Biopolymeric Membrane Enriched with Chitosan and Silver for Metallic Ions Removal

The present paper synthesized, characterized, and evaluated the performance of the novel biopolymeric membrane enriched with cellulose acetate and chitosan (CHI)-silver (Ag) ions in order to remove iron ion from the synthetic wastewater using a new electrodialysis system. The prepared membranes were characterized by Fourier Transforms Infrared Spectroscopy-Attenuated Total Reflection (FTIR-ATR), Thermal Gravimetric Analysis (TGA) and Differential Thermal Analysis (DSC), contact angle measurements, microscopy studies, and electrochemical impedance spectroscopy (EIS). The electrodialysis experiments were performed at the different applied voltages (5, 10, and 15 V) for one hour, at room temperature. The treatment rate (TE) of iron ions, current efficiency (IE), and energy consumption (Wc) were calculated. FTIR-ATR spectra evidenced that incorporation of CHI-Ag ions into the polymer mixture led to a polymer-metal ion complex formation within the membrane. The TGA-DSC analysis for the obtained biopolymeric membranes showed excellent thermal stability (>350 °C). The contact angle measurements demonstrated the hydrophobic character of the polymeric membrane and a decrease of it by CHI-Ag adding. The EIS results indicated that the silver ions induced a higher ionic electrical conductivity. The highest value of the iron ions treatment rate (>60%) was obtained for the biopolymeric membrane with CHI-Ag ions at applied voltage of 15 V.


Fourier Transforms Infrared Spectroscopy-Attenuated Total Reflection (FTIR-ATR)
Infrared spectroscopy (IR) study was performed on Perkin Elmer Spectrum 100 FTIR spectrophotometer (PerkinELMER, Ltd, London, United Kingdom) in the attenuated total reflection mode (ATR). The FTIR-ATR spectra between 4000-600 cm −1 were recorded by collecting 16 scans for each point, at a spectral resolution of 4 cm −1 .

Contact Angle and Surface Free Energy (SFE)
The wetting behavior of the obtained membranes was evaluated by contact angle measurements using a Contact Angle Meter (Kyowa Surface Chemistry Co., Ltd., Tokyo, Japan), KSV instruments CAM 100 equipment (KSV Instruments, Helsinki, Finland). For each membrane, the procedure was repeated three times at room temperature, with a constant volume of distilled water (10 µL). The static contact angle formed between membrane and different liquids drop (water, dimethyl sulfoxide,

Fourier Transforms Infrared Spectroscopy-Attenuated Total Reflection (FTIR-ATR)
Infrared spectroscopy (IR) study was performed on Perkin Elmer Spectrum 100 FTIR spectrophotometer (PerkinELMER, Ltd, London, United Kingdom) in the attenuated total reflection mode (ATR). The FTIR-ATR spectra between 4000-600 cm −1 were recorded by collecting 16 scans for each point, at a spectral resolution of 4 cm −1 .

Contact Angle and Surface Free Energy (SFE)
The wetting behavior of the obtained membranes was evaluated by contact angle measurements using a Contact Angle Meter (Kyowa Surface Chemistry Co., Ltd., Tokyo, Japan), KSV instruments CAM 100 equipment (KSV Instruments, Helsinki, Finland). For each membrane, the procedure was repeated three times at room temperature, with a constant volume of distilled water (10 µL). The static contact angle formed between membrane and different liquids drop (water, dimethyl sulfoxide, and glycerol) was measured, and surface free energy (SFE) was calculated. The contact angle value represents the average of 5 measurements ± Standard Deviation (SD).

Water Uptake
Water uptake was determined by the gravimetric method [16]. The prepared polymeric membranes were cut into small pieces with a size of 15 mm × 15 mm and kept in a beaker with 25 mL of distilled water for 24 h at room temperature (22 • C). After that, the polymeric membranes were taken out, and the excess water was removed from the polymeric membranes surface using filter paper and weighed using an analytical balance (Kern PCB 350-3, precision 0.001 g) (m wet ). The biopolymeric membranes were immediately placed in an oven (DNO20-MRC, MRC Ltd., Holon, Israel), dried at 50 • C for 3 h, and then were weighed (m dry ). The water uptake percentage was calculated applying Equation (1): Water uptake (%) = m wet − m dry m dry · 100 (1)

Thermal Gravimetric Analysis (TGA) and Differential Thermal Analysis (DSC)
Thermogravimetric analysis (TGA) and differential thermal analysis (DSC) of the prepared membranes were conducted using a Q5000IR system (TA Instruments, New Castle, DE, USA). Briefly, 4-14 mg of the sample was heated from 35 to 740 • C with a heating rate of 10 • C/min in platinum pans (100 µL), under nitrogen atmosphere (50 mL/min).

Microscopy Studies
A digital optical microscope with high-resolution and full color 3.5" TFT LCD screen (Celestron LCD Digital Microscope II) was used to investigate the surface morphology of the prepared membranes without and with chitosan-silver ions.
The surface morphologies, top and cross-section view of the prepared biopolymeric membranes, were recorded using Scanning Electron Microscopy (SEM) (S-2600N, HITACHI). The biopolymeric membranes were cut into pieces of 0.5 × 0.5 cm and then covered with a thin gold layer. The experiments were performed in liquid nitrogen.

