Treatment of coking wastewater using a needle coke electro-Fenton cathode: optimizing of COD, NH₄⁺-N, and TOC removal and characterization of pollutants

Coking wastewater is refractory industrial wastewater produced during coal coking, gas purification, and recovery and refining of chemical products (Ahmed et al. 2010; Meng et al. 2023); originates from residual ammonia recovery, gas desulfurization, and coal tar purification recovery (Garcia-Segura et al. 2012); and comprises a large number of organic and inorganic pollutants (Kim et al. 2008). Coking wastewater comprises organic compounds such as phenols, benzene series, and polycyclic aromatic compounds; heterocyclic organic compounds containing nitrogen, oxygen, and sulfur; and inorganic substances including sulfate ions, chloride ions, and ammonium salts (Saidi et al. 2019). Among them, polycyclic aromatic hydrocarbons (PAHs) and phenols are highly toxic pollutants. If coking wastewater is discharged without effective degradation, it can harm the ecological environment and human beings (Borràs et al. 2013).

Deng et al. (2019). reported on a novel tripolyphosphate (TPP)-assisted EF process for the treatment of coking wastewater. The high efficiency of the TPP-assisted EF process in neutral pH was attributed to the newly formed Fe–O–P coordination of the iron ligand complex (Fe²⁺-TPP), which activated oxygen to generate and degraded organic matter. K. Zhao et al. (2019) and Q. Zhao et al. (2019) fabricated a highly efficient fluorine-containing activated carbon catalyst to treat coking wastewater via fluoridation modification. The fluorination-modified catalyst exhibited a high H₂O₂ yield and completely degraded phenol in 15 min. Zhou et al. treated coking wastewater using an EF system comprising an Fe sheet as the anode and graphite as the cathode, followed by peroxy-coagulation (Zhou et al. 2020), and the removal rate of chemical oxygen demand (COD) in this EF system was 70%. Fe-functionalized carbonaceous-material–based cathodes are the most widely studied functionalized cathodes for the EF treatment of organic pollutants because they can in situ produce a large amount of H₂O₂ during electrolysis using carbonaceous materials (Ganiyu et al. 2018; Poza-Nogueiras et al. 2018) and provide Fe²⁺ for the Fenton reaction. Needle coke is a new type of carbon material with an excellent structure and performance. It is used as the raw material for the production of ultrahigh-power electrodes, special carbon materials, carbon fibers, composite materials, and other high-end carbon products. Needle coke exhibits advantages such as good electrical conductivity, low specific resistance, a small thermal expansion coefficient, and high mechanical strength (Yang et al. 2013; Kondrasheva et al. 2020; Gabdulkhakov et al. 2021). After modification, the electrochemical activity of carbon materials changes, and the modified needle coke exhibits increased specific surface area and electrical conductivity. In a previous study, our group fabricated a novel iron-loaded needle coke composite cathode (Fe-NCCC), and rhodamine B (RhB) was used as the target pollutant to demonstrate the powerful decolorization (∼100%) and COD removal capability (81.6%) of the Fe-NCCC (Chi et al. 2022). However, despite its complex molecular structure, RhB is a single pollutant, and coking wastewater contains many complex organic substances with high pollution indicators such as COD, ammonia nitrogen (⁠-N), and total organic carbon (TOC). The EF system constructed using the Fe-NCCC is intriguing and promising for the treatment of coking wastewater. The removal conditions of each pollution index, removal mechanism, and removal rates of various characteristic pollutants are important references for examining the performance of the Fe-NCCC electrode and can also be the theoretical basis and design parameters for the final application of this electrode in the treatment of coking wastewater and other similar wastewater.

Herein, a novel needle coke composite electrode was fabricated for the treatment of coking wastewater. Based on the single-factor experimental results, a Box–Behnken design (BBD) response surface experiment was designed. Using the applied voltage, electrode spacing, and initial pH as the influencing factors and COD, -N, and TOC as the response values, respectively, three second-order regression models were established and tested. The reaction conditions of the EF system for the treatment of coking wastewater were optimized. Furthermore, the degradation mechanism of organic matter in coking wastewater using the EF system was analyzed via gas chromatography–mass spectrometry (GC–MS) and Fourier transform infrared (FT-IR) spectroscopy.

