sign in sign up
<<
Home , Shell star

arXiv:1204.1066v1 [astro-ph.GA] 4 Apr 2012 fgalaxies. of 05 aqaie l 08 n nysihl younger slightly only H the and of age dynamical -formation 2008) estimated al. younger the al. et than et Cannon the Pasquali 2000; Walter of 2005; & broadband ages young Stewart from derived (e.g., as The those photometry with are significant consistent periphery are Myr. most events shell 10 and the the ago as along Myr with 100 that bursts nearly imaged began isolated shows formation star of wave- 2009) episodes resolved Re- radio al. 1999). of et Brinks (Weisz and & analysis Walter 2009), 1998; 2005), cent al. al. et et al. Weisz (Walter Walter et lengths 2008; & (Cannon (Stewart al. outskirts 1000 et UV current mid-infrared (Pasquali the at This optical of intensively near 2000), studied region is been . prominent has and the most star-formation of the which recent of in one and 1999), located Brinks & Walter is 1998; al. et (Walter ih wr reua aaylctd40Mcdistant Mpc gas- 4.0 2009), 1 located al. 2002, galaxy et al. Croxall et irregular (Karachentsev 1991; dwarf al. rich, et Masegosa 1996; asmn a (assuming 0.3 pedfnto PF ihasraebihns peak brightness surface H a the of with center the near (PSF) function spread msinfo h uegatselosre ihthe with the observed beyond shell tended supergiant the ROSAT from 1999). emission Brinks & Walter 1998; al. et (Walter ais atre l 19)cmueapam tempera- plasma a compute of (1998) ture al. et Walter ratios, pc lgtCne,V6,Hnsil,A,USA AL, Huntsville, VP62, Center, Flight Space USA AL, Tuscaloosa, h aayhssnmru H numerous hosts galaxy The C27 salwmtliiy(ilr&Hodge & (Miller metallicity low a is 2574 IC rpittpstuigL using Letters typeset Journal Preprint Astrophysical to Submitted atre l 19)as eotaayi fsf X-ray soft of analysis report also (1998) al. et Walter − 2 1 . e ag,ada lcrndniyo .3cm 0.03 of density electron an and range, keV 2.4 nvriisSaeRsac soito,NS Marshall NASA Association, Research Space Universities Alabama, of University Astronomy, & Physics of Department ∼ rmhtgsascae ihtercn trfrainatvt in activity star-formation recent the headings: emission Subject with X-ray diffuse associated thermal-timesc residual gas undergoing the hot of star from properties B The early overflow. an lobe from accreting hole black omto ciiysroniga1000 a surrounding activity formation on erysa lse,a es y l,baktn h ieya likely th the bracketing at old, Myr history fainte 5 formation and least and beginning at star 5 center formation cluster, The its star star nearby of from young hole. burst offset a the but indicates beyond hole and the throughout within source point-like Observatory .4kV uioiyo 1.6 of a keV, 0.54 SC hyfudta h msinwsex- was emission the that found They PSPC. h 8 ru ebrdafglx C27 ot uegatshe supergiant a hosts 2574 IC galaxy dwarf member group M81 The ∼ 5Mr h orei hslkl rgthg-asXrybnr eith — binary X-ray high-mass bright a likely thus is source The Myr. 25 1. ∼ mgn bevtosrva luminous, a reveal observations imaging A INTRODUCTION T 28 E i tl mltajv 5/2/11 v. emulateapj style X oe ae nseta hardness spectral on Based hole. -A MSINFO H UEGATSELI C2574 IC IN SHELL SUPERGIANT THE FROM EMISSION X-RAY ′′ Z/Z aais niiul(C27)—glxe:sabrt—Xry:binar X-rays: — starburst : — 2574) (IC individual galaxies: galaxies WMisrmna point- instrumental FWHM ′′ × ⊙ 1 c nteM1group M81 the in pc) =19 0 csprin shell supergiant pc 500 0 = i . iooYukita Mihoko × 5adaspherical a and 15 10 oe n shells and holes umte oAtohsclJunlLetters Journal Astrophysical to Submitted 38 i r s erg hl,1 Myr 14 shell, × − 0 chl nteabetH ambient the in hole pc 500 Hubble 1 nthe in 1 ABSTRACT − n oga .Swartz A. Douglas and ∼ 3 5Mraoadcretywaeigadteei a is there and weakening currently and ago Myr 25 L o ne .)aditisclmnst 6.5 luminosity (pho- intrinsic spectrum and law 1.6) power index ton hard mildly-absorbed a with rgtpitsuc oae ihnteH the within from arises