Performance and mechanism of sulfamethoxazole degradation on a Magnéli-phase Ti₄O₇ reactive electrochemical membrane

Abstract

Antibiotics have emerged as a new class of environmental contaminants of global concern due to their persistence and potential ecotoxicological impacts. In order to enhance antibiotic removal efficiency, it is necessary to develop cost-effective and environmentally friendly wastewater treatment technologies. Magnéli-phase titanium oxide (Ti₄O₇) presents a promising solution, combining excellent metallic-like conductivity with ceramic-like corrosion resistance. In this study, the electrochemical oxidation of sulfamethoxazole (SMX) with a Ti₄O₇ reactive electrochemical membrane (REM) as the anode and stainless steel as the cathode was systematically investigated. The effects of key operational parameters, including current density, initial pH, supporting electrolyte type and concentration, and the presence of humic acid (HA) on SMX degradation were evaluated. Under optimal conditions (current density of 0.08 mA cm⁻², pH of 7, supporting electrolyte of 0.1 mol L⁻¹ Na₂SO₄, no HA effect, and pump circulation), 99.7% SMX degradation was achieved within 2 hours. Quenching experiments and electron paramagnetic resonance (EPR) analysis confirmed that hydroxyl radicals (˙OH) played a dominant role in SMX degradation. High-performance liquid chromatography coupled with quadrupole time-of-flight tandem mass spectrometry (HPLC-Q-TOF-MS/MS) identified key degradation intermediates, revealing three primary SMX degradation pathways: (1) ˙OH-mediated amino group oxidation, (2) S–N bond cleavage, and (3) –NH₂ hydroxylation. Additionally, Ecological Structure Activity Relationships (ECOSAR) software analysis demonstrated a significant reduction in overall toxicity following Ti₄O₇ oxidation. In summary, the Ti₄O₇ REM exhibits exceptional oxidation efficiency with low energy consumption, demonstrating strong practical feasibility and promising application potential for antibiotic-contaminated wastewater treatment.

---

1 Introduction

The frequent use of antibiotics and the low removal efficiency of traditional wastewater treatment plants for antibiotics have led to their gradual accumulation in natural water bodies, which poses a threat to human health by inducing antibiotic resistance.^1,2 Therefore, it is urgent to treat antibiotics to ensure their removal before discharging into natural water bodies. Among numerous antibiotics, sulfonamide drugs are widely used in human and veterinary medicine due to their broad-spectrum antibacterial activity, and sulfamethoxazole (SMX) is the most frequently detected substance in wastewater.³ SMX exhibits high biological activity, strong bioaccumulation, difficulty in degradation, and low removal efficiency, posing significant health risks to humans, livestock, and plants.

The traditional wastewater and drinking water treatment systems exhibit poor removal efficiency of SMX.^4,5 Therefore, previous studies have mainly focused on enhancing SMX removal by using methods such as chemical oxidation,⁶ advanced oxidation processes (AOPs),^7–9 photodegradation,¹⁰ and reverse osmosis.¹¹ Although these methods indicate effective removal of SMX, some limitations and shortcomings still exist. For example, ozone oxidation demonstrates effective removal of SMX, but shows ineffective mineralization, and requires high investment costs and high ozone dosages. AOPs require continuous oxidants and catalysts for the degradation of wastewater, resulting in high operating costs and complex operations. In reverse osmosis, the membrane has a short lifespan and requires the addition of salt solution for further treatment, which is a complex and expensive operation.

Electrochemical advanced oxidation processes (EAOPs),^12,13 as an effective water and wastewater treatment technology, have been extensively researched and developed due to their versatility, non-selectivity, environmental acceptability, safety, and high energy efficiency.¹⁴ Among numerous EAOPs, anodic oxidation (AO) is the most commonly studied, and pollutants can be removed by direct electron transfer (direct oxidation) or by generating active species in situ (indirect oxidation), such as hydroxyl radicals (˙OH), hydrogen peroxide (H₂O₂), and sulfate radicals (SO₄⁻˙).¹⁵ The anode materials are the key factor determining the efficiency of the AO process, which are generally classified into two major categories based on their catalytic performance: active anodes (Pt, RuO₂, IrO₂) and inactive anodes (PbO₂, SnO₂, BDD). The active anodes have low oxygen evolution potential, which promotes the formation of oxides or superoxides on the anode surface, ultimately leading to strong adhesion of ˙OH species to the electrode surface. Conversely, the inactive anodes exhibit weak ˙OH adhesion on their surface, allowing the ˙OH to attack pollutants without selectivity, thereby augmenting the effectiveness of the oxidation.¹⁶ Considering the sufficient amounts of ˙OH production, inactive anodes are regarded as the most effective anodes in EAOPs processes. However, in inactive anodes, PbO₂ shows poor stability and short service life, and the catalytic oxides (mainly toxic Pb²⁺) will leach in solution, causing secondary pollution.¹⁷ Similarly, SnO₂ also has the disadvantages of short usage time and insufficient stability, while BDD is limited due to high manufacturing costs.¹⁸ Therefore, developing new anode materials with high catalytic activity, high stability, and environmentally friendly characteristics is the key to promoting the practical application of advanced electrochemical oxidation technology.

