Journal of Membrane Science xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci
Cardo-type random co-polyimide membranes for high pressure pure and mixed sour gas feed separations ⁎
Garba O. Yahaya , Ilham Mokhtari, Afnan A. Alghannam, Seung-Hak Choi, Husnul Maab, Ahmad A. Bahamdan Research & Development Center, P.O. Box 62, Saudi Aramco, Dhahran 31311, Saudi Arabia
A R T I C L E I N F O
A B S T R A C T
Keywords: Random co-polyimide 6FDA-Durene-CARDO Sour gas separation Permeability coefficient Selectivity coefficient
A series of aromatic random co-polyimide membranes based on 6FDA-Durene/CARDO backbone with varying content of CARDO moiety (3:1; 1:1 and 1:3) were synthesized for enhanced acid gas separation performance. Gas transport properties of pure and mixed gas streams consisting of H2S, CO2, He, CH4, N2 and C2H6 through the dense films of the co-polyimide were studied. Mixed sweet gas tests were done with feeds containing no H2S up to 55 bar, while mixed sour gas measurements were conducted with feeds containing H2S up to 34 bar for 10% H2S and 20% H2S. The membranes exhibit very attractive pure gas transport properties, as CO2 permeability and CO2/CH4 selectivity are up to 323 barrer and 35 respectively. Furthermore, the transport properties of sour gas mixture consisting of five gases were also found attractive, as the CO2/CH4 and H2S/CH4 ideal selectivities are in the range of 18–23 and 19–21 respectively; while CO2 and H2S permeabilities are in the range of 38–51 and 40–47 barrers respectively for 20% H2S in the gas mixture. These values and separation performance exhibited by the co-polyimide are comparable and very competitive even, as compared to the values obtained in some of the high performance polymeric membranes that have been reported in the literature. The stability of the random co-polyimide 6FDA-Durene/CARDO membrane under these aggressive environments is quite remarkable.
1. Introduction For many decades, natural gas (NG) has been a popular energy source and its demand as an efficient fuel is continuously increasing worldwide [1]. Most of the gas reserves around the world are of lowquality with high contents of impurities, which include acid gas (carbon dioxide (CO2) and hydrogen sulfide (H2S)), water, heavy hydrocarbons (C3+) and other contaminants such as helium, nitrogen, mercaptans etc. For NG to meet the sales gas specification, these contaminants especially acid gas that constitute the largest amount in many existing NG reserves have to be removed. One of the major requirement for NG sweetening process involve the separation of acid gases from NG and this treatment is very important in order to prevent corrosion of transportation pipeline, reduce atmospheric pollution, and avoid other detrimental effects [2]. The removal of acid gases from NG streams require very efficient separation technologies, and NG sweetening process currently constitutes a major industrial gas separation processes. Membrane-based NG separations and membrane-absorption hybrid processes have emerged as among the fastest growing technologies, due to their lower
⁎
capital cost, higher energy savings, smaller size, environmental friendly and more economically viable as compared to conventional technologies such as stand-alone pressure swing adsorption (PSA) and standalone absorption process [3,6,7]. However, the inadequate performance of the current existing polymeric membranes impedes the full utilization of the application opportunities on the industrial scale [8–10]. Some of the challenges include inability to achieve both high permeability and selectivity, selectivity-permeability trade-off, membrane plasticization and physical aging. These certainly inhibit long-term gas separation performance and membrane stability. Thus, polymeric membrane materials with high permeation properties (i.e., both high permeability and selectivity) are indispensable to the viability of membrane-based NG separations and membrane-absorption hybrid processes. While many membranes have been developed for the separation of CO2/CH4 over the past several decades, only few studies have considered the development of membranes for simultaneous separation of CO2 and H2S at moderate to high concentration of H2S and feed pressure [3–5,11]. Such studies include the development of 6FDA-DAM: DABA (3:2) for simultaneous removal of CO2 and H2S from sour natural
Corresponding author. E-mail address:
[email protected] (G.O. Yahaya).
http://dx.doi.org/10.1016/j.memsci.2017.10.063 Received 22 August 2017; Received in revised form 11 October 2017; Accepted 29 October 2017 0376-7388/ © 2017 Elsevier B.V. All rights reserved.
Please cite this article as: Yahaya, G.O., Journal of Membrane Science (2017), http://dx.doi.org/10.1016/j.memsci.2017.10.063
Journal of Membrane Science xxx (xxxx) xxx–xxx
G.O. Yahaya et al.
Nomenclature Ji l pi0 pil Di Si
Po So
volumetric flux of component i expressed as (cm3(STP)/ cm2 s) membrane thickness (cm) partial pressure of component i on the feed side (cm Hg) partial pressure of component i on the permeate side (cm Hg) diffusion coefficient of component i (cm2/s) solubility coefficient of component i (cm3(STP) of
Do R T EP ΔHs ED
penetrant gas/cm3 of polymer per pressure) permeability coefficient pre-exponential factor (Barrer) solubility coefficient pre-exponential factor (cm3(STP)/ cm3 cmHg) diffusion coefficient pre-exponential factor (cm2/s) universal gas constant (0.278 cm3 cm Hg/(cm3(STP) K)), absolute temperature (K) activation energy of permeation (kJ/mol) enthalpy of sorption or heat of solution (kJ/mol) activation energy of diffusion (kJ/mol)
certain type of homo-polyimides result in membranes with very attractive and improved gas separation performance than what is usually obtainable with homo-polyimides. These polymers can be simply prepared from existing, commercially available materials. Even though the 6FDA-based co-polyimides (i.e., containing 6FDA, mPDA and durene moieties) show high pure gas CO2/CH4 selectivity of up to 61 and attractive mixed sour gas H2S/CH4 and CO2/CH4 selectivities of about 23 and 27 respectively, however, the pure and mixed gas permeabilities are very low (i.e., pure gas CO2 permeability of 40 barrer and mixed sour gas CO2 and H2S permeabilities of 13 and 11 barrers respectively) [22,23]. Other types of 6FDA-based co-polyimide previously reported [3–5] showed very attractive pure and mixed gas permeabilities and very good H2S/CH4 and CO2/CH4 selectivities. However, in order to enhance the separation performance of the polymeric membranes even further and optimize these materials for better gas separation, chemical modifications that include substitution of other pertinent moieties or monomers and bulky functional groups in the co-polyimides backbone is necessary. This can be achieved by substituting bulky CARDO moiety with mPDA in the existing copolymers. This CARDO functionality, which is