M F Corcoran Overview Introduction Problems of Mass and Mass Loss Xray Fine Analysis as a Probe of Mass Loss Mass Outflows and Shocks a Embedded Wind Shocks in single stars b Colliding Wind Shocks in binaries ID: 363559
Download Presentation The PPT/PDF document "X-ray Line Diagnostics of Shocked Outflo..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
X-ray Line Diagnostics of Shocked Outflows in Eta Carinae and Other Massive Stars
M. F. CorcoranSlide2
Overview
Introduction: Problems of Mass and Mass Loss X-ray Fine Analysis as a Probe of Mass LossMass Outflows and Shocks:
a) Embedded Wind Shocks in single stars
b) Colliding Wind Shocks in binaries
Summary and More QuestionsSlide3
Introduction: The Masses of Massive stars
Mass is the fundamental stellar parameter
but as it increases it becomes (observationally) less well constrained
Moffat 1989Slide4
Weight Loss Secrets of the Stars
Mass is lost due toradiatively driven stellar winds
Transfer/Roche Lobe leaks
Eruptions
Explosions…
Meynet & Maeder 2003
Mass Loss: Crucial impact on Evolution Slide5
Problems Measuring Mass Loss
Smooth wind vs. clumped? (Mass-loss rates obtained from P V wind profiles are systematically smaller than those obtained from fits to Hα emission profiles or radio free-free emission by median factors of ~20–130, Fullerton, Massa & Prinja 2006)
spherical or not?
eruption: timescales & rates?
explosion: core & remnant amounts?
Beginning to End…Slide6
X-ray Studies of Mass Loss
All observed OB stars in the range B2V–O2I are X-ray emittersThermal X-rays arise from at least 3 (non-exclusive) shock processes:
embedded wind shocks from unstable line driving
wind-wind collisions in binaries
magnetically-confined wind shocks
X-rays sensitive to detailed mass loss process:
Shocks depend on density and velocity as
f(r)
continuum and lines sensitive to overlying opacitySlide7
Some Observed X-ray Properties
For single OB stars, Lx α 10
–7
L
bol
(known since days of the Einstein Observatory)
Single WR stars very weak X-ray sources
Naze et al.2011 Chandra Carina Project
O stars
B stars
Broadly Speaking:
Emission thermal; kT < 1 keV for single O stars; lots of emission lines
Single stars: X-ray emission non-variable (but see
q
2
Ori A, Schulz et al. 2996, Mitschang et al. 2009)
Binaries: may be harder (kT>2 keV); brighter; variableSlide8
Questions
What is the spatial distribution of the shocked gas in the stellar wind?
What is the temperature distribution vs. radius?
What does the shocked gas tell us about the detailed mass loss process (mass loss rates, velocity laws, unstable regions, large-scale vs. small-scale clumps)Slide9
Tools:
High Energy High Resolution Spectrometry
Stellar wind velocities (1000-3000 km/s) generate distribution of X-ray emitting gas in the Chandra/XMM band (0.2-10 keV)
Shocked gas generates thermal X-ray line emission useful for detailed measures of wind flow & densities
High-energy lines particularly important since inner winds have high optical depth at soft X-ray energies (E<1 keV)
HETG spectroscopy provides a unique tool:
energy band matches wind dynamics
spectrally resolve broad lines, esp. at high energies
spatially resolve clustered starsSlide10
Diagnostics
Process/Property
Diagnostic
Comments/Complications
wind velocity and density profile
Line profile shape
profiles sensitive to velocity, impact parameter, density
/optical depth
ionization process (collisions vs. radiation)
G=(F+I)/R (~1
collisional
;
>4
photoionized
)
resonance line scattering,
optical depth effects
Equilibrium vs. Non-
Eq
.
