Electrodialysis as a potential technology for 4-nitrophenol abatement from wastewater

4-Nitrophenol is a widely used emerging pollutant in various industries, including the production of agrochemicals, drugs, and synthetic dyes. Due to its potential environmental harmful effects, there is a need to study its reuse and removal from wastewater. This study used electrodialysis technology to separate 4-nitrophenol ions using a four-compartment stack. The effects of supporting electrolyte concentration, pH, voltages, and current density on the performance of electrodialysis for separating 4-nitrophenol were investigated. A high extraction percentage of 77% was achieved with low energy consumption (107 kWh kg−1) when high 4-nitrophenol flows and transport numbers were reached.


Introduction
For several decades, electrodialysis (ED) has been widely used for the isolation of chemical compounds dissolved in water, such as water desalination, acid concentrations, mineral recovery, and removal of harmful metallic ions (Danaeifar et al. 2023;Li et al. 2023;Lotfy et al. 2022;Martí-Calatayud et al. 2012;Pramanik et al. 2017;Tian et al. 2021), among others.In these processes, its relevance is attributed to both its removal efficiency and its main operative advantages, including compact installations and short treatment times (Chang et al. 2010;Lotfy et al. 2022).In environmental terms, it is considered advanced, since it allows the removal of organic ions by innocuous processes, carried out at room temperature and pressure and without the addition of harmful chemical compounds (Bernardes et al. 2014;Rajoria et al. 2022).
About a new promising application, the ED can be included as a technological step in novel integral processes known as minimal (MLD) or zero liquid discharge (ZLD), whose main purpose is the pollution control for environmental and human health protection (Panagopoulos and Giannika 2023).These systems address the elimination or minimization the brine discharge, from natural and industrial brine effluents, and recover valuable resources such as water, minerals, salts, and metals, among others.Technologies such as reverse osmosis (RO), forward osmosis (FO), brine concentrator (BC), and brine crystallizer (BCr), among others, also can be considered in these processes (Panagopoulos and Giannika 2022a).However, this must still overcome several drawbacks associated with the high energy consumption, emerging scale, and employment of fossil fuels as energy sources, among others (Panagopoulos and Giannika 2022b).In this context, the ED is an adequate choice to contemplate in these systems since it has been used industrially on a large scale for several decades.Also in the last decade, other novel ED applications have been tackled (Strathmann 2010), by coupling with some technologies, such as electrodeionization (Ortega et al. 2017; Van der Bruggen 2018), diffusion dialysis (Sun et al. 2020;Wang et al. 2021), reverse electrodialysis, bioreactors (Cheikh et al. 2013) or nanofiltration (Kim et al. 2013), among others.In this vein, José et al. (2021) reported a new ED application in pharmacology, i.e., the separation of ibuprofen from an ethyl esters/ibuprofen mixture, coming from the products of an enzymatic kinetic resolution of rac-ibuprofen.Consequently, using electrodialysis for the purification and concentration of drugs, herbicides, dyes, nutrients, and other emerging concern pollutants could be proposed as a novel alternative (Kamilya et al. 2022;Nguyen et al. 2019).
Regarding the separation of organic molecules by the ED method, Deng and coworkers (Wu et al. 2019) studied the phenol removal from wastewater by electro-electrodialysis (EED).These authors used a two-chamber cell with an AEM (anion exchange membrane) to separate the electrode compartments.In this process, the phenol was converted into phenoxide ions by pH adjustment, while Fehér et al. (2020) carried out the separation of phenylacetic acid from organic impurities by a two-step electrodialysis process, recovering 95% of the organic acid from an initial concentration of 4000 ppm.Li and coworkers also (Yang et al. 2021) studied the separation of halogenated organic compounds from inorganic halogens (trichlorophenol, trichloroacetic acid, and trichloroacetonitrile) (Bu et al. 2018;Zhang et al. 2018) and examined the transport of the organic halogen compounds and their retention in the ion exchange membrane (Yang et al. 2021).Between the water contaminants, the 4-nitrophenol (4-NP) is included in the US-EPA (U.S. Environmental Protection Agency) priority pollutant list (Ashok Kumar et al. 2021;Serrà et al. 2020), since it is a subproduct of several synthesis processes of pesticides, herbicides, insecticides (Abazari et al. 2019), drugs (Serrà et al. 2020), and synthetic dyes (Landge et al. 2021).
Recently, several reports focused on nitrophenol removal have been directed toward technologies of nanofiltration (Yahya et al. 2021), catalytic reduction by nanoparticles (NPs) (Gioria et al. 2020;Princy and Gopinath 2021), and advanced oxidation processes (Marghade et al. 2020;Truong et al. 2021), among others.However, these technologies require overcoming some challenges in the scale industry.
Catalytic reduction, with supported nanoparticles (NPs), must address the challenges related to their manipulation and subsequent separation from the treated media, as well as 4-NP recovery for future applications, while nanofiltration devices, operating within low-pressure ranges, require specially designed blend membranes of polymers (Yahya et al. 2021), into account the 4-NP molecular size to achieve high rejection capabilities (Sedighi et al. 2023).In the advanced oxidation process, which combines UV-H 2 O 2 and Fenton's reactions, the time required to achieve high degradation is remarkably short.However, the mineralization could result in incomplete reactions, leading to the production of ringopening products (Marghade et al. 2020).These drawbacks, combined with the challenges of separating or immobilizing the catalyst, are impediments that need to be overcome for the application of this technology shortly.
Considering the previously mentioned difficulties, electrodialysis could be an attractive technological alternative to removing 4-nitrophenol.However, an adequate choice of parameters is essential to achieve good performance.In this context, the experimental conditions related to ion transfer, flow rates, and conductivities, among others, must be studied in depth.In this vein, Roman et al. extensively studied the pH effect on the mechanisms of the adsorption and transport of organic micropollutants, and OMPs (pharmaceuticals, pesticides, personal care products, or agrochemicals) (Roman et al. 2020).They concluded that unbuffered conditions (low system pH) conducted better OMPs adsorption but negatively affected the molecule transport (Yang et al. 2021).On the other hand, the in-solution specie sizes rule up the ion mobility and the membrane ionic exchanges (Fehér et al. 2020;Liu et al. 2021;Yang et al. 2021).
In this sense, this research aims to study the separation performance of low-concentration solutions of 4-nitrophenol, by electrodialysis process employing an ED stack with four compartments.The 4-NP molecule is formed by an aromatic ring with a nitro group that at pH values higher than 7.15 (pKa) (Scheme 1) (Bhatt et al. 2021;Echaubard et al. 2020;Yu et al. 2009;Zhang et al. 2022) is at anionic form, and pH = 8 presents a dissociation degree of c.a. 87.5% of (Hasselbalch formula) (Fehér et al. 2020;Wu et al. 2019).The influence of the coexistence of different ions and the addition of NaOH to pH control, the voltage, and the current density were investigated.The final concentration of 4-nitrophenol in the dilute and concentrate, recovery percentage, energy consumption, and current efficiency were examined to obtain a better ED stack parameter set.

