Constraints on Kerr-Newman black holes from merger-ringdown gravitational-wave observations

We construct a template to model the postmerger phase of a binary black hole coalescence in the presence of a remnant $U(1)$ charge. We include the quasinormal modes typically dominant during a binary black hole coalescence, $(\ensuremath{\ell},m,n)={(2,2,0),(2,2,1)}$ and also present analytical fits for the quasinormal mode frequencies of a Kerr-Newman black hole in terms of its spin and charge, here also including the (3, 3, 0) mode. Aside from astrophysical electric charge, our template can accommodate extensions of the Standard Model, such as a dark photon. Applying the model to LIGO-Virgo detections, we find that we are unable to distinguish between the charged and uncharged hypotheses from a purely postmerger analysis of the current events. However, restricting the mass and spin to values compatible with the analysis of the full signal, we obtain a 90th percentile bound $\overline{q}<0.33$ on the black hole charge-to-mass ratio, for the most favorable case of GW150914. Under similar assumptions, by simulating a typical loud signal observed by the LIGO-Virgo network at its design sensitivity, we assess that this model can provide a robust measurement of the charge-to-mass ratio only for values $\overline{q}\ensuremath{\gtrsim}0.5$; here we also assume that the mode amplitudes are similar to the uncharged case in creating our simulated signal. Lower values, down to $\overline{q}\ensuremath{\sim}0.3$, could instead be detected when evaluating the consistency of the premerger and postmerger emission.