406 V. Ferreira and J. Cacho
d) The target odor zone corresponds to anonpolar compound of average volatility.
In this case a primary isolation by a dynamic headspace technique followed by a
normal phase fractionation is a good choice.8E.4 Quantitative Determination of Wine Active Odorants
Given the different concentration levels at which wine active odorants are found, it is
absolutely necessary to use an array of analytical methods for their determination.
Major volatiles (C>0.1 mg/L), such as fusel alcohols, fatty acids and their ethyl
esters, can be directly determined by GC-FID using many different preconcentration
techniques such as direct injection, simple solvent extraction, microextraction with
a solvent, Stir Bar Sorptive Extraction (SBSE) dynamic headspace or SPME. Minor
volatiles (1<C<0.2 mg/L) can also be determined by direct GC-MS analysis using
a similar array of sample preparation techniques. Results obtained with the different
sample preparation techniques differ in the range of analytes available with accuracy
and in the level of automation of the method. From this last point of view, solventless
fully automated sample preparation techniques such as SBSE, SPME, or its recently
introduced competitor Solid Phase Dynamic Extraction (SPDE) present some major
advantages, since they are very easy to use and are relatively cost-effective. From the
point of view of the range of analytes available, however, they cannot compete today
with the broad possibilities offered by Solid Phase Extraction. Headspace techniques
fail in the analysis of some less volatilecompounds, and SBSE at present, also has
problems in properly extracting some polar compounds.
However, is not the analysis of minor wine volatiles that still presents difficul-
ties. With the level of sensitivity and automation of the analytical techniques, the
determination of many odorants atμg/L level is a simple analysis. The difficul-
ties come when the analytes of interest cannot be easily determined using a single
non-selective-preconcentration step. This will happen when the analytes are diffi-
cult to extract because they are very polar and/or not very volatile or when they
are present at very low levels. The concentration level at which the analysis of
an aroma compound becomes difficult is related to its polarity and to the quality
of its mass spectrum. For instance, the analysis of 2,4,6-trichloroanisol (TCA) at,
let’s say, 20 ng/L is not a very difficult analysis, because this molecule is quite
nonpolar (easily extractable, relative volatile) and has a mass spectrum with abun-
dant high mass ions.In contrast, the analysis of methional or of sotolon at 1μg/L
is quite difficult because these compounds are very polar (difficult to extract, not
very volatile) and their mass spectra lack powerful ions. For these difficult analytes,
some of which are very important wine impact aromas, specific strategies must be
developed:
Carbonyls. The direct analysis of some relevant carbonyls such as methional,
phenylacetaldehyde, isobutyraldehyde, isovaleraldehyde, 2-methylbutanal or
(E)-2-nonenal, even if it can be carried out (da Costa et al. 2004), is not a very
convenient technique. Different alternatives have been proposed, most of