The experiments were carried out using a Voltalab potentiostat (PGZ100) connected to a classical electrochemical “three electrodes” cell; Metrohm titration

tanks, having various capacities (1 < V tank   cm 3   < 150) as function of the available

volume of the sample to be analyzed could be used. The working electrode used is the manufactured carbon paste electrode (CPE) of which a picture was showed in Figure 1(a) and Figure 1(b). A reference electrode of Ag/AgCl/KCl_saturated and an auxiliary platinum electrodes were also used. The response of the system was observed via a computer using the software Voltalab master 4.

A standard solution of lead 1000 ppm Pro Analysis Sigma-Aldrich was used to

prepare lead solutions at different concentrations. The acids used for the preparation of the supporting electrolyte are nitric acid 70% DAEJUNG, sulfuric acid PROLABO 98%, hydrochloric acid 37% PANREAC and perchloric acid (70% - 72%, MERCK). Each dosage was preceded by thorough cleaning of electrodes followed by electrolytic cell cleaning with ethanol and rinsing with distilled water. The graphite carbon powder was used for the preparation of the carbon paste.

The carbon paste electrode (Figure 1(a) and Figure 1(b)) was prepared by mixing of 100 mg of graphite powder (ketjen black 300 supplied by akzonobel) and 50 µL of paraffin oil. The obtained mixture was kneaded until to produce a homogeneous paste, a part of which is introduced into a cylinder cavity (internal diameter 2.5 mm) with a pencil bar (acting as electronic collector). The resulting working electrode (the external disk area) was cleaned and dried using a filter paper that removes excess oil used in the preparation of the paste carbon (in order to avoid any influence of the oil on the conductivity of the electrode).

The working electrode was immersed into 50 mL of lead solution in which 129 μL of concentrated HClO₄ was added to obtain an electrolyte with a concentration of 0.03 M. The reduction of Pb²⁺ was carried out by chronoamperometry under a constant applied potential ranging from −800 to −1300 mV (preconcentration step), under stirring (magnetic bar at 200 rpm) throughout the electrolysis. The lead deposed into the carbon paste electrode was then oxidized by linear scanning voltammetry (redissolution step) between −700 and 0 mV. The presence of an oxidation peak in the obtained voltammogram attests the presence of the lead in the analyzed solution. Several parameters such as electrolysis time, accumulation potential were studied/optimized to improve the method sensibility for the determination of lead in drinking water. Response of system is an anodic stripping peak in which the amount of lead is proportional to the magnitude of the current.

Current-potential curves (not showed here) were plotted by Linear Sweep Voltammetry under stirring and in presence of atmospheric oxygen, using aqueous solution of perchloric acid (0.03 M), in absence and in presence of lead (200 µg/L of Pb²⁺ - 10⁻⁶ M); None curve exhibits any peak or any wave for the Pb²⁺ reduction. Other experiments with the same concentration of lead, without stirring or higher potential scan rates does not lead to more satisfactory results.

In order to try to get a signal for the lead, potentiostatic elecrtolyses were achieved to reduce Pb²⁺ in various conditions, and followed by the plotting of the current potential curves to the anodic direction. The applied cathodic potential chosen ranges from −700 to −1800 mV, and the electrolyses durations vary from 1 to 30 min. During the electrolyses the metallic lead deposited accumulates to the electrode surface, and it can reoxidize during the anodic striping.

Figure 2 presents the results, i.e. the temporal evolution of the current during the potentiostatic reduction of the Pb²⁺ (inset), and the current-potential curves obtained during the stripping (1 to 3).

*curve (1), plotted in absence of lead, towards the anodic direction, with the CPE immersed in the electrolyte, does not exhibit current in the whole potential range, meaning that no reaction occurs on the CPE; specifically nor dissolved oxygen nor proton reductions were observed.

*curve (2) obtained with the same electrode, immersed in the electrolyte containing 200 µg/L of Pb²⁺, but without preelectrolysis, exhibits at −0.7/−0.6 V an exponential shaped area, attributed to the reduction of solvent at CPE and probably include the reduction of the Pb²⁺. No anodic stripping peak was observed meaning that, during the scan, there is not accumulation of enough of lead on the CPE.

