3.1 Determination of Surface Charge of Chitosan in Solution
The zeta potential (ζ)) was measured to investigate the surface charge of a 0.1% w/v chitosan solution at different pH values. Figure 1 shows that at a low pH (pH = 3), the amino groups present in the chitosan are protonated, resulting in a positive surface charge. As the pH increases, the proton is gradually released from the amino groups, leading to a decrease in surface charge. The continuous decrease in surface charge with increasing pH indicates further deprotonation of the amino groups and the acquisition of negative surface charges on chitosan. The isoelectric point is found at pH = 8, where the surface charge of chitosan stabilizes and approaches zero. This means that at this pH, the number of protonated amino groups is equal to the number of deprotonated groups. Moreover, the negative surface charge observed at pH above 8 can be attributed to the presence of deprotonated amino groups and other charged species in the chitosan structure. Since chitosan has a positive surface charge at pH less than 7, favorable electrostatic interactions may occur with negatively charged ions as phosphate species. These interactions may determine the adsorption of phosphate species in chitosan membranes through electrostatic complex formation [11].
3.2 Morphological Characterization of Chitosan Membranes
3.2.1 Atomic Force Microscopy
Characterization of the morphology and roughness of the chitosan and chitosan-molybdate membranes was performed by atomic force microscopy (AFM) in contact mode. The thickness of the membranes deposited on the stainless-steel electrode was analyzed to avoid interferences caused by the presence of a carbon film during the casting process.
The morphological characterization of the stainless steel/chitosan membrane is shown in Fig. 2(a)-(b). Figure 2(a) shows a general view of the membrane morphology, which is homogeneous with no aggregates and no perceptible particle shape at a window size of 100 micrometers. To ensure the presence of the membrane, an indentation in the membrane can be seen on the lower right side of the image to distinguish it from the stainless-steel substrate; in this image, the roughness of the membrane surface was measured and an average value (Sa) of 74.79 nm was obtained. Figure 2(b) shows a 3D projection of the film with an image size of 2.5 micrometers. This image shows a rough film with some protrusions on the order of 10 nm and an average roughness (Sa) of 1.776 nm, confirming that the film looks practically smooth at lower magnification. To determine the thickness of the chitosan membrane, a scalpel was used to scratch the surface. Figure 2(c) shows the image of the interface between steel and membrane. By measuring a line roughness in this area (Fig. 2(d)), it is possible to determine the height difference and thus the thickness of the applied chitosan film. A film thickness of 1.5 µm was determined.
The morphological characterization of the stainless steel/chitosan molybdate membrane is shown in Fig. 3 (a)-(b). Figure 3 (a) shows a general overview of the membrane morphology. It is less smooth than the chitosan membrane, but also homogeneous and exhibits some porosity. The roughness of the membrane surface was determined and gave an average value (Sa) of 81.73 nm. The 3D projection of the film with an image size of 2.5 micrometers (Fig. 3 (b)) shows the presence of a less rough film with projections in the order of 20–45 nm and an average roughness (Sa) of 13.07 nm. To determine the thickness of the chitosan-molybdate membrane, a scratch was made on the surface with a needle. Figure 3(c) shows the image of the interface between steel and membrane. By measuring a line roughness in this area (Fig. 3(d)), it is possible to determine the height difference and thus the thickness of the deposited chitosan film. A film thickness of 2.59 µm was determined. The change in morphology and roughness of the chitosan-molybdate membrane indicates that the presence of the molybdate ion during the casting process of the membrane does not increase the relative size of the particles, but it does increase their porosity, which is certainly due to the interaction of the molybdate ion with the chitosan causing segregation of the micelles during the drying process.
3.2.2 FTIR spectroscopy
The FTIR spectra of the chitosan powder (Ch_P), the chitosan membrane (Ch_M), and the chitosan-molybdate membrane (Ch/M_M) are shown in Fig. 4. In the three spectra, a peak at 3400 cm− 1 is observed, which corresponds to the O-H stretch, and low-intensity signals at 2920 and 2880 cm− 1 are assigned to the C-H stretch, which is particular to –CH2– groups. At 1650 cm-1 a stretching peak is observed that corresponds to C = O, at 1420, 1380, and 1075 cm− 1 the N-H, C-H, and C-O stretching signals are observed respectively. The FTIR spectra correspond which those reported for chitosan. In the case of the chitosan membrane, it is observed that the signals are intensified, this is attributed to greater compaction in the chitosan membrane concerning chitosan powder and chitosan-molybdate membrane, corroborating what is observed in the morphological characterization. Additionally, the chitosan-molybdate membrane does not show any important modification in its surface [20, 21].