The coking wastewater used in the experiments was obtained from a steel plant in Anshan, China. The characteristics of coking wastewater are presented in Table 1.

Table 1

Characteristics of coking wastewater

CODCr (mg·L−1) .	-N (mg·L−1) .	pH .	Color (folds) .	TDS (ms·cm−1) .	TOC (mg·L−1) .
1,761	119	8.91	250	3.105	408

CODCr (mg·L−1) .	-N (mg·L−1) .	pH .	Color (folds) .	TDS (ms·cm−1) .	TOC (mg·L−1) .
1,761	119	8.91	250	3.105	408

View Large

The experimental device includes an electrolytic cell, a DC-regulated power supply (SG1731SC3A, Shanghai Wenkai Power Equipment Co., LTD), an aeration pump (BT-100CA, Chongqing Jieheng Peristaltic Pump Co., LTD), and electrodes. The electrolytic cell is a transparent plexiglass reactor with a dimensionally stable anode (DSA) and the self-made Fe-NCCC as the cathode. The Fe-NCCC was made using iron-loaded activated carbon and modified needle coke. The cathode preparation method and its characterization are described in the article from our group (Chi et al. 2022). The aeration pump provided oxygen to the reaction system.

Coconut shell–based activated carbon, used as the Fenton catalyst carrier and electrode carbon material, was purchased from Anshan Activated Carbon Factory. Needle coke, used as the main carbon electrode material and conductive agent, was procured from Anshan Kai Carbon Thermal Energy Material Co., Ltd. Polytetrafluoroethylene was purchased from Shanxi Lizhiyuan Battery Material Co., Ltd. HNO₃, (NH₄)₂Fe(SO₄)₂·6H₂O, and other chemicals used in the experiment were of analytical grade. All solutions were prepared using deionized water.

The COD and -N were evaluated using the standard method (Huang et al. 2016). TOC was measured via the combustion–oxidation–IR absorption method with a InnovOx laboratory-type TOC analyzer (GE Analytical Instruments' Sievers) using 30% sodium persulfate as oxidant, 6 molL⁻¹ phosphate as a buffer solution, and 5% sodium carbonate as absorbent. The color was measured via the dilution method. The types of oxygen-containing functional groups present in the samples were studied using FT-IR spectroscopy (Nicolet iS 10), and the changes in the species and molecular structures of organic compounds in wastewater were investigated using a gas chromatography–mass spectrometer (7890A/5975C).

There are a number of variables that influence the treatment of coking wastewater by EF, such as applied voltage, initial pH, temperature, and organic composition. We selected three of these factors that have a more significant impact and are more controllable for field application to investigate.

While the degradation rate decreases at a neutral pH of 6–7, it is more than 65%, probably because herein, the iron salt is loaded into the Fe-NCCC electrode, and a heterogeneous EF system is formed in the reaction. The iron catalyst is loaded on the active site of the electrode to catalyze the EF reaction, which has greater resistance to the change in pH of the solution. To a certain extent, the application range of EF pH is widened (Zhao et al. 2016).

Figure 1(b) shows that the COD removal rate increases with the applied voltage, and a maximum COD removal rate of 83.7% is obtained at an applied voltage of 5 V, but the COD removal rate decreases to 70.2% when the applied voltage is increased to 6 V. This trend can be because increasing the applied voltage increases the current, which accelerates the transfer rate of electrons on the cathode surface and electrochemical reduction reaction of oxygen, thereby promoting the generation of H₂O₂ and and the decomposition of organic pollutants (Liu et al. 2018). However, an excessive applied voltage leads to side reactions such as the hydrogen reduction reaction, reducing the current efficiency and effectiveness of the COD removal. Therefore, the experimental voltage was selected as 5 V.

