This invention relates to gas-separation membranes (GSMs) and to their preparation and use.

For purifying gaseous mixtures e.g. natural gas and flue gas, the removal of undesired components can in some cases be achieved based on the relative size of the components (size-sieving).

US 8,177,891 describes GSMs comprising a continuous substantially non- porous layer comprising the polymerization product of a compound, which compound comprises at least 70 oxyethylene groups forming an uninterrupted chain of the formula -(CH2CH2O) _n - wherein n is at least 70.

US 8,303,691 describes composite membranes comprising a polymer sheet and a porous support layer for the polymer sheet, characterised in that the polymer sheet comprises at least 60wt% of oxyethylene groups and the porous support layer has defined flux properties.

Kwisnek et al. (Macromolecules 2014, 47, 3243-3253) describes the preparation of GSMs comprising a discriminating layer comprising oxyethylene groups . The GSMs are prepared from monomers in one step without the use of a thioether comprising curable copolymer and without a support.

JP2015160159 describes facilitated transport composite membranes (FTCM) for separating CO2 from CH4 comprising a discriminating layer comprising cross-linked polymers comprising oxyethylene groups and at least one liquid additive. Such liquid additive in FCTM applications usually are added at least 30wt% in discriminating gellayers compositions which has the potential of leaching out during the application of the membranes, possibly resulting in pollution of the retentate stream with said liquid.

There is a need for strong, but flexible GSMs having a high permeability and being capable of discriminating well between gases (e.g. between polar and non-polar gases such as especially H2S and CH4). Ideally such membranes can be produced efficiently at high speeds using roll-to-roll processes. In this manner the membranes could be made in a particularly cost effective manner and without the presence of any defects.

According to a first aspect of the present invention there is provided a gasseparation membrane comprising:

(i) a porous substrate; and

(ii) a discriminating layer; wherein the discriminating layer has an average thickness of 0.05 to 4.9 pm, a sulphur content above 0.001 % relative to the total number of atoms in the discriminating layer and an ethylene oxide content of at least 60 wt%. In this specification the term “comprising” is to be interpreted as specifying the presence of the stated parts, steps or components, but does not exclude the presence of one or more additional parts, steps or components.

Reference to an element by the indefinite article "a" or "an" does not exclude the possibility that more than one of the element(s) is present, unless the context clearly requires that there be one and only one of the elements. The indefinite article "a" or "an" thus usually means "at least one". The discriminating layer is often abbreviated to “DL”.

In this specification ethylene oxide groups and groups of formula - (CH2CH2O)- are sometimes abbreviated to “EO groups”. Thus a group of formula - (CH ₂ CH ₂ O) ₅ - would contain five EO groups.

Preferably the discriminating layer comprises thioether groups. Preferred thioether groups are of the formula -CH2-S-CH2-. Preferably all or substantially all of the sulphur content of the discriminating layer is provided by thioether groups.

According to a second aspect of the present invention there is provide a process for preparing a GSM according to the first aspect of the present invention which comprises the steps of:

(i) applying a curable composition to a porous substrate; and

(ii) curing the composition thereby forming the DL on the porous substrate; wherein the curable composition comprises a curable copolymer which has (a) a sulphur content above 0.001 % relative to the total number of atoms in the curable copolymer; and (b) an ethylene oxide content of at least 60 wt% (often abbreviated herein to the “curable copolymer”).

Preferably the curable copolymer has a number averaged molecular weight above 1 ,800 Da, a sulphur content above 0.001 % relative to the total number of atoms in the curable copolymer and an ethylene oxide content of at least 60 wt%.

In a preferred embodiment the curable copolymer comprises:

(a) thioether groups; and

(b) ethylene oxide groups of the formula -(CH2CH2O) _n -, wherein each n independently has a value of 5 to 22.

Preferably the curable copolymer has a numbered averaged molecular weight (NAMW) above 10,000 Da, more preferably in the range 12,500 Da to 100,000,000 Da, even more preferably in the range 20,000 Da to 50,000,000 Da and especially in the range 25,000 Da to 5,000,000 Da. The NAMW may be determined by gel permeation chromatography, e.g. as described below in more detail.

The value of n affects the polarity of the discriminating layer, with higher values of n resulting in a discriminating layer of higher polarity.

Preferably n has a value of at least 6, e.g. in a preferred embodiment n has a value or 6 to 22, more preferably 7 to 22 and especially 7 to 20. The polarity of the discriminating layer is important for achieving good separation of polar and nonpolar gases, e.g. for the removal of H2S from mixed gas streams.

When n has a value of 23 or higher crystallization of polyethylene glycol) in the discriminating layer can occur which adversely affects the performance of the GSM.

Preferably the curable copolymer has an ethylene oxide content of at least 70 wt% more preferably at least 75 wt% and especially at least 80 wt%.

Preferably the DL has an ethylene oxide content of at least 60 wt%, more preferably at least 70 wt%, especially at Ieast75 wt% and more especially at least 80 wt%.

In our experiments we observed that the NAMW of the curable copolymer can influence how much of the curable copolymer is able to impregnate into the porous substrate and how much will stay on top of the porous substrate to form a discriminating layer external to the porous substrate (after curing). Hence the curable copolymer preferably has a NAMW above 1 ,800 Da.

The NAMW of the curable copolymer is preferably measured by gel permeation chromatography.

Preferably the composition which may be used to form the DL comprises the curable copolymer in an amount of 0.1 to 25wt%, more preferably 0.5 to 20wt% and most preferably 1 .0 to 15.0 wt%, relative to the weight of the composition.

Optionally the composition which may be used to prepare the DL further comprises one or more further monomers, e.g. a crosslinking agent and/or a monomer having only one ethylenically unsaturated group.

Preferred crosslinking agents comprise two or more ethylenically unsaturated groups. The crosslinking agents can be the same as the ones used to prepare the pre-polymer composition (PPC) described below, however the crosslinking agent(s) can also be different.

Preferred ethylenically unsaturated groups include (meth)acrylic groups (e.g. CH ₂ =CR ₄ C(O)- groups), especially (meth)acrylate groups (e.g. CH2=CR4 C(O)O- groups) and (meth)acrylamide groups (e.g. CH2=CR4C(O)NR4- groups), wherein each R4 independently is H or CH3.

Examples of such further monomers having only one ethylenically unsaturated group include polyethylene glycol) (meth)acrylate, allyl (meth)acrylate and vinyl (meth)acrylate.

The amount of such further monomers which may be included in the composition used to form the DL is preferably in the range of 0 to 10wt%, more preferably 0 to 5wt%, and especially 0 to 2wt%, relative to the weight of the composition.

The curing of the curable copolymer together with the optional further monomer provides cross-linking and provides the DL as a cross-linked network on the surface of the porous substrate (with some of the curable copolymer partially impregnated into the porous substrate and cured therein).

The composition which may be used to form the DL preferably further comprises an initiator which facilitates curing of the polymerisable components present in the composition. Any initiator capable of polymerizing ethylenically unsaturated functionalities may be used, e.g. a thermal initiator, photo-initiator a (free-)radical polymerization initiator. Also the curing may comprise inter- and/or intra-molecular polymerization.

