An insight to the filtration mechanism of Pb(II) at the surface of a clay ceramic membrane through its preconcentration at the surface of a graphite/clay composite working electrode

2.1 Clay sampling

The clay material investigated in this study was collected from Wadi Haqil Ras Al-Khaimah, UAE. The sampling site is situated approximately 3.0 km from Ras Al-Khaimah town center (Geographical coordinates: Latitude: 25°48′46″ North and the Longitude: 56°2′54″ Est). The clay material was pulverized manually with a mortar then different granule sizes were obtained using ASTM (American Society for Testing and Materials) standardized sieves. Approximately, 5 sizes were collected by mounting the sieves on top of each other. The fraction size in the range of 250–425 µm was used in the fabrication of the clay membrane supports.

2.2 Reagents

All chemicals were of analytical reagent grade and were used without further purification. Lead(II) nitrate (Pb(NO₃)₂, ACS reagent, ≥99.0 %), potassium chloride (KCl, ACS reagent, ≥99.0 %), Potassium hexacyanoferrate (II) trihydrate (K₄Fe(CN)₆·3H₂O, ReagentPlus, ≥98.5 %), hydrochloric acid (HCl, 37 %), graphite powder (20 µm, synthetic), mineral oil, phosphate-buffered saline (PBS, pH = 5,7) and boric acid/potassium chloride/sodium hydroxide buffer solution (pH = 10) were purchased from Sigma Aldrich. Activated carbon (sieved over 40 × 40 mesh, loss on ignition: 10 % at 120 °C and residue on ignition: 7 % at 800 °C) used as a porosity promotor was purchased from Fisher Scientific. Acetate buffer solution (0.1 M, pH 4.5) was prepared from sodium acetate (CH₃COONa) and acetic acid (CH₃COOH). The pH of the buffer solution was then adjusted using HCl. Deionized water (resistance = 15.0 MΩ.cm) was used throughout the experiments and for the preparation of solutions obtained from Elix® Essential 5 deionizer.

2.3 Clay membranes preparation

Clay mineral was sampled from the northern region of the United Arab Emirates. Particles ranging in size between 250 and 425 μm were made using ASTM standardized sieves. In brief, 3.0 g of 3 % (w/w) (activated carbon/clay) were shacked thoroughly to assure proper homogeneity, then introduced into a stainless-steel die set and pressed under 518 bar of uniaxial pressure, were flat discs like pellets (30.0 mm diameter, 2.0 mm thickness) were obtained. The consolidation of the membrane is assured by the surface adsorbed and constitutional waters of the clay; therefore, no water was added to the powders. The raw flat-disc membranes were sintered to final temperatures ranging between 250 and 1100 °C, using an electric furnace (CARBOLITE type 11/6B) following a heating program developed based on physicochemical phenomena occurring in common clay minerals based on thermal analysis. Clay membranes are crushed and ground into fine powders for further use in thermogravimetric analysis (TGA), X-ray fluorescence (XRF), X-ray diffraction (XRD), Fourier Transform Infrared Spectroscopy (FTIR), Scanning Electron microscopy equipped with Energy Disperse Spectroscopy (SEM/EDS) and for the preparation of electrochemical sensors.

2.4 Clay fraction isolation

5.0 g of ground clay was suspended in 1.0 L of deionized water and stirred for 18 h. 10.0 % HCl solution was added dropwise to the suspension with a speed of 2 drops/hour, where a total number of 50 drops were added resulting in the final pH of the suspension of 7.30. At this stage, the suspension was allowed to settle for 50 min. 40.0 mL aliquot of the supernatant was placed in four test tubes and centrifuged at 2700 rpm for 30 min. A very small amount of solid was recovered from the bottom of each tube and combined in one test tube then centrifuged again at 1750 rpm for 30 min. The recovered solid was then collected for XRD analysis. The analyzed sample was not identified as the clay fraction. The rest of the supernatant in the 1.0 L beaker was discarded and the settled coarse material at the bottom of the beaker was collected in 80.0 mL deionized water then transferred to test tubes and centrifuged at 3500 rpm for 40 min. The supernatants were discarded, and solids were suspended again in water and centrifuged at 600 rpm for 7 min, this operation was carried out two times to wash the solid material. At the final step, the cloudy supernatant was suspected to contain particles with a size less than 2 µm (clay fraction), which was then collected in a 100 mL beaker and further centrifuged at 2750 rpm for 40 min. Finally, the solid collected from the test tubes was analyzed in the XRD machine.