As shown in Figure 1(c), in the electrode spacing range of 1–5 cm, the COD removal rate first increases and then decreases as the electrode spacing decreases. The maximum COD removal rate of 89.02% is obtained when the electrode spacing is 2 cm, but when the electrode spacing decreases to 1 cm, the COD removal rate decreases to 80.36%. This decrease in the COD removal rate with a decrease in the electrode spacing is because when the electrode spacing decreases, the current density increases (Fukui et al. 1993), which improves the electron transfer rate on the cathode surface and is conducive to generation, thus increasing the COD removal rate. However, when the electrode spacing is less than 2 cm, the COD removal rate decreases for two reasons: (1) the distance between the cathode and anode is too small, leading to the reduction of part of generated in the EF operation at the cathode, and the organic matter is not fully oxidized and decomposed. (2) If the electrode spacing is too small, the high current density promotes the hydrogen evolution side reaction (He et al. 2011), blocking the H₂O₂ source and reducing the yield of (Wang et al. 2010). However, when the electrode spacing is larger than 2 cm, the COD removal rate continuously decreases because the current density decreases with an increase in the electrode spacing (Sruthi et al. 2018), which reduces the electron transfer rate on the cathode surface and generation. Therefore, the optimal electrode spacing is 2 cm.

As shown in Figure 1(d), as the EF reaction time increases to 30 min, the COD removal rate also increases, and the degradation effect is the most significant in the first 30 min when the initial COD concentration of wastewater decreases from 1,761.6 to 933.7 mg·L⁻¹. Then, the COD removal rate slowly increases. The maximum COD removal rate of 89.13% is achieved at 180 min of EF reaction, and after 180 min, the COD removal rate slightly decreases.

This trend can be because within the first 180 min, most organic matter in wastewater is effectively oxidized and degraded, and macromolecular substances are transformed into small molecular substances via ring-opening and chain-breaking reactions, and the degradation effect of coking wastewater significantly increases. After 180 min of EF reaction, the remaining organic matter in wastewater has a stable structure and is not easily oxidized, and more energy is required to break the chemical bond. Further increasing the reaction time has no significant effect on the degradation of organic matter. Consistent with the findings of Abdulla et al., the reduction in COD removal may be due to the combination of excess and H₂O₂ (Ghjair & Abbar 2023). Therefore, the reaction time is 180 min.

To test the significance of the experimental design model and reliability of the degree of fit, analysis of variance (ANOVA) and regression equation error analyses were performed on the experimental data in Table 2. The experimental results are shown in Tables A.1–A.4.

As shown in Table A.1, the p-value of the COD model is 0.0490, which is less than 0.05, indicating that this model (Equation (10)) is significant. The lack of fit is the standard used to evaluate the reliability of the model data. The model can be proven to be valid only if the lack of fit is not significant. The p-value of the lack of fit is 0.3406, and it is not significant, indicating that the experimental model data are true and reliable, and the second-order fitting equation can accurately evaluate the data (Zhang et al. 2019).

The larger the multiple correlation coefficient R², the better the correlation. Table A.4 shows that the R² of this model is 0.8266, indicating that it can well explain the experimental prediction results. The coefficient of variation (CV) is less than 10%, showing that the model can accurately predict the experimental data. The CV of this model is 7.42%, indicating that the model has high reliability. Adeq Precision, representing the signal-to-noise ratio of the model, is greater than 4, indicating that the design of the experimental model is reasonable. The Adeq precision of the model is 7.984, that is, greater than 4, showing that the designed experimental model is reasonable and can be used to guide the experimental work.

Therefore, the model can be used to fit the COD removal rate of using EF to treat coking wastewater. The F-text was used to analyze the significance of parameters in the model (de Luna et al. 2020). It can be seen that the applied voltage and initial pH significantly affect the COD removal rate of wastewater, whereas the electrode spacing has no significant effect on the COD removal rate, and the square effect of the electrode spacing significantly affects the model. The significance order of the three variables in this model is as follows: applied voltage (X₁) > initial pH (X₃) > electrode spacing (X₂).

Table A.2 shows that the p-value of the model is less than 0.0001, indicating that the model is highly significant. The F-value is 60.76, and a high F-value indicates that the model design is meaningful (Baker et al. 2021). The lack of fit term of the model is 0.2531, which is not significant, indicating that the selected model is suitable. The higher the multivariate correlation coefficient R², the better the correlation. As shown in Table A.4, the R² of the model is 0.9874, indicating that the model can explain 98.74% of the results. The Adj-R² value of 0.9711 is close to R², and both are close to 1. The model is highly reliable and can truly reflect the relationship between the experimental and predicted values. CV is 3.49%, considerably less than 10%, further verifying the good fit of the experiment and the validity of the model.