Electrochemical Impedance Spectroscopy
The bulk resistance of the membranes was determined through the electrochemical impedance spectroscopy. The experimental set-up was performed as previously indicated in our studies [40][41][42] using a specific symmetrical four electrodes impedance cell ( Figure 2) with two pairs of Ag/AgCl electrodes. One pair of electrodes will allow the current passing, and the other will measure the potential difference across the polymeric membrane.
All electrochemical impedance experiments have been carried out at an ambient temperature, in the presence of the ferrous electrolyte (10 −2 mol/L). The electrical parameters of the membranes will be extracted from the equivalent circuits associated with the membranes.
A Parstat 2273 Electrochemical System (Princeton Applied Research University, USA) was used for all impedance measurements. The impedance cell made from Jena glass and presented in Figure 2 was manufactured at the University Politehnica of Bucharest facilities. The presented cell allowed the measurement of the voltage and the current at the membrane level using all four available channels of the electrochemical system. Figure 2 presents the four active electrodes-the working electrode (W E ) and the counter electrode (C E ) with their associated reference electrodes (W R , C R ). The reference electrodes connected through an operational amplifier gave the measured potential (V measured ). At the level of C E is applied the selected voltage (V applied ) and the total current is provided, while the W E measures the current (I measured ). The electrodes were made from 0.5 mm pure Ag wire (99.9%, Sigma Aldrich, Merck KGaA, Darmstadt, Germany), either as Ag rods or coiled AgCl coated Ag wire.

Figure 2.
Scheme of the used impedance cell with four electrodes. WE-working electrode; WRreference electrode associated with the WE; CE-counter electrode; CR-reference electrode associated with the CE; Imeasured is the measured current; Vmeasured is the measured potential; Vapplied is the applied potential.
A Parstat 2273 Electrochemical System (Princeton Applied Research University, USA) was used for all impedance measurements. The impedance cell made from Jena glass and presented in Figure  2 was manufactured at the University Politehnica of Bucharest facilities. The presented cell allowed the measurement of the voltage and the current at the membrane level using all four available channels of the electrochemical system. Figure 2 presents the four active electrodes-the working electrode (WE) and the counter electrode (CE) with their associated reference electrodes (WR, CR). The reference electrodes connected through an operational amplifier gave the measured potential (Vmeasured). At the level of CE is applied the selected voltage (Vapplied) and the total current is provided, while the WE measures the current (Imeasured). The electrodes were made from 0.5 mm pure Ag wire (99.9%, Sigma Aldrich, Merck KGaA, Darmstadt, Germany), either as Ag rods or coiled AgCl coated Ag wire.
The area of the cut polymeric membranes, used in the impedance cell, was 0.785 cm 2 , and the width 0.22 mm. The applied frequency range was from 1.0 Hz to 2.0 MHz, and the applied a.c. amplitude was 10 mV. For all experiments, the data integration was set at 5 cycles and 10 steps per decade.
The electrochemical impedance investigations allow the assessment of the main electrical parameters of the polymeric membranes studied, namely: resistance, capacitance, dielectric constant. The dielectric constant, known as well as relative electrical permittivity, is determined from the measured geometrical capacitance and the geometrical characteristics of the polymeric membranes: width and surface. The geometrical capacitance was calculated using both the Nyquist and Bode plots and was determined according to the Equation (2) [40][41][42]: where: Cg represents the capacitance (F); γ stands for the frequency at the maximum recorded on the Nyquist plot (Hz); Rct is the charge transfer resistance of the membrane (Ω). The value of Rct was determined using the Nyquist diagram as the difference between the intercept of the semicircle Figure 2. Scheme of the used impedance cell with four electrodes. W E -working electrode; W R -reference electrode associated with the W E ; C E -counter electrode; C R -reference electrode associated with the C E ; I measured is the measured current; V measured is the measured potential; V applied is the applied potential.
The area of the cut polymeric membranes, used in the impedance cell, was 0.785 cm 2 , and the width 0.22 mm. The applied frequency range was from 1.0 Hz to 2.0 MHz, and the applied a.c. amplitude was 10 mV. For all experiments, the data integration was set at 5 cycles and 10 steps per decade.
The electrochemical impedance investigations allow the assessment of the main electrical parameters of the polymeric membranes studied, namely: resistance, capacitance, dielectric constant. The dielectric constant, known as well as relative electrical permittivity, is determined from the measured geometrical capacitance and the geometrical characteristics of the polymeric membranes: width and surface. The geometrical capacitance was calculated using both the Nyquist and Bode plots and was determined according to the Equation (2) [40][41][42]: where: C g represents the capacitance (F); γ stands for the frequency at the maximum recorded on the Nyquist plot (Hz); R ct is the charge transfer resistance of the membrane (Ω). The value of R ct was determined using the Nyquist diagram as the difference between the intercept of the semicircle having the Z' axis in the region of low frequencies and the intercept of the semicircle having real axis in the region of high frequencies. The accurate value of the frequency recorded for the maximum point was established from the Bode plot. The relative electrical permittivity, which depends on the homogeneity of the surface, is given by Equation (3) [40][41][42]: where: C g represents the capacitance (F); ε r is the relative electrical permittivity of the polymeric membrane; ε 0 represents the electrical permittivity of the free space (8.854 × 10 −12 F/m); A is the surface (1.53 ×10 −4 m 2 ), while l is the width (2.2 × 10 −4 m) of the membrane. The dielectric constant gives essential information on the charge transporting ability of the polymeric membranes. Conductivity, another significant electrical parameter, was calculated using the experimental data provided by the Nyquist diagram, according to the Equation (4) [40][41][42]: where: σ is the electrical conductivity (Ω −1 ·m −1 ); R ct is the charge transfer resistance (Ω); l is the thickness (m), and A is the surface of the membrane (m 2 ). The distribution of both carrier concentration and mobility within a membrane will lead to the distribution of conductivity (σ).

