ALPHANSO: Open-Source Modeling of ($\alpha$,n) Neutron Source Terms

This paper introduces ALPHANSO, an open-source Python package that modernizes ($\alpha$,n) neutron source term calculations by incorporating updated nuclear data and offering a modular framework that matches or exceeds the accuracy of legacy tools like SOURCES-4C.

The Gist

Imagine you are trying to predict exactly how many sparks will fly off a piece of metal when you hit it with a hammer. In the world of nuclear physics, this "hammer" is an alpha particle (a tiny, fast-moving piece of a radioactive atom), and the "sparks" are neutrons.

Scientists need to know exactly how many sparks fly and how fast they go for many reasons: to keep nuclear power plants safe, to manage radioactive waste, and even to hunt for Dark Matter (the invisible stuff that makes up most of the universe). If you can't predict the sparks accurately, your experiments might get "noisy" and fail, or your safety calculations could be wrong.

For decades, scientists have used a tool called SOURCES-4C to do this math. Think of SOURCES-4C as a vintage, 1980s calculator. It works, but it has some big problems:

• It's outdated: The data inside it is from the 1980s. It's like trying to navigate a modern city with a map from 40 years ago; you'll miss new roads and buildings.
• It's fragile: It's written in an old computer language (FORTRAN 77) that is hard for modern programmers to fix or update.
• It's blind: It can't see certain types of atoms (like Nitrogen or Potassium) and it stops working if the "hammer" hits too hard (energies above 6.5 MeV).

Enter ALPHANSO: The Modern Smartphone

The authors of this paper built a new tool called ALPHANSO. If SOURCES-4C is that old calculator, ALPHANSO is a brand-new smartphone.

Here is why ALPHANSO is a game-changer, explained through simple analogies:

1. The Library of Knowledge (Data)

Imagine SOURCES-4C is a library that only has books printed before 1985. If you need information about a new discovery, the librarian just says, "I don't have that."
ALPHANSO, however, is a library that connects to the internet. It uses the latest, most accurate scientific books (called nuclear data libraries like JENDL, TENDL, and ENDF) that were written just recently. It also has a special feature: if a new book is published tomorrow, scientists can easily add it to the library without rebuilding the whole building.

2. The Recipe (The Math)

Both tools use the same basic recipe to calculate the sparks: they figure out how fast the alpha particle slows down as it travels through the material and how likely it is to hit a nucleus.

• SOURCES-4C uses an old, rough estimate for how fast things slow down.
• ALPHANSO uses a high-precision GPS (modern data) to track exactly how the particle loses energy.
  The paper found that the "GPS" matters less than the "Map" (the cross-section data). If your map is wrong, even the best GPS won't help. ALPHANSO uses the best maps available today.

3. The Open-Source Advantage

SOURCES-4C is like a locked safe. Only a few people have the key, and if the safe breaks, no one can fix it because the blueprints are lost.
ALPHANSO is like a LEGO set that is open to everyone. It is "open-source," meaning the code is free for anyone to see, use, and improve. If a scientist in Japan finds a better way to calculate a specific atom's behavior, they can snap that new piece into ALPHANSO, and everyone benefits immediately.

The Results: Does it Work?

The authors tested ALPHANSO against real-world experiments (actual sparks in a lab) and compared it to the old SOURCES-4C.

• Accuracy: ALPHANSO predicted the number and speed of sparks just as well as, or better than, the old tool.
• Range: It can handle "harder hits" (higher energy) that the old tool simply couldn't calculate.
• Versatility: It can calculate sparks for atoms that the old tool didn't even know existed.

The Bottom Line

The paper concludes that ALPHANSO is the future. It replaces the clunky, outdated 1980s tool with a modern, flexible, and transparent system.

The Analogy:
If you were building a house, you wouldn't use a hammer from 1980 that only works on soft wood and breaks if you hit a nail too hard. You would use a modern power tool that is precise, works on any material, and comes with a manual that anyone can read and update. ALPHANSO is that modern power tool for nuclear physics.

This is a big deal for scientists hunting Dark Matter or ensuring nuclear safety, because now they have a reliable, up-to-date way to predict the invisible "noise" of neutrons, allowing them to see the universe more clearly.

Technical Summary

1. Problem Statement

Accurate prediction of neutron fields generated by $(\alpha, n)$ reactions is critical for applications ranging from nuclear safeguards and waste management to dark matter detection and criticality safety. While deterministic codes like SOURCES-4C have been the industry standard for decades, they suffer from significant limitations:

• Outdated Data: Reliance on nuclear data libraries that have not been updated since the mid-1980s.
• Format Incompatibility: Data files do not conform to modern standards like the Generalized Nuclear Data Structure (GNDS), hindering updates.
• Limited Nuclide Coverage: Inability to calculate source terms for many naturally occurring target nuclides (e.g., $^6$Li, $^{14}$N, $^{35}$Cl, $^{41}$K).
• Energy Constraints: Restricted to $\alpha$-particle energies below 6.5 MeV, which excludes high-energy emissions from natural decay chains critical for low-background experiments.
• Legacy Architecture: Written in FORTRAN 77, lacking modern maintainability and transparency.

