Type Ia Supernovae: Fusion Explosion Explained

[qraal]

Hi All

Is it type Ia supernovae that detonate in a fusion explosion?

 

[marcus]

Hi All

Is it type Ia supernovae that detonate in a fusion explosion?

Yes, it is type Ia.

 

[qraal]

Yes, it is type Ia.

According to Wikipedia the Type I Supernova are the fusion explosion types, while Type II are core collapse. I guess Pair-Instability "Hypernova" are a different group. The sub-divisions are due to different spectral lines.

Fusion explosion types are modeled as either a mass-exchange scenario in a binary, where a white dwarf siphons off enough mass to collapse as it approaches the Chandrasekhar limit. That's the old scenario which I had in mind, but according to the Wikipedia the new scenario is two white dwarfs merging and pushing over the Chandrasekhar limit - which is new to me. I had heard of neutron star mergers perhaps powering GRBs, but I didn't realize there were very many close white dwarf binaries sufficient to produce the observed Supernova rate.

Is there any major observational difference between the two scenarios?

 

[marcus]

I guess I am out of date. I had not heard of the two white dwarf merger scenario. My only mental picture for Type Ia is the mass-exchange scenario where the dwarf's mass builds up to near chandrasekhar limit. Someone else needs to step in here who knows about this other scheme.

You could help by giving links to the Wikipedia articles you are using, and maybe how far down the page it is, where it gives this new scenario. Then other people could check it out and respond.

 

[Arch2008]

http://www.astronomynow.com/news/n1003/17supernova/
http://www.sciencedaily.com/releases/2010/03/100315162049.htm
http://www.physorg.com/news182077672.html
http://www.scientificamerican.com/article.cfm?id=type-ia-mergers
http://arxiv.org/abs/1002.3359


White dwarf (WD) stars are key to Type Ia SN. Either WD stars accrete hydrogen from larger companion stars or merge with neutron stars or black holes in computer models that mimic observations. However, several recent events indicate a WD-WD merger as a new explanation. The spectra of these SN show carbon and oxygen (the components of WDs), but not hydrogen from a companion. Apparently, we didn’t have the math to model these mergers, so they were excluded…until now. However, this discovery should not affect the use of Type Ia SN as standard candles or there use in determining the expansion of the universe.

 

[marcus]

http://www.astronomynow.com/news/n1003/17supernova/
http://www.sciencedaily.com/releases/2010/03/100315162049.htm
http://www.physorg.com/news182077672.html
http://www.scientificamerican.com/article.cfm?id=type-ia-mergers
http://arxiv.org/abs/1002.3359


White dwarf (WD) stars are key to Type Ia SN. Either WD stars accrete hydrogen from larger companion stars or merge with neutron stars or black holes in computer models that mimic observations. However, several recent events indicate a WD-WD merger as a new explanation. The spectra of these SN show carbon and oxygen (the components of WDs), but not hydrogen from a companion. Apparently, we didn’t have the math to model these mergers, so they were excluded…until now. However, this discovery should not affect the use of Type Ia SN as standard candles or there use in determining the expansion of the universe.

Excellent. Thanks Arch.

 

[Arch2008]

My pleasure, Marcus. You can probably skip the first links and go straight to the paper by Gilfanov, which I couldn't find until I read the SciAm article.

 

[Ken G]

That's news to me too-- I thought the whole idea of a Ia being a "standard candle" is that you know you had the Chandrasekhar mass involved in the thermonuclear runaway! If you are merging two stars, it seems like you could get more than that, but perhaps the "standard candle" has more to do with correlating the luminosity with aspects of the light curve than it does with having exactly a Chandrasekhar mass. Still, there is a significant inconsistency in the Scientific American article, where it states "Such an explosion, which would have more fuel to burn than a single detonated white dwarf, might explain certain bright supernovae that appear to be powered by an object above the Chandrasekhar mass." Now hang on, are we talking about the primary mechanism that produces type Ias, or are we talking about "might" explain "certain bright" supernovae? Could they be a little more vague? (By the way, I was never bothered that the hydrogen in the companion star did not show up in the supernova, as companion stars normally survive supernovae just fine.)

