Printer Friendly

Preparation and characterization of symmetric and asymmetric pure polysulfone membranes for C[O.sub.2] and C[H.sub.4] separation.


Nowadays, there are different methods for gas separation such as absorption, adsorption, and cryogenic processes [1, 2]. Membrane separation technology is an energy efficient and cost effective application in gas separation process [3-6]. Gas separation is one of the necessary processes for many industries such as petrochemical complexes and refineries [7]. Separation of C[O.sub.2] from gas mixtures by membranes is more attractive than other gas separations. Some examples are: C[O.sub.2] capturing from flue gases for greenhouse gas emissions control (C[O.sub.2]/[N.sub.2] separation), separation of C[O.sub.2] from C[H.sub.4] (landfill gas recovery), natural gas upgrading, etc. [8]. There are numerous studies which show that, pure polymers were blended with other polymers (blend membrane) or mixed with inorganic nanoparticles (mixed matrix membranes) for preparation of high performance gas separation membranes [9-14]. Vankelecom et al. [15] reported that polysulfone (PSf)/PI blend membrane with a weight ratio of 3/1 increases membrane selectivity and membrane stability at high C[O.sub.2] feed pressure and high temperature. Sikander et al. [12] noticed that C[O.sub.2]/C[H.sub.4] selectivity was increased to 61.0 in 2 bar feed pressure by addition of silica nanoparticle up to 15.2 wt% in PSf/silica mixed matrix membrane. In contrast, pure PSf is a less used membrane in comparison with blend and mixed matrix membranes because of its low gas selectivity during separation of C[O.sub.2] from C[H.sub.4]. Ismail et al. [16] studied the influences of nonsolvent additives, organic additives, and concentration of polymer on the morphology

and gas separation properties of pure PSf membrane. According to their results, the best C[O.sub.2]/C[H.sub.4] selectivity was 12.45 when pure PSf (30 wt%) membrane was prepared using EtOH as internal nonsolvent and mixture of NMP and THF as solvent. Effect of varying solvent (NMP/DMAC) compositions (20/80, 50/50, and 80/20) on morphology and gas permeation properties of pure PSf membrane for C[O.sub.2] separation from natural gas was examined by Sikander et al. [17], According to their results, C[O.sub.2]/C[H.sub.4] selectivity increased by enhancement of NMP concentration in the casting solution and reached to 26 for NMP/ DMAC with the ratio of 80/20. In wet/wet phase inversion technique, the exchange velocity of external nonsolvent and solvent can be controlled easily by changing coagulation bath and casting solution compositions [18]. In this technique, membranes are formed by contacting casting solution with two nonsolvent baths in series. The first bath is employed to obtain a concentrated layer of polymer at the interface. This step makes ultra-thin surface layer similar to the evaporation step in dry/wet phase separation process [19, 20]. The second bath is responsible for the actual coagulation and formation of final film. The selection of nonsolvents for the coagulation bath depends on type of solvent in casting solution. Type of solvent in polymer solution also affects membrane structure and performance. For instance, addition of nonsolvent along with volatile solvent to the casting solution change the liquid-liquid phase separation process and change membrane permeability and selectivity simultaneously. On the other hand, employing of a mixture of solvents in the casting solution can remove large pore formation during instantaneous phase separation and can convert finger like to sponge like pores [16]. Two important structures in polymeric membranes are symmetric and asymmetric structures. Symmetrical membranes consist of a single polymer layer. They can be porous or nonporous. The most important factor for gas separation by symmetric membrane is its morphology that should be nonporous dense single polymer layer, which is homogeneous in all directions. Asymmetric membranes are made up of a dense skin layer with a porous support layer. Separation is generally controlled by the dense active layer [21]. Most studies on the gas transport properties of polymers are performed on dense homogeneous symmetric membranes. Dense homogeneous symmetric membranes tend to be significantly thicker than the selective layer of asymmetric membranes. This leads to lower gas fluxes for symmetric membranes in comparison with asymmetric membrane structures. Asymmetric membranes are used for reverse osmosis, ultra-filtration and gas separation processes because of their high permeability, selectivity, and mechanical strength in high pressure application [21]. In this study, as the thickness of top layers for prepared asymmetric membranes using different nonsolvents were too thin, gas selectivity of prepared symmetric membrane was higher than asymmetric membrane. The novelty of this paper is to prepare pure PSf membrane with high C[O.sub.2]/C[H.sub.4] selectivity without blending with other polymers or the addition of inorganic nanoparticles into the polymer solution. This article explains how different types and concentrations of external nonsolvent in coagulation bath can change membrane structure from asymmetric to symmetric. Also, the effects of three different types of solvent such as: NMP, THF, and a mixture of NMP/THF with a ratio of 80/20 on morphology and performance of PSf membrane were examined. Furthermore, presence of BuOH as internal nonsolvent in casting solution and its effects on structure and gas permeability of prepared membranes were studied.



