2023MNRAS.526.6031V

Despite the increasing number of gravitational wave (GW) detections, the astrophysical origin of binary black hole (BBH) mergers remains elusive. A promising formation channel for BBHs is inside accretion discs around supermassive black holes, that power active galactic nuclei (AGN). In this paper, we test for the first time the spatial correlation between observed GW events and AGN. To this end, we assemble all sky catalogues with 1,412 (242) AGN with a bolometric luminosity greater than 10^45.5erg s^–1 ($10^{46}\, {\rm erg\, s}^{-1}$) with spectroscopic redshift of z ≤ 0.3 from the Milliquas catalogue, version 7.7b. These AGN are cross-matched with localization volumes of BBH mergers observed in the same redshift range by the LIGO and Virgo interferometers during their first three observing runs. We find that the fraction of the detected mergers originated in AGN brighter than $10^{45.5}\, {\rm erg\, s}^{-1}$ ($10^{46}\, {\rm erg\, s}^{-1}$) cannot be higher than 0.49 (0.17) at a 95 per cent credibility level. Our upper limits imply a limited BBH merger production efficiency of the brightest AGN, while most or all GW events may still come from lower luminosity ones. Alternatively, the AGN formation path for merging stellar-mass BBHs may be actually overall subdominant in the local Universe. To our knowledge, ours are the first observational constraints on the fractional contribution of the AGN channel to the observed BBH mergers.