Constraining exotic dark matter models with the dark ages 21-cm signal

Abstract: The dark ages 21-cm signal is a powerful tool for precision cosmology and probing new physics. We study two non-standard models: an excess radio background (ERB) model (possibly generated by dark matter decay) and the millicharged dark matter (mDM) model. These models were inspired by the possible EDGES detection of a strong global 21-cm absorption during cosmic dawn, but more generally they provide a way to anticipate the potential discovery space. During the dark ages the 21-cm global signal in the ERB model reaches a saturated form for an amplitude $A_{\rm r}=0.4$, where $A_{\rm r}$ is the radio background intensity at cosmic dawn relative to the cosmic microwave background. This amplitude is one-fifth of the minimum required to explain the EDGES signal, and corresponds to just 0.1% of the observed extragalactic background; it would give a signal that can be detected at 5.9$\sigma$ significance (compared to $4.1\,\sigma$ for the standard signal) and can be distinguished from the standard (no ERB) signal at $8.5\,\sigma$, all with a 1,000 hr global signal measurement. The 21-cm power spectrum has potentially more information, but far greater resources would be required for comparable constraints. For the mDM model, over a range of viable parameters, the global signal detection significance would be $4.7-7.2\,\sigma$, and it could be distinguished from the standard at $2.2-9.3\,\sigma$. With an array of global signal antennas achieving an effective 100,000 hr integration, the significance would be $10\,\times$ better. Our analysis helps motivate the development of lunar and space-based dark ages experiments.