Application of interval estimations of the total content of antioxidants for the analysis of juice products by the CUPRAC method using the Cu(II) – NC–PMM system

Сu(II) – neocuproinе indicator system (CUPRAC method, R. Apak) was used for spectrophotometric determination of the total content of antioxidant reductants (ΣAO) in juice products. Gallic acid, ascorbic acid, quercetin, luteolin, dihydroquercetin, which are typical for fruit and berry juices, as well as trolox, as a frequently used standard, were used as model compounds of antioxidant-reducing agents (AO). The polymethacrylate matrix (PMM) with immobilized reagents was immersed in the test solution for 20 minutes, and then photometrically measured at 450 nm. The aim of the work was to compare two methods for estimating ΣAO: a) in terms of the content of a standard substance (generally accepted method); b) in the form of an interval calculated taking into account the sensitivity coefficients when determining different AO at a selected wavelength (interval estimate according to V.I. Vershinin). It was established that the analytical signals of the model AO obeyed the Beer-Lambert-Bouguer law in the range of 10–6–2.10–5 mol/l, and were measured with a relative standard deviation not exceeding 10 %; deviations from the light absorption additivity of AO mixtures were statistically insignificant. As in the photometry of solutions, the sensitivity coefficients of different AO were different, and the differences decreased when using «normal» concentrations. The modulus of the relative error in estimating ΣAO in model mixtures, even when using optimal standards (luteolin), reached 17%. The possibility of interval estimation of ΣAO from generalized analytical signals measured by the CUPRAC method using PMM matrices was shown. The boundaries of the intervals practically did not depend on the standards used during the calculation. For all mixtures the actual values of ΣAO were within the boundaries of the corresponding intervals. The results of the group analysis of fruit and berry juices before and after their dilution confirmed the approximate correctness of the results obtained both by method (a), i.e. in terms of luteolin, and by method (b), i.e. in the interval form. Unfortunately, the width of the intervals that take into account the intragroup signal intensity was an order of magnitude larger than the width of the confidence intervals calculated from the results of repeated measurements of generalized signals (Student's test, n = 3, P = 0.95). This indicates the necessity of leveling the sensitivity coefficients for different AO. Thus, both methods of calculation are suitable for the group analysis of juice products, but method (b) leads to more reliable, more metrological correct, but less unambiguous estimates of the total content of the antioxidants.