(381b) Polymer-Zeolite Dense Mixed-Matrix Membrane for Carbon-Dioxide Separation

Combustion of the fossil fuels has been main energy source since the time of the industrial revolution. Carbon-dioxide, as one of the main constituent of the flue gases, is produced and emitted in significant amounts. Increased concentration of carbon-dioxide in the atmosphere has strong negative influence to the environment. Despite the increased efficiency of the combustion chambers, treatment of the flue gases remains as one of the most challenging tasks in modern chemical and environmental engineering. Traditional methods of reduction of the carbon-dioxide amounts in flue gases include adsorption and cryoscopy methods. As those methods have significant disadvantages, membrane technology emerged as a promising alternative to the flue gas treatment. Approach that is analyzed in this work is construction of the mixed-matrix membrane based on the inorganic zeolite powder dispersed in the polymer matrix. Membrane that is suitable for this purpose should have as high as possible permeability for carbon-dioxide, with the lowest possible permeability for other common constituents of the flue gas (oxygen, water vapor, nitrogen, methane...). At the same time, membrane should have sufficient mechanical stability and should maintain permeation properties over time of the exploitation. Conventional approach to the membrane technology is based on the molecular sieves approach where the separation is done based on the size of the molecules. This mechanism cannot be applied for this purpose as membrane should be permeable for relatively large molecules of carbon dioxide, but it should not be permeable for small, non-polar molecules of hydrogen. One of the possible approaches that drew most of the attention of the chemical and environmental engineering community is to use dense composite membranes. Those membranes are based on the mechanisms of solution and diffusion. The molecule of the substance that should be separated from the mixture in dissolved in the membrane matrix, and then diffuses trough the bulk of the membrane to the permeate side. Inorganic zeolite powder can me dispersed in the matrix to increase the diffusivity of the carbon dioxide, and to decrease the diffusivity of other gases. The permeation is enhanced by the pressure difference between permeate (underpressure) and retentate (overpressure) side of the membrane. The main challenge in the choice of the polymer used for matrix is that carbon-dioxide should be solvable in it, but, at the same time, it should provide good, homogenous dispersion of inorganic powder.

In this experiment, possibility and feasibility of application of two different polymers with five different zeolite powders were tested. As it was predicted that polymer matrix might be incompatible with electrically charged zeolite powder particles, two different additives were tested. The reason for application of the additives is to avoid formation of the gaps between particles and polymer matrix. Formation of gaps has been observed to occur in some of the previous experiments. Two different additives were tested: n-tetradecyldimethylamonium bromide (NTAB) and dimethylaminopyridine (DMAP). It was supposed that highly charged ammonium group of the NTAB will be compatible with zeolite powder, while long non-polar aliphatic chain will provide compatibility with polymer matrix (mechanism similar to the effect of the detergent). On the other hand, main advantage of DMAP is (besides improving zeolite-polymer compatibility) that it has weak alkali properties which should improve the solubility of the slightly acid carbon-dioxide. One series of the samples without any additive was prepared as well. Polymer that was tested for this purpose was polyether-b-amide with 60% of polyether. This polymer was supplied by Arkema. It is available under the commercial name PEBAX 1657. The second choice for the suitable polymer was block-co-polymer of polyethylene glycol (PEG) and polybutylene terephthalate (PBT) with the weight ratio PEG:PBT is 77:23 and average molar mass of 1500 g/mol. This polymer was supplied by IsoTis OrthoBiologics (commercial name Polyactive). Five different types of zeolite particles with three different channel system grid were used for the experiment. The types of zeolite are LTL (1-dimensional pores); IHW, TER (2-dimensional pores) and MFI, FAU (3-dimensional pores). The reason for application of different pore systems was to test the influence of the pore system to the possibility of the membrane synthesis, and to the permeability and selectivity properties of the membranes. Ten different sample series were synthesized, one for each polymer-zeolite combination. Additional two membrane series were made of pure each polymer as a control series. In each series, three groups of samples were made: one with untreated zeolite (without any additive), and additional two groups with one additive in each of them (one group with NTAB, another group with DMAP).

The membranes were synthesized by dissolving the polymers in suitable solvents (70/30 wt.% water/ethanol mixture for PEBAX, and chloroform for Polyactive) at appropriate temperature (353K under reflux for Pebax and room temperature for Polyactive). The inorganic zeolites were dissolved in the small amounts of the solvent, and the additive was added (if the samples contained additives), and the solution was homogenized by ultrasound mixing. The solution was mixed with polymer solution and stirred overnight in order to get homogenous solution. Resulting solution was casted to the Teflon surface, covered with non-woven textile and left overnight at the room temperature and the working fume hood for drying.

The first estimation of the membrane quality was performed based on their appearance. The promising membrane should be transparent, smooth on touch at both of sides, homogenous and without any visible damages, air bubbles or spots. Membrane of the uneven thickness indicates that the drying process was dominated by the surface tension and that the viscosity of the casting solution was too high. Low viscosity solution, on the other hand, causes rapid sedimentation of the zeolite particles so membrane self-rolls due to the inhomogeneous distribution of the particles through the thickness of the membrane. White spots and/or white areas indicate presence of the air between the particles and polymer matrix. Visible bubbles in the membrane are result of rapid evaporation due to the low pressure in the drying phase of the membrane synthesis. All of the mentioned irregularities decreases the efficiency of the membrane.

The measurements of the permeability and selectivity was performed only for the membranes that were transparent and flat. For all samples, two type of measurements were performed. In the first step, pure (“dry”) gas was measured, and in the second trial mixture of the gas and water vapour (“wet”) was measured. Permeability measurements were performed by placing the membrane on the high vacuum line for 30 minutes in order to remove any solvent residuals. For the dry measurements, the gas that was measured was applied on the one side of the membrane, while for the wet measurements, the gas was mixed with saturated water vapour at 318K, and then applied on the one side of the membrane. In both of cases, another side was initially at vacuum, so the driving force was the difference in pressure. The permeability was measured by measuring the pressure on the permeate side of the membrane.

The optical observation of the membranes has shown that not all of the combinations are suitable for the synthesis of the membranes. In general, PEBAX-based membranes were better if they were prepared with zeolite treated with NTAB, while the Polyactive-based membranes have shown better results with untreated zeolites. Regarding the additives, PEBAX improved the appearance of the membranes without negative influence to their performance. On the other hand, NTAB did not give acceptable results in membrane synthesis. Permeability measurements have shown good results both in dry and wet measurements. The permeability of carbon dioxide was around ten times higher in comparison with permeability of the hydrogen in dry measurements (this ratio is usually called selectivity of carbon-dioxide versus oxygen). In the wet measurements, lower selectivity was achieved, with average values between 7 and 8. However, as most of the separation processes should be done in the wet environment, it might be concluded that this is promising system for the gas treatment and carbon-dioxide separation.