Fingerprinting SNRs: Clues to Explosion Progenitors · Inspiration Carles Badenes Caltech 08/23/09 2 Krause et al. 08, Sci 320, 1195 SNR evolution strongly modified by progenitor

Carles BadenesCarles Badenes Princeton UniversityPrinceton UniversityWeizmann Institute of ScienceWeizmann Institute of ScienceTel-Aviv UniversityTel-Aviv University

A Festival of Cosmic Explosions (Rogerfest)A Festival of Cosmic Explosions (Rogerfest)

Caltech, Pasadena CACaltech, Pasadena CAAugust 23 2009August 23 2009

Collaborators:Collaborators:J. Hughes (Rutgers), K. Borkowski, S. Reynolds, J. Blondin (NCSU),J. Hughes (Rutgers), K. Borkowski, S. Reynolds, J. Blondin (NCSU),E. Bravo (IEEC/UPC), J.L. Prieto (OSU), D. Zaritsky (UA), J. Harris (NOAO) E. Bravo (IEEC/UPC), J.L. Prieto (OSU), D. Zaritsky (UA), J. Harris (NOAO)

Fingerprinting SNRs: Clues to Explosion Progenitors

LMC image LMC image from MCELSfrom MCELS(C. Smith et al.)(C. Smith et al.)

Inspiration Carles BadenesCaltech 08/23/09

2Krause et al. 08, Sci 320, 1195

SNR evolution strongly modified by progenitor properties : ejecta mixing, fallback onto CCO, PWN expansion, CSM interaction, AM photoionization ⇒ Cas A must have been a Type IIL or IIb SN.

Cas A

Outline Carles BadenesCaltech 08/23/09

3

Lack of SN observations is not the problem. In this talk, you'll hear about the SNR approach to Type Ia SN progenitors:

● Introduction: Type Ia SNe.

● Ejecta emission in Type Ia SNRs and comparison to light echoes.

● Structure of the CSM in Type Ia SNRs and constraints on the progenitor systems.

● Stellar populations around Type Ia SNRs in the LMC.

● A new method to measure the metallicity of Type Ia SN progenitors directly.

The explosion mechanisms and progenitor systems of Type Ia SNe are still unknown

Type Ia SNe

4

SupernovaCosmology Project

Type Ia Supernovae (SNe) are the result of the thermonuclear explosion of a C+O white dwarf

prompted by accretion in a binary system

SN 1994D inNGC 4526 (P. Challis)

REVIEWS: Branch et al. 95, PASP 107, 1019; Branch & Khokhlov 95, Phys. Rep. 265, 53; Hillebrandt & Niemeyer 00, ARA&A 38, 191.

● Fundamentals are well understood: energy budget, no H in spectra, rate of light curve decay.

● Some key details remain obscure: progenitor systems, explosion mechanism.

● Light curves and spectra are strikingly uniform ⇒ LC width / luminosity relation [Phillips 93, ApJ 4123, L105] ⇒ Cosmology.

Carles BadenesCaltech 08/23/09

The Ejecta of Type Ia SNe

5

Mazzali et al. 07, Sci 315, 825 [23 Type Ia SNe]

DET

DDTa

DDTe

Ek, ρtr

● Phenomenological 1D delayed detonation (DDT) models provide the best match to Type Ia SN observations.

IME:IME:Si, S, Ar, CaSi, S, Ar, Ca

Fe-peakFe-peaknuclei: nuclei: 5656Ni, Ni, 5454 Fe Fe


SN Ia Progenitor Systems

6

Sullivan et al. 06, ApJ 648, 868

Gallagher et al. 05, ApJ 634, 210

● Interpreted as evidence for TWO progenitor populations:

● 'Prompt' 'younger' (~100 Myr) ⇒progenitors in star forming galaxies, SN rate star formation rate, ∝brighter Type Ia SNe.

● 'Delayed' 'older' (~Gyr) ⇒progenitors in passive galaxies, SN rate total stellar mass, dimmer ∝Type Ia SNe.

● A+B models [Scannapieco & Bildsten 05, ApJ 629, L85]:

SNIa=AMstars+B(dMstars/dt)

● Both populations appear to follow the same Phillips relation! (at least to 1st order - Howell et al. 07, ApJ 667, L37).


Type Ia SNRs: Tycho

7


● Must model spectra + dynamics; M56Ni estimate based on 1D DDT models. ● Tycho SNR: M56Ni=0.74 M ⊙

[Badenes et al. 06,

ApJ 645, 1373]. Later confirmed by light echo [Rest et al. 08, ApJ 681, L81; Krause et al. 08, Nat 456, 617]. Tycho SNR

Krause et al. 08, Nat 456, 617

The X-ray spectra from SNRs can be used to type (Ia/CC) objects and estimate the peak

brightness (M56Ni) of SN Ia.

