Transcript High-Mass Star Formation

Deriving the Physical
Structure of High-mass Star
Forming Regions
Yancy L. Shirley
Collaborators: Neal Evans, Kaisa Young, Dan Jaffe, Claudia Knez, & Jingwen Wu
May 2003
SF in the Milky Way
1011 stars in the Milky Way
Evidence for SF throughout history of the galaxy (Gilmore 2001)
SF occurs in molecular gas
Molecular cloud complexes: M < 107 M0 (Elmegreen 1986)
Isolated Bok globules
M > 1 M0 (Bok & Reilly 1947)
SF traces spiral structure (Schweizer 1976)
M51 Central
Region
NASA
SF Occurs in Molecular Clouds
Lupus
Total molecular gas = 1 – 3 x 109 Mo
SF occurring throughout MW disk
(Combes 1991)
SF occurs in isolated & clustered modes
SF occurs within dense molecular cores
BHR-71
VLT
Pleiades
Orion Dense Cores
CO J=2-1
VST, IOA U Tokyo
Lis, et al. 1998
High-mass Dense Cores
Optical
RCW 38
Embedded clusters visible in Near-IR
W42
Near-IR
Blum, Conti, & Damineli 2000
J. Alves & C. Lada 2003
High-mass Cores : Complexity
S106
Near- IR
Subaru
High-Mass Star Formation
Star with M > 100 Mo appear to exist (Kudritzki et al.
1992): How do massive stars (M > few M0) form?
Basic formation mechanism debated:
Accretion (McKee & Tan 2002)
How do you form a star with M > 10 Msun before radiation pressure stops accretion?
Coalescence (Bonnell et al. 1998)
Requires high stellar density: n > 104 stars pc-3
Predicts high binary fraction among high-mass stars
Theories predict dense core structure & evolution: n(r,t) & v(r,t)
Observational complications:
Farther away than low-mass regions = low resolution
Dense cores may be forming cluster of stars = SED dominated by most massive
star = SED classification confused!
Very broad linewidths consistent with turbulent gas
Potential evolutionary indicators from presence of :
H2O, CH3OH masers
Hot core  Hyper-compact HII  UCHII regions  HII  Star ?
Hot Cores & UCHII Regions
Hot Cores & UCHII Regions observed in same highmass regions : W49A
VLA 7mm Cont.
DePree et al. 1997
BIMA
Wilner et al. 1999
Outline
What is lacking is a fundamental understanding of the
basic properties of the ensemble of high-mass star
forming cores
Texas survey of high-mass star forming cores:
Plume et al. 1992 & 1997 CS line survey
Dust Continuum 350 mm Survey
Mueller, Shirley, Evans, & Jacobson 2002, ApJS
Constrain n( r ), T ( r )
High-mass cores associated with H20 maser emission
Arectri catalog of H2O maser sources
Plume et al. 1992 & 1997 CS survey towards (0,0) position
CS J = 5 - 4 Mapping Survey
Shirley, Evans, Young, Knez, Jaffe 2003, ApJS
Dense gas properties
CS Dense Core Survey
CS J=7-6 detected 104 / 179 cores with H2O masers
Plume et al. 1992
H2O masers trace very dense gas
n > 1010 cm-3 for the 22 GHz 616-523 transition
Low J CO Surveys generally trace lower density gas.
H2O maser positions are known accurately to within a few
arcseconds. HII regions and luminous IR sources may not be
spatially coincident with dense gas.
Multi-transition study and initial mapping
Plume et al. 1997
71 cores detected in CS and C34S J = 2-1, 3-2, 5-4, and 7-6.