Another interesting thing I thought I knew about type Ias, which might now be up in the air also, is that they tend to be from carbon white dwarfs, more so than other kinds, because the fusion of carbon is so temperature sensitive that it tends to run away before gas pressure can stabilize it (the way gas pressure stabilizes the helium flash at the end of the red giant phase). If you want the whole star to blow up, you don't want part of it to blow and then the rest expands and cools below fusion temperatures, so it helps to have fusable material that is more unstable when degenerate than is the norm, which apparently carbon is (maybe oxygen too, I really don't know).

 

[Ken G]

In thinking about the Gilfanov paper, another thing that bothers me is that although it may certainly be true that two infalling white dwarfs shouldn't have much in the way of accretion disks themselves, so little X-rays, one has to ask how they got to be white dwarfs in the first place. If they are close enough to infall, presumably they experienced a whole lot of mass transfer during their evolution. One of them would have been a white dwarf first, and when the other went through an asymptotic giant phase, would it not have transferred a great deal of mass to the white dwarf? And would that not have involved something like half as much X-ray generation as a white dwarf accreting past the Chandra limit, because we know it would have had to accrete past half that limit? So I'm not sure I see their mechanism reducing X-rays from a galaxy by a factor of 30-50, rather than maybe a factor of 2. I'm worried the X-ray accounting is botched there somehow.

 

[Arch2008]

IIRC, the WD-WD SN seems to explain only three or four observed events. The larger WD shreds the smaller and you get something like 2.4 solar masses rather than 1.4 involved. One other paper that I did not list showed a statistical breakdown of likely WD binary companions and they didn’t even consider the companion being another WD. Perhaps for the reasons you explain, this is not very likely, although apparently not impossible. I, too, got the impression from the articles that this was a completely new explanation for all Type Ia, although Gilfanov only gives this process a max of 5% of all such events.

 

[ibysaiyan]

There are two ways of the formation of type one a supernova.I recently discovered.. one is by the collision of two oncoming white dwarfs reaching beyond the Chandrashekar limit for the collapse of electron degenerate pressure.

It's a little confusing for there exists two types of white dwarfes the common being Carbon-oxygen class and least common oxygen,neon and magnesium class.
For the Carbon-oxygen type in a binary star the more massive star turns into an asymptotic Branch red giant which then forms into a white dwarf as for the less mass star that turns into a red giant.The matter accretes into the white dwarf as a result of which we get a type 1a supernovae.

Other scenario which I believe is what Accretion Induced collapse states:
Less common oxygen-neon-magnesium white dwarf accretes matter from the companion star going beyond the 1.44 solar mass limit BUT this causes collapsing core to rebound which subsequently leads the formation of a proto-neutron star and a short supernovae.

 

[Chronos]

If it is simply a matter of accreting enough mass to reach the Chandrasekhar limit, it is curious why they are so rare. A white dwarf siphoning mass from a red giant companion in close orbit would seemingly reach the Chandrasekhar limit in a relatively short period of time and such systems should not be exceedigly rare. So why are SnIa so rare, or - to paraphrase Fermi - where are they?

In arXiv:0910.1288 129 white dwarfs are identified within 20 pc. In an earlier paper by Holberg, the fraction that are members of binary systems was about 25%. The number of white dwarfs in the Milky Way has been estimated to be around 12 billion [arXiv:0903.2159]. Assuming about 25% are members of binary systems, about 3 billion such stars exist. Per http://articles.adsabs.harvard.edu/full/2007MNRAS.375.1315K at least 1% are in the 1.3 solar mass range - meaning around 10 million such stars should exist. Given that no SnIa within our galaxy have been observed in at least 400 years, the maximum accretion rate for any of the 10 million candidate stars could not have exceeded about .0004 solar masses per year over the past 400 years. The SDSS survey
Parameters of DA white dwarfs in SDSS-DR1 (Hu+, 2007)" J/A+A/466/627/table2
gives 1813 white dwarf masses. 80 have masses of 1.1 solar or more. The accretion rate for any of the estimated 44 million such stars in binary systems in this galaxy cannot have averaged more than about .0009 solar masses per year for the last 400 years.

