A redox-active modified metal-organic framework

Abstract

A redox-active modified metal-organic framework There is provided an electrode comprising a substrate coated with a redox-active modified metal-organic framework (MOF), wherein the redox-active modified MOF is a ferrocene-functionalised MOF. There is also provided an electrosorption unit comprising the electrode and a method of separating at least one target molecule from a fluid source using the electrode.

Description

[0001] A redox-active modified metal-organic framework

[0002] Technical Field

[0003] The present invention relates to as redox-active modified metal organic framework (MOF).

[0004] Background

[0005] Precise and rapid separation of inorganic ions is a technical endeavour sought after in processes ranging from the production of synthetics and pharmaceuticals to water desalination and remediation. The specifics of ion separation depend greatly on the type of industry and the stage of process flow. For example, in environmental remediation, contamination of oxyanions such as NOs", SC2', PCU3-, chromate (CrC2-), arsenate (HAsC2-) and / or selenate (SeC2-) in groundwater can cause ecological damage and threaten public health. Industries are therefore required to meet stringent regulatory limits on wastewater disposal.

[0006] Current separation techniques focus on the removal of target ions or by excluding unwanted ions. These techniques include precipitation or flocculation-coagulation, which generally require the use of a precipitating agent to induce precipitation of target ionic species from solution before conventional filtration methods being used for separating the solid from liquid. However, such techniques result in the addition to cost due to additional precipitation agents, or further treatment of sludge produced.

[0007] Another common technique is the use of inexpensive and porous adsorbents such as activated carbon and zeolites. Adsorbents, such as metal organic frameworks (MOFs), may be tuned for preferential ion adsorption. However, MOFs have poor electrical conductivity and water stability, making them unsuitable for ion-selective electrochemical separation. While these shortcomings of MOFs can be mitigated by carbonising the MOFs through pyrolysis to obtain MOF-derived carbon products, pyrolysis causes concurrent precipitation of metals and partial collapse of pores within the MOF structure. The unique local chemical environment afforded by nodes and ligands of the MOF may therefore be destroyed.

[0008] There is therefore a need for an improved MOF material. Summary of the invention

[0009] The present invention seeks to address these problems, and / or to provide an improved electrode material for ion-selective electrochemical separation.

[0010] According to a first aspect, there is provided an electrode comprising a substrate coated with a redox-active modified metal-organic framework (MOF), wherein the redox-active modified MOF is a ferrocene-functionalised MOF. In particular, the electrode may be an anode.

[0011] The redox-active modified MOF may comprise any suitable MOF. In particular, the redoxactive modified MOF may be stable in water. According to a particular aspect, the ferrocene-functionalised MOF may be ferrocene-functionalised UiO-66 (Fc-UiO-66).

[0012] The substrate comprised in the electrode may be any suitable material. For example, the substrate may be a conductive material.

[0013] The electrode may be used in any suitable application. According to a particular aspect, the electron may be comprised in an electrosorption unit.

[0014] According to a second aspect, there is provided an electrosorption unit cell comprising the electrode according to the first aspect. In particular, the electrode may be an anode. The electrosorption unit may further comprises a cathode. In particular, the anode and the cathode may be separated by a separator.

[0015] According to a third aspect, there is provided a method of separating at least one target molecule from a fluid source, the method comprising: placing in a fluid a first electrode and a second electrode, wherein the first electrode is an electrode according to the first aspect; and applying an electrical potential across the first electrode and the second electrode, such that the first electrode transforms to an oxidised state and selectively binds to a target electron-donating functional group of a target molecule present in the fluid source, such that the at least one target molecule comprising the target electron-donating functional group is separated from the fluid source. The second electrode may be any suitable electrode. In particular, the second electrode may selectively capture a cationic species present in the fluid source.

[0016] The method may further comprise reversing the applied electrical potential to release the bound target molecule from the first electrode and / or the captured cationic species from the second electrode.

[0017] The target molecule may comprise any suitable molecule. According to a particular aspect, the target molecule may comprise an oxyanion. In particular, the target molecule may comprise phosphate, sulphate and / or nitrate ions.

[0018] The fluid source may be any suitable fluid source. For example, the fluid source may be an aqueous fluid, an organic fluid, or a mixture thereof.

