Transcript Slides - Agenda INFN

Nucleosynthesis & Cosmology
Infrared light from first stars:
Spitzer image ( z=11)
You are back to about 400 million years
after Big Bang
Mounib El Eid
American University of Beirut
Department of Physics
Santa Tecla: Sept. 18, 2011
program
1. General comments
2. Heavy elements in oldest Stars and Early Universe
3. Cosmological Motivations and evolutionary
scenarios. Just drawing a connection
1. General comments
 First stars formed after Big Bang were quite different from those formed later,
simply because they were composes of H, He and small fraction of Li
 Classification of the first stars is not easy.
Bromm etal, Nature 459, 49 (2009) made the following classifications:
pop III.1
stars formed under initial cosmological parameter
pop III.2
stars formed from photo-ionized gas by earlier generation
pop II stars:
EMP (extreme metal-poor)
10-4 < [Fe/H]< 10-3
UMP (ultra metal-poor)
10-4 < [Fe/H]< 10-4
HMP (hyper metal-poor)
10-6 < [Fe/H]< 10-6
It seems that such a classification is related to nature of the first galaxies, where a
galaxy is a system of stars, gas and a dark matter halo
Within this picture: pop III.1 stars are assumed to be formed in isolation in minihalos
(not in galaxies)
 Many arguments exist (Bromm et al 2009 and references therein) that the first
stars were massive and even very massive, mainly because of their formation
in a medium devoted of heavy elements such that cooling and effective
fragmentation were inhibited
If so, the first stars represent extinct generations, because they ended in supernovae
explosion, eventually not only as core collapse (later more)
 Observing the chemical abundances in the oldest surviving stars
is a way to learn about the nature of the first stars.
But chemical abundances are related to nucleosynthesis processes occurring
in stars, that is linked to stellar evolution.
The heavy elements (neutron-capture elements beyond iron) in the universe
were formed in very late evolutionary phases of stars by the
s-process, like Barium (mainly in AGB stars)
r-process, like platinum or gold (mainly supernovae)
with a time scale of the order of millions to billions of years
Early star formation not well understood .The iron abundance may tell about the
history of star formation.
Most of the iron comes from Type Ia SNe (exploding white dwarf in binary systems).
In other words, it comes from long-lived low-mass stars Thus the stars formed had
little effect on the history of metallicity. In our Galaxy these metal-poor stars are
found in the halo and the metal-rich in the galactic disk.

The metal-poor stars in the halo serve as “laboratory” for the study of the
nucleosynthesis of neutron-capture elements. Their chemical compositions are
linked to the types of synthesis processes that occurred early phase of the
Galaxy.
The presence of heavy elements in these stars indicate preceding extinct
generation of massive stars (second generation?) which have synthesized all
heavy elements
But
The r-process crisis
I have also a crises after so many years of
struggle with the r-process. But I am so
strong to understand the weak r-process
2. Heavy elements in oldest Stars and Early Universe
First example : the r-process rich star CS22892-52
show some interesting results Sneden eta al: APJ, 591, 936 (2003) r  process rich :
[ Eu / Fe]  1.6
A similar case is CS31082-001 (Hill et al. 2002)
EMP: [Fe/H]=-3.1
remarkable
n-capture elements
Main
with Z56 in this
r=process meta-poor star match
closely solar system rprocess pattern
Weak
r-process
Also remarkable
Scaled solar rprocess distribution
does not extend to the
lighter n-capture
elements below Z=56
log  ( A)  log( N A / N H )  12.0
For example: Silver
(Ag, Z=47) is deficient
Conclusion from previous Figure:
It seems that the r-process could be divided like the s-process into 2 components :
weak r-process ( so far it is called LEPP=Light Element Primary Process)
main r-process (classical r-process)
Heavy elements have been also observed in extremely metal-poor stars
with [Fe/H]=--5:
HE 0107-5420 and HE1327-2326
Both are rich in CNO but very poor in n-capture elements. This is different
from the previous cases (CS 22892-052 or CS31082-001)
Learn effect
Rapid change in nucleosynthesis in the early phase of the Galaxy.
