Speaker
Description
The Belle II experiment has accumulated 575 fb⁻¹ of data from the SuperKEKB asymmetric electron-positron collider. We present the current status of the search for low-mass dark photons (A′) from $e^+e^- \rightarrow A'\gamma \rightarrow e^+e^-\gamma $reaction based on collision data recorded by the Belle II experiment. The dark photon is a hypothetical gauge boson that would appear as a narrow resonance in the electron-positron invariant mass spectrum. This analysis targets masses ranging from the electron pair production threshold up to the muon pair production threshold, with a particular focus on the sub-20 MeV region. This low-mass range is of special interest due to the presence of the X17 anomaly and the lack of strong constraints on the kinetic mixing parameter ($\epsilon$). The major challenge in this mass region is the photon conversion background, which mimics the signal signature. Belle II’s high-luminosity dataset and excellent vertex reconstruction capabilities provide enhanced sensitivity in this difficult low-mass regime. To mitigate background contamination, we employ a Boosted Decision Tree (BDT) classifier that leverages advanced features derived from vertex fitting. The trained model was validated using radiative dimuon events with a photon conversion $e^+e^-\rightarrow \mu^+\mu^-\gamma \rightarrow \mu^+\mu^-[\gamma \rightarrow e^+e^-]$. We will present the event selection criteria, validation and optimization strategies, and preliminary sensitivity projections based on Monte Carlo simulations.
| Keyword-1 | dark photon |
|---|---|
| Keyword-2 | hidden sector |
