Francisco Förster Collaborators: Pierre Lesaffre, Philipp Podsiadlowski Padova - July 2011 Hydrostatic  12 C burning in CO WDs: the  simmering  phase
CHASE 50 cm robotic telescope  (SN search and follow up)
Start with BPS (~10 8-10  yr) Accrete to M Ch (~10 6  yr) Add ignition bubbles (few sec) DDT it (~1-0.1 s) Simulate spectrum... Compare and repeat... Cooking a Type Ia Supernova (SD M ch ) ~ Hydrostatic evolution Hydrodynamic evolution Simmer (~10 3  yr)
Path to Type Ia Supernova -  Close binary system formation:  CO WD + companion -  Mass transfer phase:  accretion + shrinking + heating -  Hydrostatic  12 C burning:  energy release + ashes pollution -  Simmering phase (C-flash):  convective core growth + steep L gradient -  Thermonuclear runaway + ignition -  Deflagration + detonation C-flash C-flash SN Ia ? CO WD CO WD CO WD
Cooling ~10 7 -10 10  yr Heating (accretion) ~10 6  yr Simmering (carbon burning) ~10 3  yr T c  [K] r c  [g cm -3 ] Hydrostatic  central evolution t burn  =  t conv e burn  =  e n
Cooling  (~ 1e8-10 yr) Simmering  (~1000 yr) Accretion  (~1e6 yr) Explosion  (~1 sec)
Cooling ~10 7 -10 10  yr Heating (accretion) ~10 6  yr Simmering (carbon burning) ~10 3  yr T -  r
s -   r
Problem 1: Diversity among galaxy types Dim, fast decliners. Early type galaxies Bright, slow decliners. Late type galaxies Progenitor evolution prediction: central density at ignition main parameter Lesaffre et al. 2006,  c.f. Townsley et al. 2009, Seitenzhal et al. 2011 C+O Si/S 56 Ni 54 Fe C+O Si/S 56 Ni 54 Fe
Lesaffre et al. 2006 Meng et al. 2009, 2011 Z Age effect Mass effect
Old progenitor Higher  r c   @ ignition Higher M WD   (BPS) Lower initial C/O ratio Less C/O~1 matter accreted to be mixed into the centre Higher Z environment Dimmer SNe Lower central C/O ratio @ignition
Problem 2:  Quantify neutronization More  metals  in the CO WD leads to more  stable NSE  matter in the ejecta (not radioactive). However, the mass of  iron group elements  (main  source of opacity ) would be constant  (Mazzali & Podsiadlowski 2008) 54 Fe,  58 Ni 56 Ni L k   (decline rate) = This could  break the one parameter  law that makes SNe Ia standardizable candles. ` Z
23 Na produced when  12 C burns At high densities (M~M Ch ), 23 Na can  become  23 Ne after capturing an electron. Add convection: fresh  23 Na can move to high density regions and be neutronized. Not enough time for  23 Ne to be de-neutronized in  low density regions. Convective Urca process Ratio between  23 Na e -  captures and  23 Ne  b  decays
t conv  ~   t EC   <  t burn Need time dependent theory of convection:  Lesaffre et al. 2005  Need accurate, but simplified nuclear physics:  Förster et al. 2010 (ApJS) 23 Ne Urca shell 23 Na Convective core Convective core Convective core Convective core e- captures n
Zingale et al. 2009 Convection in hydrodynamic simulation (MAESTRO) No e -  captures.