Existing alternatives like NeuCBOT (Python-based but reliant on TALYS, leading to lower accuracy for key materials) or NEDIS (high accuracy but not publicly available) fail to provide a complete, accessible, and accurate solution.

2. Methodology

The authors present ALPHANSO (Alpha Neutron Sources), a modern, open-source Python package designed to calculate $(\alpha, n)$ neutron source terms.

• Physics Model: ALPHANSO employs the same thick-target, continuous slowing-down approximation as SOURCES. It assumes targets are infinitely thick (all $\alpha$-particles stop within the material).
  • Neutron Yield: Calculated by integrating the reaction probability over the slowing-down path (Eq. 1), utilizing the Bragg-Kleeman relationship for material stopping power (Eq. 2).
  • Neutron Spectra: Constructed using two-body kinematics, assuming isotropic emission in the center-of-mass frame. The code accounts for branching ratios to various residual nuclear excitation states to determine the effective Q-value and kinematic bounds for neutron energy (Eqs. 4–9).
• Data Integration:
  • Cross Sections: ALPHANSO utilizes modern evaluated nuclear data libraries: ENDF/B-VIII.1, JENDL-5, and TENDL-2023. It also incorporates specific re-evaluations by Parvu et al. for light elements (C, Mg, Al, Si) to correct discrepancies in JENDL-5.
  • Selection Logic: The code automatically selects the most accurate data source for each nuclide (e.g., ENDF/B-VIII.1 for $^6$Li, $^9$Be; Parvu et al. for light elements; JENDL-5 where available; TENDL-2023 as a fallback).
  • Stopping Power: Uses the ASTAR library where available, falling back to SRIM for elements with $Z \le 92$, and Perry/Wilson coefficients for heavier elements.
• Architecture: Built in Python with a modular design, allowing users to easily swap data libraries, add custom evaluations, and extend functionality. It supports the GNDS format for future-proofing.

3. Key Contributions

• Open-Source Modernization: Provides a fully accessible, cross-platform Python replacement for legacy FORTRAN codes.
• Comprehensive Nuclide Coverage: Supports all naturally occurring target nuclides, including those missing in SOURCES-4C (e.g., $^{21}$Ne, $^{33}$S, $^{35}$Cl).
• Extended Energy Range: Capable of handling $\alpha$-particle energies significantly higher than the 6.5 MeV limit of SOURCES-4C, addressing needs for natural decay chain analysis.
• Data Transparency & Flexibility: Users can directly evaluate the impact of different nuclear data evaluations (e.g., comparing JENDL vs. TENDL) via a built-in validation suite.
• GNDS Compatibility: Adopts modern data structures, facilitating the integration of future nuclear data releases.

4. Results

The authors validated ALPHANSO against experimental data (West and Sherwood; [37]) and compared it against SOURCES-4C and an improved SOURCES-4A.

• Monoenergetic Beam Yields:
  • ALPHANSO demonstrated high accuracy, with Mean Absolute Percentage Errors (MAPE) generally lower than SOURCES-4C.
  • For Beryllium (Be), ALPHANSO achieved a 1.10% error compared to 19.68% for SOURCES-4C.
  • For Iron (Fe) and M316 Stainless Steel, SOURCES-4C produced no results (missing data), while ALPHANSO provided predictions (though with higher error due to reliance on TENDL data lacking resonance structure).
• Neutron Emission Spectra:
  • ALPHANSO reproduced experimental spectra with accuracy comparable to or exceeding SOURCES-4A and significantly better than SOURCES-4C.
  • Errors were primarily attributed to the use of TENDL-2023 data for specific nuclides, which lacks fine resonance structure compared to JENDL-5 or ENDF/B-VIII.1.
• Homogeneous Mixtures (Uranium Decay Series):
  • In secular equilibrium scenarios for $^{235}$U and $^{238}$U, ALPHANSO achieved a Mean Absolute Error (MAE) of ~0.06, whereas SOURCES-4C systematically underestimated yields with an MAE of ~0.5.
• Sensitivity Analysis:
  • The study confirmed that differences in calculated yields and spectra are driven almost entirely by $(\alpha, n)$ cross-section data, not stopping power. Stopping power variations between ASTAR, SRIM, and SOURCES had negligible impact on integral quantities.

5. Significance

ALPHANSO represents a critical advancement in nuclear modeling tools:

• Reliability: It offers a validated, high-accuracy alternative to legacy codes, essential for low-background experiments (e.g., dark matter searches) where precise neutron background estimation is vital.
• Future-Proofing: Its modular, open-source nature allows the community to rapidly incorporate new nuclear data evaluations as they are released, ensuring the tool remains current.
• Accessibility: By moving from proprietary, obsolete FORTRAN code to a transparent Python framework, ALPHANSO lowers the barrier to entry for researchers and engineers, fostering reproducibility and collaboration in nuclear safeguards and safety analysis.

In conclusion, ALPHANSO successfully addresses the data obsolescence and architectural rigidity of SOURCES-4C, establishing itself as the preferred modern tool for $(\alpha, n)$ source term calculations.