Enhanced Ozone Oxidation by a Novel Fe/Mn@γ−Al₂O₃ Nanocatalyst: The Role of Hydroxyl Radical and Singlet Oxygen

Abstract

Catalytic ozonation is a potential alternative to address the dye wastewater effluent, and developing an effective catalyst for catalyzing ozone is desired. In this study, a novel Fe/Mn@γ−Al₂O₃ nanomaterial was prepared and successfully utilized for catalytic ozonation toward dye wastewater effluent components (dimethyl phthalate and 1−naphthol). The synthesized Fe/Mn@γ−Al₂O₃ exhibited superior activity in catalytic ozonation of dimethyl phthalate and 1−naphthol in contrast to Fe@γ−Al₂O₃ and Mn@γ−Al₂O₃. Quench and probe tests indicated that HO° contributed to almost all removal of dimethyl phthalate, whereas O₃, HO°, and singlet oxygen participated in the degradation of 1−naphthol in the Fe/Mn@γ−Al₂O₃/O₃ system. The results of XPS, FT−IR, and EPR suggested that HO° and singlet oxygen were generated from the valence variations of Fe(II/III)and Mn(III/IV). Moreover, the Fe/Mn@γ−Al₂O₃/O₃ system could also have excellent efficacy in actual water samples, including dye wastewater effluent. This study presents an efficient ozone catalyst to purify dye wastewater effluent and deepens the comprehension of the role and formation of reactive species involved in the catalytic ozonation system.

1. Introduction

Advanced oxidation processes (AOPs) gained much attention as powerful techniques to mineralize the pollutants in water treatments [1]. Catalytic ozonation is a promising technology for efficiently removing refractory pollutants, especially in the advanced treatment of wastewater. Heterogeneous catalytic ozonation can minimize the dissolving of toxic metal cations in contrast to homogeneous catalyzing.

Many metal oxides have been utilized in the heterogeneous catalytic ozonation process, including manganese oxides, iron oxides/oxyhydroxide, aluminum oxides, and bimetallic/polymetallic oxides. MnO_X supported by granular activated carbon or alumina catalyzed ozone to generate hydroxyl radical (HO°) and efficiently degrade benzenes [2,3]. The surface hydroxyl groups of hydroxylated synthetic ɑ−FeOOH can promote catalyzing ozone to yield HO° [4], and surface MeO−H weak bonds were the favorable sites for accelerating HO° generation. The complex of oxalic acid and iron (Fe₂O₃/Al₂O₃) reacted with ozone and accelerated the degradation of oxalic acid [5]. In addition to Mn and Fe, TiO₂ and MgO nanoparticles were also investigated and used to catalyze ozone [6,7].

Bimetallic nanostructured materials hold promise for improving catalyst activity and selectivity [8]. Mesoporous bimetallic catalysts, such as Ru−Cu/SBA−15, Co−Mn−MCM−41 catalysts, accelerated catalytic ozonation to generate more HO° to degrade dye industry effluent or dimethyl phthalate (DMP) in contrast to single metal loading [9,10]. Fe−Ni/activated carbon also exhibited high performance for degrading 2,4−dichlorophenoxyacetic acid compared to single metal or activated carbon [11]. The micron−sized Fe⁰/Cu/O₃ process enhanced p−nitrophenol mineralization more than twice the sum of ozone and Fe⁰/Cu alone [12]. The ratios of surface defective oxygen/lattice oxygen and Mn(III)/Mn(IV) of Mn−M bimetallic HZSM−5 (M: Fe, Cu, Ru, Ag) catalysts were higher than that of Mn−only catalyst, of which Ru−Mn/HZSM−5 showed the highest efficiency in degrading toluene [13,14].

Ozonation, radical oxidation, and nonradical reaction in the catalytic ozonation system mainly contributed to contaminant degradation [15,16,17]. In addition to HO°, reactive oxygen species (ROS), including singlet oxygen (¹O₂) and superoxide radical (O₂°⁻), were also involved in contaminant degradation during catalytic ozonation [18]. Oxygen vacancies and multi−valence Mn were the main Lewis acidic sites for Mn₂O₃/LaMnO_3−δ perovskite composites during catalytic ozonation, and the generated HO°, ¹O₂, and O₂°⁻ were possibly responsible for the degradation of 1H−benzotriazole [19]. As bimetallic nanoparticles and Fe− and Mn−based materials exhibited excellent performance on catalytic ozonation, however, Fe/Mn bimetallic material in catalytic ozonation has not been reported so far. Consequently, the performance of Fe/Mn bimetallic nanoparticles on catalyzing ozone and the corresponding reactive species in such systems needs further evaluation.