Over the past decade, substoichiometric titanium oxides based on Magnéli-phase ceramic electrodes, particularly Ti₄O₇ as an inactive anode, have been studied due to their low cost and high oxidation potential.^19–24 Some research reports indicate that the electro-oxidation activity of Ti₄O₇ is significantly higher than that of active anodes, such as dimensionally stable anode (DSA) or Pt, and can even match the electro-oxidation activity of BDD.^19,21 More importantly, porous reactive electrochemical membrane (REM) utilizing Ti₄O₇ can be fabricated via a cost-effective and scalable plasma coating method. When operated in flow-through mode, the REM-based system demonstrates superior current efficiency compared to conventional systems, attributed to the enhanced mass transfer of pollutants from the bulk solution to the anode surface.^25–27 However, previous studies have mainly focused on the removal of tetracycline, amoxicillin, humic acid, etc. by the Ti₄O₇ anode, and there is still a lack of systematic research on the removal of sulfonamide antibiotics. Furthermore, most previous studies have only explored a relatively single influencing factor and have not explored the factors that affect Ti₄O₇ REM from multiple perspectives. The optimal operating conditions and degradation mechanism of SMX removal by Ti₄O₇ REM are still lacking systematic research. In addition, it is necessary to conduct specific analysis of the toxicity changes of intermediate products and combine them with the ecological environment to ensure the safety of the water body after electrochemical oxidation.

In this study, a flow-through electrochemical filtration system was developed, comprising tubular anodes made of Magnéli-phase Ti₄O₇ and stainless steel (SS) tubular cathodes, for the removal of SMX. The effects of current density, initial pH, background electrolyte and humic acid (HA) on the degradation of SMX in Ti₄O₇ REM were systematically studied. The degradation pathways of SMX were identified by high-performance liquid chromatography coupled with quadrupole time-of-flight tandem mass spectrometry (HPLC-Q-TOF-MS/MS). Furthermore, toxicity analysis during the process was conducted using the Ecological Structure Activity Relationships (ECOSAR) software. This work will provide valuable insights into the electrochemical degradation of refractory organic pollutants in water and wastewater.

2 Materials and methods

2.1. Chemicals

All chemicals are of analytical grade or above and ready for use without the need for further purification. Further details can be found in Text S1 in SI.

2.2. Fabrication and characterization of Ti₄O₇ REM

The fabrication method for Ti₄O₇ REM can be found in Text S2 in SI. The sample morphology of Ti₄O₇ REM was characterized using a Quanta-250 environmental scanning electron microscope (SEM, Phillips/FEI, USA). Crystalline phase analysis was performed with a D8 Advance X-ray diffractometer (XRD, Bruker, USA) employing Cu Kα radiation. Surface chemical composition was determined by X-ray photoelectron spectroscopy (XPS) using an Escalab 250 Xi system (Thermo Fisher Scientific, USA) under high vacuum (1 × 10⁻⁹ Torr), with all binding energies referenced to the C 1s peak at 284.8 eV.

2.3. Experimental setup and operation

All electrolysis experiments in the study were conducted in an open 1160 mL cylindrical reactor. In this reactor, Ti₄O₇ REM (diameter of 35 mm, length of 130 mm, effective area of 137 cm²) served as the anode, and a stainless steel mesh (diameter of 39 mm, length of 130 mm) functioned as the cathode. The anode was positioned centrally within the reactor and encircled by the cathode, maintaining a distance of 2 mm between the electrodes. A direct current (DC) regulated power supply was utilized to supply a constant current intensity. The schematic diagram of the reaction apparatus is shown in Fig. S1. During the experiment, 10 mg L⁻¹ SMX and a certain concentration of electrolyte were continuously circulated under a peristaltic pump to ensure the solution was thoroughly mixed at room temperature (23 ± 2 °C). The degradation experiment lasted for 120 minutes, and the samples were taken at 0, 5, 15, 30, 45, 60, 80, 100, and 120 minutes. After sampling, the samples were filtered through a mixed fiber membrane with a pore size of 0.22 µm and immediately transferred to a liquid-phase bottle containing 1 mol L⁻¹ methanol. The influencing factors of current densities (0.008–0.8 mA cm⁻²), pH values (3–11), electrolyte concentrations (0.01–0.2 mol L⁻¹), electrolyte types (NaCl, Na₂SO₄ and NaNO₃), HA (0–80 mg L⁻¹), and circulation were systematically studied.

2.4. Analytical methods

The analytical methods, including the determination of SMX concentration, identification of free radical species, measurement of intermediate products, and toxicity assessment, are detailed in Text S3.

3 Results and discussion

3.1. Characterization and electrochemical properties of Magnéli phase Ti₄O₇ REM

The microstructure of Magnéli-phase titanium oxide can be observed through rutile-type TiO₂ derivatives.²⁸ During the reorientation process of the TiO₆ octahedral structure, the partial reduction of Ti(IV) to Ti(III) generates defect sites (oxygen vacancies) that lead to the overlap of Ti 3d orbitals within the crystal lattice. This overlap of Ti 3d orbitals is considered to be the reason for the high conductivity of the oxygen-deficient Magnéli-phase.²⁹ As shown in Fig. 1(a), the SEM image of the Magnéli-phase Ti₄O₇ REM reveals a rough and dense honeycomb-like surface. During high-temperature annealing, the molten nano-TiO₂ powder tends to form Ti₄O₇ aggregates (with diameters ranging from 0.5 to 2.0 µm) during cooling, thus creating abundant pores within the Magnéli-phase Ti₄O₇. This increases the pore volume of the electrode (mesopores and macropores), thereby enhancing the effective contact between organic molecules in the solution and the electrode surface at the electrode/solution interface. The porous microstructure also facilitates flow operations, overcoming the mass transfer limitations (convection) associated with conventional batch electrochemical methods. The XRD pattern of the prepared electrode (Fig. 1(b)) shows major diffraction peaks at 20.7 nm, 26.3 nm, 40.5 nm, etc., consistent with Ti₄O₇.³⁰ The chemical composition of the Ti₄O₇ REM was further investigated through XPS (Fig. 1(c)). The XPS scanning results indicated the presence of Ti 2p, O 1s, and C 1s on the surface of the Magnéli-phase Ti₄O₇ REM. The C 1s peak is attributed to accidental carbon contamination associated with XPS technology, while the O 1s spectrum can be deconvoluted into three peaks centered at 529.9 eV, 532.6 eV, and 531.2 eV, corresponding to lattice oxygen, surface hydroxyl groups, and oxygen vacancies, respectively. The Ti 2p XPS spectrum (Fig. 1(d)) reveals that both Ti 2p3/2 (458.8 eV) and Ti 2p1/2 (464.4 eV) originate from Ti–O bonding. The binding energy difference between these two characteristic peaks (5.6 eV) suggests that the primary valence state of Ti in the electrode material is +4.³¹

