Exactly Solvable Models Hosting Altermagnetic Quantum Spin Liquids

João Augusto Sobral, Pietro M. Bonetti, Subrata Mandal, Mathias S. Scheurer · 2025

Read Exactly Solvable Models Hosting Altermagnetic Quantum Spin Liquids by João Augusto Sobral, Pietro M. Bonetti, Subrata Mandal, Mathias S. Scheurer (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-24076.txt.

Introduction

João Augusto Sobral, Pietro Bonetti, Subrata Mandal, and Mathias Scheurer build exactly solvable models of a recently named phase of matter: the altermagnetic quantum spin liquid. Spin liquids are magnetic states where spins never settle into order, yet this variety breaks time-reversal symmetry like an altermagnet while its net magnetic moment cancels out by symmetry.

Working with Majorana-based methods on two lattice designs, a square-octagon lattice and a checkerboard, the team studies spin three-halves and spin seven-halves versions. The smaller-spin model has a single ground state, a g-wave altermagnetic spin liquid. The larger-spin model is richer, with several chiral spin liquids competing against a d-wave variant. The authors also catalog the excitations of these phases, sorting which carry topological character and which do not, and showing how symmetry and topology combine in this new state of matter.

Catalog record

CREATORJoão Augusto Sobral, Pietro M. Bonetti, Subrata Mandal, Mathias S. Scheurer
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE100 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2512-24076

