Method of Capturing a Target Species From a Gas

Abstract

A method of capturing a target species from a gas comprises the steps of: contacting a gas containing a target species with a first absorbent solution comprising a capture species; dissolving the target species in the first absorbent solution to form a target anion; electrochemically separating the target anion from the first absorbent solution by contacting the first absorbent solution with one or more ion-exchange membranes, and transferring the target anion through an ion-exchange membrane into a second absorbent solution; and releasing at least some of the target species from the second absorbent solution. The one or more ion-exchange membranes are not permeable to the capture species, so the capture species does not pass through the one or more ion-exchange membranes. An apparatus for capturing a target species from a gas is also provided.

Claims

1. A method of capturing a target species from a gas comprising the steps of: contacting a gas containing a target species with a first absorbent solution comprising a capture species; dissolving the target species in the first absorbent solution to form a target anion; electrochemically separating the target anion from the first absorbent solution by contacting the first absorbent solution with one or more ion-exchange membranes, and transferring the target anion through an ion-exchange membrane into a second absorbent solution; and releasing at least some of the target species from the second absorbent solution, in which the capture species is too large to pass through the pores in the one or more ion-exchange membrane, so the capture species does not pass through the one or more ion-exchange membranes.

2. The method according to claim 1, in which the second absorbent solution does not contain the capture species.

3. The method according to claim 1, in which the capture species binds to the target anion in the first absorbent solution, and in which the target anion is electrochemically dissociated from the capture species before being transferred through the ion-exchange membrane.

4. The method according to claim 1, in which the capture species is a cationic capture species that does not comprise an alkali metal cation.

5. The method according to claim 1, in which the capture species is an ionic capture species.

6. The method according to claim 1, in which the capture species is an ionic polymer.

7. The method according to claim 1, in which the capture species is a choline-derived ionic liquid.

8. The method according to claim 1, in which the capture species is a cationic polymer.

9. The method according to claim 1, in which the capture species is a polymeric amine.

10. The method according to claim 1, in which the capture species comprises polyethyleneimine (PEI).

11. The method according to claim 1, in which the capture species has a molecular weight of greater than or equal to 200 g/mol.

12. The method according to claim 1, in which the one or more ion-exchange membranes are configured to permit passage of the target anion therethrough, and to prevent passage of capture species having a molecular weight of greater than 200 g/mol.

13. The method according to claim 1, in which the capture species is weakly basic.

14. The method according to claim 1, in which the first absorbent solution contains a hydration catalyst for accelerating the conversion of the dissolved target species into the target anion.

15. The method according to claim 1, in which at least one of the ion-exchange membranes is an anion-exchange membrane permeable to the target anion.

16. The method according to claim 1, in which the target species is dissolved in the first absorbent solution to form a target anion and a hydrogen cation (H.sup.+), and in which the one or more ion-exchange membranes comprise an anion-exchange membrane permeable to the target anion, and a cation-exchange membrane permeable to the hydrogen cation.

17. The method according to claim 1, in which the target anion is combined with a hydrogen cation to form a target acid in the second absorbent solution.

18. The method according to claim 17, in which the hydrogen cation is produced by electrolysing H.sub.2O.

19. The method according to claim 1, in which the second absorbent solution has a pH which is different from the pH of the first absorbent solution.

20. The method according to claim 1, in which the first absorbent solution is an aqueous solution and the second absorbent solution is a non-aqueous solution.

21. (canceled)

22. The method according to claim 1, comprising one or more flow electrodes in contact with the one or more ion-exchange membranes.

23. The method according to claim 1, in which the step of electrochemically separating the target anion from the first absorbent solution comprises capacitive deionisation (CDI), flow-CDI, or electrodialysis.

24. The method according to claim 1, in which the target species is released from the second absorbent solution as a gas, in order to maintain the chemical equilibrium of the target acid in the second absorbent solution.

25. The method according to claim 1, in which at least some of the target anions in the second absorbent solution are reacted with a mineral or salt to form a precipitated material that is released from the second absorbent solution.

26. The method according to claim 1, in which the target species is selected from the group consisting of CO.sub.2, H.sub.2S, SO.sub.2, NO, NO.sub.2, and N.sub.2O.

27. The method according to claim 1, in which the target species is CO.sub.2, the target anion is bicarbonate, and the target acid is carbonic acid.

28. (canceled)

29. (canceled)

30. The method according to claim 1, in which the capture species is a cationic capture species.

31. The method according to claim 12, in which the one or more ion-exchange membranes are configured to permit passage of the target anion therethrough, and to prevent passage of capture species having a cationic charge.

Description

DETAILED DESCRIPTION

[0209] Specific embodiments of the invention will now be described by way of example, with reference to the accompanying drawings, in which:

[0210] FIG. 1 is a schematic diagram of a flow electrode capacitive deionisation apparatus operating according to a preferred embodiment of the present invention;

[0211] FIG. 2A is a schematic illustration of a hybrid flow-CDI-electrodialysis apparatus usable in a preferred embodiment of the present invention;

[0212] FIG. 2B is a schematic illustration of an alternative electrodialysis apparatus usable in a preferred embodiment of the present invention;

[0213] FIG. 3 is a schematic illustration of an electrolyzer-electrodialysis apparatus usable in a preferred embodiment of the present invention;

[0214] FIG. 4 is a graph of pH change vs time, on sparging air at 1 L per minute through a solution according to a preferred embodiment of the present invention;

[0215] FIG. 5 is a graph of captured CO.sub.2 vs time, on sparging air at 1 L per minute through a solution according to a preferred embodiment of the present invention; and

[0216] FIG. 6 is a graph of salt captured vs time in a flow-CDI cell, according to a preferred embodiment of the present invention;

[0217] FIG. 7 is a graph comparing CO.sub.2 capture rate for free bovine carbonic anhydrase (free bCA), bCA immobilised on Fe.sub.3O.sub.4, and Fe.sub.3O.sub.4;

[0218] FIG. 8 is a graph comparing average capture efficiency and max capture efficiency for free bovine carbonic anhydrase (free bCA), bCA immobilised on Fe.sub.3O.sub.4, and Fe.sub.3O.sub.4;

[0219] FIG. 9 illustrates the chemical structure of an exemplary branched Polyethyleneimine (PEI) polymer chain;

[0220] FIG. 10 is a schematic illustration of an electrolyzer-electrodialysis apparatus usable in a preferred embodiment of the present invention, in which polyethyleneimine (PEI) is used as the first absorbent solution;

[0221] FIG. 11 is a graph of CO.sub.2 captured with an H.sub.2O/PEI solution;

[0222] FIG. 12 is a graph of CO.sub.2 released from an H.sub.2O/PEI solution;

[0223] FIG. 13 is a graph comparing CO.sub.2 capture rates of NaOH, bCA in H.sub.2O/PEI, and bCA immobilised on Fe.sub.3O.sub.4 in in Na.sub.2HPO.sub.4;

[0224] FIG. 14 is a graph comparing the capture rate and capture efficiency of aqueous absorbents PEI/CA, NaOH and Na.sub.2CO.sub.3;

[0225] FIG. 15 is a graph of CO.sub.2 released, and of the CO.sub.2 separation and release rate, for continuous capture of CO.sub.2 using PEI absorbent solution using the apparatus illustrated in FIG. 10;

[0226] FIG. 16 is a graph illustrating the CO.sub.2 separation and release rate for continuous capture of CO.sub.2 using PEI absorbent solution using the apparatus illustrated in FIG. 10;

[0227] FIG. 17 is a schematic illustration of an alternative electrodialysis apparatus usable in a preferred embodiment of the present invention;

[0228] FIG. 18 is a graph of the CO.sub.2 capture rate, and the energy consumption measured using the apparatus of FIG. 17;

[0229] FIG. 19 is a schematic illustration of an apparatus according to a preferred embodiment of the present invention;

[0230] FIG. 20 is a graph of the CO.sub.2 exit concentration from a falling film reactor usable in a preferred embodiment of the present invention;

[0231] FIG. 21 is a graph of the CO.sub.2 exit concentration from the falling film reactor used in Figure across a range of temperatures and for two different capture species;

[0232] FIG. 22 is a is a schematic illustration of an electrodialysis apparatus usable in a preferred embodiment of the present invention;

[0233] FIG. 23 is a photograph of an electrodialysis cell usable in an embodiment of the present invention;

[0234] FIG. 24 is a graph showing CO.sub.2 release rate as a function of voltage/current measured using the apparatus of FIG. 23;

[0235] FIG. 25 is a graph showing the total energy consumption and energy of H.sub.2 produced in the same experiment as FIG. 24;

[0236] FIG. 26 is a graph showing CO.sub.2 stability data obtained using an electrodialysis cell;

[0237] FIG. 27 is a graph of CO.sub.2 output rates in mg/hr (black) vs energy consumption for CO.sub.2 separation and release in kWh/tCO.sub.2 (red) for a 40-membrane-pair electrodialysis cell;

[0238] FIG. 28 is a graph of the modelled electrical energy demand of electrodialysis using the same 40-membrane-pair electrodialysis cell as FIG. 27 with different concentrations of capture species in the first absorbent solution.

[0239] FIG. 1 illustrates a preferred embodiment of the present invention which employs flow electrode capacitive deionisation to electrochemically separate ions as part of a gas capture process.

[0240] The flow-CDI apparatus 100 illustrated in FIG. 1 is made up of a gas contactor 102, an ion-separator 104, and a release vessel 106.

[0241] The gas contactor 102 is arranged to receive a flow of gas 108 which contains a target species to be captured, and to bring the gas into contact with a stream of a first absorbent solution 110.