Fig. 1 Characterization of Magnéli-phase Ti₄O₇ REM: (a) SEM image, (b) XRD spectrum, (c) full-spectrum XPS, and (d) Ti 2p XPS spectrum.

The electrochemical degradation reaction is essentially the transfer of charge between an electronic conductor and an ionic conductor, and the inevitable transformation of biomass or pollutants occurs at the interface between the two phases. This study analyzed the Magnéli-phase Ti₄O₇ REM through electrochemical impedance and as its high-frequency resistance, and the Nyquist spectra of Ti₄O₇ REM are shown in Fig. S2. According to the test fitting results, the internal resistance of the Ti₄O₇ REM is 14.2 Ω, indicating its good conductivity.

3.2. Effect of operational parameters on SMX degradation

3.2.1. Effect of current density. Current density is one of the most crucial factors in the electrochemical oxidation process. It governs the transfer of electrons and the formation of active functional groups, directly influencing the removal rate of organic pollutants. Fig. 2(a) illustrates the relationship between SMX variation and electrolysis time under different current densities. Experimental results indicate that the attenuation of SMX concentration is related to the applied current density, and as the current density increases, the degradation rate of SMX accelerates. This may be attributed to the fact that higher current densities can expedite the indirect electro-oxidation rate, namely, the formation rate of ˙OH on the electrode surface.²³ With the extension of electrolysis time, SMX can be nearly completely degraded at all current densities, indicating that the Ti₄O₇ electrode has an excellent treatment effect on SMX. Through linear regression fitting of experimental data, the kinetic parameters were calculated. As shown in Fig. 2(b), the electrochemical oxidation of SMX conforms to a pseudo-first-order kinetic reaction, and there is a good linear correlation between removal efficiency and electrolysis time. The apparent rate constant k can be obtained by analyzing the slope of the curve. It can be found that the value of k increases with the increase in current density. As the current density increases from 0.008 mA cm⁻² to 0.8 mA cm⁻², the values of k vary from 0.00642 to 0.09843. Furthermore, Fig. 2(c) shows the effect of current density on the removal of TOC. It can be observed that the TOC removal rate increases with the increase of applied current density and electrolysis time. However, when the current density reaches its maximum value (0.8 mA cm⁻²), the TOC removal rate is only 46.23%, indicating that SMX was not completely mineralized by Ti₄O₇ REM. In addition, the energy consumption of the electrode during the electrocatalytic oxidation process was also calculated according to eqn (S3) in the SI. As listed in Table S1, when the current density reaches 0.08 mA cm⁻², the degradation rate of SMX can reach over 99% within 2 hours. Combined with the voltage drift under different current density states during the experimental process, the calculated energy consumption is only 9.89 kWh kg⁻¹ TOC, which is much lower than the 57.63 kWh kg⁻¹ TOC at 0.8 mA cm⁻². It can be found that when the removal rates are similar, the energy consumption continues to increase with the increase of current density. Therefore, to reduce energy consumption, 0.08 mA cm⁻² is selected as the optimal current density in this study. Moreover, Table S2 compares the energy consumption of similar electrochemical systems for removing target pollutants, providing a benchmark for the energy consumption of such electrochemical systems. From Table S2, we can see that the energy consumption of this study is lower than that of most similar electrochemical oxidation systems. Therefore, we believe that the current density selected for this experiment is relatively low-energy, environmentally friendly, and efficient.

Fig. 2 Effect of current density on SMX degradation (a), quasi-first-order kinetic fitting (b), and TOC removal (c). (Conditions: initial concentration of SMX is 10 mg L⁻¹, electrolyte Na₂SO₄ concentration is 0.1 mol L⁻¹, pH is not adjusted.)

3.2.2. Effect of initial pH. The pH value is widely believed to significantly affect the effectiveness of the electrochemical oxidation process.^23,32 Therefore, the effect of initial pH (pH = 3, 7, and 11) on SMX degradation was investigated in this experiment, with a current density set at 0.08 mA cm⁻². As shown in Fig. 3(a), the final degradation effects of SMX under pH = 3 and pH = 7 conditions are similar, and much better than that under pH = 11. However, according to quasi-first-order kinetic fitting in Fig. 3(b), it can be observed that the k value reaches a maximum of 0.04830 at pH = 7. Similarly, Fig. 3(c) shows that the TOC removal at pH = 7 is higher than that at pH = 3 and 11. It has been reported that the structure of SMX can also affect the degradation process under different pH conditions.^33,34 The pK_a values of SMX are 1.8 (pK_a1) and 5.5 (pK_a2). When pH < pK_a1, the protonated structure of SMX renders it positively charged, preventing the delocalization of unshared electron pairs from the N atom to the aromatic ring, thereby reducing its reactivity with electrophilic radicals. When pH > pK_a2, the radical oxidation activity of the aniline group is enhanced due to the deprotonation of the –NH sulfonamide group. Therefore, when pH = 7, the free radical oxidation activity of the aniline group in SMX is strong, so the removal rate is the highest under this condition. Ultimately, pH = 7 was chosen as the optimal reaction condition in this study.

