No room to hide: implications of cosmic-ray upscattering for GeV-scale dark matter

TL;DR

This paper shows that cosmic-ray upscattering effectively rules out the possibility of strongly interacting sub-GeV dark matter reaching underground detectors, significantly constraining certain dark matter models.

Contribution

It provides a detailed analysis of cosmic-ray upscattering effects, including nuclear form factors and inelastic scattering, to set new limits on strongly interacting dark matter.

Findings

Cosmic-ray upscattering closes the window for strongly interacting sub-GeV dark matter.

Updated constraints significantly limit parameter space for baryonic dark matter candidates.

Momentum-dependent scattering effects are crucial for accurate dark matter detection predictions.

Abstract

The irreducible upscattering of cold dark matter by cosmic rays opens up the intriguing possibility of detecting even light dark matter in conventional direct detection experiments or underground neutrino detectors. The mechanism also significantly enhances sensitivity to models with very large nuclear scattering rates, where the atmosphere and rock overburden efficiently stop standard non-relativistic dark matter particles before they could reach the detector. In this article, we demonstrate that cosmic-ray upscattering essentially closes the window for strongly interacting dark matter in the (sub-)GeV mass range. Arriving at this conclusion crucially requires a detailed treatment of both nuclear form factors and inelastic dark matter-nucleus scattering, as well as including the full momentum-transfer dependence of scattering amplitudes. We illustrate the latter point by considering three generic situations where such a momentum-dependence is particularly relevant, namely for interactions dominated by the exchange of light vector or scalar mediators, respectively, and for dark matter particles of finite size. As a final concrete example, we apply our analysis to a putative hexaquark state, which has been suggested as a viable baryonic dark matter candidate. Once again, we find that the updated constraints derived in this work close a significant part of otherwise unconstrained parameter space.