[0019] Brief Description of the Drawings

[0020] In order that the invention may be fully understood and readily put into practical effect there shall now be described by way of non-limitative example only exemplary embodiments, the description being with reference to the accompanying illustrative drawings. In the drawings:

[0021] Figure 1 shows a schematic diagram depicting an Fc-UiO-66 anode integrated into an electrosorption unit;

[0022] Figure 2 shows a schematic diagram depicting multiple electrosorption units integrated within an electrosorption module;

[0023] Figure 3 shows a flow chart of an example of how an electrosorption module operates;

[0024] Figure 4 shows SEM images of UiO-66 and Fc-UiO-66, a schematic representing UiO- 66 with missing linker defects, and a schematic representing Fc-UiO-66 with a ferrocene pendant;

[0025] Figure 5 shows the XRD patterns for both UiO-66 and Fc-UiO-66 and a comparison between the two with a simulated pattern (COD #4132636);

[0026] Figure 6 shows EXAFS data at the Zr K-edge for Fc-UiO-66 with reference to UiO-66 and pure ferrocene, respectively; Figure 7 shows EXAFS data at the Fe K-edge for Fc-UiO-66 with reference to UiO-66 and pure ferrocene, respectively;

[0027] Figure 8 shows electron paramagnetic resonance (EPR) spectra of Fc-UiO-66 with reference to UiO-66;

[0028] Figure 9 shows pore size distribution of Fc-UiO-66 with reference to UiO-66, with inset showing N2 adsorption-desorption isotherms;

[0029] Figure 10 shows CV curves obtained in a 0.1 M NaCI solution;

[0030] Figure 11 shows CV curves of Fc-UiO-66 in concentrations of 0.001 , 0.01 , and 0.1 M for NaCI, NaNOs, Na2SO4, and NaH2PO4 (experiments were performed using a 3-electrode setup by applying a scan voltage of 5 mV s-1across a voltage window of -0.2 to 1 .0 V);

[0031] Figure 12 shows normalized redox-mediated adsorption in dual-ion solutions of 1 mM NOs", SO42’, and PC3', with 10 mM Cl' as background, and calculated selectivities;

[0032] Figure 13 shows normalized redox-mediated adsorption in multi-ion solutions of 1mM NO3- and SO42-, NO3- and PC>43', and NOs', SO42; and PCU3-, with 10 mM Cl' as background, and calculated selectivities; and

[0033] Figure 14 shows the effect of electrolyte pH on stability and performance of the Fc-UiO- 66 anode.

[0034] Detailed Description

[0035] As explained above, there is a need for an improved MOF material.

[0036] In general terms, the present invention provides an improved redox-active MOF material with ion-selective properties. The redox-active MOF material may be used in separation applications, particularly for the rapid removal or concentration of target molecules to achieve desired solution compositions. In particular, the redox-active MOF material may be used in electrosorption processes to effectively and controllably remove targeted ionic contaminants. The redox-active MOF material described herein is water-stable while being redox-active, enabling it to be suitably used in a wide range of separation applications, particularly in detection and selective removal of target ions in wastewater.

[0037] According to a first aspect, there is provided an electrode comprising a substrate coated with a redox-active modified metal-organic framework (MOF). The redox-active modified MOF may comprise any suitable MOF and may be modified by any suitable ligand. In particular, the MOF may be, but not limited to, UiO-66, ZIF-8, ZIF-67, MIL-101 , MFM- 300(M) series, or a combination thereof. Even more in particular, the MOF may be UiO- 66.

[0038] UiO-66 is constructed from a unit of hexa-zirconium(IV)-oxo, hydroxo, aqua node connected to twelve 1,4-benzene dicarboxylate (BDC) ligands. UiO-66 possesses high specific surface areas averaging at -1200 m2g-1, triangular pore windows of -6 A and high stability over a range of pH values in aqueous solution. Even in an unmodified state, the uniform, rigid pores of UiO-66 provide moderately effective size-exclusion based separation in gas and liquid mixtures. In addition, unsaturated zirconium nodes endow it with highly catalytic Lewis acid sites which can also double as binding sites to select ions. The extent of binding specificity depends on the type and oxidation state of the metal node in conjunction with the spatial arrangement of ligands.

