Electrochemical Sensing of Gallic Acid in Beverages Using a 3D Bio-Nanocomposite Based on Carbon Nanotubes/Spongin-Atacamite

Gallic acid (GA) is one of the most important polyphenols, being widely used in the food, cosmetic, and pharmaceutical industries due to its biological effects such as antioxidant, antibacterial, anticancer, antiviral, anti-inflammatory, and cardioprotective properties. Hence, simple, fast, and sensitive determination of GA is of particular importance. Considering the fact that GA is an electroactive compound, electrochemical sensors offer great potential for GA quantitation due to their fast response time, high sensitivity, and ease of use. A simple, fast, and sensitive GA sensor was fabricated on the basis of a high-performance bio-nanocomposite using spongin as a natural 3D polymer, atacamite, and multi-walled carbon nanotubes (MWCNTs). The developed sensor showed an excellent response toward GA oxidation with remarkable electrochemical features due to the synergistic effects of 3D porous spongin and MWCNTs, which provide a large surface area and enhance the electrocatalytic activity of atacamite. At optimal conditions by differential pulse voltammetry (DPV), a good linear relationship was obtained between peak currents and GA concentrations in a wild linear range of 500 nM to 1 mM. Subsequently, the proposed sensor was used to detect GA in red wine as well as in green and black tea, confirming its great potential as a reliable alternative to conventional methods for GA determination.


Introduction
Over the past few years, antioxidants have emerged as important ingredients in foods and beverages due to their numerous health benefits, such as anti-aging and antiinflammatory properties [1]. In fact, they neutralize free radicals and reactive oxygen species that damage cells and cause chronic health problems and aging. Polyphenols, secondary plant compounds with over 8000 variants, are the most common antioxidants found in vegetables, fruits, grains, and beverages. Recently, they have been widely applied as natural additives instead of synthetic ones to improve food quality and promote health benefits [2][3][4]. Among them, GA (3,4,5-trihydroxybenzoic acid) and its derivatives have received significant attention due to their extraordinary biological features. They can be found in a variety of herbs, vegetables, and fruits, including apple peel, bile seeds, grapes, bananas, strawberries, lemons, pineapple, and oak bark, as well as in wine products and in green and black tea [5,6]. The anti-cancer [7], anti-diabetic [8], anti-allergic [9], anti-fungal [10], anti-bacterial [11], anti-inflammatory [12], neuroprotective [13], and cardioprotective properties [14] have made GA as one of the most widely used phenolic components in the medicine, food, and pharmaceutical industries. Considering these diverse biological and pharmaceutical effects, the determination of GA is of great importance. High-performance liquid chromatography [15], spectrophotometry [16], flow injection analysis [17], capillary

Apparatus and Measurements
Electrochemical measurements were performed with a PalmSens 4 electrochemical analyzer system with the software PSTrace 5.8 (PalmSens BV, Houten, the Netherlands). All experiments were carried out at room temperature in an electrochemical cell with a three-electrode platform. A saturated silver/silver chloride (Ag/AgCl) (3 M KCl solution) and a platinum wire electrode were used as the reference electrode and auxiliary electrode, respectively. Bare and modified CPEs served as the working electrode. Phosphate buffer solution (0.1 M, pH 3.0-6.0) was prepared using a mixture of the stock solutions (NaH 2 PO 4 and Na 2 HPO 4 ) and employed as an electrolyte solution for all measurements. EIS measurements were carried out in 0.1 M KCl containing 5 mM [Fe(CN) 6 ] 3−/4− at the frequency range of 0.1 Hz-100 kHz, applied potential of 0.23 V, and an amplitude of 5 mV.

Preparation of the CPEs
Three-dimensional Sp-At was prepared according to our earlier reports [38,51]. It was characterized using different methods such as high-resolution transmission electron microscopy, scanning electron microscopy, X-ray photoelectron spectroscopy, and X-ray diffraction [38,51]. The bare CPE and modified CPEs were fabricated by grinding different ratios of graphite, MWCNTs, Sp-At powder, and paraffin oil as a binder in a mortar and with a grinding time of 40 min ( Table 1). The components were homogenized to shape a paste, which was then pressed into a carbon paste holder with an inner diameter of 4 mm. These electrodes were then denoted as MWCNTs/CPE (MWCNTs modified CPE), Sp-At/CPE (Sp-At modified CPE), and Sp-At/MWCNTs/CPE (Sp-At-and MWCNT-modified CPE). Before each measurement, the surface of the CPEs was regenerated by smoothing with a filter paper.

