WO2018065555A1

WO2018065555A1 - Metal-organic frameworks, methods for their manufacture and uses thereof

Info

Publication number: WO2018065555A1
Application number: PCT/EP2017/075429
Authority: WO
Inventors: David FAIREN-JIMENEZ; Tian Tian
Original assignee: Immaterial Labs Ltd
Priority date: 2016-10-06
Filing date: 2017-10-05
Publication date: 2018-04-12
Also published as: GB201617000D0

Abstract

A metal-organic framework (MOF) body is disclosed which comprises MOF crystallites adhered to each other via a MOF binder. The MOF crystallites and the MOF binder are formed of HKUST-1, i.e. Cu₃(BTC)₂-3H₂O. The MOF body has utility for storing and/or separating gases such as CH₄, CO₂, O₂, NH₃, Ar, CO, N₂ and C₂H₄ (ethylene); toxic industrial gases such as benzene, toluene, xylenes, sulphur dioxide, ethylene oxide; and warfare agents such as sarin, mustard gas and derivatives thereof.

Description

METAL-ORGANIC FRAMEWORKS. METHODS FOR THEIR MANUFACTURE AND

USES THEREOF

BACKGROUND TO THE INVENTION

Field of the invention

The present invention relates to metal-organic framework (MOF) materials, to methods for their manufacture and to practical applications of such materials. The invention has particular, but not exclusive, applicability to monolithic forms of the materials. Suitable applications include gas adsorption applications such as for gas storage, separation and filtration.

Related art

Metal-organic frameworks (MOFs) are porous crystalline materials prepared by the self- assembly of metal ions and organic ligands. MOFs can have large pore volumes and apparent surface areas as high as 8,000 m²/g. MOFs combine a structural and chemical diversity that make them attractive for many potential applications, including gas storage, gas separation and purification, sensing, catalysis and drug delivery. The most striking advantage of MOFs over more traditional porous materials is the possibility to tune the host / guest interaction by choosing the appropriate building blocks, i.e. the metal ions and organic ligands, from which the MOF is formed.

WO 2010/148463 discloses a method for synthesis of MOFs in which the synthesis conditions are mild - typically below 30°C - and the synthesis proceeds relatively quickly - typically in less than 1 hour. The synthesis takes place in a mixture of water and ethanol. The material studied in WO 2010/148463 is Cu₃(BTC)₂-type MOF. Fu et al (2013) reports on efforts to incorporate a MOF (UiO-66) into a copolymer (MAA- co-EDMA) matrix, for use in liquid chromatography. The resulting structure is described as a "monolith", and comparisons are made with a monolith formed using the copolymer only. SEM analysis shows that the microstructure of the material includes spherical MOF particles which adhere to the copolymer matrix. Huang et al (2013) provides similar disclosure.

US 2010/0181212 discloses MOF materials supported on open cell polymer foam structures, for use in gas storage applications.

Kijsgens et al (2010) discloses the manufacture of Cu3(BTC)2 MOF material in situ on cordierite monolithic honeycomb structures. The results are reported to be poor.

Additionally, Kijsgens et al (2010) discloses the manufacture of Cu3(BTC)2-based honeycomb structures, formed by mixing Cu3(BTC)2 powder with a silicone-based binder and a methyl hydroxyl propyl cellulose plasticizer. The structures were formed by extrusion and subsequent drying at 120°C.

SUMMARY OF THE INVENTION

The present inventors consider that one of the main challenges for the industrial use of MOFs is to prepare them in a suitable shape for a given application, in order to translate advantageous properties of the materials into industrial products. During the synthesis processes used to date in the art, MOFs are generally obtained in a powdered crystalline state. This makes them costly to shape for final industrial applications. Furthermore, the use of binders and high-pressure processes to pelletize the material in order to create suitable monolithic structures causes significant reductions of the porous properties (e.g. the BET surface area per unit mass and/or the degree of microporosity) of the material. Porous properties may be reduced due to collapses in the porosity when using high pressures, pore blocking caused by the binder preventing access to the porosity, and/or the presence of the binder reducing the final gravimetric amount of adsorbent in the pellet. In addition, pellets may present low densities of MOF due to the presence of interstitial spaces between the powdered crystallites of MOF, causing low volumetric adsorption capacities, as well as reduced mechanical properties compared to the MOF single crystal. The present invention has been devised in order to address at least one of the above problems. Preferably, the present invention reduces, ameliorates, avoids or overcomes at least one of the above problems.

In a first development, the present invention provides a metal-organic framework (MOF) body comprising MOF crystallites adhered to each other via a MOF binder, wherein the MOF crystallites and the MOF binder are formed of HKUST-1.

In a second development, the present invention provides a metal-organic framework (MOF) body consisting of:

MOF crystallites;

a MOF binder which binds the crystallites together in the body;

optionally, residual solvent; and

optionally, one or more additives, wherein the additives are present at a level of not more than 10% by mass,

wherein the MOF crystallites and the MOF binder are formed of HKUST-1. There will now be set out various preferred aspects of the invention that may be applied to and/or combined with the first and/or second development. In a first preferred aspect, the present invention provides a metal-organic framework (MOF) monolith having a volume of at least 10 mm³, wherein:

(i) when the monolith is formed from a composition capable of forming a MOF single crystal of the same composition, the BET surface area per unit bulk volume of the monolith is at least 0.6 times the BET surface area per unit bulk volume of said MOF single crystal of the same composition; and

(ii) when the monolith is formed from a composition not capable of forming a single crystal of the same composition, instead being capable of forming a MOF single crystal and one or more remaining components of the composition, the BET surface area per unit bulk volume of the monolith is at least 0.6 times the volumetric weighted arithmetic mean of the BET surface area per unit bulk volume of said MOF single crystal and said remaining components,

and wherein the BET surface area per unit bulk volume is determined based on the N2 adsorption isotherm at 77K. Certain optional features of the first aspect of the invention are now set out. These may be applied singly or in any combination, unless the context demands otherwise. These may also be applied, in any combination unless the context demands otherwise, to any other aspect and/or development of the invention. Preferably, the monolith has a BET surface area per unit bulk volume of at least 0.7 times, 0.8 times or 0.9 times (i) the BET surface area per unit bulk volume of the MOF single crystal of the same composition or (ii) the volumetric weighted arithmetic mean of the BET surface area per unit bulk volume of said MOF single crystal and said remaining components.

Where necessary, the BET surface area per unit bulk volume of a MOF single crystal can alternatively be determined by calculation based on knowledge of the crystal structure and the micro-pores entrained in that crystal structure. The single crystal is therefore considered to be free of meso- and macro-pores.

Preferably, the BET surface area per unit bulk volume of the monolith is at least 600 m²/cm³. The volume of the monolith (or body) may in some cases be smaller than 10 mm³. For example, the volume of the monolith may be at least 1 mm³, more preferably at least 2 mm³, more preferably at least 3 mm³, more preferably at least 4 mm³, more preferably at least 5 mm³. In a second preferred aspect, the present invention provides a metal-organic framework (MOF) monolith having a volume of at least 10 mm³ wherein the BET surface area per unit bulk volume of the monolith is at least 600 m²/cm³, wherein the BET surface area per unit bulk volume is determined based on the N2 adsorption isotherm at 77K. The volume of the monolith (or body) may in some cases be smaller than 10 mm³. For example, the volume of the monolith may be at least 1 mm³, more preferably at least 2 mm³, more preferably at least 3 mm³, more preferably at least 4 mm³, more preferably at least 5 mm³.

In describing some aspects and embodiments of the present invention, and comparative materials, it is useful to present values for porosity in terms of volume percent (vol%). This represents the ratio of the total volume of the pores (sometimes within a defined size range) to the volume of the monolith. It is possible to measure the bulk volume of a monolith by the Archimedes method in a mercury porosimeter, i.e. by determining the volume of mercury displaced by the monolith before allowing the mercury to infiltrate the pores of the monolith.

In a third preferred aspect, the present invention provides a metal-organic framework (MOF) monolith having a volume of at least 10 mm³, the monolith having a meso-porosity of at most 10 vol%, wherein meso-porosity is defined as pores with diameter in the range 2-50 nm (macro-porosity being defined as pores of greater than 50 nm diameter), the porosity and pore size distributions being determined based on the N2 adsorption isotherm at 77K.

The volume of the monolith (or body) may in some cases be smaller than 10 mm³. For example, the volume of the monolith may be at least 1 mm³, more preferably at least 2 mm³, more preferably at least 3 mm³, more preferably at least 4 mm³, more preferably at least 5 mm³. It is at present considered that the determination of porosity and pore size distributions based on the N2 adsorption isotherm at 77K is suitable for determination of micro- and meso-porosity for MOF materials. Determination of porosity and pore size distributions over 50 nm, i.e. macro-porosity, may be carried out by alternative methods, such as mercury porosimetry.

Preferably, the MOF monolith has a micro-porosity, defined as pores with diameter less than 2 nm, of at least 40 vol%. More preferably, the MOF monolith has a micro-porosity of at least 50 vol%, still more preferably at least 55 vol% and still more preferably at least 60 vol%.

In a fourth preferred aspect, the present invention provides a metal-organic framework (MOF) monolith having a volume of at least 10 mm³, wherein:

(i) when the monolith is formed from a composition capable of forming a MOF single crystal of the same composition, the monolith has a micro-porosity, defined as pores with diameter less than 2nm, of at least 0.6 times the micro-porosity of a MOF single crystal of the same composition; and

(ii) when the monolith is formed from a composition not capable of forming a single crystal of the same composition, instead being capable of forming a MOF single crystal and one or more remaining components of the composition, the monolith has a micro-porosity of at least 0.6 times the volumetric weighted arithmetic mean of the micro-porosity of said MOF single crystal and said remaining components,

and wherein the porosity and pore size distributions are determined based on the N2 adsorption isotherm at 77K. The delimitation of micro-porosity, meso-porosity and macro-porosity is as follows:

Micro-porosity: pore sizes below 2 nm

Meso-porosity: pore sizes in the range 2-50 nm

Macro-porosity: pore sizes larger than 50 nm

This approach follows that of lUPAC and is applicable to porous materials, including MOF materials [Rouquerol et al (1994) and Sing (1982)].

Where necessary, the micro-porosity of a MOF single crystal can alternatively be determined by calculation based on knowledge of the crystal structure and the micropores entrained in that crystal structure. The single crystal is therefore considered to be free of meso- and macro-pores.

Preferably, the monolith has a micro-porosity of at least 0.7 times, 0.8 times or 0.9 times (i) the micro-porosity of the MOF single crystal of the same composition or (ii) the volumetric weighted arithmetic mean of the micro-porosity of said MOF single crystal and said remaining components.

Preferably, the density of the MOF monolith is at least 90% of (i) the density of the MOF single crystal of the same composition or (ii) the volumetric weighted arithmetic mean of the density of said MOF single crystal and said remaining components. In this case, the density of the MOF single crystal of the same composition can be determined by calculation based on knowledge of the crystal structure. The single crystal is considered to be free of meso- and macro-pores. More preferably, the MOF monolith has a density of at least 95%, more preferably at least 100%, more preferably at least 105%, more preferably at least 1 10%, more preferably at least 1 15% or more preferably at least 120% of the density of (i) the MOF single crystal of the same composition or (ii) the volumetric weighted arithmetic mean of the density of said MOF single crystal and said remaining components.

In some embodiments, it is preferred for the monolith to have as low a value for the meso- and macro-porosity as possible. For example the cumulative meso- and macro-porosity may be less than 1vol%. This is advantageous where the intended application of the material is as an adsorbent material, e.g., for gases such as CO2, H2, CH4, etc., where the development of high micro-pore volumes allows the adsorption of greater amounts of the relevant gas.

However, in some embodiments, it is preferred for the monolith to have some meso- and/or macro-porosity, in order to promote flow through the monolith. In these

circumstances, the meso- and macro-pores provide flow passages to the micro-pores. Whether this is wanted again depends on the intended application of the material, the advantage of improved transport through the monolith being balanced against lower available surface area for adsorption due to proportionally smaller amount of micro- porosity per unit volume. Meso- and/or macro-porosity can be included in the monolith by the use of additives in the manufacturing process. Deliberate hierarchical porosity of this type can provide a useful balance in the properties of the monoliths for particular applications which require flow through the monolith.

For example, the monolith can be produced using a template material, in and/or around which the monolith is allowed to form. The template material can subsequently be removed to leave a suitable network of porosity through the monolith. The network of porosity can be meso- and/or macro-scale porosity. Suitable bi-continuous porosity can be formed in the context of MOFs as reported in Cao et al (2013).

Using N2 adsorption at 77K, it is possible to measure micro- and meso-porosity. Typically, the micro-pore volume is obtained at relative pressure P/Po = 0.1 , whereas adsorption at higher pressures is related to meso-porosity until P/Po = 0.98.

Mercury porosimetry can be used to measure the "bulk" density of the monolith since Hg does not penetrate any porosity at atmospheric pressure. In an alternative arrangement, mercury porosimetry can also be used to measure macro and meso-porosity, by increasing the Hg pressure and measuring the extent of intrusion of Hg into the pores of the monolith with pressure. Mercury porosimetry cannot be used to measure the micro- porosity. Some work reported for comparison purposes relates to powder materials. In this case, the bulk density is the tap bulk density, i.e. the apparent density of a powder based on causing a sample of the powder to settle in a receptacle by tapping, measuring the mass and dividing this by the apparent volume of the sample. For MOF materials in powder form, this therefore includes the volume occupied by micro-, meso- and macro-porosity and also interstitial spaces between the powder particles.

The volume of the monolith (or body) may in some cases be smaller than 10 mm³. For example, the volume of the monolith may be at least 1 mm³, more preferably at least 2 mm³, more preferably at least 3 mm³, more preferably at least 4 mm³, more preferably at least 5 mm³. Preferably, the monolith has a volume of at least 50mm³, more preferably at least 100mm³, more preferably at least 500mm³, still more preferably at least 1000mm³. The term "monolith" is intended to include self-supporting bodies. It is intended to exclude forms of material that are formed on a substrate or other support, or which rely on another structure to be supported.

Preferably, the MOF monolith or body has a smallest linear dimension of at least 1 mm. That is, assuming that the monolith is not perfectly spherical, the shortest straight line passing through the material of the monolith has a length in the monolith of at least 1 mm. This dimension may be considered to be the thickness of the monolith, depending on the overall shape of the monolith. More preferably, the MOF monolith has a smallest linear dimension of at least 5 mm. The monolith may comprise a composite material comprising particles of a first MOF composition in a matrix of a second MOF composition, as explained further below in relation to an independent aspect of the invention.

In a fifth preferred aspect, the present invention provides a metal-organic framework (MOF) monolith having a volume of at least 10 mm³ which is substantially transparent.

The volume of the monolith (or body) may in some cases be smaller than 10 mm³. For example, the volume of the monolith may be at least 1 mm³, more preferably at least 2 mm³, more preferably at least 3 mm³, more preferably at least 4 mm³, more preferably at least 5 mm³. In a sixth preferred aspect, the present invention provides a metal-organic framework (MOF) body comprising MOF crystallites adhered to each other via a MOF binder. Preferably, the MOF crystallites have an average particle size of not more than 300 nm. The particle size can be measured using SEM or, more preferably (for accurate size measurement for smaller particles) by TEM. The particle size of 15 particles is measured based on a random selection of field of view and the average taken. The MOF crystallites may have an average particle size of not more than 200 nm, more preferably not more than 150 nm, still more preferably not more than 145nm.