[0242] A variety of gas-liquid contactor designs are known in the art, such as falling-film columns, packed columns, bubble columns or spray towers, any of which would be suitable for use with the present invention.

[0243] The ion-separator 104 contains a separation chamber 112 that is filled with a porous solid electrolyte, an anion-exchange membrane 114 along one side of the separation chamber 112, and a cation-exchange membrane 116 along the opposite side of the separation chamber 112. An inlet pipe 118 connects an outlet of the gas contactor 102 to an inlet of the separation chamber, and an outlet pipe 120 connects an outlet of the separation chamber to an inlet of the gas contactor, so that a stream of first absorbent solution 110 can be pumped from the gas contactor, through the separation chamber, and then recirculated to the gas contactor.

[0244] A positive electrode 122 is connected to the ion-separator 104 on the side of the anion-exchange membrane, and a negative electrode 124 is connected to the ion-separator 104 on the side of the cation-exchange membrane.

[0245] The ion-separator comprises a first flow electrode channel 126 between the anion-exchange membrane and the positive electrode 122, and a second flow electrode channel 128 between the cation-exchange membrane and the negative electrode 124.

[0246] One end of a flow electrode outlet pipe 130 is connected to outlets of both the first flow electrode channel 126 and the second flow electrode channel 128, and the other end of the flow electrode outlet pipe 130 is connected to an inlet of the release vessel 106. A flow electrode inlet pipe 132 is connected between an outlet of the release vessel 106 and inlets of both the first flow electrode channel 126 and the second flow electrode channel 128.

[0247] Flow electrodes are formed by pumping a second absorbent solution 134 containing a suspension of electrically-conductive particles through both flow electrode channels 126, 128 and into the release vessel, and recirculating the second absorbent solution 134 from the release vessel 106 to the flow electrode channels 126, 128.

[0248] In use, a flow of gas 108 which contains a target species to be captured is introduced into the gas contactor 102, at the same time that a first absorbent solution 110 containing a capture species is introduced into the gas contactor. As the gas 108 comes into contact with the first absorbent solution 110, mass transfer of the target species into the absorbent solution takes place, so that the first absorbent solution 110 absorbs some of the target species from the gas.

[0249] The target species is dissolved in the first absorbent solution 110, optionally assisted by the presence of a hydration catalyst in the first absorbent solution, and forms a target anion and a hydrogen cation. The target anion and the hydrogen cation together form a target acid, but in the first absorbent solution 110 the target anion may bind to, or associate with, the capture species.

[0250] The first absorbent solution 110 is continuously pumped from the outlet of the gas contactor 102, through the inlet pipe 118, to the inlet of the separation chamber 112 of the ion-separator 104, from where the liquid first absorbent solution 110 flows through the porous solid electrolyte.

[0251] During operation, a potential difference is applied between the positive electrode 122 and the negative electrode 124. This potential difference across the ion-separator means that as the first absorbent solution 110 flows through the separation chamber, the negatively-charged target anions are dissociated from the capture species and attracted towards the positive electrode, while the positively-charged hydrogen cations are attracted towards the negative electrode. The target anions therefore migrate through the anion-exchange membrane 114, and the hydrogen cations flow through the cation-exchange membrane 116, so that the target acid ions are separated from the first absorbent solution. Neither the anion-exchange membrane 114 nor the cation-exchange membrane 116 is permeable to the capture species, so the capture species remains in the first absorbent solution 110.

[0252] By the time that the first absorbent solution 110 reaches the outlet of the separation chamber 112, at least some of the target acid ions (target anions and hydrogen cations) have been separated from the stream of first absorbent solution 110, and the first absorbent solution is recirculated through the outlet pipe 120 to the inlet of the gas contactor 102.

[0253] During operation, a stream of the second absorbent solution 134 containing a slurry of conductive particles is pumped through the first flow electrode channel 126 and the second flow electrode channel 128, so that target anions and hydrogen cations passing through the ion-exchange membranes are transferred into the stream of second absorbent solution 134. The target anions and hydrogen cations are recombined in the flow electrode outlet pipe 130, as they flow to the release vessel 106, and reassociate with one another so that the stream of second absorbent solution 134 contains the target acid when it reaches the release vessel 106.

[0254] Once in the release vessel 106, at least some of the target species is released from the second absorbent solution as a gas. This is preferably driven solely by equilibrium, and the target species gas preferably evolves from the second absorbent solution at room temperatures and pressures, without requiring additional heating or the use of a gas stripper.

[0255] The released gas of the target species can then be removed from the release vessel 106 and compressed, stored or reacted as desired.

[0256] In order to provide continuous flow electrodes, a stream of the second absorbent solution 134 is pumped back to the inlets of the inlets of both the first flow electrode channel 126 and the second flow electrode channel 128 through a flow electrode outlet pipe 130.

[0257] Using this system, the target species can be continuously absorbed from the flow of gas 108, transferred from the first absorbent solution to the second absorbent solution, and eventually released in the release vessel 106.

[0258] A particularly preferred embodiment of the invention that can be performed using this set-up is the capture of carbon dioxide (CO.sub.2) from air.

[0259] In this embodiment, air is used as the flow of gas 108, and the first absorbent solution 110 is an aqueous solution containing a CO.sub.2 hydration catalyst.

[0260] A particularly preferred option for the first absorbent solution and capture species in this embodiment is an aqueous solution polymer resin particles functionalised with cationic functional groups, for example Lewatit R VP OC1065, containing a hydration catalyst of carbonic anhydrase.

[0261] As air is introduced to the gas contactor 102 and brought into contact with the solution of cationic polymer particles and carbonic anhydrase, CO.sub.2 from the air is absorbed by the solution and hydrated, in order to form carbonic acid (bicarbonate anions and hydrogen cations) according to the following equilibrium:

CO.sub.2+H.sub.2OH.sub.2CO.sub.3H.sup.++HCO.sub.3.sup.

[0262] The bicarbonate anions bind to the weakly basic cationic polymer particles, while the free hydrogen cations reduce the pH of the first absorbent solution.

[0263] When the stream of first absorbent solution 110 reaches the ion-separator 104, the hydrogen cation is separated from the first absorbent solution through the cation-exchange membrane (for example Nafion), and the bicarbonate acid anion (HCO.sub.3.sup.) dissociates from the cationic polymer particles and migrates through the anion-exchange membrane (for example Sustainion). Neither ion-exchange membrane is permeable to the cationic polymer resin particles, so the capture species remains in the first absorbent solution. Both target anions and hydrogen cations are then transferred into the stream of second absorbent solution 134 flowing through the flow electrode channels, and recombined to form carbonic acid. In this embodiment, the second absorbent solution does not contain any of the capture species, nor any other cationic species to which the bicarbonate anions can bind.

[0264] A preferred second absorbent solution 134 for use in this embodiment is a non-aqueous solution of dimethyl carbonate containing a suspension of activated carbon nanoparticles to act as the flow electrode.

[0265] A particular benefit of this embodiment is that carbonic acid and its ions (H+ and HCO.sub.3.sup.) are naturally metastable at room temperature. In order to prevent the formation of gas bubbles in the ion-separator, the ion-separator is pressurised to a pressure at which gas bubbles cannot form. As the stream of second absorbent solution 134 containing carbonic acid arrives at the release vessel 106, the pressure is decreased. As the ions are electrochemically separated and introduced into the second absorbent solution, equilibrium begins to favour CO.sub.2 due to the increasing concentration of H+ and HCO.sub.3 in the second absorbent solution. The carbonic acid ions thus become naturally inclined to dehydrate to form gaseous CO.sub.2, which is then released from the second absorbent solution in the release vessel.

[0266] This release step may be carried out entirely at room temperature and pressure. Energy-intensive heating to the high temperatures used by the prior art, for example 90-100 C. for gas strippers, is not required, making the process much more environmentally-friendly. The use of a non-aqueous second absorbent solution also advantageously means that the released carbon dioxide gas does not have a high humidity and does not require the energy-intensive subsequent drying step that is part of some prior art methods.

[0267] Using this method, dilute CO.sub.2 gas in air (in quantities far below 1 vol %) may be captured out of the air and concentrated as pure CO.sub.2 gas.

[0268] The same apparatus and the same technique may alternatively be used to capture other target species from air, or from another gas source. For example the target species may be H.sub.2S, SO.sub.2, NO, NO.sub.2, and N.sub.2O. In order to capture different target species, different hydration catalysts may be used, and the target species would form the conjugate acids of the target species. For certain target species, the second absorbent solution may be heated to encourage release of the target species, or the target species may be concentrated to a predetermined molarity in the second absorbent solution and then discharged.

[0269] FIGS. 2A and 2B illustrate two preferred embodiments of the present invention which employ an electrodialysis stack 200 as an alternative ion-separator to electrochemically separate ions as part of a gas capture process.

[0270] In FIG. 2A, the apparatus is a hybrid of flow-CDI and electrodialysis which uses flow electrodes, while in FIG. 2B, the electrodialysis apparatus 200 does not use flow electrodes.

[0271] The electrodialysis stack 200 may be used with the gas contactor 102 and the release vessel 106 described above.

[0272] The electrodialysis stack 200 contains a separation chamber 212 that is filled with a porous solid electrolyte, between a positive electrode (anode) 222 and a negative electrode (cathode) 224. Three pairs of anion-exchange membranes 114 and cation-exchange membranes 116 are arranged in parallel between the electrodes, dividing the separation chamber 212 into seven adjacent compartments between the two electrodes. The two outermost compartments are formed by an electrode and an ion-exchange membrane, while the five intervening channels A, B are formed by pairs of opposing ion-exchange membranes.

