58
5 Calibration
5.2 Evaluation of Recovery with Reference Materials
In the case of chemical measurements, pure chemical substances are often used to
calibrate the measurement instrument. They can be used as a solid substance or as a
solution, the latter is commonly named the ‘standard solution.’ Thus, the metrological traceability in chemical measurements can be achieved by linking the respond of
instrument to known amount of pure chemical standards (solids or dissolve, depending on the instrumental technique used for performing measurements), knowing its
atomic and molecular mass in moles, the dimensionless SI unit. In practice, we can
see two main limitations of such an approach: first, it is not possible to acquire completely pure chemicals; second, it is not possible to isolate the analyte completely
from the samples, meaning to separate the analyte from the matrix. In addition, setting the calibration relationship for pure chemical substances (the first step described
in the definition of calibration given in VIM) and then determining it according to
the content of an analyte in a real analytical sample (the second step described in the
definition of calibration given in VIM), in a case when measurement can be interfered
with by matrix component can lead to systematic errors. In this case, it is necessary
to check the influence of the matrix of the sample on the analytical signal, since it
may alter its physical and chemical properties, or cause undesired interference and,
consequently, induce a change of values a and b of the calibration function. The
influence of the matrix can be archived via evaluation of the recovery.
Ideally, the best approach is to correlate the calibration relationship for references
which should be identical in chemical composition and physical form to the test
objects (samples). Establishing the calibration dependence with the use reference
material of identical composition to that of real test samples is in most cases difficult
to implement, due to the lack of the wide range of them. Thus, it is recommended to
use reference materials as similar as possible to test samples in respect of the content
of analyte and the matrix composition. The performance of the entire analytical
procedure is then evaluated by determining the concentration of analyte of interest
from the calibration and comparing to the value specified in the certificate completing
by calculation of the recovery. Compliance of the results within the accompanying
uncertainty is a confirmation of the absence of significant interference. If values differ
statistically from the certified value, the appropriate budget of uncertainty should be
taken into account or the corresponding recovery factor should be included into the
model equation.
5.3 Method of Standard Addition
In cases where it is not possible to apply standards so as to match, as closely as possible, the sample matrix, and if it could be expected that the components of the sample
significantly affect the detector response (causing interference), the determination of
an analyte based on the relationship of the calibration set for the references of pure
5 Calibration
5.2 Evaluation of Recovery with Reference Materials
In the case of chemical measurements, pure chemical substances are often used to
calibrate the measurement instrument. They can be used as a solid substance or as a
solution, the latter is commonly named the ‘standard solution.’ Thus, the metrological traceability in chemical measurements can be achieved by linking the respond of
instrument to known amount of pure chemical standards (solids or dissolve, depending on the instrumental technique used for performing measurements), knowing its
atomic and molecular mass in moles, the dimensionless SI unit. In practice, we can
see two main limitations of such an approach: first, it is not possible to acquire completely pure chemicals; second, it is not possible to isolate the analyte completely
from the samples, meaning to separate the analyte from the matrix. In addition, setting the calibration relationship for pure chemical substances (the first step described
in the definition of calibration given in VIM) and then determining it according to
the content of an analyte in a real analytical sample (the second step described in the
definition of calibration given in VIM), in a case when measurement can be interfered
with by matrix component can lead to systematic errors. In this case, it is necessary
to check the influence of the matrix of the sample on the analytical signal, since it
may alter its physical and chemical properties, or cause undesired interference and,
consequently, induce a change of values a and b of the calibration function. The
influence of the matrix can be archived via evaluation of the recovery.
Ideally, the best approach is to correlate the calibration relationship for references
which should be identical in chemical composition and physical form to the test
objects (samples). Establishing the calibration dependence with the use reference
material of identical composition to that of real test samples is in most cases difficult
to implement, due to the lack of the wide range of them. Thus, it is recommended to
use reference materials as similar as possible to test samples in respect of the content
of analyte and the matrix composition. The performance of the entire analytical
procedure is then evaluated by determining the concentration of analyte of interest
from the calibration and comparing to the value specified in the certificate completing
by calculation of the recovery. Compliance of the results within the accompanying
uncertainty is a confirmation of the absence of significant interference. If values differ
statistically from the certified value, the appropriate budget of uncertainty should be
taken into account or the corresponding recovery factor should be included into the
model equation.
5.3 Method of Standard Addition
In cases where it is not possible to apply standards so as to match, as closely as possible, the sample matrix, and if it could be expected that the components of the sample
significantly affect the detector response (causing interference), the determination of
an analyte based on the relationship of the calibration set for the references of pure