Sensitivity of an Early Dark Matter Search using the Electromagnetic Calorimeter as a Target for the Light Dark Matter eXperiment

TL;DR

This paper evaluates the sensitivity of the Light Dark Matter eXperiment (LDMX) to light dark matter by using the Electromagnetic Calorimeter (ECal) as an active target for a missing-energy search during early data-taking. It leverages Geant4-based simulations, with targeted biasing, to model signal via dark bremsstrahlung and dominant backgrounds (enriched nuclear and di-muon), and employs a simple, robust set of selection criteria on the total ECal energy, HCal veto, and shower morphology. A data-driven, sideband-based background estimation combined with a CLs framework yields projected 95% CL limits on the effective coupling $y$ as low as $\sim 2\times10^{-13}$ for $m_\chi = 1$ MeV and $\sim 5\times10^{-12}$ for $m_\chi = 10$ MeV, with 4 GeV and 8 GeV beam options. The study demonstrates that the ECal-channel can provide world-leading sensitivity in early LDMX operation and complements the main missing-momentum search, helping to broaden the reach for sub-GeV thermal dark matter in accelerator-based experiments.

Abstract

The Light Dark Matter eXperiment (LDMX) is proposed to employ a thin tungsten target and a multi-GeV electron beam to carry out a missing momentum search for the production of dark matter candidate particles. We study the sensitivity for a complementary missing-energy-based search using the LDMX Electromagnetic Calorimeter as an active target with a focus on early running. In this context, we construct an event selection from a limited set of variables that projects sensitivity into previously-unexplored regions of light dark matter phase space -- down to an effective dark photon interaction strength $y$ of approximately $2\times10^{-13}$ ($5\times10^{-12}$) for a 1MeV (10MeV) dark matter candidate mass.

Figures (8)

• Figure 1: A diagram of the ldmx detector apparatus, illustrating production of DM in the ecal from a scattering electron, and the corresponding response of the various sub-systems.
• Figure 2: A diagram of the ldmx ecal system. The longitudinal layer structure is shown (top) along with an exploded view of a single double-layer (bottom-right). The transverse view of the flower petal arrangement of modules as seen from the direction of the beam (bottom-left).
• Figure 3: The total reconstructed energy summed over all layers in the ecal ($E_\text{ECal}$) as a fraction of the beam energy ($E_\text{Beam}$) for selected mass points, shown for the 8GeV beam case. Heavier dark photons carry a larger amount of energy from the incident electron resulting in a shift of the total reconstructed energy in the ecal towards smaller values. The gray line indicates the trigger threshold. All distributions are normalized so their integral equals one.
• Figure 4: Total reconstructed energy in the ECal ($E_\text{ECal}$) as a fraction of the beam energy ($E_\text{Beam}$) depending on the total amount of simulated energy transferred to nuclear interactions (the "Nuclear Energy Fraction") for an unbiased sample of $10^9$eot. The maximum ECal energy allowed to pass the trigger is drawn in light gray. The elongated tail of the blue distribution (Nuclear Energy Fraction $< 10\%$) is due to photon conversion to muon pairs as evidenced by simulated energy deposits caused by muons within the hcal ("Muon Hits in hcal"). Events with $E_\text{ECal} > E_\text{Beam}$ due to resolution effects are omitted from this plot.
• Figure 5: The total reconstructed energy summed over all layers of the ECal ($E_\text{ECal}$) for all events that pass the trigger threshold. The signal and background distributions are normalized such that their integral is one. The events falling into bins with energy above the trigger threshold are omitted from this plot but included in efficiency calculations and the normalization.