Two-nucleon scattering amplitudes at light quark masses from lattice QCD

arXiv:2610.23332 · hep-lat, hep-ph, nucl-th · Submitted 2026-10-16 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: High Energy Physics: Lattice. The day in lattice field theory: what the simulations put on the table.

Kai: We're looking at the paper: "Two-nucleon scattering amplitudes at light quark masses from lattice QCD".

Mira: Here's the gist of the paper. A lattice-QCD determination of nucleon-nucleon scattering amplitudes was performed at pion masses around M pi about 200 MeV and M pi about 280 MeV, utilizing a single lattice spacing a about 0.064 fm across two CLS ensembles.

Kai: The finite-volume energy spectra were extracted from large Hermitian correlation matrices constructed within the distillation framework, employing an extensive basis of two-nucleon interpolating operators.

Mira: To reliably map these finite-volume spectra to infinite-volume scattering amplitudes, two formalisms—the plane-wave quantization condition and the relativistic left-hand cut formalism—were employed, explicitly accounting for the long-range one-pion exchange interaction.

Kai: The analysis investigated the quark mass dependence of nucleon-nucleon scattering lengths, effective ranges, shape parameters, and pole positions.

Title and authors: Kai: So we're looking at this paper, "Two-nucleon scattering amplitudes at light quark masses from lattice QCD." It’s basically using lattice quantum chromodynamics to figure out how nucleons interact when they have light quarks, specifically around pion masses of two hundred and eight hundred MeV.

Mira: Right. The authors are Jorge Baeza-Ballesteros and his team, and they're using a single lattice spacing of zero point zero six four femtometers across two different ensembles to get their results. It’s about the nucleon-nucleon scattering amplitudes as you can calculate them from first principles on the lattice.

Lev: From my side, I see they are using a distillation framework here, which means they're building large Hermitian correlation matrices with an extensive basis of two-nucleon interpolating operators to get those finite-volume energy spectra. That’s how they try to control excited state contamination and isolate the lowest energy levels.

Kai: Exactly. And then they use two formalisms—the plane-wave quantization condition and the relativistic left hand cut formalism—to map those finite volume spectra onto infinite volume scattering amplitudes, making sure they account for that long range one pion exchange interaction we need to keep in mind.

Mira: What’s really interesting is what they find regarding the quark mass dependence of things like scattering lengths, effective ranges, and shape parameters across all those pion masses they tested. They favored virtual states in both isospin channels by more than one sigma across the board.

Lev: That suggests that even with these light quark masses, we're looking at something that behaves more like a virtual state in these systems. If you're trying to build real hardware, seeing this behavior at light masses means the effective interaction is quite different than what you might see if you just used heavy quark masses.

Kai: And they compared their results to earlier work done at a much heavier pion mass, around seven hundred and ten MeV. They found that inverse scattering lengths and binding momenta approach their physical point phenomenological values in a monotonic way as the pion mass gets lower.

Mira: That’s important because it shows a clear path toward the real world, where we want those light quark masses to be relevant to our observations of nuclear matter. The paper "Two-nucleon scattering amplitudes at light quark masses from lattice QCD" is showing how these theoretical tools can bridge the gap between our simulations and what we actually measure.

Title and authors: Lev: But it’s worth remembering that they are explicitly accounting for the one pion exchange interaction in both formalisms, which is a big deal when you're trying to get consistent results where you need to go beyond simple approximations.

Kai: And they did some specific checks on the lowest partial waves, looking at the S wave for the dineutron system and the alpha wave for the deuteron system using Blatt-Biedenharn decomposition. The central finding there was that those poles reside on the second Riemann sheet, which corresponds to virtual states.

Mira: That confirms what they saw earlier about virtual states in both isospin channels across all their pion mass tests. As you lower the pion mass, the pole in the deuteron channel moves toward threshold, which is consistent with its eventual transition into a physical bound state on the first Riemann sheet later on.

Lev: So for someone building error correction codes, seeing those poles on the second sheet tells you that if you try to get a bound state out of this lattice setup directly, it's going to be tricky because the interaction isn't strong enough in these light quark environments.

Kai: The strongest dependence they saw was actually in inverse scattering lengths and pole positions as they lowered the pion mass, while things like effective ranges and shape parameters were already pretty consistent with their physical point values even at higher masses.

Mira: That mild dependence on the shape parameters is reassuring because it suggests that some of the other aspects of how these systems behave are less sensitive to the specific light quark mass we're probing right now.

