Pebax–PEG–MWCNT hybrid membranes with enhanced CO2 capture properties

Pebax–PEG–MWCNT hybrid membranes with enhanced CO2 capture properties

Journal of Membrane Science 460 (2014) 62–70 Contents lists available at ScienceDirect Journal of Membrane Science journal homepage: www.elsevier.co...

1MB Sizes 11 Downloads 90 Views

Journal of Membrane Science 460 (2014) 62–70

Contents lists available at ScienceDirect

Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci

Pebax–PEG–MWCNT hybrid membranes with enhanced CO2 capture properties Shaofei Wang a,b, Ye Liu a,b, Shixin Huang a,b, Hong Wu a,b, Yifan Li a,b, Zhizhang Tian a,b, Zhongyi Jiang a,b,n a b

Key Laboratory for Green Chemical Technology, School of Chemical Engineering and Technology, Tianjin University, Tianjin 300072, China Collaborative Innovation Center of Chemical Science and Engineering (Tianjin), Tianjin 300072, China

art ic l e i nf o

a b s t r a c t

Article history: Received 31 August 2013 Received in revised form 23 January 2014 Accepted 24 February 2014 Available online 1 March 2014

Three-component hybrid membranes comprising Pebax, poly (ethylene glycol) (PEG)-based polymers and multi-walled carbon nanotube (MWCNT) were prepared by simple physical mixing of Pebax solution and MWCNT-containing PEG-based polymer solution. Scanning electron microscope (SEM) images showed an improved dispersion of MWCNT within polymer matrix arisen from the hydrophilic modification of PEG-based polymers. The embedded MWCNT disrupted the polymer chain packing and decreased polymer crystallinity, which was confirmed by X-ray diffraction. Influences of PEG-based polymer type on CO2, N2 and CH4 permeations were investigated: incorporating high molecular weight PEG-based polymers favored CO2/CH4 separation, while incorporating low molecular weight PEG-based polymers favored CO2/N2 separation. All the hybrid membranes displayed enhanced CO2 permeability owing to the increased amorphous phase. With the incorporation of MWCNT, CO2/CH4 selectivity was decreased for the enhanced chain mobility, while CO2/N2 selectivity was increased for the increased amorphous PEO content. In particular, for P-PEGDME(40)–CNT(5) membrane, a CO2/N2 mixed gas selectivity of 108 and CO2 permeability up to 743 Barrer at 10 bar and room temperature were acquired, which showed promising prospect for CO2 capture. & 2014 Elsevier B.V. All rights reserved.

Keywords: Pebax PEG-based polymer Multi-walled carbon nanotube Hybrid membrane CO2 capture

1. Introduction In the past decades, membrane-based gas separation has garnered considerable interests in industry because of its inherent advantages such as small footprint, lower capital and operating costs compared with other traditional separation technologies [1]. Polymers are usually employed as the membrane materials owning to their low cost and ease of preparation. However, the intrinsic trade-off between permeability and selectivity of polymeric materials [2,3] undermines their further scale-up and concentrated efforts have been devoted to revealing the structure–property relationship in polymeric membranes [4–6] to improve their separation performances. The diffusion of gas molecules through polymer matrix mainly depends on the free volume characteristics: the static free volume voids created by inefficient chain packing or transient gaps generated

n Corresponding author at: Key Laboratory for Green Chemical Technology of Ministry of Education, School of Chemical Engineering and Technology, Tianjin University, 92# Weijin Road, District Nankai, Tianjin 300072, China. Tel./fax: þ 862 223 500086. E-mail address: [email protected] (Z. Jiang).

http://dx.doi.org/10.1016/j.memsci.2014.02.036 0376-7388 & 2014 Elsevier B.V. All rights reserved.

by thermally induced chain segment rearrangement [7,8]. The crystalline phases with high chain packing efficiency and low chain mobility in polymer matrix are generally regarded as the bottleneck that restricts the polymer's permeability [9]. Researchers have paid tremendous attentions to finding different ways to disturb the chain packing and decrease the crystallinity of polymer, in order to enhance membrane performances in terms of permeability and selectivity [8,10,11]. Incorporating inorganic fillers into polymer matrix to fabricate polymer–inorganic hybrid membrane has evolved an efficient, facile method to intervene polymer chain packing and improve free volume characteristics, especially when filler size is close to the characteristic size of the polymer chain packing [12–15]. Accordingly, various types of nonporous fillers (SiO2 [8], metal oxide [14], clay [16]) and porous fillers (zeolite [17], carbon molecular sieve [18], carbon nanotube (CNT) [19]) have been attempted. Among them, CNT has attracted considerable research interests [19–21] due to its porous structure and excellent mechanical properties. When embedded into a polymer matrix, the strong interaction between polymer chains and CNT may favorably mediate the polymer chain packing and significantly enhance gas diffusion behavior [20].

S. Wang et al. / Journal of Membrane Science 460 (2014) 62–70

In the preparation of CNT-based hybrid membranes, pretreatment of CNT by chemical or physical modification to acquire a homogenous dispersion in the polymer matrix has been widely exploited [20,22–24]. Ismail et al. [25] functionalized multi-walled carbon nanotube (MWCNT) with 3-aminopropyltriethoxysilane and acquired an improved dispersion in polyethersulfone. Aroon et al. [26] utilized chitosan as a hydrophilic polymer to wrap onto MWCNT in order to achieve the uniform dispersion of MWCNT in polyimide. Majeed et al. [24] employed pyrene–POSS nanohybrid as dispersant for the physical modification of MWCNT surface. In a representative preparation process, CNT was treated with the dispersing agent and then incorporated into the polymer matrix after being dried. This process may have the intrinsic limitation that the CNT often aggregates again during the drying process. Moreover, the influence of dispersing agent on membrane separation performances is seldom investigated. Priming method, coating the surface of the filler with a dilute polymer prior to dispersion in the bulk polymer, is a simple method to minimize agglomeration of the fillers [27]. The dilute polymer can be the same as the bulk polymer [28], or it can be the coupling agent [29] or other additives [30]. The priming method could be an effective alternative to overcome the aggregation because the homogenous dispersion state is well retained in solutions; moreover, the dispersing agent may be rationally designed to further assist membrane separation. In the present study, multi-walled carbon nanotube (MWCNT) was pre-dispersed in PEG-based polymers (hereafter denoted PEGs) solutions and the whole suspension was transferred into Pebax solutions to prepare the hybrid membranes. PEGs can endow hydrophilic modification to MWCNT [31] and improve its dispersion. Moreover, ethylene oxide (EO) unit in PEGs has been proved an effective group to achieve high CO2 permeability and CO2/light gas selectivity [32]. Systematic study of Pebax and low molecular PEG blend membranes has been conducted by Peinemann et al., which offers facile approach to improve CO2 separation performances [33–35]. In this study, a broader variety of the PEGs (PEG20000, PEG10000, PEG2000, PEG600, PEG400, Triton X-100, and poly (ethylene glycol) dimethylether (PEGDME)) with different molecular weights and terminal groups were utilized for preparing the hybrid membranes to investigate the molecular weight dependence. CO2, CH4 and N2 gases were chosen for membrane separation performance evaluation because they are the common gases in carbon capture (CO2/CH4 for natural gas sweetening and CO2/N2 for flue gas treatment). Additionally, the influences of PEGs and MWCNT contents, operation pressure and temperature on gas permeation performances of the resultant membranes were investigated.

