2.1. Chemicals and microorganisms

Sodium sulphate with 99% purity, potassium acetate (CH₃CO₂K), sodium hydroxide (NaOH), sulfuric acid (H₂SO₄), sodium chloride (NaCl) of analytical grade, and terephthalic acid with 99% purity were purchased from Loba Chemicals Pvt. Ltd., India. Terephthalic acid was used to estimate the hydroxyl radical formation in the EO process. Luria Bertani broth was purchased from HiMedia Laboratories Pvt. Ltd. Mumbai, India. All the bacterial strains in the present study, Escherichia coli (MTCC no. 448), Bacillus subtilis (MTCC no. 441), Staphylococcus aureus (MTCC no. 902), Salmonella enterica (MTCC no. 1165), Acinetobacter calcoacetius (MTCC no. 1948), Serratia marcescens (MTCC no. 2645), Listeria sp. (MTCC no. 4214), Enterococcus faecalis (MTCC no. 6845) were obtained from IMTech, Chandigarh, India. In this study, these eight different bacterial strains were chosen since they commonly found in hospital wastewater [36]. These bacteria were analyzed separately by varying the current density and electrolyte concentration range as reported previously [8]. All the solutions were prepared in double-distilled water of high purity.

2.2. Sample preparation and experimental setup

EO treatment of the simulated wastewater was carried out in a batch reactor. The simulated wastewater was prepared by dissolving 3.5 g of potassium acetate, 400 mg of sodium sulfate, and 5 mL of glacial acetic acid in 1 L of double-distilled water. This simulated wastewater containing inorganic salts simulates the composition of human urine. The concentration of sodium chloride was varied according to the reaction. The bacterial culture from log phase was taken. This culture was grown up to an absorbance range of 0.8–1.0 nm by inoculating 1–2 μL of bacterial strain in Luria Broth. The concentration of bacterial culture in the test solution was calculated in accordance with the absorbance of the bacterial culture. The simulated wastewater solution was freshly prepared each time. The EO treatment of simulated wastewater was carried out in a batch mode in a glass reactor having a working volume of 300 mL, as shown in Figure 1. The MMO anode used in this study was purchased from Tiaano Pvt. Ltd., Chennai, India. The MMO anode was composed of a titanium sheet coated with iridium, ruthenium, and platinum oxides. The stainless-steel cathode was purchased from a local vendor in Mohali, India. The dimension of the electrodes was (70 mm × 70 mm × 1 mm) with a surface area of 42 cm² (10 mm inter-electrode spacing). The DC power supply was purchased from (GAYATRI ENGINEERS, Maharashtra, India, Model: 0–30 V, 0–2 A) to maintain the current density for each experimental run. The magnetic stirrer was used at 500–550 RPM to maintain the homogeneity of the electrolyte concentration in the sample. All the experiments were performed three times to ensure the reproducibility of results. All the samples were collected at a specific time interval of 1 min.

2.3. Analytical methods

The analytical studies of bacterial inactivation were carried out by analyzing the bacterial samples through a UV–visible spectrophotometer at 600 nm. The inactivation of bacteria was also confirmed by plating the initial and final samples on Luria agar plates and incubating them in an incubator for 24 h. The hydroxyl radical (OH∙) study was done using a fluorescence spectrophotometer (Shimadzu RF 6000) during the EO process. The terephthalic acid (TPA) used reacts with OH∙ in the EO process [37] and produces a fluorescent compound 2-hydroxy terephthalic acid (TAOH). The intensity of this fluorescent compound was taken at an excitation wavelength of 315 nm and an emission wavelength of 425 nm. Total available chlorine (TAC) analysis was done to check the amount of chlorine produced during the EO process. TAC analysis was performed for the untreated and treated simulated wastewater samples by APHA standard (4500-CI B) method. The surface morphology and elemental composition of MMO anodes were executed by SEM-EDS (JSM-6510LV, JEOL, Japan).

