Chapter 23 — Explosive Nucleosynthesis: Supernovae, Neutron Star Mergers, and the Origin of Heavy Elements

"We are stardust, billion-year-old carbon. We are golden, caught in the devil's bargain, and we've got to get ourselves back to the garden." — Joni Mitchell, "Woodstock" (1970)

"On 2017 August 17, a new era of multi-messenger astrophysics began." — Abbott et al., The Astrophysical Journal Letters (2017)

Chapter Overview

Chapter 22 traced the nuclear life of a star from hydrogen burning to the iron peak. That story ended on a cliff: silicon burning builds iron-group nuclei, the binding energy per nucleon reaches its maximum, and no further exothermic fusion is possible. For a massive star, this is a death sentence. The iron core grows until it can no longer support itself against gravity, and the star collapses — producing a core-collapse supernova, one of the most violent events in the universe.

But the supernova is not just an ending. It is a beginning — the beginning of explosive nucleosynthesis. In the shock-heated layers of the dying star, nuclear reactions occur on timescales of seconds, producing elements that quiescent stellar burning never could. And beyond the supernova, even more extreme environments — the mergers of neutron stars — produce the heaviest elements of all.

This chapter tells the story of how the universe makes its heaviest elements. We will:

• Analyze the physics of core collapse: from the stiffening of the nuclear equation of state at nuclear density to the neutrino-driven explosion mechanism.
• Distinguish two types of supernovae — core-collapse (Type II/Ib/Ic) and thermonuclear (Type Ia) — and their different nucleosynthetic signatures.
• Derive the conditions that separate the s-process (slow neutron capture, in AGB stars) from the r-process (rapid neutron capture, in extreme explosive environments).
• Follow the 2017 observation of GW170817 — the neutron star merger detected through gravitational waves and electromagnetic radiation — that confirmed neutron star mergers as r-process sites.
• Apply nuclear cosmochronology to estimate the age of the Galaxy from thorium and uranium abundances.

This is where nuclear physics meets the cosmos at its most dramatic. The gold in your jewelry, the uranium in nuclear reactors, the thorium geologists use to date rocks — all of it was forged in events so violent that they briefly outshine entire galaxies. Understanding how requires everything we have built: the nuclear force (Chapter 3), the semi-empirical mass formula (Chapter 4), radioactive decay (Chapter 12), neutron capture reactions (Chapter 18), and stellar burning (Chapter 22).

🏃 Fast Track: If you are primarily interested in the r-process and GW170817, skim Sections 23.1–23.2 (supernovae) and begin in depth at Section 23.4 (s-process) and Section 23.5 (r-process). Section 23.6 (GW170817) is the anchor example and should be read carefully by all students.

🔬 Deep Dive: The neutrino transport problem in core-collapse supernovae (Section 23.1.3) and the nuclear physics inputs to r-process calculations (Section 23.5.3) represent active frontiers of nuclear astrophysics research. These sections connect to Chapter 33 (Frontiers).

23.1 Core-Collapse Supernovae: When Gravity Wins

23.1.1 The Iron Core Crisis

At the end of Chapter 22, we left a massive star ($M \gtrsim 8\,M_\odot$) with an onion-shell structure: concentric layers of progressively heavier elements, from hydrogen on the outside to an iron-group core at the center. The core has been built by silicon burning — the last exothermic fusion stage — and consists primarily of ${}^{56}\text{Fe}$, ${}^{52}\text{Cr}$, ${}^{58}\text{Ni}$, and other iron-peak nuclei in nuclear statistical equilibrium (NSE).

Why does fusion stop at iron? Recall from Chapter 1 that the binding energy per nucleon $B/A$ reaches its maximum near $A \approx 56$–$62$. Fusing iron-peak nuclei into heavier elements would require energy input rather than releasing energy. The nuclear furnace has run out of fuel.

The iron core is supported against gravitational collapse by electron degeneracy pressure — the same quantum-mechanical effect that supports white dwarfs. But the core is growing as silicon burning continues in the surrounding shell, adding mass at a rate of approximately $0.5\,M_\odot$ per day. The core approaches a critical mass.

23.1.2 The Chandrasekhar Mass and the Onset of Collapse

The maximum mass that electron degeneracy pressure can support is the Chandrasekhar mass:

$$\boxed{M_{\text{Ch}} = 1.44 \left(\frac{Y_e}{0.5}\right)^2 M_\odot}$$

where $Y_e$ is the electron fraction (electrons per baryon). For the iron core, $Y_e \approx 0.42$ due to the neutron-rich composition of iron-peak nuclei, giving $M_{\text{Ch}} \approx 1.02\,M_\odot$. When the core mass exceeds this value, collapse begins.

Two processes accelerate the collapse:

1. Electron capture on iron-peak nuclei. As density increases, the electron Fermi energy rises. When it exceeds the threshold for electron capture:

$$e^- + {}^{56}\text{Fe} \to {}^{56}\text{Mn} + \nu_e$$

and more generally:

$$e^- + (A, Z) \to (A, Z-1) + \nu_e$$

Each capture removes an electron (reducing the degeneracy pressure) and produces a neutrino that initially escapes, carrying energy out of the core. This is a positive feedback loop: fewer electrons $\to$ less pressure $\to$ higher density $\to$ more captures.

2. Photodisintegration of iron. At temperatures exceeding $T \gtrsim 8 \times 10^9\,\text{K}$ ($kT \gtrsim 0.7\,\text{MeV}$), photons have enough energy to break apart iron nuclei:

$$\gamma + {}^{56}\text{Fe} \to 13\,{}^{4}\text{He} + 4n \qquad Q = -124.4\,\text{MeV}$$

This is endothermic — it absorbs $124.4\,\text{MeV}$ per iron nucleus, draining thermal energy from the core. At even higher temperatures:

$$\gamma + {}^{4}\text{He} \to 2p + 2n \qquad Q = -28.3\,\text{MeV}$$

In seconds, the nuclear fusion that built the iron core over millions of years is undone. The core, now a sea of free protons, neutrons, and electrons, collapses at speeds approaching $0.25c$ — roughly 70,000 km/s.

💡 Physical Insight: Photodisintegration is the reverse of stellar nucleosynthesis. The star spent millions of years fusing hydrogen to iron, and gravitational collapse reverses that work in less than a second. The energy budget is staggering: disassembling the iron core absorbs roughly $1.7 \times 10^{52}\,\text{erg}$ — about 100 times the kinetic energy of the eventual supernova explosion.

23.1.3 The Bounce and the Neutrino-Driven Mechanism

The collapsing inner core ($\sim 0.6\,M_\odot$) reaches nuclear density ($\rho_0 \approx 2.7 \times 10^{14}\,\text{g/cm}^3$) in approximately 100 milliseconds. At this point, the nuclear equation of state stiffens dramatically — the strong nuclear force becomes repulsive at short range (Chapter 3), and the matter becomes nearly incompressible. The inner core bounces, launching a shock wave outward into the still-infalling outer core.

