3.2. Determination of Hg by CV-AFS
Different types of tree nuts samples were ground and digested
in pressurized Teflon vessels using microwave heating. In this
study, four different acid mixtures were employed to digest the
samples in order to obtain complete sample dissolution. The
mixture of acids and/or oxidizing reagents tested were HNO3:
H2O2:HCl; HNO3:H2O2; HCl:HNO3; and HNO3. As a pre-digestion
step, samples were sonicated with the reagents in an ultrasonic
water bath. The procedures allowed the digestion of 0.5 g of sample
in safe conditions. After the digestion of samples in the microwave
oven, the use of the HNO3:H2O2 mixture and that of just pure HNO3
were selected, because, in the case of employing HCl, the samples
were not totally dissolved.
Based in our previous work for determination of Hg in rice [24],
we assayed the two selected conditions for mercury cold vapour
generation: 25% (v/v) HCl and 5% (v/v) HNO3. In both cases, we
added KBr/KBrO3 to release Hg from remaining organic matter and
hydroxylamine hydrochloride to remove the excess of oxidant
agents.
To evaluate the effect of the sample matrix on the mercury
fluorescence the slope of the standard addition method applied to different volume of digested sample solutions was compared with
the slope of an external calibration equation. The fluorescence
signal was obtained with a gain of 1000 due to the high dilution of
sample solutions used to avoid the interference of nitrous acid and
nitrous oxides generated during the digestion. The external
calibration and standard addition method were both measured
under the same conditions. Table 7 summarises the values of the
slopes obtained for the external calibration and the ratio for the
standard addition curves with different digestion mixtures.
The use of 5% (v/v) HNO3 acid provided higher sensitivity than
25% (v/v) HCl with external slope values that varied from 3468.9 to
4130.9 fluorescence unit ng1 mL. The results found show a
significant difference between the slope values of the standard
addition curves with respect to the external ones, obtaining a
relative percentage of the slopes ratio from 56 to 83%, which
denoted the presence of a strong matrix effect in all the cases.
Results with different dilution factors of digested sample (see
Table 7) show that an increase in the dilution factor of the sample,
decreased the sensitivity. Moreover, among the conditions studied,
digestion with HNO3 alone and the addition of 5 mL of digested
solution provided a lower matrix effect. However, due to the low
concentration of mercury present in the tree nuts, the fluorescence
signals of the samples were of the same order as the reagent blank
signals, thus making the determination of total Hg by CV-AFS
inadequate.