Enhanced desalination efficiency of flow-through capacitive deionization cell by mesh electrode with granular aerogel carbon in the removal of ions from synthetic and real samples

Only 3% of the total surface water resources are in the form of surface water. The results of various studies show more than three billion people in 150 countries use desalination processes due to insufficient access to drinking water resources (Panagopoulos & Haralambous 2020). In this case, desalination seems to be a promising alternative. Desalination is the application of several different steps in saline water to remove or reduce excess salts and minerals (Panagopoulos 2021). Now, common technologies, such as reverse osmosis (RO), nanofiltration (NO), and thermal distillation (multi-stage flash (MSF) and multi-effect distillation (MED) are used despite high-energy consumption and environmental problems to purify and desalinate saline water sources (Panagopoulos 2020).

Although, owing to their energy consumption, brine discharge, and greenhouse gases emissions, these methods are not proper alternatives for the desalination from saline solutions (Xing et al. 2020; Ihsanullah et al. 2021). In the past decades, electrochemical technologies have been introduced as a more reliable alternative to desalination of brackish water sources due to less energy consumption compared to conventional methods (Tang et al. 2020). Capacitive deionization (CDI) is one of these emerging technologies (Xing et al. 2020). This process has various advantages compared to conventional processes, such as greater safety, no need for high-pressure pump, working at low voltage (1.8 < V), no need for a chemical to regenerate, short system wash time, energy saving, cost-effective, and very environmentally friendly (Choi et al. 2019). The most important advantage of the system compared to other systems is its energy consumption efficiency which is, on average, 0.1 to 1.5 kWh/m³ of deionized water. Such advantages have led to the proposed method, among conventional methods, as a reliable alternative to the removal of soluble substances from brackish water (Pan et al. 2020).

The most common CDI cellular structure, which is known as flow-by CDI or flow-between CDI electrodes (Baroud & Giannelis 2018), consists of a pair of electrodes (anode and cathode electrodes). To prevent short-circuiting, this pair of electrodes includes an adsorbent and a non-conducting layer installed between two electrodes with opposite charge (Tang et al. 2016; Kalfa et al. 2020). The general principles of the CDI system consist of two stages of adsorption and desorption (AlMarzooqi et al. 2014). The solution containing ions is passed between the two carbon electrodes, and the positively charged ions are absorbed by the cathode electrode and the negatively charged ions are adsorbed by the anode electrode, Finally, ion-free water (pure) is passing out between the two electrodes (Biesheuvel & Van der Wal 2010; Chen et al. 2020). Once these electrodes are saturated, they can be regenerated by zeroing the voltage or reversing the voltage current, and releasing adsorbed ions into the brine (Wang & Lin 2019). Previous studies have focused on the development of electrode materials, increasing the adsorption capacity of electrodes, improving energy efficiency, and determining optimal CDI conditions (Kim et al. 2019; Wang et al. 2019; Zhao et al. 2020). The results of these studies led to the formation of various CDI cell structures (Tang et al. 2019). One of these structures is the flow-through electrode capacitive deionization (FTCDI) cell, where the solution passes directly through the macropores of the electrode, unlike other structures (Andelman 2011; Guyes et al. 2017), allowing more contact between the dissolved ions, the adsorbent and the electrodes (Tang et al. 2017; Son et al. 2020). This procedure increases the efficiency of the electrodes in the adsorption phase compared to CDI and MCDI (Remillard et al. 2018). On the other hand, this system, unlike CDI and MCDI, cannot use the voltage reversal method in the regeneration phase because of the simultaneous occurrence of the adsorption process (re-absorption of co-ions by the opposite electrode) and desorption. It can only use the zero voltage method, but this method reduces the capacity of the electrodes over time (Zhang et al. 2019a, 2019b; Algurainy & Call 2020). Moreover, the previous researches have focused on the removal of an ion from solutions. This created a study gap in using CDI systems for real samples (Algurainy & Call 2020; Pan et al. 2020). Therefore, a basic study is needed to investigate the removal of all ions from different aqueous solutions such as saline water, municipal wastewater, industrial wastewater, etc.