PSf with the average [M.sub.n] = 22,000 was used as a base polymer and purchased from Sigma-Aldrich, USA. The organic solvents such as l-methyl-2-pyrrolidone (NMP), tetrahydrofuran (THF) and different nonsolvents such as ethanol (EtOH), isopropyl alcohol (IPA), and butanol (BuOH) were supplied from Merck (Germany). C[O.sub.2] and C[H.sub.4] gases were provided in 40-L cylinders with a purity of 99.99%. Distilled water was used as a second coagulation bath. The PSf resin was dried in an oven at 80[degrees]C for 24 h before usage. Other chemicals were consumed as received.

Preparation of PSf Membrane

In this study, flat sheet PSf membranes were fabricated by wet/wet phase inversion technique. For this reason, considered amount of PSf resin was dissolved in different solvents such as NMP, THF, and NMP/THF with a ratio of 80/20 separately to prepare casting solution. Membranes with various compositions were labeled from PSf-1 to PSf-6 (Table 1). Concentration of PSf in casting solution was kept constant at 22 wt% maintaining concentration of solvent in 78 wt%. Then, casting solution was stirred for at least 24 h. After preparing the homogeneous solution, this solution was kept at room temperature for 24 h for the removal of air bubbles and then cast on a smooth glass plate by film casting knife with a thickness of 350 pm. The color of wet film changed from transparent to white gradually after immersion in first coagulation bath containing two different alcohols (EtOH and IPA), which were used separately. Immersion time in first coagulation bath was 90 s and after replacement of solvent by nonsolvent, fabricated membrane dipped in distillate water for 24 h as a second coagulation bath. Finally, prepared membrane was dried at room temperature condition for one day.

Scanning Electron Microscopy

Membrane structure and morphology were examined by scanning electron microscopy (SEM). First, specimens were fractured cryogenically in liquid nitrogen to leave an undeformed structure and then mounted on sample stubs with double-surface scotch tape. Next, samples were deposited with gold using a sputter coater. Finally, they were imaged and photographed through a scanning electron microscope (LEO 1455 VPSEM) with potentials of 5.0 kV.

Porosity Determination

The porosity of produced membranes (e) was calculated by Eq. 1. Where is the [V.sub.void] volume, [V.sub.tot], the total specimen volume, [V.sub.pol] is polymer volume, [rho] is polymer density, L is thickness, A and m are effective area and weight of membrane, respectively. Thickness of sample was determined directly from SEM [22]:

[[epsilon] = [V.sub.void]/[V.sub.tot] = ([V.sub.tot] - [V.sub.pol])/[V.sub.tot] = [(LxA) - [(m/[rho]).sub.pol])/(LxA) (1)

Gas Permeation Evaluation

Permeance of pure C[O.sub.2] and C[H.sub.4] gases were measured using gas permeation test at the pressure range of 1-7 bar. Gas permeation mechanism in polymeric membrane is a solution-diffusion. Gases with larger molecular diameter diffuse slower through structure of prepared membrane and membrane act as molecular sieve [10, 23, 24], Accordingly, as C[H.sub.4] molecular diameter is larger than C[O.sub.2] [24], symmetric and asymmetric PSf membranes are able to separate these two gases by different selectivities. Stainless steel filter holder that was equipped with a back-pressure support screen with effective area of 13.8 [cm.sup.2] (Merck) was used for gas permeation examination. Glass soap bubble flow meter was employed for measuring rate of permeate stream. To ensure accuracy in our experiments, the gas permeation test was repeated twice in the steady state. The gas permeability (P / L) was calculated using the following equation:

(P/L) = Q/(A x [DELTA]P) (2)

where P is permeability, L is membrane top layer thickness, Q is gas flow (at standard pressure and temperature), A is the effective membrane area in [cm.sup.2], and [DELTA]P is the differential partial pressure across the membrane. The usual unit of permeance is GPU and 1 GPU is equal to 1 x [10.sup.-6] [cm.sup.3] (STP)/[cm.sup.2] s cmHg. Equation 3 can be used to calculate C[O.sub.2]/C[H.sub.4] selectivity ([alpha]), where [P.sub.i] and [P.sub.j] are C[O.sub.2] and C[H.sub.4] permeances respectively:

[alpha]=[P.sub.i]/[P.sub.j] (3)


Coagulation Bath Composition

The dependency of membrane morphology on the type of external nonsolvent (coagulation bath) for PSf-1/NMP/ nonsolvent system was considered as a first examination. The solubility parameter of pure solvents and nonsolvents are shown in Table 2. For determination of nonsolvents bath solubility parameter we multiply nonsolvents solubility parameters by nonsolvents moles [20]. As presented in Table 3, solubility parameter difference between NMP and EtOH-50% is higher than this number between NMP and EtOH-100%. When solubility parameter difference between solvent and nonsolvent is high, nonsolvent is able to diffuse more easily into the polymer film. This results in an increment in the exchange rate between solvent in the polymer film and nonsolvent in the coagulation bath (instantaneous phase demixing), which is normally accompanied with a porous membrane structure, higher gas permeability and lower gas selectivity [20, 25-28], Then, instantaneous phase demixing between solvent and nonsolvent has occurred using EtOH-50% as external nonsolvent. Therefore, a membrane containing large pores and higher porosity was produced. In contrast, less solubility parameter differences between NMP and IPA-100% [[DELTA][[delta].sub.(N-S)] = 0.9 [(MJ [cm.sup.-3]).sup.1/2]] in comparison with NMP and IPA-50% [[DELTA][[delta].sub.(N-S)] = 12.95 [(MJ [cm.sup.-3]).sup.1/2]], causes delayed phase demixing between solvent and nonsolvent. As a result, a membrane with a high symmetric structure containing sponge like pores and also lower porosity structure ([epsilon] = 8.36) was produced using pure IPA as coagulation bath, which is usually accompanied with lower gas permeability and higher gas selectivity. A comparison of membrane porosity for prepared PSf-1 membranes with various nonsolvents is presented in Table 3. As shown in this table, by increasing alcohol concentration as nonsolvent in coagulation bath, membrane porosity was reduced. Also, the variation in morphology of prepared membranes with different coagulation bath was studied by SEM micrographs. As demonstrated in Fig. 1, prepared membrane using EtOH-50% as external nonsolvent has the highest asymmetric structure including macro-voids in the substructure and the largest porosity. However, membrane asymmetry declined and macro-void structure converted to spongy structure using pure EtOH as external nonsolvent. In another word, less porous structure containing spongy pores and small tear like pores close to surface layer was obtained by using pure EtOH as nonsolvent in the coagulation bath (Fig. 1). The decrease in membrane porosity may be justified because of direct consequence of the macro-void reduction [17], Furthermore, asymmetric membrane structure containing large pores and also tear like pores close to surface layer was obtained employing IPA-50% as an external nonsolvent. Interestingly, a high symmetric membrane structure with the lowest membrane porosity among prepared membranes (Table 3) containing spongy pores and macro-void free structure was obtained using pure IPA in coagulation bath (Fig. 1). The variations in C[O.sub.2] and C[H.sub.4] permeances in different EtOH and IPA concentrations in coagulation bath are shown in Fig. 2. As cleared, a reduction in permeance of gases was observed by increasing alcohol concentration. Moreover, membrane that was fabricated using pure IPA in coagulation bath had lower C[O.sub.2] and C[H.sub.4] permeabilities in comparison with membrane prepared by pure EtOH. Influence of type and concentration of alcohol as external nonsolvent on the C[O.sub.2]/C[H.sub.4] selectivity of prepared PSf-1 membranes was shown in Table 4. According to the type and concentration of nonsolvents used, C[O.sub.2]/C[H.sub.4] selectivity of prepared membranes follows this order: IPA-100% > EtOH-100% > IPA-50% > EtOH-50%. Gas selectivity improved using pure alcohols and reached to a maximum of 36.40 at 1 bar feed pressure for pure IPA as external nonsolvent. Solubility parameter differences between NMP as solvent and EtOH and IPA [50% and 100% (v/v)] as nonsolvents can be applied for justification of membrane structure and performance variations. There are many studies that prove that even the small amount of difference between solvent and nonsolvent solubility parameters among various solvent/nonsolvent pairs causes a distinct influence on morphology and performance of prepared membranes [17, 23, 29]. For example, Iqbal et al. [29] studied the effect of different nonsolvents used in coagulation bath on morphology and performance of PC membrane using DMAC as solvent. According to their result, solubility parameter differences between DMAC with EtOH, PrOH, and BuOH are 16.24, 16.56, and 16.75, respectively. Although there are small solubility parameter differences among various DMAC/nonsolvents pairs, fabricated membranes have noticeable differences in morphology and gas separation properties. C[O.sub.2] and C[H.sub.4] permeation mechanism through fabricated membranes is a solution-diffusion. Then, C[H.sub.4] with larger molecular diameter diffuse slower through structure of prepared membranes. Therefore, membranes act as molecular sieve [10, 23, 24], Pure IPA as nonsolvent may causes membrane structure containing high quantity of pores with smaller diameter than C[H.sub.4] molecules. Consequently, C[O.sub.2]/C[H.sub.4] selectivity for prepared PSf-1 membrane using IPA-100% in coagulation bath is higher than other nonsolvents. In this paper, C[O.sub.2]/C[H.sub.4] selectivity of PSf-1 membrane prepared by IPA-100% as external nonsolvent is compared with some of the available research works in Table 5. The selection is based on the polymer similarity to this research.