2. Experimental 2.1. Materials Pebax(R) MH 1657 containing 60 wt% of poly (ethylene oxide) (PEO) and 40 wt% polyamide 6 (PA6) was purchased from Arkema (Paris, France). Multi-walled carbon nanotube (MWCNT) (inner diameter 1–7 nm, outer diameter 3–20 nm, purity 495%) with –OH on the surface was supplied from the Department of Chemical Engineering, Tsinghua University, China. Ethanol and poly (ethylene glycol) (PEG) with different molecular weights: PEG10000, PEG2000, PEG600 and PEG400 were purchased from Tianjin Guangfu Fine Chemical Research Institute (Tianjin, China). PEG20000 (M.W. 20,000) and poly (ethylene glycol) dimethylether (PEGDME) (average M.W.  500) were obtained from Sigma Aldrich. Triton X-100 (poly(ethylene glycol) tertoctylphenyl ether, Fig. 1, M.W. 625) was received from Alfa Aesar. All

63

Fig. 1. Chemical structure of Triton X-100.

the reagents were of analytical grade and used without further purification. Deionized water was used throughout this study.

2.2. Membrane preparation Seven types of PEGs (PEG20000, PEG10000, PEG2000, PEG600, PEG400, Triton X-100 and PEGDME) were employed in the experiments. Hybrid membranes containing PEG20000 were taken as the example to illustrate membrane preparation procedure and membranes containing other types of PEGs followed the same protocol. A specified amount of PEG20000 was first dissolved in water at room temperature with constant stirring to get an aqueous polymer solution. Afterwards, pre-grinded MWCNT was slowly added into the solution and dispersed by sonication for several hours to exfoliate bundles and improve the homogeneity. A certain amount of Pebax was dissolved in ethanol/water mixture (70/30 wt%) under mild mechanical stirring at 80 1C to get a 4.5 wt% homogeneous solution. Then, the Pebax solution was dropwise added to the slurry of MWCNT and stirred at high velocity for 12 h to ensure the uniform dispersion of MWCNT within polymer matrix. After removing bubbles, the final solutions were cast on Teflon molds and dried under ambient conditions for 24 h. Then the membranes were further dried in a vacuum oven at 45 1C for 24 h to remove the residual solvent. For comparison, Pebax and PEGs blend membranes were also prepared following the same procedure but without MWCNT. The resultant membranes with the thickness of 80– 100 μm were denoted as P A(x) or P A(x)–CNT(y), where P represents Pebax, A represents the type of PEGs, x (0, 10, 20, 40) denotes the wt% of polymer A to the total mass of Pebax plus A, and y (0, 1, 2, 3, 5) denotes the wt% of MWCNT to the total mass of Pebax plus A.

2.3. Membrane characterization 2.3.1. Field emission scanning electron microscope (FESEM) The cross section morphology and dispersion of MWCNT in the membrane samples were examined with a Nanosem 430 field emission scanning electron microscope operated at 10 kV. Before analyzing, membranes were cryogenically fractured in liquid nitrogen and then sputtered with a thin layer of gold.

2.3.2. Fourier transform infrared (FT-IR) spectroscopy FT-IR spectra for resultant membranes with and without MWCNT were obtained by using a BRUKER Vertex 70 Fourier transform infrared spectrometer with scan range of 4000– 400 cm  1 and resolution of 1.93 cm  1.

2.3.3. X-ray diffraction (XRD) The crystal structure and intermolecular distances between the intersegmental chains were recorded on a Rigaku D/max 2500 v/ pc X-ray diffractiometer (XRD) in the range of 5–651 at the scan rate of 31 min  1. The X-rays of 1.5406 ˚ wavelength were generated by a Cu Kα source. The average d-spacing value of the samples was calculated by using Bragg's law (d ¼ λ/2 sin θ).

64

S. Wang et al. / Journal of Membrane Science 460 (2014) 62–70

2.3.4. Differential scanning calorimeter (DSC) measurements Glass transition temperatures (Tg) of the membranes were analyzed by a Netzsch DSC 200F3 calorimeter. The measurements were performed from 100 to 250 1C at the scan rate of 10 1C /min, and nitrogen was used as a purge gas with a flow rate of 20 ml/min. Tg was determined as the midpoint temperature of the transition in the DSC curve. 2.4. Gas permeation experiments Gas permeation performances of the membranes were determined at room temperature (about 22 1C) based on the conventional constant pressure/variable volume method. Pure or mixed gas separation performances of the flat-sheet membranes were tested by using the same apparatus as that in a previous paper [36]. Pure component gases CO2, N2 and CH4 or gas mixtures of CO2/CH4 (30/70 vol%) and CO2/N2 (20/80 vol%) were used as the feed gas. N2 was employed as the sweep gas when the feed gas was CO2, CH4 or CO2/CH4 mixtures, while CH4 was employed as the sweep gas when the feed gas was N2 or CO2/N2 mixtures, and the sweep gas was kept at atmospheric pressure. A circular membrane disc with effective permeation area of 12.5 cm2 was applied. Volumetric flow rate of feed gas and sweep gas was measured by two separate mass flowmeters. The concentration of CO2, N2 and CH4 in the feed and permeate gases was measured by an Agilent 6820 gas chromatography equipped with a thermal conductive detector (TCD). After the steady state was reached (about 2 h), the permeability (Pi, Barrer, 1 Barrer equals 1  10  10 cm3 (STP) cm/(cm2 s cmHg)) of either gas was measured and each set of data was obtained from at least twice replicates. Pi was expressed by the equation: Pi ¼Qil/ΔpiA, where Qi is the volumetric flow rate of gas ‘i’ (cm3/s (SPT)), l is the thickness of the membranes measured by a micrometer calliper (μm), Δpi is the trans-membrane pressure difference (cmHg), and A is the effective membrane area. The ideal selectivity of gas ‘i and j’ (αi/j) was calculated by αi/j ¼Pi/Pj. Since the permeate side was maintained at atmospheric pressure, the mixed-gas selectivity was also calculated by αi/j ¼ Pi/Pj. To detect the effect of back-diffusion, further experiments were carried out [37]: P-PEG20000(40) or P-PEGDME(40) membrane was mounted in the membranes cell by using N2, CH4 or CO2 as the feed gas (at 1 bar gage pressure), while N2 (for CH4 and CO2 feed gas) or CH4 (for N2 feed gas) as the sweep gas was kept at atmospheric pressure. The composition of the retentate gas was measured by gas chromatography. The results showed that the sweep gas was not detectable in the retentate, indicating no backdiffusion occurred. To further elucidate the change in membrane separation performances, diffusivity and solubility coefficients of membranes were measured by the well-known “time-lag” method [38] at 25 1C and the high-pressure side was maintained at 2 bar. Before analysis, the membranes were evacuated at least 8 h to remove previously dissolved species. For each membrane the gases were tested in the order of N2, CH4, and CO2.

