$30 off During Our Annual Pro Sale. View Details »

What's Shaking with Extremely Low-Mass White Dwarfs

jjhermes
November 09, 2017

What's Shaking with Extremely Low-Mass White Dwarfs

Colloquium, 45 min. November 2017: Joint NRAO and University of Virginia Colloquium, Charlottesville, VA.

jjhermes

November 09, 2017
Tweet

More Decks by jjhermes

Other Decks in Science

Transcript

  1. http://jjherm.es
    J.J. Hermes
    Hubble Fellow
    University of North Carolina
    at Chapel Hill
    What's shaking with extremely
    low-mass white dwarfs

    View Slide

  2. View Slide

  3. All stars like the Sun eventually
    run out of fuel.
    When they do, they lose all their
    envelope and only the core remains:
    a white dwarf star.

    View Slide

  4. View Slide

  5. Tremblay et al. 2016, MNRAS, 461, 2100
    Hydrogen Balmer-line fits
    to low-resolution spectra
    allow us to estimate Teff
    and
    log(g), and thus mass
    CO-Core WDs ONe-Core WDs
    Only
    isolated
    WDs
    Most isolated white dwarfs
    ~0.62 M¤

    View Slide

  6. Bloemen et al. 2012, MNRAS, 422, 2600
    2.2 ± 0.2 M¤
    , 9500 K, 2.14 ± 0.08 R¤
    A star primary
    + 0.228 ± 0.014 M¤
    , 14500 K, 0.044 ± 0.002 R¤
    WD
    Porb
    = 5.188675 days, i = 87.0 ± 0.4 deg
    KOI-74

    View Slide

  7. View Slide

  8. The Galaxy is not old enough
    for a single star to evolve into a
    <0.35 M¤
    WD
    “Low-mass white
    dwarfs need friends”
    (Marsh et al. 1995, MNRAS,
    275, 828)
    David A. Aguilar, CfA
    0.228 ± 0.014 M¤
    , 14500 K, 0.044 ± 0.002 R¤
    WD
    KOI-74

    View Slide

  9. van Kerkwijk et al. 2015, ASPC, 328, 357
    Low-mass WDs first inferred in timing residuals around pulsars
    assuming i=60, MMSP
    =1.4 M¤

    View Slide

  10. van Kerkwijk et al. 2015, ASPC, 328, 357
    Low-mass WDs first inferred in timing residuals around pulsars
    only those with optical constraints

    View Slide

  11. van Kerkwijk et al. 2011, ApJ, 728, 95
    Not all low-mass white dwarfs have “friends” – many are being
    actively ablated from a nearby MSP: Black Widow Pulsars
    PSR B1957+20
    (Bow shock, artist’s
    rendition)

    View Slide

  12. Motivation and Outline
    • How well can we determine mass of a low-mass white dwarf?
    – Used to constrain pulsar masses when in MSP+WD binaries
    • What is the cooling age of a low-mass white dwarf?
    – WD cooling ages often compared to MSP spin-down ages
    – Observational constraints on H-envelope thickness and thus cooling rates
    • How does the galactic gravitational wave foreground sound?
    – The mHz regime it is awash in persistent sources from close WD+WD binaries
    D. Berry, GSFC

    View Slide

  13. Bassa et al. 2006, A&A, 456, 295
    PSR J1911–5958B: V = 22.1 mag
    Pulsar opticalcounterparts
    are usually very faint!
    8.2m VLT

    View Slide

  14. log(g)=
    log(g)=
    log(g)=
    log(g)=8
    “The ELM Survey”
    An SDSS color-selected search for new extremely low-mass
    (ELM, <0.3 M¤
    ) white dwarfs
    Orbits solved from
    6.5m MMT, 4m KPNO,
    4.1m SOAR
    Warren Brown, Mukremin Kilic, Alex Gianninas, et al.

