1 Introduction

Arsenic (As) is one of the most toxic elements that seriously affects human health when it exceeds the maximum allowable limit of 10 μg L−1 in drinking water, recommended by the World Health Organization (WHO) [1]. Chronic exposure to this pollutant is associated with several health problems such as arsenicosis, hyperkeratosis, and different types of cancer [2–4]. The contamination of groundwater with As, frequently intended for human consumption, is a serious problem around the world that affects approximately 50 million people [5]. In natural waters, inorganic forms of arsenic are generally found as dissolved species, forming oxyanions that involve As(V) and As(III). The former is more widespread in groundwater and soils, and is generally found in drinking water [6, 7], whereas, As(V) species are 4–10 times less soluble and 60 times less toxic in water than As(III) [8].

As a consequence, it is important to have an accurate, fast, and sensitive method to detect and monitor this toxic element in drinking water. Many analytical methods have been developed to detect low concentrations of this environmental pollutant. Inductively coupled plasma-mass spectrometry [9], atomic absorption spectroscopy [10], and hydride generation atomic fluorescence spectrometry [11] showed good sensitivity and a low arsenic detection limit. However, these techniques have high operation costs and expensive instrumentation, while it is not possible to use them in field measurements [7, 12]. Electroanalytical methods offer high sensitivity, easy operation, low cost instrumentation, and portability, enabling their use in field analysis [12–14]. Among electrochemical techniques for As detection are the voltammetric techniques [15]. Voltammetric methods for determining As in water have long existed [16], as it is the case for polarography using a mercury drop electrode (HDME), which has been used to detect this element [7]. Also, stripping techniques have been widely used for As detection since they are characterized by a high selectivity for arsenic, require minimal sample preparation, are highly sensitive, and can be used with low detection limits [17]. Detection limits are generally three or four orders of magnitude more sensitive than polarographic methods. Due to different methods of accumulation and determination of the amount of analyte immobilized on the electrode surface, there is a large family of techniques such as anodic and cathodic stripping, differential pulse, or square wave voltammetry [18, 19].

During the past decades, electrochemical sensors have been used for As detection. The hanging mercury drop electrode [20–22] and the mercury film electrode [23] have been widely used in As species detection. However, due to the rather poor solubility of elementary As in mercury, potential toxicity, and operational limitations [17], these sensors were replaced by other metallic material electrodes, such as platinum [24], silver [25], copper [26], and gold [12, 27–29]. The gold-based electrodes have been specially conducted to improve analytical ability for As detection in water [30]. Nevertheless, gold-based electrodes have some disadvantages such as the high cost of the noble metal and a difficult cleaning procedure [31]. Also, techniques such as anodic stripping voltammetry (ASV) and anodic stripping potentiometry (ASP) which use gold electrodes are often affected by interferants such as copper ions which form intermetallic compounds with arsenic [32]. However, the constant current cathodic stripping potentiometric (CSP) method for the determination of As has been reported [23], based on a glassy carbon mercury film electrode in the presence of copper(II) ions, obtaining a good linear range of 10–200 µg L−1, and interferences from organic and inorganic substances were successfully overcome by the use of l-cysteine as a reducing agent.