The physics of the bounce is fundamentally nuclear. The incompressibility of nuclear matter — quantified by the nuclear compressibility modulus $K_0$ — determines the stiffness of the bounce:

$$K_0 = 9\rho_0^2 \frac{d^2(E/A)}{d\rho^2}\bigg|_{\rho = \rho_0} \approx 230 \pm 20\,\text{MeV}$$

This value, measured in laboratory experiments on giant monopole resonances in nuclei (the "breathing mode"), directly controls the dynamics of core collapse. Here is a case where a measurement on a small nucleus in a laboratory determines the fate of a star.

The stalling shock. The bounce shock carries about $5 \times 10^{51}\,\text{erg}$ of energy, but it must fight against the infalling outer core and loses energy to photodisintegration (each nucleon that crosses the shock absorbs about $8.8\,\text{MeV}$ to disassemble iron into nucleons). Within $\sim 10\,\text{ms}$, the shock stalls at a radius of $\sim 100$–$200\,\text{km}$.

The neutrino mechanism. The proto-neutron star at the center (the collapsed inner core) is enormously hot ($T \sim 30$–$50\,\text{MeV}$) and radiates its gravitational binding energy primarily as neutrinos. The total energy budget is:

$$E_\nu \approx \frac{3}{5}\frac{GM_{\text{NS}}^2}{R_{\text{NS}}} \approx \frac{3}{5}\frac{(6.67 \times 10^{-11})(1.4 \times 1.99 \times 10^{30})^2}{12 \times 10^3} \approx 3 \times 10^{53}\,\text{erg} \approx 3 \times 10^{46}\,\text{J}$$

This is an extraordinary number. The Sun's total luminosity is $L_\odot \approx 3.8 \times 10^{33}\,\text{erg/s}$; the proto-neutron star radiates $3 \times 10^{53}\,\text{erg}$ in about 10 seconds. For those 10 seconds, the neutrino luminosity of the collapsing core exceeds the combined luminosity of every star in the observable universe.

The energy is radiated roughly equally among all six neutrino flavors ($\nu_e$, $\bar{\nu}_e$, $\nu_\mu$, $\bar{\nu}_\mu$, $\nu_\tau$, $\bar{\nu}_\tau$):

$$E_\nu \approx 3 \times 10^{53}\,\text{erg}, \qquad L_\nu \sim 3 \times 10^{52}\,\text{erg/s}$$

If even a small fraction of this neutrino energy is deposited in the material behind the stalled shock, it can revive the explosion. The mechanism is primarily:

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

These charged-current reactions have cross sections that scale as $\sigma \propto E_\nu^2$, with typical values:

$$\sigma \sim \sigma_0 \left(\frac{E_\nu}{10\,\text{MeV}}\right)^2, \qquad \sigma_0 \approx 4 \times 10^{-44}\,\text{cm}^2$$

The energy deposition rate per nucleon is:

$$\dot{q} \sim \frac{L_\nu \sigma}{4\pi r^2} \langle E_\nu \rangle$$

Modern simulations show that depositing roughly 5–10% of the neutrino energy behind the shock (the "gain region") is sufficient to revive it. This is the neutrino-driven explosion mechanism, first proposed by Wilson (1985) and Bethe & Wilson (1985).

⚠️ Honest Assessment: Despite decades of work, the details of the explosion mechanism remain one of the great unsolved problems in nuclear astrophysics. 1D simulations (spherically symmetric) generally fail to produce explosions. 2D and 3D simulations, which include convection, the standing accretion shock instability (SASI), and turbulence, can produce successful explosions — but the results are sensitive to the neutrino transport scheme, the nuclear equation of state, and the progenitor model. The field is converging, but a definitive, first-principles prediction of which stars explode and which form black holes has not yet been achieved.

The energy budget — a summary:

The supernova we see — the brilliant optical display that can outshine a galaxy for weeks — represents only about 0.01% of the total energy release. Supernovae are neutrino events. The optical display is an afterthought.

23.1.4 SN 1987A: The Neutrino Detection That Changed Everything

On February 23, 1987, a core-collapse supernova — designated SN 1987A — was observed in the Large Magellanic Cloud at a distance of $51.4\,\text{kpc}$ ($\sim 168,000\,\text{light-years}$). It was the nearest supernova since Kepler's supernova of 1604 and the first to be observed with modern detectors.

Three hours before the optical brightening was noticed, neutrino detectors on three continents registered a burst of events:

These 25 events (reduced to 24 after reanalysis) were the first detection of neutrinos from a supernova — and indeed from any astrophysical source other than the Sun. The detection confirmed the fundamental prediction: the overwhelming majority of the gravitational binding energy ($\sim 3 \times 10^{53}\,\text{erg}$) is carried away by neutrinos.

Quantitative consistency check. The Kamiokande-II detector contained 2140 tonnes of water. The dominant detection reaction was inverse beta decay:

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

with a cross section $\sigma \sim 10^{-43}\,\text{cm}^2$ for $\langle E_{\bar{\nu}_e} \rangle \approx 15\,\text{MeV}$. The expected number of events is:

$$N_{\text{det}} \approx \frac{E_{\bar{\nu}_e}}{6 \langle E_{\bar{\nu}_e} \rangle} \cdot \frac{\sigma}{4\pi D^2} \cdot N_p$$

where $E_{\bar{\nu}_e} \approx 5 \times 10^{52}\,\text{erg}$ (one-sixth of the total neutrino energy), $D = 51.4\,\text{kpc} = 1.59 \times 10^{23}\,\text{cm}$, and $N_p \approx 1.4 \times 10^{32}$ (free protons in 2140 tonnes of water). This gives $N_{\text{det}} \sim 10$–$15$, consistent with the 12 observed events. Masatoshi Koshiba received the Nobel Prize in Physics (2002) for this detection.

SN 1987A also confirmed the ${}^{56}\text{Ni}$ light curve model. The supernova's bolometric light curve showed an exponential tail with a decline rate of $0.98\,\text{mag}$ per 100 days — precisely matching the ${}^{56}\text{Co}$ half-life of $77.2\,\text{d}$. Furthermore, gamma-ray lines from ${}^{56}\text{Co}$ decay at $847\,\text{keV}$ and $1238\,\text{keV}$ were directly detected by the Solar Maximum Mission and balloon-borne germanium detectors, confirming that $\sim 0.07\,M_\odot$ of ${}^{56}\text{Ni}$ was produced. SN 1987A remains the benchmark against which all core-collapse supernova models are tested.

