functionalized multi-walled carbon nanotubes mixed matrix membranes for enhanced CO2 separation performance

functionalized multi-walled carbon nanotubes mixed matrix membranes for enhanced CO2 separation performance

Journal Pre-proof Bis(phenyl)fluorene-based polymer of intrinsic microporosity/ functionalized multi-walled carbon nanotubes mixed matrix membranes fo...

2MB Sizes 0 Downloads 2 Views

Journal Pre-proof Bis(phenyl)fluorene-based polymer of intrinsic microporosity/ functionalized multi-walled carbon nanotubes mixed matrix membranes for enhanced CO2 separation performance

Haixiang Sun, Wen Gao, Yanwei Zhang, Xingzhong Cao, Shanshan Bao, Peng Li, Zixi Kang, Q. Jason Niu PII:

S1381-5148(19)30983-6

DOI:

https://doi.org/10.1016/j.reactfunctpolym.2019.104465

Reference:

REACT 104465

To appear in:

Reactive and Functional Polymers

Received date:

17 September 2019

Revised date:

20 December 2019

Accepted date:

31 December 2019

Please cite this article as: H. Sun, W. Gao, Y. Zhang, et al., Bis(phenyl)fluorene-based polymer of intrinsic microporosity/functionalized multi-walled carbon nanotubes mixed matrix membranes for enhanced CO2 separation performance, Reactive and Functional Polymers (2019), https://doi.org/10.1016/j.reactfunctpolym.2019.104465

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier.

Journal Pre-proof

Bis(phenyl)fluorene-based polymer of intrinsic microporosity/functionalized multi-walled carbon nanotubes mixed matrix membranes for enhanced CO2 separation performance Haixiang Suna,b,* [email protected], Wen Gaoa,d, Yanwei Zhanga,

Xingzhong Caoc,

Shanshan Baoa, Peng Lib, Zixi Kanga, and Q. Jason Niub,* [email protected] School of Materials Science and Engineering, China University of Petroleum (East China),

of

a

b

ro

Qingdao 266580, P.R. China

State Key Laboratory of Heavy Oil Processing, China University of Petroleum (East China),

d

re

Institute of High Energy Physics, Chinese Academy of Sciences, Beijing 100000, P. R. China Shandong Pengbo Safety & Environmental Services Limited Company, Yantai 265600, P.R.

lP

c

-p

Qingdao 266580, P. R. China

China

ABSTRACT

ur

Corresponding authors

na

*

Jo

Polymers of intrinsic microporosity (PIM) as one of potential next generation membrane materials for gas separation has attracted great interests due to its ultra-permeable characteristics. Herein, a novel dihydroxyphenyl)

bis(phenyl)fluorene-based PIMs (Cardo-PIM-1) based on 9,9-bis(3,4fluorene

(BDPF),

2,3,5,6-tetrafluoroterephthalonitrile

(TFTPN)

and

spirocyclic 5,5’,6,6’-tetrahydroxy-3,3’,3,3’-tetramethylspirobisindane (TTSBI) were prepared vis dibenzodioxane polymerization reaction, and then the functionalized multi-walled carbon nanotubes (f-MWCNTs) were incorporated into Cardo-PIM-1 to fabricate mixed matrix membranes (MMMs) with solution mixing method for CO2 separation. The structure analysis

Journal Pre-proof indicated that the MWCNTs were cut into short ropes and the amino groups were incorporated into the nanotubes surface after treated with acid mixtures followed by ethylenediamine modification. FTIR spectroscopy and nuclear magnetic resonance (NMR) measurement confirmed the formation of Cardo-PIM-1 macromolecule. BET and positron annihilation lifetime spectroscopy (PALS) analysis exhibited that Cardo-PIM-1 contained larger pore-size distribution and fractional free volume (FFV), and preferential CO2 adsorption capacity over N2 compared

of

with that of PIM-1. This work investigated the structure of polymer as well as the effect of

ro

nanofillers in the gas separation performance. High CO2 permeability of 2.9 ×104 Barrer with a

-p

desirable CO2/N2 separation factor of 24.2 was achieved using the MMMs with 7.5 wt.% f-

re

MWCNTs loading, which were among the best performance for CO2 separation. The CardoPIM-1/f-MWCNTs MMMs will provide a promising alternative in industrial flue gas separation

lP

and CO2 capture process.

na

Keywords: Cardo polymers of intrinsic microporosity; multi-walled carbon nanotubes; mixed

Jo

ur

matrix membranes; gas separation; fractional free volume

Journal Pre-proof

1. Introduction With the change of global climate, the reduction of pollutant emission is becoming the focus of environmental issues. As the main atmospheric greenhouse gases, carbon dioxide (CO2) separation processes from flue gas (mainly nitrogen) have therefore received great attention worldwide [1]. Compared with the conventional separation technologies such as physical adsorption [2], chemical adsorption [3] and the separation at low temperature [4], membrane-

of

based processes has the advantages of high efficiency, low energy consumption, environmental

ro

protection and easy to control, which has been considered as one of the most potential methods

-p

for CO2 capture [5-7]. However, traditional polymer membranes suffer from a near-universal

re

trade-off phenomenon between gas permeability and selectivity as shown in upper bound curves

lP

developed by Robeson [8].

It is fundamentally important for the separation properties of prepared membranes to choose

na

the suitable polymer. Many types of new polymer membrane materials have been developed in recent years, for example the thermally rearranged polymers, modified polyimides, etc. Regarded

ur

as a new family of membrane material, polymers of intrinsic microporosity (PIMs) exhibit the

Jo

characteristics of a microporous material (pore size < 2 nm) [9] based on their rigid and contorted molecular configuration, wholly composed of fused-rings, such as the spirocyclic structure [10]. This special chain structure endows PIMs with higher Brunauer-Emmett-Teller (BET) surface areas and fractional free volume [11]. Therefore, PIMs offer the advantages over inorganic and metal-organic porous materials in terms of high gas separation productivity. The archetypal membrane-forming PIMs, referred to as PIM-1, is soluble polymer comprised of fused ring sequences interrupted by spiro centers. PIM-1 is soluble in common volatile solvents, such as chloroform and tetrahydrofuran, and can be cast from solution into solid membranes,

Journal Pre-proof which indicated great promise for gas separation [12]. In the recent two decades, there has been intensive effort to achieve even higher performance of PIM-1 membranes, for example, the synthesis of novel polymers as the chemical postmodification of PIM-1 [13,14] and the thermally self-crossing-linking of PIM-1[15]. Wang et al. synthesized a Tröger’s base (TB)-based copolymers with intrinsic microporosity (TBPIMs) through the reaction of TB-embedded tetrolhydroxy aromatic monomers and 2,3,5,6-tetrafluoroterephthalonitrile (TFTPN) [16].

of

Compared with PIM-1, the introduction of TB unit enhanced the CO2 selectivity due to the

(2D)

ribbon-shaped

polymer

PIM-TMN-Trip

based

on

a

fused

-p

two-dimensional

ro

increase of CO2 solubility in the copolymer membranes. Rose et al. reported an ultrapermeable

re

tetramethyltetrahydronaphthalene (TMN) unit as the extended substituent [17]. The results demonstrated that the gas permeability of PIM-TMN-Trip was superior to the structural similar

lP

polymers with three-dimensional (3D) contorted chains. Kazama et al. confirmed that the

na

bis(phenyl)fluorene-based cardo polymers containing a structure of four phenyl rings connected to a quaternary carbon resulted in the severe rotational hindrance of the phenyl groups. The

ur

bulky cardo moiety reduced the rotational mobility and increased the average distance of the

Jo

main chains, which exhibited larger gas permeabilities than those of commercial polymers [18]. Another alternative approach to improve gas separation performance of polymer membranes is to introduce inorganic materials, such as metal organic frameworks (MOFs) [19, 20], graphene [21, 22] and multi-walled carbon nanotubes (MWCNTs) [23-27], into the polymer matrix, which is termed as mixed matrix membranes (MMMs). Compared to other nanofillers, MWCNTs exhibit exceptional mechanical, thermal and electrical properties due to the unique one-dimensional carbon structure. Besides, MWCNTs seem to be more advantageous for the transport of light gases based on their inherent smooth walls and large diameter pores [23].