2.5 Characterization of clay ceramic membranes

Flat disc ceramic membrane supports prepared and sintered at different final temperatures were weighed then inserted in a beaker of water and left for a period of 48 h. Final masses of the support were recorded after being removed from the water and air dried for 2 h. The porosity of clay supports was calculated using Eq. 1.

Structural characterization of the clay ceramics membranes is performed by TGA, XRF, XRD, FTIR, and morphological study by SEM/EDS. Thermal analysis is performed using NETZSCH STA 449 F5 Jupiter operating in TGA modes under air at a linear static mode from room temperature of 25 °C to a final temperature of 1200 °C with a heating rate of 20 °C/min. XRF analysis is used to determine the chemical composition of the raw and sintered clay samples. A Philip model PW 1410 XRF machine with chromium target, PE crystal (2d = 8.742 Å) analyzer, and gas flow proportional counter was used. The tube rating was 50 kV, 40 mA, and 2θ range of 58° to 147°. The 2θ range is found to be suitable for major elements determination in clay using a recorder coupled detector. XRD was performed on the isolated clay fractions obtained using the process detailed above as well as raw and sintered clay materials to study the extent of phase transformations. Diffraction patterns were obtained with CuKα radiation (λ = 1.540 Å) on a PW1710 Philips Analytical diffractometer controlled by XPERT Quantify software (EA Almelo, The Netherlands) operating at 40 kV and 30 mA, with a copper anode and a graphite monochromator. The data was collected between 2θ = 5.0° to 85° in incremental steps of 0.05° with a rate (t) = 0.5 s/step. A monochromator with a normal divergence (1.0°) and receiving slits of 0.1 mm dimensions were used. 1300 points were recorded using continuous scans. Non-oriented preparations of the samples were used; fine powders were used in order to ensure the homogeneity of the sample. This method is very reliable and gives excellent results with regard to the total recovery of the clay mineral. The powder sample consisted of a given crystalline phase that always gave rise to diffraction peaks in the same directions, with relatively constant heights. This diffraction pattern thus formed a true signature of the clayey phase. FTIR of the clay materials was recorded using an FT-IR JASCO 6300 Spectrometer fitted with an attenuated total reflection (ATR) unit, with single reflection geometry. The clay material powders were kept at 70 °C in an oven, then taken in a desiccator for FTIR (ATR) analysis. Samples were introduced to the ATR unit with a single reflection geometry and spectra were collected in the range 4000–400 cm⁻¹ in the transmittance mode after 31 scans. The morphology and composition of the clay material were visualized and analyzed using a ThermoFisher Scientific Apreo C FESEM equipped with an EDS detector to examine the abundance of elements, presence of possible structural defects, and estimated membrane pore size. SEM micrographs were taken at an accelerating voltage of 20 kV.

2.6 Electrochemical techniques to study membranes selectivity

CV and LSASV of Fe(II) and Pb(II) on the surface of the clay-modified CPE are performed to study the selectivity of clay ceramic membranes. Carbon paste membranes/electrodes are fabricated by mixing graphite powder, clay membrane, and mineral oil serving as a binder of the paste. To ensure homogeneous and uniform mixing, the paste is mixed for a period of 10–15 min using a glass rod. The uniformly obtained pastes are then loaded into the sensing part of a 3D printed working electrode body with dimensions identical to the commercial glassy carbon electrode (7.0 cm long, 2.0 mm diameter) using a copper wire as an electric contact. Prior to any experiment, a new paste is loaded to the sensing part of the electrode and the sensing surface is smoothed on a wax paper. The bare graphite paste is prepared in the same way, without the addition of clay membrane powder, and is utilized for reference responses’ comparison. Table 1 presents the preparation of various ratios of composite membranes/electrodes with graphite powder, ceramic clay, and mineral oil ranging from 1 to 10 wt%. Electrochemical measurements are recorded on a Zahner Ennium Pro 43,243 electrochemical workstation. The electrochemical cell contained a clay-modified carbon paste membrane/electrode as a working electrode, a Pt wire counter electrode, and an Ag/AgCl (Sat. KCl) as a reference electrode.

