Dark Photon Search with PADME at LNF

Dark Matter elusiveness could be explained by speculating that it lives in a separate sector with respect to the Standard Model and that interacts with it only by means of messengers. The simplest model foresees just one messenger: a, possibly massive, vector boson given by a new U(1) symmetry. This mediator can faintly mix with the photon and, hence, interact with SM charged particles, seeing an effective charge equal to εe, with e SM charge. The PADME experiment, hosted at Laboratori Nazionali di Frascati, is designed to search for such kind of particle, looking for its production in e+ e− annihilations. Exploiting the DAΦNE linac, the collaboration aims to collect 1013 positrons on target by the end of 2018, reaching a sensitivity of ∼ 10−3 for masses up to 23.7 MeV.


Introduction
Since many decades there is a problem in particle physics that is still waiting for a solution: the Dark Matter (DM) identification. By many indications we know that it must exist, but there are no clear and incontrovertible evidences of it.  introduces a new U(1) symmetry with its, possibly massive, vector boson, indicated with the symbol A 1,2 . The SM particles are neutral under this symmetry, while the mediator can faintly mix with the standard photon and couple to SM particles with an effective charge εe, where ε is the mixing constant and e is the electromagnetic charge of the SM particle. For this reason A is generally called Dark Photon (DP). This approach results to be extremely predictive and simple since it requires to introduce only two new parameters to describe the DP: its mass m A and the constant ε.
In addition, depending on the selected model, such a particle could justify, totally or partially, the observed difference 3 between the measured and the expected value of the muon anomalous magnetic moment (g − 2) µ . 4 If there are no DM particles with mass m DM such that m DM ≤ 1 2 m A , the A will decay only into SM particles (visible decays). In the other case, the A will decay predominantly into DS particles (invisible decays), suppressing by a factor ε 2 the SM decays. In both cases, the DP coming from this simple model has been discarded as the only responsible of the (g − 2) µ discrepancy. 5,6 To have a complete scenario of the research status and of the experimental techniques presently used see Refs. 7 and 8 and references therein.

The experimental technique
Hosted at the Laboratori Nazionali di Frascati and placed in the newly redesigned Beam Test Facility 9 , the PADME (Positron Annihilation into Dark Matter Experiment) 10,11 experiment has as main goal the search for the DP, looking for the reaction: The needed positrons are produced and accelerated up to 550 MeV by the Frascati linear accelerator, while the electrons come from a diamond active target. If an A is produced, the event would appear as a single photon with missing energy in the detector due to the DP, that leaves the experiment undetected. Hence, knowing the initial positron and electron quadri-momentum ( P e + , 550 MeV in the beam direction and P e − = 0, respectively) and measuring the recoil photon in the final state ( P γ ), it is possible to evaluate the squared missing mass as: Taking as reference Fig. 1, from right to left, the detector components are: • Diamond active target (2 × 2 cm 2 area for 100 µm thickness). It allows to measure the beam intensity and position (≈ 5 mm precision), thanks to vertical and horizontal graphitic strips. The target low Z is to reduce the bremsstrahlung background, while the small thickness is to lower the probability of positron multiple interactions. • Magnetic dipole (1 m length for 23 cm gap). It is placed 20 cm after the target and provides a ≈ 0.5 T field. It is needed to bend the beam towards the exit and to send particles that loose energy, typically for bremsstrahlung, to the veto's detectors.

Backgrounds and sensitivity
There are two main SM sources of background for the DP search: the bremsstrahlung (e + N → e + N γ) and the annihilation into two γs, possibly with the emission of a initial state radiation (e + e − → γ γ (γ)). The former is important for higher values of M 2 miss and is reduced using the SAC. The latter, instead, is important for medium and small M 2 miss and is cut down by the calorimeter granularity and geometry. In addition, if the clusters in the ECal are temporally too close each other, also the pile-up could represent an important source of background, but this can be kept under control maintaining the beam intensity around 5000 positrons per bunch. In Fig. 2 it is presented the comparison between the background without any applied selection (in red) and the one obtained requiring only one cluster in the ECal (with  an energy that depends on m A ), no hits in the veto and no photons with energy > 50 MeV in the SAC (in blue). The sensitivity of the experiment to the DP has been evaluated simulating 2.5 · 10 10 550 MeV Positrons On Target (POT) and, then, extrapolating the result to the desired number of events. In Fig. 3 are shown the expected sensitivities for 10 13 and 4 · 10 13 POT in the squared coupling constant versus mass plane. With these numbers of events, it is possible to scan, in a model independent way, since no assumptions are made on the A decay chain, ε 2 of the order of 10 −6 for a DP that decays in invisible.

Conclusions
In the last years the search for DP signals is growing in importance as a possible solution to the DM detection problem. It can also explain, partially or entirely depending on the selected model, the muon anomalous magnetic moment discrepancy between the theoretical expectations and the experimental results.
In this context, the PADME collaboration aims to collect 10 13 POT by the end of 2018, probing, in a model independent way, the existence of a DP with masses up to 23.7 MeV and coupling constants ε of 10 −3 or lower, that decays into DS particles.