Chandra X-Ray Observatory Examines
                a New Kind of Black Hole

A surprising new class of middleweight black holes, perhaps 10 000 times more massive than the Sun, may have much to teach us about how galaxies form.

For decades, astronomers have been accumulating increasingly strong evidence for the existence of black holes in two distinct mass regimes: "stellar" black holes, weighing a few times the mass of the Sun (Msun), and "supermassive" black holes with masses ranging from 106­109 Msun, always sitting at the centers of galaxies. There seemed to be nothing in between.

But last year, two groups of x-ray astronomers1,2 presented tentative suggestions of something quite new: a middleweight class of black holes much more massive than the stellar black holes but distinctly lighter than the supermassive giants. And now, a British-Japanese-US collaboration, availing itself of the new Chandra X-Ray Observatory, has presented the first strong evidence3,4 of such a middleweight black hole: sitting about 500 light-years off center in the nearby galaxy M82, with a mass of at least 500 Msun. Its upper mass limit--perhaps 105 Msun--is somewhat more problematic.

The formation of stellar black holes is reasonably well understood. They seem, in all cases, to be the remnants of supernova explosions. But the formation of supermassive black holes at the centers of galaxies remains something of a puzzle. It's a chicken-and-egg problem: Do supermassive black holes act as seeds for galaxy formation, or is it the other way around?

The middleweight black holes have taken the theorists by surprise. Perhaps they result from some sort of runaway merger of stars in clusters of particularly large and dense stellar population. The study of such off-center middleweights, it is hoped, will improve our understanding of galaxy formation.

NASA's Chandra x-ray telescope was launched into orbit in July 1999 (see Physics Today, May, page 18). Its name honors the great theoretical astrophysicist Subrahmanyan Chandrasekhar, who died in 1995.

The decisive capability that allowed Chandra to unmask the middleweight black hole in M82 is the extraordinarily fine angular resolution of the new telescope's x-ray optics (see figure 1) and its high-resolution imaging camera. The camera's sub-arcsecond resolution is an order-of-magnitude improvement on the older ASCA1 and ROSAT2 orbiting x-ray telescopes whose data yielded the earlier hints.

High-resolution mirror assembly

Figure 1. High-resolution mirror assembly of the Chandra X-Ray Obervatory at Eastman Kodak before completion. The orbiting telescope's optics employ a nested set of four barrel-shaped mirrors that reflect incident x rays at glancing angles.
Photocredit: EASTMAN KODAK 

ASCA had provided the first hard-x-ray imaging observations of M82. In those data, Andrew Ptak and Richard Griffiths at Carnegie Mellon University called attention to an unusually bright and variable x-ray source within about 10 arcseconds of the galaxy's center.1 With ASCA's angular resolution, they could not determine whether the source might be off center. Nor could they verify that it was a single compact source. But Ptak and Griffiths argued that the x rays were being produced by the accretion of surrounding material onto a black hole with a mass of at least 450 Msun.

Examining the ROSAT and ASCA data from 39 nearby galaxies (not including M82), Edward Colbert and Richard Mushotsky at NASA's Goddard Space Flight Center had reported evidence of middleweight black holes near the centers of three spiral galaxies.2

A closer look at M82
The flamboyant M82, because of its evident star-forming hyperactivity, is classified as a "starburst galaxy." There is no optical evidence of a supermassive black hole at its center. Chandra observed M82 for 10 hours on 28 October 1999, and then for another 5 hours on 20 January.

To study the Chandra observations of the x-ray source flagged by Ptak and Griffiths in the ASCA data, astrophysicists at the universities of Leicester and Kyoto, MIT, and the Harvard­Smithsonian Center for Astrophysics (CFA) formed a collaboration. First of all, they were able to resolve the ASCA source into nine distinct pointlike sources.4 (See figure 2.) By far the brightest of these (indicated by the arrow) increased its x-ray luminosity by a remarkable factor of seven in the three months between the October and January observations. And it was clearly 9 arcseconds away from the galaxy's dynamic center (marked by the green cross). At M82's 12-million light-year distance from us, that corresponds to a separation of at least 500 light-years. And, indeed, none of the other observed x-ray sources sits at the center. (The dynamic center of M82 is determined by radio measurements of the circulation of atomic hydrogen.)

