1,4-dioxane degradation using a pulsed switching peroxi-coagulation process

As an important solvent stabilizer, reaction agent, and reaction media, 1,4-dioxane (C₄H₈O₂) has been widely used in the manufacturing processes of chemical products such as paints, varnishes, lacquers, cosmetics, resins, and deodorants (Clercq et al. 2010; Barndõk et al. 2016a). Being possibly carcinogenic to humans and chemically stable, 1,4-dioxane can pose a great risk to human health and the ecosystem (Clercq et al. 2010; Takeuchi & Tanaka 2020; Zhao et al. 2020). Since 1,4-dioxane is bio-refractory, indicated by the low ratio of biochemical oxygen demand (BOD) to chemical oxygen demand (COD) (i.e. 0.06), non-volatility, and its miscibility with water (Nakagawa et al. 2016; Radcliffe & Page 2020; Rossum 2020; Somda et al. 2020), it is difficult to effectively remove it in typical biological wastewater treatments (Mahendra et al. 2013; Huang et al. 2018; Xu et al. 2020). For example, complete decomposition of 100 mg/L 1,4-dioxane in the activated sludge processes required 7 days (Sei et al. 2010). Even with pure culture, more than 25 h was needed for the 1,4-dioxane removal (Sun et al. 2010; Sei et al. 2013). The biological co-metabolism can be used to enhance the 1,4-dioxane degradation. However, additional nutrients, such as tetrahydrofuran and lactate (Sekar & DiChristina 2014), may increase the treatment cost (Hand et al. 2015; Zhang et al. 2016; Chen et al. 2020; Fan et al. 2020; Lin et al. 2020). Thus, it is necessary to develop an efficient method for 1,4-dioxane degradation.

The advanced oxidation processes (AOPs) have been developed for 1,4-dioxane removal in recent years. Compared with the biological treatment process, AOPs can produce high reactive and non-selective oxidant (i.e. ·OH) (Elkacmi & Bennajah 2019), and thus are a fast and efficient alternative to 1,4-dioxane degradation (Clercq et al. 2010; Barndõk et al. 2016a). Because of its good performance, easy control, and environmental friendliness (Wang et al. 2016), the Fenton process is one of the powerful AOPs for 1,4-dioxane removal. The Fenton process could reduce 1,4-dioxane from 100 to 0.5 mg/L within 6 h (Nakagawa et al. 2016). However, the high risk of H₂O₂ transport limits application of the Fenton process (Brillas & Martínez-Huitle 2015; Gao et al. 2015; Wang et al. 2016). The risks of H₂O₂ transport can be avoided by in-situ H₂O₂ production using the electro-Fenton (EF) process (Brillas & Martínez-Huitle 2015). As a combined EF process, the peroxi-coagulation (PC) process utilizes a sacrificial iron anode and a gas diffusion cathode (GDC) for H₂O₂ generation (Brillas et al. 2009). Fe²⁺ ions can release from the sacrificial iron anode to catalyze H₂O₂ for hydroxyl radicals (·OH) production. Excess Fe²⁺ dissolution from the iron anode can form Fe(OH)₃ precipitation as a coagulant. However, excess Fe²⁺ may also result in severe OH scavenger and high iron sludge production (Sun & Pignatello 1993; Benhadji et al. 2016).

Recently, a pulsed switching peroxi-coagulation (PSPC) process was developed for effectively controlling Fe²⁺ consumption (Lu et al. 2018). The controllable Fe²⁺ release in the auxiliary anode and H₂O₂ electrosynthesis in the cathode of PSPC was useful for maximum ·OH production and minimum iron sludge (Lu et al. 2018). Under current densities of 5.0 mA/cm² (H₂O₂) and 0.5 mA/cm² (Fe²⁺), and the pulsed switching frequency of 1.0 s (H₂O₂): 0.3 s (Fe²⁺), 500 mg/L 2,4-dichlorophenoxyacetic acid was completely removed in the PSPC within 240 min (Lu et al. 2018). The energy consumption for 500 mg/L 2,4-dichlorophenoxyacetic acid removal in the PSPC was only 50% of that in the conventional EF process (68 ± 6 vs. 136 ± 10 kWh/kg TOC) (Lu et al. 2018). The iron consumption in the PSPC was only ∼5% of that in the conventional PC process. Although PSPC showed great potential in the 2,4-dichlorophenoxyacetic acid removal, the application of PSPC is still in its infancy and needs further testing for other refractory organics. Moreover, the physicochemical characteristics of different refractory organics can affect its removal in the AOPs (Clercq et al. 2010; Lu et al. 2018). For example, 1,4-dioxane has higher solubility and toxicity than 2,4-dichlorophenoxyacetic acid (Clercq et al. 2010; Lu et al. 2018). The optimal operational conditions for 1,4-dioxane removal in the PSPC may be different from those for 2,4-dichlorophenoxyacetic acid removal. The experiments on 1,4-dioxane removal in the PSPC should be not only important for enhancing 1,4-dioxane degradation in the wastewater treatment but also for expanding the potential application of PSPC. Thus, the objective of this study was to investigate the feasibility of the PSPC process for 1,4-dioxane degradation. The effect of H₂O₂ and Fe²⁺ production, pulsed switching frequency and current density were tested on the 1,4-dioxane removal. Five control experiments were carried out to distinguish the PSPC process from the EF and PC processes. The main intermediates of 1,4-dioxane degradation and the residual Fe²⁺ and H₂O₂ were identified to discuss the mechanism of efficient 1,4-dioxane degradation in the PSPC process.

