Early Galactic Nucleosynthesis: The Impact of HST/STIS Spectra of Metal-Poor Stars Chris Sneden...

Post on 16-Jan-2016

222 views 0 download

Transcript of Early Galactic Nucleosynthesis: The Impact of HST/STIS Spectra of Metal-Poor Stars Chris Sneden...

Early Galactic Nucleosynthesis: The Impact ofHST/STIS Spectra of Metal-Poor Stars

Chris SnedenUniversity of Texas

A Very Collaborative Effort

Definitions:

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

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

log N(A) = log10(NA/NSi) + 6.0 (meteoritic)

today two topics:- the neutron-capture elements- return to the Fe-peak

the common thread: nucleosynthesis at the end of stellar lives

HST/STIS spectra: relevant to entire nuclide range

Heaviest stable elementsa.k.a 3rd neutron-capture peak

Understudied lighter neutron-capture elements

Fe-group elements

Light elements (Be, B)

http://atom.kaeri.re.kr/

The ultimate question? How did our Galaxy produce the solar chemical composition?

SCG08 = Sneden, Cowan, & Gallino 2008, ARA&A, 46, 241

Most isotopes of elements with Z>30 are formed by: AZ + n A+1ZFollowed by, for unstable nuclei: A+1Z A+1(Z+1) + -

Rb Sr Y Zr Nb Mo Tc Ru Rh Pd Ag Cd In Sn Sb Te I

ScCaK

Xe

Ti V Cr Mn Fe Co Ni Cu Zn Ga Ge As Se Br Kr

H

BeLi

MgNa

Hf Ta W Re Os Ir Pt Au Hg Tl Pb Bi Po At Rn

RaFr

BaCs

Rf Db Sg Bh Hs Mt Uun Uuu Uub

CB O F Ne

ArAl Si P S Cl

N

He

La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Ac Th Pa U Np Pu Am Cm Bk Cf Es Fm Md No Lr

First discussion: n(eutron)-capture elements

• s-process: β-decays occur between successive n-captures• r-process: rapid, short-lived neutron blast temporarily • overwhelms β-decay rates• r- or s-process element: origin in solar-system • dominated by one or the other process

Rolfs & Rodney (1988)

Buildup of n-capture elements

A detailed look at the r- and s-process paths

SCG08

“s-process” element

“r-process” element

n-capture spectra of metal-poor stars

HD 122563:[Fe/H] = -2.7Teff ~ 4750LOW n-capture

CS 22892-052:[Fe/H] = -3.1Teff ~ 4750HIGH n-capture

“r-process-rich” metal-poor stars

first example, HD 115444, was reported by Griffin et al. 1982 SCG08

An important abundance ratio:log (La/Eu) = +0.6 (solar total) = +0.2 (solar r-only) = +1.5 (solar s-only)

HfBa La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Wisconsin lab studies: log gf and hyperfine/isotopic structureLa

wle

r et a

l. 20

00a

Law

ler e

t al.

2001

Law

ler e

t al.

2000

b

Law

ler e

t al.

2004

Law

ler e

t al.

2006

Law

ler e

t al.

2007

Law

ler e

t al.

2008

Den

Har

tog

et a

l. 20

03

Den

Har

tog

et a

l. 20

03

Law

ler e

t al.

2009

Sned

en e

t al.

2009

Sned

en e

t al.

2009

Sned

en e

t al.

2009

Sned

en e

t al.

2009

Sned

en e

t al.

2009

14 36 14 20 2 13 3 8 3 1 1 445 5 46Sun: # transitions used in analysis

1532154 7 29 13 21 6 1 837 55 9 32CS 22892-052 : # transitions used in analysis

why are rare earths well understood?

(uns

tabl

e el

emen

t)

(wel

l stu

died

in li

tera

ture

)

n-capture compositions of well-studied r-rich stars: Così fan tutte??

SCG08

lines are solar-system abundances

differences fromsolar-system

means of the differences

Thorium/Uranium detections promise alternate Galactic ages

Frebel et al. 2007

leading to possible radioactive decay ages

U/Th ratio should be best age

indicator, if both elements can be detected reliably

Persistent question: why is Pb usually low?

