IDM ¦ Technology/IT
Figure 2:
Aroma release of methyl salicylate5 (A) and 2-dodecenal6 (B) from foamed (white dots) and unfoamed (black dots) milk matrix
(4% milk protein, 1% gelatin, pH 4.5) and foam height (B, gray triangles) during heating to 40 °C. gray lines: Measurement
signal for 2-dodecenal below limit of quantification; *,** significant difference (* 10 %, ** 5 % probability of error).
the foamed matrix. Presumably, the changed conformation of the
proteins at the bubble interface leads to more binding sites for
these aroma compounds.
In the mouth model experiment, the foam had completely collapsed
after approx. 40 min (Fig. 2 B), so that the foamed matrix
no longer differed in texture from the reference. However, as
shown in Fig. 2, the aroma release curves still deviate after 60 min.
Equilibrium in the system does not seem to have been reached
yet. This means that – although the bubbles have disintegrated
and thus an unfoamed matrix is present – the aroma compounds
are bound to the matrix differently than in the unfoamed reference.
After extending the observation time to 24 h at 40 °C, it was
found that the effect of the introduced and collapsed gas bubbles
is still detectable for hours. This "memory effect" is probably due
to a reversible conformation of the proteins previously localized at
the bubble interface. Only when the protein is again in the conformation
as in the unfoamed matrix, the equilibrium in the aroma
distribution matrix-headspace is restored as in the reference. The
"memory effect" could be interesting for product development.
Example: Even if a foamed, fat-free, milk protein-based spread
partially loses its bubble structure during spreading, the aroma
release will differ significantly from the unfoamed spread of the
same composition.
After a change in aroma release due to foaming had been established
with the aid of the simplified mouth model, the question
44 · March/April 2022 ¦ international-dairy.com
remained whether this difference was large enough to be recognizable
to consumers. Sensory studies were carried out for this
purpose. Firstly, the odor of the ten aroma substances in foamed
and non-foamed matrix was compared. Second, the panelists
were asked to quantify the "taste" during consumption, which is
caused, among other things, by retronasal olfactory perception.
The aroma compounds were added to the matrix individually,
at concentrations previously determined to be pleasant. The attributes
that best described the aroma compounds were recorded
using descriptive analysis and the panel was trained on them. During
consumption, intensity was queried immediately after sample
intake, during chewing, and after swallowing.
Significant differences were found between the foamed and
unfoamed samples in both odor and during consumption. For example,
most panelists rated the intensity of allyl isothiocyanate (occurrence
in mustard) as higher during consumption of the foamed
milk sample compared to the reference. Camphor (occurrence in
rosemary) was perceived more intensely in the unfoamed reference.
Thus, both effects – increased as well as decreased aroma
release by foaming –, which were found analytically, could be detected
during the sensory studies. For five of the aroma substances
investigated, the results of the sensory studies were in agreement
with those determined in the mouth model. The differences can
probably be attributed to the longer observation time; a food
spends only seconds in the mouth rather than an hour. Addition-
/international-dairy.com
