Suzaku X-Ray Spectroscopy of a Peculiar Hot Star in the Galactic Center Region
Yoshiaki Hyodo, Masahiro Tsujimoto, Katsuji Koyama, Shogo Nishiyama, Tetsuya Nagata, Itsuki Sakon, Hiroshi Murakami, Hironori Matsumoto
aa r X i v : . [ a s t r o - ph ] D ec Suzaku X-Ray Spectroscopy of a Peculiar Hot Starin the Galactic Center Region
Yoshiaki
Hyodo , Masahiro
Tsujimoto , ∗ Katsuji
Koyama , Shogo
Nishiyama , Tetsuya
Nagata , Itsuki
Sakon , Hiroshi
Murakami , and Hironori Matsumoto Department of Physics, Graduate School of Science, Kyoto University,Kita-shirakawa Oiwake-cho, Sakyo, Kyoto 606-8502 Department of Astronomy & Astrophysics, Pennsylvania State University525 Davey Laboratory, University Park, PA 16802, USA Department of Physics, Rikkyo University, 3-34-1, Nishi-Ikebukuro, Toshima, Tokyo 171-8501 National Astronomical Observatory of Japan, 2-21-1, Osawa, Mitaka, Tokyo 181-8588 Department of Astronomy, Graduate School of Science, Kyoto University,Kita-shirakawa Oiwake-cho, Sakyo, Kyoto 606-8502 Department of Astronomy, School of Science, the University of Tokyo, 7-3-1, Hongo, Bunkyo, Tokyo 113-0033 Institute of Space and Astronautical Science, 3-1-1, Yoshinodai, Sagamihara, Kanagawa, [email protected] (Received 2007 June 12; accepted 2007 August 31)
Abstract
We present the results of a Suzaku study of a bright point-like source in the6.7 keV intensity map of the Galactic center region. We detected an intense Fe
XXV ∼ ∼ × cm − )thin thermal plasma model with a temperature of 3.8 ± ∼ × erg s − (2.0–8.0 keV) at 8 kpc. The absorption, temperature, luminosity,and the 6.7 keV line intensity were confirmed with the archived XMM-Newton data.The source has a very red ( J − K s = 8 . ∼ ∼
31 mag. The high plasma temperature and the large X-ray luminosityare consistent with a wind-wind colliding Wolf-Rayet binary. The similarity of theSED to those of the eponymous Quintuplet cluster members suggests that the sourceis a WC-type source.
Key words:
Galaxy: center — stars: Wolf-Rayet — X-rays: individual(CXOGC J174645.3–281546, 2XMMp J174645.2–281547) — X-rays: stars — X-rays:spectra 1 . Introduction
The distribution of the 6.7 keV X-ray emission line along the Galactic plane is stronglypeaked at the Galactic center (Koyama et al. 1989). The line originates from a K α transition ofhighly ionized iron. Together with its large penetrating power delving through a ≈ H cm − extinction, the emission is an indispensable tool to trace the high-energy activity and to revealthe X-ray source demographics in a heavily attenuated environment, such as the Galactic centerregion, where the enormous density of gas and dust, the strong magnetic fields, and the existenceof a super massive black hole influence every aspect of the transmigration of energy and matter(Morris & Serabyn 1996).Both the extended emission and the ensemble of numerous discrete sources contributeto the 6.7 keV emission. The nature of the extended emission was identified as a hot plasmahaving a temperature of ∼ ∼
250 pc (Koyama et al. 1989, 2007c).For discrete sources, several different X-ray populations with thermal emission play dominantroles. Cataclysmic variables (CV), binaries of a white dwarf and predominantly a late-typedwarf star, are major contributors to the 6.7 keV line, considering its large population andthermal plasma with emission lines (Ezuka & Ishida 1999; Muno et al. 2003, 2004b). Low-mass young stellar objects (YSO) or pre–main-sequence late-type sources are other types ofcandidates. They have elevated X-ray activity compared to their main-sequence phases, andshow a hard thermal X-ray emission with the 6.7 keV line both during the occasional flare andquiescent phases (Koyama et al. 1996a; Tsuboi et al. 1998; Imanishi et al. 2001; Ozawa et al.2005). They are expected to be numerous in the Galactic center region, where ∼
10% of thestar formation of the entire Galaxy takes place (Figer et al. 2004).In addition to these late-type populations, early-type sources are also expected to con-tribute to the 6.7 keV emission. Main-sequence field O-type stars do not show hard X-rayemission with their plasma temperatures below ∼ > ∼ θ Ori C) alone accounts for ∼
