A hypothesis or theory is clear, decisive, and positive, but it is believed by no one but the person who created it. Experimental findings, on the other hand, are messy, inexact things, which are believed by everyone except the person who did that work.

Harlow Shapley Through Rugged Ways to the Stars Smashing White Dwarfs: Book One of the Supernovae Trilogy

Frank Timmes

Starlib SPIDER isotopes 100 project

THE ASTROPHYSICAL JOURNAL White dwarf supernova play a key role in astronomy:

Distance indicators Element factories Cosmic-ray accelerators Kinetic energy sources Binary star terminus Identification of what is exploding is unknown - this is the outstanding mystery in the field.

Tycho Supernova Remnant: Age: - Nov 1572 Distance: ~ 8500 ly Diameter: ~ 18 ly (8 arc min) Expansion: ~ 0.0015 ly/yr

Blue - high energy electrons

Constellation: Cassiopeia Green & Yellow - iron and silicon Red - dust NASA’s Spitzer, Chandra, & Spain’s Calar Alto Several observational characteristics help in the hunt for the progenitors of the explosions:

1) About 90% of all white dwarf supernova form a homogeneous class in terms of their spectra and light curves. White dwarf supernovae are defined by their spectra: no hydrogen lines and a strong silicon absorption feature.

silicon II magnesium II cobalt II

sulfer “W”

iron II calcium II silicon II oxygen I calcium II

4000 5000 6000 7000 8000 Kasen 2008 wavelength in angstroms Pinwheel (M101)

SN 2011fe

Near maximum light, spectra are characterized by O-Ca at high velocity (~20k km/s).

Late nebular phase spectra are dominated by iron lines. Pereira et al 2013 Expansion and Diffusion time scales about equal

1010 )

⊙ Optical light curve

56Ni (τ ~ 6 d) + 56 1/2 Co (τ1/2~ 77 d)

56 ~0.6 M⊙ of Ni γ-ray escape Luminosity (L for a typical SNIa

109 0 20 40 60 Time (days since peak) Several observational characteristics help in the hunt for the progenitors of the explosions:

1) About 90% of all white dwarf supernova form a homogeneous class in terms of their spectra and light curves.

2) Correlations between different observables, such as the peak luminosity and width of the light curve. -20 Brighter is wider. as measured

-19 /65) h 5 log( - V

M -18 This empirical fact can Calan/Tololo SNe Ia be used to correct for -17 -20 0 20 40 60 variations in the peak days luminosity to give a -20 light-curve timescale standard candle. “stretch-factor”corrected

-19 /65) h 5 log( - V

M -18 After correction,

Kim et al 1997 distances are -17 -20 0 20 40 60 accurate to ≤ 7% ! days

Planetary Nebula: Runs out of H He Tosses off Hydrogen Helium fuel 106 Bright and Helium Layers C+O

Red Giant H He

Helium Burning Main Sequence to Carbon Helium Ignites Luminosity (Lsun) 1 Msun H → He 1 Carbon-Oxygen White Dwarf in 10 billion years

10-4 Dim Hot Warm Cool Surface Temperature (K) 50,000 6000 3000 Blue Color Yellow Red Different main-sequence stars make different white dwarfs.

C core + H envelope DA WDs SDSS DR1, Madej et al. 2004

100

50 Number of stars

0 0.2 0.4 0.6 0.8 1.0 1.2

Mass [M⊙] Pluto White Dwarf

Radius: 1185 km Radius: 1185 km

Mass: 0.18 Earth’s Mass: 1.37 Sun’s A white dwarf can only have so much mass.

1.5 n=3 polytrope

+ General Relativity ) + Coulomb Corrections sun 1

Ideal Fermi gas

n=3/2 polytrope Total Mass (M .5 White Neutron Dwarf Star Supernova

0 105 107 109 1011 Central Density g cm-3 Single-Degenerate channel Double-Degenerate channel

Mergers:

Collisions:

The relative frequency of these channels is unknown. The first white dwarf smashes were 0.6 + 0.9 Msun calculated in 1985:

3D, 5000 particles with nuclear burning done afterwards and approximate thermodynamics.

Bottom line: Tiny amounts of 56Ni produced.

Message: Nothing here, move along. Benz et al 1985 2010: 2 million particles with inline burning and realistic thermodynamics.

0.6 + 0.6 M⨀, zero impact parameter, x-y plane, temperature.