*curve (3) was obtained, under the same conditions as in (2), but after an electrolysis performed at E = −0.7 V. The cathodic part of the curve exhibits a nicely resolute diffusion wave at the potential range from −0.7 until −0.47 V. The plateau is due to the reduction of Pb²⁺ on the lead coated CPE and translates a Pb²⁺/Pb reversible system (in comparison with a Pb²⁺/Pb more irreversible system on the barre CPE). At the potential area −0.47 to −0.3, the curve presents a peak attributed to the anodic oxidation of lead previously deposited and accumulated at the CPE.

The previously obtained current-potential curves (Figure 2) does not clearly indicates a signal for the oxygen reduction on the CPE; this could mean that the O₂/H₂O system is irreversible on the CPE, but does not inform about its behavior on the lead covered CPE. Theoretically, under the required constant potential, both reductions (Pb²⁺®Pb and O₂®H₂O) could carry out independently, without mutual effects. Here, the effect of the deaeration time of the solution by nitrogen was examined in order to evaluate the influence of dissolved oxygen on the lead reoxidation current. Results were present in Figure 3 which gives the I = f (E) curves obtained for durations ranging from 2 min to 10 min, by LSASV, starting the potential-scan from −0.7 V until 0 V.

Curves clearly show that for bubbling times lower than 4 min, there is no lead deposition; the anodic striping peak of the lead oxidation appears for experiments submitted to a more than 4 min nitrogen bubbling. Reduction of oxygen could lead to various intermediates (such hydroxide radicals, hydrogen peroxide,…) which are able to modify the surface area of CPE, typically to create H-O-C bonds [39] . These terminal functions could enhance or strongly reduce the Pb²⁺ adsorption sites, and consequently the Pb²⁺ reduction current; in this case, at the applied potential of −1.3 V reduction of Pb²⁺ does not occur and all the current is used to reduce O₂.

Increasing the bubbling time before the preelectrolysis (4 to 10 min) enables the Pb²⁺ to be reduced to Pb; the absence of dissolved oxygen in the solution implies no reaction, so there is none modification of the surface area of the CPE and this enables the lead reduction and the deposit accumulation. For nitrogen

bubbling times higher than 10 minutes, curves does not show changes in the magnitude of the current of the anodic striping peak. Note that, this time can be easily and significantly reduced by decreasing the initial volume of the sample to be analyzed (a few minutes bubbling time /t < 3 min/ for a few cm³ /v < 5 cm³/ of analyzed volumes).

The applied potential must enable the reduction of Pb²⁺ on barre CPE, and also on this electrode “coated by metallic lead”. During this operation, all metallic species, having a higher potential than the potential of the system Pb²⁺/Pb, could be reduced and eventually deposited; consequently, if their adhesion is so enough, their oxidation could become possible, however at higher potential than this one observed for the lead oxidation. The effect of the applied potential during the accumulation phase (which enables the lead deposition), on the anodic stripping peak current, was examined in the range of −600 to −1800 mV_/Ag/AgCl/KCl and the results were presented in Figure 4. The anodic peak current increase when the applied cathodic potential decreases from −0.7 to −1.4 V. This strange behavior could be attributed to an irreversible behavior of the system Pb²⁺ on the CPE; in this potential range, the electronic transfer limits the reduction of Pb²⁺, thus, decreasing the applied potential increases the resulting cathodic current and consequently the accumulated metallic lead on the CPE. In fact, in this examined range, there is no a significant effect of the mass transfer limitation for the reduction of Pb²⁺ on the CPE, especially if the CPE was not entirely covered by the metallic deposit of Pb. For applied cathodic potentials, ranging from −1.4 to −1.8 V, the curve of the Figure 4 show a decrease of the magnitude of the stripping peak current. This decrease is attributed to the reduction of the solvent simultaneously to the reduction of Pb²⁺.