Composition of Casting Solution

In order to study the effect of solvent on morphology and performance of pure PSf membrane, three different membranes (PSf-1, PSf-2, and PSf 3) were fabricated. As shown in Fig. 1c, PSf-1 which was fabricated by casting solutions containing NMP as solvent and IPA-50% as external nonsolvent had an asymmetric structure containing macro-voids and large pores. Whereas, an asymmetric membrane (PSf-2) with a thick top layer and very dense support layer was obtained (Fig. 3a) using a casting solution including pure THF. When NMP/THF was used as solvent in polymer solution (Fig. 3b), asymmetric membrane (PSf-3) with lower porosity and smaller voids in the sub-layer in comparison with PSf-1 (Fig. 1c) were produced. Variations in C[O.sub.2] and C[H.sub.4] permeabilities of PSf-2 and PSf-3 membranes were shown in Fig. 4. As shown in Fig. 4a, PSf-2 membrane which was fabricated by pure THF as solvent in the casting solution did not have gas permeability at different feed pressures. This phenomenon can be justified by the formation of a thick top layer for PSf-2 (Fig. 3) that did not allow the gas molecules to pass through the membrane structure. Whereas, PSf-3 which was prepared by NMP/THF as solvent had gas permeability and amount of its gas permeability was lower than PSf-1 membrane (Fig. 2c). In contrast, as indicated in Table 6, an improvement in C[O.sub.2]/ C[H.sub.4] selectivity of produced membranes was obtained by using a mixture of NMP/THF (PSf-3) instead of NMP (PSf-1) as solvent. Therefore, C[O.sub.2]/C[H.sub.4] selectivity increased and reached to 18.50 in 1 bar feed pressure in PSf-3. Different morphologies and performances of prepared membranes using different solvents in the casting solution developed because of variation in demixing rate of polymer and solvent [20, 26, 29], It appears from Table 7 that each solvent used in casting solution has various solubility parameter differences with PSf ([DELTA][[delta].sub.(S-PSF)]). With regard to the solvents used, solubility parameter difference between solvent and polymer increased in the following order: THF < NMP/THF < NMP. Theoretically, the lower solubility parameter difference between solvent and polymer [DELTA][[delta].sub.(S-PSF)], the more time is required to eliminate solvent from polymer structure. Consequently, delayed phase demixing between solvent and nonsolvent in coagulation bath has occurred. As known, delayed phase demixing has a key role in preparation of membrane with lower porosity, lower gas permeability, and also higher gas selectivity [26, 29, 30].