[0273] First absorbent solution 210 is pumped through channels A, while second absorbent solution 234 is pumped through channels B.

[0274] During operation, a potential difference is applied between the anode 222 and the cathode 224, and liquid first absorbent solution 210 containing target anions and hydrogen cations is pumped into one end of four of the adjacent channels A.

[0275] For the purposes of illustration, FIG. 2 shows the counterions of carbonic acid (bicarbonate anions and hydrogen cations) being separated by the electrodialysis stack 200, but the same apparatus may be used with alternative target anions and target acids.

[0276] As the first absorbent solution 210 containing a capture species, target anions and hydrogen cations flows through the channels A, the electrical field between the electrodes attracts the target anions (HCO.sub.3 in the illustrated case of carbonic acid) towards the positive electrode 222, and the acid cations (H.sup.+) towards the negative electrode 224. Thus anions are dissociated from the capture species and migrate out of the channel A and into an adjacent channel B by passing through the anion-exchange membrane 114 contacting the channel A, while cations migrate in the other direction out of the channel A and into an adjacent channel B by passing through the cation-exchange membrane 116. Neither ion-exchange membrane is permeable to the capture species, so the capture species remains in the first absorbent solution.

[0277] The target anions and hydrogen cations are concentrated in the channels B, as once anions migrate into a channel B they are prevented from migrating further towards the anode as they cannot pass through the cation-exchange membrane 116 forming one side of the channel B. Likewise, cations migrating in the other direction are trapped in the channel B by the anion-exchange membrane 114. The target anions and hydrogen cations therefore associate to form a target acid that is the conjugate acid of the target species.

[0278] As described above, the electrodialysis stack 200 is maintained under an elevated pressure at which gas bubbles of the target species cannot form, as bubble formation inside the cell may damage one or more membranes and harm performance.

[0279] Using this arrangement, the target acid counterions can be concentrated in streams of the second absorbent solution 234 in channels B, which are then circulated to the release vessel, so that the target species can be released from the second absorbent solution 234.

[0280] The flow electrodes in FIG. 2A are made up of a separate third solution containing a suspension of conductive particles, which is recirculated between the electrodes and kept separate from the first and second absorbent solutions. They ensure ions can continuously migrate into each compartment by picking up target anions (bicarbonate) at the anode and dropping it off at the cathode.

[0281] FIG. 3 is a schematic illustration of an electrolyzer-electrodialysis apparatus 300 usable in a preferred embodiment of the present invention. Figure is described in relation to bicarbonate ions and carbonic acid for illustration.

[0282] The electrodialysis apparatus 300 contains a second absorbent chamber 312 that is filled with a porous solid electrolyte, between a positive electrode (anode) 322 and a negative electrode (cathode) 324. A pair of ion-exchange membranesan anion-exchange membrane 114 and a cation-exchange membrane 116are arranged in parallel between the electrodes, dividing the apparatus 300 into three compartments: a cathodic compartment 330 on an inlet side of the anion-exchange membrane 114, an anodic compartment 340 on an inlet side of the cation-exchange membrane 116, and the second absorbent chamber 312 between the two membranes.

[0283] During operation, a potential difference is applied between the anode 322 and the cathode 324. Liquid first absorbent solution 310 containing a capture species, bicarbonate anions and hydrogen cations is pumped from a capture vessel (not shown) into the cathodic compartment 330 and circulated around the cathodic chamber. A second absorbent solution 334, which in this example is an aqueous flow of electrolyte, for example sodium sulfate or sodium chloride, is pumped through the second absorbent chamber 312. H.sub.2O is pumped into the anodic compartment 340.

[0284] As the cathode is in an alkaline environment, but also contains hydrogen ions formed by dissolution of the target species in the first absorbent solution, two reactions take place. At the negatively charged cathode, a reduction reaction takes place, with electrons (e.sup.) from the cathode combining with hydrogen cations to form hydrogen gas. The reduction reaction taking place at the cathode is: 2H.sub.2O (I)+2.sub.e.sup..fwdarw.2OH.sup. (aq)+H.sub.2 (g) (OH.sup. is mostly neutralised by HCO.sub.3 to form carbonate), while the reaction 2H.sup.++2e.sup..fwdarw.H.sub.2 also occurs to evolve hydrogen gas.

[0285] At the positively charged anode, an oxidation reaction occurs, generating oxygen gas and giving electrons to the anode to complete the circuit. The reaction taking place at the anode is: 2H.sub.2O (I).fwdarw.O.sub.2 (g)+4e.sup.+4H.sup.+ (aq).

[0286] As the first absorbent solution 310 containing the target anions (HCO.sub.3 bicarbonate anions in the illustrated case of carbonic acid) flows into the cathodic compartment 330, the electrical field between the electrodes attracts the target anions through the anion-exchange membrane 114, and the hydrogen cations (H.sup.+) in the first absorbent solution are released as hydrogen gas. Thus target anions migrate out of the cathodic compartment 330 and into the second absorbent chamber 312 by passing through the anion-exchange membrane 114. At the same time, hydrogen cations (H.sup.+) formed by electrolysis at the anode are attracted through the cation exchange membrane 116, and migrate into the second absorbent chamber 312.

[0287] Using this arrangement, the target anions (HCO.sub.3.sup.) are combined with hydrogen cations to form the target acid (carbonic acid in the illustrated example) and concentrated in the streams of the second absorbent solution 334 through the second absorbent chamber 312. The second absorbent chamber 312 is maintained at a sufficiently high pressure that the target acid does not decompose to form gas bubbles inside the apparatus 300.

[0288] The stream of second absorbent solution is then circulated to the release vessel (not shown). Once in the stream of second absorbent solution 334, the pressure is reduced and the carbonic acid counterions decompose to form gaseous CO.sub.2, which is released from the second absorbent solution 334 and collected in the release vessel.

[0289] While prior art documents such as EP2163294 have employed electrodialysis for CO2 capture, in EP2163294 water dissociation is carried out by bipolar membranes (BPMs), which exhibits limited stability, having to perform highly reductive and highly oxidative reactions simultaneously.

[0290] The cell configuration in FIG. 3 also eliminates the use of bipolar membranes which are approximately ten times more expensive than anion-exchange membranes.

[0291] The cell configuration illustrated in FIG. 3 may preferably be scaled up by adding a plurality of pairs of ion-exchange membranes to increase the quantity of CO.sub.2 released per molecule of hydrogen and oxygen. The greater the number of pairs of ion-exchange membranes, the more CO.sub.2 will be released per molecule of H.sub.2/O.sub.2 generated.

[0292] The nature of the cell design shown in FIG. 3 means that hydrogen may be produced with efficiencies competitive with current PEM electrolysers, for example 50-60 kWh/kg (of H.sub.2) efficiency.

[0293] FIGS. 4 to 6 illustrate experimental data obtained by the inventors in relation to the direct air capture of carbon dioxide using the method of the present invention.

[0294] FIG. 4 illustrates the measured pH change over time of three potential first absorbent solutions, on sparging air through each solution at a rate of 1 L per minute. This experiment demonstrates the effectiveness of:

[0295] Line 400: Na.sub.2HPO.sub.4 (0.1M, 100 mL).

[0296] Line 420: Na.sub.2HPO.sub.4 (0.1 M, 100 mL)+0.2 mg mL.sup.1 of bovine carbonic anhydrase.

[0297] Line 440: Na.sub.2HPO.sub.4, (0.1 M, 100 mL)+0.2 mg mL.sup.1 equivalent immobilised bovine carbonic anhydrase on Fe.sub.3O.sub.4 particles.

[0298] Immobilised enzymes used in the present invention may preferably be immobilised onto particles that are dispersed throughout the first absorbent solution as a suspension. In particularly preferred embodiments, magnetic Fe.sub.3O.sub.4 particles have been used as carrier particles on which the enzyme, for example bCA is immobilised.

[0299] While Na.sub.2HPO.sub.4 is not a preferred capture species for the present invention, in this case Na.sub.2HPO.sub.4 was used as a control to demonstrate the relative efficacy of CA and immobilised CA on carbon dioxide capture.

[0300] As shown by FIG. 4, the pH of all three solutions decreased significantly, from around pH 9.1-9.5, to around pH 8.3 to 8.5, as air was sparged through the solutions. This results from the absorption of CO.sub.2 from the air, and the conversion of the dissolved CO.sub.2 into carbonic acid (bicarbonate anions and hydrogen cations). The acidic pH (<7) of the carbonic acid formed during this process naturally lowers the overall pH of the first absorbent solution, so that all three solutions eventually had a pH of 8.5 or less at equilibrium.

[0301] The comparative gradients of the three lines on FIG. 4 shows that the presence of carbonic anhydrase in the solution significantly increased the speed at which the solutions absorbed CO.sub.2 and converted it to the ions of carbonic acid, with the solutions 420, 440 containing carbonic anhydrase reaching equilibrium far more quickly than the control sample 400 without any hydration catalyst.

[0302] FIG. 4 also shows that immobilised carbonic anhydrase in sample 440 was more effective than sample 420 in which the carbonic anhydrase was not immobilised. Sample 440 absorbed CO.sub.2 and converted it to carbonic acid more quickly, and also reached equilibrium at a lower pH of around 8.3, suggesting that the immobilised catalyst caused the sample 440 to absorb more CO.sub.2 than the other samples 400, 420.

[0303] This experiment therefore demonstrates the improvement in CO.sub.2 capture and conversion to carbonic acid that is provided by carbonic anhydrase hydration catalyst, particularly when it is immobilised.