3.1 Characterization of clay ceramic membranes

The porosity of the flat membrane discs sintered to 800, 900, 1000 and 1100 °C was measured using the conventional Archimedes technique (Lahnafi et al., 2020) and the results are tabulated in Table S1. The membrane discs sintered to temperatures below 800 °C were very weak, therefore they were not suitable for water porosity measurements. At higher sintering temperatures, the porosity of the membranes was found to decrease, indicating that the clay is better consolidated, developing better mechanical and chemical strength making the membranes more suitable for filtration in drastic conditions. Porosity values of 10.06 ± 1.56 %, 7.75 ± 1.61 %, 4.14 ± 0.74 % and 2.70 ± 0.40 % were found for the discs sintered to 800, 900, 1000 and 1100 °C, respectively. The membranes sintered to 1000 °C were presented with the best mechanical strength and appropriate porosity as demonstrated in a previous study (Elgamouz et al., 2019). Therefore, membranes sintered to 1000 °C were further investigated for the study of Pb(II) filtration using CV and LSASV.

Thermal analysis using TGA was performed on raw clay material to identify the temperature regimes at which drastic weight losses and possible phase transformations take place. The study provided insight into the effect of various temperature regimes on the porous structure and pore diameter of the membrane. Fig. 1 displays the TGA curve of raw UAE clay material when subjected to heat from 25 to 1200 °C in an alumina crucible at a heating rate of 20 °C/min. The figure clearly shows existing complex phase transformation and interactions. The total weight loss of the sample is observed to be 17.84 %. The process could be divided into five steps. The first mass loss of 4.67 % occurring at 195 °C is attributed to the loss of water molecules. The second step observed a weight loss of approximately 2.3 % between 195 and 488 °C, which can be correlated to the pre-dehydration step of kaolin. The pre-dehydration of kaolin occurs due to the rearrangement of the octahedral layers, first occurring at the OH on the surface (Balek and Murat, 1996). Between 488 and 662 °C, the loss of hydroxyl groups takes place due to the transformation of kaolinite to metakaolinite according to the following chemical reaction (Chen et al., 2004).

Close

Fig. 1

Thermal gravimetric analysis of the raw clay material used in the preparation of the clay membranes.

Export to PPT

The calcination of CaCO₃ resulted in the production of CO₂ and therefore the improvement of the porous membrane structure within the temperature regime of 662 °C to 849 °C, corresponding to a total weight loss of 5.81 %. Lastly, the calcination of Na₂CO₃ corresponds to a very minimal loss of mass above 849 °C.

The chemical composition of the raw and sintered clay at 1000 °C samples obtained from the XRF are presented in Table 2. The loss on ignition (LOI) is calculated at 800 °C for the raw clay, while sintered LOI presents the weight difference between the sintered clay at 800 °C and 1000 °C. Alumina, silica, iron, and calcium oxides were found to be the major components of the two samples. These are followed by magnesium oxide and traces of titanium and manganese oxides. Changes in the oxides percent were recorded when the clay is sintered to 1000 °C, indicating the rearrangement of the inorganic phases present in the clay. For instance, iron is known to change its oxidation state from Fe(II) to Fe(III). While loss on ignition for raw clay was in the range of 17.50 ± 0.24 %, this loss is due to phase change, mass loss, and iron oxide transformation (Elgamouz et al., 2019). Mass loss for the sintered sample at 1000 °C was recorded in the order of 0.53 ± 0.05, indicating that the carbonaceous compounds had been removed at lower sintering temperatures.

