Electrochemical degradation of 2,4-Dinitrotoluene (DNT) from aqueous solutions using three-dimensional electrocatalytic reactor (3DER): Degradation pathway, evaluation of toxicity and optimization using RSM-CCD

2.1 Materials

2,4-Dinitrotoluene (2,4-DNT), (the chemical formula of C₈H₆Cl₂O and the purity of 99%) with a molecular weight of 182.14 g/mol, sodium hydroxide (NaOH) and Hydrochloric acid (HCl), Sodium Sulfate (Na₂SO₄, 99% purity), Lead nitrate (Pb(NO₃)₂, greater than99% purity), nitric acid (HNO₃, 95% purity), granular activated carbon (GAC, a diameter of 0.8–2.0 mm and specific surface area of 844.09 m²/g), and other chemicals used were provided by Sigma Aldrich (St. Louis, MO, USA). The chemical structure of 2,4-DNT and other information about it was represented in Table S1. It should be mentioned that the analytical grade of all chemicals used in this study was employed and further purification was not applied for them. Preparing the solutions was done using double distilled water. To regulate the solution pH, 0.1 M HCl and 0.1 M NaOH were employed.

2.2 Electrode preparation

Preparing the G/β-PbO₂ anode was done through electrochemical deposition of PbO₂ layers on graphite sheets (dimensions of 8 cm × 4 cm × 0.3 cm), which is as follows: at first, we sonicated the polished graphite sheets in 40% NaOH solution for 15 min and continued the sonication process for in a 1:1 (V/V) HNO₃/H₂SO₄ mixture another 15 min. After that, the electrochemical deposition process was done in a simple cell containing 0.5 M Pb(NO₃)₂ and 0.1 mol/ L HNO₃ at a constant current. In the studied process, the prepared graphite sheet and the stainless-steel sheet (with the same dimensions) were employed as the anode and the cathode, respectively. Afterward, at a constant current of 7.5 mA/cm² for 180 min at room temperature, the PbO₂ film was deposited on the anode, and lastly, the prepared G/β-PbO₂ anode was washed several times with deionized water (Dargahi et al., 2018; Samarghandi et al., 2020).

2.3 Characterization of G/β-PbO₂ anode

Through the employment of FESEM (model: FEI-Nova NanoSEM 450) and EDS (model: Bruker XFlash6L10) analysis, the surface morphology and chemical composition of the G/β-PbO₂ anode were evaluated. Moreover, XRD (model: Ultima IV, Rigaku) was employed for identifying the phase type and crystallite structure of PbO₂ film.

2.4 Synthesis of MCZ@Fe₃O₄ nanoparticles

2.5 Electrocatalytic degradation of 2,4-DNT

For conducting all electrocatalytic degradation experiments, a batch electrolytic cell (with an effective volume of 200 mL) equipped with a pair of anode and cathode was utilized. Inside the mentioned cell, the G/β-PbO₂ anode and the stainless-steel 316 (SS316) cathode (with equal dimensions) were parallelly placed at a distance of 4 cm. The third dimension of the 3DER was GAC (constant concentration of 24 gr/L), which was put in the space between the two electrodes. The two-dimensional system was created without GACs. The supporting electrolyte used in this study was Na₂SO₄ (in the different concentrations from 0.1 to 0.3 g/200 cc).

2.6 Analytical procedures

The leaching of Pb²⁺ was measured after completion of degradation of the 2,4-DNT in the studied processes using the inductively coupled plasma mass spectrometry (ICP–MS). Supplying the electrical current in this study was done using a direct current (DC) power supply (UNI-TUTP3315TFL, China) with an electric current of 0-5A and voltage of 0–30 V.