Fig. 3 Effect of pH on SMX degradation (a), quasi-first-order kinetic fitting (b), and TOC removal (c). (Conditions: initial concentration of SMX is 10 mg L⁻¹, electrolyte Na₂SO₄ concentration is 0.1 mol L⁻¹, current density is 0.08 mA cm⁻²)

3.2.3. Effect of electrolyte concentration. In addition to current density, electrolyte concentration is also a factor affecting electrochemical oxidation. Therefore, the effect of Na₂SO₄ concentrations on SMX removal was studied. Among them, the concentration range of Na₂SO₄ solution was from 0.01 mol L⁻¹ to 0.2 mol L⁻¹, the initial concentration of SMX was 10 mg L⁻¹, the current density was 0.08 mA cm⁻², and the pH was 7. As depicted in Fig. 4(a) and (b), when the electrolyte concentration was 0.01, 0.05, 0.1, and 0.2 mol L⁻¹, SMX was almost completely removed within 2 hours, and the k values showed an upward trend, which were 0.04730, 0.05830, 0.06470, and 0.09440, respectively. Besides, it can be observed in Fig. 4(c) that as the electrolyte concentration increased, the removal rate of TOC also increased. The above results indicate that as the electrolyte concentration increases, the degradation rate of SMX accelerates. This may be mainly due to the enhancement of SMX mass transfer rate in the system under higher ionic strength, thus accelerating the observed degradation kinetics. The reduction in ohmic losses might be another reason for the accelerated degradation rate of SMX. However, the current density of the reaction condition in this experiment is fixed at 0.08 mA cm⁻². Under this condition, the effect of changing electrolyte concentration on ohmic loss may be minor. The above result is similar to the previous study,³⁵ which enhanced the mass transfer of target pollutants through electrolytes. However, it is worth noting that when the Na₂SO₄ concentration continues to increase after 0.1 mol L⁻¹, the degradation of SMX increases very slowly. This is because there is a critical electrolyte concentration in each electrolysis system. When the concentration is below this critical concentration, the system lacks sufficient ions to conduct current, thereby reducing oxidation efficiency. Above the critical electrolyte concentration, when the electrolyte concentration is low, the conductive ions in the solution are nearly saturated. Therefore, increasing the electrolyte concentration barely enhances electrolysis efficiency. Considering both the critical electrolyte concentration and the cost-effectiveness of the experiment, it was ultimately decided to use 0.1 mol L⁻¹ Na₂SO₄ as the optimal electrolyte concentration.

Fig. 4 Effect of electrolyte concentration on SMX degradation (a), quasi-first-order kinetic fitting (b) and TOC removal (c). (Conditions: initial concentration of SMX is 10 mg L⁻¹, current density is 0.08 mA cm⁻², pH = 7.)

3.2.4. Effect of electrolyte types. The supporting electrolyte can generate oxidizing species and react with organic pollutants in the solution, which will have a great impact on the oxidation ability of this experimental system. In this study, 0.1 mol L⁻¹ Na₂SO₄, NaCl and NaNO₃ were selected as supporting electrolytes, and their effects were compared under the conditions of current density of 0.08 mA cm⁻² and pH value of 7. As shown in Fig. 5(a), the removal rate of SMX is the highest when Na₂SO₄ is used as the background electrolyte, reaching 99.7%. Meanwhile, Fig. 5(b) and (c) also show that the removal rate of quasi-first-order kinetic fitting and TOC is the best when Na₂SO₄ is used as the background electrolyte. In addition, we also compared the energy consumption and voltage drift in the process of removing SMX with three different supporting electrolytes (Table S3). The results showed that when NaCl and NaNO₃ were used as electrolyte, the energy consumption of electrolyte was much greater than that when Na₂SO₄ was used as electrolyte, which was similar to previous studies.³⁶

Fig. 5 Effect of electrolyte type on SMX degradation (a), quasi-first-order kinetic fitting (b), and TOC removal (c). (Conditions: initial concentration of SMX is 10 mg L⁻¹, current density is 0.08 mA cm⁻², pH = 7, electrolytes Na₂SO₄, NaCl and NaNO₃ are all 0.1 mol L⁻¹)

The observed efficient degradation of SMX under Na₂SO₄ supporting electrolyte could be attributed to the following factors. According to previous studies, when Na₂SO₄ was used as the supporting electrolyte, S₂O₈²⁻ may be generated at the inactive anode, as listed in eqn (1).³⁷ S₂O₈²⁻ is a strong oxidizer, with a standard redox potential E₀ of +2.01 V, which is close to that of ozone (E₀ = +1.70 V) and higher than that of permanganate (E₀ = +1.51 V), thus possessing a strong ability to oxidatively degrade organic pollutants.³⁸ Then, according to the previous study,³⁹ it would be reduced at the cathode (eqn (2)). The E₀ value of SO₄⁻˙ is 2.5–3.1 V, exceeding the E₀ value of ˙OH (E₀ = 1.9–2.7 V),⁴⁰ and its half-life is longer than that of ˙OH, indicating better stability. It can travel a longer distance during the reaction time, which is more conducive to the oxidative degradation of organic pollutants.^41,42

