Introduction
Supernovae are magnificent phenomena in the night sky, and have always been a marvel to human beings. A supernova is a stellar explosion that emits a burst of radiation resulting in an extremely luminous object that may outshine its entire host galaxy before fading from view over several weeks or months. One class of supernovae, known as Type Ia Supernovae (SNe Ia), is characterized by the absence of hydrogen emission lines in spectra and the presence of a prominent silicon Si II absorption line near maximum light (1). With uniform light curves and spectral evolution, SNe Ia have been increasingly used as reliable indicators of distance in measuring important cosmological constants (2). This usage has led to a need for a more intensive study of the nature of SNe Ia.
A SN Ia explosion is often the result of thermonuclear disruption of a carbon-oxygen white dwarf that accretes mass from its companion in a binary system, and thereby reaches the Chandrasekhar limit of 1.4 M⊙ (solar mass) (3). However, there is no simple means of identifying the immediate progenitor of a SNe Ia, nor of deriving information about its properties from observations (1). One way to determine the progenitors of SNe Ia is to eliminate the unlikely candidates from a pool of possible systems if they show any significant contradiction with the physical principles or observational data. Since there is not yet a best candidate whose properties agree with all the theoretical or observational criteria, identification of progenitor systems of SNe Ia remains difficult.
Properties of SNe Ia Progenitor Systems
The spectroscopic properties of SNe Ia give some indication of the composition of their progenitors. The absence of hydrogen emission lines indicates that the star contains little (less than 0.1 M⊙) to no hydrogen before explosion; the presence of a silicon Si II absorption line near maximum light suggests that nuclear fusion from pre-explosion matter into intermediate-mass elements like silicon takes place in the explosion (1). The observed velocity (mean v = 5000 km/s and peak v > 20000 km/s) of SNe Ia explosion ejecta agrees with the calculation result of about 1 M⊙ of C and O fusing into iron-group elements or intermediate-mass elements. This fact implies that the progenitor star is composed of mostly carbon and oxygen (1).
According to observational data, most SNe Ia share very similar peak luminosities, light curves, and spectra. This strongly indicates that a unique class of progenitor systems exists. Upon closer study of these properties, Chandrasekhar-mass (1.4 M⊙) white dwarfs are suggested to be the best-fitting model (1). Since 85% of observed white dwarfs have masses of no more than 0.8 M⊙ and large-mass white dwarfs are extremely rare, the only way for a SN Ia progenitor white dwarf to reach the Chandrasekhar limit is to be in a close binary system where it can accrete mass from the companion star (3).
The results of a radio observation program lasting over two decades at the Very Large Array, a radio observatory located in New Mexico, USA, imply a very low density for any possible circumstellar material established by the progenitor before explosion. This conclusion would rule out the possibility of white dwarf mass-accretion via stellar wind from a massive binary companion. Hence, the progenitor system could only be a white dwarf that accretes mass from a low mass companion by Roche lobe flow due to gravity, as suggested in single-degenerate models, or the merger of two white dwarfs, as suggested in double-degenerate models (4).
The Origin of Diversity of SNe Ia Luminosity
The above discussions all point to the currently favored model for SNe Ia progenitors: a relatively homogeneous class of C+O white dwarfs accreting mass from their companions in binary systems. However, SNe Ia also have many observed differences, among which the most important is the diversity of luminosity. Since SNe Ia are used as standard distance indicators in cosmology, this diversity and its origin requires an answer. Below are possible explanations based on various explosion models.
I. C/O ratio of white dwarf progenitors:
The brightness of a SN Ia is determined by the mass of 56Ni synthesized during the explosion, which ranges from 0.4 – 0.8 M⊙ for most SNe Ia (5). It is postulated that as the C/O ratio increases in the progenitors, the mass of 56Ni will increase, and this consequently causes a greater luminosity (2).
II. The age of progenitor systems:
As suggested by Nomoto et al (2003), in an older binary system, the mass of the companion star of the white dwarf is smaller, and the mass which can be transferred from the companion to the white dwarf is smaller. This implies that the original total mass of carbon and oxygen of the white dwarf is larger as the white dwarf reaches Chandrasekhar mass. By calculation, the explosion of a larger portion of carbon and oxygen will lead to smaller luminosity. Therefore older progenitor systems produce dimmer SNe Ia (2).
