Synthetic 6FDA-ODA copolyimide membranes for gas separation and pervaporation: correlation of separation properties with diamine monomers.
The first polyimides were synthesized in 1908, but more research interest in preparation and application of polyimides was attracted since 1955 when high molecular weight polyimides were successfully obtained through the two-step method . Attributed to the excellent thermal stability, the superior mechanical strength and good chemical resistance properties, polyimides found their applications in many industrial fields . The evolution of membrane separation technologies in the past few decades stimulates the investigation on the separation performances of polyimide membranes by incorporating their excellent physiochemical properties. Further studies and correlations of the chemical structures of polyimides with their separation properties have become a critical concern for the solution-diffusion membrane processes, such as gas separation and pervaporation [3, 4].
Relationships between the gas permeability and the chemical structure of membrane materials were studied in the early years [3-6], and new polyimides with special structural features were prepared for gas separation, enriching the knowledge of chemical structure versus separation properties. Kim et al.  studied the gas separation properties of a series of polyimides. The polymer segmental mobility and the packing properties were used to explain the difference in membrane permeabilities: a lower density and larger d-spacing of the membranes could lead to a higher permeability. Stem and Yamamoto  investigated gas permeabilities of different polyimide membranes, and found out that segmental mobility, inter-chain distance, and charge transfer complexes formed in membranes determined gas permeabilities. Highly soluble and [O.sub.2] permeable polyimides were synthesized from twisted biphenyl dianhydrides and spirobifluorene diamines by Kim et al. . The characteristic structures of the dianhydrides and the diamines caused a decrease in the degrees of molecular packing and crystallinity, and the restricted segmental mobility resulted in high permeabilities to [O.sub.2].
Because of the complexity of interactions between the membrane and the liquid mixture to be separated, polyimide pervaporation membranes were not explored intensively on the structure-property relationship. Xu et al.  studied the pervaporation properties of some polyimide membranes. The permeation flux was found to be in agreement with the orders of d-spacing and free volume of the polymers. High rigidity of the polymer chains was supposed to help increase pervaporation selectivity, while the low packing density of the segmental chains might lead to high flux .
Membrane permeabilities are governed by sorption and diffusivity of the penetrants in the membranes. From the results mentioned earlier, it can be summarized that diffusion properties are greatly influenced by the steric effects and the mobility of polymer chains. However, the steric effects of segmental chains may also partially account for the sorption properties, leading to a complicated mix-up. If the effects of monomer moieties on separation properties are quantitatively outlined, a better understanding and a deeper insight can be achieved.
In this work, the diamine monomers having similar main chains but different steric configurations were used to prepare copolyimides, and their moiety contributions to separation properties were investigated using the linear moiety contribution method proposed in our previous work .
4,4'-(Hexafluoroisopropylidene)diphthalic anhydride (6FDA, >98%) was obtained from TCI America. 4-Aminophenyl ether (ODA, 98%), 4-aminophenyl sulfone (DDS, 97%), 4,4'-methylenedianiline (MDA, 97%), 4,4'-bis(3-aminophenoxy)diphenyl sulfone (BADS, 97%), 2,6-difluorobenzonitrile (97%), 3-aminophenol (99%), 4,4'-difluorobenzophenone (99%), N,N-dimethylacetamide (DMAc, >99%), m-cresol (97%), tetrahydrofuran (THF, 99.5%) were provided by Acros Organics. Glycerol ([greater than or equal to]99%), formamide ([greater than or equal to]99%), and diiodomethane ([greater than or equal to]99%), isopropanol (anhydrous. 99.5%) were obtained from Sigma-Aldrich. Water (deionized, [less than or equal to]2 [micro]S/cm) was provided by the Department of Chemical Engineering, University of Waterloo. MDA was purified from recrystallization with water/ethanol. Other reagents and materials were used directly as received.
Bio-Rad Excalibur 3000MX spectrometer was used to record FTIR spectra. Monomers were tested in paraffin oil between NaCl crystals as a mull, and polymers films were casted on NaCl crystals with THF solutions. The Waters gel permeation chromatography (GPC) system with a DAWN[R] DSP-F Laser Photometer, a Waters 2410 Refractive Index Detector, and 3 PL gel 10 [micro]m Mixed-B columns (300 mm X 7.5 mm) was run to determine molecular weights and molecular weight distributions. The GPC was calibrated with polystyrene standards, THF as the eluent. [.sup.19.F] and [.sup.1.H] NMR spectra were recorded in pyridine-[d.sub.5] (for polymers) and DMSO-[d.sub.6] (for monomers) on Bruker 300 MHz Nuclear magnetic resonance spectrometer. Tantec contact angle meter (U.S. Patent 5,268,733) was used to get the contact angles images for water, glycerol, formamide, and diiodomethane on membranes at room temperature.
Synthesis of Monomers
2,6-Bis(3-aminophenoxyl)benzonitrile (DABN) and 4,4'-bis(3-aminophenoxy) benzophenone (BABP) were synthesized from nucleophilic substitution reactions of 3-aminophenol in DMAc with 2,6-difluorobenzonitrile and 4,4'-difluorobenzophenone, respectively [12, 13].
DABN IR (mull, [cm.sup.-1]): 3466(w) [v.sub.as] N-H, 3378(w) [v.sub.s] N-H, 2231(w) v C[equivalent to]N, 1625(br, m) v C[equivalent to]N, 1573(m) phenyl ring (skeletal vibrations), 1293(w) v C-N, 1027(m) [v.sub.s] C-O-C, [.sup.1.H] NMR (DMSO-[d.sub.6], [delta] ppm): 7.52 (t, J = 9 Hz, 1H), 7.08 (t, 2H), 6.60 (d, J = 9 Hz, 2H), 6.45 (d, J = 9 Hz, 2H), 6.29 (m, 4H), 5.38 (s, 4H).
BABP IR (mull, [cm.sup.-1]): 3473(w) [v.sub.as] N-H, 3370(w) [v.sub.s] N-H, 1634(m) v C=O, 1592(br, m) phenyl ring (Skeletal vibrations), 1281(w) v C-N, 1226(w) [v.sub.as] C-O-C. [.sup.1.H] NMR (DMSO-[d.sub.6], [delta] ppm): 7.76 (d, J = 9 Hz, 4H), 7.06 (m, 6H), 6.42 (d, J = 6 Hz, 2H), 6.29 (s, 2H), 6.24 (d, J = 9 Hz, 2H), 5.30 (s, 4H).
Polymers and Membranes
Polyimides (shown in Table 1) were synthesized from the one-step polymerization in m-cresol .
