The faint "heartbeats" of IGR J17091-3624: an exceptional black-hole candidate
D. Altamirano, T. Belloni, M. Linares, M. van der Klis, R. Wijnands, P. A. Curran, M. Kalamkar, H. Stiele, S. Motta, T. Munoz-Darias, P. Casella, H. Krimm
aa r X i v : . [ a s t r o - ph . H E ] D ec Draft version October 1, 2018
Preprint typeset using L A TEX style emulateapj v. 04/21/05
THE FAINT “HEARTBEATS” OF IGR J17091–3624: AN EXCEPTIONAL BLACK-HOLE CANDIDATE
D. Altamirano , T. Belloni , M. Linares , M. van der Klis , R. Wijnands , P. A. Curran , M. Kalamkar , H.Stiele , S. Motta , T. Mu˜noz-Darias , P. Casella , H. Krimm Draft version October 1, 2018
ABSTRACTWe report on the first 180 days of RXTE observations of the outburst of the black hole candidateIGR J17091–3624. This source exhibits a broad variety of complex light curve patterns includingperiods of strong flares alternating with quiet intervals. Similar patterns in the X-ray light curveshave been seen in the (up to now) unique black hole system GRS 1915+105. In the context of thevariability classes defined by Belloni et al. (2000) for GRS 1915+105, we find that IGR J17091–3624shows the ν , ρ , α , λ , β and µ classes as well as quiet periods which resemble the χ class, all occurringat 2-60 keV count rate levels which can be 10-50 times lower than observed in GRS 1915+105. Theso-called ρ class “heartbeats” occur as fast as every few seconds and as slow as ∼
100 seconds, tracinga loop in the hardness-intensity diagram which resembles that previously seen in GRS 1915+105.However, while GRS 1915+105 traverses this loop clockwise, IGR J17091–3624 does so in the oppositesense. We briefly discuss our findings in the context of the models proposed for GRS 1915+105 andfind that either all models requiring near Eddington luminosities for GRS 1915+105-like variabilityfail, or IGR J17091–3624 lies at a distance well in excess of 20 kpc or, it harbors one of the leastmassive black holes known ( < M ⊙ ). Subject headings:
X-rays: binaries — binaries: close — stars: individual (IGR J17091-3624, GRS1915+105) — Black hole physics INTRODUCTION
Observations with the
Rossi X-ray Timing Explorer (RXTE) have led to extraordinary progress in the knowl-edge of the variability properties of many different typesof sources, particularly of black hole candidates (BHCs)and neutron stars (NSs) in low-mass X-ray binaries (e.g,van der Klis 2006). Both BHCs and NS are knownto exhibit distinct “accretion states” (usually definedin terms of their X-ray spectral shape and variability),whose characteristics are thought to be intimately re-lated to the physics of the accretion flow and its in-teraction with the central compact object (e.g., Belloni2010; Remillard & McClintock 2006; van der Klis 2006;Belloni et al. 2011).Of all Galactic BHC X-ray binaries known today,one of the most prolific in terms of state transitions isGRS 1915+105. In outburst since its discovery in 1992with WATCH (Castro-Tirado et al. 1992), it is a 33 daysorbital period binary system (Greiner et al. 2001) har-boring a 14 ± . M ⊙ black hole (Greiner et al. 2001;Harlaftis & Greiner 2004). At a distance of ∼ . Email: [email protected] ; Astronomical Institute, “AntonPannekoek”, University of Amsterdam, Science Park 904, 1098XH,Amsterdam, The Netherlands INAF-Osservatorio Astronomico di Brera, Via E. Bianchi 46,I-23807 Merate (LC), Italy