Preparation of Sample Solutions
Black and green tea samples were obtained by a simple filtration method. A total of 3 g of dried and powered tea leaves were added to 25 mL of 80 • C boiled double distilled water for 10 min and then allowed to cool at room temperature. The extract was filtered with filter paper and diluted with 0.1 M phosphate buffer (pH 4). Prior to each measurement, it was spiked with standard GA solutions.
The red wine sample was directly diluted with 0.1 M phosphate buffer (pH 4), without any pretreatment, and spiked with standard GA solutions. Figure 1 shows the DPVs of bare CPE, Sp-At/CPE, MWCNTs/CPE, and Sp-At/MWCNTs/ CPE recorded in 0.1 M phosphate buffer, pH 4, in the absence and presence of 30 µM GA. In the absence of GA, no obvious oxidation peak was observed at any of the electrodes, and also the background current was very low (Figure 1a). DPVs of different CPEs in the presence of 30 µM GA revealed that the electrochemical oxidation of GA was defined by two anodic peaks ( Figure 1b). Peak I (0.3 V vs. Ag/AgCl) was ascribed to the formation of the semiquinone radical as a result of the oxidation of the galloyl group, which was then oxidized to the quinone form (peak II (0.6 V vs. Ag/AgCl)) [55]. In both reactions, one proton and one electron were involved. Scheme 1 illustrates the well-known route for the oxidation of phenols, hydroquinones, and derivatives. Since the second peak was much weaker than the first one, GA oxidation was studied by focusing on the first oxidation peak [25]. As seen in Figure 1b, Sp-At/MWCNTs/CPE exhibited a clearly enhanced oxidation current in response to GA in comparison to CPE, Sp-At/CPE, and MWCNTs/CPE. In accordance with the literature [56,57], the electrochemical oxidation of GA is initiated by the oxidation of copper oxides to their higher valence states, and then the oxidized copper oxides help to oxidize GA electrocatalytically. The 3D spongin structure provides a large surface area for the formation of nano-and microcrystalline atacamite and promotes surface adsorption and electron transfer between GA and Sp-At [38]. These features, combined with the ability of MWCNTs to facilitate electron transfer kinetics, favorably contributed to the enhancement of electron transfer between GA and the electrode surface and the resulting oxidation current [27]. Hence, the Sp-At/MWCNTs/CPE was used as a working electrode for further investigations.

Electrochemical Characterization of CPEs
EIS, as a powerful method for characterizing the electrical properties of an electrode surface, was utilized to confirm the electrode modification.

Electrochemical Characterization of CPEs
EIS, as a powerful method for characterizing the electrical properties of an electrode surface, was utilized to confirm the electrode modification. , where R s stands for solution resistance; C dl is the double-layer capacitance; and Z w and R ct represent Warburg resistance and charge transfer resistance, respectively. As illustrated in Figure 2, all electrodes exhibited semicircles of differing diameters in the high-frequency region, which corresponds to R ct and, in the low-frequency region, a line at about 45 degrees (also known as the Warburg element), which can be attributed to the probe's diffusion from the bulk solution to the electrode-solution interface. In comparison to the bare CPE (487 Ω), Sp-At/CPE (370 Ω), and MWCNTs/CPE (256 Ω), the Sp-At/MWCNTs/CPE (200 Ω) exhibited a smaller R ct value. This indicates that both Sp-At and MWCNTs accelerate electron transfer, but the combination of 3D porous Sp-At, which offers higher electron conduction paths, and MWCNTs with a specifically large surface area exhibits the highest charge transfer and electrical conductivity.

