Selective Separation of Singly Charged Chloride and Dihydrogen Phosphate Anions by Electrobaromembrane Method with Nanoporous Membranes

The entrance of even a small amount of phosphorus compounds into natural waters leads to global problems that require the use of modern purification technologies. This paper presents the results of testing a hybrid electrobaromembrane (EBM) method for the selective separation of Cl− (always present in phosphorus-containing waters) and H2PO4− anions. Separated ions of the same charge sign move in an electric field through the pores of a nanoporous membrane to the corresponding electrode, while a commensurate counter-convective flow in the pores is created by a pressure drop across the membrane. It has been shown that EBM technology provides high fluxes of ions being separated across the membrane as well as a high selectivity coefficient compared to other membrane methods. During the processing of solution containing 0.05 M NaCl and 0.05 M NaH2PO4, the flux of phosphates through a track-etched membrane can reach 0.29 mol/(m2×h). Another possibility for separation is the EBM extraction of chlorides from the solution. Its flux can reach 0.40 mol/(m2×h) through the track-etched membrane and 0.33 mol/(m2×h) through a porous aluminum membrane. The separation efficiency can be very high by using both the porous anodic alumina membrane with positive fixed charges and the track-etched membrane with negative fixed charges due to the possibility of directing the fluxes of separated ions in opposite sides.


Introduction
The development of new approaches for the extraction of valuable components from aqueous solutions is an important applied problem. A number of recent papers have noted the high potential of membrane methods for the selective extraction of nutrients (biologically significant chemical elements) and inorganic ions (e.g., lithium, cobalt, and nickel) from waste and industrial waters, leachate solution, etc. [1][2][3][4], as well as for reagentless production/recovery of acids and bases [5]. The advantage of membrane methods compared to traditional reagent-based methods is low energy and resource consumption (almost no chemical reagents are required), as well as high environmental friendliness (no additional waste streams). An important fact is also the ability to recover the majority of water for reuse. This reduces the involvement of new water resources and allows creating water recycling systems in production processes [6,7].
Waste and industrial waters are characterized by a variety of chemical composition, which is mainly determined by the source of their entry into the environment. Cations example, when separating acetic acid and monochloroacetic acid anions, which differ slightly in mobility in an electric field [37].
Thus, the hybrid EBM method can be effectively used for the separation of singly charged ions of strong (e.g., Li + /K + or Li + /Na + ) and weak (acetic acid anions) electrolytes. However, as far as is known, no attempts have been made to apply this method for the separation of phosphate species. In this regard, the purpose of this work is to expand the scope of the EBM method and test it in the separation of Cl − and H 2 PO 4 − ions. The results are obtained using two types of membranes: a polyethylene terephthalate track-etched membrane and a porous anodic alumina membrane.

Materials and Methods
For a simplified representation of the wastewater composition, a ternary feed solution containing sodium salts of chlorides and phosphates is used. These anions are present in almost all wastewaters, especially in the wastewaters of agriculture and the livestock sector. The main characteristics of the feed solution components that affect the efficiency of EBM separation are given in Table 1. Table 1. Some characteristics of ions (at 25 • C) from the feed solution [43].

Ion
Symbol Diffusion Coefficient, 10 −9 m 2 /s Stokes Radius, Å The design of a four-chamber, flow-through laboratory electrodialysis cell was used to obtain separation parameters using the hybrid EBM method (Figure 1). A solution containing a mixture of 0.05 M NaCl and 0.05 M NaH 2 PO 4 (pH = 3.8-3.9) was pumped through the left-hand (I) and right-hand (II) chambers, separated by a porous membrane, at the same flow rate (90 mL/min), as it was performed in previous experimental work on the selective separation of Li + and K + [32].
This method was first used by Brewer et al. [34,35] in the separation of potassium isotopes. It was later adapted for membrane systems by Konturri et al. [36][37][38][39][40][41]. In recent works, when using polyethylene terephthalate track-etched membranes in EBM devices, it was possible to achieve impressive results in the separation of Li + /K + and Li + /Na + pairs [31,32,42]. The EBM method was also used to extract ions of weak electrolytes, for example, when separating acetic acid and monochloroacetic acid anions, which differ slightly in mobility in an electric field [37].
Thus, the hybrid EBM method can be effectively used for the separation of singly charged ions of strong (e.g., Li + /K + or Li + /Na + ) and weak (acetic acid anions) electrolytes. However, as far as is known, no attempts have been made to apply this method for the separation of phosphate species. In this regard, the purpose of this work is to expand the scope of the EBM method and test it in the separation of Cl − and H2PO4 − ions. The results are obtained using two types of membranes: a polyethylene terephthalate track-etched membrane and a porous anodic alumina membrane.