Journal Pre-proof Recently, the addition of carbon nanotubes (CNTs) into the intermediate and final polymer to fabricate MMMs, and an improvement of the gas permeability in comparison with the neat polymer membranes was achieved. Similar to the preparation of most nanocomposite membranes, the dispersion and alignment of CNTs is crucial in improving the quality and properties of the resulting MMMs. Considering the dispersion of pristine MWCNTs (pMWCNTs) as fillers and strengthening the unique attraction of CO2, functional treatment of the

of

MWCNTs is an effective strategy to be used [26-29].

ro

In this work, we propose a simple method based on the bis(phenyl)fluorene unit to

-p

synthesize a novel “Cardo” polymer of intrinsic microporosity (Cardo-PIM-1, Scheme 1). The

re

fused-ring structures in Cardo-PIM-1 can disrupt the polymer crystallization and enhance gas molecules solubility in the polymer, which effectively improves the gas permeability. The

lP

chemical structure and functional group of the Cardo-PIM-1 were confirmed by Fourier

na

transform infrared (FTIR), 1H nuclear magnetic resonance (NMR), X-ray diffraction (XRD) and Brunauer–Emmett–Teller (BET)

analysis. The atomistic model and positron annihilation

ur

lifetime spectroscopy (PALS) was used to determine the fractional free volume (FFV) of the

Jo

polymer. The MMMs were fabricated by the incorporation of functionalized MWCNTs into the Cardo-PIM-1 polymer matrix, which were characterized for their morphology using scanning electron microscopy (SEM). A detailed investigation of the effect of fluorene-based cardo structure and the content of MWCNTs on the gas separation performance of the MMMs was undertaken by the gas permeation measurements.

Journal Pre-proof

n

F OH

HO

+

CN

F

F

F

O

CN

CN

OH

OH

CN

O

F OH

HO

OH

+

F

+

2n F

CN

i

n

O

O

CN O

O

O

O

F O

CN OH

CN

HO

TTSBI

n

PIM-1

TFTPN

TTSBI

HO

O

O

i

n

OH

n

O

CN

HO

TFTPN

CN

n

Cardo-PIM-1

BDPF

of

Scheme 1. Synthetic of PIM-1 and Cardo-PIM-1

ro

2. Experimental

-p

2.1. Materials

re

All solvents and reagents were used as purchased without further purification, unless stated

lP

otherwise. The main monomers required for Cardo-PIM-1 formation, 5,5’,6,6’-tetrahydroxy3,3,3’,3’-tetramethyl-1,1’-spirobisindane (TTSBI, 96%) and 9,9-bis(3,4-dihydroxyphenyl)

na

fluorene (BDPF, 98%) were purchased from TCI (Shanghai) Development Co., Ltd. 2,3,5,6-

ur

tetrafluoroterephthalonitrile (TFTPN, 99%) and anhydrous potassium carbonate (K2CO3,

Jo

99.99%) were provided by Sigma-Aldrich (Shanghai, China). HPLC grade dimethylformamide (DMF, 99.9%) and chloroform (CHCl3, >99%) were supplied by Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China) and used after dehydration. Methanol (MeOH, >99.5%), dichloromethane (CH2Cl2, >99.5%), ethanol (EtOH, >99.5%), methylbenzene (MB, >99.5%), ethylene diamine (EDA, >99.5%), HCl (37.5%), H2SO4 (98%), HNO3 (70%) and SOCl2 (>99%), were purchased from Xilong Chemical Co., Ltd. (Shanghai, China). High-purity MWCNTs (purity: >95 wt. %; average diameter: 8-15 nm; length: 10-30μm) were supplied by Chengdu Organic Chemicals Co., Ltd. Chinese Academy of Sciences (Chengdu, China). Pure CO2, N2 and CH4 were obtained from Tianyuan Gas Company (Qingdao, China).

Journal Pre-proof 2.2. Novel Cardo-PIM-1 synthesis The synthesis of Cardo-PIM-1 is based on the polycondensation reaction of three kinds of monomers TFTPN, TTSBI and BDPF. Firstly, a mixture of anhydrous K2CO3 (15mmol), TFTPN (5mmol), TTSBI (2.5mmol) and BDPF (2.5mmol) in dry DMF (30ml) reacted at 65 oC for 72h under a nitrogen atmosphere. After cooling to the room temperature, the mixture was added to methanol (100ml) and the crude product was collected by filtration. The product was then

of

dissolved in CH2Cl2 and reprecipitated in methanol to filter out some insoluble matter. The

ro

procedure was repeated twice. After filtration, the yellow solid was boiled in deionized water at

-p

least 2 h and vacuum-dried at 80 oC for at least 48 h to afford the Cardo-PIM-1, approximately

re

89% yield was obtained. In order to investigate the effect of cardo structure of the copolymer on the gas separation performance, PIM-1 were prepared as described previously [12].

lP

2.3. Functionalized MWCNTs preparation

na

In order to improve the dispersion of CNTs in the polymer matrix and reinforce the interfacial interactions between CNTs and PIM-1, the p-MWCNTs were modified by EDA

ur

functionalization. Firstly, 0.8g of p-MWCNTs were oxidized with a mixed acid solution of

Jo

H2SO4/HNO3 (3:1 v/v) at 60 oC for 4h, the acid-treated MWCNTs (o-MWCNTs) were filtered and washed with deionized water until pH was 7 and then dried in vacuum oven at 70 oC for 24h prior to further treatment. Secondly, 0.3g of o-MWCNTs were dispersed in a mixture of SOCl2 and DMF (20:1 v/v) at 70 oC for 24h. After the acyl chlorination, SOCl2 and DMF were removed through centrifugation, and the acyl chloride-treated MWCNTs (l-MWCNTs) were filtered and washed with MB. After drying in vacuum oven at 70 oC for 24h, EDA (50ml) was added to react with l-MWCNTs at 100 oC for 48 h until no HCl gas existed. After cooling to room temperature, the amine-functionalized MWCNTs (f-MWCNTs) were washed with EtOH for three times to

Journal Pre-proof remove excess EDA. Finally, the black solid was dried at room temperature overnight in vacuum oven [30]. Corresponding chemical reactions are illustrated in Fig. S1. 2.4. Synthesis of Cardo-PIM-1 based MMMs Cardo-PIM-1 based MMMs were fabricated via the solution-casting method. Firstly, a certain concentration of p-MWCNTs, o-MWCNTs or f-MWCNTs were dispersed in 5ml CHCl3 solvent by sonication for 2 h respectively. Then blending of the Cardo-PIM-1/MWCNTs mixture

of

was realized by addition of 150 mg Cardo-PIM-1 to the series of MWCNTs solutions prepared

ro

above. The resultant mixture solution was filtered using a 1.0μm Whatman’ filter prior to casting

-p

onto a polyacrylonitrile (PAN) substrate membrane using a casting machine. After evaporating

re

for 12h at room temperature, the obtained films were treated at 30 °C in a vacuum oven for 24 h to remove the traces of the solvent completely. As a result, the content of MWCNTs in the

lP

Cardo-PIM-1 based MMMs was 2.5, 7.5 and 12.5 wt. % respectively.

na

The preparation of PIM-1 or Cardo-PIM-1 membranes were also with the solution-casting method. 150mg PIM-1 or Cardo-PIM-1 was dissolved in 5ml CHCl3 and stirred overnight, then

ur

the pure polymer membranes were fabricated following the procedure of the Cardo-PIM-

Jo

1/MWCNTs MMMs described above. All membranes thicknesses were approximately 8 ± 0.5 μm achieved by SEM image analysis through controlling the thickness of the casting machine. 2.5. Characterization and measurement The chemical structure of modified MWCNTs and the novel Cardo polymers were characterized using the Fourier transform infrared spectroscopy (FT-IR, TENSORII, Bruker, Germany) in the wavenumber range of 4000−600 cm−1. The morphology of MWCNTs were observed by a transmission electron microscopy (TEM, JEM 2100F, JEOL, Japan). Surface chemistry of pristine and amine-functionalized MWCNTs was monitored with X-ray

Journal Pre-proof photoelectron spectroscopy (XPS, Thermo escalab 250Xi, Thermo Electron, USA). Amonochromatic Al Ka X-ray was used at 1486.6 eV. The molecular structures of the novel Cardo polymers were also performed by the 1H nuclear magnetic resonance (NMR, Avance Ш, Bruker, Switzerland). The X-ray diffraction spectra (XRD, X’Pert-Pro MPD, Panalytical, Holland) was employed to investigate the crystalline properties and the physical structure of the polymers using a voltage of 40 kV, current

of

of 40 mA. The X-rays were generated by a Cu Kα source. Each pattern was collected in the 2θ

ro

from 5° to 40° in the repetition mode with a total duration of approximately 0.4 h at selected

-p

times of hydration. The d-spacing of the PIM-1 and Cardo-PIM-1, which was the distance

re

between diffracting plane, was calculated by substituting the scattering angle, 2θ, of the perk into the Bragg equation.