One of the main problems in the detection of As(V) is its irreversible electrochemical behavior on the electrode surface [33]. To solve this problem and determine the total amount of arsenic, it is necessary to reduce As(V)–As(III). The proposed strategies, in combination with electrochemical techniques, were the use of various chemical reductants in strong acid media such as: KI [34–36], NaBr + N2H4 + H2SO4 [37], N2H4 + HCl + HBr [38], HCl + NH2OH [36, 39], Na2SO3 [12, 40], gaseous SO2 [37], and l-cysteine [23, 41]. However, the excessive use of reducing agents could contaminate and even be harmful to human health and the environment. In order to overcome these limitations, a lot of effort has been put into electrode surface modification and materials by using metal particles, carbonaceous nanomaterials (carbon nanotubes and graphene), and even enzymes (arsenite oxidase) immobilized on electrodes surface such as carbon-based [42], including carbon paste electrodes (CPE) [43]. Also gold-plated carbon paste electrodes have been reported for the electrochemical detection of arsenic, essentially using stripping techniques [44, 45]. Previous studies have reported the application of CPEs, modified with different crystalline forms of iron compounds, for the determination of As(III) and total arsenic in acidic media, mitigating the irreversibility of the reduction process of this element on the electrode surface [46]. However, it is also important to consider the chemical composition of the material and crystalline forms of iron, which can be stabilized at working pHs. Therefore, a detailed physicochemical characterization of the electrode material is fundamental for a better understanding of the detection process of As(V) in water.

The present work focuses on the characterization of the carbon paste electrode (CPE) modified with iron particles and its possible application on the electrochemical detection of As(V). The electrodes provided advantages such as very easy modification with iron particles, inexpensive construction, and simple renewal of electrode surface.

2 Materials and methods

2.1 Reagents and instruments

All chemicals used in this study were reagent grade. An iron solution of 1 mol L−1 was prepared from FeCl3·6H2O salt (Fermont) for further use in the carbon material modification. Diluted solutions of As(V) were prepared daily from a stock solution of 205 mg L−1 of Na2HAsO4·7H2O salt (Sigma-Aldrich). A potassium hexacyanoferrate solution of 10 mmol L−1 was prepared from a K4[Fe(CN)6]·3H2O salt (Sigma-Aldrich). Deionized water (18.2 MΩ cm−1) was used to prepare all solutions. The pH of the solutions was adjusted with a HNO3 and/or NaOH concentrated solution purchased from Fermont. For the electrolytic medium preparation, NaNO3 salt from Fermont was used. The graphite powder used for electrode preparation was purchased from Sigma-Aldrich, with particle diameter less than 20 µm. Paraffin oil from Fluka (ρ = 0.85 g cm−1) was used as binder agent in the preparation of carbon paste.

All voltammograms were performed using a potentiostat/galvanostat model VSP SAS Biologic controlled by EC-Lab software V 10.23, and a three-electrode cell system with an Ag/AgCl/KCl(sat) reference electrode. To avoid contamination of the reference electrode, a salt bridge containing a free analyte electrolyte solution of NaNO3 0.1 mol L−1 was used. A graphite rod was used as auxiliary electrode and the modified CPE was the working electrode.

2.2 Graphite powder modification and carbon paste electrode preparation

Iron particles were immobilized in the graphite powder by the slurry method. The iron(III) chloride (1 mol L−1) underwent hydrolysis to produce an acidic solution (at pH 2.0) which was homogeneously mixed with the graphite powder in a vortex. The mixture was dried for 72 h at 80 °C, and with this procedure it was possible to obtain iron hydro(oxide) compounds as was confirmed with X-ray diffraction studies (see Sect. 3.2). The carbon paste was prepared by carefully mixing Fe-modified graphite powder with paraffin oil (weight ration 55:45) in an agate mortar. In order to avoid impurities in the synthesis of modified graphite powder and in the carbon paste preparation, all the containers were immersed in HNO3 10 % solution for 24 h and then rinsed with deionized water before conducting experiments.

For the CPE preparation, a polypropylene tube (2.4 mm i.d.) was filled with the carbon paste and pressed onto a flat surface using a plunger. Subsequently, a copper wire was inserted inside the packed tube, which served as a connection to the potentiostat. The electrode surface was renewed after each voltammetric experiment by slightly pushing and removing the excess of carbon paste out of the tube tip.

