Aroma release from wines under dynamic conditions.

Aroma release from wines and model ethanolic solutions during dynamic headspace dilution was measured in real time using atmospheric pressure chemical ionization-mass spectrometry. Model ethanolic solutions maintained the headspace concentration of volatile compounds close to equilibrium values during gas phase dilution over 10 min. Wine samples (with the same ethanol content) did not maintain the headspace concentration of volatiles to the same extent. Wine components and acidity ((+)-catechin, glycerol; pH 3.6) in model ethanolic solutions (120 mL/L) had no effect on the volatile headspace concentration during dynamic headspace dilution. However, in the presence of certain proteins (beta-lactoglobulin, beta-casein, bovine serum albumin), the model ethanolic solutions failed to maintain their volatile headspace concentration upon headspace dilution, but other proteins (thaumatin, mucin, lysozyme) had no effect. Thermal imaging of the model ethanolic samples (with and without beta-casein) under dynamic headspace dilution conditions showed differences in surface temperatures. This observation suggested perturbation of the ethanol monolayer at the air-liquid interface and disruption of the Marangoni effect, which causes bulk convection within ethanolic solutions. Convection carries volatile compounds and warm liquid from the bulk phase to the air-liquid interface, thus replenishing the interfacial concentration and maintaining the gas phase concentration and interfacial surface temperature during headspace dilution. It is postulated that certain proteins may exert a similar effect in wine.