Capture of carbon dioxide (CO2) from air

Abstract

Disclosed is a method for removing carbon dioxide from a gas stream, comprising placing the gas stream in contact with a resin, wetting the resin with water, collecting water vapor and carbon dioxide from the resin, and separating the carbon dioxide from the water vapor. The resin may be placed in a chamber or a plurality of chambers connected in series wherein the first chamber contains resin that was first contacted by the gas, and each successive chamber contains resin which has been wetted and carbon dioxide collected from for a greater period of time than the previous chamber, and so on, until the last chamber. Secondary sorbents may be employed to further separate the carbon dioxide from the water vapor.

Claims

1. A method for removing carbon dioxide from a gas, comprising bringing said gas in contact with a resin in a chamber, evacuating said gas from the chamber, wetting said resin with a liquid, collecting liquid and carbon dioxide from said chamber, and separating said carbon dioxide from said liquid.

2. The method of claim 1, wherein the resin is wetted by immersion in a liquid.

3. The method of claim 2, wherein the liquid is water and also contains carbonic anhydrase.

4. The method of claim 1, wherein the liquid comprises a secondary sorbent.

5. The method of claim 4, wherein the secondary sorbent is a weak liquid amine.

6. The method of claim 4, wherein the carbon dioxide is released from the secondary sorbent upon an increase in temperature.

7. The method as recited in claim 1, wherein upon evacuating said gas stream, said chamber is at a near vacuum pressure.

8. The method as recited in claim 1, wherein a plurality of chambers comprising the resin are connected in series, and wherein the liquid enters the first chamber and exits through a last chamber, the carbon dioxide and liquid being collected from said last chamber.

9. A method for removing carbon dioxide from a gas, comprising bringing said gas in contact with a resin in a chamber, evacuating gas from the chamber, wetting said resin with liquid water, collecting liquid water and carbon dioxide from said resin, and separating said carbon dioxide from said liquid water.

10. The method of claim 9, wherein the resin is wetted by immersion in liquid water.

11. The method of claim 9, wherein the liquid water also contains carbonic anhydrase.

Description

(1) Further features and advantages of the present invention will be seen from the following detailed description, taken in conjunction with the accompanying drawings, wherein

(2) FIG. 1 is an illustration of an ion exchange process in accordance with the prior art;

(3) FIG. 2 is a schematic showing the covalent bonds in an ion exchange resin in accordance with the prior art;

(4) FIG. 3 is a schematic showing an ion exchange process for capturing CO.sub.2 in accordance with the present invention wherein water is added to release the CO.sub.2 from the resin;

(5) FIG. 4 is a schematic showing a process employing weak liquid amine for capturing CO.sub.2 in accordance with the present invention;

(6) FIG. 5 is a schematic showing an ion exchange process for capturing CO.sub.2 in accordance with one embodiment of present invention wherein multiple chambers are used in succession;

(7) FIG. 6 is a schematic showing an ion exchange process for capturing CO.sub.2 where valves are used to control flow between chambers in accordance with the present invention;

(8) FIG. 7 is a schematic showing an ion exchange process for capturing CO.sub.2 employing activated carbon in accordance with one embodiment of the present invention;

(9) FIG. 8 is a drawing showing upright cylindrical towers in accordance with the present invention;

(10) FIGS. 9A and 9B are drawings showing the roll geometry of an adjustable collector in accordance with one embodiment of the present invention;

(11) FIGS. 10A and 10B are drawings showing the flat sheet geometry of an adjustable collector in accordance with one embodiment of the present invention;

(12) FIGS. 11A and 11B are drawings showing a collapsible collector in accordance with one embodiment of the present invention;

(13) FIGS. 12A-12F are drawings showing the fixed geometry of a collector in accordance with other embodiments of the present invention;

(14) FIG. 13 is a schematic of an apparatus of the present invention having an electrodialysis cell according to one embodiment of the present invention; and

(15) FIG. 14 is a schematic of an apparatus of the present invention having an electrodialysis cell according to an alternative embodiment of the present invention.

(16) This invention is based on utilization of a swing in CO.sub.2 equilibrium pressure with changing moisture levels, with changing temperature and with changing pressure. The combination of these changes offers a number of approaches to the recovery of CO.sub.2 from the air exposed resin. The apparatus described in this disclosure is capable of removing carbon dioxide from ambient air under a wide range of conditions, it is also capable of removing carbon dioxide from other gas mixtures. While the implementation is preferable for use with low concentrations of CO.sub.2 (i.e., less than 1%), it is also possible to collect carbon dioxide from gas mixtures that are rich in CO.sub.2 ranging from 1% to 98%.

