sign in sign up
<<
Home , Super star cluster

arXiv:0908.2773v3 [astro-ph.SR] 11 Jan 2010 ta.20) osqety n ol xeta expect BATSE would of position one the Consequently, between correlation Lazzati strong 2005). 2005; al. al. et et (Hurley BATSE by detected 862 nJn17)t 10 to 1979) Jan in 1806–20 etdb AS vris95ya ieie repre- lifetime, 40–60% year 9.5 senting its over BATSE by tected oiyo h nes aefo G 862 in 1806–20 lumi- 2 SGR peak was from the 2004 flare but Dec intense the galax- distances, in of Mpc detected nosity at be not lying would ies 1979) Mar in 66 rmaglx 30–50 a from initial Gamma- the ∼ detected Compton have the would Observatory, aboard Ray BATSE, kpc. 15 e 0y,prMlyWyglx,attlo 200– of total a galaxy, Way Milky 300 per yr, 30 per Hre ta.20) yia ekotusof outputs peak 10 Typical range the in 2005). lie flares (GRBs) al. Bursts et for Ray (Hurley mistaken Gamma be could short – cosmological distances X-ray kpc Gamma at Anomalous Soft (AXPs) or either (SGRs) as mag- classified Repeaters extragalactic locally of – flares netars giant from spike ray lyaotddsac fSR10–0is 1806–20 SGR of distance adopted ally .INTRODUCTION 1. 1841 1E also presumably #3. and or (2009) #2 al. clusters et Davies to according a aay eas ics h ait npoeio initia progenitor in variety contaminati the the from low discuss modifying also apparent We to galaxy. the recourse Way reconciles without (2008) extragalactic al. et Bibby b a onadrvso ndsac oSR1806–20 Crowther A. SGR Signifi Paul to flares: distance magnetar in giant revision by downward GRBs short of Contamination eto hsc srnm,Uiest fSeed hffil,S 7R S3 Sheffield, Sheffield, of University Astronomy, & Physics of Dept eto hsc srnm,TeOe nvriy itnKye,M Keynes, Milton University, Open The Astronomy, & Physics of Dept . pk rmSR10–0hdi originated it had 1806–20 SGR from spike s 0.1 o naotdrt foesc nes flare, intense such one of rate adopted an For ti elkonta h nta,intense initial, the that known well is It ehglgthwtedwwr eiini h itnet the to distance the in revision downward the how highlight We d ∼ 15 3 50 ( τ/ M 30yr) ⊙ o G 862 n X XUJ6702451 nWesterlund in J164710.2-455216 CXOU AXP and 1806–20 SGR for × d − 15 3 10 a 1 oneL Bibby L. Joanne , ( 47 ae ol aebe de- been have would flares τ/ d d 15 30yr) 15 p away. Mpc r s erg 45 − r s erg 1 − falsotGRBs short all of 1 41 − hr h usu- the where 1 r s erg SR0525– (SGR a ae .Furness P. James , − 1 d (SGR 15 γ = - 1 asso antr ae pncutrae,ranging ages, cluster upon based magnetars of masses l n erI perneo trcutr sdistin- is clusters visual AF the of with 4–10 Thereafter, appearance 3603) near-IR to 1). and NGC Westerlund closer (e.g. ages (e.g. Myr for Myr dominating few to supergiants a up O which of brightest for ages visually clusters, the star character- represent youngest stars the the discusses of (2009) istics massive Walborn of models upon stars. evolutionary dependent masses somewhat from albeit turn-off isochrones derived, main-sequence be and to ages dia- tools subtypes) (colour- allow (stellar Photometric spectroscopic and grams) shown 1. is Fig. appearance in near-IR whose cluster, star incorrect. distance is kpc 1806–20 15 SGR adopted to possi- widely credible the very unless a bility, remains one which than years, lower 30 frequency much per is the flares Apparently, magnetar intense 2010). of al. inconclusive et remain (Hurley claims al. such et although (Ofek M81 2006), in GRB051103 as been such have proposed, flares giant extragalactic than candidate few greater no short was to that statistics flares concluded 30 GRB magnetar (2005) giant within of al. contribution from the et only originated Lazzati that while GRBs found Mpc, short (2005) of . al. 10% star-forming et nearby Tanvir However, and bursts short rqec foesc aepr3 er e Milky per years 30 per flare such one of frequency a .SmnClark Simon J. , no AS hr Rsb nes ae from flares intense by GRBs short BATSE of on otntl,SR10–0le oad young a towards lies 1806–20 SGR Fortunately, 05i toiiae rmoeo h asv RSG massive the of one from originated it if –045 trcutrascae ihSR10–0by 1806–20 SGR with associated cluster star ,UK H, 76A UK 6AA, K7 b to 1 ∼ 17 M ⊙ o G 1900+14 SGR for ac of cance ∼ % nya Only 4%. 2 Crowther et al.

