Liquid-liquid and solid-liquid equilibrium

Liquid-liquid equilibrium (equal activities, splitting, UCST/LCST, ternary tie lines) and solid-liquid equilibrium (ideal solubility, eutectics).

Drafted with Aria, reviewed by the AiCanCode.org team. Spotted an error? Use Give Feedback at the bottom of the page.

Why it matters

Liquid-liquid extraction, decanters after azeotropic distillation, and phase separation in reactors all depend on liquid-liquid equilibrium (LLE). Crystallisation, freeze concentration, wax deposition in pipelines and the design of solvents for solids depend on solid-liquid equilibrium (SLE). Both use the same rule as VLE, equal fugacity of each species in every phase, applied to condensed phases.

Key ideas

LLE criterion. Two liquid phases α and β at the same T and P share the same pure-liquid reference fugacity fᵢ, so equal fugacity reduces to equal activity: xᵢ^α·γᵢ^α = xᵢ^β·γᵢ^β for every species. Pressure has little effect because liquids are nearly incompressible.

When does a liquid split? Only if the Gibbs energy of mixing curve has a region of negative curvature (see the phase-equilibrium criteria topic). That needs strong positive deviations: for G^E/RT = A·x₁x₂ splitting requires A > 2. Activity-coefficient models that can describe LLE include Margules, van Laar (within limits), NRTL and UNIQUAC; Wilson's equation cannot.

Binary solubility diagrams. On a T-x diagram, the coexistence (binodal) curve encloses the two-phase region. Its top is an upper critical solution temperature (UCST): above it the liquids are fully miscible (phenol-water, near 66 °C). Some systems have a lower critical solution temperature (LCST) instead, becoming immiscible on heating (triethylamine-water), and a few show both.

Ternary LLE and extraction. On a triangular diagram, tie lines join coexisting phase compositions inside the binodal curve and shrink to a point, the plait point, where the two phases become identical. The distribution coefficient K_D = (solute fraction in extract)/(solute fraction in raffinate) and the selectivity β = K_D,solute/K_D,diluent measure how good a solvent is. Phase amounts follow from the lever rule along a tie line.

SLE criterion. For a pure solid i in equilibrium with a liquid solution, f̂ᵢ^l = fᵢ^s, giving xᵢγᵢ = fᵢ^s/fᵢ^l (ratio of pure solid to pure subcooled-liquid fugacity at T). Following a path through the melting point, neglecting the heat-capacity difference between liquid and solid, gives ln(xᵢγᵢ) = −(ΔH_fus/R)(1/T − 1/T_m).

Ideal solubility. With γᵢ = 1 the solubility depends only on the solute (its ΔH_fus and T_m), not on the solvent. It is a good estimate when solute and solvent are chemically similar (naphthalene in benzene) and poor for dissimilar pairs (naphthalene in water, where γ is huge). Higher melting point and larger heat of fusion mean lower solubility.

Simple eutectic systems. If the solids are mutually insoluble and the liquid is ideal, each component has its own liquidus (freezing-point) curve from the equation above. They intersect at the eutectic point, the lowest temperature at which liquid can exist, where both solids crystallise together. Below the eutectic temperature the mixture is entirely solid.

Formulas

xᵢ^α·γᵢ^α = xᵢ^β·γᵢ^β (LLE, each species) ln[x₁^α/(1 − x₁^α)] = A·(2x₁^α − 1) (symmetric two-suffix Margules, x₁^β = 1 − x₁^α) A > 2 (phase splitting for G^E/RT = A·x₁x₂) K_D = x_solute^E / x_solute^R, β = K_D,solute / K_D,diluent ln(xᵢ·γᵢ) = −(ΔH_fus,ᵢ/R)·(1/T − 1/T_m,ᵢ) (SLE, pure solid i, ΔCp neglected) ln xᵢ = −(ΔH_fus,ᵢ/R)·(1/T − 1/T_m,ᵢ) (ideal solubility) x₁ + x₂ = 1 on both liquidus curves (eutectic point)

Symbols: xᵢ^α, xᵢ^β mole fractions in the two liquid phases; γᵢ activity coefficient; A Margules parameter (dimensionless); K_D distribution coefficient; superscripts E and R extract and raffinate; ΔH_fus,ᵢ molar heat of fusion (J/mol); T_m,ᵢ melting point (K); T (K); R = 8.314 J/(mol·K). Take ΔH_fus and T_m from a data book.

