Durable asymmetric composite membranes and modified substrates used in their preparation

Abstract

Durable asymmetric composite membranes consisting essentially of a film of cross-linked sulfonated poly(ether ether ketone) adhered to a sheet of hydrophilicitized microporous poly(ethylene) are disclosed. The membranes have application in the recovery of water from feed streams where the ability to clean in situ is desirable, for example in dairy processing. Methods of preparing cross-linked sulfonated poly(ether ether ketone) suitable for use as the rejection layer and hydrophilicitized sheets of microporous poly(ethylene) suitable for use as the support layer of such membranes are also disclosed.

Claims

1. An asymmetric composite membrane comprising a film of crosslinked sulfonated poly(ether ether ketone) adhered to one side of a hydrophilicitized sheet of microporous polyolefin where the hydrophilicitized sheet of microporous polyolefin consists of polyolefin grafted with 4-ethenyl-benzenesulfonic acid.

2. The membrane of claim 1 where the crosslinked sulfonated poly(ether ether ketone) incorporates a crosslinking agent selected from the group consisting of: divinylbenzene; ethylene glycol dimethacrylate; and glyoxal bis(diallyl acetal).

3. The membrane of claim 2 where the crosslinked sulfonated poly(ether ether ketone) incorporates divinylbenzene as the crosslinking agent.

4. The membrane of claim 1 where the polyolefin is poly(ethylene).

5. An asymmetric composite membrane consisting of a film of crosslinked sulfonated poly(ether ether ketone) incorporating divinylbenzene as the crosslinking agent adhered to one side of a sheet of microporous poly(ethylene) grafted with 4-ethenyl-benzenesulfonic acid.

Description

BRIEF DESCRIPTION OF DRAWINGS

(1) FIG. 1. Comparison of FTIR spectra recorded for samples of cross-linked sulfonated PEEK: 150SClPEEK30 (upper trace), 150SClPEEK15 (middle trace) and control sample (lower trace). The scale of the vertical axis (absorbance; not shown) is the same for each trace.

(2) FIG. 2. Comparison of FTIR spectra recorded (2000 to 550 cm.sup.−1) for selected samples of cross-linked sulfonated PEEK: 150SClPEEK302X (upper trace), 150SCl30 (middle trace) and 150SClPEEK30 (lower trace). The scale of the vertical axis (absorbance; not shown) is the same for each trace.

(3) FIG. 3. Comparison of FTIR spectra recorded (4000 to 2000 cm.sup.−1) for selected samples of cross-linked sulfonated PEEK: 150SClPEEK302X (upper trace), 150SCl30 (middle trace) and 150SClPEEK30 (lower trace). The scale of the vertical axis (absorbance; not shown) is the same for each trace.

(4) FIG. 4. Comparison of FTIR spectra recorded (2000 to 550 cm.sup.−1) for samples of cross-linked sulfonated PEEK prepared by casting the same mixture (including the crosslinking agent succinyl chloride) and curing at 150° C. for 30 min or 15 min: 150SCl30 (upper trace) and 150SCl15 (lower trace). The scale of the vertical axis (absorbance; not shown) is the same for each trace.

(5) FIG. 5. Comparison of FTIR spectra recorded (2000 to 550 cm.sup.−1) for samples of cross-linked sulfonated PEEK prepared by casting mixtures with and without the inclusion of the crosslinking agent succinyl chloride and curing at 150° C. for 15 min: 150SCl30 (upper trace) and 50SCl15 (lower trace). The scale of the vertical axis (absorbance; not shown) is the same for each trace.

(6) FIG. 6. Exploded view of the filter assembly (Sterlitech Corp.).

(7) FIG. 7. Graph showing dependency of flux rate (L/m2h) (Y axis) on temperature (degrees C.) of the feed stream where the feed stream was milk (solid circles) or water (solid diamonds).

(8) FIG. 8. Flux rates (L/m.sup.2h) (Y axis) where the feed stream was milk at a pressure of 20 bar (solid diamonds) or 28 bar (solid squares) with in situ cleaning (vertical lines).

(9) FIG. 9. Graph showing the rejection of fat (solid diamonds), lactose (solid triangles), protein (solid squares) and total solids (crosses) determined for samples of permeate where the feed stream was skim milk at a pressure of 20 bar or 28 bar (third and fourth samples).

(10) FIG. 10. Flux rates (L/m.sup.2h) (Y axis) obtained following disinfection of the asymmetric composite membrane where the feed stream was water at a pressure of 20 bar.

(11) FIG. 11. Flux rates (L/m.sup.2h) (Y axis) obtained following disinfection of the asymmetric composite membrane where the feed stream was skim milk at a pressure of 20 bar.

(12) FIG. 12. Graph showing the rejection of fat (solid circles), lactose (solid inverted triangles), protein (open circles) and total solids (open inverted triangles) determined for samples of permeate where the feed stream was whole milk at a pressure of 20 bar and temperature of 7° C.

(13) FIG. 13. Comparison of the FTIR spectra obtained for Sample 1 (lower trace), Sample 3 (middle trace) and Sample 4 (upper trace). An FTIR spectrum was not obtained for Sample 2.

(14) FIG. 14. Comparison of the contact angles determined for Sample 1 (Ally [sic] alcohol), Sample 3 (HEMA), Sample 4 (SSS) and Sample 2 (Acrylic acid) before () and after exposure to an acid () or alkali (.square-solid.) environment.

(15) FIG. 15. Comparison of the permeability determined for Sample 1 (Allyl alcohol), Sample 3 (HEMA), Sample 4 (SSS) and Sample 2 (Acylic [sic] Acid) before () and after exposure to an acid () or an alkali (.square-solid.) environment relative to the permeability of the unmodified polyolefin substrate (□).

(16) FIG. 16. Correspondence between contact angle and permeability determined for samples before (.diamond-solid.) and after exposure to an acid (.box-tangle-solidup.) or an alkali (.square-solid.) environment. The outlier is Sample 4 (SSS) after exposure to an alkali (.square-solid.) environment.

(17) FIG. 17. Water absorption determined for Sample 1 (Allylic alcohol), Sample 2 (Arylic [sic] Acid), Sample 3 (HEMA) and Sample 4 (SSS).

(18) FIG. 18. The dry weight increase determined for Sample 5 (SSS), Sample 6 (AA), Sample 7 (HEMA) and the untreated microporous polyethylene substrate (CELGARD™ K2045).

(19) FIG. 19. The water absorption determined for Sample 5 (SSS), Sample 6 (AA), Sample 7 (HEMA) and the untreated microporous polyethylene substrate (CELGARD™ K2045).

(20) FIG. 20. The contact angles determined for Sample 5 (SSS), Sample 6 (AA), Sample 7 (HEMA) and the untreated microporous polyethylene substrate (CELGARD™ K2045).

(21) FIG. 21. The bubble points determined for Sample 5 (SSS), Sample 6 (AA), Sample 7 (HEMA) and the untreated microporous polyethylene substrate (CELGARD™ K2045).

(22) FIG. 22. The sodium rejection determined for Sample 5 (SSS), Sample 6 (AA), Sample 7 (HEMA) and the untreated microporous polyethylene substrate (CELGARD™ K2045).

(23) FIG. 23. The milk flux determined for Sample 5 (SSS), Sample 6 (AA), Sample 7 (HEMA) and the untreated microporous polyethylene substrate (CELGARD™ K2045).

(24) FIG. 24. The total milk solids rejection determined for Sample 5 (SSS), Sample 6 (AA), Sample 7 (HEMA) and the untreated microporous polyethylene substrate (CELGARD™ K2045).

(25) FIG. 25. Flux (LMH) (.diamond-solid.) and total solids rejection (%) (.square-solid.) for Sample 1 during repeated CIP protocols (10× according to the schedule provided in Table 14). The feed stream was whole milk.

(26) FIG. 26. Lactose rejection (%) detected by FTIR for Sample 2 during sequential CIP protocols (10 times according to the schedule provided in Table 7 followed by 12 times according to the schedule provided in Table 14), drying of the sample and further CIP protocols (8 times according to the schedule provided in Table 14). The feed stream was whole milk.

(27) FIG. 27. Flux (LMH) (.diamond-solid.) and total solids rejection (%) (.square-solid.) for Sample 3 during repeated CIP protocols (25× according to the Schedule provided in Table 14. The feed stream was whole milk.

(28) FIG. 28. Flux (LMH) (.diamond-solid.) and total solids rejection (%) (.square-solid.) for Sample 4 during repeated CIP protocols (17 times according to the schedule provided in Table 14). The feed stream was whole milk.

(29) FIG. 29. Flux (LMH) (x) for Sample 5 measured over a period of eight hours using raw milk as the feed stream.

(30) FIG. 30. Comparison of the total solids rejection (%) for Sample 6 and Sample 1 before (left hand bar) and after (right hand bar) a single CIP protocol according to the schedule provided in Table 14. The feed stream was whole milk.

(31) FIG. 31. Sodium chloride (NaCl) rejection (%) by Samples 7 to 10 of an asymmetric composite membrane prepared using different ratios of sPEEK and DVB in the preparation of the rejection layer. The feed stream was whole milk.

(32) FIG. 32. Lactose rejection (%) by Samples 7 to 10 of an asymmetric composite membrane prepared using different ratios of sPEEK and DVB in the preparation of the rejection layer. The feed stream was whole milk.

(33) FIG. 33. Total solids rejection (%) by Samples 7 to 10 of an asymmetric composite membrane prepared using different ratios of sPEEK and DVB in the preparation of the rejection layer. The feed stream was whole milk.

(34) FIG. 34. Flux (LMH) for Samples 7 to 10 of an asymmetric composite membrane prepared using different ratios of sPEEK and DVB in the preparation of the rejection layer. The feed stream was either deionised water (.diamond-solid.) or whole milk (⋄).

(35) FIG. 35. Comparison of the sodium chloride (NaCl) rejection (%) for Sample 11 of an asymmetric composite membrane prepared using a different combination of solvent and hydrophilicitizing agent.

(36) FIG. 36. Flux (LMH) for Sample 11 of an asymmetric composite membrane prepared using a different combination of solvent and hydrophilicitizing agent.

