Electrochemical Detection of Sulfite by Electroreduction Using a Carbon Paste Electrode Binder with N-octylpyridinium Hexafluorophosphate Ionic Liquid

Abstract

Sulfite is a widely used additive in food and beverages, and its maximum content is limited by food regulations. For this reason, determining the sulfite concentration using fast, low-cost techniques is a current challenge. This work describes the behavior of a sensor based on an electrode formed by carbon nanotubes an ionic liquid as binder, which by electrochemical reduction, allows detecting sulfite with a detection limit of 1.6 ± 0.05 mmol L⁻¹ and presents adequate sensitivity. The advantage of detecting sulfite by reduction and not by oxidation is that the presence of antioxidants such as ascorbic acid does not affect the measurement. The electrode shown here is low-cost and easy to manufacture, robust, and stable.

1. Introduction

Preservatives and food additives are substances that are added to some foods in small amounts with the aim of storing them for a long period of time. The use of additives and preservatives is essential to maintain its quality, healthiness, taste, appearance, and flavor for a given period [1]. Sulfite is widely used as a preservative in foods and beverages during their preparation, storage, and distribution processes [2]. The preservative power of sulfite is mainly due to its reducing capacity. This makes it an effective antioxidant that also has the ability to inhibit bacterial growth [2,3].

In some cases, the presence of sulfite in food and beverages has been associated with asthma and allergic symptoms [4,5,6]. Therefore, regulatory agencies set maximum allowable limits [2].

There are some available methods for detecting SO₂ and its derivates, such as titrimetric analysis, Aspiration and Ripper reference methods and the Monier–Williams (MW) method [7,8,9,10], chemiluminescence [11], physical and chemical adsorption [12,13], spectrophotometry [14,15], ion chromatography [16,17], fluorescence [18,19], biosensors [20,21], etc. Although these methods possess highly efficient detection capabilities, their intrinsic difficulties (e.g., complex operation, high equipment costs, long response times, etc.) often limit their practical applications.

Electroanalytical techniques for the determination of sulfites arise as a simpler and more attractive alternative. The main benefits of these techniques include simplicity, lower cost instrumentation, high sensitivity, and the possibility of working with colored samples, without pretreatment of the samples [22]. In addition, compared to the traditional volumetric method (MW), electroanalytical techniques allow the number of reagents to be reduced, being more respectful with the environment.

For the most part, electroanalytical methods to determine sulfite use its oxidation as an analytical signal. Using oxidation, the sulfite concentration has been determined with a nanogold electrode on gallium-doped zinc oxide [23]. Similarly, a carbon paste electrode modified with carbon nanotubes and benzoylferrocene (BF) [24] has been used. In both cases an electrocatalytic effect was found in the oxidation of sulfite. In addition, the electrodetermination of sulfite has been carried out using a carbon composite electrode modified by submicron gold particles, finding good sensitivity and precision [25]. The disadvantage of determining sulfite by oxidation is that many antioxidants, usually present in foods and beverages, can act as interferences in the measurement [3]. In this sense, the platinum electrode modified with nanostructured films of Salen-copper polymer (salicylideneiminate) that electrocatalyzes sulfite oxidation can be highlighted, as a notorious decrease in potential, which allows the analysis to be determined in the presence of some interfering antioxidants [26].

An attractive way to avoid the interference of common antioxidants in foods and beverages is the selective determination of sulfite through its electrochemical reduction. The reduction of sulfite in an acid medium has been studied using glassy carbon electrodes [27]. Copper-clad gold microelectrodes [28] and glassy carbon electrodes modified with ruthenium oxide and hexacyanoferrate [29] have also been used for its reduction. In all these studies, sulfite could be determined with good selectivity and sensitivity.

Based on the above information, it is possible to design a low-cost electrode with a renewable surface for the determination of sulfite through its electroreduction. In this context, carbon paste electrodes are made with multi-walled carbon nanotubes (MWCNTs) as highly conductive carbonaceous material and as binder, a mixture of mineral oil (MO) and an ionic liquid (IL) (N-octylpyridinium hexafluorophosphate). It allows the manufacture of high-conductivity electrodes, economical, easy to prepare, electrocatalytic, with a renewable surface, that act as excellent sensors in the detection and quantification of sulfite through its electrochemical reduction.

2. Results and Discussion

2.1. Obtaining and Characterization of MWCNT/IL/MO Carbon Paste Electrodes

Figure 1 shows the different morphologies of multiwalled carbon nanotubes bonded with ionic liquid and mineral oil (MWCNT/IL/MO), in a ratio of 70:20:10 (by mass), and on multiwalled carbon nanotubes only bound with mineral oil (MWCNT/MO), in a ratio of 70:30 (by mass). Figure 1 shows that the electrode that has IL in its structure (Figure 1B) has a higher resolution, clearly observing the nanotubes, due to its higher conductivity, compared to the electrode that only has mineral oil as a binder (Figure 1A).

