Solar neutrino physics with Borexino

Summary. — We report the direct measurements of 7 Be and 8 B solar neutrino signal rates performed with the Borexino detector at the Laboratori Nazionali del Gran Sasso.


-Introduction
Borexino [1,2] is a real-time experiment for solar low energy neutrino spectroscopy, operating since May 2007 at the underground Gran Sasso National Laboratories.Solar neutrinos are detected by means of their elastic scattering off electrons in a liquid-scintillator target: 278 tons of pseudocumene doped with 1.5 g/l of PPO.The scintillator is housed in a thin (125 μm) nylon vessel and is shielded by a buffer of 1000 tons of pseudocumene.2212 photomultiplier tubes, mounted on a stainless steel sphere (SSS), detect the scintillation light.Finally, the SSS is installed inside a 3000 m 3 water tank which provides the necessary shielding against rock-induced external backgrounds and is used as a Cerenkov detector to veto the residual muons.The Borexino purification strategy relies on filtration at the level of 0.05 μm, multistage distillation and high-purity nitrogen sparging.Borexino obtained an excellent level of radiopurity in the innermost scintillator target: 238 U contamination is at (1.6 ± 0.1) × 10 −17 g/g and the 232 Th contamination at (6.8 ± 1.5) × 10 −18 g/g.

-7 Be rate measurement
The basic signature for the mono-energetic 0.862 MeV 7 Be neutrinos is the Comptonlike edge of the recoil electrons at 665 keV.The dominant background contributions are muons, external background and fast delayed events ( 214 Bi-214 Po, 212 Bi-212 Po), rejected exploiting the Cerenkov detector, applying a radial cut of 3 m (the innermost 100 tons of scintillator), and thanks to the time and spatial correlations, respectively.
The measured spectrum in 192 days of data taking is shown in fig. 1.The peak at ∼ 400 keV is due to a 210 Po contamination still present in the liquid scintillator after purification and filling. 210Po can be statistically subtracted by use of the pulse shape discrimination.The high-energy component of the spectrum is dominated by the cosmogenic 11 C, produced underground by muons interacting with 12 C in the liquid scintillator.
Systematic uncertainties come mainly from the total scintillator mass (0.2%), the fiducial mass definition (6%) and the detector response function (6%).Taking into account systematic errors, our best value for the interaction rate of the 0.862 MeV 7 Be solar neutrinos is 49 ± 3 stat ± 4 syst c/d/100 ton.

-8 B rate measurement
This measurement is based on 245.9 live days of data-taking in the target mass of 100 tons.
The 2.8 MeV energy threshold is imposed by the 2.6 MeV γ's from the β-decay of 208 Tl, due to radioactive contamination mainly in the photomultiplier tubes.The data selection, like in the 7 Be measurement, requires rejection of muons (and secondaries), external background and fast coincidences.Short-lived (τ < 2 s) cosmogenic isotopes, 12 B, 8 B and 8 Li, produced in the scintillator by the residual muon flux, are removed The black line represents all events.The light blue-and blue-filled spectra are the samples after the muon cut and fiducial volume cut, respectively.The dark blue-filled spectrum is the final set of data after all cuts and before statistical subtraction of 208 Tl (white line).
by vetoing the detector for 5 s after each muon crossing the scintillator (dead time of 23.4%, live-time reduced to 187.9 d).Among long-lived (τ > 2 s) cosmogenic isotopes, the only contribution is given by 10 C decays, tagged and rejected by the triple coincidence with the parent muon and neutron capture on proton.The effect of each step of the analysis sequence described above is shown in fig. 2. The intrinsic 208 Tl contribution, from the 232 Th chain, is measured, by analyzing the delayed coincidences of its branching competitor, 212 Bi-212 Po, and statistically subtracted.The 8 B interaction rate has been measured in 0.26 ± 0.04 stat ± 0.02 syst c/d/100 tons.above 2.8 MeV [3,4].Both the measured rate and spectrum (fig.3) are in agreement with the rate (0.26 ± 0.04 c/d/100 tons) and the spectrum predicted by the Standard Solar Model [5], including the MSW-LMA solution (Δm 2 = 7.69×10 −5 eV 2 , tan 2 θ = 0.45 [6]).

-Conclusions
The Borexino 7 Be measured rate can be combined with the other solar neutrino measurements to constrain the flux normalization constants of the other solar neutrino fluxes.This leads to the best determination of the pp solar neutrinos flux, obtained with the assumption of the luminosity constraint: f pp = 1.005 +0.008 −0.020 , where f pp is the ratio between the measured and predicted pp neutrino fluxes.With the same technique, Borexino obtained the best limit on the CNO flux: f CNO < 6.27 (90% CL).
The low-energy solar neutrino spectrum is sensitive to the possible presence of a nonnull magnetic moment.We exploited this feature to determine the best upper limit to the neutrino magnetic moment (5.4 × 10 −11 μB, 90% CL) [2].
Electron neutrino survival probabilities have been estimated for both 7 Be neutrinos (0.56 ± 0.10 at 0.862 MeV) and 8 B (0.35 ± 0.10 at the mean energy of 8.6 MeV).The agreement between the Borexino results and the LMA-MSW oscillation prediction is shown in fig. 4 [3,4].

Fig. 2 .
Fig.2.-(Colour on-line) Energy spectra of candidate events after application of several cuts.The black line represents all events.The light blue-and blue-filled spectra are the samples after the muon cut and fiducial volume cut, respectively.The dark blue-filled spectrum is the final set of data after all cuts and before statistical subtraction of 208 Tl (white line).

Fig. 3 .Fig. 4 .
Fig.3.-Energy spectrum of the events surviving all cuts and after statistical 208 Tl subtraction.The expected electron recoil spectrum due to oscillated (not oscillated) 8 B ν interaction, as determined from the BS07(GS98) solar model, is represented by the solid (dashed) line.