Transcript Carbon Enhanced Stars in the Sloan Digital Sky Survey ( SDSS )

Carbon
Enhanced
Stars
in
the
Sloan Digital Sky Survey ( SDSS )
T. Sivarani, Young Sun Lee, B. Marsteller & T. C. Beers
Michigan State University
&
Joint Institute for Nuclear Astrophysics
Carbon Stars
Carbon stars are objects with prominent C2 molecular bands
They are known since 1884 (Duner 1884)
Carbon star:
Star with C/O > 1
Carbon stars
The observed Galactic and extra galactic carbon stars are
==> low mass stars on the AGB phase which has gone
through the 3rd dredge up of He burned products.
==> CH stars, Ba Stars from binary AGB companion
==> Dwarf carbon stars – possibly AGB mass transfer
Large frequency of carbon enhanced stars
at low metallicities
An increasing interest in this class of objects has arisen from
I
HK survey of metal-deficient stars conducted by
Beers, Preston, & Shectman.
Hamburg/ESO survey by Christlieb and Co.
revealed for the first time an unexpectedly large number
of very metal-poor stars with anomalously strong CH and CN bands
(~25% in the metallicity range [Fe/H] < -2.5 compared to few percent
among stars of higher metal abundances).
Implications
1. Is the intial mass function of the early galaxy was very different from the
present day. Lucatello et al. (2004)
2. Pre Galactic metal enrichment.
3. The evolution at low metallcities are very different.
4. These stars are been accreted from metal poor local satellite galaxies
Carbon Enhanced Metal Poor (CEMP) stars
Carbon rich [C/Fe] > 1.0 (Though the classical carbon star at solar metallicity is
about 2 times solar).
All of them are nitrogen rich (not all the carbon stars are nitrogen rich)
Carbon stars in our Galaxy are different from the C-stars in LMC and other
local groups. Does metallicity play an important role ?.
Chemical composition of CEMP stars
s-process rich
r-process rich
with no n-capture enrichment.
Alpha enhanced
mild CEMP stars.
N-capture in CEMP stars
S-process rich AGB mass transfer
One with normal Li : LP706-7
CS22892-052
The only r-process rich carbon star
r-only pattern (Sneden et al. 2003)
Figure from Ryan et al. (2002)
No n-capture elements
HE 0107-5240 Christlieb et al 2002
CS 22957-027 Aoki et al 2002
HE 0007-1832 Cohen et al 2004
12C/13C ~ 6 – 9 (CNO equilibrium)
Also have slightly higher nitrogen
abundance compared to s-process
rich stars.
n-Capture Processes in CEMP stars
r-process rich carbon star
The only known CS 22892-052
r-only pattern (Sneden et al. 2003)
s and r-process
15 objects are known.
High [Eu/Ba] compared to
pure s-process
No n-capture elements
HE0107-5240 Christlieb et al 2002
CS 22957-027 Aoki et al 2002
HE 0007-1832 Cohen et al 2004
12C/13C ~ 6 – 9 (CNO equilibrium)
s-only stars?
[C/Fe]=0.88, [N/Fe]=1.01, [O/Fe]=0.72
[Mg,Si/Fe] ~ 0.3
[Ba/Fe] = 0.99
[Eu/Fe] = 1.64
[Tb,Dy,Er…./Fe] ~ 1.6
[Ba/Fe] ~ 2.0
[Eu/Fe] ~ 1.5
[Ba/Eu] ~ -0.9 ~ pure r-process value
at [Fe/H] ~ -3.0
To have a better understanding
relative frequency among the various types and origin.
 Increasing the no. of similar objects.
 High resolution follow up.
 radial velocity monitoring.
 SDSS DR4 100,000 + SEGUE +SDSS II .....
SDSS DR4
SDSS DR4 has about 102,714 stars at a resolution R= 1800
from 3800Å to 9000Å
Already about 500 Faint high latitute carbon stars have been
identified by Morgan et al. 2002 and Downes et al. 2004.
from SDSS DR1
The propermotions indicate that most of them are near
by dwarfs.
==> Young population which had gone through binary mass
transfer from an AGB star
==> the AGB star will now be seen as a white dwarf.
There have been efforts in search for white dwarfs around
these stars.
Analysis
Most of the key problems in astrophysics rely on the determination
of accurate stellar fundamental parameters
1. Grid of synthetic spectra
wavelength range 3000-10000A and 0.01Ang spacing
==> smoothed to SDSS resolution 2.5
2. Grid of synthetic colors and lineindices
3. Empirical grid.
High resolution spectra with well estimated parameters
smoothed to SDSS resolution
==> ELODIE, VLT UVES archive, Cflib, S4N
Automated methods to estimate stellar parameters
Teff, logg, [Fe/H], [C/Fe] …
Validation of the estimation with real spectra.