1 parsec

CXOU J164710.2−455216

SGR 1806−20

Starlink GAIA::Skycat SGR1806-1p5pc.fits Starlink GAIA::Skycat Wd1-5pc.fits SGR1806-B 580.5079 157.5225

WESTERLUND_1 16:47:02.678 -45:50:59.58 J2000 pac Aug 19, 2009 at 14:11:54 pac Aug 19, 2009 at 14:03:24

Figure 1. Location of magnetars SGR 1806–20 and AXP CXOU J164710.2-455216 within visibly obscured young massive clusters: Left (NTT/SOFI, 200×200 arcsec, equivalent to 5×5pc at d=5 kpc); right Cl 1806–20 (Gemini-N/NIRI, 34×34 arcsec, equivalent to 1.5×1.5 pc at d=8.7 kpc). SGR1806–20 3 guished by red supergiants (e.g. RSG cluster #1: Figer et al. 2006; RSG cluster #2: Davies et al. 2007).

2. DISTANCE TO SGR 1806–20

The usual 15 kpc distance towards SGR 1806– 20 relies upon kinematics (Corbel et al. 1997; Eikenberry et al. 2004; Figer et al. 2004), but if we assume that it is physically associated with a visibly obscured (AK = 3 mag), massive star clus- ter (Eikenberry et al. 2004; Figer et al. 2005), we may combine their physical properties with evolutionary models to obtain the age and (spec- troscopic parallax) distance to the cluster. The likelihood of chance alignment between the clus- ter and magnetar is low. Indeed, Muno et al. (2006) obtained a confidence of >99.97% for the association of magnetar CXOU J164710.2–455217 (AXP) with the Westerlund 1 cluster (see Fig. 1). Both clusters host rare Luminous Blue Variables and Wolf-Rayet stars in addition to OB stars, al- though Westerlund 1 is an order of magnitude 4 more massive (5–10×10 M⊙ Clark et al. 2005; Mengel & Tacconi-Garman 2007; Brandner et al. 2008), so hosts many more evolved massive stars – 24 WR stars (Crowther et al. 2006b) versus 4 WR stars in Cl 1806–20 (Eikenberry et al. 2004; Figer et al. 2005). Bibby et al. (2008) presented Gemini- S/GNIRS near-IR spectroscopy of OB super- giants and non-dusty Wolf-Rayet stars in SGR +1.8 1806–20 from which a cluster distance of 8.7−1.5 kpc was obtained using a calibration of spectral type versus . Reliable OB spectral types were obtained in two cases from comparison with the near-IR atlas of Hanson et al. (2005). Dust producing WR stars are highly Figure 2. Meynet et al. (1994) isochrone fits unreliable as absolute magnitude calibrators since (shown in black) to properties of B supergiants their near-IR appearance is dictated by the prop- in SGR 1806–20 (shown in red from Bibby et al. erties of their dust rather than underlying stellar 2008) for ages of 2.8 Myr (top), 4 Myr (middle) continua (Crowther et al. 2006b). and 5 Myr (bottom), from which distances of 15, OB stars would provide superior absolute magni- 8.7 and 6.3 kpc are obtained. For the conven- tude calibrators with respect to Wolf-Rayet stars tional 15 kpc distance to SGR 1806–20, the clus- and OB supergiants, but these would require long ter OB stars would exhibit spectral near-IR integrations for reliable stellar classifica- morphologies, while WC stars are not predicted tion even for present 8–10m telescopes. to exist at such young ages. Alternatively, the simultaneous presence of 4 Crowther et al.

