⚛️ nuclear theory

Exact factorization of finite-range three-body matrix elements between Gaussian wave packets and its application to the $0^+$ states of $^{12}C$

This paper introduces an exact factorization method using graph Laplacians to efficiently compute finite-range three-body matrix elements in antisymmetrized molecular dynamics, enabling accurate descriptions of the ground and Hoyle states of $^{12}$ C that were previously unattainable without such terms.

Original authors: Takatoshi Ichikawa

Published 2026-11-03

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Takatoshi Ichikawa

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

At the heart of every atom lies the nucleus, a tiny, dense cluster of protons and neutrons held together by the strong nuclear force. For decades, physicists have tried to build a complete picture of how these particles arrange themselves, especially in light elements like carbon-12, which consists of twelve such particles. A major challenge in this field has been explaining why nuclei do not collapse into a single, infinitely dense point. Simple models that only consider how pairs of particles interact often fail to stop this collapse, leading to predictions that the nucleus would shrink far smaller than it actually is. To fix this, scientists introduced a more complex rule: a three-body interaction, where the force depends on the simultaneous presence of three particles. While this rule successfully stops the collapse and matches real-world observations, calculating its effects is notoriously difficult. The mathematics required to track how three particles influence each other across a whole nucleus is so vast that it has historically been impossible to run detailed simulations for anything but the simplest cases.

A researcher at the Yukawa Institute for Theoretical Physics in Kyoto has now solved this computational bottleneck, allowing for the first time a precise look at how these three-particle forces shape the carbon-12 nucleus. By developing a new mathematical shortcut, the team was able to calculate the energy of the nucleus with a speed that makes detailed study feasible. They applied this method to explore two specific states of carbon-12: the stable ground state that makes up ordinary matter, and a rare, excited state known as the Hoyle state, which is crucial for how stars create carbon. The study reveals that when the three-particle forces are included, the nucleus settles into a size and energy level that matches experimental measurements almost perfectly. Without these forces, the nucleus would collapse into a tiny, unstable blob, proving that the three-body interaction is not just a minor correction but a fundamental necessity for the existence of the universe as we know it.

The core of this breakthrough lies in how the researcher handled the sheer volume of data required to simulate the nucleus. In a nucleus with twelve particles, calculating the energy involves tracking the positions and movements of every possible combination of three particles at once. In a direct calculation, the number of steps required grows so rapidly that even the fastest supercomputers would take an impractical amount of time to finish a single simulation. The researcher realized that the mathematical structure of these three-particle forces shared a hidden pattern with a technique used in machine learning to group data points. By treating the distances between particles like a network of connections, they were able to break the massive calculation down into smaller, manageable pieces. Instead of calculating the interaction for every single trio of particles individually, the new method pre-calculates the relationships between pairs of particles and then combines them. This reduced the computational time from a scale that would take years to one that takes only a few hours, effectively turning an impossible task into a routine one.

With this new tool in hand, the team turned their attention to carbon-12, a nucleus that is famous in physics for having two very different personalities. The first is its ground state, a compact and stable arrangement that forms the basis of life. The second is the Hoyle state, a rare, expanded version of the nucleus that exists only fleetingly in the hearts of stars, where it allows three helium nuclei to fuse into carbon. Previous models struggled to describe both states simultaneously, often predicting that the nucleus would either collapse or expand too much. The new simulations, which included the three-body forces, showed that the ground state naturally settles at an energy level just below the point where the nucleus would break apart into three helium pieces. This matches the real-world measurement of the nucleus's binding energy with high precision. The study confirms that the three-body force is the key ingredient that prevents the nucleus from collapsing, acting as a repulsive pressure that keeps the particles at a healthy distance.

The researchers also examined the Hoyle state, finding that it sits just above the energy threshold where the nucleus would fall apart, a position that is critical for the production of carbon in the universe. When they adjusted the parameters of their model to match the ground state, the Hoyle state appeared at an energy level within a fraction of a percent of the experimental value. This success was consistent across three different versions of the nuclear force they tested, suggesting that the result is robust and not an artifact of a specific choice of numbers. The study also measured the size of the nucleus in these two states. The ground state was found to be compact, with a radius of about 2.33 femtometers, while the Hoyle state was significantly larger, with a radius of roughly 3.3 to 3.5 femtometers. This difference in size reflects the different ways the particles arrange themselves: tightly packed in the ground state and loosely clustered in the excited state.

A significant part of the work involved ensuring that the mathematical model did not produce misleading results due to hidden symmetries. In quantum mechanics, particles have properties like spin and momentum that can sometimes create artificial "dips" in energy calculations, making a state look more stable than it really is. The researcher introduced a constraint that forced the model to respect a specific symmetry called time-reversal invariance, which essentially means the physics looks the same whether time runs forward or backward. This constraint removed the artificial dips and ensured that the calculated states were physically real. The study found that this restriction raised the energy of the ground state by a small amount, bringing it even closer to the experimental value, and confirmed that the Hoyle state remains stable and correctly positioned.

The implications of this work extend beyond just carbon-12. The method developed to speed up the calculations can be applied to other nuclei and other types of three-body forces, opening the door to a new era of precision in nuclear physics. It allows scientists to test different theories of how matter holds together without being limited by the speed of their computers. The study concludes that the three-body interaction is not merely a theoretical add-on but a genuine physical force that dictates the structure of the nucleus. Without it, the carbon we are made of would not exist in its current form, and the universe would lack the stable building blocks necessary for life. By solving the mathematical puzzle that has hindered progress for so long, this research provides a clearer, more accurate map of the forces that shape the atomic world.