sometime called hinge or loop in Latin [35] and polymers containing this loop or hinge shaped functionalities in the polymer backbone are referred to as CARDO polymers [35]. These polymers exhibit high solubility and high thermal stability and because of its stiffness and bulky nature, CARDO moiety hinders the packing and restrict the rotational mobility of the polymer backbones. In view of these exceptional properties, CARDO-based polyimides are expected to have potential for enhancing gas separation performance. Several studies have reported attractive CO2 separation properties of some CARDO-based polymers [35–45], and it is pertinent to believe that substituting CARDO moiety with mPDA in the existing copolymers could significantly improve the gas permeation performance of the copolyimides. One of such studies involved the development of CARDOcopolybenzoxazole by thermal induced structural rearrangement, which resulted to an increment of 3 times in CO2 permeability as compared to the CO2 permeability in the non-cardo counterparts [43,44]. In addition, commercially available polyphenylene oxide (PPO) and CARDO-type polyimide membranes have also been investigated for gas separation, and it was observed that PPO exhibited high selectivity and moderate permeance, while the CARDO-type polyimide membrane showed moderate selectivity and high permeability for CO2/CH4 and CO2/N2 separations [39,41,42]. In light of this, the preparation of series of aromatic CARDO-type random co-polyimides containing three homo-polyimides (i.e., 6FDA, mPDA and CARDO moieties) was investigated in this work. The effects of CARDO substitution and varying segmental length of the moiety in the polymer backbone on the physical and gas transport properties were investigated via pertinent pure and gas mixture permeation measurements. Detailed studies that include preparation, pure and sour gas mixture transport properties have been conducted on the co-polyimide (6FDA–mPDA/durene) - containing no CARDO in our previous work [22,23]. The aim of this study is to achieve enhancement in the gas separation performance and transport properties and thus, in this work, gas transport properties of the CARDO type random co-polyimides were
gas streams. Under mixed gas feed conditions with 20%H2S, 20%CO2 and 60% CH4 at 35 °C and up to 62 bar, H2S/CH4 and CO2/CH4 selectivities of above 22 and 27 respectively; and H2S and CO2 permeabilities of about 40 and 60 respectively were observed [3,4]. Yi et al. also investigated the sour gas permeation properties of hydroxyl-functionalized polymer of intrinsic microporosity with H2S/CH4 and CO2/ CH4 selectivities of about 30 and 25 respectively; and H2S and CO2 permeabilities of about 60 and 50 respectively for a mixed gas feed conditions of 15%H2S, 15%CO2 and 70% CH4 at 35 °C and up to 48 bar [5]. Furthermore, Achoundong et al. studied the acid gas permeation behavior of modified cellulose acetate membranes with H2S/CH4 and CO2/CH4 selectivities of about 27.5 and 20 respectively and; H2S and CO2 permeabilities of about 190 and 136 respectively for a mixed gas feed conditions of 20%H2S, 20%CO2 and 60% CH4 at 35 °C and up to 48 bar [11]. Aside from the these few studies that have focused on moderate to high concentration of H2S and feed pressure, most other previous studies have generally focused on low concentrations of H2S and low feed pressure due to hazardous nature of H2S [8–10]. Even though, these studies, which are mainly focused on rubbery membranes, have shown good membrane performance, however, since rubbery membranes separate based on solubility selectivity, the CO2/ CH4 separation capability of the polymeric membranes decline sharply and much lower than those of other state-of-the-art glassy polymers such as polyimides and cellulose acetate (CA) [3,6,7]. In addition, the mechanical stability of rubbery polymers tends to fall significantly below that of glassy polymeric membranes. Moreover, since many gas reservoir wells can reach pressures well above 70 bars and H2S concentration of more than 20%, more aggressive feeds need to be considered, and this is a focus of our study. This study has therefore focused on very high H2S composition (up to 20%) in the gas mixture and up to 55 bar for mixed sweet gas permeation properties studies and 34 bar for mixed sour gas transport properties studies. Amongst the polymeric membrane materials that have been investigated for acid gas separations from NG over the last few decades, glassy aromatic polyimides have emerged as major state of the art membranes that have attracted a lot of attention [3–7,12–45]. Some of these polymers exhibit very good permeation properties for various gas pairs (e.g., CO2/CH4; He/CH4; N2/CH4; H2S/CH4; etc., [3–5,22,23]); chemical resistance, high mechanical strength and thermal stability. Since NG is usually treated at high pressures (up to 70 bar) and typically saturated with heavy hydrocarbons (C3+) and water vapor, membranes fabricated from polyimides can easily be used for treating NG because of its excellent properties previously described. Polyimide membranes have shown remarkable permeation properties, especially high selectivities for CO2/CH4 separation [3–5,24–29]. In light of the presence of CF3 groups in Hexafluorodianhydride (6FDA)-based polyimides that result in chain stiffness and hindrance to chain packing, which thus enhances permeability and selectivity, several studies have been carried out on these type of polyimides as gas separation membranes [3–5,22–28,33–35]. Furthermore, the permeation properties of 6FDA-based co-polyimides containing three homo-polyimides (i.e., 6FDA, mPDA and durene moieties) have also been investigated as gas separation membranes in the last few years [22,23,30–34]. Co-polymerization of 2
Journal of Membrane Science xxx (xxxx) xxx–xxx
G.O. Yahaya et al.
studied. The mixed sweet gas transport studies were conducted with feeds containing no H2S up to 55 bar, while mixed sour gas measurements were conducted with feeds containing H2S up to 34 bar for 10% H2S and 20% H2S. 2. Background and theory Gas transport [12] through dense membrane films is mostly modeled by the solution-diffusion mechanism and expressed by Eq. (1):
ji =
Di Si (pi0 − pi1 ) l
(1)
In Fick's and Henry's laws mechanisms, the product DiSi is a measure of the membrane's ability to separate gas, and can be denoted as Pi. This is also known as permeability. The unit for expressing permeability is Barrers, where 1 Barrers = 10−10 cm3(STP).cm/cm2.s.cmHg. A membrane ability to separate two or more different gaseous penetrants is also very important and this property is expressed as the ratio of permeabilities of two gases, i and j. This parameter is called the ideal selectivity or permselectivity, αij [46,47] and it is expressed as:
αij =
pi S D = i × i pj Sj Dj
Fig. 1. Schematic representations of 6FDA, Durene, CARDO and random co-polyimide (6FDA-Durene/CARDO).
