Ultraluminous X-ray Sources: The most extreme X-ray binaries Luca Zampieri INAF-Astronomical Observatory of Padova 1
Outline Recent advancements on ULXs Present understanding of ULXs Conclusions 2 ULXs Lubiana May 11, 2012- LZ
ULXs X-1 X-2 X-3 = SN 1978K 4 ULXs Lubiana May 11, 2012- LZ
ULXs X-ray observations of nearby galaxies show a population of pointlike, off-nuclear sources with L >> Ledd for 1 Msun (L>1.0e39 erg/s) UltraLuminous X-ray Sources (e.g. Feng & Soria 2011) Likely the majority of ULXs are accreting BHs in binaries (e.g. Zampieri & Roberts 2009) - High X-ray luminosity, X-ray spectra, short term variability - Often in young stellar environments, stellar optical count. - (Orbital) modulation in the X-ray (and possibly optical) flux - Timing properties (QPOs) Colbert & Mushotzky 1999; Roberts & Warwick 2000; La Parola et al. 2001; Strohmayer & Mushotzky 2003; Swart et al. 2004; Zampieri et al. 2004; Mucciarelli et al. 2005, 2006, 2007; Soria et al. 2005; Kaaret et al. 2004, 2006; Liu et al. 2002, 2004, 2007; Pakull et al. 2006; Stobbart et al. 2006; Strohmayer et al. 2007; Gladstone et al. 2009 6
Challenging ULXs Challenging ULXs: what are the masses of the BHs powering these sources? Possibility to probe existence of BHs in a mass range unexplored what are the properties of their donor stars? And those of their accretion flow? What is the relation with their environment? Formation of the compact object very intriguing issue, possibly linked to the existence of different classes of black holes: - intermediate-mass BHs of 10 2 10 4 Msun (Colbert & Mushotzky 99) - stellar-mass BHs of ~10 Msun (King et al. 01, 08) - massive stellar BHs of 30-80 Msun (Zampieri & Roberts 09; Mapelli et al. 09; Belczynski et al. 10) Or are they in a different accretion state wrt Galactic BH binaries? ULXs Lubiana 7 May 11, 2012 - LZ
Early X-ray spectral fits NGC 1313 X-1 With a BH this massive, accretion would be in the usual sub- Eddington regime 9
ULX X-ray spectra Many ULXs show either curved X-ray spectra or a turnover at ~3-5 kev, sometimes with a soft excess below 1 kev (ultraluminous state) Gladstone et al. (09) Walton et al. (11) Fitted with an optically thick corona covering or replacing the disc, or a modified accretion disc (e.g. slim disc) launching a wind (Middleton et al. 11) Unusual accretion state at Mdot >~ MdotEdd Curvature interpreted also in terms of relativistically blurred ionized reflection from the accretion disk (Caballero-Garcıa & Fabian 2010) ULXs Lubiana 10 May 11, 2012 - LZ
ULX X-ray timing Suppressed variability on short time scales in some ULXs (Heil et al. 09) QPOs+broad band noise in 2 persistent + 1 transient ULX M82 X-1 54-166 mhz NGC5408 X-1 20 mhz X42.3+59 in M82 3-4 mhz (Strohmayer & Mushotzky 03; Strohmayer et al. 07; Feng et al. 10) Mucciarelli et al. (06) For persistent ULXs rms, coherence, noise, variability similar to Type C QPOs in BHCs. An identification with the ultra-lfqpo seen occasionally in GRS 1915 recently proposed (Middleton et al. 11) Attempts to estimate Mbh by scaling characteristic frequencies, extrapolating correlations known to exist for BH binaries (Fiorito & Titarchuk 04; Mucciarelli et al. 06; Strohmayer et al. 07; Feng & Kaaret 07), using nondetection of variability power (Goad et al. 06) or the variability plane of BHCs and AGNs (Casella et al. 08) Complete characterization of the ULX QPOs still missing! 12