Materials and Methods
For a simplified representation of the wastewater composition, a ternary feed solution containing sodium salts of chlorides and phosphates is used. These anions are present in almost all wastewaters, especially in the wastewaters of agriculture and the livestock sector. The main characteristics of the feed solution components that affect the efficiency of EBM separation are given in Table 1. The design of a four-chamber, flow-through laboratory electrodialysis cell was used to obtain separation parameters using the hybrid EBM method (Figure 1). A solution containing a mixture of 0.05 M NaCl and 0.05 M NaH2PO4 (pH = 3.8-3.9) was pumped through the left-hand (I) and right-hand (II) chambers, separated by a porous membrane, at the same flow rate (90 mL/min), as it was performed in previous experimental work on the selective separation of Li + and K + [32]. In addition to the porous membrane, chambers I and II are formed using auxiliary cation-exchange (CEM) MK-40 heterogeneous membranes (JCC Shchekinoazot, In addition to the porous membrane, chambers I and II are formed using auxiliary cation-exchange (CEM) MK-40 heterogeneous membranes (JCC Shchekinoazot, Pervomayskiy, Russia). These membranes served to prevent the transfer of anions from the cathode chamber to chamber I and their exit from chamber II. A 0.1 M NaCl solution was pumped through the electrode chambers; polarizing platinum electrodes were used. On both sides of the porous membrane, the Luggin capillaries were installed to control the voltage across the membrane. A convective flow was created, directed from chamber II to chamber I, opposite to the migration of competing anions. This was achieved by increasing the pressure of the solution in the circuit passing through chamber II, using an automatic nitrogen dosing system. The composition of solutions in chambers I and II was monitored with a pH meter and a conductometer. The experiment on separation was repeated at least 5 times with the given parameters. Samples of the solution from chambers I and II were taken at the beginning and the end of the separation process to determine the concentration of separated ions using a Dionex ICS-3000 ion chromatograph with the conductometric detector (Thermo Fisher Scientific, Waltham, MA, USA).
Two types of membranes were used for separating Cl − and H 2 PO 4 − ions: a polyethylene terephthalate track-etched membrane manufactured and labeled as TEM #811 at the Joint Institute for Nuclear Research (Dubna, Russia) and an inorganic porous anodic alumina membrane (PAAM) manufactured at the Krasnoyarsk Scientific Center of the Siberian Branch of the Russian Academy of Sciences (Krasnoyarsk, Russia) ( Figure 2). Pervomayskiy, Russia). These membranes served to prevent the transfer of anions from the cathode chamber to chamber I and their exit from chamber II. A 0.1 M NaCl solution was pumped through the electrode chambers; polarizing platinum electrodes were used. On both sides of the porous membrane, the Luggin capillaries were installed to control the voltage across the membrane. A convective flow was created, directed from chamber II to chamber I, opposite to the migration of competing anions. This was achieved by increasing the pressure of the solution in the circuit passing through chamber II, using an automatic nitrogen dosing system. The composition of solutions in chambers I and II was monitored with a pH meter and a conductometer. The experiment on separation was repeated at least 5 times with the given parameters. Samples of the solution from chambers I and II were taken at the beginning and the end of the separation process to determine the concentration of separated ions using a Dionex ICS-3000 ion chromatograph with the conductometric detector (Thermo Fisher Scientific, Waltham, MA, USA).
Two types of membranes were used for separating Cl − and H2PO4 − ions: a polyethylene terephthalate track-etched membrane manufactured and labeled as TEM #811 at the Joint Institute for Nuclear Research (Dubna, Russia) and an inorganic porous anodic alumina membrane (PAAM) manufactured at the Krasnoyarsk Scientific Center of the Siberian Branch of the Russian Academy of Sciences (Krasnoyarsk, Russia) ( Figure 2). Both types of membranes have relatively close pore size (Table 2). However, in the case of PAAM, the specific number of pores and thickness determine its main difference from TEM #811. Alumina polyhydroxocomplexes [45] * Estimated by scanning electron microscopy, SEM.

