A Comprehensive Study of WIMP Models Explaining the Fermi-LAT Galactic Center Excess

Abstract

The Galactic Center excess (GCE) of GeV $γ$ rays may hint at dark matter (DM), yet its origin remains debated. Motivated by this, we survey weakly interacting massive particle (WIMP) models that can fit the GCE while satisfying relic-density, direct-detection (DD), and indirect-detection (ID) bounds. We group candidates into hadronic (Higgs portals; simplified scalar/vector mediators), leptonic ($U(1)_{L_i-L_j}$), and mixed ($U(1)_{B-L}$, $Z$-portal) classes. Across all cases, present DD and dwarf-spheroidal $γ$-ray limits exclude wide regions, leaving mainly narrow resonant funnels with $m_{\rm DM}\!\simeq\! m_{\rm med}/2$ and portal couplings $\ll 1$. In hadronic setups, scalar and vector Higgs portals survive only in a thin strip near $m_h/2\simeq62.5$ GeV with portal couplings $\sim 10^{-4}$, while the Dirac Higgs and $Z$ portals are essentially excluded. The UV-complete vector Higgs portal retains resonant bands whose viable portal strength depends on the mixing angle. Simplified scalars allow small windows for complex-scalar or vector DM; Dirac DM is strongly disfavored, whereas a pseudoscalar with Dirac DM remains viable over a broader parameter range. For a simplified $Z'$ mediator, a pure vector coupling leaves only a marginal region, while pure axial is excluded by DD/ID bounds. In leptonic scenarios, inverse-Compton emission is essential: $L_μ-L_e$ (and, to a lesser extent, $B\!-\!L$) fits the GCE with near-thermal cross sections, while $L_μ-L_τ$ is disfavored. Overall, viable WIMP explanations are constrained to finely tuned resonant regime, with leptophilic vectors and pseudoscalar portals emerging as the most robust options.

Figures (11)

• Figure 1: Constraints in the $(m_{\rm DM},\lambda_{\rm portal})$ plane for three Higgs portal models. The blue dashed line indicates the parameters for which $\Omega_{\rm DM} h^2=0.12$. The orange dot-dashed line labeled and the red dotted line labeled show SI DD and ID upper limits. The gray band marks the region favored by the GCE flux fit with $\Delta\chi^2\le 9.21$ (2 d.o.f.) relative to the best fit, corresponding to $\approx99\%$ C.L.
• Figure 2: Best–fit of the scalar Higgs portal to the Cholis+22 GCE datasets described in Section \ref{['sec:results_Higgs']}. We show the GCE data and the theoretical predictions for the DM prompt flux (blue dashed curve).
• Figure 3: Constraints in the $(m_{\rm DM},\,\lambda_{\rm portal})$ plane for the UV-complete Higgs-portal model under two parameter settings. In both cases we fix $m_{Hp}=80~\mathrm{GeV}$ and take the mixing to be $\tan{2\alpha}\equiv tg2=0.1$ or $0.45$. The curves and shaded bands follow the same definitions as in Fig. \ref{['fig:higgs_portals']}.
• Figure 4: Constraints in the $(m_{\rm DM},\,\lambda_{\rm portal})$ plane for the UV-complete Higgs-portal model under two parameter settings. Same as Fig. \ref{['fig:UV_VHP_MHp_80']}, except that $m_{Hp}$ is set to $300~\mathrm{GeV}$.
• Figure 5: Constraints in the $(m_{\rm DM},\,\lambda_{\rm portal})$ plane for the $Z$-portal with Dirac dark matter under two coupling choices. The left panel shows the axial coupling, where we adopt SD neutron limits (orange dot–dashed line); the right panel shows the vector coupling, where SI limits are used. The curves and shaded bands follow the same definitions as in Fig. \ref{['fig:higgs_portals']}.