398 V. Ferreira and J. Cacho
- Based on the assessment of intensity: OSME (McDaniel et al. 1990; Miranda-
Lopez et al. 1992), cross-modality matching (Eti ́evant et al. 1999), and posterior
intensity evaluation (Ferreira V et al. 2003b; van Ruth 2004)8E.2.1.1 Techniques Based on Threshold
EDA is one of the most frequently used methods for the screening of flavor impact
compounds, probably because of its simplicity. The theoretical background of the
technique showing practicalhints to obtain a reproducible and traceable signal have
been clearly established (Ferreira et al. 2002b); however, most researchers keep on
using the technique in an archaic and refutable way, surely because they think that
the possible errors in the ranking of odorants will be corrected in the subsequent
quantitative analysis. According to this technique, the flavor extract is sequentially
diluted (following a rate R, where R is usually 2, 3, 5 or 10, but could be as high
as 20) and each dilution is analyzed by GC-O by usually a small number of judges,
often (most incorrectly) by a single one. The Flavor Dilution of an odorant corre-
sponds to the maximum dilution at which that odorant can be perceived by at least
one of the judges. Numerically, if the last dilution at which compound c was per-
ceived was P (where P usually is 0, 1, 2,...n), its FD is RPor, more correctly, RP+^0.^5.
When several judges are used in a study, usually (and incorrectly) the maximum FD
is provided as the FD factor of that compound, although the result should be the
geometric mean of the FD factors recorded for each judge (Ferreira et al. 2002b).
Charm is similar in concept, but the recording of the signal is made with the help
of a computer and the judge records not only whether there is an odorant, but the
time that the odorant is perceived just by holding pressed one key of the keyboard.
Consequently, in this case the signal is a kind of peak called Charm peak defined by
its area and height. The height of the Charm peak is the same than the FD obtained
in an AEDA experiment. A basic scheme of the techniques is shown in Fig. 8E.1.
These techniques provide a ranking of aroma compounds relating to the ratio
concentration in the extract/odor threshold in air. Working properly (using a panel of
judges and 1/10 dilutions) it is possible to obtain, with reasonable effort, confidence
intervals for the results, which are most useful for comparative purposes, and to
relate them to the concentration of the odorant in the extract.
8E.2.1.2 Techniques Based on Frequency of Citations
In these techniques an extract is injected in the GC-O system and the effluent is
smelled by a group of judges. What is recorded is the number of people that are able
to detect the odorants. The proportion of people able to detect an odorant is called
the Nasal Impact Frequency (NIF), and if the time for which the sensation lasts is
also recorded, the product NIF by time is called Surface of Nasal Impact frequency
(SNIF). This technique obviously requires a group of judges and the experiment
is carried out at a single dilution of the extract. The theoretical background and
some interesting applications of the technique are well documented in the scientific
literature (Pollien et al. 1999; Debonneville et al., 2002). The technique is more