SGR 1806−20 Corbel et al. (1997)

McClure−Griffiths & Gaensler (2005)

Cameron et al. (2005)

Cl 1806−20 Eikenberry et al. (2004)

Figer et al. (2004)

Bibby et al. (2008)

0 2 4 6 8 10 12 14 16 18 Distance (kpc)

Figure 3. Comparison between various magnetar and cluster distances to 1806–20, uniformly adapted to a Galactic Centre distance of 8 kpc (Reid 1993). Updated from Bibby et al. (2008).

WN and WC-type Wolf-Rayet stars in SGR 1806– M⊙ stars allowing for rotational mixing and 20 suggest an age of 4±1 Myr, from which contemporary mass-loss prescriptions indi- an independent distance measurement was ob- cate that the WC phase is not predicted tained using isochrones for OB stars (Lejeune & until after 3.5 Myr for a model initially ro- Schaerer 2001) based on the evolutionary models tating at 40% of critical (maximum) veloc- of Meynet et al. (1994) and the B supergiant tem- ity (Hirschi, priv. comm.). For the 120 perature calibration of Crowther et al. (2006a). M⊙ non-rotating case, the WC phase would For ages of 2.8, 4 and 5 Myr, both cluster dis- be anticipated to commence after 2.8 Myr tances and magnetar masses were obtained, as (Meynet & Maeder 2003). presented in Fig 2. Let us consider the properties of B supergiants and WC stars for each of the • 8.7 kpc For this distance, the cluster B- three cases in turn, type stars would be possess in the range log(L/L⊙) = 5.4 – 5.7, close to • 15 kpc For this distance, the cluster B-type the log(L/L⊙) = 5.5 average of 25 O9.5– stars studied by Bibby et al. (2008) would B3 Ia supergiants studied by Crowther et al. be exceptionally luminous, with four of the (2006a), the majority of which are morpho- five cases exceeding log(L/L⊙) = 6. In- logically normal. From Meynet & Maeder stances of such luminous B supergiants are (2003), the hydrogen-burning phase of a 1 known, such as ζ Sco (Crowther et al. non-rotating 50 M⊙ star is predicted to last 2006a). However, such stars display hyper- for 4.0 Myr, followed by short WN and WC giant spectral morphologies (B1.5 Ia+ for phases, each of ∼0.1 Myr. In this case WC ζ1 Sco), whereas the B stars in Cl 1806– stars would be expected after ∼4.1 Myr. 20 appear to be morphologically normal Models initially rotating at 300 km s−1 sug- Ia, Iab or Ib supergiants. In addition, so- gest longer main-sequence and WR phases, lar evolutionary models for 120 corresponding to the onset of the WC phase SGR1806–20 5