Massachusetts Institute of Technology - Kavli Institute for As-trophysics and Space Research, Cambridge, MA 02139, USA Laboratoire AIM, CEA DSM/IRFU/SAp, Centre de Saclay,F-91191 Gif-sur-Yvette, France Instituto de Astrof´ısica de Canarias, 38200 La Laguna, Tener-ife, Spain School of Physics and Astronomy, University of Southampton,Southampton, Hampshire, SO17 1BJ, United Kingdom CRESST and NASA Goddard Space Flight Center, Greenbelt,MD 20771, USA ; Universities Space Research Association, 10211Wincopin Circle, Suite 500, Columbia, MD 21044, USA ten at Eddington or super-Eddington luminosity (e.g.,Done et al. 2004).GRS 1915+105 is known to show quasi-periodic oscil-lations (QPOs) in the 0.001–70 Hz range, some of whichare the same as those generally seen in other BHCs(Morgan et al. 1997; Markwardt et al. 1999; Reig et al.2000; Strohmayer 2001; Soleri et al. 2008). However, sofar GRS 1915+105 has been unique in that its X-raylight curves exhibit more than a dozen different pat-terns of variability usually called “classes” (which are re-ferred to with Greek letters), most of which are high am-plitude and highly-structured (e.g., Belloni et al. 2000).Most of this structured variability is thought to be dueto limit cycles of accretion and ejection in an unstabledisk (see, e.g., Belloni et al. 1997; Mirabel et al. 1998;Tagger et al. 2004; Neilsen et al. 2011, and referenceswithin). Much effort has been expended in order to un-derstand why GRS 1915+105 is so unusual among BHCs(e.g., Fender & Belloni 2004). It has been proposed thatthe high accretion rate estimated for GRS 1915+105might be the determining factor (e.g., Done et al. 2004);however, the lack of any other source showing similarcharacteristics has prevented definite conclusions.IGR J17091–3624 was discovered with INTE-GRAL/IBIS during a Galactic Center observation onApril 2003 (Kuulkers et al. 2003) and later again in 2007(Capitanio et al. 2009). Reexamination of archival datafrom different missions showed that IGR J17091–3624was also active in 1994, 1996, 2001 (Revnivtsev et al.2003; in’t Zand et al. 2003; Capitanio et al. 2006).Although a neutron star could not be excluded basedon spectral characteristics and the outburst radio/X-ray flux ratio, Capitanio et al. (2006) concluded thatIGR J17091–3624 is most probably a black hole. A newoutburst was detected with
Swift/BAT in February 2011(Krimm et al. 2011); the radio/X-ray characteristics BA T - I n t en s i t y ( c oun t s c m - s - ) R X T E - k e V I n t en s i t y ( m C r ab ) Time (MJD) R X T E - k e V F l u x ( - e r g c m - s - )
20 30 40 50 60 70 55600 55650 55700 55750 1.5 2.5 3.5
Fig. 1.—
Upper Panel : Swift/BAT light curve.
Lower Panel : Circles: RXTE Crab normalized 2-20 keV intensity. RXTE values areaveraged per observation, background subtracted and dead-time corrected. Triangles: RXTE 2-50 keV flux RXTE observations betweenday 55601 and 55615 are affected by the nearby source GX 349+2 (Sco X-2); the intensity during this period was arbitrarily fixed to 65mCrab and no X-ray flux is reported. Shaded areas mark quiet periods (see text). during the first ∼