[0039] According to a particular aspect, the MOF may be modified by a metallocene. For example, the metallocene may be, but not limited to, ferrocene, ruthenocene, cobaltocene, nickelocene, alloys or combinations thereof. In particular, the metallocene may be ferrocene.

[0040] According to a particular aspect, the redox-active modified MOF may be a ferrocene- functionalised MOF. In particular, the redox-active modified MOF may be ferrocene- functionalised UiO-66 (Fc-UiO-66).

[0041] In particular, the redox-active modified MOF may be stable in water.

[0042] The redox-active modified MOF may be formed by any suitable method. According to a particular aspect, the method of forming a ferrocene-functionalised MOF may comprise: mixing a ferrocene-based compound in a solvent to form a mixture; adding a MOF to the mixture; and heating the mixture.

[0043] The ferrocene-based compound may be any suitable compound. For example, the ferrocene-based compound may be, but not limited to, ferrocenecarboxylic acid, ferrocenesulfonic acid, ferrocenylmethylamine, azidoferrocene, or a combination thereof. In particular, the ferrocene-based compound may be ferrocenecarboxylic acid. The solvent may be an organic solvent More particularly, the organic solvent may be a polar solvent. For example, the solvent may be, but not limited to, dimethylformamide (DMF), dimethyl sulfoxide (DMSO), N-methyl-2-pyrrolidone (NMP), or a mixture thereof. In particular, the solvent may be DMF.

[0044] The MOF may be any suitable MOF as described above. The adding a MOF may be by any suitable method. According to a particular aspect, the adding a MOF may be by sonication. In particular, the adding a MOF may be by ultrasonication.

[0045] The heating may be for a suitable period of time. For example, the heating may be for a period of 1-48 hours. In particular, the heating may be for 6-36 hours, 12-24 hours, 18- 20 hours. Even more in particular, the heating may be for about 24 hours. The heating may be by any suitable means. For example, the heating may be over a water bath, oilbath, or the like. The heating may be carried out at a suitable temperature. For example, the heating may be carried out for a temperature of < 100°C. In particular, the temperature may be 45-100°C, 50-90°C, 55-85°C, 60-80°C, 70-75°C. Even more in particular, the temperature may be 60-80°C.

[0046] The method may further comprise washing the mixture following the heating. The washing may comprise washing the mixture with a suitable solvent. For example, the solvent may be, but not limited to, DMF, acetone, dimethyl sulfoxide (DMSO), N-methyl- 2-pyrrolidone (NMP), or a mixture thereof.

[0047] The method may further comprise drying the mixture. The drying may be in an oven.

[0048] According to one embodiment, the method of forming a ferrocene-functionalise MOF may be as follows. Ferrocene may be incorporated using the solvent-assisted ligand incorporation (SALI) process in small batches. 48 mg of ferrocenecarboxylic acid may first be dissolved in 3 ml_ of DMF before 32 mg of activated UiO-66 is added. The pale, chalky mixture may be sonicated for 30 min and placed in an oil bath set to 60°C for 24 h. Finally, a burgundy-coloured mixture may be obtained and washed with DMF and acetone before drying in an oven. Fc-UiO-66 may be activated prior to use. For example, the activation may comprise any suitable method. In particular, the activation may comprise heating in a vacuum oven at 120°C for at least 48 h.

[0049] The method may further comprise forming the MOF prior to the forming of the ferrocene- functionalised MOF. According to a particular embodiment, the UiO-66 may be formed by first preparing a metal precursor solution by depositing 125 mg of ZrCl4 into a glass vial containing 5 ml_ of DMF followed by the addition of 1 ml_ of 37% (w / w) HCI. The ligand solution may be separately prepared by depositing 123 mg of terephthalic acid (H2BDC) in 10 mL of DMF. Both solutions may be sonicated for at least 30 min before the ligand solution is dropwise added to the metal precursor solution. The resulting mixture may then be placed in an oven at 80°C for 12 h. After the mixture has cooled, it may then be washed with D and acetone before drying in an oven at 60°C. UiO-66 may be activated by heating it in a vacuum oven at 120°C for at least 48 h.

[0050] The substrate comprised in the electrode may be any suitable material. For example, the substrate may be a conductive material.