This study aims to investigate the performance of catalytic ozonation by synthesized Fe/Mn@γ−Al₂O₃ nanoparticles. A modified evaporation−induced self−assembly method was employed to fabricate Fe/Mn@γ−Al₂O₃ nanoparticles, and SEM, TEM, XRD, XPS, FTIR, and BET were utilized to characterize the morphology, crystal form, element distribution, and composition. DMP and 1−naphthol (1−NP), dye precursors and intermediates, were selected as model compounds. Their removal performance, reactive species, reaction mechanism, degradation pathway, and matrix factors were investigated in catalytic ozonation system in the presence of Fe/Mn@γ−Al₂O₃ nanoparticles.

2. Materials and Methods

2.1. Catalyst Preparation

Al₂O₃−based catalysts were prepared by a modified evaporation−induced self−assembly process as described previously for pure γ−Al₂O₃ [20]. Aluminum isopropoxide [Al(O_iPr)₃], Fe(NO₃)₃·9H₂O, and Mn(NO₃)₂ were selected as the sources of Al, Fe, and Mn. [Al(O_iPr)₃] (8.4 g), glucose (7.2 g), and a required dosage of Fe(NO₃)₃·9H₂O and Mn(NO₃)₂ were dissolved thoroughly in ultrapure water at 35 °C. The pH value of the mixture was adjusted to 5.5 using formic acid (10 wt.%), and the mixed solution was ultrasonicated for 4 h and heated at 105 °C. The final solid after grinding was calcined at 600 °C for 6 h, and a series of γ−Al₂O₃−based catalysts were obtained and termed as Fe@γ−Al₂O₃ Mn@γ−Al₂O₃ and Fe/Mn@γ−Al₂O₃.

2.2. Catalyst Characterization

The morphology and chemical composition of catalysts were characterized by transmission electron microscope (TEM, Talos L120C) and scanning electronic microscopy−energy dispersion spectroscopy (SEM−EDS, Merlin, Carl Zeiss AG, GER). The powder X−ray diffraction (XRD, Empyrean, NL) spectra of the obtained catalysts were collected with Cu−Kα radiation (λ = 1.5418 Å) at 40 kV and 40 mA operating from the 2θ angle of 10−90. The specific surface area and pore size distribution of the catalysts were calculated from the N₂ adsorption/desorption isotherms at 77 K (ASAP2020, Micromeritics, Norcross, GA, USA). X−ray photoelectron spectroscopy (XPS, EscaLab Xi⁺, ThemoFisher Scientific, Leicestershire, UK) spectra were recorded by using Al−Kα radiation at 15 kV and 51 W, and the binding energies were calibrated with C1s at 284.8 eV. Fourier transform infrared spectra (FT−IR) were recorded using a CCR−1 (Thermo−Nicolet, Waltham, MA, USA).

2.3. Catalytic Ozonation Procedure

The catalytic ozonation experiments of target contaminants were performed in a 1.2 L column reactor at a temperature of 25 °C. Ozone was produced in situ from pure oxygen by a ZA−1G laboratory ozone generator (Guangzhou Zeao Ozone Equipment Co., Ltd., Guangzhou, China) and continuously bubbled into the reactor through the reactor’s porous plate at a bottom 200 mL min⁻¹ flow rate (6 mg L⁻¹). Catalyst powder (200 mg L⁻¹) was mixed in the reactor under continuously magnetically stirring. Target compounds (DMP/1−NP, 50 mM) with/without quenchers (p−benzoquinone (p−BQ), nitrobenzene (NB), furfuryl alcohol (FFA)) or probes (NB, FFA) were added in the catalytic ozonation system. The same procedures were carried out for the control experiments, including ozone alone, catalyst without ozone, and catalyst with oxygen. All solutions were prepared with ultrapure water (≥18.2 MΩ×cm). Samples were taken at regular intervals, filtered by a Millipore filter (0.45 μm), and quenched by Na₂S₂O₃, then DMP, 1−NP, NB, and FFA were analyzed. All experiments were conducted in duplicate.