(17) We have found that dry ion exchange resin exposed to ambient air can absorb CO.sub.2 and after some time reach an equilibrium loading that is balanced with the CO.sub.2 partial pressure in air. Rather than air one can also use similar gas mixtures with low partial pressures of CO.sub.2 ranging from 1 to 200,000 ppm. In the case that the resin has been exposed to ambient air with about 400 ppm of CO.sub.2 in the mixture, it is possible to wet the air exposed resin material and retrieve CO.sub.2 at a partial pressure of e.g., 80,000 ppm or more. Hence the presence of humidity has greatly increased the equilibrium partial pressure of CO.sub.2 over the resin. As a result one can wash the resin in water and captures the CO.sub.2 from the gas that is in contact with the water. Alternatively, it is possible to capture the CO.sub.2 from an aqueous solution that absorbs the CO.sub.2 because it has the necessary alkalinity to bind the CO.sub.2 that has been freed from the resin. The alkalinity can be provided in the form of hydroxide solutions (e.g. NaOH, or KOH), it can be provided in the form of carbonate solutions, e.g. Na.sub.2CO.sub.3, K.sub.2CO.sub.3, or even mixtures of carbonate and bicarbonate solutions, that have less of bicarbonate loading than is possible to achieve in contact with the air exposed resin.

(18) We have also shown in experiments that it is sufficient to expose the ion exchange resins to increased levels of humidity rather than directly expose them to liquid water and effect a substantial release of CO.sub.2 from the resin. Experiments have shown that the equilibrium partial pressure of CO.sub.2 over resin that has been equilibrated in its CO.sub.2 loading against ambient air is approximately a function of the absolute humidity rather than the relative humidity over the system. In effect the partial pressure of CO.sub.2 over resin does not change much with a range of temperatures, but it is very sensitive to a change in the absolute humidity of the gas it is in contact with.

(19) Experiments have also shown that the exposure to water vapor pressure can stimulate the release of CO.sub.2 from the resin even in the absence of the other constituents of air, as for example oxygen and nitrogen. In one aspect of the invention we therefore fill the resin material in a cartridge or otherwise put it into a chamber that can be evacuated prior to exposing its content to water vapor and/or liquid water.

(20) In one aspect of our invention we expose the resin to air in a variety of geometries and designs. After the resin has been saturated or partially saturated with CO.sub.2 from air and thus is in equilibrium with a partial pressure of CO.sub.2 which is less than 400 ppm, we release the CO.sub.2 into a gas space in which the partial pressure of CO.sub.2 can reach 50,000 microbar or more.

(21) In some of our earlier patent applications we describe use of Type 1 and Type 2 strong-base resin-functionalized with tertiary and quaternary amines for absorbing CO.sub.2 from air. As described therein, CO.sub.2 is extracted and stored on the resin as a carbonate or bicarbonate. More particularly, as described in U.S. patent application Ser. No. 11/866,326, a remarkable feature of these strong-base resins is that they absorb far more CO.sub.2 when they are dry than when they are wet. While the performance depends on the humidity in the air, the temperature of the resin and the state of CO.sub.2 loading of the resin, we have found that the equilibrium partial pressure over a wet resin is roughly 200 times larger than over a dry resin with similar CO.sub.2 loading. For example, we have found a resin that in its dry state has been exposed to ambient dry air, after exposure to water can fill a small air space with a partial pressure of CO.sub.2 that exceeds 60 mbar (60,000 microbar). A resin that is in equilibrium with a sodium carbonate/bicarbonate solution that in turn is in equilibrium with the CO.sub.2 in air binds only slightly more than one carbon atom, as a carbonate or bicarbonate ion, for every two positive charges attached to the surface of the resin. On the other hand if the resin is equilibrated with air of low humidity in the absence of liquid water, the number of carbon ions that are bound per attached ion approaches one. This property of the resin can be used to achieve a humidity induced pressure swing for capturing and collecting carbon dioxide.

(22) As the humidity in the air increases, the equilibrium partial pressure of CO.sub.2 over the matrix rises drastically. For example, we have tested a sample of 6 g of resin of anion exchange membrane material available from SnowPure, LLC, of San Clemente.sup.1. The membrane was slit into noodle-like strands 1 mm thick by 1 mm wide. An amount of this material that at 3 ppt of water in the atmosphere at room temperature is in equilibrium with 40 ppm of CO.sub.2 can raise the CO.sub.2 partial pressure in the a 20 liter bell jar to above 1200 ppm when the water vapor content of the air is raised to 30 ppt. And, a membrane material of 10 g that had been made to release roughly 100 cc of CO.sub.2, absorbed another 100 cc of CO.sub.2 after drying, and then in the first cycle of CO.sub.2 release after wetting, drove the CO.sub.2 content of the bell jar to 3700 ppm. Thus, it appears that the drying resin material rearranges itself, accommodating bicarbonate ions rather than just carbonate ions. Because of this pronounced swing in carbon capacity, it is sufficient to raise the humidity over the resin to force the resin to release CO.sub.2 that has been absorbed. .sup.1 The manufacturer describes its material as comprising crushed anionic exchange resin mixed in a polypropylene matrix and extruded as a sheet according to the teachings of U.S. Pat. Nos. 6,503,957 and 6,716,888.

