Speaker
Description
Accidental coincidence events have emerged as the dominant background in low-energy rare event searches using dual-phase xenon time projection chambers. Interactions in the liquid xenon result in a primary scintillation signal (S1) and the release of electrons. These electrons then drift up under an electric field to the liquid surface where they are extracted and produce a secondary electroluminescence signal (S2) as they traverse the gaseous xenon. Accidental events arise when S1 and S2 pulses from separate energy depositions or instrumental effects are mistakenly paired and reconstructed as a single signal-like event. The rate of accidental coincidence events increases rapidly as the energy threshold is lowered, yet our ability to remove them with data quality cuts is reduced. This necessitates a robust accidental coincidence background model.
In this talk, I present a novel data-driven methodology for modelling these backgrounds in the LUX-ZEPLIN (LZ) detector. This approach crucially considers the time-varying nature of the isolated pulses that give rise to accidental coincidence events and their relation to the detector conditions. This model was successfully implemented in the most recent LZ analysis, enabling world-leading limits on spin-independent dark matter interactions for masses between 5 and 9 GeV/c$^2$, as well as the observation of boron-8 coherent elastic neutrino-nucleus scattering at 4.5𝜎 significance.