe 2+ is needed in the process owing to the conversion of Fe 2+ to Fe 3+ , and Fe 2+ cannot be easily regenerated. The treatment should be performed under acidic conditions to avoid the precipitation of iron at high pH values. Moreover, the SO 4 ⋅could be scavenged at high concentrations of Fe 2+ [21]. To overcome these drawbacks, nanoscale zero-valent iron (nZVI) has been employed as an efficient activator of PS owing to its high reactivity. nZVI is the most commonly used nanomaterial for wastewater and underground remediation owing to its reduction potential and high reactivity towards a broad range of contaminants [22,23]. nZVI could release Fe 2+ slowly to activate PS (Equations (2)-(4)). Furthermore, the recycling of Fe 3+ /Fe 0 on the nZVI surface could keep the concentration of Fe 2+ stable in the solution and reduce the precipitation of iron hydroxides [24,25]. Previous studies have reported that nZVI/PS process could degrade a wide range of contaminants [26][27][28].
Previously, Deng et al. found that 10 mg/L NOR was completely eliminated within 5 min in the nanoscale zerovalent copper-activated PS process [29]. Compared with Cu 0 , Fe 0 exhibits a higher reactivity and may remove NOR even at low concentrations. Moreover, unlike Cu, which is a heavy toxic metal and could cause adverse effects in aquatic ecosystems, Fe is not so hazardous to the environment [30]. To the best of our knowledge, the use of Fe to remove NOR in aqueous solutions has not been reported. Hence, the study on nZVI/PS systems for the degradation of NOR is of great importance. Therefore, the objectives of this study were to (1) investigate the effectiveness of NOR degradation in the nZVI/PS system, (2) evaluate different parameters of the NOR degradation, (3) determine the oxidizing radical species, and (4) propose a possible mechanism and pathway of the reaction.

Synthesis of nZVI.
The nZVI particles were prepared using the liquid-phase reduction method [31][32][33]. Briefly, 4.94 g FeSO 4 ·7H 2 O was dissolved in a 100 mL water-ethanol solution (7 : 3, v/v), and 100 mL of a 0.36 M NaBH 4 solution was added dropwise with vigorous stirring. The entire process progressed under a nitrogen atmosphere. After nZVI particles were generated, they were isolated by vacuum filtration and rinsed three times with ethanol and distilled water. Finally, the nZVI particles were dried in a vacuum oven overnight and stored in an airtight container for further use. The process could be described according to Equation (5) [34].

Batch Experiments.
All batch experiments were conducted in 300 mL conical flasks at ambient temperature (25 ± 1°C). For each experiment, 250 mL of 100 mg/L NOR solution was transferred into the flasks; then, predetermined amounts of nZVI and PS were added into the solution to initiate the reaction. The flasks were covered with aluminum foil to prevent light from potentially influencing the reaction and placed in a water bath with magnetic stirring at a speed of 300 r/min. The pH of the NOR solution was adjusted with 0.1 M HCl or NaOH. At regular intervals, 1.5 mL of the solution sample was withdrawn and passed through a 0.22 μm membrane filter. Then, sodium thiosulfate was added to the withdrawn sample immediately to quench the reaction for further analysis. The effects of nZVI dosage (0.025, 0.05, 0.10, 0.20, and 0.30 g/L), PS concentration (3, 6, 12, and 24 mM), initial pH (3.0, 4.5, 7.0, 9.5, and 11.0), and temperature (15,25,35, and 45°C) on the degradation of NOR were investigated. All these experiments were performed in duplicate.

Analytical
Methods. The concentration of NOR was determined by a high-performance liquid chromatography (LC2030c, Shimadzu) system equipped with an C18 column (150 mm × 4:6 mm, 5μm) at a wavelength of 278 nm. The mobile phase consisted of methanol and a 0.1 wt% phosphoric acid solution (25 : 75), and the flow rate was kept at 1.0 mL/min. The concentration of dissolved iron was measured using 1,10-phenanthroline with a UVspectrophotometer (D6000, Hach Company) at 510 nm [35]. The PS concentration was determined by a spectrometric method [36]. The radicals were analyzed with an electron paramagnetic resonance (EPR) spectrometer (A300 microx, 2 Journal of Nanomaterials Bruker Corporation) according to the method proposed by Hussain et al. [34]. The mineralization of NOR was determined by a Total Organic Carbon (TOC) analyzer (TOC-L, Shimadzu). The chemical oxygen demand (COD) was measured using a COD tester (DRB200, Hach, USA). The pH was measured by a Mettler-Toledo pH meter. The intermediate products of NOR were analyzed by an LC-MS (TSQ Quantum Ultra, Thermo Scientific) equipped with an Ultimate XB-C18 column (3 μm, 2:1 mm × 100 mm). The MS analysis was operated in the positive ionization mode using an electrospray ion source, and a mixture of 0.1% formic acid in ultrapure water and acetonitrile (70 : 30, v/v) was used as the mobile phase.