Cell configuration
The electrodialysis tests were performed in a stack with four compartments separated by three exchange membranes: cation-, anion-, and cation-, with an effective area of 9.6 cm 2 .The ED stack was made with polyethylene (HMWPE) and stainless steel 316 L sheets (9.6 cm 2 ) as electrodes.The cation/anion-exchange membranes, named CMX and AMX, were supplied by Eurodia Industrie SAS.The properties of these membranes are shown in Table S1 (Supplementary information).The internal volume of each compartment was 14.3 cm 3 , while the distance between consecutive parallel membranes was 15 mm.
The experiments were operated under constant current mode using a potential galvanostat (Agilent and own equipment) at room temperature for 6 h in batch mode by recycling each compartment solution (Scheme 2).The current densities adopted were 6.2, 7.3, and 10.4A m −2 .The supporting electrolytes were NaOH for the dilute (DC) and concentrate (CC), and Na 2 SO 4 , for electrode (E-Cs) compartments, respectively.In all ED experiments, the current density was kept below the limiting current density.Aliquots of 1.5 mL were collected from DC and CC at preestablished time intervals (0, 30, 60, 120, 180, 240, 300, and 360 min).The 4-NP concentration was monitored by .Table 1 summarizes the initial concentrations, current density, and nomenclatures adopted for each run.The ED experiments were named 6_50_100, 7_50_100, 6_250_100, 6_250_500, 6_70_500, and 10_70_500, where the first number is related to the current density, while the second and the third indicate the concentrations of NaOH (ppm) and Na 2 SO 4 (ppm), respectively (Table 1).In order to control the global electrical resistance, it was necessary to add gradual NaOH aliquots in the DC during the runs with low NaOH concentrations (50 and 70 ppm).However, in experiments with 250 ppm of NaOH with one-step addition, the global resistance was kept at low values.

