Two portals to GeV sterile neutrinos : dipole versus mixing

Abstract

Massive sterile neutrinos, also known as heavy neutral leptons, can have a mixing with active neutrinos, $θ$, as well as a dipole coupling to the photon, $d$. We study the interplay between these two portals, considering the production from meson decays of sterile neutrinos with mass $0.1$ GeV $\lesssim M_N \lesssim 10$ GeV, at beam-dump facilities such as NA62 and SHiP, and at the FASER2 experiment. These sterile neutrinos can be long-lived and decay into a photon in a distant detector, via the dipole operator. We find that all these experiments will be sensitive to values of $d$ which are presently unconstrained. The experimental reach varies strongly with the mass $M_N$ and the mixing $θ$, and one observes specific correlations with the flavour of active neutrinos. The SHiP experiment will mark a jump in sensitivity: (i) it will probe a sterile dipole as small as $d\sim 10^{-8}$ GeV$^{-1}$, thus testing new physics well above the electroweak scale; (ii) it may detect the active-sterile dipole to the level predicted by electroweak loops, if $θ$ is close to the present bound.

Figures (6)

• Figure 1: Regions in which $N_2$ production is dominated by the dipole, $N_d > N_\theta$ (above the blue lines), and where $N_2$ decays are dominated by the dipole, $\Gamma_d > \Gamma_\theta$ (above the orange lines), for three different values of $M_1$. The light blue/orange colour shading corresponds to the case $M_1=1$ GeV. In the wine (white) region, the dipole (the mixing) dominates both $N_2$ production and decays. We focused on the SHiP case, fixing a mass splitting $\delta=0~(0.1)$ in the upper (lower) panels, and non-zero mixing with a single flavour, $\theta_{i\tau} \neq 0$ ($\theta_{i\mu}\neq 0$) in the left (right) panels. The lower (upper) horizontal dashed gray line corresponds to a dipole generate by new physics at scale $m_*= 1$ TeV in a weak (strong) coupling regime, $g_*=1$ ($g_*=4\pi$). The vertical dashed gray line is the maximal allowed mixing for $M_1=1$ GeV (in the mixing-only scenario).
• Figure 2: Same as in Fig. \ref{['fig:dipole_dominance_delta']}, but for different choices of parameters. In the upper panels, we focus on the SHiP experiment and fix $M_1 = 1$ GeV, while considering three different values of $\delta$ (the shading corresponds to $\delta=0$). In the lower panels, we fix $M_1 = 1$ GeV and $\delta = 0.1$, and compare the SHiP and FASER2 experiments (the shading corresponds to SHiP). The curves for NA62 (not shown) basically coincide with those for SHiP.
• Figure 3: Sensitivity to the photon signal of the future experiments FASER2, NA62 and SHiP (coloured lines) and regions excluded by the past experiments CHARM-II and BEBC (coloured regions), fixing $\delta = 0.1$ and either no mixing (upper panel) or mixing with the $\tau$ flavour (middle panel). We also show the regions excluded by limits from supernovæ (SN) and Big Bang Nucleosynthesis (BBN). The lower (upper) horizontal dashed gray line corresponds to a dipole generate by new physics at scale $m_*= 1$ TeV in a weak (strong) coupling regime, $g_*=1$ ($g_*=4\pi$). The lower panel shows the number of events generated by the EW dipole (limit $d\to 0$), for the same parameters as in the middle panel. When $N_{events}>3$ (dashed horizontal line), the corresponding sensitivity extends to arbitrarily small values of $d$ (vertical region in the middle panel).
• Figure 4: Upper panel: Same as in the middle panel of Fig. \ref{['fig:sensitivity1']}, but choosing $\delta = 0$ (instead of $\delta=0.1$). Middle panel: same as the upper panel, but mixing with the $\mu$ flavour (instead of the $\tau$ flavour). Lower panel: number of events generated by the EW dipole, for the same parameters as in the middle panel, but in the limit $d\to 0$.
• Figure 5: Sensitivity of the SHiP experiment to the dipole coefficient, as a function of the lightest sterile mass. In the upper panel, we consider a mixing with the $\mu$ flavour only, and vary the mass splitting $\delta$. In the lower panel, we set $\delta$ to zero, and consider a mixing with various combinations of lepton flavours. For definiteness, the SN excluded region corresponds to the universal mixing case (black) and the BBN excluded region to the electron mixing (green).