The PSPC process was investigated in an undivided electrochemical cell with a cylindrical chamber (diameter × length = 3 × 4.5 cm). A platinum wire electrode (CHI115, CH Instrument, Inc., Shanghai, China) and an iron mesh (diameter of 3 cm, thickness of 0.3 mm, 90 meshes) were used as main and auxiliary anodes, respectively. A gas diffusion cathode (GDC) was constructed with a conductive gas diffusion layer, a catalyst layer, and a supporting layer of stainless steel mesh (90 meshes) (Wang et al. 2017). Carbon black powder (EC-300 J, Hesen, Shanghai, China) was used as a catalyst for H₂O₂ production. The effective surface area of the GDC was 7 cm². The distance between the GDC and Pt anode was 3.5 cm. The distance between the GDC and Pt anode was 2 cm. In-situ H₂O₂ and Fe²⁺ production was driven by the power supply (IT6700, ITECH Electronic Co., Ltd, Nanjing, China) with constant current. Two time relays with an accuracy of 0.1 s (ZYS48-S, Zhuoyi Electronic Co., Ltd, Shanghai, China) were used to control the pulsed switching circuits and the running time of H₂O₂ and Fe²⁺ productions in PSPC. A solution of 0.1 M Na₂SO₄ and 200 mg/L 1,4-dioxane (99%, Merck) was used as the electrolyte. The pH of the electrolyte was adjusted to 5.0 with H₂SO₄ or NaOH. The electrolyte was recycled in the electrochemical cell using a peristaltic pump (BT-100, Qite, China) with a flow rate of 36 mL/min.

The current density for H₂O₂ was set at 5.0 mA/cm² to produce H₂O₂ efficiently with low energy consumption based on the results of our preliminary tests. In order to avoid excess iron oxidization into Fe³⁺, the current density for Fe²⁺ production was controlled at 1 mA/cm². The molar concentration ratio of Fe²⁺ to total Fe in the solution was >95% under the iron anode operation with 1 mA/cm². Thus, the pulsed switching ratio of H₂O₂ and Fe²⁺ productions were tested under the current densities of 5 mA/cm² for H₂O₂ and 1 mA/cm² for Fe²⁺, including 4 s (H₂O₂): 4 s (Fe²⁺) (i.e. 1.00), 4:3 s (i.e. 1.33), 4:2 s (i.e. 2.00), 4:1 s (i.e. 4.00), and 4:0.5 s (i.e. 8.00). The pulsed switching frequency, including 4:3 s (0.14 Hz), 2:1.5 s (0.29 Hz), 1.2:0.9 s (0.48 Hz), 0.8:0.6 s (0.71 Hz), and 0.4:0.3 s (1.43 Hz), was tested under the fixed pulsed switching ratio between H₂O₂ and Fe²⁺ productions of 1.33 (4/3 s), respectively. Under the pulsed switching frequency of 1.43 Hz, the 1,4-dioxane removal was tested under different current densities, including 15 mA/cm² (H₂O₂) +3 mA/cm² (Fe²⁺), 10 mA/cm² (H₂O₂) +2 mA/cm² (Fe²⁺), 5 mA/cm² (H₂O₂) +1 mA/cm² (Fe²⁺), and 2.0 mA/cm² (H₂O₂) +0.4 mA/cm² (Fe²⁺), respectively.

