A strong broadband 21 cm cosmological signal from dark matter spin-flip interactions

In the standard cosmology, it is believed that there are two relatively weak and distinct band-limited absorption features, with the first absorption minima near 20 MHz (z similar to 70) and the other minima at higher frequencies between 50 - 110 MHz (z similar to 12 - 27) in the global cosmological 21 cm signal, which are signatures of collisional gas dynamics in the cosmic dark ages and Lyman-alpha photons from the first stars at cosmic dawn, respectively. A similar prediction of two distinct band-limited, but stronger, absorption features is expected in models with excess gas cooling, which have been invoked to explain the anomalous EDGES signal. In this work, we explore a novel mechanism, where dark matter spin-flip interactions with electrons through a light axial-vector mediator could directly induce a 21 cm absorption signal which is characteristically different from either of these. We find generically, that our model predicts a strong, broadband absorption signal extending from frequencies as low as 1.4 MHz (z similar to 1000), from early in the cosmic dark ages where no conventional signal is expected, all the way up to higher frequencies where star formation and X-ray heating effects are expected to terminate the absorption signal. We find a rich set of spectral features that could be probed in current and future experiments looking for the global 21 cm signal. In the standard cosmology and in excess gas cooling models it is expected that the gas spin temperature as inferred from the absorption signal is a tracer of the gas kinetic temperature. However, in our model we find in certain regions of parameter space that the spin temperature and kinetic temperature of the gas evolve differently, and the absorption signal only measures the spin temperature evolution. Large swathes of our model parameter space of interest are safe from existing particle physics constraints, however future searches for short range spin-dependent forces between electrons on the millimeter to nanometer scale have the potential to discover the light mediator responsible for our predicted signal.