Neutrino Astrophysics: A Research Briefing (1995)

impossible to explain simultaneously the results of the chlorine and the Kamiokande experiments by adjusting input solar data [20]. Just the ⁸B neutrinos that are observed in the Kamiokande experiment imply a minimum expected counting rate in the chlorine experiment of 3.21 ± 0.46 SNU compared with the observed rate of 2.55 ± 0.25 SNU. This comparison does not include the additional contributions to the chlorine rate from the much more reliably calculated lower-energy neutrinos (⁷Be and pep neutrinos; see Figure 1) to which the chlorine experiment is sensitive, but Kamiokande is not. The comparison does suggest that the lower-energy neutrinos might be preferentially depleted by some new weak interaction process (which could be the MSW effect discussed below).

The neutrinos produced by the most basic fusion reaction in the sun, the fusion of two protons, the so-called p-p reaction, are too low in energy to be detected by either the chlorine or the Kamiokande experiment. Since the p-p neutrinos are predicted to be the most numerous neutrinos emitted by the sun and since their flux is more reliably predicted than any other neutrino flux [2], the measurement of the terrestrial flux of p-p neutrinos has long been seen [21,22] as a method for testing whether something happens to the neutrinos once they are produced in the interior of the sun.

Two radiochemical experiments that can detect p-p neutrinos are currently in operation; they use ⁷¹Ga as their target. The detection reaction is

ν_e + ⁷¹Ga → ⁷¹Ge + e⁻, (3)

which has an energy threshold of 0.2 MeV. The gallium experiments operate similarly to the chlorine experiment, extracting and counting radioactive atoms of ⁷¹Ge. The standard model prediction is that after a month of operation about 16 atoms of ⁷¹Ge will be present in 30 tons of gallium.

One of the gallium experiments, GALLEX [6,7], is a European-U.S.-Israeli collaboration operating in the Gran Sasso tunnel near Rome, Italy. GALLEX uses 30 tons of gallium in a GaCl₃-HCl solution. The other experiment, SAGE [8,9] (the Soviet-American Gallium Experiment), is operating in a subterranean laboratory under Mount Andyrchi of the Caucasus Mountains in the southern region of Russia. SAGE currently uses 60 tons of molten gallium metal.

The fact that the detectors use gallium in two very different chemical environments, the GaCl₃-HCl solution (GALLEX) and the metallic form (SAGE), provides a consistency check on the results. The current GALLEX results are 79 ± 12 SNU and the current SAGE results are 74 ± 14 SNU. These results also are significantly below the standard model prediction, which is 132 ± 7 SNU for gallium detectors. Both experiments will be checked by measuring the event rate produced with a strong source of radioactive chromium.

Thus as time has passed and more experiments have been performed, the solar neutrino problem has not gone away. Indeed, it appears that, independent of detailed solar model calculations, there is no combination of the expected neutrino fluxes that fits the available data.

The MSW effect [15,16] (oscillations between different neutrino types enhanced by neutrino interactions with matter) can describe consistently the results of all four of the experiments (chlorine, Kamiokande, GALLEX, and SAGE) if either nμ or nτ has a mass of about 0.003 eV. Only future experiments can determine if this is indeed the correct explanation.

As interest in the solar neutrino problem intensifies, experiments in addition to the four described above are in various stages of development. Broadly speaking, solar neutrino detectors fall into two classes, “active” (like Kamiokande) and “radiochemical” (like chlorine and gallium). Active detectors can provide information about the arrival time, direction, and energy of individual neutrinos, and in some cases the neutrino type (e.g., electron or muon), but are complex and often faced with persistent backgrounds. Radiochemical detectors are

sensitive only to electron-type neutrinos (i.e., they specify uniquely the neutrino type) and provide a single number for the rate of detection of all neutrinos above an energy threshold, a number that is averaged over a period of time comparable to the mean life (typically weeks or months) of the radioisotope that is produced.

This section describes the three active and one radiochemical solar neutrino detectors that are under construction. The active detectors will have several thousand events per year, sufficient to test accurately the constancy of the rates predicted by the standard solar model. The expected high event rates will permit the observation of the expected seasonal variation of the solar neutrino fluxes due to the orbital eccentricity of the earth (~7 percent effect, peak to peak).

Deuterium (“heavy” hydrogen, with a neutron and a proton in its nucleus) is an excellent target for neutrinos. Two neutrino interactions can occur [23]:

ν_e + ²H → p + p + e⁻; (4a)

ν_x + ²H → p + n + ν_x. (4b)

Reaction (4a) can be induced only by electron-type neutrinos, whereas reaction (4b) can be caused by neutrinos of different types. If more neutrinos are detected via reaction (4b) than by reaction (4a), that would be direct evidence that some electron-type neutrinos have oscillated into neutrinos of some other type.

Deuterium is a rare isotope of hydrogen, but the Canadian nuclear power industry requires vast quantities of heavy water, nearly pure D₂O. Also, one of the deepest mines in the Western Hemisphere (a nickel mine belonging to INCO Limited) is located near Sudbury, Ontario. A large cavern has been excavated 2070 m underground to hold a detector consisting of 1000 tonnes of heavy water. The heavy water, contained in a transparent acrylic vessel, is to be surrounded by a shield of 7000 tonnes of ultrapure light water. The flashes of Cerenkov light will be recorded by more than 9000 photomultipliers, specially constructed of materials selected for low radioactivity. When the Sudbury Neutrino Observatory begins operation with heavy water in 1996, solar neutrinos are expected to be detected at the rate of more than 10 counts per day via reactions (4a) and (4b). In addition to the unique reactions of neutrinos on deuterium, SNO will observe the same neutrino-electron scattering process that is detected by Kamiokande. The SNO experiment is a collaboration between experimenters from Canada, the United States, and Great Britain.

If electron-type neutrinos oscillate into one of the other known neutrino types as they travel from the interior of the sun to the terrestrial detector, this will be revealed by the comparison of the rates in the two deuterium reactions, and, if the oscillation parameters are favorable, also by its distinctive effect on the shape of the observed electron energy spectrum in reaction (4a). These potential signatures of new physics are independent of solar models and solar physics.

The Superkamiokande experiment [24] is a natural extension of the current Kamiokande experiment with an upgraded performance as well as a much better sensitivity. A large detector of approximately 50,000 tonnes of ultrapure ordinary water is under construction in Japan. The principles of this detector are similar to those of the currently operating Kamiokande detector (see the section above on the Kamiokande detector), although the volume of water used will be much increased and the threshold for neutrino detection will be lowered (perhaps to 5 MeV). By comparison with the Kamiokande experiment, this immense detector will, beginning in 1996, provide a 30-fold increase in the observed rate of neutrino-electron scattering events [24]. The change in the shape of the neutrino energy spectrum predicted by the MSW effect may be observable in this experiment via the spectrum of recoil energies of the scattered electrons. Physicists from several U.S. institutions are collaborating with Japanese colleagues in the Super-kamiokande experiment.

BOREXINO [25] will be the first real-time detector capable of observing the intense flux of low-energy neutrinos produced by electron capture on beryllium nuclei (see the 0.86-MeV ⁷Be