λ

lP

d=

2 sin θ

(1)

na

where λ (λ=1.5418Å) is the wavelength of the Cu Kα ray (nm) and 2θ is the diffraction angle (o).

ur

The micropore and distribution, surface areas and the gas sorption of PIMs were analyzed by Brunauer-Emmett-Teller measurements (BET, ASAP2060, Micromeritics, USA). The

Jo

fractional free volume of PIMs were measured by the positron annihilation lifetime spectroscopy (PALS, EG&G ORTEC Co., Tennessee, USA) with a FWHM=195 ps for a 60Co prompt peak of 1.18 MeV and 1.33 MeV γ rays. A 6× 105 Bq of positron source (22Na) was deposited between two Kapton films, which was sandwiched between two identical composite samples. All PALS measurements were performed at room temperature. The thermostability of the polymer was characterized by thermogravimetric analysis (TGA, STAR System, METTLER TOLEDO, USA). Due to the strong dipolar interactions of nitrile groups, the Cardo-PIM-1 and PIM-1 were thermally stable up to 500 oC and 480 oC,

Journal Pre-proof respectively. The molecular weight of the Cardo-PIM-1 and PIM-1 were determined by gel permeation chromatography (GPC, M302, Visco Tec, USA) measurements. The obtained GPC value for Cardo-PIM-1 is as follows: Mn =40500, Mw =68900, PDI (Mw/Mn) =1.70, and the corresponding value of PIM-1 is Mn =55000, Mw =99200 and PDI (Mw/Mn) =1.80. The result indicates that the Cardo-PIM-1 chains are shorter than that of PIM-1 due to the incorporation of “Cardo” structure.

of

The cross-section structure of the Cardo-PIM-1/MWCNTs MMMs were characterized by

ro

scanning electron microscopy (SEM, S-4800, Hitachi, Japan). The membrane sample was

-p

prepared by peeling away from the PAN substrate membrane, and frozen in liquid nitrogen and

re

fractured, after being coated with a conductive layer of sputtered gold. 2.6. Atomistic models of the polymer

lP

The atomistic models of the PIM-1 and Cardo-PIM-1 polymers were constructed by the

na

Amorphous Cell module in Materials Studio 8.0 (MS, Accelrys Inc., San Diego, CA, USA) [31]. Each model was composed of one polymer chains in a cubic simulation box with an initial

ur

density of 0.7g/cm3. Each polymer chain consisted of 20 repeat units arranged in a random

Jo

torsional angle. The lowest-energy configurations were selected for equilibration by the following procedures: (1) 5000 steps of energy minimization at 0K; (2) 200ps NPT-MD simulation at 1000K and 1bar; (3) 200ps NVT-MD simulation at 1000K and 1bar; (4) 200ps NPT-MD simulation at 298K and 1bar; (5) 200ps NVT-MD simulation at 298K and 1bar. Through extensive tests, this equilibration protocol was found highly efficient to equilibrate model polymers [32]. 2.7. Gas permeation measurement

Journal Pre-proof The MMMs used for the gas separation measurement was set in a stainless steel cell at room temperature (25 oC) and standard atmospheric pressure [33]. The feed side of the membrane was exposed to the gas mixtures (CO2: N2=1:1) while the permeate side was swept by argon (Ar). A soap-film flow meter was used to measure the flux of the gas, and the gas that penetrated the membrane was analyzed using gas chromatograph (GC-2014C, SHMADZU, Japan). Gas permeability coefficients (Pi) can be calculated as follows:

m-2 s-1 Pa-1 ) = (mol ∆P ×A 3.35 i

Ni L

of

104

Ni

∆Pi ×A

(2)

ro

Pi =

where Ni is the permeate flow rate of component i (mol s-1), ΔPi is the trans-membrane pressure

-p

drop of i (Pa), and A and L are the effective area of the membrane (m2) and the membrane

re

thickness (m), respectively. The unit of Pi is in Barrer (1Barrer=1×10-10 cm3 cm cm-2 s-1 cmHg-1 ).

na

factor (αi/j) is defined as follows:

lP

The average value of at least three times measurement results was obtained. The separation

Xi ⁄ Xj Yi ⁄ Yj

=

Xi Xj

(3)

ur

αi/j =

where i and j represent the two components in the mixture, and X and Y are the mole fractions in

Jo

the permeate and feed side, respectively. In our work, the feed side of the membrane was exposed to the gas mixtures (CO2: N2=1:1), the separation factor can also be calculated as the ratio of gas permeability.

3. Results and discussion 3.1. MWCNTs structure analysis

Journal Pre-proof The functionalized MWCNTs were prepared according to the method of literature [30]. TEM images of p-MWCNTs and f-MWCNTs in Fig. 1 show that the p-MWCNTs are randomly and loosely entangled together, whereas the f-MWCNTs are suspended and manipulated as individual macromolecules without entanglement. The length of f-MWCNTs is much shorter than that of the pristine ones [34], and the p-MWCNTs are generally closed-ended while the fMWCNTs are open-ended. It also can be seen that both the pristine and amine-treated CNTs

of

graphitic walls are smooth, which demonstrates that the course of functionalization has no

na

lP

re

-p

ro

destructive effect on the tubular structure of carbon nanomaterials.

Jo

ur

Fig. 1. TEM images of p-MWCNTs (a) and f-MWCNTs (b)

The characterization of FTIR and XPS was investigated to obtain the chemical structure of pristine and functionalized MWCNTs. As shown in Fig. 2, compared to the pristine MWCNTs, the peaks at 1730 and 1110 cm-1 in the spectrum of o-MWCNTs are in correspondence to C=O and C-O stretching vibrations. The peaks of C=O and C-O stretching vibrations in the spectrum of l-MWCNTs shift to 1740 and 1120 cm-1 based on the negative inductive effect of chlorine atoms and the peak at 614 cm-1 attributes to the C-Cl stretching vibration of COCl groups. In the spectrum of f-MWCNTs, the peaks of C=O and C-O stretching vibration shift to 1720 and 1114 cm-1 due to the formation of amide linkages. The peak at 1180 cm-1 is ascribed to C-N stretching

Journal Pre-proof of amide groups, while the peak at 3400 cm-1 can be assigned to the N-H stretching vibrations [35,36], which indicates that the acid and amine treatment process effectively generates more functional groups on the MWCNTs surface.

f-MWCNTs 1180 1720

1740

of

l-MWCNTs

614

ro

1120

1730

o-MWCNTs

1110

-p

Transmittance(%)

1114

4000

3500

lP

re

p-MWCNTs

3000

2500

2000

1500

1000

500

-1

na

Wavenumber( cm )

Jo

ur

Fig. 2. FTIR spectra of MWCNTs

The XPS spectra in Fig. S2 indicate five kinds of C atoms in different functional groups, namely C-C (284.88 eV), C-N (285.08 eV), C-O (285.80 eV), C=O (287.18 eV) and C(O)O (289.50 eV). After the acid treatment, the peaks at 285.80, 287.18 and 289.50 eV of the oMWCNTs attribute to the C-O, C=O and C(O)O groups respectively, are much higher than those of the p-MWCNTs. The results illustrate that acid oxidation has introduced carboxylic acid groups (–COOH) and hydroxyl groups (–OH) on the surface of MWCNTs. In the spectrum of fMWCNTs, the C-O, C=O and C(O)O groups peaks are lower than those of the o-MWCNTs, and a new peak at 285.08 eV attributed to C-N groups appears, which confirms that amino-

Journal Pre-proof functionalized carbon nanotubes has formed the amido groups (-CONH-) with the reaction between the acyl chloride of l-MWCNTs and amino groups of EDA. As summarized in Fig. 3 and Table 1, the peaks at 533.08, 400.08 and 285.08 eV are attributed to O1s, N1s and C1s respectively [37]. The intensity of the O1s evidently increases from the p-MWCNTs of 5.22% to the o-MWCNTs of 15.82% after the acid treatment, further decreases to 10.74% after the amine treatment. Besides, the new peak at 285.08 eV attributed to N1s appears, which further certifies

ro

of

that the modification process has introduced amino groups on the surface of f-MWCNTs.