Preferably, the MOF body is a monolith. Preferably the body has a volume of at least 10 mm³. Further preferred ranges for the volume of the body are set out above in relation to the monolith. For example, the volume of the monolith (or body) may in some cases be smaller than 10 mm³. For example, the volume of the monolith may be at least 1 mm³, more preferably at least 2 mm³, more preferably at least 3 mm³, more preferably at least 4 mm³, more preferably at least 5 mm³,

Alternatively, the MOF body is a layer formed on a substrate. The nature of the substrate is not particularly limited. In the case where the intended application of the layer is based on substantial transparency of the layer, preferably the substrate is transparent or substantially transparent.

The crystallites typically have different orientation to each other. For example, the crystallites may be substantially randomly oriented. The MOF binder preferably has substantially the same composition as the MOF crystallites. However, the MOF binder may have a different porosity or pore size distribution to the MOF crystallites. The MOF binder may have a lower degree of crystallization than the MOF crystallites. For example, the MOF binder may be substantially amorphous.

Alternatively, the MOF binder may have a different composition to the MOF crystallites. In that case, the MOF body may be formed of a composite MOF material.

In a seventh preferred aspect, the present invention provides a metal-organic framework (MOF) body consisting of:

MOF crystallites;

a MOF binder which binds the crystallites together in the monolith;

optionally, residual solvent; and

optionally, one or more additives, wherein the additives are present at a level of not more than 10% by mass.

Optional features set out with respect to the sixth aspect are applicable also to the seventh aspect.

Preferably, if present, the additives are present at a level of not more than 5% by mass, more preferably not more than 3% by mass, more preferably not more than 2% by mass, still more preferably not more than 1 % by mass. It is permitted for unavoidable impurities to be present in the body. Preferably, if present, the residual solvent is present at a level of not more than 5% by mass, more preferably not more than 3% by mass, more preferably not more than 2% by mass, still more preferably not more than 1 % by mass.

There may also be present residual reactants, i.e. materials which could have reacted together to form MOF, but did not. Additionally, there may also be present a non-porous phase (which may be amorphous). There may also be present by-products of the reaction to form MOF. The total amount of MOF (i.e. MOF crystallites and MOF binder) may be at least 60% by mass, more preferably at least 65% by mass, more preferably at least 70% by mass and more preferably at least 75% by mass.

Preferably, the MOF body is a monolith. Preferably the body has a volume of at least 10 mm³. Further preferred ranges for the volume of the body are set out above in relation to the monolith. For example, the volume of the monolith (or body) may in some cases be smaller than 10 mm³. For example, the volume of the monolith may be at least 1 mm³, more preferably at least 2 mm³, more preferably at least 3 mm³, more preferably at least 4 mm³, more preferably at least 5 mm³, The crystallites typically have different orientation to each other. For example, the crystallites may be substantially randomly oriented.

The MOF binder preferably has substantially the same composition as the MOF crystallites. However, the MOF binder may have a different porosity or pore size distribution to the MOF crystallites. The MOF binder may have a lower degree of crystallization than the MOF crystallites, For example, the MOF binder may be substantially amorphous.

In an eighth preferred aspect, the present invention provides a metal-organic framework (MOF) monolith, or a MOF layer formed on a substrate, wherein:

(i) when the monolith or layer is formed from a composition capable of forming a MOF single crystal of the same composition, the monolith has a Young's modulus, and/or hardness, measured via nanoindentation, greater than the Young's modulus and/or hardness of a MOF single crystal of the same composition; and

(ii) when the monolith or layer is formed from a composition not capable of forming a single crystal of the same composition, instead being capable of forming a MOF single crystal and one or more remaining components of the composition, the monolith has a Young's modulus, and/or hardness, measured via nanoindentation, greater than the volumetric weighted arithmetic mean of the Young's modulus and/or hardness of said MOF single crystal and said remaining components. Where (ii) applies, the monolith or layer may comprise a composite material comprising particles of a first MOF composition in a matrix of a second MOF composition.

The monolith may have a volume of at least 10 mm³, for example, or another preferred range of volume as set out above. For example, the volume of the monolith (or body) may in some cases be smaller than 10 mm³. For example, the volume of the monolith may be at least 1 mm³, more preferably at least 2 mm³, more preferably at least 3 mm³, more preferably at least 4 mm³, more preferably at least 5 mm³,

The Young's modulus (used here interchangeably with the term "elastic modulus"), and/or hardness, may be at least 1.05 times, more preferably at least 1.5 times or at least 2 times (i) the Young's modulus and/or hardness of the MOF single crystal of the same

composition or (ii) the volumetric weighted arithmetic mean of the Young's modulus and/or hardness of the MOF single crystal and said remaining components. In a ninth preferred aspect, the present invention provides a population of monoliths or bodies according to any one of the first to eighth aspects.

Such a population is of use in various applications. For example, the monoliths may be of substantially similar shape and/or dimensions. They may be used in a column

arrangement with the spaces between them allowing for fluid (e.g. gas) flow. This is useful for gas separation applications. The number of monoliths in the population is not particularly limited, but as an example the number of monoliths may be at least 10, or at least 50, or at least 100. In a tenth preferred aspect, the present invention provides a process for manufacturing a metal-organic framework (MOF) monolith, or a MOF layer formed on a substrate, wherein the process includes the steps:

allowing the reaction of MOF precursors in a solvent to form the MOF composition; and forming a monolith or layer of the MOF composition including a drying stage to remove at least some of the solvent with a maximum temperature during the drying stage of not more than 50°C. The process may include a step of concentration of particles of the MOF composition into a concentrate of the particles and solvent. This step may be carried out, for example, by centrifugation.

The maximum temperature during the drying stage is preferably not more than 40°C, more preferably not more than 30°C.

Preferably, the monolith is formed into a desired shape by the drying stage taking place with the material in a mould. The drying material then preferably conforms to the shape of the mould.

In an eleventh preferred aspect, the present invention provides a process for

manufacturing a metal-organic framework (MOF) body, wherein the process includes the steps:

allowing the reaction of MOF precursors in a solvent to form the MOF composition;

concentration of particles of the MOF composition into a concentrate of the particles and solvent;

addition of additional MOF precursors to the concentrate; and

forming a body of the MOF composition including a drying stage to remove at least some of the solvent with a maximum temperature during the drying stage of not more than 50°C. Preferably, the MOF body is a MOF monolith, Alternatively the MOF body is a MOF layer formed on a substrate.

The step of concentrating the particles of the MOF composition into a concentrate of the particles and solvent may be carried out, for example, by centrifugation.

The maximum temperature during the drying stage is preferably not more than 40°C, more preferably not more than 30°C. Preferably, the body is formed into a desired shape by the drying stage taking place with the material in a mould. The drying material then preferably conforms to the shape of the mould.

In a twelfth preferred aspect, the present invention provides a process for manufacturing a metal-organic framework (MOF) body, wherein the process includes the steps:

providing particles of a first MOF composition;

addition of MOF precursors, corresponding to a second MOF composition, to the particles of the first MOF composition; and

forming a composite MOF body comprising said particles of said first MOF composition in a matrix of the second MOF composition.

Preferably, the MOF body is a MOF monolith. Alternatively the MOF body is a MOF layer formed on a substrate. Suitable MOF compositions for use with the present disclosure are: ZIFs, such as ZIF-4, ZIF-8, ZIF-90, ZIF-zni;

UiO-Frameworks, such as UiO-66, UiO-67, UiO-68;

HKUST-1 ;

Al-fumarate;

MOF-177;

MIL-47, MIL-53.

Of these, HKUST-1 is preferred in the present invention. HKUST-1 is also known as CuBTC MOF, where BTC is 1 ,3,5-tricarboxylic acid. It can be expressed as

Cu3(BTC)2-3H20. It is also known using various proprietary names. The three water molecules can be removed, so that the material is Cu3(BTC)2.

In the case of a composite MOF material, the first MOF composition may be any of the MOF compositions listed above and the second MOF composition may be any other of the MOF compositions listed above.

In a further aspect, the present invention provides a method for storage and/or for separation of a gas, the method including the steps:

providing at least one MOF body or monolith according to any preceding aspect of the invention;

providing a gas for storage and/or separation; and

exposing the MOF body to the gas for storage and/or separation at a pressure, to allow the gas for storage and/or separation to be adsorbed by the MOF body, wherein the gas for storage and/or separation is selected from the group consisting of: CH4, CO2, O2, NH3 , Ar, CO, N2 and C2H4 (ethylene); toxic industrial gases such as benzene, toluene, xylenes, sulphur dioxide, ethylene oxide; and warfare agents such as sarin, mustard gas and derivatives thereof,

and wherein:

when the gas for storage and/or separation is CH4, the volumetric storage capacity of the MOF body is at least 200 cm³ (STP) per cm³ of the MOF body at 298K and a gas storage pressure of 65 bar, and

when the gas for storage and/or separation is CO2, the volumetric storage capacity of the MOF body is at least 300 cm³ (STP) per cm³ of the MOF body at 298K and a gas storage pressure of 40 bar.

Preferably, when the gas for storage and/or separation is CH4, the volumetric storage capacity of the MOF body is at least 220 cm³ (STP) per cm³ of the MOF body at 298K and a gas storage pressure of 65 bar. More preferably, the volumetric storage capacity of CH4 by the MOF body under the same conditions is at least 220 cm³ (STP) per cm³, at least 240 cm³ (STP) per cm³, at least 250 cm³ (STP) per cm³, at least 260 cm³ (STP) per cm³, or at least 263cm³ (STP) per cm³.

Preferably, when the gas for storage and/or separation is CO2, the volumetric storage capacity of the MOF body is at least 320 cm³ (STP) per cm³ of the MOF body at 298K and a gas storage pressure of 40 bar. More preferably, when the gas for storage and/or separation is CO2, the volumetric storage capacity of the MOF body is at least 340 cm³ (STP) per cm³ of the MOF body at 298K and a gas storage pressure of 40 bar. Preferably, in the method, the MOF body is exposed to the gas for storage at a pressure of at least 2 bar, more preferably at least 10 bar.

Where NH3 is adsorbed by the MOF body at low pressure, preferably the mass storage capacity is at least 0.05 mmol/g (more preferably 0.1 mmol/g) of NH3 at 290 Pa and 298K. This is a comparatively low loading but the pressure is also very low. Note that 290 Pa is absolute pressure.

However, for NH3, preferably the gas is adsorbed by the MOF body at ambient pressure. Where NH3 is adsorbed by the MOF body at ambient pressure, preferably the mass storage capacity is at least 3 mmol/g (more preferably more than 5.7mmol/g or at least 6 mmol/g 2500) of N H3. This compares favourably with the work of Jasuja et al (2015) who studied adsorption of NH3 streams with NH3 concentration 2876 ppm (or 2000 mg/m³), at 20C. Maximum capacity in that work obtained was 5.7mmol/g. However, in the use of the preferred MOF body according to the present embodiments, the volumetric storage capacity for NH3 is substantially higher.

Further optional features of the invention are set out below. BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments of the invention will now be described by way of example with reference to the accompanying drawings in which:

Figs. 1-4 show optical micrographs of various materials:

Fig. 1 shows ZIF-8-HT.

Fig. 2 shows ZIF-8-LT. Fig. 3 shows ZIF-8-ER under normal illumination and Fig. 4 shows ZIF-8-ER under 365 nm UV light.

Fig. 5 shows powder X-ray diffraction (PXRD) patterns of the different samples alongside a simulated pattern for ZIF-8.

Figs. 6, 7 and 8 show SEM micrographs of ZIF-8-HT, ZIF-8-LT and ZIF-8-ER, respectively.

Figs. 9A and 9B shows results of N2 adsorption at 77 K on various samples.

Fig. 10 shows the results of thermogravimetric analysis (TGA) on various samples.

Fig. 1 1 shows a TEM image of ZIF-8-LT.

Fig. 12 shows 10 overlaid load-depth curves for nanoindentation of a ZIF-8-LT sample. Fig. 13 shows an SEM image of two rows of 1000 nm indents made on a sample (5 x 5 mm) of ZIF-8-LT.

Fig. 16 shows 20 overlaid load-depth curves for the ZIF-8-LTHT sample.

Fig. 17 shows an SEM image of two rows of 1000 nm indents made on a sample (5 x 5 mm) of ZIF-8-LTHT.

Fig. 18 shows results for the elastic modulus of ZIF-8-LTHT as a function of indentation depth.

Fig. 19 shows results for the hardness of ZIF-8-LTHT as a function of indentation depth. Fig. 20 shows 6 overlaid load-depth curves for the ZIF-8-ER sample indented to 3000nm. Fig. 21 shows a 3000 nm indent made on a sample (5 x 5mm) of ZI F-8-ER.

Fig. 22 shows results for the elastic modulus of ZIF-8-ER as a function of indentation depth.

Fig. 23 shows results for the hardness of ZIF-8-ER as a function of indentation depth. Fig. 24 shows 15 overlaid load-depth curves for the ZIF-8-ER sample indented to

1000nm. Fig. 25 shows rows of 1000 nm indents made on a sample (5 x 5mm) of ZIF-8-ER.

Fig. 26 shows results for the elastic modulus of ZIF-8-ER as a function of indentation depth.

Fig. 27 shows results for the hardness of ZIF-8-ER as a function of indentation depth. Fig. 28 shows the results of mercury porosimetry showing pore size distribution of the macro- and mesoporosity of ZIF-8-LT and ZIF-8-ER.

Fig. 29 shows the PXRD pattern of a ZIF-zni monolith, alongside a simulated pattern for ZIF-zni. The inset shows a view of the monolith, in a similar manner to Figs. 2 and 3. Fig. 30 shows PXRD patterns of UiO-66, MIL-101 and ZIF-90, synthesized in the present work along with their simulated patterns.

Figs. 31 -33 show SEM images of UiO-66, MIL-101 and ZIF-90 synthesized in the present work, respectively.

Figs. 34 and 35 show SEM images of the cross-section of the MIL-101 @ZIF-8 and ZIF- 90@ZIF-8 composite monoliths, respectively.

Fig. 36 shows N₂ isotherms of ZIF-8LT (diamonds), UiO-66 (squares) and UiO-66@ZIF-8 (triangles)

Fig. 37 shows N₂ isotherms of ZIF-8LT (diamonds), MIL-101 (squares) and MIL-101 @ZIF- 8 (triangles).

Fig. 38 shows N₂ isotherms of ZIF-8LT (diamonds), ZIF-90 (squares) and ZIF-90@ZIF-8 (triangles).

Fig. 39 shows the PXRD patterns of monolithic of ZIF-8ER after immersion in water at 100°C for 3, 5 and 7 days.

Fig. 40 shows a schematic representation of monolithic and powder MOF synthesis, the synthesis following a sol-gel process, where a dense monolith is obtained under mild drying conditions while powders are obtained when the wet gel is dried at higher temperature or vacuum.

