Carbon Nanotube Networks as Nano-Scaffolds for Fabricating

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE

Surfaces, Interfaces, and Applications

Carbon Nanotube Networks as Nano-Scaffolds for Fabricating Ultrathin Carbon Molecular Sieve Membranes Jue Hou, Huacheng Zhang, Yaoxin Hu, Xingya Li, Xiaofang Chen, Seungju Kim, Yuqi Wang, George P. Simon, and Huanting Wang ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b04481 • Publication Date (Web): 29 May 2018 Downloaded from http://pubs.acs.org on May 29, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Carbon Nanotube Networks as Nano-Scaffolds for Fabricating Ultrathin Carbon Molecular Sieve Membranes Jue Hou,1 Huacheng Zhang,1 Yaoxin Hu,1 Xingya Li,1 Xiaofang Chen,1 Seungju Kim,1 Yuqi Wang,1 George P. Simon,2 and Huanting Wang1*

1. Department of Chemical Engineering, Monash University, Clayton, Victoria 3800, Australia. 2. Department of Materials Science and Engineering, Monash University, Clayton, Victoria 3800, Australia

KEYWORDS: carbon molecular sieve membrane, gas separation, carbon nanotube, nanoscaffold, membrane preparation

ACS Paragon Plus Environment

1

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 21

ABSTRACT: Carbon molecular sieve (CMS) membranes have shown great potential for gas separation owing to their low cost, good chemical stability, and high selectivity. However, most of conventional CMS membranes exhibit low gas permeance due to their thick active layer, which limits their practical application. Herein, we report a new strategy for fabricating CMS membranes with a hundred-nanometer-thick ultrathin active layer using poly(furfuryl alcohol) (PFA) as a carbon precursor and carbon nanotubes (CNTs) as nano-saffolds. CNT networks are deposited on a porous substrate as nano-scaffolds, which guide PFA solution to effectively spread over the substrate and form a continuous layer, minimizing penetration of PFA into the pores of the substrate. After pyrolysis process, the CMS membranes with 100 to 1000 nm thick active layer can be obtained by adjusting the CNT loading. The 322 nm-thick CMS membrane exhibits the best trade-off between gas permeance and selectivity, and a H2 permeance of 4.55 × 10−8 mol m−2 s−1 Pa−1, O2 permeance of 2.1 × 10−9 mol m−2 s−1 Pa−1 and O2/N2 ideal selectivity is 10.5, which indicates the high quality of the membrane produced by this method. This work provides a simple, efficient strategy for fabricating ultrathin CMS membranes with high selectivity and improved gas flux.


2

Page 3 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1. Introduction Membrane technologies have shown great significance in gas separation due to their low cost, reliability and high energy efficiency compared with traditional separation processes, such as cryogenic distillation and pressure swing adsorption.1-6 Many carbon-based membranes have been prepared for separation applications, and they usually exhibit better chemical and thermal stabilities than polymeric membranes.7-9 Carbon molecular sieve (CMS) is a kind of microporous carbon material, which can be prepared by pyrolyzing polymers such as poly(vinylidene chloride) (PVDC),10 poly(furfuryl alcohol) (PFA),11, 12 polyacrylonitrile (PAN),13 and polyimide (PI).14-17 The CMS membranes have highly tunable transport properties which are suitable for separation of many gases18,

19

and liquids.20,

21

Like ceramic membranes, the CMS membranes can be

prepared as either non-supported or supported membranes. The non-supported CMS membranes are usually prepared by carbonizing polymeric flat membranes or hollow fibers directly;22-24 however, their mechanical strengths are low and the selection of polymeric materials is limited. The supported CMS membranes are normally composed of a thin CMS active layer and a porous substrate such as porous stainless steel and porous ceramics.25-27 These substrates provide excellent mechanical strength and stability for CMS membranes for practical gas separations at high feed pressures. Furthermore, a high membrane flux/permeance is needed for many separation processes. A high-flux membrane with a suitable selectivity is sometimes more useful than a low-flux membrane with ultrahigh selectivity. Therefore, supported CMS membranes are more suitable for industrial applications because the flux of such supported membranes can be readily improved by decreasing the thickness of active layer. However, constructing a hundrednanometer-thick CMS membrane without defects remains a challenge, despite significant research effort being made in the past decades. The difficulty in preparing ultrathin, defect-free