Preferred photo-initiators for use in the composition include, but are not limited to organic peroxides, for example ethyl peroxide and benzyl peroxide; hydroperoxides, e.g. methyl hydroperoxide; acyloins, e.g. benzoin; certain azo compounds, e.g. a,a'-azobisisobutyronitrile and Y,Y'-azobis(y-cyanovaleric acid); persulfates; peracetates, e.g. methyl peracetate and tert-butyl peracetate; peroxalates, e.g. dimethyl peroxalate and di(tert-butyl) peroxalate; disulfides, e.g. dimethyl thiuram disulfide; and ketone peroxides, e.g. methyl ethyl ketone peroxide. When the precursor polymer composition comprises a thermal initiator curing is preferably performed at a temperature in the range of from about 30°C to about 150°C, especially from about 40°C to about 110°C. Preferable temperatures will be especially 60 - 90 °C.

Photo-initiators are usually required when the curing uses light, for example ultraviolet (“UV”) light.

Preferred photo-initiators for use in free radical UV curing include, but are not limited to Radical Type I and/or type II photo-initiators.

Examples of radical type I photo-initiators are as described in W02007/018425, page 14, line 23 to page 15, line 26, which are incorporated herein by reference thereto.

Examples of radical type II photo-initiators are as described in WO 2007/018425, page 15, line 27 to page 16, line 27, which are incorporated herein by reference thereto.

A single type of initiator may be used but also a combination of several different types. The most preferred radical initiator is Omnirad™ 1173.

Preferably the composition which may be used to form the DL comprises 0.0001 to 8 wt%, more preferably 0.001 to 5 wt%, and especially 0.005 to 3 wt% of initiator.

Preferably the composition which may be used to form the DL further comprises an inert solvent. Inert solvents are not curable and do not cross-link with any component of the composition.

Preferably the composition to form the DL (or PPC described below) is free of a liquid selected from the group consisting of ionic liquids, glycerin, polyglycerin, polyethylene glycol, polypropylene glycol, polyethylene oxide, and amine compounds having 15 or less carbons. Essential free in the context of this inventions means that the amounts present in the PPC and/or DL and/or GSM compositions are far below 0.1 wt% and preferably below 0.05 wt% and even more preferably below 0.02 wt%.

Examples of inert solvents include alcohol-based solvents, ether-based solvents, amide-based solvents, ketone-based solvents, sulfoxide-based solvents, sulfone-based solvents, nitrile-based solvents and organic phosphorus-based solvents. Examples of alcohol-based solvents include methanol, ethanol, isopropanol, n-butanol, ethylene glycol, propylene glycol, diethylene glycol, dipropylene glycol and mixtures comprising two or more thereof.

Examples of inert solvents include dimethyl sulfoxide, dimethyl imidazolidinone, sulfolane, N-methyl pyrrolidone, dimethyl formamide, acetonitrile, acetone, 1 ,4-dioxane, 1 ,3-dioxolane, tetramethyl urea, hexamethyl phosphoramide, hexamethyl phosphorotriamide, pyridine, propionitrile, butanone, cyclohexanone, tetrahydrofuran, tetrahydropyran, 2-methyltetrahydrofuran, ethylene glycol diacetate, cyclopentylmethylether, methylethylketone, ethyl acetate, y-butyrolactone and mixtures comprising two or more thereof. Preferably the composition used to prepare the discriminating layer is free from water, even when the curable copolymer is water- soluble. However if necessary one may include water in the composition as part of the inert solvent, preferably 10 wt% or less, more preferably 5 wt% or less and most preferably 0 wt%, relative to the total weight of inert solvent present in the composition. Surprisingly omitting water from the composition which may be used to form the discriminating layer results in fewer defects in the resultant discriminating layer.

Especially preferred inert solvents include methyl ether ketone, n-butyl acetate, ethyl acetate, cyclopentyl methyl ether and 2-methyltetrahydrofuran.

The inert solvent optionally comprises a single inert solvent or a combination of two or more inert solvents. Preferred inert solvents include C1-4 alcohols (e.g. methanol, ethanol and propan-2-ol), diols (e.g. ethylene glycol and propylene glycol), triols (e.g. glycerol), carbonates (e.g. ethylene carbonate, propylene carbonate, dimethyl carbonate, diethyl carbonate, di-t-butyl dicarbonate and glycerin carbonate), dimethyl formamide, acetone, methyl ethyl ketone, ethyl acetate, butyl acetate, 2-methyl tetrahydrofuran, cyclopentyl methyl ether, N-methyl-2-pyrrolidinone and mixtures comprising two or more thereof. A particularly preferred inert solvent is n-butyl acetate, methyl ethyl ketone or ethyl acetate. In one embodiment the inert solvent has a low boiling point e.g. a boiling point below 120 °C. Solvents having a low boiling point can be easily removed after curing by evaporation, avoiding the need for a washing step for removal of the solvent. Being inert, the solvent does not copolymerize with any of the other components of the composition.

The amount of inert solvent present in the composition which may be used to form the DL is preferably in the range of 50 to 99.7999wt%, more preferably 60 to 99.0wt%, and most preferably 75 to 98.0wt%, relative to the weight of the composition.

Optionally the composition which may be used to form the DL comprises a surfactant, e.g. 0 to 5wt% and especially 0.1 to 2wt% of surfactant, relative to the total weight of the composition.

Preferred surfactants are non-ionic surfactants, for example ethoxylated and alkoxylated fatty acids, ethoxylated amines, ethoxylated alcohol, alkyl and nonylphenol ethoxylates, ethoxylated sorbitan esters, and EO/PO copolymers, in which EO is ethylene oxide and PO is propylene oxide (-(CH2CH2 CH2O)- ). The most preferred surfactant is a polyether-modified acryl functional polydimethylsiloxane (available as UV-3530 from Byk).

Optionally the composition which may be used to form the DL comprises one or more further adjuvants, binders, etc. (e.g. in an amount of up to 1 .0wt%, relative to the total weight of the composition).

The thickness of the DL is measured outwards from the surface of the porous substrate and does not include any of the DL material present within the porous substrate.

The thickness of the DL is preferably in the range 0.1 to 2pm, especially in the range 0.15 to 1 .5 pm and most preferably in the range 0.2 to 1 ,0pm.

More preferably the DL has an average thickness in the range 0.15 to 1 .0 pm and most preferably in the range 0.2 to 0.75 pm. By ensuring that the NAMW of the curable copolymer is above 1 ,800 Da, the retention of the DL on top of the porous substrate can be controlled and therefore the thickness of the DL can be controlled and be prepared in a form which is has the desired thickness and is defect- free. The claimed DL thickness provides membranes which are particularly useful for separating polar and non-polar gases, e.g. separating H2S gas from non-polar gases (e.g. from natural gas).