21 of the brightest cores mapped in CS 5-4
<R> = 1.0 pc, <Mvir> = 3800 Mo
LVG modeling of multiple CS transitions
CO: Molecular Cloud Tracer
Hubble
Telescope
NASA, Hubble Heritage Team
CO J=3-2
Emission
CSO
CS & HCN Trace Dense Cores
CO 1-0
CS 2-1
Helfer & Blitz 1997
HCN 1-0
CS LVG Models
Initially assumed n( r ) and
T( r ) = CONSTANT
40 sources detected in all 4
CS transitions
<log n> = 5.93 (0.23)
<log N> = 14.42 (0.49)
2-density component
model with a filling factor for
the dense component
nhigh ~ 108 cm-3
nlow ~ 104 cm-3
Typically, very high column
densities of low density gas
required (<log Nlow> = 16.16)
with f ~ 0.2
Plume et al. 1997
350 mm Survey
Mueller, Shirley, Evans, & Jacobson 2002
5 nights at the CSO 10.4-m telescope
51 high-mass (Lbol > 100 Lsun) cores associated with H2O
masers (Plume et al. 1992 sample)
850 pc < D < 14 kpc
All cores also observed in CS5-4 survey (Shirley et al. 2003)
SHARC 350 mm scan maps (4.0 x 2.7 arcmin)
qmb ~ 14 arcsec at 350 mm
100 arcsec chop throw
350 mm Images
G9.62+0.10
150,000 AU
W43
M8E
50,000 AU
W28A2
Mueller et al. 2002
W33A
10,000 AU
G23.95+0.16
Submm Continuum Emission
Submillimeter continuum emission is optically thin. The
specific intensity along a line-of-sight is given by:
Why must we model ?
Rayleigh-Jeans approximation fails in outer envelope of
low-mass cores
hn/k = 44 K at 350 mm
Heating from ISRF is very important in outer envelopes
of cores
Radiative transfer is optically thick at short l
Observed brightness distribution is convolved with
complicated beam pattern, scanning, and chopping
Radiative Transfer Procedure
nd(r)
L kn
Radiative
Transfer
Td(r)
Sn(l)
I(b)
Simulate
Obs.
Nearly orthogonal constraints:
Gas to
Dust
Physical Model
n(r)
SED
I(b)
Mass x Opacity
n(r)
Iterate
Observations
Dust Opacity
OH = Ossenkopf & Henning 1994
coagulated dust grains
Calculated Temperature Profiles
Mueller et al. 2002
Radiative Transfer Models
50,000 AU
Mueller et al. 2002
Best-fitted Power Law
Single power-law density
profiles fit observations
n( r ) = nf (r / rf) –p
p = - dln n/ dln r
Distribution of power law
indices
<p> = 1.8 (0.4)
Similar to distribution
of low-mass cores
modeled by Shirley et al.
(2002) & Young et al
(2003)
Mueller et al. 2002
Evolutionary Indicators ?
Mueller et al. 2002
“Standard” Indicators
Mueller et al. 2002
350 mm Survey Summary
Density and Temperature structure of outer envelope
characterized
<p> = 1.8 (0.4)
<n(1000 AU)> is order of magnitude higher than nearby low-mass
star-forming cores
Beuther et al. 1.2mm mapping 69 cores: <p> = 1.6 (0.5)
Single power law models fit our sample
CAVEAT: may be contribution from compact components (UCHIIs
or disks) within central beam
W3(OH) UCHII may contribute as much as 25% of the central flux assuming
optically thick free-free scaled from 3mm flux (Wilner, Welch, & Forster 1995)
<Rdec> = 0.16 (0.10) pc
<Tiso> = 29 (9) K isothermal temperature
Definitive trends lacking for evolutionary indicators
Except perhaps Tbol vs. Lbol/Lsmm
Lbol ranges from 103 to 106 Lsun
SEDs not well contrained in many cases due to lack of Far-IR
photometry
CS J = 5 - 4 Survey
Shirley et al. 2002
63 high-mass star forming cores associated with H2O
masers mapped at CSO 10.4m
<D> = 5.3 (3.7) kpc with 28 UCHII regions included
57 peak positions observed in C34S J=5-4, 9 in 13CS J=5-4
Over-sampled On-The-Fly maps in CS J=5-4
qmb ~ 25 arcsec at 245 GHz
Median peak integrated intensity S/N = 40
10 arcsec binned maps
