Precision half-life measurements using Ge(Li) detectors

Abstract

Three correction methods for counting losses - a constant pulser, a proportional feed-back pulser and a pile-up rejector technique - have been examined in conjunction with a Ge(Li) detector and a slow electronic system. Computer simulations of multi-component, multi-sample half-live data indicate that the rejector method should be preferred over the pulser feed-back method for precision measurements. The latter technique is comparable if strict pulser proportionality to the half-life under investigation is maintained. Maximum accuracy is always obtained if each data sample is corrected for losses before summation to other data. Summing of several data sets before counting loss correction always introduces an error, unless the samples contain only one half-life component. The improper statistical analysis of accumulated data has been found to give rise to errors comparable in size to counting loss correction errors. It is found that a weighted mean anomaly must be considered, distinct from the logarithmic weighting effect described by Robinson 4). This effect is found to shift the apparent half-life a few standard deviations from the true half-life. Precision half-life measurements are reported for 10 C: (19.28 ± 0.02) s, 14 O: (70.43 ± 0.18) s, and 23 Mg: (11.26 ± 0.08) s .