[0304] FIG. 5 illustrates the amount of CO.sub.2 captured modelled using the measured pH change in two solutions following sparging air at a rate of 1 L min.sup.1 in 100 mL of solution.

[0305] Line 500: NaOH (0.1 M, 100 mL)

[0306] Line 520: Na.sub.2HPO.sub.4, (0.1 M, 100 mL)+0.2 mg mL.sup.1 equivalent immobilised bovine carbonic anhydrase.

[0307] As shown by FIG. 5, the sample 520 containing immobilised carbonic anhydrase in Na.sub.2HPO.sub.4 absorbs vastly more CO.sub.2 than the NaOH sample 500 in the same amount of time. This demonstrates that the use of a hydration catalyst such as carbonic anhydrase provides significantly superior results than even highly alkaline absorbent solutions such as NaOH, which have been considered beneficial in the prior art.

[0308] In the illustrated experiments, Na.sub.2HPO.sub.4 was used as a carrier solution for the purposes of testing because it has a pH close to the optimum pH for bovine carbonic anhydrase to perform CO.sub.2 hydration.

[0309] Despite the good rate for CO.sub.2 capture demonstrated in FIG. 5, Na.sub.2HPO.sub.4 is not preferred as a first absorbent solution for the present invention as the capture capacity of Na.sub.2HPO.sub.4 is lower than desired for the present invention. The present invention also preferably avoids or at least reduces the use of membrane permeable alkali-metal salts such as NaHPO.sub.4 in the first absorbent solution.

[0310] In all experiments described herein, CO.sub.2 capture was measured using near-infrared sensors. In this case, near-infrared sensors were used to measure the background CO.sub.2 of the incoming air to the capture vessel, and to record the concentration of CO.sub.2 in the outlet from the capture vessel, so that the quantity of CO.sub.2 captured and removed from the air by the first absorbent solution could be quantified.

[0311] FIG. 6 illustrates the effectiveness of salt separation carried out by a flow-CDI cell with CO.sub.2 equivalent energy consumption of 524 kWh per tonne. In comparison to the >1500 kWh of thermal energy required for other CO.sub.2 capture processes, this is extremely energy efficient.

[0312] In the experiment behind FIG. 6, a salt inlet stream which contained 400 mg/L carbonate/bicarbonate buffer was introduced to a flow CDI cell at a rate of 15 mL min.sup.1, and flow electrode streams were pumped through the flow electrodes at a rate of 20 mL min.sup.1. The current density applied to the flow-CDI cell was 1 mA cm.sup.1 at a voltage of 1.2 V. This arrangement achieved a capture rate of 0.25 mg min.sup.1 cm.sup.2. The footprint of the flow-CDI cell was smaller than the air contactor which required 100 mL of volume, and so the flow-CDI is not rate limiting the process.

[0313] The rate of salt capture demonstrated by FIG. 6 shows that flow-CDI is an extremely effective and viable method for electrochemically separating the target acid ions from the first absorbent solution. While an amine sorbent or carbonate calciner requires between 1500-2000 kWh per tonne of CO.sub.2, this process was found to have an equivalent energy consumption of only 534 kWh per tonne of CO.sub.2.

[0314] FIG. 7 is a graph comparing average and maximum CO.sub.2 capture rates for three different catalysts in exemplary first absorbent solutions: [0315] free bovine carbonic anhydrase (free bCA), (0.2 mg/mL); [0316] bCA immobilised on Fe.sub.3O.sub.4, (0.2 mg/mL bCA immobilised on 2.5 mg/mL Fe.sub.3O.sub.4); and [0317] Fe.sub.3O.sub.4 (2.5 mg/mL).

[0318] All three absorbent solutions were made up of the catalyst in 100 mL of 0.1M Na.sub.2HPO.sub.4. Air was flowed through the absorbent solutions at a rate of 1 L per min.

[0319] FIG. 8 compares the average capture efficiency and maximum capture efficiency for the same three catalyst-containing absorbent solutions.

[0320] These results showed that the average CO.sub.2 capture rates of both free bCA and bCA immobilised on Fe.sub.3O.sub.4 are significantly higher than that of Fe.sub.3O.sub.4 alone. The average capture efficiencies of the three catalysts ranged from around 38% for Fe.sub.3O.sub.4 to around 47% for immobilised bCA on Fe.sub.3O.sub.4, while the max capture efficiency was highest for Fe.sub.3O.sub.4.

[0321] FIG. 9 illustrates the chemical structure of an exemplary Polyethyleneimine (PEI) chain, which is a cationic polymer containing branched amines. PEI is a preferred capture species, such that solutions of PEI are preferred first absorbent solutions usable in the present invention. PEI is water soluble and highly stable.

[0322] FIG. 10 is a schematic illustration of the electrolyzer-electrodialysis apparatus 300 of FIG. 3, in which polyethyleneimine (PEI) is used in the first absorbent solution for CO.sub.2 capture.

[0323] In this arrangement, the capture reactions taking place in the capture vessel (not shown) are:

CO.sub.2+H.sub.2O.fwdarw.H.sub.2CO.sub.3.fwdarw.H.sup.++HCO.sub.3.sup.

[0324] Polyethyleneimine (PEI) in the aqueous first absorbent solution reacts with the bicarbonate anions, it contains a 1:2:1 ratio of primary, secondary and tertiary amines. Primary and secondary amines react by the following mechanism:

CO.sub.2+H.sub.2O+2R.sub.2NH.fwdarw.HCO.sub.3.sup.+R.sub.2NH.sub.2.sup.++R.sub.2NH

[0325] While tertiary amines react by the following mechanism:

CO.sub.2+R.sub.3N+H.sub.2O.fwdarw.HCO.sub.3.sup.+R.sub.3NH.sup.+

[0326] The first absorbent solution containing the dissolved bicarbonate anions captured as R.sub.2NH.sup.+[HCO.sub.3].sup. is then circulated from the capture vessel to the cathodic compartment 330, from which the bicarbonate anions HCO.sub.3 are electrochemically separated and migrate through the anion-exchange membrane 114, while the carbonic acid cations in the first absorbent solution are released as hydrogen gas. The capture species PEI cations R.sub.2NH that remain are not membrane-permeable, and are recirculated from the cathodic compartment 330 back to the capture vessel to absorb more CO.sub.2.

[0327] As described above in relation to FIG. 3, the bicarbonate anions are recombined with hydrogen cations in the second absorbent chamber 312, forming carbonic acid in the second absorbent solution 334 before decomposing as CO.sub.2 gas that can be captured in the release vessel.

[0328] FIG. 11 is a graph showing the mass of CO.sub.2 captured with an H.sub.2O/PEI first absorbent solution. Three alternative aqueous absorbent solutions with a PEI concentration of 1.2 mg mL.sup.1 (0.12 wt %) were compared: [0329] PEI plus carbonic anhydrase (CA) 1110; [0330] PEI 1120; and [0331] PEI plus Fe.sub.3O.sub.4 1130.

[0332] As shown in FIG. 11, the aqueous solution of PEI plus carbonic anhydrase (CA) 1110 captured more CO.sub.2 than PEI alone 1120, and performed almost three times as well as PEI plus Fe.sub.3O.sub.4 1130. PEI plus carbonic anhydrase (CA) therefore appears to be a promising combination of capture species and catalyst for the first absorbent solution in the present invention.

[0333] FIG. 12 is a graph showing the mass of CO.sub.2 released 1210 from an H.sub.2O/PEI solution with a PEI concentration of 1.2 mg mL.sup.1 (0.12 wt %) as the temperature 1220 of the solution is increased. While thermal desorption of CO.sub.2 from PEI absorbent is not a necessary step in the present invention, the experimental results of FIG. 12 demonstrate the thermal reversibility of the PEI-CO.sub.2 absorption process. The results of FIG. 12 show that CO.sub.2 is gradually desorbed from PEI solution as the temperature of the solution increases, in particular at temperatures greater than 50 C. or 60 C. In the present invention the first absorbent solution is therefore preferably maintained at temperatures below 50 C., preferably below 40 C. or 30 C.

[0334] FIG. 13 is a graph comparing CO.sub.2 capture rates of three different first absorbent solutions: NaOH (3 Molar), bCA in H.sub.2O/PEI, and bCA immobilised on Fe.sub.3O.sub.4 in Na.sub.2HPO.sub.4. The CO.sub.2 capture in FIG. 13 was measured under fixed conditions using 50 mL of first absorbent solution and an air flow rate of 1 L min.sup.1 of air flowing through the absorbent solution.

[0335] The results of FIG. 13 showed that the capture rate of NaOH was initially the highest, but quickly reached steady state at a capture rate of around 0.008 mg s.sup.1. The capture rate of bCA immobilised on Fe.sub.3O.sub.4 in Na.sub.2HPO.sub.4 was the lowest of the three absorbents, peaking at around 0.008 mg s.sup.1 before dropping towards 0.006 mg s.sup.1 as time went on. The first absorbent solution of bCA in H.sub.2O/PEI demonstrated the highest CO.sub.2 capture rate, reaching around 0.012 mg s.sup.1 before dropping towards 0.0095 mg s.sup.1.

[0336] The concentration of NaOH tested was 120 g/L, which is the concentration that is used in the current state of the art of hydroxide based CO.sub.2 capture. The concentration of PEI in the bCA in H.sub.2O/PEI sample was however only 1300 mg/L. FIG. 13 therefore shows that the bCA in H.sub.2O/PEI sample performed best, even though the PEI concentration was 92 times lower the NaOH concentration.

