Penetrated COF Channels: Amino Environment and Suitable Size for

Jan 4, 2019 - Penetrated COF Channels: Amino Environment and Suitable Size for CO2 Preferential Adsorption and Transport in Mixed Matrix Membranes...
0 downloads 0 Views 2MB Size
Subscriber access provided by Iowa State University | Library

Functional Nanostructured Materials (including low-D carbon)

Penetrated COF channels: amino environment and suitable size for CO2 preferential adsorption and transport in mixed matrix membranes Xiaochang Cao, Zhi Wang, Zhihua Qiao, Song Zhao, and Jixiao Wang ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b16877 • Publication Date (Web): 04 Jan 2019 Downloaded from http://pubs.acs.org on January 4, 2019

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 35 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

Penetrated COF channels: amino environment and suitable size for CO2 preferential adsorption and transport in mixed matrix membranes Xiaochang Cao, Zhi Wang*, Zhihua Qiao, Song Zhao, Jixiao Wang

Chemical Engineering Research Center, School of Chemical Engineering and Technology, Tianjin University, Tianjin 300350, PR China Tianjin Key Laboratory of Membrane Science and Desalination Technology, State Key Laboratory of Chemical Engineering, Collaborative Innovation Center of Chemical Science and Engineering, Tianjin University, Tianjin 300350, PR China

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 35

ABSTRACT: Developing mixed matrix membranes (MMMs) is challenging because the interface between different matrices often forms undesirable structures. Herein, we demonstrate a method of creating suitable CO2-selective channels based on interface regulation that greatly enhances membrane separation performance. The poly(vinylamine) (PVAm), which also acts as a polymer matrix, was immobilized onto covalent organic frameworks (COFs) to obtain polymer-COF hybrid materials (COFp). The COFp and polymer matrix are highly compatible since they have the same segment. The polymer matrix was induced to penetrate the oversized COFp, resulting in an amino-environmental pore wall and appropriately sized CO2-selective channels dispersed in MMMs. The MMMs exhibited satisfactory membrane performance for CO2/N2, CO2/CH4 and CO2/H2 separation. A CO2 transport model for preferential adsorption and transport is clearly presented for the first time. The membrane separation mechanism is also discussed. This work demonstrates potential applications for material, interface and membrane investigations.

KEYWORDS: interface regulation; penetration; channel size; covalent organic frameworks; CO2 separation.

ACS Paragon Plus Environment

2

Page 3 of 35 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

1. INTRODUCTION The increasing levels of carbon dioxide (CO2) emissions into the environment have contributed to ecological damage and extreme weather, which create an urgent need to develop environment-friendly and energy-efficient technologies for CO2 separation (e.g., CO2 capture from flue gas1 and upgrading syngas2 and natural gas3). Due to advantages of low-energy consumption, environmental friendliness, smaller area requirements and flexible operating conditions, membrane technology has become one of the most effective technologies for CO2 separation.4 Mixed matrix membranes (MMMs), which consists of two phases, a polymer matrix and a dispersed phase, have the potential to combine the CO2 separation performance of a porous filler with the good mechanical properties and processability of a polymer matrix.5 Anticipative fillers, such as metal-organic frameworks (MOFs), metal-organic polyhedra (MOPs), covalent organic frameworks (COFs) and graphene oxide (GO),6,7,8,9 often possess predictable porous nanostructures and selectable pore sizes and offer partial CO2-selective channels. At the same time, the introduction of a rigid filler component into a polymer matrix leads to the formation of a polymer-filler interface, and the interfacial morphology significantly influences the membrane structure and performance.10 The interfacial morphology can be classified as interval, adjacency, penetration and blockage (Figure 1). The distinction in polarity and flexibility between the two phases often results in a heterogeneous dispersion and boundary defects (Figure 1(a)), which lead to membrane defects and degradation of membrane selectivity.10 In contrast, the filler pores may become blocked by polymeric chain segments or even sorbent, solvent, contaminants or minor components from feed gasses.11 Total pore blockage (Figure 1(d)) does not allow the gas molecules to pass through the filler pores and greatly limits the increase in the CO2 permeance of membranes. Therefore, adjacency or