Electrodialysis Equipment and Procedure
Experiments were carried out in an electrodialysis system with three circular compartments made from high-density polyethylene with coextruded polypropylene exfoliating protective layer. The electrodialysis system was equipped with two parallel circular lead electrodes (purity 99.9%, anode, and cathode), placed at the ends of the electrodialysis cell. The obtained biopolymeric membranes were positioned between the compartments ( Figure 3). The thickness of each compartment, with rubber gaskets (Yato YT-0687) on both sides of the compartment, was 11.6 mm. The effective surface area of each biopolymer membrane and each electrode, under investigation, was 14.3 cm 2 . The thickness of each electrode was 0.24 mm. The biopolymer membrane thickness was between 0.17 mm and 0.22 mm. The dimensions of electrodialysis cell, membrane, and electrodes thickness were performed with a Neiko 01407A electronic digital caliper stainless steel body with LCD screen (inch/fractions/millimeter conversion).  All experiments were accomplished with synthetic industrial wastewater prepared using ferrous sulfate heptahydrate, 10 mL of sulfuric acid 1 N, and deionized water to produce a wastewater feed of 100 ppm iron ions. The compartments were filled with a total volume of approximately 50 cm 3 feed synthetical wastewater. The electrodialysis experiments were performed at room temperature (22 ± 1 °C) and operated under different constant voltages (5, 10, and 15 V) (galvanostatic operation mode) imposed at electrodes by a power source (AXIOMET AX-3005D) for 1 h (each experiment). In the electrodialysis system, the applied voltage produces the transportation  All experiments were accomplished with synthetic industrial wastewater prepared using ferrous sulfate heptahydrate, 10 mL of sulfuric acid 1 N, and deionized water to produce a wastewater feed of 100 ppm iron ions. The compartments were filled with a total volume of approximately 50 cm 3 feed synthetical wastewater. The electrodialysis experiments were performed at room temperature (22 ± 1 • C) and operated under different constant voltages (5, 10, and 15 V) (galvanostatic operation mode) imposed at electrodes by a power source (AXIOMET AX-3005D) for 1 h (each experiment). In the electrodialysis system, the applied voltage produces the transportation of the cations (Fe 2+ ) towards the cathode electrode and the anions (SO 4 2− ) and hydroxyl ions (OH -) towards the anode electrode. Also, the migration of iron ions occurs across the membranes under the applied voltage at the electrodes. All solutions, after each experiment, were filtered using the filter paper (Whatmann paper, 0.45 µm) to minimize the interference of fine particles in suspension with the analysis. The filtrates were analyzed for residual concentrations of iron ions using a UV-visible spectrophotometer Metertech SP-830 at a wavelength of 510 nm.

FTIR-ATR Spectroscopy Characterization of Membranes
To identify the specific absorption bands of each component and to establish the interaction between CA and PEG and CA, PEG, CHI, and silver (Ag) ions, infrared spectroscopy was used. The FTIR-ATR spectra of the obtained membranes are presented in Figure 4.  FTIR-ATR spectra of polymeric membranes, with and without chitosan (CHI)-silver (Ag) ions, show various absorption bands in 4000-600 cm −1 domain ( Figure 4) [43]. The incorporation of CHI-Ag into the polymer matrix occurred successfully, as shown by the following bands:

•
The stretching vibration band related to the-OH group (νO-H) was observed at 3405 cm −1 for the polymeric membrane without chitosan-silver ions [44]. An intense absorption band was observed at 3395 cm −1 when CHI-Ag was added to polymer mixture, corresponding to νO-H overlap to the stretching vibration of N-H (νN-H) from the free amino group (-NH2) at the C2 position of glucosamine from CHI. This absorption band registered by prominent peaks suggests the hydrogen bonds formed between components. Comparing the membranes' spectra, it can be observed that the presence of CHI-Ag into the composition led to a more intense peak shifted to lower wavenumber (about ∼10 cm −1 ).

•
The aliphatic C-H stretching was registered at 2956 cm −1 and 2897 cm −1 , respectively, in the polymeric membrane without chitosan-silver ions (CA/PEG/CHI-Ag membrane) [45]. The carbonyl group stretching vibration (νC=O) specific to CA appeared in both membranes at 1733 cm −1 [46]. Also, the presence of this band suggested the formation of O-H· · ·C=O interaction FTIR-ATR spectra of polymeric membranes, with and without chitosan (CHI)-silver (Ag) ions, show various absorption bands in 4000-600 cm −1 domain ( Figure 4) [43]. The incorporation of CHI-Ag into the polymer matrix occurred successfully, as shown by the following bands:

•
The stretching vibration band related to the-OH group (ν O-H ) was observed at 3405 cm −1 for the polymeric membrane without chitosan-silver ions [44]. An intense absorption band was observed at 3395 cm −1 when CHI-Ag was added to polymer mixture, corresponding to ν O-H overlap to the stretching vibration of N-H (ν N-H ) from the free amino group (-NH 2 ) at the C 2 position of glucosamine from CHI. This absorption band registered by prominent peaks suggests the hydrogen bonds formed between components. Comparing the membranes' spectra, it can be observed that the presence of CHI-Ag into the composition led to a more intense peak shifted to lower wavenumber (about~10 cm −1 ).

•
The same band corresponding to valence's vibrations of C-H (ν asC-H ) in -CH 2 -and in -CH 3 (ν s C-H ) appeared at 2956 cm −1 , 2920 cm −1 , and 2853 cm −1 in the case of the polymeric membrane enriched with chitosan-silver ions. The new peak at 2920 cm −1 and the shifting from 2897 cm −1 to 2853 cm −1 are due to the CHI-Ag incorporation in the mixture.

