⚛️ nuclear theory

Astrophysical S-factor Calculation for p-p Fusion Reaction

This study calculates the astrophysical S-factor for the p-p fusion reaction by employing a genetic algorithm-optimized inverse scattering potential and an exact WKB action integral, yielding a value of $(0.1678\pm 0.0058)\times10^{-25}$ that is approximately one order of magnitude lower than currently accepted values.

Original authors: Arushi Sharma, Ishwar Kant, O. S. K. S. Sastri

Published 2026-07-02

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Arushi Sharma, Ishwar Kant, O. S. K. S. Sastri

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine two tiny, positively charged balls (protons) trying to hug each other to form a new particle (a deuteron). This is the very first step in how the Sun shines. But here's the problem: because they are both positively charged, they repel each other fiercely, like two strong magnets with the same pole facing in. To get close enough to "hug," they have to squeeze through a massive, invisible wall of force called the Coulomb barrier.

In the cold, dark vacuum of space, these protons don't have enough energy to climb over this wall. Instead, they rely on a weird quantum trick called tunneling, where they essentially "ghost" through the wall. The paper you provided is about calculating exactly how likely this ghosting is to happen.

Here is a breakdown of what the authors did, using simple analogies:

1. The Old Map vs. The New Map

For decades, scientists calculated the likelihood of this tunneling using a simplified map. They assumed the "wall" the protons had to cross was made of pure, unchanging electricity (the "bare Coulomb potential").

The Analogy: Imagine trying to calculate how hard it is to walk through a foggy forest. The old method assumed the fog was a solid, uniform wall that never changed thickness.
The Problem: In reality, the "fog" isn't uniform. It gets thinner and behaves differently depending on how close the protons get to each other. The old map ignored the actual shape of the forest and just drew a straight, thick wall.

2. The New Method: Reverse Engineering the Forest

The authors in this paper decided to stop guessing the shape of the wall. Instead, they used a technique called Inverse Scattering.

The Analogy: Imagine you can't see the forest, but you can hear how sound bounces off the trees (scattering data). Instead of guessing what the trees look like, you use a super-smart computer (a Genetic Algorithm, which works like natural selection) to "evolve" a model of the forest until the sound bounces off it exactly the way it does in real life.
The Result: They built a "Reference Potential" that acts like a custom-made, 3D model of the force field between the protons. This model naturally includes the repulsion, the attraction, and the "screening" effects without needing to force them in with math formulas.

3. The Tunneling Calculation: A More Accurate Tunnel

Once they had this realistic map of the force field, they calculated the tunneling probability using a method called WKB.

The Old Way: They used the "Sommerfeld factor," which is like saying, "The wall is 100 feet high, so the chance of getting through is X."
The New Way: They calculated the tunneling by looking at the actual shape of the wall at every single point. They found that because their new map showed the wall fading away sooner than the old map thought, the "tunnel" was actually much harder to get through than previously believed.

4. The Big Surprise: The Sun Might Be Hotter

When they plugged their new, more realistic numbers into the equation, they got a result that shocked them.

The Finding: Their calculated value for the "S-factor" (a number that tells us how easily the fusion happens) was about 10 times smaller than the value everyone else has been using for years.
The Implication: If the protons are having a much harder time fusing than we thought, then for the Sun to produce the amount of light and heat we see today, the core of the Sun must be much hotter than current models suggest. It's like realizing a campfire is burning much brighter than expected, which means the wood inside must be burning at a much higher temperature.

5. The Smart Extrapolation: Using AI instead of a Ruler

To get the final answer, scientists usually have to guess what happens at "zero energy" (the coldest possible state) because they can't measure it directly. They usually do this by drawing a smooth curve through their data points, like connecting dots with a ruler (polynomial fitting).

The Innovation: The authors didn't use a ruler. They used a Neural Network (a type of simple AI).
The Analogy: Instead of forcing the dots to fit a straight or curved line, the AI "learned" the pattern of the dots. It looked at the data from high energy down to extremely low energy and figured out the natural shape of the curve without forcing it into a box. This gave them a more trustworthy answer for the "zero energy" point.

Summary of the Claim

The paper claims that by building a more realistic model of the force between protons and using AI to analyze the data, they found that the Sun's primary fusion reaction is much less efficient than previously thought. Consequently, to keep the Sun shining as it does, the Sun's core temperature might need to be revised upward significantly.

Important Note: The authors explicitly state that their result is about the physics of the Sun's core and the nuclear reaction rates. They do not claim this changes how we build fusion reactors on Earth or that it has immediate clinical applications. It is a fundamental correction to our understanding of how stars work.

Technical Summary: Astrophysical S-factor Calculation for p-p Fusion Reaction

Problem Statement
The proton-proton (p-p) fusion reaction ( $p + p \rightarrow d + e^+ + \nu_e$ ) initiates the proton-proton chain, driving hydrogen burning in main-sequence stars. At the low thermal energies prevailing in stellar interiors, this reaction proceeds via quantum mechanical tunneling through the Coulomb barrier. The reaction rate is conventionally expressed via the astrophysical S-factor, $S(E)$ , which isolates the smooth nuclear contribution from the dominant energy dependence caused by Coulomb suppression.