-p

N1s

re

O1s

o-MWCNTs

na

lP

Intensity

f-MWCNTs

ur

p-MWCNTs

Jo

1200

C1s

O1s

1000

C1s

C1s

O1s

800

600

400

200

0

B.E.(e.V)

Fig. 3. XPS survey scans of MWCNTs

Table 1. Elemental composition of the MWCNTs Element (atom%) Samples

p-MWNTs

C1s

O1s

N1s

else

94.78

5.22

0

0

Journal Pre-proof o-MWNTs

83.54

15.82

0

0.64

f-MWNTs

82.59

10.74

5.74

0.93

3.2. Characterization of Cardo-PIM-1 The FTIR spectra of TTSBI, PIM-1 and Cardo-PIM-1 are illustrated in Fig. 4a. Compared with the monomer TTSBI, PIM-1 and Cardo-PIM-1 show the loss of the hydroxyl groups (-OH)

of

at 3400 cm-1 and the appearance of peak at 2239 cm-1 which is attributed to the nitrile stretching

ro

vibration. The characteristic band at 860 cm-1 is representative of the C-H stretching vibration of

-p

the phenyl group. 1H NMR spectra and assignments in a nonselective solvent CDCl3 (400 MHz)

re

for Cardo-PIM-1 and PIM-1 shown in Fig. 4b, the Cardo-PIM-1 contains all peaks of PIM-1, and there are some special peaks for the Cardo-PIM-1. For the PIM-1 and Cardo-PIM-1, the

lP

corresponding chemical shifts δ are 1.10-1.47 (br. m, 12H; 4×CH3), 2.00-2.22 (br. m, 4H;

na

2×CH2), 6.31 (br. s, 2H; Ph-H) and 6.65 (br. s, 4H; Ph-H). Beside the above shifts, special chemical shifts δ at 6 .65 (br. s, 3H; Ph-H), 7.14 (br. s, 2H; Ph-H), 7.60 (br. s, 2H; Ph-H) also

ur

appear in the Cardo-PIM-1 sample. FTIR and NMR measurement indicates that the

Jo

bis(phenyl)fluorene unit has successfully introduced into the PIM-1 material.

Journal Pre-proof

(a)

O

H O

O

CN

O H

H O

CN

O

O

O

O

O H

O

O

O

O

O

n

CN

TTSB I

CN

O

CN

PIM -1

CN

n

Cardo-PIM-1

Transmittance

TTSBI

-OH

Cardo-PIM-1

-CN

4000

3500

3000

ro

of

PIM-1

2500

2000

1500

1000

500

-1

-p

Wavenumber( cm )

re

(b)

3

1

3

O

2

CN

4 O

O

3

4 3

ur

2 3

10

11

9 2

CN

4

O

O

8 7

O

1

2

2

6

O

O

4

3

5

3

n

CN

Cardo-PIM-1

1

7

n

CN

PIM-1

3

4,5,6,7

Jo 8

2

3

1

O

8,9,10,11

1

1

na

4

lP

O

3

6

5

4

3

2

1

PPM

Fig. 4. FTIR (a) and 1H NMR (b) spectra of PIM-1 and Cardo-PIM-1

XRD analysis is employed to investigate the crystalline properties and the physical structure of the Cardo-PIM-1 and PIM-1. As shown in Fig. 5, three characteristic peaks at 2θ 13.1°, 18.1° and 23.4° are detected for the synthetic polymer, and the corresponding d-spacing of 6.7, 4.9 and

Journal Pre-proof 3.8Å, respectively. The peak appeared at 13.1° probably results from the inefficient packing of polymer chains which originates from the sites of contortion in the polymer backbone [38]. The peak at 18.1° is ascribed to the conformation of ultrafine ultramicropores [39] and the peak at 23.4° refers to the aromatic systems [40]. Compared with PIM-1, these two peaks in the CardoPIM-1 are distinctly stronger based on the bis(phenyl)fluorene unit directing motif in the

4.9Å 3.8Å

ro

re

-p

Intensity (a. u.)

6.7Å

of

polymer chain, which exhibits more intrinsic microporosity structure and phenyl groups.

na

lP

Cardo-PIM-1

20

30

40

2θ( degree)

ur

10

PIM-1

Jo

Fig. 5. XRD patterns of the PIM-1 and Cardo-PIM-1

The nitrogen adsorption/desorption curves of the two polymers at 77K for determining the pore-size distribution and the gas (CO2 or N2) adsorption analysis at 298K were performed. As shown in Fig. 6a, open hysteresis loops formed by the irreversible adsorption and desorption curves are observed, which is the characteristic of intrinsic microporous polymers [41, 42]. The nitrogen isotherm provides that the BET surface areas are 671 m2·g-1 for PIM-1 and 502 m2·g-1 for Cardo-PIM-1, and the corresponding pore volume are 0.58 and 0.37 cm3·g-1 respectively. The presence of micropores could also be verified by the pore-size distributions derived via the

Journal Pre-proof Horvath-Kawazoe method analysis of the isotherms. As illustrated in Fig. 6b, PIM-1 and CardoPIM-1 show apparent micropore distribution of less than 1 nm. Cardo-PIM-1 has more and larger microporous (0.7-0.8nm) in comparison to the structure of PIM-1, which are associated with the enhanced gas permeability. Open micropores derived from inefficient packing of rigid and contorted chains in Cardo-PIM-1 permit fast gas diffusion, which will improve the gas permeability. The CO2 adsorption capacity of PIMs is influenced by many factors, such as the

of

micropore surface area, pore volume, pore-size distributions [43] and the molecular interactions

ro

between the adsorbent and adsorbate molecules. The CO2 sorption of Cardo-PIM-1 at 298K is

-p

dramatically higher than that of PIM-1 at 298K and there’s no obvious difference of N2 sorption

re

(Fig. 6c). The higher CO2 sorption capability of Cardo-PIM-1 materials can be explained that PIMs with bis(phenyl)fluorine unit have two different kinds of micropores, small (diameter <0.7

lP

nm) and large (diameter 0.7-0.8 nm). The larger micropores possess more Langmuir adsorption

na

sites saturated at relatively low pressure [44], which is favorable for the improvement of solubility coefficient selectivity of CO2/N2 [45]. The following gas permeability test also

ur

confirms that higher pore width and CO2 sorption of the polymer is beneficial to the

400

Jo

improvement of the gas permeability and separation factor.

