Producing headspace extracts for the gas chromatography–olfactometric evaluation of wine aroma

In this work different designs of headspace devices were studied. Using these designs, 23 volatile compounds of wine aroma added to a synthetic wine were purged by an inert gas (nitrogen) and trapped in a solid-phase extraction (SPE) cartridge. Efficiency was measured as the similarity between the odour intensity in a model wine in the glass and the odour intensity recorded in the gas chromatography–olfactometry (GC–O) experiment. Using the initial design, the polar compounds were not detected in olfactometry from solutions in which the measured orthonasal intensity was intermediate; these compounds were undervalued by this technique. New headspace device designs were then tried. The least amount of under-valuation was obtained when the purging distance between the sample and the trapping system was reduced, which suggests that polar molecules were being trapped by polar surfaces in the original design. The improvement in detection was general, affecting all chemical compounds, resulting in overvaluation of non-polar compounds. A strong correlation of log P with the under-valuation/overvaluation of compounds was noted. From the comparison of headspace devices, one design was found which efficiently transferred even the high-polarity compounds from the wine to the trap. Therefore, this system is recommended for obtaining extracts of aged red wine for use in a screening system based on olfactometry.