← quantum · contents

§ Phase xv · Stellar Quanta · Chapter 04

Core-collapse supernovae

When a star ten times the mass of the sun runs out of fuel, its iron core collapses in roughly a second, electrons get crushed into the protons of the nuclei to make neutrons, and ten to the fifty-third ergs of binding energy are released, almost all of it as a burst of neutrinos. The optical supernova that astronomers see is what is left over after the neutrinos have carried away ninety-nine percent of the energy.

A massive star is a layered furnace. From the outside in, the burning shells go hydrogen, helium, carbon, neon, oxygen, silicon, and at the centre, iron. Each shell is hotter and denser than the one above it. Each fuses the ash of the layer below into the fuel of the layer above. And each layer burns through its supply more quickly than the one outside, because the binding-energy gain per nucleon is smaller as you climb the curve. The hydrogen shell of a twenty-solar-mass star lasts ten million years. The silicon shell at the bottom lasts about a day.

When the silicon shell finishes its work, the core is iron. And iron is special, because it sits exactly at the peak of the binding-energy-per-nucleon curve. Below iron, fusing nuclei together releases energy. Above iron, splitting them apart releases energy. At iron, there is nowhere to go. No fusion reaction starting from iron pays. The fire in the centre of the star goes out.

What stops the core from collapsing immediately is the same thing that holds up a white dwarf: electron degeneracy pressure. As long as the core mass stays below the Chandrasekhar limit, the electrons, packed shoulder to shoulder in their Fermi sea, push back hard enough to balance gravity. But the silicon shell above keeps dumping fresh iron ash onto the surface of the core. The mass creeps up. The temperature rises. And then, very quickly, two processes break the standoff.

The first process is photodisintegration. At about ten billion kelvin, gamma-ray photons in the core become energetic enough to break iron nuclei apart, undoing in seconds the binding work that previous stellar generations took ages to accumulate. That sounds bad, and it is. Photodisintegration is endothermic. It pulls energy out of the radiation field and stores it back as nuclear binding. The pressure drops.

The second is electron capture. As the density climbs toward the nuclear scale, the electron Fermi energy rises above the threshold for inverse beta decay. A free electron and a proton (bound inside a nucleus) merge to form a neutron and an electron neutrino. The electron is gone. The free proton is gone. A neutron remains in its place, and the neutrino streams away at almost the speed of light. With fewer free electrons, the degeneracy pressure that was holding up the core collapses too.

Both effects reinforce each other, and both are runaway. Within milliseconds the iron core, which has been a stable lump of degenerate matter for some weeks, begins to fall freely under its own gravity.

The fall lasts about a second. Imagine a region a thousand kilometres across, weighing about one and a half times the sun, collapsing at a substantial fraction of the speed of light onto a ball the size of a small city. By the time the core has shrunk to about thirty kilometres across, the nuclei are touching one another. Below that radius, the strong nuclear force, which has spent the entire prior history of the star being attractive on these scales, suddenly becomes strongly repulsive. The collapsing matter bounces.

What happens next is one of the unsolved problems of astrophysics, but the rough outline is clear. The bounce sends a shock wave outward through the infalling outer core. That shock stalls within milliseconds, choked by the energy lost to dissociating iron nuclei in its path. For the explosion to succeed, something has to revive the shock. The leading candidate is the neutrinos.

In 1985, Hans Bethe and James Wilson published a detailed numerical model of the delayed neutrino-driven explosion. Forty years later, with three-dimensional simulations on the world’s largest supercomputers, the basic picture remains intact, though the details (convection, turbulence, magnetorotational instability, neutrino flavour oscillations inside the proto-neutron star) are still under active study. What matters for the rest of this chapter is the final ledger. Of the roughly 3×10⁵³ erg of gravitational binding energy released in the collapse, about 99 percent is carried off by neutrinos, about 1 percent goes into the kinetic energy of the ejecta, and a hundredth of that, about 0.01 percent, is what we see as visible light.

A Type II supernova or SNII (plural: supernovae) results from the violent explosion of a massive star following the rapid collapse of its core. A star's mass must be at least 8 times, but no more than 40 to 50 times, the mass of the Sun () to undergo this type of explosion. Type II supernovae are distinguished from other types of supernovae by the presence of hydrogen in their spectra. They are usually observed in the spiral arms of galaxies and in H II regions, but not in…

That energy budget is why core-collapse supernovae are, briefly, the most neutrino-luminous objects in the universe. For about ten seconds, a single dying star outshines the rest of the observable universe in neutrinos. It is also why we know the basic picture is correct. On February 23, 1987, we watched it happen.

A blue supergiant called Sanduleak −69° 202 in the Large Magellanic Cloud, a satellite galaxy of the Milky Way 168,000 light-years away, finished its silicon-burning phase and collapsed. The first thing to leave the dying star, naturally, was the neutrino burst. A small fraction of those neutrinos, about ten in every ten thousand quadrillion, eventually reached Earth on February 23, 1987 at 7:35 UT. Two underground water-Cherenkov detectors saw them. Kamiokande in Japan recorded eleven events in a thirteen-second window. The Irvine-Michigan-Brookhaven detector in Ohio recorded eight in a six-second window. The Baksan scintillator in Russia recorded five. Three hours later, light from the explosion reached telescopes in the southern hemisphere, where amateur astronomer Ian Shelton spotted it from Las Campanas Observatory in Chile. SN 1987A.

The neutrino burst arrived three hours before the light. That is not because neutrinos travel faster than photons. They do not. It is because the outgoing photons had to push their way through tens of thousands of kilometres of opaque infalling stellar matter, scattering and rescattering until the shock broke through the surface. The neutrinos, which interact only weakly, walked straight out. The delay is a measure of the photosphere’s escape time, not of any superluminal effect, and it precisely matches the canonical model.