3. Results and discussion 3.1. Membrane characterization 3.1.1. FESEM Cross section images of the pristine Pebax and hybrid membranes are shown in Fig. 2 and the little white spots indicate the presence of MWCNT. As can be seen, pristine Pebax membrane has a smooth cross section while the hybrid membranes show a rough cross section. For P-CNT(2) membrane (shown in Fig. 2(b)) free of

PEGs, a few micro-sized nanotube agglomerates can be seen, which agrees with the literature [39]. Images from Fig. 2(c)– (h) verify that nanotubes are well dispersed with the aid of PEGs. Because uncharged PEG polymer chains tend to adsorb onto MWCNT, the surface hydrophilicity of MWCNT will increase [40,41]. In this experiment, at least 7 wt% of PEGs in the hybrid membranes is needed to achieve a good dispersion of MWCNT. Triton X-100, a commonly used surfactant, shows better dispersion-improving ability than other PEGs. When the MWCNT content reaches 5 wt%, small agglomerates occur (Fig. 2(f)), but the dispersion is still better than that of P-CNT(2) membrane. Also, no evident interfacial voids in these membranes indicate good compatibility between MWCNT and polymer matrix. 3.1.2. FT-IR The FT-IR spectrum of the prepared membranes is presented in Fig. 3. For pristine Pebax membrane, the characteristic peaks at 1733 and 1102 cm  1 are assigned to C ¼O and C–O stretching vibrations, respectively. Peaks at 1641 and 3300 cm  1 correspond to the H–N–C ¼O and N–H group, respectively. The peak at 2867 cm  1 is assigned to the symmetrical stretching vibration of –CH3, which becomes stronger after the addition of PEGDME (P-PEGDME(40) membrane). The peak at 844 cm  1 represents the stretching vibration of –OH and the incorporation of PEG 20000 increases the intensity (P-PEG20000(40) membrane). The peak at 2867 cm  1, attributed to C–H stretching, gets stronger with the addition of PEG20000 and PEGDME. Decreased intensity of the peaks at 3300 cm  1 and 2867 cm  1 is observed in membranes containing Triton X-100, which may be partly ascribed to the presence of benzene ring. All the membranes show no new absorbance peaks, which indicates the physical blending feature of MWCNT and PEGs within Pebax bulk. Compared with that of MWCNT-free polymeric membranes, the peak at around 1100 cm  1 of the corresponding hybrid membranes shifts to a lower frequency (because of the relative small amount of MWCNT, the shift is small). Analysis of the FT-IR spectrum indicates that MWCNT may affect the C–O and C–H stretching vibrations of EO segment through hydrogen bonding interaction [31]. 3.1.3. XRD XRD patterns with d-spacing values of the membranes are shown in Fig. 4. The pristine Pebax is a semicrystalline copolymer containing both crystalline and amorphous PEO and PA6 phases. The strong and broad peak ranging from 151 to 251may be an integration of the crystalline PA6 phase [42] and other amorphous phases [43]. The incorporation of high molecular weight PEG 20000 increases the crystallinity of PA6 and PEO phases, and other new sharp peaks at 2θ ¼261, 361 and 401 attributed to PEO crystalline phases also emerge. Moreover, the diffraction peak of PA6 at 2θ ¼22.51 shifts to a higher value of 23.21. The enhancement in the intensity of crystalline peaks indicates decreased d-spacing of Pebax (from 3.94̊to 3.80̊ ). PEGDME, as a plasticizer [35], decreases the crystalline peaks at 20.11, 38.51 and 40.11 and shifts the crystalline PA6 peak to a lower value of 22.31, which corresponds to a d-spacing of 3.99̊ . As for membranes containing low molecular weight PEG400 and PEG600, decreased crystallinity is observed. It can also be found that the d-spacing values of membranes containing low molecular weight PEGs varied only slightly. Similar results are also observed in membranes containing PEG400 and PEG600. The addition of 2 wt% MWCNT decreases the crystallinity of all the polymers or polymer blends and the crystallinity decrease of P-PEG20000(40) is more obvious than that of other polymer blends. Meanwhile, the crystalline peaks of MWCNT-containing hybrid membranes shift to lower values

S. Wang et al. / Journal of Membrane Science 460 (2014) 62–70

65

Fig. 2. FESEM images of the cross section of (a) Pebax, (b) P-CNT(2); (c) P-Triton X-100(40)–CNT(2); (d) P-PEG600(40)–CNT(2); (e) P-PEG20000(40)–CNT(2); (f) P-PEG20000 (40)–CNT(5); (g) P-PEGDME(40)–CNT(2) and (h) P-PEGDME(40)–CNT(5).

Transmittance

Table 1 Tg of Pebax, Pebax blend membranes and hybrid membranes.

P-PEG20000(40) P-PEG20000(40)-CNT(2) P-Triton X 100(40) P-Triton X 100(40)-CNT(2) Pebax P-CNT(2) P-PEGDME(40) P-PEGDME(40)-CNT(2)

4000

3500

3000

2000

1500

1000

500

P-PEG20000(40) P-PEG20000(40)-CNT(2) Pebax P-CNT(2) P-Triton X100(40) P-Triton X100(40)-CNT(2) P-PEG600(40) P-PEG400(40) P-PEGDME(40) P-PEGDME(40)-CNT(2)

Intensity(a.u)

d=3.86

d=3.98 d=3.97 d=3.99 d=4.00

10

20

30

2

Membrane

Tg [1C]

Pebax P-PEG20000(40) P-Triton X-100(40) P-PEG400(40) P-PEG600(40) P-PEGDME(40)

 51.4  46.2  55.0  62.8  62.6 –nd

P-CNT(2) P-PEG20000(40)–CNT(2) P-TritonX-100(40)–CNT(2) P-PEG400(40)–CNT(2) P-PEG600(40)–CNT(2) P-PEGDME(40)–CNT(2)

 53.5  49.0  53.7  63.9  64.1 –nd

The glass transition temperature is not detected.

