Constraints on decaying sterile neutrinos from solar antineutrinos

Solar neutrino experiments are highly sensitive to sources of $\nu \to \overline{\nu }$ conversions in the ${}^8\mathrm{B}$ neutrino flux. In this work, we adapt these searches to nonminimal sterile neutrino models recently proposed to explain the LSND, MiniBooNE, and reactor anomalies. The production of such sterile neutrinos in the Sun, followed by the decay chain ${\nu }_4\to \nu \phi \to \nu \nu \overline{\nu }$ with a new scalar $\phi$, results in upper limits for the neutrino mixing $|U_{e4}{|}^2$ at the per-mille level. We conclude that a simultaneous explanation of all anomalies is in tension with KamLAND, Super-Kamiokande, and Borexino constraints on the flux of solar antineutrinos. We then present other minimal models that violate parity or lepton number, and discuss the applicability of our constraints in each case. Future improvements can be expected from existing Borexino data, as well as from future searches at Super-Kamiokande with added gadolinium.

Figure 1

The solar neutrino energy spectrum from ${}^8\mathrm{B}$ (shaded orange) and the resulting antineutrino spectrum from the decays of ${\nu }_4$ (shaded grey). We also show the inverse beta decay (IBD) and neutrino-electron scattering cross sections on an overlaid axis.

Figure 2

The experimental limits on solar $\overline{{\nu }_e}$ at $90\%$ C.L. as a function of the neutrino energy. The shaded regions are excluded by Borexino (blue) [25], KamLAND [24] (pink), KamLAND [27] (grey), and SuperK-IV [26] (yellow). The different new physics predictions are also shown as solid curves assuming $|U_{\tau 4}|=0$ and $|U_{\mu 4}{|}^2=10^{-3}$.

Figure 3

The inverse beta decay spectrum at Borexino (top row), KamLAND 2011 (middle row), and SK-IV 2020 (bottom row), all as stacked histograms. The hatched histograms show the background estimations by the collaboration, and the filled histogram (gray) shows our prediction of visible neutrino decays. All plots assume $|U_{{\mu }_4}|=|U_{\tau 4}|=0$.

Figure 4

Limits from solar antineutrino searches on the active-heavy neutrino mixing at a 99% C.L. The regions required to explain the short-baseline anomalies in a Dirac sterile neutrino decay model are also shown (99% C.L.). For reactors, a preferred interval in $|U_{e4}{|}^2$ is shown and is independent of $|U_{\mu 4}{|}^2$. On the left, we show the $m_4{\mathrm{\Gamma }}_4=1{\mathrm{eV}}^2$ case, and on the right, $m_4{\mathrm{\Gamma }}_4=10{\mathrm{eV}}^2$, where the ${\nu }_4$ is shorter lived. Our bounds are the same for the two cases.

Figure 5

Same as Fig. 4, but for $m_{\phi }/m_4=0.5$. In this regime, LSND and MiniBooNE are harder to combine due to the softer ${\overline{\nu }}_e$ spectrum obtained via electron mixing. We show only the combination of MiniBooNE with other datasets, excluding LSND, as reported in Ref. [12]. The solar antineutrino spectrum is also softer, but it can effectively constrain the preferred combined region at a 99% C.L.

Figure 6

Limits obtained in this work in the plane of active-heavy mixing versus the ratio $R=m_{\phi }/m_4$, at a 99% C.L. On the left, we show the case of $m_4=300\mathrm{eV}$, but we note that our curves are mostly insensitive to the overall scale of $m_4$ and $m_{\phi }$ in the region of interest for short-baseline anomalies.

Figure 7

Full flavor transition probability as well as the relevant mixing matrix element in matter, all averaged over the ${}^8\mathrm{B}$ neutrino production region in the Sun and properly normalized. From left to right, we increase the muon-heavy mixing, showing neutrino transitions on the top row and antineutrino transitions on the bottom.