Metal organic framework based water capture apparatus

Abstract

An apparatus for capturing a water content from a water containing gas, the apparatus comprising: a housing having an inlet into which the water containing gas can flow; a water adsorbent located in the housing, the water adsorbent comprising at least one water adsorbent metal organic framework composite capable of adsorbing a water content from the water containing gas; and a water desorption arrangement in contact with and/or surrounding the water adsorbent, the water desorption arrangement being selectively operable between (i) a deactivated state, and (ii) an activated state in which the arrangement is configured to apply heat, a reduced pressure or a combination thereof to the water adsorbent to desorb a water content from the water adsorbent.

Claims

1. A water adsorbent configured to capture a water content from air, the water adsorbent comprising: at least one water adsorbent metal organic framework composite capable of adsorbing a water content from the air, wherein the at least one water adsorbent metal organic framework composite comprises at least 70 wt % water adsorbent metal organic framework and, wherein the water adsorbent metal organic framework comprises: a metal ion selected from the group consisting of Fe.sup.3+, Li.sup.+, Na.sup.+, Ca.sup.2+, Zn.sup.2+, Zr.sup.4+, Al.sup.3+, K.sup.+, Mg.sup.2+, Ti.sup.4+, Cu.sup.2+, Mn.sup.2+ to Mn.sup.7+, Ag.sup.+, or combinations thereof; and at least 0.1 wt % of a hydrophilic binder selected from the group consisting of an alkyl cellulose, a hydroxyalkyl cellulose, a carboxyalkyl cellulose derivative, or combinations thereof, and wherein the at least one water adsorbent metal organic framework composite has an average surface area of at least 700 m.sup.2/g and is configured to adsorb a water content from the air.

2. The water adsorbent according to claim 1, wherein the at least one metal organic framework composite comprises at least one of pellets, pills, spheres, granules, extrudates, honeycombs, meshes, hollow bodies, or combinations thereof.

3. The water adsorbent according to claim 1, wherein the hydrophilic binder is selected from at least one of hydroxypropyl cellulose (HPC), hydroxypropyl cellulose, hydroxyethyl methyl cellulose, hydroxypropyl methyl cellulose (HPMC), ethyl hydroxyethyl cellulose, methyl cellulose, or carboxymethyl cellulose (CMC).

4. The water adsorbent according to claim 1, comprising between 0.5 wt % and 3 wt % of the hydrophilic binder.

5. The water adsorbent according to claim 1, comprising between 0.8 wt % and 2 wt % of the hydrophilic binder.

6. The water adsorbent according to claim 1, wherein the at least one water adsorbent metal organic framework composite comprises at least one of a pellet, a coating, a plate, a sheet, or a strip.

7. The water adsorbent of according to claim 1, wherein the at least one water adsorbent metal organic framework composite comprises a shaped water adsorbent composite body comprising an elongate body having a circular or regular polygonal cross-sectional shape.

8. The water adsorbent according to claim 1, wherein the at least one water adsorbent metal organic framework comprises at least one of aluminium fumarate, MOF-801, MOF-841, CAU-10, MOF-303, MOF-573, MOF-802, MOF-805, MOF-806, MOF-808, MOF-812, or mixtures thereof.

9. The water adsorbent according to claim 1, wherein the at least one water adsorbent metal organic framework composite has a particle size of less than 800 μm.

10. An apparatus for capturing a water content from air, the apparatus comprising: a water adsorbent according to claim 1; and a water desorption arrangement in contact with and/or surrounding the water adsorbent, the water desorption arrangement being selectively operable between (i) a deactivated state, and (ii) an activated state in which the arrangement is configured to apply heat, a reduced pressure or a combination thereof to the water adsorbent to desorb a water content from the water adsorbent.

11. The apparatus of claim 10, wherein the at least one water adsorbent metal organic framework composite comprises a coating applied to the surface of the water desorption arrangement or is in thermal conductive contact with the water desorption arrangement.

12. The apparatus according to claim 10, wherein the water desorption arrangement includes at least one heat transfer arrangement in direct thermal conductive contact with the water adsorbent and the heat transfer arrangement is in thermal conductive contact with a heating device.

13. The apparatus according to claim 12, wherein the heat transfer arrangement includes at least one heat transfer element that extends from the heating device to the water adsorbent.

14. The apparatus according to claim 12, further including a condenser system for cooling the product gas flow from the water adsorbent, wherein the heating device comprises at least one peltier device, and each peltier device has a hot side and a cold side, with the hot side of each peltier device being in thermal communication with the at least one heat transfer arrangement, and the cold side of each peltier device forming part of the condenser system.

15. The apparatus of claim 14, wherein the cold side of each peltier device is in thermal communication with at least one heat transfer arrangement.

16. The apparatus according to claim 14, wherein the peltier device is capable of heating the packed bed to at least 50° C.

17. The apparatus according to claim 12, wherein the water adsorbent is housed within or coated on at least part of the heat transfer arrangement.

18. A method of capturing a water content from air, comprising at least one cycle of: feeding air over a water adsorbent according to claim 1 such that the water adsorbent adsorbs water from the air; and applying heat, a reduced pressure or a combination thereof to the water adsorbent so to release at least a portion of the adsorbed water therefrom.

19. The method according to claim 18, wherein the water adsorbent is a shaped water adsorbent metal organic framework having an average surface area of at least 700 m.sup.2/g.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The present invention will now be described with reference to the Figures of the accompanying drawings, which illustrate particular preferred embodiments of the present invention, wherein:

(2) FIG. 1A is a schematic of a magnetic induction swing apparatus for capturing a water content from a water containing gas according to one embodiment of the present invention.

(3) FIG. 1B is a schematic of a magnetic induction swing apparatus for capturing a water content from a water containing gas according to another embodiment of the present invention.

(4) FIG. 1C is a schematic of a temperature swing apparatus for capturing a water content from a water containing gas according to one embodiment of the present invention which includes a heat sink (CPU cooler) and peltier device.

(5) FIG. 1D is a photograph of the experimental temperature swing apparatus for capturing a water content from a water containing gas according to one embodiment of the present invention which includes a heat sink (CPU cooler) and peltier device.

(6) FIG. 1E provides schematic diagrams of operation of the thermal cycle water harvesting device shown in FIG. 1C when operated (A) during the adsorption phase; and (B) during the desorption phase.

(7) FIG. 2A is a photograph of the experimental setup used for aluminium fumarate synthesis.

(8) FIG. 2B is a photograph of the produced aluminium fumarate after washing procedure.

(9) FIG. 2C provides a schematic diagram of a shaped adsorption body cording to one embodiment of the present invention utilised in the packed bed of the apparatus shown in FIGS. 1A to 1D.

(10) FIG. 3A is photograph of the hand extruder and triangular shaped nozzle used to produce the shaped aluminium fumarate composite pellets.

(11) FIG. 3B provides a schematic diagram of the pellet forming process.

(12) FIG. 4 is a photograph showing the produced aluminium fumarate and aluminium fumarate composite pellets, being (A) Pristine MOF; (B) 1 wt % binder; (C) 1 wt % MNP; (D) 3 wt % MNP; and (E) 5 wt % MNP.

(13) FIG. 5 is a schematic of the experimental setup used for induction heating experiments.

(14) FIG. 6 is a schematic of the experimental magnetic induction swing water capturing rig.

(15) FIG. 7 provides a PXRD pattern of aluminium fumarate, simulated aluminium fumarate and aluminium fumarate with 1 wt % binder (Batch I).

(16) FIG. 8 provides a PXRD pattern of different aluminium fumarate magnetic composites, aluminium fumarate with 1 wt % binder (Batch I) and magnesium ferrite as reference.

(17) FIG. 9 provides a PXRD pattern of aluminium fumarate magnetic composite, aluminium fumarate (Batch II) and magnesium ferrite as reference.

(18) FIG. 10 provides a SEM image of aluminium fumarate metal organic framework (Batch II). Magnification: 10000 times.

(19) FIG. 11 provides a SEM image of magnesium ferrite nanoparticles. Magnification: 10.000 times.

(20) FIG. 12 provides a SEM image of aluminium fumarate magnetic framework composite (Batch II) at a magnification of 10000 times. The circled sections marked with A indicate the location of magnesium ferrite nanoparticles in the composite.

(21) FIG. 13 provides an averaged BET surface area of aluminium fumarate composites as a function of magnetic nanoparticle loading.

(22) FIG. 14 provides a plot of the pore size distribution of aluminium fumarate MOF pellets (Batch I).

(23) FIG. 15 provides a nitrogen isotherm of aluminium fumarate pellets.

(24) FIG. 16 provides a plot of the pore size distribution of aluminium fumarate composite pellets containing: (a) 1 wt % binder. (b) 1 wt % MNPs (c) 3 wt % MNPs (d) 5 wt % MNPs.

(25) FIG. 17 provides water vapour adsorption isotherms for aluminium fumarate batch I and aluminium fumarate batch I composite pellets collected at room temperature.

(26) FIG. 18 provides water vapour adsorption isotherms of aluminium fumarate batch II and aluminium fumarate batch II composite pellets collected at room temperature.

(27) FIG. 19 provides a plot of the initial heating rate of induction heating of Aluminium Fumarate magnetic framework composites with different MNP concentrations. Field strength was 12.6 mT.

(28) FIG. 20 provides a plot of the efficiency of induction heating of Aluminium Fumarate magnetic framework composites with different MNP loading. Field strength was 12.6 mT.

(29) FIG. 21 provides a plot of the normalized relative humidity over time for adsorption of water vapour from a nitrogen stream.

(30) FIG. 22 provides a plot of the temperature profile of aluminium fumarate composites during adsorption of moisture.

(31) FIG. 23 provides a plot of the normalized relative humidity over time for the out coming stream during regeneration.

(32) FIG. 24 provides a plot of the temperature profile of aluminium fumarate composites during regeneration of water vapour.

(33) FIG. 25 provides a plot comparing the water vapour uptake isotherms of (A) a first batch of AlFu (Aluminium Fumarate (I)); (B) pellets comprising Aluminium Fumarate (I) and a cellulose siloxane binder; (C) a second batch of AlFu ((Aluminium Fumarate (II)); and (D) pellets comprising Aluminium Fumarate (II) and a hydroxypropyl cellulose binder.

(34) FIG. 26 illustrates the setup of the testing rig for the temperature swing water harvesting device shown in FIGS. 1C and 1D, including with power supplies and measurement equipment.

(35) FIG. 27 illustrates the FTIR pattern of aluminium fumarate with 1 wt % binder after the pelletisation process for each of the three batches (batch_01, batch_02 and batch_03) and t the FTIR pattern of pristine aluminium fumarate.

(36) FIG. 28 illustrates the PXRD pattern of pristine aluminium fumarate and aluminium fumarate with 1 wt % binder of all three extrusions after the pelletisation process. Simulated pattern for comparison.

(37) FIG. 29 illustrates water uptake isotherms at 26° C. of aluminium fumarate pellets produced in this work (Pellets_02) (squares), and literature data of aluminium fumarate from Teo et al. [28] (diamonds).

(38) FIG. 30 illustrates mass logging of an adsorption phase with a humidity of 8:85 μm.sup.−3. This data is used to calculate theoretical adsorption times for all water harvesting cycles.

(39) FIG. 31 illustrate the optimisation of the condensation time of the water harvesting device plotting space time yield and specific energy over different condensation times corresponding to water harvesting cycles 12, 14, 15, 16 and 17.

(40) FIG. 32 illustrates optimisation of the desorption temperature of the water harvesting device plotting space time yield and specific energy over different desorption temperatures corresponding to water harvesting cycles 16, 18, 19 and 20.

(41) FIG. 33 illustrates temperature and relative humidity of the adsorption of water harvesting cycle 16.

(42) FIG. 34 illustrates the temperatures in the water harvesting device during the desorption phase of water harvesting cycle 16.

(43) FIG. 35 illustrates the relative humidity, dew point and condenser temperature in the water harvesting device during the desorption phase of water harvesting cycle 16.

(44) FIG. 36 illustrates the temperature and relative humidity of the adsorption of water harvesting cycle 24.

(45) FIG. 37 illustrates the temperatures in the water harvesting device during the desorption phase of water harvesting cycle 24.

(46) FIG. 38 illustrates the relative humidity, dew point and condenser temperature in the water harvesting device during the desorption phase of water harvesting cycle 24.

(47) FIG. 39 illustrates the temperature and relative humidity of the adsorption of water harvesting cycle 22.

(48) FIG. 40 provides two views of a prototype water capture apparatus using the temperature swing water harvesting embodiment showing (A) external housing; and (B) inner components, including louver system.

DETAILED DESCRIPTION

(49) The present invention provides an apparatus that provides selective control of the adsorbing and desorbing phases of a MOF based water adsorbents water harvesting cycle. The apparatus includes a water desorption arrangement which allows the MOF based water adsorbent to adsorb water when in a deactivated state, and then apply desorption conditions to the water adsorbent to desorb water from the water adsorbent when in an activated state. This selective operation of the water desorption arrangement between the deactivated and activated states enables the efficiency of water desorption arrangement to be optimised using more efficient energy desorption arrangements to desorb water from the metal organic framework based water adsorbent compared for example to utilising solar energy, and in some embodiments that can simultaneously condense the water content of any product gas flow.

(50) Adsorption Apparatus

(51) The water desorption arrangement can take any number of forms depending on whether heat and/or reduced pressure is being used to cause the adsorbed water to desorb from the water adsorbent. In some embodiments, the apparatus is designed for pressure swing adsorption, with desorption being achieved by reducing the pressure for example using a vacuum pump to evacuate the gas from around the water adsorbent. Adsorption would typically be undertaken at near atmospheric pressure. In other embodiments, temperature swing adsorption is undertaken to achieve water harvesting. This can be achieved using direct heating methods, or in some cases using magnetic induction swing adsorption.

