INFORMATION AMOUNT OBTAINED AT THE SEPARATION AND CHARACTERIZATION FOR SOME NATURAL FRUIT OILS BY CGC-MS TECHNIQUE

Using information theory concepts, a new method for the quantitative e stimation of analytical data from CGC-MS technique is described. This method was applied to the separation and characterization of very comp lex mixtures of some natural fruit oils, such as lemon, orange, grapef ruit, kiwi, cola and mango by capillary gas-chromatography - quadrupol e mass spectrometry technique. Information regarding several chromatog raphic determinations was computed by means of Shannon's formula, taki ng into account a maximum a priori uncertainty and experimental simila rity indexes as estimating values for the a posteriori probability ass igned to each component. For very complex mixtures experimental parame ters could be chosen, so that a maximum information amount on the dete rminations could be obtained.