Based on the above, the overall purpose of this study was to investigate the simultaneous removal of several different ions in synthetic and real samples. In this study, a new FTCDI cell structure has been designed to enhance its performance. Moreover, the designed system can use the voltage reversal method to regenerate the electrodes. To achieve this goal, an ion-exchange membrane (IEM) was used under each electrode and a protective plate in each cell. Therefore, an enhanced batch-FTCDI system with a new design of carbon aerogel electrodes was used to study its ability in the removal of sodium and calcium chloride, nitrate and phosphate ions in synthetic samples at specified concentrations and tap water.

Each mesh electrode consists of two circular mesh plates with an inner diameter of 5 cm and an outer diameter of 5.5 cm made of stainless steel with pores with a diameter of 0.9 mm, and containing 1 g of aerogel carbon granule between the two mesh plates. Under each electrode was an IEM with a diameter of 5.5 cm (manufactured by Fujifilm, the Netherlands) (Figure 2(b)). After the electrode was made, the FTCDI cell was prepared. Each cell consists of two electrodes prepared in the previous steps and a non-conductive ring made of Plexiglas, which was located to prevent short circuit current (Figure 2(c)). The next step is to build an FTCDI pilot with a diameter of 5.5 cm and a height of 15 cm in a cylindrical shape. This system consists of three FTCDI cells (six electrodes), with a total surface area of 2.934 m² and aperture volume of 1.824 cm (Figure 2(d)). Other components of the experimental FTCDI pilot (Figure 3) include a desalination solution and concentrate storage tank and a voltage source. An electric conductivity meter measures the concentration of ions. A peristaltic pump, with a capacity of 5 L/h, was used to regulate and circulate water flow regulate.

As shown in Figure 3, an IEM was placed under each electrode. In other words, it was placed under the negative electrode of the anion exchange membrane (AEM) and below the positive electrode of the cation exchange membrane (CEM). These membranes cause more ions with opposite charges to be trapped in each electrode during the adsorption phase (increasing the adsorption efficiency compared to conventional FTCDI). The membranes allow only the ions with the same charge as electrode to pass through. For this reason, the adsorption efficiency in the FTCDI used is extended to the conventional FTCDI.

The reduction phase, in the FTCDI used, was done by reverse voltage method, in which the voltage reverses. In this system, the adsorption flow valves close and the washing flow opens at the same time. The co-ions of each electrode are released into the upper space and the presence of the shield plate prevents their re-adsorption with opposite charge electrodes, at which time the trapped co-ions are transferred to the brine tank by washing flow (Bian et al. 2015).

Morphology, composition, and surface functional groups of carbon aerogel electrodes made using electron microscopy (Sarrafi et al. 2016) (Hitachi, S-4800), Raman spectroscopy (Horiba Jobin Yvon, Labram HR), and infrared spectroscopy (FTIR, Bruker Vertex 70), were determined, respectively. In addition, the surface properties of the electrode materials were estimated using Brunauer–Emmett–Teller (BET) analysis using N₂ adsorption isotherms at 77 K (Tristar II Plus, Micrometrics). For this analysis, the samples were degassed using N₂ gas at 145 °C overnight before determining their concentration. The concentration of sodium and calcium chloride ions were determined by electric conductivity meter. Also, the concentration of nitrate and phosphate ions were determined according to the Standard Methods for the Examination of Water and Wastewater using codes 4,500 and 4,110, respectively (APHA 2012).

Field emission scanning electron microscope (FESEM) was applied to study the morphology of samples (Rasoulzadeh et al. 2021a, 2021b, 2021c). Figure 4(a) (I, II, and III) displays FESEM images of the pristine aerogel carbon beads. Figure 4(b) (IV, V and VI) displays FESEM images of exhausted aerogel carbon particles. Pristine aerogel carbon displays non-sphere morphology with porous structure. In addition, compared with pristine aerogel carbon, as clearly observed in Figure 4 (IV, V and VI). Based on Figure 4 (before and after using the process) and considering the erosion and deformation that it shows after use, it can be concluded that the process of adsorption of ions from the solution is well formed.

