Publication: Polydimethylsiloxane and poly(4-methyl-1-pentene) as gutter layer and p84 polyimide coated composite hollow fiber membranes for co2n2 and co2ch4 separation
Loading...
Date
2021-12-01
Authors
Mohd Shafie, Zulfida Mohamad Hafis
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
In this work, the possibility of using poly(4-methyl-1-pentene) (PMP) as substrate-gutter layer in composite membrane was compared with commonly used polydimethylsiloxane (PDMS) as gutter layer, supported on lithium chloride (LiCl) modified polyethersulfone (PES) porous substrate of varying surface pore architectures. High permeability, glassy nature of PMP allows it to be used as a co substrate-gutter layer, which is expected to mitigate the complexity of solution intrusion and lateral diffusion at dense-porous interface incomposite membranes. Results showed that the fabricated composite PES/PDMS was able to obtain permeance as high as 26.6 ± 2.6 GPU for N2 and 354.4 ± 27.9 GPU for CO2 at about 1 µm coating thickness. Nevertheless, these values are lower than asymmetric dense skin PMP membrane at 84.6 ± 6.2 GPU for N2 and 607.3 ± 31.3 GPU for CO2. Solution
intrusion and geometric restriction at the PES/PDMS dense-porous interface reduces its permeance efficiency to as low as 4% of its supposedly ideal permeance at low coating thickness. Using modified resistances-in-series (RiS) model, a correlation was developed to predict the composite membrane’s efficiency as a factor of the pre-coat substrate’s permeance, coating layer’s intrinsic permeance, geometric factor, and penetration factor. It was elucidated that substrate surface uniformity also significantly affects the resulting composite membrane permeance, which was found to be a constant of the substrate used regardless of the coating layer thickness. In comparison, asymmetric PMP with thin dense surface layer was noted to be advantageous as the substrate-gutter layer for the current work as it possesses a lower overall resistance as compared to the PES/PDMS layers and was not affected by geometric restriction and solution intrusion. Hence, asymmetric PMP membrane was chosen as a co-substrate gutter layer while N2, CO2 and CH4 gases were chosen as the model permeants for further composite membrane fabrication with P84 polyimide (PI), a commercial polymer notable for gas separation in industry, as selective layer. Fabrication of multi layered composite membranes are usually limited by the suitability of the solvent for each layer’s fabrication, which might damage the underlying layers. As PMP is non-soluble in N-methyl-pyrrolidone (NMP) that is used for dissolving P84 PI, it was found that PMP fiber is compatible to form a bilayer through dip coating with P84 PI, without
the need for pre-treatment to overcome the low surface energy of PMP. Hence, P84 PI of various concentration was dip coated at 5 mm/s onto PMP-based dense skin hollow fiber membrane and tested for gas permeation performance. Results showed that ideal selectivity as high as 42.36 ± 19.08 for CO2/CH4 and 18.55 ± 6.06 for CO2/N2 was achieved at 14 wt.% P84 PI coating. Nevertheless, despite of PMP’s resistibility to NMP, introduction of P84 PI at low concentration (2-10 wt.%) damages the thin, dense skin layer of the PMP’s membrane surface which jeopardize the composite’s separation performance. It is hypothesised that P84 PI’s shrinkage during drying period tore the underlying dense PMP layer, exposing the porous structure underneath. Hence, there exist a minimum P84 PI polymer concentration in which defect free P84 PI/PMP composite membranes can be made, which is at about 14 wt.%. At this concentration, dip coating speed can be manipulated to obtain a thinner selective layer suitable for composite membrane fabrication, although dewetting of the coating solution still occurred and magnified as the coating thickness is reduced, reducing the gas selectivity below the intrinsic values of P84 PI and more towards the intrinsic
values noted by the PMP substrate-gutter layer.