Probing three-flavor neutrino oscillations with the complete MINOS data set

The 68 percent and 90 percent confidence limits for the mixing parameters Δm232 and sin2θ23 resulting from a combined fit to the MINOS data for muon-neutrino disappearance and electron-neutrino appearance. The best fit occurs in the inverted hierarchy and lower octant at a value of (0.41, -2.41×10-3 eV2), as indicated by the star.

For many years neutrinos were assumed to be massless. This assumption stemmed from the incredible smallness of their masses and the remarkable weakness of their interactions. The discovery of neutrino mass was possible due to one of the most powerful techniques in physics: interferometry.

Just like slightly untuned musical instruments creating an acoustic beat, neutrinos of different mass oscillate among different flavors. The beat frequency of these relativistic oscillations depends on the difference (interference) among frequencies of the neutrino states as they propagate, which is governed by the difference of the squares of their masses. However, not only neutrino masses play a role: This is also a potential window to the origin of the matter-antimatter imbalance in the universe. The interference pattern could reveal that leptons do not respect the so-called Charge-Parity (CP) symmetry.

For more than nine years, the MINOS experiment has observed neutrinos and antineutrinos produced by the Fermilab accelerator complex, located 735 km from the MINOS detector in Minnesota, and by cosmic-ray interactions in the Earth’s atmosphere. In particular, we have looked for neutrinos produced in the muon flavor that oscillated into other neutrino flavors. We observe this process through both the disappearance of muon neutrinos and the appearance of electron neutrinos in MINOS. Each of these channels has been studied separately before. Now, for the first time, we present results from a combined analysis of the complete MINOS data set.

The combination of appearance and disappearance allows us to probe aspects of neutrino oscillations that involve all three flavors of neutrinos. Such three-flavor phenomena are the main focus of many upcoming neutrino oscillation experiments and the new MINOS results take us a little closer to those goals. For example, the difference between neutrinos and antineutrinos known as CP violation cannot occur in oscillations between only two neutrino flavors. Furthermore, oscillations studied in a two-neutrino model cannot determine which neutrino is the heaviest or which neutrino has a larger muon-flavor component.

The new results, presented in the graphics above and below, constitute our best measurement of the parameters that determine the three-neutrino oscillation pattern: the mixing angle θ23, the mass difference Δm232 and the CP-violating phase δCP.

João Coelho

The 1-D likelihood profile for δCP, plotted separately for each combination of hierarchy and θ23 octant. The best fit occurs in the inverted hierarchy and lower octant; the worst fit is the normal hierarchy and upper octant and is disfavored at 81 percent confidence level. The dashed horizontal lines indicate the 68 percent (90 percent) single-parameter confidence limits, which disfavor 36 percent (11 percent) of the parameter space defined by the mass hierarchy, octant, and δCP.