The Cecilia Payne-Gaposchkin Lecture Center for Astrophysics May 9, 2002

Heavy Metal from Ancient Superstars In collaboration with….  Debra Burris (Oklahoma City CC)  John Cowan (University of Oklahoma)  Chris Sneden (University of Texas)  Taft Armandroff (NOAO)  Henry Roe (U.C. Berkeley)

Outline  The high-redshift universe  The Galaxy in time – a brief review of the formation of the Milky Way and its structural components  The origin of the elements – back to B 2 FH!  A stroll through the Periodic Table  A timeline of Galactic chemical enrichment - What does it all mean?

Metallicity at High Redshift Studies of the most metal-poor stars in the Galaxy give us access to the state of the Universe at very early times Songaila & Cowie 2002 But - the most metal-poor stars in the Galaxy have [Fe/H]=-4

Metallicity Distribution Function for Metal-Poor Stars NO stars with [Fe/H] < -4.0 Have we found the low metallicity end of the MDF? Did the first generation(s) raise the metallicity to [Fe/H] = -4? Beers 1999 (Selection effects for [Fe/H] > -2)

Circa 1990 The Milky Way…. Halo Flattened Inner Halo Thick Disk Dwarf Spheroidal Companions Dark Matter Corona Circa 1950

The Chemistry of Stellar Populations……  The chemical compositions of stars reflect the star formation histories of stellar populations  The complexity of the Milky Way’s history is reflected in the compositions of its stars

Solar Abundances from Grevesse and Sauval

Jargon  [m/H] = log N(m)/N(H) star – log N(m)/N(H) Sun  [Fe/H] = -1.0 is the same as 1/10 solar  [Fe/H] = -2.0 is the same as 1/100 solar  [m/Fe] = log N(m)/N(Fe) star – log N(m)/N(Fe) Sun  [Ca/Fe] = +0.3 means twice the number of Ca atoms per Fe atom  Log  (metal) = log n(metal)/n(H) + 12

The Fe Chronometer But [Fe/H] is not a good indicator of the age of the disk Why Iron? Fe is abundant Fe is easy Fe is made in supernovae In the halo, [Fe/H] is a function of both time since star formation began and the star formation rate

Chemical Evolution Nucleosynthesis in stars leads to chemical enrichment of the Galaxy Rate of enrichment depends on sites and mechanisms of nucleosynthesis The variables are: – Star formation rate – Initial mass function – Yields –Stellar evolution time Primordial nucleosynthesis Hydrogen burning –Proton-capture chains Helium Burning –C,Ne,O,Si burning Photodissociation burning process Neutron-capture processes Odd-ball stuff

Galactic Lithium Production: Figure from Con Deliyannis 10% of Big Bang origin 90% of Galactic origin

“Alpha” Elements: Edvardsson et al.; Pilachowski et al.; McWilliam et al. Excesses at low metallicity from C, Ne, O & Si production in SN II Decline in [alpha/Fe] due to Fe production by SN Ia

How to Make Heavy Metals: neutron-capture processes  r-process –High neutron flux –Type II Supernovae (massive stars) –No time for  -decay –Eu, Gd, Dy, some Sr, Y, Zr, Ba, La…  s-process –Low neutron flux –Time for  -decay before next neutron capture –No Eu, Gd, Dy Main s-process Low mass stars Double shell burning Makes SrYZr, Ba, etc. Weak s-process Massive stars He-core and shell Burning Lower neutron flux makes SrYZr only

Solar System r- and s-Process Elements

Isotopes built by n-capture syntheses The valley of b-stability Rolfs & Rodney (1988)

n-capture Synthesis Paths Ba La Cs Xe pps,r s r r p s ss r-process path s-process path 18% of solar system Ba is odd, but 48% of r-process Ba is odd

Spectrum of HD  Metal-poor giant  [Fe/H] =  Teff = 4910 K  r-process rich  Spectrum from the Mayall 4-m echelle

Star-to-Star [n-capture/Fe] variations Stars of similar temperature and metallicity may have very different neutron-capture element abundances Burris et al. (2001)