To sum up, this model (Equation (11)) can fit and predict the TOC removal of coking wastewater using the EF system comprising the Fe-NCCC. The interactions between the applied voltage and initial pH and electrode spacing and initial pH had significant impacts on the TOC removal rate, whereas the interaction between the applied voltage and electrode spacing had no significant impact on the TOC removal rate.

As shown in Table A.2, because p > F-value, the order of significance of the three variables in this model is as follows: initial pH (X₃) > applied voltage (X₁) > electrode spacing (X₂). The initial pH has a significant effect on the TOC removal rate, and its p-value is less than 0.0001. The square effect of the applied voltage and initial pH is significant, and its p-value is less than 0.05. However, the square effect of the electrode spacing is not significant, and its p-value is 0.0717.

According to Table A.3, the p-value of the model is 0.0051, that is, less than 0.05, and the lack of fit term is not significant, indicating that the model is significant, and the fitting degree is good in the selected fitting interval. Table A.4 shows that the R² of this model is 0.916, indicating that the model can explain 91.6% of the predicted results and can truly reflect the relationship between the experimental and predicted values. The Adeq precision value of the model is 12.409, higher than 4, further verifying that the experimental design is reasonable and that the model is well established.

Therefore, this model (Equation (12)) can fit the -N removal rate in coking wastewater treated using the EF system comprising the Fe-NCCC. Among them, the interaction between the electrode spacing and initial pH significantly affects the -N removal rate, and the square effect of the initial pH of the solution is highly significant. The three variables in this model significantly impact -N removal because their p-values are less than 0.05 (Table A.3). The significance order is as follows: electrode spacing (X₂) > applied voltage (X₁) > initial pH (X₃).

The 3D figure in Figure 4(b) has a steep slope, indicating that the changes in these two parameters strongly affect the -N removal rate in the EF reaction. When the applied voltage is constant and electrode spacing changes from 3 to 1 cm, the -N removal rate increases from 12.15 to 50.53%. The smaller the electrode spacing, the higher the current density and the better the -N removal rate. At a certain electrode spacing, the higher the voltage, the better the -N removal rate. When the applied voltage is 5 V, the removal rate of -N is the highest. Therefore, decreasing the electrode spacing and increasing the applied voltage is beneficial to removing -N.

According to the aforementioned research and analysis, the BBD models obtained in this experiment are significant and can be predicted within the range of selected experimental parameters. At the same time, under the optimal conditions (applied voltage, electrode spacing, and initial pH) optimized using the models, the experiment was performed to compare the difference between the predicted and real values, as shown in Table 3.

Table 3

Analysis of model verification

Item .	COD .	TOC .	-N .
Applied voltage (V)	4.92	4.83	4.99
Electrode spacing (cm)	2.29	1.06	1.11
Initial pH	3.01	3.29	3.00
Predictive value (%)	92.13	73.91	78
Actual value (%)	89.96	73.44	84.12

Item .	COD .	TOC .	-N .
Applied voltage (V)	4.92	4.83	4.99
Electrode spacing (cm)	2.29	1.06	1.11
Initial pH	3.01	3.29	3.00
Predictive value (%)	92.13	73.91	78
Actual value (%)	89.96	73.44	84.12

View Large

According to the experimental results obtained under the optimal conditions, the errors between the experimental and predicted values of COD, TOC, and -N are 2.17%, 0.47%, and 6.12%, respectively, indicating that the response surface models are reasonable and can be applied.

Therefore, the coking wastewater contains unsaturated C = O bonds, aromatic compounds, and compounds containing methyl before and after the EF process, and the intensities of the absorption peaks are reduced to different degrees after the treatment, indicating that the Fe-NCCC has a good removal effect on the aforementioned organic compounds. The intensity of absorption peak at 3,407.07 cm⁻¹ in the FT-IR spectrum of the water sample after treatment is significantly reduced, indicating that the system has the most significant degradation effect on phenolic hydroxyl substances. The intensity of the absorption peak at 3,407.07 cm⁻¹ in the FT-IR spectrum significantly decreases, indicating that the system has the most significant degradation effect on phenolic hydroxyl compounds.

