Tabulated emission probabilities [e.g. IAEA (1986)] for the s-particle decay of the nuclide 239pu appear to require revision in the light of measurements reported here. Although the sum of the probabilities of the two most intense lines, s 0 and st3, has been listed as ca 88% since the work of Baranov et al. (1963). However, the individual probabilities measured for these lines were 70.73 and 17.56%, and these values differ significantly from those appearing in reference tables [e.g. IAEA (1986)]. In earlier measurements, the incomplete resolution of two peaks only ~ 13 keV apart, or the effects of conversion electrons in producting energy-summing effects may have given rise to systematic errors depending on the type of apparatus used.
Measurements have been carried out under low geometry with a very thin source of purified 239pu and, for comparison, another of 24°pu. Emission probabilities were obtained from the energy spectra using a computer programme to find a best-fit to the peaks with analytical expressions. The programme is based on an iterative Simplex procedure to find a minimum of Chi-squared. A careful evaluation of the variances provides estimates of the relative uncertainty (at the 2a level) of each emission probability which are, in all cases, <2% of the measured values.