Discovery of the black hole X-ray binary transient MAXIJ1348-630
Mayu Tominaga, Satoshi Nakahira, Megumi Shidatsu, Motoki Oeda, Ken Ebisawa, Yasuharu Sugawara, Hitoshi Negoro, Nubuyuki Kawai, Mutsumi Sugizaki, Yoshihiro Ueda, Tatehiro Mihara
aa r X i v : . [ a s t r o - ph . H E ] A p r Draft version April 8, 2020
Typeset using L A TEX twocolumn style in AASTeX62
Discovery of the massive black hole transient MAXI J1348 − Mayu Tominaga, Satoshi Nakahira, Megumi Shidatsu, Motoki Oeda, Ken Ebisawa, andYasuharu Sugawara Institute of Space and Astronautical Science (ISAS), Japan Aerospace Exploration Agency (JAXA), 3-1-1 Yoshinodai, Chuo,Sagamihara, Kanagawa, 252-5210, Japan Department of Physics, Ehime University, 2-5, Bunkyocho, Matsuyama, Ehime 790-8577, Japan Department of Physics, Tokyo Institute of Technology, 2-12-1 Ookayama, Meguro-ku, Tokyo 152-8551, Japan (Received February 29, 2018; Revised March 4, 2018; Accepted April 8, 2020)
Submitted to ApJABSTRACTWe report the first half-year monitoring of the new Galactic black hole candidate MAXI J1348–630, discovered on 2019 January 26 with the Gas Slit Camera (GSC) on-board MAXI. During themonitoring period, the source exhibited two peaks, where the first peak flux (at T =14 day fromthe discovery of T =0) was ∼ T =132 day) was ∼
10 % ofthat. The source exhibited distinct spectral transitions between the high/soft state and the low/hardstate and an apparent “q”-shape curve on the hardness–intensity diagram, both of which are well-known characteristics of the black hole binaries.Compared to other bright black hole transients, MAXIJ1348–630 is characterized by its rather low disk-temperature ( ∼ ∼
16 ( D/ M ⊙ , where D is a distance to the source, assuming theface-on geometry of the accretion disk and a non-spinning black hole. The black hole is consideredeven more massive if the disk is inclined or the black hole is spinning. The source is in the directionof the Galactic Scutum-Centaurus arm at ∼ L peak ) and the Eddington luminosity ( L Edd ), as L peak /L Edd =0.2–0.4.We suggest that MAXI J1348–630 may host the most massive black hole among the known black holebinaries in our Galaxy.
Keywords:
X-rays: individual (MAXI J1348 − INTRODUCTIONBlack hole binaries (BHBs), those consisting of astellar mass black hole and a star, show two dis-tinct X-ray spectral states; the “low/hard” state andthe “high/soft” state. The low/hard state is ob-served when mass accretion rate is relatively low.When the accretion rate becomes higher than a cer-tain threshold, the disk switches its nature from theoptically-thin/geometrically-thick state to the optically-
Corresponding author: Mayu [email protected] thick/geometry-thin state. One of the most efficientways to discover BHBs is to continuously monitor theentire sky in X-rays, because most BHBs are transientsand exhibit unpredictable X-ray outbursts. The Mon-itor of All-sky X-ray Image (MAXI; Matsuoka et al.2009), which is attached to the ISS and surveying ∼
85% of the sky every ∼
92 minutes (corresponding to theISS orbital period, where each strip of the sky is exposedfor a duration of 40–100 sec), is one of the most idealinstruments for that purpose. In fact, since 2009, MAXIdiscovered 13 new BHB transients. In this letter, wereport the discovery of MAXI J1348 −
630 on January26 2019, by MAXI/GSC and the results of continuous
Tominaga et al. monitoring of the source till 2019 Aug 3 (176 days afterthe discovery). OBSERVATIONMAXI J1348 −
630 was first discovered by MAXI at03:16 UT on 2019 Jan 26 (MJD 58509, T =0; here-after, T is defined as MJD-58509) with the X-rayflux 47 ± Swift (Kennea & Negoro 2019), INTEGRAL(Lepingwell et al. 2019) and NICER (Sanna et al.2019).Optical counterpart was detected by iTelescope.NetT31 0.51-m telescope in Siding Spring, Australia(Denisenko et al. 2019), Swift/UVOT instrument (Kennea & Negoro2019) and Las Cumbres Observatory (LCO) network 2-m and 1-m telescopes (Russell et al. 2019a). The radiocounterpart was detected by Australia Telescope Com-pact Array (ATCA)(Russell et al. 2019b) with a flatspectrum consistent with a compact jet. RESULTSWe analyzed the long term data of MAXI J1348–630using the MAXI/GSC on-demand web interface . Fig-ure 1 shows the X-ray light curve of MAXI J1348 −
