In principal three different calibration methods exist: an external and internal
calibration as well as standard addition calibration. For an external calibration the
measuring system gets calibrated by external standard solutions with different
concentrations. The concentrations used should cover the expected unknown concentrations. Usually 4–5 calibration points are used, however the more the calibration points the better the calibration. With the resulting data set a linear correlation
can provide a calibration function as illustrated in Fig. 5.21. The quality of the
calibration is represented by the correlation coefficient, that should be near the value
1. Based on the calibration function, the concentration of the analyte in any further
sample solution, normally extracts, can now be determined.
The second approach uses an internal standard for calibration. The internal
standard should be a chemically highly similar substance such as labelled compounds (e.g.
13 C-labelled PCBs for PCB analyses) or unusual isomers or derivative
not expected in the extract (e.g. theobromine for caffeine analysis). As an important
prerequisite, these internal standard substances need to be detectable in parallel to the
target analytes without any interference. A defined amount of internal standard is
added to the sample solution or extract containing the target analyte with unknown
concentration. After measurement, the peak area of the internal standard with known
concentration is used as a one-point calibration to convert the peak area of the target
analyte to its concentration. This process is exemplified in Fig. 5.22 for a PCB
analyses based on GC/MS measurements.
The last calibration approach uses the analyte itself as standard but not as external
but as added standard. This approach needs several analyses. First, the untreated
extract is measured. Thereafter, a defined amount of analyte is added (usually by
adding a defined volume of standard solution with known concentration) and the
analysis is executed once again. This step of adding and measuring is repeated
h
h
b 50 %
b 85 %
b 15 %
Approximation for
symmetric peaks:
Approximation for asymmetric
peaks(after Condale-Bosch):
Fig. 5.20 Two approximations for calculating the areas of full symmetric or slightly asymmetric
peaks
5.2 Quantitation
123
calibration as well as standard addition calibration. For an external calibration the
measuring system gets calibrated by external standard solutions with different
concentrations. The concentrations used should cover the expected unknown concentrations. Usually 4–5 calibration points are used, however the more the calibration points the better the calibration. With the resulting data set a linear correlation
can provide a calibration function as illustrated in Fig. 5.21. The quality of the
calibration is represented by the correlation coefficient, that should be near the value
1. Based on the calibration function, the concentration of the analyte in any further
sample solution, normally extracts, can now be determined.
The second approach uses an internal standard for calibration. The internal
standard should be a chemically highly similar substance such as labelled compounds (e.g.
13 C-labelled PCBs for PCB analyses) or unusual isomers or derivative
not expected in the extract (e.g. theobromine for caffeine analysis). As an important
prerequisite, these internal standard substances need to be detectable in parallel to the
target analytes without any interference. A defined amount of internal standard is
added to the sample solution or extract containing the target analyte with unknown
concentration. After measurement, the peak area of the internal standard with known
concentration is used as a one-point calibration to convert the peak area of the target
analyte to its concentration. This process is exemplified in Fig. 5.22 for a PCB
analyses based on GC/MS measurements.
The last calibration approach uses the analyte itself as standard but not as external
but as added standard. This approach needs several analyses. First, the untreated
extract is measured. Thereafter, a defined amount of analyte is added (usually by
adding a defined volume of standard solution with known concentration) and the
analysis is executed once again. This step of adding and measuring is repeated
h
h
b 50 %
b 85 %
b 15 %
Approximation for
symmetric peaks:
Approximation for asymmetric
peaks(after Condale-Bosch):
Fig. 5.20 Two approximations for calculating the areas of full symmetric or slightly asymmetric
peaks
5.2 Quantitation
123