•
The carbonyl group stretching vibration (ν C=O ) specific to CA appeared in both membranes at 1733 cm −1 [46]. Also, the presence of this band suggested the formation of O-H· · ·C=O interaction between CA and PEG and CA, PEG and CHI, respectively [47].

•
Vibrations of-OH groups from PEG and CHI are evidenced by the peak at around 1431 cm −1 corresponding to δ CH2 in CH 2 OH groups.

•
The peaks at around 1370 cm −1 and 1320 cm −1 , respectively, appeared in both membranes and are related to the vibrations of -OH and -CH groups in the pyranose ring [48]. Significant changes occurred in the FTIR-ATR spectrum of the polymeric membrane enriched with chitosan-silver ions, primarily related to the modification of the absorption band intensities. The shifting of some bands reveals the incorporation of CHI-Ag into the polymer mixture.

Contact Angle and Surface Free Energy (SFE) Determination
Contact angle measurements investigated the effect of incorporation of CHI and Ag ions into the polymer mixture on the membrane's hydrophilic/hydrophobic capacity. A membrane with high hydrophobicity could lead to the deposition of particles on the surface reducing the membrane efficiency and its lifetime. The contact angle values of obtained membranes using different liquids (water, dimethyl sulfoxide and glycerol) are presented in Table 1. The values presented in Table 1 show that the higher contact angle value was registered for the CA/PEG membrane (101 • ), indicating its hydrophobic character. In the case of the CA/PEG/CHI-Ag membrane, the contact angle decreased (90.12 • ) when water was used as liquid, which demonstrates the effect of CHI-Ag presence on the decrease in polymeric membrane hydrophobicity. For CA/PEG membrane, a higher contact angle value was registered, compared with CA/PEG/CHI-Ag membrane, indicating that the amino group presence from chitosan led to a decrease in hydrophobic capacity. In this case, high values of the contact angle may suggest a decrease in membrane's reliability in time [49]. The contact angle offers a certain versatility of the prepared polymeric membrane for biomedical applications due to the antibacterial potential effect of the Ag ions trapped into the polymer matrix [21,34,50].
Moreover, to evaluate the chemical affinity between the liquid and the solid support, the surface free energy (SFE) was assessed. Using the contact angle values against three liquids (Table 1) and Owens-Wendt's theory based on the Fawkes method [44], disperse and polar fractions of the surface tension for liquids and SFE for solids were calculated (Equations (5) to (7)). The surface tension parameters of used fluids [51][52][53] are presented in Table 2, and the obtained results are shown in Table 3.
Polymers 2020, 12, 1792 11 of 22 where: θ represents the contact angle; γ S is the total surface energy of the solid; γ L is the total surface tension/surface energy of the liquid; γ d L and γ p L are dispersive and polar components of the surface tension, while γ d S and γ p S are the dispersive and polar components of the solid surface energy [51][52][53]. A decrease in contact angle value (Table 1), as well as an increase in total surface energy (Table 2) and the value of the polar components for the CA/PEG/CHI-Ag membrane, was noticed. The hydrophobic character of the CA/PEG membrane decreased due to the incorporation of CHI-Ag, and the increase of polar components favored the increase in energy values ( Table 3).
The mixing of PEG and CHI-Ag solution into a polymeric matrix allowed the obtaining of biopolymeric membranes where the silver ions were entrapped throughout the chitosan cross-linked polymeric networks via amino (-NH 2 ) and hydroxyl (-OH) functional groups. Due to its water solubility, PEG has a good dispersion capacity in polymer matrix [27]. The amino groups from chitosan are involved as chelating agent for the removal of iron ions, and the Ag ion influenced dielectric permittivity and electric conductivity [26,49].

