Physical Chemistry
Regular Solution Theory: Positive Deviations and the Interaction Parameter
Mix benzene and carbon tetrachloride, and the beaker warms by only a fraction of a degree; mix them in the wrong proportions of a more mismatched pair and the two liquids refuse to stay together at all, splitting into two layers. That difference is captured by a single number, the interaction parameter w (or its dimensionless cousin χ), at the heart of regular solution theory.
A regular solution is a liquid mixture in which molecules mix randomly, just as in an ideal solution, so the entropy of mixing is ideal, but in which the enthalpy of mixing is not zero. The theory, introduced by Joel Hildebrand in 1929, keeps the ideal entropy term and adds one correction: a symmetric excess enthalpy proportional to the product of the mole fractions, HE = w·x₁x₂. When w > 0 the mixing is endothermic and the solution shows positive deviations from Raoult's law.
- TypeNon-ideal solution model (physical chemistry)
- IntroducedJoel H. Hildebrand, 1929 (term coined earlier)
- Key equationG^E = w·x1·x2 ; RT·ln(γ1) = w·x2²
- Defining featureIdeal entropy of mixing, non-zero enthalpy of mixing
- Positive deviation whenw > 0 (endothermic mixing, γ > 1)
- Phase split abovew > 2RT (critical χ = 2)
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
What a Regular Solution Is and Where It Applies
A regular solution occupies the middle ground between the mathematically perfect ideal solution (where mixing costs no energy) and the messy reality of most liquid mixtures. Hildebrand's insight was to isolate a single deviation: keep the ideal entropy of mixing, ΔSmix = −R(x₁ln x₁ + x₂ln x₂), which assumes molecules distribute randomly, but allow a finite enthalpy of mixing.
- It applies best to mixtures of non-polar molecules of similar size with only dispersion (London) forces — e.g. benzene + carbon tetrachloride, hexane + cyclohexane.
- It underpins the Hildebrand solubility parameter δ = √(ΔHvap−RT)/Vm, the workhorse of paint, polymer, and pharmaceutical solvent selection ("like dissolves like" made quantitative).
- The Flory–Huggins χ parameter for polymers is its direct descendant.
Where it fails: strongly polar or hydrogen-bonding systems (ethanol–water), where mixing induces ordering and the ideal-entropy assumption breaks down.
The Derivation, Step by Step
Start from a lattice picture: N sites, each molecule surrounded by z nearest neighbors. Three pairwise interaction energies exist: ε₁₁ (1–1), ε₂₂ (2–2), and ε₁₂ (1–2). Assuming random mixing, the average number of unlike (1–2) contacts is proportional to x₁x₂.
- Step 1 — exchange energy. Define w = z·NA·[ε₁₂ − ½(ε₁₁ + ε₂₂)]. This measures how much a 1–2 contact costs relative to the average of the like contacts.
- Step 2 — enthalpy of mixing. HE = ΔHmix = w·x₁x₂ (per mole). If unlike attractions are weaker than like ones, w > 0 and mixing is endothermic.
- Step 3 — Gibbs energy. Since SE = 0, the excess Gibbs energy equals the excess enthalpy: GE = w·x₁x₂.
- Step 4 — activity coefficients. Differentiating GE gives the two-suffix Margules result: RT·ln γ₁ = w·x₂² and RT·ln γ₂ = w·x₁².
Because w > 0 forces γ > 1, the escaping tendency rises and the vapor pressure exceeds Raoult's prediction — a positive deviation.
Key Quantities and a Worked Example
The single parameter w controls everything. In dimensionless form, χ = w/RT. At 298 K, RT = 2.48 kJ/mol, so χ = 2 corresponds to w ≈ 4.96 kJ/mol.
Worked example. Take a mixture with w = +3.0 kJ/mol at T = 298 K, at composition x₁ = x₂ = 0.5.
- Enthalpy of mixing: HE = w·x₁x₂ = 3000 × 0.25 = +750 J/mol (endothermic — the beaker cools).
- Activity coefficient of 1: ln γ₁ = (w/RT)·x₂² = (3000/2478)·0.25 = 0.303, so γ₁ = 1.35.
- Partial pressure: P₁ = γ₁·x₁·P₁* = 1.35 × 0.5 × P₁* = 0.68·P₁*, well above the ideal 0.50·P₁* — a clear positive deviation.
The infinite-dilution activity coefficient is γ1∞ = exp(w/RT) = exp(1.21) = 3.36, the strongest non-ideality, felt when a single solute molecule is surrounded entirely by unlike neighbors.
How It's Measured and Used in Practice
The interaction parameter is not postulated blindly — it is extracted from data:
- Calorimetry. A mixing calorimeter measures HE directly; fitting HE = w·x₁x₂ to the symmetric parabola yields w. Real symmetric systems (benzene–CCl₄) give w on the order of +100 to +500 J/mol.
- Vapor–liquid equilibrium (VLE). Measuring total pressure vs. composition and back-fitting γ₁, γ₂ gives w through the Margules relations.
- Solubility parameters. Predict w a priori: w ≈ Vm·(δ₁ − δ₂)², so a large mismatch in δ (in MPa½) forces positive w. Water's δ ≈ 48 vs. hexane's ≈ 15 explains their immiscibility.