Five controls were used to distinguish the PSPC process from others, such as EF and PC processes (Brillas et al. 2009; Ahangarnokolaei et al. 2017). Control 0 was to test the 1,4-dioxane adsorption in the cell without H₂O₂ and Fe²⁺ production (Table 1). Control 1 was a conventional PC process to keep a continuous production of H₂O₂ and Fe²⁺ without the pulsed switching frequency in the cell under the current densities of 5 mA/cm² (H₂O₂) and 1 mA/cm² (Fe²⁺). Control 2 was a conventional EF process to keep a continuous H₂O₂ production of 5 mA/cm², and add 7.9 mM Fe²⁺ into the cell without iron anode operation. Control 3 was to keep a pulsed H₂O₂ production at 5 mA/cm² with the pulsed switching frequency of 0.4:0.3 s (1.43 Hz), and 7.9 mM Fe²⁺ added into the cell without iron anode operation. Control 4 was to keep a continuous H₂O₂ production of 5 mA/cm², and add 7.9 mM Fe²⁺ into the cell without iron anode operation. The addition of 7.9 mM Fe²⁺ was equal to the total Fe²⁺ production in the PSPC process under the pulsed switching frequency ratio of 0.4:0.3 s, and current densities of 5 mA/cm² (H₂O₂) and 1 mA/cm² (Fe²⁺) within 120 min.

The H₂O₂ concentration was measured using a spectrophotometer (T6, Persee, Beijing, China) according to the titanium (IV) sulfate method (Barazesh et al. 2015). Fe²⁺ and total iron concentrations were determined with the phenanthroline spectrophotometric method (Xu et al. 2013). The total organic carbon (TOC) was measured using a Shimadzu TOC-L CPH analyzer (Shimadzu Co., Japan). The pH was measured by a pH meter (FE20, Mettler-Toledo, Switzerland). The total COD (TCOD) and soluble COD (SCOD) was determined using the dichromate standard method according to the samples filtrated without or with a 0.22 μm filter, respectively (Liu et al. 2015; Ye et al. 2017). A high performance liquid chromatograph (HPLC, P230II, Dalian Yilite Analytic Instrument Co. Ltd, China) was used to determine 1,4-dioxane concentration. A 10:90 (v/v) acetonitrile/water (phosphate buffer of pH 3) solution at 1.0 mL/min was used as the mobile phase. The chromatograph was equipped with a SinoChrom ODS-BP column (5 μm, 4.6 mm × 25 cm, Dalian Yilite Analytic Instrument Co. Ltd, China). The UV detector was set at 190 nm and the temperature was maintained at 30°. The intermediates of 1,4-dioxane degradation, including oxalic, acetic and formic acids, were quantified by an ion chromatograph (IC, CIC-D100, SHINE IC Solution Experts, China).

In the PSPC process, the molar ratio of H₂O₂ and Fe²⁺ concentrations were determined by the pulsed switching ratio of H₂O₂ and Fe²⁺ production. For example, when the pulsed switching ratio of H₂O₂ and Fe²⁺ production increased from 1.00 to 8.00, the molar ratio of H₂O₂ and Fe²⁺ increased from 1.49 to 12.0. As shown in Figure 1(a), more than 80% of 1,4-dioxane was removed under different pulsed switching ratios of H₂O₂ and Fe²⁺ production within 120 min. The 1,4-dioxane removal slightly increased from 80.1 ± 1.8 to 89.6 ± 1.1% with the pulsed switching ratio increasing from 1.00 to 8.00. The low pulsed switching ratio resulted in low energy consumption (Figure 1(b)). The energy consumption under the pulsed switching ratio of 1.33 was 25.6 kWh/kg, which was only 76% of that under the pulsed switching ratio of 8.00. The energy consumption was almost kept stable at the pulsed switching ratio of 1.33 and 1.00. Therefore, the pulsed switching ratio of 1.33 could be suitable for 1,4-dioxane removal in the PSPC process. Correspondingly, the molar ratio of H₂O₂ and Fe²⁺ was 1.99, and the 1,4-dioxane removal efficiency was 84 ± 2%.

