The Other Supernova

The Universe is an astonishing place, so it is no wonder that there are two ways to make a supernova.  The two types were identified early in the 20th century, when astronomers noted a very peculiar thing:  supernova explosions either exhibit a lot of hydrogen in their spectra, or none at all.  That second observation is quite peculiar.  Lots of hydrogen is what you'd expect in a stellar explosion, because lots of hydrogen is what stars have.  The supernovas we've just discussed certainly expel great quantities of hydrogen.  Astronomers at the time weren't quite sure what to make of it, so they decided to create some labels:  they imaginatively named the explosions without hydrogen Type I, and the (more normal) ones with hydrogen Type II.

We have just seen how Type II supernova come about.  Now let's discuss Type I.

Our Sun happens to be a single star.  Our discussion of stellar evolution has taken place under the tacit assumption that all stars are similarly isolated.  This is true for many stars – but.  When stars are born, one can only get a single star if the gas cloud collapses symmetrically.  If the cloud collapses in a lumpy fashion, or with an elongated shape, then you usually end up with a multiple star system.  The most common form of multiple system is the binary, or double star system, where the stars rotate about each other rather like the weights at either end of a dumbbell.  However, multiple systems involving, three, four, even six stars, are not uncommon.  In the Sun's neighborhood, of the 148 stars known to be within 22.7 light-years (not counting brown dwarfs), 73 are single stars, 23 are double stars, 8 are triple stars, and one is a quintuple star, giving a total of 75 stars in multiple systems.  In other words, stars orbiting other stars isn't rare.  In point of fact, there are more of those than there are single stars!

Navigation Menu
Introduction
Matter Under Pressure
The Birth Of The Sun
The Sun's Evolution
The End Of The Sun
How Large Stars Evolve
Type II – The Other Supernova
After The Supernova

What does this have to do with stellar evolution?  For the majority of multiple systems, nothing.  The distances between double stars can vary wildly, from stars so far apart that the distance has to be measured in light-weeks (they can take millions of years to complete one orbit) to stars so close together that their atmospheres actually overlap!  (The latter are called contact binaries, for obvious reasons.)  But typically, double stars are spaced apart something like the distance between the Sun and the outer planets.  This corresponds to a distance of some 20 AU.  (One AU is the distance from the Earth to the Sun, or 93 million miles.)  At such distances each star can go through its normal evolution as though the other isn't there, and so they act as if they are single.  All of the multiple star systems within 40 light-years or so of Earth fall into this category.

But now and then, one finds double stars that are separated by only a few tenths of an AU.  As discussed on the previous page, when stars hit their red giant phase they turn into bloated behemoths roughly the size of the Earth's orbit, which is one AU.  This means that you can get stars which are larger than the binary system they're in, and the result is — complicated.  Entire libraries of very thick books have been written on the subject, complete with narrow margins and small type, and we still don't understand close binaries all that well.  In lieu of a mass of overwhelming detail, let me outline a couple of "representative" scenarios which capture the flavor of what can happen.

A Contact Binary

Suppose we create a double-star system consisting of our Sun and another, larger star that I will call "Jumbo".  We will give Jumbo a mass of 3.2 solar.  A main-sequence star of this mass is going to have a white-hot surface temperature some 4000 K° hotter than the Sun, a radius about 3.3 times larger, and it will be about one hundred times more luminous.  We will place the two stars into orbit around each other at a distance of 0.1 AU, or 9.3 million miles, which is about a quarter of the distance between the Sun and the planet Mercury.  This is by no means as close as stars can be, but it is close enough for our purposes.  Naturally, since they were formed from the same nebula, both stars are exactly the same age and have the same chemical composition.

In the beginning, the two stars behave as if they are single.  If made into a scale model where the Sun is represented by a four-inch ball, Jumbo would be a 13-inch ball setting three and a half feet away.  This is plenty far enough apart that their structures can be determined entirely from the dictates of ordinary hydrostatic equilibrium.  Except for their motion in space, the two stars have no significant effect on each other.  They swiftly orbit with a period of just five and a half days.

At the end of three hundred million years, however, the happy coexistence ends.  Jumbo's much greater luminosity means that its central core is already burnt-out, even though the Sun's life has barely begun.  Jumbo starts to leave the main sequence and ascend the red giant ladder.  But unlike a single star, Jumbo cannot just create a beautiful planetary nebula and retire into obscurity.  As its bloated atmosphere expands outward it must reach the orbit of the Sun, and then the Sun begins to pull it in.