The appropriate ratio of modified graphite powder and binder agent in the carbon paste elaboration was confirmed, since a well-defined voltammetric response of the typical reversible redox system made up of [Fe(CN)6]3−/[Fe(CN)6]4− 10−3 mol L−1 and NaNO3 0.1 mol L−1 at pH 2.5 was observed. Fig. S.1 (see Supplementary material) shows the voltammograms of CPE-Fe, unmodified carbon paste electrode (UM-CPE), and glassy carbon electrode (GCE) immersed in K4[Fe(CN)6] solution. A slight increase in the anodic and cathodic peak current intensity of the CPE-Fe compared with the UM-CPE is observed (Fig. S.1). This increase may be due to the electrical conductivity of iron particles immobilized on the modified carbon paste [30].

2.3 Material characterization

2.3.1 Surface area and pore volume

Surface area and pore size distribution of the Fe-modified graphite and graphite powder were determined by nitrogen adsorption–desorption isotherms at 77 K using a Micromeritics ASAP 2020 system. The specific surface area was calculated using the Brunauer–Emmett–Teller (BET) equation and the pore size distribution was determined using the density functional theory (DFT) method.

2.3.2 Surface charge distribution and point of zero charge

Surface charge distribution and point of zero charge (PZC) of the Fe-modified graphite and graphite powder were obtained by using a Mettler Toledo T70 automatic titrator following Bandosz et al. [47] procedure. For this procedure, 10 mg of the sample were placed in the reaction vessel with 50 mL of NaCl 0.01 mol L−1. The solution was continuously stirred for 24 h. Finally, the sample was titrated with NaOH 0.1 mol L−1 and the solution was continuously purged with N2 to avoid atmospheric CO2 interferences. After titration, the surface charge of materials and pKa distribution of graphite powder were calculated using the SAIEUS method [48].

The slurry pH method was used to determine the PZC of modified carbon paste. Different amounts of carbon paste (from 0.01 to 1.25 g) were added to 10 mL of deionized water in glass tubes. The solutions were deoxygenated by constant bubbling of N2 for 5 min. The tubes were sealed and stirred constantly for 72 h at 25 °C. After that, the pH of each tube was measured and plotted as a function of the amount of carbon paste used. The asymptotic part of the graph corresponds to the PZC of material.

2.3.3 Potentiometric titration of graphite powder

Classical Boehm titrations were carried out to quantify the protic and aprotic oxygenated groups of graphite powder (carboxyls, lactones, phenols, and carbonyls) [49] using an automatic titrator Mettler Toledo T70. In four reaction vessels, 10 mg of graphite were placed and 25 mL of 0.1 mol L−1 solutions of, respectively, NaHCO3, Na2CO3, NaOH, and NaOC2H5 were added. The solutions were continuously stirred for 5 days at 25 °C. After that, the samples were titrated with 0.1 mmol L−1 of HCl. Titration experiments were carried out under a N2 atmosphere to eliminate CO2 interferences. The titration with NaOH gave the concentration of all acidic groups (carboxylic, lactonic, and phenolic groups), the Na2CO3 neutralized the carboxylic and lactonic sites, while the NaHCO3 neutralized only the carboxylic sites. The amount of protic groups (meq/g) was calculated using the Eqs. (1) and (2).

$${\text{ASC}}\;{ = }\;\frac{{V_{\text{in}} (C_{\text{in}} - C_{\text{fin}} )}}{m} \times 1000$$
(1)
$$C_{\text{fin}} = \frac{{V_{\text{T}} C_{\text{T}} }}{{V_{m} }}$$
(2)

where ASC is the active sites concentration (meq g−1), V in is the initial volume of the neutralizing solution (L), C in is the initial concentration of the neutralizing solution (eqL−1), C fin is the final concentration of the neutralizing solution (eqL−1), m is the mass of graphite powder (g), V T is the volume of the titrant solution (mL), C T is the concentration of the titrant solution (eqL−1), and V m is the volume of the neutralizing solution with graphite powder (mL).

