8E Identification of Impact Odorants of Wines 403
of nitrogen (100 mL/min) during a very long period of time (200 min). Desorption
is carried out with the help of a solvent, which provides a single extract that can
be used along the whole GC-O process. This strategy has been successfully used in
the screening of very different wines (Campo et al. 2005, 2006a; Petka et al. 2006;
Gomez-Miguez et al. 2007). It would not be fair, however, to say that this strategy
represents a universal satisfactory solution: there are some evidences about difficul-
ties in the elution of some mercaptans (Ferreira et al. 2007) and about a poor transfer
of some polar compounds to the trap (Escudero et al., in preparation). However, the
technique makes it possible to get quite easilysimple extracts representing relatively
well the composition of the vapors reaching the pituitary.
The third alternative is the use of a Solid Phase Microextraction fibre (SPME)
to collect the volatiles in the headspace. The technique is clean, very easy to use
and provides a good concentration of many volatiles on the headspaces of wine.
Because of this, it has been the technique of choice in some recent works (Marti
et al. 2003; Fan and Qian 2005; Gurbuz et al. 2006; Tat et al. 2007). Nevertheless,
the use of this technique is not exempt from problems either. On the one hand the
technique does not provide an extract, and on the other hand it is quite difficult to
optimize and validate and therefore to assess the reliability of the results. It should
be concluded that, although the technique isappealing, more research is needed in
order to establish its advantages and drawbacks.
8E.3 Techniques for Further Isolation and Identification
of Aroma Compounds
The results of the GC-O screening processes are hierarchical lists of odor zones
ranked by their potential sensory importance attending to their FD factors, Charm
values, Osme areas, NIF or SNIF values or odor intensities. At this time of the
analysis, the single identity criteria are the nature of the odor and its retention time,
most likely in a single GC column. The following step is to identify the odorant
or odorants behind each of the most relevant odor zones. This is also a critical and
sometimes difficult part of the work which should be carried out with rigor and
perseverance. A recent paper gives some general guidelines that should be assumed
by all authors (Molyneux and Schieberle 2007). A general scheme, with some spe-
cific comments from our personal experience working with wine odorants, is given
below.
- Standardization of retention times. This is the first step. Retention times of our
odor zones have to be normalized by those ofn-alkanes used as standards. To
do that, the mixture of alkanes, which can be directly obtained from different
chromatographic suppliers, is injected in exactly the same conditions in which
the GC-O experiment was carried out. As nearly all the GC-O experiments
use temperature gradients, the Retention Index or Linear Retention Index – not
K ́ovats Index according to the IUPAC (IUPAC 1997) – will be determined with
the formula