Iron species activating chlorite: Neglected selective oxidation for water treatment

Abstract

Chlorite (ClO₂⁻) is the by-product of the water treatment process carried out using chlorine dioxide (ClO₂) as an effective disinfectant and oxidant; however, the reactivation of ClO₂⁻ has commonly been overlooked. Herein, it was unprecedentedly found that ClO₂⁻ could be activated by iron species (Fe_b: Fe⁰, Fe^II, or Fe^III), which contributed to the synchronous removal of ClO₂⁻ and selective oxidative treatment of organic contaminants. However, the above-mentioned activation process presented intensive H⁺-dependent reactivity. The introduction of Fe_b significantly shortened the autocatalysis process via the accumulation of Cl⁻ or ClO⁻ during the protonation of ClO₂⁻ driven by ultrasonic field. Furthermore, it was found that the interdependent high-valent-Fe-oxo and ClO₂, after identification, were the dominant active species for accelerating the oxidation process. Accordingly, the unified mechanisms based on coordination catalysis ([Fe^N(H₂O)_a(ClO_x^m−)_b]ⁿ⁺-P) were putative, and this process was thus used to account for the pollutant removal by the Fe_b-activated protonated ClO₂⁻. This study pioneers the activation of ClO₂⁻ for water treatment and provides a novel strategy for “waste treating waste”. Derivatively, this activation process further provides the preparation methods for sulfones and ClO₂, including the oriented oxidation of sulfoxides to sulfones and the production of ClO₂ for on-site use.

Graphical abstract

Highlights

• •The catalytic effect appeared under the activation of protonated ClO₂⁻ via Fe^0/II/III.
• •The active intermediates mainly contained high-valent-Fe-oxo and ClO₂.
• •The reaction mechanism revolved around coordination catalysis.
• •The Fe_b activating H⁺/ClO₂⁻ provided a strategy of “waste treating waste”.
• •The preparation methods of ClO₂ and sulfones were prepared.

1. Introduction

Chlorite (ClO₂⁻), as a redox agent with wide application prospects, can be used not only as an alkaline Cl oxidant with a high oxidation capacity [1] but also as a source of chlorine dioxide (ClO₂) via electron loss [2] (e.g., ultraviolet (UV) activation [3]). As a selective oxidant, the produced ClO₂ could be considered a stable radical species which is significantly reactive toward phenolic, aniline, olefin, and amine moieties through electron-transfer reactions [4]. Moreover, ClO₂ was used as a strong disinfectant in water treatment plants, which acted as an alternative to chlorine [5,6]. It was also found that ClO₂ was the most efficient disinfectant in killing bacteria and, in particular, successful in deactivating viruses over a wide pH range [7]. However, ClO₂⁻ has been seen as the by-product of ClO₂ treatment, and the Environmental Protection Agency has recently labeled ClO₂⁻ as a major water contaminant. ClO₂⁻ could result in suspected health risks such as childhood anemia [2]. According to “Standards for drinking water quality, GB5749-2022” in China, the limited concentration of ClO₂⁻ after ClO₂ disinfection was reported to be 0.7 mg L⁻¹ [8]. Therefore, a method to remediate ClO₂⁻ is of great practical interest from an environmental standpoint [2].

Currently, the catalysts, including various chlorite-dismutases (e.g., Fe^III-Clds) [[9], [10], [11], [12]] or non-heme metalloporphyrins/metal complexes (e.g., Mn^III-Pors) [[13], [14], [15], [16], [17], [18]], are also utilized to catalyze ClO₂⁻ detoxification, which results in the formation of harmless chloride (Cl⁻) and dioxygen (O₂) [[9], [10], [11], [12], [13], [14], [15]] or ClO₂ [[16], [17], [18]]. Moreover, the corresponding trivalent metal (M^III) gets transformed into high-valent metal (HVM), O=M^IV, or O=M^IV•+ [[9], [10], [11], [12], [13], [14], [15], [16], [17], [18]].

Recently, in advanced oxidation processes (AOPs), the formation and role of aqueous Fe^IV-oxo complex (Fe^IVO²⁺) have also attracted significant research attention in ferrous ion (Fe²⁺)-initiated AOPs by employing the oxidants including ozone, hydrogen peroxide, persulfates, peracetic acid (PAA), etc. [19]. Compared to the long-recognized free radicals, Fe^IVO²⁺ has been shown to remove various pollutants (e.g., phosphite, amoxicillin, and sulfamethoxazole) [19,20]. Moreover, Fe^IVO²⁺ could react with organic compounds through various pathways (hydrogen-atom, hydride, oxygen-atom, and electron transfer, as well as electrophilic addition) at moderate reaction rates [19]. Fe^IVO²⁺ could also present selective reactivity toward inorganic ions prevailing in natural water [19]. This hints at the potential application prospects of high-valent iron in environmental treatment due to the production of the above-mentioned HVM in the catalysis of ClO₂⁻.

$$4{\mathrm{F}\mathrm{e}}^{2+}+{\mathrm{C}\mathrm{l}\mathrm{O}}_2^{-}+10{\mathrm{H}}_2\mathrm{O}\stackrel{{\mathrm{O}\mathrm{H}}^{-}}{\to }4{\mathrm{F}\mathrm{e}\left(\mathrm{O}\mathrm{H}\right)}_3\left(\mathrm{s}\right)+{\mathrm{C}\mathrm{l}}^{-}+8{\mathrm{H}}^{+}$$ (1)

In terms of the reactions between ionic iron and ClO₂⁻, Fe^II was commonly adopted to reduce ClO₂⁻ to form the by-product Cl⁻ (95%) [21], in particular, under the alkaline condition for efficient ClO₂⁻ removal (as shown in equation (1)) [21,22]. However, a few studies indicated that the different phenomenon with a certain catalytic effect between ionic iron and ClO₂⁻ was observed under acidic conditions, which could be analogous to the conventional Fenton reaction. For instance, Fe^III could catalyze the decomposition of ClO₂⁻ to generate ClO₂ [23]. Fabian and Gordon considered the formation of FeClO₂²⁺-intermediates during the catalytic process (equation (2)) [[24], [25], [26], [27]]. Nevertheless, compared with the above-mentioned catalytic activity of metalloporphyrin, it seemed that the catalytic mechanism of producing ClO₂ further demands a lot more systematic explorations and improvement on the conversion of intermediate active substances, such as whether it could relate to the role of Fe^IVO²⁺.

