An iron plate (Fe, >99%) of 50 × 50 mm was purchased from Tengfeng Metallic Material Co. Ltd. (Hebei, China). p-nitrophenol (PNP, C₆H₅NO₃, 99%) was supplied from Shanghai Yien Chemical Technology Co. Ltd. 1,10-phenanthroline (C₁₂H₈N₂·H₂O, 99%), glacial acetic acid (C₂H₄O₂, 99%), sodium acetate (CH₃COONa, 99%), ferrous sulfate (FeSO₄.7H₂O, 99%), hydroxylamine hydrochloride (HONH₃Cl, 98.5%), sodium sulfate (Na₂SO₄, 99%), hydrogen peroxide (H₂O₂, 30%), titanium oxalate, potassium (K₂TiO(C₂O₄)₂, 98%), ferrous ammonium sulfate hexahydrate (Fe(NH₄)₂·(SO₄)₂·6H₂O, 99%), potassium permanganate (KMnO₄, 99%), sulfuric acid (H₂SO₄, 98%), and sodium hydroxide (NaOH, 98%) were provided by Sinopharm Chemical Reagent Co. Ltd., China. All chemicals were analytical grade and were used without further purification. Ultrapure water (18.2 MΩ cm) was used to prepare reaction solutions for all the experiments.

As shown in Figure 1, the combined electrochemical treatment process mainly consisted of four steps. Prior to step 1, a 100 mg/L PNP solution with pH 7.0 was performed in the beaker, 3 g/L of Na₂SO₄ was used as the electrolyte solution. In step 1, pH adjustment was carried out in the divided electrolysis cell, which used a plexiglass rectangular tank with the cathode and anode chambers isolated by a PTFE microfiltration membrane. The anode consisted of a 50 × 80 mm boron-doped diamond (BDD) electrode while the cathode consisted of titanium mesh of the same dimensions with the distance of 25 mm; the PTFE membrane was between the anode and cathode. The galvanostatic electrolysis reactions were performed by a DC power supply (voltage 0–30.0 V, electric current 0–5.0 A). In step 1, the PNP solution concurrently entered into the cathode and anode chambers, pH was reduced to near 3.0 in the anodic chamber within 1 min. In step 2, the solution with the reduced pH was added into the iron cathode electrochemical Fenton (ICEF) system for PNP degradation with a chamber volume of 500 ml using a three-electrode potentiostat (Model CHI1130C, Chenhua instrument Co. Ltd. Shanghai, China) coupled with a saturated calomel reference electrode (SCE), where the cathodic potential was precisely controlled by the potentiostat and SCE. The working electrode was a 50 × 50 mm iron plate, and the counter electrode was a 50 × 50 mm DSA (Ti-RuO₂-IrO₂) mesh; 200 mg/L of H₂O₂ was continuously added by a pump within 15 min. In step 3, after the treatment of PNP solution in the ICEF system, the solution was transferred into the cathode chamber of the divided electrolysis cell to increase the pH. In addition, the pH of the cathode effluent was further regulated to neutral (∼9.0) by NaOH solution. In step 4, the neutral solution was passed through a filter device to remove iron ions, which was carried out in a container filled with quartz sand (80–120 mm). The removal ratio of p-nitrophenol, COD, or TOC ( η , % ) was calculated via Eq. 8. η =   C 0 − C t C 0   × 100 % (6)

The concentrations of PNP were determined by high performance liquid chromatography (HPLC-1100, Agilent) with an Eclipse XDB-C18 column (4.6 mm × 150 mm, 5 µm), (Nakatsuji et al., 2015). The mobile phase was methanol/water (50/50), the flow rate was 1.0 ml/min, and the UV detector was set at 314 nm. The chemical oxygen demand (COD) was detected using a COD analyzer (ASH 6B-80, China) [37]. A total organic carbon analyzer (multi N/C® 2,100, Analytik Jena AG) was applied to monitor total organic carbon (TOC). (Kavitha and Palanivelu, 2004). The concentration of H₂O₂ was measured by the potassium titanium (IV) oxalate method at 400 nm with a UV-Vis spectrophotometer (UV 6000 PC Yuanxi instrument Co. Ltd. Shanghai, China) (Wang and Chu, 2011). Fe²⁺ and Fe (tot) were determined at 510 nm using a modified phenanthroline method with an Fe²⁺ detection limit of 0.5 μM, and Fe³⁺ concentration was estimated as the difference between Fe (tot) and Fe²⁺ (Xin et al., 2018). Solution pH and conductivity were measured by a water quality analyzer. The reaction was quenched in the collected samples by immediately adding 1.0 mol/L of NaOH since Fenton oxidation cannot occur at pH > 10.0. For COD measurements, the samples were pretreated with NaOH to remove any residual H₂O₂. All experiments were performed twice at least, with relative errors less than 3%.