Frebel et al. 2007

good abundances really confront r-process predictions

Ivans et al. 2006

But! crucial element groups for further progress

Ivans et al. 2006

Johnson & Bolte 2002

Beyond simplest results: de-coupling of the heavy/light r-

process elements

See also Aoki et al. 2005, 2007

ZY = 39ZBa = 56

r-process abundance distribution variations: "normal"

Roederer et al 2010

But various densities must contribute to r-process

Kratz et al. 2007

increasing density components

decreasing density components

Niobium (Nb, Z=41): surrogate for

remaining n-capture elements (and their

“issues”)

these are the best transitions in the most favorable detection cases

Nilsson et al. 2010

All reasonably strong lines are in the vacuum UV;

this is why STIS is good here

Niobium: here is why there are difficulties

Nilsson et al. 2010

UV spectra of 3 metal-poor giants

Roederer et al 2010

HD 122563:[Fe/H] = -2.7Teff ~ 4750LOW n-capture

HD 115444:[Fe/H] = -2.9Teff ~ 4650HIGH n-capture

BD+17 3248[Fe/H] = -2.2Teff ~ 5200HIGH n-capture

Initial STIS exploration

HD 122563:[Fe/H] = -2.7Teff ~ 4750LOW n-capture

HD 115444:[Fe/H] = -2.9Teff ~ 4650HIGH n-capture

CS 22892-052[Fe/H] = -3.1Teff ~ 4750HIGH n-capture

Cowan et al. 2005

Heaviest stable elements scale

with Eu

Cowan et al. 2005

Again, Eu is the usual “pure” r-process element

But light n-capture elements do NOT

scale with Eu

Cowan et al. 2005

Upon further review of these same STIS data …

Roederer et al 2010

Points = data

Lines = syntheses:• best fit (color)• ± 0.3dex (dotted)• no element (solid)

Helps “complete” the n-capture distribution

Roederer et al 2010

Points = observations

Blue line = SN ν windFarouqi et al. 2009

Gray lines = scaled solar-system

Second discussion: back to the Fe-peak

McWilliam 1997

sharpening & extension by “First Stars” group

Cayrel et al. 2004

elements that can be directly

predicted in SNe models

Kobayashi et al. 2006 (following Timmes et al. 1995)

crucial distinction from n-capture elements: the Fe-peak element productions can actually be predicted theoretically

for some elements, all seems well

Kobayashi et al. 2006(solid line new results

but for others, discomfort

Kobayashi et al. 2006

neutral lines

ionized lines

same predicted Cr abundances

should be simple to do Fe-peak elements:their abundances >> n-capture ones

Fe-peak

light n-cap

heavy n-cap = rare-earth

3rd peak, Thα, etc.

r-process-rich stars give false impression of # of lines in metal-poor abundance

work!

For Fe-peak elements #/species <10 in most cases

The reality is grim for metal-poor turnoff stars

AND, can you believe ANY past analysis??

the outcome for Bergemann et al.?

Comparison of low metallicity spectra in the “visible” spectral region

a new attack: UV STIS spectra

The complex UV spectrum of HD 84937

dotted line, no Fe; solid line, best fit dashed lines: ±0.5 dex from best fit

red line, perfect agreement; other lines, deviations

Why is working in the UV a problem?Opacity competition: metal-poor red giant

Opacity competition: metal-poor turnoff star

A summary of our results on Fe

probable explanation of the abundance “dip”

Other Fe-peak elements: a start

Or, [Co/Fe] ~ +0.3 from both of these neutral and ionized lines

Bergemann 2009: [Co/Fe] = +0.18 (LTE) = +0.66 (NLTE) {probably ~+0.3 (LTE) on our [Fe/H] scale}

This rough cobalt estimate is not too bad …

the red line is the [Co/Fe] trend with [Fe/H] derived by the “first stars” group (Cayrel et al. 2004)

HD 84937

work on Cr I and Cr II is underway

Matt Alvarez, Chris Sneden (UT), Jennifer Sobeck (U Chicago), Betsy den Hartog, Jim Lawler (U Wisconsin)

Summary:

metal-poor stars hold keys to understanding Galactic nucleosynthesis

Atomic physics dictates “non-standard” wavelengths for some elements

STIS in the vacuum UV is the only choice for some poorly understood

element regimes

neutron-capture elements: observers are in the lead!

STIS is crucial for both lightest and heaviest stable elements

Fe-peak elements: theorists are in the lead!

No sane person should believe Fe-peak abundances right now

but STIS data hold promise for the near-future