Spectroscopic Binaries in Planetary Nebulae

1
Spectroscopic Binaries in Planetary Nebulae

• Howard E. Bond
• Space Telescope Science Institute

2
Three Arguments that Many PNe are Ejected from
Binary Stars

• Large majority of PNe have highly non-spherical
  or bipolar shapes
• Simplest explanation PN ejection through
  common-envelope (CE) interactions,
• or at least the ejection process is strongly
  influenced by companions (tidal spin-up, dynamo
  generation of magnetic field)

3
Three Arguments

• An observed high incidence of very close binary
  PNNi
• Photometric monitoring shows that 10 of PNNi
  are binaries with periods of a few hours to a few
  days (Bond Grauer 1980s Bond Livio 1990
  Bond 2000)
• These close systems must have emerged from
  common-envelope interactions

4
Common-envelope Interactions

• Occur in binaries initially wide enough for one
  component to become RG or AGB star before
  interacting with MS companion
• Companion is engulfed, spirals in, and may eject
  the envelope, leaving a much closer binary MS
  star hot core.
• Hot core can ionize ejected envelope, producing a
  PN around the close binary

5
Common-envelope Interactions

• Final orbital period depends on efficiency with
  which orbital energy goes into ejecting material
  from the system
• denoted ?CE
• High ?CE results in long final period
• Low ?CE results in short period, or merger
• More discussion will be in following talk (De
  Marco)

6
Three Arguments

• The 10 of very close pairs may just be the
  short-period tail of a much larger fraction of
  binary PNNi
• Population-synthesis studies suggest this is true
  for a wide range of ?CE values
• e.g., Yungelson et al. (1993), next slide

7
Yungelson et al. (1993) Population Synthesis
Predicted orbital period distribution of binary
PNNi
8
Yungelson et al. (1993) Population Synthesis
Predicted orbital period distribution of binary
PNNi
Post-CE systems
9
Yungelson et al. (1993) Population Synthesis
Predicted orbital period distribution of binary
PNNi
Gap due to binaries moving to shorter P
Post-CE systems
10
Yungelson et al. (1993) Population Synthesis
Predicted orbital period distribution of binary
PNNi
Gap due to binaries moving to shorter P
Wide binaries that never interact
Post-CE systems
11
Yungelson et al. (1993) Population Synthesis
Predicted orbital period distribution of binary
PNNi
Gap due to binaries moving to shorter P
Wide binaries that never interact
Post-CE systems
A digression on the next slide
12
A digression

• Ciardullo, Bond, et al. (1999 AJ 118, 488)
  carried out an HST snapshot survey for the
  expected population of visual binaries in PNe
• We found 10 likely and 6 possible resolved
  binaries, out of 113 examined.

13
HST WFPC2 images
These wide companions are useful for deriving
distances to the PNe (MS fitting), but are
unlikely to affect PN ejection
14
Outcome of CE interaction depends on ?CE
High ?CE ? long periods
Low ?CE ? short periods mergers
Yungelson et al. 1993
15
but only short-period systems can be found
photometrically
Systems that can be found photometrically

16
The photometric search method depends on heating
effects in close binaries, so
Artist Dana Berry
17
the short-period systems could just be the tip
of an iceberg of longer-period binaries
18
the short-period systems could just be the tip
of an iceberg of longer-period binaries
Short periods
LONG PERIODS
19
and if so, the total fraction of binary central
stars is very high, and PNe are fundamentally a
binary-star phenomenon!

20
Testing the iceberg hypothesis

• Requires radial-velocity (RV) measurements, in
  order to find wider binaries that lack heating
  effects
• Results to be reported here
• De Marco, Bond, Harmer, Fleming WIYN 3.5m
  telescope at Kitt Peak
• Afsar Bond SMARTS 1.5m telescope at Cerro
  Tololo

21
WIYN 3.5-m Program

• Hydra spectrograph, 2002-04.
• Dispersion 0.33 A/pix resolution 7500
• RV precision 3-3.7 km/s
• Scheduling optimized for periods of days to
  months
• Results of 2002-03 observations were reported by
  De Marco, Bond, Harmer, Fleming (ApJ
  602,L93,2004)

22
RV variations are often unequivocal
?RV 33 km/s in 21 days
?RV 27 km/s in 21 days
23
WIYN result 10 out of 11 PNNi have variable RVs!
Probability RV is variable
24
SMARTS 1.5-m Program