In view of the foregoing, the DL is preferably obtained by curing a curable composition comprising:

(i) 0.1 to 25wt% (especially 1 .0 to 15.0 wt%) of the curable copolymer;

(ii) 0 to 10.0wt% (especially 0 to 2 wt%) of further monomers;

(iii) 0.0001 to 8.0wt% (especially 0.005 to 3 wt%) of initiator;

(iv) 50 to 99.7999wt% (especially 75 to 98.0wt%) of inert solvent; and

(v) 0 to 5.0wt% (especially 0.1 to 2wt%) of surfactant

Preferably the amount of (i) + (ii) + (iii) + (iv) + (v) adds up to 100%. This does not exclude the presence of other components other than (i), (ii), (iii), (iv), and (v) but it sets the total amount of these five components. In one embodiment the composition consists solely of components (i), (ii), (iii), (iv) and (v).

The curable copolymer may be prepared by, for example, by polymerization of a composition (from now on called a precursor polymer composition (“PPC”)) comprising the following components: i. an ethylenically unsaturated monomer comprising two or more acrylate groups (often abbreviated herein to “acrylate monomer”); ii. a compound comprising two or more groups selected from thiol (-SH) or thioether (e.g. -CH2-S-CH2-) groups; iii. a Michael-addition catalyst; optionally iv. one or more further monomers; and optionally v. and inert solvent.

The process for preparing the curable copolymer from the PPC preferably comprises one or more of the following process features:

(a) slow addition of component ii. to a mixture comprising component i., component iii. and optionally component iv. and/or component v., e.g. addition of component ii. to the other components of the PPC composition at a speed of 0.1 to 100 ml/min, more preferably at a speed of 0.5 to 80 ml/min and most preferably at a speed of 1 .0 to 50 ml/min;

(b) an elevated temperature, e.g. a temperature in the range 25 to 120°C, especially 30 to 95°C; and

(c) a curing period of 2 to 48 hours, especially 10 to 36 hours.

Typical conditions for preparing the curable copolymer are illustrated in the Examples section below. By adjusting the ratio of component ii.:i. , temperature and the amount of component iii. one may tailor the values of sulphur content, and NAMW of the curable copolymer as desired.

Preferably the component i. contains EO groups, more preferably groups of the formula -(CH2-CH2-O) _n - wherein n is as hereinbefore defined.

The ethylenically unsaturated groups present in component i. comprise are preferably each independently selected from . (meth)acrylate, (meth)acrylamide and Michael acceptor groups (e.g. a,[3-unsaturated ketones, esters, and amides).

Examples of compounds which may be used as component i. to make the curable copolymerwhich comprise an EO group include poly(ethylene glycol) diacrylate, bisphenol A ethoxylate diacrylate, neopentyl glycol ethoxylate diacrylate, propanediol ethoxylate diacrylate, butanediol ethoxylate diacrylate, hexanediol ethoxylate diacrylate, polyethylene glycol) diacrylamide, bisphenol A ethoxylate diacrylamide and combinations of two or more thereof.

The amount of component i. present in the PPC (used to make the curable copolymer) is preferably in the range of 10 to 50 wt%, more preferably 15 to 40wt%, and especially 15 to 35wt%, relative to the weight of the PPC.

Preferably component ii. contains three or more thiol or thioether groups, more preferably three or more thiol groups. Thiol (-SH) groups are preferred because these are reactive with component i.. When component ii. has less than two thiol groups then it preferably has two or more ethylenically unsaturated groups (so that it can copolymerize with component i.).

The thiol and/or thioether groups present in component ii. Are useful for providing the thioether groups in the DL.

Preferred compounds which may be used as component ii. include, for example, poly(ethylene glycol) dithiol, bisphenol A ethoxylate dithiol, trimethylolpropane tris(3-mercaptopropionate), pentaerythritol tetrakis(3- mercaptopropionate), tetra(ethylene glycol) dithiol, hexa(ethylene glycol) dithiol, 2,2'- (Ethylenedioxy)diethanethiol, trithiocyanuric acid, 2-hydroxymethyl-2-methyl-1 ,3- propanediol tris(3-mercaptopropionate), ethylene glycol bismercaptoacetate and combinations of two or more thereof.

The amount of component ii. present in the PPC (i.e. the composition used to make the curable copolymer) is preferably in the range of 0.5 to 10 wt%, more preferably 0.75 to 8.0 wt%, and especially 1 .0 to 5.0 wt%, relative to the weight of the PPC.

Preferably the Michael-addition catalyst used as component iii. comprises a base. More preferably the Michael-addition catalyst comprises an organic base.

Preferred Michael-addition catalysts include, for example, trimethylamine, triethanolamine, trimethylamine, trimethanolamine, 1 ,8-Diazabicyclo [5.4.0] undec-7- en (DBU) and combinations of two or more thereof.

The amount of component iii. present in the PPC used to make the curable copolymer is preferably in the range of 0.005 to 0.05 wt%, more preferably 0.005 to 0.03 wt%, and especially 0.005 to 0.02 wt%, relative to the weight of the PPC.

As mentioned above, optionally the PPC used to prepare the curable copolymer further comprises one or more further monomers as component iv. , e.g. a crosslinking agent and/or a monomer having only one ethylenically unsaturated group.

Preferred crosslinking agents which may be used as component iv. comprise two or more ethylenically unsaturated groups. Preferred ethylenically unsaturated groups include allyl groups and (meth)acrylic groups (e.g. CH2=CR ⁴ -C(O)- groups), especially (meth)acrylate groups (e.g. CH2=CR ⁴ -C(O)O- groups) and (meth)acrylamide groups (e.g. CH2=CR ⁴ -C(O)NR ⁴ - groups), wherein each R ⁴ independently is H or CH3.

Examples of such further monomers having only one ethylenically unsaturated group which may be used as component iv. include polyethylene glycol) (meth)acrylate, dimethyl(aminopropyl) (meth)acrylamide, allyl (meth)acrylate, vinyl (meth)acrylate and combinations of two or more thereof.

The amount of such further monomers which may be which may be used as component iv. of the PPC is preferably in the range of 0 to 10wt%, more preferably 0 to 5wt%, and especially 0 to 2wt%, relative to the weight of the PPC. Optionally the PPC used to prepare the curable copolymer further comprises one or more inert solvents as component v.. Inert solvents are not curable and do not cross-link with any component of the composition.

Examples of inert solvents which may be used as component v. include alcohol- based solvents, ether-based solvents, amide-based solvents, ketone-based solvents, sulfoxide-based solvents, sulfone-based solvents, nitrile-based solvents and organic phosphorus-based solvents. Examples of alcohol-based solvents include methanol, ethanol, isopropanol, n-butanol, ethylene glycol, propylene glycol, diethylene glycol, dipropylene glycol and mixtures comprising two or more thereof.