630 in2–6 keV and 6–20 keV, and the hardness ratio betweenthese energy bands, as well as the light curve in 15–50 keV obtained with Swift/BAT . Because the CrewDragon Spacecraft, which was launched by SpaceX, waslocated in the line of sight during T =36–41, we do notuse the data during this period.After the discovery on 2019 January 26 (MJD 58509, T =0), the hard band (6–20 keV) flux rapidly increasedand reached the peak on T =8. The source spectrashowed a distinct hard-to-soft transition between T =8and T =21. The source was initially in the hard state,where the hardness ratio was almost constant at around0.5 until the hard-band peak ( T =8). After that, thesource flux in the high energy band rapidly declined un-til T ∼
21, then slowly decreased throughout the follow-ing 70 days. On the other hand, the source flux in 2–6keV further increased after the high energy band peak( T =8), then reached its peak on Feb 9 ( T =14) at the flux ∼ × − erg cm − s − (2–20keV). The whole energyband flux observed with MAXI/GSC declined steadilyduring T =21–91 with holding the almost constant hard-ness ratio at around 0.05. Approaching the end of thefirst outburst ( T =91–95), the source went back to thehard state with the hardness ratio ∼ T =95, http://maxi.riken.jp/mxondem (last accessed on 2020-02-12) https://swift.gsfc.nasa.gov/results/transients/ the source gradually faded, and finally reached underthe detection limit of MAXI/GSC at T =104.On 2019 May 31 (MJD 58634, T =126), the sourcebrightened again and reached the peak flux ∼ × − erg cm − s − (2–20 keV) on T =132, which is about 10 %of the first peak, followed by steady decline of the flux.The hardness ratio during the re-brightening phase wasalmost constant at 0.5, thus the source was obviously inthe hard state throughout the second outburst. After T =175, the source flux was below the detection limitagain.The hardness-intensity diagram (HID) in Figure 2shows a clear “q”-curve (e.g. Nakahira et al. 2019). Ac-cording to time-history of the hardness ratio, we classi-fied the entire observation period into the three spectralstates; T =0–8 and 107–191 as the low/hard state; T =8–21 and 91–106 as the transition state; T =21–91 as thehigh/soft state.Both the distinct spectral transitions and the “q”-shaped HID are common properties of BHBs (e.g.Homan & Belloni 2005), so we applied a standard BHBspectral model. We used xspec version 12.10.0c forspectral model fitting. Daily MAXI/GSC data donot have enough photon statistics for model fitting,so we accumulated several days of data for each spectralfitting. For the high/soft state, we adopt the opti-cally thick Multi-Color Disk-blackbody model diskbb (Mitsuda et al. 1984) and inverse-Compton scatteringmodel simpl (Steiner et al. 2009). We applied theTuebingen-Boulder interstellar medium model tbabs and the solar abundance table given by Wilms et al.(2000). We fixed the neutral hydrogen column den-sity at N H = 6 × cm − (Sanna et al. 2019). Themodel is described as tbabs ∗ simpl ∗ diskbb wherethe free parameters are scattering fraction ( F scat ) of simpl , innermost temperature ( T in ) and normalization( N d isk ) of diskbb . The photon index( Γ ) is assumedto be constant at the canonical value Γ =2.5 since thehard-tail in the high/soft state is so weak and the in-dex is hardly constrained (e.g. McClintock & Remillard2009). For the low/hard state, the model is described as tbabs ∗ powerlaw where the free parameters are pho-ton index ( Γ ) and normalization of powerlaw . Duringthe transition states, we applied the same model as thehigh/soft state, but Γ was treated as a free parame-ter. After the spectral fittings, we estimate the realisticinnermost disk radius R in as follows: Definition of the diskbb normalization N disk is such that N d isk = (cid:18) r in D/
10 kpc (cid:19) cos i (1) iscovery of MAXI J1348 − Figure 1.