Lev: When I think about running this on real hardware, seeing that kind of scaling behavior in inverse scattering lengths tells you what kind of systematic error you’ll have when extrapolating those lattice results down to the physical point.

Kai: The left hand cut analysis they used was sensitive to the pion nucleon coupling constant, g pi NN, and it gave a value that is qualitatively consistent with what we expect from phenomenological models, though they did admit there was some model dependence in that extraction.

Mira: That's a fair caveat. They are using the relativistic left hand cut formalism which is sensitive to that coupling constant directly, so they can give us an idea of how strong the underlying interaction is without relying on a specific potential model input.

Title and authors: Lev: But we have to remember that their extraction of g pi NN wasn't perfect; it’s still model dependent, which means if you want to get a tight constraint on the coupling constant, you'd need to look at those other analyses they mentioned.

Kai: They did average four different determinations of each observable—two datasets times two formalisms—using equations six three through six five to get their final numbers and then added statistical and systematic uncertainties in quadrature.

Mira: And when you look at the conclusion, it consistently favored a virtual state pole in the corresponding nucleon-nucleon scattering amplitude. As they lowered the pion mass, the threshold intercept of p cot delta decreased in both isospin channels, following those expected trends we see theoretically.

Lev: So what’s the big picture for us? It tells us that at these light quark masses, we are firmly in the virtual state regime for these systems, and it sets a baseline for how to look for bound states when we try to move toward physical quark masses.

Kai: The paper "Two-nucleon scattering amplitudes at light quark masses from lattice QCD" suggests two crucial things moving forward. First, they need to perform calculations at multiple lattice spacings to nail down the continuum limit and second, they should conduct simulations directly at the physical point instead of inferring the pion mass evolution.

Mira: That's a solid roadmap for future work then. They’re essentially saying that to get the most robust picture of these interactions, we need to do both things: clean up those lattice spacing artifacts and get as close as possible to the real physical quark masses.

Lev: From an engineering standpoint, doing simulations at the physical point is always ideal, but it comes with huge computational overhead. So they’re telling us that maybe they can get a good sense of the physics now by extrapolating from these light masses while focusing on controlling those systematic errors in the extrapolation.

Kai: So to wrap up, this paper gives us a clear picture of where we are right now, showing that virtual states dominate across all tested light quark masses and pointing toward the necessary steps for future lattice QCD calculations.

Mira: It’s a useful step in understanding how nuclear forces emerge from the fundamental theory of quarks, even if it doesn't give us a bound state right now.

Lev: Yeah, it’s a good benchmark to see what kind of signal we are looking for when we finally get the computational power to simulate those lighter pion masses directly.

The paper's summary: Kai: So this paper is about using lattice quantum chromodynamics to figure out how nucleons interact when we use light quark masses, specifically around two hundred and eight hundred megaelectron volts for the pion mass.

Mira: Right, so they’re testing the nucleon-nucleon scattering amplitudes at those specific pion masses, looking for patterns in how that interaction changes.

Lev: The methodology they used involves building large Hermitian correlation matrices with a lot of different two-nucleon interpolating operators to get the finite volume energy spectra. That’s how you isolate what you want from all the noise.

Kai: And then they use two different ways—the plane wave quantization condition and the relativistic left hand cut formalism—to translate those finite volume measurements into infinite volume scattering amplitudes, making sure they account for that long range one pion exchange interaction.

Mira: What's really interesting is how they look at the quark mass dependence of things like scattering lengths, effective ranges, shape parameters, and pole positions across all those pion masses they tested.

Lev: Across all those masses considered, the analysis favored virtual states in both isospin channels by more than one sigma. That suggests that even with these light quark masses we're probing now, we're looking at something that behaves more like a virtual state in these systems.

Kai: They compared their results to earlier work done at a much heavier pion mass, around seven hundred and ten megaelectron volts. And they found that inverse scattering lengths and binding momenta approach their physical point phenomenological values in a monotonic way as the pion mass gets lower.

Mira: That’s important because it shows a clear path toward the real world, where we want those light quark masses to be relevant to our observations of nuclear matter. The paper is showing how these theoretical tools can bridge the gap between our simulations and what we actually measure.

Lev: But they did some specific checks on the lowest partial waves, looking at the S wave for the dineutron system and the alpha wave for the deuteron system using Blatt-Biedenharn decomposition. The central finding there was that those poles reside on the second Riemann sheet, which corresponds to virtual states.

The paper's summary: Kai: That confirms what they saw earlier about virtual states in both isospin channels across all their pion mass tests. As the pion mass gets lower, the pole in the deuteron channel moves toward threshold, which is consistent with its eventual transition into a physical bound state on the first Riemann sheet later on.