2SO²⁻₄ → S₂O²⁻₈ + 2e⁻ (1)

S₂O²⁻₈ + e⁻ → SO⁻₄˙ + SO²⁻₄ (2)

When NaCl is used as the electrolyte, ClO⁻ can be generated through the reaction of Cl⁻ and OH⁻ (eqn (3)). ClO⁻ is a strong oxidant with a standard redox potential E₀ of +1.67 V, thus can effectively promote SMX degradation.⁴³ However, as illustrated in eqn (4), the presence of Cl⁻ may lead to competitive reactions by scavenging ˙OH generated in the electrolyte through radical chain transfer reactions, leading to poor SMX removal efficiency under the NaCl supporting electrolyte.

Cl⁻ + OH⁻ → ClO⁻ + H₂O + 2e⁻ (3)

Cl⁻ + ˙OH → Cl˙ + OH⁻ (4)

When NaNO₃ was used as the electrolyte, the anodic reactions in eqn (5)–(8) occurred. These substances, particularly NH₃, can react with ˙OH,⁴⁴ leading to a decrease in SMX degradation. Based on comprehensive experimental results, 0.1 mol L⁻¹ Na₂SO₄ was selected as the optimal electrolyte type for the reaction system.

NO⁻₃ + 6H₂O + 8e⁻ → NH₃ + 9OH⁻ (5)

NO⁻₃ + 2H⁺ + 2e⁻ → NO⁻₂ + H₂O (6)

NO⁻₂ + 5H₂O + 6e⁻ → NH₃ + 7OH⁻ (7)

2NH₃ + 6˙OH → N₂ + 6H₂O (8)

3.2.5. Effect of HA. Natural organic matter (NOM) is widely distributed in surface water and has a significant impact on the degradation of antibiotics in electrochemical oxidation systems. This study used HA as a typical NOM and investigated its effect on the degradation of SMX in the Magnéli-phase Ti₄O₇ system at concentrations of 0, 10, 20, 40, and 80 mg L⁻¹. As shown in Fig. 6(a), the addition of HA inhibited the electrochemical degradation of SMX. After 2 hours of electrolysis, SMX was completely removed without the presence of HA. However, under the same experimental conditions, when HA is present, residual SMX was detected in the solution after the reaction was completed. Furthermore, as shown in Fig. 6(b), with the addition of HA, the k value decreased from 0.06000 to 0.02665, indicating that the presence of HA can react with ˙OH and inhibit the degradation of SMX. On the one hand, HA may complex with antibiotics, and its significant steric hindrance can affect the migration of antibiotics. On the other hand, HA will compete with antibiotics and mineralize on the anode surface.⁴⁵ Based on the analysis of the experimental results, 0 mg L⁻¹ HA was ultimately selected as the optimal reaction condition.

Fig. 6 Effect of HA on SMX degradation (a) and quasi-first-order kinetic fitting (b). (Conditions: initial concentration of SMX is 10 mg L⁻¹, electrolyte Na₂SO₄ concentration is 0.1 mol L⁻¹, current density is 0.08 mA cm⁻², pH = 7.)

3.2.6. Effect of circulation. To verify the effect of flow-through mode on electrochemical oxidation, the removal of SMX under cycle and no cycle conditions was compared, based on the optimal reaction conditions. As shown in Fig. S3, the removal efficiency of SMX in the cycle condition is significantly higher than that in the no cycle condition. This is mainly because the reaction solution circulates under the suction of the peristaltic pump. During the circulation process, dissolved oxygen is converted into ˙OH through the action of the cathode, and then injected into the Ti₄O₇ electrode. At this time, ˙OH is electrostatically attracted by the anode, which accelerates the mass transfer rate at the interface.

3.3. Intermediate products and degradation pathways of SMX

3.3.1. Identification of key substances. The free radicals generated on the anode surface are considered to be the main oxidant for pollutant degradation in EAOPs. As shown in Fig. 7(a), two obvious EPR signals of DMPO-˙OH adduct and DMPO-SO₄⁻˙ adduct were detected in the Ti₄O₇ REM system.⁴⁶ ˙OH has a high standard redox potential (E₀ = 1.9–2.7 V) and can mineralize organic pollutants non selectively. The formation of ˙OH in the system can be explained by eqn (9).⁴⁷ SO₄⁻˙ is a strong oxidant with a redox potential of 2.5–3.1 V. Previous studies have shown that SO₄⁻˙ (eqn (10)) can be formed by sulfate oxidation on inactive anode,⁴⁸ and the specific generation process of SO₄⁻˙ has been introduced in 3.2.4.