For the Fenton reaction, it is recommended to acidify the solution pH to ∼3.0 for •OH production and then neutralize the solution for the precipitation of Fe³⁺ (Chen et al., 2021). In this study, we developed a divided electrolysis cell using a hydrophilic PTFE microfilter membrane as the separated material. In this system, H₂O splitting reactions at the anode and cathode could produce H⁺ and OH⁻ (Eq. 7 and Eq. 8), respectively, which could be used to regulate the solution pH required for the Fenton reaction. Thus, the substrate solution pH could reduce to nearly 3.0 when it flowed through the anode chamber in step 1 (Figure 1). After the Fenton reaction, the solution pH was raised to promote the formation of Fe(OH)₃ in step 3. The above pH adjustment objective could be achieved by regulating the applied current density and flow velocity. Here, a neutral solution containing 100 mg/L of PNP and 3 g/L of Na₂SO₄ electrolyte was pumped into the electrolysis device at a flow rate of 20 ml/min by a peristaltic pump, where the residence time was 1 min. The current density applied is the key parameter for the electrochemical pH regulation because it fundamentally affects the yield of H⁺ in the anode compartment and OH⁻ in the cathode compartment, respectively. As shown in Figure 2A, the pH of the anode effluent continuously decreased with the increase of current density; as the current density increased from 0.5 to 6 mA/cm², the pH of anodic effluent decreased from 4.2 to 2.7, while the cathodic effluent pH increased from 9.9 to 11.5. This was consistent with what we expected, because as the current increased, effective water-splitting resulted in more H⁺ produced at the anode and more OH⁻ produced at the cathode. In particular, when the applied current density was 2 mA/cm², an acidic effluent with pH of 3–3.2 was automatically attained at the steady state (Neyens and Baeyens, 2003). It has been demonstrated that the Fenton process is inhibited at extremely acidic environments due to the formation of (Fe(H₂O))²⁺ and Fe(III)-hydroxyl complexes, Fe(OH)²⁺. Thus, to achieve the most favorable Fenton reaction condition, an applied current density of 2 mA/cm² was adopted in this study. 2 H 2 O  +   2 e - → 2 OH - + H 2 ↑ (7) 2 H 2 O  -   4 e - → 4 H + +  O 2 ↑ (8)

The results in Figure 2B show that the PNP degradation efficiency increased as the current density increased initially from 0 to 1 mA/cm², but it became nearly constant when the current density was elevated above 1 mA/cm². Specifically, as the current density increased from 0.5 to 1 mA/cm², the PNP removal efficiency increased from 5 to 10%. Nevertheless, further increasing the current density to 6 mA/cm² did not increase the further removal of PNP. In this case, the BDD anode had little effect on the degradation of pollutants in this system because of a too short residence time.

As shown in Figure 3, when the solution pH reduced to nearly 3.0, the iron plate immersed in water without power could produce 20 mg/L of Fe (tot) within 30 min, of which 89% was Fe(II). In this process, the iron plate would be dissolved by H⁺ to generate Fe²⁺ under acidic conditions through Eq. 9, and the oxidation of Fe²⁺ to Fe³⁺ by O₂ was slow and the solution pH increased slightly from 3.0 to 3.2. The average production rate of Fe (tot) (approximately 0.33 mg/min) and Fe²⁺ (approximately 0.3 mg/min) was stable and closed within 30 min, indicating the stable release of Fe²⁺ by the iron plate under the acid solution. Under the same conditions, Fe (tot) significantly increased to 52 mg/L after 200 mg/L of H₂O₂ was continuously added into the solution within 15 min. The average production rate of Fe (tot) was stable within 20 min (approximately 1.1 mg/min), which was three times higher than that without the addition of H₂O₂. But these two reactions exhibited a similar average production rate of Fe (tot), approximately 0.41 mg/min, within the reaction period of 20–30 min, which may be ascribed to the fact that most of the H₂O₂ was consumed within 20 min. As for Fe²⁺ concentration, it continuously increased within 10 min with the production rate of approximately 1.13 mg/min and decreased from 23 to 11 mg/L between 10 and 20 min, which was explained by the consumption of Fe²⁺ during the Fenton reaction process (Eq. 1). And after 20 min, the concentration continuously increased, finally reaching 18 mg/L at 30 min, accounting for 35% of the Fe (tot). Within 20–30 min, the Fe (tot) concentration increased by 8 mg/L, close to the value of Fe²⁺ increase (7 mg/L), this can also be explained by the fact that little H₂O₂ remained after 20 min and the increase of Fe (tot) concentration was mostly due to the Fe²⁺ released (Eq. 9). The above results revealed that the presence of H₂O₂ could facilitate the release of iron ions from Fe⁰ in the acid solution. In this case, a large amount of iron sludge was produced and side reactions (Eq. 3) might consume produced •OH, leading to inferior degradation of pollutants. Fe + 2 H + → Fe 2 + + H 2 (9)

