it has been reported that membranes used to be discarded in the early stage of these evaluations, as being considered as inappropriate [14]. However, with the development of both theoretical and experimental works, the solutions provided by membranes have been highlighted, and they happen now to be promising candidates for post-combustion CO 2 capture if some problems are correctly addressed and solved by involving both material science and system design engineering [2,3,[14][15][16]. The challenges listed by these reports are the following: (i) membranes must display high selectivity toward CO 2 -at least 50-, (ii) it would be suitable to overcome the low CO 2 concentration in the feed flow (~ 12%), (iii) the large amount of gaseous or aerosol water (25%) present in flue gas would foul the membranes stages if not removed or drastically reduced beforehand, (iv) the low pressure of the feed gas (~ 1 bar) reduces the CO 2 recovery, which is a major drawback for low permeance membranes, and (v) the raw flue gas, even after the gas cleaning and desulfurization through the FGD stage, still contain corrosive components that will destroy polymer membranes if directly exposed.
AbstractThis report describes how commercial tubular ceramic membranes, initially designed for liquid filtration, can be modified to provide the core separation components of a first stage of flue gas treatment and enrichment in post-combustion CO 2 separation. Commercially available tubular NanoFiltration (NF) ceramic membranes were turned into a membrane for CO 2 separation by a two-step process including additional ceramic coating and chemical grafting. The combination of ceramic coating and chemical grafting drastically modify the membrane properties and turn the membrane initially designed for liquid filtration into a membrane that displays CO 2 vs N 2 selectivity at the opposite of Knudsen-based selectivity, with a CO 2 :N 2 ideal selectivity of 2.3. A second step of this study addressed the reduction of membrane cost, by starting with a low ultrafiltration (UF) 200 nm ceramic support specifically manufactured for this application in place of a NF membrane. After successful coating of a 5 nm and a 1 nm ceramic membranes, this membrane, grafted with a commercial fluorosilane molecule was tested in pure gas permeation of CO 2 and N 2 , with an ideal selectivity CO 2 :N 2 =3. Finally, the same membrane, grafted with glymo, was tested against separation of a CO 2 (20%):N 2 (80%) mixture, and as a function of the permeation stage-cut. A CO 2 :N 2 selectivity of 4 was obtained for a stage-cut of 0.5, and even higher (CO 2 :N 2 selectivity=14) for low stage-cuts usually used for testing dense polymer membranes. These results demonstrate that commercial ceramic porous membranes can be used as starting elements for a first stage of CO 2 post-combustion gas cleaning and CO 2 enrichment.