(23) While not wishing to be bound by theory, it is believed that the resin as it dries out rearranges itself at least in part to favor accommodation of bicarbonate ions rather than carbonate ions. However, we have observed that the CO.sub.2 uptake rate gradually declines as the loading increases. There is no sharp boundary between the carbonate/hydroxide and the carbonate/bicarbonate regime for dry resins, suggesting that the uptake mechanism does not change as the system transits from one with residual hydroxides to one with residual carbonate ions. One possible explanation is that in a more or less dry system, carbonate ions can disassociate into bicarbonate ions and hydroxide ions, even though in aqueous solution the equilibrium of this reaction greatly favors carbonates over bicarbonates. One possible way of looking at this is that bicarbonate will precipitate on the resin surface touch earlier than carbonate ions and thus they could remove some of the carbonate from the solution in the form of bicarbonates. In any event the effective pH of the surface seems higher than it would be based on the level of carbonation in an aqueous system.

(24) Because of this pronounced swing in carbon dioxide storage capacity, it is in principle sufficient to raise the humidity over the resin to force the resin to release CO.sub.2 that has been absorbed. One possibility is to immerse the resin in water, in order to release the CO.sub.2. This process is shown in FIG. 3. A short wetting of the material by itself may not be sufficient to release the bulk of the bound CO.sub.2. The CO.sub.2 release from the material persists for a long duration but does release a large fraction of the CO.sub.2 that can be bound to the resin surface. Another embodiment of the present invention is to expose wetted resin to air, as opposed to immersing the resin for the duration of the CO.sub.2 release. Because time is the major constraint on this design, it is recommended that the apparatus be relatively large and allow the outgassing of the volume for an extended amount of time.

(25) Operating in a near vacuum, where the nitrogen and oxygen of the air have in effect been removed from the system, does not slow the kinetics of the process down. Thus it is possible to generate a gas stream from the wetted resin material that is a mixture of water vapor and CO.sub.2 that has been released from the resin. The partial pressure of the latter may be as high as 60 millibar. The partial pressure of the former will depend on temperature but should be about 30 millibar or less. The containment of the system should be constructed in a way to throttle the flow so that the slow release of CO.sub.2 can continue while keeping the system near equilibrium conditions. If the CO.sub.2 is evacuated too quickly, the rapid water vapor production kinetics will change the ratio of CO.sub.2 released to water vapor release in favor of excess water release. This would cool the system, requiring additional heat input. If water were caught in a condensation trap the energy for cooling would grow rapidly. Thus it is important to stay near the equilibrium point, which sets the maximum acceptable flow rate.

(26) Under dry atmospheric conditions, CO.sub.2 is collected from the air. Once the resin reaches a high level of saturation, in which it is essentially transformed from its carbonate form to a saturated bicarbonate form, the resin material is packed into a chamber, which is first evacuated and then filled with water vapor and with liquid water which is agitated to the point that all resin surfaces are thoroughly wetted. Alternatively, the resin may be contacted with water vapors only. However, the rate of CO.sub.2 release from a resin is much higher when the resin is wet than when it is just exposed to water vapors. Indeed, not only does the rate of release increase, the partial pressure CO.sub.2 also goes up. We have found that it may be useful to add small amounts of carbonic anhydrase to the water in order to improve the reaction kinetics of the CO.sub.2 transferring from the resin to the water and on to the surrounding gas space. In some designs it may be advantageous to begin by collecting the CO.sub.2 loaded resin material in a container that is first evacuated to remove excess air. Then the resin is wetted so that is releases CO.sub.2. This leads to 50 mbar of CO.sub.2 in the chamber, added to this would be around 20 to 30 mbar of H.sub.2O. It is possible to change the water content by raising and lowering the temperature in the system.

(27) It is now possible to pump the produced gas out of the system, as the gas is compressed the water in the gas will condense out, while the CO.sub.2 becomes more and more compressed. Ultimately we can drive the pressure to CO.sub.2 pipeline pressure. If desired, the CO.sub.2 may be dried before bottling or sending it to the customer.

(28) Even though wetting the resin material allows for the release of CO.sub.2 from the resin, this release is relatively slow. A short wetting of the material by itself is not sufficient to release the bulk of the bound CO.sub.2. However, we have found that the CO.sub.2 release from the material persists for a long time but it involves a large fraction of the CO.sub.2 that can be bound to the resin surface.

