Carbon hollow fibre membrane

Abstract

The present invention relates to a process for the production of asymmetric cellulose hollow fibres and the use of such fibres in the production of asymmetric carbon hollow fibre membranes (CHFMs). In particular, the present invention provides a facile and scalable process for the preparation of asymmetric CHFMs by direct pyrolysis of polymeric precursors without the need for complex pre-pyrolysis treatment steps to prevent pore collapse. The present invention also relates to the use of asymmetric CHFMs prepared according to said process in the separation of hydrogen gas from a mixed gas source, especially in the separation of hydrogen from CO.sub.2 in the steam-methane reforming reaction.

Claims

1. A process for the production of an asymmetric cellulose hollow fibre comprising the steps of: a) providing a dope solution comprising cellulose, at least one ionic liquid, and optionally one or more co-solvent(s); b) coextruding into a gaseous atmosphere: said dope solution and a bore fluid comprising water, at least one ionic liquid, and optionally one or more co-solvent(s); c) quenching the coextruded dope solution and bore fluid in at least one coagulation bath containing water to form a water-wetted fibre, wherein the temperature of the coagulation bath is greater than 40 C.; d) contacting said water-wetted fibre with at least one organic solvent having a surface tension lower than that of water; and optionally e) drying the fibre.

2. The process according to claim 1 wherein the cellulose in the dope solution is microcrystalline cellulose (MCC).

3. The process according to claim 1 wherein the amount of cellulose in the dope solution is 1.0 to 25.0 wt. %.

4. The process according to claim 1 wherein the ionic liquid of the dope solution and/or the bore fluid comprises 1-ethyl-3-methylimidazolium.

5. The process according to claim 1 wherein the co-solvent is a polar solvent.

6. The process according to claim 1 wherein the temperature of the coagulation bath is in the range of 41 C. to 80 C.

7. The process according to claim 1 wherein the organic solvent having a surface tension lower than that of water is selected from the group consisting of C1-C6 alcohols, C5-C8 linear or branched aliphatic hydrocarbons, or mixtures thereof.

8. The process according to claim 1 wherein step d) comprises contacting the water-wetted fibres sequentially with at least two different organic solvents having a surface tension lower than that of water.

9. An asymmetric cellulose hollow fibre produced by the process according to claim 1; wherein the asymmetric cellulose hollow fibre consists essentially of cellulose II.

10. A process for the production of an asymmetric carbon hollow fibre membrane (CHFM) comprising the steps of: a) providing an asymmetric cellulose hollow fibre produced by the process according to claim 1; and b) pyrolysing said asymmetric cellulose hollow fibre.

11. The process according to claim 10 wherein the asymmetric cellulose hollow fibre is pyrolysed directly.

12. A process according to claim 10, wherein the pyrolysis step b) involves heating the asymmetric cellulose hollow fibre to a temperature of at least 500 C.

13. An asymmetric carbon hollow fibre membrane (CHFM) produced by a process according to claim 10.

14. An asymmetric carbon hollow fibre membrane (CHFM) having a dense outer layer and a concentric porous inner layer, wherein the asymmetric carbon hollow fibre membrane comprises at least 85 atomic % carbon, 5 to 15 atomic % oxygen and up to 1.0 atomic % nitrogen.

15. The asymmetric CHFM as claimed in claim 14 having a silicon content determined by X-ray photoelectron spectroscopy (XPS) of less than 1.0 atomic %.

16. The asymmetric CHFM as claimed in claim 14 having an H.sub.2 permeance of at least 140 GPU and an H.sub.2/CO.sub.2 selectivity of at least 10.0 at 130 C. and a pressure of 2 bar.

17. A module comprising a plurality of CHFMs as claimed in claim 14.

18. The process according to claim 1, wherein the ionic liquid comprises 1-ethyl-3-methylimidazolium acetate, 1-ethyl-3-methylimidazolium chloride, 1-ethyl-3-methylimidazolium dicyanamide, 1-ethyl-3-methylimidazolium diethyl phosphate, 1-ethyl-3-methylimidazolium dimethyl phosphate or 1-butyl-3,5-dimethylpyridinium bromide.

