The Ultraluminous X‐Ray Source Population from the Chandra Archive of Galaxies

Abstract

One hundred fifty-four discrete non-nuclear ultraluminous X-ray ( ULX) sources, with spectroscopically determined intrinsic X-ray luminosities greater than 10 39 ergs s À1 , are identified in 82 galaxies observed with Chandra's Advanced CCD Imaging Spectrometer. Source positions, X-ray luminosities, and spectral and timing characteristics are tabulated. Statistical comparisons between these X-ray properties and those of the weaker discrete sources in the same fields (mainly neutron star and stellar-mass black hole binaries) are made. Sources above $10 38 ergs s À1 display similar spatial, spectral, color, and variability distributions. In particular, there is no compelling evidence in the sample for a new and distinct class of X-ray object such as the intermediate-mass black holes. Eighty-three percent of ULX candidates have spectra that can be described as absorbed power laws with index hÀi ¼ 1:74 and column density hN H i ¼ 2:24 ; 10 21 cm À2 , or $5 times the average Galactic column. About 20% of the ULXs have much steeper indices indicative of a soft, and likely thermal, spectrum. The locations of ULXs in their host galaxies are strongly peaked toward their galaxy centers. The deprojected radial distribution of the ULX candidates is somewhat steeper than an exponential disk, indistinguishable from that of the weaker sources. About 5%-15% of ULX candidates are variable during the Chandra observations (which average 39.5 ks). Comparison of the cumulative X-ray luminosity functions of the ULXs to Chandra Deep Field results suggests $25% of the sources may be background objects, including 14% of the ULX candidates in the sample of spiral galaxies and 44% of those in elliptical galaxies, implying the elliptical galaxy ULX population is severely compromised by background active galactic nuclei. Correlations with host galaxy properties confirm the number and total X-ray luminosity of the ULXs are associated with recent star formation and with galaxy merging and interactions. The preponderance of ULXs in star-forming galaxies as well as their similarities to less-luminous sources suggest they originate in a young but short-lived population such as the high-mass X-ray binaries with a smaller contribution ( based on spectral slope) from recent supernovae. The number of ULXs in elliptical galaxies scales with host galaxy mass and can be explained most simply as the high-luminosity end of the low-mass X-ray binary population.

High-energy neutrino fluxes from AGN populations inferred from X-ray surveys

“…X-ray binaries and other luminous objects that could also be observed at these redshifts (e.g. Swartz et al 2004;Feng & Soria 2011). These sources may also show beamed luminosities comparable to the fainter blazars, but whose intrinsic luminosity generally is found at around 10 38 erg s´1.…”

X-ray properties of radio-selected star forming galaxies in theChandra-COSMOS survey

“…However, Swartz et al (2004) found that the average slopes of X-ray binaries in nearby galaxies are Γ = 1.88 ± 0.06 (1.97 ± 0.11) for binaries with luminosity larger (smaller) than 10 39 erg s −1 , respectively. Only 7% (8%) of the binaries studied by Swartz et al (2004) have slopes Γ ≤ 1.4.…”

“…Harder spectra are sometimes indeed found in galaxies (e.g., see the slopes for single power-law fits in Dahlem et al 1998) where they can result from bright X-ray binaries (Persic & Rephaeli 2002). However, Swartz et al (2004) found that the average slopes of X-ray binaries in nearby galaxies are Γ = 1.88 ± 0.06 (1.97 ± 0.11) for binaries with luminosity larger (smaller) than 10 39 erg s −1 , respectively. Only 7% (8%) of the binaries studied by Swartz et al (2004) have slopes Γ ≤ 1.4.…”

X-ray observations of ultraluminous X-ray sources

“…Two arguments stand out as the principle reasons the ULX population is not dominated by (large) accreting IMBHs. Firstly, the luminosity function of X-ray sources in galaxies (XLF) has an unbroken power-law form for 5 decades up to a luminosity of ∼ 2 × 10 40 erg s −1 (Grimm, Gilfanov & Sunyaev 2003;Swartz et al 2004). This break occurs at ∼ 10 per cent of the Eddington luminosity for the ∼ 1000M ⊙ black holes inferred from ULX spectroscopy.…”

“…Unfortunately, ULXs do not generally show much short-term variability (Swartz et al 2004), with very few examples displaying sufficient variability power to establish a PSD from (Feng & Kaaret 2005). However, some measurements have been made.…”