Transcript No Slide Title

Observing magnetic fields
in star-forming regions
Jim Cohen
17th February 2004
The University of Manchester
Jodrell Bank Observatory
Zwolle Workshop
Outline of Talk
Introduction
Polarization Mechanisms
Zeeman Splitting
Maser Regions
Bipolar Outflows Align with Polarization of Starlight
Cohen et al. 1984, MNRAS 210, 425-438
Magnetic pressure estimated from OH maser Zeeman
splitting is significant in dynamics of bipolar outflow.
Are Cloud Cores Collapsing?
Virial equilibrium:
P s+ |W| = Ms + Mw + 2T
P s External pressure
W Gravitational energy
Ms Static B
Mw Alfven wave B
T
Internal Kinetic Energy
Evolutionary
Effects
What are the
polarization signatures
of protostellar
evolution?
Vallee & Bastien
2000, ApJ, 530,
806-816
General Remarks
There are many techniques available to estimate B
but not usually in one and the same source.
Some measurements give B  , some give B  ,
some give B magnitude, some give the direction,
some give the full vector B.
Polarized flux is often less than 1% so we are
usually struggling for sensitivity.
Stokes parameters are additive. Therefore
polarization structure that is unresolved either in
frequency or spatially will lead us to underestimate
the true degree of polarization.
Faraday Rotation
  2  ne B cos dx
Can mask true direction of B
Pulsar DM  2  ne dx
B cos   RM/DM
useful for large-scale Galactic B but not
small scale studies of star-formation
Continuum Polarization
Synchrotron
Aligned Dust Grains
EB
Emission E  B
(FIR or submm)
Extinction E  B
(optical)
Scattering E  B
(optical, NIR)
Interstellar Polarization in Taurus Dark Clouds
Messinger, Whittet & Roberge
1997, ApJ 487, 314-319
Well organized on large
scale, but only outer
layers of dust clouds
are probed.
Note wavelength
dependent PA of two
stars – dust properties
change with grain size
and location (depth) in
cloud. Field direction
twists inside cloud.
Lang et al. 1999, ApJ, 526, 727-743
Chuss et al. 2003,
ApJ, 599, 1116-1128
350m poln
(Hertz on CSO)
overlaid on 20cm continuum
Dense: B  b
(toroidal)
Rare: B  b
(poloidal)
Chuss et al. 2003,
ApJ, 599, 1116-1128
Collapse can
produce toroidal
B in mol cloud
while leaving B
poloidal outside.
Magnetic
reconnection can
produce the
energy for the
nonthermal
filaments.
OR bipolar wind
Classical Zeeman Effect
An electron in a
magnetic field B
precesses at the
Larmor frequency
L = eB/2me .
Blended: Bcos
Unblended: B
Spectral lines are
split into three
polarized
components at
(angular)
frequencies o ,
o + L and o - L
HI Zeeman
NGC6334 source E
Weak splitting, sigma
components dominate.
Stokes V = z Bcos dI/d
where z is the splitting factor.
Measures line-of-sight
component Bcos.
Instrumental issues limit
usefulness to strong fields
exceeding ~10G.
Sarma et al. 2000, ApJ 533, 271-280, VLA 35 x 20 arcsec
M17
Brogan & Troland 2001, ApJ 560, 821-840
VLA OH and HI
Bcos increases
where Bsin
(traced by 100m
poln) decreases.
Either B is
bending around
the HII region or
the dust
properties are
being changed by
the HII region.
Quantum Zeeman Effect
A magnetic dipole μ in a magnetic field B has a
potential energy μ.B that is quantized:
μ.B = B g J μB / ħ
where μB = eħ/2me is the Bohr magneton. Lande factor
g ~ 1 (paramagnetic) or ~ 10-3 (non-paramagnetic), but
depends on total angular momentum F and is different
for upper and lower states in general.
States split into 2F+1 substates with allowed transitions
Δm = +1
σ+
Δm = 0
π
Δm = -1
σ-
Linear polarization is parallel to B for π components,
perpendicular to B for σ components.
OH Zeeman
Polarization and
intensity depend
on angle of B to
line-of-sight
Splitting  B
provided hyperfine
components don’t
overlap. Otherwise
see Elitzur (1996,8).
Complete Zeeman
pattern can be
complex.