Ti₄O₇ + H₂O → Ti₄O₇(˙OH) + H⁺ + e⁻ (9)

SO²⁻₄ → SO⁻₄˙ + e⁻ (10)

In order to further explore the free radicals that play a major role in this study, quenching experiments was conducted. Tert-Butanol (TBA) and methanol (MeOH) were selected as the free radical trapping agents. The reaction rate constant of TBA with ˙OH is 2.8 × 10⁹ M⁻¹ s⁻¹,⁴⁹ which can quickly capture the ˙OH generated on the electrode surface, and no other radicals are generated during the reaction pathway. MeOH can serve as an effective scavenger for both ˙OH (k = 9.7 × 10⁸ M⁻¹ s⁻¹) and SO₄⁻˙ (k = 1.1 × 10⁷ M⁻¹ s⁻¹) radicals.⁵⁰ Therefore, to investigate the active species playing a major role in the electro-oxidation reaction, 0.1 mol L⁻¹ MeOH and 0.1 mol L⁻¹ TBA were added to the SMX electrocatalytic oxidation degradation process for capture experiments. The contributions of the two capture agents to the two types of radicals were explored and quantitatively analyzed. As can be seen from Fig. 7(b), when TBA is added to the electrochemical system, the ˙OH generated by electrocatalytic oxidation is largely captured, resulting in a sharp decrease in the removal rate of SMX from 99.3% to 30.2%. Similarly, in the presence of MeOH, both ˙OH and SO₄⁻˙ generated in the system are captured and removed, significantly inhibiting the electrocatalytic activity of the system. However, there is no significant difference in the impact of the two scavengers on the removal of SMX in the electrochemical system, indicating that SO₄⁻˙ is not the main active species in the system, while ˙OH plays a dominant role in the electrooxidative degradation of SMX.

Fig. 7 EPR spectrum of DMPO capturing ˙OH and SO₄⁻˙ at 15 minutes (a) and effects of different trapping agents on SMX degradation (b). (Conditions: initial concentration of SMX is 10 mg L⁻¹, electrolyte Na₂SO₄ concentration is 0.1 mol L⁻¹, concentrations of MeOH and TBA are 0.1 mol L⁻¹, current density is 0.08 mA cm⁻², pH = 7.)

3.3.2. Degradation pathway. Based on the analytical results of intermediates derived from mass spectrometry, the electrochemical oxidative degradation pathway of SMX in the Magnéli-phase Ti₄O₇ system was proposed, as illustrated in Fig. 8. The mass-to-charge ratios (m/z), molecular formulas, and potential structures of the intermediate products can be accessed in Table S4 in the SI. Overall, the degradation of SMX can primarily be categorized into the following three pathways.

Fig. 8 Possible electrocatalytic degradation pathways of SMX in the Magnéli-phase Ti₄O₇ system.

In pathway I, the primary reaction is the oxidation of amine groups, resulting in the formation of nitro-SMX derivatives.⁵¹ Amino is oxidized by ˙OH to produce a nitro-SMX derivative. This may be due to rapid and reversible hydration, which leads to the formation of an aniline radical at the N-center. The aniline radical is further neutralized by molecular oxygen, resulting in the hydrolysis of hydroxylamine and nitroso derivatives into nitro-SMX.⁵² Afterwards, ˙OH attacks the benzene ring, generating TP2 and TP4. Through the continuous hydroxylation of these primary by-products, SMX can ultimately be mineralized into CO₂, H₂O, SO₄²⁻, and NO₃⁻. These reaction pathways are consistent with previous studies on SMX degradation.^34,53 In pathway II, the S–N bond serves as the electron-rich site of SMX, which is attacked by ˙OH, leading to the breakage of the sulfonamide bond to generate TP5.⁵⁴ The ˙OH can further attack the isoxazole ring to produce the ring-opening product TP6, which is similar to previous studies.³⁷ In pathway III, the –NH₂ group on SMX aniline undergoes hydroxylation, leading to the formation of TP7 and intermediate product TP1, and is further transformed into TP8.

3.4. Toxicity analysis

Due to the generation of numerous intermediate products during the electrolysis process, it is imperative to assess the potential environmental risks posed by these products. The acute and chronic toxicity of SMX and its intermediate products to daphnia, green algae, and fish is estimated based on the globally harmonized system (GHS) of classification and labelling of chemicals. According to the GHS, toxicity data is categorized into four levels, namely highly toxic (LC 50/EC 50/ChV ≤ 1 mg L⁻¹), toxic (1 mg L⁻¹ < LC 50/EC 50/ChV ≤ 10 mg L⁻¹), harmful (10 mg L⁻¹ < LC 50/EC 50/ChV ≤ 100 mg L⁻¹), and harmless (LC 50/EC 50/ChV > 100 mg L⁻¹).^55,56 The results predicted by the ECOSAR program are presented in Table 1. The results demonstrated that SMX exhibits varying levels of toxicity towards the aforementioned organisms. Daphnia magna is also the most sensitive organism to SMX and its by-products. The toxicity of the degradation end-products of SMX shows a decreasing trend, except for the intermediate product TP5, whose toxicity decreases. We can also observe that the toxicity of the final products TP4, TP6 and TP8 is significantly lower than that of SMX. Among them, we can observe that each index of TP4 after treatment has reached a harmless state; TP6 only had high chronic toxicity to Daphnia, and other toxicities were reduced. TP8 has slight toxicity to Fish Daphnia Green, and other toxicity has also been well controlled. More importantly, the significant reduction in toxicity level (from high toxicity to low toxicity/non toxicity) indicates that Ti₄O₇ REM not only destroys the target pollutant structurally, but also realizes a key environmental detoxification process. This ability to convert high environmental risk parent compounds into low environmental risk conversion products is one of the core indicators to evaluate whether advanced oxidation technology can be safely applied. The toxicity of the final product is far below the threshold of chronic damage to sensitive aquatic organisms at the population level. Therefore, the Magnéli-phase Ti₄O₇ system reduces the overall toxicity of SMX and has great potential in protecting aquatic ecosystems.

