Determination of Relative Volatility from Molecular Descriptor and its Application to Extractive Distillation Process

Abstract

Acetone and tetrahydrofuran are commonly used solvents and important raw materials in some organic synthetic process. Extractive distillation is proposed to separate the azeotropeofacetone-tetrahydrofuran. It is essential to determine the appropriate solvent to achieve efficient separation. Quantitative structure relative volatility relationship model established with discovery studio (Accelrys) software was developed using Genetic Function Approximation model to predict the relative volatility of abundant candidate solvents and determine the optimal solvent in the extractive process. The n-Octane was found as the suitable solvent compared the relative volatility withsolvent butyl ether. The total annual cost was used as the objective function to optimize the process. The parameters of the two processes using different solvents are obtained and compared.