Results and Discussion
3.1. Characterization of nZVI-BC. The TEM images ( Fig.  S1(a)) revealed that nZVI particles had a roughly global morphology and aggregated to a chain-like structure owing to the magnetic interactions [37]. The size of nZVI particles varied between 50 nm and 100 nm. The BET surface area of nZVI was 5.92 m 2 /g. The XRD pattern ( Fig. S1(b)) demonstrated that nZVI particles were sufficiently crystallized. The significant peak at 44.7°was attributed to the presence of Fe 0 [38], and no other obvious peaks were observed. The XPS patterns of nZVI ( Fig. S1(c)) also illustrated the presence of Fe 0 [39], and some iron oxides had formed on the surface of the nZVI particles [40]. All these results clearly demonstrated that nZVI was synthesized successfully.

Comparison of NOR Degradation in Different Systems.
A series of control experiments were carried out to evaluate the effect of different nZVI/PS system compositions on NOR degradation. Figure 1 shows the NOR degradation perfor-mances in nZVI alone, PS alone, Fe 2+ /PS, and nZVI/PS systems under the same conditions. The removal of NOR was minimal in the nZVI system, and the observations were consistent with the report that NOR is resistant to nZVI treatment [41]. A low degree of degradation (9.5%) was observed in the PS system owing to the limited oxidation ability of PS (E 0 = 2:01 V) [29]. The removal efficiency of NOR was 69.5% in the Fe 2+ /PS system after 7 min. The degradation of NOR was rapid at the beginning but remained remarkably suppressed after 1 min to the scavenging effect of Fe 2+ [24]. By contrast, the application of nZVI significantly enhanced the activation of PS, and more than 90% of NOR was removed in 7 min. The results indicated that nZVI was a good source of Fe 2+ , which can activate PS effectively to generate SO 4 ⋅- [42], thus enhancing the degradation of NOR.
The variations in dissolved iron and PS concentrations in the nZVI/PS system were investigated. As shown in Figure 2(a), the concentration of Fe 2+ first increased rapidly and then decreased to a low concentration in the reaction time, which was well in line with the consumption of PS. Fe 2+ was an important activator for PS, and it was quickly oxidized to Fe 3+ after reacting with PS. The increase in Fe 2+ led to a fast decomposition of PS in the first stage. However, according to Al-Shamsi and Thomson, the reactivity of nZVI particles is inhibited after they make contact with PS because of the passivation of nZVI [43]. Therefore, a downtrend for Fe 2+ concentration was observed owing to the higher consumption rate than release rate of Fe 2+ . Fe 3+ accumulated and increased gradually from 0 to 90.6 mg/L, as Fe 2+ was continuously released and reacted with PS. On the other hand, the total dissolved iron concentration was closed to the initial addition of nZVI, which indicated that most of the nZVI was converted to Fe 3+ after 3 Journal of Nanomaterials the reaction. The change in PS decomposition was attributed to the passivation and depletion of nZVI [21].
The variations in pH during the degradation process were also measured ( Figure 2(b)). The pH decreased rapidly in the first 30 s and then maintained a stable value. The obvious decline of the pH was attributed to the generation of H + and is discussed in Section 3.3.3. Acid conditions favored generation of Fe 2+ ; thus, a high degradation efficiency and rate were obtained in the nZVI/PS system.
The mineralization degree of NOR in the nZVI/PS system was also evaluated, and the result is presented in Fig. S2. The TOC of NOR decreased slowly in the reaction time, and mineralization of NOR was 15.8%, indicating that most of the NOR was transformed into intermediate products, a small portion of which was then mineralized to CO 2 and H 2 O. Similar mineralization results were also obtained in the nanoscale zero-valent copper-activated PS process (21.37%) [29] and peroxymonosulfate (PMS) activation process (18%) for NOR degradation [44]. The low TOC removal may be attributed to the generation of intermediates, which may be more difficult to degrade than NOR, and NOR could not be completely mineralized within 7 min [44]. The pseudo-first-order kinetic model was used to describe the degradation process of NOR in the nZVI/PS system: where C 0 (mg/L) was the initial concentration of NOR, C t (mg/L) was the concentration of NOR at time t (min), and k obs was the observed pseudo-first-order rate constant (min -1 ). All the k obs in this study are summarized in Table 1.
The catalytic activity of different nZVI dosages was investigated, and the results are presented in Figure 3. When 0.025 g/L of nZVI was used, about 57.8% of NOR was degraded within 7 min. The degradation efficiency reached 99.1% as the dosage was increased to 0.2 g/L. The reaction rate constant also increased from 0.134 to 0.733 min -1 , as the nZVI dosage was increased from 0.025 to 0.2 g/L. The results indicated that the NOR degradation was significantly influenced by the nZVI dosage. In the nZVI/PS system, the available Fe 2+ released from nZVI played a crucial role in the activation of PS and formation of SO 4 ⋅- [24,25]. The number of active sites increased with an increase in the ZVI dose within the range of 0.025-0.2 g/L, which accelerated the reaction with PS and increased the production of SO 4 ⋅-, thereby improving the degradation efficiency and reaction rate [45]. However, the degradation efficiency and reaction rate constant decreased to 96.0% and 0.436 min −1 , when the dosage further increased to 0.3 g/L. The results were attributed to the generation of excess Fe 2+ , which induced the scavenging effect of SO 4 ⋅-(Equation (7)) [21]; the consumption of SO 4 ⋅originally for degradation, therefore, hindered the degradation of NOR. The reaction occurred too quickly for the degradation process to be explored when 0.2 g/L of nZVI was added. Consequently, a dosage of 0.1 g/L was used in this study owing to its high degradation efficiency.