To further reduce the release of iron ions from the iron cathode and improve the regeneration of Fe²⁺ from Fe³⁺, we developed a novel iron cathode electro-Fenton system, applying a potential on the iron cathode. Figure 4A shows that concentration variations of Fe (tot) under different cathodic potentials, it was found that as the applied cathodic potential increased, the produced Fe (tot) decreased continuously. At the cathodic potential of −1.0 and −1.2 V/SCE, the produced Fe (tot) concentration was 30 and 21 mg/L within 30 min, which reduced by 42 and 61%, respectively, compared with no electricity input. Further increasing the cathodic potential from −1.2 to −1.4 V/SCE resulted in the slight decrease of Fe (tot) concentration to 18 mg/L. The production rate of Fe (tot) (0.43 mg/min) at the cathodic potential of −1.2 V/SCE was significantly higher than −1.4 V/SCE (0.29 mg/min) within the initial 20 min, while the value was apparently decreased to 0.17 mg/min within 20–30 min, and the rate slightly increased to 0.33 mg/min during the same time at −1.4V/SCE. Figure 4B shows the Fe²⁺ and Fe (tot) concentrations produced at 30 min under different cathodic potentials. It was found that the ratio of Fe²⁺ to Fe (tot) concentration enhanced significantly from 0.39 to 0.77 with the decrease of cathodic potential from −1.0 to −1.4V/SCE. This can be explained by the fact that the regeneration from Fe³⁺ to Fe²⁺ increased with the decrease of cathodic potential.

The results of PNP concentration versus time are shown in Figure 5A. It was found that only 75% of PNP was degraded within 30 min under the conditions of the iron plate immersed in water with 200 mg/L of H₂O₂ without electricity (WE Fenton process). And PNP removal efficiency was enhanced to 87, 91, and 90% at cathodic potentials of −1.0, −1.2, and −1.4 V/SCE, respectively (Kavitha and Palanivelu, 2004). However, it is noted that the removal efficiency was inferior initially then reached similar levels as the others after 20 min at −1.4 V/SCE. In addition, COD removal efficiency was basically the same in the case of −1 and −1.2 V/SCE, whereas the treatment at −1.4V/SCE exhibited an inferior degradation rate. As shown in Figure 5B, the COD removal efficiency was 67% without electricity, and increased to 76 and 75% at the cathodic potentials of −1.0 and −1.2 V/SCE, respectively, with residual COD concentration below 50 mg/L. However, when the cathodic potential was decreased to −1.4 V/SCE, the COD concentration reduced from 154 to 73 mg/L within 30 min, which can be ascribed to the inferior production of Fe (tot) for the Fenton reaction. According to the above results, −1.2 V/SCE was chosen as the optimal cathodic potential, and the released Fe (tot) concentration decreased by approximately 61% with the COD concentration further decreasing from 52 to 40 mg/L as compared with the WE Fenton process. The degradation of the 100 mg/L PNP solution was carried out in the traditional Fenton system with initially adding 20 mg/L of Fe²⁺ and 200 mg/L of H₂O₂ at pH₀ 3.0. It was found that PNP was destructed quickly within 5 min and exhibited faster decay of PNP and COD than the present combined process within 20 min. However, its final COD removal efficiency was outperformed by the present combined process at 30 min and was slightly surpassed by the latter (Babuponnusami and Muthukumar, 2014).

