Pirogov–Sinai theory (low-temperature phase diagrams with multiple ground states)

A contour-based framework proving existence and stability of multiple Gibbs phases and first-order transitions for low-temperature lattice systems with finitely many competing ground states.
Pirogov–Sinai theory (low-temperature phase diagrams with multiple ground states)

Context

Many lattice models (see ) have several distinct ground states at zero temperature. Pirogov–Sinai theory provides a systematic way to analyze the corresponding low-temperature equilibrium structure: existence of multiple , stability under perturbations, and precise first-order transition lines.

It builds on contour representations and convergent expansions (see ).

Theorem (Pirogov–Sinai; standard qualitative form)

Consider a finite-range lattice spin system with:

  • a finite set of periodic ground states G={g1,,gm}\mathcal{G}=\{g_1,\dots,g_m\} at zero temperature,
  • a Peierls-type condition: excitations above each ground state have energy cost proportional to the size of their boundary (contour cost).

Then there exists β0<\beta_0<\infty such that for all ββ0\beta\ge \beta_0:

  1. Phase coexistence and pure phases: for each stable ground state gkg_k there exists a corresponding translation-invariant pure infinite-volume Gibbs state μk\mu_k solving the .

  2. Uniqueness away from coexistence: in regions of parameter space where a single ground state is stable, the infinite-volume Gibbs measure is unique and correlations decay exponentially (finite ).

  3. First-order transitions: coexistence surfaces (where two or more ground states have equal free-energy competition) correspond to first-order transitions: multiple Gibbs states exist and macroscopic observables (e.g. an ) exhibit discontinuities across the surface.

  4. Interfaces and surface tension: at coexistence, mixed boundary conditions generate stable interfaces with positive , giving control of droplet formation and providing a rigorous foundation for at low temperature.

What you can do with it

  • Prove low-temperature phase diagrams for models like Potts-type systems, Ising models with fields or competing interactions, lattice gases, and small perturbations of exactly solvable ground-state structures.
  • Establish sharp statements about stable phases, coexistence lines, and interface costs in terms of contour weights and convergent expansions.