2.4. Analytic Methods

DMP, 1−NP, NB, and FFA concentrations were quantified using high−performance liquid chromatography (UPLC, Agilent 1200) equipped with a PDA 2998 detector and a C−18 column (Agilent Eclipse, 5 μm, 4.6 mm × 250 mm). The mobile phase was acetonitrile/0.02% phosphoric acid water (60:40, v/v) with a flow rate of 0.1 mL min⁻¹. The analytical wavelengths of DMP, 1−NP, NB, and FFA were set at 228 nm, 210 nm, 265 nm, and 204 nm at 30°C. The electron paramagnetic resonance (EPR) spectra were measured by Bruker EMX plus instrument (Karlsruhe, Germany). The compound 2,2,6,6−Tetramethylpiperidine (TEMP) was used to capture ¹O_2, and 5,5−dimethyl−1−pyrroline−N−oxide (DMPO) was applied to capture HO° and O₂°⁻. In detail, the samples for EPR tests were dispersed in a 100 mM DMPO/TEMP solution (2 mL) with O₃ purging, and the solvent was H₂O for trapping ¹O₂ and HO° and methanol for trapping O₂°⁻. The concentration of leached metals was determined by graphite furnace atomic absorption spectrometry (AAS, Pin AAcle 900T, Perkin Elmer, Waltham, MA, USA). The intermediates of DMP and 1−NP during catalytic ozonation were identified by using LC−MS (Thermo Fisher Scientific, Q Exactive Plus, USA).

3. Results

3.1. Characterization

A series of γ−Al₂O₃−based nanocatalysts were synthesized using the glucose template method [20]. SEM images (Figure 1A−D) show amorphous particle structure for all materials, and Fe/Mn@γ−Al₂O₃ was more compact compared to others. TEM images (Figure S1) displayed the nanoscale structure of Fe/Mn@γ−Al₂O₃ with a thickness of 5−10 nm. The element mapping of Fe/Mn@γ−Al₂O₃ indicated that four elements were distributed uniformly (Figure 1E,F), and the element composition of Fe/Mn was 1.3, which corresponded to the nominal loading molar ratio (1.5). Figure 1G shows the XRD patterns of γ−Al₂O₃−based nanomaterials, and five broad peaks (39.4°, 45.9°, 66.9°, 80.5°, 84.9° of 2θ) were observed and related to γ−Al₂O₃ phase [21]. The peaks of Fe/Mn oxide alloy (35° of 2θ), Fe₂O₃ (33° of 2θ), and MnO₂ (21° of 2θ) were not observed, which might be affected by γ−Al₂O₃ peaks [22]. FT−IR spectroscopy at approximately 575, 748, and 819 cm⁻¹ derived from the Al−O vibrations confirmed the formation of γ−Al₂O₃ (Figure S2) [23]. Additionally, XPS full spectra of γ−Al₂O₃−based nanomaterials identified O, Al, Fe, and Mn, indicating Fe and Mn oxides grew on the γ−Al₂O₃ (Figure S3).

The nitrogen sorption isotherm of Fe/Mn@γ−Al₂O₃ is shown in Figure 1H, exhibiting a type IV N₂ sorption isotherm with an H3 hysteresis loop, which is a typical characteristic of a mesoporous material [24]. The BET surface area, total pore volume, and median mesopore diameter of Fe/Mn@γ−Al₂O₃ were 304.2 m² g⁻¹, 0.416 cm³ g⁻¹, and 3.830 nm (Figure S4), which were similar to γ−Al₂O₃−based materials (Sasol product) [21]. This result indicates that the Fe−Mn catalyst was impregnated well into the γ−Al₂O₃ without pore blocking. Overall, the Fe/Mn@γ−Al₂O₃ catalyst was a mesoporous nanomaterial with uniform iron−manganese oxide growing on the surface.

3.2. Degradation of DMP and 1−NP

DMP and 1−NP, the precursor or intermediate of dyes, were selected as model compounds (Figure S5), and Figure 2 illustrates their degradation in the catalytic ozonation system. The observed first−order rate of 1−NP (4.13 × 10⁻³ s⁻¹) under ozonation alone was higher in contrast to DMP (6.64 × 10⁻⁴ s⁻¹) (Figure 2A,B). This result corresponds with the fact that the reactivity of DMP with O₃ alone was relatively low (0.2 M⁻¹ s⁻¹) compared with 1−NP (~10² M⁻¹ s⁻¹) (

Table 1. The estimated and experimental observed removal rate constants of DMP/1−NP.