The organic composition of coking wastewater and effluent after treatment using the EF system comprising the Fe-NCCC was analyzed via GC–MS, as shown in Figure 5(b) and 5(c). Different kinds of organic compounds are present in coking wastewater. A comparison of Table A.5 shows that the types and contents of organic compounds change in coking wastewater before and after treatment.

GC–MS analysis shows that coking wastewater mainly contains phenols, nitrogen-containing heterocycles, quinolines, PAHs, oxygen-containing heterocycles, alkanes, and other organic pollutants. The contents of these substances in the wastewater are relatively high, accounting for ∼89.13% of the total organic pollutants. In addition, quinoline, indole, 5H-indeno[1, 2-B]pyridine, 1-indanone, and other pollutants are also present in the wastewater. After the EF treatment, the wastewater mainly contains styrene, dimethyl phthalate, and long-chain alkanes. By comparing the pollutants present in the wastewater before and after the EF treatment, it can be seen that after the EF treatment, the types and contents of organic compounds in the effluent significantly decrease; most of the macromolecular organic compounds with complex structures disappear, and the characteristic phenolic substances in coking wastewater are not detected. Therefore, the Fe-NCCC can effectively decompose complex macromolecules from organic matter via the EF process, weaken the toxicity of coking wastewater (Ren et al. 2019; Q. Zhao et al. 2019), and reduce pollution indicators such as COD and TOC.

The dynamic analysis of COD degradation in coking wastewater was performed using the EF system comprising the Fe-NCCC. COD removal was measured, and its removal rate was calculated every 30 min. The experimental results are shown in Table A.6. Linear fitting regression was performed on the experimental data, and the reaction order was determined (Table 4). The COD kinetic analysis diagram is shown in Figure 5(d).

According to the fitting results in Table 4 and Figure 5(d), the fitting effect of the quasi-first-order kinetic equation is better with a correlation coefficient R² of 0.9599, and the reaction conforms to the quasi-first-order kinetic equation. The reaction kinetic model is ln (C₀/C_t) = 0.01122t − 0.07486, and the reaction rate constant K is 0.01122 min⁻¹.

Figure 5(e) shows that the color of coking wastewater decreases from 250-folds to colorless in a reaction time of 120 min, demonstrating that sufficient is generated in the EF reaction to oxidize and decompose chromogenic groups, phenols, and other macromolecules in wastewater, allowing its rapid degradation and decolorization. After 180 min of the reaction, the BOD₅/COD of coking wastewater increases from 0.126 to 0.34, indicating that the EF system comprising the Fe-NCCC can improve the biodegradability of coking wastewater.

The EC of the treatment of coking wastewater using the EF system comprising the Fe-NCCC was analyzed. According to Equation (6), the EC of a unit of COD was calculated. The EC changing with reaction time is shown in Figure 5(f). The corresponding average current in the reaction is 0.97 A, and the experimental water consumption is 300 mL.

Figure 5(f) shows that as the reaction time increases, the unit EC of COD gradually increases, indicating that with an increase in reaction time, the COD removal rate of the reaction system gradually increases, and the COD concentration gradually decreases. When the reaction time is 180 min, the COD concentration decreases from 1,761.14 to 180.12 mg·L⁻¹. After the calculation, the EC is only 30 kW·h (kg COD⁻¹), which is because is produced by the EF system using the Fe-NCCC, and plays a major role in the EF reaction and decreases COD with high decomposition efficiency. At the same time, the electrode conductivity is good, the heat loss is reduced, and the current efficiency is improved; therefore, the unit EC of COD is lower than those reported in the literature (Table 5).