The electronic state of the MMO anodes was analyzed by XPS (X-ray photoelectron spectroscopy (PHI 5000 Versa-Probe III, Physical Electronics)). The potassium ion leakage test was performed by APHA: 3500 K method, which confirmed the inactivation of bacteria in untreated and treated simulated wastewater samples.

2.4. Experimental procedure

The EO treatment of the simulated wastewater containing eight different bacteria was carried out under galvanostatic conditions. The pH of the working sample was adjusted accordingly by 1N HCl and 1N NaOH. The experimental reaction was carried out at an original pH of 4.25. The conductivity of the simulated wastewater solution was improved by adding NaCl as a supporting electrolyte. The initial conductivity of the sample solution was around 2.5 mS. The NaCl concentration and current density were varied according to the experimental reaction. At specific time intervals, 100 μL of the sample solution was taken and was inoculated in 5 mL of Luria broth. The samples were then incubated for 24 h in an incubator. After 24 h, the samples were analyzed spectrophotometrically at a wavelength of 600 nm to measure the absorbance of the samples for determination of % inactivation of bacteria. After each experimental run, the electrodes were washed by dipping them in a 5% H₂SO₄ solution. The complete inactivation of the bacteria was observed within a short period of time. Thus, the efficacy of the EO process was proved in terms of the inactivation of bacteria.

2.5. Experimental design and analysis

Response Surface Methodology (RSM), a three-level Box-Behnken design (BBD) based statistical technique is used to optimize the operational parameters as well as to reduce the number of experimental runs. In addition, it is used to analyze inter-parametric interactions of the input parameters and their effects on the responses [38]. The RSM technique works according to the equation N = s² + s + m_p, where s is the number of factors and m_p is the replicating number of central points [39]. The three initial input parameters in this study are NaCl dose (n) (X₁), current density (j) (X₂), and treatment time (t) (X₃) which were coded by three levels designated as −1 (low), 0 (center), +1 (high) as shown in Table S1. All the statistical plots have been generated by the software Design-Expert V-6.0.8. The experimental range for each process parameter was determined based on a literature survey and preliminary tests [40, 10]. In this study, there were 17 experimental runs designed by BBD under a three-level factorial design, as shown in Table 1. The efficacy of the EO process was determined by analyzing the responses of % inactivation (Q₁) and energy consumption (Q₂). The energy consumption (E) in (kWhrm⁻³) and % inactivation was calculated by the Equation (1) [41].

To fit the experimental data of the input variables on the response Q, a second-order polynomial equation (2) was used by considering all square terms, linear terms, and linear by linear interaction terms, the quadratic response model [42] can be expressed as:

3.1. Statistical analysis with BBD

The BBD of RSM developed by a statistical Design-Expert software of 6.0.8 version (STAT-Ease Inc., Minneapolis, US) was used for the EO bacteria treatment. The results of % inactivation and energy consumption were analysed by the EO experimental runs suggested by the BBD design set process, as shown in Table 1. The lack of fit tests, sequential model sum of squares, and model summary statistics tests were used to select the best regression models among various other models of linear, modified, mean, cubic, 2FI, and quadratic [46, 47]. Among the nine transformations of RSM, the none transformation was used for both the responses, and the cubic and quadratic models were used to analyse % inactivation and energy consumption, respectively. The model summary statistics for % inactivation showed the cubic model’s R² and adjusted R² values to be 0.997 and 0.9987, respectively. The cubic model here is aliased. In the case of the quadratic model, the R² and adjusted R values for energy consumption were 0.9987 and 0.9971, respectively. Here, the quadratic model is suggested.

The relationship between the responses and the input parameters in terms of coded factors obtained from the RSM software were expressed by the quadratic model equation as shown in Equations (3) and (4).