Preparation of porous anodic alumina membranes
The aluminum foil of high purity (99.999%) with a thickness of 500 μm was used for the membrane preparation [46]. The foil was electrochemically polished in a solution of Both types of membranes have relatively close pore size (Table 2). However, in the case of PAAM, the specific number of pores and thickness determine its main difference from TEM #811. Table 2. Some characteristics of the porous membranes used in this study.

Preparation of porous anodic alumina membranes
The aluminum foil of high purity (99.999%) with a thickness of 500 µm was used for the membrane preparation [46]. The foil was electrochemically polished in a solution of 1. After that, the barrier layer was removed using a solution of 0.5 M H 3 PO 4 with electrochemical detection of the pore opening [47]. According to SEM images, the average pore diameter in the obtained membranes was 44 ± 2 nm.

Determination of membrane ion separation coefficient
The ion separation coefficient, S 1/2 , (also called the (specific) permselectivity coefficient between two counterions [48]) is used to quantify the selective transport of ions across a membrane. It is defined by the following equation [3,49,50]: where j k is the flux density of ions i through the membrane (where k = 1, 2 and corresponds to different types of ions of the same charge sign); c f k is the concentration of ions i in the feed solution; ∆c p k is the change in the concentration of ions k in the permeate (in the case of NF) or in the concentrate (in the case of ED); P k is the ion passage (use in pressure-driven membrane process), which is defined as P k = (c p k /c f k ) × 100%. Concentrations and fluxes must be expressed in the same units of the amount of substance (mols or equiv.).
To calculate the fluxes of ions k through the membrane, the following equation is used: where V is the volume of the processed solution or permeate (concentrate), s is the effective membrane surface area, and t is the duration of the experiment.

Results
As noted above, the polybasicity of phosphoric acid anions is the main problem for their effective extraction from wastewater by membrane methods. In the process of EBM separation, as in the case of pressure-driven membrane processes, the proportion of the predominant form of the orthophosphoric acid anion does not change with time. On the other hand, the pH of the feed solution can change when the limiting current density reached the auxiliary CEMs that form chambers I and II. This change in pH is caused by proton-transfer catalytic reactions between the fixed groups of the membrane and water molecules at the ion-exchange membrane/desalted solution interface ("water splitting" [29,51]). Water splitting takes place when the concentration of charge carriers in the electrolyte reaches close to zero value at this interface [52]. Upon reaching the limiting state on the auxiliary membranes, the pH value should decrease in chamber II and increase in chamber I. If the pH value decreases in chamber II, then the efficiency of EBM separation will be affected by the diffusion transfer of orthophosphoric acid molecules from chamber II to chamber I through the porous membrane. The appearance of additional current carriers (H + /OH − ions) in the EBM system will lead to a decrease in current efficiency due to the transfer of protons and hydroxide ions through the track membrane. In addition, the change in pH will lead to a change in the charge of phosphate acid species. In particular, an increase in pH will result in the appearance of HPO 4 2− anions. The performance of EBM separation would be reduced. In this regard, in each EBM system under study, preliminary tests to determine the limiting current on the auxiliary membranes were carried out. Then underlimiting and close to the limiting current densities were used in order to avoid the development of the described unwanted effects.
With regard to theoretical analysis, it has previously been shown that a simplified model can be used that allows for a better understanding of the principles of the method without claiming quantitative agreement [32]. This model does not provide as much detail as the countercurrent flow separation models based on the Nernst-Planck equations [33,38,53]. However, its simplicity makes it accessible to a wide range of researchers.
For the calculation of flux densities j k of competing counterions (k = 1, 2), the contributions of electromigration, diffusion, and convection (denoted below by subscripts migr, dif, and conv, respectively) were taken into account as follows: where i is the current density (in A/m 2 of the membrane surface), t k , c k , and z k are the effective transport number (dimensionless), concentration (in mol/m 3 of the pore solution), and charge number of ion k, v conv is the average convective velocity (in m/s), and γ is the surface porosity. The first term in Equations (3) and (4) describes the joint contribution of electromigration and diffusion. This is taken into account in the value of the effective transport number, t k . Thereby, t k is the fraction of electric charge carried by ion k under the action of electric current and diffusion (if any). The second term in Equations (3) and (4) describes the convection contribution. It does not depend on the set current, but is determined only by the electrolyte concentration in the membrane pores, the membrane porosity, and the flow rate of the electrolyte solution through the membrane. The velocity of pressure-driven flow through a pore of diameter d is described by the Hagen-Poiseuille equation: where ∆p is the pressure difference between chamber II and chamber I, L is the length of a pore, and η is the liquid viscosity. v conv is linked with the flow rate W of solution through the membrane with surface S: W = v conv Sγ.

