KamLAND, solar antineutrinos and their magnetic moment

We investigate the possibility of detecting solar antineutrinos with the KamLAND experiment. These antineutrinos are predicted by spin-flavor oscillations at a significant rate even if this mechanism is not the leading solution to the SNP. The recent evidence from SNO shows that a) the neutrino oscillates, only around 34% of the initial solar neutrinos arrive at the Earth as electron neutrinos and b) the conversion is mainly into active neutrinos, however a non e, mu, tau component is allowed: the fraction of oscillation into non-mu-tau neutrinos is found to be cos^2(alpha) = 0.08^{+0.20}_{-0.40}. This residual flux could include sterile neutrinos and/or the antineutrinos of the active flavors. KamLAND is potentially sensitive to antineutrinos derived from solar ^8 B neutrinos. In case of negative results, we find that KamLAND could put strict limits on the flux of solar antineutrinos, Phi(^8 B) < 1.0 times 10^4 cm^{-2} s^{-1}, more than one order of magnitude smaller than existing limits, and on their appearance probability P < 0.20-0.15% (95% CL) after 1-3 years of operation. Assuming a concrete model for antineutrino production by spin-flavor precession, this upper bound implies an upper limit on the product of the intrinsic neutrino magnetic moment and the value of the solar magnetic field mu B < 10^{-21} MeV (95% CL). For B ~ 10-100 kG, we would have mu < 10^{-11}-10^{-12} mu_B (95% CL). In the opposite case, if spin-flavor precession is indeed at work even at a non-leading rate, the additional flux of antineutrinos could strongly distort the signal spectrum seen at KamLAND at energies above 4-5 MeV and their contribution should properly be taken into account.