First Evidence of $pep$ Solar Neutrinos by Direct Detection in Borexino

We observed, for the first time, solar neutrinos in the 1.0–1.5 MeV energy range. We determined the rate of $pep$ solar neutrino interactions in Borexino to be $3.1\pm {0.6}_{\mathrm{stat}}\pm {0.3}_{\mathrm{syst}}\mathrm{counts}/(\mathrm{day}·100\mathrm{ton})$. Assuming the $pep$ neutrino flux predicted by the standard solar model, we obtained a constraint on the CNO solar neutrino interaction rate of $<7.9\mathrm{counts}/(\mathrm{day}·100\mathrm{ton})$ ($95\%$ C.L.). The absence of the solar neutrino signal is disfavored at $99.97\%$ C.L., while the absence of the $pep$ signal is disfavored at $98\%$ C.L. The necessary sensitivity was achieved by adopting data analysis techniques for the rejection of cosmogenic ${}^{11}\mathrm{C}$, the dominant background in the 1–2 MeV region. Assuming the Mikheyev-Smirnov-Wolfenstein large mixing angle solution to solar neutrino oscillations, these values correspond to solar neutrino fluxes of $(1.6\pm 0.3)\times {10}^8{\mathrm{cm}}^{-2}{\mathrm{s}}^{-1}$ and $<7.7\times {10}^8{\mathrm{cm}}^{-2}{\mathrm{s}}^{-1}$ ($95\%$ C.L.), respectively, in agreement with both the high and low metallicity standard solar models. These results represent the first direct evidence of the $pep$ neutrino signal and the strongest constraint of the CNO solar neutrino flux to date.

Figure 1

Top: energy spectra of the events in the FV before and after the TFC veto is applied. The solid and dashed blue lines show the data and estimated ${}^{11}\mathrm{C}$ rate before any veto is applied. The solid black line shows the data after the procedure, in which the ${}^{11}\mathrm{C}$ contribution (dashed black line) has been greatly suppressed. The next largest background, ${}^{210}\mathrm{Bi}$, and the $e^{-}$ recoil spectra of the best estimate of the $pep\mathrm{\text{−}}\nu$ rate and of the upper limit of the CNO-$\nu$ rate are shown for reference. Rate values in the legend are integrated over all energies and are quoted in units of $\mathrm{counts}/(\mathrm{day}·100\mathrm{metric}\mathrm{ton})$. Bottom: residual energy spectrum after best-fit rates of all considered backgrounds are subtracted. The $e^{-}$ recoil spectrum from $pep\mathrm{\text{−}}\nu$ at the best-fit rate is shown for comparison.

Figure 2

Experimental distribution of the pulse-shape parameter (black data points). The best-fit distribution (dashed black line) and the corresponding $e^{-}$ (solid red line) and $e^{+}$ (solid blue line) contributions are also shown.

Figure 3

Experimental distribution of the radial coordinate of the reconstructed position within the FV (black data points). The best-fit distribution (dashed black line) and the corresponding contributions from bulk events (solid red line) and external $\gamma$ rays (solid blue line) are also shown.

Figure 4

$\Delta {\chi }^2$ profile obtained from likelihood ratio tests between fit results where the $pep$ and CNO neutrino interaction rates are fixed to particular values (other species are left free) and the best-fit result.

Figure 5

Electron neutrino survival probability as a function of energy. The red line corresponds to the measurement presented in this Letter. The $pp$ and ${}^7\mathrm{Be}$ measurements of $P_{ee}$ given in [5] are also shown. The ${}^8\mathrm{B}$ measurements of $P_{ee}$ were obtained from [3, 4, 25], as indicated in the legend. The MSW-LMA prediction band is the $1\sigma$ range of the mixing parameters given in [22].