Type II Supernova

The default death of a massive star

A massive star is, for almost all of its life, in a quiet equilibrium between gravity pulling inward and radiation pressure from nuclear burning pushing outward. The star is built like an onion: hydrogen on the outside, helium below it, then carbon, neon, oxygen, silicon, and an iron-group core in the middle. Every shell is fusing its dominant species into the next one, with a shell-burning timescale that gets dramatically shorter as you move inward. Helium burning takes a million years; carbon burning takes a thousand; silicon burning takes a day. The iron core, by contrast, does not burn — iron has the highest binding energy per nucleon, and turning it into anything else costs energy. The core sits inert, supported only by electron degeneracy pressure, until it reaches the Chandrasekhar mass M_Ch ≈ 1.44 (Y_e / 0.5)² M☉ — at which point degeneracy can no longer hold against the weight piled on top of it, and the core collapses.

The collapse releases enough gravitational binding energy to power the most common spectacular supernova in the cosmos. With its hydrogen envelope intact, the explosion shows the Balmer lines that define the Type II classification, lingers near a billion solar luminosities for a hundred days because of hydrogen recombination, and leaves a neutron star at the centre as a permanent record. Type II supernovae account for ~ 60-70% of the core-collapse supernova rate, ~ 50% of all supernovae once Type Ia are included, and almost every neutron star and black hole in the Galactic disk traces back to one. This is the default death of every star with a zero-age main-sequence mass above ~ 8 M☉ that has not lost its envelope by the end.

The pre-supernova structure and why iron collapses

By the last day of a 15 M☉ red supergiant's life, the central density is ~ 10¹⁰ g/cm³, the temperature ~ 8 × 10⁹ K, and the inner ~ 1.4 M☉ is an Fe/Ni-group core supported by electron degeneracy pressure. The Chandrasekhar mass for this composition (Y_e ≈ 0.45) is about 1.3 M☉, so the core is sitting near the edge of stability. Two physical processes finish the job:

• Electron capture. p + e⁻ → n + ν_e becomes thermodynamically favourable at the very high Fermi energies present in the core. Each capture removes an electron and softens the equation of state, lowering Y_e and the effective Chandrasekhar mass.
• Photodisintegration. Above T ~ 5 × 10⁹ K, gamma photons begin to disassemble Fe nuclei back into α particles and free nucleons (γ + ⁵⁶Fe → 13α + 4n consumes ~ 124 MeV). This is endothermic — it withdraws thermal energy from the gas, dropping the pressure.

Both effects amplify each other. Once collapse begins, the free-fall time t_ff = √(3π / 32 G ρ) ~ 1 s sets the dynamical timescale, and the inner core falls homologously while the outer regions fall faster than supersonically. The collapse stops when the inner core reaches nuclear density ρ_nuc ≈ 2.4 × 10¹⁴ g/cm³ — the equation of state hardens abruptly because the strong force takes over from degeneracy, and the inner core bounces. A shock wave is launched outward at the boundary between the bounced inner core and the still-infalling outer regions.

The neutrino mechanism — 99% in invisibles

The shock launched by core bounce is not, by itself, energetic enough to explode the star. As it propagates outward through the still-infalling outer iron core, it photodisintegrates iron — costing ~ 8 MeV per nucleon — and loses energy at a rate comparable to its initial kinetic energy. Within milliseconds, the prompt shock stalls at ~ 100-200 km, with infalling matter piling onto it. Naïvely, the explosion fails and the entire star collapses to a black hole.

The current standard solution is the delayed neutrino mechanism (Wilson 1985; Bethe & Wilson 1985). The newborn proto-neutron star at the centre is hot and dense — T ~ 30 MeV, ρ ~ 10¹⁴ g/cm³ — and radiates the full ~ 3 × 10⁵³ erg of binding energy in neutrinos over ~ 10 seconds, with roughly equal flux in all six neutrino species. Most of these neutrinos free-stream out. A small fraction — about 1% — re-deposit their energy in the gas just behind the stalled shock through inverse beta decay (ν_e + n → p + e⁻ and ν̄_e + p → n + e⁺). That heating builds up over a few hundred milliseconds in a region called the gain region, and when it accumulates enough pressure (the local "advection-residence" time exceeds the heating timescale), the shock revives and pushes outward. Multi-dimensional simulations (Müller, Janka, Burrows, and others) show that turbulence and standing accretion shock instability (SASI) play essential roles in tipping the balance toward explosion.

The energy budget is striking. Total binding energy: 3 × 10⁵³ erg. Neutrinos: 2.97 × 10⁵³ erg (99%). Kinetic energy of ejecta: 10⁵¹ erg (0.3%). Electromagnetic radiation: 10⁴⁹ erg (0.003%). Even with the most luminous supernova in the Galaxy ever, you will see only 0.003% of the energy directly; for SN 1987A, we got an actual neutrino measurement (24 events) of the dominant channel.

The recombination plateau — why II-P holds for 100 days

By the time the shock reaches the surface of the hydrogen envelope, hours after core bounce, the entire envelope is fully ionized and expanding outward at v_ej ~ 4000 km/s. Inside this expanding sphere, the photosphere — the radius where the optical depth to outgoing photons reaches τ ~ 2/3 — sits at the outer edge initially and retreats inward in mass coordinate as the envelope cools. The luminosity at the photosphere is L = 4πR² σT⁴, where T_phot tracks the recombination temperature of hydrogen at T ≈ 6000-7000 K — below this, the opacity collapses by orders of magnitude as electrons recombine onto protons, and the photosphere essentially "sees through" to the next inner ionized layer.