Data analysis and main parameter definitions
The performance analysis of the electrodialysis process was made focusing on the following parameters: the transport number, the 4-NP flux, the extraction percent, the current efficiency, the energy consumption, and the apparent stack resistance.
The transport number of the 4-nitrophenol through the membrane t 4−NP − is calculated from Eq. ( 1): where charge of the 4-nitrophenol, F is Faraday constant, i is applied current density (A m −2 ), and J 4−NP − is 4-NP flux (mol m −2 s −1 ).
The 4-nitrophenol flux is shown in Eq. ( 2): where V is volume of the solution (L), t i is time of the experiment (s), C i 4−NP is 4-NP molarity at the time I, C 0 4−NP is 4-NP initial molarity, and S is membrane section (m 2 ).
The 4-NP extraction percentage, P ext,4−NP − is defined as follows: The current efficiency η is estimated by Eq. ( 4): (1) where U is the cell instant voltage and MW 4-NP is the molecular weight.
The cells' resistance is associated with the sum of several resistances as described by the following equation (Eq.6): Finally, the average value of the apparent overall stack resistance (Abou-Shady et al. 2012;Banasiak and Schäfer 2009;Luiz et al. 2018;Rottiers et al. 2016), R stack (Ω), can be given as follows: where U is the average voltage measured.In this work, the current was kept constant.

Results and discussion
For clarity, Scheme 2 illustrates the transport of the total ions in the ED stack.Figure 1 displays the 4-NP concentrations in the DC and CC at different times (0, 180, and 360 min) for the 6_70_500 and 10_70_500 runs.These runs showed significant 4-NP separation and an accurate mass balance.The results suggest that the 4-nitrophenol ion is not degraded or adsorbed by the AEM (Fehér et al. 2020).The 4-NP separation values achieved in the CC after 360 min in the 6_70_500 and 10_70_500 runs resulted in 38.5 and 40.0 ppm, respectively.On the other hand, the typical UV-vis spectra obtained from DC and CC during the process could be observed for the 10_70_500 run.The band intensities revealed the 4-NP concentration evolution in each compartment (Fig. 2).Several fundamental experimental parameters determine the performance of the ED process, such as the 4-NP transport flux through the AEM, the current efficiency, the energy consumption, and overall, ED electrical resistance.Consequently, the electrical conductivity, current density, and pH were monitored to set the best conditions for the 4-nitrophenol high-yield separation.