 

[ibysaiyan]

If it is simply a matter of accreting enough mass to reach the Chandrasekhar limit, it is curious why they are so rare. A white dwarf siphoning mass from a red giant companion in close orbit would seemingly reach the Chandrasekhar limit in a relatively short period of time and such systems should not be extremely rare. So why are SnIa so rare, or - to paraphrase Fermi - where are they?

In arXiv:0910.1288 129 white dwarfs are identified within 20 pc. In an earlier paper by Holberg, the fraction that are members of binary systems was about 25%. The number of white dwarfs in the Milky Way has been estimated to be around 12 billion [arXiv:0903.2159]. Infering about 25% are members of binary systems, about 3 billion such stars exist. Per http://articles.adsabs.harvard.edu/full/2007MNRAS.375.1315K at least 1% are in the 1.3 solar mass range - meaning around 10 million such stars should exist. Given that no SnIa within our galaxy have been observed in at least 400 years, the maximum accretion rate for any of the 10 million candidate stars could not have exceeded about .0004 solar masses per year over the past 400 years. That seems quite low.

Interesting point... perhaps the answer may lie in the class of WD maybe naturally oxygen-neon-magnesium WD tend to exist incredibly low?
Just a thought.

 

[Ken G]

In an earlier paper by Holberg, the fraction that are members of binary systems was about 25%. The number of white dwarfs in the Milky Way has been estimated to be around 12 billion [arXiv:0903.2159]. Infering about 25% are members of binary systems, about 3 billion such stars exist. Per http://articles.adsabs.harvard.edu/full/2007MNRAS.375.1315K at least 1% are in the 1.3 solar mass range - meaning around 10 million such stars should exist. Given that no SnIa within our galaxy have been observed in at least 400 years, the maximum accretion rate for any of the 10 million candidate stars could not have exceeded about .0004 solar masses per year over the past 400 years. That seems quite low.

Is it possible that the problem there is, the 25% binary fraction refers to binaries of all orbital separations, whereas mass transfer requires a binary separation not much more than about 1 AU? Maybe the vast majority of the binaries have separations larger than 1 AU. That would seem a little surprising though, I wasn't aware that binary interactions would be so rare.

 

[Chronos]

This remains an active area of research. An interesting paper on the subject is
The Progenitors of Type Ia Supernovae: Are They Supersoft Sources?
http://arxiv.org/abs/0912.0757
Di Stefano

A related paper also by Di Stefano:
The Progenitors of Type Ia Supernovae: II. Are they Double-Degenerate Binaries? The Symbiotic Channel
http://arxiv.org/abs/1004.1193

 

[Ken G]

Remarkable that so little is known about even the most basic attributes of the environment in the type of supernova so central to the "precision cosmology" that brought us dark energy and the accelerating expansion!

 

[Garth]

Remarkable that so little is known about even the most basic attributes of the environment in the type of supernova so central to the "precision cosmology" that brought us dark energy and the accelerating expansion!

A shrewd comment!

SNe Ia appear to be photometrically quite homogeneous, generally following standard light-curve shapes to within 0.1 mag and having an absolute magnitude scatter of no more than 0.25 mag.

However we note that the 'Standard Candle', i.e. their absolute magnitude scatter, properties have been established only for cosmologically nearby supernovae.

As well as the question of the exact model for the supernovae, there is also the question of how their absolute magnitude may vary at cosmological distances, z > 1, the observation of which is intrinsically convoluted with the cosmological parameters of acceleration/deceleration and curvature, as well as other factors such as extinction.

In order to establish the latter parameters in the standard LCDM model their luminosity is assumed to be constant. However, at these distances a secular effect such as an evolving metallicity in the early universe may introduce a systematic error in this assumption.

Just a thought...

Garth

 

[Ken G]

The possibility of secular differences in type Ia SN has certainly always been an issue in interpreting these as standard candles. I can't say why that possibility is generally discounted, and I agree that if we don't even know if these are WD-WD or WD-RG systems, it's hard to say how we know there aren't secular variations. Perhaps someone on this forum has enough knowledge about type Ias to be able to say why they can be intepreted as standard candles over cosmological times, just by looking at their light curves without actually knowing what kind of event is happening there.