(37) FIG. 37. Comparison of the sodium chloride (NaCl) rejection (%) (coarse diagonal hatching), flux (LMH) (fine diagonal hatching) and sucrose rejection (%) (medium diagonal hatching) for Sample 13 of an asymmetric composite membrane prepared using unmodified μPE as the backing layer and a sample of a symmetric composite membrane prepared using hydrophilic μPE as the backing layer.

(38) FIG. 38. Characterisation of Sample 13 using the Donnan Stearic Pore Model (DSPM) and curve fitting with sucrose as the uncharged solute.

(39) FIG. 39. The flux (L/m.sup.2/hr) of samples 090517Di, 090517Di and 090517Di over three clean-in-place (CIP) protocols.

(40) FIG. 40. The salt (NaCl) rejection (%) of samples 090517Di, 090517Di and 090517Di over three clean-in-place (CIP) protocols.

DETAILED DESCRIPTION

(41) The invention provides an asymmetric composite membrane in which a film of sulfonated poly(ether ether ketone) (sPEEK) is adhered to one side of a sheet of hydrophilicitized microporous poly(ethylene) (μPE). Various options for crosslinking sPEEK are available. In a first option a Friedel-Crafts type reaction may be utilised. In a second option photoinitiated grafting of the crosslinking agents may be utilised. Similarly, various options for hydrophilicitizing a preformed μPE are available. In a first option direct sulfonation may be utilised. In a second option photoinitiated grafting with a hydrophilicitizing agent may be utilised. For the industrial manufacture of the asymmetric composite membrane the options utilising photoinitiated grafting are anticipated to be of greatest utility.

Example A

(42) In a first option, chlorosulfonated poly(ether ether ketone) is used as the substrate for cross-linking, specifically csPEEK. It is desirable to retain the degree of hydrophilicity imparted to the PEEK substrate by chlorosulfonation when crosslinking for the purpose of preparing a polymer for use in the fabrication of membranes for use in processes driven by hydrostatic or osmotic pressure. When crosslinking via the introduced chlorosulfonyl groups the degree of hydrophilicity may be reduced. Crosslinking of the csPEEK substrate via a Friedel-Crafts type reaction promotes maintenance of the desired degree of hydrophilicity. Furthermore, curing at temperatures substantially below those known to promote thermally induced crosslinking between chlorosulfony/sulfonyl groups favours the DXL being attributable to participation of the cross-linking agent in the crosslinking reaction. Curing at lower temperatures also reduces the likelihood of thermal degradation of the substrate polymer.

(43) The microstructure of the polymer network formed by cross-linking will also be influenced by the structure of the cross-linking agent selected for use.

(44) The combination of a film of cross-linked sulfonated poly(ether ether ketone) rejection layer adhered to a sheet of sulfonated microporous polyethylene backing layer provide an asymmetric composite membrane with the advantage of resistance to chemical decomposition and hence durability in commercial processing operations. In addition, the asymmetric composite membrane is tolerant of desiccation facilitating storage and transport.

(45) Preparation of Sulfonated Microporous Poly(Ethylene) (“Support Layer”)

(46) A sheet of sulfonated microporous poly(ethylene) was prepared substantially as described in the publication of Briggs (2015). An amount of 250 g of phosphorous pentoxide was added to a volume of 469 mL of sulfuric acid to provide a mixture of 1:5 (mol/mol) phosphorous pentoxide-sulfuric acid and heated to 90° C. to dissolve the phosphorous pentoxide (“sulfonating agent”).

Example 1

(47) Sheets of microporous poly(ethylene) were cut in to 15×20 cm pieces and wetted with a mixture of DMSO-trichloromethane (9:1 (v/v)). Excess of the mixture was removed from the pieces before they were frozen by covering them in liquid nitrogen. Immediately after freezing the pieces were added to the sulfonating agent at room temperature and incubated at 80° C. for 90 minutes. The incubated pieces were removed from the sulfonating agent and allowed to sit for 3 hours in order to dilute the acid before rinsing twice with methanol to remove all residual sulfonating agent.

Example 2

(48) Sheets of microporous poly(ethylene) were cut in to 14×28 cm pieces and whetted with a mixture of 20% chloroform and 8% dimethyl sulfoxide (DMSO) before being placed between glass fibre sheets and keeping on dry ice overnight. The sheets were then immersed in the sulfonating agent before being covered and cured in an oven at 85° C. for 90 minutes. Following curing the glass fibre sheets were removed and the sulfonated microporous poly(ethylene) sheet humidified overnight. The sheet was then rinsed with methanol and stored dry before use as a backing layer.

Preparation of Cross-Linked Sulfonated PEEK (“Rejection Layer”)

Example 3

(49) Solutions of dry chlorosulfonated poly(ether ether ketone), the cross-linking agent ethanedioic acid (oxalic acid), and the chloride catalyst ferric chloride (FeCl.sub.3), were prepared in the non-aqueous reactive solvent cyclopentanone under ambient conditions (room temperature and pressure): 6.6 g of dry chlorosulfonated poly(ether ether ketone) was dissolved in 50 mL of cyclopentanone (solution 1); 0.769 g of ethanedioic acid (oxalic acid) was dissolved in 25 mL of cyclopentanone (solution 2); and 0.138 g of ferric chloride (FeCl.sub.3) was dissolved in 25 mL of cyclopentanone (solution 3). A 3 mL volume of solution 2 was added to a 6 mL volume of solution 1 and the two solutions mixed thoroughly. A 3 mL volume of solution 3 was then added and the combined solutions mixed to provide a cross-linking solution. The molar ratio of chlorosulfonated poly(ether ether ketone) to ethanedioic acid (oxalic acid) was approximately 1:0.5. The molar ratio of chlorosulfonated poly(ether ether ketone) to ferric chloride (FeCl.sub.3) was approximately 1:0.05. The cross-linking solution was incubated at a temperature of 85° C. for two hours and then allowed to cool for 30 minutes under ambient conditions. The cooled cross-linking solution was then cast onto a glass plate. The film was cast under ambient conditions and left for two to five minutes before being cured at 85° C. for 2.5 hours. During curing the upper surface of the film was protected with a second glass plate located approximately 1 cm above the surface.

Example 4

(50) Solutions of dry chlorosulfonated poly(ether ether ketone), the cross-linking agent tartaric acid, and the chloride catalyst ferric chloride (FeCl.sub.3), were prepared in the non-aqueous reactive solvent cyclopentanone under ambient conditions (room temperature and pressure): 6.6 g of dry chlorosulfonated poly(ether ether ketone) was dissolved in 50 mL of cyclopentanone (solution 1); and 1.281 g of tartaric acid and 0.138 g of ferric chloride (FeCl.sub.3) were dissolved in 25 mL of cyclopentanone (solution 2). A 3 mL volume of cyclopentanone was added to a 6 mL volume of solution 1 and mixed thoroughly. A 3 mL volume of solution 2 was then added and the combined solutions mixed to provide a cross-linking solution. The molar ratio of chlorosulfonated poly(ether ether ketone) to tartaric acid was approximately 1:0.5. The molar ratio of chlorosulfonated poly(ether ether ketone) to ferric chloride (FeCl.sub.3) was approximately 1:0.05. The cross-linking solution was incubated at a temperature of 85° C. for two hours and then allowed to cool for 30 minutes under ambient conditions. The cooled cross-linking solution was then cast onto a glass plate. The film was cast under ambient conditions and left for two to five minutes before being cured at 85° C. for 2.5 hours. During curing the upper surface of the film was protected with a second glass plate located approximately 1 cm above the surface.

Example 5

(51) Solutions of dry chlorosulfonated poly(ether ether ketone), the cross-linking agent citric acid, and the chloride catalyst ferric chloride (FeCl.sub.3), were prepared in the non-aqueous reactive solvent cyclopentanone under ambient conditions (room temperature and pressure): 6.6 g of dry chlorosulfonated poly(ether ether ketone) was dissolved in 50 mL of cyclopentanone (solution 1); and 1.640 g of citric acid and 0.138 g of ferric chloride (FeCl.sub.3) were dissolved in 25 mL of cyclopentanone (solution 2). A 3 mL volume of cyclopentanone was added to a 6 mL volume of solution 1 and mixed thoroughly. A 3 mL volume of solution 2 was then added and the combined solutions mixed to provide a cross-linking solution. The molar ratio of chlorosulfonated poly(ether ether ketone) to citric acid was approximately 1:0.5. The molar ratio of chlorosulfonated poly(ether ether ketone) to ferric chloride (FeCl.sub.3) was approximately 1:0.05. The cross-linking solution was incubated at a temperature of 85° C. for two hours and then allowed to cool for 30 minutes under ambient conditions. The cooled cross-linking solution was then cast onto a glass plate. The film was cast under ambient conditions and left for two to five minutes before being cured at 85° C. for 2.5 hours. During curing the upper surface of the film was protected with a second glass plate located approximately 1 cm above the surface.

Example 6

(52) Solutions of dried chlorosulfonated poly(ether ether ketone), the cross-linking agents ethanedioic acid (oxalic acid) and tartaric acid, and the chloride catalyst ferric chloride (FeCl.sub.3), were prepared in the non-aqueous reactive solvent cyclopentanone under ambient conditions (room temperature and pressure): 6.6 g of chlorosulfonated poly(ether ether ketone) was dissolved in 50 mL of cyclopentanone (solution 1); 0.769 g of ethanedioic acid (oxalic acid) was dissolved in 25 mL of cyclopentanone (solution 2); 1.281 g of tartaric acid and 0.138 g of ferric chloride (FeCl.sub.3) was dissolved in 25 mL of cyclopentanone (solution 3); and 0.138 g of ferric chloride (FeCl.sub.3) was dissolved in 25 mL of cyclopentanone (solution 4). A 1.5 mL volume of cyclopentanone and a 1.5 mL volume of solution 2 was added to a 6 mL volume of solution 1 and mixed thoroughly. A 1.5 mL volume of solution 3 and a 1.5 mL volume of solution 4 was then added and the combined solutions mixed to provide a cross-linking solution. The molar ratio of chlorosulfonated poly(ether ether ketone) to tartaric acid was approximately 1:0.5. The molar ratio of chlorosulfonated poly(ether ether ketone) to ferric chloride (FeCl.sub.3) was approximately 1:0.05. The cross-linking solution was incubated at a temperature of 85° C. for two hours and then allowed to cool for 30 minutes under ambient conditions. The cooled cross-linking solution was then cast onto a glass plate. The film was cast under ambient conditions and left for two to five minutes before being cured at 85° C. for 2.5 hours. During curing the upper surface of the film was protected with a second glass plate located approximately 1 cm above the surface.

