Sampling when the increment size approaches the analytical aliquot size - theoretical modifications and consequences
Received:February 07, 2020  Revised:February 07, 2020
View Full Text  View/Add Comment  Download reader
DOI:doi:10.3969/j.issn.2095-1035.2020.04.001
KeyWord:sampling; error variance; theoretical correction
     
AuthorInstitution
PenttiMinkkinen 拉彭兰塔理工大学,芬兰
Minkkinen 拉彭兰塔理工大学,芬兰
Hits: 1467
Download times: 963
Abstract:
      Sampling for chemical analysis is nearly always a multi-stage process where all stages contribute to the total uncertainty of the resultsreported. A sample, when extracted from the target, forgets its past, consequently, mistakes made in earlier stages of this process cannot be corrected at later stages, no matter how carefully and precisely they are executed. Primary sampling is the most important, and generally, its variance by far exceeds the variance of the laboratory measurement. That does not mean that, at the later stages when laboratory and analytical aliquots (or test portions) are prepared, the principles of the theory of sampling (TOS) can be neglected. Modern analytical instruments are designed to handle small samples (from milligrams to a few grams). In this case, if the sample is a mixture containing a low fraction of analyte containing particles, the constitution heterogeneity of the material may be so large that it ruins the whole analysis. Heterogeneity calculations and evaluation of the fundamental sampling error variances of the sample preparation steps are essential in developing fit-for-purpose analytical procedures. At the final steps of the sample preparation the new subsample is a significant part of parentsub-sample and account of that effect has to be taken in estimating the sampling variances. TOS offers tools for handling also these situations. Application of heterogeneity calculations are elucidated with two illustrative examples. In the first example the constitution heterogeneity is estimated for low concentration additives in chicken feed and in the second example the sample preparation is optimised for calibrating an IR instrument for estimating mineral impurities in a wollastonite concentrate. Heterogeneity assessments are important also when particle mixtures are processed and the efficiency of mixing is estimated.
Close