Signal-Enhanced Electrochemical Determination of Quercetin with Poly(chromotrope fb)-Modified Pencil Graphite Electrode in Vegetables and Fruits

A novel signal-enhanced electrochemical sensing strategy was constructed for quercetin determination with a peculiarly developed poly(chromotrope fb)-modified activated pencil graphite electrode in vegetables and fruits. The oxidation signal of quercetin at 118 mV in an alcoholic solution served as the analytical response. The produced platform, characterized by cyclic voltammetry, electrochemical impedance spectroscopy, scanning electron microscopy, energy-dispersive X-ray spectroscopy, and X-ray photoelectron spectroscopy, could detect 1.9 nM of quercetin in the range of 0.01–1.2 μM. The extracted quercetin contents of red onion, red cabbage, cranberry, black mulberry, black raisin, and carob were determined by both the developed method and UV–visible spectroscopy. The results were statistically evaluated at the 95% confidence level, and no significant difference between the results was found.


INTRODUCTION
Quercetin (Qn, 3,3′,4′,5,7-pentahydroxyflavone), also named as vitamin P, is a significant carbohydrate-free bioflavonoid that is generally found in plants, vegetables, and fruits. 1 Qn is a bioactive flavonoid with many therapeutic properties in diabetes, cardiovascular and arthritic diseases, autophagy, and Alzheimer's disease. 2 In addition, it has antioxidant, antimicrobial, anti-inflammatory, antiallergic, antiobesity, antitumor, antihypertensive, antihypercholesterolemic, antiatherosclerotic, immunomodulatory, neuroprotective, and vasodilator effects. 3 Qn has shown many biological benefits including the propitiation of morning stiffness and pain in rheumatoid arthritis patients, inhibition of cytokine production, reducing lipoxygenase and cyclooxygenase expression, providing the stability of mast cells, 4 restraint of the cytopathic effects of rhinovirus, echovirus, coxsackievirus, and poliovirus, decreasing the formation of RNA and DNA viruses involving respiratory syncytial virus, polio type 1, parainfluenza type 3, and herpes simplex virus-1, 5 protecting brain cells against oxidative stress and excitotoxicity and thus preventing Alzheimer's disease, 6 preventing cancer-with apoptosisinducing effects, promoting the insulin-sensitizing effect to hinder diabetes and inhibiting platelet aggregation, and helping to impede cardiovascular diseases by improving the health of the endothelium. 7 It is recommended to take 500−1000 mg/day, which should be included in the daily diet, 8 whereas the overdose may cause headache, inflammation, and damage to the kidney and DNA structure. 8,9 Thus, considering both the advantages and disadvantages, monitoring of the Qn level in foodstuffs is very significant. The methods based on spectrophotometry 10,11 and chromatography 12,13 are widely used for determining Qn. Although these methods yield in terms of sensitivity and selectivity, they employ labor-intensive pretreatment steps, difficult separation procedures, and sophisticated and expensive instruments. 14,15 In recent years, electrochemical sensors have come to the forefront for determining some electroactive food and biologically essential substances. 16−18 Due to some distinct features of electrochemical methods such as being simple, quick, portable, cost-effective, sensitive, and selective, they have become one of the most encouraging methods for determining Qn. 19,20 A wide variety of modified electrodes including single-or multipolymer film-modified electrodes with or without metal nanoparticles, 21−25 carbon nanotube-based electrodes with 26 or without polymer films, 27,28 molecularly imprinted polymer (MIP)-based electrodes, 29,30 and the other ones 31,32 have been reported for the determination of Qn, and they are summarized in Table S1.
The electrodes composed of a single polymer 21,24 have high limit of detection (LOD) values (i.e., 0.17 21 and 20 μM 24 ); the other electrodes have complex and time-consuming preparation steps, 22,23,25−32 and the supporting surfaces in these studies, i.e., glassy carbon electrodes (GCEs), carbon paste electrodes (CPEs), and paraffin-impregnated graphite disk electrode, have advantages such as being prone to electrocatalysis, resistant to surface contamination, and undesired electrode reactions. 33−38 However, they have disadvantages due to being expensive, nondisposable, and requiring pretreatment. On the other hand, pencil graphite electrodes (PGEs)  have some benefits involving cheapness, common technology,  good electrochemical reactivity and mechanical endurance, and  being disposable, commercially available, renewable, and  suitable for modification. 14,15,19,20 It has been reported that the polymer formed as a result of the uniform arrangement of azo dyes containing a hydroxyl group in the ortho position to the azoic group forms a stable redox-active layer, increases the electrode active surface area, and has an electrocatalytic effect toward the oxidation of small organic molecules with adjacent hydroxyl groups such as dopamine, ascorbic acid, and uric acid. 39,40 In view of these findings, a novel electrochemical sensing platform based on poly(chromotrope fb)-modified activated pencil graphite electrode (pCFB/aPGE) that overcomes the mentioned limitations was produced for the determination of Qn in vegetables and fruits. Additionally, this is the first instance in which chromotrope fb (CFB) was electropolymerized. It is also the first example of the voltammetric determination of Qn in red cabbage, cranberry, black mulberry, black raisin, and carob.