Examples of inert solvents which may be used as component v. include dimethyl sulfoxide, dimethyl imidazolidinone, sulfolane, N-methyl pyrrolidone, dimethyl formamide, acetonitrile, acetone, 1 ,4-dioxane, 1 ,3-dioxolane, tetramethyl urea, hexamethyl phosphoramide, hexamethyl phosphorotriamide, pyridine, propionitrile, butanone, cyclohexanone, tetrahydrofuran, tetrahydropyran, 2-methyltetrahydrofuran, ethylene glycol diacetate, cyclopentylmethylether, methylethylketone, ethyl acetate, y- butyrolactone and mixtures comprising two or more thereof.

Preferably the PPC is free of a liquid selected from the group consisting of ionic liquids, glycerin, polyglycerin, polyethylene glycol, polypropylene glycol, polyethylene oxide, and amine compounds having 15 or less carbons. Essential free in the context of this inventions means that the amounts present in the PPC are far below 0.1 wt% and preferably below 0.05 wt% and even more preferably below 0.02 wt%.

Preferably the PPC is free from water, even when the components used to make the curable copolymer and the curable copolymer itself are water-soluble. However, if necessary, one may include water in the PPC as part of the inert solvent, preferably in an amount of 10wt% or less, more preferably 5wt% or less and most preferably 0wt%, relative to the total weight of inert solvent present in the PPC.

Especially preferred inert solvents which may be used as component v. include methyl ether ketone, n-butyl acetate, ethyl acetate, cyclopentyl methyl ether and 2- methyltetrahydrofuran and combinations of two or more thereof.

Component v. optionally comprises a single inert solvent or a combination of two or more inert solvents. Preferred inert solvents which may be used as component v. include C1-4 alcohols (e.g. methanol, ethanol and propan-2-ol), diols (e.g. ethylene glycol and propylene glycol), triols (e.g. glycerol), carbonates (e.g. ethylene carbonate, propylene carbonate, dimethyl carbonate, diethyl carbonate, di-t-butyl dicarbonate and glycerin carbonate), dimethyl formamide, acetone, methyl ethyl ketone, ethyl acetate, butyl acetate, 2-methyl tetrahydrofuran, cyclopentyl methyl ether, N-methyl-2- pyrrolidinone and mixtures comprising two or more of the foregoing. Particularly preferred inert solvents which may be used as component v. include n-butyl acetate, methyl ethyl ketone, ethyl acetate and mixtures comprising two or more thereof. In one embodiment the component v. has a low boiling point e.g. a boiling point below 120 °C. Solvents having a low boiling point can be easily removed after the curable copolymer has been prepared by evaporation, avoiding the need for a washing step for removal of the solvent.

Being inert, component v. does not react with any of the other components of the PPC.

The amount of component v. present in the PPC (used to form the curable copolymer) is preferably in the range of 29.95 to 89.495wt%, more preferably 40 to 85wt%, relative to the weight of the PPC.

In view of the foregoing, curable copolymer is preferably obtainable by a batch process comprising reaction of a composition (PPC) comprising:

(i) 10.0 to 50.0wt% of an ethylenically unsaturated monomer comprising two or more acrylate groups (acrylate monomer);

(ii) 0.5 to 10.0wt% of a compound comprising two or more groups selected from thiol groups ;

(iii) 0.005 to 0.05wt% of a Michael-addition catalyst;

(iv) 0 to 10 wt% of further monomer(s); and

(v) 29.95 to 89.495 wt% of inert solvent(s).

Preferably the amount of i. + ii. + iii. + iv. + v. adds up to 100%. This does not exclude the presence of other components other than i., ii., iii., iv. and v. but it sets the total amount of these four components. In one embodiment the composition consists solely of components than i., ii., iii., iv. and v..

The wt% of EO groups and the sulphur content of the DL may be calculated from the amounts and identity of the components used to form the DL.

The NAMW of the curable copolymer may be measured on the curable copolymer as described above. Where the amounts and identity of the components used to form the DL (or curable copolymer) are not known, for example where the GSM has been obtained from a supplier who refuses to provide this information, one may determine the wt% of EO groups and sulphur content of the DL by analysis of the DL, e.g. using pyrolysis and gas chromatography. A more preferred technique to analyze the DL is to hydrolyze the DL and analyze the hydrolysis product by sizeexclusion chromatography, mass spectrometry, inductively coupled plasma - optical emission spectrometry (ICP-OES) or inductively coupled plasma - mass spectrometry (ICP-MS). These analytical techniques may also be used to determine the identity and amount of any other components used to form the curable copolymer and/or DL, e.g. component iv. and/or (iv). These analytical techniques are particularly useful for determining the identity and ratio of monomers used to form the DL (and the identity and ratio of monomers used to form the curable copolymer from the PPC solution). A suitable pyrolysis and gas chromatography technique which may be used to determine the composition of the DL in a GSM is described in the paper by H. Matsubara and H. Ohtani entitled “Rapid and Sensitive Determination of the Conversion of UV-cured Acrylic Ester Resins by Pyrolysis-Gas Chromatography in the Presence of an Organic Alkali” in Analytical Sciences, 2007, 23(5), 513.

The preferable method to analyze the sulphur content is by ICP-OES. The preferable method to determine the EO content is mass spectrometry.

The preferred analytical technique for determining the sulphur content of the curable copolymer and/or the DL is ICP-OES. The preferred analytical technique for determining the value of n and the ethylene oxide content (wt%) of the curable copolymer and/or DL is mass spectrometry.

In our experiments we observed that the sulphur content of the curable copolymer has a significant effect on the NAMW of the curable copolymer and therefore also on the final performance of the GSM derived from the curable copolymer. The analytical techniques for determining the sulphur content of the curable copolymer and/or DL is described above. The sulphur content of the curable copolymer can influence how much of the curable copolymer is impregnated into the porous substrate and how much will stay on top of the porous substrate to form a DL external to the porous substrate. On the one hand, when the sulphur content of the curable copolymer is 0% of all atoms present in the curable copolymer, the curable copolymer will usually fully impregnate into the porous substrate without leaving any observable DL on the surface of the porous substrate. As a result, the GSM can have performance shortcomings as described in more detail below in relation to the dying test. On the other hand, when the sulphur content of the curable copolymer is above 10% of all atoms present in the curable copolymer, the curable copolymer impregnates the porous substrate little or not at all. As a consequence the resultant DL is entirely or almost entirely on the surface of the porous substrate and the bonding between the DL and the porous substrate may, as a result, be weak. However when the sulphur content of the curable copolymer is below 10% of all atoms present in the curable copolymer and above 0% of all atoms present in the curable copolymer, the curable copolymer partially impregnates the substrate and on curing a DL is formed on a surface of and to some extent within the porous substrate leading to GSMs having advantageous properties. The analytical techniques for determining the sulphur content of the curable copolymer is described above. The sulphur content of the curable copolymer is expressed as a percent and is the % of sulphur atoms compared to the total amount of atoms present in a sample of the curable copolymer. The sulphur content of the curable copolymer may often be determined from analysis of the DL or by knowledge of the composition used to form the DL.

