GC-PPC-SAFT Equation of State for VLE and LLE of Hydrocarbons and Oxygenated Compounds. Sensitivity Analysis Thanh-Binh Nguyen, † Jean-Charles de Hemptinne,* ,† Benoit Creton, † and Georgios M. Kontogeorgis ‡ † IFP Energies nouvelles, 1-4 avenue de Bois-Pre ́ au, 92852 Rueil-Malmaison, France ‡ Center for Energy Resources Engineering (CERE), Department of Chemical and Biochemical Engineering, Technical University of Denmark, DK-2800 Lyngby, Denmark * S Supporting Information ABSTRACT: Group-contribution polar versions of SAFT equations of state are very useful for predictive calculations of mixtures containing diverse polar molecules. In this work, we have evaluated the predictive performance of one such model, the so-called polar perturbed-chain (PPC) SAFT model for phase-equilibrium properties of 290 hydrocarbons and monofunctional oxygenated compounds. Emphasis has been given on carrying out an extensive evaluation considering diverse types of phase behavior (vapor-liquid and liquid-liquid equilibria) and properties/conditions (Henry’s law constant for H 2 ,N 2 , and CH 4 ; infinite-dilution activity coefficient in water; solubility in water; infinite-dilution n-octanol/water partition coefficient). In general, considering the predictive nature of the calculations, encouraging results were obtained. For pure-component vapor pressures and liquid molar volumes, the deviations are very small, at 20% and 3%, respectively. The deviations in the prediction of the Henry’s law constants are within a factor of 2, with the best results found for the methane and nitrogen solubilities. For solubilities in water and, consequently, for infinite-dilution n-octanol/water partition coefficients, deviations are within a factor of 2 for hydrocarbons and within a factor of 4 for alcohols and aldehydes, but they are large for the other oxygenated families. To identify paths for improvement, a sensitivity analysis was performed, indicating that all of the parameters make large contributions to almost all properties. In addition, the sensitivity of the infinite-dilution activity coefficient in water to the molecular size parameters was extremely high. This suggests that a small change in these parameters might improve the results significantly. 1. INTRODUCTION The development of chemical processes for converting so-called second-generation biomass into biofuels and primary chemicals has received considerable attention, especially in the context of the depletion of fossil resources and global warming. Many processes are being developed at an industrial scale for this purpose. 1,2 The design and optimization of these industrial processes require reliable knowledge of the phase-equilibrium behavior of bioresources in a broad range of conditions. In contrast to fossil feedstocks, bioresources contain large amounts of oxygen-bearing compounds such as alcohols, acids, esters, and ketones. 3,5 For these fluids, classical models such as cubic equations of state and activity-coefficient models perform less satisfactorily because of strongly associating interactions present in these systems. 6 In an attempt to overcome this problem, many thermodynamic models have been proposed for associating fluids, such as those found in biomaterials. 7 One of the older successful models is the associated perturbed anisotropic chain theory (APACT) equation of state (EoS) developed by Ikonomou and Donohue. 8 This model was used to predict the thermodynamic behaviors of pure associating substances as well as mixtures. 9 Another model, proposed by the research group of Brignole, is the group-contribution association equation of state (GCA EoS). 10 The GCA EoS consists of an association term originating from Wertheim theory and the GC EoS developed by Skjold-Joergensen. 11 This model has been used successfully for systems containing fatty acids, 12 alcohols, esters, glycerol, and oxygenated aromatics, 13 for both liquid-liquid and liquid-vapor equilibria. 14 Coutinho’s group in Portugal used the cubic plus association (CPA) EoS 15 for calculating the phase equilibria of products originating from biomass, particularly esters. 16,17 The CPA EoS is a combination of the Soave-Redlich-Kwong (SRK) cubic equation of state with an association term. The parameters of this model were obtained from fitting experimental data. However, these equations of state have not been extensively applied to complex oxygenated fluids of relevance to biomass processing. Assisted by the development of molecular simulation tools, new equations of state that are based on statistical thermody- namic concepts have been developed. One of them is the statistical associating fluid theory SAFT EoS, which was proposed by Chapman et al. 18,19 on the basis of Wertheim’s perturbation theory. 20,21 According to this approach, equations of state are constructed on the basis of a physical description of the relevant molecules. During the past two decades, many versions of SAFT equations have been proposed such as the original SAFT, 19 Chen-Kreglewski SAFT (CK-SAFT), 22 simplified SAFT, 23 Lennard-Jones SAFT (LJ-SAFT), 24,25 soft- SAFT, 26 variable-range SAFT (SAFT-VR), 27 perturbed-chain SAFT (PC-SAFT), 28 and simplified PC-SAFT. 29 More details can be found in some reviews. 30-33 In general, the principal Received: October 14, 2012 Revised: April 24, 2013 Accepted: April 25, 2013 Article pubs.acs.org/IECR © XXXX American Chemical Society A dx.doi.org/10.1021/ie3028069 | Ind. Eng. Chem. Res. XXXX, XXX, XXX-XXX