Uncertainty calculations for amount of chemical substance measurements performed by means of isotope dilution mass spectrometry as part of the PERM project

Citation
A. Dobney et al., Uncertainty calculations for amount of chemical substance measurements performed by means of isotope dilution mass spectrometry as part of the PERM project, ANALYT CHIM, 420(1), 2000, pp. 89-94
Citations number
13
Categorie Soggetti
Spectroscopy /Instrumentation/Analytical Sciences
Journal title
ANALYTICA CHIMICA ACTA
ISSN journal
00032670 → ACNP
Volume
420
Issue
1
Year of publication
2000
Pages
89 - 94
Database
ISI
SICI code
0003-2670(20000907)420:1<89:UCFAOC>2.0.ZU;2-7
Abstract
DSM Research was a partner in the certification round of a project, organiz ed under the auspices of the European Commission's Fourth Framework Standar ds Measurement and Testing Programme, to produce two polyethylene reference materials. The project, entitled Polymer Elemental Reference Material (PER M SMT4-CT95 3024) required the determination of the concentrations of Cr, C d, Hg and ph in two polyethylene materials which were identified as PERM T3 high and PERM T3 low. Analysis was accomplished by means of an in-house de veloped 'on-line' isotope dilution mass spectrometry (IDMS) method using qu adrupole ICPMS. This paper focuses on calculating the uncertainty of these IDMS measurements. Since all analytical measurement results have an uncerta inty associated with them, it is considered best practice, within metrologi cal circles, to state an uncertainty, as an integral part of any measuremen t result. We used a Monte Carlo simulation to estimate the uncertainties as sociated with the measured concentrations of Cr, Cd, Hg and Pb in the PERM T3 (high and low) materials. By participating in an interlaboratory exercis e we have been able to compare our results with those of other laboratories that used both classical IDMS and other techniques. Our results agreed wit h the overall mean of all participating laboratories, within stated uncerta inties. Furthermore our uncertainties derived from Monte Carlo simulation w ere of the same magnitude as those obtained for the classical IDMS approach . (C) 2000 Elsevier Science B.V. All rights reserved.