Is orthonasal olfaction an equilibrium driven process? Design and validation of a dynamic purge and trap system for the study of orthonasal wine aroma

Synthetic wine models containing aroma molecules at levels producing known orthonasal odour intensity scores were used to assess the representativity of the extracts obtained with a previously developed purge and trap Gas Chromatography-Olfactometry (GC-O) strategy. The study revealed that the GC-O scores of polar odourants were very small or even null, even when these compounds were clearly perceived in the orthonasal evaluation of the synthetic wines. The different parts of the system were thoroughly evaluated and reoptimized, and although partial elution problems affecting polar odourants were detected and corrected, the underestimation of polar odourants persistently remained. A significant improvement was only achieved with a completely redesigned system in which a much larger flow of gas just flows out the sample vapours – instead of bubbling through the liquid – above the strictly non-stirred liquid sample. The GC-O scores obtained with this new design were much closer to the orthonasal odour intensity scores of the synthetic models and the underestimation of polar odourants was not observed. A final validation showed that the proposed system is able to provide significantly different GC-O signals for solutions containing increasing amounts of polar odourants at perithreshold levels. All this suggests that the transfer of aroma molecules from the sample to the nose during orthonasal olfaction is mass transfer controlled rather than equilibrium driven, which has been experimentally verified using Atmospheric Pressure Chemical Ionization-Mass Spectrometry (APCI-MS). Copyright © 2014 John Wiley & Sons, Ltd.