The high-performance liquid chromatography (HPLC) (Agilent Technologies Co. Ltd., USA) equipped with a C18 column (250 mm × 4.6 mm, 5 μm, Agilent) and a diode array ultraviolet detector at λ_max = 254 nm was employed for detecting the concentration of 2,4-DNT. The mobile phase used was a mixture of methanol/water (HPLC grade, Merck) which was used in a ratio of 70:30 at a flow rate of 1.0 mL/min with a column temperature of 25 °C. The injection volume of the 2,4-DNT was 20 μL. In addition, to measure the degradation intermediates of 2,4-DNT, the liquid chromatograph-mass spectrometer (LC-MS, Shimadzu LCMS 2010 A), which has been equipped with a C18 column (150 mm × 2.1 mm) and an electron spray ionization source, was used. a mixture of 60/40 (v/v) Acetonitrile (ACN) + 0.1% formic acid and H₂O + 0.1% formic acid was employed as mobile phase for LC-MS. Under following conditions including Mode, ESI+, Detection gain,1.8 kV, Prob Volt, 4 kV, CDL Volt, 25 V, Gas nebulizer, N₂ (grade 5), Flow gas, 1.2 L/min, CDL temperature, 270 °C, and Block temperature, 270 °C, the Mass spectra (MS) was done.

Hach pH meter (HQ430D, USA) was the device that was used to determine pH, and TOC analyzer (Elementar Analysen systeme GmbH, Germany) was employed for assessing total organic carbon (TOC). Chemical oxygen demand (COD) tests were fulfilled using COD Cell Test Method (photometric 25–1500 mg/L Spectro quant® (Merck KGaA, Darmstadt, Germany) and visible light spectrophotometry (Hach spectrophotometer DR 6000)). Other parameters, e.g., the biodegradability of 2,4-DNT, COD/TOC ratio, carbon oxidation state (COS), and the average oxidation state (AOS) have been considered in this study and were appraised under optimal conditions at room temperature.

Following equations, i.e., Eqs. (1 and 2) (Liu et al., 2020) were employed for estimating efficiency and kinetics of 2,4-DNT removal in 2D and 3D electrocatalytic oxidation systems:

In above equations, η, ${[\mathrm{D}\mathrm{N}\mathrm{T}]}_0$ , and ${[\mathrm{D}\mathrm{N}\mathrm{T}]}_{\mathrm{t}}$ indicate the removal efficiency (%), the concentration of 2,4-DNT (mg/L) at t = 0 min, and the concentration of 2,4-DNT (mg/L) t = t min, respectively. ${\mathrm{k}}_{\mathrm{o}\mathrm{b}\mathrm{s}}$ is representative of pseudo-first-order kinetic coefficient (min⁻¹) and t is reaction time (min). Calculating the synergistic effect of GAC and the electrochemical degradation process was done using Eq. (3) (Pedersen et al., 2019).

${\mathrm{k}}_{2\mathrm{D}}$ , is the kinetic coefficients of 2,4-DNT removal in 3D and ${\mathrm{k}}_{\mathrm{D}}$ is the kinetic coefficients of 2,4-DNT removal in 2D electrochemical oxidation systems.

2.7 Response surface methodology (RSM) and experimental design

The optimization of parameters considered for the present study was done using response surface methodology (RSM) with the central composite design (CCD) technique. RSM is a collection of mathematical and statistical methods. It is employed to model and analyze a process in which the desired response can be affected by a number of variables. In fact, through the employment of RSM, the optimum operating conditions or a region for the factors in which certain speciation are met is determined (Seidmohammadi et al., 2021). Routinely, the linear, interactive, and quadratic effects of independent variables on the system response at five levels (-α, −1, 0, +1, +α) (Heidari et al., 2021) are evaluated by the mentioned method. The independent variables selected in the present study were the initial solution pH (A), electrolysis time (coded as B), current density (coded as C), and 2,4-DNT concentration (coded as D) (Table 1). To determine the effect of different parameters on electrocatalytic degradation of 2,4-DNT, pH in the range of 3–11, electrolysis time in the range of 15–75 min, current density in the range of 1–5 mA/cm², and 2,4-DNT concentration in the range of 10–90 mg/L was investigated. The statistical data analysis was performed using the Design Expert Software (version 11). By assessing achieved experimental responses for fitting a second-order polynomial model, a functional numerical relationship between the process variables and responses was established as shown in Eq. (4) (Afshin et al., 2021; Dargahi et al., 2021c).

In the mentioned equation, Y is indicative of response percentage of 2,4-DNT degradation, X_i, X_j indicate the k number of variables, β_i, β_ii, β_ij represent coefficients of linear, quadratic, and interaction terms respectively, and ɛ represents the model error. Random error (ɛ) articulates the measure of the difference between observed and predicted values. The experimental results were fitted to the regression model and model adequacy, and evaluation of significant model terms was done using analysis of variance (ANOVA), Fisher’s F test value, and p-value. The coefficient of determination (R²), adjusted R², and predicted R² were employed for expressing the goodness of fit of the developed model. Using Equation (5), the number of designed experiments was determined. In the mentioned Equation, k shows the number of variables and Cp is representative of the number of repetitions at the central point (Molla Mahmoudi et al., 2020).

