1. Observational Evidence for Quasi-soft X-Ray Sources in Nearby Galaxies and the link to Intermediate-mass Black Holes Albert Kong and Rosanne Di Stefano Harvard-Smithsonian Center for Astrophysics

2. X-ray Point Sources in Nearby Galaxies Chandra and XMM-Newton found many point sources in nearby galaxies. They are mainly X-ray binaries, but some of them are SNRs or foreground/background objects. We found some extreme X-ray objects in nearby galaxies. Some off-nucleus point sources have luminosity > 10 39 erg/s, often referred as ultra-luminous X-ray sources (ULXs). We also found some very soft X-ray sources typically with temperature of ~100eV; some of them are also ULXs (e.g., Mukai et al. 2003; Kong & Di Stefano 2003; Fabbiano et al. 2003).

3. Very Soft X-ray Sources Supersoft Sources and Quasi-soft Sources SSSs are sources with kT ~ tens of eV, and with L typically between 10 36 -10 38 erg/s. Some may be nuclear-burning WDs; some of these may be progenitors of Type Ia supernovae. Because absorption is usually low comparing to our Galaxy, external galaxies are good to search for very soft sources (VSSs). We have developed a method to search for VSSs in external galaxies. (Di Stefano & Kong 2003 a,b,c)

4. Very Soft X-ray Sources Supersoft Sources and Quasi-soft Sources QSSs typically have 100eV < kT < 250 eV. Or else they have a softer dominant spectrum, but may include a small hard component. We have tested our algorithm on simulated data.

6. We have tested our algorithm on real data Chandra and some XMM- Newton from ~20 galaxies. We find many SSSs and QSSs; their luminosity is between 10 36 -10 39 erg/s. QSSs and SSSs occupy between 11% and 45% of all X-ray sources.

7. Recurrent SSS in NGC 300 kT=57 eV L bol =1.5x10 39 erg/s kT bb =91 eV kT RS =74 eV L bol =7x10 39 erg/s kT=71 eV L bol =2.3x10 38 erg/s kT=76 eV L bol =4x10 39 erg/s Supersoft Sources

8. NGC1399-162 kT MCD =0.35 keV L 0.3-7 =1.7x10 39 erg/s QSS-σ M83-35 QSS-MNOH kT=0.12 keV L 0.3-7 =10 38 erg/s M51-10 QSS-σ kT=0.24 keV L 0.3-7 =3.6x10 38 erg/s kT=0.13 keV L 0.3-7 =4.3x10 37 erg/s M101-114 QSS-MNOH Quasi-soft Sources

9. A Luminous SSSs in M31 kT=70eV, α=1.8 L 0.3-7 =4x10 38 erg/s

10. M83M101 M51 NGC4697 Red: SSSs Blue: QSSs

11. What are the Quasi-soft Sources? They are almost certainly not WDs. They may be SNRs, but variability can distinguish them from accreting objects. The most natural explanation may be that they are accreting IMBHs. If the accretion is mediated by a disk which is geometrically thin, but optically thick, the spectrum must be soft. (M/10 3 M Sun )= h (42eV/kT lso ) 2 [(ξ/0.1) L obs /3x10 37 erg/s] 1/2 where h is a factor of order unity. They may represent a simple extension of observed properties of Galactic BHs.

12. Cyg X-1 kT=0.3 keV, α=2.4 L 0.5-200 =7x10 37 erg/s M BH =6.9-13.2 M Sun XTE J1550-564 kT=0.8 keV, α=2.3 L 0.5-10 =10 38 erg/s M BH =8.4-10.8M Sun Chandra+RXTE (Miller et al. 2003)

13. IC 342 X-1 kT MCD =1.98 keV L 0.5-10 =1x10 39 erg/s Kong 2003 kT MCD =0.2 keV α=1.8 L 0.5-10 =5x10 39 erg/s Miller et al. 2003

14. A Recurrent SSS in NGC 300 Recurrent SSS in NGC 300 kT=57 eV L bol =1.5x10 39 erg/s

15. If the 5.4 hr variability is orbital origin, then the system is consistent with a ~1000 M Sun IMBH. If the donor fills its Roche lobe, then P orb (8.9 hr)(M d /M ), where M d is the mass of the donor star. If P orb = 5.4 hr, then M d = 0.61 M Sun. See Rosanne’s poster (#13) for details.

16. Alternative Models Neutron star: the very soft X-ray emission would presumably emanate from a photosphere is much larger than the neutron star itself. Stellar-mass BH: Compton thick outflow from a stellar-mass BH, accreting near the Eddington limit (King & Pounds 2003).

17. Summary Galaxies are rich in SSSs and in QSSs. We are faced with the challenge of understanding the nature(s) of the members of each of these classes of X-ray sources. Some SSSs and some QSSs may be IMBHs. Whatever the nature(s) of SSSs and QSSs (See Rosanne’s poster #13) IMBHs need not be ultraluminous. IMBHs can be found in anywhere in ellipticals and spirals. Binary evolution can provide important and testable predictions about the properties of accreting IMBHs.