Table of Contents
Fetching ...

Little Red Dots from Ultra-Strongly Self-Interacting Dark Matter

M. Grant Roberts, Lila Braff, Aarna Garg, Stefano Profumo, Tesla Jeltema

Abstract

We investigate the possibility that the recently identified population of high-redshift, obscured quasars - known as "Little Red Dots" (LRDs) - originates from early black hole seed formation driven by ultra-strongly self-interacting dark matter (uSIDM). In this framework, dark matter halos undergo gravothermal core collapse due to large self-interaction cross sections, resulting in the rapid formation of massive black hole (BH) seeds with masses $\gtrsim 10^{5} M_\odot$ at redshifts $z \gtrsim 5$. We develop a semi-analytic model that tracks the evolution of the dark matter halo population, the redshift of collapse $z_{\rm coll}$, and the corresponding BH mass function. Black hole growth is modeled stochastically via a log-normal Eddington ratio distribution and a finite duty cycle. We find that the uSIDM scenario naturally reproduces key observed properties of LRDs, including their abundance, compactness, and characteristic BH masses, while offering a mechanism for early, obscured black hole formation that is difficult to achieve in standard CDM-based models. The predicted SMBH mass function at $z \sim 5$ shows excellent agreement with LRD observational data and SIDM merger-tree simulations, particularly at the high-mass end $(m_{\rm BH} \gtrsim 10^{7} M_\odot)$. These results suggest that LRDs may serve as powerful observational tracers of exotic dark sector physics and that SMBH formation in the early universe could be significantly shaped by non-gravitational dark matter interactions.

Little Red Dots from Ultra-Strongly Self-Interacting Dark Matter

Abstract

We investigate the possibility that the recently identified population of high-redshift, obscured quasars - known as "Little Red Dots" (LRDs) - originates from early black hole seed formation driven by ultra-strongly self-interacting dark matter (uSIDM). In this framework, dark matter halos undergo gravothermal core collapse due to large self-interaction cross sections, resulting in the rapid formation of massive black hole (BH) seeds with masses at redshifts . We develop a semi-analytic model that tracks the evolution of the dark matter halo population, the redshift of collapse , and the corresponding BH mass function. Black hole growth is modeled stochastically via a log-normal Eddington ratio distribution and a finite duty cycle. We find that the uSIDM scenario naturally reproduces key observed properties of LRDs, including their abundance, compactness, and characteristic BH masses, while offering a mechanism for early, obscured black hole formation that is difficult to achieve in standard CDM-based models. The predicted SMBH mass function at shows excellent agreement with LRD observational data and SIDM merger-tree simulations, particularly at the high-mass end . These results suggest that LRDs may serve as powerful observational tracers of exotic dark sector physics and that SMBH formation in the early universe could be significantly shaped by non-gravitational dark matter interactions.

Paper Structure

This paper contains 13 sections, 14 equations, 2 figures.

Figures (2)

  • Figure 1: We show the uSIDM LRD mass function (colored dots) for different uSIDM parameters taken from Fig. 3 in Roberts_uSIDM; we also plot the Poisson error for our uSIDM LRD mass function in each mass bin. The uSIDM results are plotted against the SIDM simulation results from fangzhou_LRD (shaded red), and the LRD mass function data points from Kokorev_2024 (red diamonds). The red diamonds are the inferred LRD black hole masses assuming Eddington accretion while the error bars show the range of black hole masses for accretion down to $\lambda=0.1$. We also note that the red shaded region from fangzhou_LRD assumes all of the black holes in their simulations are active, while we assume a duty cycle of 1% based on ref. duty_cycle_z=5.
  • Figure 2: We show the resulting distributions from fitting a power-law to 1000 realizations of the uSIDM parameter distribution from ref. Roberts_uSIDM (left). And we over plot the median and $\pm 1\sigma$ bands from the power-law realizations onto Fig. \ref{['fig:uSIDM-LRD-mass-function']} (right).