(52) Magnetic Swing Water Adsorption Apparatus

(53) In some cases, the apparatus can be configured as a magnetic swing water adsorption apparatus to harvest a water content from a water containing gas, such as atmospheric air. One form of this type of apparatus 200 is illustrated in FIG. 1A or 1B.

(54) FIGS. 1A and 1B illustrate an apparatus 200 for capturing a water content from a water containing gas that uses a shaped water adsorbent composite body formulated with magnetic particles as discussed above. The apparatus 200 comprises a cylindrical housing 205 which includes inlet 208 and outlet 211. Housing 205 contains a packed bed 215 of shaped water adsorbent composite bodies 100 (see FIG. 2C), the composition of which is described in more detail below. A fluid distributor disc 210 proximate the base and lid/top of the housing 205 is used to retain the shaped adsorption material 215 between the discs 205. Each fluid distributor disc 210 comprises a metal disc with multiple holes drilled therethrough to allow fluid to flow through the packed shaped adsorption material. The shaped adsorption material forms a compressed packed bed between the discs 210, and are compressed therebetween so that the adsorbent shaped bodies 100 are tightly packed therein, thereby avoiding any flow short circuiting.

(55) In the embodiment shown in FIG. 1A, an alternating current (AC) induction coil 250 is located within and surrounded by the packed bed 215 of shaped water adsorbent composite bodies 100 (FIG. 2C). The induction coil 250 is configured to apply an AC magnetic field to the packed bed 215 of shaped water adsorbent composite bodies. The induction coil 250 is embedded within the packed bed 215 to optimise the use of the applied magnetic field when the induction coil 250 is operated.

(56) The housing 205 includes magnetic dampening material 255 to reduce magnetic field leakage from the container to the surroundings. This can be important in some applications where a magnetic field could deliriously affect the operation of proximate equipment, or irradiate people or objects.

(57) In the embodiment shown in FIG. 1B, an alternating current (AC) induction coil 250 is located external of the housing 205, but in a location around the housing which extends around the packed bed 215 of shaped water adsorbent composite bodies 100. Again, the induction coil 250 is configured to apply an AC magnetic field to the packed bed 215 of shaped water adsorbent composite bodies. However, it should be appreciated that the positioning of this induction coil 250 is not as energy efficient as shown in FIG. 1A due to losses through the material of housing. Furthermore, whilst not shown in FIG. 1B, a further housing may be used to enclose the induction coil which includes magnetic dampening material 255 to reduce magnetic field leakage to the surroundings.

(58) In use, a water containing gas is flowed through the packed bed of shaped bodies 215 such that the shaped water adsorbent composite bodies adsorb water from the water containing gas. Once the packed bed 215 reaches a desired saturation (typically 70 to 90% saturation point), the induction coil 250 is operated to apply an alternating current magnetic field thereby generating heat within the shaped water adsorbent composite bodies, so to release at least a portion of the adsorbed water therefrom into a product fluid flow. The shaped water adsorbent composite bodies therefore undergo magnetic induction vacuum swing adsorption to capture water from the water containing gas fed into the packed bed of shaped water adsorbent composite bodies 215.

(59) Whilst not shown in FIG. 1A or 1B, a condenser can be used to subsequently separate the water content of the product fluid flow (typically gas with entrained water vapour) to produce a captured water product. A low or reduced pressure (sometimes referred to as a vacuum environment), or a positive pressure gas flow, for example a flow of the water containing gas or another gas such as an inert or other dry gas, to direct the released water to the condenser.

(60) The above described method is cyclically applied, where the steps of adsorbing water in the shaped water adsorbent composite bodies 100, releasing that adsorbed water through application of the AC magnetic field and condensing that water is conducted in a repetitive cycle so to continuously produce water.

(61) Temperature Swing Water Adsorption Apparatus

(62) A temperature swing water harvesting apparatus 300 configured in according to an embodiment of the present invention is illustrated in FIGS. 1C, 1D and 1E.

(63) The apparatus 300 shown in FIGS. 1C and 1D is configured to use the waste heat of a peltier device 310 to heat up shaped MOF composite bodies 100 (the composition of which is described in more detail below) placed in thermal contact with the hot side 312 of the peltier device 310 (via a heat sink 320, discussed below) to facilitate desorption of adsorbed water in the shapes MOF composite bodies. The cold side 314 of the peltier device 310 can be simultaneously used to condense the desorbed water vapour, and that condensed water can be collected as a liquid product below the peltier device 310.

(64) Peltier Devices

(65) A peltier device is a thermoelectric device with the ability to convert electrical energy into a temperature gradient, generally termed the “peltier effect”. An electrical current applied to a pair of different metal materials leads to a hot surface on the one side and a cold surface on the other side of the semiconductors and creates a heat flow through the semiconductors perpendicular to the current flow. A single pair of a p- and n-type semiconductor material coupled in series is sufficient to create a temperature gradient when a current is applied from the n-type semiconductor to the p-type semiconductor. A cold side of a peltier element is formed where the electrons flow from p- to n-type semiconductors and a hot side with the heat flow Q.sub.dis appearing on the transition from n-type to p-type semiconductors. It should be appreciated that the dissipated heat of a peltier device is higher than the electrical power due to the absorbed heat on the cold side of the peltier device.

(66) Peltier devices are typically built of 3 up to 127 semiconductor pairs per device. The semiconductors are electrically connected in series and thermal in parallel. The heat flow in commonly available peltier devices is between 1 W to 125 W. The temperature difference between the hot and cold side of a peltier device is up to 70K for single-stage devices and up to 130K for multi-stage devices (several peltier elements connected in series).

(67) Mechanical stress can occur in a peltier device due to the high temperature difference and thus material expansion difference between the cold and hot sides. The dimensions of peltier devices are therefore typically limited to 50 by 50 mm to keep such mechanical stress issues low. Current Peltier devices also suffer from a low efficiency of about 10% of the possible Carnot efficiency due mainly to the available properties semiconductor material used in the specific peltier device.

(68) Temperature Swing Desorption

(69) FIGS. 1C and 1D illustrate an embodiment of the temperature swing water harvesting apparatus 300. As shown, the device 300 comprises a sealable container 330 having a container body 332 and sealing lid 334. The container body 332 houses which a polycarbonate plate 338 positioned and spaced away from the base of the container body 332 using spacers 339 to define within the container 330 (i) an upper water adsorption-desorption chamber 340; and (ii) a lower condenser chamber 342. The container 330 is sealable using the removable sealing lid 334. Within the container body 332 sits a water harvesting device 350. The water harvesting device 350 includes the following sections:

(70) (A). A heat sink 320 including a plurality of spaced apart fins 352. Whilst not shown in detail, the space between each of the spaced apart fins 352 is filled with shaped water adsorbent composite bodies 100 forming a packed bed 355 therein;

(71) (B). A peltier device 310 having a hot side 312 in thermal communication with the heat sink 320 and a cool side 314 in thermal communication with the gas space of the lower condenser chamber 342. The peltier device 310 is configured to heat the shaped water adsorbent composite bodies 100 in the heat sink 320 during a desorption phase of a water harvesting cycle (see below); and
(C). A condenser system 360 located in the condenser 342, which uses the cool side 314 of the peltier device 310 to cool a fluid flow of water vapour that is produced from the packed bed 355 to condense and collect the water as a liquid product at the base of the container 330.

(72) The apparatus 300 shown in FIGS. 10, 1D and 1E utilise both the cold side 314 and hot side 312 of the peltier device 310 during the desorption phase of a temperature swing water harvesting cycle. The dissipated heat of the hot side 312 of the peltier device 310 can be used in a temperature swing desorption cycle to heat up the shaped MOF composite bodies 100 during the desorption phase to desorb water from the shaped MOF composite bodies 100. The cold side 314 can be used to adsorb heat from the produced water vapour, and condense that water in a condenser system/chamber, to enable water to be collected as a liquid product.

(73) For this application, it should be appreciated that the key criteria in selecting the peltier device are: Capability to provide sufficient heating so that at the hot side of the peltier device water is desorbed from the MOF composite. Capability to provide sufficient cooling for the cold side of the peltier device to be below the dew point in the condenser system for condensation to occur. Other factors including reliability and resistance to corrosion.
The lowest powered peltier device to be able to this will result in the highest efficiency device.

(74) In the illustrated system (see FIG. 1D), the water harvesting device 350 is mounted within a 10 L sealable food container. The heat sink 320 comprises two NH-D15S (Rascom Computer distribution Ges.m.b.H., Wien (Austria)) CPU coolers. However, it should be appreciated that other suitable heat sink configurations could equally be used. This type of CPU cooler has a surface area of 1:0634 m.sup.2 and a free volume of 0:9967 L. This type of heat sink 320 is used as the dimensions of peltier devices are limited to sizes of around 40 mm by 40 mm due to heat stress issues (as discussed previously). The heat sink 320 ensures heat is distributed from the peltier device(s) 310 to a much larger surface, which can be used for conductive heat transfer to heat up the shaped MOF composite bodies in the packed bed 355.

(75) The heat sink 320 has a mounting socket that fits perfectly onto a peltier device and conducts the heat with 12 heat pipes 356 to 90 metal fins 352. 45 fins 352 are stacked on top of each other with a distance of 1:92 mm. Two of these heat sink 320 stacks are assembled side by side onto the mounting socket of the heat sink 320. A 12 V fan 370 is mounted between the two heat sink 320 stacks to provide an air flow through the free volume between the fins 352 during adsorption and desorption phase. However, it should be appreciated that the fan could be included in other locations proximate to the heat sink 320 stacks. The heat sink 320 and the peltier device 310 are mounted onto the polycarbonate plate 338. The heat sink 320 is fastened onto the peltier device 310 using screws. Heat grease is applied on the connection surfaces between the peltier device 310 and the heat sink 320 to ensure sufficient heat flow through this connection. The fan 370 is selected to produce a flow rate from 3 m.sup.3/hr to 200 m.sup.3/hr. In the illustrated embodiment, the fan comprises a 12 V fan capable of flow rates up to 140 m.sup.3/hr. In most test runs it was set on a low setting generating approximately 30 m.sup.3/hr. This flow rate can be can be tuned according to the ambient humidity conditions

(76) Whilst not illustrated, an additional small heat sink can be fixed to the cold side 314 of the peltier device 310 to increase the surface area for water condensation. It should be appreciated that the cold side 314 of the peltier device 310 with the small heat sink forms the condenser system 360 of the water harvesting device 350.

(77) As indicated above, the free volume between the fins 352 is filled with the shaped MOF composite bodies 100. In the illustrated embodiment (FIGS. 1C to 1E), the MOF composite bodies 100 comprise aluminium fumarate triangle shaped pellets with a side length S=1:5 mm and a length L=3 mm (see FIG. 1). The heat sink 320 is sealed with a netting (not illustrated) having a small enough aperture to retain the pellets between the fins 352 of the heat sink 320. The netting comprises a commercially available fly wire having an aperture of 1 mm. 200.30 g MOF pellets are packed between the fins 352. This equals a packing density of 0:20 kg/L. Thus 198:30 g of aluminium fumarate is used as adsorbent in the water harvesting device 350.

(78) In the illustrated test rig (see FIG. 1D and FIG. 26), six thermocouples 375 are fixed into the heat sink 320 to observe the temperature and the temperature distribution in the MOF packed bed 355 during the desorption phase. All six thermocouples are placed in one of the two sides of the heat sink 320. Three of the thermocouples are in the centre of the fins 352 in three different heights. The other three thermocouples are on the right side of the fins 352 in three different heights.

(79) A water harvesting cycle (WHC) using this apparatus 600 can be designed with two phases:

(80) 1. An Adsorption Phase (FIG. 1E(A)) —during which the sealable container 330 is opened to the environment (i.e. lid 334 removed) and air is blown through the MOF packed bed 355 in the heat sink 320 using the fan 370. Water in the air is adsorbed by the shaped water adsorbent composite bodies 100 of the packed bed 355. During this phase, the peltier device 310 is switched off. Once the packed bed 355 reaches a desired saturation (typically 70 to 90% saturation point), the lid 334 is placed on the container body 332 to seal the container 330 and the peltier device 310 is switched on to start the desorption phase.
2. A Desorption Phase (FIG. 1E(B)) —where the peltier device 310 is switched on and the packed bed 355 in the heat sink 320 is heated up to elevated temperatures so to release at least a portion of the adsorbed water from the shaped water adsorbent composite bodies 100 in the packed bed 355 into a product fluid flow while the container 330 is sealed closed. The relative humidity in the container 330 increases to high values and water condenses on the cold side 314 of the peltier device 310. Liquid water is collected under the cold side 314 of the peltier device 310 after each water harvesting cycle.

(81) The above described water harvesting cycle is cyclically applied, where the steps of adsorbing water in the shaped water adsorbent composite bodies 100, releasing that adsorbed water through operation of the peltier device and condensing that water is conducted in a repetitive cycle so to continuously produce water.

(82) Adsorption Medium

(83) The apparatus illustrated in FIGS. 1A to 1E each use a shaped water adsorbent composite body 100 (FIG. 2C) in a packed bed as the water adsorbent. However, it should be appreciated that the metal organic framework composite can be provided in an apparatus of the present invention in any form suitable to the particular apparatus configuration. The inventors envisage that this may be in any number of composite forms including (but not limited to) shaped bodies (for example pellets or extrusions), coatings, plates, sheets, strips or the like.