40 days, combined with the discoveryof QPOs (Rodriguez et al. 2011a), further suggestedthat IGR J17091–3624 is a black hole.Recently we reported the discovery of 10 mHzQPOs in RXTE observations of IGR J17091–3624(Altamirano et al. 2011a) very similar to those inGRS 1915+105. The suggested link between thesesources was strengthened by our discovery of (i) a contin-uous progression of regular, quasi-periodic flares occur-ring at a rate of 25–30 mHz (Altamirano et al. 2011b),and (ii) a broad variety of complex patterns alternatingwith quiet intervals (Altamirano et al. 2011b,c), resem-bling respectively, the so called ρ class (“heartbeat”) os-cillations and the complex β class patterns, both so farseen only in GRS 1915+105.The existence of a source showing similar X-ray vari-ability to that seen in GRS 1915+105 opens a new win-dow of opportunities to understand the physical mecha-nism that produces the highly structured X-ray variabil-ity. Multi-wavelength and high spectral resolution stud-ies of IGR J17091–3624 as compared with GRS 1915+105can help gaining further insights into the role of disc-jetcoupling (e.g., review by Fender & Belloni 2004) and ac-cretion disk winds (e.g, Neilsen et al. 2011) in accretingBHCs. Furthermore, it allows the possibility to test therole of the accretion disk size, evolutionary state of thecompanion star, and the evolution of the disc structureas an explanation of the long- and short-term X-ray vari- ability seen in GRS 1915+105 (e.g., Done et al. 2004).In this Letter we use RXTE data to describe thephenomenological similarities between IGR J17091–3624and GRS 1915+105. We demonstrate that IGR J17091–3624 shows the same type of high amplitude andhighly-structured variability previously observed only inGRS 1915+105, although at much lower count rates. Theanalysis presented here is the first step of a larger pro-gram to compare IGR J17091–3624 and GRS 1915+105in detail using data obtained with RXTE , Swift , XMM-Newton as well as optical facilities. OBSERVATIONS, DATA ANALYSIS
IGR J17091–3624 is observed with the ProportionalCounter Array (PCA; Zhang et al. 1993; Jahoda et al.2006) on-board RXTE almost daily since the outburstbegan in February 2011 (at the time of submission of thisLetter, IGR J17091–3624 was still active). We use thefirst 147 observations, covering ∼
180 days. We also used1700 archival RXTE observations of GRS 1915+105.Power spectra and light curves were produced fromthe PCA using standard techniques (e.g., Belloni et al.2000; Altamirano et al. 2008). Deadtime corrected en-ergy spectra averaged over each observation were cre-ated from PCU 2 data; a systematic error of 1% wasadded to all channels. Response matrices were createdwith
PCARSP (V11.7.1) using the position reported byKennea & Capitanio (2007), thereby taking into accounta constant 0.4 ◦ pointing offset used after February 23rd, Fig. 2.—
The upper and lower panels of each frame show a lightcurve of IGR J17091–3624 and GRS 1915+105, respectively. Countrates are given in 1 s bins, per PCU, 2-60 keV and background sub-tracted. The IGR J17091–3624 A-C intervals last T T R ≡ MIN
GRS /MIN
IGR J − , where MIN is an average count rate during minima in the above lightcurve and is given for each frame as an order of magnitude estimateof the flux ratio. Greek letters on the side indicate variability theclass.
Fig. 3.—
Similar to Figure 2. The IGR J17091–3624 D-F in-tervals come from observations 96420-04-03, -08-03 and -09-06, re-spectively. For GRS 1915+105 they come from observations 20187-02-01-00, 95701-01-31-00 and 10258-01-10-00, respectively.
