Configurational Integrals and Intermolecular Forces

Introduction

An ideal-gas partition function counts independent molecular positions and momenta. Real molecules interact, so some position combinations are favoured and others are strongly suppressed. The configurational integral weighs all position arrangements by their interaction energy. It is the route from a microscopic pair potential to nonideal pressure, virial coefficients and liquid structure.

Core explanation

For N classical particles with positions r₁,…,r N, let total potential energy be U(r₁,…,r N). After momentum integration, the configurational part is Z conf = ∫ V d³r₁ … ∫ V d³r N exp[−U(r₁,…,r N)/(kT)]. The full simple-particle canonical partition function is Q N = Z conf/(N!Λ³ᴺ) under the classical convention. For an ideal gas U = 0 for allowed positions, so Z conf = Vᴺ and the familiar ideal expression returns. Real interactions make the integrand depend on arrangement rather than only on container volume.

A common approximation is pairwise additivity, U ≈ Σ i<j u(r ij). A pair potential may have a steep repulsive core at short separation and attraction at moderate separation. The Boltzmann factor suppresses overlapping core configurations and favours some attractive separations. The radial distribution function g(r) compares the probability of finding another particle at distance r with an uncorrelated reference. For a very dilute gas, g(r) approaches exp[−u(r)/kT] for isolated pairs, while at higher density many-body packing changes it.

Define the Mayer function f(r) = exp[−u(r)/kT] − 1. It is zero where interactions vanish, negative in strongly repulsive regions and positive in attractive regions. Expanding products of 1 + f ij organises Z conf into clusters. The pair cluster produces the second virial coefficient B₂ = −(1/2)∫ f(r)d³r under number-density units, equivalent to −2π∫f(r)r²dr for isotropic u(r). Higher clusters produce higher-order terms. This is a controlled low-density expansion, not a direct exact description of dense liquid structure from B₂ alone.

Pressure comes from p = kT(∂ln Q N/∂V) T,N. Because Z conf changes with V through both accessible space and interactions, its derivative need not equal NkT/V. This is the statistical origin of real-gas pressure departures. Different properties probe different aspects of the same interactions, so a potential fitted to pressure should also be checked against structure or energy data where possible.

Step-by-step reasoning

Write the kinetic wavelength factor and the positional integral separately. Insert U = 0 to verify the ideal Vᴺ limit. For pairwise forces, inspect the sign and range of u(r), form f(r), and connect its integral to B₂. Use a volume derivative of ln Q for pressure. State whether the low-density pair approximation or a full many-particle calculation is being used.

Visual explanation

Sketch two particle arrangements: one with overlapping repulsive cores has tiny Boltzmann weight; another at a favourable attractive spacing has larger weight. Plot u(r) with a steep wall and shallow well, and underneath plot f(r) negative in the core and positive in the well. Arrows connect the areas under f(r) to the sign of B₂.

Real-world analogy

Imagine arranging magnets on a table. Some placements are difficult because objects collide; others are favoured by attraction. Counting every geometric placement equally would overstate the chance of forbidden overlaps and miss clustering. The configurational integral weights arrangements, although real molecular forces are not identical to tabletop magnets.

Real-world example

Molecular simulations of gases and liquids use an assumed interaction potential to generate positions and calculate pressure and pair correlations. If the potential's repulsive core is too small, the simulated material may pack too densely. If attraction is too strong, it may predict too much clustering or a more negative second virial coefficient than experiments show.

Why?

Interactions couple particle positions. Their contribution cannot be represented by raising a one-particle position count V to the Nth power alone. The Boltzmann factor assigns low weight to high-energy arrangements and high weight to favourable ones. Integrating over all positions produces a configurational contribution whose volume and temperature dependence yields nonideal thermodynamics.

Common misconception

Z conf is not the same as compressibility factor Z; the symbols are conventional but describe different quantities. Another mistake is to assume pairwise additivity is exact for every material. Polarisation and many-body interactions can matter, and even a pairwise potential creates many-particle correlations at finite density.

Worked example

For hard spheres of diameter d, f(r) = −1 for r < d and zero for r ≥ d. Then ∫f(r)d³r = −4πd³/3, and B₂ = −(1/2)(−4πd³/3) = 2πd³/3. The positive coefficient raises low-density pressure relative to ideal. The calculation demonstrates how excluded spatial arrangements become a macroscopic equation-of-state correction.

Quick check

1. What does Z conf become when U is zero for every allowed particle arrangement? Answer: The Boltzmann factor is one everywhere in the allowed positional domain, so the N integrals each contribute V and Z conf = Vᴺ, giving the ideal-gas positional factor.

Exam focus

Keep interaction energy U distinct from internal energy as a thermodynamic average when context could confuse them. Include N! and Λ³ᴺ in the full classical Q. Check f(r)'s sign in repulsive and attractive regions. State the low-density condition when using pair clusters or B₂.

Advanced insight

In dense fluids, g(r) exhibits packing shells and long-range correlations near criticality. The pair potential still informs the energy, but many-particle organisation makes the low-density approximation g(r) ≈ e^(−βu(r)) inadequate. Integral-equation methods and molecular simulations connect the configurational integral to such structure without enumerating every arrangement directly.

Summary

The classical partition function contains a configurational integral Z conf = ∫e^(−U/kT)d³r₁…d³r N. Interactions change its value from ideal Vᴺ and therefore change pressure. Pair potentials lead to Mayer functions and a second virial coefficient, while dense fluids require fuller many-particle correlations.

Practice questions

1. A pair potential is strongly repulsive at short separation. What sign has f(r) there? Answer: u(r) is large and positive, so e^(−u/kT) is near zero and f(r) ≈ −1. 2. Why is g(r) ≈ e^(−u/kT) a low-density statement rather than a universal liquid formula? Answer: At higher density, other particles influence pair separations through packing and correlations. A pair's probability depends on the surrounding many-particle environment, not just isolated u(r). 3. How does the volume dependence of Z conf enter pressure? Answer: The canonical identity p = kT(∂ln Q N/∂V) T,N includes the V dependence of Z conf. Interactions alter that derivative relative to ideal NkT/V.