Vis enkel innførsel

dc.contributor.authorBorge, Tonenb_NO
dc.date.accessioned2014-12-19T13:23:19Z
dc.date.available2014-12-19T13:23:19Z
dc.date.created2010-11-25nb_NO
dc.date.issued2010nb_NO
dc.identifier372487nb_NO
dc.identifier.isbn978-82-471-1991-4 (printed ver.)nb_NO
dc.identifier.urihttp://hdl.handle.net/11250/248216
dc.description.abstractThe existence of an upper bound for the separation performance of polymeric membranes has motivated the development of new membrane materials with improved separation performance. Inorganic porous membranes like carbon molecular sieve (CMS) or zeolites have superior separation performance compared to polymeric membranes, but a challenging and costly manufacturing process have limited the commercialization of these membranes so far. Mixed matrix membranes (MMMs) consisting of porous inorganic particle dispersed in a polymer matrix have been suggested as a way to produce a membrane with good mechanical properties and enhanced separation performance compared to polymeric membranes. Increasing amount of research during the last 20 years has revealed several factors that are important in order to produce MMMs with properties exceeding the Robeson upper line for polymer membranes. The most essential and challenging task seems to be to achieve good adhesion of polymer onto the particle surface and avoid formation of defects at the interface. In this study, two alternative methods for production of defect free MMMs were evaluated; in-situ carbonization and in-situ polymerization. In addition, the production of CMSMs from the new precursor material chitosan was evaluated, including pyrolysis at intermediate temperatures to produce CMSMs with reduced brittleness. CMSMs were prepared from the precursor material; chitosan. CMSMs prepared by pyrolysis of chitosan had separation properties exceeding Robeson upper bound for polymer membranes for several gas pairs. The effects of chitosan molecular weight, casting temperature and casting with different solvents were evaluated and based on the results a procedure for preparation of chitosan precursor films was selected. The effects of final pyrolysis temperature, heating rate and pyrolysis atmosphere were also investigated. Thermal gravimetric analysis (TGA) showed that chitosan decomposes between 240-400 °C. The permeability results showed that a pore structure was developed already at 300 °C while the highest CO2 permeability was measured for a CMSM prepared at 650 °C when prepared under vacuum pyrolysis. Increasing the heating rate resulted in an increased permeability of the CMSMs. Pyrolysis in CO2 atmosphere, a mild oxidizing atmosphere, was expected to increase the permeability of the CMS membrane, but instead the CMSMs produced in CO2 at 650 or 750 °C showed lower permeability than the membrane prepared with vacuum pyrolysis with the same pyrolysis protocol. In-situ carbonization: Comparison of the thermal stability of chitosan and the polyimide Matrimid® showed that while the main weight loss for chitosan takes place between 240 and 400 °C, Matrimid was thermally stable up to ~500 °C. This observation was used to produce a “MMM” by in-situ carbonization of chitosan in a blend of chitosan and Matrimid at a pyrolysis temperature below the thermal decomposition temperature of Matrimid. Blend membranes with 90:10. 70:30, 60:40: 20:80, 10:90 chitosan:Matrimid ratio were prepared. The 70:30 and 60:40 chitosan:Matrimid film showed large degree of phase separation and consequently had lower mechanical strength than the other precursor films. The blend films were pyrolysed at 400 °C with a soak time of 0, 50 or 150 min. The separation performances of the resulting membranes were close to or slightly above the Robeson 1991 upper bound. The only solvent that was found to dissolve both polymers was trifluoroacetic acid (TFA), but this solvent also affected the mechanical and separation properties of Matrimid. The films with 80, 90 and 100% Matrimid turned out to be non-selective. In-situ polymerization of a monomer with dispersed CMS particles was evaluated as an alternative method to produce MMMs with good polymer-sieve contact. Several liquid monomers were evaluated based on their ability to disperse the CMS particles as well as separation performance of the polymers. Poly(ethylene glycol)dimethacrylate (PEGDMA) prepared from monomers with 9 (EGDMA-9) or 14 (EGDMA-14) ethylene oxide units was found to be the most suitable monomer/polymer. CMS particles (M650) were prepared from Matrimid, since the CMSMs prepared from