Fig. 41 shows an optical picture of the monolithic MOF, _m0noHKUST-1 , maintaining the shape of the mould where it was prepared.

Fig. 42(a) shows a comparison of PXRD patterns of _p0wdHKUST-1 (cross-shaped data points) and simulated single crystal (overlaid line, Pawley fitting).

Fig. 42(b) shows a comparison of PXRD patterns of monoHKUST-1 (cross-shaped data points) and simulated single crystal (overlaid line, Pawley fitting).

Fig. 43 shows a comparison of absolute volumetric ChU adsorption isotherms at 298 K on monoHKUST-1 (red solid circles), excess volumetric uptake on _m0noHKUST-1 (red empty circles), HKUST-1 pellets under hand packing (blue diamonds), H KUST-1 pellets packed under 2 Tons (black squares), and HKUST-1 pellets under 5 Tons (green triangles) [Peng et al (2012). DOE target of 263 cm³ (STP)/cm³ is represented by the red dashed line. Fig. 44 shows the volumetric and gravimetric adsorption isotherms of CO2 on _m0noHKUST- 1 at 298 K. Absolute uptake (solid circles), excess uptake (empty circles).

Figs. 45-47 illustrate nanoindentation on _m0noHKUST-1 .

Fig. 45 shows Young's modulus and Fig. 46 shows hardness as a function of indentation depth on a monoHKUST-1 sample. Averaged properties were derived from 60 indents, using penetration depths of 200 - 2000 nm, thus ensuring results are free from surface defects and tip calibration artefacts. Error bars are standard deviations calculated from 60 measurements. The inset in Fig. 45 shows the load-displacement raw data, the inset in Fig. 46 shows the hardness of _m0noHKUST-1 is doubled that of its conventional polycrystalline counterpart (H -200 MPa) [Bunschuh (2012)].

Fig. 47, left hand side, shows an optical micrograph showing the array of residual indents, showing no evidence of radial cracking. Fig. 47, right hand side shows an AFM profile depicting the 3D topography of a representative indent, showing there is no sign of surface cracking indicating good mechanical resilience of the monolith.

Figs. 48 and 49 show the SEM images of the monoliths and corresponding _pcWdHKUST-1 (powder HKUST-1 ), respectively.

Figs. 50 and 51 show linear and semi-log plots, respectively, of N2 adsorption isotherms on _m0noHKUST-1 at 77 K.

Fig. 52 shows the determination of maximum P/Po by applying Rouquerol's consistency criteria.

Fig. 53 shows BET representation of N2 isotherms for _m0noHKUST-1 .

Fig. 54 shows a comparison of absolute gravimetric CH4 adsorption isotherms at 298 K on monoHKUST-1 (red solid circles), excess gravimetric uptake on _m0noHKUST-1 (red empty circles), and powder HKUST-1 (black squares) from Peng et al (2013).

Fig. 55 shows a comparison of gravimetric absolute CO2 adsorption isotherms at 298 K for monoHKUST-1 , red circles; and powder HKUST-1 from Liang et a/ (2009).

Fig. 56 shows indentation modulus (/) of _m0noHKUST-1 plotted as a function of surface penetration depth.

Fig. 57 shows thermogravimetric analysis of monoHKUST-1.

Fig. 58 shows FTIR of _m0n₀HKUST-1 (red) and _p0wdHKUST-1 (black).

Fig. 59(a) shows a TEM image and primary particle size distribution of HKUST-1 synthesis at room temperature, 20 °C.

Fig. 59(b) shows a TEM image and primary particle size distribution of HKUST-1 synthesis at 40 °C.

Fig. 59(c) shows a TEM image and primary particle size distribution of HKUST-1 synthesis at 60 °C. Figs. 60(a), 60(b) and 60(c) show TEM images of the aggregation of _po»dHKUST-1 particles - synthesis at room temperature.

Fig. 61 shows transmission electron microscopy (TEM) images of monolithic and powder MOF samples. Image a shows _p0wdHKUST-1. Image b shows monoHKUST-1 . Images c show EDX analysis and elemental maps of the selected areas from images a and b. Fig. 62 shows TEM images of monoHKUST-1 (images a and b), and _PovJHKUST-1 (images c and d).

Fig. 63 shows electron diffraction patterns of monoHKUST-1 and _PovJHKUST-1.

Fig. 64 shows pore size distribution (PSD) of the macroporosity for _m0noHKUST-1 (red line) and _po»dHKUST-1 (black line), obtained through mercury porosimetry. Note the absence of macroporosity in _m0noHKUST-1.

Fig. 65a shows an SEM image of monoHKUST-1 with a higher magnification inset.

Fig. 65b shows an SEM image of _p0wdHKUST-1.

Fig. 66 shows equilibrium time of methane adsorption at 298 K as a function of equilibrium pressure for monoHKUST-1 (diamonds) and _p0wdHKUST-1 (circles).

Fig. 67 shows the decay of pressure with time, at 40 bar, for methane adsorption for monoHKUST-1 (diamonds) and _p0vJHKUST-1 (circles).

Fig. 68 shows thermogravimetric analysis of monoHKUST-1 (middle line at left hand side); activated monoHKUST-1 (top line, dotted, at left hand side); and _p0vJHKUST-1 (bottom line at left hand side). The vertical dotted line shows the 8 h activation at 120°C for activated monoHKUST-1 .

Fig. 69 shows the results of mass spectroscopy from TGA for four samples:

Fig, 69a: monoHKUST-1 ;

Fig. 69b: powdHKUST-1

Fig. 69c: activated monoHKUST-1 Fig. 69d: activated monoHKUST-1

In Fig. 69, lines are shown and labelled for H2O, ethanol, CO2, BTC. The vertical dotted line in Fig. 69c and the dotted square in Fig. 69d, shows the 8 h activation at 120 °C. Figs. 70-74 show gas adsorption isotherms at 298 K on _m0noHKUST-1 , absolute (solid circles) and excess (open circles) volumetric uptake, for CH4, 02,CC>2, N2 and Ar, respectively. CH4 is shown in Fig. 70 for ease of comparison.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS. AND FURTHER OPTIONAL FEATURES OF THE INVENTION

The embodiments of the present invention provide monolithic metal-organic frameworks (MOFs). These have utility, for example, in gas adsorption applications.

In the first part of this disclosure, the present inventors disclose the synthesis of monolithic ZIF-8 structures. This is a MOF that has received lot of attention in the last years due to its high thermal and chemical stability. The synthesis process can be a one-step process and produces robust monoliths avoiding the use of binders and/or high pressure. This therefore addresses the problems discussed above related to porosity loss and reduced mechanical properties. The work reported here concentrates on ZIF-8 as a model MOF, but the invention also applies to other MOF materials which are conventionally produced in powdery form. Thus, although ZIF-8 is not the focus of the scope of protection of the present disclosure, the disclosure of the synthesis, treatment and characterization of ZIF-8 is useful for understanding the remainder of the disclosure,

In the second part of this disclosure, the present inventors disclose the synthesis of monolithic HKUST-1 structures. ZIF-8 and related work

Briefly stated, the monolithic ZIF-8 structure is produced at ambient temperature by mixing a solution of Zn²⁺ and 2-methyl imidazole in ethanol (or another suitable solvent such as methanol or DMF), and drying at ambient conditions, optionally including a centrifugation step to speed up the removal of solvent. It can be prepared with the desired shape for industrial use in a low-cost process. Mechanical assays using nano-indentation shows that the monolithic material presents more robust mechanical properties compared with the original ZI F-8 single crystals. This is significant for industrial application in which the material may be subjected to mechanical stress. Industrial applications of the material include: gas adsorption/separation technologies, such as gas storage (e.g. h storage), carbon capture, gas purification, such as hb and/or CH4 purification, gas separation, such as ethane/ethylene separation, capture of warfare and toxic industrial compounds (e.g. xylenes, SO2, ammonia, nerve agents, etc.); gas sensing; replacement of zeolites;

replacement of activated carbon; coating of activated carbon; drug delivery; catalysis; and water treatment.

In contrast with crystalline ZIF-8 powder obtained from known synthesis techniques, the monoliths disclosed here are still fully or mainly crystalline but are substantially transparent. This makes them good candidates for sensing applications.

SEM images and N2 adsorption measurements show that meso- and macro-porosity is substantially absent in the monoliths. This leads to higher densities and therefore higher volumetric capacities critical for adsorption processes in industrial applications. Described below is a one-step synthetic procedure to produce crystalline monolithic ZIF-8 [Zn(C4H5N2)2]. These macro-scale structures are substantially more rigid than single crystals of the same composition, whereas gas adsorption studies showed they retain the characteristic porosity of ZIF-8. Monolithic structures showed bulk densities three times higher than conventional ZIF-8.

Zeolitic imidazolate frameworks (ZIFs), a sub-family of metal-organic frameworks (MOFs), are crystalline materials prepared by self-assembly of metal ions and imidazolate organic linkers [Park et al (2006)]. ZIFs adopt zeolitic topologies and display some of the quintessential stability of these classic inorganic materials [Tian et al (2007)]. Their large pore volumes and surface areas, along with the possibility for chemical functionalization, have led to potential applications in gas adsorption, separation and catalysis [Furukawa et al (2013), Bennett et al (2013)]. However, the utility of ZIFs and MOFs in such

applications is currently limited by an inability to process the microcrystalline powders resulting from their synthesis. Such shaping is very important in order to reduce the existence of pressure drops of a gas flow in columns due to powder compaction. In most cases, binders and/or high-pressure processes are used to pelletize the material, though often result in either i) partial or complete collapse of the internal porosity when using high pressures [Chapman et al (2009)], or ii) pore blocking by the binder, preventing the access to the porosity. In addition, the use of a binder per se limits the amount of MOF in the final product, and hence would be expected to lead to reduced total guest capacities.

Despite of the rapid growth of MOF research, only few reports about the development of 'monolithic' structures are available in the literature. Most of the research on

circumventing these problems concentrates on the incorporation of MOFs into porous polymer monoliths [see, for example, Fu et al (2013) and Huang et al (2013)] and open- pore polymer foams [US 2010/0181212], the use of high mechanical pressure, or the use of extrusion processes [Kusgens et al (2010)]. Thus, the work reported in the literature typically does not provide monoliths of the MOF material itself, but instead monolithic supports of another material, on which the MOF is supported, or structures in which polymeric binders are required in order to hold the MOF particles together.

In the current work, we focus on ZIF-8 [Zn(mlm)2] (mlm = 2-methylimidazolate, C4H5N2^") which is a prototypical ZIF with sodalite topology. It contains large pore cavities (about 1 1.6 A diameter) interconnected by small windows (about 3.4 A diameter). Because ZIF-8 has characteristic flexibility, these windows allow guest molecules larger than themselves into the porosity through a concerted 'swinging' motion of the mlm linkers [Fairen-Jimenez et al (201 1 ), and Fairen-Jimenez et al (2012)]. According to Fairen-Jimenez et al (201 1 ), ZIF-8 has a specific BET surface area of 1750 m²/g. According to Park et al (2006), ZIF-8 has a specific pore volume of 0.663 cm³/g.

The preferred embodiments of the invention involve a low-cost method of producing robust monoliths under ambient conditions without the use of binders, high pressures or high temperatures. The monoliths produced possess higher Young's moduli than single crystals of ZIF-8 and, importantly, retain the characteristic porosity of the framework while showing higher bulk densities. Furthermore, the resultant monoliths are transparent and fluoresce (as does ZIF-8), so the work opens up a new pathway for sensing applications. Samples were prepared from a solution of Hmlm (20 ml, 0.395 M) and Zn(N0₃)-6H₂0 (20 ml, 0.049 M) in ethanol, after 2 hours stirring at room temperature. After centrifugation at 5500 rpm for 10 minutes, a white solid was collected and dried by three different methods. White pellets (ZIF-8HT) were obtained from drying under vacuum at 100 °C overnight, whilst transparent monoliths (ZIF-8LT) resulted from room temperature drying overnight. A second transparent sample (ZIF-8LTHT) was obtained through further evacuation of ZIF-8LT at 100 °C overnight.

Zn(N0₃)-6H₂0 (98%) and 2-methylimidazole (97%) were purchased from Alfa Aesar, ethanol (≥99.5%) were purchased from Sigma-Aldrich. All chemicals were used as received.

Solutions of Hmlm (20 ml, 0.395 M) and Zn(N0₃)-6H₂0 (20 ml, 0.049 M) in ethanol were mixed and stirred for 2 hours at room temperature. After centrifugation at 5500 rpm for 10 minutes, a white solid was collected and processed by four different methods. First, ZIF- 8HT (HT = high temperature) was obtained by drying a fraction of the white solid at 100 °C overnight in a vacuum oven. Second, ZIF-8LT (LT = low temperature) was obtained by drying a second fraction of the white powder at room temperature overnight. Third, ZIF-8-LTHT was obtained by further evacuation of ZIF-8LT at 100 °C overnight in a vacuum oven. Finally, ZIF-8ER (ER = extended reaction) was obtained by washing the initial white solid twice in ethanol and by adding new solutions of Hmlm (20 ml, 0.395 M) and Zn(N03)-6H₂0 (20 ml, 0.049 M) in ethanol. The mixture was ultrasonicated for 10 minutes at room temperature, centrifuged at 5500 rpm and dried at room temperature overnight. Figs. 1-4 show optical micrographs at relatively low magnification. Fig. 1 shows ZIF-8HT. Fig. 2 shows ZIF-8LT. It is possible to produce similar monoliths of ZIF-8LTHT. Fig. 3 shows ZIF-8ER under normal illumination and Fig. 4 shows ZIF-8ER under 365 nm UV light.

As shown in Fig. 1 , the white pellets of ZIF-8HT easily disaggregated into a typical white ZIF-8 powder. However, both ZIF-8LT (see Fig. 2) and ZIF-8LTHT remained as substantially transparent monolithic structures. The fact that ZIF-8LT retained its macroscopic monolithic morphology during higher temperature activation (i.e. during the treatment to form ZIF-8LTHT) is remarkable. Fig. 4 shows the fluorescence of ZIF-8ER, which sample is discussed in more detail below. The substantially transparent nature of the preferred embodiments makes the material a perfect candidate for sensing applications.

Fig. 5 shows the powder X-ray diffraction (PXRD) pattern of the different samples. The three samples are identical in crystalline structure, despite the differences in their morphologies. The present inventors investigated these differences using scanning electron microscopy (SEM). Figs. 6, 7 and 8 show SEM micrographs of ZIF-8HT, ZIF-8LT and ZIF-8ER, respectively. Figs. 6, 7 and 8 show that ZIF-8HT presents a significant volume of interstitial spaces between primary particles, associated to pores in the range of the meso- and macroporosity, whereas ZIF-8LT and ZIF-8ER present relatively flat surfaces. The different morphologies at the macro- and micro-scale resemble those previously observed for xerogels and aerogels [Lohe et al (2009), Li et al (2013)], not only for amorphous MOF-like materials [Lohe et al (2009)] but also for carbon and silica aerogels [Fairen-Jimenez et al (2008), Fairen-Jimenez et al Carbon (2006)]. Primary particle sizes of ZIF-8 from the initial mixture of precursors, obtained by transmission electron microscopy (TEM), were around 60-70 nm.

