X-raying Hot Massive Stars
Lidia Oskinova
¨
Universita t Potsdam
ESAC 7 April 2011
Massive Stars and Stellar Winds
Initial mass M∗ > 15M⊙
Main Sequence: OB-type
Fast evolution (~Myr) trace
star formation
Hot. T eff > 10 000 K →
high surface brightness
Photon momentum →
acceleration of matter
Radiative acceleration larger
than gravitation →
nasaimages.org supersonic STELLAR WIND
02
The evolution of (very) massive stars
T eff Evolution ← stellar wind (!)
100 50 20 10 5 2 O and B type stars
7
Luminous Blue Variables
6 WR OB LBV
40 M Wolf-Rayet (WR) stars
5 According to dominant spectral lines
Planetary Nebula WN (nitrogen) →
4
log L / L
WC (carbon) →
3 WO (oxygen) → SN
2
Wh
ZA
ite
M
1
Dw
S
arf
0 1M
5 4
log Teff / kK
supernova remnant G292.0+1.8
Massive stars: the cosmic engines 03
Massive stars generate most of the ultraviolet radiation of
galaxies: re-ionization of the Universe was largely due to first
(super)massive stars
Massive stars heat the dust and power infrared luminosities of
galaxies
IR: ESA/Herschel/PACS/SPIRE/J. Fritz, U. Gent; X-ray: ESA/XMM-Newton/EPIC/W. Pietsch; optical: R. Gendler
04
Massive stars: the cosmic engines
Massive stars & their SNe input metals and energy in the ISM
HST: 30 Dor in the LMC
05
Massive stars: the cosmic engines
Massive stars regulate evolution of star clusters
HST: Quintuplet cluster
06
Massive stars: the cosmic engines
Massive stars are progenitors of black holes and neutron stars born
in core-collapse SNe and/or γ-ray bursts
Artist impression of γ-ray burst
07
Massive stars are unique physical laboratories
Nucleosynthesis Stellar interiors and evolution
Interaction between radiation and matter Magnetic fields
Stellar wind hydrodynamics Radiative transfer
08
X-ray astronomy is at the frontiers of observational astrophysics
Eight active missions:
perhaps the most observed band of EM spectrum from space
XMM-Newton 2000
Chandra 1999
Suzaku 2005
09
Multiwavelength approach
IR
optical
UV
X-ray
Modern observational
data - unprecedented
quality.
New level of
sophistication in
modeling and theory is
required to understand
the data.
10
X-ray emission from massive stars: Science objectives
Physics: how X-rays are produced in massive stars?
X-ray spectroscopy is a sensitive probe of stellar winds
X-ray emission is a sensitive probe of stellar feedback
Massive star cluster Westerlund 2: Chandra NASA 11
1. How X-rays are produced in massive stars?
13
OB stars are X-ray active (Einstein observatory 1978)
Hot stars: radiatively driven stellar winds
Supersonic stellar winds are intrinsically unstable
3000
Shocks
2000
Velocity v(r)
1000
Heating
0
X-Rays
Lucy Solomon (1970) ... Feldmeier etal (1997)
Radius r/R∗
Shocks can also result from:
Collision of streams Collision of winds
in magnetically confined wind in binaries
14
Best quality X-ray spectra before year 2001 (ROSAT)
ζ Puppis
ζ Orionis
Energy (keV)
15
High-Resolution X-ray Spectra (XMM-Newton)
FeXVIII
FeXVII
FeXVII
FeXVII
MgXII
OVIII
OVIII
SiXIII
MgXI
NVII
NeX
NeX
NeIX
OVII
NVI
CVI
0.2 XMM RGS ζ Pup
O4I
0.1
0.0
0.3
XMM RGS ζ Ori
0.2 O9.7I
0.1
0.0
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
o
Wavelength (A )
* Overall spectral fitting plasma model, abundances
* Line ratios T X (r), spatial distribution
* Line profiles velocity field, wind opacity
Analyses of the X-ray spectra of O-stars 16
Temperature 0.03 v∞ = 2450 km/s ξ Per
Flux (Counts/sec/A )
O7.5III
o
Range from 2 MK to 10 MK
0.02
Emission line profiles
Broad; width scales with wind speed
Similar accross the spectrum 0.01
Clumped wind (Feldmeier etal. 2003)
OR plasma is not in CIE (Pollock 2007) 0.00
0.12
v∞ = 1550 km/s ζ Oph
Line ratios in He-like ions 0.10
O9V
Flux (Counts/sec/A )
o
Formed close to the photosphere 0.08
Temperature decreases outward
0.06
Abundances
0.04
Agree with wind abundances
0.02
X-rays can be explained by wind shocks (..?)