The variation of H₂O₂ concentration is shown in Figure 5C. It was observed that when applying potential at the iron cathode, the H₂O₂ concentration increased initially within 15 min and then decreased afterwards in the late reaction stage. As compared with the WE Fenton reaction system, the H₂O₂ concentration was higher, which was probably due to the smaller release of iron ions from the iron cathode. As the applied cathodic potential decreased from −1.0 to −1.4 V/SCE, the remaining concentration of H₂O₂ was slightly higher over the whole reaction period, which was also associated with the production of iron ions. In the traditional Fenton system, the concentration of H₂O₂ significantly decreased from 200 to 126 mg/L within 5 min, which corresponded to the quick reaction between Fe²⁺ and H₂O₂, and the remaining H₂O₂ concentration at 30 min was 13 mg/L. From Figure 6, it could be clearly observed that as the cathodic potential decreased from −1.0 to −1.2 V/SCE, the ratio of COD removal to produced Fe (tot) and H₂O₂ consumption was increased from 3.9 to 5.7 and 0.61 to 0.63, while the value significantly decreased to 4.4 and 0.45 at a higher cathode potential of 1.4 V/SCE. These results demonstrated that -1.2 V/SCE was the optimal cathodic potential with suitable Fe (tot) production for the electro-Fenton reaction, which exhibited slightly higher COD removal/Fe (tot) and COD removal/H₂O₂ consumption for the traditional Fenton reaction system. In the WE Fenton process, the much lower value, i.e., 2 for COD removal/Fe (tot) and 0.54 for COD removal/H₂O₂ consumption, was obtained, respectively.

According to the aforementioned experimental results, the optimal current density for solution pH regulation and cathodic potential for the ICEF reaction have been obtained. The results of PNP and COD concentration time are shown in Figure 7A. It was found that the PNP concentration of the anodic effluent slightly decreased from 100 to 91 mg/L while the COD removal efficiency was only 5%, owing to the oxidation at the BDD anode. During the combined process, the removal efficiency of PNP and COD attained 93 and 72% within 30 min, and the PNP pollutant was further destructed after filtration because of the coagulation of iron ions with the concentration decreased from 6 to 2 mg/L. Figure 7B shows the TOC removal efficiency of traditional Fenton and ICFE systems, the removal efficiency was 42% within 5 min by traditional Fenton but maintained relatively stable in the remaining time, finally the value reached 48%. This can be explained by the rapid consumption rate of Fenton reagents. However, the TOC removal efficiency achieved 60% in the combined system, which was much higher than traditional Fenton (Kavitha and Palanivelu, 2004). The high removal efficiency of PNP (98%), COD (79%), and TOC (60%) revealed more efficient degradation of pollutants in the combined system than the traditional Fenton system.

For traditional Fenton, pH was adjusted to the desired values with H₂SO₄ and NaOH solution while ferrous sulfate was used as the iron ions source (Gao et al., 2020). As shown in Figure 8, the conductivity was increased from 4.35 to 4.68 mS/cm after acidification during traditional Fenton, and the value continuously enhanced to 4.95 after the Fenton reaction, while the pH slightly decreased from 3.07 to 2.73, which can be explained by the large quantity of organic acids production. The pH of the effluent was neutralized to approximately 9.0 with the conductivity reduced to 4.76 mS/cm, which might be caused by acid-base neutralization. As for the present combined process, the conductivity of the anodic effluent increased to 4.55 mS/cm with the pH decreasing from 7.0 to 3.1, the cathodic effluent pH only slightly increased from 2.95 to 3.75 at the same current density (2 mA/cm²), owing to the buffering effect of organic acids on pH. During the neutralization process, pH was further adjusted to neutral (∼9.0) by the NaOH solution (Chen et al., 2021), driving the complete precipitation of Fe(OH)₃. In this case, the conductivity further decreased to 4.37 mS/cm. Generally, the salinity of traditional Fenton increased by 9% compared with the initial solution while the present combined process remained basically unchanged after the reaction, avoiding the obvious increase of water salinity caused by the addition of a large amount of chemical reagents in traditional Fenton.

A set of experiments was performed to verify the main mechanism of pollutant degradation. As shown in Figures 9A,B, the PNP removal efficiency was only 9% without front-end pH adjustment, owing to basically no Fe²⁺ production under the neutral solution, indicating that the pH-regulation process is indispensable for the ICEF system. In addition, only 7% of PNP was oxidized by H₂O₂ in solution. In the above two reaction system, COD concentration slightly increased after the reaction, which was possibly subject to formation of some intermediates which interfered with the COD concentration. The direct oxidation of PNP by the DSA anode was also examined without H₂O₂ at pH₀ = 3.0, the removal efficiency of PNP and COD concentration were 38 and 13%, respectively. Thus, the direct PNP degradation at the anode was ruled out, indicating that PNP in solution was mainly oxidized by •OH produced by the Fenton reaction.

