A tale of two redshifts
How did 3C 186 eject its supermassive black hole? With gravitational waves.
The past three years of gravitational wave astronomy have seen a number of mergers in binary systems by the LIGO and Virgo detectors. Most of them were binary stellar-mass black holes, each with components between 7 and 35 times the mass of the Sun. While ground-based interferometers currently in use can detect gravitational waves in the 10–1000 Hz range, which includes these mergers, other sources lie beyond this regime. In particular, supermassive black hole binaries should produce waves in the nanohertz range, detectable by advanced instruments like pulsar timing arrays, or space-based observatories like LISA.
While no supermassive black hole binaries have been observed merging, several close pairs have been identified, like PKS 1302–102. These candidate systems might have components separated by anywhere from a couple dozen parsecs down to a tenth of a parsec. We can learn about the evolution of close binary supermassive black holes by studying these merger progenitors or by observing their counterparts — systems that have already merged!
Since two merged supermassive black holes simply create a single, more massive, black hole, how do you tell these merger remnants from normal solitary supermassive black holes? Today’s blog post is about 3C 186, an active galactic nucleus (or AGN) that just might be what we’re looking for. What makes it stand out? It’s been ejected from its host galaxy by gravitational waves.
The observations: Imaging and spectroscopy
For some years, a small population of active galactic nuclei were known to be candidate ejected supermassive black holes. Some, like NGC 3718, are spatially offset from the centers of their host galaxies, while others show redshifts distinct from their hosts, indicating that they’re moving away from the cores. However, no AGN had been found that displayed both spatial and velocity offsets — definitive evidence of an ejection.
In 2017, a team of astronomers (Chiaberge et al. 2017) studied 3C 186 using the Hubble Space Telescope, along with data from the Sloan Digital Sky Survey. They looked at both images and spectra of the source, at optical and UV wavelengths. To confirm the spatial offset, they modeled the galaxy’s surface brightness with something called a Sersic profile, a common model for fitting images of galaxies, while modeling the AGN emission via a point spread function, which determines how a source is blurred by an imaging system.
Their image analysis showed that the AGN was separated from the host galaxy’s core by an astonishing 11 kiloparsecs — a distance greater than that between the Sun and the center of the Milky Way, and in fact about one-third the size of the Milky Way’s disk. This is a significant offset, and showed that the supermassive black hole was definitely outside the galactic nucleus. The team followed up by analyzing the spectroscopic data, taking advantage of two sections of the structure of an AGN: the narrow line region and the broad line region.
The narrow line region exists several hundred parsecs from the center of an AGN, and consists of slow-moving gas clouds. Lower velocities mean less Doppler broadening — hence the name. The broad line region, on the other hand, has a radius on the order of a light-year, and holds fast-moving gas clouds with stronger line broadening. Emission from the broad line region is likely connected to the motion of the quasar, while the narrow line region may be more related to the motion of the host galaxy.
The team chose to measure the host galaxy’s redshift using the [O II] and [Ne III] lines from the narrow line region, and found a value of z=1.068. They used Lyman-alpha, C IV, C III] and Mg II lines to study the broad line region, deriving a redshift of z=1.054, corresponding to a velocity difference of -2140 km/s from the host. This is no small discrepancy, and is indeed confirmation that the AGN was moving away from the galaxy at a significant speed.
Are the AGN and galaxy related?
As you might expect, there are a small number of scenarios that could explain both offsets. One set of possibilities is that the AGN is simply distinct from the galaxy — in other words, that the system is a chance alignment of two completely unrelated objects. From absorption in the AGN’s spectrum by the galaxy, it is clear that the AGN cannot be a foreground object. While it is still possible that the nucleus is behind the galaxy, this would require a second AGN to produce the narrow lines associated with the host, but no separate narrow lines from the known AGN were detected, so there is only one set of broad and narrow line complexes.
Another situation the astronomers considered was a related one, where there are multiple AGNs. However, data from the Chandra X-ray Observatory precluded the existence of even a partially-shrouded second AGN, and the luminosity of 3C 186 is enough to explain the observed narrow line emission. A much simpler model is the one the team favors: an offset supermassive black hole moving at a great speed away from the galactic nucleus.
The mechanism behind the ejection is simple. The host galaxy shows signs of a past merger with another galaxy — blobs identified as either tidal tails or supernova shells. (Galaxy interactions often trigger star formation, which in turn leads to supernovae, so even evidence of a recent starburst could provide supporting evidence for the theory.) Assuming both galaxies possessed supermassive black holes, the two would have merged on timescales of 1–2 billion years, producing a merger remnant of about 3 billion solar masses about 5 million years ago.
However, the merger would not necessarily have been symmetric. Anisotropic emission of gravitational waves could have “kicked” the resulting supermassive black hole in a certain direction, akin to the pulsar kicks seen in some neutron stars. We don’t have a large enough sample size to say whether this is unusual or not, but it’s definitely a plausible model — possibly the best one for 3C 186.
The team hopes to put further constraints on other models for the system. Better Hubble imaging could rule out a distinct low-surface-brightness host for the AGN, and more detailed spectroscopy could provide additional information about the narrow and broad line regions. Finally, of course, any detection of gravitational waves from such a merger remnant would be something of a holy grail for analyzing these systems — but those observations are probably a decade or two in the future.
