COMPOSITE BIOMIMETIC MEMBRANES WITH ARTIFICIAL WATER CHANNELS

Abstract

The invention relates to the field of biomimetic membranes with artificial water channels, notably the use thereof in the context of the production and management of drinking water. The present invention relates to a biomimetic membrane with artificial water channels, the method of synthesis thereof, as well as the use thereof for desalination of brackish water and seawater, production of ultra-pure water or filtration of contaminants.

Claims

1. Composite biomimetic membrane, characterized in that it comprises: an ultrafiltration supporting membrane, at least one compound of formula I: ##STR00007## in which, R represents a linear or branched C4 to C8 alkyl group, preferably selected from butyl, pentyl, hexyl, heptyl or octyl, X represents S or O, and a crosslinked polyamide film, and in that the at least one compound of formula I is in the form of supramolecular aggregates of the imidazole quartet type, distributed homogeneously in the rigid matrix formed by the crosslinked polyamide film.

2. Membrane according to claim 1, in which the ultrafiltration supporting membrane has a molecular weight cutoff within a range from 10 to 250 kD.

3. Membrane according to claim 1, in which the supramolecular aggregate is a crystalline aggregate with an average diameter from 20 to 40 nm, preferably consisting of self-organized lamellar phases containing artificial water channels of the imidazole-quartet type.

4. Membrane according to claim 1, in which the compound of formula I is selected from the compounds of formula II: ##STR00008## in which R.sup.1 represents butyl or hexyl.

5. Membrane according to claim 1, in which the polyamide film has a thickness from 0.05 to 0.4 μm, preferably from 0.08 to 0.15 μm.

6. Membrane according to claim 1, in which the polyamide film is a crosslinked polyamide of a di- or triamine monomer and of a di or tri acyl chloride monomer, at least one of the monomers being trivalent.

7. Method of manufacture of the membrane according to claim 1, comprising the steps: a) impregnating the surface of an ultrafiltration supporting membrane with a colloidal suspension comprising at least one compound of formula I in the form of supramolecular aggregates: ##STR00009## in which, R represents a C4 to C8 alkyl group, preferably selected from butyl, pentyl, hexyl, heptyl or octyl, X represents S or O; b) formation of a crosslinked polyamide film by interfacial polymerization on the surface of the impregnated membrane obtained in step a), and production of the composite biomimetic membrane.

8. Method according to claim 7, in which the colloidal suspension comprises an organic solvent and/or water.

9. Method according to claim 7, in which the interfacial polymerization comprises the substeps: i) impregnating the surface of the impregnated membrane obtained in step a) with a solution comprising a di- or triamine monomer; ii) impregnating the surface of the impregnated membrane obtained in step i) with a solution comprising a di or tri acyl chloride monomer; and iii) polymerizing the impregnated membrane obtained in step ii) by immersion in water, at a temperature greater than or equal to 50° C.

10. Method according to claim 7, further comprising a step of rinsing the composite biomimetic membrane.

11. Use of a composite biomimetic membrane according to claim 1 for desalination of drinking water, brackish water or seawater.

Description

BRIEF DESCRIPTION OF THE FIGURES

[0082] FIG. 1 shows a schematic procedure of the method according to the invention: preparation of a composite biomimetic membrane PA-AWC according to the invention (I4RO): impregnation of the ultrafiltration supporting membrane of PS20-GPET with an aqueous solution containing supramolecular aggregates (AWC) followed by impregnation with an aqueous solution comprising the amine monomer MPD, followed by impregnation with an organic solution containing the acid chloride monomer TMC followed by the reaction of interfacial polymerization (IP) resulting in the production of a polymer film of polyamide (PA) incorporating the water channels within the crosslinked biomimetic layer PA-AWC according to the invention.

[0083] FIG. 2 shows photographs for scanning electron microscopy (SEM) showing the cross-section of the reference membranes without TFC channels and biomimetic membranes I4RO according to the invention; as well as the nanometric organization of the supramolecular aggregates of the I4s within the polyamide matrix. A) TFC reference membrane (magnification×240000), B) biomimetic membrane I4RO (magnification×240000) according to the invention and C) detail of the thin film of polyamide PA showing the homogeneous distribution of the supramolecular aggregates of I4 (AWC) (30-40 nm white spots) on the entire surface of the PA-AWC polymer (magnification×480000).