Calculation of the percentage effect of each independent variable (P_i) on the degradation of 2,4-DNT was done using Pareto analysis (Eq. (6)) (Samarghandi et al., 2021b):

3.1 Characterization of the G/β-PbO₂ and MCZ@Fe₃O₄ nanoparticles

The morphology of fabricated G/β-PbO₂ electrode with different magnifications was assessed using FESEM (Fig. S1). Based on Fig.(S1), the much smaller grain size, uniform, and compact structure were detected for pyramid crystal-shaped β-PbO₂ particles. Furthermore, the uneven and stratified properties of formed β-PbO₂ on graphite interlayer is led to enhancing the specific surface area of this electrode (Dargahi et al., 2018). EDS elemental analysis was another analysis, which was employed to identify the bulk composition of the synthesized G/β-PbO₂. In Figs (S2, S3), the EDS elemental mapping has been represented based on two major elements present in the surface of the G/β-PbO₂ anode. In elemental mapping, the atomic percentages of lead (Pb) and oxygen (O) were obtained to be 9.91% and 20.09%, respectively. Moreover, based on the EDS spectrum, there is no carbon peak; this reveals the suitably covering of graphite substrate with PbO₂ layers. Through the employment of XRD analysis on anode material prepared, exploration of the phase and crystalline structure of deposited PbO₂ film (Fig. S4) was done. According to the XRD spectrum, the correspondence of all the peaks to β phase and tetragonal symmetry of PbO2 with Joint Committee on Powder Diffraction Standards (JCPDS) card no. 04–005-4491 and 00–041-1492 of the International Centre for Diffraction Data (ICDD) database were proven (Mandal et al., 2018). The diffraction peaks were detected at 2θ = 25.4°, 32°, 36.2°, 40.4°, 45°, 49.2°, 52.2°, 55°, 58.9°, 60°, 62.7°, 67° and 74.5°; these were attributed to (1 1 0), (1 0 1), (2 0 0), (1 1 2), (0 2 2), (2 1 1), (2 2 0), (0 0 2), (3 1 0), (1 1 2), (3 0 1), (2 0 2) and (3 2 1) planes of β-PbO₂ (Heidari et al., 2021; Samarghandi et al., 2020).

In Fig. S5 (a), the results related to SEM, which provides the morphological characteristics of MCZ@Fe₃O₄ nanoparticles have been demonstrated; based on this, the non-uniform surface with many pores for MCZ@Fe₃O₄ nanoparticles is detected. Moreover, it is clarified that, after loading of Fe₃O₄ nanoparticles, the surface morphology of the MCZ@Fe₃O₄ nanoparticles has not changed. The XRD pattern (Fig. S5(b)) at an angle of about 2θ was the device used to estimate the crystalline phase of nanocomposites and characterize their structure, which was indicative of this fact that, in addition to the presence of Montmorillonite (the most and main part of zeolite clinoptilolite), there are non-clay impurities such as Quartz, Calcite, and Feldspar in the structure of the Zeolite. The high crystallinity of the adsorbent was confirmed based on the appearance of sharp and stretched peaks especially at 2θ = 24.5 and 27° (which correspond to clinoptilolite and quartz). The presence of iron oxide particles in the structure of the MCZ@Fe₃O₄ nanoparticle was approved based on the peaks in the X-ray diffraction pattern, which were detected at angles of 30.6°, 35.88°, 43.53°, 53.4°, 57.1°, and 62.6°. The above results finally approve the effective synthesizing of iron oxide particles and coating on zeolite.

3.2 Model development, statistical analysis and optimization of the process

Table 1 represents the statistical combinations of the main variables (i.e., initial pH, particle electrode concentration, current density, and electrolysis time) and the maximum actual and predicted degradation efficiency. Using RSM, experimental results were modeled by second-order polynomial models. Equation (7) represents quadratic models obtained in terms of the coded factors.