The produced gas H₂ could also creates local turbulences, which could cause a certain dispersion into the bulk of the electrodeposit (especially for dendritic lead deposited on the CPE, in the limiting current conditions). In addition with the loss of current attributed to the proton reduction, that is why the quantity of “the metallic lead oxidized” decreases with the applied potential, implying the

stripping peak current magnitude to decrease. The value of −1.3 V / A g / A g C / C l −

was selected as the optimum pre-concentration potential.

Figure 5 presents the voltammograms obtained after various durations of the potentiostatic electrolyses, carried out at the previous selected potential (−1.3 V), using the CPE immersed into a 100 µg/L solution of Pb²⁺.

The magnitude of the current of the anodic striping peak, linearly increase (inset) versus the time, for electrolyses durations lower than 20 min. This means that during these electrolyses, the cathodic current remains constant, and this indirectly imply that the CPE electrodes does not passivates; its surface does not significantly change by the lead deposits (because the linear increase of the peak current). This result is very important especially for the reproducibility and the repeatability of analyses of solutions containing very lows quantities (see following sections) of Pb²⁺.

For longer electrolyses, the increase of the magnitude of the current (inset of Figure 5) is higher than the augmentation expected by a linear evolution. This could be due to an increase of the electrochemical surface by the lead electrodeposited! Longer electrolyses times lead to a significant quantity of the metallic lead, which also acts as electronic collector and enables higher quantities of Pb²⁺ to be reduced. For the present study, an electrolysis time of 10 min was retained

in order 1) to stay in the linear part of the curve I_peak = f (electrolysis time) and, 2) to propose a rapid method for the lead titration.

Various acidic solution (nitric, sulfuric, hydrochloric and perchloric) were examined as eventual supporting electrolyte for the titration of the lead. Figure 6(a) presents the results in terms of the evolution of the magnitude of the striping anodic peak (oxidation of the accumulated lead on the CPE) against the nature of the acidic solution. Hydrochloric and perchloric acidic solutions provide the higher signals.

Two reasons could explain the obtained differences in the magnitude of the peak current against the nature of the electrolyte:

1) Higher diffusion coefficients of the Pb²⁺ could lead to higher cathodic current and consequently a higher accumulated quantity of the deposit;

2) A higher adhesion energy of the deposit on the CPE in the case of perchloric acid: this enables to keep all the reduced metallic lead at the electrode surface and avoid some dispersion because the stirring or the sponges nature of the deposit. The effect of the concentration of the supporting electrolyte choses (i.e. perchloric acid) was examined, expecting to increase the answer of the system, specifically for low concentrated solutions of lead.

In absence of electrolyte, when the applied potential is so cathodic enough to overcome the activation limitation for the reduction of Pb²⁺, the transfer of Pb²⁺ to the interface is limited by both diffusion and migration. Migration flux of Pb²⁺ increases when the electrolyte added concentration decreases, implying the

cathodic current of Pb²⁺ to increase; consequently, for the same electrolysis duration, the accumulated quantity of the metallic lead increases, so the current magnitude of the anodic stripping peak also increases. In the present study, the electrolyte was involved in large excess {6000 < ([electrolyte]/[Pb²⁺]) < 20,000} meaning that the migration of the lead must be negligible, and implying a slight effect of the electrolyte concentration on the magnitude of the anodic striping peak. However, results presented in Figure 6(b) indicate a strong dependence of the peak current versus the perchloric acid concentration! Indeed, decreasing the electrolyte concentration (from 100 to 10 mM) causes the magnitude of the current to strongly increases (from 5 to 80 µA), in fact the presence of perchloric acid appears to change the behavior of the electrode; the reduction of the lead Pb²⁺ become more irreversible probably by some disaggregation of the graphite powder by the perchloric acid. The concentration of electrolyte chosen to carry out the analyses is 0.03 M which 1) provides a higher current of the anodic striping peak, 2) ensure an ionic conductivity so high enough (to suppress possible disturbances by the matrix of the sample to be analyzed) and 3) reduce a possible negative effect of the high acidic content.

The effect of the potential scan rate r on the current magnitude of the anodic stripping peak was examined, expecting to increase the answer of the system, specifically for low concentrated solutions of lead. The results are the anodic curves obtained under various potential scan rates 80 £ r (mV∙s⁻¹) £ 300; they are similar to these presented in Figure 5 (not showed here to save place).