$${\mathrm{F}\mathrm{e}}^{3+}+{\mathrm{C}\mathrm{l}\mathrm{O}}_2^{-}\leftrightarrow {\mathrm{F}\mathrm{e}\mathrm{C}\mathrm{I}\mathrm{O}}_2^{2+}\leftrightarrow {\mathrm{F}\mathrm{e}}^{2+}+\mathrm{C}\mathrm{l}{\mathrm{O}}_2$$ (2)

Noteworthy, the previous studies were limited to the detoxification of ClO₂⁻ via its decomposition. However, the potential water treatment was neglected by producing Fe^IVO²⁺ or ClO₂ with a significant oxidation effect under the detoxification process. Therefore, based on the treatment of organic pollutants with by-product ClO₂⁻ via activation by iron species (Fe_b), the strategy for simultaneous removal of ClO₂⁻ and selective removal of organic pollutants was provided. However, to the best of our knowledge, this topic has rarely been investigated to date.

Therefore, the objectives of this study include reporting the novel methods and corresponding mechanisms of applying Fe_b (e.g., Fe^{0, II, or III})-activated ClO₂⁻ under protonation by ultrasonic in the water treatment process. The ultrasonic was adopted as the field driving the diffusion. On the one hand, the study focused on the process of activating ClO₂⁻ with Fe⁰ for the synchronous removal of ClO₂⁻ and triphenylmethane (tpm) derivatives. On the other hand, the activation mechanisms of Fe⁰-activated ClO₂⁻ including the contributions of the dominant active species for pollutant removal were further modified and elucidated. Briefly, this study may provide a novel insight into the coordination mechanism in ClO₂⁻ activated by Fe_b, and further offer an insightful strategy for “waste treating waste”.

2. Experimental

2.1. Materials

Fe-based materials including Fe⁰, Fe^II, and Fe^III species were selected, i.e., zero-valent-iron foam (Fe⁰_f), iron(II) sulfate heptahydrate (FeSO₄⋅7H₂O), and iron(III) chloride hexahydrate FeCl₃⋅6H₂O, respectively. Sodium chlorite (NaClO₂ (CI), 80%) and crystal violet (tpm_CV), a type of tpm derivative, were obtained from Macklin. Other chemicals, including commercially available analytical grade methyl phenyl sulfoxide (PMSO), dimethyl sulfoxide (DMSO), anions such as Na-salts (Cl⁻, SO₄²⁻, and CO₃²⁻), sodium azide (NaN₃), p-benzoquinone (PBQ), methyl alcohol (MA), ethylene-diamine-tetraacetic acid (EDTA), and oxalic acid (H₂C₂O₄), were used without further purification. All reactions were carried out in deionized water obtained using a Heal Force water purification system (SMAR-RO).

2.2. Instrumentation

The liquid phase reactions were driven by ultrasound (US) in a self-prepared bath-ultrasonic generator (28 kHz, 30 W L⁻¹) coupled with a cylindrical acrylic reactor, as reported in our previous study [28]. The tpm_CV solution (12 mg L⁻¹) was prepared and placed in the above-mentioned reactor. Briefly, certain amounts of H₂SO₄, NaClO₂, and Fe⁰_f were added successively, and then the ultrasonic reactor was turned on to conduct the batch reaction. At the fixed reaction time, the tpm_CV concentration was determined at λ_max = 584 nm by UV–visible (vis) spectroscopy. The initial pH of solutions was adjusted by NaOH (1 M) and H₂SO₄ (1 M), and no attempt was made to maintain a constant pH during the process. UV–vis spectra were recorded using a scanning spectrophotometer (Pgeneral TU-1810S). The converted products of PMSO were detected by gas chromatograph-mass spectrometry (GC/MS, Trace 1300 ISQ-QD, Thermo Fisher). Before the GC/MS analysis, a subsample (15 mL) at a certain treating time (5, 15, and 25 min) was extracted with dichloromethane (10 mL), respectively. Next, the extracted solution was dehydrated with anhydrous sodium sulfate and concentrated to 2 mL. Before and after treatment, the micrographs of the Fe⁰_f were examined by scanning electron microscopy (SEM, Zeiss, Germany), respectively. Before the SEM test, the Fe⁰_f samples taken out from the reaction solution immediately required dehydration by adding absolute ethyl alcohol, then the samples were dried with filter papers and finally vacuum-dried.

3. Results and discussion

3.1. Fe_b activation for chlorite

3.1.1. Coupling effect

As shown in Fig. 1a, activators including Fe⁰, Fe^II, and Fe^III were utilized to catalyze the decomposition of ClO₂⁻ for removing tpm_CV at acidic pH under the ultrasonic field. The tpm_CV, as the emerging contaminant, has been widely used in medical, veterinary, and aquaculture applications. However, the “Lancet Oncology” reported that tpm_CV had been restricted in veterinary or cosmetic applications [29]. The system exhibited a coupling effect; namely, removal effect in the following: Fe^III > Fe^II > Fe⁰ > no activator. The apparent second-order kinetics constants of the H⁺-US/CI/Fe^II and H⁺-US/CI/Fe^III systems were 0.043 μM⁻¹ min⁻¹ (R_Adj² = 0.996) and 0.115 μM⁻¹ min⁻¹ (R_Adj² = 0.995), respectively. When the H⁺-US/CI/Fe⁰ system was utilized, Fe⁰, in the presence of H⁺, could be converted to Fe^II, and Fe^II could be further oxidized into Fe^III via equations (3), (4) [27,30]. Next, Fe^III could catalyze ClO₂⁻ to produce reactive species for decomposing ClO₂⁻ and tpm_CV. For the H⁺-US/CI system, ClO₂⁻ could react with H⁺ to generate ClO₂ via equation (5) [31], and then protonated ClO₂⁻ formed HClO₂, which was subjected to sonolysis to produce ClO₂. In terms of the H⁺-US/CI and H⁺-US/CI/Fe⁰ systems, both the removal processes exhibited an early stage of “build-up”. Based on the dynamics fitting, the piecewise kinetics of the above-mentioned systems (Fig. 1b) proved this point.