Table 1 Composition of monomers in different synthesized random co-polyimides.
(2)
Gas transport through dense film membranes can be represented by the Van’t Hoff-Arrhenius type of Eqs. (3)–(5) [46–48]. This implies that temperature may have a significant effect on permeation rate, and thus influence of temperature on gas permeability, diffusion and solubility coefficients are all expressed as given below:
−ΔHs ⎞ S = S0exp ⎛ ⎝ RT ⎠
(3)
−E D = D0exp ⎛ d ⎞ ⎝ RT ⎠
(4)
⎜
m (6FDA) (mmol)
m (Durene) (mmol)
m (CARDO) (mmol)
6FDA-Durene/CARDO (3:1) 6FDA-Durene/CARDO (1:1) 6FDA-Durene/CARDO (1:3)
22.5
16.9
5.63
22.5
11.3
11.3
22.5
5.63
16.9
pulse experiment was performed using a 90° pulse at 25 °C. A combined relaxation delay and acquisition time of at least 7 s was used and a line broadening of 0.5 Hz was applied before Fourier transformation.
−Ep
⎞ P = P0exp ⎛ ⎝ RT ⎠
Polymer
⎟
(5)
Furthermore, variation in gas permeability as a function of pressure in glassy polymers is often described by dual-mode and partial immobilization models [49,50]. The Langmuir model, which represent excess free volume formed in the glassy state, contribute significant pressure dependence on the permeability in glassy polymers.
3.2.1. Random co-polyimide 6FDA-Durene/CARDO (3:1) In a three-neck round bottom flask equipped with a nitrogen inlet and a mechanical stirrer, 2.7730 g (16.883 mmol) of Durene and 1.9609 g (5.628 mmol) of CARDO were dissolved in 17 mL of m-cresol, then 10 g (22.51 mmol) of 6FDA was added with 16 mL of m-cresol. The mixture was heated at 180 °C for 8 h. The solution was diluted progressively during the reaction by the addition of 24 mL of m-cresol. After cooling, the resulting highly viscous solution was poured into methanol. The fibrous polymer obtained was ground, rinsed with methanol, filtered and dried under reduced pressure for 24 h at 60 °C. 1H NMR spectrum of the copolyimides is shown Fig. 2a (δH (500 MHz, CDCl3) 8.10 – 7.79 (10 H, m, ArH), 7.47 – 7.31 (6 H, m, ArH), 2.14 (12 H, s, CH3)).
3. Experimental 3.1. Materials In this study, series of CARDO-type random co-polyimides were prepared from 2,2’-bis-(3,4’-dicarboxyphenyl) hexafluoropropane dianhydride (6FDA), 2,3,5,6-tetramethyl-1,4-phenylenediamine (durene diamine) (Aldrich) and 9,9-bis(4-aminophenyl) fluorene (CARDO) (Aldrich). Chloroform, m-cresol and methanol (Aldrich) were used as solvents.
3.2.2. Random co-polyimide 6FDA-Durene/CARDO (1:1) In a three-neck round bottom flask equipped with a nitrogen inlet and a mechanical stirrer, 1.8487 g (11.255 mmol) of Durene and 3.9219 g (11.255 mmol) of CARDO were dissolved in 18 mL of m-cresol, then 10 g (22.51 mmol) of 6FDA was added with 18 mL of m-cresol. The mixture was heated at 180 °C for 8 h. The solution was diluted progressively during the reaction by the addition of 25 mL of m-cresol. After cooling, the resulting highly viscous solution was poured into methanol. The fibrous polymer obtained was ground, rinsed with methanol, filtered and dried under reduced pressure for 24 h at 60 °C. 1H NMR spectrum of the copolyimides is shown Fig. 2b (δH (500 MHz, CDCl3) 8.08 – 7.78 (14 H, m, ArH), 7.46 – 7.29 (14 H, m, ArH), 2.12 (12 H, s, CH3)).
3.2. Preparation procedure for the CARDO-type random co-polyimides (6FDA-Durene)/CARDO) and their characterizations As shown in Fig. 1, random co-polyimide with three different ratios of 6FDA-Durene/CARDO moieties (3:1; 1:1 and 1:3) were synthesized as given below and detailed monomer compositions in each of the different random co-polyimides are given in Table 1. The chemical and thermal properties of the synthesized polymers were characterized and confirmed by 1H NMR spectra, thermogravimetric analysis (TGA), and glass transition temperatures (Tg) as described in our previous publications [22,23]. 1 H NMR spectra were typically acquired on a Varian 500 MHz NMR spectrometer, equipped with a 5 mm liquids probe and with deuterated chloroform (CDCl3) as a solvent. In other to obtain 1H spectra, a single
3.2.3. Random co-polyimide 6FDA-Durene/CARDO (1:3) In a three-neck round bottom flask equipped with a nitrogen inlet and a mechanical stirrer, 0.9243 g (5.628 mmol) of Durene and 3
Journal of Membrane Science xxx (xxxx) xxx–xxx
G.O. Yahaya et al.
Fig. 2. a, b & c. 1H NMR spectra of (a) random copolyimide 6FDA-Durene/CARDO (3:1); (b) random copolyimide 6FDA-Durene/CARDO (1:1) and (c) random co-polyimide 6FDA-Durene/CARDO (1:3).
5.8828 g (16.883 mmol) of CARDO were dissolved in 19 mL of mcresol, then 10 g (22.51 mmol) of 6FDA was added with 19 mL of mcresol. The mixture was heated at 180 °C for 8 h. The solution was
diluted progressively during the reaction by the addition of 27 mL of mcresol. After cooling, the resulting highly viscous solution was poured into methanol. The fibrous polymer obtained was ground, rinsed with 4
Journal of Membrane Science xxx (xxxx) xxx–xxx
G.O. Yahaya et al.
methanol, filtered and dried under reduced pressure for 24 h at 60 °C. 1 H NMR spectrum of the co-polyimides is shown Fig. 2c (δH (500 MHz, CDCl3) 8.04 – 7.78 (14H, m, ArH), 7.45 – 7.29 (14H, m, ArH), 2.12 (12 H, s, CH3)).