The next few hundred million years are wildly complicated.  At first the Sun devours all of the gas reaching it, but eventually the rate at which Jumbo's atmosphere is expanding overwhelms the Sun, and both stars become enveloped in a single, lozenge-shaped cloud.  In this configuration some of Jumbo's atmosphere gets tugged off, whirled about, and eventually ends up being blown away into deep space.  The possible complications that can occur in such a binary system include oscillations or undulations in the gas cloud, uneven heating of the cloud due to dust, disruptions due to solar storms on the stars, and on and on.  The chaos and confusion continue until Jumbo undergoes the helium flash, at which point its atmosphere adruptly collapses and then the mass transfer stops.  Until Jumbo expands again in its next red giant phase, when all this is repeated.  More or less.

We don't understand this process especially well.  However, observations indicate that in such a system, the Sun will probably end up absorbing about two thirds of the hydrogen/helium atmosphere that reaches it.  The shift in momentum caused by this transfer first pushes the two stars closer together, but when the Sun reaches the same mass as Jumbo, it reverses direction and acts to push them apart.  Meanwhile, the gas being blown out of the system is reducing the gravitational attraction between the two stars, thus also tending to make them spiral farther apart.  Powerfully countering this, the magnetic fields of the two stars interact with the hot gas cloud like paddles whirling through water, which acts like a brake to slow down their rotation and thus induce them to spiral closer together.  Murky though the details may be, there is no question that in many cases the stars end up much closer together at the end of the mass transfer than they were at the beginning.  Since this is exactly the scenario that interests us, this is the one we will examine.

When the dust has settled (or perhaps I should say, when the gas has cleared), Jumbo has been shorn away to a lowly white dwarf of about 0.7 solar masses.  The Sun has been pumped up to 1.5 times its original mass, and thus it has shifted along the main sequence into a completely different stellar class.  Its surface temperature is now 1000 °K hotter than it was before, and it now blazes away at over four times its former luminosity.  A scale model of this new system would use a five-inch ball to represent the Sun, a speck smaller than a period on this page to represent the now rather poorly named "Jumbo", and they would be perhaps a foot apart.  The Sun and Jumbo now whirl around each other in only two days.

For the next two billion years or so they again settle into domestic tranquility and simply orbit.  Since their orbital period is so short, they have time to complete some 360 billion cycles (as compared to the paltry 4.4 billion cycles that the Earth has so far completed around the Sun), so it is quite parochial to say that such double-star systems are unstable or short-lived.  Nonetheless, a clock is ticking. 

Two billion years is about what a 1.5 solar-mass star has available to it before it exhausts its core hydrogen and begins to evolve into a red giant.  This is exactly what the the "Sun" in our binary system does.  But turn-about is fair play, and as the Sun expands and its outer atmosphere reaches the white dwarf, the intense gravity of the dwarf begins pulling in the gas like a vacuum cleaner.  To the delight of astronomers everywhere, there are so many pathways that a double system can take at this point that we could probably keep everybody in the astro-business employed for the next twenty years just by working out the details.  Very subtle factors, including their exact separation, the precise masses of the two stars, the eccentricity of their orbit, their rotation rates, and even the strengths of their magnetic fields, can lead to dramatic changes in how the stars interact.

A double star as seen from an imaginary planet.

Let me outline a couple of the more extreme (and therefore more easily understood) possibilities.  If the white dwarf finds itself just at the outer edge of the red giant's atmosphere, then it can develop a swirling accretion disk where the hydrogen slowly spirals down and gently "soft lands" on the surface of the white dwarf – very reminiscent of water swirling down a drain, although the physics is much different.  (The illustration at right is an artist's conception of the process.)  The immense surface gravity of the dwarf compresses the hydrogen into an ultra-dense "ocean" only a few meters deep but weighing in at 50 tons per quart.  The ocean smoothly covers the entire star.

The flow into this ocean can continue for anywhere from a few thousand years to a few hundred thousand, depending on how fast the white dwarf is skimming gas off the atmosphere of the red giant.  But there is a problem:  hydrogen is not stable when compressed to white-dwarf densities.  I devoted a full paragraph on a previous page to an explanation of how difficult it is to achieve hydrogen fusion, but that's under normal circumstances.  The surface of a white dwarf is not an especially normal place.  In fact, hydrogen fusion is quite easy for a white dwarf to achieve.  When the dwarf's electron-degenerate hydrogen "ocean" reaches a depth of about 200 meters, the pressure at the bottom becomes so high that inevitably, somewhere on the planet-sized star, hydrogen fusion will spontaneously begin.