34% of the total hard-band X-ray emission integrated overthe entire Chandra field of view containing > ∗ Chandra Fellow ∼ erg s − and a plasma temperature of > ∼ ∼ erg s − with a conspicuous 6.7 keV emission. Moreover, the X-ray spectraof these stars commonly show strong metallic emission lines against continuum including the6.7 keV line, which is a consequence of extreme hydrogen depletion caused by the mass lossand hydrogen burning (van der Hucht et al. 1986).Although early-type stars known in the Galactic center region are concentrated tothe three massive young star clusters — the Arches (Nagata et al. 1995; Cotera et al.1996; Serabyn et al. 1998; Figer et al. 1999b), the Quintuplet (Kobayashi et al. 1983; Okuda et al.1990; Nagata et al. 1990; Figer et al. 1999a), and the Central cluster (Becklin & Neugebauer1968; Krabbe et al. 1995; Ghez et al. 2005) —, it is quite natural to expect that a much largernumber of early-type stars remain unidentified. Portegies Zwart et al. (2001) claimed that thenumber of young massive star clusters may exceed 50 in the Galactic center region. The incon-sistency with the observed value indicates that these clusters are too obscured to be visible inthe optical and infrared bands, or that the cluster members are dissipated before the earliestmember reaches its end, or that that number is a gross overestimate. Kim et al. (1999) showedthat the strong tidal disruption in the Galactic center region makes massive star clusters be-come unbound on a very short time scale comparable to the lifetime of an O star. Isolatedearly-type stars may be distributed throughout the region, unlike the other parts of the Galaxywhere they are found in associations.Some early attempts have been made to discover unidentified early-type stars in theGalactic center region with the combination of X-ray and infrared (IR) observations. Muno et al.(2006a) made trailblazing observations of radio, IR, and X-ray to reveal the population of youngmassive stars in the Galactic center region. They detected strong Br γ and He I lines from twosources, and classified them as either Of or candidate LBV stars. Mauerhan et al. (2007) alsoconducted K -band spectroscopy of six bright hard X-ray sources with very red IR colors. Twoof them, with IR colors of J – K = 4–5 mag, show broad hydrogen and helium emission linescharacteristic of evolved O-type stars. They are classified as a Wolf-Rayet star of a spectraltype of WN6 and an O-type Ia supergiant. Likewise, Mikles et al. (2006) performed infraredspectroscopy of CXO J174536.1–285638, which is one of new Chandra sources discovered byMuno et al. (2003). They detected P Cygni profiles of He II lines of a 170 km s − wind. Thespectral features in the X-ray and IR bands are most consistent with the colliding wind binarysystem η Car. This source is of particular interest in terms of a strong Fe
XXV emission line,which is also prominent in our source, presented below. We see a high prospect of these methods3 able 1.
Observation log.
Start Date Observatory ObsID R. A. Decl. t exp ∗ θ † (J2000.0) (ks) ( ′ )2000-09-23 XMM-Newton 0112970201 17 h m s –28 ◦ ′ ′′
11 11.72000-10-27 Chandra 1036 17 h m s –28 ◦ ′ ′′
35 10.52001-07-16 Chandra 2271 17 h m s –28 ◦ ′ ′′
10 9.52001-07-16 Chandra 2274 17 h m s –28 ◦ ′ ′′
10 5.72001-07-16 Chandra 2285 17 h m s –28 ◦ ′ ′′
10 7.92003-03-12 XMM-Newton 0144220101 17 h m s –28 ◦ ′ ′′
34 10.42006-09-21 Suzaku 501040010 17 h m s –28 ◦ ′ ′′
70 7.42006-09-24 Suzaku 501040020 17 h m s –28 ◦ ′ ′′
50 7.4 ∗ Exposure time after removing periods with high background level. The exposure times for the XMM-Newton observations refer to those obtained with the EPIC-pn camera. † Angular distance of the source from the optical axis. to find similar sources en masse.Near-IR spectroscopy may confront a challenge for even redder sources. In addition toa ∼
30 mag visual extinction ubiquitous toward the Galactic center region (Catchpole et al.1990; Schultheis et al. 1999), evolved early-type stars are behind an additional local extinctionby their own mass loss. Photospheric emission is reprocessed into longer wavelength radiationin the optically thick circumstellar matter. The IR spectra turn out to be featureless by strongveiling (Figer et al. 1999a; Crowther et al. 2006). For such sources, the spectroscopy of thehard X-ray emission directly from the vicinity of the star is the only practical tool. These starscontribute to the 6.7 keV line emission, and the diagnosis of the line emission helps to identifytheir nature.Here, we present the results of a Suzaku (Mitsuda et al. 2007) study of an X-ray point-like source located at a ∼
