| Heisenberg model (magnetism) | |
|---|---|
| Name | Heisenberg model |
| Caption | Schematic of interacting spins on a lattice |
| Domain | Condensed matter physics |
| Introduced | 1928 |
| Inventor | Werner Heisenberg |
| Notable applications | magnetism, quantum phase transition, spintronics |
Heisenberg model (magnetism)
The Heisenberg model (magnetism) is a quantum mechanical model of interacting localized spins on a lattice used to describe magnetic ordering and collective spin excitations. Introduced in the late 1920s, it underpins modern theories of ferromagnetism and antiferromagnetism and serves as a paradigmatic model in condensed matter physics and statistical mechanics for studying quantum many-body phenomena.
The model was proposed by Werner Heisenberg in 1928 to explain the quantum origin of ferromagnetism by means of an exchange interaction arising from the Pauli exclusion principle and Coulomb interaction in itinerant electron systems. It built upon earlier ideas by Pierre Curie and Wolfgang Pauli and was contemporaneous with the development of quantum mechanics. The Heisenberg model connected microscopic electronic structure to macroscopic magnetic order and informed later developments such as the Hubbard model, the Ising model, and Anderson superexchange theory.
The canonical Heisenberg Hamiltonian for spins on sites i and j is H = -Σ_{⟨i,j⟩} J_{ij} S_i · S_j, where J_{ij} is the exchange coupling and S_i are quantum spin operators. Variants include the XXX model (isotropic), the anisotropic XXZ model, and the XY model (planar limit). Extensions introduce further-neighbor couplings (J1–J2 models), Dzyaloshinskii–Moriya interaction (DM interaction) which breaks inversion symmetry, single-ion anisotropy, and coupling to external magnetic fields (Zeeman term). The model is defined on lattices such as the one-dimensional chain, square lattice, triangular lattice, honeycomb lattice, and Kagome lattice, each yielding distinct magnetic behavior.
Spin operators S_x, S_y, S_z obey the SU(2) Lie algebra [S_α, S_β] = iħ ε_{αβγ} S_γ. The model is formulated using representations of spin-S (commonly S=1/2, 1, 3/2, ...). For S=1/2 the local Hilbert space is two-dimensional and the operators are proportional to Pauli matrices. Bosonic representations such as Holstein–Primakoff transformation and Schwinger boson representation map spins to bosons for spin-wave analysis; fermionic representations include Jordan–Wigner transformation in one dimension and Abrikosov fermion methods for mean-field theories. Symmetry considerations involve global SU(2), or reduced U(1)/Z2 symmetries in anisotropic variants.
Exact solutions exist in selected cases: the one-dimensional S=1/2 Heisenberg chain is integrable via the Bethe ansatz solved by Hans Bethe (1931), yielding exact ground state and excitation spectra. The Lieb–Mattis theorem constrains ordering in bipartite lattices. The Majumdar–Ghosh model (J1–J2 chain at special ratio) admits an exact dimerized ground state. The Haldane conjecture, supported by analytic and numerical work, distinguishes integer and half-integer spin chains; the S=1 chain has a gapped ground state (the Haldane gap) with connections to the Affleck–Kennedy–Lieb–Tasaki (AKLT) model.
Heisenberg models display phases including ferromagnetic, antiferromagnetic, spin liquid, valence-bond solid, and chiral orders depending on lattice, frustration, and dimensionality. Low-energy excitations in ordered phases are magnons (spin waves) described by linear spin wave theory; in one dimension, excitations can be fractionalized into spinons. Frustrated geometries such as the Kagome lattice and triangular lattice foster quantum spin liquid behavior with emergent gauge fields and topological order. Critical behavior near quantum phase transitions is analyzed using renormalization group methods and field theories such as the nonlinear sigma model.
Because generic Heisenberg models are nonintegrable, a variety of computational methods are used: exact diagonalization for small clusters, density matrix renormalization group (DMRG) for one-dimensional and quasi-one-dimensional systems, quantum Monte Carlo (QMC) for sign-problem-free cases, and tensor network states including matrix product states and projected entangled pair states (PEPS). Mean-field approaches include spin-wave theory, Schwinger-boson mean-field, and slave-particle techniques. Numerical studies often rely on high-performance computing at institutions like Argonne National Laboratory and Max Planck Institute for the Physics of Complex Systems.
Real materials approximating Heisenberg spin models include magnetic insulators such as KCuF3, copper oxides, and MnF2; low-dimensional compounds like Sr2CuO3 realize one-dimensional S=1/2 chains. Cold-atom experiments in optical lattices and trapped-ion simulators implement tunable Heisenberg interactions for quantum simulation, demonstrated in platforms developed at groups such as Institute for Quantum Optics and Quantum Information and MIT. Applications span interpretation of neutron scattering experiments (e.g., at Oak Ridge National Laboratory), design of spintronics devices, and foundational studies of quantum entanglement and thermalization in many-body physics.
Category:Quantum magnetism Category:Statistical mechanics Category:Condensed matter physics