On the other hand, EDX studies were carried out to determine the chemical composition of each electrode, randomly selecting three areas in each case. The results are summarized in

Table 1. Percentage values of weight for MWCNT/MO electrodes and MWCNT/IL/MO electrodes.

Electrode	%C	%O	%F	%P
MWCNT/MO	96.81 ± 0.21	3.19 ± 0.26	---	---
MWCNT/IL/MO	93.55 ± 2.81	4.02 ± 0.12	1.53 ± 0.05	0.90 ± 0.03

and confirm the presence of ionic liquid [OPy]⁺[PF₆⁻] on the electrode surface.

2.2. Electrochemical Characterization of MWCNT/IL/MO Carbon Paste Electrodes

The electrical properties of MWCNT/IL/MO carbon paste electrodes in the presence of sulfite (5 mmol L⁻¹ sulfite in BRB pH 1.81) have been studied using Electrochemical Impedance Spectroscopy (EIS). The equivalent circuit that fits the Nyquist diagram, recorded in the absence and presence of IL, is depicted in Scheme 1. In this circuit, R_s, CPE, and R_ct represent the resistance of the solution, a constant phase element corresponding to double layer capacitance, and the resistance to charge transfer associated with sulfite reduction. For the electrode containing IL (see Figure 2), the diameter of the semicircle is reduced from 776.6 Ω to 99.7 Ω, confirming the electrocatalytic capacity of this electrode for sulfite reduction. At high frequencies, resistance to charge transfer appears and at low frequencies, diffusion processes are appreciated. Generally, the resistance to charge transfer (R_ct) of the electrode is equivalent to the diameter of the semicircle in the Nyquist plot [30]. Finally, the ohmic resistance of the solution (R_s) remains constant in the two paste electrodes in the presence of sulfite (see

Table 2. Values of R_s, R_ct, and CPE obtained from the equivalent circuit of Scheme 1 for the paste electrodes.

Paste Electrode	Rs (Ω)	Rct (Ω)	CPE (F cm−2)
MWCNT/MO	2926	776,610	0.89
MWCNT/IL/MO	2850	99,275	0.70

).

2.3. Electrochemical Behavior of Sulfite on MWCNT/IL/MO

Figure 3A shows a square wave voltammogram of the MWCNT/IL/MO electrode at different pHs in the presence of sulfite. Figure 3B shows the maximum cathodic currents (Ip) against pH variation. Both graphs indicate that the best pH for a catalytic reaction is near pH 2.

Considering that the best current response is observed in acid conditions, the sulfite reduction study is carried out at pH 1.81. This behavior is consistent with the chemical equilibrium of the sulfite species. At high pH values, the products of this equilibrium are favored towards the formation of sulfurous acid, whereas in acid conditions, there is a higher concentration of dissolved SO₂, which is the species that actually undergoes electrochemical reduction at the electrode surface. Therefore, the highest maximum cathodic currents are found in more acidic conditions. Based on these results, all electroanalytical studies were performed in a BRB pH 1.81 solution.

On the other hand, the electrochemical response of the electrode was studied in a potential range that goes from +0.2 V to −0.9 V vs. Ag/AgCl, and at a scan rate of 100 mV s⁻¹ in the absence and presence of sulfite. Figure 4 presents the voltammetric profile for the MWCNT/MO and MWCNT/IL/MO electrodes in the absence and presence of 5 mmol L⁻¹ of sulfite. The response of the analyte corresponds to the characteristic profile of the sulfite reduction. According to the previous literature, this reduction corresponds to the transformation of dissolved SO₂ into the radical anion SO₂⁻ [28,29]. MWCNT/IL/MO shows an Ep value for reduction of −0.56 V. When the electrode does not have IL, the voltammetric profile shows no signal and coincides with the blank, indicating that it is the IL that catalyzes the sulfite reduction (see Figure 4).

It is interesting to note that the edges of carbon nanotubes, generally very active for many reactions [31,32], are not active enough in this case to catalyze this reaction under the aforementioned conditions.

To calculate the number of electrons involved in this reaction, it is necessary to determine the electroactive area of the electrodes. For this, the effect of the scan rate (v) on the electrochemical behavior of a probe solution of ferri-ferrocyanide ions in an aqueous medium mas measured (Figure 5). With these results, the electroactive area of the MWCNT/IL/MO electrode is calculated.

