A systematic study of variability in a sample of ultraluminous X-ray sources

We present results from a study of short-term variability in 19 archival observations by XMM–Newton of 16 ultraluminous X-ray sources (ULXs). Eight observations (six sources) showed intrinsic variability with power spectra in the form of either a power-law or broken power-law-like continuum and in some cases quasi-periodic oscillations (QPOs). The remaining observations were used to place upper limits on the strength of possible variability hidden within. Seven observations (seven sources) yielded upper limits comparable to, or higher than, the values measured from those observations with detectable variations. These represented the seven faintest sources, all with   f_x < 3 × 10⁻¹² erg cm⁻² s⁻¹  . In contrast, there are four observations (three sources) that gave upper limits significantly lower than both the values measured from the ULX observations with detectable variations, and the values expected by comparison with luminous Galactic black hole X-ray binaries (BHBs) and active galactic nuclei (AGN) in the observed frequency bandpass (10⁻³–1 Hz). This is the case irrespective of whether one assumes characteristic frequencies appropriate for a stellar mass  (10 M_⊙)  or an intermediate mass  (1000 M_⊙)  black hole, and means that in some ULXs the variability is significantly suppressed compared to bright BHBs and AGN. We discuss ways to account for this unusual suppression in terms of both observational and intrinsic effects and whether these solutions are supported by our results.     

Spectral variability of ultraluminous X-ray sources

We study spectral variability of 11 ultraluminous X-ray sources (ULX) using archived XMM–Newton and Chandra observations. We use three models to describe the observed spectra: a power law, a multicolour disc (MCD) and a combination of these two models. We find that seven ULXs show a correlation between the luminosity L _X and the photon index Γ. Furthermore, four out of these seven ULXs also show spectral pivoting in the observed energy band. We also find that two ULXs show an   L _X–Γ  anticorrelation. The spectra of four ULXs in the sample can be adequately fitted with a MCD model. We compare these sources to known black hole binaries (BHB) and find that they follow similar paths in their luminosity–temperature diagrams. Finally, we show that the 'soft excess' reported for many of these ULXs at ∼0.2 keV seems to roughly follow a trend   L _soft∝ T ^−3.5  when modelled with a power law plus a 'cool' MCD model. This is contrary to the   L ∝ T ⁴  relation that is expected from theory and what is seen for many accreting BHBs. The observed trend could instead arise from disc emission beamed by an outflowing wind around a  ∼10 M_⊙  black hole.     

The ultraluminous X-ray source population of NGC 4485/4490

We report the results of spectral and temporal variability studies of the ultraluminous X-ray sources (ULXs) contained within the interacting pair of galaxies NGC 4485/4490, combining Chandra and XMM–Newton observations. Each of the four separate observations provide at least modest quality spectra and light curves for each of the six previously identified ULXs in this system; we also note the presence of a new transient ULX in the most recent observation. No short-term variability was observed for any ULX within our sample, but three out of five sources show correlated flux/spectral changes over longer time-scales, with two others remaining stable in spectrum and luminosity over a period of at least 5 yr. We model the spectra with simple power-law and multicolour disc blackbody models. Although the data are insufficient to statistically distinguish models in each epoch, those better modelled (in terms of their  χ²  fit) by a multicolour disc blackbody appear to show a disc-like correlation between luminosity and temperature, whereas those modelled by a power-law veer sharply away from such a relationship. The ULXs with possible correlated flux/spectral changes appear to change spectral form at  ∼2 × 10³⁹ erg s⁻¹  , suggestive of a possible change in spectral state at high luminosities. If this transition is occurring between the very high state and a super-Eddington ultraluminous state, it indicates that the mass of the black holes in these ULXs is around  10–15 M_⊙  .     

