Fabrication of dual-layer polyethersulfone (PES) hollow fiber membranes with an ultrathin dense-selective layer for gas separation

Journal of Membrane Science 245 (2004) 53–60

Fabrication of dual-layer polyethersulfone (PES) hollow fiber membranes with an ultrathin dense-selective layer for gas separation Yi Lia , Chun Caoa , Tai-Shung Chunga,∗ , Kumari P. Pramodab a

Department of Chemical and Biomolecular Engineering, National University of Singapore, 10 Kent Ridge Crescent, Singapore 119260, Singapore b Institute of Material Research and Engineering, 3 Research Link, Singapore 117602, Singapore Received 14 May 2004; received in revised form 16 July 2004; accepted 10 August 2004 Available online 15 September 2004

Abstract By using co-extrusion and dry-jet wet-spinning phase-inversion techniques with the aid of heat-treatment at 75 ◦ C, we have fabricated ˚ To our best knowledge, this is the dual-layer polyethersulfone (PES) hollow fiber membranes with an ultrathin dense-selective layer of 407 A. thinnest thickness that has ever been reported for dual-layer hollow fiber membranes. The dual-layer hollow fibers have an O2 permeance of 10.8 GPU and O2 /N2 selectivity of 6.0 at 25 ◦ C. The effects of post heat-treatment on the membrane morphology and gas separation performance were also studied. It was found that heat-treatment at 75 ◦ C improves the gas permeance and ideal selectivity, while heat-treatment at 150 ◦ C results in a significant reduction in both permeance and selectivity due to the enhanced substructure resistance. SEM pictures confirmed that a higher heat-treatment temperature can significantly reduce pore sizes and the amount of pores in the substructure immediately underneath the dense-selective layer. © 2004 Elsevier B.V. All rights reserved. Keywords: Dual-layer hollow fiber membrane; Polyethersulfone (PES); Ultrathin dense-selective layer; Gas separation; Heat-treatment; Macrovoids

1. Introduction In the last 20 years, the asymmetric single-layer hollow fiber membrane is always a favorable configuration in the membrane-based systems for gas separation [1–6]. The silicone rubber technology is proven an effective means to seal surface defects for gas separation membranes. To further expand membranes for industrial applications, one must find ways to enhance the gas permeation flux (productivity) and permselectivity by combining the synthesis of high-performance polymeric membrane materials with the innovation of membrane fabrication technology. Making the dense-selective layer of asymmetric hollow fiber membranes thinner is one of the most effective approaches to increase the gas permeation flux. With the development of technology, how to fabricate single-layer hollow fiber membranes with an ultrathin dense-selective layer has ∗

Corresponding author. Fax: +65 6779 1936. E-mail address: [email protected] (T.-S. Chung).

0376-7388/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.memsci.2004.08.002

been no longer a challenge [7–15]. Excellent and informative reports to prepare single-layer hollow fiber membranes with an ultrathin dense-selective layer have been published by researchers at Permea [2,7–9], at Prof. Koros’ group at the University of Texas [10–12], and at National University of Singapore [13–17]. They concluded that the keys to fabricate hollow fiber membranes with an ultrathin dense-selective layer are (1) to control the chemistry of the internal coagulant and the bore-fluid flow rate, and (2) to have a dope exhibiting significant chain entanglement. A significant advance on polymeric materials for gas separation has also been made in the last 20 years [18–21], many high-permeability and high permselectivity materials have been discovered and synthesized. However, these high performance polymeric materials are often very expensive, while some of them are brittle. As a result, the fabrication of integrally skinned asymmetric membranes is either no longer feasible or economically attractive because it is too costly to prepare the entire membrane from the same material. One of potential solutions to overcome these problems is to fabri-