Water Uptake
The percentage of water uptake values (calculated according to the Equation (1) [54]) for CA/PEG/CHI-Ag membrane was 66.67%, and for CA/PEG membrane was 64.17%. A small difference can be seen between the values of the percentage of water uptake (%) for the two membranes due to the small amount of CHI-Ag (0.5 mL) included in the polymer solution. However, the value of the water uptake percentage CA/PEG/CHI-Ag membrane was higher than that obtained for the CA/PEG membrane, due to the immobilization of chitosan in the polymer solution, which leads to some changes in the membranes. This result correlates well with the spectral data obtained by FTIR. The interaction between CA/PEG and CHI-Ag led to a more compact polymer structure with improved iron ions removal efficiency from wastewater by electrodialysis process. This fact could be due to the interactions with the hydroxyl and amino groups from chitosan on the surface. Also, the presence of Ag ion influenced the water uptake which is important for iron ion transport process [49,54] Ag ion influenced the water uptake which is important for iron ion transport process [49,54]  The thermal properties of the prepared polymer membranes are different due to the presence of CHI-Ag in the polymer matrix ( Figure 5 and Table 4). The first weight loss reduction appeared because the water contained in the sample evaporated [55]. The prepared polymer membranes show excellent thermal stability up to 350 °C, the maximum decomposition (Tmax) appearing at 367.5 °C in the case of the CHI-Ag membrane and at 359.6 °C for the polymer membrane without CHI-Ag. The degradation starts later for CA/PEG/CHI-Ag because of loose networks or moisture present due to The thermal properties of the prepared polymer membranes are different due to the presence of CHI-Ag in the polymer matrix ( Figure 5 and Table 4). The first weight loss reduction appeared because the water contained in the sample evaporated [55]. The prepared polymer membranes show excellent thermal stability up to 350 • C, the maximum decomposition (T max ) appearing at 367.5 • C in the case of the CHI-Ag membrane and at 359.6 • C for the polymer membrane without CHI-Ag. The degradation starts later for CA/PEG/CHI-Ag because of loose networks or moisture present due to the nanoparticle's incorporation process, leading to porosity. The significant weight loss was found to be 81% for CA/PEG/CHI-Ag membrane, from 251.5 to 457.5 • C. Such weight loss could be assigned to the decomposition of cellulose acetate and chitosan. The last weight loss of 11.92% was observed between 457.5 and 740 • C and can be due to the decomposition of CA/PEG/CHI-Ag. The CA/PEG/CHI-Ag membrane has higher stability than CA/PEG membrane. The difference of the residual mass at 740 • C is due to the presence of silver nanoparticles in the membrane [21,50,56]. The T max is higher for CA/PEG/CHI-Ag membrane in comparison with the CA/PEG membrane, which may result due to the incorporation of CHI-Ag in the polymer matrix ( Table 4). The thermal stability of prepared polymeric membranes is affected by the dispersion of silver ions and chitosan in the polymer matrix, hydrogen bonding, and interaction of polymer chains and silver ions [50,56]. Also, the T max increased for the membrane, which contains CHI-Ag, indicating that the Ag ions can modify the thermal stability of the CA/PEG/CHI-Ag membrane. This phenomenon can be ascribed to the fact that the overall crystallinity of the polymer is increased, leading to enhanced thermal properties.
The DSC was carried out under the continuous flow of dry nitrogen gas (50 mL/min) at 10 • C/min heating rate. Figure 5 illustrates the thermograms of CA/PEG and CA/PEG/CHI-Ag membranes, respectively. In Figure 5, the endothermic event appeared between 35-80 • C, corresponding to water evaporation. The CA/PEG/CHI-Ag membrane exhibits increased melting temperature due to the addition of CHI-Ag into the polymer matrix [50,56].

Microscopic Analysis
The surface morphology image at the macroscale (magnification 80×) for CA/PEG membrane shows a highly texturized and porous surface, with many opened pores, in comparison with the apparent homogeneous, compact, and flat surface for CA/PEG/CHI-Ag membrane ( Figure 6). The optical microscopies images highlight that the CHI-Ag molecules were included in the obtained membrane (as spherical particles), and these decreased the rate of pores formation [34,50]. properties.
The DSC was carried out under the continuous flow of dry nitrogen gas (50 mL/min) at 10 °C/min heating rate. Figure 5 illustrates the thermograms of CA/PEG and CA/PEG/CHI-Ag membranes, respectively. In Figure 5, the endothermic event appeared between 35-80 °C, corresponding to water evaporation. The CA/PEG/CHI-Ag membrane exhibits increased melting temperature due to the addition of CHI-Ag into the polymer matrix [50,56].

Microscopic Analysis
The surface morphology image at the macroscale (magnification 80×) for CA/PEG membrane shows a highly texturized and porous surface, with many opened pores, in comparison with the apparent homogeneous, compact, and flat surface for CA/PEG/CHI-Ag membrane ( Figure 6). The optical microscopies images highlight that the CHI-Ag molecules were included in the obtained membrane (as spherical particles), and these decreased the rate of pores formation [34,50].  The top and section surface morphology of the obtained polymeric membranes without and with chitosan (CHI)-silver (Ag) ions was examined by SEM (Figure 7). SEM images (Figure 7B-a,B-b) indicate the presence of CHI-Ag ions into the prepared biopolymer. The CA/PEG/CHI-Ag membrane exhibited a dense and uniform plain microstructure with many small pores throughout the membrane. The SEM results illustrate that the pores formed have an average size between 6.12-12.64 µm for CA/PEG membranes (Figure 7A-a) and between   Figure 7B-a,B-b) indicate the presence of CHI-Ag ions into the prepared biopolymer. The CA/PEG/CHI-Ag membrane exhibited a dense and uniform plain microstructure with many small pores throughout the membrane. The SEM results illustrate that the pores formed have an average size between 6.12-12.64 µm for CA/PEG membranes ( Figure 7A-a) and between 17.23-53.61 µm for CA/PEG/CHI-Ag membrane ( Figure 7B-a). The significant decrease in the pore sizes could be due to the presence of the CHI-Ag ions in the polymer matrix and due to the involvement of further reduction as well as balancing the stabilization process by chitosan and PEG polymeric chains. It was observed that the CA/PEG/CHI-Ag membrane presents macro voids that are interconnected through the walls, which have small pores that could create a barrier to the transport of water and of metallic ions. The presence of pores and of macro voids can reduce the intrinsic membrane resistance for metallic ions transport [19]. The literature studies indicate that the silver particles presented in prepared biopolymeric membranes are within the useful range for essential antibacterial applications [21,50].