Just like the core of a star undergoing a helium flash, the electron-degenerate hydrogen on the white dwarf cannot expand and cool.  So, the trapped heat from the fusing hydrogen only acts to make the reaction run even faster, and before you can say "hydrogen flash", a nuclear firestorm engulfs the entire star.  Like a planet with a gasoline ocean, the white dwarf instantly goes up in (nuclear fusion) flames.  For weeks the nuclear inferno burns, with the dwarf spectacularly leaping to a luminosity 100,000 times that of the Sun.  Such events are called novas, from the Latin for "new", because as seen from Earth they appear to be new stars that have suddenly appeared.  (It is from novas that we get the name for their even larger cousins, the supernovas.)  Novas are surprisingly common, because unlike supernovas the delivery truck that brings them doesn't blow itself up at the end of the run.  Once the firestorm has run its course, the white dwarf is virtually unaffected and the only sign left of the nova is a small, expanding shell of hot gas.  The gas is the "ashes" of the hydrogen ocean, finally heated to the point where it could escape.  But the red giant is still out there, and its atmosphere is still swirling down on the dwarf, so the whole process begins again.  Depending on the exact parameters of the system, red giant / white dwarf doubles can explode a thousand times like this over several million years.

At the opposite extreme, if the gas flow between the stars is very high, the hydrogen behaves more like the fuel for a welding torch than the water in a quiet ocean.  The key factor is this:  the reason that the gas flowing towards the white dwarf forms an accretion disk in the first place, as illustrated above, is because it has a different velocity than the dwarf.  The gas is essentially trying to go into orbit around the dwarf.  One of the major headaches that theoretical astrophysicists have with accretion disks is that things moving in a circle have a lot of momentum, and you cannot bring anything down from orbit unless you reduce its momentum.  (Isaac Newton is very insistent on this:  momentum isn't allowed to just disappear.)  Unlike the Space Shuttle, accretion disks do not come equipped with retro-rockets, so other mechanisms have to be invoked to dissipate the rotational momentum and bring the hydrogen down.  The usual one is friction.  The idea is that the gas in the disk will be moving at different speeds depending on how far it is from the dwarf, and friction between the flows can slow the gas so that it can descend.

But such mechanisms take time to work.  If the incoming gas flow is too high then you very quickly achieve the galactic version of a stopped-up drain.  As gas piles into the accretion disk faster than it can escape, the disk becomes ever thicker, more massive, and hotter.  Much hotter.  The ferocious gravity of the white dwarf creates a violently turbulent disk with gas flows faster than 1000 miles/second.  Truly galaxy-class friction causes the disk to glow at a white-hot 15,000 K° and sizzle with even hotter hot spots that can reach 70,000 K°.  Copious bursts of X-rays and hard ultraviolet radiation pour from the roiling gasses, providing nearly unlimited raw material for PhD theses in astrophysics.

Meanwhile, down the white dwarf, a super-heated, supersonic drizzle is falling from the unimaginable thunderstorm above it, the white-hot droplets streaking down at five thousand times the velocity of a rifle bullet under the dwarf's tremendous gravity.  The hot hydrogen ignites virtually on contact, creating a ring of nuclear fire all around the equator of the white dwarf.  White dwarf / accretion-disk systems such as this can "pulse" on and off, or they may sputter and cough erratically, or they may even come to equilibrium and shine fairly steadily (anything is possible).  The average luminosity of such dwarfs tends to be quite high, around 100 times solar, so they are sometimes brighter than the red giant stars fueling them!

But ominously, since the hydrogen is burning steadily as it arrives, it cannot collect into an electron-degenerate ocean and explode, throwing its helium "ashes" into space as a nova does.  If all the physical parameters are just right, and if the gas flow from the red giant continues, the white dwarf steadily becomes coated with an ever-heavier mantle of helium.

