Single and multi-objective superstructure optimization of an integrated continuous multistage reaction-crystallization-filtration process with recycles

A superstructure optimization methodology is developed to identify optimal design and operation strategies of a combined continuous multistage reaction and crystallization process. The production of aspirin is used as a case study. The main objective is to maximize yield and minimize wastes while meeting tight critical quality constraints. The synthetic pathway involves a two-step reaction, which runs continuously, followed by a multistage seeded continuous cooling antisolvent crystallization process, and finally a wash filtration stage. To reduce the environmental footprint of the process and enhance its circularity, the filtrate (mother liquor) is purified using a multistage purification process then made available as recycle streams to the reaction and crystallization steps. The superstructure also considers that part of the filtered crystals can be recycled to the crystallization stages. A purge is used to control the level of impurities present in the system. Based on the proposed approach, several optimization scenarios were developed with a large set of decision variables, including jacket temperatures, antisolvent flow rates, recycle flow rates, residence times, and recycled crystal mass flow rate. The objective functions include maximizing the mean crystal size, and yield and minimizing the coefficient of variation (CV). It was shown that a maximum mean crystal size of 628 μm, a maximum yield of 87.91% and a minimum CV of 0.3596 can be achieved using the single objective optimization options. In addition, a multi-objective mixed integer nonlinear optimization problem was proposed to identify the best design and operation compromises captured by the Pareto front. Finally, a multicriteria decision-aiding approach based on the multi-attribute utility theory was used to rank the optimal Pareto trade-offs and help identify the most suitable solution.