Under the fixed pulsed switching ratio of 1.33, different pulsed switching frequency, including 0.14, 0.29, 0.48, 0.71, and 1.43 Hz, was tested as shown in Figure 2. High pulsed switching frequency resulted in high 1,4-dioxane removal and low energy consumption. The 1,4-dioxane removal gradually increased from 84.8 ± 0.4 to 95.3 ± 2.2% with the pulsed switching frequency increasing from 0.14 to 1.43 Hz within 120 min. Based on Equation (1), the energy input for H₂O₂ and Fe²⁺ production was determined by the pulsed switching ratio of H₂O₂ and Fe²⁺ production. Under the pulsed switching ratio of 1.33, the energy input of the PSPC process was about 0.0718 W with 80% 1,4-dioxane removal, 96% of which was consumed by H₂O₂ generation. With the pulsed switching frequency increasing from 0.14 to 1.43 Hz, the energy consumption of the PSPC process with 80% 1,4-dioxane removal was decreased from 23 to 14 kWh/kg (Figure 2(b)). Therefore, the pulsed switching frequency of H₂O₂ and Fe²⁺ production was optimized to 0.4:0.3 s in the PSPC process. Correspondingly, 95.3 ± 2.2% of 1,4-dioxane was removed with the energy consumption of 22.1 ± 0.5 kWh/kg within 120 min.

Under the optimally pulsed switching frequency of 1.43 Hz, the effect of current densities in the PSPC process was tested as shown in Table 2. To keep the optimal molar ratio of H₂O₂ and Fe²⁺ production as described above, the current densities of H₂O₂ and Fe²⁺ production were changed proportionally, including 15 mA/cm² (H₂O₂) +3 mA/cm² (Fe²⁺), 10 mA/cm² (H₂O₂) +2 mA/cm² (Fe²⁺), 5 mA/cm² (H₂O₂) +1 mA/cm² (Fe²⁺), and 2.0 mA/cm² (H₂O₂) +0.4 mA/cm² (Fe²⁺). High current density of H₂O₂ and Fe²⁺ resulted in high 1,4-dioxane removal under the optimally pulsed switching frequency in the PSPC process (Table 2). However, even with low current densities such as 2.0 mA/cm² (H₂O₂) and 0.4 mA/cm² (Fe²⁺), 89.8 ± 2.2% of 1,4-dioxane was still removed within 120 min.

In terms of 1,4-dioxane removal within 120 min, the different processes were in the order as follows: PSPC ≈ Control 1 > Control 2 > Control 3≫Control 4 ≈ Control 0. Control 0 showed that the adsorption was not significant for 1,4-dioxane removal in the PSPC process. Without Fe²⁺ catalyst addition, 1,4-dioxane could be hardly removed by H₂O₂ oxidation (see Control 4). Control 1 (i.e. the conventional PC process) removed 1,4-dioxane efficiently but consumed ∼18.2 mM Fe²⁺ from the iron electrode within 120 min. Control 2 (i.e. the EF process) removed less 1,4-dioxane than PC and PSPC processes. Control 3 showed that only pulsed H₂O₂ production with 7.9 mM Fe²⁺ addition did not remove 1,4-dioxane as efficiently as that in the PSPC process within 120 min (71.7 ± 0.4% vs. 95.3 ± 2.2%).

High current densities of H₂O₂ and Fe²⁺ production resulted in high energy consumption in the PSPC process (Table 3). The energy consumption under 15 mA/cm² (H₂O₂) +3 mA/cm² (Fe²⁺) was 15 times higher than that under 2.0 mA/cm² (H₂O₂) +0.4 mA/cm² (Fe²⁺) within 120 min (121.0 ± 0.6 vs. 7.8 ± 0.1 kWh/kg). Nevertheless, the minimum energy consumption in the PSPC process under 2.0 mA/cm² (H₂O₂) +0.4 mA/cm² (Fe²⁺) was less than 20% of that in Control 1 and 2 within 120 min, respectively (Table 3).

Under the pulsed switching frequency of 1.43 Hz and current densities of 5 mA/cm² (H₂O₂) and 1 mA/cm² (Fe²⁺), the residual Fe²⁺ and H₂O₂ concentrations during 1,4-dioxane degradation was determined in the PSPC process within 120 min (Figure 3). The residual Fe²⁺ concentration was kept as low as 0.6 ± 0.5 mg/L within 120 min. The H₂O₂ concentration reached 100 mg/L and remained stable at 100 mg/L within 120 min. The residual Fe²⁺ and H₂O₂ concentrations in the PSPC process were different from that in Controls 2 and 3. With initial 7.9 mM Fe²⁺ addition, the residual Fe²⁺ concentration was decreased to <0.5 mg/L within 40 min in Controls 1 and 2. However, the residual H₂O₂ concentration was gradually increased to 280 and 250 mg/L in Controls 1 and 2, respectively. High residual Fe²⁺ concentration within 40 min and high residual H₂O₂ concentration within 40–120 min in Controls 1 and 2 may account for the lower 1,4-dioxane removal compared with that in the PSPC process. The residually soluble Fe³⁺ was not detected in all the tests, indicating that Fe(OH)₃ precipitated in the cell (Brillas et al. 2009).