G > 1
, R> R
o
, Li-like satellites
non-
eq
Abundance
Resonance line ratios of different elements
Uncertainty due to poorly constrained ion fractions
Location
R=F/I
sensitive
both to ambient particle and UV radiation density; blending with satellite lines may be importantElectron TemperatureG ratio, H/He ratiophotoionization; resonance scattering; distinct formation regionsNon-Thermal Processesratio of satellite to resonance linesline broadness
see Porquet, Dubau, Grosso 2011Slide11
a) Embedded Wind Shocks
radiative driving force which generates winds in OB and WR stars unstable to Doppler shadowing (Lucy & White 1980, Feldmeier 1997)Wind should break up into slow dense clumps embedded within a lower density, fast wind
collisions produce shock heated gas dependent on local velocity field
Temperatures typically millions of K since shocks expected to occur in wind acceleration zone close to the star
Opacity of overlying wind beyond the acceleration zone should reduce the red-shifted portion of the line relative to the blue shifted portion, dependent on wind densitySlide12
Zeta vs. ZetaSlide13
Line Profiles
Widths: narrow to broad
in general FWHM << 2x wind velocity
Centroids: zero to negative blueshifts
No redshifts?
Symmetric or Asymmetric?Slide14
Single Star Summary
Lines broad but only 0.2 < HWHM < 0.8 (Gudel & Naze 2009) EM
X
<< EM
wind
Profiles (apparently) symmetric; centroids show small blueshifts (exc. Zeta Pup).
Opacities low; wavelength-dependent?
Lines apparently form rather deep in the wind (R
o
~1.5-1.8 R
*
)
Zeta Pup: opacity wavelength dependent, but profiles require a reduction of a factor of 3 in mass loss rate (Cohen et al. 2010)Slide15
b) Eta Car and WR 140: The Thermal Spectra of Colliding Winds
Long-Period, eccentric colliding wind systems are excellent laboratories for studying:
the development of astrophysical shocks
the physics of line formation
the process of mass loss in more than 2 dimensions
Key properties:
location of X-ray emitting volume constrained to the wind-wind shock boundary
In eccentric systems, variations of density (at constant temperature) around the orbit
clumping-free mass loss rates?
Pittard 2007Slide16
Eta Carinae: LBV+?, P=5.5yr
1820 1850 1880 1910 1940 1970 2000
50 yrs
V-Band Lightcurve
“Great Eruption”
A. Damineli
5.5 yrSlide17
Eta Car: Orbital and Wind Geometry
3D SPH model
(Okazaki et al. 2008)Slide18
Eta Car: X-ray Variations
Sampling with HETGSlide19
MEG Spectral Variations
Normalized here
Hotter than single star
Highly variable
Absorption variations
Emission variationsSlide20
Cycle-to-Cycle ComparisonSlide21
Line Formation
Radial Velocities: lines become more blueshifted near periastron as the shock cone sweeps past the line of sight
Model showing the location of maximum emissivity of the Si XIV line along the shock boundary (Henley et al. 2008)Slide22
F/I ratios
Zeta Pup
Eta CarSlide23
HETG Results
Strong changes in continuum flux and lines
lines of high ionization potential show smaller blueshifts than lines of lower IP
high IP lines form close to stagnation point where electron temperature is higher
variations in centroid velocities of Si & S lines
probably due to changing orientation of bow shock to line of sight
possible transient emission associated with RXTE flares? (Behar et al. 2005)
R=F/I ratio is above the low-density limit
ionization from inner shell of Li-like ion in NIE plasma?
excitation to n>3 levels followed by radiative cascades?
Charge exchange?Slide24
Results: Iron
Broad, Variable Fe K fluorescenceX-ray scattering by wind
Fe XXV “satellite lines” which increase in strength near periastron
cooling (via conduction?) due to the growth of dense cold instabilities
dust formation?
