Depending on the molecular model, group contribution (GC) approaches based on the statistical associating fluid theory (SAFT) can be classified in homosegmented and heterosegmented GC approaches. In homosegmented GC approaches, molecules are modeled as chains composed of identical segments. Heterosegmented GC approaches, on the other hand, consider molecular chains composed of different segment types and thus maintain a more detailed picture of real molecules. Therefore, heterosegmented GC approaches are arguably more physically realistic and ought to give more accurate descriptions of thermodynamic properties. In this work, we evaluate the performance of a homosegmented and a heterosegmented GC approach based on the perturbed-chain polar SAFT (PCP-SAFT) equation of state (EoS). To ensure a meaningful comparison between both GC approaches, a dipole term for the heterosegmented GC approach is formulated. Group parameters of 22 functional groups were adjusted to pure component property data. The comparison between both GC approaches shows that the heterosegmented GC approach leads to significantly better agreement with experimental data for various chemical families.

A Helmholtz energy functional based on the Perturbed-Chain Polar Statistical Associating Fluid Theory Equation of State is proposed. A weighted density approximation is used to develop two variants of dispersion functionals. We conduct molecular simulations (Monte Carlo) in order to assess both models. The structure of thin liquid adsorbed layers of a few molecular diameters thickness are shown to be particularly meaningful to detect deficiencies of Helmholtz energy functionals and thus discriminate between models. We further compare the model to experimental data of interfacial tensions for pure components and for binary mixtures exhibiting vapor–liquid equilibria and liquid–liquid equilibria. The proposed Helmholtz energy functional is in good agreement with experimental data of pure (organic) substances and mixtures and compares well with molecular simulation data of fluids adsorbing at solid interfaces.

No abstract

This study gives an assessment of the predictive capability of classical density functional theory (DFT) for adsorption processes of pure substances and mixtures of spherical and nonspherical molecular species. A Helmholtz energy functional based on the perturbed-chain statistical associating fluid theory (PC-SAFT) is applied to calculate isotherms and selectivities of multicomponent adsorption. In order to unambiguously assess the accuracy of the DFT model, we conduct molecular simulations. Monte Carlo (MC) simulations are performed in the grand canonical ensemble using the transition matrix. Two types of systems are studied: a model system, where fluid−fluid and solid−fluid interactions are defined as (single-site) Lennard-Jones interactions, and a more realistic methane−n-butane mixture in a graphite-like pore. Differences between a slit-shaped and a cylindrical pore geometry are examined for the model system. Adsorption isotherms and selectivities obtained from DFT calculations and MC simulations are found in very good agreement, particularly at high pressures. Capillary condensation observed along adsorption isotherms containing n-butane was accurately predicted, both, in equilibrium pressure and in density-increase. Comparisons with results from the ideal adsorbed solution theory are presented, confirming powerful predictions of the DFT approach.

This study demonstrates the capability of the density functional theory (DFT) formalism to predict contact angles and density profiles of model fluids and of real substances in good quantitative agreement with molecular simulations and experimental data. The DFT problem is written in cylindrical coordinates, and the solid−fluid interactions are defined as external potentials toward the fluid phase. Monte Carlo (MC) molecular simulations are conducted in order to assess the density profiles resulting from the Helmholtz energy functional used in the DFT formalism. Good quantitative agreement between DFT predictions and MC results for Lennard-Jones and ethane nanodroplets is observed, both for density profiles and for contact angles. That comparison suggests, first, that the Helmholtz energy functional proposed in a previous study [Sauer, E.; Gross, J. Ind. Eng. Chem. Res. 56, 2017, 4119−4135] is suitable for three-phase contact lines and, second, that Lagrange multipliers can be used to constrain the number of molecules, similar to a canonical ensemble. Experiments of sessile droplets on solid surfaces are performed to assess whether a real solid with its microscopic roughness can be described through a simple model potential. Comparison of DFT results to experimental data is done for a Teflon surface because Teflon can be regarded as a substrate exhibiting only attractive interactions of van der Waals type. It is shown that the real solid can be described as a perfectly planar solid with effective solvent-to-solid interactions, defined through a single adjustable parameter for the solid. Subsequent predictions for the contact angle of eight solvents, including polar components such as water, are found in very good agreement to experimental data using simple Berthelot−Lorentz combining rules. For the eight investigated solvents, we find mean absolute deviations of 3.77°.