Predicting the Phase Behavior of Alcohols, Aromatic Alcohols, and Their Mixtures Using the Modified Group-Contribution Perturbed-Chain Statistical Associating Fluid Theory

Abstract

The modified group-contribution perturbed-chain statistical associating fluid theory (PC-SAFT) has been extended to model the phase equilibria of alcohols, branched alcohols, aromatic alcohols, and their mixtures. The parameterization has been implemented based on some physical arguments. The association energy of linear alcohols was fixed to the average experimental value, while this parameter for aromatic alcohols was directly estimated from the trimer hydrogen binding energy as evidenced by several experimental investigations. The dipolar moment of aromatic alcohols was reused from that reported experimentally. The importance of the association and the dipolar terms was investigated using the PC-SAFT equation of state by applying the model to represent liquid−liquid equilibrium (LLE) and vapor−liquid equilibrium (VLE) of several mixtures. The results obtained in this work suggest that including the dipolar term does not clearly improve the VLE prediction of linear alcohol-containing mixtures. It was possible to obtain good prediction results of VLE of alcohol-containing mixtures by using a nonzero binary interaction parameter to compensate for omitting the dipolar term (k ij = 0.012, within a 5% deviation on bubble pressure for 95 mixtures with 2616 experimental data). However, the addition of a dipolar term with the 2B association scheme to the PC-SAFT has been proven necessary to correctly describe the LLE/VLE of aromatic alcohol-containing systems. Good LLE and VLE computation results were obtained for almost considered mixtures.

Extending the Structural (s)-SAFT-γ Mie Equation of State to Primary Alcohols

“…This is because other authors may have made different choices regarding the following: temperature ranges from which data were taken, quantities of data points per species used, spacing between data points and data sources chosen. The average %AAD values of saturated vapor pressure and saturated liquid densities (or volumes) were 2.13 and 1.76% for polar GC-SAFT-0, 2.76 and 1.46% for mg-SAFT, 5.20 and 1.80% for hetero-GC-SAFT-VR, and 2.19 and 1.09% for SAFT-γ SW, respectively. Notably, SAFT-γ SW does not use DIPPR as a data source.…”

“…These are then treated as individual molecules in the respective model frameworks. For example, this was done in developing the GC-SAFT group interaction parameters for several hydrocarbon series and alcohols, GC-SAFT-VR group interaction parameters for common nonassociating and associating organic functional groups, , as well as mg-SAFT group interaction parameters for several hydrocarbon series , and alcohols …”

“…For example, this was done in developing the GC-SAFT group interaction parameters for several hydrocarbon series 52 and alcohols, 60 GC-SAFT-VR group interaction parameters for common nonassociating and associating organic functional groups, 61,62 as well as mg-SAFT group interaction parameters for several hydrocarbon series 53,63 and alcohols. 64 However, the aim of group contribution approaches is to model as wide a range of species as possible. While ethanol has been treated as a separate molecule in other GC approaches, 65 doing so reduces the general applicability of a GC approach.…”

Measurements of Isothermal Vapor–Liquid Equilibrium and Critical Point-Based Perturbed-Chain Statistical Association Fluid Theory Phase Behavior Modeling of the Propane + Phenol and Tetracosane + Propane/n-Butane Mixtures

“…Phenol and its mixtures present a major challenge for modeling. As discussed elsewhere, in addition to dipolar interactions, this compound exhibits complex association phenomena, which can be considered to be similar to carboxylic acids. Previously, Schmelzer et al demonstrated that accounting for association version of Cubic Schmidt–Wenzel EoS with system-specific and temperature-dependent binary adjustable parameter k 12 can correlate VLE and LLE in several n -alkane + phenol binary systems.…”

“…This approach resulted in reasonably accurate correlation of some data and less impressive modeling of others, such the upper critical solution temperatures (UCSTs), which were systematically overestimated. Later on, Nguyen Huynh et al obtained particularly accurate results for LLE in n -alkane–phenol systems using the modified group-contribution PC-SAFT which simultaneously accounts for dipolar and 2B association interactions. Different system-specific k 12 values were used for accurate correlation of VLE in the same systems.…”