Detection limits should be a thing of the past in gamma-ray spectrometry in general as well as in neutron activation analysis

Abstract

In gamma-ray spectrometry with high-resolution detectors, full-energy peaks are often to be detected by a peak-search algorithm, with a threshold for detection. Detection limits can be derived from this. Detection limits are often computed along with measured activities or concentrations. When an analyte is not detected, the detection limit remains as the only available information. This leads to inhomogeneous datasets that are difficult or impossible to process correctly without introducing artefacts or biases. Here, it is proposed to determine peak areas at predetermined energies. An unbiased result with its uncertainty always results, obviating the “detection limit” concept.