By employing the above equation, the response for applied levels of every variable in terms of coded variables can be determined. The high positive value obtained for variables is indicative of a highly positive effect on the response, however, the negative coefficients for the variables indicate an adverse effect on the response. The interaction of factors, which influence the performance of the 3DER process, is determined based on second-order polynomial models. In addition to the quadratic models, validation of empirical models is also done using statistical ANOVA (Bezerra et al., 2008). Through considering five parameters, i.e., the Prob > F of the model, the lack of fit test, adequate precision, adjusted regression coefficient (Adj. R²), and the regression coefficient (R²), the validation was done. The significance of the empirical model values is approved by Prob > F; the fitness of the model is assessed based on lack of fit; adequate precision is used to measure the signal-to-noise ratio, and the regression coefficient (R²) is employed for determining the variability between the actual and predicted results. The values greater than 4 for adequate precision are desirable. Moreover, the selected model is significant and accurate, when the values of R² and Adj R² values are close. Based on mentioned results in Table 2, the models for degradation of 2,4-DNT could provide results consistent with these criteria. It shows the usability for the prediction of the process performance. In addition, Table 2 clarifies the significance of the CCD model for 2,4-DNT degradation efficiency with an R² value of 0.9965. Accordingly, more than 0.96% of the variation of 2,4-DNT degradation efficiency is explained by the model and the model is not capable of<0.5% of the variations degradation efficiencies. Following conditions for validation of a model should be observed: Prob > F must be significant (i.e., <0.05), the lack of fit should be insignificant (i.e., greater than 0.05) and adequate precision should be at least 4. Since the values of P > F were<0.05, this model was considered to be significant. For 2,4-DNT degradation, A (pH), B (Electrolysis time), C (Current density), and D (2,4-DNT concentration) were significant factors. Moreover, AB (pH × Electrolysis time), AD (pH × 2,4-DNT concentration), CD (Current density × 2,4-DNT concentration), A² (pH²), and D² (2,4-DNT concentration²) were significant model interactions. The regression model variance was evaluated, results of which were described in Fig. 1 (a, b). The distribution of residual points on both sides of the straight line has been clarified, which is indicative of the ordinary distribution of modulus residuals. The error of the used model is typically system error within the permissive error, and the indicated amount of the degradation efficiency from the model can be well-matched with the true degradation efficiency. After analyzing the suitability of the empirical model, prioritization and comparison of the contribution of the operational parameters to the process performance were done using Pareto analysis and Perturbation plots. Specification of the percentage contribution of each factor on 2,4-DNT degradation efficiency as responses was done using Graphical Pareto Analysis according to Eq (6). According to Fig (S6), a significant contribution (28.54 %) for the effect of electrolysis time (B) in 2,4-DNT degradation efficiency was detected. Nonetheless, the lowest contribution (0.31%) in 2,4-DNT degradation efficiency was related to interaction effects electrolysis time × current density (BC). To compare the effect of the operational parameters at a specific point in the design space, perturbation plots were achieved. In these plots, responses are plotted by changing one factor within a selective range while keeping other factors constant. The higher the curvature of the factor is inactive of more sensitivity of the response to that factor. In this study, the plots were at the center of pH (A) = 7, electrolysis time (B) = 45 min, current density (C) = 3 mA/cm², and 2,4-DNT concentration (D) = 50 mg/L. According to perturbation plots shown in Fig. 1(c), it was found that 2,4-DNT degradation is the most sensitive to current density (C). Hence, current density has significant effects on response, but electrolysis time (B) has the least.

Close

Fig. 1

The diagnostic plots for validation of obtained model: (a) normal plot of residuals, (b) values of 2,4-DNT removal efficiency versus experimental values, and (c) Perturbation plot for 2,4-DNT degradation.

Export to PPT

In this study, the optimization of prediction of 3DER behavior was done using the RSM-CCD method. Four independent input variables were set to “in range” mode and 2,4-DNT degradation efficiency, as the model output was set to “maximum” mode. Among the provided solutions, maximum degradation efficiency for the proposed model was obtained at the first derivative solution. Fig. (S7) shows the optimum conditions obtained (pH = 4.17, electrolysis time = 50.5 min, 2,4-DNT concentration = 23.5 mg/L, current density = 4.8 mA/cm²) by the RSM-CCD method. By considering these conditions, the highest degradation efficiency of 2,4-DNT by the proposed model was predicted to be 96.05%.

