Abstract
The main focus of this Ph.D. thesis was about the preparation of 2D materials and their application in thin-film composite membranes. Thus, graphene oxide, its modifications, and 2D Zn2(bim)4 MOF particles were successfully incorporated into the PIM-1 polymer matrix, and their effect on gas transport properties was investigated. Additionally, 2D COFs were synthesized and analyzed as adsorbents for gas uptake experiments.
Graphene oxide nanosheets were synthesized via the Hummers method, and its structure was determined by means of several techniques. It was revealed that after synthesis, graphene oxide nanosheets contain some amount of soluble oxidative debris that adsorbed on the surface of the layers, and it needs to be stripped off. The amount of oxidative debris is about 30% by weight and significantly influences the stability of the GO dispersion, which was proved by the exfoliation experiments. Thus, GO dispersion without oxidative debris showed higher concentration after centrifugation than that of as-synthesized graphene oxide.
After successful exfoliation, GO dispersions were dip-coated on microporous PAN support, and the potential use of such membranes in gas separation studies was tested. The experiments showed that the increase in thickness facilitates the permeance of hydrogen over CO2, N2, and CH4 through the GO nanolayers. Due to the presence of the oxygen functional groups, GO layers adsorb CO2 much higher than N2 and CH4. The decrease in the permeability coefficients of CO2 might be attributed to such a factor. However, it was suggested that GO is a promising material for the separation of hydrogen from carbon dioxide. The experiments showed that in a 15 nm GO layer, an H2/CO2 selectivity is 5. Plotting the obtained results in the 2008 Robeson upper bound, the surpass of the bound was achieved, showing that our results correspond to the previous results reported in the literature. The incorporation of GO into the PIM-1 polymer matrix hinders the permeation of gas molecules through the mixed matrix membrane resulting in low selectivities.
However, the covalent modification of the GO layers opens up new possibilities in the synthesis of the novel graphene-like sheets. Thus, newly synthesized GO modifications (GO-AEDPPF and GO-DClBAO) demonstrated ultimately different results rather than pure GO when they were incorporated in the PIM-1 polymer matrix. Although the functionalized GO (FGO) samples were not exfoliated, they showed an increased selectivity for CO2/N2. At 9 wt.% FGO loading, the selectivity of the PIM-1 membrane was boosted from 21 to 26 or were not changed up to 50 wt.%. Unlike pure GO, FGO-loading showed exceeding gas permeances at high loadings, which is explained with the mass transfer through the graphene layers.
Graphene oxide layers were also transformed into initiators for performing surface-initiated atom transfer radical polymerization. For the first step, the GO layers were converted into surface-initiator functionalized graphene oxide (SI-GO) layers, and they were used for the polymerization of 2-diethylaminoethyl methacrylate (DEAEMA) as in bulk and in exfoliated forms. The exfoliated SI-GO is a new approach in the SI-ATRP polymerization method, and such kind of an initiator could polymerize the target monomer that was not achieved via anionic polymerization. The controlled polymerization was confirmed with several characterization techniques, and the molecular weight of the synthesized polymer was ~40000 g mol-1 with a dispersity index of 1.2. The surface-initiated polymers were soluble in THF, and it was possible to dip-coat a membrane on the PAN support. Single gas transport experiments showed that the prepared amorphous polymer membranes are defect-free and show high selectivities such as 3158 for H2O/N2, 28 for CO2/N2, and 113 for H2O/CO2. Although the CO2/N2 selectivity is promising, the permeability coefficient of CO2 is low, which fits with the results reported for other methacrylate membranes. The membranes showed high water vapor permeability coefficients, which would be beneficial for the membrane distillation in order to overcome water scarcity in the future.
The synthesis of the 2D metal-organic frameworks and their incorporation into the PIM-1 polymer matrix were also under consideration. In the experimental part of this section, ZIF-7 nanoparticles were synthesized, and they were transformed into Zn2(bim)4 nanosheets. With the pore diameter of 0.21 nm, Zn2(bim)4 nanosheets offer high hydrogen separation performance over other gases. For this reason, the Zn2(bim)4 nanosheets were incorporated into the PIM-1 polymer matrix. It was observed that the H2/N2 selectivity increased from 11 to 15 and decreased at further loadings. The prepared mixed matrix membranes could not surpass the 2008 Robeson upper bound even though such kind of a nanofiller boosted the PIM-1 performance.
Furthermore, the synthesis of two-dimensional novel covalent organic frameworks and their gas uptake experiments were realized. High-pressure adsorption experiments revealed that novel COFs show comparable gas uptakes, and CO2/N2 selectivities are in good agreement with the reported COF adsorbents. COF-HZG2 showed the exceptionally high CO2/N2, and CH4/N2 selectivities with the solubility separation factor of ~25000 and ~5820 even though the BET surface areas and the pore volumes were lower than other COFs. This related to the non-polarization nature of nitrogen under high pressure.
As an outlook of this work, it is suggested to utilize the exfoliated nanolayers. Since graphene oxide is an insulating material, using conductive single-layer graphene, the percolation index can be detected. This gives us an interesting result for the explanation of gas transport through the PIM-1/G, PIM-1/GO, and PIM-1/FGO composite membranes. Exfoliated single-layer surface-initiated graphene oxide layers are promising initiators for the SI-ATRP polymerization, and the synthesis of block copolymers from exfoliated layers can be interesting for the membrane community. Another point for the pure GO membranes is to overcome the pinholes originated from the casting process that drastically decreases the performance of these membranes. Tuning the interlayer distance and the functional groups could be the next approach for the high-performance GO membranes as well. For the future, 2D MOF nanosheets, such as Zn2(bim)4, should be exfoliated to get better results even though the exfoliation is hurdle with the MOF nanosheets. Two-dimensional covalent organic frameworks offer a new prospect for the preparation of molecular sieving membranes to separate valuable components. It is worth noting that the pore size of the covalent organic frameworks must be tuned via post functionalization regarding the target gas molecules.
