Chemistry Reference
In-Depth Information
10.6 Decaffeinated Coffee and the Need For a Better
Limit of Detection
The industrial processes for removing caffeine from coffee products are not
quantitative, and a minimum level is acceptable depending on the country. For
instance, the Brazilian norms (Anvisa Resolution RDC number 277 from
September 22, 2005) state that maximum limits for caffeine are 0.1% and 0.3%
(w/w) for decaffeinated ground and instant coffee, respectively. The above
reported limit of detection is enough attend this norm. However, for good
quantitation, a preconcentration step is needed, because otherwise only a
positive or negative answer could be given since the limits stated in the norm
and the limit of detection are too close.
The preconcentration was obtained by liquid/liquid extraction using
dichloromethane. Initially, the pH of the coffee solution was adjusted to 10
using NaOH 1 mol L
21
to dissociate organic acids also available in coffee and
diminish the preconcentration of them in the organic phase. The initial 10 mL
sample is shaken three times with 20 mL of dichloromethane. After
evaporation of the organic solvent, the residual solid is solubilized in 1 mL
of deionized water, which results in a 10 : 1 preconcentration factor.
The recovery tests were between 89 and 97%, and the electropherograms
(Figure 10.3) show that the preconcentration did not cause enrichment of co-
migrating species. In fact, the whole profile is quite similar to the one obtained
for regular coffee.
Although the improvement of sensitivity allows quantitation of caffeine in
decaffeinated products, there are two drawbacks: (1) the preconcentration step
is time consuming and (2) it needs dichloromethane, which is eliminated by
evaporation at the end of the procedure.
This subject was revisited after the release of a new version of the C
4
D
(Francisco and do Lago 2009). This new version uses a compact electronics
and allows obtaining electropherograms with a better signal-to-noise ratio.
The limit of detection was reduced from 33 mg L
21
to 5.4 mg L
21
, which is
enough to eliminate the preconcentration step. However, there is yet an
advantage in using liquid/liquid extraction: the better reproducibility in
migration time. The great number of substances present in coffee contributes
negatively in two different ways: (1) the ionic species of the beverage contribute
to the conductivity of the sample, which impairs the initial mobilization of the
analyte; and (2) cationic species (remarkably proteins and cationic amphiphilic
species) can adsorb on the inner wall of the capillary and change the EOF,
which impacts the apparent mobilities of the species. The pH elevation and
solubilization steps in dichloromethane and water diminish the content of such
species, which is beneficial to the reproducibility of the migration time.
Nevertheless, considering the trade-off between the migration time reprodu-
cibility
d
n
0
t
2
n
g
|
9
and
the
improvement
of
the
limit
of
detection
without
dichlor-
omethane, we believe that this second approach is the best one.
Search WWH ::
Custom Search