Worked example
Example: A12 1.5, A21 0.9, x1 0.3. The inputs, in the units selected above:
| Input | Value |
|---|---|
| Margules constant A₁₂ (= ln γ₁ at infinite dilution) | 1.5000 - |
| Margules constant A₂₁ (= ln γ₂ at infinite dilution) | 0.90000 - |
| Mole fraction of component 1 | 30.000 % |
Applying the formula below gives:
| Result | Value |
|---|---|
| Activity coefficient of 1 | 1.7482 |
| Activity coefficient of 2 | 1.1695 |
| Excess Gibbs energy over RT | 0.27720 |
Open this example in the calculator and change any input.
Formula
\(\ln\gamma_1 = x_2^2\left[A_{12} + 2(A_{21} - A_{12})x_1\right], \quad \ln\gamma_2 = x_1^2\left[A_{21} + 2(A_{12} - A_{21})x_2\right]\)
Variables
| Symbol | Meaning | SI unit |
|---|---|---|
| \(A_{12}\) | Margules constant A₁₂ (= ln γ₁ at infinite dilution) | - |
| \(A_{21}\) | Margules constant A₂₁ (= ln γ₂ at infinite dilution) | - |
| \(x_1\) | Mole fraction of component 1 | - |
| \(\gamma_1\) | Activity coefficient of 1 | - |
| \(\gamma_2\) | Activity coefficient of 2 | - |
| \(G^E/RT\) | Excess Gibbs energy over RT | - |
Method and standards
Margules.
Where it is used
- Non-ideal VLE and modified Raoult's law.
- Solvent mixtures.
Common mistakes
- Constants for log₁₀ used as ln.
- Extrapolating outside fitted data.
Assumptions and limits
- Binary liquid.
References
- Two-parameter Margules G^E/RT = x₁x₂(A₂₁x₁ + A₁₂x₂) (Smith, Van Ness and Abbott)
Questions and answers
A₁₂ 1.5, A₂₁ 0.9 at x₁ 0.3?
γ₁ about 1.75, γ₂ about 1.17.
Where do constants come from?
Fitting VLE data, or from infinite-dilution activity coefficients.
Strongly non-ideal mixtures?
Wilson or NRTL usually fit better.
Method and article reviewed by FireCAD engineering on 12 November 2026. Every result is checked against an independent calculation on each release: how we verify our calculators.