6FDA-8ODA-2DDS (film, [cm.sup.-1]): 3490(vw) [v.sub.as] N-H, 3076(br, w) aromatic [delta] C-H, 1785(m) [v.sub.as] C=O, 1728(s) [v.sub.s] C=O, 1594(w) phenyl ring, 1502(s) phenyl ring, 1379 (m) v C-N, 1299(m) [v.sub.s] C[F.sub.3], 1245(s) [v.sub.as] C-O-C, 1114(br, m) imide, 726(m) [delta] imide. [.sup.1.H] NMR (Pyridine-[d.sub.5], [delta] ppm): 8.47-8.45 (0.8H), 8.17-8.04 (6.8H), 7.92-7.89 (3.2H), 7.33-7.30 (3.2H).
6FDA-6ODA-4DDS (film, [cm.sup.-1]): 3494(vw) [v.sub.as] N-H, 3081(br, w) aromatic [delta] C-H, 1786(m) [v.sub.as] C=O, 1728(s) [v.sub.s] C=O, 1593(w) phenyl ring, 1502(s) phenyl ring, 1377 (s) v C-N, 1331(vw) [v.sub.as] S[O.sub.2], 1298(m) [v.sub.s] C[F.sub.3], 1244(s) [v.sub.as] C-O-C, 1101(br, m) imide, 723(m) [delta] imide. [.sup.1.H] NMR (Pyridine-[d.sub.5], [delta] ppm): 8.47-8.44 (1.6H), 8.17-8.04 (7.6H), 7.92-7.88 (2.4H), 7.33-7.30 (2.3H).
6FDA-8ODA-2BADS (film, [cm.sup.-1]): 3493(vw) [v.sub.as] N-H, 3070(br, w) aromatic [delta] C-H, 1785(m) [v.sub.as] C=O, 1727(s) [v.sub.s] C=O, 1584(w) phenyl ring, 1501(s) phenyl ring, 1380(m) v C-N, 1298(m) [v.sub.s] C[F.sub.3], 1245(m) [v.sub.as] C-O-C, 1106(m) imide, 722(m) [delta] imide. [.sup.1.H] NMR (Pyridine-[d.sub.5], [delta] ppm): 8.20-8.04 (6.8H), 7.93-7.90 (3.2H), 7.79-7.76 (0.4H), 7.71 (0.3H), 7.64-7.61 (0.8H), 7.33-7.30 (3.2H), 7.18 (0.8H).
6FDA-6ODA-4BADS (film, [cm.sup.-1]): 3493(vw) [v.sub.as] N-H, 3074(br, w) aromatic [delta] C-H, 1785(m) [v.sub.as] C=O, 1727(s) [v.sub.s] C=O, 1584(w) phenyl ring, 1501(s) phenyl ring, 1380(m) v C-N, 1326(vw) [v.sub.as] S[O.sub.2], 1297(m) [v.sub.s] C[F.sub.3], 1245 (s) [v.sub.as] C-O-C, 1106(m) imide, 722(m) [delta] imide. [.sup.1.H] NMR (Pyridine-[d.sub.5], [delta] ppm): 8.19-8.01 (7.6H), 7.92-7.88 (2.4H), 7.79-7.76 (0.7H), 7.71 (0.6H), 7.64-7.61 (0.8H), 7.33-7.30 (2.4H), 7.18 (2.4H).
6FDA-8ODA-2MDA (film, [cm.sup.-1]): 3494(vw) [v.sub.as] N-H, 3075(br, w) aromatic [delta] C-H, 2965(vw) [v.sub.as] C[H.sub.2], 1784(m) [v.sub.as] C=O, 1726(s) [v.sub.s] C=O, 1597(w) phenyl ring, 1502(s) phenyl ring, 1379(m) v C-N, 1298(m) [v.sub.s] C[F.sub.3], 1246(br, s) [v.sub.as] C-O-C, 1112(br, m) imide, 723(m) [delta] imide. [.sup.1.H] NMR (Pyridine-[d.sub.5], [delta] ppm): 8.17-8.04 (6.0H), 7.92-7.90 (3.2H), 7.85-7.82 (0.8H), 7.49-7.46 (0.8H), 7.33-7.30 (3.2H), 4.09 (0.4H).
6FDA-6ODA-4MDA (film, [cm.sup.-1]): 3492(w) [v.sub.as] N-H, 3071(br, w) aromatic [delta] C-H, 2873(vw) [v.sub.s] C[H.sub.2], 1784(m) [v.sub.as] C=O, 1728(s) [v.sub.s] C=O, 1598(w) phenyl ring, 1502(s) phenyl ring, 1377(s) v C-N, 1300(m) [v.sub.s] C[F.sub.3], 1260(br, s) [v.sub.as] C-O-C, 1095(br, m) imide, 723(m) [delta] imide. [.sup.1.H] NMR (Pyridine-[d.sub.5], [delta] ppm): 8.17-8.04 (6.1H), 7.92-7.90 (2.4H), 7.85-7.82 (1.6H), 7.49-7.46 (1.6H), 7.33-7.30 (2.4H), 4.09 (0.7H).
6FDA-8ODA-2BABP (film, [cm.sup.-1]): 3494(vw) [v.sub.as] N-H, 3073(br, w) aromatic [delta] C-H, 1784(m) [v.sub.as] C=O, 1728(s) [v.sub.s] C=O, 1658(m) v C=O ketonic, 1589(m) phenyl ring, 1502(s) phenyl ring, 1379(m) v C-N, 1300(m) [v.sub.s] C[F.sub.3], 1246(s) [v.sub.as] C-O-C, 1112(br, m) imide, 721(m) [delta] imide. [.sup.1.H] NMR (Pyridine-[d.sub.5], [delta] ppm): 8.17-8.01 (6.0H), 7.96-7.89 (4.0H), 7.79-7.77 (0.8H), 7.66-7.63 (0.4H), 7.33-7.30 (3.5H), 7.24 (0.8H).
6FDA-6ODA-4BABP (film, [cm.sup.-1]): 3493(vw) [v.sub.as] N-H, 3074(br, m) aromatic [delta] C-H, 1785(m) [v.sub.as] C=O, 1727(s) [v.sub.s] C=O, 1655(m) v C=O ketonic, 1583(s) phenyl ring, 1501(s) phenyl ring, 1378(s) v C-N, 1302(m) [v.sub.s] C[F.sub.3], 1243(s) [v.sub.as] C-O-C, 1102(br, m) imide, 726(m) [delta] imide. [.sup.1.H] NMR (Pyridine-[d.sub.5], [delta] ppm): 8.17-8.01 (6H), 7.96-7.88 (4H), 7.79-7.78 (1.5H), 7.68-7.63 (0.8H), 7.33-7.30 (3.2H), 7.24 (1.5H).