    View Slide

  15. Brown et al. 2013, ApJ, 769, 66
    “The ELM Survey”
    An SDSS color-selected search for new extremely low-mass
    (ELM, <0.3 M¤
    ) white dwarfs
    Orbits solved from
    6.5m MMT, 4m KPNO,
    4.1m SOAR:
    • 82 WD+WD Porb
    <1.1 d
    • Most 15.5 < V < 18.5
    log(g)=
    log(g)=
    log(g)=
    log(g)=8

    View Slide

  16. Kilic et al. 2016, MNRAS, 460, 4176
    No clear pulsar
    companions to
    WDs so far in ELM
    Survey: all likely
    WD+WD binaries
    Chandra x-ray
    upper limits a
    factor of 2 more
    sensitive than MSPs
    detected in x-rays
    in 47 Tuc
    Null x-ray
    Null radio

    View Slide

  17. Mean Earth--Moon
    separation
    Minimum ELM Survey
    (J0651+2844, 12.75-min)
    Median ELM Survey (≈7 hr)
    Maximum
    ELM Survey
    (J0815+2309, 25.8-hr)
    1 R¤
    µWD2
    = 0.74 ± 0.24 M¤
    (Andrews et al. 2014, ApJ, 797, L32)

    View Slide

  18. Discovery spectroscopy
    determines the binary and
    atmospheric parameters
    blue-shifted red-shifted
    Brown et al. 2012, ApJ, 744, 142
    Teff
    = 10,540 ± 170 K
    log(g) = 6.01 ± 0.06 à M1
    = 0.17 M¤
    Porb
    = 87.996 ± 0.006 min
    K1
    = 508 ± 4 km s-1
    M2
    > 1.10 M¤
    if M1
    = 0.17 M¤
    tmerge
    < 170 Myr
    SDSS J1471+6526

    View Slide

  19. Discovery spectroscopy
    determines the binary and
    atmospheric parameters
    Teff
    = 10,540 ± 170 K
    log(g) = 6.01 ± 0.06 à M1
    = 0.17 M¤

    View Slide

  20. a b c d
    d
    c
    b
    a
    Usually, isochrones
    do not radically overlap
    on an H-R diagram

    View Slide

  21. ≤ 0.18 M¤
    WDs:
    residual pp-burning;
    simple but very slow
    evolution
    >0.18 M¤
    WDs:
    undergo unstable
    CNO burning flashes:
    Their evolution is not
    simple cooling
    Panei et al. 2007, MNRAS, 382, 779
    Stars evolve from A-J

    View Slide

  22. adapted from Althaus et al. 2013, A&A, 557, A19
    • Two ELM white dwarfs of different masses often cross the same
    points in a Teff
    –log(g) diagram
    • There is a non-uniqueness to using Teff
    ,log(g) for ELM WD mass

    View Slide

  23. Adapted from Althaus et al. 2013, A&A, 557, A19
    • There is a non-uniqueness to using Teff
    ,log(g) for ELM WD mass
    • For example, take an object at 10,000 K with log(g) = 6.60:

    View Slide

  24. Bassa et al. 2006, A&A, 456, 295
    PSR J1911–5958A: MPSR
    = 1.34 ± 0.08 M¤
    8.2m VLT

    View Slide

  25. Pulsar Masses Are Often Tied to Getting
    ELM White Dwarf Masses Correct

    View Slide

  26. Alina Istrate et al. 2016, A&A, 595, 35
    Metallicity certainly affects
    severity of CNO flashes
    These are diffusion-induced,
    unstable H burning episodes
    0.28 M¤
    0.28 M¤
    0.28 M¤
    Lower
    metallicity

    View Slide

  27. Alina Istrate et al. 2016, A&A, 595, 35
    We are finding that ELM WD
    models must include rotation
    CNO flashes likely cause shear
    affecting internal angular
    momentum:
    Prot
    ~ hours (sub-Porb
    )
    0.23 M¤
    0.23 M¤
    0.23 M¤

    View Slide

  28. Lowest-Gravity White Dwarfs All Show Metals
    Hermes et al. 2014, MNRAS, 444, 1674
    Diffusion should cause
    metals to sink out in <1 Myr
    (>100 Myr cooling ages)

    View Slide

  29. Istrate et al. 2016, A&A, 595, 35
    Rotational support
    (Eddington-Sweet
    circulation) can
    support metals at
    the gravities
    observed!

