№ 179 · astronomy

A supernova is a core collapsing: the fall comes before the blast

The usual picture of a supernova is a star that runs out of fuel and blows up like a bomb. For the common kind, the order is the other way round. A star much heavier than the Sun builds a core of iron that cannot hold itself up, and that core falls in on itself in a fraction of a second. The explosion comes after the fall, and the fall pays for it: almost all the energy comes from gravity, not from burning.

Predict first

When the core of a massive star collapses, it releases an enormous amount of energy. Where does most of it go?

Check your answer

Neutrinos. The collapse releases about 3 × 1053 erg, and nearly all of it leaves as neutrinos. The blast that throws off the star carries about 1051 erg, less than one per cent of the total.

An iron core with nothing left to burn

Stars born with more than about eight times the Sun's mass go through one round of nuclear burning after another, ending with silicon. Silicon burning leaves a core of iron, with shells of lighter elements still burning around it, like the layers of an onion. Mueller and Sykes, in a 2026 overview, describe what mostly holds that iron core up: the pressure of its electrons, packed so tightly that they resist being squeezed any closer. Adam Burrows and David Vartanyan compare the core to a white dwarf, the dense remnant a smaller star leaves behind. As the burning shell keeps adding iron, the core approaches the most mass that electron pressure can support, near 1.5 times the Sun's mass, and becomes unstable.

What pulls the floor out

Two things then take away the support, and both get worse as the core shrinks. First, the squeezed electrons are pushed into atomic nuclei, where they turn protons into neutrons and release neutrinos. Each electron captured is one fewer pushing back. Second, the core is now hot enough, around 10 billion kelvin, for high-energy light to break heavy nuclei apart. That costs energy: H.-Thomas Janka's 2012 review recalls that breaking iron into helium nuclei takes about 1.7 MeV for every proton and neutron, an argument that goes back to Burbidge and colleagues. Losing pressure makes the core shrink, and shrinking makes both processes faster. The contraction turns into a runaway fall.

A ball with nothing pushing back falls in a time set only by its density. Janka gives a simple rule: about 0.21 seconds divided by the square root of the average density in units of 108 grams per cubic centimetre. Mueller and Sykes describe the fall as taking a few hundred milliseconds. Drag the figure from the start and watch the clock. Most of the time goes on the first stretch, while the core is still large. Once the ball's radius is down to a hundred kilometres, the rest of the fall takes about a millisecond.

Interactive Squeeze an iron core. Drag radius to shrink a ball of the chosen mass from its size when the collapse starts down to the size of a neutron star, or jump to a stage with the buttons. The left panel is drawn to true scale; the right one is the same ball zoomed in 100 times. The three bars show the ball's average density, how far through a free fall it is, and how much gravitational energy has been released. Every control works with Tab, Enter and the arrow keys.

radius1,880 km
mass that collapses1.40 Suns

Radius 1,880 km, before the fall

true scalegrey dashes: the core at the start, 1,880 kmzoomed in 100 timesblue dashes: 12 km, a neutron stardensitytimeenergy out108109101010111012101310141015neutrinos trappednuclearg/cm³startfully collapsed104910501051105210531054ergthe blast

The start: an iron core of the mass you chose, at an average density of 108 g/cm³. Neutrinos made as electrons are captured stream straight out.

Average density: 1.0 × 108 g/cm³, 100 million times the density of water.

With nothing pushing back, this ball would fall all the way in after 210 ms.

Gravitational energy released so far: none.

The bounce, and a shock that stalls

The fall stops when the centre reaches the density of an atomic nucleus. There the matter resists any further squeezing, overshoots a little and springs back. That bounce sends a shock wave outward through the iron still falling in. In the 1980s this shock was studied hard as the possible explosion itself. It cannot do the job. As it moves out, it spends its energy breaking up the iron nuclei it passes through, and it loses more as a burst of neutrinos escapes from behind it. Within tens of milliseconds it stops moving outward and hangs about 100 to 200 km from the centre, while matter keeps raining through it. Janka wrote in 2012 that every modern simulation agreed: the bounce alone does not blow up any star.

Neutrinos restart it

Under the stalled shock sits a hot new neutron star, still gaining mass as matter lands on it. It is far too dense for neutrinos to escape at once, so they diffuse out over seconds, carrying away nearly all the energy released by the fall. Shirley Weishi Li, John Beacom, Luke Roberts and Francesco Capozzi give the total as about 3 × 1053 erg: about half in the first second, the rest over about ten seconds.

