Electronic nose based discrimination of a perfumery compound in a fragrance

Abstract

An electronic nose (e-nose) device developed at the University of Buenos Aires was applied to detect the presence of a given perfumery compound (also so-called the perfumery note, refereed as mangone) in a fragrance, at very low weight percentages. The results were compared with sensorial analysis performed by trained panelists and gas chromatography mass spectroscopy (GCMS) measurements. The triangle test for detection of the perfumery note in the fragrance was performed by a set of 20 trained panelists. Less than 40% of the panelist could identify the presence of the strange note for concentration 10 −2% (w/w), and similar percentages were obtained for lower concentrations. Detection by CGMS was difficult at those concentrations, because of the low percentages of the perfumery note and the similar retention times obtained for the note and other compounds included in the fragrance. The developed electronic nose provided fingerprints for different odors, associated to different samples that were used to build up an odor database. Then, two different multivariate data analysis were performed, the non-supervised principal component analysis (PCA) and an artificial neural network (ANN), in order to discriminate the samples with or without mangone. Measurements of several dilutions of mangone up to 10 −4% (w/w) were performed to obtain the database. Both methods, PCA and ANN, were successful in the discrimination process of samples with from those without mangone. In particular a 100% success was obtained by using a radial basis function (RBF) artificial neural network, even when considering the more diluted samples.