Preparation of pCFB/aPGE.
The preparation procedure of pCFB/aPGE and the plausible mechanism of electrochemical detection and electropolymerization voltammograms of pCFB/aPGE are shown in Figure 1A,B, respectively. First, the bare PGE was activated in 0.1 M of pH 7 phosphate buffer solution (PBS) and 0.1 M KCl via cycling 5 times between −0.6 and 2.0 V with 50 mV/s of scan rate 41 and denoted as aPGE. Then, CFB (1 mM) was electropolymerized onto the aPGE in 0.15 M of NaOH solution via cycling 20 repetitive scans between −1.2 and 1.6 V with a scan rate of 100 mV/s. In both cases, the total volume of the solution was adjusted to 10 mL. The peak current increase and decrease in the related regions indicate that the electropolymerization has taken place, as shown in Figure  1B. The final platform was stated as pCFB/aPGE and prepared daily. 2.5. Sample Preparation Procedure. Qn extraction was performed with minor modifications to reference. 42 In summary, 200 μM of Qn (for 50 mL) was added to the samples before the extraction process. Then, the extraction of Qn was initiated by mixing 5 g of separately ground red onion, red cabbage, cranberry, black mulberry, black raisin, and carob with 40 mL of 59.3% ethanol and continued in a shaking incubator at 59.2°C for 16.5 min. After centrifugation at 1600g for 30 min, the supernatants of vegetable and fruit samples were separated, and the final volume for each supernatant was adjusted to 50 mL with ultrapure water and tested to analyze the yield of quercetin, as shown in Figure 1C. The supernatants were further diluted 500-fold (20 μL/10 mL, approx. 400 nM of Qn) and analyzed with voltammetry. On the other hand, the supernatants were further diluted 40-fold (25 μL/1 mL, approx. 5 μM of Qn) and analyzed with a UV− visible spectroscopy.  (Figure 2A-b). With the activation of the bare PGE, the reversibility of the system improved and the peak currents increased significantly (Figure 2A-c). With pCFB modified on the aPGE, the reversibility of the system improved by about 35 mV and the peak currents increased by 48% compared to aPGE (Figure 2A-d). The peak at −0.37 V resulted from the interaction between the surface of pCFB/aPGE and the redox couple.