54

Y. Li et al. / Journal of Membrane Science 245 (2004) 53–60

cate dual-layer hollow fiber membranes by the co-extrusion technology. The dual-layer hollow fiber membrane basically consists of an asymmetric separating outer layer and a microporous supporting inner layer. The main function of the outer layer is to provide the permselectivity; therefore it should be made of a high-permeability and high-selectivity polymeric material. The main function of the inner layer is to provide the necessary mechanical support for the outer layer; therefore it may be made of low-cost polymers with good mechanical properties. Ideally, the interface between the two layers and the bulk structure of the inner layer must be porous in order to have a minimal gas transport resistance; otherwise the sub-structure resistance deduced from the interface and the supporting layer may result in significant decreases in both permeance and permselectivity [22]. In addition, the duallayer hollow fiber membranes must be delamination-free at the interface and have a D/h (dual-layer fiber diameter/wall thickness) ratio less than 2 in order to withstand high-pressure feed streams. There have been a few pioneer investigations on dual-layer hollow fiber membranes since late 1980s [23–31]. Ekiner et al.’s early work at DuPont is informative because the membranes developed are delamination-free at the interface and has good gas separation performance [25]. Suzuki et al. [27] also fabricated dual-layer hollow fiber membranes composed of a dense polyimide outer layer and a sponge-like inner layer made of another polyimide. However, their membranes may not withstand high-pressure environments because of high D/h ratios and their thinnest dense-selective layer is about 0.95 ␮m. Li et al. [30] conducted the first systematic study to investigate the effects of spinning conditions on dual-layer hollow fiber membranes and the causes of interfacial delamination between the two layers. They concluded that the inner-layer dope concentration, bore fluid composition and subsequent solvent exchange play important roles to produce delamination-free dual-layer membranes because the delamination is resulted from the difference in shrinkage rate between the two layers during the precipitation. Nevertheless, their dense-selective layer thickness is ˚ which is still too high when compared with around 8500 A those of single-layer asymmetric membranes. Jiang et al. extended Li et al.’s work and developed almost defectivefree matrimid/polyethersulfone dual-layer hollow fiber mem˚ [32] branes with a dense-selective layer thickness of 2886 A by controlling the dope composition and phase-inversion conditions. Later, Li et al. modified the dope compositions and demonstrated the dual-layer hollow fiber membranes with a ˚ [33]. dense-selective layer thickness of about 1500 A To fully compete and potentially replace the single-layer hollow fiber membranes, one must develop the co-extruding technology to produce dual-layer hollow fiber membranes ˚ with a dense-selective layer thickness of less than 1000 A. To our best knowledge, so far there is no academic literature available on this subject. Therefore, the objective of this study is to fabricate dual-layer hollow fiber membranes with an ultra-thin dense-selective layer. For concept

demonstration and the elimination of interfacial delamination, polyethersulfone (PES) was chosen as the material for both outer and inner layers because of its superior mechanical and membrane forming characteristics. However, the spinning dopes for outer and inner layers have different polymer and solvent/non-solvent compositions. The O2 /N2 selectivity of PES was reported to be 6.1 with an O2 permeability of 0.51 Barrer at 30 ◦ C and to be 5.1 with an O2 permeability of 0.81 Barrer at 50 ◦ C [34]. By using an Arrhenius relationship between permeability and temperature, the O2 permeability and O2 /N2 selectivity at 25 ◦ C can be calculated to be around 0.44 Barrer and 6.3 [13–15], respectively.

2. Experimental 2.1. Materials A commercial Radel A-300 PES, purchased from Amoco Performance Products Inc., Ohio, USA, was employed as the polymeric material. It has a weight-average molecular weight of about 15 000. The solvent, NMP (>99%), supplied by Merck Schuchardt, ethanol from Merck Inc., methanol from J.T. Baker and hexanes from EM Science, an associate of Merck, were used as received. 2.2. Preparation of spinning dopes The composition of the outer-layer dope solution was 35/50/15 (in wt.%) PES/NMP/ethanol and that of the inner-layer dope solution was 25/61/14 (in wt.%) PES/NMP/ethanol. The two dope solutions were prepared by the procedure described elsewhere [15,35]. After the dope solutions became homogenous, they were degassed for several hours in the bottle and then poured into a 500 ml syringe pump (ISCO Inc.) followed by degassing for one day before spinning. 2.3. Spinning process and solvent exchange The dual-layer hollow fiber membranes were fabricated by the dry-jet wet-spinning technique using a dual-layer spinneret as shown in our previous literature [30] and the flow rates of the bore fluid and both dope solutions were controlled by three ISCO pumps. The spinning parameters were listed in Table 1. The as-spun dual-layer hollow fiber membranes were rinsed in the water bath for 3 days and then carried out solvent exchange without further drying. The procedure of solvent exchange was to immerse these fibers into methanol for 30 min/time with three times and then n-hexane for 30 min/time with three times under stirring. Finally, these fibers were dried in the air at ambient temperature for further tests and study.

