Rejection of the binary broad-line region interpretation of double-peaked emission lines in three active galactic nuclei

It has been suggested that the peculiar double-peaked Balmer lines of certain broad-line radio galaxies come from individual broad-line regions associated with the black holes of a supermassive binary. We continue to search for evidence of the radial velocity variations characteristics of a double-lined spectroscopic binary that are required in such a model. After spectroscopic monitoring of three suitable candidates (Arp 102B, 3C 390.3, and 3C 332) spanning two decades, we find no such long-term systematic changes in radial velocity. A trend noticed by Gaskell in one of the Balmer-line peaks of 3C 390.3 before 1988 did not continue after that year, invalidating his inferred orbital period and mass. Instead, we find lower limits on the plausible orbital periods that would require the assumed supermassive binaries in all three objects to have total masses in excess of 10¹⁰ M_⊙. In the case of 3C 390.3, the total binary mass must exceed 10¹¹ M_⊙to satisfy additional observational constraints on the inclination angle. We argue that such large binary black hole masses are difficult to reconcile with other observations and with theory. In addition, there are peculiar properties of the line profiles and flux ratios in these objects that are not explained by ordinary broad-line region cloud models. We therefore doubt that the doublepeaked line profiles of Arp 102B, 3C 390.3, and 3C 332 arise in pairs of broad-line regions. Rather, they are much more likely to be intimately associated with single black holes. The recent discoveries of transient but otherwise similar double-peaked emission lines in nearby active galactic nuclei bolster the view that double-peaked emission lines are commonly produced by a single compact source.