The dissertation is devoted to the development of a new voltammetric sensor based on a carbon-paste electrode modified with β-cyclodextrin, to the study of the peculiarities of redox reactions on its surface of selected food azo dyes (Tartrazine — TAR, Sunset yellow FCF — YS, Сarmoisine — CAN, Ponceau 4R — P4R, Allura Red AC — ARAC), to the development of methods for the voltammetric determination of food dyes in various food products. It was established that all the investigated dyes on the developed sensor have both the azo group reduction peak and its oxidation peak. According to the cyclic voltammograms, such azo dyes as YS and P4R are oxidized quasi-reversibly, while CAN, TAR and ARAC are completely irreversibly oxidized. The ratio of protons to electrons participating in the oxidation of dyes was established - 1:2 (YS, TAR, ARAC, P4R) and 1:1 (CAN).
When studying the effect of the speed of the potential sweep, it was established that the nature of the oxidation current for all dyes on the surface of the developed sensor has an adsorption character. Based on this, using Laviron's theory, the number of electrons participating in the process of oxidation of dyes on the surface of the electrode was calculated: 2 for YS, TAR, ARAC, P4R and 1 for CAN. During a detailed study of cyclic voltammograms, it was established that for such dyes as YS and P4R, with an increased speed of the potential sweep, the ratio of the oxidation peak current to the corresponding reduction peak decreases, which indicates the presence of an irreversible chemical reaction with an oxidation intermediate (ECir mechanism). During a detailed study of the redox behavior of dyes during a cyclic potential sweep, it was established that after the oxidation process, peaks of a new reversible redox pair appear on the cyclic voltammograms (except in the case of TAR). The corresponding experiment showed that these redox couples can be formed both after the process of oxidation of dyes and after the process of their reduction by the azo group. Based on the obtained information, a general mechanism of oxidation of food azo dyes on the surface of a carbon-paste electrode modified with β-cyclodextrin was proposed.
Using the volumetric method of modification, it was established that the maximum oxidation current of food dyes is achieved at 10 wt.% content of the modifier (β-cyclodextrin) in the carbon paste electrode. Using the Renedels-Shevchik equation, it was determined that the active surface area of the carbon-paste electrode modified with β-cyclodextrin is 0.105 cm2, and the potential difference of the peaks of the standard system is 70 mV, which indicates good conductive capabilities of the developed sensor. Using the approach of adsorption-inversion voltammetry, a modification is proposed that allows reducing the amount of solution for determination to 10 μl, which significantly reduces the amount of chemical waste, the number of uses of analytical standards, and reduces the cost of determination. The cyclic voltammetry method was used to optimize such parameters as adsorption pH (TAR, YS, CAN, P4R, ARAC – pHads 2) and pH of electrolysis (TAR, YS, P4R, ARAC –pHelec. 7; KAN – pHelec. 3). Square-wave potential sweep voltammetry was used as a quantitative method. After the optimization, the following values were selected: oscillation frequency 15 Hz for all dyes, oscillation amplitude — 50 mV (YS, CAN, ARAC) and 35 mV (P4R, TAR). 5 min was chosen as the optimal accumulation time for all dyes. The short- and long-term stability test showed satisfactory results for the determination of food azo dyes on a carbon paste electrode modified with β-cyclodextrin. Under optimal conditions for the determination of food dyes on a carbon-paste electrode modified with β-cyclodextrin, a validation experiment was conducted and the main metrological characteristics of the sensor were established: the range of linearity (YS: 4.50-0.57 μg/ml and 0.57-0.07 μg/ml; CAN: 5.00-0.30μg/ml; TAR: 5.30-0.17 μg/ml; P4R: 3.00-0.19 μg/ml; ARAC: 5.00-0.16 μg/ml), LOD (YS: 42 ng/ml; CAN: 101 ng/ml; TAR : 60 ng/ml; P4R: 102 ng/ml; ARAC: 60 ng/ml), LOQ (YS: 140 ng/ml; CAN: 337 ng/ml; TAR: 200 ng/ml; P4R: 340 ng/ml; ARAC: 200 ng/ml), reproducibility (YS: 7.10 %; CAN: 8.80 %; TAR: 7.40 %; P4R: 6.90 %; ARAC: 7.20 %), recovery rate (YS: 96%; CAN: 93%; TAR: 93%; P4R: 96%; ARAC: 94%) and relative shift (YS: -4%; CAN: -7%; TAR: -7%; P4R: -4%; ARAC: -6%). To test the developed methodology, commercially available food products were selected and divided into 5 classes: carbonated and non-carbonated sweet drinks, juices, jelly candies, low-alcohol strongly carbonated drinks and energy drinks. Using Fisher's test, it was shown that for all selected samples (except juice and jelly candy samples), the determination of dyes on a carbon paste electrode modified with β-cyclodextrin gives an error comparable to the error of HPLC determination.
