DISMOL simulator was used to determine the best possible operating conditions to guide, in future studies, experimental works. This simulator needs several physical-chemical properties and often it is very difficult to determine them because of the complexity of the involved components. Their determinations must be made through correlations and/ or predictions, in order to characterize the system and calculate it. The first try is to have simulation results of a system that later can be validated with experimental data. To implement, in the simulator, the necessary parameters of complex systems is a difficult task. In this work, we aimed to determe these properties in order to evaluate the tocopherol (vitamin E) recovery using a DISMOL simulator. The raw material used was the crude deodorizer distillate of soya oil. With this procedure, it is possible to determine the best operating conditions for experimental works and to evaluate the process in the separation of new systems, analyzing the profiles obtained from these simulations for the falling film molecular distillator.

A new calculation procedure for free-volume parameters is considered in this work by using viscosity prediction methods and the Levenberg-Marquardt calculation scheme. All parameters used in the VrentasDuda free-volume theory can be estimated from pure component properties. The prediction results are compared with experimental data for some polymer/solvent systems.The diffusion coefficient calculated by Vrentas-Duda theory can be used in the modeling of membrane separation processes.

An essentially predictive mathematical model was developed to simulate pervaporation process. The group contribution method UNIFAC was used for calculating the upstream activity coefficients. The diffusion coefficient in the membrane was predicted using free-volume theory. Free-volume parameters were determined with viscosity and temperature data, and the binary interaction solvent-polymer parameter was calculated by a group-contribution lattice-fluid equation of state (GCLF-EOS). A simulator named PERVAP was developed applying the mathematical model. Pervaporation process was simulated for separating bioethanol-water through polyetherimide membrane. The simulated results were validated using experimental data of bioethanol/water separation through polyetherimide membrane. The model presented a satisfactory performance compared to experimental data. Related to the simulation of the studied separation, a 99% molar enriched bioethanol stream was obtained with a recovery of 94%.

DISMOL simulator was used to determine the best possible operating conditions to guide, in future studies, experimental works. This simulator needs several physical-chemical properties and often it is very difficult to determine them because of the complexity of the involved components. Their determinations must be made through correlations and/or predictions, in order to characterize the system and calculate it. The first try is to have simulation results of a system that later can be validated with experimental data. To implement, in the simulator, the necessary parameters of complex systems is a difficult task. In this work, we aimed to determe these properties in order to evaluate the tocopherol (vitamin E) recovery using a DISMOL simulator. The raw material used was the crude deodorizer distillate of soya oil. With this procedure, it is possible to determine the best operating conditions for experimental works and to evaluate the process in the separation of new systems, analyzing the profiles obtained from these simulations for the falling film molecular distillator.

No abstract