Mobility controlling parameters
The effects of the electrical current and the concentration of the chosen supporting electrolytes (SEs), Na 2 SO 4 (E-Cs), and NaOH (I-Cs), on the 4-NP transport in the solution and the AEM were studied.
The initial conductivity values in the DC are related to the ionic concentration and mobility of the different ionic species (Chandra et al. 2018;Villeneuve et al. 2019;Yuan et al. 2016).In all runs, the 4-NP concentration was 50 ppm, while that of NaOH was modified (50, 70, and 250 ppm).When 50 or 70 ppm of NaOH were employed, the gradual  1).In the E-Cs, the initial conductivity values correspond to the supporting electrolyte concentration (Na 2 SO 4 ) and their mobility (Villeneuve et al. 2019).On the other hand, in the C-C, the conductivity increased with time due to the formation of OH − and the migration of some cations from DC (Figs. 3a, 4a, and 5a), while a decrease was observed in the DC due to the migration of cations (Na + ) and anions (4-NP − and OH − ) to C-C and CC, respectively, causing the increment of electrical resistance in this compartment solution.It is appropriate to consider that in the CC and A-C, the conductivities always increase with time (not shown).Furthermore, in this particular experiment, the membrane resistances had a minimal impact due to the low concentration of the electrolyte overall (Chandra et al. 2018).
The voltage in the ED stack is directly associated with the applied current density, the apparent stack resistance, and the global energy consumption.Consequently, monitoring this parameter gives rise to information concerning transport efficiencies and the operative costs of the separation process (Figs.3b, 4b, and 5b).
The presence of H + and OH − species, as well as the dissociation grade of 4-nitrophenol (Fig. S1), in the dilute compartment, is significantly influenced by pH, thereby affecting the retention capacity of 4-NP (Yang et al. 2021).Figures 3c, 4c, and 5c show the pH drop on each run.

Run conditions: low concentration of supporting electrolytes in all compartments
In runs 6_50_100 and 7_50_100, low initial concentrations of the supporting electrolytes were employed: 100 ppm in E-Cs and 50 ppm in I-Cs (Table 1), and aliquots of NaOH were added in DC along the runs to control electrical resistance.
Comparing the initial conductivities in C-C and DC in the experiments with both current densities (Fig. 3a), similar values were observed, i.e., 200 μS cm −1 (6.2 and 7.3 A m −2 ).On the other hand, it could be noted that the conductivity in the dilute decreased to lower than 100 μS cm −1 values after about three hours in both experiments.
Nevertheless, due to the gradual NaOH addition in the DC, the voltage media values resulted in 54 and 61 V when used 6.2 and 7.3 A m −2 respectively (Fig. 3b).Under these conditions, the pH in the DC evolved between 10.5 and 7.0 (Fig. 3c) despite the NaOH aliquots addition (above 50 mg of total NaOH added).This response suggests that the NaOH addition was scarce to counteract the OH − migration to the concentrate but resulted in suitable to keep stable the global electrical resistance (Table 1).
For both runs, during the first hours, the solutions of all compartments have a similar contribution to the electrical resistance, and probably at the end of the runs, the DC solution resistance could become the dominant factor (Chandra et al. 2018).It also could be noted that such as in the C-C, the A-C controls the ionic transport during the first hours, since their conductivities were lower than the C-C ones.

Effect of the concentration of Na2SO4 (E-Cs) on the ionic transport
Run conditions: high NaOH concentrations In the tests called 6_250_100 and 6_250_500, 62.5 mg (250 ppm) of NaOH was added by one step to the internal compartments (Table 1), while the current density was 6.2 A m −2 .The initial conductivities of the electrode solutions depended on the Na 2 SO 4 concentration; thus, when 100 and 500 ppm were employed, 160 and 900 μS cm −1 were registered, respectively.On the other hand, the range of NaOH conductivities in DC was from ca. 1350 and ca. 100 μS cm −1 (Fig. 4a).Consequently, the high conductivities favored the ionic transport in the DC during the first hours.These results suggest that the dominant resistance along the first hours (1 or 3 h according to the experiment) could be associated with the cathode solution resistance, where the conductivities were lower than those of DCs.However, it is important to consider that the 4-nitrophenol transport, during the first hours, could have a low contribution due to the high concentration of supporting electrolytes.
The voltage response under these conditions (Fig. 4b) initially decreased from 43 to 13 V for both runs and continued to increase to 47 V and 34 V, respectively.When Fig. 5 Effect of applied current density on the ED process.Run conditions: high Na 2 SO 4 concentrations (500 ppm) 100 ppm of Na 2 SO 4 was employed in the C-C, the minimum was achieved at 126 min (23.2V), while for 500 ppm of the supporting electrolyte concentration, the minimum voltage was 11.2 V at 62 min.In these cases, firstly, a high resistance associated with the C-C solution was observed, and next, the DC solution resistance became dominating, where the ions were removed continuously (Wang et al. 2022;Wei et al. 2021).Zhang and coworkers (Wen et al. 2022) found a similar behavior with a minimum voltage (and electrical resistance) in the middle time of the experiment, where both the feed and product solutions have high conductivities.
These results agree with Fig. 4a, in which it could be observed that the conductivities lines in DC and C-C resulted equal at approximately 60 min and 160 min when 500 and 100 ppm Na 2 SO 4 were used; respectively.Note that the conductivities decrease in the DC and the other (C-C) increase with time.However, in these runs without NaOH additions, the pH evolution in the DC exhibited a continued decrease from 11.4 to 8.5 (Fig. 4c).The observed drop in OH − concentration agrees with the significant decrease in dilute conductivities (Fig. 3b) and the increase in apparent stack resistance over time (Fig. 4b).This trend indicates that the dominant factor influencing ionic transport resistance during the later hours of operation was the DC solution (Bernardes et al. 2014).