Worked examples

Example 1 (standard): ideal solubility of a solid. Given: naphthalene, T_m = 353.4 K, ΔH_fus = 19.0 kJ/mol. Estimate its mole-fraction solubility in benzene at 25 °C assuming an ideal solution.

  1. ln x = −(ΔH_fus/R)·(1/T − 1/T_m).
  2. 1/T − 1/T_m = 1/298.15 − 1/353.4 = 0.0033540 − 0.0028297 = 5.243 × 10⁻⁴ K⁻¹.
  3. ΔH_fus/R = 19 000/8.314 = 2285.3 K.
  4. ln x = −2285.3 × 5.243 × 10⁻⁴ = −1.198, so x = 0.302. Answer: x_naphthalene ≈ 0.30 (close to the measured value because benzene and naphthalene are chemically similar).

Example 2 (GATE level): eutectic point and liquid-liquid split. (a) Benzene: T_m = 278.7 K, ΔH_fus = 9.87 kJ/mol. With naphthalene as above, find the eutectic temperature and composition (ideal liquid, immiscible solids).

  1. Liquidus curves: x_N = exp[−2285.3(1/T − 1/353.4)] and x_B = exp[−(9870/8.314)(1/T − 1/278.7)] = exp[−1187.2(1/T − 1/278.7)].
  2. The eutectic satisfies x_N + x_B = 1. Trial T = 270 K: x_N = 0.136, x_B = 0.872, sum 1.007. Trial T = 269 K: x_N = 0.131, x_B = 0.858, sum 0.989. Interpolating and refining gives T = 269.6 K.
  3. At 269.6 K: x_N = 0.134, x_B = 0.866. Answer: eutectic at about 269.6 K (−3.6 °C), x_naphthalene ≈ 0.13

(b) A binary liquid follows G^E/RT = 2.5·x₁x₂. Find the compositions of the two coexisting liquids.

  1. Because the model is symmetric, x₁^β = 1 − x₁^α and the LLE condition becomes ln[x/(1 − x)] = A(2x − 1) with x = x₁^α.
  2. Solve by trial: at x = 0.145, left side = ln(0.1696) = −1.774; right side = 2.5 × (−0.710) = −1.775. Converged.
  3. Check equal activities: x₁^αγ₁^α = 0.1448 × exp(2.5 × 0.8552²) = 0.1448 × 6.224 = 0.901; x₁^βγ₁^β = 0.8552 × exp(2.5 × 0.1448²) = 0.8552 × 1.054 = 0.901. Equal. Answer: the phases contain x₁ = 0.145 and x₁ = 0.855

Common mistakes

  • Using VLE-style Pᵢ^sat in LLE; the vapour pressure cancels because both phases are liquids at the same T.
  • Taking the trivial solution x₁^α = x₁^β (identical phases) as the answer.
  • Using °C in the solubility equation or mixing kJ and J in ΔH_fus/R.
  • Assuming ideal solubility holds in any solvent; it ignores γ, which can be very large for dissimilar pairs.
  • Thinking the eutectic is a compound; it is a mixture of two solid phases that melt together at one temperature.
  • Using Wilson's equation for a partially miscible system.

For GATE CH

  • NAT problems on ideal solubility of a solid, freezing-point depression, and eutectic temperature from given ΔH_fus and T_m.
  • LLE with a one-parameter Margules model: condition for splitting (A > 2), compositions for a symmetric system, or activities in two phases.
  • Extraction-type questions using the lever rule and distribution coefficients on ternary data.
  • Conceptual MCQs on UCST and LCST, plait points, and why Wilson cannot predict LLE.

