Physical separation processes are among the most widely used gas separation techniques in chemical engineering, but most of them can be ruled out almost immediately because of the burden of processing the non-CO2 components of air. Physical separations typically depend on temperature or pressure variations, yet the dilute concentration of CO2 makes direct application of these techniques impractical. Heating air by 1 °C or pressurizing it by 1000 Pa (1% of an atmosphere) both require approximately 1.5 GJ/tCO2, about three times the thermodynamic minimum energy required for separating CO2 from air to produce a 1-bar product; yet a temperature swing of 1 °C or a pressure swing of 1000 Pa are still grossly insufficient to drive common physical separation process. This 1.5 GJ/tCO2 also corresponds to the net energy input if there is a very high 95% energy recovery on a still-modest 20 °C temperature swing or 20 kPa pressure swing.
Cryogenic separation CO2 could be recovered by cooling air at 1 atmosphere pressure to the point that CO2 condenses as a solid. At a 400 ppm concentration, this requires an initial temperature near -160 °C and requires cooling not only the CO2 but also the mass of oxygen and nitrogen too. Conceivably the air could be maintained above the 0.53 MPa (approximately 5 atm) triple point pressure while being cooled so the CO2 can be recovered as a liquid from the system, perhaps after distillation of a condensed nitrogen/oxygen/CO2 mixture. While this type of liquefied gas processing is established technology, cryogenic separations are expensive. Moreover, since the whole air mass must be cooled, an order of magnitude estimate for the cost for capturing CO2 may be derived from the cost of cryogenic O2 separation using the 500: 1 ratio of O2 : CO2 in ambient air implying an energy cost for CO2 capture of many hundreds of GJ/tCO2.
Physisorption to a solid surface, for instance a molecular sieve, is currently used in the front end of air separation plants to remove CO2 and water to prevent them freezing out in the later cryogenic distillation. Adapting such a process for air capture would require overcoming significant problems such as preferential absorption of water over CO2. Moreover, both pressure-swing and temperature-swing adsorption are batch processes with inefficiencies in recovering the energy required to swing the beds and their contents through their operating cycles. The mechanics of moving extremely large quantities of air through packed beds of solids also presents a major design problem, since having large amounts of surface area to improve mass transfer rates also means large areas for momentum loss, i.e. pressure drop.
Membranes that separate CO2 on the basis of its molecular size or its solubility in the polymeric matrix are under active development for application to flue gases (IPCC, 2005). Using them to separate CO2 from air where the driving force for CO2 is at most 40 Pa seems implausible given the relatively low molecular fluxes through membranes. Increasing the CO2 driving force by pressurizing the air feed is not practical because of the capital costs of compression and the energy losses in recovering that compression work. Operating the downstream (CO2 collection) side of the membrane at vacuum conditions does not increase the 40 Pa driving force and therefore the flux because that value already assumes zero pressure on the collection side of the membrane. Alternatively, the membranes could pass oxygen, nitrogen and argon while leaving concentrated CO2 behind. This approach requires a tremendous membrane area because of the quantity of gases that must be transmitted, the last fraction of which has little driving force because of its low residual concentration in the CO2.
Gas centrifuges suffer from low throughputs and relatively low separation per stage, problems which are worsened by the complexity of the equipment in each separation stage. Further, advances in design and operation of these systems are subject to government classification and export control limitations because of their potential use to separate and enrich nuclear materials.
Physisorption into a liquid is the basis for processes that absorb CO2 into a simple solvent such as cold methanol. Applying them to air capture suffers penalties with incomplete energy recovery while cooling and reheating the air stream and with the costs of any volatility loss of the solvent to the extremely large flow of air through the system.
Was this article helpful?