Weighing the black holes in ultraluminous X-ray sources through timing

We describe a new method to estimate the mass of black holes in Ultraluminous X-ray Sources (ULXs). The method is based on the recently discovered 'variability plane', populated by Galactic stellar-mass black-hole candidates (BHCs) and supermassive active galactic nuclei (AGNs), in the parameter space defined by the black-hole mass, accretion rate and characteristic frequency. We apply this method to the two ULXs from which low-frequency quasi-periodic oscillations have been discovered, M82 X-1 and NGC 5408 X-1. For both sources we obtain a black-hole mass in the range  100–1300 M_⊙  , thus providing evidence for these two sources to host an intermediate-mass black hole.     

Optical and infrared signatures of ultra-luminous X-ray sources

We have constructed a model to describe the optical emission from ultra-luminous X-ray sources (ULXs). We assume a binary model with a black hole accreting matter from a Roche lobe filling companion star. We consider the effects of radiative transport and radiative equilibrium in the irradiated surfaces of both the star and a thin accretion disc. We have developed this model as a tool with which to positively identify the optical counterparts of ULXs, and subsequently derive parameters such as the black hole mass and the luminosity class and spectral type of the counterpart. We examine the dependence of the optical emission on these and other variables. We extend our model to examine the magnitude variation at infrared wavelengths, and we find that observations at these wavelengths may have more diagnostic power than in the optical. We apply our model to existing HST observations of the candidates for the optical counterpart of ULX X-7 in NGC 4559. All candidates could be consistent with an irradiated star alone, but we find that a number of them are too faint to fit with an irradiated star and disc together. Were one of these the optical counterpart to X-7, it would display a significant temporal variation.     

Stellar-mass black hole binaries as ultraluminous X-ray sources

Ultraluminous X-ray sources (ULXs) with   L _x > 10³⁹ erg s⁻¹  have been discovered in great numbers in external galaxies with ROSAT , Chandra and XMM-Newton . The central question regarding this important class of sources is whether they represent an extension in the luminosity function of binary X-ray sources containing neutron stars and stellar-mass black holes (BHs), or a new class of objects, e.g. systems containing intermediate-mass BHs  (100–1000 M_⊙)  . We have carried out a theoretical study to test whether a large fraction of the ULXs, especially those in galaxies with recent star formation activity, can be explained with binary systems containing stellar-mass BHs. To this end, we have applied a unique set of binary evolution models for BH X-ray binaries, coupled to a binary population synthesis code, to model the ULXs observed in external galaxies. We find that for donor stars with initial masses  ≳10 M_⊙  the mass transfer driven by the normal nuclear evolution of the donor star is sufficient to potentially power most ULXs. This is the case during core hydrogen burning and, to an even more pronounced degree, while the donor star ascends the giant branch, although the latter phases last only ∼5 per cent of the main-sequence phase. We show that with only a modest violation of the Eddington limit, e.g. a factor of ∼10, both the numbers and properties of the majority of the ULXs can be reproduced. One of our conclusions is that if stellar-mass BH binaries account for a significant fraction of ULXs in star-forming galaxies, then the rate of formation of such systems is  ∼3 × 10⁻⁷ yr⁻¹  normalized to a core-collapse supernova rate of 0.01 yr⁻¹.     

The radio–mid-infrared correlation and the contribution of 15-μm galaxies to the 1.4-GHz source counts

The radio counterparts to the 15-μm sources in the European Large Area ISO Survey southern fields are identified in 1.4-GHz maps down to ∼80 μJy. The radio–mid-infrared correlation is investigated and derived for the first time at these flux densities for a sample of this size. Our results show that radio and mid-infrared (MIR) luminosities correlate almost as well as radio and far-infrared (FIR), at least up to   z ≃ 0.6  . Using the derived relation and its spread together with the observed 15-μm counts, we have estimated the expected contribution of the 15-μm extragalactic populations to the radio source counts and the role of MIR starburst galaxies in the well-known 1.4-GHz source excess observed at sub-mJy levels. Our analysis demonstrates that IR emitting starburst galaxies do not contribute significantly to the 1.4-GHz counts for strong sources, but start to become a significant fraction of the radio source population at flux densities ≲0.5–0.8 mJy. They are expected to be responsible for more than 60 per cent of the observed radio counts at ≲0.05 mJy. These results are in agreement with the existing results on optical identifications of faint radio sources.     