$$5{\mathrm{F}\mathrm{e}}^{2+}+\mathrm{C}\mathrm{l}{\mathrm{O}}_2+4{\mathrm{H}}^{+}\to {\mathrm{C}\mathrm{l}}^{-}+5{\mathrm{F}\mathrm{e}}^{3+}+{2\mathrm{H}}_2\mathrm{O}$$ (3)

$$4{\mathrm{F}\mathrm{e}}^{2+}+\mathrm{H}\mathrm{C}\mathrm{l}{\mathrm{O}}_2+3{\mathrm{H}}^{+}\to {\mathrm{C}\mathrm{l}}^{-}+4{\mathrm{F}\mathrm{e}}^{3+}+{2\mathrm{H}}_2\mathrm{O}$$ (4)

$$5{\mathrm{C}\mathrm{l}\mathrm{O}}_2^{-}+4{\mathrm{H}}^{+}\to {\mathrm{C}\mathrm{l}}^{-}+4\mathrm{C}\mathrm{l}{\mathrm{O}}_2+2{\mathrm{H}}_2\mathrm{O}$$ (5)

Fig. 1.

Fe-based activation for chlorite (ClO₂⁻) by virtue of protonation (H⁺): a, system comparison; b–c, piecewise kinetics; tpm_CV (12 mg L⁻¹), US (30 W L⁻¹, 28 kHz), ClO₂⁻ dosage (3 mM), 1 M H₂SO₄ (0.2 mL L⁻¹), vortex (0 rad min⁻¹), and initial T (293 K).

3.1.2. Piecewise kinetics: transition state

Owing to the removal reaction presenting an “S”-process, namely, existing transition state, traditional one-step dynamics with integer order failed to meet the requirement of exploring such piecewise oxidation stages. Fortunately, equation (6) is universal, that is

$$\frac{d\left[{\mathrm{t}\mathrm{p}\mathrm{m}}_{\mathrm{C}\mathrm{V}}\right]}{dt}=f^{\prime }\left(c_{{\mathrm{t}\mathrm{p}\mathrm{m}}_{\mathrm{C}\mathrm{V}}}\right)$$ (6)

Multiple inflection points (IPs) directly obtained from the process of “dc/dt vs. t” further indicated the existence of the piecewise kinetics processes for both H⁺-US/CI and H⁺-US/CI/Fe⁰ systems (Fig. 1c). After logarithmic transformation, the apparent-kinetics-orders (non-integer) of the H⁺-US/CI and H⁺-US/CI/Fe⁰ systems were 0.90 and 1.56, respectively.

Moreover, in view of excess active oxidizing species (3 mM ClO₂⁻ >> 30 μM tpm_CV), equation (7) could be obtained.

$$f^{\text{'}}\left(c_{\mathrm{t}\mathrm{p}{\mathrm{m}}_{\mathrm{C}\mathrm{V}}}\right)=-k{c_{\mathrm{t}\mathrm{p}{\mathrm{m}}_{\mathrm{C}\mathrm{V}}}}^n$$ (7)

Then, by feasible approximation, the piecewise kinetics processes were more obvious when 1 and 1.5 orders were utilized to fit kinetics curves of the H⁺-US/CI and H⁺-US/CI/Fe⁰ systems, respectively (Fig. 1b).

According to the H⁺-US/CI system, two IPs at 12.5 and 22.5 min indicated the existence of three piecewise removal processes. Moreover, processes occurring in the time range of 0–12.5 min and 12.5–22.5 min with slow reaction rates indicated the potential autocatalytic feasibility, and the reaction at 22.5–37.5 min showed pseudo-1-order kinetics constant with a value of 0.113 min⁻¹. After introducing Fe⁰, the time of the stage of “build-up”, namely, IP was shortened to 7 min, and 7–25 min exhibited pseudo-1.5-order kinetics constant with a value of 0.056 μM^−0.5 min⁻¹, which indicated that non-single oxidizing-species contributed to the pollutant removal. As a result, it is speculated that multiple oxidizing species may compete for the oxidation process.

3.1.3. Surface action by US

As shown in Fig. 2, SEM images showed more features with a relatively large size (20 μm), such as the honeycomb frame and the oxidation regions. With the prolongation of reaction time, the erosion and oxidation on the surface of Fe⁰-foam became more significant. Even so, most regions on surface of Fe⁰-foam still consisted of exposed fresh Fe⁰ sources due to the “polishing” role from ultrasonic via surface heat transfer, mechanical friction, and cavitation impact [28]. Owing to the synergistic effect between H⁺-assisted corrosion and ultrasonic-assisted polishing, the catalytic reaction of ClO₂⁻ could be sustained.

Fig. 2.

Surface morphology of Fe⁰-foam under H⁺-US/CI at different treating times: a, 0 min; b, 5 min; c, 15 min; d, 25 min.

