Linearity was performed analyzing standard concentrations
from 0.1 to 100 mg/L (six concentrations-five replicates per concentration)
and fortified matrices from 0.003 to 2 mg kg1 (ten
concentrations-six replicates per concentration): the fortified
matrices were muscle fish, mussels, crustaceans and cephalopods.
Limit of quantification was evaluated at least ten times over the
signal of blank sample (0.005 mg kg1 for each metal). Specificity
was performed analyzing negative samples (n ¼ 20) and verifying
the absence of inferences. Precision was calculated fortifying
samples at three different levels (six replicates per level) in the
validation range and evaluating repeatability as coefficient of
variation (CV 7% for Cd, 5.6% for Hg and 6% for Pb). Trueness was
evaluated on the same samples recovery percentage calculated
(86.6% for Cd, 91.1% for Hg and 65.3% for Pb). Reproducibility was
evaluated on the same samples at the same concentrations
repeating the analysis on three different times during a month (CV
8.3% for Cd, 8.2% for Hg and 9.5% for Pb). Uncertaintywas calculated
with “in house data” using Bottom up approach (21.1% for Cd, 20.4%
for Hg and 21.4% for Pb). The method was continuously monitored
by analysis of BCR samples in every batch and evaluating that the
parameters were within the predicted values.