Transcript Ia 超新星的
The Explosion Models and Progenitors of Type Ia Supernovae Wang Xiao Feng NAOC 2003. 10. 21 Outlines Introduction of SN Taxonomy Observational constraints on the basic models of SNe Ia Composition of exploding WDs Mass, the birth and propagation of thermonuclear flame of exploding WDs Progenitor systems of SNe Ia SN classifications Taxonomy Chart Spectroscopic classifications Observational constraints Observational characteristics of SNe Ia The most abundant elements of hydrogen and helium in the universe fail to appear in the spectra, and a substantial wide absorption lines of intermediatemass elements (O-Ca) dominate the spectra near maximum light The light curve peak lasts for several days, and displays exponential decline at late time Most of SNe Ia show relatively similar spectra and light curves shapes, but definite departures from the canonical events have also been observed. SNe Ia explosion have been detected in galaxies of all Hubble types Light curves and spectra of different SNe Ia What can be inferred from observations? The absence of the emissions of hydrogen requires that the atmosphere of the exploding star contains no hydrogen or very few (i.e. < 0.1 ), which point towards highly evolved compact objects. The kinematic energy per mass, ½(~ 10,000 km s-1)2, inferred from the velocities of the ejecta in the explosion is the same magnitude as the energy of fusing carbon and oxygen into Fe-group elements. The shape of SN Ia light curves follow very well with the energy model of radioactive-decay (56Ni-56Co-56Fe) The appearance in elliptical galaxies with their old populations hints at significant that nuclear processing must take place before explosion. The fact that the event is explosive suggests the existence of degenerate matter SN 2002fk in NGC A pinpoint of light from a type Ia supernova that exploded more than 10 billion years ago. The supernova was revealed by digitally subtracting before and after images of a faint elliptical galaxy that appears in the HST deep Field image. SNe Ia represent the thermonuclear disruption of mass accreting WD. Almost all researchers in this field have reached an unanimous consensus on this basic model of SN Ia explosion. However, the precise nature of the hydrodynamical models and the progenitor systems are still controversial. Composition of exploding WDs WDs form at the end of the evolution of stars whose original masses are less than 8 Msun.A star can always lose a large fraction of its material by ejecting outer layers into space at the final stages of evolution. The mass of a remaining WD is always less than the Chandrasekhar limit, 1.4 Msun, above which a hydrostatic equilibrium of degenerate matter is impossible. An isolated WD is stable and almost inert, because its temperature is not high enough to induce any substantial nuclear reactions. This isolated dead star can exist almost indefinitely, slowly cooling down to black dwarf as it radiates its energy into space. No supernova explosion will ensue. A very different fate awaits a WD that accretes mass from a close binary companion star. Observations show, however, that more than 50% of all stars are not isolated. They belong to groups of two or three stars that orbit a common center of mass. In a close binary system, a WD can increase its own mass by accreting material from its companion star. Such systems are considered to be the most probable SN Ia progenitors. In principle, the WDs that accretes to the explosion could be composed of He, of C-O or of O-Ne-Mg. (i) helium (He) WDs, composed almost entirely of helium , form as the degenerate cores of low-mass giants (M <2 Msun. ) which lose their hydrogen envelope before helium can ignite; (ii) carbon-oxygen(C-O) WDs, composed of about 20% C and 80% oxygen, form as the cores of asymptotic giant branch (AGB) stars or naked helium burning stars that lose their envelope before carbon ignition; (iii) oxygen-neon-magnesium (O-Ne-Mg) WDs, composed of heavier combinations of elements, form from giants that ignite carbon in their cores. The fate of accreting WDs Mass, ignition and propagation of flames of exploding WDs Chandrasekhar-mass model When the central density approach the critical value (2109 g cm-3), the thermonuclear reaction rate exceeds the energy loss neutrino the ignition of 12C + 12C takes place in the center due to the compressional heating. The release energy further increases the temperature, thus further accelerating thermonuclear reactions. This process is slowed down by neutrino cooling and by the convective and conductive heat exchange. Nevertheless, the temperature in the WD center rises and reaches the point where the energy release overwhelms the energy outflow. Under the condition of strong electron degeneracy , the thermonuclear reaction after ignition is unstable and explosive. The basic physics mechanism is clear: the Fermi pressure of degenerate gas is not sensitive to the change of temperature. At the initial phase, all of energy released by nuclear reaction is used to increase the temperature, while the pressure remains almost unchanged. As a result the strong temperature-dependent nuclear reaction rate increases rapidly (i.e. T and is usually very large) which would further increase the temperature circularly, and the