[0084] FIG. 3 shows images of atomic force microscopy AFM of A) the TFC reference membrane (left) and the biomimetic membranes B) H-I4RO_1 and C) L-I4RO_2.

[0085] FIG. 4 shows EDX surface analysis and the elementary surface distribution a) of the TFC reference membrane and b) the hybrid biomimetic membrane H-I4RO_1-HC6.

[0086] FIG. 5 shows the laboratory system used for measuring permeability to water, flow and solute rejection.

[0087] FIG. 6 shows the performance in desalination a) of brackish water (18 bar) or b) of seawater (65 bar) using [0088] the membranes according to the invention: a) H-I4RO_1, b) H-I4RO_2, c) H-I4RO_3, d) H-I4RO_4, e) L-I4RO_2, f) L-I4RO_3, [0089] the reference membranes: g) TFC, I) DTAB (in AWC), [0090] the commercial membranes: h) BW30, i) XLE, j) T82V, k) SW30, k1) SW30HR, k2) SW30XLE, m) GE_AG, u) BW60 [0091] the nanocomposite membranes: n) TFN-ZIF8, o) TFN-silicalite, p) TFN-CTN (Duan et al. J. Membrane Sci., 2018, 361, 682-686) [0092] the functionalized polyamide membranes: r) REFPA, s) HFAPA, (La et al. J. Membrane Sci. 2013, 437 33-39, t) AQP, (Wang et al. J. Membrane Sci. 2012, 423-424, 422-428), v1) 3DPAN450TFC_1, v2) 3DPAN450TFC_2, v3) 3DPAN450TFC_3, v4) 3DPAN450TFC_4 (Chuwdhury et al. Science, 2018, 361, 682-686).

[0093] FIG. 7 illustrates the difference in structure between membranes obtained from HC8 nano-crystalline solutions: a) at room temperature (25° C.) and b) at 60° C., and membranes obtained from colloidal solutions of c) HC4, d) HC6 at room temperature according to the invention.

[0094] The invention is further illustrated by the following examples, in a non-limiting manner.

Example 1: Synthesis of Biomimetic Membranes I4RO According to the Invention

[0095] Materials and Method

[0096] The compound HC6 (compound of formula I in which X=O and R=hexyl) was synthesized according to the methods described in Le Duc and Licsandru (Y. Le Duc, et al., Angew. Chem. Int. Ed. 2011, 50(48), 11366-11372; E. Licsandru, et al., J. Am. Chem. Soc., 2016, 138, 5403-5409). Metaphenylenediamine MPD, trimesoyl chloride TMC, and hexane were purchased from ALDRICH. The commercial supporting membranes used are of type M-PS20-GPET (Nanostone Water, USA) and M-PS35-GPP (Solecta Membranes, USA).

[0097] Dissolution of a compound of HC6, from 31.6 to 65.1 mM depending on the product used (see Table 1 below) in 5 mL of ethanol.

[0098] Addition of 1 mL of deionized water (alcohol:water volume ratio 85/15 v/v) followed by homogenization with ultrasound for 15 min. The solution is then left to stand, without stirring, for a period of 1 to 3 hours before use. This solution contains aggregates of nanometric dimensions (of 300-400 nm) (14).

[0099] The surface of an ultrafiltration supporting membrane of type M-PS20-GPET or M-PS35-GPP is then impregnated with the solution comprising the 14 aggregates. Said solution is maintained on the surface of the support, at room temperature, for about 1 min to 5 min, so as to increase evaporation of the solvent and ensure uniform dispersion of the product in the pores at the surface of said support.

[0100] Formation of the polyamide film PA is then carried out, for example according to a method of interfacial polymerization (Cadotte, J. E., 1981, U.S. Pat. No. 4,277,344 A.C). The amounts of each of the precursors used for preparing the TFN-AWC membranes given below are presented in Table 1 shown hereunder.