[0337] FIG. 14 is a graph comparing the capture rate and capture efficiency of three different aqueous absorbents: PEI plus carbonic anhydrase (PEI/CA), NaOH and Na.sub.2CO.sub.3.

[0338] All three aqueous absorbent solutions tested had a liquid volume of 50 mL and an air flow rate of 1 L min.sup.1 through the solution. The PEI/CA solution contained 0.13 wt % PEI and 0.02 wt % CA. The NaOH solution contained 12 wt % NaOH. The Na.sub.2CO.sub.3 solution contained 29 wt % Na.sub.2CO.sub.3.

[0339] If these results are normalised by the quantity of absorbent in the solutions, the PEI/CA mixture achieves a capture rate per mg of absorbent that is around 92 times higher than the NaOH and Na.sub.2CO.sub.3 absorbents. This, combined with the advantage that the PEI/CA absorbent solution does not contain membrane-permeable ions that can reduce the efficiency of the electrochemical separation step, makes a first absorbent solution containing PEI and CA a promising candidate for use in the present invention.

[0340] An experiment was carried out to demonstrate the continuous capture of CO.sub.2 using PEI absorbent solution using the apparatus illustrated in FIG. 10. FIG. 15 illustrates the CO.sub.2 separation and release rate 1510, and the quantity of CO.sub.2 released 1520 during this experiment.

[0341] In the experiment of FIG. 15, 25 mL of first absorbent solution containing 3.16% PEI was saturated with CO.sub.2 from the air using a diaphragm pump at an airflow of 4 L min.sup.1. The absorbent solution was then circulated around the cathodic chamber of the electrodialysis cell 300 at a rate of 1 mL min.sup.1. Concurrently, a second solution containing 0.5 M NaCl was pumped between the anion and cation exchange membranes at a rate of 10 mL min.sup.1. A power supply was used to apply a voltage across the cell, and the current was gradually raised to a current density of 200 mA cm.sup.2 at a voltage of 4 V. At this current density, a steady stream of bubbles were observed to be exiting the cell along with the second solution, demonstrating that separation and release of CO.sub.2 from the first absorbent solution were occurring concurrently (though in a full scale version of this apparatus the cell 300 would be pressurised to prevent bubble evolution inside the cell). The second solution and the evolved CO.sub.2 bubbles were pumped to a release chamber through which a continuous stream of air was pumped to entrain the released CO.sub.2. The quantity of evolved CO.sub.2 was measured using a high speed near-IR sensor (SprintIR-W 100% CO2 Sensor|CO2Meter.com). Air was continuously pumped from the release chamber at a rate of 300 mL min and analysed in the CO.sub.2 meter.

[0342] FIGS. 15 and 16 show that the CO.sub.2 separation and release rate 1510 climbed rapidly and peaked at a rate of around 200 grams of CO.sub.2 m.sup.2 hr.sup.1, before falling gradually. This fall in separation and release rate was thought to be caused by the fact that the electrodialytic cell was separating and releasing CO.sub.2 more quickly than was being absorbed by the first absorbent solution. This experiment confirms that the apparatus of FIG. 10 achieved CO.sub.2 separation rates in the same range as typical electrodialysis rates, and is not rate-limiting the capture step.

[0343] The total amount of CO.sub.2 released 1520 was observed to be climbing continuously, reaching 50 grams of CO.sub.2 m.sup.2 after around 1400 seconds.

[0344] The quantity CO.sub.2 released and the CO.sub.2 release rates are normalised by the interfacial surface area of the ion exchange membrane.

[0345] The cell tested in this experiment and shown in FIG. 10 contained only one pair of membranes (while electrodialysis stacks typically contain up to 500 pairs) and so the energy efficiency is not representative of what a full electrodialysis cell with a typical number of membrane pairs could achieve. However, the energy efficiency is in line with what would be expected for a 2 membrane cell.

[0346] Excluding the faradaic contribution of the electrodialysis cell, which becomes negligible in a full system, the energy calculated to perform this process is 4800 kWh tonne. This is expected to be significantly reduced with higher CO.sub.2 concentration in solution, optimising the voltage/current and the use of flow electrodes and by increasing the number of cell pairs.

[0347] These results show that the aqueous first absorbent solution of PEI and CA can successfully capture CO.sub.2 and convert it to carbonic acid ions, and that the HCO.sub.3.sup. anion of carbonic acid can be transported across an anion-exchange membrane into a second absorbent solution, and subsequently decomposed back to carbonic acid and released as pure gaseous CO.sub.2.

[0348] FIG. 17 shows an alternative electrodialysis apparatus 1700 usable in a preferred embodiment of the present invention. The apparatus 1700 of FIG. 17 is similar to the apparatus of FIGS. 3 and 10, with the differences that the configuration of the electrolyser is configured so that the first absorbent solution enters the central chamber of the cell, and the polarity of the electrodes is reversed.

[0349] The electrodialysis apparatus 1700 contains a first absorbent chamber 1712 that is filled with a porous solid electrolyte, between a positive electrode (anode) 322 and a negative electrode (cathode) 324. A pair of ion-exchange membranesan anion-exchange membrane 114 and a cation-exchange membrane 116are arranged in parallel between the electrodes, dividing the apparatus 1700 into three compartments: a cathodic compartment 330 on one side of the cation-exchange membrane 116, an anodic compartment 340 on one side of the anion-exchange membrane 114, and the first absorbent chamber 1712 between the two membranes.

[0350] During operation, a potential difference is applied between the anode 322 and the cathode 324. Liquid first absorbent solution 310 containing a capture species, target anions (bicarbonate anions in the case of the illustrated example) and hydrogen cations is pumped from a capture vessel (not shown) into the first absorbent chamber 1712. A second absorbent solution 334, which may be an aqueous or non-aqueous flow of electrolyte, for example sodium sulfate or sodium chloride, is pumped through the anodic compartment 340.

[0351] As the cathode is in an alkaline environment, but also contains hydrogen ions formed by dissolution of the target species in the first absorbent solution, two reactions take place. At the negatively charged cathode, a reduction reaction takes place, with electrons (e.sup.) from the cathode combining with hydrogen cations to form hydrogen gas. The reduction reaction taking place at the cathode is: 2H.sub.2O (I)+2e.sup..fwdarw.2OH.sup. (aq)+H.sub.2 (g) (OH.sup. is mostly neutralised by HCO.sub.3 to form carbonate), while the reaction 2H.sup.++2e.sup..fwdarw.H.sub.2 also occurs to evolve hydrogen gas from the cathodic compartment 330.

[0352] At the positively charged anode, an oxidation reaction occurs, generating oxygen gas and giving electrons to the anode to complete the circuit. The reaction taking place at the anode is: 2H.sub.2O (I).fwdarw.O.sub.2 (g)+4e.sup.+4H.sup.+ (aq).

[0353] As in the embodiments described above, the first absorbent solution contains a capture species such as PEI, and the anion- and cation-exchange membranes are impermeable to the capture species, so that the capture species is kept in the first absorbent solution 310.

[0354] As the first absorbent solution 310 containing the target anions (HCO.sub.3 bicarbonate anions in the illustrated case of carbonic acid) flows into the first absorbent chamber 1712, the electrical field between the electrodes attracts the target anions through the anion-exchange membrane 114 and into the second absorbent solution 334 in the anodic compartment 340. At the same time, hydrogen cations (H.sup.+) formed by electrolysis at the anode are created in the anodic compartment 340, where they can associate with the target anions to form the target acid in the second absorbent solution. The anodic compartment 340 is maintained at a sufficiently high pressure that the target acid does not decompose to form gas bubbles inside the apparatus 300. The target acid is concentrated in the stream of the second absorbent solution 334, which is circulated to a release vessel (not shown) where a gas of the target species (CO.sub.2 gas in the illustrated example) is evolved and captured.

[0355] In this embodiment, some oxygen gas is present in the CO.sub.2 stream. This can be removed by either burning the H.sub.2+O.sub.2+CO.sub.2 stream, or by passing the gasses through a fuel cell to recover the energy.

[0356] The hydrogen cations (H.sup.+) in the first absorbent solution migrate through the cation-exchange membrane 116 into the cathodic compartment 330, from which they are released as hydrogen gas. Thus target anions migrate out of the cathodic compartment 330 and into the second absorbent chamber 312 by passing through the anion-exchange membrane 114. At the same time, hydrogen cations (H.sup.+) formed by electrolysis at the anode are attracted through the cation exchange membrane 116, and migrate into the second absorbent chamber 312.

[0357] Having lost the target bicarbonate anions and the hydrogen cations during the electrochemical separation process, the first absorbent solution 310 is recirculated to the capture vessel (not shown) still containing the capture species.

[0358] FIG. 18 illustrates CO.sub.2 capture results obtained using the apparatus of FIG. 17 for CO.sub.2 capture, using the same components and parameters described above in relation to FIGS. 15 and 16. As shown in FIG. 18, the energy efficiency of the process and the rate of CO.sub.2 capture strongly depends on the voltage at which the apparatus is operated. At roughly 30 mg per hr, the energy consumption is 900 kWh per tonne of CO.sub.2 captured. However, because electrolysis is occurring in the electrochemical cell, the process has produced 500 kWh of H.sub.2 in the same time period. Overall, therefore, CO.sub.2 has been captured and released for 400 kWh per tonne.

[0359] FIG. 19

[0360] FIG. 19 is a schematic illustration of an apparatus according to a preferred embodiment of the present invention, which employs electrodialysis to electrochemically separate ions as part of a gas capture process. The operation of the apparatus is substantially similar to that of the flow-CDI apparatus described above in relation to FIG. 1, with the difference that the ion-separator employs electrodialysis rather than flow-CDI.