(a)

3.5

PIM-1 Cardo-PIM-1

3.0

300 250 200 150

PIM-1 adsorption PIM-1 desorption Cardo-PIM-1 adsorption Cardo-PIM-1 desorption

100 50

dV/dW/(cm3g-1nm-1)

350

Vads/(cm3/g)[STP]

(b)

2.5 2.0 1.5 1.0 0.5

0

0.0 0.0

0.2

0.4

0.6

Relative Pressure (P/Po)

0.8

1.0

0.4

0.5

0.6

0.7

0.8

Pore width (nm)

0.9

1.0

1.1

Journal Pre-proof

(c)

Vads (cm3g-1) [STP]

40 CO2

30

Cardo-PIM-1

20 PIM-1

10 N2

0 0.2

0.4

0.6

0.8

1.0

ro

p/p0

of

0.0

-p

Fig. 6. (a) Nitrogen adsorption/desorption isotherms of PIM-1 and Cardo-PIM-1 measured at 77K. (b) Pore-size distributions analysis of N2 adsorption at 77 K by the Horvath–Kawazoe

lP

re

method. (c) CO2 and N2 sorption isotherms of PIM-1 and Cardo-PIM-1 at 298K. 3.3. Fractional free volume properties of the polymer

na

3.3.1. Atomistic models analysis

ur

Many physicochemical properties of PIMs are explained by the contorted structure of their macromolecules. Combined with the PIMs rigidity and inability to pack efficiently in the solid

Jo

state, the polymers generate intrinsic microporosity. As shown in Fig. S3, based on the bis(phenyl)fluorene structure of the Cardo-PIM-1, the unit molecular chain is more distorted, correspondingly the fractional free volume is higher than that of PIM-1. The voids in a polymer govern the microscopic properties of gas molecules. Molecular dynamics simulation indicates that there exist substantial interconnected voids in PIM-1, implying PIM-1 is zeolite-like from the geometrical point of view. By comparison, there are larger micropores in the Cardo-PIM-1, which is well consistent with the pore-size distributions observed in Fig. 6b.

Journal Pre-proof The fractional free volumes (FFVMS) in the PIM-1 and Cardo-PIM-1 were estimated with the Lee proposed an empirical method [46]. Vvdw/Vsp=1/FFVmax

(4)

FFVMS=1-1.3Vvdw/Vsp

(5)

With this method, the FFVMS in PIM-1 and Cardo-PIM-1 are estimated to be 29.4% and

of

32.7% respectively (Table 3), roughly in coherence with the previously reported results [38],

ro

which indicates the atomistic models of PIMs in our work is reasonable. It is attributed to the fact

-p

that the Cardo structure can disrupt the polymer chains packing and thus generate more fractional free volume [46]. The increased fractional free volume is favorable for the increase of CO2

re

permeability.

lP

3.3.2 PALS analysis

na

PALS is a well-established tool to determine the free volume in polymer membranes. The free volume of PIM-1 and Cardo-PIM-1 polymers was characterized by PALS analysis and the

ur

resulting data were compared to those calculated from the atomistic models. The quantitative

Jo

information on the free-volume size, distribution and content mainly attributes to the pick-off annihilation of long-lived orthopositronium (o-Ps, the triplet bound state of a positron and an electron) [47]. The estimated pick-off lifetime of the o-Ps (τ3 in ns) is correlated to the mean free volume radius R (Å) by a semiempirical spherical-cavity model as the following equation [48, 49]: 1

R

1

2πR

τ3 = 2 [1- R+∆R + 2π sin (R+∆R)]

-1

(6)

Journal Pre-proof Here ΔR is the thickness of the electron layer surrounding the free volume hole, which is an empirical parameter to be 1.66Å. The relative fractional free volume (FFVtest) is calculated based on the Williams-Landel-Ferry (WLF) equation. 4

Vf = 3 πR3

(7)

FFVtest = Cf Vf = CI3 Vf =0.0018 I3 Vf

(8)

where I3 is the o-Ps intensity (%) for the estimated pick-off lifetime τ3 and Vf is the mean free

of

volume (Å3) of a cavity in the polymer matrix calculated based on the mean free volume radius

ro

R. A detailed description of PALS can be obtained elsewhere [48-50].

-p

As shown in Table 2, in comparison to PIM-1 (0.4 nm), the diameter is enlarged to 0.41 nm

re

for Cardo-PIM-1 due to the larger bis(phenyl)fluorine structure. Correspondingly, the fractional

lP

free value of Cardo-PIM-1 (5.70%) is higher than that of PIM-1 (4.96%), which is remarkably smaller than that of the predicted FFVMS values of 29.4% and 32.7% calculated from the

na

atomistic models. The reason of the FFVMS higher than that of the FFVtest is that only the ultrafine micropores (i.e., < 4 ns in τ3 value) is considered in the PALS experiments, whereas all

ur

functional groups in PIM-1 polymer structures will be taken into consideration in the atomistic

Jo

models method. As a matter of fact, the long lifetimes (i.e., >4 ns in τ4 value) of PIM-1 polymer is also observed in the PALS experiments. The long lifetimes of o-Ps observed results from both the positron beam itself and the large micropores in the PIM-1 membranes based on the nature of slow positron beam [15]. To simplify the procedure in our experiment, only three lifetimes are adopted without consideration of long lifetimes of τ4 value. As the gas diffusion through the membrane is related to the microstructure of the membrane, the improvement of the fractional free volume of polymer exerts an important influence on the transport property and separation performance based on the free volume theory [51].

Journal Pre-proof Table 2. Physical and textural properties of PIM-1 and Cardo-PIM-1 polymer

τ3 (ns)

I3 (%)

R (nm)

Vf (nm3)

FFVtest (%)

FFVMS (%)

PIM-1

3.66

10.29

0.40

0.27

4.96

29.4

Cardo-PIM-1

3.68

10.97

0.41

0.29

5.70

32.7

3.4. Characterization of the membranes In the mixed matrix membranes, the morphology of the dispersed phase influences the gas

of

transport properties. The SEM images of the cross-sectional morphology of the pristine Cardo-

ro

PIM-1 membranes and the Cardo-PIM-1 membranes incorporated with pristine and

-p

functionalized MWCNTs observed are shown in Fig. 7. The smooth surface of the pristine

re

Cardo-PIM-1 membrane nearly defect-free is observed (Fig. 7a). With the p-MWCNTs incorporated into the Cardo-PIM-1 matrix, the resulting MMM exhibits a heterogeneous

lP

agglomeration and cluster forming of MWCNTs (Fig. 7b). The explanation for the

na

agglomeration of MWCNTs in the polymer matrix is that the interactions between the MWCNTs (π-π interaction) are stronger than that with the polymer matrix [52-54]. At a reasonable

ur

modified MWCNTs loading (7.5 wt. %) in the MMMs, most of the MWCNTs exhibit a

Jo

relatively uniform distribution in Cardo-PIM-1 matrix, and there is no obvious interfacial voids in the prepared MMM (Fig. 7c and 7d). The presence of -OH, -COOH and -NH2 on the surface of MWCNTs de-bundles the highly entangled MWCNTs which results in the enhanced dispersion in the Cardo-PIM-1 matrix. However, with the doping content of f-MWCNTs higher than 12.5 wt. % (Fig. 7e), the f-MWCNTs display agglomerate and heterogeneously distribute in the Cardo-PIM-1 matrix, which will influence the permeation of gas molecules. The crosssection morphology analysis indicates that the thickness of the membranes is approximate 8 μm (Fig. 7f).

Jo

ur

na

lP

re

-p

ro

of

Journal Pre-proof

Fig. 7. SEM images of the cross-section of the pristine Cardo-PIM-1 membrane (a), the MMMs containing 7.5wt. % p-MWCNTs (b), 7.5wt. % o-MWCNTs (c), 7.5wt. % f-MWCNTs (d), 12.5 wt. % f-MWCNTs (e) and the thickness of the MMMs (f)

3.5. Gas separation performances of MMMs

Journal Pre-proof The effect of MWCNTs loading in MMMs on the gas separation performance was investigated by synthesizing a series of membranes with different content of pristine and functionalized MWCNTs loadings of 0-12.5 wt. %. Table 3 lists the gas permeability coefficients calculated from the thickness of selective layer and CO2/N2 selectivities of PIM-1, Cardo-PIM-1 and MMMs incorporated with pristine and functionalized MWCNTs at 298K. The gas separation performance as a function of MWCNTs loading is presented in Fig. 8. Cardo-PIM-1 membrane

of

demonstrates exceptionally higher values of gas permeability than that of PIM-1 membrane,

ro

which confirms that the larger pore-size and fractional free volume in the Cardo-PIM-1 are

-p

propitious to enhance the CO2 and N2 permeability. After incorporation of unmodified

re

MWCNTs, the permeability of CO2 and N2 of the MMMs decreases dramatically due to the worse dispersion of the unmodified MWCNTs in the polymer matrix.

