Dimerization in the SU(4) Heisenberg model on the cubic lattice: iPEPS study

arXiv:2510.00284 · cond-mat.str-el, cond-mat.quant-gas · Submitted 2025-09-30 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Quantum Radio. Generated commentary on the latest quantum physics and condensed matter papers.

Kai: We're looking at the paper: "Dimerization in the SU(4) Heisenberg model on the cubic lattice".

Paper summary: Kai: So we're talking about this paper called "Dimerization in the SU(four) Heisenberg model on the cubic lattice: iPEPS study" <ref:2510.00284#pg1,Dimerization in the SU(4) Heisenberg model on the cubic lattice: iPEPS>. Basically, they are looking at how these dimerized color-ordered phases stick around when you change the shape of the lattice and the coupling strengths.

Mira: Right. The main idea is that they use tensor-network methods to see if these phases persist across different regimes of spatial anisotropy in a three dimensional SU(four) symmetric Heisenberg model on a cubic lattice, which is important because it suggests that three dimensional SU(N) symmetric lattice systems can still be hosts for exotic quantum phases beyond what mean field theory predicts <ref:2510.00284#pg1,SU(4) symmetric Heisenberg model on>.

Lev: From an engineering side, seeing if these phases survive across different regimes is key because it tells us where the computational overhead would actually land on real hardware when we try to simulate this stuff.

Kai: So they tackle the model by looking at three different approaches: first, they look at the SU(four) symmetric Heisenberg model on the square lattice, and then they consider a limit where one coupling is much stronger than others using an MPS plus mean field theory approach <ref:2510.00284#pg1,SU(4) symmetric Heisenberg model on the>.

Mira: And finally, they focus on the genuine three dimensional isotropic point where all couplings are equal and use a three dimensional iPEPS wave function for that analysis.

Lev: That sounds like a lot of complexity when you're trying to map this onto something we can actually run on our current quantum error correction setups.

Kai: Exactly, and looking at the results, they find that the dimerized color-ordered state shows up in the two dimensional limit when one coupling is zero and the other isn't equal, and they see that same structure in both anisotropic and fully isotropic three dimensional regimes.

Mira: They also found that dimerization develops as a consequence of interchain coupling even in the one dimensional limit when using the decoupled chain approach with MPS plus mean field theory.

Lev: That means if you try to build something simple, like just chains, the way they connect them—the interchain coupling—is what forces that dimerization to appear, which is a necessary check for any real-world simulation.

Kai: And for the isotropic point in three dimensions, they observe that the bonds can be divided into three groups based on their energy, and they find the average energy of bonds in the first group is equal to minus zero point one eight two three.

Mira: That specific number for that average energy tells us something concrete about how strongly those particular bonds are involved in forming the structure, which is a bit more detailed than just saying a phase exists.

Lev: If we were running this on hardware, we'd need to figure out how accurately we could measure that energy difference between those three groups of bonds when the system is truly isotropic.

Kai: So what this means for us in general, based on this paper about the SU(four) Heisenberg model and iPEPS study, it points toward the persistence of a dimerized color-ordered phase across both anisotropic and fully isotropic three dimensional regimes <ref:2510.00284#pg1,the SU(4) Heisenberg model>.

Mira: And this persistence suggests that three dimensional SU(N) symmetric lattice systems remain very promising as hosts for exotic quantum phases beyond what mean field theory predicts, which is a significant implication for condensed matter physics.

Lev: It means the structure they found isn't just a fluke of the specific geometry they used; it's robust enough to potentially exist in more complex materials we might encounter.

Kai: The paper itself, "Dimerization in the SU(four) Heisenberg model on the cubic lattice: iPEPS study", uses tensor-network methods to analyze these different coupling regimes, showing how these phases are stable even when you move away from simple mean field approximations <ref:2510.00284#pg1,Dimerization in the SU(4) Heisenberg model on the cubic lattice: iPEPS>.

Mira: It's about using those advanced tools to map out phase diagrams that might be too complex for simpler theories to handle, and the results show that the iPEPS approach captures this behavior well in both anisotropic and isotropic limits.

Lev: The limitation they mention is that they study these regimes excluding only the limit of completely decoupled one dimensional chains, which is a practical constraint when trying to build things from scratch.

Kai: So, overall, the whole point of this work is to show that these dimerized states are stable in three dee SU(four) systems under various conditions, which gives us confidence that searching for these kinds of exotic quantum phases in real three dimensional materials is a viable path forward <ref:2510.00284#pg1>.

Conclusion: Kai: So, we're wrapping up this look at the paper "Dimerization in the SU(four) Heisenberg model on the cubic lattice: iPEPS study".

Mira: It’s about how these dimerized color-ordered phases stick around even when you change the shape of the lattice or make it totally isotropic.

Lev: The authors are using tensor-network methods, specifically iPEPS, to study this SU(four) symmetric Heisenberg model on a cubic lattice.

Kai: They're checking if these phases persist across different spatial anisotropy regimes in three dimensional SU(N) symmetric systems.