A small share of those neutrinos is caught in the layer just behind the shock. Mueller and Sykes put the heating there at about 5 to 10 per cent of the electron-type neutrinos' power. That heats the gas from below, the way a stove heats a pan of water, and the gas starts to churn. The churning matters. In computer models of a perfectly round star, neutrino heating almost never relaunches the shock, except in the lightest stars. In three-dimensional models, rising hot plumes push the shock outward and keep matter in the heating layer for longer, and many of these models now explode. In Burrows and Vartanyan's two-dimensional runs, the shock got going again between 150 and 400 milliseconds after the bounce.

The bookkeeping is lopsided. A typical blast carries about 1051 erg, the number the figure uses for comparison, so less than one per cent of the energy released ends up moving in the blast. That does not make the mechanism fragile: Burrows and Vartanyan point out that the heating layer absorbs 4 to 10 per cent of the neutrino energy emitted while the explosion is being decided, and most of the rest leaves later. The final collapse and the restart together take about a second. In a red supergiant, the shock then needs about a day to reach the star's surface, and only then does the star brighten.

How we know, and what is still open

In 1987 a star collapsed in the Large Magellanic Cloud, a small neighbouring galaxy, and two large water detectors, Kamiokande-II and IMB, caught 19 neutrinos from it between them over about ten seconds. That signal showed a neutron star had formed, as the theory predicted. It is still the only supernova whose neutrinos have been caught. More recently, astronomers have seen signs of radiation from the neutron star it left behind.

Mueller and Sykes call neutrino heating, helped by the churning, the prevailing explanation for most core-collapse supernovae. The details are not settled: models from different groups do not agree exactly, and the behaviour of matter at nuclear density and of the neutrinos themselves still carries uncertainty. Not every collapse explodes. When the shock does not restart, the neutron star keeps gaining mass until it collapses into a black hole. That is a failed supernova, which may give off no more than a faint glow. Waly Karim, Adam Burrows and colleagues estimated in 2026, very roughly, that about 21% of collapses end in black holes. The test everyone is waiting for is a supernova in our own galaxy. Karolina Rozwadowska, Francesco Vissani and Enrico Cappellaro combined the evidence in 2021 to estimate about 1.6 per century. Today's detectors would catch far more neutrinos from one than the 19 seen in 1987, and possibly gravitational waves, which have never been detected from a supernova.

Not every supernova

One kind of supernova does work like a bomb. A Type Ia supernova is a white dwarf in a pair of stars that destroys itself in a runaway of nuclear burning. Astronomers sort supernovae by the lines in their light: Type Ia is the thermonuclear kind, while Types II, Ib and Ic come from collapsing cores and leave a neutron star or a black hole. This page is about the collapsing kind, the one that ends the life of a massive star.

Where this comes from

  1. Core-collapse supernovae and supernova neutrinos linked only, not reproduced
    B. Mueller, B. Sykes · arXiv:2605.25879 · 2026
    arxiv.org/abs/2605.25879
  2. Core-Collapse Supernova Explosion Theory linked only, not reproduced
    Adam Burrows and David Vartanyan · arXiv:2009.14157 · 2020
    arxiv.org/abs/2009.14157
  3. Explosion Mechanisms of Core-Collapse Supernovae linked only, not reproduced
    H.-Thomas Janka (Max Planck Institute for Astrophysics, Garching) · arXiv:1206.2503 · 2012
    arxiv.org/abs/1206.2503
  4. Old Data, New Forensics: The First Second of SN 1987A Neutrino Emission reuse permitted with attribution
    Shirley Weishi Li, John F. Beacom, Luke F. Roberts, Francesco Capozzi · arXiv:2306.08024 · 2023
    arxiv.org/abs/2306.08024
  5. The Mass Function of Neutron Stars from Core-Collapse Supernova Simulations reuse permitted with attribution
    Waly M Z Karim, Adam Burrows, David Vartanyan, Anthony Fisher, and James Burry · arXiv:2608.18198 · 2026
    arxiv.org/abs/2608.18198
  6. On the rate of core collapse supernovae in the Milky Way linked only, not reproduced
    Karolina Rozwadowska, Francesco Vissani, Enrico Cappellaro · arXiv:2009.03438 · 2020
    arxiv.org/abs/2009.03438