Rekindling s-Wave Dark Matter Annihilation Below 10GeV with Breit-Wigner Effects

TL;DR

The paper demonstrates that Breit-Wigner resonant enhancements can revive thermal s-wave dark matter below the electroweak scale by exploiting ultra-narrow mediators that yield efficient annihilation near DM formation while suppressing late-time annihilation during recombination. A belated freeze-out framework is developed, incorporating kinetic decoupling which shifts the relic density and can modestly suppress it, while ensuring the CMB constraints are satisfied with carefully tuned resonance parameters. The authors derive model-independent relations for the relic abundance, kinetic-decoupling corrections, and the CMB-bound on resonance width, and they analyze velocity-dependent indirect-detection signals across galactic halos, dwarfs, and clusters. They conclude that thermal s-wave DM remains viable below about 10 GeV for certain narrow resonances, with MeerKAT and other X-ray/gamma-ray observations providing strong tests, and they highlight the need for highly symmetric dark sectors to realize such ultra-narrow states. Overall, the work opens a pathway to detect or constrain light thermal DM via resonant, velocity-dependent annihilation signatures while navigating stringent CMB limits.

Abstract

Velocity-independent (s-wave) annihilation of thermal Dark Matter is ruled out by CMB data for masses below 10GeV, effectively ruling out the possibility of indirectly detecting it in this mass range. We demonstrate in a model-independent framework that Breit-Wigner effects from very narrow resonances can circumvent CMB constraints, thereby reviving the potential to detect s-wave DM annihilation in the present Universe. The density of resonant s-wave Dark Matter continues to evolve long after chemical decoupling, leading to a scenario we refer to as belated freeze-out, where kinetic decoupling plays a significant role in determing the relic density.

Figures (6)

• Figure 1: DM abundance $Y_\chi$ as function of time, parameterized by $x= m_\chi/T$, after chemical decoupling at $x_f=20$ in the belated freeze-out scenario for $\epsilon_R=10^{-5}$. The dashed line assumes kinetic equilibrium ($x_d\epsilon_R\gg 1$), while the red, orange, green and blue curves represent the suppression due to kinetic decoupling at $x_d/x_f=1$, 10, 100, and 500, respectively. The inset highlights the mild enhancement of the abundance for very late decoupling at $x_d\epsilon_R=1/2$ (magenta) and 2 (brown). Abundances are shown in units of the kinetic equilibrium value at $x\to \infty$, denoted as $Y_0$.
• Figure 2: Resonance width $\gamma_R$ as function of the DM mass $m_\chi$. The blue-shaded area is excluded by CMB observations with different boundaries representing annihilation to electrons (green), muons (magenta), photons (orange) and pions (black), assuming $\epsilon_R=10^{-5}$. The pion bound assumes annihilation into both $\pi^0\pi^0$ and $\pi^+\pi^-$ channels, assuming isospin symmetry. The red region corresponds to resonance lifetime exceeding one second, potentially affecting BBN predictions. In the brown-shaded region the resonance branching ratios are not physical ($b_R>\omega/4$, assuming $J_R=1, S_\chi=0$). Dashed and dot-dashed lines represent results for resonances with $\epsilon_R=10^{-4}$ and $10^{-6}$, respectively, showing only photons for the CMB exclusion.
• Figure 3: Boost factor $R\equiv\langle \sigma v\rangle/\sigma v_0$ of the DM annihilation cross-section as function of the DM dispersion velocity $v_{\rm DM}$, relative to its zero-velocity counterpart, i.e., the cross-section relevant during recombination. Dashed, solid and dot-dashed lines correspond to $\epsilon_R=10^{-4},\ 10^{-5}$ and $10^{-6}$, respectively, with fixed $\gamma_R=10^{-12}$ (blue) and $10^{-15}$ (black). The vertical gray line marks the typical DM velocity dispersion $v_\odot=220\,$km/s in the Milky Way's halo, excluding the Galactic center Freese:1987wuMoffat:2024psv.
• Figure 4: DM annihilation cross-section $\langle\sigma v\rangle_{220}$ as function of the DM mass $m_\chi$, predicted by the relic density for a typical velocity dispersion $v_\odot=220\,$km/s and $\gamma_R=10^{-12}$ (black lines). The shaded regions indicate exclusions (recast as described in \ref{['sec:IDsignals']}) from previous DM indirect detection searches in the $e^+e^-$ (top) and $\mu^+\mu^-$ (bottom) channels: Fermi-LAT observations of Milky Way dwarf spheroidal galaxies Fermi-LAT:2015att (blue), constraints from the gas-rich dwarf galaxy Leo T Wadekar:2021qae (brown), the galaxy cluster legacy survey by MeerKAT Lavis:2023jju and X-ray observations a latitude ring between $42^\circ$ and $48^\circ$ by XMM-Newton Cirelli:2023tnx (gray). Dashed, solid and dot-dashed lines correspond to $\epsilon_R=10^{-4},\,10^{-5}$ and $10^{-6}$, respectively.
• Figure 5: DM abundance $Y_\chi$ (top) and temperature ratio $\xi=T'/T=y_\chi/y_{\chi\,\rm{eq}}$ (bottom) as functions of $x=m_\chi/T$, from a full numerical solution of the cBE system (gray), the semi-analytic approximation (blue), the fully analytic approximation (orange), and the simplified analytic prescription with instantaneous kinetic decoupling (dashed purple). Abundances are shown in units of the kinetic equilibrium value at $x\to\infty$, denoted as $Y_0$.