3.3. Comparison with other extraction methods
Concerning semolina pasta, previous tests demonstrated how
overnight extraction with n-hexane allows for complete extraction
of superficial contamination such as that migrated from the packaging,
but does not lead to a complete extraction of deep contamination
from different sources, which, instead, can be achieved
using PLE as described by Moret et al. (2014) or the procedure
described by Biedermann-Brem and Grob (2011). Complete extraction
was also achieved when applying MAS on selected samples
with different contamination profiles, as confirmed by the perfect
overlay of the LC–GC traces with those obtained by applying the
procedure described by Biedermann-Brem and Grob (2011) or PLE.
In the case of egg pasta, mineral oil migrated from the packaging
rapidly diffuses into the whole matrix (Barp et al., 2015), and
‘‘overnight” extraction (16 h) with hexane allows for an almost
complete mineral oil extraction, independently of the contamination
sources. Complete extraction, comparable to that obtained
using MAS, was obtained by prolonging solvent extraction over
the ‘‘weekend” (62 h) or using PLE and extracting the ungroundsample with hexane at 100 C for 1 h. LC–GC–FID traces obtained
for different samples under these different extraction conditions
overlaid perfectly with those obtained with MAS. When applying
solvent extraction, mineral oil was co-extracted with the fat, which
in the case of dry egg pasta required relatively long time to diffuse
from the whole product into the solvent: lower extraction times
gave lower fat extraction yields and hence incomplete mineral oil
extraction. MAS has also the advantage of avoiding fat coextraction,
which, although present in small amount (egg pasta
contains approximately 4% fat) can rapidly overload the LC column,
negatively affecting its performance and making frequent washing
necessary to restore it.