Topic 6 - University of Sheffield/file/Topic6_SC.pdf · 04/02/2013 4 Type Ia Light Curves SN2003du Correlation of luminosity with decline rate Type Ia Supernovae • Typical white

04/02/2013

1

Topic 6

Supernovae

Introduction

• Supernovae (SNe) are the explosive deaths of stars

• Important for Nuclear Astrophysics because– extreme conditions in the SN explosion allow production

of elements not created, or much less frequently created, in normal stellar interiors

– explosion also disseminates this material into the interstellar medium where it can be incorporated into new stars

– explosion may also trigger new star formation by creating shock waves in nearby gas

04/02/2013

2

Supernova Classification• Based mainly on spectral features, with some input

from light curve shape (photometry)

No Hydrogen

I

Si II (615 nm)Typical Ia

27%Fainter, lower ejectavelocity, no 2nd max in IR Iax

No Si IIStrong He features Ib 7%

No/weak He Ic 14%

Hydrogen II

Hydrogen vanishes at late times, II � Ib IIb 4%

Hydrogen always present

Plateau in light curve II-P 43%

Linear decline II-L 2%

Narrow lines IIn 3%

Relative frequencies are quite uncertain—error in these numbers up to ±50% for smaller numbers.

SN I vs SN II light curves

Slope at late times powered by 56Ni � 56Co + e+ + ν(t1/2 = 6.08 days)56Co � 56Fe + e+ + ν(t1/2 = 77.27 days)

04/02/2013

3

Supernova rates

e.g. arXiv:1006.4613v2 (Lick survey):

2.84 ± 0.60 total number SN/100 years2.30 ± 0.48 core collapse SN/100 years… in Milky Way

Type Ia Supernovae• Diagnostic features: absence of H lines, presence of strong

ionised Si feature near 615 nm– “P Cygni profile”, showing outflow– initial ejecta velocities ≥10 000 km s−1

• Occur in all types of galaxies, including elliptical, though more common in late types– therefore not due to massive stars

• Spectra show intermediate-mass elements (Si, Ca, Mg, S, O) near maximum light; Fe II, Fe III, Co III at late times– this looks like C/O burning– relative intensities of Co III/Fe III indicate that late-time light is

powered by radioactive decay of 56Co to 56Fe– consistent with production of few tenths of solar mass of 56Ni

• Light curves and other properties all very similar�Everything points to explosive destruction of a white dwarf

04/02/2013

4

Type Ia Light CurvesSN2003du

Correlation of luminosity with decline rate

Type Ia Supernovae

• Typical white dwarf has mass 0.6M�

– need to increase this to Chandrasekhar limit, 1.4 M�, to explode– obvious route to this is mass transfer in a binary system

• Two possible mechanisms• Single Degenerate scenario

– more massive star in binary system evolves to white dwarf, which then accretes mass from companion (main sequence star or red giant—several possible scenarios)

– tricky to “tune” accretion so as to avoid classical nova (which blows off more mass than has been accreted)

• Double Degenerate scenario– two white dwarfs in binary will lose energy by gravitational radiation,

eventually spiralling inward to coalesce

• SD is preferred, but neither model is totally unproblematic, and both may occur

04/02/2013

5

Type Ia Supernovae

Type Ia Supernovae

Many possible SD evolutionary paths!(from Wang and Han 2012)

04/02/2013

6

Type Ia Supernovae

• Once WD reaches 1.4 M� it will collapse—gravity will overcome electron degeneracy pressure

• Three possible mechanisms for resulting explosion– prompt detonation: burning proceeds as supersonic shock wave

which does not allow unburnt material time to expand before being consumed

• this tends to convert entire star to 56Ni, which is not observed (intermediate-mass elements seen in spectrum)—therefore disfavoured

– pure deflagration: burning front moves outward subsonically, producing complex turbulent boundary layer propagating outward

• it is not absolutely certain that this can generate an explosion– delayed detonation: initial deflagration converts to detonation as it

propagates into lower-density regions• this definitely does explode, and is preferred scenario

Type Iax Supernovae

• New class (proposed by Foley et al.,December 2012)

• Minority of SNe Ia– have low ejecta velocity (<8000 km/s)

– lack secondary maximum in R and Ibands

– are fainter than standard SNe Ia(absolute magnitudes from −14.2 to−18.9) and do not obey standardpeak magnitude–decline rate correlation

– appear to be found principally in late-type galaxies (one in an S0)

• These may be pure-deflagration events which fail to disrupt the WD completely, leaving a surviving WD remnant– for every 100 SNe Ia, about 30(±15) SNe Iax (large minority)

