In this work, we introduce a novel method for the direct calculation of enthalpy based on segment contact probabilities of the COSMO-SAC-Phi (CSP) equation of state. The proposed formulation completely eliminates the need for temperature derivatives of the Helmholtz free energy while remaining fully consistent with classical thermodynamic relations. The model performance is evaluated by comparing the predictions of vaporization enthalpy for 97 compounds against experimental data and results obtained using the SRK equation of state with the Mathias–Copeman modification (SRK-MC) and the PC-SAFT model. The pure-component parameters of the CSP equation of state were optimized using experimental saturation pressure and liquid volume data. Two CSP versions were considered, namely CSP (2021) and CSP (2019), which employ five and four pure-component parameters, respectively. The average absolute relative deviations (AARD) were 2.63% and 2.95% for CSP (2021) and CSP (2019) versions, respectively. Notably, the CSP (2021) model showed lower deviations than PC-SAFT for 74 of the 94 compounds with data available for comparison. These findings underscore the potential of the CSP model to provide straightforward and reliable enthalpy estimates.
Predictive models of the excess Gibbs energy of liquid mixtures are paramount in chemistry research and chemical process design. The UNIQUAC pair-interaction parameters characterize the molar dissolution energy per molecular segment for each species at infinite dilution, which can be theoretically computed from the liquid-liquid interfacial tension. For mutual binary solubility estimation of water-organic systems, the root-mean square error (RMSE) in xo(w) (solubility of organic compound in water-rich phase) for UNIQUAC and modified UNIFAC are 1.283 (37 data points) and 2.690 (30 data points) correspondingly, the RMSE in xw(o) (solubility of water in organic-rich phase) for UNIQUAC and modified UNIFAC are 0.754 and 1.938 respectively. For prediction of solid solute solubility in pure liquid solvents (10 solid solutes, 96 systems, 267 data points), the total RMSE of UNIQUAC and COSMO-SAC are 0.651 and 1.436 respectively; which indicates that this new theoretical computation method of UNIQUAC binary energy parameters seems promising, and the predictive UNIQUAC exhibits better performance than COSMO-SAC in solid-liquid equilibrium (SLE) calculations. This new theoretical method couples excess Gibbs energy models with interfacial thermodynamics.
