2016ApJ...817...42L

One of the proposed explanations for the broad, double-peaked Balmer emission lines observed in the spectra of some active galactic nuclei (AGNs) is that they are associated with sub-parsec supermassive black hole (SMBH) binaries. Here, we test the binary broad-line region hypothesis through several decades of monitoring of the velocity structure of double-peaked Hα emission lines in 13 low-redshift, mostly radio-loud AGNs. This is a much larger set of objects compared to an earlier test by Eracleous et al. and we use much longer time series for the three objects studied in that paper. Although systematic changes in radial velocity can be traced in many of their lines, they are demonstrably not like those of a spectroscopic binary in a circular orbit. Any spectroscopic binary period must therefore be much longer than the span of the monitoring (assuming a circular orbit), which in turn would require black hole masses that exceed by 1-2 orders of magnitude the values obtained for these objects using techniques such as reverberation mapping and stellar velocity dispersion. Moreover, the response of the double-peaked Balmer line profiles to fluctuations of the ionizing continuum and the shape of the Lyα profiles are incompatible with an SMBH binary. The binary broad-line region hypothesis is therefore disfavored. Other processes evidently shape these line profiles and cause the long-term velocity variations of the double peaks. \subject_ section 069 Stars