100 pc projected distance from the Galactic center. The source isexceptionally bright in the 6.7 keV map of the region, and is redder ( J – K >
2. Observation and Data Reduction
As a part of the Suzaku mapping campaign of the Galactic center region, a field con-taining the Sgr B2 molecular cloud was observed twice on 2006 September 21–23 and 24–25with an aim point at (RA, Dec) = (17 h m s , –28 ◦ ′ ′′ ). The observation log is given intable 1. Suzaku observations produce two simultaneous data sets using the X-ray ImagingSpectrometer (XIS: Koyama et al. 2007a) and the Hard X-ray Detector (HXD: Kokubun et al.4007; Takahashi et al. 2007). We concentrate on the former in this paper.The XIS is equipped with four X-ray charge coupled devices (CCDs) at the foci of fourX–Ray Telescopes (XRT: Serlemitsos et al. 2007). Each CCD chip has a format of 1024 × ∼ ′ × ′ with a half-powerdiameter (HPD) of 1 . ′ . ′ ∼ ′′ . Three of the four chips(XIS0, 2, and 3) are front-illuminated (FI) CCDs superior in the hard band response coveringthe 0.4–12 keV energy range. The remaining one is a back-illuminated (BI) CCD superior inthe soft band covering 0.2–12 keV. The total effective area is ∼
550 cm at 8 keV.The energy resolution, initially of ∼
130 eV at 6 keV in the full width at half maximum, issubject to degradation due to cosmic-ray radiation in the orbit. The XIS employs two techniquesto calibrate and rejuvenate its spectral capability. One is standard radioactive sources. Two Fe sources are installed to illuminate two corners of each CCD to calibrate the absolute energygain. The other is the spaced-row charge injection (SCI). Electrons are injected at one of every54 rows, and are transferred through columns to sacrificially fill in traps caused by radiation.This reduces the number of trapped charges by X-ray events throughout the transfer, thusimproving the energy resolution (Bautz et al. 2004; Koyama et al. 2007a).The observations were conducted using the normal clocking mode with a frame time of8 s. The SCI technique was used. Data process version 1.2 was retrieved and events wereremoved during the South Atlantic Anomaly passages and at earth elevation angles below3 degrees. After the filtering, the combined net integration time is ∼
120 ks.Because the calibration database for SCI observations is not yet released as of thiswriting, we used data obtained during the ground calibration with a charge-transfer efficiencyof 1. Using the Mn K α line from the calibration sources (cid:16) Fe (cid:17) , we measured the systematicgain offset as the deviation from the theoretical value at 5.8951 keV (Bearden 1967; Krause &Oliver 1979). The offsets were found to be within ∼
10 eV for the FI and ∼
30 eV for the BIchips.
3. Analysis
Figure 1 shows XIS images of the study field in the (a) 0.7–2.0 and (b) 2.0–7.0 keVbands. The overlaid contours in (a) are the narrow-band intensity in 6.75–6.77 keV. All fourXIS images at the two observations are combined. Several distinctive features can be foundin the hard band. Two extended sources are Sgr B2 (Koyama et al. 1996b, Murakami et al.2000, 2001; Koyama et al. 2007b) and G0.61–0.01 (Koyama et al. 2007b). The brightest source,which is the main topic of this paper, is found close to the northern edge of the hard-bandimage. The source is point-like and an exceptionally intense 6.7 keV emitter in the image, and able 2. Best-fit spectral parameters. ∗ Parameter Unit Suzaku XMM-Newton N H (10 cm − ) 2.4 +0 . − . +0 . − . k B T (keV) 3.8 +0 . − . +0 . − . Z Ar (solar) 8.7 +6 . − . +7 . − . Z Ca (solar) 3.2 +2 . − . +2 . − . Z Fe (solar) 0.8 +0 . − . +0 . − . F X † (10 − erg s − cm − ) 1.00 +0 . − . +0 . − . L X ‡ (10 erg s − ) 2.8 2.9 χ /d.o.f. 89.4/106 69.9/74 ∗ The uncertainty indicates the 90% confidence ranges of the fit. † Energy flux in the 2.0–8.0 keV band. ‡ Absorption-corrected luminosity in the 2.0–8.0 keV band. A distance of8 kpc is assumed. also in the vicinity found in the Suzaku wide-field map.