(84) Shaped Metal Organic Framework Composite Body

(85) The shaped water adsorbent composite body 100 (FIG. 2C) used in the apparatus discussed in relation to the apparatus shown in FIGS. 1A to 1E comprises a mixture of water adsorbent metal organic framework (MOF), and a hydrophilic binder which is optimised for use in a packed bed adsorption system. That mixture is composed of at least 50 wt % water adsorbent metal organic framework and at least 0.1 wt % hydrophilic binder.

(86) In the embodiments shown in FIGS. 1A and 1B, the shaped water adsorbent composite body 100 is configured to harvest water using a magnetic induction swing adsorption system. In these embodiments, the shaped water adsorbent composite body additionally contains from 0.2 to 10 wt % magnetic particles having a mean particle diameter of less than 200 nm. The use of magnetic particles in the composition forms enables inductive heat generation to be used for water desorption. This type of composite, known as a magnetic framework composite, combines the exceptional adsorption performance of MOFs and the high efficiency of magnetic induction heating.

(87) The metal organic framework composite material can be shaped into any suitable configuration for use in a packed adsorption system. In the present invention, the metal organic framework composite material is exemplified as a elongate shaped water adsorbent composite body 100 having a triangular cross-section, for example as shown in FIG. 2C. However, it should be appreciated that other shapes for example spherical, cylindrical, cubic, ovoid or the like could equally be used.

(88) Referring to FIG. 2C, the shaped water adsorbent composite body 100 comprises an elongate body having an equilateral triangle cross-sectional shape. The sides S of the equilateral triangle are at least 1 mm in length, preferably between 1.0 and 1.5 mm in length. The shaped water adsorbent composite body is preferably from 1 to 5 mm in length (longitudinal length, L), more preferably 1 to 4 mm in length. The elongated triangular shape is selected to increase packing density of the shaped water adsorbent composite bodies 100 within a packed bed (for example packed bed 215 shown in FIGS. 1A and 1B). Previous studies have shown that this shape has one of the highest packing densities in packed bed configurations. A high packing density is preferred for optimum utilisation and heat generation from an applied heat source. For example, a cylindrical pellet shape has a packing density of around 0.19 kg/L. An elongated equilateral triangular shaped pellet has a packing density of around 0.29 kg/L.

(89) Water Adsorbent Metal Organic Framework

(90) The water adsorbent metal organic framework used in the shaped water adsorbent composite body 100 can be selected from a range of suitable water adsorbent MOFs. A wide variety of water adsorbent MOFs are known, for example as discussed in Furukawa et al “Water Adsorption in Porous Metal-Organic Frameworks and Related Materials” Journal of the American Chemical Society 136(11), March 2014 and H W B Teo and A Chakraborty 2017 IOP Conf. Ser.: Mater. Sci. Eng. 272 012019 the contents of which should be understood to be incorporated into this specification by these references. In selected embodiments, the water adsorbent metal organic framework comprises at least one of aluminium fumarate, MOF-303 (Al), MOF-573 (Al), MOF-801 (Zr.sub.6O.sub.4(OH).sub.4(fumarate).sub.6), MOF-841 (Zr.sub.6O.sub.4(OH).sub.4(MTB).sub.2(HCOO).sub.4(H.sub.2O).sub.4), M.sub.2Cl.sub.2BTDD (including Co.sub.2Cl.sub.2BTDD), Cr-soc-MOF-1, MIL-101(Cr), CAU-10, alkali metal (Li+, Na+) doped MIL-101(Cr), MOF-802 (Zr.sub.6O.sub.4(OH).sub.4(PZDC).sub.5(HCOO).sub.2(H.sub.2O).sub.2), MOF-805 (Zr.sub.6O.sub.4(OH).sub.4[NDC—(OH).sub.2].sub.6), MOF-806 (Zr.sub.6O.sub.4(OH).sub.4[NDC—(OH).sub.2].sub.6), MOF-808 (Zr.sub.6O.sub.4(OH).sub.4(BTC).sub.2(HCOO).sub.6), MOF-812 (Zr.sub.6O.sub.4(OH).sub.4(MTB).sub.3(H.sub.2O).sub.4) or a mixture thereof. Preferred water adsorbent metal organic frameworks are aluminium fumarate, MOF-303 (Al), MOF-801, MOF-841, M.sub.2Cl.sub.2BTDD, Cr-soc-MOF-1, and MIL-101(Cr).

(91) Optimising the selection of a water adsorbing MOF involves a number considerations, including: 1. Water stability—the MOF should be water stable. 2. Adsorption reproducibility, the MOF should retain adsorption capacity after multiple adsorption/desorption cycles, preferably at least 10 cycles, more preferably at least 100 cycles. 3. Ease of production, the MOF should be easy to produce from readily available precursor materials. 4. High water uptake from air even at low humidity values. 5. A good affinity for water. The MOF should have a good enough affinity for water to enable the MOF to adsorb the water, but not have too high affinity for water that excessive energy needs to be expended to desorb water therefrom. Here the thermodynamics of water adsorption and desorption need consideration to ensure the MOF does not require excessive energy (kJ/mol MOF) to desorb water therefrom, and thereby adversely affect the energy efficiency of the system. Typical heats of adsorption for water for the MOF range from 10 to 100 kJ/mol MOF for water adsorbed on the MOF (550 to 5500 kJ/kg). Careful MOF selection is important to the operation of the device as the cost of the water will be directly linked to the energy required to desorb the water from the MOF.

(92) Where the MOF is required for water production for human consumption, the MOF and other materials must also meet food for human consumption regulations in relevant countries. The Applicant has found that the water adsorbent MOF preferably comprises aluminium fumarate (AlFu) MOF in these embodiments. The water adsorption properties of AlFu are published in a number of research studies available in the published literature.

(93) Aluminium Fumarate

(94) Aluminium fumarate (AlFu) is used as a preferred MOF in the shaped water adsorbent composite body 100. The structure and water adsorption properties of AlFu are well known, for example as detailed in Teo et al. (2017). Experimental study of isotherms and kinetics for adsorption of water on Aluminium Fumarate. International Journal of Heat and Mass Transfer Volume 114, November 2017, Pages 621-627, the contents of which are to be understood to be incorporated into this specification by this reference. As outlined in Teo, the crystal structure of AlFu resembles MIL-53 as it also consists of infinite Al OH Al chains connected by fumarate linkers. AlFu has a permanently porous 3D structure of formula [Al(OH)(O.sub.2C—CH═CH—CO.sub.2)] with square channels.

(95) Overall, aluminium fumarate was selected as a preferred choice of MOF for the inventive water capturing apparatus and system due to: 1. Ease of manufacture—this MOF can be synthesised in water. Following synthesis, processing the MOF is simple as outlined in the Examples. 2. Good thermal stability and is highly water stable (unlike many other MOFs); 3. It is robust to handling in ambient conditions and can withstand multiple temperature cycles without degradation. 4. It has a well-studied water adsorption behaviour; 5. High water uptake from air even at low humidity values; Aluminium fumarate has a water capacity between 0.09 to 0.5 grams of water per gram of MOF depending on the relative humidity. The typical heat of adsorption of Aluminium fumarate for water is well known and ranges between 60 and 30 kJ/mol depending on the ambient humidity 6. It can be cheaply and easily produced using non-toxic constituents/precursor material—i.e. environmentally friendly synthesis and is easy to handle and process; and 7. Low cost of its constituents.

(96) Nevertheless, it should be appreciated that the MOF component of the present invention is not restricted to Aluminium fumarate, and that other water adsorbent MOFs can also be used in the composition of the water adsorbent composite body.

(97) Hydrophilic Binder

(98) The selection of the appropriate binder is also important to the overall properties of the shaped adsorption body. The inventors have surprisingly found that a hydrophilic binder must be used to impart optimal water adsorption properties to the shaped water adsorbent composite bodies. The inventors have also found that non-hydrophilic binders and in particular hydrophobic binders (for example cellulose siloxane) reduce/decrease the water adsorption properties of the shaped water adsorbent composite bodies. The use of a hydrophilic binder is therefore important for optimal moisture capture properties of the packed bed water adsorption system. However whilst other binders are also possible, it is again noted that particularly suitable hydrophilic binders can be selected from at least one of hydrophilic cellulose derivatives such as hydroxypropyl cellulose, hydroxyethyl methyl cellulose, hydroxypropyl methyl cellulose (HPMC), ethyl hydroxyethyl cellulose, methyl cellulose, or carboxymethyl cellulose (CMC) as previously set out in this specification. As indicated in the following examples, one exemplary hydrophilic binder is hydroxypropyl cellulose (HPC).

(99) Lubricant

(100) The shaped water adsorbent composite body can further comprise a lubricant content, preferably less than 0.5 wt % lubricant, and more preferably less than 0.1 wt % lubricant. Suitable lubricants include surfactants and their salts. Examples of suitable lubricants include magnesium stearate, aluminium oxide, sodium oleate, glycerides, di-glycerides, tri-glycerides, fatty acids, oils including silicon oils and mineral oils and mixtures thereof. As mentioned previously, the lubricant content can assist with the shaping and forming processes of the shaped water adsorbent composite body.

(101) Magnetic Particles

(102) The shaped water adsorbent composite bodies can be configured to harvest water using a magnetic induction swing adsorption system. In these embodiments, the shaped water adsorbent composite body 100 (FIG. 1) comprises a mixture composed of at least 50 wt % water adsorbent metal organic framework, at least 0.1 wt % hydrophilic binder and from 0.2 to 10 wt % magnetic particles having a mean particle diameter of less than 200 nm. The mixture is optimised for use in a packed bed adsorption system.

(103) As discussed previously, a wide variety of magnetic particles can be used in the inventive shaped adsorption body. In embodiments, the magnetic particles comprise a ferromagnetic, paramagnetic, or superparamagnetic particles (typically micro or nano-particle). In embodiments, the magnetic particles comprise metal chalcogenides. Suitable metal chalcogenides comprise magnetic particles comprising any combination of element or ionic form thereof M selected from at least one of Li, Na, K, Rb, Be, Mg, Ca, Sr, Ba, Sc, Y, Ti, Zr, Hf, V, Nb, Ta, Cr, Mo, W, Mn, Re, Fe, Ru, Os, Co, Rh, Ir, Ni, Pd, Pt, Cu, Ag, Au, Zn, Cd, Hg, B, Al, Ga, In, TI, Si, Ge, Sn, Pb, As, Sb, Bi, or their combinations, in combination with elements or elemental form of at least one of O, S, Se, or Te. In some embodiments, the crystallisation facilitators comprise metal chalcogenide having the formula M.sub.xN.sub.yC.sub.z, where M,N are selected from at least one of Li, Na, K, Rb, Be, Mg, Ca, Sr, Ba, Sc, Y, Ti, Zr, Hf, V, Nb, Ta, Cr, Mo, W, Mn, Re, Fe, Ru, Os, Co, Rh, Ir, Ni, Pd, Pt, Cu, Ag, Au, Zn, Cd, Hg, B, Al, Ga, In, TI, Si, Ge, Sn, Pb, As, Sb, Bi, C is selected from at least one of O, S, Se, Te, x is any number from 0 to 10, y is any number from 0 to 10 and z is any number from 0 to 10. The metal chalcogenide particles may in some embodiments have a core-shell structure in which the core comprises at least one metal chalcogenide as previously described and the shell comprises at least one metal chalcogenide as previously described. In some forms, the core-shell structure may include multiple shells. In embodiments, the magnetic particles comprise at least one of MgFe.sub.2O.sub.4, Fe.sub.3O.sub.4, C-coated Co, CoFe.sub.2O.sub.4, NiFe.sub.2O.sub.4, Pyridine-2,6-diamine-functionalized SiO.sub.2, or Pyridine-2,6-diamine-functionalized Fe.sub.3O.sub.4.

(104) The advantages of these magnetic materials are: Local heat generation—i.e. heat can be generated insitu the material by applying an AC magnetic field (as discussed previously) as opposed to using an external heating source; Fast heating of material, due to local heat generation avoiding thermal and energy loss through thermal heating of surrounding materials; and High energy conversion efficiency

(105) The combination of the magnetic particles with MOFs to form a magnetic framework composite material yields an adsorbent with exceptional adsorption behaviour as a result of the MOFs and high efficiency of induction heating as a result of the magnetic particles.

EXAMPLES

(106) The following examples use AlFu as the water adsorbent MOF in the magnetic framework composite material. It should be appreciated that the magnetic framework composite material could use any number of other water adsorbent MOFs through direct substitution of that MOF within the magnetic framework composite material pellets.

Example 1—Magnetic Induction Swing Water Harvesting

1. Magnetic Framework Composite Material

(107) The synthesis of AlFu and the preparation of shaped water adsorbent composite bodies comprising AlFu magnetic framework composite material (MFC), hereafter referred to as MFC pellets, are described. The examples demonstrate that the experimental system can produce water repeatedly, with 1.2 grams of water having been produced from roughly 3 cycles of the described method and system. Cycle times were approximately 30 minutes in duration. In the examples outlined below, 0.4 g of water was captured using 5 g of the inventive shaped composite material within 28 minutes. This provides the following production and energy use: Anticipated water production capacity: 4.3 L/kg of MOF a day with a cycle time of 28 mins; and Anticipated energy use: 12 kWh/L.

(108) As a comparison, the system described in Yaghi 1 and Yaghi 2 (referred to in the Background of the Invention section) uses sunlight as energy source for regeneration of the MOF. This device was reported as being capable of capturing 2.8 litres of water per kilogram of MOF daily at relative humidity levels as low as 20% at 35° C. in Yaghi 1. Yaghi 2 indicates that about 0.75 g of water was produced from 3 g of MOF within 16.5 hours in the same conditions. This equates to an anticipated water production capacity of 0.25 L/kg of MOF daily. The process of the present invention therefore has a significantly higher water production rate than the system taught in Yaghi.

