Haldane gap for the spin-1 Heisenberg chain

Haldane predicted in 1983 that antiferromagnetic Heisenberg chains behave qualitatively differently for integer and half-integer spin: integer-spin chains have a unique ground state separated by a spectral gap (the Haldane gap) with exponentially decaying correlations, while half-integer chains are gapless. For $S=1$ the gap is measured numerically at $\Delta\approx0.4105\,J$ and confirmed experimentally in quasi-1D magnets.

What is known: For half-integer spin the Lieb–Schultz–Mattis theorem (extended by Affleck–Lieb, and in higher dimensions by Hastings) rigorously forbids a unique gapped ground state, consistent with gaplessness. On the integer side, the AKLT model — a nearby, specially tuned Hamiltonian — was proved by Affleck, Kennedy, Lieb and Tasaki to have a unique gapped ground state with a valence-bond-solid structure and hidden (string) order; this is the paradigm of a symmetry-protected topological phase. Den Nijs–Rommelse and Kennedy–Tasaki clarified the hidden $\mathbb{Z}_2\times\mathbb{Z}_2$ order and its detection.

The gap of the pure Heisenberg $S=1$ chain remains the target: the AKLT theorem concerns a different Hamiltonian and does not by itself establish a uniformly gapped interpolation. Periodic or bulk-gap conventions avoid confusing the question with low-energy boundary spins of open chains. Higher integer spin is a related extension, not an additional condition for solving this spin-1 benchmark. The LSM theorem allows degeneracy as an alternative to gaplessness in its general setting.

Source catalogue review: 4 September 2026. Improved indicates partial progress, including qualified preprint claims.