Track-Etched Membrane
A track-etched membrane labeled as TEM #811 was used earlier for the separation of lithium and potassium by the EBM method [32]. A high separation selectivity was achieved due to a high ratio of the mobilities of these ions. The mobility of chlorides is approximately two times higher than the mobility of dihydrogen phosphate ions. Therefore, a high separation selectivity can be expected for this ion pair also. When a constant current density of 25 A/m 2 is set in the EBM system, the fluxes of competing ions through the membrane separating chambers I and II at a low pressure drop of 0.1 bar are determined by electromigration: Figure 3a). With an increase in pressure, the fluxes of both ions are slowed down by countercurrent convection and become negative with respect to the transport of ions in an electric field. When the pressure difference is close to 0.2 bar, the flux of Cl − ions is very small, and the oppositely directed migration and convection almost cancel each other out. At the same time, the flux of H 2 PO 4 − ions becomes negative and relatively large in magnitude: for this ion, convection prevails. However, this scheme of the experiment has limitations: It is more difficult to fix an exact pressure drop than to control the current. pressure, the fluxes of both ions are slowed down by countercurrent convection and be-come negative with respect to the transport of ions in an electric field. When the pressure difference is close to 0.2 bar, the flux of Cl − ions is very small, and the oppositely directed migration and convection almost cancel each other out. At the same time, the flux of H2PO4 − ions becomes negative and relatively large in magnitude: for this ion, convection prevails. However, this scheme of the experiment has limitations: It is more difficult to fix an exact pressure drop than to control the current.   Table 2.
When a constant pressure drop of 0.3 bar between chambers I and II is applied, and the current density is relatively low (e.g., 25 A/m 2 , Figure 3b), the convection transport of both chlorides and phosphates dominates over their electromigration transport. Their fluxes are negative, since the electromigration velocity is much lower than the convection velocity.  Table 2.
When a constant pressure drop of 0.3 bar between chambers I and II is applied, and the current density is relatively low (e.g., 25 A/m 2 , Figure 3b), the convection transport of both chlorides and phosphates dominates over their electromigration transport. Their fluxes are negative, since the electromigration velocity is much lower than the convection velocity.
As in the case of cations [3], there are various options for the selective separation of In the current range of 50 A/m 2 < i < 100 A/m 2 , the second option of optimal selective separation can be chosen. Here, the resulting flux of a less mobile ion is negative (controlled by convective transport), and the flux of a more mobile ion is positive (controlled by electromigration). In other words, it is possible to simultaneously enrich one solution with a more mobile ion and the second solution with a less mobile one. In this case, it makes no sense to evaluate the separation selectivity by the S Cl  . The flux of Cl − ions is controlled by electromigration from chamber I to chamber II. Nominally, the ion separation coefficient, S Cl − /H 2 PO 4 − , is −7.5, but since the separated ions are transported in different directions, the efficiency of such a process can hardly be overestimated. However, the value of the set electric current in the system is close to the limit of applicability, i.e., to the limiting current at the auxiliary membranes. In addition, the energy consumption in comparison with the first option of separation at 50 A/m 2 increases significantly (about 0.50 and 0.67 kWh/mol Cl − for the cell entirely).
The choice of one or another option of selective separation, in our opinion, depends on the concentration of the target and competing ions in the feed solution. It is less energy-consuming to remove the component, whose concentration is lower, from the solution. Probably, the first option will be more preferable for processing wastewater, since the concentration of H 2 PO 4 − is almost always much lower than Cl − . This option should also be preferable when other anions besides Cl − are present in wastewater, since the mobility of H 2 PO 4 − is usually the lowest of all anions in wastewater, with the exception of organic compounds.
The values of the transport numbers, t k , of competing anions were estimated by fitting the theory to the experimental data. As Equations (3)-(5) show, by changing t k , the theoretical straight lines are shifted up or down when treating the j k vs. ∆p data (at i = const) ( Figure 3a); the slope is determined by the known values entering the expression c k γd 2 /(ηL). When treating the j k vs. i data (at ∆p = const) (Figure 3b), the slope of the theoretical straight lines is changed. The pore diameter value of 32 nm determined by the Hagen-Poiseuille equation (Equation (5) = 0.18. A significant difference in the values of the ion transport numbers determined from two different lots of data is apparently associated with experimental errors, in particular, with a measurement error of ∆p. However, the averaged values of t k from these two lots are higher than the transport numbers in free solution (t Cl − = 0.36 and t H 2 PO − 4 = 0.16) determined from a simple relationship: t k = z 2 k D k c k / ∑ j=1,2,3 z 2 j D j c j . The average transport number of Na + found by fitting is about 0.26 which is essentially lower than its value in the free solution, t Na + = 0.48. Hydroxyl and carboxyl groups are formed on the surface of the polymer (polyethylene terephthalate) after etching the tracks of TEM#811 [54]. These fixed groups form a negative electrical charge of the external surface and the pore walls at pH ≈ 4 [55], at which the experiments were carried out. This charge has a density of about 0.3 µC/cm 2 according to Sabbatovskiy et al. [56]. Therefore, cations are concentrated near the pore walls. When an electric field is applied, electroosmosis occurs in the pores: The cations entrain the liquid along the pore walls in the direction from the anode to the cathode ( Figure 4). Hydroxyl and carboxyl groups are formed on the surface of the polymer (polyethylene terephthalate) after etching the tracks of TEM#811 [54]. These fixed groups form a negative electrical charge of the external surface and the pore walls at pH ≈ 4 [55], at which the experiments were carried out. This charge has a density of about 0.3 μC/cm 2 according to Sabbatovskiy et al. [56]. Therefore, cations are concentrated near the pore walls. When an electric field is applied, electroosmosis occurs in the pores: The cations entrain the liquid along the pore walls in the direction from the anode to the cathode (Figure 4). If no pressure difference is initially applied across the membrane, the fluid transported by electroosmosis will create excessive pressure in the cathode chamber and an induced pressure difference will appear. This induced pressure difference will cause the return flow in the opposite direction (from the cathode to the anode) in the central part of the pores (Figure 4). This return flow can entrain the anions and thus enhance their migration velocity, resulting in an apparent increase in their transport numbers. The effect is in a certain sense the opposite of the effect of inhibition of the forced convective transport of ions in a pore with charged walls described by Tang et al. [57]. This inhibition is due to electromigration caused by the streaming potential induced by the forced flow.

