Searching for dark matter signals with high energy astrophysical neutrinos in IceCube

Abstract

High-energy neutrinos provide a potentially powerful and distinctive probe for dark matter (DM) - neutrino interactions, particularly in environments with enhanced DM densities, such as the DM spikes predicted to form around supermassive black holes (SMBHs) at the center of active galactic nuclei (AGN). Recent observations by the IceCube Neutrino Observatory, which identified four AGN, namely TXS 0506+056, NGC 1068, PKS 1424+240, and NGC 4151 as neutrino sources, provide a unique opportunity to search for signatures of these interactions. In this study, we use IceCube data to derive the most stringent constraints to date on both the energy-dependent and energy-independent DM-neutrino scattering cross-sections. We perform a statistical analysis using data from individual sources as well as a combined (stacked) analysis of all four sources. Our strongest limits arise from the stacking analysis, yielding an upper bound of $σ_{0} \le 8\times 10^{-39}$ cm$^2$ for an energy-independent cross-section and $σ_{0} \le 10^{-39}$ cm$^2$ for a linearly energy-dependent cross-section, both at 90\% confidence level, particularly in scenarios involving the adiabatic growth of black holes.

Figures (6)

• Figure 1: DM column density $\Sigma$ Vs. $r_l$ (the lower limit to the integration given in Eq. (\ref{['eq:Sigma_chi']})) for each source. We considered the upper limit to the same integration $r=10^4$ pc and $m_\chi = 1$ GeV to obtain these plots. The shaded regions represent the possible range of $R_{\rm em}$ for each source.
• Figure 2: Constraints on $\sigma_0$ for the constant$\sigma_{\nu {\chi}}$ for all the benchmark models for all the sources. The blue-dotdashed line represents the bound from the stacking analysis combining the data from all four sources. The constraints obtained by Cline et al. for NGC 1068 Cline_2023 (black dashed) and TXS 0506+056 PhysRevLett.130.091402 (black dotted) are shown for comparisons. For context, we also overlay the model independent constraints on the DM-neutrino scattering cross-section coming from SN 1987A Lin:2022dbl (magenta), stellar-neutrinos Jho:2021rmn (gray solid), diffuse supernova (dSN) Xenon1T (yellow) and dSN Super-K (purple solid) Ghosh:2021vkt, CMB+BAO Mosbech:2020ahp (green dotted), Lyman-$\alpha$Hooper:2021rjc (orange dotted).
• Figure 3: Constraints on $\sigma_0$ for the energy-dependent$\sigma_{\nu \chi}$ for all the benchmark models for all the sources. The solid-blue line represents the bound from the stacking analysis combining the data from all four sources. The limits obtained previously Cline_2023PhysRevLett.130.091402 have been rescaled to $E_0=10$ TeV for comparisons.
• Figure 4: Limits on the model parameters for energy-dependent scattering cross sections are shown for pseudo-Dirac dark matter (left panel) and complex scalar dark matter (right panel). In each panel, the black solid and dot-dashed curves indicate parameter values that reproduce the observed dark matter relic abundance with $\Omega h^{2} = 0.12$Planck:2018vyg. The colored curves show the constraints obtained from our stacking analysis for the different SMBH spike density benchmark models considered. The gray shaded regions denote existing bounds on the parameter space from complementary probes, including constraints from dark matter annihilation into neutrinos Buckley:2022btu, as well as limits from early-Universe cosmology, astrophysical observations, and terrestrial accelerator searches for new gauge bosons in this class of models Bernal:2025szh.
• Figure 5: Limits on the dark matter-neutrino scattering cross-section vs dark matter mass. We include the freeze-out relic density in our two dark matter scenarios of interest. We overlay our stacking analysis results for each benchmark model spike density model.