We extracted the source and background events from the polygonal and elliptical re-gions, respectively (figure 1b). The source was superimposed upon the intense diffuse emis-sion ubiquitous in the Galactic center region (Koyama et al. 1996b). In order to maximizethe signal-to-noise ratio against the underlying emission, we simulated point spread functionsat the source position with different enclosed energy fractions using a ray-tracing simulator( xissim : Ishisaki et al. 2007). Consequently, a 60% encircled energy polygon was chosen toaccumulate source photons. The background region was selected from an ellipse adjacent toand at a similar off-axis angle with the source region.Figure 2 shows the background-subtracted spectrum. Here, we generated the telescoperesponse files using a ray-tracing simulator ( xissimarfgen : Ishisaki et al. 2007). We usedthe detector response files as of 2005 August, which best describe the spectral profiles of thecalibration sources.The spectrum is characterized by several emission lines as well as a hard continuumextending to ∼
10 keV with a strong soft-band cut-off. The lines are identified as K α emissionfrom highly ionized (He-like and H-like) ions of Ar, Ca, and Fe. The spectrum was first fittedby a bremsstrahlung model with Gaussian lines attenuated by interstellar extinction (Morrison& McCammon 1983). The strongest line is He-like Fe K α at 6 . ± .
01 keV with an energyflux of (2 . ± . × − erg s − cm − , which corresponds to an equivalent width (EW) of950 ±
100 eV.In contrast, the line feature at 6.4 keV, if present, is very weak. We added a narrowGaussian line at the energy and derived a 90% upper limit of 1.5 × − erg s − cm − , which6 .8 0.7 0.6 0.5 0.40.20.10.0-0.1 Galactic longitude (deg) G a l ac ti c l a tit ud e ( d e g ) EN (a) Galactic longitude (deg) G a l ac ti c l a tit ud e ( d e g ) BgdSource
G0.61+0.01Sgr B2 EN (b) Fig. 1.
XIS images in the (a) 0.7–2.0 and (b) 2.0–7.0 keV bands. The narrow-band intensity of6.57–6.77 keV is shown in (a) by contours. The source and background regions are respectively shown bya polygon and a dashed ellipse in (b). Both images were processed (1) to mask calibration sources, (2) tosubtract non–X-ray background signals constructed from night earth observations (Tawa et al. 2008), (3)to correct for the vignetting and the non-uniformity caused by the contamination material on the opticalblocking filter, (4) to resample adaptively to achieve a signal-to-noise ratio of > corresponds to an EW of 50 eV. The temperature derived from the continuum and those fromthe intensity ratios of He-like and H-like K α lines are consistent with ∼ −4 −3 C oun t s s − k e V −
102 5−202 χ Energy (keV)
Ar Ca Fe FeNi
Fig. 2.
Background-subtracted XIS spectra. BI spectrum is shown in red, while the merged FI spectrumis in black. The upper panel shows the data with crosses and the best-fit APEC models with solid lines,while the lower panel shows the residuals to the best-fit. Conspicuous emission lines of He-like ions (graydash lines) and H-like ions (gray solid lines) are labeled. dances of Ar, Ca, and Fe relative to solar were free parameters, while the other elements werefixed to 1 solar abundance (Anders & Grevesse 1989). A single-temperature model yielded anacceptable fit with 3.8 keV in plasma temperature ( k B T ), 1 . × − erg s − cm − in X-rayflux ( F X ) in the 2.0–8.0 keV band and 2.4 × cm − in hydrogen column extinction ( N H ).The best-fit model and values are shown in figure 2 and table 2, respectively. No systematicdeviation was found in the single temperature fit, requiring no extra components.To examine how the results are influenced by our choice of the background region, werepeated the same procedures with several different regions. All the best-fit parameters areconsistent with each other, except for the X-ray flux, which suffers < ∼
10% systematic uncer-tainty.
