Observations of Neutron-Capture Elements in the Early Galaxy

Chris Sneden

University of Texas at Austin

Involving the Efforts of Many People, Including :

John Cowan

Jim Truran

Scott Burles

Tim Beers

Jim Lawler

Inese Ivans

Jennifer Simmerer

Caty Pilachowski

Andy McWilliam

George Preston

Debra Burris

Bernd Pfeiffer

Karl-Ludwig Kratz

Francesca Primas

Rica French

Taft Armandroff

Talk outline

Reminder of solar r- and s-process breakdownGeneral n-capture trends in the Galactic halo Star-to-star scatter Shift to r-process dominance

Detailed abundance distributions in a few stars Elemental Isotopic

Radioactive element observationsThere is more to halo star life than the r-processSummary, future questions

A detailed view of part of the n-capture synthesis paths

Ba

La

Cs

139

132131130129128

130 132

133

134 136

134 135 136 137 138

138

PP s,rs,r s,r

s,r

s,r

s,r

s

rs,r r

p

s

s,r ss

r-process paths-process path

Xe

ELEMENTAL r- and s-process solar-system abundances

Data from Burris et al. (2000)

General halo n-capture “bulk” abundance trends: LARGE scatter

Large-sample surveys are needed to show this: Gilroy et al. (1988), McWilliam et al. (1995); Ryan et

al. (1996); Burris et al. (2000); Johnson & Bolte (2001)

Obvious from simple spectrum comparisons

σ[n-capture/Fe] > 1 dex

local nucleosynthesis events occurring in a poorly mixed early Galactic halo

Stellar Spectroscopic Definitions

[A/B] = log10(NA/NB)star – log10(NA/NB)Sun

log (A) = log10(NA/NH) + 12.0

Atmospheric parameters: Teff, log g, vt, [Fe/H]

Metallicity [Fe/H]

Metal-poor halo star [Fe/H] < -1.5

Very metal-poor star [Fe/H] < -2.5

Sr II line strength variations at lowest metallicities

McWilliam et al. (1995)

All three stars have similar atmospheric parameters and[Fe/H] ~ -3.4

Strontium abundance scatter at lowest metallicities

McWilliam et al. (1995): filled circles

Gratton & Sneden (1994): open squares

n-capture/Fe variations are obvious even in spectra of “higher” metallicity stars

These two metal-poor ([Fe/H]=-2.3) giants have similar atmospheric parameters

Burris et al. (2000)

n-capture abundance variations do not occur at random

Comparison with an ordinary metal

Comparison with nearby n-capture element Dy

Burris et al. (2000)

General halo n-capture abundance ratios: trend toward pure r-process

Not considered here: carbon-rich stars with/without s-process overabundances

Usual comparison: [Ba/Eu]Basolar-system > 90% s-processEusolar-system > 90% r-process

[Ba/Eu] ~ -0.9 ~ pure r-process value

at [Fe/H] ~ -3.0

Scatter is higher than desirable: blame

the Ba abundances?

The decline of Ba/Eu at lowest metallicities

The solar-system r-process-only ratio

An alternative: La/Eu

La also sensitive to s-process (70% s-process in solar system)Both elements have several useful lines at accessible ’sAtomic parameters of Eu, La lines very well knownCan determine La/Eu with higher accuracy than Ba/EuCan use same transitions over 3 dex metallicity range

Previous lanthanum work

Burris et al. (2000) ,magenta points; Johnson & Bolte (2001), black points

The La/Eu (e.g, the s-/r-) ratio is constant???

La II lines in the solar spectrum: synthetic spectra fits with new atomic data

hyperfinestructurepattern

Green lineis the solarobserved spectrum

Lawler et al. (2001)

La/Eu at low metallicity

The Ba/Eu (e.g, the s-/r-) ratio is NOT constant

Simmerer et al. (2002)

A better idea: employ abundances of more elements than just Ba and Eu

Johnson & Bolte (2001)

Four stars, withmean abundance levels scaled tothe solar-systemcurves by their average Ba, La,Ce, Sm, and Euabundances

Detailed elemental abundance distributions in a few very low metallicity stars

Stars with # of n-capture abundances > 15: CS 22892-052 (Sneden et al. 2000); HD 115444

(Westin et al. 2000); BD+17o3248 (Cowan et al. 2002); CS 31082-001 (Hill et al. 2002)

Rare earths: “perfect” agreement with solar-system r-process-only abundances

Heaviest stable elements: must use HST

Z < 56: need for another r-process?

A small spectral interval of a metal-poor but n-capture-rich star

Sneden et al. (2000)

First example: BD+17o3248

Most “metal-rich” of n-capture-enhanced stars:

[Fe/H] = -2.1

A warmer giant by about 500K than other examples

Extensive high res, high S/N HST data in hand

First metal-poor star with gold detection

Takes advantage of large sets of new atomic data La II (Lawler et al. 2001); Ce II (Palmeri et al. 2000);

Pr II (Ivarsson et al. 2001); Tb II (Lawler et al. 2001);

Eu II (Lawler et al. 2002)

Detection of n-capture elements in HST STIS spectra

HD 122563 is n-capture-poor; BD+17o3248 is n-capture-rich

Cowan et al. (2002)

Discovery of gold in a metal-poor star

Cowan et al. (2002)

n-capture abundances in BD+17o3248: 1st solar-system comparison

Scaled solar-system r-process curve: Burris et al. (2000)

Cowan et al. (2002)

The BD+17o3248 abundances are not compatible with s-process synthesis

Scaled solar-system s-process curve: Burris et al. (2000)

Cowan et al. (2002)

Second example: CS 22892-052

First metal-poor star discovered with extreme r-process:

[Fe/H] = -3.1 [Eu/Fe] = +1.6

One puzzle: also carbon-rich: [C/Fe] = +1.0

Good high res, high S/N ground-based spectra and lower quality HST data in hand

Even more exploration of atomic data (Mo, Yb, Lu, Ga, Ge, Sn, etc.)