(109) The inventors note that Yaghi 1 originally claimed that their device was capable of capturing 2.8 litres of water per kilogram of MOF daily at relative humidity levels as low as 20% at 35° C. However, this higher production rate appears to have been greatly overestimated in that paper, as further experimental work published in Yaghi 2 using the same set up reports a production rate being an order of magnitude lower at 0.25 litres of water per kilogram of MOF daily at 20% RH and 35° C. Inventors consider that the production rate published in Yaghi 2 reflects the actual production rate of this MOF-801 based system.

(110) 1.1 Preparation of Magnetic Framework Composites

(111) The synthesis of aluminium fumarate MOF and the preparation of aluminium fumarate MFC pellets are described.

(112) 1.1.1 Aluminium Fumarate Synthesis

(113) Within this work, two different scaled batches of aluminium fumarate were synthesized. For the evaluation of different contents of magnetic nanoparticles on moisture adsorption and induction heating performance of the composite material a smaller batch, designated Batch I, was prepared. While for experiments with the water capture rig 600 (see FIG. 6), a larger batch, designated Batch II, was synthesized.

(114) The experimental setup 400 of the aluminium fumarate synthesis reaction is presented in FIG. 2A.

(115) The two precursor solutions named A and B were synthesized as follows:

(116) For solution A, aluminium sulfate octadecahydrate was dissolved in deionized water using a magnetic stirrer 406. Precursor solution B was prepared by dissolving sodium hydroxide pellets and fumaric acid with a purity of 99% in deionized water under stirring with a magnetic stirrer (not shown). The composition of both solutions can be taken from Table 1.

(117) TABLE-US-00001 TABLE 1 Composition of precursor solutions for aluminium fumarate synthesis Precursor Solution A Precursor Solution B Aluminium Deionized Sodium Fumaric Deionized Sulfate Water Hydroxide Acid Water Batch I 35 g   150 ml 13.35 g  12.9 g   191 ml Batch II 90 g 385.7 ml 34.33 g 33.17 g 491.1 ml

(118) Solution B was then filled into a round bottom flask 410 and heated up to 60° C. using a heating mantel 408. A mechanical stirrer 402 was used to stir the liquid. When 60° C. were reached, precursor solution A was added. The mixture was then stirred for 20 minutes at 60° C. measured using temperature transducer 404.

(119) Afterwards, the suspension was filled into centrifuge tubes (not shown) and centrifuged for 8 to 10 minutes at 6000 rpm for Batch I and 4500 rpm for Batch II, respectively. The liquid was then removed from the sedimented MOF crystals. After that aluminium fumarate was washed using the following procedure. At first, deionized water was added to the MOF crystals. The suspension was then shaken by hand until the sediments were homogenously mixed up. Furthermore, the suspension was mixed for 15 minutes onto a roller mixer for Batch I and an orbital shaker for Batch II, respectively. Afterwards, the suspension was centrifuged using the same settings as mentioned before. After removing the liquid, the washing procedure with deionized water was repeated for another three times.

(120) Subsequently, aluminium fumarate was washed with methanol for one time following the same procedure as described before. Aluminium fumarate after the washing procedure is shown in FIG. 2B.

(121) After the washing steps, the MOF crystals were pre-dried overnight in a glove bag under nitrogen atmosphere. Afterwards, the MOF was dried overnight in an oven at 100° C. under nitrogen atmosphere. Subsequently, the temperature was increased to 130° C. and the oven was evacuated to activate the MOF for 6 to 8 hours.

(122) 1.1.2 Aluminium Fumarate Composite-Pellet Preparation

(123) A smooth paste needs to be prepared for the extrusion of MOF pellets. Therefore, the MOF was ground using a mortar and a pastel (not illustrated) for the smaller Batch I and a coffee grinder (not illustrated) for the larger Batch II, respectively. After grinding, the MOF was sieved through a 212 μm sieve. In case of Batch II, the aluminium fumarate powder was sieved through a 150 μm sieve. The MOF powder was then weighed into a jar. Afterwards, magnetic nanoparticles (MNPs) were added. In this work, magnesium ferrite was chosen as magnetic nanoparticles. However, it should be appreciated that other magnetic nanoparticles could equally be used. The powder mixture was then shook by hand for about 10 minutes until the colour of the powder was homogenously brownish. Afterwards, the powder was filled into a bowl and a hydrophilic binder (hydroxypropyl cellulose (HPC)) was added. To investigate the effect of the amount of magnetic nanoparticles on water uptake and magnetic induction heating, different composites were prepared. The composition of the prepared samples is provided in Table 2.

(124) TABLE-US-00002 TABLE 2 Composition of different aluminium fumarate composite samples Concentration of Magnesium Ferrite Concentration of Batch [wt %] Binder [wt %] Batch I 0 0 0 1 1 1 3 1 5 1 Batch II 3 1

(125) Furthermore, a solvent, in this case deionised water was added to make the mixture pastier. In case of Batch I, also small amounts of ethanol were added. The components were then well mixed until an ice cream like paste has formed.

(126) For the extrusion of composites made from Batch I, a syringe with a round nozzle was used (not illustrated). In case of Batch II, a hand extruder 500 with a triangular shaped nozzle 505 was chosen. The extruder is illustrated in FIG. 3A. The triangular shaped extrusion attachment 505 was chosen in order to increase the packing density of the produced pellets as explained previously. Furthermore, magnesium stearate was used as lubricant for preparation of pellets from Batch II. The magnesium stearate powder was greased onto the inner walls of the hand extruder 500. The paste was extruded onto filter paper and dried for at least 10 minutes. Afterwards, the extruded MOF was cut into 3 to 5 mm long pellets using a razor blade.

(127) The pellets were then dried in an oven at 100° C. under vacuum (reduced pressure of less than 100 mbar) for 24 hours. The different MOF composite pellets (Aluminium fumarate and aluminium fumarate composite pellets) are presented in FIG. 4 which show (A) Pristine MOF; (B) 1 wt % binder; (C) 1 wt % MNP; (D) 3 wt % MNP; and (E) 5 wt % MNP. A schematic of this overall process is shown in FIG. 3B.

(128) 1.2 Analysis of Aluminium Fumarate Composites

(129) The first part of this section deals with different analysis methods that have been used to characterize the structure of aluminium fumarate composites. Furthermore, methods that evaluate the performance of the composites regarding water uptake and magnetic induction heating are described.

(130) 1.2.1 X-Ray Diffraction

(131) All samples have been characterized using powder X-ray diffraction (PXRD) as well as small-angle X-ray scattering (SAXS) and wide-angle X-ray scattering (WAXS). For X-ray diffraction analysis, the pellets were ground first to fill them into the sample holder.

(132) Powder X-ray diffraction was performed employing a Bruker D8 Advance X-ray Diffractometer operating under CuKα radiation. The diffractometer was equipped with a Lynx Eye detector. All samples were scanned over the 28 range 5° to 105° with a step size of 0.02° and a count time of 1.6 seconds per step. To give an equivalent time of 284.8 seconds per step, 178/192 of the sensor strips on the Lynx Eye were used. The Bruker XRD search match program EVA™4.2 was used to perform analyses on the collected PXRD data.

(133) Aluminium Fumarate is not present in the JCPDS database. Therefore, for reference a simulation of the structure of aluminium fumarate was generated in TOPAS using a simplified model for the geometry of the mentioned diffractometer.

(134) Small- and wide-angle X-ray scattering was performed at the Australian Synchrotron. The samples were mounted onto sample holder plates. All samples despite of magnesium ferrite control samples were analysed with 1% Flux and an exposure time of 1 sec. Magnesium ferrite samples were analysed at 100% Flux.

(135) 1.2.2 Infrared Spectroscopy

(136) Infrared spectra analysis was performed using a Thermo Scientific Nicolet 6700 FT-IR spectrometer. The samples were analysed in the wavenumber range from 500 to 4000 cm.sup.−1.

(137) 1.2.3 Scanning Electron Microscope Imaging

(138) Specimens for scanning electron microscope (SEM) imaging were prepared by diluting the samples in water and then trickling the suspension onto a silicon waver. The silicon waver was then stuck onto a SEM specimen stub using carbon tape. Before scanning, the samples were coated with iridium to increase the signal to noise ration during microscopy. The SEM images were taken using a Carl Zeiss Gemini SEM 450 instrument at 10000 times magnification.

(139) 1.2.4 Surface Area and Porosity Measurements

(140) Surface area and porosity measurements of aluminium fumarate composites were analysed using a Micrometrics ASAP 2420 high throughput analysis system.

(141) At first, composite pellets were filled into pre-weighed analysis tubes and capped with Transeal caps. The samples were then degassed for 24 hours at 140° C. under vacuum. Afterwards, the tubes with the containing degassed samples were weighed to determine the mass of the dried pellets. The tubes were then transferred to the analysis ports of the instrument. Langmuir and Brunauer-Emmett-Teller (BET) surface areas as well as pore size distribution of all samples were determined by collecting nitrogen isotherms at 77K in a liquid nitrogen bath. Pore size distribution was determined using density functional theory (DFT).

(142) The BET surface area of samples made from batch I was measured three times in order to determine the variation within the analysis. The averaged surface area x was calculated using Equation 1.1:

(143) $\begin{array}{cc}\stackrel{\_}{x}=\frac{1}{n}\underset{i}{\overset{n}{.Math.}}{x}_i& \left(1.1\right)\end{array}$

(144) Where n is the total number of experiments and xi is the surface area of the experiment i.

(145) Furthermore, the standard deviation s.sub.n was calculated. This was done by using Equation 1.2.

(146) $\begin{array}{cc}{s}_n=\sqrt{\frac{1}{n}\underset{i}{\overset{n}{.Math.}}{\left(\stackrel{\_}{x}-x_i\right)}^2}& \left(1.2\right)\end{array}$
1.2.5 Water Uptake Capacity Determination

(147) Water Uptake capacity was measured using a Quantachrome Instruments Autosorb-1 analyser.

(148) The samples were filled into pre-weighed analysis tubes. After that, the material was degassed for 16 hours at 140° C. under vacuum. Afterwards, the weight of the dried pellets was determined. The tubes were then connected to the analysis port for water vapour adsorption measurement. In order to ensure a constant temperature during the analysis, the sample tubes were put into a water bath at room temperature. Water vapour uptake was measured using pure water vapour at relative pressures p/p.sub.0 from 0.1 to 0.5 with a step size of 0.1. The water vapour adsorption experiments have only been performed once because it takes almost one week to run a single isotherm.

(149) 1.2.6 Investigation of Magnetic Induction Heating

(150) To evaluate magnetic induction heating of magnetic framework composites, heating rate and efficiency of induction heating have been investigated.

(151) For the induction heating experiments an Ambrell Easy Heat 1.2 kW induction unit was used. The induction coil 560 that was attached to the work head is made from copper. It had three turns, an inner diameter of 4 cm and a length of 2.5 cm. A water chiller 562 was used to cool down the coil 560 during the experiment.

(152) The setup 550 for the induction heating and efficiency experiments is shown in FIG. 5. A certain amount of the sample was filled into a glass vial 565. To monitor the temperature increase of the sample over time during induction heating a fibre optic cable sensor 566 was introduced into the centre of the bed 567. The sensor 566 was connected to an OpSens FOTS100 temperature data logger 570. The glass vial 565 was put into the centre of the coil 560 so that the middle of the bed's height was in line with the middle of the coil's height.

(153) To monitor the energy that is consumed by the induction heating unit during the experiment a Cabac Power-Mate™ power meter 575 was used. The power that is needed to heat up the sample was calculated as following.

(154) At first, the coil was operated without any sample in the magnetic field to get a baseline. Therefore, the energy was measured for 5 minutes. The power was then calculated using Equation 1.3.

(155) $\begin{array}{cc}\mathrm{Power}\mathrm{consumed}=\frac{\mathrm{Energy}\mathrm{consumed}\mathrm{over}5\mathrm{minutes}\left[\mathrm{kWh}\right]}{\mathrm{Time}\left(5\mathrm{minutes}\right)\left[h\right]}& \left(1.3\right)\end{array}$

(156) During heating of the sample, energy was also measured for the first 5 minutes of induction heating. The power was then calculated the same way as mentioned before.

(157) For investigating the heating effect of magnetic framework composites exposed to an external magnetic field, temperature of the sample was measured over time for different composites and for different amounts of the composite pellets. All experiments were performed for more than 20 minutes. After this time the heating curve for induction heating was constant for all samples.

(158) The initial heating rate was used to quantify the induction heating effect. This rate was determined by calculating the linear slope of the temperature profile (dT/dt)t=0 at the beginning of the experiment. The heating curve is therefore approximated by Equation 1.4:

(159) $\begin{array}{cc}T\left(t\right)=T_0+\Delta T_{\max }\left[1-e^{\frac{t}{T}}\right]& \left(1.4\right)\end{array}$

(160) Where T.sub.0 is the initial temperature of the pellets, ΔT.sub.max is the saturation temperature increase (T.sub.max−T.sub.0) and T is the time constant of heating which corresponds to the time when the temperature reaches approximately 63% of ΔT.sub.max.

(161) Efficiency of induction heating of magnetic framework composites was quantified using Equation 1.5.