4. Results and discussion 4.1. Polymers characterization and properties Fig. 2a–c show typical 1H NMR spectra of the co-polyimides. The resonance signals in the range of 7.78 – 8.10 ppm were assigned to the protons of aromatic 6FDA; the signals in the range of 7.45–7.31 ppm were assigned to protons of aromatic CARDO; and the signals in the range of 2.12–2.14 ppm was assigned to the CH3 groups of Durene. The doublets around 7.79 ppm and 7.82 ppm, however, were too ambiguous to assign with confidence to either CARDO or 6FDA. Thus, 1H–1H COSY NMR was employed and it was determined that the signals around 7.79 ppm belonged to CARDO, while that of 7.82 belonged to 6FDA. The incorporation of the monomers were observed to be relatively quantitative, confirming the incorporation of all units in the block copolyimides. The random co-polyimide with three different ratios of 6FDA-Durene/CARDO (3:1; 1:1 and 1:3) were also confirmed by the 1H NMR spectrum by calculating the percentage of the incorporated monomers. The incorporated Durene/CARDO ratio matched well with the initial polymerization ratio as 74.3% Durene/25.7% CARDO was obtained for 6FDA-Durene/CARDO (3:1); 49.7% Durene/50.3% CARDO was obtained for 6FDA-Durene/CARDO (1:1) and 25.1% Durene/74.9% CARDO was obtained for 6FDA-Durene/CARDO (1:3). Fig. 5 shows TGA curves obtained for all the series of the synthesized polymers. The degradation temperature (Td) at 5% weight loss in nitrogen are shown in Table 2. The results show that all the synthesized co-polyimides have excellent thermal stability, as Td above 500 °C was obtained. It also showed no sign of any residual solvent as indicated in Fig. 5. The glass transition temperatures (Tg) of the co-polyimides were obtained using differential scanning calorimetry (DSC), as displayed in Table 2, in the temperature range of 30 °C to 450 °C. Only one Tg value was observed for all copolymers, indicating the absence of a phase separation. This could be attributed to the excellent chemical compatibility of the monomers contained in the random co-polyimides.
3.3. Membrane preparation The dense membrane films have been prepared using a similar procedure as described in our previous publications [22,23], which are basically casting and solvent evaporation methods. It involved preparation of 2–3 wt% polymer solution in chloroform and solution was filtered through a 0.45 µm filter. The solution was then casted on a dry clean petri dish and left to evaporate at room temperature under a clean nitrogen enriched environment overnight. The film was then slowly heated in an oven under a slow nitrogen flow to about 60 °C for about 24 h. After which, the film was left in the oven under vacuum at 60 °C for another 24 h. The resulting film was finally dried in a vacuum oven at 150 °C overnight to remove any residual solvent. After dried completely, the resulting membranes were cooled to room temperature and peeled off from petri dish. The membrane was then dried at ambient temperature under a clean nitrogen environment for about 8 h. 3.4. Gas permeation measurements Pure gas permeability coefficients for He, N2, CO2, and CH4 gases were analyzed for all co-polyimides prepared in this study at various operating temperatures of 25–55 °C and feed pressures of 7–28 bar using a permeation membrane analyzer displayed in Fig. 3 and described in previous publications [22,23,47,48]. Moreover, mixed gas permeability coefficients were measured using a constant pressure system as depicted in Fig. 4. Detailed experimental procedures were described in our previous publications [22,23]. All measurements were recorded after an extended period of time to make sure steady state was reached. These measurements were repeated three times at each operating condition to ensure reproducibility, and uncertainty in the measurements were better or generally less than ± 5% of the value shown. In order to ensure absence of any defect in the membrane and to make sure measurements are reproducible, three to four membrane samples were tested simultaneously at each operating condition.
4.2. Pure gas permeation properties The permeability coefficients of pure gases that include He, CO2, CH4 and N2; and ideal selectivities of gas pairs including He/CH4, N2/
Fig. 3. Schematic diagram of the continuous flow gas permeation device used for measuring single and mixed gas permeation properties. SV-1 and SV-2: air actuated valve; PT-100 and PT-101: pressure transducers; MFM: mass flow meters; NV-101: needle valves; SV-102 and SV-104: 2-way solenoid valves; and SV-101 and SV-103: 3-way solenoid valves; V-101: sample collector.
5
Journal of Membrane Science xxx (xxxx) xxx–xxx
G.O. Yahaya et al.
Fig. 4. Constant pressure/variable volume system for measuring single and mixed gas permeation properties.
Table 3 Pure gas permeability and selectivity coefficients in the random co-polyimide (6FDADurene/CARDO) membranes measured at 7 bar feed pressure and at 35 °C. Random copolyimides
6FDA-Durene/ CARDO (3:1) 6FDA-Durene/ CARDO (1:1) 6FDA-Durene/ CARDO (1:3)
Permeability, Barrer
Selectivity
N2
CH4
He
CO2
N2/CH4
He/CH4
CO2/CH4
20.0
16.1
262
323
1.24
16.3
20.1
10.2
8.86
193
196
1.15
21.8
22.1
3.76
2.57
102
90.4
1.46
39.7
35.2
Fig. 5. TGA traces of random co-polyimides (6FDA-Durene/CARDO).
Table 2 Thermal and physical properties of the synthesized random co-polyimides. Polymer
Tg (°C)
Td at 5% weight loss (N2) (°C)
6FDA-Durene/CARDO (3:1) 6FDA-Durene/CARDO (1:1) 6FDA-Durene/CARDO (1:3)
343.6 341.3 381.0
514.7 526.4 527.3
CH4 and CO2/CH4 through the series of random co-polyimide (6FDADurene/CARDO) measured at upstream pressure of 7–28 bar and at 35 −55 °C are shown in Table 3 and Figs. 6–8. The permeation properties of all penetrant gases depicted are an average of at least two or more measurements, and error in permeability coefficients is less than ± 5% of the values shown. The content of CARDO moiety in the random copolymer was varied from 25% to 75% (3:1–1:3) in order to investigate the effect of segmental moiety variation in transport properties of the copolymers. As can be observed in Table 3, all the penetrants permeabilities decrease, while the selectivities, especially CO2/CH4 and He/CH4 increase, as the CARDO moiety content increases in the copolymers (i.e., 3:1 to 1:3). This trend could be attributed to the increase in restriction in chain mobility around the polymer main backbone due to its stiffness and bulky nature, and thus less suppression in interchain packing as the CARDO moiety content increases, thereby making the membrane more
Fig. 6. CO2/CH4 permeability-selectivity trade-off curve comparison of the random copolyimide (6FDA-Durene/CARDO) (1:3) membrane and other polymeric membranes [3–5,11,15,22,23,51–53] in pure gas feeds.
rigid as evidence from higher Tg values for co-polyimide with higher CARDO moiety content (Table 2). This leads to higher selectivity and lower permeability as CARDO moiety content increases in the copolymer as a result of decline in free volume in the membrane matrix. The decline in permeability with increasing CARDO content could also be attributed to space-filling effect in view of its bulky nature. The 6
Journal of Membrane Science xxx (xxxx) xxx–xxx
G.O. Yahaya et al.