ACS Paragon Plus Environment

3

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 4 of 35

penetration (Figure 1 (b) and (c)), which can prevent the actual pore size in membranes from being too large or too small for gases to separate, may be a functional morphology for MMMs. Traditional strategies for improving the performance of MMMs focus on modifying the fillers and polymers to reduce the number of interface voids,5 but the high interface compatibility always leads to a penetration or blockage of filler pores.12,13,14 Attempting to ensure an unobstructed pore structure might result in moderate interface compatibility, making it difficult to avoid the boundary defects. Consequently, undesirable interfacial morphologies may easily form, thus limiting the improvement of the membrane separation performance. This study offers a feasible method for solving interface issues via penetration regulation, which contribute to create amino-environmental and appropriately sized channels for CO2 preferential adsorption and transport. Poly(vinylamine) (PVAm) was chosen as the polymer matrix due to its suitable cross-linking degree and amino-rich structure.15,16 A two-dimensional (2D) lamellar COF with favorable water stability was chosen as the filler due to its compatibility with the general polymer matrix17,18 and its oversized pore structure, which can be penetrated by a polymer matrix. The COF surfaces were functionalized by an isostructural chain from the polymer matrix to generate a polymer-COF hybrid material (COFp) that could confuse the polymer-filler boundary and induce PVAm chain penetration. Three molecular weights of PVAm were used to explore the appropriately sized channels because the number of PVAm segments might influence the degree of penetration.19 The penetrated PVAm chains afforded an amino environment of the pore wall and a suitable penetration degree to ensure an appropriate pore size for CO2 preferential adsorption and transport. The substrate in this study was polydimethylsiloxane-modified porous polysulfone (PSfm). Polydimethylsiloxane (PDMS) was introduced to prevent surface pore penetration20 and to ensure an ultrathin mixed matrix layer. The gas transport model and the separation mechanism of the membranes are

ACS Paragon Plus Environment

4

Page 5 of 35 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

discussed in this work. The developed membranes were investigated for the separation of CO2 from N2, CH4 and H2.

2. EXPERIMENTAL SECTION 2.1. PVAm Synthesis. PVAm was synthesized according to a previously reported method.21 The N-vinylformamide (NVF, 10 g, 0.14 mol, Sigma-Aldrich, 98%) monomer and a certain amount of 2,2’-azobis (2-methylpropion-amidine) dihydrochloride (AIBA) were dissolved in 40 mL of deionized water and stirred for 5 hours (50 °C, nitrogen atmosphere). Next, we injected a certain concentration of hydrochloric acid for 2 hours (70 °C) to proceed to the acidic hydrolysis. Then, the mixed solution was poured into plenty of ethanol to precipitate a white deposit. The deposit was dried in a vacuum oven at 40 °C for 48 hours, dissolved in deionized water and treated with an excess of strongly basic anion exchange resin (Amberlite 717). Finally, the PVAm aqueous solution was obtained by vacuum filtration of the mixture. The molecular weight (listed in Table S1, Supporting Information) was adjusted by changing the amount of initiator used in the polymerization reaction. The PVAms with low, medium and high molecular weights are denoted by PVAm(L), PVAm(M) and PVAm(H), respectively. 2.2. Synthesis of the Fillers. The COF synthesis method is shown in Scheme 1. The original COF was synthesized on the basis of the work of Ding et al26 and was improved in this study to obtain smaller particles. In a typical synthesis, 80 mg of 1,3,5-triformylbenzene (0.50 mmol, J&K Chemical, 99%) and 80 mg of 1,4-diaminobenzene (0.75 mmol, Sigma-Aldrich, 99%) were dissolved in 25 mL of dimethyl sulfoxide at 40 ºC. Then, 1.0 mL of glacial acetic acid (Sigma-Aldrich, 99.7%) was added slowly with rapid stirring. After stirring for 12 hours at 40 °C and resting for 12 hours, the resulting yellow powder was repeatedly centrifuged and washed with dimethylformamide (5 times × 15 mL) and dimethyl sulfoxide (5 times × 15 mL)

