Membranes containing polymerized ionic liquid for use in gas separation

Abstract

The invention relates to dense synthetic membranes made from polymerised phosphonium-based ionic liquids which were found to be particularly suitable for use in gas separation. The membranes are obtainable by copolymerization via UV-curing of a composition comprising a phosphonium-based ionic liquid monomer, a co-monomer, a cross-linker, a surfactant and a photo-initiator, the remainder of the polymerization mixture consisting of water. The invention also relates to a process of manufacturing said membranes, resulting in solid, dense and mechanically stable membranes, and to the use of the membranes so produced in the separation of gas mixtures, particularly gas mixtures containing carbon dioxide.

Claims

1. A dense poly(ionic liquid)-based membrane suitable for gas separation obtainable by copolymerization via UV-curing of a mixture comprising: a) 10 to 50 wt %, of a IL monomer trialkyl[(4-vinylphenyl)alkyl]phosphonium with X.sup.− as counterion ([TAVPAP]X) of the following formula: ##STR00005## wherein each R represents an alkyl chain having 3 to 20 carbon atoms, L represents a linker consisting of a simple alkanediyl C.sub.1-C.sub.20 chain, or an ethereal chain containing 1-20 carbon atoms and 1-10 oxygen atoms, and X.sup.− represents an anion selected from the group consisting of: BF.sub.4.sup.−, PF.sub.6.sup.−, SbF.sub.6.sup.−, TsO.sup.−, CF.sub.3—SO.sub.3.sup.−, NC—N.sup.−—CN, (MeO).sub.2PO.sub.2.sup.−, EtSO.sub.4.sup.− or F.sub.3CO.sub.2SN.sup.−—SO.sub.2CF.sub.3 b) 30 to 70% wt % of one or more ethylenically unsaturated co-monomers; c) 1 to 35 wt % of a surfactant; d) 0.5-20% wt %, of one or more cross-linkers, e) 0.001 to 5 wt % of a photo-initiator; f) the remainder of the mixture being water.

2. The dense poly(ionic liquid)-based membrane according to claim 1, wherein each R represents an alkyl chain having 3 to 10 carbon atoms.

3. The dense poly(ionic liquid)-based membrane according to claim 2, wherein each R represents butyl, hexyl or octyl.

4. The dense poly(ionic liquid)-based membrane according to claim 1, wherein L represents a methanediyl group.

5. The dense poly(ionic liquid)-based membrane according to claim 1, wherein the ethylenically unsaturated co-monomer is 2-hydroxyethyl acrylate (HEMA), and is present in the composition in the amount of 50-60 wt %.

6. The dense poly(ionic liquid)-based membrane according to claim 5, wherein the IL monomer is present in the composition at a concentration of 20-30 wt %, the surfactant is present at a concentration of 10-30 wt %, the cross-linker is present at a concentration of 1-10 wt % and the photo-initiator is present at a concentration of 0.1 to 1 wt %.

7. A process for producing dense poly(ionic liquid) membranes suitable for gas separation, which process includes the following steps: providing a mixture as defined under items a) to f) of claim 1; applying said mixture to a support by means of a film casting knife or casting said mixture between two plates so that the thickness of the membrane ranges between 0.02 and 200 μm; curing said curable monomer mixture by UV radiation to induce polymerization.

8. A method of gas separation comprising using a membrane according to claim 1.

9. Use according to claim 8 wherein said gases to be separated include carbon dioxide (CO.sub.2), hydrogen (H.sub.2), methane (CH.sub.4), nitrogen (N.sub.2) and oxygen (O.sub.2).

10. A process for separating two components, A and B, of a gas mixture, which process comprises: i. passing said gas mixture across a separation membrane having a feed side and a permeate side, said separation membrane having a selective layer produced as defined in claim 7; ii. providing a driving force for transmembrane permeation; iii. withdrawing from the permeate side a permeate stream enriched in component A compared to the gas mixture; and iv. withdrawing from the feed side a residue stream depleted in component A compared to the gas mixture.

11. A method of gas separation comprising using a membrane produced according to the process of claim 7.

12. The dense poly(ionic liquid)-based membrane according to claim 1, wherein the IL monomer trialkyl[(4-vinylphenyl)alkyl]phosphonium is present in the composition in the amount of 20-30 wt %.

13. The dense poly(ionic liquid)-based membrane according to claim 1, wherein the one or more ethylenically unsaturated co-monomers is present in the composition in the amount of 50-60 wt %.

14. The dense poly(ionic liquid)-based membrane according to claim 1, wherein the one or more ethylenically unsaturated co-monomers comprises acrylates.