The curable copolymer preferably has a sulphur content of up to 10.0%, more preferably in the range 0.01 to 10.0%, and most preferred 0.01 to 5.0%, relative to the total number of atoms present in the curable copolymer. The DL preferably has a sulphur content of 0.01 % or above, more preferably in the range 0.01 to 10.0%, and most preferred 0.01 to 5.0%, relative to the total number of atoms present in the DL. The amount of thioether groups present in the curable copolymer is directly related to the sulphur content of the DL.

Preferably the GSM is substantially non-porous. In other words, the GSM preferably comprises pores having an average size (i.e. average pore size) which does not exceed the kinetic diameter of the gas molecules which are desired to be retained by (i.e. not pass through) the GSM.

A suitable method to determine the average size of pores in a GSM is to inspect the surface thereof (typically the discriminating layer) by scanning electron microscope (SEM) e.g. using a Jeol JSM-6335F Field Emission SEM, applying an accelerating voltage of 2kV, working distance 4 mm, aperture 4, sample coated with Pt with a thickness of 1 .5 nm, magnification 100 000, 3° tilted view.

Preferably the DL has an average pore size of below 10 nm, more preferably below 5 nm, especially below 2 nm. The maximum preferred pore size depends on the application e.g. on the gases to be separated.

Another method to obtain an indication of the porosity of a GSM is to measure its permeance to a liquid, e.g. water. Preferably the permeance of the GSM of the present invention to liquids is very low, i.e. the average pore size of the GSM is such that its pure water permeance at 20 °C is less than 6.10 ^-8 m ³ /m ² *s*kPa, more preferably less than 3.10’8/(m2*s*kPa).

Preferably the GSM of the present invention (after curing and optional washing and/or drying) is free of liquid selected from the group consisting of ionic liquids, glycerin, polyglycerin, polyethylene glycol, polypropylene glycol, polyethylene oxide, and amine compounds having 15 or less carbons and the GSM further comprises at least on top of the substrate a thin dense DL. Essential free in the context of this inventions means that the amounts present in the PPC and/or DL and GSM are far below 0.1 wt% and preferably below 0.05 wt% and even more preferably below 0.02 wt%.

The primary purpose of the porous substrate is to provide the GSM with mechanical strength without materially reducing gas flux.

The porous substrate is typically open-pored (before it is converted into the GSM), relative to the DL. Typically the porous substrate is sufficiently porous to allow the composition from which the DL is derived to enter the pores and partially impregnate the porous substrate.

The porous substrate preferably comprises one, two or more sheet materials. For example, in a preferred embodiment the porous substrate comprises a non-woven backing sheet as porous support (e.g. to provide mechanical strength). In this embodiment, the DL is preferably in contact with the backing sheet and optionally adheres it to the non-woven backing sheet. Examples of suitable non-woven backing sheets which may be used as porous support (e.g. for providing mechanical strength) include microporous organic and inorganic membranes, woven or non-woven fabric. The non-woven backing sheet may be constructed from any suitable material. Examples of such materials include polysulfones, polyethersulfones, polyimides, polyetherimides, polyamides, polyamideimides, polyacrylonitrile, polycarbonates, polyesters, polyacrylates, cellulose acetate, polyethylene, polypropylene, polyvinylidenefluoride, polytetrafluoroethylene, poly(4-methyl 1 -pentene), polyacrylonitrile and especially polyesters. One may use a commercially available porous sheet material as the backing sheet, if desired. Alternatively one may prepare the porous sheet material using techniques generally known in the art for the preparation of microporous materials.

The porous substrate may be bonded to the DL by partial permeation and curing of the composition used to form the DL. Preferably the porous substrate comprises a coating of polyacrylonitrile (PAN), polysulphone (PSf), polyether ether ketone (PEEK) and/or polytetrafluoroethylene (PTFE).

Examples of suitable porous substrates which are particularly good at receiving the composition used to form the DL include PAN, PSf, and PTFE.

Optionally the porous substrate or a component thereof, especially the backing sheet, has been subjected to a corona discharge treatment, glow discharge treatment, flame treatment, ultraviolet light irradiation treatment or the like, e.g. for the purpose of improving its wettability and/or adhesiveness.

The porous substrate preferably comprises a porous coating layer which has pores that have an average diameter at the surface which is smaller than the average diameter half way through the porous sheet material. In this embodiment, the pores half way through the porous sheet material preferably have an average diameter of 0.001 to 10 pm, more preferably 0.01 to 1 pm, when measured half way through the porous sheet. Furthermore, the pores preferably have a smaller average diameter at the surface of the porous substrate, e.g. an average diameter of 0.001 to 0.1 pm, preferably 0.005 to 0.05 pm. The pore diameters may be determined by, for example, viewing the surface of the porous substrate before it is converted to the GSM by scanning electron microscopy (“SEM”) and by cutting through the porous substrate and measuring the diameter of the pores half way through the porous substrate, again by SEM.

The porosity at the surface of the porous substrate may also be expressed as a % porosity, i.e.

% porosity = 100% x (area of the surface which is missing due to pores)

(total surface area) The areas required for the above calculation may be determined by inspecting the surface of the porous substrate by SEM. Thus, in a preferred embodiment, the porous substrate has a % porosity >1 %, more preferably >3%, and especially >10%. In contrast, the porosity of the porous substrate comprising a gutter layer (e.g. PTMSP or PDMS) was <1 %.

The porosity of the porous substrate may also be expressed as a CO2 gas permeance (units are m ³ (STP)/m ² .s.kPa). Preferably the porous substrate has a CO2 gas permeance of 5 to 150 x 10’ ⁵ m ³ (STP)/m ² .s.kPa, more preferably of 5 to 100 x 1 O’ ⁵ m ³ (STP)/m ² .s.kPa, most preferably of 7 to 70 x 10’ ⁵ m ³ (STP)/m ² .s.kPa.

In contrast, when the porous substrate comprised a gutter layer (e.g. PTMSP or PDMS) the CO2 gas permeance of the porous substrate was much lower, typically below 5 x 10’ ⁵ m ³ (STP)/m ² .s.kPa

Alternatively the porosity of the porous substrate may be characterised by measuring the N2 gas flow rate through the porous substrate. Gas flow rate can be determined by any suitable technique, for example using a Porolux™ 1000 device, available from Porometer.com. Typically the Porolux™ 1000 is set at the maximum pressure (about 34 bar) and one measures the flow rate (L/min) of N2 gas through the porous substrate under test. The N2 flow rate through the porous substrate at a pressure of about 34 bar for an effective sample area of 2.69 cm ² (effective diameter of 18.5mm) is preferably >1 L/min, more preferably >5L/min, especially >10L/min, more especially >25L/min. The higher of these flow rates are preferred because this reduces the likelihood of the gas flux of the resultant membrane being reduced by the porous substrate.

The GSMs of the present invention are very robust, have few surface defects and retain good selectivity for long periods of time. These properties are retained even when the GSMs are used to separate gases at very high pressure and temperature and when feed gas comprises H2S. In addition, the bond between the DL and the porous substrate is very strong and the GSM has good scratch resistance. Thus the GSMs of the present invention do not need a protective layer (PL) on top of the DL, although a PL may be included if desired.