ACS Paragon Plus Environment

5

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 6 of 35

to obtain the original COF turbid solution. The PVAm(M) chain was immobilized to the original COF to obtain COFp. A total of 2.0 g of PVAm(M) solid and 0.5 mL of glacial acetic acid were dissolved in the original COF turbid solution. Subsequently, the mixture was allowed to react for 36 hours at 95 °C in a hydrothermal reactor. Finally, COFp was obtained by centrifuging the hydrothermal product and alkaline washing and drying under vacuum. 2.3. Preparation of the Membranes. The names of all membranes investigated in this study are listed in Table 1. PVAm membranes and COFp-PVAm membranes were prepared by coating the PVAm solution or the COFp-PVAm solution onto a PSfm substrate (Figure S1, Supporting Information). The PSfm substrate consisted of a PSf support and a PDMS intermediate layer. The PDMS (provided by ShinEtsu, Japan) solution was partially crosslinked by dissolving 0.5 g of PDMS and 1 g of tetraethoxysilane (TEOS, Aladdin, 99.99%) as the cross-linkers and 1 g of ditin butyl dilaurate (DBD, Aladdin, 95%) as the catalyst in heptane and vigorously stirring the mixture for 15 min at room temperature. Then, the PDMS solution was stored in an ice bath to inhibit further cross-linking. The solution was coated on top of the PSf porous supports (with an average cut-off molecular weight of 45,000 Da provided from Jozzon Membrane Technology Co. LTD., China) using a homemade applicator. The distance between coating knife and membrane surface was approximately 30 μm. The support was maintained at 303 K and 40% relative humidity in an artificial climate chamber (Climacell 222R, Germany) for at least 12 hours before the next step. The concentration of the coating solution for the PVAm membranes was 1.0 wt%. The suitable COFp/PVAm ratio (without solvent) was 10 wt%. Other COFp/PVAm ratios (5 wt% and 15 wt%) were also evaluated, , as detailed in the supporting information. The coating solution for the COFp-PVAm membranes was composed of the COFp, PVAm solution and 0.1 wt% additive polyvinyl alcohol (PVA), which enhanced the spreadability of the coating solution. The additive amount of PVA was low, and the influence of PVA addition is not discussed in this work. The coating solution

ACS Paragon Plus Environment

6

Page 7 of 35 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

preparation condition was mild, and the COFp structure was stable during the membrane preparation. The solution was coated on top of the PDMS intermediate layer with approximately 50 μm of distance between coating knife and membrane surface. The resulting membrane was maintained in an artificial climate chamber for at least 12 hours at 30 °C and 40% relative humidity. The original COF-PVAm membrane was also prepared via the abovementioned process. 2.4. Characterization Methods and Film Preparation. The X-ray diffraction (XRD) patterns of the as-prepared COF were recorded on an X-ray diffractometer (D/MAX-2500, Japan) in reflection mode at a power of 9 kW. The Fourier transform infrared (FTIR) spectra of the original COF and COFp were recorded on an FTS-6000 spectrometer (Bio-Rad of America). Potassium bromide was used as the background. The morphologies of the original COF and COFp were obtained using scanning electron microscopy (SEM, Nova Nano 430, FEI, America) and transmission electron microscopy (TEM, Tecnai G2 F20, FEI, America). The membrane morphologies were observed using SEM. The surface and cross-section images were recorded after the samples were coated by gold via sputtering. The cross-section samples were prepared by peeling off the non-woven fabric, immersing it in liquid nitrogen, and peeling off the non-woven fabric. The average molecular weight of PVAm (5 mL of a 3 wt% polymer solution in deionized water) was characterized by gel permeation chromatography (GPC; PL120, England) using high-purity water as the mobile phase. The degree of hydrolysis of PVAm was calculated based on the elemental analysis results using equations reported in our previous work.22 The pore size distributions of the COF and COF-PVAm films were obtained from the nitrogen adsorption-desorption analysis (3H-2000PM2, China). The film samples were prepared by drying the COFp-PVAm coating solutions on silicone rubber substrates and peeling the resultant films off the substrate. The names of all films are listed in Table 1. The gas adsorption capacities of COFp were characterized with a high-temperature and high-pressure

ACS Paragon Plus Environment

7

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 8 of 35

gas adsorption instrument (H-sorb 2600, Gold APP Instruments Corp, China) using pure CO2 and N2. The positron annihilation lifetimes of the membranes were investigated by bulk positron annihilation lifetime spectroscopy (PALS).23,24 22Na was used as the positron source, and the system was operated at a counting rate of approximately 160 counts s−1. All of the positron annihilation lifetime spectra were analyzed by a finite-term lifetime analysis method using a PATFIT program. X-ray photoelectron spectroscopy (XPS, PHI-1600) and SEMenergy dispersive spectroscopy (EDX, Nova NanoSEM 430, FEI, USA) were used to investigate the chemical structure of the COFs and membranes. 2.5. Gas-Permeation Experiments. The experiments to test membrane permselectivity were performed as described in the literature.25 The feed gas was successively passed through a humidifier (35 °C) and a dehumidifier (25 °C) to saturate with water vapor. Ar was used as the sweep gas to remove any components on the membrane-permeate side. The sweep gas passed through the water humidifier (25 °C) before entering the membrane cell. The effective membrane area of the membrane cell was 19.26 cm2. The outlet gas composition was analyzed by a gas chromatograph (HP 7890A, Porapak N). The fluxes of the gasses were calculated based on the outlet sweep gas flow rate and its composition. The permeance of a species was defined as the flux divided by the partial pressure difference between the upstream and downstream sides of the membrane. Gas permeance is typically expressed in units of GPU [1 GPU = 10−6 cm3 (STP)/(cm2.s.cmHg) = 3.35×10−10 mol/(m2.s.Pa)]. The selectivity was defined as the quotient of the two species’ permeances. (The downstream pressure is negligible compared to the upstream pressure.). To further elucidate the change in membrane separation performance, the diffusivity and solubility coefficients of homogeneous films were measured via the time-lag method at 25 °C, with the pressure on the upstream side was maintained at 0.15 MPa. Before the analysis, the films were evacuated for at least 8 h to remove previously dissolved species. For each membrane, the gases were tested in the order of N2 and CO2.

