IDMS with calibration by exact or approximate signal matching methods offers a number of advantages for chemical metrology applications where measurements of low uncertainty are required for analytes at low concentrations in complex matrices. These methods have been applied extensively but are unsuitable for some important applications because in many cases, the concentration of the spike isotope in each measured blend of a sample or standard should be as close as possible to the concentration of the natural analyte isotope.
In this paper we discuss recent work investigating the extent to which it is possible to move away from this ideal 1:1 blend isotope ratio whilst retaining the key benefits of the approximate matching technique. This has been used for the first time for applications where the sample has a very high analyte concentration, which would require unacceptably high amounts of spike with the existing method.
The methodology has been validated for sulfur using a well-characterised candidate matrix reference material - NIST SRM1624d. This is a diesel fuel sample that contains approximately 4000 �g g-1 sulfur. Results were obtained using a ratio for 32S/34S of 14, achieving accuracy and uncertainty comparable with the original method using a blend isotope ratio of 1.
For further information about this article contact the authors via the VAM helpdesk or visit the JAAS website.