(29) In an experiment performed with 6 g of membrane material cut into strips, and after 15 minutes of washing the resin in DI water, the resin would still exhale CO.sub.2 and raise the CO.sub.2 level in a closed bell container system to about 1200 ppm. Left alone over night, but allowed to dry, the resin recaptured all this CO.sub.2 and then some, reducing the CO.sub.2 level in the bell container to about 40 ppm. Raising the humidity in the bell container to 30 ppm lead to CO.sub.2 exhalation that reached above 1400 ppm. Releasing this CO.sub.2 and washing the resin strands in DI water, would lead to a much more rapid release of CO.sub.2 that raised the pressure in the bell to about 1800 ppm. A second experiment yielded well over 1100 ppm and a third still reached 700 ppm. In total the system exhaled at least 1000+1400+500+3003300 ppm. This translates into roughly 66 cc of CO.sub.2 or about of the CO.sub.2 capacity that was initially collected by the carbonate washed membrane. The total capacity of the membrane material should be around 6 millimole or about 120 cc. As a result, it is believed that pure water alone can cause a large swing in CO.sub.2 binding to the matrix. It is not clear how much CO.sub.2 has been released from the matrix during the wash cycle outside of the bell jar. While the system spent several minutes in the water a substantial amount of additional CO.sub.2 could have been removed. The data toward the end of the run suggest that the rate of CO.sub.2 released dropped somewhat during the wash cycle, suggesting that the wash cycle has removed additional CO.sub.2 at a somewhat faster rate. However, the total amount of CO.sub.2 that has been removed during the wash cycle is bound to be relatively small considering how much has been emitted in the quiet periods in between.

(30) The rate of decarbonization of the membrane material while it is in the water was not measured. However, it was observed that carbonic anhydrase had little effect on the rate of CO.sub.2 release from the wet membrane which suggests that the slow step in the process is the release of the CO.sub.2 from the membrane.

(31) Alternatively, as illustrated in FIG. 4, we can extract the CO.sub.2 from the gas stream using other schemes of separation, such as a weak liquid amine for a secondary absorption step. This amine must be capable of pulling the CO.sub.2 content of gas mixture down so that the CO.sub.2 partial pressure drops to about e.g., 20 to 30 mbar. Thus it can be far weaker sorbent than the primary resin we are using for the capture of CO.sub.2, and this allows us to use very weak amines, or weak amine based resins. In the case that the resin is damaged by the resin wash, it is possible to let the amine rinse flow over other surfaces which due to being wetted by the resin can absorb a substantial amount of CO2.

(32) In yet another embodiment of the invention we could use the same type of resin as on the air capture device, but operate on a different level of the loading curve. This resin would stay moist and would be subject to a thermal swing for removal of CO.sub.2.

(33) While a high level of water vapor or moisture in the air should be sufficient for CO.sub.2 release, we could use liquid water in the CO.sub.2 release. However, the use of water vapor rather than liquid water could reduce wear and tear on the resin. It also makes it possible to use water that contains salts or other impurities.

(34) It is usually advantageous to remove excess water vapor from the CO.sub.2 water vapor mixture by pressurizing the gas so that the water vapor will become supersaturated and condense out. However, it is possible to remove the water by using a cold trap, which then in principle could be followed by a second cold trap for CO.sub.2. This method is of particular interest if the water condensation can at least be partially accomplished by cooling the heated resin back to lower ambient temperatures. If the apparatus initially worked with grey water or salt water, the water condensation could provide a separate stream of clean water that could be tapped as a byproduct.

(35) There are ways of reducing water consumption. For example one could send short pulses of steam through the resin, controlling the pulses to be too short for the steam to enter deeply in the resin, but sufficient to release a pulse of CO.sub.2 from the resin surface. Later on, diffusion inside the resin will redistribute the loading internal to the resin and the steam pulse could be repeated several times until the overall CO.sub.2 content has dropped sufficiently.

(36) A large fraction of the CO.sub.2 that has been collected on a dry resin can be removed again by exposing the dry resin either to water saturated air, or by immersion into liquid water. Particularly in an implementation that relies on a water swing the wear and tear pn the material is small, and the material can be cycled many times through this water swing. The amplitude of the swing may involve as much as half of the CO.sub.2 carrying capacity of the resin but practical designs may aim for slightly lower swings in loading.

(37) Using rough numbers for the membrane the rate of exhalation is about 1 milimole of CO.sub.2 for a 10 g sample in 15 minutes, or about 1e-4 milimole/g/sec. Since for the material we used the macroscopic surface area is about 40 cm.sup.2/g we have a release rate of 2.5e-5 mole/m.sup.2/sec. Uptake rates are comparable, although usually a little slower. It also means that a 50 g sample will produce about 0.1 cc/sec or about 100 cc of gas in about 15 minutes.

