\textit{Ab initio} Gamow density matrix renormalization group for broad nuclear many-body resonances

This paper extends the \textit{ab initio} Gamow Density Matrix Renormalization Group (G-DMRG) method to accurately describe broad nuclear many-body resonances by introducing novel truncation schemes and orbital ordering strategies based on entanglement, successfully demonstrating convergence and obtaining the first direct calculation of the \isotope[5]{H} ground state.

The Gist

The Big Picture: Predicting the Edge of the Universe's Building Blocks

Imagine the atomic nucleus as a tiny, crowded dance floor filled with protons and neutrons. Most of the time, these dancers hold hands tightly and stay in a stable circle. But at the very edge of the "periodic table" (the drip lines), the music changes. The dancers are so loosely connected that they are about to fall off the floor entirely. These are called unbound nuclei.

For decades, scientists have been great at predicting the stable dancers. But predicting the ones that are about to fall off has been a nightmare. Why? Because these unstable nuclei aren't just sitting still; they are constantly leaking particles into the surrounding space. They are "open systems," constantly interacting with the outside world.

This paper introduces a new, powerful tool to simulate these chaotic, leaking dance floors. The authors successfully upgraded their computer program to handle these "broad" resonances—nuclei that are so unstable they barely exist for a moment before falling apart.

The Problem: The "Leaky Bucket" and the "Overcrowded Room"

To understand the challenge, imagine trying to predict the behavior of a bucket with a huge hole in the bottom.

1. The Leak (Continuum Coupling): In normal atoms, particles stay inside. In these exotic nuclei, particles are constantly trying to escape. This creates a "leak" that makes the math incredibly messy.
2. The Entanglement (The Tangled Yarn): When particles interact with this "leak," they get tangled up with the outside world. In quantum physics, this is called entanglement. The more the nucleus leaks, the more tangled the yarn becomes.
3. The Crash: The authors' previous computer program (called G-DMRG) was like a very smart librarian trying to organize books. But when the "leak" got too big, the library became so tangled that the librarian couldn't find the right books, and the computer crashed or gave nonsense answers.

The Solution: Three New Tricks

The authors developed three specific tricks to untangle the yarn and keep the library organized, even when the bucket is leaking badly.

1. The "Smart Filter" (New Truncation Scheme)

Imagine you are trying to describe a complex painting, but you only have time to look at the most important brushstrokes. Usually, you just ignore the tiny, faint ones.

• The Old Way: The computer tried to ignore small details based on a simple rule. But with these leaking nuclei, the "small details" were actually huge, chaotic noise that confused the computer.
• The New Trick: The authors added a "Smart Filter." This filter looks at the math and says, "Wait, this tiny detail is actually just noise caused by the leak. Let's throw it away before it breaks the calculation." This stopped the computer from getting overwhelmed by the chaos.

2. The "Seating Chart" (Orbital Ordering)

Imagine you are hosting a party. If you seat the loud, energetic guests next to the quiet, shy ones, the whole room gets chaotic. But if you group similar people together, the party runs smoothly.

• The Old Way: The computer added the "guests" (orbitals) to the calculation in a random or purely energy-based order. This caused the "tangled yarn" to get worse at every step.
• The New Trick: The authors created a new Seating Chart. They realized that in these nuclear parties, protons and neutrons behave differently. They grouped the protons together first, then the neutrons, and saved the "leaky" guests (the ones escaping) for last. This kept the party calm and allowed the computer to build a stable picture of the nucleus.

3. The "Best View" (Natural Orbitals)

Imagine looking at a 3D object through a foggy window. You can see it, but it's blurry. If you move to a different angle, the fog clears up, and the object looks sharp.

• The Old Way: The computer was looking at the nucleus through a "foggy" set of mathematical tools (orbitals) that weren't quite right for these unstable atoms.
• The New Trick: After getting a rough, blurry picture, the authors used a technique to rotate the view. They found the "Natural Orbitals"—the specific angles where the nucleus looks the clearest. Once they switched to this clear view, the calculation converged (finished) much faster and more accurately.

The Results: What Did They Actually Do?

Using these three tricks, the authors successfully simulated several "impossible" nuclei that had never been calculated directly before:

• Helium-5 and Helium-6: They confirmed they could handle these unstable helium atoms, which are known to be tricky.
• Hydrogen-4: They calculated the properties of a very broad, unstable hydrogen nucleus.
• Hydrogen-5 (The Big Win): They performed the first direct calculation of the ground state of Hydrogen-5. This nucleus is so unstable it's like a "ghost" that barely exists. Previous methods couldn't touch it, but this new approach managed to describe it.

The Conclusion

The paper doesn't claim to cure diseases or build new batteries. Instead, it claims to have solved a specific, difficult math problem in nuclear physics.

They proved that by using a Smart Filter to remove noise, a Seating Chart to organize the particles, and a Clear View to see the structure, we can finally simulate the most unstable, short-lived nuclei in the universe. This opens the door to testing our theories of how nuclear forces work in extreme conditions, helping us understand the limits of where matter can exist.