0.00
~100 papers based only on XMM data: e.g. Kahn etal. 01, 12.0 12.1 12.2
o
Leutenegger etal. 2007, Naze etal. 2010, Raassen etal. 2005, Wavelength (A )
Sana etal. 2004, Rakowski etal. 2006 ...
17
B-
18
Stationary plasma in B-stars
Wind speed is 1500 km/s
But lines are narrow Comparable to
instrumental profile!
He-like ions: f/i line ratio probes
distance to stellar photosphere
OVII α Cru
Capella
Flux (Counts/sec/Angstrom)
X-ray plasma in B-stars
Close to the
photosphere
Stationary
λR λI λF
Different from shocks
in O-type winds
0.0 Pulsations? Coronae?
21.5 21.6 21.7 21.8 21.9 22.0 22.1 22.2
19
Wolf-Rayet type stars
Image courtesy of D.Ducros and ESA
20
X-ray view on single Wolf-Rayet Stars
Not all WR stars emit X-rays.
X-ray spectra of X-ray emitting WR are harder than spectra of
O-stars
Single WR carbon stars are X-ray quiet
X-ray bright WR stars are binaries
ζ Pup (O-type) WR 1 (WN) WR 114 (WR carbon)
Oskinova etal. 2003, Ignace etal. 2004, Skinner etal. 2010, Polock and Corcoran 2006, Gosset etal. 2005 ...
21
Glimpse at the pre core-collapse star WR142 (WO)
-12
3
log F λ [erg s -1 cm -2 A -1 ]
-13
Rel. Flux
9.54
o
8.89 2
He II 5 - 4
8.60
-14
7.63 1
7.23
7.10
-15 6.69
0
10000 10500
o
λ/A
-16
5.35
-17
3.5 4.0 4.5 5.0 5.5
o
log λ / A
Requires state-of-the art non-LTE models to fit observed optical
and UV spectra. Such as PoWR code (Hamann et al. 2006)
T =160 kK, R =0.5R , wind speed v=6000 km/s
* *
Our analysis indicates that star may be a FAST ROTATOR
V rot sin i =4000 km/s. Current mass ~10 M
22
XMM-Newton discovery of X-ray emission from a WO-type star
X-rays are too hard to be explained by wind shocks
Hint on the presence of magnetic field B(r=2R ) > 7 kG
*
0.00300
0.00250
0.00200
0.00150
0.00100
WR142
0.00050
1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0
Oskinova et al. (2009) keV
23
Mystery of X-rays from WR stars: Connection With Collapsars (?)
’’A very energetic explosion of a massive star is likely to create a ... fireball....
the inner core of a massive, rapidly rotating star collapses into a ~10 M Kerr
black hole ... A superstrong ~10 15 G magnetic field is needed to make the
object ... a microquasar. Such events must be vary rare...to account for the ...
GRBs’’
Do we indeed observe in our Galaxy massive, magnetic, rapidly
rotating stars on latest stages of their evolution ?
24
Physics of X-ray emission from massive stars
‘Curiouser and curiouser!’ cried Alice
Carroll (1865)
B-stars: not clear: magnetic fields,
pulsation, winds.
O-stars: more or less clear: winds.
WR-stars: absolutely unclear (First
spectrum : XMM large program 2010)
Massive star cluster Westerlund 2: Chandra NASA 25
2. X-rays diagnostics of stellar wind
26
Image courtesy of D.Ducros and ESA
27
Microclumping vs Macroclumping.
Observations wind is inhomogeneous. Theory density
contrast in the wind
Microclumping: strong assumption -- size is smaller than
the photon free path.
Macroclumping: New "break through" motivated by X-ray
spectroscopy: clumps are realistic, i.e. allowed not to be
optically thin.
Standard situation porosity, e.g dust: Particles are opaque:
radiation cannot go through.
Our work: how does macroclumping affect spectral
analysis.
Macroclumping: τ clump >= 1
28
Microclumping: τ clump << 1
29
The impact of clumping on empirical mass-loss rates
‘‘Macroclumping’’
diagnostic line opt. thick
Porosity effect :
κ eff < κ smooth
Important
consequence for
mass-loss empirical
estimates
˙
Larger M by factor of
a few
Mass-loss: key
parameter to stellar
evolution models &
stellar feedback
31
UV diagnostics: PV resonance doublet
1.5 P V 3p 2 P - 3s 2 S
microclumping
micro- & macroclumping
1.0
Rel. Flux
0.5
0.0
1100 1110 1120 1130 1140
o
λ/A
P V resonance doublet becomes much weaker if macroclumping
is taken into account
This resolves the descrepancy between ˙
M from resonance line
and ρ 2 diagnostics!
