Dielectric barrier discharge combined with Fe(III)-NTA activated persulfate for efficient degradation of enrofloxacin in water

ABSTRACT

Enrofloxacin (ENR) is widely used in aquaculture due to its broad-spectrum activity against a large number of pathogenic bacteria. However, ENR exhibits strong adsorption and a low biodegradation rate in the environment. This study proposes a new method for degrading ENR, which combines dielectric barrier discharge (DBD) with advanced oxidation of peroxymonosulfate (PMS, HSO₅⁻) and the use of nitrilotriacetic acid (NTA) combined with Fe(III) activator to improve reaction efficiency. The effects on the degradation efficiency were experimentally investigated by adjusting the internal process parameters and adding coexisting substances. The ESR and free radical scavenging experiments showed that both hydroxyl and sulfate radicals contributed to the degradation of the pollutant ENR. Based on the identified intermediates, and three possible degradation pathways were suggested. And the intermediates were less harmful. In conclusion, the Fe(III)-NTA/PMS/DBD system may be an efficient, environmentally friendly and new technology with development potential.

KEYWORDS:

• Dielectric barrier discharge plasma
• Fe(III)-NTA complexes
• peroxymonosulfate
• advanced oxidation
• enrofloxacin

1. Introduction

Enrofloxacin (ENR) is a synthetically produced third-generation fluoroquinolone antibiotic. Due to its broad-spectrum activity against a wide range of pathogens, it is extensively used in aquaculture [1]. However, ENR is not completely absorbed by animals, and its residues not only affect the growth of animals and plants but also easily lead to water and soil pollution. Currently, the average concentration of enrofloxacin detected in the main rivers, lakes, seas, and adjacent marine waters within China is 100 ng/L [2]. Enrofloxacin not only can cause liver damage and a decline in immune system function in fish [3], but its prolonged presence in soil and aquatic environments also impacts the structure and function of microbial communities [4]. Moreover, due to their high adsorption, difficult biodegradation, and long half-life period [5], it is challenging to be removed through natural processes. The accumulation of Enrofloxacin in the environment poses a threat to human health. Studies [6] also have indicated that traditional activated sludge methods in sewage treatment plants cannot effectively remove it due to the high antimicrobial activity of antibiotics. Therefore, there is a need to seek an efficient method to remove ENR from water bodies.

Advanced Oxidation Processes (AOPs) are currently effective methods for removing antibiotics, such as Fenton [7], ozone [8], photocatalysis [9], electrochemistry [10], solar light/periodate [11] and persulfate techniques [12], all of which have strong oxidation capabilities [13,14]. Dielectric Barrier Discharge (DBD) is an emerging low-temperature plasma water treatment technology among AOPs. It integrates the advantages of advanced oxidation in treating organic substances. In addition to directly using reactive oxygen species (ROS) produced by the discharge effect, it can also degrade organic pollutants through the combination of chemically active substances, ultraviolet light, thermal effects, and shockwaves [15]. DBD is mainly applied for the removal of organic pollutants [16], heavy metals [17] and inactivation and disinfection of pathogenic microorganisms [18] in wastewater. However, compared to other advanced oxidation technologies, many energy effects produced in this technology, such as ultraviolet radiation and shockwaves, are not fully utilized, leading to low energy efficiency of the plasma [19]. To improve the energy efficiency of this treatment technology and thereby enhance the degradation efficiency of organic pollutants, the combination of DBD with other advanced oxidation technologies has become an inevitable trend. Studies [20,21] have shown that the addition of some solid catalysts can improve the energy utilization of the plasma and the removal efficiency of pollutants, but there are also difficulties in separation in the solution and medium conduction. Persulfate advanced oxidation technology can avoid these problems, and peroxymonosulfate (PMS, HSO₅⁻) can be activated to produce SO4·⁻ with stronger oxidizing properties than ·OH [22] and it can also react with H₂O to produce ·OH [23], thereby improving the energy utilization efficiency of the plasma. At the same time, the addition of transition metal Fe(III) will also enhance the ability of PMS to produce active substances, thereby jointly improving the system’s efficiency in removing organic pollutants [24]. However, the addition of Fe(III) has disadvantages, such as the easy production of iron sludge [25] and the extremely slow redox cycle of Fe(III)/Fe(II) [26]. According to a report [27], nitrilotriacetic acid (NTA), a common aminopolycarboxylate metal complexing agent, can solve the aforementioned issues, regulate the generation rate of hydroxyl radicals, and is easily biodegradable. As a result, it is widely used in the water treatment and soil remediation industries [28]. This study constructs a Fe(III)-NTA/PMS/DBD oxidation system, investigates the degradation performance of this system on ENR, and proposes the degradation pathway of ENR through radical quenching tests and intermediate product detection.