The ultraluminous X-ray sources in the high-velocity system of NGC 1275

We report the results of a study of X-ray point sources coincident with the high-velocity system (HVS) projected in front of NGC 1275. A very deep X-ray image of the core of the Perseus cluster, made with the Chandra X-ray Observatory , has been used. We find a population of ultraluminous X-ray sources [ULXs; seven sources with   L _X(0.5 − 7.0  keV) > 7 × 10³⁹ erg s^-1  ]. As with the ULX populations in the Antennae and Cartwheel galaxies, those in the HVS are associated with a region of very active star formation. Several sources have possible optical counterparts found on the Hubble Space Telescope ( HST ) images, although the X-ray brightest one does not. Absorbed power-law models fit the X-ray spectra, with most having a photon index between 2 and 3.     

Limits on the X-ray and optical luminosity of the progenitor of the Type Ia supernova 2007sr

We present Hubble Space Telescope ( HST )/Wide Field Planetary Camera 2 (WFPC2), Galaxy Evolution Explorer ( GALEX ) and Chandra observations of the position of the Type Ia supernova 2007sr in the Antennae galaxy, taken before the explosion. No source is found in any of the observations, allowing us to put interesting constraints on the progenitor luminosity. In total there is about 450 ks of Chandra data, spread over seven different observations. Limiting magnitudes of far-ultraviolet (FUV) (23.7 AB mag), near-ultraviolet (NUV) (23.8 AB mag), F555W (26.5 Vega mag) and F814W (24.5–25 Vega mag) are derived. The distance to the Antennae galaxy is surprisingly poorly known, with almost a factor of 2 difference between the latest distance based on the tip of the red giant branch (13.3 Mpc) and the distance derived from the 2007sr light curve (25 Mpc). Using these distances we derive limits on absolute optical and UV magnitudes of any progenitor but these are still above the brightest (symbiotic) proposed progenitors. From the Chandra data a 3σ upper limit to the X-ray luminosity of  0.5–8.0 × 10³⁷ erg s⁻¹  in the 0.3–1 keV range is found. This is below the X-ray luminosity of the potential progenitor of the Type Ia supernova 2007on that we recently discovered and for which we report a corrected X-ray luminosity. If that progenitor is confirmed it suggests the two supernovae have different progenitors. The X-ray limit is comparable to the brightest supersoft X-ray sources in the Galaxy, the Large Magellanic Cloud (LMC) and the Small Magellanic Cloud (SMC) and significantly below the luminosities of the brightest supersoft and quasi-soft X-ray sources found in nearby galaxies, ruling out such sources as progenitors of this Type Ia supernova.     

No periodicities in 2dF Redshift Survey data

I suggest that there are two classes of ultraluminous X-ray sources (ULXs), corresponding to super-Eddington mass inflow in two situations: (a) thermal-time-scale mass transfer in high-mass X-ray binaries, and (b) long-lasting transient outbursts in low-mass X-ray binaries. These two classes are exemplified by SS433 and microquasars like GRS 1915+105 respectively. The observed ULX population is a varying mixture of the two, depending on the star formation history of the host galaxy. ULXs in galaxies with vigorous star formation (such as the Antennae) are generally SS433-like, while ULXs in elliptical galaxies must be of the microquasar type. The latter probably have significantly anisotropic radiation patterns. They should also be variable, but demonstrating this may require observations over decades. The close analogy between models of X-ray binaries and active galactic nuclei (AGN) suggests that there should exist an apparently super-Eddington class of the latter, which may be the ultrasoft AGN, and a set of X-ray binaries with Doppler-boosted X-ray emission. These are presumably a subset of the ULXs, but remain as yet unidentified.     

