60
5 Calibration
addition, increases by more than 1%. Moreover, the method of standard addition can
be used only when a linear relationship exists between analytical signal y and the
concentration of analyte C.
In the method of single standard addition, the determination of an analyte is usually
carried out by the graphical method (Fig. 5.2). Two measured signals (sample before
and after addition of the standard) are used for plotting the graph of respond against
concentration of analyte added. Then the negative intercept on the x-axis, at y 0,
represents the concentration of the analyte in the sample solution.
The method of standard addition is a useful approach when external calibration
with the set of standards of pure chemicals is not possible because the response
is affected by the sample matrix.
Multiple standard addition method
In some cases, improved accuracy of the result can be achieved using the method of
multiple standard additions. In this case, we make a number of additions of standard
(more than one) to the test sample solution containing the analyte, and then measure
the resultant increase in the response of detector after each addition. The results can
best be presented graphically. By extrapolating the straight line back to y 0, we
can obtain the measure of the value of concentration of analyte in the test sample
before any addition.
5.4 The Advantages of the Standard Addition Method
– Using the standard addition method is advantageous when there is no information
about the matrix, which could cause interferences;
– The standard addition method can be recommended for an analytical task with
many samples of different composition of the matrix. It would then be uneconomical to prepare many calibration graphs that could mimic the composition of the
different samples.
5.5 Limitations of the Standard Addition Method
– The standard addition method can be used only if the analytical signal is directly
proportional to the concentration C or its function, for example, log C. The calibration relationship is not always linear, so extrapolation may result in overestimation
or underestimation of the results.
Précédent

- 76/201

Suivant