6FDA-8ODA-2DABN (film, [cm.sup.-1]): 3494(vw) [v.sub.as] N-H. 3075(br, w) aromatic [delta] C-H, 2231(w) v C[equivalent to]N, 1785(m) [v.sub.as] C=O, 1728(s) [v.sub.s] C=O, 1626(w) v C[equivalent to]N, 1602(br, m) phenyl ring, 1502(s) phenyl ring, 1381(s) v C-N, 1298(m) [v.sub.s] C[F.sub.3], 1249(s) [v.sub.as] C-O-C, 1114(br, m) imide, 723(m) [delta] imide. [.sup.1.H] NMR (Pyridine-[d.sub.5], [delta] ppm): 8.17-8.04 (6.0H), 7.92-7.82 (4.0H), 7.67-7.65 (0.4H), 7.38-7.30 (3.8H), 6.73-6.70 (0.4H).
6FDA-6ODA-4DABN (film, [cm.sup.-1]): 3494(vw) [v.sub.as] N-H, 3071(br, w) aromatic [delta] C-H, 2232(w) v C[equivalent to]N, 1786(m) [v.sub.as] C=O, 1728(s) [v.sub.s] C=O, 1626(w) v C[equivalent to]N, 1604(br, m) phenyl ring, 1502(s) phenyl ring, 1378(m) v C-N, 1298(m) [v.sub.s] C[F.sub.3], 1246(s) [v.sub.as] C-O-C, 1103(br, m) imide, 721(m) [delta] imide. [.sup.1.H] NMR (Pyridine-[d.sub.5], [delta] ppm): 8.17-8.01 (6.0H), 7.92-7.82 (4.0H), 7.67-7.62 (0.8H), 7.38-7.30 (3.6H), 6.73-6.70 (0.8H).
Polymers could readily dissolve in THF, and 5 (w/v) % THF solutions were obtained and filtered with Target PTFE syringe filters (National Scientific, pore size 0.45 [micro]m). Polymer solutions were cast on clean and smooth glass plates, and kept in a dry clean chamber for at least 4 h. The slow evaporation of THF was required to form good membranes. To prevent possible defects in the membranes, polymer solutions were cast onto the dried membranes on the glass plates 2-4 times. Membranes were peeled off the glass plates and were ready to use. Thicknesses of the dense membranes were measured with a micrometer to be 15-30 [micro]m.
Pure Gas Permeation
Gas permeation performance was investigated for each membrane with [N.sub.2], [O.sub.2], He, and [H.sub.2], followed by C[O.sub.2]. Membranes had an effective area of 2.38 X [10.sup.-3] [m.sup.2]. Pure gases permeated through the membranes driven by the pressure difference between the feed side (high pressure) and the permeate side (atmospheric pressure). Feed pressures were controlled to be 0.14, 0.21, 0.28, and 0.35 MPa, respectively. Permeation rates, expressed in permeation flux, J, were measured with a soap bubble flow meter, and calculated from Eq. 1:
J = q/S[DELTA]P, (1)
where q(STP) is the volumetric flux ([cm.sup.3] [s.sup.-1]) at standard conditions, S ([cm.sup.2]) is the effective area of the membrane, and [DELTA]P (cm Hg) is the pressure difference across the membrane. Therefore, the flux J has a unit of GPU, 1 GPU = 7.5 x [10.sup.12] [m.sup.3] (STP) [m.sup.-2] [s.sup.-1] [Pa.sup.-1].
Ideal separation factor [[alpha].sub.i/j] for the gas pair i and j can be calculated from Eq. 2:
[[alpha].sub.i/j] = [J.sub.i]/[J.sub.j]. (2)
Pure water permeation was conducted at 30, 40, 50, and 60[degrees]C continuously, followed by dehydration of isopropanol. The feed water content was kept at ~20 wt%, while temperatures were controlled at 30-60[degrees]C. Then the feed temperature was remained at 60[degrees]C, and the feed water contents varied from ~10 to ~50 wt%. Permeation flux F is defined as follows:
F = Q/St. (3)
where Q, S, t is total amount of permeate, membrane area, operating time, respectively.
RESULTS AND DISCUSSION
Monomers and Polymers
BABP and DABN. In the FTIR spectra of BABP and DABN, the asymmetrical and symmetrical stretching vibration of amino groups absorbs at ~3470 [cm.sup.-1] and ~3370 [cm.sup.-1], respectively. The C-N stretching absorption occurs at ~1285 [cm.sup.-1]. Skeletal vibrations of the phenyl rings are found to be in the region of 1570-1590 [cm.sup.-1]. The aryl ether bonds show asymmetrical stretching vibration at 1226 [cm.sup.-1] for BABP and symmetrical stretching vibration at 1027 [cm.sup.-1] for DABN. respectively . In the spectrum of DABN, the characteristic absorption of the nitrile groups occurs at 2231 [cm.sup.-1], and the stretching vibration of nitrile groups absorbs at 1625 [cm.sup.-1]. In FTIR of BABP, the absorption at 1635 [cm.sup.-1] may be assigned to C=O stretching vibration, and it may shift to a lower wavenumber because of the two phenyl rings [14, 15]. In the NMR spectra of BABP and DABN, the protons in the residual moieties of 3-aminophenol have their chemical shifts at the same positions, and the resonance peaks at ~5.3 ppm are assigned to the protons in amino groups in BABP and DABN, respectively.
Copolyimides. Copolyimides were synthesized from one-step high-temperature polycondensation of 6FDA and ODA with the third monomers. As shown in Fig. 1, DDS and MDA have molecular structures similar to ODA, and the only difference lies in the sulfonyl group and the methylene bond, respectively, instead of the ether bond in ODA, while BADS, BABP, and DABN have 3-amino-phenyl moieties in the molecules. Although the one-step polycondensation is different from the two-step procedures for preparing polyimides, it still can be considered to consist of two stages. In the first stage, the dianhydride monomer reacts with diamines in m-cresol at a low temperature (~80[degrees]C) to form poly(amic acid)s, and in the second stage at a high temperature (~200[degrees]C), imidization reaction occurs with isoquinoline as the catalyst to form polyimides.
In the FTIR spectra of the copolyimides, a very weak absorption band at ~3493 [cm.sup.-1] representing the "free" asymmetrical N-H stretching modes , is observed for amino groups, possibly including the uncyclized amino groups and the polyimide end groups. Aromatic C-H stretching band occurs at ~3075 [cm.sup.-1] for all polyimides. At ~1600 and ~500 [cm.sup.-1], the spectra show two absorption bands of the skeletal stretching of aromatic rings. The five characteristic absorption bands of the imide ring occur at 1785 [cm.sup.-1] (C=O asymmetrical stretching), 1728 [cm.sup.-1] (C=O symmetrical stretching), 1380 [cm.sup.-1] (C-N stretching), 1110 [cm.sup.-1] (imide III band), and 723 [cm.sup.-1] (imide ring bending vibration) [16-19]. The asymmetrical stretching vibration of -C[F.sub.3] groups produces absorption at ~1300 [cm.sup.-1], and the stretching absorption of the ether bond in ODA moieties appears at ~1245 [cm.sup.-1]. The very weak absorption at 1331 and 1326 [cm.sup.-1] for 6FDA-6ODA-4DDS and 6FDA-6ODA-4BADS, respectively, may be due to the asymmetrical stretching vibration of sulfonyl groups . The vibration of aromatic-aromatic ketone produces absorptions at 1658 and 1655 [cm.sup.-1] in the spectra of 6FDA-8ODA-2BABP and 6FDA-6ODA-4BABP, respectively. Sharp absorption peaks can be observed for 6FDA-8ODA-2DABN and 6FDA-6ODA-4DABN at 2232 [cm.sup.-1], which is attributed to the vibration of nitrile groups in the polymer chains.