Y. Li et al. / Journal of Membrane Science 245 (2004) 53–60

55

Table 1 Spinning parameters of dual-layer PES hollow fiber membranes

The permeance, P/L, was calculated by the following equation [30]:

Parameters

Values

Outer-layer dope composition Out-layer dope flow rate (ml/min) Inner-layer dope composition Inner-layer dope flow rate (ml/min) Bore fluid composition Bore fluid flow rate (ml/min) Air gap (cm) Take up rate (cm/min) External coagulant Spinning temperature (◦ C) Coagulation bath temperature (◦ C)

PES/NMP/EtOH: 35/50/15 in wt.% 0.2 PES/NMP/EtOH: 25/61/14 in wt.% 0.6 NMP/H2O: 95/5 in wt.% 0.3 1.5 307 (keep free falling) Tap water 25 25

Q Q P = = L AP nDlPπ

where P is the permeability of separating layer (Barrer), L the thickness of the apparent dense-selective layer (cm), Q the pure gas flux (cm3 /s), n the number of fibers in one testing module, D the outer diameter of the testing fibers (cm), l the effective length of the modules (cm), and P the gas pressure difference cross the membrane (cmHg). We use GPU as the unit of permeance (1 GPU = 1 × 10−6 cm3 (STP)/cm2 s cmHg). The ideal separation factor is defined as follows [30]:

2.4. Post-treatment protocols αA/B = Two post-treatment protocols were used in order to seal the defects on the outer surface of resultant dual-layer hollow fibers. One was to make modules directly from the as-spun hollow fiber membranes and then dip-coat these modules with a 3 wt.% silicone rubber solution [15]. The other was to heat-treat the hollow fiber membranes, then make them into modules, finally perform the same dip-coating process. Two heat-treatment conditions were employed. Condition A directly heated these fibers in a vacuum oven from room temperature to 75 ◦ C, held there for 3 h and then cooled down naturally. Condition B heated these fibers from room temperature to 150 ◦ C with a heating rate of 1 ◦ C/min under vacuum; held at 150 ◦ C for 1 h and then cooled down naturally. The use of these two annealing temperatures was based on our experience and literature information. The use of heat-treatment to control pore size for composite membranes has been often practiced in industry [36,37] and their effects on membrane morphology have been summarized elsewhere [38]. Annealing reduces a substrate’s pore size and improves selectivity. Without annealing, the inventors could not produce useful membranes. The annealing technology was probably developed 30 years ago for PBI membranes in RO applications [39]. Without annealing in hot ethylene glycol (140–180 ◦ C), PBI membranes had a poor RO selectivity. This technology was extended by Cabasso and Tamvakis [40] for the development of polyethyleneimine/polysulfone (PS) hollow fibers for RO. They observed intrusion of a polymeric solution into pores to a depth as great as 0.5 ␮m. They contracted surface pore sizes and reduced intrusion by means of annealing microporous PS hollow fibers at 110–150 ◦ C for less than 30 min. 2.5. Module fabrication and gas permeation tests Testing modules fabrication and pure gas permeation measurements were introduced elsewhere [15]. Modules were composed of 5 or 10 hollow fibers with a length of around 10 cm, and 3–4 modules were tested as an order of O2 and N2 at room temperature and 200 psi for each experimental condition.