RESULTS AND
EIS spectra of the aPGE and pCFB/aPGE are shown in Figure 2B, while EIS spectra of the bare PGE and pCFB/PGE are shown in Figures S1 and S2, respectively. EIS spectra with the R s (C dl [R ct W]) circuit yielded R s (i.e., solution resistance), C dl (i.e., double-layer capacitance), R ct (i.e., charge-transfer resistance), and W (i.e., Warburg impedance) values, indicating that the pCFB/aPGE had great electron-transfer rate (21.9 μF) and low charge-transfer resistance (459 Ω) compared to the aPGE (8.92 μF and 716 Ω). In addition, real impedance was plotted against the square root of the inverse of the radial frequency (ω −0.5 ), and the slope values of 0.0011 for aPGE ( Figure S3A) and 0.0016 for pCFB/aPGE ( Figure S3B) obtained from the linear parts of the curves (low-frequency region) demonstrate that the diffusion ability of pCFB/aPGE is better than that of aPGE. 43 These results confirm that the EIS results are in agreement with the CV results.
In addition to all, the absence of redox peaks or low intensity and poor reversibility in Figure 2A-a,b coincides with the EIS spectra in Figures S1 and S2, for which it is not possible to simulate the relevant circuit.
SEM images for the bare PGE, aPGE, and pCFB/aPGE appear in Figure S4. It is seen in Figure S4A that the bare PGE surface does not show a uniform distribution, while the aPGE has a more uniform and channeled appearance ( Figure S4B). After the electropolymerization, it is clearly observed that these channels are filled with pCFB ( Figure S4C).
EDX spectra of the bare PGE, aPGE, and pCFB/aPGE are shown in Figure S5. It is observed that PGE contains aluminum (1.2%) due to the aluminum oxide in the structure of pencil graphite in addition to carbon (96.2%) and oxygen (2.6%) ( Figure S5A). 44 With the activation of bare PGE, the amount of oxygen increased approximately 3 times (7.9%) and sodium (0.5%), phosphorus (0.3%), and potassium (0.5%) appeared probably due to the phosphate buffer (KH 2 PO 4 + NaOH) and KCl present in the activation solution ( Figure  S5B). As expected, the amount of carbon (97.2%) increased and the amount of oxygen (2.2%) decreased after the electropolymerization. In addition, a low amount of nitrogen (0.6%) was observed in the spectrum due to the azo group in the structure of CFB ( Figure S5C). XPS measurements were fulfilled to acquire more information regarding the characterization of each modification step, and the spectra are given in Figure 3. C 1s bonds between 283.78 and 286.28 eV belong to the C−C and C−O− C structures. 45 A more than 2-fold increase in the C 1s signal at pCFB/aPGE clearly indicates a carbon-based formation on the surface (i.e., polymer of CFB). O 1s bonds between 531.88 and 532.08 eV correspond to the aliphatic C−O−C structures, and the increase of the O 1s peak at 531.88 eV for aPGE results from the activation procedure of the bare PGE. In addition, the peak that became apparent at 530.48 eV after activation is due to Al 2 O 3 . Although this peak is not observed at the bare PGE with XPS fitting, the characteristic O 1s behavior of Al 2 O 3 is shown in Figure S6A,B with red circles for both the bare and aPGE, respectively. It is observed that the peak at 530.48 eV disappears with pCFB coating and the characteristic Al 2 O 3 behavior is lost in the survey spectrum (O 1s region) for pCFB/aPGE ( Figure S6C). 46 The bonds at 533.58 and 533.28 eV are due to the partial aromatic C−O−C structures of graphite, 47 while the small peak at 535.68 eV for pCFB/aPGE corresponds to the C−O  Considering that the sulfur in the hydroxyl-containing naphthalene ring is more partially positive, it could be deduced that the peaks at 165.08 and 164.18 eV belong to the C-SO 3 group found in this region. 51 Consequently, largely harmonized CV, EIS, SEM, EDX, and XPS results showed that the pCFB/aPGE was decently and accurately produced for determining the Qn in real samples.
Using the characterization findings and the oxidation reaction of CFB, the possible electropolymerization mechanism and also the oxidation of Qn were proposed in accordance with the literature as depicted in Figure 1A. 52 F is the Faraday constant (96,485 coulomb/mol), R is the gas constant (8.314 J/mol·K), T is a temperature (K), and Γ is a surface coverage (mol/cm 2 ), which was found as 3.48 nmol/ cm 2 .
The type of electrode reaction belonging to the Qn was investigated by performing CV measurements at increasing scan rates between 10 and 1000 mV/s to determine whether the Qn's relocation to the surface of pCFB/aPGE was diffusion-or adsorption-controlled. The plot of the logarithm of peak height (log(I p , μA)) and the logarithm of the scan rate (log(v, mV/s)) with a slope of 0.81, log(I p ) = 0.810 log(v) − 0.024 (R 2 : 0.998), showed that the electrode reaction was based on a joint adsorption-and diffusion-controlled process ( Figure S7). The adsorption accompanying diffusion was probably due to the Qn's preference of the polarized pCFB surface.
The typical plot of E p (V) versus log(v, V/s) was drawn to calculate the apparent charge-transfer coefficient (α) ( Figure  S8). The slope of E p = 0.0511 log(v) + 0.2136 (R 2 : 0.959) was equal to 2.3RT/[(1 − α)nF], and α was found as 0.42, thereby depicting how the transition state acted uniformly between the responses of Qn and Qn-O-quinone against the applied potential. 20 The ratio of proton to electron was calculated as 1 from the E p −pH curve having a slope of −68.6 mV/pH ( Figure S9). This coincides with the oxidation of Qn including the 2electron and 2-proton process reported in the literature ( Figure 1A). 9,23 In addition, cyclic voltammograms were recorded with bare PGE, aPGE, and pCFB/aPGE in solution containing Qn to indicate the electrocatalytic effect as shown in Figure 4. The peak currents of Qn oxidation were found as 9.77, 19.92, and 43.02 μA for bare PGE, aPGE, and pCFB/aPGE, respectively. These results demonstrate a peak current increase of about 104% for aPGE over bare PGE, about 116% for pCFB/aPGE over aPGE, and about 340% for pCFB/aPGE over bare PGE. Accordingly, it was found that the pCFB film has quite good electrocatalytic effect on Qn oxidation.