[0051] The electrode may be formed by any suitable method. For example, the electrode may be formed by slurry and coating method. In particular, a slurry comprising the redoxactive modified MOF may be coated on a substrate surface.

[0052] The substrate may be any suitable substrate. For example, the substrate may be a conductive material. In particular, the substrate may comprise carbon. Even more in particular, the substrate may be carbon paper.

[0053] According to one embodiment, the electrode may be prepared by preparing a slurry by mixing the redox-active modified MOF, a binder and conductive additive in a solvent. The redox-active modified MOF, binder and conductive additive may be mixed in any suitable ratio. For example, the mass ratio may be, but not limited to, 8:1 :1.

[0054] The binder may be any suitable binder. For example, the binder may be, but not limited to, polyvinylidene fluoride (PVDF), carboxymethyl cellulose (CMC), polytetrafluoroethylene (PTFE), or a mixture thereof. In particular, the binder may be PVDF. Even more in particular, the PVDF may have a molecular weight of about 180 000.

[0055] The conductive additive may be, but not limited to, carbon nanotubes. In particular, the conductive additive may comprise multi-walled carbon nanotubes, single-walled carbon nanotubes, or a mixture thereof. The solvent may be any suitable solvent. For example, the solvent may be, but not limited to, N-methyl-2-pyrrolidone (NMP, 99.5%), dimethylformamide (DMF), dimethylacetamide (DMAc), or a mixture thereof.

[0056] The slurry may then be coated on a surface of a substrate. The substrate may be any conductive substrate, as described above. The slurry coated on the substrate surface may then be dried by suitable means. For example, the drying may comprise drying the electrode at a suitable temperature, such as, but not limited to 60-80°C. The drying may be for a suitable period of time. For example, the drying may be, but not limited to, overnight.

[0057] According to a particular aspect, the electrode may be an anode. The electrode may be used in any suitable application. According to a particular aspect, the electron may be comprised in an electrosorption unit. The electrosorption unit may be for electrochemical separation of target molecules from a feed source.

[0058] According to a second aspect, there is provided an electrosorption unit cell comprising the electrode according to the first aspect. In particular, the electrode may be an anode. The electrosorption unit may be for electrochemical separation. In particular, the electrosorption unit may be for the detection and selective removal of target molecules. The target molecules may comprise any suitable molecule. According to a particular aspect, the target molecule may comprise an oxyanion. In particular, the target molecule may comprise phosphate, sulphate and / or nitrate ions.

[0059] The electrosorption unit may further comprises a cathode. In particular, the anode and the cathode may be separated by a separator.

[0060] According to a particular embodiment, the electrosorption unit may be as shown in Figure 1. There is provided an electrosorption unit 1000 comprising a pair of outer casings (1001 , 1006), a cathode (1005), a cation exchange membrane (CEM)(1004), a redoxactive modified MOF anode (1002) and a spacer (1003) embedded with a flow channel. The electrosorption unit may additional comprise current collectors.

[0061] The outer casings (1001 , 1006), current collectors (not shown), anode 1002, cathode 1005 and CEM 1004 sandwich the spacer 1003 to form a continuous pathway for a fluid source. The electrosorption unit may be connected to a microcontroller for monitoring the change in current response when fluid source is passed through and controls the charge mode and amount of charge dispensed to the electrosorption unit 1000. In particular, the anode 1002 will contact the positive terminal of a power source and the cathode 1005 and CEM 1004 will contact the negative terminal. The cathode 1005 primarily serves as a counter electrode to the anode 1002 and may be typically composed of highly porous and conductive carbon materials such as, but not limited to, activated carbon, carbon nanotubes or mesoporous carbon. CEM 1004 only allows the passage of positively charged cations. The anode 1002 and cathode 1005 are arranged perpendicular to the spacer 1003 as shown in Figure 1. The outer casings (1001 , 1006) may be made of any suitable material. For example, the outer casings (1001, 1006) may be made of, but not limited to, acrylic or stainless steel.

[0062] Ion adsorption may occur when the anode 1002 and cathode 1005 are charged with cations adsorbed by the cathode 1005 and anions adsorbed by the anode 1002. Selective ion adsorption may be achieved by applying the appropriate oxidation voltage to the anode 1002.