ULXs: Dynamical BH mass BH mass measurable directly from radial velocity variations of spectral lines (HeII 4686 line; Pakull et al. 06, Grise et al. 08) but variations do not appear sinusoidal (Roberts et al. 11) VLT/FORS1 Roberts et al. (11) Mucciarelli et al. (05) 14
ULXs: some recent advancements Optical modulation of 6 days in NGC 1313 X-2 (Liu et al. 09) - Only 3.5 sigma significance (Zampieri et al. 11) - System consistent with a 12-15 Msun main sequence donor dumping matter on a 50-100 Msun BH (Patruno & Zampieri 10) Zampieri et al. (2011) Accurate photometric monitoring of the optical counterpart (with simultaneous X-ray observations) needed to determine P (ellipsoidal modulation and X-ray irradiation equally important) 15
ULXs: some recent advancements HLX-1 in ESO243 49 (Farrell et al. 09), with an inferred isotropic L~1.0e42 erg/s, interpreted as first unambiguous identification of an IMBH (Godet et al. 09; Webb et al. 10; Wiersema et al. 10). Wiersema et al. (10) Farrell et al. (11) 16
ULXs: some recent advancements Lasota et al. (2011) Servillat et al. (2011) Light curve interpreted as due to bursts of mass transfer occurring when the donor, in an eccentric orbit, grazes the tidal radius The relatively low disk temperature in the thermal state (0.2-0.3 kev) suggests the presence of an IMBH of few x 1000 up to a few x 10000 Msun (Davis et al. 11; Godet et al. 11) 17
ULXs: some recent advancements z Broadband spectrum (from X-rays through IR) not consistent with simple and irradiated disc models Well described by a model comprised of an irradiated accretion disc plus a 1.0e6 M stellar population Old or young? A globular cluster or the nucleus of a dwarf galaxy that underwent a gas-rich minor merger (Soria et al. 11; Mapelli & Zampieri 11)? Farrell et al. (2011) 18
IMBHs: origin? Remnants from the collapse of Population III stars formed in cosmological epochs (Madau & Rees 2001) In globular clusters, through repeated mergers of stellar mass BHs (Miller & Hamilton 2002) In young, dense stellar super clusters, from the dynamical collapse of supermassive stars in their centres (e.g. Portegies Zwart et al. 2004) Apart from a handful of hyperluminous objects, observationally BHs of several hundreds to thousands Msun are not required for the majority of ULXs 20
ULXs and low metallicity environments Subsample of galaxies with X-ray pointed observations (ROSAT, Chandra, XMM) from the Catalog of Neighboring Galaxies (Karachentsev et al. 04) Specific ULX frequency decreases with increasing host galaxy mass indicating that smaller, lower metallicity systems have more ULXs per unit mass (Swartz et al. 08; Walton et al. 11) Swartz et al. (2008) 21
ULXs and low metallicity environments (+) Specific ULX frequency decreases with increasing host galaxy mass indicating that smaller, lower metallicity systems have more ULXs per unit mass (Swartz et al. 08) (+) Line intensities of HII regions in ULX host galaxies (e.g. Mapelli et al. 10, using Pilyugin et al. 04 calibration) 0.1-0.5 Zsun for a sample of 52 ULX-hosting galaxies (+) Measurement of the metallicity from the stellar environment (NGC 4559 X-7, Cropper et al. 04; NGC 1313 X-2, Grise et al. 08) or from the optical spectrum of bubble nebulae (Ho II X-1, Pakull & Mirioni 02; NGC 1313 X-2, Ripamonti et al. 11) subsolar abundance NGC 1313 X-1 (?) Oxygen abundance from K-shell photoionization edges of high S/N ratio XMM Pintore spectra & Zampieri of (11) 14 ULXs solar abundance (Winter et al. 07) Analysis repeated on two ULXs with higher counting statistics and with a different spectral model ~0.5 Zsun (Pintore & Zampieri 11) 22
