Download - Danny Steeghs – Tucson 2009 The galactic population of AM CVn binaries Danny Steeghs University of Warwick Tom Marsh (Warwick), G.Roelofs (CfA), G.Nelemans,

Danny Steeghs – Tucson 2009

The galactic population of AM CVn binaries

Danny SteeghsUniversity of Warwick

Tom Marsh (Warwick), G.Roelofs (CfA), G.Nelemans, P.Groot (Nijmegen)

Binsim courtesy R.Hynes


The AM CVn stars

• A white dwarf accreting from a Hydrogen-deficient mass donor star

– They are CVs– They need a 2nd

common envelope


Binary evolution towards AM CVn systems

from Yungelson (2008)

1)2)

3)


Formation channels and Pmin

log(

Mdo

t)

log (binary period) [s]

2 3 4

-10

-5

detached double WD

first contact

AM CVn phase

He-star

evolved CV

1) Double WD channel, Pmin ~2-3 mins2) He-star donor channel, Pmin ~ 10 mins3) Evolved CV donor channel, Pmin ~ 10-60 mins


Mass transfer and stability : merger vs accretion

• Loss of orbital angular momentum via gravitational wave radiation:dJGR/dt = -32/5 G3/c5 M1M2M/a4 Jorb

• Initial stability of mass transfer determined by the response of the mass donor to mass loss:

2 = log R2 / log M2 < 0

• Transfer of momentum via mass loss:dJ2/dt = (GM1Rh)1/2 dM2/dt

• Spin – orbit coupling : non-synchronous rotation leads to angular momentum transport via tidal/magnetic or viscous torques

Courtesy NASA/GSFC


Mass transfer between two white dwarfs at contact

dynamically unstable: mergers

‘ q> 2/3 ‘

Type Ia

M1+m2 > Mch

accretion disk formation ‘stable’

DIRECT IMPACT

accretor

donor

• Post common-envelope detached double white dwarfs driven into contact by gravitational waves (P<10hrs)


Mass transfer between two white dwarfs

Webbink & Iben 1987

Nelemans et al. 2001, Marsh & Steeghs 2002, Marsh, Nelemans & Steeghs 2004

Porb = few minutes

Porb = half an hour

• Semi-detached ‘direct-impact’ birth at P~few mins


Emission line diagnostics

• facilitate their discovery/identification

• sample the abundances of the donor star -> formation channel

• provide a detailed dynamical probe of the emission line regions


Spectroscopic periods

• Complex photometric behaviour of AM CVn stars makes a solid identification of the binary period difficult

• The stream-disk impact localisation provides a common beacon in the binary frame

• Kicked off by fast spectroscopy of AM CVn itself

Accretion disc + bright spot emission(Nelemans, Steeghs & Groot 2001)


The blue variable ES Ceti with Magellan


ES Cet, a 10.3 mins binary

HeII 4686 HeII 5411

Robust orbital periods through fast spectroscopy


Beyond orbital periods ; tracing the accretor

Central spikes in GP Com and V396 Hya

VLT/UVES data, Steeghs et al. 2009


Central spike from the accreting WD

K1 = 11.7 km/s

orb

ital phase

0 1

2


Doppler maps


• Spikes all move in phase and with the same amplitude and flare on short timescales

=> near WD (Marsh 1999)

• However, spike mean velocity changes significantly from line to line (~ 0-50 km/s )

• Some lines show double spikes, each moving together

• Both GP Com and V396 Hya show the same pattern, other AM CVn systems show spikes as well

• Spikes are narrow, though non-Gaussian

• Stark effect in high-density line formation region?

(Morales-Rueda et al. 2003)

Emission spikes from the white dwarf

appears not to be consistent with DB/EHe models for Stark-shift and split (Beauchamp et al. 1997)


Beauchamp et al. (1997) Stark profiles

velocity +100 km/s-100 km/s velocity +100 km/s-100 km/s

Both velocity of peak and overall profile shape depends on ne

ne=1015 cm-3

1016 cm-3


Spikes at the accreting WD• Single density (ne= 3-4 1015 cm-3) Stark profiles describe the spike properties

remarkably well

• Kinematics and density conditions of the spike match a slowly rotating emission line region tied to the accreting white dwarf

• Provide us with the orbital velocity K1, the absolute binary phase and the (Stark corrected) systemic velocity

• Unknown binary space velocity prevents us from converting the overall velocity offset (+16-17 km/s) into a gravitational red-shift


