If quasars form from primordial black holes

TL;DR

This paper proposes that supermassive black holes originate from primordial black holes formed before the radiation-dominated era and grow via accretion of dwarf-galaxy material. A quantitative fueling model links quasar luminosity density and the quasar luminosity function to a PBH mass-function IMF with equal numbers per mass decade, a DM PBH fraction $f$, and a collision/feeding boost factor $b$, constrained by CMB distortion limits. The resulting quasar luminosity function is Schechter-like and, when normalized at $L_{46}$, agrees with observed data; the predicted density evolution scales as approximately $(1+z)^{-3}$, consistent with observations, and the framework allows quasars to function as standard candles under certain assumptions. The model additionally implies the existence of intermediate-mass black holes and offers a unified view of quasars and radio galaxies as different accretion phases triggered by the same galaxy interactions, with testable predictions for JWST and future surveys, as well as potential imprints in the CMB and primordial helium abundances.

Abstract

We explore the consequences of a novel but increasingly well-supported hypothesis that supermassive black holes may have formed from primordial black holes form ed prior to, and rapidly growing in, the radiation-dominated universe. We show that this hypothesis can predict the luminosity of quasars and their luminosity distribution. With reasonable values of the parameters introduced, these predictions are borne out by observations. The model predicts density evolution in accordance with observations. If the same galaxy interaction rate creates quasars and radio galaxies, whose primordial black hole nuclei seem somewhat less massive, their relative number densities reflect relative lifetimes in these states.

Figures (2)

• Figure 1: Prediction of the QSO bolometric luminosity function in $\S$3 in black and the double power law fit to the data at z = 1 by Shen et al. (2020) in green. The model is normalised at L46 for appropriate values of the parameters in equation (5). In red is the prediction for PBH seeded in the RE (Musco 2019), rather than during inflation, which does not fit the data. The crosses on the right mark the density evolution from z = 1 to 6 at constant L46 luminosity.
• Figure 2: The variation of number density at L46 with redshift in the data of Shen et al. (2020). The x-axis $\sim$ (1+z), but bridges the matter dominated and the dark energy dominated eras in a $\Lambda$CDM universe. The density evolution model of $\S$2 predicts a slope like the dashed line. The red dot with Poisson error bars is from Feeney, Kavanagh & Regan (2024 ), normalized to the redshift 6 bin and corrected to L46, assuming those with z $>$ 8 are mean luminosity, also assuming that selection effects are independent of redshift.