Optimization of information properties of NAA with respect to information content and profitability of results

Abstract

Information properties of analytical results together with other important parameters especially economic ones can be used for the optimization of analytical procedures. Therefore, we have proposed a computational technique for the optimization of multielement neutron activation analysis (NAA) based on the information content and profitability. The optimization starts with the prediction of the γ-ray spectra to be expected during analysis under given experimental conditions (sample size, irradiation, decay and counting times etc.) and with the calculation of detection and determination limits. In the next step, the information contents for the determination of particular elements and for the simultaneous determination of element groups are computed. The information content depends or is closely connected with such properties of the method as selectivity, sensitivity, precision, accuracy and, as in the other cases of trace analysis, also with the detection limit. Then, the information profitability (IP) taking into account the information content and relevance (appreciation of specific information according to its contribution to the solution of a given problem) together with economic aspects can be calculated. This function can be used for the optimization of a particular NAA procedure, for the mutual comparison of different variants of NAA and also for the comparison with other analytical methods. The use of information profitability for the optimization of NAA is shown on a practical example of the INAA analysis of urban particulate matter SRM 1648 produced by NBS (USA).