Compound	Rate Constant (M−1 s−1)			Exp. k (s−1)	Est. k * (s−1)
	O3	HO°	1O2
DMP	0.2 [33]	3.7 × 109 #	<103 [34]	3.0 × 10−3	2.1 × 10−3
1−NP	~102 $	1.3 × 1010 [35]	7.6 × 106 [36]	3.3 × 10−2	2.2 × 10−2

* The observed rate constant of DMP and 1−NP (Est. k) were estimated as described in Text S1. ^# The second−order rate constant of HO° with DMP was determined using laser flash photolysis technology and competitive kinetics detailed in Text S2 and Figure S10. ^$ The second−order rate constant of O3 and 1−NP was estimated roughly based on the quench test and other naphthalenes.

). Note that the second−order rate of 1−NP with O₃ was estimated roughly using the quench test (Figure S6), and the value was in the range of ~10²–3 × 10³ M⁻¹ s⁻¹ for naphthalene, 2−methylnaphthalene, and 1−chloronaphthalene [25,26,27]. Similarly, the observed first−order rate constants of 1−NP under each catalytic ozonation treatment (i.e., Fe@γ−Al₂O₃/O₃, Mn@γ−Al₂O₃/O₃, Fe/Mn@γ−Al₂O₃/O₃) were higher than those of DMP. Specifically, the Fe/Mn@γ−Al₂O₃ catalytic was the most effective to catalyze ozone for removing DMP or 1−NP (2.04 × 10⁻³ s⁻¹ for DMP, 2.25 × 10⁻² s⁻¹ for 1−NP), followed by Fe@γ−Al₂O₃, Mn@γ−Al₂O₃. Figure S7 shows that the catalytic activity decreased less than 10% after five cycles of use, and the concentration of leached−out Fe and Mn increased slightly during the cycling use (Table S1), implying that Fe/Mn@γ−Al₂O₃ was relatively stable during the ozonation reactions. Overall, the catalytic ozonation by Fe/Mn@γ−Al₂O₃ catalytic was feasible and efficient in removing O₃−resistant compounds.

The quench and probe tests were conducted to investigate the reactive species involved in catalytic ozonation. As noted in previous studies, HO° played the dominant role in removing organic compounds [28,29,30,31]. Moreover, other ROS (e.g., ¹O₂, O₂°⁻) also participated in the chemical oxidation during catalytic ozonation [32]. Thus, two model compounds with different reactivity toward HO° and ROS (Table 1) were selected to distinguish their roles in the contaminant removal during catalytic ozonation.

NB, FFA, and p−BQ at a high level (100 mM) were chosen to selectively quench HO°, ¹O₂, and O₂°⁻, and their second−order rate constants are listed in Table S2. Figure 2C shows that NB can inhibit most DMP removal (> 99%), and the removal efficiency of DMP with the presence of FFA or p−BQ was almost the same as that with the presence of NB. This result indicated that HO° was the main contributor to DMP degradation under the Fe/Mn@γ−Al₂O₃/O₃ system, whereas ¹O₂ or O₂°⁻ was not involved in DMP degradation. Being different from DMP, the removal of 1−NP was not inhibited by NB completely (Figure 2C), and the removal rate constant decreased by 50% in contrast to that without NB. Both FFA and p−BQ almost inhibited the degradation of 1−NP completely. This result indicated that in addition to HO°, ¹O₂, or O₂°⁻ also played a significant role in the degradation of 1−NP. EPR spectroscopy was employed to characterize ROS and demonstrated that HO°,¹O₂, and O₂°⁻ were yielded under the Fe/Mn@γ−Al₂O₃/O₃ system, but the signal of O₂°⁻ was relatively low (detailed in Section 3.3).