23.1.5 Nucleosynthesis in the Supernova Shock

As the revived shock propagates outward through the star's onion-shell structure, it heats each layer to temperatures far exceeding those of quiescent burning. The timescale is seconds — too short for the nuclear reactions to reach equilibrium — so the resulting nucleosynthesis depends on both the peak temperature and the cooling rate.

Explosive oxygen and silicon burning. The shock heats the oxygen and silicon shells to $T \gtrsim 3$–$5 \times 10^9\,\text{K}$, driving reactions that produce iron-peak elements. The key products include:

• ${}^{56}\text{Ni}$: produced in explosive silicon burning via NSE (nuclear statistical equilibrium). This is the most important radioactive product — it powers the supernova light curve.
• ${}^{44}\text{Ti}$: a rare but important radioactive nucleus ($t_{1/2} = 58.9\,\text{yr}$) that traces the innermost ejecta.
• ${}^{28}\text{Si}$, ${}^{32}\text{S}$, ${}^{36}\text{Ar}$, ${}^{40}\text{Ca}$: "alpha elements" from incomplete explosive silicon burning.

The ${}^{56}\text{Ni}$ decay chain. The light curves of core-collapse supernovae are powered by the radioactive decay:

$${}^{56}\text{Ni} \xrightarrow[\text{EC}]{t_{1/2} = 6.08\,\text{d}} {}^{56}\text{Co} \xrightarrow[\beta^+/\text{EC}]{t_{1/2} = 77.2\,\text{d}} {}^{56}\text{Fe}$$

A typical core-collapse supernova produces $\sim 0.07\,M_\odot$ of ${}^{56}\text{Ni}$. The energy released by these decays — primarily through gamma rays from de-excitation of ${}^{56}\text{Co}$ and ${}^{56}\text{Fe}$ daughter states — heats the expanding ejecta and produces the characteristic optical plateau and exponential tail of the light curve.

Numerical check: The decay energy of ${}^{56}\text{Ni} \to {}^{56}\text{Co} \to {}^{56}\text{Fe}$ totals approximately $Q \approx 3.0\,\text{MeV}$ per decay. For $0.07\,M_\odot$ of ${}^{56}\text{Ni}$:

$$N_{\text{Ni}} = \frac{0.07 \times 1.99 \times 10^{33}\,\text{g}}{56 \times 1.66 \times 10^{-24}\,\text{g}} \approx 1.5 \times 10^{54}$$

$$E_{\text{total}} = N_{\text{Ni}} \times 3.0\,\text{MeV} \times 1.6 \times 10^{-6}\,\text{erg/MeV} \approx 7 \times 10^{48}\,\text{erg}$$

This is consistent with the observed radiated energy of core-collapse supernovae over weeks to months.

The neutrino-driven wind. After the explosion is launched, the proto-neutron star continues to radiate neutrinos. These neutrinos drive a wind of material off the neutron star surface — the neutrino-driven wind — that was long considered a candidate site for the r-process. The wind is proton- or neutron-rich depending on the ratio of electron neutrino to anti-electron neutrino luminosities. Modern calculations suggest the wind is generally not neutron-rich enough for a robust r-process (Section 23.5), though it may contribute to the production of some lighter heavy elements ($A \sim 80$–$100$).

23.2 Type Ia Supernovae: Thermonuclear Explosions of White Dwarfs

23.2.1 The Progenitor and the Mechanism

Type Ia supernovae are fundamentally different from core-collapse events. There is no iron core collapse, no neutron star remnant, no neutrino burst. Instead, a carbon-oxygen white dwarf — the remnant of a low- or intermediate-mass star — is driven to thermonuclear explosion.

The white dwarf accretes matter from a companion star (either a main-sequence/giant donor in the "single-degenerate" model, or merges with another white dwarf in the "double-degenerate" model). As the white dwarf approaches the Chandrasekhar mass:

$$M_{\text{Ch}} = 1.44\,M_\odot \quad \text{(for } Y_e = 0.5\text{)}$$

the central density and temperature rise until carbon ignites under degenerate conditions. Because degenerate matter does not expand when heated (the pressure is set by electron degeneracy, not temperature), there is no thermostatic regulation. The burning is a thermonuclear runaway — a nuclear bomb the mass of the Sun.

23.2.2 Nucleosynthesis in Type Ia Events

The detonation/deflagration wave propagates through the white dwarf, burning the material at different temperatures depending on the radial location:

A typical Type Ia supernova produces $\sim 0.6$–$0.8\,M_\odot$ of ${}^{56}\text{Ni}$ — roughly ten times more than a core-collapse event. This is why Type Ia supernovae are brighter and why they are the primary source of iron-peak elements in the Galaxy.

The standardizable candle. Because the ${}^{56}\text{Ni}$ mass determines the peak luminosity (more nickel = brighter supernova), and because the white dwarf progenitor mass is nearly the same ($\sim M_{\text{Ch}}$), Type Ia supernovae have a relatively uniform peak brightness. The correlation between peak luminosity and light-curve decline rate — the Phillips relation — makes them standardizable candles for cosmological distance measurements. This is how Perlmutter, Schmidt, and Riess discovered the accelerating expansion of the universe (Nobel Prize 2011) — a discovery that rests on the nuclear physics of ${}^{56}\text{Ni}$ decay.

🔗 Cross-Reference: The Chandrasekhar mass appeared in our discussion of white dwarf stability. The nuclear equation of state at white-dwarf densities ($\rho \sim 10^9\,\text{g/cm}^3$) is that of a relativistic degenerate electron gas — a very different regime from the nuclear densities encountered in core collapse ($\rho \sim 10^{14}\,\text{g/cm}^3$). Both are important applications of nuclear and particle physics to astrophysics.

23.3 Solar System Abundances: The Signature of Nucleosynthesis

Before diving into the s-process and r-process, we need to understand the data they must explain. The solar system abundance pattern — determined from meteoritic analyses (CI chondrites) and solar spectroscopy — is the Rosetta Stone of nucleosynthesis.

The key features of the abundance pattern for elements heavier than iron ($Z > 26$) are:

1. An overall steep decline with increasing mass number, reflecting the decreasing probability of building heavier nuclei.

2. Twin peaks at $A \approx 80$, $130$, and $195$. These appear as pairs: an "s-process peak" at the closed neutron shells ($N = 50, 82, 126$) and an "r-process peak" shifted to $\sim 10$ mass units lower.

3. A small but distinct set of proton-rich isotopes (the p-nuclei) that cannot be produced by either neutron capture process.

The existence of twin peaks is the key observational evidence that two distinct neutron capture processes operate in nature. The s-process peaks at $N = 50, 82, 126$ occur because nuclei with magic neutron numbers have very small neutron capture cross sections — they are "bottlenecks" where material accumulates. The r-process peaks are shifted to lower $A$ because the r-process path runs through very neutron-rich nuclei that, after the neutron flux ceases, beta-decay back to stability — and the decay path ends at the magic neutron numbers.