ACS Paragon Plus Environment

8

Page 9 of 35 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

3. RESULTS AND DISCUSSION 3.1. Structures of PVAm and COF. The PVAm structure is shown in Scheme 1. FTIR spectroscopy (Figure S2, Supporting Information) showed that rich amino groups appeared in the PVAms. The amino group content is related to the hydrolysis degree. The three PVAms of different molecular weights displayed similar hydrolysis degrees (Table S1, Supporting Information), which indicates that the amino group contents of the PVAms were relatively constant. The structures of the original COF and COFp are shown in Scheme 1. The FTIR spectra of COFp showed the characteristic peaks of the PVAm chain (Figure S1; refer to the Supporting Information for further details). The two alternative structures of the original COF and COFp are the eclipsed model (aligned pore) and staggered model (staggered pore).26 The XRD patterns (Figure 2(a)) matched the eclipsed model (calculated with Materials Studio software). Therefore, the original COF and COFp each had an aligned pore structure with a layered-sheet morphology, consistent with previous findings.26 The peaks of COFp were weaker than those of the original COF, revealing a significant decrease in crystallinity caused by the disturbance of the PVAm segments. This decrease in crystallinity would be favorable for good dispersion of COFp into the COFp-PVAm membranes. The porosity and pore size distributions were measured by means of a nitrogen adsorption-desorption analysis (Figure 2(b)), which revealed a narrow pore width distribution of approximately 1.8 nm (Figure 2(c)) for both the original COF and COFp. The PALS results for the largest sphere (R4) for the original COF and COFp were close to the pore size distribution values (Table S2, Supporting Information). The slight decrease in the pore size distribution of COFp compared with that of the original COF indicates that the presence of a polymer chain would not block the pore but could penetrate it. SEM and TEM micrographs of the original COF and COFp are shown in Figure 3. The morphology of

ACS Paragon Plus Environment

9

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 10 of 35

the particles is uniform, and the diameters of the individual particles are 50 nm. The particle edges in COFp are unclear and jagged in Figure 3 (c and d), confirming the existence of the polymer chains. 3.2. Structures of membranes and films. The COFs structure was stable during the membrane preparation.17,26 The SEM images of the COFp-PVAm membranes suggested nearly uniform filler dispersions, a non-defective membrane surface and a thickness of approximately 154 nm for the selective layer (Figure S3; refer to the Supporting Information for further details). The fillers in the surface section were indistinct because the polymer chain on COFp has attractive interactions with the polymer matrix, and COFp prefers to be covered by a polymer matrix rather than appearing on the membrane surface. The pore structure was verified by bulk PALS analyses27,28,29,30 for the PVAm films and COFp-PVAm films (Table 2). During the positron annihilation, the pick-off lifetime (τi) of the ortho-positroniums can be converted into the pore radius, and the fractional free volume (FFV) can also be calculated. The free volume in the films can be considered to be divided into two types, i.e., τ3/I3/R3 and τ4/I4/R4, respectively. The PVAm films had a pore radius (R3) that corresponded to the segment packing of PVAm. Upon the addition of COFp, a larger pore radius (R4) appeared, possibly corresponding to the emerging region of the COFp and interfaces. With the exception of these two values, no additional pore radius values were observed, indicating that the inherent pore size of approximately 1.8 nm of COFp was decreased in the membranes. This finding confirmed the penetration morphology. Due to the hydrogen bonds between the amino groups of PVAm and twelve N atoms on each unit of COFp under the alkaline aqueous coating solution, the PVAm chains could penetrate into the pores from one end to the other end (shown in Figure 4). The FTIR spectra of the COFp-PVAm(M) membrane (Figure S4, Supporting Information) revealed a right shift of the amino group peak, which was related to the stronger hydrogen bond compared with that in the PVAm(M) membrane. The stronger hydrogen bond might be