Current high-precision calculations of the zero-energy S-factor, $S(0)$ , typically rely on a hybrid approach: they utilize realistic nucleon-nucleon interaction potentials (e.g., Argonne v18) to generate scattering and bound-state wavefunctions, but evaluate the barrier penetration probability (Gamow factor) using the "bare" Coulomb potential ( $V_C = e^2/r$ ) within the WKB integral. The authors argue that this approach is inconsistent. The bare Coulomb potential does not vanish asymptotically, leading to numerical instabilities at extremely low energies and failing to account for screening effects or the finite range of the actual interaction. Consequently, the Gamow factor may be overestimated, and the resulting $S(0)$ values (currently accepted around $4.0 \times 10^{-25}$ MeV b) may be inaccurate. Furthermore, the extrapolation of $S(E)$ to zero energy is traditionally performed using polynomial fitting, which introduces approximation errors and artificial curvature.

Methodology
The authors propose a unified, self-consistent framework that avoids the bare Coulomb approximation and employs advanced optimization and machine learning techniques:

Inverse Scattering and Reference Potential Approach:
Instead of assuming a specific analytical form for the nuclear and Coulomb interactions, the authors construct the interaction potential inversely from scattering phase shifts. They employ a "Reference Potential" consisting of three smoothly joined Morse functions.
- For the neutron-proton (np) system, the potential is purely attractive at large distances.
- For the proton-proton (pp) system, the third Morse component is inverted to naturally reproduce the long-range repulsive behavior (screened Coulomb interaction) without explicitly adding a $1/r$ term.
- The potential parameters are optimized using a Genetic Algorithm (GA) to minimize the Mean Squared Error (MSE) between calculated and expected phase shifts.
Leave-One-Out Cross-Validation (LOOCV):
To quantify uncertainties and parameter sensitivity, the authors perform LOOCV. By iteratively removing one energy data point and re-optimizing, they generate an ensemble of 11 distinct potential sets. This allows for the statistical propagation of uncertainties from the input data to the final observables.
Wavefunction Construction:
- Scattering States: The proton-proton scattering wavefunctions are generated using the Phase Function Method (PFM), which transforms the Schrödinger equation into a first-order nonlinear differential equation. This ensures consistency between the optimized potential and the scattering solution, eliminating the need for external Coulomb wavefunctions.
- Bound States: The deuteron wavefunction is calculated by solving the coupled-channel ( $^3S_1 - ^3D_1$ ) Schrödinger equations using the optimized np potentials.
S-Factor Calculation and Extrapolation:
- The astrophysical S-factor is calculated using the full effective interaction potential in the WKB action integral for the tunneling probability, rather than the Sommerfeld factor derived from the bare Coulomb potential.
- Calculations are performed at energies as low as $0.0001$ MeV.
- Neural Network Extrapolation: Instead of polynomial fitting, a supervised neural network is trained on the calculated $S(E)$ data points (from 0.2 MeV down to 0.0001 MeV) to extrapolate the value of $S(0)$ . This approach captures the intrinsic non-linear energy dependence without assuming a fixed functional form.

Key Results

Gamow Factor and Turning Points: The use of the reference potential, which decays to zero at finite distances (~34 fm), allows for the reliable determination of classical turning points and WKB integrals at energies where the bare Coulomb potential fails (e.g., below 0.01 MeV). The authors find that the WKB action integral evaluated with the full potential differs significantly from the Sommerfeld approximation ( $e^{-2\pi\eta}$ ), particularly at low energies.
Gamow Peak Energy: The calculated Gamow peak energy using the full potential is $0.0094 \pm 0.0046$ keV, significantly lower than the literature value of ~5–6 keV derived from bare Coulomb approximations. The authors note that matching the literature peak energy would require a solar core temperature an order of magnitude higher than currently assumed.
Astrophysical S-Factor ( $S(0)$ ):
- Using the neural network extrapolation on the ensemble of potentials, the authors obtain:
  $S(0) = (0.1678 \pm 0.0058) \times 10^{-25} \text{ MeV b}$
- This value is nearly one order of magnitude lower than the currently accepted standard value of $S(0) \approx (4.01 \pm 0.04) \times 10^{-25}$ MeV b.
- The discrepancy is attributed to the overestimation of the tunneling probability in conventional methods that rely on the bare Coulomb approximation.

Significance and Claims
The paper claims that the conventional separation of the problem into "realistic wavefunctions" and "bare Coulomb tunneling" is physically inconsistent. By treating the interaction potential as a unified, finite-range entity that naturally incorporates screening effects, the authors provide a new estimate for the p-p fusion S-factor.

The authors assert that their result suggests:

Previous literature values for $S(0)$ may be overestimated due to the use of simplified Coulomb approximations.
The lower $S(0)$ value implies that the actual temperature of the solar core could be higher than values assumed in standard solar models to sustain the observed luminosity.
The methodology offers a robust alternative to polynomial extrapolation, utilizing neural networks to handle the non-linear behavior of the S-factor at low energies.

The work concludes that a fully consistent treatment of the interaction potential, overlap integral, and tunneling dynamics is essential for reliable estimates of stellar fusion rates. The authors suggest this methodology could be extended to other astrophysical reactions, such as deuteron-deuteron capture.