Repeated mergers of black hole binaries: implications for GW190521

TL;DR

This paper investigates the formation of the GW190521 black hole merger through hierarchical mergers in globular clusters, using N-body simulations to assess the likelihood and implications of such pathways for gravitational wave observations.

Contribution

It introduces a dynamical formation scenario for GW190521 involving multiple black hole mergers in globular clusters, supported by extensive N-body simulation analysis.

Findings

Hierarchical mergers can produce IMBHs consistent with GW190521.

Probability of IMBH formation via mergers in globular clusters is 1-10%.

Recoil effects influence the likelihood of hierarchical black hole mergers.

Abstract

The gravitational wave event GW190521 involves the merger of two black holes of $\sim 85\text{M}_\odot$ and $\sim 66\text{M}_\odot$ forming an intermediate-mass black hole (IMBH) of mass $\sim 142\text{M}_\odot$. Both progenitors are challenging to explain within standard stellar evolution as they are within the upper black-hole mass gap. We propose a dynamical formation pathway for this IMBH based on multiple mergers in the core of a globular cluster. We identify such scenarios from analysis of a set of 58 N-body simulations using \texttt{NBODY6-gpu}. In one of our simulations, we observe a stellar black hole undergoing a chain of seven binary mergers within 6 Gyr, attaining a final mass of $97.8\text{M}_\odot$. We discuss the dynamical interactions that lead to the final IMBH product, as well as the evolution of the black hole population in that simulation. We explore statistically the effects of gravitational recoil on the viability of such hierarchical mergers. From the analysis of all 58 simulations we observe additional smaller chains, tentatively inferring that an IMBH formation through hierarchical mergers is expected in the lifetime of a median mass globular cluster with probability $0.01 \lesssim p \lesssim 0.1$ without gravitational merger recoil. Using this order-of-magnitude estimate we show our results are broadly consistent with the rate implied by GW190521, assuming that gravitational recoil ejection of progenitors has a low probability. We discuss implications for future gravitational wave detections, emphasising the importance of studying such formation pathways for BHs within the upper mass gap as a means to constrain such modelling.