The First Stars and Evolution of the Early Universe Seattle July 3 7 2006 Star Formation with the First Dust Raffaella Schneider INAF/OAArcetri
Collaborators Simone Bianchi INAF/ Benedetta Ciardi MPA Andrea Ferrara SISSA Akio Inoue Roberto Maiolino Kazuyuki Omukai Toru Tsuribe Stefania Salvadori Ruben Salvaterra Evan Scannapieco KITP INAF/OAArcetri Osaka Univ. INAF/OAMonteporzio NAO Tokyo Osaka Univ. SISSA Univ.. Como
Outline motivations dust and fragmentation of star-forming clouds dust production in the early universe critical metallicity for IMF transition consequences on the MDF of halo stars
Motivations (I) Indications for an IMF transition from PopIII to PopII/I stars Population III Z 10-12 -10-10 Very Massive Stars 30 Msun < M < 600 Msun Mc 100-300 Msun Bromm et al 01 Schneider et al 02,03 Bromm & Loeb 04 Omukai et al 05 Z > Z cr Population II/I Z 10-5 -10-4 / Z 0.02 Normal stars 0.1 Msun < M < 100 Msun Mc 1 Msun (II) Discovery of Hyper Metal-Poor Stars (HMP) [Fe/H] = -5.3 [Fe/H] = -5.4 Are these stars low-mass Pop III? Their existence contradicts Z cr? Christlieb et al 02,04 Frebel et al 05
Overview of different studies Bromm,, Ferrara, Coppi & Larson (2001) SPH simulation of collapsing dark matter mini-halos no H2 or other molecules, No dust only C and O atomic cooling 10-4 Z sun < Z cr < 10-3 Z sun Z = 10-4 Z sun Z = 10-3 Z sun Bromm & Loeb (2003) one-zone model only CII and OI line cooling Initial condition: Tc=200 K nc=10 4 cm -3 cooling rate Λ (n, T ) = 1.5 n KT /t CII,OI c c c c ff compressional heating rate [C/H]cr = -3.5 ± 0.1 [O/H]cr = -3.05 ± 0.2 RS et al (2002-2003); 2003); Omukai, Tsuribe,, RS, Ferrara (2005) one-zone model H2, HD and other molecules Metal coolingin gas-phaseand dust 50 chemical species 478 reactions 10-6 Z sun < Z cr < 10-4 Z sun
Evolution of Star-Forming Clouds at low Z Z 10-6 Z sun M Jeans 10 3 M sun Z 10-5 Z sun M Jeans < 1 M sun Schneider et al 2002 Omukai et al 2005 One-zone model with simplified dynamics but detailed chemical and thermal evolution (478 reactions for 50 species) D chemistry and HD TRANSITION cooling IN FRAGMENTATION SCALES CMB as the only external source of radiation for Z cr = 10-5±1 Z sun
With CMB radiation field at z = 20 level populations of atomic/molecular lines heating of dust grains T gas and T dust T CMB The evolution is affected at low T (high Z)
Role of Metals in Gas and Dust Metals in dust: - cooling by grains -H 2 formation on grains fiducial model Metals in gas: DUST GRAINS -DOMINATE cooling by atomicthe lines EVOLUTION AT LOW METALLICITIES Z < 10-3 Z sun Schneider et al 2003; Omukai et al 2005
Dust Production in the Early Universe Early LocalUniverse ISM dust gas originates phase elemental from AGB abundances stars τ evo are 1-2 different Gyr MRN dust sizeproduction distributionon short time-scales τ evo < 50-100 Myr dustin composition SN expanding ejecta abundances with different from localsize observations distribution (Pollack and composition et al 04) Kozasa & Hasegawa 1987; Todini & Ferrara 2001; Nozawa et al 2003; Schneider, Ferrara & Salvaterra 2004 mm/submm observations of high-z QSOs M dust 10 9 M sun at z >6 Bertoldi et al 03; Priddey et al 03 near-ir observations of z=6.22 QSO extinction curve consistent with SN dust Maiolino et al 04? Direct evidence for efficient dust formation in SN ejecta is lacking Spitzer obs of SN remnants Cas A @ 24µm Hines et al 04 M dust 10-4 10-3 M sun mid-ir instruments warm dust component (50-100 K) Cas A @ 160µm M dust < 0.2M sun far-ir and submm obs cold dust component (< 30 K) Krause et al 2004 major limitation: contamination by clouds along the los
Observations of dust condensation within a few years of the explosion HST WFPC2 SST IRAC 3.6 µm, 4.5 µm, 8 µm Sugerman et al 2006 Science June 8 499 days 670 days 8 Msun SNII 2003gd in NGC 628 mid IR excess consistent with cooling dust increasing optical extinction asymmetric blue-shifted emission-line profiles same as for SN1987A used to calibrate the models M dust 0.02 M sun with T dust 500K on going continued long-term monitoring of a large sample of young typeii SNe