(38) Based on theory the kinetics of the progress are not slowed down by operating in a near vacuum where the nitrogen and oxygen of the air have in effect been removed from the system. Experiments have confirmed this observation. Under low pressure conditions it is possible to generate a gas stream from the wetted resin material that is essentially a mixture of water vapor in equilibrium with the temperature in the chamber, and CO.sub.2 that is being released from the resin. The CO.sub.2 pressure in the chamber could be as high as 60 millibar, but drops rapidly as the loading of the resin with CO.sub.2 is reduced. At room temperature partial pressure of water vapor are around 30 millibar. It is possible to raise the partial pressure of water vapor in the chamber by raising the temperature. This in effect creates a hybrid system that combines a humidity/water swing with a temperature swing. In a system that has been heated, one can reduce the water vapor pressure by cooling the output stream against ambient temperatures. This will reduce the water vapor concentration in the chamber down to the equilibrium pressure at the ambient temperature.

(39) Because it takes a long time for the CO.sub.2 release from the resin to reach equilibrium with the surrounding gas, the containment of the system should be constructed in a way to throttle the flow so that the slow release of CO.sub.2 can keep the system near CO.sub.2 partial equilibrium conditions. If the flow is allowed to move too fast then the rapid water vapor production kinetics will change the ratio of CO.sub.2 released to water vapor release in favor of excess water release. This would unnecessarily cool the system and thus would require additional heat input. If the water were to be removed in a condensation trap, the energy demand for cooling the gas stream would grow rapidly if the ratio of water vapor to CO.sub.2 in the gas mix turns unfavorable. A suitable mix of CO.sub.2 to water may have a ratio of 1:1. Thus it is preferred to maintain the system near the equilibrium point, which sets a maximum flow rate that is acceptable.

(40) Water can be removed by several ways from this low pressure mixture. The first is to cool the gas to condense the water out. As long as the temperature of the system is above ambient this incurs no practical energy penalty, but the penalty is severe if condensation is to occur under sub-ambient conditions. One way of avoiding this is to compress the H.sub.2O/CO.sub.2 gas mixture. As the partial pressure of H.sub.2O rises it will force condensation at temperatures above ambient conditions and thus it is possible to remove water with an extremely small energy penalty. Indeed in one set of implementations one can force condensations at temperatures above the (elevated) operating temperature of the chamber and thus provide a substantial fraction of the heat demand from the heat of condensation. In effect the system is run with an efficient heat pump that transfers heat of condensation into the system at elevated temperatures.

(41) Based on these conditions, we have designed a water (or humidity) swing system, where under dry conditions, CO.sub.2 is collected from the air. Based on experiments, it appears that the humidity level is mainly characterized by the absolute humidity and not the relative humidity. This is based on the observation that at constant absolute humidity, the CO.sub.2 level in a closed bell jar, where air is in contact with a resin, would not change as the temperature moves from near the dew point to much higher temperatures e.g. 10 to 20 C. higher. A change in humidity, on the other hand, would cause a large swing in the CO.sub.2 equilibrium level. This result suggests that a wet dry cycle is also possible in a moist/cool climate as well as a desert climate. Once the resin reaches a high level of saturation, which in a dry climate means it is essentially transferred from its carbonate form to a fully bicarbonated form, the resin material is packed into a chamber, which is first evacuated and then filled with water vapor and/or with liquid water which is agitated to the point that all resin surfaces are thoroughly wetted. The system then releases CO.sub.2 and, given enough material in the chamber, the system can reach a CO.sub.2 pressure that exceeds that of the water vapor in the chamber. In order to build up such a high pressure of CO.sub.2 (about 30 to 60 millibar) a chain of chambers may be provided, as in FIG. 5, that are saturated with water vapor near ambient conditions (about 30 mb of steam) but at varying temperatures so that the water saturation pressure drops from the upstream end (where steam is let in, to the downstream end, where CO.sub.2 and steam are removed.) In this system gas flow is maintained because water vapor and CO.sub.2 are extracted at the downstream end of the chain. The last chamber in the chain contains resin that has just returned from air saturation, which is capable of driving the partial pressure of CO.sub.2 as high or even higher than the water vapor pressure. Higher upstream the chambers contain resins which have been exposed for longer times to the warm humid environment.

(42) The chambers preferably will be nearly completely filled with resin material, since the larger the void space, the more difficult will it be to reach a high partial pressure of CO.sub.2 in the system. The chambers could physically move in the chain, as new resin is added at the gas outflow end of the chain, and CO.sub.2 depleted resin is removed at the other end. Rather than moving individual chambers, or cartridges that would fit into those chambers, we can also change the valving to the chambers. In this approach each chamber could be either connected to its downstream neighbor or the final extraction system. Each chamber would also be connected on the upstream side with the upstream neighbor or to the source of water vapor and sweep gas input. It is possible to valve out one unit, empty and refill it and valve it back in, so that it is now logically on the other end of the chain. In this way the resins can be moved through the chain without ever moving in location.

