Spectral States and Evolution of Ultraluminous X-ray Sources

TL;DR

This study investigates spectral evolution in ultraluminous X-ray sources, revealing two common behaviors and suggesting different accretion states and black hole masses, with implications for understanding their nature.

Contribution

It identifies two spectral behaviors in ULXs, interprets the L-Gamma correlation as an intermediate state, and suggests some ULXs contain more massive black holes than stellar-mass ones.

Findings

ULXs in starburst galaxies remain in a hard state with constant photon index.

NGC 5204 X-1 shows a correlation between luminosity and spectral index.

Evidence for super-Eddington accretion in some ULXs.

Abstract

We examined spectral evolution in ultraluminous X-ray sources (ULXs) with apparent luminosities of about 10^40 ergs/s. Based on new results in this paper and those reported in the literature, two common spectral behaviors were found. Some ULXs in starburst galaxies have varying luminosity (L) but remain in the hard state with power-law spectra and a constant, hard photon index (Gamma). Other ULXs, such as NGC 5204 X-1, show a correlation between L and Gamma. We interpret this L-Gamma correlated phase as an intermediate state with hybrid properties from the thermal dominant and steep power-law states. When the spectra of NGC 5204 X-1 are fitted with a multicolor disk blackbody plus power-law model, the X-ray luminosity increases with the effective temperature of the accretion disk in a manner similar to that found in stellar-mass black hole X-ray binaries, suggesting that the emission arises from an accretion disk. The luminosity, disk size, and temperature suggest that NGC 5204 X-1 harbors a compact object more massive than stellar-mass black holes. In contrast, the disk model in IC 342 X-1 is ruled out because the luminosity decreases as the temperature increases; sources with such behaviors may represent a class of objects with super-Eddington accretion. Also, we report a peculiar soft spectral feature from IC 342 X-2 and variability on a time scale of 20 ks from Holmberg II X-1. More observations are needed to test these results.