14B-15

S. JIAMYANGYUEN and W. J. Harper. Department of Food Science and Technology, The Ohio State University, 122 Vivian Hall, 2121 Fyffe Road, Columbus, OH 43210

Differentiation of aromas of food and related materials is now made possible by a technique of arrays of electronic sensors, generally known as "Electronic Nose" technology. This device is capable of detecting and differentiating complex mixtures of volatile compounds. One of the Electronic Nose applications being used is in the study of ice cream stick. We hypothesize that ice cream sticks obtained from different sources may exhibit different flavor profiles, which in turn possibly affect consumers' flavor perception of ice cream bars. Our objective was to differentiate flavor compounds present in different wooden ice cream sticks. Five ice cream sticks were evaluated with the Hewlett Packard headspace/mass detector (HP/ MS). This new electronic nose that combines dynamic head space sampling with a mass detector in which different mass become a sensor has been used. Varieties of ice cream stick samples came from different sources and were designated as A, B, C, D, and E. The multivariant technique and a pattern recognition are utilized as a part of data analysis. This study showed that principle component analysis (PCA) differentiated all five sticks from one another, with B being closest to A. However, cluster analysis did not separate D and E. When considering masses between 48 and 164, there were more than 50 differences, both higher and lower, between A and the others. At the specific mass of 60 (carboxyl group), the result showed that A had a higher level of volatile acid than any of the other sticks. This preliminary work showed that by using HP/ MS, not only information regarding food aroma differentiation may be provided, but also an understanding of the chemical basis for the aroma differentiation. KEYWORD: Electronic nose, ice cream sticks, aroma differentiation, multivariant analysis, a pattern recognition