On the progenitors of (Long) GRBs

Transcription

1 On the progenitors of (Long) GRBs Hideyuki Umeda (Dept. of Astronomy, Univ of Tokyo)

2 Review of other people s work + I will also show our recent calculations of evolution of massive stars, which may or may not be the GRB progenitors.

3 Contents Review of Magnetar-GRB model and progenitors Collapsar-GRB model and progenitors Binary models GRB-SN observations with minitao (with T. Minezaki) Our recent progenitor calculations (Mass-loss, Metallicity; with T. Yoshida) About the progenitors of SN2007bi (very bright, PISN? Very massive Pop III stars and GRB (with N. Yoshida, K. Nomoto, T. Ohkubo, )

4 Introduction (L-)GRB progenitor associated with Hypernovae Massive Stars Central engine (popular models): BH+DISK (Collapsar) --- progenitor M ~ 25 M Magnetar --- progenitor M~ M? (Magnetar models may have advantages in explaining long activities of GRB: X-ray shallow decay and flare emissions, e.g., Metzger 2010, Thompson this conference)

5 Magnetar Model for GRB? 22 min E.g., Bucciantini et al. (2009), MNRAS: 2D simulation, Collimated relativistic jet, 35M model assumed Several suggestions (e.g., bumps in afterglow LC) that GRBs are associated with hypernovae (likely massive SNe leaving Black holes behind), but progenitor masses have been estimated for only few cases (or possibly only one, SN2003dh-GRB (typical GRB) Others: SN1998bw-GRB (weak GRB) SN2003lw-GRB (XRF) SN1998bw

6 With X-ray Flash With GRB (Tanaka et al. 2009) SN 2003dh, 2003lw are probably too massive to have NS remnants(?). Any correlation between long activities of GRB and progenitor mass? ~ need more samples

7 Magnetar Model Even if progenitor is M>25M, if BH formation is delayed protoneutron stars (NSs) may become a central engine of a GRB (?. BH formation epoch depends on EOS and rotation, (and magnetic filed) of proto-nss quite uncertain Simulations (still long way to go example of a MHD simulation (our current, preliminary work, 3D MHD simulation for core-collapse Kuroda & Umeda (2010)

8 Or typical SNe-GRB are not so massive? If a Hypernova light curve (LC) with a GRB is powered by a magnetar (Maeda et al. 2007, Woosley 2009, Kasen & Bildsten 2009), the progenitor mass may not be determined from early LC. Later (few years) LC may distinguish Pulsar and Radioactive heating But such observations are difficult for distant supernovae Still unclear if progenitors of typical GRBs are too massive to leave Neutron Stars behind. Observations of associated SNe are quite important to determine the GRB progenitor mass (and central engine model).

9 Black hole + Disk (Collapsar) model for GRB progenitor M ~ 25 M to form a BH Pre-collapse Fe core must have sufficiently large angular momentum to form an accretion disk Associated SNe so far are all Type Ic SNe Progenitors should have lost Hydrogen and most He envelope (by mass-loss) However, this mass-loss usually causes large angular momentum loss difficult to produce GRBs (Heger & Woosley 2003, min Proposed solution: Chemically homogeneous evolution Yoon&Langer 2005, Woosley & Heger 2006, Yoon et al (by Dr. Yoon in this conference)

10 Black hole + Disk model for GRB Chemically homogeneous evolution scenario Metal poor progenitors (Z Z /5) for weak mass-loss Fast initial rotation for very efficient chemical mixing These stars remain quasi-chemically homogeneous Evolves bluewards: less mass-loss, keeping fast rotation Surface Hydrogen can be depleted without mass-loss This scenario may be the only way to provide the progenitors for collapsars from single stars, however, several uncertainties in the 1D -rotating star models: Convection, Mixing, Magnetic field, Angular momentum transport Turbulence, Meridian circulation, and Mass-loss (especially for Hydrogen-depleted Wolf-Rayet stars) All these uncertainties are complexly related 1D -rotating star models need confront with several observations