Fig. 2 summarizes XRD patterns of raw and sintered clay at 250 °C, 500 °C, 700 °C, 1000 °C and 1100 °C for 2 h in a muffle furnace. The XRD patterns of raw clay fraction presented in Fig. 2A indicate that the inorganic mixture originally consists of illite, smectite, and kaolinite minerals in the clay sample. Based on Fig. 2B, the XRD trends show no significant changes to kaolinite in the clay structure, however with increasing sintering temperatures, calcium carbonate, and quartz peaks diminish above 700 °C. This suggests that these materials either melt into the clay matrix or a phase conversion from kaolinite to metakaolinite takes place as illustrated by the TGA curves. The other existing phases that appear are quartz and chlorite. As a result of no further significant phase transformation occurring at higher temperatures, this implies that the clay ceramic membrane skeletal structure constitutes mainly metakaolinite, quartz, smectite, illite, and chlorite.

Close

Fig. 2

XRD pattern of (A) Raw UAE clay fraction and (B) Raw clay (RC) and clay sintered (CS) to various temperatures of 250 °C, 500 °C, 700 °C, 1000 °C, and 1100 °C. Abbreviations; I: illite; K: Kaolinite; Q: quartz; S: smectite; Ch: Chlorite.

Export to PPT

FTIR spectroscopy is used to analyze the composition/functional property of raw clay and sintered clay in an acquired range of 400–4000 cm⁻¹ and results are shown in Fig. 3A-C and tabulated in Table S2 (supporting document). The band at 3474 cm⁻¹ may correspond to the Al----O—H stretching mode of clay mineral (Kaolinite). Vibration peaks at 1733 and 1654 cm⁻¹ can be attributed to C⚌O stretching and H—O—H bending of water molecules which are adsorbed on the surface of the clay, respectively. The small bands at 1449 cm^-1 can be ascribed as C—H bending. The peak vibrations at 1373 and 992 cm⁻¹ can be attributed to Al-O and Si-O stretching vibrations of the clay mineral (Kaolinite). The absorption band at 1211 cm⁻¹ represents the stretching vibration of C—O of organic matter present in the clay material, which is seen to disappear after the sintering process (Zuo et al., 2018; Akanji et al., 2019). Peaks at 871 and 792 cm⁻¹ may be corroborated by the OH deformation linked to 2Al³⁺ and Si-O quartz, respectively. Furthermore, vibration peaks appearing at 645 and 517 cm⁻¹ can be correlated to Si-O-Si bending.

Close

Fig. 3

FT IR spectra of (A) overlay spectra of (a) RC, (b) cCS@250 °C, (c) CS@500 °C, (d) CS@700 °C, (e) clay CS@1000 °C and (f) CS@1100 °C, (B) zoom-in of 4000–2500 cm⁻¹ and (C) zoom-in of 4000–400 cm⁻¹.

Export to PPT

Fig. 4 (A-D) displays the SEM images of raw clay and clays sintered at 250 °C, 1000 °C, and 1100 °C. The clay samples showed a surface with a rough morphological structure. It was also noticed that the membrane porosity decreased for the sintered clay and the grains become fused. The more consolidated membrane structure is a result of particles agglomeration, creating a denser ceramic body. The observed SEM images signify that the membrane did not have any pinholes cracks, making the membrane suitable for the micro-filtration application. SEM micrographs of clay modified carbon paste membranes/electrodes presented in Fig. S1 (supplementary document) show difference in the morphology in comparison with the unmodified surfaces. The figure shows increased pore sizes and uneven distribution of the pores on the modified surface. Furthermore, Table S3 (supporting document) represents the atomic composition of powdered and paste materials for raw clay and clays sintered at 250 °C, 1000 °C, and 1100 °C. The prepared materials demonstrate the presence of O, Si, C, Al, Fe, Ca, Mg, and traces of Mg and K in clay mineral powders whereas excessive amounts of C (≥95 %) was observed in 5 % clay modified carbon paste materials.

