Abstract:
Certified reference materials (CRMs) certified for the mass fraction of different metal species have been monitored for stability over the course of many years and the results are presented in the paper.
Components tested include organotin, organomercury, organolead and organoarsenic species in different matrices like sediments or fish tissue.
The stability monitoring points were set to yearly or every 3 years, based on the quality of the original stability data. Several labs were involved in the stability testing, presenting unique challenges to the comparability of the results. The use of ratios is presented to eliminate the possible between-run bias and lab bias.
Highlights:
- Reference samples (stored at lower temperature than the normal samples) were used for comparison with the normal samples.
- Typically, 4 bottles (2 normal stock, 2 reference stock) with 3 measurements each with independent sample preparation were tested per measurement run.
- To check for stability, comparison of the mean of results with the certified value was done, under consideration of lab uncertainty. If the lab uncertainty had been underestimated, the uncertainty from the certification process was used.
- A visual check for a possible trend over time was done.
- The ratio of means of normal stock and reference stock were compared (expected value = 1) and checked if significantly different.
- Furthermore, regression analysis was used over the test series to approximate a possible degradation. The slope was tested for significant difference from zero. If it is not significant, the uncertainty of the slope was then used as ustab for a shelf life of an additional 24 months on top of the “age” of the material.
- If in the comparison, ustab is negligible (max. 1/3) compared to other uncertainty contributions, this was seen as a positive demonstration of stability and the shelf life could be derived from this approach.
- This paper shows the usefulness of the evaluation via ratios - even if absolute values differ significantly from the certified values.
Citation:
Linsinger, T.P.J., Auclair, G., Raffaelli, B., Lamberty, A., Gawlik, B.M., “Conclusions from 13 Years of Stability Testing of CRMs for Determination of Metal Species”. Trends Anal. Chem (2011) 30:6. 875–886.