Nuclear physics  -  Burning prior to ignition : approximately six  12 C nuclei become  20 Ne,  23 Na,  16 O and  13 C, with one or two electron captures (Piro et al. 2008) New nuclear network 1. Reduced number of variables 2. Accuracy in relevant quantities 3. Remove short time-scales if possible Zegers et al. 2008 Förster et al 2010 ApJS
Simplified   networks We define four types of time-scales: 1.  Trace nuclei:  n, p,   a ,  13 N 2.  23 Ne  b- decays and  23 Na e- captures 3.  12 C burning 4.  Convection 2 One can show   t 1  <  t 2 ,  t 3 ,  t 4 3 3 1 1 1 1
Detailed network abundances well reproduced. Förster et al 2010 ApJS
Stellar evolution code (preliminary)
Central temperature vs central density tracks t burn  =  t conv
Neutronization
Time scales
Implications for cosmology Simple model:  56 Ni +  54 Fe  post explosion composition X ( 56 Ni) = 1 –  X ( 54 Fe),  h f  =  X ( 54 Fe)  h ( 54 Fe) Because the explosion is very fast,  h f  ~  h 0 X ( 56 Ni) = 1 –  h 0 / h ( 54 Fe) = 1 – 27  h 0 h 0  =  h formation   +  Dh preWD   +  Dh Urca h formation  ~  X ( 56 Fe)  h ( 56 Fe)    ~ 3e-6  @ Z=0.02 Dh preWD  ~  X ( 22 Ne)  h ( 22 Ne)   ~ 22 [ X ( 12 C)/12 +  X ( 14 N)/14 +  X ( 16 O)/16)]  h ( 22 Ne)   ~ 1.8e-3  @ Z=0.02 Dh Urca  ~  X ( 13 C)  h ( 13 C) +  X ( 23 Na)  h ( 23 Na) +  X ( 23 Ne)  h ( 23 Ne)   up to ~1.2e-3 27 (3e-3) ~ 0.09 Neutronization  can change M( 56 Ni) up to 9%. 14 N( a , g ) 18 F( b +) 18 O( a , g ) 22 Ne
Implications for progenitor scenarios Type of ashes Nebular  spectra! > 10 8 > 6 x 10 9 5 x 10 7-8 5-6 x 10 9 < 5 x 10 7 2-3 x 10 9 < 5 x 10 7 3-5 x 10 9 Maeda et al. 2010 C+O IME 56 Ni Stable IGE Deflagration Detonation SD Detonation DD Detonation Sub Chandra r   [g cm -3 ] T max  [K]
Pakmor et al. 2010
~10 x M( 56 Ni) M( 54 Fe) M( 56 Ni) M(NSE) M(IME) HVG LVG faint 56 Ni IME C+O 54 Fe NSE
Conclusions / Discussion - Hydrostatic  12 C burning:   Relevant reactions understood (Piro et al. 2008, Chamulak et al. 2008 and  Förster et al. 2010 ).  Simplified nuclear network implemented in simultaneous structure+chemistry stellar evolution  code. - Neutronization: Pre WD enrichment (Timmes et al. 2003) important above solar metallicities  (Chamulak et al.  2008). Urca process important at lower metallicities (cosmological sample). Older systems would  undergo stronger neutronization. - Diversity between early and late type host galaxy SNe Ia: Progenitor evolution suggests central density as main parameter for SN Ia light curves  (Lesaffre et al. 2006, c.f. Townsley et al. 2009, but Seitenzhal et al. 2011) - Caveats: Need to include effects of heat diffusion from the accreting envelope (Nomoto 1984). Simplified energy equation, feedback into convective velocities work in progress. - If heat diffusion slow: More stable IGE in older/dimmer events (c.f. DD/sub-Chandra: no stable IGE?) Galactic SN Ia Fe enrichment faster than SN Ia Ni enrichment. More ignition points (strong def. + weaker det.) with increasing central density?
Stein & Wheeler  2006,  with 1000x Urca and nuclear rates No URCA process URCA process The Urca process can also affect the  convective velocities  (Lesaffre et al. 2005)
Neutron excess Relative excess of neutrons over protons: h  = (n - p) / (n + p) h mix  =  S   h i   X i The neutron excess can only be changed by weak forces, which are generally slower than strong forces. In fact, during a supernova explosion the neutron excess does not change, except in the inner ~0.1 Msun where the density is very high.
Detailed reaction network Flows at fixed  r , T 4 X 10 8  K, 3 X 10 9  g cm -3
Metallicity of the host decreases with redshift Sullivan et al. 2010
NSE IME C+O SN Ia ejecta composition (M Ch ) C+O:   unburnt material IME:   intermediate mass elements (Si, S, Mg), or partially burnt C NSE:   nuclear statistical equilibrium matter ( 56 Ni), or fully burnt C HD NSE:   High density NSE ( 54 Fe,  58 Ni), or fully burnt C with electron captures. HD NSE Radioactive ashes

2011 padova

  • 1.
