Two Puzzles, One Solution: Neutrino Mass and Secluded Dark Matter

We present a minimal secluded dark-matter (DM) framework based on an extra $U(1)_X$ gauge symmetry. The model contains a Dirac DM particle $\chi$, three heavy neutrinos $N_I$ with masses $M_{N,I}$, and a singlet scalar $R$ that mixes with the Standard Model Higgs doublet $\Phi$ by an angle $\alpha$. A symmetry forbids the $\Phi$-$R$ portal at tree level; the leading portal then arises at one loop from the same Yukawa structures that generate active neutrino masses $m_{\nu,I}$, implying $\tan(2\alpha) \propto \sum_I m_{\nu,I} M^2_{N,I}/(v_h m_H^2)$, where $v_h$ and $m_H$ are the SM Higgs VEV and mass. For heavy-neutrino masses in the multi-TeV range, this yields a naturally tiny mixing, $\tan(2\alpha)\sim 5\times 10^{-11}\left(M_N/10~\mathrm{TeV}\right)^2$, which strongly suppresses DM signals in direct, indirect, and collider searches. For PeV-scale heavy neutrinos the DM-nucleon cross section can instead enter the reach of direct-detection experiments. The visible and dark sectors thermalize at temperatures of order a few times the mass of the lightest heavy neutrino, then subsequently decouple, and typically evolve with a slightly hotter dark bath. In the secluded regime, with $\tan(2\alpha)\ll 1$ and $m_\chi>m_{H_p}$, the relic density is set by $p$-wave annihilation $\chi\bar\chi \to H_p H_p$ (with $H_p$ the Higgs-like particle of the dark sector), and the dark-sector Yukawa couplings required to reproduce the observed abundance are $\sim(0.1\text{-}1)$, as in the standard WIMP case. For heavy-neutrino masses $\gtrsim 10~\mathrm{TeV}$, the mediator decays before nucleosynthesis without spoiling BBN observables, while the tiny portal suppresses present-day signals below current and near-future sensitivities. This links two long-standing puzzles, the absence of DM signals and the smallness of neutrino masses, within a predictive thermal framework.