Under the pulsed switching frequency of 1.43 Hz and current densities of 5 mA/cm² (H₂O₂) and 1 mA/cm² (Fe²⁺), TCOD and SCOD in the cell were measured within 120 min as shown in Figure 4. The TCOD and SCOD concentrations decreased from 356 ± 28 to 119 ± 27 mg/L and from 356 ± 24 to 99 ± 10 mg/L within 120 min, respectively. Almost the same removals of TCOD and SCOD (67 vs. 72%) indicated that hydroxyl radical oxidation played the key role in the removal of 1,4-dioxane and its intermediates. Only 26.0% of TOC was removed in the PSPC system within 120 min, indicating that 1,4-dioxane was incompletely mineralized. High concentrations of the residual intermediates (e.g. small molecular organic acids) should account for low TOC removal in the 1,4-dioxane degradation in the PSPC process according to previous reports (Barndõk et al. 2016a). Acetic acid, formic acid, and oxalic acid were identified in the PSPC process (Figure 5). Under the pulsed switching frequency of 1.43 Hz and current densities of 5 mA/cm² (H₂O₂) and 1 mA/cm² (Fe²⁺), the maximum formic acid concentration of 51.2 ± 0.3 mg/L was determined within 80 min. The concentrations of acetic acid, formic acid and oxalic acid were 10.0 ± 4.6, 31.1 ± 9.1, and 2.7 ± 0.8 mg/L at 120 min, respectively.

The excess of Fe²⁺ or H₂O₂ may give rise to the side reactions (4), (5), (6), and (7). As previously reported, the ·OH attacks the organics to form hydroxylated derivatives with a reaction rate (k₂) of about 10⁷–10⁹ M⁻¹ s⁻¹ (Brillas & Martínez-Huitle 2015; Gao et al. 2015), which is comparable to that in the side reactions (4), (5), (6), and (7). The reaction rate of 63 M⁻¹·s⁻¹ in the ·OH production (Equation (3)) was much lower than those in Equations (4)–(7). Excess of Fe²⁺ or H₂O₂ can act as powerful competitors to the organics such as 1,4-dioxane to react with the ·OH (Sun & Pignatello 1993; Muruganandham & Swaminathan 2004). Therefore, excess of Fe²⁺ or H₂O₂ can result in low 1,4-dioxane removal and high energy consumption in the EF and PC processes. However, with the pulsed switching circuits, the excess of Fe²⁺ and H₂O₂ production was minimized in the PSPC process (Figure 3, Tables 1 and 2). The low residual Fe²⁺ concentration in the PSPC process within 120 min could be attributed to the high reduction activity of Fe²⁺ released from the iron sacrifice electrode, which can be rapidly consumed as described in Equation (3). However, low residual Fe²⁺ concentration in Controls 2 and 3 within 40–120 min indicated that not enough Fe²⁺ catalyzed H₂O₂ to produce ·OH. The Fe²⁺ in Controls 2 and 3 may be regenerated via the side reaction such as Equation (8) with low reaction rate, inhibiting efficient ·OH production. Therefore, the PSPC could have higher 1,4-dioxane removal with lower energy consumption compared with other processes.

The optimal molar ratio of H₂O₂ to Fe²⁺ (summarized as [H₂O₂]/[Fe²⁺]) was greatly changed depending on the organics, solutions, etc. The optimal [H₂O₂]/[Fe²⁺] was in the range of 5–11 for the chlorinated aliphatic organic removal in the Fenton process (Tang & Huang 1997). [H₂O₂]/[Fe²⁺] = 1000:1 had been reported to produce the highest ·OH concentration in the Fenton process among different molar ratios (Fischbacher et al. 2017). The theoretically optimal [H₂O₂]/[Fe²⁺] was 2.0 based on the pulsed switching ratio of 1.33 in this study, which was different from [H₂O₂]/[Fe²⁺] = 6 for the 2,4-dichlorophenoxyacetic acid degradation in the PSPC process (Lu et al. 2018). It indicated that the pulsed switching ratio of H₂O₂ to Fe²⁺ should be further investigated to optimize [H₂O₂]/[Fe²⁺] during the PSPC application into the real wastewater treatment.