Maser propagation/competive effects
OH Thermal Absorption
NGC6334
Sarma et al. 2000, ApJ 533, 271-280, VLA 16 x 12 arcsec
OH Thermal Emission
Crutcher & Troland 2000, ApJ 537, L139-L142
Arecibo 2.8 x 3.2 arcmin
CN Zeeman
Crutcher et al. 1999, ApJ 514, L121-L124 Pico Vateta
DR21(OH) 0.71mG
CN 1-0 at 113 GHz
Traces 105-106 cm-3
9 hyperfine components,
well separated in velocity
4 strong Zeeman, 3 weak
Zeeman effect, 2 useless
Different splitting factors
reduce systematic errors
Simultaneous fitting to 4
strong (upper) and weak
(lower) components
OMC1n 0.36 mG
Magnetic Fields in Molecular Clouds
Crutcher 1999, ApJ 520, 706-713
H2O Masers
OH Masers
Excited OH
CN
B  nH20.5
Ambipolar diffusion?
Or constant VAlfven
B(4)-1/2  0.7 km s-1
We Need More Tracers of B
OH thermal emission and absorption generally traces
the outer regions of molecular clouds but not the dense
cores.
Crutcher et al. 2004 propose use of randomness in
polarization vectors to estimate B (Chandrasekhar &
Fermi 1953) based on MHD wave argument
Bsin  n1/2 V -1
L1544 results in OH give smaller B than SCUBA
polarimetry at 850 microns which penetrates core.
Could have angle  = 16 to line of sight to be
consistent.
Prestellar Cores
Ward Thompson et al. 2000, ApJ 537, L135-L138
L183
Crutcher et al. 2004, ApJ 600, 279-285
Bsin = 80G
L1544
SCUBA 850 m 14 arcsec Bsin = 140 G
MERLIN
Multi Element Radio Linked Interferometer Network
 D
D = 218 km
0.170" 18 cm
0.042 " 4 cm
0.013" 1.4 cm
13x1612-MHz, 430x1665-MHz, 3x1667-MHz masers
Orion-KL
OH masers
trace a
rotating and
expanding
molecular
torus at the
centre of the
H2 outflow
(Gasiprong
2000, PhD
thesis).
Magnetic Beaming in
Masers
Complete Zeeman patterns
rarely observed.
σ-components grow fastest
and can suppress π-comps
(Gray & Field 1995).
100% circular polarization
most common.
Zeeman shift has same effect
as velocity shift. In a
turbulent medium LHC and
RHC trace different
molecules in general.
W75N
σ+ π
σ-
W75N
Vector B
OH maser
polarization
indicates 3-d
magnetic field
with suitable
interpretation
(need to identify
-components)
Garcia-Baretto et al.
1988 ApJ 326, 954
W75N bipolar outflow
0.6pc
•
Shepherd et al. 2003, ApJ 584, 882
Large-scale B-field parallel to outflow (submm poln).
2000AU
OH Masers
Hutawarakorn & Cohen
2002, MNRAS 330, 349
Kinematics show
a rotating and
expanding disc
(torus) orthogonal
to the outflow.
Strong linear poln
up to 100%.
Vectors are either
parallel to outflow
or perpendicular.
1665 MHz
0.010 pc
OH Masers
continued
Magnetic field
reverses on
opposite sides of
disc (toroidal
component).
Field lines twisted
up in the rotating
disc.
Uchida & Shibata
(1985) model is
supported.
1667 MHz and 1720 MHz
Twisted Magnetic Field
Uchida & Shibata 1985 hydrodynamical computation.
(a) large scale field contracts with disc
(b) disc twists field lines
(c) close-up of core
PASJ 37, 515
Model of OH masers and polarization
Synthetic maser spectra generated using polarizationdependent model of propagation, with physical
conditions taken from Uchida & Shibata (1985) model.
Gray et al. 2003, MNRAS, 343, 1067-1080.
Masers originate at different depths in disc.
IRAS 20126+4104 Bipolar Outflow
Cesaroni et al., in press Plateau de Bure
Edris et al., in preparation MERLIN
Vallee & Bastien 2000,
ApJ 530, 806-816
SCUBA
B  outflow
H2O Maser Polarization
Sarma et al. 2002, ApJ, 580, 928-937
Hyperfines?
VLA
H2O Linear Polarization
Imai et al 2003, ApJ 595, 285-293 VLBA
Where Next?
3-d magnetic field studies are sensitivity limited for
now (key polarized flux is only a small % of total).
Potential to probe range of densities to 1010cm-3.
Major new IR/submm/mm facilities are coming and
will overlap with masers at subarcsec resolution.
Some key questions:
• How to treat overlap of hyperfine components?
• Relation to galactic magnetic field?
• Magnetic field evolution, does B dominate?
• Maser lifetimes and source evolution?