Doppler maps

GP Com HeI 3888 V396 HeI 5015


Mass ratio from disk-stream impact

GP Com V396 Hya

q=M2/M1 q


Mass ratios from spike + impact spot

• GP Comq = 0.019 ± 0.002

• V396 Hyaq = 0.013 ± 0.002

• SDSS J1240 (Roelofs et al. 2005) q = 0.039 ± 0.010

• AM CVn (Roelofs et al. 2006)q = 0.18 ± 0.01

• (SDSS J1552)


Only one eclipser so far …

Marsh et al. ; P=28 min


Period distribution of current AM CVn systems

theoretical minimum for 2 WDs

(~3m)

Hydrogen rich period minimum

young DWDs

high Mdot old evolved systems

low Mdot


Beyond periods : distances• Relatively nearby sample (300 pc scaleheight, closest at ~70pc)

- parallax measurements possible with HST-FGS

- 50 orbit program on five AM CVn systems

GP Comd= 75 ± 2 pc

HP Lib d= 197 ± 13 pc

CR Boo d= 337 ± 40 pc

V803 Cen d= 347 ± 30 pc

AM CVn d= 606 ± 110 pc

Roelofs et al. 2007

- space density estimate of <10-5 pc-3

- some systems must have semi-degenerate donor

(see also Deloye et al. 2007)


Any systems undergoing direct-impact?

• Direct impact accretion first proposed to explain the 9 min variable V407 Vul (Ramsay et al. ; Marsh & Steeghs 2002)

– Short period – Luminous x-ray source with emission pulsed at full amplitude– No polarisation; non-magnetic – Out of phase optical pulsations– No emission lines (?)


The not-so-cooperative; V407 Vul

• 7 hours (53 orbits) of Gemini GMOS spectroscopy of V407 Vul

Steeghs et al. 2006


The record-holder RXJ0806 ; 5.3 mins ?

RXJ0806;

Discovered as a variable ROSAT source, 100% modulation in X-ray intensity every 5.3minutes, same period in the optical/UV.

Twin of V407 Vul, spectroscopy is also challenging (V>21)

Keck LRIS runs: weathered out

three times!Barros et al. (2007)


Gravitational wave beacons

Adapted from

Nelemans 2007

Roelofs et al. 2007

strongest known resolved LISA sources ; verification sources


The AM CVn stars and their discovery method• Ultra-compacts (3)

– HM Cnc / RXJ 0806.3 321s X-ray (ROSAT) – V407 Vul / RXJ1914+24 569s X-ray (ROSAT)– ES Cet 620s blue/UV

• High mass transfer rate systems (2)– AM CVn 1029s blue color (HZ)– HP Lib 1103s blue color (EC)

• Intermediate systems and Helium Dwarf Novae (8)– KL Dra 1502s ‘SN’– CR Boo 1471s blue (PG)– V803 Cen 1596s chance variable– SDSSJ0926 1699s emission lines (SDSS)– CP Eri 1724s blue variable (Luyten)– V406 Hya / SN2003aw 2028s ‘SN’– 2QZ J142701-0123 2194s emission lines (2QZ)– SDSSJ1240 2241s emission lines (SDSS)

• Low mass transfer rate systems (5)– SDSSJ0804 2670s blue color (SDSS) – GP Com 2790s proper motion– V396 Hya / CE-315 3906s proper motion (CE)– SDSSJ1411+4813 2760s emission lines (SDSS)– SDSSJ1552 3376s emission lines (SDSS)

• New, no periods yet (4)– SDSSJ0129 emission lines (SDSS)– SDSSJ2047 emission lines (SDSS)– SDSSJ1208 emission lines (SDSS)– SNF20060524 ‘SN’


Need for better samples

• we need large and well selected samples to determine the formation channels, the mass transfer stability and thus the actual distribution of sources

• for progenitors (detached systems), accretors as well as end-products / post-mergers

• targeted low-galactic latitude survey

-Variability ; RATS, VST

-Colours/lines ; EGAPS

-X-rays ; GBS, cross-matching

• exploit large-scale surveys

-SDSS (many filters, good coverage)

-SkyMapper, Pan-STARRS, LSST


SDSS AM CVn search

SDSS types courtesy J.Girven

• 6 AM CVn systems have been discovered serendipitously in SDSS because spectra were taken

• AM CVn systems occupy a well-defined region in color space that is sparsely populated

• Easy to recognise and distinguish from white dwarfs


SDSS AM CVn search

• N=6 implied space density is 1-3 10-6 pc-3 (Roelofs et al. 2007)

• This is close to the most pessimistic estimates from population synthesis calculations(Nelemans et al. 2001: 6 10-6 – 3 10-5 pc-3 )

• ~50 systems still lurking in this color-color region for a wide range of

Roelofs et al. 2007


Summary

• AM CVn binaries form a crucial population in testing binary evolution (abundant progenitors, 2 common envelope phases, dominate GWR)

• Initial phase of mass transfer after contact is dominated by direct-impact geometry (survival, mergers, explosions?)