15. The dense poly(ionic liquid)-based membrane according to claim 1, wherein the surfactant is present in the composition in the amount of 10-30 wt %.

16. The dense poly(ionic liquid)-based membrane according to claim 1, wherein the one or more cross-linkers is present in the composition in the amount of 2-10 wt %.

17. The dense poly(ionic liquid)-based membrane according to claim 1, wherein the photo-initiator is present in the composition in the amount of 0.1-1 wt %.

Description

BRIEF DESCRIPTION OF THE FIGURES

(1) The specific features of the invention, as well as the advantages thereof, will become more apparent with reference to the exemplary and not limiting experimental work description reported in the following, and to the relative figures, in which:

(2) FIG. 1 schematically shows an experimental setup for the mixed gas permeation measurements to test the performance of the poly(RTIL) membranes of the present invention;

(3) FIG. 2 shows the CO.sub.2/N.sub.2 selectivity vs. permeability plot for membranes with RTIL monomers having pendant C.sub.4 and C.sub.8 alkyl chains according to the invention, in comparison with membranes having no RTIL monomer in the composition (referred to as CEx1) and membranes having pendant C.sub.2 alkyl chains (CEx2);

(4) FIG. 3 shows the CO.sub.2/CH.sub.4 selectivity vs. permeability plot for membranes of the same types of FIG. 2;

(5) FIG. 4 shows the CO.sub.2/H.sub.2 selectivity vs. permeability plot for membranes of the same types of FIG. 2;

(6) FIG. 5 is a graph showing the permeability (left) and the selectivity (right) to various gases of membranes with RTIL monomers having pendant C.sub.4 and C.sub.8 alkyls according to the invention, or having pendant C.sub.2 alkyls according to the prior art, as a function of chain length; and

(7) FIG. 6 is a graph showing the selectivity vs. permeability plot for the various couples of gases tested, for the same membranes of FIG. 5.

EXAMPLES

(8) The following materials were used in the Examples:

(9) Ionic liquid monomers were prepared according to literature (Barsanti, A. C. et al., RSC Adv. 2014, 4, 38848-38854).

(10) The co-monomers (such as HEMA), the cross-linkers (such as EGDMA), the surfactants (such as docecyltrimethilammonium bromide, DTAB), the photoinitiators (such as dimethoxyphenyl acetophenone, DMPA) were purchased from Sigma-Aldrich and used as received.

(11) In the following examples all parts and percentages are by weight unless otherwise specified.

Examples and Comparative Examples

(12) The blank solution (referred to in Table 1 as CEx1) was prepared by mixing 70 wt % of a co-monomer (2-hydroxyethyl methacrylate, HEMA), 17.5 wt % of a surfactant (dodecyltrimethylammonium bromide, DTAB) and 12.5 wt % water. Then, the crosslinker (ethylene glycol dimethacrylate, EGDMA; 3 wt % with respect to the amount of HEMA) and a photoiniator (dimethoxyphenyl acetophenone, DMPA, 0.6 wt % with respect to the total weight of the mixture thus obtained) were added.

(13) Each RTIL momomer (27 wt %) was mixed with 52 wt % of a co-monomer (2-hydroxyethyl methacrylate, HEMA), 13 wt % of a surfactant (dodecyltrimethylammonium bromide, DTAB) and 8 wt % water. Then, the cross-linker (ethylene glycol dimethacrylate, EGDMA; 3 wt % with respect to the amount of RTIL+HEMA) and a photoiniator (dimethoxyphenyl acetophenone, DMPA, 0.6 wt % with respect to the total weight of the mixture thus obtained) were added (see Table 1 for the compositions of CEx2 and invention examples Ex1 to Ex6).

(14) All solutions were separately cast between two Rain-X coated quartz plates and photo-polymerised for 3 minutes under a 365 nm UV-lamp with an intensity of 8.5 mW/cm.sup.2 at the sample surface (XX-15A, Spectroline, Westbury, N.Y.).

(15) The polymerized membranes were removed in water from the glass plates and stored in water up to the gas tests.

