1. Abundance Patterns to Probe Stellar Nucleosynthesis and Chemical Evolution Francesca Primas

2. Setting the Stage DLA: dominant reservoir of neutral baryons measure the mean metallicity in HI (5< z < 0) 1. High-z universe, i.e. looking at objects in their infancy stages 2. Nearby universe, i.e. looking at the fossilized imprint left by the first generations of stars EMP: precious witnesses of the early evolutionary phases of our Galaxy Identify the imprint left by the first SNe explosions [Fe/H] = log(Fe/H) * - log(Fe/H) sun Old Metal-poor

3. Stellar Nucleosynthesis light elements BBN, CR spallation alpha and iron-group SN physics (time delays) and SN imprintheavies n-capture

4. The Light Elements Li Be B (670.7nm, 313.0nm, 250.0nm) Implications for cosmology BBN vs IBBN (Li) Implications for stellar structure LiT=2.5x10 6 K BeT=3.0x10 6 K BT=5.0x10 6 K Implications for nucleosynthesis and cosmic-ray physics classical spallation (Reeves et al. 1970) ? primary ? neutrino-spallation ( 11 B) ?

5. Latest Abundances Cayrel et al. 2004 McWilliam et al 1995

6. What Have We Learned well defined trends with low dispersion all the way to the most MP stars as quality , the dispersion  -- SMALL in most cases: 0.05-0.15dex -- THINNEST of all: Cr,Ti -- MOST DISPERSED: Mn (~0.2dex for [Fe/H]<-3) there is NO unambiguous detection of products of PISN constraints on SN II yields: M~15-50 M sun, but also up to 100M sun or hypernovae are able to reproduce the observed trends (mixing and fallback) Abundancedispersion  =0.05dex! Abundance dispersion  =0.05dex! Cayrel et al. (2004) data

7. Stellar Highlights HE 0107-5240HE 1327-2326 HE 0107-5240 ([Fe/H]=-5.3) and HE 1327-2326 ([Fe/H]=-5.5 (large[C/Fe], but different in N, Na, Mg, Al) C-rich metal-poor stars C-rich metal-poor stars: 20-30% (?) r-process rich stars

8. The Detailed Picture CS 22892-052 Galaxy at z=2.63 CS 31082-001 Sneden et al. 2003 Prochaska et al. 2003 Hill et al. 2002

9. 1. Does Os really deviate from the solar r-process pattern ? Not anymore after new gf value ( Ivarsson et al. 2003 ) ==> Os=Ir 2. Still missing input atomic physics for Ho, Lu, and Yb!

10. Asplund 2002 What’s next ? ---> ---> Atomic physics and (best) abundance indicators ---> ---> Model atmospheres and theory of line formation LTE vs NLTE 1D vs 3D Asplund 2005 (ARAA)

11. The End

13. Collet et al 2006 CNO Fe 1D vs 3D

14. age EMP = old ?? Not quite, presumably old ==> age In r-process rich stars: Dt = 46.67(log(Th/S) o - log(Th/S) obs ) Problem: evaluation of the log(Th/S) o based on the assumption that the r-process is universal [ cf. Cowan (pro) and Goriely (con) ] Warning: actinides and lower-mass r-nuclei may vary strongly (despite the constancy for Z=56-82) [ cf. Hill et al. 2002, Honda et al. 2003 ] Th (and U) seem to be over-abundant: log(Th/Eu) = -0.22 (wrt ~ -0.6! for similar stars) Cosmo-chronometry

15. Age difference ? NO, otherwise CS 31082-001 would have a negative age, compared to the other n-rich stars ! Ab initio enhancement ? IF SO, r-process may not be universal ! Th/Eu =? a reliable chronometer CS 31082-001: Th and U were detected and used Dt = 21.76(log(U/Th) o - log(U/Th) obs ) + similar ionization and excitation potentials --> errors largely cancel out + initial production ratio: more robust against variations in n-exposure (nuclear reaction network code containing more than 3500 isotopes with all relevant reactions! ) Cosmo-chronometry log(U/Th) obs = -0.74 (0.15) … -0.94 (0.11) Age = 14.0 +/- 2.4 Gyr 8 Th lines, each with gf determined better than 0.08dex 1 U line with gf determined better than 0.06dex log(U/Th) o = 0.50 (0.02) ! Nilsson et al. 2002a,bGoriely & Arnould 2001