2.3.4 X-ray diffraction studies

The iron hydro(oxide) crystalline phases presented in the modified carbon paste (CP-Fe), were obtained in a Bruker D8 Advanced diffractometer using CuKα radiation (λ = 1.5418). Before the analysis, the samples were dried for 24 h at 80 °C. The XRD patterns were obtained with a step size of 0.02° 2θ at 10 s per step.

2.4 Adsorption experiments

Adsorption experiments were carried out in a 50 mL Falcon conical flask, in which 15 mg of CP-Fe powder were in contact with initial As(V) concentration, from 20 to 1500 µg L−1 at pH 2.5 and 25 °C. Two adsorption isotherms were performed with and without electrolytic media of NaNO3 0.1 mol L−1, in order to evaluate the competition between As(V) species and NO3 − ions, for adsorption active sites. The experimental solutions were kept under constant stirring. The pH of each sample was adjusted daily with NaOH, and/or HNO3 0.1 mol L−1 until equilibrium was reached. The final arsenic concentration was determined by inductively coupled plasma atomic emission spectroscopy (ICP-AES), in a Varian spectrometer 730-ES at a wavelength of 188.98 nm. The adsorption capacity of the modified material was calculated by a mass balance, and the data were adjusted by either the Freundlich or Langmuir model.

2.5 Electrochemical analysis

The differential pulse voltammetry (DPV) technique was used for arsenic detection in a NaNO3 0.1 mol L−1 electrolytic solution, at pH 2.5 and room temperature (25 °C ± 2). The NaNO3 salt was used as supporting electrolyte because it provides the highest anodic peak currents among other electrolytes in arsenic detection [50]. Argon was bubbled through the solution for at least 10 min before the electroanalytical experiments started. In order to suppress the undesirable background currents within the potential range of interest and enhance the sensitivity of the electrodes [51], a pretreatment potential of −1.10 V for 5 s was applied in arsenic-free solutions. Afterward, the electrode was transferred to electrolytic As(V) solution and a reduction potential deposition step of −1.10 V for 180 s was applied under stirring, followed by the recording of the anodic differential pulse voltammogram by scanning from −0.70 to 0.70 V (vs. Ag/AgCl/KCl(sat)) at a scan rate of 50 mV s−1, pulse amplitude of 100 mV, and pulse period of 2.0 ms. A reduction time of 180 s showed the highest signal-to-background (S/B) ratio for stripping voltammetry of arsenic. During the experiments a continuous argon flow was maintained over the solution.

3 Results and discussion

3.1 Optimization of the instrumental parameters

In order to optimize the arsenic detection signal, the percentage of iron particles in the carbon paste was evaluated by DPV. Mixtures of carbon paste and iron particles (10, 8, 6, and 4 % w/w) were tested. The mixture of 4 % showed a well-defined arsenic voltammetric response, and was chosen to characterize the electrode material and perform the electroanalytical experiments.

3.2 Carbon material characterization

Table 1 shows that the surface area and pore volume slightly decreased after modification with iron hydro(oxide) particles due to the obstruction of some meso- and macropores. The surface area of Fe-graphite modified (GP-Fe) at 4 % decreased 11.7 % in comparison with the unmodified graphite powder (UM-GP). Similarly, the pore volume of meso- and macropores decreased by 55.4 and 6.3 %, respectively. The reduction in pore volume of mesopores suggested that the particles size of the iron hydro(oxides) immobilized on graphite powder ranged from 2 to 50 nm. The presence of meso- and macropores in the modified carbon material will favor the arsenic diffusion on the electrode surface, facilitating its electrochemical detection. The micropores in the materials were null.

