Chemistry Labs

Physical chemistry

Linking the microscopic and macroscopic

The bridge between atoms and thermometers: S = k_B ln W, the Gibbs entropy formula, ensembles and fluctuations — how every thermodynamic quantity is an average (or a fluctuation) of molecular statistics.

IntuitionIntuition: the census behind the thermometer

Thermodynamics was built in the 19th century without knowing atoms exist. Its laws work because the macroscopic state of a beaker — pressure, temperature, entropy — is compatible with an astronomical number of microscopic arrangements, and nature spends almost all its time in the most probable ones. Statistical mechanics is the census: it counts those arrangements (W), weights them by Boltzmann factors, and shows that every dial on the lab bench reads out an average of molecular chaos.

Entropy is the measure of our ignorance. S = k_B ln W says: the more microstates share the same macroscopic face, the larger S. A crystal at 0 K has W = 1 and S = 0 — nothing is uncertain. Heat it, and exponentially many arrangements become reachable; S grows logarithmically with that freedom.

Two model systems against k_BT/ε: a two-level system showing the Schottky heat-capacity peak, and the Einstein solid showing the rise of C toward the classical 3Nk limit.

SchoolSchool: why counting explains spontaneity

Definition: Microstate, macrostate, multiplicity

A microstate is a complete microscopic specification — positions, velocities, quantum numbers of every particle. A macrostate is what a lab measures: T, P, V, composition. The multiplicity W of a macrostate is the number of microstates realising it. Equilibrium is simply the macrostate with the largest accessible W: gas expands into vacuum not because molecules prefer emptiness, but because there are overwhelmingly more ways to be spread out.

A two-level system makes the counting concrete. For N molecules that can be in ground or excited state, W = N!/(N₀!N₁!) peaks when the populations equalise — but only infinite temperature can truly equalise them, because the Boltzmann weight always favours the lower level. Stirling's approximation turns W into a smooth function whose maximum the system seeks.

UndergraduateUniversity: from q to every thermodynamic dial

Once the molecular partition function q is known, the canonical partition function Q = q^N/N! (indistinguishable particles) generates everything: the mean energy, entropy, free energies, pressure and heat capacity all follow from ln Q and its derivatives. The entropy formula below contains two contributions — the energy spread over states, and the sheer counting term Nk_B — which produces the Sackur–Tetrode equation for an ideal gas.

U−U(0)=−(∂ln⁡Q∂β) ⁣V;S=U−U(0)T+kBln⁡Q;A−A(0)=−kBTln⁡Q;CV=(∂U∂T) ⁣VU-U(0)=-\left(\frac{\partial\ln Q}{\partial\beta}\right)_{\!V};\quad S=\frac{U-U(0)}{T}+k_B\ln Q;\quad A-A(0)=-k_BT\ln Q;\quad C_V=\left(\frac{\partial U}{\partial T}\right)_{\!V}
S=NkB[ln⁡ ⁣(VNΛ3)+52](Sackur–Tetrode)S=Nk_B\left[\ln\!\left(\frac{V}{N\Lambda^3}\right)+\frac{5}{2}\right]\quad(\text{Sackur--Tetrode})

The 1/N! in Q is profound. Without it, the entropy of mixing two identical gases would be nonzero — the Gibbs paradox. Dividing by N! acknowledges that swapping two identical particles is not a new microstate; it makes entropy extensive and rescues the third law. Quantum mechanics later justified the factor: truly identical particles have no "which one" label.

Example: Entropy of argon at 298 K

Estimate the molar entropy of argon gas at 298 K and 1 bar using Sackur–Tetrode (m = 39.95 u, Λ ≈ 1.6×10⁻¹¹ m at 298 K).

Solution

At 1 bar the molar volume is RT/P ≈ 24.8 L, so V/(N_AΛ³) ≈ 0.0248/(4.1×10⁻³³) ≈ 6.0×10³⁰ per mole; ln of that is ≈ 70.8... more precisely the standard result S° ≈ R(ln(V/NΛ³) + 5/2) ≈ 154.8 J K⁻¹ mol⁻¹ — matching the tabulated value. The calculation is pure counting: no empirical entropy data enters.

AdvancedAdvanced: fluctuations, ensembles, and the quantum crossover

Fluctuations are not noise to discard — they are thermodynamic quantities. The energy variance in the canonical ensemble equals k_BT²C_V, so heat capacity literally measures how much the microscopic energy jitters. For N particles fluctuations shrink as 1/√N, which is why macroscopic thermodynamics is deterministic; near a critical point, however, fluctuations span the whole system and the averages-only view breaks down.

⟨(ΔU)2⟩=kBT2CV;S=−kB∑iPiln⁡Pi(Gibbs/Shannon)\langle(\Delta U)^2\rangle=k_BT^2C_V;\qquad S=-k_B\sum_i P_i\ln P_i\quad(\text{Gibbs/Shannon})

Quantum mechanics redraws the map at low temperature or high density. When the thermal de Broglie wavelength Λ becomes comparable to the mean interparticle spacing, particles can no longer be counted independently: bosons condense into the ground state (Bose–Einstein condensation, λ_transition ≈ 2.2 K in liquid helium), fermions build a Fermi sea that explains why metals have a tiny electronic heat capacity. The classical Boltzmann weight is then replaced by Bose and Fermi occupation functions — same logic, different indistinguishability rules.

The three ensembles and what is fixed
EnsembleFixedNatural potential
MicrocanonicalN, V, ES = k_B ln W
CanonicalN, V, TA = −k_BT ln Q
Grand canonicalμ, V, TPV = k_BT ln Ξ

ResearchResearch frontier

References

  • Equilibrium Free-Energy Differences from Nonequilibrium Measurements: A Master-Equation Approach · C. Jarzynski, 1997
  • Verification of the Crooks fluctuation theorem and recovery of RNA folding free energies · D. Collin, F. Ritort, C. Jarzynski, S. B. Smith, I. Tinoco, C. Bustamante, 2005
  • Elementary Principles in Statistical Mechanics · J. W. Gibbs, 1902
  • Information Theory and Statistical Mechanics · E. T. Jaynes, 1957