The electrolyte perturbed-chain statistical associating fluid theory (ePC-SAFT) classical density functional theory (DFT) was developed to describe the behavior of pure ionic liquid (IL) and CO2/IL mixture confined in nanopores, in which a new ionic functional based on the ionic term from ePC-SAFT was proposed for electrostatic free-energy contribution. The developed model was verified by comparing the model prediction with molecular simulation results for ionic fluids, and the agreement shows that the model is reliable in representing the confined behavior of ionic fluids. The developed model was further used to study the behavior of pure IL and CO2/IL mixture in silica nanopores where the IL ions and CO2 were modeled as chains that consisted of spherical segments with the parameters taken from the bulk ePC-SAFT. The results reveal that the nanoconfinement can lead to an increased CO2 solubility, and the solubility increases with increasing pressure. The averaged density of pure IL and solubility of CO2 are strongly dependent on pore sizes and geometries. In addition, the choice of IL ions is very important for the CO2 solubility. Overall, the modeling results for silica-confined systems are consistent with available molecular simulation and experimental results.

Viscosity is one of the most important physical properties when developing ionic liquids (ILs) for industrial applications such as CO₂ separation. The viscosities of ILs have been measured experimentally, while the modeling work is still limited. In this work, the electrolyte perturbed-chain statistical associating fluid theory (ePC-SAFT) was combined with the free volume theory (FVT) to model the viscosities of pure ILs and IL mixtures up to high pressures and temperatures, in which the ePC-SAFT was used to calculate the density as inputs for modeling the viscosity of ILs with FVT. The ILs under consideration contain one of the IL cations [C_nmim]⁺, [C_npy]⁺, [C_nmpy]⁺, [C_nmpyr]⁺, or [THTDP]⁺ and one of the IL anions [Tf₂N]⁻, [PF₆]⁻, [BF₄]⁻, [tfo]⁻, [DCA]⁻, [SCN]⁻, [C₁SO₄]⁻, [C₂SO₄]⁻, [eFAP]⁻, Cl^–, [Ac]⁻, or Br^–. In total, 89 ILs were considered combined with a thorough literature survey of the available experimental viscosity data and evaluation. The comparison with the available experimental viscosities shows that the model can provide reliable representation and prediction for most of the pure ILs in a wide temperature and pressure range, and it can be further used to predict and describe the viscosity of IL mixtures reliably.

In this work, density gradient theory (DGT) was combined with electrolyte perturbed-chain (ePC)-SAFT to model the interfacial properties of pure imidazolium-based ionic liquids (ILs). The ePC-SAFT pure-component parameters for the IL-ions were taken from literature for the modelling of density and chemical potential of the pure ILs in the bulk phase. The calculated results were used as inputs for modelling surface tension using DGT. The influence parameters for DGT were obtained from the fitting of the experimental surface tensions. Application of anion-specific influence parameters linearised with the molecular weight of the IL-cation allowed to model surface tensions of pure ILs in a broad temperature range within experimental uncertainty. Surface tensions of ILs which have not been used for the fitting of the influence parameter were predicted in quantitative agreement with experimental data. DGT+ePC-SAFT was further used to predict the interfacial density profile of pure ILs.

In this work, ePC-SAFT was extended to predict the second order thermodynamic derivative properties of pure ionic liquids (ILs), such as isothermal and isentropic compressibility coefficients, thermal pressure coefficient, heat capacities, speed of sound, thermal expansion coefficient and internal pressure. ePC-SAFT predictions were compared with available experimental data of imidazolium-based ILs. The pure-component ePC-SAFT parameters for the IL-cations [C2mim]+, [C4mim]+, [C6mim]+ and [C8mim]+, and IL-anions [BF4]−, [PF6]− and [Tf2N]− were taken from literature in order to predict the thermodynamic derivative properties. The pure-component ePC-SAFT parameters for the IL-cations [C3mim]+, [C5mim]+, [C7mim]+ and [C10mim]+ were predicted based on linear molecular-weight-dependent relations. These estimated ePC-SAFT parameters were verified by comparing so-predicted pure-IL density as well as predicted CO2 solubility in ILs with respective experimental data. Further, these parameters were used to predict the second order thermodynamic derivative properties. The comparison of model prediction with experimental data showed that ePC-SAFT predictions were reliable in a wide temperature and pressure range.