    Francisco Förster Collaborators:Pierre Lesaffre, Philipp Podsiadlowski Padova - July 2011 Hydrostatic 12 C burning in CO WDs: the simmering phase
  • 2.
    CHASE 50 cmrobotic telescope (SN search and follow up)
  • 3.
    Start with BPS(~10 8-10 yr) Accrete to M Ch (~10 6 yr) Add ignition bubbles (few sec) DDT it (~1-0.1 s) Simulate spectrum... Compare and repeat... Cooking a Type Ia Supernova (SD M ch ) ~ Hydrostatic evolution Hydrodynamic evolution Simmer (~10 3 yr)
  • 4.
    Path to TypeIa Supernova - Close binary system formation: CO WD + companion - Mass transfer phase: accretion + shrinking + heating - Hydrostatic 12 C burning: energy release + ashes pollution - Simmering phase (C-flash): convective core growth + steep L gradient - Thermonuclear runaway + ignition - Deflagration + detonation C-flash C-flash SN Ia ? CO WD CO WD CO WD
  • 5.
    Cooling ~10 7-10 10 yr Heating (accretion) ~10 6 yr Simmering (carbon burning) ~10 3 yr T c [K] r c [g cm -3 ] Hydrostatic central evolution t burn = t conv e burn = e n
  • 6.
    Cooling (~1e8-10 yr) Simmering (~1000 yr) Accretion (~1e6 yr) Explosion (~1 sec)
  • 7.
    Cooling ~10 7-10 10 yr Heating (accretion) ~10 6 yr Simmering (carbon burning) ~10 3 yr T - r
  • 8.
  • 9.
    Problem 1: Diversityamong galaxy types Dim, fast decliners. Early type galaxies Bright, slow decliners. Late type galaxies Progenitor evolution prediction: central density at ignition main parameter Lesaffre et al. 2006, c.f. Townsley et al. 2009, Seitenzhal et al. 2011 C+O Si/S 56 Ni 54 Fe C+O Si/S 56 Ni 54 Fe
  • 10.
    Lesaffre et al.2006 Meng et al. 2009, 2011 Z Age effect Mass effect
  • 11.
    Old progenitor Higher r c @ ignition Higher M WD (BPS) Lower initial C/O ratio Less C/O~1 matter accreted to be mixed into the centre Higher Z environment Dimmer SNe Lower central C/O ratio @ignition
  • 12.
    Problem 2: Quantify neutronization More metals in the CO WD leads to more stable NSE matter in the ejecta (not radioactive). However, the mass of iron group elements (main source of opacity ) would be constant (Mazzali & Podsiadlowski 2008) 54 Fe, 58 Ni 56 Ni L k (decline rate) = This could break the one parameter law that makes SNe Ia standardizable candles. ` Z
  • 13.
    23 Na producedwhen 12 C burns At high densities (M~M Ch ), 23 Na can become 23 Ne after capturing an electron. Add convection: fresh 23 Na can move to high density regions and be neutronized. Not enough time for 23 Ne to be de-neutronized in low density regions. Convective Urca process Ratio between 23 Na e - captures and 23 Ne b decays
  • 14.
    t conv ~ t EC < t burn Need time dependent theory of convection: Lesaffre et al. 2005 Need accurate, but simplified nuclear physics: Förster et al. 2010 (ApJS) 23 Ne Urca shell 23 Na Convective core Convective core Convective core Convective core e- captures n
  • 15.
    Zingale et al.2009 Convection in hydrodynamic simulation (MAESTRO) No e - captures.
  • 16.
    Nuclear physics - Burning prior to ignition : approximately six 12 C nuclei become 20 Ne, 23 Na, 16 O and 13 C, with one or two electron captures (Piro et al. 2008) New nuclear network 1. Reduced number of variables 2. Accuracy in relevant quantities 3. Remove short time-scales if possible Zegers et al. 2008 Förster et al 2010 ApJS
  • 17.
    Simplified networks We define four types of time-scales: 1. Trace nuclei: n, p, a , 13 N 2. 23 Ne b- decays and 23 Na e- captures 3. 12 C burning 4. Convection 2 One can show t 1 < t 2 , t 3 , t 4 3 3 1 1 1 1
  • 18.
    Detailed network abundanceswell reproduced. Förster et al 2010 ApJS
  • 19.
