Case Study 2: The s-Process vs. the r-Process — Two Paths to Heavy Elements

The Puzzle of the Twin Peaks

In the 1950s, when Suess and Urey compiled the first accurate solar system abundance table, a striking pattern emerged in the heavy elements ($A > 60$): the abundance curve showed prominent peaks, but many of these peaks came in pairs. Near each neutron magic number ($N = 50, 82, 126$), there was a peak at a higher mass number and a second, broader peak at a lower mass number, separated by roughly 5 to 10 mass units.

This twin-peak structure was the key observational clue that led Burbidge, Burbidge, Fowler, and Hoyle (1957) and Cameron (1957) to propose two distinct neutron capture processes. The "s-process peaks" at $A \approx 88, 138, 208$ and the "r-process peaks" at $A \approx 80, 130, 195$ are among the most important features in all of nucleosynthesis, and understanding them requires comparing the two processes in detail.

This case study traces a single element — barium ($Z = 56$) — through both the s-process and the r-process, showing how the same magic neutron number ($N = 82$) produces its signature in two very different ways.

Part 1: Barium and the s-Process

The s-Process Path to Barium

In an AGB star, the s-process builds heavy elements by neutron capture on iron-group seeds. By the time the s-process flow reaches the $Z \sim 56$ region, the nuclei are following the valley of stability, stepping one neutron at a time:

$${}^{134}\text{Ba}(n,\gamma){}^{135}\text{Ba}(n,\gamma){}^{136}\text{Ba}(n,\gamma){}^{137}\text{Ba}(n,\gamma){}^{138}\text{Ba}$$

${}^{138}\text{Ba}$ has $N = 82$ — a closed neutron shell. Its neutron capture cross section is remarkably small:

$$\sigma_n({}^{138}\text{Ba}) = 4.0 \pm 0.2\,\text{mb} \quad \text{(at } kT = 30\,\text{keV)}$$

Compare this to a typical cross section just beyond the shell closure:

$$\sigma_n({}^{139}\text{La}) = 32 \pm 2\,\text{mb}$$

The factor-of-8 difference means that the s-process flow stalls at ${}^{138}\text{Ba}$. While other isotopes are quickly transmuted by neutron capture, ${}^{138}\text{Ba}$ resists, accumulating a large abundance. The mean time between neutron captures at $n_n = 10^8\,\text{cm}^{-3}$ is:

$$\tau_n({}^{138}\text{Ba}) = \frac{1}{n_n \sigma_n v_T} = \frac{1}{(10^8)(4 \times 10^{-27})(1.38 \times 10^8)} \approx 1.8 \times 10^{10}\,\text{s} \approx 570\,\text{yr}$$

This is long compared to typical AGB thermal pulse intervals ($\sim 10^4$–$10^5\,\text{yr}$), so ${}^{138}\text{Ba}$ builds up over multiple pulse cycles.

The s-Process Signature in Barium

The s-process produces a characteristic isotopic pattern in barium:

${}^{136}\text{Ba}$ is particularly important: it is an s-only isotope, shielded from the r-process beta-decay chain by the stable isobar ${}^{136}\text{Xe}$. This means the r-process decay chain at $A = 136$ ends at ${}^{136}\text{Xe}$ and never reaches ${}^{136}\text{Ba}$. The entire solar system inventory of ${}^{136}\text{Ba}$ was produced by the s-process.

Observational Evidence: s-Process Barium in AGB Stars

The s-process enrichment of barium has been directly observed in AGB stars and their descendants. Barium stars — spectral class G and K giants with anomalously strong Ba II absorption lines — are believed to be the binary companions of former AGB stars that transferred s-process-enriched material through Roche-lobe overflow or stellar winds.

The isotopic ratio ${}^{138}\text{Ba}/{}^{136}\text{Ba}$ in these stars can in principle distinguish s-process from r-process enrichment, though the measurement is extremely challenging due to the small isotope shifts. High-resolution spectroscopy with instruments like UVES on the VLT or HIRES on Keck has been used to constrain isotopic ratios in a handful of barium stars, generally confirming the expected s-process enrichment pattern.

Part 2: Barium and the r-Process

The r-Process Path to the $N = 82$ Region

In a neutron star merger, the r-process path at $N = 82$ runs through nuclei far from stability:

$$\cdots \to {}^{120}\text{Sr}(n,\gamma) {}^{121}\text{Sr}(n,\gamma) \cdots \to {}^{124}\text{Sr} \to \text{waiting point at } N = 82$$

Wait — ${}^{120}\text{Sr}$? Strontium with $N = 82$? That is $A = 120$, $Z = 38$ — a nucleus with 20 fewer protons than stable barium at the same neutron number. This extremely neutron-rich nucleus has $Z/A = 0.317$, far from the stable value of $Z/A \approx 0.41$ at $A \approx 136$.

