Magnetic Fields in the Envelopes of LateType Stars: Circular Polarization of H2O Masers

1
Magnetic Fields in the Envelopes of Late-Type
StarsCircular Polarization of H2O Masers

• Wouter Vlemmings, JBO
• Phil Diamond, JBO
• Huib Jan van Langevelde, JIVE

2
Role of Magnetic Fields

• Mass loss
• Alfvén waves can drive stellar winds and produce
  clumpy mass loss
• Outflows
• Shaped by magnetic fields ?
• Magnetic pressure dominates
• the thermal/kinetic pressure
• for high magnetic fields
• Planetary nebulae
• Distinctly a-spherical

3
Circumstellar Masers

• Onion model
• Dust at few AU
• Molecules until
• dissociation by UV
• Excitation varies
• SiO at few AU
• Water at up to few 100 AU
• OH at 500 2000 AU
• As shell acceleration decreases
• radial amplification starts to dominate
  tangential amplification

4
Previous Observations

• SiO Masers
• Highly ordered Magnetic Fields
• Field Strengths (Zeeman)
• Supergiants up to 100 G
• Miras 10-30 Gauss
• But non-Zeeman interpretation
• Fields factor 1000 less
• OH Masers
• Some indication of alignment with CSE structure.
• Field Strengths
• Both Supergiants and Miras show a few mG fields

TX Cam
VX Sgr
Kemball and Diamond, 1997, ApJ 481 L111
5
Observations

• H2O non-paramagnetic
• Very small Zeeman effect leads to S-curve in the
  V-spectrum
• Need high spectral resolution to detect this
  signature
• Need high spatial resolution to distinguish
  between maser features
• VLBA observations of 6 late type stars
• Supergiants S Per, VY CMa, NML Cyg and VX Sgr
• Mira variables U Ori and U Her (masers of R Cas
  undetected)
• U Her observed in 2 epochs 5.5 years apart
• Correlated twice
• All 4 polarizations, 0.1 km/s resolution
• RR and LL only, 0.027 km/s resolution
• Calibration
• First calibration on low spectral resolution
• Apply solutions on high resolution data

6
First H2O Magnetic Fields Results

• Using simple LTE approximation
• Clear detections on S Per
• Only up to few
• Also fit for scaled down replica of total power
• due to intrumental effects
• Rule out systematics
• Varying values and directions
• B 207 30 mG
• But
• V spectrum narrower than thermal Zeeman
• No linear polarization
• non-LTE works better

Vlemmings, Diamond, van Langevelde, 2001, AA 375
L1
7
Polarization Analysis

• Non-LTE method
• (Nedoluha Watson 1992)

• Calculate Equations of State
• Linear maser geometry
• Including interaction between
• 3 dominant Hyperfine lines
• Their magnetic substates
• Total of 99 non-linearly related equations
• Solve for various thermal line widths of the
  maser medium

• Directly fit the observations to the models
• Partly explains narrowing
• (2D or 3D could provide solution)

8
Results

• S Per
• H2O 150 mG / 200 mG
• OH 1 mG (Masheder et al. 1999)

• VY CMa
• H2O 175 mG / 200 mG
• SiO 65 G (Barvainis et al.1987)
• OH 2 mG (Cohen et al. 1987)

No linear polarization

• NML Cyg
• H2O 500 mG / 500 mG
• OH 2 mG (Cohen et al. 1987)

• U Her
• H2O 1.5 G / 2.5 G
• OH 1 mG (Palen Fix 2000)

9
New Results

• U Her
• H2O 700 mG upper limit
• Previous 1.5 G
• OH 1 mG (Palen Fix 2000)

• U Ori
• H2O 2-3 G
• OH 10 mG (Reid et al. 1979)

Again, no linear polarization

• VX Sgr
• H2O 0.3 3 G
• OH 1-2 mG
• (Szymczak et al. 2001)
• SiO 80 G (Barvainis et al. 1987)