    View Slide

  30. M1
    ~ 0.19 M¤
    (M2
    ~ 0.90 M¤
    )
    R1
    = 0.093 ± 0.013 R¤
    i = 86.9 ± 0.4 deg
    Porb
    = 5.90724895(41) hr
    -20
    vrot
    = 50+30 km s-1
    Prot
    = 2.3+2.0 hr
    Hermes et al. 2014, MNRAS, 444, 1674
    -1.0
    GALEX J1717+6757
    Most low-mass white dwarfs rotate faster than Porb
    !

    View Slide

  31. Original Kepler Mission (4 years):
    Just 20 white dwarfs observed,
    K2 through Campaign 13:
    >1250 white dwarfs observed
    K2 has given us
    thousands of candidate
    white dwarfs to observe
    k2wd.org

    View Slide

  32. Marsh et al. 2017, in prep.
    We observed the first eclipsing ELM
    WD every minute for 78+ days w/ K2
    NLTT 11748
    spectra show:
    M1
    ~ 0.162 M¤
    M2
    ~ 0.74 M¤
    (i = 89.7 deg)
    secondary
    primary
    Römer delay:
    -4.44 ± 0.39 s offset:
    q = M2
    /M1
    = 4.6 ± 0.13
    (if e < 10-4)

    View Slide

  33. Marsh et al. 2017, in prep.
    We observed the first eclipsing ELM
    WD every minute for 78+ days w/ K2
    NLTT 11748
    primary
    (massive WD
    passing in front
    of ELM WD)
    w/out gravitational
    lensing
    w/ lensing
    Light curve fits (see also
    Kaplan et al. 2014):
    R1
    ~ 0.043 R¤
    R2
    ~ 0.011 R¤
    T1
    ~ 8700 K
    T2
    ~ 7590 K
    spectra show:
    M1
    ~ 0.162 M¤
    M2
    ~ 0.74 M¤
    (i = 89.7 deg)

    View Slide

  34. Yellow:
    Tidally
    distorted
    ELM WDs
    White:
    Eclipsing
    WD+WD
    binaries
    Pink: Eclipsing
    WD+A stars
    from Kepler
    Hermes et al. 2014, ApJ, 792, 39
    The Empirical He-WD Mass-Radius Relation

    View Slide

  35. Istrate et al. 2016, A&A, 595, 35
    0.23 M¤

    View Slide

  36. Istrate et al. 2016, A&A, 595, 35
    A ‘typical’ white dwarf
    0.6 M¤
    electron degenerate
    C/O core
    non-degenerate
    He layer
    non-degenerate
    H layer
    [thermal reservoir]
    [insulating blanket]
    A low-mass white dwarf
    0.2 M¤
    electron degenerate
    He core
    non-degenerate
    H layer

    View Slide

  37. Hermes et al. 2012, ApJ, 750, L28
    Hermes et al. 2013, ApJ, 765, 102
    Hermes et al. 2013, MNRAS, 436, 3573
    Kilic et al. 2015, MNRAS, 446, 26
    We can use pulsations to check H-layer thicknesses
    and thus calibrate cooling ages: Asteroseismology!

    View Slide

  38. Kilic et al. 2015, MNRAS, 446, 26
    We have also found a pulsating
    ELM WD around a MSP!
    Antoniadis et al. 2012, MNRAS, 423, 3316
    "Looking at individual acquisition frames, the
    scatter of the magnitude difference was
    ∼0.05 mag, somewhat larger than expected
    based on measurement noise, though with
    no obvious correlation with orbital phase."
    V = 21.3 mag
    Pulsation periods of
    1790-3060 s (0.50-0.85 hr)

    View Slide

  39. McDonald 2.1m
    KPNO 2.1m
    SOAR 4m,
    PROMPT 0.41m
    SAAO 1m SSO 2m
    Mt. John 1m
    NARIT 2.4m
    Lulin 1m
    BOAO 1.8m
    Xinglong
    2.16m
    Tian-Shan 1m
    Peak Terskol 2m
    CrAO 1.25m
    Tuebingen &
    Mt. Suhora 0.6m
    Haleakala 2m
    CTAS 0.7m
    • Whole Earth Telescope run from 2013 April/May on J1518+0658
    – 17 sites in coordinated global campaign
    – 54 days, ~40% duty cycle

    View Slide

  40. • Whole Earth Telescope run from 2013 April/May on J1518+0658
    – 17 sites in coordinated global campaign
    – 54 days, ~40% duty cycle
    relative amplitude

