Dark Glueball Direct Detection

TL;DR

This work studies dark glueball DM from a confining dark SU(N) Yang–Mills sector connected to the SM via electrically charged vector-like portals with mass $m_\psi$. It develops a controlled EFT framework anchored in QCD phenomenology and a tensor–Pomeron inspired description of the two-photon couplings, enabling a first quantitative prediction for elastic $\chi A$ scattering that yields $σ_{SI} \propto Λ_D^{2.15} m_ψ^{-8}$. It identifies a light-portal window with $m_ψ \sim 3-30~\mathrm{GeV}$ and $Λ_D \sim 0.55-5.5~\mathrm{GeV}$ potentially testable by current and future Xenon experiments, while remaining consistent with collider constraints via a minimal EW completion. Overall, the work provides a realistic bridge from UV portals to direct-detection recoil spectra for confining dark sectors.

Abstract

We consider glueball dark matter (DM) in a Yang-Mills dark sector confined at $Λ_D$ scale and coupled to the Standard Model through electrically and dark-color charged vector-like fermion portals, with the mass scale $m_ψ$. In a simple case with two lightest mass-degenerate vector-like fermions with opposite electric charges the effective amplitudes with one $C$-odd glueball (oddball) and odd number of photons vanish, rendering the lightest $C$-odd spin-1 state with mass $m_χ$ a viable DM candidate provided that $m_ψ\gtrsim 5.5 Λ_D$. We develop a controlled effective field theory framework with non-perturbative information supported by QCD phenomenology leading to a quantitative prediction for coherent elastic glueball scattering off nuclei. We find a steep scaling of the spin-independent cross section $σ_{\rm SI}\propto Λ_D^{2.15} m_ψ^{-8}$. This implies that the sensitivity of the current and next-generation xenon experiments in the range of $σ_{\rm SI} \sim 10^{-46} - 10^{-48}$ cm$^2$ corresponds to $m_ψ\simeq 3-30$ GeV, respectively, for $Λ_D\simeq 0.55-5.5$ GeV. We provide a minimal UV completion of the portal sector compatible with collider phenomenology. Our results pave a quantitative foundation for testing glueball DM in direct-detection experiments.

Figures (3)

• Figure 1: UV-to-EFT matching for the effective two-photon-glueball vertex. The portal-fermion loop (left) is represented in the EFT as a dim-8 operator (right) that mediates both $\chi\to\gamma\gamma$ and $\chi A$ elastic scattering via two off-shell photons.
• Figure 2: Elastic glueball-nucleus scattering in the Coulomb regime. Left panel shows two-photon exchange mediated by portal-box, while right one illustrates the factorized form with the glueball Compton tensor $T^\chi_{\mu\nu}$, modeled as $T^\chi_{\mu\nu}\sim \Gamma^{(\gamma\gamma\mathbb{P}_D)}\,\Delta\,\Gamma^{(\chi\chi\mathbb{P}_D)}$ in the dark-Pomeron framework.
• Figure 3: Spin-independent (SI) per-nucleon cross section for glueball DM in the Coulomb regime compared to bounds from Xenon experiments. The colored domain spans over the portal-fermion mass $m_\psi$, in the color bar, and the oddball mass $m_\chi$ satisfying the glueball stability $m_\chi < 2m_\psi$. Projected detector sensitivities are given for PandaX-xT PANDA-X:2024dlo, PandaX-4T and LZ PandaX:2024qfuLZ:2024zvo with the neutrino floor OHare:2021utq.