19. The process according to claim 1, wherein the co-solvent is an aprotic polar solvent.

20. The process according to claim 1 wherein the organic solvent is selected from the group consisting of isopropanol, n-hexane, and mixtures thereof.

21. The process according to claim 1, wherein the at least one organic solvent of step d) comprises a first solvent selected from the group consisting of C1-C6 alcohols and a second solvent selected from the group consisting of C5-C8 linear or branched aliphatic hydrocarbons, wherein step d) comprises a first contacting step with the first solvent followed by a second contacting step with the second solvent.

Description

BRIEF DESCRIPTION OF FIGURES

(1) FIG. 1. A) Schematic of the fabrication process for asymmetric cellulose hollow fibres by the dry-wet spinning process; B) Schematic of dried cellulose hollow fibre precursors; C) Carbonization protocols for fabrication of CHFMs; D) Schematic of asymmetric CHFMs.

(2) A key step is the fabrication of asymmetric cellulose hollow fibers by controlling the coagulation temperature at >40 C. (60 C. was used in the examples). Step 2 is the non-solvent exchange using lower surface tension solvents such as isopropanol, n-hexane to remove residual water inside hollow fibers before drying to prevent pore morphology collapse. Step 3 is tuning the ultramicropore and micropore structure of carbon membranes by changing the final carbonization temperature from 550-850 C.

(3) FIG. 2. A+B) Cross-sectional SEM images of a cellulose hollow fibre precursor dried after anti-collapse treatment. Scale bars: A-200 m, B-50 m.

(4) FIG. 3. A+B) Cross-sectional SEM images of a cellulose hollow fibre carbon membrane precursor with ambient air drying, directly from water-wetted membranes. Scale bars: A-200 m, B-30 m.

(5) FIG. 4. Comparative cross-sectional SEM images of flat sheet membranes cast at various coagulation bath temperatures. Scale bars: 100 m.

(6) FIG. 5. The mechanism of transformation from cellulose precursors to amorphous carbon membranes with bimodal pore structure.

(7) FIG. 6. Carbonization protocols for cellulose hollow fibre precursors.

(8) FIG. 7. A-B) Cross-sectional SEM images of CHFM-700; inset bent fibre. Scale bars: a-100 m, b-20 m; C) the XRD patterns of CHFMs carbonized at different temperatures; and D) the pore size distributions of CHFMs.

(9) FIG. 8. A) Single gas performances of CHFMs. B) Single-gas permeances of CHFM-850 as a function of the gas kinetic diameter. Inset: selectivity of the membrane for H.sub.2 over CO.sub.2, N.sub.2 and CH.sub.4.

(10) FIG. 9. Illustration of the high-pressure mixed gas permeation rig.

(11) FIG. 10. Representative module used for mixed gas permeation measurements.

(12) FIG. 11. 50 mol % H.sub.2/50 mol % CO.sub.2 mixed gas measurements of CHFM-700 at different operation pressures (5-18 bar) at 70 C.

EXAMPLES

(13) Materials

(14) Microcrystalline cellulose (MCC) powder (Avicel PH-101), isopropanol (99.7%, FCC grade), n-hexane (ReagentPlus, 99%) and dimethyl sulfoxide (DMSO, FCC grade) were purchased from Sigma-Aldrich. 1-Ethyl-3-methylimidazolium acetate (EmimOAc, >95%) was purchased from IOLITEC GmbH. All chemicals were used as received. Single gas (e.g. H.sub.2, CO.sub.2) and 50 mol %-50 mol % H.sub.2/CO.sub.2 mixed gas were bought from AGA, Norway. All fittings used for the construction of membrane modules were purchased from Swagelok.

(15) Characterization

(16) SEM images were obtained using a Hitachi SU-6600 field emission scanning electron microscope (FESEM). XRD analysis of CHFMs was carried out by Bruker D8 Focus instrument operated at 45 kV and 200 mA with 2 ranging from 5 to 70 at a scan speed of 0.05 s.sup.1 (Cu-K radiation, =0.154 nm). CO.sub.2 physisorption was measured by Quantachrome ASiQwin automated gas sorption analyser at 0 C. XPS spectra were obtained by ESCALAB 250 operated at 150 W and 200 eV with monochromatic Al-K radiation. Raman analysis was conducted using Renishaw inVia Raman Microscope with a 532 nm laser.

