Clumping in Hot-Star Winds, Potsdam, June 2007 The impact of reduced mass loss rates on the evolution of massive stars Raphael HIRSCHI (KEELE University, UK)
Plan 2 Introduction Impact of reduced mass loss rates Metallicity dependence First stellar generations Gamma-ray bursts (GRBs) & magnetic fields Conclusions
Massive Stars Massive stars: M > ~10 solar masses 3 Main sequence: LBV BSG RSG hydrogen burning WR MS After Main Sequence: Helium burning Supergiant stage (red or blue) Wolf-Rayet (WR): M > 20-25 M o WR without RSG: M > 40 M o Advanced stages: carbon, neon, oxygen, silicon burning iron core Core collapse bounce supernova explosion
Mass loss by stellar winds: Importance 4 Tracks, lifetimes, surface abundances Population of massive stars Supernova type (II, Ib,c) and nature of remnant (NS or BH) Energy release into the interstellar medium Injection of newly synthesized elements Hardness of ionizing radiation... HST
Mass loss prescriptions (a few of them) 5 - de Jager et al 1988 O-type & LBV stars (bi-stab.): - Vink et al 2000, 2001 - Kudritzki & Puls 2000, Kudritzki 2002 RSG: - Observations: van Loon 2005, Crowther 2000 WR stars (incl. clumping effect?): - Nugis & Lamers 2000 - Vink et al 2005 -...
Massive Stars: Evolution of the chemical composition Burning stages (lifetime [yr]): Hydrogen (10 6-7 ): 1 H 4 He 6 & 12 C, 16 O 14 N Helium (10 5-6 ): 4 He 12 C, 16 O & 14 N 18 O 22 Ne Carbon (10 2-3 ): 12 C 20 Ne, 24 Mg Neon (0.1-1): 20 Ne 16 O, 24 Mg Oxygen (0.1-1): 16 O 28 Si, 32 S Silicon (10-3 ): 28 Si, 32 S 56 Ni http://quasar.physik.unibas.ch/~hirschi/animation/anim.html Jun 18, 2007 Raphael Hirschi University of Keele (UK)
Impact of clumping 7 Mass loss rates may be 10x smaller Fullerton et al 2006 What are the consequences of a reduction factor of 10, 2? Inflated WR stars at solar Z (Petrovic et al 06) 120 M o : Mdot ~ 2. 10-5 M o /yr, lifetime~2.5 Myr => Mass loss during MS: ~ 50 M o Mdot/10 => Mass loss during MS: ~ 5 M o How to produce WR stars?
How to produce WR stars with Mdot/10? 8 Mass loss still high in RSG, LBV? Binary interactions? All massive stars are binary stars??? (Kobulnicky, aph0605069) but RSG problem Fraction of binary stars in LMC & SMC WR stars: ~30-40% (Foellmi et al 03,03) Rotation (and magnetic fields)? WR stars produced by mixing only (Maeder 87, Yoon & Langer 05) OK only for fast rotators Mdot/10 => Strong impact on rotation
Mdot/10 impact on rotation 9 Many stars would reach critical rotation: (Ekström et al in prep) Comparison Mdot: - solid = Kudritzki & Puls 00 - dotted = Vink, de Koter & Lamers 00 Difference in Mdot only a factor ~ 2
Effect of rotation on mass loss 10 Enhancement: Maeder & Meynet 2000 Anisotropy: F rad ~ g eff : Von Zeipel, 1924 M ( ) M (0) 1 4 9 1 v v 2 2 crit,1 1 1 1 1 Maeder & Desjacques 2001 T eff =25 000 K T eff =30 000 K Mass loss at critical rotation: Rotation impact ~ Mdot *2 Meynet & Meynet 05, 94 Decressin et al 2007 Jun 18, 2007 Raphael Hirschi University of Keele (UK)
Metallicity dependence of mass loss rates Ṁ Z =Ṁ Z o Z /Z o - α = 0.5-0.6 (Kudritzki & Puls 00, Ku02) (Nugis & Lamers, Evans et al 05) - α = 0.7-0.86 (Vink et al 00,01,05) Which elements dominate Mdot? O* & WR: Z dep. / Fe dom. & plateau at low Z for WR (Vink et al 05) RSG (and LBV?): no Z-dep.; CNO? (Van Loon 05) Z(LMC)~Zo/2.5 => Mdot/1.6 Mdot/2.2 Z(SMC)~Zo/5 => Mdot/2.2 - Mdot/4 11 (Vink et al 05)
Metallicity dependence: comparison with Observations 12 Ratio WR/O: tests WR lifetime Ratio SNIbc/SNII: tests final type Meynet & Maeder 05 (Mdot ~Z**0.5) Meynet & Maeder 05 Prantzos & Boissier 03 Lower Mdot would not fit the data as well Jun 18, 2007 Raphael Hirschi University of Keele (UK)