To demonstrate the elemental compositions of materials prior and post operation, EDX was used (Gholami et al. 2020; Rasoulzadeh et al. 2021a, 2021b, 2021c). Figure 4(b) (I and II) presents the EDX spectrum of the samples. As can be seen in Figure 5(b), the peaks observed in native aerogel carbon were, also, observed in utilized aerogel carbon except some peaks appeared at 0.28, 1.2, 2.1, 2.62, and 3.7 keV which are association with adsorbed ions.

Figure 4(c) shows the Raman spectra of this electrode material. This spectra displays presence of two strong bands at 1,450 cm⁻¹ and 1,500 cm⁻¹ which assign to the D and G bonds and belong to the disordered carbon atoms in its amorphous phase.

BET analysis and BJH isotherm using N₂ adsorption/desorption were used to determine the specific surface area and pore size of aerogel carbon nanoparticles (Alipour et al. 2021; Rasoulzadeh et al. 2021a, 2021b, 2021c). N₂-sorption isotherms of aerogel carbon are shown in Figure 4(d) (I and II). According to BET, the specific surface area and the pore volume of the sample were found to be 489 m²/g and 0.3041 cm³/g, respectively. Furthermore, its average particle diameter was obtained as 158.85 nm.

One of the most important parameters in the amount of adsorption of soluble ions by various CDI cellular structures is the electrical potential. The effect of the applied voltage on the electrosorption performance is illustrated in Figure 5. Different values of the applied voltage from 0, 0.8, 1 to 1.2 V were tested. As shown in Figure 5, the amount of sodium chloride absorbed increases by increasing voltage range from 0 to 1.2 V. When voltage is not applied (V = 0), only physical adsorption occurs in the system, so that the slope of physical adsorption decreases over a period of 10 min. After the saturation of surface pores, the electrodes will no longer be absorbed during this time and the system will reach saturation. However, by applying a voltage from 0.8 to 1.2 V, in addition to physical adsorption, electrochemical adsorption is also performed. The removal efficiency of sodium chloride (Equation (1)) using voltages of 0, 0.8, 1, and 1.2 V becomes 13, 33, 40.5, and 54.5%, respectively. This type of adsorption, electrochemical adsorption, will cause all the double electrode layers and the electrode pores to be involved in the adsorption process. Thus, the saturation time of the electrode pores will be longer and more ions will accumulate in the double electrode layers. In general, it can be concluded that the effect of voltage on the rate of sodium chloride adsorption is directly related to aerogel carbon electrodes. The study of Xing et al. (2019) shows that there is a direct relationship between the amount of potential difference applied to carbon electrodes and the rate of adsorption of accumulated sodium chloride ions. Hence, by applying voltages of 0.6, 0.9, and 1.2 V, the removal efficiency was 6.4, 18, and 32%, respectively (Xing et al. 2019). Also, Mossad & Zou (2013) reported that as the voltage increases, more salt concentrations from aqueous solutions are absorbed by the system, so that at voltages of 0.6, 0.8, 1, and 1.2 V, the removal efficiencies were 51, 53, 57, and 61%, respectively.

Depending on the cellular structure of each CDI system, the way the adsorbent is used is different. For example, in conventional flow by CDI and MCDI, the used adsorbent is attached to the electrode surface in the form of a layer with a certain thickness, or in the FCDI system, the adsorbent is used as a slurry in the chamber. In the present study, aerogel carbon used in the form of granules was placed between the two mesh plates of each electrode to allow the flow of solution to pass through the electrode (Figure 2(b)). The flow rates, contact time, and the probability of collision between carbon pores and soluble ions are effective parameters in increasing the efficiency of the system. In other words, the size of the granules is significantly related to them, so it is necessary that the effect of diameter changes on performance. Figure 7(a) and 7(b) show the effect of different diameters of aerogel carbon granules on the electrochemical adsorption process of sodium chloride and the removal efficiency of sodium chloride, respectively. As shown in Figure 7, there is an inverse relationship between granule diameter and system efficiency. In constant conditions, with increasing the diameter from 0.9, 1–2, and 2–3 mm, the removal efficiency decreased by 69, 52, and 46%, respectively. The reason for this decrease can be due to the increase in the diameter of the grains, because as the diameter increases, the density of the granules decreases and the distance between them increases. It is at this point that the ions pass more easily and with less contact with the granules and, finally, the efficiency of the system decreases with increasing diameter of the granules. To increase the efficiency, the diameter of the grains cannot be significantly reduced, because in these conditions, a drop in water pressure occurs. The findings of Bian et al. (2019) show that by reducing the diameter of carbon particles from 2–5 mm to 0.4–8 mm, salt adsorption efficiency increases.

