the three Soxhlet laboratories, of which two (A and C) are official
laboratories and the third (B) is a well reputed large laboratory in a
fcooperative organization, emphasxze the difficulties in gettxng comparable
xesults using the Soxhlet method. There is no doubt, hpwever, that there is
a remarkable good correlation between all the laboratories involved in the
tests. If laboratory A, one of the official laboratories, is used as a
reference, correlation coefficients and confidence intervals can be
calculated according to table 3.
In some recent studies it has been demonstrated that the particle
size does have a significant effect on the extractability of the
fat. Both the Soxhlet and the SOXTEC methods show the same results.
In these studies it was also shown that the total content of
extractable matter is generally higher with the SOXTEC system
than with the Soxhlet method. The differences between the two
methods are of course dependent on the solvent used,' the particle
size and the duration of the extraction. For ultimate accuracy
the analysts are therefore requested to check the results achieved
on the SOXTEC system with official results on the same samples.
Based on such data the extraction time in the SOXTEC system can
be adjusted to get directly comparable results or a small correc-
tion factor can be applied. The latter alternative is to be
recommended, when speed is of importance.
The influence of different laboratory technicians on the results was
tested in a small trial, the results of which are shown in table 4.
No differences were found.
Some samples may cause frothing in the extraction cups due to their fat
composition. To avoid this the temperature of the circulating water in
the Service Unit should not exceed 75 C, (usingdiethylether) . Another
cause of frothing is a too high fat concentration in the solvent. Therefore
we recommend lower sample weight if the fat content is high.
With low density samples such as cotton seed meal, the weight might have
to be reduced in order to prevent over filling of the thimble. The sample
in the thimble should not reach above the lower edge of the thimble adaptor.
1980.01.15 RA/ML
m 23/8Q
To Next Page
Loading...