ELECTRONIC NOSE CHEMICAL SENSOR FEASIBILITY STUDY FOR THE DIFFERENTIATION OF APPLE CULTIVARS

Abstract

The ability to analytically differentiate and match intact apple (Malus domestica, Borkh) fruit and fruit juice extracts from different apple cultivars is of interest to the food industry. This study tested the feasibility of detecting the difference among volatile gases evolved from intact ‘McIntosh (Buhr),’ ‘Delicious,’ and ‘Gala’ apples and their extracted juice using a prototype 32-array polymeric detector chemical sensor. All data were first processed to obtain principal components. PCA analysis clearly separated whole ‘McIntosh,’ ‘Gala,’ and ‘Delicious’ samples from juiced on day 1. PCA analysis of day 2 samples showed clustering of whole vs. juiced for all three cultivars, although there was some overlap between the clusters. A soft independent modeling of class analogy (SIMCA) class discrimination of the sensor principal component data sets was then performed to determine the degree of difference. SIMCA analysis of the same samples showed a pronounced difference (SIMCA value >3.00) for only the ‘McIntosh’ samples. SIMCA values between 2.00 and 3.00 were found for the other two cultivars on day 1. For day 2 samples, no SIMCA values greater than 2.00 were found for any cultivar whole vs. juiced. PCA analysis showed clear separation between cultivars for day 1 whole samples. SIMCA analysis showed that there was a difference between ‘Delicious’ and ‘McIntosh’ and between ‘Delicious’ and ‘Gala.’ Neither PCA nor SIMCA showed good separation between day 2 whole cultivars, nor between juiced cultivars on either day. As a reference, the same sample headspace volatile gases were analyzed with a mass spectrometer. A hierarchical cluster analysis (HCA) of the principal components from the mass spectrometer data sets revealed five clusters that discriminated differences among intact apple and apple juice samples but did not discriminate between samples from different apple cultivars.