Low metallicity and ultra-luminous X-ray sources in the Cartwheel galaxy

1. Low metallicity and ultra-luminous X-ray sources in the Cartwheel galaxy M. Mapelli, M. Colpi, L. Zampieri

2. Abstract • Low-metallicity massive stars: maybe directly collapsing into massive BHs • In Cartwheel galaxy, >105 massive BHs generated via this mechanism probably. • Such BHs might power most of the ULXs • A possible anti-correlation between the ULX number and metallicity • Data not sufficient to draw further conclusions

3. ULXs • Characteristic • Point-like sources with isotropic X-ray luminosity Lx>1039 erg s-1: higher than the Eddington luminosity for ~ 7 Msun BH • Most of the brightest ULXs located in the starburst galaxies

4. ULXs • The origin of ULXs: still an open question • Could be associated with HMXBs powered by stellar BHs • With anisotropic X-ray emission • With super-Eddington accretion rate / luminosity • With a combination of the two mechanisms • Could be associated with HMXBs powered by IMBHs • Larger than 100 Msun not needed for most of ULXs • Required only to explain the properties of some peculiar ULXs, such as • The brightest ULXs (< 4 ULXs with Lx>1041 erg s-1) • Those which show quasi-periodic oscillations • Which are surrounded by isotropically ionized nebula and a collisional ring galaxy

5. Cartwheel Galaxy • IMBHs hardly account for all the ULXs observed in the Cartwheel galaxy • >1000 IMBHs needed to produce the 17 observed ULXs • Hard to produce such high number of IMBHs according to the most common theoretical models, such as • The runaway collapse in young stellar clusters • The repeated mergers of stellar-mass BHs in the star clusters • The remnants of population III stars

6. Model • Low-metallicity massive stars might directly collapse into massive BHs: the total number of massive BHs • A: the normalization constant • The total mass of massive BHs

7. Model • The upper limit of the fraction of massive BHs which power ULXs in a given galaxy at present • The fraction of massive BHs which are expected to power ULXs at present fMT~ 0.03 a fraction of the life of the cluster fduty~ 10-2 the fraction of time which a transient source spends its burst phase

8. Results For The Cartwheel • These numbers quite higher than those predicted by the runaway collapse • More than one massive BH may form in the same cluster and massive BHs can form also outside clusters

9. Comparison With Other Galaxies

10. Comparison With Other Galaxies

11. Comparison With Other Galaxies • AM 0644-741: the metallicity measurement comes from the bulge (dominated by old stars, while no measurements for the star forming outer ring) • NGC 4485/4490: the value of Z likely an upper limit (the method does not work for metallicities Z<0.4) • NGC 4559: undergone a strong burst of SF and hard to find HII regions • In ring galaxies or in interacting galaxies: relatively easy to find regions where the SF just started

12. Comparison With Other Galaxies • A possible anti-correlation between the number and the metallicity • A linear relation between the number and the SFR

13. Conclusions • Low-metallicity massive stars might end their life by collapsing into massive BHs • Such massive BHs may power most of the observed ULXs in the low-metallicity galaxy • An anti-correlation between the number of ULXs and metallicity of the host galaxy may exist

14. Open Questions and Uncertainties • The final stellar masses reported by M92 are still debated • The models considered in M92 and in Fryer neglect some important effects, such as: • Rotation • The possible binarity of the progenitor • The model does not include the possibility of pair instability supernovae • The lact, paucity or uncertainty in the metallicity measurements make hard to test this model