- The paper introduces a GCNN-based neural network quantum state method within VMC to capture altermagnetic order and bond-nematic phases in the spin-1/2 J1-J2-δ model.
- It reveals chiral magnon splitting and scaling of staggered magnetization, highlighting differences between conventional AFM and altermagnetic regimes.
- Results emphasize the interplay of symmetry breaking, quantum fluctuations, and topologically nontrivial orders with implications for spintronics applications.
Altermagnetism and Bond-Nematicity in the Spin-1/2 Square Lattice J1-J2-δ Model
Background and Motivation
This work analyzes the magnetic phases in the spin-1/2 square lattice J1-J2-δ model, with a focus on altermagnetic and magnetically disordered regimes. Altermagnetism, defined by non-collinear spin order not reducible to conventional AFM or FM phases, deviates from zero total magnetization due to combined action of time reversal and nontrivial point-group symmetries. Recent interest in altermagnets stems from their atypical band structure, chiral magnon splitting, and potential impact in spintronics, as well as their distinctive symmetry properties (2606.14101).
Despite substantial research on band structures and superconducting responses in altermagnetic materials, the effects of strong quantum fluctuations and frustration—especially the exotic phases resulting from order melting—remain underexplored. This manuscript offers an in-depth study of the J1-J2-δ model, employing symmetry-enhanced neural network quantum state (NQS) ansätze within a variational Monte Carlo (VMC) framework powered by group equivariant convolutional neural networks (GCNNs).
Model Specification and Computational Approach
The J1-J20-J21 Hamiltonian is
J22
where J23 modulates NNN exchange, mimicking inequivalent environments and geometric frustration. This model is experimentally relevant in iron oxychalcogenides and can be realized in ultracold atomic systems.
A NQS ansatz based on GCNNs captures both lattice and point-group symmetries (J24), as well as J25 spin parity. Parameter optimization leverages VMC combined with stochastic reconfiguration (SR) and natural gradient descent (NGD). The GCNN-VMC formalism efficiently models large lattices and treats quantum fluctuations non-perturbatively.
Altermagnetic Phase and Chiral Magnon Splitting
In the weak frustration regime (e.g., J26, J27), ground states correspond to the J28 irrep at J29 with spin parity δ0. Order is determined via finite-size scaling of staggered magnetization δ1, static spin structure factors, and energy extrapolation.
Altermagnetic order is evidenced by:
- Chiral splitting of magnon excitations: In NQS-VMC calculations, magnon modes with opposite chiralities (distinguished by δ2, δ3, δ4, δ5 irreps) exhibit energy splitting at δ6, absent in conventional AFM phases.
- Sharp structure factor peaks: δ7 and δ8 show definitive maxima at δ9 consistent with long-range order.
- Broken SU(2) symmetry: In VMC-NQS, symmetry breaking occurs even for finite clusters, mimicking thermodynamic limit behavior, with ground states generally not SU(2)-symmetric.
Energy scaling matches the expected J10 for AFM-type order. Staggered magnetization scales as J11, consistent with long-range order. Notably, J12 values are reduced compared to iPEPS due to NQS capturing stronger quantum renormalizations.
Magnon spectrum analysis reveals LSWT-consistent chiral splitting, with maximum separation at J13 and degeneracy at high-symmetry points in the thermodynamic limit.
Frustrated Regime: Bond-Nematic and SPT VBS Coexistence
In the strongly frustrated regime (J14, J15), quantum fluctuations melt altermagnetic order, yielding a magnetically disordered phase characterized by:
- Vanishing magnetic order: J16 scales critically as J17 with J18, confirming absence of long-range magnetization.
- Bond-nematicity and SPT VBS: Structure factors for nematic and dimer correlators exhibit broad, s-wave symmetry peaks at J19 and sharp VBS-associated peaks at J20 and J21. Nematic order parameters J22, J23 and J24 extrapolate to nonzero values in the thermodynamic limit, with scaling behavior indicative of a symmetry-protected topological (SPT) phase. Clusters with J25 and J26 demonstrate distinct order parameter scaling, hallmark of SPT VBS states.
- Magnon pair condensation: The ground state supports two-magnon bound pair condensation (operator J27), breaking U(1) spin rotation symmetry and generating nematic Goldstone modes.
- Broken J28 spin inversion symmetry: Finite J29 indicates spontaneous breaking of δ0 symmetry; energy level splitting for δ1 also confirms this.
Energy spectra validate coexistence, with nematic Goldstone modes corresponding to δ2, δ3 irreps (not fully symmetric under δ4) and degenerate ground state manifolds.
Implications and Theoretical Perspective
The identification of a magnetically disordered phase hosting both bond-nematicity and SPT VBS order advances understanding of quantum frustrated magnetism:
- Exotic phase formation: Demonstrates that melting altermagnetic order via frustration can yield nontrivial, symmetry-broken, topologically nontrivial quantum phases beyond spin liquids.
- Interplay of symmetry and topology: Coexistence of nematic and SPT VBS orders highlights complex symmetry fractionalization and topological sector separation, especially manifested in cluster-size dependent scaling.
- Low-energy excitation complexity: Emergence of nematic Goldstone and chiral triplon-like modes refines classification of excitations in these systems, relevant for both theoretical analysis and experimental probes.
- Quantum simulation and spintronics: The phases studied here have implications for simulating gravitational analogs and enhancing functionalities in spintronic devices utilizing altermagnets.
Future Directions
- Quantum criticality: Characterization of phase transitions, especially potential deconfined quantum critical points between Dirac spin liquid and bond-nematic/SPI VBS states.
- Superconductivity upon doping: Exploration of superconducting orders in the model via fermionic NQS approaches.
- Model extensions: Analysis of spin-1 versions and generalizations to other lattices and symmetry classes.
Conclusion
This study establishes that the spin-1/2 δ5-δ6-δ7 square lattice model supports altermagnetic order with chiral magnon splitting in weakly frustrated regimes and a magnetically disordered phase exhibiting coexisting bond-nematic and SPT VBS orders in the highly frustrated regime. The application of symmetry-adapted GCNN NQS within VMC defines a robust computational protocol for probing emergent quantum phases. These findings elucidate the nature of quantum melting in altermagnets and pave the way for future theoretical and experimental investigations into nontrivial quantum order and criticality in frustrated magnets (2606.14101).