Close

Fig. 4

SEM micrographs of (A) RC, (B) CS@250 °C, (C) CS@1000 °C, (D) CS@1100 °C.

Export to PPT

3.2 Electrochemical experiments

3.2.1

3.2.1 Effect of GR and GR/clay mass ratio

CV was employed to evaluate the electrochemical properties of the bare graphite and clay-modified CPE with different clay loadings (wt.%), where 10 mM K₄Fe(CN)₆ and 0.1 M KCl were used as the electroactive species and supporting electrolyte, respectively. The analysis was carried out at a potential range between 750 mV and −300 mV; current ranging between −500 mA and 500 mA and a scan rate of 100 mVs^-1. The CVs presented in Fig. 5 show well-defined oxidation and reduction peaks and an enhanced peak current at 0.25 V with increase in electron transfer kinetics with 5 wt% clay composite compared to 1, 3, and 10 wt% clay composite and bare graphite. This validated significant changes in the electrode surface property and showed good electrocatalytic activity of the modified electrodes towards K₄Fe(CN)₆. Hence, all further studies were carried out using a 5 wt% clay composite electrode as the optimum clay loading.

Close

Fig. 5

CVs in the presence of 10 mM K₄Fe(CN)₆ solution at a scan rate of 100 mVs^-1 for GR and GR/Clay mass ratio optimization.

Export to PPT

3.2.2

3.2.2 Effect of scan rate

Fig. S2 (supporting document) shows the influence of scan rate on the CV peak currents of the Fe(CN)₆^3-/Fe(CN)₆^4- redox couple in the scan rate range of 5 to 200 mVs^-1 for 5 % raw and sintered clay composite membranes/electrodes. Initially, CV analysis of bare graphite, 5 % raw clay, and 5 % clay sintered to 250˚C, 500˚C, 700˚C, 1000˚C, and 1100˚C were conducted at different scan rates in the presence of 10 mM K₄Fe(CN)₆ and 0.1 M KCl to examine the electroactive surface area of each membrane/electrode. Reproducible anodic and cathodic peaks were obtained owing to Fe(CN)₆^3-/Fe(CN)₆^4- redox couple at the membrane/electrode surface. The electroactive surface areas of the membranes/electrodes are calculated by applying the Randles-Sevcik equation (2) (Moudam and El Gamouz, 2016).

(2)

$$I_p=2.69\times {10}^5{n}^{\frac{3}{2}}A{D}^{\frac{1}{2}}C{\gamma }^{\frac{1}{2}}$$ where C is the concentration of the analyte (molL^-1), n is the number of electrons that contribute to the redox reaction on the membrane/electrode surface, A is the electroactive surface area of the membrane/electrode (cm²), D is the diffusion coefficient (cm² s⁻¹), and γ is the scan rate (Vs^-1). With the aid of the CVs presented in Fig. S2 (supporting document), the electroactive surface areas were calculated, then displayed as a bar graph in Fig. 6 and tabulated in Table 3.

Close

Fig. 6

Bar graph representing the calculated electroactive surface areas of different carbon paste membranes/electrodes using Randles-Sevcik equation.

Export to PPT

Table 3 The calculated electroactive surface area of each carbon paste membrane/electrode evaluated at different scan rates.

Electrode	Composition	Electroactive surface Area (cm2)
GR	Bare graphite	1.23 × 10−2
GR-5 %RC	5 % raw clay	3.07 × 10−2
GR-5 %CS@250 °C	5 % clay sintered to 250 °C	1.53 × 10−2
GR-5 %CS@500 °C	5 % clay sintered to 500 °C	1.38 × 10−2
GR-5 %CS@700 °C	5 % clay sintered to 700 °C	1.11 × 10−2
GR-5 %CS@1000 °C	5 % clay sintered to 1000 °C	7.67 × 10−3
GR-5 %CS@1100 °C	5 % clay sintered to 1100 °C	3.07 × 10−3