2. Materials and methods

2.1. Materials

The analytical grade of NTA, KHSO₅, FeCl₃, ENR, humic acids(HA),NaHCO₃, NaCl, NaOH, HCl, Na₂S₂O₃, C₄H₁₀O, C₂H₅OH, C₆H₄O₂ were purchased from the China National Pharmaceutical Group Corporation. All solutions were prepared using ultrapure water (resistivity, σ > 18.25 MΩ·cm).

2.2. Experimental methods

Figure 1 was a schematic representation of the experimental apparatus, comprising primarily four components: a pulse power supply (Model CTP-2000K, Nanjing Suman Plasma Technology Co., Ltd., China), a contact voltage regulator (Model TDGC2–1, Zhejiang Chint Electrics Co., Ltd., China), a DBD reactor system (featuring a quartz glass dielectric with an outer diameter of 100 mm, a discharge area diameter of 55 mm, a discharge gap of 8 mm, dielectric layers on the upper and lower sides each with a thickness of 2 mm, an electrode gap of 10 mm, and a working volume of 15 mL), and an electrical detection system. This detection system encompassed an oscilloscope (Model TBS1102C, Tektronix, Inc., China), a current probe (Model TPP0100, Tektronix, Inc., China), and a voltage probe (Model TPP0100, Tektronix, Inc., China).

Figure 1. Experimental setup.

A Fe(III)-NTA solution was prepared by mixing an aqueous solution of FeCl₃ with an NTA solution in a predefined molar ratio. The solution was shielded from light and allowed to stand for at least one hour to facilitate complete complexation. The experimental parameters were set to a peak voltage of 20 kV and a frequency of 50 Hz, with an electrical discharge duration of 40 minutes. Subsequently, a specified concentration of Fe(III)-NTA, PMS, and 50 mg/L ENR solution was introduced into a DBD reactor. Except for investigating the effect of initial pH value, the pH was not adjusted in all other experiments. Samples were collected at predetermined intervals. Upon completion of the reaction, it was quenched by adding 1 mL of 0.5 M Na₂S₂O₃ solution. Finally, the water samples were filtered through a 0.22 μm membrane to analyze the aqueous contaminants and assess the water quality parameters.

2.3. Analytical methods

The concentration of ENR was determined using a UV-Vis spectrophotometer (UV-1800, Shanghai Mapada Instruments Co., Ltd., China) at a detection wavelength of 271 nm. After testing, the detection limit of ENR’s UV spectrophotometer was 0.1 mg/L. The zeta potential and particle size were measured using a Zetasizer Ultra (Malvern Panalytical Limited, UK) at pH levels of 2, 4, 6, 8, and 10 to ascertain the presence of electrostatic adsorption in the oxidation system. Total Organic Carbon (TOC) was quantified using a TOC analyzer (multi N/C 2100, Analytik Jena AG, Germany). The electron spin resonance (ESR) spectra were conducted on a ESR spectrometer (Bruker EMX-PLUS, Germany). 5,5-dimethyl-1-pyrroline N-oxide (DMPO) and 2,2,6,6-tetramethyl-4-piperidinol (TEMP) were used as spin trapping agents. The degradation products of ENR were analyzed using High-Performance Liquid Chromatography-Mass Spectrometry (HPLC-MS) coupled with Quadrupole Time-of-Flight Mass Spectrometry (1290uplc-qtof6550, Agilent, U.S.A.). The chromatographic column used was a C18–2.1 × 100 mm *1.7 μm (BEH, Waters, U.S.A.) with a column temperature of 40°C.

In this study, the Toxicity Estimation Software Tool (TEST, Version 5.1.2) was utilized to predict the toxicity of the degradation intermediates of ENR. The potential toxicological risks were assessed using several evaluation metrics, including the Bioconcentration Factor, developmental toxicity, mutagenicity, and the 48hour -Lethal Concentration 50(LC50) for Daphnia magna. These indices were employed to provide a comprehensive prediction of the potential toxicological hazards presented by the ENR degradation products.

3. Results and discussion

3.1. Comparison of oxidation process for ENR degradation

To identify the optimal system for the treatment of ENR, we conducted degradation experiments under various systems to determine the degradation efficiency of ENR. Figure 2 illustrated the relationship between the performance of the treatment system and the plasma discharge time.

Figure 2. Degradation of ENR in different systems. Experimental conditions: [ENR] = 50 mg/L; [PMS] = 0.3 g/L; the molar ratio of Fe(III)-NTA = 1:1;[Fe(III)-NTA] = 0.2 mM; peak voltage = 20 kV.