Example 1: Preparation of Asymmetric Cellulose Hollow Fibres

(17) Asymmetric cellulose hollow fibres, which are precursors for the final carbonised CHFMs, were prepared by a dry-wet spinning process as illustrated in FIG. 1A. A 12 wt. % MCC/(EmimOAc+DMSO) dope solution was used in the spinning process. MCC (60 g) (Cellulose I) was gradually added into 440 g EmimOAc/DMSO (weight ratio 3:1) co-solvent with mechanical stirring in a N.sub.2 atmosphere glovebox, and kept at 50 C. overnight, to allow the cellulose to be dissolved completely. Asymmetric cellulose (Cellulose II) hollow fibres were then fabricated by a dry-wet spinning process under the conditions given in Table 1.

(18) TABLE-US-00001 TABLE 1 Cellulose hollow fibre spinning conditions Spinning conditions Value Dope solution composition 12% MCC in (75 wt. % EmimOAc + and temperature 25 wt. % DMSO), 25 C. Bore fluid composition and 20% Water in (75 wt. % EmimOAc + temperature 25 wt. % DMSO), 25 C. First coagulation bath 60 C. temperature Second coagulation bath 40 C. temperature Dope flow rate 4.4 mL min.sup.1 Bore flow rate 1.8 mL min.sup.1 Take up speed 14.6 m min.sup.1 Air gap 8 cm Spinneret OD/ID 0.7/0.5 mm

(19) The resulting spun hollow fibres were cut in ca 1.2 m long sections and placed in a deionized water bath over 48 h to fully exchange the solvent (EmimOAc+DMSO) with water. The water-wetted cellulose hollow fibres were immersed into pure isopropanol for 2 h, followed by soaking in n-hexane for 2 h, and then all the hollow fibres were allowed to dry under ambient conditions in air.

(20) Cross-sectional SEM images of the resulting dried cellulose hollow fibres are shown in FIGS. 2A and 2B. The hollow fibres present a clear asymmetric structure, with a relatively dense outer layer, a middle layer rich in macrovoids, and a more porous inner support layer. This is in stark contrast to the structure of the comparative cellulose hollow fibre shown in FIGS. 3A and 3B, which was prepared according to the same method but with ambient air drying direct from the water-wetted fibres i.e. without the solvent exchange treatment. The cellulose hollow fibre of FIGS. 3A and 3B has a dense and symmetric structure, most likely as a result of pore collapse. The solvent exchange treatment is therefore critical in order to obtain cellulose hollow fibres that are asymmetric.

Example 2: Investigating the Effect of Coagulation Bath Temperature on Cellulose Hollow Fibre Structure

(21) In order to determine the optimal conditions for the formation of cellulose hollow fibres, the effect of the coagulation bath temperature (Ta) was investigated. To this end, different flat sheet membranes were cast under various T, conditions in the range of 25 C. to 60 C., whilst the dope temperature (Td) was maintained at 25 C. No bore fluid was used in the preparation of the flat sheet membranes. The water-wetted cellulose membranes were then immersed into pure isopropanol for 2 h, followed by soaking in n-hexane for 2 h, and then all the membranes were allowed to dry under ambient conditions in air.

(22) Cross-sectional SEM images of the resulting flat sheet membranes are presented in FIG. 4 (A) 25 C., B) 35 C., C) 40 C., D) 45 C., E) 50 C. and F) 60 C.). The images show that a low coagulation bath temperature (T, =40 C. or less) gives rise to a cellulose membrane having a dense and symmetric structure (FIGS. 4A-C). In contrast, when T, 45 C., clear asymmetric structures having a dense top layer and a more porous support layer are generated (FIGS. 4D-F). In further tests, simply elevating the dope temperature whilst maintaining a coagulation bath temperature lower than 40 C. did not give rise to the desired asymmetric structure.

(23) The temperature of the coagulation bath is therefore critical in obtaining cellulose hollow fibres having an asymmetric structure. Moreover, variation of the coagulation bath temperature is shown to allow for control over the relative thickness of the dense and porous layers.

