The fading of two transient ULXs to below the stellar mass Eddington limit

TL;DR

This study reports on the behavior of two transient ULXs in NGC 5128, showing their luminosity decline below the stellar mass Eddington limit, with spectral analysis indicating they resemble Galactic black hole X-ray binaries.

Contribution

It provides new observations of transient ULXs at lower luminosities, analyzes their duty cycles and spectral states, and discusses implications for their population in early-type galaxies.

Findings

ULXs' luminosities declined to ~2E37 erg/s

Spectral analysis suggests soft spectra, inconsistent with IMBH hard state

Duty cycles estimated between 0.12 and 0.2

Abstract

We report new detections of the two transient ultraluminous X-ray sources (ULXs) in NGC 5128 from an ongoing series of Chandra observations. Both sources have previously been observed Lx (2-3)E39 erg/s, at the lower end of the ULX luminosity range. The new observations allow us to study these sources in the luminosity regime frequented by the Galactic black hole X-ray binaries (BH XBs). We present the recent lightcurves of both ULXs. 1RXH J132519.8-430312 (ULX1) was observed at Lx ~ 1E38 erg/s, while CXOU J132518.2-430304 (ULX2) declined to Lx ~ 2E37 erg/s and then lingered at this luminosity for hundreds of days. We show that a reasonable upper limit for both duty cycles is 0.2, with a lower limit of 0.12 for ULX2. This duty cycle is larger than anticipated for transient ULXs in old stellar populations. By fitting simple spectral models in an observation with ~50 counts we recover properties consistent with Galactic BH XBs, but inconclusive as to the spectral state. We utilise quantile analyses to demonstrate that the spectra are generally soft, and that in one observation the spectrum of ULX2 is inconsistent with a canonical hard state at >95% confidence. This is contrary to what would be expected of an accreting IMBH primary, which we would expect to be in the hard state at these luminosities. We discuss the paucity of transient ULXs discovered in early-type galaxies and excogitate explanations. We suggest that the number of transient ULXs scales with the giant and sub-giant populations, rather than the total number of XBs.