    View Slide

  41. ELM White Dwarfs Strong Sources of Gravity
    Waves (Pulsations) and Gravitational Waves!

    View Slide

  42. Gravitational Waves
    Gravity Waves

    View Slide

  43. Mean Earth--Moon
    separation
    Minimum ELM Survey
    (J0651+2844, 12.75-min)
    Median ELM Survey (≈7 hr)
    Maximum
    ELM Survey
    (J0815+2309, 25.8-hr)
    1 R¤

    View Slide

  44. phase = 0
    • We detected eclipses in April 2011
    • This is the most compact detached binary
    system currently known!
    J0651+2844: A 12.75-min WD+WD Binary
    Brown et al. 2011, ApJ, 737, L23

    View Slide

  45. Photo by Taylor Chonis
    Eclipses were detected on 1 April 2011 from the 2.1m telescope at
    McDonald Observatory on the first night of a 14-night run

    View Slide

  46. The Rock House fire
    started on 9 April 2011
    near Marfa, TX
    It eventually burned
    >310,000 acres
    (>1270 km2)
    Photos I took on 9 April 2011

    View Slide

  47. The Rock House fire eventually
    burned >310,000 acres
    Photo by Don Winget on 11 April 2011

    View Slide

  48. We re-started monitoring the 12.75-min binary on 13 April 2011,
    the first photons collected at McDonald after the Rock House fire started.
    Photo by Frank Cianciolo on 17 April 2011

    View Slide

  49. J0651+2844: A 12.75-min WD+WD Binary
    Average distance between the Earth and the Moon: 384,400 km
    M2
    = 0.50 ± 0.04 M¤
    M1
    = 0.25 ± 0.04 M¤

    View Slide

  50. Orbital Decay in J0651+2844
    After just 13 months we confirmed orbital decay from gravitational radiation.
    Hermes et al. 2012, ApJ, 757, L21

    View Slide

  51. We expect dPorb
    /dt = (-0.26 ± 0.05) ms/yr from GR (point masses) and
    observe (-0.2875 ± 0.0011) ms/yr!
    – a 0.4% measurement!
    Hermes et al. 2018, in prep.

    View Slide

  52. • This 12.75-min WD+WD binary is decaying > 3.8x faster than the 7.75-hr
    Hulse-Taylor binary pulsar (first indirect detection of gravitational waves)
    Weisberg et al. 2010
    J0651+2844 PSR B1913+16
    dP/dt = -0.288 ms/yr dP/dt = -0.076 ms/yr
    Orbital Decay in a 12.75-min WD+WD binary

    View Slide

  53. • Tidal torques should increase the rate of
    orbital decay in J0651+2844
    – Additional angular momentum is lost from the orbit to
    spin-up the WDs to remain synchronized, leading to
    >5% faster rate of orbital decay (e.g., Piro 2011, ApJ,
    740, L53; Fuller & Lai 2012, MNRAS, 421, 426)
    The Fate of the WDs in J0651+2844

    View Slide

  54. • J0651+2844 is an excellent verification source for direct detection of
    gravitational waves with eLISA: forb
    = 2.6136738(32) mHz
    • Expect Roche lobe contact in less than 1 million years
    • J0651+2844 should be
    detectable by eLISA
    with S/N > 7 within
    six months
    Kilic, Brown & Hermes 2013
    J0651 is an Excellent eLISA Verification Source
    eLISAplanned sensitivity
    known detached
    binaries
    known interacting
    binaries (AM CVns)

    View Slide

  55. population synthesis by Valeriya Korol et al. 2017, MNRAS, 470, 1894
    • In next decade, we expect to find 1000s of eclipsing WD+WD
    binaries (just 7 known to date)
    hundreds (GAIA) thousands (LSST)

    View Slide

  56. Conclusions
    • Determining the mass of a low-mass white dwarf has caveats
    – Complicated evolution due to CNO-flash episodes
    • Low-mass white dwarfs are likely fast rotators
    • Cooling ages strongly affected by outer H layer thickness
    – Asteroseismologyis finally getting off the ground for He-core WDs
    • 100s of close, eclipsing WD+WD binaries from Gaia/LSST
    – The mHz regime it is awash in persistent sources from close WD+WD binaries
    D. Berry, GSFC

    View Slide