Example 7

(53) An amount of 50 g of poly(ether ether ketone) (Victrex) having a density of 1.3 g cm.sup.−3 and a melt viscosity of 400° C. of 90 Pa.Math.s was added to a volume of 250 mL of chloroform (Fisher Chemicals) followed by the addition of a volume of 250 mL of chlorosulfonic acid (Nacalai Tesque Inc.). The mixture was stirred for two hours at 50° C. and then washed twice with two separate volumes of 200 mL of chloroform and thrice with three separate volumes of 400 mL of chloroform. Washing was indicated to be complete when the volume of chloroform used in the washing remained colourless.

(54) The orange coloured, viscous residue remaining following the chloroform washings was washed repeatedly with volumes of deionized water until the pH of the wash water had increased to 5. The resulting whitish coloured product was then broken into small flakes and dried in a vacuum oven at 65° C. for five days.

(55) An amount of 1.9932 g of flakes of the dried, whitish coloured product was added to a volume of 15 mL cyclopentanone (A K Scientific) to provide a 13.2% (w/v) solution of the product. An amount of 0.235 g zinc chloride (Sigma-Aldrich) was added to a volume of 5 mL cyclopentanone (A K Scientific) to provide a 0.55% (w/v) solution of the catalyst.

(56) To a volume of 15 mL of the solution of product in a vial either a volume of 0.3 mL of succinyl chloride (Sigma-Aldrich) or an amount of 0.4 g of 1,3,5-benzenetricarboxylic chloride (Sigma-Aldrich) was added. Following one of these additions a volume of 1.5 mL of the solution of the catalyst was added to provide a mixture of product, cross-linker and catalyst in a molar ratio of 1:0.5:0.1.

(57) Following shaking of the vial the mixture was cast on a glass plate and cured at a predetermined temperature for a predetermined period of time. The temperatures and times used to provide samples of membrane consisting of putatively cross-linked chlorosulfonated poly(ether ether ketone) are presented in Table 1 and Table 2.

(58) For comparative purposes, samples of membrane were also prepared without the addition of cross-linker and catalyst to the cast mixture. The curing conditions used in the preparation of these latter samples are presented in Table 3.

(59) TABLE-US-00001 TABLE 1 Curing temperatures and times used for the preparation of putatively cross-linked chlorosulfonated poly(ether ether ketone) membranes where the crosslinking agent was succinyl chloride. Molar ratio Curing Curing (product to Molar ratio Sample temp time crosslinking (product to designation (±5° C.) (min) agent) catalyst) 120SCl15 120 15 1:0.5 1:0.1 120SCl30 120 30 1:0.5 1:0.1 120SCl45 120 45 1:0.5 1:0.1 120SCl60 120 60 1:0.5 1:0.1 120SCl75 120 75 1:0.5 1:0.1 120SCl90 120 90 1:0.5 1:0.1 150SCl15 150 15 1:0.5 1:0.1 150SCl30 150 30 1:0.5 1:0.1 150SCl45 150 45 1:0.5 1:0.1 150SCl60 150 60 1:0.5 1:0.1 150SCl75 150 75 1:0.5 1:0.1 150SCl90 150 90 1:0.5 1:0.1 150SCl105 150 105 1:0.5 1:0.1 160SCl15 160 15 1:0.5 1:0.1 170SCl15 170 15 1:0.5 1:0.1 150SCl302x 150 30 1:1   1:0.1

(60) TABLE-US-00002 TABLE 2 Curing temperatures and times used for the preparation of putatively cross-linked chlorosulfonated poly(ether ether ketone) membranes where the crosslinking agent was 1,3,5-benzenetricarboxylic chloride. Molar ratio Curing Curing (product to Molar ratio Sample temp time crosslinking (product to designation (±5° C.) (min) agent) catalyst) 120TMC15 120 15 1:0.33 1:0.1 120TMC30 120 30 1:0.33 1:0.1 120TMC45 120 45 1:0.33 1:0.1 120TMC60 120 60 1:0.33 1:0.1 120TMC75 120 75 1:0.33 1:0.1 120TMC90 120 90 1:0.33 1:0.1 150TMC15 150 15 1:0.33 1:0.1 150TMC30 150 30 1:0.33 1:0.1 150TMC45 150 45 1:0.33 1:0.1 150TMC60 150 60 1:0.33 1:0.1 150TMC75 150 75 1:0.33 1:0.1 160TMC15 160 15 1:0.33 1:0.1 170TMC15 170 15 1:0.33 1:0.1

(61) TABLE-US-00003 TABLE 3 Curing temperatures and times used for the preparation of chlorosulfonated poly(ether ether ketone) membranes without the addition of catalyst or crosslinking agent. Molar ratio Curing Curing (product to Molar ratio Sample temp time crosslinking (product to designation (±5° C.) (min) agent) catalyst) 120SClPEEK15 120 15 — — 120SClPEEK30 120 30 — — 120SClPEEK45 120 45 — — 120SClPEEK60 120 60 — — 120SClPEEK75 120 75 — — 120SClPEEK90 120 90 — — 150SClPEEK15 150 15 — — 150SClPEEK30 150 30 — — 150SClPEEK45 150 45 — — 150SClPEEK60 150 60 — — 150SClPEEK75 150 75 — — 150SClPEEK90 150 90 — — 150SClPEEK105 150 105 — — 150SClPEEK120 150 120 — — 160SClPEEK15 160 15 — — 170SClPEEK15 170 15 — —

(62) Characterisation of Membranes

(63) Solubilities of samples of membrane were determined at room temperature by placing a small piece (circa 25 mm.sup.2) of sample in a volume of 3 mL of a solvent. A lack of solubility in various solvents was indicative of the sample of membrane consisting of cross-linked polymer. The appearances and solubilities of the samples in the solvent cyclopentanone are presented in Table 4. The solubility of selected samples of membrane (150SClPEEK30, 150SCl30 and 150SCl302x) in the solvents acetone, dimethylsulfoxide and methanol and acid (nitric acid) and alkali (sodium hydroxide)) are presented in Table 5.

(64) Fourier transform infrared spectra (FTIR) were recorded using a Thermo Electron Nicolet 8700 spectrometer equipped with a single bounce ATR and diamond crystal. An average of 32 scans with a 4 cm.sup.−1 resolution were recorded for each sample. For comparative purposes the FTIR spectrum of a sample of membrane prepared by the casting of a mixture without the addition of crosslinking agent or catalyst (control sample) was also recorded. All samples were washed with deionised water before recording scans. Comparisons of the FTIR spectra recorded for the selected samples of membrane and the control sample are presented in FIGS. 1 to 5. The observed solubilities and FTIR spectra of the selected samples of membrane are consistent with crosslinking of the csPEEK substrate having occurred.

(65) TABLE-US-00004 TABLE 4 Appearance and solubility in cyclopentanone (after 24 hours) of samples of membrane. Sample Appearance Solubility 120SClPEEK15 Brown film Soluble 120SClPEEK30 Dark brown film Soluble 120SClPEEK45 Dark brown film Soluble 120SClPEEK60 Dark brown film Soluble 120SClPEEK75 Dark brown film Soluble 120SClPEEK90 Black film Soluble 150SClPEEK15 Dark brown film Soluble 150SClPEEK30 Black film Soluble 150SClPEEK45 Black film Soluble 150SClPEEK60 Black film, brittle Soluble 150SClPEEK75 Black film, brittle Soluble 150SClPEEK90 Black film brittle Partially soluble 150SClPEEK105 Black film, brittle Partially soluble 150SClPEEK120 Black film brittle Partially soluble 160SClPEEK15 Black film Soluble 170SClPEEK15 Black film Soluble 120SCl15 Brown film Soluble 120SCl30 Dark brown film Soluble 120SCl45 Dark brown film Soluble 120SCl60 Dark brown fim Soluble 120SCl75 Dark brown film Soluble 120SCl90 Dark brown film Soluble 150SCl15 Black film Partially soluble 150SCl30 Black film Insoluble 150SCl45 Black film, brittle Insoluble 150SCl60 Black film, brittle Insoluble 150SCl75 Black film, brittle Insoluble 150SCl90 Black film, brittle Insoluble 150SCl105 Black film, brittle Insoluble 160SCl15 Black film Soluble 170SCl15 Black film Partially soluble 150SCl302x Black film Insoluble 120TMC15 Brown film Soluble 120TMC30 Dark brown film Soluble 120TMC45 Dark brown film Soluble 120TMC60 Dark brown film Soluble 120TMC75 Dark brown film Partially soluble 120TMC90 Dark brown film Partially soluble 150TMC15 Black film Partially soluble 150TMC30 Black film, brittle Partially soluble 150TMC45 Black film, brittle Insoluble 150TMC60 Black film, brittle Insoluble 150TMC75 Black film, brittle Insoluble 160TMC15 Black film Soluble 170TMC15 Black film Partially soluble

(66) TABLE-US-00005 TABLE 5 Solubilities of selected samples (shaded, Table 4) in different solvents. Solvent 150SClPEEK30 150SCl30 150SCl302X Dimethylsulfoxide Dissolved Undissolved, Undissolved, swollen swollen Cyclopentanone Dissolved Undissolved, Undissolved, swollen swollen Acetone Undissolved, Undissolved, Undissolved, slightly slightly slightly swollen swollen swollen Methanol Undissolved, Undissolved, Undissolved, swollen slightly slightly swollen swollen Sodium hydroxide Undissolved Undissolved Undissolved solution (pH 13) Nitric acid (pH 2) Undissolved Undissolved Undissolved Cyclopentanone Undissolved, Undissolved Undissolved (after 24 hours in slightly sodium hydroxide swollen solution (pH 13)) Cyclopentanone Dissolved Undissolved, Undissolved, (after 24 hours in partially, swollen swollen nitric acid (pH 2)) swollen

Example 8

(67) An amount of 13.97 g of poly(ether ether ketone) (Victrex) having a density of 1.3 g cm.sup.−3 and a melt viscosity at 400° C. of 90 Pa.Math.s was added to a volume of 66.5 mL of chloroform (Fisher Chemicals) mixed with 3.5 mL thionyl chloride. An amount of 70 g cholorsulfonic acid (Nacalai Tesque Inc.) was then added and the mixture stirred for two hours at 50° C. The mixture was then washed twice with two separate volumes of 200 mL of chloroform and thrice with three separate volumes of 400 mL of chloroform. Washing was indicated to be complete when the volume of chloroform used in the washing remained colourless. The orange coloured, viscous residue remaining following the chloroform washings was washed repeatedly with volumes of deionized water until the pH of the wash water had increased to 5. The resulting whitish coloured product was then broken into small flakes and dried in a vacuum oven at 65° C. for four days. Following drying the product was added to a volume of 100 mL tetrachloroethylene and a volume of 20 mL thionyl chloride and stirred for five hours. The product was then dried in a vacuum oven at 65° C. overnight.

