Constraining light dark matter in vector-scalar portals with COSI and AMEGO-X

TL;DR

This work addresses sub-GeV dark matter in vector-scalar portals with MeV-scale mediators, focusing on gamma-ray signatures that include lines, boxes, and continuum spectra. It develops a concrete framework with a Dirac DM $\chi$, a $Z'$ mediator, and a light scalar $h'$, exploring two benchmark charges, $U(1)_{B-L}$ and $U(1)_A$, and forecasting sensitivities for the COSI and AMEGO-X missions using Hazma. The authors find that COSI can provide leading indirect-detection constraints beyond CMB in many scenarios, while AMEGO-X can cover most viable parameter space through continuum gamma rays; they also map out how spectral features depend on mass hierarchies and resonance effects. The Hazma toolkit implementation and its GitHub release enable broader application to related vector-scalar portal models and motivate synergy with direct detection and collider searches.

Abstract

Detecting gamma-ray signals that could be due to dark matter (DM) particles would give us invaluable information about the nature of DM. In particular, gamma-ray lines could provide a way to measure the DM mass. The excellent energy resolution of the upcoming Compton Spectrometer and Imager (COSI) will allow us to probe underexplored regions of the DM parameter space while being sensitive to distinctive spectral features of potential DM signals. In this work, we consider a fermionic sub-GeV DM charged under a new U(1) gauge symmetry. Both the DM and the new gauge boson $Z'$ acquire mass from a new singlet scalar. The masses of the new particles in this class of vector-scalar portal models are naturally at the MeV scale, enabling detectable gamma-ray lines in the bandpasses of COSI and proposed missions such as the All-sky Medium Energy Gamma-ray Observatory eXplorer (AMEGO-X). We estimate the sensitivities of COSI and AMEGO-X to sub-GeV DM in this context, considering a B-L and a purely axial $Z'$ as benchmark examples. We find regions of the parameter space where COSI will provide leading constraints, beyond the strong CMB limits. On the other hand, AMEGO-X would probe most of the viable parameter space leading to continuum gamma rays. The implementation of our generic vector-scalar portal model in the Hazma toolkit is available at GitHub.

Figures (8)

• Figure 1: Relevant processes leading to gamma-ray signals in vector-scalar portal models. The black dots represent effective interactions and the gray tick lines represent a vector or scalar particle. For simplicity, we omit the exchanges of the SM $h$ and $Z$, since their contributions to the cross-sections are subdominant.
• Figure 2: Gamma-ray spectra emitted by the annihilation of our light dark matter for different mass hierarchies: $m_{h'}<m_\chi<m_e, m_{Z'}$ (upper left), $m_{h'}=m_\chi<m_e, m_{Z'}$ (upper right), $m_{h'}, m_e<m_\chi< m_{Z'}$ (lower left), and $m_{h'}, m_e, m_{Z'} <m_\chi$ (lower right). This figure shows the possible spectral features in vector-scalar portals as would be seen by the upcoming COSI telescope (blue) and the proposed telescope AMEGO-X (red), as well as by a telescope with a fixed energy resolution of $1\%$ FWHM (dashed gray).
• Figure 3: Velocity-averaged annihilation cross-section of our light dark matter into channels leading to $\gamma$ rays for different charge assignments ($U(1)_{B-L}$ in left panels and $U(1)_A$ in right panels) and different values of singlet scalar mass ($m_{h'} = 0.3$ MeV in top panels and $m_{h'} = 2 m_\chi$ in bottom panels). The remaining parameters are set to $g_X = 10^{-3}$, $s_\alpha = 0.1$, and $m_{Z'}=4$ MeV. The annihilation into leptons is dominated by the exchange of $Z'$ bosons, which are produced on-shell when $m_\chi = m_{Z'}/2$. This leads to the resonance enhancement observed in the $e^+ e^-$ channel. In the $U(1)_{B-L}$ case, the $\gamma \gamma$ channel is relevant only in the regime of singlet scalar resonance (bottom left), whereas in the $U(1)_A$ case, it is relevant provided $g_X$ is strong enough and for light enough $Z'$, independently of the scalar mixing.
• Figure 4: Measurements of diffuse gamma-rays from COMPTEL Strong:1998ck, INTEGRAL/SPI Siegert:2022jiiBerteaud:2022tws, and the COSI balloon flight of 2016 COSI:2023bsm, as well as the scaled point-source continuum (blue solid) and line (blue dashed) sensitivities reported for the COSI mission Tomsick:2023aue and the scaled point-source continuum sensitivity for AMEGO-X Caputo:2022xpx (red).
• Figure 5: Indirect detection constraints on the total dark matter annihilation cross-section as a function of the dark matter mass, for $m_{h'}<m_e$ (left panel) and $m_{h'}>m_e$ (right panel). The colored regions are excluded by COMPTEL, INTEGRAL/SPI, and the 2016 COSI balloon flight (green, orange, and blue regions). The regions above the magenta and pink curves are excluded by s-wave and p-wave CMB limits, respectively. The regions above the dashed and solid blue curves will be probed by COSI (line and continuum gamma-ray searches, respectively) and the regions above the red curve will be probed by AMEGO-X. The vertical lines are the thresholds of the different channels, color-coded as in Fig. \ref{['fig:CrossSections']}.