[FIGURE 1 OMITTED]
In the [.sup.1.H] NMR spectra of the copolyimides, the coupled peaks at 7.91 and 7.31 ppm are assigned to the protons of the ODA moiety. The chemical shifts of protons in the 6FDA moiety are located at 8.2-8.0 ppm. The protons attached to [alpha]-C of the sulfonyl groups in DDS and BADS show peaks at 8.45 and 8.19 ppm, respectively, while the protons attached to [beta]-C of the sulfonyl groups produce peaks at ~8.14 and ~7.18 ppm, respectively. The methylene groups of MDA moieties in 6FDA-6ODA-4MDA show a peak at 4.09 ppm, and the two peaks coupled at 7.83 and 7.48 ppm are assigned to the protons attached to the phenyl rings. In NMR spectrum of 6FDA-6ODA-4DABN and 6FDA-ODA-BABP, chemical shifts of protons in the residual moieties of 2,6-difluoro-benzonitrile and 4,4'-difluorobenzophenone do not shift too far away from their monomers DABN and DABP, but the chemical shifts of the protons in the residual moieties of 3-aminophenol show great difference between the monomers and the polymers, which also indicates the formation of imide bonds in the polymer chains.
The molecular weights and molecular weight distributions of the copolyimides are listed in Table 1. Four polyimides, 6FDA-8ODA-2BABP, 6FDA-6ODA-4BABP, 6FDA-8ODA-2BADS, and 6FDA-6ODA-4BADS, have higher molecular weights than the others. The molecular weight distributions are calculated from [bar.M.sub.n] and [bar.M.sub.w], and the values vary in the range of 1.18-1.67. Since these copolyimides were synthesized from the same procedures and under the same conditions, the difference in the molecular weight is supposed to result from the difference in the monomer reactivities. It is known that synthesis of polyimides is favored by the basicity of the diamines. Therefore, comparison of the charges on N-atom of the -[NH.sub.2] groups can lead to a better understanding of the basicity of the diamines and thus their reactivities. Using MOPAC, Wang-Ford charges were determined: ODA (-0.940) > BABP (-0.932), BADS (-0.931), DABN (-0.930) > MDA (-0.926) > DDS (-0.920). This sequence indicates the charge densities on N-atoms and the monomer reactivities. 6FDA-8ODA-2DDS and 6FDA-6ODA-4DDS have lower molecular weights, and this is also consistent with the sequence of Wang-Ford charges.
Properties of Polyimides
Solubility. All polymers are readily soluble in THF and pyridine, and they can dissolve in chloroform, DMAc, N,N-dimethylformamide, N-methylpyrrolidone, acetone, and dimethyl sulfoxide, but they are insoluble in isopropanol, toluene, and cyclohexane. THF was used as the solvent to prepare polyimide solutions when casting membranes.
Surface Free Energies. The sessile drop method was applied in the measurement of contact angles of liquids on horizontal membranes at room temperature. Digital images were recorded and analyzed with the shape analysis software. Table 2 shows the results in form of a mean with a 95% confidence interval. Since contact angles are directly related to the affinity of membranes toward the testing liquids, a rough judgment can be made that 6FDA-based membranes show great hydrophobicity based on the contact angles with water, but more evidences from contact angles with other liquids are still required.
Surface free energy was defined as the sum of the Lifshitz-van der Waals apolar (LW) component [[gamma].sub.i.sup.LW] and the Lewis acid-base (AB) component [[gamma].sub.i.sup.AB] in van Oss acid-base approach . Lifshitz-van der Waals component [[gamma].sub.i.sup.LW] is the apolar component related to Lifshitz-van der Waals interactions, the acid-base component [[gamma].sub.i.sup.AB] is associated with the acid-base interactions including hydrogen bonding, [pi]-bonds, and ligand formation. It was further split into two mono-polar surface parameters: Lewis acid (electron-acceptor) component [[gamma].sup.+] and Lewis base (electron-donor) component [[gamma].sup.-] . Accordingly the membrane-water interfacial free energies [DELTA][G.sub.sw.sup.IF] can be calculated [21, 22].
As shown in Table 3, the LW components [[gamma].sub.s.sup.LW] have much larger values than the AB components [[gamma].sub.s.sup.AB]. Thus it is reasonable to say that the apolar interactions play an important role in the surface free energies, however, the effects of acid-base interactions are more essential. A smaller value of [[gamma].sub.s.sup.+] is an indication of residual hydration , and a larger value of [[gamma].sub.s.sup.-] suggests higher hydrophilicity . The predominant monopolar Lewis base behavior of the polyimide membranes may be also due to the residual amino groups in the polymers. Among the membranes tested, 6FDA-6ODA-4DABN has a smaller [[gamma].sub.s.sup.LW], but both [[gamma].sub.s.sup.+] and [[gamma].sub.s.sup.-] are larger than the other membranes. Therefore, 6FDA-6ODA-4DABN is supposed to be the most hydrophilic membrane. The water-membrane interfacial free energies [DELTA][G.sub.sw.sup.IF] offer a direct view of the affinity of membrane surfaces toward water, van Oss et al.  found a boundary value of the interfacial free energies for clay materials, and pointed out that a negative value of [DELTA][G.sub.sw.sup.IF] represented hydrophobicity of the surface. Therefore, results from this work reveal that 6FDA-based membranes show great hydrophobicity, and this is not surprising because of the presence of F-atoms in the polymer chains. However, the hydrophobicity can be reduced by the incorporation of functional groups, such as the nitrile groups in DABN moieties.
[FIGURE 2 OMITTED]
Gas Permeation Properties
Pure gas permeation was performed at room temperature at different feed pressures. As shown in Fig. 2, the membrane selectivities at different pressures did not show substantial changes for the same membranes. Therefore, the average values of ideal separation factors were utilized in further studies. Table 4 shows the average gas permeances and the ideal separation factors for [O.sub.2]/[N.sub.2], [H.sub.2]/[N.sub.2], He/[N.sub.2], and C[O.sub.2]/[N.sub.2]. 6FDA-6ODA-4BABP and 6FDA-6ODA-4DABN showed the best selectivities toward [O.sub.2]/[N.sub.2], [H.sub.2]/[N.sub.2], and He/[N.sub.2], while 6FDA-8ODA-2DABN and 6FDA-6ODA-4DABN exhibited the best separation performances for C[O.sub.2]/[N.sub.2].