Preparation of Asymmetric Composite Membrane

Example 9

(68) The cooled crosslinking solution was cast onto a borosilicate glass plate to provide a wet film. A sheet of the sulfonated microporous poly(ethylene) was adhered by applying directly to the wet film ensuring full contact between the abutting surfaces of the film and sheet. The asymmetric composite was then transferred to an oven and cured for a period of ten to ninety minutes at a temperature of greater than 85° C., but not exceeding the melting point of the sheet. Following cooling the asymmetric composite membrane was removed from the glass plate in warm water.

Example 10

(69) The cooled crosslinking solution was cast onto a borosilicate glass plate to provide a wet film. The cast crosslinking solution was allowed to stand under ambient conditions (room temperature and non-condensing humidity) for a period of at least 30 minutes. A sheet of the sulfonated microporous poly(ethylene) was then adhered by applying directly to the film formed by coagulation ensuring full contact between the abutting film and sheet surfaces. The asymmetric composite was then transferred to an oven and cured for a period of ten to ninety minutes at a temperature of greater than 85° C., but not exceeding the melting point of the sheet. Following cooling the asymmetric composite membrane is removed from the glass plate in warm water.

Example 11

(70) A solution of chlorosulfonated poly(ether ether ketone) product obtained according to Example 8 was prepared at a concentration of 0.132 g/mL in cyclopentanone. To a volume of 15 mL of this solution an amount of 0.3 mL succinyl chloride (as cross-linking agent) and 1.5 mL zinc chloride (as catalyst) was added to provide a mixture of product, cross-linking agent and catalyst in a molar ratio of 1:0.5:0.1. The mixture was cast on a glass plate and the solvent evaporated at 85° C. before curing of the film at 120° C. for one hour. The cured film was then whetted with 50% tetrachloroethylene in chloroform before adhering a sheet of dry sulfonated microporous poly(ethylene) as the backing layer. To adhere the sheet of sulfonated microporous poly(ethylene) prepared according to Example 2, the sheet was flattened using 25% tetrachloroethylene in chloroform and 20% tetrachloroethylene in a mixture of 10% cyclopentanone and 90% chloroform followed by cyclopentanone. The asymmetric composite was then dried at 85° C. for 15 minutes and the membrane evaluated.

(71) Performance of Asymmetric Composite Membrane

(72) The performance of the asymmetric composite membranes prepared according to Examples 10 and 11 were evaluated using a reverse osmosis (RO) filter assembly of the type illustrated in FIG. 6.

(73) Flux Testing

(74) A section of the asymmetric composite membrane (1) was pre-wetted by dipping in distilled water and then placed on a coarse support mesh (2) located in the lower half (3) of the filter assembly housing, with a shim (4) optionally interposed. The section was placed with the rejection layer of the asymmetric composite membrane facing downwards. A fine mesh (5) located in the upper half of the filter assembly (6) housing was placed over the upper surface of the section of the asymmetric composite membrane (1). The filter assembly was sealed by sealing rings (7,8) and held in a hydraulic press pressurised to 60 Bar. The inlet port (9) of the lower half of the filter assembly housing (3) was in fluid connection with a feed reservoir (not shown) from which a feed stream was pumped at a rate to maintain the feed stream pressure measured on the pressure gauge (10). A pressure of 5 Bar was maintained for feed streams comprising water and salts. A pressure of 10 Bar was maintained for feed streams of milk. Feed streams were pre-chilled to 8° C. to mimic commercial processing conditions. Permeate was collected from the outlet port (11) of the upper half of the filter assembly housing (6) in a graduated cylinder (not shown). Collection was started at least 5 minutes after the commencement of permeate being discharged from the outlet port (11) in order to exclude water from the pre-wetting of the membrane or permeate from previously used feed streams.

(75) In Situ Cleaning of Membrane

(76) To mimic commercial processing operations the asymmetric composite membrane was subjected to repeated in situ washing protocols. The intermediate and subsequent flux rates were determined to assess the likely durability of the membrane in commercial processing operations. The in situ washing protocol was based on that employed in a commercial processing operation, but modified in duration to compensate for the greater exposure of the membrane to the cleaning agents (caustic and acid) in the filter assembly. Prior to the washing steps the membrane was rinsed by circulating water at an initial temperature of 65° C. through the filter assembly for a period of three minutes before draining the system.

(77) The membrane was subjected to a first wash by circulating a circa 2% (w/v) sodium hydroxide solution (“caustic wash”) through the filter assembly for a period of five minutes before draining and flushing the system by circulating water at an initial temperature of 65° C. through the filter assembly system for a period of five minutes. The membrane was subjected to a second wash by circulating a circa 2% (w/w) nitric acid solution (“acid wash”) through the filter assembly system for a period of ten minutes before draining and flushing the system of circulating water at an initial temperature of 65° C. for a period of ten minutes. The membrane was subjected to a third wash (a “caustic wash”) before flushing the system by circulating water at an initial temperature of 65° C. for a period of five minutes before circulating chilled water for a period of five minutes to cool the system. All rinsing and washing steps were performed with no pressure recorded on the pressure gauge (8).

(78) Samples of the asymmetric composite membrane were tested for tensile strength and burst strength following flux testing and in situ washing to assess the likely durability of the membrane in commercial processing operations. All testing methods were performed in accordance with ASTM standards (Anon (2009), Anon (2010) and Anon (2012)).

(79) Tensile Strength Testing

(80) Samples for testing were excised from a sheet of the asymmetric composite membrane used in flux testing using a die and hydraulic press. The die was positioned on the upper surface of the sheet of the asymmetric composite membrane supported on a wooden board. Sufficient pressure was applied to the die using the hydraulic press to cut through the sheet. The sample was carefully removed from the die and subjected to testing according to the following protocol.

(81) Samples were preconditioned at 23° C. plus or minus 2° C. and 50% plus or minus 10% relative humidity prior to testing. The thicknesses and widths of samples were measured at three points along the gage length of each sample. The calibrated load weighting system was zeroed and the machine crosshead adjusted to provide the required grip separation. Samples were placed in alignment in the grips of the universal testing machine ensuring sufficient tension on both edges of the sample. Where necessary, blotting or filter paper was used on the surface of the grips to prevent slippage. The extension indicator and recording system of the universal testing machine were reset before starting the machine and testing the samples to failure. Parts of samples tested to failure were removed from the universal testing machine and labelled. Data from testing of samples that failed outside the gage length, i.e. where contacting the grips of the universal testing machine or by tearing with an angle of separation greater than 30 degrees from the perpendicular, were excluded from analyses.

(82) Bursting Strength Testing

(83) The bursting strength of a sheet of the asymmetric composite membrane used in flux testing was measured (according to Mullen) using a tester (Burst-o-Matic™, Lorentzen and Wittre). Measurements were taken at multiple locations on the sheet. Data are presented in Table 6.

(84) TABLE-US-00006 TABLE 6 Comparison of burst pressures for the used asymmetric composite membrane and an unused commercially available porous poly(ethylene) (K2045, 20 μm thick, porosity 45%, CELGARD ™). Tensile strength (MPa) Sample Mean Standard deviation CELGARD ™ K2045 98.3 10 composite membrane 84.9 2.6

(85) Rejection Assessment

(86) Samples of permeate collected from a milk feed stream with periodic in situ cleaning of the membrane were analysed for fat, lactose, protein and total solids content. The results of these analyses are presented in FIG. 9.

(87) In Situ Disinfection of Membrane

(88) Prior to the washing steps the membrane was rinsed by circulating water at an initial temperature of 65° C. through the filter assembly for a period of three minutes before draining the system. The membrane was subjected to a first wash by circulating a circa 2% (w/v) sodium hydroxide solution (“caustic wash”) through the filter assembly for a period of five minutes before draining and flushing the system by circulating water at an initial temperature of 65° C. through the filter assembly system for a period of five minutes. The membrane was subjected to a second wash by circulating a circa 2% (w/w) nitric acid solution (“acid wash”) through the filter assembly system for a period of ten minutes before draining and flushing the system by circulating water at an initial temperature of 65° C. for a period of ten minutes. The membrane was subjected to a third wash by circulating a 2% (w/w) sodium hypochlorite solution (“disinfection wash”) through the filter assembly system for a period of five minutes before draining and flushing the system with circulating water at an initial temperature of 65° C. for a period of ten minutes. The membrane was subjected to a final wash (a “caustic wash”) before flushing the system by circulating water at an initial temperature of 65° C. for a period of five minutes before circulating chilled water for a period of five minutes to cool the system. As before all rinsing and washing steps were performed with no pressure recorded on the pressure gauge (8).

(89) Post Disinfection Flux Testing

(90) The performance of the asymmetric composite membrane was evaluated following exposure to 2% (w/w) sodium hypochlorite. Flux rates obtained for water as the feed solution at a pressure of 23 bar are provided in FIG. 10. Permeate flux rates obtained when the feed stream was homogenised milk (“blue top”) are provided in FIG. 11. Stable flux rates of ca. 15 litres per m.sup.2 per hour were obtainable. An operating pressure of 20 bar for a flux rate of 13 litres per m.sup.2 per hour was considered optimal for long-term use of the membrane.