A closer look at the X-ray transient XTE J1908+094: identification of two new near-infrared candidate counterparts

We have reported, in our previous paper, on the near-infrared (NIR) identification of a possible counterpart to the black hole candidate XTE J1908+094 obtained with the European Southern Observatory/New Technology Telescope. Here, we present new, follow-up, Canada–France–Hawaii Telescope adaptive optics observations of the XTE J1908+094 field, which resolved the previously proposed counterpart in two objects separated by about 0.8 arcsec. Assuming that both objects are potential candidate counterparts, we derive that the binary system is a low-mass system with a companion star which could be either an intermediate/late type (A–K) main-sequence star at a distance of 3–10 kpc or a late-type (>K) main-sequence star at a distance of 1–3 kpc. However, we show that the brighter of the two objects ( J ∼ 20.1,  H ∼ 18.7,  K '∼ 17.8) is more likely to be the real counterpart of the X-ray source. Its position is more compatible with our astrometric solution, and colours and magnitudes of the other object are not consistent with the lower limit of 3 kpc derived independently from the peak bolometric flux of XTE J1908+094. Further multiwavelength observations of both candidate counterparts are crucial in order to solve the pending identification.     

XMM–Newton and Chandra observations of SHEEP sources

We present Chandra and XMM–Newton observations of 12 bright  [ f (2–10 keV) > 10⁻¹³ erg cm⁻² s⁻¹]  sources from the ASCA search for the High Energy Extragalactic Population (SHEEP) survey. Most of these have been either not observed or not detected previously with the ROSAT mission, and therefore they constitute a sample biased towards hard sources. The Chandra observations are important in locating the optical counterpart of the X-ray sources with accuracy. Optical spectroscopic observations show that our sample is associated with both narrow-line (NL) (six objects) and broad-line (BL) active galactic nuclei (AGN) (five objects), with one source remaining unidentified. Our sources cover the redshift range 0.04–1.29, spanning luminosities from 10⁴² to  10⁴⁵ erg s⁻¹  (2–10 keV). The NL sources have preferentially lower redshift (and luminosity) compared to the BL ones. This can be most easily explained in a model where the NL AGN are intrinsically less luminous than the BL ones in line with the results of Steffen et al. The X-ray spectral fittings show a roughly equal number of obscured  ( N _H > 10²² cm⁻²)  and unobscured  ( N _H < 10²² cm⁻²)  sources. There is a clear tendency for obscured sources to be associated with NL AGN and unobscured sources with BL ones. However, there is a marked exception with the highest obscuring column observed at a BL AGN at a redshift of z = 0.5.     

Testing the starburst/AGN connection with SWIRE X-ray/70 μm sources

We explore the nature of X-ray sources with  70 μm  counterparts selected in the Spitzer Wide-Area Infrared Extragalactic Survey (SWIRE) fields: ELAIS-N1, Lockman Hole and Chandra Deep Field South, for which Chandra X-ray data are available. A total of 28 X-ray/  70 μm  sources in the redshift interval  0.5 < z < 1.3  are selected. The X-ray luminosities and the shape of the X-ray spectra show that these sources are active galactic nuclei (AGN). Modelling of the optical to far-infrared (IR) spectral energy distribution indicates that most of them (27/28) have a strong starburst component  (>50 M_⊙ yr⁻¹)  that dominates in the IR. It is found that the X-ray and IR luminosities of the sample sources are broadly correlated, consistent with a link between AGN activity and star formation. Contrary to the predictions of some models for the co-evolution of AGN and galaxies, the X-ray/  70 μm  sources in the sample are not more obscured at X-ray wavelengths compared to the overall X-ray population. It is also found that the X-ray/  70 μm  sources have lower specific star formation rates compared to the general  70 μm  population, consistent with AGN feedback moderating the star formation in the host galaxies.     

