First galaxy ultraviolet luminosity function limits on dark matter-proton scattering

Abstract

Scattering between dark matter (DM) and protons leads to suppressed small-scale fluctuations, with implications for a variety of cosmological observables. In this work, we search for evidence of DM-proton scattering with an interaction cross section $σ\!=\!σ_0 (\frac{v}{c})^n$ for $n=0,2$ and $4$, corresponding e.g. to velocity-independent contact interactions from heavy mediators, velocity-dependent pseudoscalar-mediated scattering, and higher-order dipole interactions, respectively, using high-redshift ($z \sim4-10$) ultraviolet galaxy luminosity functions (UVLFs) observed by Hubble Space Telescope (HST). We employ an adjusted implementation of GALLUMI combined with the modified Boltzmann solver CLASS DMeff that accounts for interacting DM, and incorporate UVLF data from both blank and lensed HST fields, alongside Planck CMB data and the Pantheon supernova catalog in a Bayesian analysis framework to set constraints on $σ_0$. Our results show that including lensed UVLF data, which probe fainter galaxies than the blank HST fields and thus smaller scales, leads to a substantial improvement in the constraints on $σ_0$ for $n>0$, surpassing existing bounds from Milky-Way (MW) satellite abundance and CMB anisotropies. For $m_χ = 1\,\rm MeV $, for example, we set the upper bounds at $8.3\times 10^{-26} \, \rm cm^2$ for $n=2$ and $1.2\times 10^{-22} \, \rm cm^2$ for $n=4$. For $n=0$, our bound is within an order of magnitude of those from the Lyman-$α$ forest and MW satellites.

Figures (3)

• Figure 1: The impact of IDM on the UVLFs at $z=6$. Left: The scattering cross section is varied while setting $n=2$ and $m_\chi = 1 \, \mathrm{MeV}$. Increasing the interaction strength impedes structure formation at increasingly larger scales. Right: Here we set $n=2$ and $\sigma_0 = 10^{-22} \, \mathrm{cm^2}$ and vary $m_\chi$. Decreasing the DM particle mass has a similar effect as increasing the cross section, both leading to a deviation from the HST measurements. The other astrophysical and cosmological parameters used to generate these figures are taken to be the best fit values from the MCMC posteriors described below. The figures show both data from the HST blank fields and the lensed fields. We point out that besides the noticeable effect of the reduction in the galaxies number density, the dark acoustic oscillations in the matter power spectrum also lead to small oscillatory features in the UVLFs, which are distinct from the small-scale suppression effect of other DM models (i.e. Fig. 2 in Ref. Lazare:2024uvj for FDM).
• Figure 2: The $\sigma_0$ posteriors from the MCMC analysis, for $n=2$, $m_\chi = 1 \, \rm MeV$, where the UVLFs are calculated using the top-hat filter (pink), the smooth-k filter (cyan) and the latter with both blank and lensed fields. The dashed lines represent the 95% highest density interval, which is the value we report as the upper bound on the interaction strength.
• Figure 3: Constraints on the DM-proton interaction cross section for velocity power index $n=0,2,4$ from UVLFs when using the smooth-k filter, with and without including data from the lensed field. The bounds obtained with the lensed fields show a major improvement over current constraints for $n=2,4$, and are only inferior to constraints for the Lyman-$\alpha$Rogers:2021byl forest for and MW satellites DES:2020fxi for $n=0$.