techniki chemometryczne; informacja; sygnały nieselektywne; przeciwutleniacze; chromatografia cieczowa
Katowice : Uniwersytet Śląski
Quality of products depends on many parameters including their unique chemical composition. By definition, the global approach for the quality evaluation is necessary. It often requires simultaneous control of more than one parameter. Thus, the evaluated product is described by instrumental signal(s) possibly reach in chemical information (containing the inputs of several chemical components included in the product). The nonselective signals (e.g. chromatograms or spectra) are used frequently. They are considered as chemical fingerprints since the information they contain is unique. Nevertheless, the complex chemical composition of products leads to overlaps of particular signals characterizing different components. Isolation of significant information from such complex data facilitates the application of chemometric approaches. Considering their advantages the
chemical fingerprints are used to evaluate quality of food products, in process analytical technology, and to monitor the environment. The aim of the research was to extract the relevant information form the nonselective signals in order to develop novel analytical approaches for the quality evaluation. Different approaches aiming at evaluation of the quality of sugar (in course of its color and ash content), diesel oil (in course of the level of the imposed excise duty) and food products as coffee, pepper mint, oregano, basil, tomato paste, and rooibos tea (in course of their total antioxidant capacity and total polyphenol content) are developed. The nonselective signals like excitation-emission fluorescence matrices, high performance liquid chromatograms detected by diode array detector (HPLC-DAD) and infrared spectra were registered for examined products. Obtained signals were efficiently modelled using selected chemometric methods. Moreover, the developed new chemical approaches have been validated according to their chemometric and analytical parameters.
The application of the nonselective signals in the developed approaches significantly reduces or fully eliminates the use of classic, laborious, time and money consuming analytical methods. All of performed experiments were designed in accordance with the green chemistry principles, i.e. to limit the amount of harmful chemical wastes. Thus, the nonselective signals were collected for raw samples (without any laboratory preparation) or
with the use of distilled water as the only solvent. Regarding to the high efficiency of the developed methodologies they can be recommended for the routine laboratory or on-line quality evaluation. It should be emphasized, the presented results refer to the preliminary studies and the application must be preceded by the
additional experiments (taking into account the various changes of chemical composition of samples).