Table 1 Physical characteristics of unmodified and modified carbon materials
Full size table

XRD diffraction analysis of the CP-Fe at 4 % before and after pretreatment potential of −1.10 V for 5 s, is shown in Fig. 1A and B respectively. The XRD pattern of the unmodified carbon paste (UM-CP) without pretreatment potential is shown in Fig. 1C. For the three cases, characteristic peaks of graphite around 27°, 54.5°, and 86.5° were observed. Also, the peak of XRD pattern of the UM-CP near 27° was very sharp and indicates an interlayer distance of 3.34 Å [52]. On the other hand, the XRD patterns of CP-Fe at 4 % without pretreatment potential (Fig. 1A) did not show an appreciable change compared to that of UM-CP, indicating that the iron hydro(oxides) particles are either amorphous or very small to diffract [53]. For better observation of small peaks in Fig. 1A and B, an approach of the region between 40° and 90° 2θ was performed (Fig. 2). The XRD pattern of CP-Fe before and after pretreatment (see Fig. 2A and B respectively) at 32°, 42.5°, and 44.5° agreed with the pattern reported for ferrihydrite. Similarly, peaks at 45.5°, 56.5°, 59°, 75°, 77°, and 84° corresponding to goethite have been reported. In addition, a decrease in peak intensity after pretreatment potential (Fig. 2B) is observed, indicating that the iron crystalline forms presented in the material suffered an electrochemical dissolution when a reduction potential is applied during the pretreatment [54]. Due to its poor structural organization, ferrihydrite is thermodynamically less stable than a more organized crystalline form, i.e., goethite [55]. Hence, the reactivity of the iron crystalline form depends on its degree of organization, which implies that the amorphous ferrihydrite exhibits greater dissolution kinetics than the goethite [54].

Fig. 1
figure 1

XRD patterns of carbon paste modified with iron particles (CP-Fe) at 4 % with a step size of 0.02° 2θ at 10 s per step. (A) Before pretreatment, (B) after pretreatment, and (C) unmodified carbon paste without pretreatment

Full size image
Fig. 2
figure 2

XRD patterns of the region between 40° and 90° 2θ of carbon paste modified with iron particles (CP-Fe) at 4 % with a step size of 0.02° 2θ at 10 s per step. (A) Before pretreatment and (B) after pretreatment

Full size image

The Boehm titration results for the UM-GP showed that the amounts of the carboxylic, lactonic, phenolic, and carbonyl groups were 0.03, 0.23, 0.39, and 1.43 meq g−1, respectively. The presence of these oxygenated groups in the material can contribute to the electron transfer in the electrochemical processes, because they exhibit redox activity when immersed in the appropriate electrolyte [56]. The pKa distribution indicated that the pKa of the functional groups predominantly present in the UM-GP was 10.9. This value suggests that the oxygenated group mostly present in the material was pyrone type, which has carbonyl groups in its chemical structure. The oxygen atoms at the edges of graphene layers, i.e., the pyrone type, can satisfactorily explain the redox behavior of carbon materials, because it is known that these oxygenated groups can catalyze the reactions involving electron transfer between electro-active species [57].

The PZC values of GP-Fe at 4 %, CP-Fe at 4 %, and UM-GP, are shown in Table 1. It can be observed that the PZC value of UM-GP decreased when it was modified with iron particles. This may be due to the material remaining acidic, consequence of the iron particle impregnation method in the graphite powder. The variation of PZC values suggests that the arsenic adsorption processes would be affected due to electrostatic interactions between CP-Fe and H2AsO4 − species, which predominate at pH 2.5. Moreover, the changes of PZC values can influence the electrochemical properties of the electrode material.