Understanding this pattern quantitatively requires understanding both processes.

23.4 The s-Process: Slow Neutron Capture

23.4.1 The Physical Setting

The s-process occurs in asymptotic giant branch (AGB) stars — evolved low- and intermediate-mass stars ($1$–$8\,M_\odot$) in the late stages of their lives. These stars have a characteristic structure:

• A degenerate carbon-oxygen core
• A helium-burning shell
• A hydrogen-burning shell
• A convective envelope

The neutron sources are:

$${}^{13}\text{C}(\alpha, n){}^{16}\text{O} \qquad \text{(main source, } T \sim 10^8\,\text{K)}$$

$${}^{22}\text{Ne}(\alpha, n){}^{25}\text{Mg} \qquad \text{(activated at higher } T \sim 3 \times 10^8\,\text{K)}$$

These reactions provide a modest neutron flux:

$$n_n \sim 10^6\text{–}10^{8}\,\text{n/cm}^3$$

This neutron density is the defining characteristic of the s-process. At these densities, the typical time between successive neutron captures on a given seed nucleus is months to years — much longer than the beta-decay half-lives of most unstable nuclei along the path.

The s-process operates during thermal pulses — periodic helium shell flashes that occur every $\sim 10^4$–$10^5$ years in AGB stars. During each pulse, the helium shell burns vigorously for $\sim 100$–$1000$ years, convectively mixing the newly synthesized material. Between pulses, the convective envelope may dip down into the processed region (third dredge-up), bringing freshly made s-process elements to the stellar surface — where they are observed by spectroscopists and eventually returned to the interstellar medium by the powerful AGB stellar wind ($\dot{M} \sim 10^{-7}$–$10^{-5}\,M_\odot/\text{yr}$).

The s-process is traditionally divided into three components based on the neutron exposure:

• Weak component ($A \lesssim 90$): operates during core helium burning in massive stars ($M > 8\,M_\odot$), driven by ${}^{22}\text{Ne}(\alpha,n)$. Produces elements from Fe to Sr.
• Main component ($90 \lesssim A \lesssim 208$): operates during thermal pulses in AGB stars ($1$–$3\,M_\odot$), driven primarily by ${}^{13}\text{C}(\alpha,n)$. Produces the bulk of s-process elements from Zr to Pb.
• Strong component ($A \sim 208$): an additional exposure in low-metallicity AGB stars that fills the lead peak at ${}^{208}\text{Pb}$.

23.4.2 The s-Process Path: Following the Valley of Stability

The competition between neutron capture and beta decay determines the path through the chart of nuclides. For a nucleus $(A, Z)$, the neutron capture rate is:

$$\lambda_n = n_n \langle \sigma v \rangle \approx n_n \sigma_n v_T$$

where $\sigma_n$ is the Maxwellian-averaged neutron capture cross section (typically evaluated at $kT = 30\,\text{keV}$, appropriate for AGB thermal pulses) and $v_T$ is the thermal velocity. The beta-decay rate is:

$$\lambda_\beta = \frac{\ln 2}{t_{1/2}}$$

The s-process condition is:

$$\boxed{\lambda_\beta \gg \lambda_n \quad \text{(s-process: beta decay is faster than neutron capture)}}$$

Numerical example. Consider neutron capture on ${}^{134}\text{Ba}$ ($\sigma_n \approx 180\,\text{mb}$ at $kT = 30\,\text{keV}$):

$$\lambda_n = n_n \sigma_n v_T = (10^8\,\text{cm}^{-3})(180 \times 10^{-27}\,\text{cm}^2)(1.38 \times 10^{8}\,\text{cm/s})$$ $$\lambda_n \approx 2.5 \times 10^{-18}\,\text{s}^{-1}$$

The mean time between captures:

$$\tau_n = 1/\lambda_n \approx 4 \times 10^{17}\,\text{s} \approx 13\,\text{yr}$$

Since most beta-unstable nuclei near the valley of stability have half-lives of minutes to days, $\lambda_\beta \gg \lambda_n$, and the unstable nucleus decays long before capturing another neutron. The s-process path therefore follows the valley of beta stability, stepping one neutron at a time and beta-decaying whenever it reaches an unstable isotope.

23.4.3 The Local Approximation and the $\sigma N_s$ Curve

In steady-state flow, the s-process flux through each isotope must be constant. This leads to the local approximation:

$$\boxed{\sigma_A \cdot N_s(A) \approx \text{constant}}$$

where $\sigma_A$ is the neutron capture cross section and $N_s(A)$ is the s-process abundance for mass number $A$. Isotopes with small cross sections (magic neutron numbers, $N = 50, 82, 126$) have large abundances — they act as bottlenecks. Conversely, isotopes with large cross sections are quickly transmuted and have low abundances.

The $\sigma N_s$ curve — a plot of $\sigma_A \cdot N_s(A)$ versus $A$ — is not perfectly flat, because: - The s-process is not truly in steady state (it operates during a finite number of thermal pulses in AGB stars) - There are branching points (see below) - The neutron exposure distribution is not a simple exponential

Nonetheless, the local approximation captures the essential physics and explains the s-process peaks at $A \approx 88$ ($N = 50$), $A \approx 138$ ($N = 82$), and $A \approx 208$ ($N = 126$).

23.4.4 Branching Points

At certain locations along the s-process path, a nucleus has a beta-decay half-life comparable to the neutron capture time ($\lambda_\beta \sim \lambda_n$). At these branching points, part of the s-process flow goes through beta decay and part goes through neutron capture, splitting the path.

A classic example is ${}^{85}\text{Kr}$ ($t_{1/2} = 10.76\,\text{yr}$):

• If $\lambda_\beta \gg \lambda_n$: ${}^{85}\text{Kr} \xrightarrow{\beta^-} {}^{85}\text{Rb}$, and the s-process bypasses ${}^{86}\text{Kr}$
• If $\lambda_n \gtrsim \lambda_\beta$: ${}^{85}\text{Kr}(n,\gamma){}^{86}\text{Kr}$, producing the s-only isotope ${}^{86}\text{Kr}$

The branching ratio depends on $n_n$ and $T$, making branching-point isotopes sensitive thermometers and neutron-density monitors for the s-process environment.

📊 Key Observation: Some isotopes are "s-only" (shielded from the r-process by a stable isobar), some are "r-only" (not reached by the s-process path), and some receive contributions from both. By measuring s-only and r-only abundances in the solar system, we can decompose the total heavy-element abundance pattern into s-process and r-process contributions. The r-process contribution is then determined as the solar r-process residual: $N_r(A) = N_\odot(A) - N_s(A)$.