Provide consistent sample from which to determine the
properties of the deeply embedded phase of high-mass star
formation
CS Rotational Transitions
Heavy linear molecule with
many rotational transitions
observable from the ground
J = 5 - 4 transition good
probe of dense gas:
mb = 1.98 Debye
nc(10K) = 8.8 x 106 cm-3
neff(10K) = 2.2 x 106 cm-3
CS J=5-4 Survey
G19.61-0.23
S231
S158
Shirley et al. 2003
M8E
W44
S76E
CS J=5-4 vs. Dust Continuum
CS J=5-4 is an excellent tracer of dense gas in highmass star forming regions
Shirley et al. 2003
Deconvolved Size vs. p
Convolution of a Gaussian beam pattern with a power law intensity
profile yields a deconvolved source size that varies with p
Shirley et al. 2003
Optical Depth Effect on Linewidth
C32S is typically optically thick, therefore must use rare
isotope (C34S) in linewidth sensitive calculations
Shirley et al. 2003
Linewidth-Size
Weak correlation with best fit: Dv ~ r0.3
C34S linewidth 4x larger than predicted linewidth from Casselli &
Myers (1995) indicating high turbulence: <Dv(C34S)> = 5.0 (2.0) km/s
Shirley et al. 2003
Size, Mass, & Pressure
Median core size: R = 0.32 pc
Alternatively Rn = 0.40 pc
Median projected aspect ratio: (a/b) = 1.2
Median virial mass: Mvir = 920 M0 corresponding to S = 0.6 g cm-2
Corrections for p and Dv broadening necessary
Mean mass per OB association ~ 440 M0 (Matzner 2002)
Median pressure <P/k> = 1.5 x 108 K cm-3
Shirley et al. 2003
Virial Mass vs. Dust Mass
The virial mass is consistently higher by a factor of 2 to 3 than the
mass determined from dust continuum modeling.
Uncertainty in dust opacity may account for difference
Shirley et al. 2003
Cumulative Mass Spectrum
Slope of mass spectrum similar to IMF and distribution of OB
associations G ~ -1.1 (0.1) (Massey 1995)
G  0.9
Shirley et al. 2003
Luminosity and Mass
Shirley et al. 2003
CS J=5-4 Survey Summary
CS J=5-4 is an excellent tracer of dense gas in high-mass
star forming cores
Aspect ratios consistent with spherical symmetry
Median size of 0.32 pc and median virial mass of 920 Msun
Virial mass a factor of 2 to 3 larger than dust-determined mass
Cumulative mass spectrum G ~ -0.9 similar to IMF of OB
associations
High median pressure of 1.5 x 108 K cm-3 ameliorates the lifetime
problem for confinement of UCHII regions
L/M is 100x higher than estimates from CO and has a
smaller dispersion
L/M 2x higher for cores with UCHII and/or HII regions
Lbol strongly correlates with Mvir. Combined with low dispersion of
L/M perhaps indicates that mass of most massive star is related to
the mass of the core
High Mass Pre-protocluster Core?
Have yet to identify initial
configuration of high-mass star
forming core!
No unbiased surveys for such
an object made yet
Based on dense gas surveys,
what would a 4500 M0, cold
core (T ~ 10K) look like?
Does this phase exist?
Evans et al. 2002
Conclusions & Future Work
Initial characterization of n( r ) indicates a power law
density structure of outer envelope
CS J=5-4 traces dense gas properties associated with
star formation
CS J=7-6 + HCN & H13CN J=3-2 Mapping Survey
(Texas Thesis projects of Jingwen Wu & Claudia Knez)
Radiative transfer modeling of dense gas & v( r )
Combination of bolometer camera + interferometric dust
continuum imaging with radiative transfer modeling is a
powerful diagnostic of the density & temperature
How much emission is coming from a compact component within central
beam?
SMA & ALMA submm continuum needed!
SOFIA & SIRTF needed to improve SED