• Sample is growing though very inhomogenous

• Fast spectroscopy on large telescopes has delivered orbital periods, mass ratios and an insight into (He) accretion geometries

• SDSS is benchmarking space densities, expectation of significant increase in sample size in the era of surveys


Is the DD channel a significant contributor to Ia’s?

• Type Ia = degenerate Carbon ignition of ~1.4 M object = accreting white dwarf

• Is the companion another degenerate object? (SD versus DD channels)

• Ia’s are observed in both young and old populations, most luminous events are young

• Delay times of few Gyr natural in DD channel, simple SD picture take longer

• Rates seem OK as well, but the std SD is not produced enough in current binary evolution models

• Mergers do not always explode even if M1+M2 > 1.4 M

• Does Type Ia diversity imply multiple channels and can this diversity be linked to the variety of expected DD configurations?


Key locations

Donor star with velocity K2 Accretor with velocity K1 (Ballistic) accretion stream Accretion disk Stream-disk interaction Stream-accretor interaction (direct impact)


Formation of compact WD binaries

• M < 8 Msol evolves into a WD• M_WD ~ 0.6 Msol , R_WD ~ 109 cm (CO)

• Main-sequence binaries with periods of 10-10,000 days evolve into compact binaries involving white dwarfs

• Common-envelope phase required to produce post CE binaries with P< days

• Single CE produces WD + star CV• After two CE sessions left with WD + WD


Rotation

• Spikes are relatively narrow, and shapes are dominated by the Stark broadening profile

• Need very little rotational broadening !

• From HeII 4686 (no Stark):

vsini = 60 – 85 km/s

• White dwarfs appear to be rotating slowly, unless the spike photosphere does not co-rotate with the white dwarf


A multi-wavelength picture

optical linesUV

optical cont?X-rays


Mass ratio from disk-stream impact

ballistic mix disk velocities


Double hot-spots ; stream overflow ?

not sensitive to q ? Roelofs et al. 2005

V396 Hya SDSS J1240 + sn2003aw


Spike velocity shifts

Decent match to the family of spike profiles (both shift and shape) for ne = 3 1015 cm-3


Observed split spike profiles


Rotation

• Spikes are relatively narrow, and shapes are dominated by the Stark broadening profile

• Need very little rotational broadening !

• From HeII 4686 (no Stark):

vsini = 60 – 85 km/s

• White dwarfs appear to be rotating slowly, unless the spike photosphere does not co-rotate with the white dwarf


Recent theoretical developments

• Hydro simulations of initial mass transfer (and merger)

D’Souza et al. 2006, Yoon et al. 2007, Wood et al. 2008

First steps …. At 104 MEdd !


Surviving the direct-impact phase

efficient spin-orbit coupling

• Survival of the initial direct impact phase uncertain

• Crucial for using accreting systems to probe DWD progenitor population


The AM CVn stars

• Classics ; AM CVn and GP Com known for decades

• 5 known


The accretion geometry in ES Cet (10.3m)

constant variable

A disk is present, though containing a strong asymmetry, most likely the stream-disk interaction

No classic direct-impact, but how far does the stream penetrate?

Strong orbital modulations in the UV? (HST/XMM)


Hot donors ; He-star vs hot WD


Growing samples; observational diagnostics

- accretion luminosity in the X-ray/UV/optical- low frequency GWR driving the orbital evolution- hot white dwarfs heated by accretion (optical/UV)- pulsed X-ray sources for direct-impact accretors- low mass and cool donor stars ; He-core or CO WD? (IR)- post-mergers & progenitors

Download - Danny Steeghs – Tucson 2009 The galactic population of AM CVn binaries Danny Steeghs University of Warwick Tom Marsh (Warwick), G.Roelofs (CfA), G.Nelemans,

Top Related