Measuring the effects of food composition on flavor release using the retronasal aroma simulator and solid phase microextraction

Foods are complex matrices containing a variety of ingredients, including nonvolatile macro-components that selectively affect the volatility of aroma compounds. Additionally, the flavor released is modulated by the experiential processes of sniffing, drinking, eating, etc. For example, the relative composition of gaseous aroma from a food when sniffed (the orthonasal route) is different from the relative composition of vapors released from the same food when eaten (the retronasal route). The Retronasal Aroma Simulator (RAS) was developed by Roberts and Acree (1994) to simulate conditions in the mouth that affect volatility. In this work, a modified RAS was attached to a modified mass spectrometer (MS-Nose developed by Taylor and Linforth) to verify that the flow from the RAS under certain conditions contains a comparable relative composition of aroma compounds to the retronasal breath from humans. Volatile release measured on the MS-Nose from imitation cheese, model chocolates, and five other foods was compared between the RAS and human panelists. Comparison of relative aroma composition produced from the RAS under conditions that approximate human mastication yielded correlation coefficients between 0.85 and 0.99. The effects of temperature, airflow rate, saliva to food ratio, and blending rate in the RAS on flavor release were also measured. Nine base ingredients and environmental conditions of model beverages containing 29 aroma compounds were varied by a one fourth replicate of a 2 full factorial experimental design consisting of 133 combinations of the factors. Ingredient effects on flavor release were found to be both selective and interactive.