Mira: So for someone building error correction codes, seeing those poles on the second sheet tells you that if you try to get a bound state out of this lattice setup directly, it's going to be tricky because the interaction isn't strong enough in these light quark environments.

Lev: The strongest dependence they saw was actually in inverse scattering lengths and pole positions as they lowered the pion mass, while things like effective ranges and shape parameters were already pretty consistent with their physical point values even at higher masses.

Kai: The left hand cut analysis they used was sensitive to the pion nucleon coupling constant, g pi NN, and it gave a value that is qualitatively consistent with what we expect from phenomenological models, though they did admit there was some model dependence in that extraction.

Mira: That's a fair caveat. They are using the relativistic left hand cut formalism which is sensitive to that coupling constant directly, so they can give us an idea of how strong the underlying interaction is without relying on a specific potential model input.

Lev: But their extraction of g pi NN wasn't perfect; it’s still model dependent, which means if you want to get a tight constraint on the coupling constant, you'd need to look at those other analyses they mentioned.

Kai: They did average four different determinations of each observable—two datasets times two formalisms—using equations six three through six five to get their final numbers and then added statistical and systematic uncertainties in quadrature.

The paper's summary: Mira: And when you look at the conclusion, it consistently favored a virtual state pole in the corresponding nucleon-nucleon scattering amplitude. As they lowered the pion mass, the threshold intercept of p cot delta decreased in both isospin channels, following those expected trends we see theoretically.

Lev: So what’s the big picture for us? It tells us that at these light quark masses, we are firmly in the virtual state regime for these systems, and it sets a baseline for how to look for bound states when we try to move toward physical quark masses.

Kai: The paper suggests two crucial things moving forward. First, they need to perform calculations at multiple lattice spacings to nail down the continuum limit and second, they should conduct simulations directly at the physical point instead of inferring the pion mass evolution.

Mira: That's a solid roadmap for future work then. They’re essentially saying that to get the most robust picture of these interactions, we need to do both things: clean up those lattice spacing artifacts and get as close as possible to the real physical quark masses.

Lev: From an engineering standpoint, doing simulations at the physical point is always ideal, but it comes with huge computational overhead. So they’re telling us that maybe they can get a good sense of the physics now by extrapolating from these light masses while focusing on controlling those systematic errors in the extrapolation.

Kai: So to wrap up, this paper gives us a clear picture of where we are right now, showing that virtual states dominate across all tested light quark masses and pointing toward the necessary steps for future lattice quantum chromodynamics calculations.

Mira: It’s a useful step in understanding how nuclear forces emerge from the fundamental theory of quarks, even if it doesn't give us a bound state right now.

Lev: Yeah, it’s a good benchmark to see what kind of signal we are looking for when we finally get the computational power to simulate those lighter pion masses directly.

Kai: So next up, we’re going to talk about how they improved their uncertainty estimation using a parametric bootstrap method and what that means for getting tighter error bars on these results.

The paper's improvements: Tom: So we're looking at how the authors are suggesting ways to make their measurements more reliable and less prone to error.

Kai: They talk about improving their uncertainty estimation by implementing a parametric bootstrap that uses posterior variance multiplied by a widening factor, which they observe reduces outlier results in Figure seventeen when compared to linear error propagation from boot0.

Mira: That sounds like they’re trying to get better control over the statistical noise in those measurements, but you have to watch how that widening factor interacts with the actual data points.

Lev: From my side, I see that's a way to handle those outliers more intelligently than just using linear error propagation from boot0. That sounds like something you’d need when you’re building hardware where you expect some noise spikes.

Kai: And they also have this new way to assess systematic uncertainty by performing model and data-subset averaging based on the Akaike Information Criterion, which allows for a controlled assessment of systematic variations across different analysis variants.

Mira: That’s smart because it lets them weigh which analysis setup is actually giving them the most reliable information, instead of just picking one way to do everything.

Lev: It takes more work to set up those different analysis variants, but if you can use the Akaike Information Criterion to guide your choice, that’s a solid way to manage that complexity.

Kai: Then they have this new capability where they can directly constrain the one pion exchange coupling constant g pi NN by using the relativistic left hand cut formalism, because it’s sensitive to that quantity and its extracted value is qualitatively consistent with the expected phenomenological value.

Mira: That’s a big deal for theorists because if you can get a constraint on that coupling constant from lattice QCD without relying on a specific potential model input, it tells us something fundamental about how the quarks interact.