Moreover, some

lP

MWCNTs even block holes in the Cardo-PIM-1 matrix which exhibits lower CO2 permeability.

na

The MMMs incorporated with acid-treated and amine-functionalized MWCNTs represent better gas separation performance. It can be seen that the CO2 permeability exhibits an evident increase

ur

with the increase of o-MWCNTs or f-MWCNTs loading in the polymer matrix. The CO2

Jo

permeability increases from 1.3×104 Barrer with the pure Cardo-PIM-1 membrane to 1.9×104 Barrer with the MMMs containing 7.5 wt% o-MWCNTs loading and 2.9×104 Barrer with the MMMs containing 7.5 wt% f-MWCNTs loading without compromising gas separation factor, with the potential to resist the trade-off effects between the permeance and the selectivity. It is mainly based on the more carboxyl groups (-COOH), hydroxyl groups (-OH) on the o-MWCNTs and amino groups (-NH2) on the f-MWCNTs surface than that of the p-MWCNTs. These groups especially -NH2 have strong affinity force with the CO2 molecules, which enhances CO2 permeability and gas separation factor. However, with the increase of the o-MWCNTs and f-

Journal Pre-proof MWCNTs loading to 12.5 wt%, the CO2 permeability of the MMMs shows a downward trend, still higher than that of pure Cardo-PIM-1 membrane. It is worth mentioning that the dispersion state of CNTs in the polymer matrix plays an important role in determining the gas separation performance. From the above SEM images, it can be seen that f-MWCNTs are well dispersed in the polymer matrix at a certain f-MWCNTs loadings, which suggests that acid and amine treatment impels the nanotubes to disintegrate into

of

small fragments. In this case, the well-dispersed f-MWCNTs could insert into the polymers

ro

chains more effectively and enhance the gas permeability with the reinforced interfacial

-p

interactions between CNTs and Cardo-PIM-1. While when the f-MWCNTs content in the mixed

re

matrix membrane reaches 12.5 wt. %, there are agglomerations and clusters of nanotubes formed in the polymer matrix, which leads to a slight deterioration of the gas permeability. The more

lP

evident decrease of CO2 permeability than that of N2 results in the reduction of the CO2/N2

na

separation factor.

ur

Table 3. Gas permeability coefficient and separation factor of PIM-1, Cardo-PIM-1 and mixed

Jo

matrix membranes at 25 oC Membrane

PaCO2/Barrerb

PN2/Barrer

αCO2/N2

PIM-1

(4.4±0.22)×103

(2.9±0.15)×102

15.2

Cardo-PIM-1

(1.3±0.07)×104

(8.9±0.45)×102

14.6

Cardo-PIM-1+2.5 wt.% p-MWCNTs

(8.0±0.40)×103

(5.6±0.28)×102

14.3

Cardo-PIM-1+7.5 wt.% p-MWCNTs

(5.3±0.27)×103

(4.8±0.24)×102

11.0

Cardo-PIM-1+12.5 wt.% p-MWCNTs

(1.1±0.06)×104

(8.6±0.43)×102

12.8

Cardo-PIM-1+2.5 wt.% o-MWCNTs

(1.8±0.09)×104

(9.8±0.49)×102

18.4

Journal Pre-proof (1.0±0.05)×103

19.0

Cardo-PIM-1+12.5 wt.% o-MWCNTs

(1.8±0.09)×104

(9.6±0.48)×102

18.8

Cardo-PIM-1+2.5 wt.% f-MWCNTs

(2.3±0.12)×104

(1.1±0.06)×103

20.9

Cardo-PIM-1+7.5 wt.% f-MWCNTs

(2.9±0.15)×104

(1.2±0.06)×103

24.2

Cardo-PIM-1+12.5 wt.% f-MWCNTs

(1.6±0.08)×104

(1.0±0.05)×103

16.0

selective thin layer thickness considered at 8 μm from SEM images. 3

∙(STP)∙cm cm2 ∙s∙cmHg

1Barrer=1×10

3.0x104

-p

30

re lP

2.0x104 1.5x104

na

20

1.0x104 5.0x103

Jo

ur

α (CO2/N2)

25

15

2.5x104

pCO2 (Barrer)

-10 cm

of

b

(1.9±0.10)×104

ro

a

Cardo-PIM-1+7.5 wt.% o-MWCNTs

0.0

10

0

2.5

7.5

12.5

MWCNTs content (wt%)

Fig. 8. Effects of MWCNTs loading level on the gas separation performance at 25oC (hollow for CO2/N2 separation factor; solid for CO2 permeability coefficient): Cardo-PIM-1/p-MWCNTs MMMs (▽and▼), Cardo-PIM-1/o-MWCNTs MMMs (☆ and ★), Cardo-PIM-1/f-MWCNTs MMMs (○ and ●)

Journal Pre-proof The most informative method of demonstrating the potential of a novel polymer for the gas separation is to compare the permeability data on a Robeson plot to assess its position relative to the benchmark upper bound. Fig. 9 shows the CO2/N2 separation performance (that is, PCO2 versus αCO2/N2) in our experiment in comparison with the current advanced related membranes reported in the literatures. In contrast, the bulky bis(phenyl)fluorene unit in Cardo-PIM-1 enhance the gas permeability relative to the PIM-1, correspondingly the gas permeability data for

of

Cardo-PIM-1 are closer to the 2008 upper bounds. With the optimal content of f-MWCNTs

ro

incorporated into the Cardo-PIM-1 polymer, the separation performance of CO2/N2 are exceed

-p

Robeson’s trade-off upper bound limit. For the comparison of most pure PIM-1 membrane, the

re

gas separation performance of the obtained PIM-1 membrane in our work is similar to the literatures [38, 46]. After incorporation of functionalized MWCNTs into the Cardo-PIM-1

lP

matrix, the MMMs exhibits outstanding CO2/N2 separation efficiency, which is superior to most

Jo

ur

na

PIM-based mixed matrix membrane reported in the literatures.

Journal Pre-proof 45 40 35

2008 upper bound

30

20 15

5 1000

of ro -p

10

PIM-1(our work) PIM-1[46] PIM-1[38] Cardo-PIM-1(our work) PIM-SBF[56] TBPIM25[16] PIM-PI-SBI[57] PIM-TMN-SBI[17] PIM-TMN-Trip[17] Cardo-PIM-1/f-MWCNTs(our work) f-MWCNTs/PIM-1[29] f-MWCNT/PIM-1[55]

re

α (CO2/N2)

25

10000

50000

lP

PCO2 (Barrer)

literatures.

Jo

4. Conclusions

ur

na

Fig. 9. Comparison of CO2 separation performance of Cardo-PIM-1/ MWCNTs MMMs with the

In summary, the MMMs were successfully fabricated by comprising MWCNTs as dispersed phase and Cardo-PIM-1 as polymer matrix. The MWCNTs were functionalized via acidtreatment, chloride-treated and amine-treated. FTIR and XPS confirmed that functional treatment had introduced carboxylic acid groups (-COOH), hydroxyl groups (-OH) and amino groups(NH2) on the surface of MWCNTs. TEM and SEM images showed that the carbon nanotubes were cut into short ropes and the f-MWCNTs are well dispersed throughout the Cardo-PIM-1 matrix compared to the unmodified MWCNTs. The FTIR and NMR patterns obtained for the polymers indicated that the bis(phenyl)fluorene unit was successfully connected to the structure

Journal Pre-proof of PIM-1 to fabricate the Cardo-PIM-1. The BET, atomistic models and PALS patterns demonstrated that the Cardo-PIM-1 contained higher pore-size distribution and fractional free volume, and stronger CO2 sorption capability than that of PIM-1. Gas separation performance showed that the MMMs with the incorporation of f-MWCNTs represented the optimal gas separation performance than other MMMs. In specific, there was 2.9×104 Barrer in the CO2 permeability and 24.2 in the selectivity of CO2/N2 at 7.5wt.% f-MWCNTs loading based on the

of

special microporous structure of the Cardo-PIM-1, uniform dispersion of the functionalized

ro

carbon nanotubes and the presence of functional groups such as the hydroxyl groups, carboxyl

-p

groups and amino group on the surface of f-MWCNTs. The Cardo-PIM-1/MWCNTs MMMs

re

show high CO2 permeability and a desirable CO2/N2 separation factor, which will provide a

na

Declaration of Competing Interest

lP

promising alternative in the industrial flue gas separation and CO2 capture process.

