An interactive guide to the theory and application of the UNIFAC model.
The UNIFAC (UNIQUAC Functional-group Activity Coefficients) model is a powerful semi-empirical system used in chemical engineering to predict activity coefficients in non-electrolyte, non-ideal liquid mixtures. Unlike models that require experimental data for every binary pair, UNIFAC leverages a "group-contribution" approach, estimating molecular properties based on the functional groups that constitute the molecules.
Activity coefficients ($\gamma_i$) are fundamental in chemical thermodynamics for describing deviations from ideal behavior in real solutions. They are crucial for accurately calculating phase equilibria (e.g., vapor-liquid equilibrium, VLE; liquid-liquid equilibrium, LLE), which is essential for designing and optimizing separation processes.
For an ideal solution, $\gamma_i = 1$, meaning the component behaves as if there are no interactions other than those between like molecules. In non-ideal solutions, $\gamma_i$ deviates from 1, indicating attractive ($\gamma_i < 1$) or repulsive ($\gamma_i > 1$) interactions between different types of molecules.
The UNIFAC model elegantly decomposes the total activity coefficient for each component ($i$) into two distinct contributions, reflecting different aspects of molecular interaction:
This term primarily accounts for the entropic effects arising from differences in molecular size and shape between the components in the mixture. It reflects the statistical probability of finding molecules in a certain arrangement due to their physical dimensions, rather than specific chemical interactions. It is calculated based on the volume ($R_k$) and surface area ($Q_k$) parameters of the functional groups that make up the molecules.
This term captures the energetic interactions between the different functional groups present in the mixture. It is the more chemically specific part, accounting for non-ideal interactions such as hydrogen bonding, dipole-dipole forces, and dispersion forces between the various functional groups. The residual contribution is derived from the concept of a "solution-of-groups," where the mixture is treated as a solution of functional groups rather than individual molecules.
The total activity coefficient for component $i$ is then given by the product of these two contributions:
This decomposition allows UNIFAC to predict properties for a vast number of mixtures by using a relatively small set of group-specific parameters, making it incredibly versatile for chemical process design where experimental data might be limited or unavailable.
The power of UNIFAC lies in its group contribution approach. Instead of needing parameters for every single compound in a mixture, UNIFAC breaks down molecules into their constituent functional groups. The model then uses parameters associated with these groups and their binary interactions to predict mixture properties.
UNIFAC classifies functional groups into "main groups" and "subgroups." For example, an alkane main group might include subgroups like CH₃, CH₂, and CH. Each subgroup has defined volume ($R_k$) and surface area ($Q_k$) parameters, which are derived from Van der Waals volumes and surface areas.
The residual contribution depends on "group interaction parameters" ($a_{mn}$ and $b_{mn}$) between pairs of main groups ($m$ and $n$). These parameters are typically temperature-dependent and are determined from experimental phase equilibrium data. The accuracy of UNIFAC predictions heavily relies on the completeness and quality of these interaction parameters.
This approach allows UNIFAC to predict properties for a vast number of mixtures, even those for which no experimental data exists, as long as the functional groups and their interactions are defined within the model's parameter tables.
The UNIFAC model is widely used in chemical engineering and related fields for its ability to predict thermodynamic properties of complex mixtures, especially when experimental data is scarce. Its group contribution nature makes it highly versatile. Click on each category to explore specific applications.
The UNIFAC model was a groundbreaking development in chemical thermodynamics, offering a practical way to predict mixture properties. This timeline highlights its key milestones.
The UNIFAC model is first published by Fredenslund, Jones, and Prausnitz, building upon the UNIQUAC model and introducing the concept of functional group contributions to activity coefficients.
Several revisions and extensions to the original UNIFAC model are published, refining parameters and expanding its applicability to more chemical systems.
The UNIFAC Consortium is established at the University of Oldenburg to support the continuous development, revision, and extension of UNIFAC and its variants (e.g., modified UNIFAC (Dortmund)).
UNIFAC is integrated into most commercial process simulators. Advanced variants like Modified UNIFAC 2.0 (incorporating machine learning) and UNIFAC-vdW-FV continue to expand its accuracy and range of applications, including ionic liquids and polymers.