Fig. 8. Temperature dependence of pure gas permeability coefficients of random copolyimide (6FDA-Durene/CARDO) (1:3) membrane measured at 7 bar (permeability P is in the unit of Barrer).
Table 4 Activation energies of permeation for the random co-polyimide (6FDA-Durene/CARDO) membranes at 7 bar. Properties
CO2
He
N2
CH4
EP (kJ/mol) Po (Barrer)
6.75 1.28 × 103
7.29 1.77 × 103
8.87 1.20 × 102
13.44 4.89 × 102
constant or slight increase (especially CO2/CH4) in most of the penetrants selectivities with respect to CH4 as depicted in Fig. 7b. Furthermore, in addition to being selective to both CO2 and He, this co polyimide is also selective to N2 as compare to methane and thus could simultaneously permeate both acid gas and N2, while keeping methane in the high pressure feed stream.
Fig. 7. a & b. (a) Pure gas permeability and (b) selectivity coefficients in the random copolyimide (6FDA-Durene/CARDO) (1:3) membrane as function of feed pressure measured at 35 °C.
permeability values of 90.4 and 102.1 barrers for CO2 and He respectively and CO2/CH4 and He/CH4 selectivities of 35.18 and 39.73 respectively obtained for the random copolymer (6FDA-Durene/CARDO (1:3)) with the highest content of CARDO moiety (75%) are very close to the target performance being sought for acid gas separations from natural gas application. As shown in Figs. 6 and 7a, an order of magnitude increment in all the penetrants permeabilities was achieved in the random co-polyimide 6FDA-Durene/CARDO (1:3) membrane as compared to the penetrants permeabilities obtained in the non-CARDO counterparts [22,23]. In addition, the selectivities of most of the penetrant gases with respect to CH4 showed insignificant changes, as compared to those obtained in the previous studies [22,23]. The improved gas separation performance could be attributed to suppression or loosening of interchain packing by the substitution of the bulky CARDO moiety as well as some possible conformational changes in the backbone. This can result in the simultaneous inhibition of intrachain motion around flexible hinge points, which tends to increase permeability without unacceptable losses in selectivity. These values of permeability and selectivity obtained are comparable and very competitive even, as compared to the values obtained in some of the high performance polymeric membranes that have been reported in the literature [3–5,11,15,22,23,51–53] (Fig. 6). In light of this, detailed pure gas permeation properties that include influence of pressure and temperature were investigated on the random copolymer 6FDA-Durene/CARDO (1:3) membrane.
4.2.2. Effect of operating temperature on pure gas permeabilities The influence of temperature on pure gas permeabilities and selectivities of the random co-polyimide (1:3) membrane was investigated over varying operating temperature of 35–55 °C and at 7 bar feed pressure, and the results are shown in Fig. 8. Fig. 8 shows plots of T−1 vs. natural log of permeability by applying Eq. (5), and using the least square fitting method, the Arrhenius parameters were obtained, and tabulated in Table 4. As depicted in the figure, the permeability coefficients of all gases increase with increase in temperature. This trend that has also been observed in our previous studies [22,23], could be attributed to differences in permeation activation energies for various gas pairs. The trend in activation energies of permeation Ep obtained for the permeation process as depicted in Table 4, i.e., EP (CO2) < EP (He) < EP (N2) < EP (CH4) is in good agreement with their permeabilities. Ep also increases with increase in kinetic diameter of penetrant gases in most cases. The Ep values estimated in this study and the trend observed are in close agreement with those reported in our previous publications [22,23]. Even though there is close agreement between this study and literatures, however, previous studies focused on random and block copolyimides (6FDA-Durene/6FDA-mPDA), while this study was focused exclusively on the random co-polyimides 6FDA-Durene/CARDO membranes.
4.2.1. Effect of pressure on pure gas permeabilities As shown in Fig. 7a, permeability coefficients of most of the penetrants that include He, CO2, CH4 and N2 stay relatively constant or slightly increases (especially CO2) with increase in feed pressure, indicating insignificant plasticization effect up to a feed pressure of about 28 bar. Similar trends have been observed in our previous studies involving block co-polyimide (6FDA-durene)−(6FDA-mPDA) membranes [22,23]. The random co-polyimide membrane also showed almost
4.3. Mixed gas permeation and separation properties The permeation properties of sweet gas mixtures consisting of 10, 59, 30 and 1 vol% CO2, CH4, N2 and C2H6 respectively through random co-polyimide (6FDA-Durene/CARDO) (1:3) membrane were studied at upstream pressure of up to 55 bar and at 22 °C (Fig. 9a & b). In addition, the effect of H2S in the feed mixture on the transport 7
Journal of Membrane Science xxx (xxxx) xxx–xxx
G.O. Yahaya et al.
As can be observed in Fig. 9a, the permeability coefficient of CO2 decrease with increasing feed pressure and the minimum permeability of about 74 barrer was obtained at upstream pressure of 55 bar in the membranes. This value is quite attractive especially at this elevated pressure of 55 bar. However, permeabilities of CH4, N2, and C2H6 were observed to be constant or slightly decrease with increase in feed pressure. The decrease in permeability of CO2 may be attributed to penetrant gases competition for sorption sites in the membrane matrix, and CO2 loses sorption sites to other gases with increasing pressure. Furthermore, unlike the pure gas studies (Fig. 7b); the mixed gas studies with this co-polyimide did exhibit trade off. This is indicated in Fig. 9b, where CO2/CH4 selectivity decline with increasing pressure and the minimum selectivity of 31.5 was obtained at upstream pressure of 55 bar. However, C2H6/CH4 and N2/CH4 selectivities remain constant or slightly decrease with increase in feed pressure. 4.3.2. Sour gas mixture permeation properties In order to assess the real performance of the random co-polyimide membrane, especially under an aggressive H2S environment (i.e., H2S concentration of up to 20 vol%), sour gas mixture tests were clearly needed. Thus, sour gas mixture consisting of five gases were made containing 10; 57–60; 10–20; 3 and 10–20 vol% of CO2, CH4, N2, C2H6 and H2S respectively. As shown in Table 5, permeability coefficients of all the penetrants that include CO2, CH4, N2, C2H6 and H2S stay relatively constant or slightly decrease with increase in pressure for the H2S composition of 10 and 20 vol%. Furthermore, the general enhancement in gas separation performance and very attractive transport properties exhibited by this membrane as compared to non-CARDO counterparts [22,23], could be attributed to a suppression or loosening of interchain packing by the presence of the bulky CARDO group as well as some possible conformational changes in the backbone. This can result in the simultaneous inhibition of intrachain motion around flexible hinge points due to its stiffness, which tends to increase permeability without unacceptable losses in selectivity. The CO2/CH4 and H2S/CH4 ideal selectivities are in the range of 18–23 and 19–21 respectively; while CO2 and H2S permeabilities are in the range of 38–51 and 40–47 barrers respectively at 20 vol% H2S. These values are comparable and very competitive even, as compared to the values obtained in some of the high performance polymeric membranes that have been reported in the literature [3–5,8–11,22,23] as shown in Fig. 10a & b.