Jo

Acknowledgments

ur

The authors declare no conflict of interest.

The authors gratefully acknowledge the financial support from the National Key Research and Development Plan of China (2016YFE0106700), National Natural Science Foundation of China (U1862120), Major Science and Technology Innovation Project of Shandong

(2018CXGC1002),

Natural

Science

Foundation

of

Shandong

Province

(ZR2019MB012) and the Fundamental Research Funds for the Central Universities (18CX05006A).

Journal Pre-proof Author Statement Haixiang Sun: Conceptualization, Methodology, Writing-Original draft preparation Wen Gao: Investigation, Writing-Original draft preparation Yanwei Zhang: Investigation, Visualization Xingzhong Cao: Resources, Visualization Shanshan Bao: Data Curation Peng Li : Resources, Visualization Zixi Kang: Writing-Reviewing and Editing Q. Jason Niu:

of

Supervision

ro

References

-p

[1] T. Dasgupta, S. N. Punnathanam, K. G. Ayappa, Effect of functional groups on separating

lP

Chem. Eng. Sci. 121(2015) 279-291.

re

carbon dioxide from CO2/N2 gas mixtures using edge functionalized graphene nanoribbons,

[2] S. Satyapal, T. Filburn, J. Trela, J. Strange, Performance and properties of a solid amine

ur

250-255.

na

sorbent for carbon dioxide removal in space life support applications, Energy Fuels 15 (2001)

[3] M. B. Yue, Y. Chun, Y. Cao, X. Dong, J. H. Zhu, CO2 capture by as-prepared SBA-15 with

Jo

an occluded organic template, Adv. Funct. Mater. 16 (2006) 1717-1722. [4] A. C. Lua, Y. Shen, Preparation and characterization of polyimide-silica composite membranes and their derived carbon–silica composite membranes for gas separation, Chem. Eng. J. 220 (2013) 441-451. [5] Z. Qiao, M. Sheng, Wang J, Z. Song, Z. Wang. Metal-induced polymer framework membrane with high performance for CO2 separation, AIChE J. 65 (2019) 239-249. [6] B. Zornoza, S. Irusta, C. Téllez, J. Coronas , Mesoporous silica sphere-polysulfone mixed matrix membranes for gas separation, Langmuir 25 (2009) 5903-5909.

Journal Pre-proof [7] V. M. Vorotyntsev, P. N. Drozdov, I. V. Vorotyntsev, High purification of substances by a gas separation method, Desalination 240 (2009) 301-305. [8] L. M. Robeson, The upper bound revisited, J. Membr. Sci. 320 (2008) 390-400. [9] D. H. Everett, Manual of symbols and terminology for physicochemical quantities and units, Appendix II: Definitions, terminology and symbols in colloid and surface chemistry, Pure Appl. Chem. 31 (1972) 4577-4638.

of

[10] P. M. Budd, B. S. Ghanem, S. Makhseed, N. B. McKeown, K. J. Msayib, C. E. Tattershall,

ro

Polymers of intrinsic microporosity (PIMs): Robust, solution-processable, organic

-p

nanoporous materials, Chem. Commun. 2 (2004) 230-231.

re

[11] C. Ma, J. J. Urban, Polymers of intrinsic microporosity (PIMs) gas separation membranes: A mini review, Proc. Nat. Res. Soc. 2 (2018) 02002.

lP

[12] P. M. Budd, K. J. Msayib, C. E. Tattershall, B.Ghanem, Gas separation membranes from

na

polymers of intrinsic microporosity, J. Membr. Sci. 251 (2005) 263-269. [13] C. R. Mason, L. Maynard-Atem, N. M. Al-Harbi, P. M. Budd,P. Bernardo,F. Bazzarelli,

ur

G. Clarizia , J. C. Jansen, Polymer of intrinsic microporosity incorporating thioamide

6479.

Jo

functionality: Preparation and gas transport properties, Macromolecules 44 (2011) 6471-

[14] C. R. Mason, L. Maynard-Atem, K.W.J. Heard, B. Satilmis,P. M. Budd,K. Friess, M.Lanc, P. Bernardo,G. Clarizia, J. C. Jansen, Enhancement of CO2 affinity in a polymer of intrinsic microporosity by amine modification, Macromolecules 47 (2014) 1021-1029. [15] F.Y. Li, Y. Xiao, T. Chung, S. Kawi, High-performance thermally self-cross-linked polymer of intrinsic microporosity (PIM-1) membranes for energy development, Macromolecules 45 (2012) 1427-1437.

Journal Pre-proof [16] Z. G. Wang, X. Liu, D. Wang, J. Jin, Tröger’s base-based copolymers with intrinsic microporosity for CO2 separation and effect of Tröger’s base on separation performance, Polym. Chem. 5 (2014) 2793-2800. [17] I. Rose, C. G. Bezzu, M. Carta, B. Comesana-Gandara, E. Lasseuguette, M. C. Ferrari, P. Bernardo, G. Clarizia, A. Fuoco, J. C. Jansen, K. E. Hart, T. P. Liyana-Arachchi, C. M. Colina,N. B. Mckeown, Polymer ultrapermeability from the inefficient packing of 2D

of

chains, Nat. Mater. 16 (2017) 932-937.

ro

[18] S Kazama, T Teramoto, K Haraya. Carbon dioxide and nitrogen transport properties of bis

-p

(phenyl) fluorene-based cardo polymer membranes. J. Membr. Sci. 2002, 207(1):91-104.

re

[19] S. Basu, A. Cano-Odena, I.F.J. Vankelecom, MOF-containing mixed-matrix membranes for CO2/CH4 and CO2/N2 binary gas mixture separations, Sep. Purif. Technol. 81 (2011) 31-40.

lP

[20] W. Zhang, D. Liu, X. Guo, H. L. Huang, C. L. Zhong. Fabrication of mixed-matrix

3400-3409.

na

membranes with MOF-derived porous carbon for CO2 separation, AIChE J, 64 (2018)

ur

[21] J. Shen, G. P. Liu, K. Huang, W. Jin,K. R. Lee,N. Ping, Membranes with fast and

Jo

selective gas-transport channels of laminar graphene oxide for efficient CO2 capture, Angew. Chem. 54 (2015) 588-592. [22] X. Q. Li, L. Ma, H. Y. Zhang, S. F. Wang,Z. Y. Jiang,R. L. Guo, H. Wu,Efficient CO2 capture by functionalized graphene oxide nanosheets as fillers to fabricate multipermselective mixed matrix membranes, Appl. Mater. Interfaces 7 (2015) 5528-5537. [23] X. Q. Li, L. Ma, H. Y. Zhang, S. F. Wang, Z. Y. Jiang, R. L. Guo, H. Wu,X. Z. Cao, J. Yang, B. Y. Wang, X. Z. Cao, J. Yang, B. Y. Wang, Synergistic effect of combining carbon

Journal Pre-proof nanotubes and graphene oxide in mixed matrix membranes for efficient CO2 separation, J. Membr. Sci. 479 (2015) 1-10. [24] Y. N. Zhao, B. T. Jung, L. Ansaloni, W. S. Wiston, Multiwalled carbon nanotube mixed matrix membranes containing amines for high pressure CO2/H2 separation, J. Membr. Sci. 459 (2014) 233-243. [25] A. F. Ismail, N. H. Rahim, A. Mustafa, T. Matsuura, B. C. Ng, S. Abdullah, S. A.

of

Hashemifard, Gas separation performance of polyethersulfone/multi-walled carbon

ro

nanotubes mixed matrix membranes, Sep. Purif. Technol. 80 (2011) 20-31.