3.1.4. Modeling

Complex reactions tended to multistage processes during the heterogeneous system (i.e., the abovementioned system) at a certain treating time. It may be unsuitable to describe reaction processes under overall evaluation only with a single quasi-dynamic equation. Here, a novel indicator was used to evaluate the removal effect in a heterogeneous system. In particular, at a certain time, the degraded momentum, namely, the concentration movement of removing contaminant μ (M min M⁻¹), could be defined to determine the overall trends of the removal [32].

$$\mu ={\int }_{t_0}^{t_{\mathrm{e}}}\frac{c_0-c_t}{c_0}t$$ (8)

where c_t/c₀ was the concentration gradient of contaminant (c_t/c₀ ∈ [0,1]), t₀ and t_e represented the start and end times of the reaction. Theoretically, the value μ_max equals the value t_e, namely, c_t/c₀ reached 0 at t₀.

$$r_{\mathrm{a}\mathrm{p}\mathrm{p}}=\frac{\mu }{{\mu }_{\max }}$$ (9)

where r_app was the apparent ratio of concentration movement.

$$\phi =\frac{{r_{\mathrm{a}\mathrm{p}\mathrm{p}}}_i}{{r_{\mathrm{a}\mathrm{p}\mathrm{p}}}_{\mathrm{con}}}$$ (10)

where φ was the relative factor, i represented one influencing factor, and con represented the control group. Moreover, the t_e was selected at 17.5 or 22.5 min in the follow-up evaluation.

3.2. Influencing factors

3.2.1. pH

As shown in Fig. 3a, the removal reaction could be improved by introducing H⁺ into the H⁺-US/CI/Fe⁰ system (φ increased from 0.856 to 1.331). H⁺ could cause the conversion of Fe⁰ to ionic Fe for activation. However, once the system presented a neutral or basic state, it became fatal to Fe⁰-activation. The abovementioned system was close to only US/CI or OH⁻-US/CI, leading to a very limited oxidation effect. However, it seemed that the removal effect under alkaline conditions (φ = 0.119) was slightly better than that under neutral conditions (φ = 0.090). The sonolysis of ClO₂⁻ could generate reactive radicals (e.g., •ClO_xⁿ⁻), which was supported by our previous study [28]. Photolysis of ClO₂⁻ was ever investigated to obtain that OH⁻ was found to be conducive to the transformation of free radicals [1], which may be analogous to the above-mentioned result. Moreover, OH⁻ may prevent the organic substances from evaporating into the cavitation bubble for pyrolysis, leading to the occurrence of the free radical oxidation only at the gas–liquid interface of the cavitation bubble.

Fig. 3.

Influencing factors for H⁺-US/CI/Fe⁰ system: a, pH; b, ClO₂⁻ dosage; c, anion. Conditions ibid., Fig. 1.

3.2.2. ClO₂⁻ dosage

The solution produces OH⁻ due to the hydrolysis of ClO₂⁻ [33]. More ClO₂⁻ dosage leads to the generation of more OH⁻, which can increase the consumption of H⁺. As a result, the amount of H⁺ actually acting on Fe⁰ and ClO₂⁻ decreases. However, more reactant ClO₂⁻ could bring more effective collisions and improved utilization efficiency of other reactants (e.g., H⁺). The removal reaction in system H⁺-US/CI/Fe⁰ maintained a balance (Fig. 3b) (φ around 0.959–1.008).

3.2.3. Anions

The anions, including SO₄²⁻ (φ = 0.981) and Cl⁻ (φ = 1.002) (3 mM), would not interfere with the solution-phase oxidation; however, CO₃²⁻ (1 mM) exhibited significant influence on the removal of tpm_CV (Fig. 3c) in the H⁺-US/CI/Fe⁰ system (φ = 0.140). The main reason for the phenomenon was that the formation of OH⁻ via hydrolysis of CO₃²⁻ (equation (11)) resulted in competitive consumption of H⁺ [34]. Furthermore, OH⁻ led to the formation of the colloidal or precipitated Fe(OH)_n, which adversely affected the activation of ClO₂⁻ with Feⁿ⁺. Therefore, low-dosage CO₃²⁻ (0.1 mM) showed a slight influence on removing tpm_CV (φ = 0.837), which indicated that the reaction was not interfered with by the ions themselves except the hydrolysis.

$${\mathrm{H}}_2\mathrm{C}{\mathrm{O}}_3\stackrel{\mathrm{pKa}=6.34}{\iff }{\mathrm{H}}^{+}+\mathrm{H}\mathrm{C}{\mathrm{O}}_3^{-}\stackrel{\mathrm{pKa}=10.37}{\iff }2{\mathrm{H}}^{+}+\mathrm{C}{\mathrm{O}}_3^{2-}$$ (11)

3.3. Inhibition

The reaction rate of MA vs. •OH was 9.7 × 10⁸ M⁻¹ s⁻¹; therefore, MA could be used as a scavenger for •OH [35]. However, when MA (1 M) was added to the solution, the removal trend seemed to remain constant (φ = 0.982) (Fig. 4a). Therefore, the •OH (at least in the liquid phase) showed little contribution to the removal of tpm_CV, illustrating that ultrasonic provided other roles rather than •OH role in the catalytic system.

Fig. 4.

Possible inhibition (a), competition and coordination (b) from two reactants for H⁺-US/CI/Fe⁰ system. Conditions ibid., Fig. 1.

Notably, DMSO has recently been used to identify oxygen atom transfer reactions, such as reaction with Fe^IVO²⁺ to form DMSO₂ (equation (12)) [36]. However, DMSO could also react with •OH to generate methyl radical (•CH₃) (equation (13)) [37].

$${\mathrm{F}\mathrm{e}}^{\mathrm{I}\mathrm{V}}{\mathrm{O}}^{2+}+\mathrm{D}\mathrm{M}\mathrm{S}\mathrm{O}\to {\mathrm{D}\mathrm{M}\mathrm{S}\mathrm{O}}_2+{\mathrm{F}\mathrm{e}}^{2+}$$ (12)

$$\bullet \mathrm{O}\mathrm{H}+\mathrm{D}\mathrm{M}\mathrm{S}\mathrm{O}\to \bullet {\mathrm{C}\mathrm{H}}_3+{\mathrm{C}\mathrm{H}}_3{\mathrm{S}\mathrm{O}}_2\mathrm{H}$$ (13)

The screening test of MA ruled out the role of •OH; thus, the decrease in the removal of tpm_CV after adding DMSO should be attributed to the active species causing an oxygen atom ([O]) transfer step. Moreover, it was demonstrated that DMSO could be an excellent masking agent for aqueous chlorine (e.g., Cl₂ or ClO⁻ (HClO)), without affecting ClO₂, ClO₂⁻, and ClO₃⁻ by the co-existence of excess DMSO [30,38,39].