The results related to the reusability of G/β-PbO₂ anode in the studied process used for electrodegradation of 2,4-DNT were presented in Figure (S8).. According to the results presented in Figure (S8), the degradation efficiency in the first oxidation cycle is 98.6%, while it was 94.6% in fifteenth oxidation cycles, which is indicative of only a 4 percent difference between the first and fifteenth cycles; based on this, the good stability and reusability of the prepared anode can be concluded. The higher stability of graphite electrodes can be due to several effective factors as follows: the penetration of lead dioxide particles into the inner layers of the graphite substrate, which causes more interaction and adhesion of the lead dioxide film with the graphite substrate is one of these factors. Another factor, as shown in FESEM (Figures S1) and XRD (Figure S4), is that the reduction in the particle size of β-PbO₂ on the graphite surface can reduce the defect density at the electrode surface and create a compact and uniform layer with a micro and dense structure (Dargahi et al., 2019). These results indicate the high electrochemical stability of G/β-PbO₂ electrodes. Therefore, modifying the surface of graphite electrodes and other electrodes prepared with the compact β-PbO₂ film not only eliminates the possibility of electrolyte penetration through cracks and pores but also increases the internal pressure caused by the production of oxygen gas inside the electrode. Therefore, the probability of corrosion, rupture, and mechanical collapse of the electrodes is reduced (Ansari and Nematollahi, 2018; Dargahi et al., 2018).

Moreover, based on results for leaching of lead ion obtained by ICP-MS analysis during the process, the 0.0051 mg/L of lead leaching in the sample solution was detected. This value was much lower than the standards announced by WHO and EPA for drinking water.

In optimum conditions (2,4-DNT concentration = 23.5 mg/L, j = 4.8 mA/cm², pH = 4.1, electrolysis time = 50 min, GAC dose = 6 g/250 cc), the value of energy consumption as a function of treated solution volume for 2DER and 3DER-GAC processes was also calculated by Eq. (8) (64, 65). The results showed that the energy consumption for 2DER and 3DER-GAC processes was 5.3 and 2.6 kwh/m³, respectively. According to the results, the 3DER-GAC has a lower energy consumption compared to 2DER due to high efficiency in the 2,4-DNT degradation, which was even lower than the energy consumption reported in studies conducted by Pipi et al. (2014) (Pipi et al., 2014), Souza et al. (2015) (Souza et al., 2015), Hashim et al. (2017 (Hashim et al., 2017) and Kobya et al. (2016) (Kobya et al., 2016). In the mentioned studies, the energy consumptions were observed to be 16.9, 455.5, 6.21 and 11.17 kwh/m³, respectively.

Where, E_EC is the electrical energy in kWh/m³, U is the cell voltage in volt (V), I represents the current in ampere (A), t_EC is the electrolysis time of the electrochemical process (hr).

3.3 The effect of various parameters

3.3.1

3.3.1 The effect of initial pH

The results of this part of the study, i.e., the effect of pH on the 2,4-DNT removal, were represented in Fig. 2. Based on mentioned Figure, the best DNT removal efficiency (94.85% after 75 min) by the studied system was achieved at a pH of 3, which is agreed with the study conducted by Dargahi et al. (Dargahi et al., 2019). However, an increase in the pH of the samples has led to a remarkable reduction in the removal efficiency. The significant effect of pH on the removal of pollutants has been confirmed by several studies (Ansari and Nematollahi, 2018; Mengelizadeh et al., 2019). This variable can affect the production of various radicals, e.g., hydroxyl radicals (Zhang et al., 2013a). Increasing the production of hydroxyl radical is observed for low values of pH. Enhancing the production of these radicals leads to development in pollutant degradation, which is due to increasing the possibility of contact between these radicals with DNT molecules and higher oxidation power. However, the predominant radicals in very alkaline conditions, especially pH above 12, were the hydroxide radical. The oxidation potential of the hydroxide is harshly declined in alkaline conditions, which is led to diminish the efficiency of the process, even in the presence of the predominant radical. Also, in terms of the degree of radical stability in the aqueous medium, the stability of hydroxide radicals gets much lower (Xie et al., 2021). In the study conducted by Xiu-Yan Li et al., TiO₂-SiO₂/GAC was prepared and used as particle electrodes in a 3DER for the degradation of textile wastewater; in their study (Li et al., 2017), the optimum pH of 3 was also found to be an important parameter for removal of pollutant. Also, in a study conducted by Rahmani et al. for degradation of diuron herbicide in a 3DER with felt/PbO₂ anode, the results showed that with increasing pH, the diuron degradation efficiency decreased (Rahmani et al., 2021), which can be interpreted as follows: As the pH of the solution decreases, the number of produced hydroxyl radicals increases, which have a higher ability for degradation of pollutant in an acidic environment, and therefore the oxidation of organic matter occurs at a higher rate which is consistent with the results of the present study.