(43) Unless the CO.sub.2 is absorbed from the stream while leaving the water vapor unperturbed or unless both gas components are simply pumped out and compressed, the water vapor will have to be condensed out to be separated from the CO.sub.2 stream. Condensation could occur at the downstream end of the chain or more likely in the intermediate chambers where the temperature drops from chamber to chamber leading to lower water vapor pressures as the gas flow proceeds. Condensation of the water vapor will produce a flow toward the low pressure end of the system.

(44) Downstream from the last chamber, the CO.sub.2 may be removed through condensation/sublimation at extremely low temperatures (between 80 and 100 C.) or with the help of compressor pumps which move the gas into another reservoir at higher pressure. Alternatively, the CO.sub.2 may be removed by a sorbent, like cold liquid methanol, liquid amines, or sodium carbonate. Other options include hydrophobic versions of activated carbon.

(45) Referring to FIG. 6, in a single chamber with 6 g of SnowPure Anion Exchange material and after 15 minutes of washing the resin in DI water, the resin can still exhale CO.sub.2 and raise the CO.sub.2 level in the system to about 1200 ppm. Left alone over night, but allowed to dry, the resin recaptured all this CO.sub.2 and more, reducing the CO.sub.2 level in the container to about 40 ppm. Raising the humidity in the container to 30 ppm would lead to CO.sub.2 exhalation that reached above 1400 ppm. Releasing this CO.sub.2 and washing the resin strands in DI water leads to an even higher rate of release of CO.sub.2, raising the pressure in the bell to about 1800 ppm. A second iteration still achieved well over 1100 ppm and a third iteration still reached 700 ppm. In total the system exhaled approximately 3300 ppm. This translates into roughly 66 cc of CO.sub.2 or about of the CO.sub.2 capacity that was initially collected by the carbonate washed membrane. The total capacity of the membrane material is around 6 millimole or about 120 cc. As a result, it is believed that pure water alone can cause a large swing in CO.sub.2 binding to the matrix. It is not clear how much CO.sub.2 has been released from the matrix during the wash cycle outside of the bell jar measurement system. While the system spent several minutes in the water where a substantial amount of additional CO.sub.2 could have been removed, the data obtained toward the end of the run suggested that the rate of CO.sub.2 released dropped somewhat during the wash cycle, suggesting that the aqueous wash cycle has removed additional CO.sub.2 at a somewhat faster rate. However, the total amount of CO.sub.2 that has been removed during the wash cycle is bound to be relatively small considering how much has been emitted in the quiet periods in between.

(46) Using rough numbers for the membrane, the rate of exhalation is about 1 millimole of CO.sub.2 for a 10 g sample in 15 minutes, or about 1e-4 millimole/g/sec. For material having a macroscopic surface area of about 40 cm.sup.2/g we achieve a release rate of 2.5e-5 mole/m2/sec. Uptake rates are comparable, although usually a little slower. Experiments also show that a 50 g sample will produce about 0.1 cc/sec or about 100 cc of gas in about 15 minutes.

(47) A design feature is that there are a number of chambers in this sequence. The resins in the different chambers have been exposed for different length of time to the outgassing process. Resins may move from chamber to chamber, or more likely, the valving is changed to take a chamber from the purged end of the chain, remove its charge and fill it with a resin which is now put on the unpurged end of the chin. The gas in each chamber is composed of water vapor, CO.sub.2 and possibly an inert sweep gas. The sum of the three partial pressures is monotonically declining from the upstream end of the system to the downstream end of the system. The sweep gas pressure can be reduced by increasing the flow speed, but the water vapor pressure is more or less in equilibrium with the liquid water at this point. The CO.sub.2 pressure should increase in the direction of the sweep. If the water vapor is a large part of the total pressure, the water vapor pressure gradient controls the flow and it would be established by a temperature drop from one chamber to the next, while the CO.sub.2 pressure will rise from one chamber to the next, as each chamber is adding some CO.sub.2 to the flow. The contributions of each chamber will be limited by the rate at which the material can release CO.sub.2 and the equilibrium pressure that particular resin can reach. The resin on the upstream end has been nearly exhausted of all its CO.sub.2, thus adding very little to the total product; the last downstream unit can in principle add as much as 60 millibars. For this to happen the upstream end either has a steam pressure in excess of 70 millibars, or there is a substantial contribution of a sweep-gas that still would have to be separated from the CO.sub.2.

(48) One such sweep-gas that would be condensed out of the system might be methanol vapor, assuming that the last step of the system is a collection of CO.sub.2 in cold methanol. The other alternative is to rely on water, in which case the upstream temperature of the water has to be large enough to generate a partial pressure of at least 60 milibars. As the steam moves from one chamber to the next it will encounter cooler temperatures and thus it will partially condense out, creating a pressure profile that drives a net gas flow from the low CO.sub.2 end to the high CO.sub.2 end of the system.