located emission source the of point bulk bright the supergiant a that the show We of shell. observations imaging trophotometric mtigvlm f70p diameter). pc 700 of volume emitting oiy2.7 0.3 nosity keV, 0.5 (temperature plasma thermal ( emission 0.3 the in u n eodteH the beyond and out neryBsa opno ( companion with (HMXB) star binary B X-ray early high-mass a an is source point ad eea on trcutr nieado h su- the on and inside ( clusters shell star pergiant young several wards activity. formation star recent bevto saaye eas h H the because analyzed is observation 0k uainadaesniiet nai point-like on-axis to sensitive as are faint and as 30 sources duration approximately June of 2008 ks were on observations 10 again Both and 9541). 792) (ObsID (ObsID 7 January 2000 vatory h asv trcutri h H the in in supernovae cluster and star stars massive massive the by produced than energy more ical no for accounts emission pia lcigFles h bevto a aein made was was observation their CCDs The on ACIS material Filters. of the the buildup Blocking in of the Optical by that enough response compromised note early yet soft not also taken the We was that 2000 observation. mission in ar- latter observation ACIS-I the front-illuminated the for the used than from expected bubble, ray is gas emission below X-ray hot area the a of effective most CCD higher where S3 has keV, 1 observation back-illuminated this the for and used aimpoint the to close X ee eaayehge resolution higher analyze we Here, ete eemn h as g,adetnto to- and age, mass, the determine then We C27 a bevdwt the with observed was 2574 IC ∼ 6 . dacdCDIaigSetoee AI)on (ACIS) Spectrometer Imaging CCD Advanced 5 × × − § 10 10 X-ray residual remains There range. keV 8.0 )wihcnb hrceie sadiffuse a as characterized be can which 4) 38 h region. the 37 r ossetwt hs expected those with consistent are i 2. eo h -a orea between at source X-ray the of ge § l astase hog Roche through transfer mass ale r s erg a distribution. gas r s erg )adueti oageta h bright the that argue to this use and 5) oaino h on source point the of location e 2 -A MG ANALYSIS IMAGE X-RAY lo urn n eetstar- recent and current of ll ∼ iueeiso extending emission diffuse r − − 1.5 1 i 1 oeadi ieytersl fthe of result the likely is and hole .Teeiso xed through- extends emission The ). nte0.3 the in × 10 37 ranurnsa or star neutron a er § r s erg ) h eiulX-ray residual The 6). − i Chandra Chandra hole. . e band, keV 8.0 e X-rays: — ies − ∼ 1 %o h mechan- the of 1% ee h earlier the Here, . i oewsimaged was hole Chandra − × X-ray . e lumi- keV 8.0 -a Obser- X-ray 10 i oe( hole 38 r s erg spec- § − 3) 1 2

Fig. 1.— Left panel: the diffuse X-ray emission in a 100′′×100′′ region of the supergaint shell of IC 2574. The ellipse denotes the Hi hole as defined by Walter et al. (1998) and the small circle the position of CXOU J102843.0+682816. The other symbols denote regions identified in Figure 2. The contour levels are at 1 and 2×10−9 photon cm−2 s−1 pixel−2 (1 pixel=0′′. 492). Right panel: deprojected radial profile of X-ray events (unsmoothed; in the 0.3−2.0 keV energy range). The deprojection assumed that the elliptical Hi hole is circular. the full-frame timed exposure mode using the standard rounding the source for spectral analysis. This encircles 3.2 s frame time and the VFAINT telemetry format. roughly 99% of the 0.3−8.0 keV flux from the source. The data were reprocessed beginning with the Level A background spectrum was extracted from the sur- 1 event list. A time-dependent gain correction was rounding region. An absorbed power law model, in- applied and pixel randomization was removed using cluding a multiplicative pileup correction, was fit to the the Chandra X-ray Center’s CIAO ver. 4.3 tool background-subtracted spectrum using the XSPEC ver. acis process events. The standard ASCA event 11.3.2ag analysis tool. grades were selected and hot pixels, columns, and