Electrochemical Impedance Spectroscopy Tests
The a.c. impedance spectroscopy represents one of the essential techniques that allow the study of the electrical response of the membranes in terms of complex impedance, dielectric constant, or conductivity, among others [40][41][42]. The impedance configuration of the system under study could be explained as a sum of RC circuits in parallel or series combination, being presented in the complex form of Z'-resistive and Z"-capacitive. The cell with four electrodes, and presented in Figure 2, allows the minimization of the impedance artifacts in the measurement of the biopolymeric membranes studied. During the standard impedance measurement of a polymer/composite material, placed between the two electrodes, the complex physical parameters (impedance, conductivity) are assessed from the current I measured and the potential difference V measured = V + − V − (see Figure 2). The classical experiments are based on the assumption that the potential dropped homogeneously across the polymeric film. However, this hypothesis is not valid in real systems when the studied materials are inhomogeneous, with ionic insertions, for instance, that generate important potential changes with significantly different profiles across the membranes. Such specificity of the system studied together with the necessity to avoid the measurement impedance artifacts, sustain the usage of the four electrodes cell for the electrochemical investigations. Much more, at low values for the frequency, below 100 Hz, the silver electrodes present higher values for the impedance compared to the silver-silver chloride electrodes, whose impedance decrease with frequency increasing.
As a consequence, the used silver-silver chloride C R and W R electrodes presented low impedance, comparable with the W E and C E impedances. During the experimental setup, it was taken into account that an applied voltage with high amplitude could cause a significant movement of ions inside/out the biopolymeric membrane that might lead to erroneous experimental data. Therefore, the used 0.01 V for voltage amplitude permitted a proper recording of the impedance spectra.
The electrochemical impedance spectra (Figures 8 and 9) obtained for both biopolymeric membranes (without and with CHI-Ag), were characterized through fully resolved semicircles. The absence of any linear part shows that the exchange process at the level of the membranes occurs without diffusional control. For the prepared membranes, the observed single semicircle indicates a unique relaxation process. the used 0.01 V for voltage amplitude permitted a proper recording of the impedance spectra.
The electrochemical impedance spectra (Figures 8 and 9) obtained for both biopolymeric membranes (without and with CHI-Ag), were characterized through fully resolved semicircles. The absence of any linear part shows that the exchange process at the level of the membranes occurs without diffusional control. For the prepared membranes, the observed single semicircle indicates a unique relaxation process. The displacement of the semicircle on the diagram towards origin with the simultaneously decreasing in the semicircle size upon adding CHI-Ag in the membrane composition signifies that the membrane conductivity is not electronic, but ionic.  The set up in the four electrodes investigation cell ( Figure 2) allowed us to assess the main electrical parameters with analytical interest for our study. The electrochemical behavior of the membranes was modeled using the equivalent circuits presented in Figure 9. For all samples studied, a single semicircle fit was recorded on the Nyquist diagram, indicating a single relaxation process. The Z′′ peak on the Nyquist plot shifted towards a higher frequency region when the membrane contained CHI-Ag. Such behavior suggests that electrical relaxation depends on membrane composition.
As shown in Figure 10, the two equivalent circuits are similar. Compared to the membranes CA/PEG, in the equivalent circuit for the membrane with CHI-Ag, it was introduced in parallel to a capacitance, which is clearly due to the presence of CHI-Ag. The bulk resistances for the studied membranes varied quite significantly from 1160 Ω for the membrane CA/PEG to 10.9 Ω for the CA/PEG/CHI-Ag membrane. Therefore, it could be concluded that adding the chitosan-Ag into the membrane composition, the bulk resistance of the membrane was lowered by two orders of magnitude. The changes in the membrane resistance could be related to the variations in ion mobility within the composite membrane. The displacement of the semicircle on the diagram towards origin with the simultaneously decreasing in the semicircle size upon adding CHI-Ag in the membrane composition signifies that the membrane conductivity is not electronic, but ionic.
The set up in the four electrodes investigation cell ( Figure 2) allowed us to assess the main electrical parameters with analytical interest for our study. The electrochemical behavior of the membranes was modeled using the equivalent circuits presented in Figure 9. For all samples studied, a single semicircle fit was recorded on the Nyquist diagram, indicating a single relaxation process. The Z peak on the Nyquist plot shifted towards a higher frequency region when the membrane contained CHI-Ag. Such behavior suggests that electrical relaxation depends on membrane composition.
As shown in Figure 10, the two equivalent circuits are similar. Compared to the membranes CA/PEG, in the equivalent circuit for the membrane with CHI-Ag, it was introduced in parallel to a capacitance, which is clearly due to the presence of CHI-Ag. The bulk resistances for the studied membranes varied quite significantly from 1160 Ω for the membrane CA/PEG to 10.9 Ω for the CA/PEG/CHI-Ag membrane. Therefore, it could be concluded that adding the chitosan-Ag into the membrane composition, the bulk resistance of the membrane was lowered by two orders of magnitude. The changes in the membrane resistance could be related to the variations in ion mobility within the composite membrane. The diffusion coefficients of ions in membranes are strongly dependent upon the dielectric constant. A higher value for the membrane dielectric constant suggests a better ability of the membrane to hold a more substantial charge onto the membrane surface. The dielectric constant calculated according to the procedure presented in the experimental part (Equations (2) and (3)) was 10.41 for the CA/PEG membrane. In contrast, a value of 28.40 for the dielectric constant of the CA/PEG/CHI-Ag membrane was determined. Previously, literature data reported dielectric constant values for cellulose acetate membrane around 3.6 [57,58].
The linear amino polysaccharide compound that is chitosan (CHI) presents various functional groups, namely: -NH2, -C=O-NHR, or -OH. In complex polymeric matrix as cellulose acetate-PEG-CHI, the presence of silver particles influences the dielectric permittivity and electric conductivity that have a significant role in the ferrous ion transport mechanism.
The addition of CHI-Ag to the initial polymeric matrix cellulose acetate-PEG increases the conductivity of the final membranes. Such an increase of the conductivity values could be assigned to the possible movement of Ag charge carriers. Therefore, we might state that there are two routes for cation transport through the membrane: continuous narrow water channels and the bulk membrane hydrophobic phase. For the initial membrane, cellulose acetate-PEG, a value of 5.49 × 10 −4 S/cm has been obtained for the electric conductivity. In other papers, for cellulose acetate-PEG membrane, values of the electrical conductivity between 2.12 × 10 −5 S/cm and 9.5 × 10 −4 S/cm [59] have been reported, but through another experimental procedure. Nevertheless, based on the impedance/diffusion cell used in the impedance investigations, the determined experimental values are realistic for the chemical species transportation.
The rapid movements of the polymer matrix segment, and the silver particles, generate a higher ionic electrical conductivity, as it is the case of the membrane with CHI-Ag when σ = 5.89 × 10 −2 S/cm. The inclusion of an inert filler as silver in the polymeric matrix resulted in a higher ionic electrical conductivity. The diffusion coefficients of ions in membranes are strongly dependent upon the dielectric constant. A higher value for the membrane dielectric constant suggests a better ability of the membrane to hold a more substantial charge onto the membrane surface. The dielectric constant calculated according to the procedure presented in the experimental part (Equations (2) and (3)) was 10.41 for the CA/PEG membrane. In contrast, a value of 28.40 for the dielectric constant of the CA/PEG/CHI-Ag membrane was determined. Previously, literature data reported dielectric constant values for cellulose acetate membrane around 3.6 [57,58].
The linear amino polysaccharide compound that is chitosan (CHI) presents various functional groups, namely: -NH 2 , -C=O-NHR, or -OH. In complex polymeric matrix as cellulose acetate-PEG-CHI, the presence of silver particles influences the dielectric permittivity and electric conductivity that have a significant role in the ferrous ion transport mechanism.
The addition of CHI-Ag to the initial polymeric matrix cellulose acetate-PEG increases the conductivity of the final membranes. Such an increase of the conductivity values could be assigned to the possible movement of Ag charge carriers. Therefore, we might state that there are two routes for cation transport through the membrane: continuous narrow water channels and the bulk membrane hydrophobic phase. For the initial membrane, cellulose acetate-PEG, a value of 5.49 × 10 −4 S/cm has been obtained for the electric conductivity. In other papers, for cellulose acetate-PEG membrane, values of the electrical conductivity between 2.12 × 10 −5 S/cm and 9.5 × 10 −4 S/cm [59] have been reported, but through another experimental procedure. Nevertheless, based on the impedance/diffusion cell used in the impedance investigations, the determined experimental values are realistic for the chemical species transportation.
The rapid movements of the polymer matrix segment, and the silver particles, generate a higher ionic electrical conductivity, as it is the case of the membrane with CHI-Ag when σ = 5.89 × 10 −2 S/cm. The inclusion of an inert filler as silver in the polymeric matrix resulted in a higher ionic electrical conductivity.
The electrochemical impedance analysis, applied in this study, has the important advantage of an accurate assessment. The influence that the interaction between polymer matrix and the analytical aqueous solution has on the membrane's charge transfer ability was quantified through the dielectric constant.