The Scatter is in the Stars! The r-process elements vary together Burris et al. (2001)

Abundance Data Sources  -1.0 < [Fe/H] < 0.0 Edvardsson et al Jehin et al  -3.0 < [Fe/H] < -1.0 Burris et al  -4.0 < [Fe/H] < -2.0 McWilliam et al. 1995, 1998

Heavy Metal Abundances Note:  Scatter  Deficiencies at low metallicity  Excesses at intermediate metallicity

n-capture Abundances in BD+17 o 3248 Scaled solar-system r-process curve: Sneden 2002

Solar-System s-process Abundances DON’T Fit Sneden (2002), Burris et al. (2000)

BD is typical of very metal poor stars Sneden et al. (2000); Westin et al. (2000); Cowan et al. (2002)

r-Process vs. s-Process Transition from r-process only to r+s process at log  (Ba)=+0.5 and [Fe/H] = -2.0

La/Eu at low metallicity s-process seen at [Fe/H]=2.1 Simmerer et al. (2002)

When does the s-process start?  Main s-process occurs during thermal pulses in AGB stars of 2-4 solar masses  H mixes inward, giving 12 C(p,e+) 13 C 13 C( ,n) 16 O  t ~ 10 8 years  s-process elements do not appear before this

r-process appears at [Fe/H]=-2.9

New r-process elements come from deep in the Supernova This may be part of the reason for the n-capture scatter. Not all SN II produce lots of r-process Rolfs & Rodney (1988)

The “light” heavy metals Production of Sr, Y, and Zr requires an additional neutron capture process

Heavy metals at [Fe/H] = -4 At very low metallicity, the production of heavy metals is dominated by an unknown process

What came before the r-process? Identify “weak r-process stars” to see yields of very early nucleosynthesis

The Earliest Star Formation Formation of stars as “pre-galactic” objects from small density fluctuations H 2 provides cooling Masses from a few tens to a few hundred solar masses Low mass star formation is suppressed by reionization Provides early chemical enrichment Abel, Bryan, & Norman 2002

Theoretical Framework  Stochastic model for early chemical evolution (Travaglio et al. 1999)  Coalescing and fragmenting clouds  Homogenization time scale ~ few x 10 8 years reduces scatter  Suggests r-process from 8-10 M Sun  s-process elements from 1-3 M Sun  AGB stars after homogenization  Stochastic model for early chemical evolution (Travaglio et al. 1999)  Coalescing and fragmenting clouds  Homogenization time scale ~ few x 10 8 years reduces scatter  Suggests r-process from 8-10 M Sun  s-process elements from 1-3 M Sun  AGB stars after homogenization

Theoretical Models of Chemical Evolution Stochastic models of Travaglio et al. for r-process production by 8-10 solar mass SN II

The scatter in the abundances of all of the n-capture elements from star-to-star is of astrophysical origin, and the scatter increases as metallicity decreases. The heavy n-capture elements were formed predominantly by the r-process at metallicities below [Fe/H] = Significant production of r-process elements began when the metallicity of the Galaxy reached [Fe/H] = -3. Elements from the s-process appear at a metallicity of [Fe/H] = -2.1, when low-mass AGB stars begin to contribute from double shell burning. The s-process then dominates Ba production. The origin of heavy metals at the lowest Galactic metallicity ([Fe/H] = -4) is still not understood, but may be dominated by the weak s-process, or by a separate r-process in massive stars. Conclusions

The Epochs of Galactic Chemical Evolution  Primordial Epoch -The Big Bang  Epoch of Massive [Fe/H] ~ -4 –Ca, O, Sr-Y-Zr + ?  r-process Epoch - r-process elements from 8-10 MSun SNII  The Double Shell Epoch yields s-process [Fe/H]=-2.1 (~ 10 9 years)  The Iron Epoch – from SN [Fe/H]=-2  The Lithium [Fe/H]=-1.0 from ??? Key Concept – Stellar evolution timescales are important

The Oxygen Abundance Oxygen abundances are still uncertain, with inconsistencies between the triplet and forbidden lines