[0361] The electrodialysis apparatus 1900 illustrated in FIG. 19 is made up of a gas contactor 1902, an ion-separator 1904, and a release vessel 1906.

[0362] The gas contactor 1902 is arranged to receive a flow of gas 1908 which contains a target species to be captured, and to bring the gas into contact with a stream of a first absorbent solution 1910. A variety of gas-liquid contactor designs are known in the art, such as falling-film reactors, packed columns, bubble columns or spray towers, any of which would be suitable for use with the present invention.

[0363] The ion-separator 1904 contains a separation chamber 1912, an anion-exchange membrane 1914 along one side of the separation chamber 1912, and a cation-exchange membrane 1916 along the opposite side of the separation chamber 1912. An inlet pipe and an outlet pipe connect the gas contactor 1902 to the separation chamber, so that a stream of first absorbent solution 1910 can be pumped from the gas contactor, through the separation chamber, and then recirculated to the gas contactor.

[0364] A positive electrode 1922 (anode) is connected to the ion-separator 1904 on the side of the anion-exchange membrane, and a negative electrode 1924 (cathode) is connected to the ion-separator 1904 on the side of the cation-exchange membrane.

[0365] The ion-separator comprises a second absorbent channel 1926 between the anion-exchange membrane and the positive electrode 1922. The second absorbent channel is connected in a loop with the release vessel 1906, and a second absorbent solution 1934 is circulated between the release vessel 1906 and the second absorbent channel 1926.

[0366] For electrodialysis, the second absorbent solution may be H.sub.2O, or another aqueous solution, for example an organic acid such as sulfonic acid. In one experiment, the inventors have successfully used 0.18% poly-4-styrene sulfonic acid as the second absorbent solution.

[0367] In use, a flow of gas 1908 which contains a target species to be captured is introduced into the gas contactor 1902, at the same time that a first absorbent solution 1910 containing a capture species is introduced into the gas contactor. As the gas 1908 comes into contact with the first absorbent solution 1910, mass transfer of the target species into the absorbent solution takes place, so that the first absorbent solution 1910 absorbs some of the target species from the gas.

[0368] The target species is dissolved in the first absorbent solution 1910, optionally assisted by the presence of a hydration catalyst in the first absorbent solution, and forms a target anion and a hydrogen cation. The target anion and the hydrogen cation associate with and are stabilised by the capture species in the first absorbent solution.

[0369] The first absorbent solution 1910 is continuously pumped from the outlet of the gas contactor 1902 to the inlet of the separation chamber 1912 of the ion-separator 1904, from where the liquid first absorbent solution 1910 flows through the separation chamber.

[0370] During operation, a potential difference is applied between the positive electrode 1922 and the negative electrode 1924. This potential difference across the ion-separator means that as the first absorbent solution 1910 flows through the separation chamber, the negatively-charged target anions (bicarbonate anions HCO.sub.3.sup. in the illustrated embodiment) are dissociated from the capture species and attracted towards the positive electrode, while the positively-charged hydrogen cations are attracted towards the negative electrode. The target anions therefore migrate through the anion-exchange membrane 1914, and the hydrogen cations flow through the cation-exchange membrane 1916, so that the target anions are separated from the first absorbent solution. Due to the large hydrodynamic radius and high molecular weight of the capture species, neither the anion-exchange membrane 1914 nor the cation-exchange membrane 1916 is permeable to the capture species, so the capture species remains in the first absorbent solution 1910.

[0371] By the time that the first absorbent solution 1910 reaches the outlet of the separation chamber 1912, at least some of the target acid ions (target anions and hydrogen cations) have been separated from the stream of first absorbent solution 1910, and the first absorbent solution is recirculated to the inlet of the gas contactor 1902.

[0372] During operation, a stream of the second absorbent solution 1934 containing a slurry of conductive particles is pumped through the second absorbent channel 1926, so that target anions and hydrogen cations passing through the ion-exchange membranes are transferred into the stream of second absorbent solution 1934. The target anions and hydrogen cations are recombined in the second absorbent channel 1926, as they flow to the release vessel 1906, and reassociate with one another so that the stream of second absorbent solution 1934 contains the target acid when it reaches the release vessel 1906.

[0373] The stream of second absorbent solution 1934 is maintained under pressure in the second absorbent channel 1926, which prevents bubble formation within the ion-separator, and is then depressurised in the release vessel 1906 where the target gas (CO.sub.2 in the illustrated embodiment) spontaneously evolves from the solution.

[0374] The stream of the second absorbent solution 1934 is then recirculated back to the second absorbent channel 1926 in a continuous process.

[0375] Using this system, the target species can be continuously absorbed from the flow of gas 1908, transferred from the first absorbent solution to the second absorbent solution, and eventually released in the release vessel 1906.

[0376] Similarly to the embodiments described above, a particularly preferred embodiment of the invention that can be performed using this set-up is the direct capture of carbon dioxide (CO.sub.2) from air.

[0377] In this direct air capture (DAC) embodiment, air is used as the flow of gas 1908, and the first absorbent solution 1910 is an aqueous solution containing a CO.sub.2 hydration catalyst. As illustrated in FIG. 19, air 1908 flows into the gas contactor with a temperature of 298 K and a pressure of 1 bar with a CO.sub.2 concentration of 400 ppm. After passing through the gas contactor and having some of its CO.sub.2 content absorbed by the first absorbent solution, the air has a CO.sub.2 concentration of only 100 ppm.

[0378] A particularly preferred option for the capture species in this embodiment is an aqueous solution of polyethyleneimine (PEI) having a molecular weight of greater than 800. The high molecular weights and hydrodynamic radii of the PEI means that these components are excluded from passage through the ion-exchange membranes, and therefore remain in the first absorbent solution to be recirculated.

[0379] FIG. 20

[0380] While many of the experiments described above were obtained using a gas sparger as a gas contactor, a falling film reactor was assembled as an alternative gas contactor for use with the present invention. A falling film reactor may advantageously allow for a more accurate determination of the CO.sub.2 capture rate of different capture species as a function of surface area, gas: liquid ratio and residence time.

[0381] FIG. 20 is a graph of the CO.sub.2 exit concentration from a falling film reactor usable in a preferred embodiment of the present invention. The CO.sub.2 exit concentration from the falling film reactor at demonstrated a capture rate of 0.83 g CO.sub.2 hr.sup.1 m.sup.2 for 3 molar NaOH, compared to a capture rate of 0.6 g CO.sub.2 hr.sup.1 m.sup.2 with a first absorbent solution which is a 10 wt % aqueous solution of PEI (M.sub.w 1800).

[0382] The direct comparison of NaOH and polymeric amine (PA) capture species can be seen in FIG. 20, which shows a 10 wt % solution of PEI (M.sub.w 1800) and achieves a capture rate within 78% of the capture rate achieved by 3M NaOH. Temperature sensitivity experiments were also conducted with 10 wt % PEI (M.sub.w 1800) and can be seen in Table 1 below. The data shows that while solution temperature does impact the rate of CO.sub.2 capture, only a 17% drop in rate is observed from reducing the temperature from 20 C. to 5 C.

TABLE-US-00001 TABLE 1 Effect of temperature on the CO.sub.2 capture rate with 10 wt % PEI (M.sub.w 1800) with a gas:liquid ratio 10.6, air residence time 1.2 s, fluid residence time 4.3 s. Temperature, C. ppm mg.sub.CO2 hr.sup.1 m.sup.2.sub.film surface area 30 163 678.32 20 144 599.25 10 123 511.86 5 120 499.38

[0383] In further experiments, the effect of concentration and the nature of polymeric amines (PA) having different molecular weights was investigated in CO.sub.2 capture experiments, as shown in FIG. 21.

[0384] FIG. 21 is a graph of the CO.sub.2 exit concentration from the falling film reactor used in Figure across a range of temperatures and for two different capture species. PA #1 is PEI (M.sub.w 1800), which is the capture species used in FIG. 20, while PA #2 is PEI (M.sub.w 800), which is a related variant polymeric amine with a lower molecular weight.

[0385] The results indicate that by increasing the concentration of PA #1 (PEI, M.sub.w 1800) from 10 wt % to 15 wt %, the performance could be improved across a wide temperature range, while increasing to 20 wt % was less optimal, particularly at higher temperatures. Another polymeric amine labelled PA #2 (PEI, M.sub.w 800) was tested as an alternative variant with a different molecular weight. PA #2 (PEI M.sub.w 800) was highly effective across the temperature range, outperforming all other solvent combinations. The CO.sub.2 capture rate performance of PA #2 (PEI M.sub.w 800) at 10 wt % was significantly higher than PA #1 (PEI M.sub.w 1800) at concentrations of 10 wt %, 15 wt % or 20 wt %.

[0386] FIG. 22 is a is a schematic illustration of an electrodialysis cell apparatus usable for direct air capture of CO.sub.2 in a preferred embodiment of the present invention, in which X is a capture species that cannot permeate either anion- or cation-exchange membranes. The operation of the electrodialysis cell is based on that described above in relation to FIG. 2B.