(49) One implementation would condense out all remaining water and then pump away the CO.sub.2. Other implementations would rely on CO.sub.2 absorption and subsequent sorbent recovery. These sorbents could be cold methanol, liquid amines, or ammonia, or sodium carbonate. Other options include hydrophobic versions of activated carbon.

(50) Yet another invention is illustrated in FIG. 7. This invention is based on observation that activated carbon has been observed to absorb CO.sub.2 when wet, and release the absorbed CO.sub.2 as it dries. This is the opposite behavior from ion exchange resins. Accordingly, this makes it possible to couple ion exchange resins and activated carbon in sequence. Starting with dry activated carbon and moistened resin materials air is passed through the system. As the air dries the resin, it transports the water vapor to the carbon. The resin picks up CO.sub.2 as it dries, and the activated carbon picks up CO.sub.2 as it accepts moisture. Once the resin is dry, the system is reversed, and fresh air is flowed through the activated carbon, and releases moisture back to the ion exchange resins. As the carbon dries it gives off CO.sub.2, raising the CO.sub.2 partial pressure where it can be concentrated and removed. A feature and advantage of coupling ion exchange material and activated carbon in this manner is that water is preserved, and is a simple matter of valving to reverse air flow.

(51) The invention is susceptible to modification. For example, other materials such zeolite may be used in place of the activated carbon.

(52) It is thus seen that using ion exchange resins, moisture swing can be used to separate and capture ambient CO.sub.2. By stringing several air capture devices together, the ambient CO.sub.2 removed may be concentrated.

(53) Using ion exchange resins to remove CO.sub.2 alone or in combination with activated carbon or zeolite in sequence, has the advantage that harsh chemicals such as sodium hydroxide typically used in CO.sub.2 scrubbers in the past, can be avoided. Also, using humidity swing in the essentially closed systems, water is preserved.

(54) Referring to FIG. 8, a CO.sub.2 air capture device could be comprised of a plurality of upright cylindrical towers 100 containing CO.sub.2 capture material. Preferably, but not necessarily, the CO.sub.2 capture material comprises humidity sensitive ion exchange material as previously described. Each tower 100 has a plurality of air inlets 102 adjacent the lower end thereof and a plurality of air outlets 104 adjacent a top end. Alternatively, the walls of the towers may be open to let air pass through at any level above the ground. Air is taken in the inlets 102 where it flows upwardly in the towers 100 in contact with the CO.sub.2 capture material. Airflow may be by natural convective flow, or, if desired, powered by fans. Inlets 102 and outlets 104 may be selectively closed off by shutters (not seen) so that humidity within the towers may be raised and the captured CO.sub.2 removed from the resin, and concentrated. Alternatively, the resin may be washed with liquid water to remove captured CO.sub.2 which is concentrated in the base in which is also contained plumbing, and valving, etc., (not shown). Alternatively, the CO.sub.2 capture material may comprise a wet sorbent such as an amine or carbonate or other CO.sub.2 sorbent, which is pumped up and dribbled or flowed down a matrix material in the tower. In a preferred embodiment of the invention, several towers 100 are connected together underground where liquid effluent from the towers is processed to remove the CO.sub.2, and the removed CO.sub.2 is concentrated. Yet other changes are possible.

(55) In another aspect of the present invention, the functionalized solid structure in the collector may be optimized for gas exchange, steam exchange, liquid exchange and mechanical performance. For gas exchange, the pore-spaces/tunnels will be large enough to keep the pressure drop low, minimize turbulence, and to prevent the boundary layer from increasing to a thickness that hinders CO.sub.2 supply to the functional groups. Where the gas stream is ambient air, for example, if one liter of functionalized solid has one mol of exchange sites, it will take about 22.4 liters of pure CO.sub.2 to saturate those sites with bicarbonate. If we consider an average uptake of 100 parts per million Volume (ppmV) of CO.sub.2 per air volume, we will need to move 224,000 liters of air through one liter of solid to saturate the functional groups. This would require that the air flow is decreased as the solid saturates, since the CO.sub.2 uptake decreases with saturation.

(56) For steam and liquid exchange, the solid to void ratio typically is larger. For example, in commercially available water treatment systems, the typical ion exchange resin bed has an approximate void space of 30 to 40%. The beads are about 0.5 millimeters in size, leaving even smaller interstitial pore spaces. Therefore, for many applications, the collector should be a convertible structure with a primary and secondary porosity, which differ in size.

(57) The design also includes a fixed power size system for cases where gas, steam, and liquid exchange use the same pore-size or when mechanical considerations favor a fixed power-space system.

(58) In another aspect, the present invention provides improvements in the geometry of a collector that utilizes a functionalized solid collector medium to remove CO.sub.2 from gas streams. The capture medium can for example consist of amine groups immobilized on/in a solid.