[0101] The ultrafiltration supporting membranes M-PS20-GPET and M-PS35-GPP are fixed with adhesive tape on a stainless steel base so that the polymerization reaction is only produced on the available surface.

[0102] The metaphenylenediamine MPD is dissolved in deionized water. The use of a solution of MPD of 0.32 M leads to formation of the high-density membranes (H-I4RO), the use of a solution of MPD of 0.19 M leads to formation of the low-density membranes (L-I4RO), whereas the use of a solution of MPD of 0.13 M and 0.17 M leads to formation of the very low density membranes (XL-I4RO). The surface of the ultrafiltration membrane M-PS20-GPET or M-PS35-GPP (Table 1) is impregnated by immersion for 120 seconds in the solution of MPD to obtain uniform and homogeneous dispersion.

[0103] The membrane is dried using an air blower to remove any excess solution on the surface of the membrane or trace of residual compounds.

[0104] Trimesoyl chloride TMC is dissolved in hexane (of 3.7 mM for the high-density membranes (H-I4RO) and low-density membranes (L-I4RO), at 2.5 mM for the very low density membranes (XL-I4RO)) by uniform dispersion on the surface of the ultrafiltration membrane M-PS20-GPET or M-PS35-GPP (Table 1), by immersion of the membrane in the solution for 60 seconds.

[0105] Interfacial polymerization takes place and the thin film of polyamide (of reference TFC without AWC or of biomimetic membrane I4RO) is formed.

[0106] Formation of the thin films PA-TFC or I4RO on the surface of the supporting membranes, with a thickness of about 0.1 to 0.4 μm, is observed.

[0107] The membranes obtained are then immersed in deionized water at 95° C. for about 120 seconds, followed by rinsing with an aqueous solution of 200 ppm NaOCl for about 120 seconds, followed by immersion in an aqueous solution of 1000 ppm Na.sub.2S.sub.2O.sub.5 for about 30 seconds, finally being immersed in deionized water at 95° C. for about 120 seconds.

[0108] The membranes are stored in deionized water at 4° C. or protected with a protective layer of glycerin prior to any use for filtration.

TABLE-US-00001 TABLE 1 Biomimetic composite membranes, high (H-I4RO), low (L-I4RO) or very low (XL-I4RO) density Membrane I4 conc. MPD conc. TMC conc. I4RO I4 (mM) (mol/L) (mM) support H-I4RO_1 HC6 31.6 0.32 3.7 M-PS20- H-I4RO_2 HC6 47.5 0.32 3.7 GPET H-I4RO_3 HC6 54.6 0.32 3.7 H-I4RO_4 HC6 65.1 0.32 3.7 L-I4RO_1 HC6 47.5 0.19 3.7 L-I4RO_2 HC6 54.6 0.19 3.7 L-I4RO_3 HC6 54.6 0.19 3.7 M-PS35- L-I4RO_4 HC6 57.7 0.19 3.7 GPP XL-I4RO_1 HC6 54.6 0.17 2.5 XL-I4RO_2 HC6 54.6 0.13 2.5

Example 2: Methods for Static Characterization and Characterization in Filtration Processes, of the Composite Biomimetic Membranes

[0109] Analyses in transmission electron microscopy revealed a very high degree of organization at the nanometric level (see FIG. 2). In fact, these photographs show the formation of the thin films with a thickness of 300-350 nm for the reference membranes TFC in polyamide PA, which is greatly reduced to 70-150 nm for the biomimetic membranes I4RO. These structures have a morphology of the crests and valleys with a thickness of 30 nm for the reference membranes TFC in polyamide PA, which is greatly reduced to 15-20 nm for the biomimetic membranes I4RO. It is also remarkable to observe the distribution of the artificial water channels AWCs within the polymer matrix, for which the SEM images highlight the formation of the supramolecular clusters at a scale of 30-40 nm, forming regions of nanometric channels arranged parallel to one another and very uniformly within the polyamide matrix.