The thickness of the various layers (e.g. the porous substrate, the DL and the optional ((though unnecessary) protective layer) may be determined by cutting through the GSM and examining the cross section by SEM. The part of the DL which is present within the pores of the porous substrate is not taken into account when defining the thickness of the DL.

Preferably the GSM of the present invention has an average dry thickness (including the porous substrate) in the range 0.1 to 1000 pm, more preferably 5 to 500 pm and especially 10 to 200 pm.

Preferably the GSM has a H2S/CH4 selectivity (aH2S/CH4) ^30. Preferably the selectivity is determined by a process comprising exposing the membrane to a C02/CH4/nC4Hio/H2S = 47/33/0.5/0.1 (amounts by volume) of H2S and CH4 respectively at a feed pressure of 3760 kPa at 30°C.

Preferably the GSM has a permeability to H2S of at least 300 Barren Preferably the GSM has a permeability to CH4 of at most 10 Barren The permeability may be measured by the method described below.

In a preferred embodiment the composition used to prepare the DL is applied to the porous substrate by a coating process. Examples of coating processes include slot die coating, slide coating, air knife coating, roller coating, screen-printing, and dipping. Depending on the used technique and the desired end specifications, it might be necessary to remove excess composition from the porous substrate by, for example, roll-to-roll squeeze, roll-to-blade or blade-to-roll squeeze, blade-to-blade squeeze or removal using coating bars.

Advantageously the GSMs of the present invention avoid the need for a gutter layer between the DL and the porous substrate. This results in lower raw materials costs and a simpler, cheaper manufacturing process for the GSMs.

Thus the present invention enables the preparation of GSMs in which the DL is not in contact with a gutter layer, e.g. not in contact with any layer comprising Si-CHs groups. Gas-separation membranes free from Si-CHs groups and in fact free from organosilicon groups may be prepared by the present invention. Thus the process according to the second aspect of the present invention enables one to use a porous substrate which is free from gutter layers. The DL may be in (direct) contact with the porous substrate.

The target value of H2S permeance for the GSMs of the present invention is below 700 GPU.

Preferably the curable composition used to prepare the DL is applied to the porous substrate in a roll-to-roll process having high tension forces at unrolling and/or rolling the porous substrate of at least 50 N/m ² In even more preferred process the tension forces of unrolling or rolling the porous substrate are at least 100N/m ² .

The composition used to form the DL is preferably radiation-curable. Preferably irradiation to cure the composition begins within 7 seconds, more preferably within 5 seconds, most preferably within 3 seconds, of the composition used to prepare the DL being applied to the porous substrate.

Suitable sources of radiation include mercury arc lamps, carbon arc lamps, low pressure mercury lamps, medium pressure mercury lamps, high pressure mercury lamps, swirl flow plasma arc lamps, metal halide lamps, xenon lamps, tungsten lamps, halogen lamps, lasers and ultraviolet light emitting diodes. Particularly preferred are UV emitting lamps of the medium or high pressure mercury vapor type. In addition, additives such as metal halides may be present to modify the emission spectrum of the lamp. In most cases lamps with emission maxima between 200 and 450 nm are particularly suitable.

The energy output of the irradiation source is preferably from 20 to 1000 W/cm, preferably from 40 to 500 W/cm but may be higher or lower as long as the desired exposure dose can be realized.

Irradiation in order to cure the composition used to prepare the DL may be performed once or more than once.

In order to produce sufficiently flowable composition for use in a high speed coating machine, the composition used to form the DL preferably has a viscosity below 4000 mPa s when measured at 25°C, more preferably from 0.4 to 1000 mPa s when measured at 25°C. Most preferably the viscosity of the composition is from 0.4 to 500 mPa.s when measured at 25°C. For coating methods such as slide bead coating the preferred viscosity is from 1 to 100 mPa.s when measured at 25°C. The desired viscosity is preferably achieved by controlling the amount of inert solvent in the composition and/or by appropriate selection of the components of the composition and their amounts.

With suitable coating techniques, coating speeds of at least 5 m/min, e.g. at least 10 m/min or even higher, such as 15 m/min, 20 m/min, 25 m/min or even up to 100 m/min, can be reached. In a preferred embodiment the composition used to form the DL is applied to the porous substrate at a coating speed as described above.

While it is possible to prepare the GSMs of the present invention on a batch basis with a stationary porous substrate, it is much preferred to prepare them on a continuous basis using a moving porous substrate, e.g. the porous substrate may be in the form of a roll which is unwound continuously or the porous substrate may rest on a continuously driven belt. Using such techniques the composition used to form the DL can be applied on a continuous basis or it can be applied on a large batch basis. Removal of any inert solvent present in the composition used to form the DL can be accomplished at any stage after the composition has been applied to the porous substrate, e.g. by evaporation or drying.

Thus in a preferred process for making the GSMs of the present invention, the composition used to form the DL is applied continuously to a porous substrate by means of a manufacturing unit comprising one or more composition application stations, one or more curing stations and a GSM collecting station, wherein the manufacturing unit comprises a means for moving the porous substrate from the first to the last station (e.g. a set of motor driven pass rollers guiding the substrate through the coating line). The manufacturing unit optionally comprises one composition application station which applies the composition, e.g. a slide bead coater. The unit optionally further comprises one or more drying stations or IR-heating stations, e.g. for drying the final GSM. The GSM is preferably in tubular form or, more preferably, in sheet form. Tubular forms of GSMs are sometimes referred to as being of the hollow fibre type. GSMs in sheet form are suitable for use in, for example, spiral-wound, plate-and-frame and envelope cartridges.

While this specification emphasizes the usefulness of the GSMs of the present invention for separating gases, especially polar and non-polar gases, it will be understood that the GSMs can also be used for other purposes, for example providing a reducing gas for the direct reduction of iron ore in the steel production industry, dehydration of organic solvents (e.g. ethanol dehydration), pervaporation, oxygen enrichment, solvent resistant nanofiltration and vapour separation.

The GSMs of the present invention are particularly suitable for separating a feed gas containing a target gas into a gas stream rich in the target gas and a gas stream depleted in the target gas. For example, a feed gas comprising polar and nonpolar gases may be separated into a gas stream rich in polar gases and a gas stream depleted in polar gases. In many cases the GSMs have a high permeability to polar gases, e.g. CO2, H2S, NH3, SO _X , and nitrogen oxides, especially NO _X , relative to nonpolar gases, e.g. alkanes, H2, N2, and water vapour.

The target gas may be, for example, a gas which has value to the user of the GSM and which the user wishes to collect. Alternatively the target gas may be an undesirable gas, e.g. a pollutant or ‘greenhouse gas’, which the user wishes to separate from a gas stream in order to meet product specification or to protect the environment.

Preferably the GSM has a H2S/CH4 selectivity (a H2S/CH4) ^30. Preferably the selectivity is determined by a process comprising exposing the GSM to a CO2/CH4/nC4H /H2S = 77/22/0.7/0.3 (amounts by volume) of H2S and CH4 respectively at a feed pressure of 6000 kPa at 40°C.