Abundances or significant upper limits for 57 elements

Abundance Summary

Colors identify different element groups

Sneden et al. (2002), in preparation

Li and Be values are w.r.t. to unevolved stars of similar metallicity

Terbium in the Sun and CS 22892-052

0.80.80.80.80.80.80.80.8

0.8

0.9

1.0

1.1

Rel

ativ

e F

lux

Sun

This is the cleanest Tb II feature in the solar spectrum

n-capture-rich metal-poor stars are good “laboratories” for these lines

CS22892-052

Summary of the latest n-capture abundances for CS 22892-052

Sneden et al. (2003), in preparation

Z56 stable n-capture elements: excellent match to solar r-process

Sneden et al. (2003), in preparation

Z<56 n-capture elements: some deviations, some questions

The upper limits for Sn and especially for Ga, Ge are significant

Ga and Ge share the metal poverty of Fe-peak and lighter elements

Sneden et al. (2003), in preparation

Comparison with CS 22892-052 abundances

Note differenceof HD 122563: real or needing better data?

Perfect agreement with CS22892-052 would be a horizontal line

Some attempts to get isotopic abundances

Need large hyperfine and/or isotopic splitting

Rare-earth lines provide best opportunity

Some elements have only one stable isotope

Barium and now europium have been studied in metal-poor stars

See Ivans et al. poster at this meeting

An example of Eu II syntheses: the 4205.05A line

The Eu abundance is altered by 0.2 dex for each synthesis

Eu isotopic fractions are very similar to solar-system values

%(151Eu):0

355065

100

%(153Eu) = 100 - %(151Eu)

Solar system:%(151Eu) = 47.8%(153Eu) = 52.2

Sneden et al. (2002)

Barium isotopic mixes

134 135

s & rs only

137136 138

s only s & r s & r

135

25.7%0.0%

137134 138

0.0% 20.4% 53.9%

134 135

yesno

137136 138

no yes no

134 135

6.6%2.4%

137136 138

8.0% 11.2% 71.8%

136

synthesiscause

solar systemabundances

r-process abundances

hyperfinesplitting?

odd isotopes18%

odd isotopes 46%

odd isotopes are only11% of solar system s-process material

Barium Isotopic Abundances in HD 140283

Lambert & Allende Prieto (2002)

odd isotopes:10%

31%

52%

Solar system:total = 18%

r-only = 46%s-only = 11%

31% is best fit

Radioactive cosmochronometry for metal-poor stars

Galactic chemical evolution effects do not matter for radioactive elements Th and U “frozen” into metal-poor stars born near the start of the Galaxy.

?Daughter product Pb is also a direct n-capture synthesis product

Rolfs & Rodney (1988)

Best Th II and U II lines

Cowan et al. (2002) Cayrel et al. (2001)

BD +17o3248 CS 31082-001

Age computations for halo stars

1/2(Th) = 14.0 Gyr; 1/2(U) = 4.5 GyrSo for thorium:

NTh,now/NTh,start = exp(-t/mean)= exp(-t/20.3Gyr)Cannot know NTh,start assume NTh,start/NEu and compare that to N Th,observed/NEu IF solar-system r-process abundances can be assumed to extend to U, then can use [Thobserved/Eu ] as a measure of Th decay<[Thobserved/Eu ]> = -0.58 +/- 0.02 ( = 0.07, # = 10)

<t> = 13.6 +/-1.0 Gyr ( ~ 3.6 Gyr)But in CS 31082-001 the [Th/Eu] ratio is much larger:

[Th/U] t = 12.5 Gyr [Th/Eu] t = 4 to 5 Gyr

Thorium-to-europium ratios in some halo stars

Open circles:new data

Filled squares:Johnson &Bolte (2001)

The curious chemical composition of CS 29497-030

M68

[M68 diagram from Walker 1994]

It is like a “blue straggler”

It is a binary (companion undetected)

Preston & Sneden ( 2000)

CS 22947-030 is another example of lead-enriched metal-poor stars

These are s-process enrichments!

Log (Pb)solar system = 1.9

All data for CS 29497-030 point to mass transfer from former AGB companion

Summary, future work

Large star-to-star scatter in n-capture levels below [Fe/H] ~ -2: established but not well interpretedSwitch from r,s-process contributions to r-only abundances is seen in many low metallicity starsTh, U radioactive element chronometry is in its nfancy, but is a promising techniqueExtreme s-process stars may be understood?

Do [Th/Eu] ratios always yield “same” ages?Are there more U detections be had?Can the abundances of Z<56 n-capture elements be understood?

Total r- and s-process synthesis paths

The r-process alone makes radioactive chronometer elements Thand U

Bi is the end of the s-process

Rolfs & Rodney (1988)

What are s-/r- trends in the Galactic disk?

Woolf et al. (1995) derived [Eu/Fe] in disk dwarf stars with [Fe/H] > -1

Woolf spectra also contain 4123Å La II line

One La II and one Eu II line used to derive La/Eu for “disk” metallicity stars

Complements Mashonkina & Gehren study of Ba/Eu

Europium in Galactic disk stars

Woolf et al. 1995

Results confirmed by Koch & Evardsson (2002)

La/Eu at high metallicity

Does La/Eu have a break at [Fe/H] -0.4 ?

Simmerer et al. (2002)

La/Eu and space velocity

The s-/r- process abundance ratio correlates with space velocity as much as (more than?) [Fe/H]

Simmerer et al. (2002)

s.s. r-process

s.s. total