1,3-Dioxolane production via reactive distillation combined with vapor permeation: Experiments and modelling

Abstract

To overcome the problem of difficult separation of the 1,3-dioxane (DOL)–H2O azeotrope in the production of DOL, a new process via reactive distillation (RD) combined with vapor permeation (VP) was established. The reliability of the RD model has been reported in the literature. The VP membrane used an Aspen Custom Modeler (ACM) self-built model, with the permeation parameters of H2O, DOL, and formaldehyde (FA) in the membrane flux equation fitted through the H2O-DOL-FA dehydration experiments by NaA zeolite membrane. The results showed that the relative error between the mass of H2O in a series of permeates in experiments and simulated by the VP model was mainly within 20%, and R2 is 0.94. For DOL and FA, which have very small permeation, the value was within 50%, implying the reliability of the VP model. On this basis, a two-dimensional VP model was established. Finally, a new RD-VP-D process for producing DOL was designed. The VP was used to overcome the azeotropic point of DOL and H2O by dehydration. Then 99.9 wt% DOL can be obtained at the bottom of the subsequent DOL purification column, with DOL- H2O azeotrope at the top recycled back to the VP membrane. The effects of the degree of VP dehydration and column pressure on membrane area, reboiler duty, and TAC were investigated, meanwhile introducing a new concept: logarithmic mean partial-pressure difference (LMPD). Under the optimal parameter, TAC was reduced by 20.0% compared with the process reported in the literature, and CO2 emissions were reduced by 35.8%.