404 V. Ferreira and J. Cacho
I= 100 ×
[
tRi−tRz
tR(z+1)−tRz+z
]
where tRiis the Retention time of the odor zone i, tRzand tR(z+1)are the reten-
tion times of then-alkanes eluting immediately before (z) and after (z+1) the
odor zone, and z is the number of Carbon atoms in suchn-alkane. It should
be noted that if the extract is very concentrated, such as those obtained from
total extraction, the retention times of some compounds eluting near of a major
compound, such as isobutanol, isoamyl alcohol, ethyl lactate, diethyl succinate or
β-phenylethyl alcohol, will be significantly shifted. In these cases it is advisable
to spike the alkane solution to the extract. In the case an AEDA experiment has
been carried out, the retention times ofthe higher dilutions are more accurate.- Getting retention data from our odor zones in a second chromatographic
column with a phase of different polarity. Most typically the main column
will be a Carbowax-20M type phase and the second one a 5% phenyl-
dimethylpolysiloxane. In these last phases, many alcohols and particularly acids
have problems of tailing and fronting, respectively, which complicates the cor-
rect identification of odor zones. This problem will be particularly acute, again,
in concentrated extracts obtained by totalextraction. In any case, particularly if
the extracts have many odors, it will not be possible to establish a clear rela-
tion between the identities of the odor zones in the two columns. It should be
noted that many relevant aroma compounds are really “difficult” molecules from
the chromatographic point of view. This is particularly the case of aldehydes
(especially methional and phenylacetaldehyde) mercaptans and some polar com-
pounds such as Furaneol or Sotolon. These compounds are particularly sensitive
to the existence of specific adsorption sites on the chromatographic inlets. The
inertness of our columns is something that should be ensured before the GC-O
experiment via the analysis of Grob-type mixtures (Grob et al. 1981).- Running the extract on a GC-MS system. The extract will be injected in a GC-MS
system in scan mode using a column similar to the main column for the GC-O
experiment. Then-alkane mixture will also be injected under the same condi-
tions, and a screening will be carried out to obtain the MS spectrum and potential
identification of those peaks eluting at approximately the Retention Index of the
target odor zones.- Compiling data. Bringing together data of retention times and mass spectra to
propose candidates for the different odor zones. At this stage it is convenient
to do the search in “reverse” mode, i.e. to look specifically for those odorants
that according to the literature are more likely to be present in the sample.
There are several quite complete lists about the odorant composition of many
wines in the literature (Aznar et al. 2001; Ferreira et al. 2001a) and a review
(Ferreira et al. 2001b). Table 8E.1 presents a summary of some of the most
recent references including the type ofwine studied, the type of extract used
and the olfactometric technique.- Confirmation of candidates. This will be done by the injection of the pure stan-
dards in the two columns in the GC-O system and in the main column in the