Endothermic dark matter with a light dark photon and the LUX--ZEPLIN high-energy nuclear-recoil candidate

Abstract: The LUX-ZEPLIN (LZ) experiment has reported a single nuclear-recoil candidate at $$E_{\rm nr}=248\pm23_{\rm stat}\pm23_{\rm sys}\,{\rm keV}$$. We investigate whether this event can be explained by endothermic inelastic dark matter coupled to a kinetically mixed dark photon, while reproducing the observed dark-matter relic abundance. Performing a global scan of the five model parameters, combining an energy-only recast of the LZ high-energy likelihood with a relic-density likelihood, we find a preferred region with TeV-scale dark matter masses, mass splittings of a few hundred keV, and a GeV-scale dark photon. The high recoil energy requires the splitting to lie close to the kinematic threshold, so that the signal is supplied by the high-velocity tail of the halo, while the secluded annihilation mechanism fixes the dark gauge coupling, largely independently of the kinetic mixing. The benchmark point predicts $1$ accepted event at the candidate energy with $Ωh<sup>2=0.120$. The preferred splittings are below the $e<sup>+e<sup>-$ threshold, closing the fastest decay channels and leaving a long-lived excited state. Its surviving population is subject to stringent cosmological constraints from energy injection and can also produce an additional exothermic scattering signal, making the late-time abundance an important consistency condition for the minimal model. A dimension-five transition dipole provides a simple way to efficiently deplete $χ_2$ without modifying either the relic abundance or the endothermic LZ signal. The corresponding light-dark-photon scenario remains testable in accelerator searches, including future LHCb, Belle II, and SHiP experiments.