X-RAY TIME AND SPECTRAL VARIABILITY AS PROBES OF ULTRALUMINOUS X-RAY SOURCES

Abstract

A long-standing debate in the field of ultraluminous X-ray sources (ULXs: luminosities > 3&times;10<super>39</super> ergs s<super>-1</super>) is whether these objects are powered by stellar-mass black holes (mass range of 3-25 solar masses) undergoing hyper-accretion/emission or if they host the long-sought after class of intermediate-mass black holes (mass range of a few 100-1000 solar masses) accreting material at sub-Eddington rates. We present X-ray time and energy spectral variability studies of ULXs in order to understand their physical environments and accurately weigh their compact objects.

A sample of ULXs exhibit quasi-periodic oscillations (QPOs) with centroid frequencies in the range of 10-200 mHz. The nature of the power density spectra (PDS) of these sources is qualitatively similar to stellar-mass black holes when they exhibit the so-called type-C low-frequency QPOs (frequency range of 0.2-15 Hz). However, the crucial difference is that the characteristic frequencies within the PDS of ULXs, viz., the break frequencies and the centroid frequencies of the QPOs, are scaled down by a factor of approximately 10-100 compared to stellar-mass black holes. It has thus been argued that the ULX mHz QPOs are the type-C low-frequency QPO analogs of stellar-mass black holes and that the observed difference in the frequencies (a few&times;0.01 Hz compared with a few Hz) is due to the presence of intermediate-mass black holes (M_ULX = (QPO_{stellar-mass black hole}}/QPO_ULX)&times;M_{stellar-mass black hole}, where M and QPO are the mass and the QPO frequency, respectively) within these ULXs. We analyzed all the archival XMM-Newton X-ray data of ULXs NGC 5408 X-1 and M82 X-1 in order to test the hypothesis that the ULX mHz QPOs are the type-C analogs by searching for a correlation between the mHz QPO frequency and the energy spectral power-law index as type-C QPOs show such a dependence. From our multi-epoch timing and spectral analysis of ULXs NGC 5408 X-1 and M82 X-1, we found that the mHz QPOs of these sources vary in frequency by factors of approximately 4 and 6, respectively. However, we did not find evidence for changes in their energy-spectral indices. The apparent lack of a correlation--unlike the type-C QPOs--implies that either the ULX mHz QPOs are fundamentally different compared to the stellar-mass black hole low-frequency QPOs or they are indeed analogous to the low-frequency QPOs but with the observed dependence corresponding to the saturated portion of the correlation seen in stellar-mass black holes.

We analyzed all the archival Swift data of ULX NGC 5408 X-1 and found evidence for a 243&plusmn;23 day X-ray period. Based on its variation profile, energy dependence and transient nature, we argue that this period represents the orbital period of the black hole binary. We revisit the previously reported 62 day X-ray period of M82 X-1 and found evidence that the accretion disk's flux varies with this period's phase and also noted that the period's phase changed unusually fast during a certain epoch. Based on this we argue that this period might not be orbital but instead be due to a precessing accretion disk.

By combining and averaging all the archival RXTE/PCA data of M82 we detect stable, 3:2 frequency ratio QPOs (>4.7 &sigma; statistical significance) which we argue represent the high-frequency QPO analogs of stellar-mass black holes. Unlike the low-frequency QPOs, the high-frequency QPOs of stellar-mass black holes are stable, often occur in frequency ratios of 3:2 and scale inversely with black hole mass. Using the most recent mass estimates of stellar-mass black holes which show high-frequency QPOs and the detected 3:2 pair frequencies of 3.32&plusmn;0.06 and 5.07&plusmn;0.06 Hz from M82 X-1, we were able to accurately weigh its black hole to 428&plusmn;105 solar masses. This detection presents a unique technique to weigh the black holes in variable ULXs. Similar oscillations in other ULXs should be detectable with future X-ray observatories.

Finally, we conclude by discussing our preliminary results from the first X-ray -- optical reverberation mapping of a ULX and also describe future prospects of detecting intermediate-mass black holes using tidal disruption flares and by searching for high-frequency QPOs in ULXs.