3


Page 4 of 21

CMS membranes can be explained as follows. Firstly, during carbonization the polymer layer shrinks on the substrate, easily generating defects. Therefore, the preparation of the CMS membrane usually requires multiple coating and pyrolysis processes, which inevitably results in a micrometer-thick active layer.28, 29 Secondly, the properties of the porous support affect the formation of an ultrathin and dense active layer. The polymer solution tends to wet the porous substrate and penetrate into the pores of the support, which increases the actual thickness of active layer. To solve this problem, a thin and dense intermediate layer with extremely small pores, such as a γ-Al2O3 layer, can be fabricated on the surface of porous α-Al2O3 support before deposition of a polymer layer.30 however, this would add a significant cost to CMS membrane fabrication process. Herein, we report on a novel strategy for fabricating CMS membranes with ultrathin active layers by forming carbon nanotube (CNT) network nano-scaffold on a porous substrate to guide polymer solution to spread over the surface of the porous substrate and thus minimize the penetration of polymer solution into the substrate. To demonstrate this strategy, a commonly used CMS precursor, PFA is chosen as precursor; an anodized aluminum oxide (AAO) substrate with straight nanoporous channels is used to easily observe the formation of ultrathin CMS layer and polymer penetration. . The thickness of CMS/CNT hybrid active layer ranging from 100 to 1000 nm can be prepared and controlled by varying the CNT loadings. The resulting CMS membrane shows excellent molecular sieving properties. Particularly, a 322 nm-thick PFAderived CMS membrane exhibits the best trade-off between the O2 flux and the O2/N2 selectivity, and its gas permeance reachs 4.55 × 10−8 mol m−2 s−1 Pa−1 for H2 and 2.1 × 10−9 mol m−2 s−1 Pa−1 at an ideal selectivity of O2/N2 = 10.5. The strategy reported in this work is simple and effective


4

Page 5 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


for the fabrication of supported membranes with an ultrathin CMS active layer for potential practical applications.

2. Results and Discussion Figure 1a illustrates the preparation of the ultrathin CMS/CNT membrane on the AAO substrate. Firstly, the CNTs (dispersed in ethanol, 5 µg/mL) were vacuum filtrated on the AAO substrate. By varying the volume of the CNT solution vacuumed on AAO, CNT networks with different loadings could be prepared. Then, a 30 wt% PFA acetone solution was spin-coated (1000 rpm) on the surface. When the PFA was completely dried, the membrane was pyrolyzed at 600ºC under argon atmosphere. To ensure the formation of CMS/CNT hybrid membrane with minimal defects, the spin coating and carbonization processes were repeated. Figure 1b-d show the scanning electron microscopy (SEM) images of bare AAO support of a 100 nm pore diameter, and CNT covered AAO substrates coated with one PFA coating before and after carbonization, respectively. Note that the loading of CNT was 6.4 × 10−8 mg/cm2. The CNTs were uniformly deposited on the surface of AAO, the resulting network showed good mechanical stability (Figure S1). After the CNT-deposited AAO support was spin-coated with PFA once, the CNT network was clearly visible (Figure 1c); after carbonization of PFA, the CNT network was exposed, indicating the amount of PFA was not enough to form a continuous CMS layer.

. After the second PFA coating and carbonization was conducted, a uniform

CMS/CNT layer was successfully fabricated on the top of the AAO, which was seen from SEM images of the surface and cross section of CMS/CNT/AAO membrane (Figures 1e and 1f). The surface of this membrane was flat and uniform without visible defects. The average thickness of this CMS/CNT layer on the AAO support was 322 ± 6.3 nm. TEM image also confirmed that


5


Page 6 of 21

CNTs were well embedded inside the CMS layer (Figure 1g). ATR-FTIR spectra proved the formation of CMS/CNT hybrid membrane (Figure 1h). The peaks of CMS/CNT/AAO at 2918 cm−1 and 2850 cm−1 arose from the C-H stretch of CNT network; and the peaks at 3240 cm−1, 1660 cm−1, 1420 cm−1 and 1060 cm−1 assigned to PFA structure significantly decreased compared with PFA/CNT/AAO because the furan ring vibrations became weak after carbonization. A slight peak shift was shown in on the XRD patterns of those samples (Figure S2). Thermogravimetric analysis (TGA) results showed the CMS/CNT hybrid membrane was as stable as CMS itself at 600ºC (Figure S3).