Lev: I’m interested in that part because constraining g pi NN directly is what we need to move beyond just looking at scattering lengths and getting closer to the underlying interaction strength.

Kai: They also have this new way to distinguish between virtual and bound state poles by analytically continuing constrained amplitudes, which allows for a determination of pole structure on the second Riemann sheet.

Mira: That’s useful because it gives them a way to be more certain about what those poles actually represent in terms of physics, rather than just seeing where they fall on the mathematical sheets.

Lev: It's hard work to do that analytical continuation, but if you can pin down the pole structure reliably, that helps you predict what kind of physical state might actually show up in a real system.

Kai: So they’re showing them how to refine their error bars and constrain fundamental parameters like g pi NN, and they're also giving them tools to better interpret the mathematical structure of the results.

Mira: It sounds like they are building a much more robust framework for interpreting these lattice QCD calculations.

Lev: It means when you eventually get the computational power to run simulations at physical pion masses, you won’t be guessing too much about the systematic errors in your extrapolation.

Kai: So what we're hearing is that the authors are taking their results and applying more sophisticated statistical tools to make sure those results actually mean something concrete for our understanding of nuclear forces.

Mira: It sounds like they're setting a really high bar for how we interpret these kinds of lattice QCD calculations moving forward.

Lev: And that sets a good benchmark, right? So next up, we’re going to talk about the implications of these findings for real-world nuclear physics and what it means if they can nail down those coupling constants.

Conclusion: Kai: So to wrap up, we saw that the paper, "Two-nucleon scattering amplitudes at light quark masses from lattice QCD," consistently favored virtual state poles across all their tested pion masses.

Mira: That means for systems with light quarks, like nucleons interacting, we are definitely in a virtual state regime right now.

Lev: It sets a baseline for us when we eventually try to simulate things at the physical point, showing us what kind of signal we should expect to look for.

Kai: The authors pointed out that they need to do more calculations at different lattice spacings to get the continuum limit and also really push toward simulations at the physical point instead of just inferring the evolution from these light masses.

Mira: It’s a solid roadmap then. They’re saying we have a clear path forward, but it requires both better control over those lattice artifacts and more computational power for real physics.

Lev: I agree with that extrapolation idea, but the cost of running simulations at physical pion masses is huge, so they're right to suggest using these light masses as a stepping stone while managing those systematic errors in the way we extrapolate.

Kai: So essentially, this paper gives us a very clear picture of where we are today on the lattice and what steps are needed to get closer to real physics.

Mira: It’s a useful step in understanding how nuclear forces emerge from the fundamental theory of quarks, even if it doesn't give us a stable bound state right now.

Lev: Yeah, it’s a good benchmark to see what kind of signal we are looking for when we finally get the computational power to simulate those lighter pion masses directly.

Kai: We'll keep an eye on those continuum limit studies and hope the next set of simulations gets us closer to that physical point.

Mira: Definitely, because seeing how these amplitudes evolve with quark mass is crucial for our theoretical models.

Lev: And I’ll be thinking about how they might try to incorporate these coupling constant constraints into error analysis for future work.

John von Neumann-Institut für Computing NIC · Fakultńt fuer Physik und Astronomie Institut für Theoretische Physik II Ruhr-Universitä Bochum · Department of Physics University of California Berkeley · Nuclear Science Division Lawrence Berkeley National Laboratory · Institute for Theoretical Physics Albert Einstein Center for Fundamental Physics University of Bern · Department of Physics Kent State University · Department of Physics Carnegie Mellon University Department of Physics and Astronomy University of North Carolina Chapel Hill Institute für Kernphysik Technische Universitat Darmstadt Helmholtz-Institut Mainz Johannes Gutenberg-Universitä Mainz

hep-lat, hep-ph, nucl-th

Submitted: 2026-10-16

Updated: 2026-11-13

Comments: 48 pages, 21 figures, 24 tables; v2: author list amended, references added, energy levels included as ancillary file

Code: https://github.com/gplepage/gvar

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 95/100

The gist: A lattice-QCD determination of nucleon-nucleon scattering amplitudes was performed at pion masses around M pi about 200 MeV and M pi about 280 MeV…

Key concepts

Lattice-QCD Determination
This method uses numerical simulations on a discretized spacetime grid (lattice) to calculate particle properties like scattering amplitudes. It involves discretizing space and time, then solving quantum field theory equations on this grid to extract physical observables from the simulation data.
Finite-Volume Energy Spectra
In lattice simulations, particles are confined within a finite box. The energy levels of these confined systems (spectra) are extracted from correlation matrices built using interpolating operators. These spectra must then be carefully mapped to describe the behavior of scattering amplitudes in infinite volume.
Virtual States and Riemann Sheets
In scattering theory, poles in the complex energy plane determine the nature of a state. Poles on the second Riemann sheet correspond to virtual states—unbound systems that are not true physical particles. The study found poles residing here for both isospin channels across all studied pion masses.