-p

[26] S. M. Sanip, A. F. Ismail, P. S. Goh, T. Soga, M. Tanemura, H. Yasuhiko, Gas separation

re

properties of functionalized carbon nanotubes mixed matrix membranes, Sep. Purif. Technol. 78 (2011) 208-213.

lP

[27] L. Ge, Z. H. Zhu, V. Rudolph, Enhanced gas permeability by fabricating functionalized

na

multi-walled carbon nanotubes and polyethersulfone nanocomposite membrane, Sep. Purif. Technol. 78 (2011) 76-82.

ur

[28] H. Sun, T. Wang, Y. Xu, W. Gao, P. Li, Q. J. Niu, Fabrication of polyimide and

Jo

functionalized multi-walled carbon nanotubes mixed matrix membranes by in-situ polymerization for CO2 separation, Sep. Purif. Technol. 177 (2017) 327-336. [29] M. M. Khan, V. Filiz, G. Bengtson, S. Shishatskiy, M. Rahma, V. Abetz, Functionalized carbon nanotubes mixed matrix membranes of polymers of intrinsic microporosity for gas separation, Nanoscale. Res. Lett. 7 (2012) 504. [30] J. F. Shen, W. S. Huang, L. P. Wu, Y. Z. Hu, M. X. Ye, Thermo-physical properties of epoxy nanocomposites reinforced with amino-functionalized multi-walled carbon nanotubes, Compos. Part A: Appl. Sci. Manufac. 38 (2007) 1331-1336.

Journal Pre-proof [31] D. N. Theodorou, U. W. Suter, Detailed molecular structure of a vinyl polymer glass, Macromolecules 18 (1985) 1467-1478. [32] W. Fang, L. Zhang, J. Jiang, Gas permeation and separation in functionalized polymers of intrinsic microporosity: A combination of molecular simulations and Ab initio calculations, J. Phys. Chem. C 115 (2011) 14123-14130. [33] Z. Kang, L. Fan, S. Wang, D. F. Sun, M. Xue, S. L. Qiu, In situ confinement of free linkers

of

within a stable MOF membrane for highly improved gas separation properties,

ro

CrystEngComm 19 (2017) 1601-1606.

-p

[34] M. A. Aroon, A. F. Ismail, M. M. Montazer-Rahmati, T. Matsuura, Effect of chitosan as a

re

functionalization agent on the performance and separation properties of polyimide/multiwalled carbon nanotubes mixed matrix flat sheet membranes, J. Membr. Sci. 364 (2010) 309-

lP

317.

na

[35] Y. Wang, Z. Iqbal, S. V. Malhotra, Functionalization of carbon nanotubes with amines and enzymes, Chem. Phys. Lett. 402 (2005) 96-101.

ur

[36] C. Zhao, L. Ji, H. Liu, G.Hu, S. Zhang, M. Yang, Z. Yang , Functionalized carbon

Jo

nanotubes containing isocyanate groups, J. Solid State Chem. 177 (2004) 4394-4398. [37] J. Cui, W. P. Wang, Y. Z. You, C. Liu, P. Wang, Functionalization of multiwalled carbon nanotubes by reversible addition fragmentation chain-transfer polymerization, Polymer 45 (2004) 8717-8721. [38] N. Du, G. P. Robertson, J. Song, I. Pinnau, S. Thomas, M. D. Guiver, Polymers of intrinsic microporosity containing trifluoromethyl and phenylsulfone groups as materials for membrane gas separation, Macromolecules 41(2008) 9656-9662.

Journal Pre-proof [39] J. Weber, Q. Su, M. Antonietti, A. Thomas, Exploring polymers of intrinsic microporositymicroporous, soluble polyamide and polyimide, Macromol. Rapid Commun. 28 (2007) 1871-1876. [40]

U.H.F.

Bunz,

V.

Enkelmann,

L.

Kloppenburg,

Solid-state

structures

of

phenyleneethynylenes:  Comparison of monomers and polymers, Chem. Mater. 11 (1999) 1416-1424.

of

[41] X. Ma, R. Swaidan, Y. Belmabkhout, Y. Zhu, Synthesis and gas transport properties of

ro

hydroxyl-functionalized polyimides with intrinsic microporosity, Macromolecules 45

-p

(2012) 3841-3849.

re

[42] B. S. Ghanem, N. B. McKeown, P. M. Budd, J. D. Selbie, D. Fritsch, High-performance membranes from polyimides with intrinsic microporosity, Adv. Mater. 20 (2008) 2766-

lP

2771.

na

[43] W. A. Elmehalmey, R. A. Azzam, Y. S. Hassan, M. H. Alkordi, T. M. Madkour, Imidebased polymers of intrinsic microporosity: probing the microstructure in relation to CO2

ur

sorption characteristics, ACS Omega 3 (2018) 2757-2764.

Jo

[44] M. Lanč, K. Pilnáček, C. R. Mason, P. M. Budd, Y. Rogan, R. Malpass-Evans , M. Carta, B. C. Gándara, N. B. McKeown, J. C. Jansen, O. Vopička, K. Friess,

Gas sorption in

polymers of intrinsic microporosity: The difference between solubility coefficients determined via time-lag and direct sorption experiments, J. Membr. Sci. 570-571 (2019) 522-536. [45] R. Swaidan, B. Ghanem, I. Pinnau, Fine-tuned intrinsically ultramicroporous polymers redefine the permeability/selectivity upper bounds of membrane-based air and hydrogen separations, Acs Macro Lett. 4 (2015) 947-951.

Journal Pre-proof [46] C. L. Staiger, S. J. Pas, A. J. Hill, C.J. Cornelius, Gas separation, free volume distribution, and physical aging of a highly microporous spirobisindane polymer, Chem. Mater. 20 (2008) 2606-2608. [47] Y. C. Jean, P. E. Mallon, D. M. Schrader, Positron and positronium chemistry, World Scientific (2003). [48] S. J. Tao, Positronium annihilation in molecular substances, Chem. Phys. 56 (1972) 5499-

of

5510.

ro

[49] M. L. Williams, R. F. Landel, J. D. Ferry, The temperature dependence of relaxation

-p

mechanisms in amorphous polymers and other glass-forming liquids, J. Am. Chem. Soc. 77

re

(1955) 3701-3707.

[50] Y. C. Jean, Positron annihilation spectroscopy for chemical analysis: A novel probe for

lP

microstructural analysis of polymers, Microchem. J. 42 (1990) 72-102.

na

[51] G. Choudalakis, A.D. Gotsis, Free volume and mass transport in polymer nanocomposites, Curr. Opin. Colloid Interface SCI. 17 (2012) 132-140.

ur

[52] D. M. Delozier, K. T. Watson, J. G. Smith, D. J. Klein, P. T. Lillehei, J. W. Connell,

Jo

Investigation of aromatic/aliphatic polyimides as dispersants for single wall carbon nanotubes, Macromolecules 39 (2006) 1731-1739. [53] H. P. Chen, Z. Liu, P. Cebe, Chain confinement in electrospun nanofibers of PET with carbon nanotubes, Polymer 50 (2009) 872-880. [54] M. Chen, H. W. Yu, J. H. Chen, H. S. Koo, Effect of purification treatment on adsorption characteristics of carbon nanotubes, Diamond Relat. Mater. 16 (2007) 1110-1115. [55] M. M. Khan, V. Filiz, G. Bengtson, S. Shishatskiy, M. Rahma, J. Lillepaerg, V. Abetz, Enhanced gas permeability by fabricating mixed matrix membranes of functionalized

Journal Pre-proof multiwalled carbon nanotubes and polymers of intrinsic microporosity (PIM), J. Membr. Sci. 436 (2013) 109-120. [56] C. G. Bezzu, M. Carta, A. Tonkins, J. C. Jansen, P. Bernardo, F. Bazzarelli, N. B. McKeown, A spirobifluorene-based polymer of intrinsic microporosity with improved performance for gas separation, Adv Mater. 24 (2012) 5930-5933. [57] Y. Rogan, L. Starannikova, V. Ryzhikh, Y. Yampolskii, P. Bernardo, F. Bazzarelli, J. C.

of

Jansen, N, B. McKeown, Synthesis and gas permeation properties of novel spirobisindane-

-p

ro

based polyimides of intrinsic microporosity, Polym. Chem. 4 (2013) 3813-3820.

Jo

ur

na

lP

re

Highlights:  A novel bis(phenyl)fluorene-based PIMs (Cardo-PIM-1) with dibenzodioxin polymerization reaction was prepared.  Cardo-PIM-1/functionalized MWCNTs MMMs were fabricated by the solution mixing method.  Cardo-PIM-1 membrane exhibited excellent CO2/N2 separation performance.  Functionalized MWCNTs in the MMMs further improved the CO2 separation performance.