Fig. 9. a & b. (a) Sweet mixed gases permeability and (b) selectivity coefficients in the random co-polyimide (6FDA-Durene/CARDO) (1:3) membrane as function of feed pressure at 22 °C using gas mixture containing 10, 59, 30 and 1 vol% of CO2, CH4, N2 and C2H6 respectively.
properties was investigated. The H2S concentration was varied from 10 to 20 vol%, while CO2 concentration was kept at 10 vol% in the simulated sour gas mixtures that contain 10; 57–60; 10–20; 3 and 10–20 vol% of CO2, CH4, N2, C2H6 and H2S respectively. The study was conducted at different gas feed pressure as shown in Table 5.
4.3.1. Sweet gas mixture permeation properties As shown in Fig. 9a & b, an order of magnitude increment in permeabilities and selectivities with respect to CH4 for most of the penetrant gases were observed in the random co-polyimide 6FDA-Durene/ CARDO (1:3) membrane, as compared to the penetrants permeabilities obtained in the non-CARDO counterparts [22,23]. The improved gas separation performance could be attributed to a suppression or loosening of interchain packing by the substitution of the bulky CARDO group as well as some possible conformational changes in the backbone. This can result in the simultaneous inhibition of intrachain motion around flexible hinge points due to its stiffness, which tends to increase permeability without unacceptable losses in selectivity.
5. Conclusion This study presents the transport properties and separation performance of pure and gas mixture consisting of CO2, H2S, CH4, N2 and C2H6 through the dense films of a series of aromatic random 6FDA-Durene/ CARDO co-polyimides with varying content of CARDO moiety (3:1; 1:1 and 1:3) for simultaneous separation of CO2, H2S and N2 from sour gas streams. The membranes exhibit very attractive pure and mixed gas permeation properties. For pure gas, the CO2 permeability is up to 323
Table 5 Mixed gases permeability and selectivity coefficients in the random co-polyimide (6FDA-CARDO/Durene) (1:3) membrane as function of H2S composition, acid gas (CO2+H2S) partial pressure and total feed pressure at 22 °C using gas mixture containing 10; 57–60; 10–20; 3 and 10–20 vol% of CO2, CH4, N2, C2H6 and H2S respectively. H2S comp. vol %
10.0
20.0
Total feed pres. (bar)
21 31 34 14 21
Partial pres. CO2 (bar)
2.10 3.10 3.45 1.38 2.10
Partial pres. H2S (bar)
2.10 3.10 3.45 2.76 4.13
Partial Press. (CO2+H2S) (bar)
4.20 6.20 6.90 4.14 6.23
Gas mixture composition: 10% H2S: including 20% N2; 60% CH4; 10%CO2. 20% H2S: including 10% N2; 57% CH4; 10%CO2; 3% C2H6.
8
Permeability (Barrer)
Ideal selectivity
N2
CH4
C2H6
CO2
H2S
CO2/CH4
H2S/CH4
1.73 1.54 1.53 2.46 1.88
1.95 1.94 2.07 2.27 2.12
0.00 0.00 0.00 2.18 2.03
42.4 37.2 37.5 51.4 38.1
41.4 40.0 42.5 47.3 40.4
21.8 19.2 18.2 22.7 18.0
21.2 20.6 20.6 20.9 19.1
Journal of Membrane Science xxx (xxxx) xxx–xxx
G.O. Yahaya et al.
Saudi Arabia (OGT-NGUS-143-15) is greatly acknowledged. The authors would also like to thank Donya A. Sewdan and Qasim Saleem of technical support division, R & DC, Saudi Aramco for their help and valuable contribution with regards to NMR characterization of the polymers. References [1] B.D. Bhide, A. Voskericyan, S.A. Stern, Hybrid processes for the removal of acid gases from natural gas, J. Membr. Sci. 140 (1998) 27–49. [2] R.W. Baker, Future directions of membrane gas separation technology, Membr. Technol. 138 (2001) 5–10. [3] B. Kraftschik, W.J. Koros, J.R. Johnson, O. Karvan, Dense film polyimide membranes for aggressive sour gas feed separations, J. Membr. Sci. 428 (2013) 608–619. [4] B. Kraftschik, W.J. Koros, Cross-linkable polyimide membranes for improved plasticization resistance and permselectivity in sour gas separations, Macromolecules 46 (2013) 6908–6921. [5] S. Yi, X. Ma, I. Pinnau, W.J. Koros, A high-performance hydroxyl-functionalized polymer of intrinsic microporosity for an environmentally attractive membranebased approach to decontamination of sour gas natural gas, J. Mater. Chem. A. 3 (2015) 22794–22806. [6] W.J. Koros, R.T. Chern, Separation of gaseous mixtures using polymer membranes, in: R.W. Rousseau (Ed.), Handbook of Separation Process Technology, John Wiley & Sons, Ltd., New York, 1987. [7] W.J. Koros, G.K. Flemming, Review: membrane-based gas separation, J. Membr. Sci. 83 (1993) 1–80. [8] C.J. Orme, F.F. Stewart, Mixed gas hydrogen sulfide permeability and separation using supported polyphosphazene membranes, J. Membr. Sci. 253 (2005) 243–249. [9] G. Chatterjee, A.A. Houde, S.A. Stern, Poly(etherurethane) and poly (ether urethane urea) membranes with high H2S/CH4 selectivity, J. Membr. Sci. 135 (1997) 99–106. [10] T. Mohammadi, M.T. Moghadam, M. Saeidi, M. Mahdyarfar, Acid gas permeation behavior through poly (ester urethane urea) membrane, Ind. Eng. Chem. Res. 47 (2008) 7361–7367. [11] C.S.K. Achoundong, N. Bhuwania, S.K. Burgess, O. Karvan, J.R. Johnson, W.J. Koros, Silane modification of cellulose acetate dense films as materials for acid gas removal, Macromolecules 46 (2013) 