23.4.5 The s-Process Endpoint

The s-process builds elements from iron-group seeds (${}^{56}\text{Fe}$) all the way to bismuth. At ${}^{209}\text{Bi}$ (the heaviest nucleus with a bound ground state that is effectively stable, $t_{1/2} = 1.9 \times 10^{19}\,\text{yr}$), neutron capture produces ${}^{210}\text{Bi}$, which beta-decays to ${}^{210}\text{Po}$, which alpha-decays to ${}^{206}\text{Pb}$. The s-process therefore cycles through a Pb-Bi loop and cannot build the transactinides. The s-process cannot make thorium, uranium, or any element heavier than bismuth. For those, we need the r-process.

23.5 The r-Process: Rapid Neutron Capture and the Heaviest Elements

23.5.1 The r-Process Condition

The r-process requires an environment so neutron-rich that neutron captures occur much faster than beta decays — the exact opposite of the s-process:

$$\boxed{\lambda_n \gg \lambda_\beta \quad \text{(r-process: neutron capture is faster than beta decay)}}$$

This demands neutron densities of:

$$n_n \gtrsim 10^{20}\,\text{n/cm}^3$$

— twelve or more orders of magnitude higher than in the s-process. At such extraordinary densities, a seed nucleus captures neutrons in a rapid burst, running far to the neutron-rich side of the chart of nuclides until the process is halted by either:

1. Photodisintegration equilibrium $(n,\gamma) \rightleftharpoons (\gamma,n)$: at high temperature, photodisintegration competes with neutron capture and establishes equilibrium along isotopic chains. The r-process path is then determined by the condition $S_n \approx 2$–$3\,\text{MeV}$ (where $S_n$ is the neutron separation energy).

2. The neutron drip line: at sufficiently neutron-rich compositions, the neutron separation energy drops to zero and additional neutrons are unbound. The nucleus literally cannot hold more neutrons.

23.5.2 The r-Process Path

The r-process path runs through nuclei far from stability — species with $10$–$30$ more neutrons than the most stable isotope of the same element. Most of these nuclei have never been produced in a laboratory and their properties (masses, half-lives, neutron capture cross sections) are unknown experimentally. This is one of the great challenges of r-process theory: the nuclear physics inputs are largely theoretical.

Waiting points. At magic neutron numbers ($N = 50, 82, 126$), the neutron separation energy drops sharply (the shell closure makes the nucleus tightly bound). At these locations: - Photodisintegration $(\gamma, n)$ prevents further neutron capture - The r-process "waits" for beta decay: $(A, Z) \to (A, Z+1) + e^- + \bar{\nu}_e$ - Beta decay increases $Z$ by one, allowing neutron capture to resume

The waiting-point nuclei have the longest effective half-lives along the path, and the r-process flow piles up at these locations. After the neutron flux ceases, the extremely neutron-rich nuclei beta-decay back toward stability, and the accumulated material at the waiting points produces the r-process abundance peaks at $A \approx 80$ ($N = 50$), $A \approx 130$ ($N = 82$), and $A \approx 195$ ($N = 126$).

The shift between s-process and r-process peaks is now clear: - s-process peaks at $A \approx 88, 138, 208$: built up at the magic neutron numbers in stable nuclei on the valley of stability - r-process peaks at $A \approx 80, 130, 195$: built up at the magic neutron numbers in neutron-rich nuclei far from stability, then shifted to lower $A$ by beta-decay back to stability

23.5.3 Nuclear Physics Inputs to the r-Process

Calculating the r-process abundance pattern requires knowledge of nuclear properties for thousands of nuclei far from stability:

Facilities like FRIB (Facility for Rare Isotope Beams, operational 2022 at Michigan State University), RIKEN's RIBF (Japan), and FAIR (under construction, Germany) are designed specifically to produce and study the neutron-rich nuclei on the r-process path. FRIB, in particular, is expected to reach approximately 80% of the r-process path nuclei — a transformative advance.

🔗 Cross-Reference: The exotic nuclei near the neutron drip line discussed in Chapter 10 are precisely the nuclei traversed by the r-process. The shell structure far from stability — whether the traditional magic numbers persist, shift, or are replaced — directly impacts the r-process abundance pattern. Changes in shell structure can shift the r-process peaks by several mass units.

23.5.4 The r-Process Endpoint: Fission Recycling

Unlike the s-process, which ends at bismuth, the r-process can build nuclei well beyond $A = 209$. As the path reaches the actinide region ($Z \gtrsim 90$), the nuclei become susceptible to neutron-induced fission:

$$n + (A, Z) \to \text{fission fragments}$$

and beta-delayed fission (the daughter of beta decay is produced in a state above its fission barrier). The fission fragments, typically with $A \sim 110$–$150$, are themselves neutron-rich and can undergo additional neutron captures. This fission recycling creates a feedback loop that:

1. Limits the maximum mass produced by the r-process
2. Contributes to the rare-earth peak ($A \sim 160$)
3. Establishes a "fission cycling" steady state if the neutron exposure is long enough

Fission recycling is a major source of uncertainty in r-process calculations because fission barriers and fragment distributions for the relevant nuclei ($A \gtrsim 240$, far from stability) are poorly known.

23.5.5 The Sixty-Year Mystery: Where Does the r-Process Happen?

The existence of the r-process was recognized by Burbidge, Burbidge, Fowler, and Hoyle (1957, the legendary B$^2$FH paper) and independently by Cameron (1957). But identifying the astrophysical site proved extraordinarily difficult. The requirements are stringent:

1. Extremely high neutron density ($n_n > 10^{20}\,\text{n/cm}^3$)
2. High temperature ($T \sim 10^9$–$10^{10}\,\text{K}$)
3. Rapid expansion (to freeze out the neutron-rich conditions before everything decays back to iron)
4. Ejection into the interstellar medium (to enrich the next generation of stars)

For decades, the leading candidates were:

• The neutrino-driven wind in core-collapse supernovae (Section 23.1.4) — attractive because supernovae are common, but modern calculations show the wind is generally not neutron-rich enough.
• Neutron star mergers — neutron-rich by construction (the material starts as neutron star matter, $Y_e \sim 0.01$–$0.1$), but considered too rare and too delayed (the binary must spiral together over millions to billions of years).
• Collapsars / jet-driven supernovae — the accretion disk around a newly formed black hole may produce r-process elements in a neutron-rich outflow.

The question was finally answered — at least in part — on August 17, 2017.

Why neutron star mergers are ideal r-process sites. Consider the conditions: neutron star matter has $Y_e \approx 0.01$–$0.05$ in the crust and outer core — an electron fraction far below that of any other astrophysical environment. When this material is ejected during a merger (through tidal tails, disk winds, and dynamical ejection), it decompresses from nuclear density to sub-nuclear density in milliseconds to seconds. During decompression, the free neutron fraction is enormous — neutron-to-seed ratios of $100:1$ or higher are achieved naturally. The temperature is high enough ($T \sim 5$–$10 \times 10^9\,\text{K}$) to establish $(n,\gamma) \rightleftharpoons (\gamma,n)$ equilibrium, and the expansion is rapid enough to freeze out the neutron-rich conditions before the material decays back to iron.