We retrieved the archived Chandra (Weisskopf et al. 2002) and XMM-Newton(Jansen et al. 2001) data to study the long-term behaviors of the source, and to locate itsposition more precisely using telescopes of smaller HPDs than that of Suzaku. Four Chandra ob-servations using the Advanced CCD Imaging Spectrometer (ACIS: Garmire et al. 2003) and twoXMM-Newton observations using the European Photon Imaging Camera (EPIC: Turner et al.2001; Str¨uder et al. 2001), which is comprised of two MOS and a PN, were found to cover theSuzaku source (table 1).Within the Suzaku positional uncertainty of 50 ′′ , we found only one Chandrasource (CXOGC J174645.3–281546) in Muno et al. (2006b) and one XMM-Newton source(2XMMp J174645.2–281547) in the Second XMM-Newton Serendipitous Source Pre-release8atalogue, XMM-Newton Survey Science Centre (2006). Also, hereafter we refer the Suzakusource as CXOGC J174645.3–281546. −4 −3 C oun t s s − k e V − χ Energy (keV) Fe Fig. 3.
Background-subtracted EPIC spectra. The pn spectrum is shown in red, while the merged MOSspectrum is in black. The symbols follow figure 2.
Since no spectroscopic and temporal behaviors are presented for both of the Chandraand XMM-Newton sources in the literature, we analyzed the data and present the results here.One of the two XMM-Newton observations showed a net exposure of ∼
34 ks after removingdata during high background. Spectra with high statistics were obtained from MOS and PN,for which we conducted spectral fits in a similar manner as with the Suzaku spectrum. Thebest-fit model and parameters are shown in figure 3 and table 2, respectively. Although theemission lines are less conspicuous in the XMM-Newton data, the Suzaku and XMM-Newtonresults are consistent with each other.For the remaining one XMM-Newton and four Chandra observations, the photon statis-tics were too poor for a detailed spectral analysis, due to short exposures, chip gaps, and largeoff-axis angles of the source position. We therefore applied the best-fit Suzaku model to derivetheir flux.Figure 4 shows a long-term flux variation, which spans ∼ ∼ We retrieved other databases to characterize the multi-wavelength features ofCXOGC J174645.3–281546. Because of the extreme extinction, the source is inaccessible inthe optical bands. In the near-infrared (NIR) bands, we examined the Two Micron All-SkySurvey (2MASS: Cutri et al. 2003; Skrutskie et al. 2006) and the NIR Galactic center survey9
000 2002 2004 2006 2008051015 F l ux ( − e r g s − c m − ) Observation date (year) (cid:1632)
Fig. 4.
Long-term trend of X-ray flux in the 2.0–8.0 keV using XMM-Newton (squares), Chandra (circles),and Suzaku (a diamond). Error bars on the data points are plotted at the 90% confidence level. (Nishiyama et al. 2006) using Simultaneous Infrared Imager for Unbiased Survey (SIRIUS:Nagashima et al. 1999; Nagayama et al. 2003) on the Infrared Survey Facility (IRSF) telescope.In the mid-infrared (MIR) bands, we used the Midcourse Space Experiment (MSX: Egan et al.2003) and the Galactic Legacy Infrared Mid-Plane Survey Extraordinaire program (GLIMPSE:Benjamin et al. 2003) using the Spitzer Space Telescope (Werner et al. 2004).
Galacti c longitude (deg) G a l ac ti c l a tit ud e ( d e g ) EN Fig. 5.
NIR pseudo-color image using 2MASS J - (blue), H - (green), and K s - (red) band images. Thesource positions with the uncertainty of Chandra and MSX are shown with a small and large circle,respectively. Figure 5 shows an NIR pseudo-color image, in which an isolated point-like source(2MASS J17464524–2815476 and MSX C6 G000.7036+00.1375) is found in the 2MASS andMSX images within the positional uncertainty range of the Chandra source. The source is verybright and red with 2MASS magnitudes of ( J , H , K s ) = ( > > J - and K s -band magnitudes are upper limits due to nearby source contamination. In the SIRIUS data,while the K s -band is unavailable due to saturation, the magnitudes of ( J , H ) = (15.53, 10.05)10ere derived with photometric errors of ∼ µ m band. We fittedthe data with a single-temperature blackbody and an assumed extinction proportional tothe inverse square of the wavelength. We found a best-fit blackbody temperature ( T BB ) of980 ±
20 K, with a visual extinction ( A V ) of 31 ± L bol ) of(8.3 ± × L ⊙ ( d/ . .
2 50.11101001000 I n t e n s it y ( − e r g s − c m − µ m − ) Wavelength ( µ m) SIRIUSMSX
Fig. 6.
SED constructed from SIRIUS (open squares) and MSX (open triangles) photometry. The solidcurve shows the best-fit blackbody model with T BB = 980 K, A V =31 mag and L bol = 8 . × L ⊙ . SEDsof the eponymous Quintuplet cluster members (Okuda et al. 1990; Figer et al. 1999a) are also shown withcolors for comparison.