after 5 Myr. (Tanvir et al. 2005); (iii) < 15% from the lack of local host galaxies for several well-localized short • 6.3 kpc For this distance, the cluster B- GRBs (Nakar et al. 2006); (iv) a few percent type stars would be possess luminosities in contamination from the absence of giant flares the range log(L/L⊙) = 5.1 – 5.5, some- from galaxies in the Virgo cluster (Popov & Stern what below the average of the Ia super- 2006). giants studied by Crowther et al. (2006a), and more representative of Iab or Ib super- 4. MAGNETAR MASSES FROM ASSO- giants (Searle et al. 2008). Our GNIRS CIATED CLUSTERS near-IR spectroscopic datasets does not al- low us to discriminate between Ia and Ib su- As discussed above, associated clusters allow pergiants. Again, turning to the WR pop- progenitor masses of magnetars to be estimated, ulation, from Meynet & Maeder (2003), a since magnetars are believed to be young objects. non-rotating 40 M⊙ solar metallicity model For SGR 1806–20, Bibby et al. (2008) infer a would not be expected to remain as a WN +20 progenitor mass of 48−8 M⊙, adding to evidence star prior to core-collapse, whereas a model −1 linking some magnetars to very massive stars. initially rotating at 300 km s is predicted Identical conclusions were reached by Muno et al. ∼ to advance to the WC phase only after 5.7 (2006) for the AXP CXOU J164710.2-455216 in Myr. the Westerlund 1 (Clark et al. 2005) Overall, a spectroscopic distance of 7–10 kpc is whose age and stellar content is remarkably sim- favoured from the simultaneous presence of nor- ilar to Cl 1806–20. Other observations have also mal B supergiants, plus WN and WC stars in Cl been used to link magnetars to high mass pro- 1806–20. This cluster distance is compared to genitors. For example, Gaensler et al. (2005) previous work in Fig. 3, and is consistent with have interpreted a hydrogen shell centred upon both the Cameron et al. (2005) magnetar dis- the AXP 1E 1048.1–5937 as a wind blown bubble tance and McClure-Griffiths & Gaensler (2005) from its 30–40 M⊙ progenitor. reanalysis thereof. However, lower progenitor mass magnetars are also known. SGR 1900+14 possesses a pair of M5 3. CONTRIBUTION OF EXTRA- supergiants (Vrba et al. 1996; 2000). Davies et GALACTIC GIANT FLARES TO al. (2009) have recently presented deep K-band BATSE SHORT BURSTS imaging of these red supergiants, in which faint members of the associated cluster Cl 1900+14 Overall, we find a reduced cluster distance of can be identified. Davies et al. (2009) have ob- d15 =0.58 ± 0.1, and since the contribution of gi- tained a kinematic distance of 12.5 kpc to these ant flares to BATSE short GRB statistics scales M supergiants, and an evolutionary age of 14±1 3 +5 with d15 we find that giant flares contribute 8−4% Myr, based upon contemporary solar metallicity of all BATSE short bursts, for the canonical rate evolutionary models of Meynet & Maeder (2000) of one giant flare per 30 yr per galaxy. plus bolometric corrections of Levesque et al. Of course, the frequency of once per 30 years is (2005). If one assumes that the progenitor of merely a crude upper limit, since it originates SGR 1900+14 originated from this cluster, an ini- from the fact that solely one giant flare (from tial mass of 17±2 M⊙ can be deduced. Similar SGR 1806–20) has been detected from a Milky results would be obtained for the AXP 1E 1841– Way source since the advent of gamma-ray tele- 045 (within Kes 73) if it originated from either scopes. Nevertheless, our result largely resolves RSG cluster #2 (Davies et al. 2007) or RSG clus- the previous (low) rate problem, namely (i) ∼4% ter #3 (Clark et al. 2009). These are located at from the scarcity of BATSE sources with spectral a similar distance to 1E 1841–045, within a few properties similar to giant flares (Lazzati et al. degrees of its sight-line, although the likelyhood 2005); (ii) ∼10% from giant flares within 30 Mpc of physical assocation is naturally much weaker 6 Crowther et al.