Table 1 Predicted toxicity of SMX and its TPs by ECOSAR program

---

4 Conclusions

In this study, a stable and effective electrochemical filtration system using Magnéli-phase Ti₄O₇ REM was successfully established for the removal of SMX from water. The effect of current density, pH, electrolyte concentration and types, HA, and circulation on SMX degradation was systematically investigated. The results indicate that the degradation of SMX follows first-order reaction kinetics. Within the current density range of 0.008–0.8 mA cm⁻², the degradation rate of SMX increases with increasing current density. The optimal parameters for SMX degradation by Ti₄O₇ are under cyclic conditions, current density is 0.08 mA cm⁻², electrolyte is 0.1 mol L⁻¹ Na₂SO₄, pH = 7, and no HA. Under this condition, the removal efficiency of SMX reaches over 99.7%, which depends on the ˙OH generated at the anode, indicating that this technology has good electro-oxidation efficiency. It is inferred that the degradation pathway of SMX can be divided into three types: oxidation of the amino group by ˙OH, breakage of the S–N bond, and hydroxylation of –NH₂. Meanwhile, a toxicity evaluation of the SMX degradation process was conducted, and the results showed that the overall toxicity decreased after 2 hours of SMX treatment. In summary, the Magnéli-phase Ti₄O₇ REM may provide a new effective strategy for the removal of refractory organic pollutants in water and wastewater.

Author contributions

Qiang Liu: methodology, validation. Yinuo Wang: formal analysis, writing-original draft, writing-review & editing. Boxuan Yang: methodology, figure supplementation and test polishing. Jinwei Li: conceptualization. Chengchun Jiang: visualization, supervision. Huan Wang: investigation.

Conflicts of interest

The authors declare that there are no conflicts of interest.

Data availability

The data supporting the findings of this study are available from the corresponding author upon reasonable request.

Supplementary information (SI) is available. See DOI: https://doi.org/10.1039/d5ra06264k.

Acknowledgements

This work is supported by Shenzhen Science and Technology Program (RCBS20231211090623038), the National Natural Science Foundation of China (No. 52170007), Shenzhen Polytechnic University Research Fund (6023310040K), and Natural Science Foundation of Guangdong Province of China (No. 2023A1515012063).