Figure 1. a) Schematic illustration of the fabrication process of carbon molecular sieve/carbon nanotube (CMS/CNT) hybrid membrane on AAO substrate. b) SEM images of bare AAO substrate, c) single coated PFA/CNT/AAO membrane, d) CMS/CNT/AAO membrane prepared with single PFA coating and carbonization, and e) CMS/CNT/AAO membrane prepared wit


6

Page 7 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


double FA coating and carbonization. f) Cross-sectional SEM image of double coated andcarbonized CMS/CNT/AAO membrane. g) TEM image of CNT network in PFA-derived CMS. h) ATR-FTIR spectra of bare AAO (orange), CNT/AAO (red), single coated PFA/CNT/AAO (green) and double coated and carbonized CMS/CNT/AAO (blue) membranes.

To prove the key role of CNT network in the preparation of continuous ultrathin CMS layer, a bare AAO substrate without CNT network deposited was coated by PFA and carbonized twice for comparison (Figure 2). For the AAO with CNT network, the PFA solution mostly spread over the substrate surface without significantly penetrating into the AAO channels; so after carbonization, the AAO channels were very clean without carbon deposit and a continuous CMS/CNT layer was obtained (Figure 2a-c). By contrast, most PFA solution flowed into the AAO channels after being spread on the bare AAO. As a result, it was hard to get a continuous CMS layer after carbonization, and most CMS remained in the channels, forming irregular blocks and defects (Figure 2d-f); such a membrane showed no gas selectivity. The differences in the CMS layer formation on the AAO with/without CNT network were also clearly observed by EDX mapping (Figure 2c, 2f and S4). A continuous carbon layer with a clear interface with the AAO was observed on the cross-sectional EDX mapping image of CMS/CNT/AAO membrane; whereas the carbon was formed inside the AAO channels without CNT network. The role of the CNT network becomes more significant when the substrates with larger pore size and higher porosity were used. As the pore size increased, forming a thin continuous PFA layer on the substrate with the CNT network became more difficult even after double coating and carbonization processes, resulting in less CMS and more membrane defects (Figure S5). In contrast, a continuous CMS/CNT layer was readily obtained on a 200 nm AAO substrate with a


7


Page 8 of 21

much higher porosity with the help of CNT network (Figure S6). The SEM image showed that a dense CMS layer was formed one the area with CNT network deposited, and no continuous CMS layer was seen on the area without CNT deposited, further demonstrating the important role of CNT network in the CMS membrane formation (Figure S7). Therefore, our strategy of using CNT network would be potentially useful for fabrication of thin CMS membrane on low-cost porous substrates with non-uniform pores and rough surface.

Figure 2. The key role of CNT network in the preparation of ultrathin, continuous CMS membrane. a,d) Schematic illustration of spreading of PFA on AAO substrate (a) with or (d) without CNT network . The blue regions illustrate PFA. With the help of CNTs, PFA spreads over the substrate surface along the CNT network and forms a countinuous layer; while without CNT network, PFA penetrates into the AAO channels, resulting in irregular structures. b) Crosssectional and top-view (inset) SEM images and (c) EDX mapping image of CMS/CNT hybrid membrane with a continuous layer on the AAO substrate. Due to the dense CMS layer, CNT network and AAO can not be seen from the top-view SEM image. e) Cross-sectional and topview (inset) SEM images and (f) EDX mapping image of the carbonized PFA-coated AAO


8

Page 9 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


without CNT network. Porous structure of AAO can be seen from the top-view SEM image due to the defects after carbonization.