The neutrino observation also placed strict bounds on the neutrino mass. If neutrinos had a substantial rest mass, the higher-energy ones would have arrived earlier than the lower-energy ones, spreading the burst out in time. The burst arrived essentially simultaneously across the energy range observed. From this, theorists could place an upper bound of about 30 eV on the electron neutrino mass, much tighter than any laboratory result available in 1987. (The current best limits, from KATRIN and from cosmology, are below an eV, but for thirteen years SN 1987A held the record.)

What is left behind after a core-collapse supernova depends on the mass of the progenitor. For stars in the rough range of 8 to 25 solar masses, the proto-neutron star cools and contracts into a stable neutron star. The next chapter is about it. For stars more massive than about 25 solar masses, the proto-neutron star is too heavy for neutron degeneracy to support, and it collapses further into a black hole. Sometimes the collapse fails to produce a visible explosion at all (a “failed supernova”), and the star vanishes from view while leaving a black hole behind. Telescope surveys have identified a handful of candidate failed supernovae over the past decade, including the disappearance of N6946-BH1, a 25-solar-mass red supergiant in NGC 6946 that flashed faintly in 2009 and then was simply gone.

The other thing left behind is the heavy elements. The explosion blasts the outer shells of the star outward, but it also forges new nuclei in the process. The shock heats material along its path to temperatures and densities at which nuclear reactions proceed on millisecond timescales. Free neutrons, copiously released by electron capture and neutrino spallation, get captured by seed nuclei faster than the resulting beta decays can occur. This is the r-process (for “rapid neutron capture”), and it is one of the two main pathways by which the periodic table is filled in beyond iron.

For decades, supernovae were considered the dominant site of the r-process. The discovery of GW170817 in 2017 (Chapter 06 of this phase) showed that neutron-star mergers are at least equally important and may dominate for the heaviest elements. The current consensus is that core-collapse supernovae and compact-object mergers together account for the gold in your jewellery and the uranium in the Earth’s crust, with the relative contributions still under study. The 1957 paper of Burbidge, Burbidge, Fowler, and Hoyle that founded the field of stellar nucleosynthesis (universally called B²FH) anticipated almost everything that is now known.

In nuclear astrophysics, the rapid neutron-capture process, also known as the '''r-process', is a set of nuclear reactions that is responsible for the creation of approximately half of the atomic nuclei heavier than iron, the "heavy elements", with the other half produced largely by the s-process. The r-process synthesizes the more neutron-rich of the stable isotopes of even elements, and those separated from the beta-stable isotopes by those…

↘ Why the neutrino flux escapes when nothing else can

The proto-neutron star, in the first second after core bounce, has a density of about 10¹⁴ g/cm³ in its inner regions, exceeding the density of nuclear matter. Photons cannot stream freely through this medium; they are scattered, absorbed, and re-emitted on length scales much shorter than the radius of the star, and the radiation field is in thermal equilibrium with the matter. Neutrinos, however, interact via the weak force with a cross-section roughly σ ≈ 10⁻⁴⁴ (E_ν / MeV)² cm². For typical 10-MeV neutrinos, this gives a mean free path of about ten metres inside the proto-neutron star and several kilometres in its envelope. The “neutrinosphere” is the surface from which a neutrino can finally stream away without further interaction. Different neutrino species (νₑ, ν̄ₑ, ν_μ + ν_τ + their antiparticles) have slightly different neutrinospheres because their cross-sections differ. Roughly equal energy is released in all six species, and the cooling timescale is set by neutrino diffusion through the bulk of the proto-neutron star, lasting about ten seconds. Photons, by contrast, take days to weeks to climb out of the expanding ejecta. The two timescales differ by orders of magnitude because they are governed by entirely different cross-sections.

The neutron remnant that survives the explosion arrives, briefly, glowing at a temperature near 10¹¹ kelvin and rotating with whatever angular momentum the iron core started with. Within seconds it begins to cool by surface neutrino emission. Within hours it has cooled by orders of magnitude. Within a few hundred thousand years it has reached the surface temperature of a typical observed pulsar. The crust solidifies. The interior superfluids and superconducts. A pulsed beam of radio emission sweeps the galaxy at the rotation period. A new neutron star has begun its life.

A few of the brightest galactic supernova remnants we observe today, including the Crab Nebula (SN 1054), the Vela remnant (about 11,000 years old), and Cassiopeia A (about 340 years old), still host their pulsars. Many more remnants exist without an obvious central source, presumably because the beam misses Earth, or because the remnant collapsed to a black hole instead, or because the progenitor was kicked at hundreds of kilometres per second during the asymmetric explosion and has wandered far from the visible remnant. SN 1987A’s remnant was finally identified in 2024, after decades of searching, hiding inside a dense cloud of recently formed dust.

The chapter that follows turns from supernovae to the most extreme outcome of stellar collapse: a black hole. We will see that the same quantum vacuum that participates in pair production at every QED vertex of Chapter 13 also participates at the event horizon of a black hole, with the result that a black hole, against every Newtonian intuition, slowly glows.

§

A core-collapse supernova converts a star into a flash of neutrinos that briefly outshines the rest of the cosmos, plus a layer of newly forged heavy nuclei drifting outward at thousands of kilometres per second, plus, in the centre, a quantum sea of degenerate neutrons one teaspoon of which weighs as much as a continent. When the star is heavy enough that even the neutrons cannot hold, the collapse continues past the point where general relativity has anything left to say about ordinary matter. What it leaves behind, against every expectation, is not silent.

next chapter → Hawking radiation