In the process without DBD discharge, persulfates generated sulfate radicals (SO₄^·-) with high oxidative potential (Equation (1)). However, the degradation efficiency of persulfates alone was relatively low. Typically, transition metal ions such as Fe(III)/Fe(II) were added to disrupt the -O-O- bond in persulfates (Equation (2)) [29,30]. As shown in Figure 2, the addition of Fe(III) resulted in an approximate 3% increase in the degradation efficiency of ENR within 40 minutes. Nevertheless, this system also faced issues of low degradation rates. The reason was that the regeneration rate of Fe(II) was significantly lower than the oxidation rate of Fe(III), while Fe(II) played a crucial role in the activation process [31]. When the organic chelating agent NTA was added, it participated in the electron transfer oxidation reaction mediated by Fe(III), converting Fe(III) to Fe(II). The reason for this transformation was as follows: The complex formed by the chelator with Fe(III) had a lower redox potential than free Fe(III). A lower redox potential indicates stronger reducing properties. Therefore, the complex formed by the chelator with Fe(III) had a greater reducing capacity than free Fe(III). Thermodynamically, the complex formed by the chelator with Fe(III) was more easily reduced to Fe(II), thereby increasing the concentration of Fe(II) in the reaction and consequently enhancing the degradation of the pollutant ENR. This enhancement generated more free radicals in the system, thereby facilitating the degradation of target pollutants, as detailed in reactions (Equations (3–6)) [32,33]. As indicated in Figure 2, the Fe(III)-NTA/PMS system increased the degradation rate by 10% within 40 minutes of treatment compared to the single PMS system.

During the single DBD discharge process, an increase in discharge voltage caused a breakdown of the air layer between the high-voltage electrode and the ground electrode. This transformed electrons receiving electrical energy into high-energy electrons. At the gas-liquid interface, O₂, N₂, and H₂O underwent a series of physicochemical reactions with these high-energy electrons. These reactions produced a substantial amount of reactive species, ultraviolet radiation, and localized high temperatures [34,35] Additionally, the discharge process was accompanied by a certain intensity of ultraviolet light and thermal effects produced by the high-voltage discharge (Equations (7–11)). These chemical and physical reactions collectively attacked the organic pollutants in the solution. This led to the oxidative decomposition of the target pollutants, resulting in a 63% degradation rate of ENR within 40 minutes of experimental time [36].

Upon integrating PMS during the plasma discharge process, radicals such as O₃ and H₂O₂, generated by the plasma discharge, enhanced the oxidative capacity of the reaction system. These radicals, possessing higher oxidation-reduction potentials, contributed to this enhancement. For instance, the ·OH and SO₄·⁻ generated by PMS, along with the ·OH produced during the plasma discharge process, collaboratively degraded organic pollutants. SO₄·⁻ selectively degraded organic pollutants containing electron-donating groups, while ·OH non-selectively degraded those with electron-withdrawing groups. This effectively improved the treatment efficiency of organic pollutants [37]. Concurrently, PMS was also activated by the thermal effects of the DBD. The e⁻ and O₂·⁻ generated during the DBD discharge played a significant role in the formation of radicals within the DBD/PMS system [24,37] (Equations (12–18)). As illustrated in Figure 2, the degradation rate of ENR reached 78% within 40 minutes of experimental time.

However, the addition of Fe(III), in the Fe(III)/PMS/DBD system, did not show a significant difference in degradation rate within 40 minutes of experimental time compared to the PMS/DBD system, which aligns with the analysis of the PMS and Fe(III)/PMS systems. In the Fe(III)-NTA/PMS/DBD system, the generated Fe(II)-NTA· not only reacts with PMS and H₂O₂ to produce ·OH and SO₄^·-, but also competes with ENR molecules for these radicals, reducing the consumption of ·OH and SO₄^·- [24] (Equations (2, 18–21)).This promotes the system’s degradation efficiency for the target pollutant ENR, as shown in Figure 2, where the final degradation rate of ENR reached 92% within 40 minutes.

In summary, without DBD, systems such as Fe(III), Fe(III)-NTA, Fe(III)/PMS, and Fe(III)-NTA/PMS were ineffective in substantially degrading ENR. The most effective among these, Fe(III)-NTA/PMS, achieved only a 29% degradation rate in 40 minutes. However, integrating the DBD system with the PMS oxidation process significantly enhanced ENR’s degradation efficiency. The Fe(III)-NTA/PMS/DBD system achieved a 92% degradation efficiency in 40 minutes. This indicates a synergistic effect between the DBD plasma and PMS in degrading ENR.Additionally, the degradation of ENR through different systems all follow pseudo-first-order kinetics (R² >0.95), with the reaction rates ranked as follows: Fe(III)-NTA/PMS/DBD (0.06456 min⁻¹) > DBD (0.02821 min⁻¹) > Fe(III)-NTA/PMS (0.00925 min⁻¹). Table 1 showed the comparison of the degradation efficiency of this experimental system with other advanced oxidation techniques. This indicated that the synergistic system was a promising technology for the degradation of organic pollutants.