  • 20.
    Central temperature vscentral density tracks t burn = t conv
  • 21.
  • 22.
  • 23.
    Implications for cosmologySimple model: 56 Ni + 54 Fe post explosion composition X ( 56 Ni) = 1 – X ( 54 Fe), h f = X ( 54 Fe) h ( 54 Fe) Because the explosion is very fast, h f ~ h 0 X ( 56 Ni) = 1 – h 0 / h ( 54 Fe) = 1 – 27 h 0 h 0 = h formation + Dh preWD + Dh Urca h formation ~ X ( 56 Fe) h ( 56 Fe) ~ 3e-6 @ Z=0.02 Dh preWD ~ X ( 22 Ne) h ( 22 Ne) ~ 22 [ X ( 12 C)/12 + X ( 14 N)/14 + X ( 16 O)/16)] h ( 22 Ne) ~ 1.8e-3 @ Z=0.02 Dh Urca ~ X ( 13 C) h ( 13 C) + X ( 23 Na) h ( 23 Na) + X ( 23 Ne) h ( 23 Ne) up to ~1.2e-3 27 (3e-3) ~ 0.09 Neutronization can change M( 56 Ni) up to 9%. 14 N( a , g ) 18 F( b +) 18 O( a , g ) 22 Ne
  • 24.
    Implications for progenitorscenarios Type of ashes Nebular spectra! > 10 8 > 6 x 10 9 5 x 10 7-8 5-6 x 10 9 < 5 x 10 7 2-3 x 10 9 < 5 x 10 7 3-5 x 10 9 Maeda et al. 2010 C+O IME 56 Ni Stable IGE Deflagration Detonation SD Detonation DD Detonation Sub Chandra r [g cm -3 ] T max [K]
  • 25.
  • 26.
    ~10 x M(56 Ni) M( 54 Fe) M( 56 Ni) M(NSE) M(IME) HVG LVG faint 56 Ni IME C+O 54 Fe NSE
  • 27.
    Conclusions / Discussion- Hydrostatic 12 C burning: Relevant reactions understood (Piro et al. 2008, Chamulak et al. 2008 and Förster et al. 2010 ). Simplified nuclear network implemented in simultaneous structure+chemistry stellar evolution code. - Neutronization: Pre WD enrichment (Timmes et al. 2003) important above solar metallicities (Chamulak et al. 2008). Urca process important at lower metallicities (cosmological sample). Older systems would undergo stronger neutronization. - Diversity between early and late type host galaxy SNe Ia: Progenitor evolution suggests central density as main parameter for SN Ia light curves (Lesaffre et al. 2006, c.f. Townsley et al. 2009, but Seitenzhal et al. 2011) - Caveats: Need to include effects of heat diffusion from the accreting envelope (Nomoto 1984). Simplified energy equation, feedback into convective velocities work in progress. - If heat diffusion slow: More stable IGE in older/dimmer events (c.f. DD/sub-Chandra: no stable IGE?) Galactic SN Ia Fe enrichment faster than SN Ia Ni enrichment. More ignition points (strong def. + weaker det.) with increasing central density?
  • 28.
    Stein & Wheeler 2006, with 1000x Urca and nuclear rates No URCA process URCA process The Urca process can also affect the convective velocities (Lesaffre et al. 2005)
  • 29.
    Neutron excess Relativeexcess of neutrons over protons: h = (n - p) / (n + p) h mix = S h i X i The neutron excess can only be changed by weak forces, which are generally slower than strong forces. In fact, during a supernova explosion the neutron excess does not change, except in the inner ~0.1 Msun where the density is very high.
  • 30.
    Detailed reaction networkFlows at fixed r , T 4 X 10 8 K, 3 X 10 9 g cm -3
  • 31.
    Metallicity of thehost decreases with redshift Sullivan et al. 2010
  • 32.
    NSE IME C+OSN Ia ejecta composition (M Ch ) C+O: unburnt material IME: intermediate mass elements (Si, S, Mg), or partially burnt C NSE: nuclear statistical equilibrium matter ( 56 Ni), or fully burnt C HD NSE: High density NSE ( 54 Fe, 58 Ni), or fully burnt C with electron captures. HD NSE Radioactive ashes