1. Nucléosynthèse dans les supernovae Ia J. Isern Institut d’Estudis Espacials de Catalunya CSIC Toulouse, April 20st, 2004

2. Exploding stars They play a important role in shaping the galaxy –They inject 10 51 ergs/explosion in the form of kinetic energy –They inject several M o of freshly synthesized chemical elements –In particular radioactive elements that can be used as a diagnostic tools of the explosion

3. How to use them? Depends on their lifetime: –Prompt: Short lived radioisotopes emit γ rays that thermalize or escape. The outcome depends on M i, M eject, K eject They provide direct information about the ejecta. 56 Ni  56 Co  56 Fe ( ~ 100 days in total and 158 - 1238 keV) SN1987A –SNR: Long lived emit lower fluxes but can be detected in SNRs for centuries ( 44 Ti with 89 years and 68, 78, 157 keV) Cas A –Diffuse: They can be detected in young SNRs also, but they mix with the ISM and contribute to the diffuse galactic emission ( 26 Al 1.1x10 6 yr and 1809 keV) Galactic plane In all cases the abundances of the radioactive isotopes provide information about the pre-explosion state and the explosion itself But, why stars explode? How they explode?

4. Hydrostatic Equilibrium Characteristic times Hydrodynamic time:  HD  440  -1/2 Thermal time: 10 7 yr Nuclear time: 10 9 yr

5. Electron degeneracy At high densities e - are dominant If Even at T=0 electrons (and other fermions) are able to exert pressure! Zero temperature structures can exist

6. The virial theorem P=2/3 eP=1/3 e E i = -1/2 E G E i = -E G Non Relativistic Particles Extremely Relativistic Particles During a gravitational transition from an equilibrium configuration to another one, half of the energy is radiated away and half is invested in internal energy. Relativistic stars are not bounded M Ch =1.44 2 M o

7. Nuclear reactions Virial theorem  E i  E E i ~ MT E G ~ M 2 R -1 T ~ M/R  ~ T 3 M -2  ~ M R -3 Each burning phase occurs at a fixed temperature  ~M -2 Light stars ignite nuclear reactions at high densities Electron degeneracy can stop the nuclear burning process M<0.08 Mo, H is never ignited M<0.5 Mo, He is never ignited M<8-9 Mo, C is never ignited M<10-12 Mo, Ne is never ignited M>10-12 Mo, Fe cores are formed

8. Non relativistic electrons If electrons are non relativistic Hydrostatic equilibrium: It is always possible to find an equilibrium structure The star only needs to contract R decreases when M increases

9. Relativistic electrons If electrons are relativistic Hydrostatic equilibrium: It is not possible to find an equilibrium structure There is not a length scale If  E < 0  R < 0 The star contracts If  E > 0  R > 0 The star contracts The ideal scenario for catastrophic events !

10. Electron captures Nuclear energy release # The energy losses by electron captures depend on the ignition density # The injected energy depends on the velocity of the burning front He cores always explode CO cores can explode or collapse ONe cores always collapse Fe cores always collapse

11. Explosive sources of energy Gravitational collapse Electron degenerate core Neutron star M ~ 1.4 Mo R ~ 10 8 -10 9 cm M ~ 1.4 Mo R ~ 10 6 cm  E G ~ 10 53 erg Thermonuclear explosion { 12 C, 16 O}  { 56 Ni} q ~ 7x10 17 erg/g 1 Mo x q ~ 10 51 erg

12. Supernova taxonomy Fireworks are determined by: The amount of radioactive elements The size and extension of the envelope surrounding the degenerate core

13. Observational constraints (SNIa). I H must be absent at the moment of the explosion –There are some evidences of H-lines before maximum or at late epochs Progenitors should be long lived to account for their presence in all galaxies, including ellipticals The explosion should produce at least ~ 0.3 M 0 of 56 Ni to account for the light curve and late time spectra SNIa are caused by the explosion of a C/O white dwarf in binary systems (He white dwarfs detonate and are completely incinerated ONe collapse to a neutron star)