Performance Evaluation of Electrodialysis Cell
Based on preliminary data, which showed that a low applied voltage (<5 V) leads to a low treatment rate of iron ions, and a voltage over 15 V leads to cracking of polymeric membranes and increases energy consumption, there were chosen the values of applied voltages of 5, 10, and 15 V, respectively. Also, a concentrated synthetic industrial wastewater (over 100 ppm) was not studied due to the high energy consumption, damage, and clogging of membranes, which eventually increases the total cost. Moreover, after 1 h of treatment, it was observed that the energy consumption is higher and, also, the prepared membranes shows cracks. These effects lead to higher total costs, which are not an advantage. In this paper, we have included the best results obtained using the electrodialysis system with the biopolymeric membranes without and with chitosan (CHI)-silver (Ag).
The treatment rate of the iron (II) ions (T R %) in the solutions was calculated by the Equation (8) [13][14][15][16][17]: where: C i and C f , were the initial and final iron (II) ions concentrations (g/L). The current efficiency (I E %) was calculated based on the concentrations of solutions using Equation (9) [15][16][17]: where: I E (%) is current efficiency; z is the valence of the iron ion; V is the total volume from all compartments (L); F is the Faraday constant (96,486 C/mol); N is the number of repeat cells; − I is the average current intensity (A); t is the time of experience (s).
The energy consumption (W c , kWh/m 3 ) of treatment by electrodialysis cell was calculated using the following Equation (10) [13][14][15][16][17]: where: W c is the energy consumption (kWh/m 3 ); U is the applied voltage of the cell (V); I is the current intensity (A); t is the final of experience (h), and V is the total volume from all compartments (m 3 ). The calculated results for iron ions, after 1 h of electrodialysis cell, at different applied voltages, are indicated in Table 5. The data presented in Table 5 highlight that the applied voltage that increased from 5 to 15 V causes an increase of the T R from 15.86 to 33.71% for the polymeric membrane without chitosan-silver ions (CA/PEG membrane), and from 30.35 to 63.70% for the polymeric membrane enriched with chitosan-silver ions (CA/PEG/CHI-Ag membrane), after 1 h of treatment. The higher value of T R resulted from a more intensive ionic migration with a higher voltage applied. At the end of each experiment, the formation of wet iron deposits on the cathode surface was observed. After drying, this deposit (iron powder) can be used in many different industries for different applications, such as brazing (sandpaper), chemicals, metallurgy (iron wires, iron cylinders, car parts, power tools), printing (toners, electrophotographic process), surface coating (iron plates, engine valves, steel rollers), dyes (if the iron powder is soaked in vinegar, a black, gray or brownish-red dye could be obtained, useful for craft makers and woodworkers) [16,60]. If, after the treatment of wastewater by the electrodialysis system, the water has the concentration allowed by the legislation imposed by the Environmental Protection Agency, then it can be released into the environment or it can be used as washing water, irrigation, or as rinsing water.
The increased treatment rate is due to the protons and hydroxyl ions (HO -) resulted from the water dissociation at the membrane surface [26,27]. Under the influence of electric fields, the protons move towards the cathode and hydroxyl ions (HO -) to the anode where they met with the Fe 2+ through the membrane. Therefore, the hydroxide precipitate is formed onto the polymeric membrane surface [16]. In the literature, it was reported that under acidic conditions, the mobility of the adsorbate to the bio-sorbent surface is inhibited because the amino group (RNH 3 + ) and the M 2+ exhibit an electrostatic repulsion. The adsorption process of metal ions on chitosan is controlled by the transport of these species in solution [61]. Analyzing Table 5, it could be seen that the W c increases when the applied voltage increases due to the amount of electrical intensity flowing between the electrodialysis cells. At 15 V, it was recorded a higher value of W c (48.73 kWh/m 3 ) for the polymeric membrane without chitosan-silver ions (CA/PEG membrane), compared to 37.30 kWh/m 3 for polymeric membrane enriched with chitosan-silver (CA/PEG/CHI-Ag membrane). The presence of the iron ions causes a decrease in current efficiency for the polymeric membrane without chitosan-silver ions (CA/PEG membrane) and an increase in the energy consumption when the applied voltage increases. The T R increased with the increasing applied voltage, but the water dissociation took place when the voltage increased to 15 V. This fact leads to decrease the current efficiency with increase of the applied voltage [17]. The I E decreases as the iron ions concentration starts to increase in comparison with the concentration from the central compartment due to diffusion of these ions and osmotic water flow. At 15 V, the value of I E (9.42%) was higher for the polymeric membrane enriched with chitosan-silver ions (CA/PEG/CHI-Ag membrane) compared to 7.71% for the polymeric membrane without chitosan-silver ions (CA/PEG membrane). This fact may be attributed to the competing effects of higher dissociation degree of water at the biodegradable polymer membrane level when the applied voltage was 15 V. After 1 h of electrodialysis, the total voltage and the resistance increased possible due to the depletion of iron ions in the dilute solution and due to the heating of the solution from electrodialysis cell. It had been reported in the literature that the constant voltage applied to the electrodes crosses the electrodialysis system. Due to the applied voltage as a function of the total resistance of the system, the electrodialysis system can be compared to a series of resistors, which includes the electrode rinsing solution resistance, the electrode reaction resistance, the membrane resistance, and finally, the dilute and concentrated solution resistances [17]. As iron ion transport is produced during the electrodialysis process, a dilute solution was flowing across the system (central compartment). It was observed that the resistance continuously increases, and the total resistance increases in the electrodialysis system. Due to this fact, the total voltage increases under galvanostatic control.