5584–5594. [12] J.G. Wijmans, R.W. Baker, The solution-diffusion model: a review, J. Membr. Sci. 107 (1995) 1–21. [13] B.D. Freeman, I. Pinnau, Polymeric materials for gas separations, in Polymeric membranes for gas and vapor separations: chemistry and materials science, in: B.D. Freeman, I. Pinnau (Eds.), ACS Symposium Series, 733 American Chemical Society, Washington, D.C., 1999, p. 1. [14] L.M. Robeson, Polymer membranes for gas separation, Curr. Opin. Solid State Mater. Sci. 4 (1999) 549–552. [15] J.D. Wind, D.R. Paul, W.J. Koros, Natural gas permeation in polyimide membranes, J. Membr. Sci. 228 (2004) 227–236. [16] S.A. Stern, Polymers for gas separations: the next decade, J. Membr. Sci. 94 (1994) 1–65. [17] P.M. Budd, N.B. McKeown, Highly permeable polymers for gas separation membranes, Polym. Chem. 1 (2010) 63–68. [18] Y. Xiao, B.T. Low, S.S. Hosseini, T.S. Chung, D.R. Paul, The strategies of molecular architecture and modification of polyimide-based membranes for CO2 removal from natural gas - a review, Prog. Polym. Sci. 34 (2009) 561–580. [19] J.T. Hupp, C.A. Mirkin, O.K. Farha, A.M. Spokoyny, Gas adsorption and gas mixture separation using porous organic polymer. US 2011/0160511 A1. [20] S. Nakanishi, K. Ito, Y. Kusuki, Aromatic polyimide and gas separation. EP 0747418, 1996. [21] Y.F. Ozcayir, G. Goetz, B. Bikson, Novel sulfonated polyimide gas separation membranes. EP 0750939, 1997. [22] R. Heck, M.S. Qahtani, G.O. Yahaya, I. Tanis, D. Brown, A.A. Bahamdan, A.W. Ameen, M.M. Vaidya, J.-P.R. Ballaguet, R.H. Alhajry, E. Espuche, R. Mercier, Block co-polyimide membranes for pure- and mixed- gas separation, Sep. Purif. Technol. 173 (2017) 183–192. [23] G.O. Yahaya, M.S. Qahtani, A.Y. Ammar, A.A. Bahamdan, A.W. Ameen, R.H. Alhajry, M.M. Ben Sultan, F. Hamad, Aromatic block co-polyimide membranes for sour gas feed separations, Chem. Eng. J. 304 (2016) 1020–1030. [24] T.H. Kim, W.J. Koros, G.R. Husk, K.C. O'Brien, Relationship between gas separation properties and chemical structure in a series of aromatic polyimides, J. Membr. Sci. 37 (1988) 45–62. [25] S.A. Stern, Y. Mi, H. Yamamoto, A.K. St. Clair, Structure/permeability relationships of polyimide membranes. Applications to the separation of gas mixtures, J. Polym. Sci. Part B: Polym. Phys. 27 (1989) 1887–1909. [26] M.R. Coleman, W.J. Koros, Isomeric polyimides based on fluorinated dianhydrides and diamines for gas separation applications, J. Membr. Sci. 50 (1990) 285–297. [27] C. Bas, R. Mercier, J. Sanchez-Marcano, S. Neyertz, N.D. Albe´rola, E. Pinel, Copolyimides containing alicyclic and fluorinated groups: solubility and gas separation properties, J. Polym. Sci., Part B: Polym. Phys. 43 (2005) 2413–2426. [28] Y.-C. Wang, S.-H. Huang, C.-C. Hu, C.-L. Li, K.-R. Lee, D.-J. Liaw, J.-Y. Lai, Sorption and transport properties of gases in aromatic polyimide membranes, J. Membr. Sci. 248 (2005) 15–25. [29] M.K. Ghosh, K.L. Mittal, Polyimides: Fundamentals and Applications, Marcel Dekker Inc, 1996. [30] K. Tanaka, M. Okano, H. Toshino, H. Kita, K.-I. Okamoto, Effect of methyl
Fig. 10. a & b. (a) CO2/CH4 and (b) H2S/CH4 permeability-selectivity trade-off comparison of random co-polyimide (6FDA-Durene/CARDO) (1:3) membrane and other polymeric membranes [3–5,11,22,23] in ternary H2S/CO2/CH4 feed mixtures.
barrer and CO2/CH4 selectivity is up to 35 obtained at 35 °C and feed pressure of up to 28 bar. In addition, for sweet gas mixture (i.e., no H2S), the random co-polyimide 6FDA-Durene/CARDO (1:3) membrane shows CO2 permeability and CO2/CH4 selectivity of about 74 barrer and 31 respectively at an elevated pressure of 55 bar. In case of high content of sour gas mixtures (20 vol% H2S in the feed gas), the membrane exhibits CO2/CH4 and H2S/CH4 ideal selectivities in the range of 18–23 and 19–21 respectively; and CO2 and H2S permeabilities in the range of 38–51 and 40–47 barrers respectively. These values and separation performance exhibited by the random co-polyimide are comparable and very competitive even, as compared to the values obtained in some of the high performance polymeric membranes that have been reported in the literature. One significant point to note is that at moderate feed pressure and up to 20 vol% H2S in feed gas mixture, ideal selectivities and permeabilities are still very attractive and remarkably stable, even under these much more aggressive environments. With this remarkable membrane performance, it indicates that this membrane is potentially a very promising candidate in combination with amine absorption for aggressive sour gas separations. With this potentially attractive technology, significant cost savings could be achieved when compared to using conventional technology such as stand-alone amine absorption systems for acid gas separations from natural gas. Acknowledgment The support provided by the Research & Development Center (R & DC) of Saudi Arabian Oil Company (Saudi Aramco), Dhahran, 9
Journal of Membrane Science xxx (xxxx) xxx–xxx
G.O. Yahaya et al.