 

[Chronos]

The evidence is pretty good so far, as is the modeling, that SnIa are reliable 'standard candles'. It is less clear why they are so rare. Despite millions upon millions of promising candidates in our galaxy, not a single one has erupted in over 400 years. I find that curious - not that I'm complaining. Local SnIa tend to depress property values.

 

[Ken G]

But how can the modeling be telling us that they are good standard candles if we don't know if we should be modeling WD-WD or WD-RG systems? And how can we have empirical evidence they are standard candles if cosmologically old ones could be different from the ones we get independent checks on?

 

[Astronuc]

But how can the modeling be telling us that they are good standard candles if we don't know if we should be modeling WD-WD or WD-RG systems? And how can we have empirical evidence they are standard candles if cosmologically old ones could be different from the ones we get independent checks on?

Perhaps this will partially answer the questions - "A massive supernova variety - Type Ia - brightens and dims so predictably that astronomers use them to measure the universe's expansion -called a "standard candle."" But this doesn't address the cause, but only the effect.

http://www.dailygalaxy.com/my_weblog/2011/02/type-ia-supernova-one-of-the-great-unsolved-mysteries-in-astronomy-.html

It has been ~400 years since the last SN was observed in this galaxy. Perhaps one happened on the other side of the Milky Way nucleus, so we didn't see it. And perhaps one will happen in the MW next year or next decade or next century or next millenia.

http://online.itp.ucsb.edu/online/snovae07/distefano/
http://online.itp.ucsb.edu/online/snovae_c07/distefano/

Apparently the two papers cited by Chronos are the two main papers by Prof. DiStefano.

There's this summary of observation -
http://en.wikipedia.org/wiki/History_of_supernova_observation

A lecture on Type Ia
http://www.pha.jhu.edu/~bfalck/SeminarPres.html#sneIa

DiStefano talk on Sne progenitors - http://online.itp.ucsb.edu/bblunch/distefano/

 

[Ken G]

Perhaps this will partially answer the questions - "A massive supernova variety - Type Ia - brightens and dims so predictably that astronomers use them to measure the universe's expansion -called a "standard candle."" But this doesn't address the cause, but only the effect.

http://www.dailygalaxy.com/my_weblog/2011/02/type-ia-supernova-one-of-the-great-unsolved-mysteries-in-astronomy-.html

That is a well-written article. I note it raises the same objection I did above, in terms of the problems in arguing that absence of X-rays adjudicates between WD-WD and WD-RG: "Since both scenarios - an accretion-driven explosion and a merger-driven explosion - involve accretion and fusion at some point, the lack of super-soft X-ray sources would seem to rule out both types of progenitor." So much for the WD-WD claim. Still, that article focuses on the "missing progenitor" problem, more so than the "how can we know how bright cosmologically old versions of these things would be if we don't even know what they are" issue.

The lecture focuses on WD-RG, where one can expect very nearly the Chandra mass. But if we can't rule out WD-WD as the main progenitor, and if the cosmologically important signal is a 0.5 mag extra brightness around z=1, then we have the issue of, how do we know there wasn't a higher proportion of WD-WD at z=1 than is typical now? A WD-WD scenario can go up to almost twice the Chandra mass in principle, and 0.5 mag is only about a 40% excess, so seems like it could be accommodated if most Ia's today are WD-RG, but most at z=1 were WD-WD. How do we know we can rule that out, if the community is not even in agreement if they are WD-WD or WD-RG today?

 

[Astronuc]

I picked the article for the statement on 'standard candles', and it also referenced Prof DiStefano's work. She seems to be a major investigator of Snes at the moment.

Her lecture mentions that binary systems are ubiquitous, and more than 50% of stars are in binary systems (at ~11 min in DiStefano's lecture), so in order to understand Sne evolution one must understand binary evolution. However, the trick is to understand the evolution of the star in the pair that evolves into SN. The challenge is to determine what binaries look like in terms of the mass and composition of star 1 and star 2 of a given binary. We can only go on local binary systems.

DiStefano also talks about stars with mass > 1.4 M_S, 1.2 M_S < M < 1.4 M_S, and < 1.2 M_S.


The OP is addressing type Ia Sne. However it might be worthwhile to look at the classification of Sne and try to understand why a given star or binary system evolves in the way it does.