Table 5

Comparison of energy consumption of coking wastewater treated by electrochemical method

Processing method .	Electrode/catalyst .	EC (kW·h (kgCOD) −1) .	Treatment effect .	References .
Electrochemical oxidation	Anode: Ti/RuO2–IrO2 Cathode: Mesh titanium	114.4	The COD and -N concentration degraded 86 and 100% in 80 min at the current density of 5 mA·cm−2.	Wang et al. (2014)
Electrochemical oxidation	Catalyst: Black titanium oxide nanotube array	63–68	The BOD/COD was from 0.19 to above 0.3 in 4 h at a current density of 2 mA·cm−2.	Liu et al. (2021)
Electrochemical oxidation	Anode: BDD Cathode: Stainless steel	67.9	The COD, BOD5, -N, and N removal efficiency was 48.74%, −63.82%, 21.15%, and −2.57%, respectively.	Wang et al. (2015)
Electro-Fenton	Anode: Ti/SnO2–Sb Cathode: Steel	396.56	The COD and TOC removal yields of the coking wastewater were 83.05% and 74.56%, respectively.	He et al. (2022)
Electro-Fenton	Anode: Fe-NCCC Cathode: DSA	30	The COD, -N, and TOC removal efficiency was 89.96%, 84.12%, and 73.44%, respectively.	This study

Processing method .	Electrode/catalyst .	EC (kW·h (kgCOD) −1) .	Treatment effect .	References .
Electrochemical oxidation	Anode: Ti/RuO2–IrO2 Cathode: Mesh titanium	114.4	The COD and -N concentration degraded 86 and 100% in 80 min at the current density of 5 mA·cm−2.	Wang et al. (2014)
Electrochemical oxidation	Catalyst: Black titanium oxide nanotube array	63–68	The BOD/COD was from 0.19 to above 0.3 in 4 h at a current density of 2 mA·cm−2.	Liu et al. (2021)
Electrochemical oxidation	Anode: BDD Cathode: Stainless steel	67.9	The COD, BOD5, -N, and N removal efficiency was 48.74%, −63.82%, 21.15%, and −2.57%, respectively.	Wang et al. (2015)
Electro-Fenton	Anode: Ti/SnO2–Sb Cathode: Steel	396.56	The COD and TOC removal yields of the coking wastewater were 83.05% and 74.56%, respectively.	He et al. (2022)
Electro-Fenton	Anode: Fe-NCCC Cathode: DSA	30	The COD, -N, and TOC removal efficiency was 89.96%, 84.12%, and 73.44%, respectively.	This study

View Large

At an industrial electricity price of 0.5 RMB Yuan per KWh, the electricity consumption cost of the EF system for treating 1 m³ of coking wastewater is 23.5 RMB Yuan.

The Fe-NCCC is used as the cathode and DSA as the anode to form an EF system to treat coking wastewater. The COD removal rate was 89.96% under optimum conditions of an applied voltage of 4.92 V, an electrode spacing of 2.29 cm, and an initial pH of 3.01; the optimized removal rate of -N was 84.12% under the condition of an applied voltage 4.99 V, an electrode spacing 1.11 cm, and an initial pH of 3; and TOC reached 73.44% under optimum conditions of an applied voltage 4.83 V, an electrode spacing 1.06 cm, and an initial pH of 3.29, respectively. The applied voltage and initial pH influence the COD, -N, and TOC removal. The electrode spacing significantly affects the -N and TOC removal. The initial pH has the greatest effect on TOC removal, followed by COD removal, and the least effect on -N because the initial pH mainly affects the generation of in the Fenton reaction, and the thorough mineralization of organic compounds is more dependent on ⁠, while the indirect oxidation of the anode also plays a major role in the removal of -N. The degradation mechanism of the EF system for organic pollutants in coking wastewater is as follows: the Fe-NCCC generates via electrocatalysis to decompose phenols, quinolines, PAHs, and oxygen- and nitrogen-containing heterocyclic compounds. The cost of treating 1 m³ of coking wastewater is ∼23.5 RMB Yuan. The EF system comprising the Fe-NCCC can effectively remove pollutants from coking wastewater with low power consumption while reducing various pollution indicators, demonstrating that the Fe-NCCC can serve as a cost-effective EF cathode in refractory industrial wastewater treatment.

This work was supported by the Research Foundation of the University of Science and Technology Liaoning (2019FW03).

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

The authors declare there is no conflict.