Optimization Studies.
The parameters affecting the electropolymerization process including the pencil graphite grade, the concentration of CFB and NaOH, cycle number, and scan rate, and the analysis involving pH, the concentration of the buffer solution, the methanol ratio, and ionic strength were examined with pCFB/aPGE using 0.6 μM of Qn in the respective ranges and specified to be 2H, 1 mM, 0.15 M, 20, and 50 mV/s, and 7, 0.03 M, 20%, and 0, respectively ( Figure  S10).

Method Validation.
The DPV voltammograms and calibration curve belonging to the Qn appear in Figure 5. The peak height of pCFB/aPGE increased proportionally with Qn in 0.03 M (pH 7) PBS solution due to the oxidation of the adjacent hydroxyl groups. 9,23 The LOD and analytical range for Qn were obtained as 1.9 nM (i.e., from the blank signal, LOD = 3s/m, where s is the standard deviation of the blank solutions (n = 6) and m is the slope of the calibration curve; the standard error limit is 1.96 for LOD and LOQ calculation) and 0.01−1.2 μM, respectively. The sensitivity of the pCFB/ aPGE was calculated as 31.9 μA·μM −1 ·cm −2 , indicating a better value compared to the literature. 32 By using six different electrodes at each concentration level, the relative standard deviation (RSD%) values were found as 8, 5, and 3% in the presence of 40 nM, 0.6 μM, and 1.2 μM Qn, respectively. In addition, the repeatability of the pCFB/aPGE was investigated by using the same concentration levels and RSD % values were obtained between 2.2 and 5.4%. The results showed that the produced platform has remarkable reproducibility and repeatability.
Interference studies of various compounds, cations, and anions were carried out in a solution containing 200 nM of Qn using pCFB/aPGE with a criterion of ±5% change in peak height, as detailed in Table S2. Consequently, it was observed that the developed method could tolerate interfering components at least 20-fold more than Qn.
3.5. Sample Application. The method was applied to samples of red onion, red cabbage, cranberry, black mulberry, black raisin, and carob using pCFB/aPGE, while the UV− visible spectroscopic method was used as a comparative method. The voltammograms, UV−visible spectra, and calibration curves of real samples, and the results of both voltammetry and UV−visible spectroscopy are shown in Figures S11 and S12 and Table 1, respectively. The recoveries and RSD % values varied from 94.10 to 102.07% and from 0.86 to 3.04%, respectively. The high recovery values in red onion and cranberry are due to their naturally high Qn content. The values calculated by multiplying the results obtained from both  methods by the dilution factors were in agreement with the literature (∼250 mg/kg for red onion and ∼130 mg/kg for cranberry). 54,55 The precision and the trueness of the proposed method were evaluated statistically by using UV−visible spectroscopy, and the results showed that the developed method had good accuracy at a 95% confidence interval for all samples because the critical F (F critical ) and t (t critical ) values exceeded the experimental F (F experimental ) and t (t experimental ) values (Table 1).

CONCLUSIONS
To the best of our knowledge, a novel, disposable, and costeffective polymer film-modified electrode, pCFB/aPGE, was designed and produced using CFB as a monomer in the presence of NaOH for the first time and comprehensively characterized by various techniques including CV, EIS, SEM, EDX, and XPS. In addition, a simple, cheap, durable, disposable, and commercially available supporting surface (i.e., PGE) was proposed as an alternative to supporting surfaces such as CPE and GCE, which are more expensive, nondisposable, and require pretreatment steps for Qn determination. The developed platform, pCFB/aPGE, exhibited an extremely electrocatalytic effect toward the oxidation of Qn by enhancing the peak current by about 340% over bare PGE. The pCFB/aPGE achieved an excellent sensitivity (31.9 μA·μM −1 ·cm −2 , LOD = 1.9 nM) and a wide linear range (0.01−1.2 μM) and had good reproducibility and repeatability between 3−8 and 2.2−5.4% at different concentration levels, respectively. Besides, the electrochemical sensor platform displayed good applicability in detecting Qn in real samples such as red onion, red cabbage, cranberry, black mulberry, black raisin, and carob without being affected by the sample matrix. Voltammetric determination of Qn in these vegetables and fruits other than red onion was made for the first time. The recovery and RSD % values were obtained in the range of 94.10−102.07 and 0.86−3.04%, respectively. The results of the voltammetric method were compared to those of UV−visible spectroscopy, and no statistical difference was found at 95% confidence interval, depicting the excellent accuracy of the proposed method.