With a continuous consumption of H₂O₂ and Fe²⁺ as shown in Equation (3), lower H₂O₂ and Fe²⁺ concentrations were accumulated under higher pulsed switching frequency in the PSPC process. Similar results had been reported in the Fenton process with the stepwise addition of H₂O₂ to avoid excess H₂O₂ accumulation and decrease H₂O₂ scavenging ·OH (Zhang et al. 2012; Gan & Li 2013). With the pulsed switching frequency increasing from 0.71 to 1.43 Hz, the 1,4-dioxane removal was not significantly improved (93.1 ± 2.6 vs. 95.4 ± 2.2%), indicating the limited effect of high pulsed switching frequency on the 1,4-dioxane removal. In addition, higher pulsed switching frequency will require higher precision of time relays and higher investment. Although the highest pulsed switching frequency of 1.43 Hz had the highest 1,4-dioxane removal with the lowest energy consumption in our PSPC process, the optimal pulsed switching frequency still needs further study in the PSPC process in future.

Because the residual H₂O₂ concentration was much higher than that of the residual Fe²⁺ concentration in the PSPC process, the ·OH production and 1,4-dioxane removal should occur close to the sacrifice iron electrode. However, with 7.9 mM Fe²⁺ addition in Control 2 (i.e. the EF process), the ·OH production and 1,4-dioxane removal could occur close to the cathode. Compared with the PSPC, the EF process may suffer from Fe(OH)₃ precipitation on the cathode surface, which can decrease the reaction area of H₂O₂ electro-generation in the cathode. Because of high solubility of 1,4-dioxane, the Fe(OH)₃ coagulation had no apparent effect on the 1,4-dioxane removal. The 1,4-dioxane degradation in the PSPC process was proposed as shown in Figure 6, based on the quantified intermediates. Various short chain organic acids have been identified in the intermediates of 1,4-dioxane degradation in AOPs (Barndõk et al. 2016a). In the UV photo-Fenton process, the intermediates such as acetic, oxalic, methoxyacetic and glycolic acids were identified during 7.3 g/L 1,4-dioxane degradation (Barndõk et al. 2016a). The maximum acetic acid concentration reached ∼200 mg/L within 120 min (Barndõk et al. 2016a). In the solar photocatalysis using an NF-TiO₂ composite with TiO₂ nanoparticles, 140 mg/L 1,4-dioxane degradation resulted in the maximum formic acid of ∼60 mg/L within 7 h (Barndõk et al. 2016b). Different AOPs including the PSPC showed that complete mineralization of 1,4-dioxane was greatly postponed due to high concentrations of short chain organic acids as intermediates (Barndõk et al. 2016a, 2016b). The short chain organic acids can be easily degraded by microbiology (Huang et al. 2017; Wu et al. 2020), thus the PSPC combined with a biological treatment process may be a potential way to mineralize 1,4-dioxane efficiently in future.

The pulsed switching peroxi-coagulation (PSPC) process was used to remove 1,4-dioxane in this study. The pulsed switching ratio of H₂O₂ to Fe²⁺ production was optimized at 1.33. Under the optimized pulsed switching frequency of 1.43 Hz, the maximum 1,4-dioxane removal reached 98.8 ± 0.5% with the current densities of 15 mA/cm² (H₂O₂) +3 mA/cm² (Fe²⁺) within 120 min. The minimum energy consumption reached 7.8 ± 0.1 kWh/kg with the current densities of 2.0 mA/cm² (H₂O₂) +0.4 mA/cm² (Fe²⁺). At the end of 120 min operation, the residual Fe²⁺ concentration was kept as low as 0.6 ± 0.5 mg/L. Under the current densities of 5.0 mA/cm² (H₂O₂) +1 mA/cm² (Fe²⁺), the residual H₂O₂ concentration reached 100 mg/L and remained stable at 100 mg/L. Low residual H₂O₂ and Fe²⁺ concentrations were attributed to high pulsed switching frequency in the PSPC process, resulting in effectively inhibiting the side reaction during the ·OH production and improving the 1,4-dioxane removal with low energy consumption. The quantified intermediates showed that the formic acid reached the maximum concentration of 51.2 ± 0.3 mg/L within 80 min. High concentrations of intermediates significantly hindered the mineralization of 1,4-dioxane.

This work was partly supported by grants from the Science and Technology Program of Guangzhou, China (No. 201804010450), the National Natural Science Foundation of China (Nos. 51608547, 51978676 and 51308557), the Guangdong Provincial Key Laboratory Project (2019B121203011), and the Fundamental Research Funds for the Central Universities (No. 19lgjc08).

All relevant data are included in the paper or its Supplementary Information.