While the two-stage alkaline extraction at pH 11 (AEP (v)) was
essentially equivalent to the sequential pH 4.8 and 11 extraction
(AEP (i)) in terms of overall extraction efficiency, it was preferable
due to the desirable distribution of protein between the emulsion
and skim fractions. EAEP (ii) had the highest oil and protein extraction
yields but with 20.3% of the total protein in the emulsion fraction.
Unfortunately, the use of carbohydrases did not improve the distribution
of oil and protein between the emulsion and skim fractions. AEP
(i) and EAEP (ii) resulted in the formation of stable emulsions due to
the adsorption of protein emulsifiers of high molecular weight along
with the mixed phospholipid–oleosin layer to the emulsions interface.
Although the alkaline (pH 11) extraction of the emulsions produced
with AEP (i) and EAEP (ii) as the third stage produced very
unstable emulsions that were easily demulsified by acidification and
enzyme treatments, it increased the dispersion of oil droplets into
the skim fraction to ~26%. Therefore, AEP (v) not only compensated
for the good distribution of protein between skim and emulsion fractions
but also resulted in a simpler and more cost effective process
design. While the stable emulsion produced during this process could
be destabilized using 2.5% Protex 6 L, the use of phospholipases would
be more desirable since they will not affect the recovery of the proteins
from the skim fraction using isoelectric precipitation. Therefore, more
studies should be performed to determine the distribution of phospholipids
between skim, emulsion, and solid residual fractions and to evaluate
the major phospholipid components and their role in stabilizing
the emulsion.
Based on the results of this study, an aqueous process consisting
of two-stage alkaline extraction followed by protease assisted demulsification
is a potentially viable process for the extraction and
recovery of oil and protein from oilseeds, such as mustard