Neutrino-Driven Nucleosynthesis: How Supernovae Build Heavy Elements

Iron is the heaviest element that stellar fusion, on its own, can produce. Beyond iron, the binding energy per nucleon decreases; fusion becomes endothermic and does not release the energy that powers the star. Yet the universe contains copious elements heavier than iron — including all the metallic precious metals, the lanthanides and actinides, uranium, thorium, all the way to the fission limit around mass 260.

Where do these elements come from? The answer involves two nucleosynthesis processes that bypass the iron barrier by using free neutrons as the tool. The s-process (slow neutron capture), operating in the interiors of asymptotic-giant-branch stars over timescales of thousands of years, produces about half the heavy-element inventory. The r-process (rapid neutron capture), operating in the extreme outflows of core-collapse supernovae and neutron-star mergers over timescales of seconds, produces the other half — specifically the neutron-rich isotopes above mass ~130 and the actinides.

Both processes require neutrons. The s-process finds them in the radiative helium-burning zones of red giants. The r-process finds them in the hot, dense, neutron-rich material ejected by catastrophic stellar endings. And it is in these extreme environments that neutrinos — the particle that seems so passive at Earth-scale energies — become dominant agents in setting the composition of cosmic matter.

This post walks through how neutrino physics shapes what elements the universe produces, from the nuclear reactions in the supernova outflow to the kilonova light we now see from neutron-star mergers.

The r-process requirement

The r-process synthesises a nucleus by rapid addition of neutrons: a seed (typically iron-peak) absorbs neutrons one after another, faster than the intermediate nuclei can beta-decay. The result is a highly neutron-rich nucleus that eventually, once the neutron flux subsides, beta-decays back toward the stable isotopes.

For this to work:

• Very high neutron density: 10²⁰–10²² neutrons per cubic centimetre, sustained for fractions of a second
• High temperature: 1–5 × 10⁹ K, so that thermonuclear reactions are in quasi-statistical equilibrium with the neutron flux
• Seed abundance: enough iron-peak nuclei to start the chain (~10⁻⁶ of total mass suffices)
• Neutron-to-seed ratio of ~100 or more, so each seed can capture enough neutrons to reach the r-process terminus

These conditions are not found anywhere in stable stars. They are found in two places:

1. The outflow of a core-collapse supernova, particularly in the neutrino-driven wind from the proto-neutron star surface
2. Neutron-star mergers, where tidal disruption ejects highly neutron-rich material

Both pathways produce r-process elements, and for most of the past sixty years the community has debated which is dominant. Recent evidence — especially the GW170817 kilonova observation in 2017 — suggests that neutron-star mergers are the primary source of heavy r-process elements (A > 130), while supernovae may contribute the lighter r-process and lanthanides.

How neutrinos set the neutron-proton ratio

In the first seconds after the core of a massive star collapses, the proto-neutron star produced is hot — several tens of MeV — and emits neutrinos of all six species (ν_e, ν_μ, ν_τ and their antiparticles) at approximately $10^{53}$ erg over 10 seconds. The neutrinos escape from a “neutrinosphere” at a radius of tens of kilometres, with ν_e and ν̄_e decoupling at slightly different temperatures because of their different interaction cross-sections with the protons and neutrons in the neutrinosphere.

In the layer immediately above the neutrinosphere — the region that will be blown outward in the “neutrino-driven wind” — the ν_e and ν̄_e interact with the nucleons:

${\nu }_e+n\to e^{-}+p$ ${\bar{\nu }}_e+p\to e^{+}+n$

These two reactions continuously convert neutrons to protons and protons to neutrons. The equilibrium ratio depends on the relative rates of the two reactions, which in turn depend on the spectra of the escaping ν_e and ν̄_e. If the ν̄_e are hotter than the ν_e (as is typical because they decouple at higher density), the second reaction outpaces the first, and the outflow becomes progressively more neutron-rich.

The electron fraction $Y_e$ — the number of protons per nucleon — is the key parameter: $Y_e=\frac{n_p}{n_p+n_n}$ A proton-rich outflow has $Y_e>0.5$; a neutron-rich outflow has $Y_e<0.5$.

