Optimization of liquid–liquid equilibria of the type 2 ternary systems (water + valeric acid + aromatic solvent): Modeling through SERLAS

• LLE data of water + valeric acid + aromatic solvent systems were determined.
• An optimization algorithm is applied to the prediction of optimum LLE.
• The derivative variation method is used to identify the optimum point.
• The SERLAS model is able to simulate satisfactorily the observed performance.
• Two six-parameter models are used for predicting optimum equilibrium properties.

The study covers the liquid–liquid equilibrium (LLE) of the type 2 ternary systems (water + valeric acid + aromatic solvent) measured at T = (298.2 ± 0.1) K and P = (101.3 ± 0.7) kPa. The extraction efficiency of valeric acid by the aromatic solvents is better for xylene and chlorobenzene as compared to benzyl ether and 1-phenyl ethanol. An optimization algorithm utilizing the derivative variation method has been applied to the prediction of the optimization range of a type 2 LLE system. The capability of the proposed six optimization factors and two six-parameter models to represent conformably the optimum extraction field has been rigorously tested regarding the variation profile of the derivatives of the optimized quantity. A solvation energy relation SERLAS involving six physical descriptors has been implemented on the relevant systems, and checked for consistency in reproducing the observed performance. The deviation statistics obtained for SERLAS testify its ability to simulate accurately the observed performance with a mean error of 5.1%. The predictive capability of the UNIFAC-original model has been also studied.