Notes and references

1. R. Meffe and I. de Bustamante, Sci. Total Environ., 2014, 481, 280–295.
2. I. Ali, A. Barros de Souza, S. De Laet, K. Van Eyck and R. Dewil, Chemosphere, 2023, 319, 137984.
3. W. Yang, H. Dong, Y. Tian, Y. Tong, J. Lu, L. Liao and J. Niu, Sep. Purif. Technol., 2025, 357, 130044.
4. S. N. Misal, M.-H. Lin, S. Mehraeen and B. P. Chaplin, J. Hazard. Mater., 2020, 384, 121420.
5. S. D. Richardson, M. J. Plewa, E. D. Wagner, R. Schoeny and D. M. DeMarini, Mutat. Res., Rev. Mutat. Res., 2007, 636, 178–242.
6. S. Gao, Z. Zhao, Y. Xu, J. Tian, H. Qi, W. Lin and F. Cui, J. Hazard. Mater., 2014, 274, 258–269.
7. F. J. Beltrán, A. Aguinaco, J. F. García-Araya and A. Oropesa, Water Res., 2008, 42, 3799–3808.
8. R. F. Dantas, S. Contreras, C. Sans and S. Esplugas, J. Hazard. Mater., 2008, 150, 790–794.
9. A. G. Gonçalves, J. J. M. Órfão and M. F. R. Pereira, J. Hazard. Mater., 2012, 239–240, 167–174.
10. A. G. Trovó, R. F. P. Nogueira, A. Agüera, C. Sirtori and A. R. Fernández-Alba, Chemosphere, 2009, 77, 1292–1298.
11. A. J. Watkinson, E. J. Murby and S. D. Costanzo, Water Res., 2007, 41, 4164–4176.
12. K. P. de Amorim, L. L. Romualdo and L. S. Andrade, Sep. Purif. Technol., 2013, 120, 319–327.
13. A. Fabiańska, A. Białk-Bielińska, P. Stepnowski, S. Stolte and E. M. Siedlecka, J. Hazard. Mater., 2014, 280, 579–587.
14. N. Oturan, S. O. Ganiyu, S. Raffy and M. A. Oturan, Appl. Catal., B, 2017, 217, 214–223.
15. R. Fu, P.-S. Zhang, Y.-X. Jiang, L. Sun and X.-H. Sun, Chemosphere, 2023, 311, 136993.
16. C. A. Martínez-Huitle and M. Panizza, Curr. Opin. Electrochem., 2018, 11, 62–71.
17. Y. Zhang, P. He, L. Zhou, F. Dong, D. Yang, H. Lei, L. Du, L. Jia and S. Zhou, Ecotoxicol. Environ. Saf., 2020, 188, 109921.
18. W. Wu, Z.-H. Huang and T.-T. Lim, J. Environ. Chem. Eng., 2016, 4, 2807–2815.
19. H. Olvera-Vargas, J.-C. Rouch, C. Coetsier, M. Cretin and C. Causserand, Sep. Purif. Technol., 2018, 203, 143–151.
20. S. O. Ganiyu, N. Oturan, S. Raffy, M. Cretin, C. Causserand and M. A. Oturan, Sep. Purif. Technol., 2019, 208, 142–152.
21. S. O. Ganiyu, N. Oturan, S. Raffy, G. Esposito, E. D. van Hullebusch, M. Cretin and M. A. Oturan, Electrochim. Acta, 2017, 242, 344–354.
22. T. Maqbool, Q. V. Ly, K. He, L. Cui, Y. Zhang, M. Sun and Z. Zhang, J. Membr. Sci., 2022, 651, 120460.
23. W. Yang, N. Oturan, S. Raffy, M. Zhou and M. A. Oturan, Chem. Eng. J., 2020, 383, 123155.
24. Y. Zhang, J. Ding, Q. Gao, B. Jiang, C. Li and Q. Zhao, Process Saf. Environ. Prot., 2022, 159, 931–943.
25. H. Lin, H. Peng, X. Feng, X. Li, J. Zhao, K. Yang, J. Liao, D. Cheng, X. Liu, S. Lv, J. Xu and Q. Huang, Water Res., 2021, 190, 116790.
26. S. O. Ganiyu, N. Oturan, S. Raffy, M. Cretin, R. Esmilaire, E. van Hullebusch, G. Esposito and M. A. Oturan, Water Res., 2016, 106, 171–182.
27. C. Trellu, M. Rivallin, S. Cerneaux, C. Coetsier, C. Causserand, M. A. Oturan and M. Cretin, Chem. Eng. J., 2020, 400, 125936.
28. A. F. Arif, R. Balgis, T. Ogi, F. Iskandar, A. Kinoshita, K. Nakamura and K. Okuyama, Sci. Rep., 2017, 7, 3646.
29. C.-H. Lee, Y.-K. Lim, E.-S. Lee, H.-J. Lee, H.-D. Park and D.-S. Lim, RSC Adv., 2018, 8, 11102–11108.
30. S. Liang, H. Lin, X. Yan and Q. Huang, Chem. Eng. J., 2018, 332, 628–636.
31. M. P. Casaletto, G. M. Ingo, S. Kaciulis, G. Mattogno, L. Pandolfi and G. Scavia, Appl. Surf. Sci., 2001, 172, 167–177.
32. F. Li, X. Peng, Y. Liu, J. Mei, L. Sun, C. Shen, C. Ma, M. Huang, Z. Wang and W. Sand, Chemosphere, 2019, 229, 383–391.
33. Y. Fan, Y. Ji, D. Kong, J. Lu and Q. Zhou, J. Hazard. Mater., 2015, 300, 39–47.
34. H. Lin, J. Niu, J. Xu, Y. Li and Y. Pan, Electrochim. Acta, 2013, 97, 167–174.
35. H. Yun, C. Lim, M. Kwon, D. Lee, Y. Yun, D.-H. Seo and K. Yong, Adv. Mater., 2024, 36, 2408280.
36. P. Saha, J. Wang, Y. Zhou, L. Carlucci, A. W. Jeremiasse, H. H. M. Rijnaarts and H. Bruning, Environ. Res., 2022, 211, 113057.
37. H. Hai, X. Xing, S. Li, S. Xia and J. Xia, Sci. Total Environ., 2020, 738, 139909.
38. C. Liang, C. Yi-Chi and Y.-L. Lin, Soil Sediment Contam.: Int. J., 2012, 21, 701–719.
39. J. Liu, S. Zhong, Y. Song, B. Wang and F. Zhang, J. Electroanal. Chem., 2018, 809, 74–79.
40. X. Chen, M. Murugananthan and Y. Zhang, Chem. Eng. J., 2016, 283, 1357–1365.
41. A. Rastogi, S. R. Al-Abed and D. D. Dionysiou, Appl. Catal., B, 2009, 85, 171–179.
42. C. Liang, C.-F. Huang and Y.-J. Chen, Water Res., 2008, 42, 4091–4100.
43. P. Britto-Costa and L. Ruotolo, Braz. J. Chem. Eng., 2012, 29, 763–773.
44. C. Wang, L. Yin, Z. Xu, J. Niu and L.-A. Hou, Chem. Eng. J., 2017, 326, 911–920.
45. C. Trellu, Y. Péchaud, N. Oturan, E. Mousset, D. Huguenot, E. D. van Hullebusch, G. Esposito and M. A. Oturan, Appl. Catal., B, 2016, 194, 32–41.
46. Y. Gao, Q. Yue, B. Gao and A. Li, Sci. Total Environ., 2020, 746, 141094.
47. J. Wang and R. Zhuan, Sci. Total Environ., 2020, 701, 135023.
48. Y. Yao, C. Huang, Y. Yang, M. Li and B. Ren, Chem. Eng. J., 2018, 350, 960–970.
49. G. V. Buxton, C. L. Greenstock, W. P. Helman and A. B. Ross, J. Phys. Chem. Ref. Data, 1988, 17, 513–886.
50. P. Xie, J. Ma, W. Liu, J. Zou, S. Yue, X. Li, M. R. Wiesner and J. Fang, Water Res., 2015, 69, 223–233.
51. S. Wang and J. Wang, Chem. Eng. J., 2018, 351, 688–696.
52. S. Wang, W. Xu, J. Wu, Q. Gong and P. Xie, Sep. Purif. Technol., 2020, 235, 116170.
53. S. Feng, J. Lin, S. Sun and F. Zhang, Chin. J. Oceanol. Limnol., 2017, 35, 153–162.
54. S. Zeb, S. Hussain, H. A. Khan, Z. Ali, N. Khan, K. I. Khan, F. Ali, S. khan, M. del Pilar Taboada Sotomayor and S. Gul, Int. J. Electrochem. Sci., 2018, 13, 9428–9440.
55. X. Ao, W. Liu, W. Sun, C. Yang, Z. Lu and C. Li, Chemosphere, 2018, 212, 365–375.
56. K. S. Tay and N. Madehi, Sci. Total Environ., 2015, 520, 23–31.

---