Table 1. A summary of plasma water treatment technology and other advanced oxidation technology applications.

Process	Conditions	Targets	Removal rate (%)	Ref.
Plasma-PMS	Fe(III)-NTA/PMS/DBD	ENR (50 mg/L)	92 (40 min)	this study
PMS	Co (OH)2/CNH; 0.64 mM PMS	ENR (20 ppm)	95.6 (30 min)	[38]
Ozone	0.19 mM O3	CIP (3 μM)	100 (30 min)	[8]
Photocatalytic	3.18 mW/cm^2 UV; n ZnO : nTiO2 (1:1)	ENR (50 mg/L)	92 (120 min)	[9]
Electrochemical	GNP-PbO2; Co = 100 mg/L; Current density = 20 mA/cm^2	ENR (100 mg/L)	100 (120 min)	[10]
Plasma	Strong ionization DBD	Nonylphenol (5 mg/L)	99 (60 min)	[39]
Plasma	Strong ionization DBD	Dimethyl phthalate (20 mg/L)	93 (60 min)	[40]
Plasma	Strong ionization DBD	Aniline (20 mg/L)	87 (60 min)	[41]
Plasma	Strong ionization DBD	Nitrobenzene (20 mg/L)	100 (55 min)	[42]
Plasma	DBD	Nitrobenzene (20 mg/L)	75 (60 min)	[43]
PMS- Plasma	PMS/DBD	Benzotriazole (10 mg/L)	97 (20 min)	[37]
Photo-Plasma	graphene-TiO2/DBD	FLU (40 mg/L)	99.4 (60 min)	[15]
Photo-Plasma	TiO2/Activated-carbon-fibers/DBD	Triclocarban (10 mg/L)	84.9 (30 min)	[44]

HSO₅⁻$\to$SO₄^·-$+$OH⁻ (1)

Fe (II)$+$HSO₅⁻$\to$SO₄^·-$+$OH⁻$+$Fe(III) (2)

Fe(III)$-$NTA$\to$Fe(II)$-$NTA· (3)

NTA·$+$O₂$\to$NTA$+$O₂·⁻ (4)

Fe(III)$-$NTA$+$O₂·⁻$\to$Fe(II)$-$NTA·$+$O₂ (5)

Fe(II)$-$NTA·$+$HSO₅⁻$+$H⁺$\to$Fe(III)$-$NTA$+$SO₄·⁻$+$H₂O (6)

^*e⁻$+$O₂$\to$2·O$+$e⁻ (7)

O₂$+$·O$\to$O₃ (8)

·O$+$H₂O$\to$2·OH (9)

2·OH$\to$H₂O₂ (10)

O₃$+$HO₂·$\to$HO·+O₂$+$⁻O₂· (11)

HSO₅⁻ $\stackrel{\mathit{plasma}}{\to }$SO₄·⁻$+$OH⁻ (12)

HOOSO₃⁻$+$e⁻$\to$SO₄^2-$+$HO· (13)

HOOSO₃⁻$+$e⁻$\to$SO₄·⁻$+$HO· (14)

-O₃SO$-$OSO₃⁻$+$e⁻$\to$SO₄·⁻$+$SO₄^2- (15)

HOOSO₃⁻$+$O₂·⁻$\to$SO₄^2-$+$HO·⁻$+$O₂ (16)

HOOSO₃⁻$+$O₂·⁻$\to$SO₄^·-$+$HO⁻$+$O₂ (17)

-O₃SO$-$OSO₃⁻$+$O₂·⁻$\to$SO₄^·-$+$SO₄^2-$+$O₂ (18)

Fe(II)$-$NTA·$+$H₂O₂$\to$Fe(III)$-$NTA$+$·OH$+$OH⁻ (19)

Fe(II)$-$NTA·$+$·OH$\to$Fe(III)$-$NTA$+$OH⁻ (20)

Fe(II)$-$NTA·$+$SO₄·⁻$\to$Fe(III)$-$NTA$+$SO₄^2- (21)