Effect of applied current density on the ED process
Run conditions: high Na 2 SO 4 concentrations In these tests, 6_70_500 and 10_70_500, the employed supporting electrolyte concentrations were 500 ppm of Na 2 SO 4 (E-Cs) and 70 ppm of NaOH (I-Cs), with the gradual addition of aliquots in the DC (Table 1).
Figure 5a displays the conductivities evolution with time.In the beginning, the conductivities in the C-C were nearby 900 μS cm −1 for both runs, resulting in 1800 and 2500 μS cm −1 at the end when 6.2 and 10.4A m −2 were applied, respectively.It can be observed that the main resistance could be associated with the DC solutions since always the conductivity in the C-Cs was higher than in the dilutes, while the voltages were kept around a median value of 29 and 55 V, respectively, due to the gradual addition of NaOH aliquots (Fig. 5b).
In these runs, a slight decrease of the pH in the DC was observed (Fig. 5c), suggesting that the OH − ions appeared and migrated from the DC to the CC at very similar rates.Under these conditions, the predominant contribution of the stack resistance could be associated with the DC solution (Fig. 5a), and the pH ranges were between 11.2 and 9.8.These results agree with those reported by Verliefde and coworkers (Roman et al. 2020), who improved the transport of various organic ions using buffered solutions.Note that when 10.4 A m −2 was applied, a higher amount of NaOH (103.8 mg, Table 1) was necessary to keep buffered conditions (pH 11-10).

Global analysis of electrodialysis performance
From the main results emerging from Figs. 3, 4 and 5, a global analysis could be done as follows.
The conductivities observed were directly related to the overall resistance of ionic transport (Figs. 3a,4a,and 5a).A similar contribution to the electrical resistance was observed in each compartment solution when a low concentration of Na 2 SO 4 (100 ppm) or a high concentration of NaOH (250 ppm) was employed in the solutions of the external and internal compartments, respectively (Fig. 3a  and 4a).In these cases, for the first hours, the dominant contribution was associated with the resistances of electrode compartment solutions and finally with those of DCs.However, when a high concentration of ES (500 ppm) in E-Cs and a low concentration of NaOH (70 ppm) were used, the DC solution resistance resulted in the dominant ionic transport in concordance with the low registered conductivities (Fig. 5a).
On the other hand, different evolutions of the registered parameters were observed (Figs.3b and 5b).The average voltage in the 6_50_100 and 6_70_500 runs (Figs.3b, 4b,  and 5b) were 54 and 29, respectively.This fact could be associated with the higher concentration of Na 2 SO 4 in the E-Cs (experiment 6_70_500), which conduced to an apparent overall resistance drop (Table 2).
Regarding the pH evolution in each experimental condition, different responses were observed.A significant pH variation was detected when the ion transport was governed by comparable resistances in each solution (Figs. 3c  and 4c).However, if the transport control was related to the DCs, a slight pH drop was observed (Fig. 5c).For the last run, 10_70_500, the major OH − addition (103.8 mg, Table 1) is related to the higher OH − transport compared with the other runs.

