⚛️ phenomenology

BSFfast: Rapid computation of bound-state effects on annihilation in the early Universe

This paper introduces BSFfast, a lightweight numerical tool that provides precomputed, tabulated effective bound-state formation cross sections for various phenomenological models, enabling efficient and rapid incorporation of complex bound-state effects into Boltzmann solvers for dark matter parameter scans.

Original authors: Tobias Binder, Mathias Garny, Jan Heisig, Stefan Lederer

Published 2026-07-30

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Tobias Binder, Mathias Garny, Jan Heisig, Stefan Lederer

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 the early Universe as a chaotic, super-hot dance floor just after the Big Bang. In this swirling crowd, invisible particles are constantly bumping into each other, sometimes sticking together to form temporary couples, and sometimes bouncing apart. Physicists call these "dark matter" particles, and they are the mysterious stuff that holds galaxies together, even though we can't see them. To figure out how much dark matter exists today, scientists have to calculate how often these particles annihilate (destroy each other) as the Universe cools down. But here's the tricky part: if these particles have a long-range force between them (like a magnetic pull that gets stronger the closer they get), they can get stuck in "bound states"—like dancers spinning in a tight embrace before they finally let go and vanish. These "embrace" moments, especially when the dancers are in excited, high-energy spins, can change the entire outcome of the dance, drastically reducing the number of particles left over.

For a long time, calculating these "embrace" effects was like trying to count every single step of a million dancers in real-time while the music was playing. It was so computationally heavy that scientists had to skip the most interesting parts or run simulations that took forever. This is where the new paper comes in. The authors, a team of physicists from Germany and Poland, have built a clever shortcut called BSFfast. Instead of doing the heavy lifting every time they run a simulation, they pre-calculated the dance steps for a wide variety of scenarios and stored them in a neat, easy-to-read table. They also discovered a mathematical "magic trick" (rescaling) that lets them use one set of calculations to predict results for many different particle masses and force strengths without doing the work twice.

The Big Discovery: A Shortcut for Cosmic Dancers

The main finding of this paper is that you don't need to re-calculate the complex physics of bound-state formation from scratch every time you want to study the early Universe. The team created a lightweight computer tool, BSFfast, that acts like a fast-forward button for these calculations. They pre-computed the "effective annihilation cross section"—a fancy way of saying "how likely these particles are to destroy each other"—for a huge range of temperatures and particle masses.

What makes this tool special is that it doesn't just look at the basic, low-energy "hugs" between particles. It includes the effects of highly excited bound states, where particles are spinning wildly in high-energy orbits (up to a principal quantum number of $n=100$ ). The paper shows that ignoring these excited states is a mistake; they can actually dominate the process, making particles disappear much faster than previously thought. By including these states, the tool reveals that the "freeze-out" process (when particles stop annihilating and stick around) happens differently and earlier than simple models predict.

How the Tool Works: The Magic of Rescaling

The authors realized that the physics of these interactions follows a set of rules that allow for rescaling. Think of it like a recipe for cookies. If you have a recipe for a batch of cookies with a certain amount of sugar and flour, you don't need to bake a whole new batch to see what happens if you double the sugar. You can just mathematically adjust the numbers. Similarly, the team found that if they calculated the results for one specific particle mass and force strength, they could mathematically "stretch" or "shrink" those results to fit almost any other mass or strength they wanted.

This means they only had to do the hard, time-consuming calculations once for a few reference points. For everything else, the tool simply looks up the reference data and applies the scaling math. This makes the tool incredibly fast, allowing scientists to plug it directly into their standard Universe-simulation software (like Boltzmann solvers) and run thousands of scenarios in the time it used to take to run just one.

Testing the Tool: The SuperWIMP Scenario

To prove their tool works, the authors applied it to a specific scenario called the superWIMP model. In this story, a heavy, colored particle (like a heavy version of a quark) freezes out of the early Universe and then slowly decays into dark matter. The team used BSFfast to see how the "embrace" effects of these heavy particles would change the final amount of dark matter.

They found that when they included the excited bound states and the transitions between them, the results changed dramatically. The heavy particles were depleted much more efficiently than in older models. This meant that to get the right amount of dark matter we see today, the mass of the dark matter particle had to be different than previously calculated. Specifically, for a "top-philic" mediator (one that likes to interact with top quarks), the transitions were so efficient that the dark matter mass had to be significantly larger to match observations. They also checked these results against constraints from the Lyman- $\alpha$ forest (a way of looking at ancient gas clouds to see how "warm" or "fast" the dark matter was). They found that the new, faster depletion rates pushed the top-philic model further away from the "forbidden zone" of these constraints, making it a more viable candidate for dark matter.

Safety Checks: When the Math Gets Too Wild

The authors were careful to check if their math was breaking the laws of physics. In quantum mechanics, there's a rule called unitarity that says particles can't interact with a probability greater than 100%. If you calculate a rate that's too high, it means your math has gone off the rails. The team checked their tool against this rule. They found that for particles interacting with Standard Model forces (like the strong force in QCD), their results stayed safely below the limit, even when including those crazy high-energy excited states. However, for "dark" forces where the strength is unknown, they noted that if the force is too strong or the particles are moving too slowly, the math could hit the unitarity wall. To be safe, their tool now includes a warning system: if a user tries to run a simulation in a region where the math might break the rules, the tool flashes a warning.

