ИСТИНА

The tasks of recognition and classification of samples are solved by obtaining their "fingerprints" (spectra, chromatograms, voltammograms, etc.) and their analysis by chemometric methods. Fluorimetric methods are successfully used to discriminate and classify samples of various nature. In this work, we studied the feasibility of using fluorescent fingerprinting to establish the fact of irradiation of food samples with high energy electrons with sterilization purposes. We studied potato samples irradiated on an UELR-1-25-T-001 industrial electron accelerator with an energy of 1 MeV**. The extracts of potatoes were prepared by cutting the central part of the tuber into cubes (about 5 mm) and shaking with water or alcohol for 24 h. Extracts of potatoes with irradiation doses 1, 5 and 10 kGy were used for the analysis along with non-irradiated samples as blank. The experiment was carried out in 96-well plates in 6 parallel runs. The cell was filled with the potato sample extract, the carbocyanine as fluorophore, surfactant, necessary for the implementation of the aggregation mechanism of dye fluorescence, and buffer with pH 7,4. The images of the plates were digitized using ImageJ software. The obtained digital data were processed by the PСA method, followed by the construction of confidence ellipses for a confidence level of 95% (by using XLStat software). For the analysis, PС-1 and PС-2 components were used as the most informative. Based on fluorimetry data and using the PCA method, it was possible to divide samples of potatoes into classes according to the level of irradiation. Complete separation was not achieved; however, in the case of aqueous extracts, only classes with close levels of irradiation overlapped; the groups of samples with irradiation of 1 and 10 kGy do not overlap. In the case of alcoholic extracts, all groups strongly overlapped with no separation. In addition, extracts of non-irradiated potatoes of a different variety were studied, which formed a separate class, different from non-irradiated samples. This is a serious problem of the described approach, since the difference in the type of non-irradiated matrix (in this case, the potato variety) has an effect on thediscrimination along with the level of irradiation.