On the progenitors of (Long) GRBs
梅田 秀之
(東京大学 天文学専攻)
• (L-)GRB progenitor – associated with Hypernovae
 Massive Stars
• Central engine (popular models):
– BH+DISK (Collapsar) --- progenitor M ≥ ~ 25 M
– Magnetar --- progenitor M~ 15-20 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)
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-GRB030329 (typical GRB)
SN1998bw-GRB980425 (weak GRB)
SN2003lw-GRB031203 (XRF)
GRB011121 (z = 0.36)
z = 0.695
Galama et al. 2000
Bloom et al. 2001
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
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)
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
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 2003,2004
• Proposed solution: Chemically homogeneous evolution
– Yoon&Langer 2005, Woosley & Heger 2006, Yoon et al. 2006
(by Dr. Yoon in this conference)
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
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)
Rotating single star or Binary interaction ?
Eldridge et al. 2009
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
2. Time scales for envelope stripping
is shorter than single star cases.
 Less angular momentum loss
Roche Lobe Overflow
Fryer, Woosley & Hartmann 1999
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
4. Main product of close WR binaries with
compact companions is a He star –
compact object merger (Detmers et al.
He star – He star (or compact object)
Progenitors can have large angular
momentum relatively easily
He-He merger can be GRB (Fryer &
Heger 2005)
He – compact merger: likely GRB but
haven’t been studied much yet.
Common envelope
Fryer & Heger 2005
Binary interactions and progenitors for collapsars
• Common envelope (CE) evolution is
complicated and the results are often
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. 2010
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
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
• Theory
– CE Ejection
– Origin of Magnetars
– Convection, magnetic filed, anugular momentum transfer, and mixing in
the progenitors
Early Black Hole Formation by
Accretion of Gas and
Dark Matter (annihilation)
H. Umeda (Univ. of Tokyo),
N.Yoshida, K. Nomoto (IPMU),
S. Tsuruta, M. Sasaki, T. Ohkubo
• It is not known how super massive blackholes (SMBH)
~109M were formed as early as z~6 as observed.
• A popular scenario: (e.g., Li et al. 2007; Tanaka & Haiman 2008
– Bondi accretion onto a Pop III (z~30) seed BH ~100M
– Eddington accretion rates is enough?
– Or Super Eddington accretion?
• Pop III seed BH ≳100M is required but the mass function of the first stars are
not well known.
Bondi 降着
• Bondi 降着率(1)は中心天体の質量Mの2乗で増える
⇒ seed BH mass が重いほど有利 (速く成長
• BH 質量が Medd ~ 103-4Mに達した後はEddington rate で成長
Evolution of First Stars
(~1000M molecular cloud in a ~106M dark halo)
No Metal
⇒ radiation pressure to
prevent mass accretion
is lower.
First stars may become very
massive (over 100 or even
300M) ?
e.g., Omukai & Palla 2003,
Tan & McKee 2004
proto star core
Pop III BH ≳100M  really existed?
• Stellar mass and fate (without Mass-loss)
– ~8-140M: Fe Core collapse (SNe)
– ~140-280M: e+-e- Pair Instability (PISNe)
– > ~280M: Fe core collapse
• It was once considered that most PopIII stars became PISNe
– PISNe do not leave BHs
– No evidence of PISNe in the abundance patters of metalpoor stars (e.g., Umeda & Nomoto 2002)
Purpose of This Work
• BH ≳100M really existed?
– Mass of First stars and their fate
• Stellar Evolutionary calculations with mass accretion
– Realistic accretion rates from cosmological simulations
– Mass of seed BHs
• Effects of dark matter annihilation on Pop III star evolution
• Related papers:
– H. Umeda et al. : Journal of Cosmology and Astroparticle Physics, 08, 024
– T. Ohkubo et al.: ApJ accepted (2009), arXiv0902.4573
Mass Accretion Rates from Cosmological
Yoshida et al & Gao et al. rates
• Yoshida et al. 2006 rates:
– Without Feed back --- Mfinal ~1000M BH (Pop III.1)
• Maybe Mfinal ≲ 200M  with Tan&McKee like feedback
– Typical formation epoch z~10
• Too late and too many to explain Z~6 SMBHs
• Gao et al. 2007 (model R5wt) :corresponding to firstest stars
in the universe (z~50)
– Compared with Z~10 objects
• Located in a denser halo
⇒ temperature of the gas cloud is higher
⇒ larger mass accretion rates ⇒heavier stars
• Rarer objects
– May avoid over production of high-z SMBH
Stellar Radius
Results: Evolution of Accreting Pop III stars
Fe core collapse
Without the DM annihilation energy
Stellar Mass
• All models evolve to Fe-core collapse
• Final Mass: 916 M (Y), 3901 M (G), 856 M (F)
• Life time: few million years
Effects of DM annihilation onto a PopIII star
• If (self-annihilating) WIMP (weakly interacting massive particles)
exist, the annihilation energy may overcome the nuclear energy
in Pop III stars:
– E.g., Spolyar et al. (2008), Freese et al. (2008), Iocco et al. (2008), Taoso et
al. (2008), Yoon et al. (2008)
⇒ The star is sustained by the DM annihilation energy
⇒ called “Dark stars”
• If DM density is sufficiently high (or ρx σ mx-1is large), stellar
evolution is “stalled” until the DM is exhausted.
