Future CMB tests of dark matter: ultra-light axions and massive neutrinos

Measurements of cosmic microwave background (CMB) anisotropies provide strong evidence for the existence of dark matter and dark energy. They can also test its composition, probing the energy density and particle mass of different dark-matter and dark-energy components. CMB data have already shown that ultra-light axions (ULAs) with mass in the range $10^{-32}~{\rm eV} \to 10^{-26}~{\rm eV}$ compose a fraction $< 0.01$ of the cosmological critical density. Here, the sensitivity of a proposed CMB-Stage IV (CMB-S4) experiment (assuming a 1 arcmin beam and $< 1~\mu K{\rm-arcmin}$ noise levels over a sky fraction of 0.4) to the density of ULAs and other dark-sector components is assessed. CMB-S4 data should be $\sim 10$ times more sensitive to the ULA energy-density than Planck data alone, across a wide range of ULA masses $10^{-32}< m_{a}< 10^{-23}~{\rm eV}$, and will probe axion decay constants of $f_{a}\approx 10^{16}~{\rm GeV}$, at the grand unified scale. CMB-S4 could improve the CMB lower bound on the ULA mass from $\sim 10^{-25}~{\rm eV}$ to $10^{-23}~{\rm eV}$, nearing the mass range probed by dwarf galaxy abundances and dark-matter halo density profiles. These improvements will allow for a multi-$\sigma$ detection of percent-level departures from CDM over a wide range of masses. Much of this improvement is driven by the effects of weak gravitational lensing on the CMB, which breaks degeneracies between ULAs and neutrinos. We also find that the addition of ULA parameters does not significantly degrade the sensitivity of the CMB to neutrino masses. These results were obtained using the axionCAMB code (a modification to the CAMB Boltzmann code), presented here for public use.