Optimization of Sp-At/MWCNTs/CPE Composition
It is well known that the electrochemical behavior of modified carbon paste electrodes strongly depends on the mixing ratio, the way of preparing the carbon pastes, the duration of grinding, and homogenization [58,59]. In this regard, to improve the efficiency of the proposed sensor based on Sp-At/MWCNTs/CPE for detecting GA, the ratio of MWCNTs and Sp-At powder as modifiers was optimized (Table 1). Six electrodes

Optimization of Sp-At/MWCNTs/CPE Composition
It is well known that the electrochemical behavior of modified carbon paste electrodes strongly depends on the mixing ratio, the way of preparing the carbon pastes, the duration of grinding, and homogenization [58,59]. In this regard, to improve the efficiency of the proposed sensor based on Sp-At/MWCNTs/CPE for detecting GA, the ratio of MWCNTs and Sp-At powder as modifiers was optimized (Table 1). Six electrodes with different ratios of MWCNTs and Sp-At powder were prepared, and the DPVs of the prepared electrodes were recorded in 0.1 M phosphate buffer, pH 4, in the presence of 30 µM GA. Figure 3 shows the responses of different electrode compositions in the similar electrochemical conditions. As can be seen, the amount of the two modifiers played an important role in the response of the sensors to GA. MWCNTs as modifiers enhanced the oxidation current of GA due to their large surface area and electrical and electrocatalytic properties. On the other hand, the 3D microporous sponges provided a larger specific surface area for atacamite as an electrocatalyst, facilitating the diffusion of GA toward atacamite and improving the electron transfer between GA and atacamite. Thus, the synergistic effects of MWCNTs and 3D Sp-At lead to a remarkable improvement in the sensing performance of Sp-At/MWCNTs/CPE. According to the obtained results, the CPE containing 10% (in a ratio of 1:1) MWCNTs and Sp-At powder showed the maximum oxidation current.

Influence of pH on GA Oxidation
The effect of pH on the oxidation of 30 μM GA was investigated in the pH range of 3-6 in 0.1 M phosphate buffer. It can be seen that GA oxidation was affected by the pH of phosphate buffer, which was due to the fact that GA oxidation was a protonation reaction ( Figure 4). GA with a pKa of 4.4 was a weak acid and existed in the undissociated form in solution at lower pH values (pH < 4.4) and dissociated as the solution pH increased. Therefore, the maximum adsorption of GA on the surface of Sp-At/MWCNTs/CPE was expected to occur when GA was undissociated [30]. As shown in the inset of Figure 4 (blue), the anodic peak current (Ipa) reached a maximum value at pH 4 and then decreased with increasing pH to 6. This implies that the species were gradually deprotonated. Moreover, the peak anodic potentials shifted negatively with an average shift of 30 mV when the pH was increased, confirming that the electrode process was affected by the protonation reactions, as observed by other authors [55,60]. A linear regression line was obtained for the changing of peak potential (Ep) as a function of pH (Figure 4, inset; red)

Influence of pH on GA Oxidation
The effect of pH on the oxidation of 30 µM GA was investigated in the pH range of 3-6 in 0.1 M phosphate buffer. It can be seen that GA oxidation was affected by the pH of phosphate buffer, which was due to the fact that GA oxidation was a protonation reaction ( Figure 4). GA with a pKa of 4.4 was a weak acid and existed in the undissociated form in solution at lower pH values (pH < 4.4) and dissociated as the solution pH increased. Therefore, the maximum adsorption of GA on the surface of Sp-At/MWCNTs/CPE was expected to occur when GA was undissociated [30]. As shown in the inset of Figure 4 (blue), the anodic peak current (I pa ) reached a maximum value at pH 4 and then decreased with increasing pH to 6. This implies that the species were gradually deprotonated. Moreover, the peak anodic potentials shifted negatively with an average shift of 30 mV when the pH was increased, confirming that the electrode process was affected by the protonation reactions, as observed by other authors [55,60]. A linear regression line was obtained for the changing of peak potential (E p ) as a function of pH (Figure 4, inset; red) with a regression equation of E p = −0.0661 pH + 0.5603 (R 2 = 0.9929). The slope of −66 mV was in good agreement with the theoretical Nernstian systems with a mono-electron transfer step followed by single protonation [61].

