🔬 condensed matter

Altermagnetism and bond-nematicity in the spin- $1/2$ square lattice $J_1-J_2-\delta$ model

Using a novel machine learning approach combining symmetry-enhanced neural networks and variational Monte Carlo, the study reveals that melting altermagnetic order in the frustrated spin- $1/2$ square lattice $J_1-J_2-\delta$ model via increased frustration induces an exotic phase characterized by coexisting symmetry-protected topological valence bond solid and bond-nematic orders driven by magnon pair condensation.

Original authors: Tanja {\DJ}urić, Shenlong Yu, Pinaki Sengupta

Published 2026-06-15

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Tanja {\DJ}urić, Shenlong Yu, Pinaki Sengupta

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 a vast, flat dance floor made of a grid of tiny dancers (these are the "spins" in the material). In the world of physics, how these dancers move and hold hands determines the material's magnetic personality. This paper explores what happens when we change the rules of their dance, specifically in a material called an altermagnet.

Here is a simple breakdown of their discovery:

1. The Starting Point: The "Altermagnet" Dance

Usually, magnets are either Ferromagnets (everyone faces the same way, like a marching band) or Antiferromagnets (neighbors face opposite ways, like a checkerboard).

Altermagnets are a newly discovered "hybrid" dancer. They look like a checkerboard (neighbors face opposite ways), but they have a special twist: their dance moves create a "chiral" (handed) effect.

The Analogy: Imagine two groups of dancers. Group A spins clockwise, and Group B spins counter-clockwise. In a normal checkerboard, these spins cancel out perfectly. But in an altermagnet, the way they are arranged creates a hidden "spin current" or a specific directionality, even though the total spin is zero.
The Paper's Finding: The researchers confirmed that in a "weakly frustrated" state (where the dancers mostly agree on the rules), this material behaves exactly like an altermagnet. The "music" (energy waves called magnons) splits into two distinct tracks based on their direction, a unique signature of this magnetic type.

2. The Twist: Adding "Frustration"

The researchers then cranked up the difficulty. They introduced frustration.

The Analogy: Imagine the dancers are told to hold hands with their neighbors, but the rules are contradictory. "Hold hands with the person to your right, but also with the person to your left, but don't let go of the person behind you." The dancers get confused. They can't all satisfy the rules at once.
The Result: As the confusion (frustration) increases, the neat altermagnetic dance breaks down. The dancers stop marching in a perfect pattern. This is called "melting" the order.

3. The Surprise Discovery: A New "Bond-Nematic" Phase

When the altermagnetic order melted, the researchers expected the dancers to just become a chaotic, liquid mess (a "spin liquid"). Instead, they found something much stranger and more organized: a Bond-Nematic phase mixed with a Topological Valence Bond Solid (SPT VBS).

Let's break down these fancy terms with analogies:

Bond-Nematicity (The "Directional Hug"):
In a normal liquid, molecules move randomly. In this new phase, the dancers aren't holding hands in a fixed pattern (like a grid), but they have agreed on a direction to face.
- The Analogy: Imagine a crowd of people in a park. They aren't holding hands in a grid, but everyone has unconsciously decided to lean slightly toward the East. They aren't pointing at a specific person, but the relationship between them has a preferred direction. This is "nematicity." It breaks the symmetry of the room (it's no longer the same in all directions) without creating a rigid grid.
- The Paper's Claim: The researchers found that the "magnons" (energy waves) paired up and condensed into this state, breaking the usual rules of spin rotation.
SPT Valence Bond Solid (The "Secret Handshake"):
This is a phase where the dancers form pairs (valence bonds) that are "protected" by the rules of the universe (topology).
- The Analogy: Imagine the dancers form pairs. If you try to swap two pairs, the whole system "feels" the change, even if you can't see the individual dancers moving. It's like a secret handshake that only works if the group stays in a specific formation. The paper found that this phase has a "topological" nature, meaning it has a hidden robustness that depends on the size of the dance floor (whether the number of dancers is divisible by 4 or not).