Table 1 Progenitor masses of magnetars based on associated star clusters. SGR/AXP Cluster d Age Mass Reference kpc Myr M⊙ +1.8 +20 1806–20 Cl1806–20 8.7−1.5 4±1 48−8 Bibby et al. (2008) CXOU J164710.2-455216 Westerlund 1 5 ∼4.5 ∼55 Clark et al. (2008) 1900+14 Cl1900+14 12.5 14±1 17±2 Davies et al. (2009) 1E1841–045 RSGC#2or#36 16±4 ∼15 Clark et al. (2009) than for SGR 1806–20, CXOU J164710.2-455216 lar content of clusters 1806–20 and 1900+14 un- and SGR 1900+14. doubtedly demonstrate distinct channels leading Table 1 provides a summary of magnetar pro- to the production of magnetars. genitor masses inferred from potentially associ- ated star clusters. In the Milky Way, it is ap- REFERENCES parent that magnetars can be produced by both ∼ ∼ 15M⊙ and 50M⊙ stars. The former pre- 1. Bibby, J.L., Crowther, P.A., Furness, J.P., sumably originated in Type II-P SN following Clark, J.S. A downward revision to the dis- the red supergiant (RSG) phase, while the lat- tance of the 1806-20 cluster and associated ter are likely to have originated in a Type Ibc magnetar from Gemini Near-Infrared Spec- SN following the WR phase. From comparison troscopy, Mon. Not. R. Astron. Soc., 386, with Heger et al. (2003), certainly SGR 1900+14 L23-L27, 2008. and possibly SGR 1806–20 would be expected to 2. Brandner, W., Clark, J.S., Stolte, A. et al. produce remnants at Solar metallic- Intermediate to low-mass stellar content of ity. For metal poor populations the latter channel Westerlund 1, Astron. & Astrophys., 478, is unlikely to be available since reduced mass- 137-149, 2008. loss rates during the main sequence and post- 3. Cameron P.B., Chandra, P., Ray, A. et al. De- main sequence phases would likely lead to the tection of a radio counterpart to the 27 De- formation of a black hole for such high initial cember 2004 giant flare from SGR 1806 - 20, mass stars. No direct evidence for rem- Nat. 434, 1112-1115, 2005. nants is known in Cl 1806–20, Westerlund 1 or Cl 4. Clark J.S. Negueruela, I., Crowther, P.A., 1900+14. Presumably this is because the com- Goodwin, S.P. On the massive stellar popu- bined stellar winds from other cluster members lation of the super star cluster Westerlund 1, will blow a sufficiently large cavity that a super- Astron. & Astrophys., 434, 949-969, 2005. remnant could expand into without showing 5. Clark J.S., Muno, M.P., Negueruela, I. et al. any observable emission. Unveiling the X-ray point source population Finally, it should be bourne in mind that clus- of the Young Massive Cluster Westerlund 1, ter ages and magnetar progenitor masses rely Astron. & Astrophys., 477, 147-163, 2008. upon the reliability of evolutionary models. Val- 6. Clark J.S., Negueruela, I., Davies, B. et al. ues quoted herein are based upon a set of assump- A third red supergiant rich cluster in the tions which have subsequently been improved Scutum-Crux arm, Astron. & Astrophys. 498, upon through mass-loss rate prescriptions and al- 109-114, 2009. lowance for rotational mixing (Meynet & Maeder 7. Corbel S., Wallyn, O., Dame, T.M. et al. The 2000). Close binary evolution also further com- Distance of the SGR plicates inferred main-sequence turn-off masses. 1806-20, Astrophys. J., 478, 624-630, 1997. Therefore, one should not necessarily treat spe- 8. Crowther P.A., Lennon, D.J., Walborn, N.R., cific progenitor masses or cluster ages as robust, Physical parameters and wind properties of although the clear differences between the stel- galactic early B supergiants, Astron. & As- SGR1806–20 7