3.3 Electrochemical Characterization
The first electrochemical characterization of the prepared chitosan membranes was performed using cyclic voltammetry at a scan rate of 25 mV/s to determine the oxidation-reduction signals of the molybdate ion MoO4= in the presence of orthophosphate ions, reported in previous electrochemical studies [12]. Electrodes with membranes of different thicknesses were analyzed to observe their electrochemical performance
Figure 6 (a)-(b) shows the cyclic voltammetry results of 2 different electrodes. Figure 6 (a) shows the voltammetric response of a thin film of chitosan (150 nm) deposited on a stainless-steel substrate immersed in a solution of 640 mg/L (NH4)6Mo7O24 at pH = 1. The positive scan without the presence of phosphate ions shows two peaks, the first at -0.17 V vs Ag/AgCl is related to the electrochemical oxidation of Mo(II) → Mo(IV) and the second for the oxidation of Mo(IV) → Mo(VI), for the negative scan the two signals corresponding to the molybdenum reduction processes are detected at -0.22 V and − 0.07 V vs. Ag/AgCl, which are associated with the electrochemical reduction of Mo(VI) → Mo(IV) and Mo(IV) → Mo(II), respectively [23]. However, the presence of the chitosan membrane reduces the intensity of the signals associated with the molybdenum oxidation-reduction processes by gradually increasing the phosphate concentration. This can be attributed to the interaction between the amino groups in the chitosan membrane and the orthophosphate ion in the analyte, which reach equilibrium when adsorbed on the membrane surface, limiting the interaction of the molybdate ion with the surface. As the thickness of the chitosan membrane increases, the voltammetric analysis shows a capacitive response in all cases, with only the signals related to the oxidation-reduction processes of Mo(II) ↔ Mo(IV) being minimally observed. Moreover, the reduction of the capacitive current confirms the favoring of the adsorption of the HPO4= species on the surface of the chitosan membrane, avoiding the interaction of the MoO4= ion, for which the voltammetric analysis is not the most appropriate way to quantify available (dissolved) phosphate species.
At this point, the polarization of either chitosan and chitosan-molybdate membranes have shown their sensitivity to phosphate species; although a proper quantification is still a challenge. Considering that the biopolymeric membrane tends to accumulate phosphate species, it is natural to investigate if an electrochemical impedance analysis could give a quantitative evaluation of chitosan membranes capacity to accumulate the available phosphorous species.
The determination of the phosphate sensor capacitance of the fabricated electrodes was then performed using electrochemical impedance in a frequency range from 100 kHz to 1 Hz. For this purpose, the electrodes were analyzed in a 3-electrode cell with a reference electrode of Ag/AgCl and a platinum wire as a counter electrode.
The Nyquist plots of eS_Ch, eSC_Ch, eS_Ch/M, and eSC_Ch/M are shown in Fig. 7(a)-(d). The characteristic shape of the spectra corresponds to a process where at low frequencies the diffusion of the analyte into the solution is not evident, while at high and medium frequencies the formation of the characteristic semicircle can be observed, which results from the parallel response of the charge transfer resistance (Rct) and the chemical capacitance (Cµ) corresponding to an equivalent circuit R1 + CPC2/R2. Analysis of the Nyquist diagrams shows in all cases a decrease in Rct with increasing concentration of HPO4= ion in the solution. This effect can be attributed to ion absorption by the chitosan membrane. In addition, an increase in the Cµ of the electrodes is observed as the concentration of HPO4= in the solution increases, which is related to the increase in the number of ions adsorbed on the membrane surface.
The Nyquist plots for the chitosan membranes deposited on both electrodes (steel and steel/carbon) show a higher Rct resistance than the membranes containing molybdate, which reach values close to 50 kΩ and 70 kΩ at a concentration of 1 mg/L, while a decrease in Rct is observed as the concentration increases, promoted by the interaction of the HPO4= ion with the membrane surface, increasing the conductivity. The change in Rct value shows a rather linear trend for the eS_Ch electrode, while in the case of the chitosan-molybdate membranes, lower Rct values are observed at low concentrations, with values between 40 kΩ and 30 kΩ for membranes deposited on steel and Steel/carbon electrodes are determined, in this case the linear trend in terms of decrease of Rct with the increase of concentration of HPO4= ion is more evident for the carbon surface electrode, indicating that the presence of molybdate in the film improves the conductivity and that this ion also has a better interaction with the carbon film. This suggests that ammonium molybdate acts as a promoter for the uptake of HPO4= by the membrane and facilitates the charge transfer at the electrodes [22].
Figure 8 (a)-(d) show the values of Rct and Cµ obtained by fitting the equivalent circuit (R1 + CPC2/R2) according to the shape of the impedance spectrum. It can be observed that the best linear fit for Rct is obtained from eS_Ch and eSC_Ch/M, while the eSC_Ch electrode shows a semi-logarithmic trend. The chemical capacitance values in the case of the chitosan membranes show opposite trends, while for the chitosan-molybdate membranes they show a "more "linear" behavior. This indicates that the HPO4= ions are more homogeneously distributed on the surface of the membrane and this phenomenon can be related to its porosity. Therefore, the eSC_Ch/M electrode is considered to have better performance as a means of detecting phosphate in agricultural soils.