Title: Star formation
1Star formation
Suzanne Ramsay UK Astronomy Technology
Centre, Royal Observatory Edinburgh
UKIRTWFCAM infrared image of Orion
2The Challenge
- A theory of star formation requires to explain
the origins of stars over four orders of
magnitude in mass - From 0.01 M? brown dwarfs powered only by
gravitational energy - To gt100 M? stars with lifetimes around 1million
years - The typical star has mass 1 M?
- So, what do we know and how do we know it?
3Stars form in molecular clouds
- although stars are generally not in clusters
young stars are and so these are identified as
the sites of star formation
From Dame, Hartmann and Thaddeus 2001.
4GMC chemistry
- gt100 molecules discovered in MCs
- H2 most abundant
- CO commonly studied at 10-4 of H2 abundance
since it emits from cold GMC which H2 does not - complex molecules detected include formaldehyde,
amino acids. - Important constituent (1 of ISM) is dust (C,
Si) - much cloud chemistry takes place on dust grains
- Most dust mass is in grains size 1000A, 109
atoms
5GMC chemistry
- Cosmic ray flux is key parameter in cloud
chemistry and provide ionisation deep in clouds,
though ionisation fraction is always low - Magnetic field acts on ions, indirectly on gas
and dust - Strength of B-field 1-10 mGauss
6Properties of GMCs
- 2-4 of interstellar volume
- The rest is the atomic interstellar medium
- Lifetime, debatable but lt 107 years
- Free-fall timescale 106 years
- Typically dispersed by radiation from massive
stars, timescale 107 years - Supported by magnetic fields and turbulence due
to motion of clumps - Observed galactic star formation rate 3 M?yr-1
- Star formation in clouds is relatively
inefficient 1-3 of the cloud ends up as stars
7- Within the Orion molecular cloud higher density
clumps are readily identifiable
8- Stars form from yet smaller structures - cores
- OMC has stars of various ages
- At 460pc, the Orion Nebula is our closest
laboratory for studying massive star formation
9Star formation in clusters
- Embedded clusters
- T associations e.g. Taurus
- R associations (AB stars) e.g. Mon R1
- OB associations (massive stars e.g. BN-KL in
Orion) - Open clusters (e.g. Hyades, Pleiades) can be very
old
10dense cores
- Bok globule b335
- typical formation site for an individual star
11Phase GMCs Clumps Cores
Mass (M?) 6x104 2x106 102 1-10
Size (pc) 20-100 0.2-4 0.1-0.4
Density (cm3) 100-300 103-104 104-105
Temp (K) 15-40 7-15 10
B (mG) 1-10 3-30 10-50
Line width (kms-1) 6-15 0.5-4 0.2-0.4
Dynamical life (years) 3 x 106 106 6x105
12Extinction
- Some values
- AV20mag
- AK2mag
- Much higher for dense cores
13Star formation requires long wavelength astronomy
- High obscuration means that many starformation
phenomena require long wavelength observations - mm, submm and infrared
- Youngest sources are the most deeply embedded and
therefore the hardest to study
14Evolution of a (low mass) protostar
Evolutionary sequence From Andre, Ward-Thompson
Barsony 1993 Extended from original by Lada
1987
15Starless cores
- Starless core or pre-stellar core
- Cold (lt15K)
- Sufficient mass for protostar envelope (0.05-30
M?) - Gravitationally bound, but no protostar
16Core collapse
- Considering the core as an isothermal sphere
- Density ? 1/r2
- Maximum mass for such a sphere is the Bonor Ebert
mass - M gt MBE, collapse starts with central core
M?
Balances surface pressure from the cloud,
velocity dispersion from temperature and gravity.
17Core collapse
- If unmediated, free fall collapse with Density ?
