Uli Heber Subluminous O stars Origin and evolutionary links Hydrogen-Deficient Stars, Tübingen 20.9.2007.

Uli Heber

Subluminous O starsOrigin and evolutionary links

Hydrogen-Deficient Stars, Tübingen 20.9.2007

Outline Early results Atmospheric parameters Evolutionary scenarios

- Close binary evolution (RLOF, CEE & WD mergers)

vs

- Delayed core helium flashers

- (non-core helium-burning stars)

Kinematics Summary & Outlook

sdO vs. sdB stars

sdO

sdB

sdB stars: - helium-deficient- „cool“: 20-40kK

sdO stars:- H-deficient- Hot > 40kK- He-sdOs: No hydrogen

Subluminous O and B stars

Greenstein & Sargent (1974)

sdB stars:He-deficiency from diffusion

Metal abundances HST/STIS UV spectra

- Enrichment of heavy elements (>100 times) except Fe

- Radiative levitation

Fe

PbO´Toole & Heber 2006

CPD-64 461 207 207PB/208PB =solar

Post-EHB vs post-AGB evolution

sdB = Extended Horizontal Branch stars:

- He-burning core & inert H-envelope (<0.01 Msun)

- How to loose the envelope?

- sdO stars=post-EHB?

Post-AGB Objects:-rare-linked to RCrB/EHe stars

sdB sdO

Convective transformation(Wesemael et al. 1982)

Groth et al.(1985):Convection occursin He-rich atmo-spheres only

Convection He/H=1

sdO

sdB

Hot subdwarfs from UVX SurveysLTE- spectroscopic analyses of sdB stars:

- Palomar Green survey: Saffer et al. 1994, Maxted et al. 2000

- Hamburg Quasar Survey: Edelmann et al. 2003

- ESO-Supernova Progenitor Survey (SPY): Lisker et al. 2005

NLTE spectroscopic analyses of sdO-stars:

- SPY (Ströer et al. 2007)

- Sloan Digital Sky Survey (Hirsch et al. 2007)

atmospheric parameters for >200 sdB

atmospheric parameters for 130 sdO

Fits of UVES-spectra (SPY): sdOhigh resolution spectra, UVES@VLT, TMAP NLTE models, H&He only

sdO He sdO

Fits of SDSS-spectra: sdO

sdO He sdO

SPY:C & N lines

solar

solar

- C&N strong: diamonds- C strong: triangle- N strong- no C or N: open - All He-rich sdOs

have C and/or N

-None of the He-poor have C/N

Carbon III/IV

SPY: n(C)=0.13%(Hirsch et al. 2007)

vrot sin i=0 km/s

Carbon III/IV

SPY: n(C)=0.25%(Hirsch et al. 2007)

vrot sin i=20 km/s

solar

SPY

He-poor He-rich

The canonical picture

He-ZAMS

Smooth evolutionary time scales:- He-poor scattered in diagram progeny of sdB stars- Clumping of He-rich sdOs can not be explained

SPY&SDSS: sdB, sdO & He-sdO

sdO stars:

He-sdO: clumping at- Teff = 45000K- log g = 5.8

sdB

He-sdOclump

SPY-sds: without error bars

Post EHB

EHB

He-ZAMS

Sub-He-ZAMS

SPY&SDSS: sdB, sdO & He-sdO

Hot He flashers

Delayed He core flash

Canonical evolutionSweigart, 1987

Core flash

Core flash

Delayed helium shell flash

He sdO

Very late helium core flash

He sdO He/C

Could explain He-sdOs below the Helium ZAMS

sds in binaries

mostly single-lined: RV curve:mass function

2

33

2 )MM(

)i(sinM

G

PK)M(f

invisvis

vis

SPY: fraction of close binaries: radial velocity variables with P<10dsdBs; :40% (Napiwotzki et al.,2005)

Minimum mass of companion

Napiwotzki et al. 2007

sdOs: 4% RVV (from SPY)

Period distributionNature of companions: white dwarf or low mass m.s. stars

WD

MS

unknown

Morales-Rueda (2006)

Binary Population Synthesis (BPS)

Han et al. (2003)

a: 1. CE ejection

b: 1. stable RLOF

c: 2. CE ejection

d: merger of two

helium white dwarfs

Comparison to Han et al. (HPMM)

sdBs: best match: models with

correlated masses and low CEE efficiency

Poor match: models with 100% CEE efficiency

O-types: He-rich sdOs: stars clump at

45000K, too hot for any HPMM simulation set

He-poor sdO: scattered in (Teff, log g) diagram

Ströer et al. 2007

Non core helium-burning evolution

Castellani, Castellani & Moroni (2006)

M=0.8 Msunη=0.75

Star leaves RGB Before helium ignites in the core (e.g. by mass tranfer to a companion)

Cooling tracks to formhelium white dwarfs

Non-core helium-burning sdB starsHD 188112 (V=10.2) (Heber et al., 2001)

- Hipparcos parallax

- distance = 80 pc

- mass = 0.22 Msun

No helium burning

- companion: M>0.72Msun

Tracks: Driebe et al.