III. Morphology of the Host Galaxy:
It is observed that the most luminous SNe Ia occur only in spiral galaxies. Both spiral and elliptical galaxies can have dimmer SNe Ia (6). This may due to the different ages of the companion stars. As suggested above, SNe Ia that occur in older progenitor systems have smaller luminosities. In elliptical galaxies, star formation has long since ceased, and most of the progenitor systems are too old to produce very luminous SNe Ia. However, in spiral galaxies, star formation continues to occur, and so these galaxies can have both old and young progenitor systems. They can host luminous SNe Ia as well as dim ones (2).
Models of Pre-Supernova Evolution
Two ways by which white dwarfs in binary systems can accrete mass toward Chandrasekhar mass and cause SNe Ia have been proposed: single-degenerate and double-degenerate. Models for both scenarios have some elements that explain the observed data, and some that do not.
Double-degenerate models
I. Mechanism:
Two C+O white dwarfs in a close binary system are brought together by the emission of gravitational radiation. When the lighter white dwarf with the larger radius fills its Roche lobe, the Roche lobe is dissipated within a few orbital periods and forms a massive and hot disk configuration around the heavier white dwarf. Then the two merge into one, reaching Chandrasekhar mass and giving rise to SN Ia explosion (7).
II. The Weaknesses of Double-Degenerate Models:
(i) When the lighter white dwarf forms a disk configuration around the primary white dwarf, the disk is rotationally supported, and so carbon ignition cannot happen immediately. The most likely result of this scenario is off-center carbon ignition if the mass-accretion rate is higher than 2.7×10-6 M⊙/year. This reaction will convert the composition of the white dwarf from C-O to O-Ne-Mg. The consequence, however, is more likely to be an accretion-induced collapse to a neutron star rather than a SN Ia explosion (7).
(ii) Galactic chemical evolution results do not agree with the double-degenerate models. In particular, Kobayashi et al. (1998) performed the chemical evolution calculations for both double-degenerate and single-degenerate models and argued that the early heavy element production of double-degenerate models, which is formulated as O/Fe as a function of Fe/H, is inconsistent with the observations (7,8).
(iii) The observed SNe Ia have a similar amount of 56Ni as a production of explosion. The merging of two white dwarfs of different mass, composition, and angular momentum with different impact parameters will lead to very different burning conditions with a different amount of 56Ni produced, which disagrees with the observations (1).
III. The Strengths of Double-Degenerate Models:
(i) The absence of hydrogen lines in SNe Ia spectra can be naturally explained by double-degenerate models since only C+O white dwarfs with little or no hydrogen are involved in this scenario (1).
(ii) Merging white dwarfs can reach Chandrasekhar mass easily, while in single-degenerate models, achieving a sufficient mass-accretion rate is a major difficulty.
(iii) Many binary systems with two white dwarfs are identified. Among the eight known systems with orbital periods of less than half a day, there is one system [KPD 0422+5421 (9)] whose mass could exceed Chandrasekhar mass. Population synthesis predicts that there could be more sufficiently massive merger candidates found in short-period white dwarf binary systems (1).
Single-degenerate models
I. Mechanism:
A C+O low mass white dwarf in a binary system accumulates hydrogen-rich or helium-rich matter from the companion star by mass overflow, reaches a critical mass near the Chandrasekhar mass and explodes due to thermonuclear disruption. Another model, known as the sub-Chandrasekhar model, suggests an alternative road of evolution: before the white dwarf reaches a critical mass limit, a layer of helium forms on top of C+O core and ignites the C+O fuel (1).
II. The Weaknesses of Single-Degenerate Models:
(i) According to single-degenerate models, since a great portion of the mass that the white dwarf accumulates is hydrogen, hydrogen lines should be seen in the spectra of SNe Ia. However, hydrogen has not yet been found in any SNe Ia. The failure to detect hydrogen in SNe Ia is a factor that may rule out single-degenerates as appropriate candidates for SNe Ia progenitor systems (8).