Scan Rate Influence on the Oxidation of GA
The GA oxidation reaction kinetics on the surface of Sp-At/MWCNTs/CPE was further evaluated by conducting CV measurements at different scan rates in 0.1 M phosphate buffer (pH 4) containing 30 μM GA solution. As can be observed, at low scan rates, a cathodic peak with very low current on the reverse sweep appeared, which indicated that the oxidized form of GA was not reduced at the surface of Sp-At/MWCNTs/CPE or maybe underwent a further chemical reaction. However, with increasing scan rate, the cathodic current increased, indicating that the oxidized form of GA was quasi-reversibly reduced before the chemical reaction [20,32]. From Figure 5a, it can be seen that the Ipa for GA oxidation increased as the scan rate rose from 20 to 300 mV•s −1 . As exhibited in the Figure 5b, Ipa increased linearly in proportion to the square root of the scan rate with a linear regression equation of Ipa (μA) = 0.2147 ν 1/2 (mV 1/2 s −1/2 ) + 0.5186 (R 2 = 0.9975). The R 2 value of Ipa vs. the square root of the scan rate was almost equal to the theoretical value of 1.0, which was generally expected for a diffusion-controlled redox system, thus verifying that electrochemical oxidation of GA at the surface of Sp-At/MWCNTs/CPE is entirely diffusion controlled [62].

Scan Rate Influence on the Oxidation of GA
The GA oxidation reaction kinetics on the surface of Sp-At/MWCNTs/CPE was further evaluated by conducting CV measurements at different scan rates in 0.1 M phosphate buffer (pH 4) containing 30 µM GA solution. As can be observed, at low scan rates, a cathodic peak with very low current on the reverse sweep appeared, which indicated that the oxidized form of GA was not reduced at the surface of Sp-At/MWCNTs/CPE or maybe underwent a further chemical reaction. However, with increasing scan rate, the cathodic current increased, indicating that the oxidized form of GA was quasi-reversibly reduced before the chemical reaction [20,32]. From Figure 5a, it can be seen that the I pa for GA oxidation increased as the scan rate rose from 20 to 300 mV·s −1 . As exhibited in the Figure 5b, I pa increased linearly in proportion to the square root of the scan rate with a linear regression equation of I pa (µA) = 0.2147 ν 1/2 (mV 1/2 s −1/2 ) + 0.5186 (R 2 = 0.9975). The R 2 value of I pa vs. the square root of the scan rate was almost equal to the theoretical value of 1.0, which was generally expected for a diffusion-controlled redox system, thus verifying that electrochemical oxidation of GA at the surface of Sp-At/MWCNTs/CPE is entirely diffusion controlled [62].

Quantification of GA by DPV
GA quantification was performed under the optimum experimental conditions (phosphate buffer (pH 4); Sp-At/MWCNTs/graphite/paraffin oil, 5:5:70:20 (W/W/W/W)). While the peak current increased by increasing the GA concentration, shift of the peak potentials towards more positive potentials was observed due to the generation of more electroinactive oxidation species that block the electrode surface [61]. The calibration curves for the peak current versus the concentrations of GA were obtained at Sp-At/MWCNTs/CPE with two different linear parts. Figure 6b illustrates the first linear part, which was observed between 500 nM and 230 µM for lower GA concentrations with a linear regression equation as I pa (µA) = 234.1 C GA (mM) + 0.299 (R 2 = 0.9991), and Figure 6c shows the second linear part for higher GA concentrations between 230 µM and 1 mM with a linear regression equation of I pa (µA) = 88.8 C GA (mM) + 32.024 (R 2 = 0.9916). The sensitivity of the prepared GA sensor based on Sp-At/MWCNTs/CPE was calculated to be 18.6 µA mM −1 cm −2 at low concentrations and 7 µA mM −1 cm −2 at high concentrations of GA. The detection limit (LOD) and limit of quantitation (LOQ) were calculated from the LOD = 3 σ/m and LOQ = 10 σ/m equations, respectively (σ is the standard deviation of the signal in blank solution, and m is the slope of the calibration curve for lower GA concentrations). The LOD and LOQ values of Sp-At/MWCNTs/CPE were obtained as 5.4 nM and 17.9 nM, respectively, which were quite low in comparison to the other reported GA sensors [26,36,63]. The shift of the peak potential towards more positive potentials was due to the generation of electro-inactive oxidation species on increasing the GA concentration. These products blocked the electrode surface.