The performance of a CDI system is affected by changes in the input concentration. The experiments were performed under optimal conditions. Removal efficiency and adsorption capacity for concentrations of 150, 300, and 450 mg/L were 75.5%, 82.5%, and 82.5%, respectively. Also, adsorption capacity for concentrations of 150, 300, and 450 mg/L were 15.45, 33, and 55.68 mg/m, respectively. Figure 8(a) and 8(b) show the effect of different concentrations of sodium chloride on the electrochemical adsorption process and the removal efficiency of sodium chloride in the presence of optimal voltage conditions, respectively. According to Figure 8(a), the results show that the removal efficiency increases with increasing concentration of the solution. By increasing salinity, the ionic resistance of the solution decreases and, as a result, the efficiency increases (Yeo & Choi 2013). As the concentration of ions in the solution increases, the conductivity of the solution increases and causes the electrode to access the ions with less electrical resistance. However, some studies have shown that the effluent concentration increases with increasing inlet concentration, due to the rapid saturation of the double electric layers (Huyskens et al. 2013; Jiang et al. 2018). The reason for such a discrepancy can be explained by the difference in the range of concentrations of the solution used. According to Figure 8(b), the ion adsorption rate decreases over time. The maximum amount of adsorption was observed in 20 min. After that, the adsorption process is almost stopped, which indicates the saturation time of the electrodes. Electrode placement can be described as the solution flow will cause the adsorbed ions to separate again from the electrode and enter the effluent, thus reducing system performance after saturation time. The study of Chang et al. (2020) shows that by increasing the input concentration from 250 to 1,000 mg/L, the removal efficiency increases. Tang et al. (2015) reported that with increasing input concentration, the removal efficiency increases so that the adsorption rate in the initial times is much higher than the final times. Wang et al. (2015) reported that system efficiency decreases with increasing input concentration.

Calcium chloride is one of the effective ions in the salinity of solutions, so it is necessary to know how much success the system can have in reducing its concentration. As shown in Figure 9, the maximum removal efficiency was 49% and the maximum adsorption capacity was 33.15 mg/g in 15–20 min. Then, the amount of salt absorbed decreases as the carbon pores become saturated. The slope of the adsorption efficiency up to 10 min indicated that the adsorption efficiency speed was much higher during this time. It can be justified in two ways. First, the concentration of the solution is still high during this period, according to the cases mentioned earlier, the electrical resistance is low. Second, the pores of the carbon electrodes are not yet fully saturated, thus, the adsorption rate will be faster than the remaining time. The optimal conditions for determining the adsorption efficiency of calcium chloride were similar to sodium chloride. It is observed that sodium chloride at a concentration of 450 mg/L, adsorption efficiency was 82.5% but for calcium chloride was 49%. Due to the use of various samples and high concentrations as well as the repetition of experiments, over time, the absorption or removal capacity of the system decreases compared to the initial time of the system. Hou & Huang (2013) reported that the amount of calcium chloride absorbed was less than sodium chloride. Hassanvand et al. (2018) reported that calcium chloride adsorption efficiencies are lower than sodium chloride.

