IGR J17091-3624

• confirm the low absorption (Nh<1e22cm-2)
• Comptonised spectrum,
• second epoch shows the presence of bright thermal component in soft X-rays.
• transition form a hard state to a soft state.
• 3rd epoch => spectra can be fitted with a Comptonised model: transition back to the hard state.
• Hysteresis in transitions

• 11.6+/-0.6 mCrab in 100-150 keV.

• Swift observations => absorbed disk blackbody, with N_H=0.51e22 cm-2, and kT = 2.1 keV.
• 0.2-10 keV unabsorbed flux from the source ~1e-10erg/s/cm2
• power law yields a good fit, with parameters similar to those observed in the past during a hard state.
• 0.2-10 keV unabsorbed flux ~9 E-10 erg/s/cm2
• Not associated with tentative radio counterpart of Rupen et al., Pandey et al.

• Infrared observations
• 2MASS source formerly suggested to be counterpart is outside Swift error box. Source is NOT the F8 V star.
• 2 candidates in NIR images, favour 1 as the counterpart
• J=16.73, H=15.83, K_s=15.36
• R_star/D_star very low => maximum radius 12.8 R_sun
• SED fitting => Av=6.8, T=6900 K
• Not a Giant, nor a superGiant
• Probable LMXB in Gal. Bulge

• Quiescence with upper limit on the luminosity of L< 7e32 erg/s => source most probably hosts a black hole.
• Begining of a new outburst seen with Swift.
• Refute the previously suggested radio counterpart for this source.
• Re analysis of VLA data => new radio counterpart at a position consistent with that of the Swift source.
• Radio spectrum is inverted (ie self absorbed): typical of the hard states of black hole.
• Swift spectrum: hard spectral state
• Source then softens, black body component needed in fits

• Renewed activity seen with Swift/BAT
• Swift observation => No flux from IGR J17098-3628
• IGR J17091-3624 is strongly detected in the XRT image
• Spectrum: absorbed power law N_H = 1.10 x 10²² cm-2 and Γ= 1.73
• The source is not detected by UVOT

• Detection of the optical and near-infrared counterparts
• I = 18.35 during outburst while in the 2005 images I = 20.32
• Detection of a Ks=16.65 source consistent with the optical counterpart.
• Finding chart here (courtesy M.A.P. Torres)

• ATCA observation => presence of a single radio source within the X-ray error circle consistent with the optical counterpart
• Flux densities of ~1.3 mJy at 5.5 and 9 GHz => consistent with flat radio spectrum, usually associated self absorbed compact jets
• Swift XRT spectra closest to the radio observation: 2011-02-09
  N_H = 0.72 x 10²² cm-2, Γ = 1.41
  2011-02-10
  N_H = 0.74 x 10²² cm-2, Γ = 1.52
• Parameters + inferred luminosity (at 8 kpc) => in the radio/X-ray correlation diagram between the standard track and the track possibly indicative of a radiatively efficient black hole in the hard state

• Timing analysis of 4 RXTE observations
• Data contaminated by GX 349+2, but manage to roughly correct its influence
• All PDS but the one from the first observation => ~0.1 Hz QPO, with an increasing RMS amplitude (from 3.2% to 4.5%).
• Lend further credence to the fact that IGR J17091-3624 is a black hole binary in a hard state

• Fast evolution of the QPO seen in this source: 0.1 Hz QPO is observed at 0.34, 0.45 and 1.66 Hz in the three most recent observations (Feb 18,19 and 20, 2011)
• QPO harmonic also observed at ~3.3 Hz (last observation)
• Hardness is seen to decrease=> spectrum is getting softer
• Behavior consistent with transition to the soft state

• Swift and INTEGRAL observations of Feb 28, 2011
• XRT data alone need a disc in the model fit
• Joint 0.8-200 keV XRT+IBIS/ISGRI spectrum <=> absorbed (disc + powerlaw) with N_H~1.00 x 10²² cm-2, kT_disc~1.0 keV and Γ~2.2
• No high-energy cut-off requested up to 200 keV
• Presently in the canonical high/soft spectral state of BHCs

• Discovery of a ~10 mHz QPO most likely from IGR J17091-3624
• QPO is only intermittently visible with an average 2-60 keV fractional rms amplitude of 5.3%
• Fractional rms amplitude increases from ~3.5% below 5 keV to a maximum of ~10% in the 15-30 keV range
• 2 QPOs at 2.3 Hz and 8.0Hz, with fractional rms amplitudes of 10.9% and 6.7%, are found in the same obs.

• INTEGRAL => IGR J17098-3628 is not active, while J17091-3624 is
• mHz QPO is from IGR J17091-3624

• mHz QPOs have increased in frequency by a factor of ~3
• PCA light curve displays flaring events occurring at a rate of 25-30 mHz. Wide diversity in the flare profiles
• Flares are also observed with Swift/XRT (IGR J17098-3628 not detected by Swift)
• Reminiscent (while of lower flux) of the "heartbeat" oscillations seen in the black hole candidate GRS 1915+105.

• Multiple peaks with evolving frequencies in RXTE observations
• Comparison to the class rho of GRS 1915+105

• Heartbeat oscillations observed in all (but 3) observations performed since March 19th 2011
• Frequency of the (heartbeat) oscillations increased with time (up to ~100 mHz 2011/04/18)
• Also associated with a loss of regularity.
• On 04/19: broad variety of complex behavior similar to the class "beta" of GRS 1915+105 (different timescale)
• QPO at 5-8 Hz occurring only during low intensity periods are also present in this obs.