PBQ could be used to scavenge superoxide radical (O₂^•−) with a reaction rate at 8 × 10⁹ M⁻¹ s⁻¹ [40], but the result indicated the absence of O₂^•− and the addition of PBQ appeared to promote the removal of tpm_CV (φ = 1.159) (Fig. 4a). The PBQ acted as an important redox reagent. Its applications range from organic synthesis to mediation of electron transfer in biochemical systems and solar cells [41]. On the one hand, the PBQ-based derivatives with electron-withdrawing substituents were employed as electron acceptors, and the quinones could serve as catalysts to shuttle electrons; thus, both facilitated the degradation of the starting aromatic compound [42]. On the other hand, the mixing of solutions of halide anions (I⁻, Br⁻, or Cl⁻) with PBQ acceptors resulted in the instantaneous formation of anion−π complexes [41]. This pointed out the charge-transfer character of these associates (equation (14)) [41].

$$\mathrm{B}\mathrm{Q}+{\mathrm{X}}^{-}\leftrightarrow {\mathrm{B}\mathrm{Q}}^{\bullet -}+{\mathrm{X}}^{\bullet }$$ (14)

3.4. Competition and coordination

Oxalic acid and EDTA, as the chelating agents for metal ions, bonded with Feⁿ⁺, which impeded Feⁿ⁺-induced activation of ClO₂⁻ (Fig. 4). Oxalic acid could react with Feⁿ⁺ to form some complexes, i.e., [Fe^III(C₂O₄)_n]³⁻²ⁿ and [Fe^II(C₂O₄)_n]²⁻²ⁿ [43], leading to the competing reaction (φ = 0.756). In particular, EDTA could combine with [Fe^II/III(H₂O)₆]^2+/3+ to form the coordination complex. For instance, [Fe^II(EDTA)H₂O]²⁻ was the seven-coordinated species with a monocapped trigonal-prismatic geometry [44]. On acidification, EDTA chelate could be protonated to form a monoprotonated species [Fe^II(EDTAH)H₂O]⁻ with pentagonal–bipyramidal geometry [44]. The autoxidation of [Fe^II(EDTA)]: the [Fe^II(EDTA)H₂O]²⁻ (20-electron) could start the reaction with O₂ to form H₂O₂, and the reaction was controlled by electron transfer and the associated reorganization barrier. The reactions were presented as equations (15), (16), (17), (18), (19), (20) [44]. Thus, with the increase in the EDTA dosage, the autoxidation effect surpassed the metal chelation effect, resulting in promoting the removal (φ increased from 0.291 to 0.666). Furthermore, more EDTA resulted in the decrease of the pH of the solution, which was also beneficial for the removal (Fig. 4b).

$$\mathrm{F}{\mathrm{e}}^{2+}+\mathrm{E}\mathrm{D}\mathrm{T}\mathrm{A}\to \left[\mathrm{F}{\mathrm{e}}^{\mathrm{I}\mathrm{I}}\left(\mathrm{E}\mathrm{D}\mathrm{T}\mathrm{A}\right)\right]\stackrel{{\mathrm{H}}_2\mathrm{O}}{\to }{\left[{\mathrm{Fe}}^{\mathrm{II}}\left(\mathrm{EDTA}\right){\mathrm{H}}_2\mathrm{O}\right]}^{2-}$$ (15)

$${\left[{\mathrm{F}\mathrm{e}}^{\mathrm{I}\mathrm{I}}\left(\mathrm{E}\mathrm{D}\mathrm{T}\mathrm{A}\right){\mathrm{H}}_2\mathrm{O}\right]}^{2-}+{\mathrm{H}}^{+}\leftrightarrow {\left[{\mathrm{F}\mathrm{e}}^{\mathrm{I}\mathrm{I}}\left(\mathrm{E}\mathrm{D}\mathrm{T}\mathrm{A}\mathrm{H}\right){\mathrm{H}}_2\mathrm{O}\right]}^{-}$$ (16)

$${\left[{\mathrm{F}\mathrm{e}}^{\mathrm{I}\mathrm{I}}\left(\mathrm{E}\mathrm{D}\mathrm{T}\mathrm{A}\mathrm{H}\right){\mathrm{H}}_2\mathrm{O}\right]}^{-}+{\mathrm{O}}_2\leftrightarrow {\left[{\mathrm{F}\mathrm{e}}^{\mathrm{I}\mathrm{I}}\left(\mathrm{E}\mathrm{D}\mathrm{T}\mathrm{A}\mathrm{H}\right){\mathrm{O}}_2\right]}^{-}+{\mathrm{H}}_2\mathrm{O}$$ (17)

$${\left[{\mathrm{F}\mathrm{e}}^{\mathrm{I}\mathrm{I}}\left(\mathrm{E}\mathrm{D}\mathrm{T}\mathrm{A}\mathrm{H}\right){\mathrm{O}}_2\right]}^{-}\to {\left[{\mathrm{F}\mathrm{e}}^{\mathrm{I}\mathrm{I}\mathrm{I}}\left(\mathrm{E}\mathrm{D}\mathrm{T}\mathrm{A}\mathrm{H}\right){\mathrm{O}}_2^{-}\right]}^{-}$$ (18)

$${\left[{\mathrm{F}\mathrm{e}}^{\mathrm{I}\mathrm{I}}\left(\mathrm{E}\mathrm{D}\mathrm{T}\mathrm{A}\mathrm{H}\right){\mathrm{H}}_2\mathrm{O}\right]}^{-}+{\left[{\mathrm{F}\mathrm{e}}^{\mathrm{I}\mathrm{I}\mathrm{I}}\left(\mathrm{E}\mathrm{D}\mathrm{T}\mathrm{A}\mathrm{H}\right){\mathrm{O}}_2^{-}\right]}^{-}\to {\left[\left(\mathrm{E}\mathrm{D}\mathrm{T}\mathrm{A}\mathrm{H}\right){\mathrm{F}\mathrm{e}}^{\mathrm{I}\mathrm{I}\mathrm{I}}\left({\mathrm{O}}_2^{2-}\right){\mathrm{F}\mathrm{e}}^{\mathrm{I}\mathrm{I}\mathrm{I}}\left(\mathrm{E}\mathrm{D}\mathrm{T}\mathrm{A}\mathrm{H}\right)\right]}^{2-}+{\mathrm{H}}_2\mathrm{O}$$ (19)