cosmic- The best-fitting model column density is ray afterglows were removed to create a Level 2 event list (5.4±5.8)×1020 cm−2 which is roughly the Galac- for analysis. An ∼800 s period of high background near tic line-of-sight value of 2.2×1020 cm−2, the best-fitting the end of the observation was omitted leaving a usable power law index is 1.6±0.6 (χ2 = 20.1 for 31 dof), and exposure time of ∼9 ks. the intrinsic X-ray luminosity is (6.5±1.5)×1038 erg s−1 after correcting for ∼10% pileup. A thermal plasma 3. CXOU J102843.0+682816 model also provides an acceptable fit to the point source There is a bright X-ray source located inside the su- spectrum but the resulting temperature, kTe ∼ 5 to pergiant shell. Walter et al. (1998) judged it to be ex- 18 keV, and X-ray luminosity are much too high to be tended based on a ROSAT PSPC observation. How- representative of a hot gas phase associated with star ever, it is clear in the Chandra image that most of formation. the X-ray emission is contributed from a single point- 4. like source. An elliptical Gaussian model (plus a con- DIFFUSE X-RAY EMISSION stant term) was fit to the spatial distribution of X- There remains an excess of X-ray events within ray events in the 0.3−8.0 keV range to determine and around the Hi hole (66′′×32′′) after removing the source centroid, Gaussian width, and approximate CXOU J102843.0+682816 from the image. This soft X- source asymmetry. The best-fitting source position is ray diffuse emission has low surface brightness, which 10h28m43.08s, +68◦28′16.37′′. The source is symmet- leads to a grainy, low S/N, image. The emission can rical within measurement errors with an rms Gaussian be seen more easily if the image is artificially smoothed. width of 0.315′′±0.004′′. The point source was removed using a 3′′ radius circle, A radial profile of the data was compared to a simu- which is large enough to avoid leaving residual emission lation of the PSF built using the Chandra Ray Tracer from the wings of the point source. Then, the excluded (ChaRT) and MARX suites of programs available from region was filled using the local background level Poisson the Chandra X-ray Center. Fitting a Gaussian to the method of the CIAO tool dmfilth. The filled image was profile of the data and of the simulated PSF (plus a then divided by a 0.5 keV monochromatic exposure map, Lorentzian for the wings and a constant for the back- then smoothed with the CIAO tool aconvolve using a ground) results in a best-fit width of the simulated PSF Gaussian smoothing function with a width of 10 pixels slightly larger than the observed source width. The (∼5′′). source is clearly point-like. The left panel of Figure 1 shows the smoothed soft Events were extracted from a 3′′ radius region sur- X-ray diffuse map of a 100′′×100′′ region around the su- 3

TABLE 1 Properties of Star-forming Regions

int Region Age Mass AV LW Net Counts LX 5 38 −1 35 −1 (Myr) (10 M⊙) (mag) (10 erg s ) (10 erg s ) C1 7 0.8 0.0 27 6.8±2.8 38±16 C2 17 1.5 0.0 29 9.3±3.3 52±19 C3 17 0.9 0.1 16 1.0±1.4 6±8 C4 7 0.2 0.1 4.7 −0.4±1.0 −2±6 pergiant shell. The Hi hole is denoted by an ellipse and in the IR band. Finally, the (dust-modified) emergent the location of CXOU J102843.0+682816 by a small cir- model spectra are convolved with instrumental response cle. There is a net of 28.1±5.3 X-ray events in the energy functions appropriate to the various instruments and range of 0.3−8.0 keV inside the Hi hole after subtracting the best-fitting solution to the observed broadband pho- background. (The background was defined as emission in tometry is determined. (The observations are corrected the S3 chip outside the optical extent of IC 2574, defined for a line-of-sight Galactic extinction of AV = 0.120 by an ellipse approximating the 25 mag s−1 blue light (Schlegel et al. 1998) using the Cardelli et al. (1989) dust isophote, after excluding any detected point sources.) We