[Badenes et al. 03 ApJ 593, 358; 05 ApJ 624, 198.]

Badenes et al. 06,ApJ 645, 1373

Ejecta+FSFS

Best DDT

Type Ia SNRs: Tycho

8


Ek, ρtr

● Only DDT models at can reproduce BOTH X-ray spectrum and SNR dynamics.

● Mildly energetic explosion ⇒ Model DDTc: Ek=1.2x1051 erg; M56Ni=0.74 M ⊙

[Badenes et al.

06, ApJ 645, 1373].

● WARNING: Comparison of synthetic & observed spectra is NOT trivial! (MUST be quantitative).

Age is known (SN1572). SNR size imposesa correlation between D and ρAM:

D vs. ρAM

DDTmodelsρAM = 2x10-24

8

Bright SNIa

Dim SNIa

Normal SNIa

1D, No CR acceleration

Type Ia SNRs: SNR 0509-67.5

9


● Method can be applied to any young, X-ray bright, ejecta-dominated Type Ia SNR. Need at least a good estimate of the age.

● SNR 0509-67.5 in the LMC ⇒ M56Ni=0.97 M

⊙ [Badenes et al. 08, ApJ 680, 1149]. Also confirmed by the light echo [Rest et al. 08, ApJ 680, 1137]. [see also Kosenko et al. 08 A&A 490, 223].

3 SN1991T-like Type Ia SNe

Rest et al. 08, ApJ 680, 1137

Badenes et al. 08ApJ 680, 1149

Mass Loss From SN Ia Progenitors

10


Accretion Winds [Hachisu et al. 96, ApJ 470, L97]: Fast (~1000 km s-1), optically thick outflows that stabilize accretion onto the WD for ~1 Myr before the SN explosion ⇒ Allow WD to grow to MCh.

● Only way to avoid a common envelope phase.

● The details can be quite complex, but accretion winds are a fundamental ingredient of ALL published binary evolution models for SN Ia progenitors in the single degenerate channel [Langer et al. 00, A&A 362, 1046; Han & Podsiadlowski 04, MNRAS 350, 1301, ...].

● Fast outflows must excavate large energy-driven cavities around the SN progenitors [Koo & McKee 92, ApJ 388, 93]. Hachisu et al. 99,

ApJ 522, 487

The CSM of Type Ia SNRs: No Fast Outflows

11

● The cavities strongly modify the dynamics and X-ray emission from Type Ia SNRs.

● SNRs that evolve in these cavities are large and have low net ejecta, unlike the known examples of Type Ia SNRs.


Kepler: A SN Ia With N-enriched CSM

12

Kepler: A Type Ia SNR with circumstellar interaction

● Optical: dense knots (N enriched), radiative shocks. ~500 pc above the Galactic plane, high systemic velocity (>200 km.s-1) ⇒ Massive runaway progenitor interacting with a bow shock CSM [Bandiera 87, ApJ 319, 885].

● X-rays: lots of Fe in the ejecta, but no detectable O. No compact object (>10-2

LCas A). Balmer shocks (require partially neutral CSM) ⇒ Thermonuclear SN.750 ks Chandra exposure [Reynolds et al.

07, ApJ 668, L135]

● The Kepler SNR is is a Type Ia SN expanding into N-enriched CSM. ● Either the progenitor or the companion must have lost CNO-enriched material before the explosion.● No detailed models for Kepler's ejecta emission + CSM interaction exist.


Type Ia SNRs vs. Extragalactic SN Ia

13


● Much of what we know about Type Ia SN progenitors comes from studies of the stellar populations in their host galaxies ⇒ metallicities, ages, SFRs, SFHs. Issues:

● Unresolved stellar populations ⇒ luminosity weighted spectra/SED.

● Measurements are not local ⇒ age and metallicity gradients.

● Can we use resolved stellar populations at the location of Type Ia SN progenitors?

● YES! ⇒ Star Formation History (SFH) maps for the LMC [Harris & Zaritsky 09, arXiv:2009, arXiv:0908.1422] and SMC [Harris & zaritsky 04, AJ 127, 1531].

● What about the SNe? ⇒ Use Supernova Remnants (SNRs)!

Detail of the LMC(MCELS)

Typical Ia SN hosts (SDSS)

Type Ia SNRs in the LMC

14


Continuum ImageHα Contours

CC SNRs: N158A, SN1987A, N63A, N49

Type Ia SNRs: 0519-69.0, N103B, DEM L71, 0509-67.5

● 47 known SNRs in the LMC [Williams et al. 99, ApJS 123, 467].

