How Massive are the First Stars? Statistical Study of the primordial star formation M popIII ALMA 北海道大学 / Jan , 2013 ○ Shingo Hirano 1 Takashi Hosokawa 1, Naoki Yoshida 1, Kazuyuki Omukai 2, H.W.Yorke 3 1 University of Tokyo, 2 University of Kyoto, 3 JPL/Caltech Variety of PopIII protostellar evolution 3 protostellar accretion paths M popIII = 10 – a few 100 M sun
How Massive are the First Stars? 2 Primordial Halo Cosmological Simulation z =17 Protostar Core ( ~ 0.01 [M sun ] ) 600 kpc/h (comving) Accretion Phase of the Primordial Protostar ■ Different thermal evolution (main coolant is H 2 molecular) M cloud ~ 1000 [M sun ] ZERO metallicity ■ No Metal & Dust No radiation pressure (?) (cf, PopII, I star formation) M popIII ~ 1000 [M sun ] (?) UV Radiative Feedback Stalls mass-accretion
UV Radiative Feedback 3 Ultraviolet (UV; hν > 13.6 [eV]) radiation from the protostar Ionizing infalling neutral gas & creating HII region Thermal pressure of the ionized region (high temperature) is much greater than that in neutral gas of the same density McKee & Tan (2008) Gas on the circumstellar disk is photo-ionized & heated photo-evaporation Growth of HII region Breakout & Expansion Accreting star emits the ionizing UV photons
Accretion History of Protostar 5 M popIII = 43 [M sun ] moderate massive Accretion Rate [M sun /yrs] … however, M popIII depend on the initial quantities : Primordial Star–Forming Cloud Can be calculated by Cosmological Simulations Can be calculated by Cosmological Simulations Hosokawa et al. (2011) Radiative Hydrodynamics (RHD) Protostar Evolution UV radiative feedback Mass Accretion M star [M sun ]
Aim & Method 6 Determining the initial mass distribution of the PopIII stars (massive side; in case of the single-star formation) ■ Cosmological Simulation primordial star-forming halos ■ RHD + Stellar Evolution accretion histories Cosmological Simulation Accretion Histories M popIII Distribution M popIII Distribution Primordial Gas Clouds Primordial Gas Clouds
Cosmological Simulation 7
8 GADGET-2 : parallel SPH+N-body code (Springel 2005) + Primordial Chemistry (Yoshida et al. 2006, 2007) Initial Condition : z ini = 99, WMAP-7 (Komatsu et al. 2011) + zoom-in re-simulation technique M resolve, init < 500 [M sun ] < M cloud Stop calculations when the collapsing center becomes : n cen ~ [cm -3 ] (L resolve ~ ー [pc] ~ 2 ー 20 [AU]) N sample L box [kpc/h] (comving) N zoom L soft [pc/h] (comving) L soft [pc] (z=19) m sph [M sun ]
Primordial Star-Forming Clouds N cen ~ [cm -3 ] R [pc] N H [cm -3 ] Gao et al. (2007) Density profiles evolve self-similarly
Infall Rate of Collapsing Cloud Infall Rate [M sun /yrs] = 10 M enclosed [M sun ] Infall Rate [M sun /yrs] ~ – M enclosed [M sun ] V rad [km/sec] N H [cm -3 ] Characteristic quantities of clouds :
Protostellar Accretion Phase 11
Protostellar Accretion 12 Using the setting & method in Hosokawa et al. (2011) Radiative Hydrodynamics (RHD) ■ 2D-axsymmetric ■ Self-gravity, Hydro ■ Primordial Chemistry (15 reactions with H, H +, H 2, H -, e) ■ Radiative-transfer : cooling, feedback ■ L cell,min ~ 25 [AU], L box = 1.2 [pc], M total ~ few 1000 [M sun ] Protostar Evolution Mass Accretion UV radiative feedback * For calculating the case of the high mass accretion rate, we adopt a simple model of the stellar evolution
“Super-Giant” Protostar 13 Hosokawa et al. (2012) M star [M sun ] R star [R sun ] M enclosed [M sun ] Infall Rate [M sun /yrs] dM/dt > 0.04 [M sun /yrs] No KH contraction (“Super-Giant” Protostar ) dM/dt > [M sun /yrs] L tot (M)| ZAMS > L edd, cannot reach ZAMS
Model of “Rebound” Phase 14 Hosokawa et al. (2012) M star [M sun ] R star [R sun ] 1 1 ①②①② 2 2 * Ignore expansion phenomena By expansion, the effective temperature, T eff, decreases this phase is not important for the UV radiative feedback L tot ~ L edd Scaling : R star // R ZAMS L star // L ZAMS dM/dt < 4E–3 [M sun /yrs] Contraction to ZAMS (KH timescale)
Accretion History : one sample 15 ZAMS Mass Accretion KH Contraction ZAMS
16 M star [M sun ] R star [R sun ] Accretion Histories M star [M sun ] Accretion Rate [M sun /yrs] 10 0 Super-Giant / Rebound / Fiducial Three paths exist
17 M star [M sun ] T eff [K] 5000 [K] Effective Temperature × UV Radiation
Accretion History onto Protostar 18 M star [M sun ] Accretion Rate [M sun /yrs] dM/dt > 0.04 [M sun /yrs] dM/dt > [M sun /yrs] dM/dt < [M sun /yrs] 11 / 108 … “Super-Giant” Phase 36 / 108 … “Rebound” Phase 61 / 108 … Become ZAMS 11 / 108 … “Super-Giant” Phase 36 / 108 … “Rebound” Phase 61 / 108 … Become ZAMS Hosokawa et al. (2012) Star cannot become the Zero-Age Main-Sequence (ZAMS) structure Omukai&Palla (2003) KH contraction & ZAMS directly KH contraction stage disappears entirely
Initial infall rate v.s Final M popIII 19 Good Correlation : (4πR 2 ρv rad ) Jeans M popIII Simple Estimation : M popIII ∝ (4πR 2 ρv rad ) Jeans Decide M popIII without the calculation of accretion history (* Not consider fragmentation) M popIII [M sun ] (4πR 2 ρv rad ) Jeans [M sun /yrs]
Count 20 M popIII [M sun ] Heger & Woosley’02 Final fate of the non-rotating PopIII stars ■ 15 < M PopIII < 40 Core Collapse SNe ■ 40 < M PopIII < 140 Black Hole ■ 140 < M PopIII < 260 Pair-Instability SNe ■ 260 < M PopIII Black Hole * with rapid rotation M PISN > 65 [M sun ] Chatzopoulos&Wheeler(2012) M popIII Distribution
Summary ■ more than 100 primordial halos show the wide range of accretion history ■ Three type of accretion histories (1) low dM/dt KH contraction UV radiative feedback (2) High dM/dt cannot reach ZAMS mass accretion continues (3) HUGE dM/dt “supergiant” protostar mass accretion continues M popIII = 10 – a few 100 [M sun ] □ Correlation between (4πR 2 ρv rad ) Jeans – M popIII Can estimate M popIII by using Jeans quantity 21 M popIII [M sun ] (4πR 2 ρv rad ) Jeans [M sun /yrs] M star [M sun ] R star [R sun ] 10 0