Agriculture Reference
In-Depth Information
involve removal of, for example, soil, crowns, stems, roots, decayed outer leaves, and
so on for certain fruits and vegetables and removal of trimmable fat for meats.
3.3.2 Effects of Sample Processing
Method validation is usually performed using laboratory spiked samples, and these
samples are normally spiked following sample preparation and processing. Sample
processing can affect the quality of the
final analytical results for two possible reasons.
First, if the laboratory sample is poorly homogenized, then representative analytical
portions cannot be abstracted. This is particularly signi
cant when using modern
MRMs where small analytical portions (often 10 g or less) are used. It is therefore
recommended that the homogeneity of the laboratory sample be checked by under-
taking replicate analyses on two or more analytical portions. Second, there could be
stability issues with regard to certain analytes and/or analyte/commodity combina-
tions. In particular for fruits and vegetables, homogenization at room temperature will
disrupt plant cells and release enzymes that can react with and/or degrade certain
analytes. In such cases, cryogenic milling [14] can not only signi
cantly reduce
analyte losses but also improve the homogeneity of the laboratory sample.
3.3.3 Extraction Efficiency
Spiked samples may not properly represent
samples containing incurred
residues of the same analyte(s) because spiked analyte molecules are unlikely to
have been in contact with the sample for long enough to allow any possible
“
real
”
”
to the matrix to occur. For example, Matthews [15] demonstrated that only 52% of
radiolabeled chlorpyrifos-methyl could be recovered as the parent compound from
wheat grain after 5 months of storage. Obtaining extraction ef
“
binding
ciency data to
demonstrate the validation of MRMs is, however, not straightforward. This is due
to the limited availability of samples containing incurred residues for many analyte/
commodity combinations. Simple spiked extraction experiments are generally the
only option available. It should also be noted that the effects of sample processing and
extraction ef
ciency are not assessed by most pro
ciency tests.
3.4
INITIAL METHOD VALIDATION
In order to obtain results that are reportable, a SDL has to be determined and checked
regularly. Initial method validation, as for quantitative MRMs, has to be performed
for each analyte that is to be sought as well as for a typical example taken from each
commodity group. Commodity groups have been de
ned in a number of AQC
documents based on similarity in chemical composition (e.g., water, sugar, protein,
fat/oil) and characteristics such as pH (Table 3.1). Some foods or crops, such as tea,
coffee, cocoa, spices, and hops, are considered unique in terms of their composition
and therefore need to be validated individually [16,17].