Terminology

Summary

A lattice-QCD determination of nucleon-nucleon scattering amplitudes was performed at pion masses around M pi about 200 MeV and M pi about 280 MeV, utilizing a single lattice spacing a about 0. p. 0064 fm across two CLS ensembles. p. 0 The finite-volume energy spectra were extracted from large Hermitian correlation matrices constructed within the distillation framework, employing an extensive basis of two-nucleon interpolating operators. p. 0 To reliably map these finite-volume spectra to infinite-volume scattering amplitudes, two formalisms—the plane-wave quantization condition and the relativistic left-hand cut formalism—were employed, explicitly accounting for the long-range one-pion exchange interaction. p. 0

The analysis investigated the quark mass dependence of nucleon-nucleon scattering lengths, effective ranges, shape parameters, and pole positions. p. 0 Across all pion masses considered, the analysis favored virtual states in both isospin channels by more than 1 sigma. p. 0 Comparison with previous results at M pi about 710 MeV showed that inverse scattering lengths and binding momenta approach their physical-point phenomenological values monotonically as the pion mass is lowered. p. 0

The numerical methodology involved using the distillation framework, which utilized smeared quark propagators (perambulators) computed via software like QUDA-LapH. p. 1 The formal treatment incorporated two distinct finite-volume frameworks to ensure consistent relation to infinite-volume amplitudes where one-pion exchange cannot be neglected. p. 1 Pole positions were determined at M pi about 200 MeV and M pi about 280 MeV, and these were compared against results from the higher pion mass calculation. p. 1

The focus was on the lowest partial waves: the S-wave in the dineutron system and the alpha-wave in the deuteron system, analyzed via Blatt-Biedenharn decomposition. p. 8 The central finding indicated that poles reside on the second Riemann sheet, corresponding to virtual states. p. 8 As M pi decreased, the pole in the deuteron channel moved toward threshold, consistent with its eventual transition into a physical bound state on the first Riemann sheet. p. 8 Conversely, the dineutron remained a virtual state throughout the considered pion mass range. p. 5 The strongest dependence on pion mass was observed in inverse scattering lengths and pole positions, while effective ranges and shape parameters exhibited milder dependence and were already broadly consistent with their physical-point values. p. 0

The left-hand-cut analysis proved sensitive to the pion-nucleon coupling g pi NN, yielding a value qualitatively consistent with phenomenological expectations, though significant model dependence was noted. p. 0 The final results were obtained by averaging four determinations—two datasets times two formalisms—of each observable, using Eqs. p. 6 (63) to (65). This procedure yielded statistical and systematic uncertainties associated with the cutoff choice and parametrization, which were added in quadrature to obtain the total error. p. 6

The results consistently favored a virtual-state pole in the corresponding NN scattering amplitude. p. 0 As M pi was lowered, the threshold intercept of p delta decreased in both isospin channels, following expected trends. p. 0 The study highlighted two crucial future directions: performing calculations at multiple lattice spacings to determine the continuum limit and conducting simulations directly at the physical point to eliminate the need to infer final pion-mass evolution. p. 1

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved uncertainty estimation by implementing a parametric bootstrap that uses posterior variance multiplied by a widening factor, which we observe reduces outlier results in Figure 17 when compared to linear error propagation from boot0.

  2. Enhanced ability to assess systematic uncertainty by performing model and data-subset averaging based on the Akaike Information Criterion (AIC) weights, allowing for a controlled assessment of systematic variations across different analysis variants.

  3. Capacity to directly constrain the one-pion-exchange coupling gpiNN by using the relativistic left-hand cut formalism, as it is sensitive to this quantity and its extracted value is qualitatively consistent with the expected phenomenological value.

  4. Capability to distinguish between virtual and bound state poles by analytically continuing constrained amplitudes, which allows for a determination of pole structure on the second Riemann sheet.

  5. Increased precision in extracting scattering parameters by utilizing a robust fitting procedure that minimizes a chi squared function relating lattice-QCD energies to quantization condition predictions, with systematic uncertainties assessed through derivative methods rather than relying solely on bootstrap resampling.

Sources

Related papers