Detecting the gravitational wave background from primordial black hole dark matter

TL;DR

This paper models the stochastic gravitational-wave background from primordial black hole dark matter, showing it could be detected by future space interferometers and pulsar timing arrays, and distinguishes it from other sources.

Contribution

It provides a comprehensive calculation of the gravitational-wave background from PBH dark matter, considering effects like eccentricity, velocities, mass distribution, and clustering, and discusses detection prospects.

Findings

LISA can potentially detect the PBH gravitational-wave background.

Current PTA constraints exclude broad-mass PBH models over three decades.

The gravitational wave spectrum is enhanced by the PBH mass distribution width.

Abstract

The black hole merging rates inferred after the gravitational-wave detection by Advanced LIGO/VIRGO and the relatively high mass of the progenitors are consistent with models of dark matter made of massive primordial black holes (PBH). PBH binaries emit gravitational waves in a broad range of frequencies that will be probed by future space interferometers (LISA) and pulsar timing arrays (PTA). The amplitude of the stochastic gravitational-wave background expected for PBH dark matter is calculated taking into account various effects such as initial eccentricity of binaries, PBH velocities, mass distribution and clustering. It allows a detection by the LISA space interferometer, and possibly by the PTA of the SKA radio-telescope. Interestingly, one can distinguish this background from the one of non-primordial massive binaries through a specific frequency dependence, resulting from the maximal impact parameter of binaries formed by PBH capture, depending on the PBH velocity distribution and their clustering properties. Moreover, we find that the gravitational wave spectrum is boosted by the width of PBH mass distribution, compared with that of the monochromatic spectrum. The current PTA constraints already rule out broad-mass PBH models covering more than three decades of masses, but evading the microlensing and CMB constraints due to clustering.