According to the solution-diffusion model, gas solubility, and diffusivity are two parameters that control gas permeation properties in polymer membranes, and they reflect the interactions between penetrants and membranes. Therefore, properties of gases and membrane materials should be studied first for gas permeation in membranes.
The gas solubility is basically related to its condensability . The gas that has a higher critical temperature (more condensable) is more soluble in membranes. Therefore, the permeation of C[O.sub.2] is probably controlled by its solubility in membranes because of the high critical temperature.
Penetrant diffusivity is directly related to the mobility of the penetrant in the polymer matrix, which is regulated by the size of the penetrant as well as the stereostructure of the polymer matrix . Although there is no apparent path for gas molecules in the nonporous dense membranes, the thermal movement of the chain segments can produce transient gaps, and penetrant molecules can jump between these gaps . In glassy polymers, polymer matrix is in a "frozen" state, and it behaves like a "molecular sieve" for the penetrants. Therefore, the mobility of gases in the matrix can be decided in the following sequence based on their kinetic diameters: [N.sub.2] < [O.sub.2] < C[O.sub.2] < [H.sub.2] < He . On the other hand, the structure and rigidity of the polymer also show great influence on the mobility of gases. Large side groups and flexible chains are factors that help to improve the diffusion of gases in polymer membranes.
An empirical quantitative method on linear contributions of monomer moieties was proposed in our previous work for [[alpha].sub.[O.sub.2]/[N.sub.2]], [[alpha].sub.[H.sub.2]/[N.sub.2]], [[alpha].sub.He/[N.sub.2]] and [[alpha].sub.C[O.sub.2]/[N.sub.2]], respectively. The linear moiety contribution method is expressed as follows :
f = [summation over i] ([a.sub.i][X.sub.i]) + [summation over j] ([b.sub.j][Y.sub.j]), (4)
where f is the parameter characterizing the membrane property (it can be the ideal separation factor for gas separation), [a.sub.i] and [b.sub.j] are the molar fractions of dianhydride moieties and diamine moieties, respectively, [summation over i] [a.sub.i] = 1, [summation over j] [b.sub.j] = 1, [X.sub.i] and [Y.sub.j] are moiety contribution factors for dianhydride moieties and diamine moieties, respectively .
From least square regression of the ideal separation factors of 6FDA-based polyimide membranes , contributions of monomer moieties to ideal separation factors were calculated and are listed in Table 5. The contributions of monomer moieties to [[alpha].sub.C[O.sub.2]/[N.sub.2]] are very complicated because they are combinations of diffusivity effects and solubility effects. It is observed that the BADS moiety shows a negative effect on [[alpha].sub.C[O.sub.2]/[N.sub.2]], and the DABN moiety is very helpful in promoting [[alpha].sub.C[O.sub.2]/[N.sub.2]].
From Table 5, it is observed that 6FDA, MDA, BABP, and DABN moieties have positive contributions to all the ideal separation factors, while ODA, DDS, and BADS contribute negatively to [[alpha].sub.[H.sub.2]/[N.sub.2]] and [[alpha].sub.He/[N.sub.2]]. It is easy to track the contributions of monomer moieties simply from their structures. It is believed that -C[F.sub.3] (in 6FDA) and methylene (in MDA) groups are beneficial to increase the flexibility of the polymer main chains, and C=O (in BABP), -CN (in DABN) as well as -C[F.sub.3] (in 6FDA) groups may provide more space for gas transport in the membranes. The contribution of the ODA moiety indicates that the flexible ether bond may produce transient gaps that are only more favorable to [O.sub.2], but the great change in local concentration of 6FDA moieties offsets the increase in gas diffusivity. The effects of the BADS moiety are analogous with the ODA moiety, but the sulfonyl groups have an entirely different impact on the diffusion of gases. The moiety of phenyl sulfone exerts enormous steric effect, and huge inter-chain space may be created within the matrix, resulting in a great increase in the diffusivities of gases including [N.sub.2]. The flexibility of the ether bonds in the BADS moiety may counteract part of the steric effects from the moiety of phenyl sulfone (i.e., the DDS moiety) in regards to the contributions to the selectivities toward [O.sub.2]/[N.sub.2], [H.sub.2]/[N.sub.2], and He/[N.sub.2].
[FIGURE 3 OMITTED]
Comparisons of the membrane selectivities with the calculated ideal separation factors from the linear moiety contribution method are made for the copolyimide membranes, as shown in Fig. 3. The data calculated can match the experimental data, and [[alpha].sub.[O.sub.2]/[N.sub.2]], and [[alpha].sub.C[O.sub.2]/[N.sub.2]], show better agreement than [[alpha].sub.He/[N.sub.2]], and [[alpha].sub.[H.sub.2]/[N.sub.2]].
Effects of the feed concentration and the operating temperature were studied through dehydration of isopropanol and pure water permeation processes. According to the solution-diffusion model, vacuum pervaporation can be divided into three steps: (1) sorption of separating species into the membrane at the feed side surface, (2) diffusion of the sorbed penetrants between the upstream surface and the downstream surface by the driving force of chemical potential gradients, and (3) desorption/evaporation of the permeates at the vacuum side [28, 29]. Therefore, sorption/desorption and diffusion properties have to be investigated for the mass transport of pervaporation.
Sorption properties describe solubility of the sorbed components in the polymer matrix. Interactions between the penetrants and the polymer, such as van der Waals forces and hydrogen bonding, are generally involved in the course of sorption. Sulfonyl groups in DDS and BADS and nitrile groups in DABN are polar groups, and they are favorable to the adsorption of polar species, e.g., water. The acyl groups in BABP can form hydrogen bonds with water, and hence the permeation of water in the BABP moiety-containing membranes will benefit more from sorption. The packing density of the polymer chains, or the size of the inter-chain space in the matrix that can accommodate the sorbed molecules, is another factor to determine the sorption properties. The desorption step is considered to be very fast and nonselective in vacuum pervaporation .
Diffusion of the penetrants in a polymer matrix is a course in which penetrant molecules jump between the thermally formed transient gaps between polymer chains, the same as described in the gas permeation mechanism. Therefore, the size of the penetrants and the size of the gaps including the inter-chain space, will control the diffusion properties. Water is much smaller in size than isopropanol, and it will diffuse faster. Flexible polymer chains can form larger gaps for the penetrants to pass through.
Effects of Monomer Structure on Selectivity. Figure 4 shows the permeate water content for pervaporation with various feed water contents and at different operating temperatures. The membrane selectivity can be directly compared from the permeate water contents.