4. Discussion
Adopting N H /A V = 1 . × cm − mag − (Predehl & Schmitt 1995), A V = 31 magis converted to N H ∼ × cm − . This is far smaller than that determined with the X-rayof N H ∼ × cm − . Since the visual extinction of 31 mag is the typical value toward theGalactic center (Catchpole et al. 1990), the excess X-ray absorption is probably due to localobscuring matter that radiates the detected IR emission. If such an obscuring matter with solarabundance spherically surrounds the X-ray source, the 6.4 keV line with an EW of ∼
200 eVshould be detected (Inoue 1985). The lack of 6.4 keV line (EW <
50 eV) requires either thatthe iron abundance is < ∼ .
25 solar or that the local matter is concentrated in front of the sourcealong the line of sight. 11 .2. Nature of the Source4.2.1. WR Binary Origin
The SED of CXOGC J174645.3–281546 in the IR bands is very similar to those of theeponymous Quintuplet cluster members (figure 6). The five stars have cool ( T ∼ ∼ K -band). The deep absorption at ∼ µ m is due to in-terstellar silicate. The lack of either emission lines or intrinsic absorption features (Okuda et al.1989; Figer et al. 1999a) allows no spectral classification, though their luminosity of ∼ L ⊙ corresponds to those of supergiant stars. These intriguing stars were recently spatially resolved(Tuthill et al. 2006). Two (GCS 3–2 and GCS 4 of Nagata et al. 1990, or Q2 and Q3 of Monetiet al. 2001) out of five showed beautiful pinwheel nebulae of dust plume, which are seen in a fewWC (carbon rich Wolf-Rayet) stars (Tuthill et al. 1999; Monnier et al. 1999). Circumstellar dustemission with temperatures of 700–1700 K is a common character of WC stars (Williams et al.1987). Interestingly, the two sources that exhibit the pinwheel morphology are the bright-est hard X-ray sources (Law & Yusef-Zadeh 2004; Wang et al. 2006) among the Quintupletmembers. Since both the pinwheel dust plumes and the hard X-ray emissions indicate fastwind-wind collision (Tuthill et al. 1999; Monnier et al. 1999; Oskinova et al. 2003), this co-incidence is not probably accidental. The summed spectrum of three X-ray emitting sources(GCS 3-2, GCS 4, and source D of Nagata et al. 1990) in the Quintuplet cluster is very hard( kT ∼ ∼ × erg s − in the 0.3–8.0 keV band; Wang et al. 2006) thanCXOGC J174645.3–281546. In the colliding wind binary scenario, the X-ray luminosity scalesas L X ∝ D − (Stevens et al. 1992), where D is the binary orbital separation. The large X-rayluminosity of CXOGC J174645.3–281546 indicates the close separation, and the moderate fluxvariation having a factor of ∼ ∼ erg s − in the 0.5–10.0 keV band with thermal spectra with a temperature of > ∼ η Carinae: Tsuboi et al. 1997, Hamaguchi et al. 2007; WR 140:Koyama et al. 1994, Pollock et al. 2005, A1 S, A1 N, and A2 in the Arches cluster; Wang et al.2006). The X-ray spectral property of CXOGC J174645.3–281546 coincides with those of thesemassive star binaries, though near- and mid-IR spectroscopy and high spatial resolution obser-vation are essential for further constraint. Taking the discussion in § ∼ M ⊙ and the age is less than ∼ H ~ 6e22 cm -2 N H ~ 2e23 cm -2
980 KWC star?massive star?
ISM
IRX-ray
50 MK
Fig. 7.
Cartoon of the source system. Winds from the primary WC star and the companion massivestar collide to produce X-rays. The obscuring material of N H ∼ × cm − around the binary system isheated to T ∼
980 K by photospheric emission to emit IR. Whereas the IR emission suffers only interstellarmedium (ISM) extinction, the X-ray emission is subject to additional intrinsic absorption. (figure 5). Muno et al. (2006a) and Mauerhan et al. (2007) also recently discovered suchisolated massive stars in the Galactic center region. One possibility is that these stars wereformed initially in a cluster, and the cluster dissipated due to the strong tidal force of theGalactic center (Portegies Zwart et al. 2002a). Future IR and X-ray observations with a higherspatial resolution and sensitivity may reveal that either this source is really a cluster member,or truly an isolated star.