The performance of the system was tested in electrochemical adsorption of nitrate under conditions of 2 V, 80 mL/min, and carbon diameter of 0.9–1 mm with different concentrations of 30, 50, and 70 mg/L. In this experiment, similar to in Section 3.2.4, the efficiency increases with increasing initial concentration. Figure 10(a) and 10(b) presents the effect of different concentrations of nitrate on the process of electrochemical adsorption and nitrate removal efficiency in the presence of optimal voltage conditions, respectively. According to Figure 10(a), the maximum removal efficiencies were 74, 79, and 85%, respectively. Also, the maximum adsorption capacity of each electrode was 3.15, 5.85, and 8.85 mg/g. The aim of this study was to reduce the nitrate concentration to below its standard level set by WHO and EPA (50 mg/L and 45 mg/L, respectively) in drinking water. Based on this study, it can be concluded that the designed system is able to reduce the nitrate concentration to the desired level. Depending on the concentration of nitrate in the water to be treated, any amount of soluble nitrate can be reduced to less than the standard by increasing the number of electrodes and thus increasing the adsorption capacity of the system. Uzun & Debik (2019) reported that the CDI system can reduce the concentration of soluble nitrate in water resources (river) at a lower cost than other methods (Uzun & Debik 2019). Tang et al. (2015) reported that under different conditions and with the presence of other cations and anions, CDI systems could be used to significantly reduce nitrate concentrations.

According to EPA standards, the maximum concentration of phosphate ions in drinking water is less than 0.03 mg/L. Based on this value, performance of the system was tested at different initial concentrations (1, 2, and 3 mg/L) and under constant conditions with a flow rate of 80 mL/min, voltage of 2 V, and carbon diameter of 0.9–1 mm. As can be seen from Figure 11, the system has a good performance in removing the system in 40 min. At concentrations of 1, 2, and 3 mg/L, the removal efficiencies were 77, 87, and 90%, respectively. Adsorption capacity, also, was 0.108, 0.225, and 0.39 mg/g, decreasing or increasing the phosphate concentration in solution depends on the pH. Therefore, its solution pH was adjusted close to the drinking water pH so that the system would not be too affected by temperature changes. Findings of a study by Bian et al. (2019) showed that the maximum removal efficiency of phosphate of 83% was reported after 7 h.

Ramhormoz city is one of the cities of Khuzestan province in Iran, which is located in the northeast of the province and situated in the Khuzestan plain at an altitude of 179 meters above sea level (Khavarian-Garmsir et al. 2019). The sources of drinking water supply in this city are rivers and deep wells. Unfortunately, due to the TDS and the high total hardness, its consumption is impossible for the general population of the city, so that the people spend a lot of money on their drinking water every year (Abtahi et al. 2015). Therefore, to determine the physicochemical characteristics of the city's water and also to evaluate the ability of the system to deionize this type of water resource, sampling was performed from three points, and the samples were collected and transferred to the water and wastewater laboratory. After determining the physicochemical characteristics (pH, TDS, turbidity, electrical conductivity, hardness, chlorine, nitrate, and phosphate) of the samples taken, process efficiency was evaluated to remove contaminants.

After determining the physicochemical characteristics of samples taken, a volume of 150 mL from each sample was injected into the system. In this stage, the conditions applied to the system were similar to the ones during the treatment of NaCl (flow rate: 80 mL/min, voltage: 1.2 V, carbon aerogels diameter: 0.9–1 mm).

To further reduce the salinity of water, this condition were selected. In this study, the final goal was to use this system in real conditions, thus, the voltage selected should have the least effect on causing Faradaic reactions and oxidation of electrodes. The voltage 1.2 V, also, achieved an acceptable efficiency in the removal of nitrate and phosphate.