[FIGURE 4 OMITTED]
On the basis of Fig. 4, a rough sequence of monomer contributions to the pervaporation selectivity can be obtained for the monomers used in this work: DDS(-) < MDA < BADS, and DABN < BABP, the negative sign for DDS means that a negative contribution was made by the DDS moiety to the pervaporation selectivity. It is known that the MDA moiety has more affinity to isopropanol but less to water. However, the flexible methylene groups in the MDA moiety may cause the polymer chains to pack more densely if compared with ODA. Thus, the MDA moieties in membranes will lead to a high permeate water content. However, the selectivity of MDA moiety-containing membranes is still lower than the other membranes. As discussed in gas separation, larger inter-chain space in the polymer matrix can be produced because of the steric effects of the sulfonyl groups in DDS and BADS. More isopropanol and water can pass through the membranes containing DDS and BADS moieties, even though the sulfonyl groups may interact with water molecules and the better sorption can be achieved for water. In BADS moieties, however, the flexible ether bonds reduce the steric effect of the sulfonyl groups. Therefore, the BADS moiety still has a positive contribution to the membrane selectivity, while the DDS moiety shows a negative contribution. Hydrogen bonding between water and BABP moieties is stronger than the van der Waals forces between water and the nitrile groups in DABN moieties. Accordingly, higher selectivities are observed for the membranes having BABP moieties than those containing DABN moieties.
Effects of the Feed Concentration. As shown in Fig. 4a, with a higher feed water content, most membranes produce higher permeation flux and a higher water content in the permeate, but the DDS moiety-containing membranes show the highest permeate water contents at the feed water contents of 40 wt%.
On the basis of the experimental data, the effect of feed concentration on permeation flux was described by an empirical equation when the feed concentration is below 50 wt% :
F([x.sub.w]) = k[x.sub.w.sup.n](0.1 < [x.sub.w] < 0.5), (5)
where F([x.sub.w]) is the total permeation flux at a feed water content of [x.sub.w] (mass fraction), n is an exponential parameter, and it was defined as the "concentration coefficient" for total flux, and k was defined as a parameter related to the membrane material and thickness as well as operation temperatures .
[FIGURE 5 OMITTED]
The concentration coefficient reflects the intrinsic properties of pervaporation membranes. Good regressions were achieved in curve fitting with Eq. 5, and Table 4 shows the n values and the coefficients of determination [r.sup.2]. Obviously, the n values are related to the numbers of the monomer moieties. Therefore, the linear moiety contribution method was applied to the n values, and least square regression was made with the n values of 6FDA-based membranes. The moiety contribution factors are listed in Table 5. From the obtained moiety contribution factors, concentration coefficients were recalculated using Eq. 5, and were compared with those from curve fitting. As shown in Fig. 5a, the two sets of n values can match well for most of the membranes.
ODA, MDA, and DABN moieties have smaller n values, which means that the permeation flux of the membranes containing these moieties are more sensitive to the change in the feed water content. This is probably attributed to the moiety contributions to the sorption properties of the membranes. It is known that the improved sorption properties can promote the permeation of penetrants. BABP moieties can form favorable interactions with water, and a better sorption can be achieved. However, the sorption saturation in the polymer matrix may be easy to reach, and it can lead to no significant increase in total flux. BABP moieties may do so, resulting in a larger n value. The stereostructures of DDS and BADS moieties provide more inter-chain space, and their membranes can accommodate the more sorbed penetrants. But similar to BABP moieties, the saturation in sorption may lead to larger n values for DDS and BADS moieties. As for 6FDA moieties, the combination of the favorable effect from imide rings and the unfavorable effect from -C[F.sub.3] groups contributes to the n value of 0.25.
Effects of the Operating Temperature. As shown in Fig. 4b, permeate water contents decrease at a higher temperature when the feed concentration is kept unchanged. 6FDA-60DA-4BABP has the best pervaporation selectivity in dehydration of isopropanol. The permeation activation energies were calculated for water permeation and dehydration of isopropanol, as listed in Table 4.
The moiety contribution factors were calculated from least square regression of the permeation activation energies for 6FDA-based polyimide membranes using the linear moiety contribution method, as expressed in Eq. 4, and results are listed in Table 5. Comparison is made in Fig. 5b between the activation energies from Arrhenius equation and those recalculated from the linear moiety contribution method.
With regard to the 6FDA and ODA moieties, activation energies for pure water permeation are lower than those for total flux, while the other moieties act in a completely different way, which may be an indication of the hydrophobicity of 6FDA and ODA moieties in the membranes. The negative contributions to activation energies from DDS and BADS moieties possibly benefit from their steric effects. The flexibility of MDA, BABP, and DABN moieties, as well as the favorable interactions with water, accounts for the large contribution factors to the activation energies.
Copolyimides were synthesized from 6FDA and diamine monomers with different stereostructures. The diamine monomers showed different reactivities in polycondensation and led to great difference in molecular weights of the copolyimides. Good solubility was observed, and the membranes were cast from their THF solutions. 6FDA-60DA-4DABN showed the best hydrophilicity among the copolyimide membranes from contact angles and surface free energies.
In gas permeation, by employing the linear moiety contribution method to the ideal separation factors, the contributions of the monomer moieties were obtained. Attributed to steric effects of the sulfonyl groups, DDS and BADS moieties had negative contributions to the selectivities of [O.sub.2]/[N.sub.2], [H.sub.2]/[N.sub.2], and He/[N.sub.2]. The DABN moiety was found to be favorable for improving the C[O.sub.2]/[N.sub.2] selectivity.
In pervaporation, the linear moiety contribution method was applied to the concentration coefficients and activation energies for the permeation flux. DDS and BADS moieties might influence the sorption properties when changing feed concentrations. The interactions between water and the moieties of BABP and DABN might contribute to the high activation energies.
1. H. Ohya, V.V. Kudryavtsev, and S.I. Semenova, Eds., Polyimide Membranes, Applications, Fabrications, and Properties, Gordon and Breach Publishers, Amsterdam (1996).
2. D. Wilson, H.D. Stenzenberger, and P.M. Hergenrother, Eds., Polyimides, Blackie & Son, Glasgow and London (1990).
3. J.Y. Park and D.R. Paul, J. Membr. Sci., 125, 23 (1997).
4. L.M. Robeson, C.D. Smith, and M. Langsam, J. Membr. Sci., 132, 33 (1997).
5. M. Salame, Polym. Eng. Sci., 26, 1543 (1986).
6. L. Jia and J. Xu, Polym. J., 23, 1 (1991).
7. T.H. Kim, W.J. Koros, G.R. Husk, and K.C. O'Brien, J. Membr. Sci., 37, 45 (1988).
8. S.A. Stern, Y. Mi, and H. Yamamoto, J. Polym. Sci. Part B: Polym. Phys., 27, 1887 (1989).
9. Y.-H. Kim, H.-S. Kim, and S.-K. Kwon, Macromolecules, 38, 7950 (2005).
10. Y. Xu, C. Chen, P. Zhang, B. Sun, and J. Li, J. Chem. Eng. Data, 51, 1841 (2006).
11. S. Xiao, R.Y.M. Huang, and X. Feng, Polymer, 48, 5355 (2007).
12. Y. Tanabe, K. Yamaguchi, Y. Yoshikawa, K. Sugimoto, and T. Yamaguchi, Jpn. Patent 61,221,158 (1986).
13. D.R. Heath and J.G. Wirth, DE Patent 2,202,167 (1972).
14. R.M. Silverstein, G.C. Bassler, and T.C. Morrill, Spectrometric Identification of Organic Compounds, Wiley-Interscience, New York (1997).