Why This Matters

Before this paper, studying these complex bound-state effects was like trying to solve a Rubik's cube while running a marathon; it was possible, but it took so much time and effort that many scientists just avoided it. With BSFfast, the Rubik's cube is solved and handed to you on a silver platter. The tool is publicly available and can be used by anyone with a computer to explore new theories about dark matter. It allows researchers to quickly scan through millions of possibilities to see which ones fit our Universe, turning a computationally prohibitive problem into a routine calculation. The authors emphasize that while their tool is a massive step forward, it is a simulation-based tool based on pre-computed tables, not a new law of physics, but it opens the door to exploring the early Universe in ways that were previously too slow to attempt.

Technical Summary of "BSFfast: Rapid computation of bound-state effects on annihilation in the early Universe"

Problem Statement
In the early Universe, the formation of metastable bound states (BSF) for new physics particles interacting via long-range forces can significantly enhance annihilation rates, thereby altering the thermal freeze-out dynamics and the resulting relic abundance of dark matter (DM). Recent studies have highlighted that including excited bound states can drastically reduce DM abundance and qualitatively modify freeze-out scenarios, such as in superWIMP models or conversion-driven freeze-out.

While the effect of these states can be encapsulated in an effective thermally averaged annihilation cross section, $\langle \sigma v \rangle_{\text{eff,BSF}}$ , its explicit computation is numerically prohibitive. Calculating this quantity requires evaluating a vast network of complex scattering-to-bound and bound-to-bound transition rates, followed by the inversion of large transition network matrices. Performing these "on-the-fly" calculations within Boltzmann solvers or during parameter scans is computationally too expensive for practical phenomenological studies.

Methodology
The authors present BSFfast, a lightweight numerical tool designed to bypass the need for repeated, expensive integral computations. The methodology relies on three core components:

Precomputation and Tabulation: The tool provides precomputed, tabulated effective BSF cross sections for a wide class of phenomenologically relevant models. These models include particles charged under SM QCD/QED and dark sector $U(1)$ or $SU(N)$ groups, covering both scalar and fermionic constituents. The calculations explicitly include bound-state formation, dissociation, decays, and the full network of radiative bound-to-bound transitions, accounting for highly excited states up to principal quantum numbers $n \le 100$ .
Rescaling Relations: To efficiently cover models with varying masses ( $m$ $m$ ) and coupling strengths ( $\alpha$ $α$ ) without generating new tables for every parameter point, the authors exploit analytic rescaling properties of the cross section.
- For frozen couplings (e.g., dark QED), the cross section satisfies exact scaling laws: $\langle \sigma v \rangle \propto m^{-2}$ and specific dependencies on $\alpha$ that allow mapping results from a reference parameter set $(m_0, \alpha_0)$ to any other $(m, \alpha)$ via simple rescaling of the temperature parameter $x = m/T$ and multiplicative factors.
- For running couplings (e.g., SM QCD), exact scaling breaks down. However, the authors demonstrate that evaluating the frozen-coupling result at a characteristic potential scale $\mu \sim \sqrt{mT}$ provides a robust approximation (within $\sim 15\%$ ) of the full running-coupling result.
Interpolation and Implementation: The precomputed data is stored in tables covering $x \in [10, 10^6]$ and masses $m \in [100, 10^7]$ GeV. The tool provides fast interpolation routines in Python, C, and Mathematica using linear interpolation in logarithmic space. This allows for seamless integration into standard Boltzmann solvers like micrOMEGAs and MadDM.

Key Contributions and Results

Efficient Tool Development: BSFfast enables the inclusion of highly excited bound states ( $n$ up to 100) and full transition networks in cosmological calculations, a feature previously limited by computational cost.
Model Coverage: The tool covers a broad spectrum of models, including "stop-like" scalar mediators, fermionic mediators, and dark sector particles. It distinguishes between models requiring full 2D interpolation (SM QCD with running coupling) and those amenable to 1D rescaling (dark sectors with frozen couplings).
Unitarity Analysis: The authors perform a detailed analysis of perturbative unitarity violations. They show that for SM QCD models within the perturbative regime ( $\alpha_s \le \alpha_s(1 \text{ GeV})$ ) and the tabulated $x$ range, the effective cross section remains safely below unitarity bounds (by at least an order of magnitude). For dark QCD models with constant couplings, the tool includes a warning system that flags parameter regions where the cross section approaches 10% or 100% of the s-wave unitarity limit.
Phenomenological Application: The tool is applied to a superWIMP scenario involving a colored scalar mediator. The study demonstrates that bound-state effects (specifically transitions among excited levels) can lead to a "super-critical" effective annihilation rate, causing the mediator abundance to continue decreasing well after standard freeze-out. This significantly alters the predicted dark matter mass required to match the observed relic density, with differences of an order of magnitude compared to calculations including only Sommerfeld enhancement. Furthermore, the study shows that electromagnetic transitions in top-philic models lead to more efficient depletion, shifting the parameter space away from Lyman- $\alpha$ constraints compared to bottom-philic models.

Significance
The paper claims that BSFfast fills a critical gap in early-Universe phenomenology by making the inclusion of complex bound-state dynamics computationally feasible for repeated evaluations in Boltzmann solvers. By replacing expensive on-the-fly computations with fast interpolation and rescaling, the tool enables phenomenological studies that were previously "computationally prohibitive." The authors emphasize that the tool is publicly available and designed to be readily integrated into existing frameworks, facilitating more accurate predictions for dark matter production mechanisms, particularly in scenarios involving long-lived mediators and non-Abelian gauge interactions. The work does not propose new experimental signatures but rather provides the necessary computational infrastructure to rigorously test existing theoretical models against cosmological data.

More like this