[0110] Atomic force microscopy (AFM): Nanoman with electronic system Nanoscope 5 (Bruker Instruments) was used for measuring the surface roughness of the TFC and I4RO membranes. The sample was placed directly on the motorized XY stage, using the Tapping mode. The measuring tips for digitization of the samples were from Nanosensors, PPP NCSTR, with a nominal spring constant of 7 N/m and a typical radius less than 5 nm. All the images were acquired with a sampling resolution of at least 512 pts/512 lines using a scanning frequency of 5.5 Hz for 5 μm.sup.2 (size of the image) and of 0.65 Hz for smaller sizes. These AFM photographs (see FIG. 3) show formation of the thin films with considerable roughness for the reference membranes TFC in PA, which is greatly reduced the I4RO hybrid biomimetic membranes.

[0111] Energy-dispersive X-ray spectroscopy (EDX): Energy-dispersive X-ray spectroscopy (EDX) of active layers of polyamide was carried out with an AZTEC system, Oxford Instruments, UK. The samples were transferred onto a carbon strip of 0.5 cm.sup.2 and an accelerating voltage of 10 kV at a working distance of 8.5 mm was used for EDX. At least three spectra were captured at different positions for calculating an average atomic percentage of C, N and O in the active layers of polyamide. After each scan, the EDX analysis software created a table containing the elemental composition of C, N and O. In fact, these analyses show (see FIG. 4 and Tables 2 and 3 below) enrichment of C % and N % for the I4RO biomimetic membranes in comparison with the PA-TFC reference membrane

TABLE-US-00002 TABLE 2 EDX surface analysis and the elementary surface distribution of the TFC reference membrane. Element wt % at % C 67.50 76.38 N 3.70 3.59 O 18.39 15.62 S 10.41 4.41 Total 100.00 100.00

TABLE-US-00003 TABLE 3 EDX surface analysis and the elementary surface distribution of the L-I4RO_4 biomimetic membrane. Element wt % at % C 76.00 82.51 N 3.90 3.63 O 13.92 11.35 S 6.17 2.51 Total 100.00 100.00

[0112] Filtration system: Measurements of the coefficient of permeability to pure water (also called permeability to water), flow of permeate and solute rejection, as well as experiments of countercurrent washing, were carried out using a cross-flow laboratory system for all the membranes, including the TFC reference membrane and the I4RO hybrid biomimetic membranes with different natures: I4RO_1, as well as the commercial reverse osmosis membranes (FIG. 5). The laboratory system comprises a high-pressure pump (Hydra-cell pump, Wanner Engineering, Inc., Minneapolis, Minn.), a feed tank, a flat-membrane receiving cell, systems for temperature control and data acquisition. The receiving cell consists of a rectangular channel with a length of 7.6 cm, width of 2.8 cm and height of 0.3 cm. The active surface area of the membrane sample is therefore 23 cm.sup.2. The transverse flow rate was controlled by a floating-disk rotameter and adjusted, with the operating pressure, by means of a by-pass valve and a back pressure regulator (Swagelok, Solon, Ohio), whereas the flow rate of permeate was measured automatically for 60 s using a balance with a computer interface. The temperature was controlled by a recirculating cooler (model MC 1200, Lauda, Lauda-Königshofen) with a stainless steel coil immersed in the feed tank.

[0113] Measurements of permeability to water (A), flow and solute rejection: Before each experiment, the membrane was immersed in water overnight. After loading the membrane sample in the cell, the sample was compacted for 6 hours at an applied pressure of 20 bar (290 psi) or 70 bar (1015 psi) until stabilization to a state of equilibrium of the membrane. The applied pressure was then lowered to a value of 18 bar (261 psi) or 65 bar (943 psi) and the flow of pure water, Jw,0, was calculated by dividing the volumetric level of permeate obtained in the state of equilibrium by the surface area of the membrane. The membrane's coefficient of permeability to pure water, “A”, was calculated from the value of Jw,0, as indicated in Table 4.