To further investigate the effect of cathode potential on iron ion concentration, we conducted the following experiments. As shown in Figure 10A, at the cathode potential of −1.2V/SCE without adding H₂O₂, the produced Fe (tot) concentration was 4 mg/L, which reduced by 80% compared with no electricity as shown in Figure 3 (20 mg/L). And in the case of adding 200 mg/L of H₂O₂ and applying the cathode potential of -1.2 V/SCE, the Fe (tot) concentration increased to 17 mg/L. There are two reasons for this phenomenon, one is that the iron plate would react with H₂O₂ in the presence of H⁺ through Eq. 10 (Pan et al., 2019); the other is that the presence of H₂O₂ promoted the conversion of Fe²⁺ to Fe³⁺ and the generated Fe³⁺ may further react with the iron plate to produce Fe²⁺ through Eq. 11. As shown in Figure 10B, 30 mg/L of Fe³⁺ was also added into the solution at the cathodic potential of -1.2 V/SCE without H₂O₂, the Fe (tot) concentration slightly increased to 35 mg/L within 30 min and Fe³⁺ concentration reduced to 26 mg/L. The results demonstrated that the reaction between Fe³⁺ and Fe⁰ contributed marginally to the release of iron ions during the ICEF process, while the direct oxidation of Fe⁰ by H₂O₂ probably dominated the release of iron ions from the iron plate. As a result, in the WE Fenton system, H₂O₂ was not only consumed for the Fenton reaction but also for the release of Fe²⁺ by the reaction with the iron plate. Thus, a part of H₂O₂ cannot be utilized for •OH production via the Fenton reaction, leading to the inferior degradation efficiency of PNP. Fe + H 2 O 2 +   2 H + →  Fe 2 + + 2 H 2 O (10) Fe 3 + +  Fe → Fe 2 + (11)

In the present combined system, the cost for electricity played an important role in the overall operating cost of the process. Apparently, the overall energy consumption is the sum of the pH adjustment unit and Fenton reaction process, the power consumption of per ton of wastewater was calculated based on Eq. 12 (Wang and Chu, 2011). E =   UIT V   × 10 − 3 (12) where E is the energy consumption, U is the voltage measured during the reaction (V), I is the applied current (A), t is the electrolysis time (h), and V is the volume of reaction solution (m³).

As shown in Figure 11A, during the pH-regulation process, current density had a significant impact on energy consumption, the value increased from 0.075 to 2.385 kWh/m³ as the current density increased from 0.5 to 6 mA/cm². In the present combined system, a pH of 3.0–3.2 was automatically attained by electrolysis with the energy consumption of 0.435 kWh/m³. Figure 11B shows that the energy consumption was increased from 0.115 to 0.235 kWh/m³ with the cathodic potential decreasing from −1.0 to −1.4 V/SCE. We can infer that the pH-regulation took up most energy consumption (72%) in our system, and the whole system power consumption of per ton of wastewater was 0.605 kWh/m³, which translates into a cost of $0.04/m³ based on the average US industrial electricity rate ($0.0653/kWh). Iron sheet is very cheap and highly reusable ($1.4), which can counteract the cost. The cost of H₂O₂ was $0.16 with 0.6 L of consumption per ton of water, while the unit price of H₂O₂ was $0.27/L. Furthermore, in situ electrochemical synthesis of highly concentrated H₂O₂ (Yamanaka et al., 2003; Chen et al., 2017) could be used as a replacement for externally supplied H₂O₂ in the future, eliminating the chemical cost.

The specific energy consumption was calculated in terms of the removal of 1 kg of COD from PNP wastewater by the Fenton process (kWh/kg COD) using Eq. 13, where U, I, and t are the average voltage (V), applied current (A), and electrolysis time (s), respectively (Kurt et al., 2007; Zhao et al., 2016). SEC = U × I × T ( COD 0 × V 0 − COD t × V t ) × 3.6 (13)

After pH adjustment and the 30-min ICEF process, the PNP mineralization shows an SEC value of 4.92 kWh/kg COD, the results in this study demonstrated that the ICEF system was environmentally friendly, efficient, and inexpensive in comparison. It was proven that the operation of the system greatly enhanced the treatment of PNP wastewater. Compared to traditional Fenton and other electro-Fenton systems, the ICEF system has its own merits. Firstly, the pH-regulation divided electrolysis cell with the PTFE membrane as the separating material was simple and convenient compared to others, an acidic pH of 3.0–3.2 was automatically attained at a steady state within 1 min, which is suitable for most Fenton-like reactions (Babuponnusami and Muthukumar, 2014). In contrast, ∼60 min was required in the divided electrolysis cell using an ion exchange membrane as the separating material (Liu et al., 2007). Secondly, the cathode potential applied on the iron plate can decrease amounts of iron irons released under acid solution and precisely control the Fe (tot) production, avoiding the addition of iron salt and increase of water salinity. All these advantages together suggest that the ICEF system has potential for cost-effective and efficient degradation of recalcitrant organic pollutants. Future work should also focus on improving the efficiency of cathode reduction of ferric iron and improving the efficiency of H₂O₂ utilization.