At the $N = 82$ waiting point, the r-process flow stalls. The neutron separation energy drops sharply at the shell closure (just as in stable nuclei), and photodisintegration $(\gamma,n)$ prevents further captures. The nucleus waits for beta decay:

$${}^{120}\text{Sr} \xrightarrow{\beta^-} {}^{120}\text{Y} \xrightarrow{\beta^-} {}^{120}\text{Zr} \xrightarrow{\beta^-} \cdots$$

Each beta decay increases $Z$ by 1 (at constant $A$), moving the nucleus toward the valley of stability.

After Freeze-Out: The Beta-Decay Cascade

When the neutron flux ceases (freeze-out), the extremely neutron-rich nuclei at $N = 82$ undergo a series of beta decays. Consider the flow at $A = 124$:

$${}^{124}\text{Mo} \xrightarrow{\beta^-} {}^{124}\text{Tc} \xrightarrow{\beta^-} {}^{124}\text{Ru} \xrightarrow{\beta^-} {}^{124}\text{Rh} \xrightarrow{\beta^-} {}^{124}\text{Pd} \to \cdots \to {}^{124}\text{Sn} \text{ (stable)}$$

And at $A = 130$:

$${}^{130}\text{Cd} \xrightarrow{\beta^-} {}^{130}\text{In} \xrightarrow{\beta^-} {}^{130}\text{Sn} \xrightarrow{\beta^-} \cdots \to {}^{130}\text{Te} \text{ (stable)}$$

The accumulation of material at the $N = 82$ waiting point, spread over a range of mass numbers (roughly $A \sim 118$–$132$), produces the second r-process peak centered at $A \approx 130$.

Why the r-Process Peak is at Lower $A$ Than the s-Process Peak

This is the origin of the twin-peak structure:

• s-process peak at $A \approx 138$: Material accumulates at ${}^{138}\text{Ba}$ ($Z = 56$, $N = 82$), which is on the valley of stability. The peak is at $A = Z + N = 56 + 82 = 138$.

• r-process peak at $A \approx 130$: Material accumulates at the $N = 82$ waiting point, but with $Z \sim 38$–$48$ (far from stability). After beta-decay to stability, $A$ is unchanged, so the peak is at $A = Z_{\text{waiting}} + 82 \approx 48 + 82 = 130$.

The $\sim 8$ mass-unit shift between the two peaks is a direct measure of how far from stability the r-process path extends at $N = 82$.

Part 3: Decomposing the Solar Abundance Pattern

The s-Process Contribution

The s-process contribution to the solar abundance pattern can be calculated from first principles using: 1. The network of neutron capture reactions along the valley of stability 2. Measured neutron capture cross sections (from KADoNiS) 3. The neutron exposure distribution in AGB stars (from stellar models)

The s-process calculation is constrained by the s-only isotopes — isotopes that receive no r-process contribution. For each s-only isotope, the product $\sigma_A N_s(A)$ can be measured directly, providing anchor points for the s-process calculation.

The r-Process Residual

Once the s-process contribution $N_s(A)$ is known, the r-process contribution is obtained by subtraction:

$$N_r(A) = N_\odot(A) - N_s(A)$$

This solar r-process residual is the primary observational constraint on r-process models. Its main features:

1. Three peaks at $A \approx 80$ ($N = 50$), $A \approx 130$ ($N = 82$), $A \approx 195$ ($N = 126$)
2. A rare-earth peak at $A \approx 164$
3. An actinide component ($A > 206$) — thorium and uranium
4. An overall decline with increasing $A$, modulated by the peaks

The Universality Test

A remarkable feature of the r-process is the universality of its abundance pattern. When the r-process abundances in extremely metal-poor stars (such as CS 22892-052, with $[\text{Fe/H}] = -3.1$) are compared to the solar r-process residuals, the agreement is striking — especially for elements in the range $56 \leq Z \leq 76$ (barium through osmium).

This universality suggests that the r-process conditions are similar in every event — or at least that the nuclear physics (masses, half-lives) at the waiting points is robust enough to produce a characteristic pattern regardless of the astrophysical details.

However, universality does not extend perfectly to the actinides. Some r-process-enriched stars show an "actinide boost" — thorium and uranium abundances 2–5 times higher than expected from the solar r-process pattern. This suggests that different r-process events may produce different amounts of the heaviest elements, possibly reflecting variations in the neutron-to-seed ratio, entropy, or expansion timescale.

A Worked Example: The Second Peak Decomposition

To illustrate the decomposition quantitatively, consider the isotopes near $A = 130$:

• ${}^{130}\text{Te}$ ($Z = 52$, $N = 78$): solar abundance $N_\odot = 4.32$. This isotope is r-only — it is shielded from the s-process because ${}^{130}\text{Xe}$ ($Z = 54$) is stable, and the s-process path goes through ${}^{130}\text{Xe}$ rather than reaching ${}^{130}\text{Te}$ via beta decay. Therefore $N_r({}^{130}\text{Te}) = 4.32$.

• ${}^{136}\text{Ba}$ ($Z = 56$, $N = 80$): solar abundance $N_\odot = 2.46$. This isotope is s-only — it is shielded from the r-process because ${}^{136}\text{Xe}$ ($Z = 54$) is stable, and the r-process beta-decay chain from $A = 136$ neutron-rich nuclei terminates at ${}^{136}\text{Xe}$, never reaching ${}^{136}\text{Ba}$. Therefore $N_s({}^{136}\text{Ba}) = 2.46$.