The analysis of variance (ANOVA) results is shown in Table 2, obtained from the second-order polynomial equation for both the responses for the EO treatment of bacteria using MMO. The model F-values 1011.24 and 619.97 for % inactivation and energy consumption, respectively, imply that the model is significant. Values of “Prob > F” < 0.05 indicate that the model terms are significant [48]. There is only a 0.01% chance that the “Model-F values” this large could occur due to the noise. In case of % inactivation, the ANOVA results suggests that the model terms, NaCl dose, current density, time, NaCl dose², current density², time², NaCl dose × current density, NaCl dose × time, NaCl dose² current density, NaCl dose² time, NaCl dose current density² are significant terms. For energy consumption, the significant model terms are NaCl dose, current density, time, current density², NaCl dose × current density, and current density × time. The adequate precision ratio was 79.548 and 91.025 for % inactivation and energy consumption, respectively. A ratio greater than 4 indicates that the model is desirable and can be used to navigate the design space [49].

The diagnostic plots for the actual versus predicted experimental values were analysed to evaluate the data points and check the mathematical model’s accuracy. These plots indicate the relation between the predicted and the actual experimental values [50]. The data points for this plot are very close to the straight diagonal line, thus suggesting a good relationship between the actual experimental values and the predicted data values by the mathematical model, as shown in Figure S1 [51]. To study the effect of the operational parameters on both the responses, the 3D response surface graphs were analysed for each parameter.

3.2. Effect of process parameters

3.2.1. Effect of current density (j) on % inactivation

Current density is one of the significant process parameters in the EO treatment from both cost-effectiveness and mechanistic study point of view. The generation of ROS and RCS on the electrodes [52] and the electron transfer mechanism depend on the current density values by following reactions (5)–(11) [30]. The maximum inactivation efficiency at acidic pH was due to the adsorption rate of ∙OH on the surface of MMO anode was high which leads to direct oxidation of contaminants, as at basic pH, the adsorption rate of ∙OH decreases due to transformation of ∙OH into lower potential oxidants like H₂O₂ and HO₂∙. Moreover, at acidic pH, the production rate of high potential oxidants like HOCl, Cl₂, ClO⁻ was maximum. Furthermore, the pH of the solution during the EO process was ∼4.5 [53]. The production rate of chloro-oxidant species increases by increasing the j value up to a certain limit which also depends on the n value [54]. The EO treatment process works best at low j values with MMO because of the low oxygen evolution potential, as there is no side reaction and therefore increases the process efficiency. From Figure 2a, the effect of j and n on % inactivation can be analyzed. With the increase in j values from 2.38 to 11.90 mA/cm², there is a gradual decrease in the % inactivation of bacteria. High j values would not effectively inactivate the bacteria at high electrolyte (NaCl) concentration because the effect of NaCl dose is very low which is again a significant parameter for RCS generation. Although higher j values may lead to an increase in temperature of the solution, even using only bacterial consortium alone without any organic load may also result in an increase in temperature, but the temperature increase in this situation is not enough to denature the bacterial proteins or inactivate the bacterial enzyme. Hence, it may be concluded that electrolyte (NaCl) concentration plays an important role at high and low current densities [55]. At a low j value of 2.38 mA/cm² and a minimum n value of 0.5 g/L, the % inactivation is maximum at around 96% because of low oxygen evolution potential which disfavours other side reaction and increases the inactivation rate. Further, with a gradual increase in n value from 0.5 to 1.50 g/L, the % inactivation remains constant. But from 2.0 to 2.50 g/L, there is a gradual decrease in % inactivation.

2C l − →C l 2 + 2 e −

(5)

C l 2 + H 2 O→HOCl+ H + +C l −

(6)

HOCl → H + +OC l −

(7)

H 2 O→ ∙OH+ H + + e −

(8)

2OH → H 2 O 2

(9)

H 2 O 2 → O 2 + 2 H + + 2 e −

(10)

O 2 + ∙O→ O 3

(11)

3.2.4. Effect of current density (j) and NaCl dose (n) on energy consumption

Energy consumption depends on the process parameters and the type of electrode used. From Figure 3a, it can be seen that with an increase in j at lower values of n, the energy consumption was maximum due to the lower solution conductivity and the increase in voltage drop. But with an increase in n at lower values of j, the energy consumption was constant. At lower j values from 2.38 to 4.76 mA/cm², the energy consumption was maximum with an increase inn. At higher j values with increasing n values from 0.5 to 1.0 g/L, the energy consumption was increasing. Further, increasing the n from 1.50 to 2.50 g/L decreased the energy consumption at higher j values because of ease in flow of current through the electrolytic solution.