3.3.2. Effect of PS Concentration. The effect of PS concentration on NOR degradation was investigated with a fixed nZVI dosage. Figure 4 shows that the rate constant increased from 0.168 to 0.539 min −1 when the PS concentration increased from 3 to 24 mM. The degradation efficiency also increased from 71.2% to 98.8% in the range of 3-24 mM. This was expected because PS was the only source of SO 4 ⋅-, and more SO 4 ⋅could be generated at a higher PS concentration [46]. When the nZVI dosage was fixed, the amount of SO 4 ⋅was insufficient for NOR degradation at a low PS concentration. With increasing PS concentration, more PS molecules could make contact and react with nZVI, resulting in a higher production of SO 4 ⋅and increased NOR degradation. However, SO 4 ⋅could react with itself and PS according to Equations (8) and (9), causing a slight increase in k obs (0.453 and 0.539 min -1 ) when the PS concentration increased from 12 to 24 mM. For the 24 mM system, the degradation efficiency increased slightly from 93.8% to 98.1 with a low PS consumption (23.7%). In view of economic efficiency, 3.3.3. Effect of Initial pH. The pH was regarded as an important factor in the iron-mediated and PS activation reaction [47,48]. Figure 5 shows that the initial pH and the acidic conditions favored the degradation of NOR. It can be seen that NOR could be effectively degraded at pH < 7, and the degradation was slightly retarded at pH 9.5 (89.9%) and 11.0 (80.8%) within 7 min. The reaction rate constant also decreased from 0.736 to 0.203 min -1 with the rise in pH from 3.0 to 11.0. The corrosion of Fe 0 could be accelerated at a lower pH [49], which pro-duced more Fe 2+ that was responsible for the generation of SO 4 ⋅-. However, the dissolved Fe 2+ tended to form iron hydroxides under alkaline conditions. The hydroxides were deposited on the surface of nZVI and inhibited further reactions. Therefore, NOR degradation was suppressed under alkaline conditions especially at a pH of 11.0.
The final pH values of different experiments are shown in Table 2. The solutions became acidic (pH < 4:0) within 10 min of adding nZVI. As a comparison, PS alone had a little effect on pH, resulting in a slight pH decrease (7.0 to 5.8) [50]. In the nZVI system, the corrosion of Fe 0 would consume H + ions, which resulted in an increase in the pH value to 8.5. The decrease in pH in the nZVI/PS system could be attributed to the following: (1) SO 4 ⋅reacted with H 2 O or OH -, which led to the formation of H + or the consumption of OH -(Equations (10) and (11)) [51]; (2) the hydrolysis of PS generated H + (Equations (12) and (13)) [34]; and (3) some short-chain organic acids (e.g., formic acid, lactic acid, acetic acid, and oxalic acid) were produced in the degradation process of NOR, leading to a release in H + [52,53]. All these facilitate the corrosion of Fe 0 and thus enhance the degradation of NOR [54]. As reported, PS can be activated to generate SO 4 ⋅even at a pH of 12 [55]. Therefore, when the pH was higher than 8.5, the reaction of SO 4 ⋅with H 2 O  C/C 0 Figure 6: Effects of temperature on NOR degradation in the nZVI/PS system. Operating conditions: C 0 = 100 mg/L, ½PS 0 = 12 mM, and nZVI = 0:1 g/L. 6 Journal of Nanomaterials (Equation (11)) (k = 6:5 × 10 7 M -1 s -1 ) decreased the pH value and reduced the inhibiting effect of iron precipitation, which had a positive effect on the degradation of NOR [56]. The degradation efficiency was not significantly influenced by the initial pH and reached 80.8% when the pH was 11.0. The high degradation efficiency indicated that NOR could be degraded over a wide pH range in the nZVI/PS system.
All pHs : Alkaline pH : 3.3.4. Effect of Temperature. Figure 6 shows NOR degradation in the PS (control experiments) and nZVI/PS systems under different temperatures. The results indicated that the higher temperatures favored degradation of NOR. The degradation efficiency of NOR increased from 76.9% to 99.8% as the temperature increased from 15 to 45°C and the corresponding kinetic rate constants also increased from 0.211 to 1.202 min -1 . As expected, the degradation efficiency increased owing to the increasing production of SO 4 ⋅at higher temperatures [34]. The higher temperature could also increase the frequency of molecular collisions and accelerated the removal rate of NOR as a result. Considering the possibility of NOR degradation in thermally activated PS [16], control experiments were performed. The results (Fig. S3) revealed that an insignificant amount of NOR was removed (<15%) in the PS system after 30 min, thus indicating the inefficiency in terms of activating PS. It had been reported that PS can be effectively activated only when the temperature is over 50°C, which also supported our findings [57].
The activation energy of NOR degradation could be calculated by the Arrhenius equation: where k was the rate constant, E a was the Arrhenius activation energy (kJ/mol), A was the frequency factor, R was the universal gas constant (8.314 J/(mol K)), and T was the absolute temperature. The E a was calculated as 42.94 kJ/mol, which indicated that the degradation of NOR could be regarded as a surface-controlled reaction [58]. It could be concluded that the degradation of NOR in the nZVI/PS system could be easily achieved owing to its relatively low activation energy [59].