The concentration of PFA solution also affects the quality of the CMS/CNT membrane. PFA solutions with different concentrations (40, 30, 20, 10, 5 wt% in acetone) were spin-coated on the AAO surfaces coated with CNT network (Figure S8). As the PFA concentration decreased, more defects appeared on the surface. A 30 wt% PFA solution was used in the following experiments because the 40 wt% PFA solution was too viscous. The PFA volume decrease during the carbonization process led to more defects and the appearance of the CNT network in the top-view SEM images (Figure S8). The PFA coating and carbonization process was repeated again in order to reduce defects. After the second coating and pyrolysis, the CNT network was not observed in SEM images and only flat surfaces without obvious defects were obtained (Figure S10). The PFA-derived CMS/CNT membranes with ultrathin and uniform active layer have excellent gas selectivity. The single-gas permeation properties of the CMS/CNT membranes were investigated for H2, CO2, O2, N2, and CH4 gases by the constant-volume/variable-pressure method. Figure 3a shows the permeance of the CMS/CNT membrane with 6.4 × 10−8 mg/cm2 CNTs carbonized at 500ºC as a function of the kinetic diameter of gas molecule at 25ºC and 1 bar. The permeance clearly depended on the gas molecular size. The H2 permeance of 1.89 × 10−8 mol m−2 s−1 Pa−1 was the highest compared with other gases (Table S1). There was a clear cut-off between O2 and N2 with an ideal selectivity of 13.4, which strongly demonstrated that the CMS/CNT membrane produced by our method possessed a high-quality active layer. The ideal


9


Page 10 of 21

selectivities of this membrane for H2/N2, H2/CH4, H2/CO2, CO2/N2, and O2/N2 were 141.2, 264.3, 5.9, 23.8, and 13.4, respectively, much greater than Knudsen selectivities (Figure 3b).

Figure 3. The gas separation performance of the CMS/CNT hybrid membrane with 6.4 × 10−8 mg/cm2 CNTs. a) Permeance of H2, CO2, O2, N2, CH4 for the CMS/CNT membrane carbonized at 500ºC. b) Ideal selectivities (blue bars) of CMS/CNT membrane carbonized at 500℃ and Knudsen selectivities (red bars) for H2/N2, H2/CH4, H2/CO2, CO2/N2, and O2/N2.

The loading of CNT network not only affects the quality and thickness of the active layer in CMS/CNT membrane, but also determines the gas separation performance. To optimize the gas separation performance of CMS/CNT membrane, we systematically investigated the permeance and selectivity of CMS/CNT membranes with different CNT’s loadings (Figure 4a,b). Different


10

Page 11 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


volumes of CNT solution were vacuum-filtrated on the AAO substrates, followed by double spin-coating PFA and carbonization at 500℃. The H2 permeance increased when the CNT’s loading increased from 3.2 × 10−8 mg/cm2 to 6.4 × 10−8 mg/cm2, and then gradually decreased as the CNT’s loading continued to increase (Figure 4a). The gas flux reached its highest value when the CNT’s loading was 6.4 × 10−8 mg/cm2. For the CMS/CNT membrane with CNTs only 3.2 × 10−8 mg/cm2, the CNT network was too sparse to support PFA solution, so some PFA still penetrated into the AAO channels and formed CMS blocks after carbonization, which could be clearly seen in the cross-sectional SEM image (Figure S9) and caused large flux decrease. As the CNT’s loading increased, less PFA penetrated into the AAO channels, while more PFA was supported by CNT network and then remained on the top of the substrate during spin-coating. As a result, the thickness of the active layer increased (Figure S10) and the gas flux decreased accordingly. The ideal selectivity of H2/N2 increased from 140 to 170 when the CNT’s loading was lower than 9.6 × 10−8 mg/cm2; and it sharply rose to 400~500 when the CNT’s loading was more than 12.8 × 10−8 mg/cm2. The trend of O2 permeance was similar to N2 permeance and reached the maximum when the CNT loading was 6.4 × 10−8 mg/cm2. The ideal selectivity of O2/N2 reached the maximum 13.4 when the CNT loading was 6.4 × 10−8 mg/cm2 and then gradually decreased as the CNT’s loading increased (Figure 4b). So the CNT loading of 6.4 × 10−8 mg/cm2 was chosen in following experiments because it had a good balance of flux and selectivity. The carbonization temperature also significantly affects the performance of CMS/CNT hybrid membrane. The gas permeances of CMS/CNT hybrid membranes with a CNT loading of 6.4 × 10−8 mg/cm2 increased as the temperature was raised (Figure 4c, Table S2), while the ideal selectivity varied (Figure 4d). When the carbonization temperature increased from 500ºC