chitosan had too low permeability to match the high permeability of the PEGDMA matrix. Separation into a polymer rich layer at the top and a CMS rich layer at the bottom during polymerization was found to be a problem. Faster formation of the membranes by increasing the polymerization rate was investigated as a method to fix the particles in the film and improve CMS distribution. Parameters like initiator type, initiator concentration, initiation method and polymerization temperature were evaluated. For UV initiated in-situ polymerization of PEGDMA9/M650, two of the initiators (2,2'-Azobisisobutyronitrile (I-1) and 4-(dimethylamino)-benzophenone (I-2)) resulted in a layered structure while (2-benzyl-2-dimethylamino-4-morpholinobutyrophenone (I-3)) resulted in MMMs with the CMS particle concentrated in the middle of the film. This film had no defects or holes and the permeability tests showed an increase in separation performance compared to the pure PEGDMA-9 membrane. No significant difference in distribution of CMS particles was observed when different amount of initiator was used (0.3, 0.6 to 1.2 wt%). For thermally initiated polymerization, increasing the amount of initiator from 0.3 to 0.6 wt% and polymerization temperature from 80 to 100 °C, resulted in a more uniform distribution of M650 within the PEGDMA-9 matrix. MMMs with good distribution of CMS particles were also produced by in-situ polymerization of PEGDMA-14 under the same conditions. However, all the MMMs prepared by heat initiated in-situ polymerization contained small holes and defects. Changing initiator to from I-1 to I-4 (benzoyl peroxide), did not changed the distribution of CMS particles, despite the longer half-time of this initiator, indicating the initiation rate and polymerization rate is fast enough to fix the particles in the polymer and prevent formation of a layered structure. Addition of CMS particles resulted in a higher H2 permeability and H2/CH4 selectivity for all the MMMs compared to the pure PEGDMA films. However, for the gas pair CO2/CH4, separation performance was more influenced by the preparation conditions and the distribution of CMS particles within the film. While a uniform distribution of particle was expected to give the highest raise in separation performance, the PEGDMA-9/17 vol%M650 MMM with best distribution of M650 particles showed only a neglectable change in separation performance for the gas pair CO2/CH4 indicating that other factors are influencing the transport. For MMMs prepared from PEGDMA-14, the separation performance for the gas pair H2/CH4 agreed very well with the predictions by Maxwell’s model based on the separation properties of the pure phases. The addition of M650 to PEGDMA-14 also resulted in a higher CO2/CH4 selectivity that generally increased with increasing loading up to 39 vol%, while the CO2 permeability was approximately unchanged. Rigidified zones of polymer around the particles with reduced mobility and/or partial blockage of the pores are suggested as reasons for why the permeability was lower than expected for CO2, assuming that H2 is too small to be influenced by any of the two phenomena to a large degree. The PEGDMA-14/M650 MMMs had separation properties close to the Robeson 1991 upper bound and the best MMM has separation performance within the zone that is considered industrial applicable. For the best MMM, the H2 permeability (35 °C, 2 bar) increased with 241% (from 11 to 39 Barrer) and the H2/CH4 selectivity increased with 416% (from 3.1 to 16), while the CO2 permeability increased with 15% (from 68 to 78) and the CO2/CH4 selectivity increased with 74% (from 19 to 33) compared to the pure PEGDMA-14 membrane. The selectivity increased further when the temperature for the permeability test was decreased from 35 to ~23°C. For CO2/CH4 the selectivity increased from 33 at 35 °C to 46 at 23 °C, while the CO2 permeability decreased from 78 to 55 Barrer.              nb_NO
dc.languageengnb_NO
dc.publisherNTNUnb_NO
dc.relation.ispartofseriesDoktoravhandlinger ved NTNU, 1503-8181; 2010:18nb_NO
dc.titleDevelopment of hybride membrane materials (mixed matrix membranes)nb_NO
dc.typeDoctoral thesisnb_NO
dc.contributor.departmentNorges teknisk-naturvitenskapelige universitet, Fakultet for naturvitenskap og teknologi, Institutt for kjemisk prosessteknologinb_NO
dc.description.degreePhD i kjemisk prosessteknologinb_NO
dc.description.degreePhD in Chemical Process Engineeringen_GB


Tilhørende fil(er)

Thumbnail

Denne innførselen finnes i følgende samling(er)

Vis enkel innførsel