Mira: What’s genuinely interesting is that they show this persistence holds even in the fully isotropic three dimensional case, which was hard to predict with simpler theories.

Lev: The implication is that these three dimensional SU(N) symmetric lattice systems are still promising hosts for exotic quantum phases beyond what mean field theory predicts.

Kai: They found that the dimerized color-ordered states show up in both the limit of very strong coupling along one axis and the opposite, when you go to zero on another axis.

Mira: And they found that dimerization develops as a consequence of interchain coupling even when you look at the one dimensional limit using their decoupled chain method.

Lev: That means if you try to build something simple, like just chains, the way they connect them—the interchain coupling—is what forces that structure to appear.

Kai: They divide the lattice bonds into three groups based on their energy in the isotropic point, and the average energy of those bonds in the first group is around minus zero point one eight two three.

Mira: That specific number tells us something concrete about how strongly those particular bonds are involved in forming that structure.

Lev: If we were running this on hardware, we’d need to figure out how accurately we could measure that energy difference between those three groups of bonds when the system is truly isotropic.

Kai: Overall, it points toward the persistence of a dimerized color-ordered phase across both anisotropic and fully isotropic three dimensional regimes.

Mira: This suggests that searching for these kinds of exotic quantum phases in real three dimensional materials is a viable path forward.

Lev: The next step would be to see if this method can be extended to different lattice geometries, like the pyrochlore lattice.

I.V. Lukin, A.G. Sotnikov

Akhiezer Institute for Theoretical Physics · Haiqu, Inc. · Education and Research Institute “School of Physics and Technology” Karazin Kharkiv National University

cond-mat.str-el, cond-mat.quant-gas

Submitted: 2025-09-30

Updated: 2026-11-17

Comments: 10 pages, 7 figures, published version

Journal ref: Physical Review B 114, 274423 (2026)

DOI: 10.1103/7k6p-rld9

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

Importance score: 76/100

The gist: Dimerization in the SU(4) Heisenberg model on the cubic lattice: iPEPS study is a study of how dimerized color-ordered phases persist throughout various regimes of spatial anisotropy in the SU(4)…

Key concepts

SU(4) Symmetric Heisenberg Model
This is a specific type of quantum spin model defined on a cubic lattice. The 'SU(4)' part relates to the internal degrees of freedom (spins), and the 'Heisenberg' part describes how these spins interact. It's a complex system used to explore exotic quantum states.
iPEPS
iPEPS stands for imaginary-time Projected Entangled Pair States, which is a sophisticated tensor-network method. This technique is used to efficiently represent and calculate the ground state properties of strongly correlated quantum systems on lattices, especially in three dimensions.
Dimerized Color-Ordered Phase
This refers to a specific pattern of spin arrangement where neighboring spins pair up (dimerize) and exhibit a specific color order. The research shows this ordered state is robust, meaning it survives even when the lattice structure becomes highly stretched or perfectly uniform.
Anisotropic Limits
These are different physical scenarios where the coupling strengths between neighboring sites on the cubic lattice are not all equal. The study examines limits where one coupling (like Jz) is much stronger than others, allowing researchers to see how the system behaves when it's stretched or highly directional.

Terminology

Summary

Dimerization in the SU(4) Heisenberg model on the cubic lattice: iPEPS study is a study of how dimerized color-ordered phases persist throughout various regimes of spatial anisotropy in the SU(4) symmetric Heisenberg model on a cubic lattice using tensor-network methods, which is important because it suggests that three-dimensional SU(N)-symmetric lattice systems remain promising hosts for exotic quantum phases beyond mean-field theory results

Model Description and Limits

The main focus of the current study is the SU(4)-symmetric Heisenberg model on the simple cubic lattice <ref:2510.00284#pg3>. The Hamiltonian of this model has the following form H = X ⟨ij⟩γ JγPi,j, (1) <ref:2510. p. 100284#pg2>. In the limit Jz → 0, the layers in xy-planes decouple from one another, and it is known that in this limit, the ground state of the SU(4) Heisenberg model on the square lattice spontaneously breaks both translational and cyclic rotational symmetry, as well as continuous SU(4) spin symmetry <ref:2510. p. 100284#pg3>. In the opposite limit Jz → ∞, the chains of sites in the z direction also decouple from each other and are described separately by the SU(4)-symmetric Heisenberg models on the chain <ref:2510. p. 100284#pg3>.

Methodology for Analysis

The theoretical analysis consists of three parts:

  1. We first briefly return to the SU(4)-symmetric Heisenberg model on the square lattice, which was already studied in Ref. p. 1 [26] <ref:2510. p. 000284#pg2>.

  2. We consider another limit, where Jz ≫ Jx = Jy and work in the framework of the decoupled MPS (decoupled chains) ansatz to study this regime with the MPS plus mean-field theory (MPS + MF) approach [55–57] <ref:2510. p. 100284#pg2>.