11 Rotating single star or Binary interaction? 16 min Several puzzles that can not be explained by the standard (1D spherical, non-rotating) stellar evolution models. (e.g., surface abundance anomaly, ratio of blue stars to red stars) have been attempted to be explained by the rotation effects (e.g., Geneva group. However, it is not clear if all (or most of) the puzzles should be explained by the rotation effects, because binary interaction sometimes may lead similar results. E.g., anisotropic mass-loss by eta carinae Relative numbers of O stars, Red stars, Wolf-Rayet etc. Binary interaction > hydrogen envelope removed Fewer RSGs, More WR, More SNe Ib/c as observed (next page)

12 Rotating single star or Binary interaction? Eldridge et al (rotation) binary Single

13 Binary interactions and progenitors for collapsars Binary evolution is very complicated and various possibilities. But, binaries certainly exist and are very important. Roche Lobe Overflow/ Common envelope mass ejection 1. Stripping H (& He )envelopes efficiently Making SNe Ic progenitor easier than single star models. 2. Time scales for envelope stripping is shorter than single star cases. Less angular momentum loss Roche Lobe Overflow Fryer, Woosley & Hartmann 1999

14 Binary interactions and progenitors for collapsars Mass transfer from the companion / tideal interaction 3. Spin-up by gaining mass (Petrovic et al. 2005), or by tidal interaction (Detmers et al. 2010) are not significant for most cases. 4. Main product of close WR binaries with compact companions is a He star compact object merger (Detmers et al. 2010). He star He star (or compact object) merger 5. Progenitors can have large angular momentum relatively easily 6. He-He merger can be GRB (Fryer & Heger 2005) 7. He compact merger: likely GRB but haven t been studied much yet. Fryer & Heger 2005 Common envelope phase tidal interaction

15 Binary interactions and progenitors for collapsars Common envelope (CE) evolution is complicated and the results are often controversial 8. A new mechanism for the ejection of a CE (Explosive CE ejection, Podsiadlowski et al. 2010) to explain short-period blackhole low-mass binaries. Podsiadolowski et al Orbital energy release during spiral-in is too small Explosive hydrogen burning may be strong enough to remove H & He envelope progenitor of SN Ic CE ejection occurs late angular momentum loss is small GRB Low mass BH binaries are progenitors of LGRB (see also, Brown et al. 2007) Rate ~ 10-6 yr -1 (significant fraction of all LGRBs

16 Rotating single star or Binary interaction? How can we distinguish these scenarios? Metallicity distribution Binary model can occur even in super-solar metal (but more common at low metallicity, Podsiadlowski et al. 2010) (already found?, e.g., Levesque et al. 2010) Properties of associated SNe Especially the amount of He (any associated SN Ib? (single star models tend to predict larger amount of He in the ejecta) Ejecta mass and Ni56 mass (to constrain magnetar models) Finding any evidence of chemically homogeneous WR stars without mass-loss Theory CE Ejection Origin of Magnetars Convection, magnetic filed, anugular momentum transfer, and mixing in the progenitors

17 GRB-SN observations with minitao (Univ. of Tokyo) GRB-SNe observations ( & , Minezaki & Umeda) financially supported by Grant-in-Aid for Priority Research Area "Deciphering the Ancient Universe with Gamma-Ray Bursts Thank you very much! In , MAGNAM 2m telescope in Hawaii: SN2008D-XRF Unfortunately MAGNAM has been removed from Hawaii , New telescope minitao (Chile, 1m but comparable performance with MAGNUM because of the good location North > South hemisphere Up to Z~0.025 (~100Mpc) for SN2008D type ~ 10 yr -1 SN I b/c observable GRB100316D/SN2010bh z=0.059 barely possible, if working 9 min

18 Site Studies since 1999 Satellite data analysis (with other institutes) Northern Chile West (Ocean) side: larger clear fraction but no high (>4000m) mountain Compromise for IR astronomy Cerro Chajnantor (5640m) Atacama Northern Chile

19 The minitao telescope Primary mirror Φ1.0m Cassegrain focus focal ratio F/12 field of view Φ10 plate scale 16.6 /mm Vendor Nishimura Co.