The probe test was conducted to specify the contributions of diverse reactive species to DMP and 1−NP degradation under the Fe/Mn@γ−Al₂O₃/O₃ system. NB and FFA were selected as the probes of HO° and ¹O₂, and the modeled steady−state concentrations of HO° and ¹O₂ ([HO°]_SS, [¹O₂]_SS) are listed in Table S3. The estimated observed first−order rate constants of DMP (3.0 × 10⁻³ s⁻¹) or 1−NP (3.3 × 10⁻² s⁻¹) were calculated based on Equations (1)–(3), which were nearly the same as the experimental ones (2.1 × 10⁻³ s⁻¹ or 2.2 × 10⁻² s⁻¹) (Figure 2D).

$${\mathrm{k}}^{\prime }{=\mathrm{k}}_{{\mathrm{O}}_3 }^{\prime }{+\mathrm{k}}_{{\mathrm{HO}}^{\text{°} }}^{\prime }{+\mathrm{k}}_{{}_{}{}^1\mathrm{O}{}_2}^{\prime }$$

(1)

$${\mathrm{k}}^{\prime }{=\mathrm{k}}_{{\mathrm{O}}_3,\mathrm{DMP}}{[\mathrm{O}}_3{]+\mathrm{k}}_{{\mathrm{HO}}^{°},\mathrm{DMP}}{[\mathrm{HO}}^{°}{]+\mathrm{k}}_{{}_{}{}^1\mathrm{O}{}_2,\mathrm{DMP}}\left[{}_{}{}^1\mathrm{O}{}_2\right]$$

(2)

$${\mathrm{k}}^{\prime }{=\mathrm{k}}_{{\mathrm{O}}_3,1-\mathrm{NP}}{[\mathrm{O}}_3{]+\mathrm{k}}_{{\mathrm{HO}}^{°},1-\mathrm{NP}}{[\mathrm{HO}}^{°}{]+\mathrm{k}}_{{}_{}{}^1\mathrm{O}{}_2,1-\mathrm{NP}}\left[{}_{}{}^1\mathrm{O}{}_2\right]$$

(3)

Specifically, [HO°]_SS and [¹O₂]_SS were 8.1 × 10⁻¹³ M and 1.8 × 10⁻⁹ M during the degradation of DMP in catalytic ozonation system. The estimated first−order rate constant of DMP with HO° was 3.0 × 10⁻³ s⁻¹. Note that [¹O₂]_SS was as high as [HO°]_SS but can rarely contribute to the degradation of DMP (1.8 × 10⁻⁶ s⁻¹) due to the relatively low reactivity (${\mathrm{k}}_{{}_{}{}^1\mathrm{O}{}_2,\mathrm{DMP}}$< 10³ M⁻¹ s⁻¹). The DMP degradation during catalytic ozonation was consequent on the HO°−mediated oxidation. As for the 1−NP−involved system, [HO°]_SS and [¹O₂]_SS were 1.2 × 10⁻¹² M and 2.0 × 10⁻⁹ M (Table S3), which were comparable with the DMP involved system. However, the degradation of 1−NP was not only contributed to by HO°, but ¹O₂ and O₃ also participated in its removal. The estimated rate constant of each reactive species (${\mathrm{k}}_{{\mathrm{O}}_3,1-\mathrm{NP}}^{\prime }$, ${\mathrm{k}}_{{\mathrm{HO}}^{°},1-\mathrm{NP}}^{\prime }$, ${\mathrm{k}}_{{}_{}{}^1\mathrm{O}{}_2,1-\mathrm{NP}}^{\prime }$) accounted for 7.51%, 46.85%, and 45.65% of the total 1−NP removal rate constant, respectively.

The above results suggested that the Fe/Mn@γ−Al₂O₃/O₃ system can provide comparable levels of HO° (10⁻¹² − 10⁻¹³ M) and ¹O₂ (10⁻⁹ M) that were involved in removing different pollutants. HO° (E⁰ = 2.7 V) is non−selective thus reacts with both DMP and 1−NP at considerably high second−order rate constants (10⁹ − 10¹⁰ M⁻¹ s⁻¹) [37,38,39,40]. O₃ and ¹O₂ are more selective to naphthalenes (e.g., naproxen and propranolol) than DMP. Therefore, the inherent O₃ and the yielded ¹O₂ in the Fe/Mn@γ−Al₂O₃/O₃ system made contributions to the degradation of 1−NP (> 40%).

3.3. Catalytic Ozonation Mechanism

HO° and ¹O₂ were identified and involved in degrading pollutants in the Fe/Mn@γ−Al₂O₃/O₃ system. XPS and EPR spectra were utilized to explore their formation and catalytic mechanisms in such a system.