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Exactly Solvable Models Hosting Altermagnetic Quantum Spin Liquids João Augusto Sobral, Pietro M. Bonetti, Subrata Mandal, Mathias S. Scheurer Abstract We construct spin-3/2 and spin-7/2 models on the square-octagon and checkerboard lattices that are exactly solvable with Majorana representations. They give rise to spin-liquid phases with full spin-rotation and lattice-translational symmetries but broken time-reversal symmetry. Although non-zero on elementary plaquettes, the net orbital magnetic moment is guaranteed to vanish as a result of point symmetries; due to the analogy to long-range ordered altermagnets, these types of phases were dubbed altermagnetic spin liquids in [ Phys. Rev. Research 7 , 023152 ]. For the spin-3/2 model, we find that a g g -wave altermagnetic spin liquid emerges as the unique ground state. In contrast, the spin-7/2 model exhibits a significantly richer phase diagram, involving different types of chiral spin liquids competing with a d d -wave altermagnetic spin liquid. Finally, we identify and characterize the topological and non-topological excitations, illustrating the rich physics of altermagnetic spin liquids resulting from the interplay of non-trivial topological and symmetry aspects of this novel phase of matter. I Introduction Altermagnetism refers to a class of magnetic states with vanishing net magnetization (unlike ferromagnets) which is enforced not as a result of translational symmetry combined with time-reversal (like in anti-ferromagnets) but follows from a combination of the latter with a point-group operation 23 . Although initially driven by their relevance to spintronics, altermagnets have also become subject of intense research in the field of strongly correlated physics and superconductivity 9 ; 12 . For instance, the fate of these long-range ordered altermagnetic phases in regimes dominated by strong quantum fluctuations—particularly in proximity to, or within, quantum spin liquids (QSLs) 21 ; 33 —has been recently explored in distinct scenarios 24 ; 26 ; 15 . Ref. 24 showed that a variety of exotic QSLs can descend from long-range ordered parent altermagnets, with some retaining the same (alter-)magnetic point-group symmetries in experimentally accessible observables, in both metallic and insulating regimes. Notably, the electronic spectral functions of the doped system can still exhibit the characteristic altermagnetic momentum-dependent splitting, even though the topologically ordered phase maintains full SU(2) spin-rotational invariance. Among the identified phases, the “orbital altermagnet” (see also 31 ; 20 ; 3 ; 16 for a discussion of orbital altermagnetism in different contexts) exhibits a pattern of orbital currents arranged in an altermagnetic fashion. The descendant QSL inherits a corresponding structure in its emergent gauge-flux configuration and scalar spin chiralities, leading to its designation as an orbital altermagnetic quantum spin liquid (OAMSL). Starting from a Kitaev-type 10 gapless \mathbb{Z}_{2} QSL defined on a bilayer square–octagon lattice, Ref. 26 showed that increasing the strength of additional interactions that commute with both intra- and interlayer flux operators of the original model drives a first-order transition into a \boldsymbol{q}=0 ordered state. In this phase, a finite local symmetry-breaking order parameter coexists with remnant global \mathbb{Z}_{2} topological order inherited from the parent gapless QSL—an example of magnetic fragmentation 27 ; 2 ; 19 . This coexistence motivated the designation of the phase as a \mathbb{Z}_{2} topological altermagnet. The bands of the fractionalized excitations exhibit a d -wave momentum-dependent spin-splitting protected by the combination of a diagonal mirror symmetry and time reversal. A related phenomenon was identified in an exactly solvable \mathbb{Z}_{2} quantum-spin(-orbital) liquid on the square lattice 15 . Small perturbations (compared to the flux gap) which obey altermagnetic projective symmetry requirements, and that do not break the integrability of the model, lead to an effective low-energy theory for fermionic partons whose particle–hole asymmetry generates momentum-dependent band splittings that are analogous to altermagnets. These were dubbed fractionalized altermagnets, akin to fractionalized antiferromagnets 14 ; 22 ; 8 , and were classified according to their projective symmetry group 28 . Taken together, these advances reveal that altermagnetic symmetry breaking can appear in QSLs either as a descendant of an ordered parent phase with fully preserved spin-rotation symmetry and potentially additional altermagnetic scalar spin chiralities 24 or through coexistence of an altermagnetic spin texture with remnant \mathbb{Z}_{2} topological order 26 ; 15 . Motivated by these developments, and building on the Kitaev 10 and Yao–Kivelson 30 constructions, we introduce two exactly solvable models that intrinsically host OAMSLs in the sense defined in Ref. 24 . The ground state of the spin-3/2 model is a g -wave OAMSL, in sharp contrast with the more complex phase diagram arising for the spin-7/2 model. We further address the different types of excitations in the flux sector. II Spin-3/2 model As already outlined above, our goal is to construct an exactly solvable model hosting an OAMSL, which is characterized by the following properties: while spin-rotation invariance is preserved, time-reversal symmetry is spontaneously broken. However, as opposed to a chiral spin liquid (which can be thought of as the ferromagnetic analog of the OAMSL), there are still residual magnetic point group symmetries that guarantee that the associated net (orbital) magnetic moment vanishes. Similarly to the difference between an antiferromagnet and an altermagnet, we emphasize that the vanishing of the moment is not the consequence of the product of time-reversal and translation; the OAMSLs we will discuss here all preserve the translational symmetry of the underlying lattice. In the spirit of the Kitaev 10 and Yao–Kivelson 30 constructions, the coordination number of the lattice determines the “length” S of the spin operators on every site. Furthermore, to break time-reversal symmetry, we require elementary loops with an odd number of sites 29 ; 5 . Focusing on tetragonal two-dimensional systems, as in Ref. 24 , the checkerboard lattice seems like a natural starting point to ensure the second constraint. However, the coordination number is six, which requires S=7/2 . While we will discuss such a model in Sec. III below, we will first reduce the coordination number to five by working on the square-octagon lattice (see Fig. 1 a). In this case, S=3/2 suffices as its four-dimensional onsite Hilbert space allows for five \Gamma^{a} matrices which obey the Clifford algebra \{\Gamma^{a},\Gamma^{b}\}=2\delta^{ab}\mathbb{I}_{4} . The Hamiltonian is defined by \begin{gathered}{\cal H}_{3/2}=\,\sum_{\boldsymbol{R}}\left[J_{1}\left(\Gamma^{3}_{\boldsymbol{R},A}\Gamma^{1}_{\boldsymbol{R},D}+\Gamma^{4}_{\boldsymbol{R},B}\Gamma^{2}_{\boldsymbol{R},A}+\Gamma^{2}_{\boldsymbol{R},D}\Gamma^{4}_{\boldsymbol{R},C}+\Gamma^{1}_{\boldsymbol{R},C}\Gamma^{3}_{\boldsymbol{R},B}\right)+J_{1}^{\prime}\left(\Gamma^{5}_{\boldsymbol{R},D}\Gamma^{5}_{\boldsymbol{R}_{x},B}+\Gamma^{5}_{\boldsymbol{R}_{y},C}\Gamma^{5}_{\boldsymbol{R},A}\right)\right]+\\ +\sum_{\boldsymbol{R}}J_{2}\left(\Gamma^{3}_{\boldsymbol{R}_{y},C}\Gamma^{1}_{\boldsymbol{R}_{x},A}+\Gamma^{4}_{\boldsymbol{R},A}\Gamma^{2}_{\boldsymbol{R}_{x+y},C}+\Gamma^{4}_{\boldsymbol{R},D}\Gamma^{2}_{\boldsymbol{R}_{x+y},B}+\Gamma^{1}_{\boldsymbol{R}_{x},B}\Gamma^{3}_{\boldsymbol{R}_{y},D}\right),\end{gathered} (1) with nearest-neighbor couplings J_{1} along the edges of the square , J_{1}^{\prime} along the horizontal and vertical edges of the octagon , and second-neighbor interactions J_{2} inside the octagons. Here, \boldsymbol{R} labels the Bravais lattice sites (gray square in Fig. 1 a) and the shorthands \boldsymbol{R}_{\delta}:=\boldsymbol{R}+\boldsymbol{e}_{\delta} , \delta=x,y , and \boldsymbol{R}_{x+y}:=\boldsymbol{R}+\boldsymbol{e}_{x}+\boldsymbol{e}_{y} are used to refer to neighboring unit cells. The second subscript of the \Gamma -matrices refers to the four sublattices \alpha=A,B,C,D in each unit cell. There are many different ways of representing the \Gamma^{a} matrices acting on the four-dimensional Hilbert space on each site (suppressing the site index here). For instance, they can be expressed in terms of two spin- 1/2 or two qubits per site. Denoting the Pauli matrices by \sigma^{j} , one possible representation is given by \begin{gathered}\Gamma^{1}=\sigma^{z}\otimes\sigma^{y},\,\,\Gamma^{2}=\sigma^{z}\otimes\sigma^{x},\,\Gamma^{3}=\sigma^{y}\otimes\mathbb{I}_{2},\\ \Gamma^{4}=\sigma^{x}\otimes\mathbb{I}_{2},\,\Gamma^{5}=\sigma^{z}\otimes\sigma^{z}.\end{gathered} (2) Alternatively, the \Gamma^{a} matrices can also be expressed as spin-3/2 operators S^{j} (see Appendix B ), such that Eq. ( 1 ) can be reinterpreted as a Hamiltonian with interacting quadrupole moments 6 ; 13 . The point group of the lattice is C_{4v} , which is generated by 90 ∘ anti-clockwise rotations ( C_{4z} )

Read the complete text