What and Where are the Lowest

1
What and Where are the Lowest
Metallicity Stars? Kim A. Venn Univ. of Victor
ia First Stars III, Santa Fe, NM July 16, 2
007
2

• Review some basic assumptions
• Where?
• Why is the Galactic halo a good place
• to search for metal-poor stars?
• Are there other good places to look?
• 2. What?
• How do we know the metallicity of a
• star and interpret its abundance ratios?
• Are there other interpretations?

3

• Why search the halo?
• in monolithic collapse scenario, oldest
• most metal-poor stars expected in the outer
  halo.
  
• location of metal-poor globular clusters.
• proper motion surveys in the 1960s found
• metal-poor stars with halo kinematics,
• e.g., HD 122563 by Wallerstein et al. 1963,
  with Fe/H -2.6

4

• Targetted searches for metal-poor stars in the
• halo ongoing for decades, e.g., Bond (1970,
  1980)
  
• Stromgren m1 index, and/or objective prism
  surveys (CH),
  
• then high res spectroscopy
• since magnitude limited, then look for giants
• Now, the gold standard are the HK and
• Hamberg/ESO (HES) surveys for the Galaxy
• regularly find stars below Fe/H -2.5
• see talks by Beers, Christlieb, Frebel

5
Johnston Bullock 2003
Now that we live in a ?CDM Universe,
galaxy formation happens through mergers
the halo is thought to be formed from disrupted
dwarf galaxies (of some sort, lots of questions
still) what about searching in the dwarfs?
6
models with strong feedback Pop III wind eff.
models with weak feedback
Scannapieco et al. (2006) first stars first

enriched only in first products second
7
DART Dwarf Abundances Radial velocities Team
ESO VLT FLAMES LP (PI Eline Tolstoy) The MD
Fs are from CaT spectroscopy, and show wide
variety due to diff.s in SFHs (also seen in th
eir CMDs)
Helmi et al. 2006 Note Carina CaT data from ESO
archive
for a different LP (PI Glmore)
8
Comparison of MDFs Normalized each at Fe/H
?2.5 not same distributions Small stat
istics of course, but checks on our methods sugg
est that we are sensitive to stars with Fe/H
e surviving dwarfs are not similar to MWG buildi
ng blocks, even at earliest epochs.
HES
Helmi et al. 2006
9
IMF? Did the lowest mass stars form in the fi
rst SF episodes? Pre-enrichment? Were the
dwarf galaxies pre-enriched to Fe/H -3 befor
e SF began? What does this have to do with fi
rst stars? if pre-enrichment, then should have
been from the first stars true for metal-poo
r stars in dwarfs and for those in Galactic halo
from earliest collapsed dwarfs.
Helmi et al. 2006
10
Tumlinson 2005 Below Fe/H ???ish
, polluted only by first stars.
Notice only ONE SN progenitor for Fe/H Number of first stars contributing per
Fe/H B Number of SN progenitor contribut
ers per Fe/H D
11
If stars with metallicities of -3 are polluted by only a few first stars, then what
are the Fe/H -5 stars? e.g., HE1327-2326
(Frebel et al. 2004, Aoki et al. 2006)
Comparison of chemistry with othe
r known Fe/H -5 stars shows similariti
es to HE1327 (Venn Lambert 2007)
These are chemically peculliar
stars - post AGB stars that show
separation of dust-gas, with dust
seen in circumbinary disks
(Lamers 1986, Waelkens et al.
1991 van Winckel et al. 1992)
The elements with intermediate
Tcs are critical tests, e.g., S,
Zn, Na
Tc condensation temperature
12
Chemical peculiarities due to dust-gas separation
are not unique to post-AGB stars Lambda B
ootis stars (Venn Lambert 1990, Hei
ter et al. 2002, Paunzen et al. 2003)
young, single, A-type stars (Vega),
found in open clusters short lived phas
e of of accretion. RV Tau stars
(Giridhar et al. 2005 Gonzalez Lambert
1997) evolved red giants, found in the
field and globular clusters, single and bin
ary stars. Thus, more than one mechan
ism to separate gas and dust.
Tc condensation temperature
13
What about the other Fe/H HE0107-5240 (Christlieb et al. 2002, Bessell et a
l. 2002) Similar pattern seen.
esp.ly due to Na. Really need S or Zn
abundances to confirm!!. (Venn
Lambert 2007)
C N O
S Zn
Sr Ba
Na
Fe Mg Ni
Si Ti Ca
Tc condensation temperature
14
Other elemental abundances and upper limits for
HE0107 still resemble the dust-gas separation pat
terns expected. Lambda Bootis stars
RV Tau stars Implication is
that intrinsic metallicities for both
HE0107 and HE1327 are Fe/H ???to???.
S and Zn are critical tests and would
establish intrinsic metallicities accurate
ly.
Tc condensation temperature
15
Another test dust can be associated with IR
excesses
Details of dust temperature, composition, and
distance from host star all affect IR emission
and its detectability. Not all dusty stars WIT
H IR excesses show dust-gas pattern either though
still lets look for it anyways.
16
No IR excesses found in HE0107 nor HE1327,
however dust could be too cool or too distant fro
m the stars
Venn Lambert 2007
UBVRI from Christlieb et al. 2002 JHK from 2MASS
UBVRI from Aoki et al. 2006 JHK from 2MASS
17
If HE1327 and HE0107 are chemically peculiar
stars, then those with -4 important ones to study nucleosynthesis and/or po
llution from first stars e.g., HE1300
0157 with Fe/H ???? N
otice that while (LTE) carbon oxygen are
high, Na is not (Frebel et al. 2
007) Thus, not obviously the dust-
gas separation pattern (again, S, Zn would c
onfirm)
18

• Summary
• Searches for first stars (or pollution) in the
  halo and
  
• dwarf galaxies both require first stars to form
  over
  
• several epochs, e.g., z 10 to 5.
• Few stars with Fe/H galaxies.
  
• Probably pre-enrichment from earliest epochs.
• Possibly those stars do sample first star
  pollution.
  
• Hyper metal-poor stars in the MW with Fe/H
  
• could be chemically peculiar stars.
• Appear metal-poor due to dust-gas separation.
• If so, then -4 (or pollution)