Internal Nonsolvent in the Casting Solution

In this study, in order to find the effects of internal nonsolvent on the morphology and gas permeability of prepared membranes, four samples (PSf-3, PSf-4, PSf-5, and PSf-6) including different amounts of BuOH (0, 2, 6, and 10 wt%) in the casting solution were examined while IPA-50% was employed as external nonsolvent. It appears from cross-section photographs taken by SEM, different concentrations of BuOH in the casting solution cause different membrane morphologies in terms of membrane thickness and porosity (Figs. 3b and 5). According to SEM analysis, membrane thickness of PSf-3, PSf-4, PSf-5, and PSf-6 are 124.8, 123.4, 97.59, and 72.11 [micro]m respectively. It shows that the higher concentration of BuOH in casting solution, the lower membrane thickness. Then, C[O.sub.2] and C[H.sub.4] permeances increased significantly for casting solution including higher concentration of BuOH (Fig. 6). The same observation was reported by Iqbal et al. [29] by the addition of BuOH as an internal nonsolvent to the polycarbonate solution. Solubility parameter differences between solvent and PSf ([DELTA][[delta].sub.(S-PSf)]) for PSf-3, PSf-4, PSf-5, and PSf-6 are 1.78, 2.35, 3.50, and 4.65 [(MJ [cm.sup.-3]).sup.1/2] respectively. Higher BuOH concentration causes higher solubility parameter differences between solvent and PSf. Hence, lower time is required to eliminate solvent from polymer and then instantaneous phase demixing has occurred between solvent and nonsolvent by immersing the wet film in the coagulation bath. Finally, membranes with lower thicknesses (Fig. 5) and higher gas permeances were produced by the addition of BuOH to the casting solution (Fig. 6).




Pure PSf membranes with different coagulation bath and casting solution compositions were prepared for separation of C[O.sub.2] from C[H.sub.4]. Solubility parameter differences between NMP as solvent and alcohols as external nonsolvents are 14.05, 3.10, 12.95, and 0.90 [(MJ/[cm.sup.3]).sup.1/2] for EtOH-50%, EtOH-100%, IPA-50%, and IPA-100%. As lower solubility parameter difference between solvent and nonsolvent results in delayed phase demixing, PSf-1 which was prepared by pure IPA as external nonsolvent had a less porous structure and high C[O.sub.2]/C[H.sub.4] selectivity of 36.40 in 1 bar feed pressure. PSf-1, PSf-2, and PSf-3 which were prepared by different solvents (NMP, THF, and NMP/THF respectively) and same external nonsolvent (IPA-50%) had different morphologies and gas separation properties. C[H.sub.4] permeance decreased from 12.95 to 0.85 GPU at 1 bar feed pressure using a mixture of NMP/ THF instead of NMP as solvent in the casting solution. In contrast, C[O.sub.2]/C[H.sub.4] selectivity increased from 7.10 for PSf-1 to 18.5 for PSf-3 at 1 bar feed pressure. PSf-2 which was prepared by THF as solvent was not able to have gas permeability at different feed pressures. Low solubility parameter difference between PSf and THF results in delayed phase demixing and caused a membrane with a thick top layer without gas permeability. A reduction in membrane thickness was observed by the addition of BuOH as internal nonsolvent to the casting solution. Wheras, C[O.sub.2] and C[H.sub.4] permeances increased from 15.70 to 44.33 GPU and 0.85 to 34.50 GPU, respectively, at 1 bar feed pressure. Instantaneous phase demixing between solvent and nonsolvent in higher BuOH concentrations results in membrane with lower thickness and higher gas permeability.


[1.] A. Ebadi, H. Sanaeepur, A. Kargari, and A. Moghadassi, Sep. Purif Technol., 82, 102 (2012).

[2.] S.S. Madaeni, M.M.S. Badieh, and V. Vatanpour, Polym. Eng. Sci., 10.1002/pen.23456, 1 (2013).

[3.] D. Bastani, N. Esmaeili, and M. Asadollahi, J. Ind. Eng. Chem., 19, 375 (2013).

[4.] M. Takhtravanchi, T. Kaghazchi, and A. Kargari, Desalination., 235, 199 (2009).

[5.] P. Ji, Y. Cao, X. Jie, M. Li, and Q. Yuan, Sep. Purif. Technol., 71, 160 (2010).

[6.] Y. Li and T.S. Chung, J. Membr. Sci., 350, 226 (2010).