X-ray /pt/vol-53/iss-11/images of the central region of the galaxy

Figure 2. X-ray /pt/vol-53/iss-11/images of the central region of the galaxy M82, recorded by the Chandra telescope's high-resolution camera in October 1999 (left) and January 2000 (right). Each panel is 35 arcseconds wide. The brightest source (indicated by the arrow) is about 500 light-years from the galaxy's dynamic center (green cross). Its x-ray luminosity increased sevenfold in the three months between exposures. 

The luminosity of a radiating celestial source provides a minimum estimate of its mass--the so-called Eddington limit. In the 1920s, Arthur Eddington pointed out that a source cannot maintain itself if its outward radiation pressure exceeds its self-gravitation. The January 2000 luminosity measurement of the brightest point source implies an Eddington lower mass limit of 500­900 Msun.

Could the bright x-ray source be simply a tight cluster of distinct sources rather than a single massive black hole? The sevenfold luminosity increase in three months argues strongly for its being a single, coherent object, as does the fact that the high-resolution x-ray camera sees no evidence of spatial extension. "And if it is a single object more than 500 times heavier than the Sun," says Andrea Prestwich of the collaboration's CFA group, "it could hardly be anything but a black hole."

Furthermore, the bright x-ray source's spectral features, as measured by Chandra's imaging spectrometer in September and December 1999, and then again in March of this year, are strikingly reminiscent of stellar black holes accreting material from binary partners. Between September and December, as the overall luminosity of the source was increasing, its hard multi-keV spectrum developed a pronounced soft x-ray component at photon energies around 1 keV. By March, this soft component was gone again and the source's integrated x-ray luminosity had also fallen from its January high. "Much the same thing happens in the variable x-ray emission from stellar black holes in our own galaxy," explains Martin Ward, who leads the collaboration's British contingent: A soft x-ray component appears when there is a pronounced increase in the rate at which the black hole is swallowing material from the accretion disk that surrounds it. (See Physics Today, April 1997, page 20.) 

"Chandra was crucial in showing us that the source is not a dim supermassive black hole at the center of M82," says Philip Kaaret (CFA). The object's off-center location also provides the best upper limit of its mass. In 1975, theorists Jeremiah Ostriker and Scott Tremaine formulated a relation between the mass of a celestial object and the rate at which "dynamical friction" (random gravitational encounters) would cause it to migrate toward the center of its galaxy. The heaviest migrate fastest. From the M82 x-ray source's off-center position and the galaxy's measured dispersion of stellar velocities, Kaaret and company calculate that the source's mass must be less than 105 Msun if it is nearly as old (1010 years) as the galaxy itself. Anything heavier would long since have settled in at the galaxy's dynamic center. But if the object is only a billion years old, its mass could be as large as 106 Msun.

A cautionary tale
For a time, the Chandra group thought it had been granted an additional argument for an upper mass limit of something like 105 Msun. A Fourier analysis of the short-term variability of the bright point source's x-ray luminosity yielded a prominent peak in the power spectrum, corresponding to a period of about 10 minutes. It looked like a quasiperiodic oscillation of the kind that has been seen, at higher frequencies, in stellar black hole candidates and, at lower frequencies, in active galactic nuclei.

This 10-minute oscillation would have imposed an absolute size limit of 10 light-minutes (2 X 108 km) on the x-ray source. And it would have provided an argument for an upper mass limit that did not depend on assumptions about the source's age. But there was, alas, a nagging problem. The 10-minute peak was suspiciously suggestive of one of the "dither" frequencies to which Chandra is purposely subjected. Dithering is a regular oscillation imposed on the telescope's pointing direction in order to spread the incident x-ray flux over the camera's pixel array and thus avoid degradation.

Although the power spectrum with the spectacular 10-minute peak appeared in early versions of reference 3, the group ultimately convinced itself that this was not an astrophysical phenomenon. The 10-minute oscillation period was, apparently, an artifact due to the interaction of Chandra's dither with an algorithm designed to screen out spurious events.

In any case, the evidence for an accreting intermediate-mass black hole in M82 remains strong. If this new class of exotic objects is as abundant as the Colbert­Mushotsky survey of nearby galaxies suggests,2 they may soon teach us much about how galaxies form. 

Bertram Schwarzschild
1. A. Ptak, R. Griffiths, Astrophys. J. 517, L85 (1999).
2. E. Colbert, R. Mushotsky, Astrophys. J. 519, 89 (1999).
3. P. Kaaret et al.,, M. Not. R. Astron. Soc. (in press).
4. H. Matsumoto et al.,, Astrophys. J. Lett. (in press). 
© 2001 American Institute of Physics