CO2 separation plays an important role in greenhouse gas emission mitigation, in bio-fuel production via biomass gasification as well as in biogas upgrading. The current CO2 separation technologies are energy intensive, and new cost-effective CO2 separation technologies are needed. Ionic liquids (ILs) are promising absolvents for CO2 separation due to very low vapor pressurehigh solubility and selectivity for CO2 as well as low energy consumption for solvent regeneration. However, the absorption capacity of CO2 by ILs at low pressure is not high enough for practical application, and the capacity can be increased by incorporating functional groups such as amine group in ILs (task-specific ILs). In the last decades, a huge number of ILs has been synthesized in order to improve the IL performance for CO2 separation. Another drawback of using IL for CO2separation is the high viscosities, and using supported ILs has been proved to be a promising solution. Using supported ILs can take advantage of the high selectivity of gas in IL, and the high surface area of materials can reduce the impact of viscosity and improve the gas transfer, and hence increase the absorption rate. Considering the potential industrial applications and scientific interests, studying the thermophysical and transport properties of ILs and IL-containing mixtures both in bulk phase and in porous materials is crucial. The work of this thesis aims to develop tools for the estimation of important bulk and inhomogeneous properties of fluids related to CO2 separation using ILs. A first step of this thesis was to develop a density functional theory (DFT) based on perturbed-chain statistical associating fluid theory (PC-SAFT) equation of state (EoS) to describe the properties of inhomogeneous fluids in the pores of materials, which can be later used to describe the gas absorption on the materials supported with ILs. Compared to the molecular simulation results, the developed DFT model can rigorously provide the micro-structure of inhomogeneous fluids, and the model results are in good agreement with the simulation data for various “model” systems. The developed DFT model was further extended to model the gas adsorption on porous materials by assuming that the material has single-size pores with simple shape (e.g. slit or cylinder). It was found that the model with the parameters fitted to pure-gas adsorption at one temperature can be used to predict the pure- and mixed-gas adsorption isotherms at other temperatures with satisfied accuracy. As many porous materials have a wide pore size distribution (PSD), the influence of the PSD on the DFT model performance was investigated. In general, PSD can improve the accuracy of model results. In the modeling of transport properties, the viscosity models (friction theory (FT) and free volume theory (FVT)) were combined with electrolyte PC-SAFT (ePC-SAFT) to represent the viscosities of pure ILs and IL/CO2 mixtures. The viscosity model parameters of FT and FVT were obtained by fitting to the experimental viscosity data of ILs and linearized with the molecular weight of the IL-cation. It was found that FT can provide accurate results for both pure and binary systems, and FVT can provide satisfied results with few parameters. Hence FVT can be recommended for general estimation of viscosity containing ILs, and FT can be used for accurate calculation. Finally, ePC-SAFT was extended to describe the thermodynamic derivative properties such as heat capacities, isothermal and isentropic compressibilities, thermal pressure coefficient, speed of sound, thermal expansion coefficient and internal pressure. It is shown that the model with the parameters obtained from the easily-accessible experimental data (pure-IL density) can be used to predict the thermodynamic derivative properties over a wide range of temperature and pressure.

The developed hybrid PC-SAFT-DFT model, a coupling of density functional theory (DFT) with perturbed-chain statistical associating fluid theory (PC-SAFT), was used to study the adsorption of pure- and mixed-fluids on nano-porous materials, and carbons and zeolites were chosen as examples of nano-porous materials in this work for model performance evaluation. In the PC-SAFT-DFT model, the modified fundamental measure theory was used for the hard sphere contribution, the dispersion free energy functional was represented with a weighted density approximation, and the chain free energy functional from interfacial SAFT was used to account for the chain connectivity. The fluid was modeled as a chain molecule with molecular parameters taken from those in the bulk PC-SAFT. The external force field was used to describe the interaction between the solid surface of a nano-porous material and fluid. Application of this model was demonstrated on the gas adsorption on porous carbons and zeolites which were assumed to have slit- and cylinder-shaped pores with mean pore sizes, respectively. The parameters of the adsorption model were obtained by fitting to the pure-gas adsorption isotherms measured experimentally. With parameters of the model fitted to the pure-gas adsorption at one temperature, the model was used to predict the pure-gas adsorption at other temperatures as well as the adoption isotherms of mixtures. The model prediction was compared with the available experimental data, which shows that the predictions are reliable for most of the systems studied in this work. The effect of the pore size distribution on the model performance was further investigated, and it was found that the consideration of the pore size distribution (PSD) can improve the accuracy of the model results but the PSD analysis requires much more computing time.

In this work, the friction theory (FT) and free volume theory (FVT) were combined with the electrolyte perturbed-chain statistical association fluid theory (ePC-SAFT) in order to model the viscosity of pure ionic liquids (ILs) and IL/CO2 mixtures in a wide temperature and pressure (up to 3000 bar) range and with viscosities up to 4000 mPa·s. The ePC-SAFT pure-component parameters for the considered imidazolium-based ILs were adopted from our previous work. These parameters were used to calculate the density and residual pressure of the pure ILs. The density and pressure were then used as inputs for pure-IL viscosity modeling using FVT or FT, respectively. The viscosity-model parameters of FT and FVT were obtained by fitting to experimental viscosity data of imidazolium-based ILs and linearized with the molecular weight of the IL-cation. As a result, the FT viscosity model can more accurately describe the experimental viscosity data of pure ILs than the FVT model, at the cost of an increased number of parameters used in the FT viscosity model. Finally, FT and FVT were applied to model the viscosities of IL/CO2 mixtures in good agreement to experimental data by adjusting one binary viscosity-model parameter between the IL-anion and CO2. The application of FT required fitting the viscosity model parameters of pure ILs to experimental viscosity data of pure ILs and of IL/CO2 mixtures. In contrast, the FVT viscosity model parameters were adjusted to the experimental viscosity data of pure ILs only.

The perturbed-chain statistical associating fluid theory (PC-SAFT) density functional theory developed in our previous work was extended to the description of inhomogeneous confined behavior in nanopores for mixtures. In the developed model, the modified fundamental measure theory and the weighted density approximation were used to represent the hard-sphere and dispersion free energy functionals, respectively, and the chain free energy functional from interfacial statistical associating fluid theory was used to account for the chain connectivity. The developed model was verified by comparing the model prediction with molecular simulation results, and the agreement reveals the reliability of the proposed model in representing the confined behaviors of chain mixtures in nanopores. The developed model was further used to predict the adsorption of methane-carbon dioxide mixtures on activated carbons, in which the parameters of methane and carbon dioxide were taken from the bulk PC-SAFT and those for solid surface were determined from the fitting to the pure-gas adsorption isotherms measured experimentally. The comparison of the model prediction with the available experimental data of mixed-gas adsorption isotherms shows that the model can reliably reproduce the confined behaviors of physically existing mixtures in nanopores