(91) A section of the asymmetric composite membrane was cut to size and mounted in the assembly illustrated in FIG. 6 as described above. A feed stream of homogenised whole milk (blue top) at a temperature of 7° C. was pumped at a rate (600 rpm) to maintain a feed stream pressure of 20 bar. Collection of permeate was alternated with the following clean-in-place (CIP) protocol: Two volumes of two litres tap water at a temperature of 35° C. (flushed for 5 minutes each); Circulating sodium hydroxide (pH 12) at a temperature of 35° C. for a period of 10 minutes; One volume of one litre of tap water at a temperature of 35° C. (flushed for 5 minutes); Circulating hydrochloric acid (pH 1.522) at a temperature of 35° C. for a period of 10 minutes; One volume of one litre of tap water at a temperature of 35° C. (flushed for 5 minutes each); Circulating sodium hydroxide (pH 12) at a temperature of 35° C. for a period of 10 minutes; Two volumes of two litres tap water at a temperature of 35° C. (flushed for 5 minutes each); Rinsing with cold tap water for a period of 5 minutes.

(92) Ten samples of permeate from the milk feed stream were collected and independently analysed (Livestock Improvements Corporation, Hamilton) for fat, lactose, protein and total solids content. The results of these analyses are presented in FIG. 12. The asymmetric composite membrane provided consistently high rejection of fat, lactose, protein and total solids despite the repeated application of the CIP protocol. The performance of the membrane is consistent with its proposed use in food processing applications.

Example B

(93) In a second option, the backing layer is prepared by the photoinitiated graft polymerisation of a sheet of μPE with selected hydrophilicitizing agents (Table 7). The hydrophilicitizing agent is selected to provide graft polymers with the chemical and physical properties dictated by the intended use of the asymmetric composite membrane. The irradiation of the microporous substrate should be for a period of time sufficient to allow for completion of the crosslinking or grafting. The intensity of the irradiation is typically sufficient to provide for a period of time of less than 5 minutes. In the context of preparing water permeable membranes using acrylic acid as the hydrophilicitizing agent this selection of parameters has been found to provide a PE sheet suitable for use in ultrafiltration of feed streams such as whole milk. In the context of preparing a durable (i.e. chlorine tolerant) asymmetric composite membrane, the selection of 4-ethenyl-benzenesulfonic acid (SSS) as the hydrophilicitizing agent has been found to provide a hydrophilicitized sheet of μPE particularly suited for use in the preparation of the asymmetric composite membrane.

(94) Preparation of Hydrophilicitized Microporous Polyolefin (“Support Layer”)

(95) A microporous polyolefin substrate is contacted with a solution of 1% (w/v) photoinitiator and 6% (w/v) hydrophilicitizing agent in 1:1 (v/v) acetone-water. The contacted substrate is then irradiated at a wavelength optimal for the photoinitiator. When benzophenone is used as the photoinitiator a wavelength of around 270 nm is used. An intensity that is sufficient for a period of time of 5 minutes irradiation to be sufficient for the grafting to be complete is typically used. The irradiated substrate is finally washed using ultrasound in an excess of water followed by soaking in water. It was observed that a lower contact angle was achievable when irradiation of the contacted substrate occurred with the photoinitiator in solution (as opposed to being dried on the surface of the substrate).

(96) For the preparation of samples A to D of modified polyolefin substrate according to the general method, sheets (20 μm thickness) of porous (45% porosity, 0.08 μm average pore diameter) poly(ethylene) (CELGARD™ K2045, Celgard LLC) were used as the polyolefin substrate. Other sources of sheets of porous poly(ethylene) may be used, e.g. TARGRAY™ SW320H. The solution was prepared by mixing benzophenone (photoinitiator) with acetone before adding water and then the selected hydrophilicitizing agent. The polyolefin substrate was contacted with the solution by casting the hydrophilicitizing solution on the sheet and removing excess fluid on the surface of the sheet. Irradiation was for three and a half minutes using UV fluorescent lamps (250 nm) having a bulb irradiance of 0.1 mW m.sup.−1 at a distance of 50 mm. The ultrasound washing was for five minutes followed by soaking at 45° C. for three hours.

(97) For the preparation of Sample E amounts of 0.6 g of the hydrophilicitizing agent sodium 4-vinylbenzene sulphonate and 0.1 g of the photoinitator benzophenone were dissolved in water (5 mL) and acetone (5 mL). The solution was then applied to a microporous polyethylene sheet on a glass plate using a threaded rod. Three applications were made until the polyethylene was wetted out. The glass plate and sample were then placed in a polyethylene plastic bag then clamped to a frame and cured using fluorescent UV lamps at a distance of 5 cm on both sides of the sample. The peak wavelength of the lamps was 250 nm and an irradiance power of 0.2 to 0.4 mW/m for each lamp. The lamps were placed in a line with 50 mm centres. The time the samples were irradiated was 210 seconds. The samples were then washed in 45° C. water for 10 seconds to removed excess polymer and unreacted hydrophilicitizing agent and put in an oven to dry for 30 minutes at 65° C. The samples were then removed from the glass plate by immersion in a water bath and extracted in a beaker of deionised water for three hours. Sample F was prepared by the same method as used for the preparation of Sample E, but with a volume of 0.6 mL of the hydrophilicitizing agent acrylic acid being substituted for the hydrophilicitizing agent sodium 4-vinylbenzene sulphonate and added after the benzophenone was dissolved in the solvent. Sample G was prepared by the same method as used for the preparation of Sample E, but with a volume of 0.6 mL of the hydrophilicitizing agent 2-hydroxyethyl methacrylate being substituted for the hydrophilicitizing agent sodium 4-vinylbenzene sulphonate and added after the benzophenone was dissolved in the solvent. The properties of samples of modified polyolefin substrate prepared using different hydrophilicitizing agents were assessed.

(98) TABLE-US-00007 TABLE 7 Structure of AMPS, SSS and alternative hydrophilicitizing agents. Hydrophilicitizing agents Structure 2-acrylamido-1-methyl- 2-propanesulfonic acid (AMPS) 2-propen-1-ol (allyl alcohol) 2-propenoic acid (acrylic acid) 0 2-hydroxyethyl 2-methyl-2-propenoic acid ester (HEMA) 4-ethenyl-benzenesulfonic acid (as the sodium salt) (SSS)

(99) Characterization of Hydrophilic Microporous Polyolefin Samples

(100) Fourier Transform Infrared (FTIR)

(101) Spectra of the samples were recorded using a Thermo Electron Nicolet 8700 FTIR spectrometer equipped with a single bounce ATR and diamond crystal. An average of 32 scans with a 4 cm.sup.−1 resolution was taken for all samples.

(102) Surface Analysis

(103) The contact angles for the surfaces of the asymmetric composite membrane were determined in using the captive bubble method as described in the publication of Causserand and Aimar (2010). The samples were immersed in deionized water with the surface to be analysed facing downwards. An air bubble was trapped on the lower surface of the sample using a syringe. An image of the bubble was captured and the contact angle was calculated from its geometrical parameters.

(104) Permeability and Flux Testing

(105) Permeability was determined by measuring the flux in deionized water at various pressures starting at 20 bar and decreasing in 4 bar iterations. Flux J.sub.v was then graphed against effective pressure difference across the membrane, p.sub.eff, and the slope of the permeability L.sub.p.

(106) $L_p=\frac{J_V}{\Delta p_{\mathrm{eff}}}$

(107) Initial flux rates under pressure (20 bar) and no pressure were determined using the Sterlitech flux rig illustrated in FIG. 1 equipped with a PolyScience cooling unit. The samples were mounted in the flux cell and bolted. Deionized water was fed into the rig at 2.5 L min.sup.−1 and 4 to 8° C. The time to collect a predetermined volume of permeate was noted. The flux rate (J) was calculated according to the following equation:

(108) $J=\frac{V}{t\times A}$
where V is the permeate volume (L), t is the time (h) for the collection of V and A is area of the sample (m.sup.2) which was determined to be 0.014 m.sup.2.

(109) To assess durability in different environments tests were also performed on samples immersed for 60 to 70 hours in aqueous solutions of either 30% (w/v) potassium hydroxide (“alkali environment”) or 33% (w/v) hydrochloric acid (“acid environment”).

(110) Dry weight increases were calculated by taking the dry weight of the sample after it had dried in an oven for half an hour and comparing the weight to the initial weight of the porous polyethylene before grafting. Dry weights were taken after loose polymer had been extracted from the membrane and at the end of testing after a clean in place.

(111) $\Delta m_{\mathrm{dry}}=\frac{m_{\mathrm{dry}}-m_{\mathrm{initial}}}{m_{\mathrm{initial}}}\times 100$

(112) Water absorption was measured after loose polymer from the membrane had been extracted. The wet membranes were dabbed dry with a paper towel to remove surface moisture and weighed.

(113) $\Delta m_{\mathrm{Wet}}=\frac{m_{\mathrm{wet}}-m_{\mathrm{initial}}}{m_{\mathrm{initial}}}\times 100$

(114) Total solids rejection for whole milk samples was measured by pouring 20 mL of sample from the feed in a petri dish and measuring the dry weight after 2 hours in a 100° C. oven.

(115) $\%R_{\mathrm{TS}}=\left(1-\frac{m_{p,\mathrm{TS}}}{m_{f,\mathrm{TS}}}\right)\times 100$
where m.sub.p,TS is total milk solids in the permeate and m.sub.f,TS is the mass of milk total solids in the feed.

(116) Sodium chloride rejection was measured using a 2 g/L solution with a feed pressure of 16 bar. The conductivities from the feed and permeate were compared.

(117) $\%R_{\mathrm{NaCl}}=\left(1-\frac{{\sigma }_p}{{\sigma }_f}\right)\times 100$

(118) Where σ.sub.p is the conductivity of permeate and σ.sub.f is the conductivity of the feed.

(119) The bubble point of the dry membranes was determined by gradually increasing the pressure of the feed until permeate started to flow through the membrane.

(120) Results

(121) The FTIR spectra for samples A to D generally showed faint peaks compared to the peaks observed in the FTIR spectrum of the unmodified polyolefin substrate (CELGARD™ K2045, Celgard LLC) (see FIG. 13). However, the ester and carbonyl groups of Sample C were clearly discernible. The hydroxyl group peaks of Sample A and Sample D were barely evident. The FTIR spectrum for Sample B was not determined.

