Seconds after a neutron-star merger, the ejected matter is so neutron-rich (nn ∼ 10²&sup8; cm⁻³) that iron-group seed nuclei capture neutrons far faster than they can beta-decay. Each nuclide (N neutrons, Z protons) climbs in N until neutron capture and photodissociation balance — the (n,γ)⇆(γ,n) equilibrium, a Saha-like condition:
n_n · λ(n,γ) = λ(γ,n)
→ waiting point set by neutron separation energy S_n(Z,N)
S_n drops sharply at closed neutron shells N = 50, 82, 126
At that "waiting point" the nuclide can only advance by beta-minus decay (n → p + e⁻ + ν̄), which raises Z by 1 and lets neutron capture resume toward the new, higher equilibrium N. Near a closed shell the beta half-life jumps by orders of magnitude, so nuclei pile up there before decaying — and after the neutron flux runs out ("freeze-out"), that pile-up decays back down isobaric chains (A = N+Z stays fixed) to stable nuclei. The result is exactly the three abundance peaks seen in real r-process yields, near mass numbers A ≈ 80, 130 and 195, tied to N = 50, 82, 126.
- Neutron flux — how fast nuclides climb toward each equilibrium N (higher nn = more neutron-rich path, further from stability).
- Ejecta temperature — higher T9 strengthens photodissociation (γ,n), pulling the equilibrium back toward lower N (a cooler wind reaches further out).
- Seed nucleus — the iron-group starting point; the same physics runs regardless of seed, since it is a “waiting-point” process, not seed-dependent.
- Neutron-to-seed ratio — the free neutrons available per seed nucleus; once it hits zero the flux is spent and freeze-out beta decay takes over, exactly as in a real kilonova wind lasting roughly a second.
This process is how essentially all of the universe's gold, platinum and uranium were made — confirmed observationally by the kilonova AT2017gfo that followed the gravitational-wave event GW170817.