trophys., 446, 279-293, 2006a. tic SGR giant flare, Mon. Nat. Roy. Astron. 9. Crowther P.A., Hadfield, L.J., Clark, J.S. et Soc. submitted (arXiv:0907.2462) al. A census of the Wolf-Rayet content in 20. Lazzati D., Ghirlanda, G., Ghisellini, G. Westerlund 1 from near-infrared imaging and Soft gamma-ray repeater giant flares in the spectroscopy, Mon. Not. Roy. Astron. Soc., BATSE short gamma-ray burst catalogue: 372, 1407-1424, 2006b. constraints from spectroscopy, Mon. Not. 10. Davies B., Figer, D.F., Kudritzki, R.-P. et al. Roy. Asron. Soc., 362, L8-L12, 2005. A Massive Cluster of Red Supergiants at the 21. Lejeune, T., Schaerer, D., Database of Base of the Scutum-Crux Arm, Astrophys. J., Geneva tracks and ′ 671, 781-801, 2007. isochrones for (UBV)J(RI)CJHKLL M, 11. Davies B., Figer, D.F., Kudritzki, R.-P. et al. HST-WFPC2, Geneva and Washington The progenitor mass of the magnetar SGR photometric systems, Astron. & Astrophys., 1900+14. Astrophys. J., 707, 844-851, 2009. 366, 538-546, 2001. 12. Eikenberry S.S., Matthews, K., LaVine, J.L. 22. Levesque, E.M., Massey, P.M., Olsen, K.A.G. et al. Infrared Observations of the Candidate et al. he Effective Temperature Scale of LBV 1806-20 and Nearby Cluster Stars, As- Galactic Red Supergiants: Cool, but Not As trophys. J., 616, 506-518, 2004. Cool As We Thought, Astrophys. J., 628, 973- 13. Figer D.F., Najarro, F., Kudritzki, R.-P. The 985, 2005. Double-lined Spectrum of LBV 1806-20, As- 23. McClure-Griffiths, N.M., Gaensler, B.M. trophys. J, 610, L109-L112, 2004. Constraints on the Distance to SGR 1806-20 14. Figer D.F. Najarro, F., Geballe, T.R. et from H I Absorption, Astrophys. J. 630, L161- al. Massive Stars in the SGR 1806-20 Clus- L163, 2005. ter, Astrophys. J., 622, L49-L52, 2005. bib- 24. Mengel, S., Tacconi-Garman, L.E. Medium itemFiger06 Figer D.F., MacKenty, J.W., resolution 2.3 µm spectroscopy of the massive Roberto, M. et al. Discovery of an Extraor- Galactic Westerlund 1, Astron. dinarily Massive Cluster of Red Supergiants, & Astrophy. 466, 151-155, 2007. Astrophys. J., 643, 1166-1179, 2006. 25. Meynet G., Maeder A., Stellar evolution with 15. Gaensler, B.M., McClure-Griffiths, N.M., rotation. V. Changes in all the outputs of Oey, M.E. et al. A bubble co- massive star models, Astron. & Astrophys. incident with the Anomalous X-ray 361, 101-120, 2000. 1E 1048.1–5937: Are magnetars formed from 26. Meynet G., Maeder A., Stellar evolution with massive progenitors? Astrophys. J., 620, L95- rotation. X. Wolf-Rayet star populations at L99, 2005. solar metallicity, Astron. & Astrophys. 404, 16. Hanson, M.M., Kudritzki, R.-P., Kenwor- 975-990, 2003. thy, M.A. et al. A Medium Resolution Near- 27. Muno, M.P., Clark, J.S., Crowther, P.A. et Infrared Spectral Atlas of O and Early-B al. A Neutron Star with a Massive Progenitor Stars, Astrophys. J. Suppl., 161, 154-170, in Westerlund 1, Astrophys. J., 636, L41-L44, 2005. 2006. 17. Heger A., Fryer, C.L., Woosley, S.E. et al. 28. Nakar, E. , Gal-Yam, A., Piran, T., Fox, How Massive Single Stars End Their Life, As- D.B. The Distances of Short-Hard Gamma- trophys. J., 591, 288-300, 2003. Ray Bursts and the Soft Gamma-Ray Re- 18. Hurley, K., Boggs, S.E., Smith, D.M. et al. An peater Connection, Astrophys. J., 640, 849- exceptionally bright flare from SGR 1806-20 83, 2006. and the origins of short-duration γ-ray bursts, 29. Ofek, E.O. Kulkarni, S.R., Nakar, E. et al. Nat. 434, 1098-1103, 2005. The Short-Hard GRB 051103: Observations 19. Hurley, K., Rowlinson, A., Bellm, E. et al. and Implications for Its Nature, Astrophys. A new analysis of the short-duration, hard- J., 652, 507-511, 2006. spectrum GRB 051103, a possible extragalac- 30. Popov, S.B., Stern, B.E., Soft gamma re- 8 Crowther et al.

peaters outside the Local Group, Mon. Nat. Roy. Astron. Soc., 365, 885-890, 2006. 31. Reid, M.J., The distance to the center of the Galaxy, Ann. Rev. Astron. & Astrophys., 31, 345-372, 1993. 32. Searle, S.C., Prinja, R.K., Massa, D., Ryans, R. Quantitative studies of the optical and UV spectra of Galactic early B supergiants I. Fundamental parameters, Astron. Astrophys. 481, 777-797, 2008. 33. Tanvir N.R., Chapman, R., Levan, A.J., Pridey, R.S. An origin in the local Universe for some short γ-ray bursts, Nat., 438, 991- 993, 2005. 34. Vrba, F.J., Luginbuhl, C.B., Hurley, K.C. et al. The Double Infrared Source toward the Soft Gamma-Ray Repeater SGR 1900+14. Astrophys. J., 648, 225-230, 1996. 35. Vrba, F.J., Henden, A.A., Luginbuhl, C.B. et al. The Discovery of an Embedded Cluster of High-Mass Stars near SGR 1900+14, Astro- phys. J., 533, L17-L20, 2000. 36. Walborn, N.R., Age Paradigms for Massive Young Clusters, in Proc: Hot and Cool: Bridging Gaps in Massive Star Evolution, ASP Conf. Ser (eds. C. Leitherer et al.), in press, 2009.