Matter- and magnetically-driven flavor conversion of neutrinos in magnetorotational collapses

Abstract

The magnetorotational collapse of massive stars copiously emits neutrinos of all flavors, with a prominent hierarchy between the non-electron and electron flavor average energies. Relying on a three-dimensional neutrino-magnetohydrodynamic simulation of a $13 M_\odot$ progenitor, we investigate flavor conversion in matter. We find that, in addition to resonant flavor conversion of neutrinos and antineutrinos in matter, (anti)neutrinos experience chirality-flipping interactions due to their non-zero magnetic moment ($μ\lesssim 10^{-12} μ_B$) and large magnetic field in the source ($B \simeq 10^{15}$ G). For Majorana neutrinos, this leads to resonant flavor-changing neutrino-antineutrino mixing. The event rate expected from a Galactic collapse at current and next-generation neutrino telescopes, such as IceCube and Hyper-Kamiokande, strongly depends on the orientation of the magnetorotational collapse with respect to the observer direction and flavor conversion scenario. The event rate is expected to be larger for an observer facing head on the jet launched during the stellar collapse and peaks around $400$-$600$ ms after bounce. Our work highlights that understanding the rich phenomenology of flavor conversion in magnetorotational collapses is essential to take full advantage of the joint detection of neutrinos and gravitational waves from these sources.

Figures (8)

• Figure 1: Radial profiles of the baryon density ($\rho$, top row), electron fraction ($Y_e$, middle row), and perpendicular magnetic-field strength ($B_\perp$, bottom row) at two post-bounce times ($t=0.3\,\mathrm{s}$ and $t=1\,\mathrm{s}$), shown along the pole ($\theta=0$ and $\varphi =0$, left column) and the equator ($\theta=\pi/2$ and $\varphi =0$, right column). The dotted lines in the bottom panel represent the output of the neutrino-magnetohydrodynamic simulation, but are unphysical and due to the progenitor model employed to initialize the simulation. The solid lines represent our interpolated profiles that smoothly bridge these unphysical gaps (see main text for more details).
• Figure 2: Time evolution of the neutrino luminosity (top row) and mean energies (bottom row) for electron neutrinos ($\nu_e$), electron antineutrinos ($\bar{\nu}_e$), and heavy-lepton species ($\nu_x=\nu_\mu$, $\nu_\tau$, $\bar{\nu}_\mu$, or $\bar{\nu}_\tau$). The left and right panels represent the neutrino emission properties as measured by a distant observer with angular coordinates close to the pole ($\theta=0$ and $\varphi =0$) and equator ($\theta=\pi/2$ and $\varphi = 0$), respectively.
• Figure 3: Radial profiles of $\rho Y_e$ for our magnetorotational collapse model at $t=0.3$ s (blue) and $t=1$ s (red), shown along the pole ($\theta=0$ and $\varphi = 0$, left panel) and the equatorial direction ($\theta=\pi/2$ and $\varphi = 0$, right panel). The horizontal lines indicate the MSW(H) and MSW(L) resonance conditions (cf. Eqs. \ref{['eq:msw_conditions1']} and \ref{['eq:msw_conditions2']}) in green and light blue, respectively. The resonance conditions have been computed using a representative neutrino energy corresponding to the time-averaged mean energy of all species in each direction: $E \simeq 14.4$ MeV for the polar direction and $E = 14.2$ MeV for the equator (see Fig. \ref{['fig:lum_en']}). The radii where the MSW resonances are expected are determined by the intersection between $\rho Y_e$ and the MSW(H) and MSW(L) resonance conditions. The resonance conditions are not significantly different for the selected post-bounce times and emission directions. The resonances occur in the proximity of $r_{\mathrm{MSW(H)}} \simeq 5\times10^{4}$ km and $r_{\mathrm{MSW(L)}} \sim 10^{5}$ km.
• Figure 4: Level-crossing diagrams for NO (left) and IO (right). The plots are computed for the equatorial direction ($\theta=\pi/2,\, \varphi = 0$) at $t = 0.3\,\text{s}$ using a neutrino energy $E = 14.2\,\text{MeV}$, which represents the time averaged mean energy across all species (see Fig. \ref{['fig:lum_en']}). The features of the level-crossing diagrams are qualitatively invariant across different propagation directions, time snapshots, or energies within the relevant range. For the calculation of the eigenstates relevant for the B-res resonances, a magnetic coupling of $\mu = 10^{-13} \mu_B$ has been assumed. For $Y_e<0.5$, the MSW(L) resonance occurs in the neutrino channel for both NO and IO. On the other hand, the MSW(H) resonance takes place in the neutrino (antineutrino) channel for NO (IO), as highlighted by the corresponding level crossings.
• Figure 5: Radial profiles of $\rho (1 - 2 Y_e)$ (top panels) and $\rho$ (bottom panels) for our magnetorotational model at $t = 0.3$ s (blue) and $t = 1$ s (red), along the polar ($\theta=0,\, \varphi = 0$, left) and equatorial ($\theta=\pi/2,\, \varphi = 0$, right) directions. In the top panels, the horizontal lines mark the B-res(H) and B-res(L) conditions (see Eqs. \ref{['eq:B-res_conditions1']} and \ref{['eq:B-res_conditions2']}). Negative values of $\rho(1 - 2Y_e)$ indicate regions with $Y_e > 0.5$. In the bottom panels, the purple and gray lines mark the B-res$^*$ condition (Eq. \ref{['eq:B-resstar']}). The resonance conditions are computed for a representative neutrino energy equal to the time-averaged mean energy of all species: $E \simeq 14.4$ MeV for the polar direction and $E = 14.2$ MeV for the equatorial direction (see Fig. \ref{['fig:lum_en']}).