Close

Fig. 2

(a) 3D and contour response surface plots showing the interaction of electrolysis time and pH (2,4-DNT concentration = 50 mg/L, j = 3 mA/cm², activated carbon = 6 g/250 cc); (b) 3D and contour response surface plots showing the interaction of Current density and pH (2,4-DNT concentration = 50 mg/L, Electrolysis time = 45 min, Activated carbon = 6 g/250 cc).

Export to PPT

3.3.2

3.3.2 The effect of electrolysis time

Increasing electrolysis time in many treatment methods can lead to greater contact between the pollutant and the treatment agent and increase process efficiency. To study the effect of electrolysis time on the degradation of 2,4-DNT, electrolysis time was selected in the range of 15–75 min (Figs. 2 and 3). According to the results, the electrolysis time was directly correlated with 2,4-DNT degradation efficiency. As the electrolysis time increased, the degradation efficiency of the 2,4-DNT also increased. Because increasing the electrolysis time is led to increasing the amount of hydroxyl radical produced, thereby increasing the efficiency of 2,4-DNT degradation by the electrochemical process (Chabi et al., 2014; Fang et al., 2015; Oudenhoven et al., 2011). Increasing electrolysis time enhances the opportunity to perform Eqs. 9–11, which is led to increasing the contact time of organic pollutants and the intermediates caused by its degradation with ^°OH (Samarghandi et al., 2021b). Improving the performance of 3DER systems in removing organic pollutants by increasing the reaction time has also been reported in studies conducted by Samarghandi et al. (Samarghandi et al., 2021a), Dargahi et al. (Dargahi et al., 2021a; Dargahi et al., 2021d), and Khosravi et al. (Khosravi et al., 2015).

(9)

$${\mathrm{O}}_2+{2\mathrm{H}}^{+}+{2\mathrm{e}}^{-}\to {\mathrm{H}}_2{\mathrm{O}}_2$$

(10)

$${\mathrm{P}\mathrm{b}\mathrm{O}}_2[{\mathrm{H}\mathrm{O}}^{·}]+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{P}\mathrm{b}\mathrm{O}}_2[{\mathrm{H}\mathrm{O}}^{·}]+{\mathrm{H}}^{+}+{\mathrm{e}}^{-}$$

(11)

$${\mathrm{P}\mathrm{b}\mathrm{O}}_2[{\mathrm{H}\mathrm{O}}^{·}]+\mathrm{o}\mathrm{r}\mathrm{g}\mathrm{a}\mathrm{n}\mathrm{i}\mathrm{c}\mathrm{s}\to {\mathrm{P}\mathrm{b}\mathrm{O}}_2[{\mathrm{H}\mathrm{O}}^{·}]+{\mathrm{m}\mathrm{C}\mathrm{O}}_2+{\mathrm{n}\mathrm{H}}_2\mathrm{O}+{\mathrm{H}}^{+}+{\mathrm{e}}^{-}$$

Close

Fig. 3

(a) 3D and contour response surface plots showing the interaction of Current density and 2,4-DNT concentration (pH = 7, electrolysis time = 45 min, activated carbon = 5 g/250 cc), (b) 3D and contour response surface plots showing the interaction of Current density and electrolysis time (2,4-DNT concentration = 50 mg/L, pH = 7, GAC dose = 6 g/250 cc).

Export to PPT

3.4 Synergy of 2,4-DNT degradation and kinetics studies

To clarify the synergy of 2,4-DNT degradation, a comparative study was conducted using 3DER-G/β-PbO₂, 2DER-G/β-PbO₂ anode, 2DER-Graphite anode, and the efficiencies of 2,4-DNT of these systems were compared (Fig. 4a). Based on results obtained under optimal conditions, the 2DER-Graphite anode and 2DER-G/β-PbO₂ anode after 50 min could provide 2,4-DNT degradation efficiencies of 54.5% and 79.5%, respectively. However, under similar operating conditions, the 2,4-DNT degradation efficiency was enhanced to 98.6% and 96.5% in the 3DER-G/β-PbO₂ with GAC and MCZ@Fe₃O₄ nanoparticles as particle electrodes, respectively.

