Transcript AGB Stars

Introduction: AGB Asymptotic Giant Branch
Ryszard Szczerba
Centrum Astronomiczne im. M. Kopernika, Toruń
[email protected]
1
„Asymptotic Giant Branch”
Harm Habing, Hans Olofsson (Eds.)
A&A Library, 2004 Springer-Verlag
2
Introduction
3
NCAC
TORUN
Tubingen, March 2009
POLAND
4
NCAC
TORUN
Tubingen, March 2009
POLAND
5
NCAC
TORUN
Tubingen, March 2009
TORUŃ
6
NCAC
TORUN
Tubingen, March 2009
TORUŃ
7
NCAC
TORUN
Tubingen, March 2009
TORUŃ
8
NCAC
TORUN
N. COPERNICUS (1473-1543)
9
NCAC
TORUN
TORUŃ
10
NCAC
TORUN
N. Copernicus Astronomical Center,Toruń
11
NCAC
TORUN
Tubingen, March 2009
CAMK – opening ceremony: 24.05.1978 r.
12
What I will talk about?
• Stellar evolution of low- and intermediate-mass stars;
• mass loss and the end of evolution during AGB;
• molecules and dust formation (circumstellar envelopes);
• dynamics and instabilities in dusty winds;
• post-AGB stellar evolution;
• observations and theory related to this phase of stellar
evolution (results from IRAS, ISO, HST, SST, HSO)
• about future of the Sun.
13
AGB Stars: History
•At the beginning of the XX centuary dwarfs and giants were
discovered in the Henry Draper (HD) catlogue. (Hertzsprung
1911, Russell 1914=> H-R diagram).
• The reason what causes a star to be either a dwarf or a giant
was unknown until 1960’s.
• Many AGB stars are Long Period Variables (LPVs):
M-stars:
1596 Fabricius discoverd that one 3rd magnitude star
disappeared! (note that Tycho Brahe – discovered „Tycho’s
supernova” in 1572);
Fabricius -> Brahe -> published by Kepler.
In 1638 the star re-appeared (seen by Dutch astronomer
Holwarda). He established period of this phenomena for
about 1 year! (Stella Mira „the wonderful star” Hewelius)
14
AGB Stars: History
•Classification based on the appearance of the light curves:
Miras, Semiregulars, Irregulars does not allow to understand
physical reasons of the variablility.
•Glass & Lloyd Evans (1981) discovered a linear relation
between K-mag and log(P) in Mira variables.
C-stars:
 Kirchoff and Bunsen (1860) had published correct
interpretation of spectral lines;
 1868 - Father Secchi (Vatican Observatory) clasified
spectra of ~4000 stars. He recognized a small group of very
red stars with spectra „similar” to that of the ligth in
carbon arcs.
15
AGB Stars: History
• Why there are 2 very different classe of red stars (C- and Mtype: O-rich)? The question answered in 1934.
• Russell (1934) showed that high binding energy of CO
molecule (11.09 eV) leads to:
 M-type spectra when O > C
 C-type spectra when C > O
Stellar models (Main Sequence):
• Eddington (1926) „The internal constitution of the Stars” –
he stated that H->He is (probably) the source of the stellar
energy! but he didn’t know how the mechanism works.
• It was assumed that the atomic composition of the Sun was
the same as that of the Earth- ~TRUE! if one ignore H and He.
• Payne-Gaposchkin (1925) had found the large relative
abundances of H and He, but she rejected this result!
• Russel (1929) draw the correct conclussion about chemical
16
composition of the Sun.
AGB Stars: History
• Bethe (1939) shows that pp – reactions works in ~1 Mo stars
(T< 15 milion K), while in more massive stars the CNO-cycle
dominates.
17
AGB Stars: History
Stellar models (Red Giants):
• Progress possible because: development of observational
techniques (photometry) and development of „electronic
devices” – analytical solutions => the numerical ones.
 color-mag diagrams in globular clusters:
 Arp et al. (1953) „bifurcation of the red giant branch”
18
AGB Stars: History
•Sandage and Walker (1966) – were the first authors to use
term AGB.
• The term AGB originted as a description of the sequence of
stars in the HR diagrams, the term AGB is now used to
describe all stars with M < 8Mo that are on the second ascent
(asymptotic) into the RG region of the HR-diagram.
19
AGB Stars: History
• Hoyle & Schwarzschild (1955) showed that evolution of
stars through the RGB to max L and then down to HB can be
understood.
20
AGB Stars: History
• Merrill (1952) discovered lines of 99Tc, t~2 105 years!! (sprocess element). The short half-life time showed that Tc has
been recently dredge-up to the surface.
• Iben (1975) showed models which produce C in He-burning
shell by the triple-a process (formation of Father Secchi’s star
has been explained).
There are no stable isotopes with Atomic Mass 5 or 8 (i.e such
that reactions like:
4He + 1H --> 5X or 4He + 4He --> 8X may occur).
The next stage is the triple-a process:
4He + 4He + 4He --> 12C
This reaction requires both very high T (> 100 milion K) and
very high densities which will occur only after the star has
exhausted its H and has a core of nearly pure He. Only stars
with masses > 0.4 Mo will can ignite 3-a process.
21
AGB Stars: History
New results from new observing techniques - IR astronomy:
•Infrared astronomy started in 1960’s due to strong intrest
from military.
• ~1970 observations were made in all telluric windows from
1-20 mm.
22
Transmission on Mauna Kea: 4.2 km.