The results show that the peak current (Ip) increases linearly with the square root of the scan rate of the potential (v^1/2). The electroactive area is calculated from the slope of the linear regression curve (R² = 0.99874) (inset of Figure 5) using Equation (1):

$$\frac{\mathit{I}\mathit{p}}{\sqrt{\mathit{v}}}=\left(\mathbf{2.65}\times {\mathbf{10}}^{\mathbf{5}}\right)\times \sqrt{{\mathit{n}}^{\mathbf{3}}}\times \mathit{A}\times \sqrt{\mathit{D} }\times \mathit{C}$$

(1)

where A is the electroactive area, n corresponds to the number of electrons transferred (n = 1), D is the diffusion coefficient of ferrocyanide 6.5 × 10⁻⁶ cm² s⁻¹, and C corresponds to the concentration ferrocyanide probe 5.0 × 10⁻⁶ mol m⁻³), obtaining a value of electroactive area A = 0.016 cm².

Taking the electroactive area, the number of transferred electrons can be calculated through the equations and if the reaction is irreversible. The irreversibility of the reaction (mass transfer-controlled) is clearly observed in Figure 6.

$$\mathit{I}\mathit{p}=\mathbf{2.99}\times \mathbf{10}{ }^{\mathbf{5}} \mathit{n}\sqrt{\left(\mathbf{1}-\mathit{\alpha }\right){\mathit{n}}_{\mathit{\alpha }}} \mathit{A} \sqrt{\mathit{D}} \mathit{C}\sqrt{\mathit{v}}$$

(2)

$$\frac{\mathbf{0.0477} \mathit{V}}{{\mathit{E}}_{\mathit{p}}-{\mathit{E}}_{\mathit{p}/\mathbf{2}}}=\left(\mathbf{1}-\mathit{\alpha }\right){\mathit{n}}_{\mathit{\alpha }}$$

(3)

$${\mathit{E}}_{\mathit{p}}-{\mathit{E}}_{\mathit{p}/\mathbf{2}}=-\mathbf{0.543}-\left(-\mathbf{0.461}\right)=\left|\mathbf{0.082}\right|$$

A corresponds to the electroactive area of the electrode (0.016 cm²), and (1 − α)n_α = 0.5817 is obtained from Equation (3), where α is the transfer coefficient and nα is the number of electrons transferred in the rate-determining step. The Ep value corresponds to 100 mV s⁻¹. D is the diffusion coefficient of the analyte [33] and C is the sulfite concentration (5 × 10⁻⁶ mol cm⁻³). The number of electrons transferred in the reduction of sulfite (n) is obtained from the slope of the graph in Figure 6. Using Equation (2), n = 0.5 is obtained, which indicates that the electrochemical reduction of sulfite is a transfer process half an electron for sulfite. It has been described in the literature that the equilibrium anion radical SO₂⁻ dithionite is established very quickly, with dimerization rate constants close to those controlled by diffusion [33] at a pH close to 2, which explains the number of electrons found in this work.

2.4. Voltammetric Determination of Sulfite

A calibration curve of peak reduction current versus analyte concentration was prepared. The technique used was square wave voltammetry and the optimized parameters were the following: frequency = 50 Hz, amplitude = 0.025 V, increment = 0.004 V. The calibration curve (six replicates) was prepared with concentrations of 5 ± 0.2 mmol L⁻¹, 10 ± 0.3 mmol L⁻¹, 20 ± 0.6 mmol L⁻¹, 30 ± 0.8 mmol L⁻¹, and 50 ± 1.5 mmol L⁻¹ of a certified standard of sodium sulfite at pH 1.81, at BRB (see Figure 7).

2.5. Analytical Parameters

Figure 7 shows the voltammetric response of the electrode at different of sulfite. The linear regression of Ip_c versus sulfite concentration R² = 0.99355.

The calibration curve of sulfite in MWCNT/IL/MO is linear in the range of 5.0 mmol L⁻¹ to 50 mmol L⁻¹, with a limit of detection of 1.6 ± 0.05 mmol L⁻¹ (102.5 mg SO₂ L⁻¹) (S/N = 3). It is noted that the limit of detection of the MWCNT/IL/MO system 1.6 ± 0.05 mmol L⁻¹ is higher compared to other published studies [22,34]. This value indicates that with this system, it could be detected below the maximum concentrations established for sulfite present in some alcoholic beverages (200–300 mg L⁻¹) [9]. In addition, the simplicity of the sulfite detection method is highlighted.

The linear regression equation is: I (μA) = 1.02656 + 0.41192 C (mmol L⁻¹). The sensitivity is 0.41192 μA mmol⁻¹ L.

From these data we obtain that MWCNT/IL/MO electrode exhibits a one order of magnitude range of linear concentration, a lower limit of detection, and a higher sensitivity compared to most of electrochemical sulfite sensors reported [22,28,29,34].