Dust production in the first SNe SNII Z=0 M=22 M sun PISN Z=0 M=195 M sun Todini & Ferrara 2002 Schneider, Salvaterra & Ferrara 2004 M met,in = 2.8 M sun M met,fin = M O = 1.47 M sun f dep =M dust /M met = 0.38 M met,in = 87.75 M sun M met,fin =M O = 17.23 M sun f dep =M dust /M met = 0.65 M dust 5% M prog f dep > 20% 12 M sun < M prog < 40 M sun M dust 20-30% M prog f dep = 60-70% 140 M sun < M prog < 260 M sun
Destruction in the reverse shock - semi-analytic model - only thermal sputtering Main results: - small grains are evaporated but partially replaced by the erosion of larger ones - size distribution function is flatter > 20% of dust mass survives Bianchi, Schneider & Pinna in prep Destruction in high-v interstellar shocks Nozawa, Kozasa & Habe 2006 efficiency of dust destruction depends on: - initial grain size distributions - density of the ISM 0.3 < ε < 0.4 -energyof SN explosions n = 1 cm -3 and 1 < E 51 < 30 > 30% of the dust survives
Fragmentation of pre-stellar clouds enriched by the first SNe Schneider, Omukai, Inoue & Ferrara 2006 line cooling dust cooling line cooling dust cooling - dust-induced fragmentation leads to subsolar mass fragments at Z cr 10-6 Z sun - strong depletion of C atoms on graphite grains OI is the main coolant in gas-phase - line-induced cooling leads to fragment masses 100 M sun up to Z 10-2 Z sun
Fragmentation of pre-stellar clouds enriched by the first SNe Schneider, Omukai, Inoue & Ferrara 2006 - dust-induced fragmentation leads to subsolar mass fragments at Z cr 10-6 Z sun - strong depletion of C atoms on graphite grains OI is the main coolant in gas-phase - line-induced cooling leads to fragment masses 100 M sun up to Z 10-2 Z sun
Constraints from the observed MDF i. What are the implications of the observed low-z tail of the halo MDF? ii. iii. A sharp cut-off can be related to the shape of primordial IMF and Zcr? What is the statistical impact of second-generation stars in the observed sample? Chemical evolution and stellar population history along the hierarchical evolution of the Galaxy Salvadori et al 2006 Stars form in Lya cooling halos T vir >10 4 K feedback E sn >E b Z =20 Galactic medium M < M res M acc Z<Z cr PopIII m 200 M sun Z>Z cr PopII/I Salpeter IMF 0.1 M sun <m<100 M sun MW 10 12 M sun Monte Carlo SN wind merger efficiency treeε wind algorithm star formation Cole et al. efficiency 2000 f *
Good agreement with EPS theory
Reproducing the MW average properties close box model model with feedback
Results Close Box models do not reproduce the observed MDF All PopII/I stars form with [Fe/H] > -1.8 Z second stars M met,popiii /M gas = 0.45 M * /M gas = 0.45 f * 0.2 Z sun Metals can be diluted only through mergers with unpolluted progenitors feedback The lowest-z PopII stars form from gas Z Z cr accreted from the galactic medium onto an unpolluted progenitor The analyzed sample does not allow to discriminate among values of Z cr Z cr = 10-4 Z sun Z cr = 10-6 Z sun Z cr = 0 - A Salpeter-like IMF (Z cr = 0) is still compatible with the observed MDF Ryan & Norris 1991 -Z cr < 10-5 Z sun is required to account for the 2 HMP stars [Fe/H] < -5 averaged over 200 realizations
Galactic medium enrichment
Cumulative MDF Z cr = 10-4 Z sun Z cr = 10-6 Z sun Ryan & Norris 1991 averaged over 200 realizations
Results (II) Fraction of observed PopIII stars: F 0 < 1/N obs ([Fe/H]<-2.5) = 1.9 10-3 Tumlinson 2005 Z cr F 0 10-4 Z sun <10-8 Z 10-6 Z sun <10-8 cr MUST BE > 0 0 2.8 10-2 PopIII stars must have m > 0.9 M sun Statistics of second-generation stars: enriched by yields of PopIII SN Z cr = 10-4 Z sun Z cr = 10-6 Z sun Z cr = 0 PopII 1st PopII all 20% Cayrel et al (2004): 30 stars with -4.1 <[Fe/H] < -2.7 Z cr Expected 57% number of second-generation stars 0.05% 10-4 Z sun 0.34 10-6 Z sun 0 0 0 0.004% 2% accretion mode self-enrichment mode
Conclusions The observed MDF does not constrain the value of Z cr but the existence of 2 HMP stars implies Z cr 10-5 Z sun The observed value of F 0 (fraction of PopIII stars [Fe/H]< -2.5) implies that Z cr > 0 and that m PopIII > 0.9 M sun For 0 < Z cr 10-5 Z sun dust condensed in PopIII SN explosions has a fundamental role to form low-mass second generation stars The statistical impact of second generation stars in the observed MDF is very small: < 1 for the Cayrel sample