Search

The need for more-efficient and environmentally-friendly CO2 separation techniques is becoming more and more evident for a carbon-constrained world. High performance membrane materials, such as Polymers of Intrinsic Microporosity (PIMs) and specifically PIM-1, can promote faster growth of the technology of membrane gas separation. However, despite the potential of high free volume polymers such as PIM-1, the decay in their performance over time has inhibited a full exploitation of these materials and this has been the motive for the current project. This research explores a variety of methods to improve the initial and/or the long term performance of PIM-1-based membranes for CO2 separation. Expanding on the current studies on PIM-1-based membranes, it has been tried to focus on novel approaches for fabrication of membranes and development of materials. The explored routes in this research include: in situ PIM-1 synthesis, i.e. integration of PIM-1 synthesis and thin film composite (TFC) membrane fabrication, synthesis of novel crosslinked materials for incorporation into composite membranes based on PIM-1 and grafting reactions to improve the compatibility between the two phases of PIM-1-based composite membranes. In situ synthesis of PIM-1 and highly crosslinked network-PIM-1 was investigated on porous organic and inorganic supports with high thermal and chemical stability. Interfacial polymerisation of an octafluoro and a tetrahydroxy monomer led to promising CO2/CH4 selectivity (~ 15) for the in situ generated highly crosslinked TFC membrane. A highly-crosslinked network variant of PIM-1 with (ultra-)microporous nanoplatelet structure was successfully synthesized here for the first time. A noticeable improvement in the CO2/CH4 mixed-gas (1:1, v:v) separation performance of PIM-1 membranes at 2 atm transmembrane pressure difference and 298 K was obtained upon incorporation of network-PIM-1 nanoplatelets in PIM-1 continuous matrix. Low-crosslink-density (LCD) network-PIM-1 was also synthesised and was incorporated into PIM-1-based Mixed Matrix Membranes (MMMs). Conventional mixing/casting method or grafting PIM-1 onto LCD network-PIM-1 was explored as two separate approaches for the preparation of MMMs. The obtained MMMs based on grafting PIM-1 onto LCD network-PIM-1 showed a noticeable improvement in the aging behaviour, compared to PIM-1 membranes and the conventionally-fabricated MMMs.

There were three main driving forces behind this thesis: global concern over climate change mainly due to uncontrolled carbon dioxide (CO2) emissions, the excitement over the discovery of graphene and its versatile potential, and the potential to design three-dimensional (3D) or two-dimensional (2D) structures, in our case using unique triptycene molecule. We examined two polymeric materials for CO2 adsorption and suggested simple design of disordered carbons suitable for gas adsorption studies. The approach in each task was to examine structural and adsorption properties of materials using detailed atomistic modelling employing Monte Carlo and Molecular Dynamics techniques and where possible provide experimental measurements to validate the simulations. The thesis is presented as a collection of papers and the work can be divided into three independent projects. The aim of the first project is to utilize graphene as an additive in polymer composites in order to increase separation between the polymer chains increasing available surface area. The matrix used is a polymer of intrinsic microporosity (PIM-1), which possess large surface area and narrow nano-sized ( > 2nm) pore distribution attractive for gas separation membrane applications. Adding a filler can reduce aging of the polymer, and enhance permeability across the membrane, often to the expense of loosing selectivity. Therefore, we investigated the packing of PIM-1 chains in presence of discrete 2D graphene platelets and 3D graphene-derived structures and its effect on composite structure and adsorption properties. We found that additives do not alter structural polymer properties at the molecular level preserving the same adsorption capacity and affinity. Potential permeability increase would benefit from the retention of selectivity in the material. Building on design philosophy of materials with intrinsic microporosity we continued further investigation of 3D graphene-derived structures. The idea is that highly concave molecules or polymer chains pack inefficiently creating microporous materials with sufficient surface area for gas adsorption. 3D propeller-like structures were derived from graphene arms connected through the rigid triptycene and other types of cores. The resulting structures created a large amount of micropores and showed similar CO2/CH4 selectivity to activated carbons reported in the literature. It was shown that rigid triptycene core leads to more open structures. The model was also applied to model commercially available activated carbon to predict n- perfluorohexane adsorption. The fitting to experimental structural information proved to be challenging due to trial and error nature of the approach. Nevertheless, the simple packing procedure and diverse structure design have a great potential to serve as a virtual model for porous carbons that possess pore complexity and does not require any previous experimental data to be build on. The last project concerns CO2 adsorption and selectivity over CH4 and N2 in recently reported triptycene-based polymer. The triptycene shape polymer can form a porous 2D network that can be exfoliated into free-standing sheets and potentially used as a membrane. Sheets stack in the bulk material forming anisotropic channel pores. Additionally it contains fluoro- functional groups, which are known to have a high CO2 affinity. We explored pore structure and chemistry of stacked material for gas adsorption and predicted comparable capacity and CO2 selectivity to other microporous covalent materials such as activated carbons and PIMs. The CH4/N2 selectivity was similar to currently most selective material belonging to MOF family. We showed that fluoro-group have a positive effect on CO2 affinity, however predictions are sensitive to the charges of fluorine atoms assigned by different methods.

