| Abstract
| - We carried out a QSPR (quantitative structure−property relationships) approach to evaluate theinfluence of the chemical structure of aqueous matrixes over the partition coefficient between thegas phase and the matrix. The determination of the partition coefficient of flavor ingredients wasperformed by headspace analysis at equilibrium for both saline solution and ι-carrageenan gel. Startingfrom an initial list of 90 descriptors, we selected 10 descriptors to perform equation generation by theGFA (genetic function approximation) method available in the Cerius2 package. The best obtainedequations involve only five descriptors, which encode electronic properties of charges repartition onthe molecule (Jurs-RNCS and Dipole-Z) and molecules' shapes (PMI-Y, Shadow-XY, and RadOfGyration), both for saline solution and for ι-carrageenan gel. However, the best-fitting equation forcarrageenan gel is obtained with a quadratic relation, suggesting that the effect of carrageenanpolymers only modulates but does not change the interaction of aroma compounds with watermolecules. Keywords: ι-Carrageenan; gels; aroma release; partition coefficient; interaction; headspace analysis;QSAR; QSPR; genetic function approximation
|