(122) The contact angles for samples A to D showed an inverse relationship with the permeability determined for the same sample (see FIGS. 14 to 16). Sample C was observed to have the lowest contact angle and the highest permeability prior to exposure to an acid or alkali environment. Following exposure to an acid environment the contact angle for Sample D increased. The contact angle of the unmodified polyolefin substrate (CELGARD™ K2045, Celgard LLC) was determined to be 89°, so modification of the surface tension is shown for all the samples despite the absence of definitive FTIR spectra. The observed initial flux rates were also consistent with modification of the polyolefin substrate (see Table 8).

(123) TABLE-US-00008 TABLE 8 Initial flux rates of samples of modified polyolefin substrate (CELGARD ™ K2045, Celgard LLC). Initial flux (Lm.sup.2min.sup.−1) Sample No pressure Pressure (20 bar) A (Allyl alcohol) 50 484 B (Acrylic acid) 43 555 C (HEMA) 61 772 D (SSS) 44 577

(124) All of samples A to D showed an increase in permeability compared to the unmodified membrane which measured 2.56 m s.sup.−1 Pa.sup.−1. When soaked for 66 hours in 30% (w/v) potassium hydroxide Sample A was stable based on a comparison of the permeability determined before and after exposure to the alkali environment. By comparison Sample D showed a large increase in permeability when exposed to the same alkali environment indicating the importance of the selection of the hydrophilicitizing agent when preparing modified polyolefin substrates for particular applications, e.g. alkaline battery separators. Furthermore, when immersed in 33% (w/v) hydrochloric acid Sample D turned the acid environment yellow and a strong odour of chlorine was detected, indicating oxidation of the modified polyolefin substrate. Notwithstanding this observation, the permeability of Sample D following exposure to the acid environment remained stable suggesting that the polyolefin substrate was not being degraded. When Sample B was exposed to the acid environment no colour change was observed, but the permeability decreased to less than the permeability of the polyolefin substrate, i.e. less than 2.56 m s.sup.−1 Pa.sup.−1).

(125) As a general rule the higher the observed contact angle the lower the permeability determined for a sample. After Sample D was exposed to an alkali environment the sample developed a high initial flux even though the contact angle was determined to remain high. This observation indicates that the structure of the modified polyolefin is degraded. Water absorption was observed to be greatest for Sample B and Sample D, and of these two samples, Sample D had the largest water absorption. Sample A had a larger water absorption than Sample C (see FIG. 17).

(126) Based on the assessment the preparation of modified polyolefin substrates according to the general method using 2-hydroxyethyl 2-methyl-2-propenoic acid ester as the hydrophilicitizing agent is selected for use as a backing or support layer in osmosis membranes. Sample C has been determined to provide high initial flux and the ability to let permeate through at low pressure differentials. Use of this class of modified polyolefin is indicated for medical applications.

(127) Based on the assessment the preparation of modified polyolefin substrates according to the general method using 2-propen-1-ol as the hydrophilicitizing agent is selected for use in applications having an alkali environment. Sample A maintained a relatively high permeability under these conditions.

(128) Based on the assessment the preparation of modified polyolefin substrates according to the general method using 4-ethenyl-benzenesulfonic acid as the hydrophilicitizing agent is selected for use in applications having an acid environment. Under these conditions Sample D maintained a more stable flux than Sample B exposed to the same conditions.

(129) The assessments of replicates (i, ii, iii, . . . ) of samples E, F and G are presented in Table 9 and FIGS. 18 to 24.

(130) TABLE-US-00009 TABLE 9 Assessments of replicates of Samples E, F and G. Hydrophilicitizing Sample B. Pt FluxMilk agent (replicate) Δm.sub.dry Δm.sub.wet bar B. Pt.sub.CIP 1 Θ Flux.sub.DI % R.sub.NaCl (Lm.sup.−2hr.sup.−1) % RTS 4-ethyenyl- E(i)  9%  10% 4 0 32 429 2% 16 66% benzenesulfonic E(ii)  7%  13% 4 4 60 114 3% 15 71% acid, Na salt E(iii)  7% 155% 4 0 55 213 5% 15 65% (SSS) Acrylic acid F(i) 10% 158% 0 0 33 208 9% 13 72% (AA) F(ii) 13% 165% 0 0 32 167 13%  147  8% F(iii) 16% 158% 0 0 30 208 41%  12 71% 2-hydroxyethyl G(i) 13%  64% 0 0 32 303 4% 20 50% 2-methyl- G(ii) 14%  57% 0 0 35 405 3% 44 46% 2-propenoic acid G(iii) 10%  68% 4 4 27 147 4% 51 46% ester G(iv) 10%  68% 0 0 31 385 2% 97 16% (HEMA)

(131) Sample F was observed to provide a water permeable membrane with the highest rejection of salt (sodium chloride) (FIG. 10) combined with a relatively high flux (FIG. 23) and rejection of total milk solids (FIG. 24). Based on this assessment the preparation of modified polyolefin substrates according to the general method using acrylic acid as the hydrophilicitizing agent is indicated for use as a membrane in the ultrafiltration of feed streams such as milk.

(132) The combination of a cross-linked poly(ether ether ketone) rejection layer and a hydrophilic microporous polyethylene backing layer provides a durable asymmetric composite membrane suitable for use in commercial processing operations.

(133) Preparation of the Asymmetric Composite Membrane (“Two-Step Method”)

(134) The membrane is prepared by adhering a sheet of hydrophilic microporous poly(ethylene) (μPE) to a film of putatively cross-linked sulfonated poly(ether ether ketone) (sPEEK). The adherence is augmented by the interpenetration of the two polymers. In the laboratory the membrane may be prepared according to the following method in which the sheet of hydrophilic μPE is nominally referred to as the ‘backing layer’ and the film of putatively cross-linked sPEEK is nominally referred to as the ‘rejection layer’. (The backing layer may alternatively be referred to as the ‘support layer’ and the rejection layer alternatively referred to as the ‘barrier layer’.) The method provides the advantage of being adaptable to the continuous production of the asymmetric composite membrane. The method is described in detail in respect of the preparation of a single sample.

(135) Rejection Layer

(136) Poly(ether ether ketone) (PEEK) (VICTREX™ 450P, Victrex, England) was sulfonated by heating to 50° C. in concentrated sulfuric acid (95%) for 1 and one half hours. The sulfonated PEEK (sPEEK) was precipitated and washed in ice water several times before being dried in a vacuum oven. Without wishing to be bound by theory it is believed the small amount of water present in the concentrated sulfuric acid prevents cross-linking attributable to the formation of sulfone bridges. The degree of sulfonation of the sPEEK was determined by titration according to a modified form of the method disclosed in the publication of Drioli et al (2003). The sPEEK was leached for three days in a 3M solution of sodium chloride (NaCl) and the resulting solution titrated against a 0.2 M solution of sodium hydroxide (NaOH) using phenolphthalein as indicator. An amount of sPEEK (0.2 g) with a 69% DS was then added to a volume of dimethylacetamide (DMAc) (2.7 mL) and sonicated until a clear to slightly cloudy dispersion was obtained.

(137) A volume (0.1 mL) of divinylbenzene (DVB) as crosslinking agent and an amount (0.14 g) of sodium styrene sulfonate (SSS) as hydrophilicitizing agent were added to a dispersion of sPEEK in DMAc. The dispersion contained 8% (w/w) sPEEK (0.216 mol/L) to provide a mixture containing a molar ratio of DVB to sPEEK of 1:2 and a molar ratio of SSS to sPEEK of 1:2. To increase the rate of the photoinitiated reaction an amount of benzophenone (BP) (8 μg) was added to the mixture before pouring onto aluminium foil on a glass plate, directly onto a glass plate or directly onto a stainless steel surface. The poured mixture was then exposed to 0.1 mW m.sup.−1 UVA fluorescent lamps (368 nm) at a distance of 50 mm for a limited time of 60 to 90 seconds to provide a semi-cured film. The photoinitiated reaction is conveniently performed under an atmosphere of air (without the need to provide an inert, e.g. nitrogen (N.sub.2), atmosphere). The structures of DVB and alternative di- and tetra-ethenyl cross-linking agents are provided in Table 10.

(138) TABLE-US-00010 TABLE 10 Structure of cross-linking agents. Cross-linking agents Structure o-Divinylbenzene (o-DVB) m-Divinylbenzene (m-DVB) p-Divinylbenzene (p-DVB) Ethylene glycol dimethacrylate (EGDMA) glyoxal bis (diallyl acetal) (GBDA)

(139) Backing Layer

(140) The sheet of sμPE to which a film of xsPEEK is adhered was prepared from a preformed sheet of microporous poly(ethylene) (μPE). The formation of sheets μPE is described, for example, in the publications of Fisher et al (1991) and Gillberg-LaForce (1994). In the present studies a preformed sheet of μPE (20 μm thickness, 45% porosity, 0.08 μm average pore diameter) (CELGARD™ K2045, Celgard LLC) was contacted with a solution of 1% (w/v) benzophenone and 6% (w/v) 4-ethenyl-benzenesulfonic acid (as the sodium salt) (SSS) as hydrophilicitizing agent in 1:1 (v/v) acetone-water. The solution was prepared by mixing benzophenone with acetone before adding water and then the hydrophilicitizing agent. The use of SSS is preferred due to the greater chlorine tolerance of membranes prepared using this hydrophilicitizing agent. This advantage applies to both the preparation of the hydrophilicitized backing layer and the asymmetric composite membrane.

(141) Asymmetric Composite Membrane

(142) The sheet of μPE contacted with the solution was laid on top of the semi-cured film (the nascent ‘rejection layer’). The composite of μPE contacted with the solution and semi-cured film of putative xsPEEK was then exposed as before to 0.1 mW m.sup.−1 UVA fluorescent lamps (368 nm) at a distance of 50 mm, but for a limited time of 210 seconds. The UVA-irradiated composite was then dried in an oven at 60° C. for 30 minutes to promote adherence of the film and sheet before releasing the composite membrane from the aluminium foil by immersion in a solution of 2% w/w sodium hydroxide or, if cured on a glass plate, by immersing the membrane in a water bath at room temperature until the membrane releases and floats to the surface (typically for 10 to 15 minutes). Where the nascent rejection layer is cured on a stainless steel surface it may be necessary to soak in water overnight. The structures of AMPS, SSS and alternative mono-ethenyl hydrophilicitizing agents are provided in Table 7. Before evaluation the laboratory prepared composite membrane was rinsed at 50° C. with a large excess of deionised (DI) water.

