Search for MeV dark photons in a light-shining-through-walls experiment at CERN

In addition to gravity, there might be another very weak interaction between the ordinary and dark matter transmitted by $U^{\prime }(1)$ gauge bosons $A^{\prime }$ (dark photons) mixing with our photons. If such $A^{\prime }$’s exist, they could be searched for in a light-shining-through-a-wall experiment with a high-energy electron beam. The electron energy absorption in a calorimeter (CAL1) is accompanied by the emission of bremsstrahlung $A^{\prime }$’s in the reaction $eZ\to eZ{A}^{\prime }$ of electrons scattering on nuclei due to the $\gamma -A^{\prime }$ mixing. A part of the primary beam energy is deposited in the CAL1, while the rest of the energy is transmitted by the $A^{\prime }$ through the “CAL1 wall” and deposited in another downstream calorimeter CAL2 by the $e^{+}{e}^{-}$ pair from the $A^{\prime }\to e^{+}{e}^{-}$ decay in flight. Thus, the $A^{\prime }$’s could be observed by looking for an excess of events with the two-shower signature generated by a single high-energy electron in the CAL1 and CAL2. A proposal to perform such an experiment to probe the still unexplored area of the mixing strength ${10}^{-5}\lesssim \epsilon \lesssim {10}^{-3}$ and masses $M_{A^{\prime }}\lesssim 100\mathrm{MeV}$ by using 10–300 GeV electron beams from the CERN SPS is presented. The experiment can provide complementary coverage of the parameter space, which is intended to be probed by other searches. It has also a capability for a sensitive search for $A^{\prime }$’s decaying invisibly to dark-sector particles, such as dark matter, which could cover a significant part of the still allowed parameter space.

Figure 1

Schematic illustration of the setup to search for dark photons in a light-shining-through-a-wall-type experiment at high energies. The incident electron energy absorption in the CAL1 is accompanied by the emission of bremsstrahlung $A^{\prime }$’s in the reaction $eZ\to eZ{A}^{\prime }$ of electrons scattering on nuclei due to the $\gamma -A^{\prime }$ mixing, as shown in Fig. 2. The part of the primary beam energy is deposited in the CAL1, while the rest of the total energy is transmitted by the $A^{\prime }$ through the CAL1 wall. The $A^{\prime }$ penetrates the CAL1 and veto V1 without interactions and decays in flight in the DV into a narrow $e^{+}{e}^{-}$ pair, which generates the second electromagnetic shower in the CAL2 resulting in the two-shower signature in the detector. The sum of energies deposited in the $\mathrm{CAL}1+\mathrm{CAL}2$ is equal to the primary beam energy.

Figure 2

Diagram illustrating the massive $A^{\prime }$ production in the reaction $e^{-}Z\to e^{-}Z{A}^{\prime }$ of electrons scattering off a nuclei $(A,Z)$ with the subsequent $A^{\prime }$ decay into an $e^{+}{e}^{-}$ pair.

Figure 3

Expected distributions of energy deposition for selected events (i) in the CAL1 (shaded) and (ii) in the CAL2 from the bremsstrahlung $A^{\prime }\to e^{+}{e}^{-}$ decays in flight in the DV region. The spectra are calculated for the 10 MeV $A^{\prime }$’s produced by 100 GeV $e^{-}$’s in the CAL1 with momentum pointing towards the CAL2 fiducial area and the mixing strength $\epsilon =3\times {10}^{-4}$. For this mixing value, most of the $A^{\prime }$’s decay outside of the CAL1 in the DV. The distributions are normalized to a common maximum.

Figure 4

Exclusion region in the $(M_{A^{\prime }};\epsilon )$ plane obtained in the present work from the expected results of the experiments accumulated ${10}^9$ $e^{-}$’s at 300 GeV (H4-a), ${10}^{11}$ $e^{-}$’s at 300 GeV (H4-b), and ${10}^{13}$ $e^{-}$’s at 10 GeV (H4-c). Shown are also areas excluded from the electron (g-2) considerations ($a_e$) [15], by the results of the electron beam-dump experiments E141 [16] and E774 [17], by searches at LAL Orsay [18], U70 (Protvino) [19], and KLOE [20], from kaon decays [21] and data of the experiment SINDRUM [22, 23], and by the WASA-at-COSY Collaboration [24]. Expected sensitivities of the planned APEX (full run), HPS, and DarkLight experiments are also shown for comparison [2]. For a review of all experiments, which intend to probe a similar parameter space, see Ref. [2] and references therein. In addition, the light grey area shows the $\pm 2\sigma$ preferred band from the muon g-2 anomaly consideration.

Figure 5

Diagram illustrating the massive $A^{\prime }$ production in the reaction $e^{-}Z\to e^{-}Z{A}^{\prime }$ of electrons scattering off a nuclei $(A,Z)$ with the subsequent $A^{\prime }$ decay into a $\chi \overline{\chi }$ pair.

Figure 6

The scheme of the additional tagging of high-energy electrons in the beam by using the electron synchrotron radiation in the banding magnetic dipole. The synchrotron radiation photons are detected by a $\gamma$ detector by using the LYSO inorganic crystal (Sc) capable for the work in vacuum. The crystal is viewed by a high quantum efficiency photodetector, e.g., PMT, SiPM, or APD. The beam defining counters S and veto V are also shown.

Figure 7

Constraints in the $\epsilon$ vs $M_{A^{\prime }}$ plane for invisibly decaying $A^{\prime }$ into a pair of light dark-matter particles $\chi \overline{\chi }$ provided $M_{A^{\prime }}>2m_{\chi }$. The light blue and blue areas show the expected 90% C.L. exclusion areas corresponding, respectively, to ${10}^9$ and ${10}^{12}$ accumulated electrons at 100 GeV for the background-free case. Other existing constraints, mostly adapted from Ref. [27], are also shown. The constraint from the BABAR monophoton search is given as the light grey shaded region. Further limits are shown from the anomalous magnetic moment of the electron [$(g-2{)}_e$], and DarkLight, the rare kaon decay $K^{+}\to {\pi }^{+}{A}^{\prime }$ (E787), and LSND experiments. The LSND area is determined assuming $A^{\prime }-\chi$ coupling ${\alpha }_D=0.1$, and that $\chi$ cannot decay to other light dark-sector states which do not interact with $A^{\prime }$’s [30]. The red shaded region is preferred in order to explain the discrepancy between the measured and the predicted value of the anomalous magnetic moment of the muon. A more complete plot including various other constraints from performed and planned experiments can be found in Ref. [29].