In the food, Leverage and cosmetic industry and so on, odor sensing systems
instead of human sensory test are demanded. MB have developed odor sensing
systems using QCM (quartz crystal microbalance) sensor array and pattern r
ecognition method. Since the properties of the sensors depend on the gas so
rption characteristics of the sensing films coated on them, the optimum fil
ms according to target odors should Le selected. In this study, we tried to
select sensing films appropriate for discrimination of slightly different
apple flavors. The examples of typical apple flavors were prepared blending
9 compounds. The sensing films were extracted from various kinds of materi
als such as lipid, stationary phase material of CC (gas chromatography) and
cellulose. The selection method under the condition of tho small number of
measurements was studied. We analyzed the data of steady-state sensor resp
onses in terms of the Euclidean distance, and the: films appropriate for ap
ple flavor discrimination were successfully selected.