3.2. Effect of internal factors of the system

3.2.1. Effect of molar ratio of Fe(III)-NTA

Figure 3(a) showed the impact of the molar ratio of Fe(III)-NTA on the degradation of ENR. When the molar ratio increased from 1:0.5 to 1:1, the degradation rate of ENR rose from 87% to 92%, and the reaction rate increased from 0.05779 min⁻¹ to 0.06456 min⁻¹. However, further increasing the concentration of NTA reduced the efficiency. Specifically, as the Fe(III)-NTA molar ratio rose from 1:1 to 1:1.5 and 1:2, the degradation efficiency of ENR decreased from 92% to 90% and 87%, respectively. The trend’s potential reason is that typically, one NTA molecule binds with one Fe(II) or Fe(III) ion, forming a 1:1 Fe-NTA complex [26]. Studies [45,46] suggest that more NTA could mean more complete Fe(III) complexation, but NTA also competes with target reactants for radicals. With increased NTA concentration, its scavenging effect on radicals becomes more pronounced, affecting the degradation efficiency of ENR.

Figure 3. Effect of oxidation system process parameters on ENR degradation: (a) Fe(III)-NTA molar ratio, (b) Fe(III)-NTA dosage, (c) PMS dosage,(b) ENR dosage. Experimental conditions: [ENR] = 50 mg/L; peak voltage = 20 kV; pH = 4.53(initial).

3.2.2. Effect of Fe(III)-NTA dosages

A degradation experiment was conducted using the Fe(III)-NTA molar ratio of 1:1, which was selected for its optimal degradation efficiency. Figure 3(b) showed that as the dosage of Fe(III)-NTA increased from 0.02 mM to 0.2 mM, both the degradation efficiency and reaction rate of ENR progressively improved. The degradation rate rose from 86% (0.04221 min⁻¹) to 92% (0.06456 min⁻¹). However, increasing the Fe(III)-NTA dosage to 0.5 mM paradoxically decreased the ENR degradation efficiency to 89%. This decrease could be due to high doses of NTA inducing competitive reactions, excessively consuming active species like ·OH and SO₄·⁻ [47]. Therefore, 0.2 mM Fe(III)-NTA was identified as the optimal dosage for this oxidation system.

3.2.3. Effect of PMS dosages

Figure 3(c) illustrated that the degradation effect of the Fe(III)-NTA/PMS/DBD system on ENR initially increased and then decreased with rising PMS dosage. When the concentration of PMS went up from 0.1 g/L to 0.3 g/L, ENR’s removal rate climbed from 86% to 92%. Yet, a further increase in PMS dosage (from 0.4 g/L to 0.5 g/L) led to no additional improvement in degradation efficiency, with the reaction rate dropping from 0.05706 min⁻¹ to 0.05363 min⁻¹. This trend could be explained by the fact that a moderate PMS concentration boost helped generate more reactive species, thereby aiding in ENR degradation. However, a higher PMS concentration might have turned PMS into a scavenger for SO4·⁻, reacting with SO₄^·- and ⋅OH to form SO₅·⁻ with a lower oxidation-reduction potential (E⁰ = 1.1 eV) than SO4·⁻ [48,49] (Equations (22, 23)), thereby reducing the degradation efficiency of the system.

HSO₅⁻$+$SO₄·⁻$\to$SO₅·⁻$+$SO₄^2-$+$H⁺ (22)

HSO₅⁻$+$HO·$\to$SO₅·⁻$+$OH⁻$+$H⁺ (23)

3.2.4. Effect of initial ENR concentrations

Figure 3(d) showed how varying initial ENR concentrations affected the degradation efficiency. Higher efficiency was observed at lower ENR concentrations. At an initial concentration of 5 mg/L, the degradation rate reached approximately 91% in 20 minutes. For concentrations of 10 mg/L and 20 mg/L, the rate was about 90% after 30 minutes. At 50 mg/L, efficiency reached 92% by 40 minutes. As the initial ENR concentration rose, the time for degradation extended, and reaction rates dropped. The rates were 0.12789 min⁻¹ for 5 mg/L, 0.09278 min⁻¹ for 10 mg/L, 0.07553 min⁻¹ for 20 mg/L, and 0.06456 min⁻¹ for 50 mg/L. This pattern occurred because the total amount of reactive substances in the system was fixed. With higher initial ENR levels, the limited reactive substances couldn’t degrade all ENR molecules effectively, leading to slower degradation rates at higher ENR concentrations [42].

3.3. Effect of initial pH value

After pH testing, the pH of the initial solution was 4.53. The initial low pH observed in the system can be attributed to the inherently acidic nature of the experimental reagents utilized in the system. Figure 4(a) showed that the degradation efficiency of ENR gradually decreased with increasing pH. This decline was due to the higher binding affinity of OH⁻ to Fe than Fe to the chelating agent at higher pH levels, leading Fe(III)/Fe(II) to form hydroxides more readily [50]. This transition reduced the concentration of Fe(III)/Fe(II) available for activating PMS, decreasing free radical concentration and thus lowering ENR degradation efficiency. Additionally, studies [51] indicated that under alkaline conditions, SO₄·⁻ in the oxidation system was more likely to react with OH⁻ and H₂O (Equations (24, 25)), producing ·OH, which then became the dominant radical in the system. The generated ·OH (with an oxidation-reduction potential of E⁰ = 2.5–3.1 eV) had a lower potential than SO₄·⁻ (E⁰ = 2.8 eV) [37]. Compared to ·OH, SO₄·⁻ was more selective and effective in electron transfer with unsaturated bonds in organic compounds and had a longer lifespan, potentially enhancing degradation efficiency. Moreover, DBD generated more reactive oxygen species (ROS) under acidic conditions, activating PMS to produce SO₄·⁻ [52]. Thus, acidic conditions were more conducive to ENR removal.

Figure 4. (a)Effect of initial pH on ENR degradation, (b) Zeta potential of Fe(III)-NTA(III). Experimental conditions: [ENR] = 50 mg/L; the molar ratio of Fe(III)-NTA = 1:1; [Fe(iii)-NTA] = 0.2 mM; peak voltage = 20 kV.

SO₄·⁻$+$OH⁻$\leftrightarrow$SO₄^2-$+$·OH (24)

SO₄·⁻$+$H₂O$\leftrightarrow$HSO₄⁻$+$·OH (25)

The Zeta potential of the Fe(III)-NTA solution at different pH values was measured (Figure 4(b)), determining the point of zero charge (pH_zpc) for the activator to be 3.31 (pH_zpc = 3.31). When the pH was less than pH_zpc, the activator carried a positive charge, making it more likely to adsorb anions from the solution. In contrast, at a pH greater than pH_zpc, the activator had a negative charge, leading to electrostatic repulsion with anions. This resulted in varying pollutant degradation rates within the system under different pH conditions [53]. Regarding PMS, it predominantly existed as HSO₅⁻ below a pH of 9.3 and as SO₅^2- above it, showing its anionic nature over a wide pH range. For ENR, it was cationic when the pH was below its pK_a1 (3.85) and anionic above its pK_a2 (9.86) [53]. Therefore, at an initial experimental pH of 3.5, with pH_zpc (3.31) < pH < pK_a (3.85) [53], ENR was positively charged, Fe(III)-NTA negatively charged, and PMS existed as an anion. This facilitated the adsorption of ENR molecules onto the activator’s surface and PMS, thereby accelerating ENR’s degradation.

3.4. Effects of coexisting substances in surface water on ENR degradation

In practical applications, the composition of surface water bodies is complex, with co-existing substances like natural HA and inorganic anions impacting the stability of oxidants and altering the degradation products of target pollutants. This affected the generation of reactive oxygen species during oxidation, thereby influencing the degradation efficiency of organic pollutants [54]. Consequently, this study investigated the impact of co-existing substances on the degradation of ENR by introducing varying concentrations of HA, Cl⁻, and CO₃^2- into the oxidation system.

When 1 mg/L of HA was added, as depicted in Figure 5(a), the degradation rate of ENR decreased from 92% to 89% within 40 minutes. Increasing HA concentrations to 5 mg/L, 10 mg/L, and 20 mg/L led to further declines in ENR degradation to 87%, 85%, and 82%, respectively. This indicated that the presence of HA inhibited the removal of ENR. HA, a polyelectrolyte macromolecule with functional groups like phenolic, carboxylic, hydroxyl, and carbonyl, tended to react with reactive species such as ·OH and SO₄^·-, acting as a radical scavenger and reducing degradation efficiency. Moreover, HA could affect the generation of ·OH from hydrogen peroxide, further impacting the system’s efficiency [55,56].

Figure 5. Effects of coexisting substance concentrations and free radical quenchers in water on ENR degradation. (a) HA concentration, (b) Cl⁻ concentration, (c) CO₃^2- concentration. Experimental conditions: [ENR] = 50 mg/L; [PMS] = 0.3 g/L; the molar ratio of Fe(III)-NTA = 1:1; [Fe(iii)-NTA] = 0.2 mM; peak voltage = 20 kV.

Figure 5(b) showed that adding Cl⁻ significantly inhibited ENR degradation, with higher Cl⁻ concentrations worsening this effect. At 100 mM Cl⁻, ENR degradation efficiency fell to 56%. Cl⁻reacted with radicals like ·OH and SO4·- in the system, forming radicals such as Cl·、Cl₂·⁻、ClOH·⁻、HClO and ClO₃⁻ with lower oxidative capacities than ·OH and SO₄^·- This was because of their lower oxidation-reduction potentials E⁰(Cl₂^·-/Cl^·) = 2.41 V, E⁰(Cl₂/Cl⁻) = 1.36 V，E⁰(HOCl/Cl⁻) = 1.48 V compared to ·OH and SO₄^·-(E⁰ = 2.5 V–3.1 V) [28,57–59]. As a result, ENR degradation efficiency decreased with increasing Cl⁻ concentration.

The addition of CO₃^2- notably reduced the degradation efficiency of ENR; efficiency dropped from 92% to 52% as CO₃^2- concentration increased from 0 mM to 100 mM. This occurred because CO₃^2- in the system reacted with radicals like OH and SO₄·⁻ to form CO₃·⁻ (E⁰ = 1.78 V) and HCO₃·, which had significantly lower oxidation-reduction potentials than ·OH and SO₄·⁻. CO₃·⁻ and HCO₃· [59,60], being electrophilic radicals with higher selectivity towards organics, led to a reduced oxidative capacity and diminished degradation efficiency of ENR.

3.5. Role of active species and mineralization removal

The radicals produced by this system in the degradation process and the degree of contribution of each radical were determined by adding different concentrations of ESR spectra and free radical scavenger analysis. In the Figure 6(a, b), DMPO-·O₂⁻ produced a distinct 1:1:1:1 characteristic peak, and a distinct 1:1:1 characteristic peak was observed for TEMP-¹O₂. A 1:2:2:1 signaling characteristic peak was observed for DMPO-·OH and a 1:2:2:1 signaling characteristic peak was observed for DMPO-SO₄·⁻ of six characteristic peaks. To investigate the contribution of radicals to ENR degradation, tert-butanol (TBA, ·OH radical scavengers, 10 mM), ethanol (EtOH, ·OH and SO₄·⁻ radical scavengers, 10 mM), benzoquinone (BQ, ·O₂⁻ radical scavengers, 10 mM) were added to quench radicals during the reaction. As shown in Figure 6(c), the addition of tert-butanol, ethanol, and benzoquinone reduced ENR removal efficiency to 45%, 19%, and 63%, respectively. Previously, it was indicated that ·OH, SO₄·⁻, and O₂⁻ all participated in ENR degradation, with the extent of inhibition from scavengers in descending order: EtOH>TBA>BQ, suggesting the contribution order to ENR removal was ·OH radical > SO₄·⁻ radical > ·O₂⁻ radical.

Figure 6. (a, b) ESR spectra showing DMPO and TEMO adducts formed with SO₄·⁻, ·OH, ·O₂⁻ and¹O₂, (c) the effect of free radical scavengers on ENR degradation, (d) removal rate of ENR and TOC. Experimental conditions: [ENR] = 50 mg/L; [PMS] = 0.3 g/L; the molar ratio of Fe(III)-NTA = 1:1; [Fe(iii)-NTA] = 0.2 mM; peak voltage = 20 kV.

Furthermore, by monitoring TOC changes during ENR degradation (Figure 6(d)), it was observed that after 40 minutes of reaction, the removal rate of ENR reached 91%, but the TOC removal rate was only 49%. It was indicated that while ENR was removed, mineralization was relatively low, suggesting incomplete degradation by this oxidation system. Therefore, it was essential to analyze the main degradation products of ENR.