BeppoSAX observations of 1-Jy BL Lacertae objects – I

We present new BeppoSAX observations of seven BL Lacertae objects selected from the 1-Jy sample plus one additional source. The collected data cover the energy range     (observer's frame), reaching ∼50 keV for one source (BL Lac). All sources characterized by a peak in their multifrequency spectra at infrared/optical energies (i.e., of the low-energy peaked BL Lac type, LBL) display a relatively flat     X-ray spectrum, which we interpret as inverse Compton emission. Four objects (two-thirds of the LBLs) show some evidence for a low-energy steepening, which is probably due to the synchrotron tail merging into the inverse Compton component around ∼     . If this were generally the case with LBLs, it would explain why the     ROSAT spectra of our sources are systematically steeper than the BeppoSAX ones     . The broad-band spectral energy distributions fully confirm this picture, and a synchrotron inverse Compton model allows us to derive the physical parameters (intrinsic power, magnetic field, etc.) of our sources. Combining our results with those obtained by BeppoSAX on BL Lacs covering a wide range of synchrotron peak frequency, ν _peak, we confirm and clarify the dependence of the X-ray spectral index on ν _peak originally found in ROSAT data.     

X-ray spectral evolution of the extragalactic Z source LMC X-2

We present the results obtained by a detailed study of the extragalactic Z source LMC X-2, using broad-band Suzaku data and a large (∼750 ks) data set obtained with the proportional counter array (PCA) onboard the Rossi X-ray Timing Experiment ( RXTE ). The PCA data allow the study of the complete spectral evolution along the horizontal, normal and flaring branches of the Z track. Comparison with previous studies shows that the details of spectral evolution (like the variation of Comptonizing electron temperature) are similar to those of GX 17+2 but unlike those of Cyg X-2 and GX 349+2. This suggests that Z sources are a heterogeneous group, with perhaps LMC X-2 and GX 17+2 being members of a subclass. However, non-monotonic evolution of the Compton y parameter seems to be generic to all sources. The broad-band Suzaku data reveal that the case in which the additional soft component of the source is modelled as disc blackbody emission is strongly preferred over the one where it is taken to be a blackbody spectrum. This component, as well as the temperature of seed photons, does not vary when the source goes into flaring mode, and the entire variation can be ascribed to the Comptonizing cloud. The bolometric unabsorbed luminosity of the source is constrained to be  ∼2.23 × 10³⁸ erg s⁻¹  , which, if the source is Eddington-limited, implies a neutron star mass of  1.6 M_⊙  . We discuss the implications of these results.     

The distribution of active galactic nuclei in a large sample of galaxy clusters

We present an analysis of the X-ray point source populations in 182 Chandra images of galaxy clusters at   z > 0.1  with exposure time >10 ks, as well as 44 non-cluster fields. The analysis of the number and flux of these sources, using a detailed pipeline to predict the distribution of non-cluster sources in each field, reveals an excess of X-ray point sources associated with the galaxy clusters. A sample of 148 galaxy clusters at  0.1 < z < 0.9  , with no other nearby clusters, shows an excess of 230 cluster sources in total, an average of ∼1.5 sources per cluster. The lack of optical data for these clusters limits the physical interpretation of this result, as we cannot calculate the fraction of cluster galaxies hosting X-ray sources. However, the fluxes of the excess sources indicate that over half of them are very likely to be active galactic nuclei (AGN), and the radial distribution shows that they are quite evenly distributed over the central 1 Mpc of the cluster, with almost no sources found beyond this radius. We also use this pipeline to successfully reproduce the results of previous studies, particularly the higher density of sources in the central 0.5 Mpc of a few cluster fields, but show that these conclusions are not generally valid for this larger sample of clusters. We conclude that some of these differences may be due to the sample properties, such as the size and redshift of the clusters studied, or a lack of publications for cluster fields with no excess sources. This paper also presents the basic X-ray properties of the galaxy clusters, and in subsequent papers in this series the dependence of the AGN population on these cluster properties will be evaluated.
In addition the properties of over 9500 X-ray point sources in the fields of galaxy clusters are tabulated in a separate catalogue available online or at http://www.sc.eso.org~rgilmour .     