(16) TABLE-US-00001 TABLE 1 Composition of the tested membranes Type of IL HEMA DTAB EGDMA DMPA Concentration (co-monomer) H.sub.2O (surfactant) (cross-linker) (photoinitiator) Thickness Membrane (wt %) (wt %) (wt %) (wt %) (wt %) (wt %) (μm) CEx1 — 70 12.5 17.5 3 0.6 150 CEx2 triethyl(4-vinylbenzyl) 52 8 13 3 0.6 200 phosphonium tetrafluoroborate 27 wt % Ex1 tributyl(4-vinylbenzyl) 52 8 13 3 0.6 98 phosphonium tetrafluoroborate 27 wt % Ex2 tributyl(4-vinylbenzyl) 52 8 13 3 0.6 51 phosphonium tetrafluoroborate 27 wt % Ex3 tributyl(4- 52 8 13 3 0.6 69 vinylbenzyl)phosphonium tetrafluoroborate 27 wt %) Ex4 trioctyl(4-vinylbenzyl) 52 8 13 3 0.6 200 phosphonium tetrafluoroborate (27 wt %) Ex5 trioctyl(4-vinylbenzyl) 52 8 13 3 0.6 110 phosphonium tetrafluoroborate (27 wt %) Ex6 trioctyl(4-vinylbenzyl) 52 8 13 3 0.6 47 phosphonium tetrafluoroborate (27 wt %)

(17) In the frame of the research that led to the present invention some tests were performed adding methylmethacrylate (MMA) (at 10 and 13.5 wt %) as a second co-monomer in the system. The solution was co-polymerised within the membrane together with PILs. However, it was found that the presence of MMA caused a decrease in mechanical resistance of the resulting membrane.

(18) Further, some membranes were prepared avoiding the use of the cross-linker EGDMA. However, when EGDMA was not present in the system, the polymerization was not complete and the resulting membrane was very fragile.

(19) The amount of water was also varied (8, 20, 25, and 30 wt %). However when the concentration of water was increased to more than 8 wt % the membrane resulted porous, and its mechanical properties were drastically decreased.

(20) The following Table 2 shows, by way of comparative examples, the compositions of some membranes outside the scope of the present invention, together with a short comment on their performances.

(21) TABLE-US-00002 TABLE 2 Composition of membranes outside the scope of this invention Mem- HEMA MMA (2.sup.nd DTAB EGDMA DMPA brane (co-monomer) co-monomer) H.sub.2O (surfactant (cross-linker) (initiator) code IL (wt %) (wt %) (wt %) (wt %) (wt %) (wt %) (wt %) Comments IL 21 Trioctyl(4-vinylbenzyl)phos- 52 wt %   10 wt % 8 13 wt % 3 wt % on 0.6 wt % The membrane phonium bis(trifluorometh- the weight of is fragile anesulfonamide) IL + HEMA 17 wt % IL 17 Trioctyl(4-vinylbenzyl)phos- 52 wt % 13.5 wt % 8 13 wt % 3 wt % on 0.6 wt % The membrane phonium bis(trifluorometh- the weight of is fragile anesulfonamide) IL + HEMA 13.5 wt % IL 37 1-Hexyl-3-(4-vinylbenzyl)- 18 wt % — 36.6 27.7 wt % — 0.6 wt % The membrane 1H-3-imidazolium is fragile bis(trifluorometh- anesulfonamide) 18.8 wt % IL 43 1-Hexyl-3-(4-vinylbenzyl)- 29 wt % — 20 29 wt % 5 wt % on 0.6 wt % The membrane 1H-3-imidazolium the weight of is fragile bis(trifluorometh- IL + HEMA anesulfonamide) 22 wt % IL 42 1-Hexyl-3-(4-vinylbenzyl)- 26 wt % — 25 28 wt %) 5 wt % 0.6 wt % The membrane 1H-3-imidazolium on the is permeable bis(trifluorometh- weight of to water. It anesulfonamide) IL + HEMA is porous 21 wt % IL 41 1-Hexyl-3-(4-vinylbenzyl)- 23 wt % — 30 27 wt % 5 wt % 0.6 wt % The membrane 1H-3-imidazolium on the is fragile bis(trifluorometh- weight of anesulfonamide) IL + HEMA 20 wt %
Evaluation of Gas Permeability and Selectivity

(22) The transport properties of the membranes were investigated by feeding single gases, in saturated conditions (RH=99%) at a transmembrane pressure difference of 10 bar, and measuring the membrane properties such as permeance (flux) and ideal selectivity. Table 3 below reports the operating conditions adopted during the experiments.

(23) TABLE-US-00003 TABLE 3 Operating parameters Temperature 25° C. Feed pressure 10 bar Permeate pressure 1 bar Relative humidity, % 99 No sweep gas Feed composition single gases: CO.sub.2, N.sub.2, CH.sub.4, H.sub.2

(24) The experimental apparatus used for carrying out the gas permeation experiments is schematically shown in FIG. 1.