reaction becomes eventually thermal runaway until degeneracy disappears. In the C-O core of electron degeneracy, it is expected that nuclear burning is explosive after carbon ignition since the energy released at the burning point can not be taken away immediately. when the degenerate C-O is ignited, the burning is explosive. The burning front will propagate into the surrounding fuels by subsonic deflagration or supersonic detonation, which depends on the overpressure produced by the burning material. Based on the above two basic propagation modes of thermonuclear burning, different hydrodynamical models have been proposed: Prompt detonation : Arnett proposed the first hydodynamical model that the thermonuclear burning start from a detonation wave, which burns the whole star with a supersonic velocity. The resulting nuclear synthesis is contradict with observations (without intermediate-mass elements) Pure deflagration: (Nomoto W7 model, 20%-30% of Vc, one of the most successful model, which can give reasonable nuclear synthesis results, e.g. Large amounts of intermediate mass elements Ca-S-Si, ONe-Mg etc. problems: overproduction of neutron-rich isotopic elements (58Ni, 54Fe, 54Cr), e.g. W7 model gives 58Ni, 54Cr, 4-5 times higher than solar isotopes. The overproduction of Fe-isotopes Delayed detonation Inspired by the terrestrial combustion experiments that turbulence deflagrations can sometimes be observed to undergo spontaneous transitions to detonations (DDT), it was suggested that this process may occur in the late phase of a Mch-explosion. The delayed detonation models assume that the early propagation of the deflagration is as low as a few percent of the sound speed required to preexpand the star, followed by a transition to a shock-driven, supersonic detonation mode that produces large amounts of high-velocity intermediate-mass elements. Many 1D simulations have demonstrated that the delayed detonation models can reproduce well the features of observed SN Ia spectra and light curves, as well as reasonable nucleosynthesis. However, the physical mechanisms by such DDTs occur are unclear. Pulsational delayed detonation If the initial deflagration phase fails to release energy to unbind the star and no DDT takes place during the expansion, the star undergoes a large amplitude of pulsation by one or more times. The following contraction may trigger a detonation by compression heating, eventually the WD is completely disrupted. Litte Ni but a subtantially amount of Si and Ca. Weak SN Ia explosion SubChandrasekhar-mass model C-O WD below the Chandradekhar mass do not reach the critical density and temperature for explosive carbon burning by accretion, and therefore need to be ignited by an external trigger. In this model, also known as Edge Lit Detonation(ELD), the first nuclear ignition takes place near the bottom of the accretion helium layer of about 0.15-0.20Msun. A prompt detonation propagates outwards through the helium, while an inward non-burning pressure wave compressed the C-O core which ignites off-center and derives a second detonation outwards through the C-O core. Owing to the difference between the nuclear kinematics of carbon and helium burning these models have a composition structure that is fundamentally different from that of carbon ignitors. 4He burns to 12C by the slow triple alpha process and as soon as 12C is formed it rapidly captures alpha particles to form 56Ni, so the original helium layer ends up as a high-velocity mixture of 56Ni and leftover 4He. The ELD models are mainly favored by required statistics, since less mass needs to be accreted, and WD does not need to be extremely massive. Turbulent flame surface The WD Near the Chandrasekhar limit When the mass of the WD approaches the Mch, the pressure of degenerate electrons could not resist the gravity. Any small mass increase results in a substantial contraction of the star, and this increases the density and temperature in the center of C-O core rapidly. The energy balance near the center is determined by the neutrino losses and the compressional heating. The evolution of entropy and temperature near the core of the WDs may be affected by the convective URCA process (a convectively driven electron capture-beta decay cycle leading to neutrinoantineutrino losses). Progenitors of SN Ia What are the progenitor systems (or pre-supernovae) of exploding SNe Ia, and how they evolve towards explosion? This is the key problem in stellar evolution and remains unresolved yet. In contrast to SN II from collapsing of massive stars for which in two cases the progenitor Stars were identified and some of its properties could be inferred directly from observations before explosion, e.g. SN 1987A in LMC and SN 1993J in M81, attempting to identify the progenitors of Sne Ia is a difficult task since they are most likely faint compact dwarf stars. Therefore we can but surmise their progenitors and give the potential candidates by indirect means, namely, matching some parameters indirectly derived from the explosion to the observations. Two evolutionary scenarios have been proposed, including: A single degenerate (SD) scenario, i.e., accretion of hydrogen-rich matter via mass transfer from binary companion (Nomoto 1982). The strong wind from accreting WD plays a