The XPS spectra of Fe 2p, Mn 2p, and O 1s for Fe/Mn@γ−Al₂O₃ before/after catalytic ozonation are shown in Figure 3A−C. As described in previous studies, the Fe 2p_3/2 and Fe 2p_1/2 peaks could be deconvoluted into three peaks, including Fe(II) (711.7 eV and 725.3 eV), Fe(III) (712.9 and 726. eV), and satellite (719.6 and 732.8 eV), respectively (Figure 3A) [41,42]. The Mn 2p_3/2 and Mn 2p_1/2 peaks could be deconvoluted into two peaks, including Mn(III) (642.2 eV and 653.9 eV) and Mn(IV) (643.8 eV and 654.8 eV) (Figure 3B) [22,43]. The contents of the high valence of both Fe and Mn (i.e., Fe(III) and Mn(IV)) increased slightly after ozonation (Table S4), indicating the conversion between Fe(II))/Mn(III) and Fe(III)/Mn(IV) in the Fe/Mn@γ−Al₂O₃/O₃ system. The previous studies also found valence variation of doping metals during ozonation [44,45]. The formation of high−valence metals resulted in the transformation of oxygen vacancy (V_O) to lattice oxygen (O_lat) due to the strong storing/releasing oxygen nature of the metal redox couple [46]. Moreover, oxygen mobility by electron transfer caused the valence variation of doping metals [47,48,49].

Three characteristic peaks were identified in the O 1s spectrum of Fe/Mn@γ−Al₂O₃ and regarded as adsorbed oxygen (O_ads, 530.3 eV), O_lat, (532.4 eV), and oxygen of metal oxide (O_MO, 531.2 eV), respectively (Figure 3C) [42,50]. After ozonation, the relative contents of O_lat/O_ads significantly increased from 1.32 to 1.64 (Table S4), whereas a subtle increase in that of Mn(IV) or Fe(II) was observed, indicating the prominent role of O_ads in reacting with pollutants rather than high valence metal. Specifically, the increase in O_lat contributed to the metal valence change and the formation of Mn(IV) or Fe(III). However, O_ads declined slightly after ozonation, which might have resulted from the replenishment of V_O [51,52]. Overall, the continuous catalytic activity of Fe/Mn@γ−Al₂O₃ originated from the interaction among V_O, O_lat, and the transformation of Fe(II)/(III)and Mn(III)/(IV) redox couple.

Solid EPR was applied to directly and accurately detect the unpaired electron information of Fe/Mn@γ−Al₂O₃ for identifying V_O (Figure 3D). Note that the EPR signal of Fe/Mn@γ−Al₂O₃ at g = 2.095 was more intensive than others, demonstrating that the introduction of Fe and Mn led to more formation of V_O on the Fe/Mn@γ−Al₂O₃ surface [45]. V_O contained many unpaired electrons and preferred to serve as active sites for the adsorption of oxygen (e.g., molecule O₂, molecule O₃), thereby facilitating the formation of ROS (HO°, ¹O₂, O₂°⁻, etc.) [53].

The reactive species in the Fe/Mn@γ−Al₂O₃/O₃ system was identified by EPR tests, as illustrated in Figure 3D. Generally, DMPO and TEMP are typical trapping chemicals for HO°, O₂°⁻, and ¹O₂. The characteristic signal of DMPO−HO° (A_H = A_N = 14.9 G), with a peak intensity ratio of 1:2:2:1, was obtained in both ozonation alone and Fe/Mn@γ−Al₂O₃/O₃ systems, but no signal was observed in pure water. Moreover, the intensity of DMPO−HO° adduct elevated significantly under the Fe/Mn@γ−Al₂O₃/O₃ system, implying Fe/Mn@γ−Al₂O₃ dramatically enhanced ozone decomposition to generate more HO°. TEMP was applied as the capture agent for ¹O₂. The typical signal of TEMP−¹O₂ (A_N = 16.9 G) in the Fe/Mn@γ−Al₂O₃/O₃ catalytic ozonation system was more intensive than that in the ozonation alone system. The results implied that ¹O₂ was generated in the ozone−based systems and enhanced in the Fe/Mn@γ−Al₂O₃ catalytic ozonation process. O₂°⁻ in the ozonation alone and Fe/Mn@γ−Al₂O₃/O₃ systems were also captured using DMPO to form DMPO−O₂°⁻ (A_H = 10.2 G,  A_N = 12.9 G), but the intensity was very weak and not promoted in the catalytic system, indicating the neglectable role of ¹O₂ in the catalytic ozonation system.

