A study of the extraction kinetics of the minerals from the white lady's bedstraw (Galium mollugo L.) by using an artificial neural network

摘要

The present work deals with the mineral composition of the extracts obtained from lady's bedstraw (Galium mollugo L.) by maceration, extraction under reflux (extraction at the boiling temperature of solvent) and ultrasonic extraction, using atomic absorption spectrometry. The main goals of the work were to evaluate the operation of an artificial neural network (ANN) of the selected topology, to determine the parameters of a kinetic model of unsteady diffusion of minerals through plant material, and to define the yield of minerals as a function of the time of extraction and the yield of total extract (resinoid). The ANN results showed a positive correlation with the experimental data, so they could be used to examine the kinetics of extraction of minerals from lady's bedstraw (G. mollugo), regardless of the extraction technique. The yield of minerals (K, Ca and Mg) was correlated with the time of extraction and the yield resinoid by a polynomial equation of the first order in both variables.