[0387] In the preferred embodiment of FIG. 22, CO.sub.2 captured by the first absorbent solution containing polymeric amine (PA) is regenerated through electrodialysis. The first absorbent solution, which is customarily referred to as the diluate in electrodialysis terminology, contains a PA capture species X and stabilised H.sup.+ and HCO.sub.3.sup. ions, and is flowed through the electrodialysis cell via the diluate chambers 2100. Each chamber is separated by cation exchange membranes 2110 and anion exchange membranes 2120 which selectively allow ions of the complementary charge to migrate. The large size of the PA capture species molecules prevents them from migrating through any ion-exchange membrane, so the PA capture species remains in the diluate chambers 2100. A voltage is applied across the electrodialysis cell, which causes anions to migrate towards the anode and cations to migrate towards the cathode. This process causes ions from the first absorbent capture solution to concentrate in the second absorbent release solution, customarily referred to as the concentrate, in the concentrate chambers 2200. Alone, these ions are unstable in solution and decompose into CO.sub.2, whose solubility is also low in H.sub.2O and transfers to the gas phase.

[0388] As the large size of the capture species excludes them from passage through the ion-exchange membranes into the second absorbent release solution, this process may be termed size-exclusion electrodialysis (SEED).

[0389] This size-exclusion electrodialysis (SEED) process was carried out with a small scale, three-chambered electrochemical cell 2300 purchased from Dioxide Materials and shown in FIG. 23. This cell features two titanium current collectors with serpentine, 1 mm wide channels and a 2 mm separator that acts as the internal chamber where the first absorbent solution flows between the ion-exchange membranes. In this configuration, there is a single pair of cation and anion exchange membranes. The HCO.sub.3 target anion is transported to the anode and decomposed to CO.sub.2 by the acidic environment of the anodic chamber caused by water oxidation, while the H.sup.+ from PA-H.sup.+ is transported to the cathode and reduced at the electrode to form hydrogen.

[0390] Experiments were conducted with a 3.6 wt % polymeric amine capture species (Polyethyleneimine, 1800 molecular weight) aqueous solution pre-saturated with air and continuously bubbled throughout the experiment to maximise the concentration of HCO.sub.3.sup. and H.sup.+ in solution. The second absorbent solution was pure H.sub.2O. A range of voltages were applied from 5.2 V to 2.6 V. At higher voltages, the stability of the membranes and electrodes would have been significantly compromised, as well as a significant concentration deficit of CO.sub.2 would have occurred in the first absorbent capture solution. At voltages less than 2.6 V, the CO.sub.2, output became challenging to measure. CO.sub.2 measurements were taken continuously using high-speed near-infrared sensors, the results of which can be seen in FIG. 24.

[0391] Calculating the power input and dividing by the amount of CO.sub.2 captured per hour yielded the specific energy consumption of the cell in kWh/tCO.sub.2, as shown in FIG. 25. It should be noted that a large proportion of the total energy consumption in the single membrane-pair cell is related to the production of hydrogen and oxygen at the electrodes. This parasitic loss is greatly diminished in systems that contain >10 pairs of membrane channels and losses are negligible by 40 pairs. After the experiment, a sample of the concentrate solution was dried to detect any residue that may have resulted from PA capture species migrating through the membrane; no sign of this was observed. Overall, these results prove that bicarbonate target anions can be separated from the polymeric-amine-containing first absorbent solution without any noticeable transfer of PA into the second absorbent release solution. It also demonstrated an energy consumption as low as 400 kWh/tCO.sub.2 (omitting hydrogen/oxygen production) is possible in such a system without any optimisation. Corresponding voltages, current densities, and surface area normalised capture rates are tabulated in Table 2.

TABLE-US-00002 TABLE 2 Voltage and current densities of the 5 cm.sup.2 cell of FIG. 23, with corresponding CO.sub.2 production rates. Voltage, V Current, A cm.sup.2 g.sub.CO2 hr.sup.1 cm.sup.2.sub.membrane area 5.2 0.057 0.038 4.5 0.0418 0.028 4 0.0216 0.020 3 0.0072 0.0088 2.6 0.002 0.0057

[0392] FIG. 26 is a graph showing CO.sub.2 stability data obtained using the falling film reactor and the electrodialysis cell of FIG. 23, which contains a single cell pair of 37 cm 2 interfacial membrane area. Voltage 4 V; current 7 mA; flow rate 15 mL/min. The first absorbent solution was an aqueous solution of PEI, with a PEI concentration of 10 wt %. The ion-exchange membranes used in the experiment were Fumatech FKS-PET-130 (cation exchange) and Fumatech FAS-PET-130 (anion exchange).

[0393] As shown in FIG. 26, the CO.sub.2 concentration in the release vessel stabilised after a period of testing and pretreatment, and remained stable at around 1750 ppm for around 10 hours until a sensor test was carried out.

[0394] Forty-Membrane-Pair Electrodialysis

[0395] As mentioned above, typical electrodialysis modules contain a larger number of membrane pairs to increase the energy efficiency of separation by reducing the proportion of energy that relates to water splitting reactions, and so a lab-scale electrodialysis cell was purchased from Fumatech, which contained 40 membrane pairs. The ED-40 cell from Fumatech contains standard Fumasep FAS and FKS ion-exchange membranes used in electrodialysis applications and are separated by 450 m spacers.

[0396] The electrodialysis cell was operated under conditions of constant current, at a current density of 0.5 mA cm.sup.2 of electrode area requiring an applied voltage of 21.5 V. Assuming the voltage necessary to drive the faradaic process was 1.5 V would indicate a cell pair voltage of 0.5 V. With these conditions, the electrodialysis cell produced up to 0.7 g of CO.sub.2 per hour, resulting in specific energy consumption of 510 kWh/tCO.sub.2 with an associated current efficiency of 69%.

[0397] FIG. 27 is a graph of CO.sub.2 output rates in mg/hr (black) vs energy consumption for CO.sub.2 separation and release in kWh/tCO.sub.2 (red). Membrane surface area, 0.145 m.sup.2. Electrode area, 0.0036 m.sup.2. Voltage=21.5 V, Current=0.0017 A, current efficiency=69%. Spacer thickness, 450 micron.

[0398] FIG. 28 is a modelled graph of the electrical energy demand of electrodialysis using the same 40-membrane-pair electrodialysis cell as FIG. 26 with different concentrations of PEI capture species in the first absorbent solution.

[0399] Electrical energy demand of electrodialysis, 40 cell pairs, PA loading variation (3.6 wt % PA; 7.2 wt % PA; 14.4 wt % PA). Spacer thickness 100 microns. Diluate flow rate 2 L/hr/cell, saturated with 400 ppm CO.sub.2 at 1 atm.

[0400] Using an electrochemical computational model, the effect of capture loading on the productivity of the cell and the required energy demand can be predicted, as shown in FIG. 28. The modelling predicts a significant improvement in the CO.sub.2 capture rate for a given energy demand at higher solvent loadings due to increased conductivity and concentration of dissolved HCO.sub.3 and H.sup.+. Furthermore, the model considered the effect of thinner membrane spacers revealing a non-linear relationship, suggesting an improvement of between 10-100 could be achieved vs the current experimental data.