11


Page 12 of 21

to 600ºC, the permeances of H2, O2, N2 increased from 1.89 × 10−8 mol m−2 s−1 Pa−1, 1.80 × 10−9 mol m−2 s−1 Pa−1, 1.35 × 10−10 mol m−2 s−1 Pa−1 to 4.56 × 10−8 mol m−2 s−1 Pa−1, 2.10 × 10−9 mol m−2 s−1 Pa−1, 2.01 × 10−10 mol m−2 s−1 Pa−1, respectively. The H2 flux of CMS/CNT membrane carbonized at 600ºC was about 2.4 times than the one carbonized at 500ºC. The fluxes of O2 and N2 also improved and the ideal selectivity of O2/N2 only slightly decreased and kept as high as 10.5.

Figure 4. Gas performances of CMS/CNT hybrid membranes. a) Permeance of H2, N2 and ideal selectivity of H2/N2 of CMS/CNT membranes with different loadings of CNT prepared by double coating and carbonization at 500ºC. b) Permeance of O2, N2 and ideal selectivity of O2/N2 of CMS/CNT membranes with different loadings of CNT prepared by double coating and carbonization at 500ºC. c) Gas permeance of CMS/CNT membranes carbonized at different


12

Page 13 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


temperatures. d) Ideal selectivities of H2/N2, H2/CH4, H2/CO2, CO2/N2, and O2/N2 of CMS/CNT membranes carbonized at 400ºC, 500ºC, and 600ºC, respectively.

The effect of carbonization temperature on gas permeation can be explained by the change of CMS porous structure with carbonization temperature. It is known that the molecular structure and pore size distribution of PFA-derived carbon depend on carbonization conditions.

12

When

the pyrolysis temperature increases from 400ºC to 600ºC, the polymer chains further crosslink and transfer into conjugated structures, and microporous structure in PFA-derived carbon is formed gradually. As shown in Figure 4c, the permeance for all kinds of gases of the membrane carbonized at 400ºC, 500ºC and 600ºC gradually increase means that the porosity and pore size of CMS increases. The O2/N2 selectivity of the CMS membrane carbonized at 600ºC is lower than the one at 500ºC because the pore size of the former is slightly greater than that of the latter. Importantly the extent of permeance increase is different for different gases: the smaller the gas molecule, the larger the permeance. For some gas pairs such as H2/N2 and H2/CH4, their selectivity initially decreases and then increases, when the carbonization temperature increases from 400 to 500ºC, and then to 600ºC. This result indicates that the pore size increases more significantly initially (400 to 500ºC) and then the increase in porosity (number of pores) becomes dominant (500 to 600ºC) during carbonization. The CMS/CNT membranes in this work were compared with the previously reported CMS membranes derived from PFA (Figure 5 and Table S3).25, 26, 28, 31-33 The CMS/CNT membranes (carbonized at 500ºC and 600ºC) in this work demonstrated an improved trade-off between the permeance and selectivity in O2/N2 separation, which had the highest O2 flux among the CMS membranes. Due to the assistance of the CNT network, the preparation of CMS required only


13


Page 14 of 21

two times of polymer solution coating and carbonization process to obtain high-quality membranes, compared with many of those membranes reported in literature that required coating and pyrolyzing at least 3 times. As discussed, this strategy also worked on the substrates with higher porosity (e.g. 200 nm AAO disk, Figure S7). Therefore, although there is a lot of research needed to develop and scale up CMS membranes for practical application, our new strategy provides an alternative technique for controlling deposition of carbon precursor and for ensuring better structural integrity during carbonization.

Figure 5. Comparison of the ideal O2/N2 selectivity of CMS/CNT membrane with the previously reported PFA-derived CMS membranes. Detailed data are shown in Table S3 (ESI).