• Cassegrain spectrograph, 2003-04.
• Dispersion 0.77 A/pix resolution 2000
• RV precision 10 km/s
• Scheduling (as with WIYN) optimized for periods
  of days to months
• Part of PhD thesis of Melike Afsar, Ege
  University (Turkey) STScI

25
SMARTS again many RV variables
P gt 0.99
26
SMARTS again many RV variables
Found var. in WIYN pgm.
27
RV survey results

• WIYN (?3.5 km/s) 10 out of 11 PNNi variable
• SMARTS (? 10 km/s) 7 out of 19
• Sorensen Pollacco (Asymm PNe III, 2003) (? 5
  km/s?) found 13 out of 33 PNNi have variable RVs
  (incl. NGC 6891, also found by WIYN SMARTS)

28
Caveats

• Measurements are difficult in some PNNi due to
  few suitable absorption lines (free of nebular
  contamination, etc) and low RV amplitudes
• Wind variations could produce line-profile
  variations that mimic RV variability
• But we tried to select against PNNi with strong
  UV P Cygni profiles

29
Searching for periods

• Finding orbital periods would strengthen the case
  for binarity, but we have not been able to fit a
  binary period to any of our objects

30
with one possible exception
f(m) 0.006, implying m2gt0.13Msun if m10.6Msun
31
Searching for periods

• Finding orbital periods would strengthen the case
  for binarity, but we have not been able to fit a
  binary period to any of our objectswith the
  possible exception of IC 4593
• But our observations cover 2-3 years sparsely,
  which is highly non-optimum for finding periods
  that now appear to be short (few days)

32
What next?

• We need an intensive campaign on a few objects
  with a large telescope, high resolution, and high
  S/N
• in order to distinguish between binary orbital
  motion and wind-profile variations
• Bond De Marco had successful 5-night run in May
  2005 with Kitt Peak 4m echelle spectrograph.
• Concentrated on IC 4593, BD33 2642, LS IV-12
  111, NGC 6210
• Stay tuned for results!

33
Other implications of binaries in PNe

• May explain existence of PNe in globular clusters
• Post-AGB remnants of low-mass stars evolve too
  slowly to produce ionized PNe
• But binaries can merge or transfer matter before
  the PN stage, producing higher-mass remnants
• PNe at the bright end of the PNLF may be
  descended from binaries (Ciardullo poster)

34
The PN in M15 (HST)
35
Other implications of binaries in PNe, contd.

• Most of the classes of compact binaries are
  probably descended from binary PNNi via common
  envelopes
• Pre-cataclysmic red-dwarf/white-dwarf binaries
  like V471 Tau
• Cataclysmic variables
• Low-mass X-ray binaries
• SN Ia progenitors

36
Other implications of binaries in PNe, contd.

• Knowing the overall orbital period distribution
  of PNNi would help constrain the typical value of
  ?CE, which is needed in population-synthesis
  calculations.

37
Summary

• 10 of PNNi are very short-period binaries
  (hours to a few days) that must have been ejected
  from CE interactions
• Resolved visual binary PNNi occur about as often
  as expected
• RV observations are now suggesting that a large
  population of spectroscopic binaries exists among
  PNNi, making the total binary fraction very high

38
Summary contd.

• However, additional spectroscopic observations
  with large telescopes are needed to verify the
  suspected high spectroscopic binary fraction
• At present, it appears very plausible that
  binary-star ejection is a major formation channel
  for planetary nebulae

39
Thanks to collaborators!

• Orsola De Marco
• Di Harmer
• Andrew Fleming
• Melike Afsar
• Robin Ciardullo
• Al Grauer
• Telescope operators, funding agencies, SMARTS
  Consortium, STScI DDRF!

40
The Hubble Heritage Program
And now for a commercial announcement

• Founded in 1998 to bring the most compelling
  Hubble images to the public
• Main criterion is pictorial beauty, with
  scientific interest also considered
• Images are taken from archive, sometimes
  supplemented by new observations obtained by the
  Heritage team through Directors Discretionary
  time, and processed for release

41
The Hubble Heritage Program

• Some observations are entirely new images
  obtained by Heritage team
• Prizes honors
• Images on US British postage stamps
• 2003 Klumpke-Roberts Award of the Astronomical
  Society of the Pacific for contributions to
  public appreciation of astronomy

42
The Hubble Heritage Gallery

• Visit our website http//heritage.stsci.edu