4-NP transport
Concerning the 4-nitrophenol separation, Fig. 6 shows the 4-NP concentration evolution in the CC in each run, while Fig. 7 reports the transport numbers and fluxes.
When a low concentration was employed, the modification of the current density did not influence the 4-NP concentrations (Fig. 6).Therefore, the current increment led to a depletion in the process yield.During the first 2 h, the 4-NP flux in the 6_50_100 run was lower (1.4 × 10 −6 mol m −2 s −1 ) than in the 7_50_100 run (2 × 10 −6 mol m −2 s −1 ) (Fig. 7), probably because of the lower transport contribution observed in the first run ( t 4−NP − = 0.022).It could be noted that the stack resistances (Fig. 3) resulted similarly in both cases (9.1 and 8.8 × 10 3 Ω), although the different applied current densities (6.2 and 7.3 A m −2 ) (Table 2).
In the runs with one-step NaOH addition, the evolution of the 4-NP concentration exhibited the lowest rates (Fig. 6).These results could be associated with the low contribution of the 4-NP transport; their transport number ranges were between 0.011 and 0.035 and 0.014 and 0.037 when 100 and 500 ppm of Na 2 SO 4 were employed, respectively.Therefore, the 4-NP fluxes were also low (Fig. 7).Clearly, the lowest extraction percent was obtained in these experiments (c.a.55%, see Table 2).Wu et al. (2019) studied the phenol removal (100 ppm) by EED, employing Na 2 SO 4 (500-2500 ppm) as a supporting electrolyte in the dilute chamber.They reported that an increase in sodium sulfate produced two effects: firstly, an increase in the current densities (constant voltage), and secondly, a decrease in the rate of phenoxide ion migration, since all negatively charged ions in this compartment shared the electric current.This phenomenon could be associated Figure 6 shows that the highest values of 4-NP concentrations in the CC (at the end) were obtained in 6_70_500 and 10_70_500 runs, being very similar to each other (Fig. 6).Consequently, increasing the current density does not improve the 4-NP separation; therefore, using a highcurrent density (10.4A m −2 ), the process yield decreased, even though the total flux grew with the current.These results agree with the report in Fig. 7, where it could be observed that the transport numbers were between 0.047 to 0.055 and 0.025 to 0.032 when the chosen current densities were 6.2 and 10.4A m −2 , respectively, leading to higher 4-NP fluxes for the 6_70_500 run in comparison with the 10_70_500 run.The explanation of this behavior could be supported by the major addition of NaOH required (103.8 mg, see Table 1) to keep the voltage when the current density was 10.4 A m −2 , which induced higher OH − fluxes in comparison.The fluxes were ranked from 3 to 3.5 × 10 −6 and 2.7-3.5 × 10 −6 mol m −2 s −1 , respectively.It could be due to the DC solution being the main resistance during both complete experiments because in these tests, an ES high concentration (Na 2 SO 4 , 500 ppm) was employed for E-Cs.
On the other hand, it is important to consider that the highest transport numbers obtained, in these runs, leading to high extraction per cents, 77 and 82%, were the values registered after 6 h, for 6_70_500 and 10_70_500, respectively (Table 2).
Finally, comparing the runs of 50_100 and 70_500, where the current densities were varied, lower fluxes were observed for the first cases (1.4-2.6 × 10 −6 mol m −2 s −1 ) in comparison with the last ones (2.7-3.5 × 10 −6 mol m −2 s −1 ), probably because of the higher apparent stack resistance compared with the last ones (9 and 5 × 10 3 Ω, respectively) (Table 2).An increment in the overall resistance leads to higher costs because a greater permeation area or voltage was required to achieve the same fluxes (Ledingham et al. 2022).

