Space / Stars

More than a decade ago, Fabian and his colleagues discovered a way of seeing into this shimmering cloud, almost to the edge of its black heart; and it was this strange, subtle feature of the X-ray spectrum that the Wilms group went looking for. Some of the X rays from the corona, the Cambridge researchers realized, would shine back onto the accretion disk and excite iron atoms there. And some of those iron atoms would thereupon fluoresce, emitting X rays of their own—not over the whole band this time but at a single precise line in the energy spectrum: 6.4 kilo-electron volts, which is the energy an electron loses when it falls from one shell in an iron atom to a lower one.

An "emission line" like that in the hands of an astronomer is like a radar gun in the hands of a cop: It reveals how fast the X-ray-emitting iron atoms are traveling. Because the iron atoms in MCG-6-30-15 are moving, astronomers don't see the line right at 6.4 kilo-electron volts. Instead the X rays are Doppler-shifted, like a radar beam bouncing off a speeding car (the radar waves hit the gun more often if the car is moving toward the gun and less often if the car is moving away from the gun). The X rays are thus shifted toward the blue side of the spectrum, or "blue-shifted" to higher energies, and also intensified on the side of the accretion disk that is moving toward Earth; they are "red-shifted" to lower energies on the side that is moving away. When astronomers record a single spectrum for the whole galaxy, the iron line is smeared in both directions by this Doppler effect. At the same time it is also gravitationally red-shifted, because some of the iron atoms are very close to the black hole, where time itself and thus all light waves are stretched.

The net result is that the sharp emission line is smeared into a broad, asymmetrical hump—and the broader the hump, the faster the iron must be moving and the closer it must be to the black hole. Astronomer Andrew Fabian predicted all this in 1989. In 1994, working with Japanese researchers and the Japanese X-ray satellite ASCA, he found evidence for a broad iron line in MCG-6-30-15. Wilms and his colleagues hoped for more conclusive results with the more sensitive XMM-Newton.

"And what we saw right from the beginning was that the iron line was wrong," says Wilms. "It was much broader than what we thought it should be." The initial excitement was followed by fretting about whether they understood their own telescope. "Almost monthly, we'd have these panic attacks where someone would raise a calibration problem," recalls Chris Reynolds of the University of Maryland in College Park, who worked with Fabian on the earlier study and with Wilms on this one. "We'd have to do the whole analysis again."

The analysis consisted in building, brick by theoretical brick, a model of MCG-6-30-15 that would explain the data they got from the real galaxy. The researchers started with a model that included only continuum X rays from the corona; they found that it produced too many X rays at low energies and not enough at high ones. They added a cloud of warm haze a few light-years from the black hole to absorb some of those low-energy X rays. (That haze really seems to exist; it's what makes MCG-6-30-15 look dull in visible light.) They supplemented the X rays coming directly from the corona with ones that had reflected first off the disk—and found they were still coming up short. Finally, they added a fluorescent iron line, amazingly bright and red-shifted so strongly that it had to be coming from iron atoms streaking just over the event horizon at near light speed. Bingo.

"It was almost like a blazing ring right around the black hole," says Reynolds.

For the iron atoms to get that bright so close to the black hole in MCG-6-30-15, the hole has to be rotating rapidly. By dragging space-time around with it, a rotating hole allows gas to orbit closer to the event horizon without falling in. And if the iron atoms are fluorescing that brightly, it means something is wrong with the standard model of black-hole accretion disks. In that view the disk is lit up by only gravitational energy, which is converted into heat and light by means of friction. But it's hard to generate blazing rings that way. "The gravitational energy is released gradually, so the glowing region of the accretion disk is fairly extended," says Reynolds.

"There's no way you can produce more energy, say, by throwing the stuff down the black hole faster," says Wilms. "You really need some other mechanism."

"If what we're seeing is what we think we're seeing," says Mitch Begelman, "then it's very significant."
The new mechanism for getting energy out of a black hole is not really new. Roger Blandford and Roman Znajek of Cambridge University proposed it in 1977. And the reason you can get energy out of a black hole, that swallower of all things, is that the energy you detect never really got into the black hole to begin with—it's associated with the space-time whirlpool created outside the event horizon by the black hole's rotation.