Multimessenger Constraints on Supermassive Dark Stars and Their Black Hole Remnants

TL;DR

This work addresses whether DM-annihilation-powered supermassive dark stars (DSs) could seed the first supermassive black holes and how their cosmological population would imprint diffuse electromagnetic backgrounds. It develops a population-level model linking DS formation in early DM halos, DS evolution with DM heating, adiabatic contraction halos, and post-collapse BH-DM spikes, then computes the resulting diffuse photon and neutrino signals across redshift, including attenuation. The main finding is that, for thermal-relic DM, external-halo and BH-spike annihilations can exceed the Fermi-LAT extragalactic gamma-ray background for $m_\chi$ below about 1 TeV (hadronic/electroweak channels) or below about 100 GeV (leptonic channels), with BH-spike uncertainties dominating at higher masses. The results provide the first population-integrated multimessenger constraints on DSs as SMBH progenitors and demonstrate that diffuse backgrounds offer a powerful complement to JWST-era searches for DM-related early structure formation.

Abstract

Dark matter (DM) annihilation can power the first generation of stars as long lived dark stars (DSs) that grow to supermassive scales $M_{\rm DS}\gtrsim 10^{5} M_{\odot}$ and eventually collapse into heavy black holes that could seed the supermassive black holes observed at high redshifts. We compute the diffuse electromagnetic emission from a cosmological population of such supermassive DSs and their black hole remnants, tracking the entire DS history and including thermal surface radiation, DM annihilation in adiabatically contracted halos as well as late-time emission from DM overdensity spikes around the resulting black holes. After accounting for photon attenuation, we find that DS related contributions can exceed the Fermi-LAT extragalactic $γ$-ray background for thermal relic annihilation cross-sections and DM masses below $\sim 1$ TeV. Our results constitute the first population integrated diffuse multimessenger constraints on supermassive DSs as progenitors of early black holes and demonstrate that diffuse photon and neutrino backgrounds offer a powerful and complementary avenue for probing the role of DM in the formation of the earliest massive structures.

Figures (18)

• Figure 1: Illustration of the considered electromagnetic emission components. A supermassive DS forming at high redshift $z \sim 25$ emits thermal radiation from its surface and is embedded in an adiabatically contracted DM overdensity, where DM annihilations outside the DS generate an additional non-thermal component. At a characteristic redshift $z_{\rm lim} \sim 15$ (vertical dashed line), the DS exhausts its DM fuel and gravitationally collapses into a SMBH. The resulting SMBH is surrounded by a DM density spike from the progenitor halo, with DM annihilations in the spike producing a non-thermal emission that persists from the time of collapse to low redshifts $z\to0$.
• Figure 2: Photon spectra per DM annihilation for the channels $b\bar{b}$, $W^+W^-$, $\mu^+\mu^-$, and $e^+e^-$. Dashed curves correspond to a DM mass $m_\chi = 100~\mathrm{GeV}$, while solid curves correspond to $m_\chi = 10~\mathrm{TeV}$. For $m_\chi\gg m_W$ internal bremsstrahlung leads to a final state radiation enhancement near the kinematic endpoint in the $W^+W^-$ channel Bergstrom:2008gr. Computed using data of Ref. Cirelli:2010xx.
• Figure 3: DM density profile of the spike surrounding a BH of mass $M_{\rm BH}=10^{6} M_\odot$ embedded in a $M_h = 10^{8} M_\odot$ halo. The solid curve denotes the power-law spike $\rho_{\rm sp}\propto r^{-\gamma_{\rm sp}}$, while the green horizontal line marks the annihilation plateau at $\rho_{\rm max}=m_\chi/(\langle\sigma v\rangle,t_{\rm BH})$, where $t_{\rm BH}\simeq1.3\times10^{10}~$yr is the time elapsed since collapse at $z_{\rm lim}=15$.
• Figure 4: Luminosity evolution of a DS forming at $z_{\rm form}=25$ and collapsing at $z_{\lim}=15$. Left: Benchmark model with halo mass $M_h = 10^8 M_\odot$, DM mass $m_\chi = 100$ GeV, and annihilation channel $\chi\chi \to b\bar{b}$, assuming a thermal relic cross section $\langle\sigma v\rangle = 3\times10^{-26} {\rm cm}^3 {\rm s}^{-1}$. The DS surface luminosity during its accreting phase (purple), the luminosity from DM annihilation in the DS-induced contracted halo (red), and the post-collapse luminosity from the BH DM spike (green) are shown, together with a conservative possible limit imposed by $\rho_{\max}$ (light blue). Right: Same as the left panel, but for $M_h = 10^6 M_\odot$, $m_\chi = 10$ TeV, considering leptonic DM annihilation channel $\chi\chi \to e^+e^-$.
• Figure 5: Predicted diffuse thermal fluxes from a cosmological population of supermassive DSs, shown for representative DM masses $m_\chi=100$ GeV and $m_\chi=10$ TeV, assuming the benchmark annihilation channel $\chi\chi \to b\bar{b}$ and the reference relic cross-section $\langle\sigma v\rangle = 3\times10^{-26}~\mathrm{cm^3~s^{-1}}$. Observational data from JWST/PEARLS Windhorst:2022rrv, $\gamma$-ray constraints Greaux:2024wyc, New Horizons LORRI measurements Postman:2024erl, ESO VLT/FORS dark cloud observations Mattila:2017zrp, CIBER near-infrared background data Matsuura:2017lub, and integrated galaxy light (IGL) estimates from Driver:2016krv and Koushan2021 are shown. The predicted supermassive DS emission peaks in the near-infrared and remains below current observational limits.