X-ray and Observational Astronomy

News August 2002: Nearby interacting galaxy pair has a rich set of ultra-luminous X-ray sources

The optical image of the galaxies NGC 4485 & 4490 is shown in black and white, the Chandra X-ray colours are also shown: red for soft X-rays, blue for hard. Bright X-ray point sources and diffuse emission can be seen.	The same optical image (this time shown negative) with the locations of the detected Chandra X-ray sources marked. The six labelled X-ray sources are all ultra-luminous. A normal galaxy the size of the NGC 4490 (the larger of the two) would usually only have one such source. The outer limits of the optical galaxies are shown as ellipses.

An X-ray source that is not the massive black hole at the centre of a galaxy, yet still has a luminosity above 10^39 erg/sec is called an ultra-luminous X-ray source (ULX). At a luminosity of more than a quarter of a million times that of the sun, they are of great interest because such an extraordinarily high power surely indicates an exceptional environment.

One possible origin of such a high luminosity is a supernova explosion in a dense interstellar medium. Stars more massive than 10 times the mass of the sun explode when the nuclear fuel runs out. The remains of such an explosion are rammed at very high speed into any surrounding gas causing high temperature shocks radiating X-rays (and visible light). Dense gas radiates X-rays efficiently, leading to the possibility of very bright X-ray emission from the remnants of some supernovae. Nearby supernova remnants appear as beautiful X-ray 'nebulae', and such remnants will always appear to be extended if they are not to far away. They are also visible as radio sources.

Another possible origin of ULX emission is the accretion of gas onto a collapsed star such as a black hole or neutron star. When such an object is in a close binary system with another star it can pull gas off its companion and down onto itself. Because of the extraordinarily high densities and so high gravitational fields of these objects, gas falling towards them reaches very high speeds and very high temperatures. At slightly lower luminosities, these are the familiar X-ray binaries, of which there are around 200 in our own Galaxy. Because the X-ray emission from these system comes from such a small region near the black hole or neutron star, they will always as simple point-like sources of X-rays.

Most supernova remnants have X-ray luminosities less than 1% of a ULX, while X-ray binaries have a range of luminosities below that of a ULX. However, for an X-ray binary to shine the gas has to reach the collapsed star, and at very high luminosities, above the so-called 'Eddington limit', the force imparted on the in-falling gas by the emitted radiation is greater than the gravitational attraction. The problem with a ULX being an X-ray binary it that its luminosity, which is more than 10^39 erg/sec, is above this Eddington limit (~ 2 10^38 erg/sec for a neutron star).

Tim Roberts, Bob Warwick and Martin Ward from the University of Leicester, with Steven Murray of the Center for Astrophysics in Cambridge USA, have used the NASA Chandra X-ray satellite observatory to find 6 ULX in the nearby interacting galaxy pair NGC 4485/4490. While these are all point-like X-ray sources, one is coincident with a radio source, suggesting that this is a supernova remnant. The rest are almost certainly accreting binary sources.

The infra-red brightness and colour of these galaxies shows that they are undergoing moderate star formation, strengthening the growing link between the ULX phenomenon and star formation. Indeed the brightest sources show the greatest low energy absorption, implying that they are buried in regions of dense gas such as are expected where stars a forming and young, massive stars are exploding.

Tim's observation shows that most ULX are accreting binary stars, so how is the Eddington limit to be avoided? It has been suggested that the accreting objects in ULX are black holes of 100 to 10,000 times the mass of the sun. For such massive black holes the Eddington limit is high enough to avoid the paradox, but they take around a billion years to grow to this size, much too long to be associated with active star formation. An alternative is that the X-ray emission is beamed, in this way most of the accreting gas does not feel the force of the radiation. This explanation requires there to be many more ULX than we know about because most would be beamed away from us. Other possibilities include the accreting gas from the companion star being rich in helium due to nuclear burning, as this reduces the effective radiation pressure; or genuine super-Eddington accretion such as has recently been shown to be possible due to the presence of photon bubbles in the highly inhomogeneous, radiation dominated slim accretion disks around black holes.

The galaxy pair NGC 4485/4490 offer the nearest sample of ULX uncontaminated by the usual diffuse emission associated with star formation. They are likely to be the key in solving the mystery of the ULX phenomenon.

This work was recently accepted for publication in Monthly Notices of the Royal Astronomical Society.

Other news items are available here.

Back to group page.

Last updated: 2002 August 12 by Julian Osborne