Math 2.0

Statistical mechanics, structure of matter Foundations of equilibrium statistical mechanics Theorem

Gibbs–Bogoliubov inequality

Builds on Canonical ensembleJensen's inequalityExpected value

Statement

Let (Ω,ν)(\Omega, \nu) be a σ\sigma-finite measure space and let β>0\beta > 0. Let H,H0 ⁣:Ω→ℝH, H_0 \colon \Omega \to \R be measurable with Z0=∫Ωe−βH0 dν∈(0,∞)Z_0 = \int_\Omega e^{-\beta H_0} \, d\nu \in (0, \infty) and Z=∫Ωe−βH dν<∞Z = \int_\Omega e^{-\beta H} \, d\nu < \infty. Let ⟨⋅⟩0\langle \cdot \rangle_0 be the expected value under the canonical ensemble e−βH0 dν/Z0e^{-\beta H_0} \, d\nu / Z_0, and assume ⟨∣H−H0∣⟩0<∞\langle \abs{H - H_0} \rangle_0 < \infty. Let F=−β−1ln⁡ZF = -\beta^{-1} \ln Z and F0=−β−1ln⁡Z0F_0 = -\beta^{-1} \ln Z_0 be the free energies. Then F≤F0+⟨H−H0⟩0,F \le F_0 + \langle H - H_0 \rangle_0, with equality if and only if H−H0H - H_0 is constant almost everywhere for the measure e−βH0dνe^{-\beta H_0} d\nu.

Remarks

Proof idea: Z/Z0=⟨e−β(H−H0)⟩0Z / Z_0 = \langle e^{-\beta (H - H_0)} \rangle_0, and Jensen's inequality for the convex function e−βxe^{-\beta x} gives Z/Z0≥e−β⟨H−H0⟩0Z / Z_0 \ge e^{-\beta \langle H - H_0 \rangle_0}. Taking −β−1ln⁡-\beta^{-1} \ln of both sides gives the claim. The same inequality holds for Hermitian matrices, with Z=tr⁡e−βHZ = \tr e^{-\beta H} and ⟨A⟩0=tr⁡(e−βH0A)/Z0\langle A \rangle_0 = \tr(e^{-\beta H_0} A) / Z_0; this is the Peierls–Bogoliubov inequality. Example: in mean-field theory one takes H0H_0 to be a sum of independent single-site terms with free parameters. Minimizing the right side over these parameters gives the best such upper bound on FF. It gives a variational principle: F=min⁡H0(F0+⟨H−H0⟩0)F = \min_{H_0} \left( F_0 + \langle H - H_0 \rangle_0 \right), attained at H0=HH_0 = H.

Going deeper

For a student

This inequality is the rigorous basis of mean-field and variational approximations, such as the Curie–Weiss treatment of the Ising model. It is equivalent to the nonnegativity of the Kullback–Leibler divergence between the trial ensemble and the true canonical ensemble. It is also the finite-volume form of the Gibbs variational principle, which characterizes Gibbs measures as maximizers of entropy minus energy.

Reading path

To understand this page from the ground up, read these 5 pages in order. The foundations come first.

Show the path as a graph
Each page uses the pages above it. The colour of the bar on a box shows how strong the evidence is, and the box also gives it as text. The list has the same pages.
  • no evidence level
  1. Canonical ensemble concept
  2. Convex function definition
  3. Expected value concept
  4. Jensen's inequality statement
  5. Gibbs–Bogoliubov inequality statement

Further reading

  1. Web https://en.wikipedia.org/wiki/Helmholtz_free_energy
About this page
TagWR3RBEH4QQ
URIm2://WR3RBEH4QQ
Kindstatement, theorem
WritingW0 stub
MSC82B05, 82B10, 82B03
StatusThis copy of the site has no build data. Run m2 status WR3RBEH4QQ.
AboutCanonical ensemble
UsesCanonical ensemble, Jensen's inequality, Expected value
Changed3 November 2026

Markdown · JSON · Edit on GitHub

What this page needs for W1

  • an inventory row: the row of the inventory that the node accounts for (0 of 1)
  • a source with a locator (0 of 1)

The writing levels say how complete the text of a page is. They do not measure its truth.

What links here

No page links to this page yet.