It seems: first stars were massive able to produce CNO elements
but not the heavy n-capture elements
Heavy n-capture elements
10 r-process rich stars
CS22892-052
HD115444
BD+173248
CS310802-001
[Fe/H]=-3.1
[Fe/H]=-3.1
HD221170
HE13250901
CS22953-003
He2327-5642
Cs29491-069
HE1219-0312
Figure indicate:
All heavy n-capture
elements (Ba and above)
consistent with solar
system r-process
distribution (Sneden et al
2009)
Light n-capture Elements
The lighter n-capture elements (Z<56) seem to fall below the solar system curve
Observation of four metal-poor r-enriched stars (Grawford et al 1998)
indicates: Ag (Z=47) produced in proportion of the heavier elements in
stars with -2.2 <[Fe/H] <-1.2.
Wasserburg et al (1996) proposed multiple processes of the heavy elements.
Travaglio et al (2004) suggested: not all of the Sr-Zr solar system abundances can
be explained by the classical r-process, or s-process (main and weak).
The term LEPP has been invented to give such a process a name.(not really the
best description).
I would say: ENCP (early n-capture process)
Montes et al (2007) extended the range of the LEPP up to Ba , Z=38- 56.
Their suggestion: LEPP may have been important in synthesizing the abundances
in the r-process poor star HD122563 (see next slide)
r-process rich
32 elements
[ Eu / Fe]  0.90
Solar system
normalized to Eu
Farouqi et al (2009)
HEW
Relatively flat
abundance distribution
consistent with scaled
solar system r-process
for the heavies
r-process poor
[Eu/Fe]=-0.5
Solar system
normalized to Sr
Solar system
normalized to Eu
Here: sharp decrease with increasing atomic number ;
looks like incomplete r-process (or weak r-process)
Another comparison
r-process rich: [Eu/Fe]=+1.6
r-process poor: [Eu/Fe]=-0.5
Comparison of the abundances in the stars
BD+173248 and HD122563
shows that the third peak of the r-process (Os, Ir, Pt: Z=76, 77, 78 ) is
not formed
High-entropy wind model
in SNII: Farouqi et al (2009)
Deviation starting Z=45:
Ag (Z=46), Cd (Z=48) and
also Pd (Z=46)
insight
Such comparison
argue for a
combination of
processes: LEPP, -p
process, ???
r-process throughout the Galaxy
16 stars
The difference between BD+173248 and HD122563 discussed above is found
to be a general behavior as shown in the following Figure:
r-process rich
flat
Nothing
here
HD122563
References to this Figure: Cowan eta l (2011) preprint
r-process poor
No heavy ncapture elements
The abundances in HD122563 suggest an incomplete r-process, or let
us say an n-capture process with a neutron density between 10201024 cm-3, since the synthesis of the heavy n-capture elements needs
1024 – 1028 cm-3
\in the following some results by Kratz et al (2007):
Log nn=20-22
Log nn=20-24
Log nn=20-26
Log nn=20-28
Conclusions
 clear presence of n-capture element in atmospheres of metal-poor stars
and globular cluster stars
The comparison between r-process rich ([Eu/Fe]> 1.0) and r-process poor
([Eu/Fe] < 1 indicates :
abundances of the heavy elements (Ba and above, Z=56)
consistent with solar system r-process distribution. This seems
to be the main r-process.
The distribution of the lighter (Z<56) n-capture elements is not
conform to solar pattern. New detection of Pd, Ag, Cd (Z=46,
47,48) suggest a weak r-process not yet identied:
LEEP
-p process in core collapse SNe
High Entropy Wind in core collapse SNe
Exotic mixing in late phases of massive EMP stars
Do different mass region (Ge, Sr-Zr Pd-Ag-Cd) require different processes?
3. Cosmological Motivation & Evolutionary Scenarios
Timeline of light in the universe
The oldest light we can see today is the
cosmic background radiation . It came
from the time 380,000 yrs after Big
Bang when the universe became
transparent. This light had a redshift of
z=1100 and appears in the microwave
WMAP data indicate:
Some 400 million years later at
z=11 the first stars appeared .
They reionized the universe,
and their light is now shifted
to infrared wavelength .
Galaxies formed
more recently
and can be seen
at visible
wavelength.
Dark ages: era from recombination (at 380,00 yr) to the first stars at 400 million yr. This
dark ages ended when the universe was filled for the first time with light from stars
Evolutionary Scenario: overview
1 7 ?
AGB
8 140
Iron core collapse
Shock/neutrino
driven
Planetary
Nebula
White Dwarf
Neutron stars
Black holes
 140 260
 260
Massive C/O
Cores: 64- 133 MSun
Explosive oxygen
burning
Pair Creation
Supernovae
Black holes
The Central Evolution of Stars
ElectronPositron –
Creation.