(ii) Theoretically, few mass-accretion rates can lead to thermonuclear explosion. For low accretion rates below 10-8 M⊙/year, repeated nova outbursts will occur before the white dwarf reaches Chandrasekhar mass, and in these eruptions, more mass will be lost than accreted between eruptions. On this track the white dwarf will never reach Chandrasekhar mass (8). For higher rates (10-8 – a few ×10-7 M⊙/yr), the white dwarf will lose mass due to helium shell flashes (1). For even higher rates of accretion above a few ×10-7 M⊙/year, a hydrogen-rich red-giant envelope will form outside the white dwarf and mass will be lost due to winds. Moreover, no observation has given evidence to the existence of the debris of this red-giant envelope in SNe Ia explosion (1).
III. The Strengths of Single Degenerate Models
(i) A class of binary systems, namely the Supersoft X-ray Sources, has been identified. In this system hydrogen-rich matter is being transferred from the companion star at so a high a rate that hydrogen burns steadily outside the C+O core of the white dwarf (10). If the accreted mass can be retained, the mass of the white dwarf can actually increase toward Chandrasekhar mass. These systems may serve as strong candidates for SNe Ia progenitors in the single-degenerate scenario (1).
(ii) There are other good candidates that exist, such as symbiotic systems or recurrent novae (8).
Summary and Conclusions
Based on current knowledge of observational evidence and physical principles, it can be confidently concluded that the progenitors of SNe Ia are a homogeneous class of compact white dwarfs composed of carbon and oxygen that accrete mass from binary companion stars.
The luminosity of a SN Ia can offer some indications about its progenitor. Generally, the progenitor white dwarf of a brighter SN Ia has higher C/O ratio and a younger age, and appears in a spiral galaxy; the progenitor white dwarf of a dimmer SN Ia has lower C/O ratio and an older age, and appears in either a spiral or an elliptical galaxy.
Two kinds of models, double-degenerate and single-degenerate, are proposed to explain pre-supernova evolution. As discussed, there are observational and theoretical arguments that support and contradict each. However, double-degenerate models have more significant conflicts with theories, and with the discovery of Supersoft X-ray Sources, single-degenerate models are favored today.
As new observation technologies in x-ray, radio, and high resolution optical spectroscopy are being developed, more information concerning the properties of SNe Ia progenitor systems will be obtained. In particular, an unambiguous choice between single-degenerate and double-degenerate models can be made if the absence or presence of hydrogen in SNe Ia is determined conclusively by observations.
References
1. W. Hillebrandt and J. Niemeyer, Annual Review of Astronomy and Astrophysics 38, 191 (2000).
2. K. Nomoto, et al., in From Twilight to Highlight: The Physics of Supernovae, W. Hillebrandt & B. Leibundgut, Eds., ESO/Springer Series “ESO Astrophysics Symposia” (Springer, Berlin, 2003).
3. M. Partharsarathy, D. Branch, D. Jeffery, & E. Baron, New Astronomy Reviews 51, 524 (2007).
4. N. Panagia, et al., in American Institute of Physics Conference Proceedings, Cefalu’, Italy, 11-24 June 2006, (American Institute of Physics, Melville, NY, 2006).
5. P. Mazzali and L. Lucy, Monthly Notice of the Royal Astronomical Society, 295, 428 (1998).
6. M. Hamuy, M. M. Phillips, R. Schommer, and N. B. Suntzeff, The Astronomical Journal, 112, 2391 (1996).
7. C. Kobayashi, T. Tsujimato, K. Nomoto, I. Hachisu, and M. Kato, Astrophysical Journal, 503, L155 (1998).
8. M. Livio, Type Ia Supernovae: Theory and Cosmology, J. Niemeyer, & J. Truran, Ed. (Cambridge Univ. Press, Cambridge, 1999).
9. C. Koen, J. Orosz, and R. Wade, Monthly Notice of the Royal Astronomical Society, 200, 695 (1998).
10. P. Kahabka, E. Van Den Heuvel, Astronomy and Astrophysics, 35, 69 (1997).