Degradation Mechanism of NOR.
As mentioned previously, OH ⋅ could be generated from SO 4 ⋅through different ways (Equations (13) and (14)); therefore, it was necessary to explore the potential effect of OH ⋅ on the degradation of NOR owing to its strong oxidation ability. The radical quenching experiments were carried out to evaluate the contribution of different radicals in the degradation process. MeOH was regarded as an effective scavenger for both SO 4 ⋅and OH ⋅ (k SO4⋅− = 0:9-1:3 × 10 7 M -1 S -1 , k OH⋅ = 8:0-10 × 10 8 M -1 S -1 ), and TBA was a strong quenching agent selective for OH ⋅ (k OH⋅ = 3:8 -7:6 × 10 8 M -1 S -1 , k SO4⋅− = 4:0-9:1 × 10 5 M -1 S -1 ) [57,60]. Therefore, MeOH and TBA were employed as the quenching agents. However, the reaction rate constants of the reaction between NOR and radicals were very high (k = 10 7 -10 9 M -1 s -1 for SO 4 ⋅and 1 × 10 9 M -1 s -1 for OH ⋅ ) [57]. To ensure that the scavengers had priority in consuming radicals, an overwhelming concentration of MeOH and TBA was used in the quenching study.   Figure 8: Proposed degradation pathways of NOR in the nZVI/PS system.

Journal of Nanomaterials
As depicted in Figure 7(a), the addition of scavengers had a different inhibiting effect on NOR degradation. The degradation efficiency decreased from 93.8% to 64.3% and 46.1% with 1 M MeOH and TBA, respectively. With increases in MeOH and TBA concentrations to 2 M, the degradation efficiency further decreased to 52.8% and 38.3%, respectively. The results demonstrated that both SO 4 ⋅and OH ⋅ were responsible for NOR degradation in the nZVI/PS system [34,61]. The inhibiting effect of TBA was greater than that of MeOH, indicating that OH ⋅ played a more important role in the process. These results were also consistent with previous reports [62][63][64].
EPR analysis coupled with DMPO as the radical spin trap was conducted to identify the generation of free radicals. As shown in Figure 7(b), the signals of DMPO-SO 4 ⋅and DMPO-OH ⋅ were recognized by their hyperfine splitting peaks (six lines, 1 : 1 : 1 : 1 : 1 : 1 for DMPO-SO 4 ⋅and four lines, 1 : 2 : 2 : 1 for DMPO-OH ⋅ ), which were direct evidences that SO 4 ⋅and OH ⋅ coexisted in the nZVI/PS system [29,62]. However, the peaks of these radicals could not be identified in the PS system owing to their low concentrations, confirming that PS could not be effectively activated at 25°C to generate sufficient radicals. The signal of DMPO-OH ⋅ was stronger than that of DMPO-SO 4 ⋅in the nZVI/PS system, revealing that more OH ⋅ was generated in the degradation process [29]. The results were highly consistent with the results of quenching experiments.
Based on the above analyses, the possible mechanism for NOR degradation by the nZVI/PS system was proposed. First, Fe 2+ was formed on the surface of nZVI owing to the corrosion of Fe 0 . Second, PS was activated by Fe 2+ , and the highly reactive radicals SO 4 ⋅-/OH ⋅ were generated continuously. Finally, NOR was attacked by SO 4 ⋅-/OH ⋅ and transformed into its intermediates with a series of reactions. Meanwhile, a quantity of NOR was mineralized to CO 2 and H 2 O.

