POST COMMON ENVELOPE BINARIES

1
POST COMMON ENVELOPE BINARIES FROM THE SLOAN
DIGITAL SKY SURVEY
Alberto Rebassa-Mansergas Supervisor Dr. Boris
Gaensicke Co-supervisor Dr. Pablo
Rodríguez-Gil Working with Dr. Linda
Schmidtobreick Dr. Matthias Schreiber
2
INTRODUCTION
PCEBs wide MS binaries CE phase Friction
within the envelope leads to a rapid decrease of
the binary orbit E and J extracted from the orbit
ejects the CE WDMS binaries WD MS (no CE)
3

• PCEBs are the progenitors of the following
  fascinating systems
• - double degenerates
• - gamma ray bursts
• - super soft sources
• - black-hole candidates
• - CVs and X-Ray binaries
• - milisecond pulsars

4
18 PCEBs identified from RVs

• Population models are available
  clear lack of observational
• 
  constraints
• We need to establish a large sample of one type
  of close compact binaries
• PCEBs consisting of a WD and a MS are the best
  systems because they are
• - numerous (population studies are feasible)
• - well understood in terms of single star
  evolution
• - nearby and easily accessible with 2-8m
  telescopes
• - no mass transfer systems
• SDSS 1500 WDMS

5
IDENTIFYING PCEBs IN THE SDSS
WD clearly visible in the blue The MS
dominates the red
6
IDENTIFYING PCEBs IN THE SDSS
WD clearly visible in the blue The MS dominates
the red Na ?? 8183.27,8194.81 doublet Ha
emission (if present)
10 of the spectroscopic SDSS objects are
observed more than once RV variations will
identify such a system as a strong PCEB candidate
7
SDSSJ02460041 display an extremely large radial
velocity variation Ha emission
Gaussian
parabola Na doublet
double-Gaussian fixed
separation
parabola
8

• 18 strong PCEB candidates imply 15 in our
  WDMS sample. However
• In most cases only two spectra are available
• The low spectral resolution of SDSS limit the
  detection of significant radial
• velocity change to 10 km/s
• Na doublet will smear in binaries with extremely
  short orbital periods
• PCEB fraction among the SDSS WDMS might be higher
  than predicted,
• probably in agreement with the 20 obtained by
  the population models
• Follow-up with higher spectral resolution will be
  necessary to confirm
• this hypothesis

9
STELLAR PARAMETERS
Decompose the WDMS into
its WD and MS components M-dwarf
templates, a grid of observed WD templates and a
grid of WD model spectra
Two steps (1) Fit the WDMS spectra model
M-dwarf Sp the flux scaling factor between
the M star template and the observed spectrum
10
(2) M-dwarf template subtracted Residual line
profiles in the WD fitted with the grid of WD
models WD Teff and log(g) the flux
scaling factor between the WD model and the WD
observed spectrum WD
Mass from Bergeron et al's (1995) tables
Teff and log(g) obtained from the fit to the
whole spectrum were select the hot or cold
solutions from the line profiles
11

• Histograms consistent in broad terms
• with other authors
• WD mass peaks at 0.6 solar masses
• The most common Sp are M3-M4
• The most frequent Teff are between
• 10000-20000 k.
• - log(g) peaks at log(g) 8

12
Distances estimated from the best-fit flux
scaling factors of the two spectral
components For the WD For the M-star
It is necessary to assume a radius for the
secondary star This requires a spectral
type-radius relation for M stars Problem! Lack
of observational work Compile Sp and R from the
literature empirical Sp-R relation
for M stars
13
Average relation irrespective of - ages -
metallicities - activity levels

• The Sp-R relation is compared to
• Theoretical models
• Directly measured radii from
• eclipsing binaries and
• interferometry
• Directly measured radii from
• eclipsing WDMS binaries
• (RR Cae, NN Ser, DE CVn,
• RXJ2130.64710,
• EC 13471-1258)

14

• 2/3 of the systems have d(sec) d(wd) within
  their errors. However, there is a clear
• trend for outliers where d(sec) gt d(wd)
• Systematic problems in the WD fits?
• A relationship with close binarity?
• Problems in determining the Sp of the secondary
  star?
• Problems in the Sp-R relation?

15
Could magnetic activity affect the Spectral type
of the secondary? We assume that the secondary
star appears hotter that it should for its given
mass. This implies a change of 1-2 Sp subclasses,
and hence a change in the Teff and the radius.
16
Could magnetic activity affect the Spectral type
of the secondary? We assume that the secondary
star appears hotter that it should for its given
mass. This implies a change of 1-2 Sp subclasses,
and hence a change in the Teff and the radius.
17
CONCLUSIONS

• We have identified 18 PCEBs and PCEB candidates
  among a sample of 101
• WDMS for which repeat SDSS spectroscopic
  observations are available.
• From the SDSS spectra we determine Sp of the
  companions, Teff, M, log(g) of
• the WDs, as well as distance estimates to the
  systems. Even though some of
• the stellar parameters obtained from our
  decomposing/fitting technique differ
• from those obtained from other authors, our
  results agree in broad terms.
• In about 1/3 of the WDMS studied, the SDSS
  spectra suggest that the secondaries
• have Sp types too early for their masses. This
  behaviour could be explained by
• magnetic activity if covering a significant
  fraction of the star by cool dark spots will
• raise the temperature of the inner spots
  regions.
• The fraction of PCEBs among the WDMS population
  is 15, However, our data
• suggest a higher fraction, probably in
  agreement with the results obtained from
• population models.

18
SUPPORTING MATERIAL
19
Ha emission radial velocities
Na doublet radial velocities.
20
ORBITAL PERIODS OF THE PCEBs
We assume i 90 degrees and also that the radial
velocities sample the maximum quadrature of the
radial velocity amplitud. Thus we get absolute
maximum periods of the PCEBs, which range
between 0.46d 7880d. The actual periods are
likely to be susbtancially shorter, especially
for those systems where only two SDSS spectra are
available and the phase sampling is
correspondingly poor.
21
Comparison with Raymond et al. (2003)
22
Comparison with Silvestry et al. (2006)
23
(No Transcript)
24
(No Transcript)
25

• 2/3 of the systems have d(sec) d(wd) within
  their errors. However, there is a clear
• trend for outliers where d(sec) gt d(wd). We
  considered
• Systematic problems in the WD fits?
• A relationship with close binarity?

26

• Problems in determining the Sp of the secondary
  star?
• Problems in the Sp-R relation?

27
For Sp later than M3 the theoretical Sp-R
relation is not sufficient enough to shift the
outliers. For Sp earlier than 2.5 the theoretical
relation exacerbates the d(sec) gt d(wd) problem.