100 105 110 115 120 125 0.32 0.33 0.34 0.35 0.36 0.37 0.38 0.39 0.4 0.41 I n t en s i t y ( c t s / s e c / P CU ) Hard color 2000 3000 4000 5000 6000 7000 8000 9000 10000 0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85 I n t en s i t y ( c t s / s e c / P CU ) Hard color
GRS 1915+105IGR J17091−3624
Fig. 4.—
Left panels show the hardness-intensity diagram for flares observed during the ρ variability class in IGR J17091–3624 (top;ObsID 96420-01-04-03) and GRS 1915+105 (bottom; ObsID 96378-01-01-00) occurring at an average period of T1=70.96 seconds andT2=63.72 seconds, respectively. Arrows mark the time evolution. Inset shows representative flares. Light curves and colors are estimatedfrom 1 sec averages. Intensity is the count rate in the 2-60 keV range (absolute channels 0-240) and hard color is the 6.5–15.0 keV /2–6.5 keV count rate ratio (channels 15-35 and 0-14, respectively). Right panels show representative power spectra from averages of 512 secsegments during the ρ variability class for IGR J17091–364 (top, ObsID:96420-01-05-000, MJD 55647.9) and for GRS 1915+105 (bottom,ObsID:40703-01-07-00, MJD 51235.3). phabs*(diskbb + powerlaw )) using Xspec12 (Arnaud 1996) with N h fixed to 1 . × cm − (Krimm et al. 2011); some spectra required a Gaussianat 6.5 keV for a good fit. RESULTS
Figure 1 (upper panel) shows the Swift/BAT(Barthelmy et al. 2005) daily light curve of the first200 days of the outburst (assuming outburst onset onMJD ∼ /s (15-50 keV) ∼
27 days after onset;then decreased and remained below 0.01 counts/cm /s.The lower panel of Figure 1 shows the 2-20 keV, Crab-normalized intensity and the 2-50 keV un-absorbed fluxfrom each RXTE/PCA pointed observations. RXTEobservations started 21 days after onset. The first10 observations (until MJD 55615) are affected by thecontribution of the bright, variable source GX 349+2 (Sco X-2) which was within the 1 ◦ PCA FoV (see, also,Rodriguez et al. 2011a). For reference, in Figure 1 weshow these observations at an arbitrarily fixed intensityof 65 mCrab; we do not report their X-ray flux.From the variability point of view, we find that dur-ing MJD 55601–55632 , the X-ray light curves were wellcharacterized by broad band noise and a QPO movingin time from 0.1 Hz up to ∼ ∼ ν class, and panels Band C two different varieties of ρ class. The ν and ρ variability classes in GRS 1915+105 are characterized byquasi-periodic “flares” recurring on time scales between ∼
40 (e.g., Massaro et al. 2010; Neilsen et al. 2011) and ∼
120 seconds (Belloni et al. 2000). The main differencesbetween the ν and ρ classes are that (i) the shape andperiod of the flares in ν can be more irregular than in ρ and (ii) that the ν flares show a characteristic structurein their profiles, notably a dip followed by a secondarypeak after the main one (see Figure 2).Panel D shows segments of the α class, in which“rounded-bumps” are sometimes accompanied by sharppeaks which last a few seconds. Note that in our ob-servations of IGR J17091–3624 we do not see the typical ∼ α class of GRS 1915+105 (not shown, see, e.g., Figure 19din Belloni et al. 2000). Panel E shows light curves fromeither the β or the λ class. These light curves consistof quasi-periodic alternation of low-quiet periods withhighly-variable and oscillating ones. Panel F shows seg-ments of the µ class, where we observe periods of veryrapid (factor of 2-4 changes in count rate in less than 5seconds) and almost incoherent variability.In addition to the light curves shown in Figures 2 and 3, IGR J17091–3624 displays complex light curves com-bining characteristics of different classes. It also shows quiet periods in which 1-sec light curves are flat ( < < Swift/BAT , i.e. when the spectrum ishard. In terms of the GRS 1915+105 variability classes,these periods could correspond to the χ class.Figures 2 and 3 not only shows the remarkable simi-larities in the shapes of light curves from both sources,it also shows two clear differences: (i) the time scalescan be different (IGR J17091–3624 tends to be faster,see below) and (ii) the average count rate (or flux)of the source can be much higher (factor 10-50) inGRS 1915+105. Some of the spectral and timing anal-ysis performed for GRS 1915+105 on time scales of sec-onds (e.g., Markwardt et al. 1999; Belloni et al. 2000;Done et al. 2004; Soleri et al. 2008) is prevented by thecombination of relatively low count rate and faster vari-ability in IGR J17091–3624.Given that we find the ν and ρ variability classes (pan-els A-C) in about 35% of our observations (another ∼ ∼