(1)

(P/L)A (P/L)B

(2)

The dense-selective layer thickness is estimated by the following equation [30]: L=

PO2 (dense film) (P/L)O2 (dual-layer hollow fiber membrane)

(3)

2.6. Characterization of dual-layer hollow fiber membranes The morphology of dual-layer hollow fiber membranes with different post-treatment protocols was characterized by SEM using a JEOL JSM-5600LV and JSM-6700F. Membrane samples were prepared with the procedure depicted elsewhere [15]. 3. Results and discussion 3.1. The effect of different post-treatment protocols on gas separation performance Table 2 summarizes the gas separation performance and the apparent dense-selective layer thickness of dual-layer hollow fiber membranes with different post-treatment protocols. Before the silicone rubber coating, it can be found that the permeance decreases with an increase in heat-treatment temperature. This is due to the fact that the heat-treatment induces relaxation of the stresses imposed in hollow fibers when they were fabricated. Hollow fibers shrink inwards gradually (i.e., reduced outer diameter), leading to higher packing density of polymer chains. Therefore, the heat-treatment not only reduces the surface defects but also the bulk porosity. However, all membranes with or without post-treatment exhibit an O2 /N2 selectivity of 0.93–0.94, indicating that Knudsen diffusion is still dominant and the surface pore size cannot be not reduced to follow the sorption–diffusion mechanism solely by the heat-treatment. After dip-coating in the silicone rubber solution, it can be found that the as-spun membranes have an O2 /N2 selectivity of 5.26 which is 83% of the intrinsic selectivity (i.e., 6.3) of

56

Y. Li et al. / Journal of Membrane Science 245 (2004) 53–60

Table 2 Gas separation performances of dual-layer PES hollow fiber membranes with different post-treatment protocols Dual-layer hollow fiber membrane name

As-spun fibers

Heat-treated fibers with Condition A

Heat-treated fibers with Condition B

Before Permeance (O2 ) (GPU) Permeance (N2 ) (GPU) Ideal selectivity

513 544 0.94

394 422 0.93

86.8 93.6 0.93

After coatingb Permeance (O2 ) (GPU) Permeance (N2 ) (GPU) Ideal selectivity ˚ Skin thickness (A)

11.3 2.15 5.26 389

10.8 1.80 6.00 407

4.04 1.14 3.54 1089

coatinga

a b

Tested at 25 ◦ C, 70 psi. Tested at 25 ◦ C, 200 psi.

˚ PES. The apparent dense-selective layer thickness is 389 A calculated from Eq. (3). The O2 /N2 selectivity increases impressively to 6.0 after heat-treating the dual-layer hollow fiber membranes at the Condition A (i.e., 75 ◦ C for 3 h). This selectivity is very close to the intrinsic selectivity of 6.3. In addition, the permeance only dips slightly to 10.8 GPU after heat-treatment at 75 ◦ C. This permeance is equivalent to an ˚ which is the thinnest apparent dense-selective layer of 407 A, one ever reported in the literature for the dual-layer hollow fiber membranes and is also comparable to that of state-ofthe-art single-layer asymmetric hollow fibers. Further increasing the heat-treatment temperature to 150 ◦ C not only lowers the O2 /N2 selectivity to 3.54, but ˚ also thickens the apparent dense-selective layer to 1089 A. This phenomenon may be arisen from the fact that a high-

temperature heat-treatment may induce densification of both selective skin and substructure. The densification of a selective skin may result in a reduction of permeance, while the densification of the substructure is manifested by a lower selectivity. 3.2. Membrane morphology Fig. 1 shows a typical morphology of the as-spun duallayer hollow fiber membrane produced in this study. Its outer diameter is around 560 ␮m and inner diameter is around 290 ␮m. Because of using the same PES and similar solvents in both layers, it significantly enhances the interfacial diffusion as pointed out in our previous paper [32], thus no clear delamination can be observed between the inner and outer

Fig. 1. Integrity of as-spun dual-layer hollow fiber membranes. (A) Overall profile; (B) membrane wall; (C) outer layer’s outer edge; (D) inner layer’s inner bulk; (E) inner layer’s inner skin; (F) outer layer’s outer skin.