Table 1 below shows the mechanical properties of elastic modulus (Young's modulus) and hardness for different ZIF-8 structures.

Table 1

Elastic Modulus Hardness

Material

GPa GPa

ZIF-8 single crystal [Tan et al (2010)] 2.973 ± 0.051 0.501 ± 0.023

ZIF-8-LT 3.66 ± 0.18 0.417 ± 0.038

ZIF-8-LTHT 3.57 ± 0.22 0.429 ± 0.026

ZIF-8-ER 7.04 ± 0.13 0.643 ± 0.021

The data for the single crystal was obtained in the {1 , 0, 0} facet.

The microcrystalline nature of ZIF-8HT precluded investigation of the Young's modulus, E, and hardness, H, by nano-indentation, though monoliths of ZIF-8 LTHT were of sufficient size to allow characterisation.

Table 1 shows comparable H values to those seen before, though Young's moduli were significantly higher. In some cases, measurements could only be performed on one face of the monoliths because of the small area available on others. In order to use MOF-monoliths in e.g. column beds or fuel tanks, they must have appropriate mechanical properties to support mechanical stresses, which come from the weight of the adsorbent inside the columns and from vibrations or movements of the bed.

Without wishing to be bound by theory, the present inventors consider that the formation of the monolithic structures stems from the existence of small primary particles and the mild drying conditions. The fact that ZIF-8LT and ZIF-8LTHT are transparent and therefore do not show light scattering is presumably related to the absence of electronic contrast between phases [Fairen-Jimenez et al J. Phys. Chem. (2006),] or the existence of primary particles smaller than the light wavelength [Apetz and van Bruggen (2003)]. At the time of writing, the present inventors hypothesise that the existence of residuary reactants (Zn ions and mlm) within the sample and the mild drying process allows extension of the polymerisation reaction and the formation of the monolithic structure. In this case, new ZIF-8 is formed during the drying process of ZIF-8LT at room temperature, acting as a binder of the primary ZIF-8 particles.

To investigate this hypothesis, the present inventors proceeded with the synthesis of a new sample where the initial precipitate, immediately after centrifugation, was included in a new solution of mlm and Ζη(Νθ3)·6Η2θ in ethanol. This mixture was ultrasonicated for 10 minutes at room temperature, centrifuged at 5500 rpm and dried at room temperature overnight. The resulting white but partially transparent monolithic structure was named the extended-reaction sample, ZIF-8ER. As can be seen in Table 1 , ZIF-8ER is significantly more rigid than the previous monoliths and the ZIF-8 single crystal. The high values of E reported can be compared to thin films (i.e. not monoliths) of ZIF-8 (3.5 GPa) prepared by Eslava et al (2012), where the deviation in moduli from single crystals was assigned to surface roughness effects.

Notable differences in E between thin films of HKUST-1 [Bunschuh et al (2012)] (9.3 GPa and 3.5 GPa) have been noted before and ascribed to elastic anisotropic effects.

The porosity of the prepared samples was analysed using N2 adsorption at 77 K.

Figs. 9A and 9B show the results in a semi-logarithmic and linear scale, respectively. The data points in Figs. 9A and 9B are: ZIF-8-LT - squares; ZIF-8-HT - triangles; and ZIF-8- LTHT - diamonds. Note the use of semi-logarithmic scale allows more detail to be seen for the low pressure range. Table 2 reports the main results. For comparison with Figs. 9A and 9B, the theoretical single crystal capacity would be represented by a horizontal line at about 420 cm³/g STP. The effect of the density differences between samples on the volumetric adsorption is very significant. First, the low density of powder ZIF-8 means that the volumetric capacity, BET volumetric area and micropore volume are very low. Then, the monolithic materials prepared here present an outstanding enhancement of the conventional, powder ZIF-8, with values more than 3 times higher: 1660 vs.485 m² cm^-3 for ZIF-8ER and powder ZIF-8, respectively, due to the high densities. The fact that the volumetric adsorption capacity is higher than the theoretical single crystal capacity, which is calculated from the 18 mmol g^_1 N2 capacity and a crystal density of 0.95 g cm^-3, could be related to the existence of issues when calculating bulk densities or the existence of impurities. Efforts towards MOF densification have been addressed before for MOF-177 [Zacharia (2010)], where the density of MOF-177 increased from 0.1 g cm^-3 up to 1.40 g cm^-3. However, in all the cases the volumetric capacities were below the theoretical single crystal capacities. The fact that the maximum volumetric capacity for MOF-177 was obtained for pellets with a density of 0.53 g cm^-3 before decreasing for higher preparation pressures suggest a gradual amorphization when using higher pressures, causing the collapse of the porosity. This was indeed confirmed by XRD studies on the pellets obtained at very high pressures.

All the samples presented the typical step-wise adsorption mechanism of N2 in ZIF-8, which indicate the samples were indeed microporous. In addition, ZIF-8-HT showed higher adsorption at higher pressures, close to saturation pressure, which is consistent with the existence of meso- and macroporosity observed in the SEM. The gravimetric BET areas were around 1390 m²/g, whereas gravimetric BET areas of ZIF-8 are generally in the range of 1300-1600 m²/g [Song et al (2012)], meaning that the monolithic materials retained the characteristic porosity of ZIF-8. Moreover, volumetric adsorption capacities (i.e. the amount of gas that can be adsorbed per cm³ of a specific material) and volumetric BET areas, which can be obtained by multiplying gravimetric data with bulk density of the sample, are especially important from an applied point of view in most industrial applications when the adsorbent material has to be confined in a fixed given volume.

These differences were studied by measuring the density of the samples using mercury porosimetry. Since mercury does not penetrate the porosity of the materials at atmospheric pressure, it allows the measurement of the bulk density of the samples by applying Archimedes' method, which in turn facilitates calculation of their bulk densities. Table 2, below, shows data for BET area (SBET), micropore volume (Wo), meso-pore volume (V2) (note that N2 adsorption analysis typically probes porosity only up to 50nm), total pore volume (V-rot) and bulk density (pb) for the different ZIF-8 structures.

Table 2

SBET Wo^a v₂ V_Tot^b Pb^c SBET(VOI) Wo(vol) V₂(vol) Viot(vol)

Material

m²/g cm³/g cm³/g cm³/g g/cm³ m²/cm³ cm³/cm³ cm³/cm³ cm³/cm³

ZI F-8HT 1387 0.552 0.277 0.829 0.35^d 485 0.193 0.097 0.29

ZI F-8LT 1359 0.532 0.01 1 0.543 1 .14 1549 0.606 0.013 0.619

ZI F-8LTHT 1423 0.543 0.003 0.546 1.05 1494 0.570 0.003 0.573

ZI F-8ER 1395 0.535 0.010 0.545 1.19 1660 0.637 0.012 0.648

ZI F-8 single

1706 0.74 - 0.74 0.95 1620 0.70 - 0.70 crystal ^e ^a Obtained at P/P₀ = 0.1 ^b Obtained at P/P₀ = 0.98 ^c Bulk density quantified by measurement of weight and volume using mercury porosimetry ^d ZIF-8 tap bulk density as reported by BASF ^e Reference Fairen-Jimenez et al [201 1]

The single crystal density of ZIF-8 is high (about 0.95 cm³/g). However, of course in a particulate format, the inter-particle space takes up a substantial portion of the bulk volume of a powder material. In the case of ZIF-8, it is expected that the inter-particle space results in the powder having a tap bulk density of about 50% of the single crystal density [Juan-Juan et al (2010)]. Indeed, commercial ZIF-8 from BASF (Basolite® Z1200) presents a bulk density of 0.35 g/cm³ [see

URL: httpi//www.siqmaaldrichxom/cataloq/product/aldrich/891348?[anq=en&reqion=GB accessed 10 June 2014]. In contrast, all the monolithic structures of the preferred embodiments revealed very high bulk densities, which is especially important from an applied point of view when the adsorbent material has to be confined in a given volume. The fact that measured densities are higher than the crystal density of ZIF-8 suggests the presence of denser, amorphous phases or a non-complete activation, and therefore slightly lower gravimetric surface areas. When translating the BET areas and micropore volumes into volumetric, the monolithic materials present more than 3 times higher values than conventional, powder ZIF-8. The stability of the monolithic ZIF-8ER was tested in water at 100 °C for 7 days. Fig. 39 shows the PXRD patterns of the samples at 3, 5 and 7 days. After being immersed in boiling water, ZIF-8ER was able to keep the monolithic morphology and the crystalline structure of ZIF-8, similar to previous reported data for standard, powder ZIF-8. In the examples reported above, the samples were manufactured without the deliberate addition of components other than solvent and the components needed to form the MOF. In other embodiments, it is possible to include other components. Such other components may be included, for example, to increase the meso- or macro-porosity of the monoliths, where that is wanted for a particular application. In this case, it is typical that the composition of the monolith cannot be considered to be equivalent to the composition of a MOF single crystal. Instead, the composition of the monolith can be considered to be equivalent to the composition of a MOF single crystal and one or more remaining components of the composition (i.e. the additives). In order to make a fair assessment of the properties of the monolith for a particular property (e.g. BET surface area, porosity, pore size distribution, Young's modulus, hardness, etc.), the property of the monolith is compared with a volumetric weighted arithmetic mean of the corresponding property of the MOF single crystal and said remaining components. Thus, for example, when the composition of the monolith can be considered to be 80% by volume of a MOF material capable of forming a MOF single crystal having Young's modulus Eo and 15% by volume of a first additive having Young's modulus Ei and 5% by volume of a second additive having Young's modulus E2, then the volumetric weighted arithmetic mean of the Young's modulus of the MOF single crystal and the remaining components is:

[0.8Eo + 0.15Ei +0.05E₂]. Fig. 10 shows the results of thermogravimetric analysis (TGA) of the different samples. This provides information on the stability of the samples. Fig. 10 shows weight losses of 12, 7 and 8 % between 150-300 °C for ZIF-8LT, ZIF-8LTHT and ZIF-8ER respectively. These losses are attributed to the residual ethanol and water in the materials. ZIF-8-HT shows about 4 % weight loss between 150-300°C, indicating that the residual solvent molecules in monoliths are more difficult to be removed than in powders. A second weight loss step due to thermal degradation is observed at 600 °C for all three samples, which is consistent with the previous literature reports [e.g. Park et al (2006)].

In summary, disclosed above is the synthesis of transparent monoliths of ZIF-8 by a one step process using mild conditions. Monolithic materials retained the characteristic porosity of ZIF-8 while showing bulk densities three times higher than conventional ZIF-8. In addition, samples were substantially more rigid than single crystals of the same composition. All these characteristics make the reported process ideal for industrial applications where optimal materials need to present high volumetric adsorption capacities and satisfactory mechanical properties. MOF layers/coatings

The same synthesis principle as for the monolithic MOFs reported above has been used to create MOF layers (also called coatings here) coatings on substrates. The synthesis method of the ZIF-8 coating, named here ZIF-8N, was similar to that of ZIF-8LT reported above. After 2 hours reaction under stirring, the white solution was allowed to settle for 30 minutes. After most of the white solids precipitate, 30 ml of the supernatant were carefully removed using a pipette. The rest of the solvent as well as the solids were poured into a petri dish and dried for 24 hours at ambient conditions. The resultant ZIF-8 coating formed on the base of the petri dish (the substrate) was firmly attached to the petri dish (it was difficult to remove it) and was substantially transparent. A suitable determination of substantial transparency applicable to layers of embodiments of the present invention can be made as based on the approach set out with respect to the fifth or sixth aspect of the invention.

Examples based on other MOF compositions

The work reported above relates primarily to ZIF-8. Now disclosed are applications of different MOF compositions. The present example relates to the MOF material ZIF-zni. Switching the organic linker from 2-methylimidazole (the one used in ZIF-8) to imidazole allows the preparation of ZIF-zni. In the synthesis method here, a ZIF-zni monolith was obtained using a similar method to that of ZIF-8 MOF.

1 ml of NaOH (1 M) (in order to improve the deprotonation of the organic linker) was added to a solution of imidazole (20 ml, 0.395 M) in ethanol. It was then mixed with a solution of Zn(N03)-6H₂0 (20 ml, 0.049 M) in ethanol, and stirred for 1 hour at room temperature. After centrifugation of the solution at 5500 rpm for 10 minutes, a white solid was collected and dried at room temperature overnight to form a monolith. In this case, a substantially identical monolithic material was observed as for ZIF-8ER.

Fig. 29 shows the PXRD pattern of the monolith (upper PXRD trace), confirming that the structure was ZIF-zni by comparison with a known ZIF-zni PXRD trace (lower PXRD trace). The inset shows a view of the monolith, in a similar manner to Figs. 2 and 3.

MOF(g⁾ZIF-8 composite monoliths

In order to combine the properties of MOFs with, for example, different selectivity, different hydrophilicity / hydrophobicity, etc., a series of hydrophilic MOFs were prepared. These were UiO-66, MIL-101 and ZIF-90. These were subsequently embedded in a matrix of hydrophobic ZIF-8, in different MOF:ZIF-8 proportions, where ZIF-8 is working as a binder. In this way, composite MOF monoliths were formed. This approach can be considered to be a generalisation from the special case of the extended reaction samples discussed above.

UiO-66 was synthesised by using the method reported by Katz et al. (2013). 0.75 g of zirconium chloride were dissolved in 30 ml DMF with 6 ml HCI (37%). A solution of 0.738 g of terephthalic acid in 60 ml DMF was then added. The mixture was heated at 80°C overnight. The obtained solid was washed with hot DMF (70°C, 100 ml, 3 times) and ethanol (20 ml, 3 times) respectively, and dried at 80°C under vacuum. MIL-101 was synthesised following the procedures from Khan et al. (201 1 ). 0.532 g of CrC -ehbO, 0.332 g of terephthalic acid and 20 ml of deionized water were mixed in a 45 ml autoclave. The mixture was ultrasonicated for 20 minutes before being heated at 210°C for 24 hours. The obtained solid was washed in hot DMF (70°C, 100 ml, 3 times) and ethanol (20 ml, 3 times) respectively, and dried at 80°C under vacuum.

ZIF-90 was synthesised following the method reported by Shieh et al. (2013). 0.48 g of imidazole-2-carboxaldehyde (ICA) and 0.5 g of polyvinylpyrrolidone (PVP, MW: 40000) were dissolved in 12.5 ml of deionized water. A solution of 0.371 g of Zn(N03)-6H20 in 12.5 ml of ethanol was then added. The mixture was stirred at room temperature for 3 minutes. The solid obtained after centrifugation was washed 3 times with ethanol and dried at 80°C under vacuum.

In a typical reaction to form a MOF composite, 0.04 g of UiO-66, MIL-101 or 0.02 g of ZIF- 90 were dispersed in 40 ml of ethanol, respectively, and ultrasonicated for 20 minutes. For each MOF solution, 0.649g of 2-methylimidazole was added and ultrasonicated for 3 additional minutes until it was all dissolved. Then, 0.293 g of Zn(N03)-6H20 were added. The mixtures were ultrasonicated for 10 minutes and the obtained solids were washed in ethanol (20 ml, 3 times) under ultrasonication for 3 minutes and dried at room temperature overnight. The derived samples were labelled as UiO-66@ZIF-8 monolith, MIL-101 @ZIF- 8 monolith and ZIF-90@ZIF-8 monolith respectively.