J:1.25, H:1.65, K:2.2 mm
Water vapour: 1.6 mm
23
Transmission on Mauna Kea: 4.2 km.
L:3.5, M:4.7 mm
Water vapour: 1.6 mm
24
Transmission on Mauna Kea: 4.2 km.
N:10.5 mm
25
Transmission on Mauna Kea: 4.2 km.
Q:19.5 mm
26
Sir Frederick William Herschel
• F.W. Herschel (1738 1822) was born in
Hanover.
• From 1757 he lived in
England.
• A musician and an
astronomer.
• In 1781 he discovered
Uranus;
• He created catalogs of
double stars and
nebulae;
• In 1800 he discovered
infrared radiation.....
27
Discovery of IR radiation.
28
AGB Stars: History
New results from new observing techniques –
IR astronomy:
• Neugebauer & Leighton (1969) – 2.2 mm survey (IRC). About
5000 sources were detected north of d=-33o , e.g. IRC+10 216
(the nearest C-star), sources associated with Sgr A. Most of the
sources were red giants.
•Price & Walker (1976) – RAFGL – Revised Air Force ... ~2400
sources with photometry at 4 bands between 4 and 28 mm, e.g.
AFGL 2688 (Egg Nebula); AFGL 915 (Red Rectangle).
•IRAS (1983) – photometry @ 12, 25, 60 and 100 mm (~250000) +
LRS spectra (~10000) for the brightest sources.
• ASTRO-F 2006!!!
•ISO (1995-1998), SST (2003-...), HSO (2009-2013), ....
29
AGB Stars: History
Mass loss on AGB:
•Deutsch (1956) noticed that circumsttellar absorption lines in
the MII component of the binary system a Her were seen also
in the spectrum of the companion GI star => Renv ~ 2 105 Ro.
With Vexp~10 km/s he estimated Mloss ~3 10-8 Mo/yr.
•Reimers (1975) collected data for many such systems and
concluded that Mloss ~ L R / M (Reimer’s formula).
• Gillet et al. (1968) identified emission band ~ 10 mm at
spectra of M-type giants as due to silicate dust.
•Hachwell (1972) discovere 11.5 mm band in C-stars (SiC)
• Gilman (1969) explained the observed dust dichotomy as
due to the high binding energy of CO molecule (like Russell
30
1934 for stellar spectra!).
AGB Stars: History
•At the begining of 1970’s it was clear that AGB stars produce
dust (a question of dust origin was open from 1930’s when
interstellar extinction was discovered).
• Goldreich and Scoville (1976) developed a model of mass
loss due to radiation pressure on dust and momentum
exchange between dust and gas.
•In all calculations of stellar evolution before ~1980 the
assumption was made that M* did not change!
•Schoenberner (1979, 1980) was first who employed the
Reimer’s formula for the stellar evolutionary calculations.
• However, it was aalready then clear that Reimer’s formula
predict too small mass loss rates for the AGB phase of stellar
evolution (observations suggested Mloss up to ~10-4 Mo/yr).
• The life of AGB stars is cut off by mass loss!!! (Iben &
Renzini 1983).
31
AGB Stars: Overview
32
AGB Stars: observational characteristics
•The most important spectral classes of AGB stars are M, S
and C. MS –top: dominated by TiO (VO – in very cold stars);
C- bottom: C2 and CN molecules dominate. S-stars have ZrO;
Zr is s-process element.
33
AGB Stars: observational characteristics
•M-type stars: O > C; TiO, VO (very cold stars)
 MS-type
•S-type stars O ~ C; ZrO (Zr – is s-process element
 SC-type
•C-type stars O< C; C2, CN....
•A particularly interesting s-process element found in the
atmospheres of some AGB stars is 99Tc, t~2 105 years. Its
presence means tht it has been brought to the stellar surface
in the last few times 105 years.
•This is direct observational evidence for the production of
new elements inside stars.
34
AGB Stars: how to recognize them?
•Other properties:
 TP – thermal pulse;
 presence of s-process elements (the efect of dredge-up
after TP): Zr, V, ... and especially 99 Tc;
S- and C-stars are AGB, but ... a care should be taken of a
(possible) binarity;
Mass loss > 10-7 Mo/yr is typical for AGB (supergiants,
LBVs have also large mass loss rates – but they are rare);
Long-period pulsations (AGB stars are Long Period
Variables – LPVs).
35
AGB Stars: variability
•Classification of the light curves of LPVs (as defined in the
General Catalogue of Variable Stars: GCVS):
 Mira-like „M”: regular variations with a large amplitude
in the V-band (DV > 2.5);
Semiregular variables of type a „SRa”: relatively regular
with a smaller amplitude in the V-band (DV < 2.5);
Semiregular variables of type b „SRb”: poor regularity
with a small amplitude in the V-band (DV < 2.5);
 Irregular „L”: irregular variations of low amplitude in
the V-band.
•The high quality data are available now from microlensing
surveys: MACHO (Alcock et al. 1995); EROS (Aubourg et al.
1993); OGLE (Udalski et al. 1993). See also:
http://www.aavso.org/adata/curvegenerator.shtml or
http://www.vsnet.kusastro.kyoto-u.ac.jp/vsnet/gcvs
36
AGB Stars: variability
•The V-variations of Miras
can reach 6 mag, but
bolometric variability is
smaller (most of the energy
is emitted in the IR).
•The large amplitude in the
shorter l’s is a result of:
Strong variations of the
TiO bands during pulsation
cycle
A large change of flux in
short l’s with Teff.
37
Thermal radiation
l[mm] * T[K]=3000
38
Useful relations
Bv 
Bl 
2 hv
c
2
2 hc
l
3
5
2
erg


