Surface modification of silicalite-1 with alkoxysilanes to improve the performance of PDMS/silicalite-1 pervaporation membranes: Preparation, characterization and modeling

Abstract

The objective of this work was to study the effect of silane coupling agent on the performance of mixed matrix membranes (MMMs) and to develop a mathematical model to analyze the performance of the MMMs. The silicalite-1 modified by various alkoxysilanes was incorporated into polydimethysiloxane (PDMS) to prepare dense MMMs. The modified silicalite-1 and corresponding MMMs were characterized by FT-IR, CA, TGA, DSC and SEM. These results confirmed that the silicalite-1 was successfully modified by various alkoxysilanes. The effects of alkoxysilane chain length and chemical structure on the pervaporation (PV) performance of MMMs were discussed in detail. All the silane modification did not change the framework of silicalite-1. Better dispersion of silicalite-1 in PDMS was attained after modification and all the modified membranes could eliminate the nonselective voids inside the membranes. Moreover, the silane coupling agent had a contradictory effect on the PV performance of the membranes. On the one hand, the silane coupling agent improved the compatibility between silicalite-1 and PDMS. On the other hand, the silane coupling agent introduced a silane phase around the silicalite-1, which depressed the selectivity of the pristine silicalite-1. Silicalite-1 grafted with vinyl group showed the best compatibility with PDMS, and the resulting MMMs had the highest separation factor to ethanol because strong chemical bonds between vinyl-silicalite-1 and PDMS were formed with the thinnest silane film. For the separation of ethanol from a dilute solution, a separation factor of 34.3 was obtained with the vinyl-MMMs, which was 49% higher than that of unmodified membrane. As the silicalite-1 content increased, the separation factor increased. Furthermore, a mathematical model was proposed to predict the performance of MMMs. And both the permeability and selectivity obtained by the proposed model were in good agreement with the experiment data.