Clusters, Standard stars .
Synthetic grid
Teff = 3500K – 10000K , logg = 0.00-5.0
[Fe/H] = -5.00 – 0.00
Carbon Enhanced grid.
[C/Fe]=0.5,1.0,1.5,2.0,2.5,3.0
At low temperature the
stellar atmospheric
structure itself changes
due to high carbon
abundance.
==> Solar scaled models
are not appropriate.
Masseron et al. 2005
Synthetic spectra
Results from Artificial neural networks
[Fe/H]
logg
Teff
---------------------------------------- ----------------------------------------- ------------------------------------------------Com_cls
True Diff
IntErr
Com_cls
True
Diff
IntErr
Com_cls
True
Diff
IntErr
---------------------------------------- ----------------------------------------- -------------------------------------------------0.0524 -0.0000 -0.0524 0.0976 2.9984 3.0000 -0.0016 0.0054 4997.8731 5000.0000 -2.1269 11.4377
0.2335 0.2500 -0.0165 0.1056 2.9678 3.0000 -0.0322 0.0123 5000.5895 5000.0000 0.5895 8.7776
-0.3378 -0.2500 -0.0878 0.0758 3.0561 3.0000 0.0561 0.0203 4998.8447 5000.0000 -1.1553 10.4254
0.5235 0.5000 0.0235 0.0539 3.9910 4.0000 -0.0090 0.0039 5239.0910 5250.0000 -10.9090 5.4740
-0.5048 -0.5000 -0.0048 0.0573 4.0301 4.0000 0.0301 0.0229 5242.5087 5250.0000 -7.4913 3.2398
0.7738 0.7500 0.0238 0.0534 3.9812 4.0000 -0.0188 0.0116 5239.0285 5250.0000 -10.9715 6.4912
0.9883 1.0000 -0.0117 0.0594 4.0040 4.0000 0.0040 0.0236 5238.7821 5250.0000 -11.2179 4.3407
-0.9611 -1.0000 0.0389 0.0325 3.9825 4.0000 -0.0175 0.0449 5241.2428 5250.0000 -8.7572 4.3664
0.2654 0.2500 0.0154 0.0824 3.9911 4.0000 -0.0089 0.0086 5501.4074 5500.0000 1.4074 1.8401
-0.2632 -0.2500 -0.0132 0.0776 4.0072 4.0000 0.0072 0.0115 5504.5832 5500.0000 4.5832 3.8452
0.5351 0.5000 0.0351 0.0816 3.9833 4.0000 -0.0167 0.0189 5500.9626 5500.0000 0.9626 6.4307
-0.5114 -0.5000 -0.0114 0.0690 4.0111 4.0000 0.0111 0.0135 5504.9800 5500.0000 4.9800 5.3205
-1.4869 -1.5000 0.0131 0.0446 4.5127 4.5000 0.0127 0.0193 5985.8269 6000.0000 -14.1731 7.4680
-1.9174 -2.0000 0.0826 0.0849 4.4932 4.5000 -0.0068 0.0196 5987.6177 6000.0000 -12.3823 2.6852
-3.0891 -3.0000 -0.0891 0.0777 4.4984 4.5000 -0.0016 0.0275 5987.3293 6000.0000 -12.6707 4.6541
Now we are ready for the science
From the SDSS spectra
1. [C/Fe] versus [Fe/H]
2. [alpha/Fe] versus [Fe/H]
3. s-process and C abundance
4. relative frequency of s-process rich and no
n- enrichment
How they are distributed in metallicity
High resolution follow up
C,N and isotopic abundance ==> hydrogen and helium burning
s-process abundance (light and heavy) ==> metal poor AGB evolution
and nucleosynthesis.
r-process abundance ==> SN neucleosynthesis.
S-process among the C-enhanced stars
In the figure the straight lines
indicate the equivalent width
for solar composition of Ba
in
The metallicity range.
The symbols corresponds to
Ba equivalent width for the
SDSS-DR3 C-enhanced stars
Conclusions
●
●
●
●
From SDSS data we can get [C/Fe],
[alpha/Fe], [Sr/Fe] and [Ba/Fe] can be
obtained.
Relative frequency of various types of CEMP
stars can be found from the Sr and Ba
abundances.
Several hundreds will be suitable for follow
up high resolution studies.
Radial velocity monitoring.