Raskin et al 2010 Message in 2010: Lots of interesting possibilities!

0.6 ]

๏ 0.5

0.4 0.6 + 0.6 M⨀ 0.3 Equal mass 0.2 Equal h

Ni Yield [M Yield Ni 1985

56 0.1 result Raskin et al 2010 0 0 500 1000 1500 2000 Particle Count [x1000] Collisions can cover the observed range of 56Ni masses.

1.5 ) ⊙ Range of observed 56Ni masses Ni (M

56 1 Average observed

Mass mass

.5 Raskin et al 2010 Kushnir et al 2013

0 .4 .6 .8 1 1.2 Average White Dwarf Mass (M⊙) Exceptionally bright white dwarf supernova have been interpreted as double-degenerate events.

2.0

SN 2009dc SN 2007if 1.5 ) SN 2003fg sun SN 2006gz

1.0 SN 2005hj Nickel Mass (M

0.5

Howell et al 2006

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Luminosity (1043 erg) Observations suggest about 5 million white dwarf supernova per year within a redshift of one.

SDSS David Kirkby An objection to the collision scenario is the perception that such collisions are extremely rare.

Eugene Onegin and Vladimir Lensky’s duel.

Watercolor by Ilya Repin (1899) Collisions have been believed to predominantly occur in dense stellar environments, such as cores of globular clusters. Even accounting for gravitational focusing, the collision rate is ~5000 white dwarf supernovae per year within a redshift of one.

2 2 2 vesc σ = πb = πR 1 + v Wait! There is a 3rd body in this duel.

Eugene Onegin and Vladimir Lensky’s duel.

Watercolor by Ilya Repin (1899) Hierarchical triple star systems with white dwarf binary orbital separations of 1-300 AU are known to exist. The white dwarf binary’s ellipticity can be driven to large values in a triple star system because ellipticity can be traded for inclination in the conservation of angular momentum

2 Lz = √1 e cos(i) − in Kozai-Lidov oscillations.

e � 0.999999

3 body problem

15 10 m1 = m2 = 0.5 M⊙ white dwarf binary + m3 = 0.5 M⊙ perturber semimajor axis 14 10

13 10

12 10

11 10 Distance (cm) pericenter seperation 10 10

9 10 Radii of white dwarfs Head-on Collision Katz & Dong 2013 8 10 0 20 40 60 80 100 120 time (x 1000 yr) White dwarfs in a triple star system have a ~3% chance of experiencing a collision within 5 billion years.

If ~20% of white dwarfs are in triplets, the calculated supernova rate is about the same as the inferred rate. 15-20 % of 1M⊙ ≤ M ≤ 8M⊙ stars with a M > 1M⊙ companion makes the collision scenario dominant.

AO Measurements of A-stars, N=121 RV Measurements of red giants in open clusters, N=797

All binaries All Red Giants 0.5 < P < 5 yr secondary ⊙ 15 M > 1 M 0.5 < P < 5 yr, Msecondary > 1M⊙

100 10

5 10 Number of binaries Number of Red Giants

0 5 50 500 5,000 50,000 1 Period (years) 1 2 4 8 16 Mprimary (M⊙)

Klein & Katz 2016 Prediction: GAIA will find ~10 new wide orbit double degenerates within 20 pc from the Sun.

This puts a strong constraint on the “triple-assisted” collision model for white dwarf supernovae.

Advances (plus a little serendipity) over the next decade should enable us to decipher the progenitors of white dwarf supernovae:

1) Silicon, Sulfur, Calcium ratios 2) Unburned carbon and oxygen 3) Tidal tails 4) Sufficient number of binary WDs 5) Early gamma-rays 6) Narrow hydrogen emission or absorption 7) Circumstellar interaction in radio or x-rays 8) Gravitational wave signatures 9) Frequency as a function of redshift A successful model starting from a carbon+oxygen white dwarf must make

0.1 - 1.0 M☉ 56Ni for the light curve

0.2 - 0.4 M☉ Si, S, Ar, Ca for the spectrum

< 0.1 M☉ 54Fe + 58Ni for the nucleosynthesis

Allow for some diversity for the population Raskin et al 2010 0.8 + 0.6 M⨀, zero impact parameter, density

Raskin et al 2010 0.8 + 0.6 M⨀, zero impact parameter, density, zoomed

Raskin et al 2010