3. Conclusion


14

Page 15 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


In this work, by constructing CNT networks on the porous substrates as nano-scaffolds and adjusting the loading of CNT, CMS/CNT hybrid membranes were successfully prepared with a controllable active layer thicknesses ranging from 100 to 1000 nm. The CMS/CNT membrane with 322 ± 6.3 nm-thick active layer exhibited a better trade-off between the permeance and selectivity, whose O2 flux was 2.1 × 10−9 mol m−2 s−1 Pa−1 and O2/N2 ideal selectivity was greater than 10. This work provides an attractive strategy to effectively prepare an ultrathin CMS active layer on various substrates to improve membrane flux whilst retaining its high selectivity. This strategy is suited to the preparation of CMS membranes on substrates with large pores and porosity for further flux improvement for gas separation and many other potential applications such as desalination, osmotic power, and acid separation.

4. Experimental Section Preparation of CMS/CNT/AAO membrane: CNT was deposited on AAO substrate (pore size: 100 nm or 200 nm) by vacuum filtration of 5 µg/mL CNT ethanol solution with a different volume on the surface. After the CNT/AAO was dried at 60ºC for 2 h, PFA acetone solution (30 wt%) was spin-coated on the surface and dried at 60ºC for another 2 h to evaporate the acetone. When the acetone was completely dried, the PFA coated CNT/AAO membrane was carbonized at 500ºC in argon atmosphere. The double coated CMS/CNT/AAO was obtained by coating PFA and carbonizing the membrane again. Characterization: Scanning electron microscopy (SEM) images and EDX mapping images were taken with a field-emission scanning electron microscope (FEI Nova NanoSEM 420 and 450). Transmission electron microscopy (TEM) image was taken by FEI Tecnai G2 T20 operated at an accelerating voltage of 200 kV. X-ray diffraction (XRD) patterns were recorded in


15


Page 16 of 21

the 2θ range of 2-60° at room temperature using a Miniflex 600 diffractometer (Rigaku, Japan) in transmission geometry using Cu Kα radiation (15 mA, 40 kV) at a scan rate of 2°/min and a step size of 0.02°. Thermogravimetric analysis (TGA) measurements were performed on a flow of air with a ramp of 10ºC/min from room temperature to 700ºC. Attenuated Total Reflection Fourier Transform Infrared (ATR-FTIR) spectra of the samples were taken by an attenuated total reflectance ATR-FTIR (PerkinElmer, U.S.A.) at an average of 20 scans with a resolution of 2 cm–1. Gas permeation tests: The single gas permeance of membranes was measured using a constantvolume/variable-pressure apparatus using the same method described by Yao et al.34, 35 First, a piece of flat membrane was fixed on a stainless steel porous sample holder by a Varian Torr Seal vacuum sealant. Then the holder was placed inside a larger Pyrex tube with a feed gas flowing through, and the end connected to a MKS 628B Baratron pressure transducer and a vacuum pump. The gas permeance experiments were performed using the steady-state gases H2, CO2, O2, N2, CH4, at room temperature (25ºC). For each single gas measurement, the permeate side of the membrane was degassed under vacuum for at least half an hour, allowing enough time to reach steady-state permeation condition. At least three samples were prepared under the same conditions for gas permeation measurements. The data presented in this work were the average value with ±10% error. The molar flow rate (Ni) of the permeating gas was calculated from the linear pressure rise, and its coefficient was calibrated using a digital flowmeter (ADM2000, Agilent, California, USA). The feed gas is supplied at room temperature (25ºC) under atmospheric pressure (1 bar). The permeate side in a vacuum condition, providing a driving force for permeation. The effective


16

Page 17 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


membrane area was measured. Membrane permeance, Pi (mol m−2 s−1 Pa−1), is defined as equation (1).

Pi =

Ni

∆pi ⋅ A

(1)

where Ni (mol·s-1) is the molar flow rate of component i, ∆pi (Pa) is the transmembrane pressure difference of component i, and A (m2) is the effective membrane area for testing. The ideal selectivity Si/j is calculated from the relation between the permeance of component i and component j.