(143) Samples of the asymmetric composite membrane were prepared according to the foregoing method consisting of a rejection layer and a backing layer prepared using the compositions and conditions provided in Table 11 and Table 12.

(144) TABLE-US-00011 TABLE 11 Rejection layer formulations and cure conditions used in the preparation of each of the samples. The rejection layer of sample 12 was prepared using 1:1 (v/v) acetone-water as solvent. sPEEK DVB SSS BP % solids Cure time Number of Sample DS % of solids Solvent (w/w) (s) applications 1 69 45 22 31 2 DMAc 12 90 1 2 69 45 22 31 2 DMAc 12 60 1 3 69 45 15 33 6 DMAc 15 90 2 4 >80 41 17 30 11 DMAc 15 90 1 5 69 45 15 33 6 DMAc 15 90 2 6 69 98 0 0 2 DMAc 15 90 1 7 69 70 21 0 9 DMAc 9 90 1 8 69 57 35 0 8 DMAc 9 90 1 9 69 47 46 0 6 DMAc 9 90 1 10 69 42 52 0 6 DMAc 9 90 1 11 >80 63 32 0 5 MeOH 29 90 1 12 >80 15 10 70 5 acetone/water 6 300 1 13 69 45 19 34 2 DMAc 15 90 2

(145) TABLE-US-00012 TABLE 12 Backing layer formulations used in the preparations of each of the samples. All backing layers (except for sample 11 and sample 12) were prepared using 1:1 (v/v) acetone-water as solvent. Hydrophilicitizing H.A. BP % Cure agent % of solids time Number of Sample (H.A.) solids (w/w) (s) applications 1 AMPS 86 14 7 90 1 2 AMPS 86 14 7 600 2 3 SSS 86 14 7 90 1 4 SSS 86 14 7 90 1 5 SSS 86 14 7 90 1 6 SSS 86 14 7 90 1 7 SSS 86 14 7 90 1 8 SSS 86 14 7 90 1 9 SSS 86 14 7 90 1 10 SSS 86 14 7 90 1 11 n.a. n.a. n.a. n.a. n.a. n.a. 12 n.a. n.a. n.a. n.a. n.a. n.a. 13 SSS 86 14 7 90 1

(146) Preparation of the Asymmetric Composite Membrane (“One-Step Method”)

(147) The membrane may also be prepared by contacting one side of a sheet of microporous poly(ethylene) (TARGRAY™ SW320H) with a single mixture of the components of the formulations presented in Table 11 and Table 12. In this “one-step method” the hydrophilicitization of the poly(ethylene) of the microporous support layer occurs in conjunction with the crosslinking of the sPEEK of the rejection layer. According to this method a solution in a volume of 15 mL of dimethylacetamide (DMAc) or dimethylsulfoxide (DMSO) of 1.2 g 4-ethenyl-benzenesulfonic acid (as the sodium salt) (SSS), 0.2 g of benzophenone (BP), 0.8 g sPEEK and 0.14 g divinylbenzene is prepared and a volume of about 5 mL of the solution applied uniformly to the one side of a pre-weighed sheet of microporous poly(ethylene) (185 mm×135 mm) supported on a glass plate. The wetted sheet is then quickly transferred to a sealed poly(ethylene) bag to minimise evaporation of solvent and irradiated at a wavelength around 250 nm for a period of time of around 2 minutes. The irradiated sheet is then oven dried at 65° C. for a period of time of 30 to 45 minutes before being allowed to cool to room temperature and washed in a deionised water bath maintained at 50° C. for a period of time of 3 hours. The washed sheet is finally air dried at room temperature before storage. When preparing the solution dissolution of the sPEEK may require heating of the solvent, e.g. DMSO, to 60° C. Accordingly, the sPEEK should be added first and the heated volume of solvent cooled to room temperature prior to addition of the other components of the mixture. Exposure of the solution to light is avoided following addition of the photoinitiator, e.g. BP.

(148) Evaluation of the Asymmetric Composite Membrane (“Two-Step Method”)

(149) The performance of the asymmetric composite membrane was evaluated using a reverse osmosis (RO) filter assembly of the type illustrated in FIG. 6 as described above for the characterisation of hydrophilicitized microporous polyolefin sample. Flow rates of approximately 2 L/min were obtained.

(150) To mimic commercial processing operations the asymmetric composite membrane was subjected to ‘clean-in-place’ (CIP) protocols between each use of milk as the feed stream. The CIP protocols were based on those employed in a commercial processing operation for reverse osmosis (RO) membranes (Anon (2014)) and summarised in Table 7. The CIP protocols were repeated alternating with the use of milk as a feed stream. Samples were taken from the feed and permeate for each intervening use of milk as a feed stream to determine any deterioration in the performance of the membrane attributable to repeated CIP protocols. The asymmetric composite membrane was also evaluated for its tolerance to a CIP protocol including sodium hypochlorite (Table 14).

(151) TABLE-US-00013 TABLE 13 Clean-in-place (CIP) protocol adapted from Anon (2014). Time Temperature Step Wash.sup.1 (min) (° C.) 1 Water 5 Ambient 2 Water 5 35 3 Alkali 10 35 4 Water 5 35 5 Acid 10 35 6 Water 5 Ambient 7 Alkali 10 35 8 Water 5 Ambient .sup.1alkali (2% (w/v) NaOH) and acid (1.9% (w/v) H.sub.2NO.sub.3 and 0.6 (w/v) H.sub.3PO.sub.4).

(152) TABLE-US-00014 TABLE 14 Clean-in-place (CIP) protocol including 200 ppm free chlorine (as sodium hypochlorite). Time Temperature Step Wash.sup.1 (min) (° C.) 1 Water 5 Ambient 2 Water 5 35 3 Alkali 10 35 4 Water 5 35 5 Acid 10 35 6 Water 5 Ambient 7 Chlorine 10 35 8 Water 5 35 9 Water 1-2 35 10 Water 1-2 Ambient .sup.1alkali (2% (w/v) NaOH), acid (1.9% (w/v) H.sub.2NO.sub.3 and 0.6 (w/v) H.sub.3PO.sub.4) and chlorine (0.05% (w/v) sodium hydroxide and 0.09% (w/v) sodium hypochlorite).

(153) The following measurements relating to the performance of the asymmetric composite membrane before and after repeated application of the CIP protocols were made: 1. initial flux rates with water or whole milk as the feed stream after equilibration for 30 minutes; 2. rejection levels for fat, lactose and protein; 3. total solids content; 4. salt (NaCl or Na2SO4) retention; and 5. Sucrose retention.

(154) The total solids content was determined gravimetrically for both the feed and permeate. Samples were weighed in Petri dishes and dried in an oven at 60° C. for two hours and then 102° C. for a further two hours. The results are summarised in Table 15.

(155) Comparative Studies

(156) Sample 1

(157) The sample was subjected to repeated CIP protocols according to the schedule provided in Table 14 with the exception that Step 1 and Step 6 were also performed at 35° C. The maximum total solids rejection (standard milk) was observed after three CIP protocols with flux and total solids rejection stabilising after four to five CIP protocols (FIG. 25). Microscopic examination of the surface of the sample exposed to repeated CIP protocols indicated an increase in crystallinity of the membrane. It was found that increasing the concentration of the photoinitiator benzophenone (BP) used in the subsequent preparation of samples improved the reproduceability of these observations.

(158) TABLE-US-00015 TABLE 15 Performance of the samples of the asymmetric composite membrane measured at 20 bar. Standard milk Rejec- Deionised water Rejec- Rejec- tion Flux tion tion (total Flux Sample L/m.sup.2/h (gfd) (NaCl) (lactose) solids) L/m.sup.2/h (gfd) 1 40 (11.7) 52 99 99 12.1 (3.5) 2 18.1 (5.3) 47 98 99 10.1 (3.0) 3 9.5 (2.8) 46 90 97 9.4 (2.8) 4 50 (14.7) 64 75 97 14.7 (4.3) 5 9.5 (2.8) 46 91 6 (1.8) 6 1051 (308) 82 13.5 (4.0) 7 3.3 (1.0) 19 42 73 8.7 (2.6) 8 56 (16) 17 91 83 12.4 (3.6) 9 65 (19) 13 59 79 14 (4.1) 10 107 (31) 5 32 71 12.7 (3.7) 11 1.6 (0.5) 50 n.a. n.a. n.a. 12 83 (24) 25 13 100 (29) 38

(159) Sample 2

(160) The sample was subjected to repeated sequential CIP protocols according to the schedules provided in Table 13 (10×) and Table 14 (12×). The sample was then dried for several days before being subjected to further CIP protocols. The lactose rejection remained high throughout the sequential CIP protocols, the moderate decline in performance being recoverable following drying of the sample (FIG. 26).

(161) Sample 3

(162) The sample was subjected to repeated CIP protocols (25×) according to the schedule provided in Table 14. A total solids rejection (standard milk) comparable with that obtained for sample 1 was observed. A greater variability in flux was observed (FIG. 27).

(163) Sample 4

(164) The sample was subjected to repeated CIP protocols (17×) and exhibited an unacceptable decline in the rejection of total solids (FIG. 28). The unacceptable performance of this sample was attributed to the high DS (greater than 80%) of the sPEEK used in the preparation of the rejection layer.

(165) Sample 5

(166) The performance of the sample was evaluated when used to recover permeate from fresh raw milk over a prolonged period of time (18 hours) at a constant pressure of 16 bar. A performance comparable with that of existing commercial operations was observed.

(167) Sample 6

(168) The sample was prepared to demonstrate the advantage provided by the inclusion of both cross-linking and hydrophilicitizing agents in the preparation of the rejection layer. The performance of the sample before and after a single CIP protocol according to the schedule provided in Table 14 was compared with that of Sample 1. Whereas the performance of the latter in terms of total solids rejection improved, the performance of Sample 6 deteriorated. The poor durability of the sample is attributed to the absence of cross-linking and interpenetration of the polymers of the backing layer and rejection layer of the composite membrane.