Fig. 30 shows the PXRD pattern of UiO-66, MIL-101 , and ZIF-90 compared with simulated patterns. The agreement indicates the success in the synthesis of the MOFs. Figs. 31 -33 show the SEM images of the synthesised UiO-66, MIL-101 and ZIF-90, respectively. UiO-66 has a particle size of 100-150 nm; MIL-101 has a size of 400-500 nm; and ZIF-90 has a size around 2 μηη. The MOFs are synthesised as small particles to inhibit the precipitation during the synthesis of MOF@ZIF-8 monoliths.

Figs. 34 and 35 show SEM images of the cross-section of the interior of a MIL-101 @ZIF-8 and a ZIF-90@ZIF-8 monolith composite, respectively. In the case of MIL-101 @ZIF-8 there are some holes on the cross-section as well as some large particles isolated on the flat surface. The diameter of the holes and particles are around 400-500 nm, similar to the size of the synthesised MIL-101 particles. In the case of ZIF-90@ZIF-8 the effect is even clearer, with particles of size about 2 μηη - similar to the size of the synthesised ZIF- 90 particles - embedded in a matrix.

Figs. 36-38 Table 3 show the N2 adsorption isotherms and BET areas for different composites, respectively. As showed by the N2 isotherms, the porosity properties of the composites are a combination of ZIF-8 and the included MOF. However, the BET areas of the composites are reduced compared with the theoretical value. For UiO-66@ZIF-8 and MIL-101 @ZIF-8 composites, the experimental BET areas are around 500 m²/g lower than the theoretical values. In the case of the ZIF-90@ZIF-8 composite, the BET area is only 250 m²/g lower than the theoretical value. It is speculated, without wishing to be bound by theory, that this might be because of pore blocking effects.

Table 3. BET areas of different MOFs and MOF-ZIF8 composites

Material SBET

m²/g

ZIF-8 LT Ϊ387 Material SBET

m²/g

UiO-66 1 173

MIL-101 2426

ZIF-90 1036

UiO-66@ZIF-8 710

MIL-101 @ZIF-8 1455

ZIF-90@ZIF-8 946

Table 4 shows the mechanical properties of UiO-66@ZIF8 and MIL-101 @ZIF-8. Both elastic modulus (E) and hardness (H) of the MOF@ZIF-8 composites are of the same order as those of ZIF-8ER.

Table 4 Mechanical properties of different MOF-ZIF8 composite and ZIF-8-ER with indentation depth of 500nm.

Materials Elastic modulus (E) Hardness (H)

GPa GPa

ZIF-8ER 7.04 ± 0.13 0.643 ± 0.021

UiO-66@ZIF-8 6.864 ± 0.498 0.490 ± 0.052

MIL-101 ©ZIF-8 6.915 ± 0.394 0.531 ± 0.046

Measurement of porosity and BET specific surface area by N? adsorption N2 adsorption isotherms were recorded at 77 K using a Micromeritics ASAP 2020 instrument. Prior to the measurements, the samples were degassed at 425 K using heating rate of 5 K min^-1 for 4 h.

The BET equation was applied to experimental N2 isotherms using the consistency criteria suggested in the literature and carefully assuring that the BET constant remains positive [Rouquerol et al (1999)]. In the experimental isotherm, the BET equation was fitted over a broad range of pressures.

Nanoindentation

Nanoindentation experiments were performed using an MTS Nanoindenter XP, located in an isolation cabinet to shield against thermal fluctuations and acoustic interference.

Before indentation, monolith surfaces were first cold-mounted using an epoxy resin and then carefully polished using increasingly fine diamond suspensions. Indentations were conducted under the dynamic displacement-controlled "continuous stiffness

measurement" mode. E (Elastic modulus) and H (Hardness) mechanical properties were subsequently determined as a function of the surface penetration depth. A 2-nm sinusoidal displacement at 45 Hz was superimposed onto the system's primary loading signal, and the loading and unloading strain rates were set at 5x10^~2 s^~1. All tests were performed to a maximum indentation depth of 1 ,000 nm using a Berkovich (i.e. three- sided pyramidal) diamond tip of radius about 100 nm. The raw data (load-displacement curves) obtained were analysed using the Oliver and Pharr (2004) method, and Poisson's ratio set at 0.2, in accordance with prior work on zeolitic imidazolate frameworks [Tan et al (2010)]. Data resulting from surface penetrations of less than 100 nm were discarded due to imperfect tip-surface contacts.

For ZIF-8LT, 45 indents of 1000 nm were performed. For illustration, 10 overlaid load- depth curves are shown in Fig. 12, showing the consistency of behaviour for the ZIF-8LT sample. Fig. 13 shows an SEM image of two rows of 1000 nm indents made on a sample (5 x 5 mm) of ZIF-8LT. Fig. 14 shows results for the elastic modulus of ZIF-8LT as a function of indentation depth. Fig. 15 shows results for the hardness of ZIF-8LT as a function of indentation depth. In each of Figs. 14 and 15, each error bar arises from the standard deviation of 45 indents.

Fig. 16 shows 20 overlaid load-depth curves for the ZIF-8LTHT sample. Fig. 17 shows an SEM image of two rows of 1000 nm indents made on a sample (5 x 5 mm) of ZIF-8LTHT. Fig. 18 shows results for the elastic modulus of ZIF-8LTHT as a function of indentation depth. Fig. 19 shows results for the hardness of ZIF-8LTHT as a function of indentation depth. In each of Figs. 18 and 19, each error bar arises from the standard deviation of 45 indents.

For ZIF-8-ER, two nanoindentation analyses were carried out. In the first, a monolith of ZIF-8ER was subjected to 6 indents of depth 3000nm. In the second, another monolith of ZIF-8ER was subjected to 15 indents of depth 1000nm.

Fig. 20 shows 6 overlaid load-depth curves for the ZIF-8ER sample indented to 3000nm. Fig. 21 shows a 3000 nm indent made on a sample (5 x 5mm) of ZIF-8ER. Fig. 22 shows results for the elastic modulus of ZIF-8ER as a function of indentation depth. Fig. 23 shows results for the hardness of ZIF-8ER as a function of indentation depth. In each of Figs. 22 and 23, each error bar arises from the standard deviation of 6 indents.

Fig. 24 shows 15 overlaid load-depth curves for the ZIF-8ER sample indented to 1000nm. Fig. 25 shows rows of 1000 nm indents made on a sample (5 x 5mm) of ZIF-8ER. Fig. 26 shows results for the elastic modulus of ZIF-8ER as a function of indentation depth. Fig. 27 shows results for the hardness of ZIF-8ER as a function of indentation depth. In each of Figs. 26 and 27, each error bar arises from the standard deviation of 15 indents.

Further experimental details

Powder X-ray diffraction (XRD) patterns were recorded with a Bruker D8 diffracto meter using CUKCH (λ=1.5405 A^-1) radiation with a step of 0.02° at a scanning speed of 0.1 ° s^"1.

Scanning electron microscope (SEM) images were taken by FEI XL30 FEGSEM with an accelerating voltage of 5 kV.

TEM images were obtained using a FEI Tecnai G2 with a 200 kV voltage. 1 ml of the mother solution was taken and diluted 10 times by ethanol before centrifugation. 50 μΙ of the solution was dripped on a copper grid. The TEM image shown in Fig. 1 1 was taken after the ethanol evaporated at room temperature.

Thermogravimetric analysis (TGA) was performed using a Pyris 1 TGA under N2 atmosphere, from room temperature to 750 °C, using a ramp rate of 10 °C min^-1.

Mercury porosimetry was obtained up to a final pressure of 2000 bar using an AutoPore IV 9500 instrument. With this technique, the volume of the pores greater than 100 nm and the bulk particle density at atmospheric pressure were obtained. Fig. 28 shows the results of mercury porosimetry showing pore size distribution of the macro- and mesoporosity of ZIF-8LT and ZIF-8ER. The chemical stability of ZIF-8ER was tested in refluxing water at 100 °C for seven days. The stability was monitored using XRD every 48 hours from day 3.

HKUST-1 and related work

A critical bottleneck for the use of natural gas as a transportation fuel has been the development of materials capable of storing it in a sufficiently compact form at ambient temperature. The U.S. Department of Energy (DOE) set in 2012 the ambitious volumetric storage target for adsorbed natural gas (ANG) to 263 cm³ (STP)/cm³ at room temperature and 65 bar, equivalent to the storage capacity of an empty tank at 250 bar - targets which, prior to the present disclosure, have not been met. Since the highest reported values were of 180 cm³ (STP)/cm³, it has so far been unclear whether a material able to reach DOE targets could even be developed. Here we report the synthesis of a porous monolithic metal-organic framework (MOF), which after successful packing and densification reaches 267 cm³ (STP)/cm³ capacity, surpassing the DOE targets. Notably, this is the highest value reported to date for porous solids in a conformed shape, and a >50% improvement over any previously reported experimental value. Nanoindentation tests on the monolithic MOF showed robust mechanical properties with a hardness at least 130% greater than that previously measured in its conventional MOF counterparts. Natural gas (NG), mainly composed of CH₄, has long been considered as a preferable energy alternative to traditional fossil fuels due to its high hydrogen to carbon ratio and lower CO2 emissions. However, the low energy density of CH₄ compared with traditional fossil fuels restrains its on-board applicability. A long-standing challenge has been to design storage systems that efficiently and safely store CH₄ at a realistic volume and that allow it to be easily extracted at reasonable pressures and temperatures. Achieving DOE targets is critical for use in wider vehicular/transportation applications or NG transoceanic shipping.

From all the existing adsorbents, MOFs, obtained by the self-assembly of metal clusters and organic linkers, are arguably the most promising class of CH₄ storage materials due to their large surface areas and pore volumes [He et al (201 14]. MOFs are one of the most exciting advances in recent porous materials science, symbolising the beauty of self- assembled chemical structures and the possibility of modifying their individual chemical and physical properties. From all the multiple structural possibilities, a careful

examination of 137,953 different materials using molecular simulations have predicted a maximum ChU adsorption capacity of 267 cm³ (STP)/cm³ [Wilmer et al (2012)]. However, volumetric capacities obtained computationally are almost always calculated using the ideal single-crystal density of MOFs and, due to the existence of packing issues in real scenarios, these values are, in the best case, only theoretical [Mason et al (2014) and Casco et al (2015)].

In a recent work, Peng et al. (2013) studied experimentally the effect of MOF shaping and densification on CH₄ adsorption. They identified HKUST-1 [Cu₃(BTC)2(H₂0)₃] (BTC = 1 ,3,5-benzenetricarboxylate) as the only material capable of achieving the volumetric DOE target - again, using the theoretical crystal density of the MOF (i.e. 0.883 g/cm³) - and with a maximum uptake of 270 cm³ (STP)/cm³ at 65 bar. However, when HKUST-1 was experimentally packed and densified, the volumetric adsorption capacity was reduced down to 180 (STP) v/v (i.e. 35% loss compared to the theoretical maximum value) because of the partial mechanical collapse of the internal pore structure.

Densification and pelletization of MOFs is indeed one of the main challenges for MOF applications in industry, since conventional synthesis methods produce MOFs as powders with very low packing density, generally 3-4 times lower than theoretical crystal one [Tian et al (2015)]. In the present disclosure, we aimed to determine whether there exists a plausible synthetic protocol for a MOF that would meet the DOE volumetric targets. We used our recent developments in advanced synthesis, engineering and densification of MOFs to produce pure monolithic structures of up to about 1 cm³ size without using high pressures or binders [Tian et al (2015)]. We focus here on HKUST-1 as the - theoretically - top- performing MOF for ChU adsorption synthesized to date, but the present disclosure is not necessarily limited to this material. The principles of the present disclosure may be applied to any suitable MOF composition and may find application in the adsorption and storage of any suitable gas composition. The synthesis of the high-density _m0noHKUST-1 (i.e. monolithic HKUST-1 ) described here follows a sol-gel process similar to the synthesis of organic and inorganic aero/xerogels [Tan et al (2006), Fairen-Jimenez et al (2008 and 2006), Pekala (1989) and Dorcheh et al (2008)]. Fig. 40 shows the proposed synthetic mechanism followed in this work, Fig. 41 shows an optical image of monoHKUST-1 and Fig. 42 shows the PXRD patterns of the samples. After the formation of the crystalline, primary MOF particles at the beginning of the reaction, the mother solution was centrifuged and the resulting densified solid, i.e. the gel, was washed to remove at least some unreacted precursors. We found the drying process to be important for the final morphology of the material. On the one hand, if the dense gel was dried at high temperature, the fast removal of the solvent from the interstitial spaces between primary particles does not allow maintaining the gel macrostructure, and therefore only a powder was obtained (named here _p0wdHKUST-1 ). On the other hand, if the dense gel is dried at mild conditions (e.g. room temperature), the retained precursors start nucleating at the interface, experiencing an epitaxial growth within the existing primary particles. In this way, the MOF, a polymer, acts as a binder, closely connecting the existing primary particles together, and leading to a dense, glassy- looking monolith.

Elemental analysis for experimental monoHKUST-1 and _p0wdHKUST-1 samples did not show any significant differences between them and theoretical composition of hydrated HKUST-1 , with ca. 1 water molecule per Cu atom. This is demonstrated in Table 5 below.

Table 5: Elemental analysis of monoHKUST-1 compared with calculated HKUST-1.

Sample % C % H % 0 % Cu monoHKUST-1 33 3 37 27 m_o/loHKUST-1 (repeat) 35 2.5 33.5 29 powc_fHKUST-1 32 3 37 28

Calc. HKUST-1 36 1 32 31

Calc. HKUST-1 - hydrated 33 2 36 29

When comparing the composition with the molecular formula of HKUST-1 , it is clear that the analysed materials were hydrated, with ca. one molecule of water per Cu atom. A repeated experiment on a dry sample was closer to the calculated composition of HKUST- 1 .

High-resolution PXRD analysis and Pawley fitting showed that the crystalline phase of both _mo/loHKUST-1 (Fig. 42(b)) and _p0wdHKUST-1 (Fig. 42(a)) was the same as the predicted HKUST-1 single-crystal. No extra crystalline phases were observed although at this point the existence of amorphous phases inside the MOF primary particles or acting as a binder cannot be discarded.

FTIR showed essentially identical peaks in both _m0noHKUST-1 and _p0wdHKUST-1 samples (Fig. 58), indicating that there were no new chemical functionalities in the monolith. A careful examination of TEM images showed that the powder sample is formed by an aggregation of primary particles of ca. 51 nm size (Fig. 59(a) and Figs. 60(a)-60(c)), whereas the monolithic sample is made of a continuous phase where the primary particles cannot be observed any more (Figs. 61 and 62), i.e. there is no boundary or interphase between primary particles. Electron diffraction also showed same results for powder and monolithic samples (Fig. 63).

Both the powder and the monoHKUST-1 samples show some bright spots of ca. 5 nm uniformly distributed in the samples (Fig. 62). These bright spots have been observed previously in HKUST-1 samples. EDX elemental analysis (Fig. 61 ) in both _p0wdHKUST-1 and monoHKUST-1 confirmed the existence of Cu (mainly), O and C in the bright spots (Fig. 61 ). We attribute these spots to denser, non-crystalline defects probably caused by the fast synthesis of HKUST-1 - this would explain the higher density of the monolith.