15. K. Nakanishi and P.H. Solomon, Infrared Absorption Spectroscopy, 2nd ed., Holden-Day, San Francisco (1977).
16. D.L. Dunson, Synthesis and Characterization of Thermosetting Polyimide Oligomers for Microelectronics Packaging, PhD dissertation, Virginia Polytechnic Institute and State University, Blacksburg, VA (2000).
17. K. Pramoda, S. Liu, and T.-S. Chung, Macromol. Mater. Eng., 287, 931 (2002).
18. C.E. Sroog, Prog. Polym. Sci., 16, 561 (1991).
19. K.W. Kim, S.C. Choi, S.S. Kim, S.J. Cho, and C.N. Whang, J. Mater. Sci., 28, 1537 (1993).
20. R.J. Good and C.J. van Oss, "The Modem Theory of Contact Angles and the Hydrogen Bond Components of Surface Energies," in Modern Approach to Wettability: Theory and Applications, M.E. Schrader and G. Loeb, Eds., Plenum, New York, Chapter 1 (1991).
21. L. de Bartolo, A. Gugliuzza, S. Morrelli, B. Cirillo, A. Gordano, and E. Drioli, J. Mater. Sci.: Mater. Med., 15, 877 (2004).
22. E. Oh and P.E. Luner, Int. J. Pharm., 188, 203 (1999).
23. C.J. van Oss, R.F. Giese, and W. Wu, J. Adhesion, 63, 71 (1997).
24. C.J. van Oss and R.F. Giese, Clays Clay Miner., 43, 474 (1995).
25. W.J. Koros and G.K. Fleming, J. Membr. Sci., 83, 1 (1993).
26. P. Pandey and R.S. Chauhan, Prog. Polym. Sci., 26, 853 (2001).
27. J.D. Perry, K. Nagai, and W.J. Koros, MRS Bull., 31, 745 (2006).
28. W.S.W. Ho and K.K. Sirkar, Membrane Handbook, Van Nostrand Reinhold, New York (1992).
29. P. Shao and R.Y.M. Huang, J. Membr. Sci., 287, 162 (2007).
30. R.Y.M. Huang and X. Feng, Sep. Sci. Technol., 27, 1583 (1992).
Shude Xiao, Xianshe Feng, Robert Y.M. Huang
Department of Chemical Engineering, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1
Correspondence to: R.Y.M. Huang; e-mail: email@example.com
Contract grant sponsor: Natural Sciences and Engineering Research Council (NSERC) of Canada.
TABLE 1. 6FDA-ODA copolyimides and molecular weights. Polyimides Monomers molar ratio [bar.M.sub.n] (a) 6FDA-8ODA-2DDS 6FDA:ODA:DDS = 1:0.8:0.2 44,510 6FDA-6ODA-4DDS 6FDA:ODA:DDS = 1:0.6:0.4 50,410 6FDA-8ODA-2MDA 6FDA:ODA:MDA = 1:0.8:0.2 54,320 6FDA-6ODA-4MDA 6FDA:ODA:MDA = 1:0.6:0.4 63,460 6FDA-8ODA-2BADS 6FDA:ODA:BADS = 1:0.8:0.2 1,33,900 6FDA-6ODA-4BADS 6FDA:ODA:BADS = 1:0.6:0.4 1,18,200 6FDA-8ODA-2BABP 6FDA:ODA:BABP = 1:0.8:0.2 1,64,300 6FDA-6ODA-4BABP 6FDA:ODA:BABP = 1:0.6:0.4 1,28,300 6FDA-8ODA-2DABN 6FDA:ODA:DABN = 1:0.8:0.2 46,990 6FDA-6ODA-4DABN 6FDA:ODA:DABN = 1:0.6:0.4 89,560 [bar.M.sub.w]/ Polyimides [bar.M.sub.w] (a) [bar.M.sub.n] (a) [bar.DP] 6FDA-8ODA-2DDS 58,270 1.31 72 6FDA-6ODA-4DDS 62,810 1.25 80 6FDA-8ODA-2MDA 74,930 1.38 89 6FDA-6ODA-4MDA 74,680 1.18 104 6FDA-8ODA-2BADS 1,71,500 1.28 205 6FDA-6ODA-4BADS 1,60,800 1.36 169 6FDA-8ODA-2BABP 2,75,100 1.67 254 6FDA-6ODA-4BABP 2,04,500 1.59 187 6FDA-8ODA-2DABN 62,280 1.33 74 6FDA-6ODA-4DABN 1,24,800 1.39 137 (a) Molecular weights and molecular weight distributions were obtained from GPC in THF at room temperature. TABLE 2. Contact angles of liquids on polyimide membranes from the sessile drop method at room temperature. Membranes Water ([degrees]) (a) Glycerol ([degrees]) (a) 6FDA-6ODA-4DDS 78.6 [+ or -] 0.9 66.0 [+ or -] 1.9 6FDA-6ODA-4MDA 75.9 [+ or -] 2.1 65.0 [+ or -] 0.9 6FDA-6ODA-4BADS 77.3 [+ or -] 1.1 65.5 [+ or -] 0.7 6FDA-6ODA-4BABP 82.9 [+ or -] 0.5 67.4 [+ or -] 0.6 6FDA-6ODA-4DABN 68.4 [+ or -] 2.1 61.2 [+ or -] 2.7 Diiodomethane Membranes Formamide ([degrees]) (a) ([degrees]) (a) 6FDA-6ODA-4DDS 53.8 [+ or -] 1.2 21.3 [+ or -] 1.2 6FDA-6ODA-4MDA 54.0 [+ or -] 0.8 28.0 [+ or -] 0.5 6FDA-6ODA-4BADS 59.9 [+ or -] 0.2 34.1 [+ or -] 0.8 6FDA-6ODA-4BABP 56.5 [+ or -] 1.9 26.7 [+ or -] 0.8 6FDA-6ODA-4DABN 45.8 [+ or -] 2.1 32.2 [+ or -] 0.5 (a) All results are given as means with 95% confidence intervals. TABLE 3. Surface free energy components (mJ/[m.sup.2]) and membrane- water interfacial free energies (mJ/[m.sup.2]) of polyimides. Membranes [[gamma].sub.s.sup.LW] [[gamma].sub.s.sup.