Low Mass WIMP Searches with a Neutrino Experiment: A Proposal for Further MiniBooNE Running

TL;DR

This work proposes repurposing MiniBooNE to search for sub-GeV WIMPs by operating the proton beam off target into a 25m absorber, dramatically reducing neutrino backgrounds and enabling NC-like WIMP scattering searches in a previously inaccessible mass range ($\sim$10–200 MeV). It adopts a minimal dark-sector model with a light vector mediator that kinetically mixes with the photon, producing WIMPs via neutral meson decays and allowing detection through elastic scattering on nucleons or electrons, with timing and kinematic cuts enhancing discrimination. The authors provide sensitivity projections for both NC nucleon and NC electron channels, showing substantial improvements over on-target runs and potential coverage of the muon g-2 favored region, while also outlining practical deployment steps (e.g., a $25$ m absorber) and potential expansions to future fixed-target facilities. If successful, this program would deliver competitive WIMP limits or discovery within a year of data, and establish a blueprint for light dark-matter searches at existing and future beamlines. Overall, the proposal highlights a timely, cost-effective path to probing light dark matter and dark forces in a neutrino-beam setting with broad implications for particle physics and cosmology.

Abstract

A proposal submitted to the FNAL PAC is described to search for light sub-GeV WIMP dark matter at MiniBooNE. The possibility to steer the beam past the target and into an absorber leads to a significant reduction in neutrino background, allowing for a sensitive search for elastic scattering of WIMPs off nucleons or electrons in the detector. Dark matter models involving a vector mediator can be probed in a parameter region consistent with the required thermal relic density, and which overlaps the region in which these models can resolve the muon g-2 discrepancy. Estimates of signal significance are presented for various operational modes and parameter points. The experimental approach outlined for applying MiniBooNE to a light WIMP search may also be applicable to other neutrino facilities.

Figures (18)

• Figure 1: The contribution to the anomalous magnetic moment of SM fermions from the vector mediator. The crosses represent the kinetic mixing $\kappa$ of the vector $V$ with the photon.
• Figure 2: An illustration of the dark matter production modes and elastic scattering signatures.
• Figure 3: Top: The production of a WIMP pair through neutral meson decay. Bottom: The scattering of a WIMP in the MiniBooNE detector. The cross again represents the kinetic mixing between the vector mediator V and the photon.
• Figure 4: Regions of nucleon-WIMP scattering cross section (corresponding to dark matter in the halo moving with $v\sim10^{-3}c$) vs WIMP mass. The plot uses $m_V=300$ MeV and $\alpha^\prime=\alpha$. Constraints are shown from dark force searches (labeled $e^+e^-$, and including from left-to-right limits from KLOE, APEX, MAMI and BaBar) Hewett:2012ns, limits on $pp\to j+{\rm inv.}$monojet (labeled Monojet), limits on $J/\psi\to{\rm inv.}$ decays J/Psi_invis, excessive contributions to $(g-2)_\mu$Pospelov:2008zw, together with low-mass limits from the direct detection experiments CRESST CRESST (1-4 GeV) and XENON10 XENON10 (4-10 GeV). Note that a similar, but slightly stronger, exclusion contour to CRESST has also been obtained by DAMIC DAMIC. The light blue band indicates the region where the current $\sim3\sigma$ discrepancy in $(g-2)_\mu$ is alleviated by 1-loop corrections from the vector mediator Pospelov:2008zw. The solid black line shows points where the present relic density of the WIMP matches observations---the structure in this occurs when the WIMP mass is such that its annihilation during freeze-out through an off-shell $s$-channel $V^\ast$ is resonantly enhanced. This relationship only applies for $m_\chi<m_V$. The left panel shows regions where we expect 1--10 (light green), 10--1000 (green), and more than 1000 (dark green) elastic scattering events off nucleons in the MiniBooNE detector with $2\times10^{20}$ POT. The right panel shows the same for elastic scattering off electrons.
• Figure 5: Regions of mixing angle $\kappa$ ($\epsilon$) vs vector mass $m_V$ ($m_{A'}$), contrasting the existing sensitivity to light vectors in the two scenarios where the dominant decay mode of the vector is either visible or invisible. On the left we show the sensitivity to the light dark matter model considered here, in which $m_V > 2 m_\chi$ so that the vector predominantly decays invisibly and Br$(V\rightarrow {\rm SM})\sim \kappa^2\alpha'/\alpha$ with $\alpha'=\alpha$. On the right (reproduced from Hewett:2012ns, figure courtesy of R. Essig), this is contrasted with with the sensitivity in the absence of light dark matter, or with $m_V < 2m_\chi$, so that Br$(V\rightarrow {\rm SM})\sim {\cal O}(1)$. The shaded regions are existing limits, while the open contours are current and planned searches. Note that in the light dark matter scenario, many of the existing dark force constraints shown on the right are weakened by the reduced leptonic branching ratio, while beam dump limits that rely on a long lifetime for the $V$ are removed entirely. In the left-hand plot, as in Fig. \ref{['fig:sigmaNvsmchi']}, constraints from dark force searches (labeled $e^+e^-$) Hewett:2012ns, $pp\to j+{\rm inv.}$monojet (labeled Monojet), $J/\psi\to{\rm inv.}$ decays J/Psi_invis, and excessive contributions to $(g-2)_\mu$Pospelov:2008zw are shown, along with limits from $\pi^0\to\gamma+{\rm inv.}$pi) and $K^+\to\pi^++{\rm inv.}$Kplus decays. The light blue band again indicates the region where the current $\sim3\sigma$ discrepancy in $(g-2)_\mu$ is alleviated Pospelov:2008zw, and the solid black line shows the parameters required to reproduce the observed relic density of dark matter.