Importantly, both the powder and the monolith present them, and therefore cannot be related to the "binder" itself.

We have identified three key factors in the synthesis of the monolithic MOF. First, the primary particles of the MOF are preferably small (roughly, below 200 nm, and specifically they are here shown to be about 51 ± 10 nm, see Fig. 59(a)) [Tian et al (2015)]. This was also suggested by Horcajada et al. (2009) when preparing MIL-89 aero/xerogels. Second, the nucleation and crystal growth processes between MOF primary particles during the drying process are preferably fast under the selected conditions. If the conditions are unfavourable, the weak interactions will induce a mismatched growth, resulting in a noncrystalline gel [Li et al (2013)]. Third, the drying process is preferably achieved under mild conditions, i.e. typically low temperature (lower than or not more than 40°C) and avoiding vacuum. With a slow drying process, it is possible to avoid the mechanical stress at the vapour-liquid meniscus interface of the solvent inside the porosity, and to get a dense monolithic structure instead of a powdered MOF.

To probe the synthesis mechanism, we prepared two additional HKUST-1 samples with larger particle size: 73 ± 18 and 145 ± 60 nm (Figs. 59(b) and 59(c)), and then dried the samples at different temperatures in the range of 20 to 80°C. Table 6 shows the effect of particle size and drying temperature on the formation of either the monolith or the powder. We found that samples with particle size of 51 nm can be successfully dried up to 40°C in order to get the monolithic structure. On the other hand, samples with particle size of 73 nm can only be successfully dried up to 30°C, whereas samples with particle size of 145 nm cannot conform the monolithic structure at any temperature.

Table 6: Effects of particle size and drying conditions on the morphology of obtained materials. M represents monolithic structure, P represents powdered structure, M/P represents partial monolithic partial powdered structure.

Drying Temp (°C) Particle size (nm)

51 73 145

20 M M P

30 M M P

40 M M/P P Drying Temp (°C) Particle size (nm)

50 M/P M/P P

60 P P P

70 P P P

80 P P P

During activation of _m0noHKUST-1 at high temperature, we found that the sample was able to retain the macroscopic monolithic morphology of the mould, whereas the match in the PXRD with the simulated structure confirms the successful synthesis of _m0noHKUST-1 (Fig. 42). Figs. 48 and 49 show the SEM images of the monoliths and corresponding pow_dHKUST-1 (powder HKUST-1 ), respectively. The powdered sample showed significant amounts of interstitial spaces between primary particles, whereas the monoliths presented a flat surface. We found the sizes of the primary HKUST-1 particles that conformed monoHKUST-1 to be of size about 180 nm. The porosity was first evaluated using N2 adsorption at 77 K (Figs. 50-53); it showed a typical Type I isotherm shape, indicative of the microporous character of the monoliths and the absence of any meso-/macroporosity [Tian et al (2015)]. Table 7 compares the bulk densities, gravimetric and volumetric BET areas and pore volumes of monoHKUST-1 with the HKUST-1 powder samples from Peng et al. (2013). In spite of the lower gravimetric BET areas compared with previously reported data (i.e. in the range of 1500-1850 m²/g) [Peng et al. (2013), Kim et al (2013) and Getzshmann et al (2010)], the critical advantage of monolithic MOFs is their high bulk density, and therefore the higher volumetric BET areas, pore volumes and adsorption capacities than traditional powdered counterparts. The measured bulk density (i.e. 1 .08 g/cm³) of monoHKUST-1 was higher than the hand packed and, remarkably, than the crystal densities of HKUST-1 (i.e. 0.430 and 0.883 g/cm³, respectively). The larger density of monoHKUST-1 could be due to the presence of denser phases (e.g. non-porous phases or crystalline defects) or MOF precursors trapped in the porosity, something that is supported by the relatively lower gravimetric BET area.

Table 7. BET areas (SBET), micropore volume (Wo), total pore volume (V_tot) and bulk density (pb) for _m0noHKUST-1 .

SBET Wo^a V_tot^b p_b ^c °^{BET VU|} Vtot(vol) CH₄ uptake . . . . , uptake Materials ^

m²/g cm³/g m³/g g/cm³ m²/cm³ m³/cm³ g/g v/v g/g v/v

_MONOHKUST-1 1 193 0.51 0.52 1 .06 1288 0.56 0.177 259 0.62 344 Peng et al

1850 0.68 0.78 0.43^D 796 0.33 0.21 6 130 N/A

(2013) ^aObtained at P/Po = 0.1 ; ^bobtained at P/Po = 0.99; ^cbulk density quantified by measuring of weight and volume using mercury porosimetry; ^dhand packing density.

Fig. 64 shows the pore size distributions obtained from mercury porosimetry up to 206 MPa (i.e. equivalent to 60 A). The volume of mercury intruded for _m0noHKUST-1 and powdHKUST-1 were 0.037 and 1 .922 cm³/g, respectively. In particular, the volume of mercury intruded for _p0wdHKUST-1 is attributed to the interparticle space rather than any real porosity.

Mercury porosimetry is a well-established method to determine envelope (i.e. bulk) densities and macro- and mesoporosity. In contrast to capillary condensation where the pore fluid wets the pore walls (i.e. the contact angle is <90°), mercury is a non-wetting liquid (i.e. contact angle >90°) that must be forced to enter a pore by application of external pressure. The surface tension of mercury and the interfacial tension between mercury and the solid surface results in mercury bridging the openings to pores, cracks, and crevices until sufficient pressure is applied to force entry. At atmospheric pressure, mercury will resist entering pores smaller than ca. 6 μηη diameter and therefore can be used to calculate envelope (bulk) volume. Therefore, when an object is surrounded by mercury, the mercury forms a closely fitting liquid envelope around the object. Thus, a progressive increase in hydrostatic pressure is applied to enable the mercury to enter the pores in decreasing order of width (i.e. first large macropores, then mesopores). At a pressure of 60,000 psi (414 MPa) mercury has been forced to enter pores of diameters down to 0.003 μηη. In a typical mercury porosimetry experiment, the exact volume of the sample cell is known. The cell containing the activated sample is evacuated and filled with mercury. In the case of a monolithic sample, mercury surrounds the sample, but, at ambient pressure, does not enter small cracks and crevices in the surface smaller than ca. 6 μηη diameter, nor into pores in the structure of the material. Reweighing the filled sample containers and subtracting from this the weight of the empty sample cell plus sample, yields the weight of the surrounding mercury from which the volume of mercury is to be calculated. The fact that mercury presents a high density allows minimising errors in the evaluation of volume from mass. At ambient pressure, the difference in the volume of the empty sample cell and the calculated volume of mercury is equal to the envelope volume of the sample. When increasing the pressure, the mercury will start invading the open pore space, starting with larger macropores and following with mesopores. This allows the evaluation of a pore size distribution of the macro- and mesopore region.

In the case of a powdered sample, the procedure follows essentially the same preliminary steps as when the sample is a single piece (monolith). A powdered sample is a bulk mass of grains, in which the bulk of the sample also contains interparticle space as void volume. Initially, the mercury envelope forms around the bulk mass and not around the individual particles, so the bulk or envelope volume of the entire sample mass is displaced. Only when the pressure is increased will mercury invade the interparticle space and envelope individual particles. A further increase in pressure will force mercury into the voids within the individual particles (i.e. the macro- and mesopores).

To calculate the bulk density of both monoliths and powder samples, we therefore used the volume of mercury displaced at ambient pressure, i.e. before mercury penetrates any kind of interparticle space and/or porosity. Whereas the "particle density" of the monolithic sample can be the exactly the same to the one of a powder particle, the "bulk density" (i.e. the envelop density or mercury density) of the powder will be much smaller than the monolith due to the existence of interparticle spaces. In this work, we used in the same activation conditions for density and gas adsorption: i.e. vacuum oven at 120 ^°C overnight to fully dry the samples. This allows measurement of the weight unequivocally. Note that high temperature and lengthy in situ activation is not necessary for mercury porosimetry since the technique measures the macro- and mesoporosity as well as the envelope volume, which is not affected by the presence of adsorbed gas/moisture in the microporosity. However, high vacuum and long times are required for gas adsorption in order to start the adsorption isotherms at really low pressure (1 E-8 P/Po in the case of N2 isotherms at 77 K). This is common for the characterisation of MOFs and other porous materials, and although some gas molecules (N2, CO2) can be removed during this extra step, we can also confirm that the weight of samples before and after using high vacuum was not affected. In this work, we first activated the samples (vacuum oven, rotatory pump, at 120 °C overnight) before measuring the mass, and then we degassed the samples in situ thoroughly before the mercury porosimetry. Maximum pressure used in the mercury porosimetry was 206 MPa. The density of monoHKUST-1 at ambient pressure was 1.04 ± 0.06 g/cm³.

Fig. 64 shows the pore size distributions obtained from mercury porosimetry up to 206 MPa (i.e. equivalent to 60 A). From the pressure versus intrusion data, the instrument generates volume and size distributions using the Washburn equation. The volume of mercury intruded for _m0noHKUST-1 and _p0wdHKUST-1 were 0.037 and 1 .922 cm³/g, respectively. In particular, the volume of mercury intruded for _p0wdHKUST-1 is due primarily to the interparticle space rather than any real porosity.

Alternative methods for density evaluation include i) tap density, valid for powders, and ii) geometrical density and iii) Archimedes' principle, valid for monolithic samples. However, these methods cannot be used in both monoliths and powders, and therefore the comparison is not straightforward. In our case, when measuring the geometrical density assuming that the monolith had a truncated cone shape, the derived density was 1.26 ± 0.06 g/cm³, higher than the bulk density from mercury porosimetry (1 .04 ± 0.06 g/cm³). We assume these important differences are a result of the uncertainties in the

measurement of height and radii, and the existence of minor cracks that are difficult to take into account in the real shape compared with an ideal body. For Archimedes' principle, we immersed our monolith in silicone oil - not able to penetrate the

microporosity network - with a 0.967 g/cm³ density, allowing the monolith to sink. The sample was preactivated at 120°C under vacuum, overnight, and the weight of the samples was measured before immersion. The obtained density of monolithic HKUST-1 by this method is 1 .10 g/cm³, confirming the high value obtained through mercury.

Taking into account the density measured using the mercury method (1.04 ± 0.04 g/cm³) and the Archimedes' principle using silicon oil (1.10 g/cm³), we get an average value of 1.06 ± 0.05 g/cm³ (Supplementary Table 3);note that we are notincludingthe very large density obtained through the geometrical method (1.26 DO.06 g/cm3) due to the large discrepancies. Table 8: Density of monoHKUST-1 measured by different methods

Mercury - Mercury - Archimedes' Archimedes' Average

Cambridge Micromeritics Principle Principle

Density (g/cm³) 1 .08 1.00 1 .10 1.095 1.06 ± 0.05

Degas overnight 3 h 3 h overnight

conditions^*

^*AII samples were activated at 120^°C and vacuum overnight. The degas conditions only describe the second stage. An additional sample (not included) was degassed under high vacuum at 120^°C overnight to measure any potential changes in the weight; it is confirmed that no changes were observed.

In Figs. 65a and 65b, there is a clear difference in the way the material is packed. powdHKUST-1 is a simple agglomeration of particles with a large amount of interstitial space, whereas monoHKUST-1 surface is much more compact and with minimal amount of interstitial space.

To probe the improved performance of densified MOFs in NG storage, we ran ChU adsorption isotherms at room temperature and up to 70 bar. Fig. 43 compares the absolute volumetric adsorption isotherms of CH4 in _m0noHKUST-1 at 298 K with previous powdered and densified HKUST-1 samples [Peng et al (2013)]. We also included the 263 (STP) v/v target from DOE for benchmark comparison. The gravimetric uptake is shown in Fig. 54 for comparison. It is important to note that the experimentally measured values are excess amounts adsorbed {NE_XC), which are transformed into absolute uptakes (A ) by using equation [1 ]:

N_Abs = N_Exc + pV_pore [1 ]

where p is the density of the gas at the given adsorption pressure and temperature, obtained from the National Institute of Standards and Technology (NIST) [Lemmon et al (2005)], and V_p0re is the pore volume of the adsorbent [Fairen-Jimenez (2012)].

Interestingly, the volumetric CH4 storage capacity of the monoHKUST-1 , i.e. 259 cm³ (STP)/cm³ at 65 bar, virtually matches the DOE target due to the high bulk density of the monolith. To the best of our knowledge, this is the first example of an adsorbent - including MOFs but also other traditional porous materials such as activated carbons and zeolites - that can achieve the DOE target after successful packing [Casco et al (2015)]. Remarkably, the high CH4 adsorption capacity of monoHKUST-1 matches the theoretical, but previously unachievable volumetric CH4 uptake for HKUST-1 when ideal crystal density was assumed [Peng et al (2013)]. When applying pressures in the range of 0.5 to 5 Tons to densify powdered HKUST-1 in previous works [Peng et al (2013)], the MOF density increased but the total pore volume was reduced due to partial collapse of the

MOF structure. As a result, the CH4 capacity was only increased up to 180 cm³ (STP)/cm³ (i.e. 35 % loss compared with the theoretical value without collapse). Overall, we found that monoHKUST-1 shows an enhancement of the CH4 adsorption capacity of, at least, ca. 99% over previously reported experimental values on a powder and 45% over previously reported experimental values on densified powders. Taking into account previous accurate computational models for methane storage [Wilmer et al (2012)], we reasonably believe this value to be, within a small range of statistical error, the physical limit of ambient temperature methane storage capacity in porous materials. Densification of strictly microporous materials has previously usually been found to come at the expense of slower adsorption kinetics. We have measured the kinetics for methane adsorption, and the evolution of the pressure decay with time, in _p0wdHKUST-1 and monoHKUST-1 samples at 5, 10, 20, 40 and 60 bar (Figs. 66 and 67). Interestingly, we noticed no important differences in the adsorption kinetics between the two samples, showing very fast equilibrium between 75 and 200 s for both _p0wdHKUST-1 and monoHKUST-1 . Small differences are due to the absence of mesoporosity in _m0noHKUST-1 and therefore slightly slower transport diffusivity compared with the nanometre-sized particles of _p0wdHKUST-1. Figs. 66 and 67 follow on from Fig. 43. Fig. 66 shows equilibrium time of methane adsorption at 298 K as a function of equilibrium pressure for _m0noHKUST-1 (diamonds) and powdHKUST-1 (circles). Fig. 67 shows the decay of pressure with time, at 40 bar, for methane adsorption for _m0noHKUST-1 (diamonds) and _PovJHKUST-1 (circles). In addition to DOE targets described above, the Advanced Research Projects Agency- Energy (ARPA-E) has also set the ChU deliverable capacity to 315 cm³ (STP)/cm³, where the deliverable capacity is defined as the uptake at the storage pressure of 65 bar subtracted by the uptake at the depletion pressure of 5.8 bar. This ARPA-E ChU deliverable target is often considered too high to be reached [Gomez-Gualdron et al (2014) and Simon et al (2015)]. Indeed, current theoretical - using crystal density - maximum delivery capacities of top-performing adsorbents are around 200 cm³ (STPycm³: e.g. 190, 208 and 183 cm³ (STP)/cm³ for HKUST-1 [Peng et al (2013)], MOF- 519 Gandara et al (2014)], and NU-125 [Peng et al (2013)], etc. [Gomez-Gualdron et al (2014) and Simon et al (2015)]. Taking into account previous packing loses due to MOF densification, all these delivery capacities are in practice likely to decrease down to ca.

