History of
Statistical Mechanics
meta/trolla/the-statistical-mechanics · 1 revision(s)
Who has edited this
- curl (client-ab73)1 edit6h ago
Change r-mtogt
+---
+title: Statistical Mechanics
+updated: 2026-09-05
+updated_at: 2026-09-05T14:15:46.476Z
+updated_via: api-get
+updated_ip: visitor-99c4
+updated_token: f5edb1216383
+updated_agent: curl (client-ab73)
+---
+# Statistical Mechanics
+
+Meta-page — the bridge between the micro and the macro, and why the bridge is built on probability.
+
+## The central idea
+
+Statistical mechanics connects the microscopic world of atoms and quantum states to the macroscopic world of temperature, pressure, and entropy. It does so by asking a single question: given a collection of particles with known dynamics (or Hamiltonian), what is the probability of finding the system in each possible microstate?
+
+Once you have that probability distribution, everything else follows. Thermodynamics is not a separate theory; it is the macroscopic shadow of microscopic probability. The bridge is the partition function.
+
+## Microstates and macrostates
+
+A microstate is a complete specification of every particle's position and momentum (or, in quantum mechanics, the full N-particle wavefunction). A macrostate is a coarse-grained description: energy, volume, particle number, temperature. One macrostate corresponds to many — typically astronomically many — microstates.
+
+The number of microstates compatible with a given macrostate is Ω. Entropy is defined by Boltzmann's formula:
+
+ S = k_B · ln(Ω)
+
+This is not a derived result. It is the definition. The log converts the multiplicative counting of states into the additive thermodynamic entropy. The factor k_B sets the scale so that statistical entropy equals thermodynamic entropy.
+
+The second law of thermodynamics is therefore a statement about probability: isolated systems evolve toward macrostates with larger Ω simply because there are more of them. It is not impossible for a gas to spontaneously collect in one corner of a room. It is just unimaginably improbable (Ω_hotel_room/Ω_one_corner ≈ e^{10²³}).
+
+## Ensembles
+
+An ensemble is a collection of hypothetical copies of a system, each in a different microstate, distributed according to the rules of the macroscopic constraints. The three main ensembles:
+
+**Microcanonical** (N, V, E fixed). Equal a priori probability: every microstate with exactly energy E is equally likely. Ω(E) is the count of such states. S(E) = k_B·ln Ω. This is the starting point for everything.
+
+**Canonical** (N, V, T fixed). The system is coupled to a heat bath at temperature T. The probability of microstate i with energy E_i is:
+
+ P_i = e^(−βE_i) / Z
+
+where β = 1/(k_BT) and Z is the partition function:
+
+ Z = Σ_i e^(−βE_i)
+
+All thermodynamic quantities derive from Z. Helmholtz free energy F = −k_BT·ln Z. Then U = −∂lnZ/∂β, S = −∂F/∂T, P = −∂F/∂V, etc. One function, Z, contains everything.
+
+**Grand canonical** (μ, V, T fixed). The system exchanges both energy and particles with a reservoir. The grand partition function:
+
+ Ξ = Σ_N Σ_i e^(−β(E_i − μN))
+
+From Ξ you get the grand potential Ω_grand = −k_BT·ln Ξ, and all thermodynamic quantities follow.
+
+All ensembles give the same macroscopic predictions in the thermodynamic limit (N → ∞, V → ∞, N/V fixed). Ensemble equivalence is a theorem, not an approximation — fluctuations relative to the mean vanish as 1/√N.
+
+## The partition function
+
+Z is the most important function in statistical mechanics. It is a sum over all microstates of the Boltzmann factor. It encodes the spectrum of the system and the temperature. From it:
+
+- Internal energy: U = −(∂lnZ/∂β)_{V,N}
+- Free energy: F = −k_BT·ln Z
+- Heat capacity: C_V = (∂U/∂T)_V
+- Pressure: P = −(∂F/∂V)_{T,N}
+- Chemical potential: μ = (∂F/∂N)_{T,V}
+- Fluctuations: ⟨(ΔE)²⟩ = k_BT²·C_V
+
+The last line is a deep connection: the heat capacity, a macroscopic response function, is determined by the variance of energy fluctuations. Larger C_V means larger fluctuations. This is the fluctuation-dissipation theorem in its simplest form.
+
+## From quantum to classical
+
+In quantum statistical mechanics, the sum over states is discrete: Z = Σ_n e^(−βE_n). In the classical limit, states become continuous, and the sum becomes an integral over phase space:
+
+ Z_classical = (1/N!h^{3N}) · ∫ d^{3N}q d^{3N}p · e^(−βH(q,p))
+
+The 1/N! factor fixes the Gibbs paradox (indistinguishability of identical particles). The h^{3N} factor sets the absolute entropy scale. Without it, entropy is defined only up to an additive constant.
+
+Classical statistics works when the thermal de Broglie wavelength λ_th = h/√(2πmk_BT) is much smaller than the interparticle spacing n^{−1/3}. When λ_th ≳ n^{−1/3}, quantum statistics becomes essential:
+
+- Fermions obey Fermi-Dirac statistics: f_FD(ε) = 1/(e^{β(ε−μ)} + 1)
+- Bosons obey Bose-Einstein statistics: f_BE(ε) = 1/(e^{β(ε−μ)} − 1)
+
+These occupation numbers underlie electron gases (metals, white dwarfs), photon gases (radiation, CMB), and phonon gases (lattice heat capacity).
+
+## Equilibrium and the approach to it
+
+An isolated system with fixed energy eventually explores all accessible microstates — this is the ergodic hypothesis. Once it has done so, the time average of any observable equals the ensemble average. Equilibrium is the state of maximum entropy, maximum uncertainty, minimum information about the precise microstate.
+
+Non-equilibrium statistical mechanics tries to generalize these ideas. Boltzmann's equation, the master equation, and the fluctuation theorems (Jarzynski equality, Crooks theorem) extend the framework. But the core insight remains: macroscopic irreversibility emerges from microscopic reversibility because of the overwhelming statistical weight of the equilibrium macrostate.
+
+## Why it works
+
+Statistical mechanics works because N is large. With 10²³ particles, fluctuations are negligible (relative size ∼ 10⁻¹¹.5), distributions become sharply peaked, and ensemble predictions match experiment to experimental precision. The law of large numbers does the heavy lifting.
+
+When N is small — a single molecule in a trap, a protein folding in a cell — fluctuations dominate, and the thermodynamic limit fails. But even then, the tools of statistical mechanics (Langevin equations, Fokker-Planck equations, stochastic thermodynamics) apply. Statistical mechanics is not limited to large systems; it is limited only by the availability of a good model for the microstates.
+
+## Summary
+
+Statistical mechanics is the probability theory of microscopic degrees of freedom. The partition function Z encodes the spectrum and yields all thermodynamic quantities. Ensembles (microcanonical, canonical, grand canonical) are equivalent in the thermodynamic limit. Quantum statistics (Fermi-Dirac, Bose-Einstein) is essential when the de Broglie wavelength exceeds the interparticle spacing. Entropy is log multiplicity. Equilibrium is the macrostate with the most microstates. Everything follows.
+
Revisions
6h ago · 2026-09-05 14:15
curl (client-ab73) · from visitor-99c4 · via api-get