1/r3/2 and vff2? 1/r1/2 - Requires additional support otherwise
- Timescales too fast
- Velocities become supersonic and core fragments
18Magnetic Support
- Clouds are known to contain magnetic fields
- These support the cloud against collapse
- Mechanism to allow slow collapse required
- Ambipolar diffusion
- Neutral particles immune to magnetic field drift
to the centre of the core - Ionised particles remain fixed by the field lines
- Once the core mass reaches critical level,
collapse proceeds - AD timescales are too long for standard initial
conditions - Effect of AD increased by turbulence
19Starless cores
- Observed magnetic fields inadequate for ambipolar
diffusion model - Turbulent support of the core required
Ward-Thompson, Motte, Andre 1999
20Class 0 sources
- Sources with a central protostar that are very
faint/undetectable in the optical/NIR - Lsubmm/Lbol gt 0.5
- Menvelopegtm
- Tbol lt 70K
21Class 0 sources
- First Class 0 source, VLA1623, discovered in Rho
Ophiucus (1993)
- Andre, Ward-Thompson, Barsony 1993
22Class 0 sources
- Sources with a central protostar that are very
faint/undetectable in the optical/NIR - Lsubmm/Lbol gt 0.5
- Menvelopegtm
- Tbol lt 70K
- The deeply embedded protostar acquires most of
its mass during this phase - Bipolar molecular outflows are associated with
Class 0 sources - Mechanism for removing angular momentum
23B335 revisted
- Contains embedded source of 3 L?
- Contains a disk, radius 100AU
- Density profile inner region of r-1.5 and outer
envelope r-2 (to 5000AU) - Inner density profile consistent with
gravitational free fall
H2CO map from Choi. A bipolar outflow is
detected from the embedded young source
Harvey et al 2003 sub-mm imaging reveals. Disk of
radius 100AU.
24Protostellar evolution
- Most of the core mass must be ejected to evolve
from Class 0 to Class I - During their evolution, Class 0 sources
- Increase mass from 0.3 M? to 3 M?
- Mass accretion regulated by deuterium burning
- Luminosity reaches 10-100 L?
25Class I sources IR visible protostars
- Sources with air gt 0 over the wavelength range
from 2.2 to 10-25mm - air is the slope on the spectral energy
distribution - These sources have both disks and envelopes
- 70K lt Tbol lt 650K
- Identifiable by their large infrared excess
- Infrared emission lines detectable
- Outflows, less energetic than those from Class 0
26Class 0/I sourcestimescales
- Time spent in Class I phase 1-5 105 years from
statistical arguments on source numbers - This works under assumption that the various
classes are an evolutionary trend - 10 times fewer than Class II
- Timescale for Class 0 - 104 years in Rho Oph
- 10 times fewer than Class I
- Implies mass accretion rate of 10-5 M?Yr-1 to
form half solar mass star
27Class II sourcesClassical T Tauris
- Sources with -1.5 lt airlt0 pre-main sequence
sources with large circumstellar disks - Optically visible
- H-alpha and forbidden lines from outflow
- Stellar photospheric features, but often veiled
by disk/dust continuum - Ages 1-4 x 106yr
T Tauri. 2MASS Atlas Image mosaics by E. Kopan,
R. Cutri, and S. Van Dyk (IPAC).
28- Strong infrared excess initially hypothesised as
an obsuring disk, with later observational
confirmation
29Class III sourcesWeak line T Tauris
- Sources with airlt-1.5 pre-main sequence stars
that are no longer strongly accreting - Disks disspipated, so optically visible
- weak-lined - H-alpha equivalent width lt 10 Å
- Ages 1-20 x 106yr
- Final state for our low mass protostar
- Somewhat ambiguous definition as e.g. not all
stars with disks have strong H-alpha and vice
versa
30Accretion and outflow
- Outflows and jets are a ubiquitous phenomenon
associated with star formation - They appear during all phases, but with trends in
their evolution with protostellar class - Class 0 highly collimated, luminous
- Class 1, lower collimation, less energetic
- Momentum flux of outflow predicted by modelling
to be proportional to mass accretion so Class 0
sources have higher accretion than Class 1
31Accretion and outflow
HH212 (above) and HH211 (below) are class 0
sources high collimation, highly luminous
molecular outflow
32HH-30
HH-47
33Outflows and angular momentum transport
- Preferred launching mechanism for outflows is
magnetic - Capable of explaining high degree of collimation
and outflow strength - Material ejected along magnetic field lines from
the disk - Field geometry is crucial, but a succesful model
can remove a large fraction of angular momentum
with a small amount of material - Launch sites disk disk-star interface stars
surface
34High mass star formation
- Stars above 8 M? cant form by the same process
as low mass - Hydrogen burning ignites during accretion phase
- Yet they conspicuously exist, though in small
numbers compared with low mass stars - Extreme examples
- Eta Carinae 100 M? the Pistol 150 M?