Example 3: Preparation of Carbon Hollow Fibre Membranes

(24) The dried cellulose hollow fibres prepared in Example 1 were carbonized in a tubular furnace (Horizontal Split Tube Furnace, Carbolite Gero Limited) by applying the specific carbonization protocols depicted in FIG. 6, under high purity argon (Ar, 99.999%) purge gas under a continuous flow of 80 mL min.sup.1. A dwell-time of 2 h at 300 C. was employed to take into account the significant weight loss at this temperature due to cellulose depolymerisation. Three types of carbon membranes were obtained at different final temperatures of 550, 700 and 850 C. (denoted as CHFM-550, CHFM-700, and CHFM-850 respectively) while all other carbonization parameters (e.g., heating rate, dwell time, etc.) were the same. The tubular furnace was evacuated down to 3 mbar overnight before being purged with Ar. The system was cooled down naturally after the carbonization process was completed, and the resulting carbon hollow fibre membranes (CHFMs) were removed when the temperature had cooled to below 50 C.

(25) Cross-sectional SEM images of CHFM-700 are presented in FIGS. 7A and 7B. The asymmetric structure of the hollow fibres was well maintained with an outer selective layer of ca. 3 m and an integral porous inner support layer. The prepared CHFMs also exhibit good mechanical flexibility with a bend radius of <1.5 cm, as indicated in the inset of FIG. 7A.

(26) The XRD patterns for these CHFMs are shown in FIG. 7C. The patterns reveal the characteristic peak for 20 at around 24, which corresponds to the (002) plane of the graphite phase (sp.sup.2 carbon). The d-space was calculated from the Bragg equation. The peak shift to a higher 20 indicates that the average inter-plane distance (d.sub.002) decreases from 3.78 to 3.50 when the carbonization temperature increases from 550 to 850 C. This indicates that the carbon membranes prepared at higher carbonization temperatures tend to form graphitic carbon (3.4 ) with a more ordered graphitic structure and smaller pores.

(27) The pore size distribution shown in FIG. 7D, calculated by the NLDFT model from CO.sub.2 physisorption at 0 C. in the range of 3-10 , confirms the narrowing of the pore width of CHFM-850 compared to CHFM-550. The present CHFMs exhibit a strong peak for the ultramicropores in the range of 3-4 , which is in the size range needed to allow molecular sieving between H.sub.2 (2.9 ) and other larger gas molecules (e.g., CO.sub.2, N.sub.2, and CH.sub.4). With the increase of the carbonization temperature, the micropore peaks (>5 ) are weakened, while that of the ultramicropores (<5 ) increases, which indicates that the average pore size decreases for the CHFMs carbonized at higher temperatures.

(28) The CHFMs were characterized by XPS, and the elemental compositions of different carbon membranes are given in Table 2. The carbon content increases with the increase of carbonization temperature.

(29) TABLE-US-00002 TABLE 2 Elemental composition of the CHFMs from XPS analysis. C (at. %) O (at. %) N (at. %) CHFM-550 90.08 9.26 0.67 CHFM-700 91.25 8.10 0.65 CHFM-850 92.41 7.04 0.55

(30) In order to establish the suitability for gas separation, single and mixed gas permeation experiments were performed on the CHFMs prepared in Example 3. Single gas permeation measurements were conducted by applying a constant permeate volume method using a feed pressure of 2 bar. The gas permeance and selectivity are calculated using eq. (1):

(31) $\begin{array}{cc}\frac{P}{l}=\frac{273.15.Math.{10}^{3}V}{76T.Math.A}.Math.\frac{{}_{p_1}^{p_2}\frac{dp}{P_F-p}}{t}& \left(1\right)\end{array}$
where P/I (GPU, 1 GPU=110.sup.6 cm.sup.3(STP).Math.cm.sup.2.Math.s.sup.1.Math.cm Hg.sup.1=3.3510.sup.10 mol.Math.s.sup.1.Math.m.sup.2.Math.Pa.sup.1) is the single gas permeance. V (cm.sup.3) is the downstream (permeate) volume (predetermined using He calibration), and T(K) is the experimental temperature. A (cm.sup.2) is the hollow fibre membrane outer active surface area (shell-side feed). P.sub.F and p (bar) are the pressures in the feed side and permeate side, respectively. t (s) is the steady state testing time. The H.sub.2/CO.sub.2 ideal selectivity is calculated by the ratio of H.sub.2 permeance to CO.sub.2 permeance.