model.) This method self-consistently overcomes the note that 90% of the emission is found in the soft band age-extinction degeneracy when fitting models to UV- (0.3−2.0 keV). to-IR broadband photometry. The right panel of Figure 1 displays the radial profile Here, we use calibrated GALEX FUV and NUV obser- of events in the 0.3−2.0 keV energy range of the diffuse vations available through the GalexView data interface3 emission after deprojecting the image. The deprojection and Spitzer 3.6, 4.5, and 24 µm observations processed was made assuming that the 66′′×32′′ elliptical Hi hole is and made available as part of the Spitzer Infrared Nearby actually circular. The radial profile is centered on the Hi Galaxies Survey (Kennicutt et al. 2003). Images of the hole center. It clearly shows the excess emission within supergiant shell at several passbands are displayed in Fig- the Hi hole and perhaps extending slightly beyond the ure 2. A continuum-subtracted Hα image observed with hole. Note that there is no emission enhancement within the Kitt Peak 2.1 m telescope using the T2KB CCD Im- the current star-forming regions in the shell surround- ager in 2008 March (1.3′′ seeing) is also shown. ing the hole. The broad energy distribution of the X- Our cluster selection is to be both FUV and 24 µm ray events within the Hi hole was compared to a grid of bright sources. Three clusters around the shell were models of an absorbed thermal plasma. Specifically, the identified from both FUV and 24 µm images. The UV ratios of observed events in the 0.3−1.0 to 0.3−8.0 and bright cluster located in the Hi hole has been well studied in the 0.3−1.0 to 0.3−2.0 bands most closely match a (Stewart & Walter 2000; Pasquali et al. 2008) therefore model of temperature 0.5±0.1 keV and intervening ab- it is also investigated further, although it lacks 24 µm sorption column of (4±2)×1020 cm−2 where the errors emission. Overall, the total of four isolated star clusters denote the model grid spacing. This model results in a were analyzed. These are marked in the figure along with luminosity of (2.7±0.5)×1037 erg s−1 in the 0.3−8.0 keV the position of CXOU J102843.0+682816 and the ellip- range where the error is based on the net source counts. tical boundary of the Hi hole. The best-fit age, mass, and extinction derived for these clusters are listed in 5. MASS, AGE, AND EXTINCTION IN SUPERGIANT Table 1. Also listed are the current mechanical lumi- SHELL STAR-FORMING REGIONS nosities from stellar winds and supernovae, LW, corre- As mentioned in the introduction, several previous sponding to the best-fitting Starburst99 model, X-ray studies have provided estimates of the mass and/or age of background-subtracted net counts in circular apertures star-forming regions associated with the supergiant shell. of size equal to the 99% FUV encircled energy region We have performed our own multi-wavelength analysis in the 0.5−2.0 keV energy range, and the corresponding which produces self-consistent age, mass, and extinction X-ray luminosity, LX, in the 0.5−2.0 keV energy range estimates for individual (assumed co-eval) star clusters. assuming a 0.5 keV thermal plasma X-ray spectral shape. All clusters are young and the masses of the clusters The method is presented in detail in Yukita (2010). 4 Briefly, the analysis uses the Starburst99 stellar syn- are on the order of 10 M⊙ with the exception of C2. C2 i thesis program (Leitherer et al. 1999) to compute star is the central stellar cluster located within the H hole cluster model spectra as a parameterized function of clus- surrounded by the supergiant shell. The age of this clus- ter age, history, and metallicity from a ter has been estimated as 11 Myr from the FUV observa- combination of tracks and a grid of stel- tion (Stewart & Walter 2000) and 5 Myr from the optical lar model atmospheres weighted by an initial mass func- data (Pasquali et al. 2008). We obtain a slightly older tion (IMF). We assume a (single) instantaneous burst age. Stewart & Walter (2000) also estimate the mass of 5 history, solar metallicity, and a Salpeter IMF (Salpeter the cluster as 1.4×10 M⊙ consistent with the present re- 5 1955). Transmission of the composite stellar spectra sult of 1.5×10 M⊙. Both authors also found that the Hii through the surrounding ISM is modeled using a stan- regions in the vicinity of the cluster are younger, having dard starburst dust extinction model (Calzetti et al. ages of 1−3 Myr, which is less than the present results. 