424 M.A. Pozo-Bay ́ ́ on and G. Reineccius
Fig. 8F.2Dynamic headspace dilution profile of four volatiles in aqueous () and ethanol ()
solutions (relative values). Each point is the mean of three replicates,error barsshow stan-
dard deviation (reprinted with permission from Tsachaki et al. (2005) J Agric Food Chem 53:
8328–8333. Copyright (2005) American Chemical Society)
static conditions, the absolute volatile concentrations above ethanolic solutions were
greater than those found above a water solution. This effect could not be completely
explained by the log P value since some compounds with very different log P values
showed very similar behaviours (Fig. 8F.2). Tsachaki et al. (2005) explained the
effect observed during dynamic studies as a direct result of the properties of ethanol.
Since ethanol is surface active, it will concentrate preferentially at the solution
vapour-interface. When ethanol evaporates, some areas of the interface are depleted
of ethanol, creating a surface tension gradient at the interface. Ethanol moves from
the bulk phase to replenish the depletedsurface areas carrying along an appreciable
volume of underlying liquid (i.e. aroma compounds). This phenomenon is called the
Marangoni effect (Spedding et al. 1993).
Concerning the impact of ethanol on aroma perception, Pet’ka et al. (2003)
showed that ethanol at low concentrations (under 10%) could decrease aroma
compound detection threshold. Nevertheless, Grosch (2001) observed that the less
ethanol present in a complex wine model mixture, the greater the intensity of
the fruity and floral odours. Although this effect could be easily explained by
the increased partial pressure of the odorants with reduced ethanol concentration,
they showed in GC-O (gas chromatography-olfactometry) experiments that ethanol
strongly increased the odour threshold of wine volatiles. In fact the reduction in
odour activity of the wine volatiles when ethanol was added was much larger than
the reduction in their partial pressure.