30% area mix of the α , β , µ , λ and unclassified ones) and that ρ is one of the best studied classes in GRS 1915+105, inthe rest of this Letter we constrain ourselves on furthercomparison of the ν and ρ classes between sources. Moredetailed comparison of the other variability classes willbe presented in upcoming papers.Some of the observations clearly show only the ρ orthe ν types of flares, some show a mix and sometimesdifferentiating the two classes is difficult due to the lowstatistics. In any case, these flares can occur as fastas every few (2-5) seconds (e.g., ObsID: 96420-01-22-04,MJD 55768), and as slow as every ∼
100 seconds (e.g., ObsID 96420-01-03-01, MJD 55634). This means thatthe oscillations in IGR J17091–3624 can be faster thanthose in GRS 1915+105, but not as slow. The fractionalrms amplitude of the flares covers a range from ∼
2% toup to 50-60%. If one assumes that the minimum pe-riod that a quasi-periodic feature can reach scales pro-portional to some power of the mass of the compact ob-ject (see, e.g., Belloni et al. 1997; Frank et al. 2002), andthat a 2-5 s recurrence time of the ρ / ν flares IGR J17091–3624 versus ∼
40 s in GRS 1915+105 is due to a differ-ence in mass, then our results suggest the black hole inIGR J17091–3624 could be a factor of a few less massivethan the 14 ± . M ⊙ of GRS 1915+105.The variability classes are known to exhibit distinc-tive spectral evolution (e.g., Belloni et al. 2000). Ina color-color diagram (CD) or hardness-intensity dia-gram (HID), one observes that each flare from the ρ (and sometimes ν ) class traces a loop (or “ring”, seeVilhu & Nevalainen 1998; Belloni et al. 2000). Figure 4(bottom-left) shows the HID for ten consecutive flaresfrom a single observation of GRS 1915+105 (inset showsa representative flare). The loop in the HID is alwaystraversed clockwise.The HIDs and CDs from single flares of IGR J17091–3624 are dominated by low statistics, so we could nei-ther exclude nor confirm loop-like patterns similar tothose observed in GRS 1915+105. To improve statistics,we used intervals where more than 10 flares occurredapproximately periodically and folded each interval atthe best average period. Although this process washedout some of the structure in the light curves (due tothe quasi-periodicity of the signal and the profile differ-ences between flares), all the HIDs we created using thesame energy bands as in GRS 1915+105 (which are fixedby the observing modes used) always showed a loop re-sembling that seen in GRS 1915+105. In the upper-leftpanel of Figure 4 we show a representative example (in-set shows the average profile of the flare). However, inall cases the loop is traversed in an anti-clockwise sense,i.e., opposite to what we see in GRS 1915+105. As thehardness ratios are a crude characterization of the spec-trum, detailed spectral modeling of these loops (e.g.,Neilsen et al. 2011) are needed to understand whetherthere is a physical difference between the ρ class seen inboth sources.Figure 4 (right panels) shows representative powerspectra of the ρ variability class in IGR J17091–3624 andGRS 1915+105 . The power spectra share the same mainfeatures: (i) a low-frequency QPO (with high harmoniccontent) due to the “flares” and (ii) a QPO with char-acteristic frequency between 6-10 Hz (e.g., Muno et al.1999). In addition, for both sources we sometimes finda “bump” with characteristic frequency between 1 and 5Hz. DISCUSSION
An extensive literature exists attempting tounderstand the complex variability observed inGRS 1915+105. Some authors propose that thehigh luminosity (close to, or super-Eddington) of GRS 1915+105 is the determining factor(e.g., Belloni et al. 1997; Vilhu & Nevalainen 1998;Belloni et al. 2000; Nayakshin et al. 2000; Janiuk et al.2002; Done et al. 2004; Neilsen et al. 2011, and refer- C o m pa c t ob j e c t m a ss ( M o ) Distance (kpc)F bol = 3 x F F bol = 2 x F F bol = 1 x F Fig. 5.—