Fig. 3. FT-IR spectra of Pebax, Pebax blend membranes and hybrid membranes.

d=3.94 d=3.98 d=3.96 d=3.98

Tg [1C]

nd

Wave number (cm-1)

d=3.80

Membrane

40

3.1.4. DSC Glass transition temperature of PEO segments is shown in Table 1. As can be seen, the incorporation of high molecular PEG20000 increased the Tg from  51.4 1C to  42.6 1C, which indicated the reduced chain mobility. Low molecular weight PEGs with high chain mobility could act as plasticizer, and increase the chain mobility and decrease Tg values of the blend membranes. Different from Peinemann's work [35], the Tg of PEGDME-containing membranes could not be detected in this study. We suppose the difference in membrane preparation procedure may result in the different self-assembly behaviors of polymer blends especially in membranes containing block copolymers. Decreased Tg is observed in all the hybrid membranes with MWCNT, indicating MWCNT may effectively disrupt the chain packing efficiency. 3.2. Gas separation performances

50

deg

Fig. 4. XRD patterns of Pebax, Pebax blend membranes and hybrid membranes.

compared with those of their polymeric counterparts. These shifts indicate an increase in intersegmental spacing which is anticipated to create more free volume cavities for molecule diffusion.

3.2.1. Pure gas permeation performances Permeability of single gas (CO2, N2 or CH4) was studied for the polymeric and hybrid membranes containing different PEGs at 1 bar and room temperature. The polymeric membranes are Pebax–PEGs blend membranes, of which the PEGs content ranges from 0 to 40 wt %, and hybrid membranes are the polymeric membranes with 2 wt% MWCNT. Fig. 5 shows the single gas permeability and ideal CO2/N2 and CO2/CH4 selectivities of the polymeric (open symbols) and hybrid membranes (solid symbols) as a function of the PEGs content. There is a distinct correlation between CO2 permeability and the content/ molecular weight of PEGs. It can be observed in Fig. 5(a) that high molecular weight PEGs in polymeric membranes decrease CO2 permeability, while low molecular weight PEGs increase the CO2 permeability compared with pristine Pebax membrane. Two extremes are

66

S. Wang et al. / Journal of Membrane Science 460 (2014) 62–70

PEGDME PEG400 PEG600 Triton X-100 PEG2000 PEG10000 PEG20000

400

PEGDME PEG400 PEG600 Triton X-100 PEG2000 PEG10000 PEG20000

6

N2 permeability (Barrer)

CO2 permeability (Barrer)

500

300

200

4

2

100

0

0

0

10

20

30

40

0

10

20

PEG content (wt. %)

40

30

40

120

40 PEGDME PEG400 PEG600 Triton X-100 PEG2000 PEG10000 PEG20000

30

PEGDME PEG400 PEG600 Triton X-100 PEG2000 PEG10000 PEG20000

100

CO2/N2 selectivity

CH4 permeability (Barrer)

30

PEG content (wt. %)

20

10

80

60

40

20 0 0

10

20

30

0

40

10

PEGDME PEG400 PEG600 Triton X-100 PEG2000 PEG10000 PEG20000

30

CO2/CH4 selectivity

20

PEG content (wt. %)

PEG content (wt. %)

20

10 0

10

20

30

40

PEG content (wt. %) Fig. 5. Pure gas (a) CO2 permeability; (b) N2 permeability; (c) CH4 permeability; (d) CO2/N2 ideal selectivity and (e) CO2/CH4 ideal selectivity of polymer blend membranes (open symbols) and hybrid membranes (solid symbols) containing different PEGs at 1 bar feed pressure and room temperature.

found in P-PEG20000(40) ðP CO2 ¼ 35 BarrerÞ and P-PEGDME(40) ðP CO2 ¼ 553 BarrerÞ. The complex phase behaviors of Pebax are found to retard the CO2 capture performance because of the crystalline PEO and PA6 phases. The opposite trends of CO2 permeability can be ascribed to the different micro-phase structures of polymer blends influenced by various PEGs, which is reflected in the XRD pattern. High molecular weight PEGs with lower chain mobility tend to crystallize [10] and once added to the Pebax matrix, they will cause the crystallization of both PA6 and PEO segments (confirmed by XRD pattern of P-PEG20000(40)), thus decrease the fraction of amorphous region for gas transport. The diffusivity and solubility data of certain membranes were measured by using the time-lag method to give further explanations. As shown in Table 2, decreased CO2, CH4 and N2 diffusivities were found in P-PEG20000(40) compared with pristine

Pebax membrane. PEG20000 and PEG10000 are two typical crystalline polymers in this study; the gas permeabilities of membranes containing the same mass ratio of PEG20000 and PEG10000 display no significant difference. However, low molecular weight PEGs with higher chain mobility can be located between Pebax polymer chains and may effectively inhibit crystallization [38]. The increase of fractional free volume takes places in both PEO and PA phases, resulting in a notable increase in gas diffusivity. It should be mentioned that PEGDME, with almost the same PEO content as PEG400 or PEG600, has notable effect on membrane permeability. Because the end group –OCH3 in PEGDME other than –OH in PEGs will not form hydrogen bond between polymer chains [44], which will lead to larger increase in free volume. Moreover, the solubility could be further increased through the incorporation of alkyl end groups [45,46]. However, as for

S. Wang et al. / Journal of Membrane Science 460 (2014) 62–70

67

Table 2 Gas diffusivity and solubility coefficients of the membranes (membranes were tested at 2 bar, 25 1C). Membrane

DCO2 a

SCO2 b

DN2 a

SN 2 b

DCO2 =DN2

SCO2 =SN2

DCH4 a

SCH4 b

DCO2 =DCH4

SCO2 =SCH4

Pebax P-CNT(2) P-PEG20000(40) P-PEG20000(40)–CNT(2) P-PEGDME(40) P-PEGDME(40)–CNT(2) P-PEG2000(40) P-PEG2000(40)–CNT(2) P-PEG400(40) P-PEG400(40)–CNT(2) P-PEG600(40) P-PEG600(40)–CNT(2) P-TX-100(40) P-TX-100(40)–CNT(2)

1.35 1.58 0.96 1.02 3.75 4.02 0.98 1.11 3.45 3.65 3.56 3.64 2.89 3.05

59.36 76.80 24.40 34.29 113.53 138.15 17.87 23.37 71.40 93.53 71.01 88.32 63.57 77.89

0.58 0.82 0.35 0.45 2.38 2.66 0.45 0.52 2.03 2.35 2.12 2.17 1.89 1.88

2.59 2.88 3.58 3.43 2.81 2.53 1.67 1.84 2.72 2.59 2.46 2.53 2.28 2.66

2.33 1.93 2.74 2.27 1.58 1.51 2.17 2.13 1.70 1.55 1.68 1.68 1.53 1.62

22.87 26.68 6.82 10.01 40.39 54.52 10.71 12.71 26.24 36.09 28.88 34.85 27.90 29.28

0.48 0.59 0.18 0.29 1.67 1.85

10.31 11.49 4.45 5.08 19.57 21.28

– – – – – – – –

– – – – – – – –

2.81 2.68 5.49 3.52 2.25 2.17 – – – – – – – –

5.76 6.68 5.49 6.75 5.80 6.49 – – – – – – – –

a b

Diffusivity coefficient [cm2/s]  106. Solubility coefficient [cm3 (STP)/cm3 cmHg]  104.