Furthermore, the ejecta mass per event ($\sim 0.01$–$0.05\,M_\odot$) is far larger than what the neutrino-driven wind produces ($\sim 10^{-4}\,M_\odot$). Even though neutron star mergers are rare (estimated rate $\sim 10$–$1000\,\text{Gpc}^{-3}\,\text{yr}^{-1}$, or roughly $10^{-5}$–$10^{-4}\,\text{yr}^{-1}$ in the Milky Way), the large ejecta mass per event can potentially account for the entire r-process inventory of the Galaxy.

23.6 GW170817: When Neutron Stars Collide

23.6.1 The Detection

On August 17, 2017, at 12:41:04 UTC, the LIGO (Hanford and Livingston, USA) and Virgo (Italy) gravitational-wave detectors recorded a signal designated GW170817. The signal lasted approximately 100 seconds — far longer than the $\lesssim 1\,\text{s}$ signals from binary black hole mergers — and swept upward in frequency from about 24 Hz to $\sim 600\,\text{Hz}$ as the two objects spiraled together.

The gravitational-wave signal encoded the chirp mass:

$$\mathcal{M} = \frac{(m_1 m_2)^{3/5}}{(m_1 + m_2)^{1/5}} = 1.188 \pm 0.001\,M_\odot$$

and constrained the individual component masses to $m_1 \in [1.36, 1.60]\,M_\odot$ and $m_2 \in [1.17, 1.36]\,M_\odot$. These masses are squarely in the neutron star range — too light for black holes and consistent with the mass distribution of neutron stars observed in binary pulsars.

Crucially, the tidal deformability parameter $\tilde{\Lambda}$ — which measures how much each neutron star is distorted by the gravitational field of its companion — was constrained to $\tilde{\Lambda} = 300^{+420}_{-230}$. This parameter depends directly on the nuclear equation of state (Chapter 25): stiffer equations of state produce larger, more deformable neutron stars. The measurement provided the first direct constraint on the equation of state from gravitational waves.

💡 Physical Insight: The tidal deformability of a neutron star is determined by its internal structure — which is governed by the nuclear force at extreme densities. When LIGO measures $\tilde{\Lambda}$, it is measuring nuclear physics. This is a profound example of the connection between the very small (nuclear forces) and the very large (neutron star mergers).

23.6.2 The Electromagnetic Counterpart

Just 1.7 seconds after the gravitational-wave signal, the Fermi Gamma-ray Burst Monitor detected a short gamma-ray burst — GRB 170817A. This confirmed a long-standing hypothesis: that short gamma-ray bursts are produced by neutron star mergers.

But the truly revolutionary observation came in the hours and days that followed. The merger was localized to the galaxy NGC 4993 at a distance of $\sim 40\,\text{Mpc}$ ($\sim 130$ million light-years), and telescopes around the world and in space turned to observe it.

What they found was a kilonova — a thermal transient powered by the radioactive decay of freshly synthesized r-process elements. The term had been coined by Metzger et al. (2010), who predicted that neutron star mergers should produce observable electromagnetic radiation through precisely this mechanism:

1. The merger ejects $\sim 0.01$–$0.05\,M_\odot$ of extremely neutron-rich material ($Y_e \sim 0.01$–$0.4$)
2. This material undergoes r-process nucleosynthesis as it decompresses
3. The r-process produces a broad distribution of heavy elements, many of which are radioactive
4. The radioactive decay heats the ejecta, which emits thermal radiation

23.6.3 The Kilonova: AT 2017gfo

The kilonova associated with GW170817, designated AT 2017gfo, was one of the most intensively observed transients in the history of astronomy. Key observations:

Early emission (first 1–2 days): A "blue" component peaking at optical wavelengths, with $T \sim 10^4\,\text{K}$. This was attributed to the decay of relatively light r-process elements ($A \lesssim 140$) in the polar ejecta, which has lower opacity due to fewer lanthanide-series elements.

Late emission (days 3–10+): A "red" component peaking in the near-infrared ($1$–$2\,\mu\text{m}$), with $T \sim 2000$–$3000\,\text{K}$. This was attributed to the decay of heavy r-process elements including lanthanides ($57 \leq Z \leq 71$) and actinides ($89 \leq Z \leq 103$), whose complex atomic structures produce enormous opacities in the UV and optical.

The nuclear physics of the opacity is striking. Lanthanide elements have partially filled $4f$ electron shells, producing millions of spectral lines that effectively block optical light and force the emission into the infrared. The presence of lanthanides in the ejecta is therefore diagnosed by the color evolution of the kilonova — a transition from blue to red over $\sim 1$ week.

The heating rate. The luminosity of the kilonova is set by the radioactive heating rate from r-process decays. For a broad distribution of r-process nuclei, the heating rate follows an approximate power law:

$$\dot{Q}(t) \approx 2 \times 10^{10}\,\text{erg}\,\text{s}^{-1}\,\text{g}^{-1} \times \left(\frac{t}{1\,\text{day}}\right)^{-1.3}$$

This power law arises because the r-process produces thousands of radioactive species with a broad distribution of half-lives; at any given time $t$, the heating is dominated by nuclei with $t_{1/2} \sim t$. This was derived analytically by Metzger et al. (2010) and confirmed by detailed nucleosynthesis calculations.

For AT 2017gfo, the observed luminosity was consistent with $\sim 0.04\,M_\odot$ of r-process ejecta — enough to produce approximately $10\,M_\oplus$ (ten Earth masses) of material heavier than iron, including an estimated $\sim 3$–$13\,M_\oplus$ of gold and $\sim 3$–$13\,M_\oplus$ of platinum.

23.6.4 Spectroscopic Identification of Strontium

The definitive nuclear physics confirmation came from spectroscopy. Watson et al. (2019) identified absorption features of strontium (${}^{88}\text{Sr}$, $Z = 38$) in the early spectra of AT 2017gfo. Strontium is a light r-process element ($A = 88$, at the first r-process peak, $N = 50$) with relatively simple atomic spectra that can be confidently identified.

The identification was made by comparing the observed spectra to synthetic spectra computed with the atomic transition data for Sr II (singly ionized strontium). The P Cygni profile — an absorption dip blueshifted by $\sim 0.2$–$0.3c$ from the emission peak — is characteristic of rapidly expanding ejecta. The best-fit expansion velocity was $v \approx 0.2$–$0.3c$, consistent with the expected ejecta speeds from neutron star mergers.

This was the first unambiguous spectroscopic identification of an r-process element in a kilonova. Combined with the gravitational-wave identification of the source as a neutron star merger, it provided the most direct evidence yet that neutron star mergers are r-process sites.