14. Supernova light curve 56 Ni(8.8d) 56 Co(113.7d) 56 Fe 56 Ni 56 Co

15. Observational constraints. II Intermediate elements must be present in the outer layers to account for the spectrum at maximum lig ht The burning must be subsonic. It can be supersonic only if ρ < 10 7 g/cm 3 The abundances of the iron peak elements ( 54 Fe, 58 Ni, 54 Cr) must be compatible with the Solar System abundances after mixing with gravitational supernova products Neutron excesses have to be avoided: Post-burning e - -captures Neutrons stored as 22 Ne Decrease ignition density Decrease 22 Ne content Reduce the SNIa galactic contribution

16. Observational constraints. III Homogeneity? –Differences in brightness: Overluminous (SN 1991T), underluminous (SN1991bg) –Differences in the expansion velocity (v exp ~ 10,000-15,000 km/s) Two points of view: –There is a bulk of homogeneous supernovae plus some peculiar ones –SNIa display a continuous range of values Is there a unique scenario & unique mechanism able to accomodate the normal behavior plus that of dissidents? Is there a mechanism able to produce a continuous range of situations? Can both mechanisms coexist? Evrything able to explode eventually do it !!!

17. Thermonuclear runaways The necessary condition is that the energy must be released in a time shorter than the dynamical time Nuclear heating time: The hydrodynamical time: In hydrostatic equilibrium The instability condition is: Deflagration temperature

18. Stars stabilize their fuel by means of adiabatic expansions The efficiency of the adiabatic cooling is defined as the expansion, δρ, experienced to restore pressure equilibrium In the gas ideal case: Since Q ~ 1 MeV and kT ~ 1-100 keV adiabatic cooling is very efficient and stars are stable Where ΔX is the amount of burned fuel If the electronic degenerate component is dominant: Cooling is only efficient if Thermonuclear runaways occur if: T def < T F H is not a good explosive because it needs weak interactions to convert p in n (novae) He and C are good explosives (supernovae)

19. The burning front l Δ (ρ,ε) 0 (ρ,ε) 1 Burned Unburned Δ << l Mass, momentum and energy conservation Two types of solutions Detonation: v front supersonic versus the unburned material sonic or subsonic versus the burned material Deflagration: v front always subsonic versus the unburned & burned material

20. Deflagrations In spherical symmetry burned material is at rest at the center, v 1 = 0 Assume unburned material at rest, v 0 = 0 Mass and momentum conservation demands: V 0 (P 1 - P 0 ) = u 0 (u 0 - u 1 ) or V 0 (P 1 - P 0 ) = u 1 D in the frame at rest But if P 1 - P 0 0 then v 1 < 0 in contradiction with the hypothesis A deflagration can only exist if it generates a shock precursor that burst matter outwards! The star expands and the flame is quenched If  < 10 7 g/cm 3 the hypothesis of a thin burning front is no longer valid and matter is not completely incinerated

21. Which are the progenitor systems? How a WD manages to reach the Chandrasekhar limit? Which is the evolution of the WD before the thermonuclear runaway? How the flame propagates throughout the WD? –How much nuclear energy is converted into kinetic energy during the flame quenching (  < 10 7 g/cm 3 ) –Is it possible a deflagration detonation transition? How the light curves, spectra & remnants look like in 3D? Are rotation and magnetic fields relevant? SNIa are caused by the explosion of a C/O white dwarf in a binary system with a mass near the Chandrasekhar’s limit (He white dwarfs detonate and are converted in Fe and ONe collapse to a neutron star)

22. The outcome depends on: Accretion rate Chemical composition of accreted matter CO accreting WD Merging of two CO WD. The outcome depends on the rate. AIC? He accreting WD 10 -9 < M t < 5x10 -8 M o yr -1 off center detonation CO+He star, normal or degenerte H accreting WD M t < 10 -9 Mo yr -1 Nova 1o -9 < M t <10 -6 M o yr -1 Steady burning or weak flashes. He detonation in some cases 10 -6 < M t < M t (Eddington) Red giant and common envelope