TABLE-US-00004 TABLE 4 Formulae used for calculating membrane performance Intrinsic permeability to A = J.sub.w.0/ΔP [LMH/bar] water Solute permeability B = J.sub.w((1 − R)/R)exp(−J.sub.w/k) [LMH] coefficient Rejection R = 1 − c.sub.p/c.sub.b. [—] Flow of permeate with a Jw, 0 [LMH] feed solution of Flow of permeate with a J.sub.w [LMH] feed solution containing 35 g/L of NaCl at pH 8 Pressure difference applied ΔΠ [Bar] Mass transfer coefficient in k [m/s] cross flow operating mode Concentration of the c.sub.p [mg/L] permeate Concentration of the c.sub.b [mg/L] feed phase

[0114] Solute rejection tests were then carried out with a constant cross flow of 4.5 L/min (transverse flow velocity of 0.89 m/s), and an applied pressure of 65 bar (943 psi). The feed stream consisted of an aqueous solution of NaCl 100 mM (3000 ppm) for filtration at 18 bar (Table 5) or 600 mM (35000 ppm) for filtration at 65 bar (Table 6). The pH (8.0±0.1) is adjusted with sodium bicarbonate. The total volume of the feed solution at the start of the rejection test was of 5 L. The total duration of each rejection experiment was about 8 hours. After reaching the state of equilibrium, the flow of permeate, Jw, was calculated by dividing the volumetric level of permeate by the surface area of the membrane. The observed value of solute rejection was therefore calculated from the concentrations of solute in the feed stream and permeate stream, as indicated in Table 4. The concentrations of solute in the feed and permeate streams were found from the electrical conductivity measured using a calibration line. Three different observed rejection values were obtained, one every 30 minutes, and the three values were averaged. The permeability coefficient of solute “B” was also calculated as indicated in Table 4. For all the experiments, the temperature of the feed water was kept constant at 27±1° C.

[0115] Mechanical resistance to countercurrent washing: The membrane sample was first compacted for 12 h at an applied pressure of 70 bar (1015 psi) using pure water as feed solution. This step was followed by tests of permeability to water, flow rate and rejection carried out in the same operating conditions and at the same concentrations as described in the preceding section. Consequently, the values of A, B and the rejections observed were calculated for each membrane. Cycles of countercurrent washing were then carried out by inducing reverse flow of the solution of permeate from the support to the feed side of the membrane. After depressurization of the feed solution, the permeate flow pressure was increased to a value of 116 psi (6 bar), thus inducing reverse flow of permeate, for 20 min. The feed solution was allowed to flow with a transverse flow of 4.5 L/min during the countercurrent washing. After 20 minutes, the permeate solution was depressurized, and the feed solution was repressurized: tests of permeability to water, flow and rejection carried out in the same operating conditions and at the same concentrations as described previously, followed by determination of the values of A, B and observed rejections. Three cycles of countercurrent washing were carried out.

[0116] Chemical stability: In order to test the chemical stability of the membranes, they were stored for 30 minutes in different solutions containing: (i) sodium dodecyl sulphate 34 mM (at pH 11.5) or (ii) ethanol. After this step, the membranes were carefully rinsed with deionized water and tested as described above.

[0117] The results obtained for the saline solutions are shown in Tables 5 to 8 and in FIG. 6. Filtration results for organic contaminants such as paracetamol or urea are also shown.