3.3. Optimization

This study optimized the process parameters such as current density, time, and NaCl dose to get the maximum bacterial inactivation and minimum energy consumption by using MMO anodes. The optimum conditions for the two responses, Q₁ and Q₂, were not the same. Using the desirability function approach, the optimum conditions for Q₁ were selected as maximum, and for Q₂ as a minimum. Some constraints were applied to input process parameters to obtain this integer, as shown in Table S2.

The optimized values of the process parameters and the responses and the desirability values for individual and simultaneous optimization are shown in Table S3. The optimum conditions for the process variables were NaCl dose = 1.77 g/L, current density = 6.34 mA/cm², time = 10 min. At this optimized condition, the responses Q₁ and Q₂, as proposed by BBD, were 99.96% and 0.58 kWh/m³, respectively, along with a desirability value of D = 0.884. The experiment was performed under these optimized conditions to check the developed model’s efficacy. The responses, Q₁ and Q₂ at these optimized conditions, came out to be 99.2% and 0.42 kWh/m³, respectively. These values are very close to the ones predicted by the model, as shown in Table S4. During electrolysis, various ions like OH∙, HOCl, Cl₂, and Cl⁻ were generated, leading to direct and indirect oxidation.

Further, the OH∙ generated in the process helps in direct oxidation and readily gets converted to H₂O₂ and HO₂∙, thus leading to indirect oxidation. Further, the presence of RCS also helps in indirect oxidation. Hence, bacterial inactivation occurs due to direct and indirect oxidation at optimized conditions. Moreover, the obtained results have been critically analyzed with the reported studies as depicted in Table 3. The present study uses the novel composition of MMO anode (TiO₂/IrO₂/RuO₂/PtO₂) with low current density of 6.34 mA/cm² for the complete inactivation of bacterial consortium within a treatment time of 10 min. Thus, it is concluded that the technique is economically feasible at lower current density and treatment time.

3.4. Characterization of MMO anodes

3.4.1. SEM/EDS analysis

To characterize the surface morphology and elemental composition of MMO anodes, SEM-EDS was performed for both fresh and recycled anodes Figure 4a,b. The SEM images for both fresh and recycled anodes show almost similar structural characteristics even after 50 cycles, thus indicating the uniform layer of metal oxides. The incorporation of Ir, Ru, and Pt oxides in the titanium metal sheet has made the electrode surface porous and smooth with slight cracks, thus preventing the electrodes from corrosion and increasing the stability for a more extended period [60]. The pronounced peaks of all the three metals, i.e., Ir, Ru, and Pt in the recycled anode, along with the peaks of Ti, and O, have proved the durability of electrodes even after 50 cycles. The quantitative analysis of the electrodes was performed by EDS. It depicts that the atomic concentration of the metal oxides of Ti, Ru, Ir, Pt, and O were almost similar for both fresh and recycled anodes after 50 cycles. Such results confirm the stability of electrodes after 50 experimental runs as shown in Figure 4a,b.