See - The classification of supernovae, by da Silva, L. A. L.
Astrophysics and Space Science (ISSN 0004-640X), vol. 202, no. 2, p. 215-236.
Bibliographic Code: 1993Ap&SS.202..215D

I was looking for a comparable paper more recent, but haven't found it.
Edit/Update: Perhaps this one - Foundations of Supernova Cosmology
Authors: Robert P. Kirshner - http://arxiv.org/abs/0910.0257


There are some more papers of interest:

Chemically homogeneous evolution in massive binaries
http://arxiv.org/abs/1010.2177
http://adsabs.harvard.edu/abs/2010arXiv1010.2177D

Type Ia Supernovae and Accretion Induced Collapse
Authors: A. J. Ruiter, K. Belczynski, S. A. Sim, W. Hillebrandt, M. Fink, M. Kromer
http://arxiv.org/abs/1009.3661

Evolution of Binaries in Dense Stellar Systems
Authors: Natalia Ivanova
http://arxiv.org/abs/1101.2864

And this conference. The proceedings are forthcoming, but some papers are available through ArXiV (I haven't gotten around to collecting them all).
http://ciera.northwestern.edu/Ron-fest2010/sch.php
http://ciera.northwestern.edu/Ron-fest2010/conference-proceedings.php


Perhaps we need a thread on binary evolution
http://www.astro.psu.edu/users/rbc/a1/lec16n.html

and this might be of interest
http://www.cfa.harvard.edu/supernova/HighZ.html

 

[Chronos]

Your input is greatly appreciated, Astronuc. I am content with the notion SNeIa qualify as 'standard candles'. Spectral analysis shows undeniable similarities in their properties and I very much doubt the age of the universe is a relevant variable. The progenitor part is, however, debatable. These exceedingly rare events demand a better explanation. Perhaps accreting WD's normally expel or burn off excess mass, and only special case WD's permit enough to accumulate to reach critical mass. That pretty much summarizes my best wild guess.

 

[Ken G]

The problem is, the soft X-rays from the accretion phase are themselves not seen. Independently of any issues involving supernovae, there is apparently a significant absence of accretion in the first place. What's more, science is not about what we doubt, or what we are content with, it is about what we can establish as true against skeptical challenge. I don't actually know what evidence there is that cosmological evolution of type Ia's is not possible, I only know that no evidence has been presented in this thread. If anyone has any idea what that evidence is, it would make a nice contribution.

 

[Astronuc]

Interestingly, one group posits a third possibility for Sne Ia progenitors.

The recurrent nova T Pyx: a progenitor of a type Ia supernova ?
http://sait.oat.ts.astro.it/MSAIS/3/PDF/129.pdf


I'm assuming that folks are waiting for a nearby type Ia, i.e. one in this galaxy for which we know the precursors. One will have to follow the research on binary evolution to see what evidence has been collected on WD-WD pairs, WD-RD pairs, and other combinations.

 

[Garth]

I don't actually know what evidence there is that cosmological evolution of type Ia's is not possible,

Such evidence cannot exist, you cannot prove their cosmological evolution is not possible!

What we may be able to show, if we have an independent measure of distance at high z (>1) (independent of luminosity and z) is that the data is consistent with the SN Ia's absolute magnitude M remaining constant at those distances. Such an independent measure might be a standard ruler for example.

If the standard candle and the standard ruler (or whatever) measures of the SNIa's distance are both consistent with the LCDM model then that would be a strong verification that their absolute magnitudes are constant and the model is correct.

Garth

 

[Astronuc]

This highlights the difficulty associated with observing SN in fine detail, i.e., being able to observe the progenitor and it's mate:

http://seds.org/messier/more/mw_sn.html

Only two supernovae have been discovered in other galaxies of the Local Group: SN 1885 or S Andromedae in the Andromeda Galaxy M31, and SN 1987A in the Large Magellanic Cloud.

see also - http://stupendous.rit.edu/richmond/answers/historical.html

Otherwise they are in galaxies, far, far away.
http://www.supernovae.net/

 

[Ken G]

If the standard candle and the standard ruler (or whatever) measures of the SNIa's distance are both consistent with the LCDM model then that would be a strong verification that their absolute magnitudes are constant and the model is correct.