Si / j =

Pi

Pj

(2)

ASSOCIATED CONTENT Supporting Information. The following files are available free of charge. Figures showing SEM images, XRD, TGA of the CMS membranes. Tables showing comparisons of the CMS/CNT membranes and the pervious reported PFA-derived CMS membranes. (PDF) AUTHOR INFORMATION Corresponding Author Email: [email protected] Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes


17


Page 18 of 21

The authors declare no competing financial interest. ACKNOWLEDGMENT This work is supported by the Australian Research Council (Discovery project DP 170102964). The authors thank the staff of the Monash Centre for Electron Microscopy for their technical assistance with SEM and TEM. This study was partly performed at the Melbourne Centre for Nanofabrication (MCN) in the Victorian Node of the Australian National Fabrication Facility (ANFF). REFERENCES 1. Snyder, M. A.; Tsapatsis, M., Hierarchical nanomanufacturing: from shaped zeolite nanoparticles to high-performance separation membranes. Angew. Chem. Int. Ed. 2007, 46, 7560-7573. 2. Gin, D. L.; Noble, R. D., Chemistry. Designing the next generation of chemical separation membranes. Science 2011, 332, 674-676. 3. Yao, J.; Wang, H., Zeolitic imidazolate framework composite membranes and thin films: synthesis and applications. Chem. Soc. Rev. 2014, 43, 4470-4793. 4. Rangnekar, N.; Mittal, N.; Elyassi, B.; Caro, J.; Tsapatsis, M., Zeolite membranes - a review and comparison with MOFs. Chem. Soc. Rev. 2015, 44, 7128-7154. 5. Park, H. B.; Kamcev, J.; Robeson, L. M.; Elimelech, M.; Freeman, B. D., Maximizing the right stuff: The trade-off between membrane permeability and selectivity. Science 2017, 356, eaab0530. 6. Koros, W. J.; Zhang, C., Materials for next-generation molecularly selective synthetic membranes. Nat. Mater. 2017, 16, 289-297. 7. Sakintuna, B.; Yürüm, Y., Templated Porous Carbons: A Review Article. Ind. Eng. Chem. Res. 2005, 44, 2893-2902. 8. Paul, D. R., Creating new types of carbon-based membranes. Science 2012, 335, 413-414. 9. Liu, G.; Jin, W.; Xu, N., Two-Dimensional-Material Membranes: A New Family of High-Performance Separation Membranes. Angew. Chem. Int. Ed. 2016, 55, 13384-13397. 10. Centeno, T. A.; Fuertes, A. B., Carbon molecular sieve gas separation membranes based on poly(vinylidene chloride-co-vinyl chloride). Carbon 2000, 38, 1067-1073. 11. Lafyatis, D. S.; Tung, J.; Foley, H. C., Poly(furfuryl alcohol)-derived carbon molecular sieves: dependence of adsorptive properties on carbonization temperature, time, and poly(ethylene glycol) additives. Ind. Eng. Chem. Res. 1991, 30, 865-873. 12. Wang, H.; Yao, J., Use of Poly(furfuryl alcohol) in the Fabrication of Nanostructured Carbons and Nanocomposites. Ind. Eng. Chem. Res. 2006, 45, 6393-6404.