Chandra monitoring observations of the ultraluminous X-ray source NGC 5204 X-1

We report the results of a two-month campaign conducted with the Chandra X-ray observatory to monitor the ultraluminous X-ray source (ULX) NGC 5204 X-1. This was composed of a 50-ks observation, followed by ten 5-ks follow-ups spaced initially at ∼3, then at ∼10-d intervals. The ULX flux is seen to vary by factors ∼5 on time-scales of a few days, but no strong variability is seen on time-scales shorter than an hour. There is no evidence for a periodic signal in the X-ray data. An examination of the X-ray colour variations over the period of the campaign shows the ULX emission consistently becomes spectrally harder as its flux increases. The X-ray spectrum from the 50-ks observation can be fitted by a number of disparate spectral models, all of which describe a smooth continuum with, unusually for a ULX, a broad emission feature evident at 0.96 keV. The spectral variations, both within the 50-ks observation and over the course of the whole campaign, can then be explained solely by variations in the continuum component. In the context of an optically thick corona model (as found in other recent results for ULXs) the spectral variations can be explained by the heating of the corona as the luminosity of the ULX increases, consistent with the behaviour of at least one Galactic black hole system in the strongly Comptonized very high state. We find no new evidence supporting the presence of an intermediate-mass black hole in this ULX.     

Multifrequency studies of the enigmatic gamma-ray source 3EG J1835+5918

The EGRET telescope aboard the NASA Compton Gamma-Ray Observatory ( CGRO ) has repeatedly detected 3EG J1835+5918, a bright and steady source of high-energy gamma-ray emission which has not yet been identified. The absence of any likely counterpart for a bright gamma-ray source located 25° off the Galactic plane initiated several attempts of deep observations at other wavelengths. We report on counterparts in X-rays on a basis of a 60-ks ROSAT HRI image. In order to conclude on the plausibility of the X-ray counterparts, we reanalysed data from EGRET at energies above 100 MeV and above 1 GeV, including data up to CGRO observation cycle 7. The gamma-ray source location represents the latest and probably the final positional assessment based on EGRET data. We especially address the question of flux and spectral variability, here discussed using the largest and most homogeneous data set available at high-energy gamma-rays for many years. The results from X-ray and gamma-ray observations were used in a follow-up optical identification campaign at the 2.2-m Guillermo Haro Telescope at Cananea, Mexico. VRI imaging has been performed at the positions of all of the X-ray counterpart candidates, and spectra were taken where applicable. The results of the multifrequency identification campaign toward this enigmatic unidentified gamma-ray source are given, especially on the one object which might be associated with the gamma-ray source 3EG J1835+5918. This object has the characteristics of an isolated neutron star and possibly of a radio-quiet pulsar.     

On the maximum amplitude and coherence of the kilohertz quasi-periodic oscillations in low-mass X-ray binaries

I study the behaviour of the maximum rms fractional amplitude, r _max, and the maximum coherence, Q _max, of the kilohertz quasi-periodic oscillations (kHz QPOs) in a dozen low-mass X-ray binaries. I find that (i) the maximum rms amplitudes of the lower- and upper-kHz QPOs,   r ^ℓ_max  and   r ^u_max  , respectively, decrease more or less exponentially with increasing luminosity of the source; (ii) the maximum coherence of the lower-kHz QPO,   Q ^ℓ_max  , first increases and then decreases exponentially with luminosity, at a faster rate than both   r ^ℓ_max  and   r ^u_max  ; (iii) the maximum coherence of the upper-kHz QPO,   Q ^u_max  , is more or less independent of luminosity; and (iv) r _max and Q _max show the opposite behaviour with hardness of the source, consistent with the fact that there is a general anticorrelation between luminosity and spectral hardness in these sources. Both r _max and Q _max in the sample of sources, and the rms amplitude and coherence of the kHz QPOs in individual sources show a similar behaviour with hardness. This similarity argues against the interpretation that the drop of coherence and rms amplitude of the lower-kHz QPO at high QPO frequencies in individual sources is a signature of the innermost stable circular orbit around a neutron star. I discuss possible interpretations of these results in terms of the modulation mechanisms that may be responsible for the observed variability.