November 10, 2009
Astronomers have uncovered compelling evidence for the existence of at least two classes of black holes: the super-massive variety that lurk at the centers of most galaxies - including the 4 million solar mass behemoth at the heart of the Milky Way - and the solar mass variety that are the product of the evolution of massive stars. This latter group have weighed-in in the range from about 4 - 25 solar masses. But what about the middle ground? Where are the black holes with masses of about a thousand Suns or more? Do they even exist? Theory suggests that the first generation of star formation in the early universe could produce black holes of several hundred solar masses, and that, due to mergers of these "primordial" holes, there should have been a whole spectrum of masses produced that would populate the present black hole "desert" in the range from 100 - 10,000 solar masses. Still other thinking suggests that such black holes could form in the present day universe in very dense star clusters. If either of these ideas are valid, then where are these black holes hiding, and how can we find and weigh them?
The answer to this puzzle may tie in with a current conundrum in observational high energy astrophysics. The question concerns the nature of a population of extremely bright X-ray sources found in nearby galaxies. These objects have been dubbed Ultra-luminous X-ray sources, or ULXs for short, because of their extreme luminosities. These objects have also shown short time-scale variability of the kind that can only be produced by a physically compact source, like a black hole. Because of this and other arguments, many astronomers believe they are black holes, but the question is how massive are they? Because in some cases they can be up to 100 times brighter than the stellar black holes we observe in the Milky Way, one suspicion is that some of them could be the missing middle-weight black holes predicted by theory.
Weighing black holes is no simple proposition. For black holes in the Galaxy the most reliable method is to observe the motion of its binary companion star. By observing its orbital motion astronomers can use Kepler's laws to deduce the mass of the black hole. For the ULXs in nearby galaxies this proposition is extremely daunting because the companion stars are so very faint, other means must be devised. Tod Strohmayer and Richard Mushotzky have been exploring new ways to estimate the masses of ULXs by measuring their X-ray timing properties and comparing them with the population of Galactic black holes of known mass. Many black holes in the Galaxy, whose masses we know, show periodic variations in their X-ray fluxes.
The periods of these oscillations - technically called quasiperiodic oscillations, or QPO for short - appear to correlate with the mass of the black hole. This is perhaps not surprising, as the time it takes for an oscillatory wave to propagate in a cavity depends on the size of the cavity, and bigger, more massive black holes basically have bigger resonance cavities. Strohmayer and Mushotzky have been using the XMM/Newton observatory to search for similar QPOs in ULXs and have found these QPOs in two objects so far; M82 X-1, and NGC 5408 X-1. In each case the periods of the QPOs are longer than seen in Galactic black holes, suggesting more massive black holes in the ULXs.
The most intriguing case is that of NGC 5408 X-1. Recent observations spaced two years apart found that the QPO period in NGC 5408 X-1 had increased (see Figure 1). The period increase was accompanied by a drop in the X-ray flux from the accretion disk, the same behavior that has been observed in the Galactic black holes, suggesting that the same physical processes are being observed. Whereas the typical oscillation period in the Galactic black holes is 1 second, in NGC 5408 X-1 it is 100 seconds. Moreover, the X-ray luminosity is also about a 100 times greater in NGC 5408 X-1 than the comparable Galactic black holes. These relatively simple arguments suggest that NGC 5408 X-1 is at least 100 times more massive than the Galactic black holes, and would make it one of the best candidates for a middle weight black hole of more than 1,000 solar masses.
The only way NGC 5408 X-1 can be so bright is that it must be devouring lots of matter. This amount of matter could only come from a binary companion star, and likely a rather massive one. Strohmayer has been using the SWIFT observatory to monitor the X-ray flux from NGC 5408 X-1 about twice per week for almost two years now. SWIFT is the only current observatory with both the imaging sensitivity (provided by the XRT), and the observing flexibility to carry out such a monitoring campaign. Enough data has now been collected to show that the flux from NGC 5408 X-1 varies with a 115-day period (see Figure 2). This is very likely the orbital period of the black hole binary. The relatively long orbital period indicates that the companion star must be a "bloated" star, a giant or supergiant. Only such a star would be big enough to be able to lose the requisite amount of matter to the black hole. More XMM/Newton observations are in the works to further pin down the mass of NGC 5408 X-1, but so far it represents one of the best cases for the elusive middle-weight black holes.