The primary source of high energy neutrinos from the sun is the beta decay of ⁸B, which is produced by nucleosynthesis deep in the solar interior. The exact spectral shape of the neutrinos from ⁸B is of current interest because of recent results from the SuperKamiokande and Sudbury Neutrino Observatory collaborations. These results may be explained by neutrino oscillations which are accompanied by a distortion in the observed neutrino spectrum.

The beta decay of ⁸B is allowed and the decay proceeds primarily to the unbound first excited state of ⁸Be at 2.9 MeV, which decays totwo alpha particles. The ⁸B beta and neutrino spectra deviate from the allowed shape due to the broad and complicated energy profile of the final state.

In the past, the neutrino spectrum has been calculated form measurements of the delayed alpha spectrum and in one case from constraints imposed form measurements of the beta spectrum¹. In 1996, Bahcall et al.² have reviewed all the experimental data currently available attempting to predict the ⁸B neutrino spectrum. It is pointed out in that work that the five existing measurements of the alpha spectrum are in poor agreement. While the basic shape of the spectra are similar there are discrepancies of about 80 keV in the energy scales, which made it difficult to assess the overall systematic uncertainty. The experimental data became even more discrepant following the recent measurement of the alpha spectrum by Ortiz et al.³, whose data indicates an alpha spectrum peaked at lower energies (by nearly 100 keV) than the previous measurements, and hence predicts an energetically harder ⁸B neutrino spectrum.

We are currently working on an experiment to measure the beta-delyaed alpha spectrum from ⁸B using the novel technique of implanting ⁸B in a silicon detector. Previous experiments used ⁸B implanted externally in thin metallic foils and the uncertainties of the foil and source thicknesses are suspected to be the source of the discrepancies. The technique involves the production of a ⁸B beam using the ATLAS accelerator at Argonne National Lab to bombard a ³He gas cell with a ⁶Li beam, using the ³He(⁶Li,n)⁸B reaction. Boron nuclei with the proper energy for implantation are selected by the ATLAS Spectrograph and deposited into the center of a 91 micrometer Silicon detector located in the Spectrograph focal plane. This setup eliminates systematic errors associated with alpha energy loss in catcher foils and detector dead layers, and minimizes the error due to beta energy deposition by requiring a coincidence hit with a 0.8 mm plastic scintillator located 4 cm away from the silicon detector, and sharing its central axis. We use ²⁰Na for calibration, which decays by beta-delayed alpha emission about 20% of the time and produces alpha lines with energies above and below the peak of the ⁸B spectrum. The ²⁰Na was produced using the ³He(¹⁹F,2n)²⁰Na reaction, and implanted in the detector iun a similar manner. External sources are also used for calibration. 

In 2001, three runs at ATLAS took place. The first was in May and ⁸B ions were successfully implanted into the Si detector and a preliminary spectrum taken. The second, in October, was a test run for ²⁰Na production. We attained a beam of 2-3 ions per second through the magnet, sufficent to perform the calibration. 

The final run of 2001 was in December and implantation of both ⁸B and ²⁰Na was achieved, and the coincidence scintillator was utilized for the first time. We observed 50,000 coincident events in the alpha spectrum, and an anlaysis of the data is being carried out at present. A final measurement, which will eliminate small problems with the experimental setup and provide adequate statistics, is being planned for summer 2002.

Figure 1. Data from the most recent run. The blue histogram is the 8B alpha spectrum, while the red is the 20Na calibration. Both are in beta-coincidence.

[1] J. Napolitano, et al., Phys Rev C36, 298 (1987)
[2] J.N Bahcall, et al., Phys Rev C58, 411 (1996)
[3] C. Ortiz, et al., Phys Rev Lett 85, 2909 (2000)