The linear increase of the cathodic peak current versus the square root of the scan rate indicates that the diffusion of Fe(CN)₆^3-/Fe(CN)₆^4- redox couple on the membrane/electrode surface took place (Raghavendra, 2021; Moutcine, 2020). In addition, a decrease in the electroactive surface of the membrane/electrode with increasing sintering temperature was noticed. This behavior can be explained by the decreased membrane porosity for the sintered clay and fusion of the grains. At higher sintering temperatures, the membrane becomes more consolidated due to particle agglomeration, generating a denser ceramic body and thus decreasing the electroactive area of the membrane/electrode surface. Hence, GR-5 %CS@1000 °C membrane/electrode was selected as the optimal membrane/electrode for subsequent studies.

3.2.3

3.2.3 Effect of supporting electrolyte and pH on Pb(II) ions filtration

The influence of various supporting electrolytes and pHs on Pb(II) ions filtration was evaluated using GR-5 %CS@1000 °C via a LSASV. A 0.1 M concentration of different supporting electrolyte such as HCl, acetate buffer, PBS buffer and boric acid/potassium chloride/sodium hydroxide buffer was used to investigate the effect of supporting electrolyte on the filtration of Pb(II) ions on the membrane/electrode surface. The pH of the medium was varied from 1.5 to 10.0 as shown in Fig. 7A and 7B. The high peak current at low pHs could be attributed to the presence of ionized Pb(II) species and charge carriers, large enough to increase the current. In acetate buffer at pH = 4.5, the decrease in the current could be due to the hydrolysis of the metal ions. During the hydrolysis process, hydroxide ions are produced which react with the metal ions to give the Pb(OH)⁺ species. At higher pHs (>7), the Pb(OH)⁺ species continue to react to give the precipitates Pb(OH)₂, causing little or no accumulation of Pb(II) ions by LSASV. Later, the Pb(OH)₃^- anions are generated. It is also worth noting that at this extent both the GR-5 %CS@1000 °C membrane/electrode surface as well as the hydroxides hold an overall negative charge, triggering electrostatic repulsion and causing lack of interaction between the clay minerals and the Pb(II) ions. Similar results were reported in previous studies (Akanji et al., 2019; Blaise, 2022; Ns et al., 2018). Moreover, based on these outcomes, 0.1 M acetate buffer at pH 4.5 was selected an optimal supporting medium and pH for further studies. This is explained by the destruction of the membrane/electrode surface at acidic mediums despite the high peak current responses and the precipitation of Pb(II) ions under basic conditions.

Close

Fig. 7

The effect of supporting electrolyte and pH on the filtration of Pb(II) ions using GR-5 %CS@1000 °C via LSASV (A), and peak current of Pb(II) ions at different pH (B), at accumulation time 60 s, accumulation potential −1.2 V and scan rate of 100 mVs^-1.

Export to PPT

3.2.4

3.2.4 Selectivity measurements via LSASV

Additional scan rate optimization studies were conducted with 5 % RC and GR-5 %CS@1000 °C in the presence of 1.0 mM Pb(II) in 0.1 M acetate buffer of pH = 4.5 as the target analyte and supporting electrolyte, respectively. The CVs of both 5 % RC and GR-5 %CS@1000 °C represented in Fig. 8A and 8B show clear peak current consistency with increasing scan rates. The existence of an anodic peak at −0.5 V affirmed the filtration of Pb(II) ions in the system. A slight shift in the potential was observed with an increasing scan rate from −0.5 V at 5 mVs^-1 to −0.4 V at 100 mVs^-1. Typically, the change in peak potentials with faster scan rates in irreversible waves usually gives information about the nature of the electron transfer, such that more potential is needed to induce the electron transfer when the electron rate constant is slow. With increasing scan rates from 5 to 100 mVs^-1, the concentration gradient takes longer time to react to the changes in the potential, causing the maximum concentration gradient to take place at potentials further from the initial potential at 5 mVs^-1. Another study reported that, the potential shift in the forward direction at fast scan rates is mainly attributed to the behavior of the heterogeneous kinetic control (Aoki et al., 2020). Moreover, a positive linear relationship was exhibited when plotting the anodic peak currents against the square root of the scan rates with R² values of 0.9927 and 0.9975 for 5 % RC and GR-5 %CS@1000 °C, respectively (Fig. 8C).