23. Exploding mechanisms Detonation: supersonic flame If ρ > 10 7 g/cc  C,O  Ni If ρ  10 7 g/cc  C,O  Si,Ca, S,... Deflagration: subsonic velocity laminar flame: v ~ 0.01 c s Turbulent flame: v ~ 0.1 - 0.3 c s Other possibilities: Deflagration + detonation Pulsating delayed detonation The laminar flame becomes turbulent: * Rayleigh-Taylor instability * Kelvin- Helmholtz Flame surface increases effective velocity increases

24. Exploding mechanisms: Off center ignition detonation Shock wave M CO ~ 0.5 - 1.1 M 0 M He ~ 0.2 - 0.3 M 0

25. Main properties of the ejecta Can γ-ray astronomy provide new information? DET DEL SUB DEF Additionally: Each one of these models rely on several parameters that introduce further variety in the nucleosynthesis products Sub-Chandrasekhar models produce large amounts of 44 Ti

26. 20d D = 5 Mpc Spectral evolution for: Deflagration Delayed detonation Detonation Sub-Chandrasekhar DET DEL DEF SUB DEF only shows the continuum DEL,DET,SUB display strong lines 56 Ni still present 56 Ni & 56 Co prominent in SUB/DET Because of the presence of low Z elements, continuum extends to lower energies in DEF & DEL

27. 60d D = 5 Mpc Spectral evolution for: Deflagration Delayed detonation Detonation Sub-Chandrasekhar DET DEL DEF SUB 56 Ni lines have disappeared 122 - 136 keV 57 Co lines appear the line intensity in DEL,DET, SUB  m of radioactive elements the energy cut-off of DEF is still low

28. 120d D = 5 Mpc Spectral evolution for: Deflagration Delayed detonation Detonation Sub-Chandrasekhar Optically thin Continuum dominated by annihilation DET DEL DEF SUB

29. DET DEL SUB 158 keV 56 Ni DET DEL SUB Light curves 1 Mpc

30. DET DEL DEF SUB 1 Mpc Light curves 847 keV 56 Co

31. DET DEL DEF SUB Light curves 1238 keV 56 Co 1 Mpc

32. SN 1998bu M96 d = 8- 12 Mpc (From Diehl’02)

33. Line profiles for the 847 keV line D = 1 Mpc t = 120 d DET DEF DEL SUB

34. Narrow line sensitivity for SPI (3σ in 10 6 s). The 511 keV line is not shown Degradation factor for broad lines

35. D = 5 Mpc t int =10 6 s Simulated observational spectra

36. Original linewidth for each model Maximum distances at which the linewidths can be distinguished DET DEL DEF SUB Given a distance and a model a signal is generated and convolved with the response of SPI. A random background fluctuation is added, and a line width is assigned. The red double arrow contains the 90% of the line widths obtained in this way.

37. CONTINUUM Distance for detecting differences in the flux for several bands at 30 d (Mpc)

38. SNIa models Deflagration k is a free parameter Delayed detonation Flame artificially accelerated until a detonation occurs Pulsating delayed detonation The same but after bounce Additionally: Each one of these models rely on several parameters that introduce further variety in the nucleosynthesis products

39. r(cm)  (g/cm 3 )

41. Dotted line is the normalized density:  /  max M ex = 1.38 M  M ni = 1.22 M  M ex = 0.97 M  M ni = 0.50 M 

42. Dotted line is the normalized density:  /  max  tr = 3.9 x 10 7 g/cc, M ni = 1.03 M   tr = 1.3 x 10 7 g/cc, M ni = 0.56 M 

43. Dotted line is the normalized density:  /  max k = 0.06, M ni = 0.55 M  k = 0.16, M ni = 0.75 M 

44. Dotted line is the normalized density:  /  max  tr = 4.4 x 10 7 g/cc, M ni = 1.11 M   tr = 7.7 x 10 7 g/cc, M ni = 0.58 M 