(25) In the experiments reported herein, symmetric flat sheet membranes were mounted in a stainless steel membrane module opportunely dimensioned. The membrane module can host membranes from 1 mm.sup.2 to 100 cm.sup.2. In this specific case, the effective membrane area available for permeation was ranging between 3.8 and 19.2 cm.sup.2.

(26) Once the membranes were sealed in, the permeation module constituted by four ends was placed in a furnace with a PID controller for controlling the temperature during the experiments. The four ends of the module were: feed, retentate, permeate, sweep. No sweep gas was applied in the present measures; therefore, this exit was kept closed during all the experiments.

(27) The method used for permeation measurements was the concentration gradient method, consisting in forcing a part of the feed stream to permeate the membrane under a pressure gradient and measuring both the permeate and retentate flow rates. Mass transport properties were measured by single gas experiments. Each gas (contained in single cylinders with a purity 5.0) was fed to the membrane module with a mass flow controller (Brooks Instrument.sup.AM, 5860S) positioned at the feed line to manipulate the feed flow rates. A back pressure regulator (Swagelok, KBP) and a pressure gauge on the retentate line were used to keep the required trans-membrane pressure difference in the module. The retentate and permeate flow rates were measured by two bubble flow meters.

(28) To perform the experiments in wet conditions, a stainless steel humidifier was placed before the module inlet and was set at the same temperature and pressure of the membrane module. The dry feed gas was forced to enter the humidifier were ultrapure water was contained. Since this humidifier was set at the same temperature and pressure of the membrane module, after bubbling in the water, the gaseous stream exiting the humidifier and fed to the module was saturated, as confirmed by the humidity sensor placed before the module feed line.

(29) In principle, this apparatus allows to modulate the relative humidity by mixing two streams of the same gas or mixture: the first stream saturated with water (relative humidity=100%) and the other stream completely dry. The value of relative humidity can be tuned by changing the flow rates of the two streams.

(30) All the experimental measurements carried out on the membranes of the present invention were performed at 99% of relative humidity in the feed stream to assure the proper level of membrane hydration. Three humidity sensors measured the relative humidity of the feed, retentate and permeate.

(31) The separation performance of the membrane was evaluated by the permeance and selectivity in the gas mixture. The permeance (Eq. 1) is the permeate flow rate normalized by the membrane area and the partial pressure differences through the membrane.

(32) $\begin{array}{cc}\mathrm{Permean}c{e}_i=\frac{\frac{\mathrm{Permeate}\mathrm{flow}\mathrm{rate}}{\mathrm{Membrane}\mathrm{area}}}{\left[\mathrm{Pressur}{e}^{Feed}\left(\frac{x_i^{Feed}+x_i^{\mathrm{Retentate}}}{2}\right)-{\mathrm{Pressure}}^{\mathrm{Permeate}}{x}_i^{\mathrm{Permeate}}\right]},\mathrm{GPU}& \left(1\right)\end{array}$
where x is the molar fraction of gas i. (Units are in barrer, where 1 barrer=10.sup.−10 cm.sup.3 (STP) cm/cm.sup.2 s cm Hg)

(33) As in the case of the membranes of the present invention, the flat module and the high flow rates used allow the complete mixing in the feed side; therefore, no profiles exited between feed/retentate ends.

(34) Permeability was used (Eq. 2) to compare membranes performance having different thicknesses.
Permeability.sub.i=Permeance*membrane thickness  (2)

(35) The selectivity (Eq. 3) is the ratio of the membrane permeance of two gases.

(36) $\begin{array}{cc}Se{\mathrm{lectivity}}_{\mathrm{ij}}=\frac{Permeanc{e}_i}{Permeanc{e}_j}& \left(3\right)\end{array}$

(37) Tables 4, 5 and 6 below shows the results of the permeability and selectivity tests on membranes having no RTIL monomer in the composition (CEx1), membranes having pendant C.sub.2 alkyl chains (CEx2) and membranes according to the invention (butyl.sub.3P.sup.+vinylbenzene and octyl.sub.3P.sup.+vinylbenzene), respectively for the couples of gases CO.sub.2/N.sub.2, CO.sub.2/CH.sub.4 and CO.sub.2/H.sub.2. The same results are presented in graphic form in FIGS. 2, 3 and 4.