2000) that modifies the blue portion of the spectrum (We note that we obtain ages of 2−3 Myr if we assume a and we assume that the dust re-emits this radiation as a 3 blackbody (fixed at T = 30 K, see Cannon et al. 2005) http://galex.stsci.edu/GalexView/ 4

Fig. 2.— Clockwise from top left: FUV, Hα, 24 µm, and 3.6 µm images of the same supergiant shell region as depicted in Figure 1. Four star clusters identified in the 24 µm and FUV are denoted by white circles depicting 3σ aperture sizes defined by circular Gaussian fits to the FUV data. canonical dust temperature of 70 K.) This age gradient emission throughout the hole and in the surround- supports the premise of propagating star formation such ing shell. Analysis of the UV-to-IR spectral energy that the mechanical energy from the central cluster com- distributions of several representative star clusters in presses the surrounding ISM and triggers subsequent star and around the Hi hole results in age and mass es- formation at the rim. The mass of C2 is large enough timates consistent with previous studies. In particu- 5 to drive star formation in the surroundings: The derived lar, there is a massive, 1.5×10 M⊙, intermediate-age, mechanical luminosity for C2 is 2.9 × 1039 erg s−1. The 5−17 Myr, cluster near the center of the hole that has cumulative energy released is 2× 1054 erg over the life- released enough mechanical energy from massive stars time of the cluster, which is much more than needed to and supernovae over its lifetime to have created the create the hole (2 × 1053 erg; Walter et al. 1998). hole. Our analysis is consistent with previous anal- yses of the supergiant shell in IC 2574 (Walter et al. 6. SUMMARY AND DISCUSSION 1998; Walter & Brinks 1999; Stewart & Walter 2000; High-angular resolution Chandra spectrophotomet- Pasquali et al. 2008; Weisz et al. 2009) and supports ric imaging of the Hi hole and surrounding super- the classical picture of bubble formation and evolu- giant shell in IC 2574 reveal a bright point source, tion such that younger star clusters on the periph- CXOU J102843.0+682816, near the center of the Hi ery of the hole likely formed as a consequence of this hole and residual low surface brightness soft X-ray stellar feedback sweeping ambient material into dense 5 clouds that collapse under self- and subsequently mate the age of the binary system at between 5 Myr, form new stars (e.g., Larson 1974; Weaver et al. 1977; the youngest age of the nearby massive , and McCray & Kafatos 1987; Mac Low & McCray 1988). 25 Myr, the age of the peak in the star-formation rate We have shown that the ratio of X-ray to mechanical at the location of CXOU J102843.0+682816 as deter- luminosity in these young star clusters in the region is . mined by Weisz et al. (2009). The compact object must 0.2%. Observed ratios of 0.1% to 1% are typical of many then have had a minimum initial mass between about 10 star-forming regions (Yukita 2010) but are poorly con- and 40 M⊙ to have already evolved to a end- strained theoretically because of uncertainties in ambi- point. The companion is likely also of roughly this mass ent conditions, starburst histories, and three-dimensional in order to provide the required high mass rate. 2 geometries. For example, we note that the central clus- This rate can be estimated assuming LX ∼ Lbol ∼ ηmc˙ −7 −1 ter alone could be responsible for all the X-ray diffuse with an efficiency, η ∼ 0.1, givingm ˙ ∼ 10 M⊙ yr . emission within the Hi hole; the mechanical luminosity According to Podsiadlowski et al. (2003), this requires a 39 −1 from the central cluster is 2.9×10 erg s and the dif- companion star mass of at least 8−10 M⊙ and higher if fuse X-ray luminosity from the hole is 2.7×1037 erg s−1. the system is older, if the bolometric luminosity exceeds Weisz et al. (2009) estimate that the cumulative energy the X-ray luminosity, or if mass is lost from the sys- released from the whole region in the past ∼25 Myr is tem. Podsiadlowski et al. (2003) also estimate the rate ∼1055 erg. Even at a constant (current) rate, less than of formation of systems of this type scaled to a fiducial 0.1% of the total energy from stars has been lost to the supernova rate. Their rate can be converted to a total X-ray radiation. Thus, the central cluster in the Hi hole number of systems in the central star cluster as there are alone can account for the diffuse X-ray emission and the about one M > 8 M⊙ star (i.e., stars that eventually creation of the Hi hole. Perhaps we are witnessing pow- undergo supernovae) per 50 M⊙ of star formation for a erful positive feedback from this central cluster in the Salpeter IMF. Thus, the Podsiadlowski et al. (2003) rate −4 −1 form of subsequent star formation in the surrounding su- of <10 yr per 8 M⊙ star translates to 0.3 luminous 5 pergiant shell. HMXBs for the central star cluster of mass 1.5×10 M⊙. Most of the X-ray emission in the Hi hole This is a high estimate because not all the potential com- in IC 2574 comes from the single bright source pact objects have yet been created by supernova events. CXOU J102843.0+682816. The presence of such a Thus, objects like CXOU J102843.0+682816 are indeed luminous X-ray point source in even a massive su- rare even for massive star clusters. pergiant shell is uncommon. Recent supernovae can We gratefully acknowledge the referee, Leisa Townsley, be this luminous but they, too, are rare and typi- for her expert critique and especially for alerting us to the cally much softer than CXOU J102843.0+682816. The possibility of pile-up of the point source. Support for this high luminosity, hard spectral shape, and large num- research was provided in part by NASA through an As- bers of recently-formed stars in the vicinity suggest that trophysics Data Analysis Program grant NNX08AJ49G. CXOU J102843.0+682816 is an HMXB. We can esti-

REFERENCES Calzetti, D., Armus, L., Bohlin, R. C., Kinney, A. L., Koornneef, Salpeter, E. E. 1955, ApJ, 121, 161 J., & Storchi-Bergmann, T. 2000, ApJ, 533, 682 Schlegel, D. J., Finkbeiner, D. P., & Davis, M. 1998, ApJ, 500, Cannon, J. M., et al. 2005, ApJ, 630, L37 525 Cardelli, J. A., Clayton, G. C., & Mathis, J. S. 1989, ApJ, 345, Stewart, S. G., & Walter, F. 2000, AJ, 120, 1794 245 Walter, F., & Brinks, E. 1999, AJ, 118, 273 Croxall, K. V., van Zee, L., Lee, H., Skillman, E. D., Lee, J. C., Walter, F., Kerp, J., Duric, N., Brinks, E., & Klein, U. 1998, Cˆot´e, S., Kennicutt, R. C., & Miller, B. W. 2009, ApJ, 705, 723 ApJ, 502, L143+ Karachentsev, I. D., et al. 2002, A&A, 383, 125 Weaver, R., McCray, R., Castor, J., Shapiro, P., & Moore, R. Kennicutt, Jr., R. C., et al. 2003, PASP, 115, 928 1977, ApJ, 218, 377 Larson, R. B. 1974, MNRAS, 169, 229 Weisz, D. R., Skillman, E. D., Cannon, J. M., Walter, F., Brinks, Leitherer, C., et al. 1999, ApJS, 123, 3 E., Ott, J., & Dolphin, A. E. 2009, ApJ, 691, L59 Mac Low, M.-M., & McCray, R. 1988, ApJ, 324, 776 Yukita, M. 2010, PhD thesis, The University of Alabama in Masegosa, J., Moles, M., & del Olmo, A. 1991, A&A, 249, 505 Huntsville McCray, R., & Kafatos, M. 1987, ApJ, 317, 190 Miller, B. W., & Hodge, P. 1996, ApJ, 458, 467 Pasquali, A., et al. 2008, ApJ, 687, 1004 Podsiadlowski, P., Rappaport, S., & Han, Z. 2003, MNRAS, 341, 385