[31]
[32]
[33]
[34]
[35] [36]
[37]
[38]
[39]
[40]
[41]
[42] M.P. Chenar, M. Soltanieh, T. Matsuura, A. Tabe-Mohammadi, K.C. Khulbe, The effect of water vapor on the performance of commercial polyphenylene oxide and cardo-type polyimide hollow fiber membranes in CO2/CH4 separation applications, J. Membr. Sci. 285 (2006) 265–271. [43] Y.F. Yeong, H. Wang, K.P. Pramoda, T.-S. Chung, Thermally induced structure rearrangement of cardo-copolybenzoxazole membranes for enhanced gas transport properties, J. Membr. Sci. 397-398 (2012) 51–65. [44] Y. Lu, J. Hao, L. Li, J. Song, G. Xiao, H. Zhao, Z. Hu, Preparation and gas transport properties of thermally induced rigid membranes of copolyimide containing cardo moities, React. Funct. Polym. 119 (2017) 134–144. [45] X. Sun, C. Ma, B. Yuan, T. Wang, T. Xu, Q. Xue, P. Li, Y. Kong, Cardo polyimides/ TiO2 mixed matrix membranes: synthesis, characterization, and gas separation property improvement, Sep. Purif. Technol. 122 (2014) 367–375. [46] R.D. Raharjo, B.D. Freeman, D.R. Paul, G.C. Sarti, E.S. Sanders, Pure and mixed gas CH4 and n-C4H10 permeability and diffusivity in poly(dimethylsiloxane), J. Membr. Sci. 306 (2007) 75–92. [47] C.K. Yeom, J.M. Lee, Y.T. Hong, K.Y. Choi, S.C. Kim, Analysis of permeation transients of pure gases through dense polymeric membranes measured by a new apparatus, J. Membr. Sci. 166 (2000) 71–83. [48] C.K. Yeom, S.H. Lee, J.M. Lee, Study of transport of pure and mixed CO2/N2 gases through polymeric membrane, J. Appl. Polym. Sci. 78 (2000) 179–189. [49] H. Park, Y. Lee, Polymeric membrane materials and potential use in gas separation, in: N. Li, A. Fane, Ho.W. Winston, T. Matsuura. (Eds.), Advanced Membrane Technology and Applications, John Wiley & Sons, Hoboken, 2008, pp. 633–664. [50] T.M. Murphy, G.T. Offord, D.R. Paul, Fundamentals of membrane gas separation, in: E. Drioli, L. Giorne (Eds.), Membrane Operations: Innovative Separations and Transformations, Wiley –VCH Verlag GmbH & Co KGaA, Weinheim, 2009, pp. 63–82. [51] M.D. Guiver, G.P. Robertson, Y. Dai, F. Bilodeau, Y.S. Kang, K.J. Lee, J.Y. Jho, J. Won, Structure characterization and gas-transport properties of brominated matrimid polyimide, J. Polym. Sci.: Part A: Polym. Chem. 40 (2002) 4193–4204. [52] A.M.W. Hillock, S.J. Miller, W.J. Koros, Crosslinked mixed matrix membranes for the purification of natural gas: effect of sieve surface modification, J. Membr. Sci. 314 (2008) 193–199. [53] J.T. Vanghn, W.J. Koros, J.R. Johnson, O. Karvan, Effect of thermal annealing on a novel polyamide-imide polymer membrane for aggressive acid gas separations, J. Membr. Sci. 401–402 (2012) 163–174.
substituents on permeability and permselectivity of gases in polyimides prepared from methyl-substituted phenylenediamines, J. Polym. Sci. Part B: Polym. Phys. 30 (1992) 907–914. L. Wang, Y. Cao, M. Zhou, X. Ding, Q. Liu, Q. Yuan, The gas permeation properties of 6FDA-2, 4, 6-trimethyl-1, 3-phenylenediamine (TMPDA)/1, 3-phenylenediamine (mPDA) co-polyimides, Polym. Bull. 60 (2008) 137–147. W.-H. Lin, T.-S. Chung, Gas permeability, diffusivity, solubility, and aging characteristics of 6FDA-durene polyimide membranes, J. Membr. Sci. 186 (2001) 183–193. T.-S. Chung, W.-H. Lin, R.H. Vora, Gas transport properties of 6FDA-durene/1,3 phenylenediamine (mPDA) copolyimides, J. Appl. Polym. Sci. 81 (2001) 3552–3564. S.-X. Cheng, T.-S. Chung, R. Wang, R.H. Vora, Gas-sorption properties of 6FDA–durene/1,4-phenylenediamine (pPDA) and 6FDA–durene/1,3-phenylenediamine (mPDA) copolyimides, J. Appl. Polym. Sci. 90 (2003) 2187–2193. V. Korshak, S. Vinogradova, Y. Vygodskii, J. Macromol. Sci. - Rev. Macromol. Chem. C11 (1) (1974) 45–53. S. Karashima, Y. Tokuda, A. Tachiki, K. Haraya, C. Kamizawa, Development of Cardo-type polyimide hollow fiber membranes for CO2 separation, in: P. Riemer, B. Eliasson, A. Wokaun (Eds.), Greenhouse Gas Control Technologies, Elsevier Sci. Ltd., 1990, p. 1032. S. Kazama, T. Teramoto, K. Haraya, Carbon dioxide and nitrogen transport properties of bis(phenyl)fluorine-based CARDO polymer membranes, J. Membr. Sci. 207 (2002) 91–104. Y. Tokuda, E. Fujisawa, N. Okabyashi, N. Matsumiya, K. Takagi, H. Mano, K. Haraya, M. Sato, Development of hollow fiber membranes for CO2 separation, Energy Convers. Manag. (1997) 38 (Sill). M.P. Chenar, M. Soltanieh, T. Matsuura, A. Tabe-Mohammadi, C. Feng, Gas permeation properties of commercial polyphenylene oxide and Cardo-type polyimide hollow fiber membranes, Sep. Purif. Technol. 51 (2006) 359–366. S. Kazama, S. Morimoto, S. Tanaka, H. Mano, T. Yashima, K. Yamada, K. Haraya, CARDO polyimide membranes for CO2 capture from flue gases, Greenhouse Gas Control Technologies, 1 Elsevier Sci. Ltd., 2005, pp. 75–82. M.P. Chenar, M. Soltanieh, T. Matsuura, A. Tabe-Mohammadi, M. Sadeghi, Application of cardo-type polyimide (PI) and polyphenylene oxide (PPO) hollow fiber membranes in two-stage membrane systems for CO2/CH4 separation, J. Membr. Sci. 324 (2008) 85–94.
10