Pair Creation
supernovae
25
AGB stars
WD
Non-degenerate
Iron
Disintegration
Core collapse
supernova
Rapid
Electron
Capture
Core
collapse
Supernva
What is a Pair Instability Supernova (PCSNe)?
PCSNe represents the final state evolution of a very massive stars which develops a very
massive carbon-oxygen cores
Such massive core composed mainly of oxygen and are largely supported by the radiation
pressure.
As seen in the Figures below, the central evolution of the core proceeds toward higher
temperature and relatively low density such that electron-positron pairs are created in
equilibrium by the radiation field according to
2  e   e 
Although the mean energy photons is about kT, there are enough photons in the tail of the
Planck’s distribution that can create these pairs even at 109 K
Example follows
Example: several cores of different masses are shown. The core of mass 112 Msun
corresponding to an initial mass of  200 Msun undergoes collapse and explosion
induced by explosive oxygen burning leading to a very brilliant supernova.
Central
evolution
About 25 Msun of
oxygen needed
to explode such
a core with
explosion energy
> 1052 erg
Temperature-density profile
Ober, El Eid & Fricke: A&A , 119, 61 (1983)
T- profile for a 112 Msun star at time
reversal of the collapse. Note that
significant part of the core is inside the
instability region.
More details on pair creation:
What really happens is:
 Radiation pressure and entropy decreases
srad

4aT 3

,
3k B b
Prad
a 4
 T
3
Prad  ( srad  ) 4 / 3
 Electron-positron pairs are created and their
entropies increase along that of nuclei
 When this happens, the adiabatic index
s=10
drops below 4/3 and not only at center as we
have see on the previous slide
Entropy per unit volume divided
by baryon density
Why does <4/3 has a finite range?
The decrease of  below 4/3 is a consequence of the new particles which
do not immediately add their contribution to the total pressure..
At high densities: >4/3, because the electron becomes more degenerate
At high temperatures: >4/3 because the particles become relativistic such that
the energy gap for pair creation is no more important
Here is a case
This figure tells that the fate of a
very massive stars depends on its
mass and initial metallicity
The lines are schematic and not
well determined
It may be interesting that the
PCSN may be found in the
local universe and at
metallicity up to ZSolar/3.
While the common view is that
these events may be
associated to Pop Iii stars
Langer, N. Nature 462, 579 (2009)
The PCSNe are usually associated with early stars, or Pop III stars
Is there any evidence of this unusual supernova type?
The observations of the SN 2007bi ( Gal-Yam et al, Nature ,462, 624, 2009) in a
dwarf galaxy argue for this with an estimated core mass is about 100 MSun
Another object is SN2006gy (Smith et al , Apj 666, 1116 (2007))
Light curves of super luminous SNe
Gal-Yam et al (2009)
R-band light curve of SN2007bi. With
a peak magnitude of -21.3 mag
If this light curve is radioactively driven,
> 3 Msun of 56Ni are needed. The slow
rise time (70 days) and photospheric
velocity of 12,000 km/s indicate an
exploding very massive object of about
100 Msun and very high explosion
energy > 1052 erg
Implication of the discovery of SN2007 bi
 The estimated high core mass is in conflict with the commonly used mass loss
rates as a function of metallicity
 Regardless the correct description of mass loss, the data indicate that an
extremely massive stars (>150 Msun ) are formed in the local universe in a dwarf
galaxy with a metallicity
12+log[O/H]=8.25 (less than 1/10 of the Sun’s metalicity
Can the dwarf galaxies serve as fossil laboratories for studying the earl universe?
 Future missions like the
NASA’s James Webb Space telescope
will help to estimate the contribution of these events to the chemical evolution in
the early universe.