Possible
Pathways of NOR Degradation. The intermediate products of NOR degradation in the nZVI/PS system were analyzed by LC-MS, and the MS spectra were compared with the results of previous reports [29,57]. Except for NOR, eight intermediates or degradation products were identified, and their structures are listed in the Fig. S3. The structure of the piperazinyl ring and nalidixic rings, which seem to be easily attacked by free radicals, was relatively unstable according to the results of a quantitative calculation [41]. Moreover, the standard redox potentials of SO 4 ⋅and OH ⋅ (2.60 V and 2.80 V) were higher than those of the activation energies of hydroxylation on the piperazinyl ring, furthermore, the defluorination and hydroxylation reactions on the quinolone group of NOR (1.55-2.35 eV) [65]. SO 4 ⋅tended to attack the electron-rich sites via electron transfer, and OH ⋅ was selective for hydrogen abstraction or addition reaction [66]. On the basis of the structure of the degradation products and literature reports [53,67], three possible degradation pathways of NOR are proposed in Figure 8: pathway 1: defluorination. The C-F bond was directly attacked by radicals due to the nucleophilicity of the fluorine atom, and F was replaced by the hydroxyl group, which yielded com-pound A [11]; pathway 2: piperazine ring opening and cleavage. The reaction was initiated by carbonylation and hydroxylation of NOR, and compounds B and C were successively generated [68]. Then, cleavage of the piperazine ring could occur during the oxidation of compound C followed by structural rearrangement, producing compound D. Compound D could lose -CH 2 and -CO groups successfully to form compounds E and F; compound F could continue to lose -C 2 H 5 N to yield compound G, which finally underwent decarbonylation to form compound H; and pathway 3: quinolone group transformation. The quinolone moieties were attacked by radicals at the C=C bond adjacent to the carboxylic acid group; thus, compound I was generated. Several studies have reported the formation of short-chain organic compounds in the degradation of NOR by AOPs [52,53]. Finally, some of the transformation products were further mineralized to CO 2 and H 2 O.

Conclusions
This study provided a simple and effective method for the removal of NOR in aqueous solutions by using nZVI as the catalyst to activate PS. The key findings are as follows: (1) nZVI demonstrated excellent catalytic activity, and a high removal efficiency (93.8%) of NOR was achieved in the nZVI/PS system (2) NOR degradation was significantly influenced by nZVI dosage, PS concentration, initial pH, and temperature (3) Both SO 4 ⋅and OH ⋅ were the dominant radical species based on the results of quenching experiments and EPR analysis, while OH ⋅ played a more important role (4) The intermediates were analyzed, and three degradation pathways including defluorination, piperazinyl ring opening, and quinolone group transformation were proposed This study suggested that nZVI/PS is a promising pretreatment process for the antibiotic pollution caused by NOR. Figure S1: characterizations of the nZVI. (a) TEM, (b) XRD, (c) XPS analysis. Figure S2: change profiles of TOC and NOR concentration versus time in the nZVI/PS system. Operating conditions: C 0 = 100 mg/L, pH = 7:0, T = 25°C, ½PS 0 = 12 mM, and nZVI = 0:1 g/L. Figure S3: the degradation efficiency of NOR in the PS system. Operating conditions: C 0 = 100 mg/L, pH = 7:0, and ½PS 0 = 12 mM. Figure S4: MS spectra of degradation products of NOR in the nZVI-BC/PS system. Graph abstract: the whole process of NOR degradation in the nZVI/PS system. The picture demonstrated the corrosion of nZVI, the activation of PS, and the potential pathway in the process. (Supplementary Materials)