4. The "Chiral" Ghost in the Machine

Here is the most mind-bending part of the discovery. Even though the neat altermagnetic order was destroyed, the remnants of its "handedness" (chirality) survived in the new phase.

The Analogy: Even though the marching band broke up and started leaning in a new direction, the music still has a "left-handed" and "right-handed" echo. The energy levels of the new "triplon" particles (excited states) split apart based on their chirality, just like they did in the original altermagnet.
The Paper's Claim: The new phase breaks two specific symmetries:
1. U(1) Spin Rotation: The dancers can no longer rotate freely; they are locked into a specific orientation.
2. Z2 Spin Inversion: The "mirror image" of the dance is no longer identical to the original.

5. How They Found It

The researchers didn't just guess this; they used a powerful new tool: Machine Learning.

The Method: They built a "neural network" (a type of AI) that was taught the rules of symmetry (like rotation and reflection). They used this AI to simulate the dance floor with millions of possible moves to find the lowest energy state.
The Result: The AI confirmed that when frustration is high, the system settles into this exotic "Bond-Nematic + Topological" state, which is a mix of directional order and hidden topological protection.

Summary

The paper claims that if you take a specific magnetic material (the $J_1-J_2-\delta$ model) and make it "frustrated" enough to break its altermagnetic order, it doesn't just turn into a messy liquid. Instead, it transforms into a complex, exotic state where:

The spins form a directional "lean" (Bond-Nematic).
They form protected pairs (Topological VBS).
They still retain a "handed" split in their energy levels, a ghost of the original altermagnetic order.

This discovery is significant because it identifies a new "exotic phase of matter" that exists right next to altermagnets, showing that these materials are far more complex and capable of hosting strange quantum states than previously thought.

Technical Summary: Altermagnetism and Bond-Nematicity in the Spin-1/2 Square Lattice $J_1-J_2-\delta$ Model

Problem Statement
The paper investigates the phase diagram of the spin-1/2 square lattice $J_1-J_2-\delta$ model, specifically focusing on the transition from an altermagnetic (AM) ordered state to magnetically disordered phases driven by increased geometric frustration and quantum fluctuations. While recent literature has extensively explored the impact of altermagnetic moments on electronic band structures and superconductivity, the influence of strong quantum fluctuations on the melting of altermagnetic order and the resulting exotic phases remains underexplored. The study addresses the regime of intermediate to strong frustration ( $J_2/J_1 \approx 0.5$ ) and significant exchange modulation ( $\delta$ ), where the competition between nearest-neighbor ( $J_1$ ) and next-nearest-neighbor ( $J_2$ ) interactions, modulated by $\delta$ , is expected to yield complex ground states beyond standard antiferromagnetism or simple spin liquids.

Methodology
The authors employ a novel machine learning (ML) approach combining symmetry-enhanced Neural Network Quantum States (NQS) with Variational Monte Carlo (VMC).

Model: The Hamiltonian includes nearest-neighbor coupling $J_1$ and next-nearest-neighbor couplings $J_2(1 \pm \delta)$ , introducing geometric frustration and breaking the equivalence of sublattice environments.
Ansatz: The wavefunction is represented by a Group Equivariant Convolutional Neural Network (GCNN). This architecture explicitly incorporates the space-group symmetries of the Hamiltonian (lattice translations and $C_{4v}$ point group symmetries) and $Z_2$ spin parity.
Optimization: The network parameters are optimized using the Stochastic Reconfiguration (SR) method, equivalent to Natural Gradient Descent (NGD), to minimize a free energy loss function ($F = E - TS$) that includes a pseudo-entropy term to stabilize training.
Simulation: Calculations are performed on $L \times L \times 4$ clusters with periodic boundary conditions using the NetKet, JAX, FLAX, and OPTAX libraries on GPUs. Results are extrapolated to the thermodynamic limit.