04/02/2013

7

Nucleosynthesis in SNe Ia

• Nuclear statistical equilibrium � mostly iron peak elements• Some intermediate-mass elements from partially burned

C/O

1.E-10

1.E-08

1.E-06

1.E-04

1.E-02

1.E+00

10 20 30 40 50 60 70A

SN II

SN Ia

Core Collapse Supernovae

• All other supernova classes (Ibc, II) are caused by the core collapse of a massive star

• We saw in Topic 4 that stars >~10 M� will fuse successively heavier elements– each successive stage requires higher temperature, is less

efficient in generating energy, and hence lasts for less time– star develops “onion” structure with layers of increasingly

heavy elements from hydrogen envelope to heavy core– this continues until Si fusion creates (in a few days) an iron

core exceeding 1.4 M�

• Stars of ~8-10 M� form an O-Ne-Mg core– this will also collapse and generate a supernova

04/02/2013

8

Core Collapse Supernovae

• Collapse of the iron core– As usual, as the fuel is exhausted gravitational collapse

occurs causing heating. Usually this ignites the next fuel.– However in this case the mean photon energy becomes

such that photodissociation of Iron can take place. This is the so-called Iron-Helium phase transition

γ + 56Fe � 13 4He + 4n −124.4 MeV– Note that this process is endothermic– This energy is provided at the expense of the

gravitational field– This accelerates the core collapse

Core Collapse Supernovae

• Neutron star formation– As the core further collapses the transition to a neutron

star commences– Density reaches 1017 kg.m-3 and electrons are forced

into protons causing neutrons to form– This generates a burst of neutrinos from p + e− � n + νe

– In fact neutrinos carry off ~99% of the energy of the collapse (few x 1046J)—but at this early stage they are trapped, as collapse is faster than neutrino diffusion time

– Collapse is rapid, neutron star forms, outer core of star is not collapsing (infalling) as fast and so hits the neutron star and “bounces off”

04/02/2013

9

• Prompt shock– Bounce creates a shock wave

which propagates outward– Shock dissociates nuclei into

free protons and neutrons• using up 8-9 MeV/nucleon

– Neutrino burst emerges at shock breakout

– Energy loss due to dissociation of nuclei and neutrino cooling will cause shock to stall

�Initial shock will not cause explosion

Markus Weiland, LMU

• Delayed neutrino heating mechanism– Neutrinos carry 99% of the supernova energy– Neutrino cooling

• inverse β decay: e− + p � n + νee+ + n � p + ν̅e

• pair production: e− + e+ � ν + ν̅

• Bremsstrahlung: N + N � N + N + ν + ν̅

– Neutrino heating• n + νe � e− + p

• p + ν̅e � e+ + n

– If a few % of neutrino energygoes into heating, shock willrevive

�Explosion!!

Markus Weiland, LMU

04/02/2013

10

Nucleosynthesis in CCSNe

• Explosive fusion– Pre-existing iron-peak elements are mostly dissociated

during core collapse– However, propagating shock wave will initiate explosive

burning of Si, O, Ne and C• Si burning produces iron-peak elements including 56Ni

whose decay powers late-time exponential decay of light curve

• O burning produces many α-particles, which will enhance production of “α-process isotopes” such as 16O, 20Ne, 24Mg (effectively bound states of α-particles)

• this occurs at temperatures of several ×109 K and high densities—conditions for nuclear statistical equilibrium

Nucleosynthesis in CCSNe

• p and r processes– Early ejecta are proton rich

• this may be where “bypassed nuclei” (p-process nuclei) are formed by proton capture (rp-process)

– Later ejecta may be neutron rich• high neutron number densities� possible site for r-process

• however unclear if this reallyhappens

• for low-metallicity SNe r-processmay occur further out, in He-shell of original star(cold r-process)

Markus Weiland, LMU

04/02/2013

11

Electron Capture Supernovae

• Stars of 8-10 M� form an O-Ne-Mg core from carbon burning– this is supported by electron degeneracy pressure (i.e.

all accessible electron quantum states are full)

– at 1.4 M� and 4.5×1012 kg m−3 electron capture on 24Mg becomes energetically favoured: e− + 24Mg � 24Na + νe

– this removes electron pressure support � core collapses

– simulations indicate fairly weak explosion forming sub-Chandrasekhar mass neutron star, with much less heavy element ejection than standard CCSN (<0.001M�

56Ni, compared to 0.01−0.1 M� for typical CCSN)