Conclusions
The results showed that the biopolymeric membrane enriched with the chitosan-silver ions can be successfully used for the removal of iron ions from synthetic industrial wastewater using a new electrodialysis system. The obtained data indicated that the efficiency increased due to the incorporation of CHI-Ag into the CA/PEG membrane. The highest value of the treatment rate of the iron ion (>60%) was obtained for biopolymeric membranes enriched with the chitosan-silver ions, at 15 V, after one hour of treatment.
FTIR-ATR study evidenced that hydrogen bonds' interactions between the components in the mixtures occurred, and a complex of the polymeric matrix and metal ion was obtained. The TGA-DSC analysis for the obtained membranes showed excellent thermal stability (>350 • C), the maximum decomposition appearing at 367.5 • C for the biopolymeric membrane enriched with the chitosan-silver ions and at 359.6 • C for the biopolymeric membrane without chitosan-silver ions. The microscopic studies showed that the presence of CHI-Ag ions into the prepared biopolymeric membrane determined a decrease in the pore sizes.
The contact angle measurements evidenced different values for the prepared membranes with different compositions. The amino groups from chitosan presented in CA/PEG/CHI-Ag membrane composition have the key role in decreasing the hydrophobic capacity. Also, the hydrophobicity decreased with the increase of the polar components, favoring the increase in surface energy values.
The impedance spectroscopy studies allowed the assessment of the electrical conductivity of the prepared membranes. The value of this electrical parameter was improved from an initial 5.49 × 10 −4 S/cm for the biopolymeric membrane without chitosan-silver ions to 5.89 × 10 −2 S/cm for biopolymeric membrane enriched with the chitosan-silver ions. Such an increase in conductivity was assigned to the movement of the silver charge carriers or polymer matrix segments. The electrical characteristics of the biopolymeric membrane enriched with the chitosan-silver ions highlighted the suitability of the system for retaining iron ions from wastewaters.
The value of current efficiency was higher for the polymeric membrane enriched with chitosan-silver ions (9.42%) in comparison with the polymeric membrane without chitosan-silver ions (7.71%), and may be due to the competing effects of higher dissociation degree of water at the biodegradable polymer membrane level when the applied voltage was 15 V. The total voltage and the resistance increased was possible due to the depletion of iron ions in the dilute solution and due to the heating of the solution from electrodialysis cell.
The obtained biopolymeric membranes, due to their remarkable properties, could be used in different applications such as fuel cells, membrane processes (microfiltration, ultrafiltration, nanofiltration, reverse osmosis, ultrafiltration), and lately for thorough cleaning of clinical wastewaters.