Another important electroanalytical study is repeatability. The electrochemical response of four (MWCNT/IL/MO) paste electrodes was assessed, with different triplicate, performed the same day at pH 1.81 in BRB, using a concentration of sodium sulfite equivalent to 5 ± 0.2 mmol L⁻¹. The relative standard deviation (RSD) of the peak current response for the four different electrodes was of 3.55% (n = 3), demonstrating that the MWCNT/IL/MO sensor presents good repeatability (Figure 8).

3. Experimental

3.1. Materials, Reagents, and Equipment

Multi-walled carbon nanotubes (MWCNTs), diethyl ether (DEE), and potassium chloride (KCl) were purchased from Merck (Santiago, Chile) and used as received. Acetic acid (CH₃COOH), boric acid (H₃BO₃), orthophosphoric acid (H₃PO₄), mineral oil (MO), sodium sulfite (98% pure), certified sulfite standard, potassium ferricyanide (K₃[Fe(CN)₆]) (98.5% pure), and potassium ferrocyanide trihydrate (K₄[Fe(CN)₆]) were purchased from Sigma Aldrich (Santiago, Chile), and used as received. Ultrapure water (>17 MΩ-cm) was obtaining from a Direct-Q^®® water purification system (Model ZRQSVP3WW) and argon gas (99.99% pure) was purchased from AGA (Santiago, Chile). N-octylpyridinium hexafluorophosphate [OPy]⁺[PF₆]⁻ ionic liquid was previously synthetized and characterized [35,36] obtaining a yield of 91%.

Electrochemical studies were performed in a conventional three-electrode electrochemical cell which contains: an Ag/AgCl (3 mol L⁻¹ KCl) reference electrode (RE), a platinum wire counter electrode (CE), and a hollow Teflon working electrode (WE) of 2 mm diameter, which is filled with the carbon pastes specified in Table 1. The cell contains a 0.1 mol L⁻¹ Britton–Robinson (BRB) buffer aqueous solution, and the pH is adjusted by additions of 5 mol L⁻¹ NaOH and 5 mol L⁻¹ HCl. The solution is maintained in an inert atmosphere of Ar.

Morphological characterization was performed using a JEOL 6300 scanning electron microscope (SEM) coupled to 6699 ATW X-ray microanalysis elemental system (Oxford Instruments Ltd., Abingdon, UK). Data were analyzed using the associated software (version 4.0). Spectra were recorded using a 15 keV accelerating voltage. Cyclic voltammetry (CV), square wave voltammetry (SWV) and electrochemical impedance spectroscopy (EIS) were performed using a model 750D workstation (CH Instruments, Austin, TX, USA). All measurements were performed at room temperature, 21 ± 1 °C.

3.2. Paste Electrode Preparation

The 70:30 (w/w) ratio of carbon materials and binder was used after experimentally checking that at different binder ratios (both higher and lower), the electrodes lose their mechanical and structural consistency. For the manufacture of carbon paste electrode, the bare carbon paste electrode was made by mixing MWCNT and oil mineral in a 70:30 percent weight-to-weight ratio [37]. On the other hand, to manufacture the carbon paste electrode with MWCNT binder with oil mineral and ionic liquid, a 70:20:10 (w/w) ratio of MWCNT, oil mineral, and ionic liquid was used, respectively.

The 70: 30 (w/w) ratio of carbon and binder materials was used after experimentally verifying that at different binder ratios (both higher and lower), the electrodes lose their mechanical and structural consistency. For the fabrication of the carbon paste electrode, the bare carbon paste electrode was fabricated by mixing MWCNT and mineral oil in a 70:30 percent weight-to-weight ratio [35]. On the other hand, for the fabrication of the MWCNT binder carbon paste electrode with mineral oil and ionic liquid, a ratio of 70:20:10 (w/w) of MWCNT, mineral oil, and ionic liquid, respectively, was used. The mixtures are homogenized in a mortar by adding diethyl ether. After homogenization, the diethyl ether is evaporated until a carbon paste is obtained. The generated paste is used to fill hollow Teflon electrodes. After compacting by hand, the electrodes are left at 90 °C in an oven. After 2 h, the electrodes are cooled to room temperature and polished with weighing paper to a smooth surface. The obtained electrodes were electrochemically stabilized by cycling the potential in the range of 200 to −900 mV in BRB pH 1.81. The obtained electrodes were characterized by their voltammetric response using the potassium ferrocyanide/potassium ferricyanide redox couple, showing that they are very reproducible systems.

4. Conclusions

This work shows that the electrode prepared with carbon nanotubes bonded with a mixture of LI (N-octylpyridinium hexafluorophosphate) and mineral oil allows obtaining an electrode that is easy to prepare, low-cost, and eco-friendly. The electrode is active for the reduction of sulfite, being able to become a sensor to determine its concentration in an acid medium.