Viabilidade técnica da síntese de epicloridrina a partir de glicerol: simulação comparativa da reação crítica em três processos

Abstract

Glycerol is currently a low-value commodity co-produced in large quantities in the synthesis of biodiesel, which renders its production mostly renewable. In the search for more valued substances that can be made from glycerol, epichlorohydrin stands out as the main upgrading option. This substance is a monomer of epoxy resins traditionally synthesized from propene in a particularly polluting process. The successful replacement of propene with glycerol in this synthesis, with economic and environmental gains, has motivated studies dedicated to further understand and optimize the process. This work takes place in the current research scenario as an evaluation and comparison of the technical feasibility of the synthesis through the simulation of its critical reaction, the hydrochlorination of glycerol to 1,3-dichloropropanol (1,3-DCP), in three different processes based on available literature and patents. After studying the kinetics of the reaction and choosing the best reactional model and kinetic data, the software Aspen Plus was used to simulate a process using a reactor in the CSTR model, a second process using a reactive column, and a third one combining both of these equipments in series. Departing from the operational conditions described in the references, the processes were optimized for 1,3-DCP production and equipment feasibility. The process with a CSTR proved to be the most feasible, in spite of the inferior results when compared to the claims of the reference patent, such as the yield of 68% for 1,3-DCP that was strongly limited by the volume of the reactor. The process that used exclusively the reactive column was shown to have no feasibility whatsoever due to the residence time in the column being much inferior to what would be required to yield noticeable amounts of product, which was proved by sizing the column internals. The simulation of the process that combines the two reactive equipments showed once again that the reaction virtually does not happen in the column, even when it is dimensioned beyond realistic sizes; the results got increasingly better as the production shifted towards being fully done in the CSTR, using the column solely for separation of the products. The three processes showed high performance in the purification of 1,3-DCP. The conclusions were that although feasible, the CSTR process lacks the information needed to lead the simulations to achieve the efficiency claimed in the reference patent; and that the feasibility of using a reactive column in this process depends on technical advances that allow for a larger residence time of the reagents within reasonable column sizes.