ULX population N ULX Statistical analysis of a sample of galaxies hosting ULXs plus a number of galaxies of the Local Group (excluding ellipticals) 66 galaxies observed in X-rays with measured SFR and Z red: Z > 0.2 Zsun black: Z < 0.2 Zsun Mean Z = 0.22+/-0.11 Strong correlation between Nulx and SFR (Grimm et al. 03; Mineo et al. 10) Nulx ~ A*SFR A=1.20+/-0.2 Mapelli et al. (2010) Fit of Nulx/SFR vs Z using a power-law wrt a fit with a constant: improvement significant at the 96% conf. level 23
ULXs from massive low-z stars Massive BH formation through direct collapse of massive low-z stars Mbh~30-80 Msun Bressan et al. (2011) Fryer (99) Zampieri & Roberts (2009) Small (~0.5) beaming or L ~ 2-3 Ledd for bright (> 10 40 erg/s) ULXs 24
Dynamics of MSBHs and ULX displacements (1) Simulations of 25 realizations of isolated star clusters hosting MSBH binaries with massive donors Starlab code (Portegies Zwart et al. 2001) Nstars=5000, Mstars=4000 Msun, initial binary fraction=0.1, Salpeter IMF (0.08-120 Msun) Stellar and binary evolution not considered (work in progress) 1 MSBH in each cluster (as expected at Z=0.1 Zsun; Mapelli et al. 10) with Mbh=50 Msun Donor masses uniformly distributed between 10 Msun (young systems) and 50 Msun (stable mass transfer) a>0.1 AU (to avoid merger) and a<10 AU (only hard binaries) Snapshot of a Starlab simulation (256-body, W0=12 King model, Salpeter mass function ) 25
Dynamics of MSBHs and ULX displacements (3) Mapelli et al. (11) Mapelli et al. (11) Mapelli et al. (11) Evolution after 10 Myr MSBH of 50 Msun 14/25 (56%) MSBHs bound to the cluster, 70% of them in dynamically unstable triple systems 11/25 (44%) MSBHs ejected, 80% of them in binary systems (30% exchanged companion) Most ejections between 2.5 and 6.5 Myr (companion on the MS) Kicks smaller than assumed natal kicks of stellar-mass BHs (Linden et al. 10), but significant for direct-collapse BHs Fractional decrease of a and remarkable agreement with observed offset of ULXs 27
Dynamics of IMBHs and ULX displacements Mapelli et al. (11) Simulations repeated with an IMBH of 300 Msun in each cluster Evolution after 10 Myr Distance from the center larger than in the initial conditions IMBHs pushed at the outskirts because of three body encounters However, only 1 IMBH (5% of the sample) ejected Might be stripped by tidal external fields on longer timescales, but no significant recoil velocity Large observed ULX offsets not reproduced If HLX-1 is in a cluster, we do not expect it to be significantly displaced from its center 28
Conclusions ULXs are accreting BHs in binaries: either a different class of BHs (wrt Galactic BHs) or a different accretion state The brightest ULXs may contain IMBHs (e.g. HLX1 in ESO 243-49) Direct collapse of massive stars formed in low-metallicity environments may produce MSBHs that power a fraction of ULXs Crucial contribution of future X-ray missions and optical follow-up 36
ULXs: future developments in X-rays X-ray spectral and timing results (XMM, Chandra, Suzaku, Swift) remain consistent with various alternative scenarios What can we look forward to with future/planned X-ray missions? Characterize the spectral states and spectral evolution of ULXs (spectral measurement above 10 kev needed) ASTRO-H, NuSTAR Investigate the timing properties of bright ULXs that show short term variability LOFT Survey the X-ray sky for new ULXs and HLXs e-rosita Identify bright, transient ULXs through monitoring LOFT (MAXI) Transient ULX detection capability of LOFT/WFM in 50 ks: - ULX with L~1.0e41 erg/s at D~5 Mpc - HLX at D~15 Mpc 37