Current efficiency and energy consumption
The current efficiency (η) is related to the number transport since it involves the 4-NP flux and the total ionic flux.The energy consumption (E) is the ratio between the energy employed in the 4-NP separation and the total energy consumed.On the other hand, E is directly related to the average voltage and the electrical current, while it is inversely proportional to the 4-NP quantity transported (kg).Consequently, for the highest transport numbers of 4-NP, the current efficiency improves, and at the same time, if the voltage and the electrical current remained at low values, the energy consumption decrease.Thus, an improvement in the process's global efficiency could be observed.The current efficiency and energy consumption calculated at the end of each experiment are shown in Fig. 8.All runs display a range of η between 3.2% and 4.0%, which agrees with the lower values of 4-NP transport numbers (Fig. 7) in comparison with the 6_70_500 run (η = 5.1%).Instead, the energy consumption range was more fluctuating (84-346 kWh kg −1 ), because of the different combinations of 4-nitrophenol extraction percentages: media voltages, and electrical currents.For the 6_70_500 and 7_50_100 runs, the voltages were high, and the extraction percent was media, so their energy consumptions were relatively high, 262 and 346 kWh kg −1 , respectively (Fig. 7).For 6_250_100 and 6_250_500 runs, the energy consumptions were low.However, the low 4-NP transport number did not favor the current efficiency.During the 10_70_500 run, the energy consumption resulted in 330 kWh kg −1 , and the electrical current keeps at the highest value (10 mA) with a media stack resistance of 5.5 × 10 3 Ω.Deng and coworkers (Wu et al. 2019) reported energy consumption in similar ranges.They found about 230 kWh kg −1 for phenol removal (100 ppm) by EED with Na 2 SO 4 (500 ppm) as a supporting electrolyte, in the dilute chamber, with 9 V applied voltage, and the estimated current efficiency was about 1.3%.The authors described that the phenol concentration only had a slight influence on the solution conductivity, but it affected the electrical current.
Instead, Fehér et al. (2020) obtained low energy consumption (6.9 kWh kg −1 ) by employing a two steps process (S1 and S2) with high concentrations of phenolic acid (4000 ppm, S1), buffer solution (0.1 M, S1), and Na 2 SO 4 (SE,18,300 ppm,S2).With this method, they achieved Fig. 8 Current efficiency and energy consumption at the end of the experiments for all runs displacement of the phenol ions by SO 4 −2 ions, which were previously adsorbed in the site actives of the anion exchange membrane.They reported an electrical efficiency of 28.5%, employing five-unit repeated cells, which means an efficiency of 5.7% per unit cell.The low energy consumption could be associated with the high phenolate migration obtained in these experimental conditions, due to the high concentrations and therefore high conductivities registered (c.a.12,500 μS cm −1 ) in the DC, at the beginning of the experiment.
Figure 8 shows that the best combination of both parameters (η and E) was found in the 6_70_500 run, in which the highest transport numbers for 4-NP were obtained, together with the lowest electrical current requirement and minimum average voltage.This experiment showed that η and E resulted in 5.1% and 107 kWh kg −1 , respectively.
The results discussed in previous paragraphs (Sect.3) demonstrated that the parameters such as electrical current and average voltages, transport contribution, and others associated with those affect the performance of ED experiments.While, when the current increase, the fluxes and the extraction percentages also increase, together with the energy consumption.On the other hand, OH − concentration is another important parameter.Consequently, continued monitoring and control are essential for improving electrodialysis separation.