(59) Referring to FIGS. 9A and 9B, the present invention provides an adjustable-pore-size collector in roll-geometry, comprising an inert center shaft 205 with locking mechanism; inert mesh or other porous medium in sheet form 206; and functionalized collector medium in sheet form 207. One porous sheet and one collector medium sheet are fastened onto and wound around the shaft such that the innermost and the outermost layer of the roll consist of a porous sheet. This ensures that the collector medium is always sandwiched between porous sheets to allow contact with gas, steam or liquid on both sides of the collector medium sheet. Depending diameter of the roll, more than one membrane and porous layer are used, so that widening of the pore space requires less rotation by the center shaft. E.g. three layers require a third of the rotation of one layer.

(60) The porous sheet provides the pathway for the gas, steam, and/or liquid stream. This flow path can be straight, spiraling, tortuous, laminar or turbulent design. The porous sheet also sets the minimum width of the pore space, which is achieved when the roll is tightened around the shaft. The pore space can be widened by turning the shaft against the direction of the coil or narrowed by turning the shaft in the direction of the coil. The shaft and porous sheet are manufactured from materials that are inert with regard to the gas, steam and liquid streams to which they are exposed.

(61) In an alternative embodiment, shown in FIGS. 10A and 10B, the present invention provides an adjustable-pore-size collector in flat-sheet geometry comprising inert mesh or other porous medium in sheet form 211 or as a compressible layer 212 and functionalized collector medium in sheet form 213. The flat-sheet geometry collector has alternating layers of mesh or other porous medium and collector medium. During fluid flow/exchange the layers are pressed together to achieve the smallest flow channel thickness between the collector medium sheets. The minimum thickness of the flow channel and the void space within this channel is determined by the porous medium that is used. During gas flow/exchange the flow channel can be adjusted to the optimum thickness by pulling the sheets apart. The optimum spacing is determined by the exchange rate of the collector medium and the liquid or gas stream, as well as the flow rate of the liquid and gas stream.

(62) In another alternative embodiment shown in FIGS. 11A and 11B, the present invention provides a collapsible collector having a primary and a secondary porosity. The primary porosity is optimized for gas flow. The walls of the channels used for gas flow are constructed of the collector medium. The collector medium itself has a secondary porosity that is optimized for fluid flow. Once gas-exchange is complete and the collector medium is saturated with CO.sub.2, the primary porosity is collapsed, leaving only the secondary porosity for steam or fluid exchange.

(63) The simplest geometry for this collector is a series of medium collector sheets with open-cell pores. For gas exchange the sheets are spaced apart at a distance optimized for the CO.sub.2 uptake rate of the collector medium and the length of the sheets. For steam and fluid exchange the sheets are pushed together and fluid flows through the porosity of the collector medium. Other geometries are shown in FIGS. 12A-12F, and include collapsible square or rectangular channels, etc.

(64) The fixed collector will employ one or several fixed pore sizes. The pore geometry varies. Examples of pore geometry are round, square, rectangular, triangular, hexagonal, or differently shaped channels. The design also includes open cell foam geometry, or other irregular connected shapes determined by the form of the collector medium.

(65) In another aspect of the present invention shown in FIG. 13, the ion exchange process may be completed using an electrodialysis (ED) cell. The ion exchange resin may be washed using a basic solution, such as sodium carbonate (Na.sub.2CO.sub.3), preferably having a pH of 11-12. The resulting effluent, which in the example of a sodium carbonate wash will be primarily sodium bicarbonate (NaHCO.sub.3), will preferably have a pH of 9-10.

(66) The effluent is then supplied to the acid side of an ED cell, where the reaction is controlled through bipolar and cationic membranes. After an initial run, the acidic side of the cell stabilizes at a near neutral pH, at which point CO.sub.2 evolves and is captured. Osmotic pressure drives water towards the base side of the cell. The basic solution is maintained near a pH of 12 and may also be used to replenish the wash fluid.

(67) Because of the differences in pH and composition of the wash fluid and effluent, separate basins may be used for each.

(68) Alternatively, as shown in FIG. 14, the basic solution for the ED cell and the wash fluid may be contained in the same basin. This basin may contain a separator for separating two basic solutions having two distinct pH levels. In a preferred embodiment, the pH for the wash fluid is 12 and the pH for the basic solution supplied to the ED cell is 11. The effluent may be contained in a separate basin prior to entering the acid side of the ED cell. Also, the effluent, which preferably is at a pH of 9-10, may be used to dilute the portion of the basic solution to be used for the wash to a pH of 11.

(69) It should be emphasized that the above-described embodiments of the present process, particularly, and preferred embodiments, are merely possible examples of implementations and merely set forth for a clear understanding of the principles of the invention. Many different embodiments of the processes described herein may be designed and/or fabricated without departing from the spirit and scope of the invention. All these and other such modifications and variations are intended to be included herein within the scope of this disclosure and protected by the following claims. Therefore the scope of the invention is not intended to be limited except as indicated in the appended claims.