Industrially this drives solvent selection for coatings and drug crystallization, extraction design (choosing solvents that phase-split cleanly), and the Flory–Huggins theory that predicts polymer blend miscibility from χ.
How It Compares to Related Models
Regular solution theory sits in a family of increasingly flexible models:
- Ideal solution: the limit w = 0. Both γ = 1 and Raoult's law is exact (benzene–toluene).
- Athermal solution: the opposite limit — HE = 0 but SE ≠ 0, dominated by size/shape (polymer–solvent, Flory's entropic correction). Regular theory ignores exactly this.
- Margules / van Laar: add a second parameter for asymmetry; regular theory is the one-parameter, symmetric special case.
- NRTL and UNIQUAC: introduce local composition (non-random mixing) and temperature dependence, the modern engineering standard.
The key conceptual distinction: regular theory assumes random mixing even when HE ≠ 0, which is internally inconsistent (if unlike contacts cost energy, molecules should avoid them). This is why it captures the sign and rough magnitude of deviations but not their asymmetry or strong temperature dependence.
Exceptions, Phase Splitting, and Significance
The most dramatic consequence of a positive interaction parameter is liquid–liquid phase separation. The molar Gibbs energy of mixing is ΔGmix = RT(x₁ln x₁ + x₂ln x₂) + w·x₁x₂. When w exceeds a critical value, this curve develops two minima and the mixture splits.
- Critical condition: setting ∂²ΔG/∂x² = 0 at x = 0.5 gives the threshold w = 2RT, i.e. χc = 2. Above it, an upper critical solution temperature (UCST) appears.
- At 298 K that critical w is ≈ 4.96 kJ/mol — modest chemistry (a δ mismatch of a few MPa½) is enough to force demixing.
Where it breaks: negative deviations (w < 0, γ < 1) arise when unlike attractions dominate — acetone–chloroform (hydrogen bonding) is the classic exothermic case, giving a minimum-boiling... actually a maximum-boiling azeotrope. And systems with a lower critical solution temperature (nicotine–water) invert the sign of the temperature effect entirely, something the simple w-model cannot reproduce because it forces w to be temperature-independent.
| Property | Ideal solution | Regular solution | Real solution |
|---|---|---|---|
| Entropy of mixing | Ideal: ΔS = −R Σ xᵢ ln xᵢ | Ideal (random mixing assumed) | Non-ideal (ordering, size effects) |
| Enthalpy of mixing | ΔH = 0 | ΔH = w·x₁x₂ ≠ 0 | ΔH ≠ 0, generally asymmetric |
| Activity coefficient γ | γ = 1 | ln γ₁ = (w/RT)·x₂² | Fitted (Margules, NRTL, UNIQUAC) |
| Raoult's law | Obeyed exactly | Positive dev. if w>0, negative if w<0 | Both signs, asymmetric |
| Miscibility limit | Always miscible | Splits when w > 2RT | Complex, T-dependent |
| Example | Benzene–toluene | Benzene–CCl₄, hexane–cyclohexane | Ethanol–water, acetone–chloroform |
Frequently asked questions
What is the interaction parameter in regular solution theory?
It is a single energy, symbolized w (or its dimensionless form χ = w/RT), that measures the energetic cost of forming an unlike (1–2) molecular contact relative to the average of like contacts: w = z·N_A·[ε₁₂ − ½(ε₁₁+ε₂₂)]. A positive w means unlike neighbors are less favorable than like ones, so mixing is endothermic and the solution deviates positively from Raoult's law.
Why do positive deviations from Raoult's law occur?
When w > 0, mixing weakens the average attractive interactions, so each component's molecules escape more easily into the vapor. This raises the activity coefficient above one (γ > 1) and pushes the partial pressures above the ideal Raoult's-law line. Physically, the components 'prefer their own kind,' as with hexane and ethanol.
What distinguishes a regular solution from an ideal solution?
Both assume random mixing and therefore share the same ideal entropy of mixing, ΔS = −R Σ xᵢ ln xᵢ. The difference is the enthalpy: an ideal solution has ΔH_mix = 0 (γ = 1), whereas a regular solution has ΔH_mix = w·x₁x₂ ≠ 0. So the excess Gibbs energy of a regular solution is purely enthalpic: G^E = H^E.
When does a regular solution split into two phases?
Phase separation occurs when the interaction parameter exceeds the critical value w = 2RT, i.e. χ_c = 2. Above this threshold the Gibbs-energy-of-mixing curve develops two minima and the mixture demixes below its upper critical solution temperature (UCST). At 298 K this critical w is about 4.96 kJ/mol.
How is the interaction parameter measured or predicted?
It is measured by fitting the symmetric parabola H^E = w·x₁x₂ to calorimetric heat-of-mixing data, or by back-fitting activity coefficients from vapor–liquid equilibrium. It can be predicted a priori from Hildebrand solubility parameters via w ≈ V_m·(δ₁−δ₂)², so a large δ mismatch forces a large positive w.
What are the main limitations of regular solution theory?
It assumes random mixing even when H^E ≠ 0 (an internal inconsistency), forces the excess entropy to zero, and makes w temperature-independent and the model symmetric in composition. This means it cannot capture asymmetric deviations, strong hydrogen-bonding systems like ethanol–water, or lower critical solution temperatures. Local-composition models like NRTL and UNIQUAC were developed to fix these gaps.