45. Light curves obtained from the different families of solutions overlap

46. Line profiles of the 847 keV line 70 days after the explosion

47. INTEGRAL can provide useful information for SNIa, provided they occur close enough, on the basis of: –Line light curves –Properties of the continuum –Line profiles Because of the poor understanding of the flame properties we have to use parametrized values. In this case there are ambiguities in the information provided by the gamma- rays alone. Independent information is necessary

48. Test for Sub-Chandrasekahr’s models The decay of surface 56 Ni produces Fe-Co-Ni K  emission in Sub-Ch models but not in the other models 20d A typical event placed at 15 Mpc produces a flux of ~ 1.82 x 10 -7 ph/s/cm 2 which is on the limit of capacity of XMM

49. Departures from sphericity? The physics of the flames indicates they are unstable and departures from sphericity easily appear Nevertheless, observations suggest that departures from sphericity are small: –The homogeneity of the light curve s & spectra indicates photospheric perturbations < 10% –Standard SNIa show small polarizations (Wang et al 2001) –But subluminous display polarizations ~ 0.7% (Howell et al 2001)

50. Temperature evolution on the plane xy for six different times (Bravo et al IEEC/UPC)

51. T & P at the end of the deflagration phase (t=1.55 s) The front structure is not homogeneous but the envelope evolves spherically. Large pockets of unburned matter are left that introduce irregularities in the line profiles If the deflagration turns out into a detonation these pockets disappear

52. Ignition with 30 random bubbles: T, from 0 to 1.1 s

53. C-O Mg-Si 56 Ni

54. In general are not energetic enough!

55. 3D deflagration model: Line profiles for 847 keV 25 days 70 days

56. Central deflagration: T; time = 0 – 1.55 s; = 400 km

57. The detonation starts: t = 1.55 -2.06 s

58. Model B Model D Model A Model C Final configuration: C-O, Mg-Si, 56 Ni Central ignition: Delayed Detonation Mass fraction vs velocity M C-O M 56Ni EKEK Model 0.650.320.33D 0.510.430.51C 0.550.400.48B 0.390.540.75A

61. Devices with high spectral resolution and enough sensitivity can provide information about 3D effects: Inhomogeneities Large scale structure Gamma lens

62. Can  -rays reveal large scale properties? If the initial structure clearly departs from sphericity, important variations in the intensity of the lines should appear

63. Off center detonations (E. Bravo et al IEEC/UPC) (temperature)

64. 56 Ni distribution on 6 perpendicular directions to the axis

65. Five ignition points at random The 56 Ni distribution is slightly more isotropic One ignition point The 56 Ni distribution is not isotropic

66. Sub-Ch: 1 ignition point, 847 keV line, 40 days Solid : 0 dashed: 90 dotted: 180 Solid: 1D Dashed: 3D-180

67. Departures from sphericity? Since SNIa occur in binary systems, large scale departures from sphericity can occur (0.8,0.6) M o

68. at = 8 s Bravo, García-Senz, Serichol (IEEC/UPC)

69. At = 21 s

70. At = 99 s

71. At = 160 s

72. At = 209 s

73. At = 277 s

74. At = 337 s

75. At = 404 s

76. At = 531 s

77. At =609 s

78. At = 918 s

82. At = 1696 s

83. Single scenario

84. No hole Hole 847 keV light curve

85. Hole, day 20

86. Hole

88. Interaction with the interstellar medium SNIa are not associated with dense regions of ISM, but they can meet clouds (l) line (i) ejecta (j) ISM ---- DET _._ SUB 4439 keV 12 C If 100 cm -3 ~ 1.5 kpc during 150 yr If 10 cm -3 ~ 0.5 kpc during 150 yr 10 kpc

89. Interaction with the Interstellar Wind Symbiotic scenario This scenario seems not to be a source of detectable emission by nuclear excitations! D = 10 kpc