+] 6FDA-6ODA-4DDS 47.39 0.0002 6FDA-6ODA-4MDA 45.03 0.0086 6FDA-6ODA-4BADS 42.44 0.0043 6FDA-6ODA-4BABP 45.53 0.0182 6FDA-6ODA-4DABN 43.29 0.1427 Membranes [[gamma].sub.s.sup.-] [[gamma].sub.s.sup.AB] 6FDA-6ODA-4DDS 5.09 0.06 6FDA-6ODA-4MDA 7.13 0.50 6FDA-6ODA-4BADS 7.38 0.35 6FDA-6ODA-4BABP 2.99 0.47 6FDA-6ODA-4DABN 11.54 2.57 Membranes [[gamma].sub.s] [DELTA][G.sub.sw.sup.IF] 6FDA-6ODA-4DDS 47.45 -66.07 6FDA-6ODA-4MDA 45.52 -55.51 6FDA-6ODA-4BADS 42.80 -53.34 6FDA-6ODA-4BABP 45.99 -73.91 6FDA-6ODA-4DABN 45.85 -38.20 All results were calculated from contact angles of water, glycerol, formamide, and diiodomethane on membranes at room temperature. TABLE 4. Gas separation properties at room temperature and parameters calculated for pervaporation. Permeance (GPU) (a) Membranes [J.sub.N.sub.2] [J.sub.O.sub.2] [J.sub.H.sub.2] 6FDA-8ODA-2DDS 0.027 0.098 0.473 6FDA-6ODA-4DDS 0.030 0.100 0.485 6FDA-8ODA-2MDA 0.021 0.084 0.467 6FDA-6ODA-4MDA 0.057 0.269 2.750 6FDA-8ODA-2BADS 0.013 0.050 0.315 6FDA-6ODA-4BADS 0.037 0.156 0.816 6FDA-8ODA-2BABP 0.028 0.100 0.517 6FDA-6ODA-4BABP 0.013 0.075 0.728 6FDA-8ODA-2DABN 0.094 0.496 3.113 6FDA-6ODA-4DABN 0.021 0.121 0.957 Permeance (GPU) (a) Membranes [J.sub.He] [J.sub.C[O.sub.2]] 6FDA-8ODA-2DDS 0.518 1.367 6FDA-6ODA-4DDS 0.545 1.410 6FDA-8ODA-2MDA 0.494 0.905 6FDA-6ODA-4MDA 3.116 2.001 6FDA-8ODA-2BADS 0.321 0.503 6FDA-6ODA-4BADS 0.975 1.505 6FDA-8ODA-2BABP 0.545 1.289 6FDA-6ODA-4BABP 0.948 0.533 6FDA-8ODA-2DABN 3.031 5.679 6FDA-6ODA-4DABN 1.285 1.183 Ideal separation factor [[alpha].sub.[O.sub.2]/ [[alpha].sub.[H.sub.2]/ Membranes [N.sub.2]] [N.sub.2]] 6FDA-8ODA-2DDS 3.6 17.4 6FDA-6ODA-4DDS 3.4 16.3 6FDA-8ODA-2MDA 4.1 22.7 6FDA-6ODA-4MDA 4.8 48.6 6FDA-8ODA-2BADS 4.0 25.1 6FDA-6ODA-4BADS 4.2 21.8 6FDA-8ODA-2BABP 3.6 18.7 6FDA-6ODA-4BABP 5.7 54.9 6FDA-8ODA-2DABN 5.3 33.3 6FDA-6ODA-4DABN 5.7 44.8 Ideal separation factor [[alpha].sub.He/ [[alpha].sub.C[O.sub.2]/ Membranes [N.sub.2]] [N.sub.2]] 6FDA-8ODA-2DDS 19.0 50.1 6FDA-6ODA-4DDS 18.3 47.4 6FDA-8ODA-2MDA 24.1 44.1 6FDA-6ODA-4MDA 55.0 35.3 6FDA-8ODA-2BADS 25.5 40.0 6FDA-6ODA-4BADS 26.0 40.2 6FDA-8ODA-2BABP 19.7 46.7 6FDA-6ODA-4BABP 71.5 40.2 6FDA-8ODA-2DABN 32.4 60.7 6FDA-6ODA-4DABN 60.1 55.3 [E.sub.p] [E.sub.p] (pure) (c) (total) (d) Membranes n ([r.sup.2]) (b) (kJ/mol) (kJ/mol) 6FDA-8ODA-2DDS 0.50 (0.99) 21.6 36.2 6FDA-6ODA-4DDS 0.60 (0.99) 21.6 34.0 6FDA-8ODA-2MDA 0.31 (0.98) 26.6 38.5 6FDA-6ODA-4MDA 0.41 (0.99) 30.0 48.7 6FDA-8ODA-2BADS 0.37 (0.97) 16.8 30.8 6FDA-6ODA-4BADS 0.44 (0.92) 19.5 38.7 6FDA-8ODA-2BABP 0.27 (0.99) 22.1 39.1 6FDA-6ODA-4BABP 0.41 (0.96) 30.0 49.1 6FDA-8ODA-2DABN 0.28 (0.96) 27.8 41.9 6FDA-6ODA-4DABN 0.30 (0.97) 30.3 43.5 (a) Average data of gas permeances at 0.14, 0.21, 0.28, and 0.35 MPa. 1GPU = 7.5 X [10.sup.-12] [m.sup.3] (STP) [m.sup.-2] [s.sup.-1] [Pa.sup.-1]. (b) Concentration coefficients were obtained from curve fitting. (c) Activation energies for pure water permeation. (d) Activation energies for total permeation flux with feed water content 20 wt%. TABLE 5. Contributions of monomer moieties to the ideal separation factors, concentration coefficients, and permeation activation energies. [[alpha].sub.[O.sub.2]/ [[alpha].sub.[H.sub.2]/ Moieties of monomers [N.sub.2]] [N.sub.2]] 6FDA 3.6 29.4 ODA 0.6 -0.4 MDA 1.7 9.7 DDS -1.7 -37.3 BADS 0.3 -18.6 BABP 2.9 41.2 DABN 4.6 35.6 [[alpha].sub.He/ [[alpha].sub.C[O.sub.2]/ Moieties of monomers [N.sub.2]] [N.sub.2]] n 6FDA 36.4 33.8 0.25 ODA -2.7 9.6 0.03 MDA 5.1 12.1 0.07 DDS -48.3 24.3 0.88 BADS -26.4 -0.2 0.43 BABP 58.8 6.5 0.27 DABN 48.7 50.7 0.06 [E.sub.p] (pure) (a) [E.sub.p] (total) (b) Moieties of monomers (kJ/mol) (kJ/mol) 6FDA 24.1 35.6 ODA -6.7 7.1 MDA 16.9 6.3 DDS 5.7 -17.0 BADS -3.3 -13.0 BABP 23.0 16.1 DABN 29.3 7.7 (a) Activation energies for pure water permeation. (b) Activation energies for total permeation flux with feed water content 20 wt%.