Key Results

Weak Frustration Regime (Altermagnetism):
- For $J_2/J_1 = 0.2$ and $\delta = 0.5$ , the ground state is confirmed to be altermagnetic.
- The order is characterized by a non-zero squared staggered magnetization in the thermodynamic limit.
- Chiral Splitting: The magnon excitation spectrum exhibits a distinct splitting between modes of opposite chirality. This splitting is reflected in the energy levels of irreducible representations (irreps) $A_1$ and $B_1$ (and $A_2$ and $B_2$ ) at high-symmetry points in the Brillouin zone. The splitting vanishes at high-symmetry points in the thermodynamic limit but is maximal at $\vec{q} = (\pm \pi/2, \pi/2)$ , consistent with $d_{xy}$ symmetry.
Strong Frustration Regime (Bond-Nematicity and SPT VBS):
- For $J_2/J_1 = 0.5$ and $\delta = 0.5$ , the altermagnetic order melts due to strong quantum fluctuations. The squared staggered magnetization vanishes in the thermodynamic limit, indicating a magnetically disordered phase.
- Coexisting Orders: The ground state is identified as a phase hosting coexisting Symmetry Protected Topological (SPT) Valence Bond Solid (VBS) order and Bond-Nematic (BN) order.
- Bond-Nematicity: The phase is characterized by the condensation of magnon pairs ( $\hat{S}^-_i \hat{S}^-_j$ ), leading to bond-nematic order. This breaks $U(1)$ spin rotation symmetry and $Z_2$ spin inversion symmetry. The bond-nematic structure factors show peaks at $\vec{q}=(0,0)$ (nematic order) and $\vec{q}=(\pi, 0), (0, \pi)$ (coexisting VBS order).
- SPT VBS Order: The dimer order parameter exhibits distinct finite-size scaling behaviors for clusters with linear dimensions $2L = 4n$ (trivial sector) and $2L = 4n+2$ (topological sector), a hallmark of SPT VBS order. The extrapolated values differ, confirming the topological nature of the phase.
- Excitation Spectrum: The low-energy spectrum features:
  - Goldstone Modes: Massless quadrupolar excitations corresponding to the spontaneous breaking of $C_4$ rotational and reflection symmetries (irreps $B_1$ and $B_2$ ).
  - Chiral Triplon-like Modes: Excitations with $S^z_{tot} = \pm 1$ and $P=-1$ exhibit chiral splitting similar to the altermagnetic regime, indicating the persistence of chirality in the disordered phase.
  - Symmetry Breaking: The splitting between $S^z_{tot} = +1$ and $-1$ levels confirms the breaking of $Z_2$ spin inversion symmetry, consistent with the non-zero expectation value of the off-diagonal bond-nematic operator $\hat{O}_{yz}$ .

Significance and Claims
The authors claim this study presents an important step in identifying exotic phases of matter emerging from the melting of altermagnetic order. Specifically:

The identification of a novel phase where SPT VBS and bond-nematic orders coexist is a significant finding, distinct from the gapless $Z_2$ Dirac spin liquid found in the $\delta \to 0$ limit.
The work highlights the complex interplay between symmetry and topology in magnetically disordered phases, noting that the bond-nematic phase breaks $U(1)$ and $Z_2$ symmetries while retaining chiral characteristics in its excitation spectrum.
The study suggests the potential existence of deconfined quantum critical (DQC) points between the identified SPT/BN phase and the $Z_2$ Dirac spin liquid.
The authors position their findings as a foundation for future research into unconventional quantum phase transitions and the nature of superconductivity in doped versions of this model.

The paper does not propose specific experimental realizations beyond noting the model's relevance to iron oxychalcogenides, ultracold fermion atoms in optical lattices, and monolayer vanadium oxides ( $V_2Se_2O$ , $V_2Te_2O$ ), nor does it speculate on applications beyond the theoretical identification of these phases.