[7.] A. Arabi, A. Kargari, M. Bahrami, S. Laki, and H. Ajami, J. Ind. Eng. Cltem., 19, 677 (2013).

[8.] H. Sanaeepur, A. Ebadi, A. Moghadassi, and A. Kargari, Sep. Purif. Technol., 80, 499 (2011).

[9.] A. Jomekian, M. Pakizeh, A. Reza, S. Ali, and A. Mansoori, Sep. Purif. Technol., 80, 556 (2011).

[10.] F. Dorosti, M.R. Omidkhah, M.Z. Pedram, and F. Moghadam, Chem. Eng. J., 171, 1469 (2011).

[11.] C.P. Leo, W.P.C. Lee, A.L. Ahmad, and A.W. Mohammad, Sep. Purif. Technol., 89, 51 (2012).

[12.] S. Rafiq, Z. Man, A. Maulud, N. Muhammad, and S. Maitra, Sep. Purif. Technol., 90, 162 (2012).

[13.] M. Esmaeili, S.S. Madaeni, and J. Barzin, J. Appl. Polym. Sci., 10.1002/APP.38749, 1 (2012).

[14.] S.S. Madaeni, M.M.S. Badieh, V. Vatanpour, and N. Ghaemi, Polym. Eng. Sci., 10.1002/pen.23223, 2664 (2012).

[15.] S. Basu, A. Cano-Odena, and J.I.F. Vankelecom, Sep. Purif. Technol., 75, 15 (2010).

[16.] M.A. Aroon, A.F. Ismail, M.M. Montazer-Rahmati, and T. Matsuura, Sep. Purif. Technol., 72, 194 (2010).

[17.] S. Rafiq, Z. Man, A. Maulud, N. Muhammad, and S. Maitra, J. Membr. Sci., 378, 444 (2011).

[18.] J. Carolus, M. Giovanna, A. Figoli, and E. Drioli, J. Membr. Sci., 272, 188 (2006).

[19.] S.A. Mousavi, M. Sadeghi, M.M.Y. Motamed-Hashemi, M. Pourafshari Chenar, R. Roosta-Azad, and M. Sadeghi, Sep. Purif. Techno!., 62, 642 (2008).

[20.] S.S. Madaeni and P. Moradi, J. Appl. Polym. Sci., 121, 2157 (2011).

[21.] C.E. Powell and G.G. Qiao, J. Membr. Sci., 279, 1 (2006).

[22.] J. Jansen, M. Macchione, and E. Drioli, J. Membr. Sci., 255, 167 (2005).

[23.] S.S. Madaeni, E. Enayati, and V. Vatanpour, J. Appl. Polym. Sci., 122, 827 (2011).

[24.] C.A. Scholes, S.E. Kentish, and G.W. Stevens, Recent. Pat. Cliem. Eng., 1, 52 (2008).

[25.] C. Torras, F. Ferrando, J. Paltakari, and R. Garciavalls, J. Membr. Sci., 282, 149 (2006).

[26.] S.S. Madaeni, A. Farhadian, and V. Vatanpour, Polymer, 00, 1 (2011).

[27.] I. Soroko, M.P. Lopes, and A. Livingston, J. Membr. Sci., 381, 152 (2011).

[28.] N. Riyasudheen, and A. Sujith, Desalination., 294, 17 (2012).

[29.] M. Iqbal, Z. Man, H. Mukhtar, and B.K. Dutta, J. Membr. Sci., 318, 167 (2008).

[30.] P. Moradihamedani, N.A. Ibrahim, W.Z.W. Yunus, and N.A. Yusof, J. Appl. Polym. Sci., 10.1002/APP.39288, 1 (2013).

[31.] L.A. Utracki, Polymer, 42, 2909 (2004).

Pourya Moradihamedani, (1) Nor Azowa Ibrahim, (1) Wan Md Zin Wan Yunus, (1) Nor Azah Yusof (2)

(1) Department of Chemistry, Faculty of Science, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia

(2) Department of Defence Science, Faculty of Defence Science and Technology, National Defence University of Malaysia, Sungai Besi Camp, 57000 Kuala Lumpur, Malaysia

Correspondence to: Nor Azowa Ibrahim; e-mail: norazowa@

DOI 10.1002/pen.23706

Published online in Wiley Online Library (
TABLE 1. Composition of casting solutions.