ACS Paragon Plus Environment

10

Page 11 of 35 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

attributable at least in part to penetration inside the COFp pores. The XPS results confirmed the rich amino groups in COFp and PVAm and supported hydrogen bond deduction (Figure S5; refer to the Supporting Information for further details). The pore distribution of COFp was only slightly smaller than that of the original COF, and the R4 value of COFp-PVAm membranes was much lower than the pore distribution of COFp, because the solvent (DMSO) under the acid catalyst presented electrophilicity during COFp synthesis, which blocked penetration by weakening the hydrogen bond between the PVAm segment and COF framework.31 After washing with water and drying, the pore distribution of COFp was decreased to approximately 1.6 nm (Figure S6, Supporting Information), which verified that the aqueous solution environment was beneficial to the penetration. After dipping with PVAm aqueous solution and drying, the pore distribution of COFp was widened to 1.4 nm ~ 1.7 nm, which verified the PVAm was easy to penetrate into the shell side of COFp. The EDX mapping of oxygen supported the PVAm penetration morphology in COFp (Figure S7; refer to the Supporting Information for further details). Moreover, the flexibility of the polymer matrix could be enhanced by increasing the PVAm molecular weight, resulting in an enhancement in the polymer-filler compatibility. A larger pore radius (R4) was obtained for higher-molecularweight COFp-PVAm film, as reported in Table 2; this trend could not attributed to the polymerfiller voids. PVAm modification of COF could greatly enhance the polymer-filler compatibility and avoid non-selective interface,28 as demonstrated by the smooth SEM surface image of the COFp-PVAm(M) membrane. Based on the characterization results, a schematic of the COFp-PVAm membrane selective layer structure is presented in Figure 4. The fillers are typically dispersed and embedded in the polymer matrix according to the SEM results. The polymer matrix penetrates the filler pores and decreases the intrinsic pore size.

ACS Paragon Plus Environment

11

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 12 of 35

3.3. Transport model for membranes. To identify the rules governing their gas separation performance, the COFp-PVAm membranes can be simplified to a 2D model, as shown in Figure 5. The behavior of the gas molecules is limited to unidimensional transport in the model. The filler pores may be regarded as uniformly distributed nanochannels with a width of W and a length of L. W is the penetrated pore size of COFp, and L is the total equivalent thickness of COFp particles along the vertical direction. The nanochannels are in the middle of the membrane, and the sum distance from the nanochannel to the top and bottom is calculated by subtracting L from the membrane thickness (T). All CO2 transport behavior can be classified into two forms in this model. In the first, the polymer matrix is the only route for CO2 molecules. In the second, CO2 molecules pass through the nanochannel with length L and through the polymer matrix with length T-L. The membrane performance could be regarded as the result of the parallel connections of these two forms (Figure 5). For the PVAm membranes, the performance can be regarded as reflecting a parallel connection of two of the first forms. When the transmembrane pressure is similar, the fluxes of the COFp-PVAm membranes (Qm) and the PVAm membranes (Qp) can be related via the following equation:

Qp Qm

=

2 R2 2Q1  R1  R2 Q1  Q2

(1)

where Q1 and Q2 represent the fluxes of the first and second forms, respectively, and R1 and R2 represent the gas transport resistances of the first and second forms, respectively. If it is assumed that (1) the N2 molecules do not pass through the nanochannels at low pressure (refer to Section 3.4 for further details) and (2) the pressure difference and membrane area are constant, then the proportions of Q1 and Q2 can be calculated from the membrane performance. 3.4. Membrane separation mechanisms. For the PVAm membranes, the PVAm segment spaces, which were in the range of 0.388 to 0.44 nm (Table 2), were slightly larger

ACS Paragon Plus Environment

12

Page 13 of 35 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

than the kinetic diameters of CO2 (0.33 nm), H2 (0.289 nm), N2 (0.364 nm) and CH4 (0.38 nm). CO2 molecules were preferentially adsorbed into the PVAm membranes because the rich amino groups in PVAm served as CO2 carriers.15 As a result, the permeance of CO2 was higher than that of H2, N2 or CH4.32 For the COFp-PVAm membranes, the separation mechanism in the polymer matrix was equivalent to the separation mechanism in the PVAm membranes. Regarding the nanochannels, the CO2 adsorption capacity of COFp was much higher than those of N2, CH4 and H2, as shown in Figure 2(d). With increasing pressure, the growth trends of the four gases were consistent the Langmuir equation. In the membranes, the amino groups of the penetrated PVAm were gathered inside the pore walls, which caused the CO2 molecules to be preferentially adsorbed into the pores. The suitable pore size (d) was more than two times the kinetic diameter of CO2 (dCO2) and less than two times dCO2 plus one kinetic diameter of another component (dS), such as H2, N2 or CH4. The relationship is expressed as follows: 2dCO2< d