The MAXI/GSC 2–6 keV and 6–20 keV light curves, hardness ratio between 6–20 keV and 2–6 keV band, and the15–50 keV
Swift /BAT light curve, from top to bottom. Grey points indicate individual scans and colored points are adaptivelybinned data. The difference in color represents the initial outburst (red) and the re-brightening (blue). Black dashed lines showthe following key dates; T =8 (hard band peak), T =14 (soft band peak), T =21 (end of the hard-soft transition), T =91 (startof soft-hard transition), T =95 (end of the soft-hard transition), T =104 (disappearance), T =126 (reappearance), T =132(peak after reappearance), and T =175 (second disappearance). Figure 2.
Hardness Intensity Diagram (HID) between thehardness ratio (6–20 keV/2–6 keV) vs. 2–20 keV intensity.The grey points and the color points connected by line areproduced in the same manner as in Figure 1. where D is the source distance, i is the disk inclination,and r in is an apparent disk radius. We estimate therealistic inner radius R in as R in = ξκ r in (2)where ξ =0.41 is a correction factor for the inner bound-ary condition (Kubota et al. 1998), κ =1.7 is the colorhardening factor (Shimura & Takahara 1995). Figure 3shows a time-history of the spectral fitting parameters,where R in is calculated assuming D =5 kpc and i =0.Here, we point out that the maximum disk tempera-ture is as low as T in ≈ T =14). This is remarkably lowercompared to other luminous black hole transients, wherethe maximum disk temperature almost always exceeds ∼ R T . The large in-nermost radius suggests that MAXI J1348–630 harborsa relatively massive black hole compared to other blackhole binaries (see the following discussion). Tominaga et al.
Figure 3.
Time evolution of the spectral parameters derived from model fits to the 2–20 keV MAXI/GSC spectra. R in iscalculated assuming D =5 kpc and i =0 (face-on geometry). The background colors represent each of the spectral state; low/hardstate (red; T =0–8 and 107–191), transition state (blue; T =8–21 and 91–106) and high/soft state (green; T =21–91). Accordingto the spectral states, we used different spectral models (see texts for details).4. DISCUSSIONWe discuss nature of the new X-ray transient MAXIJ1348–630 from the data analysis results above. Thesource has indicated clear spectral transitions (Figure1) and a q-shaped track on the HID (Figure 2), both ofwhich are well-known characteristics of BHBs. Further-more, we successfully fitted energy spectra in all thethree spectral states assuming typical models for dif-ferent spectral states of BHBs (Figure 3). These factsstrongly suggest that MAXI J1348–630 is a new blackhole binary.In particular, innermost radius of the accretion disk( R in ) is nearly constant during the high/soft state whilethe 2–20 keV flux and the disk temperature were sig-nificantly variable (Figure 3). This is a remarkableproperty in the high/soft state BHBs, such that theconstant radius is believed to be the Innermost Sta-ble Circular Orbit around the black hole (ISCO; e.g., Tanaka 1989; Ebisawa et al. 1993; Steiner et al. 2010),which is determined by only black hole mass and spin.In the case of non-rotating black holes, ISCO is equalto three times the Schwarzschild radius. Thus, fromthe average innermost radius of the disk, R in ≈ ±
17 ( D/ i ) − / km, the black hole mass is esti-mated as M BH = c R in G ≈ ± (cid:18) D (cid:19) (cos i ) − M ⊙ . (3)This distance-mass relationship is indicated in Figure 4for i = 0 ◦ and 60 ◦ .In addition, BHBs often show similar luminosity de-pendence of the spectral states. For instance, in thecurrent case of MAXI J1348–630, the flux at the soft-to-hard transition ( T =91) is ∼
10 % of the peak fluxin the high/soft state ( T =14); this value is consistentwith those of other BHBs discovered by MAXI, such asMAXI J1820+070 ( ∼
12 %; Shidatsu et al. 2019) and iscovery of MAXI J1348 − Figure 4.