Close

Fig 4

(a) The degradation of 2,4-DNT by the Electrochemical processes (2,4-DNT concentration = 23.5 mg/L, j = 4.8 mA/cm², pH = 4.1, GAC and MCZ@Fe₃O₄ nanoparticles dose as particle electrodes = 6 g/250 cc) (■ 3DER with G/β-PbO₂ anode and GAC as particle electrode; ▴3DER with G/β-PbO₂ anode and MCZ@Fe₃O₄ nanoparticles as particle electrode‌; ♦ 2DER with G/β-PbO₂ anode; ● 2DER with Graphite anode), (b) Kinetics of 2,4-DNT degradation, (2,4-DNT concentration = 23.5 mg/L, current density = 4.8 mA/cm², pH = 4.1, electrolysis time = 50 min, GAC dose = 6 g/250 cc).

Export to PPT

The degradation of 2,4-DNT efficiency using 2DER-G/β-PbO₂ and 2DER-G/β-PbO₂ processes was also estimated based on the rate constants (K), which were studied under optimal conditions (Fig. 4b). Fig. 4b and table (3,) with high R² values (R² greater than 0.99), was confirmed the applicability of the pseudo-first-order model for explaining the kinetic behavior, which highlighted the remarkable degradation ability of 2DER-G/β-PbO₂ and 2DER-G/β-PbO₂ processes in degradation of 2,4-DNT. Table 3 also represents the half-life (t_1/2) of 2,4-DNT degradation by evaluated systems. As can be seen its values for the 2DER-G/β-PbO₂ and 2DER-G/β-PbO₂ processes are different and equal to 25.95 and 15.53 min, respectively. Calculation of degradation synergies for the 3DER-G/β-PbO₂ system is performed by placing the kinetic coefficients of 2,4-DNT removal obtained in the 3DER and 2DER oxidation system (represented in Table 3) in the above-mentioned equation (Eq.12) as follow:

Based on the above, the synergy of evaluated contaminant degradation in 3DER with G/β-PbO₂ anode was 51.34%.

3.5 Mineralization, 2,4-DNT degradation pathway and toxicity testing using bioassay

Evaluation of 2,4-DNT mineralization in the 3DER-GAC system was carried out using G/β-PbO₂ anode based on COD and TOC removal efficiencies at three concentrations of 20, 50, and 80 mg/L 2,4-DNT under optimum conditions, which include the current density of 4.8 mA/cm², pH of 4.1, and GAC of 6 g/250 cc (Fig. 5(a, b)). According to the results, increasing electrolysis time has led to significant development of the COD and TOC removal efficiency by the G/β-PbO₂ electrode; so that at the electrolysis time of 10 min, the removal efficiencies of COD and TOC for a 2,4-DNT concentration of 20 mg/L were 28.9% and 19.5%, respectively. However, at the end of 50 min, the removal efficiencies for COD and TOC were 88.5% and 80.9%, respectively. Mentioned results clarified that the electrolysis time had a notable effect on the mineralization of 2,4-DNT. Furthermore, by calculation of parameters, including COD/TOC, AOS, and COS under optimum conditions (2,4-DNT concentration = 20.0 mg/L, current density = 4.8 mA/cm², pH = 4.1, electrolysis time = 50 min, and GAC dose = 6 g/250 cc), the biodegradability of 2,4-DNT was assessed. Results related to this part (Fig. 5c) was indicative of an enhancement in the values of AOS (from 1.27 to 2.36) and COS (from 1.27 to 3.68) in the outlet of the 3DER-GAC process and a reduction in the COD/TOC ratio (from 1.81 to 1.09), which confirms the biodegradability of 2,4-DNT by the 3DER-GAC process.