[0063] Multiple electrosorption units may be integrated within an electrosorption module as shown in Figure 2. The anode and cathode of each unit may be connected together with the respective cathode current collector 1007 and anode current collector 1008. The arrangement laid out in Figure 2 shows a series arrangement where a fluid source enters from the left and exits from the right of the Figure 2. As explained below, ion selectivity may be the result of a positively charged pore environment and may be tuned by increasing or decreasing the concentration of dopant ferrocene in the redox-active modified MOF anode. By changing the anode material with MOF materials with different dopant concentrations, ion-selective sections 1009, 1010 and 1011 may be created. When a target molecule is present in the fluid source, this registers as a current transient. Multiple current transients may be registered in each of sections 1009, 1010 and 1011 as the fluid flows through to achieve ion detection. The adsorption of ions can also be tuned by changing the charging parameters of the electrosorption unit.

[0064] An example of how an electrosorption module of Figure 2 operates in practice is outlined in Figure 3. Allowed concentration limits for target oxyanions are inputted via a computer linked to the electrosorption module and these will be translated to charging parameters such as current value, voltage value and charging duration for the module. Fluid flows into the module and when ions contact the electrodes, a small current transient is detected and interpreted according to the section 1009, 1010, 1011 in contact. This is translated to a concentration value, c*. The difference between the concentration limit and the detected concentration is then determined by the computer and the charging parameters are then modified to optimised values. Charging begins and will end when the concentration is below the limit. Fluid may exit the device with concentration levels equal to or below the limits set.

[0065] Electrochemical separation techniques, particularly those of electrosorption, involve the removal of ions due to electrostatic attraction between the ions and the charged electrodes. The electrosorption unit and its method of use as described above may be entirely modular, thereby providing fast ion removal kinetics and electrodes can be regenerated over multiple charge-discharge cycles. Another advantage is the ease at which the electrosorption unit may be integrated into wider smart, decentralised networks. The electrosorption unit may be implemented as part of water quality monitoring networks in agriculture, manufacturing or mining industries, just to name a few.

[0066] According to a third aspect, there is provided a method of separating at least one target molecule from a fluid source, the method comprising: placing in a fluid a first electrode and a second electrode, wherein the first electrode is an electrode according to the first aspect; and applying an electrical potential across the first electrode and the second electrode, such that the first electrode transforms to an oxidised state and selectively binds to a target electron-donating functional group of a target molecule present in the fluid source, such that the at least one target molecule comprising the target electron-donating functional group is separated from the fluid source.

[0067] According to a particular aspect, the first electrode may be an anode. The second electrode may be any suitable electrode. In particular, the second electrode may be a cathode selectively capture a cationic species present in the fluid source.

[0068] The method may further comprise reversing the applied electrical potential to release the bound target molecule from the first electrode and / or the captured cationic species from the second electrode. The target molecule may comprise any suitable molecule. The target molecules may be contaminants in a fluid source. According to a particular aspect, the target molecule may comprise an oxyanion. In particular, the target molecule may comprise phosphate ions (PC>43'), sulphate ions (SCu2') and / or nitrate ions (NOs-)-

[0069] The fluid source may be any suitable fluid source. For example, the fluid source may be an aqueous fluid, an organic fluid, or a mixture thereof. In particular, the fluid source may include wastewater. The wastewater may be from any source, such as, but not limited to, agricultural, pharmaceutical, sewage and manufacturing of synthetics. The method described above provides a convenient way to regulate wastewater ion composition for regulatory compliance prior to discharging the wastewater.

[0070] The specificity of the results shown below is by no means limiting on the aspects described above. The results were obtained as follows.

[0071] Material characterisation

[0072] Scanning electron micrographs (SEMs) were taken using a field emission scanning electron microscope (FE-SEM, JEOL JSM-7600F), while transmission electron micrographs (TEMs) were obtained with a transmission electron microscope (TEM, FEI Talos F200X) equipped with an energy-dispersive X-ray spectrometer (EDS). Electron paramagnetic resonance (EPR) spectroscopy was performed using a Bruker EMXplus- 6 / 1 spectrometer with a 100 kHz modulation frequency.