The performance of a water treatment system in the face of water resources is different from synthetic samples that have only one type of ion. Because the interaction of other ions with each other in real samples may increase or decrease the efficiency of the system, the ultimate goal was to use the CDI system in the purification and desalination of real samples. In this study, the water resources of Ramhormoz city were used, because their water resources contain high salts. Table 1 summarizes the water features of this city. Based on Table 1, TDS, EC, hardness, chlorine, and turbidity are much higher than the standard. According to Table 2, it can be seen that the system has a removal efficiency of 54.4% nitrate, 66.6% phosphate, 36.5% chlorine, and 33.3% total hardness and finally reduced them to less than the standard level. However, in terms of reducing the concentration of TDS, EC, and turbidity, despite a decrease of 23.5, 39 and 25%, respectively (TDS absorption capacity 60 mg/g) the levels are still higher than the standard. Therefore, it was concluded that this pilot was not able to reduce TDS and EC to the desired level. The reason for such performance against high concentrations of TDS and EC can be attributed to the low number of electrodes. Owing to the limited adsorption capacity of the electrodes, more carbon and electrodes are needed to further reduce the concentration. In this case, with increasing the number of electrodes and, finally, the amount of carbon, the adsorption capacity of the system increases and its saturation time will be longer. Therefore, it can be expected that this system will be able to place a higher concentration of soluble solutes between its pores. Thanks to repeated experiments during the study (more than three months of continuous testing with different concentrations), the adsorption capacity of the electrodes has decreased compared to the starting time. Therefore, it is not possible to expect the EC and TDS to fall below the desired level. A study was conducted by Lee et al. (2019) for desalination of domestic wastewater. The results show that the MCDI pilot under conditions of 1.2 V and flow rate of 10 mL/min in contact with the synthetic sample of sodium chloride with an initial concentration of 1,200 μs/cm at the duration of 180 min decreased to less than 1.27 μs/cm (more than 99% efficiency). However, in contact with domestic wastewater containing TDS of 276.7 mg/L, EC 437 μs/cm (much less than the actual sample in the present study), turbidity 0.27 NTU, hardness 118.2 mg/L, nitrate 57 mg/L and phosphate 11.6 mg/L the efficiency reduced to 48.6 mg/L, 99.1 μs/cm, 0.20 NTU, 9.3 mg/L, 5.9 mg/L, and 8/7 mg/L, respectively. The reason for such a decrease in efficiency compared to the synthetic state is the existence of interactions of other ions, including calcium, magnesium, sodium, potassium, sulfate, carbonate, nitrate, and phosphate (Lee et al. 2019). Also, a study conducted by Jiang et al. (2019) to remove phosphate from domestic wastewater using a conventional MCDI system showed that only the maximum efficiency of 57% was reached under conditions with voltage of 1.2 V, flow rate of 24 mL/min, and pH = 7.2. If, as in the present study, this efficiency has reached 66.6%, the reason for such an increase in efficiency compared to the Jiang et al. (2019) study, can be described because of the modified system compared to the conventional MCDI and the type of sample used. On the other hand, the previous studies have shown that the transfer of different ions under the same conditions from the solution to the electrodes and their adsorption can be different. Ions with smaller hydrated radii move faster towards the electrodes. On the other hand, high ion capacity is also important to remove more ions (Liang et al. 2013). Four studies have emphasized that monovalent ions are more efficiently adsorbed than divalent ions (Johnson & Newman 1971; Gabelich et al. 2002; Xu et al. 2008; Martínez-Landeros et al. 2019). The reason for this result in those studies is that the pore size of their electrodes varies from micropores (<2 nm) to mesopores (2–50 nm). This distribution of pore sizes causes ions to have very different tendencies to be absorbed by an electrode (Seo et al. 2010). As a result, the reason for the removal of more monovalent ions than divalent ions is the fact that divalent ions have a hydrated diameter between 6 and 7 Å.

Monovalent ions have a considerably lower hydrated diameter (4 Å) compared to the multivalent ions (6–7 Å). Therefore, electrode pores are occupied with monovalent ions in a higher rate than multivalent ions. This leads to the lower adsorption of multivalent ions. However, in our study, different results were obtained, so that under the same conditions (80 mL/min, 1.2 V, and 1.9 mm), the removal efficiency of PO₄^3- ion compared to NO₃⁻ ion and nitrate ion compared to ion Cl⁻ is higher. This discrepancy can be seen in the size of the pores of the electrodes, since the carbons are concentrated in the form of granules in the electrode and a great deal of porosity is obtained. There is no competition between ions with low or high hydrated radius. Phosphate and nitrate tends to be more adsorbed by the electrodes due to their higher charge capacity. These results are consistent with studies by Fan et al. (2016), Kim et al. (2013), and Uzun & Debik (2019), because they also emphasized that the adsorption efficiency of nitrate, phosphate, and sulfate ions was higher than that of chloride and fluoride ions under constant conditions. Therefore, it can be concluded from the above that the removal of less salinity than the removal of phosphate and nitrate is somewhat justifiable.