WR subtypes: comparison with Observations 13 Rotation does not help for WN star properties (Hamann,Graefener,Liermann 06) WC/WN is Z-dep Z-dep. WR Mdot fits better the data Mdot/2 (Eldridge & Vink 06) Mdot/2 does not fit as well But binary could compensate (Vanbeveren et al 07) WR/O=? Eldridge & Vink 06
Jun 18, NASA 2007- WMAP science team Raphael Hirschi University of Basel (CH) First Stellar Generations in Cosmology 14 BIG BANG NOW
Physics @ Very Low Metallicities (Z) Stars are more compact: R~R(Z o )/4 (lower opacities) 15 Z=0 (<10-10 ): only pp chains; no CNO cycle Mass loss ~ Z (0.5-0.86) => weaker winds - main contributors: CNO, Fe? Mdot ~ (CNO+Fe)**0.5 assumed Vink & De Koter 2005, Vink, de Koter & Lammers 2001, Kudritzki 2002, Van Loon 05 Higher initial masses? <M>10 ~100 M o? (Bromm 2005,...) El Eid et al 1983; Chiosi et al 1983; Ober et al 1983; Bond et al 1984; Klapp 1984; Arnett 1996; Limongi et al. 2000; Chieffi et al. 2000; Chieffi and Limongi 2002,4; Siess et al. 2002; Heger and Woosley 2002; Umeda and Nomoto 2003; Nomoto et al. 2003; Picardi et al. 2004; Gil-Pons et al. 2005 Chemical signature of Pair Instability SN (M:140-260 Mo) not observed!
Mass Loss @ Z=0? Up to 10 % (M ini ~200 M o ) mass lost due to break-up: 16 Ekström et al 2006
Mass Loss @ Very Low Z? Strong mass loss during RSG stage (M ini >~60 M o ): 17 Meynet et al 06, Hirschi 07 Mixing of CNO into envelope after MS Break-up M ini [M o ] M final [M o ] V ini [km/s] 20 19.95 600 40 35.8 700 60 48.98 800 85 19.87 800 RSG 85 M o model WO star & possible GRB progenitor?
CEMP Star: HE 1327-2326 (Frebel) (Carbon-rich Extremely Metal Poor stars) 18 Ejecta from winds match observed abundance pattern Meynet et al 06, Hirschi 05, 07 Other models: - Pop III Limongi et al 2003 - Asym. Expl. Observations: Frebel et al 04 & Aoki et al 05 (stars), Plez & Cohen 05 (triangle), Christlieb et al 04, Norris et al 04, Depagne et al 02 Umeda & Nomoto 2003 - AGBs Suda et al 2004 [X/Y]=log(X/Y)-log(X o /Y o ) Jun 18, 2007 Raphael Hirschi University of Keele (UK)
Long & Soft Gamma-Ray Bursts (GRBs) Long soft GRB-SN Ic(b) connection: GRB060218/SN2006aj Tagliaferri, G et al 2004 / Matheson 2003,... / Iwamoto, K. 1999,... GRB 031203-SN 2003lw / GRB 030329-SN 2003dh / GRB 980425-SN 1998bw,... Collapsar progenitors must: (Woosley 1993, A. Mc Fadyen) Form a BH Lose their H-rich envelope WR star Core w. enough angular momentum Observational info: Z of close-by GRBs is low 19 Cusumano et al 2006,... ~ Z (Magellanic clouds) (Stanek et al 06, Le Floc'h er al 2003, Fruchter et al 2006) (simulation by Mc Fadyen) GRB from runaway massive stars? (Hammer et al 2006) Jun 18, 2007 Raphael Hirschi University of Keele (UK)
85 M o Model @ Z = 10-8 20 WO type WR star High j final GRB
Jun 18, 2007 Raphael Hirschi University of Basel (CH) GRB progenitors with B-Fields Taylor-Spruit dynamo (Spruit 2002) : better for NS (Heger et al 2005) A BH ~1 Quasi chemically-homog. evol. of fast rot. stars (avoid RSG) (Yoon & Langer 06, Woosley & Heger 2006) 21 40 M o models V ini [km/s] Z o Z(SMC) Z=10-3 Z=10-5 Z=10-8 ~230 - - - No - ~300 - WR - - - 400-500 WR WR WR WR No 700 - - - - WR WR (SNIb,c) & GRBs predicted down to Z=~0 (Yoon et al 06) Question: This study GRBs around Z(LMC) & Z(SMC)? Dep. On mass loss / NO GRB @ Z o (Meynet & Maeder 2007)
Conclusions 22 Mass loss is important for massive stars Reduction factor ~ 10 unlikely (WR/O ratio, critically rotating *) Rotation and binarity may compensate Mdot/2 Z=0 models up to 10 % mass loss break-up WR, SNIb,c and GRBs predicted from second stellar gen. Max. Z(GRB) ~ Z(MC), dep. on mass loss, solar Z and B-fields