• ${}^{138}\text{Ba}$ ($Z = 56$, $N = 82$): solar abundance $N_\odot = 2.52 \times 10^1$ (by far the most abundant barium isotope). This receives both s-process and r-process contributions. Using the s-process model: $N_s({}^{138}\text{Ba}) \approx 2.07 \times 10^1$, giving $N_r({}^{138}\text{Ba}) \approx 4.5$. The s-process dominates ($\sim 82$%).

The pattern is clear: near the s-process peak at $N = 82$ ($A \approx 138$), the s-process contribution dominates. Near the r-process peak ($A \approx 130$), the r-process dominates. Between the peaks, both contribute. This decomposition, repeated across the entire heavy-element region, provides the detailed r-process residual pattern that r-process models must reproduce.

Part 4: A Quantitative Comparison

Part 5: Experimental Nuclear Physics Behind the Processes

Measuring s-Process Cross Sections

The s-process is the better-constrained of the two processes because its path follows the valley of stability, where nuclei can be studied in the laboratory. The key experimental input is the Maxwellian-averaged neutron capture cross section $\langle \sigma v \rangle / v_T$, evaluated at stellar temperatures (typically $kT = 25$–$30\,\text{keV}$ for the main s-process).

These measurements are performed at neutron facilities worldwide: - n_TOF at CERN: A pulsed spallation neutron source with a 185-meter flight path, providing white-spectrum neutrons that cover the full energy range relevant for astrophysics ($1\,\text{eV}$ to $1\,\text{MeV}$). - ORELA at Oak Ridge: A linear electron accelerator producing neutrons via photonuclear reactions on a tantalum target, with excellent energy resolution. - FRANZ at Frankfurt: Specifically designed for stellar neutron capture measurements at $kT = 25\,\text{keV}$.

The experimental technique typically involves irradiating a sample with quasi-Maxwellian neutron spectra (produced by the ${}^{7}\text{Li}(p,n){}^{7}\text{Be}$ reaction at proton energies near threshold) and measuring the capture gamma rays or the activation products. Precisions of $1$–$5$% have been achieved for many s-process nuclei, making the s-process abundance calculation remarkably accurate.

The r-Process Challenge: Measuring Nuclei That Have Never Been Made

The r-process presents a fundamentally different experimental challenge. The nuclei on the r-process path have $10$–$30$ more neutrons than the most stable isotope and have lifetimes of milliseconds to seconds. Most have never been produced in a laboratory.

The strategy is to produce these nuclei at radioactive beam facilities (FRIB, RIKEN RIBF, FAIR) using projectile fragmentation or fission of heavy beams, then measure their properties before they decay: - Masses are measured using Penning traps (precision $\sim 10\,\text{keV}$) or time-of-flight spectrometers - Beta-decay half-lives are measured by implanting ions in segmented silicon detectors and correlating implantation with subsequent decay radiation - Beta-delayed neutron emission probabilities are measured using neutron detectors surrounding the implantation station

FRIB, which achieved first beams in 2022, is expected to produce approximately 80% of the nuclei on the r-process path — a transformative advance that will replace theoretical predictions with experimental data for most of the key r-process nuclei.

Lessons and Key Takeaways

1. The same magic number ($N = 82$) produces two different abundance peaks through two different mechanisms: the s-process bottleneck (small capture cross section at stability) and the r-process waiting point (drop in $S_n$ far from stability followed by beta-decay back to stability).

2. s-only and r-only isotopes are the Rosetta Stone of nucleosynthesis. They allow us to decompose the solar abundance pattern into its constituent processes and constrain the conditions in each.

3. The shift between twin peaks directly measures how far from stability the r-process operates. Larger shifts imply more extreme neutron-rich conditions. Precise measurements of peak positions constrain the r-process neutron density and temperature.

4. Universality of the r-process pattern suggests robust nuclear physics at the waiting points. The magic numbers and the associated nuclear structure properties (shell gaps, beta-decay rates) are the primary determinants of the r-process abundance pattern, not the specific astrophysical conditions — at least for elements with $56 \leq Z \leq 76$.

Discussion Questions

1. If the neutron shell closure at $N = 82$ were weaker in neutron-rich nuclei than in stable nuclei (a "quenched" shell gap), how would the r-process abundance pattern change? Would the r-process peak be broader, narrower, or shifted?

2. The s-process and r-process produce barium with different isotopic compositions. In principle, high-precision isotopic measurements of barium in metal-poor stars could distinguish the two contributions. What observational challenges make this measurement difficult?

3. The p-process isotopes ${}^{130}\text{Ba}$ and ${}^{132}\text{Ba}$ are extremely rare (together $< 0.3$% of natural barium). Why are they so hard to produce, and what does their rarity tell us about the p-process mechanism?

4. If you could measure the beta-decay half-life of one currently unmeasured nucleus to improve r-process abundance predictions, which nucleus would you choose, and why? (Consider the waiting-point nuclei at $N = 50$, $82$, and $126$.)