3.3 Adsorption isotherms

The affinity between CP-Fe and As(V) in solution was evaluated by adsorption experiments at pH 2.5 and 25 °C, which was the same acid media used for the electrochemical detection of this pollutant. The adsorption isotherms of the CP-Fe material in As(V) solution with and without NaNO3 0.1 mol L−1, which was used as supporting electrolyte during the electrochemical detection, are shown in Fig. 3. A further comparative point of UM-GP in As(V) solutions is also shown. Experimental data were fitted with the Langmuir model (Table S.1, see Supplementary material), since this adjusted better than the Freundlich model. It can be observed (Fig. 3) that the As(V) adsorption capacity decreased 25 %, from 612.7 to 459.6 µg g−1 in the presence of electrolytic medium, due to competition between NO3 − and H2AsO4 − anions for the active adsorption sites [58, 59], since it has been reported that NO3 − ion interacts well with the iron hydro(oxide) surface by outer-sphere complexes [58]. Moreover, the UM-GP showed a very low As(V) adsorption capacity of 1.9 µg g−1 (equivalent to 0.3 %) compared with the CP-Fe without electrolytic medium (612.7 µg g−1). This substantial increase in the As(V) adsorption capacity is due to the high affinity between iron hydro(oxide) complexes and H2AsO4 − anions [60]. The XRD analysis showed that ferrihydrite and goethite were present in the modified carbon paste, which has been reported to have a high adsorption capacity of arsenic species [58, 61, 62]. It is known that arsenic is adsorbed on iron hydro(oxides) by inner-sphere surface complexes, which may be monodentate, bidentate-mononuclear, and bidentate-binuclear [63–66].

Fig. 3
figure 3

As(V) adsorption isotherms of CP-Fe at 4 % at pH 2.5 and 25 °C

Full size image

In the electrochemical detection, the adsorption process may be advantageous because it provides an improved electroanalytical response by analyte pre-concentration on adsorbent material [67]. The affinity between arsenic species and iron hydro(oxide) complexes is an important factor which favors the electrochemical detection of As(V) by the CPE-Fe electrode, since the functionalization of carbon materials (i.e., activated carbon) with iron (Fe) hydro(oxide) particles and their application to remove As(V) from water have been reported [53, 68, 69]. Due to low costs, high selectivity, and adsorption capacity, iron(III) compounds and elemental iron are widely used for water purification from arsenic because As(V) and As(III) are well adsorbed on Fe(III) hydroxides and oxides in the pH range of natural water in the presence of oxygen [68].

3.4 Electrochemical detection of As(V) on the CPE-Fe

Figure 4A shows the arsenic voltammetric response on CPE-Fe at 4 % at pH 2.5 with a peak around Epa ≈ −0.08 V, which may involve the oxidation process of As(0)–As(III). In the same figure, one second and less intense peak around Epa ≈ 0.1 V was observed, which may correspond to the oxidation process of complexes of arsenic not defined. Furthermore, it can be seen that the current intensity of both peaks was sensitive to the concentration of arsenic in solution, i.e., Curves (a) and (b) corresponded to 10 and 25 µg L−1 of As(V), respectively. The voltammetric response of Curve (c), after exposing the UM-CPE to the As(V) solution, did not show a defined peak, suggesting that the iron particles on the electrode surface improved the arsenic detection of CPE-Fe. In addition, Fig. 4B shows the voltammetric response of CPE-Fe in NaNO3 0.1 mol L−1 solution without arsenic (Curve a). Broad peaks with high current were observed. This response is associated with the oxidation process of iron particles immobilized on carbonaceous material, indicating that there is a contribution of their oxidation process in arsenic detection. As mentioned above, in electrochemical arsenite detection studies it was found that the NaNO3 electrolyte provided the highest well-defined anodic peak currents among other electrolytes [50].

Fig. 4
figure 4

(A) Voltammetric response of CPE-Fe at 4 % exposed to different As(V) concentrations in 0.1 mol L−1 NaNO3 solutions at pH 2.5: (a) 10 μg L−1, (b) 25 μg L−1, and (c) UM-CPE exposed to a solution of As(V) 25 μg L−1. (B) Voltammetric response of CPE-Fe at 4 % exposed to: (a) only 0.1 mol L−1 NaNO3 solution, (b) As(V) 25 μg L−1, and (c) UM-CPE exposed to a solution of As(V) 25 μg L−1. Deposition step −1.10 V for 180 s. Anodic stripping scan −0.70 to 0.70 V (vs. Ag/AgCl/KCl (sat)). Scan rate 50 mVs−1. Pretreatment potential −1.10 V for 5 s. The solutions were saturated with argon at 25 °C

