Spin space
A spin space is the set of allowed values for a single spin variable at one lattice site. In a lattice model on the integer lattice (see integer lattice ), each site carries a spin .
For probabilistic and Gibbsian formulations, one typically equips with a -algebra (see sigma-algebra ) so that local observables are measurable functions of . Often one also fixes an a priori measure on (see measure ), which is counting measure for discrete spins and a reference (e.g. Haar/surface) measure for continuous spins.
The spin space determines what a spin configuration is and underlies the full configuration space .
Common examples
- Ising spins: (see Ising model ).
- Potts spins: (see Potts model ).
- Lattice gas/occupancy: (see lattice gas ).
- XY spins: is the unit circle (see XY model ).
- Heisenberg spins: is a unit sphere in (see Heisenberg model ).
Key properties
- Discrete vs continuous: If is finite/countable, sums replace integrals in the partition function ; if is continuous, the choice of a priori measure is part of the model.
- Compactness/topology (often used): When is finite or compact, the product configuration space can inherit useful compactness properties (see configuration space ).
- Symmetries: Many models have a symmetry group acting on (e.g. spin-flip for Ising, rotations for XY/Heisenberg). Symmetries on induce symmetries of the Hamiltonian and Gibbs measures, relevant for spontaneous symmetry breaking .
Physical interpretation
The spin space represents the local microscopic degrees of freedom:
- orientations of a magnetic moment (continuous ),
- discrete “up/down” magnetic moments (Ising),
- internal states/colors (Potts),
- presence/absence of a particle (lattice gas).
Changing the choice of changes the nature of fluctuations and possible ordered phases (e.g. continuous symmetries vs discrete symmetries), influencing whether and how phase transitions can occur.