A grand product design model for crystallization solvent design

摘要

Solvents play an important role in crystallization processes. The screening/design of solvents for crystallization (crystallization solvents) is still of great concern in research and development. At present, most of the design/screening methods of crystallization solvents are still based on the trial-and-error approach. Even though some computer-aided design methods have been proposed, the process model as well as economic/environmental/health issues have not been fully established for a wide application range. In this paper, the Grand Product Design (GPD) model is applied for the screening/design of crystallization solvents. The GPD-model includes process sub-model, property sub-model, quality sub-model, cost submodel, pricing sub-model, economic sub-model and environmental sub-model as well as other factors such as company strategy, government policies and regulations. The general equations of the GPD-model for crystallization solvent screening/design are presented in this paper. Solution strategies are given for three cases: solvent design for a fixed process, process design for a fixed solvent and simultaneous design of solvent and process. Taking 2-Mercapotobenzothiazole (MBT) as an example, the solvent design for a fixed (existing) process design is carried out by using the problem specific GPD-model, in which the GPD-model is formulated as Mixed-Integer Non-Linear Programming (MINLP) model with objective function, process sub-model, property sub-model, quality sub-model, pricing sub-model, cost sub-model, economic sub-model and environmental sub-model. The established MINLP model is then solved by the decomposition-based approach. Experiments are carried out to verify the candidate solvents, which is found to perform better in terms of product purity and recovery than the best-known solvents in use.