Energy analysis of 4-nitrophenol removal under the different used operational conditions
The energy analysis of 4-nitrophenol removal under different conditions revealed the following trends: When pH (DC) was plotted against extraction per cent (Fig. 9a), it became evident that when the pH decreased below 8.5, the 4-NP removal remained limited to 54-64% after 6 h.However, in the pH range of 11.3-9.8, the 4-NP extraction percentages reached higher values (77-82%) within the same period.
Upon examining the energy consumption per unit mass of 4-NP transport versus extraction per cent (Fig. 9b), it was observed that employing a high concentration of Na 2 SO 4 and an applied current density of 6 A m −2 (6_70_500 and 6_250_500) resulted in low energy consumption values.The ranges fell between 151 and 107 kWh kg −1 and 151-84 kWh kg −1 for 6_70_500 and 6_250_500, respectively.
In these conditions, the control of ionic transport could be associated with the dilute compartment.Notably, the 6_70_250 run achieved a higher 4-NP extraction of 77% (6 h) compared to 6_250_500 (55%).In the remaining runs, energy consumption was high, ranging between 341 and 793 kWh kg −1 at the beginning and 158-346 kWh kg −1 at the end of the experiments (Fig. 9b).
The electrical efficiency was also analyzed concerning 4-NP removal (Fig. 9c).The highest values were observed at a 4-NP extraction per cent of 41% for the 6_70_500 condition.For this condition, the energy consumption reached a minimum of 102 kWh kg −1 at a pH of 10.3.In the 10_70_500 run, the current efficiency ranged between 2.1% and 3.2%, achieving an extraction percentage of 82%.However, increasing the current density did not improve 4-NP transport and significantly increased energy consumption (481-330 kWh kg −1 ).

Conclusions
The stack ED configuration with four compartments is adequate for the transport of 4-NP in low concentrations.The ED technology allows the purification and removal of this EC, with high extraction per cent, by a process environmentally friendly and economically feasible.The study investigated the effects of various factors on the performance of electrodialysis in separating 4-nitrophenol.
During the runs with one-step NaOH (500 ppm) addition in the internal compartments, a change in the dominant resistance was observed.Initially, a high resistance was associated with the C-C solution, and subsequently, the DC solution resistance became dominant.
The NaOH gradual addition was necessary to keep constant the electrical resistance.However, its concentration must be controlled to avoid excessively high fluxes and, therefore, low contributions of 4-NP in the total ionic transport, inducing small 4-NP transport numbers.The highest extraction percent was achieved in alkaline media between 11 and 10.
The concentration of electrolyte used in the E-Cs, Na 2 SO 4 , notoriously affects the global resistance.High values (500 ppm) were required to guarantee lower resistance in the electrode solutions compared with the DC solution during the entire test.When 100 ppm was employed, all resistances have a comparable contribution over the ion transport.In the first hours, the predominant resistance to the ionic transport was associated with the E-Cs solutions.
When the controlling step for the ionic transport was associated with the DC, employing a high concentration of Na 2 SO 4 (500 ppm) and low concentrations of NaOH (gradual addition) the extraction percentages rose.The highest extraction percentages (> 75%) were achieved with high 4-NP transport numbers.
In this proposed system, for higher 4-NP extraction, achieving high ED energy efficiency required low initial concentrations of NaOH with gradual additions in the DC compartments and high concentrations of Na 2 SO 4 in the external compartments when applying low current density.After 3 h, the highest 4-NP extraction (41%) with high electrical efficiency (5.5%) and low energy consumption (102 kWh kg −1 ) was achieved at a pH of approximately 10.3, an applied current density of 6 A m −2 , and 500 ppm Na 2 SO 4 .
After achieving these promising results, the future aim is to contribute to the advancement of ED by separating different emerging concern contaminants in real conditions.Alternatively, the use of bipolar membranes is also a future aim.This design would enhance the control of OH − concentration and improve the overall ED performance.Additionally, it would reduce the necessary concentration of supporting electrolytes and could even eliminate the need for them.However, the main drawback of employing bipolar membranes is their cost.
Scheme 1The 4-nitrophenol forms for different pH media

anode solution Scheme 2
Scheme 2 Schematic representation of the ion transport in the ED stack operated in batch mode

Fig. 1
Fig. 1 Evolution 4-NP in the dilute and concentrate chambers for the 70-500 runs

Fig. 2
Fig. 2 Absorption spectra of 4-nitrophenol in alkaline media at different times for the 10_70_500 run

Fig. 6
Fig. 6 Evolution of 4NP concentration in the concentrate for all runs

Fig. 9
Fig. 9 Global analysis of the effect of pH (DC) (a), energy consumption (b), and current efficiency (c) on 4-NP extraction percentage