key role, which yields important age and metallicity effects on the evolution. A double degenerate (DD) scenario, i.e., merging of double C-O WDs with a combined mass exceeding the Chandrasekhar mass limit Mch (Iben & Tutukov 1984; Webbink 1984) Two evolutionary channels for SD model WD+RG (symbiotic system) WD+MS system(super-soft system) SD progenitors are theoretically favored even though it is very difficult for hydrogen-accreting WDs to reach the Chandrasekhar limit. They consist of a low-mass WD accreting material (H or He) from the companion star until either it reaches Mch or a layer of helium has formed on top of C-O core that can ignite and possibly drive a burning front into carbon and oxygen fuels. Critical accretion rate(steady hydrogen burning): ． ． MWD M M b 0.75 10 ( 0.40) M 6 M⊙/yr If the accretion rate exceeds the critical value, it would lead to form an extended H-rich common envelope around the WD since the material accreted is larger than that consumed. If the accretion rate is low, undergo repeated nova outburst. The mass of WD will not grow at all. At a moderate accretion rate, helium flash and give rise to sub-Ch explosions. The main problem with this scenario At SN Ia explosion, ejecta would collide with the CSM, which produce shock waves propagating both outward and inward. At the shock front, particle accelerations take place to cause radio emissions. Hot plasmas in the shocked materials emit X-rays. The CSM ahead of the shock is ionized by Xrays and produce recombination Ha emissions, which is inconsistent with SN Ia spectral observations. Accretion problem? The first case for the detection of Ha emission in SN Ia (2002ic by Hamuy et al) Spectroscopic evolution of SN 2002ic. a, This sequence shows five spectra of SN 2002ic (in AB magnitudes) obtained between 2002 Nov. 29 and 2003 Feb. 1 UT with the Las Campanas Observatory Baade 6.5-m and du Pont 2.5-m telescopes, and the Steward Observatory Bok 2.3-m telescope. Arbitrary offsets have been added to the spectra for clarity. The spectra are +6, +10, +34,+47, and +70 days from estimated maximum light. We attempted to remove the 8 two most prominent telluric lines (indicated with the circled plus signs symbols), but some residuals are evident. The top spectrum shows the Si II λ6355 feature that defines the Ia class, as well as prominent Fe III absorption features at 4200 and 4900 Å. The absence of the He/Na feature at 5900 Å in the spectral evolution rules out a type Ib/c classification. b, A comparison between the 29 Nov 2002 (+6 days) observation of SN 2002ic and the spectrum of the type Ia SN 1991T obtained at an epoch of +4 days, shows that both spectra are quite similar, except that the features in SN 2002ic are all diluted in strength. Light curves of SN 2002ic Spectroscopic comparison between SN 2002ic and SN 1997cy.spectrum of SN 2002ic taken on Jan. 9 (~47 days after maximum light, which corresponds to ~67 days after explosion for an assumed time of 20 days between explosion and peak brightness) compared to that of the type IIn SN1997cy taken 71 days after explosion13, which is assumed to coincide with thedetection of GRB 97051412. The striking similarity between these two objects suggests that some SNe IIn are the result of thermonuclear explosions of white dwarfs surrounded by a dense CSM instead of core collapse in massive stars. Why was hydrogen not detected before? One of the key questions posed by the recent observations of Hamuy et al. (2003): why hydrogen has been detected only in case of SN 2002ic since there exist about 100 spectra of Sne Ia? The main problem of this scenario is that one would expect to observe a range of Ha lines in Sne Ia, depending on the amount of circunstellar material (in turn, determined primarily by the mass of the AGB star), rather than detecting a relatively strong line in only one case (it is also hard to believe that this is the first progenitor system containing an AGB star). Instead, it might be proposed that the total absence of Ha lines in all pre-SN2002ic Sne Ia observed to date argues that SN 2002ic represents rather rare cirmustances, not a WD accreting from the wind of an AGB star. Merging of two WDs There is no question that binary WD systems are an expected outcome of binary star evolution. Double WD systems having short enough periods are expected to merge as a result of angular momentum losses via gravitational wave radiation (GWR) in a time: t GW R ( yr ) 1.5 10 8 A 4ff M 1R M 2 R ( M 1R 8 10 7 P 8 / 3 ( M 1R M 2 R )1 / 3 M 2R ) M 1R M 2 R Evolutionary scenarios for two merging WDs Problems with DD model Probabilities of realization Some progress has been made in the search of DD binary systems. For example, Saffer etal. (1998) found 18 in 153 field WDs and subdwarf B stars. Based on N-body simulations, Shara & Hurley (2002) find a remarkably enhanced production rate(~15 times) in star clusters of very short period, massive DD systems due to dynamical interactions. If the enhancement mechanism works, the frequency number would not be a problem. Accretion-induced collapse ? Compressional heating effect will trigger C-ignition off center of C-O core (makes c-o core become mixture of ONe-Mg). Electrons has been captured by 24Mg, decreasing the electron pressure . The End Thanks for listening!