Preferably the GSM has a permeability to H2S of at least 300 Barren Preferably the GSM has a permeability to CH4 of at most 10 Barren

The permeability may be measured by the method described below.

Preferably the GSM is gas permeable and liquid impermeable.

According to a third aspect of the present invention there is provided a process for separating a feed gas comprising polar and non-polar gases into a gas stream rich in polar gas and a gas stream depleted in polar gas comprising bringing the feed gas into contact with a GSM according to the first aspect of the present invention.

Thus the GSMs of the present invention may be used for the separation of gases and/or for the purification of a gas

According to a fourth aspect of the present invention there is provided a gas-separation module comprising a GSM according to the first aspect of the present invention. In the modules of the fourth aspect of the present invention the GSM is preferably in the form of a flat sheet, a spiral-wound sheet or takes the form of a hollowfibre membrane.

The invention will now be illustrated by the following non-limiting Examples in which all parts are by weight unless specified otherwise.

The following materials were used in the Examples (all without further purification):

The following materials were used to prepare the GSMs described below:

Omnirad 1173 is a photoinitiator from IGM resins.

MEK is methylethylketone (an inert solvent) from Sigma-Aldrich/Merck.

DBU is 1 ,8-Diazabicyclo [5.4.0] undec-7-en (a Michael-addition catalyst) from Sigma-Aldrich/Merck.

A-BPE-30 is ethoxylated bisphenol A diacrylate (a crosslinking agent from Sartomer). is polyether modified acryl functional polydimethylsiloxane BYK UV-

BYK 3530 (an surfactant from BYK Chemie GmbH).

PET-4SH Is pentaerythritol tetrakis(3-mercaptopropionate) from Sigma- Aldrich/Merck.

PAN is a polyacrylonitrile porous substrate (MN PAN an ultrafiltration membrane from MicrodynNadir) comprising a PET sheet backing material coated with PAN having a CO2 permeance 11 x 10’ ⁵ m ³ (STP)/m ² .s.kPa and a N2 gas flow of 2 L/min.

TMP-3SH is trimethylolpropane tris(3-mercaptopropionate) from Sigma- Aldrich/Merck.

PEG1000DA is polyethylene glycol) diacrylate with NAMW of 1000 Da from Sigma Aldrich/Merck.

PEG700DA is poly(ethylene glycol) diacrylate with NAMW of 700 Da from Sigma Aldrich/Merck. is tetraethylene glycol diacrylate with NAMW of 302 Da from Sigma

TEGDA Aldrich/Merck. The performance of the GSMs and (C)PPCs were evaluated in the following tests: Gas Selectivity and Permeability The gas selectivity and permeability of each of the GSMs under test were measured using a feed gas having the composition CO2/CH4/nC4H10/H2S in the ratio 47/33/0.5/0.1 (by volume). The feed gas was passed through each GSM under test at a temperature of 30 °C and feed pressure of 3760 kPa using a circular gas permeation cell having a measurement diameter of 1.5 cm. The flow rate, pressure, and gas composition of each feed gas, permeate gas, and retentate gas was calculated according formulation described in “Calculation Methods for Multicomponent Gas Separation by Permeation” (Y. Shindo et al, Separation Science and Technology, Vol.20, Iss.5-6, 1985) with “countercurrent flow” mode. Permeability The permeability (Pi) shown in Table D1 was measured as follows: The permeability (Pi) of CO2, H2S, CH4 and nC4H10 was determined using the following equation: For example: wherein Pi = Permeability of the relevant gas (i.e. is CO2, H2S, CH4 or nC4H10) (m ³ (STP)·m/m ² ·kPa·s); ^Perm = Permeate flow rate (m ³ (STP)/s); XPerm,i = Volume fraction of the relevant gas in the permeate gas; A = Membrane area (m²); PFeed = Feed gas pressure (kPa); XFeed,i = Volume fraction of the relevant gas in the feed gas; PPerm = Permeate gas pressure (kPa); and STP is standard temperature and pressure, which is defined here as 25.0 °C and 1 atmosphere pressure (101.325 kPa). The Barrer (P) was then determined by 1 Barrer = 1×10 ^-10 cm ³ (STP)·cm/(s・ cm ² ・cmHg). Selectivity The selectivity (Sel) shown in Table D1 was measured as follows: The membrane patch selectivity (H2S/CH4 selectivity; α(H2S/CH4)) of the membrane under test for the gas mixture described in Table 2 was calculated from respectively from P(H2S) and P(CH4) calculated as described in (A) above based on following equations: H2S/CH4 selectivity : α (H2S/CH4)= P(H2S)/P(CH4) The permeance (Q) shown in Table C was measured as follows: The permeance (Qi) of CO2, H2S, CH4 and nC4H10 was determined using the following equation: Qi = Pi · L Qi = Permeance of the relevant gas (i is CO2, H2S, CH4 or nC4H10) (m ³ (STP) / m ² · kPa·s); L = Thickness of discriminating layer in membrane [µm] The Barrer (Q) was then determined by 1 GPU = 1×10 ^-6 cm ³ (STP) / (s・cm ² ・ cmHg). Test for Defects in the Discriminating Layer - The Dying test Defects in the DL of the GSMs under test were identified as follows: 6 drops of a dye solution (a 1wt% solution of 1,4-bis[(2-ethylhexyl)amino]-anthraquinone (solvent blue) in n-heptane) were applied to the DL at room temperature and left in contact with the DL for 30 seconds. Then the excess of dye solution was removed from the DL and the DL was washed with n-heptane. The DL was then examined visually with the naked eye. If blue spots were visible on the DL this indicated the presence of defects (e.g. pinholes) and the GSM was scored "Yes". If no blue spots were visible this indicated the absence of defects and the GSM was scored "No". Sulphur amount determination on GSM 12.57 cm ² of GSMs under test were digested in 10 ml concentrated nitric acid. Digestion was done by means of microwave digestion (Digestion program 1200W ramp 15', Hold 30', Cool 30'.) The digested solutions were diluted up to 50ml with milli-Q water and sulphur amount was determined using Thermo ICAP-PRO ICP-OES. A standard concentric nebulizer is used in conjunction with a cyclonic spray chamber. All samples were prepared in duplicate.

Sulphur amount determination on (C)PPC

1 g of each (C)PPC was diluted up to 250 g with THF (tetrahydrofuran for HPLC (>99,9% inhibitor free)) from Sigma Aldrich/Merck in duplicate and sulphur amount was determined using Thermo ICAP-PRO ICP-OES. A standard concentric nebulizer is used in conjunction with a cyclonic spray chamber on both solutions for each (C)PPC.

NAMW determination on GSM

12.57cm ² of GSMs under test were degraded in 10 ml 1 M NaOH solution in water. Degradation was done by heating the solutions to 60 degrees Celsius for 16 hours. The degraded solutions were diluted up to 50ml with milli-Q water in duplicate. Both solutions were subjected to GPC analysis with THF (tetrahydrofuran for HPLC (>99,9% inhibitor free)) from Sigma Aldrich/Merck as eluent at 60 degrees Celcius to obtain the NAMW molecular weight distribution of the polymers. The GPC system used was a Viscotek TDA 305 GPC coupled to a Viscotek VE 2001 GPC solvent/sample module.