(32) FIG. 8A shows the single gas performances of the CHFMs prepared in Example 3 at 25 C., 60 C., 100 C. and 130 C. with 2 bar feed pressure. Hollow symbols represent predicted performance at 200 C. The membranes prepared at higher carbonization temperatures provide higher H.sub.2/CO.sub.2 selectivity, but with the sacrifice of some H.sub.2 permeance. For instance, CHFM-850 has a H.sub.2/CO.sub.2 selectivity of 46.2 at 25 C., which is 4 times higher than that of CHFM-550, while H.sub.2 permeance is decreased from 102.1 GPU to 16.2 GPU concomitantly. The solid and dashed lines drawn in a) are based on the 2008 Robeson upper bound line by converting permeability to permeance, assuming a membrane selective layer thickness of 1 and 3 m, respectively.

(33) FIG. 8 B shows the single-gas permeances of CHFM-850 as a function of the gas kinetic diameter at 130 C. and 2 bar. There is a clear cut-off of gas permeance between the smaller molecules (148.2 GPU for H.sub.2 and 139.6 GPU for He) and that of the larger molecules, which indicates that gas permeation is dominated mainly by kinetic diameter of the gas molecules, i.e. using a molecular sieving transport mechanism. The inset figure shows the selectivity of H.sub.2 over CO.sub.2, N.sub.2 and CH.sub.4.

(34) Gas permeance and selectivity vary significantly with temperature. Significant increases of gas permeance and selectivity are observed by increasing the temperature from 25 to 130 C. (FIG. 8A), particularly for the membranes prepared at higher carbonization temperatures. At 130 C., the H.sub.2/CO.sub.2 selectivity and H.sub.2 permeance of the CHFM-850 increased to 83.9 and 148.2 GPU, respectively, which are approximately 2 times and 9 times higher than the results obtained at a temperature of 25 C. Higher temperatures accelerate gas diffusion, which enhances gas permeation. Conversely, the lower CO.sub.2 adsorption at higher temperatures improves the H.sub.2/CO.sub.2 selectivity. Therefore, considering practical industrial applications, for example, H.sub.2 purification from natural gas-derived syngas (which is usually operated at 150 C. or above), a higher operating temperature is preferable to enhance the H.sub.2/CO.sub.2 separation performance.

(35) To test the potential of CHFMs for H.sub.2 purification in a steam methane reforming process (usually performed at pressures of up to 15-20 bar), a lab-scale hollow fiber module containing CHFM-700 was tested using a 50/50 mol. % H.sub.2/CO.sub.2 mixed gas at 70 C. at different feed pressures (5-18 bar) using a high-pressure gas permeation rig (FIGS. 9 and 10). All the tube lines and the membrane module were pre-heated to a set temperature during gas permeation testing. The feed flow is controlled at 150 NmL min.sup.1 during the testing. Argon was used as sweep gas. The permeate gas flow and composition were measured by a bubble flow meter and a gas chromatograph (GC, 8610C, SRI Instruments Inc.), respectively. Three CHFM-700 membrane modules (8 carbon hollow fibres per module) were tested to determine experimental error. Gas was fed to the shell side, and the permeate gas exited from the bore side, with argon as sweep gas operated in a counter-current flow pattern. The selectivity is calculated by

(36) $=\frac{y_{H_2}/y_{{\mathrm{CO}}_2}}{x_{H_2}/x_{{\mathrm{CO}}_2}},$
where y.sub.i and x.sub.i are the concentration of the components in the permeate and feed, respectively.

(37) The results of the mixed gas test are shown in FIG. 11. There is a gradual decrease in H.sub.2 permeance (ca. 15.8%) with an increase in the total feed pressure from 5 to 18 bar, but the H.sub.2/CO.sub.2 selectivity increases from 31.8 to 37.7 (18.6% increase). The CHFMs according to the present invention are thus particularly suited for high pressure gas separation conditions, such as in the steam-methane reforming reaction.