                   Solvent (wt%)       Internal
Membrane                              nonsolvent
(PSF 22 wt%)    NMP   THF   NMP/THF   (wt%) BuOH

PSf-1           78    --      --          --
PSf-2           --    78      --          --
PSf-3           --    --      78          --
PSf-4           --    --      76           2
PSf-5           --    --      72           6
PSf-6           --    --      68          10

TABLE 2. Solubility parameters for different solvents and
nonsolvents [20].

Solvent and nonsolvent      Solubility parameter
                         [(MJ [cm.sup.-3]).sup.1/2]

NMP                                22.90
THF                                18.60
EtOH                               26.00
IPA                                23.80
Distilled water                    47.90

TABLE 3. Effects of concentration and type of nonsolvent on
membrane porosity and solubility parameter differences between
nonsolvent (N) and solvent (NMP).

           Nonsolvent   [delta][[delta].sub.(N-NMP)]   porosity
Membrane     (v/v%)     (MJ/[cm.sup.3]).sup.1/2] (%)   [epsilon]

PSf-1      EtOH-50%                14.05                 26.76
PSf-1      EtOH-100%                3.10                 11.90
PSf-1      IPA-50%                 12.95                 18.34
PSf-1      IPA-100%                 0.90                  8.36

TABLE 4. Effects of concentration and type of nonsolvents on
C[O.sub.2]/C[H.sub.4] selectivity.

                        Feed pressure (bar)

Membrane   Nonsolvent     1       2       3       4

C[O.sub.2]/C[H.sub.4] selectivity ([+ or -] 0.10)
PSf-1       EtOH-50%     4.62    3.51    3.20    2.90
PSf-1      EtOH-100%    12.20    9.80    7.10    6.00
PSf-1       IPA-50%      7.10    6.23    4.46    3.64
PSf-1       IPA-100%    36.40   28.10   19.40   12.30

           Feed pressure (bar)

Membrane     5       6      7

C[O.sub.2]/C[H.sub.4] selectivity ([+ or -] 0.10)
PSf-1       2.45    1.68   1.53
PSf-1       4.23    3.50   2.40
PSf-1       2.91    2.08   1.78
PSf-1      10.25   10.98   9.50

TABLE 5. C[O.sub.2]/C[H.sub.4] selectivity comparison of the present
research work with previous studies.

Sample       C[O.sub.2]     C[H.sub.4]    C[O.sub.2]/C[H.sub.4]
             permeance      permeance          selectivity

This work   15.20 (GPU)     0.41 (GPU)            36.40
PSf [291    23.00 (GPU)     4.00 (GPU)             5.75
PSf [16]     4.11 (GPU)     0.33 (GPU)            12.45
PSf [17]    17.40 (GPU)     0.89 (GPU)            19.55

TABLE 6. Effect of solvent type on C[O.sub.2]/C[H.sub.4] selectivity

                      Feed pressure (bar)

Membrane  Nonsolvent    1       2       3      4      5      6      7

C[O.sub.2]/C[H.sub.4] selectivity ([+ or -] 0.10)
PSf-1     IPA-50%      7.10    6.23    4.46   3.64   2.91   2.08   1.78
PSf-3     IPA-50%     18.50   12.40    9.71   6.80   4.40   3.50   2.72

TABLE 7. Difference of solvent solubility parameter with PSf [20, 31].

Components     Solubility parameter          Difference with PSf
             [(MJ/[cm.sup.3]).sup.1/2]   [DELTA][[delta].sub.(S-PSF)

THF                    18.60                        1.66
NMP/THD                22.04                        1.78
NMP                    22.90                        2.64
PSf                    20.26                         --
COPYRIGHT 2014 Society of Plastics Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2014 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Moradihamedani, Pourya; Ibrahim, Nor Azowa; Yunus, Wan Md Zin Wan; Yusof, Nor Azah
Publication:Polymer Engineering and Science
Article Type:Report
Geographic Code:9MALA
Date:Jul 1, 2014
Previous Article:One-pot synthesis of hyperbranched poly(aryl ether ketone)s for the modification of bismaleimide resins.
Next Article:Synthesis and properties of poly(aryl ether ketone)-based phthalonitrile resins.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters