Mechanistic Insights into Radical-Mediated Moxifloxacin Degradation Using Ultrasound-Assisted Persulfate Activation by Iron-Rich Soil

Abstract

Fluoroquinolones are a major issue in aquatic ecosystems due to their persistence, potential to induce antibiotic resistance, and inability to be effectively removed using conventional treatment methods. Several advanced oxidation processes have been studied for their degradation; however, there is still a lack of knowledge about their degradation mechanisms and the precise roles played by reactive species. In this context, the study investigated the heterogeneous activation of persulfate (PS) to degrade fluoroquinolones (FQs), such as moxifloxacin (MFX), in iron-rich soil (Cat) under ultrasound irradiation (US). The analysis of the soil catalyst revealed the presence of quartz (35%), iron oxides (33%), and alumina (26%) as the predominant constituents of the sample. The mineral phase analysis indicated the presence of magnetite, hematite, and alumina. Then, the outcomes of the specific surface area, micropore volume, and total pore volume were determined to be 19 m² g⁻¹, 6 m³ g⁻¹ and 9.10 m³ g⁻¹, respectively. The MFX/PS/US/Cat system demonstrated 89% degradation and 56% mineralization after 300 min. However, the optimized concentrations of i-PrOH, t-BuOH, and CHCl₃ were 50, 100, and 50 mM, respectively, in order to trap the radicals SO₄^•−, OH^•, and O₂^•−. The study examined the individual contributions of SO₄^•−, OH^•, and O₂^•− radicals to the overall process of MFX degradation. The results indicated that SO₄^•− was the primary radical, with a contribution of 52%, followed by OH^• with 43%, and O₂^•− with 5%. Finally, the investigation revealed that laterite exhibited both good catalytic activity and reusability over several cycles. The development of this new process could stimulate the creation of cost-effective technology for water remediation through the effective removal of fluoroquinolones.

1. Introduction

Moxifloxacin (MFX), a synthetic fluoroquinolone antibiotic, is widely employed for the treatment of diverse bacterial infections due to its broad-spectrum efficacy [1,2,3]. Reported concentrations of fluoroquinolones (FQs) in aquatic environments range from 3 ng L⁻¹ to 240 µg L⁻¹ in hospital effluents and from 0.5 ng L⁻¹ to 6.5 mg L⁻¹ in surface waters, depending on geographical location and sampling period [4]. However, the extensive usage of MFX has contributed to its ubiquitous occurrence in aquatic environments, particularly in wastewater systems [5]. This persistence raises serious concerns due to the potential development of antibiotic-resistant bacteria and ecological disruption [6,7,8]. Therefore, the effective removal of MFX from wastewater represents a significant environmental challenge.

Conventional wastewater treatment processes are inefficient at eliminating such recalcitrant pharmaceutical pollutants, necessitating the development of advanced treatment strategies [9,10,11]. Among them, advanced oxidation processes (AOPs), which are characterized by the generation of highly reactive species, have shown great potential for the degradation of persistent organic contaminants [12,13,14]. Specifically, the persulfate/iron (PS/Fe) system is a promising technology because of its capacity to produce both sulfate (SO₄^•−) and hydroxyl (OH^•) radicals. These oxidants are capable of effectively transforming complex antibiotic molecules into less harmful by-products [15,16,17]. A critical step in optimizing the PS/Fe system lies in evaluating the role of selective radical quenchers (scavengers). The systematic variation in scavenger concentrations (e.g., tert-butanol, isopropanol, chloroform) allows for clarification of the contribution of SO₄^•−, OH^•, and O₂^•− radicals to pollutant degradation. Understanding these contributions is essential for enhancing mineralization efficiency and for designing sustainable water remediation technologies [18,19].

This study, therefore, aimed to assess the actual contribution of radical species under optimized operating conditions (i.e., pollutant concentration, catalyst dose, and oxidant concentration were optimized). Specifically, scavenger concentrations of chloroform (Chl), tert-Butanol (t-BuOH), and isopropanol (i-ProH), responsible for trapping superoxide or hydroperoxides, sulfate anions, and hydroxyl radicals, respectively, were varied under optimized operating conditions. The percentage contribution of each radical species to MFX degradation was then quantified.

2. Results and Discussion

2.1. Laterite Characterization

2.1.1. Morphology and Composition

The XRF analysis revealed quartz (35.99%), iron oxides (33.59%), and alumina (26.97%) as the predominant constituents in the sample (Figure 1b). This finding was corroborated by XRD, which confirmed the presence of magnetite, hematite, and alumina (see XRD spectrum, Figure 1a). Indeed, a significant proportion of ferric iron would contribute to the oxidation–reduction process of the oxidation mechanism [20]. SEM/EDS analysis further supported these results, showing the presence of elements such as iron, silica, potassium, aluminum, titanium, and oxygen (see EDS spectrum and image) (Figure 1c,d). Overall, the most prevalent phases identified in this study are consistent with those reported in previous works [21,22].

2.1.2. Surface Texture and Properties

As demonstrated in Figure 2a, the specific surface area, micropore volume, and total pore volume were found to be 19 m² g⁻¹, 6 m³ g⁻¹, and 5 cm³ g⁻¹, respectively, based on the multipoint adsorption isotherm of N₂. Subsequently, the BJH Adsorption Pore Distribution method was employed, revealing an average pore diameter of 1.19 nm (see Figure 2b). All of the above demonstrate that the adsorption phenomenon is minimal in the pollutant elimination process.

2.2. Oxidation and Adsorption Processes

As illustrated in Figure 3, the removal of MFX was evaluated under different systems as a function of time. The single systems (MFX/US, MFX/Cat, and MFX/PS) achieved only about 10% removal. When combined in dual systems (MFX/US/Cat, MFX/PS/Cat, and MFX/US/PS), the elimination efficiency remained below 20%. In contrast, the complete MFX/PS/US/Cat system achieved almost total removal, reaching ~92% after 300 min. To distinguish between degradation and adsorption, a desorption test was performed by increasing the pH of the reaction medium to 12 after the final sampling. Results indicated a degradation rate of 89% and a mineralization extent of 56%, leaving approximately 3% of MFX adsorbed on the catalyst surface. This low adsorption of the pollutant makes sense because of the low surface area (approximately 19 m²/g), which indicates that degradation controls the elimination process. The enhanced efficiency of laterite in conjunction with ultrasound irradiation can be attributed to the increased generation of reactive oxygen species (SO₄^•−, OH^•, and HO₂^•/O₂^•−). These radicals are produced from persulfate activation under the redox cycling of surface-bound ≡Fe^III+/≡Fe^II+ on laterite, as previously reported [23,24]. This process is promoted by sonication activity. The following equation is derived:

≡Fe^III+ + H₂O → ≡Fe^II+ + OH^• + H⁺

(1)

It has been demonstrated that the generation of ≡Fe^(II) can be achieved through the process of sono-reduction of Fe^(III) sites present on the surfaces of laterite. After this, a reaction occurs between the generated Fe^(II) and PS, resulting in the formation of SO₄^•. The subsequent sequence of events is as follows [23,24]:

≡Fe^II + S₂O₈²⁻ → ≡Fe^III + SO₄^•− + SO₄²⁻ k = 2.7 × 10¹ M⁻¹ s⁻¹

(2)

Then, OH^• could also be produced when SO₄^•− reacts with H₂O/HO^• [25,26,27]:

SO₄^•− + H₂O → SO₄²⁻ + OH^• + H⁺   k [H₂O] = 1.3 × 10³ s ⁻¹

(3)

SO₄^•− + OH⁻ → SO₄²⁻ + OH^• (pH > 8.5) k = 6.5 ± 1 × 10⁷ M ⁻¹ s ⁻¹

(4)

Furthermore, the formation of SO₄^•− and OH^• can also be attributed to chain reactions predominantly governed by the superoxide anion radicals (O₂^•−) [28]:

≡Fe^II + O₂ → ≡Fe^III + O₂⁻

(5)

S₂O₈²⁻ + O₂^•− → SO₄^•− + O₂ + SO₄²⁻

(6)

2O₂^•− + 2H⁺ → H₂O₂ + O₂

(7)

≡Fe^II + H₂O₂ → ≡Fe^III + OH^• + OH⁻

(8)

2.3. Chloroform Effect on MFX Degradation

Figure 4 illustrates the effect of chloroform (Chl) concentration on MFX degradation kinetics. The degradation rate was 92 ± 2% without scavengers after 300 min, while the inhibition effects at 10, 50, and 100 mM of Chl were 5%, 7%, and 8 ± 2%, respectively, resulting in a reduced degradation rate to 84 ± 2%. The mineralization rate (i.e., [TOC]/[TOC]₀) accordingly decreased from 56 ± 2% to 48 ± 2% after 300 min. This inhibition corresponded to the reduced contribution of superoxide radicals to MFX oxidation. In contrast, superoxide (O₂^•−), sulfate (SO₄^•−), and hydroxyl (OH^•) radicals are known to react with chloroform at very high second-order rate constants. Notably, chloroform exhibits high reactivity with O₂^•− (k_CHCl3/O2^•− = 2.3 × 10⁸ M⁻¹ S⁻¹) [29] and low reactivity with OH^• (k_CHCl3/OH^• = 5.4 × 10⁷ M⁻¹ S⁻¹) [30] and SO₄^•− (k_CHCl3/SO4•⁻ = 1.5 × 10⁵ M⁻¹ S⁻¹) [31]. It was therefore determined that the optimum concentration of chloroform required to quench the superoxide radicals was 50 mM. Indeed, the concentration of Chl used to trap superoxide radicals can vary, with some researchers opting for 10 mM and others choosing 5 mM.

2.4. Tert-Butanol Impact on MFX Degradation

As illustrated in Figure 5, the impact of tert-Butanol (t-BuOH) concentration on the degradation kinetics of MFX is evident. The results showed that the degradation rate was 92 ± 2% without scavengers after 300 min, while the inhibition effects at 10, 50, and 100 mM of t-BuOH were 15%, 26%, and 30 ± 2%, respectively, resulting in a reduced degradation rate of 62 ± 2%. The mineralization rate (i.e., [TOC]/[TOC]₀) also decreased from 56 ± 2% to 40 ± 2% after 300 min. This reduction in degradation efficiency was attributed to the scavenging of hydroxyl radicals (OH^•) by t-BuOH [16,30]. Despite the high reaction rate between OH^• and MFX, t-BuOH exhibits a stronger reactivity with hydroxyl radicals (OH^•) (i.e., k_t-BuOH/OH^• = (3.8–7.6) × 10⁸ M⁻¹ S⁻¹) [23] and a weaker reactivity with sulfate radicals (SO₄^•−) (i.e., k_t-BuOH/SO4^•− = (4–9.1) × 10⁵ M⁻¹ S⁻¹) [32]. However, t-BuOH does not react with O₂^•− [29]. The best concentration of t-BuOH needed to stop the hydroxyl radicals was found to be 100 mM.

2.5. Isopropanol Effect on MFX Degradation

Figure 6 illustrates the effect of varying isopropanol (i-PrOH) concentration on the degradation kinetics of MFX. Without scavengers, the degradation rate is 92 ± 2% after 300 min. Meanwhile, the inhibition rates of 10, 50, and 100 mM i-PrOH are 5%, 82%, and 84% ± 2%, respectively. This demonstrated that the degradation rate has decreased by only 8% ± 2%. The mineralization rate accordingly fell from 56 ± 2% to 3 ± 2% after 300 min. This strong inhibition was attributed to the simultaneous scavenging of both sulfate (SO₄^•−) and hydroxyl (OH^•) radicals by i-PrOH. Indeed, i-PrOH simultaneously traps sulfate and hydroxyl radicals with high rate constants, i.e., k_i-PrOH/HO^•= 1.9 × 10⁹ M⁻¹ S⁻¹ [30] et k_i-PrOH/SO4^•− = 7.42 × 10⁷ M⁻¹ S⁻¹ [33]. Unlike the other radicals above-mentioned, the reactivity between i-PrOH and O₂^•− is low in the range of 1.0 × 10⁶ M⁻¹ S⁻¹ [34]. It should be noted that the concentration of i-PrOH required to trap radicals is 50 mM.

2.6. Scavenging Tests on MFX Degradation

Figure 7 illustrates the degradation and inhibition percentages of MFX as a function of various scavengers, as well as the relative contribution of individual radicals to the overall degradation. The overall degradation efficiency was 89%, while the total elimination was 92%, indicating approximately 3% of MFX removal was due to adsorption (sorption). After the addition of i-PrOH, only 5% degradation of MFX was observed, while 84% was inhibited. In the case of t-BuOH, 51% of degradation and 38% of inhibition were recorded. Similarly, in the presence of Chl, 73% degradation and 16% inhibition were observed (Figure 7a,b). Based on these results, the relative degradation percentages assigned to the main reactive species O₂^•−, OH^•, and SO₄^•− were 5%, 38%, and 46%, respectively. These contributions were further refined to 5.62%, 42.7%, and 51.69%, respectively, indicating their quantitative roles in overall MFX degradation (89%) (see Figure 7c) [16].

2.7. Summary of Second-Order Rate Constants of ROS Relative to Some Scavengers

A categorized list of scavengers that can be used for different radicals is presented in

Table 1. Second-order rate constants of scavengers towards different radicals.

Compounds	Formular	pKa	MW	M−1 s−1
				kO2•−	kOH•	kSO4•−
Chloroform	CHCl3	15.5	119.37	2.3 × 108	5.4 × 107	1.5 × 105
Isopropanol	C3H80	16.5	60.10	1.0 × 106	1.9 × 109	7.42 × 107
Terbutanol	C4H10O	16.54	74.12	-	(3.8–7.6) × 108	(4–9.1) × 105
Methanol	CH3OH	15.54	32.04	-	9.7 × 108	2.5 × 107
Phenol	C6H6O	10.0	94.11	-	6.0 × 108	-
Benzoquinone	C6H4O2	9.91	108.10	9.6 × 108	7.4 × 106	-
Potassium iodide	KI	0.06	166	-	1.1 × 1010	>1.0 × 1010

. The scavenger’s effectiveness depends on its second-order rate constant towards the radicals in the degradation process [31,32,33,34,35,36].

Among these scavengers, Chl and BQ are used to catch superoxides due to their high reactivity, while t-BuOH and i-PhOH commonly capture hydroxyl radicals. Subsequently, i-PrOH and MeOH, as well as KI, serve to trap sulfate and hydroxyl radicals in a simultaneous manner [37,38,39,40,41,42]. According to literature [43], MFX (m/z = 401) is capable of undergoing decomposition into either m/z = 360 and m/z = 320, or m/z = 337 and m/z = 125 and m/z = 101 (Figure 8).

2.8. Effect of Hospital Wastewater Components on Removal Efficiency

The kinetics of MFX removal from Hospital Wastewater (HW) and Ultra Pure Water (UPW) are illustrated in Figure 7. A reduction in elimination performance was observed for HW. The inhibitory effect in HW can be ascribed to the abundant dissolved components in HW, such as chlorides, phosphates, nitrates, sulfates, and dissolved organic matter (Figure 9). It is important to note that these wastewater components exist in very high concentrations compared to the target compounds (e.g., the molar ratio of [Phosphate]/[MFX] = 3 and [Chloride]/[MFX] = 4). This may produce strong competitive and/or radical scavenging impacts [44]. Indeed, phosphates have a strong inhibitory effect, which can be explained by their ability to adsorb strongly onto metal oxides, reducing the interaction between the oxidant and the solid surface. Phosphate really sticks to Fe^III oxides, through monodentate and/or multidentate–mononuclear surface complexes. There are a bunch of different models that try to explain how phosphate is eliminated by mineral surfaces [45]. Additionally, phosphate can also effectively react to SO₄^•− and OH^• according to the following equations [30,33]:

SO₄^•− + HPO₄²⁻ → HPO₄^•− + SO₄²⁻  k = 1.2 × 10⁶ M⁻¹ s⁻¹

(9)

OH^• + HPO₄²⁻ → HPO₄^•− + OH⁻   k = 1.5 × 10⁵ M⁻¹ s⁻¹

(10)

2.9. Stability and Reusability of the Catalyst

From an eco-economic point of view, whether the catalyst can be used more than once is a key factor in its use in large-scale reactors. The reusability of the catalyst was then subjected to rigorous testing involving successive oxidation cycles, with the optimal conditions previously outlined (Figure 10). At the end of each oxidation process, the solid is carefully removed from the reactor. It is then washed with very pure water. After that, it dries at 50 °C overnight. Finally, it is used in the next experiment. The results demonstrated that the catalyst exhibited slight variations when used for the second time, yet retained its activity after five cycles. The removal yield decreased from 92% ± 2% to 86% ± 2% from the first oxidation cycle to the last cycle, representing a decrease of 6% ± 2%. The TOC then confirmed that the system was very stable, which was also shown before by US/PS/Cat. This showed that the mineralization (56 ± 2%) stayed the same even after five oxidation cycles [46,47]. The catalytic activity is very stable because not much iron is lost during the oxidation cycles, and the structure of the solid is stable [48,49].

3. Materials and Methods

3.1. Chemicals

All chemicals were used at analytical grade without further purification. Moxifloxacin (MFX, C₂₁H₂₄FN₃O₄, ≥99% purity) and potassium persulfate (K₂S₂O₈, ≥99% purity) were obtained from Sigma-Aldrich (Saint-Quentin-Fallavier, France). Hydrochloric acid (HCl, 37% v/v, ≥98%), tert-Butanol (t-BuOH, C₄H₁₀O ≥ 99%), isopropanol (i-PrOH, C₃H₈O, ≥99.8%), and chloroform (Chl, CHCl₃, ≥99.5%) were also purchased from Sigma-Aldrich (Saint-Quentin-Fallavier, France). Other chemicals, including sodium hydroxide (NaOH, ≥97%), acetonitrile (CH₃CN, ≥99.99%), and formic acid (≥99%) were supplied via Sigma-Aldrich (Merck), St. Louis, MO, USA. FerroVer^® iron reagent pillows and the LCK 380 kit were obtained from Hach Company (Loveland, CO, USA).

3.2. Catalyst Soil and Characterization Methods

Laterite soil, found in tropical regions that are both humid and hot, was used as a catalyst in this study. The soil was collected from Man, western Côte d’Ivoire, at coordinates 5°44′06″ N and 3°24′12″ W. Samples were crushed to less than 50 µm mesh size, washed with ultrapure water (UPW) three times, and dried at 50 °C. The chemical composition was determined by X-ray fluorescence (XRF, Horiba MESA-50, Tokyo, Japan). Morphology and surface microanalysis were examined by SEM–EDX (Hirox, SH4000 M, Tokyo, Japan). Crystalline phases were identified via X-ray diffraction (GBC Emma, Hampshire, IL, USA). The specific surface area is measured using a Brunauer–Emmett–Teller (BET) analyzer (V-Sorb 2800, Gold APP Instruments, Beijing, China). Zeta potential is measured using a Zeta Compact CAD, Naucelle, France. Ultrapure water (resistivity 18.2 MΩ cm, Millipore Milli-Q, Loveland, CO, USA) was used for all experiments. The Regional Public Hospital Establishment (RPHE) of Man is the Public Regional Hospital of Man City in Côte d’Ivoire. The RPHE of Man encompasses a wide spectrum of clinical services and health requirements, boasting a substantial capacity for hospitalization (131 beds; 20,000 admissions and 50,000 consultations in 2025) across diverse specialties, including medicine, surgery, obstetrics, and gynecology. The Hospital Wastewater (HW) was supplied by RPHE. Consequently, the wastewater sample examined in this study is a suitable representation of the HW in Côte d’Ivoire. A tabular presentation has been provided which details the physico-chemical characteristics and inorganic species of HW (

Table 2. Physico-chemical characteristics and inorganic species of HW.

Parameters	Values
pH	6.91 ± 0.2
Turbidity (NTU)	186 ± 2
Conductivity (µS Cm−1)	1432 ± 5
Total Dissolved Solids (TDS) (mg L−1)	270 ± 5
Total Suspended Solids (TSS) (mg L−1)	110 ± 2
Carbon Oxygen Demand (COD) (mg O2 L−1)	196 ± 2
Nitrates (mg L−1)	69 ± 2
Chloride (mg L−1)	72 ± 2
Sulfate (mg L−1)	47 ± 3
Phosphate (mg L−1)	179 ± 5

).

3.3. Ultrasonic Oxidation Tests and Analytical Methods

The experimental setup consisted of a stainless-steel ultrasonic tank (10 L, 240 W, 40 kHz, 15 × 24 × 30 cm). Batch experiments were performed in 250 mL borosilicate reactors immersed in the ultrasonic bath under ambient temperature and neutral pH. Each reaction mixture contained pollutant solution (MFX), laterite catalyst, and PS oxidant. The solution pH was adjusted using HCl or NaOH (0.01–1 M) [11]. Two sets of pre-tests were conducted: (i) adsorption equilibrium of MFX onto laterite in the absence of light, and (ii) oxidation reaction under ultrasonic conditions with catalyst and oxidant. Samples (3 mL) were withdrawn every 30 min, filtered, and analyzed.

MFX concentrations were determined by HPLC (Waters 600 controller, binary pump, autosampler Waters 717 plus, UV detector Waters 2489 set at 294 nm, Milford, MA, USA). Separation was achieved using a reverse-phase C18 column (250 mm × 4.6 mm, 5 µm) under isocratic elution with acetonitrile/water (30:70 v/v) acidified with 0.1% formic acid at 1.0 mL min⁻¹. The degradation efficiency at time t was calculated by:

$$\mathrm{Degradation} (\%) ={\displaystyle \frac{C_0-C_t}{{\mathrm{C}}_0}}\ast 100$$

(11)

where C₀ and C_t are the initial and residual concentrations of MFX, respectively.

Leached iron concentrations were measured colorimetrically using a DR 3900 spectrophotometer (Hach, 10 mL cuvette tests), while total organic carbon (TOC) was measured using the same platform. Potential leaching of trace metals (Cr, Mn, Cu, Co, Ni, Zn) from the laterite was monitored by XRF [48,49,50,51]. To determine the role of SO₄^•−, OH^•, and HO₂^•/O₂^•− radicals in the MFX elimination process, quenching experiments were performed using Chl, i-PrOH, and t-BuOH, respectively. Indeed, i-PrOH and t-BuOH were used to scavenge SO₄^•− and OH^• radicals, while Chl was employed to trap HO₂^•/O₂^•− radicals [11,13,19]. Each scavenger concentration was varied from 10 to 100 mM. Experiments were conducted in triplicate, and results showed a high reproducibility with a relative standard deviation below 3%.

4. Conclusions

The objective of this research was to investigate how PS is activated heterogeneously and generate radicals to degrade MFX in iron-rich soil when subjected to ultrasound. XRF analysis of the soil catalyst revealed the presence of quartz (35%), iron oxides (33%), and alumina (26%) as the predominant constituents. Subsequently, XRD confirmed the presence of magnetite, hematite, and alumina. The MFX/PS/US/Cat system demonstrated an effective degradation of approximately 89% with a 56% mineralization after 300 min. An optimization for the concentration of i-PrOH (50 mM), t-BuOH (100 mM), and Chl (50 mM) was performed to selectively quench SO₄^•−, OH^•, and O₂^•− radicals, respectively. After that, the relative contributions of these radicals to MFX degradation have been investigated, with results indicating that SO₄^•− was the dominant reactive species. The contributions of SO₄^•−, OH^•, and O₂^•− were quantified as 51.69%, 42.7%, and 5.62%, respectively. This cost-effective process has the potential to facilitate the development of innovative technologies for water remediation and the removal of fluoroquinolones from contaminated sources.