Technical Summary

Technical Summary: Ab Initio Gamow Density Matrix Renormalization Group for Broad Nuclear Many-Body Resonances

Problem Statement
The prediction of nuclear drip lines and the description of exotic, unbound nuclei remain significant challenges in ab initio nuclear theory. While ab initio methods have successfully described bound states and narrow resonances, they struggle with broad many-body resonances. These are unstable quantum states where the lifetime is barely larger than the interaction timescale of constituents, and their dynamics cannot be reduced to effective few-body interactions. Standard approaches often fail because:

1. Continuum Couplings: Broad resonances involve strong couplings to scattering states, which significantly increase many-body entanglement, destabilizing renormalization group methods.
2. Identification Issues: In non-Hermitian frameworks (like the Gamow formalism), distinguishing physical discrete states from a background of scattering states is difficult, particularly when the wave function is not dominated by a single configuration.
3. Computational Cost: The necessity of large model spaces to describe the asymptotic behavior of broad resonances leads to exponential scaling, making convergence difficult with existing truncation schemes.

Methodology
The authors extend the Gamow Density Matrix Renormalization Group (G-DMRG) method, which utilizes the Berggren basis (including bound, resonant, and scattering states) to treat open quantum systems. To address the specific challenges of broad resonances, the paper introduces several theoretical and computational developments:

1. Reference Space Construction and Truncation:

   • A new truncation scheme is applied during the construction of the reference space (the initial set of orbitals). Instead of building a full Configuration Interaction (FCI) space and then truncating, the authors apply particle-hole (p-h) truncations (e.g., 2p2h) while building the space. This leverages a "cascade effect" where suppressing excitations in lower orbitals suppresses higher-order excitations, significantly reducing the number of matrix elements while maintaining accuracy.

2. Orbital Ordering Strategies:

   • The authors identify that standard energy-based ordering fails to stabilize calculations for broad resonances due to the rapid increase in entanglement from continuum couplings.
   • They propose a hybrid ordering scheme:
     • Stabilization Phase: Orbitals are grouped by partial wave $(l, j)$, with proton orbitals added first in neutron-rich systems, followed by neutron orbitals relevant to decay added last. This leverages proton-neutron factorization and shell structure to stabilize the initial renormalization.
     • Optimization Phase: Once a stable basis is established, the authors switch to Natural Orbitals (NAT) ordered by occupation number ($\omega$-ordering). This redefines the reference space based on actual occupations rather than mean-field expectations, allowing for efficient truncation.

3. Stabilization of Renormalization (The $\kappa$ Truncation):

   • The authors observe that strong continuum couplings can lead to the emergence of eigenvalues with negative real parts (negative probabilities) in the reduced density matrix, causing the DMRG ansatz to collapse.
   • A new truncation parameter $\kappa$ is introduced to discard eigenvalues satisfying $|\omega_\alpha| < \kappa$ (where $\kappa \ll \epsilon$, the standard DMRG truncation). This removes a "tail" of small, complex eigenvalues associated with numerical noise and excessive entanglement, stabilizing the diagonalization of complex-symmetric matrices without significantly affecting the physical energy or width.

4. Physical State Identification:

   • The standard "overlap method" (selecting the state with the largest overlap with a pole-space solution) is adapted. The authors emphasize that for broad resonances, the reference space must be carefully constructed to include relevant discrete states, and the use of NATs helps maintain the physical character of the state during renormalization.

Key Results
The authors apply this refined G-DMRG framework to several light exotic nuclei, demonstrating convergence and control over renormalization:

• $^5$He ($J^\pi = 3/2^-$): A typical single-particle resonance. The method reproduces near-exponential convergence in energy and stable width behavior, validating the approach against No-Core Gamow Shell Model (NCGSM) benchmarks.
• $^6$He ($J^\pi = 0^+, 2^+$): The ground state (halo) and the first excited state (resonance) are calculated. The NAT basis significantly improves convergence, with the $2^+$ state showing saturation in energy and width.
• $^4$H ($J^\pi = 2^-$): A broad resonance ($\Gamma/2 \sim E_r$). The study demonstrates that the new ordering and truncation schemes are essential to obtain stable results. The NAT basis reveals emergent orbitals (e.g., $0p_{1/2}$) not present in the original basis, lowering the energy by over 1 MeV.
• $^5$H ($J^\pi = 1/2^+$): The authors present the first direct ab initio calculation of the ground state of $^5$H. Using the proposed recipe, they obtain a converged energy and width, demonstrating the method's capability to handle extreme unbound systems.
• $^{12}$C Benchmark: To demonstrate computational efficiency, the authors successfully calculated the ground state of $^{12}$C using modest resources (128 cores), showing that the new truncation schemes allow for large-scale ab initio calculations previously considered unfeasible.

Significance and Claims
The paper claims to successfully extend the reach of ab initio G-DMRG into the regime of broad many-body resonances, a domain where previous methods often failed due to numerical instability and lack of convergence.

• Control of Entanglement: The work demonstrates that entanglement induced by continuum couplings can be controlled through specific orbital ordering and density matrix truncation.
• Systematic Recipe: The authors propose a concrete "recipe" for converging G-DMRG calculations: (1) stabilize with a specific ordering and reference space, (2) generate Natural Orbitals, (3) reorder by occupation, and (4) remove p-h truncations in the NAT basis.
• First Principles Calculation: The calculation of the $^5$H ground state represents a significant milestone, providing a direct ab initio result for a system that had previously only been studied via indirect extrapolation methods (e.g., complex-scaled Faddeev-Yakubovsky with confinement).

The authors remain modest regarding the absolute precision of their results, noting that discrepancies with experiment or other high-precision calculations (e.g., in $^4$H and $^5$H) are likely due to limited model spaces and the absence of explicit three-body forces in their current interaction. However, they assert that the developed methodology provides a robust framework for systematic tests of nuclear forces in light exotic nuclei.