Full size image

It has been documented that the electrochemical detection of this pollutant is carried out in acid media. The main problem with arsenic electrochemistry is the electrochemically irreversible nature of its reduction [70]. For this reason, it is necessary to use strongly acidic media to produce the arsenic reduction process on the electrode surface [33]. In this study, the effect of pH solution on the electrochemical response of As(V) was studied (see Fig. S.2 of Supplementary material). It was observed that the peak current increased by increasing the pH in the range of 1.0–3.0. At pH 2.5, the peak reached a maximum, and at pH 3.0 there was no signal. Therefore, the solution of pH 2.5 value was used for all the studies reported here, which was a bit higher than the working pH using paraffin-impregnated graphite electrodes modified with iron oxides such as limonite, goethite, or hematite [71]. On the other hand, it was expected that the interface of the CPE-Fe at 4 % electrode surface to be positively charged, due to the PZC of the modified carbon paste, from pH 1 to 2, in which the detection was performed. This implies that there is an electrostatic attraction between the H2AsO4 − species and the surface of the material. This attraction decreases with increasing pH, due to electrostatic repulsions, affecting the adsorption process of this contaminant. Therefore, it would be expected that the current intensity decreases as the pH of the solution increases. However, the opposite occurs, suggesting that the redox process occurring on the surface determines the detection process. Moreover, the predominant As(V) species at working pH 2.5 are H2AsO4 − and H3AsO4. Also, the As(III) species which is H3AsO3 at working pH can be present (see Fig. S.3 Supplementary material). Hence, these arsenic species participate in the adsorption and reduction process on the electrode interface.

The behavior of arsenic species on the electrode surface suggests an interaction between this pollutant and iron particles (Fig. 4). Also, during the electrochemical pretreatment of the modified carbon paste electrode, Fe(II) species are generated on the electrode surface due to electrochemical reduction from Fe(III) to Fe(II) [54]. Therefore, an arsenic reduction mechanism can be proposed. Figure 5 shows the arsenic detection mechanism on the electrode surface. Firstly, a chemical reaction can be carried out in which H2AsO4 − and H3AsO4 species are adsorbed by iron crystalline forms (ferrihydrite and goethite) present on the electrode surface, through inner-sphere surface complexes. Afterward, As(V) species can be chemically reduced to H3AsO3 by Fe(II) generated during the electrochemical pretreatment of the electrode. It has been observed that not only Fe(III) can oxidize chemically As(III)–As(V), but also Fe(II) is able to reduce chemically As(V)–As(III) at low pH [72]. These reactions can be coupled to an electrochemical step in which H3AsO3 is reduced to zero-valent arsenic on the electrode surface during electrochemical deposition at −1.10 V. Finally, zero-valent arsenic is stripped during the anodic scan. The whole reduction process can be described as summarized below:

Fig. 5
figure 5

Arsenic detection mechanism. Stage C: adsorption of H2AsO4 − ions on the electrode surface, (1) electro-dissolution of Fe(III)–Fe(II) and (2) chemical reduction of As(V)–As(III). Stage E: electrochemical reduction of As(III) to zero-valent arsenic (3)

Full size image
  1. 1.

    Electro-dissolution of Fe(III)–Fe(II) in the electrode surface.

    $${\text{Fe}}_{2}{\text{O}}_{3}\,{+}\,{\text{6H}}^{+}+{2}{\rm e^{-}} \rightleftharpoons 2 {\text{Fe}}^{2+}{+}\,{\text{3H}}_{2}{\text{O}}$$
  2. 2.

