| Zugriffsnummer | 12416 |
| Dokumenttyp | Konferenzartikel |
| Sprache | Englisch |
| Titel | Application of a primary method to amount-of-substance measurements in real life samples |
| Autor(in); Institution |
Dube, Gunther; 3.21, Metrologie in der Chemie, PTB-Braunschweig
|
| Quelle/Jahr | International Conference on Metrology: Proceedings:(2000), 122 - 126 |
| Verlag | Jerusalem: NCSL |
| Konferenzangaben | International Conference on Metrology - Trends and applications in calibration and testing laboratories, Jerusalem, 16-18, May, 2000, Israel |
| Freie Schlagworte | Klinische Chemie ; CCQM ; CIPM ; Traceability ; Isotope Dilution Mass Spectrometry ; Sample preparation ; glucose ; cholesterol ; cortisol ; uncertainty of measurement |
| Zusammenfassung | Isotope dilution mass spectrometry (IDMS) is one of the methods considered to be primary by the Consultative Committee of Amount-of-Substance (CCQM) of the CIPM. Application of this method to the analysis of simple mixtures of organic chemistry allows results to be obtained which exhibit an uncertainty of about 10-3. If amount-of-substance measurements are carried out with complex matrices, before the measurement by means of mass spectrometry different steps of sample preparation are necessary, for example extraction, reaction, and chromatography, in order to separate the analyte from the matrix. Isotope effects can occur here, which lead to an increase of the uncertainty of the measurement result. Examples are given which demonstrate the occurence of these effects in clinical chemistry in the determination of glucose, cholesterol and cortisol in human serum. In these cases the combined standard uncertainty amounts to about 10-2. The contribution of sample preparation to the uncertainty of the measurement result is estimated by model calculations as follows: The uncertainty of the result of the amount-of-substance measurement is calculated from from the uncertainties of the input quantities given by the basis equation of IDMS following the Guide to the Expression of Uncertainty in Measurement. In the model calculation the input quantities are divided into 3 groups: uncertainty of the weighing of sample and spike (group1), uncertainty of measurement (group 2), and uncertainty of sample preparation (group 3). An attempt is made to bring the results of model calculations in agreement with the combined uncertainties calculated from the experimental results. This is attained by carrying out a series of calculations, in the course of which the uncertainties of group 1 and 2 are estimated and kept constant whereas the uncertainty of group 3 is varied until the standard uncertainty of the results matches the value of the combined standard uncertainty calculated from the experimental data. |