 2

hv
 
exp    1  cm s Hz ster 
 kT 
1
1 [Jy] = 10-23 [erg/cm2/s/Hz]
erg


 2

 hc   cm s cm ster 
exp 
 1
 l kT 
1
Bn dn = Bl dl ; n l  c
1 [W] = 107 [erg/s]
39
AGB Stars: Period-Luminosity relations
•Glass & Lloyd Evans (1981) discovered a linear relation
between K-magnitude and log(P) in Mira variables.
• Hipparcos distances have been used to look for P-L
relations (e.g. Bedding and Zijlstra 1998; Whitelock & Feast
2000)
•However, the most exciting results have been obtained from
studies of AGB stars in the LMC, where the distance is known
and the reddening is small.
40
Stellar evolution
•Schematic evolution of a
star of 1 Mo mass:
1-4 core H-burning
5-8 shell H-burning
(He core becomes
electron degenerate)
8 convection => the 1st
dredge-up: 4He, 14N, 13C
(CN + ON cycling) are
mixed to the surface
 9 Core He Flash
10-14 Core Helium
burning
After 14 E-AGB
41
H mass profile during evolution of the MS
42
H and He mass profiles
43
H mass profile during shell H-burning
44
T and density during shell H-burning
45
CNO cycle: CN; ON
12
C+ H →
13
N+γ
→
13
C + e + νe +1,37 MeV
13
1
C+ H →
14
N+γ
+7,54 MeV
14
N+ H →
1
15
O+γ
+7,35 MeV
→
15
N+ H →
12
13
15
15
1
N
O
1
+1,95 MeV
+
+
N + e + νe +1,86 MeV
4
C + He
+4,96 MeV
15
• CNO cycle (99.96 % up;
0.04% right).
16
17
17
1
N+ H
1
O+ H
F
1
O+ H
→
16
O+γ
→
17
F+γ
→
17
→
14
+
O + e + νe
4
N + He
46
Convection and the 1st dredge-up
47
Convection and the 1st dredge-up
48
Core He-burning
49
Stellar evolution
•Schematic evolution of a star of 5 Mo mass:
50
AGB stars: structure
•A schematic view of a 1Mo star. The structure is similar
regardless of the stellar mass: CO degenerate core + He- and Hburning shells. Pulsations take place in the convective env.
51
AGB Stars: structure
•Comparison between structure of 1 and 5 Mo stars.
52
AGB: the unstable He-burning shell
•Physical reasons for the thermally unstable He-shell burning
were recognized by Schwarzschild and Harm (1965). (The high
temperature sensitivity of the 3-a reaction and the thinness of
the shell).
53
AGB Stars: TP-phase
•Computations of TP-AGB
phase is difficult and time
consuming (synthetic AGB
calculations are involved).
•The most extensive sets of
full AGB calculations are
those of Bloecker (1995) and
Vassiliadis & Wood (1993).
•L=59250(Mc-0.522) and
tIP=3.05-4.5(Mc-1.0) the
Paczynski’s relations. They
are results of the presence of
a radiative layer betwen Hburning shell and the
convective envelope.
• A typical liftime on the
AGB is 106 years.
54
Post-AGB Stars
•When „superwind”
reduces the mass of H-rich
envelope below ~10-3 Mo the
star begins to shrink.
• In this phase of stellar
evolution both: mass loss
and nuclear reaction (~10-7
Mo/yr) lead to the reduction
of the H-rich envelope.
55
AGB Stars: TP-phase
•The energy production by He-shell flash is very rapid (~106
Lo in this case).
• When the He-shell flash energy escapes from the core, it
leads to a peak in surface luminosity lasting several hundred
years.!!
56
AGB Stars: characteristic „atmospheric” phenomena?
•low gravity – assumption of spherical symmetry less realistic.
•instability against convection in deeper layers – due to H
disociation and absorption (giant granular cells).
•instability against pulsations – generation of shock fronts
(complex dynamics)
•Molecule and dust formation - complex radiation transfer &
wind generation.
•Interaction between: 1.) convection, 2.) pulsations, 3.) radiation,
4.) molecular and dust formation and absorption, 5.) acceleration
of the stellar wind.
57