Observations of CCSNe

• Progenitors are massive stars– therefore should expect SNe to occur in association with star

formation: not in elliptical galaxies, S0 galaxies, spiral galaxies bulges

• Most evolved massive stars should have an outer envelope of hydrogen– therefore expect strong hydrogen lines in spectrum

• These features match observations of Type II supernovae• Type Ib/c supernovae are also associated with star

formation, but don’t have strong hydrogen– progenitors’ hydrogen envelopes lost through stellar winds or binary

mass transfer

– Distinction is not sharp: Type IIb appears to be genuine intermediate (very low-mass H envelope, quickly dispersed)

04/02/2013

12

Type II-P Supernovae• Strong hydrogen features at maximum light �

hydrogen envelope in progenitor star• Plateau is caused as visible photosphere (at

~5000K) migrates backwards through expanding envelope as density decreases– implies fairly extended, massive envelope– most likely progenitor is therefore red supergiant– several candidate RSG progenitors for SNe II-P have

now been detected (for SN2003gd, SN2005cs, SN2008bk, SN2012aw)

– first 3 have masses around 8 M�, which indicates that stars of this mass can produce “normal” CCSNe

• apparently not ECSN? or ECSN explosion stronger than expected?

04/02/2013

13

SN 1987A in the LMC• SN II-P in Large Magellanic

Cloud at ~50 kpc• Progenitor definitely the B

supergiant Sk −69 202– unexpected: most II-P arise from

red supergiants

– more compact star led to atypicalearly light curve

• 24 neutrinos detected ~3 hours before light– time delay, number, and energies of neutrinos consistent with

expectations from core collapse (total energy emitted ~ 2×1046 J)

• No neutron star detected (no pulsar or energy source)– various ways to explain this, depending on assumed properties of NS

– also possible that enough material fell back onto NS to make BH

04/02/2013

14

Other CCSNe

• Type II-P accounts for ~60% of all CCSNe– RSGs of ~8-18 M�

– wide range of brightnesses, plateau widths, ejectavelocities, etc., probably reflects range of masses/radii of progenitor stars

– no evidence of very massive (>20M�) progenitors

• Type II-L (somewhat brighter, no plateau) is comparatively rare– lack of plateau suggests lower mass hydrogen

envelope—mass loss?– higher luminosity and lower frequency � possibly more massive progenitors?

– One candidate detection suggesting RSG of 15-20 M�

Type IIb/Ibc Supernovae• Progenitors must have been stripped or almost

stripped of their hydrogen envelopes (and in case of Ic also helium), but from location must be massive stars

• Wolf-Rayet stars?– very massive stars that lose outer layers via stellar wind– these are very luminous: non-detection of progenitor for

several nearby SNe Ibc more or less rules them out as usual progenitors (though they may account for some)

• Luminous Blue Variables?• Interacting binaries?

– hydrogen envelope lost through mass transfer

04/02/2013

15

SN 1993J in M81

• Well observed SN IIb in nearby galaxy• Progenitor identified in HST imaging—from

photometry apparently binary consisting of stripped K supergiant with OB supergiant companion– B supergiant apparently still there in post-explosion

photometry

• This is the best characterised SN progenitor apart from Sk −69 202 (SN 1987A)– by coincidence, the surviving B supergiant might well

later explode as a BSG, like Sk −69 202

SN 1993J in M81

Image courtesy of NRAO/AUI and N. Bartel, M. Bietenholz, M. Rupen, et al.

04/02/2013

16

Pair Instability Supernovae

• Stars with masses >20-25M� probably do not explode as core-collapse supernovae– they collapse directly to a black hole, so there is no surface

for the outer layers to bounce off

• However, extremely massive stars (>150M�) can explode by a very different mechanism– very massive stars are supported mainly by radiation

pressure (not gas pressure as in Sun)– temperature during C/O burning exceeds threshold for e+e−

pair production, reducing number of photons and hence decreasing pressure

– this causes core collapse and runaway thermonuclear fusion, disrupting star completely (no remnant)

Pair Instability Supernovae

• This is very important for first-generation (Pop III) stars, which are expected to be very massive– PISNe therefore responsible for first injection of heavy

elements into interstellar medium

Ultraluminous Type Ic SN2007bi may have been a PI SN

04/02/2013

17

Solved Supernova Problem

• A 10 solar mass star explodes leaving behind a neutron star of radius 10 km and density 6.7 x 1017 kg.m-3. A theorist proposes that 10% of the mass difference is released as high energy (10 TeV) photons. Assuming the photons are released isotropically and that the supernovae takes place on the other side of the galaxy at 50 kpc from Earth estimate the flux of high energy photons incident on the Earth after the explosion. (1 TeV = 1012eV)

Solved Problem solution