Fe XXV & Fe K profile nearly identical in 2 near periastron observations separated by 1 cycle.Slide25
WR140: Shock Physics Lab
Courtesy P.M. WilliamsSlide26
2000-12-29
(
,D/
a,
)=(1.987,0.23,+44
)
periastron
2006-04-01
(
,D/a,
)=(2.649,1.77,-36
)
apastron
2008-08-22
(
,D/a,
)=(2.951,0.59,+2
)
O-star
Chandra
phase-dependent grating spectra of WR140
WC
O
(2009-01-25)
Courtesy Andy PollockSlide27
WR140 phase-dependent MEG spectra
T~5keV electron continuum 80%
lines 20%
WC abundances
periastron
=1.987
O-star
=2.951
apastron
=2.649
Courtesy Andy PollockSlide28
XUVOIR : WR140 NeX MEG line profiles
periastron
=1.987
O-star
=2.951
apastron
=2.649
Courtesy Andy Pollock
Apastron: view flow from both sides of shock cone; velocity equilibrium?
Periastron: emission from leading arm suppressed – due to changes in cooling?
O-star conjunction: emission from near side of shock cone dominates
Courtesy P.M. WilliamsSlide29
Conclusions & Questions
Other Issues:magnetic fields & collisionless plasmas
close, late-type companions
satellite lines
charge exchange
Embedded shock emission from single stars suggest that most of the shocked gas exists deep in the wind near the wind acceleration region
Mass loss rates need to be reduced
How general is the wavelength dependence of X-ray line opacity?
Importance of radiative instabilities/NEI effects in X-ray line formation in binaries
change in shock physics/cooling near periastron in Eta Car and WR 140
(Very) small R ratios in long-period binaries vs. large R ratios in single stars
Evidence for ionization stratification along the shock boundary in colliding-wind systems
No double-peaked profiles in CWBs: simple conical picture too simple?Slide30
ENDSlide31
X-ray Line Emission: Observed Trends
Declining trend of X-ray ionization with stellar spectral type and weakening of H- to He-like ratios (Walborn et al. 2009)
dwarfs
giants
supergiants
pec.
g
Cas
q
1
Ori C
t
Sco
z
Oph
HD 93250
z
Pup
Walborn et al. 2009Slide32
Summary of Grating Observations
how many massive stars have grating observations? (hetgs, letgs, rgs)how many need these observations?
Relatively small number of stars observed at high resolution, so hard to make firm conclusions regarding trends in line formation properties and connection to winds and stellar properties
Tools:
TGCAT:
XATLAS:http://cxc.cfa.harvard.edu/XATLASSlide33
Dichotomy: 1 vs 2
Single and Binaries: different sources of emission:
Single stars: embedded shocks at some fraction of the wind terminal velocity. Produced by instabilities in radiative wind driving force; “clumping”
Binaries: Shocked gas produced by the collision of the wind of one star and that of its companion; also (I) above
Studies of X-ray emission from single and binary stars provide complementary information regarding stellar mass loss (and more generally the production of X-rays in shocks)Slide34
Examples: Zeta Pup & Zeta Ori
Perhaps the best-studied massive star at high X-ray spectral resolution
Zeta Pup: O4If, N overabundance compared to C, O
Zeta Ori: O9.7Ib
Strong thermal line emission
Mdot ~ 10
-5
M
yr
-1
based on H-α line
Issues:
Lines much less blueshifted and more symmetric than expected given high Mdot
Also high-temperature X-ray lines (S XV) deep in wind; but how deep?Slide35
Zeta Pup Line Analysis
X-ray wind opacity: grey (large, dense clumps) vs. wavelength-dependent (small, optically thin clumps)
Kramer et al. (2003) analysis of HETG spectrum: X-ray optical depth nearly independent of wavelength: large, dense clumps
Cohen et al. (2010) re-analysis, including short-wavelength lines: X-ray optical depth IS dependent on wavelength;
requires ~factor of 3 reduction in Mdot
lines form near 1.5 R
*
Short wavelength lines at high S/N key probes of the inner wind