[0073] Structural properties were investigated via powder X-ray diffraction spectroscopy (PXRD) on desolvated MOF samples in Bragg-Brentano geometry using a Bruker D8 Advance diffractometer equipped with a Ni filtered Cu Ka radiation (A = 1 .5406 A, 40 kV, and 40 mA) source and a 1 D LynxEye detector. Diffraction patterns were collected over a 20 range of 5-80° at increments of 0.02°. Rietveld refinement of the full spectral range was performed using Profex (Ver. 5.2.4) with instrument specific parameters.

[0074] X-ray absorption near edge structure (XANES) and extended X-ray absorption fine structure (EXAFS) spectra were obtained in transmission mode using Si(111) crystal monochromators at the BL11B beamline at the Shanghai Synchrotron Radiation Facility (SSRF, Shanghai, China). A four-channel silicon drift detector (SDD) Bruker 5040 was used to record the spectra at room temperature. All samples were prepared by first compressing the materials into thin sheets of ~1 cm and sealing with Kapton tape. Spectral data were subsequently processed and analyzed using Athena and Artemis from the Demeter software package.

[0075] Tests and results

[0076] Figure 4 shows the SEM images of both UIO-66 and Fc-UiO-66 with particle size approximately the same at -200 pm. The molecular structures of both materials were presented to show the position of the doped ferrocene unit in Fc-UiO-66.

[0077] Figure 5 shows the XRD patterns for both materials and a comparison between the two with a simulated pattern (COD #4132636), which shows no major differences except for an increase in peak intensity at 12° corresponding to the (220) plane for Fc-UiO-66. This was likely due to the occupancy of leftover solvent molecules and / or ferrocene pendants. The remarkable similarity in diffraction patterns indicates that crystalline structure was largely preserved after ferrocene doping.

[0078] Figures 6 and 7 shows EXAFS data at the Zr K-edge and Fe K-edge for Fc-UiO-66 in reference to UiO-66 and pure ferrocene, respectively. In Figure 6, two prominent features were detected at 1.6 A and 3.1 A with little deviation in radial distance. Using crystal data from the UiO-66 structure, first shell scattering was attributed to single scattering paths Zr-OP3 and Zr-0 whereas second shell scattering was largely due to backscattering from closest coordinated Zr neighbours. Spectra of Fc-UiO-66 notably differed from UiO-66 in two ways: sharper and more intense features at 1.6 A and a decrease in peak intensity at 3.1 A. Sharp features at 1.6 A indicates increased ligand coordination due to functionalisation with ferrocenecarboxylic acid while a decreased intensity at 3.1 A was consistent with loss of coordinated Zr neighbours (from CNZr4.17 to 3.33). An average loss of Zr neighbours could be, in part, due to elongation of Zr-0 bond tethered to ferrocene. In Figure 7, first shell scattering of Fe in pure ferrocene was fitted to Fe-C single scattering paths corresponding cyclopentadienyl carbons with secondary contribution from a multiple scattering path originating from two adjacent carbons. The average Fe-C bond distance was determined to be 2.063 A. In contrast, spectral features of ferrocene in UiO-66 were more complex and far less intense. The broad feature between 1.2-2.2 A was associated with two Fe-C scattering paths corresponding to effective radii of 2 and 2.1 A, while the sharp feature at 2.6 A was due to carbon scatterers situated on an outstretched region of the cyclopentadienyl ring. The distortion of the ferrocene geometry was brought about from a weakening of the ligand field due to carboxylate tether with a Zr node. This was further supported by the presence of an EPR signal as shown in Figure 8. A smaller, broad feature at 3.4 A was associated with carbon in the aforementioned carboxylate tether.

[0079] The occupancy of ferrocene decreased the specific pore volume and surface area. The surface area of UiO-66 in particular, decreased by about 23% after ferrocene doping, whereas its total pore volume suffered a loss of about 13%. While the isotherms of both samples were predominantly microporous (Figure 9 inset), there were significant differences in pore size distribution. The pore size distribution of UiO-66 in Figure 9 shows a sharp feature culminating at approximately 6.07 A accompanied by a broad feature between 12-23 A. A primary pore width of 6.07 A is consistent with aperture sizes of UiO-66 while the broad feature suggests a distribution of pore sizes enlarged by missing linker defects. Fc-UiO-66 on the other hand, showed a slight increase in primary pore width (8.04 A) while the accompanying broad feature became significantly less intense.