[0401] Preferred Aspects

[0402] Preferred aspects of the present invention are defined in the following numbered clauses: [0403] 1. A method of capturing a target species from a gas comprising the steps of: [0404] contacting a gas containing a target species with a first absorbent solution comprising a capture species; [0405] dissolving the target species in the first absorbent solution to form a target anion; [0406] electrochemically separating the target anion from the first absorbent solution by contacting the first absorbent solution with one or more ion-exchange membranes, and transferring the target anion through an ion-exchange membrane into a second absorbent solution; and [0407] releasing at least some of the target species from the second absorbent solution, [0408] in which the one or more ion-exchange membranes are not permeable to the capture species, so the capture species does not pass through the one or more ion-exchange membranes. [0409] 2. A method according to clause 1, in which the capture species binds to the target anion in the first absorbent solution, and in which the target anion is electrochemically dissociated from the capture species before being transferred through the ion-exchange membrane. [0410] 3. A method according to clause 1 or 2, in which the capture species is a non-alkali-metal capture species. [0411] 4. A method according to clause 1, 2 or 3, in which the capture species is an ionic capture species, preferably a cationic capture species. [0412] 5. A method according to any preceding clause, in which the capture species is an ionic polymer. [0413] 6. A method according to any preceding clause, in which the capture species is a cationic capture species that does not comprise an alkali metal cation. [0414] 7. A method according to any preceding clause, in which the capture species is a cationic organic capture species. [0415] 8. A method according to any preceding clause, in which the capture species is a choline-derived ionic liquid, preferably a cationic choline-derived ionic liquid containing the conjugate base of an organic acid such as carboxylic acid or propanoic acid. [0416] 9. A method according to any preceding clause, in which the capture species is a cationic polymer, preferably a cationic polymer having a repeat unit which comprises a plurality of amine groups. [0417] 10. A method according to clause 9, in which the capture species comprises a plurality of polymer resin particles functionalised with cationic functional groups. [0418] 11. A method according to clause 9, in which the capture species comprises a slurry of anion-exchange resin particles functionalised with cationic functional groups. [0419] 12. A method according to any preceding clause, in which the capture species is weakly basic, preferably in which the capture species has a pK.sub.a of less than 10, preferably less than 8.5, particularly preferably less than 7.5. [0420] 13. A method according to any preceding clause, in which the capture species is a polymeric amine, preferably a cationic polymeric amine. [0421] 14. A method according to any preceding clause in which the capture species comprises polyethyleneimine (PEI). [0422] 15. A method according to any preceding clause, in which the capture species has a molecular weight of greater than or equal to 200, or 400, or 500, or 600, or 700, or 800 g/mol. [0423] 16. A method according to any preceding clause, in which the one or more ion-exchange membranes are configured to permit passage of the target anion therethrough, and to prevent passage of capture species having a cationic charge and/or a molecular weight of greater than 200, or 250, or 300, or 400, or 500, or 600 g/mol. [0424] 17. A method according to any preceding clause, in which the first absorbent solution contains no inorganic salts, or contains less than 2 wt % inorganic salt. [0425] 18. A method according to any preceding clause, in which the first absorbent solution contains a hydration catalyst for accelerating the conversion of the dissolved target species into the target anion. [0426] 19. A method according to clause 18, in which the catalyst comprises an enzyme, for example carbonic anhydrase, organometallic compounds of zinc (zinc cyclen), and/or metallic or metal-oxide particles or nanoparticles. [0427] 20. A method according to any preceding clause, in which the first absorbent solution is maintained at a temperature of between 15 C. and 60 C., preferably between 18 C. and 40 C., particularly preferably between 30 C. and 40 C., and/or at a pressure of less than 2 bar, preferably at atmospheric pressure. [0428] 21. A method according to any preceding clause, in which at least one of the ion-exchange membranes is an anion-exchange membrane permeable to the target anion, preferably in which the anion-exchange membrane is a monovalent-anion-exchange membrane. [0429] 22. A method according to any preceding clause, in which the target species is dissolved in the first absorbent solution to form a target anion and a target counterion, preferably in which the target counterion is H.sup.+. [0430] 23. A method according to clause 22, in which the one or more ion-exchange membranes comprise an anion-exchange membrane permeable to the target anion, and a cation-exchange membrane permeable to the target counterion. [0431] 24. A method according to clause 23, in which the target counterion is electrochemically separated from the first absorbent solution and transferred through the cation-exchange membrane into the second absorbent solution. [0432] 25. A method according to any of clauses 22 to 24, in which the target anion associates with the target counterion in the second absorbent solution, preferably to form a target acid. [0433] 26. A method according to any of clauses 22 to 24, in which the target anion is combined with a hydrogen cation to form a target acid in the second absorbent solution, the hydrogen cation being produced by electrolysing H.sub.2O. [0434] 27. A method according to clause 25 or 26, in which the target acid is the conjugate acid of the target species. [0435] 28. A method according to any preceding clause, in which the second absorbent solution has a different composition from the first absorbent solution. [0436] 29. A method according to any preceding clause, in which the second absorbent solution has a pH which is different from the pH of the first absorbent solution, preferably in which the pH of the second absorbent solution is less than 7. [0437] 30. A method according to any preceding clause, in which the first absorbent solution is an aqueous solution and the second absorbent solution is a non-aqueous solution. [0438] 31. A method according to any preceding clause, in which the second absorbent solution does not contain the capture species. [0439] 32. A method according to any preceding clause, in which the second absorbent solution is non-aqueous, preferably in which the second absorbent solution comprises or consists of an organic carbonate solvent such as ethylene carbonate, propylene carbonate or dimethyl carbonate. [0440] 33. A method according to any preceding clause, in which the second absorbent solution comprises one or more catalysts for the in-situ reduction of the target anion, comprising of metallic catalysts or metal chalcogenides (oxides, nitrides, sulphides, phospides) of a metal selected from the list: Pt, Pd, Fe, Mo, Mn, Cu, Zn, V, W. [0441] 34. A method according to clause 29, in which the second absorbent solution contains less than 5 wt % of an inorganic salt, preferably less than 2 wt % of an inorganic salt. [0442] 35. A method according to any preceding clause, comprising one or more flow electrodes in contact with the one or more ion-exchange membranes, preferably in which a first flow electrode comprises a stream of second absorbent solution in contact with an output side of the ion-exchange membrane through which the target anion is transferred. [0443] 36. A method according to clause 35, in which each flow electrode comprises a stream of absorbent solution comprising a suspension of electrically or ionically-conductive particles selected from the group of: carbon- or metal-based particles or nanoparticles such as activated carbon; oxides, hydroxides, and/or oxyhydroxides of platinum, silver, iron, nickel, manganese, and/or titanium; or redox species such as riboflavin 5-monophosphate sodium salt hydrate, anthraquinone, polyoxometalates. [0444] 37. A method according to any preceding clause, in which the step of electrochemically separating the target anion from the first absorbent solution comprises capacitive deionisation (CDI), preferably flow-CDI, or electrodialysis. [0445] 38. A method according to any preceding clause, in which the target species is released from the second absorbent solution as a gas, in order to maintain the chemical equilibrium of the target acid in the second absorbent solution, preferably at room temperature and atmospheric pressure. [0446] 39. A method according to clause 38, in which the step of releasing at least some of the target species from the target acid in the second absorbent solution comprises the step of heating the second absorbent solution via means such as photothermal, magnetic induction, resistive or dielectric, and/or reducing the pressure above the second absorbent solution. [0447] 40. A method according to any of clauses 1 to 39, in which at least some of the target anions in the second absorbent solution are reacted with a mineral or salt to form a precipitated material that is released from the second absorbent solution. [0448] 41. A method according to any preceding clause, in which a concentration of the target species in the gas is less than 50 vol %, or 45 vol %, or 25 vol %, or 15 vol %, or 10 vol %, or 5 vol %, or 1 vol %, preferably less than 0.5 vol %. [0449] 42. A method according to any preceding clause, in which the gas containing the target species is air, flue gas from fossil fuel combustion, industrial gas, or any combination thereof. [0450] 43. A method according to any preceding clause, in which the target species is selected from the group consisting of CO.sub.2, H.sub.2S, SO.sub.2, NO, NO.sub.2, and N.sub.2O. [0451] 44. A method according to any preceding clause, in which the target species is CO.sub.2, the target anion is bicarbonate, and the target acid is carbonic acid. [0452] 45. A method according to clause 40, in which the first absorbent solution contains a catalyst for converting CO.sub.2 into bicarbonate, preferably in which the catalyst is carbonic anhydrase or a Zn 2+ containing compound such as zinc cyclen. [0453] 46. An apparatus for capturing a target species from a gas, comprising: [0454] a gas contactor configured to contact a gas containing a target species with a first absorbent solution containing a capture species, dissolving the target species in the first absorbent solution to form target anions; [0455] an ion-separator comprising one or more ion-exchange membranes for electrochemically separating the target anions from the first absorbent solution and transferring at least some of the target anions to a second absorbent solution; and [0456] a release vessel for releasing at least some of the target species from the second absorbent solution, [0457] in which the one or more ion-exchange membranes are not permeable to the capture species, in use. [0458] 47. An apparatus according to clause 46, in which the one or more ion-exchange membranes are configured to transfer the target anions from the first absorbent solution into the second absorbent solution, and to retain the capture species in the first capture solution. [0459] 48. An apparatus according to clause 46 or 47, in which the ion-separator is configured to operate under a hydrostatic pressure of greater than 2 atm, preferably greater than 3 atm or 5 atm or 7 atm, or even 30 atm or higher. [0460] 49. An apparatus according to clause 46, 47 or 48, in which the ion-separator is configured to transfer only the target anions into the second absorbent solution. [0461] 50. An apparatus according to clause 46, 47 or 48, in which the ion-separator is configured to transfer both the target anions and a plurality of hydrogen cations from the first absorbent solution into the second absorbent solution. [0462] 51. An apparatus according to any of clauses 46 to 50, in which the one or more ion-exchange membrane comprises, or consists of, an anion-exchange membrane configured to permit passage of the target anion therethrough. [0463] 52. An apparatus according to any of clauses 46 to 50, in which the ion-separator comprises two or more ion-exchange membranes, preferably an anion-exchange membrane and a cation-exchange membrane. [0464] 53. An apparatus according to any of clauses 46 to 52, in which the ion-separator comprises a chamber with an anion-exchange membrane, in which the ion-separator is configured to receive a stream of the first absorbent solution, and to electrochemically separate the target anions through the anion-exchange membrane into the second absorbent solution. [0465] 54. An apparatus according to any of clauses 46 to 53, in which the ion-separator comprises a separation chamber with a pair of opposing ion-exchange membranes, one of which is permeable to the target anion, and the other of which is permeable to hydrogen cations. [0466] 55. An apparatus according to any of clauses 46 to 54, in which the ion-separator comprises one or more, or two or more, flow electrodes in contact with output sides of the one or more ion-exchange membranes. [0467] 56. An apparatus according to clause 55, in which the flow electrode(s) comprises a stream of second absorbent solution, so that target anions passing through the one or more ion-exchange membranes are transferred into the stream of second absorbent solution. [0468] 57. An apparatus according to any of clauses 46 to 56, in which the apparatus is configured to electrolyse water, and to introduce the resulting hydrogen cations into the second absorbent solution. [0469] 58. An apparatus according to any of clauses 46 to 57, in which the apparatus comprises means for transferring first absorbent solution from the gas contactor to the ion-separator, and means for recirculating first absorbent solution from the ion-separator to the gas contactor. [0470] 59. An apparatus according to any of clauses 46 to 58, in which the apparatus comprises means for transferring second absorbent solution from the ion-separator to the release vessel, and means for recirculating second absorbent solution from the release vessel to the ion-separator. [0471] 60. An apparatus according to any of clauses 46 to 59, in which the ion-separator is a capacitive deionisation (CDI) ion-separator, or a CDI cell, or in which the ion-separator is an electrodialysis ion-separator, or an electrodialysis cell. [0472] 61. An apparatus according to any of clauses 46 to 60, in which The apparatus is configured to operate continuously. [0473] 62. An apparatus according to any of clauses 46 to 61, in which the ion-separator is a flow electrode capacitive deionisation (FCDI) ion-separator, or a continuous-flow electrodialysis ion-separator.