Y. Li et al. / Journal of Membrane Science 245 (2004) 53–60

57

Fig. 2. Visual estimation of the dense-selective layer thickness of dual-layer hollow fiber membranes with different heat-treatment methods. Condition A (left); Condition B (right).

layers as shown in Fig. 1B and C. Based on the mass balance calculation using the information provided in Table 1, the thickness of the outer layer is less than 15 ␮m. The outer layer has an asymmetric structure, while the inner layer is fully porous. In addition, the bulk of the inner layer (Fig. 1D) and the inner surface of the inner layer (Fig. 1E) are fully porous, while the outer skin of the outer layer (Fig. 1F) is fully dense.

There exist multiple-layer macrovoids in the inner layer as shown in Fig. 1B and the macrovoid structures of these two layers are somewhat different. The macrovoids near the outer layer have a finger-like structure and have a smooth and less porous skin, meanwhile, most pores are created at the bottom of these macrovoids; while the macrovoids near the inner cavity of the fiber show a structure more like tear drops than fingers and have a fully porous skin interconnecting with

Fig. 3. A comparison of the cross-section morphology of dual-layer hollow fiber membranes with different post-treatment protocols. (A,D) As-spun; (B,E) Condition A; (C,F) Condition B.

58

Y. Li et al. / Journal of Membrane Science 245 (2004) 53–60

Fig. 4. Pore size comparison of inner layer’s inner skin of dual-layer hollow fiber membranes with different post-treatment protocols. As-spun (left); Condition A (middle); Condition B (right).

surrounding spongy cells. The macrovoid formation is a common phenomenon in phase-inversion membranes; however, its mechanisms are very complicated and have been heavily debated in the last 40 years [5,32,33,41–48]. It is generally accepted that fast precipitation with the aid of unbalanced localized stresses and solvent diffusion favors the formation of finger like macrovoids through the nucleation and growth of the polymer lean phase as well as polymer rich phase and the spinodal separation. The multiple-layer macrovoids can often be observed in fibers spun with low polymer concentrations possibly due to a rapid solvent exchange. The finger-like macrovoids near the outer layer start just beneath the outer skin of inner layer as illustrated in Fig. 1B and C. This phenomenon may suggest that the external strong coagulant, that is tap water, can move into the thin outer layer dope solution during the formation of nascent membranes and result in the formation of macrovoids in the inner layer with the aid of unbalanced localized stresses and rapid solvent diffusion. The tear droplike macrovoids near the inner cavity of the fiber may be induced by the combination of the finger-like macrovoids with a slower precipitation deduced by the internal coagulant. The structural difference between two types of macrovoids may be resulted from different coagulation rates and solvent exchange rates. The external coagulant, tap water, is a strong coagulant so that there is no enough time for macrovoids near the outer layer to propagate before the precipitation and is also such a powerful penetrant that it can diffuse downwards through a long distance in a short time; therefore macrovoids near the outer layer form a long and narrow finger-like structure. In contrast, macrovoids near the inner cavity of the fiber have plenty of time to relax outwards because the internal

coagulant is composed of 95/5 (in wt.%) NMP/H2 O, but the penetration rates of the coagulant greatly decrease simultaneously; thus leading to a short and wide tear drop-like structure. Fig. 2 displays the cross-section morphology near the outer edge of the outer layer after various heat-treatments. Interestingly, the calculated thicknesses of the apparent dense-selective layers are in agreement with the SEM observation. They are approximately 40 and 110 nm for membranes heat-treated at Conditions A and B, respectively. It can also be found from this figure that the Condition B significantly reduces the size and the amount of pores immediately beneath the dense-selective layer because of the high heattreatment temperature. Fig. 3 exhibits a comparison of cross-section morphology of dual-layer hollow fiber membranes with different posttreatment protocols. Macroscopically, there is no sharp difference in the overall cross-section morphology (Fig. 3A–C) except that membranes treated with a higher temperature (Condition B) apparently have slightly smaller pores and less slim finger-like macrovoids. Future studies will be focused on if a higher heat-treatment temperature can lead to the disappearance of small-size macrovoids. Fig. 3D–F show the substructure morphology underneath the outer layer, while Fig. 4 shows a comparison of porous structure in the inner surface of the inner layer as a function of post-treatment temperature. The as-spun fibers and the fiber heat-treated at Condition A (i.e., 75 ◦ C) apparently have similar substructure morphology and inner surface pores, while the fiber heat-treated at Condition B (i.e., 150 ◦ C) has denser substructure and smaller pores in the inner surface. This observation is consistent with our previous explanation why both permeance and selectivity decrease for membranes