$${\left[\left(\mathrm{E}\mathrm{D}\mathrm{T}\mathrm{A}\mathrm{H}\right){\mathrm{F}\mathrm{e}}^{\mathrm{I}\mathrm{I}\mathrm{I}}\left({\mathrm{O}}_2^{2-}\right)\mathrm{F}{\mathrm{e}}^{\mathrm{I}\mathrm{I}\mathrm{I}}\left(\mathrm{E}\mathrm{D}\mathrm{T}\mathrm{A}\mathrm{H}\right)\right]}^{2-}\underset{{\mathrm{H}}_2\mathrm{O}}{\overset{{\mathrm{H}}^{+}}{\to }}2\left[\mathrm{F}{\mathrm{e}}^{\mathrm{I}\mathrm{I}\mathrm{I}}\left(\mathrm{E}\mathrm{D}\mathrm{T}\mathrm{A}\mathrm{H}\right){\mathrm{H}}_2\mathrm{O}\right]+{\mathrm{H}}_2{\mathrm{O}}_2$$ (20)

When NaN₃ was used (Fig. 4b), the study reported that without spin trap agent N-tert-butyl-α-phenylnitrone, the reaction of ClO₂ with azide (N₃⁻) was strongly inhibited by ClO₂⁻ due to rapid back-electron transfer. This led to additional mechanistic pathways, complex kinetics, and the formation of an unusual set of products, including N₂, N₂O, NO₃⁻, Cl⁻, and ClO₂⁻ as represented in equations (21), (22), (23), (24), (25), (26) [45]. Furthermore, in the presence of an iron(III)-azide system, the complex formation reaction occurred via the conjugate acid/conjugate base pathway (equation (27)) [46].

$${\mathrm{C}\mathrm{l}\mathrm{O}}_2+{\mathrm{N}}_3^{-}\leftrightarrow {\mathrm{N}}_3+{\mathrm{C}\mathrm{l}\mathrm{O}}_2^{-}$$ (21)

$${\mathrm{N}}_3+\mathrm{P}\mathrm{B}\mathrm{N}\to \mathrm{P}\mathrm{B}\mathrm{N}{\mathrm{N}}_3$$ (22)

$${10\mathrm{C}\mathrm{l}\mathrm{O}}_2+{10\mathrm{N}}_3^{-}+{4\mathrm{H}}_2\mathrm{O}\to 10{\mathrm{C}\mathrm{l}}^{-}+{8\mathrm{N}\mathrm{O}}_3^{-}+11{\mathrm{N}}_2+{8\mathrm{H}}^{+}$$ (23)

$${8\mathrm{C}\mathrm{l}\mathrm{O}}_2+{8\mathrm{N}}_3^{-}+{\mathrm{H}}_2\mathrm{O}\to 8{\mathrm{C}\mathrm{l}}^{-}+{2\mathrm{N}\mathrm{O}}_3^{-}+{11\mathrm{N}}_2\mathrm{O}+{2\mathrm{H}}^{+}$$ (24)

$${2\mathrm{C}\mathrm{l}\mathrm{O}}_2+{2\mathrm{N}}_3^{-}\to 2{\mathrm{C}\mathrm{l}\mathrm{O}}_2^{-}+3{\mathrm{N}}_2$$ (25)

$${\mathrm{C}\mathrm{l}\mathrm{O}}_2+{\mathrm{N}}_3^{-}+{0.278\mathrm{H}}_2\mathrm{O}\to 0.833{\mathrm{C}\mathrm{l}}^{-}+0.167{\mathrm{C}\mathrm{l}\mathrm{O}}_2^{-}+0.555{\mathrm{N}\mathrm{O}}_3^{-}+{0.278\mathrm{N}}_2\mathrm{O}+0.943{\mathrm{N}}_2+{0.555\mathrm{H}}^{+}$$ (26)

$$\mathrm{F}\mathrm{e}{\left({\mathrm{H}}_2\mathrm{O}\right)}_5{\mathrm{O}\mathrm{H}}^{2+}+\mathrm{H}{\mathrm{N}}_3\to \mathrm{F}\mathrm{e}{\left({\mathrm{H}}_2\mathrm{O}\right)}_5{\mathrm{N}}_3^{2+}+{\mathrm{H}}_2\mathrm{O}$$ (27)

The above-mentioned competitive competition and coordination reaction indicated the occurrence of a transfer process from primary reactants (ClO₂⁻ and Fe⁰) to secondary reactants (ClO₂ or Feⁿ⁺). The deprivation of ClO₂ or Feⁿ⁺ led to poor removal performance, demonstrating the enhancement of effective active species around them.

3.5. Mechanism discussion

3.5.1. Identification

The identification of active species involved the investigation of Cl-containing species and Fe-containing species. The ClO₂⁻ was consumed at 260 nm (ε = 154 M⁻¹ cm⁻¹) [[16], [17], [18]], and the formation of ClO₂ was monitored by the characteristic absorption band at 360 nm (ε = 1200 M⁻¹ cm⁻¹) [2,[16], [17], [18]] with a width at half-maximum of about 80 nm. The spectrum was unusual for the condensed phase in that it showed a vibrational structure, making this a very characteristic spectrum [40]. Moreover, the introduction of Fe⁰ made the production of ClO₂ extremely pronounced (Fig. 5b and c). Based on the measurement of ClO₂ concentrations at 360 nm, the ClO₂ produced by the H⁺-US/CI system exhibited linear relation with the action time, i.e., A_{360 nm} = 0.00443t_min (R_Adj² > 0.999) (Fig. 5a). However, the generation of ClO₂ in the H⁺-US/CI/Fe⁰ system, in contrast, increased rapidly. This could explain why the apparent activation of the H⁺-US/CI/Fe⁰ system was more effective from a certain perspective. Moreover, the studies reported that λ at around 290 nm was responsible for the presence of chlorine (as ClO⁻) [47,48], and the corresponding λ could be detected for both the H⁺-US/CI and H⁺-US/CI/Fe⁰ systems with the hidden absorption.