For the r-process to occur, $Y_e$ must be roughly 0.3 or lower. Whether the supernova wind achieves this depends sensitively on the neutrino spectra. Early supernova simulations (1990s) predicted $Y_e\approx 0.49$ in the wind, just barely on the proton-rich side of the threshold — not low enough for r-process. More recent simulations with improved neutrino transport give $Y_e$ in the range 0.43–0.48 — still probably not low enough, at least not for the heavy r-process.

This is one of the reasons the community has gradually shifted toward neutron-star mergers as the primary heavy-r-process site: the tidal-disruption ejecta of a NS-NS merger has $Y_e\approx 0.1$ naturally, well below the r-process threshold, without needing fine-tuning of the neutrino spectra.

The GW170817 kilonova confirmation

On 17 August 2017, the LIGO and Virgo gravitational-wave detectors observed a clear signal of a neutron-star merger — designated GW170817. Within 1.7 seconds, Fermi satellite and INTEGRAL observatories detected a short gamma-ray burst coincident with the merger. Within hours, optical observatories (Swope, VISTA, DLT40) identified an electromagnetic counterpart in the NGC 4993 galaxy. Over the following weeks, the spectrum and light-curve of this counterpart were matched to kilonova models — the predicted signature of r-process nucleosynthesis in merger ejecta.

The kilonova observation allowed direct inference of the amount of r-process material produced: approximately 0.05 solar masses of heavy elements, including gold (several Earth masses worth), platinum, and the lanthanides. Multiplying this yield by the inferred galactic merger rate (~100 per million years), the kilonova yield accounts for essentially the full galactic inventory of heavy r-process elements.

This was the first direct confirmation that a specific astrophysical process produces r-process material at the right abundance. Neutron-star mergers as the dominant heavy-r-process site went from theoretical hypothesis to observed fact in a single day.

What remains open

Several aspects of the neutrino-nucleosynthesis connection are active research questions:

Flavour-dependent effects in supernova outflows. The neutrino-driven wind’s electron fraction depends on the specific spectra of ν_e and ν̄_e, which in turn depend on non-linear effects in the neutrinosphere — “collective oscillations”, fast flavour conversions, matter-induced mixing. These effects are not yet fully incorporated into supernova simulations, and small changes in their treatment can change the predicted r-process yield by orders of magnitude.

The role of supernovae in lighter r-process production. While neutron-star mergers dominate the heavy r-process (A > 130), supernovae may still contribute the lighter r-process (A = 90–130) through their magnetorotational variants or through proton-rich “νp-process” nucleosynthesis. The observational distinction between these sites is still being worked out.

Primordial r-process contribution. Some of the oldest r-process elements observed in metal-poor halo stars require a production event in the very early universe — earlier than neutron-star mergers could have occurred. Whether this came from primordial supernovae or from more exotic early-universe sites is unresolved.

Why this matters

Neutrinos are often described as passive particles that barely interact with matter. In the specific environment of a core-collapse supernova or neutron-star merger, that description is exactly backwards. Neutrinos are the dominant energy carrier (carrying away 99% of the gravitational binding energy of the collapse), the dominant driver of the outflow (the “neutrino wind” that launches the ejecta), and the dominant factor determining what elements the ejecta will contain.

Every atom of gold on Earth, every milligram of platinum, the uranium in every nuclear reactor — all of it was forged in an event where neutrinos shaped the isotopic composition of the material that eventually condensed into our solar system. The bridge from particle physics to cosmic chemistry passes directly through the physics of these tiny, seemingly inconsequential particles in their one environment of genuine relevance.

The next galactic supernova, whenever it occurs, will produce a neutrino burst observable in IceCube, DUNE, Super-Kamiokande, KamLAND, and other detectors — giving us direct sensitivity to the neutrino spectra that set the r-process conditions. Combined with gravitational-wave and electromagnetic follow-up of the next neutron-star merger, we will finally close the loop between fundamental neutrino physics and the origin of the elements we are made of.