Y. Li et al. / Journal of Membrane Science 245 (2004) 53–60

heat-treated at Condition B (i.e., 150 ◦ C). Based on the resistance model [49], an increased substructure resistance decreases the permeance of a fast gas more than that of a slow gas. As a result, the overall selectivity also decreases. 4. Conclusions Dual-layer PES hollow fiber membranes with an ultra˚ have successfully been thin dense-selective layer of 407 A fabricated by using co-extrusion and dry-jet wet-spinning phase-inversion techniques with the aid of heat-treatment. The effects of heat-treatment on membrane morphology and gas separation performances have also been investigated. The following conclusions can be drawn from this work: (1) Heat-treatment at 75 ◦ C for 3 h can effectively improve gas separation performance. The as-spun dual-layer hollow fiber membranes have an apparent dense-selective ˚ and an O2 /N2 selectivity of 5.26 layer thickness of 389 A after silicone coating. The O2 /N2 selectivity increases to 6.00 after heat-treatment at 75 ◦ C, while the apparent ˚ dense-selective layer thickness increases to 407 A. ◦ (2) Heat-treatment at 150 C for 1 h can lower both selectivity and permeance values due to the enhanced substructure resistance induced by the heat-treatment. SEM observation confirms this phenomenon. (3) Macrovoids formed just beneath the outer skin of inner layer have a long and narrow finger-like structure, while macrovoids near the inner cavity of the fiber have a short and wide tear drop-like structure. The structural difference between these two types of macrovoids may be resulted from different coagulation rates and solvent exchange rates. Acknowledgements The authors would like to thank NUS for funding this research with the grant number of R-279-000-108-112. References [1] W.J. Koros, I. Pinnau, Membrane formation for gas separation process, in: D.R. Paul, Y.P. Yampol’skii (Eds.), Polymeric Gas Separation Membranes, CRC Press, Boca Raton, 1994, Chapter 5. [2] R.E. Kesting, A.K. Fritzsche, Polymeric Gas Separation Membranes, Wiley, New York, NY, 1993. [3] I.M. Wienk, R.M. Boom, M.A.M. Beerlage, A.M.W. Bulte, C.A. Smolders, H. Strathmann, Recent advances in the formation of phase inversion membranes made from amorphous or semi-crystalline polymers, J. Membr. Sci. 113 (1996) 361. [4] T.S. Chung, A review of microporous composite polymeric membrane technology for air-separation, Polym. Polym. Comp. 4 (1996) 269. [5] H. Strathmann, K. Kock, P. Amar, R.W. Baker, The formation mechanism of asymmetric membranes, Desalination 16 (1975) 179. [6] W.S.W. Ho, K.K. Sirkar, Membrane Handbook, Van Nostrand Reinhold, New York, 1992.

