Direct detection of sub-GeV dark matter with semiconductor targets

Dark matter in the sub-GeV mass range is a theoretically motivated but largely unexplored paradigm. Such light masses are out of reach for conventional nuclear recoil direct detection experiments, but may be detected through the small ionization signals caused by dark matter-electron scattering. Semiconductors are well-studied and are particularly promising target materials because their $$ \mathcal{O} $$ (1 eV) band gaps allow for ionization signals from dark matter particles as light as a few hundred keV. Current direct detection technologies are being adapted for dark matter-electron scattering. In this paper, we provide the theoretical calculations for dark matter-electron scattering rate in semiconductors, overcoming several complications that stem from the many-body nature of the problem. We use density functional theory to numerically calculate the rates for dark matter-electron scattering in silicon and germanium, and estimate the sensitivity for upcoming experiments such as DAMIC and SuperCDMS. We find that the reach for these upcoming experiments has the potential to be orders of magnitude beyond current direct detection constraints and that sub-GeV dark matter has a sizable modulation signal. We also give the first direct detection limits on sub-GeV dark matter from its scattering off electrons in a semiconductor target (silicon) based on published results from DAMIC. We make available publicly our code, QEdark , with which we calculate our results. Our results can be used by experimental collaborations to calculate their own sensitivities based on their specific setup. The searches we propose will probe vast new regions of unexplored dark matter model and parameter space.