(162) $\begin{array}{cc}\mathrm{Efficiency}\%=\frac{P_{\mathrm{Coil}}}{P_{\mathrm{SAR}}}\times 100\%& \left(1.5\right)\end{array}$

(163) In this equation, P.sub.Coil is the power that is consumed by the coil during induction heating and it is calculated by using Equation 1.6.
P.sub.Coil=P.sub.consumed,heating MFC−P.sub.consumed, without MFC in field  (1.6)

(164) The specific adsorption rate (SAR) is usually used to estimate the heating effect of magnetic nanoparticles exposed to an external magnetic field. The SAR was determined by dispersing 10 mg of magnetic nanoparticles in 100 ml of deionized water. The suspension was then put into the centre of the induction coil and triggered with a magnetic field. The temperature increase of the suspension was measured using an optic fibre cable and a temperature data logger. The specific adsorption rate can then be calculated using Equation 1.7.

(165) $\begin{array}{cc}\mathrm{SAR}=C_{\mathrm{water}}\times \frac{m_{\mathrm{water}}}{m_{\mathrm{nanoparticles}}}\times {\left(\frac{\mathrm{dT}}{\mathrm{dt}}\right)}_{t=0}& \left(1.7\right)\end{array}$

(166) In this equation, C.sub.water is the specific heat capacity of water, m.sub.water is the mass of water, m.sub.nanoparticles is the mass of magnetic nanoparticles in the suspension and (dT/dt)t=0 is the initial heating rate. The initial gradient of the heating curve was calculated as mentioned before.

(167) Finally, the power that is generated by the magnetic nanoparticles in the composite PSAR can then be calculated using Equation 1.8.
P.sub.SAR=SAR×Magnetic nanoparticle content of MFC pellet (g)  (1.8)

(168) All experiments for determination of the initial heating rate and efficiency of induction heating were carried out for three times to determine a standard variation.

(169) 1.3 Proof of Concept Experiments for Magnetic Induction Vacuum Swing Adsorption

(170) This subsection deals with different methods that have been used to evaluate the performance of a magnetic induction vacuum swing adsorption process for water capture from ambient air. These experiments were carried out with a self-constructed water capturing rig on bench scale. The schematic process flow diagram of the Water capturing rig 700 is provided in FIG. 6.

(171) Moisturized nitrogen was used as test gas for water capture and breakthrough experiments. A nitrogen stream from gas supply 702 at 1 bar was split up into a dry gas stream 704 and a wet gas stream 706 that was moisturized by bubbling it through deionized water in bubbler 708. Flow of each stream were measured using flowmeters 710 and 711. The desired humidity of the feed stream for the adsorption column was reached by setting the ratio of the dry gas stream 704 and the wet gas stream 706.

(172) A vertically orientated adsorption column 720 was used comprising a 1 inch polyether ether ketone tube. A perforated Teflon spacer was glued in the bottom part of the tube to hold the adsorption bed thereon within the tube (not illustrated but enclosed within adsorption column 720). Furthermore, glass wool was used to prevent pellets from falling through the holes of the spacer. The tube was connected to the feed and outlet pipes using stainless steel ultra-Torr vacuum fittings purchased from Swagelok.

(173) In these experiments an Ambrell EasyHeat 3542 LI induction coil 725 with a system power of 4.2 kW was used. A copper coil with 5 turns, an inner diameter of 4 cm and a length of 5 cm was connected to the work head of the induction coil 725. The feed and outlet pressure were monitored using manometers 730. To measure the temperature in the adsorption bed, a fibre optic cable 732 was introduced into the middle of the packed bed and connected to a temperature data logger. A RS 1365 Data logging Humidity-Temperature Meter 735 was used to monitor the humidity of the outlet stream 745 of the adsorption column 720.

(174) For water capture, the out-coming stream 645 was lead through a cold trap 740 containing dry ice to condensate the moisture to produce water 755. A vacuum pump 750 was used to generate the driving force for the desorbed gas stream 745.

(175) 1.3.1 Water Collection Experiments

(176) For water collection experiments, aluminium fumarate magnetic framework composite pellets made from batch I containing 3 wt % magnetic nanoparticles were filled into the adsorption column. In order to increase the packing density, the pipe was tapped onto the bench for a few times.

(177) A dry nitrogen stream with different volume flow rates was flown through the packed adsorption column in order to determine the back pressure as a function of the flow rate.

(178) The determined back pressure was then used to calculate the desired relative humidity of the feed stream. For the desired humidity, desert like conditions were chosen. The relative humidity in these areas is about 35% at 35° C. and atmospheric pressure. To simplify the experimental set up, the water uptake experiments in this work were carried out at room temperature. This simplification is justified because the temperature dependence of water vapour uptake on aluminium fumarate in this temperature region is negligible. However, in order to get comparable results, calculations of humidity were based on water content in the desert like air. The dessert like conditions correspond to a water content of 11273 ppmV in a pure nitrogen atmosphere (calculated with Michel Instruments Humidity Calculator—http://www.michell.com/us/calculator/).

(179) Based on this water content, the ratio of the dry and the moisturized gas stream can then be calculated.

(180) For water capturing experiments, MFC pellets were activated by triggering them with an alternating current magnetic field and directing a flow of dry nitrogen stream through the column. Activation of the material was performed until the humidity of the out coming gas stream was zero. It is noted that this activation step was used for experimental date collection purposes only in order to obtain a dry MOF for measurement accuracy. A commercial system would not generally require this activation/drying step to be performed as any preadsorbed moisture in the MOF material in the system would simply be desorbed in the first cycle of operation of the system.

(181) Before charging the MFC pellets with water vapour, volume flow rates of the dry and the moisturized streams were set. The resulting feed stream was first vented to allow the humidification of the gas stream to stabilize. After three minutes, the valve in front of the adsorption column was switched to enable the feed to flow through the column.

(182) During moisture adsorption, the humidity of the outlet stream of the column was measured and noted down every 30 seconds in order to determine breakthrough curves. After the humidity of the vented stream had stabilized, the feed stream was turned off and the valve was closed towards the column.

(183) For regeneration, in a first experiment only humidity was measured during desorption in order to determine the duration of the regeneration step. Therefore, during induction heating, a dry nitrogen stream was lead through the adsorption bed to flush the column.

(184) In order to determine breakthrough curves, the relative humidity that was measured in the outlet stream of the column has been normalized. Therefore, the measured humidity was divided by the humidity of the out coming stream that is reached when the adsorption bed is saturated.

2. Results and Discussion

(185) In this section, results of characterisation and performance analysis of aluminium fumarate composites and its ability for capturing water from ambient air using a magnetic induction vacuum swing adsorption process are presented and discussed.

(186) 2.1. Structural Characterisation of Composite Material

(187) The PXRD pattern of aluminium fumarate and aluminium fumarate with 1 wt % of binder is presented in FIG. 7. Furthermore, a simulated pattern of aluminium fumarate is plotted. It can be seen that most of the peaks in the trace of the pristine MOF and the MOF with 1 wt % binder are a reasonable match to the simulated phase.

(188) The PXRD pattern of different aluminium fumarate magnetic framework composites made from Batch I can be taken from FIG. 8. Furthermore, the trace of magnesium ferrite is plotted as reference.

(189) The traces of all composites match the PXRD pattern of aluminium fumarate containing only the binder.

(190) For the composite containing 1 wt % of magnetic nanoparticles, the PXRD pattern does not reveal a significant evidence of magnesium ferrite in the sample. Presumably, any magnetic nanoparticles present in this sample are below the detection limit.

(191) The PXRD pattern of composites with 3 wt % and 5 wt % of magnesium ferrite look fairly similar with a trace of magnetic nanoparticles being visible at the 2θ angle 35°.

(192) The PXRD trace of the magnetic framework composite made from Batch II are presented in FIG. 9. The pattern matches the trace of the pristine MOF of this batch. Similar to the composites made from Batch I, there is a trace of magnesium ferrite visible at the 2θ angle 35°.

(193) The surface morphology of aluminium fumarate has been studied using scanning electron microscopy and is presented in FIG. 10. It can be observed from this Figure that this MOF has a quite rough surface and a poorly crystalline structure.

(194) The narrow particle size distribution of magnesium ferrite nanoparticles are presented in FIG. 11. The average particle size of this sample is about 150 nm.

(195) The incorporation of aluminium fumarate and magnesium ferrite nanoparticles can be observed from FIG. 12. The circled sections marked with A indicate examples of the location of magnesium ferrite nanoparticles in the composite. It can be seen, that the nanoparticles are fairly equally distributed within the MOF structure.

(196) The average BET surface area of aluminium fumarate magnetic framework composites as a function of magnetic nanoparticle loading is shown in FIG. 13. The sample containing no magnesium ferrite refers to composite pellets that were made only from the MOF and binder. The average surface area of pristine aluminium fumarate pellets is 976 m.sup.2/g. The standard deviation of the measurements for this sample is 33 m.sup.2/g.

(197) From the plot and the surface area of the pristine MOF it can be taken, that there is not a significant change in the BET surface area for adding the binder and for an increasing magnetic nanoparticle concentration. Even though there is a slight decrease in surface area visible for higher nanoparticle loadings, the average surface area is still in the same order of magnitude. Furthermore, the standard deviation for the different experiments does not substantiate the decreasing trend.

(198) The BET surface area of the second batch and composite pellets prepared from this batch is shown in Table 3. It can be seen that there is also not a significant difference in the surface area between the pristine MOF and its composite. However, the surface area of MOF pellets from batch II is lower that on pellets prepared from the first batch.

(199) TABLE-US-00003 TABLE 3 BET surface area of aluminium fumarate batch II and aluminium fumarate batch II composite Concentration MNPs BET Surface Area Sample [wt %] [m.sup.2/g] Pristine MOF 0 876 Composite 3 849

(200) The pore size distribution of aluminium fumarate pellets are presented in FIG. 14. The main pores size distribution is in the microporous area which is below 2 nm. The microporosity of this sample can also be confirmed with the nitrogen adsorption-desorption isotherm which is shown in FIG. 15. According to IUPAC classification of adsorption isotherms, the shape of this isotherm corresponds to a physisorption isotherm type H4. This shape is typical for microporous materials where the high uptake at low relative pressures is associated with the filling of micropores.

(201) FIG. 16 illustrates that most of the pores of aluminium fumarate composite pellets are also present in the microporous are below 2 nm. Only for the sample containing 3 wt % of magnesium ferrite nanoparticles, there are pores in the mesoporous area between 2 nm and 50 nm.

(202) 2.2 Performance Analysis of Aluminium Fumarate Composites

(203) In this subsection, water uptake results and magnetic induction heating performance of the prepared material are presented and discussed. Furthermore, the results of induction heating experiments are compared to a conventional heating method.

(204) 2.2.1 Water Uptake Performance

(205) The water vapour adsorption isotherms of samples prepared for pre studies on the effect of an increasing amount of nanoparticles in the MOF can be taken from FIG. 17. All isotherms were collected at room temperature. The plots show that there is not a significant difference between the moisture uptake capacities for the different composites. This was already shown with nitrogen isotherms in section 2.1. In order to check the accuracy of the measurements, the isotherm of the composite containing 5 wt % magnesium ferrite was collected two times. The Figure shows that the values of the second measurement vary by approx. 70% from the first measurement. Regarding this accuracy, it can be confirmed that moisture uptake of the composite pellets does not differ strongly from vapour uptake from pristine aluminium fumarate pellets.

(206) The water vapour isotherms of aluminium fumarate and aluminium fumarate composite pellets prepared from the larger batch are presented in FIG. 18. Regarding the experimental variance of moisture adsorption experiments, it can be indicated that the water vapour uptake of the composite does not significantly differ to the one of the pristine MOF.

(207) 2.2.2 Induction Heating Performance and Comparison to a Conventional Heating Method

(208) The induction heating performance was evaluated, using the initial heating rate. The initial heating rate was determined as described in Section 1.2.6. Results of these experiments are shown in FIG. 19. The field strength in these experiments was 12.6 mT. It can be seen, that the initial heating rate increases with the concentration of magnetic nanoparticles incorporated into the metal organic framework. Furthermore, the heating rate increases fairly linear with the amount of sample that is triggered by the alternating current magnetic field.

(209) In addition to the initial heating rate, the energy conversion efficiency of induction heating of the prepared composites is shown in FIG. 20. The field strength was 12.6 mT. Similar to the initial heating rate, the efficiency also increases with an increase in magnetic nanoparticle loading and an increase in sample weight. The increase in efficiency with an increasing sample mass is counterintuitive as one would know from conventional heating methods. However, with an increasing amount of magnetic framework composite pellets, the amount of magnetic nanoparticles triggered by the magnetic field also increases. Therefore, the coupling between the nanoparticles is improved.

(210) For applications on industrial scale, where much larger amounts of MFC pellets are used, the energy conversion efficiency is expected to be even higher than shown in this experiment. That is because of a loss of heat that is caused by non-existent insulation of the heated sample. The loss is not considered in the calculation of the SAR value. To minimize this heat loss, an adiabatic experimental set up needs to be used. However, the non-adiabatic system delivers quick and reliable SAR values without the need for extensive, time-consuming and expensive adiabatic measurements.

(211) In addition to that, the efficiency could be further improved by utilizing induction heating systems that are not water cooled. Water cooled systems require separate support systems with pumps and connections that increase complexity and costs of the system. Induction heating systems that do not need direct cooling have been reported to achieve up to 90% energy efficiency.