Fig. 5.

UV–vis spectroscopy for the ClO₂⁻ and ClO₂ during catalysis/activation: a, concentration variation at 360 nm; b, H⁺-US/CI system; c, H⁺-US/CI/Fe⁰ system. Conditions: a, ibid., Fig. 1; b–c, Fe⁰_f (80 mg L⁻¹), CI (12 mM), 1 M H₂SO₄ (0.8 mL L⁻¹). Detection: a–b, undiluted; c, diluted, sample:water (1:2, v/v); b–c, detection time (30 min).

Recently, PMSO has been used as a chemical probe to identify Fe^IV/V-oxo (HVI-oxo) species by turning itself into sulfones (PMSO₂) due to the aforementioned [O] transfer process. However, noteworthy, it was important to discuss whether PMSO was suitable for HVI-oxo identification. The studies reported that RMSO (e.g., DMSO) could mask aqueous chlorine (e.g., Cl₂ or ClO⁻) [38,39]; however, Cl-containing species by identification were ClO₂ (the difference was obvious after the introduction of Fe⁰) in this study. Except under specific catalysts, the ClO₂ showed a limited ability to directly oxidize sulfoxides to sulfones [49,50]. Unlike sulfide oxidation processes, part of the sulfones was generated [49]. Therefore, by elimination process, the ClO⁻ was also likely to cause the generation of PMSO₂. As a result, this might be a good explanation for why a small amount of PMSO₂ (retention time (t_R) = 9.27 min, m/z of 156, 141, 125, 94, and 77) was detected by GC/MS (Fig. 6), and the remaining most PMSO (t_R = 8.43 min, m/z of 140, 125, 97, and 77) was present in the H⁺-US/CI system. Then the introduction of Fe⁰ led to the production of PMSO₂ in large quantities (Fig. 6), which should be attributed to HVI-oxo species due to the (relatively stable) imperceptible change in ClO⁻ after Fe-activation. Briefly, the [O] of newly formed PMSO₂ was derived from ClO₂⁻ rather than from ⋅⋅⋅OH/H₂O. Moreover, this process clearly brought forward that the H⁺-US/CI/Fe⁰ system provided an effective preparation method for directional oxidation from PMSO to PMSO₂.

Fig. 6.

Comparisons of GC/MS detection with PMSO as a chemical probe: PMSO dosage (0.3 mM); other conditions ibid., Fig. 1.

3.5.2. Transformation

Consequently, the results further indicated that produced HVI-oxo species and ClO₂ exhibited a certain transforming relationship. Previously, Fabian and Gordon proposed the formation of FeClO₂²⁺ (Fe³⁺ + ClO₂⁻) intermediates, but they did not deny the existence of Fe^IV [[24], [25], [26]]. According to FeClO₂²⁺, Wang et al. proposed the combination of “Fe²⁺ + ClO₂”, which involved the reduction of ClO₂ to ClO₂⁻ by aqueous Fe^II. This proceeded by both outer-sphere (86%) and inner-sphere (14%) electron-transfer pathways [27]. Moreover, the traditional mainstream views of the catalysis of ClO₂⁻ by metalloporphyrin or Clds were as follows (considering Fe-Pors as an example, Scheme 1), heterolytic (Route 1, to obtain Compound I [Por^+•⋅⋅⋅Fe^IV=O] and hypochlorite (HOCl/⁻OCl)) and homolytic (Route 2, to obtain chlorine monoxide (ClO•) and Compound II [Por⋅⋅⋅Fe^IV=O]) cleavage of ClO₂⁻ [10,11].

Scheme 1.

The possible mechanisms of ClO₂⁻ cleavage.

Recently, a study used the natural substrate, ClO₂⁻, and the model substrate, PAA, under x-Cld role to reveal the formation of distinct intermediates; i.e., the transient triplet-state biradical species distinct from Compound I and Compound II [11]. Moreover, the studies also offered convenient routes to produce ClO₂ in water by catalyzing ClO₂⁻ via non-heme Fe^III- or Mn^II/III-complexes. The processes underwent the proton-coupled electron transfer (Mn^II) [2] or initiated the formation of high-valent Fe and Mn^IV–OH intermediates with oxidant PAA (Mn^III) [18].

However, combined with the identifications in this study and based on the analogy of ClO₂⁻ catalysis of Mn^II-complexes [2], a possible Fe⁰ catalytic mechanism for ClO₂⁻ to form interdependent high-valent-Fe-oxo and ClO₂ was proposed under protonation (Scheme 2). The proposed Fe⁰ catalytic mechanism also involved Fe^II and Fe^III activation processes. Relying on the role of H⁺, the surface corrosion of Fe⁰ to form ionic Fe led to interfacial catalysis, and then the ultrasonic field provided the driving force for the mass transfer to undergo diffuse catalysis (a transition process from interfacial catalysis to liquid phase (bulk phase) catalysis). Furthermore, the ultrasonic cavitation could provide the activation of HClO₂ to produce ClO₂ and polish the surface of Fe⁰ to continuously expose fresh Fe⁰ sources (Fig. 2). Scheme 2 presented that the ions release of Fe⁰ led to subsequent chain transfer processes of [Fe^II(H₂O)₆]²⁺ [51] and [HO–Fe^III(H₂O)₅]²⁺, which sourced from the hydrolysis of [Fe^III(H₂O)₆]³⁺ [52,53]. Direct addition of Fe^II and Fe^III afforded the pseudo-second-order kinetics dependence, while Fe⁰ expressed the pseudo-1.5-order kinetics dependence after the ion accumulation stage (0–7 min, Fig. 1b), which indicated that the ion leaching of Fe⁰ affected reaction equilibrium. The as-formed HVI-oxo species, O=Fe^IV or [O=Fe^IV]^•+, corresponding to the catalytic transfer station: one pathway generated ClO₂, which involved protonation of the oxo ligand on Fe^IV and thus accounted for the observed pH dependence (pH with great effect) for ClO₂ production (Fig. 5c) [2]. Other pathways led to the formation of ClO₃⁻ and Fe^II, which rapidly reacted with ClO₂ to form [HO–Fe^III]²⁺, thus re-entering the catalytic cycle [2] (macroscopically being consistent with the process: Fe²⁺ + ClO₂ → FeClO₂²⁺ → Fe³⁺ + ClO₂⁻ [[24], [25], [26]]). The dominant species, namely, [HO–Fe^III]²⁺, reacted with ClO₂⁻ in the presence of protons via homolytic Cl–O bond cleavage to form ClO• and regenerate the O=Fe^IV species or via heterolytic Cl–O bond cleavage to form ClO⁻ and regenerate the [O=Fe^IV]^•+ species [2,10]. The generated ClO• or ClO⁻ with strong reactivity entered reaction toward the low-valence Cl⁻. Furthermore, the previous research also reported and supported that the [Fe^II(H₂O)₆]²⁺ under HClO could allow the production of O=Fe^IV [54].