Quick check

  1. What is the equilibrium condition for species i between two liquid phases?
  2. For G^E/RT = A·x₁x₂, what is the minimum A for phase splitting?
  3. Does ideal solubility of a solid depend on the solvent?
  4. Raising the melting point of a solute, at fixed ΔH_fus and T, makes its ideal solubility higher or lower?
  5. What is the plait point on a ternary LLE diagram?

Answers: 1. xᵢ^αγᵢ^α = xᵢ^βγᵢ^β. 2. A = 2 (splitting for A > 2). 3. No, only on the solute's ΔH_fus, T_m and the temperature. 4. Lower. 5. The point where the tie lines shrink to zero length and the two liquid phases become identical.

Try answering each one aloud before you open it.

  1. 1.What is the equilibrium criterion for liquid-liquid equilibrium, and why does vapour pressure not appear in it?Concept

    Each species must have equal fugacity in both liquid phases. Both phases are liquids at the same T and P, so they share the same pure-liquid reference fugacity fᵢ, and the condition reduces to equal activities: xᵢ^α·γᵢ^α = xᵢ^β·γᵢ^β. The vapour pressure would appear on both sides and cancels, which is why LLE depends only on the activity-coefficient model and is nearly independent of pressure.

  2. 2.Why do some liquid mixtures split into two phases while others stay miscible?Concept

    A single phase is stable only if its Gibbs energy of mixing curve is convex in composition. Strong positive deviations from ideality (large positive G^E) can create a region of negative curvature, and the mixture then lowers its Gibbs energy by forming two liquids. For the one-parameter model G^E/RT = A·x₁x₂ this happens when A exceeds 2. Mixtures such as hydrocarbons with water have very large activity coefficients and split; similar molecules do not.

  3. 3.What are UCST and LCST?Concept

    The upper critical solution temperature is the highest temperature at which two liquid phases can coexist; above it the liquids are completely miscible, as in phenol-water. The lower critical solution temperature is the lowest temperature at which they split; below it they mix, as in triethylamine-water, where hydrogen bonding that favours mixing weakens on heating. Some systems show both, giving a closed solubility loop.

  4. 4.What is the ideal solubility of a solid, and when is it a good estimate?Concept

    For a pure solid in equilibrium with an ideal liquid solution, ln x = −(ΔH_fus/R)(1/T − 1/T_m), neglecting the heat-capacity difference between liquid and solid. It depends only on the solute's heat of fusion and melting point, not on the solvent. It works well when solute and solvent are chemically similar, such as naphthalene in benzene, and badly for dissimilar pairs where γ is far from 1.

  5. 5.Explain a simple eutectic phase diagram.Concept

    When two components are miscible as liquids but form pure, mutually insoluble solids, each has a freezing-point (liquidus) curve that falls from its pure melting point as the other component is added. The curves meet at the eutectic point, the lowest temperature at which liquid can exist, where both solids crystallise together. At fixed pressure this point is invariant, and below the eutectic temperature the mixture is completely solid.

  6. 6.How are LLE data used in designing an extraction process?Concept

    Tie lines from LLE data give the compositions of the coexisting extract and raffinate phases. The distribution coefficient K_D of the solute shows how much solvent is needed, and the selectivity (ratio of K_D for solute to K_D for the diluent) shows how clean the separation can be. The lever rule on a ternary diagram gives phase amounts, and stage-by-stage calculations use these to find the number of stages for a given recovery.

  7. 7.Why can't the Wilson equation be used for liquid-liquid equilibrium?Concept

    The Wilson form of G^E/RT always gives a Gibbs energy of mixing curve that is convex across the whole composition range for any positive parameters, so it never predicts a region of instability. It therefore cannot represent two liquid phases, even though it fits VLE of strongly non-ideal miscible mixtures well. NRTL or UNIQUAC, which can describe splitting, are used for LLE instead.

Finished this topic? Mark it so your progress, study plan and readiness keep up.

Stuck on something here?