INVESTIGATIONS INTO THE ORIGIN OF THE SPURIOUS 17 KEV NEUTRINO SIGNAL OBSERVED IN S-35 BETA-DECAY

An exhaustive study has been made of the beta spectrum of S-35, recorded with a Si(Li) detector. The object was to identify the origin of a distortion in the S-35 beta spectrum some 17 keV below the end point, reported over three years ago and interpreted then as evidence for a 17 keV neutrino. Measurements with different source-detector spacings and with varied collimation have shown that there is a long range curvature in the Kurie plot which is a sensitive function of configuration, but the principal origin of the distortion is energy loss in the S-35 sources. The S-35 sources, prepared by chemical adsorption of (BaSO4)-S-35 On a gold substrate, are clumped and locally thick. Electrons near the end point lose similar to 0.3 keV in the source material and if this is taken into account the spectra are well fitted without any admixture of 17 keV neutrino. The source thickness has been investigated with a proton microprobe and determined from both source tilting and the yield of barium K X-rays; these studies are discussed in detail. The uncertainties in and justification for the form of the electron response function employed are also thoroughly discussed. If there is no systematic error common to the majority of 14 independent sets of S-35 data, the admixture of 17 keV neutrino is < 10(-3) (95% CL). A simple search for a kink at 150 keV in the combined data from all 14 runs yielded a limit of 1.8 x 10(-3) (95% CL). The end point of the S-35 beta spectrum is found to be 167.60 +/- 0.05 keV.