Probing MeV Dark Matter at Low--Energy $e^+e^-$ Colliders

Abstract

It has been suggested that the annihilation of Dark Matter particles χwith mass between 0.5 and 20 MeV into e^+e^- pairs could be responsible for the excess flux of 511 keV photons coming from the central region of our galaxy that has been detected by the IINTEGRAL satellite. The simplest way to achieve the required cross section for χpair annihilation while respecting existing constraints is to introduce a new vector boson U with mass M_U below a few hundred MeV. In this Letter we point out that over most of the allowed parameter space, the process e^+e^- to U γ, followed by the decay of U into either an e^+e^- pair or into an invisible (ν\bar νor χ\bar χ) channel, should lead to signals that can be detected by the B-factory experiments BaBar and Belle. A smaller, but still substantial, region of parameter space can also be probed at the Φfactory DAFNE.

Figures (2)

• Figure 1: Parameter space for a Majorana $\chi$ with $g_{e_L} = g_\nu= 0$ and $g_\chi = 10 g_{e_R}$. In between the solid curves, $\chi$ has the correct relic density, and the correct cross section to explain the flux of 511 keV photons emerging from the galactic center; the red (black) curves correspond to $2 m_\chi < (>) M_U$. The dotted (blue) line indicates the upper bound on $g_{e_R}$ from $(g_e - 2$) measurements. The dashed curves show the maximal sensitivity of DA$\Phi$NE to $e^+e^- \rightarrow U \gamma$ production, for $U \rightarrow e^+ e^-$ (upper, dark green, curve) and $U$ decaying invisibly (lower, magenta curve). The DM constraints are essentially independent of the ratio $g_\chi / g_{e_R}$, whereas the $g_e-2$ constraint as well as the sensitivity limits are independent of $g_\chi$. Results for $g_{e_R} = g_{e_L}$ are similar to those for scalar $\chi$ (Fig. 2).
• Figure 2: Parameter space of the model with a complex scalar as MeV Dark Matter $\chi$ annihilating through the exchange of spin--1 $U$ bosons, for $g_{e_L} = g_\nu= 0$ and $g_\chi = 1$. Notation is as in Fig. 1, except that the indicated sensitivities are now those that can be achieved at the $B-$factories.