  3. We focus on the genuine 3d isotropic point with Jx = Jy = Jz and employ 3d iPEPS <ref:2510. p. 100284#pg2>.

Analysis of Anisotropic Limits

A. 2d iPEPS

We begin our theoretical analysis from the SU(4)-symmetric Heisenberg model on the square lattice with anisotropic couplings (Jx ̸= Jy) <ref:2510.00284#pg2>. We take the elementary unit cells of size 4 × 2 and 4 × 4 <ref:2510. p. 200284#pg2>. The iPEPS is initialized with a random product state and then optimized using the Simple Update method with the gradually decreased time step [18, 19, 58] <ref:2510. p. 200284#pg2>. After the optimization, we compute energies for different lattice bonds using the double-layer CTMRG method [59, 60] <ref:2510. p. 200284#pg2>.

B. MPS + MF

We are also interested in the limit where the coupling along one crystallographic axis is much stronger than the couplings in the transverse plane <ref:2510.00284#pg2>. In this limit, we propose to use the ansatz consisting of separate matrix product states (MPS) along different chains, as shown in Fig. p. 2 2(c) <ref:2510. p. 000284#pg2>. Such a combination of MPS along the chain and mean-field theory treatment of the interaction between the chains is known as the MPS + MF framework [55–57] <ref:2510. p. 200284#pg2>.

Analysis of Isotropic Point

C. 3d iPEPS

In the proximity of the isotropic point of the model (1), we employ 3d iPEPS wave functions, as shown in Fig. 2(a) <ref:2510. p. 000284#pg2>. The 3d iPEPS wave function is optimized using imaginary time evolution with the Simple Update method [18, 19, 58] <ref:2510. p. 200284#pg2>. After this initial step, the observables of iPEPS are computed with a more accurate CTMRG method <ref:2510. p. 200284#pg2>.

Key Findings

A. 2d limit

In particular, we analyze the 2d limit by taking Jz = 0 and Jx ̸= Jy <ref:2510.00284#pg2>. In the isotropic case, it is known that the model is in the color-ordered dimerized phase [26] <ref:2510. p. 400284#pg2>. We observe the same structure of the dimerized phase as in the isotropic limit <ref:2510. p. 400284#pg2>.

B. 1d limit

We study the behavior in this limit within the decoupled chain MPS + MF ansatz <ref:2510.00284#pg2>. It is clear that, in both cases, dimerization develops as a consequence of interchain coupling <ref:2510. p. 500284#pg2>.

C. Isotropic point in 3d

From the two previous subsections, it is clear that dimerized color-ordered states appear in the cubic lattice both in the limit of large Jz and in the opposite limit Jz → 0 <ref:2510.00284#pg2>. We observe that the bonds can be divided into three groups according to their respective energies <ref:2510. p. 500284#pg2>. The average energy of bonds in the first group is equal to Ex ≈ Ey = −0. p. 51823 <ref:2510. p. 000284#pg2>.

Conclusion and Outlook

Our numerical results point towards the persistence of the dimerized color-ordered phase known from the 2d limit [26] in both anisotropic and completely isotropic 3d regimes <ref:2510.00284#pg2>. This, in turn, is a sign that the three-dimensional SU(N)-symmetric lattice systems remain very promising as hosts for exotic quantum phases beyond the mean-field theory results [66] <ref:2510. p. 600284#pg2>. One of them is the extension of the method to different lattice geometries <ref:2510. p. 600284#pg2>. The most straightforward direction to proceed is to study the Heisenberg model on the pyrochlore lattice [68] <ref:2510. p. 600284#pg2>. A lighter version of this approach is the implementation of a 3d extension of the neighborhood tensor update from Ref. p. 6 [72] <ref:2510. p. 000284#pg2>. One can also investigate issues with potential CTMRG ill-convergence and compare the observable calculations from the rotated tensor networks to analyze algorithmic convergence <ref:2510. p. 600284#pg2>.

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. More accurately predict ground-state properties of SU(4) symmetric Heisenberg models on cubic lattices across various spatial anisotropy regimes by employing 3d iPEPS variational wave functions optimized with boundary iPEPS plus CTMRG.

  2. Analyze the persistence of dimerized color-ordered phases throughout the whole range of magnetic couplings, including the limit of decoupled one-dimensional chains, by studying how coupling SU(4) chains in transverse directions induces dimerization.

  3. Determine whether dimerized color-ordered states in cubic lattices are adiabatically connected between the limit of large Jz and the limit Jz to zero by studying the system at the isotropic point using 3d iPEPS with bond dimension D = 5.

  4. Calculate local onsite reduced density matrices for all sites inside a unit cell and energies on all bonds in 3d isotropic SU(4) Heisenberg model, achieving accuracy below 10-5 for onsite elements and up to 10-4 for bond energies.

  5. Identify the groups of bonds based on their respective energies (x-y bonds, even z-bonds, odd z-bonds) and characterize their average energy values in the isotropic 3d case using optimized 3d iPEPS results.

Sources

Related papers