● Avoid mistyping ⇒ young SNRs (ejecta-dominated or with a compact object).

● Four objects: DEM L71, 0509-67.5, 0519-69.0, N103B

● Three w/ light echoes (0509-67.5, 0519-69.0, N103B) [Rest et al. 05, Nat 438, 1132].

SFHs of the CC SNRs

15


SFHs of the Ia SNRs

16


SNRs 0509-67.5 and N103B

17


● SNR 0509-67.5 was originated by a very bright Type Ia SN (M56Ni~1M

⊙).

Δm15<0.9 [Rest et al. 08, ApJ 680, 1137] ⇒ MV

~ -19.5. Yet, it is embedded in an old (7.9 Gyr) and metal-poor (Z=0.0014) stellar population.

● SNR N103B, on the other hand, is associated with vigorous SF in the recent past (t<100 Myr). Its morphology shows signs of interaction [Lewis et al. 03, ApJ 582, 770]. Maybe it had a young(er), more massive progenitor that lost a large amount of mass before the explosion.

No definitive claims about the progenitorscan be made from the SFHs alone!

A Surprise From Suzaku

18


● Tamagawa et al. [PASJ]: Suzaku detection of Cr (>10σ) and Mn (~7σ) Kα emission lines from ejecta in the Tycho SNR ⇒ 10 elements detected in X-rays: O, (Ne), Mg, Si, S, Ar, Ca, Fe + Cr, Mn.● The upper limit on Ni is important, but the detection of Mn is more interesting (only odd-Z element!).

Suzaku

No Ni!

Zoom Around Fe Kα

XMM-NewtonIntegrated Spectrum

Metallicity in Type Ia Supernovae

19


● Type Ia SNe ⇒ C+O WDs born from intermediate mass stars that burn H through the CNO cycle.

● The main contributors to metallicity are C, N, O (and Fe): Z ⇔ C,N,O.

● The bottleneck in the CNO cycle is 14N(p,γ): Z ⇔ 14N.

● During He burning 14N is converted to 22Ne:

14N(α,γ)18F(β+,νe)18O(α,γ)22Ne

CNO cycle(F. Timmes)

● When the WD is born, Z ⇔ 22Ne.

● This results in a neutron excess that scales with metallicity [Timmes et al. 03, ApJ 590, L83]:

η =1-2[ZA/A

A]= 0.101xZ

(Super)solar Metallicity for Tycho

20


● Outside the n-rich NSE region, MMn/MCr traces η in the progenitor ⇔Metallicity, WD simmering [Timmes et al. 03 ApJ 590, L83; Chamulak et al. 08, ApJ 677, 160].● In Tycho, MMn/MCr=0.74±0.47, which yields a supersolar metallicity:

Z=0.048+0.051

Z=4.0+4.1 Z⊙ (Asplund)

Z=2.6+2.7 Z⊙ (Grevesse)

-0.036

-3.0

-2.0

⇒ Large uncertainty dominated by statistics and poor atomic data.⇒ Highly subsolar values seem unlikely even with a completely convective WD.⇒ Suggests a young age for the progenitor. 20

Tycho SNR

Nordström et al. 04 A&A 418, 989

Summary

21

● Fundamental properties (X-ray spectrum and dynamics) of SNRs well reproduced by 1D DDT models (in INDEPENDENT agreement with SN spectra).

● Tycho ⇒ DDTc: Ek=1.2x1051 erg; M56Ni=0.74 M⊙; 0509-67.5 DDTa: ⇒

Ek=1.4x1051 erg; M56Ni=0.97 M⊙. Results can be verified using light echoes.

● SNR dynamics can put constraints on the mass loss from SN Ia progenitors.

● Studies of SFH in the LMC give important clues on SN Ia progenitors.

● MMn/MCr is a promising tracer of progenitor metallicity ⇒ (Super)solar metallicity for Tycho's progenitor, more to come...


X-ray observations of young Type Ia SNRs open a new window onto thephysics of the SN explosions and the properties of their progenitor systems

Gory details: Badenes et al. 06, ApJ 645, 1373; Badenes et al. 07, ApJ 662, 472; Badenes et al. 08, ApJ 680, 1149; Badenes et al. 08, ApJ 680, L33; Badenes et al. 09, ApJ 700, 727.