Scheme 2.

Proposed reaction mechanism for Fe_b-activated ClO₂⁻ by the role of protonation.

3.5.3. Action

All these abovementioned catalysis/activation processes were based on Fe_b and ClO₂⁻; therefore, how to relate to the degradation process for removing contaminants?

In contrast, with iron-free catalysis, the H⁺-US/CI system for pollutant removal underwent autocatalysis due to the formation process of ClO₂. Kieffer and Gordon reported that Cl⁻ could catalyze the disproportionation of HClO₂ [31]. Cl⁻ could alter the reaction from equation (5) to equation (28) [31]. Furthermore, Horvath et al. considered that the autocatalysis concerning H⁺, Cl⁻, and key intermediate, HOCl, led to the unique redox reaction [55]. Olagunju et al. reported that ClO₂ production after the induction period was due to the reaction of the intermediate HOCl species with ClO₂⁻ [56]. The apparent piecewise and non-integer kinetics observed in this study also supported the abovementioned result. The accumulation of intermediates, Cl⁻ or ClO⁻ led to a self-catalyzed reaction to boost the yield of ClO₂, following the “S”-curve of the removal process by using the H⁺-US/CI system (Fig. 1a). Attention: in general, purity of commercial sodium chlorite is around 80%, which is accompanied by part of chloride ions; therefore, the commodity might have itself contributed to autocatalysis.

$${4\mathrm{C}\mathrm{l}\mathrm{O}}_2^{-}+2{\mathrm{H}}^{+}\to {{\mathrm{C}\mathrm{l}\mathrm{O}}_3^{-}+\mathrm{C}\mathrm{l}}^{-}+2\mathrm{C}\mathrm{l}{\mathrm{O}}_2+{\mathrm{H}}_2\mathrm{O}$$ (28)

However, in terms of the H⁺-US/CI/Fe_b system, the coordination catalysis, i.e., effective contact catalysis/activation, could be proposed based on the abovementioned analyses presented in sections 3.5.1 and 3.5.2. The as-formed coordination center [Fe^N(H₂O)_x]ⁿ⁺ could be obtained by complexation of reactive intermediates such as ClO₂ (ClO₂⁻) or ClO⁻ by substituting H₂O, and its valence state could also be affected to form the possible [Fe^N(H₂O)_a(ClO_x^m−)_b]ⁿ⁺ complex, briefly, Fe^NClO_x^M+ species. This process could be analogous to hemoglobin carrying oxygen, which could act as one transmission type. The [Fe^N(H₂O)_x]ⁿ⁺ could carry ClO₂ throughout the liquid phase to increase the chance of collision or contact among the molecules, particularly when driven by the US. Owing to the abovementioned coordination, the solute contaminant macromolecules were caught by the water molecules with intermolecular forces (similarly, pollutant molecules were trapped by the friction) to form localized/limited catalysis or activation, namely, [Fe^N(H₂O)_a(ClO_x^m−)_b]ⁿ⁺-P_tpm complexes, which easily led to the outer-sphere and inner-sphere electron transfers. Furthermore, the processes of free radical and non-free radical oxidation and active species for removing contaminants could all be produced. The above-mentioned catalytic processes were generally selective for contaminant structures.

4. Conclusions

• (1)The system, namely, Fe_b activating protonated ClO₂⁻, could be utilized to synchronously remove inorganic ions (ClO₂⁻) and organic pollutants (tpm-derivatives) by interdependent high-valent-Fe-oxo and ClO₂. However, the redox system presented selective oxidation rather than non-selective oxidation for contaminants removal.
• (2)Differences were observed between metal porphyrin/complexes and ion catalysis. Fe_b-induced activation of protonated ClO₂⁻ seemed to approach Mn-porphyrin/complexes catalysis of ClO₂⁻; both processes produced ClO₂. In contrast, Fe-Cld/porphyrin catalysis more tended to catalyze ClO₂⁻ to produce Cl⁻ and O₂.
• (3)Interestingly, this study further provided the preparation methods of two products, i.e., ClO₂ and sulfones. The Fe_b-activated ClO₂⁻ could correspond to a method of oriented oxidation of sulfoxides to sulfones and environmentally allow for ClO₂ preparation for the on-site use (e.g., disinfection or sterilization) and further ClO₂ production.
• (4)Owing to the micro/small dosage of activators used in this study, the autocatalytic process became explicit. In general, the studies were based on a high dosage of activators or homogeneous activation except for the autocatalysis caused by the pollutant structures (e.g., hydroquinone or benzoquinone), which shortened the autocatalytic accumulation processes. However, it has also made a preliminary exploration for the controlled release-catalysis of ionic Fe with small/micro dosages.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.