(38) TABLE-US-00004 TABLE 4 CO.sub.2/N.sub.2 selectivity vs. CO.sub.2 permeability CO.sub.2 permeability, Standard CO.sub.2/N.sub.2 Standard barrer deviation selectivity deviation CEx1 34.5 n.a. 2.5 n.a. CEx2 101.6 n.a. 4.1 n.a. butyl.sub.3P.sup.+vinylbenzene 37.9 4.15 6.0 0.00 octyl.sub.3P.sup.+vinylbenzene 33.6 7.88 29.9 2.54

(39) TABLE-US-00005 TABLE 5 CO.sub.2/N.sub.2 selectivity vs. CO.sub.2 permeability CO.sub.2 permeability, Standard CO.sub.2/CH.sub.4 Standard barrer deviation selectivity deviation CEx1 34.5 n.a. 4.1 n.a. CEx2 101.6 n.a. 3.4 n.a. butyl.sub.3P.sup.+vinylbenzene 37.9 4.15 7.8 0.00 octyl.sub.3P.sup.+vinylbenzene 33.6 7.88 31.4 3.15

(40) TABLE-US-00006 TABLE 6 CO.sub.2/N.sub.2 selectivity vs. CO.sub.2 permeability CO.sub.2 permeability, Standard CO.sub.2/H.sub.2 Standard barrer deviation selectivity deviation CEx1 34.5 n.a. 1.8 n.a. CEx2 101.6 n.a. 1.04 n.a. butyl.sub.3P.sup.+vinylbenzene 37.9 4.15 11.2 0.00 octyl.sub.3P.sup.+vinylbenzene 33.6 7.88 17.3 2.54

(41) FIGS. 2, 3 and 4 clearly show the improvement found (particularly in terms of selectivity) for the membranes of the compositions of Examples 1 to 6 according to to the invention against CEx1 (blank without the IL monomer) and CEx2 having an ethyl-chain connected to the phosphonium group.

(42) All the results have been normalized in terms of permeability, meaning to that the thickness has been already taken into account in the graphics. The thickness was measured just before experimental measurements on the membrane completely wet, since it is not possible to measure the thickness on the dry sample.

(43) Table 7 below shows the permeability of the membranes in this study to the various gases, and the results of selectivity tests for the various couples of gases.

(44) TABLE-US-00007 TABLE 7 Permeability and selectivity as a function of alkyl chain length CO.sub.2 N.sub.2 CH.sub.4 H.sub.2 permeability, permeability, permeability, permeability, CO.sub.2/N.sub.2 CO.sub.2/CH.sub.4 CO.sub.2/H.sub.2 barrer barrer barrer barrer selectivity selectivity selectivity CEx1 34.5 14.0 8.4 18.7 2.5 4.1 1.8 CEx2 101.6 24.7 29.5 97.3 4.1 3.4 1.04 butyl.sub.3P.sup.+ 37.9 ± 4.15 6.5 5.0 3.5 ± 0.95  6.0 ± 0.00  7.8 ± 0.00 11.2 ± 0.00 vinylbenzene octyl.sub.3P.sup.+ 33.6 ± 7.88 1.1 1.1 1.9 ± 0.20 29.9 ± 2.54 31.4 ± 3.15 17.3 ± 3.95 vinylbenzene

(45) The same results of the table above are presented in graphic form in FIG. 5, where the permeability of the various gases (left) and the selectivity of the membranes for various couples of gases (right) are plotted as a function of the pendant alkyl chains length. As it can be seen in FIG. 5, the permeability of all the investigated gases decreases as longer is the ionic liquid chain length.

(46) The triethyl(4-vinylbenzyl)phosphonium tetrafluoroborate has thus the highest permeability which is more than three times greater than the one of trioctyl(4-vinylbenzyl)phosphonium tetrafluoroborate.

(47) The functionality of selectivity with ionic liquid chain length follows a positive trend, therefore the trioctyl(4-vinylbenzyl)phosphonium tetrafluoroborate shows the highest selectivity with respect to the other membranes prepared with a shorter ionic liquid chain length. The membranes result to be selective toward CO.sub.2, and this can be mainly ascribed to the presence of ionic liquids which promote the solubility of CO.sub.2, preferentially favouring its permeability, with respect to the other gases for which the transport is mainly influenced by diffusivity.

(48) From the data of Table 7 above it is also possible to draw the graph of FIG. 6, where the CO.sub.2/N.sub.2, CO.sub.2/CH.sub.4 and CO.sub.2/H.sub.2 selectivities are reported vs. the CO.sub.2 permeability for the membranes examined. Basically, as shown in this figure, a sort of trade-off between permeability and selectivity was observed, as function of the chain length: the higher the permeability the lower the selectivity and vice versa.

(49) The present invention has been described with reference to a few specific embodiments, but it is to be understood that variations and modifications may be made by those skilled in the art without departing from the scope of the invention as is clear from the appended claims.