A Hyper-velocity star (HVS) amongst sdO stars from SDSS

HVS

-500 0 +500 km/s

Galactic restframe velocity

SMBH Slingshot

Hills (1988): Disruption of a

binary near a Super-Massive Black Hole

releases companion at up to

1000 km/s or more. Detection of a

single HVS: evidence for a

SMBH

Gualandris et al. (2005)

Summary & Conclusion Origin of sdB/sdO stars? (i) delayed core helium flash (ii) close binary evolution (RLOF & CEE ejection), mergers of He-WDsHe-poor sdOs are the progeny of sdB starsHe-rich sdO stars are hotter than predicted by (i) & (ii) atmospheres: No metal line blanketing metalicity effects evolution (Brown et al. 2007)Post-AGB-evolution & Non-core He-burning evolution: rare due to short evolutionary time scales

Outlook: A pulsating sdO star

Strongest mode: P=119.3 s A=38.6 mmag plus- First Harmonic plus- 8 modes: 62 ... 118s

Woudt et al. (2001)

Stellar & Envelope Masses

Masses: 0.45 to 0.55 Msun Envelope masses: 10-3.... 10-5 Msun

sdB

Thank You!

sdB Asteroseismology

Multi-periodic light variations (few mmag) at periods from 2 to 10min.

Østensen et al. (2001)

Carbon and Nitrogen

SPY: C and/or N linesDetected - in all helium-rich- In none of the helium-poor ones(Ströer et al. 2007)

Carbon abundances

Challenges Observations: better statistics, better data: the quest for high resolution. metal abundances (see Poster 25) Evolution theory: Prediction of surface abundances for late hot

flasher (Cassisi et al. 2003) & He WD mergers Angular momentum and stellar rotation Stellar atmospheres & envelopes: diffusion (rad. levitation) &

metal line blanketing, see talk by G. Michaud Mass loss and diffusion The role of magnetic fields (O´Toole et al. 2005)

Grazie!

Blue Hook stars

HD128220B: Fe & Ni

Fe/H=1/100solar

Ni/H=1/10 solar

US 708: Keck LRIS spectrum

• Teff = 45500K,

• log g = 5.23,

• mass = 0.5 Mo

• B=19.0 mag

• Distance: 19 kpc

Run-away stars

Ejection scenario:

born in the plane and ejected (Blaauw, 1961)

- binary supernova ejection

- 3 body interaction in an open cluster Calculate path and time of flight:

- radial velocities, distances & proper motion

- orbit integrator: Odenkirchen & Brosche (1992)

- Galactic potential: Allen & Santillan (1991)

BD+75 325 (Lanz et al. 1997)

- Slight enrichment of Fe&Ni

- fully metal line blanketed models:

Teff lower by 6000K than metal free models

Metallicity effects on atmospheric parameters for the sdB SB 707

Solar ([m/H]=0.0):Teff = 33940Klog g= 5.82log He/H=-2.95

10*solar ([m/H]=+1.0) :Teff = 35380Klog g= 5.90log He/H=-2.91

Metal line blanketed LTE models

Summary II

Heavy metals in sdO and sdB stars: Non solar abundances of Fe & Ni in sdO stars Non solar Ni/Fe (>solar) Strong enrichment of many iron group elements in hot

sdB stars (except Fe), about solar in “cool” sdBs (<30000K):

FUV flux suppression UV upturn Teff scale significantly changed by supersolar metal

abundances (line blanketing)

Outlook: Radial velocities

Vrad=700Km/s

Hypervelocity star

Cosmic accelerator?

Ejection from a cluster by three body interaction?

SN II in a binary release companion at orbital velocity?

Supermassive black hole in the Galactic center?

Better ideas??

HQS-sdB: comparison with Han et al.

Trends of helium abundance

sdB

sdO

He sdO

solar

sdB stars: - 2 sequences

sdO stars: - Spread by 6 orders of magnitude - 1/3 helium- deficient!

sdB Helium abundances

Edelmann et al. 2003

Two sequences:He/H vs. Teff

Hamburger Quasar Survey

sdB stars:

Edelmann et al.