20 minitao 1.0m telescope Performance of the minitao telescope Pointing accuracy σ=1.8 arcsec Hartmann constant σ=0.19 arcsec Seeing FWHM= arcsec

21

22 Access Road to the Chajnantor summit ~5.6km Univ. of Tokyo constructed in 2006

23 The minitao telescope a Pathfinder 1-m Telescope construction Started in Dec.2008 Great Local workers Completed in March 2009

24 Near Infrared Camera : ANIR a NIR/opt. imager with Hawaii-2 array 9arcmin FoV June, 2009: First Light Run (a few nights) Nov. 2009: the 2 nd commissioning Run (12 nights) First imaging of hydrogen Paa Line from the ground! Good image quality ( FWHM ~50%) opened to Chilean Astronomers from S10A through CNTAC (~30nights of scientific operation -> ~3 nights) Galactic Center NGC6926 ANIR on the minitao 1m Telescope

25 Please see the Poster #61 by K. Motohara for the detail of minitao and Near Infrared Camera : ANIR

26 Our recent progenitor calculations with T. Yoshida (Univ. of Tokyo) 5 min Investigating mass-loss effects and abundance in the ejecta for large parameter space e.g., M=13, 15, 18, 20, 25, 30, 35, 40, 50, 70, 100M Z=10-4, 0.001, 0.004, 0.01, 0.02 SN Ib/c, GRB Mass-loss rate (for our standard) OB stars Vink et al Red giant de Jager et al Wolf-Rayet Nugis & Lamers 2000 with metallicity dependence dm/dt Z 0.5 For this standard non-rotating model, GRB progenitors (SNe Ic) are M>35 and leave BHs.

27 Non-rotating standard Mass-loss

28 Is mass-loss correct?: SN2007bi case (T. Yoshida & HU, in preparation) SN2007bi (Type Ic!) was extremely bright SN and suggested that ~ M of Ni56 was produced (Gal-Yam et al. 2009; Young et al. 2010), no evidence for GRB PISN (M(CO)> ~60M ) or very massive Fe-core collapse M(CO)> ~ M (Umeda & Nomoto 2008; Moriya et al. 2010) Metallicity of the host galaxy, 12+log(O/H)=8.15±0.15 Z We calculate stellar models for Z=0.004

29 Is mass-loss correct?: SN2007bi case : Results Z=0.004 = Z /5 PISN Core-collapse M(CO) ~ 40 M is possible for M ini >~ 100M if mass-loss rate is reduced by factor 2. To realize PISN, M ini >~ 250M and mass-loss <~ 1/3 is required.

30 Is mass-loss correct?: SN2007bi case : Can these become SNe Ic? Z=0.004 = Z /5 Total He mass in the envelope Probably yes for M > ~ 140 M

31 Is mass-loss correct?: SN2007bi case : Discussion If mass-loss for very massive (metal-poor) star correct? Or any other effects, such as rotation (chemically homogeneous evolution? If true mass-loss rate for massive metal-poor stars are much smaller than the one we usually use, single star may become GRB progenitors more easily. Or SN2007bi was not a simple single star, but merger of several massive stars in a dense star cluster (?

32 Very massive Pop III GRB? 0 min We have calculated evolution of accreting first stars taking into account of realistic mass accretion rates obtained by cosmological simulations (Ohkubo et al. 2009, Umeda et al. 2009) First star is likely M > ~ 3000M and form intermediate mass BH. Such objects may become like a collapsar and GRB (Meszaros & Rees 2010, arxiv: ) With dark matter annihilation M fin ~ 6000M Formed z ~20, final mass ~ 4000M Umeda et al. 2009, JCAP

33 Summary We don t know exactly what are GRB progenitors yet. We should consider magnetar and binary models seriously, as well as popular collapsar Observations will tell you the answer. GRB-SNe observations are important (M, M(He).