Close

Fig. 8

Effect of scan rate on 1.0 mM (Pb(NO₃)₂ in 0.1 M acetate buffer solution at pH = 4.5 using 5 %-GR-5 %RC membrane/electrode (A), using GR-5 %CS@1000˚C (B), Correlation between anodic peak currents against the square root of the scan rates (C), and Linear scan anodic stripping voltammogram of lead using various membranes/electrodes: GR, GR-5 %RC, GR-5 %CS@250 °C, GR-5 %CS@500 °C, GR-5 %CS@700 °C, GR-5 %CS@1000 °C and GR-5 %CS@1100 °C, at accumulation time 60 s, accumulation potential −1.2 V and scan rate of 100 mVs^-1 (D).

Export to PPT

To further understand the selectivity of the membrane towards Pb(II), the GR-5 %RC and the GR-5 %CS@1000 °C electroactive surface areas were determined at scan rates spanning between 5 and 200 mVs^-1, using the Randles-Sevcik equation (1). The Pb(II) diffusion coefficient value used was 9.45 × 10^-6 cm² s⁻¹ (Sato et al., 1975). The GR-5 %CS@1000 °C was presented with higher electroactive surface area of 2.03 × 10^-2 cm² towards Pb(II), as compared to GR-5 %CS@1000 °C electroactive surface area of 7.67 × 10^-3 cm², obtained by using the Fe(CN)₆^3-/Fe(CN)₆^4- redox couple. The significant enhancement of GR-5 %CS@1000 °C electrochemical surface area towards Pb(II) is mainly due to doping of GR with sintered membrane at 1000 °C (CS@1000 °C). This is due to the fact that at higher sintering temperatures, some of the clay mineral structures are changed and destroyed, creating a more consolidated and denser form of the membrane support. The fusion of the grains at this temperature plays a role in the selectivity of Pb(II) and Fe(II)/Fe(III) ions, such that GR-5 %CS@1000 °C is more selectivity towards Pb(II) ions than Fe(II)/Fe(III) ions. Another study reported that the hydrated radii, the hydration enthalpy of the cations as well as the space requirements in the pores of the clay minerals play a role in the selectivity order. It was reported that the high selectivity of Pb(II) ions compared to Fe(II)/Fe(III) ions is due to their complex formation with water molecules, such that Fe(II)/Fe(III) ions are more likely to form octahedral complexes, making them difficult to pass through the consolidated membrane support (Inglezakis et al., 2003).

The Pb(II) ions selectivity was further investigated by linear scan anodic stripping voltammetry. Voltammograms of Pb(II) obtained from GR, GR-5 %RC, GR-5 %CS@250 °C, GR-5 %CS@500 °C, GR-5 %CS@700 °C,GR-5 %CS@1000 °C and GR-5 %CS@1100 °C membranes/electrodes immersed in 1.0 mM Pb(NO₃)₂ in 0.1 M acetate buffer solution of pH = 4.5 as a function of accumulation potential are shown in Fig. 8D. An accumulation time of 60 s, accumulation potential of −1.2 V and scan rate of 100 mVs^-1 were used. It can be seen that the stripping peak of Pb(II) was the highest with GR-5 %CS@1000 °C. The mobility of the Pb(II) ions is found to increase when 5 % clay sintered to 1000 °C was utilized as the membrane/electrode, leading to an increase in the amount of reduced Pb(II) ions on the surfaces of the clay membrane/electrode. However, a drastic decrease in the peak current is observed using 5 % clay sintered to 1100 °C. This electrochemical behavior is confirmed by the increased stripping peak currents of Pb(II) in Fig. 8D.