Ultraluminous X-ray Sources in Interacting Galaxies

I give a brief review of how X-rays from nearby galaxies are used as direct tracers of recent star formation. This leads to the conclusion that it is the most luminous point-like sources that are associated with star formation and that the majority of these are high-mass X-ray binaries. I then discuss a recent study that shows that ULXs are preferentially found in regions as young as or younger than typical H ii regions in their host galaxies. Finally, I describe a new study that attempts to determine the maximum luminosity of ULXs in the local universe by searching for them in interacting galaxies where the star formation rate is high. 1 X-ray Sources as Tracers of Recent Star Formation David, Jones, & Forman (1992) performed the first large survey of galaxies that showed a correlation between the total X-ray luminosity and galaxy-wide starformation rate as measured by FIR (IRAS ) luminosities. Although the X-ray data were not of a quality sufficient to distinguish among different types of Xray sources, they argued on energetic grounds that most of the X-ray light came from high-mass X-ray binaries, with typical X-ray luminosities on the order of 10 ergs s and lifetimes of 10 to 10 yr rather than OB-star winds and SNe (both of which are much fainter though late O stars live longer than XRBs; see also Helfand & Moran (2001); Dalton & Sarazin (1995)). The next major step forward awaited Chandra’s superb angular resolution and broadband (CCD) spectral senstivity to isolate individual X-ray sources in nearby galaxies. The definitive work is Grimm et al. (2003). They showed that the point-source X-ray luminosity functions (XLFs) of individual normal and starburst galaxies all had roughly the same power law index over a wide range of source luminosities from <10 ergs s up to about the highest luminosities observed of >10 ergs s (Figure 1, left panel). Furthermore, they showed that the overall normalization of these XLFs scaled with the total star-formation rates (SFRs) of the individual host galaxies. Thus, by rescaling observed XLFs to their universal XLF (Figure 1, right panel) one could deduce the star formation rate of the observed galaxy. Since the total X-ray luminosity is just the integral of the XLF, it follows (rather circularly back to the work of David, Jones, & Forman (1992)) that the SFR can also be deduced from a measure of the total X-ray luminosity even when individual sources cannot be resolved such as at high