This work explores the fabrication and permeability testing of mixed matrix membranes (MMM) utilising a polymer of intrinsic microporosity (PIM-1) and various fillers. PIM-1 has been chosen for this work due to its high apparent surface area and high sorption of gases. PIM-1 also is a good candidate for gas sorption applications due to the film forming properties of the polymer. The fillers utilised in this work are Metal Organic Frameworks (MOFs) and organic cages, which have been chosen due to the gas sorption properties they exhibit. The MOFs used are micro and nanoparticles of Zeolitic Imidazole Framework-8 (ZIF-8), copper based MOF HKUST-1 and chromium based MOF MIL-101. Micro particles of magnesium based MOF Mg-MOF-74 were also looked at as well as cage 3, nano cage 3 and reduced cage 3. Comparable surface areas of the MOFs compared to those quoted in the literature have been obtained. Successful PIM-1/Filler MMMs were synthesised utilising PIM-1 and the fillers outlined above with various loadings of filler. The highest loading achieved was with a 10:6.4 PIM-1/nanoZIF-8 ratio. All MMMs apart from PIM-1/Mg-MOF-74 MMM were homogenous on a macroscale with scanning electron microscopy proving the dispersion of fillers. Gas transport properties of the MMMs were determined using predominantly a time lag method. PIM-1/ZIF-8 MMMs were also tested using a chromatographic method and using a gas sorption experiment. A range of gases were tested including CO2, N2, CH4, O2, He and H2. Ideal selectivities were also calculated with focus on the gas pairs O2/N2, CO2/CH4 and CO2/N2.When comparing the two permeability methods using the PIM-1/nanoZIF-8 MMM, lower permeability results were found from the time lag method. This was concluded to be due to the aging effect brought about by the vacuum used in the time lag method. The chromatographic method produced positive results with high selectivities, breaking Robeson’s upper bound, for the gas pair O2/N2. All other fillers tested showed an increase in permeability and stable selectivity with an increase in the amount of filler. MIL-101 and Cage 3 were the most successful fillers with high permeabilities of 35600 and 37400 Barrer respectively, encroaching on that of PTMSP. Mg-MOF-74 and reduced cage 3 MMM however, had a detrimental effect on the permeability. Aging data was also investigated which showed that for the majority of MMM the permeability followed the trend of PIM-1. microHKUST-1 and cage 3 of 10:3 loading were shown to give promising results with 10000 and 14300 Barrer respectively compared to 7200 Barrer for PIM-1. Although a loss in permeability is seen, it is still above that of PIM-1 at the same point of aging. These results give a positive indication that MMMs have the potential to provide resistance against aging, a major problem in using high free volume polymers in industrial applications.