Table 3 Degree of CO2 permeability increase, CO2/CH4 selectivity decrease and CO2/N2 selectivity increase of hybrid membranes (with 2 wt% MWCNT) compared with the corresponding polymeric membranes (i.e. without MWCNT). Polymeric membrane

P CO2 a

αCO2 =CH4 a

αCO2 =N2 a

P CO2 increase (%)

αCO2 =CH4 decrease (%)

αCO2 =N2 increase (%)

Pebax P-PEGDME(10) P-PEG600(10) P-PEG10000(10) P-PEG20000(10) P-PEG20000(20) P-PEG20000(40)

119.3 7 4.2/88.4 7 2.6 196.7 7 6.1/162.0 7 6.4 179.0 7 3.7/144.9 7 2.2 89.6 7 1.2/66.27 0.8 102.8 7 1.8/67.7 7 2.6 72.2 7 2.4/48.17 0.5 35.07 0.9/23.5 7 0.7

17.6 7 0.3/20.4 7 0.6 15.4 7 0.2/16.5 70.2 18.9 7 0.2/20.487 0.2 17.8 70.1/21.3 7 0.5 17.9 7 0.6/21.7 7 0.2 21.0 7 0.2/25.4 7 0.2 23.7 7 0.6/30.08 7 0.2

51.5 7 0.8/49.47 0.5 62.0 7 1.1/56.22 7 0.7 52.3 7 0.8/48.467 1.2 49.27 3.5/44.6 7 1.9 49.97 2.7/46.17 1.5 32.2 7 0.4/28.2 7 0.8 22.7 7 0.2/18.7 7 0.1

35.01 21.38 23.46 35.42 51.75 49.97 49.28

14.02 6.76 7.79 16.40 17.09 17.10 21.10

4.35 10.34 8.01 10.43 8.19 12.89 21.24

a

The first data are from hybrid membranes and the second data are from polymeric membranes.

Triton X-100, a molecule with bulkier group at one end, is expected to increase the permeability more effectively than PEG400/PEG600 in blend membranes. However, the P-Triton X-100 membranes show lower gas permeability compared with P-PEG400/PEG600; this maybe because the bulky tert-octylphenyl group is not easily inserted to the Pebax chains, thus the miscibility between Pebax and Triton X-100 is weakened. The higher Tg values of P-Triton X-100 (40) than that of P-PEG400(40) may support such hypothesis. The ideal selectivities of CO2/N2 (Fig. 5(d)) and CO2/CH4 (Fig. 5(e)) also show dependence on PEGs molecular weight. The incorporation of high molecular weight PEGs increases CO2/CH4 selectivity and decreases CO2/N2 selectivity; in contrast, the incorporation of low molecular PEGs decreases CO2/CH4 selectivity and increases CO2/N2 selectivity. For P-PEG20000(40) membrane, the CO2/CH4 selectivity increases by 51% compared with pristine Pebax membrane. In this study, it is found that the blending of PEGDME increases CO2/N2 selectivity. This data is on the contrary to the experimental result from one paper of Peinemann's research group, where the blending of PEGDME decreases CO2/N2 selectivity [35]. As is known, multiple phases exist in a block copolymer like Pebax; the membrane preparation process and temperature conditions play significant roles in resultant membrane micro-phase structures [47], which can be also verified by the different XRD patterns of Pebax-1657 in previous studies [39,43]. In Peinemann's work, permeabilities of gases were measured by the constant volume/variable pressure method with absolute feed pressure of 0.3 bar at 30 1C. Whereas, in this study, permeabilities of gases were measured based on the constant pressure/variable volume method with absolute feed pressure of 2 bar at 23 1C. The feed gas pressure and temperature would have notable influence on membrane nanostructures. Moreover, the increase in feed gas pressure and decrease in temperature would effectively facilitate CO2 sorption in membranes. As an illustration, Pebax and

PEG200 blending membranes tested with different feed pressures showed opposite tendency of CO2/N2 selectivity [13,38]. According to the solution–diffusion model, for a given polymeric membrane, the solution coefficient mainly depends on the condensability of gas molecules and the diffusion coefficient mainly depends on the kinetic diameter of gas molecules. Ideal selectivity of the gas pairs is the product of solubility selectivity and diffusivity selectivity. The kinetic diameters of CO2, CH4 and N2 are 0.33 nm, 0.38 nm and 0.364 nm, respectively. Due to the larger size difference and moderate critical temperature difference of CO2/CH4 than that of CO2/N2, it would be effective to separate CO2 and CH4 by diffusivity selectivity. The smaller size difference between CO2 and N2 renders the diffusivity selectivity close to unity [48]. However, the critical temperature of CO2 (304.1 K) is much higher than that of N2 (126.2 K), which means solubility selectivity may be favorable for CO2/N2 separation. In membranes containing high molecular weight PEGs, the decreased chain spacing can effectively elevate the membrane size sieving ability, thus the ideal selectivity for CO2/CH4. As shown in Table 2, with the incorporation of 40 wt% PEG20000 the CO2/CH4 diffusivity selectivity increases from 2.81 to 5.48, whereas the solubility changes only slightly. Adequate amorphous EO units with favorable affinity toward CO2 in membranes should significantly enhance solubility selectivity [42]. However, the addition of high molecular weight PEGs decreases the amount of amorphous EO units and the solubility selectivity for CO2/N2. Higher amorphous fraction and larger chain spacing along with the incorporation of low molecular weight PEGs render the lower diffusivity selectivity for CO2/CH4, while the increased amorphous EO units render the higher solubility selectivity for CO2/N2. With the incorporation of MWCNT, CO2, N2 and CH4 permeabilities of all these membranes increase with varying degrees (Fig. 5, solid

S. Wang et al. / Journal of Membrane Science 460 (2014) 62–70

80

50 20 40 30

CO2/X selectivity

30

60

10

20

CO2 permeability (Barrer)

70 CO2 permeability (Barrer)

120

600

100

500

80 400 60 300 40 200

20 0

100

10

0

1

2 3 4 MWCNT content (wt. %)

CO2 /X selectivity

68

0

5

1

2

3

4

5

MWCNT content (wt. %)

Fig. 6. Effect of MWCNT content on CO2 permeability, CO2/CH4 selectivity and CO2/N2 selectivity of (a) P-PEG20000(40) and (b) P-PEGDME(40) membranes at 2 bar mixed gas feed pressure and room temperature. (The pure gas data are displayed with open symbols.)