Theoretically, the peak current increases linearly with the r for adsorption (or a thin layer cell or a monolayer) limited systems and linearly with the square root of r for diffusion limited systems. Here nor I_peak = f (r^0.5) nor I_peak = f (r) exhibit a linear evolution, probably because a limitation of the Pb oxidation caused by both phenomena: desorption and then surface migration of the metallic lead into the pores of the CPE, until its surface, where take places the Pb oxidation.

Note that four-fold increase of the potential scan rate causes a slight increase of the peak potential (ΔE_peak = 60 mV) of the Pb oxidation system, and this translates that the Pb®Pb²⁺ is a slightly irreversible system. The scan rate of 200 mV∙s⁻¹ was chosen as a satisfactory value of the potential scan rate to carry out the study of the effect of the various operating parameters.

The influence of the Pb²⁺ concentration (ranging from 0.1 to 100 μg∙L⁻¹) on the magnitude of the current of the anodic stripping peak was studied, using the CPE. A part of the results were present in Figure 7(a) and the I = f (E) curve obtained for [Pb²⁺] = 3 µg/L was indicated Figure 7(b).

A satisfactory linear dependence was observed between the oxidation peak current and the Pb²⁺ concentration:

I peak ( in   A ) = 7.5 × 10 − 7 ×   [ P b 2 + ] ( in   μg / L ) R 2 = 0.99

The detection limit of the elaborated carbon paste electrode is 0.138 μg∙L⁻¹. The reproducibility was studied for three concentrations of Pb²⁺: 3, 10 and 20 μg∙L⁻¹ by repeating the experiments 6 times, and comparable results were obtained with a uncertainties lower than 5%.

The previous experiments were carried out in order 1) to validate the ability of the carbon paste electrode in the Pb²⁺-traces electrochemical analysis and 2) to select the optimum operating parameters enabling to obtain the highest answer

(e.g. the magnitude of the current) produced by the LSASV technique (thus enabling lowering the detection limit concentration). The results show that all the examined operating parameters strongly affect the shape and the size of the anodic striping peak. The optimum values of these parameters for the analysis of lead were summarized in Table 1. Besides, the previous results, specifically the calibration curve (Figure 7(a)) shows that, using the CPE (and the previously indicated electrochemical device), it is possible to detect and to quantify lead (Pb²⁺) at concentration equal to 3 µg∙L⁻¹ with uncertainties less than 5% (detection limit with the available apparatus: 0.1 µg/L of Pb²⁺); this concentration is lower than the accepted concentration fixed by WHO (10 μg∙L⁻¹).

The proposed “simple” analytical method was involved to directly determine lead (Pb²⁺) in various aqueous samples collected in Senegal (irrigation and wastewater samples), as well as in samples of drinking water of some various kind of buildings in Dakar City.

More than ten samples, were analyzed without any further treatment and the concentrations of Pb²⁺ were deduced from the stretched range of the regression equation (§3.7.). The Pb²⁺ concentrations found reaches for some samples until 270 μg∙L⁻¹, contents significantly higher than the limit of the WHO (World Health Organization) standards (10 µg∙L⁻¹) in tap water. The high concentrations of lead in tap water are mainly due to the presence of lead in the “house-drinking water network”. In the same order, to assess to the practical analytical utility of the method, distilled water samples that have been kept in contact for two hours with a pure lead material at the pH 7, were analyzed by LSASV at the unmodified carbon paste electrodes: result show a concentration of Pb²⁺, higher than the limit of the WHO (14.48 µg/L).

Figure 7(c) presents a reoxidation voltammogram obtained, under the optimum values of the operating parameters, with a tap water collected in a recent building in Dakar city. Two peeks were observed, the first one (E_peek~ −0.4 V) corresponds to the oxidation of lead electrodeposited (2 < Pb²⁺_µg/L < 3) and the second (E_peek~ −0.1 V) corresponds to the oxidation of an another heavy metal electrodeposited simultaneously to the lead, thus confirming the selectivity of the proposed method