Probing Non-Minimal Dark Sectors via the 21 cm Line at Cosmic Dawn

TL;DR

This work demonstrates that the sky-averaged 21 cm signal from cosmic dawn can constrain metastable particles and non-minimal dark sectors by analyzing energy deposition into the IGM and its impact on the spin temperature $T_S$. It develops a model- and parameter-space–driven framework that combines DarkHistory energy-deposition modeling with back-reaction effects and a threshold-based bound approach, and it applies this framework to two explicit microscopic realizations: an ALP coupled to photons and a pseudo-Dirac DM system with vector portals or dipole operators. The authors obtain leading bounds on lifetimes, abundances, and mass splittings across broad DM mass ranges (down to $m_\chi \sim 20.4~\mathrm{eV}$) and show that 21 cm constraints can surpass CMB limits in many regions, while remaining applicable to non-minimal dark sectors. Collectively, the results provide a robust methodological foundation for using 21 cm cosmology to probe non-minimal DM scenarios and motivate future 21 cm monopole and power-spectrum analyses.

Abstract

Observations of the hydrogen hyperfine transition through the 21 cm line near the end of the cosmic dark ages provide unique opportunities to probe new physics. In this work, we investigate the potential of the sky-averaged 21 cm signal to constrain metastable particles produced in the early universe that decay at later times, thereby modifying the thermal and ionization history of the intergalactic medium. The study begins by extending previous analyses of decaying dark matter (DM), incorporating back-reaction effects and tightening photon decay constraints down to DM masses as low as 20.4 eV. The focus then shifts to non-minimal dark sectors with multiple interacting components. The analysis covers two key scenarios: a hybrid setup comprising a stable cold DM component alongside a metastable sub-component, and a two-component dark sector of nearly degenerate states with a metastable heavier partner. A general parameterization based on effective mass spectra and fractional densities allows for a model-independent study. The final part presents two explicit realizations: an axion-like particle coupled to photons, and pseudo-Dirac DM interacting via vector portals or electromagnetic dipoles. These scenarios illustrate how 21 cm cosmology can set leading bounds and probe otherwise inaccessible regions of parameter space.

Figures (14)

• Figure 1: Simulated monopole signal $\overline{\delta T_b}$ in the redshift range $8 < z < 25$, computed using the publicly available code Zeus21mun:ane, shown for two astrophysical scenarios. Left panel: Default fiducial parameters. Right panel: Maximally efficient Ly$\alpha$ coupling with no astrophysical heating.
• Figure 2: Bounds on the DM lifetime $\tau$ as a function of the DM mass $m_\chi$ in the single-component scenario. These limits assume a signal at redshift $z_t = 15$ with a threshold $\mathcal{T}/\mathcal{A}_{\rm std} = 0.25$. Left panel: Mass range $2\mathcal{R} < m_\chi < 2m_\mu$, where the only allowed final states are photons and $e^\pm$ pairs. Right panel: Mass range $2m_\mu < m_\chi < 2$ TeV. The vertical gray dashed line at $10$ GeV marks the lower bound for hadronically decaying particles to avoid perturbativity issues.
• Figure 3: Bounds on the DM lifetime $\tau$ as a function of the DM mass $m_\chi$ for the channel $\chi \to \gamma \gamma$. We investigate the dependence of the bounds on the redshift of the trough $z_t$ and the threshold $\mathcal{T}/\mathcal{A}_{\Lambda\rm CDM}^{\rm max}$. Existing constraints are shown shaded in light gray wad:strArias:2012azEssig:2013goaHoriuchi:2013noaNg:2019gchSicilian:2020glgFoster:2021ngmFoster:2022nvaRoach:2022lgoCirelli:2023tnxMassari:2015xeaGiesen:2015ufaBoudaud:2016mos. Left panel: The redshift is fixed at $z_t = 15$, and three different values of the threshold are shown. Right panel: The threshold is fixed at $\mathcal{T}/\mathcal{A}_{\Lambda\rm CDM}^{\rm max} = 0.50$, while the redshift $z_t$ is varied.
• Figure 4: Same as Fig. but for the decay channel $\chi\to e^+ e^-$.
• Figure 5: Maximal allowed fraction $F^{\rm max}_\chi$ of the unstable component $\chi$ in the $(m_\chi, \tau)$ plane, as constrained by a 21 cm monopole observation with $z_t = 15$ and $\mathcal{T}/\mathcal{A}_{\Lambda\rm CDM}^{\rm max} = 0.5$. Left panel: Decay mode $\chi \to \gamma\gamma$. Right panel: Decay mode $\chi \to e^+ e^-$.