59

[7] R.E. Kesting, A.K. Fritzsche, M.K. Murphy, C.A. Cruse, A.C. Handermann, R.F. Malon, M.D. Moore, The second-generation polysulfone gas-separation membrane. I. The use of Lewis acid: base complexes as transient templates to increase free volume, J. Appl. Polym. Sci. 40 (1990) 1557. [8] R.E. Kesting, A.K. Fritzsche, M.K. Murphy, A.C. Handermann, C.A. Cruse, R.F. Malon, Process for forming asymmetric gas separation membranes having graded density skins, US Patent 4,871,494 (1989). [9] A.K. Fritzsche, M.K. Murphy, C.A. Cruse, R.F. Malon, R.E. Kesting, Characterization of asymmetric hollow fiber membranes with gradeddensity skin, Gas Sep. Purif. 3 (1989) 106. [10] I. Pinnau, W.J. Koros, Defect-free ultrahigh flux asymmetric membranes, US Patent 4,902,422 (1990). [11] S.C. Pesek, W.J. Koros, Aqueous quenched asymmetric polysulfone hollow fibers prepared by dry/wet phase separation, J. Membr. Sci. 88 (1994) 1. [12] D.T. Clausi, W.J. Koros, Formation of defect-free polyimide hollow fiber membranes for gas separations, J. Membr. Sci. 167 (2000) 79. [13] T.S. Chung, E.R. Kafchinski, P. Foley, Development of asymmetric hollow fibers from polyimide for gas separation, J. Membr. Sci. 75 (1992) 181. [14] D.L. Wang, K. Li, W.K. Teo, Polyethersulfone hollow fiber gas separation membranes prepared from NMP/alcohol systems, J. Membr. Sci. 115 (1996) 85. [15] D.L. Wang, W.K. Teo, K. Li, Preparation and characterization of high-flux polysulfone hollow fibre gas separation membranes, J. Membr. Sci. 204 (2002) 247. [16] T.S. Chung, S.K. Teoh, X.D. Hu, Formation of ultrathin highperformance polyethersulfone hollow-fiber membranes, J. Membr. Sci. 133 (1997) 161. [17] J.J. Qin, T.S. Chung, Effects of orientation relaxation and bore fluid chemistry on morphology and performance of polyethersulfone hollow fibers for gas separation, J. Membr. Sci. 229 (2004) 1. [18] W.J. Koros, G.K. Fleming, S.M. Jordan, T.H. Kim, H.H. Hoehn, Polymeric membrane materials for solution-diffusion based permeation separations, Prog. Polym. Sci. 13 (1988) 339. [19] S.A. Stern, Polymers for gas separations: the next decade, J. Membr. Sci. 94 (1994) 1. [20] L.M. Robeson, W.F. Burgoyne, M. Langsam, A.C. Savoca, C.F. Tien, High performance polymers for membrane separation, Polymer 35 (1994) 4970. [21] B.D. Freeman, Basis of permeability/selectivity tradeoff relations in polymeric gas separation membranes, Macromolecules 32 (1999) 375. [22] I. Pinnau, W.J. Koros, Relationship between substructure resistance and gas separation properties of defect-free integrally-skinned asymmetric membranes, Ind. Eng. Chem. Res. 30 (1991) 1837. [23] E. Kuzumoto, K. Nitta, Manufacture of permselective hollow membranes, Japanese Patent 01,015,104 (1989). [24] T. Yoshinaga, H. Shimazaki, Y. Kusuki, Polyimide bilayer hollow membranes and their manufacture, Japanese Patent 02,169,019 (1990). [25] O.M. Ekiner, R.A. Hayes, P. Manos, Novel multicomponent fluid separation membranes, US Patent 5,085,676 (1992). [26] S. Nago, Y. Mizutani, Microporous polypropylene hollow fibers with double layers, J. Membr. Sci. 116 (1996) 1. [27] H. Suzuki, K. Tanaka, H. Kita, K. Okamoto, H. Hoshino, T. Yoshinaga, Y. Kusuki, Preparation of composite hollow fiber membranes of poly (ethylene oxide)-containing polyimide and their CO2 /N2 separation properties, J. Membr. Sci. 146 (1998) 31. [28] S.H. Yang, W.K. Teo, K. Li, Formation of annular hollow fibers for immobilization of yeast in annular passages, J. Membr. Sci. 184 (2001) 107. [29] T. He, M.H.V. Mulder, H. Strathmann, M. Wessling, Preparation of composite hollow fiber membranes: co-extrusion of hydrophilic coating onto porous hydrophobic support structures, J. Membr. Sci. 207 (2002) 143.