3.4.2. XPS analysis

The XPS spectra of both fresh and recycled anodes Figure 5 was performed in order to analyze the metal elements’ oxidation state and to study the molecular information associated with electrode surface chemistry [61, 62]. The results indicate no drastic change in the oxidation state of the metals as well as each metal element present after 50 cycles. As shown in Figure 5a, in the Ti 2p spectrum, the two peaks at 458.5 eV and 464.3 eV correspond to the binding energies of Ti 2p_3∕2 and Ti 2p_1∕2 states for both the fresh and recycled anodes [63]. Figure 5e, shows the spectrum of O 1s signal at 531.47 eV, which corresponds to a broad peak due to the formation of Ti–O bonds. Figure 5b,c and d, shows the spectrum signal of Ir 4f, Ru 3d, and Pt 4f for both fresh and recycled anodes. Pt signal spectrum is indicated at 74.2 eV. It shows the signal of the Pt (v) oxidation state at the 4f_5∕2 level, which further corresponds to the formation of Pt–Ti bonds. However, for Ru, the peak was observed at 284.8 eV while for Ir, the peaks were observed at 61.6 eV and 63.2 eV. The signals obtained from the Ru and Ir spectrum were linked with Ru 3d_3∕2, and Ir 4f_7∕2, Ir 4f_5∕2, respectively, indicating the formation of hydrated metal oxides of RuO₂ and IrO₂. The studies reported in the previous literature correspond to the results in the present study [64, 65, 66].

3.5. Durability study

The durability and the stability of MMO anodes have been studied in order to check the economic cost and the practical feasibility of the EO process on a commercial scale. In particular, the durability of the electrodes has been evaluated in terms of the number of recycles and the inactivation efficiency of bacteria, as shown in Figure 6c. The electrodes were thoroughly used for 50 cycles without any significant loss of the metal oxides. Conditions for 50 cycles were according to the experimental runs as suggested in Table 1. The high activeness of the MMO anodes was due to the minimum amount of NaCl concentration used at lower current density values. Due to the presence of IrO₂ in the MMO anodes, it can be used for a maximum of 5 years. These MMO anodes can produce a magnificent amount of RCS and ROS due to TiO₂, IrO₂, and Pt, leading to direct and indirect oxidation [70]. The durability of MMO after 50 cycles was further confirmed through SEM-EDS, and XPS analysis, as discussed in Section 3.4.

3.6. TAC analysis (total available chlorine)

To check the quality of the untreated and treated simulated wastewater, analytical test such as TAC was performed under optimized conditions. From Figure 6a, it was observed that the TAC level increased during the EO treatment of simulated wastewater after a treatment time of 10 min. The TAC is a mixture of reactive intermediates comprising chloramines such as (NH₂Cl) and free chlorine (Cl₂, HOCl) [71]. Thus, it is concluded that the RCS produced during electrolysis using MMO anode is responsible for the inactivation of bacteria present in the simulated wastewater.

3.7. Bacterial cell damage checked by potassium ion leakage test

The outer membrane of the bacterial cell serves an essential function of barrier to the permeability of the intracellular substances. The bacterial cell damage was due to RCS, which damages the cell wall and the outer membrane of bacteria produced during the EO treatment of simulated wastewater. Thus, It leads to nd increased permeability and leakage of intracellular substances like K⁺ from the cell [11]. The leakage of K⁺ ions from the damaged bacterial cell was studied for 10 min. The K⁺ ion concentration increased from 1.63 ppm to 1.78 ppm Figure 6b. The K⁺ ion was released entirely from the inactivated bacterial cell after a treatment of 10 min. Thus, it is concluded that the membrane permeability of bacterial cells was disrupted during the EO process.

3.8. Operating cost analysis

To commercialize the EO treatment technology at large scale, it is mandatory to estimate the total cost of the treatment process, which should be economically feasible. Hence, the electrical energy consumed and the cost of the electrodes have been considered for the economic evaluation of the treatment process. The total operating cost for the inactivation of bacteria in simulated wastewater was determined to be 0.189 $/m³ (Table S5). Such value indicates that the EO process by MMO is economically feasible. Few researchers have reported the operating cost analysis for the treatment of different wastewaters [71, 72, 73, 74]. Further, the operating cost can be minimized during the scale-up studies by modifying the design of the electrolytic reactor and the operating conditions.