SNRs: The Persistence of Memory

22

The Persistence of Memory.Salvador Dalí (1931), now at MoMA (NYC)

SNR observations can probe regimes that are NOT available to SNe :● Explosion physics: Chandra can resolve the structure of the SN ejecta in Galactic SNRs (NOT equivalent to probing lines of sight!). ● Progenitors: We can identify the SNRs from dim/bright SNe and study their stellar environment and (maybe) the metallicities of their progenitors.

Tycho SNRChandra

Without SNR studies, our understanding of Type Ia SNe will never be complete.


Models vs. Observations: SNR 0509-67.5

23


Other SN Ia Explosion Models (Tycho SNR)

24


Best sub-Chandrasekhar

Ejecta+FSFS

Ejecta+FSFS

Best DDT

Ejecta+FSFS

Ejecta+FSFS

Best 3D Deflagration

Best 1D Deflagration

(Beyond the) SFHs of the Ia SNRs

25


Sullivan et al. 06, ApJ 648, 868

● Type Ia SNe explode in a variety of environments.

● Age and metallicity of the stellar populations can be measured locally (with caveats mixing of old ⇒populations due to dynamics).

● Fraction of prompt and delayed progenitors can be calculated for each SNR [assume A+B model, with B 64<t<180 Myr] ⇒ specific kind of progenitor is hard to determine!

Mn/Cr Mass Ratio as a Metallicity Tracer

26


● Variations in η do not affect the synthesis of the most abundant nuclei (with the exception of 56Ni at high Z [Timmes et al. 03, ApJ 590, L83]). ● Some trace nuclei act as 'neutron excess sinks' ⇒ Mn.● The MMn/MCr ratio is an excellent tracer of Z:

MMn/MCr = 5.3xZ0.65

DDTcZ=2x10-4

DDTcZ=9x10-3

(inner 0.2 M☉

removed)

Caveat: WD Convection and C Simmering

27


● Slowly accreting WDs ⇒ C simmering ~1000 yr before the explosion.

● Can increase η [Piro & Bildsten 08, ApJ 673, 1009], but only by Δη≃0.0015 [Chamulak et al. 08, ApJ 677, 160].

● Effect on MMn/MCr ⇒ overlap between Si-rich ejecta and convective WD core.

● The extent of the convective core is not known [Piro & Chang 08, ApJ 678, 1158]:

● The entire WD is convective ⇒ MMn/MCr has a lower limit of ~0.4.

● Limited to the C-depleted core created during He-burning [Höflich & Stein 02, ApJ 568, 779] ⇒ Dim SNe at low Z [Domínguez et al. 01, ApJ 557, 279].

Mn/Cr as Ia Neutron Excess Tracer

28


● Mn/Cr is a tracer of η in SN Ia ⇒ Window onto the progenitors.● Several datapoints ⇒ discriminate simmering and metallicity effects. Potential targets include Kepler, G337, and (maybe) the LMC SNRs.● Present X-ray missions have limitations ⇒ Con-X and XEUS (IXO) will open new possibilities.

XEUS Con-X

Kepler SNR(Suzaku)Park et al., in prep.

Cosmic Ray Acceleration at SNR Shocks

29

● FS in Tycho is very close to CD (RCD ≃

0.93RFS) ⇒ Cosmic Rays are being accelerated at the FS [Warren et al. 05, ApJ 634, 376; Cassam-Chenaï et al. 07, ApJ 665, 315].

● CR-modified dynamics cannot be studied with γ=5/3 hydro [Ellison et al. 04, A&A 413, 189].

● RS is NOT accelerating CRs:

● Not close to CD.

● Traced by hot Fe Kα

● CR acceleration at the FS does not disturb the dynamics of the shocked ejecta [Ellison et al. 07, ApJ 661, 879].

⇒ γ=5/3 HD+NEI models are appropriate for the shocked ejecta

Warren et al. 05, ApJ 634, 376

●Ellison et al. 04, A&A 413, 189


RS

CD

Shocked Ejecta

β=βmin

β=0.1

β=βmin

β=βminβ=0.1

β=0.1

CR Acceleration

30


Warren et al. 05, ApJ 634, 376

● Hot Fe Kα emission and RS/CD ratio are hard to explain if CR acceleration affects the dynamics of the shocked ejecta, but arise naturally in γ=5/3 HD+NEI when β≠0

The way ahead

31

● Multi-D HD+NEI simulations are necessary to interpret the spatially resolved spectroscopy from Chandra ⇒


Radiative losses and thermal conduction

32


DDTc ρAM=2x10-24 β=0.01

Documents

Fingerprinting SNRs: Clues to Explosion Progenitors · Inspiration Carles Badenes Caltech 08/23/09 2 Krause et al. 08, Sci 320, 1195 SNR evolution strongly modified by progenitor