(212) 2.3 Proof of Concept: Magnetic Induction Vacuum Swing Adsorption

(213) The normalized humidity in the outlet stream of the column during adsorption of moisture is presented in FIG. 21. For this experiment, the relative humidity in the feed stream was set to 50% at a surrounding temperature of 22° C. This corresponds to the same moisture concentration that is present in the driest areas of the world. The volume flow rate for the moisturized and the dry nitrogen stream were both set to 4 SLPM. With these settings, the adsorption bed is fully saturated after approximately one hour. After about 17 minutes the humidity of the out coming stream stabilizes for approx. 8 minutes. This might be due to a sectional higher packing density along the column length which is caused by the inhomogeneity of the pellet length.

(214) In order to reduce the cycle time, for further experiments the breakthrough point where adsorption is stopped was set to the time when 90% of the maximum outlet humidity is reached. This is after approx. 27 minutes.

(215) The temperature that was measured during adsorption of water vapour is shown in FIG. 22. The temperature increases in the beginning due to the released heat of adsorption.

(216) The normalized humidity during regeneration is shown in FIG. 23. Due to the height of the adsorption bed, the induction coil needed to be placed at two different positions in order to heat up the whole material. First, the coil was placed at the upper part of the column. After approx. 2 minutes, the humidity increases drastically due to the rapid heating rate of magnetic induced heating. Almost 20 minutes later, the humidity of the out coming stream decreases as the water amount captured in the MOF also decreases. Right before the humidity in the outlet stream settles, the coil was moved to the lower part. The water that is still adsorbed on the material in the lower part of the column is therefore released. The power of the induction coil was shut off when the humidity reached zero.

(217) The temperature over time during regeneration of water vapour can be taken from FIG. 24. It can be seen from that Figure and the plot of the normalized humidity during regeneration that as soon as the temperature reaches about 50° C., water release starts. The temperature decreases as the coil was moved to the lower part of the column. That is because the temperature sensor sits above the middle of the adsorption bed.

(218) Based on the adsorption isotherms for moisture and the pre studies on the behaviour of the column regarding water adsorption and regeneration, a theoretical yield for the rig can be calculated.

(219) In order to evaluate the energy consumption and efficiency of a magnetic induction vacuum swing adsorption process, energy was measured during regeneration of the MOF composite. For these measurements the 1.2 kW induction heating system was chosen. The parameters for energy efficiency experiments can be taken from Table 4. The energy consumption was monitored using an energy data logger.

(220) TABLE-US-00004 TABLE 4 Parameter energy consumption measurements Parameter Value Unit Mass MFC pellet 5 g Flow rate dry N.sub.2 stream 4 SLPM Flow rate wet N.sub.2 stream 4 SLPM Surrounding temperature 22 °C. Current induction heating system 225.6 A Frequency induction heating system 268 kHz

(221) Before the actual experiment a breakthrough curve for the set up described in Table 4 was determined. Therefore, the activated MOF composite pellets were charged with water vapour for twenty minutes. After this time the adsorption bed was fully loaded with moisture. The breakthrough curve is presented in FIG. 21.

(222) However, in order to increase the overall efficiency of the process, the breakthrough point where adsorption is stopped for the experiments was chosen to be when 90% of the maximum outlet humidity was reached.

(223) After adsorption of water vapour, the regeneration was started and energy consumption of the induction heating system was monitored. Regeneration of the adsorption bed was performed for twenty minutes. This experiment was repeated three times. The results can be taken from Table 5. In this table, the cycle time is the total time for adsorption and regeneration. The capture efficiency is calculated as the ratio between the amount of moisture that is fed into the column and the amount of water that is captured by the absorbance. The calculated price per litre shows there are reasonable prices for water captured from air using this methodology.

(224) TABLE-US-00005 TABLE 5 Results water capture experiments Yield Energy Energy Price Cycle [L Capture Con- Conversion per Cycle Time kg.sup.−1 Efficiency sumption Efficiency Litre* No [min] day.sup.−1] [%] [kWh/L] [%] [$/L] 1 28 4.1 57.3 12.8 98.3 3.5 2 28 4.6 64.4 10.4 106.7 2.9 3 28 4.1 57.3 13.0 96.4 3.6 *Excluding capital costs

3. Water Analysis

(225) An IPC analysis was conducted on a comparative Milli-Q water sample and a Milli-Q water sample mixed with water captured using the inventive method from cycle 1 (Table 5) in with a dilution ratio of 1:15 of inventive water to Milli-Q. Water collected from cycle 1 was analysed to test the water for its suitability as potable water. The sample was diluted with ultrapure water with a dilution rate of 1:15. The water sample was analysed for cations (Ca.sup.+, K.sup.+, Mg.sup.+, Na.sup.+, S.sup.+) and metals (Al, As, B, Cd, Co, Cr, Cu, Fe, Mn, Mo, Ni, P, Pb, Sb, Se, Si, Sr, Zn) using inductively coupled plasma mass spectrometry. Additionally, ion chromatography was performed to analyse the water for anions (F.sup.−, Cl.sup.−, Br.sup.−, NO.sub.3.sup.−, SO.sub.4.sup.−).

(226) IPC analysis of both samples followed testing standards as follows: Fluoride, bromide, sulfate [APHA method 4110]. These common anions are determined by ion chromatography using a Dionex ICS-2500 system with 2 mm AS19 anion separation column and potassium hydroxide eluent generated on line, followed by conductivity detection after chemical suppression. With a flow rate of 0.25 mL per minute the anions F.sup.−, Cl.sup.−, Br.sup.−, NO.sub.3.sup.− and SO.sub.4.sup.2− are eluted between 3.5 and 25 minutes. Each ion concentration is calculated from peak areas using a 25 μL injection and compared to calibration graphs generated from a set of mixed standards with a range of concentrations. Cations and metals [APHA method 3120]. A range of elements are determined by Inductively Coupled Plasma Optical Emission Spectroscopy (ICPOES). The sample is nebulised into the plasma of an Agilent 5100 ICPOES. The emission spectra of the elements of interest are measured simultaneously. This determines the major cations (Ca, K, Mg and Na) along with trace elements (Al, B, Cu, Fe, Mn, Sr and Zn) and the non-metallic elements P, S and Si.

(227) The results of the IPC analysis are provided in Tables 6A and 6B.

(228) TABLE-US-00006 TABLE 6A Results of IPC Analysis of water samples part 1 |----Ion Chromatography---| ICP Majors F.sup.— Cl.sup.— Br.sup.— NO.sub.3.sup.— SO.sub.4.sup.═ Ca K Mg Na S Sample # mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L 1 MIlli-Q <0.05 <0.05 <0.05 1.6 <0.05 <0.1 <0.2 <0.1 <0.2 <0.2 Water 2 Sample + <0.05 <0.05 <0.05 4.1 <0.05 <0.1 <0.2 <0.1 <0.2 <0.2 MIlli-Q Water

(229) TABLE-US-00007 TABLE 6B Results of IPC Analysis of water samples—part 2 ICP Minors Al As B Cd Co Cr Cu Fe Mn Mo Ni P Pb Sb Se Si Sr Zn # mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L 1 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.1 <0.05 <0.05 <0.05 <0.2 <0.05 <0.1 <0.05 <0.2 <0.05 <0.05 2 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.1 <0.05 <0.05 <0.05 <0.2 <0.05 <0.1 <0.05 <0.2 <0.05 <0.05

(230) The results indicate that sample water produced using the method and apparatus of the present invention have a similar content to Milli-Q water, i.e. ultrapure water as defined by a number of authorities such as ISO 3696. Thus, apart from nitrate concentrations (NO.sub.3.sup.−), the concentration of all compounds in each sample (reference and inventive cycle 1) is below the detection limit.

(231) The only significant difference is the nitrate concentrations (NO.sub.3.sup.−), concentrations. The concentration of nitrate in the water sample collected from cycle 1 is about 60.8 mg/L. In the “Guidelines for drinking-water quality” the World Health Organisation (WHO) has restricted nitrate concentration in potable water to 50 mg/L. The concentration of nitrate in the control/reference ultrapure water sample is also elevated being 1.6 mg/L. The concentration of NO.sub.3.sup.− in ultrapure water type I however should be lower than 0.2 mg/L according to ISO 3696. It is thought that the abnormal nitrate concentrations of both water samples may be the result of contamination either during sample preparation or during sample analysis.

Example 2—Binders

4. Comparative Example—Binders

(232) The following provides a comparative example of the water adsorption properties of a water adsorption body/pellet formed using a hydrophobic binder. The inventors have surprisingly found that a hydrophilic binder must be used to impart optimal water adsorption properties to the shaped water adsorbent composite bodies. Non-hydrophilic binders such as hydrophobic binders deliriously affect the water adsorption properties of the shaped water adsorbent composite bodies compared to pellets formed using hydrophobic binders.

(233) A study was conducted on the effect on adsorption properties of Aluminium Fumarate pellets using different binders in aluminium fumarate pellet preparation.

(234) Pellets were prepared following the methodology set out in section 1.1.2. However, the binder composition was varied between two batches of pellets. A first batch of pellets was made using the first batch (batch I discussed above) of AlFu (designated Aluminium Fumarate (I)) and a cellulose siloxane binder, which is a hydrophobic binder. A second batch of pellets was made using the second batch (batch II discussed above) of AlFu (designated Aluminium Fumarate (II)) and a hydroxypropyl cellulose binder, which is a hydrophilic binder. The water uptake capacity of each batch of pellets was determined following the methodology herein outlined.

(235) The results of the water uptake capacity determination are provided in FIG. 25. A comparison of each batch to the Water Uptake Capacity isotherm for the comprising Aluminium Fumarate batch—i.e. Aluminium Fumarate (I) and Aluminium Fumarate (II) is also shown. It is noted that the adsorption isotherms between these batched differed due to differences in the properties of the formed Aluminium Fumarate MOF.

(236) It is also noted from FIG. 25 that aluminium fumarate has a water capacity between 0.09 to 0.5 g of water per gram of MOF depending on the relative humidity. The typical heat of adsorption of aluminium fumarate for water is well known and ranges between 60 and 30 kJ/mol depending on the ambient humidity

(237) The water vapour uptake isotherms shown in FIG. 25 clearly indicate that using cellulose siloxane decreases the performance of the MOF. However, when using hydroxypropyl cellulose as a binder, there is no decrease win moisture uptake visible.

Example 3—Temperature Swing Water Harvesting

5. Experimental

(238) 5.1 Testing Rig

(239) FIG. 26 shows the testing rig 600 for the temperature swing water harvesting device 350 (as illustrated in FIGS. 1C to 1E), with the water harvesting device 350, power supplies and measurement equipment. The illustrated testing rig 600 comprises the previously described apparatus 300 (FIGS. 1C and 1D) equipped with the following measurement devices: Relative humidity data logger 682 “EL-USB-2-LCD+” (Lascar Electronics Ltd., Wiltshire (United Kingdom)) to measure relative humidity and air temperature within the container 630 during adsorption and desorption phase of each water harvesting cycle (WHC). Power meter 683 “Digital DC Watt Meter” (KickAss®, Australia) to measure the energy consumption of the peltier device 610 during each desorption phase. Temperature data logger 684 “TC-08” (Pico Technology Ltd., Cambridgeshire (United Kingdom)) to log the temperature data of all thermocouples 375 during each desorption phase.

(240) A regulated DC power supply 685 (0-16 V, 0-10 A) was used to power the peltier device 610, and a digital-control power supply (0-30 V, 0-3 A) 686 was used to power the fan 670. Furthermore, a laptop 687 was used to collect the data from the relative humidity data logger 682 and the temperature data logger 684. A precision bench scale 689 “EK-15KL” (A&D Co. Ltd., Tokyo (Japan)) was also used to weigh the water harvesting device 350 before and after each adsorption or desorption cycle. The difference was calculated to determine the amount of water adsorbed during the adsorption phase or desorbed during the desorption phase.

(241) 5.2 Water Harvesting Experiments

(242) Each water harvesting cycle was run with the same setup and using the following protocol:

(243) The adsorption phase of the ith water harvesting cycle commences using an assembled water harvesting device 350 and a desorbed MOF packed bed 355. All thermocouples 375 were disconnected from the data logger 684 and the power supplies 685, 686 were disconnected from the peltier device 310 and the fan 370. The lid 334 of the container 330 was removed and the relative humidity data logger 682 was removed from the container 330. The whole water harvesting device 350 was put on a scale 689 to determine the weight m.sub.des,i-1 with the desorbed MOF packed bed 355. After the RH data logger 682 was set up (sampling rate: 5 min) and started on the laptop 687, the fan 370 was connected to its power supply 686 and switched on to start the air flow through the heat sinks 320 and packed bed 355 and thus commencing the adsorption phase. The fan 370 was switched off and disconnected from its power supply 686 after a set adsorption time. The water harvesting device 350 was put on the scale 689 again to determine the weight m.sub.ads,i with a water adsorbed MOF packed bed 355. The adsorbed amount of water m.sub.water,ads can be calculated using:
m.sub.water,ads=m.sub.des,i-1−m.sub.ads,i  (4.1)
where i is the number of the present water harvesting cycle and i−1 is the number of the previous water harvesting cycle. Furthermore, the RH data logger 682 was synchronised with the laptop 687 and the data collected during adsorption was saved. The RH data logger 687 was prepared (sampling rate: 15 s) for the desorption phase.

(244) The thermocouples 375 were connected to their data logger 684, and the peltier device 310 and the fan 370 were connected to the power supplies 685, 686 to start the desorption phase. The RH data logger 682 was started and placed in the container 330. The lid 334 of the container 330 was placed over the container body 332 and was sealed in place with electrical tape. The power supply 685 of the peltier element 310 was switched on after starting the temperature data logger 684 on the laptop 687. The fan 370 was started either together with the peltier element 310 or after the MOF packed bed 355 was heated up to a set temperature, depending on the experiment.