60

Y. Li et al. / Journal of Membrane Science 245 (2004) 53–60

[30] D.F. Li, T.S. Chung, R. Wang, Y. Liu, Fabrication of fluoropolyimide/polyethersulfone (PES) dual-layer asymmetric hollow fiber membranes for gas separation, J. Membr. Sci. 198 (2002) 211. [31] C.C. Pereira, R. Nobrega, K.V. Peinemann, C.P. Borges, Hollow fiber membranes obtained by simultaneous spinning of two polymer solutions: a morphological study, J. Membr. Sci. 226 (2003) 35. [32] L.Y. Jiang, T.S. Chung, D.F. Li, C. Cao, Fabrication of matrimid/ polyethersulfone dual-layer hollow fiber membranes for gas separation, J. Membr. Sci. 240 (2004) 91. [33] D.F. Li, T.S. Chung, R. Wang, Morphological aspects and structure control of dual-layer asymmetric hollow fiber membranes formed by a simultaneous co-extrusion approach, J. Membr. Sci., in press. [34] D.L. Wang, Polyethersulfone hollow fiber gas separation membranes prepared from solvent systems containing nonsolvent-additives, Ph.D. Dissertation, National University of Singapore, 1996. [35] T.S. Chung, E.R. Kafchinski, The effects of spinning conditions on asymmetric 6FDA/6FDAM polyimide hollow fibers for air separation, J. Appl. Polym. Sci. 65 (1997) 1555. [36] B. Bikson, J.K. Nelson, Composite membranes and their manufacture and use, US Patent 4,826,599 (1989). [37] A.Z. Gollan, Anisotropic membranes for gas separation, US Patent 4,681,605 (1987). [38] T.S. Chung, A review of microporous composite polymeric membrane technology for air-separation, Polym. Polym. Comp. 4 (1996) 269. [39] F.S. Model, L.A. Lee, in: H.K. Lonsdale, H.E. Podall (Eds.), Reverse Osmosis Membrane Research, Plenum, New York, 1972, p. 285.

[40] I. Cabasso, A.P. Tamvakis, Composite hollow fibre membranes, J. Appl. Polym. Sci. 23 (1979) 1509. [41] L. Broens, F.W. Altena, C.A. Smolders, D.M. Koenhen, Asymmetric membrane structures as a result of phase separation phenomena, Desalination 32 (1980) 33. [42] H. Strathmann, K. Kock, The formation mechanism of phase inversion membranes, Desalination 21 (1977) 241. [43] A.J. Reuvers, C.A. Smolders, Formation of membranes by means of immersion precipitation. Part II. The mechanism of formation of membranes prepared from the system cellulose acetate–acetone–water, J. Membr. Sci. 34 (1987) 67. [44] C.A. Smolders, A.J. Reuvers, R.M. Boom, I.M. Wienk, Microstructures in phase-inversion membranes. 1. Formation of macrovoids, J. Membr. Sci. 73 (1992) 259. [45] S.S. Shojaie, W.B. Krantz, A.R. Greenberg, Dense polymer film and membrane formation via the dry-cast process. 2. Model validation and morphological study, J. Membr. Sci. 94 (1994) 281. [46] S.A. Mckelvey, W.J. Koros, Phase separation, vitrification, and the manifestation of macrovoids in polymeric asymmetric membranes, J. Membr. Sci. 112 (1996) 29. [47] R. Matz, The structure of cellulose acetate membranes. 1. The development of porous structures in anisotropic membranes, Desalination 10 (1972) 1. [48] M.R. Pekny, J. Zartman, W.B. Krantz, A.R. Greenberg, P. Todd, Flow-visualization during macrovoid pore formation in dry-cast cellulose acetate membranes, J. Membr. Sci. 211 (2003) 71. [49] J.M.S. Henis, M.K. Tripodi, Composite hollow fiber membranes for gas separation: the resistance model approach, J. Membr. Sci. 8 (1981) 233.