[0080] In Figure 10, the CV curve for Fc-UiO-66 shows a pair of redox peaks centred at about 0.31 V, whereas no discernible electrochemical activity was observed for UiO-66. The highly symmetric peaks indicated a highly reversible redox process not limited by poor electrical conductivity. This contrasts with prior art, in which the CV curve for a ferrocene doped sample of UiO-66 showed highly asymmetric redox behaviour and an arduous oxidation process. The differences in electrochemical response are attributed to different electrode fabrication techniques (slurry coating vs. film deposition), where the use of conductive multi-walled carbon nanotubes (MWCNTs) enhanced electron transport.

[0081] Figure 11 depicts the results of further investigation conducted in various salt solutions representing the oxyanions. At 0.1 M, CV curve of Cl shows a typical pair of symmetric redox peaks centred at E1 / 2 = 0.31 V. When the concentration was decreased by a factor of 10, oxidative and reductive peaks started to broaden and shift. A further 10-fold reduction in concentration resulted in severely diminished current densities and a heavily distorted voltammogram. At high Cl’ concentration, diffusion was fast enough to counterbalance the positively charged ferrocenium on the surface and within MOF crystallites. However, as anion concentration was decreased, bulk diffusion slowed and ferrocene embedded deep within the MOF cannot be accessed efficiently. These observations were also made across NOs', SO42; and PC>43', yet there appeared to be other factors at play. While voltammogram distortion was only visible at 0.001 M for Cl it became visible at 0.01 M for oxyanion solutions. At 0.01 M, shifts in oxidative and reductive peaks had caused voltammogram shearing, and E1 / 2 shifted to lower potentials. This could be caused by steric hindrance and pore blockage since oxyanions are larger and bulkier. However, N2 sorption results showed the existence of sufficiently large pores (pore width: -8.04 A; hydrated radii of Cl" 3.32 A, NO3" 3.35 A, SO42" 3.79 A and PC3" 3.39 A). If size effects are apparent, CV curves should exhibit much lower current densities (especially between Cl' and SC>42'). Instead, similar current densities were observed which indicated that an equivalent amount of ferrocene / ferrocenium was accessed. Hence, the inefficiency in ion transport was likely caused by ion association with ferrocenium within the confined space of the pore interior. The ease at which an anion is adsorbed or released directly influenced the rate at which the redox reaction proceeds. In the case of the highly charged trivalent PC3', PC3' was strongly associated to the positively charged pore environment of both the Zr node and the ferrocenium. This hampered diffusion and resulted in a lower current density.

[0082] Figure 12 shows experiments performed on a Fc-UiO-66 anode in dual-ion solution mixtures with a 1 mM sodium-based oxyanion salt along with 10 mM NaCI as background. To adsorb the ions, Fc-UiO-66 anode was oxidised at constant potentials of 0.4 - 1 V for 15 minutes with an activated carbon / cation-exchange membrane serving as the cathode. To regenerate the electrodes, Fc-UiO-66 was reduced at -0.2 V for 15 minutes. A preferential removal of oxyanions over Cl’ was observed across all experiments. A notable feature was the lack of selectivity at 0.4 V which is attributed to incomplete conversion of ferrocene to ferrocenium. At potentials above 0.4 V, all accessible ferrocene would have been converted to ferrocenium and oxyanion uptake was therefore higher.

[0083] Within the limits of experimental error, the amount of Cl’ adsorbed was fairly consistent across all potentials in all solution mixtures. Oxyanion adsorption when competing against to a 10-fold background of Cl’ generally followed a trend of PO43' > SO42' > NOs' , although the calculated selectivities indicated a greater preference for SO42' between the potentials of 0.6-0.8 V. The highest ion selectivities were recorded at 2.53 for SO42', 2.21 for PO43' and 1.44 for NOs'. The reason for enhanced oxyanion adsorption can be deduced by considering the adsorption of NOs' and Cl'. Since both NOs' and Cl' possess similar hydrated radii and hydration enthalpies, a higher adsorption of Cl' was expected given the 10-fold concentration advantage. However, experimental evidence depicted only a marginally higher adsorption of NC ' over Cl' across oxidation potentials of 0.5- 1.0 V, yielding an average selectivity of about 1.3. This result was attributed to favourable attractive interactions between oxygen constituents of nitrate and the positively charged environment of ferrocenium and zirconium. The effects of ion association manifested more strongly in species carrying a more negative charge such as SOU2' and PO '. While strong ion association promoted ion competitiveness and displace weakly associated Cl' , it also resulted in slower charge transport and retarded regeneration of the electrode.