Nucleosynthesis in PCSNe
Updated calculation by Heger & Woosley (2002) yield:
 No heavier elements than Zinc, no r-process, no s-process
 Mainly products of explosive oxygen burning. Even nuclear charges (Si, S, Ar, Ca,, ...)
in almost solar distribution
 Element of odd nuclear charges (Na, Al, P, V, Mn,...) are deficient. The explanation of
this is because the massive C/O massive evolves almost directly to oxygen burning
without creating a neutron excess
Production factors of C/O cores of
masses 64 to 130 Msun which
undergo PCSNe, with different
assumption of the exponent of a
Salpeter-like IMF
Heger & Woosley (2002)
Interesting evolutionary scenario of extremely meta-poor massive stars
Motivation: for example Barium Isotopes in the meta-poor subgiant HD
140283 : ( [Fe/H]=-2.6, [Ba/Eu]=-0.66 and [Eu/H]< -2.8 (r-poor)
Author
f
odd
Magain (1995)
0.08  0.06
Ghallagher et al. (2010)
0.02  0.06
Lambert et al (2002)
0.30  0.21
purleyS-process signature
Shielding of the Barium isotopes
Different mixture of odd an even Ba (Z=56) isotopes are produced by the
r-process and s-process
134Ba and 136 Ba are produced by the s-process only, since they are
shielded by the Xenon (Z=54) isotopes made in the r-process.
A challenging question (K.L. Kratz, private communication) :
How can one get for a star like HD 140283:
fodd = 0.1-0.2 (s-process) and [Ba/Eu}=-0.66 r-process
Exotic n-capture scenario
(1) Zero-age mains sequence shifted to higher effective temperature as Z decreases.
This is a consequence of reduced metallicity.
Recall that the energy generation via the CN cycle:
( )CN  8 10  xH xCN T6
27
2 / 3 152.3 / T61 / 3
e
( ) CN  ( ) CN T 17
At T=25x106 K
( ) pp   ) CN T 4
At T=15x106 K
Lacking of heavy elements, the star has to contract and heats up to burn hydrogen at
high temperature.
(2) As consequence of the compactness of the star , it cannot evolve to become
red giants. They remain confined to the blue part of the HR diagram, when
Z<10-3 . As seen on next page, the hydrogen-burning shell remains
convective all the time.
Z=2x10-2 : Solar-like
Convective
Envelope
El Eid, The , Meyer
Space Sci. Rev. , 147, 2009
Many references there
blue
red
Z=10-3
No Convective Envelope
34
Z=10-3
Possible mixing of protons into the
helium shell. It works only at low metallicity.
Here: [Fe/H]=-4.5
35
Let’s believe it……… then
What is the advantage of this game?
36
If this would be true?, we make
Primary Sr for example
Result after one time step of proton mixing into the helium convective shell. Strong
enhancement near Z=38 to more than 50
37
Arnett, D. 1996: Supernovae and Nucleosynthesis, Princeton Univ. press, p. 244
Does this game work?
We need his optimism like this star
From FC Barcelona (and FC Libya) to get an answer
JJ with American muscles
Final Words
 Evolution of early stars linked to nucleosynthesis of heavy elements turns out to be
a link to Near-field cosmology (understanding galaxy formation); It is a challenging
topic and a revival of the importance of stellar evolution as a fundamental
cornerstone of modern Astrophysics
 The evolution of the neutron-capture elements traces back the chemical
evolution of the galaxy and is bring us back to the dark ages where in order to
become more enlightened and overcome our ignorance
Thank you for your attention
Imbriani et al (2001), ApJ 558, 903
Rate of
12
C    16O  
CF85
X 12=0.18
CF85 > CF88
X12 lower
25 Msun star
No convective
carbon-burning
core
No convective core
Remaining
car bon
mass
fraction
CF88
X12=0.42
But here
La: mainly s-process
Eu: mainly r-process
Elemental ratio La/Eu for large number of stars
s-process rich
s  process / r  process
Total solar system
 General increase in the
ratio La/Eu ratio as the sprocess contribution to La
production rises with
metallicity. That is after the
low mass stars had time to
evolve
 Only the most metalpoor stars seem to have
La/Eu ratio consistent with
r-process-only ratio
r-process
enhanced
 Some s-processing
below [Fe/H]=-2.0
r-process only
Filled circles for halo stars: Simmerer et al : APJ, 617, 1091 (2004)
Filled diamonds (disk starts): Woolf et al , APJ, 453, 660 (1995)
References
Cowan, J.J xiv:1106.11091/1
Cowan, J.J, Sneden, C. heavy element synthesis in the old and the early
universe. Nature, 440, 1151 (2006)
Montes et al, APJ, 671, 1685 (2007)
Heger, A, Woosle, S.E. 2002, ApJ 567, 532
We need his
optimism like
this start
From FC
Barcelona
To get the
answer