23.6.5 The Significance of GW170817

The observation of GW170817 was one of the great scientific events of the 21st century. A single event combined:

1. Gravitational-wave astronomy — measuring the masses, spins, and equation of state of neutron stars
2. Gamma-ray astronomy — confirming the origin of short gamma-ray bursts
3. Optical and infrared astronomy — observing the kilonova and its spectral evolution
4. Nuclear astrophysics — confirming r-process nucleosynthesis in neutron star mergers
5. Cosmology — providing an independent measurement of the Hubble constant from the gravitational-wave "standard siren"

More than 4,500 physicists and astronomers from 70+ observatories contributed to the discovery papers. The event was detected across the entire electromagnetic spectrum — from radio waves to gamma rays — plus gravitational waves. This is multi-messenger astronomy at its finest.

🔴 Critical Assessment: GW170817 demonstrated that neutron star mergers can produce r-process elements. It did not prove that they are the only site, or even the dominant site. The r-process enrichment of very old, metal-poor stars in the Galactic halo — some of which formed within the first billion years of the Galaxy — is difficult to explain with neutron star mergers alone, because the time delay for binary inspiral may be too long. The debate between neutron star mergers, rare supernovae (collapsars, magneto-rotational supernovae), and other sites continues. What GW170817 established beyond doubt is that the merger of two neutron stars produces an environment capable of forging the heaviest elements.

23.7 The p-Process: Proton-Rich Heavy Nuclei

Not all heavy nuclei can be produced by neutron capture. Approximately 35 proton-rich isotopes — the p-nuclei — are shielded from both the s-process and r-process paths by stable isotopes. Examples include ${}^{92}\text{Mo}$, ${}^{96}\text{Ru}$, ${}^{144}\text{Sm}$, and ${}^{196}\text{Hg}$.

The p-nuclei are the rarest heavy isotopes (typically $0.01$–$1$% of the element's total abundance) and are thought to be produced primarily by photodisintegration of pre-existing s-process and r-process seed nuclei during explosive burning in supernovae:

$$(\gamma, n), \quad (\gamma, p), \quad (\gamma, \alpha) \quad \text{reactions at } T \sim 2\text{–}3 \times 10^9\,\text{K}$$

This is sometimes called the $\gamma$-process. The sequence begins with $(\gamma, n)$ reactions that strip neutrons from heavy seed nuclei, moving them to the proton-rich side. At certain points, $(\gamma, \alpha)$ and $(\gamma, p)$ reactions redirect the flow.

Two p-nuclei are particularly problematic: ${}^{92}\text{Mo}$ and ${}^{94}\text{Mo}$ are severely underproduced in $\gamma$-process calculations. Their production may require the $\nu p$-process — proton captures in the neutrino-driven wind of core-collapse supernovae, where neutrino interactions maintain a proton-rich composition ($Y_e > 0.5$).

23.8 Cosmochronology: Dating the Galaxy with Nuclear Physics

23.8.1 The Principle

Long-lived radioactive nuclei produced by the r-process serve as nuclear clocks. If we know the initial production ratio of two r-process nuclei with different half-lives, and we measure their present-day ratio, we can infer the time elapsed since production.

The most useful pair is ${}^{232}\text{Th}$ ($t_{1/2} = 14.05\,\text{Gyr}$) and ${}^{238}\text{U}$ ($t_{1/2} = 4.468\,\text{Gyr}$). Both are produced exclusively by the r-process, and their long half-lives make them sensitive to timescales comparable to the age of the Galaxy.

23.8.2 The Cosmochronology Equation

If the production ratio at time $t = 0$ was $({}^{232}\text{Th}/{}^{238}\text{U})_0$ and the present ratio is $({}^{232}\text{Th}/{}^{238}\text{U})_\text{now}$, then:

$$\frac{N_{232}(t)}{N_{238}(t)} = \frac{N_{232}(0)}{N_{238}(0)} \cdot \exp\left[-\left(\lambda_{232} - \lambda_{238}\right)t\right]$$

where $\lambda = \ln 2 / t_{1/2}$. Solving for $t$:

$$\boxed{t = \frac{1}{\lambda_{238} - \lambda_{232}} \ln\left[\frac{({}^{232}\text{Th}/{}^{238}\text{U})_0}{({}^{232}\text{Th}/{}^{238}\text{U})_\text{now}}\right]}$$

Numerical application. The present-day solar system ratio is $({}^{232}\text{Th}/{}^{238}\text{U})_\text{now} \approx 3.9$ (by number). The r-process production ratio, estimated from nucleosynthesis calculations, is $({}^{232}\text{Th}/{}^{238}\text{U})_0 \approx 1.5 \pm 0.3$.

The decay constants are:

$$\lambda_{232} = \frac{\ln 2}{14.05\,\text{Gyr}} = 0.0493\,\text{Gyr}^{-1}, \qquad \lambda_{238} = \frac{\ln 2}{4.468\,\text{Gyr}} = 0.1551\,\text{Gyr}^{-1}$$

$$t = \frac{1}{0.1551 - 0.0493}\ln\left(\frac{1.5}{3.9}\right) = \frac{1}{0.1058}\ln(0.385) = \frac{-0.955}{0.1058} \approx 9.0\,\text{Gyr}$$

Wait — a negative argument of the logarithm? No: $({}^{232}\text{Th}/{}^{238}\text{U})_\text{now} > ({}^{232}\text{Th}/{}^{238}\text{U})_0$ because ${}^{238}\text{U}$ decays faster (shorter half-life), enriching the ratio in ${}^{232}\text{Th}$ over time. The equation gives:

$$t = \frac{1}{0.1058}\ln\left(\frac{1.5}{3.9}\right) = \frac{-0.9555}{0.1058} \approx -9.0\,\text{Gyr}$$

The negative sign arises because we defined the equation with $(\lambda_{238} - \lambda_{232})$ in the denominator and the ratio in the logarithm has $(R_0 / R_\text{now}) < 1$. Re-deriving carefully:

$$R(t) = R_0 \cdot e^{-(\lambda_{232} - \lambda_{238})t} = R_0 \cdot e^{+(\lambda_{238} - \lambda_{232})t}$$

Since $\lambda_{238} > \lambda_{232}$, $R(t)$ grows with time (more ${}^{238}\text{U}$ decays away relative to ${}^{232}\text{Th}$). Then:

$$t = \frac{1}{\lambda_{238} - \lambda_{232}} \ln\left(\frac{R_\text{now}}{R_0}\right) = \frac{1}{0.1058}\ln\left(\frac{3.9}{1.5}\right) = \frac{0.956}{0.1058} \approx 9.0\,\text{Gyr}$$

This is the time since the r-process material in the solar system was produced. Adding the age of the solar system ($4.57\,\text{Gyr}$), the r-process production event occurred about $13.6\,\text{Gyr}$ ago — consistent with the age of the universe ($13.8\,\text{Gyr}$ from Planck CMB measurements). In reality, r-process production was not a single event but occurred over the history of the Galaxy, requiring a more sophisticated chemical evolution model — but the simple calculation gives a remarkably good first estimate.