(169) Samples 7 to 10

(170) These samples were prepared to evaluate the influence the proportion of SPEEK used in the preparation of the rejection layer had on performance (in the absence of the hydrophilicitizing agent SSS). The non-linear relationship between the proportion of SPEEK used and sodium chloride rejection is consistent with an expected increase in the electric field gradient of the membrane and corresponding rejection of charged species (FIG. 31). The optimal lactose and total solids rejection was obtained for the sample with a molar ratio of sPEEK:DVB of 0.6 (FIGS. 32 and 33). The molar ratio of sPEEK:DVB that provided optimal flux was dependent on the feed stream (FIG. 34). For water the flux was highest for the sample with the lowest molar ratio of 0.3. For milk the flux was highest for the samples with the lower molar ratios. For both feed streams a high molar ratio of sPEEK:DVB was incompatible with a high flux.

(171) Sample 11

(172) The sample was prepared using a high (greater than 80%) solids content when preparing the rejection layer. In addition, HEMA was substituted for SSS as the hydrophilicitizing agent due to the poor solubility of the latter in methanol. An extended curing period of 10 minutes was employed. At a pressure of 20 bar the sample provided a comparable sodium chloride rejection (FIG. 24) but at a negligible flux (FIG. 36).

(173) Sample 12

(174) The sample was prepared using an unmodified μPE as the backing layer. This necessitated the use of acetone/water as the solvent for the rejection layer formulation. Pursuant to the use of this solvent the proportion of sPEEK was reduced and the proportion of SSS increased with a total solid content of 6% (w/w). The curing was performed in a sealed polyethylene bag to prevent flush evaporation of acetone during the curing period of five minutes. The performance of the sample at 20 bar in terms of flux and sodium chloride and sucrose rejection was poor when compared with the performance of an analogous sample prepared using a grafted, hydrophilicitized backing layer.

(175) Evaluation of the Asymmetric Composite Membrane (“One-Step Method”)

(176) The performance over three clean-in-place (CIP) protocols of samples (090517Di, 090517Di and 090517Di) of the membrane prepared by contacting one side of a sheet of microporous poly(ethylene) (TARGRAY™ SW320H) with a single mixture of the components of the formulations presented in Table 11 and Table 12 was evaluated and the results presented in FIGS. 39 and 40.

(177) Although the invention has been described with reference to embodiments or examples it should be appreciated that variations and modifications may be made to these embodiments or examples without departing from the scope of the invention. Where known equivalents exist to specific elements, features or integers, such equivalents are incorporated as if specifically referred to in this specification. In particular, variations and modifications to the embodiments or examples that include elements, features or integers disclosed in and selected from the referenced publications are within the scope of the invention unless specifically disclaimed. The advantages provided by the invention and discussed in the description may be provided in the alternative or in combination in these different embodiments of the invention.

REFERENCED PUBLICATIONS

(178) Allmer et al (1988) Surface modification of polymers. I. Vapor-phase photografting with acrylic acid Journal of Polymer Science, Part A: Polymer Chemistry, 26(8), 2099-111. Allmer et al (1989) Surface modification of polymers. II. Grafting with glycidyl acrylates and the reactions of the grafted surfaces with amines Journal of Polymer Science: Part A: Polymer Chemistry, 27, 1641-1652. Ang et al (1980) Photosensitized grafting of styrene, 4-vinylpyridine and methyl methacrylate to polypropylene Journal of Polymer Science: Polymer Letters Edition, 18, 471-475. Anon (2009) Standard Practice for Cutting Film and Sheeting Test Specimens (D6287) ASTM International, 100 Barr Harbour Drive, PO Box C700, West Conshohocken, Pa. 19428-2959, United States. Anon (2010) Standard Test Method for Tensile Properties of Plastics (D638) ASTM International, 100 Barr Harbour Drive, PO Box C700, West Conshohocken, Pa. 19428-2959, United States. Anon (2012) Standard Test Method for Tensile Properties of Thin Plastic Sheeting (D882) ASTM International, 100 Barr Harbour Drive, PO Box C700, West Conshohocken, Pa. 19428-2959, United States. Anon (2014) DOW FILMTEC™ Membranes—Cleaning procedures for DOW FILMTEC FT30 elements Tech Fact (Form No. 609-23010-0211). Bai et al (2011) Surface UV photografting of acrylic acid onto LDPE powder and its adhesion Shenyang Huagong Daxue Xuebao 25(2), 121-125. Callahan et al (1990) Composite porous membranes and methods of making the same U.S. Pat. No. 4,976,897. Callahan et al (1992) Membranes from UV-curable resins U.S. Pat. No. 5,102,552. Causserand and Aimar (2010) 1.15 Characterisation of filtration membranes In Comprehensive membrane science and engineering Drioli, E; Giorna, L. eds. Oxford Elsevier. Choi (2002) Graft polymerisation, separators, and batteries including the separators U.S. Pat. No. 6,384,100. Choi (2004) Battery separator U.S. Pat. No. 6,680,144. Choi (2005) Graft polymerisation, separators, and batteries including the separators U.S. Pat. No. 6,955,865. Cussler et al (1992) Process for making microporous membranes having gel-filled pores and separations methods using such membranes U.S. Pat. No. 5,160,627. Deb et al (2007) Modification of sulfonated poly(ether ether ketone) with phenolic resin Polym. Adv. Technol. 18, 419-426. Di Vona et al (2008) Synthetic strategies for the preparation of proton-conducting hybrid polymers based on PEEK and PPSU for PEM fuel cells C. R. Chimie, 11, 1074-1081. Di Vona et al (2009) Analysis of Temperature-Promoted and Solvent-Assisted Cross-Linking in Sulfonated Poly (ether ether ketone) (SPEEK) Proton-Conducting Membranes J. Phys. Chem. B 2009, 113, 7505-7512. Donato et al (1993) Composite porous membranes U.S. Pat. No. 5,266,391. Donato (1994) Composite porous membranes U.S. Pat. No. 5,294,342. Donato et al (1994) Composite microporous membranes U.S. Pat. No. 5,294,346. Drioli et al (2003) Sulfonated PEEK-WC membranes for possible fuel cell applications Journal of Membrane Science 228 (2004) 139-148. Edge et al (1993) Surface modification of polyethylene by photochemical grafting with 2-hydroxyethylmethacrylate Journal of Applied Polymer Science, 47, 1075-1082. El Kholdi et al (2004) Modification of adhesive properties of a polyethylene film by phtografting Journal of Applied Polymer Science 92(5), 2803-2811. Fisher et al (1991) Microporous membranes having increased pore densities and process for making the same U.S. Pat. No. 5,013,439. Gao et al (2013) Radiation cross-linked lithium-ion battery separator with high rupture temperature and high tensile strength and manufacture method Chinese patent application no. 2013-10196439 (publ. no. CN 103421208). Gillberg-LaForce et al (1991) Modified microporous structures U.S. Pat. No. 5,049,275. Gillberg-LaForce (1994) Microporous membrane from cold-rolled precursor film U.S. Pat. No. 5,328,760. Hande et al (2008) Crosslinking of sulfonated poly (ether ether ketone) using aromatic bis (hydroxyethyl) compound, J. Membr. Sci., 322, 67-73. Hou et al (2012) Building bridges: Crosslinking of sulfonated aromatic polymers-a review J. Membr. Sci., 423-424, 113-127. Jones (1990) Polybenzimidazole thin film composite membranes U.S. Pat. No. 4,933,083. Kubota and Hata (1990a) Distribution of methacrylic acid-grafted chains introduced into polyethylene film by photografting Journal of Applied Polymer Science, 41, 689-695. Loeb (1981) The Loeb-Sourirajan Membrane: How It Came About In Synthetic Membranes, ACS Symposium Series, American Chemical Society, Washington, D.C. McCutcheon and Wang (2013) Osmotic Processes for a Sustainable 21st Century-Guest Editorial Desalination, 312, 1. Merle et al (2014) Friedel Crafts crosslinked highly sulfonated polyether ether ketone (SPEEK) membranes for a vanadium/air redox flow battery, Membranes 4, 1-19. Mikhailenko et al (2004) Proton conducting membrane based on cross-linked sulfonated poly(ether ether ketone) (SPEEK) J. Membr. Sci. 233, 93-99. Mikhailenko et al (2006) Properties of PEMs based on crosslinked sulfonated poly(ether ether ketone) J. Membr. Sci. 285, 306-316. Ogiwara et al (1981) Photosensitized grafting on polyolefin films in vapor and liquid phases Journal of Polymer Science: Polymer Letters Edition, 19, 457-462. Rhoden et al (2011) Low equivalent weight Friedel-Crafts cross-linked sulfonated poly(ether ether ketone) J. Membr. Sci., 376, 290-301. Shentu et al (2002) Factors affecting photo-grafting on low density polyethylene Hecheng Suzhi Ji Suliao 19(3), 5-8. Singleton et al (1993) Polymeric sheet International Application No. PCT/GB92/01245 (publ. no. WO 93/01622). Tazuke and Kimura (1978) Surface photografting. I. Graft polymerization of hydrophilic hydrophilicitizing agents onto various polymer films Journal of Polymer Science: Polymer Letters Edition, 16, 497-500. Xu and Yang (2000) Study on the mechanism of LDPE-AA vapor-phase photografting system Gaofenzi Xuebao (2000), 5, 594-598. Yao and Ranby (1990a) Surface modification by continuous graft copolymerization. I. Photoinitiated graft copolymerization onto polyethylene tape film surface Journal of Applied Polymer Science, 40, 1647-1661. Yao and Ranby (1990b) Surface modification by continuous graft copolymerization. III. Photoinitiated graft copolymerization onto poly(ethylene terephthalate) fiber surface Journal of Applied Polymer Science, 41, 1459-1467. Yao and Ranby (1990c) Surface modification by continuous graft copolymerization. IV. Photoinitiated graft copolymerization onto polypropylene fiber surface Journal of Applied Polymer Science, 41, 1469-1478. Ye et al (2009) Sulfonated poly (ether ether ketone) membranes crosslinked with sulfonic acid containing benzoxazine monomer as proton exchange membranes Polymer, 50, 3196-3203. Zhang and Ranby (1991) Surface modification by continuous graft copolymerisation. II. Photoinitiated graft copolymerization onto polypropylene film surface Journal of Applied Polymer Science, 43, 621-636.