Returning for a moment to our model system, recall that Jumbo is now a 0.7 solar-mass white dwarf and the Sun is a 1.5 solar-mass star attempting to become a red giant.  Or to put the emphasis differently, the Sun is a star attempting to shed enough gas to join Jumbo as a white dwarf, because that is the natural end for a red giant.  As a single star, the Sun's evolution would lead it to emit a planetary nebula with a mass of approximately 1.5 - 0.6 = 0.9 solar masses.  But as a double star, the gas that would have formed the nebula is being wolfed down by Jumbo.

The efficiency of the gas transfer in white dwarf / normal star binaries is very nearly 100%.  When you add the 0.9 solar masses of gas that the Sun is trying to liberate to Jumbo's mass of 0.7 solar you have — 1.6 solar masses.  Which is far too much.

Already dangerously compressed to the size of Mars, Jumbo cannot absorb all the hydrogen swirling off the Sun.  It must eventually reach the critical limit of 1.4 solar masses predicted by Chandrasekhar in 1931.  After hundreds of thousands of years of mass accretion, the day must come when, in less time than it takes a candle to flicker, Jumbo finally and catastrophically collapses.

And then it stops!  Unlike the center of a red supergiant, Jumbo is not made up of 1.4 solar masses of iron.  Jumbo consists almost entirely of helium, carbon, and oxygen, all of which (unlike iron) are perfectly willing to liberate fusion energy.  The terrific pressure of the Chandrasekhar collapse instantly ignites the entire star as though it is the heaviest thermonuclear bomb in the galaxy – which, in point of fact, it is.  For a fraction of a second the matter hangs in the balance as gravity tries to crush Jumbo into a neutron star and a fury of nuclear fusion tries to vaporize Jumbo into incandescent gas.

And the winner is — nuclear fusion!  In a single apocalyptic blast, Jumbo is completely shattered and ceases to exist.  The entire bulk of the star is turned into a radioactive cloud so hot, it literally shines with the light of a 100 billion stars.  All the gas is hurled into space at tens of thousands of meters per second.  About half of it is now iron, because that's the fraction of the star that was able to fuse all the way down to the bottom of the "nuclear well" in the few seconds of the explosion.  (Type I supernovas are why there is so much iron on the Earth, as compared to other metals.  Type II supernovas, by contrast, crush most of their iron into neutron stars and do not share it with the rest of the galaxy.)

It is remarkable that something with less than half the diameter of the Earth can produce such an explosion.  It is also remarkable that Type I and Type II supernovas are almost equally luminous and create dazzling light shows that last for very nearly the same length of time, which is why they were so easily confused by turn-of-the-century astronomers.  The coincidence is made more remarkable when you stop to consider that Type II supernovas are in fact about 100 times more powerful than Type I supernovas!  But since approximately 99% of the energy in a Type II supernova is emitted as invisible neutrinos, which streak away from the star and race into space, never to be seen again, the detectable energy from a Type II is almost exactly the same as a Type I.  The two types are even roughly equal in frequency:  observed supernovas consist of roughly 60% Type I and 40% Type II.

Which brings us back to the question that started this, the mysterious difference in the hydrogen spectra of the two types of supernovas.  No doubt the astute reader has already realized how it is that Type I supernovas can explode and yet show no hydrogen in the spectrometer:  the collapsing white dwarfs don't have any.  The fusion energy is supplied exclusively by helium and heavier elements.1

As for the Sun, it is surprisingly unfazed by Jumbo's splashy departure.  One might think that an explosion so furious it can outshine entire galaxies, taking place literally just outside its atmosphere, would evaporate the Sun to nothing but a memory.  However, this is not so.  Stars are very massive and (already) very hot; even a supernova explosion right beside them cannot do more than than blow away a bit of their outer atmosphere.  The Sun will lose only perhaps 15% of its mass, and most of that it would have lost anyway in the final stages of its life as a red giant.  Odd as it may seem, the partner star in a Type I supernova is almost unaffected by the explosion.

Except for the fact that it no longer has a partner.  With all the complicated gas transfer phenomena at an end, the Sun rearranges its affairs and becomes a perfectly normal post-red-giant star.  It eventually retires as a staid and respectable white dwarf, not the disreputable kind which explode.
1 – Actually, the fact that Type I supernovas show NO hydrogen at all means that this scenario may not be entirely correct, because even the small amount of hydrogen blown off the partner star ought to be detectable.  One theory is that the partner star is in fact stripped of all its hydrogen before the supernova explosion, and the final coup de gras is delivered by helium streaming off the core of the partner star, not hydrogen.


Next:  After The Supernova