Quantification of GA by DPV
GA quantification was performed under the optimum experimental conditions (phosphate buffer (pH 4); Sp-At/MWCNTs/graphite/paraffin oil, 5:5:70:20 (W/W/W/W)). While the peak current increased by increasing the GA concentration, shift of the peak potentials towards more positive potentials was observed due to the generation of more electro-inactive oxidation species that block the electrode surface [61]. The calibration curves for the peak current versus the concentrations of GA were obtained at Sp-At/MWCNTs/CPE with two different linear parts. Figure 6b [26,36,63]. The shift of the peak potential towards more positive potentials was due to the generation of electro-inactive oxidation species on increasing the GA concentration. These products blocked the electrode surface.

Quantification of GA by Amperometry
In amperometric studies, mass transfer occurs via diffusion and convection processes, which facilitate the quantitation of analytes at lower concentrations. The amperometric results under the optimized conditions for GA oxidation at 0.35 V are presented in Figure 7a. By increasing the GA concentration, the oxidation current increased. Figure 7b

Quantification of GA by Amperometry
In amperometric studies, mass transfer occurs via diffusion and convection processes, which facilitate the quantitation of analytes at lower concentrations. The amperometric results under the optimized conditions for GA oxidation at 0.35 V are presented in Figure 7a. By increasing the GA concentration, the oxidation current increased. Figure 7b,c illustrates the corresponding calibration plots for GA consisting of two different slopes. The first linear section was found between 250 nM and 60 µM with an equation of I pa (µA) = 20.7 C GA (mM) + 0.0238 (R 2 = 0.9954), and the second linear part was found between 60 µM and 550 µM with an equation of I pa (µA) = 3.3 C GA (mM) + 1.1295 (R 2 = 0.9841). The sensitivity of the proposed GA sensor based on Sp-At/MWCNTs/CPE was evaluated as 1.64 µA mM −1 cm −2 at low concentrations and 0.26 µA mM −1 cm −2 at high concentrations, with a detection limit of 3.6 nM. These quantitative results clearly show that the fabricated Sp-At/MWCNTs/CPEbased sensor has excellent performance for GA determination. The extraordinary electronic properties of MWCNTs promote electron transfer in electrochemical reactions, and the large surface area and good crystalline structure of the 3D Sp-At porous scaffold offer a considerable number of interaction sites for such reactions, leading to the high-performance properties of the GA sensor. Table 2 summarizes the analytical performance characteristics of the Sp-At/MWCNTs/CPE with several previously reported GA sensors.
In amperometric studies, mass transfer occurs via diffusion and convection processes, which facilitate the quantitation of analytes at lower concentrations. The amperometric results under the optimized conditions for GA oxidation at 0.35 V are presented in Figure 7a. By increasing the GA concentration, the oxidation current increased. Figure 7b,c illustrates the corresponding calibration plots for GA consisting of two different slopes. The first linear section was found between 250 nM and 60 μM with an equation of Ipa (μA) = 20.7 CGA (mM) + 0.0238 (R 2 = 0.9954), and the second linear part was found between 60 μM and 550 μM with an equation of Ipa (μA) = 3.3 CGA (mM) + 1.1295 (R 2 = 0.9841). The sensitivity of the proposed GA sensor based on Sp-At/MWCNTs/CPE was evaluated as 1.64 μA mM −1 cm −2 at low concentrations and 0.26 μA mM −1 cm −2 at high concentrations, with a detection limit of 3.6 nM. These quantitative results clearly show that the fabricated Sp-At/MWCNTs/CPE-based sensor has excellent performance for GA determination. The extraordinary electronic properties of MWCNTs promote electron transfer in electrochemical reactions, and the large surface area and good crystalline structure of the 3D Sp-At porous scaffold offer a considerable number of interaction sites for such reactions, leading to the high-performance properties of the GA sensor. Table 2 summarizes the analytical performance characteristics of the Sp-At/MWCNTs/CPE with several previously reported GA sensors.    Figure S1). The relative standard deviation (RSD) of recorded anodic peak currents of four independently prepared electrodes was 4.22%, indicating that the fabricated sensors are excellent in terms of reproducibility. Similarly, in the repeatability test, eight consecutive measurements of 30 µM GA on one Sp-At/MWCNTs/CPE resulted in an RSD of 0.62%, confirming the favorable repeatability of the fabricated sensor ( Figure S2). The long-term stability test was additionally assessed by DPV measurements of the Sp-At/MWCNTs/CPE in the presence of 30 µM GA. After one month, the current was still maintained at over 87% of its original value, signifying that the proposed sensor is notably stable ( Figure S3). Moreover, the current response of Sp-At/MWCNTs/CPE was monitored at intervals of 2000 s after the injection of 30 µM GA. The amperometric current remained at 85% of its initial value after testing for 2000 s (Figure 8a). The specificity of the Sp-At/MWCNTs/CPE sensor was evaluated for 60 µM GA in the presence of 100-fold of possible coexistence species (sucrose, glucose, ascorbic acid (AA), and uric acid (UA)), which are usually presented along with GA [26,63]. As shown in Figure 8b, it can be found that these organic substances did not significantly impact the determination of GA. The Sp-At/MWCNTs/CPE sensor showed a relatively small peak current change, which was less than ±5%, suggesting excellent selectivity for the determination of GA. Considering that the oxidation of AA and UA occurred in the same potential window of GA, this selectivity could have been due to the structure of GA and the number of its hydroxyl groups. GA with three ionizable hydroxyl groups (compared to AA with two hydroxyl groups and uric acid without hydroxyl groups) forms both intra-(between hydroxyl groups) and intermolecular hydrogen bonds [68]. This hydrogen bonding ability contributes to GA diffusing and adsorbing more strongly than other coexisting species on the spongin scaffold (with different amino acids and consequently multiple functional groups such as SH, OH, NH 2 , and COOH) [51].
Biosensors 2023, 13, x 12 o This hydrogen bonding ability contributes to GA diffusing and adsorbing more stron than other coexisting species on the spongin scaffold (with different amino acids a consequently multiple functional groups such as SH, OH, NH2, and COOH) [51].