40

40 20 30 10 20 10

0 0

4

8

12

16

20

120

900 800

100

700 600

80

500

CO2/N2 selectivity

30

CO2 permeability (Barrer)

50

1000

CO2/CH4 selectivity

CO2 permeability (Barrer)

60

400 300

0

2

Feed pressure (bar)

4

6

8

10

60

Feed gas pressure (bar)

Fig. 7. Effect of mixed gas feed pressure on (a) CO2 permeability and CO2/CH4 selectivity of P-PEG20000(40)–CNT(3) membrane; (b) CO2 permeability and CO2/N2 selectivity of P-PEGDME(40)–CNT(5) membrane (pure gas data shown as open symbols).

100

CO2 /N 2 selectivity

symbols). As shown in Table 2, the CO2 diffusivity and solubility both increase with MWCNT incorporation. XRD patterns have shown that incorporating MWCNT decreases crystallinity of both PA6 and PEO phases and increases the chain mobility, thus increased gas diffusivity is observed compared with the polymeric membranes. Moreover, the increased amorphous PEO regions can further enhance the CO2 solubility. Percentage of CO2 permeability increases for different hybrid membranes compared with their polymeric counterparts as listed in Table 3 in order to acquire more insight into the effect of MWCNT in different polymer blends. The CO2 permeability increases of P-PEGDME(10) and P-PEG600(10) are lower than that of pristine Pebax membrane. As stated above, the low molecular weight PEGs tend to inhibit crystallization of Pebax; the effect of MWCNT on crystallinity decrease is not prominent. However, in polymers with high degree of crystallinity such as P-PEG10000 and P-PEG20000, the effect of MWCNT is more pronounced. The increase of CO2 permeability in P-PEG20000(10)–CNT(2) reaches a high value of 51.7%. With the further increase of PEG20000 content, the CO2 permeability increase would stay at about 50%. Increased CO2/N2 selectivities (Fig. 5(d)) are found in all these hybrid membranes compared with their polymeric counterparts. The major reason attributes to the increased amount of amorphous CO2-philic EO units by MWCNT incorporation [49]. Decreased CO2/N2 diffusivity selectivity is observed in MWCNTcontaining membranes compared with their polymeric counterparts (Table 2). However, the decrease could be offset by the increased solubility selectivity, thus an overall increase in permeability selectivity could be obtained. The increase in CO2/N2 selectivity of P-PEG20000(40) is more significant than that of

10 P-PEGDME membranes P-PEGDME-CNT membranes P-PEGDME-CNT membranes with increased MWCNT content P-PEGDME-CNT membranes under different feed pressure

1 10

100

1000

CO2 permeability (Barrer) Fig. 8. CO2/N2 separation performances of some membranes in Robeson upper bond plot.

other polymer blends (Table 3). Decreased CO2/CH4 selectivities are observed in MWCNT-containing membranes (Fig. 5(e)), and this should be attributed to the enhanced polymer chain mobility. As shown in Table 2, due to the smaller difference in critical temperatures, the CO2/CH4 solubility selectivity increased with increased amorphous PEO content but within a small range. However, the enhanced chain mobility decreases CO2/CH4 diffusivity selectivity and the overall permeability selectivity. As seen in Table 3, the decrease of CO2/CH4 selectivity becomes more obvious

S. Wang et al. / Journal of Membrane Science 460 (2014) 62–70

20

140

16

12

600

8

400 4

120

CO2/N2 selectivity

800

N2 permeability (Barrer)

CO2 permeability (Barrer)

1000

69

100

80

60

200 270

280

290

300

310

320

330

40 270

280

Operating temperature (K)

290

300

310

320

Operating temperature (K)

Fig. 9. Effect of operating temperature on (a) CO2 and N2 permeability and (b) CO2/N2 selectivity of P-PEGDME(40)–CNT(5) membrane at 2 bar mixed gas feed pressure.

for membranes with high crystallinity. The different trends of CO2/ N2 and CO2/CH4 selectivities in hybrid membranes also indicate that polymers with better size sieving ability are more suitable for CO2/CH4 separation while polymers with better solubility selectivity are more suitable for CO2/N2 separation.

display the membrane performances, some of the pure gas data have been encompassed in the Robeson upper bound plot (Fig. 8), which shows the promising prospects of our membranes for CO2/ N2 separation.

3.2.2. Effect of the MWCNT content To further confirm the effect of the MWCNT content on the separation performance of hybrid membranes, increased loading was conducted in P-PEG20000(40) and P-PEGDME (40) membranes for CO2/CH4 (30/70 vol%) and CO2/N2 (20/80 vol%) mixed gas separation, respectively. The feed pressure of mixed gas was set to 2 bar. The pure gas permeation data for membranes containing 2 wt% MWCNT are also included in Fig. 6 and shown with open symbols. No significant difference between the mixed and pure gas performance is observed for the low CO2 partial pressure [33]. Enhanced permeabilities are observed in both membranes containing high molecular PEG20000 and low molecular PEGDDME with the MWCNT content increasing from 1 wt% to 5 wt%. When the MWCNT content reaches to 5 wt%, CO2 permeability increases by 150% for P-PEG20000 membrane. However, decreased CO2/CH4 selectivity by MWCNT loading is found for both kinds of membranes. With the increased MWCNT loading, the increase in CO2 permeability and CO2/N2 selectivity exists in both high molecular and low molecular weight PEGs (Fig. 6(b)). Decreased CO2/CH4 selectivity and increased CO2/N2 selectivity for mixed gas tests by MWCNT loading show good agreement with the data obtained from single gas measurements.

3.2.4. Effect of operating temperature Operating temperature has great influences on the transport of small molecules in membranes. From the previous results, P-PEGDME(40)–CNT(5) demonstrates high CO2 permeability and CO2/N2 selectivity. Therefore, CO2/N2 (20/80 vol%) mixture was applied as the feed gas to study the influence of operating temperature. CO2, N2 permeability and CO2/N2 selectivity data are shown in Fig. 9 with feed gas pressure of 2 bar. As the temperature increases from 275 K to 323 K, both CO2 and N2 permeabilities increase. Because the mobility of CO2 and N2 molecules will increase at elevated temperature, which will enhance the driving force for diffusion. In addition, the increase of temperature will lead to more flexible polymer chains, thus creates more free volume cavities for molecule transport. However, CO2/N2 selectivity displays a decrease trend as temperature increases. This should be attributed to the decreased CO2 solubility at higher temperatures, which decreases the CO2/N2 solubility selectivity and the overall selectivity [50].