TABLE-US-00005 TABLE 5 LPRO I4RO MEMBRANES (NaCl 3000 ppm, ΔP = 18 bar) Average (AVG) NaCl Flow of Permeability to Flow, NaCl Permeability Actual and standard rejection deionized water, A solution to NaCl B rejection Membrane deviation (STD) observed (%) water (LMH) (LMH/bar) 100 mM (mL/min) (LMH) (Rr, %) H-I4RO_1 AVG 99.62 32.76 1.82 1.31 0.07 99.74 STD ±0.18 ±2.55 ±0.14 ±0.01 ±0.03 ±0.12 H-I4RO_2 AVG 99.31 51.52 2.86 2.04 0.15 99.63 STD ±0.24 ±8.25 ±0.46 ±0.38 ±0.06 ±0.14 H-I4RO_3 AVG 98.92 69.35 3.85 2.61 0.26 99.51 STD ±0.28 ±6.8 ±0.38 ±0.17 ±0.07 ±0.14 H-I4RO_4 AVG 95.17 50.31 2.8 1.93 1.11 97.16 STD ±1.97 ±4.64 ±0.26 ±0.16 ±0.44 ±1.28 L-I4RO_2 AVG 99.07 61.45 3.41 2.16 0.21 99.52 H-I4RO_3 AVG 99.1 85.54 1.65 2.8 0.22 99.62 TFC AVG 98.62 24.4 0.36 1.04 0.22 98.98 STD ±0.07 ±2.89 ±0.16 ±0.02 ±0.01 ±0.06 SW30HR AVG 98.52 38.14 2.12 1.51 0.3 99.06 BW60 AVG 97.17 107.34 5.96 3.47 0.71 99.01 T82V AVG 96.9 29.68 1.65 1.24 0.56 97.82

TABLE-US-00006 TABLE 6 SWRO AWC MEMBRANES (NaCl 35000 ppm, ΔP = 65 bar) Average (AVG) NaCl Flow of Permeability Flow, NaCl Permeability Actual and standard rejection deionized to water, A solution to NaCl B rejection Membrane deviation (STD) observed (%) water (LMH) (LMH/bar) 600 mM (mL/min) (LMH) (Rr, %) H-I4RO_1 AVG 99.67 111.94 1.72 1.49 0.07 99.82 H-I4RO_2 AVG 99.32 119.36 1.84 2.26 0.17 99.73 STD ±0.24 8.49 0.46 0.27 0.06 0.12 H-I4RO_3 AVG 99.62 140.23 2.16 2.61 0.09 99.87 STD ±0.03 21.43 0.33 0.17 0.01 0.02 TFC AVG 99.25 80.36 1.24 1.66 0.18 99.64 STD ±0.54 17.99 0.28 0.25 0.25 0.25 SW30HR AVG 99.72 103.59 1.59 1.85 0.07 99.87 STD ±0.02 11.78 0.18 0.23 0.01 0.02 L-I4RO_2 AVG 99.49 144.92 2.23 2.09 0.13 99.79 L-I4RO_3 AVG 99.53 185.1 2.84 3.16 0.11 99.88

TABLE-US-00007 TABLE 7 LPRO(2) I4RO MEMBRANES (NaCl 2000 ppm, ΔP = 15.5 bar) Average (AVG) NaCl Permeability Permeability Actual and standard rejection to water, A FLOW to NaCl B rejection Membrane deviation (STD) observed (%) (LMH/bar) (LMH) (LMH) (Rr, %) XL-I4RO_2 AVG 99.1 5.73 78.64 0.634 99.2 STD 0.26 0.17 5.33 0.107 0.29

TABLE-US-00008 TABLE 8 XLPRO I4RO MEMBRANES (TAP WATER, ΔP = 6 bar) Observed Rejection of Average rejection of Ca2+ (AVG) and salts, determined by standard determined by ion-exchange deviation measurements FLOW chromatography Membrane (STD) of conductivity (LMH) (%) L-I4RO_3 AVG 98.96 25.5 98.96 STD 0.03 2.12 0.39 L-I4RO_4 AVG 99.38 29 98.68 XL-I4RO_1 AVG 99.39 29.1 99.46 XL-I4RO_2 AVG 98.38 30.68 98.88 STD 0.5 1.81 0.27 TW30 AVG 95.57 30.09 96.68

Example 3: Comparison of the Efficiency of a Biomimetic Membrane Comprising I-Quartet (According to the Invention) and a Membrane Comprising Nano-Crystals

[0118] The efficiency of a biomimetic membrane prepared according to the invention is compared with a biomimetic membrane prepared according to a conventional method, which does not use a colloidal suspension, but a nano-crystallization method: 50 μL 100 or 200 mM concentrated methanol solution of HC8 is added to a 10 ml aqueous solution of MPD to obtain a suspension of HC8 nanocrystals (concentration: 0.49 mM or 0.99 mM) for a total volume of 10.05 mL; (MeOH:H2O=0.5:99.5) according to the procedure described in Itsvan Kocsis, Supramolecular artificial water channels: from molecular design to membrane materials, Thesis to obtain the degree of Doctor of the University of Montpellier in Chemistry and Physicochemistry of Materials, 6 May 2018 (http://hal.umontpellier.fr/tel-01684404/document).