Anodic Alumina Membrane
The porous anodic alumina membrane (PAAM) is less permeable than TEM #811. Its measured hydraulic permeability is half that of TEM #811 (Table 2). In this regard, when separating Cl − and H2PO4 − ions using PAAM, the current value of 50 A/m 2 was fixed and the pressure drop was varied. Figure 5 shows that the varying pressure drop has almost no effect on the flux of chlorides through the PAAM. Obviously, the chloride flux will reach zero at a pressure If no pressure difference is initially applied across the membrane, the fluid transported by electroosmosis will create excessive pressure in the cathode chamber and an induced pressure difference will appear. This induced pressure difference will cause the return flow in the opposite direction (from the cathode to the anode) in the central part of the pores (Figure 4). This return flow can entrain the anions and thus enhance their migration velocity, resulting in an apparent increase in their transport numbers. The effect is in a certain sense the opposite of the effect of inhibition of the forced convective transport of ions in a pore with charged walls described by Tang et al. [57]. This inhibition is due to electromigration caused by the streaming potential induced by the forced flow.

Anodic Alumina Membrane
The porous anodic alumina membrane (PAAM) is less permeable than TEM #811. Its measured hydraulic permeability is half that of TEM #811 (Table 2). In this regard, when separating Cl − and H 2 PO 4 − ions using PAAM, the current value of 50 A/m 2 was fixed and the pressure drop was varied. Figure 5 shows that the varying pressure drop has almost no effect on the flux of chlorides through the PAAM. Obviously, the chloride flux will reach zero at a pressure drop outside the investigated range. However, it has been experimentally established that at a pressure drop above 0.5 bar, PAAM cannot be used, since the risk of its destruction increases significantly. Above 0.6 bar, the membrane cracks.  It is known from earlier works that the porous anodic alumina membrane should exhibit anion-exchange properties at pH ≈ 3.9 of the feed solution [58,59]. In this regard, higher k t  values were expected for both competing anions than were obtained by fitting It was found that at a constant current density of 50 A/m 2 and a pressure drop of 0.3 bar, the flux of H 2 PO 4 − ions is close to zero. Within the measurement error, j H 2 PO 4 − ≈ −0.005 mol/(m 2 ×h), while j Cl − ≈0.33 mol/(m 2 ×h). A high separation selectivity (S Cl − /H 2 PO 4 − = −66) can be achieved. However, the logical value of this coefficient tends to infinity, since the fluxes of separated ions are directed in the opposite direction, as in the case of using TEM #811 at currents of 75 and 100 A/m 2 .
It is known from earlier works that the porous anodic alumina membrane should exhibit anion-exchange properties at pH ≈ 3.9 of the feed solution [58,59]. In this regard, higher t k values were expected for both competing anions than were obtained by fitting ( t Cl − = 0.38 and t H 2 PO − 4 = 0.21), taking also into account the results for the TEM #811 membrane. The pore diameter value of 38 nm, determined by the Hagen-Poiseuille equation (Equation (5)) from the experimental hydraulic permeability, was used to fitting the transport numbers.
There are two reasons for these low transport numbers. First, although the pore walls are charged, the pore diameter is relatively large compared to the size of the separated ions. Hence, the transport numbers should be close to those in the free solution. Second, the stability of the PAAM properties during long-term operation is rather low. After the pretreatment of the membrane and pre-used with pressure to flush particles out of the etched channels, its hydraulic permeability is still not constant. Apparently, this causes a specific shape of the j Cl − vs. ∆p dependence, and this complicates the correct determination of the parameters for the separation of Cl − and H 2 PO 4 − ions. Under experimental conditions (pH = 3.8-3.9), the protonation of polymerized polyhydroxocomplexes of aluminum oxide, [Al 13 O 4 (OH) 24  Thus, the EBM separation method can be effectively used for the selective separation of Cl − and H2PO4 − singly charged anions. Despite the limitations associated with the allowable range of the currents, high selectivity can be achieved both when using the tracketched membrane and the membrane from porous alumina. However, in the latter case, there are additional limitations associated with the stability of the membrane characteristics.
Let us make a brief analysis of the obtained separation characteristics and compare them with similar characteristics found by other membrane methods. Table 3 presents the results from some recent papers on the selective recovery of phosphates in the presence of other anions using membrane technologies. Table 3. Comparison of recovery/rejections of phosphates using different membrane methods. Thus, the EBM separation method can be effectively used for the selective separation of Cl − and H 2 PO 4 − singly charged anions. Despite the limitations associated with the allowable range of the currents, high selectivity can be achieved both when using the tracketched membrane and the membrane from porous alumina. However, in the latter case, there are additional limitations associated with the stability of the membrane characteristics.
