New Multi-messenger Probe of Dark Matter-Nucleon Interactions from Ultra-high Energy Cosmic Ray Acceleration

TL;DR

The study investigates how dense DM spikes around supermassive black holes could influence the acceleration and survival of ultra-high-energy cosmic rays, particularly iron nuclei, by enabling DM–nucleon interactions that fragment heavy nuclei. By requiring that such fragmentation not occur during acceleration or escape from AGN environments, the authors derive upper bounds on the DM–nucleon cross section $\sigma_{\chi p}$ across multiple DM-density profiles (Burkert, NFW, and spikes) and source types (Milky Way-like and NGC1068-like). They obtain bounds such as $\sigma_{\chi p} \lesssim 7\times 10^{-29}\left(m_\chi/\mathrm{GeV}\right)\ \mathrm{cm^2}$ for spike scenarios, tightening to $\lesssim 9\times 10^{-31}\left(m_\chi/\mathrm{GeV}\right)\ \mathrm{cm^2}$ for NGC1068-like spikes, with a non-conservative bound as strong as $\sigma_{\chi p} \lesssim 8\times 10^{-35}\left(m_\chi/\mathrm{GeV}\right)\ \mathrm{cm^2}$. The results are relevant for momentum transfers $\gtrsim 10\ \mathrm{MeV}$ and provide complementary constraints to direct detection and cosmological probes, highlighting the potential of multi-messenger CR observations to test DM-spike physics. The work also discusses caveats related to the acceleration region, possible baryonic feedback, and the TXS 0506+056 neutrino source, and calls for more detailed CR-acceleration modeling that incorporates spikes to further sharpen these limits.

Abstract

It has been suggested that the density of dark matter (DM) halo can be highly enhanced around supermassive black holes at the centers of massive galaxies. If real, these DM \emph{spikes} would offer new opportunities to probe the properties of DM. In this work, we point out that DM spikes can significantly impact the composition and survivability of ultra-high-energy cosmic rays accelerated near supermassive black holes. A large DM-nucleon cross section would fragment heavy nuclei into lighter elements and prevent them from attaining the energies observed at Earth. While the origin of cosmic rays remains a mystery, we show that if the highest-energy cosmic rays on Earth come from sources like NGC1068, then cross sections of size $σ_{χp} \leq 3 \times 10^{-34} \left( \frac{m_χ}{\mathrm{GeV}}\right)\;\mathrm{cm^{2}}$ would be excluded by cosmic ray data. These bounds can be competitive with other existing probes in the DM mass region $m_χ\in [5\;\mathrm{MeV}, 50\;\mathrm{MeV}]$. While the uncertainties on the acceleration mechanism of cosmic rays prevent us from setting robust limits, our study highlights an important connection between DM spikes and cosmic ray physics that is complementary to existing cosmological and direct detection constraints.

Figures (4)

• Figure 1: The approximate bounds set here (pink region) using a spiked DM-density profile Gondolo:1999ef in an NGC 1068-type galaxy compared to other proposed bounds using CRs (black lines). Specifically, we are showing scattering Cappiello:2018hsu, and cooling Herrera:2023nww bounds. We also show DM-proton cross section values that can produce the correct relic abundance via freeze-in (green region) Elor:2021swjBhattiprolu:2022sdd. Our potential bounds using a Burkert Burkert:1995yz, NFW Navarro:1995iw, and spiked-density profiles for Milky-Way-like galaxies are shown as dotted, dashed-dotted, and dashed lines, respectively.
• Figure 2: The DM density as predicted by Gondolo:1999ef combined with the NFW Navarro:1995iw profile starting at $r = 4R_\mathrm{S} = 1.6\times 10^{-9}\;\mathrm{kpc}$. The orange region shows where DM self-interactions (e.g., annihilation) could reduce the expected density. The black dashed line denotes the expected DM density when accounting for significant stellar heating Balaji:2023hmy, and the dotted black line indicates the plateau induced by an annihilation of $\langle\sigma v \rangle \sim 10^{-25}$ cm$^3$/s. The pink vertical lines indicate the conservative acceleration regions of CRs for a Milky-Way-like or NGC1068-like galaxy, respectively, based on multi-messenger observations.
• Figure 3: The resulting bounds using the NFW and Spike profiles for the Milky-Way (marked MW) and assuming NGC1068-like population. These bounds are constructed under the assumption of a GZK-like process. The experimental bounds are from LZ LZ:2025iaw. Direct detection bounds adopted from Billard:2021uyg with bounds from CRESST CRESST:2017uesCRESST:2022dtl, DAMIC DAMIC:2020cut, DarkSide DarkSide:2018bpjDarkSide:2018kuk, and XENON1TXENON:2018vocXENON:2019gfn. BBN Sabti:2019mhnKrnjaic:2019dzcSabti:2021reh.
• Figure 4: The bounds for strongly interacting DM compared to current constraints. Current constraints are from XQC Erickcek:2007jv, CMB Gluscevic:2017ywp, Upscattered DM (LZ-Boosted LZ:2025iaw and PROSPECT PROSPECT:2021awi), EDELWEISS EDELWEISS:2019vjv, CRESST 2017 (surface) CRESST:2017uesEmken:2018run, XENON1T XENON:2017vdwXENON:2018vocXENON:2019gfn, CRESST-III CRESST:2019jnqCRESST:2019axx, DAMIC DAMIC:2020cut, DarkSide DarkSide:2018bpjDarkSide:2018kuk, and XENON1TXENON:2018vocXENON:2019gfn. BBN Sabti:2019mhnKrnjaic:2019dzcSabti:2021reh. Milky-Way Sattelites Nadler:2019zrb, Lyman-$\alpha$Rogers:2021byl, and re-interpretations of existing data Emken:2019tni.