Why Are Galaxies without Black Holes Uncommon? (Forwarded)

Office of University Communications
University of Maryland

Contacts:
Lee Tune, 301 405 4679

For Immediate Release: May 30, 2007

Why Are Galaxies without Black Holes Uncommon?

The Answer is Spin Alignment, Say UM researchers

COLLEGE PARK, Md. -- Recent calculations indicate that when two galaxies,
and the supermassive black holes that lie at their centers, merge, these
galactic "marriages" frequently produce gravitational forces strong enough
to kick the new combined black hole right out of its merged galaxy. However,
so far, none of the many "empty nest" galaxies predicted by such
calculations have been found.

Now researchers at the University of Maryland say merged black holes
probably are kicked out far less often than predicted because torques from
the vast accretion disk of parental galactic material that spins around and
feeds merging black holes act to align their spins in a way that reduces the
kick force.

In findings presented today during a press briefing at the American
Astronomical Society's meeting in Honolulu Hawaii, Maryland astronomers
Tamara Bogdanovic and Christopher Reynolds propose that in the majority of
gas-rich galactic mergers, torques from gas accretion align the spins of
supermassive black holes and their orbital axis with a large-scale gas disk.
This mechanism, they say, helps explain the ubiquity of black holes at the
center of galaxies despite the potentially large kicks from gravitational
radiation recoil.

"While we expect a black hole ejection to be uncommon in the aftermath of
gas-rich mergers, it is still possible that it may happen, especially in
merging galaxies that are relatively gas-poor. Future observations of such
gas-poor mergers may point to a class of massive galaxies without a central
supermassive black hole," said Bogdanovic.

"What we have uncovered here is a remarkable interaction between the
galactic scale gas disk and the comparatively tiny black holes," added
Reynolds.

The Power of Spin

It has recently been shown using relativistic calculations that when
galaxies -- and the supermassive black holes at their center -- merge, a
large pulse of gravitational radiation is produced. This radiation pulse,
essentially ripples in the fabric of space itself, can kick the final black
hole up to speeds of 3000 km/s. But these large kicks require the black
holes to be rapidly spinning, and for the spin axes to be oppositely
directed and tilted over into the plane of the orbit. The difficulty this
poses, say Bogdanovic and Reynolds is that the escape speed from most
galaxies is less that 1000 km/s. Therefore, if large recoil speeds are
typical, one might expect that many galaxies that have undergone major
mergers would be without a black hole.

Current estimates are that over the past 6 billion years, about a half of
all galaxies have undergone mergers. And yet, in contradiction to what the
models predict, all observed galaxies with bulges indicative of mergers
appear to have central supermassive black holes. "It therefore seems that
there is an astrophysical avoidance of the types of supermassive black hole
coalescences that would lead to kicks beyond galactic escape speeds," write
Bogdanovic, Reynolds and fellow Maryland astronomer Coleman Miller in a
paper (to be published in the June 1 2007 issue of Astrophysical Journal
Letters) explaining their proposal.

The lower than expected kicks can be explained if (1) the spins of merging
black holes are all small, (2) there is a large difference in mass between
most merging black holes or (3) the spins tend to align with each other and
be oriented perpendicular to the orbit of the pair of black holes.

Observations do not favor the low spin mechanism and evidence suggests
strongly that coalescence of comparable-mass black holes (which would enable
high kick out speeds) should be common. "The most likely solution therefore
seems to be that astrophysical processes tend to align the spins of
supermassive black holes with the orbital axis," write Bogdanovic, Reynolds
and Miller.

This work was supported by the University of Maryland, College Park -
Astronomy Center for Theory and Computation Prize Fellowship program and by
funding from the National Science Foundation.