Lattice gas
A lattice gas is a statistical-mechanical model of particles on a lattice, where each site is either empty or occupied. It can be formulated as a lattice spin system with spin space and is closely related to the Ising model .
Let be finite. A configuration is with , where means “occupied.” A common nearest-neighbor attractive lattice gas Hamiltonian is
where is the chemical potential (an analogue of an external field ) and makes occupied neighbors energetically favorable. More general lattice gases can be defined using an interaction potential beyond nearest neighbors.
At inverse temperature , the finite-volume Gibbs distribution is the finite-volume Gibbs measure
normalized by the partition function . The associated pressure is typically defined from (and its infinite-volume limit by thermodynamic limit of the pressure ).
Key properties
Density as the main observable. The natural order parameter is the particle density
or its infinite-volume expectation. Phase coexistence appears as multiple possible limiting densities at the same .
Equivalence to Ising in a field. Define Ising spins by . For nearest-neighbor interactions on (degree ), one obtains (up to an additive constant)
Thus an attractive lattice gas is equivalent to a ferromagnetic Ising model with an effective field, making lattice gases a convenient physical reinterpretation of Ising phase transitions .
Gas–liquid transition and coexistence line. In the Ising correspondence, the “coexistence” regime (two densities) corresponds to the regime with multiple infinite-volume Gibbs states. The symmetry point (Ising zero field) corresponds to a particular chemical potential.
Boundary conditions and phase separation. Different boundary conditions can select low-density (“gas”) or high-density (“liquid”) phases in finite volume, and interfaces/domain walls appear in mixed boundary conditions.
Physical interpretation
The lattice gas is a coarse-grained model of a fluid on discrete sites, useful for:
- gas–liquid coexistence and criticality,
- adsorption in porous materials (occupied vs empty sites),
- phase separation and interfaces.
Because of its exact mapping to the Ising model, it provides a direct bridge between magnetic language (spins, fields, magnetization) and fluid language (occupations, chemical potential, density).