(245) The peltier device 310 and fan 370 were switched off after a set desorption time. The lid 334 of the container 330 was removed from the container body 332 and the data from both data loggers 682, 684 was saved on the laptop 687. The energy consumption of the peltier device 310 was read off the power meter 683. The energy consumption of the fan W.sub.Fan was calculated using:
W.sub.Fan=I.sub.Fan×U.sub.Fan×t.sub.run  (4.2)

(246) where I.sub.Fan the operating current of the fan, U.sub.Fan is the operating voltage is of the fan 370, and t.sub.run is the runtime of the fan 370. Condensed water was collected with a syringe (not illustrated) from the base of the container 330. The water volume was measured and the water harvesting device 350 was put on the scale 689 after removing the power supplies 685, 686 and the thermocouples 375 were disconnected. The weight of the water harvesting device 350 after desorption m.sub.des,i was used to calculate the amount of desorbed water m.sub.water,des with:
m.sub.water,des=m.sub.ads,i−m.sub.des,i  (4.3)

(247) The water harvesting cycle was now completed and the water harvesting device 350 was ready to start the next adsorption phase.

(248) 5.3 Performance of Peltier-Heated Water Harvesting Device

(249) The aluminium fumarate pellets were characterised to evaluate the performance of the water harvesting device. The pellets were characterised again after the water harvesting cycles to show the usability of aluminium fumarate pellets for water harvesting purposes. Subsequently, 24 water harvesting cycles were run with the device to determine optimised operation conditions.

(250) 5.3.1 Characterisation of MOF Pellets

(251) The MOF pellets, containing 99 wt % aluminium fumarate and 1 wt % binder, was characterised using FTIR, PXRD, nitrogen sorption isotherms, water sorption isotherms, and calculated BET surface area.

(252) Infrared spectroscopy was used to investigate changes in the aluminium fumarate due to mixing with binder and solvent during the pelletisation process. FIG. 27 shows the FTIR patterns of aluminium fumarate with 1 wt % hydroxypropyl cellulose (HPC) binder. Three extrusion batches were made to produce 200 g of pellets. Aluminium fumarate for all three batches was prepared as discussed in Example section 1.1.1. The pellets were prepared following a similar procedure outlined in Example section 1.1.2, though in this case the composition of the pellets was formulated using aluminium fumarate with 1 wt % hydroxypropyl cellulose (HPC) binder with no magnetic nanoparticle content.

(253) The first batch of pellets, designated “Pellets_01” were mixed with water and ethanol as solvents to get the required consistency of the paste for extrusion. However, this solvent formulation leads to a paste strand that dried too slowly for desired cutting behaviour in the pelletisation process (as described in section 1.1.2). The subsequent pellet batches, Pellets_02 and Pellets_03, were made with pure ethanol as solvent. This results in a quicker drying of the paste during the pelletisation. Furthermore, as FIG. 27 shows, the use of the different solvents results in different FTIR patterns. The pellets made with water and ethanol showed a stronger peak at wave numbers 3400 cm.sup.−1 and 1150 cm.sup.−1 and a weaker peak at 980 cm.sup.−1 compared to the other batches made with pure ethanol as solvent.

(254) FIG. 27 also compares the FTIR pattern of Pellets_01, Pellets_02 and Pellets_03 to pristine aluminium fumarate. The pattern of Pellets_01 was in good accordance with the pattern of pristine aluminium fumarate with both showing a strong and wide peak around a wave number of 3400 cm.sup.−1. The pristine MOF was not activated prior to the infrared spectroscopy. The pellets made with just ethanol as solvent show a slightly different FTIR pattern compared to the pristine MOF. The pellets were dried over night at 100° C. The pristine MOF and the pellets with water and ethanol as solvent exhibit strong peaks at 3400 cm.sup.−1 and 1150 cm.sup.−1 as well as a weaker peak around 980 cm.sup.−1. This difference may be attributable to water adsorbed to the MOF.

(255) A powder X-ray diffraction analysis was run on the samples, to evaluate the crystallinity of the produced aluminium fumarate pellets. The pattern was compared to the pristine aluminium fumarate as well as a simulated PXRD pattern. FIG. 28 shows the PXRD patterns. The strong peaks in the patterns of all three pellet batches indicate the crystallinity of the material. The three PXRD patterns were very similar, confirming that the difference in the FTIR pattern was due to the adsorbed water in one of the pellet batches. The PXRD pattern of the pellets also matches the pristine MOF well. Similarly, the simulated pattern matches the other patterns satisfactorily.

(256) Samples from all extrusion batches were analysed with a nitrogen uptake measurement, to characterise the aluminium fumarate pellets in regard to their adsorption characteristics. The BET surface area was determined as a quantitative value to compare the adsorption capability. Table 5.4 shows the BET and Langmuir surface areas of all three pellet extrusion batches and the surface areas of the pristine aluminium fumarate.

(257) TABLE-US-00008 TABLE 5.4 BET and Langmuir surface areas calculated from nitrogen sorption isotherms of pristine aluminium fumarate and aluminium fumarate pellets. BET surface area Langmuir surface area Sample (m.sup.2 g.sup.−1) (m.sup.2 g.sup.−1) Pristine aluminium fumarate 884 1064 Pellets_01 805 1042 Pellets_02 817 1061 Pellets_03 824 1071

(258) The surface areas of the pellets were 9%, 8% and 7% lower for Pellets_01, Pellets_02 and Pellets_03, respectively. This is likely a result of the added binder and the processing with solvents during the pelletisation process. Furthermore, the MOF was packed in a rigid shape and not in a powder like the pristine aluminium fumarate. The pellets extruded with a mixture of ethanol and water show a slightly lower surface area. It was decided that pellet extrusion should use pure ethanol as solvent for enhanced pellet quality.

(259) Water isotherms were measured for the batch Pellets_02, to determine the water uptake capacity of this batch. FIG. 29 shows the water uptake isotherm of the aluminium fumarate pellets compared to a water uptake isotherm for pristine aluminium fumarate were reported in the literature by Teo et al. (2017). Experimental study of isotherms and kinetics for adsorption of water on Aluminium Fumarate. International Journal of Heat and Mass Transfer Volume 114, November 2017, Pages 621-627.

(260) Firstly looking at the experimental isotherm (noting that this isotherm only shows the adsorption phase as desorption was not measured due to the very long equilibrate phases during the isothermal desorption): Aluminium fumarate pellets follow a type-IV isotherm. A first increase of water uptake is shown in the relative pressure range of 0:01 to 0:03, followed by a plateau with a lower slope. A second steep increase was observed at relative pressures between 0:2 and 0:4, again followed by a plateau region. The water uptake at a relative pressure of 0:4 was around 0:3 g.sub.water=g.sub.MOF and a maximal water uptake of 0:34 g.sub.water=g.sub.MOF was observed at a relative pressure of 0:6.

(261) Now looking at the comparison to the Teo isotherm in FIG. 29, it can be seen that Teo's reported water uptake isotherms of aluminium fumarate were of type-IV as well and have the steep increases of water uptake in the same relative pressure range. The water uptake in low relative pressure ranges was lower in the data of Teo, compared to the pellets of this work. In high relative pressure ranges the water uptake reported by Teo was higher compared to the pellets of this work. This might be due to the rigid pellet form as well as the 1 wt % binder in the aluminium fumarate pellets. Thus, the isotherm of the pellets show water uptake in g per g of MOF pellets, containing 99 wt % aluminium fumarate.

(262) In conclusion, the produced aluminium fumarate pellets were of high crystallinity and structure comparable to simulated data of X-ray diffraction analysis. The use of ethanol as solvent during the pelletisation process provided the best surface area results. Furthermore, the produced pellets were of still a good quality in regards to water uptake behaviour with a maximum water uptake of 0:34 g/g.

(263) 5.3.2 Water Harvesting Experiments

(264) Water harvesting cycles with the water harvesting device were run in the testing rig. For each cycle, containing an adsorption and a desorption phase the following date was recorded: the relative humidity during adsorption and desorption; the temperature in the MOF bed during desorption; the energy consumption of the peltier device and the fan; and the desorbed amount of water.

(265) Twenty four water harvesting cycles were run with the device. The goal of the experiments was to select a suitable peltier device and determine an optimal temperature range for desorption of the MOF bed. Once the best peltier device was determined the operating parameters were optimised with respect to energy consumption and water yield per day. All experiments were run with ambient air in March 2019 in Melbourne (Clayton), Australia. Table 5.1 shows all water harvesting cycles.

(266) TABLE-US-00009 TABLE 5.1 List of all water harvesting cycles with operating parameters (Fan parameter: HF - high flow rate [I = 0:09 A], LF - low flow rate [I = 0:02 A], ‘*’ fan switched on during heating and condensation, in all other experiments the fan was just switched on during condensation). Peltier Heating Condensation Condensation No. device current (A) current (A) time (h) Fan  1  29 W 3.0 3.0 3.0 (incl. heating) —  2  29 W 3.5 3.5 3.0 (incl. heating) —  3  29 W 4.0 4.0 3.0 (incl. heating) —  4  29 W 4.0 4.0 3.0 (incl. heating)  HF*  5  29 W 4.0 4.0 2.0 (incl. heating) —  6  29 W 4,0 4.0 1.5 (incl. heating) —  7  29 W 4.0 4.0 1.0 (incl. heating) —  8 110 W 5.0 4.5 2.0 —  9 110 W 5.5 4.5 2.0 — 10 110 W 6.0 4.5 2.0 — 11 110 W 6.5 4.5 2.0 — 12 110 W 6.5 6.0 2.0  LF* 13 110 W 6.5 6.5 2.0 HF 14 110 W 6.5 6.5 . . . 5.5 1.5 LF 15 110 W 6.5 6.5 . . . 6.0 1.0 LF 16 110 W 6.5 6.5 . . . 6.0 0.5 LF 17 110 W 6.5 6.5 0.083 LF 18 110 W 6.5 6.5 . . . 5.5 0.5 LF 19 110 W 6.5 6.5 . . . 5.5 0.5 LF 20 110 W 6.5 6.5 . . . 4.0 0.5 LF 21 110 W 6.5 6.5 . . . 5.5 0.5 LF 22 110 W 6.5 6.5 . . . 5.5 0.5 LF 23 110 W 6.5 6.0 0.5 LF 24 110 W 6.5 6.5 0.5 LF

(267) The first tested peltier device (Adaptive [53] AP2-162-1420-1118 Max Current 7:8 A) had a maximal temperature difference of 95° C. and a maximal heat flow of 29.3 W. The second tested peltier device (Multicomp [54] MCTE1-12712L-S Max Current of 12:0 A) had a maximal temperature difference of 68° C. and a maximal heat flow of 110 W. The tested electrical currents for the peltier device in water harvesting cycle 1 to 3 and 8 to 11 were selected by heating experiments, run prior to the water harvesting cycles.

(268) The heating experiments were run with the peltier device and an unloaded heat sink to determine the maximal temperatures in the heat sink in the current range of the peltier devices. Based on this data the initially tested currents with the loaded heat sink were elected. Further experiments were run with currents based on the previous water harvesting cycles. All experiments were evaluated with respect to the space time yield (STY) of the present operating parameters and the energy consumption of the device per kg of harvested water. Table 5.2 shows the results of the experiments.

(269) TABLE-US-00010 TABLE 5.2 Evaluation of all water harvesting cycles. Cycle times marked with ‘*’ were calculated with a theoretical adsorption time, calculated from Adsorption_14. Water Water Cycle Specific desorbed collected time STY energy No. (g) (mL) (hh:mm) (L kg.sup.−1 d.sup.−1) (kW h kg.sup.−1)  1 9.3 4.1 20:05 0.025 21.95  2 32.4 17.5 20:20 0.106 7.43  3 44.8 28.1 25:30 0.170 5.34  4 11.5 7.9 *5:00 0.192 18.34  5 34.9 21.9 *4:00 0.664 4.67  6 22.5 12.3  3:30 0.426 5.87  7 13.1 4.9 *3:00 0.198 8.60  8 35.2 29.1 *5:00 0.705 4.20  9 33.4 27.6  4:42 0.712 4.43 10 32.0 24.5 *4:30 0.660 4.58 11 26.9 23.1  4:21 0.664 4.42 12 57.9 51.7 *7:31 0.833 3.06 13 49.2 36.7 *5:24 0.824 5.61 14 52.7 42.8 *5:32 0.938 3.37 15 44.8 37.2 *4:31 0.998 3.05 16 32.3 26.3 *3:08 1.018 2.75 17 19.1 11.2 *1:37 0.838 3.67 18 17.9 10.7 *1:51 0.701 3.85 19 8.8 2.9 *1.21 0.260 7.12 20 3.0 0.6 *1:21 0.054 17.77 21 10.2 2.8  2:50 0.120 18.67 22 11.3 6.6  2:55 0.274 7.92 23 27.4 21.6 *3:09 0.833 3.35 24 30.7 25.2  3:04 0.998 3.27

(270) As the reported, desorption temperature of aluminium fumarate was 110° C., the peltier device with the greatest temperature difference was tested in the first place. The device was tested with three different currents in WHC 1 to 3. The temperature in the MOF bed during desorption was up to 65° C., 70° C. and 80° C. in WHC 1, 2 and 3, respectively.

(271) With increasing temperature in the MOF bed, the amount of desorbed water increases as well. The space time yield in the first three cycles was calculated with the actual adsorption time prior to the desorption phase. The adsorption was carried out overnight. Thus, the cycle times were around 24 h. The peltier device was switched on for three hours for the desorption phase in these experiments after the adsorption was completed. The water was collected afterwards. The highest STY of 0:170 L kg.sup.−1 d.sup.−1 and lowest specific energy consumption of 5:34 kW h kg.sup.−1 were measured in WHC 3.

