A highly selective sorption process in POSS-g-PDMS mixed matrix membranes for ethanol recovery via pervaporation

Abstract

Incorporating nanofillers into polymer matrix to construct mixed matrix membranes (MMMs) is an effective approach to improve either sorption selectivity or diffusion selectivity or both. In this contribution, octa-vinyl polyhedral oligomeric silsesquioxanes (V-POSS) particles were chemically incorporated into polydimethylsiloxane (PDMS) matrix to fabricate POSS-g-PDMS MMMs. POSS-g-PDMS MMMs exhibited a homogeneous phase and achieved good integration between V-POSS and PDMS. Compared with pure PDMS membrane, POSS-g-PDMS MMMs possessed higher crosslinking density and gel content as well as enhanced ethanol affinity and hydrophobicity, which contributed to the significant improvement of ethanol sorption selectivity with a little sacrifice of diffusion selectivity. The chemical incorporation of V-POSS into PDMS matrix significantly enhanced both separation factor and permeation flux, which showed reversal trade-off effect. The best pervaporation performance of POSS-g-PDMS MMMs was achieved at 5 wt% POSS loading, with separation factor of 17.7 and permeation flux of 536 g/(m2 h), increased by 130% and 260% compared with pure PDMS membrane, respectively. An excessive POSS loading sharply depressed the fast transport of small molecules and finally led to the decline of diffusion selectivity and separation factor. The sorption and diffusion selectivity results suggested selective sorption was a major contributor to the high separation factor of MMMs at 5 wt% POSS loading, and diffusion was a rate-controlling process. The effect of variation of external driving forces and intrinsic membrane properties induced by increasing feed temperature and ethanol concentration on pervaporation performance was also investigated in detail. This work may provide useful insights of PDMS based mixed matrix membranes, especially those with covalent bonding between organic and inorganic phases for pervaporation recovery applications.