[0119] Table 9 below shows a comparison of the performance (WP=water permeability and salt rejection, %) of the AWC-HC4 and AWC-HC6 membranes according to the invention and of a reference thin-film composite membrane with the AWC-HC8-25 and AWC-HC8-60 membranes repaired according to the method of Itsvan Kocsis document and a reference thin-film composite membrane (TFC for “thin-film composite membrane”).

TABLE-US-00009 TABLE 9 MEMBRANE PERFORMANCE Brackish water reverse osmosis Observed salt Membrane WP/LMH/bar rejection, %. AWC-HC4 2.09 99.25 AWC-HC6 3.85-6.6 99.51-98.73 TFC (reference) 1.36 98.62 AWC-HC8-25 1.15-1.2 97.1 AWC-HC8-60 1.35 95.05 TFC (reference) 1.17 97.06

[0120] For the membranes obtained by the nano-crystallisation method (not included in the invention), a slight increase in permeability with a decrease in selectivity is observed compared to the TFC reference membrane. These results are related to the creation of micro defects in the polyamide film structure at the interface with the nanocrystals.

[0121] On the contrary, the membranes according to the invention AWC-HC4 and AWC-HC6 show clear improvements in permeabilities and with significant selectivities.

[0122] FIG. 7 illustrates the difference in structure between the membranes obtained from the nanocrystalline solutions and the membranes obtained from the colloidal solutions. Thus, it can be seen that the membranes according to the invention are homogeneous, with very homogeneous ridge-and-valley protrusions arranged over the entire surface of the membrane. The AWC particles are smaller and interact continuously on the surface of the PA protrusions.

[0123] The composite biomimetic membranes according to the invention thus offer the following advantages: [0124] they are easy to prepare by an inexpensive method of synthesis. They can easily be produced in large quantities without any expensive equipment, and do not require particular safety conditions. Moreover, the strategy of synthesis by supramolecular self-assembly and interfacial polymerization leads to the production of highly organized membrane films with inclusion, within the polyamide PA structure, of active regions from 30 to 40 nm enclosing artificial water channels; [0125] they have excellent permeability, 4 to 5 times greater (in LPRO-brackish water filtration mode) or 2.8 times greater (in SWRO-seawater filtration mode) than the TFC reference membrane while maintaining selectivity >99.5%—NaCl rejection. They have been tested in the same operating conditions of filtration that are usually employed in the methods used conventionally in desalination. This consequently makes them a material of choice for use as RO membrane in desalination modules; [0126] they also have remarkable properties of homogeneity due to a controlled nanostructuring during their synthesis. These membranes also display good stability in a basic hydrolytic environment and/or dissolution in the presence of organic solvents such as ethanol or methanol or surfactants such as dodecyl sulphate.

[0127] The membranes according to the invention have a biomimetic architecture in the form of the nanometric supramolecular aggregates of the AWC channels with a structure of soft material within a rigid polymer matrix of PA. This particular architecture of the membranes according to the invention is favourable to water permeability without allowing the transport of cations and anions. This offers an advantage in terms of water permeability, thus increasing their capacity for transport while maintaining high ionic retention.

[0128] Interesting selectivities obtained for rejection of organic contaminants may also be reported: [0129] a solution of 2 g/L of urea (pH=7) was filtered with an H-I4RO_3 membrane with a permeability of 4.29 LMH/bar and a rejection of urea of 70%; [0130] paracetamol solutions of 100 ppm, 200 ppm, 300 ppm were filtered with an H-I4RO_3 membrane with a permeability of 3.76 LMH/bar and rejections of paracetamol of 99.51, 99.14 and 98.98% respectively.