Dark stars
• 典型的WIMP mass ~100GeV, 対消滅断面積<σ v>=3x10-26
cm3/s を採ると、DM密度が充分濃い(ρx~10 11GeV/cm3など)場
合には DM annihilation energy が核燃焼によるエネルギー生
成を卓越する (e.g., Spolyar et al. 2008)
⇒ 星がDM対消滅によって支えられる
⇒ このような星を Dark stars と呼ぶ(人がいる
(ちなみに見た目は暗くない– 宇宙で最も明るいかも
• これまでの研究の多くは一定の星質量の場合:
DMが濃い場合(ρx σ mx-1が大きい)星の進化はDMが消費され尽
Previous work:
dark matter density and dark star evolution
of DM
Lifetime of
dark stars
with a constant
stellar mass
e.g., Spolyar et al. 2008, Iocco et al. 2008,
Freese et al. 2009
Dark matter density
Dark matter density
Taoso et al. 2008
This work (Umeda et al. 2009,JCAP)
• We have investigated the evolution of mass accreting dark star models
up to the onset of gravitational core-collapse,
• using realistic mass accretion rates based on cosmological simulations
(Yoshida et al 2006 & Gao et al. 2007).
Gas accretion+
DM capture
Captured DM
concentrates at the center
of the star, and annihilates.
Accreting “Dark Star”
Pop III star
DM annihilation
Parameters & Assumptions
• WIMP mass =100GeV,
annihilation cross section <σ v>=3x10-26 cm3/s、
DM density ρx=1011 GeV/cm3
• Only consider captured DM
– DM by adiabatic contraction (c.f. Spolyar et al 2009) is neglected
• DM Capture rate : according to Iocco et al. 2008
• Gas (baryon) mass accretion rates
dM/dt = 10 -2, -3, -4 M /yr (constant) &
Time dependent (from cosmological simulations)
DM capture & annihilation energy generation rate
(capture rate)
Maxwell-Boltzman distribution
(in thermal equilibrium)
Energy generation rate
Stellar Luminosity (approximately)
∝ρx σ mx-1
σ : DM-baryon elastic scattering
Mx : DM mass
Gould 1987; Iocco et al. 2008; Yoon et al.2008
Results (constant dM/dt)
Without DM
With DM heating
Fe core
Fe core
dM/dt=10-2 M /yr
> critical rate
⇒stellar envelope
expand during
⇒may disturb mass
Fe core
dM/dt=10-4 M /yr:
Fe core formation
dM/dt=10-3 M /yr:
DM annihilation
effect is very large
Time dependent dM/dT (model Fd)
This star is sustained
mostly by the DM
annihilation energy
~dark star~
Energy generation rate
at center
However, its
appearance is not
much different from
an ordinal star for
Final Mass and Stellar Luminosity
L ≈ Ledd for all models with M ≳1000M
d: with DM heating
B: dM/dt =10-2 M/yr
C: dM/dt =10-3 M/yr
(model F)
Final mass
(without DM)
>1150 M
2920 M
418 M
860 M
Final Mass
(with DM)
>850 M
>105 M
515 M
988 M
(the masses of the 10-2 models are still increasing )
X(H) initial =0.753
He mass fraction
Gravitational collapse of the Model (Bd)
Energy generation
Rate @center
Model Bd (dM/dt=1e-2)
with DM heating:
stalls during H-burning
(Left Figure)
DM density is reduced
by a factor of 3
⇒H-burning resumed
(DM annihilation)
Central density
⇒Gravitational Collapse
during He-burning stage