18

Page 19 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


13. David, L. I. B.; Ismail, A. F., Influence of the thermastabilization process and soak time during pyrolysis process on the polyacrylonitrile carbon membranes for O2/N2 separation. J. Memb. Sci. 2003, 213, 285-291. 14. Jones, C. W.; Koros, W. J., Carbon molecular sieve gas separation membranes-I. Preparation and characterization based on polyimide precursors. Carbon 1994, 32, 1419-1425. 15. Suda, H.; Haraya, K., Gas Permeation through Micropores of Carbon Molecular Sieve Membranes Derived from Kapton Polyimide. J. Phys. Chem. B 1997, 101, 3988-3994. 16. Qiu, W.; Zhang, K.; Li, F. S.; Zhang, K.; Koros, W. J., Gas separation performance of carbon molecular sieve membranes based on 6FDA-mPDA/DABA (3:2) polyimide. ChemSusChem 2014, 7, 1186-1194. 17. Ning, X.; Koros, W. J., Carbon molecular sieve membranes derived from Matrimid® polyimide for nitrogen/methane separation. Carbon 2014, 66, 511-522. 18. Yoshino, M., Olefin/paraffin separation performance of carbonized membranes derived from an asymmetric hollow fiber membrane of 6FDA/BPDA–DDBT copolyimide. J. Memb. Sci. 2003, 215, 169-183. 19. Saufi, S. M.; Ismail, A. F., Fabrication of carbon membranes for gas separation––a review. Carbon 2004, 42, 241-259. 20. Liao, K.-S.; Fu, Y.-J.; Hu, C.-C.; Chen, J.-T.; Lin, D.-W.; Lee, K.-R.; Tung, K.-L.; Jean, Y. C.; Lai, J.-Y., Microstructure of carbon molecular sieve membranes and their application to separation of aqueous bioethanol. Carbon 2012, 50, 4220-4227. 21. Koh, D. Y.; McCool, B. A.; Deckman, H. W.; Lively, R. P., Reverse osmosis molecular differentiation of organic liquids using carbon molecular sieve membranes. Science 2016, 353, 804-807. 22. Tanihara, N.; Shimazaki, H.; Hirayama, Y.; Nakanishi, S.; Yoshinaga, T.; Kusuki, Y., Gas permeation properties of asymmetric carbon hollow fiber membranes prepared from asymmetric polyimide hollow fiber. J. Memb. Sci. 1999, 160, 179-186. 23. Vu, D. Q.; Koros, W. J.; Miller, S. J., High Pressure CO2/CH4Separation Using Carbon Molecular Sieve Hollow Fiber Membranes. Ind. Eng. Chem. Res. 2002, 41, 367-380. 24. Jiang, L. Y.; Chung, T.-S.; Rajagopalan, R., Dual-layer hollow carbon fiber membranes for gas separation consisting of carbon and mixed matrix layers. Carbon 2007, 45, 166-172. 25. Shiflett, M. B.; Foley, H. C., Ultrasonic Deposition of High-Selectivity Nanoporous Carbon Membranes. Science 1999, 285, 1902-1905. 26. Wang, H.; Zhang, L.; Gavalas, G. R., Preparation of supported carbon membranes from furfuryl alcohol by vapor deposition polymerization. J. Memb. Sci. 2000, 177, 25-31. 27. Dashti, A.; Asghari, M., Recent Progresses in Ceramic Hollow-Fiber Membranes. ChemBioEng Rev. 2015, 2, 54-70. 28. Shiflett, M. B.; Foley, H. C., On the preparation of supported nanoporous carbon membranes. J. Memb. Sci. 2000, 179, 275-282. 29. Strano, M. S.; Zydney, A. L.; Barth, H.; Wooler, G.; Agarwal, H.; Foley, H. C., Ultrafiltration membrane synthesis by nanoscale templating of porous carbon. J. Memb. Sci. 2002, 198, 173-186. 30. Richter, H.; Voss, H.; Kaltenborn, N.; Kamnitz, S.; Wollbrink, A.; Feldhoff, A.; Caro, J.; Roitsch, S.; Voigt, I., High-Flux Carbon Molecular Sieve Membranes for Gas Separation. Angew. Chem. Int. Ed. 2017, 56, 7760-7763. 31. Acharya, M.; Foley, H. C., Spray-coating of nanoporous carbon membranes for air separation. J. Memb. Sci. 1999, 161, 1-5.


19


Page 20 of 21

32. Shiflett, M. B.; Pedrick, J. F.; McLean, S. R.; Subramoney, S.; Foley, H. C., Characterization of Supported Nanoporous Carbon Membranes. Adv. Mater. 2000, 12, 21-25. 33. Wang, H.; Huang, L.; Holmberg, B. A.; Yan, Y., Nanostructured zeolite 4A molecular sieving air separation membranes. Chem. Commun. 2002, 1708-1709. 34. Yao, J.; Dong, D.; Li, D.; He, L.; Xu, G.; Wang, H., Contra-diffusion synthesis of ZIF-8 films on a polymer substrate. Chem. Commun. 2011, 47, 2559-2561. 35. He, M.; Yao, J.; Low, Z.-X.; Yu, D.; Feng, Y.; Wang, H., A fast in situ seeding route to the growth of a zeolitic imidazolate framework-8/AAO composite membrane at room temperature. RSC Adv. 2014, 4, 7634-7639.


20

Page 21 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TOC Graphic


21

Carbon Nanotube Networks as Nano-Scaffolds for Fabricating

Recommend Documents