2003:

100 sdB stars

sdB = Extreme HB stars

Saffer et al. 1994

EHBPost-EHB

The lower sequence

Tracks from Driebe et al. (1998)

Mcore

sdB and sdO stars from SPYSPY: ESO-VLT+UVES:High-res. Spectra of >1000 Double degenerate

candidates- sdB: 79 (Lisker et al. 2005)

- He-sdO: 30 (Ströer et al. sdO: 28 2007)

- fraction of RV variables (P<10d):

sdB: 39% He-sdO: 4% (1 SB2 binary)

sdB

Trends and Sequences

Combining all studies

Neglecting

selection bias

SDSS sdBs:

To be donesdB

Gap?

SPY-sds: no error bars shown

BPS

Han et al:

Binary population synthesis

a) Without GK selection

b) With GK selection

M 15

UV

Post EHB & post-AGB evolution

Post-AGBPost EHB

UV spectroscopy of HB stars

Caloi, Castellani et al. 1986 Heber et al. 1986

IUE

SDSS-sdOs

Atmospheric models:

- NLTE:

- H+He, no metals

- PRO2 code

(Dreizler &Werner)

- improved He atomic models

- temperature correction scheme (Dreizler, 2003)

sdOHe sdO

Globular Cluster CMDs

Moehler (2000) NGC 2808 (Walker , 1999)

Blue hoo

k

NGC 6752: HB &EHB starsMoni-Bidin et al. (2007)

LTE spectral analyses: Teff, logg g match (E)HB prediction

Helium subsolar

EHB Models

Castellani et al. 1994

Helium core mass: 0.47 Msun depen-ding on He and metal abundance(fixed by onset of He core flash)

Horizontal Branch=sequence of envelope mass Menv,

EHB=very low Menv (0.01 Msun), inert H-rich envelope

avoids AGB evolutions

Origin of EHB stars

Castellani & Castellani 1993

M=0.8 Msunη=0.75

EHB-progenitor stars must loosealmost their entire envelope bythe time of the helium core flash strong RGB mass loss;

Low mass stars (Pop. II, globular cluster): Very efficient RGB Reimers windmay be sufficient.

Younger populations, i.e. more massive progenitors (field): ?

KPD 1930+2752: sdB + massive WD

Billeres et al, 2000

Maxted et al. 2000, Geier et al. 2007

Candidate SN Ia ProgenitorKPD 1930+2752:Total mass=1.4 Msun(Chandrasekhar mass)

-Double degenerate-System merges within 2 108 years-SN Ia explosion?(Geier et al. 2007)

More on massive compaions: talk by Stephan Geier

sdB Asteroseismology

Non-radial p-modePulsationa driven by Iron opacity bump:

Predicted instabilityStrip matches Observations

Charpinet et al. (2001)

sdBAsteroseismology

Period matching technique:

Linear theory:Amplitudes can not be predicted

(PG1325+, Charpinet et al. 2006)

Metal abundances: Fe & Ni

Feige 34:

He-poor sdO

Teff=60kK

Fe/H=10*solarNi/

H=70*solar

sdB Asteroseismology

(PG1325, Charpinet et al. 2006)

Model parameters: Teff, log g, Mtotal, Menv

sdB Asteroseismology

PG 1605+072: Time resolved spectroscopy (9000 spectra)Radial velocity variations (O´Toole et al. 2005): 20 periods (few km/s)Line profile variations (phase folded, Tillich et al. 2007):

sdB Asteroseismology

Dominant mode:Teff semi-amplitude: 800KLog gsemi-amplitude: 0.08

First harmonic detected

Cleaning for dominant mode:8 weaker modesdetected

sdO stars from SDSS

candidates selected from

all releases according to

colour: u-g<0.2 (0.4)

g-r<0.1 11000 spectra: 40 sdO + 43 He sdO

(Hirsch, Dipl. Thesis)

Fits with NLTE models

He sdO

The two sequences

The two sequences

Tracks from Dorman et al. (2003) with Z=0.02

The upper sequence

Tracks from Dorman et al. (2003)

The lower sequence

Tracks from Dorman et al. (2003)

Early NLTE Analyses : sdO

Classification: He II > He I

He-sdO: no Balmer detectable to the eyeC and/or N strong

sdO: otherwise

Hunger et al. 1980Heber (1987)

Post-EHB

Post-AGB

Evolution of hot subluminous stars: the canonical picture

SdB + sdO stars: Extreme Horizontal Branch stars

EHBHB

sdBsdO

Dorman et al. (1993, ApJ 419, 596)

He-rich sdOs:

- diamonds: C&N strong

- C strong

triangles

- N strong

- (triangles)

The lower sequence

Tracks from Driebe et al. (1998)

Mcore