135 cm³ (STP)/cm³ assuming a typical 35 % packing loss [Peng et al (2013)]. Our

monoHKUST-1 shows a delivery capacity, using real bulk density, of 172 cm³ (STP)/cm³ (i.e. the difference between 259 and 87 cm³ (STP)/cm³, for the uptakes obtained at 65 and 5.8 bar, respectively). Again, to the best of our knowledge, this is the highest deliverable capacity achieved by any adsorbent after successful pelletization and shaping.

Table 9 shows the methane adsorption isotherm on _m0noHKUST-1 at 298 K.

Table 9: methane adsorption isotherm on monoHKUST-l at 298 K.

Excess adsorption Absolute adsorption

Pressure wt. % Gravimetric Volumetric Methane density Gravimetric Volumetric bar (g/g) (cm³/cm³) g/ml (g/g) (cm³/cm³)

0.36 0.6 0.006 8 0.0002 0.006 9

0.72 1.0 0.010 15 0.0005 0.010 16

1.52 1.9 0.019 28 0.0010 0.019 29

3.08 3.3 0.033 50 0.0020 0.034 52

4.10 4.2 0.042 63 0.0026 0.043 65

5.10 4.9 0.049 74 0.0033 0.051 77

7.10 6.2 0.062 94 0.0046 0.065 98

10.12 7.9 0.079 1 19 0.0066 0.082 124

15.1 1 9.8 0.098 148 0.0100 0.103 156

20.08 1 1 .2 0.1 12 170 0.0134 0.1 19 180

25.04 12.4 0.124 188 0.0169 0.133 201

30.08 13.2 0.132 199 0.0205 0.142 215

39.96 14.1 0.141 213 0.0278 0.155 234

49.88 14.5 0.145 219 0.0353 0.163 246

59.79 14.6 0.146 220 0.0430 0.168 254 Excess adsorption Absolute adsorption

69.76 15.1 0.151 227 0.0510 0.177 267

79.57 15.3 0.153 231 0.0592 0.184 278

89.35 15.4 0.154 232 0.0676 0.189 285

99.38 14.3 0.143 215 0.0761 0.182 275

The exceptional volumetric CH4 storage capacities of the monoHKUST-1 inspired us to extend the study to CO2 adsorption. Fig. 44 shows the absolute and excess uptakes for both volumetric and gravimetric adsorption isotherms for CO2 on _m0noHKUST-1 at 298 K. Fig. 55 shows the excellent match of the gravimetric uptake for CO2 on _m0noHKUST-1 compared with previously reported values [Liang et al (2009)]. Previous highest records for theoretical CO2 volumetric capacity of MOFs were hold by MOF-177 (i.e. 320 cm³

(STPycm³ at 42 bar and 298 K) [Millward et al (2005)], and MIL-101 Llewellyn et al

(2008)] (390 cm³ (STP)/cm³ at 50 bar, 303 K). Again, these studies used ideal crystal densities to calculate volumetric capacities, and therefore it is expected to experience typical packing loses of about 35 % during pelletization [Peng et al (2013) and Marco- Lozar et al (2012)]. In this case, the volumetric capacities are very likely to be reduced to ca. 208 and 255 cm³ (STP)/cm³ for MOF-177 and MIL-101 , respectively. Remarkably, the real CO2 volumetric adsorption capacity of monoHKUST-1 shows an exceptional high value of 344 cm³ (STP)/cm³ at 40 bar and 298 K, surpassing any other porous material reported so far.

High pressure gas adsorption (methane, CH4, carbon dioxide, CO2; oxygen, O2; nitrogen, N₂; Argon, Ar) at 298 K was conducted using an HPVA II from Micromeritics. The

temperature was controlled by using a Julabo F25 HE bath circulator. Prior to the analyses, the samples were activated overnight at 120 °C (vacuum) before measuring the mass, and then degassed in situ thoroughly before the gas adsorption. The results are shown in Figs. 70-74. Once a MOF has been shaped and densified, one of the main challenges that pellets and monolithic structures need to face in industrial settings is the necessity to support mechanical stress from e.g. friction against the tank walls, vibrations within a column, the weight of the adsorbent, external pressurization etc. To determine the mechanical properties of the synthesised monoHKUST-1 , we have measured the elastic modulus and hardness of the monolithic samples using nanoindentation technique (Figs.45-47), and compared them with theoretical calculations [Ryder et al (2016)]. Our results show that the indentation modulus (/) of monoHKUST-1 is 1 1.5 ± 0.4 GPa (Fig. 56), from which we established its Young's modulus (£) to be 9.3 ± 0.3 GPa (taking Poisson's ratio, v = 0.433 from Woll et al) [Bundschuh (2012)]. Interestingly, our current Young's modulus is -15% higher than that recently predicted by density functional theory (DFT, E = 8.1 GPa) [Ryder et al (2016)] for an isotropic polycrystalline HKUST-1 where its isotropic v = 0.45. Another revealing comparison can be made against the mechanical properties of an epitaxially grown HKUST-1 polycrystalline film by Woll et al (i.e. E = 9.3 GPa, H = 198 ± 19 MPa), [Bundschuh (2012)] also measured by nanoindentation method. Most remarkably, while the Young's modulus of our monolithic HKUST-1 is matching the conventional HKUST-1 (see above), [Bundschuh (2012)] the mechanical hardness of HKUST-1 monolith (H = 460 ± 30 MPa) is in fact more than 130% surpassing that of its conventional counterpart. Significantly, this meant that the monolithic version of HKUST-1 has improved mechanical durability against permanent plastic deformation, simply ascribed to its high bulk density (Table 7). Likewise, monolithic HKUST-1 will have a significantly greater yield strength (θγ), because σγ κ H [Tabor (1996)]. Indeed we have carried out atomic force microscopy (AFM) imaging of the residual indents, verifying their good resistance against surface cracking, evidenced in Fig. 47. In addition to providing high mechanical strength, the high bulk density of the monolith is expected to lead to higher thermal conductivity, providing further benefits in practical applications.

Note that in Fig. 56, the average value determined from 200 to 2000 nm is 1 1 .5 ± 0.4 GPa. Each error bar arises from the standard deviation of 60 indents. Note that the indentation modulus, /, was obtained by assuming the sample Poisson's ratio v to be zero. This meant that the values shown here are representing the upper bound of the Young's modulus (£). HKUST-1 single crystal has not being included here due to its high anisotropy in terms of elastic response.

In conclusion, we have synthesised a monolithic MOF, _m0noHKUST-1 , using a sol-gel process and without using binders and/or high pressures. The mild conditions of the synthetic protocol and soft drying process lead to a dense monolithic structure. Small primary particles and a mild drying process are preferred to allow the successful synthesis. monoHKUST-1 was able to retain the characteristic structure and porosity of the powder, while showing 3 times higher density and therefore volumetric gas adsorption capacity. _m0noHKUST-1 showed an outstanding ChU capacity of 267 cm³ (STP)/cm³ at 65 bar, becoming the first conformed adsorbent, after successful densification and shaping, to achieve the volumetric DOE target (i.e. 263 cm³ (STP)/cm³). Taking into account earlier accurate computational models for methane storage, we reasonably believe this value to be, within a very small range of statistical error, the physical limit of ambient temperature methane storage capacity in porous materials. From the mechanical point of view, it is striking to discover that the hardness of monoHKUST-1 is exceeding twice that of a conventional HKUST-1 material published to date. This work represents a significant step forward in the shaping and densification of MOFs, and therefore opens the gate towards their applicability in ANG and other real-world industrial applications where high volumetric adsorption capacities and resilient mechanical properties are critical.

Methods

Materials. Cu(N0₃)2-2.5H₂0 (98 %), BTC (95 %), and ethanol (≥99.5 %) were purchased from Sigma Aldrich and used as received.

Synthesis of HKUST-1 samples. _m0noHKUST-1 was synthesised based on a

modification of the synthesis method of HKUST-1 reported by Kim et al. (2013). Solutions of BTC (10 ml, 0.062 M) and Cu(N0₃)₂-2.5H₂0 (10 ml, 0.064 M) in ethanol were mixed and stirred for 10 min at room temperature. Note that the concentration of particles is twice that found in the ZIF-8 manufacture reported above. After centrifugation, the collected monolithic solid was washed in ethanol (15 ml, 3 times) and then dried at room temperature (20 ± 1 °C) overnight. The solid was transferred to a galls vial and was further dried at 120 °C in an incubator under vacuum overnight. _pcWdHKUST-1 was obtained by drying the washed solid, after centrifugation, at high temperature (120 °C) rather than allowing them to dry first at room temperature. The yields of _m0noHKUST-1 and

powdHKUST-1 were 53 % after activation. Two samples of HKUST-1 with larger particle size were synthesised following the previous method but at 40 °C and 60 °C. The yields were 53 % and 55 %, respectively, after activation. These new samples were dried at different temperatures from 20 to 80°C to provide either a powder or a monolithic sample. Characterisation of _m0n_oHKUST-1. Powder X-ray diffraction (PXRD) patterns were recorded with a Bruker D8 diffractometer using CUKCH (λ = 1.5405 A) radiation with a step of 0.02° at a scanning speed of 8 s per step. IR was performed on PerkinElmer spectrum 100 FT-IR spectrometer. Scanning electron microscope (SEM) images were taken using a FEI XL30 FEGSEM with an accelerating voltage of 5 kV. Transmission Electron

Microscopy (TEM) was carried out using a FEI Osiris S/TEM operated at 200 kV, operated in scanning mode. Elemental analysis in the TEM was performed with a Bruker Super-X EDX detector. C, H, N analysis was performed on an Exeter analytical CE 440 elemental analyser at a combustion temperature of 975 °C while Cu analysis was performed on a Thermo Scientific iCAP 7400 ICP-OES analyser against 1 ppm and 10 ppm standards. Thermogravimetric analysis - mass spectrometry (TGA-MS) was performed using TGA Q500 from TA Instruments in nitrogen, from room temperature to 900 °C, using a ramp rate of 5 °C min^-1 (Fig. 57). An additional TGA-MS was performed using a ramp rate of 5 °C min^-1 up to 120°C, keeping the sample for 8 h at this temperature, and then heated again (ramp rate of 5 °C min^"1) up to 900°C.

N2 adsorption isotherms were undertaken at 77 K using a Micromeritics 3Flex instrument. High pressure CH₄ adsorption at 298 K was conducted using an HPVA II from

Micromeritics. Adsorption kinetics for methane were measured at different pressures by recording the manifold pressure versus time until equilibrium has been reached. The temperature was controlled by using a Julabo F25 HE bath circulator. Prior to all analyses, the samples were activated overnight at 120°C (vacuum) before measuring the mass, and then degassed in situ thoroughly before the gas adsorption.

Mercury porosimetry was obtained up to a final pressure of 2000 bar using an AutoPore IV 9500 instrument from Micromeritics. This technique was used to estimate the particle density of the monoHKUST-1 at atmospheric pressure, as well as the volume of the pores larger than 100 nm. Prior to the analysis, all samples were degassed overnight at 120 °C (vacuum).

Nanoindentation study was performed using an MTS Nanoindenter XP instrument, equipped with a continuous stiffness measurement module. HKUST-1 monoliths were mounted on an epoxy resin (Struers Epofix) and the surface was carefully prepared using established methodology designed for studying MOF crystals. The prepared monolith surface was cleaned with isopropanol and then desolvated at 100 °C. The evacuated sample was secured in a desiccator until testing. A Berkovich diamond tip was used to measure load-displacement data to a surface penetration depth of 2000 nm, from which the hardness values and Young's moduli were derived in accordance to the Oliver-Pharr method [Tan et al (2009)].

N2 adsorption and BET representation

BET area was calculated by Rouquerol's consistency criteria [Rouquerol et al (2013) and Gomez-Gualdron et al (2016)].

Mercury porosimetry

Based on Archimedes' method, it allows measuring the total volume including the volume of the skeleton and the porosity of the MOF samples since mercury at ambient pressure does not penetrate any micro-, meso- or macroporosity. Bulk densities of the samples can be calculated by dividing the mass of the sample by the total volume displaced.

High-pressure adsorption The experimental data of high-pressure gas adsorption was obtained as excess gravimetric adsorption capacity (NE_XC). The gravimetric uptake can be transferred to volumetric capacity by multiplying the bulk density of the adsorbent. The experimentally measured values are excess amounts adsorbed {NE_XC), which are transformed into absolute uptakes (A ) by using equation [1 ]:

N_Abs = N_Exc + pVp_0re [1 ]

where p is the density of the gas at the given adsorption pressure and temperature, obtained from NIST, and V_p0re is the pore volume of the adsorbent.

Nanoindentation experiments

See Fig. 56. The average value determined from 200 to 2000 nm is 1 1 .5 ± 0.4 GPa. Each error bar arises from the standard deviation of 60 indents. Note that the indentation modulus, /, was obtained by assuming the sample Poisson's ratio v to be zero. This meant that the values shown here are representing the upper bound of the Young's modulus (£).

Thermogravimetric analysis (TGA)

Fig. 57 shows thermogravimetric analysis of monoHKUST-1.

TGA-MS of powdHKUST-1 shows that the degradation takes place at the same temperature as for monoHKUST-1 (i.e. 330 °C); the small differences are in the initial stage at low temperature, where povJHKUST-1 and monoHKUST-1 lose ca. 16 and 20 wt.%, respectively (Fig. 68). Mass spectroscopy analysis of the gases show that the initial step is provoked by loss of water and ethanol adsorbed, whereas the 330 °C step is related to the decomposition of the sample, BTC and CO2 (Fig. 69). A further TGA-MS was carried on a monoHKUST-1 sample previously activated at 120 °C under vacuum for 8 hours. This time, after activation the sample was heated up to 120 °C, kept for 8 hours at this temperature, and then heated again up to 900 °C. This experiment tried to replicate the heating activation procedure - with the obvious limitation of the absence of vacuum and the inevitable presence of small amounts of moisture in the chamber of the TGA at room temperature. The TGA showed a very small initial mass loss (ca. 1.5 %) related to water desorption below or at 120 °C (Fig. 68). Mass spectroscopy proved the absence of ethanol after the activation sample, and the successful removal of water at 120 °C without vacuum (Fig. 68).

While the invention has been described in conjunction with the exemplary embodiments described above, many equivalent modifications and variations will be apparent to those skilled in the art when given this disclosure. Accordingly, the exemplary embodiments of the invention set forth above are considered to be illustrative and not limiting. Various changes to the described embodiments may be made without departing from the spirit and scope of the invention.

All references referred to above and/or listed below are hereby incorporated by reference.