Observational constrains on the distance-massdiagram. Two solid colored lines represent Equation 3 when i = 0 ◦ (blue) and 60 ◦ (cyan). The dashed lines with thesame colors denote the error region due to systematic error.The pink curves give the empirical relation that the peakluminosity in the high/soft state is 0.2–0.4 L Edd . The pink-hatched region represents the plausible mass and distancerange. The vertical lines on d =4–8 kpc indicate distancerange to the Scutum-Centaurus arm in our line of sight. MAXI J1910-057 ( ∼ L Edd ; Maccarone2003). Namely, the high/soft state peak luminos-ity is presumably between 0.1 to 0.4 L Edd . In fact,McClintock & Remillard (2009) found that the peak lu-minosity in the high/soft state corresponds to 0.2–0.4 L Edd for the established black hole binaries with knownmasses and distances, GRS1915+105, GRO J1655–40and XTE J1550–564. Assuming the same peak lumi-nosity for MAXI J1348–630 (0.2–0.4 L Edd ), we may put another constrain to the mass-distance relation. Thepink curves in Figure 4 give the distance-mass relation-ships for L peak /L Edd = 0.2 and 0.4, where we took thepeak flux on T =14 and L Edd = 1 . × ( M/M ⊙ ) ergs − .Next, we try to estimate distance to the source. Thegalactic coordinate of the source is ( l, b ) = (309 . , − . −
630 locates in the Scutum-Centaurus arm, the distance is estimated as ∼ N H = 6 × cm determined by NICERobservations (Sanna et al. 2019) does not contradictwith these distances, assuming . − and littlecircum-stellar absorption.From these arguments, we may constrain the likelyblack hole mass and distance range as the pink-hatchedarea in Figure 4, where we assumed the face-on diskgeometry. Note that the mass is greater than ∼ M ⊙ at ∼ ∼ M ⊙ ; Reid et al. 2014). Ifthe black hole is spinning (=ISCO becomes smaller forthe same mass) or the disk is inclined, the black holemass will become still larger. Existence of such massiveblack holes has been confirmed via gravitational wavedetection due to black hole merger (e.g., Abbott et al.2019). However, they have never been observed as X-ray binaries. Follow-up optical/infrared spectroscopicobservations are strongly encouraged to constrain theblack hole mass of MAXI J1348–630 more precisely fromdynamical motion measurement.This research has made use of MAXI data provided byRIKEN, JAXA and the MAXI team, and software pro-vided by the High Energy Astrophysics Science ArchiveResearch Center (HEASARC), which is a service of theAstrophysics Science Division at NASA/GSFC.REFERENCES Abbott, B. P., Abbott, R., Abbott, T. D., et al. 2019, Phys.Rev. X, 9, 031040.https://link.aps.org/doi/10.1103/PhysRevX.9.031040Denisenko, D., Denisenko, I., Evtushenko, M., et al. 2019,The Astronomer’s Telegram, 12430, 1Ebisawa, K., Makino, F., Mitsuda, K., et al. 1993, ApJ,403, 684Homan, J., & Belloni, T. 2005, The evolution of black holestates, Tech. rep.Kennea, J. A., & Negoro, H. 2019, The Astronomer’sTelegram, 12434, 1 Kubota, A., Tanaka, Y., Makishima, K., et al. 1998,Publications of the Astronomical Society of Japan, 50,667Lepingwell, A. V., Fiocchi, M., Bird, A. J., et al. 2019, TheAstronomer’s Telegram, 12441, 1Maccarone, T. J. 2003, Astronomy and Astrophysics, 409,697Matsuoka, M., Kawasaki, K., Ueno, S., et al. 2009,Publications of the Astronomical Society of Japan, 61,999