NAMW determination on (C)PPC

1 g of each (C)PPC was diluted up to 250 g with THF (tetrahydrofuran for HPLC (>99,9% inhibitor free)) from Sigma Aldrich/Merck in duplicate.

Both solutions were subjected to GPC analysis with the same THF as eluent at 60 degrees Celsius to obtain the NAMW molecular weight distribution of the polymers. The GPC system used was a Viscotek TDA 305 GPC coupled to a Viscotek VE 2001 GPC solvent/sample module.

EQ content determination on GSM

12.57cm ² of GSMs under test were degraded in 10 ml 1 M NaOD sodium deuteroxide (40wt% in deuterium oxide, 99,5 atom% D in deuterated water solution) from Sigma Aldrich/Merck in deuterated water by heating the solutions to 60 degrees Celsius for 16 hours. The degraded solutions were diluted up to 20ml with deuterated water (D2O) (deuterated water, 99,9% atom% D) from Sigma Aldrich/Merck containing 1 wt% of calcium formate as internal standard. 1 H-NMR analysis using the signal of calcium formate (singlet at 8.30 ppm) was correlated to the signal of ethylene oxide (singlet at 3.5 ppm) and the EO content was quantified in duplicate. The 1 H-NMR spectra were recorded using a Magritek Spinsolve 60 MHz table-top NMR (32 scans, 90o pulse angle, 3 sec acquisition time, 30 sec relaxation time). EO content determination on (C)PPC

1 g of each (C)PPC was diluted up to 25 g with D2O (deuterated water, 99,9% atom% D) from Sigma Aldrich/Merck, containing 1 wt% of calcium formate as internal standard. 1 H-NMR analysis using the signal of calcium formate (singlet at 8.30 ppm) was correlated to the signal of ethylene oxide (singlet at 3.5 ppm) and the EO content was quantified in duplicate. The 1 H-NMR spectra were recorded using a Magritek Spinsolve 60 MHz table-top NMR (32 scans, 90o pulse angle, 3 sec acquisition time, 30 sec relaxation time).

Examples and comparative examples:

(a) Preparation of Precursor Polymer Compositions (PPC)

PPC’s PPC1 to PPC9 and comparative PPC’s CPPC1 to CPPC7 were prepared by mixing the ingredients shown in Table A1 below at room temperature. 10 wt% solution of Monomer B in inert solvent (component (v)) was added to a mixture of Monomer A and Michael addition catalyst in the same inert solvent at 15 ml/min and the mixture was heated at 60 °C for 18 hours. The mixture was cooled down to room temperature. Subsequently the NAMW, the sulphur content and the EO content of the resultant curable copolymer was determined and the mixture was directly used later on for preparation of the Polymer Composition (PC).

The results of the NAMW, the sulphur content and the EO content of the curable copolymers are shown in Table A2. Table A1 : PPC compositions

* the value of n in the groups of the formula -(CH2CH2)nO- present in the resultant curable copolymer (and also in the final DL).

Table A2: Measured NAMW results of curable co-polymer made from the PPC Compositions described in Table A1

*N.A. = Not available, could not be determined due to gel formation.

(b) Preparation of Polymer Compositions (PC) used to Make the DLs

PCs PC1 to PC9 and comparative PCs CPC1 to CPC7 were prepared from curable copolymers to obtain the corresponding compositions PPC1 to PPC9 and CPPC1 to CPPC7 described in Table A1 above. The PCs and CPCs were obtained by mixing of each PPC and CPPC (34.4wt%) with MEK (64.9wt%), BYK (0.5wt%) and Omnirad™ 1173 (0.2wt%) and stirring the mixture for at least 15 min. to obtain a homogeneous mixture. This created PCs PC1 to PC9 and comparative PCs CPC1 to CPC7.

(c) Coating of Curable Polymer Compositions (PC) on Substrates

PCs PC1 to PC9 and comparative PCs CPC1 to CPC7. obtained in step (b) above were each independently coated on top of a porous substrate and cured to polymerise the compositions and form the discriminating layers. Membranes Ex1 to Ex9 and Comparative Membranes CEx1 to CEx8 were prepared by coating each of the compositions PC1 to PC9 and CPC1 to CPC8 as described in Table C1 below, continuously and at 20°C onto a porous support using just one slot of a slide bead coating machine. Comparative membrane CEx9 was prepared by coating composition PC1 same way as membrane of Ex1 on a substrate except that a non- porous PET substrate was used. The resultant, coated supports were dried at 30 °C in a drying zone and subsequently cured by passing them under an irradiation source (a Light Hammer LH6 from Fusion UV Systems fitted with a H-bulb working at 100% intensity). The resultant GSMs then travelled to the collecting station. A cross-section through the resultant membranes was examined by SEM and the thicknesses of the resultant DLs are shown in Table C1 below: Table C1 : H ₂ S GSM

*A thickness of 0.0 indicates that the composition used to form the DL permeated entirely into the porous substrate and consequently no DL was formed on the surface of the porous substrate. All DL thicknesses are measured from the surface of the porous substrate outwards.

(d) Testing of the GSMs

The membranes obtained in step (C) were tested using the methods described above to determine their H ₂ S permeability (Q(H ₂ S)) and their H ₂ S/CH4 selectivity (a(H ₂ S/CH4)). The results are shown in Table D1 below. A H ₂ S permeability (Q(H ₂ S)) above 150 GPU was deemed to be good. A H ₂ S/CH4 selectivity (a(H ₂ S/CH4)) from 30 was deemed to be good. The defects on the sample were visually checked by dying test as described earlier. When no visual defects are found, the sample was deemed to be good according to the present invention.

From Table D1 it can be seen that the membranes according to the present invention have good H2S permeability (>300 GPU) and very good H2S/CH4 selectivity (^30). The comparative membranes CEx1 to CEx6 and CEx8 had an unmeasurably high H2S permeability and no H2S/CH4 selectivity due to the detrimental effect on the membranes caused by defects in the DL. Furthermore, the sulphur content and EO content of the DLs in CExIto CEx6 and CEx8 could not be determined because there was no DL present on the surface of the substrate (instead the CPCs had permeated entirely within the porous substrate).

The membranes were tested for their adhesive strength (i.e. the strength of the adhesion of the DL to the porous substrate). In particular the strength between the porous substrate and the discriminating layer was evaluated by adhesion testing according to above method and all examples of the invention (Ex1 - Ex9) showed good adhesion too. As no DL was present on the surface of CEx1 to CEx6 and CEx8, it was not possible to measure DL adhesion for these comparative examples. The adhesive strength of CEx7 was evaluated and was found to be very poor - the DL was removed very easily from the non-porous PET substrate.

Table D1 - Results

*ND= Not determined