[0084] Figure 13 shows experiments performed on a Fc-UiO-66 anode in multi-ion solution mixtures where the concentration of oxyanions was set at 1 mM and a background of 10 mM Ch was used. Figure 13 shows that the adsorption of PO43' was distinctly higher across all applied potentials and an apparent selectivity of 3.80 was reached at 0.7 V. In the absence of PO43', SO42' was preferentially adsorbed but to a lesser degree. The apparent selectivity of SO42' hovered at an average of about 1.4 which is considerably less than what was achieved in the dual-ion solution mixture. The adsorption of NOs' in all mixtures mimicked its performance in the dual-ion solution experiment and is only slightly more adsorbed compared to the Cl' background.

[0085] The stability and performance of the Fc-UiO-66 anode was further investigated by considering the effects of electrolyte pH from 4-10 (Figure 14). There appeared to be no significant change in adsorbed amounts at low pH, yet decreasing adsorption could be observed as pH was increased. Loss of selectivity set in as early as pH 7.6 for NOs' and at pH 8.5 for SO42'. This was due to gradual degradation of the MOF framework as a result of increasing linker solubility, as terephthalate linkers of UiO-66 start to leach as pH is increased. The leaching of terephthalate linkers is closely linked to its pKa value (-3.51 at 298 K) where linker dissociation is expected at pH values above 5. The dissociation of linkers destabilised pore structure and can lead to pore collapse. However, this is less impactful for the adsorption of PO43' since PCU3' directly coordinates with the zirconium node and helped to stabilise pore structure.

[0086] Whilst the foregoing description has described exemplary embodiments, it will be understood by those skilled in the technology concerned that many variations may be made without departing from the present invention.

Claims

Claims1. An electrode comprising a substrate coated with a redox-active modified metalorganic framework (MOF), wherein the redox-active modified MOF is a ferrocene- functionalised MOF.

2. The electrode according to claim 1 , wherein the ferrocene-functionalised MOF is ferrocene-functionalised UiO-66 (Fc-UiO-66).

3. The electrode according to claim 1 or claim 2, wherein the electrode is an anode.

4. The electrode according to any preceding claim, wherein the redox-active modified MOF is stable in water.

5. The electrode according to any preceding claim, wherein the substrate is a conductive material.

6. The electrode according to any preceding claim, wherein the electron is comprised in an electrosorption unit.

7. An electrosorption unit cell comprising an electrode according to any preceding claim.

8. The electrosorption unit according to claim 7, wherein the electrode is an anode.

9. The electrosorption unit according to claim 8, wherein the electrosorption unit further comprises a cathode.

10. The electrosorption unit according to claim 9, wherein the anode and the cathode are separated by a separator.

11. A method of separating at least one target molecule from a fluid source, the method comprising: placing in a fluid source a first electrode and a second electrode, wherein the first electrode is an electrode according to any of claims 1 to 5; andapplying an electrical potential across the first electrode and the second electrode such that the first electrode transforms to an oxidised state and selectively binds to a target electron-donating functional group of a target molecule present in the fluid source, such that the at least one target molecule comprising the target electron-donating functional group is separated from the fluid source.

12. The method according to claim 11 , wherein the second electrode selectively captures a cationic species present in the fluid source.

13. The method according to claim 12, further comprising reversing the applied electrical potential to release the bound target molecule from the first electrode and / or the captured cationic species from the second electrode.

14. The method according to any of claims 11 to 13, wherein the target molecule comprises an oxyanion.

15. The method according to any of claims 11 to 14, wherein the target molecule comprises phosphate, sulphate and / or nitrate.

16. The method according to any of claims 11 to 15, wherein the fluid source is an aqueous fluid and / or an organic fluid.

Images