Let us make a brief analysis of the obtained separation characteristics and compare them with similar characteristics found by other membrane methods. Table 3 presents the results from some recent papers on the selective recovery of phosphates in the presence of other anions using membrane technologies.
It is known that selective electrodialysis or selectrodialysis (S-ED), using special-grade monovalent-ion-selective ion-exchange membranes, makes it possible to successfully separate monovalent and multivalent ions of the same charge sign [49,69]. Using conventional ED with monopolar single-layer ion-exchange membranes, it is possible to concentrate certain types of ions. Rotta et al. [63] reported that during an ED process, the concentration of H x PO 4 (3-x)− ions in the concentration chamber increased by about 10 times. However, the concentration of the competing SO 4 2− ions increased approximately by the same factor. The low selectivity of the used anion-exchange membrane for doubly charged SO 4 2− ions (S is explained by electrostatic interactions [70] and takes place only at low current densities [71]. The performance of the conventional ED depends on the applied current/voltage [64] and is limited due to the pH variation in the ED chambers. Bipolar membrane electrodialysis (BMED) typically also uses monopolar ion-exchange membranes for separation [65]. At the same time, the use of bipolar membranes makes it possible to generate H + and OH − ions without reagents and to obtain different products with relatively high productivity in separated chambers. For example, a reagentless pH shift allows selective extraction of N III and P V from the feed solution in the form of NH 4 + ions and H 2 PO 4 − ions. In this mode, the NH 3 was concentrated up to 16 g/L in the base solution [65].
S-ED uses monovalent-ion permselective IEMs, which help to solve the problem of separation of ions of the same charge sign [66]. Neosepta AMS, CMS, ACS and CIMS (Astom Corp., Shunan, Japan); Selemion ASV and CSO (AGC Engineering Co., Ltd., Chiba, Japan); Fumasep FAA, FKL and FKE (FuMA-Tech, Bietigheim-Bissingen, Germany) are among commercial membranes of this special grade. They pass singly charged ions but reject multiply charged ones. Table 3 presents our estimates of the selective separation parameters of HPO 4 2− , Cl − , and NO 3 − ions using a special-grade Neosepta ACS membrane [66]. transferred through this membrane is 2. Therefore, selectrodialysis can be used to enrich the phosphate proportion even in the presence of high chloride concentrations [64,67]. The presence of competing anions in the feed solution has little effect on the selectivity coefficient, but significantly increases the solution processing time to the same degree as the phosphate removal [67]. As noted in the introduction, commercial nanofiltration membranes can also be used to concentrate H x PO 4 (3-x)− ions due to the high retention rate [22,68]. Our estimates ( .3 × 10 −3 mol/(m 2 ×h)) [22]. The hybrid EBM method used in this work, as well as nanofiltration, makes it possible to separate both ions of the same charge and different charge values, in contrast to the above-mentioned electromembrane (ED) methods of selective separation [3]. The EBM method is intensively studied now. In earlier papers, the effectiveness of the method has been proven in the separation of binary mixtures of Li + /Na + , Li + /K + , and Li + /Ca 2+ ions for lithium extraction [38][39][40][41]. The researchers selected such separation parameters (current density and pressure drop) so that competing ions (Na + , K + , and Ca 2+ ) were transported through the membrane, while lithium ions remained in the feed solution. The fluxes of Na + , K + , and Ca 2+ were 0.28, 0.44, and 0.44 mol/(m 2 ×h), and the separation coefficients S Li + /M n+ were 0.35, 0.085, and 0.27, respectively. In the case of processing the Li + /Na + /K + ternary mixture, the separation efficiency decreased by 1.5-2 times [36].  In recent works, the ion separation coefficient for the Li + /K + pair can vary from 59 [32] to 150 [31,42]. When Li + /Na + ions are separated, the selective permeability coefficient is somewhat lower and reaches 30 [31]. The flux of lithium through the membrane under optimal conditions can be~0.5 mol/(m 2 ×h) [3]. If lithium remains in the feed solution and the competing K + ion is transferred through the membrane, as in the works of Konturri et al. [38][39][40][41], its flux (j K + ) can be up to 2.1 mol/(m 2 ×h) [3,32].
Of course, the EBM method has limitations, similar to all membrane methods. As discussed above, this is primarily due to the processes occurring at auxiliary ion-exchange membranes. However, the method demonstrates high performance for the separation of ionic components compared to other membrane methods (Table 3).
Generally, the application of membrane methods can significantly reduce or exclude the use of chemicals. Nowadays, in industry, ion separation is carried out using reagentbased technologies (hydrometallurgy). For example, by combining the EBM method with reverse osmosis, conventional and selective ED, it is possible to arrive at a technology for the reagentless extraction of valuable components [3].