The most popular class of hard X-ray sources that exhibits iron K-shell emission lines isCV. The X-ray luminosity of CXOGC J174645.3–281546 of 3 × erg s − is at the brightest endof CVs. The bolometric luminosity of ∼ L ⊙ ( d/ . is, however far larger than CVs,because most of the optical companions of CVs are late-type main-sequence stars. The spectraof CVs are characterized by hard continuum, with three iron K-shell emission lines at 6.4 keV,6.7 keV, and 6.97 keV. The equivalent widths of both the 6.4 keV and 6.7 keV lines are, however,around 100–200 eV, although in some cases that of the 6.7 keV line shows an exceptionally highvalue possibly due to resonance scattering (Terada et al. 2001). The 6.97 keV/6.7 keV flux ratioof CVs is larger than 0.1, indicating that the ionization temperature is higher than ∼ ∼ ± L X /L bol ) of 10 − is in the range oflow-mass YSOs (Imanishi et al. 2001; Feigelson et al. 2002), the absolute luminosities are toohigh for a YSO if CXOGC J174645.3–281546 is located at 8 kpc. The possibility still remainsthat CXOGC J174645.3–281546 is a foreground YSO in a dense molecular cloud. The 6.7 keV line flux from CXOGC J174645.3–281546 is 2 . × − erg s − cm − . The to-tal flux of iron K-lines in the Galactic center region is ∼ . × − erg s − cm − (Yamauchi et al.1990). Since this value includes the 6.4 keV and 6.97 keV lines, it overestimates the 6.7 keV fluxby a factor of 2–3. The detected point sources account for ∼
10% of the total flux of the 6.7 keVline (Wang et al. 2002; Muno et al. 2004a). Therefore, the 6.7 keV line of CXOGC J174645.3–281546 alone accounts for ∼
4% of that of the detected point sources and ∼ .
4% of that ofthe total diffuse flux. Such X-ray sources would comprise a substantial fraction of 6.7 keVline-emitting point sources in the Galactic center.
5. Summary
1. We detected a strong 6.7 keV line with an equivalent width of ∼ N H ∼ × cm − ) 3.8 keV thermal plasma with an iron abundance of ∼ ∼ × erg s − in the 2.0–8.0keV band, assuming a distance of8 kpc.2. We also analyzed the archived data of Chandra and XMM-Newton, and found that theX-ray flux spanning ∼ ∼ L bol ∼ . L ⊙ at 8 kpc), andhas a cool ( T BB ∼ References
Albacete Colombo, J. F., M´endez, M., & Morrell, N. I. 2003, MNRAS, 346, 704Albacete Colombo, J. F., Flaccomio, E., Micela, G., Sciortino, S., & Damiani, F. 2007, A&A, 464, 211Anders, E., & Grevesse, N. 1989, Geochim. Cosmochim. Acta, 53, 197Bautz, M. W., Kissel, S. E., Prigozhin, G. Y., LaMarr, B., Burke, B. E., & Gregory, J. A. 2004, Proc.SPIE, 5501, 111Bearden, J. A. 1967, Reviews of Modern Physics, 39, 78Becklin, E. E., & Neugebauer, G. 1968, ApJ, 151, 145Benjamin, R. A., et al. 2003, PASP, 115, 953Berghoefer, T. W., Schmitt, J. H. M. M., Danner, R., & Cassinelli, J. P. 1997, A&A, 322, 167Broos, P. S., Feigelson, E. D., Townsley, L. K., Getman, K. V., Wang, J., Garmire, G. P., Jiang, Z.,& Tsuboi, Y. 2007, ApJS, 169, 353Cappellari, M., & Copin, Y. 2003, MNRAS, 342, 345Catchpole, R. M., Whitelock, P. A., & Glass, I. S. 1990, MNRAS, 247, 479Cotera, A. S., Erickson, E. F., Colgan, S. W. J., Simpson, J. P., Allen, D. A., & Burton, M. G. 1996,ApJ, 461, 750Crowther, P. A., Hadfield, L. J., Clark, J. S., Negueruela, I., & Vacca, W. D. 