Absolutely. The problem now is that the SNIa's are the standard ruler, so can't be checked by a standard ruler. There is some X-ray data that was consistent with the SNIa interpretations, but it's not clear how standalone that X-ray data really is-- could be a kind of bandwagon effect. It is certainly true that constant-magnitude SNIa's form part of an overall observationally consistent picture, but it involves dark energy. One can't help wonder if it happened that some new interpretation of the SNIa data implied a different expansion history, then it might need to be made consistent with the other observations by invoking some similarly unknown new physics. But dark energy is certainly the only game in town that succeeds in marrying all the current interpretations.

 

[Ken G]

This highlights the difficulty associated with observing SN in fine detail, i.e., being able to observe the progenitor and it's mate:

http://seds.org/messier/more/mw_sn.html

And note that one of them, 1987A, occurred in a progenitor that at the time was not even known to be susceptible to going supernova! It seems that every time we get a nearby supernova, we find out something new about supernovae progenitors, and that might be true for the next nearby Ia also. One can't help wondering how supernova physics might change in the next few decades, and what the cosmological consequences might be.

 

[Astronuc]

And note that one of them, 1987A, occurred in a progenitor that at the time was not even known to be susceptible to going supernova! It seems that every time we get a nearby supernova, we find out something new about supernovae progenitors, and that might be true for the next nearby Ia also. One can't help wondering how supernova physics might change in the next few decades, and what the cosmological consequences might be.

Speaking of which,

SN 1987A was classified as a Type II supernova (SN II) in view of the strong hydrogen lines in its optical spectrum, but because it was the explosion of a blue supergiant (BSG) rather than a red one (RSG), it was an atypical SN II: its light curve did not reach maximum until three months after core collapse and at maximum it was only about 10 percent as luminous as most SNe II.

http://arxiv.org/abs/astro-ph/0611033

See also - http://sn1987a-20th.physics.uci.edu/1330-Podsiadlowski.pdf

Let's be careful to distinguish type Ia from II, and let's keep in mind there are Ib and Ic.

Still we need a nearby type Ia to with some pre-characterization of it's progenitor to provide direct observational evidence of Ia development.

 

[Chronos]

We should limit discussion to Ia supernova. Type II supernova are core collapse events occurring in massive stars - not the detonation event believed responsible for Ia supernova. Anyways, I wonder why WD's are the usual suspects in Ia events. Has any consideration been given to neutron stars as possible Ia progenitors? The mass range of neutron stars vary widely - from less than a solar mass [4U1656+35] to 2 solar masses [PSR J1614-2230]. Do they have 'critical' mass limits? What happens when whatever mass limits they may have is exceeded?

 

[Ken G]

Anyways, I wonder why WD's are the usual suspects in Ia events. Has any consideration been given to neutron stars as possible Ia progenitors? The mass range of neutron stars vary widely - from less than a solar mass [4U1656+35] to 2 solar masses [PSR J1614-2230]. Do they have 'critical' mass limits? What happens when whatever mass limits they may have is exceeded?

Neutron stars do have mass limits, usually between 2 and 3 solar masses, and would collapse if they exceeded that limit. But they wouldn't make Ia SNs, because they would look like a bare core collapse (no envelope to blow off), and would not have time to do what would have to be done to those neutrons to get them to fuse. The collapse time would be less than a second once the mass limit was exceeded, so it would just fall right into a black hole, I imagine without much ado, beyond a burst of X-rays from the accretion disk.

 

[ibysaiyan]

We should limit discussion to Ia supernova. Type II supernova are core collapse events occurring in massive stars - not the detonation event believed responsible for Ia supernova. Anyways, I wonder why WD's are the usual suspects in Ia events. Has any consideration been given to neutron stars as possible Ia progenitors? The mass range of neutron stars vary widely - from less than a solar mass [4U1656+35] to 2 solar masses [PSR J1614-2230]. Do they have 'critical' mass limits? What happens when whatever mass limits they may have is exceeded?

Yes,they do have a mass limit which I believe is called : Tolman–Oppenheimer–Volkoff limit.
Ranging about 1.5-3 solar masses .Above this limit neutron degeneracy pressure is unable to withstand the inward gravitational force as a result it collapses.