Conclusions
In this work, the capability of the hybrid electrobaromembrane (EBM) method for the separation of singly charged Cl − and H 2 PO 4 − anions was studied. The EBM technology differs from other membrane methods where the separation occurs under the action of an electric field and pressure simultaneously. Separation selectivity is achieved due to the difference in the electrical mobility of the ions being separated. It has been shown that at least three options of separation of this pair are possible: to reduce the flux of the most mobile ion to zero, to reduce the flux of the less mobile ion to zero, or to organize the process so that the separated ions move through the porous membrane in different directions. In the latter case, the H 2 PO 4 − flux of about −0.055 mol/(m 2 ×h) and maximum Cl − flux of about 0.40 mol/(m 2 ×h) through a track-etched membrane can be expected at 0.3 bar and 100 A/m 2 . Dihydrogen phosphate ions can be removed from the solution at lower currents (50-55 A/m 2 ), when their resulting flux is controlled by convective transport (−0.23-0.29 mol/(m 2 ×h)), while the chloride flux through the membrane is close to zero. Efficient separation of these ions using a porous anodic alumina membrane is achieved with the same separation parameters (0.3 bar, 50 A/m 2 ). Under these conditions, chloride flux can be estimated as 0.33 mol/(m 2 ×h). The separation efficiency can be high when using both types of porous membranes due to the possibility of directing the fluxes of separated ions in opposite sides, which is unattainable when using other membrane methods.