- LBV 180620 130-190 M?
35High mass star formation
- Fundamental difficulties in observing high mass
star formation is due to the rarity of the
sources, the distance of the nearest examples - Recent intense effort is providing larger samples
of candidate HMYSOs based on infrared colours,
radio data
36High mass star formation
- Basic problem Kelvin Helmholtz timescale
exceeds the free fall timescale - tKH104 years for an O star (107 for the Sun)
- Contraction proceeds faster than accretion of
material from the cloud and hydrogen burning
begins while still embedded in the cloud - Alternative formation mechanism? E.g. coagulation
from lower mass stars
37HII regions as signposts
- HII regions form once Hydrogen burning ignites
producing Lyman continuum photons - Electron free-free emission detected in radio
- Embedded HII regions are constrained as compact
or ultra-compact HII regions
38High mass young stellar objects
- hot cores (T100K) observed associated with or
as precursors to UCHII regions
39High mass young stellar objects
- Sub-mm imaging reveals dense cluster of sources
analogous to the Trapezium cluster in Orion
Outflow activity in the region SiO jet
Beuther et al. 2007
40Outflows from HMYSOs
- Well know examples of high mass outflows have
suggested low collimation compared with low mass
sources - Different mechanism for generation or low spatial
resolution?
41IRAS201264104Varricatt et al. 2008
Outflows from high mass sources
IRAS18151-1208Davis et al. 2004
42Brown dwarfs
- Stars with insufficient mass to star hydrogen
burning - Mass limit 0.011-0.013 M? (12-14MJup)
- Brown dwarfs represent bridge the gap between
stars and planets - Stars form from collapsing cloud cores
- Planets from coagulation of material in
circumstellar disks (during the Class II stage)
43Formation of the lowest mass stars
- Brown dwarf discoveries
- L and T dwarfs now numerous, identified from
their very red colours through 2MASS and Sloan
surveys - T dwarfs M 80MJup-10MJup, Temp800K
- Surveys with e.g. WFCAM on UKIRT, VISTA promise
the discovery of yet cooler, lower mass objects
the (as yet) mythical Y dwarf - NB 2-3 objects for 100s sq degrees of sky
44Formation of the lowest mass stars
- Statistics suggest that brown dwarfs have much in
common with stars - Possible formation mechanisms include
- photo-evaporation of cores by HII regions
- ejection from star forming cores
- fragmentation of low mass prestellar cores
- All supported by modelling which dominates?
45Outflow from 2MASSW J1207334-393254
Subarcsecond outflow detected from a 24 Jupiter
Mass brown dwarf (Whelan et al. 2007, ApJ, 659,
L45.
46The initial mass function
- From Salpeter (1955)
- The relative number of stars produced per unit
mass interval - Derived from the observed luminosity function
- Power law function of Mg, slope g -2.35
47Initial mass function
Example observed IMF
Salpeter mass function
48The initial mass function
- Salpeter power law slope g -2.35
- Now updated
- C(M/ M? )-1.2 0.1 lt M/ M? lt 1.0
- C(M/ M? )-2.7 1 lt M/ M? lt 10
- 0.4C(M/ M? )-2.3 10 lt M/ M?
49Determining the Initial mass function using
clusters
Low end of the IMF needs deep IR observations and
observations of open clusters
Establishing slope for high Mass stars requires
observations Of OB associations
50IMF in clusters
51The initial mass function
- The IMF for field stars and those in clusters
shows it to be the same - confirmation that the stars did form in clusters.
- More recently, the core mass function found to be
consistent with the stellar IMF - The IMF is robust to a variety of clusters and
environments, but so far lacking theoretical
basis
52The end
These stars provide most of the mass in the galaxy
These stars dominate energy feedback and
chemical enrichment
These stars provide most of the luminosity in the
galaxy.