Gapped Spin Liquids and Positive Diagnostics

A gapped quantum spin liquid is a symmetry-preserving, long-range-entangled phase with deconfined spin-carrying excitations and topological sectors. The absence of a magnetic Bragg peak is necessary but not sufficient: a valence-bond solid, trivial paramagnet, disorder, or a finite-size singlet can also lack magnetic order.

Required background. Symmetry fractionalization supplies spinon and vison quantum numbers; emergent Z2 gauge fields supplies the gapped deconfined benchmark.

Helpful background. Tensor-network entanglement ansätze supplies finite-cylinder and virtual-symmetry diagnostics.

Positive data for a gapped spin liquid

A coherent identification specifies:

1. a nonzero singlet and spin gap after two-dimensional finite-size scaling;
2. no magnetic or valence-bond order in structure factors and correlation-length scaling;
3. topology-dependent sectors with exponentially small splitting;
4. spinon and vison excitations, including mutual statistics or equivalent string algebra;
5. symmetry fractionalization compatible with the microscopic spin per unit cell;
6. long-range-entanglement data such as minimally entangled states or $\gamma=\ln\mathcal D$.

These tests are linked. For a Z2 liquid, four torus sectors, total quantum dimension $\mathcal D=2$, and $e$–$m$ mutual phase $\pi$ must describe one phase. A numerical continuum without a sector or excitation interpretation is weaker.

Valence bonds versus a liquid

A valence-bond solid breaks lattice symmetry and has a local bond-energy order parameter. On finite clusters its symmetry-related states form a low-energy tower that can resemble topological degeneracy. Their momenta, splitting with volume, response to boundary pinning, and dimer structure factor distinguish them from topology sectors. A resonating-valence-bond wavefunction can represent either a spin liquid or a valence-bond solid after projection; “resonating” is not a phase label.

For a local finite-range spin Hamiltonian with primitive-cell translation and unbroken SO(3) spin rotation, half-odd-integer spin per primitive cell triggers a Lieb–Schultz–Mattis-type obstruction to a unique, fully gapped, symmetry-preserving short-range-entangled ground state. A symmetric gapped phase must have topological order, while the alternatives include gaplessness or breaking translation or spin symmetry Hastings 2004. Variants protected by U(1), time reversal, or other internal symmetry require their own hypotheses; the theorem narrows possibilities but does not identify which spin liquid occurs.

Numerical control

On cylinders, compare several circumferences, orientations, bond dimensions, and initialized sectors. Extrapolate energy variance and correlation length before the two-dimensional limit. Boundary conditions can pin a valence-bond pattern or select a topological sector. Exact diagonalization needs shape/aspect-ratio comparisons; a gap on one cluster is not a thermodynamic gap.

The kagome Heisenberg model illustrates both the power and limits of such inference: large-scale DMRG found a gapped-liquid candidate Yan, Huse, and White 2011, while determining the precise infrared order requires continuing cross-method and size tests. The durable lesson is the multi-diagnostic method, not a frozen platform verdict.

Exercise

Why can four nearly degenerate singlets on a torus fail to establish Z2 order?

Solution

A fourfold valence-bond crystal can also produce four symmetry-related singlets. One must compare their momenta and symmetry quantum numbers, dimer structure factor, boundary pinning, topology-sector response, vison/spinon data, and size scaling.

References

• Matthew B. Hastings, “Lieb–Schultz–Mattis in Higher Dimensions,” Physical Review B 69 (2004) 104431, doi:10.1103/PhysRevB.69.104431.
• Simeng Yan, David A. Huse, and Steven R. White, “Spin-Liquid Ground State of the S=1/2 Kagome Heisenberg Antiferromagnet,” Science 332 (2011) 1173–1176, doi:10.1126/science.1201080.