Mass of the compact object versus distance to the bi-nary system assuming IGR J17091-3624 is emitting at Eddington.The three curves correspond to different (1, 2 and 3) correctionfactors between the 2-50 keV and the bolometric flux. ences therein). Although it is generally accepted thatthe complex X-ray variability in GRS 1915+105 resultsfrom disk instabilities, the exact nature of the instabilityremains unknown. The lack of at least a second sourceshowing similar characteristics has prevented definiteconclusions.In this Letter we show for the first time that anothersource, the BHC IGR J17091-3624, can show the samebroad variety of complex light curves as GRS 1915+105(at least 7 of the 12 variability classes observed inGRS 1915+105, in addition to unclassified ones; notethat at the time of submission of this Letter IGR J17091-3624 is still active). Although the comparisons presentedin this paper constitute only a first step, the observedsimilarities suggest that the complex light curves of thetwo sources are produced by the same physical mecha-nisms. If true, the low flux of IGR J17091-3624 comparedwith GRS 1915+105 combined with the circumstancethat currently neither the distance to IGR J17091-3624nor the mass of its compact object are known, raisesthe fundamental question: is IGR J17091-3624 close toEddington or not at times when showing the same char-acteristic X-ray variability?A scenario in which IGR J17091-3624 is not emit-ting at close to the Eddington but only at a few per-cent, is at variance with models where the variabilityis explained as due to disk instabilities that only oc-cur at high luminosity (e.g., Vilhu & Nevalainen 1998;Belloni et al. 2000; Nayakshin et al. 2000; Janiuk et al.2002; Done et al. 2004; Neilsen et al. 2011, and refer-ences therein). However, in this scenario IGR J17091-3624 would follow radio/X-ray correlation of some ormost BHCs (depending if the distance is closer to ∼ ∼
17 kpc, respectively; Rodriguez et al. 2011b). A scenario in which IGR J17091-3624 is emitting atclose to the Eddington limit puts constraints on its massand distance. Although this scenario would imply that,independently of the distance, IGR J17091-3624 doesnot follow the standard Radio/X-ray luminosity planefor BHCs (Rodriguez et al. 2011b), this would be similarto GRS 1915+105, as it does not follow the radio/X-rayflux plane either (see, e.g., figure 4 in Rodriguez et al.2011b, and references therein).We therefore made a rough estimate of the sourceflux in each observation by calculating the average-per-observation energy spectrum. In the upper panel ofFigure 1 we plot the 2-50 keV unabsorbed flux. InIGR J17091-3624 we observe the α , β , ν , ρ and µ classeswhen the flux is between ∼ ∼ − erg cm − s − (the diskbb temperature kT vary between 1 and 2 keVwhile the power law index between 2 and 3). As-suming that IGR J17091-3624 is emitting at Eddingtonrates, we derived the mass of the compact object as afunction of distance (Figure 5) for a 2-50 keV flux of4 · − erg cm − s − . Given that the correction fac-tor linking the 2-50 keV and the bolometric flux is notexactly known, we plot curves for factors of 1, 2 and 3.Figure 5 implies that if IGR J17091-3624 emits at Ed-dington, then either it harbors the lowest mass black holeknown today ( < M ⊙ for distances lower than 17 kpc),or, it is very distant. Such a large distance, together withits b ≃ ◦ Galactic latitude, would imply a significant,but not necessarily implausible, altitude above the disk(e.g., b ≃ -2.8 ◦ and d > ∼ × cm − , see Rodriguez et al. 2011b), imply-ing that optical studies when the binary system is inquiescence (e.g., Casares et al. 2004) might be challeng-ing. However, accurate parallax distance estimates likethose reported for the black hole X-ray binary V404 Cyg(Miller-Jones et al. 2009) could still be possible. Acknowledgments:
We thank Dave Russell andJamesMiller-Jones for insightful discussions. M.L. ac-knowledges support from an NWO Rubicon fellowship.TB has received funding from the European Commu-nitys Seventh Framework Programme (FP7/2007-2013)under grant agreement number ITN 215212 Black HoleUniverse, and TMD from the Spanish MEC undertheConsolider-Ingenio 2010 Programme grant CSD2006-00070: First Science with the GTC.