3.2.3. Effect of feed gas pressure The influence of feed gas pressure was studied by using P-PEG20000(40)–CNT(3) and P-PEGDME(40)–CNT(5) membranes for CO2/CH4 (30/70 vol%) and CO2/N2 (20/80 vol%) mixed gas separation and shown in Fig. 7. For comparison, the pure gas data for CO2/N2 are also shown as open symbols in Fig. 7(b). From Fig. 7 (a) it can be found that for P-PEG20000(40)–CNT(3) membrane both CO2 permeability and CO2/CH4 selectivity fluctuate in a narrow range. However, the increased feed gas pressure greatly enhances the CO2 permeability and CO2/N2 selectivity in P-PEGD ME(40)–CNT(5) (Fig. 7(b)). This is attributed to the enhanced CO2 sorption at evaluated pressures [50]. The pure gas permeability and selectivity are both larger than those of the mixed gas. In particular, P-PEGDME(40)–CNT(5) membrane shows a high CO2 permeability (743 Barrer) and CO2/N2 mixed gas selectivity (108) at 10 bar, which is above the 2008 upper bound line [4]. To clearly

In the present study, three-component hybrid membranes containing Pebax, PEG-based polymers and MWCNT were prepared and their CO2 capture properties were evaluated. PEG-based polymers were used to improve the MWCNT dispersion within hybrid membranes. MWCNT interacted with the polymer chains and effectively decreased the crystallinity of PA6 and PEO. The incorporation of high molecular weight PEGs led to an increased CO2/CH4 selectivity whereas the incorporation of low molecular weight PEGs led to an increased CO2/N2 selectivity. The incorporation of MWCNT increased the permeability of CO2, N2 and CH4 gases. All the hybrid membranes exhibited decreased CO2/CH4 selectivity for the enhanced chain mobility, while increased CO2/N2 selectivity for increased amorphous PEO content. In particular, P-PEGDME(40)– CNT(5) membrane acquires a CO2 permeability of 743 Barrer and CO2/N2 mixed gas selectivity of 108 at 10 bar and room temperature, which transcends the Robeson upper bound curve as desired.

4. Conclusion

70

S. Wang et al. / Journal of Membrane Science 460 (2014) 62–70

Acknowledgments The authors gratefully acknowledge the financial support from the National High Technology Research and Development Program of China (2012AA03A611), the National Science Fund for Distinguished Young Scholars (No. 21125627), and the Program of Introducing Talents of Discipline to Universities (B06006). References [1] R. Rousseau, Handbook of Separation Process Technology, Wiley Interscience, Paris, 1987. [2] T. Kim, W. Koros, G. Husk, K. O'brien, Relationship between gas separation properties and chemical structure in a series of aromatic polyimides, J. Membr. Sci. 37 (1988) 45–62. [3] C.-L. Lee, H.L. Chapman, M.E. Cifuentes, K.M. Lee, L.D. Merrill, K.L. Ulman, K. Venkataraman, Effects of polymer structure on the gas permeability of silicone membranes, J. Membr. Sci. 38 (1988) 55–70. [4] B.D. Freeman, Basis of permeability/selectivity tradeoff relations in polymeric gas separation membranes, Macromolecules 32 (1999) 375–380. [5] L.M. Robeson, The upper bound revisited, J. Membr. Sci. 320 (2008) 390–400. [6] B.W. Rowe, L.M. Robeson, B.D. Freeman, D.R. Paul, Influence of temperature on the upper bound: theoretical considerations and comparison with experimental results, J. Membr. Sci. 360 (2010) 58–69. [7] V. Shantarovich, I. Kevdina, Y.P. Yampolskii, A.Y. Alentiev, Positron annihilation lifetime study of high and low free volume glassy polymers: effects of free volume sizes on the permeability and permselectivity, Macromolecules 33 (2000) 7453–7466. [8] T.C. Merkel, B.D. Freeman, R.J. Spontak, Z. He, I. Pinnau, P. Meakin, A.J. Hill, Ultrapermeable, reverse-selective nanocomposite membranes, Science 296 (2002) 519–522. [9] W.J. Koros, Barrier Polymers and Structures, American Chemical Society, Washington, DC, 1990. [10] H.Q. Lin, E. Van Wagner, B.D. Freeman, L.G. Toy, R.P. Gupta, Plasticizationenhanced hydrogen purification using polymeric membranes, Science 311 (2006) 639–642. [11] Y. Shen, A.C. Lua, Preparation and characterization of mixed matrix membranes based on PVDF and three inorganic fillers (fumed nonporous silica, zeolite 4A and mesoporous MCM-41) for gas separation, Chem. Eng. J. 192 (2012) 201–210. [12] Y. Zhang, K.J. Balkus, I.H. Musselman, J.P. Ferraris, Mixed-matrix membranes composed of Matrimids and mesoporous ZSM-5 nanoparticles, J. Membr. Sci. 325 (2008) 28–39. [13] S. Kim, E. Marand, High permeability nano-composite membranes based on mesoporous MCM-41 nanoparticles in a polysulfone matrix, Microporous Mesoporous Mater. 114 (2008) 129–136. [14] S. Matteucci, V.A. Kusuma, S.D. Kelman, B.D. Freeman, Gas transport properties of MgO filled poly (1-trimethylsilyl-1-propyne) nanocomposites, Polymer 49 (2008) 1659–1675. [15] A.C. Lua, Y. Shen, Influence of inorganic fillers on the structural and transport properties of mixed matrix membranes, J. Appl. Polym. Sci. 128 (2013) 4058–4066. [16] J.P.G. Villaluenga, M. Khayet, M.A. López-Manchado, J.L. Valentin, B. Seoane, J.I. Mengual, Gas transport properties of polypropylene/clay composite membranes, Eur. Polym. J. 43 (2007) 1132–1143. [17] S. Husain, W.J. Koros, Mixed matrix hollow fiber membranes made with modified HSSZ-13 zeolite in polyetherimide polymer matrix for gas separation, J. Membr. Sci. 288 (2007) 195–207. [18] D.Q. Vu, W.J. Koros, S.J. Miller, Mixed matrix membranes using carbon molecular sieves – I. Preparation and experimental results, J. Membr. Sci. 211 (2003) 311–334. [19] S. Kim, L. Chen, J.K. Johnson, E. Marand, Polysulfone and functionalized carbon nanotube mixed matrix membranes for gas separation: theory and experiment, J. Membr. Sci. 294 (2007) 147–158. [20] A.F. Ismail, P.S. Goh, S.M. Sanip, M. Aziz, Transport and separation properties of carbon nanotube-mixed matrix membrane, Sep. Purif. Technol. 70 (2009) 12–26. [21] M.M. Khan, V. Filiz, G. Bengtson, S. Shishatskiy, M.M. Rahman, J. Lillepaerg, V. Abetz, Enhanced gas permeability by fabricating mixed matrix membranes of functionalized multiwalled carbon nanotubes and polymers of intrinsic microporosity (PIM), J. Membr. Sci. 436 (2013) 109–120. [22] S. Bose, R.A. Khare, P. Moldenaers, Assessing the strengths and weaknesses of various types of pre-treatments of carbon nanotubes on the properties of polymer/carbon nanotubes composites: a critical review, Polymer 51 (2010) 975–993. [23] G. Dong, H. Li, V. Chen, Challenges and opportunities for mixed-matrix membranes for gas separation, J. Mater. Chem. A 1 (2013) 4610–4630.