3.6. Proposed pathways for ENR degradation

HPLC-MS was employed to analyze the intermediate products formed during the experimental process. The principal intermediates are presented in Table 2. Based on these findings, three possible pathways were proposed, as illustrated in Figure 7.According to the studies [61,62], Pathway 1 involved ENR undergoing defluorination to form P1 (C₁₉H₂₃N₃O₃, m/z = 340), which then opened the benzene ring and undergoes hydroxylation to yield P2 (C₂₀H₂₇N₃O₅, m/z = 390). Further oxidation led to side-chain cleavage, destroying the ENR structure and forming P3 (C₆H₁₄N₂, m/z = 115). Due to the use of FeCl₃ as an activator and the potential participation of free Cl⁻ in water, Pathways 2 and 3 were proposed. In Pathway 2, under radical influence, the piperazine group opened, leading to C-N bond cleavage and forming intermediate A (C₂₀H₂₄FN₃O₄), which was not detected in this experiment but was reported in other studies [63]. This intermediate could potentially have been formed under UV light produced by the plasma used. Further radical-induced C-F and C-C cleavage and oxidation formed P4 (C₁₉H₂₁N₃O₅, m/z = 369) or opened the piperazine ring in A, substituting Cl for F to yield P5 (C₁₉H₂₀ClN₃O₅, m/z = 405). Pathway 3, as reported in the paper [64], indicates that during the degradation of ENR using UV/Fe(II)/H2O2, ciprofloxacin (Intermediate B, C₁₇H₁₈FN₃O₃) is formed through OH substitution and subsequent decarboxylation. However, it was not detected in this experiment, likely due to its low concentration. Subsequently, Cl replaces F and -OH in ciprofloxacin, resulting in the formation of P6 (C₁₆H₁₇Cl₂N₃O, m/z = 316). Additionally, the C-N bond between the piperazine group and the aromatic ring breaks, leading to the formation of P7 (C₁₃H₁₀FNO₃, m/z = 248). Although Intermediate C, formed by Cl substitution on the carboxyl group of P7 [65], was not detected in this study, it has been reported in similar degradation processes of fluoroquinolone antibiotics like fleroxacin under Cl and ClO₂ conditions [65]. Finally, under oxidative conditions, Intermediate C (C₁₂H₉ClFNO) undergoes cyclopropane ring cleavage and defluorination, resulting in the formation of P8 (C₉H₆ClNO, m/z = 180). Under the influence of radicals, these products further mineralize into substances such as H₂O and CO₂.

Figure 7. Proposed ENR decomposition pathways in the Fe(III)-NTA/PMS/DBD system.

Table 2. M/Z, chemical mechanism of degradation products produced in the reaction process.

No.	m/z	Retention time (min)	abb7d998”>Chemical formula	Molecular structure
P1	340	10.984	C19H23N3O3
P2	390	5.411	C20H27N3O5
P3	115	0.601	C6H14N2
P4	369	10.377	C19H21N3O5
P5	405	4.249	C19H20ClN3O5
P6	316	4.200	C16H17Cl2N3O
P7	248	0.654	C13H10FNO3
P8	180	4.499	C9H6ClNO

3.7. Toxicity of transformation products

The T.E.S.T. software was used to calculate and predict the biotoxicity of ENR and its potential intermediates, employing bioaccumulation factors, developmental toxicity, mutagenicity induction, and Daphnia magna 48 h-LC50 as assessment indicators. As shown in Figure 8, Figure 8(a) revealed that most intermediates exhibited weaker bioaccumulation than ENR, except for P6 and P7, which had higher bioaccumulation potential, particularly P6 with stronger bioaccumulation. However, the overall intermediates showed reduced bioaccumulation capacity. According to Figure 8(b), ENR was classified as ‘toxic’ in terms of developmental toxicity, whereas most degradation intermediates, except for P1, P4, P5, and P7, exhibited lower toxicity compared to ENR. Most intermediates were positively mutagenic (Figure 8(c)), albeit less toxic than ENR, indicating even low-accumulating products like P1, P4, and P5 posed environmental risks post-treatment. The 48 h-LC50 for Daphnia magna indicated acute toxicity, and as shown in Figure 8(d), only P6 exhibited higher acute toxicity than ENR. Overall, most degradation intermediates were less toxic than ENR, but the bioaccumulation and acute toxicity of P6 were higher and warranted attention.

Figure 8. Ecotoxicity analysis of ENR and its intermediate products.

4. Conclusions

The Fe(III)-NTA/PMS/DBD oxidation system constructed in this study can effectively and stably remove ENR, and this study can draw the following conclusions:

1. Compared to other systems, the Fe(III)-NTA/PMS/DBD system increased the degradation rate of ENR to 92%. The optimal conditions obtained in the experiments were: Fe(III)-NTA molar ratio of 1:1, Fe(III)-NTA concentration of 0.2 mM, PMS concentration of 0.3 g/L, PMS concentration of 50 mg/L, and initial solution pH of 3.5.

2. Coexisting substances in surface water, such as HA, Cl⁻, and CO₃^2-, all had inhibitory effects on the degradation of ENR, with CO₃^2- having the most significant inhibitory effect.

3. Free radicals ·OH, SO₄·⁻, and ·O₂⁻ all participated in the degradation of ENR, with their contribution rates being in the order of OH > SO₄·⁻>·O₂⁻.

4. HPLC-MS identified the products formed during the degradation of ENR, indicating three potential degradation pathways: ① Defluorination; ② Pyrazole ring opening; ③ Hydroxyl substitution. The toxicity analysis shows that, the degradation products are difficult to be bio-accumulate and have lower developmental toxicity， mutagenicity, and exhibit weaker acute toxicity.

Disclosure statement

No potential conflict of interest was reported by the author(s).