23.8.3 Stellar Cosmochronology

The same principle can be applied to individual stars. In 2001, Cayrel et al. measured the thorium abundance in the ultra-metal-poor star CS 22892-052 (a star in the Galactic halo with iron abundance $[\text{Fe/H}] = -3.1$, meaning $\sim 1/1000$ of the solar iron abundance). The measured ${}^{232}\text{Th}/{}^{238}\text{U}$ ratio implied an age of $12.5 \pm 3\,\text{Gyr}$, consistent with other age determinations for the oldest stars.

More recently, the discovery of actinide-boost r-process stars — stars with unusually high thorium and uranium abundances relative to lighter r-process elements — has provided additional cosmochronometric targets. The star J0954+5246, with an actinide boost factor of $\sim 3$, was dated to $13.0 \pm 1.5\,\text{Gyr}$, making it one of the oldest known stars.

💡 Physical Insight: Nuclear cosmochronology is remarkable: by measuring the ratio of two radioactive isotopes produced by nuclear reactions inside a neutron star merger billions of years ago, we can determine when that merger occurred. The clock is set by the nuclear force (which determines the half-lives) and read by atomic spectroscopy (which measures the abundances). Nuclear physics dates the Galaxy.

23.9 Putting It All Together: The Origin of the Elements

We can now summarize the origin of every element in the periodic table — a synthesis of Chapters 22 and 23:

🔗 Cross-Reference: This table integrates material from Chapters 21 (fusion), 22 (stellar nucleosynthesis), 23 (this chapter), and 24 (Big Bang nucleosynthesis). The cosmic ray spallation of Li, Be, B is a topic we touch on only briefly — it is a non-thermal process that operates in the interstellar medium, not in stellar interiors.

23.10 Project Checkpoint: r-Process Path Visualization

This chapter's contribution to the Nuclear Data Analysis Toolkit is r_process_path.py, which visualizes the s-process and r-process paths on the chart of nuclides.

What the Script Does

The script:

1. Plots a simplified chart of nuclides ($N$ vs. $Z$) showing stable nuclei (valley of stability)
2. Overlays the s-process path, which follows the valley of stability, stepping one neutron at a time and beta-decaying at unstable isotopes
3. Overlays the r-process path, which runs along contours of constant neutron separation energy ($S_n \approx 2$–$3\,\text{MeV}$), far to the neutron-rich side
4. Marks the magic neutron numbers ($N = 50, 82, 126$) and shows how they create waiting points on the r-process path and bottleneck peaks on the s-process path
5. Indicates the beta-decay flow from the r-process path back to stability after freeze-out

Key Physics Illustrated

• The s-process hugs the valley of stability; the r-process runs $10$–$30$ neutron numbers beyond it
• Both processes pile up at magic neutron numbers, producing the twin peaks in the abundance pattern
• The r-process peaks are shifted to lower $A$ than the s-process peaks because the waiting-point nuclei have fewer protons (for the same $N$) and subsequently beta-decay to lower $A$

See code/r_process_path.py for the complete script and code/project-checkpoint.md for documentation.

💻 Computational Note: The script requires numpy and matplotlib. Run it with: python r_process_path.py. The nuclear data (stable isotopes, approximate drip line, and representative s-process and r-process paths) are embedded directly in the script using published nuclear data values.

Chapter Summary

1. Core-collapse supernovae occur when the iron core of a massive star ($M \gtrsim 8\,M_\odot$) exceeds the Chandrasekhar mass. Electron capture and photodisintegration accelerate the collapse. The core bounces at nuclear density, and the neutrino-driven mechanism (depositing $\sim 5$–$10$% of the $3 \times 10^{53}\,\text{erg}$ neutrino energy) revives the stalled shock.

2. Explosive nucleosynthesis in the supernova shock produces iron-peak elements (especially ${}^{56}\text{Ni}$, which powers the light curve through its decay chain ${}^{56}\text{Ni} \to {}^{56}\text{Co} \to {}^{56}\text{Fe}$) and alpha elements (${}^{28}\text{Si}$, ${}^{32}\text{S}$, ${}^{40}\text{Ca}$).

3. Type Ia supernovae — thermonuclear explosions of carbon-oxygen white dwarfs near the Chandrasekhar mass — produce $\sim 0.6$–$0.8\,M_\odot$ of ${}^{56}\text{Ni}$ and are the dominant source of iron in the Galaxy.

4. The s-process (slow neutron capture) occurs in AGB stars at neutron densities $n_n \sim 10^{6}$–$10^{8}\,\text{cm}^{-3}$. Because $\lambda_\beta \gg \lambda_n$, the path follows the valley of stability. The local approximation $\sigma_A N_s(A) \approx \text{const}$ explains the s-process abundance peaks at magic neutron numbers. The s-process builds elements from Fe to Bi.

5. The r-process (rapid neutron capture) requires neutron densities $n_n \gtrsim 10^{20}\,\text{cm}^{-3}$. Because $\lambda_n \gg \lambda_\beta$, the path runs far to the neutron-rich side. Waiting points at magic neutron numbers create the r-process abundance peaks, shifted $\sim 10$ mass units below the s-process peaks. The r-process builds elements through and beyond uranium.

6. GW170817 (August 17, 2017) was the first observed neutron star merger detected through gravitational waves (LIGO/Virgo) with an electromagnetic counterpart (kilonova AT 2017gfo). Spectroscopic identification of strontium confirmed r-process nucleosynthesis. This multi-messenger event established neutron star mergers as sites of heavy-element production.

7. The p-process ($\gamma$-process) produces $\sim 35$ proton-rich isotopes through photodisintegration of heavy seeds in supernova explosive layers.

8. Cosmochronology uses the ${}^{232}\text{Th}/{}^{238}\text{U}$ ratio (and similar pairs) to date the r-process production events. The simple Th/U chronometer gives an age of $\sim 13$–$14\,\text{Gyr}$ for the Galaxy, consistent with independent determinations.

What's Next

In Chapter 24, we go back to the very beginning: Big Bang nucleosynthesis in the first three minutes of the universe. The primordial abundances of hydrogen, deuterium, helium-3, helium-4, and lithium-7 are determined by nuclear reaction rates at temperatures of $\sim 10^9$–$10^{10}\,\text{K}$ — and they provide one of the most precise determinations of the baryon density of the universe. The nuclear physics is simpler than the r-process (a network of only $\sim 12$ reactions), but the cosmological implications are profound.