Scattering meets absorption in dark matter detection

Direct detection experiments have started to explore dark matter scattering off electrons and nucleons through light mediators. Mediators with sub-keV masses are efficiently produced in the Sun and can be absorbed in the same detectors that probe dark matter scattering. We investigate the interplay of dark matter scattering and mediator absorption for two models with a dark photon as mediator. For Dirac dark matter, we find that scattering and absorption can be simultaneously observed at direct detection experiments in the near future. For atomic dark matter, we predict additional signals due to scattering of both dark atoms and constituents from ionized dark atoms. In both models, we determine the parameter space that respects bounds from cosmology and astrophysics, where the strongest constraints come from dark matter self-interactions. In this way, we identify viable targets for dark matter with light mediators at upcoming direct detection experiments. Distinguishing between the various signals, for instance by measuring energy distributions, will be crucial to reveal the underlying model in case of a discovery.

💡 Research Summary

This paper investigates the simultaneous appearance of two distinct signals—dark‑matter scattering and solar‑origin dark‑photon absorption—in the same direct‑detection experiments, focusing on sub‑keV vector mediators. The authors consider two concrete dark‑sector frameworks. The first, “Dirac dark matter,” features a single light Dirac fermion (the dark electron χ = e′) as the dark‑matter candidate, coupled to a massive dark photon A_d via a U(1)_d gauge interaction. The second, “atomic dark matter,” contains two dark fermions (e′ and p′) that form bound “dark hydrogen” atoms H′; the same dark photon mediates interactions both within the dark sector and with the Standard Model through kinetic mixing. The kinetic‑mixing parameter ε is taken to be ≲10⁻⁸, a regime that evades collider and beam‑dump limits while still allowing appreciable absorption rates in underground detectors.

The authors first review cosmological and astrophysical constraints. Cosmic‑microwave‑background (CMB) and big‑bang‑nucleosynthesis (BBN) limits on extra relativistic degrees of freedom impose a dark‑sector temperature ratio ξ ≲ 0.3; the paper adopts ξ = 0.3 as a conservative benchmark. Dark‑acoustic‑oscillation (DAO) effects are shown to be subdominant for the parameter space of interest. The most stringent bounds arise from dark‑matter self‑interactions mediated by the dark photon. Using the Bullet‑Cluster limit σ_T/m_DM ≲ 1 cm² g⁻¹ (v ≈ 1000 km s⁻¹) and small‑scale‑structure limits σ_T/m_DM ≲ 1–100 cm² g⁻¹ (v ≈ 10 km s⁻¹), the authors delineate viable regions where the self‑interaction cross section is velocity‑dependent, allowing sizable effects on dwarf‑galaxy scales while satisfying cluster constraints.

The paper then details the production and detection mechanisms. In the Sun, thermal electrons and ions generate dark photons via kinetic mixing; for sub‑keV masses the plasma frequency suppresses emission only modestly, leading to a solar dark‑photon flux that can be absorbed in terrestrial detectors. The absorption cross section scales as ε² α m_d²/(m_d² + ω_p²)², where ω_p is the solar plasma frequency. Current experiments (XENONnT, LZ, SuperCDMS) already probe ε ≈ 10⁻⁸, surpassing stellar‑cooling bounds for the chosen mass range.

Scattering of the dark matter itself proceeds through t‑channel exchange of the same dark photon. Because the mediator is light, the differential cross section scales as 1/q⁴, dramatically enhancing low‑momentum transfers. Electron scattering yields recoil energies in the eV–keV range, accessible to Skipper‑CCD, SENSEI, and other low‑threshold technologies, while nuclear scattering produces keV–MeV recoils detectable in liquid‑xenon or cryogenic crystal detectors.

For Dirac dark matter, the authors find a region where both scattering and absorption generate comparable event rates (∼1–10 events per tonne‑year). The absorption signal appears as a mono‑energetic line at the solar dark‑photon energy (∼keV), whereas the scattering signal is a continuous recoil spectrum. Distinguishing the two requires good energy resolution and, ideally, complementary detectors covering both low‑energy electron recoils and higher‑energy nuclear recoils.

Atomic dark matter introduces additional complexity. Dark hydrogen atoms can scatter off nuclei, while any ionized fraction f_ion of dark electrons and dark protons can scatter individually. The ionization fraction depends on the dark‑sector temperature ratio ξ and the dark fine‑structure constant α_d, ranging from 10⁻⁴ to 10⁻¹ in the viable parameter space. Consequently, three overlapping recoil spectra may be present: (i) atomic‑nucleus scattering, (ii) dark‑electron scattering, and (iii) dark‑proton scattering. The authors compute the combined rate and show that, for reasonable choices of α_d (10⁻³–10⁻¹) and mediator mass (meV–eV), the total signal remains within reach of upcoming experiments while respecting all astrophysical bounds.

A key part of the analysis is the proposal of experimental strategies to disentangle the overlapping signatures. Time‑modulation studies exploit the varying solar dark‑photon flux as the Earth orbits the Sun, providing a distinctive annual modulation for the absorption line. Directional information, obtainable in low‑pressure gas TPCs or anisotropic crystal detectors, can help separate isotropic scattering from the solar‑direction‑biased absorption. Moreover, a joint analysis of data from low‑threshold CCDs (sensitive to the electron‑recoil component) and high‑mass liquid‑xenon detectors (sensitive to nuclear recoils) can isolate the contributions from dark‑electron, dark‑proton, and atomic scattering.

In summary, the paper demonstrates that models with a sub‑keV dark photon mediator naturally predict concurrent scattering and absorption signals in direct‑detection experiments. By systematically incorporating cosmological, astrophysical, and laboratory constraints, the authors identify a well‑defined target region—ε ≈ 10⁻⁹–10⁻⁸, α_d ≈ 10⁻³–10⁻¹, m_DM ≈ MeV–GeV, m_d ≈ meV–eV—where upcoming experiments could observe both phenomena. The work expands the conventional dark‑matter search paradigm, showing that a single experiment can simultaneously probe two complementary interaction channels, thereby offering a powerful avenue for model discrimination in the event of a discovery.