As illustrated in Figure S2, FT−IR spectroscopy of Fe/Mn@γ−Al₂O₃ at ~3460 cm⁻¹ and ~1634 cm⁻¹ corresponded to the surface hydroxyl group and the chemisorbed water, respectively [24,54]. Metal oxides adsorbed water molecules further dissociated to hydroxyls, generating the surface hydroxyl groups at Lewis acid sites of metal oxide surfaces [55,56,57,58,59]. The surface hydroxyl group functioned as Brønsted acid for catalytic ozone decomposition and promoted HO° generation [55,56,57,58,59].

The catalytic ozonation mechanism of Fe/Mn@γ−Al₂O₃ was proposed and illustrated in Figure 3F. First, O₃ was adsorbed at the surface hydroxyl group of metal oxide and interacted with Vo. The V_O was replenished with O donated by O₃ and produced an O₂ molecule releasing into the environment and active oxygen (O²⁻) to form adsorbed oxygen (O_ads) [60,61]. The O²⁻ further reacted with O₃ to generate O₂²⁻ (peroxide) [50]. The yielded O²⁻/O₂²⁻ (i.e., O_ads) then reacted with water molecules to generate ROS, including HO°, ¹O₂, and O₂°⁻ via charge−transfer interactions. Thereinto, O₂°⁻ underwent self−quenching to form ¹O₂ or react with O₃ to yield HO° (k = 1.6 × 10⁹ M⁻¹ s⁻¹) [62]. V_O also converted to O_lat, leading to the metal valence increase. The interaction among V_O, Fe(II/III)/Mn(III/IV), and O_lat of Fe/Mn@γ−Al₂O₃ resulted in the circulation of metal valence, thus constantly catalyzing O₃ [63].

3.4. Proposed Degradation Pathways of DMP and 1−NP

The intermediates of DMP and 1−NP during Fe/Mn@γ−Al₂O₃ catalytic ozonation were analyzed using LC−MS, and their identified intermediates are detailed in Figures S8 and S9. As previously noted, the degradation of DMP was dominated by HO°−induced reaction, and that of 1−NP resulted from O₃, HO°, and ¹O₂ oxidation. Their degradation pathways are proposed and illustrated in Figure 4.

The degradation of DMP during Fe/Mn@γ−Al₂O₃ catalytic ozonation (Figure 4A) was initiated by HO° to attack the ester group to form P₁₈₀ and further oxidized to generate 1,2−benzenedicarboxylicacid [64], or to undergo addition on the benzene ring to yield HO−adducts (P₂₁₀, P₂₂₆, P₂₄₂) [65]. The formed 1,2−benzenedicarboxylicacid reacted with HO° to yield HO−adducts (P₁₈₂) or transform to phthalic anhydride via dehydration. The adducts P₂₁₀, P₂₂₆, and P₂₄₂ were oxidized by HO° to form P₁₉₆, P₁₈₂, and P₁₉₈. 1,2−benzenedicarboxylicacid and its HO−adducts (P₁₈₂ and P₁₉₈) underwent a ring−opening reaction to generate maleic acid and phenols. Phthalic anhydride and maleic acid were decomposed to yield low molecular weight organic acids (e.g., acetic acid, oxalic acid) via ring−opening reactions, which were finally mineralized to CO₂ and H₂O [66].

Figure 4B shows the degradation pathway of 1−NP under the Fe/Mn@γ−Al₂O₃/O₃ system. Ozone quickly reacted with the naphthalene group as electron−rich moieties to produce hydroxy−1,4−naphthoquinone (P₁₇₄) that was oxidized to yield 1,4−naphthoquinone (P₁₅₈) [67]. The initial attack of HO° on 1−naphthol generated transient naphthyloxy radical (1−NP°(−H)) via H−abstraction and transformed to 1,4−naphthoquinone (P₁₅₈) [68]. ¹O₂ readily added to 1−NP to form P_176, and P₁₇₆ rearranged and dehydrated to yield 1,4−naphthoquinone (P₁₅₈) [69]. The generated 1,4−naphthoquinone was further oxidized to 1,2−benzenedicarboxylicacid and phthalic anhydride by ozone and HO°, which were finally mineralized to CO₂ and H₂O [67].

3.5. Effects of Water Matrix

The performance of the synthesized catalyst in the complex water matrix was evaluated to investigate its potential application, especially for dyeing wastewater effluent. Thus, the effects of ions, pH, and actual water samples on the degradation efficacy of DMP and 1−NP were investigated (Figure 5).

Carbonate, sulfate, chloride, and nitrate were widely detected in dyeing wastewater. Figure 5A,D show their effects on the performance of catalytic ozonation in the ultrapure water system. As seen, sulfate, chloride, and nitrate cannot distinctly inhibit the degradation of DMP and 1−NP (< 5%) because sulfate and nitrate barely consumed the reactive species (<10 M⁻¹ s⁻¹) involved in the Fe/Mn@γ−Al₂O₃/O₃ system. Chloride reacts with HO° but cannot accelerate forward reaction at a low concentration (2 mM) (k_+/− = 3.0 × 10⁹ M⁻¹ s⁻¹/6.10 × 10⁹ M⁻¹ s⁻¹) [70,71]. Being different from chloride, carbonate as HO° scavenger (k = 1.0 × 10⁷ M⁻¹ s⁻¹) intensively scavenged [HO°], thus impeding their removal (> 50%). The formed CO₃°⁻ (E⁰ = 1.63 V) was much less reactive than HO° (E⁰ = 2.7 V) [37,72].

The pH value affects the concentration and speciation of reactive species during catalytic ozonation and thus leads to the variation in pollutant removal efficiency. The removal rate increased from 1.83 × 10⁻³ s⁻¹ to 2.46 × 10⁻³ s⁻¹ for DMP and from 1.90 × 10⁻² to 2.72 × 10⁻² s⁻¹ for 1−NP within pH 5−8 (Figure 5B,E). Their removal efficiencies exceeded 90% within 30 min, indicating that the synthesized Fe/Mn@γ−Al₂O₃ catalyst had a good ozone−catalyzing performance at a wide pH variation. Specifically, the alkali condition was more suitable for the contaminant degradation in the Fe/Mn@γ−Al₂O₃/O₃ system as OH⁻ is mainly hydroxyl radical initiator in ozonation [73], corresponding with previous studies [74,75].

To further investigate the effect of the water matrix on the removal of 1−NP and DMP with the Fe/Mn@γ−Al₂O₃/O₃ system, DMP/1−NP (50 μM) was spiked in printing and dyeing industrial park (PDIP) effluent, wastewater treatment plant (WWTP) effluent, and river water (RW). The water quality parameters of three water samples are listed in Table S5. The removal rate constants of 1−NP and DMP in all actual water samples decreased by 50–60% in contrast to ultrapure water (Figure 5C,F). This decrease resulted from the competition of reactive species by dissolved organic matter (DOM) in the actual water samples. In contrast, the Fe/Mn@γ−Al₂O₃/O₃ system accelerated the DOC and UV₂₅₄ decrease in the actual water samples compared with ozonation alone (Figure S11). Overall, the synthesized Fe/Mn@γ−Al₂O₃ can be potentially applied in the catalytic ozonation of actual water.

4. Conclusions

A novel Fe/Mn@γ−Al₂O₃ nanocatalyst was successfully synthesized and utilized for catalytic ozonation of DMP and 1−NP. Their removal rate constants could elevate 3−5 times with Fe/Mn@γ−Al₂O₃/O₃ system compared to ozonation alone due to more yield of HO° and ¹O₂ involved in their degradation. The characterization of Fe/Mn@γ−Al₂O₃ suggested that the formation of HO° and ¹O₂ possibly resulted from the transformation of O₃ from V_O to O_ads with the circulation of Mn(III)/(IV) and Fe(II)/(III). The degradation pathways of DMP and 1−NP were proposed based on the mechanisms and the identified intermediates. Additionally, the Fe/Mn@γ−Al₂O₃/O₃ system could be applied in removing pollutants at pH 6−9, in the presence of different ions, and dealing with actual dye wastewater effluent samples. The performance and practicability of the Fe/Mn@γ−Al₂O₃/O₃ system on the treatment of dye wastewater effluent should be further evaluated in a pilot−/full−scale experiment.