Close

Fig. 5

(a) COD removal, (b) TOC removal (current density = 4.8 mA/cm², pH = 4.1, electrolysis time = 50 min, GAC dose = 6 g/250 cc), (c) AOS, COS and COD/TOC of 2,4-DNT in the electrochemical processes process at the optimum conditions (2,4-DNT concentration = 20.0 mg/L, current density = 4.8 mA/cm², pH = 4.1, electrolysis time = 50 min, GAC dose = 6 g/250 cc).

Export to PPT

The experiments for determination of the mechanism and intermediates of electrocatalytic oxidation of 2,4-DNT in the studied system were conducted using the pollutant concentration of 25 mg/L in an 0.1 M electrolyte solution (Na₂SO₄) at pH = 5 and a current density of 4.5 mA/cm² for 50 min. Through following up the electrolysis process by a Liquid chromatography-mass spectrometry (LC-MS) (Fig. 6), the intermediates were identified and reported in Table 4.

Close

Fig. 6

LC/MS chromatographs and proposed pathway for degradation of 2,4-DNT by 3DER at the optimum condition.

Export to PPT

Table 4 Identified intermediates by LC-MS during 2,4-DNT degradation using 3DER with GAC particle electrodes and G/β-PbO₂ anode.

Molecular structure	Chemical name	m/z (Da)
	1-methyl-2,4-dinitrobenzene	182
	2-methyl-3,5-dinitrophenol	198
	2-hydroxy-4,6-dinitrobenzoic acid	228
	1,3-dinitrobenzene	168
	nitrobenzene	123
	4-nitrosophenol	123
	nitrosobenzene	107
	oxalic acid	90
	acetic acid	60

According to the results, masses above 182 are related to the binding of hydroxide to the compound and the production of 2-methyl-3,5-dinitrophenol. Also, by oxidation of the methyl functional group, a combination of 2-hydroxy-4,6-dinitrobenzoic acid with a mass of 228 is produced. In the main process, a methyl group is first isolated and a compound of 1, 3-dinitrobenzene with a mass of 168 is formed. By removal of the nitro group, compounds, i.e., nitrobenzene or 4-nitrosophenol compounds are produced. The opening of the nitrosobenzene compound with a mass of 107 produces oxalic acid and then acetic acid with a mass of 60. The end products of this process are water and carbon dioxide.

To conduct the toxicity tests, the use of standard strains of Gram-negative Escherichia coli (E. coli) and Gram-positive Staphylococcus aureus (S. aureus) were considered (Hasani et al., 2020). Moreover, the broth lactose culture medium was used to perform the toxicity test for degradation of 2,4-DNT by 3DER under the optimal conditions obtained. The accuracy of the study was confirmed by conducting the experiments in two steps. Lastly, the bacteria growth inhibition in 5 different time intervals were estimated by GI (%) = [(1 – OD_600S/OD_600B) × 100] for the input and output solutions (Fig. 7). In this equation, GI, OD_600S, and OD_600B represent the growth inhibition (%), the optical density of the main and control samples at the wavelength of 600 nm, respectively. As reported in Fig. 7, the growth rate of studied bacteria in the control sample and the main effluent was developed after 10 hr, and a decrease in their toxicity was detected. It should be noted that lower bacteria growth was observed in the input solution. Based on the results obtained under optimal conditions including 2,4-DNT concentration = 23.5 mg/L, current density = 4.8 mA/cm², pH = 4.1, electrolysis time = 50 min, and GAC dose = 5 g/250 cc, the toxicity rate for E. coli in the reactor inlet solution 55.68% and was decreased to 10.85% for the output solution (80.5% reduction of toxicity) after process. Moreover, the toxicity for S. aureus was 39.82% in the input solution. However, it was reduced to 12.58% in the output solution (68.4% reduction of toxicity) (Table S2).

Close

Fig. 7

(a) Growth trend of Escherichia coli (Gram-negative) and (b) Growth trend of Staphylococcus aureus (Gram positive) in toxicity bioassay.

Export to PPT

The results showed that 3DER effectually diminish the toxicity of solution containing 2,4-DNT after treatment, and the bioassay method using microorganisms could be used as an efficient and cost-effective technique to evaluate the toxicity of aqueous solutions. In this study, the reduction of toxicity in gram-positive bacteria was less than gram-negative. In other words, Gram-positive Staphylococcus aureus was less sensitive compared to Escherichia coli in toxicity tests, which may be related to the ability to form spores and cell wall structure in Gram-positive bacteria (Hasani et al., 2020).