Determination of GA in Real Samples
Determination of GA content in black and green tea and red wine accomplish utilizing the Sp-At/MWCNTs/CPE sensor. The samples were prepared in 0.1 M phosph buffer (pH 4) and then were spiked with standard GA solutions to obtain a GA range fr 0 to 30 μM, followed by recording their corresponding DPV voltammograms (Figure In all real samples (red wine, black tea, and green tea), GA was detected and quantif on the basis of the generated Ipa (Table 3), which is generally attributed to the to antioxidant capacity of the sample, as described [55]. More oxidation peaks w produced with increasing spiked GA concentrations in all samples. The obtained excell recovery resulted in 95.2 to 102% (Table 3), confirming the applicability of the develop sensor for determining GA in beverages. Table 3. Results of the analysis of GA in black/green tea and red wine.

Determination of GA in Real Samples
Determination of GA content in black and green tea and red wine accomplished utilizing the Sp-At/MWCNTs/CPE sensor. The samples were prepared in 0.1 M phosphate buffer (pH 4) and then were spiked with standard GA solutions to obtain a GA range from 0 to 30 µM, followed by recording their corresponding DPV voltammograms ( Figure S4). In all real samples (red wine, black tea, and green tea), GA was detected and quantified on the basis of the generated I pa (Table 3), which is generally attributed to the total antioxidant capacity of the sample, as described [55]. More oxidation peaks were produced with increasing spiked GA concentrations in all samples. The obtained excellent recovery resulted in 95.2 to 102% (Table 3), confirming the applicability of the developed sensor for determining GA in beverages.

Conclusions
In summary, for the first time, renewable biopolymer spongin and atacamite in combination with MWCNTs (SP-At/MWCNTs) was successfully applied as a sensitive composite to fabricate a low-cost, easy, and reliable GA sensor. The SP-At/MWCNT composite enabled fast electron transfer, charge conduction, and efficient mass transport of GA on the surface of electrode. Further, the fabricated sensor based on SP-At/MWCNTs/CPE showed improved GA quantitation at a lower potential. Elevated sensitivity; acceptable selectivity among several interferences; a low detection limit; and a wide linear range with excellent stability, good repeatability, and reproducibility were realized. The developed sensor was applied successfully for determining GA in beverage samples such as black/green tea and red wine with satisfactory recovery results. The low cost, incorporation of a renewable biopolymer, ease of fabrication, and good stability make this approach attractive for the development of new sensing materials to quantify other phenolic compounds in food and beverage matrices.

Conflicts of Interest:
The authors declare no conflict of interest.