List of non-patent document references:

J. Rouquerol, D. Avnir, C. W. Fairbridge, D. H. Everett, J. H. Haynes, N. Pernicone, J. D. F. Ramsay, K. S. W. Sing, and K. K. Unger Pure and Applied Chemistry, vol. 66, p. 1739, 1994.

K. S. W. Sing Pure & Appl. Chem., 4, 1 1 , 2201 -2218, 1982. S. Cao, G. Gody, W. Zhao, S. Perrier, X. Peng, C. Ducati, D. Zhao and A. K. Cheetham Chem. Sci., 2013, 4, 3573-3577

K. S. Park, Z. Ni, A. P. Cote, J. Y. Choi, R. Huang, F. J. Uribe-Romo, H. K. Chae, M.

O'Keeffe, and O. M. Yaghi, Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 10186-91. Y.-Q. Tian, Y.-M. Zhao, Z.-X. Chen, G.-N. Zhang, L.-H. Weng, and D.-Y. Zhao, Chemistry, 2007, 13, 4146-54.

H. Furukawa, K. E. Cordova, M. O'Keeffe, and O. M. Yaghi, Science, 2013, 341 , 1230444.

T. D. Bennett, P. J. Saines, D. A. Keen, J.-C. Tan, and A. K. Cheetham, Chemistry, 2013, 19, 7049-55. K. W. Chapman, G. J. Haider, and P. J. Chupas, J. Am. Chem. Soc, 2009, 131 , 17546-7.

Y.-Y. Fu, C.-X. Yang, and X.-P. Yan, Chem. Commun., 2013, 49, 7162-4.

Hsi-Ya Huang, Cheng-Lan Lin, Cheng-You Wu, Yi-Jie Cheng, Chia-Her Lin Analytica Chimica Acta Volume 779, 2013, Pages 96-103

P. Kijsgens, A. Zgaverdea, H.-G. Fritz, S. Siegle, and S. Kaskel, J. Am. Ceram. Soc, 2010, 93, 2476-2479.

D. Fairen-Jimenez, S. Moggach, M. T. Wharmby, P. Wright, S. Parsons and T. Dijren, J. Am. Chem. Soc, 201 1 , 133, 8900-2.

David Fairen-Jimenez, Raimondas Galvelis, Antonio Torrisi, Alistair D. Gellan, Michael T. Wharmby, Paul A. Wright, Caroline Mellot-Draznieks and Tina Dijrena Dalton

Transactions Issue 35, 2012

M. R. Lohe, M. Rose, and S. Kaskel, Chem. Commun., 2009, 6056-8.

L. Li, S. Xiang, S. Cao, J. Zhang, G. Ouyang, L. Chen, and C.-Y. Su, Nat. Commun., 2013, 4, 1774.

D. Fairen-Jimenez, F. Carrasco-Marin, and C. Moreno-Castilla, Langmuir, 2008, 24,

2820-5.

D. Fairen-Jimenez, F. Carrasco-Marin, C. Moreno-Castilla Carbon Volume 44, Issue 1 1 , September 2006, Pages 2301-2307 J. C. Tan, T. D. Bennett, and A. K. Cheetham, PNAS 2010, 107, 22, 9938-9943.

D. Fairen-Jimenez, F. Carrasco-Marin, D. Djurado, F. Bley, F. Ehrburger-Dolle, and C. Moreno-Castilla, J. Phys. Chem. B, 2006, 110, 8681-8.

R. Apetz and M. P. B. Bruggen, J. Am. Ceram. Soc, 2003, 86, 480-486. S. Eslava, L. Zhang, S. Esconjauregui, J. Yang, K. Vanstreels, M. R. Baklanov, and E. Saiz, Chem. Mater., 2013, 25, 27-33.

S. Bundschuh, O. Kraft, H. K. Arslan, H. Gliemann, P. G. Weidler, and C. Woll, Appl. Phys. Lett, 2012, 101 , 101910.

Q. Song, S. K. Nataraj, M. V. Roussenova, J. C. Tan, D. J. Hughes, W. Li, P. Bourgoin, M. A. Alam, A. K. Cheetham, S. A. Al-Muhtaseb, and E. Sivaniah, Energy Environ. Sci., 2012, 5, 8359.

J. Juan-Juan, J. P. Marco-Lozar, F. Suarez-Garcia, D. Cazorla-Amoros, and A. Linares- Solano, Carbon, 2010, 48, 2906-2909.

Rouquerol, J.; Rouquerol, F.; Sing, K. S. W., Adsorption by powders and porous solids Academic Press: San Diego, 1999.

W. C. Oliver and G. M. Pharr, J Mater Res, 2004, 19, 3-20.

D. Fairen-Jimenez, R. Galvelis, A. Torrisi, A. D. Gellan, M. T. Wharmby, P. a Wright, C. Mellot-Draznieks, and T. Dijren, Dalton Trans., 2012, 41 , 10752-62.

M. J. Katz, Z. J. Brown, Y. J. Colon, P. W. Siu, K. a Scheldt, R. Q. Snurr, J. T. Hupp, and O. K. Farha, Chem. Commun., 2013, 49, 9449-51 .

N. A. Khan, I. J. Kang, H. Y. Seok, and S. H. Jhung, Chem. Eng. J., 201 1 , 166, 1 152- 1 157.

F.-K. Shieh, S.-C. Wang, S.-Y. Leo, and K. C.-W. Wu, Chem. Eur. J., 2013, 19, 1 1 139-42.

R. Zacharia, D. Cossement, L. La · and R. Chahine, J. Mater. Chem., 2010, 20, 2145. He, Y., Zhou, W., Qian, G. & Chen, B. Methane storage in metal-organic frameworks. Chem. Soc. Rev. 43, 5657-5678 (2014). Peng, Y. et al. Methane storage in metal-organic frameworks: current records, surprise findings, and challenges. J. Am. Chem. Soc. 135, 1 1887-1 1894 (2013).

Makal, T. A., Li, J.-R., Lu, W. & Zhou, H.-C. Methane storage in advanced porous materials. Chem. Soc. Rev. 41 , 7761-7779 (2012). Wilmer, C. E. et al. Large-scale screening of hypothetical metal-organic frameworks. Nature Chemistry A, 83-89 (2012).

Mason, J. A., Veenstra, M. & Long, J. R. Evaluating metal-organic frameworks for natural gas storage. Chemical Science 5, 32-51 (2014).

Casco, M. E. et al. High-Pressure Methane Storage in Porous Materials: Are Carbon Materials in the Pole Position? Chem. Mater. 27, 959-964 (2015).

Tian, T., Velazquez-Garcia, J., Bennett, T. D. & Fairen-Jimenez, D. Mechanically and chemically robust ZIF-8 monoliths with high volumetric adsorption capacity. Journal of Materials Chemistry A 3, 2999-3005 (2015).

Tan, G., John, V. T. & McPherson, G. L. Nucleation and growth characteristics of a binary low-mass organogel. Langmuir 22, 7416-7420 (2006).

Fairen-Jimenez, D., Carrasco-Marin, F. & Moreno-Castilla, C. Inter-and intra-primary- particle structure of monolithic carbon aerogels obtained with varying solvents. Langmuir 24, 2820-2825 (2008).

Fairen-Jimenez, D. et al. Surface area and microporosity of carbon aerogels from gas adsorption and small-and wide-angle X-ray scattering measurements. The Journal of Physical Chemistry B 110, 8681-8688 (2006).

Pekala, R. Organic aerogels from the polycondensation of resorcinol with formaldehyde. Journal of Materials Science 24, 3221 -3227 (1989).

Dorcheh, A. S. & Abbasi, M. Silica aerogel; synthesis, properties and characterization. J. Mater. Process. Technol. 199, 10-26 (2008).

Horcajada, P. et al. Colloidal route for preparing optical thin films of nanoporous metal- organic frameworks. Adv. Mater. 21 , 1931-1935 (2009). Li, L. et al. A synthetic route to ultralight hierarchically micro/mesoporous Al (III)- carboxylate metal-organic aerogels. Nature communications 4, 1774 (2013).

Kim, K.-J. et al. High-rate synthesis of Cu-BTC metal-organic frameworks. Chem.

Commun. 49, 1 1518-1 1520 (2013). Getzschmann, J. et al. Methane storage mechanism in the metal-organic framework

Cu3(btc)2: An in situ neutron diffraction study. Microporous Mesoporous Mater. 136, 50-58 (2010).

Lemmon, E., McLinden, M. & Friend, D. Thermophysical properties of fluid systems. NIST chemistry webbook, NIST standard reference database 69 (2005). Fairen-Jimenez, D. et al. Understanding excess uptake maxima for hydrogen adsorption isotherms in frameworks with rht topology. Chem. Commun. 48, 10496-10498 (2012).

Gomez-Gualdron, D. A., Wilmer, C. E., Farha, O. K., Hupp, J. T. & Snurr, R. Q. Exploring the limits of methane storage and delivery in nanoporous materials. The Journal of Physical Chemistry C 118, 6941 -6951 (2014). Simon, C. M. et al. The materials genome in action: identifying the performance limits for methane storage. Energy & Environmental Science 8, 1 190-1 199 (2015).

Gandara, F., Furukawa, H., Lee, S. & Yaghi, O. M. High methane storage capacity in aluminum metal-organic frameworks. J. Am. Chem. Soc. 136, 5271-5274 (2014).

Liang, Z., Marshall, M. & Chaffee, A. L. Comparison of Cu-BTC and zeolite 13X for adsorbent based CO2 separation. Energy Procedia 1 , 1265-1271 (2009).

Millward, A. R. & Yaghi, O. M. Metal-organic frameworks with exceptionally high capacity for storage of carbon dioxide at room temperature. J. Am. Chem. Soc. 127, 17998-17999 (2005).

Llewellyn, P. L. et al. High Uptakes of CO2 and CH₄ in Mesoporous Metal-Organic Frameworks MIL-100 and MIL-101. Langmuir l , 7245-7250 (2008).

Marco-Lozar, J., Juan-Juan, J., Suarez-Garcia, F., Cazorla-Amoros, D. & Linares-Solano, A. MOF-5 and activated carbons as adsorbents for gas storage. Int. J. Hydrogen Energy 37, 2370-2381 (2012). Ryder, M. R., Civalleri, B., Cinque, G. & Tan, J.-C. Discovering connections between terahertz vibrations and elasticity underpinning the collective dynamics of the HKUST-1 metal-organic framework. CrystEngComm (2016).

Bundschuh, S. et al. Mechanical properties of metal-organic frameworks: An indentation study on epitaxial thin films. Appl. Phys. Lett. 101 , 101910 (2012).

Tabor, D. Indentation hardness: fifty years on a personal view. Philos. Mag. A 74, 1207- 1212 (1996).

Tan, J., Merrill, C, Orton, J. & Cheetham, A. Anisotropic mechanical properties of polymorphic hybrid inorganic-organic framework materials with different dimensionalities. Acta Mater. 57, 3481 -3496 (2009).

Rouquerol, J., Rouquerol, F., Llewellyn, P., Maurin, G. & Sing, K. S. Adsorption by powders and porous solids: principles, methodology and applications. (Academic press, 2013).

Gomez-Gualdron, D. A., Moghadam, P. Z., Hupp, J. T., Farha, O. K. & Snurr, R. Q.

Application of consistency criteria to calculate BET areas of micro- and mesoporous metal-organic frameworks. J. Am. Chem. Soc. 138, 215-224 (2016).

H. Jasuja, G. W. Peterson, J. B. Decoste, M. A. Browe and K. S. Walton, Chem. Eng. Sci., 2015, 124, 1 18

Claims

1. A metal-organic framework (MOF) body comprising MOF crystallites adhered to each other via a MOF binder, wherein the MOF crystallites and the MOF binder are formed of HKUST-1 .

2. A metal-organic framework (MOF) body consisting of:

MOF crystallites;

a MOF binder which binds the crystallites together in the body;

optionally, residual solvent; and

wherein the MOF crystallites and the MOF binder are formed of HKUST-1.

3. A MOF body according to claim 1 or claim 2 wherein the MOF crystallites have an average particle size of not more than 300 nm.

4. A MOF body according to any one of claims 1 to 3 wherein the body is a monolith.

5. A MOF body according to any one of claims 1 to 4 wherein the body has a volume of at least 1 mm³.

6. A MOF body according to any one of claims 1 to 5 wherein the body has a volume of at least 10 mm³.

7. A MOF body according to any one of claims 1 to 6 having a smallest linear dimension of at least 1 mm.

8. A MOF body according to any one of claims 1 to 7, further comprising an adsorbed gas stored in the MOF body,

wherein the stored gas is selected from the group consisting of:

CH4, CO2, O2, NH3, Ar, CO, N2 and C2H4 (ethylene); toxic industrial gases such as benzene, toluene, xylenes, sulphur dioxide, ethylene oxide; and warfare agents such as sarin, mustard gas and derivatives thereof, and:

when the stored gas is CH4, the stored gas is present in an amount of at least 200 cm³ (STP) per cm³ of the MOF body at 298K, based on a gas storage pressure of

65 bar, and

when the stored gas is CO2, the stored gas is present in an amount of at least 300 cm³ (STP) per cm³ of the MOF body at 298K, based on a gas storage pressure of 40 bar.

9. A population of bodies according to any one of claims 1 to 8.

10. A process for manufacturing a metal-organic framework (MOF) body, wherein the process includes the steps:

allowing the reaction of MOF precursors in a solvent to form crystallites of the MOF composition; and

forming a body of the MOF composition including a drying stage to remove at least some of the solvent with a maximum temperature during the drying stage of not more than 50°C, the drying stage leaving a MOF binder adhering the MOF crystallites in the body,

wherein the MOF crystallites and the MOF binder are formed of HKUST-1.

1 1. A process according to claim 10, including a step of concentration of particles of the MOF composition into a concentrate of the particles and solvent.

12. A process according to claim 1 1 wherein the step of concentrating the particles of the MOF composition into a concentrate of the particles and solvent is carried out by centrifugation.

13. A process according to any one of claims 10 to 12 wherein the MOF body is a MOF monolith.

14. A process according to claim 13, said monolith being formed into a desired shape by said drying stage taking place with the material in a mould.

15. A method for storage and/or separation of a gas, the method including the steps: providing at least one MOF body;

providing a gas for storage and/or separation; and

exposing the MOF body to the gas for storage and/or separation at a pressure, to allow the gas for storage and/or separation to be adsorbed by the MOF body, wherein the gas for storage and/or separation is selected from the group consisting of: CH4, CO2, O2, NH3, Ar, CO, N2 and C2H4 (ethylene); toxic industrial gases such as benzene, toluene, xylenes, sulphur dioxide, ethylene oxide; and warfare agents such as sarin, mustard gas and derivatives thereof,

and the MOF body comprises MOF crystallites adhered to each other via a MOF binder, wherein the MOF crystallites and the MOF binder are formed of HKUST-1 , and wherein: when the gas for storage and/or separation is CH4, the volumetric storage capacity of the MOF body is at least 200 cm³ (STP) per cm³ of the MOF body at 298K and a gas storage pressure of 65 bar, and

16. A method according to claim 15 wherein the gas for storage and/or separation is CH4, and the volumetric storage capacity of the MOF body is at least 260 cm³ (STP) per cm³ of the MOF body at 298K and a gas storage pressure of 65 bar.