The cost analysis, in terms of capacity adsorption and energy consumption of the FTCDI system, at working voltage of 1.2 V for different pollutants is shown in Table 3. Based on Table 3, the presence of sodium and calcium ions in the solution reduces the energy consumption of the FTCDI system because they have a higher adsorption capacity than nitrate and phosphate. The maximum consumption energy of FTCDI was nearly 276.92 Wh/g for nitrate which was 138 and 70 times sodium and calcium, respectively. It is concluded that the FTCDI tends to adsorb sodium and calcium more than nitrate. On the other hand, the FTCDI has high selectivity for sodium and calcium. Table 4 shows comparison of adsorption performance of different pollutants using aerogel carbon granules containing mesh electrode. Based on Table 4, the FTCDI shows the good efficiency in NaCl removal with reasonable consumption of energy.

The performance of the spent carbon aerogel regeneration experiment was studied using the FTCDI system with optimum result (voltage 1.2 V, carbon particle diameter between 0.9–1 mm, and flow rate 80 mL/min). After each adsorption test, the adsorption flow valves close and the flushing flow opens at the same time as the voltage is reversed. Then, the ions adsorbed were desorbed by injecting 0.24 mL/s distilled water (0 μs/cm) to the column for 5–10 min until the EC of distilled water attained the removed EC in the adsorption phase. Subsequently, the next adsorption experiment started. The mentioned cycle was repeated for ten series. The system removal efficiency for the first experiment was 92.1% and the tenth experiment showed a removal efficiency of 88.3%. Therefore, the obtained sorbents can lose up to ∼4% of their removal efficiency of the first cycle after ten experiments. On this basis, carbon aerogel was found to be a stable material as FTCDI electrode and can be reused multiple times with good performance. This confirms a significant balance between its affordability and its ecological impact.

CDI is an electrochemical technology which uses electric fields between two electrodes to temporarily remove ionic species from water. Excess ions in water and wastewater include a variety of salts (sodium, calcium chloride, etc.), nitrates, phosphates, sulfates, heavy metals, colloids, organic compounds, biomolecules, and a variety of bacteria and viruses. Other important ions can also be removed and placed in porous electrodes as long as these materials are stored in the pores of the used electrodes. Based on the results obtained and comparing them with other studies, the MCDI system has been modified, in which permeable mesh electrodes enable the solution current to pass through the electrode itself and ultimately increase the level of contact of ions with carbon and use the maximum capacity (2,934 m²/g). In synthetic solutions, sodium chloride (82.5%), calcium chloride (49%), nitrate (85%), and phosphate (90%) can be removed. Also, in contact with the real sample the following were reduced: turbidity (25%), total hardness (33.3%), phosphate (66.6%), nitrate (51.4%), electrical conductivity (29%), and total solids (11.8%). Therefore, if the general purpose of using this process is to remove high concentrations of ions from water sources, it is necessary that the number of electrodes and the amount of adsorbent for this system are designed and set up properly, then it is possible to remove target pollutants with high efficiency by consuming low energy. Such a system will be able to remove any concentration from solutions with very low energy consumption and high efficiency.

Based on the results of the findings of this study, these cases can also be considered for future studies:

• The synthesis of an adsorbent with low electrical resistance, high specific surface area, and oxidation resistance to enable the process to perform better and more efficiently.

• In the case of research in this field, first, determine the amount of adsorbent and the number of electrodes before starting the system by determining the amount of concentrations of contaminants under study and determining the area of the specific adsorbent surface. These items must be carefully checked before starting the system: absorbent dose required, number of electrodes used, concentration of target contaminant to be removed.

• Study the efficiency of the system in removing organic matter such as drugs, heavy metals, bacteria, and viruses.

• The operating time of the system and the adsorbent used, after which time the electrodes should be replaced, should be carefully calculated.

• Study of chemical and electrochemical reactions between adsorbent and adsorbed material.

• The possibility of exploitation at real scale and, finally, the possibility of its commercialization should be examined.

The authors thank the Environmental Health Engineering Department, School of Public Health & safety of Shahid Beheshti University of Medical Science, Tehran, Iran for financial support (IR.SBMU.PHNS.REC.1398.100) and providing instrumentation facilities.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

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