2006, MNRAS, 372,1407Cutri, R. M., et al. 2003, The IRSA 2MASS All-Sky Point Source Catalog, NASA/IPAC InfraredScience Archive. http://irsa.ipac.caltech.edu/applications/Gator/,Diehl, S., & Statler, T. S. 2006, MNRAS, 368, 497Egan, M. P., et al. 2003, VizieR Online Data Catalog, 5114, 0Ezuka, H., & Ishida, M. 1999, ApJS, 120, 277Feigelson, E. D., Broos, P., Gaffney, J. A., III, Garmire, G., Hillenbrand, L. A., Pravdo, S. H.,Townsley, L., & Tsuboi, Y. 2002, ApJ, 574, 258Feigelson, E. D., et al. 2005, ApJS, 160, 379Figer, D. F., McLean, I. S., & Morris, M. 1999a, ApJ, 514, 202Figer, D. F., Kim, S. S., Morris, M., Serabyn, E., Rich, R. M., & McLean, I. S. 1999b, ApJ, 525, 750Figer, D. F., Rich, R. M., Kim, S. S., Morris, M., & Serabyn, E. 2004, ApJ, 601, 319Garmire, G. P., Bautz, M. W., Ford, P. G., Nousek, J. A., & Ricker, G. R., Jr. 2003, Proc. SPIE,4851, 28 hez, A. M., Salim, S., Hornstein, S. D., Tanner, A., Lu, J. R., Morris, M., Becklin, E. E., & Duchˆene,G. 2005, ApJ, 620, 744Hamaguchi, K., et al. 2007, ApJ, 663, 522Hellier, C., & Mukai, K. 2004, MNRAS, 352, 1037Hyodo, Y., et al. 2008, PASJ, 60, agashima, C., et al. 1999, Proceedings of Star Formation 1999, ed. T. Nakamoto, Nobeyama RadioObservatory, 397Nagata, T., Woodward, C. E., Shure, M., Pipher, J. L., & Okuda, H. 1990, ApJ, 351, 83Nagata, T., Woodward, C. E., Shure, M., & Kobayashi, N. 1995, AJ, 109, 1676Nagayama, T., et al. 2003, Proc. SPIE, 4841, 459Nishiyama, S., et al. 2006, ApJ, 638, 839Okuda, H., et al. 1989, The Center of the Galaxy, 136, 281Okuda, H., et al. 1990, ApJ, 351, 89Oskinova, L. M., Ignace, R., Hamann, W.-R., Pollock, A. M. T., & Brown, J. C. 2003, A&A, 402, 755Ozawa, H., Grosso, N., & Montmerle, T. 2005, A&A, 438, 661Pollock, A. M. T., Corcoran, M. F., Stevens, I. R., & Williams, P. M. 2005, ApJ, 629, 482Portegies Zwart, S. F., Makino, J., McMillan, S. L. W., & Hut, P. 2001, ApJ, 546, L101Portegies Zwart, S. F., Makino, J., McMillan, S. L. W., & Hut, P. 2002a, ApJ, 565, 265Portegies Zwart, S. F., Pooley, D., & Lewin, W. H. G. 2002b, ApJ, 574, 762Predehl, P., & Schmitt, J. H. M. M. 1995, A&A, 293, 889Rana, V. R., Singh, K. P., Schlegel, E. M., & Barrett, P. E. 2006, ApJ, 642, 1042Schultheis, M., et al. 1999, A&A, 349, L69Serabyn, E., Shupe, D., & Figer, D. F. 1998, Nature, 394, 448Serlemitsos, P., et al. 2007, PASJ, 59, S9Skrutskie, M. F., et al.2006, AJ, 131, 1163Smith, R. K., Brickhouse, N. S., Liedahl, D. A., & Raymond, J. C. 2001, ApJ, 556, L91Stevens, I. R., Blondin, J. M., & Pollock, A. M. T. 1992, ApJ, 386, 265Str¨uder, L., et al. 2001, A&A, 365, L18Takahashi, T., et al. 2007, PASJ, 59, S35Tawa, N., et al. 2008, PASJ, 60, illiams, P. M., van der Hucht, K. A., & The, P. S. 1987, A&A, 182, 91Yamauchi, S., Kawada, M., Koyama, K., Kunieda, H., & Tawara, Y. 1990, ApJ, 365, 532Yusef-Zadeh, F., Law, C., Wardle, M., Wang, Q. D., Fruscione, A., Lang, C. C., & Cotera, A. 2002,ApJ, 570, 665illiams, P. M., van der Hucht, K. A., & The, P. S. 1987, A&A, 182, 91Yamauchi, S., Kawada, M., Koyama, K., Kunieda, H., & Tawara, Y. 1990, ApJ, 365, 532Yusef-Zadeh, F., Law, C., Wardle, M., Wang, Q. D., Fruscione, A., Lang, C. C., & Cotera, A. 2002,ApJ, 570, 665