Valence-Bond Solids and Quantum Paramagnets

A quantum paramagnet has no magnetic dipole order at zero temperature, but that absence does not determine the phase. A valence-bond solid (VBS) breaks lattice symmetry through an ordered pattern of singlets; a featureless paramagnet preserves symmetry; an SPT paramagnet has protected boundary structure; and a spin liquid has intrinsic topological order or fractionalization. These alternatives require positive diagnostics, not merely a small ordered moment.

Required background. Frustration and Order by Disorder supplies competing singlet and ordered states. Helpful background. Spin Liquids and Valence Bonds supplies fractionalization and emergent-gauge alternatives.

Singlets and valence-bond order

For two spin-$1/2$ moments, the singlet

$$\lvert s_{ij}\rangle=\frac1{\sqrt2} (|\uparrow_i\downarrow_j\rangle-|\downarrow_i\uparrow_j\rangle)$$

has $\langle\mathbf S_i\cdot\mathbf S_j\rangle=-3/4$. A product of singlets on selected bonds is magnetically invariant but can distinguish bonds or unit cells. A columnar VBS on a square lattice, for example, breaks translations and rotations and therefore has several symmetry-related ground states in the thermodynamic limit.

A local dimer operator is

$$B_{i,\boldsymbol\delta}=\mathbf S_i\cdot\mathbf S_{i+\boldsymbol\delta}.$$

VBS order is established by long-range correlations of the appropriate Fourier component of $B$, a nonzero order parameter after the thermodynamic limit, and the expected finite-size pattern of quasi-degenerate states. A gap and short-ranged spin correlations alone do not establish lattice-symmetry breaking. Large-$N$ antiferromagnets connect Berry phases of monopoles to VBS selection Read and Sachdev 1989, pp. 1694–1697.

Quantum paramagnets and microscopic constraints

A featureless quantum paramagnet is short-range entangled, symmetric, and unique on a closed manifold. Whether it is allowed depends on the representation per unit cell. With a half-odd-integer spin per primitive cell, translation and spin rotation impose a Lieb–Schultz–Mattis–Oshikawa–Hastings obstruction: a symmetric gapped phase must either have ground-state degeneracy/topological order or otherwise evade an assumption. Oshikawa’s flux-insertion argument exposes the filling constraint Oshikawa 2000, pp. 1535–1538.

With an integer representation per unit cell, a symmetric product state may be allowed, but a given Hamiltonian need not realize it. An SPT paramagnet is also short-range entangled in the bulk yet cannot be connected to a trivial product state while its protecting symmetry and gap are preserved. Its diagnostics include projective boundary representations, symmetry action on entanglement states, and quantized responses—not bulk topological degeneracy.

A topologically ordered spin liquid instead supports long-range entanglement, fractional quasiparticles, and characteristic ground-state sectors on nontrivial spatial topology. A broad continuum in a spin response is compatible with such fractionalization but is not sufficient, because multiparticle magnons, disorder, or phonons can also produce continua.

Transitions and finite-size evidence

An ordinary Landau transition between a Néel state and a VBS would involve unrelated order parameters and generically allow coexistence or a first-order boundary. Deconfined-critical scenarios instead describe a continuous transition in terms of fractional spinons and an emergent gauge field, with monopoles controlling lattice symmetry. This is a sharp field-theory proposal, not a label to infer from an apparently smooth finite-size crossover.

Numerical discrimination should combine magnetic and dimer correlation ratios, stiffness, excitation gaps, histogram structure, entanglement information where controlled, and drift with system size. A low singlet gap can be a VBS tower of states, a critical mode, or topological sector splitting; its momentum and scaling distinguish them. Boundary conditions can pin one VBS pattern and conceal the degeneracy, so the boundary field must be included in interpretation.

Validity boundaries

Nearest-neighbor valence-bond pictures provide useful variational states but are not an orthogonal basis and need not capture longer bonds. A measured absence of static magnetism is negative evidence only. Establishing a phase requires the positive signatures appropriate to symmetry breaking, SPT order, or intrinsic topological order, together with finite-size and disorder controls.

Exercises

1. Compute $\langle s_{ij}|\mathbf S_i\cdot\mathbf S_j|s_{ij}\rangle$.

Solution

Use $\mathbf S_i\cdot\mathbf S_j=[\mathbf S_{ij}^2-\mathbf S_i^2-\mathbf S_j^2]/2$. The singlet has total spin zero and each constituent has $S(S+1)=3/4$, giving $(0-3/4-3/4)/2=-3/4$.

2. Why does a columnar VBS on an even periodic square lattice have more than one thermodynamic ground state?

Solution

Translations and $90^\circ$ rotations map a columnar pattern to distinct patterns of equal energy. A finite system forms symmetry eigenstate superpositions with exponentially small splittings; in the thermodynamic limit the patterns become distinct symmetry-broken ground states.

References

• Oshikawa, M. “Commensurability, Excitation Gap, and Topology in Quantum Many-Particle Systems on a Periodic Lattice.” Physical Review Letters 84 (2000): 1535–1538. DOI.
• Read, N., and S. Sachdev. “Valence-Bond and Spin-Peierls Ground States of Low-Dimensional Quantum Antiferromagnets.” Physical Review Letters 62 (1989): 1694–1697. DOI.