[24] S. Majeed, V. Filiz, S. Shishatskiy, J. Wind, C. Abetz, V. Abetz, Pyrene–POSS nanohybrid as a dispersant for carbon nanotubes in solvents of various polarities: its synthesis and application in the preparation of a composite membrane, Nanoscale Res. Lett. 7 (2012) 1–11. [25] A.F. Ismail, N.H. Rahim, A. Mustafa, T. Matsuura, B.C. Ng, S. Abdullah, S.A. Hashemifard, Gas separation performance of polyethersulfone/multiwalled carbon nanotubes mixed matrix membranes, Sep. Purif. Technol. 80 (2011) 20–31. [26] M.A. Aroon, A.F. Ismail, M.M. Montazer-Rahmati, T. Matsuura, Effect of chitosan as a functionalization agent on the performance and separation properties of polyimide/multi-walled carbon nanotubes mixed matrix flat sheet membranes, J. Membr. Sci. 364 (2010) 309–317. [27] M.A. Aroon, A.F. Ismail, T. Matsuura, M.M. Montazer-Rahmati, Performance studies of mixed matrix membranes for gas separation: a review, Sep. Purif. Technol. 75 (2010) 229–242. [28] R. Mahajan, W.J. Koros, Factors controlling successful formation of mixedmatrix gas separation materials, Ind. Eng. Chem. Res. 39 (2000) 2692–2696. [29] T.T. Moore, W.J. Koros, Non-ideal effects in organic–inorganic materials for gas separation membranes, J. Mol. Struct. 739 (2005) 87–98. [30] D. Şen, H. Kalıpçılar, L. Yilmaz, Development of polycarbonate based zeolite 4A filled mixed matrix gas separation membranes, J. Membr. Sci. 303 (2007) 194–203. [31] X. Zheng, Q. Xu, Comparison study of morphology and crystallization behavior of polyethylene and poly (ethylene oxide) on single-walled carbon nanotubes, J. Phys. Chem. B 114 (2010) 9435–9444. [32] H.Q. Lin, B.D. Freeman, Materials selection guidelines for membranes that remove CO2 from gas mixtures, J. Mol. Struct. 739 (2005) 57–74. [33] S.R. Reijerkerk, M.H. Knoef, K. Nijmeijer, M. Wessling, Poly(ethylene glycol) and poly(dimethyl siloxane): combining their advantages into efficient CO2 gas separation membranes, J. Membr. Sci. 352 (2010) 126–135. [34] A. Car, C. Stropnik, W. Yave, K. Peinemann, Pebaxs/polyethylene glycol blend thin film composite membranes for CO2 separation: performance with mixed gases, Sep. Purif. Technol. 62 (2008) 110–117. [35] W. Yave, A. Car, K.V. Peinemann, Nanostructured membrane material designed for carbon dioxide separation, J. Membr. Sci. 350 (2010) 124–129. [36] Y. Li, S. Wang, H. Wu, J. Wang, Z. Jiang, Bioadhesion-inspired polymer– inorganic nanohybrid membranes with enhanced CO2 capture properties, J. Mater. Chem. 22 (2012) 19617–19620. [37] X. Yu, Z. Wang, Z. Wei, S. Yuan, J. Zhao, J. Wang, S. Wang, Novel tertiary amino containing thin film composite membranes prepared by interfacial polymerization for CO2 capture, J. Membr. Sci. 362 (2010) 265–278. [38] A. Car, C. Stropnik, W. Yave, K.-V. Peinemann, PEG modified poly(amide-bethylene oxide) membranes for CO2 separation, J. Membr. Sci. 307 (2008) 88–95. [39] R.S. Murali, S. Sridhar, T. Sankarshana, Y. Ravikumar, Gas permeation behavior of Pebax-1657 nanocomposite membrane incorporated with multiwalled carbon nanotubes, Ind. Eng. Chem. Res. 49 (2010) 6530–6538. [40] S. Nuriel, L. Liu, A.H. Barber, H.D. Wagner, Direct measurement of multiwall nanotube surface tension, Chem. Phys. Lett. 404 (2005) 263–266. [41] L. Vaisman, G. Marom, H.D. Wagner, Dispersions of surface-modified carbon nanotubes in water-soluble and water-insoluble polymers, Adv. Funct. Mater. 16 (2006) 357–363. [42] V. Bondar, B. Freeman, I. Pinnau, Gas sorption and characterization of poly (ether-b-amide) segmented block copolymers, J. Polym. Sci., Part B: Polym. Phys. 37 (1999) 2463–2475. [43] S. Sridhar, R. Suryamurali, B. Smitha, T. Aminabhavi, Development of crosslinked poly (ether-block-amide) membrane for CO2/CH4 separation, Colloids Surf. A 297 (2007) 267–274. [44] W.-C. Lai, W.-B. Liau, T.-T. Lin, The effect of end groups of PEG on the crystallization behaviors of binary crystalline polymer blends PEG/PLLA, Polymer 45 (2004) 3073–3080. [45] W. Yave, A. Car, S.S. Funari, S.P. Nunes, K.-V. Peinemann, CO2-philic polymer membrane with extremely high separation performance, Macromolecules 43 (2010) 326–333. [46] M.K. Barillas, R.M. Enick, M. O'Brien, R. Perry, D.R. Luebke, B.D. Morreale, The CO2 permeability and mixed gas CO2/H2 selectivity of membranes composed of CO2-philic polymers, J. Membr. Sci. 372 (2011) 29–39. [47] S.L. Liu, L. Shao, M.L. Chua, C.H. Lau, H. Wang, S. Quan, Recent progress in the design of advanced PEO-containing membranes for CO2 removal, Prog. Polym. Sci. 38 (2013) 1089–1120. [48] K. Ramasubramanian, Y. Zhao, W.S. Winston Ho, CO2 capture and H2 purification: prospects for CO2-selective membrane processes, AIChE J. 59 (2013) 1033–1045. [49] J.H. Kim, S.Y. Ha, Y.M. Lee, Gas permeation of poly (amide-6-b-ethylene oxide) copolymer, J. Membr. Sci. 190 (2001) 179–193. [50] H.Q. Lin, B.D. Freeman, S. Kalakkunnath, D.S. Kalika, Effect of copolymer composition, temperature, and carbon dioxide fugacity on pure- and mixedgas permeability in poly(ethylene glycol)-based materials: free volume interpretation, J. Membr. Sci. 291 (2007) 131–139.