10
VX Sgr

• Fit dipole magnetic field
• using VX Sgr stellar outflow
• model (Chapman Cohen, 1986)
• Vexp 20 km/s

• Fit results
• Polar angles
• - ? 35 deg
• - ? 230 deg
• Surface field
• - B 2 kG
• indication of
• elongated outflow
• along the equator
• But need better
• outflow model

11
Interludemagnetic field lines
field lines for increasing ratio between stellar
rotation and outflow velocity. -from dipole
(black) to toroidal
12
VX Sgr (2)

• Previous results
• -H2O maser show equatorial expansion with polar
  axis at ?60 30, f200 20 deg (Murakawa et
  al.2003)
• -OH masers show similar negative/positive
  polarization structure (Zell Fix, 1996)
• -OH maser structure consistent with dipole field
  with ?25 5, f210 30 deg (Szymczak et
  al. 2001)
• -Field strength interpolation consistent with
  dipole field

-Our fit - ? 35 10 deg - ? 230 20 deg
13
Magnetic Fields in CSEs

• Observations trace
• Inner edge of the maser region
• High density clumps
• Solar Type (r -2) and dipole magnetic fields fit
  data
• Dipole field favored for VX Sgr and U Ori
• Surface field of 100 G (Miras) to several kG
  (Supergiants)
• Magnetic pressure can drive outflows and help
  shape nebulae
• Real fields likely more complex than simple power
  law (e.g. Pascoli, 1997)

Vlemmings, Diamond, van Langevelde, 2002, AA
394, 589
14
Planetary Nebulae

• Magnetic pressure in the H2O maser region
• ? ? 8 ? nH k T / B²
• (ratio of thermal and magnetic pressure)
• ? ? 0.05 the magnetic pressure dominates by a
  factor of 20 for B ? 250 mG.
• Asymmetric nebulae possibly due to
• magnetic shaping of the outflow (García-Segura,
  1999)
• binary Interaction (Soker, 2002)
• wind interaction with a warped circumstellar disk
  (Icke, 2003)
• disk results from only moderately strong dipole
  field (Matt, 2000)
• additional warping may be caused by high magnetic
  fields (Lai, 1999)

15
Planetary Nebulaerecent discussions

• Central stars of PNe
• VLT polarization measurements indicate fields of
  1-3 kG (Jordan, Werner OToole, 2005)
• fairly consistent with maser measurements
• Large scale magnetic fields
• maintaining large-scale magnetic fields might be
  impossible as it will quench rotation and thus
  dynamo action (Soker, 2005)
• would need companion star
• but localized fields cannot explain observed
  structure in H2O and OH maser polarization.

16
Conclusions

• Zeeman interpretation is favored
• No linear polarization
• LTE models appear too simple
• Coupled transfer models (non-LTE) promising
• Constraints on saturation beaming
• Inferred magnetic fields fit nicely
• Compared to OH SiO values and dipole or solar
  type magnetic field
• Indicate surface fields of ? 1-2 kG
  (supergiants) 100 G (Miras)
• Comparable to dynamo-produced fields (Blackman et
  al. 2001)
• Comparable to CS-PN (1-3 kG Jordan et al. 2005)
• Mira stars in sample appears to have a stronger
  field
• H2O maser in thick shell, closer to the star
• Masers in P-PNe can provide clues on evolution of
  the magnetic fields
• observations performed on IRAS 192962227, K3-35
  and W43A

17
Proto-Planetary Nebulae

• K3-35
• H2O masers in a magnetized torus at 85 AU
• Expect fields of several hundred mG
• (several mG fields observed in OH masers at 800
  AU)
• Also at 5000 AU at the tip of the bi-polar
  outflow
• (Miranda et al. 2001)

18
Proto-Planetary Nebulae (2)

• Collimated and precessing jet from W43a
• Formed approximately 28 years ago
• H2O masers at 500 AU
• (Imai et al. 2002)