(272) WHC 4 was run with the same current as WHC 3. The influence of the fan during the desorption phase was tested in this cycle. The fan was switched on simultaneously with the peltier device. A high air convection was created in the container 630. Hence, the MOF bed heats up more slowly during the desorption phase compared to the third cycle. As a result the highest temperature in the MOF bed after three hours was just 50° C. and the resulting amount of harvested water was 7:9 mL, significantly lower than the 28:1 mL in WHC 3. The maximal dissipated heat of the 29.3 W peltier device was not sufficient to heat up the MOF bed with a high convection. The subsequent experiments with this device were therefore carried out without a fan in the container 630 during desorption.

(273) Beginning with WHC 4, the space time yield was calculated with a theoretical adsorption time. To calculate a comparable cycle time for all experiments, the adsorption behaviour was logged for one adsorption phase. The weight of the device was logged during the adsorption of WHC 14. The adsorption starts with a MOF bed temperature of 60° C. to take the cooling from the previous WHC into account. Thus, a water adsorption curve over time was created as shown in FIG. 30. The adsorption time could then be calculated from the desorbed amount of water.

(274) In WHC 5 to 7 the runtime of the desorption phase was shortened. Consequently, less water needed to be adsorbed allowing more cycles to be run per day, increasing the STY. However, the results reveal a decline in the STY with shorter desorption phases. This might be a result of the fact that a majority of the time was used to heat up the MOF bed to elevated temperatures and that with shorter desorption phases the maximal reachable temperature decreases. For example, the maximal temperature in WHC 7 with just 1 h desorption time was 65° C. The highest space time yield of 0:664 L kg.sup.−1 d.sup.−1 was thus reached with a desorption time of 2 h and a corresponding adsorption time of 2 h. Furthermore, the specific energy consumption per harvested litre of water was greater even though the runtime of the peltier device was shorter as the amount of harvested water was significantly lower.

(275) The peltier device was resultantly changed to the 110 W peltier device for the subsequent experiments, which was found to be sufficient to heat the MOF bed to a temperature a temperature of around 70° C. At 70° C. in WHC 3, 82% of the adsorbed water is desorbed after 3 h. Again the first experiments were run with different currents of the peltier device between 5:0 A and 6:5 A. The MOF bed was heated up with this current. As soon as the MOF bed reached a temperature of 68° C. the current was lowered to maintain the temperature. A new time parameter designated ‘condensation time’ was used which correlated to the time after which the MOF bed had reached the temperature of 68° C. The current to maintain the temperature during the condensation time was set to 4:5 A in WHC 8 to 11, based on the data from the heating experiments.

(276) As expected, the highest current of 6:5 A lead to the shortest heating time for the MOF and thus to the highest space time yield of harvested water. At the same time the specific energy consumption was lowest at this current. The subsequent experiments were carried out with a heating current of 6:5 A.

(277) In the subsequent two experiments the influence of the fan, mounted between the fins of the heat sink, was investigated with the 110 W peltier device. As the change in the temperature in the MOF bed was very high in the first experiment with the fan (WHC 4), two different flow rates of the fan were tested: A high flow rate with a fan current of 0:09 A and a low flow rate with a fan current of 0:02 A. Besides, the temperature to change to condensing current was set to 75° C. to enhance the amount of desorbed water. The amount of harvested water in WHC 12 with the low flow rate was 51:7 mL, the highest amount so far. As a result, this experiment has the highest space time yield and lowest energy consumption thus far with 0:833 L kg.sup.−1 d.sup.−1 and 3:06 kW h kg.sup.−1, respectively. The amount of harvested water in WHC 13 with the high flow rate was just 36:7 mL, even though the fan was switched on after the MOF bed was heated up. In the previous experiment the fan was running the whole time, including during the heating phase. Resultantly, the subsequent experiments were carried out using the fan with a low flow rate. Furthermore, the fan was switched on when the MOF bed reaches the final desorption temperature, as the initial heating rate was 2:73° C. min.sup.−1 compared to 2:13° C. min.sup.−1 when the fan was switched on during the heating phase. The higher initial heating rate leads to shorter desorption times and thus to shorter cycle times.

(278) Based on these thirteen experiments, the 110 W peltier device was used for the further optimisation of the water harvesting device using the following parameters for water harvesting cycle 14 to 24: Peltier device MCTE1-12712L-2, 110 W. Heating current 6:5 A. Fan flow rate Low (I=0:02 A). Fan run time Start after MOF bed heated up.
5.3.3 Optimisation of Operating Parameters

(279) In the next experiments the condensation time; and desorption temperature were optimised. For this the condensation time was varied between 2 h and 5 min with a desorption temperature of 75° C. in WHC 14 to 17, respectively. Subsequently the desorption temperature was varied between 75° C. and 45° C. with the best desorption time, determined in the previous experiments, in WHC 18 to 20, respectively.

(280) TABLE-US-00011 TABLE 5.3 Optimisation of the condensation time of the water harvesting device. Collected water over different condensation times. Condensation time Water collected WHC (min) (mL) 12 120 51.7 14  90 42.8 15  60 37.2 16  30 26.3 17  5 11.2

(281) Table 5.3 and FIG. 31 show the collected water, space time yield and specific energy consumption of the device over different condensation times. As expected, the amount of water collected was greater for higher condensation times. However, the space time yield was better in a range of low condensation times with a maximum at 30 min. This was due to the significantly lower adsorption time that was necessary to adsorb the amount of water which was desorbed during the desorption phase. Only for very short desorption times of 5 min, the space time yield was lower as the time was not long enough to desorb a significant amount of water from the MOF bed. The specific energy consumption also had a local minimum at a condensation time of 30 min. Thus the optimal condensation time for the water harvesting device was determined to be 30 min. Based on this condensation time the optimal desorption temperature was determined in the subsequent experiments.

(282) TABLE-US-00012 TABLE 5.4 Optimisation of the desorption temperature of the water harvesting device. Collected water over different desorption temperatures. WHC Desorption Temp (° C.) Water collected (mL) 16 75 26.3 18 65 10.7 19 55 2.9 20 45 0.6

(283) Table 5.4 and FIG. 32 show the collected water, space time yield and specific energy consumption of the device over different desorption temperatures. All three parameters were greater with higher desorption temperatures. The most important factor was the specific energy consumption as depicted in FIG. 32. The energy consumption was greater for lower desorption temperatures as the energy was used to heat up the MOF and the heat sink but not for the desorption of water. In conclusion, the best solution in terms of energy efficiency and space time yield was to heat the device to 75° C. before switching on the fan and condense for 30 min.

(284) Water harvesting cycle 16 was the experiment with the highest space time yield and the lowest specific energy consumption. The logged data of this cycle during adsorption and desorption is shown in FIGS. 33, 34 and 35.

(285) The adsorption phase, shown in FIG. 33, was under isothermal conditions with a varying relative humidity between 50% and 70%. The average loading of water in the air was 11:11 gm.sup.−3. The changes in relative humidity were caused by the air conditioning system in the lab and the weather outside. The high temperature in the beginning was caused by the hot MOF bed and heat sink from the previous water harvesting cycle. The data logger was placed next to the heat sink, thus the air in the beginning was hotter than the ambient air.

(286) FIG. 34 shows the temperatures in the water harvesting device during the desorption phase. The MOF bed was heated up to 75° C. before the fan was switched on after 33 min. The temperature in the MOF bed decreases after the fan was switched on, due to the higher convection in the container. After the first decrease, the temperature increases slowly and the current of the peltier device was changed to 6:0 A after 53 min to maintain a temperature of slightly above 70° C. The change in the current was shown by the bend in the temperature curve of the MOF bed and the ΔT. The ΔT plot shows the temperature difference between the MOF bed and the condenser. This difference changes due to the higher convection when the fan was switched on and due to the current change after 53 min. The air temperature in the container increased at nearly a constant rate until the fan was switched on. The higher convection also results in a higher air temperature as the heat was transported from the MOF bed into the air in the container.

(287) FIG. 35 shows the temperature of the condenser as well as the relative humidity and dew point in the container. The increase of relative humidity was caused by desorption of water from the MOF bed. The relative humidity decreases afterwards as liquid water condenses, and the amount of new desorbing water decreases as well. The peak in the relative humidity graph might be a result of a droplet on the humidity meter probe and was thus ignored in the discussion. This graph shows that the condenser temperature was always lower than the dew point in the system. Hence water vapour will condense on the condenser and can be collected afterwards. It was also observed that the walls of the device act as a second condenser, as a lot of droplets were forming on the walls during the experiments.

(288) With only the adsorption changed, the experiment was repeated under the same conditions in WHC 24 to show the reproducibility of the results of WHC 16. The adsorption phase was started with a desorbed and hot MOF bed from the previous experiment. The duration of the adsorption phase was set to the length of the theoretical adsorption time used to calculate the space time yield of WHC 16. Hence the adsorption time in WHC 24 was 2:0 h and the desorption time was 1:06 h, resulting in a real cycle time of 3:06 h.

(289) FIG. 36 shows the adsorption phase of WHC 24. Compared to WHC 16 the water loading in the air was lower with just 9:27 gm.sup.−3. 25:2 mL of water was collected after the desorption phase. This resulted in a space time yield of 0:998 L kg.sup.−1 d.sup.−1. Compared to WHC 16 the space time yield deviates by 2%. This demonstrates the reproducibility of the experiments with high space time yields and confirms the calculation of the theoretical adsorption time, used to calculate the space time yield in most of the experiments.

(290) As shown in FIGS. 37 and 38 the temperature graph of the MOF bed has a good match with the graph of WHC 16. The temperature after the fan was switched on was increasingly slower compared to WHC 16. Due to this the current was not lowered in WHC 24 to maintain temperatures slightly above 70° C. Additionally, ΔT was higher in WHC 24 than in WHC 16. This was due to the replacement of the heat grease between the heat sink and the peltier device after WHC 20 affecting the thermal conductivity between the peltier device and the heat sink.

(291) Furthermore, two water harvesting cycles, WHC 21 and 22, were run on the same dry day (Aug. 3, 2019, Melbourne (AUS)) under very dry conditions during the adsorption phase. The relative humidity was between 25% and 30% with a temperature of slightly about 20° C. The average water loading in the air was 5:50 gm.sup.−3. FIG. 39 shows the adsorption phase of WHC 22. The experiments of WHC 21 and WHC 22 showed that the low relative humidity leads to much longer adsorption times to adsorb the same amount of water. As the adsorption time was constant, the amount of water adsorbed was lower compared to a higher relative humidity. As a consequence the amount of harvested water in WHC 21 and 22 was lower than in the previous cycles, with a space time yield of 0:120 L kg.sup.−1 d.sup.−1 and 0:274 L kg.sup.−1 d.sup.−1 for WHC 21 and 22, respectively. This was a deviation of 88% and 73%, respectively. Nevertheless, these results demonstrate that the water harvesting device stills works in very dry and desert-like conditions.

(292) Finally, FIG. 40 provides two views of a prototype water capture apparatus 800 that uses the temperature swing water harvesting apparatus 300 shown in FIGS. 1C to 1E. FIG. 40(A) illustrates the external housing 802 including the water dispensing outlet 805 activated by control panel 808; and FIG. 40(B) illustrates the inner components, which essentially show the fan 810 and louver system 820 for creating a flow of atmospheric air into and over the water adsorbent, i.e. fan is operated, and pivots open the louvers of the louver system 820 allowing air to flow over and through the packed bed of MOF adsorbent (not shown in FIG. 40) inside the apparatus packed into the heat sink (not shown in FIG. 40), and when the fan is inactive, the louvers of the louver system 820 pivot closed to create a closed environment, allowing the desorption phase and condensation processes of the water harvesting cycle to take place in a closed/sealed environment.

(293) Comparison Between Devices

(294) Table 5.8 below provides a comparison between the water harvesting device as developed in this work in accordance with embodiments of the present invention and to Yaghi's MOF based water harvesting devices as described in the background of the invention section. STY.sub.device is the space time yield with regard to the device's volume. X.sub.min provides a measure of the environmental conditions, i.e. the humidity (minimum water content) of the air fed over the MOF adsorbent.

(295) TABLE-US-00013 TABLE 5.8 Comparison between the water harvesting devices developed in this work to other MOF based water harvesting devices. Energy STY mass Output consumption X.sub.min device of Device (L/day) (kWh/L) (g m.sup.−3) (L/m.sup.3/d) MOF Yaghi Prototype* 0.078 sunlight 4.6 1.77 825 Inventive Induction 0.23 10.4 9.7 ~0.05 28 device (Example 1) Inventive Peltier device 0.202 2.75 9.3 4.59 ~200 (high RH) (Example 3) Inventive Peltier device 0.054 7.92 5.5 1.23 ~200 (low RH) (Example 3) *F. Fathieh, M. J. Kalmutzki, E. A. Kapustin, P. J. Waller, J. Yang, and O. M. Yaghi. “Practical water production from desert air”. In: Science Advances 4.6 (2018).

(296) The comparison indicate that both the tested embodiments of induction and Peltier device water capture apparatus of the present invention have a better water output compared to the Yaghi devices.

(297) Those skilled in the art will appreciate that the invention described herein is susceptible to variations and modifications other than those specifically described. It is understood that the invention includes all such variations and modifications which fall within the spirit and scope of the present invention.

(298) Where the terms “comprise”, “comprises”, “comprised” or “comprising” are used in this specification (including the claims) they are to be interpreted as specifying the presence of the stated features, integers, steps or components, but not precluding the presence of one or more other feature, integer, step, component or group thereof.