Two effects then conspire. R (the radius of the recombination front in physical space) grows with v_ej · t. But the recombination front in mass coordinate retreats inward at a rate that closely cancels the geometric expansion. The result is a luminosity that stays nearly constant — the plateau — for the time it takes the recombination front to walk through the entire ionized envelope. For a 10 M☉ envelope expanding at 4000 km/s, this is about 100 days.

Phase	Duration	Source of light	Characteristic L
Shock breakout	~ 1 hour	Shock crosses photosphere — UV/X-ray flash	10⁴⁵ erg/s (transient)
Rising photosphere	~ 5-15 days	Adiabatically cooling, expanding fireball	10⁸-10⁹ L☉ rising
Plateau (II-P)	~ 80-120 days	Hydrogen recombination front	~ 10⁹ L☉ flat
Photosphere drop	~ 20-40 days	Recombination reaches envelope base	10⁸-10⁹ L☉ falling
Radioactive tail	Years	⁵⁶Co → ⁵⁶Fe decay (t½ = 77 d)	10⁷-10⁸ L☉, 0.0098 mag/d slope

The II-L (linear) subclass is the same explosion in a thinner envelope. With only 1-3 M☉ of hydrogen left, the recombination front sweeps through it in 30-50 days, the plateau is truncated, and the light curve declines almost linearly from peak through the tail. II-P and II-L sit on a continuum parameterised by envelope mass, with II-L progenitors generally interpreted as RSGs that lost more mass through winds or binary interaction.

The teaching example: SN 1987A

SN 1987A in the Large Magellanic Cloud (d ≈ 50 kpc, magnitude 3 at peak — the brightest Galactic-neighbourhood supernova since Kepler's SN of 1604) remains the cleanest direct test of the Type II mechanism. Five hours before the optical brightening, Kamiokande-II (Japan), Irvine-Michigan-Brookhaven (US), and Baksan (USSR) recorded a total of 25 neutrino events in two 13-second bursts. The mean energy (~ 15 MeV) and total inferred energy budget (~ 3 × 10⁵³ erg) confirmed the basic picture: the binding energy of the new neutron star, radiated in neutrinos, on the predicted timescale. The discovery was the first observed astrophysical neutrino source other than the Sun.

The optical light curve was unusual: a slow rise to a relatively faint peak at M_V ≈ −15.5, no extended classical plateau, and a smooth fall onto the ⁵⁶Co decay tail. The reason was the progenitor: Sk −69 202 was a 20 M☉ blue supergiant of only ~ 50 R☉ rather than the expected ~ 1000 R☉ red supergiant. With a compact envelope, recombination produced only a stunted plateau, and the early-time luminosity was suppressed because much of the shock-deposited internal energy was converted to expansion work rather than radiation.

The post-supernova remnant has been continuously imaged for nearly four decades. A central neutron star is now expected (after JWST and ALMA detected dust-shrouded emission consistent with one) but never definitively imaged. The famous triple-ring system around the remnant — a luminous inner ring at r ≈ 0.6 ly and two fainter outer rings — was ionized by the shock breakout flash and subsequently slammed into by the expanding ejecta starting around 1995, producing an X-ray and optical light show that continues today.

Back-of-envelope: the plateau duration

If you take a hydrogen-rich envelope of mass M_env expanding at velocity v_ej and ask how long the recombination front takes to cross it inward in mass coordinate, you get the plateau duration. Using Arnett's analytic recombination model (Arnett 1980; Popov 1993), the result is

t_plat ≈ 100 d · (M_env / 10 M☉)^(1/2) · (E / 10⁵¹ erg)^(-1/4) · (R₀ / 500 R☉)^(1/6)

where R₀ is the pre-explosion radius and E is the kinetic energy of the ejecta. The mild scaling with R₀ is why even quite different progenitor radii give roughly 100-day plateaus, and the inverse scaling with E is why more energetic explosions — at fixed envelope mass — have shorter plateaus. The peak luminosity on the plateau scales as L ~ 10⁹ L☉ · (E / 10⁵¹)^(5/6) (R₀ / 500 R☉)^(2/3), which is comfortably consistent with the M_V ≈ −16.5 ± 1 distribution observed in II-P samples.

Common pitfalls

• Calling SN 1987A a "Type II-P". Strictly it was a Type II-pec (peculiar): the unusual blue supergiant progenitor produced a light curve that did not match either II-P or II-L templates. Its scientific value lies precisely in being a counterexample that tested the dependence of the light curve on envelope structure.
• Confusing Type II energetics with Type II light. The visible luminosity is 0.003% of the total release. Most of the energy is in neutrinos. The visible explosion is just the byproduct of a tiny fraction of the binding energy heating the envelope.
• Assuming every massive star ends as a Type II. Mass loss by line-driven winds and binary stripping removes the hydrogen envelope from a substantial fraction of massive stars before they explode. Wolf-Rayet stars produce Type Ib (no H, with He) and Ic (no H, no He) supernovae instead. The Type II classification depends on the envelope surviving, which depends sensitively on metallicity, rotation, and binary history.
• Treating the plateau as a single-mechanism feature. The flat plateau is the combination of geometric expansion and recombination-front retreat; either alone would not produce it. Models that ignore one or the other miss the shape.
• Forgetting the neutron star. Almost every Type II event leaves a neutron star. Failure to find a compact remnant in many post-supernova nebulae (especially SN 1987A, where the search has been intense) is a real puzzle, not evidence that no neutron star formed — dust obscuration and pulsar beaming away from us are the most likely explanations.
• Saying "Type II supernovae are exploding red supergiants". Most are, but a meaningful minority (perhaps 10%) come from yellow or blue supergiants, especially in binary systems where the envelope is partially stripped before collapse. The category is defined by the spectrum, not the progenitor.