    Chemical reduction of As(V) to As(III) on the electrode surface

    $${\text{H}}_{2}{\text{AsO}}_{4}^{-}{+}{\text{3H}}^{+}+{\text{2Fe}}^{2+} \rightleftharpoons{\text{H}}_{3}{\text{AsO}}_{3}\,{+}\,{\text{2Fe}}^{3+}+{\text{H}}_{2}{\text{O}}$$
    $${\text{H}}_{3} {\text{AsO}}_{4} + {\text{2H}}^{ + } + {\text{2Fe}}^{{2 + }} \rightleftharpoons {\text{H}}_{3} {\text{AsO}}_{3} + {\text{2Fe}}^{{3 + }} + {\text{H}}_{2} {\text{O}}$$
  3. 3.

    Electrochemical reduction of As(III) to As(0) on the electrode surface

    $${\text{H}}_{3}{\text{AsO}}_{3}+{\text{3H}}^{+}+{\text{3e}}^{-} \rightleftharpoons {\text{As}}^{0}+{\text{3H}}_{2}{\text{O}}$$

3.5 Calibration and limit of detection (LOD)

In order to determine the sensitivity of CPE-Fe electrode, a calibration curve for the As(V) detection was established using the optimal voltammetric conditions described above. Figure 6 shows that the anodic current height for arsenic signal increased linearly with its concentration using the CPE-Fe electrode at 4 %. The slope and correlation coefficient (R 2) were 0.035 µA µg−1 L and 0.995, respectively. The good linear fit of the electrode calibration from As(V) 25 to 1 × 103 µg L−1 confirmed that the voltammetric response corresponded to the arsenic stripping processes after its electrodeposition at the electrode. The LOD was determined to be 19.4 µg L−1. It was also possible to detect As(V) at a concentration of 10 µg L−1 as it can be observed in the voltammetric response of Fig. 4A (curve a). But, if this value of concentration is considered there is not a good linear correlation between As(V) concentration and the anodic current height. Also, the ratio of the intensity of the current peaks of curves a and b of Fig. 4A suggests that at low As(V) concentrations, the kinetics of the reduction process of this pollutant is different. However, the detection limit for As(V) species is found to be appropriate for the WHO guidelines in drinking water. Additionally, this electrode possesses some key advantages such as ability to control iron content, economical use of raw materials, ease of fabrication, and ambient process conditions. The electrode was stored at room temperature without any special care. Also, 1 year of storage did not shorten the lifetime of the electrode. It can have practical utility in arsenic sensing. Further studies to apply this electrode in model solutions containing other metal ions and to monitor tap water are still in progress, and they can be part of another report.

Fig. 6
figure 6

Calibration curve of CPE-Fe at 4 % exposed to different As(V) concentrations in 0.1 mol L−1 NaNO3 solutions at pH 2.5 and 25 °C

Full size image

4 Conclusions

The results demonstrated that the As(V) adsorption capacity of the carbonaceous material increases once it is modified with iron particles, due to the high selectivity of its coordination compounds for arsenic. The electrolytic media significantly decrease the As(V) adsorption capacity due to the competition between NO3 − and As(V) ions for active adsorption sites. The voltammetry response of arsenic is better defined in a CPE-Fe due to affinity between iron hydro(oxides) and this element. The crystalline forms of iron hydro(oxide) complexes identified were ferrihydrite and goethite, which were effective to pre-concentrate As at the electrode surface. The reduction mechanism of As(V) is given by chemical and electrochemical reactions, in which the arsenate ions interacts with iron species immobilized on the electrode surface. The CPE-Fe exhibited good response toward total As, with a minimum detection limit around 10 µg L−1 (133.3 nM) which is the maximum allowable limit recommended by the World Health Organization (1993). Accordingly, this electrode could be potentially applied as a sensor for the detection of this pollutant in natural water samples. However, considerable efforts are still needed to develop electrode materials and analytical procedures for reliable detection of As with sub-ppb levels in the presence of endogenous toxic metals and organics in water matrices.