Technical Summary: Exact Factorization of Finite-Range Three-Body Matrix Elements and Application to $^{12}\text{C}$

Problem Statement
Microscopic nuclear structure models, such as Antisymmetrized Molecular Dynamics (AMD), require genuine three-body interactions to reproduce the saturation property of nuclear matter, which two-body effective interactions alone cannot achieve. While density-dependent terms in energy density functionals offer saturation, they introduce complications when projecting onto good angular momentum or particle number, often requiring ad-hoc prescriptions that lead to spurious divergences. Phenomenological finite-range Gaussian three-body interactions (e.g., Tohsaki F1, Kanada-En'yo–Akaishi F3B) avoid these issues as genuine operators. However, their direct application in $\beta$ – $\gamma$ -constrained Generator Coordinate Method (GCM) calculations has been computationally prohibitive. The direct evaluation of three-body matrix elements between Slater determinants of Gaussian wave packets scales as $O(A^6)$ in the number of complex exponential evaluations and relies on short loops that fail to utilize modern processor architectures efficiently. This bottleneck prevents the study of saturation properties in light nuclei like $^{12}\text{C}$ using these rigorous interactions.

Methodology
The paper introduces a method to exactly factorize the matrix elements of finite-range three-body Gaussian interactions between Slater determinants, reducing the computational cost to approximately one order of magnitude above that of two-body terms. The approach involves two primary steps:

Graph Laplacian Factorization: Drawing on spectral clustering techniques from machine learning, the author utilizes the graph Laplacian of the interaction ranges. The exponent of a Gaussian three-body kernel is a quadratic form weighted by this Laplacian. By integrating over the three nucleon coordinates, the exponent of the kernel between two Slater determinants is shown to retain this quadratic form but with new weights. This allows the exponential factor to be exactly factorized into pairwise terms, each depending on only two of the six single-particle indices. Consequently, the number of required exponential evaluations drops from $O(A^6)$ to $O(A^2)$ , stored in pairwise tables.
Matrix-Matrix Contraction: The remaining contraction of the three-body density (a $3 \times 3$ determinant of the transition density matrix) with the factorized kernel is rewritten using a cofactor expansion along one ket index. This reformulates the summation over the remaining indices as a series of complex matrix-matrix products. While the operation count remains $O(A^6)$ , these products are executed using optimized BLAS routines (specifically zgemm), which achieve a significantly higher fraction of the processor's peak performance compared to the direct evaluation of individual tuples.

Additionally, the study restricts the intrinsic states to those invariant under time reversal ( $\hat{T}$ ). This constraint eliminates spurious energy dips caused by intrinsic states with net spin or orbital currents that satisfy deformation constraints but possess negligible $0^+$ projection components. This restriction ensures the stability of the variation after projection (VAP) without removing the physical model space, as the superposition of time-reversal-invariant states still spans the full space.

Key Results
The method was applied to the $0^+$ states of $^{12}\text{C}$ using three interactions: Volkov No. 2 (two-body only), Tohsaki F1, and Kanada-En'yo–Akaishi F3B. The Majorana parameters were adjusted to reproduce the experimental ground-state energy relative to the $3\alpha$ threshold ($-7.275$ MeV).

Computational Performance: For $^{12}\text{C}$ , the factorization reduced the time required for projected three-body kernels from 530 times to 24 times that of the two-body kernels. The speed-up relative to direct evaluation grows with mass number $A$ , reaching a factor of 93 for $^{40}\text{Ca}$ . Parallelization over Euler nodes provided an additional speed-up factor of 43–56.
Energy Surfaces: The energy surfaces on the $\beta$ – $\gamma$ plane for all three interactions exhibited similar valley structures. The minima were located at different deformation parameters but lay 1.8–2.9 MeV above the experimental ground state, with the correlation energy recovered by the GCM superposition.
Spectroscopy: When the ground state was fixed to the experimental position, the Hoyle state ( $0^+_2$ $0_{2}^{+}$ ) was found within 0.6 MeV of the experimental value (0.38 MeV above threshold) for all three interactions:
- Volkov No. 2: $+0.97$ MeV
- F3B*: $+0.96$ MeV
- F1: $+0.22$ MeV
  The Hoyle state radius was found to be 3.3–3.5 fm, consistent with the position relative to the threshold, though the monopole transition matrix elements ($6.5 $–$ 8.1$ fm $^2$ ) exceeded the experimental value ( $5.47 \pm 0.09$ fm $^2$ ).
Role of Three-Body Terms: Without the three-body term, $^{12}\text{C}$ collapses to energies 47–81 MeV below the $3\alpha$ threshold. The three-body term acts as a repulsive force at short distances that prevents this collapse, raising the compact ground state more than the dilute Hoyle state. The difference in the Hoyle state positions between F1 and F3B* was attributed not to the three-body term itself, but to the contributions of the rest of the Hamiltonian (kinetic and two-body terms).

Significance and Claims
The paper claims that the proposed factorization method makes $\beta$ – $\gamma$ -constrained GCM calculations with genuine finite-range three-body interactions computationally feasible for light nuclei. It demonstrates that these interactions, when calibrated to the ground state, naturally reproduce the position of the Hoyle state near the $3\alpha$ threshold without the need for artificial prescriptions required by density-dependent functionals.

The study highlights that the three-body interaction is not merely a correction to a bound system but a fundamental component required to prevent nuclear collapse. Furthermore, the results suggest that the specific position of the Hoyle state relative to the threshold is sensitive to the details of the two-body interaction and the spin-orbit term, rather than being solely determined by the three-body term. The author concludes that their method enables the consistent determination of interaction parameters by fitting projected energies of closed-shell nuclei and scattering data, a task previously hindered by computational cost. The restriction to time-reversal-invariant states is presented as a necessary step to avoid spurious solutions in the variation-after-projection framework, though the physical nature of the excluded correlations (displacement of time-reversed pairs) remains an open question.

More like this