32
X-ray diagnostics: X-ray emission lines
Model emission lines
clumped wind ˙
Wind opacity for X-ray drastic-
same M
ally reduced by clumping
0.2
Normalized Flux
Opacity becomes "grey"
0.1
⇓
Similar line profiles accross
smooth
wind the spectrum
0.0
-1.0 -0.5 0.0 0.5 1.0
Normalized Frequency
33
Observed and model lines of ζ Puppis (no fitting!)
SiXIV ζ Pup MgXII ζ Pup NeX ζ Pup
Normalized flux
λ0
λ0 λ0
6.12 6.14 6.16 6.18 6.20 6.22 8.36 8.38 8.40 8.42 8.44 8.46 12.05 12.10 12.15 12.20
FeXVII ζ Pup OVIII ζ Pup NVII λ0 ζ Pup
Normalized flux
λ0 λ0
14.9 15.0 15.1 18.8 18.9 19.0 19.1 24.6 24.7 24.8 24.9
o o o
Wavelength (A ) Wavelength (A ) Wavelength (A )
Oskinova et al. (2006)
34
High-Mass X-ray Binaries as stellar wind probes
Compact object embedded in stellar wind of OBI star
separation ∼ 1R*
Stellar wind accretion on neutron star
high L X , power-law spectrum
X-rays photoionize small part of stellar wind: recombination
X-rays suffer absorption in stellar wind
Vela X-1
Kretschmar etal. 2008, 2009, 2010, Kreykenbohm etal. 2010
35
Fast temporal variability in X-rays - High Mass X-ray Binaries
X-ray light curve: strong variability
Optical donor star O-type supergiant
LX ≈ 1035 erg/s - accreting black hole
36
Accretion in clumped wind
37
Stellar winds and X-ray spectroscopy
Stellar wind is
clumped
Clumps are
optically thick at
some λ
Clumps are
most likely
pancakes!
Stellar mass
loss rate is
quite high
New radiative transfer technique!
Massive star cluster Westerlund 2: Chandra NASA 38
3. X-rays diagnostics of massive star feedback
38a
S308 wind blown bubble around WR6, DSS
38b
top: XMM-Newton image (Chu etal. 2003)
40
X-ray observations help to understand feedback
Spitzer, HST, CXO, LMC 30 Dor (Townsley+’06)
Evolution of L X from cluster wind
Quintuplet
NGC3603
NGC2100
NGC1818
NGC1850
NGC1860
NGC1831
6
Arches
R136
4
log( LX [L⊙ ] )
2
Solar Z
LMC Z
0
stellar winds supernovae
-2
6.0 6.4 6.8 7.2 7.6 8.0 8.4
log ( Time [yr] )
Oskinova 2005
Spatial correlation of YSO and diffuse X-ray emission Chemical
gradients Evolution of kinetic energy input X-ray dating of low-mass
stars (perhaps high-mass, too? )
41
Cosmic archaeology
NGC 602 a massive star cluster (HST image)
Example of triggered secondary star formation with a large yield
X-rays trace hot plasma: how it is connected with star formation?
NGC 602 is at the edge of a SUPERGIANT SHELL. Largest
structures in the interstellar medium
41a
Supergiant shell in the SMC
41b
Part of the supergiant shell in the X-rays
42
From massive stars to structuring galaxies
Supergiant shells are formed by massive star
feedback?
They provide chimneys for hot gas to escape to
intergalactic space
X-ray trace this hot gas and stellar feedback
43
X-ray emission from massive stars: Summary
Physics: how X-rays are produced in massive stars?
Non-stationary processes in stellar winds: shocks, magnetic
fields...
X-ray spectroscopy is a sensitive probe of stellar winds
Mass- loss from massive stars is prodigious 10−4...−7 M⊙ /yr
Poorely known: standard methods need to be improved
X-rays: Winds are not stationary and not homogeneous -
clumping
X-ray emission is a sensitive probe of stellar feedback
Massive stars strongly affect the ISM by radiative (UV photons)
and mechanic (winds) energy input.
How kinetic energy feedback affects the ISM and star
formation ?