• In rho-type oscillations both small and large amplitude oscillations are seen
• Intervals of large bursts and short bursts appearing alternatively
• Characteristic time scale is ~40.25 sec
• Peak-to-dip count rate ratio ~3-5 for large bursts ; 1.5-2 for short bursts
• From 26 May 2011 to 03 June 2011 # of large burst intervals and their maximum peaks seen to decrease=>disappeared on 04 Jun 2011.

• 1^st four simultaneous multi-wavelength observations (ATCAn Swift, INTEGRAL, RXTE) of this source, during the 2011 outburst
• Hard shard state in the first 2 observations, soft intermediate state later
• Luminosity at the state transition => distance to the source of ~11 to ~17 kpc (assuming BH mass of 10 Msun).
• The radio vs X-ray luminosity properties similar to other black hole binaries if distance greater than 11 kpc
• Radio => Preferred distance of 17 kpc.
• Observed radio-X-ray behaviour more typical to that of GX 339-4 or H1743-322 than that of GRS 1915+105

• RXTE light curves: dips lasting 20-30s similar to GRS 1915+105 in ω and θ classes
• Peak in hard color coincident with an intensity dip, instead of usual (in BHB) accompanying dip in hard color

• Detailed analysis of the 2011 RXTE observations
• Identify 7 patterns of variability in light curves, similar to classes seen in GRS 1915+105
• Typical timescales of variability & brightness respectively shorter and fainter compared to 1915+105
• Either models based on Eddington Luminosity fail for IGR J17091-3624, or distance well above 20 kpc, or harbours a very light black holes

• Observations with Chandra
• High resolution spectra =>presence of absorption line attributed to a fast disk wind component
• Simultaneous EVLA observation => jet activity quenched during Chandra observations
• Presence of a disk wind when in soft state compatible with what observed in other such sources

• Discovery of a high frequency QPO
• Frequency = 64.8 Hz very similar to the 67 HZ QPO of GRS 1915+105
• Potential feature at 164 Hz, strengthening the similarity to the 1915's HFQPOs
• Implication of these findings are discussed in the view of the potential high distance, or low mass BH in the IGR source

• HFQPO due to non-linear resonances of oscillating modes in the disc, in 5:2 ratio => BH mass ~ 6 M_sun (comparison with 1915+105)

• Analysis of RXTE data of the 2011 outburst
• Identification and comprehensive characterisation of 2 variability classes never seen in any other black hole binaries, including GRS 1915+105

• analysis of the long term Swift and INTEGRAL monitoring
• Up to the transition to the soft state => HID and rms-intensity diagram show similar pattern to other sources
• After behaviour more similar to that of GRS 1915+105
• Spectral analysis of the data => heartbeat pattern corresponds to rapid emptying and refiling of the inner region of an accretion disc.
• In addition to being at a large distance, the source may also have a large inclination, which would explain its faintness.

• Chandra observation => refined X-ray position
• Source position compatible with all previously suggested counterparts => securing source as a μquasar

• Re-analysis of 2 XMM observations during quiescence
• Contrary to what previously claimed source detected at 0.5-10 keV fluxes of 9 and 12 e-14erg/cm2/s.
• In both observations Γ~1.6 (absorption fixed).
• Assuming relation between quiescent luminosity and orbital period holds => orbital period between >4 d (at 10 kpc) and tens of days in case of a larger distance

• Analysis of simultaneous RXTE and XMM observations
• Phase resolved spectroscopy of the ρ-like type of variability
• Comparison to GRS 1915+105 and investigation of differences
• Source is a high inclination system (> 53deg)
• Low luminosity explained only if the spin of the BH has a low (or even negative) value, in a distant and light (<5 M_sun) BH

• XMM spectral analysis => presence of a number of absorption lines and a distinct feature at 7.1 keV
• Photoionization model provides a good fit, but a relativistic iron line model provides a better modeling
• System probably in a soft intermediate state during this obs
• Disk temperature ~1.1 keV and a power law Γ~1.9
• Analysis of the broadened iron line => inclination to the disc < 62°ree; with emissivity index Q<7.4
• Inner radius is constrained to be <2.3GM/c2 => black hole with a spin a*>~0.9 (in contradiction with Rao & Vadawale)

• Spectral and timing studies of the RXTE observations with cycles and in depth comparison with GRS 1915+105
• The properties of hard X-ray dips of 17091 consistent with those of 1915's abrupt transitions
• Correlation between X-ray cycle time and time spent in hard dip in the two sources
• Difference between hard dips and peak spectra due to variation of the (p-free) accretion disk
• Oscillations due to limit cycle beaviour of unstable radiation pressure dominated inner disk

• Heartbeat oscillations => radiation pressure instability of accretion disk
• Disc outflow (wind) => plausible mechanism to regulate (and stop) oscillations
• Wind properties (Chandra) => 2 wind components with rates ~2.7 e17 g/s and 4.2e17 g/s
• Launching sites => between 380 and 4700-5900 R_G and 950 and 4200-4900R_G
• Modeling gives parameter (outflow rates) compatible with observations
• Geometry of disc-corona-wind presented

• Spectro-temporal study of RXTE and Swift data
• 3 different methods => 8.7 M_Sun < M_BH < 15.6 M_Sun @ 90% confidence